Intel(R) Communications Chipset 89xx Series Datasheet October 2012 Order Number: 327879-001US INFORMATION IN THIS DOCUMENT IS PROVIDED IN CONNECTION WITH INTEL(R) PRODUCTS. NO LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. EXCEPT AS PROVIDED IN INTEL'S TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, INTEL ASSUMES NO LIABILITY WHATSOEVER, AND INTEL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY, RELATING TO SALE AND/OR USE OF INTEL PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT. Intel products are not intended for use in medical, life saving, life sustaining, critical control or safety systems, or in nuclear facility applications. Legal Lines and Disclaimers Intel may make changes to specifications and product descriptions at any time, without notice. Intel Corporation may have patents or pending patent applications, trademarks, copyrights, or other intellectual property rights that relate to the presented subject matter. The furnishing of documents and other materials and information does not provide any license, express or implied, by estoppel or otherwise, to any such patents, trademarks, copyrights, or other intellectual property rights. Designers must not rely on the absence or characteristics of any features or instructions marked "reserved" or "undefined." Intel reserves these for future definition and shall have no responsibility whatsoever for conflicts or incompatibilities arising from future changes to them. Intel processor numbers are not a measure of performance. Processor numbers differentiate features within each processor family, not across different processor families. See http://www.intel.com/products/processor_number for details. Contact your local Intel sales office or your distributor to obtain the latest specifications and before placing your product order. Copies of documents which have an order number and are referenced in this document, or other Intel literature may be obtained by calling 1-800-548-4725 or by visiting Intel's website at http://www.intel.com. Intel, the Intel logo and Pentium are registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries. *Other names and brands may be claimed as the property of others. Copyright (c) 2012, Intel Corporation. All Rights Reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 2 October 2012 Order Number: 327879-001US Revision History Revision History Date Revision October 2012 001 Description * October 2012 Order Number: 327879-001US First release Intel(R) Communications Chipset 89xx Series - Datasheet 3 Product Features Platform Controller Hub (PCH) -- Integrated Intel(R) Platform Controller Hub Complex technology -- Intel Architecture Processor Companion -- Asynchronous DIMM Refresh (ADR) -- Intel(R) QuickAssist Technology -- Extensive integration of standard Intel architecture communications interfaces provide cost, power and board area savings Intel(R) QuickAssist Integrated Accelerator -- Symmetric Cryptographic Functions -- Public Key Functions -- Compression/Decompression Direct Media Interface (DMI) Gen1 -- 10 Gb/s each direction, full duplex -- Transparent to software PECI Interface PCI Express* Gen1 -- 4 PCI Express* Root Complex Ports -- PCIe Gen1 speed (2.5GT/s) -- Compliant to Gen2 messaging protocol -- Ports can be independently configured to support 4x1, 2x2, 1x2 + 2x1, or 1x4 -- Supports lane reversal with x4 configuration -- Module based Hot-Plug supported (e.g., ExpressCard*) Integrated Serial ATA Host Controller -- Two SATA ports -- SATA Gen2 Data transfer rates up to 3.0 Gb/s (300 MB/s). -- One activity LED -- Multiple MSI Message vectors. -- Integrated AHCI controller USB 2.0/1.1 -- Six USB 1.1 or USB 2.0 -- One EHCI Host Controller, supporting up to six external ports. -- Per-Port-Disable Capability -- Includes two USB 2.0 High-speed Debug Ports -- Supports wake-up from sleeping states S1-S4 -- Supports legacy Keyboard/Mouse software UART -- Two integrated UARTs 16550 compatible Intel(R) Trusted Execution Technology Support Intel(R) Communications Chipset 89xx Series - Datasheet 4 SMBus -- SMBus Interfaces -- One Host SMBus (Master) -- One SMLINK (Slave) -- One EndPoint SMBus (Slave) -- One GbE SMBus (Master/Slave) -- SMBus Max speed, up to 100 KHz -- Supports SMBus 2.0 Specification -- Host interface allows processor to communicate via SMBus -- Slave interface allows an internal or external microcontroller to access system resources -- Compatible with most two-wire components that are also I2C compatible High Precision Event Timers -- Advanced operating system interrupt scheduling Timers Based on 82C54 -- System timer, Refresh request, Speaker tone output Real-Time Clock -- 256-byte battery-backed CMOS RAM -- Integrated oscillator components -- Lower Power DC/DC Converter implementation System TCO Reduction Circuits -- Timers to generate SMI# and Reset upon detection of system hang -- Timers to detect improper processor reset -- Integrated processor frequency strap logic -- Supports ability to disable external devices Serial Peripheral Interface (SPI) -- Supports up to two SPI devices -- Two Chip Select pins, up to 16MB per memory device. -- Supports 20/33/50 MHz SPI devices. -- Support up to two different erase granularities. Interrupt Controller -- Supports up to eight PCI interrupt pins -- Supports PCI 2.3 Message Signaled Interrupts -- Two cascaded 82C59 with 15 interrupts -- Integrated I/O APIC capability with 24 interrupts -- Supports Processor System Bus interrupt delivery DMA Controller -- Two cascaded 8237 DMA controllers -- Supports LPC DMA October 2012 Order Number: 327879-001US Revision History Power Management Logic -- Supports ACPI 3.0b -- ACPI-defined power states (processor driven C states) -- ACPI Power Management Timer -- SMI# generation -- All registers readable/restorable for proper resume from 0 V suspend states Support for APM-based legacy power management for non-ACPI implementations Integrated Gigabit Ethernet Controllers -- Four Integrated IEEE 802.3 MACs -- Four SGMII/SerDes interface Ports -- Four I2C/MDIO Ports for External PHY configuration1 -- Serial EEPROM interface -- 10/100/1000 Mbps Ethernet Support -- Jumbo Frame Support PCI Express* Gen2 End Point -- SR-IOV support for Intel(R) QuickAssist Technology -- PCI Express* 2.0 specification running at 5GT/s. -- Supports lane reversal Low Pin Count (LPC) I/F -- Supports two Master/DMA devices -- Support for Security Device (Trusted Platform Module) connected to LPC 68 GPIO pins (multiplexed or dedicated) -- TTL, Open-Drain, Inversion -- GPIO lock down Package -- 27 mm x 27 mm FCBGA -- 942 pin JTAG 1149.1 -- Boundary Scan for testing during board manufacturing 1. See the Supported Ethernet PHY Devices for the Intel(R) Communications Chipset 89xx Series Application Note for more information. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 5 Contents Overview and PCH Interfaces - Volume 1 of 4 ............................. 81 1.0 Introduction ............................................................................................................82 1.1 Introduction ......................................................................................................82 1.2 Intel(R) Communications Chipset 89xx Series SKU Definition .....................................83 1.3 Document Organization ......................................................................................83 1.4 Referenced Documents and Related Websites ........................................................84 1.5 Acronyms .........................................................................................................85 1.6 Glossary ...........................................................................................................87 2.0 Architecture Overview .............................................................................................93 2.1 Introduction ......................................................................................................93 2.2 PCH Architecture Overview..................................................................................95 2.2.1 PCH Block Summary .................................................................................96 2.2.2 PCH External Interfaces.............................................................................96 2.2.3 IA Compatibility .......................................................................................97 3.0 PCH Platform Memory and Device Configuration ......................................................98 3.1 Overview ..........................................................................................................98 3.1.1 Configuration Objectives............................................................................98 3.1.2 Terminology and Conventions ....................................................................99 3.2 IA Platform Infrastructure ...................................................................................99 3.2.1 General IA Platform View of the Physical Address Space.................................99 3.2.2 IA Platform View of Configuration ...............................................................99 3.3 High-Level Views ............................................................................................. 100 3.3.1 Characteristics of External System Memory (DRAM) .................................... 100 3.3.2 Characteristics of Internal and External Memories ....................................... 100 3.3.3 Characteristics of Device Configuration ...................................................... 101 3.4 Memory Map for PCIe* Endpoint-Attached Devices ............................................... 101 3.5 PCH Endianness............................................................................................... 101 3.6 PCI Configuration............................................................................................ 102 3.6.1 Overview............................................................................................... 102 3.6.2 Device Tree ........................................................................................... 103 3.6.3 Materializing Device Structures................................................................. 105 3.6.4 PCI Configuration Headers ....................................................................... 105 4.0 Functional Description ........................................................................................... 108 4.1 PCI Express* Root Ports ................................................................................... 108 4.1.1 Interrupt Generation ............................................................................... 108 4.1.2 Power Management ................................................................................ 109 4.1.2.1 S3/S4/S5 Support ........................................................................... 109 4.1.2.2 Resuming from Suspended State....................................................... 109 4.1.2.3 Device Initiated PM_PME Message ..................................................... 109 4.1.2.4 SMI/SCI Generation......................................................................... 110 4.1.3 SERR# Generation.................................................................................. 110 4.1.4 Hot-Plug................................................................................................ 110 4.1.4.1 Presence Detection .......................................................................... 110 4.1.4.2 Message Generation ........................................................................ 110 4.1.4.3 Attention Button Detection ............................................................... 111 4.1.4.4 SMI/SCI Generation......................................................................... 111 4.2 LPC Bridge (with System and Management Functions) (B0:D31:F0)........................ 112 4.2.1 LPC Interface ......................................................................................... 112 4.2.1.1 LPC Cycle Types.............................................................................. 113 4.2.1.2 Start Field Definition ........................................................................ 113 Intel(R) Communications Chipset 89xx Series - Datasheet 6 October 2012 Order Number: 327879-001US Contents 4.2.1.3 Cycle Type / Direction (CYCTYPE + DIR) ............................................ 114 4.2.1.4 Size .............................................................................................. 114 4.2.1.5 SYNC ............................................................................................ 114 4.2.1.6 SYNC Time-Out .............................................................................. 115 4.2.1.7 SYNC Error Indication...................................................................... 115 4.2.1.8 LFRAME# Usage ............................................................................. 115 4.2.1.9 I/O Cycles ..................................................................................... 115 4.2.1.10 Bus Master Cycles........................................................................... 115 4.2.1.11 LPC Power Management................................................................... 116 4.2.1.12 Configuration and Implications ......................................................... 116 4.3 DMA Operation (B0:D31:F0) ............................................................................. 116 4.3.1 Channel Priority ..................................................................................... 117 4.3.1.1 Fixed Priority.................................................................................. 117 4.3.1.2 Rotating Priority ............................................................................. 117 4.3.2 Address Compatibility Mode..................................................................... 117 4.3.3 Summary of DMA Transfer Sizes .............................................................. 118 4.3.3.1 Address Shifting When Programmed for 16-Bit I/O Count by Words ....... 118 4.3.4 Autoinitialize ......................................................................................... 118 4.3.5 Software Commands .............................................................................. 119 4.4 LPC DMA ........................................................................................................ 119 4.4.1 Asserting DMA Requests ......................................................................... 119 4.4.2 Abandoning DMA Requests ...................................................................... 120 4.4.3 General Flow of DMA Transfers ................................................................ 120 4.4.4 Terminal Count ...................................................................................... 121 4.4.5 Verify Mode ........................................................................................... 121 4.4.6 DMA Request Deassertion ....................................................................... 121 4.4.7 SYNC Field / LDRQ# Rules....................................................................... 122 4.5 8254 Timers (B0:D31:F0) ................................................................................ 122 4.5.1 Timer Programming................................................................................ 123 4.5.2 Reading from the Interval Timer .............................................................. 124 4.5.2.1 Simple Read................................................................................... 124 4.5.2.2 Counter Latch Command ................................................................. 124 4.5.2.3 Read Back Command ...................................................................... 124 4.6 8259 Interrupt Controllers (PIC) (B0:D31:F0) ..................................................... 125 4.6.1 Interrupt Handling.................................................................................. 126 4.6.1.1 Generating Interrupts...................................................................... 126 4.6.1.2 Acknowledging Interrupts ................................................................ 126 4.6.1.3 Hardware/Software Interrupt Sequence ............................................. 126 4.6.2 Initialization Command Words (ICWx)....................................................... 127 4.6.2.1 ICW1 ............................................................................................ 127 4.6.2.2 ICW2 ............................................................................................ 127 4.6.2.3 ICW3 ............................................................................................ 127 4.6.2.4 ICW4 ............................................................................................ 128 4.6.3 Operation Command Words (OCW)........................................................... 128 4.6.4 Modes of Operation ................................................................................ 128 4.6.4.1 Fully Nested Mode........................................................................... 128 4.6.4.2 Special Fully Nested Mode................................................................ 128 4.6.4.3 Automatic Rotation Mode (Equal Priority Devices)................................ 128 4.6.4.4 Specific Rotation Mode (Specific Priority)............................................ 129 4.6.4.5 Poll Mode....................................................................................... 129 4.6.4.6 Cascade Mode ................................................................................ 129 4.6.4.7 Edge and Level Triggered Mode ........................................................ 129 4.6.4.8 End of Interrupt (EOI) Operations ..................................................... 130 4.6.4.9 Normal End of Interrupt................................................................... 130 4.6.4.10 Automatic End of Interrupt Mode ...................................................... 130 4.6.5 Masking Interrupts ................................................................................. 130 4.6.5.1 Masking on an Individual Interrupt Request........................................ 130 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 7 4.6.5.2 Special Mask Mode .......................................................................... 130 4.6.6 Steering PCI Interrupts ........................................................................... 130 4.7 Advanced Programmable Interrupt Controller (APIC) (B0:D31:F0).......................... 131 4.7.1 Interrupt Handling .................................................................................. 131 4.7.2 Interrupt Mapping................................................................................... 131 4.7.3 PCI Express* Message-Based Interrupts .................................................... 132 4.7.4 IOxAPIC Address Remapping.................................................................... 132 4.8 Serial Interrupt (B0:D31:F0) ............................................................................. 132 4.8.1 Start Frame ........................................................................................... 133 4.8.2 Data Frames .......................................................................................... 133 4.8.3 Stop Frame............................................................................................ 134 4.8.4 Specific Interrupts Not Supported via SERIRQ ............................................ 134 4.8.5 Data Frame Format................................................................................. 134 4.9 Real Time Clock (B0:D31:F0) ............................................................................ 135 4.9.1 Update Cycles ........................................................................................ 136 4.9.2 Interrupts.............................................................................................. 136 4.9.3 Lockable RAM Ranges ............................................................................. 136 4.9.4 Century Rollover..................................................................................... 136 4.9.5 Clearing Battery-Backed RTC RAM ............................................................ 136 4.10 Processor Interface (B0:D31:F0)........................................................................ 138 4.10.1 Processor Interface Signals and VLW Messages........................................... 138 4.10.1.1 A20M# (Mask A20) / A20GATE ......................................................... 138 4.10.1.2 INIT (Initialization).......................................................................... 138 4.10.1.3 FERR# (Numeric Coprocessor Error) .................................................. 139 4.10.1.4 NMI (Non-Maskable Interrupt) .......................................................... 139 4.10.1.5 Processor Power Good (PROCPWRGD) ................................................ 139 4.10.2 Dual-Processor Issues ............................................................................. 140 4.10.2.1 Usage Differences ........................................................................... 140 4.10.3 Virtual Legacy Wire (VLW) Messages ......................................................... 140 4.11 Power Management (B0:D31:F0) ....................................................................... 140 4.11.1 Features................................................................................................ 140 4.11.2 System Power States .............................................................................. 141 4.11.3 System Power Planes .............................................................................. 142 4.11.4 SMI#/SCI Generation.............................................................................. 142 4.11.4.1 PCI Express* SCI ............................................................................ 144 4.11.4.2 PCI Express* Hot-Plug ..................................................................... 144 4.11.5 C-States................................................................................................ 145 4.11.6 Sleep States .......................................................................................... 145 4.11.6.1 Sleep State Overview....................................................................... 145 4.11.6.2 Initiating Sleep State ....................................................................... 145 4.11.6.3 Exiting Sleep States......................................................................... 145 4.11.6.4 PCI Express* WAKE# Signal and PME Event Message ........................... 147 4.11.6.5 Sx-G3-Sx, Handling Power Failures.................................................... 147 4.11.7 Event Input Signals and Their Usage ......................................................... 147 4.11.7.1 PWRBTN# (Power Button) ................................................................ 147 4.11.7.2 RI# (Ring Indicator) ........................................................................ 149 4.11.7.3 SYS_RESET# Signal ........................................................................ 149 4.11.7.4 THRMTRIP# Signal .......................................................................... 149 4.11.8 ALT Access Mode .................................................................................... 150 4.11.8.1 Write Only Registers with Read Paths in ALT Access Mode..................... 150 4.11.8.2 PIC Reserved Bits............................................................................ 152 4.11.8.3 Read Only Registers with Write Paths in ALT Access Mode..................... 153 4.11.9 System Power Supplies, Planes, and Signals............................................... 153 4.11.9.1 Power Plane Control with SLP_S3#, SLP_S4# and SLP_S5#.................. 153 4.11.9.2 SLP_S4# and Suspend-To-RAM Sequencing........................................ 153 4.11.9.3 PWROK Signal ................................................................................ 154 Intel(R) Communications Chipset 89xx Series - Datasheet 8 October 2012 Order Number: 327879-001US Contents 4.11.9.4 BATLOW# (Battery Low) ................................................................. 154 4.11.10 Legacy Power Management Theory of Operation......................................... 154 4.11.10.1 APM Power Management ................................................................. 154 4.11.11 Reset Behavior ...................................................................................... 154 4.12 System Management ....................................................................................... 156 4.12.1 SMBus Host Controller (B0:D31:F3).......................................................... 156 4.12.1.1 Host Controller Interface.................................................................. 157 4.12.1.2 Command Protocols ........................................................................ 158 4.12.1.3 Bus Arbitration ............................................................................... 161 4.12.1.4 Bus Timing .................................................................................... 161 4.12.1.5 Clock Stretching ............................................................................. 161 4.12.1.6 Bus Time Out (Host as SMBus Master) ............................................... 161 4.12.1.7 Interrupts / SMI# ........................................................................... 162 4.12.1.8 SMBALERT#................................................................................... 163 4.12.1.9 SMBus CRC Generation and Checking ................................................ 163 4.12.2 TCO Slave SMBus Interface ..................................................................... 163 4.12.2.1 Format of Slave Write Cycle ............................................................. 164 4.12.2.2 Format of Read Command................................................................ 165 4.12.2.3 Behavioral Notes ............................................................................ 167 4.12.2.4 Slave Read of RTC Time Bytes .......................................................... 168 4.12.2.5 Format of Host Notify Command ....................................................... 168 4.12.2.6 TCO Slave Functions ....................................................................... 169 4.12.2.7 Theory of Operation ........................................................................ 169 4.12.3 EndPoint (EP) Slave SMBus ..................................................................... 170 4.12.3.1 SMBus Supported Transactions ......................................................... 171 4.12.3.2 Addressing..................................................................................... 172 4.12.3.3 SMBus Initiated Southbound Configuration Cycles ............................... 173 4.12.3.4 SMBus Error Handling...................................................................... 174 4.12.3.5 SMBus Interface Reset .................................................................... 174 4.12.3.6 Configuration and Memory Read Protocol ........................................... 175 4.12.3.7 Configuration and Memory Write Protocol........................................... 178 4.12.4 GbE SMBus (Master/Slave) ...................................................................... 180 4.12.5 SMLINK1 Interface ................................................................................. 181 4.13 Serial I/O Unit and Watchdog Timer (SIW) (B0:D31:F0) ....................................... 181 4.13.1 Overview .............................................................................................. 181 4.13.2 Features ............................................................................................... 181 4.13.3 Functional Description............................................................................. 182 4.13.3.1 Host Processor Interface (LPC) ......................................................... 182 4.13.3.2 LPC Interface ................................................................................. 182 4.13.3.3 LPC Cycles ..................................................................................... 182 4.13.3.4 I/O Read and Write Cycles ............................................................... 182 4.13.3.5 Policy ............................................................................................ 182 4.13.3.6 LPC Transfers ................................................................................. 183 4.13.4 LPC Logical Devices 4 and 5: Serial Ports (UART1 and UART2) ..................... 183 4.13.4.1 UART Feature List ........................................................................... 184 4.13.4.2 UART Operational Description ........................................................... 185 4.13.4.3 Programmable Baud Rate Generator.................................................. 186 4.13.4.4 FIFO Operation ............................................................................... 186 4.13.4.5 FIFO Polled Mode Operation ............................................................. 187 4.13.5 LPC Logical Device 6: Watchdog Timer ...................................................... 188 4.13.5.1 Overview ....................................................................................... 188 4.13.5.2 Theory Of Operation........................................................................ 189 4.14 General Purpose I/O (B0:D31:F0)...................................................................... 189 4.14.1 Power Wells........................................................................................... 190 4.14.2 SMI# SCI and NMI Routing ..................................................................... 190 4.14.3 Triggering ............................................................................................. 190 4.14.4 GPIO Registers Lockdown........................................................................ 190 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 9 4.14.5 Serial POST Codes Over GPIO .................................................................. 191 4.14.5.1 Theory of Operation......................................................................... 191 4.14.5.2 Serial Message Format ..................................................................... 192 4.15 SATA Host Controller (B0:D31:F2, F5) ................................................................ 193 4.15.1 SATA Ports 4 and 5 Numbering................................................................. 193 4.15.2 SATA Feature Support ............................................................................. 194 4.15.3 Theory of Operation ................................................................................ 194 4.15.3.1 Standard ATA Emulation .................................................................. 194 4.15.3.2 48-Bit LBA Operation ....................................................................... 194 4.15.4 SATA Swap Bay Support.......................................................................... 195 4.15.5 Hot Plug Operation ................................................................................. 195 4.15.5.1 Low Power Device Presence Detection ................................................ 195 4.15.6 Power Management Operation .................................................................. 195 4.15.6.1 Power State Mappings...................................................................... 195 4.15.6.2 Power State Transitions ................................................................... 196 4.15.6.3 SMI Trapping (APM) ........................................................................ 197 4.15.7 TA LED.................................................................................................. 197 4.15.8 AHCI Operation ...................................................................................... 197 4.15.9 SGPIO Signals........................................................................................ 197 4.15.9.1 Mechanism ..................................................................................... 198 4.15.9.2 Message Format.............................................................................. 199 4.15.9.3 LED Message Type .......................................................................... 199 4.15.9.4 SGPIO Waveform ............................................................................ 201 4.16 High Precision Event Timers .............................................................................. 201 4.16.1 Timer Accuracy ...................................................................................... 202 4.16.2 Interrupt Mapping................................................................................... 202 4.16.3 Periodic Versus Non-Periodic Modes .......................................................... 203 4.16.4 Enabling the Timers ................................................................................ 203 4.16.5 Interrupt Levels...................................................................................... 203 4.16.6 Handling Interrupts................................................................................. 204 4.16.7 Issues Related to 64-Bit Timers with 32-Bit Processors ................................ 204 4.17 USB EHCI Host Controllers (B0:D29:F0) ............................................................. 204 4.17.1 EHC Initialization .................................................................................... 205 4.17.1.1 BIOS Initialization ........................................................................... 205 4.17.1.2 Driver Initialization .......................................................................... 205 4.17.1.3 EHC Resets .................................................................................... 205 4.17.2 Data Structures in Main Memory ............................................................... 205 4.17.3 USB 2.0 Enhanced Host Controller DMA ..................................................... 206 4.17.4 Data Encoding and Bit Stuffing ................................................................. 206 4.17.5 Packet Formats ...................................................................................... 206 4.17.6 USB 2.0 Interrupts and Error Conditions .................................................... 206 4.17.6.1 Aborts on USB 2.0-Initiated Memory Reads......................................... 207 4.17.7 USB 2.0 Power Management .................................................................... 207 4.17.7.1 Pause Feature................................................................................. 207 4.17.7.2 Suspend Feature ............................................................................. 207 4.17.7.3 ACPI Device States.......................................................................... 207 4.17.7.4 ACPI System States......................................................................... 208 4.17.8 USB 2.0 Legacy Keyboard Operation ......................................................... 208 4.17.9 USB 2.0 Based Debug Port....................................................................... 208 4.17.9.1 Theory of Operation........................................................................ 209 4.17.10 EHCI Caching ......................................................................................... 213 4.17.11 USB Pre-Fetch Based Pause ..................................................................... 213 4.17.12 USB Overcurrent Protection ..................................................................... 213 4.18 Integrated USB 2.0 Rate Matching Hub ............................................................... 213 4.18.1 Architecture ........................................................................................... 214 4.19 Thermal Management....................................................................................... 214 Intel(R) Communications Chipset 89xx Series - Datasheet 10 October 2012 Order Number: 327879-001US Contents 4.19.1 Modes of Operation ................................................................................ 214 4.19.2 Thermal Reporting Over System Management Link 1 Interface (SMLink1) ...... 215 4.20 WatchDog Timer (WDT) Host Controller (B0:D31:F7) ........................................... 215 4.20.1 Theory Of Operation ............................................................................... 215 4.20.2 Watchdog Timer Behavior ....................................................................... 216 4.20.3 Pending Bit Array ................................................................................... 216 4.20.4 MSI Interrupt Formation ......................................................................... 216 4.20.5 Usage Models ........................................................................................ 217 4.21 Serial Peripheral Interface (SPI) (B0:D31:F0) ..................................................... 217 4.21.1 Descriptor Mode..................................................................................... 218 4.21.2 SPI Flash Regions................................................................................... 218 4.21.2.1 Device Partitioning .......................................................................... 219 4.21.3 Flash Descriptor ..................................................................................... 219 4.21.3.1 Descriptor Master Region ................................................................. 221 4.21.4 Flash Access.......................................................................................... 221 4.21.4.1 Direct Access Security ..................................................................... 221 4.21.4.2 Register Access Security .................................................................. 222 4.21.5 Serial Flash Device Compatibility Requirements .......................................... 222 4.21.5.1 SPI Based BIOS Requirements.......................................................... 222 4.21.5.2 Intel(R) Management Engine Firmware SPI Flash Requirements............... 223 4.21.5.3 Hardware Sequencing Requirements ................................................. 223 4.21.6 Multiple Page Write Usage Model .............................................................. 224 4.21.6.1 Soft Flash Protection ....................................................................... 224 4.21.6.2 BIOS Range Write Protection ............................................................ 225 4.21.6.3 SMI# Based Global Write Protection .................................................. 225 4.21.7 Flash Device Configurations ..................................................................... 225 4.21.8 SPI Flash Device Recommended Pinout ..................................................... 225 4.21.9 Serial Flash Device Package..................................................................... 226 4.21.9.1 Common Footprint Usage Model........................................................ 226 4.21.9.2 Serial Flash Device Package Recommendations ................................... 226 4.22 Feature Capability Mechanism ........................................................................... 227 5.0 Register and Memory Mappings ............................................................................. 228 5.1 I/O Map ......................................................................................................... 228 5.1.1 Fixed I/O Address Ranges ....................................................................... 228 5.1.2 Variable I/O Decode Ranges .................................................................... 230 5.2 Memory Map................................................................................................... 231 5.3 Boot-Block Update Scheme............................................................................... 232 6.0 LPC Interface Bridge Registers (B0:D31:F0) .......................................................... 234 6.1 Overview ....................................................................................................... 234 6.1.1 PCI Configuration Registers (LPC I/F--B0:D31:F0) ...................................... 234 6.1.1.1 Offset 00h: VID--Vendor Identification Register ................................. 236 6.1.1.2 Offset 02h: DID--Device Identification Register ................................. 236 6.1.1.3 Offset 04h: ID--PCICMD--PCI COMMAND Register .............................. 237 6.1.1.4 Offset 06h: ID--PCISTS--PCI Status Register .................................... 238 6.1.1.5 Offset 08h: RID--Revision Identification Register ............................... 239 6.1.1.6 Offset 09h: PI--Programming Interface Register ................................ 239 6.1.1.7 Offset 0Ah: SCC--Sub Class Code Register ........................................ 239 6.1.1.8 Offset 0Bh: BCC--Base Class Code Register ....................................... 240 6.1.1.9 Offset 0Dh: PLT--Primary Latency Timer Register ............................... 240 6.1.1.10 Offset 0Eh: HEADTYP--Header Type Register ..................................... 240 6.1.1.11 Offset 2Ch: SS--Sub System Identifiers Register ................................ 241 6.1.1.12 Offset 40h: PMBASE--ACPI Base Address Register .............................. 241 6.1.1.13 Offset 44h: ACPI_CNTL--ACPI Control Register .................................. 242 6.1.1.14 Offset 48h: GPIOBASE--GPIO Base Address Register .......................... 243 6.1.1.15 Offset 4Ch: GC--GPIO Control Register ............................................. 244 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 11 6.1.1.16 Offset 60h: PIRQ[n]_ROUT--PIRQ[A,B,C,D] Routing Control Register .... 245 6.1.1.17 Offset 64h: SIRQ_CNTL--Serial IRQ Control Register ........................... 246 6.1.1.18 Offset 68h: PIRQ[n]_ROUT--PIRQ[E,F,G,H] Routing Control Register ..... 247 6.1.1.19 Offset 6Ch: LPC_IBDF--IOxAPIC Bus:Device:Function ......................... 248 6.1.1.20 Offset 70h: LPC_HnBDF - HPET n Bus:Device:Function ....................... 249 6.1.1.21 Offset 80h: LPC_I/O_DEC--I/O Decode Ranges Register ...................... 250 6.1.1.22 Offset 82h: LPC_EN--LPC I/F Enables Register ................................... 251 6.1.1.23 Offset 84h: GEN1_DEC--LPC I/F Generic Decode Range 1 Register ....... 253 6.1.1.24 Offset 88h: GEN2_DEC--LPC I/F Generic Decode Range 2 Register ....... 254 6.1.1.25 Offset 8Ch: GEN3_DEC--LPC I/F Generic Decode Range 3 Register ....... 255 6.1.1.26 Offset 90h: GEN4_DEC--LPC I/F Generic Decode Range 4 Register ....... 256 6.1.1.27 Offset 94h: ULKMC - USB Legacy Keyboard / Mouse Control ................ 257 6.1.1.28 Offset 98h: LGMR -- LPC I/F Generic Memory Range ........................... 259 6.1.1.29 Offset DCh: BIOS_CNTL--BIOS Control Register ................................. 260 6.1.1.31 Offset E2h: FDLEN--Feature Detection Capability Length ..................... 261 6.1.1.32 Offset E3h: FDVER--Feature Detection Version ................................... 262 6.1.1.33 Offset E4h: FDVCT--Feature Vector .................................................. 262 6.1.1.34 Offset F0h: RCBA--Root Base Address Register .................................. 263 6.1.2 DMA I/O Registers (LPC I/F--B0:D31:F0)................................................... 263 6.1.2.1 Offset 00h: DMABASE_CA--DMA Base and Current Address Registers .... 264 6.1.2.3 Offset 87h: DMAMEM_LP--DMA Memory Low Page Registers ................ 266 6.1.2.4 Offset 08h: DMACMD--DMA Command Register .................................. 267 6.1.2.6 Offset 0Ah: DMA_WRSMSK--DMA Write Single Mask Register ............... 269 6.1.2.7 Offset 0Bh: DMACH_MODE--DMA Channel Mode Register .................... 270 6.1.2.8 Offset 0Ch: DMA Clear Byte Pointer Register ...................................... 271 6.1.2.9 Offset 0Dh: DMA Master Clear Register ............................................. 271 6.1.2.10 Offset 0Eh: DMA_CLMSK--DMA Clear Mask Register ............................ 271 6.1.2.11 Offset 0Fh: DMA_WRMSK--DMA Write All Mask Register ...................... 272 6.2 Timer I/O Registers (LPC I/F--B0:D31:F0) .......................................................... 272 6.2.1 Timer I/O Registers ................................................................................ 272 6.2.1.1 Offset 43h: TCW--Timer Control Word Register .................................. 273 6.2.1.2 Offset 40h: SBYTE_FMT--Interval Timer Status Byte Format Register .... 275 6.2.1.3 Offset 40h: Counter Access Ports Register ......................................... 276 6.3 8259 Interrupt Controller (PIC) Registers ............................................................ 277 6.3.1 Interrupt Controller I/O MAP .................................................................... 277 6.3.1.1 Offset 20h: Master PIC ICW1--Master Initialization Command Word 1 Register .............................................................................. 278 6.3.1.2 Offset 21h: Master PIC ICW2--Master Initialization Command Word 2 Register .............................................................................. 279 6.3.1.3 Offset 21h: Master PIC ICW3--Master Initialization Command Word 3 Register .............................................................................. 280 6.3.1.4 Offset A1h: Slave PIC ICW3--Slave Initialization Command Word 3 Register .............................................................................. 280 6.3.1.5 Offset 021h: Master PIC ICW4--Master Initialization Command Word 4 Register .............................................................................. 281 6.3.1.6 Offset 021h: Master PIC OCW1--Master Operational Control Word 1 (Interrupt Mask) Register ...................................................... 282 6.3.1.7 Offset 020h: Master PIC OCW2--Master Operational Control Word 2 Register .............................................................................. 283 6.3.1.8 Offset 020h: Master PIC OCW3--Master Operational Control Word 3 Register .............................................................................. 284 6.3.1.9 Offset 4D0h: Master PIC ELCR1--Master Controller Edge/Level Triggered Register ......................................................................................... 285 6.3.1.10 Offset 4D1h: Slave PIC ELCR2--Slave Controller Edge/Level Triggered Register ......................................................................................... 286 6.4 Advanced Programmable Interrupt Controller (APIC) ............................................ 287 6.4.1 APIC Register Map .................................................................................. 287 6.4.2 APIC Direct Registers .............................................................................. 287 6.4.2.1 IND--Index Register ........................................................................ 287 Intel(R) Communications Chipset 89xx Series - Datasheet 12 October 2012 Order Number: 327879-001US Contents 6.4.2.2 DAT--Data Register......................................................................... 288 6.4.2.3 EOIR--EOI Register......................................................................... 288 6.4.3 APIC Indirect Registers ........................................................................... 289 6.4.3.1 Offset 00h: ID--Identification Register............................................... 290 6.4.3.2 Offset 01h: VER--Version Register .................................................... 291 6.4.3.3 Offset 10h: REDIR_TBL0--Redirection Table 0 .................................... 291 6.5 Real Time Clock Registers................................................................................. 294 6.5.1 I/O Register Address Map........................................................................ 294 6.5.2 Indexed Registers .................................................................................. 295 6.5.2.1 Offset 0Ah: RTC_REGA--Register A ................................................... 296 6.5.2.2 Offset 0Bh: RTC_REGB--Register B (General Configuration) ................. 297 6.5.2.3 Offset OCh: RTC_REGC--Register C (Flag Register) ............................. 298 6.5.2.4 Offset ODh: D--Register D (Flag Register).......................................... 299 6.6 Processor Interface Registers (LPC I/F--B0:D31:F0)............................................. 300 6.6.1 Processor Interface PCI Registers Address Map .......................................... 300 6.6.1.1 Offset 61h: NMI_SC--NMI Status and Control Register ........................ 301 6.6.1.2 Offset 70h: NMI_EN--NMI Enable and Real Time Clock Index Register .. 302 6.6.1.3 Offset 92h: PORT92--Fast A20 and Init Register ................................ 302 6.6.1.4 Offset F0h: COPROC_ERR--Coprocessor Error Register ....................... 303 6.6.1.5 Offset CF9h: RST_CNT--Reset Control Register .................................. 303 6.7 Power Management Registers (PM--B0:D31:F0) .................................................. 305 6.7.1 Power Management PCI Configuration Registers (PM--B0:D31:F0)................ 305 6.7.1.1 Offset A0h: GEN_PMCON_1--General PM Configuration 1 Register ........ 306 6.7.1.2 Offset A2h: GEN_PMCON_2--General PM Configuration 2 Register ........ 308 6.7.1.3 Offset A4h: GEN_PMCON_3--General PM Configuration 3 Register ........ 310 6.7.1.4 Offset A6h: GEN_PMCON_LOCK--General Power Management Configuration Lock Register.............................................................. 313 6.7.1.5 Offset A9h: Chipset Initialization Register 4 ....................................... 313 6.7.1.6 Offset ABh: BM_BREAK_EN Register ................................................. 314 6.7.1.7 Offset ACh: PMIR--Power Management Initialization Register ............... 314 6.7.1.8 Offset B8h: GPIO_ROUT--GPIO Routing Control Register ..................... 315 6.7.2 APM I/O Decode..................................................................................... 315 6.7.2.1 Offset B2h: APM_CNT--Advanced Power Management Control Port Register ........................................................................................ 316 6.7.2.2 Offset B3h: APM_STS--Advanced Power Management Status Port Register ........................................................................................ 316 6.7.3 Power Management I/O Registers............................................................. 317 6.7.3.1 Offset PMBASE+00h: PM1_STS--Power Management 1 Status Register .. 318 6.7.3.2 Offset PMBASE + 02h: PM1_EN--Power Management 1 Enable Register . 321 6.7.3.3 Offset PMBASE + 04h: PM1_CNT--Power Management 1 Control ........... 322 6.7.3.4 Offset PMBASE + 08h: PM1_TMR--Power Management 1 Timer Register ........................................................................................ 323 6.7.3.5 Offset PMBASE + 20h: GPE0_STS--General Purpose Event 0 Status Register ........................................................................................ 324 6.7.3.6 Offset PMBASE + 28h: GPE0_EN--General Purpose Event 0 Enables Register ........................................................................................ 326 6.7.3.7 Offset PMBASE + 30h: SMI_EN--SMI Control and Enable Register ......... 328 6.7.3.8 Offset PMBASE + 34h: SMI_STS--SMI Status Register ......................... 330 6.7.3.9 Offset PMBASE +38h: ALT_GP_SMI_EN--Alternate GPI SMI Enable Register ........................................................................................ 333 6.7.3.10 Offset PMBASE +3Ah: ALT_GP_SMI_STS--Alternate GPI SMI Status Register ........................................................................................ 333 6.7.3.11 Offset PMBASE +3Ch: UPRWC--USB Per-Port Registers Write Control .... 334 6.7.3.12 Offset PMBASE +42h: GPE_CNTL--General Purpose Control Register ..... 335 6.7.3.13 Offset PMBASE +44h: DEVACT_STS--Device Activity Status Register ..... 336 6.7.3.14 Offset PMBASE +50h: PM2_CNT--Power Management 2 Control ............ 337 6.8 System Management TCO Registers (B0:D31:F0) ................................................ 337 6.8.1 TCO Register I/O Map ............................................................................. 337 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 13 6.8.1.1 Offset TCOBASE +00h: TCO_RLD--TCO Timer Reload and Current Value Register ......................................................................................... 338 6.8.1.2 Offset TCOBASE +02h: TCO_DAT_IN--TCO Data In Register ................. 338 6.8.1.3 Offset TCOBASE +03h: TCO_DAT_OUT--TCO Data Out Register ............ 339 6.8.1.4 Offset TCOBASE +04h: TCO1_STS--TCO1 Status Register .................... 339 6.8.1.5 Offset TCOBASE +06h: TCO2_STS--TCO2 Status Register .................... 341 6.8.1.6 Offset TCOBASE +08h: TCO1_CNT--TCO1 Control Register................... 343 6.8.1.7 Offset TCOBASE +0Ah: TCO2_CNT--TCO2 Control Register................... 344 6.8.1.8 Offset TCOBASE +0Ch and Offset TCOBASE +0Dh: TCO_MESSAGE1 and TCO_MESSAGE2 Registers................................................................ 345 6.8.1.9 Offset TCOBASE + 0Eh: TCO_WDCNT--TCO Watchdog Control Register.. 345 6.8.1.10 Offset TCOBASE + 10h: SW_IRQ_GEN--Software IRQ Generation Register ......................................................................................... 346 6.9 General Purpose I/O Registers (B0:D31:F0) ........................................................ 347 6.9.1 General Purpose I/O Signals .................................................................... 347 6.9.1.1 Offset GPIOBASE + 00h: GPIO_USE_SEL--GPIO Use Select Register ...... 348 6.9.1.2 Offset GPIOBASE + 04h: GP_IO_SEL--GPIO Input/Output Select Register ......................................................................................... 349 6.9.1.3 Offset GPIOBASE + 0Ch:GP_LVL--GPIO Level for Input or Output Register ......................................................................................... 349 6.9.1.4 Offset GPIOBASE + 18h: GPO_BLINK--GPO Blink Enable Register .......... 350 6.9.1.5 Offset GPIOBASE + 1Ch: GP_SER_BLINK--GP Serial Blink Data ............. 351 6.9.1.6 Offset GPIOBASE + 20h: GP_SB_CMDSTS--GP Serial Blink Command Status ........................................................................................... 352 6.9.1.7 Offset GPIOBASE + 24h: GP_SB_DATA--GP Serial Blink Data................ 353 6.9.1.8 Offset GPIOBASE + 28h: GPI_NMI_EN--GPI NMI Enable ....................... 353 6.9.1.9 Offset GPIOBASE + 2Ah: GPI_NMI_STS--GPI NMI Status ..................... 354 6.9.1.10 Offset GPIOBASE + 2Ch: GPI_INV--GPIO Signal Invert Register ............ 354 6.9.1.11 Offset GPIOBASE + 30h: GPIO_USE_SEL2--GPIO Use Select 2 Register.. 355 6.9.1.12 Offset GPIOBASE + 34h: GP_IO_SEL2--GPIO Input/Output Select 2 Register ......................................................................................... 356 6.9.1.13 Offset GPIOBASE + 38h: GP_LVL2--GPIO Level for Input or Output 2 Register ......................................................................................... 356 6.9.1.14 Offset GPIOBASE + 40h: GPIO_USE_SEL3--GPIO Use Select 3 Register.. 357 6.9.1.15 Offset GPIOBASE + 44h: GP_IO_SEL3--GPIO Input/Output Select 3 Register ......................................................................................... 358 6.9.1.16 Offset GPIOBASE + 48h: GP_LVL3--GPIO Level for Input or Output 3 Register ......................................................................................... 359 6.9.1.17 Offset GPIOBASE + 60h: GP_RST_SEL1--GPIO Reset Select ................. 360 6.9.1.18 Offset GPIOBASE + 64h: GP_RST_SEL2--GPIO Reset Select ................. 361 6.9.1.19 Offset GPIOBASE + 68h: GP_RST_SEL3--GPIO Reset Select ................. 361 7.0 Chipset Configuration Registers (B0:D31:F0) ......................................................... 362 7.1 Chipset Configuration Registers ......................................................................... 362 7.1.1 Chipset Configuration Register Memory Map (Memory Space) ....................... 362 7.1.1.1 Offset 0014h: V0CTL--Virtual Channel 0 Resource Control Register ....... 365 7.1.1.2 Offset 001Ah: V0STS--Virtual Channel 0 Resource Status Register ........ 366 7.1.1.3 Offset 001Ch: V1CAP--Virtual Channel 1 Resource Capability Register ... 366 7.1.1.4 Offset 0020h: V1CTL--Virtual Channel 1 Resource Control Register ....... 367 7.1.1.5 Offset 0026h: V1STS--Virtual Channel 1 Resource Status Register ........ 367 7.1.1.6 Offset 0050h: CIR0--Chipset Initialization Register 0 .......................... 368 7.1.1.7 Offset 0088h: CIR1--Chipset Initialization Register 1 .......................... 368 7.1.1.8 Offset 00ACh: REC--Root Error Command Register ............................. 369 7.1.1.9 Offset 01A0h: ILCL--Internal Link Capabilities List Register .................. 369 7.1.1.10 Offset 01A4h: LCAP--Link Capabilities Register ................................... 370 7.1.1.11 Offset 01A8h: LCTL--Link Control Register ......................................... 370 7.1.1.12 Offset 01AAh: LSTS--Link Status Register ......................................... 371 7.1.1.13 Offset 0220h: BCR--Backbone Configuration Register ......................... 371 7.1.1.14 Offset 0224h: RPC--Root Port Configuration Register .......................... 372 7.1.1.15 Offset 0234h: DMIC--DMI Control Register ........................................ 373 Intel(R) Communications Chipset 89xx Series - Datasheet 14 October 2012 Order Number: 327879-001US Contents 7.1.1.16 7.1.1.17 7.1.1.18 7.1.1.19 7.1.1.20 7.1.1.21 7.1.1.22 7.1.1.23 7.1.1.24 7.1.1.25 7.1.1.26 7.1.1.27 7.1.1.28 7.1.1.29 7.1.1.30 7.1.1.31 7.1.1.32 7.1.1.33 7.1.1.34 7.1.1.35 7.1.1.36 7.1.1.37 7.1.1.38 7.1.1.39 7.1.1.40 7.1.1.41 7.1.1.42 7.1.1.43 7.1.1.44 7.1.1.45 7.1.1.46 7.1.1.47 7.1.1.48 7.1.1.49 7.1.1.50 7.1.1.51 7.1.1.52 7.1.1.53 7.1.1.54 7.1.1.55 7.1.1.56 7.1.1.57 7.1.1.58 7.1.1.59 7.1.1.60 7.1.1.61 7.1.1.62 7.1.1.63 7.1.1.64 7.1.1.65 7.1.1.66 7.1.1.67 7.1.1.68 7.1.1.69 8.0 Offset 0238h: RPFN--Root Port Function Number and Hide for PCI Express* Root Ports ..................................................................................... 373 Offset 0290h: Reserved Register ...................................................... 374 Offset 1D40H: CIR5--Chipset Initialization Register 5 ......................... 375 Offset 1E00h: TRSR--Trap Status Register ........................................ 375 Offset 1E10h: TRCR--Trapped Cycle Register ..................................... 376 Offset 1E18h: TWDR--Trapped Write Data Register ............................ 376 Offset 1E80h: IOTRn--I/O Trap Register (0-3) ................................... 377 Offset 2010h: DMC--DMI Miscellaneous Control Register ..................... 378 Offset 2024h: CIR6--Chipset Initialization Register 6 .......................... 378 Offset 2324h: DMC2--DMI Miscellaneous Control Register 2 ................ 379 Offset 60h: SBI Unified AFE Address Register ..................................... 379 Offset 64h: SSBI Unified AFE Data Register ........................................ 380 Offset 68h: SBI Unified AFE Status Register ....................................... 381 Offset 3000h: TCTL--TCO Configuration Register ................................ 382 Offset 3100h: D31IP--Device 31 Interrupt Pin Register ....................... 383 Offset 3108h: D29IP--Device 29 Interrupt Pin Register ....................... 384 Offset 310Ch: D28IP--Device 28 Interrupt Pin Register ....................... 385 Offset 3124h: D22IP--Device 22 Interrupt Pin Register ....................... 386 Offset 3140h: D31IR--Device 31 Interrupt Route Register ................... 387 Offset 3144h: D29IR--Device 29 Interrupt Route Register ................... 388 Offset 3146h: D28IR--Device 28 Interrupt Route Register ................... 389 Offset 315Ch: D22IR--Device 22 Interrupt Route Register ................... 390 Offset 31FEh: OIC--Other Interrupt Control Register .......................... 391 Offset 3310h: PRSTS--Power and Reset Status .................................. 392 Offset 3314h: CIR7--Chipset Initialization Register 7 .......................... 393 Offset 3324h: CIR8--Chipset Initialization Register 8 .......................... 393 Offset 3330h: CIR9--Chipset Initialization Register 9 .......................... 393 Offset 3340h: CIR10--Chipset Initialization Register 10 ...................... 394 Offset 3350h: CIR13--Chipset Initialization Register 13 ...................... 394 Offset 3368h: CIR14--Chipset Initialization Register 14 ...................... 394 Offset 3378h: CIR15--Chipset Initialization Register 15 ...................... 395 Offset 3388h: CIR16--Chipset Initialization Register 16 ...................... 395 Offset 33A0h: CIR17--Chipset Initialization Register 17 ...................... 395 Offset 33A8h: CIR18--Chipset Initialization Register 18 ...................... 396 Offset 33C0h: CIR19--Chipset Initialization Register 19 ...................... 396 Offset 33CCh: CIR20--Chipset Initialization Register 20 ...................... 396 Offset 33D0h: CIR21--Chipset Initialization Register 21 ...................... 397 Offset 33D4h: CIR22--Chipset Initialization Register 22 ...................... 397 Offset 3400h: RC--RTC Configuration Register ................................... 398 Offset 3404h: HPTC--High Precision Timer Configuration Register ........ 399 Offset 3410h: GCS--General Control and Status Register .................... 400 Offset 3414h: BUC--Backed Up Control Register ................................ 402 Offset 3418h: FD--Function Disable Register ..................................... 403 Offset 341Ch: CG--Clock Gating ...................................................... 405 Offset 3420h: FDSW--Function Disable SUS Well ............................... 406 Offset 3428h: FD2--Function Disable 2 ............................................. 406 Offset 3500h: USBIR[0:5]--USB Initialization Register [0-5] ................ 407 Offset 3564h: USBIRC--USB Initialization Register C .......................... 408 Offset 3570h: USBIRA--USB Initialization Register A .......................... 408 Offset 357Ch: USBIRB--USB Initialization Register B .......................... 409 Offset 3590h: MISCCTL--Miscellaneous Control Register ..................... 410 Offset 359Ch: PDO--USB Port Disable Override .................................. 411 Offset 35A0h: USBOCM1--Overcurrent MAP Register 1 ....................... 412 Offset 35B0h: RMHWKCTL--Rate Matching Hub Wake Control Register .. 413 SATA Controller Registers (B0:D31:F2) ................................................................. 415 8.1 PCI Configuration Registers (SATA-B0:D31:F2)................................................... 415 8.1.1 SATA Controller PCI Register Address Map................................................. 415 8.1.1.1 Offset 00h: Vendor Identification Register ......................................... 418 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 15 8.1.1.2 Offset 02h: Device Identification Register .......................................... 418 8.1.1.3 Offset 04h: PCI Command Register ................................................... 418 8.1.1.4 Offset 06h: PCI Status Register ........................................................ 420 8.1.1.5 Offset 08h: Revision Identification Register ........................................ 421 8.1.2 Programming Interface Register (SATA-B0:D31:F2).................................... 421 8.1.2.1 Offset 0Ah: When Sub Class Code Register = 01h................................ 421 8.1.2.2 Offset 0Ah: When Sub Class Code Register = 06h................................ 422 8.1.2.3 Offset 0Ah: Sub Class Code Register ................................................. 423 8.1.2.4 Offset 0Bh: Base Class Code Register ................................................ 423 8.1.2.5 Offset 0Dh: Primary Master Latency Timer Register ............................ 423 8.1.2.6 Offset 0Eh: Header Type ................................................................. 424 8.1.2.7 Offset 10h: Primary Command Block Base Address Register ................. 424 8.1.2.8 Offset 14h: Primary Control Block Base Address Register ..................... 425 8.1.2.9 Offset 18h: Secondary Command Block Base Address Register ............. 425 8.1.2.10 Offset 1Ch: Secondary Control Block Base Address Register ................. 426 8.1.2.11 Offset 20h: Legacy Bus Master Base Address Register ......................... 426 8.1.3 AHCI Base Address Register/Serial ATA Index Data Pair Base Address (SATA- B0:D31:F2) ........................................................................................... 427 8.1.3.1 Offset 24h: When SCC is Not 01h ..................................................... 427 8.1.3.2 Offset 24h: When SCC is 01h............................................................ 428 8.1.3.3 Offset 2Ch: Subsystem Vendor Identification Register ......................... 428 8.1.3.4 Offset 2Eh: Subsystem Identification Register .................................... 429 8.1.3.5 Offset 34h: Capabilities Pointer Register ............................................ 429 8.1.3.6 Offset 3Ch: Interrupt Line Register ................................................... 429 8.1.3.7 Offset 3Dh: Interrupt Pin Register .................................................... 430 8.1.3.8 Offset 40h: IDE Timing Register ....................................................... 430 8.1.3.9 Offset 48h: Synchronous DMA Control Register .................................. 431 8.1.3.10 Offset 4Ah: Synchronous DMA Timing Register ................................... 431 8.1.3.11 Offset 54h: IDE I/O Configuration Register ........................................ 432 8.1.4 PCI Power Management Capabilities .......................................................... 432 8.1.4.1 Offset 70h: PCI Power Management Capability Identification Register .... 432 8.1.4.2 Offset 72h: PCI Power Management Capabilities Register ..................... 433 8.1.4.3 Offset 74h: PCI Power Management Control and Status Register ........... 434 8.1.5 Message Signaled Interrupt Capability ....................................................... 435 8.1.5.1 Offset 80h: Message Signaled Interrupt Capability Identification ........... 435 8.1.5.2 Offset 82h: Message Signaled Interrupt Message Control ..................... 436 8.1.5.3 Offset 84h: Message Signaled Interrupt Message Address .................... 437 8.1.5.5 Offset 90h: MAP--Address Map Register ............................................ 438 8.1.5.6 Offset 92h: PCS - Port Control and Status Register ............................. 439 8.1.5.7 Offset 94h: SCLKCG--SATA Clock Gating Control Register .................... 441 8.1.5.8 Offset 9Ch: SCLKGC--SATA Clock General Configuration Register ......... 442 8.1.5.9 Offset A0h: SIRI--SATA Indexed Registers Index ................................ 443 8.1.5.10 Offset A4h: STRD--SATA Indexed Register Data ................................. 443 8.1.5.11 Offset A8h: SATACR0--SATA Capability Register 0 .............................. 444 8.1.5.12 Offset ACh: SATACR1--SATA Capability Register 1 .............................. 445 8.1.5.13 Offset B0h: FLRCID--FLR Capability ID .............................................. 446 8.1.5.14 Offset B2h: FLRCLV--FLR Capability Length and Version ...................... 446 8.1.5.15 Offset B4h: FLRC--FLR Control ......................................................... 447 8.1.5.16 Offset C0h: ATC--APM Trapping Control Register ................................ 448 8.1.5.17 Offset C4h: ATS--APM Trapping Status Register ................................. 448 8.1.5.18 Offset D0h: SP--Scratch Pad Register ............................................... 449 8.1.5.19 Offset E0h: BFCS--BIST FIS Control/Status Register ........................... 450 8.1.5.20 Offset E4h: BFTD1--BIST FIS Transmit Data1 Register ........................ 451 8.1.5.21 Offset E8h: BFTD2--BIST FIS Transmit Data2 Register ........................ 452 8.1.6 SATA Indexed Registers .......................................................................... 453 8.1.6.1 Offset 18h: SATA Indexed Registers Index.......................................... 454 8.1.6.2 Offset 1Ch: SATA Indexed Registers Index ......................................... 454 8.1.6.3 Offset 28h: SATA Indexed Registers Index.......................................... 455 8.1.6.4 Offset 3Eh: SATA Indexed Registers Index.......................................... 455 Intel(R) Communications Chipset 89xx Series - Datasheet 16 October 2012 Order Number: 327879-001US Contents 8.1.6.5 Offset 54h: SATA Indexed Registers Index ......................................... 456 8.1.6.6 Offset 64h: SATA Indexed Registers Index ......................................... 456 8.1.6.7 Offset 68h: SATA Indexed Registers Index ......................................... 457 8.1.6.8 Offset 78h: SATA Indexed Registers Index ......................................... 457 8.1.6.9 Offset 84h: SATA Indexed Registers Index ......................................... 458 8.1.6.10 Offset 88h: SATA Indexed Registers Index ......................................... 458 8.1.6.11 Offset 8Ch: SATA Indexed Registers Index ......................................... 459 8.1.6.12 Offset 94h: SATA Indexed Registers Index ......................................... 459 8.1.6.13 Offset A0h: SATA Indexed Registers Index ......................................... 460 8.1.6.14 Offset A8h: SATA Indexed Registers Index ......................................... 460 8.1.6.15 Offset C4h: SATA Indexed Registers Index ......................................... 461 8.1.6.16 Offset C8h: SATA Indexed Registers Index ......................................... 461 8.2 Bus Master IDE I/O Registers (B0:D31:F2) ......................................................... 462 8.2.1 Bus Master IDE I/O Register Address Map ................................................. 462 8.2.1.1 Offset 00h: Bus Master IDE Command Register Primary ...................... 463 8.2.1.2 Offset 02h: Bus Master IDE Status Register Primary ........................... 464 8.2.1.3 Offset 04h: Bus Master IDE Descriptor Table Pointer Register Primary ... 465 8.2.1.4 Offset 08h: Bus Master IDE Command Register Secondary .................. 466 8.2.1.5 Offset 0Ah: Bus Master IDE Status Register Secondary ....................... 467 8.2.1.6 Offset 0Ch: Bus Master IDE Descriptor Table Pointer Register Secondary .................................................................................... 468 8.2.1.7 Offset 10h: AHCI Index Register ...................................................... 468 8.2.1.8 Offset 14h: AHCI Index Data Register ............................................... 469 8.3 Serial ATA Index/Data Pair Superset Registers .................................................... 470 8.3.1 Superset Registers ................................................................................. 470 8.3.1.1 Offset SIDPBA + 00h: Serial ATA Index ............................................ 470 8.3.1.2 Offset 04h: Serial ATA Data ............................................................ 471 8.4 AHCI Registers (B0:D31:F2) ............................................................................. 471 8.4.1 AHCI Generic Host Control Registers (B0:D31:F2) ...................................... 472 8.4.1.1 Offset 00h: Host Capabilities Register ............................................... 472 8.4.1.2 Offset 04h: Global PCH Control Register ............................................ 477 8.4.1.3 Offset 08h: Interrupt Status Register ................................................ 478 8.4.1.4 Offset 0Ch: Ports Implemented Register ........................................... 479 8.4.1.5 Offset 10h: Serial AHCI Version ....................................................... 479 8.4.1.6 Offset 14h: Serial Command Completion Coalescing Control Register .... 480 8.4.1.7 Offset 18h: Serial Command Completion Coalescing Ports Register ....... 481 8.4.1.8 Offset 1Ch: Serial Enclosure Management Location Register ................ 481 8.4.1.9 Offset 20h: Serial Enclosure Management Control Register .................. 482 8.4.1.10 Offset 24h: Serial Extended Host Capabilities ..................................... 483 8.4.1.11 Offset 6Ch: Serial NVMHCI Ports Implemented .................................. 484 8.4.1.12 Offset 70h: Serial AHCI Version ....................................................... 484 8.4.1.13 Offset A0h: Serial Vendor Specific .................................................... 485 8.4.2 Port Registers (B0:D31:F2) ..................................................................... 486 8.4.2.1 Offset 300h: Port [4:5] Command List Base Address Register .............. 487 8.4.2.2 Offset 304h: Port [4:5] Command List Base Address Upper 32-Bits Register ........................................................................................ 487 8.4.2.3 Offset 308h: Port [4:5] FIS Base Address Register ............................. 488 8.4.2.4 Offset 30Ch: Port [4:5] FIS Base Address Upper 32-Bits Register ......... 488 8.4.2.5 Offset 310h: Port [4:5] Interrupt Status Register ............................... 489 8.4.2.6 Offset 314h: Port [4:5] Interrupt Enable Register ............................... 491 8.4.2.7 Offset 318h: Port [4:5] Command Register ....................................... 493 8.4.2.8 Offset 320h: Port [4:5] Task File Data Register .................................. 496 8.4.2.9 Offset 324h: Port [4:5] Signature Register ........................................ 497 8.4.2.10 Offset 328h: Port [4:5] Serial ATA Status Register ............................. 498 8.4.2.11 Offset 32Ch: Port [4:5] Serial ATA Control Register ............................ 500 8.4.2.12 Offset 330h: Port [4:5] Serial ATA Error Register ............................... 502 8.4.2.13 Offset 334h: Port [4:5] Serial ATA Active .......................................... 503 8.4.2.14 Offset 338h: Port [4:5] Command Issue Register ............................... 504 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 17 9.0 SATA Controller Registers (B0:D31:F5) .................................................................. 505 9.1 PCI Configuration Registers (SATA-B0:D31:F5) ................................................... 505 9.1.1 SATA Controller PCI Register Address Map ................................................. 505 9.1.1.1 Offset 00h: Vendor Identification Register ......................................... 506 9.1.1.2 Offset 02h: Device Identification Register .......................................... 507 9.1.1.3 Offset 04h: PCI Command Register ................................................... 507 9.1.1.4 Offset 06h: PCI Status Register ........................................................ 508 9.1.1.5 Offset 08h: RID--Revision Identification Register ................................ 509 9.1.1.6 Offset 09h: Programming Interface Register ...................................... 510 9.1.1.7 Offset 0Ah: Sub Class Code Register ................................................. 510 9.1.1.8 Offset 0Bh: BCC--Base Class Code Register ....................................... 511 9.1.1.9 Offset 0Dh: Primary Master Latency Timer Register ............................ 511 9.1.1.10 Offset 10h: Primary Command Block Base Address Register ................. 512 9.1.1.11 Offset 14h: Primary Control Block Base Address Register ..................... 513 9.1.1.12 Offset 18h: SCMD-Secondary Command Block Base Address Register ... 513 9.1.1.13 Offset 1Ch: SCNL- Secondary Control Block Base Address Register ....... 514 9.1.1.14 Offset 20h: Legacy Bus Master Base Address Register ......................... 514 9.1.1.15 Offset 24h: SATA Index/Data Pair Base Address Register ..................... 515 9.1.1.16 Offset 2Ch: Subsystem Vendor Identification Register ......................... 515 9.1.1.17 Offset 2Eh: Subsystem Identification Register .................................... 516 9.1.1.18 Offset 34h: Capabilities Pointer Register ............................................ 516 9.1.1.19 Offset 3Ch: Interrupt Line Register ................................................... 516 9.1.1.21 Offset 40h: IDE Timing Register ....................................................... 517 9.1.1.22 Offset 48h: Synchronous DMA Control Register .................................. 517 9.1.1.23 Offset 4Ah: Synchronous DMA Timing Register ................................... 518 9.1.1.24 Offset 54h: IDE I/O Configuration Register ........................................ 519 9.1.1.25 Offset 70h: PCI Power Management Capability Identification Register .... 520 9.1.1.26 Offset 72h: PC--PCI Power Management Capabilities Register .............. 520 9.1.1.27 Offset 74h: PCI Power Management Control and Status Register ........... 521 9.1.1.28 Offset 90h: MAP--Address Map Register ............................................ 522 9.1.1.29 Offset 92h: Port Control and Status Register ...................................... 523 9.1.1.30 Offset A8h: SATA Capability Register 0 .............................................. 524 9.1.1.31 Offset ACh: SATA Capability Register 1 ............................................. 524 9.1.1.32 Offset B0h: FLR Capability ID ........................................................... 525 9.1.1.33 Offset B2h: FLR Capability Length and Value ...................................... 525 9.1.1.34 Offset B4h: FLR Control ................................................................... 526 9.1.1.35 Offset C0h: APM Trapping Control Register ........................................ 527 9.1.1.36 Offset C4h: APM Trapping Control Register ........................................ 527 9.1.2 Bus Master IDE I/O Registers (B0:D31:F5)................................................. 528 9.1.2.1 Offset 00h: Bus Master IDE Command Register .................................. 528 9.1.2.2 Offset 02h: Bus Master IDE Status Register ....................................... 530 9.1.2.3 Offset 04h: Bus Master IDE Descriptor Table Pointer Register ............... 531 9.1.3 Serial ATA Index/Data Pair Superset Registers............................................ 531 9.1.3.1 Offset 00h: SINDX--SATA Index Register .......................................... 532 9.1.3.2 Offset 04h: SDATA--SATA Index Data Register .................................. 532 9.1.3.3 Offset 04h: PxSSTS--Serial ATA Status Register ................................. 533 9.1.3.4 Offset 04h: PxSCTL--Serial ATA Control Register ................................ 535 9.1.3.5 Offset 04h: PxSERR--Serial ATA Error Register ................................... 537 10.0 EHCI Controller Registers (B0:D29:F0) .................................................................. 539 10.1 USB EHCI Configuration Registers (USB EHCI--B0:D29:F0) ................................... 539 10.1.1 USB EHCI PCI Register Address Map ......................................................... 539 10.1.1.1 Offset 00h: Vendor Identification Register ......................................... 542 10.1.1.2 Offset 02h: Device Identification Register .......................................... 542 10.1.1.3 Offset 04h: PCI Command Register ................................................... 543 10.1.1.4 Offset 06h: PCI Status Register ........................................................ 545 10.1.1.5 Offset 08h: RID--Revision Identification Register ................................ 546 10.1.1.6 Offset 09h: Programming Interface Register ...................................... 546 10.1.1.7 Offset 0Ah: Sub Class Code Register ................................................. 546 Intel(R) Communications Chipset 89xx Series - Datasheet 18 October 2012 Order Number: 327879-001US Contents 10.1.1.8 10.1.1.9 10.1.1.10 10.1.1.11 10.1.1.12 10.1.1.13 10.1.1.14 10.1.1.15 10.1.1.16 10.1.1.17 10.1.1.18 10.1.1.19 10.1.1.20 10.1.1.21 10.1.1.22 10.1.1.23 10.1.1.24 10.1.1.25 10.1.1.26 10.1.1.27 10.1.1.28 10.1.1.29 10.1.1.30 10.1.1.31 10.1.1.32 10.1.1.33 10.1.1.34 Offset 0Bh: BCC--Base Class Code Register ....................................... 547 Offset 0Dh: Primary Master Latency Timer Register ............................ 547 Offset 0Eh: Header Type Register .................................................... 548 Offset 10h: Memory Base Address Register ....................................... 548 Offset 2Ch: Subsystem Vendor ID Register ....................................... 549 Offset 2Eh: Subsystem ID Register .................................................. 549 Offset 34h: Capabilities Pointer Register ........................................... 550 Offset 3Ch: Interrupt Line Register ................................................... 550 Offset 3Dh: Interrupt Pin Register .................................................... 550 Offset 50h: PCI Power Management Capability ID Register .................. 551 Offset 51h: Next Item Pointer #1 Register ........................................ 551 Offset 52h: Power Management Capabilities Register .......................... 552 Offset 54h: Power Management Control/Status Register ...................... 553 Offset 58h: Debug Port Capability ID Register .................................... 554 Offset 59h: Next Item Pointer #2 Register ........................................ 554 Offset 5Ah: Debug Port Base Offset Register ..................................... 554 Offset 60h: USB Release Number Register ......................................... 555 Offset 61h: Frame Length Adjustment Register .................................. 555 Offset 62h: Port Wake Capability Register ......................................... 557 Offset 68h: Legacy Support Extended Capability Register .................... 558 Offset 6Ch: Legacy Support Extended Control/Status Register ............. 559 Offset 70h: SPECIAL_SMI--Intel Specific USB 2.0 SMI Register ............ 562 Offset 80h: ACCESS_CNTL--Access Control Register ........................... 564 Offset 84h: EHCIIR1--EHCI Initialization Register 1 ............................ 564 Offset 98h: FLR_CID--Function Level Reset Capability ID .................... 565 Offset 99h: FLR_NEXT--Function Level Reset Next Capability Pointer .... 565 Offset 9Ah: FLR_CLV--Function Level Reset Capability Length and Version ......................................................................................... 566 10.1.1.35 Offset 9Ch: FLR_CTRL--Function Level Reset Control Register .............. 567 10.1.1.36 Offset 9Dh: FLR_STS--Function Level Reset Status Register ................ 567 10.2 Memory-Mapped I/O Registers .......................................................................... 568 10.2.1 Host Controller Capability Registers .......................................................... 568 10.2.1.1 Offset 00h: CAPLENGTH--Capability Registers Length Register ............. 569 10.2.1.2 Offset 02h: HCIVERSION--Host Controller Interface Version Number Register ........................................................................................ 569 10.2.1.3 Offset 04h: HCSPARAMS--Host Controller Structural Parameters .......... 570 10.2.1.4 Offset 08h: HCCPARAMS--Host Controller Capability Parameters Register ........................................................................................ 571 10.2.2 Host Controller Operational Registers........................................................ 572 10.2.2.1 Offset 20h: USB2.0_CMD--USB 2.0 Command Register ....................... 573 10.2.2.2 Offset 24h: USB2.0_STS--USB 2.0 Status Register ............................. 576 10.2.2.3 Offset 28h: USB2.0_INTR--USB 2.0 Interrupt Enable Register ............. 579 10.2.2.4 Offset 2Ch: FRINDEX--Frame Index Register ..................................... 581 10.2.2.5 Offset 30h: CTRLDSSEGMENT--Control Data Structure Segment Register ........................................................................................ 582 10.2.2.6 Offset 34h: PERIODICLISTBASE--Periodic Frame List Base Address Register ........................................................................................ 583 10.2.2.7 Offset 38h: ASYNCLISTADDR--Current Asynchronous List Address Register ........................................................................................ 584 10.2.2.8 Offset 60h: CONFIGFLAG--Configure Flag Register ............................. 585 10.2.2.9 Offset 64h: PORTSC--Port N Status and Control Register .................... 586 10.2.3 USB 2.0-Based Debug Port Registers ........................................................ 591 10.2.3.1 Offset A0h: CNTL_STS--Control/Status Register ................................ 592 10.2.3.2 Offset A4h: USBPID--USB PIDs Register ........................................... 594 10.2.3.3 Offset A8h: DATABUF[7:0]--Data Buffer Bytes[7:0] Register ............... 595 10.2.3.4 Offset B0h: CONFIG--Configuration Register ..................................... 595 11.0 SMBus Controller Registers (B0:D31:F3) ............................................................... 596 11.1 PCI Configuration Registers (SMBus--B0:D31:F3) ................................................ 596 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 19 11.1.1 SMBus Controller PCI Register Address Map ............................................... 596 11.1.1.1 Offset 00h: Vendor Identification Register ......................................... 597 11.1.1.2 Offset 02h: Device Identification Register .......................................... 597 11.1.1.3 Offset 04h: PCI Command Register ................................................... 598 11.1.1.4 Offset 06h: PCI Status Register ........................................................ 599 11.1.1.5 Offset 08h: Revision Identification Register ........................................ 600 11.1.1.6 Offset 09h: Programming Interface Register ...................................... 600 11.1.1.7 Offset 0Ah: Sub Class Code Register ................................................. 600 11.1.1.8 Offset 0Bh: Base Class Code Register ................................................ 601 11.1.1.9 Offset 10h: SMBus Memory Base Address 0 ....................................... 601 11.1.1.10 Offset 14h: SMBus Memory Base Address 1 ....................................... 602 11.1.1.11 Offset 20h: SMBus Base Address Register .......................................... 602 11.1.1.12 Offset 2Ch: Subsystem Vendor Identification Register ......................... 603 11.1.1.13 Offset 2Eh: Subsystem Identification Register .................................... 603 11.1.1.14 Offset 3Ch: Interrupt Line Register ................................................... 604 11.1.1.15 Offset 3Dh: Interrupt Pin Register .................................................... 604 11.1.1.16 Offset 40h: Host Configuration Register ............................................. 605 11.2 SMBus I/O and Memory Mapped I/O Registers ..................................................... 606 11.2.1 SMBus Registers..................................................................................... 606 11.2.1.1 Offset 00h: Host Status Register ...................................................... 607 11.2.1.2 Offset 02h: Host Control Register ..................................................... 609 11.2.1.3 Offset 03h: Host Command Register ................................................. 611 11.2.1.4 Offset 04h: Transmit Slave Address Register ...................................... 611 11.2.1.5 Offset 05h: Host Data 0 Register ...................................................... 612 11.2.1.6 Offset 06h: Host Data 1 Register ...................................................... 612 11.2.1.7 Offset 07h: Host Block Data Byte Register ......................................... 613 11.2.1.8 Offset 08h: Packet Error Check (PEC) Register ................................... 614 11.2.1.9 Offset 09h: Receive Slave Address Register ....................................... 614 11.2.1.10 Offset 0Ah: Receive Slave Data Register ............................................ 615 11.2.1.11 Offset 0Ch: Auxiliary Status Register ................................................ 615 11.2.1.12 Offset 0Dh: Auxiliary Control Register ............................................... 616 11.2.1.13 Offset 0Eh: SMLink Pin Control Register ............................................ 617 11.2.1.14 Offset 0Fh: SMBus Pin Control Register ............................................. 618 11.2.1.15 Offset 10h: Slave Status Register ..................................................... 619 11.2.1.16 Offset 11h: Slave Command Register ................................................ 620 11.2.1.17 Offset 14h: Notify Device Address Register ........................................ 621 11.2.1.18 Offset 16h: Notify Data Low Byte Register ......................................... 621 11.2.1.19 Offset 17h: Notify Data High Byte Register ........................................ 622 12.0 PCI Express* Configuration Registers (B0:D28:F0/1/2/3) .................................... 623 12.1 PCI Express* Configuration Registers (PCI Express* -- B0:D28:F0/F1/F2/F3) .......... 623 12.1.1 PCI Express* Configuration Registers Address Map...................................... 623 12.1.1.1 Offset 00h: Vendor Identification Register ......................................... 626 12.1.1.2 Offset 02h: DID--Device Identification Register .................................. 626 12.1.1.3 Offset 04h: PCI COMMAND Register .................................................. 627 12.1.1.4 Offset 06h: PCI Status Register ........................................................ 628 12.1.1.5 Offset 08h: RID--Revision Identification Register ................................ 629 12.1.1.6 Offset 09h: Programming Interface Register ...................................... 630 12.1.1.7 Offset 0Ah: Sub Class Code Register ................................................. 630 12.1.1.8 Offset 0Bh: Base Class Code Register ................................................ 631 12.1.1.9 Offset 0Ch: Cache Line Size Register ................................................ 631 12.1.1.10 Offset 0Dh: Primary Latency Timer Register ....................................... 632 12.1.1.11 Offset 0Eh: Header Type Register ..................................................... 632 12.1.1.12 Offset 18h: Bus Number Register ..................................................... 633 12.1.1.13 Offset 1Bh: Secondary Latency Timer ............................................... 633 12.1.1.14 Offset 1Ch: IOBL--I/O Base and Limit Register ................................... 634 12.1.1.15 Offset 1Eh: Secondary Status Register .............................................. 635 12.1.1.16 Offset 20h: Memory Base and Limit Register ...................................... 636 12.1.1.17 Offset 24h: Prefetchable Memory Base and Limit Register .................... 637 12.1.1.18 Offset 28h: PMBU32--Prefetchable Memory Base Upper 32 Bits Register 637 Intel(R) Communications Chipset 89xx Series - Datasheet 20 October 2012 Order Number: 327879-001US Contents 12.1.1.19 12.1.1.20 12.1.1.21 12.1.1.22 12.1.1.23 12.1.1.24 12.1.1.25 12.1.1.26 12.1.1.27 12.1.1.28 12.1.1.29 12.1.1.30 12.1.1.31 12.1.1.32 12.1.1.33 12.1.1.34 12.1.1.35 12.1.1.36 12.1.1.37 12.1.1.38 12.1.1.39 12.1.1.40 12.1.1.41 12.1.1.42 12.1.1.43 12.1.1.44 12.1.1.45 12.1.1.46 12.1.1.47 12.1.1.48 12.1.1.49 12.1.1.50 12.1.1.51 12.1.1.52 12.1.1.53 12.1.1.54 12.1.1.55 12.1.1.56 12.1.1.57 12.1.1.58 12.1.1.59 12.1.1.60 12.1.1.61 12.1.1.62 12.1.1.63 12.1.1.64 12.1.1.65 12.1.1.66 Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset 2Ch: Prefetchable Memory Limit Upper 32 Bits Register ............. 638 34h: Capabilities List Pointer Register ...................................... 638 3Ch: Interrupt Information Register ........................................ 639 3Eh: Bridge Control Register .................................................. 639 40h: Capabilities List Register ................................................ 641 42h: PCI Express* Capabilities Register ................................... 641 44h: Device Capabilities Register ............................................ 642 48h: Device Control Register .................................................. 643 4Ah: Device Status Register ................................................... 644 4Ch: Link Capabilities Register ............................................... 645 50h: Link Control Register ..................................................... 647 52h: Link Status Register ...................................................... 648 54h: Slot Capabilities Register ................................................ 649 58h: Slot Control Register ...................................................... 650 5Ah: Slot Status Register ....................................................... 651 5Ch: Root Control Register .................................................... 652 60h: Root Status Register ...................................................... 653 64h: Device Capabilities 2 Register ......................................... 653 68h: Device Control 2 Register ............................................... 654 70h: Link Control 2 Register ................................................... 655 80h: Message Signaled Interrupt Identifiers Register ................. 655 82h: Message Signaled Interrupt Message Control Register ........ 656 84h: Message Signaled Interrupt Message Address Register ....... 657 88h: Message Signaled Interrupt Message Data Register ............ 657 90h: Subsystem Vendor Capability Register ............................. 658 94h: Subsystem Vendor Identification Register ......................... 658 A0h: Power Management Capability Register ............................ 659 A2h: PCI Power Management Capabilities Register .................... 659 A4h: PCI Power Management Control and Status Register .......... 660 D4h: Miscellaneous Port Configuration Register 2 ...................... 661 D8h: Miscellaneous Port Configuration Register ........................ 662 DCh: SMI/SCI Status Register ................................................ 665 E1h: Root Port Dynamic Clock Gating Enable ............................ 666 E8h: PCI Express* Configuration Register 1 ............................. 667 104h: Uncorrectable Error Status Register ............................... 667 108h: Uncorrectable Error Mask ............................................. 669 10Ch: Uncorrectable Error Severity ......................................... 671 110h: Correctable Error Status Register ................................... 672 114h: Correctable Error Mask Register .................................... 673 118h: Advanced Error Capabilities and Control Register ............. 674 130h: Root Error Status Register ............................................ 675 180h: Root Complex Topology Capability List Register ............... 676 184h: Element Self Description Register .................................. 676 190h: Upstream Link Description Register ................................ 677 198h: Upstream Link Base Address Register ............................. 678 300h: PCI Express* Configuration Register 2 ............................ 678 318h: PCI Express* Extended Test Mode Register ..................... 679 324h: PCI Express* Configuration Register 1 ............................ 679 13.0 High Precision Event Timer Registers .................................................................... 680 13.1 High Precision Behavioral Rules ......................................................................... 680 13.1.1 Memory Mapped Registers....................................................................... 680 13.1.1.1 Offset 00h: General Capabilities and Identification Register .................. 681 13.1.1.2 Offset 010h: General Configuration Register ....................................... 682 13.1.1.3 Offset 020h: General Interrupt Status Register ................................... 683 13.1.1.4 Offset 0F0h: Main Counter Value Register .......................................... 684 13.1.1.5 Offset 100h: Timer n Configuration and Capabilities Register ................ 685 13.1.1.6 Offset 108h: Timer n Comparator Value Register................................. 690 14.0 Serial Peripheral Interface (SPI) ........................................................................... 691 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 21 14.1 Serial Peripheral Interface Memory Mapped Configuration Registers........................ 691 14.1.1 Serial Peripheral Interface (SPI) Register Address Map ................................ 691 14.1.1.1 Offset 00h: BIOS Flash Primary Region Register ................................. 693 14.1.1.2 Offset 04h: Hardware Sequencing Flash Status Register ...................... 694 14.1.1.3 Offset 06h: Hardware Sequencing Flash Control Register ..................... 696 14.1.1.4 Offset 08h: Flash Address Register ................................................... 697 14.1.1.5 Offset 10h: Flash Data 0 Register ..................................................... 698 14.1.1.6 Offset 14h: Flash Data [N] Register .................................................. 698 14.1.1.7 Offset 50h: Flash Regions Access Permissions Register ........................ 699 14.1.1.8 Offset 54h: Flash Region 0 (Flash Descriptor) Register ........................ 700 14.1.1.9 Offset 58h: Offset 00h: Flash Region 1 (BIOS Descriptor) Register ........ 700 14.1.1.10 Offset 5Ch: Flash Region 2 (Intel ME) Register ................................... 701 14.1.1.11 Offset 60h: Flash Region 3 Register .................................................. 701 14.1.1.12 Offset 64h: Flash Region 4 (Platform Data) Register ............................ 702 14.1.1.13 Offset 74h: Protected Range 0 Register ............................................. 702 14.1.1.14 Offset 78h: Protected Range 1 Register ............................................. 703 14.1.1.15 Offset 7Ch: Protected Range 2 Register ............................................. 704 14.1.1.16 Offset 80h: Protected Range 3 Register ............................................. 705 14.1.1.17 Offset 84h: Protected Range 4 Register ............................................. 706 14.1.1.18 Offset 90h: Software Sequencing Flash Status Register ....................... 707 14.1.1.19 Offset 91h: Software Sequencing Flash Control Register ...................... 708 14.1.1.20 Offset 94h: Prefix Opcode Configuration Register ................................ 710 14.1.1.21 Offset 96h: Opcode Type Configuration Register ................................. 711 14.1.1.22 Offset 98h: Opcode Menu Configuration Register ................................ 712 14.1.1.23 Offset A0h: BIOS Base Address Configuration Register ........................ 713 14.1.1.24 Offset B0h: Flash Descriptor Observability Control Register .................. 714 14.1.1.25 Offset B4h: Flash Descriptor Observability Data Register ..................... 714 14.1.1.26 Offset C0h: Additional Flash Control Register ..................................... 715 14.1.1.27 Offset C4h: Host Lower Vendor Specific Component Capabilities Register ........................................................................................ 716 14.1.1.28 Offset C8h: Host Upper Vendor Specific Component Capabilities Register ........................................................................................ 718 14.1.1.29 Offset D0h: Flash Partition Boundary ................................................ 720 14.2 Flash Descriptor Records................................................................................... 720 14.2.1 OEM Section .......................................................................................... 720 15.0 UART / WDT (SIW) ............................................................................................... 721 15.1 Overview ........................................................................................................ 721 15.2 LPC Logical Devices 4 and 5: Serial Ports (UART1 and UART2) ............................... 721 15.2.1 UART Register Details ............................................................................. 721 15.2.1.1 Offset 00h: RBR--Receive Buffer Register ........................................... 723 15.2.1.2 Offset 00h: THR--Transmit Holding Register ....................................... 723 15.2.1.3 Offset 01h: IER--Interrupt Enable Register ......................................... 724 15.2.1.4 Offset 02h: IIR--Interrupt Identification Register................................. 725 15.2.1.5 Offset 02h: FCR--FIFO Control Register.............................................. 727 15.2.1.6 Offset 03h: LCR--Line Control Register............................................... 728 15.2.1.7 Offset 04h: MCR--Modem Control Register ......................................... 730 15.2.1.8 Offset 05h: LSR--Line Status Register................................................ 731 15.2.1.9 Offset 06h: MSR--Modem Status Register........................................... 735 15.2.1.10 Offset 07h: SCR--Scratchpad Register ............................................... 736 15.2.1.11 Offset 00h: DLL--Programmable Baud Rate Generator Divisor Latch Register Low................................................................................... 736 15.2.1.12 Offset 01h: DLH--Programmable Baud Rate Generator Divisor Latch Register High.................................................................................. 737 15.3 Logical Device 6: Watchdog Timer ..................................................................... 737 15.3.1 Watchdog Timer Register Details .............................................................. 738 15.3.1.1 Offset 00h: PV1R0--Preload Value 1 Register 0 ................................... 738 15.3.1.2 Offset 01h: PV1R1--Preload Value 1 Register 1 ................................... 739 15.3.1.3 Offset 02h: PV1R2--Preload Value 1 Register 2 ................................... 739 Intel(R) Communications Chipset 89xx Series - Datasheet 22 October 2012 Order Number: 327879-001US Contents 15.3.1.4 Offset 04h: PV2R0--Preload Value 2 Register 0 ................................... 740 15.3.1.5 Offset 05h: PV2R1--Preload Value 2 Register 1 ................................... 740 15.3.1.6 Offset 06h: PV2R2--Preload Value 2 Register 2 ................................... 741 15.3.1.7 Offset 08h: GISR--General Interrupt Status Register ........................... 741 15.3.1.8 Offset 0Ch: RR0--Reload Register 0 .................................................. 742 15.3.1.9 Offset 0Dh: RR1--Reload Register 1 .................................................. 742 15.3.1.10 Offset 10h: WDTCR--WDT Configuration Register................................ 743 15.3.1.11 Offset 18h: WDTLR--WDT Lock Register ............................................ 744 15.4 SIW Configuration .......................................................................................... 745 15.4.1 SIW Configuration Register Summary ....................................................... 745 15.4.1.1 Global Control/Configuration Registers [00h - 2Fh].............................. 746 15.4.1.2 Logical Device Configuration Registers [30h - FFh] .............................. 747 16.0 Per Thread WDT (B0:D31:F7) ................................................................................ 752 16.1 PCI Registers ................................................................................................. 752 16.2 Register Attribute Legends................................................................................ 752 16.3 Configuration Space ........................................................................................ 753 16.3.1 PCI Header (B0:D31:F7) ......................................................................... 753 16.3.1.1 Offset 00h: ID--Identifiers ............................................................... 753 16.3.1.2 Offset 04h: CMD--Command ............................................................ 754 16.3.1.3 Offset 06h: DSTS--Device Status ...................................................... 755 16.3.1.4 Offset 08h: RID--Revision ID ........................................................... 756 16.3.1.5 Offset 09h: CC--Class Code ............................................................. 756 16.3.1.6 Offset 0Ch: CLS--Cache Line Size ..................................................... 757 16.3.1.7 Offset 0Dh: MLT--Master Latency Timer............................................. 757 16.3.1.8 Offset 0Eh: HTYPE--Header Type ...................................................... 757 16.3.1.9 Offset 10h: TBAR--WDT Table Base Address....................................... 758 16.3.1.10 Offset 14h: PBAR--WDT PBA Base Address......................................... 758 16.3.1.11 Offset 18h: CBAR--WDT CFG Base Address ........................................ 759 16.3.1.12 Offset 2Ch: SSVID--Subsystem Vendor ID ......................................... 759 16.3.1.13 Offset 2Eh: SSID Subsystem ID........................................................ 760 16.3.1.14 Offset 34h: CAP--Capabilities Pointer ................................................ 760 16.3.2 PCI MSI-X Capability .............................................................................. 760 16.3.2.1 Offset 80h: MID--MSI-X Capability ID................................................ 761 16.3.2.2 Offset 82h: MC--MSI-X Control & Status ............................................ 761 16.3.2.3 Offset 84h: MT--MSI-X Table ........................................................... 761 16.3.2.4 Offset 88h: MP--MSI-X Message PBA ................................................. 762 16.4 MMIO Space ................................................................................................... 762 16.4.1 Table BAR Range ................................................................................... 762 16.4.1.1 Offset 00h: WDT0MAW--WDT 0 Msg Add WARN.................................. 764 16.4.1.2 Offset 04h: WDT0MUAW--WDT 0 Msg Upper Add WARN....................... 764 16.4.1.3 Offset 08h: WDT0MDW--WDT 0 Msg Data WARN ................................ 765 16.4.1.4 Offset 0Ch: WDT0VCW--WDT 0 Vector Ctrl WARN ............................... 765 16.4.1.5 Offset 10h: WDT0MAR--WDT 0 Msg Add RESET .................................. 765 16.4.1.6 Offset 14h: WDT0MUAR--WDT 0 Msg Upper Add RESET ....................... 765 16.4.1.7 Offset 18h: WDT0MDR--WDT 0 Msg Data RESET................................. 766 16.4.1.8 Offset 1Ch: WDT0VCR--WDT 0 Vector Ctrl RESET ............................... 766 16.4.2 PBA BAR Range ..................................................................................... 767 16.4.2.1 Offset 00h: WDT_PBA0--Pending Bit Array 0 ...................................... 768 16.4.2.2 Offset 04h: WDT_PBA1--Pending Bit Array 1 ...................................... 769 16.4.2.3 Offset 08h: WDT_PBA2--Pending Bit Array 2 ...................................... 770 16.4.2.4 Offset 0Ch: WDT_PBA3--Pending Bit Array 3 ...................................... 771 16.4.3 CFG BAR Range ..................................................................................... 772 16.4.3.1 Offset 00h: WDT0COUNT--WDT Count Register .................................. 772 16.4.3.2 Offset 04h: WDT0CMD--WDT Cmd/Status Register.............................. 773 16.4.3.3 Offset 200h: WDT_PSCALE............................................................... 773 16.4.3.4 Offset 204h: WDT_GBLCFG .............................................................. 774 17.0 Thermal Sensor Registers (B0:D31:F6).................................................................. 775 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 23 17.1 Thermal Configuration Registers ........................................................................ 775 17.1.1 PCI Configuration Topology ...................................................................... 775 17.1.2 Configuration Register Access Restrictions ................................................. 775 17.1.3 Temperature Alert .................................................................................. 775 17.1.4 Register Attribute Legend ........................................................................ 776 17.2 PCI Configuration Registers (Thermal Sensor - B0:D31:F6) ................................... 776 17.2.1 Thermal Sensor PCI Register Address Map ................................................. 776 17.2.1.1 Offset 00h: VID--Vendor Identification ............................................... 778 17.2.1.2 Offset 02h: DID--Device Identification ............................................... 778 17.2.1.3 Offset 04h: CMD--Command ............................................................ 778 17.2.1.4 Offset 06h: STS--Status................................................................... 779 17.2.1.5 Offset 08h: RID--Revision Identification ............................................. 780 17.2.1.6 Offset 09h: PI--Programming Interface .............................................. 781 17.2.1.7 Offset 0Ah: SCC--Sub Class Code...................................................... 781 17.2.1.8 Offset 0Bh: BCC--Base Class Code .................................................... 781 17.2.1.9 Offset 0Ch: CLS--Cache Line Size ..................................................... 782 17.2.1.10 Offset 0Dh: LT--Latency Timer.......................................................... 782 17.2.1.11 Offset 0Eh: HTYPE--Header Type....................................................... 782 17.2.1.12 Offset 10h: TBAR--Thermal Base ...................................................... 783 17.2.1.13 Offset 14h: TBARH--Thermal Base High DWord .................................. 783 17.2.1.14 Offset 2Ch: SVID--Subsystem Vendor ID ........................................... 784 17.2.1.15 Offset 2Eh: SID--Subsystem ID ........................................................ 785 17.2.1.16 Offset 34h: CP--Capabilities Pointer ................................................... 785 17.2.1.17 Offset 3Ch: INTLN--Interrupt Line ..................................................... 786 17.2.1.18 Offset 3Dh: INTPN--Interrupt Pin ...................................................... 786 17.2.1.19 Offset 40h: TBARB--BIOS Assigned Thermal Base Address .................. 787 17.2.1.20 Offset 44h: TBARBH--BIOS Assigned Thermal Base High DWord ........... 787 17.2.1.21 Offset 50h: PID--PCI Power Management Capability ID ........................ 788 17.2.1.22 Offset 52h: PC--Power Management Capabilities ................................. 788 17.2.1.23 Offset 54h: PCS--Power Management Control And Status ..................... 789 17.2.1.24 Offset 80h: MID--Message Signaled Interrupt Identifiers ...................... 790 17.2.1.25 Offset 82h: MC--Message Signaled Interrupt Message Control............... 790 17.2.1.26 Offset 84h: MA--Message Signaled Interrupt Message Address .............. 791 17.2.1.27 Offset 88h: MD--Message Signaled Interrupt Message Data .................. 791 17.2.1.28 Offset 94h: CCTS0RD--Thermal Sensor 0 Remote Diode....................... 791 17.2.1.29 Offset 96h: CCTS0SSI--Thermal Sensor 0 Sense Stage Inputs .............. 792 17.2.1.30 Offset 98h: CCTS0DSAC0--Thermal Sensor 0 Delta Sigma ADC Control 0........................................................................................ 793 17.2.1.31 Offset 9Ch: CCTS0DSAC 1--Thermal Sensor 0 Delta Sigma ADC Control 1........................................................................................ 793 17.2.1.32 Offset A0h: CCTS0C--Thermal Sensor 0 Calibration ............................. 794 17.2.1.33 Offset A4h: CCTS0TL0--Thermal Sensor 0 Top Logic 0 ......................... 794 17.2.1.34 Offset A8h: CCTS0TL1--Thermal Sensor 0 Top Logic 1 ......................... 796 17.2.1.35 Offset ACh: CCTS0BP--Thermal Sensor 0 Bandgap Pins........................ 797 17.2.1.36 Offset B4h: CCTS1RD--Thermal Sensor 1 Remote Diode....................... 797 17.2.1.37 Offset B6h: CCTS1SSI--Thermal Sensor 1 Sense Stage Inputs .............. 798 17.2.1.38 Offset B8h: CCTS1DSAC0--Thermal Sensor 1 Delta Sigma ADC Control 0........................................................................................ 798 17.2.1.39 Offset BCh: CCTS1DSAC1--Thermal Sensor 1 Delta Sigma ADC Control 1........................................................................................ 799 17.2.1.40 Offset C0h: CCTS1C--Thermal Sensor 1 Calibration ............................. 799 17.2.1.41 Offset C4h: CCTS1TL0--Thermal Sensor 1 Top Logic 0 ......................... 800 17.2.1.42 Offset C8h: CCTS1TL1--Thermal Sensor 1 Top Logic 1 ......................... 801 17.2.1.43 Offset CCh: CCTS1BP--Thermal Sensor 1 Bandgap Pins........................ 802 17.2.1.44 Offset F0h: Reserved ....................................................................... 802 17.2.1.45 Offset F8h: MANID--Manufacturing/Process ....................................... 802 17.2.2 PCH Thermal Reporting ........................................................................... 803 17.2.2.1 DIMM Thermal Reporting Configurations............................................. 803 17.2.3 Thermal Reporting Registers .................................................................... 803 Intel(R) Communications Chipset 89xx Series - Datasheet 24 October 2012 Order Number: 327879-001US Contents 17.2.3.1 17.2.3.2 17.2.3.3 17.2.3.4 17.2.3.5 17.2.3.6 17.2.3.7 17.2.3.8 17.2.3.9 17.2.3.10 17.2.3.11 17.2.3.12 17.2.3.13 17.2.3.14 17.2.3.15 17.2.3.16 17.2.3.17 17.2.3.18 17.2.3.19 17.2.3.20 17.2.3.21 17.2.3.22 17.2.3.23 17.2.3.24 17.2.3.25 17.2.3.26 17.2.3.27 17.2.3.28 17.2.3.29 17.2.3.30 17.2.3.31 17.2.3.32 17.2.3.33 17.2.3.34 17.2.3.35 17.2.3.36 17.2.3.37 17.2.3.38 17.2.3.39 17.2.3.40 17.2.3.41 17.2.3.42 17.2.3.43 17.2.3.44 17.2.3.45 17.2.3.46 17.2.3.47 17.2.3.48 17.2.3.49 17.2.3.50 17.2.3.51 18 Offset 00h: TS0IU--Thermal Sensor 0 In Use ..................................... 805 Offset 01h: TS0C--Thermal Sensor 0 Control...................................... 806 Offset 02h: TS0S--Thermal Sensor 0 Status ....................................... 807 Offset 03h: TS0TR--Thermal Sensor 0 Thermometer Read ................... 808 Offset 04h: TS0TTP--Thermal Sensor 0 Temperature Trip Point............. 808 Offset 08h: TS0CO--Thermal Sensor 0 Calibration Offset ..................... 810 Offset 0Bh: TS0TTPC--Thermal Sensor 0 Temperature Trip Point Cont'd. 810 Offset 0Ch: TS0ES--Thermal Sensor 0 Error Status ............................. 811 Offset 0Dh: TS0GPEN--Thermal Sensor 0 General Purpose Event Enables ......................................................................................... 812 Offset 0Eh: TS0PC--Thermal Sensor 0 Policy Control ........................... 813 Offset 10h: CPEC--CPU Power Error Correction Data............................ 814 Offset 12h: C0TA--CPU0 Temperature Adjust ..................................... 814 Offset 16h: C1TA--CPU1 Temperature Adjust ..................................... 815 Offset 1Ah: MC--MPC Control .......................................................... 815 Offset 22h: Timestamp .................................................................... 816 Offset 24h: DIMMEN--DIMM Enable................................................... 817 Offset 30h: C0TV--CPU0 Temperature Value ...................................... 818 Offset 32h: C1TV--CPU1 Temperature Value ...................................... 819 Offset 34h: C0EV--CPU0 Energy Value............................................... 819 Offset 38h: C1EV--CPU1 Energy Value............................................... 819 Offset 3Fh: AE--Alert Enable ............................................................ 820 Offset 40h: TS1IU--Thermal Sensor 1 In Use ..................................... 821 Offset 41h: TS1C--Thermal Sensor 1 Control...................................... 821 Offset 42h: TS1S--Thermal Sensor 1 Status ....................................... 823 Offset 43h: TS1TR--Thermal Sensor 1 Thermometer Read ................... 824 Offset 44h: TS1TTP--Thermal Sensor 1 Temperature Trip Point............. 824 Offset 48h: TS1CO--Thermal Sensor 1 Calibration Offset ..................... 826 Offset 4Bh: TS1TTPC--Thermal Sensor 1 Temperature Trip Point Cont'd. 827 Offset 4Ch: TS1ES--Thermal Sensor 1 Error Status ............................. 827 Offset 4Dh: TS1GPEN--Thermal Sensor 1 General Purpose Event Enables ......................................................................................... 829 Offset 4Eh: TS1PC--Thermal Sensor 1 Policy Control ........................... 830 Offset 50h: HTS--HOST Turbo Status ............................................... 831 Offset 56h: MTL--MCP Temperature Limit .......................................... 831 Offset 58h: MTV--MCH Temperature Value ......................................... 832 Offset 60h: MCPTV--MCP Temperature Value...................................... 832 Offset 64h: MMPC--Max MCH Power Clamp ........................................ 832 Offset 66h: MMCPPC--Max MCP Power Clamp ..................................... 833 Offset 82h: TS0PIEN--Thermal Sensor 0 PCI Interrupt Event Enables .... 833 Offset 83h: TS0LOCK--Thermal Sensor 0 Register Lock Controls ........... 834 Offset 98h: STS--SMBus Turbo Status ............................................... 834 Offset 9Ch: SEC--SMBus Event Clear................................................. 835 Offset A4h: TC3--Thermal Compares 3 .............................................. 835 Offset A8h: TC1--Thermal Compares 1 .............................................. 836 Offset ACh: TC2--Thermal Compares 2 .............................................. 836 Offset B0h: DIMM0--CPU0 DIMM Data ............................................... 837 Offset B4h: DIMM1--CPU1 DIMM Data .............................................. 839 Offset B8h: DIMMID--DIMM ID ......................................................... 840 Offset BCh: ECPCLAMP--EC Power Clamp Data ................................... 842 Offset C2h: TS1PIEN--Thermal Sensor 1 PCI Interrupt Event Enables .... 842 Offset C3h: TS1LOCK--Thermal Sensor 1 Register Lock Controls ........... 843 Offset D8h: ITV--Internal Temperature Values.................................... 843 Intel(R) Management Engine Interface (MEI - B0:D22:F0/F1) ................................. 844 18.1 First Intel Management Engine Interface (MEI1) Configuration Registers (MEI1 -- B0:D22:F0) .................................................................................................... 844 18.1.1 VID--Vendor Identification Register .......................................................... 845 18.1.2 DID--Device Identification Register .......................................................... 845 18.1.3 PCICMD--PCI Command Register ............................................................. 846 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 25 18.1.4 PCISTS--PCI Status Register .................................................................... 846 18.1.5 RID--Revision Identification Register......................................................... 847 18.1.6 CC--Class Code Register.......................................................................... 847 18.1.7 HTYPE--Header Type Register .................................................................. 847 18.1.8 MEI1_MBAR--MEI1 MMIO Base Address Register ........................................ 848 18.1.9 SVID--Subsystem Vendor ID Register ....................................................... 848 18.1.10 SID--Subsystem ID Register .................................................................... 848 18.1.11 CAPP--Capabilities List Pointer Register ..................................................... 848 18.1.12 INTR--Interrupt Information Register ........................................................ 849 18.1.13 MLMG--Maximum Latency/Minimum Grant Register..................................... 849 18.1.14 HFS--Host Firmware Status Register ......................................................... 849 18.1.15 ME_UMA--Management Engine UMA Register ............................................. 850 18.1.16 GMES--General ME Status ....................................................................... 850 18.1.17 H_GS--Host General Status ..................................................................... 850 18.1.18 PID--PCI Power Management Capability ID Register .................................... 851 18.1.19 PC--PCI Power Management Capabilities Register ....................................... 851 18.1.20 PMCS--PCI Power Management Control and Status Register ......................... 852 18.1.21 MID--Message Signaled Interrupt Identifiers Register .................................. 852 18.1.22 MC--Message Signaled Interrupt Message Control Register .......................... 853 18.1.23 MA--Message Signaled Interrupt Message Address Register.......................... 853 18.1.24 MUA--Message Signaled Interrupt Upper Address Register ........................... 853 18.1.25 MD--Message Signaled Interrupt Message Data Register.............................. 853 18.1.26 HIDM--MEI Interrupt Delivery Mode .......................................................... 854 18.1.27 HERES--MEI Extend Register Status.......................................................... 854 18.1.28 HERX--MEI Extend Register DWX.............................................................. 855 18.2 MEI1_MBAR: MEI1 MMIO Registers .................................................................... 855 18.2.1 H_CB_WW--Host Circular Buffer Write Window ........................................... 855 18.2.2 H_CSR--Host Control Status .................................................................... 856 18.2.3 ME_CB_RW--ME Circular Buffer Read Window ............................................ 856 18.2.4 ME CSR_HA--ME Control Status Host Access .............................................. 857 18.3 Second Management Engine Interface (MEI2) Configuration Registers (MEI2--B0:D22:F1) ......................................................................................... 858 18.3.1 VID--Vendor Identification Register .......................................................... 858 18.3.2 DID--Device Identification Register ........................................................... 859 18.3.3 PCICMD--PCI Command Register.............................................................. 859 18.3.4 PCISTS--PCI Status Register .................................................................... 860 18.3.5 RID--Revision Identification Register......................................................... 860 18.3.6 CC--Class Code Register.......................................................................... 860 18.3.7 HTYPE--Header Type Register .................................................................. 860 18.3.8 MEI_MBAR--MEI MMIO Base Address Register ............................................ 861 18.3.9 SVID--Subsystem Vendor ID Register ....................................................... 861 18.3.10 SID--Subsystem ID Register .................................................................... 861 18.3.11 CAPP--Capabilities List Pointer Register ..................................................... 861 18.3.12 INTR--Interrupt Information Register ........................................................ 862 18.3.13 MLMG--Maximum Latency/Minimum Grant Register..................................... 862 18.3.14 HFS--Host Firmware Status Register ......................................................... 862 18.3.15 GMES--General ME Status ....................................................................... 862 18.3.16 H_GS--Host General Status ..................................................................... 863 18.3.17 PID--PCI Power Management Capability ID Register .................................... 863 18.3.18 PC--PCI Power Management Capabilities Register ....................................... 863 18.3.19 PMCS--PCI Power Management Control and Status Register ......................... 864 18.3.20 MID--Message Signaled Interrupt Identifiers Register .................................. 864 18.3.21 MC--Message Signaled Interrupt Message Control Register .......................... 865 18.3.22 MA--Message Signaled Interrupt Message Address Register.......................... 865 18.3.23 MUA--Message Signaled Interrupt Upper Address Register ........................... 865 Intel(R) Communications Chipset 89xx Series - Datasheet 26 October 2012 Order Number: 327879-001US Contents 18.3.24 MD--Message Signaled Interrupt Message Data Register ............................. 865 18.3.25 HIDM--MEI Interrupt Delivery Mode ......................................................... 866 18.3.26 HERES--MEI Extend Register Status ......................................................... 866 18.3.27 HERX--MEI Extend Register DWX ............................................................. 867 18.4 MEI2_MBAR--MEI2 MMIO Registers ................................................................... 867 18.4.1 H_CB_WW--Host Circular Buffer Write Window .......................................... 867 18.4.2 H_CSR--Host Control Status .................................................................... 868 18.4.3 ME_CB_RW--ME Circular Buffer Read Window ............................................ 868 18.4.4 ME CSR_HA--ME Control Status Host Access.............................................. 869 PCIe EndPoint & GbE - Volume 2 of 4 ......................................... 870 19.0 PCI Express* EndPoint Introduction ...................................................................... 871 19.1 PCI Express* EndPoint (EP) .............................................................................. 871 19.1.1 Overview .............................................................................................. 871 19.1.2 Feature List ........................................................................................... 871 19.1.3 EP Interface Block Diagram ..................................................................... 872 19.1.4 EP Functional Description ........................................................................ 873 19.1.4.1 Transmit Interface (TI) .................................................................... 873 19.1.4.2 Receive Interface (RI) ..................................................................... 873 19.1.4.3 PCIe* EP Functions ......................................................................... 874 19.1.4.4 EP Function Mapping ....................................................................... 874 19.1.4.5 EP Mapping of BARs to MMIO............................................................ 875 19.1.5 PCIe* EP Interrupts................................................................................ 875 19.1.5.1 INTx ............................................................................................. 875 19.1.5.2 MSI-X ........................................................................................... 875 19.1.6 PCIe* EP Errors ..................................................................................... 876 19.1.6.1 PCIe* Error Management ................................................................. 876 19.1.6.2 PCIe* Error Reporting Mechanisms.................................................... 876 19.1.6.3 PCIe* Error Handling and Signalling .................................................. 877 19.1.6.4 PCIe* Error Sources ........................................................................ 877 19.1.7 Device Specific Error Management............................................................ 881 19.1.7.1 Memory Error Poisoning................................................................... 881 19.1.8 PCIe* Request Generation....................................................................... 881 19.1.8.1 64-bit Memory Accesses .................................................................. 883 19.1.8.2 Completions................................................................................... 883 19.1.8.3 Messages....................................................................................... 885 19.2 Intel(R) QuickAssist Technology .......................................................................... 886 19.2.1 Features ............................................................................................... 886 19.2.2 EP Intel(R) QuickAssist Integrated Accelerator (IQIA).................................... 887 19.2.3 Ring Controller....................................................................................... 887 19.2.3.1 Features ........................................................................................ 887 19.2.3.2 Functionality .................................................................................. 887 20.0 PCIe Endpoint Function 0 Registers ...................................................................... 889 20.1 EP PF PCI Configuration Space .......................................................................... 889 20.2 Detailed Register Summaries ............................................................................ 891 20.2.1 PCI Views.............................................................................................. 891 20.2.2 PCI Standard Header Registers ................................................................ 896 20.2.2.1 PVID--PF Vendor Identification Register ............................................. 896 20.2.2.2 PDID--PF Device Identification Register ............................................. 896 20.2.2.3 PPCICMD--PF Device Command Register............................................ 897 20.2.2.4 PPCISTS--PF Device Status Register.................................................. 898 20.2.2.5 PCC--PF Class Code Register ............................................................ 900 20.2.2.6 PHDR--PF Header Type Register ....................................................... 901 20.2.2.7 PeQATLBAR--PF QAT Lower Base Address Register .............................. 901 20.2.2.8 QATUBAR--PF QAT Upper Base Address Register................................. 902 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 27 20.2.2.9 PMISCLBAR--PF Miscellaneous Lower Base Address Register ................. 902 20.2.2.10 PMISCUBAR--PF Miscellaneous Upper Base Address Register................. 903 20.2.2.11 PETRINGCSRLBAR--PF Ring CSR Lower Base Address Register .............. 903 20.2.2.12 PETRINGCSRUBAR--PF Ring CSR Upper Base Address Register.............. 903 20.2.2.13 PSVID--PF Subsystem Vendor ID Register .......................................... 904 20.2.2.14 PSID--PF Subsystem ID Register....................................................... 904 20.2.2.15 PCP--PF Capabilities Pointer Register ................................................. 905 20.2.2.16 PIRQL--PF Interrupt Line Register ..................................................... 905 20.2.2.17 PIRQP--PF Interrupt Pin Register ....................................................... 906 20.2.2.18 SKU: SKU Register .......................................................................... 906 20.2.2.19 SSKU Register 2.............................................................................. 907 20.2.3 MSI-X Capability Structure....................................................................... 908 20.2.3.1 PMSI-X--PF Message Signalled Interrupt X Capability ID Register .......... 908 20.2.3.2 PMSIXNCP--PF MSIX Next Capability Pointer Register........................... 908 20.2.3.3 PMSIXCNTL--PF Message Signalled Interrupt X Control Register ............ 909 20.2.3.4 PMSIXTBIR--PF MSI-X Table Offset & Table BIR Register ...................... 909 20.2.3.5 PMSIXPBABIR--PF MSI-X Pending Bit Array & BIR Offset Register .......... 910 20.2.4 Power Management Capability Structure .................................................... 910 20.2.4.1 PPMCAP--PF Power Management Capabilities ID Register...................... 910 20.2.4.2 PPMCP--PF Power Management Next Capability Pointer Register ............ 911 20.2.4.3 PPMC--PF Power Management Capabilities Register ............................. 911 20.2.4.4 PPMCSR--PF Power Management Control and Status Register................ 912 20.2.5 PCI Express Capability Structure............................................................... 914 20.2.5.1 PPCID--PF PCI Express Capability ID Register ..................................... 914 20.2.5.2 PPCP--PF PCI Express Next Capability Pointer Register ......................... 914 20.2.5.3 PPCR--PF PCI Express Capabilities Register......................................... 915 20.2.5.4 PPDCAP--PF PCI Express Device Capabilities Register........................... 915 20.2.5.5 PPDCNTL--PF PCI Express Device Control Register ............................... 917 20.2.5.6 PPDSTAT--PF PCI Express Device Status Register ................................ 918 20.2.5.7 PLCAPR--PF Link Capabilities Register ................................................ 919 20.2.5.8 PLCNTLR--PF Link Control Register .................................................... 921 20.2.5.9 PLSR--PF Link Status Register........................................................... 923 20.2.5.10 PDCAPR2--PF Device Capabilities 2 Register ....................................... 924 20.2.5.11 PDCNTR2--PF Device Control 2 Register ............................................. 925 20.2.5.12 PLCNTLR2--PF Link Control 2 Register................................................ 926 20.2.5.13 PLSR2--PF Link Status 2 Register ...................................................... 928 20.2.6 MSI Capability Structure.......................................................................... 928 20.2.6.1 PMSICID--PF Message Signalled Interrupt Capability ID Register ........... 928 20.2.6.2 PMSINCP--PF Message Signalled Interrupt Next Capability Pointer Register ......................................................................................... 929 20.2.6.3 PMSICTL--PF Message Signalled Interrupt Control Register ................... 929 20.2.6.4 PMSILADDR--PF Message Signalled Interrupt Lower Address Register .... 930 20.2.6.5 PMSIUADDR--PF Message Signalled Interrupt Upper Address Register .... 930 20.2.6.6 PMSIDATA--PF Message Signalled Interrupt Data Register .................... 931 20.2.6.7 PMSIMSK--PF Message Signalled Interrupt Mask Register ..................... 931 20.2.6.8 PMSIPND--PF Message Signalled Interrupt Pending Register ................. 932 20.2.7 PF Advanced Error Reporting Capability Structure ....................................... 932 20.2.7.1 PPCIEAERCAPID--PF PCI Express AER Capability ID Register ................. 932 20.2.7.2 PPAERUCS--PF PCI Express AER Uncorrectable Error Status Register...... 933 20.2.7.3 PPAERUCM--PF PCI Express AER Uncorrectable Error Mask Register ....... 934 20.2.7.4 PPAERUCSEV--PF PCI Express AER Uncorrectable Error Severity Register ......................................................................................... 935 20.2.7.5 PPAERCS--PF PCI Express AER Correctable Error Register..................... 936 20.2.7.6 PPAERCM--PF PCI Express AER Correctable Error Mask Register ............ 937 20.2.7.7 PPAERCTLCAP--PF PCI Express AER Control and Capability Register ....... 938 20.2.7.8 PPAERHDRLOG0--PF PCI Express AER Header Log 0 Register ................ 938 20.2.7.9 PPAERHDRLOG1--PF PCI Express AER Header Log 1 Register ................ 939 20.2.7.10 PPAERHDRLOG2--PF PCI Express AER Header Log 2 Register ................ 939 20.2.7.11 PPAERHDRLOG3--PF PCI Express AER Header Log 3 Register ................ 940 Intel(R) Communications Chipset 89xx Series - Datasheet 28 October 2012 Order Number: 327879-001US Contents 20.2.8 PF Alternative Routing-ID Extended Capability Structure.............................. 940 20.2.8.1 PARIDHDR--PF Alternative Routing ID Capability Header ...................... 940 20.2.8.2 PFARICAP--PF ARI Capabilities Register ............................................. 941 20.2.8.3 PARIDCTL--PF Alternative Routing ID Control Register ......................... 942 20.2.9 PF SR-IOV Extended Capability Structure .................................................. 942 20.2.9.1 PSRIOVCAPID--PF SR-IOV Capability ID Register ................................ 942 20.2.9.2 PSRIOVCVNC--PF SRIOV Capability Version and Next Capability Pointer Register ........................................................................................ 943 20.2.9.3 PSRIOVCAP--PF SRIOV Capabilities Register....................................... 944 20.2.9.4 PSRIOVCS--PF SRIOV Control and Status Register .............................. 944 20.2.9.5 PSRIOVMTOTINI--PF SRIOV Initial and Total VFs Register .................... 944 20.2.9.6 PSRIOVNUMVF--PF SRIOV Number of VFs Register.............................. 945 20.2.9.7 PSRIOVFVFO--PF SRIOV First VF Offset Register ................................. 945 20.2.9.8 PSRIOVVFS--PF SRIOV VF Stride Register .......................................... 946 20.2.9.9 PSRIOVFDID--PF SRIOV VF Device ID Register ................................... 946 20.2.9.10 PSRIOVPAGESIZE--PF SRIOV Supported Page Size Register ................. 947 20.2.9.11 PSRIOVSYSPS--PF SRIOV System Page Size Register .......................... 947 20.2.9.12 PSRIOVLBAR0--SRIOV Lower BAR0 Register....................................... 948 20.2.9.13 PSRIOVUBAR0--SRIOV Upper BAR0 Register ...................................... 949 20.2.9.14 PSRIOVLBAR1--SRIOV Lower BAR1 Register....................................... 950 20.2.9.15 PSRIOVUBAR1--SRIOV Upper BAR1 Register ...................................... 951 20.2.9.16 PSRIOVVFMA--PF SRIOV VF Migration Array Register........................... 952 20.3 EP VF PCI Configuration Space .......................................................................... 953 20.3.1 PCI Standard Header Registers ................................................................ 953 20.3.2 VF PCI Express Capability Structure .......................................................... 959 20.3.2.10 VDCAPR2[0:15]--VF Device Capabilities 2 Register.............................. 966 20.3.3 VF MSI Capability Structure ..................................................................... 967 20.3.3.7 VMSIMSK--VF Message Signalled Interrupt Mask Register .................... 969 20.3.3.8 VMSIPND--VF Message Signalled Interrupt Pending Register................. 970 20.3.4 VF Advanced Error Reporting Capability Structure....................................... 970 20.3.5 VF Alternative Routing ID Extended Capability Structure.............................. 980 20.4 EP Memory Mapped Registers ........................................................................... 981 20.4.1 Detailed Register Summary ..................................................................... 981 20.4.2 CSRs .................................................................................................... 981 20.4.2.1 ERRSOU2--Error Source Register 2 ................................................... 982 20.4.2.2 ERRMSK2--Error Source Mask Register 2 ........................................... 982 20.4.2.3 SINTPF--Signal Raw PF Interrupt Register .......................................... 983 20.4.2.4 SMIAPF--Signal IA PF Interrupt Mask Register .................................... 983 20.4.2.5 GBECFGMMIOV--GBE Configuration and MMIO Valid Register................ 984 20.4.2.6 SINTGBE[0:3]--Signal Raw PF Interrupt GBE Register ......................... 985 20.4.2.7 SMIAGBE[0:3]--Signal IA Interrupt Mask GBE Register ........................ 985 21.0 GbE Controller Overview........................................................................................ 987 21.1 Feature Summary ........................................................................................... 987 21.2 Scope ............................................................................................................ 988 21.3 Terminology and Acronyms............................................................................... 988 21.3.1 External Specification and Documents....................................................... 989 21.3.1.1 Network Interface Documents .......................................................... 990 21.3.1.2 Host Interface Documents................................................................ 990 21.3.1.3 Networking Protocol Documents........................................................ 990 21.3.1.4 Manageability Document.................................................................. 990 21.4 Product Overview ............................................................................................ 990 21.5 External Interface ........................................................................................... 991 21.5.1 PCIe Interface (Connected to PCI Express Through PCH PCIe End Point) ........ 991 21.5.2 Network Interfaces ................................................................................. 991 21.5.3 EEPROM Interface .................................................................................. 991 21.5.4 SMBus Interface .................................................................................... 992 21.5.5 MDIO/I2C Two-Wire Interfaces................................................................. 992 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 29 21.5.6 Software-Definable Pins (SDP) Interface (General-Purpose I/O) .................... 992 21.5.7 LED Interface......................................................................................... 993 21.6 GbE Controller Features and Capabilities ............................................................. 993 21.6.1 Network Interface................................................................................... 996 21.6.1.1 Quad Port ...................................................................................... 996 21.6.1.2 Shared MDIO Support ...................................................................... 996 21.6.1.3 I2C Clock Stretching ........................................................................ 996 21.6.1.4 1000BASE-KX Backplane Ethernet Interface........................................ 996 21.6.2 HOST Interface ...................................................................................... 996 21.6.2.1 PCIe Connectivity through EndPoint Interface...................................... 996 21.6.2.2 64-bit BAR Support ......................................................................... 996 21.6.3 Performance Features ............................................................................. 997 21.6.4 Receive and Transmit Queues .................................................................. 997 21.6.4.1 Unused Receive and Transmit Ports Buffer Sharing .............................. 997 21.6.5 Virtualization ......................................................................................... 997 21.6.5.1 Malicious Driver Detection ................................................................ 997 21.6.5.2 2-tuple Filtering .............................................................................. 997 21.6.6 Security Offload ..................................................................................... 997 21.6.6.1 Transmit Queue Prioritization ............................................................ 998 21.6.7 Manageability......................................................................................... 998 21.6.7.1 SMBus Interface to External BMC ...................................................... 998 21.6.7.2 Exclusive Management Filtering......................................................... 998 21.6.8 Time SYNC (IEEE1588 and IEEE 802.1AS) ................................................. 998 21.6.8.1 Per Packet Timestamp ..................................................................... 998 21.6.8.2 Improved System Time Accuracy....................................................... 998 21.6.8.3 SYSTIM Synchronized Pulse Generation on SDP Pins ............................ 998 21.6.8.4 Time SYNC Interrupts ...................................................................... 999 21.7 Device Data Flows ........................................................................................... 999 21.7.1 Transmit Data Flow................................................................................. 999 21.7.2 Rx Data Flow ....................................................................................... 1000 22.0 GbE Interconnects ............................................................................................... 1001 22.1 PCIe Interface ............................................................................................... 1001 22.1.1 Special GbE Controller Features .............................................................. 1001 22.1.2 Relaxed Ordering.................................................................................. 1001 22.1.3 Snoop Not Required .............................................................................. 1002 22.1.4 No Snoop and Relaxed Ordering for LAN Traffic ........................................ 1002 22.1.4.1 No-Snoop Option for Payload .......................................................... 1003 22.1.5 PCIe Power Management ....................................................................... 1003 22.2 Management Interfaces .................................................................................. 1003 22.2.1 SMBus ................................................................................................ 1003 22.2.1.1 Channel Behavior .......................................................................... 1003 22.3 EEPROM ....................................................................................................... 1013 22.3.1 EEPROM Interface................................................................................. 1013 22.3.1.1 General Overview.......................................................................... 1013 22.3.1.2 EEPROM Device............................................................................. 1014 22.3.1.3 HW Initial Load Process.................................................................. 1014 22.3.1.4 Software Accesses......................................................................... 1017 22.3.1.5 Signature Field ............................................................................. 1018 22.3.1.6 EEPROM Recovery ......................................................................... 1018 22.3.2 Shared EEPROM ................................................................................... 1018 22.3.2.1 EEPROM Deadlock Avoidance .......................................................... 1018 22.3.2.2 EEPROM Map Shared Words............................................................ 1019 22.3.3 Vital Product Data (VPD) Support............................................................ 1019 22.4 Configurable I/O Pins ..................................................................................... 1020 22.4.1 General-Purpose I/O (Software-Definable Pins)......................................... 1020 Intel(R) Communications Chipset 89xx Series - Datasheet 30 October 2012 Order Number: 327879-001US Contents 22.4.2 Software Watchdog ...............................................................................1021 22.4.2.1 Watchdog Rearm ...........................................................................1021 22.4.3 LEDs ...................................................................................................1021 22.5 Network Interfaces .........................................................................................1022 22.5.1 Overview .............................................................................................1022 22.5.2 MAC Functionality .................................................................................1023 22.5.2.1 Internal GMII/MII Interface .............................................................1023 22.5.2.2 MDIO/MDC PHY Management Interface .............................................1023 22.5.3 SerDes, SGMII and 1000BASE-KX Support ...............................................1025 22.5.3.1 SerDes, SGMII and 1000BASE-KX PCS Block .....................................1025 22.5.3.2 GbE Physical Coding Sub-Layer (PCS)...............................................1025 22.5.4 Auto-Negotiation and Link Setup Features ................................................1026 22.5.4.1 SerDes Link Configuration ...............................................................1027 22.5.4.2 1000BASE-KX Link Configuration .....................................................1029 22.5.4.3 SGMII Link Configuration ................................................................1029 22.5.4.4 Loss of Signal/Link Status Indication ................................................1030 22.5.5 Ethernet Flow Control (FC) .....................................................................1030 22.5.5.1 MAC Control Frames and Receiving Flow Control Packets.....................1032 22.5.5.2 PAUSE and MAC Control Frames Forwarding ......................................1033 22.5.5.3 Transmission of PAUSE Frames ........................................................1034 22.5.5.4 IPG Control and Pacing ...................................................................1035 22.5.6 Loopback Support .................................................................................1035 22.5.6.1 General ........................................................................................1035 22.5.6.2 MAC Loopback...............................................................................1036 22.5.6.3 SerDes, SGMII and 1000BASE-KX Loopback ......................................1036 23.0 GbE Initialization..................................................................................................1038 23.1 Power Up ......................................................................................................1038 23.1.1 Power-Up Sequence ..............................................................................1038 23.1.2 Power-Up Timing Diagram ......................................................................1039 23.2 Reset Operation .............................................................................................1040 23.2.1 Hardware-Based Reset Sources...............................................................1040 23.2.1.1 GBE_AUX_PWR_OK........................................................................1040 23.2.1.2 PCIE_EP_RST#..............................................................................1040 23.2.1.3 Function Level Reset (FLR)..............................................................1040 23.3 Software Based Reset Sources .........................................................................1041 23.3.1 Full Software Reset (DEV_RST and RST) ..................................................1041 23.3.2 BME (Bus Master Enable) .......................................................................1042 23.3.2.1 Transaction Flow for BME when GIO Master Disable Algorithm is Used...1042 23.3.2.2 Transaction Flow for BME when GIO Master Disable Algorithm is Not Used ............................................................................................1042 23.3.3 Force TCO ............................................................................................1043 23.3.4 Firmware Reset.....................................................................................1043 23.3.5 EEPROM Reset ......................................................................................1043 23.3.6 External Phy Reset ................................................................................1043 23.3.7 CSR Access Flow Following Power-on and Reset Events ..............................1043 23.3.8 Reset Effects ........................................................................................1044 23.4 Software Initialization and Diagnostics ..............................................................1047 23.4.1 Power Up State.....................................................................................1048 23.4.2 Initialization Sequence ...........................................................................1048 23.4.3 Interrupts During Initialization ................................................................1048 23.4.4 Global Reset and General Configuration....................................................1049 23.4.5 Flow Control Setup ................................................................................1049 23.4.6 Link Setup Mechanisms and Control/Status Bit Summary ...........................1049 23.4.6.1 MAC/SERDES Link Setup (CTRL_EXT.LINK_MODE = 011b) ..................1049 23.4.6.2 MAC/SGMII Link Setup (CTRL_EXT.LINK_MODE = 010b).....................1050 23.4.6.3 MAC/1000BASE-KX Link Setup (CTRL_EXT.LINK_MODE = 001b) ..........1051 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 31 23.4.7 Initialization of Statistics ....................................................................... 1051 23.4.8 Receive Initialization ............................................................................. 1052 23.4.8.1 Initialize the Receive Control Register .............................................. 1052 23.4.8.2 Dynamic Enabling and Disabling of Receive Queues ........................... 1052 23.4.9 Transmit Initialization ........................................................................... 1053 23.4.9.1 Dynamic Queue Enabling and Disabling ............................................ 1053 23.4.10 Virtualization Initialization Flow .............................................................. 1054 23.4.10.1 VMDq Mode .................................................................................. 1054 23.5 Access to Shared Resources ............................................................................ 1055 23.5.1 Acquiring Ownership Over a Shared Resource........................................... 1055 23.5.2 Releasing Ownership Over a Shared Resource .......................................... 1056 24.0 Non-Volatile Memory Map - EEPROM .................................................................... 1057 24.1 EEPROM General Map ..................................................................................... 1057 24.2 EEPROM Configuration Notes ........................................................................... 1060 24.3 Hardware Accessed Words .............................................................................. 1061 24.3.1 Ethernet Address (LAN Base Address + Offsets 0x00-0x02)........................ 1061 24.3.2 Initialization Control Word 1 (Word 0x0A) ................................................ 1061 24.3.3 Subsystem ID (Word 0x0B) ................................................................... 1062 24.3.4 Subsystem Vendor ID (Word 0x0C)......................................................... 1062 24.3.5 Device ID (LAN Base Address + Offset 0x0D) ........................................... 1063 24.3.6 Vendor ID (Word 0x0E) ......................................................................... 1063 24.3.7 Initialization Control Word 2 (Word 0x0F) ................................................ 1063 24.3.8 EEPROM Sizing (Word 0x12) .................................................................. 1064 24.3.9 Initialization Control 4 (LAN Base Address + Offset 0x13) .......................... 1065 24.3.10 PHY_RST_CONTROL (LAN Base Address + Offset 0x16) ............................. 1066 24.3.11 Device Rev ID (Word 0x1E) ................................................................... 1066 24.3.12 LED 0 Configuration Defaults (LAN Base Address + Offset 0x1F) ................. 1066 24.3.13 Software Defined Pins Control (LAN Base Address + Offset 0x20)................ 1067 24.3.14 CGB Functions Control (Word 0x21) ........................................................ 1068 24.3.15 Initialization Control 3 (LAN Base Address + Offset 0x24) .......................... 1069 24.3.16 PCIe Control 3 (Word 0x29) ................................................................... 1070 24.3.17 Watchdog Configuration (Word 0x2E) ...................................................... 1071 24.3.18 VPD Pointer (Word 0x2F) ....................................................................... 1071 24.4 CSR Auto Configuration Pointer (LAN Base Address + Offset 0x17) ....................... 1071 24.4.1 CSR Configuration Section Length - Offset 0x0 ......................................... 1072 24.4.2 Block CRC8 (Offset 0x1) ........................................................................ 1072 24.4.3 CSR Address - Offset 0x2 ...................................................................... 1072 24.4.4 CSR Data LSB - Offset 0x3..................................................................... 1072 24.4.5 CSR Data MSB - Offset 0x4 .................................................................... 1072 24.5 CSR Auto Configuration Power-Up Pointer (LAN Base Address + Offset 0x27) ........ 1073 24.5.1 CSR Configuration Power-Up Section Length - Offset 0x0 ........................... 1073 24.5.2 Block CRC8 (Offset 0x1) ........................................................................ 1073 24.5.3 CSR Address - Offset 0x2 ...................................................................... 1074 24.5.4 CSR Data LSB - Offset 0x3..................................................................... 1074 24.5.5 CSR Data MSB - Offset 0x4 .................................................................... 1074 24.5.6 EICT Information Structure .................................................................... 1074 24.6 PHY Configuration Structure ............................................................................ 1075 24.6.1 PHY Configuration Section Length - Offset 0x0.......................................... 1075 24.6.2 Block CRC8 (Offset 0x1) ........................................................................ 1075 24.6.3 PHY Number and PHY Address - (Offset 2*n; [n = 1... Section Length])....... 1076 24.6.4 PHY Data (Offset 2*n + 1; [n = 1... Section Length]) ................................ 1076 24.7 Pointers and Control Words Used By Firmware ................................................... 1076 24.7.1 Pass Through LAN Configuration Pointer (LAN Base Address + Offset 0x11) ........................................................................................ 1076 Intel(R) Communications Chipset 89xx Series - Datasheet 32 October 2012 Order Number: 327879-001US Contents 24.7.2 PHY Configuration Pointer (Word 0x50) ....................................................1076 24.7.3 Firmware Patch Pointer (Word 0x51) .......................................................1077 24.7.4 Sideband Configuration Pointer (Word 0x57).............................................1077 24.7.5 Manageability Capability/Manageability Enable (Word 0x54) .......................1077 24.7.6 Management HW Config Control (Word 0x23) ...........................................1078 24.8 Firmware Patch Structure ................................................................................1078 24.8.1 Firmware Patch Data Size (Offset 0x0) .....................................................1078 24.8.2 Block CRC8 (Offset 0x1).........................................................................1078 24.8.3 Patch Ram Address Word (Offset 0x2) .....................................................1079 24.8.4 Patch Version 1 Word (Offset 0x3)...........................................................1079 24.8.5 Patch Version 2 Word (Offset 0x4)...........................................................1079 24.8.6 Patch Version 3 Word (Offset 0x5)...........................................................1079 24.8.7 Patch Version 4 Word (Offset 0x6)...........................................................1079 24.8.8 Patch Data Words (Offset 0x7, Block Length) ............................................1079 24.9 PT LAN Configuration Structure ........................................................................1080 24.9.1 PT LAN Configuration Structure Section Length - Offset 0x0 ........................1080 24.9.2 Block CRC8 (Offset 0x1).........................................................................1080 24.9.3 CSR Address - (Offset 2*n; [n = 1... Section Length]) ...............................1080 24.9.4 CSR Data LSB - (Offset 0x1 + 2*n; [n = 1... Section Length]) ....................1081 24.9.5 CSR Data MSB - (Offset 0x2 + 2*n; [n = 1... Section Length])....................1081 24.9.6 Manageability Filters..............................................................................1081 24.10 Sideband Configuration Structure .....................................................................1082 24.10.1 Section Length (Offset 0x0)....................................................................1082 24.10.2 Block CRC8 (Offset 0x1).........................................................................1082 24.10.3 SMBus Max Fragment Size (Offset 0x2)....................................................1082 24.10.4 SMBus Notification Timeout (Offset 0x3) ..................................................1082 24.10.5 SMBus Slave Address 0 1 (Offset 0x4) .....................................................1082 24.10.6 SMBus Slave Address 2 3 (Offset 0x5) .....................................................1083 24.10.7 Reserved (Offset 0x6)............................................................................1083 24.10.8 LAN Receive Enable 1 (Offset 0x7) ..........................................................1083 24.10.9 LAN Receive Enable 2 (Offset 0x8) ..........................................................1083 24.10.10 SMBus Flags (Offset 0x9) .......................................................................1084 24.10.11 LAN Receive Enable 3 (Offset 0xA) ..........................................................1084 24.11 Software Accessed Words ................................................................................1088 24.11.1 Compatibility (Word 0x03) .....................................................................1088 24.11.2 Port Identification LED Blinking (Word 0x04).............................................1089 24.11.3 OEM Specific (Word 0x06, 0x07) .............................................................1089 24.11.4 EEPROM Image Revision (Word 0x05)......................................................1089 24.11.5 PBA Number (Word 0x09) ......................................................................1090 24.11.5.1 PBA Structure ...............................................................................1090 24.11.6 PXE Main Setup Options (Word 0x30/0x34/0x38/0x3A) ..............................1090 24.11.7 PXE Configuration Customization Options (Word 0x31/0x35/0x39/0x3B) ......1091 24.11.8 PXE Boot Agent Version Number (Word 0x32) ...........................................1092 24.11.9 IBA Capabilities (Word 0x33) ..................................................................1092 24.11.10 PXE Reserved Words (Words 0x36, 0x3C, 0x3E)........................................1093 24.11.11 Alternate MAC Address Pointer (Word 0x37) .............................................1093 24.11.12 Checksum Word (Offset 0x3F) ................................................................1093 24.11.13 Image Unique ID (Word 0x42, 0x43) .......................................................1093 25.0 GbE Power Management .......................................................................................1094 25.1 General Power State Information......................................................................1094 25.2 Power States .................................................................................................1094 25.2.1 D0 Uninitialized (D0u) State ...................................................................1095 25.2.1.1 Entry Into D0u State ......................................................................1095 25.2.2 D0 Active (D0a) State............................................................................1095 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 33 25.2.2.1 Entry to D0a State ........................................................................ 1095 25.2.3 D3 State (PCI-PM D3hot)....................................................................... 1095 25.2.3.1 Entry to D3 State .......................................................................... 1096 25.2.3.2 Master Disable Via CTRL Register .................................................... 1096 25.2.4 Dr State (D3cold) ................................................................................. 1097 25.2.4.1 Dr Disable Mode............................................................................ 1097 25.2.4.2 Entry to Dr State........................................................................... 1098 25.2.4.3 Auxiliary Power Usage.................................................................... 1098 25.2.5 Device Power-Down State...................................................................... 1098 25.3 Timing of Power-State Transitions .................................................................... 1098 25.3.1 Power Up (Off to Dup to D0u to D0a) ...................................................... 1099 25.3.2 Transition From D0a to D3 and Back Without PCIE_EP_RST# ..................... 1100 25.3.3 Transition From D0a to D3 and Back With PCIE_EP_RST# .......................... 1101 25.3.4 Transition From D0a to Dr and Back Without Transition to D3 ..................... 1102 25.4 Wake Up ...................................................................................................... 1103 25.4.1 Advanced Power Management Wake Up................................................... 1103 25.4.2 PCIe Power Management Wake Up .......................................................... 1104 25.4.3 Wake-Up Packets.................................................................................. 1105 25.4.3.1 Pre-Defined Filters......................................................................... 1105 25.4.3.2 Flexible Filters .............................................................................. 1108 25.4.3.3 Wake Up Packet Storage ................................................................ 1110 26.0 GbE Inline Functions ............................................................................................ 1111 26.1 Receive Functionality...................................................................................... 1111 26.1.1 Receive Queues Assignment................................................................... 1111 26.1.1.1 Queuing in a Non-Virtualized Environment ........................................ 1113 26.1.1.2 Receive Queuing in a Virtualized Environment ................................... 1113 26.1.1.3 Queue Configuration Registers ........................................................ 1114 26.1.1.4 L2 Ether-Type Filters ..................................................................... 1115 26.1.1.5 2-Tuple Filters .............................................................................. 1115 26.1.1.6 Flex Filters ................................................................................... 1116 26.1.1.7 SYN Packet Filters ......................................................................... 1117 26.1.1.8 Receive-Side Scaling (RSS) ............................................................ 1117 26.1.2 L2 Packet Filtering ................................................................................ 1123 26.1.2.1 MAC Address Filtering .................................................................... 1124 26.1.2.2 VLAN Filtering............................................................................... 1126 26.1.2.3 Platform Manageability Filtering ...................................................... 1127 26.1.3 Receive Data Storage............................................................................ 1128 26.1.3.1 Host Buffers ................................................................................. 1128 26.1.3.2 On-Chip Receive Buffers................................................................. 1129 26.1.3.3 On-Chip Descriptor Buffers ............................................................. 1129 26.1.4 Legacy Receive Descriptor Format .......................................................... 1129 26.1.5 Advanced Receive Descriptors ................................................................ 1132 26.1.5.1 Advanced Receive Descriptors (RDESC) - Read Format....................... 1132 26.1.5.2 Advanced Receive Descriptors (RDESC) - Writeback Format ................ 1133 26.1.6 Receive Descriptor Fetching ................................................................... 1137 26.1.7 Receive Descriptor Write-Back................................................................ 1138 26.1.8 Receive Descriptor Ring Structure ........................................................... 1138 26.1.8.1 Low Receive Descriptors Threshold .................................................. 1140 26.1.9 Header Splitting and Replication ............................................................. 1140 26.1.9.1 Purpose ....................................................................................... 1140 26.1.9.2 Description................................................................................... 1141 26.1.10 Receive Packet Timestamp in Buffer ........................................................ 1143 26.1.11 Receive Packet Checksum Off Loading ..................................................... 1144 26.1.11.1 Filters Details ............................................................................... 1145 26.1.11.2 Receive UDP Fragmentation Checksum............................................. 1147 26.1.12 SCTP Offload........................................................................................ 1148 Intel(R) Communications Chipset 89xx Series - Datasheet 34 October 2012 Order Number: 327879-001US Contents 26.2 Transmit Functionality.....................................................................................1148 26.2.1 Packet Transmission ..............................................................................1148 26.2.1.1 Transmit Data Storage ...................................................................1149 26.2.1.2 On-Chip Transmit Buffers................................................................1149 26.2.1.3 On-Chip Descriptor Buffers..............................................................1149 26.2.1.4 Transmit Contexts .........................................................................1149 26.2.2 Transmit Descriptors .............................................................................1150 26.2.2.1 Legacy Transmit Descriptor Format ..................................................1151 26.2.2.2 Advanced Transmit Context Descriptor .............................................1154 26.2.2.3 Advanced Transmit Data Descriptor..................................................1156 26.2.2.4 Transmit Descriptor Ring Structure ..................................................1159 26.2.2.5 Transmit Descriptor Fetching ...........................................................1161 26.2.2.6 Transmit Descriptor Write-Back .......................................................1162 26.2.3 Transmit Completions Head Write Back ....................................................1163 26.2.3.1 Description ...................................................................................1163 26.2.4 TCP/UDP Segmentation..........................................................................1164 26.2.4.1 Assumptions .................................................................................1164 26.2.4.2 Transmission Process .....................................................................1164 26.2.4.3 TCP Segmentation Performance .......................................................1166 26.2.4.4 Packet Format ...............................................................................1167 26.2.4.5 TCP/UDP Segmentation Indication....................................................1167 26.2.4.6 Transmit Checksum Offloading with TCP/UD Segmentation ..................1169 26.2.4.7 TCP/UDP/IP Header Update .............................................................1170 26.2.4.8 IP/TCP/UDP Checksum Offloading ....................................................1172 26.2.4.9 Data Flow .....................................................................................1172 26.2.5 Checksum Offloading in Non-Segmentation Mode ......................................1173 26.2.5.1 IP Checksum .................................................................................1173 26.2.5.2 TCP Checksum ..............................................................................1174 26.2.5.3 SCTP CRC Offloading ......................................................................1174 26.2.5.4 Checksum Supported Per Packet Types.............................................1175 26.2.6 Multiple Transmit Queues .......................................................................1175 26.3 Interrupts .....................................................................................................1176 26.3.1 Mapping of Interrupt Causes...................................................................1176 26.3.1.1 Legacy and MSI Interrupt Modes......................................................1176 26.3.1.2 MSI-X Mode - VMDq Mode...............................................................1177 26.3.2 Legacy Interrupt Registers .....................................................................1178 26.3.2.1 Interrupt Cause Register (ICR) ........................................................1179 26.3.2.2 Interrupt Cause Set Register (ICS)...................................................1180 26.3.2.3 Interrupt Mask Set/Read Register (IMS) ...........................................1180 26.3.2.4 Interrupt Mask Clear Register (IMC) .................................................1180 26.3.2.5 Interrupt Acknowledge Auto-Mask Register (IAM)...............................1180 26.3.2.6 Extended Interrupt Cause Registers (EICR) .......................................1180 26.3.2.7 Extended Interrupt Cause Set Register (EICS) ...................................1181 26.3.2.8 Extended Interrupt Mask Set and Read Register (EIMS) & Extended Interrupt Mask Clear Register (EIMC) ...............................................1181 26.3.2.9 Extended Interrupt Auto Clear Enable Register (EIAC) ........................1181 26.3.2.10 Extended Interrupt Auto Mask Enable Register (EIAM) ........................1181 26.3.2.11 GPIE Register ................................................................................1181 26.3.3 MSI-X and Vectors ................................................................................1182 26.3.4 Interrupt Moderation .............................................................................1182 26.3.5 Clearing Interrupt Causes.......................................................................1185 26.3.5.1 Auto-Clear ....................................................................................1185 26.3.5.2 Write to Clear................................................................................1185 26.3.5.3 Read to Clear ................................................................................1185 26.3.6 Rate Controlled Low Latency Interrupts (LLI) ............................................1186 26.3.6.1 Rate Control Mechanism .................................................................1186 26.3.7 TCP Timer Interrupt...............................................................................1187 26.3.7.1 Introduction ..................................................................................1187 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 35 26.3.7.2 Description................................................................................... 1187 26.3.8 Special Cases About Interrupts ............................................................... 1188 26.3.8.1 Ordering Issue Between CSR Rd/Wr Received With Error by the Controller in the PCH and its Generated Interrupt .............................. 1188 26.4 802.1q VLAN Support ..................................................................................... 1188 26.4.1 802.1q VLAN Packet Format................................................................... 1188 26.4.2 802.1q Tagged Frames.......................................................................... 1188 26.4.3 Transmitting and Receiving 802.1q Packets.............................................. 1189 26.4.3.1 Adding 802.1q Tags on Transmits.................................................... 1189 26.4.3.2 Stripping 802.1q Tags on Receives .................................................. 1189 26.4.4 802.1q VLAN Packet Filtering ................................................................. 1189 26.4.5 Double VLAN Support............................................................................ 1190 26.4.5.1 Transmit Behavior With External VLAN ............................................. 1191 26.4.5.2 Receive Behavior With External VLAN............................................... 1191 26.5 Configurable LED Outputs ............................................................................... 1192 26.5.1 MODE Encoding for LED Outputs............................................................. 1192 26.6 Error Correction and Detection ........................................................................ 1193 26.7 CPU affinity Features ..................................................................................... 1193 26.7.1 Direct Cache Access (DCA) .................................................................... 1193 26.7.1.1 DCA Description ............................................................................ 1193 26.7.1.2 Details of Implementation .............................................................. 1194 26.8 Virtualization................................................................................................. 1195 26.8.1 Assignment of MSI-X Vectors to VM ........................................................ 1195 26.8.2 VM Resource Summary ......................................................................... 1196 26.8.3 Packet Switching (VMDq) Model.............................................................. 1196 26.8.3.1 VMDq Assumptions........................................................................ 1196 26.8.3.2 VM Selection................................................................................. 1196 26.8.3.3 L2 Filtering ................................................................................... 1196 26.8.3.4 VMDq Receive Packets Switching ..................................................... 1197 26.8.3.5 Mirroring Support .......................................................................... 1201 26.8.3.6 VMDq Offload Support ................................................................... 1201 26.8.3.7 Security Features .......................................................................... 1202 26.8.3.8 External Switch Loopback Support ................................................... 1204 26.8.3.9 Switch Control .............................................................................. 1205 26.8.4 Virtualization of the Hardware ................................................................ 1205 26.8.4.1 Per Pool Statistics.......................................................................... 1205 26.9 Time SYNC (IEEE1588 and IEEE 802.1AS)......................................................... 1206 26.9.1 Flow and Hardware/Software Responsibilities ........................................... 1206 26.9.1.1 TimeSync Indications in Receive and Transmit Packet Descriptors ........ 1208 26.9.2 Hardware Time Sync Elements ............................................................... 1208 26.9.2.1 System Time Structure and Mode of Operation .................................. 1208 26.9.2.2 Time Stamp Mechanism ................................................................. 1209 26.9.2.3 Time Adjustment Mode of Operation ................................................ 1210 26.9.3 Time Sync Related Auxiliary Elements ..................................................... 1210 26.9.3.1 Target Time.................................................................................. 1210 26.9.3.2 Configurable Frequency Clock ......................................................... 1211 26.9.3.3 Time Stamp Events ....................................................................... 1212 26.9.4 Time SYNC Interrupts ........................................................................... 1212 26.9.5 PTP Packet Structure............................................................................. 1212 26.10 Statistic Counters .......................................................................................... 1215 26.10.1 IEEE 802.3 Clause 30 Management ......................................................... 1215 26.10.2 OID_GEN_STATISTICS .......................................................................... 1216 26.10.3 RMON ................................................................................................. 1217 26.10.4 Linux net_device_stats .......................................................................... 1218 27.0 GbE Platform Manageability ................................................................................. 1220 27.1 Platform Configurations .................................................................................. 1220 Intel(R) Communications Chipset 89xx Series - Datasheet 36 October 2012 Order Number: 327879-001US Contents 27.1.1 On-Board BMC Configurations .................................................................1220 27.1.2 GbE Controller ......................................................................................1221 27.2 Pass Through Functionality ..............................................................................1221 27.2.1 SMBus Pass Through (PT) Functionality ....................................................1221 27.2.1.1 Pass Through (PT) Modes................................................................1221 27.3 Programming Interfaces ..................................................................................1221 27.3.1 SMBus Programming .............................................................................1221 27.3.1.1 Write SMBus Transactions (BMC to GbE Controller) ............................1222 27.3.1.2 Receive Enable Command ...............................................................1223 27.3.1.3 Read SMBus Transactions (Controller to BMC) ...................................1227 27.3.1.4 SMBus ARP Transactions.................................................................1236 27.4 Manageability Receive Filtering.........................................................................1239 27.4.1 Overview and General Structure..............................................................1239 27.4.2 L2 Filters .............................................................................................1240 27.4.2.1 MAC and VLAN filters......................................................................1240 27.4.2.2 EtherType Filters ...........................................................................1240 27.4.3 L3 and L4 Filters ...................................................................................1241 27.4.3.1 ARP Filtering .................................................................................1241 27.4.3.2 Neighbor Discovery Filtering............................................................1241 27.4.3.3 RMCP Port Filtering ........................................................................1241 27.4.3.4 Flex Port Filtering...........................................................................1241 27.4.3.5 The Controller IP Address Filtering ...................................................1241 27.4.3.6 Checksum Filter.............................................................................1241 27.4.4 Manageability Decision Filters .................................................................1241 27.4.4.1 Exclusive Traffic ............................................................................1244 27.4.5 Possible Configurations ..........................................................................1244 27.4.5.1 Dedicated MAC Packet Filtering ........................................................1244 27.4.5.2 Broadcast Packet Filtering ...............................................................1244 27.4.5.3 VLAN Packet Filtering .....................................................................1245 27.4.5.4 IPv6 Filtering ................................................................................1245 27.4.5.5 Receive Filtering with Shared IP - CPMP ............................................1245 28.0 GbE Programming Interface .................................................................................1246 28.1 Introduction ..................................................................................................1246 28.1.1 Memory, I/O Address, and Configuration Decoding ....................................1246 28.1.1.1 Memory-Mapped Access to Internal Registers and Memories ................1246 28.1.1.2 Memory-Mapped Access to MSI-X Tables...........................................1246 28.1.1.3 I/O-Mapped Access to Internal Registers and Memories ......................1246 28.1.1.4 Configuration Access to Internal Registers and Memories ....................1248 28.1.2 Register Conventions .............................................................................1249 28.1.2.1 Registers Byte Ordering ..................................................................1250 28.1.3 Detailed Register Summary ....................................................................1252 28.1.3.1 PCI Views .....................................................................................1252 28.1.4 Alias Addresses.....................................................................................1262 28.1.4.1 MSI-X BAR Register Summary .........................................................1264 28.1.4.2 Device Control Register--CTRL [0:3] (0x00000; R/W).........................1264 28.1.4.3 Device Status Register--STATUS [0:3] (0x0008; R)............................1268 28.1.4.4 Extended Device Control Register--CTRL_EXT [0:3] (0x0018; R/W)......1269 28.1.4.5 MDI Control Register--MDIC [0:3] (0x0020; R/W)..............................1272 28.1.4.6 MDC/MDIO Configuration Register--MDICNFG [0:3] (0x0E04; R/W)......1273 28.1.4.7 Copper/Fiber Switch Control--CONNSW [0:3] (0x0034; R/W) ..............1274 28.1.4.8 VLAN Ether Type--VET [0:3] (0x0038; R/W) .....................................1274 28.1.4.9 LED Control--LEDCTL [0:3] (0x0E00; RW) ........................................1275 28.2 Internal Packet Buffer Size Registers.................................................................1275 28.2.1 Detailed Register Descriptions.................................................................1276 28.2.1.1 Internal Receive Packet Buffer Size--IRPBS [0:3](0x2404; RO)............1276 28.2.1.2 Internal Transmit Packet Buffer Size--ITPBS [0:3] (0x3404; RO) .........1277 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 37 28.3 EEPROM Registers.......................................................................................... 1277 28.3.1 Detailed Register Descriptions ................................................................ 1277 28.3.1.1 EEPROM Control and Data Register--EEC [0:3] (0x0010; R/W)............ 1277 28.3.1.2 EEPROM Read Register - EERD [0:3] (0x0014; RW) ........................... 1279 28.3.1.3 EEPROM Diagnostic--EEDIAG [0:3] (0x1038; RO).............................. 1280 28.3.1.4 VPD Diagnostic Register--VPDDIAG [0:3] (0x1060; RO) ..................... 1281 28.3.1.5 Management--EEPROM CSR I/F....................................................... 1282 28.3.1.6 Management EEPROM Control Register--EEMNGCTL [0:3] (0x1010; RO) ............................................................................... 1282 28.3.1.7 Management EEPROM Read/Write Data--EEMNGDATA [0:3] (0x1014; RO) ............................................................................... 1283 28.4 Flow Control Registers .................................................................................... 1284 28.4.1 Detailed Register Descriptions ................................................................ 1284 28.4.1.1 Flow Control Address Low--FCAL [0:3] (0x0028; RO)......................... 1284 28.4.1.2 Flow Control Address High--FCAH [0:3] (0x002C; RO) ....................... 1285 28.4.1.3 Flow Control Type--FCT [0:3] (0x0030; R/W) ................................... 1285 28.4.1.4 Flow Control Transmit Timer Value--FCTTV [0:3] (0x0170; R/W)......... 1286 28.4.1.5 Flow Control Receive Threshold Low--FCRTL0 [0:3] (0x2160; R/W) ..... 1287 28.4.1.6 Flow Control Receive Threshold High--FCRTH0 [0:3] (0x2168; R/W) .... 1288 28.4.1.7 Flow Control Refresh Threshold Value--FCRTV[0:3] (0x2460; R/W)...... 1289 28.4.1.8 Flow Control Status--FCSTS0 [0:3] (0x2464; RO) ............................. 1290 28.5 GbE PCIe Registers ........................................................................................ 1290 28.5.1 Detailed Register Description ................................................................. 1290 28.5.1.1 Function Active and Power State to MNG--FACTPS[0:3] (0x5B30; RO) ............................................................................... 1290 28.5.1.2 Mirrored Revision ID--MREVID[0:3] (0x5B64; R/W)........................... 1292 28.5.1.3 PCIE Control Extended Register--GCR_EXT[0:3] (0x5B6C; R/W) ......... 1293 28.6 Semaphore Registers ..................................................................................... 1294 28.6.1 Detailed Register Descriptions ................................................................ 1294 28.6.1.1 Software Semaphore--SWSM[0:3] (0x5B50; R/W) ............................ 1294 28.6.1.2 Firmware Semaphore--FWSM[0:3] (0x5B54; R/WS) .......................... 1295 28.6.1.3 Software-Firmware Synchronization--SW_FW_SYNC[0:3] (0x5B5C; RWS) Check with LAD for Phy Related Bits.......................... 1297 28.6.1.4 Software Mailbox Write--SWMBWR[0:3] (0x5B04; R/W)..................... 1298 28.6.1.5 Software Mailbox 0--SWMB0[0:3] (0x5B08; RO) ............................... 1298 28.6.1.6 Software Mailbox 1--SWMB1[0:3] (0x5B0C; RO) ............................... 1298 28.6.1.7 Software Mailbox 2--SWMB2[0:3] (0x5B18; RO) ............................... 1299 28.6.1.8 Software Mailbox 3--SWMB3[0:3] (0x5B1C; RO) ............................... 1299 28.7 Interrupt Register Descriptions ........................................................................ 1299 28.7.1 Detailed Register Descriptions ................................................................ 1299 28.7.1.1 Extended Interrupt Cause--EICR[0:3] (0x1580; RC/W1C) .................. 1299 28.7.1.2 EICR Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) .............................................................. 1300 28.7.1.3 EICR Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1).. 1301 28.7.1.4 Extended Interrupt Cause Set--EICS [0:3] (0x1520; WO) .................. 1301 28.7.1.5 EICS Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) .............................................................. 1301 28.7.1.6 EICS Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) .. 1302 28.7.1.7 Extended Interrupt Mask Set/Read--EIMS [0:3] (0x1524; RWS).......... 1302 28.7.1.8 EIMS Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) .............................................................. 1302 28.7.1.9 EIMS Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) .............................................................. 1303 28.7.1.10 Extended Interrupt Mask Clear - EIMC [0:3] (0x1528; WO) ................ 1303 28.7.1.11 EIMC Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) .............................................................. 1304 28.7.1.12 EIMC Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) .............................................................. 1304 28.7.1.13 Extended Interrupt Auto Clear--EIAC [0:3] (0x152C; R/W)................. 1305 Intel(R) Communications Chipset 89xx Series - Datasheet 38 October 2012 Order Number: 327879-001US Contents 28.7.1.14 28.7.1.15 Extended Interrupt Auto Mask Enable--EIAM [0:3] (0x1530; R/W) .......1306 EIAM Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) ...............................................................1306 28.7.1.16 EIAM Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) ...............................................................1306 28.7.1.17 Interrupt Cause Read Register--ICR [0:3] (0x1500; RC/W1C) .............1307 28.7.1.18 Interrupt Cause Set Register--ICS [0:3] (0x1504; WO) ......................1309 28.7.1.19 Interrupt Mask Set/Read Register--IMS [0:3] (0x1508; R/W) ..............1311 28.7.1.20 Interrupt Mask Clear Register--IMC [0:3] (0x150C; WO) ....................1313 28.7.1.21 Interrupt Acknowledge Auto Mask Register--IAM [0:3] (0x1510; R/W) ..............................................................................1315 28.7.1.22 Interrupt Throttle--EITR [0:3] (0x1680 + 4*n [n = 0...9]; R/W) ..........1315 28.7.1.23 Interrupt Vector Allocation Registers--IVAR [0:3] (0x1700 + 4*n [n=0...3]; RW)........................................................1317 28.7.1.24 Interrupt Vector Allocation Registers--MISC IVAR_MISC [0:3] (0x1740; RW) ...............................................................................1318 28.7.1.25 General Purpose Interrupt Enable--GPIE [0:3] (0x1514; RW) ..............1319 28.8 MSI-X Table Register Descriptions ....................................................................1319 28.8.1 Detailed Register Descriptions.................................................................1320 28.8.1.1 MSIX Table Entry Lower Address--MSIXTADD [0:3][0:9] (BAR3: 0x0000 + 0x10*n [n=0...9]; R/W) .................................................................1320 28.8.1.2 MSIX Table Entry Upper Address--MSIXTUADD [0:3][0:9] (BAR3: 0x0004 + 0x10*n [n=0...9]; R/W) .................................................................1321 28.8.1.3 MSIX Table Entry Message--MSIXTMSG [0:3][0:9] (BAR3: 0x0008 + 0x10*n [n=0...9]; R/W) ............................................................................1321 28.8.1.4 MSIX Table Entry Vector Control--MSIXTVCTRL [0:3][0:9] (BAR3: 0x000C + 0x10*n [n=0...9]; R/W) .................................................................1322 28.8.1.5 MSIXPBA Bit Description--MSIXPBA [0:3] (BAR3: 0x2000; RO)............1322 28.8.1.6 MSIX PBA Clear--PBACL [0:3] (0x5B68; R/W1C)................................1323 28.9 Receive Registers ...........................................................................................1324 28.9.1 Detailed Register Descriptions.................................................................1324 28.9.1.1 Receive Control Register--RCTL [0:3] (0x0100; R/W) .........................1324 28.9.1.2 Split and Replication Receive Control--SRRCTL [0:3][0:7] (0xC00C + 0x40*n [n=0...7]; R/W) .................................................1327 28.9.1.3 Packet Split Receive Type--PSRTYPE [0:3][0:7] (0x5480 + 4*n [n=0...7]; R/W).......................................................1330 28.9.1.4 Replicated Packet Split Receive Type--RPLPSRTYPE [0:3] (0x54C0; R/W)..............................................................................1331 28.9.1.5 Receive Descriptor Base Address Low--RDBAL [0:3] [0:7] (0xC000 + 0x40*n [n=0...7]; R/W) .................................................1332 28.9.1.6 Receive Descriptor Base Address High--RDBAH [0:3] [0:7] (0xC004 + 0x40*n [n=0...7]; R/W) .................................................1332 28.9.1.7 Receive Descriptor Ring Length--RDLEN [0:3][0:7] (0xC008 + 0x40*n [n=0...7]; R/W) ............................................................................1333 28.9.1.8 Receive Descriptor Head--RDH [0:3][0:7] (0xC010 + 0x40*n [n=0...7]; RO) ...................................................1333 28.9.1.9 Receive Descriptor Tail--RDT [0:3][0:7] (0xC018 + 0x40*n [n=0...7]; R/W) .................................................1334 28.9.1.10 Receive Descriptor Control--RXDCTL [0:3][0:7] (0xC028 + 0x40*n [n=0...7]; R/W) .................................................1334 28.9.1.11 Receive Queue Drop Packet Count--RQDPC [0:3][0:7] (0xC030 + 0x40*n [n=0...7]; RC/W) ...............................................1336 28.9.1.12 Receive Checksum Control--RXCSUM [0:3] (0x5000; R/W) .................1337 28.9.1.13 Receive Long Packet Maximum Length--RLPML [0:3] (0x5004; R/W) ....1339 28.9.1.14 Receive Filter Control Register--RFCTL [0:3] (0x5008; R/W) ...............1339 28.9.1.15 Multicast Table Array--MTA [0:3][0:127] (0x5200 + 4*n [n=0...127]; R/W) ...................................................1340 28.9.1.16 Receive Address Low 0--RAL0 [0:3][0:15] (0x5400 + 8*n [n=0...15]; R/W).....................................................1341 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 39 28.9.1.17 Receive Address Low 1--RAL1 [0:3][0:7] (0x54E0 + 8*n [n=0...7]; R/W) ...................................................... 1342 28.9.1.18 Receive Address High 0--RAH0 [0:3][0:15] (0x5404 + 8*n [n=0...15]; R/W) .................................................... 1342 28.9.1.19 Receive Address High 1--RAH1 [0:3][0:7] (0x54E4 + 8*n [n=0...7]; R/W) ...................................................... 1344 28.9.1.20 VLAN Filter Table Array--VFTA [0:3][0:127] (0x5600 + 4*n [n=0...127]; R/W) .................................................. 1345 28.9.1.21 Multiple Receive Queues Command Register--MRQC [0:3] (0x5818; R/W) ............................................................................. 1346 28.9.1.22 RSS Random Key Register--RSSRK [0:3][0:9] (0x5C80 + 4*n [n=0...9]; R/W) ...................................................... 1348 28.9.1.23 Redirection Table--RETA [0:3][0:31] (0x5C00 + 4*n [n=0...31]; R/W) .................................................... 1348 28.10 Filtering Register ........................................................................................... 1350 28.10.1 Detailed Register Descriptions ................................................................ 1350 28.10.1.1 Immediate Interrupt RX--IMIR [0:3][0:7] (0x5A80 + 4*n [n=0...7]; R/W) ...................................................... 1350 28.10.1.2 Immediate Interrupt Rx Ext.--IMIREXT [0:3][0:7] (0x5AA0 + 4*n [n=0...7]; R/W)...................................................... 1351 28.10.1.3 2tuples Queue Filter--TTQF [0:3][0:7] (0x59E0 + 4*n[n=0..7];RW) .... 1352 28.10.1.4 Immediate Interrupt Rx VLAN Priority--IMIRVP [0:3] (0x5AC0; R/W) ... 1352 28.10.1.5 SYN Packet Queue Filter--SYNQF [0:3] (0x55FC; RW)........................ 1353 28.10.1.6 EType Queue Filter--ETQF [0:3][0:7] (0x5CB0 + 4*n[n=0...7]; RW) ... 1354 28.11 Transmit Registers ......................................................................................... 1355 28.11.1 Detailed Register Descriptions ................................................................ 1355 28.11.1.1 Transmit Control Register--TCTL [0:3] (0x0400; R/W) ....................... 1355 28.11.1.2 Transmit Control Extended--TCTL_EXT [0:3] (0x0404; R/W) .............. 1356 28.11.1.3 Transmit IPG Register--TIPG (0x0410; R/W)..................................... 1357 28.11.1.4 Retry Buffer Control--RETX_CTL [0:3] (0x041C; RW)......................... 1358 28.11.1.5 DMA Tx Control--DTXCTL [0:3] (0x3590; R/W) ................................. 1358 28.11.1.6 DMA TX TCP Flags Control Low--DTXTCPFLGL [0:3] (0x359C; RW) ...... 1359 28.11.1.7 DMA TX TCP Flags Control High--DTXTCPFLGH [0:3] (0x35A0; RW) ..... 1360 28.11.1.8 DMA TX Max Total Allow Size Requests--DTXMXSZRQ [0:3] (0x3540; RW)............................................................................... 1360 28.11.1.9 DMA TX Maximum Packet Size--DTXMXPKTSZ [0:3] (0x355C; RW)...... 1361 28.11.1.10 Transmit Descriptor Base Address Low--TDBAL [0:3][0:7] (0xE000 + 0x40*n [n=0...7]; R/W) ................................................. 1361 28.11.1.11 Transmit Descriptor Base Address High--TDBAH [0:3][0:7] (0xE004 + 0x40*n [n=0...7]; R/W) ................................................. 1362 28.11.1.12 Transmit Descriptor Ring Length--TDLEN [0:3][0:7] (0xE008 + 0x40*n [n=0...7]; R/W) ................................................. 1362 28.11.1.13 Transmit Descriptor Head--TDH [0:3][0:7] (0xE010 + 0x40*n [n=0...7]; RO)................................................... 1363 28.11.1.14 Transmit Descriptor Tail--TDT [0:3][0:7] (0xE018 + 0x40*n [n=0...7]; R/W) ................................................. 1363 28.11.1.15 Transmit Descriptor Control--TXDCTL [0:3][0:7] (0xE028 + 0x40*n [n=0...7]; R/W) ................................................. 1364 28.11.1.16 Tx Descriptor Completion Write-Back Address Low--TDWBAL [0:3][0:7] (0xE038 + 0x40*n [n=0...7]; R/W) ................................................. 1366 28.11.1.17 Tx Descriptor Completion Write-Back Address High - TDWBAH [0:3][0:7] (0xE03C + 0x40*n [n=0...7];R/W).................................................. 1366 28.12 DCA and TPH Registers................................................................................... 1367 28.12.1 Detailed Register Descriptions ................................................................ 1367 28.12.1.1 Rx DCA Control Registers--RXCTL [0:3][0:7] (0xC014 + 0x40*n [n=0...7]; R/W)................................................. 1367 28.12.1.2 Tx DCA Control Registers--TXCTL [0:3][0:7] (0xE014 + 0x40*n [n=0...7]; R/W) ................................................. 1369 28.12.1.3 DCA Requester ID Information--DCA_ID [0:3] (0x5B70; RO) .............. 1370 Intel(R) Communications Chipset 89xx Series - Datasheet 40 October 2012 Order Number: 327879-001US Contents 28.12.1.4 DCA Control--DCA_CTRL [0:3] (0x5B74; R/W) ..................................1370 28.13 Virtualization Registers ...................................................................................1371 28.13.1 Detailed Register Descriptions.................................................................1371 28.13.1.1 VMDq Control Register--VT_CTL [0:3] (0x581C; R/W) ........................1371 28.13.1.2 Malicious Driver Free Block--MDFB [0:3] (0x3558; RWS) ....................1372 28.13.1.3 Last VM Misbehavior Cause--LVMMC [0:3] (0x3548; RC) ....................1372 28.13.1.4 VM Offload Register--VMOLR [0:3][0:7] (0x5AD0 + 4*n [n=0...7]; RW) .......................................................1373 28.13.1.5 Replication Offload Register--RPLOLR [0:3] (0x5AF0; RW) ..................1374 28.13.1.6 VLAN VM Filter--VLVF [0:3][0:31] (0x5D00 + 4*n [n=0...31]; RW) .....1375 28.13.1.7 Unicast Table Array--UTA [0:3][0:127] (0xA000 + 4*n [n=0...127]; R/W) ...................................................1376 28.13.1.8 Storm Control Control Register--SCCRL [0:3] (0x5DB0; RW) ...............1376 28.13.1.9 Storm Control Status--SCSTS [0:3] (0x5DB4;RO)..............................1377 28.13.1.10 Broadcast Storm Control Threshold--BSCTRH [0:3] (0x5DB8;RW)........1378 28.13.1.11 Multicast Storm Control Threshold--MSCTRH [0:3] (0x5DBC; RW)........1378 28.13.1.12 Broadcast Storm Control Current Count - BSCCNT [0:3] (0x5DC0;RO) ................................................................................1378 28.13.1.13 Multicast Storm Control Current Count--MSCCNT [0:3] (0x5DC4;RO) ................................................................................1379 28.13.1.14 Storm Control Time Counter--SCTC [0:3] (0x5DC8; RO).....................1379 28.13.1.15 Storm Control Basic Interval--SCBI [0:3] (0x5DCC; RW) ....................1380 28.13.1.16 Virtual Mirror Rule Control--VMRCTL [0:3][0:3] (0x5D80 + 0x4*n [n= 0...3]; RW) ...................................................1380 28.13.1.17 Virtual Mirror Rule VLAN--VMRVLAN [0:3][0:3] (0x5D90 + 0x4*n [n= 0...3]; RW) ...................................................1381 28.13.1.18 Virtual Mirror Rule VM--VMRVM [0:3][0:3] (0x5DA0 + 0x4*n [n= 0...3]; RW) ...................................................1381 28.14 Power Management Registers ..........................................................................1382 28.14.1 Detailed Register Descriptions.................................................................1382 28.14.1.1 DMA Receive Power Saving Register--DMARPS [0:3] (0x2500; R/W) ....1382 28.14.1.2 DMA Transmit Power Saving Register--DMATPS [0:3] (0x2504; R/W) ...1382 28.15 Timer Registers..............................................................................................1384 28.15.1 Detailed Register Descriptions.................................................................1384 28.15.1.1 Watchdog Setup--WDSTP [0:3] (0x1040; R/W) .................................1384 28.15.1.2 Watchdog Software Device Status--WDSWSTS [0:3] (0x1044; R/W) ....1385 28.15.1.3 Free Running Timer--FRTIMER [0:3] (0x1048; RWS) ..........................1385 28.15.1.4 TCP Timer--TCPTIMER [0:3] (0x104C; R/W)......................................1386 28.16 Time Sync Register Descriptions.......................................................................1387 28.16.1 Detailed Register Descriptions.................................................................1387 28.16.1.1 RX Time Sync Control Register--TSYNCRXCTL [0:3] (0xB620;RW) .......1387 28.16.1.2 RX timestamp Low--RXSTMPL [0:3] (0xB624; RO) .............................1388 28.16.1.3 RX Timestamp High--RXSTMPH [0:3] (0xB628; RO) ...........................1388 28.16.1.4 RX Timestamp Attributes Low--RXSATRL[0:3] (0xB62C; RO)...............1388 28.16.1.5 RX timestamp Attributes High--RXSATRH [0:3] (0xB630; RO) .............1389 28.16.1.6 TX Time Sync Control Register--TSYNCTXCTL [0:3] (0xB614; RW) .......1389 28.16.1.7 TX Timestamp Value Low--TXSTMPL [0:3] (0xB618;RO) .....................1390 28.16.1.8 TX Timestamp Value High--TXSTMPH[0:3] (0xB61C; RO)....................1390 28.16.1.9 System Time Register Residue--SYSTIMR [0:3] (0xB6F8; RW).............1390 28.16.1.10 System Time Register Low--SYSTIML [0:3] (0xB600; RW) ..................1391 28.16.1.11 System Time Register High--SYSTIMH [0:3] (0xB604; RW) .................1391 28.16.1.12 Increment Attributes Register--TIMINCA [0:3] (0xB608; RW)..............1392 28.16.1.13 Time Adjustment Offset Register Low--TIMADJL [0:3] (0xB60C; RW) ...1392 28.16.1.14 Time Adjustment Offset Register High--TIMADJH [0:3] (0xB610;RW) ...1392 28.16.1.15 TimeSync Auxiliary Control Register--TSAUXC [0:3] (0xB640; RW) ......1393 28.16.1.16 Target Time Register 0 Low--TRGTTIML0 [0:3] (0xB644; RW) .............1395 28.16.1.17 Target Time Register 0 High--TRGTTIMH0 [0:3] (0xB648; RW)............1395 28.16.1.18 Target Time Register 1 Low--TRGTTIML1 [0:3] (0xB64C; RW) .............1395 28.16.1.19 Target Time Register 1 High--TRGTTIMH1 [0:3] (0xB650; RW)............1396 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 41 28.16.1.20 Frequency Out 0 Control Register--FREQOUT0 [0:3] (0xB654; RW) ..... 1396 28.16.1.21 Frequency Out 1 Control Register--FREQOUT1 [0:3] (0xB658; RW) ..... 1397 28.16.1.22 Auxiliary Time Stamp 0 Register Low--AUXSTMPL0 [0:3] (0xB65C; RO) ............................................................................... 1397 28.16.1.23 Auxiliary Time Stamp 0 Register High--AUXSTMPH0 [0:3] (0xB660; RO) ............................................................................... 1398 28.16.1.24 Auxiliary Time Stamp 1 Register Low--AUXSTMPL1 [0:3] (0xB664; RO) ............................................................................... 1398 28.16.1.25 Auxiliary Time Stamp 1 Register High--AUXSTMPH1 [0:3] (0xB668; RO) ............................................................................... 1399 28.16.1.26 Time Sync RX Configuration--TSYNCRXCFG [0:3] (0x5F50; R/W) ........ 1399 28.16.1.27 Time Sync SDP Configuration Register--TSSDP [0:3] (0x003C; R/W) ... 1400 28.16.2 Time Sync Interrupt Registers ................................................................ 1401 28.16.2.1 Time Sync Interrupt Cause Register--TSICR [0:3] (0xB66C; RC/W1C) . 1401 28.16.2.2 Time Sync Interrupt Mask Register--TSIM [0:3] (0xB674; RW) ........... 1402 28.16.2.3 Time Sync Interrupt Set Register--TSIS [0:3] (0xB670; WO) .............. 1403 28.17 PCS Registers................................................................................................ 1404 28.17.1 Detailed Register Descriptions ................................................................ 1404 28.17.1.1 PCS Configuration--PCS_CFG [0:3] (0x4200; R/W)............................ 1404 28.17.1.2 PCS Link Control--PCS_LCTL [0:3] (0x4208; RW).............................. 1404 28.17.1.3 PCS Link Status--PCS_LSTS [0:3] (0x420C; RO) ............................... 1406 28.17.1.4 AN Advertisement--PCS_ANADV [0:3] (0x4218; R/W) ....................... 1408 28.17.1.5 Link Partner Ability--PCS_LPAB [0:3] (0x421C; RO)........................... 1409 28.17.1.6 Next Page Transmit--PCS_NPTX [0:3] (0x4220; RW) ......................... 1410 28.17.1.7 Link Partner Ability Next Page--PCS_LPABNP [0:3] (0x4224; RO) ........ 1411 28.17.1.8 SFP I2C Command--I2CCMD [0:3] (0x1028; R/W) ............................ 1411 28.17.1.9 SFP I2C Parameters--I2CPARAMS [0:3] (0x102C; R/W) ..................... 1412 28.18 Statistics Register Descriptions ........................................................................ 1413 28.18.1 Detailed Register Descriptions ................................................................ 1413 28.18.1.1 CRC Error Count--CRCERRS [0:3] (0x4000; RC)................................ 1413 28.18.1.2 Alignment Error Count--ALGNERRC [0:3] (0x4004; RC) ..................... 1414 28.18.1.3 Symbol Error Count--SYMERRS [0:3] (0x4008; RC) ........................... 1414 28.18.1.4 Missed Packets Count--MPC [0:3] (0x4010; RC)................................ 1415 28.18.1.5 Single Collision Count--SCC [0:3] (0x4014; RC)................................ 1415 28.18.1.6 Excessive Collisions Count--ECOL [0:3] (0x4018; RC) ........................ 1416 28.18.1.7 Multiple Collision Count--MCC [0:3] (0x401C; RC) ............................. 1416 28.18.1.8 Late Collisions Count--LATECOL [0:3] (0x4020; RC) .......................... 1417 28.18.1.9 Collision Count--COLC [0:3] (0x4028; RC) ....................................... 1417 28.18.1.10 Defer Count--DC [0:3] (0x4030; RC)............................................... 1418 28.18.1.11 Host Transmit Discarded Packets by MAC Count--HTDPMC [0:3] (0x403C; RC) ............................................................................... 1418 28.18.1.12 Receive Length Error Count--RLEC [0:3] (0x4040; RC) ...................... 1419 28.18.1.13 XON Received Count--XONRXC [0:3] (0x4048; RC) ........................... 1419 28.18.1.14 XON Transmitted Count--XONTXC [0:3] (0x404C; RC) ....................... 1420 28.18.1.15 XOFF Received Count--XOFFRXC [0:3] (0x4050; RC) ......................... 1420 28.18.1.16 XOFF Transmitted Count--XOFFTXC [0:3] (0x4054; RC) ..................... 1421 28.18.1.17 FC Received Unsupported Count--FCRUC [0:3] (0x4058; RC).............. 1421 28.18.1.18 Packets Received [64 Bytes] Count--PRC64 [0:3] (0x405C; RC).......... 1422 28.18.1.19 Packets Received [65-127 Bytes] Count--PRC127 [0:3] (0x4060; RC) ............................................................................... 1422 28.18.1.20 Packets Received [128-255 Bytes] Count--PRC255 [0:3] (0x4064; RC) ............................................................................... 1423 28.18.1.21 Packets Received [256-511 Bytes] Count--PRC511 [0:3] (0x4068; RC) ............................................................................... 1424 28.18.1.22 Packets Received [512-1023 Bytes] Count--PRC1023 [0:3] (0x406C; RC) ............................................................................... 1424 28.18.1.23 Packets Received [1024 to Max Bytes] Count--PRC1522 [0:3] (0x4070; RC) ............................................................................... 1425 28.18.1.24 Good Packets Received Count--GPRC [0:3] (0x4074; RC)................... 1425 Intel(R) Communications Chipset 89xx Series - Datasheet 42 October 2012 Order Number: 327879-001US Contents 28.18.1.25 28.18.1.26 28.18.1.27 28.18.1.28 28.18.1.29 28.18.1.30 28.18.1.31 28.18.1.32 28.18.1.33 28.18.1.34 28.18.1.35 28.18.1.36 28.18.1.37 28.18.1.38 28.18.1.39 28.18.1.40 28.18.1.41 28.18.1.42 28.18.1.43 28.18.1.44 28.18.1.45 28.18.1.46 28.18.1.47 28.18.1.48 28.18.1.49 28.18.1.50 28.18.1.51 28.18.1.52 28.18.1.53 28.18.1.54 28.18.1.55 28.18.1.56 28.18.1.57 28.18.1.58 28.18.1.59 28.18.1.60 28.18.1.61 28.18.1.62 28.18.1.63 28.18.1.64 28.18.1.65 28.18.1.66 28.18.1.67 28.18.1.68 28.18.1.69 28.18.1.70 28.18.1.71 28.18.1.72 October 2012 Order Number: 327879-001US Broadcast Packets Received Count--BPRC [0:3] (0x4078; RC) .............1426 Multicast Packets Received Count--MPRC [0:3] (0x407C; RC) ..............1426 Good Packets Transmitted Count--GPTC [0:3] (0x4080; RC) ...............1427 Good Octets Received Count--GORCL [0:3] (0x4088; RC) ...................1427 Good Octets Received Count--GORCH [0:3] (0x408C; RC) ..................1428 Good Octets Transmitted Count--GOTCL [0:3] (0x4090; RC) ...............1428 Good Octets Transmitted Count--GOTCH [0:3] (4094; RC)..................1429 Receive No Buffers Count--RNBC [0:3] (0x40A0; RC) .........................1429 Receive Undersize Count--RUC [0:3] (0x40A4; RC)............................1430 Receive Fragment Count--RFC [0:3] (0x40A8; RC) ............................1430 Receive Oversize Count--ROC [0:3] (0x40AC; RC) .............................1431 Receive Jabber Count--RJC [0:3] (0x40B0; RC) .................................1431 Management Packets Received Count--MNGPRC [0:3] (0x40B4; RC) ....1432 Management Packets Dropped Count--MPDC [0:3] (0x40B8; RC).........1432 Management Packets Transmitted Count--MNGPTC [0:3] (0x40BC; RC)................................................................................1433 Total Octets Received--TORL [0:3] (0x40C0; RC)...............................1433 Total Octets Received--TORH [0:3] (0x40C4; RC) ..............................1434 Total Octets Transmitted--TOTL [0:3] (0x40C8; RC)...........................1434 Total Octets Transmitted--TOTH [0:3] (0x40CC; RC) ..........................1435 Total Packets Received--TPR [0:3] (0x40D0; RC)...............................1435 Total Packets Transmitted--TPT [0:3] (0x40D4; RC)...........................1436 Packets Transmitted [64 Bytes] Count--PTC64 [0:3] (0x40D8; RC) ......1436 Packets Transmitted [65-127 Bytes] Count--PTC127 [0:3] (0x40DC; RC) ...............................................................................1437 Packets Transmitted [128-255 Bytes] Count--PTC255 [0:3] (0x40E0; RC) ................................................................................1437 Packets Transmitted [256-511 Bytes] Count--PTC511 [0:3] (0x40E4; RC) ................................................................................1438 Packets Transmitted [512-1023 Bytes] Count--PTC1023 [0:3] (0x40E8; RC) ................................................................................1438 Packets Transmitted [1024 Bytes or Greater] Count--PTC1522 [0:3] (0x40EC; RC)................................................................................1439 Multicast Packets Transmitted Count--MPTC [0:3] (0x40F0; RC) ..........1439 Broadcast Packets Transmitted Count--BPTC [0:3] (0x40F4; RC) .........1440 TCP Segmentation Context Transmitted Count--TSCTC [0:3] (0x40F8; RC) ................................................................................1440 Interrupt Assertion Count--IAC [0:3] (0x4100; RC)............................1441 Rx Packets to Host Count--RPTHC [0:3] (0x4104; RC)........................1441 Debug Counter 1--DBGC1 [0:3] (0x4108; RC)...................................1441 Debug Counter 2--DBGC2 [0:3] (0x410C; RC) ..................................1442 Debug Counter 3--DBGC3 [0:3] (0x4110; RC)...................................1443 Debug Counter 4--DBGC4 [0:3] (0x411C; RC) ..................................1443 Host Good Packets Transmitted Count--HGPTC [0:3] (0x4118; RC) ......1444 Receive Descriptor Minimum Threshold Count--RXDMTC [0:3] (0x4120; RC) ................................................................................1444 Host Good Octets Received Count--HGORCL [0:3] (0x4128; RC)..........1445 Host Good Octets Received Count--HGORCH [0:3] (0x412C; RC) .........1445 Host Good Octets Transmitted Count--HGOTCL [0:3] (0x4130; RC)......1446 Host Good Octets Transmitted Count - HGOTCH [0:3] (0x4134; RC).....1446 Length Error Count--LENERRS [0:3] (0x4138; RC) .............................1447 SerDes/SGMII/KX Code Violation Packet Count--SCVPC [0:3] (0x4228; RW) ...............................................................................1447 Switch Drop Packet Count--SDPC [0:3] (0x41A4; RC) ........................1448 Virtualization Statistical Counters .....................................................1448 Per Queue Good Packets Received Count--VFGPRC [0:3][0:7] (0x10010 + n*0x100 [n=0...7]; RO)................................................1448 Per Queue Good Packets Transmitted Count--VFGPTC [0:3][0:7] (0x10014 + n*0x100 [n=0...7]; RO)................................................1449 Intel(R) Communications Chipset 89xx Series - Datasheet 43 28.18.1.73 Per Queue Good Octets Received Count--VFGORC [0:3][0:7] (0x10018 + n*0x100 [n=0...7]; RO) ............................................... 1450 28.18.1.74 Per Queue Good Octets Transmitted Count--VFGOTC [0:3][0:7] (0x10034 + n*0x100 [n=0...7]; RO) ............................................... 1451 28.18.1.75 Per Queue Multicast Packets Received Count--VFMPRC [0:3][0:7] (0x10038 + n*0x100 [n=0...7]; RO) ............................................... 1451 28.19 Manageability Statistics .................................................................................. 1452 28.19.1 Detailed Register Descriptions ................................................................ 1452 28.19.1.1 BMC Management Packets Dropped Count--BMPDC [0:3] (0x4140; RC) ............................................................................... 1452 28.19.1.2 BMC Management Packets Transmitted Count--BMNGPTC [0:3] (0x4144; RC) ............................................................................... 1452 28.19.1.3 BMC Management Packets Received Count--BMNGPRC [0:3] (0x413C; RC) ............................................................................... 1453 28.19.1.4 BMC Total Unicast Packets Received--BUPRC [0:3] (0x4400; RC) ........ 1453 28.19.1.5 BMC Total Multicast Packets Received--BMPRC [0:3] (0x4404; RC) ...... 1454 28.19.1.6 BMC Total Broadcast Packets Received--BBPRC [0:3] (0x4408; RC) ..... 1454 28.19.1.7 BMC Total Unicast Packets Transmitted--BUPTC [0:3] (0x440C; RC) .... 1454 28.19.1.8 BMC Total Multicast Packets Transmitted--BMPTC [0:3] (0x4410; RC).. 1455 28.19.1.9 BMC Total Broadcast Packets Transmitted--BBPTC [0:3] (0x4414; RC) ............................................................................... 1455 28.19.1.10 BMC FCS Receive Errors--BCRCERRS [0:3] (0x4418; RC) ................... 1456 28.19.1.11 BMC Alignment Errors--BALGNERRC [0:3] (0x441C; RC) .................... 1456 28.19.1.12 BMC Pause XON Frames Received--BXONRXC [0:3] (0x4420; RC) ....... 1456 28.19.1.13 BMC Pause XOFF Frames Received--BXOFFRXC [0:3] (0x4424; RC) ..... 1457 28.19.1.14 BMC Pause XON Frames Transmitted--BXONTXC [0:3] (0x4428; RC) ... 1457 28.19.1.15 BMC Pause XOFF Frames Transmitted--BXOFFTXC [0:3] (0x442C; RC) ............................................................................... 1458 28.19.1.16 BMC Single Collision Transmit Frames--BSCC [0:3] (0x4430; RC)........ 1458 28.19.1.17 BMC Multiple Collision Transmit Frames--BMCC [0:3] (0x4434; RC) ..... 1458 28.20 Wake Up Controls Registers ............................................................................ 1459 28.20.1 Detailed Register Descriptions ................................................................ 1459 28.20.1.1 Wakeup Control Register--WUC [0:3] (0x5800; R/W) ........................ 1459 28.20.1.2 Wakeup Filter Control Register--WUFC [0:3] (0x5808; R/W) ............... 1460 28.20.1.3 Wakeup Status Register--WUS [0:3] (0x5810; R/W1C) ...................... 1461 28.20.1.4 Wakeup Packet Length--WUPL [0:3] (0x5900; RO)............................ 1462 28.20.1.5 Wakeup Packet Memory--WUPM [0:3][0:31] (0x5A00 + 4*n [n=0...31]; RO) ...................................................... 1462 28.20.1.6 IP Address Valid--IPAV [0:3] (0x5838; R/W) .................................... 1463 28.20.1.7 IPv4 Address Table--IP4AT [0:3][0:3] (0x5840 + 8*n [n=0...3]; R/W) ...................................................... 1463 28.20.1.8 IPv6 Address Table--IP6AT [0:3][0:3] (0x5880 + 4*n [n=0...3]; R/W) ...................................................... 1464 28.20.2 Flexible Host Filter Table Registers--FHFT [0:3] (0x9000 - 0x93FC; RW)...... 1464 28.20.2.5 Flex Filter 0--Example ................................................................... 1468 28.20.3 Flexible Host Filter Table Extended Registers--FHFT_EXT [0:3] (0x9A00 - 0x9DFC; RW) ........................................................................ 1468 28.21 Management Registers ................................................................................... 1471 28.21.1 Detailed Register Descriptions ................................................................ 1471 28.21.1.1 Management VLAN TAG Value--MAVTV [0:3][0:7] (0x5010 +4*n [n=0...7]; RW) ........................................................ 1471 28.21.1.2 Management Flex UDP/TCP Ports--MFUTP[0:3][0:3] (0x5030 + 4*n [n=0...3]; RW) ............................................................................. 1471 28.21.1.3 Management Ethernet Type Filters--METF [0:3][0:3] (0x5060 + 4*n [n=0...3]; RW) ............................................................................. 1472 28.21.1.4 Management Control Register--MANC [0:3] (0x5820; RW) ................. 1472 28.21.1.5 Management Only Traffic Register--MNGONLY [0:3] (0x5864; RW)...... 1474 28.21.1.6 Manageability Decision Filters--MDEF [0:3][0:7] (0x5890 + 4*n [n=0...7]; RW) ....................................................... 1474 Intel(R) Communications Chipset 89xx Series - Datasheet 44 October 2012 Order Number: 327879-001US Contents 28.21.1.7 Manageability Decision Filters--MDEF_EXT [0:3][0:7] (0x5930 + 4*n[n=0...7]; RW).........................................................1476 28.21.1.8 Manageability IP Address Filter--MIPAF [0:3][0:15] (0x58B0 + 4*n [n=0...15]; RW) ......................................................1476 28.21.1.9 Manageability MAC Address Low--MMAL [0:3] (0x5910 + 8*n [n= 0...1]; RW) .......................................................1479 28.21.1.10 Manageability MAC Address High--MMAH [0:3][0:1] (0x5914 + 8*n [n=0...1]; RW)........................................................1480 28.21.1.11 Flexible TCO Filter Table Registers--FTFT [0:3] (0x9400-0x94FC; RW) ...................................................................1480 28.22 Memory Error Registers Description ..................................................................1482 28.22.1 Parity and ECC Error Indication--PEIND [0:3] (0x1084; RC)........................1482 28.22.2 Parity and ECC Indication Mask--PEINDM [0:3] (0x1088; RW) ....................1483 28.22.3 DMA Transmit Descriptor Parity Status--DTPARS [0:3] (0x3510; RW1C) ......1483 28.22.4 DMA Receive Descriptor Parity Status--DRPARS [0:3] (0x3514; RW1C)........1484 28.22.5 Dhost Parity Status--DDPARS [0:3] (0x3518; RW1C).................................1484 28.22.6 Tx Packet Buffer ECC Status--TPBECCSTS [0:3] (0x345C; RW) ...................1485 28.22.7 LAN Port Parity Error Control Register--LANPERRCTL [0:3] (0x5F54; RW) .....1485 28.22.8 LAN Port Parity Error Status Register--LANPERRSTS [0:3] (0x5F58; RO) ......1486 29.0 Function 1-4 (GbE) ...............................................................................................1487 29.1 Introduction ..................................................................................................1487 29.2 Detailed Register Summary .............................................................................1490 29.2.1 PCI Views.............................................................................................1490 29.3 Mandatory PCI Configuration Registers..............................................................1493 29.3.1 Detailed Register Descriptions.................................................................1493 29.3.1.1 PVID[0:3]--PF Vendor Identification Register.....................................1493 29.3.1.2 PDID0--PF Device Identification Register (GbE0)................................1494 29.3.1.3 PDID1--PF Device Identification Register (GbE1)................................1494 29.3.1.4 PDID2--PF Device Identification Register (GbE2)................................1495 29.3.1.5 PDID3--PF Device Identification Register (GbE3)................................1495 29.3.1.6 PPCICMD[0:3]--PF Device Command Register ...................................1496 29.3.1.7 PPCISTS[0:3]--PF Device Status Register .........................................1497 29.3.1.8 PRID[0:3]--PF Revision ID Register..................................................1499 29.3.1.9 PCC[0:3]--PF Class Code Register....................................................1499 29.3.1.10 PHDR[0:3]--PF Header Type Register ...............................................1500 29.3.2 Base Address Registers (0x10...0x27; R/W)..............................................1500 29.3.2.1 Base Address Register Fields ...........................................................1500 29.3.2.2 GbEPCIBAR0[0:3]--BAR0 Base Address Register ................................1501 29.3.2.3 GbEPCIBAR1[0:3]--BAR1 Base Address Register ................................1502 29.3.2.4 GbEPCIBAR2[0:3]--BAR2 Base Address Register ................................1502 29.3.2.5 GbEPCIBAR3[0:3]--BAR3 Base Address Register ................................1503 29.3.2.6 GbEPCIBAR4[3:0]--BAR4 Base Address Register ................................1504 29.3.2.7 GbEPCIBAR5[0:3]--BAR5 Base Address Register ................................1505 29.3.2.8 PSVID[0:3]--PF Subsystem Vendor ID Register .................................1505 29.3.2.9 PSID[0:3]--PF Subsystem ID Register ..............................................1506 29.3.2.10 Expansion ROM Base Address (0x30; RO) (Flash Not Implemented)......1506 29.3.2.11 PCP[0:3]--PF Capabilities Pointer Register.........................................1506 29.3.2.12 PIRQL[0:3]--PF Interrupt Line Register .............................................1507 29.3.2.13 PIRQP0--PF Interrupt Pin Register 0 .................................................1507 29.3.2.14 PIRQP1--PF Interrupt Pin Register 1 .................................................1508 29.3.2.15 PIRQP2--PF Interrupt Pin Register 2 .................................................1508 29.3.2.16 PIRQP3--PF Interrupt Pin Register 3 .................................................1509 29.4 PCI Capabilities ..............................................................................................1509 29.4.1 Power Management Capability Structure ..................................................1509 29.4.1.1 PPMCAP[0:3]--PF Power Management Capabilities ID Register .............1510 29.4.1.2 PPMCP[0:3]--PF Power Management Next Capability Pointer Register ...1510 29.4.1.3 PPMC[0:3]--PF Power Management Capabilities Register .....................1511 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 45 29.4.1.4 29.4.2 MSI 29.4.2.1 29.4.2.2 PPMCSR--PF Power Management Control and Status Register.............. 1512 Capability Structure........................................................................ 1513 PMSICID[0:3]--Message Signalled Interrupt Capability ID Register ...... 1513 PMSINCP[0:3]--Message Signalled Interrupt Next Capability Pointer Register ............................................................................ 1513 29.4.2.3 PMSICTL[0:3]--Message Signalled Interrupt Control Register .............. 1514 29.4.2.4 PMSILADDR[0:3]--Message Signalled Interrupt Lower Address Register ....................................................................................... 1514 29.4.2.5 PMSIUADDR[0:3]--Message Signalled Interrupt Upper Address Register ....................................................................................... 1515 29.4.2.6 PMSIDATA[0:3]--Message Signalled Interrupt Data Register ............... 1515 29.4.2.7 PMSIMSK[0:3]--Message Signalled Interrupt Mask Register ................ 1515 29.4.2.8 PMSIPND[0:3]--Message Signalled Interrupt Mask Pending Register .... 1516 29.4.3 MSI-X Configuration.............................................................................. 1516 29.4.3.1 PMSI-X[0:3]--PF Message Signalled Interrupt X Capability ID Register . 1517 29.4.3.2 PMSIXNCP[0:3]--PF MSIX Next Capability Pointer Register ................. 1518 29.4.3.3 PMSIXCNTL[0:3]--PF Message Signalled Interrupt X Control Register ... 1518 29.4.3.4 PMSIXTBIR[0:3]--PF MSI-X Table Offset & Table BIR Register ............. 1519 29.4.3.5 PMSIXPBABIR[0:3]--PF MSI-X Pending Bit Array & BIR Offset Register ....................................................................................... 1519 29.4.4 CSR Access Via Configuration Address Space............................................ 1520 29.4.4.1 IOADDR[0:3]--IOADDR Register ..................................................... 1520 29.4.4.2 IODATA[0:3]--IODATA Register ...................................................... 1520 29.4.5 PCIe Configuration Registers .................................................................. 1521 29.4.6 PCI Express Capability Structure............................................................. 1521 29.4.6.1 PPCID[0:3]--PF PCI Express Capability ID Register ............................ 1521 29.4.6.2 PPCP[0:3]--PF PCI Express Next Capability Pointer Register ................ 1522 29.4.6.3 PPCR[0:3]--PF PCI Express Capabilities Register ............................... 1522 29.4.6.4 PPDCAP[0:3]--PF PCI Express Device Capabilities Register ................. 1523 29.4.6.5 PPDCNTL[0:3]--PF PCI Express Device Control Register...................... 1524 29.4.6.6 PPDSTAT[0:3]--PF PCI Express Device Status Register....................... 1525 29.4.6.7 PLCAPR[0:3]--PF Link Capabilities Register ....................................... 1526 29.4.6.8 PLCNTLR[0:3]--PF Link Control Register ........................................... 1528 29.4.6.9 PLSR[0:3]--PF Link Status Register ................................................. 1529 29.4.6.10 PDCAPR2[0:3]--PF Device Capabilities 2 Register .............................. 1531 29.4.6.11 PDCNTR2[0:3]--PF Device Control 2 Register.................................... 1532 29.4.6.12 PLCNTLR2[0:3]--PF Link Control 2 Register ...................................... 1533 29.4.6.13 PLSR2[0:3]--PF Link Status 2 Register ............................................. 1535 29.4.7 Vital Product Data (VPD) Registers.......................................................... 1535 29.4.7.1 VPDCID[0:3]--VPD Capability ID Register......................................... 1536 29.4.7.2 VPDNCP[0:3]--VPD Next Capability Pointer Register .......................... 1536 29.4.7.3 VPDADDR[0:3]--VPD Address Register............................................. 1537 29.4.7.4 VPDDATA[0:3]--VPD Data .............................................................. 1537 29.5 PCIe Extended Configuration Space.................................................................. 1538 29.5.1 Advanced Error Reporting (AER) Capability .............................................. 1538 29.5.2 PF Advanced Error Reporting Capability Structure ..................................... 1539 29.5.2.1 PPCIEAERCAPID[0:3]--PF PCI Express AER Capability ID Register........ 1539 29.5.2.2 PPAERUCS[0:3]--PF PCI Express AER Uncorrectable Error Status Register ....................................................................................... 1539 29.5.2.3 PPAERUCM[0:3]--PF PCI Express AER Uncorrectable Error Mask Register ....................................................................................... 1540 29.5.2.4 PPAERUCSEV[0:3]--PF PCI Express AER Uncorrectable Error Severity Register ....................................................................................... 1541 29.5.2.5 PPAERCS[0:3]--PF PCI Express AER Correctable Error Register............ 1542 29.5.2.6 PPAERCM[0:3]--PF PCI Express AER Correctable Error Mask Register ... 1543 29.5.2.7 PPAERCTLCAP[0:3]--PF PCI Express AER Control and Capability Register ....................................................................................... 1543 29.5.2.8 PPAERHDRLOG0[0:3]--PF PCI Express AER Header Log 0 Register ....... 1544 29.5.2.9 PPAERHDRLOG1[0:3]--PF PCI Express AER Header Log 1 Register ....... 1544 Intel(R) Communications Chipset 89xx Series - Datasheet 46 October 2012 Order Number: 327879-001US Contents 29.5.2.10 PPAERHDRLOG2[0:3]--PF PCI Express AER Header Log 2 Register .......1545 29.5.2.11 PPAERHDRLOG3[0:3]--PF PCI Express AER Header Log 3 Register .......1545 29.5.3 PF Alternative Routing-ID Extended Capability Structure.............................1546 29.5.3.1 PARIDHDR[0:3]--PF Alternative Routing ID Capability Header..............1546 29.5.3.2 PFARICAP0--PF ARI Capabilities Register ..........................................1547 29.5.3.3 PFARICAP1--PF ARI Capabilities Register ..........................................1548 29.5.3.4 PFARICAP2--PF ARI Capabilities Register ..........................................1549 29.5.3.5 PFARICAP3--PF ARI Capabilities Register ..........................................1549 29.5.3.6 PARIDCTL[0:3]--PF Alternative Routing ID Control Register.................1550 29.5.4 TLP Processing Hint Requester (TPH) Capability.........................................1550 29.5.4.1 TPH CAP ID (0x1A0; RO) ................................................................1550 29.5.4.2 TPH Requester Capabilities (0x1A4; RO) ...........................................1550 29.5.4.3 TPH Requester Control (0x1A8; R/W) ...............................................1551 29.5.4.4 TPH Steering table (0x1AC - 0x1B8; R/W).........................................1551 29.5.4.5 Latency Tolerance Requirement Reporting (LTR) Capability..................1551 29.5.4.6 LTR CAP ID (0x1C0; RO) ................................................................1551 29.5.4.7 LTR Capabilities (0x1C4; RW) ..........................................................1551 29.5.5 I/O Space ............................................................................................1551 Test Features - Volume 3 of 4 .................................................. 1552 30.0 PCH Global Test Features......................................................................................1553 30.1 JTAG ............................................................................................................1553 30.1.1 JTAG Functions Overview .......................................................................1553 30.1.2 PCH TAP Interfaces ...............................................................................1553 30.1.3 TAP Controller Operation and State Diagram .............................................1554 Technical Specifications - Volume 4 of 4 .................................. 1556 31.0 System Clocks ......................................................................................................1557 31.1 External Clock Requirements ...........................................................................1557 32.0 Signal Descriptions ...............................................................................................1558 32.1 Name Convention...........................................................................................1558 32.2 JTAG Boundary Scan Chain (BSC) 1149.1 and 1149.6 Chain ................................1560 32.3 Direct Media Interface.....................................................................................1560 32.4 PCI Express* EndPoint ....................................................................................1561 32.5 Gigabit Ethernet Interface ...............................................................................1563 32.6 EndPoint Management SMBus Interface (Slave)..................................................1570 32.7 PCI Express* Root Complex .............................................................................1570 32.8 Serial ATA Interface........................................................................................1571 32.9 LPC Interface.................................................................................................1573 32.10 USB Interface ................................................................................................1574 32.11 UART Interface ..............................................................................................1575 32.12 Host SMBus (Master) Interface.........................................................................1578 32.13 Serial Peripheral Interface Boot Interface ..........................................................1579 32.14 Interrupt Interface .........................................................................................1579 32.15 Processor Interface.........................................................................................1580 32.16 Power Management Interface...........................................................................1581 32.17 Thermal Sensor Current Reference ...................................................................1583 32.18 Miscellaneous Interface ...................................................................................1584 32.19 General Purpose Input/Output Interface ............................................................1584 32.20 Real Time Clock (RTC) Interface .......................................................................1592 32.21 System Input Clock ........................................................................................1594 32.22 JTAG Interface ...............................................................................................1595 32.23 Strapping Signals ...........................................................................................1597 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 47 32.24 Reserved Signals ........................................................................................... 1601 32.25 Power and Ground Signals .............................................................................. 1604 33.0 Electrical Characteristics...................................................................................... 1606 33.1 Absolute Maximum and Minimum Ratings.......................................................... 1606 33.2 Recommended Power Supply Range ................................................................. 1607 33.3 Maximum ICC Supply Current.......................................................................... 1608 33.4 Power Supply Pin Groupings ............................................................................ 1609 33.5 General DC Characteristics.............................................................................. 1611 33.6 PCI Express* Root Complex and DMI Gen1 DC/AC Characteristics ........................ 1616 33.7 PCIe* Gen2 EndPoint DC/AC Specification ......................................................... 1619 33.7.1 PCIe* Specification - Input Clock ............................................................ 1621 33.8 SATA Gen1/Gen2 DC/AC Characteristics ........................................................... 1622 33.9 Gigabit Ethernet SGMII DC/AC Characteristics ................................................... 1623 33.10 SerDes DC/AC Specification ............................................................................ 1624 33.11 CRU Clock DC/AC Specification ........................................................................ 1625 33.12 SATA, DMI, and PCIe* Clocks DC/AC Specification ............................................. 1627 33.13 AC Characteristics.......................................................................................... 1631 33.13.1 I2C/SFP AC Specification ....................................................................... 1638 33.13.2 MDIO DC Specification .......................................................................... 1639 33.13.3 MDIO AC Specification........................................................................... 1640 33.13.4 EEPROM AC Specification ....................................................................... 1641 33.13.5 UART AC Specification ........................................................................... 1643 33.13.6 JTAG AC Specification ........................................................................... 1644 33.14 AC Timing Diagrams....................................................................................... 1644 34.0 Thermal Specifications and Design Considerations............................................... 1650 35.0 Packaging Information ........................................................................................ 1651 35.1 Package Introduction ..................................................................................... 1651 35.2 Ball Map Information ...................................................................................... 1651 35.2.1 Ball Map Pin Lists.................................................................................. 1652 35.2.2 Ball Map Illustrations............................................................................. 1668 35.3 Package Mechanical Information ...................................................................... 1682 Figures 2-1 2-2 3-1 3-2 4-1 4-2 4-3 4-4 4-5 4-6 4-7 4-8 4-9 4-10 4-11 4-12 4-13 4-14 PCH Modes ........................................................................................................94 PCH Block Diagram .............................................................................................95 Device-Centric Logical View of PCH Devices ...........................................................98 Attaching PCH to the PCI Fabric (Logical Perspective) ............................................ 103 LPC Interface Diagram ...................................................................................... 112 DMA Controller ................................................................................................. 116 DMA Request Assertion through LDRQ#............................................................... 119 System Management Bus (SMBus) Interface ........................................................ 156 SMBus Block-Size Configuration Register Read ..................................................... 175 SMBus Block-Size Memory Register Read ............................................................. 176 SMBus Word-Size Configuration Register Read...................................................... 176 SMBus Word-Size Memory Register Read ............................................................. 177 SMBus Byte-Size Configuration Register Read....................................................... 177 SMBus Byte-Size Memory Register Read .............................................................. 178 SMBus Block-Size Configuration Register Write ..................................................... 179 SMBus Block-Size Memory Register Write ............................................................ 179 SMBus Word-Size Configuration Register Write ..................................................... 179 SMBus Word-Size Memory Register Write............................................................. 179 Intel(R) Communications Chipset 89xx Series - Datasheet 48 October 2012 Order Number: 327879-001US Contents 4-15 4-16 4-17 4-18 4-19 4-20 4-21 4-22 4-23 4-24 4-25 19-1 19-2 19-3 19-4 21-1 22-7 22-9 23-1 23-2 25-1 25-2 25-3 25-4 25-5 26-1 26-2 26-3 26-12 26-13 26-14 26-15 26-16 26-17 26-18 26-19 26-20 26-21 26-22 26-23 26-24 26-25 26-26 26-27 26-28 26-29 26-32 26-33 27-1 27-2 28-1 29-1 30-1 30-2 32-1 SMBus Configuration (Byte Write, PEC Enabled) ................................................... 180 SMBus Memory (Byte Write, PEC Enabled)........................................................... 180 Example UART Data Frame ................................................................................ 185 WDT Block Diagram.......................................................................................... 188 Serial Post over GPIO Reference Circuit ............................................................... 191 1-byte Serial Write with a Data Byte of 5Ah ......................................................... 193 Serial Data Transmitted Over the SGPIO Interface ................................................ 201 EHCI with USB 2.0 with Rate Matching Hub ......................................................... 214 MSI Packet Header ........................................................................................... 217 Flash Partition Boundary ................................................................................... 219 Flash Descriptor Sections .................................................................................. 220 PCIe* EP Interface Block Diagram ...................................................................... 872 EP Functional Description Block Diagram ............................................................. 873 EP Block Diagram ............................................................................................. 887 Ring Block Diagram .......................................................................................... 888 High-Level View of GbE Interface Connectivity in PCH Implementation .................... 991 SMBus ARP Flow .............................................................................................1011 GbE Controller Loopback Modes ........................................................................1035 Power-Up - General Flow..................................................................................1038 Power-Up Timing Diagram ................................................................................1039 Power Management State Diagram ....................................................................1094 Power Up (Off to Dup to D0u to D0a) .................................................................1099 Transition From D0a to D3 and Back Without PCIE_EP_RST# ................................1100 Transition From D0a to D3 and Back With PCIE_EP_RST#.....................................1101 Transition From D0a to Dr and Back Without Transition to D3 ...............................1102 Stages in Packet Filtering .................................................................................1111 Receive Queuing Flow (Virtualization) ................................................................1114 RSS Block Diagram..........................................................................................1118 Receive Filtering Flow Chart..............................................................................1124 Host MAC Address Receive Filtering Flow Chart ...................................................1125 VLAN Filtering.................................................................................................1127 Platform Manageability Filtering ........................................................................1128 Receive Descriptor Ring Structure .....................................................................1139 Header Splitting ..............................................................................................1141 Header Replication ..........................................................................................1142 Transmit Descriptor Ring Structure ....................................................................1160 Cause Mapping in Legacy Mode .........................................................................1177 Cause Mapping in MSI-X Mode ..........................................................................1178 Interrupt Throttle Flow Diagram ........................................................................1184 Case A: Heavy Load, Interrupts Moderated .........................................................1184 Light Load, Interrupts Immediately on Packet Receive..........................................1185 Packet Reception Decision Table........................................................................1190 Diagram of DCA Implementation on FSB System .................................................1193 PCIe Message Format for DCA...........................................................................1194 Pool List Selection - Replication Enabled .............................................................1199 Pool List Selection - Replication Disabled ............................................................1200 Sync Flow and Offset Calculation .......................................................................1207 Time Stamp Point............................................................................................1210 Controller to BMC Connectivity Through a SMBus-only Connection .........................1220 Manageability Decision Filters ...........................................................................1243 Multicast Table Array .......................................................................................1341 GbE Controller PCI Configuration Registers .........................................................1488 TAP Connectivity .............................................................................................1554 TAP Controller State Diagram ...........................................................................1555 Interface Signals Block Diagram ........................................................................1559 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 49 33-1 33-2 33-3 33-4 33-5 33-6 33-7 33-8 33-9 33-10 33-11 33-12 33-13 33-14 33-15 33-16 33-17 33-18 33-19 33-20 33-21 33-22 33-23 33-24 33-25 33-26 33-27 33-28 33-29 33-30 33-31 33-32 33-33 33-34 35-1 Tables 1-1 1-2 1-3 1-4 1-5 2-1 3-1 3-2 3-3 3-4 3-5 3-6 4-1 4-2 4-3 4-4 4-5 PCI Express Gen1 Transmitter Eye .................................................................... 1617 PCI Express Gen1 Receiver Eye ........................................................................ 1618 PCI Express Transmitter Eye ............................................................................ 1620 PCI Express Receiver Eye ................................................................................ 1621 CRU Differential Clock Waveform ...................................................................... 1626 CRU Differential Clock Cross-Point Specification .................................................. 1626 SATA, DMI, and PCIe* Clocks Differential Clock Waveform ................................... 1629 Single-Ended Measurement Point for Absolute Cross Point and Swing .................... 1629 Single-Ended Clock Measurement Points for Delta Cross Point............................... 1629 Differential Clock Cross Point Specification ......................................................... 1630 Differential Measurement Point for Duty Cycle and Period..................................... 1630 Differential Measurement Point for Rise and Fall Time .......................................... 1630 Differential Measurement Point for Ringback....................................................... 1631 Digital I/O Output Timing Diagram .................................................................... 1632 Digital I/O Input Timing Diagram ...................................................................... 1632 I2C I/F Timing Diagram ................................................................................... 1638 MDIO Input AC Timing Diagram ........................................................................ 1640 MDIO Output AC Timing Diagram...................................................................... 1641 EEPROM Timing Diagram ................................................................................. 1642 JTAG AC Timing Diagram ................................................................................. 1644 Clock Cycle Time ............................................................................................ 1644 Clock Timing .................................................................................................. 1645 Valid Delay from Rising Clock Edge ................................................................... 1645 Setup and Hold Times ..................................................................................... 1645 Float Delay .................................................................................................... 1645 Pulse Width ................................................................................................... 1646 Output Enable Delay ....................................................................................... 1646 USB Rise and Fall Times .................................................................................. 1646 USB Jitter ...................................................................................................... 1647 USB EOP Width .............................................................................................. 1647 SMBus Transaction ......................................................................................... 1647 SMBus Timeout .............................................................................................. 1648 SPI Timings ................................................................................................... 1648 Measurement Points for Differential Waveforms .................................................. 1649 Mechanical Package ........................................................................................ 1682 Intel(R) Communications Chipset 89xx Series SKUs...................................................83 Referenced Documents .......................................................................................84 Related Websites ................................................................................................84 Acronym Table ...................................................................................................85 Glossary Table ..................................................................................................87 PCH External Interface Summary .........................................................................96 Supported Operations by Memory Type ............................................................... 100 Address Space Sizes of PCIe* Endpoint-Attached Devices ...................................... 101 PCH PCI Device Summary .................................................................................. 104 EP PCI Device Summary .................................................................................... 104 GbE PCI Device ID Summary.............................................................................. 105 PCI Configuration Header Support for Type 0 Headers in PCIe* Endpoint Devices ..... 106 MSI vs. PCI IRQ Actions .................................................................................... 108 LPC Cycle Types Supported ................................................................................ 113 Start Field Bit Definitions ................................................................................... 113 Cycle Type Bit Definitions .................................................................................. 114 Transfer Size Bit Definition................................................................................. 114 Intel(R) Communications Chipset 89xx Series - Datasheet 50 October 2012 Order Number: 327879-001US Contents 4-6 4-7 4-8 4-9 4-10 4-11 4-12 4-13 4-14 4-15 4-16 4-17 4-18 4-19 4-20 4-21 4-22 4-23 4-24 4-25 4-26 4-27 4-28 4-29 4-30 4-31 4-32 4-33 4-34 4-35 4-36 4-37 4-38 4-39 4-40 4-41 4-42 4-43 4-44 4-45 4-46 4-47 4-48 4-49 4-50 4-51 4-52 4-53 4-54 4-55 4-56 4-57 4-58 4-59 4-60 4-61 4-62 4-63 5-1 SYNC Bit Definition ........................................................................................... 114 DMA Transfer Size ............................................................................................ 118 Address Shifting in 16-Bit I/O DMA Transfers ....................................................... 118 Counter Operating Modes .................................................................................. 123 Interrupt Controller Core Connections ................................................................. 125 Interrupt Status Registers ................................................................................. 126 Content of Interrupt Vector Byte ........................................................................ 126 APIC Interrupt Mapping1 ................................................................................... 131 Stop Frame Explanation .................................................................................... 134 Data Frame Format .......................................................................................... 134 Configuration Bits Reset by RTCRST# Assertion.................................................... 137 INIT# Going Active........................................................................................... 139 NMI Sources .................................................................................................... 139 General Power States for Platform Systems ......................................................... 141 State Transition Rules ....................................................................................... 141 System Power Plane ......................................................................................... 142 Causes of SMI and SCI...................................................................................... 143 Sleep Types..................................................................................................... 145 Causes of Wake Events ..................................................................................... 146 GPI Wake Events ............................................................................................. 146 Transitions Due to Power Failure ........................................................................ 147 Transitions Due to Power Button ........................................................................ 148 Transitions Due to RI# Signal ............................................................................ 149 Write Only Registers with Read Paths in ALT Access Mode...................................... 151 PIC Reserved Bits Return Values ........................................................................ 152 Register Write Accesses in ALT Access Mode ........................................................ 153 Causes of Host and Global Resets ....................................................................... 155 I2C Block Read................................................................................................. 160 Enable for SMBALERT# ..................................................................................... 162 Enables for SMBus Slave Write and SMBus Host Events ......................................... 162 Enables for the Host Notify Command ................................................................. 163 Slave Write Registers........................................................................................ 164 Command Types .............................................................................................. 165 Slave Read Cycle Format................................................................................... 165 Data Values for Slave Read Registers.................................................................. 166 Host Notify Format ........................................................................................... 168 Event Transitions that Cause Messages ............................................................... 170 SMBus Command Encoding................................................................................ 171 Internal SMBus Protocol Stack ........................................................................... 172 SMBus Slave Address Format............................................................................. 173 Memory Region Address Field ............................................................................ 173 Status Field Encoding for SMBus Reads ............................................................... 174 Address Map.................................................................................................... 182 Supported LPC Cycle Types ............................................................................... 182 I/O Sync Bits Description .................................................................................. 183 UART Clock Divider Support .............................................................................. 184 Baud Rate Example .......................................................................................... 184 UART Register/Signal Reset States ..................................................................... 185 Multi-Activity LED Message Type ........................................................................ 200 Legacy Replacement Routing ............................................................................. 202 USB EHCI Features........................................................................................... 205 Debug Port Behavior......................................................................................... 209 Region Size Versus Erase Granularity of Flash Components .................................... 218 Region Access Control Table .............................................................................. 221 Hardware Sequencing Commands and Opcode Requirements ................................. 223 Flash Protection Mechanism Summary................................................................. 224 Recommended Pinout for 8-Pin Serial Flash Device ............................................... 225 Recommended Pinout for 16-Pin Serial Flash Device ............................................. 226 Fixed I/O Ranges Decoded ............................................................................... 228 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 51 5-2 5-3 6-1 6-2 6-3 6-4 6-5 6-6 6-7 6-8 6-9 6-10 6-11 6-12 6-13 6-14 6-15 6-16 6-17 6-18 6-19 6-20 6-21 6-22 6-23 6-24 6-25 6-26 6-27 6-28 6-29 6-30 6-31 6-32 6-33 6-34 6-35 6-36 6-37 6-38 6-39 6-40 6-41 6-42 6-43 6-44 6-45 6-46 6-47 6-48 Variable I/O Decode Ranges............................................................................... 230 Memory Decode Ranges (From CPU Perspective)................................................... 231 PCI Configuration Registers (LPC I/F--B0:D31:F0) ................................................ 234 Offset 00h: ID: Vendor Identification Register ...................................................... 236 Offset 02h: DID--Device Identification Register (LPC I/F--B0:D31:F0) ..................... 236 Offset 04h: ID--PCICMD--PCI COMMAND Register (LPC I/F--B0:D31:F0) ................. 237 Offset 06h: ID--PCISTS--PCI Status Register (LPC I/F--B0:D31:F0) ........................ 238 Offset 08h: RID--Revision Identification Register (PCI-PCI--B0:D31:F0) .................. 239 Offset 09h: PI--Programming Interface Register (PCI-PCI--B0:D31:F0) ................... 239 Offset 0Ah: SCC--Sub Class Code Register (PCI-PCI--B0:D31:F0)........................... 239 Offset 0Bh: BCC--Base Class Code Register (LPC I/F--B0:D31:F0) .......................... 240 Offset 0Dh: PLT--Primary Latency Timer Register (LPC I/F--B0:D31:F0) .................. 240 Offset 0Eh: HEADTYP--Header Type Register (LPC I/F--B0:D31:F0) ........................ 240 Offset 2Ch: SS--Sub System Identifiers Register (LPC I/F--B0:D31:F0) ................... 241 Offset 40h: PMBASE--ACPI Base Address Register (LPC I/F--B0:D31:F0) ................. 241 Offset 44h: ACPI_CNTL--ACPI Control Register (LPC I/F--B0:D31:F0)...................... 242 Offset 48h: GPIOBASE--GPIO Base Address Register (LPC I/F--B0:D31:F0).............. 243 Offset 4Ch: GC--GPIO Control Register (LPC I/F--B0:D31:F0) ................................ 244 Offset 60h: PIRQ[n]_ROUT--PIRQ[A,B,C,D] Routing Control Register (LPC I/F-- B0:D31:F0) ..................................................................................................... 245 Offset 64h: SIRQ_CNTL--Serial IRQ Control Register (LPC I/F--B0:D31:F0).............. 246 Offset 68h: PIRQ[n]_ROUT--PIRQ[E,F,G,H] Routing Control Register (LPC I/F--B0:D31:F0) ....................................................................................... 247 Offset 6Ch: LPC_IBDF--IOxAPIC Bus:Device:Function (LPC I/F--B0:D31:F0) ............ 248 Offset 70h: LPC_HnBDF - HPET n Bus:Device:Function (LPC I/F--B0:D31:F0) .......... 249 Offset 80h: LPC_I/O_DEC--I/O Decode Ranges Register (LPC I/F--B0:D31:F0) ......... 250 Offset 82h: LPC_EN--LPC I/F Enables Register (LPC I/F--B0:D31:F0) ...................... 251 Offset 84h: GEN1_DEC--LPC I/F Generic Decode Range 1 Register (LPC I/F--B0:D31:F0) ....................................................................................... 253 Offset 88h: GEN2_DEC--LPC I/F Generic Decode Range 2 Register (LPC I/F--B0:D31:F0) ....................................................................................... 254 Offset 8Ch: GEN3_DEC--LPC I/F Generic Decode Range 3 Register (LPC I/F--B0:D31:F0) ....................................................................................... 255 Offset 90h: GEN4_DEC--LPC I/F Generic Decode Range 4 Register (LPC I/F--B0:D31:F0) ....................................................................................... 256 Offset 94h: ULKMC - USB Legacy Keyboard / Mouse Control (LPC I/F--B0:D31:F0).... 257 Offset 98h: LGMR -- LPC I/F Generic Memory Range (LPC I/F--B0:D31:F0) .............. 259 Offset DCh: BIOS_CNTL--BIOS Control Register (LPC I/F--B0:D31:F0) .................... 260 Offset E0h: FDCAP--Feature Detection Capability ID (LPC I/F--B0:D31:F0) .............. 261 Offset E2h: FDLEN--Feature Detection Capability Length (LPC I/F--B0:D31:F0) ........ 261 Offset E3h: FDVER--Feature Detection Version (LPC I/F--B0:D31:F0) ...................... 262 Offset E4h: FDVCT--Feature Vector (LPC I/F--B0:D31:F0) ..................................... 262 Offset F0h: RCBA--Root Complex Base Address Register (LPC I/F--B0:D31:F0) ........ 263 DMA I/O Registers (LPC I/F--B0:D31:F0)............................................................. 263 Offset 00h: DMABASE_CA--DMA Base and Current Address Registers (LPC I/F --B0:D31:F0) ...................................................................................... 264 Offset 01h: DMABASE_CC--DMA Base and Current Count Registers (LPC I/F--B0:D31:F0) ....................................................................................... 265 Offset 87h: DMAMEM_LP--DMA Memory Low Page Registers (LPC I/F--B0:D31:F0) ... 266 Offset 08h: DMACMD--DMA Command Register (LPC I/F--B0:D31:F0)..................... 267 Offset 08h: DMASTA--DMA Status Register (LPC I/F--B0:D31:F0)........................... 268 Offset 0Ah: DMA_WRSMSK--DMA Write Single Mask Register (LPC I/F--B0:D31:F0) ....................................................................................... 269 Offset 08h: DMACH_MODE--DMA Channel Mode Register (LPC I/F--B0:D31:F0)........ 270 Offset 0Ch: DMA Clear Byte Pointer Register (LPC I/F--B0:D31:F0) ......................... 271 Offset 0Dh: DMA Master Clear Register (LPC I/F--B0:D31:F0) ................................ 271 Offset 0Eh: DMA Master Clear Register (LPC I/F--B0:D31:F0)................................. 271 Offset 0Fh: DMA_WRMSK--DMA Write All Mask Register (LPC I/F--B0:D31:F0) ......... 272 Timer I/O Registers .......................................................................................... 272 Intel(R) Communications Chipset 89xx Series - Datasheet 52 October 2012 Order Number: 327879-001US Contents 6-49 6-50 6-51 6-52 6-53 6-54 6-55 6-56 6-57 6-58 6-59 6-60 6-61 6-62 6-63 6-64 6-65 6-66 6-67 6-68 6-69 6-70 6-71 6-72 6-73 6-74 6-75 6-76 6-77 6-78 6-79 6-80 6-81 6-82 6-83 6-84 6-85 6-86 6-87 6-88 6-89 6-90 6-91 6-92 6-93 6-94 6-95 6-96 6-97 6-98 6-99 Offset 43h: TCW--Timer Control Word Register (LPC I/F--B0:D31:F0)..................... 273 Offset 40h: SBYTE_FMT--Interval Timer Status Byte Format Register (LPC I/F--B0:D31:F0) ....................................................................................... 275 Offset 40h: Counter Access Ports Register (LPC I/F--B0:D31:F0) ............................ 276 PIC Registers................................................................................................... 277 Offset 20h: Master PIC ICW1--Master Initialization Command Word 1 Register ......... 278 Offset 21h: Master PIC ICW2-- Master Initialization Command Word 2 Register ........ 279 Offset 21h: Master PIC ICW3--Master Initialization Command Word 3 Register ......... 280 Offset A1h: Slave PIC ICW3--Slave Initialization Command Word 3 Register ............ 280 Offset 021h: Master PIC ICW4--Master Initialization Command Word 4 Register ....... 281 Offset 021h: Master PIC OCW1--Master Operational Control Word 1 (Interrupt Mask) Register .................................................................................. 282 Offset 020h: Master PIC OCW2--Master Operational Control Word 2 Register ........... 283 Offset 020h: Master PIC OCW3--Master Operational Control Word 3 Register ........... 284 Offset 4D0h: Master PIC ELCR1--Master Controller Edge/Level Triggered Register .... 285 Offset 4D1h: Slave PIC ELCR2--Slave Controller Edge/Level Triggered Register ........ 286 APIC Direct Registers........................................................................................ 287 IND--Index Register ......................................................................................... 287 DAT--Data Register .......................................................................................... 288 EOIR--EOI Register .......................................................................................... 289 APIC Indirect Registers ..................................................................................... 289 Offset 00h: ID--Identification Register ................................................................ 290 Offset 01h: VER--Version Register...................................................................... 291 Offset 10h: REDIR_TBL0--Redirection Table 0...................................................... 292 RTC I/O Registers ............................................................................................ 294 RTC (Standard) RAM Bank................................................................................. 295 Real Time Clock Registers ................................................................................. 295 Offset 0Ah: RTC_REGA--Register A..................................................................... 296 Offset 0Bh: RTC_REGB--Register B (General Configuration) ................................... 297 Offset OCh: RTC_REGC--Register C (Flag Register)............................................... 298 Offset ODh: D--Register D (Flag Register) ........................................................... 299 Processor Interface PCI Register Address Map (LPC I/F--B0:D31:F0)....................... 300 Offset 61h: NMI_SC--NMI Status and Control Register (LPC I/F--B0:D31:F0)........... 301 Offset 70h: NMI_EN--NMI Enable (and Real Time Clock Index) Register (LPC I/F--B0:D31:F0) ....................................................................................... 302 Offset 92h: PORT92--Fast A20 and Init Register (LPC I/F--B0:D31:F0) ................... 302 Offset F0h: COPROC_ERR--Coprocessor Error Register (LPC I/F--B0:D31:F0)........... 303 Offset CF9h: RST_CNT--Reset Control Register (LPC I/F--B0:D31:F0) ..................... 303 Power Management PCI Register Address Map (PM--B0:D31:F0) ............................ 305 Offset A0h: GEN_PMCON_1--General PM Configuration 1 Register (PM--B0:D31:F0) ............................................................................................. 306 Offset A2h: GEN_PMCON_2--General PM Configuration 2 Register (PM--B0:D31:F0) ............................................................................................. 308 Offset A4h: GEN_PMCON_3--General PM Configuration 3 Register (PM--B0:D31:F0) ............................................................................................. 310 Offset A6h: GEN_PMCON_LOCK- General Power Management Configuration Lock Register .......................................................................................................... 313 Offset A9h: Chipset Initialization Register 4 (PM--B0:D31:F0) ................................ 313 Offset ABh: BM_BREAK_EN Register (PM--B0:D31:F0) .......................................... 314 Offset ACh: PMIR--Power Management Initialization Register (PM--B0:D31:F0) ............................................................................................. 314 Offset B8h: GPIO_ROUT--GPIO Routing Control Register (PM--B0:D31:F0) .............. 315 APM Register Map ............................................................................................ 315 Offset B2h: APM_CNT--Advanced Power Management Control Port Register ............. 316 Offset B3h: APM_STS--Advanced Power Management Status Port Register .............. 316 ACPI and Legacy I/O Register Map...................................................................... 317 Offset PMBASE+00h: PM1_STS--Power Management 1 Status Register ................... 318 Offset PMBASE + 02h: PM1_EN--Power Management 1 Enable Register................... 321 Offset PMBASE + 04h: PM1_CNT--Power Management 1 Control ............................ 322 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 53 6-100 6-101 6-102 6-103 6-104 6-105 6-106 6-107 6-108 6-109 6-110 6-111 6-112 6-113 6-114 6-115 6-116 6-117 6-118 6-119 6-120 6-121 6-122 6-123 6-124 6-125 6-126 6-127 6-128 6-129 6-130 6-131 6-132 6-133 6-134 6-135 6-136 6-137 6-138 6-139 6-140 6-141 6-142 7-1 7-2 7-3 7-4 7-5 7-6 7-7 7-8 7-9 7-10 7-11 7-12 7-13 7-14 7-15 Offset PMBASE + 08h: PM1_TMR--Power Management 1 Timer Register .................. 323 Offset PMBASE + 20h: GPE0_STS--General Purpose Event 0 Status Register ............ 324 Offset PMBASE + 28h: GPE0_EN--General Purpose Event 0 Enables Register ............ 326 Offset PMBASE + 30h: SMI_EN--SMI Control and Enable Register ........................... 328 Offset PMBASE + 34h: SMI_STS--SMI Status Register........................................... 330 Offset PMBASE +38h: ALT_GP_SMI_EN--Alternate GPI SMI Enable Register ............. 333 Offset PMBASE +3Ah: ALT_GP_SMI_STS--Alternate GPI SMI Status Register............ 333 Offset PMBASE +3Ch: UPRWC--USB Per-Port Registers Write Control ...................... 334 Offset PMBASE +42h: GPE_CNTL--General Purpose Control Register ....................... 335 Offset PMBASE +44h: DEVACT_STS--Device Activity Status Register....................... 336 Offset PMBASE +50h: PM2_CNT--Power Management 2 Control.............................. 337 TCO I/O Register Address Map ........................................................................... 337 Offset TCOBASE +00h: TCO_RLD--TCO Timer Reload and Current Value Register ..... 338 Offset TCOBASE +02h: TCO_DAT_IN--TCO Data In Register .................................. 338 Offset TCOBASE +03h: TCO_DAT_OUT--TCO Data Out Register.............................. 339 Offset TCOBASE +04h: TCO1_STS--TCO1 Status Register...................................... 339 Offset TCOBASE +06h: TCO2_STS--TCO2 Status Register...................................... 341 Offset TCOBASE +08h: TCO1_CNT--TCO1 Control Register .................................... 343 Offset TCOBASE +0Ah: TCO2_CNT--TCO2 Control Register .................................... 344 Offset TCOBASE +0Ch and Offset TCOBASE +0Dh: TCO_MESSAGE1 and TCO_MESSAGE2 Registers ................................................................................. 345 Offset TCOBASE + 0Eh: TCO_WDCNT--TCO Watchdog Control Register ................... 345 Offset TCOBASE + 10h: SW_IRQ_GEN--Software IRQ Generation Register............... 346 Offset TCOBASE + 12h:TCO_TMR--TCO Timer Initial Value Register ........................ 346 General Purpose I/O Signals............................................................................... 347 Offset GPIOBASE + 00h: GPIO_USE_SEL--GPIO Use Select Register ....................... 348 Offset GPIOBASE + 04h: GP_IO_SEL--GPIO Input/Output Select Register ................ 349 Offset GPIOBASE + 0Ch:GP_LVL--GPIO Level for Input or Output Register ............... 349 Offset GPIOBASE + 18h: GPO_BLINK--GPO Blink Enable Register ........................... 350 Offset GPIOBASE + 1Ch: GP_SER_BLINK--GP Serial Blink Data .............................. 351 Offset GPIOBASE + 20h: GP_SB_CMDSTS--GP Serial Blink Command Status............ 352 Offset GPIOBASE + 24h: GP_SB_DATA--GP Serial Blink Data ................................. 353 Offset GPIOBASE + 28h: GPI_NMI_EN--GPI NMI Enable ........................................ 353 Offset GPIOBASE + 2Ah: GPI_NMI_STS--GPI NMI Status ....................................... 354 Offset GPIOBASE + 2Ch: GPI_INV--GPIO Signal Invert Register ............................. 354 Offset GPIOBASE + 30h: GPIO_USE_SEL2--GPIO Use Select 2 Register ................... 355 Offset GPIOBASE + 34h: GP_IO_SEL2--GPIO Input/Output Select 2 Register ........... 356 Offset GPIOBASE + 38h: GP_LVL2--GPIO Level for Input or Output 2 Register.......... 356 Offset GPIOBASE + 40h: GPIO_USE_SEL3--GPIO Use Select 3 Register ................... 357 Offset GPIOBASE + 44h: GP_IO_SEL3--GPIO Input/Output Select 3 Register ........... 358 Offset GPIOBASE + 48h: GP_LVL3--GPIO Level for Input or Output 3 Register.......... 359 Offset GPIOBASE + 60h: GP_RST_SEL1--GPIO Reset Select ................................... 360 Offset GPIOBASE + 64h: GP_RST_SEL2--GPIO Reset Select ................................... 361 Offset GPIOBASE + 68h: GP_RST_SEL3--GPIO Reset Select ................................... 361 Chipset Configuration Registers .......................................................................... 362 Offset 0014h: V0CTL--Virtual Channel 0 Resource Control Register ......................... 365 Offset 001Ah: V0STS--Virtual Channel 0 Resource Status Register .......................... 366 Offset 001Ch: V1CAP--Virtual Channel 1 Resource Capability Register ..................... 366 Offset 0020h: V1CTL--Virtual Channel 1 Resource Control Register ......................... 367 Offset 0026h: V1STS--Virtual Channel 1 Resource Status Register .......................... 367 Offset 0050h: CIR0--Chipset Initialization Register 0 ............................................. 368 Offset 0088h: CIR1--Chipset Initialization Register 1 ............................................. 368 Offset 00ACh: REC--Root Error Command Register ............................................... 369 Offset 01A0h: ILCL--Internal Link Capabilities List Register .................................... 369 Offset 01A4h: LCAP--Link Capabilities Register ..................................................... 370 Offset 01A8h: LCTL--Link Control Register ........................................................... 370 Offset 01AAh: LSTS--Link Status Register............................................................ 371 Offset 0220h: BCR--Backbone Configuration Register ............................................ 371 Offset 0224h: RPC--Root Port Configuration Register............................................. 372 Intel(R) Communications Chipset 89xx Series - Datasheet 54 October 2012 Order Number: 327879-001US Contents 7-16 7-17 7-18 7-19 7-20 7-21 7-22 7-23 7-24 7-25 7-26 7-27 7-28 7-29 7-30 7-31 7-32 7-33 7-34 7-35 7-36 7-37 7-38 7-39 7-40 7-41 7-42 7-43 7-44 7-45 7-46 7-47 7-48 7-49 7-50 7-51 7-52 7-53 7-54 7-55 7-56 7-57 7-58 7-59 7-60 7-61 7-62 7-63 7-64 7-65 7-66 7-67 7-68 7-69 7-70 8-1 8-2 8-3 Offset 0234h: DMIC--DMI Control Register .......................................................... 373 Offset 0238h: RPFN--Root Port Function Number and Hide for PCI Express* Root Ports ....................................................................................................... 373 Offset 0290h: Reserved .................................................................................... 374 Offset 1D40H: CIR5--Chipset Initialization Register 5............................................ 375 Offset 1E00h: TRSR--Trap Status Register........................................................... 375 Offset 1E10h: TRCR--Trapped Cycle Register ....................................................... 376 Offset 1E18h: TWDR--Trapped Write Data Register............................................... 376 Offset 1E80h: IOTRn--I/O Trap Register (0-3) ..................................................... 377 Offset 2010h: DMC--DMI Miscellaneous Control Register ....................................... 378 Offset 2024h: CIR6--Chipset Initialization Register 6 ............................................ 378 Offset 2324h: DMC2--DMI Miscellaneous Control Register 2................................... 379 Offset 60h: SBI Unified AFE Address Register (SATA-B0:D31:F2) ........................... 379 Offset 64h: SBI Unified AFE Data Register (SATA-B0:D31:F2)................................ 380 Offset 68h: SBI Unified AFE Status Register (SATA-B0:D31:F2) ............................. 381 Offset 3000h: TCTL--TCO Configuration Register .................................................. 382 Offset 3100h: D31IP--Device 31 Interrupt Pin Register ......................................... 383 Offset 3108h: D29IP--Device 29 Interrupt Pin Register ......................................... 384 Offset 310Ch: D28IP--Device 28 Interrupt Pin Register ......................................... 385 Offset 3124h: D22IP--Device 22 Interrupt Pin Register ......................................... 386 Offset 3140h: D31IR--Device 31 Interrupt Route Register ..................................... 387 Offset 3144h: D29IR--Device 29 Interrupt Route Register ..................................... 388 Offset 3146h: D28IR--Device 28 Interrupt Route Register ..................................... 389 Offset 315Ch: D22IR--Device 22 Interrupt Route Register ..................................... 390 Offset 31FEh: OIC--Other Interrupt Control Register............................................. 391 Offset 3310h: PRSTS--Power and Reset Status .................................................... 392 Offset 3314h: CIR7--Chipset Initialization Register 7 ............................................ 393 Offset 3324h: CIR8--Chipset Initialization Register 8 ............................................ 393 Offset 3330h: CIR9--Chipset Initialization Register 9 ............................................ 393 Offset 3340h: CIR10--Chipset Initialization Register 10 ......................................... 394 Offset 3350h: CIR13--Chipset Initialization Register 13 ......................................... 394 Offset 3368h: CIR14--Chipset Initialization Register 14 ......................................... 394 Offset 3378h: CIR15--Chipset Initialization Register 15 ......................................... 395 Offset 3388h: CIR16--Chipset Initialization Register 16 ......................................... 395 Offset 33A0h: CIR17--Chipset Initialization Register 17......................................... 395 Offset 33A8h: CIR18--Chipset Initialization Register 18......................................... 396 Offset 33C0h: CIR19--Chipset Initialization Register 19......................................... 396 Offset 33CCh: CIR20--Chipset Initialization Register 20......................................... 396 Offset 33D0h: CIR21--Chipset Initialization Register 21......................................... 397 Offset 33D4h: CIR22--Chipset Initialization Register 22......................................... 397 Offset 3400h: RC--RTC Configuration Register ..................................................... 398 Offset 3404h: HPTC--High Precision Timer Configuration Register ........................... 399 Offset 3410h: GCS--General Control and Status Register....................................... 400 Offset 3414h: BUC--Backed Up Control Register................................................... 402 Offset 3418h: FD--Function Disable Register........................................................ 403 Offset 341Ch: CG--Clock Gating......................................................................... 405 Offset 3420h: FDSW--Function Disable SUS Well.................................................. 406 Offset 3428h: FD2--Function Disable 2 ............................................................... 406 Offset 3500h: USBIR[0:5]--USB Initialization Register [0-5] .................................. 407 Offset 3564h: USBIRC--USB Initialization Register C............................................. 408 Offset 3570h: USBIRA--USB Initialization Register A............................................. 408 Offset 357Ch: USBIRB--USB Initialization Register B............................................. 409 Offset 3590h: MISCCTL--Miscellaneous Control Register........................................ 410 Offset 359Ch: PDO--USB Port Disable Override .................................................... 411 Offset 35A0h: USBOCM1--Overcurrent MAP Register 1 .......................................... 412 Offset 35B0h: RMHWKCTL--Rate Matching Hub Wake Control Register .................... 413 SATA Controller PCI Register Address Map ........................................................... 415 Offset 00h: Vendor Identification Register (SATA-B0:D31:F2)................................ 418 Offset 02h: Device Identification Register (SATA-B0:D31:F2) ................................ 418 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 55 8-4 8-5 8-6 8-7 8-8 8-9 8-10 8-11 8-12 8-13 8-14 8-15 8-16 8-17 8-18 8-19 8-20 8-21 8-22 8-23 8-24 8-25 8-26 8-27 8-28 8-29 8-30 8-31 8-32 8-33 8-34 8-35 8-36 8-37 8-38 8-39 8-40 8-41 8-42 8-43 8-44 8-45 8-46 8-47 8-48 8-49 8-50 8-51 8-52 8-53 8-54 8-55 8-56 Offset 04h: PCI Command Register (SATA-B0:D31:F2) ......................................... 418 Offset 06h: PCI Status Register (SATA-B0:D31:F2)............................................... 420 Offset 08h: Revision Identification Register (SATA-B0:D31:F2)............................... 421 Offset 0Ah: When Sub Class Code Register (B0:D31:F2:Offset 0Ah) = 01h............... 421 Offset 0Ah: When Sub Class Code Register (B0:D31:F2:Offset 0Ah) = 06h............... 422 Offset 0Ah: Sub Class Code Register (SATA-B0:D31:F2)........................................ 423 Offset 0Bh: Base Class Code Register (B0:D31:F2) ............................................... 423 Offset 0Dh: Primary Master Latency Timer Register (B0:D31:F2) ............................ 423 Offset 0Eh: Header Type (SATA-B0:D31:F2) ........................................................ 424 Offset 10h: Primary Command Block Base Address Register (SATA-B0:D31:F2)........ 424 Offset 14h: Primary Control Block Base Address Register (SATA-B0:D31:F2)............ 425 Offset 18h: Secondary Command Block Base Address Register (IDE D31:F1)............ 425 Offset 1Ch: Secondary Control Block Base Address Register (IDE B0:D31:F2)........... 426 Offset 20h: Legacy Bus Master Base Address Register (SATA-B0:D31:F2)................ 426 Offset 24h: When SCC is Not 01h (SATA-B0:D31:F2)............................................ 427 Offset 24h: When SCC is 01h (SATA-B0:D31:F2).................................................. 428 Offset 2Ch: Subsystem Vendor Identification Register (SATA-B0:D31:F2) ................ 428 Offset 2Eh: Subsystem Identification Register (SATA-B0:D31:F2) ........................... 429 Offset 34h: Capabilities Pointer Register (SATA-B0:D31:F2)................................... 429 Offset 3Ch: Interrupt Line Register (SATA-B0:D31:F2) .......................................... 429 Offset 3Dh: Interrupt Pin Register (SATA-B0:D31:F2) ........................................... 430 Offset 40h: IDE Timing Register (SATA-B0:D31:F2) .............................................. 430 Offset 48h: Synchronous DMA Control Register (SATA-B0:D31:F2) ......................... 431 Offset 4Ah: Synchronous DMA Timing Register (SATA-B0:D31:F2).......................... 431 Offset 4Ah: IDE I/O Configuration Register (SATA-B0:D31:F2) ............................... 432 Offset 70h: PCI Power Management Capability Identification Register (SATA-B0:D31:F2) ........................................................................................... 432 Offset 72h: PCI Power Management Capabilities Register (SATA-B0:D31:F2) ........... 433 Offset 74h: PCI Power Management Control and Status Register (SATA-B0:D31:F2) ........................................................................................... 434 Offset 80h: Message Signaled Interrupt Capability Identification (SATA-B0:D31:F2) ........................................................................................... 435 Offset 82h: Message Signaled Interrupt Message Control (SATA-B0:D31:F2)............ 436 Offset 84h: Message Signaled Interrupt Message Address (SATA-B0:D31:F2)........... 437 Offset 88h: Message Signaled Interrupt Message Data (SATA-B0:D31:F2) ............... 437 Offset 90h: MAP--Address Map Register (SATA-B0:D31:F2) ................................... 438 Offset 92h: PCS--Port Control and Status Register (SATA-B0:D31:F2) .................... 440 Offset 94h: SCLKCG--SATA Clock Gating Control Register (SATA-B0:D31:F2) .......... 441 Offset 9Ch: SCLKGC--SATA Clock General Configuration Register (SATA-B0:D31:F2) ........................................................................................... 442 Offset A0h: SIRI--SATA Indexed Registers Index (SATA-B0:D31:F2) ...................... 443 Offset A4h: STRD--SATA Indexed Register Data (SATA-B0:D31:F2)........................ 443 Offset A8h: SATACR0--SATA Capability Register 0 (SATA-B0:D31:F2)..................... 444 Offset ACh: SATACR1--SATA Capability Register 1 (SATA-B0:D31:F2)..................... 445 Offset B0h: FLRCID--FLR Capability ID (SATA-B0:D31:F2) .................................... 446 Offset B2h: FRLCLV--FLR Capability Length and Version (SATA-B0:D31:F2) ............. 446 Offset B2h: FLRCLV--FLR Capability Length and Version (SATA-B0:D31:F2) ............. 447 Offset B4h: FLRC--FLR Control (SATA-B0:D31:F2)................................................ 447 Offset C0h: ATC--APM Trapping Control Register (SATA-B0:D31:F2)....................... 448 Offset C4h: ATS--APM Trapping Status Register (SATA-B0:D31:F2) ........................ 448 Offset D0h: SP--Scratch Pad Register (SATA-B0:D31:F2) ...................................... 449 Offset E0h: BFCS--BIST FIS Control/Status Register (SATA-B0:D31:F2) .................. 450 Offset E4h: FBTD1--BIST FIS Transmit Data1 Register (SATA-B0:D31:F2)............... 451 Offset E8h: BFTD2--BIST FIS Transmit Data2 Register (SATA-B0:D31:F2)............... 452 SATA Indexed Registers .................................................................................... 453 Offset 18h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) ........................................................................................... 454 Offset 1Ch: SATA Indexed Registers Index (SATA Test Mode Enable Register) (SATA-B0:D31:F2) ........................................................................................... 454 Intel(R) Communications Chipset 89xx Series - Datasheet 56 October 2012 Order Number: 327879-001US Contents 8-57 8-58 8-59 8-60 8-61 8-62 8-63 8-64 8-65 8-66 8-67 8-68 8-69 8-70 8-71 8-72 8-73 8-74 8-75 8-76 8-77 8-78 8-79 8-80 8-81 8-82 8-83 8-84 8-85 8-86 8-87 8-88 8-89 8-90 8-91 8-92 8-93 8-94 8-95 8-96 8-97 8-98 8-99 Offset 28h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 455 Offset 3Eh: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 455 Offset 64h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 456 Offset 64h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 456 Offset 68h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 457 Offset 78h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 457 Offset 84h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 458 Offset 88h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 458 Offset 8Ch: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 459 Offset 94h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 459 Offset A0h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 460 Offset A8h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 460 Offset C4h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 461 Offset C8h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2)........................................................................................... 461 Bus Master IDE I/O Register Address Map............................................................ 462 Offset 00h: Bus Master IDE Command Register Primary (B0:D31:F2)...................... 463 Offset 02h: Bus Master IDE Status Register Primary (B0:D31:F2) ........................... 464 Offset 04h: Bus Master IDE Descriptor Table Pointer Register Primary (B0:D31:F2)... 465 Offset 08h: Bus Master IDE Command Register Secondary (B0:D31:F2) .................. 466 Offset 0Ah: Bus Master IDE Status Register Secondary (B0:D31:F2) ....................... 467 Offset 0Ch: Bus Master IDE Descriptor Table Pointer Register Secondary (B0:D31:F2).................................................................................................... 468 Offset 10h: AHCI Index Register (B0:D31:F2) ...................................................... 468 Offset 14h: AHCI Index Data Register (B0:D31:F2) .............................................. 469 Superset Registers ........................................................................................... 470 Offset SIDPBA + 00h: Serial ATA Index (B0:D31:F2) ............................................ 470 Offset SIDPBA + 04h: Serial ATA Data (B0:D31:F2).............................................. 471 AHCI Register Address Map ............................................................................... 471 AHCI Generic Host Control Registers ................................................................... 472 Offset 00h: Host Capabilities Register (B0:D31:F2)............................................... 472 Offset 04h: Global PCH Control Register (B0:D31:F2)............................................ 477 Offset 08h: Interrupt Status Register (B0:D31:F2) ............................................... 478 Offset 0Ch: Ports Implemented Register (B0:D31:F2) ........................................... 479 Offset 10h: Serial AHCI Version (B0:D31:F2) ....................................................... 479 Offset 14h: Serial Command Completion Coalescing Control Register (B0:D31:F2).... 480 Offset 18h: Serial Command Completion Coalescing Ports Register (B0:D31:F2)....... 481 Offset 1Ch: Serial Enclosure Management Location Register (B0:D31:F2) ................ 481 Offset 20h: Serial Enclosure Management Control Register (B0:D31:F2) .................. 482 Offset 24h: Serial Extended Host Capabilities (B0:D31:F2) .................................... 483 Offset SIDPBA + 04h: Serial ATA Index (B0:D31:F2) ............................................ 484 Offset 70h: Serial AHCI Version (B0:D31:F2) ....................................................... 484 Offset A0h: Serial Vendor Specific (B0:D31:F2).................................................... 485 Port [4:5] DMA Register Address Map ................................................................. 486 Offset 300h: Port [4:5] Command List Base Address Register (B0:D31:F2) .............. 487 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 57 8-100 8-101 8-102 8-103 8-104 8-105 8-106 8-107 8-108 8-109 8-110 8-111 8-112 9-1 9-2 9-3 9-4 9-5 9-6 9-7 9-8 9-9 9-10 9-11 9-12 9-13 9-14 9-15 9-16 9-17 9-18 9-19 9-20 9-21 9-22 9-23 9-24 9-25 9-26 9-27 9-28 9-29 9-30 9-31 9-32 9-33 9-34 9-35 9-36 9-37 9-38 9-39 9-40 9-41 9-42 9-43 Offset 304: Port [4:5] Command List Base Address Upper 32-Bits Register (B0:D31:F2) .................................................................................................... 487 Offset 308h: Port [4:5] FIS Base Address Register (B0:D31:F2).............................. 488 Offset 30Ch: Port [4:5] FIS Base Address Upper 32-Bits Register (B0:D31:F2) ......... 488 Offset 310h: Port [4:5] Interrupt Status Register (B0:D31:F2) ............................... 489 Offset 314h: Port [4:5] Interrupt Enable Register (B0:D31:F2) ............................... 491 Offset 318h: Port [4:5] Command Register (B0:D31:F2)........................................ 493 Offset 320h: Port [4:5] Task File Data Register (B0:D31:F2) .................................. 496 Offset 324h: Port [4:5] Signature Register (B0:D31:F2) ........................................ 497 Offset 328h: Port [4:5] Serial ATA Status Register (B0:D31:F2).............................. 498 Offset 32Ch: Port [4:5] Serial ATA Control Register (B0:D31:F2) ............................ 500 Offset 330h: Port [4:5] Serial ATA Error Register (B0:D31:F2)................................ 502 Offset 334h: Port [4:5] Serial ATA Active (B0:D31:F2) .......................................... 503 Offset 338h: Port [4:5] Command Issue Register (B0:D31:F2) ............................... 504 SATA Controller PCI Register Address Map ........................................................... 505 Offset 00h: VID: Vendor Identification Register (SATA--B0:D31:F5) ........................ 506 Offset 02h: Device Identification Register (SATA--B0:D31:F5) ................................ 507 Offset 04h: PCI Command Register (SATA--B0:D31:F5) ........................................ 507 Offset 06h: PCI Status Register (SATA--B0:D31:F5).............................................. 508 Offset 08h: RID--Revision Identification Register (SATA--B0:D31:F5)...................... 509 Offset 09h: Programming Interface Register (SATA--B0:D31:F5) ............................ 510 Offset 0Ah: Sub Class Code Register (SATA--B0:D31:F5)....................................... 510 Offset 0Bh: BCC--Base Class Code Register (SATA-B0:D31:F5) .............................. 511 Offset 0Dh: Primary Master Latency Timer Register (SATA-B0:D31:F5) ................... 511 Offset 10h: Primary Command Block Base Address Register (SATA-B0:D31:F5)........ 512 Offset 14h: Primary Control Block Base Address Register (SATA-B0:D31:F5)............ 513 Offset 18h: Secondary Command Block Base Address Register (IDE D31:F1)............ 513 Offset 1Ch: Secondary Control Block Base Address Register (IDE B0:D31:F5)........... 514 Offset 20h: Legacy Bus Master Base Address Register (SATA-B0:D31:F5)................ 514 Offset 24h: SATA Index/Data Pair Base Address Register (SATA-B0:D31:F5) ........... 515 Offset 2Ch: Subsystem Vendor Identification Register (SATA-B0:D31:F5) ................ 515 Offset 2Eh: Subsystem Identification Register (SATA-B0:D31:F5) ........................... 516 Offset 34h: Capabilities Pointer Register (SATA-B0:D31:F5)................................... 516 Offset 3Ch: Interrupt Line Register (SATA-B0:D31:F5) .......................................... 516 Offset 3Dh: Interrupt Pin Register (SATA-B0:D31:F5) ........................................... 517 Offset 40h: IDE Timing Register (SATA-B0:D31:F5) .............................................. 517 Offset 48h: Synchronous DMA Control Register (SATA-B0:D31:F5) ......................... 518 Offset 4Ah: Synchronous DMA Timing Register (SATA-B0:D31:F5).......................... 518 Offset 4Ah: IDE I/O Configuration Register (SATA-B0:D31:F5) ............................... 519 Offset 70h: PCI Power Management Capability Identification Register (SATA-B0:D31:F5) ........................................................................................... 520 Offset 72h: PC--PCI Power Management Capabilities Register (SATA-B0:D31:F5) ..... 520 Offset 74h: PCI Power Management Control and Status Register (SATA-B0:D31:F5) ........................................................................................... 521 Offset 90h: MAP--Address Map Register (SATA-B0:D31:F5) ................................... 522 Offset 92h: Port Control and Status Register (SATA-B0:D31:F5)............................. 523 Offset A8h: SATA Capability Register 0 (SATA-B0:D31:F5) .................................... 524 Offset ACh: SATA Capability Register 1 (SATA-B0:D31:F5) .................................... 524 Offset B0h: FLR Capability ID (SATA-B0:D31:F5) ................................................. 525 Offset B2h: FLR Capability Length and Value (SATA-B0:D31:F5)............................. 525 Offset B2h: FLR Capability Length and Value (SATA-B0:D31:F5)............................. 526 Offset B4h: FLR Control (SATA-B0:D31:F5) ......................................................... 526 Offset C0h: APM Trapping Control Register (SATA-B0:D31:F5) ............................... 527 Offset C4h: APM Trapping Control Register (SATA-B0:D31:F5) ............................... 527 Bus Master IDE I/O Registers (B0:D31:F5)........................................................... 528 Offset 00h: Bus Master IDE Command Register (B0:D31:F5) .................................. 528 Offset 02h: Bus Master IDE Status Register (B0:D31:F5) ....................................... 530 Offset 04h: Bus Master IDE Descriptor Table Pointer Register (B0:D31:F5)............... 531 Serial ATA Index/Data Pair Superset Registers...................................................... 531 Intel(R) Communications Chipset 89xx Series - Datasheet 58 October 2012 Order Number: 327879-001US Contents 9-44 9-45 9-46 9-47 9-48 10-1 10-2 10-3 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 10-12 10-13 10-14 10-15 10-16 10-17 10-18 10-19 10-20 10-21 10-22 10-23 10-24 10-25 10-26 10-27 10-28 10-29 10-30 10-31 10-32 10-33 10-34 10-35 10-36 10-37 10-38 10-39 10-40 10-41 10-42 10-43 Offset 00h: SINDX--SATA Index Register (B0:D31:F5).......................................... 532 Offset 04h: SDATA--SATA Index Data Register (B0:D31:F5) .................................. 532 Offset 04h: PxSSTS--Serial ATA Status Register (B0:D31:F5) ................................ 533 Offset 04h: PxSCTL--Serial ATA Control Register (B0:D31:F5) ............................... 535 Offset 04h: PxSERR--Serial ATA Error Register (B0:D31:F5) .................................. 537 USB EHCI PCI Register Address Map ................................................................... 539 Offset 00h: VID: Vendor Identification Register (USB EHCI--B0:D29:F0) ................. 542 Offset 02h: DID--Device Identification Register (USB EHCI--B0:D29:F0) ................. 542 Offset 04h: PCI Command Register (USB EHCI--B0:D29:F0)................................. 543 Offset 06h: PCI Status Register (USB EHCI--B0:D29:F0) ....................................... 545 Offset 08h: RID--Revision Identification Register (USB EHCI--B0:D29:F0) ............... 546 Offset 09h: Programming Interface Register (USB EHCI--B0:D29:F0) ..................... 546 Offset 0Ah: Sub Class Code Register (USB EHCI--B0:D29:F0) ................................ 546 Offset 0Bh: BCC--Base Class Code Register (USB EHCI--B0:D29:F0) ...................... 547 Offset 0Dh: Primary Master Latency Timer Register (USB EHCI--B0:D29:F0) ........... 547 Offset 0Eh: Header Type Register (USB EHCI--B0:D29:F0) .................................... 548 Offset 10h: Memory Base Address Register (USB EHCI--B0:D29:F0)....................... 548 Offset 2Ch: Subsystem Vendor ID Register (USB EHCI--B0:D29:F0) ....................... 549 Offset 2Eh: Subsystem ID Register (USB EHCI--B0:D29:F0) .................................. 549 Offset 34h: Capabilities Pointer Register (USB EHCI--B0:D29:F0) ........................... 550 Offset 3Ch: Interrupt Line Register (USB EHCI--B0:D29:F0) .................................. 550 Offset 3Dh: Interrupt Pin Register (USB EHCI--B0:D29:F0) ................................... 550 Offset 50h: PCI Power Management Capability ID Register (USB EHCI--B0:D29:F0) ................................................................................... 551 Offset 51h: Next Item Pointer #1 Register (USB EHCI--B0:D29:F0) ........................ 551 Offset 52h: Power Management Capabilities Register (USB EHCI--B0:D29:F0) ......... 552 Offset 54h: Power Management Control/Status Register (USB EHCI--B0:D29:F0) ..... 553 Offset 58h: Debug Port Capability ID Register (USB EHCI--B0:D29:F0) ................... 554 Offset 59h: Next Item Pointer #2 Register (USB EHCI--B0:D29:F0) ........................ 554 Offset 5Ah: Debug Port Base Offset Register (USB EHCI--B0:D29:F0) ..................... 554 Offset 60h: USB Release Number Register (USB EHCI--B0:D29:F0) ........................ 555 Offset 61h: Frame Length Adjustment Register (USB EHCI--B0:D29:F0) ................. 556 Offset 62h: Port Wake Capability Register (USB EHCI--B0:D29:F0)......................... 557 Offset 68h: Legacy Support Extended Capability Register (USB EHCI--B0:D29:F0) ... 558 Offset 6Ch: Legacy Support Extended Control/Status Register (USB EHCI--B0:D29:F0) .................................................................................. 559 Offset 70h: SPECIAL_SMI--Intel Specific USB 2.0 SMI Register (USB EHCI--B0:D29:F0) ................................................................................... 562 Offset 80h: ACCESS_CNTL--Access Control Register (USB EHCI--B0:D29:F0) .......... 564 Offset 84h: EHCIIR1--EHCI Initialization Register 1 (USB EHCI--B0:D29:F0) ........... 564 Offset 98h: FLR_CID--Function Level Reset Capability ID (USB EHCI--B0:D29:F0).... 565 Offset 99h: FLR_NEXT--Function Level Reset Next Capability Pointer (USB EHCI--B0:D29:F0) ................................................................................... 565 Offset 9Ah: FLR_CLV--Function Level Reset Capability Length and Version (USB EHCI--B0:D29:F0) .................................................................................. 566 Offset 9Ah: FLR_CLV--Function Level Reset Capability Length and Version (USB EHCI--B0:D29:F0) .................................................................................. 566 Offset 9Ch: FLR_CTRL--Function Level Reset Control Register (USB EHCI--B0:D29:F0) ................................................................................... 567 Offset 9Dh: FLR_STS--Function Level Reset Status Register (USB EHCI--B0:D29:F0) ................................................................................... 567 Enhanced Host Controller Capability Registers ...................................................... 568 Offset 00h: CAPLENGTH--Capability Registers Length (USB EHCI--B0:D29:F0)......... 569 Offset 02h: HCIVERSION--Host Controller Interface Version Number Register (USB EHCI--B0:D29:F0) ................................................................................... 569 Offset 04h: HCSPARAMS--Host Controller Structural Parameters (USB EHCI--B0:D29:F0) ................................................................................. 570 Offset 08h: HCCPARAMS--Host Controller Capability Parameters Register (USB EHCI--B0:D29:F0) ................................................................................... 571 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 59 10-44 10-45 10-46 10-47 10-48 10-49 10-50 10-51 10-52 10-53 10-54 10-55 10-56 10-57 10-58 11-1 11-2 11-3 11-4 11-5 11-6 11-7 11-8 11-9 11-10 11-11 11-12 11-13 11-14 11-15 11-16 11-17 11-18 11-19 11-20 11-21 11-22 11-23 11-24 11-25 11-26 11-27 11-28 11-29 11-30 11-31 11-32 11-33 11-34 11-35 11-36 11-37 12-1 12-2 Enhanced Host Controller Operational Register Address Map................................... 572 Offset 20h: USB2.0_CMD--USB 2.0 Command Register (USB EHCI--B0:D29:F0) ...... 573 Offset 24h: USB2.0_STS--USB 2.0 Status Register (USB EHCI--B0:D29:F0)............. 576 Offset 28h: USB2.0_INTR--USB 2.0 Interrupt Enable Register (USB EHCI--B0:D29:F0).................................................................................... 579 Offset 2Ch: FRINDEX--Frame Index Register (USB EHCI--B0:D29:F0) ..................... 581 Offset 30h: CTRLDSSEGMENT--Control Data Structure Segment Register (USB EHCI--B0:D29:F0) ................................................................................... 582 Offset 34h: PERIODICLISTBASE--Periodic Frame List Base Address Register (USB EHCI--B0:D29:F0) ....................................................................... 583 Offset 38h: ASYNCLISTADDR--Current Asynchronous List Address Register (USB EHCI--B0:D29:F0).................................................................................... 584 Offset 60h: CONFIGFLAG--Configure Flag Register (USB EHCI--B0:D29:F0) ............. 585 Offset 64h: PORTSC--Port N Status and Control Register (USB EHCI--B0:D29:F0) .... 586 Debug Port Register Address Map ....................................................................... 591 Offset A0h: CNTL_STS--Control/Status Register (USB EHCI--B0:D29:F0) ................ 592 Offset A4h: USBPID--USB PIDs Register (USB EHCI--B0:D29:F0) ........................... 594 Offset A8h: DATABUF[7:0]--Data Buffer Bytes[7:0] Register (USB EHCI--B0:D29:F0).................................................................................... 595 Offset B0h: CONFIG--Configuration Register (USB EHCI--B0:D29:F0) ..................... 595 SMBus Controller PCI Register Address Map ......................................................... 596 Offset 00h: Vendor Identification Register (SMBus--B0:D31:F3) ............................. 597 Offset 02h: Device Identification Register (SMBus--B0:D31:F3) .............................. 597 Offset 04h: PCI Command Register (SMBus--B0:D31:F3) ...................................... 598 Offset 06h: PCI Status Register (SMBus--B0:D31:F3)............................................ 599 Offset 08h: Revision Identification Register (SMBus--B0:D31:F3)............................ 600 Offset 09h: Programming Interface Register (SMBus--B0:D31:F3) .......................... 600 Offset 0Ah: Sub Class Code Register (SMBus--B0:D31:F3)..................................... 600 Offset 0Bh: Base Class Code Register (SMBus--B0:D31:F3) ................................... 601 Offset 10h: SMBus Memory Base Address 0 (SMBus--B0:D31:F3) ........................... 601 Offset 14h: SMBus Memory Base Address 1 (SMBus--B0:D31:F3) ........................... 602 Offset 20h: SMBus Base Address Register (SMBus--B0:D31:F3) ............................. 602 Offset 2Ch: Subsystem Vendor Identification Register (SMBus--B0:D31:F3) ............. 603 Offset 2Eh: Subsystem Identification Register (SMBus--B0:D31:F3) ........................ 603 Offset 3Ch: Interrupt Line Register (SMBus--B0:D31:F3) ....................................... 604 Offset 3Dh: Interrupt Pin Register (SMBus--B0:D31:F3) ........................................ 604 Offset 40h: Host Configuration Register (SMBus--B0:D31:F3) ................................ 605 SMBus Bus Registers......................................................................................... 606 Offset 00h: Host Status Register (SMBus--B0:D31:F3) .......................................... 607 Offset 02h: Host Control Register (SMBus--B0:D31:F3) ......................................... 609 Offset 03h: Host Command Register (SMBus--B0:D31:F3) ..................................... 611 Offset 04h: Transmit Slave Address Register (SMBus--B0:D31:F3).......................... 611 Offset 05h: Host Data 0 Register (SMBus--B0:D31:F3).......................................... 612 Offset 06h: Host Data 1 Register (SMBus--B0:D31:F3).......................................... 612 Offset 07h: Host Block Data Byte Register (SMBus--B0:D31:F3) ............................. 613 Offset 08h: Packet Error Check (PEC) Register (SMBus--B0:D31:F3) ....................... 614 Offset 09h: Receive Slave Address Register (SMBus--B0:D31:F3) ........................... 614 Offset 0Ah: Receive Slave Data Register (SMBus--B0:D31:F3) ............................... 615 Offset 0Ch: Auxiliary Status Register (SMBus--B0:D31:F3) .................................... 615 Offset 0Dh: Auxiliary Control Register (SMBus--B0:D31:F3) ................................... 616 Offset 0Eh: SMLink Pin Control Register (SMBus--B0:D31:F3) ................................ 617 Offset 0Fh: SMBus Pin Control Register (SMBus--B0:D31:F3) ................................. 618 Offset 10h: Slave Status Register (SMBus--B0:D31:F3) ......................................... 619 Offset 11h: Slave Command Register (SMBus--B0:D31:F3).................................... 620 Offset 14h: Notify Device Address Register (SMBus--B0:D31:F3) ............................ 621 Offset 16h: Notify Data Low Byte Register (SMBus--B0:D31:F3) ............................. 621 Offset 17h: Notify Data High Byte Register (SMBus--B0:D31:F3) ............................ 622 PCI Express* Configuration Registers Address Map................................................ 623 Offset 00h: Vendor Identification Register (PCI Express*--B0:D28:F0/F1/F2/F3)....... 626 Intel(R) Communications Chipset 89xx Series - Datasheet 60 October 2012 Order Number: 327879-001US Contents 12-3 12-4 12-5 12-6 12-7 12-8 12-9 12-10 12-11 12-12 12-13 12-14 12-15 12-16 12-17 12-18 12-19 12-20 12-21 12-22 12-23 12-24 12-25 12-26 12-27 12-28 12-29 12-30 12-31 12-32 12-33 12-34 12-35 12-36 12-37 12-38 12-39 12-40 12-41 12-42 12-43 12-44 12-45 12-46 Offset 02h: DID--Device Identification Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 626 Offset 04h: PCI COMMAND Register (PCI Express*--B0:D28:F0/F1/F2/F3)............... 627 Offset 06h: PCI Status Register (PCI Express*--B0:D28:F0/F1/F2/F3)..................... 628 Offset 08h: RID--Revision Identification Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 629 Offset 09h: Programming Interface Register (PCI Express*--B0:D28:F0/F1/F2/F3) ... 630 Offset 0Ah: Sub Class Code Register (PCI Express*--B0:D28:F0/F1/F2/F3).............. 630 Offset 0Bh: Base Class Code Register (PCI Express*--B0:D28:F0/F1/F2/F3) ............ 631 Offset 0Ch: Cache Line Size Register (PCI Express*--B0:D28:F0/F1/F2/F3) ............. 631 Offset 0Dh: Primary Latency Timer Register (PCI Express*--B0:D28:F0/F1/F2/F3) ... 632 Offset 0Eh: Header Type Register (PCI Express*--B0:D28:F0/F1/F2/F3).................. 632 Offset 18h: Bus Number Register (PCI Express*--B0:D28:F0/F1/F2/F3) .................. 633 Offset 1Bh: Secondary Latency Timer (PCI Express*--B0:D28:F0/F1/F2/F3) ............ 633 Offset 1Ch: IOBL--I/O Base and Limit Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 634 Offset 1Eh: Secondary Status Register (PCI Express*--B0:D28:F0/F1/F2/F3)........... 635 Offset 20h: Memory Base and Limit Register (PCI Express*--B0:D28:F0/F1/F2/F3)... 636 Offset 24h: Prefetchable Memory Base and Limit Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 637 Offset 28h: PMBU32--Prefetchable Memory Base Upper 32 Bits Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 637 Offset 28h: PMBU32--Prefetchable Memory Base Upper 32 Bits Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 638 Offset 34h: Capabilities List Pointer Register (PCI Express*--B0:D28:F0/F1/F2/F3) ... 638 Offset 3Ch: Interrupt Information Register (PCI Express*--B0:D28:F0/F1/F2/F3) ..... 639 Offset 3Eh: Bridge Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) ............... 639 Offset 40h: Capabilities List Register (PCI Express*--B0:D28:F0/F1/F2/F3).............. 641 Offset 42h: PCI Express* Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 641 Offset 44h: Device Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/F3) ......... 642 Offset 48h: Device Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) ............... 643 Offset 4Ah: Device Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................ 644 Offset 4Ch: Link Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/F3) ............ 645 Offset 50h: Link Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) .................. 647 Offset 52h: Link Status Register (PCI Express*--B0:D28:F0/F1/F2/F3).................... 648 Offset 54h: Slot Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/F3) ............. 649 Offset 58h: Slot Control Register (PCI Express*--B0:D28:F0/F1/F2/F3)................... 650 Offset 5Ah: Slot Status Register (PCI Express*--B0:D28:F0/F1/F2/F3).................... 651 Offset 5Ch: Root Control Register (PCI Express*--B0:D28:F0/F1/F2/F3).................. 652 Offset 60h: Root Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................... 653 Offset 64h: Device Capabilities 2 Register (PCI Express*--B0:D28:F0/F1/F2/F3) ...... 653 Offset 68h: Device Control 2 Register (PCI Express*--B0:D28:F0/F1/F2/F3) ............ 654 Offset 70h: Link Control 2 Register (PCI Express*--B0:D28:F0/F1/F2/F3)................ 655 Offset 80h: Message Signaled Interrupt Identifiers Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 655 Offset 82h: Message Signaled Interrupt Message Control Register (PCI Express*-- B0:D28:F0/F1/F2/F3) ....................................................................................... 656 Offset 84h: Message Signaled Interrupt Message Address Register (PCI Express*-- B0:D28:F0/F1/F2/F3) ....................................................................................... 657 Offset 88h: Message Signaled Interrupt Message Data Register (PCI Express*-- B0:D28:F0/F1/F2/F3) ....................................................................................... 657 Offset 90h: Subsystem Vendor Capability Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 658 Offset 94h: Subsystem Vendor Identification Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 658 Offset A0h: Power Management Capability Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 659 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 61 12-47 12-48 12-49 12-50 12-51 12-52 12-53 12-54 12-55 12-56 12-57 12-58 12-59 12-60 12-61 12-62 12-63 12-64 12-65 12-66 12-67 13-1 13-2 13-3 13-4 13-5 13-6 13-7 14-1 14-2 14-3 14-4 14-5 14-6 14-7 14-8 Offset A2h: PCI Power Management Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 659 Offset A4h: PCI Power Management Control and Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 660 Offset D4h: Miscellaneous Port Configuration Register 2 (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 661 Offset D8h: Miscellaneous Port Configuration Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 662 Offset DCh: SMI/SCI Status Register (PCI Express*--B0:D28:F0/F1/F2/F3).............. 665 Offset E1h: Root Port Dynamic Clock Gating Enable (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 666 Offset E8h: PCI Express* Configuration Register 1 (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 667 Offset 104h: Uncorrectable Error Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 667 Offset 108h: Uncorrectable Error Mask (PCI Express*--B0:D28:F0/F1/F2/F3) ........... 669 Offset 10Ch: Uncorrectable Error Severity (PCI Express*--B0:D28:F0/F1/F2/F3)....... 671 Offset 110h: Correctable Error Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 672 Offset 114h: Correctable Error Mask Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 673 Offset 118h: Advanced Error Capabilities and Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 674 Offset 130h: Root Error Status Register (PCI Express*--B0:D28:F0/F1/F2/F3).......... 675 Offset 180h: Root Complex Topology Capability List Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 676 Offset 184h: Element Self Description Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 676 Offset 190h: Upstream Link Description Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 677 Offset 198h: Upstream Link Base Address Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 678 Offset 300h: PCI Express* Configuration Register 2 (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 678 Offset 318h: PCI Express* Extended Test Mode Register (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 679 OFfset 324h: PCI Express* Configuration Register 1 (PCI Express*--B0:D28:F0/F1/F2/F3) ................................................................. 679 Memory Mapped Registers ................................................................................. 680 Offset 00h: General Capabilities and Identification Register .................................... 681 Offset 010h: General Configuration Register......................................................... 682 Offset 020h: General Interrupt Status Register ..................................................... 683 Offset 0F0h: Main Counter Value Register ............................................................ 684 Offset 100h: Timer n Configuration and Capabilities Register .................................. 686 Offset 108h: Timer n Comparator Value Register .................................................. 690 Serial Peripheral Interface (SPI) Register Address Map .......................................... 691 Offset 00h: BIOS Flash Primary Region Register (SPI Memory Mapped Configuration Registers)........................................................................................................ 693 Offset 04h: Hardware Sequencing Flash Status Register (SPI Memory Mapped Configuration Registers) .................................................................................... 694 Offset 06h: Hardware Sequencing Flash Control Register SPI Memory Mapped Configuration Registers) .................................................................................... 696 Offset 08h: Flash Address Register (SPI Memory Mapped Configuration Registers) .... 697 Offset 14h: Flash Data [N] Register (SPI Memory Mapped Configuration Registers) ... 698 Offset 50h: Flash Regions Access Permissions Register (SPI Memory Mapped Configuration Registers) .................................................................................... 699 Offset 54h: Flash Region 0 (Flash Descriptor) Register (SPI Memory Mapped Configuration Registers) .................................................................................... 700 Intel(R) Communications Chipset 89xx Series - Datasheet 62 October 2012 Order Number: 327879-001US Contents 14-9 14-10 14-11 14-12 14-13 14-14 14-15 14-16 14-17 14-18 14-19 14-20 14-21 14-22 14-23 14-24 14-25 14-26 14-27 14-28 14-29 15-1 15-2 15-3 15-4 15-5 15-6 15-7 15-8 15-9 15-10 15-11 15-12 15-13 15-14 15-15 15-16 15-17 15-18 Offset 58h: Offset 00h: Flash Region 1 (BIOS Descriptor) Register (SPI Memory Mapped Configuration Registers).................................................................................... 700 Offset 5Ch: Flash Region 2 (Intel ME) Register (SPI Memory Mapped Configuration Registers) ....................................................................................................... 701 Offset 60h: Flash Region 3 Register (SPI Memory Mapped Configuration Registers) ... 701 Offset 64h: Flash Region 4 (Platform Data) Register (SPI Memory Mapped Configuration Registers).................................................................................... 702 Offset 74h: Protected Range 0 Register (SPI Memory Mapped Configuration Registers) ....................................................................................................... 702 Offset 78h: Protected Range 1 Register (SPI Memory Mapped Configuration Registers) ....................................................................................................... 703 Offset 7Ch: Protected Range 2 Register (SPI Memory Mapped Configuration Registers) ....................................................................................................... 704 Offset 80h: Protected Range 3 Register (SPI Memory Mapped Configuration Registers) ....................................................................................................... 705 Offset 84h: Protected Range 4 Register (SPI Memory Mapped Configuration Registers) ....................................................................................................... 706 Offset 90h: Software Sequencing Flash Status Register (SPI Memory Mapped Configuration Registers).................................................................................... 707 Offset 91h: Software Sequencing Flash Control Register (SPI Memory Mapped Configuration Registers).................................................................................... 708 Offset 94h: Prefix Opcode Configuration Register (SPI Memory Mapped Configuration Registers) ....................................................................................................... 710 Offset 96h: Opcode Type Configuration Register (SPI Memory Mapped Configuration Registers) ....................................................................................................... 711 Offset 98h: Opcode Menu Configuration Register (SPI Memory Mapped Configuration Registers) ....................................................................................................... 712 Offset A0h: BIOS Base Address Configuration Register (SPI Memory Mapped Configuration Registers).................................................................................... 713 Offset B0h: Flash Descriptor Observability Control Register (SPI Memory Mapped Configuration Registers).................................................................................... 714 Offset B4h: Flash Descriptor Observability Data Register SPI Memory Mapped Configuration Registers).................................................................................... 714 Offset C0h: Additional Flash Control Register (SPI Memory Mapped Configuration Registers) ....................................................................................................... 715 Offset C4h: Host Lower Vendor Specific Component Capabilities Register (SPI Memory Mapped Configuration Registers) ........................................................................ 716 Offset C8h: Host Upper Vendor Specific Component Capabilities Register (SPI Memory Mapped Configuration Registers) ........................................................................ 718 Offset D0h: Flash Partition Boundary (SPI Memory Mapped Configuration Registers) ....................................................................................................... 720 Summary of UART Registers in I/O Space (DLAB=0) ............................................. 721 Summary of UART Registers in I/O Space (DLAB=1) ............................................. 721 Summary of UART Timer Registers in I/O Space ................................................... 722 Internal Register Descriptions ........................................................................... 722 Offset 00h: RBR--Receive Buffer Register ........................................................... 723 Offset 00h: THR--Transmit Holding Register ...................................................... 723 Offset 01h: IER--Interrupt Enable Register ........................................................ 724 Interrupt Conditions ........................................................................................ 725 Offset 02h: IIR--Interrupt Identification Register ................................................ 725 Interrupt Identification Register Decode ............................................................. 726 Offset 02h: FCR--FIFO Control Register ............................................................. 727 Offset 03h: LCR--Line Control Register .............................................................. 728 Offset 04h: MCR--Modem Control Register ........................................................ 730 Offset 05h: LSR--Line Status Register ............................................................... 732 Offset 06h: MSR--Modem Status Register .......................................................... 735 Offset 07h: SCR--Scratchpad Register .............................................................. 736 Offset 00h: DLL--Programmable Baud Rate Generator Divisor Latch Register Low ... 736 Offset 01h: DLH--Programmable Baud Rate Generator Divisor Latch Register High ... 737 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 63 15-19 15-20 15-21 15-22 15-23 15-24 15-25 15-26 15-27 15-28 15-29 15-30 15-31 15-32 15-33 15-34 16-1 16-2 16-3 16-4 16-5 16-6 16-7 16-8 16-9 16-10 16-11 16-12 16-13 16-14 16-15 16-16 16-17 16-18 16-19 16-20 16-21 16-22 16-23 16-24 16-25 16-26 16-27 16-28 16-29 16-30 16-31 16-32 16-33 16-34 16-35 16-36 17-1 17-2 17-3 17-4 17-5 17-6 Summary of Watchdog Timer Registers in I/O Space ............................................. 738 Offset 00h: PV1R0--Preload Value 1 Register 0 ................................................... 738 Offset 01h: PV1R1--Preload Value 1 Register 1 ................................................... 739 Offset 02h: PV1R2--Preload Value 1 Register 2 ................................................... 739 Offset 04h: PV2R0--Preload Value 2 Register 0 ................................................... 740 Offset 05h: PV2R1--Preload Value 2 Register 1 ................................................... 740 Offset 06h: PV2R2--Preload Value 2 Register 2 ................................................... 741 Offset 08h: GISR--General Interrupt Status Register ........................................... 741 Offset 0Ch: RR0--Reload Register 0 .................................................................. 742 Offset 0Dh: RR1--Reload Register 1 .................................................................. 742 Offset 10h: WDTCR--WDT Configuration Register ................................................ 743 Offset 18h: WDTLR--WDT Lock Register ............................................................ 744 Configuration Register Summary ....................................................................... 745 Logical Device 4 (Serial Port 1) .......................................................................... 748 Logical Device 5 (Serial Port 2) .......................................................................... 749 Logical Device 6 (Watch Dog Timer) ................................................................... 751 PCI Header (B0:D31:F7).................................................................................... 753 Offset 00h: ID--Identifiers ................................................................................. 753 Offset 04h: CMD--Command .............................................................................. 754 Offset 06h: DSTS--Device Status........................................................................ 755 Offset 08h: RID--Revision ID ............................................................................. 756 Offset 09h: CC--Class Code ............................................................................... 756 Offset 0Ch: CLS--Cache Line Size ....................................................................... 757 Offset 0Dh: MLT--Master Latency Timer............................................................... 757 Offset 0Eh: HTYPE--Header Type ........................................................................ 757 Offset 10h: TBAR--WDT Table Base Address ........................................................ 758 Offset 14h: PBAR--WDT PBA Base Address........................................................... 758 Offset 18h: CBAR--WDT CFG Base Address .......................................................... 759 Offset 2Ch: SSVID--Subsystem Vendor ID ........................................................... 759 Offset 2Eh: SSID Subsystem ID.......................................................................... 760 Offset 34h: CAP--Capabilities Pointer .................................................................. 760 PCI MSI-X Capability ......................................................................................... 760 Offset 80h: MID--MSI-X Capability ID ................................................................. 761 Offset 82h: MC--MSI-X Control & Status .............................................................. 761 Offset 84h: MT--MSI-X Table ............................................................................. 761 Offset 88h: MP--MSI-X Message PBA................................................................... 762 Table BAR Range .............................................................................................. 763 Offset 00h: WDT0MAW--WDT 0 Msg Add WARN.................................................... 764 Offset 08h: WDT0MDW--WDT 0 Msg Data WARN .................................................. 765 Offset 0Ch: WDT0VCW--WDT 0 Vector Ctrl WARN ................................................. 765 Offset 18h: WDT0MDR--WDT 0 Msg Data RESET................................................... 766 Offset 1Ch: WDT0VCR--WDT 0 Vector Ctrl RESET ................................................. 766 PBA BAR Range ................................................................................................ 767 Offset 00h: WDT_PBA0--Pending Bit Array 0 ........................................................ 768 Offset 04h: WDT_PBA1--Pending Bit Array 1 ........................................................ 769 Offset 08h: WDT_PBA2--Pending Bit Array 2 ........................................................ 770 Offset 0Ch: WDT_PBA3--Pending Bit Array 3 ........................................................ 771 CFG BAR Range ................................................................................................ 772 Offset 00h: WDT0COUNT--WDT Count Register .................................................... 772 Offset 04h: WDT0CMD--WDT Cmd/Status Register ............................................... 773 Offset 200h: WDT_PSCALE ................................................................................ 773 Offset 204h: WDT_GBLCFG ................................................................................ 774 Bus 0, Device 31, Function 6: Summary of Thermal Reporting PCI Configuration Registers ......................................................................................................... 776 Offset 00h: ID: Vendor Identification Register ...................................................... 778 Offset 02h: DID--Device Identification Register .................................................... 778 Offset 04h: ID--Command Register..................................................................... 778 Offset 06h: ID--STS--Status Register.................................................................. 779 Offset 08h: RID--Revision Identification Register .................................................. 780 Intel(R) Communications Chipset 89xx Series - Datasheet 64 October 2012 Order Number: 327879-001US Contents 17-7 17-8 17-9 17-10 17-11 17-12 17-13 17-14 17-15 17-16 17-17 17-18 17-19 17-20 17-21 17-22 17-23 17-24 17-25 17-26 17-27 17-28 17-29 17-30 17-31 17-32 17-33 17-34 17-35 17-36 17-37 17-38 17-39 17-40 17-41 17-42 17-43 17-44 17-45 17-46 17-47 17-48 17-49 17-50 17-51 17-52 17-53 17-54 17-55 17-56 17-57 17-58 17-59 17-60 17-61 17-62 17-63 17-64 Offset 09h: PI--Programming Interface Register................................................... 781 Offset 0Ah: SCC--Sub Class Code Register .......................................................... 781 Offset 0Bh: BCC--Base Class Code Register ......................................................... 781 Offset 0Ch: Cache Line Size............................................................................... 782 Offset 0Dh: Latency Timer................................................................................. 782 Offset 0Eh: Header Type ................................................................................... 782 Offset 10h: Thermal Base.................................................................................. 783 Offset 14h: TBARH--Thermal Base High DWord .................................................... 783 Offset 2Ch: Subsystem Vendor ID ...................................................................... 784 Offset 2Eh: Subsystem ID ................................................................................. 785 Offset 34h: Capabilities Pointer .......................................................................... 785 Offset 3Ch: Interrupt Line ................................................................................. 786 Offset 3Dh: Interrupt Pin................................................................................... 786 Offset 40h: BIOS Assigned Thermal Base Address ................................................ 787 Offset 44h: BIOS Assigned Thermal Base High DWord ........................................... 787 Offset 50h: PCI Power Management Capability ID ................................................. 788 Offset 52h: Power Management Capabilities......................................................... 788 Offset 54h: Power Management Control And Status .............................................. 789 Offset 80h: MID--Message Signaled Interrupt Identifiers ....................................... 790 Offset 82h: MC--Message Signaled Interrupt Message Control................................ 790 Offset 84h: MA--Message Signaled Interrupt Message Address ............................... 791 Offset 88h: MD--Message Signaled Interrupt Message Data ................................... 791 Offset 94h: CCTS0RD--Thermal Sensor 0 Remote Diode........................................ 791 Offset 96h: CCTS0SSI--Thermal Sensor 0 Sense Stage Inputs ............................... 792 Offset 98h: CCTS0DSAC0--Thermal Sensor 0 Delta Sigma ADC Control 0 ................ 793 Offset 9Ch: CCTS0DSAC1--Thermal Sensor 0 Delta Sigma ADC Control 1 ................ 793 Offset A0h: CCTS0C--Thermal Sensor 0 Calibration .............................................. 794 Offset A4h: CCTS0TL0--Thermal Sensor 0 Top Logic 0 .......................................... 794 Offset A8h: CCTS0TL1--Thermal Sensor 0 Top Logic 1 .......................................... 796 Offset ACh: CCTS0BP--Thermal Sensor 0 Bandgap Pins......................................... 797 Offset B4h: CCTS1RD--Thermal Sensor 1 Remote Diode........................................ 797 Offset B6h: CCTS1SSI--Thermal Sensor 1 Sense Stage Inputs ............................... 798 Offset B8h: CCTS1DSAC0--Thermal Sensor 1 Delta Sigma ADC Control 0 ................ 798 Offset BCh: CCTS1DSAC1--Thermal Sensor 1 Delta Sigma ADC Control 1 ................ 799 Offset C0h: CCTS1C--Thermal Sensor 1 Calibration .............................................. 799 Offset C4h: CCTS1TL0--Thermal Sensor 1 Top Logic 0 .......................................... 800 Offset C8h: CCTS1TL1--Thermal Sensor 1 Top Logic 1 .......................................... 801 Offset CCh: CCTS1BP--Thermal Sensor 1 Bandgap Pins......................................... 802 Offset F8h: MANID--Manufacturing/Process ........................................................ 802 Bus 0, Device 31, Function 6: Summary of Thermal Reporting Registers Mapped Through TBARB BAR ......................................................................................... 803 Offset 00h: TS0IU--Thermal Sensor 0 In Use ....................................................... 805 Offset 01h: TS0C--Thermal Sensor 0 Control ....................................................... 806 Offset 02h: TS0S--Thermal Sensor 0 Status ........................................................ 807 Offset 03h: TS0TR--Thermal Sensor 0 Thermometer Read..................................... 808 Offset 04h: TS0TTP--Thermal Sensor 0 Temperature Trip Point .............................. 808 Offset 08h: TS0CO--Thermal Sensor 0 Calibration Offset ....................................... 810 Offset 0Bh: TS0TTPC--Thermal Sensor 0 Temperature Trip Point Cont'd .................. 810 Offset 0Ch: TS0ES--Thermal Sensor 0 Error Status .............................................. 811 Offset 0Dh: TS0GPEN--Thermal Sensor 0 General Purpose Event Enables ................ 812 Offset 0Eh: TS0PC--Thermal Sensor 0 Policy Control............................................. 813 Offset 10h: CPEC--CPU Power Error Correction Data ............................................. 814 Offset 12h: C0TA--CPU0 Temperature Adjust....................................................... 814 Offset 16h: C1TA--CPU1Temperature Adjust ....................................................... 815 Offset 1Ah: MC--MPC Control............................................................................. 815 Offset 22h: Timestamp ..................................................................................... 816 Offset 24h: DIMMEN--DIMM Enable .................................................................... 817 Offset 30h: C0TV--CPU0 Temperature Value ........................................................ 818 Offset 32h: C1TV--CPU1 Temperature Value ........................................................ 819 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 65 17-65 17-66 17-67 17-68 17-69 17-70 17-71 17-72 17-73 17-74 17-75 17-76 17-77 17-78 17-79 17-80 17-81 17-82 17-83 17-84 17-85 17-86 17-87 17-88 17-89 17-90 17-91 17-92 17-93 17-94 17-95 17-96 17-97 18-1 18-2 18-3 18-4 19-1 19-2 19-3 19-4 19-5 19-6 19-7 19-8 19-9 19-10 19-11 19-12 19-13 20-1 20-2 20-3 20-4 20-5 Offset 34h: C0EV--CPU0 Energy Value ................................................................ 819 Offset 38h: C1EV--CPU1 Energy Value ................................................................ 819 Offset 3Fh: AE--Alert Enable .............................................................................. 820 Offset 40h: TS1IU--Thermal Sensor 1 In Use ....................................................... 821 Offset 41h: TS1C--Thermal Sensor 1 Control........................................................ 821 Offset 42h: TS1S--Thermal Sensor 1 Status......................................................... 823 Offset 43h: TS1TR--Thermal Sensor 1 Thermometer Read ..................................... 824 Offset 44h: TS1TTP--Thermal Sensor 1 Temperature Trip Point .............................. 824 Offset 48h: TS1CO--Thermal Sensor 1 Calibration Offset ....................................... 826 Offset 4Bh: TS1TTPC--Thermal Sensor 1 Temperature Trip Point Cont'd................... 827 Offset 4Ch: TS1ES--Thermal Sensor 1 Error Status ............................................... 827 Offset 4Dh: TS1GPEN--Thermal Sensor 1 General Purpose Event Enables ................ 829 Offset 4Eh: TS1PC--Thermal Sensor 1 Policy Control ............................................. 830 Offset 50h: HTS--HOST Turbo Status ................................................................. 831 Offset 56h: MTL--MCP Temperature Limit ............................................................ 831 Offset 58h: MTV--MCH Temperature Value........................................................... 832 Offset 60h: MCPTV--MCP Temperature Value........................................................ 832 Offset 64h: MMPC--Max MCH Power Clamp .......................................................... 832 Offset 66h: MMCPPC--Max MCP Power Clamp ....................................................... 833 Offset 82h: TS0PIEN--Thermal Sensor 0 PCI Interrupt Event Enables ...................... 833 Offset 83h: TS0LOCK--Thermal Sensor 0 Register Lock Controls ............................. 834 Offset 98h: STS--SMBus Turbo Status ................................................................. 834 Offset 9Ch: SEC--SMBus Event Clear .................................................................. 835 Offset A4h: TC3--Thermal Compares 3 ................................................................ 835 Offset A8h: TC1--Thermal Compares 1 ................................................................ 836 Offset ACh: TC2--Thermal Compares 2................................................................ 836 Offset B0h: DIMM0--CPU0 DIMM Data ................................................................. 837 Offset B4h: DIMM1--CPU1 DIMM Data ................................................................. 839 Offset B8h: DIMMID--DIMM ID ........................................................................... 840 Offset BCh: ECPCLAMP--EC Power Clamp Data ..................................................... 842 Offset C2h: TS1PIEN--Thermal Sensor 1PCI Interrupt Event Enables ....................... 842 Offset C3h: TS1LOCK--Thermal Sensor 1 Register Lock Controls ............................. 843 Offset D8h: ITV--Internal Temperature Values ..................................................... 843 MEI1 Configuration Registers Address Map (MEI1--B0:D22:F0)............................... 844 MEI1 MMIO Register Address Map ....................................................................... 855 MEI2 Configuration Registers Address Map (MEI2--B0:D22:F1)............................... 858 MEI2 MMIO Register Address Map ....................................................................... 867 PCIe* Commands Supported by RI...................................................................... 874 PCIe* EP Function Mapping ................................................................................ 874 EP Mapping of BARs to MMIOs ............................................................................ 875 PCIe* Error Sources.......................................................................................... 877 Advisory Error Cases ......................................................................................... 878 Device Specific Errors........................................................................................ 881 Supported PCIe* Transactions ............................................................................ 882 PCIe* Request Header Format for 32-bit Memory Addressing.................................. 883 PCIe* Request Header Format for 64-bit Memory Addressing.................................. 883 PCIe* Completion Header Format ....................................................................... 883 PCI Completion Request Generation .................................................................... 884 PCIe* Request Header Format for Messages ......................................................... 885 PCIe* Message Generation................................................................................. 885 EP PCI Configuration Registers ........................................................................... 889 Bus M, Device 0, Function 0: Summary of PCIe Intel(R) QuickAssist Configuration Registers ......................................................................................................... 891 Bus M, Device 0, Function 0: Summary of PCIe Intel(R) QuickAssist Registers Mapped Through PMISCRBAR+1A000h Memory BAR ......................................................... 894 Bus M, Function 0: Summary of PCIe Intel(R) QuickAssist Registers Mapped Through PMISCSBAR+1A000h Memory BAR...................................................................... 894 Bus M, Function 8+Index 1: Summary of PCIe Intel(R) QuickAssist Configuration Registers ......................................................................................................... 894 Intel(R) Communications Chipset 89xx Series - Datasheet 66 October 2012 Order Number: 327879-001US Contents 20-6 20-7 20-8 20-9 20-10 20-11 20-12 20-13 20-14 20-15 20-16 20-17 20-18 20-19 20-20 20-21 20-22 20-23 20-24 20-25 20-26 20-27 20-28 20-29 20-30 20-31 20-32 20-33 20-34 20-35 20-36 20-37 20-38 20-39 20-40 20-41 20-42 20-43 20-44 20-45 20-46 20-47 20-48 20-49 20-50 20-51 20-52 20-53 20-54 20-55 20-56 20-57 20-58 20-59 20-60 20-61 20-62 20-63 20-64 PVID--PF Vendor Identification Register .............................................................. 896 PDID--PF Device Identification Register............................................................... 896 PPCICMD--PF Device Command Register ............................................................. 897 PPCISTS--PF PCI Device Status Register ............................................................. 898 PRID--PF Revision ID Register ........................................................................... 900 PCC--PF Class Code Register ............................................................................. 900 PHDR--PF Header Type Register......................................................................... 901 PeSRAMLBAR--PF Memory Lower Base Address Register........................................ 901 PeQATUBAR--PF QAT Upper Base Address Register ............................................... 902 PMISCLBAR--PF Miscellaneous Lower Base Address Register .................................. 902 PMISCUBAR--PF Miscellaneous Upper Base Address Register.................................. 903 PETRINGCSRLBAR--PF Ring CSR Lower Base Address Register ............................... 903 PSVID--PF Subsystem Vendor ID Register ........................................................... 904 PSID--PF Subsystem ID Register........................................................................ 904 PCP--PF Capabilities Pointer Register .................................................................. 905 PIRQL--PF Interrupt Line Register ...................................................................... 905 PIRQP--PF Interrupt Pin Register ........................................................................ 906 SKU Register ................................................................................................... 906 SKU Register 2 ................................................................................................ 907 PMSI-X--PF Message Signalled Interrupt X Capability ID Register ........................... 908 PMSIXNCP--PF MSIX Next Capability Pointer Register............................................ 908 PMSIXCNTL--PF Message Signalled Interrupt X Control Register ............................. 909 PMSIXTBIR--PF MSI-X Table Offset & Table BIR Register ....................................... 909 PMSIXPBABIR--PF MSI-X Pending Bit Array & BIR Offset Register ........................... 910 PPMCAP--PF Power Management Capabilities ID Register....................................... 910 PPMCP--PF Power Management Next Capability Pointer Register ............................. 911 PPMC--PF Power Management Capabilities Register .............................................. 911 PPMCSR--PF Power Management Control and Status Register................................. 912 PPCID--PF PCI Express Capability Register .......................................................... 914 PPCP--PF PCI Express Next Capability Pointer Register .......................................... 914 PPCR--PF PCI Express Capabilities Register.......................................................... 915 PPDCAP--PF PCI Express Device Capabilities Register............................................ 915 PPDCNTL--PF PCI Express Device Control Register ................................................ 917 PPDSTAT--PF PCI Express Device Status Register ................................................. 918 PLCAPR--PF Link Capabilities Register ................................................................. 919 PLCNTLR--PF Link Control Register ..................................................................... 921 PLSR--PF Link Status Register............................................................................ 923 PDCAPR2--PF Device Capabilities 2 Register ........................................................ 924 PDCNTR2--PF Device Control 2 Register .............................................................. 925 PLCNTLR2--PF Link Control 2 Register................................................................. 926 PLSR2--PF Link Status 2 Register ....................................................................... 928 PMSICID--PF Message Signalled Interrupt Capability ID Register ............................ 928 PMSINCP--PF Message Signalled Interrupt Next Capability Pointer Register .............. 929 PMSICTL--PF Message Signalled Interrupt Control Register .................................... 929 PMSILADDR--PF Message Signalled Interrupt Lower Address Register ..................... 930 PMSIUADDR--PF Message Signalled Interrupt Upper Address Register ..................... 930 PMSIDATA--PF Message Signalled Interrupt Data Register ..................................... 931 PMSIMSK - PF Message Signalled Interrupt Mask Register...................................... 931 PMSIPND--PF Message Signalled Interrupt Pending Register .................................. 932 PPCIEAERCAPID--PF PCI Express AER Capability ID Register .................................. 932 PPAERUCS--PF PCI Express AER Uncorrectable Error Status Register....................... 933 PPAERUCM--PF PCI Express AER Uncorrectable Error Mask Register ........................ 934 PPAERUCSEV--PF PCI Express AER Uncorrectable Error Severity Register ................ 935 PPAERCS--PF PCI Express AER Correctable Error Register...................................... 936 PPAERCM--PF PCI Express AER Correctable Error Mask Register ............................. 937 PPAERCTLCAP--PF PCI Express AER Control and Capability Register ........................ 938 PPAERHDRLOG0--PF PCI Express AER Header Log 0 Register ................................. 938 PPAERHDRLOG1--PF PCI Express AER Header Log 1 Register ................................. 939 PPAERHDRLOG2--PF PCI Express AER Header Log 2 Register ................................. 939 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 67 20-65 20-66 20-67 20-68 20-69 20-70 20-71 20-72 20-73 20-74 20-75 20-76 20-77 20-78 20-79 20-80 20-81 20-82 20-83 20-84 20-85 20-86 20-87 20-88 20-89 20-90 20-91 20-92 20-93 20-94 20-95 20-96 20-97 20-98 20-99 20-100 20-101 20-102 20-103 20-104 20-105 20-106 20-107 20-108 20-109 20-110 20-111 20-112 20-113 20-114 20-115 20-116 20-117 20-118 20-119 20-120 20-121 20-122 20-123 PPAERHDRLOG3--PF PCI Express AER Header Log 3 Register.................................. 940 PARIDHDR--PF Alternative Routing ID Capability Header........................................ 941 PFARICAP--PF ARI Capabilities Register ............................................................... 941 PARIDCTL--PF Alternative Routing ID Control Register........................................... 942 PSRIOVCAPID--PF SR IOV Capability ID Register .................................................. 942 PSRIOVCVNC--PF SRIOV Capability Version and Next Capability Pointer Register....... 943 PSRIOVCAP--PF SRIOV Capabilities Register......................................................... 944 PSRIOVCS--PF SRIOV Control and Status Register ................................................ 944 PSRIOVMTOTINI--PF SRIOV Initial and Total VFs Register ...................................... 945 PSRIOVNUMVF--PF SRIOV Number of VFs Register................................................ 945 PSRIOVFVFO--PF SRIOV First VF Offset Register ................................................... 946 PSRIOVVFS--PF SRIOV VF Stride Register ............................................................ 946 PSRIOVFDID--PF SRIOV VF Device ID Register ..................................................... 947 PSRIOVPAGESIZE--PF SRIOV Supported Page Size Register ................................... 947 PSRIOVSYSPS--PF SRIOV System Page Size Register ............................................ 948 PSRIOVLBAR0--SRIOV Lower BAR0 Register ........................................................ 949 PSRIOVUBAR0--SRIOV Upper BAR0 Register ........................................................ 950 PSRIOVLBAR1--SRIOV Lower BAR1 Register ........................................................ 951 PSRIOVUBAR1--SRIOV Upper BAR1 Register ........................................................ 952 PSRIOVVFMA--PF SRIOV VF Migration Array Register ............................................ 952 VVID[0:15]--VF Vendor Identification Register ..................................................... 953 VDID[0:15]--VF Device Identification Register...................................................... 953 VPCICMD[0:15]--VF Device Command Register .................................................... 954 VPCISTS[0:15]--VF PCI Device Status Register .................................................... 955 VRID[0:15]--VF Revision ID Register .................................................................. 956 VCC[0:15]--VF Class Code Register .................................................................... 956 VHDR[0:15]--VF Header Type Register................................................................ 956 VSVID[0:15]--VF Subsystem Vendor ID Register .................................................. 957 VSID[0:15]--VF Subsystem ID Register............................................................... 957 VCP[0:15]--VF Capabilities Pointer Register ......................................................... 958 VIRQL[0:15]--VF Interrupt Line Register ............................................................. 958 VIRQP[0:15]--VF Interrupt Pin Register ............................................................... 958 VPCID[0:15]--VF PCI Express Capability ID Register ............................................. 959 VPCP[0:15]--VF PCI Express Next Capability Pointer Register ................................. 959 VPCR[0:15]--VF PCI Express Capabilities Register................................................. 960 VPDCAP[0:15]--VF PCI Express Device Capabilities Register................................... 960 VPDC[0:15]--VF PCI Express Device Control Register ............................................ 961 VPDS[0:15]--VF PCI Express Device Status Register ............................................. 962 VLCR[0:15]--VF Link Capabilities Register............................................................ 963 VLCNTLR[0:15]--VF Link Control Register ............................................................ 965 VLSR[0:15]--VF Link Status Register................................................................... 966 DCAPR2[0:15]--Device Capabilities 2 Register...................................................... 966 VMSICID[0:15]--Message Signalled Interrupt Capability ID Register........................ 967 VMSINCP[0:15]--Message Signalled Interrupt Next Capability Pointer Register ......... 967 VMSICTL[0:15]--Message Signalled Interrupt Control Register ............................... 968 VMSILADDR[0:15]--Message Signalled Interrupt Lower Address Register ................. 968 VMSIUADDR[0:15]--Message Signalled Interrupt Upper Address Register ................ 969 VMSIDATA[0:15]--Message Signalled Interrupt Data Register................................. 969 VMSIMSK--VF Message Signalled Interrupt Mask Register ...................................... 969 VMSIPND--VF Message Signalled Interrupt Pending Register .................................. 970 VPCIEAERCAPID[0:15]--VF PCI Express AER Capability ID Register ......................... 970 VPAERUCS[0:15]--VF PCI Express AER Uncorrectable Error Status Register.............. 971 VPAERUCM[0:15]--VF PCI Express AER Uncorrectable Error Mask Register ............... 972 VPAERUCSEV[0:15]--VF PCI Express AER Uncorrectable Error Severity Register ....... 974 VPAERCS[0:15]--VF PCI Express AER Correctable Error Status Register ................... 975 VPAERCM[0:15]--VF PCI Express AER Correctable Error Mask Register .................... 976 VPAERCTLCAP[0:15]--VF PCI Express AER Control and Capability Register ............... 977 VPAERHDRLOG0[0:15]--VF PCI Express AER Header Log 0 Register ........................ 978 VPAERHDRLOG1[0:15]--VF PCI Express AER Header Log 1 Register ........................ 979 Intel(R) Communications Chipset 89xx Series - Datasheet 68 October 2012 Order Number: 327879-001US Contents 20-124 20-125 20-126 20-127 20-128 20-129 20-130 20-131 20-132 20-133 20-134 20-135 21-1 21-2 21-3 21-4 21-5 21-6 21-7 21-8 21-9 22-1 22-2 22-3 22-4 22-5 22-6 22-7 22-8 22-9 22-10 22-11 22-12 22-13 22-14 22-15 22-16 22-17 22-18 22-19 22-20 22-21 22-22 22-23 23-1 23-2 23-3 24-1 24-2 24-3 24-4 24-5 24-6 24-7 24-8 24-9 24-10 24-11 24-12 VPAERHDRLOG2[0:15]--VF PCI Express AER Header Log 2 Register........................ 979 VPAERHDRLOG3[0:15]--VF PCI Express AER Header Log 3 Register........................ 980 VARIDHDR[0:15]--VF Alternative Routing ID Capability Header .............................. 980 VFARICAP[0:15]--VF ARI Capabilities Register ..................................................... 981 VARIDCTL[0:15]-- VF Alternative Routing ID Control Register ................................ 981 ERRSOU2--Error Source Register 2..................................................................... 982 ERRMSK2--Error Source Mask Register 2............................................................. 982 SINTPF--Signal Raw PF Interrupt Register ........................................................... 983 SMIAPF--Signal IA PF Interrupt Mask Register...................................................... 983 GBECFGMMIOV - GBE Configuration and MMIO Valid Register................................. 984 SINTGBE[0:3]--Signal Raw PF Interrupt for GBE Register ...................................... 985 SMIAGBE[0:3]--Signal IA PF Interrupt Mask GBE Register ..................................... 985 Glossary ......................................................................................................... 988 GbE Controller Features .................................................................................... 993 GbE Controller Network Features........................................................................ 993 GbE Controller Host Interface Features................................................................ 994 GbE Controller LAN Functions Features................................................................ 994 GbE Controller LAN Performance Features ........................................................... 995 GbE Controller Virtualization Features ................................................................. 995 Transmit Data Flow .......................................................................................... 999 Receive Data Flow ...........................................................................................1000 LAN Traffic Attributes.......................................................................................1002 SMBus ARA Cycle Format .................................................................................1005 Asynchronous Notify Command Format ..............................................................1005 Direct Receive - LAN Packet Receive Transaction Format ......................................1006 Direct Receive - Status Change Transaction Format .............................................1006 Flags in Transmit Packet Transactions ................................................................1008 Unique Device Identifier (UDID) ........................................................................1012 Dynamic and Persistent Address, PEC Support Bit................................................1012 Version/Revision: UDID Version 1, Silicon Revision ..............................................1012 Silicon Revision ID...........................................................................................1012 Vendor Specific ID...........................................................................................1013 EEPROM Structure ...........................................................................................1013 EEPROM Auto-Load Sequence ...........................................................................1014 Notes on EEPROM Words ..................................................................................1019 Small Resource Structure .................................................................................1019 Large Resource Structure .................................................................................1020 VPD Structure.................................................................................................1020 Encoding on LINK_MODE Field for Each Mode......................................................1022 Ordered Sets ..................................................................................................1026 802.3z Advertised Base Page Mapping ...............................................................1028 802.3X Packet Format......................................................................................1032 Forwarding of PAUSE Packet to Host (DPF Bit).....................................................1033 Transfer of Non-PAUSE Control Packets to Host (PMCF Bit) ...................................1033 Notes to Power-Up Timing Diagram ...................................................................1039 GbE Controller Reset Effects - Common Resets....................................................1044 GbE Controller Reset Effects - Per Function Resets ...............................................1045 EEPROM Top Level Partitioning ..........................................................................1057 Common and Lan Port 0 EEPROM Map................................................................1057 LAN Ports 1, 2, and 3 EEPROM Map ..................................................................1059 Supported MAC Connection Types .....................................................................1061 CSR Auto Configuration Structure Format ...........................................................1071 CSR Auto Configuration Power-Up Structure Format.............................................1073 EICT Information Structure Format ....................................................................1074 PHYAuto Configuration Structure Format ............................................................1075 HY Configuration Section Length .......................................................................1075 Block CRC8 structure .......................................................................................1075 PHY Number and PHY Address ..........................................................................1076 PHY Data .......................................................................................................1076 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 69 24-13 24-14 25-1 25-2 25-3 25-4 25-5 25-6 25-7 25-8 25-9 25-10 26-1 26-2 26-3 26-4 26-5 26-6 26-7 26-8 26-9 26-10 26-11 26-12 26-13 26-14 26-15 26-16 26-17 26-18 26-19 26-20 26-21 26-22 26-23 26-24 26-25 26-26 26-27 26-28 26-29 26-30 26-31 26-32 26-33 26-34 26-35 26-36 26-37 26-38 26-39 26-40 26-41 26-42 26-43 26-44 26-45 26-46 26-47 PT LAN Configuration Structure Format.............................................................. 1080 Alternate MAC Address Block............................................................................ 1093 Power Up (Off to Dup to D0u to D0a) ................................................................ 1099 Transition From D0a to D3 and Back Without PCIE_EP_RST# ............................... 1100 Transition From D0a to D3 and Back With PCIE_EP_RST# .................................... 1101 Transition From D0a to Dr and Back Without Transition to D3 ............................... 1102 Magic Packet Structure .................................................................................... 1106 ARP Packet Structure and Processing................................................................. 1107 IPv4 Packet Structure and Processing ................................................................ 1107 IPv6 Packet Structure and Processing ................................................................ 1108 IPX Diagnostic Responder Request Packet Structure and Processing ...................... 1109 IPX Packet Structure and Processing ................................................................. 1109 IPv4 ............................................................................................................. 1122 IPv6 ............................................................................................................. 1122 Legacy Receive Descriptor (RDESC) Layout ........................................................ 1129 Receive Status (RDESC.STATUS) Layout ............................................................ 1130 Receive Status Bits ......................................................................................... 1130 IPCS, L4CS, and UDPCS .................................................................................. 1131 RXE, IPE and L4E............................................................................................ 1131 VLAN Tag Field Layout (for 802.1q Packet)......................................................... 1132 RDESC Descriptor Read Format ........................................................................ 1132 RDESC Descriptor Write-Back Format ................................................................ 1133 RSS Type ...................................................................................................... 1133 Packet Type LSB Bits (11:10) ........................................................................... 1134 Receive Status (RDESC.STATUS) Layout of Last Descriptor .................................. 1135 Receive Status (RDESC.STATUS) Layout of Non-Last Descriptor ............................ 1135 Receive Errors (RDESC.ERRORS) Layout ............................................................ 1136 GBE Controller Split/Replicated Header Behavior ................................................. 1142 Timestamp Layout in Buffer ............................................................................. 1143 Supported Receive Checksum Capabilities .......................................................... 1145 Typical IPv6 Extended Header Format (Traditional Representation)........................ 1146 Header Type Encoding and Lengths ................................................................... 1147 Descriptor Fields............................................................................................. 1148 SCTP Header.................................................................................................. 1148 Transmit Descriptor (TDESC) Fetch Layout - Legacy Mode.................................... 1151 Transmit Descriptor (TDESC) Write-Back Layout - Legacy Mode ............................ 1151 Transmit Command (TDESC.CMD) Layout .......................................................... 1152 VLAN Tag Insertion Decision Table .................................................................... 1153 Transmit Status (TDESC.STA) Layout ................................................................ 1153 VLAN Field (TDESC.VLAN) Layout...................................................................... 1154 Transmit Context Descriptor (TDESC) Layout - (Type = 0010b) ............................ 1154 MACLEN Values .............................................................................................. 1154 Transmit Command (TDESC.TUCMD) Layout ...................................................... 1155 Valid Field in Context vs. Required Offload ......................................................... 1156 Advanced Transmit Data Descriptor (TDESD) Layout - (Type = 0011b) .................. 1156 Advanced Tx Descriptor Write-back Format ........................................................ 1156 Transmit Data (TDESD.MAC) Layout.................................................................. 1157 Transmit Data (TDESD.DCMD) Layout ............................................................... 1157 Transmit Data (TDESD.POPTS) Layout............................................................... 1158 Write Back Options For Large Send ................................................................... 1166 TCP/IP or UDP/IP Packet Format Sent by Host .................................................... 1167 TCP/IP or UDP/IP Packet Format Sent by the Controller ....................................... 1167 TCP Partial Pseudo-Header Sum for IPv4............................................................ 1168 TCP Partial Pseudo-Header Sum for IPv6............................................................ 1168 Supported Transmit Checksum Capabilities ........................................................ 1169 Conditions for Checksum Off Loading ................................................................ 1169 Checksum Per Packet Type .............................................................................. 1175 Cause Allocation in the IVAR Registers - MSI and Legacy Mode ............................. 1177 Cause Allocation in the IVAR Registers .............................................................. 1178 Intel(R) Communications Chipset 89xx Series - Datasheet 70 October 2012 Order Number: 327879-001US Contents 26-48 26-49 26-50 26-51 26-52 26-53 26-54 26-55 26-56 26-57 26-58 26-59 26-60 26-61 26-62 26-63 26-64 26-65 26-66 26-67 26-68 26-69 27-1 27-2 27-3 27-5 27-6 27-7 27-4 27-8 27-9 28-1 28-2 28-3 28-4 28-5 28-6 28-7 28-8 28-9 28-10 28-11 28-12 28-13 28-14 28-15 28-16 28-17 28-18 28-19 28-20 28-21 28-22 28-23 28-24 28-25 28-26 Interrupt Registers - Legacy Mode .....................................................................1178 Interrupt Registers - Extended Mode..................................................................1179 Settings for Different Interrupt Modes ................................................................1182 Comparing Packets..........................................................................................1188 TCI Bit Ordering..............................................................................................1189 Receive Processing in Double VLAN Mode ...........................................................1191 Mode Encoding for LED Outputs ......................................................................1192 Legacy DCA Systems .......................................................................................1194 DCA Systems..................................................................................................1194 Storm Control Interval by Speed .......................................................................1204 Chronological Order of Events for Sync and Path Delay.........................................1207 V1 and V2 PTP Message Structure .....................................................................1212 PTP Message Over Layer 2................................................................................1214 PTP Message Over Layer 4................................................................................1214 Message Decoding for V1 (Control Field at Offset 32) ...........................................1214 Message Decoding for V2 (Message Id Field at Offset 0) .......................................1214 IEEE 802.3 Mandatory Package Statistics ...........................................................1215 IEEE 802.3 Recommended Package Statistics......................................................1215 IEEE 802.3 Optional Package Statistics...............................................................1216 Microsoft* OID_GEN_STATISTICS .....................................................................1216 RMON Statistics ..............................................................................................1217 Linux net_device_stats ....................................................................................1218 SMBus Write Transactions ................................................................................1222 TCO LAN Packet Status Data.............................................................................1229 Packet Status Info ...........................................................................................1229 Security errors................................................................................................1229 Packet Type....................................................................................................1230 MNG Status ....................................................................................................1230 Error Status Info .............................................................................................1230 Registers Written by the BMC............................................................................1239 Exclusive Traffic Behavior .................................................................................1244 Address Space Regions ....................................................................................1246 IOADDR and IODATA in I/O Address Space.........................................................1247 IOADDR and IODATA in Configuration Address Space...........................................1248 GbE Controller Register Field Attributes ..............................................................1250 Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR0[03]B:0:1+ Index 1 Registers ............................................................................................1252 Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR3[03]B:0:1+ Index 1 Registers ............................................................................................1261 Register Summary...........................................................................................1262 MSI-X Register Summary .................................................................................1264 Device Control Register--CTRL [0:3] (0x00000; R/W) ..........................................1264 Device Status Register-- STATUS[0:3] (0x0008; R) .............................................1268 Extended Device Control Register--CTRL_EXT [0:3] (0x0018; R/W) .......................1269 Mappings for SDI Pins Used as GPI ....................................................................1271 MDI Control Register--MDIC[0:3] (0x0020; R/W) ................................................1272 MDC/MDIO Configuration Register--MDICNFG [0:3] (0x0E04; R/W) .......................1273 Copper/Fiber Switch Control--CONNSW[0:3] (0x0034; R/W).................................1274 VLAN Ether Type--VET [0:3] (0x0038; R/W).......................................................1274 LED Control--LEDCTL [0:3](0x0E00; RW)...........................................................1275 Internal Receive Packet Buffer Size--IRPBS [0:3] (0x2404; RO) ............................1276 Internal Transmit Packet Buffer Size--ITPBS [0:3] (0x3404; RO)...........................1277 EEPROM Control Register--EEC [0:3] (0x0010; R/W) ...........................................1278 EEPROM Read Register--EERD [0:3] (0x0014; RW) .............................................1279 EEPROM Diagnostic--EEDIAG [0:3] (0x1038; RO)................................................1280 VPD Diagnostic Register--VPDDIAG [0:3] (0x1060; RO) .......................................1281 Management EEPROM Control Register--EEMNGCTL [0:3] (0x1010; RO) ................1282 Management EEPROM Read/Write Data--EEMNGDATA [0:3] (0x1014; RO) .............1283 Flow Control Address Low--FCAL [0:3] (0x0028; RO)...........................................1284 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 71 28-27 28-28 28-29 28-30 28-31 28-32 28-33 28-34 28-35 28-36 28-37 28-38 28-39 28-40 28-41 28-42 28-43 28-44 28-45 28-46 28-47 28-48 28-49 28-50 28-51 28-52 28-53 28-54 28-55 28-56 28-57 28-58 28-59 28-60 28-61 28-62 28-63 28-64 28-65 28-66 28-67 28-68 28-69 28-70 28-71 28-72 28-73 28-74 28-75 28-76 28-77 Flow Control Address High--FCAH [0:3] (0x002C; RO)......................................... 1285 Flow Control Type--FCT [0:3] (0x0030; R/W)..................................................... 1285 Flow Control Transmit Timer Value--FCTTV [0:3] (0x0170; R/W) .......................... 1286 Flow Control Receive Threshold Low--FCRTL0 [0:3] (0x2160; R/W)....................... 1287 Flow Control Receive Threshold High--FCRTH0 [0:3] (0x2168; R/W) ..................... 1288 Flow Control Refresh Threshold Value--FCRTV[0:3](0x2460; R/W) ........................ 1289 Flow Control Status--FCSTS0 [0:3] (0x2464; RO) ............................................... 1290 Function Active and Power State to MNG--FACTPS[0:3] (0x5B30; RO) ................... 1290 Mirrored Revision ID--MREVID[0:3] (0x5B64; R/W) ............................................ 1292 PCIE Control Extended Register--GCR_EXT[0:3] (0x5B6C; R/W) ........................... 1293 Software Semaphore--SWSM[0:3] (0x5B50; R/W) .............................................. 1294 Firmware Semaphore--FWSM[0:3] (0x5B54; R/WS)............................................ 1295 Firmware Synchronization--SW_FW_SYNC[0:3] (0x5B5C; RWS) ........................... 1297 Software Mailbox Write--SWMBWR[0:3] (0x5B04; R/W) ...................................... 1298 Software Mailbox 0--SWMB0[0:3] (0x5B08; RO) ................................................ 1298 Software Mailbox 1--SWMB1[0:3] (0x5B0C; RO) ................................................ 1298 Software Mailbox 2--SWMB2[0:3] (0x5B18; RO) ................................................ 1299 Software Mailbox 3--SWMB3[0:3] (0x5B1C; RO) ................................................ 1299 EICR Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0)............. 1300 EICR Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) ................... 1301 EICS Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0)............. 1301 SEICS Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) ................. 1302 EIMS Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) ............ 1302 EIMS Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) ................... 1303 EIMC Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) ............ 1304 EIMC Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1)................... 1304 Extended Interrupt Auto Clear--EIAC [0:3] (0x152C; R/W) .................................. 1305 IAM Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) .............. 1306 EIAM Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) ................... 1306 Interrupt Cause Read Register--ICR [0:3] (0x1500; RC/W1C) .............................. 1307 Interrupt Cause Set Register--ICS [0:3] (0x1504; WO) ....................................... 1309 Interrupt Mask Set/Read Register--IMS [0:3] (0x1508; R/W) ............................... 1311 Interrupt Mask Clear Register--IMC [0:3] (0x150C; WO) ..................................... 1313 Interrupt Acknowledge Auto Mask Register--IAM [0:3] (0x1510; R/W) .................. 1315 Interrupt Throttle--EITR [0:3][0:9] (0x1680 + 4*n [n = 0...9]; R/W).................... 1316 Interrupt Vector Allocation Registers--IVAR [0:3][0:3] (0x1700 + 4*n [n=0...3]; RW)......................................................................... 1317 Interrupt Vector Allocation Registers--MISC IVAR_MISC[0:3] (0x1740; RW)........... 1318 General Purpose Interrupt Enable--GPIE [0:3] (0x1514; RW) ............................... 1319 MSI-X Table Structure ..................................................................................... 1319 MSI-X PBA Structure ....................................................................................... 1320 MSIX Table Entry Lower Address--MSIXTADD [0:3][0:9] (BAR3: 0x0000 + 0x10*n [n=0...9]; R/W)......................................................... 1320 MSIX Table Entry Upper Address--MSIXTUADD [0:3][0:9] (BAR3: 0x0004 + 0x10*n [n=0...9]; R/W)......................................................... 1321 MSIX Table Entry Message--MSIXTMSG [0:3][0:9] (BAR3: 0x0008 + 0x10*n [n=0...9]; R/W)......................................................... 1321 MSIX Table Entry Vector Control--MSIXTVCTRL [0:3][0:9] (BAR3: 0x000C + 0x10*n [n=0...9]; R/W) ........................................................ 1322 MSIXPBA Bit Description--MSIXPBA [0:3] (BAR3: 0x2000; RO)............................. 1322 MSIX PBA Clear--PBACL [0:3] (0x5B68; R/W1C) ................................................ 1323 Receive Control Register--RCTL[0:3] (0x0100; R/W) ........................................... 1324 Split and Replication Receive Control--SRRCTL [0:3][0:7] (0xC00C + 0x40*n [n=0...7]; R/W) .................................................................. 1327 Split and Replication Receive Control--SRRCTL [0:3][1:7] (0xC00C + 0x40*n [n=0...7]; R/W) .................................................................. 1328 Packet Split Receive Type--PSRTYPE [0:3][0:7] (0x5480 + 4*n [n=0...7]; R/W)..... 1330 Replicated Packet Split Receive Type--RPLPSRTYPE [0:3] (0x54C0; R/W)............... 1331 Intel(R) Communications Chipset 89xx Series - Datasheet 72 October 2012 Order Number: 327879-001US Contents 28-78 28-79 28-80 28-81 28-82 28-83 28-84 28-85 28-86 28-87 28-88 28-89 28-90 28-91 28-92 28-93 28-94 28-95 28-96 28-97 28-98 28-99 28-100 28-101 28-102 28-103 28-104 28-105 28-106 28-107 28-108 28-109 28-110 28-111 28-112 28-113 28-114 28-115 28-116 28-117 28-118 28-119 28-120 28-121 28-122 Receive Descriptor Base Address Low--RDBAL [0:3][0:7] (0xC000 + 0x40*n [n=0...7]; R/W)...................................................................1332 Receive Descriptor Base Address High--RDBAH [0:3][0:7] (0xC004 + 0x40*n [n=0...7]; R/W)...................................................................1332 Receive Descriptor Ring Length--RDLEN [0:3][0:7] (0xC008 + 0x40*n [n=0...7]; R/W)...................................................................1333 Receive Descriptor Head--RDH [0:3][0:7] (0xC010 + 0x40*n [n=0...7]; RO) .........1333 Receive Descriptor Tail--RDT [0:3][0:7] (0xC018 + 0x40*n [n=0...7]; R/W) ..........1334 Receive Descriptor Control--RXDCTL [0:3][0:7] (0xC028 + 0x40*n [n=0...7]; R/W)...................................................................1334 Receive Queue Drop Packet Count--RQDPC [0:3][0:7] (0xC030 + 0x40*n [n=0...7]; RC/W).................................................................1336 Receive Checksum Control--RXCSUM [0:3] (0x5000; R/W)...................................1337 Receive Long Packet Maximum Length--RLPML [0:3] (0x5004; R/W) .....................1339 Receive Filter Control Register--RFCTL [0:3] (0x5008; R/W) .................................1339 Multicast Table Array--MTA [0:3][0:127] (0x5200 + 4*n [n=0...127]; R/W)...........1340 Receive Address Low 0--RAL0 [0:3][0:15] (0x5400 + 8*n [n=0...15]; 0x54E0 + 8*n [n=0...7]; R/W) ..................................1341 Receive Address Low 1--RAL1 [0:3][0:7] (0x54E0 + 8*n [n=0...7]; R/W) ..............1342 Receive Address High 0--RAH0 [0:3][0:15] (0x5404 + 8*n [n=0...15]; 0x54E04 + 8*n [n=0...7]; R/W).................................1343 Receive Address High 1--RAH1 [0:3][0:7] (0x54E4 + 8*n [n=0...7]; R/W).............1344 VLAN Filter Table Array--VFTA [0:3][0:127] (0x5600 + 4*n [n=0...127]; R/W).......1345 Multiple Receive Queues Command Register--MRQC [0:3](0x5818; R/W) ...............1346 RSS Random Key Register--RSSRK [0:3][0:9] (0x5C80 + 4*n [n=0...9]; R/W).......1348 Redirection Table--RETA [0:3][0:31] (0x5C00 + 4*n [n=0...31]; R/W) ..................1348 Immediate Interrupt RX--IMIR [0:3][0:7] (0x5A80 + 4*n [n=0...7]; R/W) .............1350 Immediate Interrupt Rx Ext.--IMIREXT [0:3][0:7] (0x5AA0 + 4*n [n=0...7]; R/W)........................................................................1351 2tuples Queue Filter--TTQF[0:3][0:7] (0x59E0 + 4*n[n=0..7]; R/W).....................1352 Immediate Interrupt Rx VLAN Priority--IMIRVP [0:3] (0x5AC0; R/W) .....................1352 SYN Packet Queue Filter--SYNQF [0:3] (0x55FC; RW)..........................................1353 EType Queue Filter--ETQF [0:3][0:7] (0x5CB0 + 4*n[n=0...7]; RW) .....................1354 Transmit Control Register--TCTL [0:3] (0x0400; R/W) .........................................1355 Transmit Control Extended--TCTL_EXT [0:3] (0x0404; R/W) ................................1356 Transmit IPG Register--TIPG [0:3] (0x0410; R/W) ..............................................1357 Retry Buffer Control--RETX_CTL [0:3] (0x041C; RW)...........................................1358 DMA Tx Control--DTXCTL [0:3] (0x3590; R/W) ...................................................1358 DMA TX TCP Flags Control Low--DTXTCPFLGL [0:3] (0x359C; RW) ........................1359 DMA TX TCP Flags Control High--DTXTCPFLGH [0:3] (0x35A0; RW) .......................1360 DMA TX Max Total Allow Size Requests--DTXMXSZRQ [0:3] (0x3540; RW) .............1360 DMA TX Maximum Packet Size--DTXMXPKTSZ [0:3] (0x355C; RW)........................1361 Transmit Descriptor Base Address Low--TDBAL [0:3][0:7] (0xE000 + 0x40*n [n=0...7]; R/W) ...................................................................1361 Transmit Descriptor Base Address High--TDBAH [0:3][0:7] (0xE004 + 0x40*n [n=0...7]; R/W) ...................................................................1362 Transmit Descriptor Ring Length--TDLEN [0:3][0:7] (0xE008 + 0x40*n [n=0...7]; R/W) ...................................................................1362 Transmit Descriptor Head--TDH [0:3] [0:7] (0xE010 + 0x40*n [n=0...7]; RO) .......1363 Transmit Descriptor Tail--TDT [0:3][0:7] (0xE018 + 0x40*n [n=0...7]; R/W) .........1363 Transmit Descriptor Control--TXDCTL [0:3][0:7] (0xE028 + 0x40*n [n=0...7]; R/W) ...................................................................1364 Tx Descriptor Completion Write-Back Address Low--TDWBAL [0:3][0:7] (0xE038 + 0x40*n [n=0...7]; R/W) ...................................................................1366 Tx Descriptor Completion Write-Back Address High - TDWBAH [0:3][0:7] (0xE03C + 0x40*n [n=0...7];R/W)....................................................................1366 Rx DCA Control Registers--RXCTL [0:3][0:7] (0xC014 + 0x40*n [n=0...7]; R/W) ...1367 Tx DCA Control Registers--TXCTL [0:3][0:7] (0xE014 + 0x40*n [n=0...7]; R/W) ....1369 DCA Requester ID Information--DCA_ID [0:3] (0x5B70; RO)................................1370 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 73 28-123 28-124 28-125 28-126 28-127 28-128 28-129 28-130 28-131 28-132 28-133 28-134 28-135 28-136 28-137 28-138 28-139 28-140 28-141 28-142 28-143 28-144 28-145 28-146 28-147 28-148 28-149 28-150 28-151 28-152 28-153 28-154 28-155 28-156 28-157 28-158 28-159 28-160 28-161 28-162 28-163 28-164 28-165 28-166 28-167 28-168 28-169 28-170 28-171 28-172 28-173 28-174 28-175 28-176 28-177 28-178 28-179 28-180 DCA Control--DCA_CTRL [0:3] (0x5B74; R/W) ................................................... 1370 VMDq Control Register--VT_CTL [0:3] (0x581C; R/W) ......................................... 1371 Malicious Driver Free Block--MDFB [0:3] (0x3558; RWS) ..................................... 1372 Last VM Misbehavior Cause--LVMMC[0:3] (0x3548; RC) ..................................... 1372 VM Offload Register--VMOLR [0:3][0:7] (0x5AD0 + 4*n [n=0...7]; RW) ................ 1373 Replication Offload Register--RPLOLR [0:3] (0x5AF0; RW .................................... 1374 VLAN VM Filter--VLVF [0:3][0:31] (0x5D00 + 4*n [n=0...31]; RW) ...................... 1375 Unicast Table Array--UTA [0:3][0:127] (0xA000 + 4*n [n=0...127]; R/W)............. 1376 Storm Control Control Register--SCCRL [0:3] (0x5DB0; RW)................................ 1376 Storm Control Status--SCSTS [0:3] (0x5DB4;RO)............................................... 1377 Broadcast Storm Control Threshold--BSCTRH [0:3] (0x5DB8;RW)......................... 1378 Multicast Storm Control Threshold--MSCTRH [0:3] (0x5DBC; RW) ........................ 1378 Broadcast Storm Control Current Count - BSCCNT [0:3] (0x5DC0;RO)................... 1378 Multicast Storm Control Current Count--MSCCNT [0:3] (0x5DC4;RO) .................... 1379 Storm Control Time Counter--SCTC [0:3] (0x5DC8; RO)...................................... 1379 Storm Control Basic Interval--SCBI [0:3] (0x5DCC; RW) ..................................... 1380 Virtual Mirror Rule Control--VMRCTL [0:3][0:3] (0x5D80 + 0x4*n [n= 0...3]; RW) .................................................................... 1380 Virtual Mirror Rule VLAN--VMRVLAN [0:3][0:3](0x5D90 + 0x4*n [n= 0...3]; RW)... 1381 Virtual Mirror Rule VM--VMRVM [0:3][0:3] (0x5DA0 + 0x4*n [n= 0...3]; RW) ........ 1381 DMA Receive Power Saving Register--DMARPS [0:3] (0x2500; R/W) ..................... 1382 DMA Transmit Power Saving Register--DMATPS [0:3] (0x2504; R/W) .................... 1382 Watchdog Setup--WDSTP [0:3] (0x1040; R/W) .................................................. 1384 Watchdog Software Device Status--WDSWSTS [0:3] (0x1044; R/W) ..................... 1385 Free Running Timer--FRTIMER [0:3] (0x1048; RWS)........................................... 1385 TCP Timer--TCPTIMER [0:3] (0x104C; R/W) ...................................................... 1386 RX Time Sync Control Register--TSYNCRXCTL [0:3] (0xB620;RW) ........................ 1387 RX timestamp Low--RXSTMPL [0:3] (0xB624; RO).............................................. 1388 RX Timestamp High--RXSTMPH [0:3] (0xB628; RO)............................................ 1388 RX Timestamp Attributes Low--RXSATRL[0:3] (0xB62C; RO)................................ 1388 RX timestamp Attributes High--RXSATRH [0:3] (0xB630; RO) .............................. 1389 TX Time Sync Control Register--TSYNCTXCTL [0:3] (0xB614; RW)........................ 1389 TX Timestamp Value Low--TXSTMPL [0:3] (0xB618;RO) ...................................... 1390 TX Timestamp Value High--TXSTMPH[0:3] (0xB61C; RO) .................................... 1390 System Time Register Residue--SYSTIMR [0:3] (0xB6F8; RW) ............................. 1390 System Time Register Low--SYSTIML [0:3] (0xB600; RW) ................................... 1391 System Time Register High--SYSTIMH [0:3] (0xB604; RW).................................. 1391 Time Adjustment Offset Register Low--TIMADJL [0:3] (0xB60C; RW) .................... 1392 Time Adjustment Offset Register High--TIMADJH [0:3] (0xB610;RW) .................... 1392 TimeSync Auxiliary Control Register--TSAUXC [0:3] (0xB640; RW) ....................... 1393 Target Time Register 0 Low--TRGTTIML0 [0:3] (0xB644; RW) .............................. 1395 Target Time Register 0 High--TRGTTIMH0 [0:3] (0xB648; RW)............................. 1395 Target Time Register 1 Low--TRGTTIML1 [0:3] (0xB64C; RW) .............................. 1395 Target Time Register 1 High--TRGTTIMH1 [0:3] (0xB650; RW)............................. 1396 Frequency Out 0 Control Register--FREQOUT0 [0:3] (0xB654; RW)....................... 1396 Frequency Out 1 Control Register--FREQOUT1 [0:3] (0xB658; RW)....................... 1397 Auxiliary Time Stamp 0 Register Low--AUXSTMPL0 [0:3] (0xB65C; RO)................. 1397 Auxiliary Time Stamp 0 Register High--AUXSTMPH0 [0:3] (0xB660; RO) ............... 1398 Auxiliary Time Stamp 1 Register Low--AUXSTMPL1 [0:3] (0xB664; RO) ................. 1398 Auxiliary Time Stamp 1 Register High--AUXSTMPH1 [0:3] (0xB668; RO) ............... 1399 Time Sync RX Configuration--TSYNCRXCFG [0:3] (0x5F50; R/W).......................... 1399 Time Sync SDP Configuration Register--TSSDP [0:3] (0x003C; R/W)..................... 1400 Time Sync Interrupt Cause Register--TSICR [0:3] (0xB66C; RC/W1C) ................... 1401 Time Sync Interrupt Mask Register--TSIM [0:3] (0xB674; RW)............................. 1402 Time Sync Interrupt Set Register--TSIS [0:3] (0xB670; WO) ............................... 1403 PCS Configuration--PCS_CFG [0:3] (0x4200; R/W) ............................................. 1404 PCS Link Control--PCS_LCTL [0:3] (0x4208; RW) ............................................... 1404 PCS Link Status--PCS_LSTS [0:3] (0x420C; RO) ................................................ 1406 AN Advertisement--PCS_ANADV [0:3] (0x4218; R/W) ......................................... 1408 Intel(R) Communications Chipset 89xx Series - Datasheet 74 October 2012 Order Number: 327879-001US Contents 28-181 28-182 28-183 28-184 28-185 28-186 28-187 28-188 28-189 28-190 28-191 28-192 28-193 28-194 28-195 28-196 28-197 28-198 28-199 28-200 28-201 28-202 28-203 28-204 28-205 28-206 28-207 28-208 28-209 28-210 28-211 28-212 28-213 28-214 28-215 28-216 28-217 28-218 28-219 28-220 28-221 28-222 28-223 28-224 28-225 28-226 28-227 28-228 28-229 28-230 28-231 28-232 28-233 28-234 28-235 Link Partner Ability--PCS_LPAB [0:3] (0x421C; RO).............................................1409 Next Page Transmit--PCS_NPTX [0:3] (0x4220; RW) ...........................................1410 Link Partner Ability Next Page--PCS_LPABNP [0:3] (0x4224; RO) ..........................1411 SFP I2C Parameters--I2CPARAMS [0:3] (0x102C; R/W) .......................................1412 CRC Error Count--CRCERRS [0:3] (0x4000; RC) .................................................1413 Alignment Error Count--ALGNERRC [0:3] (0x4004; RC) .......................................1414 Symbol Error Count--SYMERRS [0:3] (0x4008; RC).............................................1414 Missed Packets Count--MPC [0:3] (0x4010; RC)..................................................1415 Single Collision Count--SCC [0:3] (0x4014; RC)..................................................1415 Excessive Collisions Count--ECOL [0:3] (0x4018; RC) ..........................................1416 Multiple Collision Count--MCC [0:3] (0x401C; RC) ...............................................1416 Late Collisions Count--LATECOL [0:3] (0x4020; RC) ............................................1417 Collision Count--COLC [0:3] (0x4028; RC) .........................................................1417 Defer Count--DC [0:3] (0x4030; RC).................................................................1418 Host Transmit Discarded Packets by MAC Count--HTDPMC [0:3] (0x403C; RC)........1418 Receive Length Error Count--RLEC [0:3] (0x4040; RC) ........................................1419 XON Received Count--XONRXC [0:3] (0x4048; RC) .............................................1419 XON Transmitted Count--XONTXC [0:3] (0x404C; RC) .........................................1420 XOFF Received Count--XOFFRXC [0:3] (0x4050; RC) ...........................................1420 XOFF Transmitted Count--XOFFTXC [0:3] (0x4054; RC) .......................................1421 FC Received Unsupported Count--FCRUC [0:3] (0x4058; RC)................................1421 Packets Received [64 Bytes] Count--PRC64 [0:3] (0x405C; RC)............................1422 Packets Received [65-127 Bytes] Count--PRC127 [0:3] (0x4060; RC)....................1422 Packets Received [128-255 Bytes] Count--PRC255 [0:3] (0x4064; RC) ..................1423 Packets Received [256-511 Bytes] Count--PRC511 [0:3] (0x4068; RC) ..................1424 Packets Received [512-1023 Bytes] Count--PRC1023 [0:3] (0x406C; RC) ..............1424 Packets Received [1024 to Max Bytes] Count--PRC1522 [0:3] (0x4070; RC)...........1425 Good Packets Received Count--GPRC [0:3] (0x4074; RC).....................................1425 Broadcast Packets Received Count - BPRC [0:3] (0x4078; RC) ..............................1426 Multicast Packets Received Count--MPRC [0:3] (0x407C; RC) ...............................1426 Good Packets Transmitted Count--GPTC [0:3] (0x4080; RC).................................1427 Good Octets Received Count--GORCL [0:3] (0x4088; RC) ....................................1427 Good Octets Received Count--GORCH [0:3] (0x408C; RC)....................................1428 Good Octets Transmitted Count--GOTCL [0:3] (0x4090; RC) ................................1428 Good Octets Transmitted Count--GOTCH [0:3] (4094; RC) ...................................1429 Receive No Buffers Count--RNBC [0:3] (0x40A0; RC) ..........................................1429 Receive Undersize Count--RUC [0:3] (0x40A4; RC) .............................................1430 Receive Fragment Count--RFC [0:3] (0x40A8; RC) ..............................................1430 Receive Oversize Count--ROC [0:3] (0x40AC; RC)...............................................1431 Receive Jabber Count--RJC [0:3] (0x40B0; RC) ..................................................1431 Management Packets Received Count--MNGPRC [0:3] (0x40B4; RC)......................1432 Management Packets Dropped Count--MPDC [0:3] (0x40B8; RC) ..........................1432 Management Packets Transmitted Count--MNGPTC [0:3] (0x40BC; RC) .................1433 Total Octets Received--TORL [0:3] (0x40C0; RC) ................................................1433 Total Octets Received--TORH [0:3] (0x40C4; RC) ...............................................1434 Total Octets Transmitted--TOTL [0:3] (0x40C8; RC) ............................................1434 Total Octets Transmitted--TOTH [0:3] (0x40CC; RC) ...........................................1435 Total Packets Received--TPR [0:3] (0x40D0; RC) ................................................1435 Total Packets Transmitted--TPT [0:3] (0x40D4; RC .............................................1436 Packets Transmitted [64 Bytes] Count--PTC64 [0:3] (0x40D8; RC) .......................1436 Packets Transmitted [65-127 Bytes] Count--PTC127 [0:3] (0x40DC; RC) ...............1437 Packets Transmitted [128-255 Bytes] Count--PTC255 [0:3] (0x40E0; RC) ..............1437 Packets Transmitted [256-511 Bytes] Count--PTC511 [0:3] (0x40E4; RC) ..............1438 Packets Transmitted [512-1023 Bytes] Count--PTC1023 [0:3] (0x40E8; RC) ..........1438 Packets Transmitted [1024 Bytes or Greater] Count--PTC1522 [0:3] (0x40EC; RC) .................................................................................................1439 28-236 Multicast Packets Transmitted Count--MPTC [0:3] (0x40F0; RC)............................1439 28-237 Broadcast Packets Transmitted Count--BPTC [0:3] (0x40F4; RC)...........................1440 28-238 TCP Segmentation Context Transmitted Count--TSCTC [0:3] (0x40F8; RC).............1440 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 75 28-239 28-240 28-241 28-242 28-243 28-244 28-245 28-246 28-247 28-248 28-249 28-250 28-251 28-252 28-253 28-254 28-255 28-256 28-257 28-258 28-259 28-260 28-261 28-262 28-263 28-264 28-265 28-266 28-267 28-268 28-269 28-270 28-271 28-272 28-273 28-274 28-275 28-276 28-277 28-278 28-279 28-280 28-281 28-282 28-283 28-284 28-285 28-286 28-287 28-288 28-289 Interrupt Assertion Count--IAC [0:3] (0x4100; RC)............................................. 1441 Rx Packets to Host Count--RPTHC [0:3] (0x4104; RC)......................................... 1441 Debug Counter 1--DBGC1 [0:3] (0x4108; RC).................................................... 1441 DBGC1 Values ................................................................................................ 1442 Debug Counter 2--DBGC2 [0:3] (0x410C; RC) ................................................... 1442 DBGC2 Values ................................................................................................ 1442 Debug Counter 3--DBGC3 [0:3] (0x4110; RC).................................................... 1443 DBGC3 Values ................................................................................................ 1443 Debug Counter 4--DBGC4 [0:3] (0x411C; RC) ................................................... 1443 DBGC4 Values ................................................................................................ 1444 Host Good Packets Transmitted Count--HGPTC [0:3] (0x4118; RC) ....................... 1444 Receive Descriptor Minimum Threshold Count--RXDMTC [0:3] (0x4120; RC) .......... 1444 Host Good Octets Received Count--HGORCL [0:3] (0x4128; RC)........................... 1445 Host Good Octets Received Count--HGORCH [0:3] (0x412C; RC) .......................... 1445 Host Good Octets Transmitted Count--HGOTCL [0:3] (0x4130; RC) ...................... 1446 Host Good Octets Transmitted Count - HGOTCH [0:3] (0x4134; RC)...................... 1446 Length Error Count--LENERRS [0:3] (0x4138; RC).............................................. 1447 SerDes/SGMII/KX Code Violation Packet Count--SCVPC [0:3] (0x4228; RW) .......... 1447 Switch Drop Packet Count--SDPC [0:3] (0x41A4; RC) ......................................... 1448 Per Queue Good Packets Received Count--VFGPRC [0:3][0:7] (0x10010 + n*0x100 [n=0...7]; RO)................................................................. 1448 Per Queue Good Packets Transmitted Count--VFGPTC [0:3][0:7] (0x10014 + n*0x100 [n=0...7]; RO)................................................................. 1449 Per Queue Good Octets Received Count--VFGORC [0:3][0:7] (0x10018 + n*0x100 [n=0...7]; RO)................................................................. 1450 Per Queue Good Octets Transmitted Count--VFGOTC [0:3][0:7] (0x10034 + n*0x100 [n=0...7]; RO)................................................................. 1451 Per Queue Multicast Packets Received Count--VFMPRC [0:3][0:7] (0x10038 + n*0x100 [n=0...7]; RO)................................................................. 1451 BMC Management Packets Dropped Count--BMPDC [0:3] (0x4140; RC)................. 1452 BMC Management Packets Transmitted Count--BMNGPTC [0:3] (0x4144; RC) ........ 1452 BMC Management Packets Received Count--BMNGPRC [0:3] (0x413C; RC) ............ 1453 BMC Total Unicast Packets Received--BUPRC [0:3] (0x4400; RC).......................... 1453 BMC Total Multicast Packets Received--BMPRC [0:3] (0x4404; RC) ....................... 1454 BMC Total Broadcast Packets Received--BBPRC [0:3] (0x4408; RC) ...................... 1454 BMC Total Unicast Packets Transmitted--BUPTC [0:3] (0x440C; RC) ..................... 1454 BMC Total Multicast Packets Transmitted--BMPTC [0:3] (0x4410; RC) ................... 1455 BMC Total Broadcast Packets Transmitted--BBPTC [0:3] (0x4414; RC) .................. 1455 BMC FCS Receive Errors--BCRCERRS [0:3] (0x4418; RC) .................................... 1456 BMC Alignment Errors--BALGNERRC [0:3] (0x441C; RC) ..................................... 1456 BMC Pause XON Frames Received--BXONRXC [0:3] (0x4420; RC) ........................ 1456 BMC Pause XOFF Frames Received--BXOFFRXC [0:3] (0x4424; RC) ...................... 1457 BMC Pause XON Frames Transmitted--BXONTXC [0:3] (0x4428; RC) .................... 1457 BMC Pause XOFF Frames Transmitted--BXOFFTXC [0:3] (0x442C; RC) .................. 1458 BMC Single Collision Transmit Frames--BSCC [0:3] (0x4430; RC) ......................... 1458 BMC Multiple Collision Transmit Frames--BMCC [0:3] (0x4434; RC) ...................... 1458 Wakeup Control Register--WUC [0:3] (0x5800; R/W) .......................................... 1459 Wakeup Filter Control Register--WUFC [0:3] (0x5808; R/W) ................................ 1460 Wakeup Status Register--WUS [0:3] (0x5810; R/W1C) ....................................... 1461 Wakeup Packet Length--WUPL [0:3] (0x5900; RO) ............................................. 1462 Wakeup Packet Memory--WUPM [0:3][0:31] (0x5A00 + 4*n [n=0...31]; RO)......... 1462 IP Address Valid--IPAV [0:3] (0x5838; R/W)...................................................... 1463 IPv4 Address Table--IP4AT [0:3][0:3] (0x5840 + 8*n [n=0...3]; R/W).................. 1463 IPv6 Address Table--IP6AT [0:3][0:3] (0x5880 + 4*n [n=0...3]; R/W).................. 1464 Flex Filter Even Data Register Fields--FEDR [0:3][0:3][0:15] (0x9000 +16*n[0..15]; RW) ............................................................................ 1465 Flex Filter Odd Data Register Fields--FODR [0:3][0:3][0:15] (0x9000 +16*n[0..15];RW) ............................................................................. 1466 Intel(R) Communications Chipset 89xx Series - Datasheet 76 October 2012 Order Number: 327879-001US Contents 28-290 Flex Filter Mask Field Register--FMFR [0:3][0:3][0:15] (0x9008 +16*n[0..15];RW)..............................................................................1466 28-291 Flex Filter Queueing Field--FQFR [0:3][0:3][0:15] (0x90FC + 16*n[n=0..15] ;RW)........................................................................1467 28-292 Flex Filter Even Data Register Extended--FHFT_EXT_FEDR [0:3][0:3][0:15] (0x9A00 +16*n[0..15]; RW) ............................................................................1468 28-293 Flex Filter Odd Data Register Extended--FHFT_EXT_FODR [0:3][0:3][0:15] (0x9A00 +16*n[0..15];RW) .............................................................................1469 28-294 Flex Filter Mask Field Extended--FHFT_EXT_FMFR [0:3][0:3][0:15] (0x9A08 +16*n[0..15];RW) .............................................................................1469 28-295 Flex Filter Queueing Extended--FHFT_EXT_FQFR [0:3][0:3] (0x9AFC;RW) ..............1470 28-296 Management VLAN TAG Value--MAVTV [0:3] [0:7] (0x5010 +4*n [n=0...7]; RW) ..........................................................................1471 28-297 Management Flex UDP/TCP Ports--MFUTP [0:3][0:3] (0x5030 + 4*n [n=0...3]; RW) .........................................................................1471 28-298 Management Ethernet Type Filters--METF [0:3][0:3] (0x5060 + 4*n [n=0...3]; RW) .........................................................................1472 28-299 Management Control Register--MANC [0:3] (0x5820; RW) ...................................1472 28-300 Management Only Traffic Register--MNGONLY [0:3] (0x5864; RW)........................1474 28-301 Manageability Decision Filters--MDEF [0:3][0:7] (0x5890 + 4*n [n=0...7]; RW) .....1474 28-302 Manageability Decision Filters--MDEF_EXT [0:3][0:7] (0x5930 + 4*n[n=0...7]; RW) ..........................................................................1476 28-303 Manageability IP Address Filter--MIPAF [0:3][0:15] (0x58B0 + 4*n [n=0...15]; RW) .......................................................................1479 28-304 Manageability MAC Address Low--MMAL [0:3][0:1] (0x5910 + 8*n [n= 0...1]; RW) ........................................................................1479 28-305 Manageability MAC Address High--MMAH [0:3][0:1] (0x5914 + 8*n [n=0...1]; RW) .........................................................................1480 28-306 Parity and ECC Error Indication--PEIND [0:3] (0x1084; RC)..................................1482 28-307 Parity and ECC Indication Mask--PEINDM [0:3] (0x1088; RW) ..............................1483 28-308 DMA Transmit Descriptor Parity Status--DTPARS [0:3] (0x3510; RW1C) ................1483 28-309 DMA Receive Descriptor Parity Status--DRPARS [0:3] (0x3514; RW1C) ..................1484 28-310 Dhost Parity Status--DDPARS [0:3] (0x3518; RW1C)...........................................1484 28-311 Tx Packet Buffer ECC Status--TPBECCSTS [0:3] (0x345C; RW) .............................1485 28-312 LAN Port Parity Error Control Register--LANPERRCTL [0:3] (0x5F54; RW) ...............1485 28-313 LAN Port Parity Error Status Register--LANPERRSTS [0:3] (0x5F58; RO) ................1486 29-1 Bus B, Device D, Function 1: Summary of PCIe GigE Configuration Registers...........1490 29-2 Bus B, Device D, Function 2: Summary of PCIe GigE Configuration Registers...........1490 29-3 Bus B, Device D, Function 3: Summary of PCIe GigE Configuration Registers...........1490 29-4 Bus B, Device D, Function 4: Summary of PCIe GigE Configuration Registers...........1490 29-5 Bus B, Device D, Function 1+Index1: Summary of PCIe GigE Configuration Registers .......................................................................................................1490 29-6 Bus B, Device D, Function 1+Index 1: Summary of GigE Registers Mapped Through GbEPCI Memory BAR .......................................................................................1492 29-7 PVID[0:3]--PF Vendor Identification Register ......................................................1493 29-8 PDID0--PF Device Identification Register (GbE0 ) ................................................1494 29-9 PDID1--PF Device Identification Register (GbE1 ) ................................................1494 29-10 PDID2--PF Device Identification Register (GbE2 ) ................................................1495 29-11 PDID3--PF Device Identification Register (GbE3 ) ................................................1495 29-12 PPCICMD[0:3]--PF Device Command Register .....................................................1496 29-13 PPCISTS[0:3]--PF PCI Device Status Register .....................................................1497 29-14 PRID[0:3]--PF Revision ID Register ...................................................................1499 29-15 PCC[0:3]--PF Class Code Register .....................................................................1499 29-16 PHDR[0:3]--PF Header Type Register ................................................................1500 29-17 GbE Controller Base Address Registers Description - LAN 0...3 ..............................1500 29-18 Base Address Registers' Fields ..........................................................................1500 29-19 GbEPCIBAR0[0:3]--BAR0 Base Address Register .................................................1501 29-20 GbEPCIBAR1[0:3]--BAR1 Base Address Register .................................................1502 29-21 GbEPCIBAR2[0:3]--BAR2 Base Address Register .................................................1502 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 77 29-22 29-23 29-24 29-25 29-26 29-27 29-28 29-29 29-30 29-31 29-32 29-33 29-34 29-35 29-36 29-37 29-38 29-39 29-40 29-41 29-42 29-43 29-44 29-45 29-46 29-47 29-48 29-49 29-50 29-51 29-52 29-53 29-54 29-55 29-56 29-57 29-58 29-59 29-60 29-61 29-62 29-63 29-64 29-65 29-66 29-67 29-68 29-69 29-70 29-71 29-72 29-73 29-74 29-75 29-76 29-77 29-78 29-79 29-80 GbEPCIBAR3[0:3]--BAR3 Base Address Register................................................. 1503 GbEPCIBAR4[0:3]--BAR4 Base Address Register................................................. 1504 GbEPCIBAR5[0:3]--BAR5 Base Address Register................................................. 1505 PSVID[0:3]--PF Subsystem Vendor ID Register .................................................. 1505 PSID[0:3]--PF Subsystem ID Register ............................................................... 1506 PCP[0:3]--PF Capabilities Pointer Register ......................................................... 1506 PIRQL[0:3]--PF Interrupt Line Register.............................................................. 1507 PIRQP0--PF Interrupt Pin Register 0 .................................................................. 1507 PIRQP1--PF Interrupt Pin Register 1 .................................................................. 1508 PIRQP2--PF Interrupt Pin Register 2 .................................................................. 1508 PIRQP3--PF Interrupt Pin Register 3 .................................................................. 1509 PPMCAP[0:3]--PF Power Management Capabilities ID Register .............................. 1510 PPMCP[0:3]--PF Power Management Next Capability Pointer Register .................... 1510 PPMC[0:3]--PF Power Management Capabilities Register...................................... 1511 PPMCSR[0:3]--PF Power Management Control and Status Register ........................ 1512 PMSICID[0:3] - Message Signalled Interrupt Capability ID Register ....................... 1513 PMSINCP[0:3]--Message Signalled Interrupt Next Capability Pointer Register ......... 1513 PMSICTL[0:3]--Message Signalled Interrupt Control Register ............................... 1514 PMSILADDR[0:3]--Message Signalled Interrupt Lower Address Register ................. 1514 PMSIUADDR[0:3]--Message Signalled Interrupt Upper Address Register ................ 1515 PMSIDATA[0:3]--Message Signalled Interrupt Data Register................................. 1515 PMSIMSK[0:3]--Message Signalled Interrupt Mask Register ................................. 1515 PMSIPND[0:3]--Message Signalled Interrupt Pending Register .............................. 1516 MSI-X Capability Structure............................................................................... 1517 PMSI-X[0:3]--PF Message Signalled Interrupt X Capability ID Register................... 1517 PMSIXNCP[0:3]--PF MSIX Next Capability Pointer Register ................................... 1518 PMSIXCNTL[0:3]--PF Message Signalled Interrupt X Control Register .................... 1518 PMSIXTBIR[0:3]--PF MSI-X Table Offset & Table BIR Register .............................. 1519 PMSIXPBABIR[0:3]--PF MSI-X Pending Bit Array & BIR Offset Register .................. 1519 IOADDR[0:3]--IOADDR Register....................................................................... 1520 IODATA[0:3]--IODATA Register........................................................................ 1520 PPCID[0:3]--PF PCI Express Capability Register.................................................. 1521 PPCP[0:3]--PF PCI Express Next Capability Pointer Register ................................. 1522 PPCR[0:3]--PF PCI Express Capabilities Register ................................................. 1522 PPDCAP[0:3]--PF PCI Express Device Capabilities Register ................................... 1523 PPDCNTL[0:3]--PF PCI Express Device Control Register ....................................... 1524 PPDSTAT[0:3]--PF PCI Express Device Status Register ........................................ 1525 PLCAPR[0:3]--PF Link Capabilities Register ........................................................ 1526 PLCNTLR[0:3]--PF Link Control Register ............................................................ 1528 PLSR[0:3]--PF Link Status Register................................................................... 1529 PDCAPR2[0:3]--PF Device Capabilities 2 Register................................................ 1531 PDCNTR2[0:3]--PF Device Control 2 Register ..................................................... 1532 PLCNTLR2[0:3]--PF Link Control 2 Register ........................................................ 1533 PLSR2[0:3]--PF Link Status 2 Register .............................................................. 1535 VPDCID[0:3]--VPD Capability ID Register .......................................................... 1536 VPDNCP[0:3]--VPD Next Capability Pointer Register ............................................ 1536 VPDADDR[0:3]--VPD Address Register .............................................................. 1537 VPDDATA[0:3]--VPD Data ............................................................................... 1537 PCIe Extended Configuration Space................................................................... 1538 PCIe Extended Capability Structure ................................................................... 1538 PPCIEAERCAPID[0:3]--PF PCI Express AER Capability ID Register ......................... 1539 PPAERUCS[0:3]--PF PCI Express AER Uncorrectable Error Status Register .............. 1539 PPAERUCM[0:3]--PF PCI Express AER Uncorrectable Error Mask Register ............... 1540 PPAERUCSEV[0:3]--PF PCI Express AER Uncorrectable Error Severity Register........ 1541 PPAERCS[0:3]--PF PCI Express AER Correctable Error Register ............................. 1542 PPAERCM[0:3]--PF PCI Express AER Correctable Error Mask Register .................... 1543 PPAERCTLCAP[0:3]--PF PCI Express AER Control and Capability Register ............... 1543 PPAERHDRLOG0[0:3]--PF PCI Express AER Header Log 0 Register ........................ 1544 PPAERHDRLOG1[0:3]--PF PCI Express AER Header Log 1 Register ........................ 1544 Intel(R) Communications Chipset 89xx Series - Datasheet 78 October 2012 Order Number: 327879-001US Contents 29-81 29-82 29-83 29-84 29-85 29-86 29-87 29-88 29-89 30-1 30-2 31-1 32-1 32-2 32-3 32-4 32-5 32-6 32-7 32-8 32-9 32-10 32-11 32-12 32-13 32-14 32-15 32-16 32-17 32-18 32-19 32-20 32-21 32-22 32-23 32-24 32-25 32-26 32-27 32-28 32-29 33-1 33-2 33-3 33-4 33-5 33-6 33-7 33-8 33-9 33-10 33-11 33-12 33-13 33-14 33-15 33-16 33-17 33-18 PPAERHDRLOG2[0:3]--PF PCI Express AER Header Log 2 Register .........................1545 PPAERHDRLOG3[0:3]--PF PCI Express AER Header Log 3 Register .........................1545 PARIDHDR[0:3]--PF Alternative Routing ID Capability Header ...............................1546 PFARICAP0--PF ARI Capabilities Register ............................................................1547 PFARICAP1--PF ARI Capabilities Register ............................................................1548 PFARICAP2--PF ARI Capabilities Register ............................................................1549 PFARICAP3--PF ARI Capabilities Register ............................................................1549 PARIDCTL[0:3]--PF Alternative Routing ID Control Register ..................................1550 IOBAR Register Map ........................................................................................1551 PCH Die-level Legacy I/O (DMI)TAP Pin Interface.................................................1553 PCH Die-Level Endpoint TAP Pin Interface ...........................................................1554 Platform External Clock Interface ......................................................................1557 Signal Type Definitions.....................................................................................1558 Direct Media Interface Signals...........................................................................1560 PCI Express* Endpoint Interface Signals ............................................................1561 SerDes/SGMII Interface Signals .......................................................................1563 SFP Interface Signals .......................................................................................1565 GbE EEPROM SPI Interface Signals ....................................................................1566 GbE SMBus Interface Signals (Master/Slave).......................................................1566 LED, Software Defined, Miscellaneous Signals .....................................................1567 EndPoint Management SMBus Interface Signals (Slave) ........................................1570 PCI Express* Root Complex Interface Signals .....................................................1570 SATA Interface Signals ....................................................................................1571 LPC Interface Signals .......................................................................................1573 USB Interface Signals ......................................................................................1574 UART Interface Signals ....................................................................................1575 Host SMBus (Master) Interface Signals...............................................................1578 SPI Boot Interface Signals ................................................................................1579 Interrupt Interface Signals ...............................................................................1579 Processor Interface Signals...............................................................................1580 Power Management Interface Signals.................................................................1581 Thermal Sensor Current Reference ....................................................................1583 Miscellaneous Interface Signals .........................................................................1584 General Purpose I/O Interface Signals................................................................1584 Real Time Clock (RTC) Interface Signals .............................................................1592 Clock Input Interface Signals ............................................................................1594 PCH JTAG Interface Signals ..............................................................................1595 EndPoint JTAG Interface Signals .......................................................................1595 Strap Signals ..................................................................................................1597 Reserved Signals.............................................................................................1601 Power and Ground Signals................................................................................1604 PCH Absolute Maximum and Minimum Ratings ....................................................1606 Operating Conditions Power Supply Rails ............................................................1607 Maximum Supply Current .................................................................................1608 Power for Standby Support Designs ...................................................................1609 Power for Non-Standby Support Designs ...........................................................1610 DC Input Characteristic Signal Association ..........................................................1611 DC Input Characteristics ..................................................................................1612 DC Output Characteristic Signal Association ........................................................1614 DC Output Characteristics ................................................................................1615 PCI Express* and DMI Gen1 TX Specification Interface Timings .............................1616 PCI Express* and DMI Gen1 RX Specification Interface Timings .............................1617 PCI Express* Gen2 TX Specification Interface Timings..........................................1619 PCI Express* Gen2 RX Specification Interface Timings .........................................1620 SATA Specification Interface Timings .................................................................1622 GbE SGMII Driver DC Specification Interface.......................................................1623 GbE SGMII Receiver DC Specification Interface (AC Coupled) ................................1623 GbE SGMII Driver AC Specification Interface .......................................................1624 GbE SGMII Receiver AC Specification Interface....................................................1624 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 79 33-19 33-20 33-21 33-22 33-23 33-24 33-25 33-26 33-27 33-28 33-29 33-30 33-31 33-32 33-33 33-34 33-35 33-36 33-37 33-38 33-39 33-40 33-41 35-1 35-2 35-3 35-4 35-5 35-6 35-7 SerDes GbE 1000 Base RX Specifications ........................................................... 1624 SerDes GbE 1000 Base TX Specifications ........................................................... 1624 CRU Differential Input Clock DC Specifications .................................................... 1625 CRU Differential Input Clock Timing Specifications............................................... 1627 SATA, DMI, and PCIe* Differential Input Clock DC Specification ............................ 1627 SATA, DMI, and PCIe* Differential Input Clock AC Specification ............................ 1628 Digital 3.3V I/O Timing Characteristics .............................................................. 1631 Clock Timings................................................................................................. 1633 Universal Serial Bus (USB) Timing..................................................................... 1635 SMBus Timing ................................................................................................ 1636 LPC Timing .................................................................................................... 1636 SPI Timings (20 MHz)...................................................................................... 1636 SPI Timings (33 MHz)...................................................................................... 1637 SPI Timings (50 MHz)...................................................................................... 1637 I2C Timing Parameters .................................................................................... 1638 DC Input Characteristics: MDIO Mode of Operation.............................................. 1639 DC Output Characteristics: MDIO Mode of Operation ........................................... 1639 MDIO I/F Timing Parameters ............................................................................ 1640 EEPROM I/F Timing Parameters ........................................................................ 1641 PECI Timings ................................................................................................. 1642 UART Received Timing..................................................................................... 1643 UART Transmit Timing..................................................................................... 1643 JTAG I/F Timing Parameters............................................................................. 1644 Alphabetical Ball Listing ................................................................................... 1652 Alphabetical Signal Listing................................................................................ 1660 Ball Map Color Code ........................................................................................ 1668 Bottom View Left ............................................................................................ 1668 Bottom View Right .......................................................................................... 1671 Top View Left ................................................................................................. 1675 Top View Right ............................................................................................... 1678 Intel(R) Communications Chipset 89xx Series - Datasheet 80 October 2012 Order Number: 327879-001US Overview and PCH Interfaces Volume 1 of 4 October 2012 Order Number: 327879-001US 1.0 Introduction 1.1 Introduction Intel(R) Communications Chipset 89xx Series can be most easily described as a PCH that includes standard PC interfaces (e.g., PCI Express, SATA, USB, etc.) along with Intel(R) QuickAssist Technology and I/O interfaces. This document is intended as a reference for architects, hardware/software designers and engineers, or others who need technical information for silicon development and for programming the hardware. It is expected that the readers of this document have a understanding of: * IA-64 microprocessor * Memory controller * I/O architecture (PCI Express) * Intel(R) QuickAssist Technology * A basic understanding of system software (operating system and BIOS) internals Note: Throughout this document, all references to "PCH" apply to "Intel(R) Communications Chipset 89xx Series" or "Chipset 89xx Series" PCH. Intel(R) Communications Chipset 89xx Series - Datasheet 82 October 2012 Order Number: 327879-001US 1.0 1.2 Intel(R) Communications Chipset 89xx Series SKU Definition Table 1-1 provides a high-level summary of the Chipset 89xx Series PCH SKUs. Table 1-1. Intel(R) Communications Chipset 89xx Series SKUs DH8900CC SKU1 DH8903CC SKU2 DH8910CC SKU3 DH8920CC SKU4 PCIe EndPoint (max width) x4 x4 x8 x16 Virtual Function (SR-IOV) for Intel(R)QuickAssist Technology Services 0 16 16 16 Devices GbE 4 4 4 4 SATA 2 2 2 2 USB 6 6 6 6 Symmetric Cryptographic Functions 0 2 4 4 Public Key Functions 0 2 4 4 Compression/Decompress ion 0 1 2 2 Normal Mode TDP (W) 8.5 9.5 11 12 End Point Mode TDP 5.7 6.7 8.2 9.2 Non-End Point Mode TDP 2.8 2.8 2.8 2.8 TJUNCTION-MAX (C) 103 103 103 103 TJUNCTION-MIN (C) 0 0 0 0 Notes: 1. Contact your local Intel Field Sales Representative for currently available PCH SKUs. 2. The above shows feature differences between the PCH SKUs. If a feature is not listed in the table, it is considered as a base feature that is included in all SKUs. 1.3 Document Organization This document first describes the Intel(R) Communications Chipset 89xx Series PCH product line architecture, building blocks and usage models for the product line as well as their high-level programming model and memory map. The overview chapter also reviews the block diagram and defines external and internal interfaces. The document is organized into volumes: -- Volume 1 - Overview and PCH Interfaces -- Volume 2 - PCIe Endpoint and Gigabit Ethernet -- Volume 3 - Test Features -- Volume 4 - Technical Specifications October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 83 1.4 Referenced Documents and Related Websites Table 1-2. Referenced Documents Document Title Table 1-3. Location Number Advanced Configuration and Power Interface (ACPI) Specification http://www.acpi.info/ Intel Corporation, Enhanced Host Controller Interface Specification for Universal Serial Bus http://www.intel.com/technology/usb/ ehcispec.htm Intel Corporation, High Precision Event Timers (HPET) Specification http://www.intel.com/hardwaredesign/ hpetspec.htm Intel Corporation, Low Pin Count (LPC) Interface Specification http://www.intel.com/design/chipsets/ industry/lpc.htm Intel Corporation, Enhanced Host Controller Interface (EHCI) Specification http://www.intel.com/technology/usb/ ehcispec.htm Intel Corporation, Universal Serial Bus (USB) Specification http://www.intel.com/technology/usb/ spec.htm Intel Corporation, USB2 Debug Device Functional Specification http://www.intel.com/technology/usb/ download/DebugDeviceSpec_R090.pdf JEDEC Specification http://www.jedec.org/default.cfm Serial ATA Specification https://www.sata-io.org/developers/pu rchase_spec.asp SMBus Specification http://www.smbus.org/specs/ Universal Serial Bus Specification 1.1 and 2.0 http://www.usb.org/developers/docs/ Related Websites Specification or Technology PIRQ routing table information ACPI and related specifications Website http://www.microsoft.com/whdc/archive/pciirq.mspx http://www.acpi.info/spec.htm AT Attachment-6 with Packet Interface (ATA/ATAPI-6) http://T13.org (T13 1410D) BIOS boot specifications http://www.phoenix.com/en/customer+services/whit e+papers-specs/ PCI and PCI Express* related specifications Power management specifications Intel(R) Communications Chipset 89xx Series - Datasheet 84 Order http://www.pcisig.com/specifications http://www.microsoft.com/whdc/resources/respec/s pecs/pmref/default.mspx October 2012 Order Number: 327879-001US 1.0 1.5 Acronyms This section describes acronyms that are used throughout this document. Table 1-4. Acronym Table (Sheet 1 of 3) Term Description ACPI Advanced Configuration and Power Interface Specification, an industry specification of the common interfaces enabling robust operating system (OS)-directed motherboard device configuration and power management of both devices and entire systems. AHCI Advanced Host Controller Interface, an industry specification of the interface between memory and SATA devices. ARP Address resolution protocol ASF Alert Specification Format. This is the next generation of "Alert on LAN*" implementation. BAR PCI Base Address Register used to define the base and limit of an I/O or memory region assigned to a PCI device. BER Bit Error Rate BGA Ball Grid Array CMC Common Mode Choke CM Coherent Memory CNR Communications and Networking Riser CRC See Cyclic Redundancy Check in Table 1-5. DDP Direct Data Placement Protocol DDR DDR SDRAM (Double Data Rate Synchronous Dynamic Random Access Memory) is a system memory technology. DED Double-bit Error Detect DMA See Direct Memory Access in Table 1-5. DW Double Word. A legacy reference to 32 bits of data on a naturally aligned four-byte boundary (i.e. the least significant two bits of the byte address are b00). This is a legacy term used by PCI and must not be used other than in that context. ECC Error Checking and Correction EMI Electro Magnetic Interference EMTS Electrical Mechanical Thermal Specification used for processor specifications. ESD Electrostatic Discharge FRU Field Replaceable Unit FS Full-speed. Refers to USB. GbE Gigabit Ethernet GigE Gigabit Ethernet HBA Host Bus Adapter - necessary when connecting a peripheral to a computer that doesn't have native support for that peripheral's interface. HCD Host Controller Device - USB interface for programmers HECBASE HPET HSI IA IA-CPU October 2012 Order Number: 327879-001US PCI Express Enhanced Configuration Base Register High Precision Event Time (HPET) - The IA-PC HPET Architecture defines a set of timers that can be used by the operating system. The timers are defined such that the OS may be able to assign specific timers to be used directly by specific applications. Each timer can be configured to generate a separate interrupt. High Speed Interface. Refers to USB. Intel Architecture instruction set commonly known as "x86" IA-CPU, IA Complex and IA Processor are the same terminology Intel(R) Communications Chipset 89xx Series - Datasheet 85 Table 1-4. Acronym Table (Sheet 2 of 3) Term Description INTx Legacy PCI interrupt architecture that encodes interrupts on one of four side-band signals (INTA, INTB, INTC, and INTD). I/O IP ISA 1. Input/Output. 2. When used as a qualifier to a transaction type, specifies that transaction targets Intel(R) architecture-specific I/O space (e.g., I/O read). Internet Protocol See Industry Standard Architecture in Table 1-5 LML Latency Measurement Logic LPC Low Pin Count LS Low-speed. Refers to USB. LSb Least Significant Bit LSB Least Significant Byte ME Management Engine MMIO MSb Memory Mapped I/O Most Significant Bit MSB Most Significant Byte MSI Message-signaled interrupt that encodes interrupts as an in-band 32-bit write transaction. MTBF Mean Time Between Failures NCM Non Coherent Memory NIC Network interface controller OS Operating system. OSPM P2P Operating system directed Power Management See Peer-to-Peer in Table 1-5 PCH Platform Controller Hub PCI Peripheral Component Interconnect Local Bus. A 32- or 64-bit bus with multiplexed address and data lines that is primarily intended for use as an interconnect mechanism within a system between processor/memory and peripheral components or add-in cards. PCM PDP Platform Pulse Code Modulation PECI-to-DIMM Processor (PDP) based platform: Platforms based on CPUs with integrated SMBuses to DIMMs. The PCH collects DIMM Thermal Data via the PECI interface to the Processor PEC Packet Error Checking. This is an SMBUS 2.0 feature. POC Power-on-configuration QAT Intel QuickAssist Technology RASUM Reliability, Availability, Serviceability, Usability, and Manageability, which are important characteristics of servers. RCBA Root Complex Base Address. RCRB Root Complex Register Block, as defined in the PCI Express Specification v1.0a. RDMA RFL Remote Direct Memory Access Receive FIFO Level RMW Read-Modify-Write operation RTC Real-Time Clock RTCRESET# SATA Signal that resets the RTC well (but does not clear the RTC RAM memory contents). Serial Advanced Technology Attachment Intel(R) Communications Chipset 89xx Series - Datasheet 86 October 2012 Order Number: 327879-001US 1.0 Table 1-4. Acronym Table (Sheet 3 of 3) Term Description SATA* Serial ATA, an industry specification of the interface for storage controllers and devices. SDP Platform SEC Single-bit Error Correct SEC/DED Single Error Correct/Double Error Detect - A specific data protection algorithm that distributes data and ECC across 144 bits. Enables correction of single bit errors. Allows detection of double bit errors. SMM System Management Mode SPD Serial Presence Detect STR Suspend To Ram TAP Test Access Port used for testability and debug of the component. TCO Total Cost of Ownership TCP Transmission Control Protocol TDR Time Domain Reflectometry TFL Transit FIFO Level TID See Transaction Identifier in Table 1-5 USB Universal Serial Bus VSCC Vendor Specific Component Capabilities WDT Watch Dog Timer Classification used to describe a register's reset value when the value of the register is indeterminate. Certain registers' value at reset does not have a default value, therefore it is unknown what the value will be. XX 1.6 SMBus-to-DIMM Processor (PDP) based platform: Platforms based on CPUs without integrated SMBuses to DIMMs. The PCH collects DIMM Thermal Data via the platform internal Host SMBus interface to the DIMMs Glossary This section presents a glossary for this document. Table 1-5. Glossary Table (Sheet 1 of 6) Term Agent ALT Access Mode Definition A logical device connected to a bus or shared interconnect that can either initiate accesses or be the target of accesses. Mode to allow the reading of write-only registers, usually used when saving/restoring register content for power management sleep state implementations. Anti-Etch Any plane-split, void or cutout in a VCC or GND plane is referred to as an anti-etch. Asserted Signal is set to a level that represents logical true. Asynchronous Atomic operation October 2012 Order Number: 327879-001US 1. An event that causes a change in state with no relationship to a clock signal. 2. When applied to transactions or a stream of transactions, a classification for those that do not require service within a fixed time interval. A series of two or more transactions to a device by the same initiator which are guaranteed to complete without intervening accesses by a different master. Most commonly required for a read-modify-write (RMW) operation. Intel(R) Communications Chipset 89xx Series - Datasheet 87 Table 1-5. Glossary Table (Sheet 2 of 6) Term Definition Buffer 1. A random access memory structure. 2. The term I/O buffer is also used to describe a low-level input receiver and output driver combination. Cx States Processor power states (Cx states) are processor power consumption and thermal management states within the global working state, G0. * C0: Processor power state - While the processor is in this state, it executes instructions. * C1: Processor power state - This power state has the lowest latency. The hardware latency in this state must be low enough that the operating software does not consider the latency aspect of the state when deciding whether to use it. * C2: Processor power state - This state offers improved power savings over the C1 state. The worst-case hardware latency for this state is provided via the ACPI system firmware and operating software can use this information to determine when the C1 state should be used instead of the C2 state. * C3: Processor power state - This state is not supported. The C3 state offers improved power savings over the C1 and C2 states. The worst-case hardware latency for this state is provided via the ACPI system firmware and the operating software can use this information to determine when the C2 state should be used instead of C3 state. While in the C3 state, the processor's caches maintain state but ignore any snoops. Cache Line The unit of memory that is copied to and individually tracked in a cache. Specifically, 64 bytes of data or instructions aligned on a 64-byte physical address boundary. Cfg Used as a qualifier for transactions that target PCI configuration address space. Character The raw data byte in an encoded system (i.e., the 8b value in a 8b/10b encoding scheme). This is the meaningful quantum of information to be transmitted or that is received across an encoded transmission path. Coherent (C) Transactions that ensure that the processor's view of memory through the cache is consistent with that obtained through the I/O subsystem. In Chipset 89xx Series PCH, Coherent (C) memory regions are coherent with IA caches when accessed from AIOC agents. Accesses to these memory regions enter the memory system through the IMCH. Memory regions that are coherent with IA caches must be accessible to the IA CPU Command The distinct phases, cycles, or packets that make up a transaction. Requests and Completions are referred to generically as Commands. Completion A packet, phase, or cycle used to terminate a Transaction on a interface, or within a component. A Completion will always refer to a preceding Request and may or may not include data and/or other information. Core Power Well Cyclic Redundancy Check Deasserted Main system power, turns off in S3 - S5 A number derived from, and stored or transmitted with, a block of data in order to detect corruption. By recalculating the CRC and comparing it to the value originally transmitted, the receiver can detect some types of transmission errors. Signal is set to a level that represents logical false. Deferred Transaction A processor bus Split Transaction. The requesting agent receives a Deferred Response which allows other transactions to occur on the bus. Later, the response agent completes the original request with a separate Deferred Reply transaction. Delayed Transaction A transaction where the target retries an initial request, but unknown to the initiator, forwards or services the request on behalf of the initiator and stores the completion or the result of the request. The original initiator subsequently reissues the request and receives the stored completion. Direct Memory Access Method of accessing memory on a system without interrupting the processors on that system. Intel(R) Communications Chipset 89xx Series - Datasheet 88 October 2012 Order Number: 327879-001US 1.0 Table 1-5. Glossary Table (Sheet 3 of 6) Term Definition Downstream Describes commands or data flowing away from the processor-memory complex and toward I/O. The terms Upstream and Downstream are never used to describe transactions as a whole. (e.g. Downstream data may be the result of an Outbound Write, or an Inbound Read. The Completion to an Inbound Read travels Downstream.) Full Duplex A connection or channel that allows data or messages to be transmitted in opposite directions simultaneously. Gb/s Gigabits per second (109 bits per second) GB/s Gigabytes per second (109 bytes per second) Global visibility An operation is said to be globally visible when all side-effects of the operation are visible to every observer in the system. For example, a write to some resource (e.g., memory location, control register, etc.) R achieves global visibility when a read of R by all other agents is guaranteed to return the new value. Gx States Global system states (Gx states) apply to the entire system and are visible to the user. * G3: Mechanical off - A computer state that is entered and left by a mechanical switch. It is implied by the entry of this off state through a mechanical means that no electrical current is running through the circuitry and that it can be worked on without damaging the hardware or endangering service personnel. * G2/S5: Soft Off - A computer state where the computer consumes a minimal amount of power. * G1: Sleeping - A computer state where the computer consumes a small amount of power, user mode threads are not being executed, and the system "appears" to be off (from an end user's perspective, the display is off, and so on). * G0: Working - A computer state where the system dispatches user mode (application) threads and they execute. In this state, peripheral devices are having their power state changed dynamically. Half Duplex A connection or channel that allows data or messages to be transmitted in either direction, but not simultaneously. Implicit Writeback A snoop-initiated data transfer from the bus agent with the modified Cache Line to the memory controller due to an access to that line. Inbound A transaction where the request destination is the processor-memory complex and is sourced from I/O. The terms Inbound and Outbound refer to transactions as a whole and never to Requests or Completions in isolation. (e.g., an Inbound Read generates Downstream data, whereas an Inbound Write has Upstream data. Even more confusing, the Completion to an Inbound Read travels Downstream.) Industry Standard Architecture (ISA) A 16-bit bus architecture associated with the IBM AT motherboard designed to connect motherboard circuitry to expansion card devices that is now considered Legacy. Initiator The source of requests. [IBA] An agent sending a request packet on 3GIO is referred to as the Initiator for that Transaction. The Initiator may receive a completion for the Request. [3GIO] ISA Regime A special legacy mode to support ISA-based devices which have been integrated into the chipset. It opens a dedicated channel from the peripheral device to the processor bus. While in this mode, the legacy device is granted exclusive accesses to memory and the ability to use Tenured Transactions. Isochronous A classification of transactions or a stream of transactions that require service within a fixed time interval. October 2012 Order Number: 327879-001US Lane A set of differential signal pairs, one pair for transmission and one pair for reception. A by-N Link is composed of N Lanes. Layer A level of abstraction commonly used in interface specifications as a tool to group elements related to a basic function of the interface within a layer and to identify key interactions between layers. Intel(R) Communications Chipset 89xx Series - Datasheet 89 Table 1-5. Glossary Table (Sheet 4 of 6) Term Legacy Link LPC Bus Master Mbyte/s Definition Functional requirements handed down from previous chipsets, or PC compatibility requirements from the past. The collection of two Ports and their interconnecting Lanes. A Link is a dual simplex communications path between two components. Low Pin Count connection used to connect to the super I/O device. A device or logical entity that is capable of initiating transactions. A Master is any potential Initiator. Megabytes per second (106 bytes per second) Mem Used as a qualifier for transactions that target memory space. (For example, a Mem read to I/O.) Metastability A characteristic of flip flops that describes the state where the output becomes non-deterministic. Most commonly caused by a setup or hold time violation. Multi Media Timer (MMT) Non-Coherent North See High Precision Event Timer (HPET) in Table 1-4. Transactions that may cause the processor's view of memory through the cache to be different than that obtained through the I/O subsystem. Usually refers to bridges. The bridge or device that is closer to the processor-memory complex. Ordering Refers to the order in which signals and/or memory accesses to different locations must reach global visibility to ensure some behavior. This excludes the "ordering" necessary to prevent data hazards which are accesses to the same location. Outbound A transaction where the request destination is I/O and is sourced from the processor-memory complex. The terms Inbound and Outbound refer to transactions as a whole and never to Requests or Completions in isolation. (For example, an Outbound Read generates Upstream data, whereas an Outbound Write has Downstream data. Even more confusing, the Completion to an Outbound Read travels Upstream.) OWord 128 bits of data on a naturally aligned sixteen-byte boundary (e.g., the least significant four bits of the byte address are b"0000"). This is the native size of the IMCH datapath. Packet The indivisible unit of data transfer and routing, consisting of a header, data, and CRC. PCI Reset PCIRST#. This is the secondary PCI Bus reset signal. It is a logical OR of the primary interface PLTRST# signal and the state of the Secondary Bus Reset bit of the Bridge Control register (D30:F0:Reg3Eh[6]). Peer-to-Peer Transactions that occur between two devices independent of memory or the processor. Platform Reset ILB asserts PLTRST# to reset devices that reside on the primary PCI bus. The ILB asserts PLTRST# during power-up and when a hard reset sequence is initiated through the CF9h register. PLTRST# is driven inactive a minimum of 1 ms after both PWROK and VGATE are driven high. PLTRST# is driven for a minimum of 1 ms when initiated through the CF9h register. Plesiochronous From Greek, meaning almost synchronous. Describes signals that have the same nominal digital rate, but are synchronized on different clocks. Any variation in rate is constrained within specified limits, which allows a device to process the data signal without buffer underflow or overflow by making periodic compensating adjustments that repeat or delete dummy data bits. However, there is no limit to the phase difference that can accumulate between the signals over time. Port 1. Logically, an interface between a component and a PCI Express Link. 2. Physically, a group of Transmitters and Receivers located on the same chip that define a Link. Intel(R) Communications Chipset 89xx Series - Datasheet 90 October 2012 Order Number: 327879-001US 1.0 Table 1-5. Glossary Table (Sheet 5 of 6) Term Posted Push Model Queue Definition A Transaction that is considered complete by the initiating agent or source before it actually completes at the Target of the Request or destination. All agents or devices handling the Request on behalf of the original Initiator must then treat the Transaction as being system visible from the initiating interface all the way to the final destination. Commonly refers to memory writes. Method of messaging or data transfer that predominately uses writes instead of reads. A first-in first-out (FIFO) structure. Receiver 1. The Agent that receives a Packet across an interface regardless of whether it is the ultimate destination of the packet. 2. More narrowly, the circuitry required to convert incoming signals from the physical medium to more perceptible forms. Request A packet, phase, or cycle used to initiate a Transaction on a interface, or within a component. Reserved The contents or undefined states or information that are not defined at this time. Using any reserved area is not permitted. Reserved register bits must be set to 0. However, when stated, there may be specific instances where a reserved register is either non-zero, or there may be a requirement to make it non-zero. Resume Power Well Resume Reset RTC Power Well Trickle from power supply, only turns off when power is disconnected from wall. Signal that resets the parts of the PCH in the resume power well, generated when the trickle supply turns on. Powered by a coin cell battery and only turns off when the battery is drained. Powers the RTC and some resume events. Sx States Sleeping states (Sx states) are types of sleeping states within the global sleeping state, G1. * S5: Soft Off state. The main memory power plane is shut down in addition to the clock synthesizer and core well power planes for the processor and CMI. The CMI resume well is still powered. * S4: Sleeping state - This state is only used to transition to or from the S5 state. The S4 state is not a supported power management state in CMI. * S3: Suspend to RAM (STR) state - The clock synthesizer and core well power planes for the processor and CMI are shut down, but the main memory power plane and the CMI resume well remain active. All clocks from synthesizers are shut down during the S3 state. * S0: Awake state - Power Management state when all power planes are active. Simplex A connection or channel that allows data or messages to be transmitted in one direction only. SMBus Snooping South South Port Split Lock Sequence Split Transaction October 2012 Order Number: 327879-001US System Management Bus. A two-wire interface through which various system components may communicate. A means of ensuring cache coherency by monitoring all memory accesses on a common multi-drop bus to determine if an access is to information resident within a cache. Usually refers to bridges. The bridge or device that is further from the processor-memory complex. The PCI Express downstream root port(s) on the ILB. A sequence of transactions that occurs when the target of a lock operation is split across a processor bus data alignment or Cache Line boundary, resulting in two read transactions and two write transactions to accomplish a read-modify-write operation. A transaction that consists of distinct Request and Completion phases or packets that allow use of bus, or interconnect, by other transactions while the Target is servicing the Request. Intel(R) Communications Chipset 89xx Series - Datasheet 91 Table 1-5. Glossary Table (Sheet 6 of 6) Term Definition Symbol An expanded and encoded representation of a data Byte in an encoded system (e.g., the 10b value in a 8b/10b encoding scheme). This is the value that is transmitted over the physical medium. Symbol Time The amount of time required to transmit a symbol. Target A device that responds to bus Transactions. The agent receiving a request packet is referred to as the Target for that Transaction. Tenured Transaction A transaction that holds the bus or interconnect until complete, effectively blocking all other transactions while the Target is servicing the Request. Transaction An overloaded term that represents an operation between two or more agents that can be comprised of multiple phases, cycles, or packets. Transaction Identifier A multi-bit field used to uniquely identify a transaction. Commonly used to relate a Completion with its originating Request in a Split Transaction system. Transmitter 1. The Agent that sends a Packet across an interface regardless of whether it was the original generator of the packet. 2. More narrowly, the circuitry required to drive signals onto the physical medium. Upstream Describes commands or data flowing toward the processor-memory complex and away from I/O. The terms Upstream and Downstream are never used to describe transactions as a whole. (For example, Upstream data may be the result of an Inbound Write, or an Outbound Read. The Completion to an Outbound Read travels Upstream.) Intel(R) Communications Chipset 89xx Series - Datasheet 92 October 2012 Order Number: 327879-001US 2.0 2.0 Architecture Overview 2.1 Introduction The PCH targets communications and embedded platforms. These product's stringent performance, power, and cost targets can be met by integrating common platform components, adding on-chip hardware accelerators. Systems based on the PCH include four major blocks: * One or more IA CPU cores * IA-compatible memory and I/O controller hubs * On-chip memory and controllers for external memory * A model-specific communications complex Examples of such communications-oriented functions are: * PCI Express target * Ethernet * Protocol acceleration hardware : -- Intel(R) QuickAssist Integrated Accelerators -- Security acceleration or bulk encryption, hashing, public/private key generation, and compression. From a system standpoint, Chipset 89xx Series can be most easily described as a PCH that includes both standard PC interfaces (e.g., PCI Express Root Complex, SATA, USB, etc.) along with Intel(R) QuickAssist Technology and GbE interfaces. The PCH can be accessed by the IA processor via two interfaces: a DMI interface that provides connectivity to the standard PC interfaces and a PCI Express interface that provides connectivity to the PCH's PCI Express End Point (EP) for Intel(R) QuickAssist Technology and access to the GbE MACs. These interfaces are not dependent on one another and can be used independently in a system. This offers three modes in which the PCH can be used: * Normal Mode (via DMI and PCI Express EP Interfaces) For PC Interfaces, GbE & Intel(R) QuickAssist Technology * End Point Mode (via PCI Express EP Interface Only) For GbE & Intel(R) QuickAssist Technology * Non-End Point Mode (via DMI Interfaces Only) For PC Interfaces October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 93 Figure 2-1. PCH Modes Normal Mode PCI Express EP Intel(R) QuickAssist Technology & GbE MACs To CPU PC Interfaces DMI Interface 89xx Series PCH End Point Mode PCI Express EP Intel(R) QuickAssist Technology & GbE MACs To CPU PC Interfaces Not Powered DMI Interface (N/C) 89xx Series PCH Non-End Point Mode (R) PCI Express EP (N/C) Intel QuickAssist Technology & GbE MACs Not Powered To CPU PC Interfaces DMI Interface 89xx Series PCH Intel(R) Communications Chipset 89xx Series - Datasheet 94 October 2012 Order Number: 327879-001US 2.0 2.2 PCH Architecture Overview This section provides an overview of the PCH architecture. Section 2.2.1, "PCH Block Summary" gives a high-level summary for each of the major blocks and their internal interfaces. Section 2.2.2, "PCH External Interfaces" reviews PCH's external chip interfaces and internal block gear ratios. The following fgure shows the major PCH blocks. Figure 2-2. PCH Block Diagram EndPoint Unit Primary: 3.3V LVIO SMBus EndPoint (Slave) SGMII GbE MAC SFP/I2C/MDIO Port 0 Function 1 SGMII GbE MAC SFP/I2C/MDIO Port 1 Function 2 PCIe Endpoint 1.25GT/s Primary: 1.0V Support: 3.3V PLL/BG: 1.8V SGMII SGMII Primary: 3.3V LVIO (x4, x8, or x16) 5.0GT/s PCIe G2 Gen2 Primary: 1..0V PLL/BG: 1.8V Intel(R) QuickAssist Technology GbE MAC SFP/I2C/MDIO Port 2 Function 3 Function 0 GbE MAC SFP/I2C/MDIO Port 3 Function 4 GbE SMBus (Mater/Slave) DEVICE 0 125MHz 480MT/s Primary: 1.0V Compat: 3.3V PLL/BG: 1.8V 3.0GT/s Primary: 1.0V PLL/BG: 1.8V 2.5GT/s Primary: 1.0V PLL/BG: 1.8V Primary: 3.3V RTC: 3.3V USB SATA USB 1.1 or 2.0 (6 Ports) DMI G1 (x4) (2 Ports) 68 General Purpose I/O LVIO Primary: 3.3V LPC 1.1, SPI Boot, SMLink1, SMBus (Host) LVIO Primary: 3.3V (4x1, 2x2, 1x2 + 2x1, or 1x4) Gen1 LVIO Timers, WDT, RTC, HPET (x3) APIC, PIC, DMA UART (2 Ports) October 2012 Order Number: 327879-001US 2.5GT/s Gen1 SATA 2.0 PCIe Root Complex PCIe G1 DMI South LVIO Primary: 3.3V Intel(R) Communications Chipset 89xx Series - Datasheet 95 2.2.1 PCH Block Summary The PCH has two major interfaces to the IA processor: the PCIe interface for Intel(R) QuickAssist Technology and GbE MACs and the DMI interface or standard PCIO interfaces. 2.2.2 PCH External Interfaces Table 2-1 summarizes the key features of PCH's external interfaces. Specific products may elect to make a subset of these interface visible based on their needs and requirements. Table 2-1. PCH External Interface Summary Name Qty. Description PCI Express (Gen2, Endpoint) 1 Supports x1, x4, x8, or x16 widthsa and provides the primary interface into the PCH's PCIe Endpoint for GbE and services. This interface is Gen2 capable. The PCH has no parallel PCI interface. DMI 2.0 Gen1 1 Supports x4 width and provides the primary interface into the PCH's Legacy I/O. Gigabit Ethernet 4 10/100/1000 Gigabit Ethernetb MACs with SGMII interface. Each of the ports support IEEE 1588 time synchronization. MDIO 4 MDIO/I2C support the configuration of PHY or SFP. PCI Express Gen1 (root) 1-4 USB 2.0 1 Universal Serial Bus 2.0 host controller interface, supports six USB ports (shared with USB 1.1 ports) USB 1.1 1 Universal Serial Bus 1.1 host controller interface, supports six USB ports (shared with USB 2.0 ports) SATA 2 SATA 1.0 or 2.0 used to attached external hard drives. Supports 4x1, 2x2, 1x2 + 2x1, or 1x4 configurations. This interface is PCI Express 2.0 Compliant, operating at Gen1 speed (2.5 GT/s). The PCH has no parallel PCI interface. LPC 1 Low Pin Count Bus (LPC) interface. SPI 1 Serial Peripheral Interface (SPI). Used for boot device. GPIO 20c SMBus/I2C 2 I2C compatible SMBus2.0 connections. UART 2 16550 compatible asynchronous serial ports that support data rate of at least 115Kbits/sec. Programmable General Purpose I/O (GPIO) pins. Of the pins, many have alternate functions defined. a. SKU dependent b. See the Supported Ethernet PHY Device for the PCH. c. Dedicated. Intel(R) Communications Chipset 89xx Series - Datasheet 96 October 2012 Order Number: 327879-001US 2.0 2.2.3 IA Compatibility 100% IA platform compatibility allows applications and operating systems to scaledown from mainstream IA products to Chipset 89xx Series-based products. The ability to deploy an unmodified binary IA operating system with existing IA device drivers for third party PCI Express devices on a Chipset 89xx Series-based product is critical. From an operating system developer's perspective, a Chipset 89xx Series-based product manifests itself like a new IA chipset and not as an entirely different platform. That is, devices can be enumerated and configured via existing PCI configuration mechanisms, signaling and ordering follow established PCI rules, expected legacy IA chipset features such as interrupt and DMA controllers, timers, real-time clock and ACPI power management interfaces are supported. While most communications and embedded applications live on modern 32-bit operating systems (Linux, BSD, VxWorks, QNX, etc.), support for 8/16 bit environments seems unnecessary. However, many users of the IA legacy functionality exist. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 97 3.0 PCH Platform Memory and Device Configuration 3.1 Overview This chapter presents the views of the major address spaces and device configuration structures as seen by various internal and external agents in a PCH system. Three related aspects are covered: * The memory maps seen by various internal and external agents in a PCH system. * The endianess seen by various agents in a PCH system and mechanisms PCH uses to allow communication between agents with different endianess expectations. * The PCI configuration infrastructure for the PCH, which the PCH exposes through its memory maps. 3.1.1 Configuration Objectives The PCH device and configuration model blends the architectures of many disparate components into a unified whole. The major goals for the device configuration and access architecture include: * Provide a configuration and access model that is aligned with existing IA platform algorithms. * Support a unified address space model. The IA is the primary agent responsible for device configuration. This is true across all supported SKU configurations. To provide device configuration and operation capabilities that are aligned with the IA platform, the PCH exposes the devices through PCI devices or functions. Figure 3-1. Device-Centric Logical View of PCH Devices PCIe* Root Bus M PCI Bus 0 PCI Host Bridge SATA USB Legacy PCIe* EndPoint Intel(R) QuickAssist Accelerator GbE IA CPU Intel(R) Communications Chipset 89xx Series - Datasheet 98 October 2012 Order Number: 327879-001US 3.0 In this approach, the PCH extends the hardware blocks that make up the PCIe* Endpoint to allow them to materialize as a collection of PCI functions. This allows the IA to use existing system software to discover, enumerate, and operate all devices in the chip. The PCIe* Endpoint is not built around a PCI fabric, but rather, layers PCI abstractions on top of existing infrastructure. 3.1.2 Terminology and Conventions Throughout this chapter, the generic term "device" refers to either a PCI device or a function of a PCI device. The text will be explicit when the distinction between device and function is important. Addresses are always in hexadecimal and broken into 16-bit segments, such as 0_FEED_BEEF. When the distinction is important and not obvious, addresses are subscripted with "V", "P", or "S" for virtual, physical, or system address spaces. 3.2 IA Platform Infrastructure The PCH provides an IA platform infrastructure with respect to endianess, address spaces and memory maps, configuration, etc. This section focuses on the IA views and expectations around the endianess, address spaces and memory maps, and configuration for a basic IA platform. The PCH operates within this framework. Later sections in this chapter and volumes discuss the specifics of the PCH implementation. These discussions highlight how and where the PCH differs from the framework. 3.2.1 General IA Platform View of the Physical Address Space The PCH operates within a standard IA physical address space. To support specific processors, the PCH supports at least 46-bit physical address spaces. For more information on the specific layout of the address space, consult the appropriate IA documentation for the core that is used with the PCH. 3.2.2 IA Platform View of Configuration The PCH is comprised of IA functional blocks that are exposed through a PCI infrastructure. The PCH extends this PCI infrastructure to expose the functionality in the PCIe* Endpoint, as Section 3.6, "PCI Configuration" on page 102 describes. Before describing how the PCIe* Endpoint integrates with the IA-based PCH blocks, it is helpful to consider how PCI exposes PCH functionality. Logically, the software-visible sub-blocks of the PCH materialize as PCI devices and functions on PCI bus 0 of the system through three independent address spaces: * Configuration Space: Each function of each device has at least 256B of configuration space that is mapped to a fixed location based on bus, device and function number (PCI Express devices can provide for larger configuration spaces). This space provides system software with basic information on the device and allows for device-independent configuration. * Memory-Mapped I/O (MMIO) and I/O Spaces: Each function of each device can request up to six MMIO or I/O regions of device-specified sizes to be mapped into physical address space through base address registers in the configuration header. System software selects the base address of each region. These spaces support device-specific operation such as access to device-specific control registers. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 99 In general, the IA uncore claims configuration accesses (i.e., those accesses that target configuration space) to some devices on bus 0 and routes configuration accesses to the remaining devices to the PCH over the DMI interface using Type 0 PCI configuration transactions. 3.3 High-Level Views This section presents an overview of some of the general characteristics of the agents that the various PCH memory maps expose. 3.3.1 Characteristics of External System Memory (DRAM) The PCH operates in the standard IA physical address space. See the relevant IA CPU documentation for more details on address space size and layout. 3.3.2 Characteristics of Internal and External Memories Table 3-1 defines the supported operations by memory type. The table uses the following notation to indicate the behavior of the PCH: * "-" means the operation is not supported by the PCH. * "S" implies that the operation happens as a single atomic1 update to memory. In other words, either the update is observable in its entirety or not at all. * "M" implies that the operation may happen as multiple updates to memory. In other words, other agents can observe different parts of the affected memory location change values in any order but the end state of the memory location will be the desired value. This "flickering lights" effect makes such memory accesses useless for multi-agent synchronization unless a semaphore or flag variable is used to guard access to the shared location2. This table only applies to aligned-to-size operations; that is, a 4-byte operation is aligned to a 4-byte boundary, an 8-byte operation is aligned to an 8-byte boundary, etc.) Table 3-1. Supported Operations by Memory Type Operation Type Read, Write IA CPU Size [Bytes] IA WB Cacheable IA UC MMIO 1 S S S 2 S S S 4 S S S 8 S S S 16 S M M 32 - - - 64 S - - 128 - - - 1. In the sense that it cannot be divided into multiple smaller writes. 2. In guarding the location, visibility of the new flag must imply that the "flickering" has stopped. Intel(R) Communications Chipset 89xx Series - Datasheet 100 October 2012 Order Number: 327879-001US 3.0 Table 3-1. Supported Operations by Memory Type Operation Type Atom ReadModify-Write (Semaphore) IA CPU Size [Bytes] IA WB Cacheable IA UC MMIO 1 S M - 2 S M - 4 S M - 8 S M - . 3.3.3 Characteristics of Device Configuration To be able to leverage existing IA BIOS, Operating Systems, and power management software, PCH's configuration mechanisms follow existing IA platform approaches. Of the two major complexes of the PCH: * The PCH can use normal IA platform configuration algorithms. * To interoperate with normal IA platform configuration algorithms, the PCIe* Endpoint must be configured by the IA processor. To the extent possible, the configuration algorithms should be independent of the specific topology to allow for maximum re-use of hardware and software designs. Further, they should be forward-looking to support future platforms with address spaces larger than 32 bits. This implies that control registers should be sized to allow the device to perform an access anywhere in a 64-bit address space. The goal does not require the device to be able to be a 64-bit target (i.e., it need not support 64-bit BARs to allow its MMIO regions to materialize in PCI H). With the PCH, the boot and configuration process is: 1. The IA boots from a FLASH device on the PCH SPI interface. 2. System software discovers and configures the devices and function on PCI bus 0 in the PCH and PCIe* Endpoint. 3. System software configures other buses on the system. Once this process completes, the PCH and IA are ready for operation. 3.4 Memory Map for PCIe* Endpoint-Attached Devices All PCIe* Endpoint-attached agents support independent target IDs that provide independent address spaces. Agents that do not natively support 64-bit addressing are mapped into the full 64-bit address space as defined in the sections below. Table 3-2. Address Space Sizes of PCIe* Endpoint-Attached Devices Address Space Size [b] Devices 50 Gigabit Ethernet MACs, PCIe* Endpoint All addresses are zero-extended to 64-bits by definition. 3.5 PCH Endianness The PCH operates in an IA platform environment that is little-endian. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 101 3.6 PCI Configuration This section presents an overview of the implementation that integrates the PCIe* Endpoint with the IA PCI infrastructure for configuration. * PCI mechanisms (configuration space, memory-mapped I/O spaces, and I/O spaces) expose state for configuration. * The IA-32 core performs all system configuration and initialization. * The IA-32 core configures the PCIe* Endpoint using standard IA platform algorithms (i.e., PCI-based discovery, enumeration, and configuration) with modifications for the specific mix of functionality that the PCH instantiations provide. * The GbE MACs and other sub-blocks appear as PCI functions of a PCI Endpoint and allow the use of the standard PCI discovery, enumeration, and configuration algorithms (this implies, for example, corresponding PCI configuration headers, etc.). * A PCH-specific user driver handles interaction with external non-PCI agents attached to PCH through its PCIe* Endpoint I/O interfaces. That is, the PCH user driver, not BIOS, will "discover" and operate any such devices. Since these devices do not implement PCI semantics, it is expected that they will not allocate MMIO regions beyond those already allocated for the PCI view of the appropriate PCIe* Endpoint function. * The O/S always allocates an aperture in the memory map for any PCI device that defines one or more BARs, even if the device is unknown to the O/S at discovery. 3.6.1 Overview The PCH integrates the legacy PC Interfaces (DMI section) and PCIe* Endpoint into the PCI fabric as Figure 3-2 shows. This figure presents a logical view of the system and does not show exact physical connectivity nor the exact fabrics. Intel(R) Communications Chipset 89xx Series - Datasheet 102 October 2012 Order Number: 327879-001US 3.0 Figure 3-2. Attaching PCH to the PCI Fabric (Logical Perspective) PCH Die DMI QuickAssist Accelerator Devices GigE PCI Bus 0 DMI Root PCI Bus 0 CPU Devices Core Die PCIe Root Endpoint PCIe up to x16 (Bus M) PCIe External PCIe Devices In the figure above, PCI bus 0 originates in the CPU die and reaches internal PCI devices through internal paths. Bus 0 is bridged into the PCH via an DMI interconnect and into the PCIe* Endpoint via a subset of the PCI Express lanes on the CPU die. The PCI Express Endpoint materializes the individual functions of the endpoint device on bus M as the figure illustrates. 3.6.2 Device Tree The PCIe* Endpoint implements the function semantics, effectively mediating accesses between PCI bus M and its functions. Devices and functions can request space in the system memory and I/O address maps through BARs in the configuration header. In general, the PCH materializes most device control and status registers in memory-mapped regions allocated by a BAR. The only exception lies in the standard PCI configuration, status, and capability registers that PCI requires which materialize only in PCI configuration space. The remainder of this section summarizes the device tree that the PCH implements. This summary includes a mapping between PCI devices and the PCH blocks along with the value of the device ID, class code, and a summary of the resources (i.e., registers, memory, etc.) that each device requests. The following table summarizes the PCI devices that the PCH materializes. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 103 Table 3-3. PCH PCI Device Summary PCI Device Name B/D/Fa PCIe* Function Subtractive decode BSPb Device ID B0:D28:Fn 0x244E PCIe* Function Subtractive decode B0:D28:Fn 0x2448 LPC Interface B0:D31:F0 0x2310 SATA Controller 1 Desktop: in IDE, supporting 2 ports B0:D31:F2 0x2321 SATA Controller 1 Desktop: in AHCI supporting 4 ports B0:D31:F2 0x2323 SATA Controller 2 Desktop: in IDE, supporting 2 ports B0:D31:F5 0x2326 SMBus Controller B0:D31:F3 0x2330 Thermal Subsystem B0:D31:F6 0x2332 USB 2.0 Controller B0:D29:F0 0x2334 0x2335 PCI-Express Root Port #1 B0:D28:F0 0X2342 0x2343 PCI-Express Root Port #2 B0:D28:F1 0x2344 0x2345 PCI-Express Root Port #3 B0:D28:F2 0x2346 0x2347 PCI-Express Root Port #4 B0:D28:F3 0x2348 0x2349 WDT Timer for per core reset B0:D31:F7 0x2360 Intel(R) Management Engine Interface #1 (MEI #1) B0:D22:F0 0x2364 Intel(R) Management Engine Interface #2 (MEI #2) B0:D22:F1 0x2365 a. PCI bus number, device number, and function number. b. PCI base class code, subclass code, and programming interface PCI configuration register values. Table 3-4. EP PCI Device Summary PCI Device Name B/D/Fa BSPb Device ID PCIe* Endpoint & IQAT BM:DD:F0 0434h GbE, Port 0 BM:DD:F1 GbE, Port 1 BM:DD:F2 GbE, Port 2 BM:DD:F3 GbE, Port 3 BM:DD:F4 0436h, 0438h, 043Ah, 043Ch, 0440hc a. PCI bus number, device number, and function number. b. PCI base class code, subclass code, and programming interface PCI configuration register values. c. All GbE functions use one of the DIDs at a given time. A DID is assigned to the GbEs as shown in Table 3-5. Intel(R) Communications Chipset 89xx Series - Datasheet 104 October 2012 Order Number: 327879-001US 3.0 Table 3-5. GbE PCI Device ID Summary GbE DID Usage 0436h Default DID - this is not handled in the OS network driver. 0438h DH8900CC Series Gigabit Network Connection. 043Ah DH8900CC Series Gigabit Fiber Network Connection. 043Ch DH8900CC Series Gigabit Backplane Network Connection. 0440h DH8900CC Series Gigabit SFP Network Connection. Table 3-3 and Table 3-5 imply a logical view of the configuration registers. They are not meant to imply physical location. 3.6.3 Materializing Device Structures The PCH exposes PCIe* Endpoint resources through standard PCI abstractions: devices/functions, configuration spaces and memory-mapped I/O spaces. Access to MMIO and I/O spaces are provided through memory and I/O read/write instructions, respectively. The addressing of these spaces for a given device depends on the specific mapping that the PCI configuration header establishes through BARs. PCI defines two mechanisms for accessing the 256B of each device/function configuration registers located in PCI configuration space. * PCI Mechanism: The header is accessed using 1-, 2-, or 4-byte IN and OUT instructions that access the PCI configuration address and data I/O ports at addresses 0CF8h - 0CFBh and 0CFCh - 0CFFh, respectively, in the IA I/O space. This mechanism allows access only to the 256B PCI-compatible configuration space. * PCI Express Enhanced Mechanism: The header is accessed using 1-, 2-, or 4-byte memory accesses to the 256MB region starting at HECBASE. This mechanism allows access to an expanded 4KB configuration space that PCI Express defines (the first 256B are, by definition, the PCI-compatible configuration space). The PCH supports both mechanisms. These mechanisms differ in the address space they use to access the header. The PCI mechanism travels through IA I/O space while the PCI Express Enhanced mechanism travels through IA memory space. The address format that the mechanisms use is identical to the standard IA platform format that encodes the PCI bus, device, and function numbers along with a register offset or number. For either access method, the hardware in the PCIe* Endpoint that implements the configuration headers must be able to process accesses of the appropriate sizes. 3.6.4 PCI Configuration Headers The PCI specification requires each PCI function to provide a 256B configuration space. The first 64B of this space contains a standard PCI configuration header and the remaining 192B contains any device-specific registers, capabilities records, etc. needed by the function. There are two flavors of configuration headers: * All non-bridge devices provide a PCI type 0 configuration headers. This form of header is used to represent devices on the PCI fabric. * All bridge devices provide a PCI type 1 configuration header. This form of header is used to represent bridge devices in the PCI fabric. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 105 Since the PCIe* Endpoint devices that the PCH exposes through PCI are not, strictly speaking, PCI devices or functions. The following tables describe the support in greater detail. Table 3-6 summarizes the fields in a PCI type 0 header (i.e., header for non-bridge devices) and identifies which fields the PCH implements for Endpoint functions. The PCH hardware implements the appropriate PCI semantics for all supported registers and fields in this table. Table 3-6. PCI Configuration Header Support for Type 0 Headers in PCIe* Endpoint Devices (Sheet 1 of 2) Register and Field Bit(s) Supt.a Acc.b 00h - 01h Vendor ID 15:0 Y RO Required by PCI. 02h - 03h Device ID 15:0 Y RO Required by PCI. Interrupt Disable 10 Y RW Supported in devices that can use INTxc. Fast Back-to-Back Enable 9 N RO Not supported. SERR# Enable 8 Y RW Supported. 06h - 07h Status Register 04h - 05h Command Register Offset Notes Parity Error Response 6 Y RW Supported. VGA Palette Snoop 5 N RO Not supported. Mem. Write & Inval. Enable 4 N RO Not supported. Special Cycles 3 N RO Not supported. Bus Master Enable 2 Y RW QAT, GbE, devices. Memory Space Enable 1 Y RW All devices: QAT, GbE. I/O Space Enable 0 Detected Parity Error Signalled System Error Y RW For GbE; GbE materializes in I/O space. N RO All devices: QAT except GbE. 15 Y RW1C Supported. 14 Y RW1C Supported. Received Master-Abort 13 Y RW1C Supported. Received Target-Abort 12 Y RW1C Supported. Signalled Target-Abort 11 Y RW1C DEVSEL Timing 10:9 N RO Master Data Parity Error 8 Y RW1C Fast Back-to-Back Capable 7 N RO Not supported. 66MHz Capable 5 N RO Not supported. Capabilities List 4 Y RO Setup based on capabilities exposure by device. Interrupt Status Supported. Not supported. Supported. 3 Y RO Supported. Revision ID 7:0 Y RO Required by PCI. 09h - 0Bh Class Code 23:0 Y RO Required by PCI. 0Ch Cache Line Size 7:0 N RO Not supported. 08h Intel(R) Communications Chipset 89xx Series - Datasheet 106 October 2012 Order Number: 327879-001US 3.0 Table 3-6. PCI Configuration Header Support for Type 0 Headers in PCIe* Endpoint Devices (Sheet 2 of 2) Offset Register and Field Bit(s) Supt.a Acc.b 0Dh Latency Timer 7:0 N RO 0Eh Header Type 7:0 Y RO Required by PCI. 0Fh BIST 7:0 N RO Not supported. 10h - 27h Base Address (x6) 6x 31:0 Y RW Devices that materialize in I/O or memory spaces will populate these slots as necessary based on address space needs. 28h - 2Bh CIS Pointer 31:0 N RO Not supported. 2Ch - 2Dh Subsystem VID 15:0 Y RO Required by PCI. 2Eh - 2Fh Subsystem ID 15:0 Y RO Required by PCI. Notes Not supported. 34h Capability Pointer 7:0 Y RO Setup based on capabilities exposure by device. 3Ch Interrupt Line 7:0 Y RW Supported in devices that can use INTx. 3Dh Interrupt Pin 7:0 Y RO Supported in devices that can use INTx. 3Eh Min_Gnt 7:0 N RO Not supported. 3Fh Max_Lat 7:0 N RO Not supported. a. Supported fields provide appropriate PCI semantics. Unsupported fields always return zero on reads unless otherwise noted. b. RW1, RO and RW access types indicate that the register or field supports read-only access and read/write access c. The specific portion of the 256B PCI configuration space that is active in a device depends on the needs of the specific device. In general, a device requires far less than 256B of storage to implement a typical configuration space. Regions of the 256B configuration space that are not required are reserved and need only support default behavior compliant with the PCI specification: * All PCI devices must treat Configuration Space write operations to reserved registers as no-ops; that is, the access must be completed normally on the bus and the data discarded. Read accesses to reserved or unimplemented registers must be completed normally and a data value of 0 returned. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 107 4.0 Functional Description This chapter describes system functions and interfaces. 4.1 PCI Express* Root Ports There are four PCIe* Root Ports (Ports 1 - 4). The root ports operate at Gen1 speed (2.5GT/s) but are compliant to PCIe* Gen2 specification Messaging Protocol. The ports all reside in bus 0:device 28, and take function 0 - 3. Port 1 is function 0, port 2 is function 1, port 3 is function 2, and port 4 is function 3. The PCI Express* Root Ports can be independently configured to support four x1s, two x2s, one x2 + two x1 or one x4 port widths. The port configuration is set by soft straps in the Flash Descriptor. 4.1.1 Interrupt Generation The root port generates interrupts on behalf of hot-plug and power management events, when enabled. These interrupts can be pin based or can be MSIs, when enabled. An interrupt can be generated via a legacy mechanism that is based on the setting of the chipset configuration registers. Specifically, the chipset configuration registers used are the D28IP (Base address + 310Ch) and D28IR (Base address + 3146h) registers. The following table summarizes interrupt behavior for MSI and wire-modes. Bits refer to the hot-plug and PME interrupt bits. Table 4-1. MSI vs. PCI IRQ Actions Interrupt Register MSI Action All bits 0 Wire inactive No action One or more bits set to 1 Wire active Send message One or more bits set to 1, new bit gets set to 1 Wire active Send message One or more bits set to 1, software clears some (but not all) bits Wire active Send message One or more bits set to 1, software clears all bits Wire inactive No action Software clears one or more bits, and one or more bits are set on the same clock Wire active Send message Intel(R) Communications Chipset 89xx Series - Datasheet 108 Wire-Mode Action October 2012 Order Number: 327879-001US 4.0 4.1.2 Power Management 4.1.2.1 S3/S4/S5 Support Software initiates the transition to S3/S4/S5 by performing an IO write to the Power Management Control register. After the IO write completion has been returned to the processor, each root port sends a PME_Turn_Off TLP (Transaction Layer Packet) message on its downstream link. The device attached to the link eventually responds with a PME_TO_Ack TLP message followed by sending a PM_Enter_L23 DLLP (Data Link Layer Packet) request to enter the L2/L3 Ready state. When all of the root ports links are in the L2/L3 Ready state, the power management control logic proceeds with the entry into S3/S4/S5. Prior to entering S3, software is required to put each device into D3HOT. When a device is put into D3HOT it initiates entry into a L1 link state by sending a PM_Enter_L1 DLLP. Thus under normal operating conditions when the root ports sends the PME_Turn_Off message the link is in state L1. However, when the root port is instructed to send the PME_Turn_Off message, it sends it whether or not the link was in L1. Attached EndPoints can make no assumptions about the state of the link prior to receiving a PME_Turn_Off message. 4.1.2.2 Resuming from Suspended State The root port contains enough circuitry in the suspend well to detect a wake event through the WAKE# signal and to wake the system. When WAKE# is detected asserted, an internal signal is sent to the power management controller to cause the system to wake up. This internal message is not logged in any register, nor is an interrupt/GPE generated due to it. 4.1.2.3 Device Initiated PM_PME Message When the system has returned to a working state from a previous low power state, a device requesting service sends a PM_PME message continuously, until acknowledge by the root port. The root port takes different actions depending upon whether this is the first PM_PME has been received, or whether a previous message has been received but not yet serviced by the operating system. If this is the first message received (RSTS.PS - B0:D28:F0/F1/F2/F3:Offset 60h:bit 16 is cleared), the root port sets RSTS.PS, and log the PME Requester ID into RSTS.RID (B0:D28:F0/F1/F2/F3:Offset 60h:bits 15:0). If an interrupt is enabled via RCTL.PIE (B0:D28:F0/F1/F2/F3:Offset 5Ch:bit 3), an interrupt is generated. This interrupt can be either a pin or an MSI if MSI is enabled via MC.MSIE (B0:D28:F0/F1/F2/F3:Offset 82h:bit 0). See Section 4.1.2.4 for SMI/SCI generation. If this is a subsequent message received (RSTS.PS is already set), the root port sets RSTS.PP (B0:D28:F0/F1/F2/F3:Offset 60h:bit 17) and log the PME Requester ID from the message in a hidden register. No other action is taken. When the first PME event is cleared by software clearing RSTS.PS, the root port sets RSTS.PS, clear RSTS.PP, and move the requester ID from the hidden register into RSTS.RID. If RCTL.PIE is set, an interrupt is generated. If RCTL.PIE is not set, a message is sent to the power management controller so that a GPE can be set. If messages have been logged (RSTS.PS is set), and RCTL.PIE is later written from a 0 to a 1, and interrupt is generated. This last condition handles the case where the message was received prior to the operating system re-enabling interrupts after resuming from a low power state. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 109 4.1.2.4 SMI/SCI Generation Interrupts for power management events are not supported on legacy operating systems. To support power management on non-PCI Express* aware operating systems, PM events can be routed to generate SCI. To generate SCI, MPC.PMCE must be set. When set, a power management event causes SMSCS.PMCS (B0:D28:F0/F1/ F2/F3:Offset DCh:bit 31) to be set. BIOS workarounds for power management can be supported by setting MPC.PMME (B0:D28:F0/F1/F2/F3:Offset D8h:bit 0). When this bit is set, power management events sets SMSCS.PMMS (B0:D28:F0/F1/F2/F3:Offset DCh:bit 0), and SMI # is generated. This bit is set regardless of whether interrupts or SCI is enabled. The SMI# may occur concurrently with an interrupt or SCI. 4.1.3 SERR# Generation SERR# may be generated through PCI Express mechanisms involving bits in the PCI Express capability structure. 4.1.4 Hot-Plug Each root port implements a hot-plug controller that performs the following: * Messages to turn on / off / blink LEDs * Presence and attention button detection * Interrupt generation The root port only allows hot-plug with modules (e.g., ExpressCard*). Edge-connector based hot-plug is not supported. 4.1.4.1 Presence Detection When a module is plugged in and power is supplied, the physical layer detects the presence of the device, and the root port sets SLSTS.PDS (B0:D28:F0/F1/F2/F3:Offset 5Ah:bit 6) and SLSTS.PDC (B0:D28:F0/F1/F2/F3:Offset 6h:bit 3). If SLCTL.PDE (B0:D28:F0/F1/F2/F3:Offset 58h:bit 3) and SLCTL.HPE (B0:D28:F0/F1/F2/F3:Offset 58h:bit 5) are both set, the root port also generates an interrupt. When a module is removed (via the physical layer detection), the root port clears SLSTS.PDS and sets SLSTS.PDC. If SLCTL.PDE and SLCTL.HPE are both set, the root port also generates an interrupt. 4.1.4.2 Message Generation When system software writes to SLCTL.AIC (B0:D28:F0/F1/F2/F3:Offset 58h:bits 7:6) or SLCTL.PIC (B0:D28:F0/F1/F2/F3:Offset 58h:bits 9:8), the root port sends a message down the link to change the state of LEDs on the module. Writes to these fields are non-postable cycles, and the resulting message is a postable cycle. When receiving one of these writes, the root port performs the following: * Changes the state in the register. * Generates a completion into the upstream queue. * Formulates a message for the downstream port if the field is written to regardless of if the field changed. * Generates the message on the downstream port. Intel(R) Communications Chipset 89xx Series - Datasheet 110 October 2012 Order Number: 327879-001US 4.0 * When the last message of a command is transmitted, sets SLSTS.CCE (D28:F0/F1/ F2/F3:Offset 58h:bit 4) to indicate the command has completed. If SLCTL.CCE and SLCTL.HPE (D28:F0/F1/F2/F3:Offset 58h:bit 5) are set, the root port generates an interrupt. The command completed register (SLSTS.CC) applies only to commands issued by software to control the Attention Indicator (SLCTL.AIC), Power Indicator (SLCTL.PIC), or Power Controller (SLCTL.PCC). However, writes to other parts of the Slot Control Register would invariably end up writing to the indicators, power controller fields; Hence, any write to the Slot Control Register is considered a command and if enabled, results in a command complete interrupt. The only exception is a write to disable the command complete interrupt which does not result in a command complete interrupt. A single write to the Slot Control register is considered as a single command, and hence receives a single command complete, even if the write affects more than one field in the Slot Control Register. 4.1.4.3 Attention Button Detection When an attached device is ejected, the user can press an attention button. This results in a the PCI Express message "Attention_Button_Pressed" from the device. Upon receiving this message, the root port will set SLSTS.ABP (B0:D28:F0/F1/F2/ F3:Offset 5Ah:bit 0). If SLCTL.ABE (B0:D28:F0/F1/F2/F3:Offset 58h:bit 0) and SLCTL.HPE (B0:D28:F0/F1/ F2/F3:Offset 58h:bit 5) are set, the hot-plug controller will also generate an interrupt. The interrupt is generated on an edge-event. If SLSTS.ABP is already set, a new interrupt is not generated. 4.1.4.4 SMI/SCI Generation Interrupts for hot-plug events are not supported on legacy operating systems. To support hot-plug on non-PCI Express* aware operating systems, hot-plug events can be routed to generate SCI. To generate SCI, MPC.HPCE (B0:D28:F0/F1/F2/F3:Offset D8h:bit 30) must be set. When set, enabled hot-plug events will cause SMSCS.HPCS (B0:D28:F0/F1/F2/F3:Offset DCh:bit 30) to be set. Additionally, BIOS workarounds for hot-plug can be supported by setting MPC.HPME (B0:D28:F0/F1/F2/F3:Offset D8h:bit 1). When this bit is set, hot-plug events can cause SMI status bits in SMSCS to be set. Supported hot-plug events and their corresponding SMSCS bit are: * Command Completed - SCSCS.HPCCM (B0:D28:F0/F1/F2/F3:Offset DCh:bit 3) * Presence Detect Changed - SMSCS.HPPDM (B0:D28:F0/F1/F2/F3:Offset DCh:bit 1) * Attention Button Pressed - SMSCS.HPABM (B0:D28:F0/F1/F2/F3:Offset DCh:bit 2) * Link Active State Changed - SMSCS.HPLAS (B0:D28:F0/F1/F2/F3:Offset DCh:bit 4) When any of these bits are set, SMI # is generated. These bits are set regardless of whether interrupts or SCI is enabled for hot-plug events. The SMI# may occur concurrently with an interrupt or SCI. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 111 4.2 LPC Bridge (with System and Management Functions) (B0:D31:F0) The LPC bridge function resides in PCI Bus 0:Device 31:Function 0. In addition to the LPC bridge function, B0:D31:F0 contains other functional units including: * DMA * Interrupt controllers * Timers * Power Management * System Management * GPIO * UART * WDT * RTC In this chapter, registers and functions associated with other functional units (power management, GPIO, USB, etc.) are described in their respective sections. 4.2.1 LPC Interface The LPC interface as described in the Low Pin Count Interface Specification, Revision 1.1. The LPC interface is shown in Figure 4-1. The LPC interface implements all of the signals that are shown as optional, but peripherals are not required to do so. Figure 4-1. LPC Interface Diagram LPC Control Bus Intel(R) Communications Chipset 89xx Series - Datasheet 112 LPC PME# GPI . SERIRQ SUS_STAT# RST# PCI CLK LPC Controller LAD [3:0] LFRAME# LDRQ[1:0]# (Optional) LPCPD# (Optional) LPC Device LSMI# (Optional) October 2012 Order Number: 327879-001US 4.0 4.2.1.1 LPC Cycle Types The LPC controller implements all of the cycle types described in the Low Pin Count Interface Specification, Revision 1.1. Table 4-2 shows the cycle types supported by the LPC controller. Table 4-2. LPC Cycle Types Supported Cycle Type Memory Read Memory Write Comment 1 byte only. (See Note 1 below) 1 byte only. (See Note 1 below) I/O Read 1 byte only. LPC Controller breaks up 16- and 32-bit processor cycles into multiple 8-bit transfers. I/O Write 1 byte only. LPC Controller breaks up 16- and 32-bit processor cycles into multiple 8-bit transfers. DMA Read Can be 1, or 2 bytes DMA Write Can be 1, or 2 bytes Bus Master Read Can be 1, 2, or 4 bytes. (See Note 2 below) Bus Master Write Can be 1, 2, or 4 bytes. (See Note 2 below) Notes: 1. The LPC Controller provides a single generic memory range (LGMR) for decoding memory cycles and forwarding them as LPC Memory cycles on the LPC bus. The LGMR memory decode range is 64 KB in size and can be defined as being anywhere in the 4 GB memory space. This range needs to be configured by BIOS during POST to provide the necessary memory resources. BIOS should advertise the LPC Generic Memory Range as Reserved to the OS in order to avoid resource conflict. For larger transfers, The LPC Controller performs multiple 8-bit transfers. If the cycle is not claimed by any peripheral, it is subsequently aborted, and the LPC Controller returns a value of all 1s to the processor. This is done to maintain compatibility with ISA memory cycles where pull-up resistors would keep the bus high if no device responds. 2. Bus Master Read or Write cycles must be naturally aligned. For example, a 1-byte transfer can be to any address. However, the 2-byte transfer must be word-aligned (i.e., with an address where A0=0). A dword transfer must be dword-aligned (i.e., with an address where A1 and A0 are both 0). 4.2.1.2 Start Field Definition Table 4-3. Start Field Bit Definitions Bits[3:0] Encoding Definition 0000 Start of cycle for a generic target 0010 Grant for bus master 0 0011 Grant for bus master 1 1111 Stop/Abort: End of a cycle for a target. Note: All other encodings are RESERVED. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 113 4.2.1.3 Cycle Type / Direction (CYCTYPE + DIR) The LPC controller always drives bit 0 of this field to 0. Peripherals running bus master cycles must also drive bit 0 to 0. Table 4-4 shows the valid bit encodings. Table 4-4. 4.2.1.4 Cycle Type Bit Definitions Bits[3:2] Bit1 Definition 00 0 I/O Read 00 1 I/O Write 01 0 Memory Read 01 1 Memory Read 10 0 DMA Read 10 1 DMA Write 11 x Reserved. If a peripheral performing a bus master cycle generates this value, The LPC Controller aborts the cycle. Size Bits[3:2] are reserved. The LPC controller always drives them to 00. Peripherals running bus master cycles are supposed to drive 00 for bits 3:2, but the LPC controller ignores those bits. Bits[1:0] are encoded as listed in Table 4-5. Table 4-5. Transfer Size Bit Definition Bits[1:0] 4.2.1.5 Size 00 8-bit transfer (1 byte) 01 16-bit transfer (2 bytes) 10 Reserved. The LPC Controller never drives this combination. If a peripheral running a bus master cycle drives this combination, the Controller may abort the transfer. 11 32-bit transfer (4 bytes) SYNC Valid values for the SYNC field are shown in Table 4-6. Table 4-6. SYNC Bit Definition Bits[3:0] Indication 0000 Ready: SYNC achieved with no error. For DMA transfers, this also indicates DMA request deassertion and no more transfers desired for that channel. 0101 Short Wait: Part indicating wait-states. For bus master cycles, the controller does not use this encoding. Instead, the controller uses the Long Wait encoding (see next encoding below). 0110 Long Wait: Part indicating wait-states, and many wait-states is added. This encoding driven by the controller for bus master cycles, rather than the Short Wait (0101). 1001 Ready More (Used only by peripheral for DMA cycle): SYNC achieved with no error and more DMA transfers desired to continue after this transfer. This value is valid only on DMA transfers and is not allowed for any other type of cycle. 1010 Error: Sync achieved with error. This is generally used to replace the SERR# or IOCHK# signal on the PCI/ISA bus. It indicates that the data is to be transferred, but there is a serious error in this transfer. For DMA transfers, this not only indicates an error, but also indicates DMA request deassertion and no more transfers desired for that channel. Notes: 1. All other combinations are RESERVED. Intel(R) Communications Chipset 89xx Series - Datasheet 114 October 2012 Order Number: 327879-001US 4.0 4.2.1.6 SYNC Time-Out Several error cases can occur on the LPC interface. The controller responds as defined in section 4.2.1.9 of the Low Pin Count Interface Specification, Revision 1.1 to the stimuli described therein. There may be other peripheral failure conditions, but these are not handled by the controller. 4.2.1.7 SYNC Error Indication The LPC controller responds as defined in section 4.2.1.10 of the Low Pin Count Interface Specification, Revision 1.1. Upon recognizing the SYNC field indicating an error, the controller treats this as an SERR by reporting this into the Device 31 Error Reporting Logic. 4.2.1.8 LFRAME# Usage The controller follows the usage of LFRAME# as defined in the Low Pin Count Interface Specification, Revision 1.1. The controller performs an abort for the following cases (possible failure cases): * The controller starts a Memory, I/O, or DMA cycle, but no device drives a valid SYNC after four consecutive clocks. * The controller starts a Memory, I/O, or DMA cycle, and the peripheral drives an invalid SYNC pattern. * A peripheral drives an illegal address when performing bus master cycles. * A peripheral drives an invalid value. 4.2.1.9 I/O Cycles For I/O cycles targeting registers specified in the controller's decode ranges, the controller performs I/O cycles as defined in the Low Pin Count Interface Specification, Revision 1.1. These are 8-bit transfers. If the processor attempts a 16-bit or 32-bit transfer, the controller breaks the cycle up into multiple 8-bit transfers to consecutive I/ O addresses. Note: If the cycle is not claimed by any peripheral (and subsequently aborted), the controller returns a value of all 1s (FFh) to the processor. This is to maintain compatibility with ISA I/O cycles where pull-up resistors would keep the bus high if no device responds. 4.2.1.10 Bus Master Cycles The LPC interface supports Bus Master cycles and requests (using LDRQ#) as defined in the Low Pin Count Interface Specification, Revision 1.1. The interface has two LDRQ# inputs, and thus supports two separate bus master devices. It uses the associated START fields for Bus Master 0 (0010b) or Bus Master 1 (0011b). Note: The LPC interface does not support LPC Bus Masters performing I/O cycles. LPC Bus Masters should only perform memory read or memory write cycles. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 115 4.2.1.11 LPC Power Management LPCPD# Protocol Same timings as for SUS_STAT#. Upon driving SUS_STAT# low, LPC peripherals drive LDRQ# low or tri-state it. The interface shuts off the LDRQ# input buffers. After driving SUS_STAT# active, The LPC controller drives LFRAME# low, and tri-states (or drive low) LAD[3:0]. Note: The Low Pin Count Interface Specification, Revision 1.1 defines the LPCPD# protocol where there is at least 30 s from LPCPD# assertion to LRST# assertion. This specification explicitly states that this protocol only applies to entry/exit of low power states which does not include asynchronous reset events. The controller asserts both SUS_STAT# (connects to LPCPD#) and PLTRST# (connects to LRST#) at the same time during a global reset. This is not inconsistent with the LPC LPCPD# protocol. 4.2.1.12 Configuration and Implications LPC I/F Decoders To allow the I/O cycles and memory mapped cycles to go to the LPC interface, the interface includes several decoders. During configuration, the controller must be programmed with the same decode ranges as the peripheral. The decoders are programmed via the Device 31:Function 0 configuration space. Bus Master Device Mapping and START Fields Bus Masters must have a unique START field. In the case where the LPC interface supports two LPC bus masters, it drives 0010 for the START field for grants to bus master #0 (requested via LDRQ0#) and 0011 for grants to bus master #1 (requested via LDRQ1#.). Thus, no registers are needed to configure the START fields for a particular bus master. 4.3 DMA Operation (B0:D31:F0) LPC DMA operations are supported using an internal DMA controller. The DMA controller has registers that are fixed in the lower 64 KB of I/O space. The DMA controller is configured using registers in the PCI configuration space. These registers allow configuration of the channels for use by LPC DMA. The DMA circuitry incorporates the functionality of two 82C37 DMA controllers with seven independently programmable channels (Figure 4-2). DMA controller 1 (DMA-1) corresponds to DMA channels 0-3 and DMA controller 2 (DMA-2) corresponds to channels 5-7. DMA channel 4 is used to cascade the two controllers and defaults to cascade mode in the DMA Channel Mode (DCM) Register. Channel 4 is not available for any other purpose. In addition to accepting requests from DMA slaves, the DMA controller also responds to requests that software initiates. Software may initiate a DMA service request by setting any bit in the DMA Channel Request Register to a 1. Figure 4-2. DMA Controller Channel 4 Channel 0 Channel 1 Channel 5 DMA-1 Channel 2 Channel 6 Channel 3 Channel 7 Intel(R) Communications Chipset 89xx Series - Datasheet 116 DMA-2 October 2012 Order Number: 327879-001US 4.0 Each DMA channel is hardwired to the compatible settings for DMA device size: channels [3:0] are hardwired to 8-bit, count-by-bytes transfers, and channels [7:5] are hardwired to 16-bit, count-by-words (address shifted) transfers. The DMA controller provides 24-bit addressing in compliance with the ISA-Compatible specification. Each channel includes a 16-bit ISA-Compatible Current Register which holds the 16 least-significant bits of the 24-bit address, an ISA-Compatible Page Register which contains the eight next most significant bits of address. The DMA controller also features refresh address generation, and auto-initialization following a DMA termination. 4.3.1 Channel Priority For priority resolution, the DMA consists of two logical channel groups: channels 0-3 and channels 4-7. Each group may be in either fixed or rotate mode, as determined by the DMA Command Register. DMA I/O slaves normally assert their DREQ line to arbitrate for DMA service. However, a software request for DMA service can be presented through each channel's DMA Request Register. A software request is subject to the same prioritization as any hardware request. 4.3.1.1 Fixed Priority The initial fixed priority structure is as follows: High priority Low priority 0, 1, 2, 3 5, 6, 7 The fixed priority ordering is 0, 1, 2, 3, 5, 6, and 7. In this scheme, channel 0 has the highest priority, and channel 7 has the lowest priority. Channels [3:0] of DMA-1 assume the priority position of channel 4 in DMA-2, taking priority over channels 5, 6, and 7. 4.3.1.2 Rotating Priority Rotation allows for "fairness" in priority resolution. The priority chain rotates so that the last channel serviced is assigned the lowest priority in the channel group (0-3, 5-7). Channels 0-3 rotate as a group of 4. They are always placed between channel 5 and channel 7 in the priority list. Channel 5-7 rotate as part of a group of 4. That is, channels (5-7) form the first three positions in the rotation, while channel group (0-3) comprises the fourth position in the arbitration. 4.3.2 Address Compatibility Mode When the DMA is operating, the addresses do not increment or decrement through the High and Low Page Registers. Therefore, if a 24-bit address is 01FFFFh and increments, the next address is 010000h, not 020000h. Similarly, if a 24-bit address is 020000h and decrements, the next address is 02FFFFh, not 01FFFFh. When the DMA is operating in 16-bit mode, the addresses still do not increment or decrement through the High and Low Page Registers, but the page boundary is 128 K. Therefore, if a 24-bit address is 01FFFEh and increments, then the next address is 000000h, not 0100000h. Similarly, if a 24-bit address is 020000h and decrements, then the next address is 03FFFEh, not 02FFFEh. This is compatible with the 82C37 and Page Register implementation used in the PC-AT. This mode is set after CPURST is valid. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 117 4.3.3 Summary of DMA Transfer Sizes Table 4-7 lists the DMA device transfer sizes. "Current Byte/Word Count Register" indicates that the register contents represents either the number of bytes to transfer or the number of 16-bit words to transfer. "Current Address Increment/Decrement" indicates the number added to or taken from the Current Address register after each DMA transfer cycle. The DMA Channel Mode Register determines if the Current Address Register is incremented or decremented. 4.3.3.1 Address Shifting When Programmed for 16-Bit I/O Count by Words Table 4-7. DMA Transfer Size Current Byte/Word Count Register DMA Device Date Size And Word Count Current Address Increment/Decrement 8-Bit I/O, Count By Bytes Bytes 1 16-Bit I/O, Count By Words (Address Shifted) Words 1 The DMA interface maintains compatibility with the implementation of the DMA in the PC AT that used the 82C37. The DMA shifts the addresses for transfers to/from a 16-bit device count-by-words. Note: The least significant bit of the Low Page Register is dropped in 16-bit shifted mode. When programming the Current Address Register (when the DMA channel is in this mode), the Current Address must be programmed to an even address with the address value shifted right by one bit. The address shifting is shown in Table 4-8. Table 4-8. Address Shifting in 16-Bit I/O DMA Transfers Note: 4.3.4 Output Address 8-Bit I/O Programmed Address (Ch 0-3) 16-Bit I/O Programmed Address (Ch 5-7) (Shifted) A0 A[16:1] A[23:17] A0 A[16:1] A[23:17] 0 A[15:0] A[23:17] The least significant bit of the Page Register is dropped in 16-bit shifted mode. Autoinitialize By programming a bit in the DMA Channel Mode Register, a channel may be set up as an autoinitialize channel. When a channel undergoes autoinitialization, the original values of the Current Page, Current Address and Current Byte/Word Count Registers are automatically restored from the Base Page, Address, and Byte/Word Count Registers of that channel following TC. The Base Registers are loaded simultaneously with the Current Registers by the microprocessor when the DMA channel is programmed and remain unchanged throughout the DMA service. The mask bit is not set when the channel is in autoinitialize. Following autoinitialize, the channel is ready to perform another DMA service, without processor intervention, as soon as a valid DREQ is detected. Intel(R) Communications Chipset 89xx Series - Datasheet 118 October 2012 Order Number: 327879-001US 4.0 4.3.5 Software Commands There are three additional special software commands that the DMA controller can execute. The three software commands are: * Clear Byte Pointer Flip-Flop * Master Clear * Clear Mask Register They do not depend on any specific bit pattern on the data bus. 4.4 LPC DMA DMA on LPC is handled through the use of the LDRQ# lines from peripherals and special encodings on LAD[3:0] from the host. Single, Demand, Verify, and Increment modes are supported on the LPC interface. Channels 0-3 are 8 bit channels. Channels 5-7 are 16-bit channels. Channel 4 is reserved as a generic bus master request. 4.4.1 Asserting DMA Requests Peripherals that need DMA service encode their requested channel number on the LDRQ# signal. To simplify the protocol, each peripheral on the LPC I/F has its own dedicated LDRQ# signal (they may not be shared between two separate peripherals). The LPC interface has two LDRQ# inputs, allowing at least two devices to support DMA or bus mastering. LDRQ# is synchronous with LCLK (PCI clock). As shown in Figure 4-3, the peripheral uses the following serial encoding sequence: * Peripheral starts the sequence by asserting LDRQ# low (start bit). LDRQ# is high during idle conditions. * The next three bits contain the encoded DMA channel number (MSB first). * The next bit (ACT) indicates whether the request for the indicated DMA channel is active or inactive. The ACT bit is 1 (high) to indicate if it is active and 0 (low) if it is inactive. The case where ACT is low is rare, and is only used to indicate that a previous request for that channel is being abandoned. * After the active/inactive indication, the LDRQ# signal must go high for at least 1 clock. After that one clock, LDRQ# signal can be brought low to the next encoding sequence. If another DMA channel also needs to request a transfer, another sequence can be sent on LDRQ#. For example, if an encoded request is sent for channel 2, and then channel 3 needs a transfer before the cycle for channel 2 is run on the interface, the peripheral can send the encoded request for channel 3. This allows multiple DMA agents behind an I/O device to request use of the LPC interface, and the I/O device does not need to selfarbitrate before sending the message. Figure 4-3. DMA Request Assertion through LDRQ# LCLK LDRQ# October 2012 Order Number: 327879-001US Start MSB LSB ACT Start Intel(R) Communications Chipset 89xx Series - Datasheet 119 4.4.2 Abandoning DMA Requests DMA Requests can be deasserted in two fashions: on error conditions by sending an LDRQ# message with the `ACT' bit set to 0, or normally through a SYNC field during the DMA transfer. This section describes boundary conditions where the DMA request needs to be removed prior to a data transfer. There may be some special cases where the peripheral desires to abandon a DMA transfer. The most likely case of this occurring is due to a floppy disk controller which has overrun or underrun its FIFO, or software stopping a device prematurely. In these cases, the peripheral wishes to stop further DMA activity. It may do so by sending an LDRQ# message with the ACT bit as 0. However, since the DMA request was seen by the controller, there is no guarantee that the cycle has not been granted and will shortly run on LPC. Therefore, peripherals must take into account that a DMA cycle may still occur. The peripheral can choose not to respond to this cycle, in which case the host will abort it, or it can choose to complete the cycle normally with any random data. This method of DMA deassertion should be prevented whenever possible, to limit boundary conditions both on the controller and the peripheral. 4.4.3 General Flow of DMA Transfers Arbitration for DMA channels is performed through the 8237 within the host. Once the host has won arbitration on behalf of a DMA channel assigned to LPC, it asserts LFRAME# on the LPC I/F and begins the DMA transfer. The general flow for a basic DMA transfer between the controller and an LPC peripheral is as follows: 1. The controller starts transfer by asserting 0000b on LAD[3:0] with LFRAME# asserted. 2. The controller asserts `cycle type' of DMA, direction based on DMA transfer direction. 3. The controller asserts channel number and, if applicable, terminal count. 4. The controller indicates the size of the transfer: 8 or 16 bits. 5. If a DMA reads: -- The controller drives the first 8 bits of data and turns the bus around. -- The peripheral acknowledges the data with a valid SYNC. -- If a 16-bit transfer, the process is repeated for the next 8 bits. 6. If a DMA writes: -- The controller turns the bus around and waits for data. -- The peripheral indicates data ready through SYNC and transfers the first byte. -- If a 16-bit transfer, the peripheral indicates data ready and transfers the next byte. 7. The peripheral turns the bus around. Intel(R) Communications Chipset 89xx Series - Datasheet 120 October 2012 Order Number: 327879-001US 4.0 4.4.4 Terminal Count Terminal count is communicated through LAD[3] on the same clock that DMA channel is communicated on LAD[2:0]. This field is the CHANNEL field. Terminal count indicates the last byte of transfer, based upon the size of the transfer. For example, on an 8-bit transfer size (SIZE field is 00b), if the TC bit is set, then this is the last byte. On a 16-bit transfer (SIZE field is 01b), if the TC bit is set, then the second byte is the last byte. The peripheral, therefore, must internalize the TC bit when the CHANNEL field is communicated, and only signal TC when the last byte of that transfer size has been transferred. 4.4.5 Verify Mode Verify mode is supported on the LPC interface. A verify transfer to the peripheral is similar to a DMA write, where the peripheral is transferring data to main memory. The indication from the host is the same as a DMA write, so the peripheral is driving data onto the LPC interface. However, the host does not transfer this data into main memory. 4.4.6 DMA Request Deassertion An end of transfer is communicated to the controller through a special SYNC field transmitted by the peripheral. An LPC device must not attempt to signal the end of a transfer by deasserting LDREQ#. If a DMA transfer is several bytes (e.g., a transfer from a demand mode device) the controller needs to know when to deassert the DMA request based on the data currently being transferred. The DMA agent uses a SYNC encoding on each byte of data being transferred, which indicates to the controller whether this is the last byte of transfer or if more bytes are requested. To indicate the last byte of transfer, the peripheral uses a SYNC value of 0000b (ready with no error), or 1010b (ready with error). These encodings tell the controller that this is the last piece of data transferred on a DMA read (controller to peripheral), or the byte that follows is the last piece of data transferred on a DMA write (peripheral to controller). When the controller sees one of these two encodings, it ends the DMA transfer after this byte and deasserts the DMA request to the 8237. Therefore, if the controller indicated a 16-bit transfer, the peripheral can end the transfer after one byte by indicating a SYNC value of 0000b or 1010b. The controller does not attempt to transfer the second byte, and deasserts the DMA request internally. If the peripheral indicates a 0000b or 1010b SYNC pattern on the last byte of the indicated size, then the controller only deasserts the DMA request to the 8237 since it does not need to end the transfer. If the peripheral wishes to keep the DMA request active, then it uses a SYNC value of 1001b (ready plus more data). This tells the 8237 that more data bytes are requested after the current byte has been transferred, so the controller keeps the DMA request active to the 8237. Therefore, on an 8-bit transfer size, if the peripheral indicates a SYNC value of 1001b to the controller, the data is transferred and the DMA request will remain active to the 8237. The controller will return lager with another START- CYCTYPE-CHANNEL-SIZE etc. combination to initiate another transfer to the peripheral. The peripheral must not assume that the next START indication from the controller is another grant to the peripheral if it had indicated a SYNC value of 1001b. On a single mode DMA device, the 8237 will re-arbitrate after every transfer. Only demand mode DMA devices can be guaranteed that they will receive the next START indication from the controller. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 121 Note: Indicating a 0000b or 1010b encoding on the SYNC field of an odd byte of a 16-bit channel (first byte of a 16-bit transfer) is an error condition. Note: The host stops the transfer on the LPC bus as indicated, fills the upper byte with random data on DMA writes (peripheral to memory), and indicates to the 8237 that the DMA transfer occurred, incrementing the 8237's address and decrementing its byte count. 4.4.7 SYNC Field / LDRQ# Rules Since DMA transfers on LPC are requested through an LDRQ# assertion message, and are ended through a SYNC field during the DMA transfer, the peripheral must obey the following rule when initiating back-to-back transfers from a DMA channel. The peripheral must not assert another message for eight LCLKs after a deassertion is indicated through the SYNC field. This is needed to allow the 8237, that typically runs off a much slower internal clock, to see a message deasserted before it is re-asserted so that it can arbitrate to the next agent. Under default operation, the host only performs 8-bit transfers on 8-bit channels and 16-bit transfers on 16-bit channels. The method by which this communication between host and peripheral through system BIOS is performed is beyond the scope of this specification. Since the LPC host and LPC peripheral are motherboard devices, no "plug-n-play" registry is required. The peripheral must not assume that the host is able to perform transfer sizes that are larger than the size allowed for the DMA channel, and be willing to accept a SIZE field that is smaller than what it may currently have buffered. To that end, it is recommended that future devices that may appear on the LPC bus, that require higher bandwidth than 8-bit or 16-bit DMA allow, do so with a bus mastering interface and not rely on the 8237. 4.5 8254 Timers (B0:D31:F0) There are three counters that have fixed uses. All registers and functions associated with the 8254 timers are in the core well. The 8254 unit is clocked by a 14.31818 MHz clock. Counter 0, System Timer This counter functions as the system timer by controlling the state of IRQ0 and is typically programmed for Mode 3 operation. The counter produces a square wave with a period equal to the product of the counter period (838 ns) and the initial count value. The counter loads the initial count value 1 counter period after software writes the count value to the counter I/O address. The counter initially asserts IRQ0 and decrements the count value by two each counter period. The counter negates IRQ0 when the count value reaches 0. It then reloads the initial count value and again decrements the initial count value by two each counter period. The counter then asserts IRQ0 when the count value reaches 0, reloads the initial count value, and repeats the cycle, alternately asserting and negating IRQ0. Counter 1, Refresh Request Signal This counter provides the refresh request signal and is typically programmed for Mode 2 operation and only impacts the period of the REF_TOGGLE bit in Port 61. The initial count value is loaded one counter period after being written to the counter I/O address. Intel(R) Communications Chipset 89xx Series - Datasheet 122 October 2012 Order Number: 327879-001US 4.0 The REF_TOGGLE bit has a square wave behavior (alternate between 0 and 1) and will toggle at a rate based on the value in the counter. Programming the counter to anything other than Mode 2 results in undefined behavior for the REF_TOGGLE bit. 4.5.1 Timer Programming The counter/timers are programmed as follows: 1. Write a control word to select a counter. 2. Write an initial count for the selected counter. 3. Load the least and/or most significant bytes (as required by Control Word bits 5, 4) of the 16-bit counter. 4. Repeat with other counters. Two conventions must be observed when programming the counters. First, for each counter, the control word must be written before the initial count is written. Second, the initial count must follow the count format specified in the control word (least significant byte only, most significant byte only, or least significant byte and then most significant byte). A new initial count may be written to a counter at any time without affecting the counter's programmed mode. Counting is affected as described in the mode definitions. The new count must follow the programmed count format. If a counter is programmed to read/write two-byte counts, the following precaution applies: A program must not transfer control between writing the first and second byte to another routine which also writes into that same counter. Otherwise, the counter is loaded with an incorrect count. The Control Word Register at port 43h controls the operation of all three counters. Several commands are available: * Control Word Command. Specifies which counter to read or write, the operating mode, and the count format (binary or BCD). * Counter Latch Command. Latches the current count so that it can be read by the system. The countdown process continues. * Read Back Command. Reads the count value, programmed mode, the current state of the OUT pins, and the state of the Null Count Flag of the selected counter. Table 4-9 lists the six operating modes for the interval counters. Table 4-9. Counter Operating Modes Mode Function Description 0 Out signal on end of count (=0) Output is 0. When count goes to 0, output goes to 1 and stays at 1 until counter is reprogrammed. 1 Hardware retriggerable one-shot Output is 0. When count goes to 0, output goes to 1 for one clock time. 2 Rate generator (divide by n counter) Output is 1. Output goes to 0 for one clock time, then back to 1 and counter is reloaded. 3 Square wave output Output is 1. Output goes to 0 when counter rolls over, and counter is reloaded. Output goes to 1 when counter rolls over, and counter is reloaded, etc. 4 Software triggered strobe Output is 1. Output goes to 0 when count expires for one clock time. 5 Hardware triggered strobe Output is 1. Output goes to 0 when count expires for one clock time. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 123 4.5.2 Reading from the Interval Timer It is often desirable to read the value of a counter without disturbing the count in progress. There are three methods for reading the counters: a simple read operation, counter Latch command, and the Read-Back command. With the simple read and counter latch command methods, the count must be read according to the programmed format; specifically, if the counter is programmed for two byte counts, two bytes must be read. The two bytes do not have to be read one right after the other. Read, write, or programming operations for other counters may be inserted between them. 4.5.2.1 Simple Read With a simple read operation, the counter is selected through port 40h (counter 0), 41h (counter 1), or 42h (counter 2). Performing a direct read from the counter does not return a determinate value, because the counting process is asynchronous to read operations. However, in the case of counter 2, the count can be stopped by writing to the GATE bit in port 61h. 4.5.2.2 Counter Latch Command The Counter Latch command, written to port 43h, latches the count of a specific counter when the command is received. This command ensures that the count read from the counter is accurate, particularly when reading a two-byte count. The count value is then read from each counter's Count register as was programmed by the Control register. The count is held in the latch until it is read or the counter is reprogrammed. The count is then unlatched. This allows reading the contents of the counters on the fly without affecting counting in progress. Multiple Counter Latch Commands may be used to latch more than one counter. Counter Latch commands do not affect the programmed mode of the counter in any way. If a Counter is latched and then latched again before the count is read, the second Counter Latch command is ignored. The count read is the count when the first Counter Latch command was issued. 4.5.2.3 Read Back Command The Read Back command, written to port 43h, latches the count value, programmed mode, and current states of the OUT pin and Null Count flag of the selected counter or counters. The value of the counter and its status may then be read by I/O access to the counter address. The Read Back command may be used to latch multiple counter outputs at one time. This single command is functionally equivalent to several counter latch commands, one for each counter latched. Each counter's latched count is held until it is read or reprogrammed. Once read, a counter is unlatched. The other counters remain latched until they are read. If multiple count Read Back commands are issued to the same counter without reading the count, all but the first are ignored. The Read Back command may additionally be used to latch status information of selected counters. The status of a counter is accessed by a read from that counter's I/O port address. If multiple counter status latch operations are performed without reading the status, all but the first are ignored. Intel(R) Communications Chipset 89xx Series - Datasheet 124 October 2012 Order Number: 327879-001US 4.0 Both count and status of the selected counters may be latched simultaneously. This is functionally the same as issuing two consecutive, separate Read Back commands. If multiple count and/or status Read Back commands are issued to the same counters without any intervening reads, all but the first are ignored. If both count and status of a counter are latched, the first read operation from that counter returns the latched status, regardless of which was latched first. The next one or two reads, depending on whether the counter is programmed for one or two type counts, returns the latched count. Subsequent reads return unlatched count. 4.6 8259 Interrupt Controllers (PIC) (B0:D31:F0) The system incorporates the functionality of two 8259 interrupt controllers that provide interrupts for ISA compatible interrupts. These interrupts are: system timer, keyboard controller, serial ports, parallel ports, floppy disk, mouse, and DMA channels. In addition, this interrupt controller can support the PCI based interrupts, by mapping the PCI interrupt onto the compatible ISA interrupt line. Each 8259 core supports eight interrupts, numbered 0-7. Table 4-10 shows how the cores are connected. . Table 4-10. Interrupt Controller Core Connections 8259 8259 Input 0 Master Slave Note: Typical Interrupt Source Internal Connected Pin / Function Internal Timer / Counter 0 output / HPET #0 1 Keyboard IRQ1 via SERIRQ 2 Internal Slave controller INTR output 3 Serial Port A IRQ3 via SERIRQ, PIRQ# 4 Serial Port B IRQ4 via SERIRQ, PIRQ# 5 Parallel Port / Generic IRQ5 via SERIRQ, PIRQ# 6 Floppy Disk IRQ6 via SERIRQ, PIRQ# 7 Parallel Port / Generic IRQ7 via SERIRQ, PIRQ# 0 Internal Real Time Clock Internal RTC / HPET #1 1 Generic IRQ9 via SERIRQ, SCI, TCO, or PIRQ# 2 Generic IRQ10 via SERIRQ, SCI, TCO, or PIRQ# 3 Generic IRQ11 via SERIRQ, SCI, TCO, or PIRQ#, or HPET #2 4 PS/2 Mouse IRQ12 via SERIRQ, SCI, TCO, or PIRQ#, or HPET #3 5 Internal State Machine output based on processor FERR# assertion. May optionally be used for SCI or TCO interrupt if FERR# not needed. 6 SATA IRQ14 via SERIRQ or PIRQ# 7 SATA IRQ15 via SERIRQ or PIRQ# SERIRQ and PIRQ are not externally routed. They are used to reference internally routed PCIe* interrupts. The system cascades the slave controller onto the master controller through master controller interrupt input 2. This leaves 15 possible interrupts for the PIC. Interrupts can individually be programmed to be edge or level, except for IRQ0, IRQ2, IRQ8#, and IRQ13. Active-low interrupt sources (e.g., the PIRQ#s) are inverted internally. In the following descriptions of the 8259s, the interrupt levels are in reference to the signals at the internal interface of the 8259s after the required inversions have occurred. Therefore, the term "high" indicates "active," which means "low" on an originating PIRQ#. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 125 4.6.1 Interrupt Handling 4.6.1.1 Generating Interrupts The PIC interrupt sequence involves three bits, from the IRR, ISR, and IMR, for each interrupt level. These bits are used to determine the interrupt vector returned, and status of any other pending interrupts. Table 4-11 defines the IRR, ISR, and IMR. Table 4-11. Interrupt Status Registers Bit 4.6.1.2 Description IRR Interrupt Request Register - This bit is set on a low to high transition of the interrupt line in edge mode, and by an active high level in level mode. This bit is set whether or not the interrupt is masked. However, a masked interrupt does not generate INTR. ISR Interrupt Service Register - This bit is set, and the corresponding IRR bit cleared, when an interrupt acknowledge cycle is seen, and the vector returned is for that interrupt. IMR Interrupt Mask Register - This bit determines whether an interrupt is masked. Masked interrupts do not generate INTRs. Acknowledging Interrupts The processor generates an interrupt acknowledge cycle that is translated by the host bridge into a PCI Interrupt Acknowledge Cycle. The PIC translates this command into two internal INTA# pulses expected by the 8259 cores. The PIC uses the first internal INTA# pulse to freeze the state of the interrupts for priority resolution. On the second INTA# pulse, the master or slave sends the interrupt vector to the processor with the acknowledged interrupt code. This code is based upon bits [7:3] of the corresponding ICW2 register, combined with three bits representing the interrupt within that controller. Table 4-12. Content of Interrupt Vector Byte Master, Slave Interrupt Bits [2:0] IRQ7,15 111 IRQ6,14 110 IRQ5,13 101 IRQ4,12 IRQ3,11 4.6.1.3 Bits [7:3] ICW2[7:3] 100 011 IRQ2,10 010 IRQ1,9 001 IRQ0,8 000 Hardware/Software Interrupt Sequence 1. One or more of the Interrupt Request lines (IRQ) are raised high in edge mode, or seen high in level mode, setting the corresponding IRR bit. 2. The PIC sends INTR active to the processor if an asserted interrupt is not masked. 3. The processor acknowledges the INTR and responds with an interrupt acknowledge cycle. The cycle is translated into a PCI interrupt acknowledge cycle by the host bridge. This command is broadcast over PCI. 4. Upon observing its own interrupt acknowledge cycle on PCI, the cycle is converted into two cycles that the internal 8259 pair can respond to. Each cycle appears as an interrupt acknowledge pulse on the internal INTA# pin of the cascaded interrupt controllers. Intel(R) Communications Chipset 89xx Series - Datasheet 126 October 2012 Order Number: 327879-001US 4.0 5. Upon receiving the first internally generated INTA# pulse, the highest priority ISR bit is set and the corresponding IRR bit is reset. On the trailing edge of the first pulse, a slave identification code is broadcast by the master to the slave on a private, internal three bit wide bus. The slave controller uses these bits to determine if it must respond with an interrupt vector during the second INTA# pulse. 6. Upon receiving the second internally generated INTA# pulse, the PIC returns the interrupt vector. If no interrupt request is present because the request was too short in duration, the PIC returns vector 7 from the master controller. 7. This completes the interrupt cycle. In AEOI mode the ISR bit is reset at the end of the second INTA# pulse. Otherwise, the ISR bit remains set until an appropriate EOI command is issued at the end of the interrupt subroutine. 4.6.2 Initialization Command Words (ICWx) Before the operation can begin, each 8259 must be initialized. This is a four-byte sequence. The four initialization command words are referred to by their acronyms: ICW1, ICW2, ICW3, and ICW4. The base address for each 8259 initialization command word is a fixed location in the I/O memory space: 20h for the master controller and A0h for the slave controller. 4.6.2.1 ICW1 An I/O write to the master or slave controller base address with data bit 4 equal to 1 is interpreted as a write to ICW1. Upon sensing this write, the PIC expects three more byte writes to 21h for the master controller, or A1h for the slave controller, to complete the ICW sequence. A write to ICW1 starts the initialization sequence during which the following automatically occur: 1. Following initialization, an interrupt request (IRQ) input must make a low-to-high transition to generate an interrupt. 2. The Interrupt Mask Register is cleared. 3. IRQ7 input is assigned priority 7. 4. The slave mode address is set to 7. 5. Special mask mode is cleared and Status Read is set to IRR. 4.6.2.2 ICW2 The second write in the sequence (ICW2) is programmed to provide bits [7:3] of the interrupt vector that is released during an interrupt acknowledge. A different base is selected for each interrupt controller. 4.6.2.3 ICW3 The third write in the sequence (ICW3) has a different meaning for each controller: * For the master controller, ICW3 is used to indicate which IRQ input line is used to cascade the slave controller. In this case, IRQ2 is used. Therefore, bit 2 of ICW3 on the master controller is set to a 1, and the other bits are set to 0s. * For the slave controller, ICW3 is the slave identification code used during an interrupt acknowledge cycle. On interrupt acknowledge cycles, the master controller broadcasts a code to the slave controller if the cascaded interrupt won arbitration on the master controller. The slave controller compares this October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 127 identification code to the value stored in its ICW3, and if it matches, the slave controller assumes responsibility for broadcasting the interrupt vector. 4.6.2.4 ICW4 The final write in the sequence (ICW4) must be programmed for both controllers. At the very least, bit 0 must be set to a 1 to indicate that the controllers are operating in an Intel Architecture-based system. 4.6.3 Operation Command Words (OCW) These command words reprogram the Interrupt controller to operate in various interrupt modes. * OCW1 masks and unmasks interrupt lines. * OCW2 controls the rotation of interrupt priorities when in rotating priority mode, and controls the EOI function. * OCW3 sets up ISR/IRR reads, enables/disables the special mask mode (SMM), and enables/disables polled interrupt mode. 4.6.4 Modes of Operation 4.6.4.1 Fully Nested Mode In this mode, interrupt requests are ordered in priority from 0 through 7, with 0 being the highest. When an interrupt is acknowledged, the highest priority request is determined and its vector placed on the bus. Additionally, the ISR for the interrupt is set. This ISR bit remains set until. The processor issues an EOI command immediately before returning from the service routine, or if in AEOI mode, on the trailing edge of the second INTA#. While the ISR bit is set, all further interrupts of the same or lower priority are inhibited, while higher levels generate another interrupt. Interrupt priorities can be changed in the rotating priority mode. 4.6.4.2 Special Fully Nested Mode This mode is used in the case of a system where cascading is used, and the priority has to be conserved within each slave. In this case, the special fully nested mode is programmed to the master controller. This mode is similar to the fully nested mode with the following exceptions: * When an interrupt request from a certain slave is in service, this slave is not locked out from the master's priority logic and further interrupt requests from higher priority interrupts within the slave are recognized by the master and initiate interrupts to the processor. In the normal-nested mode, a slave is masked out when its request is in service. * When exiting the Interrupt Service routine, software has to check whether the interrupt serviced was the only one from that slave. This is done by sending a NonSpecific EOI command to the slave and then reading its ISR. If it is 0, a nonspecific EOI can also be sent to the master. 4.6.4.3 Automatic Rotation Mode (Equal Priority Devices) In some applications, there are a number of interrupting devices of equal priority. Automatic rotation mode provides for a sequential 8-way rotation. In this mode, a device receives the lowest priority after being serviced. In the worst case, a device requesting an interrupt has to wait until each of seven other devices are serviced at most once. Intel(R) Communications Chipset 89xx Series - Datasheet 128 October 2012 Order Number: 327879-001US 4.0 There are two ways to accomplish automatic rotation using OCW2; the rotation on NonSpecific EOI Command (R=1, SL=0, EOI=1) and the rotate in automatic EOI mode that is set by (R=1, SL=0, EOI=0). 4.6.4.4 Specific Rotation Mode (Specific Priority) Software can change interrupt priorities by programming the bottom priority. For example, if IRQ5 is programmed as the bottom priority device, then IRQ6 is the highest priority device. The Set Priority command is issued in OCW2 to accomplish this, where: R=1, SL=1, and LO-L2 is the binary priority level code of the bottom priority device. In this mode, internal status is updated by software control during OCW2. However, it is independent of the EOI command. Priority changes can be executed during an EOI command by using the Rotate on Specific EOI command in OCW2 (R=1, SL=1, EOI=1 and LO-L2=IRQ level to receive bottom priority. 4.6.4.5 Poll Mode Poll mode can be used to conserve space in the interrupt vector table. Multiple interrupts that can be serviced by one interrupt service routine do not need separate vectors if the service routine uses the poll command. Poll mode can also be used to expand the number of interrupts. The polling interrupt service routine can call the appropriate service routine, instead of providing the interrupt vectors in the vector table. In this mode, the INTR output is not used and the microprocessor internal Interrupt Enable flip-flop is reset, disabling its interrupt input. Service to devices is achieved by software using a Poll command. The Poll command is issued by setting P=1 in OCW3. The PIC treats its next I/O read as an interrupt acknowledge, sets the appropriate ISR bit if there is a request, and reads the priority level. Interrupts are frozen from the OCW3 write to the I/O read. The byte returned during the I/O read contains a 1 in bit 7 if there is an interrupt, and the binary code of the highest priority level in bits 2:0. 4.6.4.6 Cascade Mode The PIC has one master 8259 and one slave 8259 cascaded onto the master through IRQ2. This configuration can handle up to 15 separate priority levels. The master controls the slaves through a three bit internal bus. When the master drives 010b on this bus, the slave controller takes responsibility for returning the interrupt vector. An EOI command must be issued twice: once for the master and once for the slave. 4.6.4.7 Edge and Level Triggered Mode In ISA systems this mode is programmed using bit 3 in ICW1, which sets level or edge for the entire controller. This bit is disabled and a new register for edge and level triggered mode selection, per interrupt input, is included. This is the Edge/Level Control Registers ELCR1 and ELCR2. If an ELCR bit is 0, an interrupt request is recognized by a low-to-high transition on the corresponding IRQ input. The IRQ input can remain high without generating another interrupt. If an ELCR bit is 1, an interrupt request is recognized by a high level on the corresponding IRQ input and there is no need for an edge detection. The interrupt request must be removed before the EOI command is issued to prevent a second interrupt from occurring. In both the edge and level triggered modes, the IRQ inputs must remain active until after the falling edge of the first internal INTA#. If the IRQ input goes inactive before this time, a default IRQ7 vector is returned. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 129 4.6.4.8 End of Interrupt (EOI) Operations An EOI can occur in one of two fashions: by a command word write issued to the PIC before returning from a service routine, the EOI command, or automatically when AEOI bit in ICW4 is set to 1. 4.6.4.9 Normal End of Interrupt In normal EOI, software writes an EOI command before leaving the interrupt service routine to mark the interrupt as completed. There are two forms of EOI commands: Specific and Non-Specific. When a Non-Specific EOI command is issued, the PIC clears the highest ISR bit of those that are set to 1. Non-Specific EOI is the normal mode of operation of the PIC, as the interrupt being serviced currently is the interrupt entered with the interrupt acknowledge. When the PIC is operated in modes that preserve the fully nested structure, software can determine which ISR bit to clear by issuing a Specific EOI. An ISR bit that is masked is not cleared by a Non-Specific EOI if the PIC is in the special mask mode. An EOI command must be issued for both the master and slave controller. 4.6.4.10 Automatic End of Interrupt Mode In this mode, the PIC automatically performs a Non-Specific EOI operation at the trailing edge of the last interrupt acknowledge pulse. From a system standpoint, this mode should be used only when a nested multi-level interrupt structure is not required within a single PIC. The AEOI mode can only be used in the master controller and not the slave controller. 4.6.5 Masking Interrupts 4.6.5.1 Masking on an Individual Interrupt Request Each interrupt request can be masked individually by the Interrupt Mask Register (IMR). This register is programmed through OCW1. Each bit in the IMR masks one interrupt channel. Masking IRQ2 on the master controller masks all requests for service from the slave controller. 4.6.5.2 Special Mask Mode Some applications may require an interrupt service routine to dynamically alter the system priority structure during its execution under software control. For example, the routine may wish to inhibit lower priority requests for a portion of its execution but enable some of them for another portion. The special mask mode enables all interrupts not masked by a bit set in the Mask register. Normally, when an interrupt service routine acknowledges an interrupt without issuing an EOI to clear the ISR bit, the interrupt controller inhibits all lower priority requests. In the special mask mode, any interrupts may be selectively enabled by loading the Mask Register with the appropriate pattern. The special mask mode is set by OCW3 where: SSMM=1, SMM=1, and cleared where SSMM=1, SMM=0. 4.6.6 Steering PCI Interrupts The PIC can be programmed to allow PIRQA#-PIRQH# to be routed internally to interrupts 3-7, 9-12, 14 or 15. The assignment is programmable through the PIRQx Route Control registers, located at 60-63h and 68-6Bh in Device 31:Function 0. One or more PIRQx# lines can be routed to the same IRQx input. If interrupt steering is not required, the Route registers can be programmed to disable steering. Intel(R) Communications Chipset 89xx Series - Datasheet 130 October 2012 Order Number: 327879-001US 4.0 The PIRQx# lines are defined as active low, level sensitive to allow multiple interrupts on a PCI board to share a single line across the connector. When a PIRQx# is routed to specified IRQ line, software must change the IRQ's corresponding ELCR bit to level sensitive mode. Active-low interrupts (PIRQx#) can be internally inverted to send an active high level to the PIC. When a PCI interrupt is routed onto the PIC, the selected IRQ can no longer be used by an active high device (through SERIRQ). However, active low interrupts can share their interrupt with PCI interrupts. Internal sources of the PIRQs, including SCI and TCO interrupts, cause the external PIRQ to be asserted. The PIRQ input, like all of the other external sources, are routed accordingly. 4.7 Advanced Programmable Interrupt Controller (APIC) (B0:D31:F0) In addition to the standard ISA-compatible PIC described in the previous chapter, the system incorporates the APIC. While the standard interrupt controller is intended for use in a uni-processor system, APIC can be used in either a uni-processor or multiprocessor system. 4.7.1 Interrupt Handling The I/O APIC handles interrupts differently than the 8259. Briefly, these differences are: * Method of Interrupt Transmission - The I/O APIC transmits interrupts through memory writes on the normal data path to the processor, and interrupts are handled without the need for the processor to run an interrupt acknowledge cycle. * Interrupt Priority - The priority of interrupts in the I/O APIC is independent of the interrupt number. For example, interrupt 10 can be given a higher priority than interrupt 3. * More Interrupts - The I/O APIC supports a total of 24 interrupts. * Multiple Interrupt Controllers - The I/O APIC architecture allows for multiple I/O APIC devices in the system with their own interrupt vectors. 4.7.2 Interrupt Mapping The I/O APIC supports 24 APIC interrupts. Each interrupt has its own unique vector assigned by software. The interrupt vectors are mapped as follows and match "Config 6" of the Multi-Processor Specification. Table 4-13. APIC Interrupt Mapping1 (Sheet 1 of 2) IRQ # Via SERIRQ Via PCI Message Internal Modules 0 No No 1 Yes Yes 2 No No 3 Yes Yes 4 Yes Yes 5 Yes Yes 6 Yes Yes 7 Yes Yes 8 No No RTC, HPET #1 (legacy mode) 9 Yes Yes Option for SCI, TCO October 2012 Order Number: 327879-001US Cascade from 8259 #1 8254 Counter 0, HPET #0 (legacy mode) Intel(R) Communications Chipset 89xx Series - Datasheet 131 Table 4-13. APIC Interrupt Mapping1 (Sheet 2 of 2) IRQ # Via SERIRQ Via PCI Message 10 Yes Yes 11 Yes Yes HPET #2, Option for SCI, TCO (Note 2) 12 Yes Yes HPET #3 (Note 3) Internal Modules Option for SCI, TCO 13 No No FERR# logic 14 Yes Yes SATA 15 Yes Yes SATA 16 PIRQA# Yes Internal devices are routable Yes Option for SCI, TCO, HPET #0,1,2, 3. Other internal devices are routable 17 PIRQB# 18 PIRQC# 19 PIRQD# 20 N/A 21 N/A 22 N/A 23 N/A Notes: 1. PIRQ is not externally routed rather they are used to reference internally routed PCIe* interrupts. 2. When programming the polarity of internal interrupt sources on the APIC, interrupts 0 through 15 receive active-high internal interrupt sources, while interrupts 16 through 23 receive active-low internal interrupt sources. 3. If IRQ 11 is used for HPET #2, software should ensure IRQ 11 is not shared with any other devices to guarantee the proper operation of HPET #2. The hardware does not prevent sharing of IRQ 11. 4. If IRQ 12 is used for HPET #3, software should ensure IRQ 12 is not shared with any other devices to guarantee the proper operation of HPET #3. The hardware does not prevent sharing of IRQ 12. 4.7.3 PCI Express* Message-Based Interrupts When external devices through PCI Express* wish to generate an interrupt, they send the message defined in the PCI Express* Base Specification, Revision 1.0a for generating INTA# - INTD#. These are translated internal assertions/de-assertions of INTA# - INTD#. 4.7.4 IOxAPIC Address Remapping To support Intel(R) Virtualization Technology, interrupt messages are required to go through similar address remapping as any other memory request. Address remapping allows for domain isolation for interrupts, so a device assigned in one domain is not allowed to generate an interrupt to another domain. The address remapping is based on the Bus: Device: Function field associated with the requests. The internal APIC is required to initiate the interrupt message using a unique Bus: Device: function. The system BIOS allows for the programming of the unique Bus: Device: Function address for the internal APIC. This address field does not change the APIC functionality and the APIC is not promoted as a stand-alone PCI device. See Bus 0: Device 31: Function 0 Offset 6Ch for additional information. 4.8 Serial Interrupt (B0:D31:F0) Serial Interrupt protocol is supported. This allows a single signal to be used to report interrupt requests. The signal used to transmit this information is shared between the host and all peripherals that support serial interrupts. The signal line, SERIRQ, is Intel(R) Communications Chipset 89xx Series - Datasheet 132 October 2012 Order Number: 327879-001US 4.0 synchronous to PCI clock, and follows the sustained tri-state protocol that is used by all PCI signals. This means that if a device has driven SERIRQ low, it first drives it high synchronous to PCI clock and release it the following PCI clock. The serial IRQ protocol defines this sustained tri-state signaling in the following fashion: * S - Sample Phase. Signal driven low * R - Recovery Phase. Signal driven high * T - Turn-around Phase. Signal released The SERIRQ interface supports a message for 21 serial interrupts. These represent the 15 ISA interrupts (IRQ0-1, 2-15), the four PCI interrupts, and the control signals SMI# and IOCHK#. The serial IRQ protocol does not support the additional APIC interrupts (20-23). 4.8.1 Start Frame The serial IRQ protocol has two modes of operation which affect the start frame. These two modes are: * Continuous - SERIRQ controller is responsible for generating the start frame. * Quiet - a serial IRQ peripheral is responsible for beginning the start frame. The mode that must first be entered when enabling the serial IRQ protocol is continuous mode. In this mode, SERIRQ Controller asserts the start frame. This start frame is 4, 6, or 8 PCI clocks wide based upon the Serial IRQ Control Register, bits 1:0 at 64h in Device 31:Function 0 configuration space. This is a polling mode. When the serial IRQ stream enters quiet mode (signaled in the Stop Frame), the SERIRQ line remains inactive and pulled up between the Stop and Start Frame until a peripheral drives the SERIRQ signal low. The SERIRQ Controller senses the line low and continues to drive it low for the remainder of the Start Frame. Since the first PCI clock of the start frame was driven by the peripheral in this mode, the SERIRQ Controller drives the SERIRQ line low for 1 PCI clock less than in continuous mode. This mode of operation allows for a quiet, and therefore lower power, operation. 4.8.2 Data Frames Once the Start frame has been initiated, all of the SERIRQ peripherals must start counting frames based on the rising edge of SERIRQ. Each of the IRQ/DATA frames has exactly 3 phases of 1 clock each: * Sample Phase - During this phase, the SERIRQ device drives SERIRQ low if the corresponding interrupt signal is low. If the corresponding interrupt is high, then the SERIRQ devices tri-state the SERIRQ signal. The SERIRQ line remains high due to pull-up resistors (there is no internal pull-up resistor on this signal, an external pull-up resistor is required). A low level during the IRQ0-1 and IRQ2-15 frames indicates that an active-high ISA interrupt is not being requested, but a low level during the PCI INT[A:D], SMI#, and IOCHK# frame indicates that an active-low interrupt is being requested. * Recovery Phase - During this phase, the device drives the SERIRQ line high if in the Sample Phase it was driven low. If it was not driven in the sample phase, it is tristated in this phase. * Turn-around Phase - The device tri-states the SERIRQ line. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 133 4.8.3 Stop Frame After all data frames, a Stop Frame is driven by the SERIRQ controller. The SERIRQ signal is driven low by the SERIRQ controller for two or three PCI clocks. The number of clocks is determined by the SERIRQ configuration register. The number of clocks determines the next mode: Table 4-14. Stop Frame Explanation Stop Frame Width 4.8.4 Next Mode 2 PCI clocks Quiet Mode. Any SERIRQ device may initiate a Start Frame 3 PCI clocks Continuous Mode. Only the host (SERIRQ Controller) may initiate a Start Frame Specific Interrupts Not Supported via SERIRQ There are three interrupts seen through the serial stream that are not supported. These interrupts are generated internally, and are not sharable with other devices within the system. These interrupts are: * IRQ0 - Heartbeat interrupt generated off of the internal 8254 counter 0. * IRQ8# - RTC interrupt can only be generated internally. * IRQ13 - Floating point error interrupt generated off of the processor assertion of FERR#. The SERIRQ controller ignores the state of these interrupts in the serial stream, and does not adjust their level based on the level seen in the serial stream. 4.8.5 Data Frame Format Table 4-15 shows the format of the data frames. Table 4-15. Data Frame Format (Sheet 1 of 2) Data Frame # Interrupt Clocks Past Start Frame 1 IRQ0 2 2 IRQ1 5 3 SMI# 8 4 IRQ3 11 5 IRQ4 14 6 IRQ5 17 7 IRQ6 20 8 IRQ7 23 Ignored. IRQ0 can only be generated via the internal 8524 Causes SMI# if low. Sets the SERIRQ_SMI_STS bit. 9 IRQ8 26 10 IRQ9 29 11 IRQ10 32 12 IRQ11 35 13 IRQ12 38 14 IRQ13 41 Ignored. IRQ13 can only be generated from FERR# 15 IRQ14 44 Not attached to SATA logic 16 IRQ15 47 Not attached to SATA logic Intel(R) Communications Chipset 89xx Series - Datasheet 134 Comment Ignored. IRQ8# can only be generated internally. October 2012 Order Number: 327879-001US 4.0 Table 4-15. Data Frame Format (Sheet 2 of 2) Data Frame # 4.9 Interrupt Clocks Past Start Frame 17 IOCHCK# 50 Same as ISA IOCHCK# going active. 18 PCI INTA# 53 PIRQA# Comment 19 PCI INTB# 56 PIRQB# 20 PCI INTC# 59 PIRQC# 21 PCI INTD# 62 PIRQD# Real Time Clock (B0:D31:F0) The Real Time Clock (RTC) module provides a battery backed-up date and time keeping device with two banks of static RAM with 128 bytes each, although the first bank has 114 bytes for general purpose usage. Three interrupt features are available: * Time of day alarm with once a second to once a month range * Periodic rates of 122 s to 500 ms * End of update cycle notification Seconds, minutes, hours, days, day of week, month, and year are counted. Daylight savings compensation is no longer supported. The hour is represented in 12- or 24hour format, and data can be represented in BCD or binary format. The design is functionally compatible with the Motorola* MS146818B. The time keeping comes from a 32.768 kHz oscillating source, which is divided to achieve an update every second. The lower 14 bytes on the lower RAM block have specific functions. The first 10 are for time and date information; the next four (0Ah to 0Dh) are registers, which configure and report RTC functions. The time and calendar data should match the data mode (BCD or binary) and hour mode (12- or 24-hour) as selected in register B. The programmer must make sure that the data stored in these locations is within the reasonable values ranges and represents a possible date and time. The exception to these ranges is to store a value of C0-FFh in the Alarm bytes to indicate a don't care situation. All Alarm conditions must match to trigger an Alarm Flag, which could trigger an Alarm Interrupt if enabled. The SET bit must be 1 while programming these locations to avoid clashes with an update cycle. Access to time and date information is done through the RAM locations. If a RAM read from the ten time and date bytes is attempted during an update cycle, the value read do not necessarily represent the true contents of those locations. Any RAM writes under the same conditions are ignored. Note: The leap year determination for adding a 29th day to February does not take into account the end-of-the-century exceptions. The logic assumes that all years divisible by 4 are leap years. According to the Royal Observatory Greenwich, years that are divisible by 100 are typically not leap years. In every fourth century (years divisible by 400, like 2000), the 100-year-exception is over-ridden and a leap-year occurs. The year 2100 is the first time in which the current RTC implementation would incorrectly calculate the leap-year. The month/year alarms are not implemented. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 135 4.9.1 Update Cycles An update cycle occurs once a second, if the SET bit of register B is not asserted and the divide chain is properly configured. During this procedure, the stored time and date are incremented, overflow is checked, a matching alarm condition is checked, and the time and date are rewritten to the RAM locations. The update cycle starts at least 488 s after the UIP bit of register A is asserted, and the entire cycle does not take more than 1984 s to complete. The time and date RAM locations (0-9) are disconnected from the external bus during this time. To avoid update and data corruption conditions, external RAM access to these locations can safely occur at two times. When a updated-ended interrupt is detected, almost 999 ms is available to read and write the valid time and date data. If the UIP bit of Register A is detected to be low, there is at least 488 s before the update cycle begins. Warning: The overflow conditions for leap years adjustments are based on more than one date or time item. To ensure proper operation when adjusting the time, the new time and data values should be set at least two seconds before leap year occurs. 4.9.2 Interrupts The real-time clock interrupt is internally routed both to the I/O APIC and the 8259. It is mapped to interrupt vector 8. This interrupt is not shared with any other interrupt. IRQ8# from the SERIRQ stream is ignored. However, the High Performance Event Timers can also be mapped to IRQ8#; in this case, the RTC interrupt is blocked. 4.9.3 Lockable RAM Ranges The RTC's battery-backed RAM supports two 8-byte ranges that can be locked via the configuration space. If the locking bits are set, the corresponding range in the RAM is not readable or writable. A write cycle to those locations has no effect. A read cycle to those locations does not return the location's actual value (resultant value is undefined). Once a range is locked, the range can be unlocked only by a hard reset, which invokes the BIOS and allow it to relock the RAM range. 4.9.4 Century Rollover A century rollover is detected when the Year byte (RTC I/O space, index offset 09h) transitions form 99 to 00. Upon detecting the rollover, the NEWCENTURY_STS bit (TCOBASE + 04h, bit 7). If the system is in an S0 state, this causes an SMI#. The SMI# handler can update registers in the RTC RAM that are associated with century value. If the system is in a sleep state (S1-S5) when the century rollover occurs, the NEWCENTURY_STS bit is set, but no SMI# is generated. When the system resumes from the sleep state, BIOS should check the NEWCENTURY_STS bit and update the century value in the RTC RAM. 4.9.5 Clearing Battery-Backed RTC RAM CMOS RAM is cleared by using a jumper on RTCRST# or GPI. Implementations should not attempt to clear CMOS by using a jumper to pull VccRTC low. Using RTCRST# to Clear CMOS A jumper on RTCRST# can be used to clear CMOS values, as well as reset to default, the state of those configuration bits that reside in the RTC power well. When the RTCRST# is strapped to ground, the RTC_PWR_STS bit (B0:D31:F0:A4h bit 2) is set Intel(R) Communications Chipset 89xx Series - Datasheet 136 October 2012 Order Number: 327879-001US 4.0 and those configuration bits in the RTC power well is set to their default state. BIOS can monitor the state of this bit, and manually clear the RTC CMOS array once the system is booted. The normal position would cause RTCRST# to be pulled up through a weak pull-up resistor. Table 4-16 shows which bits are set to their default state when RTCRST# is asserted. This RTCRST# jumper technique allows the jumper to be moved and then replaced--all while the system is powered off. Then, once booted, the RTC_PWR_STS can be detected in the set state. Table 4-16. Configuration Bits Reset by RTCRST# Assertion Bit Name Register Location Bit(s) Default State Alarm Interrupt Enable (AIE) Register B (General Configuration) (RTC_REGB) I/O space (RTC Index + 0Bh) 5 X Alarm Flag (AF) Register C (Flag Register) (RTC_REGC) I/O space (RTC Index + 0Ch) 5 X SWSMI_RATE_SEL General PM Configuration 3 Register GEN_PMCON_3 B0:D31:F0:A4h 7:6 0 SLP_S4# Minimum Assertion Width General PM Configuration 3 Register GEN_PMCON_3 B0:D31:F0:A4h 5:4 0 SLP_S4# Assertion Stretch Enable General PM Configuration 3 Register GEN_PMCON_3 B0:D31:F0:A4h 3 0 RTC Power Status (RTC_PWR_STS) General PM Configuration 3 Register GEN_PMCON_3 B0:D31:F0:A4h 2 0 Power Failure (PWR_FLR) General PM Configuration 3 Register (GEN_PMCON_3) B0:D31:F0:A4h 1 0 AFTERG3_EN General PM Configuration 3 Register GEN_PMCON_3 B0:D31:F0:A4h 0 0 Power Button Override Status (PRBTNOR_STS) Power Management 1 Status Register (PM1_STS) PMBase + 00h 11 0 RTC Event Enable (RTC_EN) Power Management 1 Enable Register (PM1_EN) PMBase + 02h 10 0 Sleep Type (SLP_TYP) Power Management 1 Control (PM1_CNT) PMBase + 04h 12:10 0 PME_EN General Purpose Event 0 Enables Register (GPE0_EN) PMBase + 2Ch 11 0 BATLOW_EN General Purpose Event 0 Enables Register (GPE0_EN) PMBase + 2Ch 10 0 RI_EN General Purpose Event 0 Enables Register (GPE0_EN) PMBase + 2Ch 8 0 NEWCENTURY_STS TCO1 Status Register (TCO1_STS) TCOBase + 04h 7 0 Intruder Detect (INTRD_DET) TCO2 Status Register (TCO2_STS) TCOBase + 06h 0 0 Top Swap (TS) Backed Up Control Register (BUC) Chipset Config Registers:Offset 3414h 0 X October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 137 Using a GPI to Clear CMOS A jumper on a GPI can also be used to clear CMOS values. BIOS would detect the setting of this GPI on system boot-up, and manually clear the CMOS array. Note: The GPI strap technique to clear CMOS requires multiple steps to implement. The system is booted with the jumper in new position, then powered back down. The jumper is replaced back to the normal position, then the system is rebooted again. Warning: Do not implement a jumper on VccRTC to clear CMOS. 4.10 Processor Interface (B0:D31:F0) The processor interface is based on the following pin-based signals other than DMI: * Standard Outputs to processor: A20GATE, INIT3_3V#, CPUPWRGD, PMSYNC#, PECI * Standard Input from processor: THRMTRIP# Most outputs to the processor use standard buffers. The system has separate V_CPU_IO signals that are pulled up at the system level to the processor voltage, and thus determines VOH for the outputs to the processor. The following processor interface legacy pins were removed: * IGNNE#, STPCLK#, DPSLP#, are DPRSLPVR are no longer required for systems. * A20M#, SMI#, NMI, INIT#, INTR, FERR#: Functionality has been replaced by inband Virtual Legacy Wire (VLW) messages. See Section 4.10.3. 4.10.1 Processor Interface Signals and VLW Messages This section describes each of the processor interface signals. The behavior of some signals may vary during processor reset because the signals are used for frequency strapping. 4.10.1.1 A20M# (Mask A20) / A20GATE The A20M# VLW message is asserted when both of the following conditions are true: * The ALT_A20_GATE bit (Bit 1 of PORT92 register) is a 0 * The A20GATE input signal is a 0 The A20GATE input signal is expected to be generated by the external microcontroller (KBC). 4.10.1.2 INIT (Initialization) The INIT# VLW Message is asserted based on any one of several events described in Table 4-17. When any of these events occur, INIT# is asserted for 16 PCI clocks, then driven high. Intel(R) Communications Chipset 89xx Series - Datasheet 138 October 2012 Order Number: 327879-001US 4.0 Table 4-17. INIT# Going Active Cause of INIT3_3V# Going Active Comment Shutdown special cycle from processor observed on the Processor interconnect. INIT assertion based on value of Shutdown Policy Select register (SPS) PORT92 write, where INIT_NOW (bit 0) transitions from a 0 to a 1. PORTCF9 write, where SYS_RST (bit 1) was a 0 and RST_CPU (bit 2) transitions from 0 to 1. 0 to 1 transition on RCIN# must occur before INIT3_3V# can be re-generated. Note: RCIN# signal is expected to be low during S3, S4, and S5 states. Transition on the RCIN# signal in those states (or the transition to those states) may not necessarily cause the INIT3_3V# signal to be generated to the processor. RCIN# input signal goes low. RCIN# is expected to be driven by the external microcontroller (KBC). 4.10.1.3 FERR# (Numeric Coprocessor Error) The system supports coprocessor error function with the FERR# message. The function is enabled via the COPROC_ERR_EN bit. If FERR# is driven active by the processor, IRQ13 goes active (internally). When it detects a write to the COPROC_ERR register (I/O Register F0h), the internal IRQ13 is negated and IGNNE# is activated. IGNNE# remains active until FERR# is driven inactive. IGNNE# is never driven active unless FERR# is active. Note: IGNNE# (Ignore Numeric Error is internally generated by the processor.) 4.10.1.4 NMI (Non-Maskable Interrupt) Non-Maskable Interrupts (NMIs) can be generated by several sources, as described in Table 4-18. Table 4-18. NMI Sources Cause of NMI 4.10.1.5 Comment SERR# goes active (either internally, externally via SERR# signal, or via message from Processor) Can instead be routed to generate an SCI, through the NMI2SCI_EN bit (Device 31:Function 0, TCO Base + 08h, bit 11). IOCHK# goes active via SERIRQ# stream (ISA system Error) Can instead be routed to generate an SCI, through the NMI2SCI_EN bit (Device 31:Function 0, TCO Base + 08h, bit 11). Processor Power Good (PROCPWRGD) This signal is connected to the processor's VCCPRWGOOD_0 and VCCPRWGOOD_1 input. This signal represents a logical AND of PWROK and SYS_PWROK signals. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 139 4.10.2 Dual-Processor Issues 4.10.2.1 Usage Differences In dual-processor designs, some of the processor signals are unused or used differently than for uniprocessor designs. * A20M#/A20GATE and FERR# are generally not used, but still supported. * I/O APIC and SMI# are assumed to be used. 4.10.3 Virtual Legacy Wire (VLW) Messages The system supports VLW messages as alternative method of conveying the status of the following legacy sideband interface signals to the CPU: * A20M# * INTR * SMI# * INIT# * NMI Note: IGNNE# VLW message is not required to be generated as it is internally emulated by the Processor. VLW are inbound messages to the CPU. They are communicated via Vendor Defined Message over the DMI link. Legacy processor signals can only be delivered via VLW. Delivery of legacy processor signals (A20M#, INTR, SMI#, INIT# or NMI) via I/O APIC controller is not supported. 4.11 Power Management (B0:D31:F0) 4.11.1 Features * Support for Advanced Configuration and Power Interface, Version 3.0b (ACPI) providing power and thermal management -- ACPI 24-Bit Timer SCI and SMI# Generation * PCI PME# signal for Wake Up from Low-Power states * System Sleep State Control -- ACPI S3 state -- Suspend to RAM (STR) -- ACPI S4 state -- Suspend-to-Disk (STD) -- ACPI G2/S5 state -- Soft Off (SOFF) -- Power Failure Detection and Recovery * Management Engine Power Management Support -- Wake events from the Management Engine (enabled from all S-States including Catastrophic S5 conditions) Intel(R) Communications Chipset 89xx Series - Datasheet 140 October 2012 Order Number: 327879-001US 4.0 4.11.2 System Power States Table 4-19 shows the power states defined for system platforms. The state names generally match the corresponding ACPI states. Table 4-19. General Power States for Platform Systems State/ Substates Legacy Name / Description G0/S0/C0 Full On: Processor operating. Individual devices may be shut down or be placed into lower power states to save power. G0/S0/Cx Cx State: Cx states are processor power states within the S0 system state that provide for various levels of power savings. The processor initiates C-state entry. The system's behavior is based on the processor state. G1/S1 S1: The system provides the S1 messages and the S0 messages on a wake event. It is preferred for systems to use C-states than S1. G1/S3 Suspend-To-RAM (STR): The system context is maintained in system DRAM, but power is shut off to non-critical circuits. Memory is retained and refreshes continue. All external clocks stop except RTC. G1/S4 Suspend-To-Disk (STD): The context of the system is maintained on the disk. All power is then shut off to the system except for the logic required to resume. G2/S5 Soft Off (SOFF): System context is not maintained. All power is shut off except for the logic required to restart. A full boot is required when waking. G3 Mechanical OFF (MOFF): System context not maintained. All power is shut off except for the RTC. No "Wake" events are possible. This state occurs if the user removes the main system batteries in a mobile system, turns off a mechanical switch, or if the system power supply is at a level that is insufficient to power the "waking" logic. When system power returns, transition depends on the state just prior to the entry to G3 and the AFTERG3 bit in the GEN_PMCON3 register (B0:D31:F0, offset A4). See Table 4-26 for more details. Table 4-20 shows the transitions rules among the various states. Transitions among the various states may appear to temporarily transition through intermediate states. For example, in going from S0 to S3, it may appear to pass through the G1/S1 states. These intermediate transitions and states are not listed in the table. Table 4-20. State Transition Rules Present State Transition Trigger Next State G0/S0/C0 * * * * DMI Msg SLP_EN bit set Power Button Override Mechanical Off/Power Failure * * * * G0/S0/Cx G1/Sx or G2/S5 state G2/S5 G3 G0/S0/Cx * * * DMI Msg Power Button Override Mechanical Off/Power Failure * * * G0/S0/C0 S5 G3 G1/S1, G1/S3, or G1/S4 * * * Any Enabled Wake Event Power Button Override Mechanical Off/Power Failure * * * G0/S0/C0 (See Note 2) G2/S5 G3 G2/S5 * * Any Enabled Wake Event Mechanical Off/Power Failure * * G0/S0/C0 (See Note 2) G3 G3 * Power Returns * Optional to go to S0/C0 (reboot) or G2/S5 (stay off until power button pressed or other wake event). (See Notes 1 and 2) Notes: 1. Some wake events can be preserved through power failure. 2. Transitions from the S1-S5 or G3 states to the S0 state are deferred until BATLOW# is inactive in mobile configurations. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 141 4.11.3 System Power Planes The system has several independent power planes, as described in Table 4-21. When a power plane is shut off, it should go to a 0 V level. s Table 4-21. System Power Plane Plane Controlled By CPU SLP_S3# signal The SLP_S3# signal can be used to cut the power to the processor completely. SLP_S3# signal When SLP_S3# goes active, power can be shut off to any circuit not required to wake the system from the S3 state. Since the S3 state requires that the memory context be preserved, power must be retained to the main memory. The processor, devices on the PCI bus, LPC I/F, and graphics will typically be shut off when the Main power plane is off, although there may be small subsections powered. MEMORY SLP_S4# signal SLP_S5# signal When SLP_S4# goes active, power can be shut off to any circuit not required to wake the system from the S4. Since the memory context does not need to be preserved in the S4 state, the power to the memory can also be shut down. When SLP_S5# goes active, power can be shut off to any circuit not required to wake the system from the S5 state. Since the memory context does not need to be preserved in the S5 state, the power to the memory can also be shut. DEVICE[n] Implementation Specific Individual subsystems may have their own power plane. For example, GPIO signals may be used to control the power to disk drives, audio amplifiers, or the display screen. AUX Implementation Specific GbE MAC port 0 may have its own power plane for WOL implementation. SUS Implementation Specific A sustained power plane used for S5 and RTC. MAIN 4.11.4 Description SMI#/SCI Generation Upon any enabled SMI event taking place while the End of SMI (EOS) bit is set, the EOS bit is cleared and SMI is asserted to the processor, which causes it to enter SMM space. SMI assertion is performed using a Virtual Legacy Wire (VLW) message. Prior system generations (those based upon legacy processors) used an actual SMI# pin. Once the SMI VLW has been delivered, no action is taken on behalf of active SMI events until Host software sets the End of SMI (EOS) bit. At that point, if any SMI events are still active, the system sends another SMI VLW message. The SCI is a level-mode interrupt that is typically handled by an ACPI-aware operating system. In non-APIC systems (which is the default), the SCI IRQ is routed to one of the 8259 interrupts (IRQ 9, 10, or 11). The 8259 interrupt controller must be programmed to level mode for that interrupt. In systems using the APIC, the SCI can be routed to interrupts 9, 10, 11, 20, 21, 22, or 23. The interrupt polarity changes depending on whether or not it is on an interrupt shareable with a PIRQ. The interrupt remains asserted until all SCI sources are removed. Table 4-22 shows which events can cause an SMI and SCI. Some events can be programmed to cause either an SMI or SCI. The usage of the event for SCI (instead of SMI) is typically associated with an ACPI-based system. Each SMI or SCI source has a corresponding enable and status bit. Intel(R) Communications Chipset 89xx Series - Datasheet 142 October 2012 Order Number: 327879-001US 4.0 Table 4-22. Causes of SMI and SCI (Sheet 1 of 2) Cause SCI SMI Additional Enables Where Reported PME# Yes Yes PME_EN=1 PME_STS PME_B0 (Internal, Bus 0, PMECapable Agents) Yes Yes PME_B0_EN=1 PME_B0_STS PCI Express* PME Messages Yes Yes PCI_EXP_EN=1 (Not enabled for SMI) PCI_EXP_STS PCI Express* Hot Plug Message Yes Yes HOT_PLUG_EN=1 (Not enabled for SMI) HOT_PLUG_STS Power Button Press Yes Yes PWRBTN_EN=1 PWRBTN_STS Power Button Override (Note 7) Yes No None PRBTNOR_STS RTC Alarm Yes Yes RTC_EN=1 RTC_STS Ring Indicate Yes Yes RI_EN=1 RI_STS USB#1 wakes Yes Yes USB1_EN=1 USB1_STS USB#2 wakes Yes Yes USB2_EN=1 USB2_STS USB#3 wakes Yes Yes USB3_EN=1 USB3_STS USB#4 wakes Yes Yes USB4_EN=1 USB4_STS USB#5 wakes Yes Yes USB5_EN=1 USB5_STS USB#6 wakes Yes Yes USB6_EN=1 USB6_STS ACPI Timer overflow (2.34 sec.) Yes Yes TMROF_EN=1 TMROF_STS GPI[x]_STS ALT_GPI_SMI[x]_STS Any GPI[15:0] Yes Yes GPI[x]_Route=10; GPI[x]_EN=1 (SCI) GPI[x]_Route=01; ALT_GPI_SMI[x]_EN=1 (SMI) GPIO[27] Yes Yes GP27_EN=1 GP27_STS TCO SCI Logic Yes No TCOSCI_EN=1 TCOSCI_STS TCO SCI message from CPU Yes No none CPUSCI_STS TCO SMI Logic No Yes TCO_EN=1 TCO_STS TCO SMI -- Year 2000 Rollover No Yes none NEWCENTURY_STS TCO SMI -- TCO TIMEROUT No Yes none TIMEOUT TCO SMI -- OS writes to TCO_DAT_IN register No Yes none OS_TCO_SMI TCO SMI -- Message from CPU No Yes none CPUSMI_STS TCO SMI -- NMI occurred (and NMIs mapped to SMI) No Yes NMI2SMI_EN=1 NMI2SMI_STS TCO SMI -- INTRUDER# signal goes active No Yes INTRD_SEL=10 INTRD_DET TCO SMI -- Change of the BIOSWE (B0:D31:F0:DCh, bit 0) bit from 0 to 1 No Yes BLE=1 BIOSWR_STS TCO SMI -- Write attempted to BIOS No Yes BIOSWE=1 BIOSWR_STS BIOS_RLS written to Yes No GBL_EN=1 GBL_STS GBL_RLS written to No Yes BIOS_EN=1 BIOS_STS Write to B2h register No Yes APMC_EN = 1 APM_STS October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 143 Table 4-22. Causes of SMI and SCI (Sheet 2 of 2) Cause SMI Additional Enables Where Reported Periodic timer expires No Yes PERIODIC_EN=1 PERIODIC_STS 64 ms timer expires No Yes SWSMI_TMR_EN=1 SWSMI_TMR_STS Enhanced USB Legacy Support Event No Yes LEGACY_USB2_EN = 1 LEGACY_USB2_STS Enhanced USB Intel Specific Event No Yes INTEL_USB2_EN = 1 INTEL_USB2_STS Serial IRQ SMI reported No Yes none SERIRQ_SMI_STS Device monitors match address in its range No Yes none SMBus Host Controller No Yes SMB_SMI_EN Host Controller Enabled SMBus host status reg. SMBus Slave SMI message No Yes none SMBUS_SMI_STS SMBus SMBALERT# signal active No Yes none SMBUS_SMI_STS DEVTRAP_STS SMBus Host Notify message received No Yes HOST_NOTIFY_INTREN SMBUS_SMI_STS HOST_NOTIFY_STS SLP_EN bit written to 1 No Yes SMI_ON_SLP_EN=1 SMI_ON_SLP_EN_STS SPI Command Completed No Yes None SPI_SMI_STS Software Generated GPE Yes Yes SWGPE=1 SWGPE_STS USB2_STS, Write Enable Status GPIO_UNLOCK_SMI_STS USB Per-Port Registers Write Enable bit changes to 1 No Yes USB2_EN=1, Write_Enable_SMI_Enable =1 GPIO Lockdown Enable bit changes from `1' to `0' No Yes GPIO_UNLOCK_SMI_EN=1 Notes: 1. 2. 3. 4. 5. 6. 7. 8. 4.11.4.1 SCI SCI_EN must be 1 to enable SCI, except for BIOS_RLS. SCI_EN must be 0 to enable SMI. SCI can be routed to cause interrupt 9:11 or 20:23 (20:23 only available in APIC mode). GBL_SMI_EN must be 1 to enable SMI. EOS must be written to 1 to re-enable SMI for the next 1. The system must have SMI fully enabled when it is also enabled to trap cycles. If SMI is not enabled in conjunction with the trap enabling, then hardware behavior is undefined. Only GPI[15:0] may generate an SMI or SCI. When a power button override first occurs, the system transitions immediately to S5. The SCI only occurs after the next wake to S0 if the residual status bit (PRBTNOR_STS) is not cleared prior to setting SCI_EN. GBL_STS being set causes an SCI, even if the SCI_EN bit is not set. Software must take great care not to set the BIOS_RLS bit (which causes GBL_STS to be set) if the SCI handler is not in place. PCI Express* SCI PCI Express ports and the processor (via DMI) have the ability to cause PME using messages. When a PME message is received, the PCI_EXP_STS bit is set. If the PCI_EXP_EN bit is also set, this causes an SCI via the GPE1_STS register. 4.11.4.2 PCI Express* Hot-Plug PCI Express has a hot-plug mechanism and is capable of generating a SCI via the GPE1 register. It is also capable of generating an SMI. It is not capable of generating a wake event. Intel(R) Communications Chipset 89xx Series - Datasheet 144 October 2012 Order Number: 327879-001US 4.0 4.11.5 C-States Systems implement C-states by having the processor control the states. Messages are exchanged with the processor as part of the C-state flow, but the system no longer directly controls any of the processor impacts of C-states, such as voltage levels or processor clocking. In addition to the new messages, the system also provides additional information to the processor using a sideband pin (PM_SYNC). The legacy Cstate related pins (STPCLK#, STP_CPU#, DPRSLP#, DPRSLPVR#, etc.) do not exist. 4.11.6 Sleep States 4.11.6.1 Sleep State Overview The system directly supports different sleep states (S1-S5), which are entered by methods such as setting the SLP_EN bit or due to a Power Button press. The entry to the Sleep states is based on several assumptions: * The G3 state cannot be entered via any software mechanism. * The G3 state indicates a complete loss of power. 4.11.6.2 Initiating Sleep State Sleep states (S1-S5) are initiated by: * Masking interrupts, turning off all bus master enable bits, setting the desired type in the SLP_TYP field, and then setting the SLP_EN bit. The hardware then attempts to gracefully put the system into the corresponding Sleep state. * Pressing the PWRBTN# Signal for more than 4 seconds to cause a Power Button Override event. In this case the transition to the S5 state is less graceful, since there are no dependencies on DMI messages from the processor or on clocks other than the RTC clock. * Assertion of the THRMTRIP# signal will cause a transition to the S5 state. This can occur when system is in S0 or S1 state. * Internal watchdog timer timeout events Table 4-23. Sleep Types Sleep Type 4.11.6.3 Comment S1 System lowers the processor's power consumption. No snooping is possible in this state. S3 The system asserts SLP_S3#. The SLP_S3# signal controls the power to non-critical circuits. Power is only retained to devices needed to wake from this sleeping state, as well as to the memory. S4 The system asserts SLP_S3# and SLP_S4#. The SLP_S4# signal shuts off the power to the memory subsystem. Only devices needed to wake from this state should be powered. S5 The system asserts SLP_S3#, SLP_S4# and SLP_S5#. Exiting Sleep States Sleep states (S1-S5) are exited based on Wake events. The Wake events forces the system to a full on state (S0), although some non-critical subsystems might still be shut off and have to be brought back manually. For example, the hard disk may be shut off during a sleep state and have to be enabled via a GPIO pin before it can be used. Upon exit from system-controlled Sleep states, the WAK_STS bit is set. The possible causes of Wake Events (and their restrictions) are shown in Table 4-24. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 145 Table 4-24. Causes of Wake Events States Can Wake From Cause S1-S5 (Note 1) RTC Alarm Power Button S1-S5 How Enabled Set RTC_EN bit in PM1_EN register Always enabled as Wake event. (Note 2). GPE0_EN register Note: GPIs that are in the core well are not capable of waking the system from sleep states when the core well is not powered. GPI[0:15] S1-S5 (Note 1) GPIO[27] S1-S5 Set GP27_EN in GPE0_EN Register USB S1-S4 Set USB1_EN, USB 2_EN, USB3_EN, USB4_EN, USB5_EN, and USB6_EN bits in GPE0_EN register RI# S1-S5 (Note 1) Secondary PME# S1-S5 PCI_EXP_WAKE# S1-S5 Set RI_EN bit in GPE0_EN register Set PME_EN bit in GPE0_EN register. PCI_EXP_WAKE bit (Note 3) SATA S1 Set PME_EN bit in GPE0_EN register. (Note 4) PCI_EXP PME Message S1 Must use the PCI Express* WAKE# pin rather than messages for wake from S3,S4, or S5. SMBALERT# S1-S5 Always enabled as Wake event SMBus Slave Wake Message (01h) S1-S5 Wake/SMI# command always enabled as a Wake event. Note: SMBus Slave Message can wake the system from S1-S5, as well as from S5 due to Power Button Override. (Note 2). SMBus Host Notify message received S1-S5 HOST_NOTIFY_WKEN bit SMBus Slave Command register. Reported in the SMB_WAK_STS bit in the GPEO_STS register. Intel(R) ME NonMaskable Wake S1-S5 Always enabled as a wake event. (Note 2). Notes: 1. This is a wake event from S5 only if the sleep state was entered by setting the SLP_EN and SLP_TYP bits via software, or if there is a power failure. 2. If in the S5 state due to a power button override or THRMTRIP#, the possible wake events are due to Power Button, Wake SMBus Slave Message (01h), the ME-Initiated non-maskable Wake, the Integrated WOL Enable Override, Hard Reset Without Cycling (See Command Type 3 in Table 4-49), Hard Reset System (See Command Type 4 in Table 4-49). 3. When the WAKE# pin is active and the PCI Express* device is enabled to wake the system, the system will wake the platform. 4. SATA can only trigger a wake event in S1, but if PME is asserted prior to S3/S4/S5 entry and software does not clear the PME_B0_STS, a wake event would still result. The varous GPIs have different levels of functionality when used as wake events. The GPIs that reside in the core power well can only generate wake events from sleep states where the core well is powered. Only certain GPIs are ACPI-compliant, meaning that their Status and Enable bits reside in ACPI I/O space. Table 4-25 summarizes the use of GPIs as wake events. Table 4-25. GPI Wake Events GPI Power Well Wake From Notes GPI[7:0] Core S1 ACPI Compliant GPI[15:8] Suspend S1-S5 ACPI Compliant The latency to exit the vaious Sleep states varies and is dependent on power supply design to the extent that the exit latencies are insignificant. Intel(R) Communications Chipset 89xx Series - Datasheet 146 October 2012 Order Number: 327879-001US 4.0 4.11.6.4 PCI Express* WAKE# Signal and PME Event Message PCI Express ports can wake the platform from any sleep state (S1, S3, S4, or S5) using the WAKE# pin. WAKE# is treated as a wake event, but does not cause any bits to go active in the GPE_STS register. PCI Express ports and the processor (via DMI) have the ability to cause PME using messages. When a PME message is received, the PCI_EXP_STS bit is set. 4.11.6.5 Sx-G3-Sx, Handling Power Failures Depending on when the power failure occurs and how the system is designed, different transitions could occur due to a power failure. The AFTER_G3 bit provides the ability to program whether or not the system should boot once power returns after a power loss event. If the policy is to not boot, the system remains in an S5 state (unless previously in S4). Three possible events will wake the system after a power failure: * PWRBTN#: PWRBTN# is always enabled as a wake event. When RSMRST# is low (G3 state), the PWRBTN_STS bit is reset. When the system exits G3 after power returns (RSMRST# goes high), the PWRBTN# signal is already high (because VCCstandby goes high before RSMRST# goes high) and the PWRBTN_STS bit is 0. * RI#: RI# does not have an internal pull-up. Therefore, if this signal is enabled as a wake event, it is important to keep this signal powered during the power loss event. If this signal goes low (active), when power returns the RI_STS bit is set and the system interprets that as a wake event. * RTC Alarm: The RTC_EN bit is in the RTC well and is preserved after a power loss. Like PWRBTN_STS the RTC_STS bit is cleared when RSMRST# goes low. The system monitors both PWROK and RSMRST# to detect for power failures. If PWROK goes low, the PWROK_FLR bit is set. If RSMRST# goes low, PWR_FLR is set. Note: Although PME_EN is in the RTC well, this signal cannot wake the system after a power loss. PME_EN is cleared by RTCRST#, and PME_STS is cleared by RSMRST#. Table 4-26. Transitions Due to Power Failure 4.11.7 State at Power Failure AFTERG3_EN bit Transition When Power Returns S0, S1, S3 1 0 S5 S0 S4 1 0 S4 S0 S5 1 0 S5 S0 Event Input Signals and Their Usage There are various input signals that trigger specific events. This section describes those signals and how they should be used. 4.11.7.1 PWRBTN# (Power Button) The PWRBTN# signal operates as a "Fixed Power Button" as described in the Advanced Configuration and Power Interface, Version 2.0b. PWRBTN# signal has a 16 ms debounce on the input. The state transition descriptions are included in Table 4-27. The transitions start as soon as the PWRBTN# is pressed (but after the debounce logic), and does not depend on when the power button is released. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 147 Note: During the time that the SLP_S4# signal is stretched for the minimum assertion width (if enabled), the power button is not a wake event. See Power Button Override Function below. Table 4-27. Transitions Due to Power Button Present State Event Transition/Action Comment S0/Cx PWRBTN# goes low SMI or SCI generated (depending on SCI_EN, PWRBTN_EN and GLB_SMI_EN) Software typically initiates a Sleep state S1-S5 PWRBTN# goes low Wake Event. Transitions to S0 state Standard wakeup G3 PWRBTN# pressed None No effect since no power Not latched nor detected PWRBTN# held low for at least 4 consecutive seconds Unconditional transition to S5 state No dependence on processor (DMI Messages) or any other subsystem S0-S4 Power Button Override Function If PWRBTN# is observed active for at least four consecutive seconds, the state machine should unconditionally transition to the G2/S5 state, regardless of present state (S0- S4), even if PWROK is not active. In this case, the transition to the G2/S5 state should not depend on any particular response from the processor (e.g., a DMI Messages), nor any similar dependency from any other subsystem. The PWRBTN# status is readable to check if the button is currently being pressed or has been released. The status is taken after the de-bounce, and is readable via the PWRBTN_LVL bit. Note: The 4-second PWRBTN# assertion should only be used if a system lock-up has occurred. The 4-second timer starts counting when the system is in a S0 state. If the PWRBTN# signal is asserted and held active when the system is in a suspend state (S1-S5), the assertion causes a wake event. Once the system has resumed to the S0 state, the 4-second timer starts. During the time that the SLP_S4# signal is stretched for the minimum assertion width (if enabled by B0:D31:F0:A4h bit 3), the Power Button is not a wake event. As a result, it is conceivable that the user will press and continue to hold the Power Button waiting for the system to awake. Since a 4-second press of the Power Button is already defined as an Unconditional Power down, the power button timer is forced to inactive while the power-cycle timer is in progress. Once the power-cycle timer has expired, the Power Button awakes the system. Once the minimum SLP_S4# power cycle expires, the Power Button must be pressed for another 4 to 5 seconds to create the Override condition to S5. Sleep Button The Advanced Configuration and Power Interface, Version 2.0b defines an optional Sleep button. It differs from the power button in that it only is a request to go from S0 to S1-S4 (not S5). In an S5 state, the Power Button can wake the system, but the Sleep Button cannot. Although there is no specific signal designated as a Sleep Button, one of the GPIO signals can be used to create a "Control Method" Sleep Button. See the Advanced Configuration and Power Interface, Version 2.0b for implementation details. Intel(R) Communications Chipset 89xx Series - Datasheet 148 October 2012 Order Number: 327879-001US 4.0 4.11.7.2 RI# (Ring Indicator) The Ring Indicator can cause a wake event (if enabled) from the S1-S5 states. Table 4-28 shows when the wake event is generated or ignored in different states. If in the G0/S0/Cx states, the system generates an interrupt based on RI# active, and the interrupt is set up as a Break event. Table 4-28. Transitions Due to RI# Signal Present State Event RI_EN S0 RI# Active X Ignored RI# Active 0 1 Ignored Wake Event S1-S5 Event Note: Filtering/Debounce on RI# is not implemented. If required, implement externally. 4.11.7.3 SYS_RESET# Signal When the SYS_RESET# pin is detected as active after the 16 ms debounce logic, the system attempts to perform a "graceful" reset by waiting up to 25 ms for the SMBus to go idle. If the SMBus is idle when the pin is detected active, the reset occurs immediately; otherwise, the counter starts. If at any point during the count the SMBus goes idle the reset occurs. If, however, the counter expires and the SMBus is still active, a reset is forced upon the system even though activity is still occurring. Once the reset is asserted, it remains asserted for 5 to 6 ms regardless of whether the SYS_RESET# input remains asserted or not. It cannot occur again until SYS_RESET# has been detected inactive after the debounce logic, and the system is back to a full S0 state with PLTRST# inactive. If bit 3 of the CF9h I/O register is set then SYS_RESET# results in a full power cycle reset. 4.11.7.4 THRMTRIP# Signal If THRMTRIP# goes active, the processor is indicating an overheat condition, the system immediately transitions to an S5 state, driving SLP_S3#, SLP_S4#, SLP_S5# low, and setting the CTS bit. The transition looks like a power button override. When a THRMTRIP# event occurs, the system will power down immediately without following the normal S0 -> S5 path. The system will immediately drive SLP_S3#, SLP_S4#, and SLP_S5# low after sampling THRMTRIP# active. If the processor is running extremely hot and is heating up, it is possible (although unlikely) that components around it are no longer executing cycles properly. Therefore, if THRMTRIP# goes active, and the system is relying on state machine logic to perform the power down, the state machine may not be working, and the system does not power down. The system provides filtering for short low glitches on the THRMTRIP# signal to prevent erroneous system shut downs from noise. Glitches shorter than 25 nsec are ignored. During boot, THRMTRIP# is ignored until SLP_S3#, PWROK, and PLTRST# are all `1'. During entry into a powered-down state (due to S3, S4, S5 entry, power cycle reset, etc.) THRMTRIP# is ignored until either SLP_S3# = 0, or PWROK = 0, or SYS_PWROK = 0. A thermal trip event will: * Set the AFTERG3_EN bit * Clear the PWRBTN_STS bit October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 149 * Clear all the GPE0_EN register bits * Clear the SMB_WAK_STS bit only if SMB_SAK_STS was set due to SMBus slave receiving message and not set due to SMBAlert 4.11.8 ALT Access Mode Before entering a low power state, several registers from powered down parts may need to be saved. In the majority of cases, this is not an issue, as registers have read and write paths. However, several of the ISA compatible registers are either read only or write only. To get data out of write-only registers, and to restore data into read-only registers, the system implements an ALT access mode. If the ALT access mode is entered and exited after reading the registers of the timer (8254), the timer starts counting faster (13.5 ms). The following steps can cause problems: 1. BIOS enters ALT access mode for reading timer related registers. 2. BIOS exits ALT access mode. 3. BIOS continues through the execution of other needed steps and passes control to the operating system. After getting control in step #3, if the operating system does not reprogram the system timer again, the timer ticks may be happening faster than expected. For example Microsoft MS-DOS* and its associated software assume that the system timer is running at 54.6 ms and as a result the time-outs in the software may be happening faster than expected. Operating systems (e.g., Microsoft Windows* 98 and Windows* 2000) reprogram the system timer and therefore do not encounter this problem. For other OS's (e.g., Microsoft MS-DOS*) the BIOS should restore the timer back to 54.6 ms before passing control to the operating system. If the BIOS is entering ALT access mode before entering the suspend state it is not necessary to restore the timer contents after the exit from ALT access mode. 4.11.8.1 Write Only Registers with Read Paths in ALT Access Mode The registers described in Table 4-29 have read paths in ALT access mode. The access number field in the table indicates which register is returned per access to that port. Intel(R) Communications Chipset 89xx Series - Datasheet 150 October 2012 Order Number: 327879-001US 4.0 Table 4-29. Write Only Registers with Read Paths in ALT Access Mode (Sheet 1 of 2) Restore Data I/O Addr # of Rds 00h 2 01h 02h 03h 04h 05h 06h 07h 08h Restore Data Access Data 1 I/O Addr # of Rds Access Data DMA Chan 0 base address low byte 1 Timer Counter 0 status, bits [5:0] 2 DMA Chan 0 base address high byte 2 Timer Counter 0 base count low byte 1 DMA Chan 0 base count low byte 3 Timer Counter 0 base count high byte 2 DMA Chan 0 base count high byte 4 Timer Counter 1 base count low byte 1 DMA Chan 1 base address low byte 5 Timer Counter 1 base count high byte 2 DMA Chan 1 base address high byte 6 Timer Counter 2 base count low byte 1 DMA Chan 1 base count low byte 7 Timer Counter 2 base count high byte 2 DMA Chan 1 base count high byte 41h 1 Timer Counter 1 status, bits [5:0] 1 DMA Chan 2 base address low byte 42h 1 Timer Counter 2 status, bits [5:0] 2 DMA Chan 2 base address high byte 70h 1 Bit 7 = NMI Enable, Bits [6:0] = RTC Address 1 DMA Chan 2 base count low byte 2 DMA Chan 2 base count high byte 2 2 40h 7 2 2 2 2 C4h 1 DMA Chan 3 base address low byte 2 DMA Chan 3 base address high byte 1 DMA Chan 3 base count low byte 2 DMA Chan 3 base count high byte 1 DMA Chan 0-3 Command2 2 DMA Chan 0-3 Request 3 DMA Chan 0 Mode: Bits(1:0) = 00 4 DMA Chan 1 Mode: Bits(1:0) = 01 5 DMA Chan 2 Mode: Bits(1:0) = 10 6 DMA Chan 3 Mode: Bits(1:0) = 11. 2 6 C6h C8h October 2012 Order Number: 327879-001US CAh CCh CEh 1 DMA Chan 5 base address low byte 2 DMA Chan 5 base address high byte 1 DMA Chan 5 base count low byte 2 DMA Chan 5 base count high byte 1 DMA Chan 6 base address low byte 2 DMA Chan 6 base address high byte 1 DMA Chan 6 base count low byte 2 DMA Chan 6 base count high byte 1 DMA Chan 7 base address low byte 2 DMA Chan 7 base address high byte 1 DMA Chan 7 base count low byte 2 DMA Chan 7 base count high byte 2 2 2 2 2 2 Intel(R) Communications Chipset 89xx Series - Datasheet 151 Table 4-29. Write Only Registers with Read Paths in ALT Access Mode (Sheet 2 of 2) Restore Data I/O Addr # of Rds 20h 12 Access Restore Data I/O Addr Data # of Rds Access Data DMA Chan 4-7 Command2 1 PIC ICW2 of Master controller 1 2 PIC ICW3 of Master controller 2 DMA Chan 4-7 Request 3 DMA Chan 4 Mode: Bits(1:0) = 00 4 DMA Chan 5 Mode: Bits(1:0) = 01 3 PIC ICW4 of Master controller 4 PIC OCW1 of Master controller1 5 PIC OCW2 of Master controller 5 DMA Chan 6 Mode: Bits(1:0) = 10 6 PIC OCW3 of Master controller 6 DMA Chan 7 Mode: Bits(1:0) = 11. 7 PIC ICW2 of Slave controller 8 PIC ICW3 of Slave controller 9 PIC ICW4 of Slave controller 10 PIC OCW1 of Slave controller1 11 PIC OCW2 of Slave controller 12 PIC OCW3 of Slave controller D0h 6 Notes: 1. The OCW1 register must be read before entering ALT access mode. 2. Bits 5, 3, 1, and 0 return 0. 4.11.8.2 PIC Reserved Bits Many bits within the PIC are reserved, and must have certain values written for the PIC to operate properly. Therefore, there is no need to return these values in ALT access mode. When reading PIC registers from 20h and A0h, the reserved bits return the values listed in Table 4-30. Table 4-30. PIC Reserved Bits Return Values PIC Reserved Bits Value Returned ICW2(2:0) 000 ICW4(7:5) 000 ICW4(3:2) 00 ICW4(0) 0 OCW2(4:3) 00 OCW3(7) 0 OCW3(5) Reflects bit 6 OCW3(4:3) 01 Intel(R) Communications Chipset 89xx Series - Datasheet 152 October 2012 Order Number: 327879-001US 4.0 4.11.8.3 Read Only Registers with Write Paths in ALT Access Mode The registers described in Table 4-31 have write paths to them in ALT access mode. Software restores these values after returning from a powered down state. Software must handle these registers. When in normal mode, writing to the base address/count register also writes to the current address/count register. Therefore, the base address/ count must be written first, then the part is put into ALT access mode and the current address/count register is written. Table 4-31. Register Write Accesses in ALT Access Mode I/O Address Register Write Value 08h DMA Status Register for channels 0-3. D0h DMA Status Register for channels 4-7. 4.11.9 System Power Supplies, Planes, and Signals 4.11.9.1 Power Plane Control with SLP_S3#, SLP_S4# and SLP_S5# The SLP_S3# output signal can be used to cut power to the system core supply, since it only goes active for the Suspend-to-RAM state (typically mapped to ACPI S3). Power must be maintained to the suspend well, and to any other circuits that need to generate Wake signals from the Suspend-to-RAM state. During S3 (Suspend-to-RAM) all signals attached to powered down plans is tri-stated or driven low, unless they are pulled via a pull-up resistor. Cutting power to the core may be done via the power supply, or by external FETs on the motherboard. The SLP_S4# or SLP_S5# output signal can be used to cut power to the system core supply, as well as power to the system memory, since the context of the system is saved on the disk. Cutting power to the memory may be done via the power supply, or by external FETs on the motherboard. The SLP_S4# output signal is used to remove power to additional subsystems that are powered during SLP_S3#. SLP_S5# output signal can be used to cut power to the system core supply, as well as power to the system memory, since the context of the system is saved on the disk. Cutting power to the memory may be done via the power supply, or by external FETs on the motherboard. 4.11.9.2 SLP_S4# and Suspend-To-RAM Sequencing The system memory suspend voltage regulator is controlled by the Glue logic. The SLP_S4# signal should be used to remove power to system memory rather than the SLP_S5# signal. The SLP_S4# logic provides a mechanism to fully cycle the power to the DRAM and/or detect if the power is not cycled for a minimum time. Note: To utilize the minimum DRAM power-down feature that is enabled by the SLP_S4# Assertion Stretch Enable bit (B0:D31:F0:A4h bit 3), the DRAM power must be controlled by the SLP_S4# signal. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 153 4.11.9.3 PWROK Signal When asserted, PWROK is an indication that its core well power rails have been stable for at least 1 ms. PWROK can be driven asynchronously. When PWROK is low, the system asynchronously asserts PLTRST#. PWROK must not glitch, even if RSMRST# is low. Power associated with PCI/PCIe* must be valid for 99 ms prior to PWROK assertion to comply with the 100 ms PCI 2.3 / PCIe* 1.1 specification on PLTRST# deassertion. Note: SYS_RESET# is recommended for implementing the system reset button. This saves external logic that is needed if the PWROK input is used, and it allows for better handling of the SMBus and processor resets and avoids improperly reporting power failures. 4.11.9.4 BATLOW# (Battery Low) The BATLOW# input can inhibit waking from S3, S4, and S5 states if power is not sufficient. It also causes an SMI if the system is already in an S0 state. 4.11.10 Legacy Power Management Theory of Operation Instead of relying on ACPI software, legacy power management uses BIOS and various hardware mechanisms. The scheme relies on the concept of detecting when individual subsystems are idle, detecting when the whole system is idle, and detecting when accesses are attempted to idle subsystems. However, the operating system is assumed to be at least APM enabled. Without APM calls, there is no quick way to know when the system is idle between keystrokes. The system does not support burst modes. 4.11.10.1 APM Power Management There is a timer that, when enabled by the 1MIN_EN bit in the SMI Control and Enable register, generates an SMI once per minute. The SMI handler can check for system activity by reading the DEVTRAP_STS register. If none of the system bits are set, the SMI handler can increment a software counter. When the counter reaches a sufficient number of consecutive minutes with no activity, the SMI handler can then put the system into a lower power state. If there is activity, various bits in the DEVTRAP_STS register is set. Software clears the bits by writing a 1 to the bit position. The DEVTRAP_STS register allows for monitoring various internal devices, or Super I/O devices (SP, PP, FDC) on LPC or PCI, keyboard controller accesses, or audio functions on LPC or PCI. Other PCI activity can be monitored by checking the PCI interrupts. 4.11.11 Reset Behavior When a reset is triggered, the system will send a warning message to the CPU to allow the CPU to attempt to complete any outstanding memory cycles and put memory into a safe state before the platform is reset. When the CPU is ready, it will send an acknowledge message to the system. Once the message is received the system asserts PLTRST#. The system does not require an acknowledge message from the CPU to trigger PLTRST#. A global reset will occur after 4 seconds if an acknowledge from the CPU is not received. Intel(R) Communications Chipset 89xx Series - Datasheet 154 October 2012 Order Number: 327879-001US 4.0 When the system causes a reset by asserting PLTRST# its output signals will go to their reset states. A reset in which the host platform is reset and PLTRST# is asserted is called a Host Reset or Host Partition Reset. Depending on the trigger a host reset may also result in power cycling. See the following table for details. If a host reset is triggered and the timer times out before receiving an acknowledge message from the CPU, a Global Reset with power cycle will occur. A reset in which the host and ME partitions of the platform are reset is called a Global Reset. Table 4-32. Causes of Host and Global Resets Host Reset without Power Cycle Host Reset with Power Cycle Global Reset with Power Cycle Write of 0Eh to CF9h Register when Global Reset Bit=0b No Yes No (Note 1) Write of 06h to CF9h Register when Global Reset Bit=0b Yes No No (Note 1) Write of 06h or 0Eh to CF9h Register when Global Reset Bit=1b No No Yes SYS_RESET# Asserted and CF9h Bit 3 = 0 Yes No No (Note 1) SYS_RESET# Asserted and CF9h Bit 3 = 1 No Yes No (Note 1) SMBus Slave Message received for Reset with Power Cycle No Yes No (Note 1) SMBus Slave Message received for Reset without Power Cycle Yes No No (Note 1) TCO Watchdog Timer reaches zero two times Yes No No (Note 1) Power Failure: PWROK signal or SYS_PWROK signal goes inactive or RSMRST# asserts No No Yes (Note 2) Special shutdown cycle from CPU causes CF9h-like PLTRST# and CF9h Global Reset Bit = 1 No No Yes Special shutdown cycle from CPU causes CF9h-like PLTRST# and CF9h Bit 3 = 1 No Yes No (Note 2) Special shutdown cycle from CPU causes CF9h-like PLTRST# and CF9h Global Reset Bit = 0 Yes No No (Note 1) Intel Management Engine Triggered Host Reset without power cycle Yes No No (Note 1) Intel(R) Management Engine Triggered Host Reset with power cycle No Yes No (Note 1) No No Yes Management Engine Initiated Host Reset with power down No Yes (Note 3) No (Note 1) PLTRST# Entry Timeout No No Yes Trigger (R) Intel (R) Management Engine Triggered Global Reset Intel(R) Intel(R) Management Engine Watchdog Timer Power Management Watchdog Timer No Yes (Note 4) No (Note 1) No Yes (Note 4) No (Note 1) PROCPWRGD Stuck Low No No Yes PROCPWRGD-to CPURST# Violation No No Yes Notes: 1. 2. 3. 4. Trigger results in Global Reset with power cycle if the acknowledge message is not received. The system does not send warning message to CPU, reset occurs without delay. The system waits for enabled wake event to complete reset. The system allowed to drop this type of reset request if received while the system is in S3/S4/S5. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 155 The system must not drop this type of reset request if received while system is in a software-entered S3/S4/S5 state. However, the system is allowed to perform the reset without executing the RESET_WARN protocol in these states. 4.12 System Management The System Management Bus (SMBus) 2.0 Host Controller and SMBus Slave Interfaces make it easier to manage and lower Total Cost of Ownership (TCO) of the system. Figure 4-4 provides a high-level block of the System Management Interfaces. The following sections provides a description of the functions and protocols of the SMBus interfaces. Figure 4-4. System Management Bus (SMBus) Interface System Management Interface SMLINK1 SMT3 BMC / EC ME Host SMBus CK420 MST_SMBus TPM Internal TCO TCO Slave LPC (Master) BMC_SMBus Host DIMM Thermal Sensors PCIe Slot SMT1 EP/GbE/QAT Regs EP_SMBus EP Slave GbE Master/ Slave IPMI Packets (GbE) GbE_SMBus Optional (Connect per Design Requirements) 4.12.1 SMBus Host Controller (B0:D31:F3) The host controller provides a mechanism for the processor to initiate communications with SMBus peripherals (slaves). The Host Controller is also capable of operating in a mode in which it can communicate with I2C compatible devices. See Section 11.0, "SMBus Controller Registers (B0:D31:F3)" for SMBus Host Controller Configuration register descriptions. Intel(R) Communications Chipset 89xx Series - Datasheet 156 October 2012 Order Number: 327879-001US 4.0 SMBus messages can be performed with either packet error checking (PEC) enabled or disabled. The actual PEC calculation and checking is performed in hardware. The Slave Interface allows an external master to take control of the bus and perform read or write transactions. Write cycles can be used to cause certain events or pass messages, and the read cycles can be used to determine the state of various status bits. The system internal host controller cannot access the internal Slave Interface. The SMBus exists in Bus 0:Device 31:Function 3 configuration space, and consists of a transmit data path, and host controller. The transmit data path provides the data flow logic needed to implement the seven different SMBus command protocols and is controlled by the host controller. The SMBus host controller logic is clocked by RTC clock. The SMBus Address Resolution Protocol (ARP) is supported by using the existing host controller commands through software, except for the new Host Notify command (which is actually a received message). The programming model of the host controller is combined into two portions: a PCI configuration portion, and a system I/O mapped portion. All static configuration, such as the I/O base address, is done via the PCI configuration space. Real-time programming of the Host interface is done in system I/O space. The SMBus host controller checks for parity errors as a target. If an error is detected, the detected parity error bit in the PCI Status Register (Device 31:Function 3:Offset 06h:bit 15) is set. If bit 6 and bit 8 of the PCI Command Register (Device 31:Function 3:Offset 04h) are set, an SERR# is generated and the signaled SERR# bit in the PCI Status Register (bit 14) is set. 4.12.1.1 Host Controller Interface The SMBus host controller is used to send commands to other SMBus slave devices. Software sets up the host controller with an address, command, and, for writes, data and optional PEC; and then tells the controller to start. When the controller has finished transmitting data on writes, or receiving data on reads, it generates an SMI# or interrupt, if enabled. The host controller supports 8 command protocols of the SMBus interface (See System Management Bus (SMBus) Specification, Version 2.0): Quick Command, Send Byte, Receive Byte, Write Byte/Word, Read Byte/Word, Process Call, Block Read/Write, Block Write-Block Read Process Call, and Host Notify. The SMBus host controller requires that the various data and command fields be setup for the type of command to be sent. When software sets the START bit, the SMBus Host controller performs the requested transaction, and interrupts the processor (or generates an SMI#) when the transaction is completed. Once a START command has been issued, the values of the "active registers" (Host Control, Host Command, Transmit Slave Address, Data 0, Data 1) should not be changed or read until the interrupt status message (INTR) has been set (indicating the completion of the command). Any register values needed for computation purposes should be saved prior to issuing of a new command, as the SMBus host controller updates all registers while completing the new command. The Host Controller supports the System Management Bus (SMBus) Specification, Version 2.0. Slave functionality, including the Host Notify protocol, is available on the SMBus pins. Note: Using the SMBus host controller to send commands to any SMBus slave port is not supported. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 157 4.12.1.2 Command Protocols In all of the following commands, the Host Status Register (offset 00h) is used to determine the progress of the command. While the command is in operation, the HOST_BUSY bit is set. If the command completes successfully, the INTR bit is set in the Host Status Register. If the device does not respond with an acknowledge, and the transaction times out, the DEV_ERR bit is set. If software sets the KILL bit in the Host Control Register while the command is running, the transaction will stop and the FAILED bit is set. Quick Command When programmed for a Quick Command, the Transmit Slave Address Register is sent. The PEC byte is never appended to the Quick Protocol. Software should force the PEC_EN bit to 0 when performing the Quick Command. Software must force the I2C_EN bit to 0 when running this command. See section 5.5.1 of the System Management Bus (SMBus) Specification, Version 2.0 for the protocol format. Send Byte / Receive Byte For the Send Byte command, the Transmit Slave Address and Device Command Registers are sent. For the Receive Byte command, the Transmit Slave Address Register is sent. The data received is stored in the DATA0 register. Software must force the I2C_EN bit to 0 when running this command. The Receive Byte is similar to a Send Byte, the only difference is the direction of data transfer. See sections 5.5.2 and 5.5.3 of the System Management Bus (SMBus) Specification, Version 2.0 for the protocol format. Write Byte/Word The first byte of a Write Byte/Word access is the command code. The next 1 or 2 bytes are the data to be written. When programmed for a Write Byte/Word command, the Transmit Slave Address, Device Command, and Data0 Registers are sent. In addition, the Data1 Register is sent on a Write Word command. Software must force the I2C_EN bit to 0 when running this command. See section 5.5.4 of the System Management Bus (SMBus) Specification, Version 2.0 for the protocol format. Read Byte/Word Reading data is slightly more complicated than writing data. First, the host must write a command to the slave device. Then it must follow that command with a repeated start condition to denote a read from that device's address. The slave then returns 1 or 2 bytes of data. Software must force the I2C_EN bit to 0 when running this command. When programmed for the read byte/word command, the Transmit Slave Address and Device Command Registers are sent. Data is received into the DATA0 on the read byte, and the DAT0 and DATA1 registers on the read word. See section 5.5.5 of the System Management Bus (SMBus) Specification, Version 2.0 for the protocol format. Process Call The process call is so named because a command sends data and waits for the slave to return a value dependent on that data. The protocol is simply a Write Word followed by a Read Word, but without a second command or stop condition. When programmed for the Process Call command, the host transmits the Transmit Slave Address, Host Command, DATA0 and DATA1 registers. Data received from the device is stored in the DATA0 and DATA1 registers. The Process Call command with Intel(R) Communications Chipset 89xx Series - Datasheet 158 October 2012 Order Number: 327879-001US 4.0 I2C_EN set and the PEC_EN bit set produces undefined results. Software must force either I2C_EN or PEC_EN to 0 when running this command. See section 5.5.6 of the System Management Bus (SMBus) Specification, Version 2.0 for the protocol format. Note: For process call command, the value written into bit 0 of the Transmit Slave Address Register (SMB I/O register, offset 04h) needs to be 0. Note: If the I2C_EN bit is set, the protocol sequence changes slightly: the Command Code (bits 18:11 in the bit sequence) are not sent. As a result, the slave does not acknowledge (bit 19 in the sequence). Block Read/Write The system contains a 32-byte buffer for read and write data which can be enabled by setting bit 1 of the Auxiliary Control register at offset 0Dh in I/O space, as opposed to a single byte of buffering. This 32-byte buffer is filled with write data before transmission, and filled with read data on reception. The interrupt is generated only after a transmission or reception of 32 bytes, or when the entire byte count has been transmitted/received. The byte count field is transmitted but ignored as software will end the transfer after all bytes it cares about have been sent or received. For a Block Write, software must either force the I2C_EN bit or both the PEC_EN and AAC bits to 0 when running this command. The block write begins with a slave address and a write condition. After the command code, a byte count is issued to indicate how many more bytes will follow in the message. If a slave had 20 bytes to send, the first byte would be the number 20 (14h), followed by 20 bytes of data. The byte count may not be 0. A Block Read or Write is allowed to transfer a maximum of 32 data bytes. When programmed for a block write command, the Transmit Slave Address, Device Command, and Data0 (count) registers are sent. Data is then sent from the Block Data Byte register; the total data sent being the value stored in the Data0 Register. On block read commands, the first byte received is stored in the Data0 register, and the remaining bytes are stored in the Block Data Byte register. See section 5.5.7 of the System Management Bus (SMBus) Specification, Version 2.0 for the protocol format. Note: For Block Write, if the I2C_EN bit is set, the format of the command changes slightly. The host will still send the number of bytes (on writes) or receive the number of bytes (on reads) indicated in the DATA0 register. However, it does not send the contents of the DATA0 register as part of the message. Also, the Block Write protocol sequence changes slightly: the Byte Count (bits 27:20 in the bit sequence) are not sent - as a result, the slave does not send an acknowledge (bit 28 in the sequence). I2C Read This command allows the interface to perform block reads to certain I2C devices, such as serial E2PROMs. The SMBus Block Read supports the 7-bit addressing mode only. However, this does not allow access to devices using the I2C "Combined Format" that has data bytes after the address. Typically these data bytes correspond to an offset (address) within the serial memory chips. Note: This command is supported independent of the setting of the I2C_EN bit. The I2C Read command with the PEC_EN bit set produces undefined results. Software must force both the PEC_EN and AAC bit to 0 when running this command. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 159 For I2C Read command, the value written into bit 0 of the Transmit Slave Address Register (SMB I/O register, offset 04h) needs to be 0. The format that is used for the command is shown in Table 4-33. Table 4-33. I2C Block Read Bit 1 8:2 9 10 18:11 Description Start Slave Address -- 7 bits Write Acknowledge from slave Send DATA1 register 19 Acknowledge from slave 20 Repeated Start 27:21 Slave Address -- 7 bits 28 Read 29 Acknowledge from slave 37:30 38 46:39 47 Data byte 1 from slave -- 8 bits Acknowledge Data byte 2 from slave -- 8 bits Acknowledge - Data bytes from slave / Acknowledge - Data byte N from slave -- 8 bits - NOT Acknowledge - Stop The host will continue reading data from the peripheral until the NAK is received. Block Write-Block Read Process Call The block write-block read process call is a two-part message. The call begins with a slave address and a write condition. After the command code the host issues a write byte count (M) that describes how many more bytes is written in the first part of the message. If a master has 6 bytes to send, the byte count field has the value 6 (0000 0110b), followed by the 6 bytes of data. The write byte count (M) cannot be 0. The second part of the message is a block of read data beginning with a repeated start condition followed by the slave address and a Read bit. The next byte is the read byte count (N), which may differ from the write byte count (M). The read byte count (N) cannot be 0. The combined data payload must not exceed 32 bytes. The byte length restrictions of this process call are summarized as follows: * M 1 byte * N 1 byte * M + N 32 bytes Intel(R) Communications Chipset 89xx Series - Datasheet 160 October 2012 Order Number: 327879-001US 4.0 The read byte count does not include the PEC byte. The PEC is computed on the total message beginning with the first slave address and using the normal PEC computational rules. It is highly recommended that a PEC byte be used with the Block Write-Block Read Process Call. Software must do a read to the command register (offset 2h) to reset the 32 byte buffer pointer prior to reading the block data register. There is no STOP condition before the repeated START condition, and a NACK signifies the end of the read transfer. Note: E32B bit in the Auxiliary Control register must be set when using this protocol. See section 5.5.8 of the System Management Bus (SMBus) Specification, Version 2.0 for the protocol format. 4.12.1.3 Bus Arbitration Several masters may attempt to get on the bus at the same time by driving the SMBDATA line low to signal a start condition. The host continuously monitors the SMBDATA line. When the host is attempting to drive the bus to a 1 by letting go of the SMBDATA line and it samples SMBDATA low, then some other master is driving the bus and the host stops transferring data. If the host sees that it has lost arbitration, the condition is called a collision. The host sets the BUS_ERR bit in the Host Status Register, and if enabled, generate an interrupt or SMI#. The processor is responsible for restarting the transaction. When the host is the SMBus master, it drives the clock. When the host is sending address or command as an SMBus master, or data bytes as a master on writes, it drives data relative to the clock it is also driving. It does not start toggling the clock until the start or stop condition meets proper setup and hold time. the host also guarantees minimum time between SMBus transactions as a master. The system supports the same arbitration protocol for both the SMBus and the System Management Link (SMLink) interfaces. 4.12.1.4 Bus Timing 4.12.1.5 Clock Stretching Some devices may not be able to handle their clock toggling at the rate that the host, as an SMBus master, would like. They have the capability of stretching the low time of the clock. When the host attempts to release the clock, allowing the clock to go high, the clock remains low for an extended period of time. The host monitors the SMBus clock line after it releases the bus to determine whether to enable the counter for the high time of the clock. While the bus is still low, the high time counter must not be enabled. Similarly, the low period of the clock can be stretched by an SMBus master if it is not ready to send or receive data. 4.12.1.6 Bus Time Out (Host as SMBus Master) If there is an error in the transaction such that an SMBus device does not signal an acknowledge or holds the clock lower than the allowed time-out time, the transaction times out. the host discards the cycle and set the DEV_ERR bit. The time out minimum is 25 ms (800 RTC clocks). The time-out counter inside the host starts after the last bit of data is transferred by the host and it is waiting for a response. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 161 The 25 ms timeout counter does not count under the following conditions: 1. BYTE_DONE_STATUS bit (SMBus I/O Offset 00h, bit 7) is set 2. The SECOND_TO_STS bit (TCO I/O Offset 06h, bit 1) is not set (this indicates that the system has not locked up). 4.12.1.7 Interrupts / SMI# The Host's SMBus controller uses PIRQB# as its interrupt pin. However, the system can alternatively be set up to generate SMI# instead of an interrupt, by setting the SMBUS_SMI_EN bit (Device 31:Function 0:Offset 40h:bit 1). Table 4-34 and Table 4-35 specify how the various enable bits in the SMBus function control the generation of the interrupt, Host and Slave SMI, and Wake internal signals. The rows in the tables are additive, which means that if more than one row is true for a particular scenario then the results for all of the activated rows occurs. Table 4-34. Enable for SMBALERT# Event INTREN (Host Control I/O Register, Offset 02h, Bit 0) SMB_SMI_EN (Host Configuration Register, D31:F3:Offset 40h, Bit 1) SMBALERT_DIS (Slave Command I/ O Register, Offset 11h, Bit 2) X X X Wake generated X 1 0 Slave SMI# generated (SMBUS_SMI_STS) 1 0 0 Interrupt generated SMBALERT# asserted low (always reported in Host Status Register, Bit 5) Result Table 4-35. Enables for SMBus Slave Write and SMBus Host Events INTREN (Host Control I/O Register, Offset 02h, Bit 0) SMB_SMI_EN (Host Configuration Register, D31:F3:Offset 40h, Bit1) Slave Write to Wake/ SMI# Command X X Wake generated when asleep. Slave SMI# generated when awake (SMBUS_SMI_STS). Slave Write to SMLINK_SLAVE_SMI Command X X Slave SMI# generated when in the S0 state (SMBUS_SMI_STS) 0 X None 1 0 Interrupt generated 1 1 Host SMI# generated Event Any combination of Host Status Register [4:1] asserted Intel(R) Communications Chipset 89xx Series - Datasheet 162 Event October 2012 Order Number: 327879-001US 4.0 Table 4-36. Enables for the Host Notify Command 4.12.1.8 HOST_NOTIFY_INTREN (Slave Control I/O Register, Offset 11h, bit 0) SMB_SMI_EN (Host Config Register, D31:F3:Off40h, Bit 1) HOST_NOTIFY_WKEN (Slave Control I/O Register, Offset 11h, bit 1) 0 X 0 None X X 1 Wake generated 1 0 X Interrupt generated 1 1 X Slave SMI# generated (SMBUS_SMI_STS) Result SMBALERT# SMBALERT# is multiplexed with GPIO[11]. When enable and the signal is asserted, the system can generate an interrupt, an SMI#, or a wake event from S1-S5. 4.12.1.9 SMBus CRC Generation and Checking If the AAC bit is set in the Auxiliary Control register, the host automatically calculates and drives CRC at the end of the transmitted packet for write cycles, and checks the CRC for read cycles. It does not transmit the contents of the PEC register for CRC. The PEC bit must not be set in the Host Control register if this bit is set or it results in unspecified behavior. If the read cycle results in a CRC error, the DEV_ERR bit and the CRCE bit in the Auxiliary Status register at offset 0Ch is set. 4.12.2 TCO Slave SMBus Interface The internal TCO Slave shares the external pins with the Host SMBus. The TCO Slave SMBus logic does not generate or handle receiving the PEC byte and will only act as a Legacy Alerting Protocol device. The slave interface decodes address cycles, and allows an external microcontroller to perform specific actions. The TCO Slave is connected internally to the host SMBus by setting the soft trap TCO Slave Select in the flash descriptor. Key features and capabilities of the TCO Slave include: * Supports decode of three types of messages: Byte Write, Byte Read, and Host Notify. * Receive Slave Address register: This is the address that the slave decodes. A default value is provided so that the slave interface can be used without the processor having to program this register. * Receive Slave Data register in the SMBus I/O space that includes the data written by the external microcontroller. * Registers that the external microcontroller can read to get the state of the system. * Status bits to indicate that the SMBus slave logic caused an interrupt or SMI# due to the reception of a message that matched the slave address: -- Bit 0 of the Slave Status Register for the Host Notify command -- Bit 16 of the SMI Status Register for all others October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 163 The external microcontroller should not attempt to access the TCO SMBus slave logic until either: * One second after both: RTCRST# is high and RSMRST# is high, OR * PLTRST# de-asserts If a master leaves the clock and data bits of the SMBus interface at 1 for 50 s or more in the middle of a cycle, the slave logic's behavior is undefined. This is interpreted as an unexpected idle and should be avoided when performing management activities to the slave logic. Note: When an external microcontroller accesses the SMBus Slave Interface over the SMBus a translation in the address is needed to accommodate the least significant bit used for read/write control. For example, if the slave address (RCV_SLVA) is left at 44h (default), the external micro controller would use an address of 88h/89h (write/read). 4.12.2.1 Format of Slave Write Cycle The external master performs Byte Write commands to the SMBus Slave I/F. The "Command" field (bits 11:18) indicate which register is being accessed. The Data field (bits 20:27) indicate the value that should be written to that register. Table 4-37 provides the values associated with the Slave Write registers. Table 4-37. Slave Write Registers Register 0 1-3 Note: Function Command Register. See this register. Reserved 4 Data Message Byte 0 5 Data Message Byte 1 6-7 Reserved 8 Reserved 9-FFh Reserved The external microcontroller is responsible for making sure it does not update the contents of the data byte registers until they have been read by the system processor. The system overwrites the old value with any new value received. A race condition is possible where the new value is being written to the register just at the time it is being read. The system does not attempt to cover this race condition (i.e., unpredictable results in this case). Intel(R) Communications Chipset 89xx Series - Datasheet 164 Table 4-38, "Command Types" for legal values written to October 2012 Order Number: 327879-001US 4.0 . Table 4-38. Command Types Command Type 0 Reserved 1 WAKE/SMI# - This command wakes the system if it is not already awake. If system is already awake, an SMI# is generated. Note: The SMB_WAK_STS bit is set by this command, even if the system is already awake. The SMI handler should then clear this bit. 2 Unconditional Powerdown - This command sets the PWRBTNOR_STS bit, and has the same effect as the Power button Override occurring. 3 HARD RESET WITHOUT CYCLING - This command causes a hard reset of the system (does not include cycling of the power supply). This is equivalent to a write to the CF9h register with bits 2:1 set to 1, but bit 3 set to 0. 4 HARD RESET SYSTEM - This command causes a hard reset of the system (including cycling of the power supply). This is equivalent to a write to the CF9h register with bits 3:1 set to 1. 5 Disable the TCO Messages - This command will disable the Slave from sending Heartbeat and Event messages (as described in Section 4.12). Once this command has been executed, Heartbeat and Event message reporting can only be re-enabled by assertion and deassertion of the RSMRST# signal. 6 WD RELOAD - Reload watchdog timer. 7 Reserved 8 SMLINK_SLV_SMI - When the Slave detects this command type while in the S0 state, it sets the SMLINK_SLV_SMI_STS bit. This command should only be used if the system is in an S0 state. If the message is received during S1-S5 states, the system acknowledges it, but the SMLINK_SLV_SMI_STS bit does not get set. Note: It is possible that the system transitions out of the S0 state at the same time that the SMLINK_SLV_SMI command is received. In this case, the SMLINK_SLV_SMI_STS bit may get set but not serviced before the system goes to sleep. Once the system returns to S0, the SMI associated with this bit would then be generated. Software must be able to handle this scenario. 9-FFh 4.12.2.2 Description Reserved. Format of Read Command The external master performs Byte Read commands to the TCO SMBus Slave interface. The Command field (bits 18:11) indicate which register is being accessed. The Data field (bits 30:37) contain the value that should be read from that register. Table 4-39. Slave Read Cycle Format Bit 1 Description Driven by Comment Start External Microcontroller Slave Address - 7 bits External Microcontroller Must match value in Receive Slave Address register 9 Write External Microcontroller Always 0 10 ACK Slave Command code - 8 bits External Microcontroller 19 ACK Slave 20 Repeated Start External Microcontroller Slave Address - 7 bits External Microcontroller Must match value in Receive Slave Address register 28 Read External Microcontroller Always 1 29 ACK Slave 2-8 11-18 21-27 October 2012 Order Number: 327879-001US Indicates which register is being accessed. See Table 4-38 for list of implemented registers. Intel(R) Communications Chipset 89xx Series - Datasheet 165 Table 4-39. Slave Read Cycle Format Bit Description Driven by 30-37 Data Byte Slave 38 NOT ACK External Microcontroller 39 Stop External Microcontroller Comment Value depends on register being accessed. See Table 4-40 for list of implemented registers. Table 4-40. Data Values for Slave Read Registers (Sheet 1 of 2) Register Bits 0 07:00 Reserved for capabilities indication. Should always return 00h. Future chips may return another value to indicate different capabilities. 02:00 System Power State: 000 = S0 001 = S1 010 = Reserved 011 = S3 100 = S4 101 = S5 110 = Reserved 111 = Reserved 07:03 Reserved 03:00 Reserved 1 2 3 07:04 Reserved 05:00 Watchdog Timer current value. Watchdog Timer has 10 bits, but this field is only 6 bits. If the current value is greater than 3Fh, the system will always report 3Fh in this field. 07:06 4 Description Reserved 00 1 = The Intruder Detect (INTRD_DET) bit is set. This indicates that the system cover has probably been opened. 01 1 = BTI Temperature Event occurred. This bit is set if the system's THRM# input signal is active. Else this beat will read "0." 02 DOA CPU Status. This bit is 1 to indicate that the processor is dead 03 1 = SECOND_TO_STS bit set. This bit is set after the second time-out (SECOND_TO_STS bit) of the Watchdog Timer occurs. 06:04 07 Reserved. Will always be 0, but software should ignore. Reflects the value of the GPIO[11]/SMBALERT# pin (and is dependent upon the value of the GPI_INV[11] bit. If the GPI_INV[11] bit is 1, then the value in this bit equals the level of the GPI[11]/SMBALERT# pin (high = 1, low = 0). If the GPI_INV[11] bit is 0, then the value of this bit will equal the inverse of the level of the GPIO[11]/SMBALERT# pin (high = 0, low = 1). Intel(R) Communications Chipset 89xx Series - Datasheet 166 October 2012 Order Number: 327879-001US 4.0 Table 4-40. Data Values for Slave Read Registers (Sheet 2 of 2) Register Bits 00 Reserved 01 Reserved 02 CPU Power Failure Status: - `1' if the CPUPWR_FLR bit in the GEN_PMCON_2 register is set. 03 INIT3_3V# due to receiving Shutdown message - This event is visible from the reception of the shutdown message until a platform reset is done if the Shutdown Policy Select bit (SPS) is configured to drive INIT3_3V#. When the SPS bit is configured to generate PLTRST# based on shutdown, this register bit will always return 0. Events on signal do not create a event message 04 Reserved 05 POWER_OK_BAD - Indicates the failure core power well ramp during boot/resume. This bit is active if the SLP_S3# pin is de-asserted and PWROK pin is not asserted. 06 Thermal Trip - This bit will shadow the state of processor Thermal Trip status bit (CTS) (16.2.1.2, GEN_PMCON_2, bit 3). Events on signal does not create an event message. 07 Reserved - Default value is "X" Note: Software should not expect a consistent value when this bit is read through SMBUS/SMLink 6 07:00 Contents of the Message 1 register. See Section 6.8.1.8 for the description of this register. 7 07:00 Contents of the Message 2 register. See Section 6.8.1.8 for the description of this register. 8 07:00 Contents of the TCO_WDCNT register. See Section 6.8.1.9 for the description of this register. 9 07:00 Seconds of the RTC A 07:00 Minutes of the RTC B 07:00 Hours of the RTC 5 4.12.2.3 Description C 07:00 "Day of Week" of the RTC D 07:00 "Day of Month" of the RTC E 07:00 Month of the RTC F 07:00 Year of the RTC 10h-FFh 07:00 Reserved Behavioral Notes According to the SMBus protocol, Read and Write messages begin with a Start bit - Address- Write bit sequence. When the Slave detects that the address matches the value in the Receive Slave Address register, it assumes that the protocol is always followed and ignores the Write bit (bit 9) and signal an Acknowledge during bit 10. In other words, if a Start-Address-Read occurs (which is illegal for SMBus Read or Write protocol), and the address matches TCO Slave Address, the Slave still grabs the cycle. Also according to SMBus protocol, a Read cycle contains a Repeated Start-Address- Read sequence beginning at bit 20. If the Address matches the Slave's Receive Slave Address, it will assume that the protocol is followed, ignore bit 28, and proceed with the Slave Read cycle. Note: An external microcontroller must not attempt to access SMBus TCO Slave logic until at least 1 second after both RTCRST# and RSMRST# are deasserted (high). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 167 4.12.2.4 Slave Read of RTC Time Bytes The TCO SMBus slave interface allows external SMBus master to read the internal RTC's time byte registers. The RTC time bytes are internally latched by the system's hardware whenever RTC time is not changing and SMBus is idle. This ensures that the time byte delivered to the slave read is always valid and it does not change when the read is still in progress on the bus. The RTC time will change whenever hardware update is in progress, or there is a software write to the RTC time bytes. The SMBus slave interface only supports Byte Read operation. The external SMBus master will read the RTC time bytes one after another. Software must check and manage the possible time rollover when subsequent time bytes are read. For example, assuming the RTC time is 11 hours, 59 minutes, 59 seconds: When the external SMBus master reads the hour as 11, and then reads the minute, it is possible that the rollover happens between the reads and the minute is read as 0. This results in 11 hours, 0 minutes instead of the correct time of 12 hours, 0 minutes. Unless it is certain that rollover does not occur, software is required to detect the possible time rollover by reading multiple times such that the read time bytes can be adjusted accordingly if needed. 4.12.2.5 Format of Host Notify Command The system tracks and responds to the standard Host Notify command as specified in the System Management Bus (SMBus) Specification, Version 2.0. The host address for this command is fixed to 0001000b. If the system already has data for a previouslyreceived host notify command which has not been serviced yet by the host software (as indicated by the HOST_NOTIFY_STS bit), then it NACKs following the host address byte of the protocol. This allows the host to communicate non-acceptance to the master and retain the host notify address and data values for the previous cycle until host software completely services the interrupt. Note: Host software must always clear the HOST_NOTIFY_STS bit after completing any necessary reads of the address and data registers. Table 4-41. Host Notify Format Bit 1 Description Comment Start External Master SMB Host Address -- 7 bits External Master Always 0001_000 Write External Master Always 0 ACK (or NACK) Slave Slave NACKs if HOST_NOTIFY_STS is 1 Device Address - 7 bits External Master Indicates the address of the master; loaded into the Notify Device Address Register 18 Unused -- Always 0 External Master 7-bit-only address; this bit is inserted to complete the byte 19 ACK Slave Data Byte Low -- 8 bits External Master ACK Slave Data Byte High -- 8 bits External Master 37 ACK Slave 38 Stop External Master 8:2 9 10 17:11 27:20 28 36:29 Intel(R) Communications Chipset 89xx Series - Datasheet 168 Driven By Loaded into the Notify Data Low Byte Register Loaded into the Notify Data High Byte Register October 2012 Order Number: 327879-001US 4.0 4.12.2.6 TCO Slave Functions The TCO Slave can be assessed by an external microcontroller. The following features and functions are supported by the TCO Slave: * Processor present detection: Detects if processor fails to fetch the first instruction after reset * Various Error detection (such as ECC Errors) indicated by host controller: Can generate SMI#, SCI, SERR, NMI, or TCO interrupt * Intruder Detect input: -- Can generate TCO interrupt or SMI# when the system cover is removed -- INTRUDER# allowed to go active in any power state, including G3 * Detection of bad BIOS Flash (Flash on SPI) programming: Detects if data on first read is FFh (indicates that BIOS flash is not programmed) * Ability to hide a PCI device: Allows software to hide a PCI device in terms of configuration space through the use of a device hide register. 4.12.2.7 Theory of Operation The System Management functions are designed to allow the system to diagnose failing subsystems. The intent of this logic is that some of the system management functionality can be provided without the aid of an external microcontroller. 4.12.2.7.1 Detecting a System Lockup When the processor is reset, it is expected to fetch its first instruction. If the processor fails to fetch the first instruction after reset, the TCO timer times out twice and the system asserts PLTRST#. 4.12.2.7.2 Handling an Intruder The system has an input signal, INTRUDER#, that can be attached to a switch that is activated by the system's case being open. This input has a two RTC clock debounce. If INTRUDER# goes active (after the debouncer), this will set the INTRD_DET bit in the TCO_STS register. The INTRD_SEL bits in the TCO_CNT register can cause an SMI# or interrupt. The BIOS or interrupt handler can then cause a transition to the S5 state by writing to the SLP_EN bit. The software can also directly read the status of the INTRUDER# signal (high or low) by clearing and then reading the INTRD_DET bit. This allows the signal to be used as a GPI if the intruder function is not required. If the INTRUDER# signal goes inactive some point after the INTRD_DET bit is written as a 1, then the INTRD_DET signal will go to a 0 when INTRUDER# input signal goes inactive. This is slightly different than a classic sticky bit, since most sticky bits would remain active indefinitely when the signal goes active and would immediately go inactive when a 1 is written to the bit. Note: The INTRD_DET bit resides in the RTC well, and is set and cleared synchronously with the RTC clock. Thus, when software attempts to clear INTRD_DET (by writing a 1 to the bit location) there may be as much as two RTC clocks (about 65 s) delay before the bit is actually cleared. Also, the INTRUDER# signal should be asserted for a minimum of 1 ms to guarantee that the INTRD_DET bit is set. Note: If the INTRUDER# signal is still active when software attempts to clear the INTRD_DET bit, the bit remains set and the SMI is generated again immediately. The SMI handler October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 169 can clear the INTRD_SEL bits to avoid further SMIs. However, if the INTRUDER# signal goes inactive and then active again, there is not further SMIs, since the INTRD_SEL bits would select that no SMI# be generated. 4.12.2.7.3 Detecting Improper Flash Programming The system can detect the case where the BIOS flash is not programmed. This results in the first instruction fetched to have a value of FFh. If this occurs, the system sets the BAD_BIOS bit. The BIOS flash resides in SPI flash. 4.12.2.7.4 TCO Trigger Events Table 4-42 provides the events that can trigger TCO messages. The external BMC has to poll to detect pending messages. Table 4-42. Event Transitions that Cause Messages Event 4.12.3 Assertion? Deassertion? Comments INTRUDER# pin yes no Must be in "S1 or hung S0" state THRM# pin yes yes Must be in "S1 or hung S0" state. The THRM# pin is isolated when the core power is off, thus preventing this event in S3-S5. Watchdog Timer Expired yes no (NA) GPIO[11]/SMBALERT# pin yes yes Must be in "S1 or hung S0" state BATLOW# yes yes Must be in "S1 or hung S0" state CPU_PWR_FLR yes no "S1 or hung S0" state entered "S1 or hung S0" state entered EndPoint (EP) Slave SMBus This SMBus slave port provides System management (SM) visibility into all configuration registers in the EndPoint. The EP SMBus operations may be split into two upper level protocols: writing information to configuration registers and reading configuration registers. This section describes the required protocol for an SMBus master to access the EP's internal configuration registers. See the SMBus Specification, Revision 2.0 for the specific bus protocol, timings, and waveforms. EP SMBus Features: * The EP SMBus allows access to any register within the EP whether the CSR exists in PCI (bus, device, function) space or in memory mapped space. * The EP SMBus interface acts as a side-band configuration access to the internal configuration space, therefore, the config block must service all SMBus config transactions even in the presence of a deadlock condition where the system cannot respond to configuration requests on the PCIe* links. * The EP SMBus supports Packet Error Checking (can be disabled) as defined in the SMBus 2.0 specification. * The EP SMBus requires the SMBus master to poll the busy bit to determine if the previous transaction has completed. For reads, this is after the repeated start sequence. Intel(R) Communications Chipset 89xx Series - Datasheet 170 October 2012 Order Number: 327879-001US 4.0 4.12.3.1 SMBus Supported Transactions The EP SMBus supports six SMBus individual commands associated into two groups (read/write) with three data sizes. The read transactions require two SMBus sequences, one is writing the requested read address to the internal register stack and the other is the read command to extract the data once it is available. The write transactions are a single sequence that contains both the address and the data. Supported transactions: * Block Write (Dword sized data packet) * Block Read (Dword sized data packet) * Word Write (Word sized data packet) * Word Read (Word sized data packet) * Byte Write (Byte sized data packet) * Byte Read (Byte sized data packet) To support longer PCIe* timeouts the SMBus master is required to poll the busy bit to know when the data in the stack contains the desired data. This applies to both reads and writes. The protocol diagrams (Figure 4-6 through Figure 4-16) only shows the polling in read transactions. The reason for this is due to the length of PCIe* timeouts which may be as long as several seconds. This violates the SMBus spec of a maximum of 25 ms. To overcome this limitation, the SMBus slave requests the configuration master for access. Once granted, the slave asserts its busy bit and releases the link. The SMBus master is free to address other devices on the link or poll the busy bit until any pending transaction is completed. The command format is illustrated in Table 4-43 and the subsequent bulleted list of sub-field encodings. Table 4-43. SMBus Command Encoding 7 Begin 6 End 5 MemTrans 4 PEC_en 3:2 Internal Command: 00 - Read DWord 01 - Write Byte 10 - Write Word 11 - Write DWord 1:0 SMBus command: 00 - Byte 01 - Word 10 - Block 11 - Reserved. Block command is selected. * The Begin bit indicates the first transaction of the read or write sequence. * The End bit indicates the last transaction of the read or write sequence. * The MemTrans bit indicates the configuration request is a memory mapped addressed register or a PCI (bus, device, function, offset) addressed register. A logic 0 addresses a PCI configuration register. A logic 1 addresses a memory mapped register. When this bit is set it enables the designation memory address type. * The PEC_en bit enables the 8-bit packet error checking (PEC) generation and checking logic. For the examples below, if PEC was disabled, then there would be no PEC generated or checked by the slave. * The Internal Command field specifies the internal command to be issued by the SMBus slave logic. The EP SMBus supports dword reads and byte, word, and dword writes to configuration space. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 171 * The SMBus Command field specifies the SMBus command to be issued on the bus. This field is used as an indication of the length of transfer so that the slave knows when to expect the PEC packet (if enabled). The SMBus interface uses an internal register stack that is filled by the SMBus master before a request to the config master block is made. Table 4-44 provides a list of the bytes in the stack and their descriptions. Table 4-44. Internal SMBus Protocol Stack SMBus Stack usage for bus/dev/func commands (cmd[5] = 0) Description Command Command Command byte Byte Count Byte Count The number of bytes for this transaction when Block command is used. Bus Number Memory region Bus number for bus/dev config space command type. Memory region for memory config space command type. Device/Function Address [23:16] Device[4:0] and Function[2:0] for cmd[5] = 0 type of config transaction. Address[23:16] for cmd[5] = 1 type of memory config transaction. Address [15:8] The following fields are further defined for cmd[5]=0: Address High[7:4] = Reserved[3:0] Address High [3:0] = Address[11:8]: This is the high order PCIe* address field. The following fields are further defined for cmd[5]=1: Address[15:8] Register Offset Address [7:0] The following fields are further defined for cmd[5]=0: Lower order 8-bit register offset (Address[7:0]) The following fields are further defined for cmd[5]=1: Address [7:0] Data3 Data3 Data byte 3 Data2 Data2 Data byte 2 Data1 Data1 Data byte 1 Data0 Data0 Data byte 0 Address High 4.12.3.2 SMBus Stack usage for memory region commands (cmd[5] = 1) Addressing The EP SMBus slave address, which each component claims, is dependent on the GBE_LED0_0 and GBE_LED1_0 pin straps (sampled on the assertion of GBE_AUX_PWR_OK). The EP SMBus slave is accessed with address [7:1] = 1110_XX0. The X's represent GBE_LED0_0 and GBE_LED1_0 strap pins. See Table 4-45 for the mapping of strap pins to the bit positions of the slave address. Note: The slave address is dependent on the GBE_LED0_0 and GBE_LED1_0 strap pins only and cannot be reprogrammed. Intel(R) Communications Chipset 89xx Series - Datasheet 172 October 2012 Order Number: 327879-001US 4.0 Table 4-45. SMBus Slave Address Format Slave Address Field Bit Position Slave Address Source [7] 1 [6] 1 [5] 1 [4] 0 [3] GBE_LED1_0 strap pin [2] GBE_LED0_0 strap pin [1] 0 [0] Read/Write# bit. This bit is in the slave address field to indicate a read or write operation. It is not part of the SMBus slave address. If the Mem/Cfg (MemTrans) bit as described in Table 4-43 is cleared then the address field represents the standard PCI register addressing nomenclature namely; bus, device, function and offset. If the Mem/Cfg bit is set, the address field has a new meaning. Bits [23:0] hold a linear memory address and bits [31:24] is a byte to indicate which memory region it is. Table 4-46 describes the selections available. A logic one in a bit position enables that memory region to be accessed. If the destination memory byte is zero then no action is taken (no request is sent to the configuration master). If a memory region address field is set to a reserved space, the RHSL slave performs the following actions: * The transaction is not executed * The slave releases the SCL (Serial Clock) signal. * The master abort error status is set. Table 4-46. Memory Region Address Field Bit field 0Fh 0Eh-0Ah 4.12.3.3 Memory Region Address Field Internal SMBus Register Reserved 09h GbE3 08h GbE2 07h GbE1 06h GbE0 05h Reserved 04h Reserved 03h Reserved 02h Reserved 01h EndPoint Memory-Mapped CSRs 00h Reserved SMBus Initiated Southbound Configuration Cycles The platform SMBus master agent that is connected to the EP Slave SMBus can request a configuration transaction to a downstream PCI-Express* device. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 173 4.12.3.4 SMBus Error Handling SMBus Error Handling feature list: * Errors are reported in the status byte field. * Errors in Table 4-47 are also collected in the FERR and NERR registers. The EP SMBus slave interface handles two types of errors: internal and PEC. For example, internal errors can occur when the system issues a configuration read on the PCI-Express* port and that read terminates in error. These errors manifest as a NotAcknowledge (NACK) for the read command (End bit is set). If an internal error occurs during a configuration write, the final write command receives a NACK just before the stop bit. If the master receives a NACK, the entire configuration transaction should be reattempted. If the master supports packet error checking (PEC) and the PEC_en bit in the command is set, then the PEC byte is checked in the slave interface. If the check indicates a failure, then the slave NACKs the PEC packet. Each error bit must be routed to the FERR and NERR registers for error reporting. The status field encoding is defined in the Table 4-47. This field reports if an error occurred. If bits[2:0] are 000b then transaction was successful. A successful indication here does not necessarily mean that the transaction was completed correctly for all components in the system. The busy bit is set whenever a transaction is accepted by the slave. This is true for reads and writes but the affects may not be observable for writes. This means that since the writes are posted and the communication link is so slow the master should never see a busy condition. A timeout is associated with the transaction in progress. When the timeout expires a timeout error status is asserted. Table 4-47. Status Field Encoding for SMBus Reads Bit 07 4.12.3.5 Description Busy 06:03 Reserved 02:00 101-111: Reserved 100: Timeout Error 011: Master Abort. An error that is reported with respect to this transaction. 010: Completer Abort. An error is reported by downstream PCIe* device with respect to this transaction. 001: Memory Region encoding error. This bit is set if a memory region is not valid. 000: Successful SMBus Interface Reset The EP Slave interface state machine can be reset in several ways. The first two items are defined in the SMBus rev 2.0 specification. * The master holds SCL low for 25 ms cumulative. Cumulative in this case means that all the "low time" for SCL is counted between the Start and Stop bit. If this totals 25 ms before reaching the Stop bit, the interface is reset. * The master holds SCL continuously high for 50 us. * Force a platform reset. Intel(R) Communications Chipset 89xx Series - Datasheet 174 October 2012 Order Number: 327879-001US 4.0 4.12.3.6 Configuration and Memory Read Protocol Configuration and memory reads are accomplished through an SMBus write(s) and later followed by an SMBus read. The write sequence is used to initialize the Bus Number, Device, Function, and Register Number for the configuration access. The writing of this information can be accomplished through any combination of the supported SMBus write commands (Block, Word or Byte). The Internal Command field for each write should specify Read DWord. After all the information is set up, the last write (End bit is set) initiates an internal configuration read. The slave asserts a busy bit in the status register and release the link with an acknowledge (ACK). The master SMBus performs the transaction sequence for reading the data, however, the master must observe the status bit [7] (busy) to determine if the data is valid. This is due to the PCIe* timeouts which may be long causing an SMBus spec violation. The SMBus master must poll the busy bit to determine when the pervious read transaction has completed. If an error occurs then the status byte reports the results. This field indicates abnormal termination and contains status information such as target abort, master abort, and time-outs. Examples of configuration reads are illustrated below. All of these examples have PEC (Packet Error Code) enabled. If the master does not support PEC, then bit 4 of the command would be cleared and no PEC byte exists in the communication streams. For the definition of the diagram conventions below, see the SMBus Specification, Revision 2.0. For SMBus read transactions, the last byte of data (or the PEC byte if enabled) is NACKed by the master to indicate the end of the transaction. 4.12.3.6.1 SMBus Configuration and Memory Block-Size Reads Figure 4-5. SMBus Block-Size Configuration Register Read S Write address for a Read sequence Read data sequence 1110_1X0 W A Rsv[3:0] & Addr[11:8] A S 1110_1X0 W A Sr 1110_1X0 R Data [15:8] A A Cmd = 11010010 A Byte cnt = 4 A Regoff [7:0] A PEC A Cmd = 11010010 A Byte cnt = 5 A Status A Bus Num A Dev / Func A Data [31:24] A Data [23:16] A P Poll until Status[7] = 0 October 2012 Order Number: 327879-001US Data [7:0] A PEC N P Intel(R) Communications Chipset 89xx Series - Datasheet 175 Figure 4-6. SMBus Block-Size Memory Register Read S 1110_1X0 Write address for a Read sequence Read data sequence W A Addr off[15:8] A S 1110_1X0 W A Sr 1110_1X0 R Data [15:8] A A Cmd = 11110010 A Byte cnt = 4 A Addr off[7:0] A PEC A Cmd = 11110010 A Byte cnt = 5 A Status A MemRegion A Addr off[23:16] A Data [31:24] A Data [23:16] A P Poll until Status[7] = 0 Figure 4-7. Write address for a Read sequence Read Sequence Poll until Status[7] = 0 Data [7:0] PEC N P SMBus Word-Size Configuration Register Read S 1110_1X0 W A Cmd = 10010001 A Bus Num A Dev / Func A PEC A P S 1110_1X0 W A Cmd = 01010001 A Rsv[3:0] & Addr[11:8] A Regoff [7:0] A PEC A P S 1110_1X0 W A Cmd = 10010001 A Sr 1110_1X0 R Status A Data [31:24] A PEC N P S 1110_1X0 W A Cmd = 00010001 A Sr 1110_1X0 R Data [23:16] A Data [15:8] A PEC N P S 1110_1X0 W A Cmd = 01010000 A Sr 1110_1X0 R Data [7:0] A PEC N A A A Intel(R) Communications Chipset 89xx Series - Datasheet 176 A P October 2012 Order Number: 327879-001US 4.0 Figure 4-8. SMBus Word-Size Memory Register Read Write address for a Read sequence Read Sequence Poll until Status[7] = 0 S 1110_1X0 W A Cmd = 10110001 A Mem region A Addr off[23:16] A PEC A P S 1110_1X0 W A Cmd = 01110001 A Addr off[15:8] A Addr off[7:0] A PEC A P S 1110_1X0 W A Cmd = 10110001 A Sr 1110_1X0 R Status A Data [31:24] A PEC N P S 1110_1X0 W A Cmd = 00110001 A Sr 1110_1X0 R Data [23:16] A Data [15:8] A PEC N P S 1110_1X0 W A Cmd = 01110000 A Sr 1110_1X0 R Data [7:0] A PEC N A A A 4.12.3.6.2 SMBus Configuration and Memory Byte Reads Figure 4-9. SMBus Byte-Size Configuration Register Read Write address for a Read squence Read Sequence Poll until Status[7] = 0 P S 1110_1X0 W A Cmd = 10010000 A Bus Num A PEC A P S 1110_1X0 W A Cmd = 00010000 A Dev / Func A PEC A P S 1110_1X0 W A Cmd = 00010000 A Rsv[3:0] & Addr[11:8] A PEC A P S 1110_1X0 W A Cmd = 01010000 A Regoff [7:0] A PEC A P Cmd = 10010000 A Status A PEC N P Cmd = 00010000 A Data [31:24] A PEC N P Cmd = 00010000 A Data [23:16] A PEC N P Cmd = 00010000 A Data [15:8] A PEC N P Cmd = 01010000 A Data [7:0] A PEC N P S 1110_1X0 W A Sr 1110_1X0 R A S 1110_1X0 W A Sr 1110_1X0 R A S 1110_1X0 W A Sr 1110_1X0 R A S 1110_1X0 W A Sr 1110_1X0 R S 1110_1X0 W A Sr 1110_1X0 R October 2012 Order Number: 327879-001US A A Intel(R) Communications Chipset 89xx Series - Datasheet 177 Figure 4-10. SMBus Byte-Size Memory Register Read Write address for a Read squence Read Sequence Poll until Status[7] = 0 4.12.3.7 S 1110_1X0 W A Cmd = 10110000 A Bus Num A PEC A P S 1110_1X0 W A Cmd = 00110000 A Dev / Func A PEC A P S 1110_1X0 W A Cmd = 00110000 A Rsv[3:0] & Addr[11:8] A PEC A P S 1110_1X0 W A Cmd = 01110000 A Regoff [7:0] A PEC A P Cmd = 10110000 A Status A PEC N P Cmd = 00110000 A Data [31:24] A PEC N P Cmd = 00110000 A Data [23:16] A PEC N P Cmd = 00110000 A Data [15:8] A PEC N P Cmd = 01110000 A Data [7:0] A PEC N P S 1110_1X0 W A Sr 1110_1X0 R A S 1110_1X0 W A Sr 1110_1X0 R A S 1110_1X0 W A Sr 1110_1X0 R A S 1110_1X0 W A Sr 1110_1X0 R S 1110_1X0 W A Sr 1110_1X0 R A A Configuration and Memory Write Protocol Configuration and memory writes are accomplished through a series of SMBus writes. As with configuration reads, a write sequence is first used to initialize the Bus Number, Device, Function, and Register Number for the configuration access. The writing of this information can be accomplished through any combination of the supported SMBus write commands (Block, Word or Byte). Note: The SMBus has no concept of byte enables. Therefore, the Register Number written to the slave is assumed to be aligned to the length of the Internal Command. In other words, for a Write Byte internal command, the Register Number specifies the byte address. For a Write DWord internal command, the two least-significant bits of the Register Number or Address Offset are ignored. This is different from PCI where the byte enables are used to indicate the byte of interest. After all the information is set up, the SMBus master initiates one or more writes which sets up the data to be written. The final write (End bit is set) initiates an internal configuration write. The slave interface could potentially clock stretch the last data write until the write completes without error. If an error occurred, the SMBus interface NACKs the last write operation just before the stop bit. The busy bit is set for the write transaction. A config write will likely complete before the SMBus master can poll the busy bit. If the transaction is destined to a chip on a PCIe* link then it could take several more clock cycle to complete the outbound transaction being sent. Examples of configuration writes are illustrated below. For the definition of the diagram conventions below, see the SMBus Specification, Revision 2.0 Intel(R) Communications Chipset 89xx Series - Datasheet 178 October 2012 Order Number: 327879-001US 4.0 4.12.3.7.1 SMBus Configuration and Memory Block Writes Figure 4-11. SMBus Block-Size Configuration Register Write S 1110_1X0 W A Rsv[3:0] & Addr[11:8] A Data [7:0] A Cmd = 11011110 Regoff [7:0] A A PEC Byte cnt = 4 Data [31:24] A A A Bus Num Data [23:16] A A Dev / Func Data [15:8] A A P Figure 4-12. SMBus Block-Size Memory Register Write S 1110_1X0 W A Addr [15:8] A Data [7:0] A 4.12.3.7.2 Cmd = 11111110 Addr [7:0] PEC A A Byte cnt = 4 Data [31:24] A A A Mem Region Data [23:16] A A Addr [23:16] Data [15:8] A A P SMBus Configuration and Memory Word Writes Figure 4-13. SMBus Word-Size Configuration Register Write S 1110_1X0 W A Cmd = 10011001 A Bus Num A Dev / Func A PEC A P S 1110_1X0 W A Cmd = 00011001 A Rsv[3:0] & Addr[11:8] A Regoff [7:0] A PEC A P S 1110_1X0 W A Cmd = 00011001 A Data [31:24] A Data [23:16] A PEC A P S 1110_1X0 W A Cmd = 01011001 A Data [15:8] A Data [7:0] A PEC A P Figure 4-14. SMBus Word-Size Memory Register Write S 1110_1X0 W A Cmd = 10111001 A Mem Region A Addr [23:16] A PEC A P S 1110_1X0 W A Cmd = 00111001 A Addr [15:8] A Addr [7:0] A PEC A P S 1110_1X0 W A Cmd = 00111001 A Data [31:24] A Data [23:16] A PEC A P S 1110_1X0 W A Cmd = 01111001 A Data [15:8] A Data [7:0] A PEC A P October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 179 4.12.3.7.3 SMBus Configuration and Memory Byte Writes Figure 4-15. SMBus Configuration (Byte Write, PEC Enabled) S 1110_1X0 W A Cmd = 10010100 A Bus Num A PEC A P S 1110_1X0 W A Cmd = 00010100 A Dev / Func A PEC A P S 1110_1X0 W A Cmd = 00010100 A Rsv[3:0] & Addr[11:8] A PEC A P S 1110_1X0 W A Cmd = 00010100 A Regoff [7:0] A PEC A P S 1110_1X0 W A Cmd = 00010100 A Data [31:24] A PEC A P S 1110_1X0 W A Cmd = 00010100 A Data [23:16] A PEC A P S 1110_1X0 W A Cmd = 00010100 A Data [15:8] A PEC A P S 1110_1X0 W A Cmd = 01010100 A Data [7:0] A PEC A P Figure 4-16. SMBus Memory (Byte Write, PEC Enabled) S 1110_1X0 W A Cmd = 10110100 A Mem Region A PEC A P S 1110_1X0 W A Cmd = 00110100 A Addr[23:16] A PEC A P S 1110_1X0 W A Cmd = 00110100 A Rsv[3:0] & Addr[11:8] A PEC A P S 1110_1X0 W A Cmd = 00110100 A Regoff [7:0] A PEC A P S 1110_1X0 W A Cmd = 00110100 A Data [31:24] A PEC A P S 1110_1X0 W A Cmd = 00110100 A Data [23:16] A PEC A P S 1110_1X0 W A Cmd = 00110100 A Data [15:8] A PEC A P S 1110_1X0 W A Cmd = 01110100 A Data [7:0] A PEC A P 4.12.4 GbE SMBus (Master/Slave) See Chapter 22.0, "Management Interfaces" and Chapter 27.0, "GbE Platform Manageability" for the GbE SMBus interface description. Intel(R) Communications Chipset 89xx Series - Datasheet 180 October 2012 Order Number: 327879-001US 4.0 4.12.5 SMLINK1 Interface SMLINK1 is used by an external controller to obtain system thermal data from sensors integrated into the components on the platform. See Section 4.19, "Thermal Management" for information about Thermal Reporting using SMLINK1. 4.13 Serial I/O Unit and Watchdog Timer (SIW) (B0:D31:F0) 4.13.1 Overview The Serial I/O unit and Watchdog Timer (SIW) is similar to currently available Super I/O controllers. It is specifically designed to be integrated into the PCH and consists of two UARTs, a serial interrupt controller, a watchdog timer, and the LPC interface. 4.13.2 Features LPC Interface Multiplexed Command, Address, and Data Bus * 8-bit I/O transfers * 16-bit address qualification for I/O transactions * Serial IRQ interface compatible with serialized IRQ support for PCI systems Serial Port * Two full-function 16550-compatible serial ports * Configurable I/O addresses and interrupts * 16-byte FIFOs * Supports up to 115 Kbps * Programmable baud rate generator * Modem control circuitry * 14.7456 MHz and 48 MHz supported for UART baud clock input Watchdog Timer (WDT) Selectable Prescaler - about 1 MHz (1 s to 1 s) and about 1 KHz (1 ms to 10 min) * 33 MHz Clock (30 ns Clock Ticks) * Multiple Modes (WDT and Free-Running) * Free-Running Mode: One-stage timer - Toggles WDT_TOUT# after programmable time. * WDT Mode: Two-stage timer (First stage generates interrupt, second stage drives WDT_TOUT# low) -- First stage generates an SERIRQ interrupt (if enabled) after Programmable time. -- Second stage drives WDT_OUT# low or inverts the previous value. -- Used only after first timeout occurs. -- Status bit preserved in RTC well for possible error detection and correction. -- Drives WDT_TOUT# if OUTPUT is enabled. * Timer can be disabled (default state) or Locked (Hard reset required to disable WDT) * WDT Automatic Reload of Preload value when WDT Reload Sequence is performed. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 181 4.13.3 Functional Description 4.13.3.1 Host Processor Interface (LPC) The host processor communicates with the SIW via the LPC bus. Access is through a series of read/ write registers and accomplished through I/O cycles. All registers are eight bits wide. The SIW registers include global configuration space and device specific regions accessed by setting the Logical Device Number in the SIW Configuration Register 07H (SCR7). Table 4-48. Address Map Address Block Name Logical Device 04Eh or 2Eh (SIU1_DTR# dependent) Configuration Index 04Fh or 2Fh (SIU1_DTR# dependent) Configuration Data Base+(0-7) Serial Port 1 04H Base+(0-7) Serial Port 2 05H Base+(0-18) Watchdog Timer 06H See Section 7.1 for configuration register descriptions and information on setting the base address. 4.13.3.2 LPC Interface The LPC interface is used to control all the logical blocks on the SIW. LPC bus signals use PCI 33 MHz electrical signal characteristics. See the Low Pin Count (LPC) Interface Specification Rev 1.0. 4.13.3.3 LPC Cycles The following cycle types are supported by the LPC protocol. Table 4-49. Supported LPC Cycle Types Cycle Type Transfer Size I/O Write 1 Byte I/O Read 1 Byte The SIW ignores cycles that it does not support. 4.13.3.4 I/O Read and Write Cycles The SIW is the target for I/O cycles. I/O cycles are initiated by the host for register or FIFO accesses and generally have minimal synchronization times. Data transfers are assumed to be exactly 1-byte. If the CPU requested a 16-bit or 32bit transfer, the host must divide it up into 8-bit transfers. See the LPC Interface Specification for the sequence of cycles for the I/O Read and Write cycles. 4.13.3.5 Policy The following rules govern the reset policy: Intel(R) Communications Chipset 89xx Series - Datasheet 182 October 2012 Order Number: 327879-001US 4.0 SIW_RESET# is tied to the internal PCI bus reset. When SIW_RESET# goes active (low): * The host drives the LFRAME# signal high, tri-states the LAD[3:0] signals. * The SIU ignores LFRAME#, tri-states the LAD[3:0] pins. Note: LPC bus signals from SIW are internally tied to the primary LPC interface of the PCH. Host LPC and SIW LPC names are used interchangeably throughout. 4.13.3.6 LPC Transfers 4.13.3.6.1 I/O Transfers These are generally used for register or FIFO accesses, and generally have minimal synchronization times. The minimum number of wait-states between bytes is one. Data transfers are assumed to be exactly one byte. The host is responsible for breaking up larger data transfers into 8-bit cycles. Table 4-50. I/O Sync Bits Description Bits 4.13.4 Indication 0000 Synchronization achieved with no error. 0101 Indicates that synchronization not achieved yet, but the part is driving the bus. 0110 Indicates that synchronization not achieved yet, but the part is driving the bus and expects long synchronization 1010 Special Case: peripheral indicating errors. LPC Logical Devices 4 and 5: Serial Ports (UART1 and UART2) This section describes the Universal Asynchronous Receiver/Transmitter (UART) serial port used for the two UART integrated into the SIW. The UART can be controlled via programmed I/O. The basic programming model is the same for both UARTs with the only difference being the Logical Device Number assigned to each. The serial port consists of a UART which supports all the functions of a standard 16550 UART including hardware flow control interface. The UART performs serial-to-parallel conversion on data characters received from a peripheral device or a modem and parallel-to-serial conversion on data characters received from the processor. The processor can read the complete status of the UART at any time during the functional operation. Available status information includes the type and condition of the transfer operations being performed by the UART and any error conditions (parity, overrun, framing, or break interrupt). The serial port can operate in either FIFO or non-FIFO mode. In FIFO mode, a 16-byte transmit FIFO holds data from the processor to be transmitted on the serial link and a 16-byte Receive FIFO buffers data from the serial link until read by the processor. Each UART includes a programmable baud rate generator which is capable of dividing the baud clock input by divisors of one to (2 16 -1) and producing a 16X clock to drive the internal transmitter and receiver logic. Each UART has complete modem control capability and a processor interrupt system. Interrupts can be programmed to the user's requirements, minimizing the computing required to handle the communications link. Each UART can operate in a polled or an interrupt driven environment as configured by software. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 183 The baud rate generator input is a function of the UART_CLK and a configurable predivide of 1, 8, or 26. See SIW Configuration (address 29h) in Section 15.4.1, "SIW Configuration Register Summary" on page 745. The output of the baud rate generator is 16 times the baud rate. Table 4-51. UART Clock Divider Support Clock Frequency 1.8432 MHz 14.7456 MHz 48.0 MHz Predivide Value 1 8 26 Generator Frequency 1.8432 MHz 1.8432 MHz 1.8462 MHz Table 4-52. Baud Rate Example 4.13.4.1 Desired Baud Rate Divisor % error @ 1.8432 % error @ 1.8462* 300 384 0.16 1200 96 0.16 2400 48 0.16 4800 24 0.16 9600 12 0.16 19200 6 0.16 38400 3 56000 2 115200 1 0.16 2.8 3 0.16 UART Feature List * Functionally compatible with National Semiconductor's PC16550D * Adds or deletes standard asynchronous communications bits (start, stop, and parity) to or from the serial data * Independently controlled transmit, receive, line status and data set interrupts * Programmable baud rate generator allows division of clock by 1 to (216 -1) and generates an internal 16X clock * Modem control functions (CTS#, RTS#, DSR#, DTR#, RI#, and DCD#) * Fully programmable serial-interface characteristics: * 5, 6, 7, or 8-bit characters * Even, odd, or no parity detection * 1, 1-1/2, or 2 stop bit generation * Baud rate generation (up to 115 Kbps) * False start bit detection * 16-byte Receive FIFO * Complete status reporting capability * Line break generation and detection * Internal diagnostic capabilities include: -- Loopback controls for communications link fault isolation Intel(R) Communications Chipset 89xx Series - Datasheet 184 October 2012 Order Number: 327879-001US 4.0 -- Break, parity, overrun, and framing error simulation -- Fully prioritized interrupt system controls 4.13.4.2 UART Operational Description The format of a UART data frame is shown in Figure 4-17. Figure 4-17. Example UART Data Frame Each data frame is between seven bits and 12 bits long depending on the size of data programmed, if parity is enabled and if two stop bits is selected. The frame begins with a start bit that is represented by a high to low transition. Next, 5 to 8 bits of data are transmitted, beginning with the least significant bit. An optional parity bit follows, which is set if even parity is enabled and an odd number of ones exist within the data byte, or if odd parity is enabled and the data byte contains an even number of ones. The data frame ends with one, one and a half or two stop bits as programmed by the user, which is represented by one or two successive bit periods of a logic one. The unit is disabled upon reset, the user needs to enable the unit by setting bit six of Interrupt Enable Register. When the unit is enabled, the receiver starts looking for the start bit of a frame; the transmitter starts transmitting data to the transmit data pin if there is data available in the transmit FIFO. Transmit data can be written to the FIFO before the unit is enabled. When the unit is disabled, the transmitter/receiver finishes the current byte being transmitted/received if it is in the middle of transmitting/ receiving a byte and stops transmitting/receiving more data. An SIU_RESET# to the SIU forces the internal register and output signals on the serial port to the values listed in Table 4-53. Table 4-53. UART Register/Signal Reset States Register/Signal Reset Control Interrupt Enable Register RESET All bits are low. Interrupt ID Register RESET Bit 0 is forced high. Bits 1-3 and 6-7 are forced low. Bits 4-5 are permanently low. Line Control Register RESET All bits are forced low. Line Status Register RESET Bits 0-4, 7 are forced low. Bits 5 and 6 are forced high. Modem Control Register RESET Bits 0, 1, 2, 3, 4 are forced low. Bits 5, 6, 7 are permanently low. October 2012 Order Number: 327879-001US Reset State Intel(R) Communications Chipset 89xx Series - Datasheet 185 Table 4-53. UART Register/Signal Reset States 4.13.4.3 Register/Signal Reset Control Reset State Modem Status Register RESET/Modem signal, read MSR for bits 3-0. Infrared Selection Register RESET All bits are permanently low. Txd RESET High Int RESET/ clear LINE STATUS REG Low rts_n RESET High dtr_n RESET High Low Programmable Baud Rate Generator The UART contains a programmable Baud Rate Generator that is capable of taking the UART_CLK input and dividing it by any divisor from 1 to (2 16 -1). The output frequency of the Baud Rate Generator is 16 times the baud rate. Two 8-bit latches store the divisor in a 16-bit binary format. These Divisor Latches must be loaded during initialization to ensure proper operation of the Baud Rate Generator. If both Divisor Latches are loaded with 0, the 16X output clock is stopped. Upon loading either of the Divisor latches, a 16-bit baud counter is immediately loaded. This prevents long counts on initial load. Access to the Divisor latch can be done with a word write. Note: The UART_CLK is the SIW_CLK input divided by the prescaler set by the SIW Configuration Register (Offset 29h). The baud rate of the data shifted in/out of the UART is given by: Baud Rate = UART_CLK(MHz)/[16X Divisor] For example, if UART_CLK is 14.7456 MHz and the divisor is 96, the baud rate is 9600. A divisor value of 0 in the Divisor Latch Register is not allowed. The reset value of the divisor is 02. 4.13.4.4 FIFO Operation 4.13.4.4.1 FIFO Interrupt Mode Operation 4.13.4.4.2 Receiver Interrupt When the Receive FIFO and receiver interrupts are enabled (FCR[0]=1 and IER[0]=1), receiver interrupts occur as follows: * The receive data available interrupt is invoked when the FIFO has reached its programmed trigger level. The interrupt is cleared when the FIFO drops below the programmed trigger level. * The IIR receive data available indication also occurs when the FIFO trigger level is reached, and like the interrupt, the bits are cleared when the FIFO drops below the trigger level. * The receiver line status interrupt (IIR = C6H), as before, has the highest priority. The receiver data available interrupt (IIR=C4H) is lower. The line status interrupt occurs only when the character at the top of the FIFO has errors. Intel(R) Communications Chipset 89xx Series - Datasheet 186 October 2012 Order Number: 327879-001US 4.0 * The data ready bit (DR in LSR register) is set to 1 as soon as a character is transferred from the shift register to the Receive FIFO. This bit is reset to 0 when the FIFO is empty. 4.13.4.4.3 Character Timeout Interrupt When the receiver FIFO and receiver time out interrupt are enabled, a character timeout interrupt occurs when all of the following conditions exist: * At least one character is in the FIFO. * The last received character was longer than four continuous character times ago (if two stop bits are programmed the second one is included in this time delay). * The most recent processor read of the FIFO was longer than four continuous character times ago. * The receive FIFO trigger level is greater than one. The maximum time between a received character and a timeout interrupt is 160 ms at 300 baud with a 12-bit receive character (i.e., one start, eight data, one parity, and two stop bits). When a time out interrupt occurs, it is cleared and the timer is reset when the processor reads one character from the receiver FIFO. If a timeout interrupt has not occurred, the timeout timer is reset after a new character is received or after the processor reads the receiver FIFO. 4.13.4.4.4 Transmit Interrupt When the transmitter FIFO and transmitter interrupt are enabled (FCR[0]=1, IER[1]=1), transmit interrupts occur as follows: The transmitter holding register interrupt occurs when the transmit FIFO is empty; it is cleared as soon as the transmitter holder register is written to (1 to 16 characters may be written to the transmit FIFO while servicing this interrupt) or the IIR is read. The transmitter FIFO empty indications are delayed one character time minus the last stop bit time whenever the following occurs: THRE = 1 and there have not been at least two bytes at the same time in the transmit FIFO since the last THRE = 1. The first transmitter interrupt after changing FCRO is immediate if it is enabled. 4.13.4.5 FIFO Polled Mode Operation With the FIFOs enabled (TRFIFOE bit of FCR set to 1), setting IER[3:0] to all zeros puts the serial port in the FIFO polled mode of operation. Since the receiver and the transmitter are controlled separately, either one or both can be in the polled mode of operation. In this mode, software checks receiver and transmitter status via the LSR. As stated in the register description: * LSR[0] is set as long as there is one byte in the receiver FIFO. * LSR[1] through LSR[4] specify which error(s) has occurred for the character at the top of the FIFO. Character error status is handled the same way as interrupt mode. The IIR is not affected since IER[2] = 0. * LSR[5] indicates when the transmitter FIFO needs data. * LSR[6] indicates that both the transmitter FIFO and shift register are empty. * LSR[7] indicates whether there are any errors in the receiver FIFO. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 187 4.13.5 LPC Logical Device 6: Watchdog Timer 4.13.5.1 Overview This device is a Watchdog timer that provides a resolution that ranges from 1 s to 10 minutes. The timer uses a 35-bit down-counter. The counter is loaded with the value from the first Preload register. The timer is then enabled and it starts counting down. The time at which the WDT first starts counting down is called the first stage. If the host fails to reload the WDT before the 35-bit down counter reaches zero the WDT generates an internal interrupt. After the interrupt is generated the WDT loads the value from the second Preload register into the WDT's 35-bit Down-Counter and starts counting down. The WDT is now in the second stage. If the host still fails to reload the WDT before the second timeout, the WDT drives the WDT_TOUT# pin low and sets the timeout bit (WDT_TIMEOUT). This bit indicates that the system has become unstable. The WDT_TOUT# pin is held low until the system is reset or the WDT times out again (Depends on TOUT_CNF). The process of reloading the WDT involves the following sequence of writes: 1. Write "80" to offset BAR1 + 0Ch 2. Write "86" to offset BAR1 + 0Ch 3. Write `1' to WDT_RELOAD in Reload Register. The same process is used for setting the values in the preload registers. The only difference exists in step 3. Instead of writing a `1' to the WDT_RELOAD, you write the desired preload value into the corresponding Preload register. This value is not loaded into the 35-bit down counter until the next time the WDT reenters the stage. For example, if Preload Value 2 is changed, it is not loaded into the 35-bit down counter until the next time the WDT enters the second stage. Figure 4-18. WDT Block Diagram LPC LPC Interface Preload Value 1 Preload Value 2 Down-Counter Timeout/Interrupt Control Logic WDT_TOUT# (External) Intel(R) Communications Chipset 89xx Series - Datasheet 188 WDT_INT (Internal) October 2012 Order Number: 327879-001US 4.0 4.13.5.2 Theory Of Operation 4.13.5.2.1 RTC Well and WDT_TOUT# Functionality The WDT_TIMEOUT bit is set to a `1' when the WDT 35-bit down counter reaches zero for the second time in a row. Then the WDT_TOUT# pin is toggled LOW by the WDT from the ILB. The board designer must attach the WDT_TOUT# to the appropriate external signal. If WDT_TOUT_CNF is a `1' the WDT toggles WDT_TOUT# again the next time a time out occurs. Otherwise WDT_TOUT# is driven low until the system is reset or power is cycled. 4.13.5.2.2 Register Unlocking Sequence The register unlocking sequence is necessary whenever writing to the RELOAD register or either PRELOAD_VALUE registers. The host must write a sequence of two writes to offset BAR1 + 0Ch before attempting to write to either the WDT_RELOAD and WDT_TIMEOUT bits of the RELOAD register or the PRELOAD_VALUE registers. The first writes are "80" and "86" (in that order) to offset BAR1 + 0Ch. The next write is to the proper memory mapped register (e.g., RELOAD, PRELOAD_VALUE_1, PRELOAD_VALUE_2). Any deviation from the sequence (writes to memory-mapped registers) causes the host to have to restart the sequence. When performing register unlocking, software must issue the cycles using byte access only. Otherwise the unlocking sequence does not work properly. The following is an example of how to prevent a timeout: 1. Write "80" to offset BAR1 + 0Ch 2. Write "86" to offset BAR1 + 0Ch 3. Write a `1' to RELOAD [8] (WDT_RELOAD) of the Reload Register Note: Any subsequent writes require that this sequence be performed again. 4.13.5.2.3 Reload Sequence To keep the timer from causing an interrupt or driving WDT_TOUT#, the timer must be updated periodically. Other timers refer to "updating the timer" as "kicking the timer". The frequency of updates required is dependent on the value of the Preload values. To update the timer the Register Unlocking Sequence must be performed followed by writing a `1' to bit 8 at offset BAR1 + 0Ch within the watchdog timer memory mapped space. This sequence of events is referred to as the "Reload Sequence". 4.13.5.2.4 Low Power State The Watchdog Timer does not operate when PCICLK is stopped. 4.14 General Purpose I/O (B0:D31:F0) The system contains up to 67 General Purpose Input/Output (GPIO) signals. Each GPIO can be configured as an input or output signal. The number of inputs and outputs varies depending on the configuration. The following list is a brief summary of GPIO features: * Capability to mask Suspend well GPIOs from CF9h events configured via GP_RST_SEL registers) * Added capability to program GPIO prior to switching to output October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 189 4.14.1 Power Wells Some GPIOs exist in the suspend power plane. Care must be taken to make sure GPIO signals are not driven high into powered-down planes. Some GPIOs may be connected to pins on devices that exist in the core well. If these GPIOs are outputs, there is a danger that a loss of core power (PWROK low) or a Power Button Override event results in the system driving a pin to a logic 1 to another device that is powered down. 4.14.2 SMI# SCI and NMI Routing The routing bits for GPIO[15:0] allow an input to be routed to SMI#, SCI, NMI or neither. A bit can be routed to either an SMI# or an SCI, but not both. 4.14.3 Triggering GPIO[15:0] have "sticky" bits on the input. See the GPE0_STS register and the ALT_GPI_SMI_STS register. As long as the signal goes active for at least 2 clock cycles, the system keeps the sticky status bit active. The active level can be selected in the GP_INV register. This does not apply to GPI_NMI_STS residing in GPIO IO space. If the system is in an S0 or an S1 state, the GPI inputs are sampled at 33 MHz, so the signal only needs to be active for about 60 ns to be latched. In the S3-S5 states, the GPI inputs are sampled at 32.768 kHz, and thus must be active for at least 61 microseconds to be latched. GPIs that are in the core well are not capable of waking the system from sleep states where the core well is not powered. If the input signal is still active when the latch is cleared, it sets again. Another edge trigger is not required. This makes these signals "level" triggered inputs. 4.14.4 GPIO Registers Lockdown The following GPIO registers are locked down when the GPIO Lockdown Enable (GLE) bit is set. The GLE bit resides in B0:D31:F0:GPIO Control (GC) register. * Offset 00h: GPIO_USE_SEL[31:0] * Offset 04h: GP_IO_SEL[31:0] * Offset 0Ch: GP_LVL[31:0] * Offset 28h: GPI_NMI_EN[15:0] * Offset 2Ch: GPI_INV[31:0] * Offset 30h: GPIO_USE_SEL2[63:32] * Offset 34h: GPI_IO_SEL2[63:32] * Offset 38h: GP_LVL2[63:32] * Offset 40h: GPIO_USE_SEL3[95:64] * Offset 44h: GPI_IO_SEL3[95:64] * Offset 48h: GP_LVL3[95:64] * Offset 60h: GP_RST_SEL[31:0] * Offset 64h: GP_RST_SEL2[63:32] * Offset 68h: GP_RST_SEL3[95:64] Once these registers are locked down, they become Read-Only registers and any software writes to these registers has no effect. To unlock the registers, the GPIO Lockdown Enable (GLE) bit is required to be cleared to `0'. When the GLE bit changes from a `1' to a `0' a System Management Interrupt (SMI#) is generated if enabled. Intel(R) Communications Chipset 89xx Series - Datasheet 190 October 2012 Order Number: 327879-001US 4.0 Once the GPIO_UNLOCK_SMI bit is set, it can not be changed until a PLTRST# occurs. This ensures that only BIOS can change the GPIO configuration. If the GLE bit is cleared by unauthorized software, BIOS will set the GLE bit again when the SMI# is triggered and these registers will continue to be locked down. 4.14.5 Serial POST Codes Over GPIO The system provides an extended capability which allows system software to serialize POST or other messages on GPIO. This capability negates the requirement for dedicated diagnostic LEDs on the platform. Additionally, based on the newer BTX form factors, the PCI bus as a target for POST codes is increasingly difficult to support as the total number of PCI devices supported are decreasing. 4.14.5.1 Theory of Operation The POST code serialization logic is shared with GPIO. These GPIOs will likely be shared with LED control offered by the Super I/O (SIO) component. Figure 4-19 shows a likely configuration. Figure 4-19. Serial Post over GPIO Reference Circuit V_3P3_STBY R Post Code Logic SIO LED Note: The pull-up value is based on the brightness required. The anticipated usage model is that either the system or the SIO can drive a pin low to turn off an LED. In the case of the power LED, the SIO would normally leave its corresponding pin in a high-Z state to allow the LED to turn on. In this state, the system can blink the LED by driving its corresponding pin low and subsequently tristating the buffer. The I/O buffer should not drive a `1' when configured for this functionality and should be capable of sinking 24mA of current. An external optical sensing device can detect the on/off state of the LED. By externally post-processing the information from the optical device, the serial bit stream can be recovered. The hardware will supply a `sync' byte before the actual data transmission to allow external detection of the transmit frequency. The frequency of transmission should be limited to 1 transition every 1s to ensure the detector can reliably sample the on/off state of the LED. To allow flexibility in pull-up resistor values for power optimization, the frequency of the transmission is programmable via the DRS field in the GP_GB_CMDSTS register. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 191 The serial bit stream is Manchester encoded. This choice of transmission ensures that a transition is seen on every clock. The 1 or 0 data is based on the transmission happening during the high or low phase of the clock. As the clock is encoded within the data stream, hardware must ensure that the Z-0 and 0-Z transitions are glitch-free. Driving the pin directly from a flop or through glitch-free logic are possible methods to meet the glitch-free requirement. A simplified hardware/software register interface provides control and status information to track the activity of this block. Software enabling the serial blink capability should implement an algorithm referenced in the following list to send the serialized message on the enabled GPIO: 1. Read the Go/Busy status bit in the GP_GB_CMDSTS register and verify it is cleared. This will ensure that the GPIO is idled and a previously requested message is still not in progress. 2. Write the data to serialize into the GP_GB_DATA register. 3. Write the DLS and DRS values into the GP_GB_CMDSTS register and set the Go bit. This may be accomplished using a single write. The reference diagram shows the LEDs being powered from the suspend supply. By providing a generic capability that can be used both in the main and the suspend power planes maximum flexibility can be achieved. A key point to make is that the system does not unintentionally drive the LED control pin low unless a serialization is in progress. System board connections utilizing this serialization capability are required to use the same power plane controlling the LED as system GPIO pin. Otherwise, the system GPIO may float low during the message and prevent the LED from being controlled from the SIO. The hardware will only be serializing messages when the core power well is powered and the processor is operational. Care should be taken to prevent the system from driving an active `1' on a pin sharing the serial LED capability. Since the SIO could be driving the line to 0, having the system drive a 1 would create a high current path. A recommendation to avoid this condition involves choosing a GPIO defaulting to an input. The GP_SER_BLINK register should be set first before changing the direction of the pin to an output. This sequence ensures the open-drain capability of the buffer is properly configured before enabling the pin as an output. 4.14.5.2 Serial Message Format In order to serialize the data onto the GPIO, an initial state of high-Z is assumed. The SIO is required to have its LED control pin in a high-Z state as well to allow the system to blink the LED (see the reference diagram). The three components of the serial message include the sync, data, and idle fields. The sync field is 7 bits of `1' data followed by 1 bit of `0' data. Starting from the high-Z state (LED on) provides external hardware a known initial condition and a known pattern. In case one or more of the leading 1 sync bits are lost, the 1s followed by 0 provide a clear indication of `end of sync'. This pattern is used to `lock' external sampling logic to the encoded clock. The data field is shifted out with the highest byte first (MSB). Within each byte, the most significant bit is shifted first (MSb). The idle field is enforced by the hardware and is at least 2 bit times long. The hardware does not clear the Busy and Go bits until this idle time is met. Supporting the idle time in hardware prevents time-based counting in BIOS as the hardware is immediately ready for the next serial code when the Go bit is cleared. The idle state is represented as a high-Z condition on the pin. If the last transmitted bit is a 1, returning to the idle Intel(R) Communications Chipset 89xx Series - Datasheet 192 October 2012 Order Number: 327879-001US 4.0 state results in a final 0-1 transition on the output Manchester data. Two full bit times of idle correspond to a count of 4 time intervals (the width of the time interval is controlled by the DRS field). The waveform in Figure 4-20 shows a 1-byte serial write with a data byte of 5Ah. The internal clock and bit position are for reference purposes only. The Manchester D is the resultant data generated and serialized onto the GPIO. Since the buffer is operating in open-drain mode, the transitions are from high-Z to 0 and back. Figure 4-20. 1-byte Serial Write with a Data Byte of 5Ah Bit 7 6 5 4 3 2 1 0 Internal Clock Manchester D 8-bit sync field (1111_1110) 4.15 5A data byte 2 clk idle SATA Host Controller (B0:D31:F2, F5) The SATA Interface contains two controllers (B0:D31:F2 and B0:D31:F5) to support ACHI/IDE modes of operation for different operating systems. Each controller is configured to support two physical ports. These are physically numbered as Port 4 and Port 5. 4.15.1 SATA Ports 4 and 5 Numbering The SATA Interface is adopted from a legacy SATA Interface which supported six interface ports (ports 0-5) in both the IDE and AHCI modes. In the legacy interface usage, the controllers were used to support 6 ports as follows: * In ACHI mode, controller #1 supported all 6 ports (ports 0-5). * In Native IDE mode, controller #1 supported 4 ports (ports 0-3) and controller #2 supported 2 ports (ports 4 & 5) In this product, the controllers are used to support 2 ports as follows: * In ACHI mode, controller #1 supports 2 ports (ports 4 and 5). * In Native IDE mode, controller #2 supports 2 ports (ports 4 and 5) October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 193 4.15.2 SATA Feature Support Feature Description System Support Native Command Queuing (NCQ) Allows the device to reorder commands for more efficient data transfers Supported Auto Activate for DMA Collapses a DMA Setup then DMA Activate sequence into a DMA Setup only Supported Hot Plug Support Allows for device detection without power being applied and ability to connect and disconnect devices without prior notification to the system Supported Asynchronous Signal Recovery Provides a recovery from a loss of signal or establishing communication after hot plug Supported 3 Gb/s Transfer Rate Capable of data transfers up to 3Gb/s Supported ATAPI Asynchronous Notification A mechanism for a device to send a notification to the host that the device requires attention Supported Host & Link Initiated Power Management Capability for the host controller or device to request Partial and Slumber interface power states Supported Staggered Spin-Up Enables the host the ability to spin up hard drives sequentially to prevent power load problems on boot Supported Legacy Compatibility Mode Compatibility with Legacy Modes Not Supported Command Completion Coalescing Reduces interrupt and completion overhead by allowing a specified number of commands to complete and then generating an interrupt to process the commands Not Supported Port Multiplier A mechanism for one active host connection to communicate with multiple devices Not Supported External SATA Technology that allows for an outside the box connection of up to 2 meters (when using the cable defined in SATA-IO) Not Supported 4.15.3 Theory of Operation 4.15.3.1 Standard ATA Emulation The system contains a set of registers that shadow the contents of the legacy IDE registers. The behavior of the Command and Control Block registers, PIO, and DMA data transfers, resets, and interrupts are all emulated. Note: The system asserts INTR when the master device completes the EDD command regardless of the command completion status of the slave device. If the master completes EDD first, an INTR is generated and BSY remains '1' until the slave completes the command. If the slave completes EDD first, BSY is '0' when the master completes the EDD command and asserts INTR. Software must wait for busy to clear (0) before completing an EDD command, as required by the ATA5 through ATA7 (T13) industry standards. 4.15.3.2 48-Bit LBA Operation The SATA host controller supports 48-bit LBA through the host-to-device register FIS when accesses are performed via writes to the task file. The SATA host controller ensures that the correct data is put into the correct byte of the host-to-device FIS. Intel(R) Communications Chipset 89xx Series - Datasheet 194 October 2012 Order Number: 327879-001US 4.0 There are special considerations when reading from the task file to support 48-bit LBA operation. Software may need to read all 16-bits. Since the registers are only 8-bits wide and act as a FIFO, a bit must be set in the device/control register, which is at offset 3F6h for primary and 376h for secondary (or their native counterparts). If software clears bit 7 of the control register before performing a read, the last item written is returned from the FIFO. If software sets bit 7 of the control register before performing a read, the first item written is returned from the FIFO. 4.15.4 SATA Swap Bay Support The system provides for basic SATA swap bay support using the PSC register configuration bits and power management flows. A device can be powered down by software and the port can then be disabled, allowing removal and insertion of a new device. Note: This SATA swap bay operation requires board hardware (implementation specific), BIOS, and operating system support. 4.15.5 Hot Plug Operation The system supports Hot Plug Surprise removal and Insertion Notification in the PARTIAL, SLUMBER and Listen Mode states when used with Low Power Device Presence Detection. Software can take advantage of power savings in the low power states while enabling hot plug operation. See the AHCI specification for details. 4.15.5.1 Low Power Device Presence Detection Low Power Device Presence Detection enables SATA Link Power Management to coexist with hot plug (insertion and removal) without interlock switch or cold presence detect. The detection mechanism allows Hot Plug events to be detectable by hardware across all link power states (Active, PARTIAL, SLUMBER) as well as AHCI Listen Mode. If the Low Power Device Presence Detection circuit is disabled, the system reverts to Hot Plug Surprise Removal Notification (without an interlock switch) mode that is mutually exclusive of the PARTIAL and SLUMBER power management states. 4.15.6 Power Management Operation Power management of SATA controller and ports covers operations of the host controller and the SATA wire. 4.15.6.1 Power State Mappings The D0 PCI power management state for device is supported by SATA controller. SATA devices may also have multiple power states. From parallel ATA, three device states are supported through ACPI: * D0 - Device is working and instantly available. * D1 - device enters when it receives a STANDBY IMMEDIATE command. Exit latency from this state is in seconds * D3 - from the SATA device's perspective, no different than a D1 state, in that it is entered via the STANDBY IMMEDIATE command. However, an ACPI method is also called which resets the device and then cut its power. Each of these device states are subsets of the host controller's D0 state. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 195 SATA defines three PHY layer power states, which have no equivalent mappings to parallel ATA. They are: * PHY READY - PHY logic and PLL are both on and active * Partial - PHY logic is powered, but in a reduced state. Exit latency is no longer than 10 ns * Slumber - PHY logic is powered, but in a reduced state. Exit latency can be up to 10 ms. Since these states have much lower exit latency than the ACPI D1 and D3 states, the SATA controller defines these states as sub-states of the device D0 state. 4.15.6.2 Power State Transitions 4.15.6.2.1 Partial and Slumber State Entry/Exit The partial and slumber states save interface power when the interface is idle. It would be most analogous to PCI CLKRUN# (in power savings, not in mechanism), where the interface can have power saved while no commands are pending. The SATA controller defines PHY layer power management (as performed via primitives) as a driver operation from the host side, and a device proprietary mechanism on the device side. The SATA controller accepts device transition types, but does not issue any transitions as a host. All received requests from a SATA device is ACKed. When an operation is performed to the SATA controller such that it needs to use the SATA cable, the controller must check whether the link is in the Partial or Slumber states, and if so, must issue a COM_WAKE to bring the link back online. Similarly, the SATA device must perform the same action. 4.15.6.2.2 Device D1, D3 States These states are entered after some period of time when software has determined that no commands is sent to this device for some time. The mechanism for putting a device in these states does not involve any work on the host controller, other then sending commands over the interface to the device. The command most likely to be used in ATA/ATAPI is the "STANDBY IMMEDIATE" command. 4.15.6.2.3 Host Controller D3HOT State After the interface and device have been put into a low power state, the SATA host controller may be put into a low power state. This is performed via the PCI power management registers in configuration space. There are two important aspects to note when using PCI power management: 1. When the power state is D3, only accesses to configuration space are allowed. Any attempt to access the memory or I/O spaces results in master abort. 2. When the power state is D3, no interrupts may be generated, even if they are enabled. If an interrupt status bit is pending when the controller transitions to D0, an interrupt may be generated. When the controller is put into D3, it is assumed that software has properly shut down the device and disabled the ports. Therefore, there is no need to sustain any values on the port wires. The interface is treated as if no device is present on the cable, and power is minimized. When returning from a D3 state, an internal reset is not performed. Intel(R) Communications Chipset 89xx Series - Datasheet 196 October 2012 Order Number: 327879-001US 4.0 4.15.6.2.4 Non-AHCI Mode PME# Generation When in non-AHCI mode (legacy mode) of operation, the SATA controller does not generate PME#. This includes attach events (since the port must be disabled), or interlock switch events (via the SATAGP pins). 4.15.6.3 SMI Trapping (APM) Bus0:Device 31:Function2:Offset C0h contains control for generating SMI# on accesses to the IDE I/O spaces. These bits map to the legacy ranges (1F0-1F7h, 3F6h, 170-177h, and 376h) and native IDE ranges defined by PCMDBA, PCTLBA, SCMDBA an SCTLBA. If the SATA controller is in legacy mode and is using these addresses, accesses to one of these ranges with the appropriate bit set causes the cycle to not be forwarded to the SATA controller, and for an SMI# to be generated. If an access to the Bus-Master IDE registers occurs while trapping is enabled for the device being accessed, then the register is updated, an SMI# is generated, and the device activity status bits are updated indicating that a trap occurred. 4.15.7 TA LED The SATALED# output is driven whenever the BSY bit is set in any SATA port. The SATALED# is an active-low open-drain output. When SATALED# is low, the LED should be active. When SATALED# is high, the LED should be inactive. 4.15.8 AHCI Operation The system provides hardware support for Advanced Host Controller Interface (AHCI), a programming interface for SATA host controllers developed through a joint industry effort. AHCI defines transactions between the SATA controller and software and enables advanced performance and usability with SATA. Platforms supporting AHCI may take advantage of performance features such as no master/slave designation for SATA devices--each device is treated as a master--and hardware assisted native command queuing. AHCI also provides usability enhancements such as hot-plug. AHCI requires appropriate software support (e.g., an AHCI driver) and for some features, hardware support in the SATA device or additional platform hardware. The system supports all of the mandatory features of the Serial ATA Advanced Host Controller Interface Specification, Revision 1.2 and many optional features, such as hardware assisted native command queuing, aggressive power management, LED indicator support, and hot-plug through the use of interlock switch support (additional platform hardware and software may be required depending upon the implementation). Note: For reliable device removal notification while in AHCI operation without the use of interlock switches (surprise removal), interface power management should be disabled for the associated port. See Section 7.3.1 of the AHCI Specification for more information. 4.15.9 SGPIO Signals The SGPIO signals, in accordance to the SFF-8485 specification, support per-port LED signaling. These signals are not related to SATALED#, which allows for simplified indication of SATA command activity. The SGPIO group interfaces with an external controller chip that fetches and serializes the data for driving across the SGPIO bus. The output signals then control the LEDs. This feature is only valid in AHCI mode. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 197 4.15.9.1 Mechanism The enclosure management for SATA Controller 1 (Bus0:Device 31: Function 2) involves sending messages that control LEDs in the enclosure. The messages for this function are stored after the normal registers in the AHCI BAR, at Offset 580h bytes from the beginning of the AHCI BAR as specified by the EM_LOC global register. Software creates messages for transmission in the enclosure management message buffer. The data in the message buffer should not be changed if CTL.TM bit is set by software to transmit an update message. Software should only update the message buffer when CTL.TM bit is cleared by hardware otherwise the message transmitted is indeterminate. Software then writes a register to cause hardware to transmit the message or take appropriate action based on the message content. The software should only create message types supported by the controller, which is LED messages. If the software creates other non LED message types (e.g. SAF-TE, SES-2), the SGPIO interface may hang and the result is indeterminate. During reset all SGPIO pins is in tri-state state. The interface will continue to be in tristate state after reset until the first transmission occurs when software programs the message buffer and sets the transmit bit CTL.TM. The SATA Host controller will initiate the transmission by driving SCLOCK and at the same time drive the SLOAD to `0' prior to the actual bit stream transmission. The Host will drive SLOAD low for at least 5 SCLOCK then only start the bit stream by driving the SLOAD to high. SLOAD is driven high for 1 SCLOCK follow by vendor specific pattern that is default to "0000" if software has yet to program the value. A total of 21-bit stream from 7 ports (Port0, Port1, Port2, Port3, Port4 Port5 and Port6) of 3-bit per port LED message is transmitted on SDATAOUT0 pin after the SLOAD is driven high for 1 SCLOCK. Only 3 ports (Port4, Port5 and Port6) of 9 bit total LED message follow by 12 bits of tri-state value is transmitted out on SDATAOUT1 pin. All the default LED message values is high prior to software setting them, except the Activity LED message that is configured to be hardware driven that is generated based on the activity from the respective port. All the LED message values is driven to `1' for the port that is unimplemented as indicated in the Port Implemented register regardless of the software programmed value through the message buffer. There are two ways to reset the system's SGPIO interface: asynchronous reset and synchronous reset. Asynchronous reset is caused by platform reset to cause the SGPIO interface to be tri-state asynchronously. Synchronous reset is caused by setting the CTL.RESET bit, clearing the GHC.AE bit or HBA reset, where Host Controller completes the existing full bit stream transmission then only tri-state all the SGPIO pins. After the reset, both synchronous and asynchronous, the SGPIO pins stays tri-stated. Note: The Host Controller does not guarantee to cause the target SGPIO device or controller to be reset. Software is responsible to keep the SGPIO interface in tri-stated state for 2 seconds to cause a reset on the target of the SGPIO interface. Intel(R) Communications Chipset 89xx Series - Datasheet 198 October 2012 Order Number: 327879-001US 4.0 4.15.9.2 Message Format Messages are constructed with a one Dword header that describes the message to be sent followed by the actual message contents. The first Dword is constructed as follows: Bit 31:28 Description Reserved Message Type (MTYPE): Specifies the type of the message. The message types are: 0h = LED 27:24 1h = SAF-TE 2h = SES-2 3h = SGPIO (register based interface) All other values reserved 23:16 Data Size (DSIZE): Specifies the data size in bytes. If the message (enclosure services command) has a data buffer that is associated with it that is transferred, the size of that data buffer is specified in this field. If there is no separate data buffer, this field has have a value of `0'. The data directly follows the message in the message buffer. This value should always be `0'. 15:08 Message Size (MSIZE): Specifies the size of the message in bytes. The message size does not include the one Dword header. A value of `0' is invalid. The message size is always 4 bytes. 07:00 Reserved The SAF-TE, SES-2, and SGPIO message formats are defined in the corresponding specifications. The LED message type is defined in Section 4.15.9.3. It is the responsibility of software to ensure the content of the message format is correct. If the message type is not programmed as 'LED' for this controller, the controller shall not take any action to update its LEDs. For LED message type, the message size is always 4 bytes. 4.15.9.3 LED Message Type The LED message type specifies the status of up to three LEDs. Typically, the usage for these LEDs is activity, fault, and locate. Not all implementations necessarily contain all LEDs (for example, some implementations may not have a locate LED). The message identifies the HBA port number that the slot status applies to. The format of the LED message type is defined in Table 4-54. The LEDs shall retain their values until there is a following update for that particular slot. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 199 Table 4-54. Multi-Activity LED Message Type Byte Description Value (VAL) - This field describes the state of each LED for a particular location. There are three LEDs that may be supported by the HBA. Each LED has 3 bits of control. LED values are: 000b - LED shall be off 001b - LED shall be solid on as perceived by human eye All other values reserved 3-2 The Bits Bits Bits Bits LED bit locations are: 2:0 - Activity LED (may be driven by hardware) 5:3 - Vendor Specific LED (e.g. locate) 8:6 - Vendor Specific LED (e.g. fault) 15:9 - Reserved Vendor specific message is: Bit 3:0 - Vendor Specific Pattern Bit 15:4 - Reserved Note: If Activity LED Hardware Driven (ATTR.ALHD) bit is set, host outputs the hardware LED value sampled internally and ignores software written activity value on bit [2:0]. Since the Enclosure Management does not support port multiplier based LED message, the LED message is generated independently based on respective port's operation activity. Vendor specific LED values Locate (Bits 5:3) and Fault (Bits 8:6) always are driven by software. 1 Port Multiplier Information - Port Multiplier not supported. 0 HBA Information - Specifies slot specific information related to the HBA. Note: This feature is Not Supported. Bits 4:0 - HBA port number for the slot that requires the status update. Bit 5 - If set to '1', Value is a vendor specific message that applies to the entire enclosure. If cleared to '0', Value applies to the port specified in bits 4:0. Bits 7:6 - Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 200 October 2012 Order Number: 327879-001US 4.0 4.15.9.4 SGPIO Waveform Figure 4-21. Serial Data Transmitted Over the SGPIO Interface 4.16 High Precision Event Timers This function provides a set of timers that can be used by the operating system. The timers are defined such that in the future, the operating system may be able to assign specific timers to be used directly by specific applications. Each timer can be configured to cause a separate interrupt. The system provides eight timers. The timers are implemented as a single counter each with its own comparator and value register. This counter increases monotonically. Each individual timer can generate an interrupt when the value in its value register matches the value in the main counter. The registers associated with these timers are mapped to a memory space (much like the I/O APIC). However, it is not implemented as a standard PCI function. The BIOS reports to the operating system the location of the register space. The hardware can October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 201 support an assignable decode space; however, the BIOS sets this space prior to handing it over to the operating system. It is not expected that the operating system moves the location of these timers once it is set by the BIOS. 4.16.1 Timer Accuracy 1. The timers are accurate over any 1 ms period to within 0.05% of the time specified in the timer resolution fields. 2. Within any 100 microsecond period, the timer reports a time that is up to two ticks too early or too late. Each tick is less than or equal to 100 ns, so this represents an error of less than 0.2%. 3. The timer is monotonic. It does not return the same value on two consecutive reads (unless the counter has rolled over and reached the same value). The main counter is clocked by the 14.31818 MHz clock, synchronized into the 66.666 MHz domain. This results in a non-uniform duty cycle on the synchronized clock, but does have the correct average period. The accuracy of the main counter is as accurate as the 14.31818 MHz clock. 4.16.2 Interrupt Mapping Mapping Option #1 (Legacy Replacement Option) In this case, the Legacy Replacement Rout bit (LEG_RT_CNF) is set. This forces the mapping found in Table 4-55. Table 4-55. Legacy Replacement Routing Timer 8259 Mapping APIC Mapping Comment 0 IRQ0 IRQ2 In this case, the 8254 timer does not cause any interrupts 1 IRQ8 IRQ8 In this case, the RTC does not cause any interrupts. 2&3 Per IRQ Routing Field. Per IRQ Routing Field 4, 5, 6, 7 not available not available Mapping Option #2 (Standard Option) In this case, the Legacy Replacement Rout bit (LEG_RT_CNF) is 0. Each timer has its own routing control. The interrupts can be routed to various interrupts in the 8259 or I/O APIC. A capabilities field indicates which interrupts are valid options for routing. If a timer is set for edge-triggered mode, the timers should not be share with any PCI interrupts. The only supported interrupt values are as follows: * Timer 0 and 1: IRQ20, 21, 22 & 23 (I/O APIC only). * Timer 2: IRQ11 (8259 or I/O APIC) and IRQ20, 21, 22 & 23 (I/O APIC only). * Timer 3: IRQ12 (8259 or I/O APIC) and IRQ 20, 21, 22 & 23 (I/O APIC only). * Interrupts from Timer 4, 5, 6, 7 can only be delivered via direct FSB interrupt messages. Intel(R) Communications Chipset 89xx Series - Datasheet 202 October 2012 Order Number: 327879-001US 4.0 4.16.3 Periodic Versus Non-Periodic Modes Non-Periodic Mode Timer 0 is configurable to 32 (default) or 64-bit mode, whereas Timers 1, 2, and 3 only support 32-bit mode. All of the timers support non-periodic mode. See Section 2.3.9.2.1 of the IA-PC HPET Specification for a description of this mode. Periodic Mode Timer 0 is the only timer that supports periodic mode. See Section 2.3.9.2.2 of the IAPC HPET Specification for a description of this mode. The following usage model is expected: 1. Software clears the ENABLE_CNF bit to prevent any interrupts. 2. Software clears the main counter by writing a value of 00h to it. 3. Software sets the TIMER0_VAL_SET_CNF bit. 4. Software writes the new value in the TIMER0_COMPARATOR_VAL register 5. Software sets the ENABLE_CNF bit to enable interrupts. The Timer 0 Comparator Value register cannot be programmed reliably by a single 64-bit write in a 32-bit environment except if only the periodic rate is being changed during run-time. If the actual Timer 0 Comparator Value needs to be reinitialized, then the following software solution always works regardless of the environment: 1. Set TIMER0_VAL_SET_CNF bit. 2. Set the lower 32 bits of the Timer0 Comparator Value register. 3. Set TIMER0_VAL_SET_CNF bit. 4. Set the upper 32 bits of the Timer0 Comparator Value register. 4.16.4 Enabling the Timers The BIOS or operating system PnP code should route the interrupts. This includes the Legacy Rout bit, Interrupt Rout bit (for each timer), interrupt type (to select the edge or level type for each timer) The Device Driver code should do the following for an available timer: 1. Set the Overall Enable bit (Offset 10h, bit 0). 2. Set the timer type field (selects one-shot or periodic). 3. Set the interrupt enable. 4. Set the comparator value. 4.16.5 Interrupt Levels Interrupts directed to the internal 8259s are active high. See Section 4.7 for information regarding the polarity programming of the I/O APIC for detecting internal interrupts. If the interrupts are mapped to the 8259 or I/O APIC and set for level-triggered mode, they can be shared with PCI interrupts. They may be shared although it's unlikely for the operating system to attempt to do this. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 203 If more than one timer is configured to share the same IRQ (using the TIMERn_INT_ROUT_CNF fields), then the software must configure the timers to leveltriggered mode. Edge-triggered interrupts cannot be shared. 4.16.6 Handling Interrupts If each timer has a unique interrupt and the timer has been configured for edgetriggered mode, then there are no specific steps required. No read is required to process the interrupt. If a timer has been configured to level-triggered mode, then its interrupt must be cleared by the software. This is done by reading the interrupt status register and writing a 1 back to the bit position for the interrupt to be cleared. Independent of the mode, software can read the value in the main counter to see how time has passed between when the interrupt was generated and when it was first serviced. If Timer 0 is set up to generate a periodic interrupt, the software can check to see how much time remains until the next interrupt by checking the timer value register. 4.16.7 Issues Related to 64-Bit Timers with 32-Bit Processors A 32-bit timer can be read directly using processors that are capable of 32-bit or 64-bit instructions. However, a 32-bit processor may not be able to directly read 64-bit timer. A race condition comes up if a 32-bit processor reads the 64-bit register using two separate 32-bit reads. The danger is that just after reading one half, the other half rolls over and changes the first half. If a 32-bit processor needs to access a 64-bit timer, it must first halt the timer before reading both the upper and lower 32-bits of the timer. If a 32-bit processor does not want to halt the timer, it can use the 64-bit timer as a 32-bit timer by setting the TIMERn_32MODE_CNF bit. This causes the timer to behave as a 32-bit timer. The upper 32-bits are always 0. Alternatively, software may do a multiple read of the counter while it is running. Software can read the high 32 bits, then the low 32 bits, the high 32 bits again. If the high 32 bits have not changed between the two reads, then a rollover has not happened and the low 32 bits are valid. If the high 32 bits have changed between reads, then the multiple reads are repeated until a valid read is performed. On a 64-bit platform, if software attempts a 64 bit read of the 64-bit counter, software must be aware that some platforms may split the 64 bit read into two 32 bit reads. The read maybe inaccurate if the low 32 bits roll over between the high and low reads. 4.17 USB EHCI Host Controllers (B0:D29:F0) The system contains one Enhanced Host Controller Interface (EHCI) host controller which support up to six USB 2.0 high-speed root ports. USB 2.0 allows data transfers up to 480 Mb/s. USB 2.0 based Debug Port is also implemented. The following table summarizes the key features of the EHCI host controller. Intel(R) Communications Chipset 89xx Series - Datasheet 204 October 2012 Order Number: 327879-001US 4.0 Table 4-56. USB EHCI Features Parameter 4.17.1 USB EHCI Accessible by Memory Space Memory Data Structure Separated into Periodic and Asynchronous lists Differential Signaling Voltage 400 mV # of Ports 6 EHC Initialization Section 4.17.1.1, "BIOS Initialization" through Section 4.17.1.3, "EHC Resets" describe the expected Enhanced Host Controller (EHC) initialization sequence in chronological order, beginning with a complete power cycle in which the suspend well and core well have been off. 4.17.1.1 BIOS Initialization BIOS performs a number of platform customization steps after the core well has powered up. Note: Contact your Intel Field Representative for additional BIOS information. 4.17.1.2 Driver Initialization See Chapter 4 of the Enhanced Host Controller Interface Specification for Universal Serial Bus, Revision 1.0. 4.17.1.3 EHC Resets In addition to the standard hardware resets, portions of the EHC are reset by the HCRESET bit and the transition from the D3HOT device power management state to the D0 state. The effects of each of these resets are: Reset Does Reset Does not Reset Comments HCRESET bit set. Memory space registers except Structural Parameters (which is written by BIOS). Configuration registers. The HCRESET must only affect registers that the EHCI driver controls. PCI Configuration space and BIOS-programmed parameters can not be reset. Software writes the Device Power State from D3HOT (11b) to D0 (00b). Core well registers (except BIOSprogrammed registers). Suspend well registers; BIOSprogrammed core well registers. The D3-to-D0 transition must not cause wake information (suspend well) to be lost. It also must not clear BIOS-programmed registers because BIOS may not be invoked following the D3-to-D0 transition. If the detailed register descriptions give exceptions to these rules, those exceptions override these rules. This summary is provided to explain the reasons for the reset policies. 4.17.2 Data Structures in Main Memory See Section 3 and Appendix B of the Enhanced Host Controller Interface Specification for Universal Serial Bus, Revision 1.0 for details. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 205 4.17.3 USB 2.0 Enhanced Host Controller DMA The USB 2.0 EHC implements three sources of USB packets. They are, in order of priority on USB during each microframe: 1. The USB 2.0 Debug Port 2. The Periodic DMA engine 3. The Asynchronous DMA engine The system always performs any currently-pending debug port transaction at the beginning of a microframe, followed by any pending periodic traffic for the current microframe. If there is time left in the microframe, then the EHC performs any pending asynchronous traffic until the end of the microframe (EOF1). The debug port traffic is only presented on Port #1. The other ports are idle during this time. 4.17.4 Data Encoding and Bit Stuffing See Chapter 8 of the Universal Serial Bus Specification, Revision 2.0. 4.17.5 Packet Formats See Chapter 8 of the Universal Serial Bus Specification, Revision 2.0. The EHCI allows entrance to USB test modes, as defined in the USB 2.0 specification, including Test J, Test Packet, etc. However, Test Packet test mode interpacket gap timing may not meet the USB 2.0 specification. 4.17.6 USB 2.0 Interrupts and Error Conditions Section 4 of the Enhanced Host Controller Interface Specification for Universal Serial Bus, Revision 1.0 goes into detail on the EHC interrupts and the error conditions that cause them. All error conditions that the EHC detects can be reported through the EHCI Interrupt status bits. Only specific interrupt and error-reporting behavior is documented in this section. The EHCI Interrupts section must be read first, followed by this section of the datasheet to fully comprehend the EHC interrupt and error-reporting functionality. * Based on the EHC's buffer sizes and buffer management policies, the Data Buffer Error may never occur. * Master Abort and Target Abort responses from hub interface on EHC-initiated read packets is treated as Fatal Host Errors. The EHC halts when these conditions are encountered. * The system may assert the interrupts which are based on the interrupt threshold as soon as the status for the last complete transaction in the interrupt interval has been posted in the internal write buffers. The requirement in the Enhanced Host Controller Interface Specification for Universal Serial Bus, Revision 1.0 (that the status is written to memory) is met internally, even though the write may not be seen on DMI before the interrupt is asserted. * Since the system supports the 1024-element Frame List size, the Frame List Rollover interrupt occurs every 1024 milliseconds. * The system delivers interrupts using PIRQH#. * The system does not modify the CERR count on an Interrupt IN when the "Do Complete-Split" execution criteria are not met. * For complete-split transactions in the Periodic list, the "Missed Microframe" bit does not get set on a control-structure-fetch that fails the late-start test. If subsequent accesses to that control structure do not fail the late-start test, then the "Missed Microframe" bit is set and written back. Intel(R) Communications Chipset 89xx Series - Datasheet 206 October 2012 Order Number: 327879-001US 4.0 4.17.6.1 Aborts on USB 2.0-Initiated Memory Reads If a read initiated by the EHC is aborted, the EHC treats it as a fatal host error. The following actions are taken when this occurs: * The Host System Error status bit is set * The DMA engines are halted after completing up to one more transaction on the USB interface * If enabled (by the Host System Error Enable), then an interrupt is generated * If the status is Master Abort, then the Received Master Abort bit in configuration space is set * If the status is Target Abort, then the Received Target Abort bit in configuration space is set * If enabled (by the SERR Enable bit in the function's configuration space), then the Signaled System Error bit in configuration bit is set. 4.17.7 USB 2.0 Power Management 4.17.7.1 Pause Feature This feature allows platforms to dynamically enter low-power states during brief periods when the system is idle (i.e., between keystrokes). This is useful for enabling power management features. The policies for entering these states typically are based on the recent history of system bus activity to incrementally enter deeper power management states. Normally, when the EHC is enabled, it regularly accesses main memory while traversing the DMA schedules looking for work to do; this activity is viewed by the power management software as a non-idle system, thus preventing the power managed states to be entered. Suspending all of the enabled ports can prevent the memory accesses from occurring, but there is an inherent latency overhead with entering and exiting the suspended state on the USB ports that makes this unacceptable for the purpose of dynamic power management. As a result, the EHCI software drivers are allowed to pause the EHC's DMA engines when it knows that the traffic patterns of the attached devices can afford the delay. The pause only prevents the EHC from generating memory accesses; the SOF packets continue to be generated on the USB ports (unlike the suspended state). 4.17.7.2 Suspend Feature The Enhanced Host Controller Interface (EHCI) For Universal Serial Bus Specification, Section 4.3 describes the details of Port Suspend and Resume. 4.17.7.3 ACPI Device States The USB 2.0 function only supports the D0 and D3 PCI Power Management states. Notes regarding the implementation of the Device States: * The EHC hardware does not inherently consume any more power when it is in the D0 state than it does in the D3 state. However, software is required to suspend or disable all ports prior to entering the D3 state such that the maximum power consumption is reduced. * In the D0 state, all implemented EHC features are enabled. * In the D3 state, accesses to the EHC memory-mapped I/O range performs a master abort. Since the Debug Port uses the same memory range, the Debug Port is only operational when the EHC is in the D0 state. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 207 * In the D3 state, the EHC interrupt must never assert for any reason. The internal PME# signal is used to signal wake events, etc. * When the Device Power State field is written to D0 from D3, an internal reset is generated. See section EHC Resets for general rules on the effects of this reset. * Attempts to write any other value into the Device Power State field other than 00b (D0 state) and 11b (D3 state) completes normally without changing the current value in this field. 4.17.7.4 ACPI System States The EHC behavior as it relates to other power management states in the system is summarized in the following list: * The system is always in the S0 state when the EHC is in the D0 state. However, when the EHC is in the D3 state, the system may be in any power management state (including S0). * When in D0, the Pause feature (See Section 4.17.7.1) enables dynamic processor low-power states to be entered. * The PLL in the EHC is disabled when entering the S3/S4/S5 states (core power turns off). * All core well logic is reset in the S3/S4/S5 states. 4.17.8 USB 2.0 Legacy Keyboard Operation The system must support the possibility of a keyboard downstream from either a fullspeed/low-speed or a high-speed port. The description of the legacy keyboard support is unchanged from USB 1.1. The EHC provides the basic ability to generate SMIs on an interrupt event, along with more sophisticated control of the generation of SMIs. 4.17.9 USB 2.0 Based Debug Port The system supports the elimination of the legacy COM ports by providing the ability for new debugger software to interact with devices on a USB 2.0 port. High-level restrictions and features are: * * * * Operational before USB 2.0 drivers are loaded. Functions even when the port is disabled. Allows normal system USB 2.0 traffic in a system that may only have one USB port. Debug Port device (DPD) must be high-speed capable and connect directly to Port #1 of the system (e.g., the DPD cannot be connected to Port #1 through a hub. When a DPD is detected, the EHCI bypasses the integrated Rate Matching Hub and connect directly to the port and the DPD.). * Debug Port FIFO always makes forward progress (a bad status on USB is simply presented back to software). * The Debug Port FIFO is only given one USB access per microframe. The Debug port facilitates operating system and device driver debug. It allows the software to communicate with an external console using a USB 2.0 connection. Because the interface to this link does not go through the normal USB 2.0 stack, it allows communication with the external console during cases where the operating system is not loaded, the USB 2.0 software is broken, or where the USB 2.0 software is being debugged. Specific features of this implementation of a debug port are: * Only works with an external USB 2.0 debug device (console) Intel(R) Communications Chipset 89xx Series - Datasheet 208 October 2012 Order Number: 327879-001US 4.0 * Implemented for a specific port on the host controller * Operational anytime the port is not suspended AND the host controller is in D0 power state. * Capability is interrupted when port is driving USB RESET 4.17.9.1 Theory of Operation There are two operational modes for the USB debug port: 1. Mode 1 is when the USB port is in a disabled state from the viewpoint of a standard host controller driver. In Mode 1, the Debug Port controller is required to generate a "keepalive" packets less than 2 ms apart to keep the attached debug device from suspending. The keepalive packet should be a standalone 32-bit SYNC field. 2. Mode 2 is when the host controller is running (i.e., host controller's Run/Stop# bit is 1). In Mode 2, the normal transmission of SOF packets keeps the debug device from suspending. Behavioral Rules 1. In both Modes 1 and 2, the Debug Port controller must check for software requested debug transactions at least every 125 microseconds. 2. If the debug port is enabled by the debug driver, and the standard host controller driver resets the USB port, USB debug transactions are held off for the duration of the reset and until after the first SOF is sent. 3. If the standard host controller driver suspends the USB port, then USB debug transactions are held off for the duration of the suspend/resume sequence and until after the first SOF is sent. 4. The ENABLED_CNT bit in the debug register space is independent of the similar port control bit in the associated Port Status and Control register. Table 4-57 shows the debug port behavior related to the state of bits in the debug registers as well as bits in the associated Port Status and Control register. Table 4-57. Debug Port Behavior OWNER_CNT ENABLED_CT Port Enable Run / Stop Suspend Debug Port Behavior 0 X X X X Debug port is not being used. Normal operation. 1 0 X X X Debug port is not being used. Normal operation. 1 1 0 0 X Debug port in Mode 1. SYNC keepalives sent plus debug traffic 1 1 0 1 X Debug port in Mode 2. SOF (and only SOF) is sent as keepalive. Debug traffic is also sent. No other normal traffic is sent out this port because the port is not enabled. 1 1 1 0 0 Illegal. Host controller driver should never put controller into this state (enabled, not running and not suspended). 1 1 1 0 1 Port is suspended. No debug traffic sent. 1 1 1 1 0 Debug port in Mode 2. Debug traffic is interspersed with normal traffic. 1 1 1 1 1 Port is suspended. No debug traffic sent. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 209 4.17.9.1.1 OUT Transactions An Out transaction sends data to the debug device. It can occur only when the following are true: * The debug port is enabled * The debug software sets the GO_CNT bit * The WRITE_READ#_CNT bit is set The sequence of the transaction is as follows: 1. Software sets the appropriate values in the following bits: -- USB_ADDRESS_CNF -- USB_EndPoint_CNF -- DATA_BUFFER[63:0] -- TOKEN_PID_CNT[7:0] -- SEND_PID_CNT[15:8] -- DATA_LEN_CNT -- WRITE_READ#_CNT: (This is always 1 for OUT transactions) -- GO_CNT: (This is always 1 to initiate the transaction) 2. The debug port controller sends a token packet consisting of the following content: -- SYNC -- TOKEN_PID_CNT field -- USB_ADDRESS_CNT field -- USB_EndPoint_CNT field -- 5-bit CRC field 3. After sending the token packet, the debug port controller sends a data packet consisting of the following content: -- SYNC -- SEND_PID_CNT field -- The number of data bytes indicated in DATA_LEN_CNT from the DATA_BUFFER -- 16-bit CRC Note: A DATA_LEN_CNT value of 0 is valid in which case no data bytes would be included in the packet. 4. After sending the data packet, the controller waits for a handshake response from the debug device. * If a handshake is received, the debug port controller: a. Places the received PID in the RECEIVED_PID_STS field b. Resets the ERROR_GOOD#_STS bit c. Sets the DONE_STS bit * If no handshake PID is received, the debug port controller: a. Sets the EXCEPTION_STS field to 001b b. Sets the ERROR_GOOD#_STS bit c. Sets the DONE_STS bit Intel(R) Communications Chipset 89xx Series - Datasheet 210 October 2012 Order Number: 327879-001US 4.0 4.17.9.1.2 IN Transactions An IN transaction receives data from the debug device. An IN transaction can occur only when the following states are true: * The debug port is enabled * The debug software sets the GO_CNT bit * The WRITE_READ#_CNT bit is reset The sequence of the transaction takes place as follows: 1. Software sets the appropriate values in the following bits: -- USB_ADDRESS_CNF -- USB_EndPoint_CNF -- TOKEN_PID_CNT[7:0] -- DATA_LEN_CNT -- WRITE_READ#_CNT: (This is always 0 for IN transactions) -- GO_CNT: (This always 1 to initiate the transaction) 2. The debug port controller sends a token packet consisting of the following content: -- SYNC -- TOKEN_PID_CNT field -- USB_ADDRESS_CNT field -- USB_EndPoint_CNT field -- 5-bit CRC field. 3. After sending the token packet, the debug port controller waits for a response from the debug device. If a response is received: -- The received PID is placed into the RECEIVED_PID_STS field -- Any subsequent bytes are placed into the DATA_BUFFER -- The DATA_LEN_CNT field is updated to show the number of bytes that were received after the PID. 4. If a valid packet was received from the device that was one byte in length (indicating it was a handshake packet), then the debug port controller performs the following actions: -- Resets the ERROR_GOOD#_STS bit -- Sets the DONE_STS bit 5. If a valid packet was received from the device that was more than one byte in length (indicating it was a data packet), then the debug port controller performs the following actions: -- Transmits an ACK handshake packet -- Resets the ERROR_GOOD#_STS bit -- Sets the DONE_STS bit 6. If no valid packet is received, then the debug port controller performs the following actions: -- Sets the EXCEPTION_STS field to 001b -- Sets the ERROR_GOOD#_STS bit -- Sets the DONE_STS bit. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 211 4.17.9.1.3 Debug Software Enabling the Debug Port Debug software must address two mutually exclusive conditions as part of its startup processing: * The EHCI has been initialized by system software * The EHCI has not been initialized by system software Debug software can determine the current `initialized' state of the EHCI by examining the Configure Flag in the EHCI USB 2.0 Command Register. If this flag is set, then system software has initialized the EHCI. Otherwise the EHCI should not be considered initialized. Debug software initializes the debug port registers depending on the state of the EHCI. Before this can be accomplished, debug software must determine which root USB port is designated as the debug port. Determining the Debug Port Debug software can determine which USB root port has been designated as the debug port by examining bits 20:23 of the EHCI Host Controller Structural Parameters register. This 4-bit field represents the numeric value assigned to the debug port (i.e., 0001=port 1). Debug Software Startup with Non-Initialized EHCI Debug software can attempt to use the debug port if after setting the OWNER_CNT bit, the Current Connect Status bit in the appropriate (See Determining the Debug Port) PORTSC register is set. If the Current Connect Status bit is not set, then debug software may choose to terminate or it may choose to wait until a device is connected. If a device is connected to the port, then debug software must reset/enable the port. Debug software does this by setting and then clearing the Port Reset bit the PORTSC register. To guarantee a successful reset, debug software should wait at least 50 ms before clearing the Port Reset bit. Due to possible delays, this bit may not change to 0 immediately; reset is complete when this bit reads as 0. Software must not continue until this bit reads 0. If a high-speed device is attached, then the EHCI automatically sets the Port Enabled/ Disabled bit in the PORTSC register and the debug software can proceed. Debug software should set the ENABLED_CNT bit in the Debug Port Control/Status register, and then reset (clear) the Port Enabled/Disabled bit in the PORTSC register so the system host controller driver does not see an enabled port when it is first loaded. Debug Software Startup with Initialized EHCI Debug software can attempt to use the debug port if the Current Connect Status bit in the appropriate (See Determining the Debug Port) PORTSC register is set. If the Current Connect Status bit is not set, then debug software may choose to terminate or it may choose to wait until a device is connected. If a device is connected, then debug software must set the OWNER_CNT bit and then the ENABLED_CNT bit in the Debug Port Control/Status register. Determining Debug Peripheral Presence After enabling the debug port functionality, debug software can determine if a debug peripheral is attached by attempting to send data to the debug peripheral. If all attempts result in an error (Exception bits in the Debug Port Control/Status register Intel(R) Communications Chipset 89xx Series - Datasheet 212 October 2012 Order Number: 327879-001US 4.0 indicates a Transaction Error), then the attached device is not a debug peripheral. If the debug port peripheral is not present, then debug software may choose to terminate or it may choose to wait until a debug peripheral is connected. 4.17.10 EHCI Caching EHCI Caching is a power management feature in the USB (EHCI) host controllers which enables the controller to execute the schedules entirely in cache and eliminates the need for the DMA engine to access memory when the schedule is idle. EHCI caching allows the processor to maintain longer C-state residency times and provides substantial system power savings. 4.17.11 USB Pre-Fetch Based Pause The Pre-Fetch Based Pause is a power management feature in USB (EHCI) host controllers to ensure maximum C3/C4 processor power state time with C2 popup. This feature applies to the period schedule, and works by allowing the DMA engine to identify periods of idleness and preventing the DMA engine from accessing memory when the periodic schedule is idle. Typically in the presence of periodic devices with multiple millisecond poll periods, the periodic schedule is idle for several frames between polls. The USB Pre-Fetch Based Pause feature is disabled by setting bit 4 of EHCI Configuration Register. 4.17.12 USB Overcurrent Protection The system has implemented programmable USB Overcurrent signals. There are a total of four overcurrent pins shared across the six ports. Four overcurrent signals have been allocated to the ports in each USB device: * OC[3:0]# for Device 29 (Ports 0-5) Each pin is mapped to one or more ports by setting bits in the USBOCM1 register. See Section 7.1.1.68. It is system BIOS' responsibility to ensure that each port is mapped to only one over current pin. Operation with more than one overcurrent pin mapped to a port is undefined. It is expected that multiple ports are mapped to a single overcurrent pin, however they should be connected at the port and not at the pin. Shorting these pins together may lead to reduced test capabilities. 4.18 Integrated USB 2.0 Rate Matching Hub The system has an integrated USB 2.0 Rate Matching Hub (RMH). The Hub is connected to the EHCI controller as shown in Figure 4-22. The Hub convert low and full-speed traffic into high-speed traffic. When the RMH is enabled, it appears to software like an external hub is connected to Port 0 of each EHCI controller. In addition, port 1 the RMH is muxed with Port 1 of the EHCI controller and is able to bypass the RMH for use as the Debug Port. The hub operates like any USB 2.0 Discrete Hub and consumes one tier of hubs allowed by the section 4.1.1 in the USB 2.0 specification. A maximum of four additional nonroot hubs can be supported on any of the USB Ports. The RMH reports Vendor ID = 0x8087 and Product ID = 0x0020. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 213 Figure 4-22. EHCI with USB 2.0 with Rate Matching Hub ECHI USB 2.0 Rate Matching Hub (RMH) Port 0 4.18.1 Port 1 Port 2 Port 3 Port 4 Port 5 Architecture A hub consists of three components: the Hub Repeater, the Hub Controller, and the Transaction Translator. * The Hub Repeater is responsible for connectivity setup and tear-down. It also supports exception handling, such as bus fault detection and recovery and connect/ disconnect detect. * The Hub Controller provides the mechanism for host-to-hub communication. Hubspecific status and control commands permit the host to configure a hub and to monitor and control its individual downstream facing ports. * The Transaction Translator (TT) responds to high-speed split transactions and translates them to full-/low-speed transactions with full-/low-speed devices attached on downstream facing ports. There is 1 TT per RMH. See Chapter 11 of the USB 2.0 Specification for more details on the architecture of the hubs. 4.19 Thermal Management 4.19.1 Modes of Operation The PCH thermal sensor is available in Non-EndPoint and Normal modes. It has two modes of operation: * the analog mode (i.e. sequencer disabled) * the four-function mode, which includes a digital thermometer. The analog mode is kept primarily as a backup in case thermal time constants are found to be significantly faster than the temperature trip point detection rate of the newer four-function mode or the four-function mode has functional problems. Also, it is possible that the four-function mode has no utility beyond that of catastrophic/hot/aux detection, depending on the thermal sensor accuracy requirements of other usage models. The analog mode uses two comparators and only reports catastrophic and/or hot trips. The four-function mode uses one comparator that is time-multiplexed to perform 4 functions. A state machine switches the specific function of the thermal sensor and logic at a rate dependent on the Thermal Sensor Control [Sequencer Enable and Rate] Intel(R) Communications Chipset 89xx Series - Datasheet 214 October 2012 Order Number: 327879-001US 4.0 register setting, giving a separate output reading for each function. The sequencer can be configured in different modes to support various validation scenarios and fall-back modes. See the Thermal Sensor Control register (TSC) for detailed descriptions of the various modes and associated thermal sensor settling times. The five functions of the "four-function mode" are: * Catastrophic Trip Point - This trip point is set at the temperature at which the chip must be shut down immediately without any software support. For this logic to function, the catastrophic trip point must correspond to a temperature guaranteed to be functional. Special care must be taken to make sure that the sequencer logic, Power Management logic, and THRMTRIP# are functional at the catastrophic trip point. * Hot Temperature Trip Point - This trip point may be set dynamically if desired and provides an interrupt when it is crossed in either direction. Software could optionally set this as an "Interrupt me when the temperature goes above this level" setting. * Auxiliary Temperature Trip Point - This trip point is set below the Hot Temperature Trip Point and provides the same response options. However, the responses are separately programmable from the Hot Temp in order to provide incrementally more aggressive actions. The Aux trip point is fully software-programmable during system run time. * Auxiliary2 (Aux2) Temperature Trip Point - This trip point is typically set below the Aux Temperature Trip Point and provides interrupt generation capability. The Aux2 trip point is fully software-programmable during system run time. * Thermometer - The thermometer is implemented via a counter that starts at 0 and increments during each sample point until the comparator indicates that the temperature is above the current value. The value of the counter is loaded into a read-only register when the comparator first trips. 4.19.2 Thermal Reporting Over System Management Link 1 Interface (SMLink1) SMLink1 is the SMBus interface utilized by an External Management Controller (BMC/ EC) for Platform Thermal Reporting. It is used by the BMC/EC to control or obtain thermal sensor data from DIMMs, CPU, PCH, and other components integrated in the system. The Thermal Reporting features, functions, and interface configurations are described in the SMBus-to-PECI Bridge Protocol Application Note (Doc# 460415). 4.20 WatchDog Timer (WDT) Host Controller (B0:D31:F7) The Watchdog Host Controller device implements per-thread reset capability. The controller manifests itself as a PCI Express* legacy EndPoint. This device is MSI-X compatible. 4.20.1 Theory Of Operation The reset WDTs provide a mechanism for threads to be woken from a hung condition. A WDT is provisioned for each thread on the platform. Some platforms does not use/ enable all the WDTs. Each WDT can be loaded with a specific value and allowed to decrement. The time taken to fully decrement is referred to as the "interval" hereafter. The WDTnCOUNT registers stores the 32 bit reload value for the counter. The lower 10 bits of the reload value will always be zero. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 215 4.20.2 Watchdog Timer Behavior Each WDT behaves in an identical fashion. Each is clocked off a 10 bit prescaler counter. The prescaler counter is free running and cannot be disabled. The prescaler will run off the 125MHz backbone clock allowing a prescaler rollover time of up to 1K * 8ns = 8us. If the prescaler value is set to zero the individual WDTs will decrement at the 125MHz rate, otherwise the WDTs will decrement when the prescaler rolls over. Each 32 bit WDT counter can extend the interval to several hours (8e-06 * 232 ~ 32 Ksec). Expected usage is in the ms/second range. Given that the minimum valid value of a reload register is 0x400, the minimum interval that can be generated is ~8 us. When enabled properly and reloaded the WDT will start decrementing (under control of the prescaler) immediately and will continue to do so until it reaches zero unless it is reloaded (by a write to WDTnCMD) during that time. A reload will restart the WDT immediately if enabled. There are two events (and corresponding interrupts) that the unit generates when counting down: 1. The WARNing event is generated at approximately the halfway stage (when the counter value = WDTnCOUNT[31:1]). This results in an interrupt that should be sufficient to cause most threads to act. 2. The RESET event is generated when the WDT counter transitions from 0x1 to 0x0. this is intended to be a more severe interrupt that possibly resets a hung thread. 4.20.3 Pending Bit Array Each event will set a corresponding bit in the PBA. If the interrupt is not masked the device should attempt to deliver the corresponding interrupt as soon as possible after the event occurs. Interrupt delivery may of course be delayed by arbitration on the backbone bus. If a RESET event occurs before the earlier WARN interrupt has been delivered, the WARN PBA bit is cleared and the WARN interrupt can be cancelled. The RESET bit is set and the RESET interrupt send. When an interrupt is sent or cancelled, the PBA bit is cleared. Both the RESET and WARN PBA bits are cleared when the WDT is reloaded. The interrupt mask does not affect the setting of the PBA bits. 4.20.4 MSI Interrupt Formation The interrupts are generated as MSI interrupts. An MSI interrupt is a memory write PCIe* transaction. The device must deliver the interrupt as soon as possible after the corresponding PBA bit is set (unless the interrupt is masked by the Vector Control register or CMD.BME). An MSI packet is a memory write txn. For a given RESET/WARN interrupt, the packet is built as a 3DW header (because the upper half of the address is 0x00000000) with data as shown in Figure 4-23. Intel(R) Communications Chipset 89xx Series - Datasheet 216 October 2012 Order Number: 327879-001US 4.0 Figure 4-23. MSI Packet Header 7 6 5 10 4 3 2 1 0 7 0 0 0 0 0 R 6 5 0 R TC=0 4 Req ID = 0x00FF (Bus 0, Dev 31, Fn 7) 8 Lower Address - Contents of WDTnMAW/R 4 3 2 1 R 0 7 6 0 0 5 4 00 3 2 R 1 0 7 6 5 4 3 2 1 0 Length = 01 Tag = 0 0x0 0x3 The single DW of data is the contents of the WDTnMDW/R register. 4.20.5 Usage Models There are two models for setting up and initiating timer activity: * The first involves BIOS in setting timer values and all timers are enabled simultaneously: -- BIOS configures the TBAR content & required WDTnCOUNT/WDTnCFG registers. Values can be locked if necessary. -- Unused WDTs is left with WDTnCOUNT = 0x00, disabling them. These values can be locked if necessary. -- BIOS sets up the prescaler WDT_PSCALE. Lock if necessary -- BIOS reloads (write to WDTnCMD) all operating timers - no counting because of global enable. -- Later, OS globally enables all timers simultaneously using WDT_GBLCFG.GEN -- Threads are thereafter expected to avoid interrupts by reloading the WDT. * The second model allows threads to set their timer values & kick off timers individually: -- BIOS configures TBAR. Lock if necessary -- BIOS locks any unused WDTnCOUNT registers at 0x00 if necessary. -- BIOS sets up the prescaler WDT_PSCALE. Lock if necessary. -- BIOS globally enables using WDT_GBLCFG.GEN -- Later on individual threads can set the WDTnCOUNT values (locking if necessary). The threads can then initiate timer activity by writing to WDTnCMD. -- Threads are thereafter expected to avoid interrupts by reloading the WDT. 4.21 Serial Peripheral Interface (SPI) (B0:D31:F0) The Serial Peripheral Interface (SPI) is a 4-pin interface. This 4-pin SPI interface consists of clock (CLK), master data out (Master Out Slave In (MOSI)), master data in (Master In Slave Out (MISO)) and an active low chip select (SPI_CS[1:0]#). The system supports up to two SPI flash devices using two separate Chip Select pins. Each SPI flash device can be up to 16 MBytes. The SPI interface supports 20 MHz, 33 MHz and 50 MHz SPI devices. A SPI Flash device controlled with Chip Select 0 and with a valid descriptor MUST be attached directly to the system. Communication on the SPI bus is done with a Master-Slave protocol. The Slave is connected to the system and is implemented as a tri-state bus. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 217 SPI Flash has two operational modes, descriptor and non-descriptor. Non-Descriptor Mode is not supported. A valid flash descriptor is required for all platforms. 4.21.1 Descriptor Mode Descriptor Mode is required for all platforms. It enables many system features: * Intel(R) Management Engine Firmware * PCI Express* root port configuration * Supports up to two SPI components using two separate chip select pins -- Tow SPI Flash components or -- One SPI Flash and one user authentication device. * Hardware enforced security restricting master accesses to different regions * Soft Strap regions provides the ability to use Flash Non-Volatile Memory (NVM). * Supports the SPI Fast Read instruction and frequencies of up to 50 MHz * Uses standardized Flash Instruction Set 4.21.2 SPI Flash Regions In Descriptor Mode the Flash is divided into five separate regions: Region Content 0 Flash Descriptor 1 BIOS 2 Management Engine 3 Reserved 4 Platform Data Only two masters can access the four accessible regions: Host processor running BIOS code and Management Engine. The only required region is Region 0, the Flash Descriptor. Region 0 must be located in the first sector of device 0 (offset 0). Flash Region Sizes SPI flash space requirements differ by platform and configuration. The Flash Descriptor requires one 4 KB or larger block. The amount of flash space consumed is dependent on the erase granularity of the flash part and the platform requirements for the ME and BIOS regions. Table 4-58. Region Size Versus Erase Granularity of Flash Components Region Size with 4 KB Blocks Size with 64 KB Blocks Descriptor 4 KB 8 KB 64 KB BIOS Varies by Platform Varies by Platform Varies by Platform ME 128 KB 128 KB 128 KB Intel(R) Communications Chipset 89xx Series - Datasheet 218 Size with 8 KB Blocks October 2012 Order Number: 327879-001US 4.0 4.21.2.1 Device Partitioning The SPI Flash controller supports two sets of attributes in SPI flash space. This allows for supporting an asymmetric flash component that has two separate sets of attributes in the upper and lower part of the memory array. An example is a flash part that has different erase granularities in two different parts of the memory array. This allows for the usage of two separate flash vendors if using two different flash parts. Figure 4-24. Flash Partition Boundary Upper Flash Partition ... Flash Partition Boundary Lower Flash Partition 4.21.3 ... Flash Descriptor The maximum size of the Flash Descriptor is 4 KB. If the block/sector size of the SPI flash device is greater than 4 KB, the flash descriptor will only use the first 4 KB of the first block. The flash descriptor requires its own block at the bottom of memory (0x00h). The information stored in the Flash Descriptor can only be written during the manufacturing process as its read/write permissions must be set to Read only when the system leaves the manufacturing floor. The Flash Descriptor is made up of eleven sections: October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 219 Figure 4-25. Flash Descriptor Sections 4KB OEM Section Descriptor Upper MAP Management Engine VSCC Table Reserved System Soft Straps Master Region Component Descriptor MAP Signature 0 1. The Flash signature selects Descriptor Mode as well as verifies if the flash is programmed and functioning. The data at the bottom of the flash (offset 0) must be 0FF0A55Ah in order to be in Descriptor mode. 2. The Descriptor map has pointers to the other five descriptor sections as well as the size of each. 3. The component section has information about the SPI flash in the system including: the number of components, density of each, illegal instructions (such as chip erase), and frequencies for read, fast read and write/erase instructions. 4. The Region section points to the three other regions as well as the size of each region. 5. The master region contains the security settings for the flash, granting read/write permissions for each region and identifying each master by a requestor ID. See Section 4.21.3.1 for more information. 6. The Processor and System Soft Strap sections contain processor and system configurable parameters. 7. Same as #6. 8. The Reserved region between the top of the Processor Strap section and the bottom of the OEM Section is reserved for future chipset usages. 9. The Descriptor Upper MAP determines the length and base address of the Management Engine VSCC Table. Intel(R) Communications Chipset 89xx Series - Datasheet 220 October 2012 Order Number: 327879-001US 4.0 10. The Management Engine VSCC Table holds the JEDEC ID and the VSCC information of the entire SPI Flash supported by the NVM image. 11. OEM Section is 256 bytes reserved at the top of the Flash Descriptor for OEM use. 4.21.3.1 Descriptor Master Region The master region defines read and write access setting for each region of the SPI device. The master region recognizes two masters: BIOS and Management Engine. Each master is only allowed to do direct reads of its primary regions. Table 4-59. Region Access Control Table Master Read/Write Access Region 4.21.4 CPU and BIOS ME Descriptor N/A N/A BIOS CPU and BIOS can always read from and write to BIOS Region Read / Write Management Engine Read / Write ME can always read from and write to ME Region Platform Data Region N/A N/A Flash Access There are two types of flash accesses: * Direct Access: -- Masters are allowed to do direct read only of their primary region -- Master's Host or Management Engine virtual read address is converted into the SPI Flash Linear Address (FLA) using the Flash Descriptor Region Base/Limit registers * Program Register Access: -- Program Register Accesses are not allowed to cross a 4KB boundary and can not issue a command that might extend across two components -- Software programs the FLA corresponding to the region desired -- Software must read the devices Primary Region Base/Limit address to create a FLA. 4.21.4.1 Direct Access Security * Requester ID of the device must match that of the primary Requester ID in the Master Section * Calculated Flash Linear Address must fall between primary region base/limit * Direct Write not allowed * Direct Read Cache contents are reset to 0's on a read from a different master -- Supports the same cache flush mechanism in ICH7 which includes Program Register Writes October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 221 4.21.4.2 Register Access Security * Only primary region masters can access the registers * Masters are only allowed to read or write those regions they have read/write permission * Using the Flash Region Access Permissions, one master can give another master read/write permissions to their area * Using the five Protected Range registers, each master can add separate read/write protection above that granted in the Flash Descriptor for their own accesses -- Example: BIOS may want to protect different regions of BIOS from being erased -- Ranges can extend across region boundaries 4.21.5 Serial Flash Device Compatibility Requirements A variety of serial flash devices exist in the market. For a serial flash device to be compatible with the SPI bus, it must meet the minimum requirements detailed in the following sections. Note: All system platforms have the required Intel(R) Management engine firmware. 4.21.5.1 SPI Based BIOS Requirements A serial flash device must meet the following minimum requirements when used explicitly for system BIOS storage. * Erase size capability of at least one of the following: 64 Kbytes, 8 Kbytes, 4 Kbytes, or 256 bytes. * Device must support multiple writes to a page without requiring a preceding erase cycle (See Section 4.21.6) * Serial flash device must ignore the upper address bits such that an address of FFFFFFh aliases to the top of the flash memory. * SPI Compatible Mode 0 support (clock phase is 0 and data is latched on the rising edge of the clock). * If the device receives a command that is not supported or incomplete (less than 8 bits), the device must complete the cycle gracefully without any impact on the flash content. * An erase command (page, sector, block, chip, etc.) must set all bits inside the designated area (page, sector, block, chip, etc.) to 1 (Fh). * Status Register bit 0 must be set to 1 when a write, erase or write to status register is in progress and cleared to 0 when a write or erase is NOT in progress. * Devices requiring the Write Enable command mst automatically clear the Write Enable Latch at the end of Data Program instructions. * Byte write must be supported. The flexibility to perform a write between 1 byte to 64 bytes is recommended. * Hardware Sequencing requirements are optional in BIOS only platforms. * SPI flash parts that do not meet Hardware sequencing command set requirements may work in BIOS only platforms via software sequencing. Intel(R) Communications Chipset 89xx Series - Datasheet 222 October 2012 Order Number: 327879-001US 4.0 4.21.5.2 Intel(R) Management Engine Firmware SPI Flash Requirements Intel(R) Management Engine Firmware must meet the SPI flash based BIOS Requirements plus: * Hardware Sequencing. * Flash part must be uniform 4 KB erasable block throughout the entire device or have 64 KB blocks with the first block (lowest address) divided into 4 KB or 8 KB blocks. * Write protection scheme must meet SPI flash unlocking requirements for Management Engine. 4.21.5.2.1 SPI Flash Unlocking Requirements for Management Engine Flash devices must be globally unlocked (read, write and erase access on the ME region) from power on by writing 00h to the flash's status register to disable write protection. If the status register must be unprotected, it must use the enable write status register command 50h or write enable 06h. Opcode 01h (write to status register) must then be used to write a single byte of 00h into the status register. This must unlock the entire part. If the SPI flash's status register has non-volatile bits that must be written to, bits [5:2] of the flash's status register must be all 0h to indicate that the flash is unlocked. If bits [5:2] return a non zero values, the Intel(R) ME firmware will send a write of 00h to the status register. This must keep the flash part unlocked. If there is no need to execute a write enable on the status register, then opcodes 06h and 50h must be ignored. After global unlock, BIOS has the ability to lock down small sections of the flash as long as they do not involve the ME region. 4.21.5.3 Hardware Sequencing Requirements The following table provides a list of commands and the associated opcodes that a SPIbased serial flash device must support in order to be compatible with hardware sequencing. Table 4-60. Hardware Sequencing Commands and Opcode Requirements Commands Opcode Notes Write to Status Register 01h Writes a byte to SPI flash's status register. Enable Write to Status Register command must be run prior to this command. Program Data 02h Single byte or 64 byte write as determined by flash part capabilities and software. Read Data 03h Write Disable 04h Read Status 05h Write Enable 06h Fast Read 0Bh Enable Write to Status Register 50h or 60h October 2012 Order Number: 327879-001US Outputs contents of SPI flash's status register Enables a bit in the status register to allow an update to the status register Intel(R) Communications Chipset 89xx Series - Datasheet 223 Table 4-60. Hardware Sequencing Commands and Opcode Requirements Commands 4.21.5.3.1 Opcode Erase Programm able Full Chip Erase C7h JEDEC ID 9Fh Notes 256B, 4 Kbyte, 8 Kbyte or 64 Kbyte See Section 4.21.5.3.1. JEDEC ID Since each serial flash device may have unique capabilities and commands, the JEDEC ID is the necessary mechanism for identifying the device so the uniqueness of the device can be comprehended by the controller (master). The JEDEC ID uses the opcode 9Fh and a specified implementation and usage model. This JEDEC Standard Manufacturer and Device ID read method is defined in Standard JESD21-C, PRN03-NV. 4.21.6 Multiple Page Write Usage Model The system BIOS usage model require that the serial flash device support multiple writes to a page (minimum of 512 writes) without requiring a preceding erase command. BIOS commonly uses capabilities such as counters that are used for error logging and system boot progress logging. These counters are typically implemented by using byte-writes to `increment' the bits within a page that have been designated as the counter. Note: This usage model requirement is based on any given bit only being written once from a `1' to a `0'without requiring the preceding erase. An erase would be required to change bits back to the 1 state. 4.21.6.1 Soft Flash Protection There are two types of flash protection that are not defined in the flash descriptor but supported by the system: 1. BIOS Range Write Protection 2. SMI#-Based Global Write Protection Both mechanisms are logically OR'd together such that if any of the mechanisms indicate that the access should be blocked, then it is blocked. Table 4-61 provides a summary of the mechanisms. Table 4-61. Flash Protection Mechanism Summary Accesses Blocked Mechanism Range Specific? Reset-Override or SMI#Override? BIOS Range Write Protection Writes Yes Reset Override Write Protect Writes No SMI# Override A blocked command will appear to software to finish, except that the Blocked Access status bit is set in this case. Intel(R) Communications Chipset 89xx Series - Datasheet 224 October 2012 Order Number: 327879-001US 4.0 4.21.6.2 BIOS Range Write Protection The system provides a method for blocking writes to specific ranges in the SPI flash when the Protected BIOS Ranges are enabled. This is achieved by checking the Opcode type information (which can be locked down by the initial Boot BIOS) and the address of the requested command against the base and limit fields of a Write Protected BIOS range. Note: Once BIOS has locked down the Protected BIOS Range registers, this mechanism remains in place until the next system reset. 4.21.6.3 SMI# Based Global Write Protection The system provides a way to block writes to the SPI flash when the Write Protected bit is cleared (i.e., protected). This is achieved by checking the Opcode type information (which can be locked down by the initial boot BIOS) of the requested command. 4.21.7 Flash Device Configurations All platforms must have a SPI flash connected directly with a valid descriptor and Intel(R) Management Engine Firmware. 4.21.8 SPI Flash Device Recommended Pinout The table below provides the recommended serial flash device pin-out for an 8-pin device. Using the recommended pin-out on an 8-pin device reduces complexities involved with designing the serial flash device onto a motherboard and allows for support of a common footprint usage model (see Section 4.21.9.1). Table 4-62. Recommended Pinout for 8-Pin Serial Flash Device Pin # Signal 1 Chips Select 2 Data Output 3 Write Protect 4 Ground 5 Data Input 6 Serial Clock 7 Hold / Reset 8 Supply Voltage Although an 8-pin device is preferred due to footprint compatibility, the following table contains the recommended serial flash device pinout for a 16-pin SOIC. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 225 4.21.9 Serial Flash Device Package Table 4-63. Recommended Pinout for 16-Pin Serial Flash Device 4.21.9.1 Pin # Signal Pin # Signal 1 Hold / Reset 9 Write Protect 2 Supply Voltage 10 Ground 3 No Connect 11 No Connect 4 No Connect 12 No Connect 5 No Connect 13 No Connect 6 No Connect 14 No Connect 7 Chip Select 15 Serial Data In 8 Serial Data Out 16 Serial Clock Common Footprint Usage Model To minimize platform motherboard redesign and to enable platform Bill of Material (BOM) selectability, many PC System OEMs design their motherboard with a single common footprint. This common footprint allows population of a soldered down device or a socket that accepts a leadless device. This enables the board manufacturer to support, via selection of the appropriate BOM, either of these solutions on the same system without requiring any board redesign. The common footprint usage model is desirable during system debug and by flash content developers since the leadless device can be easily removed and reprogrammed without damage to device leads. When the board and flash content is mature for highvolume production, both the socketed leadless solution and the soldered down leaded solution are available through BOM selection. 4.21.9.2 Serial Flash Device Package Recommendations It is highly recommended that the common footprint usage model be supported. An example of how this can be accomplished is as follows: * The recommended pinout for 8-pin serial flash devices is used (see Section 4.21.8). * The 8-pin device is supported in either an 8-contact VDFPN (6x5 mm MLP) package or an 8-contact WSON (5x6 mm) package. These packages can fit into a socket that is land pattern compatible with the wide body SO8 package. * The 8-pin device is supported in the SO8 (150 mil) and in the wide-body SO8 (200 mil) packages. The 16-pin device is supported in the SO16 (300 mil) package. Intel(R) Communications Chipset 89xx Series - Datasheet 226 October 2012 Order Number: 327879-001US 4.0 4.22 Feature Capability Mechanism A set of registers is included in the LPC Interface (Bus0:Device 31, Function 0, offset E0h - EBh) that allows the system software or BIOS to easily determine system supported features. These registers can be accessed through LPC PCI configuration space, thus allowing for convenient single point access mechanism for chipset feature detection. This set of registers consists of: * Capability ID (FDCAP) * Capability Length (FDLEN) * Capability Version * Vendor-Specific Capability ID (FDVER) Feature Vector (FVECT) October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 227 5.0 Register and Memory Mappings The system contains registers that are located in the processor's I/O space and memory space and sets of PCI configuration registers that are located in PCI configuration space. This chapter describes the I/O and memory maps at the registerset level. Register access is also described. Register-level address maps and individual register bit descriptions are provided in the following chapters. 5.1 I/O Map The I/O map is divided into separate types. Fixed ranges cannot be moved, but in some cases can be disabled. Variable ranges can be moved and can also be disabled. 5.1.1 Fixed I/O Address Ranges Table 5-1 shows the fixed I/O decode ranges from the CPU perspective. For each I/O range, there may be separate behavior for reads and writes. DMI cycles that go to target ranges that are marked as Reserved will not be decoded, and will be passed to the subtractive decode agent. Address ranges that are not listed or marked RESERVED are NOT positively decoded by (unless assigned to one of the variable ranges). In subtractive mode, I/O ranges that are not otherwise decoded will be forwarded to PCIe. Table 5-1. Fixed I/O Ranges Decoded (Sheet 1 of 3) I/O Address Read Target 00h - 08h DMA Controller 09h - 0Eh RESERVED 0Fh 10h - 18h Write Target Separate Enable/ Disable Internal Unit DMA Controller DMA None DMA Controller DMA Controller DMA None DMA Controller DMA Controller DMA None 19h - 1Eh RESERVED 1Fh DMA Controller DMA Controller DMA None 20h - 21h Interrupt Controller Interrupt Controller Interrupt None 24h - 25h Interrupt Controller Interrupt Controller Interrupt None 28h - 29h Interrupt Controller Interrupt Controller Interrupt None 2Ch - 2Dh Interrupt Controller Interrupt Controller Interrupt None 2E-2F LPC SIO LPC SIO Forwarded to LPC Yes 30h - 31h Interrupt Controller Interrupt Controller Interrupt None 34h - 35h Interrupt Controller Interrupt Controller Interrupt None 38h - 39h Interrupt Controller Interrupt Controller Interrupt None 3Ch - 3Dh Interrupt Controller Interrupt Controller Interrupt None Intel(R) Communications Chipset 89xx Series - Datasheet 228 October 2012 Order Number: 327879-001US 5.0 Table 5-1. Fixed I/O Ranges Decoded (Sheet 2 of 3) I/O Address Read Target 40h - 42h Timer/Counter 43h RESERVED 4E-4F Write Target Separate Enable/ Disable Internal Unit Timer/Counter PIT (8254) None LPC SIO LPC SIO Forwarded to LPC Yes 50h - 52h Timer/Counter Timer/Counter PIT None 53h RESERVED 60h Micocontroller Microcontroller Forwarded to LPC Yes w/ 64h 61h NMI Controller NMI Controller CPU I/F None Yes w/ 66h 62h Microcontroller Microcontroller Forwarded to LPC 63h NMI Controller1 NMI Controller1 CPU I/F Yes, alias to 61h Yes w/ 60h 64h Micocontroller Microcontroller Forwarded to LPC 65h NMI Controller1 NMI Controller1 CPU I/F Yes, alias to 61h Yes w/ 62h Yes, alias to 61h 66h Microcontroller Microcontroller Forwarded to LPC 67h NMI Controller1 NMI Controller1 CPU I/F 70h RESERVED3 71h RTC Controller RTC Controller RTC None 72h RTC Controller NMI and RTC Controller RTC Yes, w/ 73h 73h RTC Controller RTC Controller RTC Yes, w/ 72h 74h RTC Controller NMI and RTC Controller RTC None 75h RTC Controller RTC Controller RTC None 76h RTC Controller NMI and RTC Controller RTC None 77h RTC Controller RTC Controller RTC None 80h DMA Controller, or LPC, or PCI DMA Controller, or LPC, or PCI DMA None 81h - 83h DMA Controller DMA Controller DMA None 84h - 86h DMA Controller DMA Controller and LPC or PCI DMA None 87h DMA Controller DMA Controller DMA None 88h DMA Controller DMA Controller and LPC or PCI DMA None 89h - 8Bh DMA Controller DMA Controller DMA None 8Ch - 8Eh DMA Controller DMA Controller and LPC or PCI DMA None 8Fh DMA Controller DMA Controller DMA None 90h - 91h DMA Controller DMA Controller DMA Yes, alias to 8xh 92h Reset Generator Reset Generator CPU I/F None 93h - 9Fh DMA Controller DMA Controller DMA Yes, alias to 8xh A0h - A1h Interrupt Controller Interrupt Controller Interrupt None A4h - A5h Interrupt Controller Interrupt Controller Interrupt None A8h - A9h Interrupt Controller Interrupt Controller Interrupt None ACh - ADh Interrupt Controller Interrupt Controller Interrupt None October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 229 Table 5-1. Fixed I/O Ranges Decoded (Sheet 3 of 3) I/O Address B0h - B1h Read Target Interrupt Controller Write Target Separate Enable/ Disable Internal Unit Interrupt Controller Interrupt None None B2h - B3h Power Management Power Management Power Management B4h - B5h Interrupt Controller Interrupt Controller Interrupt None B8h - B9h Interrupt Controller Interrupt Controller Interrupt None BCh - BDh Interrupt Controller Interrupt Controller Interrupt None C0h - D1h DMA Controller DMA Controller DMA None D2h - DDh RESERVED DEh - DFh DMA Controller DMA Controller DMA None F0h FERR#/IGNNE# / Interrupt Cont. FERR#/IGNNE# / Interrupt Cont. CPU I/F None 170h - 177h Reserved 1F0h - 1F7h Reserved 200-207h Gameport Low Gameport Low Forwarded to LPC Yes 208-20Fh Gameport High Gameport High Forwarded to LPC Yes 376h Reserved 3F6h Reserved 4D0h - 4D1h Interrupt Controller Interrupt Controller Interrupt None CF9h Reset Generator Reset Generator CPU I/F None 1. 5.1.2 Only if the Port 61 Alias Enable bit (GCS.P61AE) bit is set. Variable I/O Decode Ranges Table 5-2 shows the variable I/O Decode ranges. They are set using Base Address Registers (BARs) or other config bits in the various configuration spaces. The PnP software (ACPI) can use their configuration mechanisms to set and adjust these values. Warning: The variable I/O ranges should not be set to conflict with the fixed I/O ranges. Unpredictable results if the configuration software allows conflicts to occur. Table 5-2. Variable I/O Decode Ranges Range Name Mappable Size (Bytes) ACPI Anywhere in 64K I/O Space 64 IDE Bus Master Anywhere in 64K I/O Space 1. 16 or 32 Bytes 2. 16 Bytes Target Power Management 1. IDE (SATA Host Controller #1, #2) SMBus Anywhere in 64K I/O Space 32 SMBus Unit TCO 96 bytes above ACPI base 32 TCO Unit GPIO Anywhere in 64K I/O space 128 GPIO Unit Parallel Port 3 ranges in 64K I/O Space 8 (See Note 3) LPC Peripheral Serial Port 1 8 Ranges in 64K I/O Space 8 LPC Peripheral Serial Port 2 8 Ranges in 64K I/O Space 8 LPC Peripheral Intel(R) Communications Chipset 89xx Series - Datasheet 230 October 2012 Order Number: 327879-001US 5.0 Table 5-2. Variable I/O Decode Ranges Range Name Mappable Size (Bytes) Target Floppy Disk Controller 2 Ranges in 64K I/O Space 8 LPC Peripheral LPC Generic 1 Anywhere in 64K I/O Space 4 to 256 Bytes LPC LPC Generic 2 Anywhere in 64K I/O Space 4 to 256 Bytes LPC LPC Generic 3 Anywhere in 64K I/O Space 4 to 256 Bytes LPC LPC Generic 4 Anywhere in 64K I/O Space 4 to 256 Bytes LPC IO Trapping Ranges Anywhere in 64K I/O Space 1 to 256 Bytes Trap on Backbone Native IDE Cmd Anywhere in 64K I/O Space1 8 Native IDE Ctrl Anywhere in 64K I/O Space1 4 Serial ATA Index/Data Pair Anywhere in 64K I/O Space 16 SATA Host Controller #1, #2 PCI-Express Root Ports Anywhere in 64K I/O Space I/O Base/Limit PCIE Root Ports Keyboard and Text (KT) Anywhere in 64K I/O Space 8 Keyboard and Text IDE (SATA Host Controller #2) IDE (SATA Host Controller #2) Notes: 1. All ranges are decoded directly from DMI. 2. There is also an alias 400h above the parallel port range that is used for ECP parallel ports. 5.2 Memory Map Table 5-3 shows (from the CPU perspective) the memory ranges that will decode. Cycles that arrive from DMI that are not directed to any of the internal memory targets that decode directly from DMI (see Table 5-3) will be passed to the subtractive decode agent. Table 5-3. Memory Decode Ranges (From CPU Perspective) (Sheet 1 of 2) Memory Range Target Dependency/Comments FECXX000 - FECXX040 I/O(x)APIC X controlled via APIC Range Select (ASEL) field and APIC Enable (AEN) bit. FEC10000 - FEC17FFF PCIe port 1 PCIe root port 1 I/OxApic Enable (PAE) set FEC18000 - FEC1FFFF PCIe port 2 PCIe root port 2 I/OxApic Enable (PAE) set FEC20000 - FEC27FFF PCIe port 3 PCIe root port 3 I/OxApic Enable (PAE) set FEC28000 - FEC2FFFF PCIe port 4 PCIe root port 4 I/OxApic Enable (PAE) set 1 KB anywhere in 4GB range USB2 Host Controller Enable via standard PCI mechanism (Device 29, Function 0) HPET BIOS determines "fixed" location which is one of four 1KB ranges where X (in the first column) is 0h, 1h, 2h, or 3h. LPC LPC Generic Memory Range. FED0 X000h-FED0 X3FFh FED4_0000h - FED4_BFFFh October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 231 Table 5-3. Memory Decode Ranges (From CPU Perspective) (Sheet 2 of 2) Memory Range Target Dependency/Comments 64KB anywhere in 4GB range LPC LPC Generic Memory Range. Enable via setting bit[0] of the LPC Generic Memory Range register (B0:D31:F0:offset 98h). 32 Bytes anywhere in 64-bit address range SMBus Enable via standard PCI mechanism (Device 31: Function 3) 2KB anywhere above 64KB to 4GB range SATA Host Controller #1 AHCI memory-mapped registers. Enable via standard PCI mechanism (Device 31: Function 2) Memory Base/Limit anywhere in 4GB range PCI-Express Root Ports 1-4 Enable via standard PCI mechanism (Device 28: Function 03) Prefetchable Memory Base/ Limit anywhere in 64-bit address range PCI-Express Root Ports 1-4 Enable via standard PCI mechanism (Device 28: Function 03) 4KB anywhere in 4GB range Thermal Reporting Enable via standard PCI mechanism (Device 31: Function 6) 16KB anywhere in 4GB range Root Complex Register Block (RCRB) Enable via setting bit[0] of the Root Complex Base Address register (B0:D31:F0:offset F0h). 5.3 Boot-Block Update Scheme The system supports a "Top-Block Swap" mode that has the system swap the top block in the SPI (the boot block) with another location. This allows for safe update of the Boot Block (even if a power failure occurs). When the "top-swap" enable bit is set, the System will invert A16 for cycles going to the upper two blocks in the SPI. Specifically, in this mode accesses to FFFF_0000h-FFFF_FFFFh are directed to FFFE_0000hFFFE_FFFFh and vice versa. When the Top Swap Enable bit is 0, the System will not invert A16. This bit is automatically set to 0 by RTEST#, but not by PLTRST#. The scheme is based on the concept that the top block is reserved as the "boot" block, and the block immediately below the top block is reserved for doing boot-block updates. The algorithm is: 1. Software copies the top block to the block immediately below the top 2. Software checks that the copied block is correct. This could be done by performing a checksum calculation. 3. Software sets the "Top-Block Swap" bit. This will invert A16 for cycles going to the SPI. 4. Software erases the top block 5. Software writes the new top block 6. Software checks the new top block 7. Software clears the top-block swap bit 8. Software sets the Top_Swap Lock-Down bit If a power failure occurs at any point after step 3, the system will be able to boot from the copy of the boot block that is stored in the block below the top. This is because the top-swap bit is backed in the RTC well. There is one remaining unusual case that could occur if the RTC battery is not sufficiently high to maintain the RTC well. To avoid the potentially fatal case (where the Top-Swap bit is NOT set, but the top block is not valid), a pinstrap will allow forcing the top-swap bit to be set. This would be a last resort to allow the user to get the system to boot (and avoid having to de-solder the SPI). Intel(R) Communications Chipset 89xx Series - Datasheet 232 October 2012 Order Number: 327879-001US 5.0 See the strapping section above for the signal used as the emergency strap. When the top-swap strap is used, the top-swap bit will be forced to 1 (cannot be cleared by software). The algorithm to put in the BIOS spec is as follows: 1. If an RTC well power failure is experienced during a boot block update, the system will probably not be able to boot at that point. 2. The user can set the Top-Swap strap and force the system to boot from the second block. The code in the second block should read the valid BIOS image from disk (probably a floppy or CD-ROM) and put it into the top-swap. 3. The BIOS will not be able to clear the Top-Swap bit because the jumper is in place. The user should then remove the jumper and reboot. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 233 6.0 LPC Interface Bridge Registers (B0:D31:F0) 6.1 Overview The LPC bridge function of the PCH resides in PCI Bus 0:Device 31:Function 0. This function contains many other functional units, such as DMA and Interrupt controllers, Timers, Power Management, System Management, GPIO, RTC, and LPC Configuration Registers. Registers and functions associated with other functional units (EHCI,... etc.) are described in their respective sections. 6.1.1 PCI Configuration Registers (LPC I/F--B0:D31:F0) Note: Address locations that are not shown should be treated as Reserved. Table 6-1. PCI Configuration Registers (LPC I/F--B0:D31:F0) (Sheet 1 of 2) Offset Start 00h Offset End Default Value Register ID - Description 01h "Offset 00h: ID: Vendor Identification Register" on page 236 8086h 02h 03h "Offset 02h: DID--Device Identification Register (LPC I/F--B0:D31:F0)" on page 236 See register description 04h 05h "Offset 04h: ID--PCICMD--PCI COMMAND Register (LPC I/F--B0:D31:F0)" on page 237 0007h 06h 07h "Offset 06h: ID--PCISTS--PCI Status Register (LPC I/F--B0:D31:F0)" on page 238 0210h 08h 08h "Offset 08h: RID--Revision Identification Register (PCI-PCI--B0:D31:F0)" on page 239 See register description 09h 09h "Offset 09h: PI--Programming Interface Register (PCI-PCI--B0:D31:F0)" on page 239 00h 0Ah 09h "Offset 0Ah: SCC--Sub Class Code Register (PCI-PCI--B0:D31:F0)" on page 239 01h 0Bh 0Bh "Offset 0Bh: BCC--Base Class Code Register (LPC I/F--B0:D31:F0)" on page 240 06h 0Dh 0Dh "Offset 0Dh: PLT--Primary Latency Timer Register (LPC I/F--B0:D31:F0)" on page 240 00h 0Eh 0Eh "Offset 0Eh: HEADTYP--Header Type Register (LPC I/F--B0:D31:F0)" on page 240 80h 2Ch 2Fh "Offset 2Ch: SS--Sub System Identifiers Register (LPC I/F--B0:D31:F0)" on page 241 00000000h 40h 43h "Offset 40h: PMBASE--ACPI Base Address Register (LPC I/F--B0:D31:F0)" on page 241 00000001h 44h 44h "Offset 44h: ACPI_CNTL--ACPI Control Register (LPC I/F--B0:D31:F0)" on page 242 00h 48h 4Bh "Offset 48h: GPIOBASE--GPIO Base Address Register (LPC I/F--B0:D31:F0)" on page 243 00000001h 4Ch 4Ch "Offset 4Ch: GC--GPIO Control Register (LPC I/F--B0:D31:F0)" on page 244 00h 63h "Offset 60h: PIRQ[n]_ROUT--PIRQ[A,B,C,D] Routing Control Register (LPC I/F-- B0:D31:F0)" on page 245 80h 60h Intel(R) Communications Chipset 89xx Series - Datasheet 234 October 2012 Order Number: 327879-001US 6.0 Table 6-1. Offset Start PCI Configuration Registers (LPC I/F--B0:D31:F0) (Sheet 2 of 2) Offset End Register ID - Description Default Value 64h 64h "Offset 64h: SIRQ_CNTL--Serial IRQ Control Register (LPC I/F--B0:D31:F0)" on page 246 10h 68h 6Bh "Offset 68h: PIRQ[n]_ROUT--PIRQ[E,F,G,H] Routing Control Register (LPC I/F-- B0:D31:F0)" on page 247 80h 6Ch 6Dh "Offset 6Ch: LPC_IBDF--IOxAPIC Bus:Device:Function (LPC I/F--B0:D31:F0)" on page 248 00F8h 70h 7Fh "Offset 70h: LPC_HnBDF - HPET n Bus:Device:Function (LPC I/F--B0:D31:F0)" on 00F8h page 249 80h 80h "Offset 80h: LPC_I/O_DEC--I/O Decode Ranges Register (LPC I/F--B0:D31:F0)" on page 250 0000h 82h 83h "Offset 82h: LPC_EN--LPC I/F Enables Register (LPC I/F--B0:D31:F0)" on page 251 0000h 84h 87h "Offset 84h: GEN1_DEC--LPC I/F Generic Decode Range 1 Register (LPC I/F-- B0:D31:F0)" on page 253 00000000h 88h 8Bh "Offset 88h: GEN2_DEC--LPC I/F Generic Decode Range 2 Register (LPC I/F-- B0:D31:F0)" on page 254 00000000h 8Ch 8Eh "Offset 8Ch: GEN3_DEC--LPC I/F Generic Decode Range 3 Register (LPC I/F-- B0:D31:F0)" on page 255 00000000h 90h 93h "Offset 90h: GEN4_DEC--LPC I/F Generic Decode Range 4 Register (LPC I/F-- B0:D31:F0)" on page 256 00000000h 94h 97h "Offset 94h: ULKMC - USB Legacy Keyboard / Mouse Control (LPC I/F-- B0:D31:F0)" on page 257 00002000h 98h 9Bh "Offset 98h: LGMR -- LPC I/F Generic Memory Range (LPC I/F--B0:D31:F0)" on page 259 00000000h DCh DCh "Offset DCh: BIOS_CNTL--BIOS Control Register (LPC I/F--B0:D31:F0)" on page 260 20h E0h E1h "Offset E0h: FDCAP--Feature Detection Capability ID (LPC I/F--B0:D31:F0)" on page 261 0009h E2h E2h "Offset E2h: FDLEN--Feature Detection Capability Length (LPC I/F--B0:D31:F0)" on page 261 0Ch E3h E3h "Offset E3h: FDVER--Feature Detection Version (LPC I/F--B0:D31:F0)" on page 262 10h E4h EBh "Offset E4h: FDVCT--Feature Vector (LPC I/F--B0:D31:F0)" on page 262 See register description F0h F3h "Offset F0h: RCBA--Root Complex Base Address Register (LPC I/F--B0:D31:F0)" on page 263 00000000h A0h A0h "Offset A0h: GEN_PMCON_1--General PM Configuration 1 Register (PM-- B0:D31:F0)" on page 306 0000h A2h A2h "Offset A2h: GEN_PMCON_2--General PM Configuration 2 Register (PM-- B0:D31:F0)" on page 308 00h A4h A4h "Offset A4h: GEN_PMCON_3--General PM Configuration 3 Register (PM-- B0:D31:F0)" on page 310 00h A6h A6h "Offset A6h: GEN_PMCON_LOCK- General Power Management Configuration Lock Register" on page 313 00h A9h A9h "Offset A9h: Chipset Initialization Register 4 (PM--B0:D31:F0)" on page 313 01h ABh ABh "Offset ABh: BM_BREAK_EN Register (PM--B0:D31:F0)" on page 314 01h ACh ACh "Offset ACh: PMIR--Power Management Initialization Register (PM--B0:D31:F0)" on page 314 00000000h B8h BBh "Offset B8h: GPIO_ROUT--GPIO Routing Control Register (PM--B0:D31:F0)" on page 315 00000000h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 235 . 6.1.1.1 Offset 00h: VID--Vendor Identification Register (LPC I/F--B0:D31:F0) Table 6-2. Offset 00h: ID: Vendor Identification Register Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 8086h Bit Range Bit Acronym 15 :00 VID Offset Start: 00h Offset End: 01h Power Well: Core Bit Description Sticky Vendor Identification -- This 16-bit value is assigned to Intel. Intel VID = 8086h Bit Reset Value Bit Access 8086h RO 6.1.1.2 Offset 02h: DID--Device Identification Register (LPC I/F--B0:D31:F0) Table 6-3. Offset 02h: DID--Device Identification Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 16 bit Bus:Device:Function: B0:D31 :F0 BAR: Configuration Default: See register description Bit Range Bit Acronym 15 :00 DID Bit Description Device Identification -- This is a 16-bit value assigned to the PCH's LPC bridge PCI device Intel(R) Communications Chipset 89xx Series - Datasheet 236 Offset Start: 02h Offset End: 03h Power Well: Core Sticky Bit Reset Value Bit Access RO October 2012 Order Number: 327879-001US 6.0 6.1.1.3 Table 6-4. Offset 04h: ID--PCICMD--PCI COMMAND Register (LPC I/F-- B0:D31:F0) Offset 04h: ID--PCICMD--PCI COMMAND Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 0007h Bit Range Bit Acronym 15 :10 Reserved Offset Start: 04h Offset End: 05h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00h RO 0h RO 09 FBE Fast Back to Back Enable -- Hardwired to `0' as per PCI Express* Specification. 08 SEE SERR# Enable: 0 = LPC bridge does not generates SERR# 1 = LPC bridge generates SERR# 0h RW 07 WCC Wait Cycle Control - Hardwired to `0' as per PCI Express* Specification. 0h RO 06 PERE Parity Error Response Enable: 0 = No action is taken when detecting a parity error. 1 = Enables the LPC bridge to respond to parity errors detected on ILB interface. 0h RW 05 VGA_PSE VGA Palette Snoop -- Hardwired to `0' as per PCI Express* Specification. 0h RO 04 MWIE Memory Write and Invalidate Enable -- Hardwired to `0'as per PCI Express* Specification. 0h RO 03 SCE Special Cycle Enable -- Hardwired to `0' as per PCI Express* Specification. 0h RO 02 BME Bus Master Enable -- Bus Masters cannot be disabled. 1 RO 01 MSE Memory Space Enable -- Memory space cannot be disabled on LPC. 1 RO 00 IOSE I/O Space Enable -- I/O space cannot be disabled on LPC. 1 RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 237 6.1.1.4 Offset 06h: ID--PCISTS--PCI Status Register (LPC I/F--B0:D31:F0) Note: For the writable bits, software must write a 1 to clear bits that are set. Writing a 0 to the bit has no effect. Table 6-5. Offset 06h: ID--PCISTS--PCI Status Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: Default: 0210h Offset Start: 06h Offset End: 07h Power Well: Bit Reset Value Bit Access DPE Detected Parity Error: 0 = Parity error not detected. 1 = Indicates that the PCH detected a parity error on the internal backbone. This bit gets set even if the Parity Error Response bit (D30:F0:04 bit 6) is not set. 0h RWC 14 SSE Signaled System Error -- Set when the LPC bridge signals a system error to the internal SERR# logic. 0h RWC 13 RMA Received Master Abort: 0 = No master abort received. 1 = Set when the bridge receives a master abort status from the I/O data bus. 0h RWC 12 RTA Received Target Abort: 0 = No target abort received. 1 = Set when the bridge receives a target abort status from the I/O data bus. 0h RWC 11 STA Signaled Target Abort: 0 = No signaled target abort. 1 - Set when the bridge generates a completion packet with target abort status on the I/O data bus. 0h RWC Bit Range Bit Acronym Bit Description 15 10 :09 08 07 :05 Reserved DPD Reserved 04 CLIST 03 IS 02 :00 Reserved Reserved Sticky 00h RO Data Parity Error Detected: 0 = Data parity error not detected. 1 = Set when the bridge receives a completion packet from the I/O data bus from a previous request, and detects a parity error, and CMD.PER is set (D30, F0, 04, bit 6). 0h RWC Reserved 0h RO Capabilities List -- Capability list exist on the PCI bridge. 1h RO Interrupt Status -- The PCI bridge does not generate interrupts. 0h RO Reserved 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 238 B0:D31 :F0 October 2012 Order Number: 327879-001US 6.0 6.1.1.5 Offset 08h: RID--Revision Identification Register (LPC I/F-- Table 6-6. Offset 08h: RID--Revision Identification Register (PCI-PCI--B0:D31:F0) B0:D31:F0) Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: See register description Bit Range Bit Acronym 07 :00 RID Offset Start: 08h Offset End: 08h Power Well: Bit Description Sticky Revision ID: Indicates stepping of the host controller hardware Bit Reset Value Bit Access Variable RO 6.1.1.6 Offset 09h: PI--Programming Interface Register (PCI-PCI-- B0:D31:F0) Table 6-7. Offset 09h: PI--Programming Interface Register (PCI-PCI--B0:D31:F0) Description: View: PCI Size: 8 bit Bus:Device:Function: B0:D31 :F0 BAR: Configuration Default: 00h Bit Range Bit Acronym 07 :00 PI Offset Start: 09h Offset End: 09h Power Well: Bit Description Sticky Bit Reset Value Bit Access Programming Interface RO 6.1.1.7 Offset 0Ah: SCC--Sub Class Code Register (LPC I/F--B0:D31:F0) Table 6-8. Offset 0Ah: SCC--Sub Class Code Register (PCI-PCI--B0:D31:F0) Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 01h Bit Range Bit Acronym 07 :00 SCC October 2012 Order Number: 327879-001US Offset Start: 0Ah Offset End: 09h Power Well: Bit Description Sticky Sub Class Code -- 8-bit value that indicates the category of bridge for the LPC bridge. 01h = PCI-to-ISA bridge. Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 239 6.1.1.8 Offset 0Bh: BCC--Base Class Code Register (LPC I/F--B0:D31:F0) Table 6-9. Offset 0Bh: BCC--Base Class Code Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 8 bit BAR: Configuration B0:D31 :F0 Default: 06h Bit Range Bit Acronym 07 :00 BCC 6.1.1.9 Bus:Device:Function: Offset Start: 0Bh Offset End: 0Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Base Class Code -- 8-bit value that indicates the type of device for the LPC bridge. 06h = Bridge device. RO Offset 0Dh: PLT--Primary Latency Timer Register (LPC I/F-- B0:D31:F0) Table 6-10. Offset 0Dh: PLT--Primary Latency Timer Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 8 bit Default: 00h Bit Range Bit Acronym 07 :03 MLC 02 :00 Reserved 6.1.1.10 Bus:Device:Function: B0:D31 :F0 BAR: Configuration Offset Start: 0Dh Offset End: 0Dh Power Well: Bit Description Sticky Bit Reset Value Bit Access Master Latency Count -- Reserved RO Reserved RO Offset 0Eh: HEADTYP--Header Type Register (LPC I/F--B0:D31:F0) Table 6-11. Offset 0Eh: HEADTYP--Header Type Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 8 bit Bit Range Default: 80h Bit Acronym HEADTYP Bit Description Header Type -- This 7-bit field identifies the header layout of the configuration space. Intel(R) Communications Chipset 89xx Series - Datasheet 240 Offset Start: 0Eh Offset End: 0Eh Power Well: Multi-Function Device -- This bit is 1 to indicate a multifunction device. 07 06 :00 Bus:Device:Function: B0:D31 :F0 BAR: Configuration Sticky Bit Reset Value Bit Access 1b RO 00h RO October 2012 Order Number: 327879-001US 6.0 6.1.1.11 Offset 2Ch: SS--Sub System Identifiers Register (LPC I/F--B0:D31:F0) This register is initialized to logic 0 by the assertion of PLTRST#. This register can be written only once after PLTRST# de-assertion. Table 6-12. Offset 2Ch: SS--Sub System Identifiers Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 32 bit BAR: Configuration B0:D31 :F0 Default: 00000000h Bit Range Bit Acronym 31 :16 SSID 15 :00 SSVID 6.1.1.12 Bus:Device:Function: Offset Start: 2Ch Offset End: 2Fh Power Well: Bit Reset Value Bit Access Subsystem ID -- This is written by BIOS. No hardware action taken on this value. 0000h RWO Subsystem Vendor ID -- This is written by BIOS. No hardware action taken on this value. 0000h RWO Bit Description Sticky Offset 40h: PMBASE--ACPI Base Address Register (LPC I/F-- B0:D31:F0) Sets base address for ACPI I/O registers, GPIO registers and TCO I/O registers. These registers can be mapped anywhere in the 64-K I/O space on 128-byte boundaries. Table 6-13. Offset 40h: PMBASE--ACPI Base Address Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000001h Bit Range Bit Acronym 31 :16 Reserved 15 :07 BA 06 :01 Reserved 00 Bus:Device:Function: B0:D31 :F0 RTE October 2012 Order Number: 327879-001US Offset Start: 40h Offset End: 43h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved 0000h RO Base Address -- This field provides 128 bytes of I/O space for ACPI, GPIO, and TCO logic. This is placed on a 128byte boundary. 0000h RW Reserved 0000h RO 1b RO Resource Type Indicator -- Hardwired to 1 to indicate I/O space. Intel(R) Communications Chipset 89xx Series - Datasheet 241 6.1.1.13 Offset 44h: ACPI_CNTL--ACPI Control Register (LPC I/F--B0:D31:F0) Table 6-14. Offset 44h: ACPI_CNTL--ACPI Control Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 8 bit Bit Range 07 06 :03 BAR: Configuration Bus:Device:Function: B0:D31 Offset Start: 44h :F0 Offset End: 44h Default: 00h Power Well: Core Bit Reset Value Bit Access ACPI_EN ACPI Enable: 0 =Disable. 1 = Decode of the I/O range pointed to by the ACPI base register is enabled, and the ACPI power management function is enabled. NOTE: The APM power management ranges (B2/B3h) are always enabled and are not affected by this bit. 0b RW Reserved Reserved 0h Bit Acronym Bit Description Sticky SCI IRQ Select: Specifies on which IRQ the SCI will internally appear. If not using the APIC, the SCI must be routed to IRQ9-11, and that interrupt is not sharable with the SERIRQ stream, but is shareable with other PCI interrupts. If using the APIC, the SCI can also be mapped to IRQ20-23, and can be shared with other interrupts. Bits 02 :00 SCI_IRQ_SEL SCI Map 000b IRQ9 001b IRQ10 010b IRQ11 011b Reserved 100b IRQ20 (Only available if APIC enabled) 101b IRQ21 (Only available if APIC enabled) 110b IRQ22 (Only available if APIC enabled) 111b IRQ23 (Only available if APIC enabled) 00h RW When the interrupt is mapped to APIC interrupts 9, 10 or 11, the APIC should be programmed for active-high reception. When the interrupt is mapped to APIC interrupts 20 through 23, the APIC should be programmed for active-low reception. Intel(R) Communications Chipset 89xx Series - Datasheet 242 October 2012 Order Number: 327879-001US 6.0 6.1.1.14 Offset 48h: GPIOBASE--GPIO Base Address Register (LPC I/F-- B0:D31:F0) Table 6-15. Offset 48h: GPIOBASE--GPIO Base Address Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 00000001h Bit Range Bit Acronym 31 :16 Reserved 15 :07 BA 06 :01 Reserved 00 October 2012 Order Number: 327879-001US Offset Start: 48h Offset End: 4Bh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved. Always 0. Base Address -- Provides the 128 bytes of I/O space for GPIO. RW Reserved. Always 0. Hardwired to 1 to indicate I/O space. Intel(R) Communications Chipset 89xx Series - Datasheet 243 6.1.1.15 Offset 4Ch: GC--GPIO Control Register (LPC I/F--B0:D31:F0) Table 6-16. Offset 4Ch: GC--GPIO Control Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 8 bit BAR: Configuration Bit Acronym 07 :05 Reserved 03 :01 B0:D31 :F0 Default: 00h Bit Range 04 Bus:Device:Function: EN Reserved Offset Start: 4Ch Offset End: 4Ch Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved GPIO Enable -- This bit enables/disables decode of the I/O range pointed to by the GPIO Base Address register (B0:D31:F0:48h) and enables the GPIO function. 0 = Disable. 1 = Enable. RW Reserved GPIO Lockdown Enable -- This bit enables lockdown of the following GPIO registers: 00 GLE Offset Offset Offset Offset Offset Offset Offset Offset Offset Offset 00h: GPIO_USE_SEL 04h: GP_IO_SEL 0Ch: GP_LVL 30h: GPIO_USE_SEL2 34h: GP_IO_SEL2 38h: GP_LVL2 40h: GPIO_USE_SEL3 44h: GP_IO_SEL3 48h: GP_LVL3 60h: GP_RST_SEL RW 0 = Disable. 1 = Enable. When this bit is written from a `1' to a `0' an SMI# is generated if enabled. This ensures that only SMM code can change the above GPIO registers after they are locked down. Intel(R) Communications Chipset 89xx Series - Datasheet 244 October 2012 Order Number: 327879-001US 6.0 6.1.1.16 Offset 60h: PIRQ[n]_ROUT--PIRQ[A,B,C,D] Routing Control Register (LPC I/F--B0:D31:F0) Offset Address: PIRQA - 60h PIRQB - 61h PIRQC - 62h PIRQC - 63h Table 6-17. Offset 60h: PIRQ[n]_ROUT--PIRQ[A,B,C,D] Routing Control Register (LPC I/ F--B0:D31:F0) Description: View: PCI Size: 8 bit Bit Range 07 Bus:Device:Function: B0:D31 :F0 BAR: Configuration Default: 80h Power Well: Core Bit Acronym IRQEN Offset Start: 60h Offset End: 63h Bit Description Sticky Interrupt Routing Enable 0 = The corresponding PIRQ is routed to one of the ISAcompatible interrupts specified in bits[3:0]. 1 = The PIRQ is not routed to the 8259. Bit Reset Value Bit Access RW BIOS must program this bit to 0 during POST for any of the PIRQs that are being used. The value of this bit may subsequently be changed by the OS when setting up for I/O APIC interrupt delivery mode. 06 :04 Reserved Reserved IRQ Routing -- (ISA compatible) 03 :00 October 2012 Order Number: 327879-001US Value IRQ Value IRQ 0000b Reserved 1000b Reserved 0001b Reserved 1001b IRQ9 0010b Reserved 1010b IRQ10 0011b IRQ3 1011b IRQ11 0100b IRQ4 1100b IRQ12 0101b IRQ5 1101b Reserved 0110b IRQ6 1110b IRQ14 0111b IRQ7 1111b IRQ15 RW Intel(R) Communications Chipset 89xx Series - Datasheet 245 6.1.1.17 Offset 64h: SIRQ_CNTL--Serial IRQ Control Register (LPC I/F-- B0:D31:F0) Table 6-18. Offset 64h: SIRQ_CNTL--Serial IRQ Control Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 8 bit Bit Range 07 06 BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 10h Power Well: Core Bit Acronym Bit Description SIRQEN Serial IRQ Enable: The buffer is input only and internally SERIRQ will be a 1. 1 = Serial IRQs will be recognized. The SERIRQ pin will be configured as SERIRQ. SIRQMD Offset Start: 64h Offset End: 64h Sticky Bit Reset Value Bit Access RW Serial IRQ Mode Select: The serial IRQ machine will be in quiet mode when this bit is 0. The serial IRQ machine will be in continuous mode when this bit is 1. RW For systems using Quiet Mode, this bit should be set to 1 (Continuous Mode) for at least one frame after coming out of reset before switching back to Quiet Mode. Failure to do so results in the PCH not recognizing SERIRQ interrupts. 05 :02 01 :00 SIRQSZ Serial IRQ Frame Size -- Fixed field that indicates the size of the SERIRQ frame as 21 frames. SFPW Start Frame Pulse Width -- This is the number of PCI clocks that the SERIRQ pin will be driven low by the serial IRQ machine to signal a start frame. In continuous mode, the PCH will drive the start frame for the number of clocks specified. In quiet mode, the PCH will drive the start frame for the number of clocks specified minus one, as the first clock was driven by the peripheral. 00 = 4 clocks 01 = 6 clocks 10 = 8 clocks 11 = Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 246 4h RO RW October 2012 Order Number: 327879-001US 6.0 6.1.1.18 Offset 68h: PIRQ[n]_ROUT--PIRQ[E,F,G,H] Routing Control Register (LPC I/F--B0:D31:F0) Offset Address: PIRQE - 68h PIRQF - 69h PIRQG - 6Ah PIRQH - 6Bh Table 6-19. Offset 68h: PIRQ[n]_ROUT--PIRQ[E,F,G,H] Routing Control Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 8 bit Bit Range 07 06 :04 Bus:Device:Function: B0:D31 :F0 BAR: Configuration Default: 80h Power Well: Core Bit Acronym IRQEN Reserved Offset Start: 68h Offset End: 6Bh Bit Description Sticky Interrupt Routing Enable: The corresponding PIRQ is routed to one of the ISAcompatible interrupts specified in bits[3:0]. The PIRQ is not routed to the 8259. BIOS must program this bit to 0 during POST for any of the PIRQs that are being used. The value of this bit may subsequently be changed by the OS when setting up for I/O APIC interrupt delivery mode. Bit Reset Value Bit Access RW Reserved IRQ Routing -- (ISA compatible.) 03 :00 October 2012 Order Number: 327879-001US Value IRQ Value IRQ 0000b Reserved 1000b Reserved 0001b Reserved 1001b IRQ9 0010b Reserved 1010b IRQ10 0011b IRQ3 1011b IRQ11 0100b IRQ4 1100b IRQ12 0101b IRQ5 1101b Reserved 0110b IRQ6 1110b IRQ14 0111b IRQ7 1111b IRQ15 RW Intel(R) Communications Chipset 89xx Series - Datasheet 247 6.1.1.19 Offset 6Ch: LPC_IBDF--IOxAPIC Bus:Device:Function (LPC I/F-- B0:D31:F0) Table 6-20. Offset 6Ch: LPC_IBDF--IOxAPIC Bus:Device:Function (LPC I/F--B0:D31:F0) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 00F8h Bit Acronym Offset Start: 6Ch Offset End: 6Dh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access IOxAPIC Bus:Device:Function -- This field specifies the bus:device:function that PCH's IOxAPIC will be using for the following: As the Requester ID when initiating Interrupt Messages to the processor. As the Completer ID when responding to the reads targeting the IOxAPIC's Memory-Mapped I/O registers. The 16-bit field comprises the following: 15 :00 IBDF Bits Description 15:8 Bus Number 7:3 Device Number 2:0 Function Number RW This field defaults to Bus 0: Device 31: Function 0 after reset. BIOS can program this field to provide a unique bus:device:function number for the internal IOxAPIC. Intel(R) Communications Chipset 89xx Series - Datasheet 248 October 2012 Order Number: 327879-001US 6.0 6.1.1.20 Offset 70h: LPC_HnBDF - HPET n Bus:Device:Function (LPC I/F-- B0:D31:F0) Offset: H0BDF H1BDF H2BDF H3BDF H4BDF H5BDF H6BDF H7BDF 70h-71h 72h-73h 74h-75h 76h-77h 78h-79h 7Ah-7Bh 7Ch-7Dh 7Eh-7Fh Table 6-21. Offset 70h: LPC_HnBDF - HPET n Bus:Device:Function (LPC I/F--B0:D31:F0) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 00F8h Bit Acronym Offset Start: 70h Offset End: 7Fh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access HPET n Bus:Device:Function -- This field specifies the bus:device:function that the PCH's HPET n will be using in the following: -As the Requester ID when initiating Interrupt Messages to the CPU -As the Completer ID when responding to the reads targeting the corresponding HPET's MemoryMapped I/O registers. The 16-bit field comprises the following: 15 :00 HnBDF Bits Description 15:08 Bus Number 07:03 Device Number 02:00 Function Number RW This field is default to Bus 0: Device 31: Function 0 after reset. BIOS shall program this field accordingly if unique bus:device:function number is required for the corresponding HPET. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 249 6.1.1.21 Offset 80h: LPC_I/O_DEC--I/O Decode Ranges Register (LPC I/F-- B0:D31:F0) Table 6-22. Offset 80h: LPC_I/O_DEC--I/O Decode Ranges Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 16 bit BAR: Configuration Bit Acronym 15 :13 Reserved Reserved Reserved 02 :00 Reserved Bit Reset Value Bit Access RW Reserved RW Reserved RW Reserved COMA Decode Range -- This field determines which range to decode for the COMA Port. 000 = 3F8h - 3FFh (COM1) 001 = 2F8h - 2FFh (COM2) 010 = 220h - 227h 011 = 228h - 22Fh 100 = 238h - 23Fh 101 = 2E8h - 2EFh (COM4) 110 = 338h - 33Fh 111 = 3E8h - 3EFh (COM3) Intel(R) Communications Chipset 89xx Series - Datasheet 250 Sticky Reserved COMB Decode Range -- This field determines which range to decode for the COMB Port. 000 = 3F8h - 3FFh (COM1) 001 = 2F8h - 2FFh (COM2) 010 = 220h - 227h 011 = 228h - 22Fh 100 = 238h - 23Fh 101 = 2E8h - 2EFh (COM4) 110 = 338h - 33Fh 111 = 3E8h - 3EFh (COM3) 06 :04 03 Bit Description LPT Decode Range -- This field determines which range to decode for the LPT Port. 00 = 378h - 37Fh and 778h - 77Fh 01 = 278h - 27Fh (port 279h is read only) and 678h - 67Fh 10 = 3BCh -3BEh and 7BCh - 7BEh 11 = Reserved 09 :08 Offset Start: 80h Offset End: 80h Power Well: Core FDD Decode Range -- Determines which range to decode for the FDD Port 3F0h - 3F5h, 3F7h (Primary) 1 = 370h - 375h, 377h (Secondary) 12 07 B0:D31 :F0 Default: 0000h Bit Range 11 :10 Bus:Device:Function: RW October 2012 Order Number: 327879-001US 6.0 6.1.1.22 Offset 82h: LPC_EN--LPC I/F Enables Register (LPC I/F--B0:D31:F0) Table 6-23. Offset 82h: LPC_EN--LPC I/F Enables Register (LPC I/F--B0:D31:F0) (Sheet 1 of 2) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 0000h Bit Range Bit Acronym 15 :14 Reserved Offset Start: 82h Offset End: 83h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved CNF2_LPC_EN Microcontroller Enable # 2.: 0 = Disable. 1 = Enables the decoding of the I/O locations 4Eh and 4Fh to the LPC interface. This range is used for a microcontroller. RW CNF1_LPC_EN Super I/O Enable: 0 = Disable. 1 = Enables the decoding of the I/O locations 2Eh and 2Fh to the LPC interface. This range is used for Super I/O devices. RW MC_LPC_EN Microcontroller Enable # 1.: 0 = Disable. 1 = Enables the decoding of the I/O locations 62h and 66h to the LPC interface. This range is used for a microcontroller. RW KBC_LPC_E Keyboard Enable: 0 = Disable. 1 = Enables the decoding of the I/O locations 60h and 64h to the LPC interface. This range is used for a microcontroller. RW 09 High Gameport Enable: 0 = Disable. GAMEH_LPC_EN 1 = Enables the decoding of the I/O locations 208h to 20Fh to the LPC interface. This range is used for a gameport. RW 08 Low Gameport Enable: 0 = Disable. 1 = Enables the decoding of the I/O locations 200h to 207h to the LPC interface. This range is used for a gameport. RW 13 12 11 10 07 :04 03 GAMEL_LPC_E Reserved FDD_LPC_EN October 2012 Order Number: 327879-001US Reserved Floppy Drive Enable 0 = Disable. 1 = Enables the decoding of the FDD range to the LPC interface. This range is selected in the LPC_FDD/LPT Decode Range Register (B0:D31:F0:80h, bit 12). RW Intel(R) Communications Chipset 89xx Series - Datasheet 251 Table 6-23. Offset 82h: LPC_EN--LPC I/F Enables Register (LPC I/F--B0:D31:F0) (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range 02 01 00 BAR: Configuration Bus:Device:Function: Default: 0000h Bit Acronym Offset Start: 82h Offset End: 83h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Parallel Port Enable 0 = Disable. 1 = Enables the decoding of the LPT range to the LPC interface. This range is selected in the LPC_FDD/LPT Decode Range Register (B0:D31:F0:80h, bit 9:8). RW COMB_LPC_EN Comm Port B Enable 0 = Disable. 1 = Enables the decoding of the COMB range to the LPC interface. This range is selected in the LPC_COM Decode Range Register (B0:D31:F0:80h, bits 6:4). RW COMA_LPC_EN Comm Port A Enable 0 = Disable. 1 = Enables the decoding of the COMA range to the LPC interface. This range is selected in the LPC_COM Decode Range Register (B0:D31:F0:80h, bits 3:2). RW LPT_LPC_EN Intel(R) Communications Chipset 89xx Series - Datasheet 252 B0:D31 :F0 October 2012 Order Number: 327879-001US 6.0 6.1.1.23 Offset 84h: GEN1_DEC--LPC I/F Generic Decode Range 1 Register (LPC I/F--B0:D31:F0) Table 6-24. Offset 84h: GEN1_DEC--LPC I/F Generic Decode Range 1 Register (LPC I/F-- B0:D31:F0) Description: View: PCI Size: 32 bit BAR: Configuration B0:D31 :F0 Default: 00000000h Bit Range Bit Acronym 31 :24 Reserved 23 :18 GEN1_MASK 17 :16 Reserved 15 :02 Bus:Device:Function: GEN1_BASE Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved Generic I/O Decode Range Address[7:2] Mask -- A `1' in any bit position indicates that any value in the corresponding address bit in a received cycle will be treated as a match. The corresponding bit in the Address field, below, is ignored. The mask is only provided for the lower 6 bits of the DWord address, allowing for decoding blocks up to 256 bytes in size. RW Reserved Generic I/O Decode Range 1 Base Address -- This address is aligned on a 128-byte boundary, and must have address lines 31:16 as 0. Note: The PCH does not provide decode down to the word or byte level 01 Reserved Reserved 00 GEN1_EN Generic Decode Range 1 Enable 0 = Disable. 1 = Enable the GEN1 I/O range to be forwarded to the LPC I/F October 2012 Order Number: 327879-001US Offset Start: 84h Offset End: 87h RW RW Intel(R) Communications Chipset 89xx Series - Datasheet 253 6.1.1.24 Offset 88h: GEN2_DEC--LPC I/F Generic Decode Range 2 Register (LPC I/F--B0:D31:F0) Table 6-25. Offset 88h: GEN2_DEC--LPC I/F Generic Decode Range 2 Register (LPC I/F-- B0:D31:F0) Description: View: PCI Size: 32 bit BAR: Configuration Bit Acronym 31 :24 Reserved 23 :18 GEN2_MASK 17 :16 Reserved GEN2_BASE Offset Start: 88h Offset End: 8Bh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved Generic I/O Decode Range Address[7:2] Mask -- A `1' in any bit position indicates that any value in the corresponding address bit in a received cycle will be treated as a match. The corresponding bit in the Address field, below, is ignored. The mask is only provided for the lower 6 bits of the DWord address, allowing for decoding blocks up to 256 bytes in size. RW Reserved Generic I/O Decode Range 2 Base Address -- The PCH does not provide decode down to the word or byte level 01 Reserved Reserved 00 GEN2_EN Generic Decode Range 2 Enable 0 = Disable. 1 = Enable the GEN2 I/O range to be forwarded to the LPC I/F Intel(R) Communications Chipset 89xx Series - Datasheet 254 B0:D31 :F0 Default: 00000000h Bit Range 15 :02 Bus:Device:Function: RW RW October 2012 Order Number: 327879-001US 6.0 6.1.1.25 Offset 8Ch: GEN3_DEC--LPC I/F Generic Decode Range 3 Register (LPC I/F--B0:D31:F0) Table 6-26. Offset 8Ch: GEN3_DEC--LPC I/F Generic Decode Range 3 Register (LPC I/F-- B0:D31:F0) Description: View: PCI Size: 32 bit BAR: Configuration B0:D31 :F0 Default: 00000000h Bit Range Bit Acronym 31 :24 Reserved 23 :18 GEN3_MASK 17 :16 Reserved 15 :02 Bus:Device:Function: GEN3_BASE Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved Generic I/O Decode Range Address[7:2] Mask -- A `1' in any bit position indicates that any value in the corresponding address bit in a received cycle will be treated as a match. The corresponding bit in the Address field, below, is ignored. The mask is only provided for the lower 6 bits of the DWord address, allowing for decoding blocks up to 256 bytes in size. RW Reserved Generic I/O Decode Range 3 Base Address -- The PCH does not provide decode down to the word or byte level 01 Reserved Reserved 00 GEN4_EN Generic Decode Range 4 Enable: 0 = Disable. 1 = Enable the GEN3 I/O range to be forwarded to the LPC I/F October 2012 Order Number: 327879-001US Offset Start: 8Ch Offset End: 8Eh RW RW Intel(R) Communications Chipset 89xx Series - Datasheet 255 6.1.1.26 Offset 90h: GEN4_DEC--LPC I/F Generic Decode Range 4 Register (LPC I/F--B0:D31:F0) Table 6-27. Offset 90h: GEN4_DEC--LPC I/F Generic Decode Range 4 Register (LPC I/F-- B0:D31:F0) Description: View: PCI Size: 32 bit BAR: Configuration Bit Acronym 31 :24 Reserved 23 :18 GEN4_MASK 17 :16 Reserved GEN4_BASE Offset Start: 90h Offset End: 93h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved Generic I/O Decode Range Address[7:2] Mask -- A `1' in any bit position indicates that any value in the corresponding address bit in a received cycle will be treated as a match. The corresponding bit in the Address field, below, is ignored. The mask is only provided for the lower 6 bits of the DWord address, allowing for decoding blocks up to 256 bytes in size. RW Reserved Generic I/O Decode Range 4 Base Address -- The PCH does not provide decode down to the word or byte level 01 Reserved Reserved 00 GEN4_EN Generic Decode Range 4 Enable: 0 = Disable. 1 = Enable the GEN3 I/O range to be forwarded to the LPC I/F Intel(R) Communications Chipset 89xx Series - Datasheet 256 B0:D31 :F0 Default: 00000000h Bit Range 15 :02 Bus:Device:Function: RW RW October 2012 Order Number: 327879-001US 6.0 6.1.1.27 Offset 94h: ULKMC - USB Legacy Keyboard / Mouse Control (LPC I/F-- B0:D31:F0) Table 6-28. Offset 94h: ULKMC - USB Legacy Keyboard / Mouse Control (LPC I/F-- B0:D31:F0) (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR: Configuration Bit Acronym 31 :16 Reserved 14 :12 11 10 09 08 07 B0:D31 :F0 Default: 00002000h Bit Range 15 Bus:Device:Function: SMIBYENDPS Reserved Offset Start: 94h Offset End: 97h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved SMI Caused by End of Pass-Through -- This bit indicates if the event occurred. Even if the corresponding enable bit is not set in bit 7, then this bit will still be active. It is up to the SMM code to use the enable bit to determine the exact cause of the SMI#. 0 = Software clears this bit by writing a 1 to the bit location in any of the controllers. 1 = Event Occurred RWC Reserved TRAPBY64W SMI Caused by Port 64 Write -- This bit indicates if the event occurred. Even if the corresponding enable bit is not set in bit 3, this bit will still be active. It is up to the SMM code to use the enable bit to determine the exact cause of the SMI#. The A20Gate Pass-Through Logic allows specific port 64h writes to complete without setting this bit. 0 = Software clears this bit by writing a 1 to the bit location in any of the controllers. 1 = Event Occurred. RWC TRAPBY64R SMI Caused by Port 64 Read -- This bit indicates if the event occurred. Even if the corresponding enable bit is not set in bit 2, this bit will still be active. It is up to the SMM code to use the enable bit to determine the exact cause of the SMI#. 0 = Software clears this bit by writing a 1 to the bit location in any of the controllers. 1 = Event Occurred. RWC TRAPBY60W SMI Caused by Port 60 Write -- This bit indicates if the event occurred. Even if the corresponding enable bit is not set in bit 1, this bit will still be active. It is up to the SMM code to use the enable bit to determine the exact cause of the SMI#. The A20Gate Pass-Through Logic allows specific port 64h writes to complete without setting this bit. 0 = Software clears this bit by writing a 1 to the bit location in any of the controllers. 1 = Event Occurred. RWC TRAPBY60R SMI Caused by Port 60 Read -- This bit indicates if the event occurred. Even if the corresponding enable bit is not set in the bit 0, then this bit will still be active. It is up to the SMM code to use the enable bit to determine the exact cause of the SMI#. 0 = Software clears this bit by writing a 1 to the bit location in any of the controllers. 1 = Event Occurred. RWC SMIATENDPS SMI at End of Pass-Through Enable -- This bit enables SMI at the end of a pass-through. This can occur if an SMI is generated in the middle of a pass-through, and needs to be serviced later. 0 = Disable 1 = Enable RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 257 Table 6-28. Offset 94h: ULKMC - USB Legacy Keyboard / Mouse Control (LPC I/F-- B0:D31:F0) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: Configuration B0:D31 :F0 Default: 00002000h Offset Start: 94h Offset End: 97h Power Well: Core Bit Description PSTATE Pass Through State: 0 = If software needs to reset this bit, it should set bit 5 in all of the host controllers to 0. 1 = Indicates that the state machine is in the middle of an A20GATE pass-through sequence. RO 05 A20PASSEN A20Gate Pass-Through Enable: 0 = Disable. 1 = Enable. Allows A20GATE sequence Pass-Through function. A specific cycle sequence involving writes to port 60h and 64h does not result in the setting of the SMI status bits. RW 04 USBSMIEN SMI on USB IRQ Enable: 0 = Disable 1 = Enable. USB interrupt will cause an SMI event. RW 03 64WEN SMI on Port 64 Writes Enable: 0 = Disable 1 = Enable. A 1 in bit 11 will cause an SMI event. RW 02 64REN SMI on Port 64 Reads Enable: 0 = Disable 1= Enable. A 1 in bit 10 will cause an SMI event. RW 01 60WEN SMI on Port 60 Writes Enable: 0 = Disable 1 = Enable. A 1 in bit 9 will cause an SMI event. RW 00 60REN SMI on Port 60 Reads Enable: 0 = Disable 1 = Enable. A 1 in bit 8 will cause an SMI event. RW Intel(R) Communications Chipset 89xx Series - Datasheet Sticky Bit Reset Value Bit Acronym 06 258 Bus:Device:Function: Bit Access October 2012 Order Number: 327879-001US 6.0 6.1.1.28 Offset 98h: LGMR -- LPC I/F Generic Memory Range (LPC I/F-- B0:D31:F0) Table 6-29. Offset 98h: LGMR -- LPC I/F Generic Memory Range (LPC I/F--B0:D31:F0) Description: View: PCI Size: 32 bit Bit Range BAR: Configuration B0:D31 :F0 Default: 00000000h Bit Acronym Reserved 00 October 2012 Order Number: 327879-001US Offset Start: 98h Offset End: 9Bh Power Well: Bit Description Sticky Memory Address [31:16] -- This field specifies a 64 KB memory block anywhere in the 4 GB memory space that will be decoded to LPC as standard LPC memory cycle if enabled. 31 :16 15 :01 Bus:Device:Function: Bit Reset Value Bit Access RW Reserved LPC Memory Range Decode Enable -- When this bit is set to `1', then the range specified in bits 31:16 of this register is enabled for decoding to LPC RW Intel(R) Communications Chipset 89xx Series - Datasheet 259 6.1.1.29 Offset DCh: BIOS_CNTL--BIOS Control Register (LPC I/F--B0:D31:F0) Table 6-30. Offset DCh: BIOS_CNTL--BIOS Control Register (LPC I/F--B0:D31:F0) Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 20h Bit Range Bit Acronym 07 06 Reserved Offset Start: DCh Offset End: DCh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 05 SMM_BWP SMM BIOS write protect disable -- This bit set defines when the BIOS region can be written by the host. BIOS region SMM protection is disabled. The BIOS Region is writable regardless if Processors are in SMM or not. (Set this field to `0' for legacy behavior) BIOS region SMM protection is enabled. The BIOS Region is not writable unless all Processors are in SMM. 04 TSS Top Swap Status -- This bit provides a read-only path to view the state of the Top Swap bit that is at offset 3414h, bit 0. RWLO SPI Read Configuration -- This 2-bit field controls two policies related to BIOS reads on the SPI interface: Bit 3- Prefetch Enable Bit 2- Cache Disable Settings are summarized as follows: 03 :02 01 00 SRO Bits 3:2 Description 00b No prefetching, but caching enabled. 64B demand reads load the read buffer cache with "valid" data, allowing repeated code fetches to the same line to complete quickly 01b No prefetching and no caching. One-to-one correspondence of host BIOS reads to SPI cycles. This value can be used to invalidate the cache. 10b Prefetching and Caching enabled. This mode is used for long sequences of short reads to consecutive addresses (i.e., shadowing). 11b Reserved. This is an invalid configuration, caching must be enabled when prefetching is enabled. BLE BIOS Lock Enable: Setting the BIOSWE will not cause SMIs. Enables setting the BIOSWE bit to cause SMIs. Once set, this bit can only be cleared by a PLTRST# RWLO BIOSWE BIOS Write Enable: Only read cycles result in Firmware Hub I/F cycles. Access to the BIOS space is enabled for both read and write cycles. When this bit is written from a 0 to a 1 and BIOS Lock Enable (BLE) is also set, an SMI# is generated. This ensures that only SMI code can update BIOS. RW Intel(R) Communications Chipset 89xx Series - Datasheet 260 RW October 2012 Order Number: 327879-001US 6.0 6.1.1.30 Offset E0h: FDCAP--Feature Detection Capability ID (LPC I/F-- B0:D31:F0) Table 6-31. Offset E0h: FDCAP--Feature Detection Capability ID (LPC I/F--B0:D31:F0) Description: View: PCI Size: 16 bit BAR: Configuration B0:D31 :F0 Default: 0009h Bit Range Bit Acronym 15 :08 NEXT 07 :00 6.1.1.31 Bus:Device:Function: Offset Start: E0h Offset End: E1h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Next Item Pointer -- Configuration offset of the next Capability Item. 00h indicates the last item in the Capability List. RO Capability ID -- Indicates a Vendor Specific Capability RO Offset E2h: FDLEN--Feature Detection Capability Length (LPC I/F -- B0:D31:F0) Table 6-32. Offset E2h: FDLEN--Feature Detection Capability Length (LPC I/F--B0:D31:F0) Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 0Ch Bit Range Bit Acronym 07 :00 FDLEN October 2012 Order Number: 327879-001US Offset Start: E2h Offset End: E2h Power Well: Core Bit Description Sticky Capability Length -- Indicates the length of this Vendor Specific capability, as required by PCI Spec. Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 261 6.1.1.32 Offset E3h: FDVER--Feature Detection Version (LPC I/F--B0:D31:F0) Table 6-33. Offset E3h: FDVER--Feature Detection Version (LPC I/F--B0:D31:F0) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 10h Bit Acronym Offset Start: E3h Offset End: E3h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 07 :04 Vendor-Specific Capability ID -- A value of 1h in this 4-bit field identifies this Capability as Feature Detection Type. This field allows software to differentiate the Feature Detection Capability from other Vendor-Specific capabilities RO 03 :00 Capability Version -- This field indicates the version of the Feature Detection capability RO 6.1.1.33 Offset E4h: FDVCT--Feature Vector (LPC I/F--B0:D31:F0) Table 6-34. Offset E4h: FDVCT--Feature Vector (LPC I/F--B0:D31:F0) Description: View: PCI Size: 64 bit Bus:Device:Function: B0:D31 :F0 BAR: Configuration Default: See register description Bit Range Bit Acronym 63 :00 Reserved Bit Description Power Well: Core Sticky Bit Reset Value Bit Access Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 262 Offset Start: E4h Offset End: EBh October 2012 Order Number: 327879-001US 6.0 6.1.1.34 Offset F0h: RCBA--Root Base Address Register (LPC I/F --B0:D31:F0) Table 6-35. Offset F0h: RCBA--Root Complex Base Address Register (LPC I/F--B0:D31:F0) Description: View: PCI BAR: Configuration Size: 32 bit B0:D31 :F0 Offset Start: F0h Offset End: F3h Default: 00000000h Bit Range Bit Acronym 31 :14 BA 13 :01 Reserved 00 6.1.2 Bus:Device:Function: EN Power Well: Bit Description Sticky Bit Reset Value Base Address -- Base Address for the root complex register block decode range. This address is aligned on a 16-KB boundary. Bit Access RW Reserved Enable -- When set, enables the range specified in BA to be claimed as the Root Complex Register Block RW DMA I/O Registers (LPC I/F--B0:D31:F0) Table 6-36. DMA I/O Registers (LPC I/F--B0:D31:F0) Offset Start Offset End Register ID - Description Default Value 00h CCh "Offset 00h: DMABASE_CA--DMA Base and Current Address Registers (LPC I/F -- B0:D31:F0)" on page 264 Undefined 01h CEh "Offset 01h: DMABASE_CC--DMA Base and Current Count Registers (LPC I/F-- B0:D31:F0)" on page 265 Undefined 87h 8Ah "Offset 87h: DMAMEM_LP--DMA Memory Low Page Registers (LPC I/F-- B0:D31:F0)" on page 266 Undefined Ch.#0 - 3 = 08h Ch.#4 - 7 = D0h "Offset 08h: DMACMD--DMA Command Register (LPC I/F--B0:D31:F0)" on page 267 Undefined Ch.#0 - 3 = 08h Ch.#4 - 7 = D0h "Offset 08h: DMASTA--DMA Status Register (LPC I/F--B0:D31:F0)" on page 268 Undefined Ch.#0 - 3 = 0Ah Ch.#4 - 7 = D4h "Offset 0Ah: DMA_WRSMSK--DMA Write Single Mask Register (LPC I/F-- B0:D31:F0)" on page 269 0000 01xx Ch.#0 - 3 = 0Bh Ch.#4 - 7 = D6h "Offset 08h: DMACH_MODE--DMA Channel Mode Register (LPC I/F--B0:D31:F0)" on page 270 0000 00xx Ch.#0 - 3 = 0Ch Ch.#4 - 7 = D8h "Offset 0Ch: DMA Clear Byte Pointer Register (LPC I/F--B0:D31:F0)" on page 271 XXXX XXXX Ch.#0 - 3 = 0Ch Ch.#4 - 7 = DAh "Offset 0Dh: DMA Master Clear Register (LPC I/F--B0:D31:F0)" on page 271 XXXX XXXX Ch.#0 - 3 = 0Eh Ch.#4 - 7 = DCh "Offset 0Eh: DMA Master Clear Register (LPC I/F--B0:D31:F0)" on page 271 XXXX XXXX Ch.#0 - 3 = 0Fh Ch.#4 - 7 = DEh "Offset 0Fh: DMA_WRMSK--DMA Write All Mask Register (LPC I/F--B0:D31:F0)" on 0000 1111 page 272 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 263 6.1.2.1 Offset 00h: DMABASE_CA--DMA Base and Current Address Registers (LPC I/F--B0:D31:F0) I/O Address: Ch. #0 = Ch. #1 = Ch. #2 = Ch. #3 = Ch. #5 = Ch. #6 = Ch. #7 = 00h 02h 04h 06h C4h C8h CCh Table 6-37. Offset 00h: DMABASE_CA--DMA Base and Current Address Registers (LPC I/F --B0:D31:F0) Description: View: IA F Size: 16 bita Bit Range 15 :00 Bus:Device:Function: B0:D31 :F0 BAR: 000h (IO) Default: Undefined Bit Acrony m Offset Start: 00h Offset End: CCh Power Well: Core Bit Description Base and Current Address -- This register determines the address for the transfers to be performed. The address specified points to two separate registers. On writes, the value is stored in the Base Address register and copied to the Current Address register. On reads, the value is returned from the Current Address register. The address increments/decrements in the Current Address register after each transfer, depending on the mode of the transfer. If the channel is in auto-initialize mode, the Current Address register will be reloaded from the Base Address register after a terminal count is generated. For transfers to/from a 16-bit slave (channel's 57), the address is shifted left one bit location. Bit 15 will be shifted into Bit 16. The register is accessed in 8 bit quantities. The byte is pointed to by the current byte pointer flip/ flop. Before accessing an address register, the byte pointer flip/flop should be cleared to ensure that the low byte is accessed first Sticky Bit Reset Value Bit Access RW a. per channel but accessed in two 8-bit quantities Intel(R) Communications Chipset 89xx Series - Datasheet 264 October 2012 Order Number: 327879-001US 6.0 6.1.2.2 Offset 01h: DMABASE_CC--DMA Base and Current Count Registers (LPC I/F--B0:D31:F0) I/O Address: Ch. #0 = Ch. #1 = Ch. #2 = Ch. #3 = Ch. #5 = Ch. #6 = Ch. #7 = 01h 03h 05h 07h C6h CAh CEh Table 6-38. Offset 01h: DMABASE_CC--DMA Base and Current Count Registers (LPC I/F-- B0:D31:F0) Description: View: IA F Size: 16 bita Bit Range BAR: 000h (IO) B0:D31 Bus:Device:Function: :F0 Default: Undefined Bit Acronym 15 :00 Offset Start: 01h Offset End: CEh Power Well: Core Bit Description Sticky Bit Reset Value Base and Current Count -- This register determines the number of transfers to be performed. The address specified points to two separate registers. On writes, the value is stored in the Base Count register and copied to the Current Count register. On reads, the value is returned from the Current Count register. The actual number of transfers is one more than the number programmed in the Base Count Register (i.e., programming a count of 4h results in 5 transfers). The count is decrements in the Current Count register after each transfer. When the value in the register rolls from 0 to FFFFh, a terminal count is generated. If the channel is in auto-initialize mode, the Current Count register will be reloaded from the Base Count register after a terminal count is generated. For transfers to/from an 8-bit slave (channels 0-3), the count register indicates the number of bytes to be transferred. For transfers to/from a 16-bit slave (channels 5-7), the count register indicates the number of words to be transferred. The register is accessed in 8 bit quantities. The byte is pointed to by the current byte pointer flip/ flop. Before accessing a count register, the byte pointer flip/flop should be cleared to ensure that the low byte is accessed first. Bit Access RW a.per channel but accessed in two 8-bit quantities October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 265 6.1.2.3 Offset 87h: DMAMEM_LP--DMA Memory Low Page Registers (LPC I/F --B0:D31:F0) I/O Address: Ch. #0 = Ch. #1 = Ch. #2 = Ch. #3 = Ch. #5 = Ch. #6 = Ch. #7 = 87h 83h 81h 82h 8Bh 89h 8Ah Table 6-39. Offset 87h: DMAMEM_LP--DMA Memory Low Page Registers (LPC I/F-- B0:D31:F0) Description: View: IA F BAR: 000h (IO) Size: 8 bit Default: Undefined Bit Range 07 :00 Bit Acronym Bus:Device:Function: B0:D31 :F0 Power Well: Core Bit Description DMA Low Page (ISA Address bits [23:16]) -- This register works in conjunction with the DMA controller's Current Address Register to define the complete 24-bit address for the DMA channel. This register remains static throughout the DMA transfer. Bit 16 of this register is ignored when in 16 bit I/O count by words mode as it is replaced by the bit 15 shifted out from the current address register. Intel(R) Communications Chipset 89xx Series - Datasheet 266 Offset Start: 87h Offset End: 8Ah Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 6.0 6.1.2.4 Offset 08h: DMACMD--DMA Command Register (LPC I/F--B0:D31:F0) Table 6-40. Offset 08h: DMACMD--DMA Command Register (LPC I/F--B0:D31:F0) Description: View: IA F BAR: 000h (IO) Size: 8 bit Default: Undefined Bit Range 07 :05 Bit Acronym Offset Start: Ch.#0 - 3 = 08h Offset End: Ch.#4 - 7 = D0h Power Well: Core Bit Description Sticky Bit Reset Value DMA Group Arbitration Priority -- Each channel group is individually assigned either fixed or rotating arbitration priority. At part reset, each group is initialized in fixed priority. 0 = Fixed priority to the channel group 1 = Rotating priority to the group. Bit Access WO Reserved Reserved. Must be 0. DMA Channel Group Enable -- Both channel groups are enabled following part reset. 0 = Enable the DMA channel group. 1 = Disable. Disabling channel group 4-7 also disables channel group 0-3, which is cascaded through channel 4. 02 01 :00 B0:D31 :F0 Reserved Reserved. Must be 0. 04 03 Bus:Device:Function: WO Reserved Reserved. Must be 0. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 267 6.1.2.5 Offset 08h: DMASTA--DMA Status Register (LPC I/F--B0:D31:F0) Table 6-41. Offset 08h: DMASTA--DMA Status Register (LPC I/F--B0:D31:F0) Description: View: IA F BAR: 000h (IO) Size: 8 bit Default: Undefined Bit Range Bit Acronym Bus:Device:Function: Offset Start: Ch.#0 - 3 = 08h Offset End: Ch.#4 - 7 = D0h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 07 :04 Channel Request Status -- When a valid DMA request is pending for a channel, the corresponding bit is set to 1. When a DMA request is not pending for a particular channel, the corresponding bit is set to 0. The source of the DREQ may be hardware or a software request. Channel 4 is the cascade channel, so the request status of channel 4 is a logical OR of the request status for channels 0 through 3. 4 = Channel 0 5 = Channel 1 (5) 6 = Channel 2 (6) 7 = Channel 3 (7) RO 03 :00 Channel Terminal Count Status -- When a channel reaches terminal count (TC), its status bit is set to 1. If TC has not been reached, the status bit is set to 0. Channel 4 is programmed for cascade, so the TC bit response for channel 4 is irrelevant: 0 = Channel 0 1 = Channel 1 (5) 2 = Channel 2 (6) 3 = Channel 3 (7) RO Intel(R) Communications Chipset 89xx Series - Datasheet 268 B0:D31 :F0 October 2012 Order Number: 327879-001US 6.0 6.1.2.6 Offset 0Ah: DMA_WRSMSK--DMA Write Single Mask Register (LPC I/F --B0:D31:F0) Table 6-42. Offset 0Ah: DMA_WRSMSK--DMA Write Single Mask Register (LPC I/F-- B0:D31:F0) Description: View: IA F Size: 8 bit Bit Range 07 :03 BAR: 000h (IO) Bus:Device:Function: Default: 0000 01xx Bit Acronym B0:D31 :F0 Offset Start: Ch.#0 - 3 = 0Ah Offset End: Ch.#4 - 7 = D4h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved. Must be 0. 02 01 :00 October 2012 Order Number: 327879-001US Channel Mask Select: 0 = Enable DREQ for the selected channel. The channel is selected through bits [1:0]. Therefore, only one channel can be masked / unmasked at a time. 1 = Disable DREQ for the selected channel. WO DMA Channel Select -- These bits select the DMA Channel Mode Register to program. 00 = Channel 0 (4) 01 = Channel 1 (5) 10 = Channel 2 (6) 11 = Channel 3 (7) WO Intel(R) Communications Chipset 89xx Series - Datasheet 269 6.1.2.7 Offset 0Bh: DMACH_MODE--DMA Channel Mode Register (LPC I/F-- B0:D31:F0) Table 6-43. Offset 08h: DMACH_MODE--DMA Channel Mode Register (LPC I/F-- B0:D31:F0) Description: View: IA F Size: 8 bit Bit Range BAR: 000h (IO) Bus:Device:Function: Default: 0000 00xx Bit Acronym Offset Start: Ch.#0 - 3 = 0Bh Offset End: Ch.#4 - 7 = D6h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access DMA Transfer Mode -- Each DMA channel can be programmed in one of four different modes: 00 = Demand mode 01 = Single mode 10 = Reserved 11 = Cascade mode WO 05 Address Increment/Decrement Select -- This bit controls address increment/decrement during DMA transfers. 0 = Address increment. (default after part reset or Master Clear) 1 = Address decrement. WO 04 Autoinitialize Enable: 0 = Autoinitialize feature is disabled and DMA transfers terminate on a terminal count. A part reset or Master Clear disables autoinitialization. 1 = DMA restores the Base Address and Count registers to the current registers following a terminal count (TC). WO 03 :02 DMA Transfer Type -- These bits represent the direction of the DMA transfer. When the channel is programmed for cascade mode, (bits[7:6] = 11) the transfer type is irrelevant. 00 = Verify - No I/O or memory strobes generated 01 = Write - Data transferred from the I/O devices to memory 10 = Read - Data transferred from memory to the I/O device 11 = Invalid WO 01 :00 DMA Channel Select -- These bits select the DMA Channel Mode Register that will be written by bits [7:2]. 00 = Channel 0 (4) 01 = Channel 1 (5) 10 = Channel 2 (6) 11 = Channel 3 (7) WO 07 :06 Intel(R) Communications Chipset 89xx Series - Datasheet 270 B0:D31 :F0 October 2012 Order Number: 327879-001US 6.0 6.1.2.8 Offset 0Ch: DMA Clear Byte Pointer Register (LPC I/F--B0:D31:F0) Table 6-44. Offset 0Ch: DMA Clear Byte Pointer Register (LPC I/F--B0:D31:F0) Description: View: IA F BAR: 000h (IO) Size: 8 bit B0:D31 :F0 Default: XXXX XXXX Bit Acronym Bit Range Offset Start: Ch.#0 - 3 = 0Ch Offset End: Ch.#4 - 7 = D8h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Clear Byte Pointer -- No specific pattern. Command enabled with a write to the I/O port address. Writing to this register initializes the byte pointer flip/flop to a known state. It clears the internal latch used to address the upper or lower byte of the 16-bit Address and Word Count Registers. The latch is also cleared by part reset and by the Master Clear command. This command precedes the first access to a 16-bit DMA controller register. The first access to a 16-bit register will then access the significant byte, and the second access automatically accesses the most significant byte. 07 :00 6.1.2.9 Bus:Device:Function: WO Offset 0Dh: DMA Master Clear Register (LPC I/F--B0:D31:F0) Table 6-45. Offset 0Dh: DMA Master Clear Register (LPC I/F--B0:D31:F0) Description: View: IA F BAR: 000h (IO) Size: 8 bit Default: XXXX XXXX Bit Acronym Bit Range Offset Start: Ch.#0 - 3 = 0Ch Offset End: Ch.#4 - 7 = DAh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Master Clear -- No specific pattern. Enabled with a write to the port. This has the same effect as the hardware Reset. The Command, Status, Request, and Byte Pointer flip/flop registers are cleared and the Mask Register is set. 07 :00 6.1.2.10 B0:D31 Bus:Device:Function: :F0 WO Offset 0Eh: DMA_CLMSK--DMA Clear Mask Register (LPC I/F-- B0:D31:F0) Table 6-46. Offset 0Eh: DMA Master Clear Register (LPC I/F--B0:D31:F0) Description: View: IA F Size: 8 bit Bit Range BAR: 000h (IO) B0:D31 Bus:Device:Function: :F0 Default: XXXX XXXX Bit Acronym 07 :00 October 2012 Order Number: 327879-001US Offset Start: Ch.#0 - 3 = 0Eh Offset End: Ch.#4 - 7 = DCh Power Well: Core Bit Description Sticky Bit Reset Value Clear Mask Register -- No specific pattern. Command enabled with a write to the port. Bit Access WO Intel(R) Communications Chipset 89xx Series - Datasheet 271 6.1.2.11 Offset 0Fh: DMA_WRMSK--DMA Write All Mask Register (LPC I/F-- B0:D31:F0) Table 6-47. Offset 0Fh: DMA_WRMSK--DMA Write All Mask Register (LPC I/F--B0:D31:F0) Description: View: IA F BAR: 000h (IO) Size: 8 bit Bus:Device:Function: B0:D31 :F0 Default: 0000 1111 Bit Acronym Bit Range 07 :04 Offset Start: Ch.#0 - 3 = 0Fh Offset End: Ch.#4 - 7 = DEh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved. Must be 0. Channel Mask Bits --This register permits all four channels to be simultaneously enabled/disabled instead of enabling/disabling each channel individually, as is the case with the Mask Register - Write Single Mask Bit. In addition, this register has a read path to allow the status of the channel mask bits to be read. A channel's mask bit is automatically set to 1 when the Current Byte/Word Count Register reaches terminal count (unless the channel is in auto-initialization mode). Setting the bit(s) to a 1 disables the corresponding DREQ(s). Setting the bit(s) to a 0 enables the corresponding DREQ(s). Bits [3:0] are set to 1 upon part reset or Master Clear. When read, bits [3:0] indicate the DMA channel [3:0] ([7:4]) mask status. Bit 0 = Channel 0 (4)1 = Masked, 0 = Not Masked Bit 1 = Channel 1 (5)1 = Masked, 0 = Not Masked Bit 2 = Channel 2 (6)1 = Masked, 0 = Not Masked Bit 3 = Channel 3 (7)1 = Masked, 0 = Not Masked 03 :00 RW Disabling channel 4 also disables channels 0-3 due to the cascade of channel's 0 - 3 through channel 4. 6.2 Timer I/O Registers (LPC I/F--B0:D31:F0) 6.2.1 Timer I/O Registers Table 6-48. Timer I/O Registers Offset Start Offset End Default Value Register ID - Description 43h 43h "Offset 43h: TCW--Timer Control Word Register (LPC I/F--B0:D31:F0)" on page 273 Bits[7:1]= undefined, Bit 0=0 40h 42h "Offset 40h: SBYTE_FMT--Interval Timer Status Byte Format Register (LPC I/F-- B0:D31:F0)" on page 275 Bits[6:0] = undefined, Bit[7]=0 40h 42h "Offset 40h: Counter Access Ports Register (LPC I/F--B0:D31:F0)" on page 276 Undefined Intel(R) Communications Chipset 89xx Series - Datasheet 272 October 2012 Order Number: 327879-001US 6.0 6.2.1.1 Offset 43h: TCW--Timer Control Word Register (LPC I/F--B0:D31:F0) This register is programmed prior to any counter being accessed to specify counter modes. Following part reset, the control words for each register are undefined and each counter output is 0. Each timer must be programmed to bring it into a known state. Table 6-49. Offset 43h: TCW--Timer Control Word Register (LPC I/F--B0:D31:F0) Description: View: IA F Size: 8 bit Bit Range BAR: 000h (IO) Default: Bus:Device:Function: Bits[7:1] = undefined, Bit[0]=0 Bit Acronym B0:D31 :F0 Offset Start: 43h Offset End: 43h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 07 :06 Counter Select -- The Counter Selection bits select the counter the control word acts upon as shown below. The Read Back Command is selected when bits[7:6] are both 1. 00 = Counter 0 select 01 = Counter 1 select 10 = Counter 2 select 11 = Read Back Command WO 05 :04 Read/Write Select -- These bits are the read/write control bits. The actual counter programming is done through the counter port (40h for counter 0, 41h for counter 1, and 42h for counter 2). 00 = Counter Latch Command 01 = Read/Write Least Significant Byte (LSB) 10 = Read/Write Most Significant Byte (MSB) 11 = Read/Write LSB then MSB WO Counter Mode Selection -- These bits select one of six possible modes of operation for the selected counter. Bit Value 03 :01 Mode 000b Mode 0 Out signal on end of count (=0) 001b Mode 1 Hardware retriggerable one-shot x10b Mode 2 Rate generator (divide by n counter) x11b Mode 3 Square wave output 100b Mode 4 Software triggered strobe Binary/BCD Countdown Select: 0 = Binary countdown is used. The largest possible binary count is 216 00 WO WO 1 = Binary coded decimal (BCD) count is used. The largest possible BCD count is 104 There are two special commands that can be issued to the counters through this register, the Read Back Command and the Counter Latch Command. When these commands are chosen, several bits within this register are redefined. These register formats are described below: October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 273 RDBK_CMD--Read Back Command (LPC I/F--B0:D31:F0) The Read Back Command is used to determine the count value, programmed mode, and current states of the OUT pin and Null count flag of the selected counter or counters. Status and/or count may be latched in any or all of the counters by selecting the counter during the register write. The count and status remain latched until read, and further latch commands are ignored until the count is read. Both count and status of the selected counters may be latched simultaneously by setting both bit 5 and bit 4 to 0. If both are latched, the first read operation from that counter returns the latched status. The next one or two reads, depending on whether the counter is programmed for one or two byte counts, returns the latched count. Subsequent reads return an unlatched count. Bit 07:06 Description Read Back Command. Must be 11 to select the Read Back Command 05 Latch Count of Selected Counters. 0 = Current count value of the selected counters will be latched 1 = Current count will not be latched 04 Latch Status of Selected Counters. 0 = Status of the selected counters will be latched 1 = Status will not be latched 03 Counter 2 Select. 1 = Counter 2 count and/or status will be latched 02 Counter 1 Select. 1 = Counter 1 count and/or status will be latched 01 Counter 0 Select. 1 = Counter 0 count and/or status will be latched. 00 Reserved. Must be 0. LTCH_CMD--Counter Latch Command (LPC I/F--B0:D31:F0) The Counter Latch Command latches the current count value. This command is used to insure that the count read from the counter is accurate. The count value is then read from each counter's count register through the Counter Ports Access Ports Register (40h for counter 0, 41h for counter 1, and 42h for counter 2). The count must be read according to the programmed format, i.e., if the counter is programmed for two byte counts, two bytes must be read. The two bytes do not have to be read one right after the other (read, write, or programming operations for other counters may be inserted between the reads). If a counter is latched once and then latched again before the count is read, the second Counter Latch Command is ignored. Bit Description 07:06 Counter Selection. These bits select the counter for latching. If "11" is written, then the write is interpreted as a read back command. 00 = Counter 0 01 = Counter 1 10 = Counter 2 05:04 Counter Latch Command. 00 = Selects the Counter Latch Command. 03:00 Reserved. Must be 0. Intel(R) Communications Chipset 89xx Series - Datasheet 274 October 2012 Order Number: 327879-001US 6.0 6.2.1.2 Offset 40h: SBYTE_FMT--Interval Timer Status Byte Format Register (LPC I/F--B0:D31:F0) Each counter's status byte can be read following a Read Back Command. If latch status is chosen (bit 4=0, Read Back Command) as a read back option for a given counter, the next read from the counter's Counter Access Ports Register (40h for counter 0, 41h for counter 1, and 42h for counter 2) returns the status byte. Table 6-50. Offset 40h: SBYTE_FMT--Interval Timer Status Byte Format Register (LPC I/F--B0:D31:F0) Description: View: IA F 8 bit per Size: counter Bit Range BAR: 000h (IO) Bus:Device:Function: Bits[6:0] = undefined, Default: Bit[7]=0 Bit Acronym B0:D31 :F0 Offset Start: 40h Offset End: 42h Power Well: Bit Description Sticky Bit Reset Value Bit Access 07 Counter OUT Pin State: 0 = OUT pin of the counter is also a 0 1 = OUT pin of the counter is also a 1 RO 06 Count Register Status -- This bit indicates when the last count written to the Count Register (CR) has been loaded into the counting element (CE). The exact time this happens depends on the counter mode, but until the count is loaded into the counting element (CE), the count value will be incorrect. 0 = Count has been transferred from CR to CE and is available for reading. 1 = Null Count. Count has not been transferred from CR to CE and is not yet available for reading. RO 05 :04 Read/Write Selection Status -- These reflect the read/write selection made through bits[5:4] of the control register. The binary codes returned during the status read match the codes used to program the counter read/write selection. 00 = Counter Latch Command 01 = Read/Write Least Significant Byte (LSB) 10 = Read/Write Most Significant Byte (MSB) 11 = Read/Write LSB then MSB RO 03 :01 Mode Selection Status -- These bits return the counter mode programming. The binary code returned matches the code used to program the counter mode, as listed under the bit function above. 000 = Mode 0 -- Out signal on end of count (=0) 001 = Mode 1 -- Hardware retriggerable one-shot x10 = Mode 2 -- Rate generator (divide by n counter) x11 = Mode 3 -- Square wave output 100 = Mode 4 -- Software triggered strobe 101 = Mode 5 -- Hardware triggered strobe RO Countdown Type Status -- This bit reflects the current countdown type. 0 = Binary countdown 1 = Binary Coded Decimal (BCD) countdown. RO 00 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 275 6.2.1.3 Offset 40h: Counter Access Ports Register (LPC I/F--B0:D31:F0) I/O Address: Counter 0 = 40h Counter 1 = 41h Counter 2 = 42h Table 6-51. Offset 40h: Counter Access Ports Register (LPC I/F--B0:D31:F0) Description: View: IA F Size: 8 bit Bit Range 07 :00 B0:D31 Bus:Device:Function: :F0 BAR: 000h (IO) Default: All bits undefined Bit Acronym Power Well: Bit Description Counter Port -- Each counter port address is used to program the 16-bit Count Register. The order of programming, either LSB only, MSB only, or LSB then MSB, is defined with the Interval Counter Control Register at port 43h. The counter port is also used to read the current count from the Count Register, and return the status of the counter programming following a Read Back Command. Intel(R) Communications Chipset 89xx Series - Datasheet 276 Offset Start: 40h Offset End: 42h Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 6.0 6.3 8259 Interrupt Controller (PIC) Registers 6.3.1 Interrupt Controller I/O MAP The interrupt controller registers are located at 20h and 21h for the master controller (IRQ 0-7), and at A0h and A1h for the slave controller (IRQ 8-13). These registers have multiple functions, depending upon the data written to them. Table 6-52 shows the register possibilities for each address. Table 6-52. PIC Registers Port 20h Aliases Register Name Default Value Type 24h, 28h, "Offset 20h: Master PIC ICW1--Master Initialization Command Word 1 Register" Undefined WO "Offset 020h: Master PIC OCW2--Master Operational Control Word 2 Register" 001XXXXXb WO "Offset 020h: Master PIC OCW3--Master Operational Control Word 3 Register" X01XXX10b WO "Offset 21h: Master PIC ICW2--Master Initialization Command Word 2 Register" Undefined WO 2Dh, 31h, "Offset 21h: Master PIC ICW3--Master Initialization Command Word 3 Register" Undefined WO 35h, 39h, 3Dh "Offset 021h: Master PIC ICW4--Master Initialization Command Word 4 Register" 01h WO "Offset 021h: Master PIC OCW1--Master Operational Control Word 1 (Interrupt Mask) Register" 00h R/W Undefined WO 2Ch, 30h, 34h, 38h, 3Ch 25h, 29h, 21h A0h A4h, A8h, Slave PIC ICW1 - Slave Initialization Command Word 1 ACh, B0h, Slave PIC OCW2 - Slave Operational Control Word 2 001XXXXXb WO Slave PIC OCW3 - Slave Operational Control Word 3 X01XXX10b WO Slave PIC ICW2 - Slave Initialization Command Word 2 Undefined WO ADh, B1h, Slave PIC ICW3 - Slave Initialization Command Word 3 Undefined WO B5h, B9h, BDh Slave PIC ICW4 - Slave Initialization Command Word 4 01h WO Slave PIC OCW1 - Slave Operational Control Word 1 00h R/W B4h, B8h, BCh A5h, A9h, A1h 4D0h - "Offset 4D0h: Master PIC ELCR1--Master Controller Edge/ Level Triggered Register" 00h R/W 4D1h - "Offset 4D1h: Slave PIC ELCR2--Slave Controller Edge/Level Triggered Register" 00h R/W Note: See the note addressing active-low interrupt sources in 8259 Interrupt Controllers section (Chapter 4.6). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 277 6.3.1.1 Offset 20h: Master PIC ICW1--Master Initialization Command Word 1 Register Offset Address: Master Controller - 20h Slave Controller - A0h Default Value: All bits undefined Attribute: Size: WO 8 bit /controller A write to Initialization Command Word 1 starts the interrupt controller initialization sequence, during which the following occurs: 1. The Interrupt Mask register is cleared. 2. IRQ7 input is assigned priority 7. 3. The slave mode address is set to 7. 4. Special mask mode is cleared and Status Read is set to IRR. Once this write occurs, the controller expects writes to ICW2, ICW3, and ICW4 to complete the initialization sequence. Table 6-53. Offset 20h: Master PIC ICW1--Master Initialization Command Word 1 Register Description: View: IA F bit per Size: 8 controller Bit Range Default: Undefined Bit Acronym 07 :05 04 03 Bus:Device:Function: B0:D31 :F0 BAR: 000h (IO) LTIM 02 01 SNGL 00 IC4 Power Well: Bit Description Sticky Bit Reset Value Bit Access ICW/OCW Select -- These bits are MCS-85 specific, and not needed. 000 = Should be programmed to "000" WO ICW/OCW Select -- This bit must be a 1 to select ICW1 and enable the ICW2, ICW3, and ICW4 sequence. WO Edge/Level Bank Select -- Disabled. Replaced by the edge/level triggered control registers (ELCR, B0:D31:F0:4D0h, B0:D31:F0:4D1h). WO ADI -- Ignored for the PCH. Should be programmed to 0. WO Single or Cascade: Must be programmed to a 0 to indicate two controllers operating in cascade mode. WO ICW4 Write Required --This bit must be programmed to a 1 to indicate that ICW4 needs to be programmed. WO Intel(R) Communications Chipset 89xx Series - Datasheet 278 Offset Start: 20h Offset End: 20h October 2012 Order Number: 327879-001US 6.0 6.3.1.2 Offset 21h: Master PIC ICW2--Master Initialization Command Word 2 Register Offset Address: Master Controller - 21h Slave Controller - A1h Default Value: All bits undefined Attribute: Size: WO 8 bit /controller ICW2 is used to initialize the interrupt controller with the five most significant bits of the interrupt vector address. The value programmed for bits[7:3] is used by the processor to define the base address in the interrupt vector table for the interrupt routines associated with each IRQ on the controller. Typical ISA ICW2 values are 08h for the master controller and 70h for the slave controller. Table 6-54. Offset 21h: Master PIC ICW2-- Master Initialization Command Word 2 Register Description: View: IA F Size: 8 bit per controller Bit Range Bus:Device:Function: B0:D31 :F0 BAR: 000h (IO) Default: Undefined Power Well: Bit Acronym 07 :03 Offset Start: 21h Offset End: 21h Bit Description Sticky Bit Reset Value Interrupt Vector Base Address -- Bits [7:3] define the base address in the interrupt vector table for the interrupt routines associated with each interrupt request level input. Bit Access WO Interrupt Request Level -- When writing ICW2, these bits should all be 0. During an interrupt acknowledge cycle, these bits are programmed by the interrupt controller with the interrupt to be serviced. This is combined with bits [7:3] to form the interrupt vector driven onto the data bus during the second INTA# cycle. The code is a three bit binary code. 02 :00 October 2012 Order Number: 327879-001US Code Master Interrupt Slave Interrupt 000b IRQ0 IRQ8 001b IRQ1 IRQ9 010b IRQ2 IRQ10 011b IRQ3 IRQ11 100b IRQ4 IRQ12 101b IRQ5 IRQ13 110b IRQ6 IRQ14 111b IRQ7 IRQ15 WO Intel(R) Communications Chipset 89xx Series - Datasheet 279 6.3.1.3 Offset 21h: Master PIC ICW3--Master Initialization Command Word 3 Register Offset Address: 21h Default Value: All bits undefined Attribute: Size: WO 8 bits Table 6-55. Offset 21h: Master PIC ICW3--Master Initialization Command Word 3 Register Description: View: IA F Size: BAR: 000h (IO) 8 bit per controller 07 :03 Offset Start: 21h Offset End: 21h Power Well: Bit Description Sticky Bit Reset Value 0 = These bits must be programmed to 0. 01 :00 Bit Access O Cascaded Interrupt Controller IRQ Connection -- This bit indicates that the slave controller is cascaded on IRQ2. When IRQ8#-IRQ15 is asserted, it goes through the slave controller's priority resolver. The slave controller's INTR output onto IRQ2. IRQ2 then goes through the master controller's priority solver. If it wins, the INTR signal is asserted to the processor, and the returning interrupt acknowledge returns the interrupt vector for the slave controller. 1 = This bit must always be programmed to a 1. 02 6.3.1.4 B0:D31 :F0 Default: Undefined Bit Acronym Bit Range Bus:Device:Function: WO 0 = These bits must be programmed to 0. O Offset A1h: Slave PIC ICW3--Slave Initialization Command Word 3 Register Offset Address: A1h Default Value: All bits undefined Attribute: Size: WO 8 bits Table 6-56. Offset A1h: Slave PIC ICW3--Slave Initialization Command Word 3 Register Description: View: IA F BAR: 000h (IO) Size: 8 bit Default: Undefined Bit Range Bit Acronym Bus:Device:Function: B0:D31 :F0 Power Well: Bit Description 07 :03 0 = These bits must be programmed to 0. 02 :00 Slave Identification Code -- These bits are compared against the slave identification code broadcast by the master controller from the trailing edge of the first internal INTA# pulse to the trailing edge of the second internal INTA# pulse. These bits must be programmed to 02h to match the code broadcast by the master controller. When 02h is broadcast by the master controller during the INTA# sequence, the slave controller assumes responsibility for broadcasting the interrupt vector. Intel(R) Communications Chipset 89xx Series - Datasheet 280 Offset Start: A1h Offset End: A1h Sticky Bit Reset Value Bit Access O WO October 2012 Order Number: 327879-001US 6.0 6.3.1.5 Offset 021h: Master PIC ICW4--Master Initialization Command Word 4 Register Offset Address: Master Controller - 021h Slave Controller - 0A1h Default Value: 01h Attribute: Size: WO 8 bits Table 6-57. Offset 021h: Master PIC ICW4--Master Initialization Command Word 4 Register Description: View: IA F Size: 8 bit Bit Range BAR: 000h (IO) Bus:Device:Function: Default: 01h Bit Acronym 07 :05 B0:D31 :F0 Offset Start: 021h Offset End: 021h Power Well: Bit Description Sticky Bit Reset Value 0 = These bits must be programmed to 0. Bit Access O 04 SFNM Special Fully Nested Mode: 0 = Should normally be disabled by writing a 0 to this bit. 1 = Special fully nested mode is programmed. 03 BUF Buffered Mode: 0 = Must be programmed to 0 for the PCH. This is non-buffered mode. WO 02 Master/Slave in Buffered Mode -- Not used. 0 = Should always be programmed to 0. WO 01 Automatic End of Interrupt: 0 = This bit should normally be programmed to 0. This is the normal end of interrupt. 1 = Automatic End of Interrupt mode is programmed. WO Microprocessor Mode --Must be programmed to 1 to indicate that the controller is operating in an Intel Architecture-based system. WO AEOI 00 October 2012 Order Number: 327879-001US WO Intel(R) Communications Chipset 89xx Series - Datasheet 281 6.3.1.6 Offset 021h: Master PIC OCW1--Master Operational Control Word 1 (Interrupt Mask) Register Offset Address: Master Controller - 021h Slave Controller - 0A1h Default Value: 00h Attribute: Size: R/W 8 bits h Table 6-58. Offset 021h: Master PIC OCW1--Master Operational Control Word 1 (Interrupt Mask) Register Description: View: IA F Size: 8 bit Bit Range 07 :00 BAR: 000h (IO) Bus:Device:Function: Default: 00h Bit Acronym Offset Start: 021h Offset End: 021h Power Well: Bit Description Interrupt Request Mask -- When a 1 is written to any bit in this register, the corresponding IRQ line is masked. When a 0 is written to any bit in this register, the corresponding IRQ mask bit is cleared, and interrupt requests will again be accepted by the controller. Masking IRQ2 on the master controller will also mask the interrupt requests from the slave controller. Intel(R) Communications Chipset 89xx Series - Datasheet 282 B0:D31 :F0 Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 6.0 6.3.1.7 Offset 020h: Master PIC OCW2--Master Operational Control Word 2 Register Offset Address: Master Controller - 020h Attribute: Slave Controller - 0A0h Size: Default Value: Bit[4:0]=undefined, Bit[7:5]=001 WO 8 bits Following a part reset or ICW initialization, the controller enters the fully nested mode of operation. Non-specific EOI without rotation is the default. Both rotation mode and specific EOI mode are disabled following initialization. Table 6-59. Offset 020h: Master PIC OCW2--Master Operational Control Word 2 Register Description: View: IA F Size: 8 bit Bit Range 07 :05 BAR: 000h (IO) Bus:Device:Function: B0:D31 :F0 Default: 001XXXXXb Bit Acronym Power Well: Bit Description Sticky Rotate and EOI Codes --These three bits control the Rotate and End of Interrupt modes and combinations of the two. 000 = Reserved 001 = Non-specific EOI command 010 = No Operation R, SL, EOI 011 = *Specific EOI Command 100 = Reserved 101 = Rotate on Non-Specific EOI Command 110 = *Set Priority Command 111 = *Rotate on Specific EOI Command *L0 - L2 Are Used OCW2 Select -- When selecting OCW2, bits 4:3 = "00" 04 :03 Offset Start: 020h Offset End: 020h Bit Reset Value Bit Access WO WO Interrupt Level Select -- L2, L1, and L0 determine the interrupt level acted upon when the SL bit is active. A simple binary code, outlined below, selects the channel for the command to act upon. When the SL bit is inactive, these bits do not have a defined function; programming L2, L1 and L0 to 0 is sufficient in this case. 02 :00 L2, L1, L0 October 2012 Order Number: 327879-001US Code Interrupt Level Code Interrupt Level 000b IRQ0/8 000b IRQ4/12 001b IRQ1/9 001b IRQ5/13 010b IRQ2/10 010b IRQ6/14 011b IRQ3/11 011b IRQ7/15 WO Intel(R) Communications Chipset 89xx Series - Datasheet 283 6.3.1.8 Offset 020h: Master PIC OCW3--Master Operational Control Word 3 Register Offset Address: Master Controller - 020h Attribute:WO Slave Controller - 0A0h Size: 8 bits Default Value: Bit[6,0]=0, Bit[7,4:2]=undefined, Bit[5,1]=1 Table 6-60. Offset 020h: Master PIC OCW3--Master Operational Control Word 3 Register Description: View: IA F Size: 8 bit Bit Range 07 BAR: 000h (IO) Bus:Device:Function: Default: X01XXX10b Bit Acronym Offset Start: 020h Offset End: 020h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved. Must be 0. 06 SMM Special Mask Mode: 1 = The Special Mask Mode can be used by an interrupt service routine to dynamically alter the system priority structure while the routine is executing, through selective enabling/disabling of the other channel's mask bits. Bit 5, the ESMM bit, must be set for this bit to have any meaning. 05 ESMM Enable Special Mask Mode. 0 = Disable. The SMM bit becomes a "don't care". 1 = Enable the SMM bit to set or reset the Special Mask Mode. WO OCW3 Select -- When selecting OCW3, bits 4:3 = 01 WO Poll Mode Command: 0 = Disable. Poll Command is not issued. 1 = Enable. The next I/O read to the interrupt controller is treated as an interrupt acknowledge cycle. An encoded byte is driven onto the data bus, representing the highest priority level requesting service. WO Register Read Command -- These bits provide control for reading the In-Service Register (ISR) and the Interrupt Request Register (IRR). When bit 1=0, bit 0 will not affect the register read selection. When bit 1=1, bit 0 selects the register status returned following an OCW3 read. If bit 0=0, the IRR will be read. If bit 0=1, the ISR will be read. Following ICW initialization, the default OCW3 port address read will be "read IRR". To retain the current selection (read ISR or read IRR), always write a 0 to bit 1 when programming this register. The selected register can be read repeatedly without reprogramming OCW3. To select a new status register, OCW3 must be reprogrammed prior to attempting the read. 00 = No Action 01 = No Action 10 = Read IRQ Register 11 = Read IS Register WO 04 :03 02 01 :00 Intel(R) Communications Chipset 89xx Series - Datasheet 284 B0:D31 :F0 WO October 2012 Order Number: 327879-001US 6.0 6.3.1.9 Offset 4D0h: Master PIC ELCR1--Master Controller Edge/Level Triggered Register Offset Address: 4D0h Default Value: 00h Attribute: Size: R/W 8 bits In edge mode, (bit[x] = 0), the interrupt is recognized by a low to high transition. In level mode (bit[x] = 1), the interrupt is recognized by a high level. The cascade channel, IRQ2, the heart beat timer (IRQ0), and the keyboard controller (IRQ1), cannot be put into level mode Table 6-61. Offset 4D0h: Master PIC ELCR1--Master Controller Edge/Level Triggered Register Description: View: IA F Size: 8 bit Bit Range Bus:Device:Function: B0:D31 :F0 BAR: 000h (IO) Default: 00h Bit Acronym Offset Start: 4D0h Offset End: 4D0h Power Well: Bit Description Sticky Bit Reset Value Bit Access 07 IRQ7 ECL: 0 = Edge. 1 = Level. RW 06 IRQ6 ECL: 0 = Edge. 1 = Level. RW 05 IRQ5 ECL: 0 = Edge. 1 = Level. RW 04 IRQ4 ECL: 0 = Edge. 1 = Level. RW 03 IRQ3 ECL: 0 = Edge. 1 = Level. RW 02 :00 Reserved Reserved. Must be 0. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 285 6.3.1.10 Offset 4D1h: Slave PIC ELCR2--Slave Controller Edge/Level Triggered Register Offset Address: 4D1h Default Value: 00h Attribute: Size: R/W 8 bits In edge mode, (bit[x] = 0), the interrupt is recognized by a low to high transition. In level mode (bit[x] = 1), the interrupt is recognized by a high level. The real time clock, IRQ8#, and the floating point error interrupt, IRQ13, cannot be programmed for level mode. Table 6-62. Offset 4D1h: Slave PIC ELCR2--Slave Controller Edge/Level Triggered Register Description: View: IA F Size: 8 bit Bit Range Bus:Device:Function: B0:D31 :F0 BAR: 000h (IO) Default: 00h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access 07 IRQ7 ECL: 0 = Edge. 1 = Level. RW 06 IRQ6 ECL: 0 = Edge. 1 = Level. RW 05 IRQ5 ECL: 0 = Edge. 1 = Level. RW 04 IRQ4 ECL: 0 = Edge. 1 = Level. RW 03 IRQ3 ECL: 0 = Edge. 1 = Level. RW 02 :00 Reserved Reserved. Must be 0. Intel(R) Communications Chipset 89xx Series - Datasheet 286 Offset Start: 4D1h Offset End: 4D1h October 2012 Order Number: 327879-001US 6.0 6.4 Advanced Programmable Interrupt Controller (APIC) 6.4.1 APIC Register Map The APIC is accessed via an indirect addressing scheme. Two registers are visible by software for manipulation of most of the APIC registers. These registers are mapped into memory space. The address bits 19:12 of the address range are programmable through bits 7:0 of OIC register (Chipset Config Registers (RCBA); Offset 31FEh) The registers are shown in Table 6-63. 6.4.2 APIC Direct Registers Table 6-63. APIC Direct Registers Memory Address Note: 6.4.2.1 Register ID - Description Default Value FECx y000h "IND--Index Register" on page 287 00h FECx y010h "DAT--Data Register" on page 288 00000000h FECx y040h "EOIR--EOI Register" on page 288 N/A Memory address nibbles: x,y = bits[19:12] are programmable bits[7:0] of the OIC register IND--Index Register Memory Address FECx y000h Default Value: 00hSize: Attribute: 8 bits R/W The Index Register will select which APIC indirect register to be manipulated by software. The selector values for the indirect registers are listed in Table 6-67. Software will program this register to select the desired APIC internal register . Table 6-64. IND--Index Register Description: View: IA F Size: 8 bit Bit Range Memory FECx y000h Address B0:D31 Bus:Device:Function: :F0 Default: 00h Bit Acronym 07 :00 October 2012 Order Number: 327879-001US Offset Start: y000h Offset End: y000h Power Well: Bit Description Sticky APIC Index -- This is a 8-bit Index to select which APIC indirect register (Table 6-67) to write or read Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 287 6.4.2.2 DAT--Data Register Memory Address FECx y010h Default Value: 00000000h Attribute: Size: R/W 32 bits This is a 32-bit register specifying the data to be read or written to the register pointed to by the Index register. This register can only be accessed in DWord quantities. Table 6-65. DAT--Data Register Description: View: IA F Size: 32 bit Bit Range 31 :00 6.4.2.3 Memory FECx y010h Address B0:D31 Bus:Device:Function: :F0 Default: 00000000h Bit Acronym Offset Start: y010h Offset End: y013h Power Well: Bit Description Sticky Bit Reset Value APIC Data -- This is a 32-bit register for the data to be read or written to the APIC indirect register (Table 6-67) pointed to by the Index register (Memory Address FEC0 0000h, if OIC Register Bits[7:0] = 00h) Bit Access RW EOIR--EOI Register Memory Address FECx y040h Default Value: N/A Attribute: Size: R/W 32 bits The EOI register is present to provide a mechanism to maintain the level triggered semantics for level-triggered interrupts issued on the parallel bus. When a write is issued to this register, the I/O APIC will check the lower 8 bits written to this register, and compare it with the vector field for each entry in the I/O Redirection Table. When a match is found, the Remote_IRR bit (Index Offset 10h, bit 14) for that I/O Redirection Entry will be cleared. Note: If multiple I/O Redirection entries, for any reason, assign the same vector for more than one interrupt input, each of those entries will have the Remote_IRR bit reset to 0. The interrupt which was prematurely reset will not be lost because if its input remained active when the Remote_IRR bit is cleared, the interrupt will be reissued and serviced at a later time. Only bits 7:0 are used. Bits 31:8 are ignored by the PCH. Note: To provide for future expansion, the processor should always write a value of 0 to Bits 31:8. Intel(R) Communications Chipset 89xx Series - Datasheet 288 October 2012 Order Number: 327879-001US 6.0 Table 6-66. EOIR--EOI Register Description: View: IA F Size: 32 bit Bit Range 31 :08 Memory FECx y040h Address B0:D31 :F0 Default: N/A Offset Start: y040h Offset End: y043h Power Well: Bit Acronym Bit Description Sticky Bit Reset Value Bit Access Reserved. To provide for future expansion, the Reserved processor should always write a value of 0 to Bits 31:08. Redirection Entry Clear -- When a write is issued to this register, the I/O APIC will check this field, and compare it with the vector field for each entry in the I/O Redirection Table. When a match is found, the Remote_IRR bit for that I/O Redirection Entry will be cleared. 07 :00 6.4.3 Bus:Device:Function: WO APIC Indirect Registers Table 6-67 lists the registers that can be accessed within the APIC via the Index Register defined in Table 6-63. When accessing these registers, accesses must be done one dword at a time. For example, software should never access byte 2 from the Data register before accessing bytes 0 and 1. The hardware will not attempt to recover from a bad programming model in this case. Table 6-67. APIC Indirect Registers (Sheet 1 of 2) Index 00h 01h 02-0F Mnemonic & Register Name "Offset 00h: ID--Identification Register" on page 290 "Offset 01h: VER--Version Register" on page 291 Size Type 32 bits R/W 32 bits RO Reserved RO 10h-11h "Offset 10h: REDIR_TBL0--Redirection Table 0" on page 291 64 bits R/W, RO 12h-13h REDIR_TBTL1 - Redirection Table 1 64 bits R/W, RO 14h-15h REDIR_TBTL2 - Redirection Table 2 64 bits R/W, RO 16h-17h REDIR_TBTL3 - Redirection Table 3 64 bits R/W, RO 18h-19h REDIR_TBTL4 - Redirection Table 4 64 bits R/W, RO 1Ah-1Bh REDIR_TBTL5 - Redirection Table 5 64 bits R/W, RO 1Ch-1Dh REDIR_TBTL6 - Redirection Table 6 64 bits R/W, RO 1Eh-1Fh REDIR_TBTL7 - Redirection Table 7 64 bits R/W, RO 20h-21h REDIR_TBTL8 - Redirection Table 8 64 bits R/W, RO 22h-23h REDIR_TBTL9 - Redirection Table 9 64 bits R/W, RO 24h-25h REDIR_TBTL10 - Redirection Table 10 64 bits R/W, RO 26h-27h REDIR_TBTL11 - Redirection Table 11 64 bits R/W, RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 289 Table 6-67. APIC Indirect Registers (Sheet 2 of 2) Index 6.4.3.1 Mnemonic & Register Name Size Type 28h-29h REDIR_TBTL12 - Redirection Table 12 64 bits R/W, RO 2Ah-2Bh REDIR_TBTL13 - Redirection Table 13 64 bits R/W, RO 2Ch-2Dh REDIR_TBTL14 - Redirection Table 14 64 bits R/W, RO 2Eh-2Fh REDIR_TBTL15 - Redirection Table 15 64 bits R/W, RO 30h-31h REDIR_TBTL16 - Redirection Table 16 64 bits R/W, RO 32h-33h REDIR_TBTL17 - Redirection Table 17 64 bits R/W, RO 34h-35h REDIR_TBTL18 - Redirection Table 18 64 bits R/W, RO 36h-37h REDIR_TBTL19 - Redirection Table 19 64 bits R/W, RO 38h-39h REDIR_TBTL20 - Redirection Table 20 64 bits R/W, RO 3Ah-3Bh REDIR_TBTL21 - Redirection Table 21 64 bits R/W, RO 3Ch-3Dh REDIR_TBTL22 - Redirection Table 22 64 bits R/W, RO 3Eh-3Fh REDIR_TBTL23 - Redirection Table 23 64 bits R/W, RO 40h-FFh Reserved Offset 00h: ID--Identification Register Index Offset: Default Value: 00h 00000000h Attribute: Size: R/W 32 bits The APIC ID serves as a physical name of the APIC. The APIC bus arbitration ID for the APIC is derived from its I/O APIC ID. This register is reset to 0 on power-up reset. Table 6-68. Offset 00h: ID--Identification Register Description: View: IA I Size: 32 bit Bit Range 31 :28 27 :24 23 :16 15 14 :00 Win:Idx: APIC_WDW:APIC_IDX Default: 00000000h Bit Acronym Offset Start: 00h Offset End: 00h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved APIC ID -- This field contains the APIC Identification. Software must program this value before using the APIC. RW Reserved Reserved Scratchpad Bit. RW Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 290 B0:D31 Bus:Device:Function: :F0 October 2012 Order Number: 327879-001US 6.0 6.4.3.2 Offset 01h: VER--Version Register Index Offset: Default Value: 01h 00170020h Attribute: Size: RO, RWO 32 bits Each I/O APIC contains a hardwired Version Register that identifies different implementation of APIC and their versions. The maximum redirection entry information also is in this register, to let software know how many interrupt are supported by this APIC. Table 6-69. Offset 01h: VER--Version Register Description: View: IA I Size: 32 bit Bit Range 31 :24 23 :16 15 14 :08 07 :00 6.4.3.3 Win:Idx: APIC_WDW:APIC_IDX Bus:Device:Function: B0:D31 :F0 Default: 00170020h Bit Acronym Offset Start: 01h Offset End: 01h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved MRE Maximum Redirection Entries -- This is the entry number (0 being the lowest entry) of the highest entry in the redirection table. It is equal to the number of interrupt input pins minus one and is in the range 0 through 239. In the PCH this field is hardwired to 17h to indicate 24 interrupts. BIOS must write to this field after PLTRST# to lockdown the value. this allows BIOS to utilize some of the entries for its own purpose and thus advertising fewer IOxAPIC Redirection Entries to the OS. RWO PRQ Pin Assertion Register Supported -- Indicate that the IOxAPIC does not implement the Pin Assertion Register. RO Reserved Reserved VS Version -- This is a version number that identifies the implementation version. RO Offset 10h: REDIR_TBL0--Redirection Table 0 Index Offset: Default Value: 10h-11h (vector 0) 12h-13h (vector 1) ----3E-3Fh (vector 23) Bit 16 = 1. All other bits undefined Attribute:R/W, RO Size: 64 bits each, (accessed as two 32 bit quantities) The Redirection Table has a dedicated entry for each interrupt input pin. The information in the Redirection Table is used to translate the interrupt manifestation on the corresponding interrupt pin into an APIC message. The APIC will respond to an edge triggered interrupt as long as the interrupt is held until after the acknowledge cycle has begun. Once the interrupt is detected, a delivery status bit internally to the I/O APIC is set. The state machine will step ahead and wait for an acknowledgment from the APIC unit that the interrupt message was sent. Only then will the I/O APIC be able to recognize a new edge on that interrupt pin. That new October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 291 edge will only result in a new invocation of the handler if its acceptance by the destination APIC causes the Interrupt Request Register bit to go from 0 to 1. (In other words, if the interrupt was not already pending at the destination.) Table 6-70. Offset 10h: REDIR_TBL0--Redirection Table 0 (Sheet 1 of 2) Description: View: IA I Size: 64 bits Win:Idx: APIC_WDW:APIC_IDX Default: B0:D31 Offset Start: 10h :F0 Offset End: 11h Bit 16 = 1. All other bits undefined Power Well: Bit Acronym Bit Description 63 :56 DEST Destination -- If bit 11 of this entry is 0 (Physical), then bits 59:56 specifies an APIC ID. In this case, bits 63:59 should be programmed by software to 0. If bit 11 of this entry is 1 (Logical), then bits 63:56 specify the logical destination address of a set of processors. RW 55 :48 EDID Extended Destination ID -- These bits are sent to a local APIC only when in Processor System Bus mode. They become bits 11:4 of the address. RO 47 :17 Reserved Bit Range Sticky Bit Reset Value Bit Access Reserved Mask: 0 = Not masked: An edge or level on this interrupt pin results in the delivery of the interrupt to the destination. 1 = Masked: Interrupts are not delivered nor held pending. Setting this bit after the interrupt is accepted by a local APIC has no effect on that interrupt. This behavior is identical to the device withdrawing the interrupt before it is posted to the processor. It is software's responsibility to deal with the case where the mask bit is set after the interrupt message has been accepted by a local APIC unit but before the interrupt is dispensed to the processor. RW 15 Trigger Mode -- This field indicates the type of signal on the interrupt pin that triggers an interrupt. TRIGMOD 0 = Edge triggered. 1 = Level triggered. RW 14 Remote IRR -- This bit is used for level triggered interrupts; its meaning is undefined for edge triggered interrupts. 0 = Reset when an EOI message is received from a local APIC. 1 = Set when Local APIC/s accept the level interrupt sent by the I/O APIC. RO 13 INTPOL Interrupt Input Pin Polarity -- This bit specifies the polarity of each interrupt signal connected to the interrupt pins. 0 =Active high. 1 = Active low. RW DELIVS Delivery Status -- This field contains the current status of the delivery of this interrupt. Writes to this bit have no effect. 0 = Idle. No activity for this interrupt. 1 = Pending. Interrupt has been injected, but delivery is not complete. RO 16 12 MASK Intel(R) Communications Chipset 89xx Series - Datasheet 292 Bus:Device:Function: October 2012 Order Number: 327879-001US 6.0 Table 6-70. Offset 10h: REDIR_TBL0--Redirection Table 0 (Sheet 2 of 2) Description: View: IA I Size: 64 bits Bit Range 11 Note: Win:Idx: APIC_WDW:APIC_IDX Default: Bus:Device:Function: B0:D31 Offset Start: 10h :F0 Offset End: 11h Bit 16 = 1. All other bits undefined Bit Acronym Power Well: Bit Description Sticky Destination Mode -- This field determines the interpretation of the Destination field. 0 = Physical. Destination APIC ID is identified by bits DESTMOD 59:56. 1 = Logical. Destinations are identified by matching bit 63:56 with the Logical Destination in the Destination Format Register and Logical Destination Register in each Local APIC 10 :08 DELMOD Delivery Mode -- This field specifies how the APICs listed in the destination field should act upon reception of this signal. Certain Delivery Modes will only operate as intended when used in conjunction with a specific trigger mode. These encodings are listed in the note below. 07 :00 INTVEC Interrupt Vector -- This field contains the interrupt vector for this interrupt. Values range between 10h and FEh. Bit Reset Value Bit Access RW RW RW Bits[10:8] - Delivery Mode Encoding 000 = Fixed. Deliver the signal on the INTR signal of all processor cores listed in the destination. Trigger Mode can be edge or level. 001 = Lowest Priority. Deliver the signal on the INTR signal of the processor core that is executing at the lowest priority among all the processors listed in the specified destination. Trigger Mode can be edge or level. 010 = SMI (System Management Interrupt). Requires the interrupt to be programmed as edge triggered. The vector information is ignored but must be programmed to all 0s for future compatibility: not supported 011 = Reserved 100 = NMI. Deliver the signal on the NMI signal of all processor cores listed in the destination. Vector information is ignored. NMI is treated as an edge triggered interrupt even if it is programmed as level triggered. For proper operation this redirection table entry must be programmed to edge triggered. The NMI delivery mode does not set the RIRR bit. If the redirection table is incorrectly set to level, the loop count will continue counting through the redirection table addresses. Once the count for the NMI pin is reached again, the interrupt will be sent again: not supported 101 = INIT. Deliver the signal to all processor cores listed in the destination by asserting the INIT signal. All addressed local APICs will assume their INIT state. INIT is always treated as an edge triggered interrupt even if programmed as level triggered. For proper operation this redirection table entry must be programmed to edge triggered. The INIT delivery mode does not set the RIRR bit. If the redirection table is incorrectly set to level, the loop count will continue counting through the redirection table addresses. Once the count for the INIT pin is reached again, the interrupt will be sent again: not supported 110 = Reserved 111 = ExtINT. Deliver the signal to the INTR signal of all processor cores listed in the destination as an interrupt that originated in an externally connected 8259A compatible interrupt controller. The INTA cycle that corresponds to this ExtINT delivery will be routed to the external controller that is expected to supply the vector. Requires the interrupt to be programmed as edge triggered. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 293 6.5 Real Time Clock Registers 6.5.1 I/O Register Address Map The RTC internal registers and RAM are organized as two banks of 128 bytes each, called the standard and extended banks. The first 14 bytes of the standard bank contain the RTC time and date information along with four registers, A-D, that are used for configuration of the RTC. The extended bank contains a full 128 bytes of battery backed SRAM, and will be accessible even when the RTC module is disabled (via the RTC configuration register). Registers A-D do not physically exist in the RAM. All data movement between the host processor and the real-time clock is done through registers mapped to the standard I/O space. The register map appears in Table 6-71. Table 6-71. RTC I/O Registers I/O Locations If U128E bit = 0 70h and 74h Also alias to 72h and 76h Real-Time Clock (Standard RAM) Index Register 71h and 75h Also alias to 73h and 77h Real-Time Clock (Standard RAM) Target Register Function 72h and 76h Extended RAM Index Register (if enabled) 73h and 77h Extended RAM Target Register (if enabled) Notes: 1. I/O locations 70h and 71h are the standard legacy location for the real-time clock. The map for this bank is shown in Table 6-72. Locations 72h and 73h are for accessing the extended RAM. The extended RAM bank is also accessed using an indexed scheme. I/O address 72h is used as the address pointer and I/O address 73h is used as the data register. Index addresses above 127h are not valid. If the extended RAM is not needed, it may be disabled. 2. Software must preserve the value of bit 7 at I/O addresses 70h and 74h. When writing to this address, software must first read the value, and then write the same value for bit 7 during the sequential address write. Port 70h is not directly readable. The only way to read this register is through Alt Access mode. Although RTC Index bits 6:0 are readable from port 74h, bit 7 will always return 0. If the NMI# enable is not changed during normal operation, software can alternatively read this bit once and then retain the value for all subsequent writes to port 70h. Intel(R) Communications Chipset 89xx Series - Datasheet 294 October 2012 Order Number: 327879-001US 6.0 6.5.2 Indexed Registers The RTC contains two sets of indexed registers that are accessed using the two separate Index and Target registers (70/71h or 72/73h), as shown in Table 6-72. Table 6-72. RTC (Standard) RAM Bank Index Name 00h Seconds 01h Seconds Alarm 02h Minutes 03h Minutes Alarm 04h Hours 05h Hours Alarm 06h Day of Week 07h Day of Month 08h Month 09h Year 0Ah Register A 0Bh Register B 0Ch Register C 0Dh Register D 0Eh-7Fh 114 Bytes of User RAM Table 6-73. Real Time Clock Registers Offset Start 0Ah Offset End 0Ah Register ID - Description "Offset 0Ah: RTC_REGA--Register A" on page 296 Default Value Undefined 0Bh 0Bh "Offset 0Bh: RTC_REGB--Register B (General Configuration)" on page 297 X0X00XXX 0Ch 0Ch "Offset OCh: RTC_REGC--Register C (Flag Register)" on page 298 00X00000 0Dh 0Dh "Offset ODh: D--Register D (Flag Register)" on page 299 10XXXXXX October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 295 6.5.2.1 Offset 0Ah: RTC_REGA--Register A This register is used for general configuration of the RTC functions. None of the bits are affected by RSMRST# or any other PCH reset signal. Table 6-74. Offset 0Ah: RTC_REGA--Register A Description: View: IA F Size: 8 bit BAR: RTC Standard RAM Bank B0:D31 :F0 Default: Undefined Offset Start: 0Ah Offset End: 0Ah Power Well: RTC Bit Reset Value Bit Acronym Bit Description UIP Update In Progress -- This bit may be monitored as a status flag. 0 =The update cycle will not start for at least 488 s. The time, calendar, and alarm information in RAM is always available when the UIP bit is 0. 1 = The update is soon to occur or is in progress. RW 06 :04 Division Chain Select -- These three bits control the divider chain for the oscillator, and are not affected by RSMRST# or any other reset signal. 010 = Normal Operation 11X = Divider Reset DV[2:0] 101 = Bypass 15 stages (test mode only) 100 = Bypass 10 stages (test mode only) 011 = Bypass 5 stages (test mode only) 001 = Invalid 000 = Invalid RW 03 :00 Rate Select -- Selects one of 13 taps of the 15 stage divider chain. The selected tap can generate a periodic interrupt if the PIE bit is set in Register B. Otherwise this tap will set the PF flag of Register C. If the periodic interrupt is not to be used, these bits should all be set to 0. RS3 corresponds to bit 3. 0000 = Interrupt never toggles 0001 = 3.90625 ms 0010 = 7.8125 ms 0011 = 122.070 s 0100 = 244.141 s 0101 = 488.281 s 0110 = 976.5625 s 0111 = 1.953125 ms 1000 = 3.90625 ms 1001 = 7.8125 ms 1010 = 15.625 ms 1011 = 31.25 ms 1100 = 62.5 ms 1101 = 125 ms 1110 = 250 ms 1111= 500 ms RW Bit Range 07 :00 RS[3:0] Intel(R) Communications Chipset 89xx Series - Datasheet 296 Bus:Device:Function: Sticky Bit Access October 2012 Order Number: 327879-001US 6.0 6.5.2.2 Offset 0Bh: RTC_REGB--Register B (General Configuration) Table 6-75. Offset 0Bh: RTC_REGB--Register B (General Configuration) Description: View: IA F Size: 8 bit Bit Range 07 BAR: RTC Standard RAM Bank Bus:Device:Function: B0:D31 :F0 Default: X0X00XXX Bit Acronym SET Offset Start: 0Bh Offset End: 0Bh Power Well: RTC Bit Description Sticky Update Cycle Inhibit -- Enables/Inhibits the update cycles. This bit is not affected by RSMRST# nor any other reset signal. 0 = Update cycle occurs normally once each second. 1 = A current update cycle will abort and subsequent update cycles will not occur until SET is returned to 0. When set is one, the BIOS may initialize time and calendar bytes safely. Bit Reset Value Bit Access RW This bit should be set then cleared early in BIOS POST after each powerup directly after coin-cell battery insertion. 06 05 04 03 02 01 00 PIE Periodic Interrupt Enable -- This bit is cleared by RSMRST#, but not on any other reset. 0 = Disable. 1 = Enable. Allows an interrupt to occur with a time base set with the RS bits of register A. RW AIE Alarm Interrupt Enable -- This bit is cleared by RTCRST#, but not on any other reset. 0 = Disable. 1 = Enable. Allows an interrupt to occur when the AF is set by an alarm match from the update cycle. An alarm can occur once a second, one an hour, once a day, or one a month. RW UIE Update-Ended Interrupt Enable -- This bit is cleared by RSMRST#, but not on any other reset. 0 = Disable. 1 = Enable. Allows an interrupt to occur when the update cycle ends. RW SQWE Square Wave Enable -- This bit serves no function in the PCH. It is left in this register bank to provide compatibility with the Motorola 146818B. The PCH has no SQW pin. This bit is cleared by RSMRST#, but not on any other reset. RW DM Data Mode -- This bit specifies either binary or BCD data representation. This bit is not affected by RSMRST# nor any other reset signal. 0 = BCD 1 = Binary RW Hour Format --This bit indicates the hour byte format. This bit is not affected by RSMRST# nor any other reset signal. HOURFORM 0 = Twelve-hour mode. In twelve-hour mode, the seventh bit represents AM as 0 and PM as one. 1 = Twenty-four hour mode. RW DSLSWS October 2012 Order Number: 327879-001US Daylight Savings Legacy Software Support -- Daylight savings functionality is no longer supported. This bit is used to maintain legacy software support and has no associated functionality. If BUC.DSO bit is set, the DSLSWS bit continues to be R/W. RW Intel(R) Communications Chipset 89xx Series - Datasheet 297 6.5.2.3 Offset OCh: RTC_REGC--Register C (Flag Register) Writes to Register C have no effect. Table 6-76. Offset OCh: RTC_REGC--Register C (Flag Register) Description: View: IA F Size: 8 bit Bit Range 07 06 05 04 03 :00 BAR: RTC Standard RAM Bank B0:D31 :F0 Default: 00X00000 Bit Acronym Offset Start: 0Ch Offset End: 0Ch Power Well: RTC Bit Description Sticky Bit Reset Value Bit Access IRQF Interrupt Request Flag -- IRQF = (PF * PIE) + (AF * AIE) + (UF *UFE). This bit also causes the RTC Interrupt to be asserted. This bit is cleared upon RSMRST# or a read of Register C. RO PF Periodic Interrupt Flag -- This bit is cleared upon RSMRST# or a read of Register C. 0 = If no taps are specified via the RS bits in Register A, this flag will not be set. 1 = Periodic interrupt Flag will be 1 when the tap specified by the RS bits of register A is 1. RO AF Alarm Flag: 0 = This bit is cleared upon RTCRST# or a read of Register C. 1 = Alarm Flag will be set after all Alarm values match the current time. RO UF Update-Ended Flag: 0 = The bit is cleared upon RSMRST# or a read of Register C. 1 = Set immediately following an update cycle for each second. RO Reserved Reserved. Will always report 0. Intel(R) Communications Chipset 89xx Series - Datasheet 298 Bus:Device:Function: October 2012 Order Number: 327879-001US 6.0 6.5.2.4 Offset ODh: D--Register D (Flag Register) Table 6-77. Offset ODh: D--Register D (Flag Register) Description: View: IA F Size: 8 bit Bit Range 07 06 BAR: RTC Standard RAM Bank Bus:Device:Function: B0:D31 :F0 Default: 10XXXXXX Offset Start: 0Dh Offset End: 0Dh Power Well: RTC Bit Acronym Bit Description VRT Valid RAM and Time Bit: 0 = This bit should always be written as a 0 for write cycle, however it will return a 1 for read cycles. 1 = This bit is hardwired to 1 in the RTC power well. Sticky Bit Reset Value Bit Access RW Reserved Reserved. This bit always returns a 0 and should be set to 0 for write cycles. 05 :00 October 2012 Order Number: 327879-001US Date Alarm -- These bits store the date of month alarm value. If set to 000000b, then a don't care state is assumed. The host must configure the date alarm for these bits to do anything, yet they can be written at any time. If the date alarm is not enabled, these bits will return 0s to mimic the functionality of the Motorola 146818B. These bits are not affected by any reset assertion. RW Intel(R) Communications Chipset 89xx Series - Datasheet 299 6.6 Processor Interface Registers (LPC I/F--B0:D31:F0) 6.6.1 Processor Interface PCI Registers Address Map The following table provides he register address map for the processor interface registers. Table 6-78. Processor Interface PCI Register Address Map (LPC I/F--B0:D31:F0) Offset Start Offset End Default Value Register ID - Description 61h 61h "Offset 61h: NMI_SC--NMI Status and Control Register (LPC I/F--B0:D31:F0)" on 00h page 301 70h 70h "Offset 70h: NMI_EN--NMI Enable (and Real Time Clock Index) Register (LPC I/F-- 80h B0:D31:F0)" on page 302 92h 92h "Offset 92h: PORT92--Fast A20 and Init Register (LPC I/F--B0:D31:F0)" on page 302 F0h F0h "Offset F0h: COPROC_ERR--Coprocessor Error Register (LPC I/F--B0:D31:F0)" on 00h page 303 CF9h CF9h "Offset CF9h: RST_CNT--Reset Control Register (LPC I/F--B0:D31:F0)" on page 303 00h B2h B2h "Offset B2h: APM_CNT--Advanced Power Management Control Port Register" on page 316 00h B3h B3h "Offset B3h: APM_STS--Advanced Power Management Status Port Register" on page 316 00h Intel(R) Communications Chipset 89xx Series - Datasheet 300 00h October 2012 Order Number: 327879-001US 6.0 6.6.1.1 Offset 61h: NMI_SC--NMI Status and Control Register (LPC I/F-- B0:D31:F0) Table 6-79. Offset 61h: NMI_SC--NMI Status and Control Register (LPC I/F--B0:D31:F0) Description : View IA F : Size: 8 bit Bit Range B0:D31 Bus:Device:Function: :F0 BAR: 0000h (IO) Default: 00h Offset Start: 61h Offset End: 61h Power Well: Core Bit Description SERR#_NMI_STS SERR# NMI Source Status: 1 = Bit is set if a PCI agent detected a system error and pulses the PCI SERR# line and if bit 2 (PCI_SERR_EN) is cleared. This interrupt source is enabled by setting bit 2 to 0. To reset the interrupt, set bit 2 to 1 and then set it to 0. When writing to port 61h, this bit must be 0. Note: This bit is set by any of the PCH internal sources of SERR; this includes SERR assertions forwarded from the secondary PCI bus, errors on a PCI Express* port, or other internal functions that generate SERR#. RO IOCHK_NMI_STS IOCHK# NMI Source Status: 1 = Bit is set if an LPC agent (via SERIRQ) asserted IOCHK# and if bit 3 (IOCHK_NMI_EN) is cleared. This interrupt source is enabled by setting bit 3 to 0. To reset the interrupt, set bit 3 to 1 and then set it to 0. When writing to port 61h, this bit must be a 0. RO 05 TMR2_OUT_STS Timer Counter 2 OUT Status: This bit reflects the current state of the 8254 counter 2 output. Counter 2 must be programmed following any PCI reset for this bit to have a determinate value. When writing to port 61h, this bit must be a 0. RO 04 REF_TOGGLE Refresh Cycle Toggle: This signal toggles from either 0 to 1 or 1 to 0 at a rate that is equivalent to when refresh cycles would occur. When writing to port 61h, this bit must be a 0. RO 03 IOCHK_NMI_EN IOCHK# NMI Enable: 0 = Enabled. 1 = Disabled and cleared. RW 02 PCI_SERR_EN PCI SERR# Enable: 0 = SERR# NMIs are enabled. 1 = SERR# NMIs are disabled and cleared. RW 01 SPKR_DAT_EN Speaker Data Enable: 0 = SPKR output is a 0. 1 = SPKR output is equivalent to the Counter 2 OUT signal value. RW 00 TIM_CNT2_EN Timer Counter 2 Enable: 0 = Disable 1 = Enable. RW 07 06 October 2012 Order Number: 327879-001US Sticky Bit Reset Value Bit Acronym Bit Access Intel(R) Communications Chipset 89xx Series - Datasheet 301 6.6.1.2 Offset 70h: NMI_EN--NMI Enable (and Real Time Clock Index) Register (LPC I/F--B0:D31:F0) Note: The RTC Index field is write-only for normal operation. This field can only be read in AltAccess Mode. This register is aliased to Port 74h (documented in Table 6-71), and all bits are readable at that address. Table 6-80. Offset 70h: NMI_EN--NMI Enable (and Real Time Clock Index) Register (LPC I/F--B0:D31:F0) Description: View: IA F Size: 8 bit Bit Range 07 06 :00 6.6.1.3 BAR: 0000h (IO) Bus:Device:Function: B0:D31 :F0 Default: 80h Bit Acronym NMI_EN Offset Start: 70h Offset End: 70h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access NMI Enable (special): 0 = Enable NMI sources. 1 = Disable All NMI sources. RW Real Time Clock Index Address (special): RTC_INDX This data goes to the RTC to select which register or CMOS RAM address is being accessed. RW Offset 92h: PORT92--Fast A20 and Init Register (LPC I/F--B0:D31:F0) Table 6-81. Offset 92h: PORT92--Fast A20 and Init Register (LPC I/F--B0:D31:F0) Description: View: IA F Size: 8 bit Bus:Device:Function: B0:D31 :F0 BAR: 0000h (IO) Default: 00h Bit Range Bit Acronym 07 :02 Reserved 01 ALT_A20_GATE 00 INIT_NOW Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved Alternate A20 Gate -- This bit is Or'd with the A20GATE input signal to generate A20M# to the processor. 0 = A20M# signal can potentially go active. 1 = This bit is set when INIT# goes active. RW When this bit transitions from a 0 to a 1, the PCH will force INIT# active for 16 PCI clocks. RW Intel(R) Communications Chipset 89xx Series - Datasheet 302 Offset Start: 92h Offset End: 92h October 2012 Order Number: 327879-001US 6.0 6.6.1.4 Offset F0h: COPROC_ERR--Coprocessor Error Register (LPC I/F-- B0:D31:F0) Table 6-82. Offset F0h: COPROC_ERR--Coprocessor Error Register (LPC I/F--B0:D31:F0) Description: View: IA F Size: 8 bit BAR: 0000h (IO) Bus:Device:Function: Default: 00h Offset Start: F0h Offset End: F0h Power Well: Core Bit Range Bit Acronym Bit Description 07 :00 COPROC_ERR Coprocessor Error -- Any value written to this register will cause IGNNE# to go active, if FERR# had generated an internal IRQ13. For FERR# to generate an internal IRQ13, the COPROC_ERR_EN bit must be 1. 6.6.1.5 B0:D31 :F0 Sticky Bit Reset Value Bit Access WO Offset CF9h: RST_CNT--Reset Control Register (LPC I/F--B0:D31:F0) Table 6-83. Offset CF9h: RST_CNT--Reset Control Register (LPC I/F--B0:D31:F0) Description: View: IA F Size: 8 bit BAR: 0000h (IO) Default: 00h Bit Range Bit Acronym 07 :04 Reserved 03 Bus:Device:Function: B0:D31 :F0 FULL_RST Offset Start: CF9h Offset End: CF9h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved Full Reset -- This bit is used to determine the states of SLP_S3#, SLP_S4#, and SLP_S5# after a CF9 hard reset (SYS_RST =1 and RST_CPU is set to 1), after PWROK going low (with RSMRST# high), or after two TCO timeouts. 0 = PCH will keep SLP_S3#, SLP_S4# and SLP_S5# high. 1 = PCH will drive SLP_S3#, SLP_S4# and SLP_S5# low for 3 - 5 seconds. RW Note: When this bit is set, it also causes the full power cycle (SLP_S3/4/5# assertion) in response to SYS_RESET#, PWROK#, and Watchdog timer reset sources. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 303 Table 6-83. Offset CF9h: RST_CNT--Reset Control Register (LPC I/F--B0:D31:F0) Description: View: IA F Size: 8 bit Bit Range BAR: 0000h (IO) Bus:Device:Function: Default: 00h Offset Start: CF9h Offset End: CF9h Power Well: Core Bit Description RST_CPU Reset CPU -- When this bit transitions from a 0 to a 1, it initiates a hard or soft reset, as determined by the SYS_RST bit (bit 1 of this register). RW 01 SYS_RST System Reset -- This bit is used to determine a hard or soft reset to the processor. 0 = When RST_CPU bit goes from 0 to 1, the PCH performs a soft reset by activating INIT# for 16 PCI clocks. 1 = When RST_CPU bit goes from 0 to 1, the PCH performs a hard reset by activating PLTRST# and SUS_STAT# active for a minimum of about 1 milliseconds. In this case, SLP_S3#, SLP_S4# and SLP_S5# state (assertion or de-assertion) depends on FULL_RST bit setting. The PCH main power well is reset when this bit is 1. It also resets the resume well bits (except for those noted throughout the EDS). RW 00 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet Sticky Bit Reset Value Bit Acronym 02 304 B0:D31 :F0 Bit Access October 2012 Order Number: 327879-001US 6.0 6.7 Power Management Registers (PM--B0:D31:F0) The power management registers are distributed within the PCI Bus 0:Device 31: Function 0 space, as well as a separate I/O range. Each register is described below. Unless otherwise indicate, bits are in the main (core) power well. Bits not explicitly defined in each register are assumed to be reserved. When writing to a reserved bit, the value should always be 0. Software should not attempt to use the value read from a reserved bit, as it may not be consistently 1 or 0. 6.7.1 Power Management PCI Configuration Registers (PM-- B0:D31:F0) Table 6-84 shows a small part of the configuration space for PCI Bus 0:Device 31: Function 0. It includes only those registers dedicated for power management. Some of the registers are only used for legacy power management schemes. Table 6-84. Power Management PCI Register Address Map (PM--B0:D31:F0) Offset Start Offset End Register ID - Description Default Value A0h A0h "Offset A0h: GEN_PMCON_1--General PM Configuration 1 Register (PM-- B0:D31:F0)" on page 306 0000h A2h A2h "Offset A2h: GEN_PMCON_2--General PM Configuration 2 Register (PM-- B0:D31:F0)" on page 308 00h A4h A4h "Offset A4h: GEN_PMCON_3--General PM Configuration 3 Register (PM-- B0:D31:F0)" on page 310 00h A6h A6h "Offset A6h: GEN_PMCON_LOCK- General Power Management Configuration Lock Register" on page 313 00h A9h A9h "Offset A9h: Chipset Initialization Register 4 (PM--B0:D31:F0)" on page 313 01h ABh ABh "Offset ABh: BM_BREAK_EN Register (PM--B0:D31:F0)" on page 314 01h ACh ACh "Offset ACh: PMIR--Power Management Initialization Register (PM--B0:D31:F0)" on page 314 00000000h B8h BBh "Offset B8h: GPIO_ROUT--GPIO Routing Control Register (PM--B0:D31:F0)" on page 315 00000000h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 305 6.7.1.1 Offset A0h: GEN_PMCON_1--General PM Configuration 1 Register (PM--B0:D31:F0) Table 6-85. Offset A0h: GEN_PMCON_1--General PM Configuration 1 Register (PM-- B0:D31:F0) (Sheet 1 of 2) Description: View: PCI Size: 16 bit Bit Range 15 :11 BAR: Configuration B0:D31 Offset Start: A0h :F0 Offset End: A0h Default: 0000h Bit Acronym Reserved 10 BIOS_PCI_EXP_EN 09 PWRBTN_LVL 08 :05 Bus:Device:Function: Reserved 04 SMI_LOCK 03 Reserved Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved This bit acts as a global enable for the SCI associated with the PCI Express* ports. 0 = The various PCI Express* ports and Processor cannot cause the PCI_EXP_STS bit to go active. 1 = The various PCI Express* ports and Processor can cause the PCI_EXP_STS bit to go active. RW This bit indicates the current state of the PWRBTN# signal. 0 = Low. 1 = High. RO Reserved When this bit is set, writes to the GLB_SMI_EN bit (PMBASE + 30h, bit 0) will have no effect. Once the SMI_LOCK bit is set, writes of 0 to SMI_LOCK bit will have no effect (i.e., once set, this bit can only be cleared by PLTRST#). RWO Reserved 0 = Disable. 1 = Enable internal CLKRUN# logic to allow DMI PLL shutdown. This bit has no impact on state of external CLKRUN# pin. 03 (server only) Notes: PSEUDO_CLKRUN_EN 1. PSEUDO_CLKRUN_EN bit does not result in STP_PCI# assertion to actually stop the external PCICLK. 2. This bit should be set mutually exclusive with the CLKRUN_EN bit. Setting PSEUDO_CLKRUN_EN in a mobile sku could result in unspecified behavior. Intel(R) Communications Chipset 89xx Series - Datasheet 306 RW October 2012 Order Number: 327879-001US 6.0 Table 6-85. Offset A0h: GEN_PMCON_1--General PM Configuration 1 Register (PM-- B0:D31:F0) (Sheet 2 of 2) Description: View: PCI BAR: Configuration Size: 16 bit Bit Range Bus:Device:Function: B0:D31 Offset Start: A0h :F0 Offset End: A0h Default: 0000h Bit Acronym Power Well: Core Bit Description Sticky Bit Reset Value Bit Access PCI CLKRUN# Enable: 0 = Disable. Drives the CLKRUN# signal low. 1 = Enable CLKRUN# logic to control the system PCI clock via the CLKRUN# and STP_PCI# signals. 02 CLKRUN_EN 02 Reserved 01 :00 PER_SMI_SEL October 2012 Order Number: 327879-001US Notes: 1. When the SLP_EN# bit is set, the PCH drives the CLKRUN# signal low regardless of the state of the CLKRUN_EN bit. This ensures that the PCI and LPC clocks continue running during a transition to a sleep state. 2. This bit should be set mutually exclusive with the PSEUDO_CLKRUN_EN bit. Setting CLKRUN_EN in a non-mobile sku could result in unspecified behavior. RW Reserved Periodic SMI# Rate Select: Set by software to control the rate at which periodic SMI# is generated. 00 = 64 seconds 01 = 32 seconds 10 = 16 seconds 11 = 8 seconds RW Intel(R) Communications Chipset 89xx Series - Datasheet 307 6.7.1.2 Offset A2h: GEN_PMCON_2--General PM Configuration 2 Register (PM--B0:D31:F0) Table 6-86. Offset A2h: GEN_PMCON_2--General PM Configuration 2 Register (PM--B0:D31:F0) (Sheet 1 of 2) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration 05 B0:D31 :F0 Default: 00h Bit Acronym Offset Start: A2h Offset End: A2h Power Well: Resume Bit Description DRAM Initialization Bit -- This bit does not effect hardware functionality in any way. BIOS is expected to set this bit prior to starting the DRAM initialization sequence and to clear this bit after completing the DRAM initialization sequence. BIOS can detect that a DRAM initialization sequence was interrupted by a reset by reading this bit during the boot sequence. - If the bit is 1, then the DRAM initialization was interrupted. - This bit is reset by the assertion of the RSMRST# pin. 07 06 Bus:Device:Function: Reserved Reserved MEM_SR Memory Placed in Self-Refresh: - If the bit is 1, DRAM should have remained powered and held in Self-Refresh through the last power state transition (i.e. the last time the system left S0). - This bit is reset by the assertion of the RSMRST# pin. Sticky Bit Reset Value Bit Access RW RO System Reset Status -- Software clears this bit by writing a 1 to it. 0 = SYS_RESET# button Not pressed. 1 = PCH sets this bit when the SYS_RESET# button is pressed. BIOS is expected to read this bit and clear it, if it is set. 04 SRS Notes: 1. This bit is also reset by RSMRST# and CF9h resets. 2. The SYS_RESET# is implemented in the Main power well. This pin must be properly isolated and masked to prevent incorrectly setting this Suspend well status bit. RWC CPU Thermal Trip Status: 0 = Software clears this bit by writing a 1 to it. 1 = This bit is set when PLTRST# is inactive and THRMTRIP# goes active while the system is in an S0 or S1 state. 03 CTS Notes: 1. This bit is also reset by RSMRST#, and CF9h resets. It is not reset by the shutdown and reboot associated with the CPUTHRMTRIP# event. 2. The CF9h reset in the description refers to CF9h type core well reset which includes SYS_RST#, PWROK/SYS_PWROK low, SMBus hard reset, TCO Timeout. This type of reset will clear CTS bit. Intel(R) Communications Chipset 89xx Series - Datasheet 308 RWC October 2012 Order Number: 327879-001US 6.0 Table 6-86. Offset A2h: GEN_PMCON_2--General PM Configuration 2 Register (PM--B0:D31:F0) (Sheet 2 of 2) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration 00 B0:D31 :F0 Default: 00h Bit Acronym Reserved PWROK_FLR October 2012 Order Number: 327879-001US Offset Start: A2h Offset End: A2h Power Well: Resume Bit Description Sticky Minimum SLP_S4# Assertion Width Violation Status: 0 = Software clears this bit by writing a 1 to it. 1 = Hardware sets this bit when the SLP_S4# assertion width is less than the time programmed in the SLP_S4# Minimum Assertion Width field (B0:D31:F0:Offset A4h:bits 5:4). The PCH begins the timer when SLP_S4# is asserted during S4/S5 entry, or when the RSMRST# input is deasserted during G3 exit. NOTE: This bit is functional regardless of the value in the SLP_S4# Assertion Stretch Enable (B0:D31:F0:Offset A4h:bit 3). NOTE: This bit is reset by the assertion of the RSMRST# pin, but can be set in some cases before the default value is readable. 02 01 Bus:Device:Function: Bit Reset Value Bit Access RWC Reserved PWROK Failure: 0 = Software clears this bit by writing a 1 to it, or when the system goes into a G3 state. 1 = This bit will be set any time PWROK goes low, when the system was in S0, or S1 state. RWC Intel(R) Communications Chipset 89xx Series - Datasheet 309 6.7.1.3 Offset A4h: GEN_PMCON_3--General PM Configuration 3 Register (PM--B0:D31:F0) Table 6-87. Offset A4h: GEN_PMCON_3--General PM Configuration 3 Register (PM-- B0:D31:F0) (Sheet 1 of 3) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 00h Offset Start: A4h Offset End: A4h Power Well: RTC, SUS Bit Acronym Bit Description Sticky Bit Reset Value Bit Access PME B0 S5 Disable --When set to '1', this bit blocks wake events from PME_B0_STS in S5, regardless of the state of PME_B0_EN. When cleared (default), wake events from PME_B0_STS are allowed in S5 if PME_B0_EN = '1'. Wakes from power states other than S5 are not affected by this policy bit. The net effect of setting PME_B0_S5_DIS = '1' is described by the truth table below: Y = Wake; N = Don't wake; B0 = PME_B0_EN 15 PME_B0_S5_DIS B0/OV S1/S3/S4 S5 00 N N 01 N N 11 Y (all PME B0 sources) N 01 Y (all PME B0 sources) N RW This bit is cleared by the RTCRST# pin. 14 Reserved Reserved 13 Reserved Reserved 12 11 :10 Disable SLP_S4# Stretching after G3: 0 = Enables stretching on SLP_S4# in conjunction with SLP_S4# Assertion Stretch Enable (bit 3) and the Minimum Assertion Width (bits 5:4) 1 = Disables stretching on SLP_S4# regardless of the state of the SLP_S4# Assertion Stretch Enable (bit 3). This bit is cleared by the RTCRST# pin. Note: This field is RO when the SLP_Sx# Stretching Policy Lock- Down bit is set. RW SLP_S3# Minimum Assertion Width -- This 2-bit value indicates the minimum assertion width of the SLP_S3# signal to guarantee that the Main power supplies have been fully power-cycled. Valid Settings are: 00: 60-100 us 01: 1-1.2 ms 10: 50-50.2 ms 11: 2-2.0002 s This bit is cleared by the RSMRST# pin. This field is RO when the SLP_Sx# Stretching Policy Lock-Down bit is set. RW Intel(R) Communications Chipset 89xx Series - Datasheet 310 October 2012 Order Number: 327879-001US 6.0 Table 6-87. Offset A4h: GEN_PMCON_3--General PM Configuration 3 Register (PM-- B0:D31:F0) (Sheet 2 of 3) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: Default: 00h Bit Description 09 GEN_RST_STS General Reset Status -- This bit is set by hardware whenever PLTRST# asserts for any reason other than going into a software-entered sleep state (via PM1CNT.SLP_EN write) or a suspend well power failure (RSMRST# pin assertion). BIOS is expected to consult and then write a `1' to clear this bit during the boot flow before determining what action to take based on PM1_STS.WAK_STS = 1. If GEN_RST_STS = `1', the cold reset boot path should be followed rather than the resume path, regardless of the setting of WAK_STS. This bit is cleared by the RSMRST# pin. 08 Reserved SWSMI_RATE_SEL 05 :04 03 October 2012 Order Number: 327879-001US Offset Start: A4h Offset End: A4h Power Well: RTC, SUS Bit Acronym 07 :06 B0:D31 :F0 Sticky Bit Reset Value Bit Access RWC Reserved This field indicates when the SWSMI timer will time out. Valid values are: 00 = 1.5 ms 0.6 ms 01 = 16 ms 4 ms 10 = 32 ms 4 ms 11 = 64 ms 4 ms These bits are not cleared by any type of reset except RTCRST#. RW SLP_S4# Minimum Assertion Width -- This field indicates the minimum assertion width of the SLP_S4# signal to ensure that the DRAMs have been safely power-cycled. Valid values are: 11 = 1 second 10 = 2 seconds 01 = 3 seconds 00 = 4 seconds This value is used in two ways: 1. If the SLP_S4# assertion width is ever shorter than this time, a status bit is set for BIOS to read when S0 is entered. 2. If enabled by bit 3 in this register, the hardware will prevent the SLP_S4# signal from deasserting within this minimum time period after asserting. RTCRST# forces this field to the conservative default state (00b). This field is RO when the SLP_S4# Stretching Policy Lock-Down bit is set. RW/RO SLP_S4# Assertion Stretch Enable: 0 = The SLP_S4# minimum assertion time is 1 to 2 RTCCLK. 1 = The SLP_S4# signal minimally assert for the time specified in bits 5:4 of this register. This bit is cleared by RTCRST#. This bit is RO when the SLP_S4# Stretching Policy Lock-Down bit is set. RW/RO Intel(R) Communications Chipset 89xx Series - Datasheet 311 Table 6-87. Offset A4h: GEN_PMCON_3--General PM Configuration 3 Register (PM-- B0:D31:F0) (Sheet 3 of 3) Description: View: PCI Size: 16 bit Bit Range 02 01 BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 00h Power Well: RTC, SUS Bit Acronym Bit Description RTC_PWR_STS RTC Power Status -- This bit is set when RTCRST# indicates a weak or missing battery. The bit is not cleared by any type of reset. The bit will remain set until the software clears it by writing a 0 back to this bit position. PWR_FLR Offset Start: A4h Offset End: A4h Power Failure -- This bit is in the RTC well, and is not cleared by any type of reset except RTCRST#. 0 = Indicates that the trickle current has not failed since the last time the bit was cleared. Software clears this bit by writing a 1 to it. 1 = Indicates that the trickle current (from the main battery or trickle supply) was removed or failed. Sticky Bit Reset Value Bit Access RW RWC Clearing CMOS in a PCH-based platform can be done by using a jumper on RTCRST# or GPI, or using SAFEMODE strap. Implementations should not attempt to clear CMOS by using a jumper to pull VccRTC low. 00 Note: AFTERG3_EN This bit determines what state to go to when power is re-applied after a power failure (G3 state). This bit is in the RTC well and is only cleared by writes of 06h or 0Eh to CF9h (when the CF9h global reset bit is clear), receiving hard reset command with or without power cycle from SMBus or RTCRST#. 0 = System will return to S0 state (boot) after power is re-applied. 1 = System will return to the S5 state (except if it was in S4, in which case it will return to S4). In the S5 state, the only enabled wake event is the Power Button or any enabled wake event that was preserved through the power failure. This bit is set any time a Power Button Override occurs (i.e., the power button is pressed for at least 4 consecutive seconds), due to the corresponding bit in the SMBus unconditional power down message, due to an internal thermal sensor catastrophic condition and the assertion of THRMTRIP#. RSMRST# is sampled using the RTC clock. Therefore, low times that are less than one RTC clock period may not be detected by the PCH. Intel(R) Communications Chipset 89xx Series - Datasheet 312 RW October 2012 Order Number: 327879-001US 6.0 6.7.1.4 Offset A6h: GEN_PMCON_LOCK- General Power Management Configuration Lock Register C Table 6-88. Offset A6h: GEN_PMCON_LOCK- General Power Management Configuration Lock Register Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 00h Offset Start: A6h Offset End: A6h Power Well: Core 07 :03 Reserved Reserved SLP_S4# Stretching Policy Lock-Down -- When set to 1, this bit locks down the SLP_S4# Minimum Assertion Width, the SLP_S4# Assertion Stretch Enable, the Disable SLP_S4# Stretching after G3 and SLP_S4# Assertion Stretch Enable bits in the GEN_PMCON_3 register, making them read-only. This bit becomes locked when a value of 1b is written to it. Writes of 0 to this bit are always ignored. This bit is cleared by platform reset. RWLO When set to 1, this bit locks down the ACPI Base Address Register (ABASE) at offset 40h. The Base Address Field becomes read-only. ACPI_BASE_LOCK This bit becomes locked when a value of 1b is written to it. Writes of 0 to this bit are always ignored. Once locked by writing 1, the only way to clear this bit is to perform a platform reset. RWLO 01 00 6.7.1.5 Reserved Sticky Bit Access Bit Acronym 02 Bit Description Bit Reset Value Bit Range Reserved Offset A9h: Chipset Initialization Register 4 (PM--B0:D31:F0) Table 6-89. Offset A9h: Chipset Initialization Register 4 (PM--B0:D31:F0) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D31 :F0 Default: 01h Bit Acronym 07 :00 October 2012 Order Number: 327879-001US Offset Start: A9h Offset End: A9h Power Well: Core Bit Description Sticky CIR4 Field 1 -- BIOS must program this field to 45h. Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 313 6.7.1.6 Offset ABh: BM_BREAK_EN Register (PM--B0:D31:F0) Table 6-90. Offset ABh: BM_BREAK_EN Register (PM--B0:D31:F0) Description: View: PCI BAR: Configuration Size: 8 bit Bit Range 07 B0:D31 :F0 Default: 01h Offset Start: ABh Offset End: ABh Power Well: Core Bit Acronym Bit Description Sticky Bit Reset Value Bit Access 0 = Serial ATA traffic will not act as a break event. STORAGE_BREAK_EN 1 = Serial ATA traffic acts as a break event, Serial ATA master activity will cause BM_STS to be set and will cause a break from C3/C4. 06 PCIE_BREAK_EN 05 Reserved 04 :03 Bus:Device:Function: RW 0 = PCI Express* traffic will not act as a break event. 1 = PCI Express* traffic acts as a break event, PCI Express* master activity will cause BM_STS to be set and will cause a break from C3/C4. RW Reserved Reserved Reserved 0 = EHCI traffic will not act as a break event. 1 = EHCI traffic acts as a break event, EHCI master activity will cause BM_STS to be set and will cause a break from C3/C4. RW 01 CIR4 Field 1 -- BIOS must program this field to 45h. RW 00 CIR4 Field 1 -- BIOS must program this field to 45h. RW 02 6.7.1.7 EHCI_BREAK_EN Offset ACh: PMIR--Power Management Initialization Register (PM--B0:D31:F0) Table 6-91. Offset ACh: PMIR--Power Management Initialization Register (PM--B0:D31:F0) Description: View: PCI Size: 32 bit Bit Range Default: 00000000h Bit Acronym 31 :30 29 :21 20 19 :00 Bus:Device:Function: B0:D31 :F0 BAR: Configuration Power Well: Bit Description PMIR Field 1-- BIOS must program these bits to 11b. Reserved CF9GR Sticky Bit Reset Value Bit Access RW Reserved CF9h Global Reset: When set, a CF9h write of 6h or Eh will cause a Global reset of both the Host and Intel(R) ME partitions. If this bit is cleared, a CF9h write of 6h or Eh will only reset the host partition. This bit field is not reset by a CF9h reset RW PMIR Field 0 -- BIOS must program these bits to 00300h. RW Intel(R) Communications Chipset 89xx Series - Datasheet 314 Offset Start: ACh Offset End: ACh October 2012 Order Number: 327879-001US 6.0 6.7.1.8 Offset B8h: GPIO_ROUT--GPIO Routing Control Register (PM--B0:D31:F0) Table 6-92. Offset B8h: GPIO_ROUT--GPIO Routing Control Register (PM--B0:D31:F0) Description: View: PCI BAR: Configuration Size: 32 bit Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Range Bit Acronym 31 :30 Offset Start: B8h Offset End: BBh Power Well: Bit Description Sticky Bit Reset Value Bit Access GPIO15 Route -- See bits 1:0 for description. RW Same pattern for GPIO14 through GPIO3. 05 :04 GPIO2 Route -- See bits 1:0 for description. RW 03 :02 GPIO1 Route -- See bits 1:0 for description. RW 01 :00 GPIO0 Route -- GPIO can be routed to cause an NMI, SMI# or SCI when the GPIO[n]_STS bit is set. If the GPIO0 is not set to an input, this field has no effect. If the system is in an S1-S5 state and if the GPE0_EN bit is also set, then the GPIO can cause a Wake event, even if the GPIO is NOT routed to cause an NMI, SMI# or SCI. 00 = No effect. 01 = SMI# (if corresponding ALT_GPI_SMI_EN bit is also set) 10 = SCI (if corresponding GPE0_EN bit is also set) 11 = NMI (If corresponding GPI_NMI_EN is also set) RW Note: GPIOs that are not implemented will not have the corresponding bits implemented in this register. 6.7.2 APM I/O Decode Table 6-93 shows the I/O registers associated with APM support. This register space is enabled in the PCI Bus 0:Device 31: Function 0 space (APMDEC_EN), and cannot be moved (fixed I/O location). Table 6-93. APM Register Map Offset Start Offset End Register ID - Description Default Value B2h B2h "Offset B2h: APM_CNT--Advanced Power Management Control Port Register" on page 316 00h B3h B3h "Offset B3h: APM_STS--Advanced Power Management Status Port Register" on page 316 00h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 315 6.7.2.1 Offset B2h: APM_CNT--Advanced Power Management Control Port Register Table 6-94. Offset B2h: APM_CNT--Advanced Power Management Control Port Register Description: View: IA F Size: 8 bit Bit Range BAR: 0000h (IO) B0:D31 :F0 Default: 00h Bit Acronym Offset Start: B2h Offset End: B2h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Used to pass an APM command between the OS and the SMI handler. Writes to this port not only store data in the APMC register, but also generates an SMI# when the APMC_EN bit is set. 07 :00 6.7.2.2 Bus:Device:Function: RW Offset B3h: APM_STS--Advanced Power Management Status Port Register Table 6-95. Offset B3h: APM_STS--Advanced Power Management Status Port Register Description: View: IA F Size: 8 bit Bit Range 07 :00 Bus:Device:Function: B0:D31 :F0 BAR: 0000h (IO) Default: 00h Bit Acronym Power Well: Core Bit Description Used to pass data between the OS and the SMI handler. Basically, this is a scratchpad register and is not affected by any other register or function (other than a PCI reset). Intel(R) Communications Chipset 89xx Series - Datasheet 316 Offset Start: B3h Offset End: B3h Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 6.0 6.7.3 Power Management I/O Registers The following table shows the registers associated with ACPI and Legacy power management support. These registers are enabled in the PCI Bus 0: Device 31: Function 0 space (PM_IO_EN), and can be moved to any I/O location (128-byte aligned). The registers are defined to support the ACPI 3.0a specification, and use the same bit names. Note: All reserved bits and registers will always return 0 when read, and will have no effect when written. Table 6-96. ACPI and Legacy I/O Register Map Offset Start Offset End Register ID - Description Default Value 00h B3h "Offset B3h: APM_STS--Advanced Power Management Status Port Register" on page 316 00h 3642537: Offset PMBASE+00h : PM1_STSoPo wer Management 1 Status Register (Sheet 1 of 3) 0000h 02h 02h "Offset PMBASE + 02h: PM1_EN--Power Management 1 Enable Register" on page 321 0000h 04h 04h "Offset PMBASE + 04h: PM1_CNT--Power Management 1 Control" on page 322 00000000h 08h 08h "Offset PMBASE + 08h: PM1_TMR--Power Management 1 Timer Register" on page 323 xx000000h 20h 20h "Offset PMBASE + 20h: GPE0_STS--General Purpose Event 0 Status Register" on page 324 00000000000 00000h 28h 28h "Offset PMBASE + 28h: GPE0_EN--General Purpose Event 0 Enables Register" on page 326 00000000000 00000h 30h 30h "Offset PMBASE + 30h: SMI_EN--SMI Control and Enable Register" on page 328 00000002h 34h 34h "Offset PMBASE + 34h: SMI_STS--SMI Status Register" on page 330 00000000h 38h 38h "Offset PMBASE +38h: ALT_GP_SMI_EN--Alternate GPI SMI Enable Register" on page 333 0000h 3Ah 3Ah "Offset PMBASE +3Ah: ALT_GP_SMI_STS--Alternate GPI SMI Status Register" on page 333 0000h 3Ch 3Ch "Offset PMBASE +3Ch: UPRWC--USB Per-Port Registers Write Control" on page 334 0000h 42h 42h "Offset PMBASE +42h: GPE_CNTL--General Purpose Control Register" on page 335 00h 44h 44h "Offset PMBASE +44h: DEVACT_STS--Device Activity Status Register" on page 336 0000h 50h 50h "Offset PMBASE +50h: PM2_CNT--Power Management 2 Control" on page 337 00h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 317 6.7.3.1 Offset PMBASE+00h: PM1_STS--Power Management 1 Status Register If bit 10 or 8 in this register is set, and the corresponding _EN bit is set in the PM1_EN register, then the PCH will generate a Wake Event. Once back in an S0 state (or if already in an S0 state when the event occurs), the PCH will also generate an SCI if the SCI_EN bit is set, or an SMI# if the SCI_EN bit is not set. Note: Bit 5 does not cause an SMI# or a wake event. Bit 0 does not cause a wake event but can cause an SMI# or SCI.. Table 6-97. Offset PMBASE+00h: PM1_STS--Power Management 1 Status Register (Sheet 1 of 3) Description: View: PCI Size: 16 bit Bit Range 15 14 13 :12 B0:D31 Offset Start: 00h Bus:Device:Function: :F0 Offset End: 00h BAR: PMBASE (IO) Bits 0-7: Core, Bits Power Well: 8-15: Resume, except Bit 11 in RTC Default: 0000h Bit Description WAK_STS Wake Status -- This bit is not affected by hard resets caused by a CF9 write, but is reset by RSMRST#. 0 = Software clears this bit by writing a 1 to it. 1 = Set by hardware when the system is in one of the sleep states (via the SLP_EN bit) and an enabled wake event occurs. Upon setting this bit, the PCH will transition the system to the ON state. If the AFTERG3_EN bit is not set and a power failure (such as removed batteries) occurs without the SLP_EN bit set, the system will return to an S0 state when power returns, and the WAK_STS bit will not be set. If the AFTERG3_EN bit is set and a power failure occurs without the SLP_EN bit having been set, the system will go into an S5 state when power returns, and a subsequent wake event will cause the WAK_STS bit to be set. Any subsequent wake event would have to be caused by either a Power Button press, or an enabled wake event that was preserved through the power failure (enable bit in the RTC well). RWC PCI Express* Wake Status: 0 = Software clears this bit by writing a 1 to it. If the WAKE# pin is still active during the write or the PME message received indication has not been cleared in the root port, then the bit will remain active (i.e. all inputs to this bit are level-sensitive). 1 = This bit is set by hardware to indicate that the system woke due to a PCI Express* wakeup event. This wakeup event can be caused by the PCI PCIEXPWAK_STS Express* WAKE# pin being active or receipt of a PCI Express* PME message at a root port. This bit is set only when one of these events causes the system to transition from a non-S0 system power state to the S0 system power state. This bit is set independent of the state of the PCIEXP_WAKE_DIS bit. Note: This bit does not itself cause a wake event or prevent entry to a sleeping state. Thus if the bit is 1 and the system is put into a sleeping state, the system will not automatically wake. RWC Reserved Bit Access Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 318 Sticky Bit Reset Value Bit Acronym October 2012 Order Number: 327879-001US 6.0 Table 6-97. Offset PMBASE+00h: PM1_STS--Power Management 1 Status Register (Sheet 2 of 3) Description: View: PCI Size: 16 bit Bit Range BAR: PMBASE (IO) Bus:Device:Function: B0:D31 Offset Start: 00h :F0 Offset End: 00h Bits 0-7: Core, Bits Power Well: 8-15: Resume, except Bit 11 in RTC Default: 0000h Bit Acronym Bit Description Sticky Bit Reset Value Bit Access 11 Power Button Override Status: 0 = Software clears this bit by writing a 1 to it. 1 = This bit is set any time a Power Button Override occurs (i.e., the power button is pressed for at least 4 consecutive seconds), due to the corresponding bit in the SMBus slave message, Intel(R) ME Initiated Power Button Override, Intel(R) ME Initiated Host Reset with Power down or due to an internal thermal PWRBTNOR_STS sensor catastrophic condition. The power button override causes an unconditional transition to the S5 state, as well as sets the AFTERG3_EN bit. The BIOS or SCI handler clears this bit by writing a 1 to it. This bit is not affected by hard resets via CF9h writes, and is not reset by RSMRST#. Thus, this bit is preserved through power failures. If this bit is still asserted when the global SCI_EN is set then an SCI will be generated. RWC 10 RTC_STS RTC Status -- This bit is not affected by hard resets caused by a CF9 write, but is reset by RSMRST#: 0 = Software clears this bit by writing a 1 to it. 1 = Set by hardware when the RTC generates an alarm (assertion of the IRQ8# signal). Additionally if the RTC_EN bit (PMBASE + 02h, bit 10) is set, the setting of the RTC_STS bit will generate a wake event. RWC ME_STS Management Engine Status -- This bit is set when the Intel(R) Management Engine generates a NonMaskable wake event, and is not affected by any other enable bit. When this bit is set, the Host Power Management logic wakes to S0. This bit is only set by hardware and can only be reset by writing a one to this bit position. This bit is not affected by hard resets caused by a CF9 write, but is reset by RSMRST#. RWC 09 Power Button Status -- This bit is not affected by hard resets caused by a CF9 write: 0 = If the PWRBTN# signal is held low for more than 4 seconds, the hardware clears the PWRBTN_STS bit, sets the PWRBTNOR_STS bit, and the system transitions to the S5 state with only PWRBTN# enabled as a wake event. 08 PWRBTN_STS 1 = This bit can be cleared by software by writing a one to the bit position. This bit is set by hardware when the PWRBTN# signal is asserted Low, independent of any other enable bit. RWC In the S0 state, while PWRBTN_EN and PWRBTN_STS are both set, an SCI (or SMI# if SCI_EN is not set) will be generated. In any sleeping state S1-S5, while PWRBTN_EN (PMBASE + 02h, bit 8) and PWRBTN_STS are both set, a wake event is generated. If the PWRBTN_STS bit is cleared by software while the PWRBTN# signal is sell asserted, this will not cause the PWRBN_STS bit to be set. The PWRBTN# signal must go inactive and active again to set the PWRBTN_STS bit. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 319 Table 6-97. Offset PMBASE+00h: PM1_STS--Power Management 1 Status Register (Sheet 3 of 3) Description: View: PCI Size: 16 bit BAR: PMBASE (IO) Bus:Device:Function: B0:D31 Offset Start: 00h :F0 Offset End: 00h Bits 0-7: Core, Bits Power Well: 8-15: Resume, except Bit 11 in RTC Default: 0000h Bit Acronym 07 :06 Reserved Reserved GBL _STS Global Status: 0 = The SCI handler should then clear this bit by writing a 1 to the bit location. 1 = Set when an SCI is generated due to BIOS wanting the attention of the SCI handler. BIOS has a corresponding bit, BIOS_RLS, which will cause an SCI and set this bit. RWC BM_STS Bus Master Status -- This bit will not cause a wake event, SCI or SMI#. 0 = Software clears this bit by writing a 1 to it. Set by the PCH when a PCH-visible bus master requests access to memory or the BM_BUSY# signal is active. RWC 05 04 03 :01 00 Reserved TO_STS Bit Description Bit Access Reserved Timer Overflow Status: 0 = The SCI or SMI# handler clears this bit by writing a 1 to the bit location. 1 = This bit gets set any time bit 22 of the 24-bit timer goes high (bits are numbered from 0 to 23). This will occur every 2.3435 seconds. When the TMROF_EN bit (PMBASE + 02h, bit 0) is set, then the setting of the TMROF_STS bit will additionally generate an SCI or SMI# (depending on the SCI_EN). Intel(R) Communications Chipset 89xx Series - Datasheet 320 Sticky Bit Reset Value Bit Range RWC October 2012 Order Number: 327879-001US 6.0 6.7.3.2 Offset PMBASE + 02h: PM1_EN--Power Management 1 Enable Register Table 6-98. Offset PMBASE + 02h: PM1_EN--Power Management 1 Enable Register Description: View: PCI Size: 16 bit Bit Range 15 14 13 :11 BAR: PMBASE (IO) Bit Acronym Reserved Reserved 09 Reserved 05 04 :01 Bit Description Sticky PWRBTN_EN Power Well: Bits 0-7: Core, Bits 8-9, 11-15: Resume, Bit 10: RTC Bit Reset Value Bit Access Reserved PCI Express* Wake Disable -- Modification of this bit has no impact on the value of the PCIEXP_WAKE_STS bit. PCIEXPWAK_DIS 0 = Inputs to the PCIEXP_WAKE_STS bit in the PM1 Status register enabled to wake the system. 1 = Inputs to the PCIEXP_WAKE_STS bit in the PM1 Status register disabled from waking the system. RTC_EN 07 :06 B0:D31 Offset Start: 02h :F0 Offset End: 02h Default: 0000h 10 08 Bus:Device:Function: RW Reserved RTC Event Enable -- This bit is in the RTC well to allow an RTC event to wake after a power failure. This bit is not cleared by any reset other than RTCRST# or a Power Button Override event. 0 = No SCI (or SMI#) or wake event is generated then RTC_STS (PMBASE + 00h, bit 10) goes active. 1 = An SCI (or SMI#) or wake event will occur when this bit is set and the RTC_STS bit goes active. RW Reserved Power Button Enable -- This bit is used to enable the setting of the PWRBTN_STS bit to generate a power management event (SMI#, SCI). PWRBTN_EN has no effect on the PWRBTN_STS bit (PMBASE + 00h, bit 8) being set by the assertion of the power button. The Power Button is always enabled as a Wake event. 0 = Disable. 1 = Enable. Reserved Reserved GBL_EN Global Enable -- When both the GBL_EN and the GBL_STS bit (PMBASE + 00h, bit 5) are set, an SCI is raised. 0 = Disable. 1 = Enable SCI on GBL_STS going active. Reserved Reserved RW RW Timer Overflow Interrupt Enable -- Works in conjunction with the SCI_EN bit (PMBASE + 04h, bit 0) as described below: 00 TMROF_EN October 2012 Order Number: 327879-001US TMROF_ EN SCI_EN Effect when TMROF_STS is set 0 X No SMI# or SCI 1 0 SMI# 1 1 SCI RW Intel(R) Communications Chipset 89xx Series - Datasheet 321 6.7.3.3 Offset PMBASE + 04h: PM1_CNT--Power Management 1 Control Table 6-99. Offset PMBASE + 04h: PM1_CNT--Power Management 1 Control Description: View: PCI Size: 32 bit Bit Range 31 :14 13 BAR: PMBASE (IO) Bus:Device:Function: B0:D31 Offset Start: 04h :F0 Offset End: 04h Bits 0-7: Core, Power Well: Bits 8-12: RTC, Bits 13-15: Resume Default: 00000000h Bit Acronym Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved SLP_EN Sleep Enable -- Setting this bit causes the system to sequence into the Sleep state defined by the SLP_TYP field. WO Sleep Type -- This 3-bit field defines the type of Sleep the system should enter when the SLP_EN bit is set to 1. These bits are only reset by RTCRST#. Code 12 :10 09 :03 SLP_TYP Master Interrupt 000b ON: Typically maps to S0 state. 001b Puts CPU in S1 state. 010b Reserved 011b Reserved 100b Reserved 101b Suspend-To-RAM. Assert SLP_S3#: Typically maps to S3 state. 110b Suspend-To-Disk. Assert SLP_S3#, and SLP_S4#: Typically maps to S4 state. 111b Soft Off. Assert SLP_S3#, SLP_S4#, and SLP_S5#: Typically maps to S5 state. Reserved Reserved 02 Global Release: 0 = This bit always reads as 0. GBL_RLS 1 = ACPI software writes a 1 to this bit to raise an event to the BIOS. BIOS software has a corresponding enable and status bits to control its ability to receive ACPI events. WO 01 Bus Master Reload -- This bit is treated as a scratchpad bit. This bit is reset to 0 by PLTRST# 0 = Bus master requests will not cause a break from the C3 state. BM_RLD 1 = Enables Bus Master requests (internal or external) to cause a break from the C3 state. If software fails to set this bit before going to C3 state, the PCH will still return to a snoopable state from C3 or C4 states due to bus master activity. RW 00 SCI Enable -- Selects the SCI interrupt or the SMI# interrupt for various events including the bits in the PM1_STS register (bit 10, 8, 0), and bits in GPE0_STS. 0 = These events will generate an SMI#. 1 = These events will generate an SCI. RW SCI_EN Intel(R) Communications Chipset 89xx Series - Datasheet 322 RW October 2012 Order Number: 327879-001US 6.0 6.7.3.4 Offset PMBASE + 08h: PM1_TMR--Power Management 1 Timer Register Table 6-100. Offset PMBASE + 08h: PM1_TMR--Power Management 1 Timer Register Description: View: PCI Size: 32 bit BAR: PMBASE (IO) Bus:Device:Function: Default: xx000000h Bit Acronym 31 :24 Reserved Reserved TMR_VAL Timer Value -- Returns the running count of the PM timer. This counter runs off a 3.579545 MHz clock (14.31818 MHz divided by 4). It is reset to 0 during a PCI reset, and then continues counting as long as the system is in the S0 state. After an S1 state, the counter will not be reset (it will continue counting from the last value in S0 state. Anytime bit 22 of the timer goes HIGH to LOW (bits referenced from 0 to 23), the TMROF_STS bit (PMBASE + 00h, bit 0) is set. The High-to-Low transition will occur every 2.3435 seconds. If the TMROF_EN bit (PMBASE + 02h, bit 0) is set, an SCI interrupt is also generated. October 2012 Order Number: 327879-001US Offset Start: 08h Offset End: 08h Power Well: Core Bit Range 23 :00 B0:D31 :F0 Bit Description Sticky Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 323 6.7.3.5 Offset PMBASE + 20h: GPE0_STS--General Purpose Event 0 Status Register This register is symmetrical to the General Purpose Event 0 Enable Register. Unless indicated otherwise below, if the corresponding _EN bit is set, then when the _STS bit is set, the PCH will generate a Wake Event. Once back in an S0 state (or if already in an S0 state when the event occurs), the PCH will also generate an SCI if the SCI_EN bit is set, or an SMI# if the SCI_EN bit (PMBASE + 04h, bit 0) is not set. Bits 31:16 are reset by a CF9h write; bits 63:32 and 15:00 are not. All are reset by RSMRST#. Table 6-101. Offset PMBASE + 20h: GPE0_STS--General Purpose Event 0 Status Register (Sheet 1 of 3) Description: View: PCI Size: 64 bit Bit Range 63 :36 35 34 :32 Default: 0000000000000000h Bit Acronym Reserved GPIO27_STS Reserved 31 :16 GPIOn_STS 15 :14 Reserved 13 Bus:Device:Function: B0:D31 :F0 BAR: PMBASE (IO) PME_B0_STS Bit Description Offset Start: 20h Offset End: 20h Power Well: Resume Sticky Bit Reset Value Bit Access Reserved 0 = Disable. 1 = Set by hardware and can be reset by writing a one to this bit position or a resume well reset. This bit is set at the level specified in GP27IO_POL. GPIO27 is always monitored as an input for the purpose of setting this bit, regardless of the actual GPIO configuration. RWC Reserved 0 = Software clears this bit by writing a 1 to it. 1 = These bits are set any time the corresponding GPIO is set up as an input and the corresponding GPIO signal is high (or low if the corresponding GP_INV bit is set). If the corresponding enable bit is set in the GPE0_EN register, then when the GPIO[n]_STS bit is set: - If the system is in an S1-S5 state, the event will also wake the system. - If the system is in an S0 state (or upon waking back to an S0 state), a SCI will be caused depending on the GPIO_ROUT bits (B0:D31:F0:B8h, bits 31:30) for the corresponding GPI. Mapping is as follows: bit 31 corresponds to GPIO[15]... and bit 16 corresponds to GPIO[0]. RWC Reserved This bit will be set to 1 by the PCH when any internal device with PCI Power Management capabilities on bus 0 asserts the equivalent of the PME# signal. Additionally, if the PME_B0_EN bit is set, and the system is in an S0 state, then the setting of the PME_B0_STS bit will generate an SCI (or SMI# if SCI_EN is not set). If the PME_B0_STS bit is set, and the system is in an S1-S4 state (or S5 state due to SLP_TYP and SLP_EN), then the setting of the PME_B0_STS bit will generate a wake event, and an SCI (or SMI# if SCI_EN is not set) will be generated. If the system is in an S5 state due to power button override, then the PME_B0_STS bit will not cause a wake event or SCI. RWC The default for this bit is 0. Writing a 1 to this bit position clears this bit. Note: HD audio wake events are reported in this bit. Intel(R) Management Engine "maskable" wake events are also reported in this bit. 12 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 324 October 2012 Order Number: 327879-001US 6.0 Table 6-101. Offset PMBASE + 20h: GPE0_STS--General Purpose Event 0 Status Register (Sheet 2 of 3) Description: View: PCI Size: 64 bit Bit Range BAR: PMBASE (IO) Bus:Device:Function: B0:D31 Offset Start: 20h :F0 Offset End: 20h Default: 0000000000000000h Power Well: Resume Bit Acronym Bit Description 11 PME_STS 0 = Software clears this bit by writing a 1 to it. 1 = Set by hardware when the PME# signal goes active. Additionally, if the PME_EN bit is set, and the system is in an S0 state, then the setting of the PME_STS bit will generate an SCI or SMI# (if SCI_EN is not set). If the PME_EN bit is set, and the system is in an S1-S4 state (or S5 state due to setting SLP_TYP and SLP_EN), then the setting of the PME_STS bit will generate a wake event, and an SCI will be generated. If the system is in an S5 state due to power button override or a power failure, then PME_STS will not cause a wake event or SCI. 10 Reserved Reserved Sticky Bit Reset Value Bit Access RWC 0 = Software clears this bit by writing a 1 to it. 1 = Set by hardware to indicate that: - The PME event message was received on one or more of the PCI Express* ports - An Assert PMEGPE message received from the Processor via DMI 09 PCI_EXP_STS 08 RI_STS 07 Notes: 1 - The PCI WAKE# pin has no impact on this bit. 2 - If the PCI_EXP_STS bit went active due to an Assert PMEGPE message, then a Deassert PMEGPE message must be received prior to the software write in order for the bit to be cleared. 3 - If the bit is not cleared and the corresponding PCI_EXP_EN bit is set, the level-triggered SCI will remain active. 4 - A race condition exists where the PCI Express* device sends another PME message because the PCI Express* device was not serviced within the time when it must resend the message. This may result in a spurious interrupt, and this is comprehended and approved by the PCI Express* Specification, Revision 1.0a. The window for this race condition is approximately 95-105 milliseconds. 0 = Software clears this bit by writing a 1 to it. 1 = Set by hardware when the RI# input signal goes active. SMBus Wake Status -- The SMBus controller can independently cause an SMI# or SCI, so this bit does not need to do so (unlike the other bits in this register). Software clears this bit by writing a 1 to it. 0 = Wake event Not caused by the PCH's SMBus logic. 1 = Set by hardware to indicate that the wake event was caused by the PCH's SMBus logic.This bit will be set by the WAKE/SMI# command type, even if the system is already awake. The SMI handler should then clear this bit. Notes: SMB_WAK_STS 1. The SMBus controller will independently cause an SMI# so this bit does not need to do so (unlike the other bits in this register). 2. This bit is set by the SMBus slave command 01h (Wake/SMI#) even when the system is in the S0 state. Therefore, to avoid an instant wake on subsequent transitions to sleep states, software must clear this bit after each reception of the Wake/SMI# command or just prior to entering the sleep state. 3. The SMBALERT_STS bit (D31:F3:I/O Offset 00h:Bit 5) should be cleared by software before the SMB_WAK_STS bit is cleared. October 2012 Order Number: 327879-001US RWC RWC RWC Intel(R) Communications Chipset 89xx Series - Datasheet 325 Table 6-101. Offset PMBASE + 20h: GPE0_STS--General Purpose Event 0 Status Register (Sheet 3 of 3) Description: View: PCI BAR: PMBASE (IO) Size: 64 bit B0:D31 Offset Start: 20h :F0 Offset End: 20h Default: 0000000000000000h Bit Range Bit Acronym 06 TCOSCI_STS 05 :03 Reserved 02 SWGPE_STS Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access Software clears this bit by writing a 1 to it. 0 = TOC logic or thermal sensor logic did Not cause SCI. 1 = Set by hardware when the TCO logic or thermal sensor logic causes an SCI. RWC Reserved The SWGPE_CTRL bit (bit 1 of GPE_CTRL reg) acts as a level input to this bit. RWC 0 = This bit is cleared by writing a 1 to this bit position. HOT_PLUG_STS 1 = When a PCI Express* Hot-Plug event occurs. This will cause an SCI if the HOT_PLUG_EN bit is set in the GEP0_EN register. 01 00 6.7.3.6 Bus:Device:Function: Reserved RWC Reserved Offset PMBASE + 28h: GPE0_EN--General Purpose Event 0 Enables Register This register is symmetrical to the General Purpose Event 0 Status Register. All the bits in this register should be cleared to 0 based on a Power Button Override or processor Thermal Trip event. The resume well bits are all cleared by RSMRST#. The RTC well bits are cleared by RTCRST#. Table 6-102. Offset PMBASE + 28h: GPE0_EN--General Purpose Event 0 Enables Register (Sheet 1 of 2) Description: View: PCI Size: 64 bit Bit Range 63 :36 35 34 :32 31 :16 B0:D31 Offset Start: 28h Bus:Device:Function: :F0 Offset End: 28h BAR: PMBASE (IO) Default: 0000000000000000h Bit Acronym Reserved GPIO27_EN Bit Description Sticky Bit Reset Value Bit Access Reserved 0 = Disable. 1 = Enable the setting of the GPIO27_STS bit to generate a wake event/SCI/SMI#. Reserved Reserved GPIn_EN These bits enable the corresponding GPI[n]_STS bits being set to cause a SCI, and/or wake event. These bits are cleared by RSMRST#. Mapping is as follows: bit 31 corresponds to GPIO15... and bit 16 corresponds to GPIO0. Intel(R) Communications Chipset 89xx Series - Datasheet 326 Bits 0-7, 9, 12, 14-63 Power Well: Resume, Bits 8, 10-11, 13 RTC RW RW October 2012 Order Number: 327879-001US 6.0 Table 6-102. Offset PMBASE + 28h: GPE0_EN--General Purpose Event 0 Enables Register (Sheet 2 of 2) Description: View: PCI Size: 64 bit Bit Range 15 :14 13 BAR: PMBASE (IO) Bits 0-7, 9, 12, 14-63 Power Well: Resume, Bits 8, 10-11, 13 RTC Bit Acronym Reserved PME_B0_EN Bit Description Sticky Reserved 11 PME_EN 10 Reserved Reserved 0 = Disable SCI generation upon PCI_EXP_STS bit being set. PCI_EXP_EN 1 = Enables PCH to cause an SCI when PCI_EXP_STS bit is set. This is used to allow the PCI Express* ports, including the link to the Processor, to cause an SCI due to wake/PME events. RI_EN 07 Reserved 06 05 :03 02 01 00 TCOSCI_EN Reserved Bit Access RW It is only cleared by Software or RTCRST#. It is not cleared by CF9h writes. 0 = Disable. 1 = Enables the setting of the PME_STS to generate a wake event and/or an SCI. PME# can be a wake event from the S1 - S4 state or from S5 (if entered via SLP_EN, but not power button override). 08 Bit Reset Value Reserved 0 = Disable 1 = Enables the setting of the PME_B0_STS bit to generate a wake event and/or an SCI or SMI#. PME_B0_STS can be a wake event from the S1-S4 states, or from S5 (if entered via SLP_TYP and SLP_EN) or power failure, but not Power Button Override. This bit defaults to 0. Reserved 09 B0:D31 Offset Start: 28h :F0 Offset End: 28h Default: 0000000000000000h Note: 12 Bus:Device:Function: The value of this bit will be maintained through a G3 state and is not affected by a hard reset caused by a CF9h write. 0 = Disable. 1 = Enables the setting of the RI_STS to generate a wake event. RW RW RW Reserved 0 = Disable. 1 = Enables the setting of the TCOSCI_STS to generate an SCI. RW Reserved This bit allows software to control the assertion of SWGPE_STS bit. This bit This bit, when set to 1, enables the SW GPE function. If SWGPE_CTRL is written to a 1, hardware will set SWGPE_STS (acts as a level input) If SWGPE_STS, SWGPE_EN, and SCI_EN are all 1's, an SCI will be generated If SWGPE_STS = 1, SWGPE_EN = 1, SCI_EN = 0, and GBL_SMI_EN = 1 then an SMI# will be generated. RW 0 = Disables SCI generation upon the HOT_PLUG_STS bit being set. HOT_PLUG_EN 1 = Enables the PCH to cause an SCI when the HOT_PLUG_STS bit is set. This is used to allow the PCI Express* ports to cause an SCI due to hot-plug events. RW SWGPE_EN Reserved October 2012 Order Number: 327879-001US Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 327 6.7.3.7 Offset PMBASE + 30h: SMI_EN--SMI Control and Enable Register Note: This register is symmetrical to the SMI status register. Table 6-103. Offset PMBASE + 30h: SMI_EN--SMI Control and Enable Register (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR: PMBASE (IO) Bit Acronym 31 :28 Reserved 26 :19 Reserved INTEL_USB2_EN 17 LEGACY_USB2_EN 14 13 Reserved PERIODIC_EN TCO_EN Offset Start: 30h Offset End: 30h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved Setting this bit will cause the PCH to generate an SMI# when the GPIO_UNLOCK_SMI_STS bit is set GPIO_UNLOCK_SMI in the SMI_STS register. _EN Once written to `1', this bit can only be cleared by PLTRST#. 18 16 :15 B0:D31 :F0 Default: 00000002h Bit Range :27 Bus:Device:Function: RWO Reserved 0 = Disable 1 = Enables Intel-Specific USB2 SMI logic to cause SMI#. RW 0 = Disable 1 = Enables legacy USB2 logic to cause SMI#. RW Reserved 0 = Disable. 1 = Enables the PCH to generate an SMI# when the PERIODIC_STS bit (PMBASE + 34h, bit 14) is set in the SMI_STS register (PMBASE + 34h). 0 = Disables TCO logic generating an SMI#. If the NMI2SMI_EN bit is set, SMIs that are caused by rerouted NMIs will not be gated by the TCO_EN bit. Even if the TCO_EN bit is 0, NMIs will still be routed to cause SMIs. 1 = Enables the TCO logic to generate SMI#. RW RW This bit cannot be written once the TCO_LOCK bit is set. 12 11 10 :08 07 Reserved MCSMI_EN Reserved MCSMI_EN Microcontroller SMI Enable: 0 = Disable. 1 = Enables PCH to trap accesses to the microcontroller range (62h or 66h) and generate an SMI#. "Trapped" cycles will be claimed by the PCH on PCI, but not forwarded to LPC. Reserved Reserved BIOS_RLS BIOS Release: 0 = This bit will always return 0 on reads. Writes of 0 to this bit have no effect. 1 = Enables the generation of an SCI interrupt for ACPI software when a one is written to this bit position by BIOS software. Note: GBL_STS being set will cause an SCI, even if the SCI_EN bit is not set. Software must take great care not to set the BIOS_RLS bit (which causes GBL_STS to be set) if the SCI handler is not in place. Intel(R) Communications Chipset 89xx Series - Datasheet 328 RW WO October 2012 Order Number: 327879-001US 6.0 Table 6-103. Offset PMBASE + 30h: SMI_EN--SMI Control and Enable Register (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: PMBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: 00000002h Offset Start: 30h Offset End: 30h Power Well: Core Bit Description SWSMI_TMR_EN Software SMI# Timer Enable: 0 = Disable. Clearing the SWSMI_TMR_EN bit before the timer expires will reset the timer and the SMI# will not be generated. Starts Software SMI# Timer. When the SWSMI timer expires (the timeout period depends upon the SWSMI_RATE_SEL bit setting), SWSMI_TMR_STS is set and an SMI# is generated. SWSMI_TMR_EN stays set until cleared by software. RW APMC_EN 0 = Disable. Writes to the APM_CNT register will not cause an SMI#. Enables writes to the APM_CNT register to cause an SMI#. RW 04 SLP_SMI_EN 0 = Disables the generation of SMI# on SLP_EN. This bit must be 0 before the software attempts to transition the system into a sleep state by writing a 1 to the SLP_EN bit. 1 = A write of 1 to the SLP_EN bit (bit 13 in PM1_CNT register) will generate an SMI#, and the system will not transition to the sleep state based on that write to the SLP_EN bit. RW 03 LEGACY_USB_EN 0 = Disable. 1 = Enables legacy USB circuit to cause SMI#. RW BIOS_EN 0 = Disable. 1 = Enables the generation of SMI# when ACPI software writes a 1 to the GBL_RLS bit (B0:D31:F0:PMBase + 04h:bit 2). If the BIOS_STS bit (B0:D31:F0:PMBase + 34h:bit 2), which gets set when software writes 1 to GBL_RLS bit, is already a 1 at the time that BIOS_EN becomes 1, an SMI# will be generated when BIOS_EN gets set. RW EOS End of SMI -- (special). This bit controls the arbitration of the SMI signal to the processor. This bit must be set for the PCH to assert SMI# low to the processor after SMI# has been asserted previously. 0 = Once the PCH asserts SMI# low, the EOS bit is automatically cleared. 1 = When this bit is set to 1, SMI# signal will be deasserted for 4 PCI clocks before its assertion. In the SMI handler, the processor should clear all pending SMIs (by servicing them and then clearing their respective status bits), set the EOS bit, and exit SMM. This will allow the SMI arbiter to reassert SMI upon detection of an SMI event and the setting of a SMI status bit. Note: PCH is able to generate first SMI after reset even though EOS bit is not set. Subsequent SMI require EOS bit is set. RW GBL_SMI_EN 0 = No SMI# will be generated by PCH. This bit is reset by a PCI reset event. 1 = Enables the generation of SMI# in the system upon any enabled SMI event. When the SMI_LOCK bit is set, this bit cannot be changed. RW 06 05 02 01 00 October 2012 Order Number: 327879-001US Sticky Bit Reset Value Bit Acronym Bit Access Intel(R) Communications Chipset 89xx Series - Datasheet 329 6.7.3.8 Offset PMBASE + 34h: SMI_STS--SMI Status Register Note: If the corresponding _EN bit is set when the _STS bit is set, the PCH will cause an SMI# (except bits 8-10 and 12, which do not need enable bits since they are logic ORs of other registers that have enable bits). The PCH uses the same GPE0_EN register (I/O address: PMBase+2Ch) to enable/disable both SMI and ACPI SCI general purpose input events. ACPI OS assumes that it owns the entire GPE0_EN register per ACPI spec. Problems arise when some of the general-purpose inputs are enabled as SMI by BIOS, and some of the general purpose inputs are enabled for SCI. In this case ACPI OS turns off the enabled bit for any GPIx input signals that are not indicated as SCI generalpurpose events at boot, and exit from sleeping states. BIOS should define a dummy control method which prevents the ACPI OS from clearing the SMI GPE0_EN bits. Table 6-104. Offset PMBASE + 34h: SMI_STS--SMI Status Register (Sheet 1 of 3) Description : View PCI : Size: 32 bit Default: 00000000h Bit Range Bit Acronym 31 :28 Reserved 27 26 25 :22 B0:D31 Bus:Device:Function: :F0 BAR: PMBASE (IO) Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved bit will be set if the GPIO registers lockdown logic GPIO_UNLOCK_SMI_ This is requesting an SMI#. Writing a `1' to this bit position STS clears this bit to `0'. SPI_STS This bit will be set if the SPI logic is generating an SMI#. This bit is read only because the sticky status and enable bits associated with this function are located in the SPI registers. Reserved Reserved RWC RO 21 MONITOR_STS This bit will be set if the Trap/SMI logic has caused the SMI. This will occur when the processor or a bus master accesses an assigned register (or a sequence of accesses). RO 20 PCI_EXP_SMI_STS PCI Express* SMI event occurred. This could be due to a PCI Express* PME event or Hot-Plug event. RO 19 Reserved 18 17 Reserved This non-sticky read-only bit is a logical OR of each of the SMI status bits in the Intel-Specific USB2 SMI Status Register ANDed with the corresponding enable bits. Additionally, the Port Disable Write Enable SMI is reported in this bit; the specific status bit for this event is contained in the USB Per-Port Registers Write Control Register in this I/O space. This bit will not be active if the enable bits are not set. Writes to this bit will have no effect. All integrated USB2 Host Controllers are represented with this bit. RO This non-sticky read-only bit is a logical OR of each of the SMI status bits in the USB2 Legacy Support Register ANDed with the corresponding enable bits. LEGACY_USB2_STS This bit will not be active if the enable bits are not set. Writes to this bit will have no effect. All integrated USB2 Host Controllers are represented with this bit. RO INTEL_USB2_STS Intel(R) Communications Chipset 89xx Series - Datasheet 330 Offset Start: 34h Offset End: 34h October 2012 Order Number: 327879-001US 6.0 Table 6-104. Offset PMBASE + 34h: SMI_STS--SMI Status Register (Sheet 2 of 3) Description : View PCI : Size: 32 bit Bit Range BAR: PMBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym Offset Start: 34h Offset End: 34h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access SMBus SMI Status -- Software clears this bit by writing a 1 to it. 0 = This bit is set from the 64 kHz clock domain used by the SMBus. Software must wait at least 15.63 s after the initial assertion of this bit before clearing it. 1 = Indicates that the SMI# was caused by: 16 SMBUS_SMI_STS - The SMBus Slave receiving a message that an SMI# should be caused, or - The SMBALERT# signal goes active and the SMB_SMI_EN bit is set and the SMBALERT_DIS bit is cleared, or - The SMBus Slave receiving a Host Notify message and the HOST_NOTIFY_INTREN and the SMB_SMI_EN bits are set, or - The PCH detecting the SMLINK_SLAVE_SMI command while in the S0 state. RWC SERIRQ_SMI_STS 0 = SMI# was not caused by the SERIRQ decoder. 1 = Indicates that the SMI# was caused by the SERIRQ decoder. This is not a sticky bit RO 14 PERIODIC_STS Software clears this bit by writing a 1 to it. 0 = Software clears this bit by writing a 1 to it. 1 = This bit is set at the rate determined by the PER_SMI_SEL bits. If the PERIODIC_EN bit (PMBASE + 30h, bit 14) is also set, the PCH generates an SMI#. RWC 13 TCO_STS Software clears this bit by writing a 1 to it. 0 = SMI# not caused by TCO logic. 1 = Indicates the SMI# was caused by the TCO logic. This is not a wake event. RWC 15 12 11 10 DEVMON_STS Device Monitor Status: 0 = SMI# not caused by Device Monitor. 1 = Set if bit 0 of the DEVACT_STS register (PMBASE + 44h) is set. The bit is not sticky, so writes to this bit will have no effect. RO MCSMI_STS Microcontroller SMI# Status -- R/WC. Software clears this bit by writing a 1 to it. 0 = indicates that there has been no access to the power management microcontroller range (62h or 66h). 1 = Set if there has been an access to the power management microcontroller range (62h or 66h) and the Microcontroller Decode Enable #1 bit in the LPC Bridge I/O Enables configuration register is 1 (B0:D31:F0:Offset 82h:bit 11). This implementation assumes that the Microcontroller is on LPC. If this bit is set, and the MCSMI_EN bit is also set, the PCH will generate an SMI#. RWC GPE0_STS This bit is a logical OR of the bits in the ALT_GP_SMI_STS register that are also set up to cause an SMI# (as indicated by the GPI_ROUT registers) and have the corresponding bit set in the ALT_GP_SMI_EN register. Bits that are not routed to cause an SMI# will have no effect on this bit. 0 = SMI# was not generated by a GPI assertion. 1 = SMI# was generated by a GPI assertion. RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 331 Table 6-104. Offset PMBASE + 34h: SMI_STS--SMI Status Register (Sheet 3 of 3) Description : View PCI : Size: 32 bit Bit Range BAR: PMBASE (IO) Bus:Device:Function: Default: 00000000h Offset Start: 34h Offset End: 34h Power Well: Core Bit Access Bit Description 09 GPE0_STS This bit is a logical OR of the bits 47:32, 14:10, 8, 6:2 and 0 in the GPE0_STS register (PMBASE + 28h) that also have the corresponding bit set in the GPE0_EN register (PMBASE + 2Ch). 0 = SMI# was not generated by a GPE0 event. 1 = SMI# was generated by a GPE0 event. RO 08 PM1_STS_REG This is an ORs of the bits in the ACPI PM1 Status Register (offset PMBASE+00h) that can cause an SMI#. 0 = SMI# was not generated by a PM1_STS event. 1 = SMI# was generated by a PM1_STS event. RO 07 Reserved 06 SWSMI_TMR_STS 04 03 02 01 :00 Sticky Bit Reset Value Bit Acronym 05 Reserved Software clears this bit by writing a 1 to it. 0 = Software SMI# Timer has Not expired. 1 = Set by the hardware when the Software SMI# Timer expires. RWC APM_STS Software clears this bit by writing a 1 to it. 0 = No SMI# generated by write access to APM Control register with APMCH_EN bit set. 1 = SMI# was generated by a write access to the APM Control register with the APMC_EN bit set. RWC SLP_SMI_STS Software clears this bit by writing a 1 to the bit location. 0 = No SMI# caused by write of 1 to SLP_EN bit when SLP_SMI_EN bit is also set. 1 = Indicates an SMI# was caused by a write of 1 to SLP_EN bit when SLP_SMI_EN bit is also set. RWC LEGACY_USB_STS This bit is a logical OR of each of the SMI status bits in the USB Legacy Keyboard/Mouse Control Registers ANDed with the corresponding enable bits. This bit will not be active if the enable bits are not set. 0 = SMI# was not generated by USB Legacy event. 1 = SMI# was generated by USB Legacy event. RO BIOS_STS 0 = No SMI# generated due to ACPI software requesting attention. 1 = This bit gets set by hardware when a 1 is written by software to the GBL_RLS bit (B0:D31:F0:PMBase + 04h:bit 2). When both the BIOS_EN bit (B0:D31:F0:PMBase + 30h:bit 2) and the BIOS_STS bit are set, an SMI# will be generated. The BIOS_STS bit is cleared when software writes a 1 to its bit position. RWC Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 332 B0:D31 :F0 October 2012 Order Number: 327879-001US 6.0 6.7.3.9 Offset PMBASE +38h: ALT_GP_SMI_EN--Alternate GPI SMI Enable Register Table 6-105. Offset PMBASE +38h: ALT_GP_SMI_EN--Alternate GPI SMI Enable Register Description: View: PCI Size: 16 bit Bit Range 15 :00 6.7.3.10 BAR: PMBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: 0000h Bit Acronym Offset Start: 38h Offset End: 38h Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access Alternate GPI SMI Enable -- These bits are used to enable the corresponding GPIO to cause an SMI#. For these bits to have any effect, the following must be true. - The corresponding bit in the ALT_GP_SMI_EN register ALT_GP_SMI_EN is set. - The corresponding GPI must be routed in the GPI_ROUT register to cause an SMI. - The corresponding GPIO must be implemented. Mapping is as follows: bit 15 corresponds to GPIO15... bit 0 corresponds to GPIO0. RW Offset PMBASE +3Ah: ALT_GP_SMI_STS--Alternate GPI SMI Status Register Table 6-106. Offset PMBASE +3Ah: ALT_GP_SMI_STS--Alternate GPI SMI Status Register Description: View: PCI Size: 16 bit Bit Range 15 :00 BAR: PMBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: 0000h Bit Acronym Power Well: Resume Bit Description Sticky Alternate GPI SMI Status -- These bits report the status of the corresponding GPIOs. 0 = Inactive. Software clears this bit by writing a 1 to it. 1 = Active These bits are sticky. If the following conditions are true, then an SMI# will be generated and the GPE0_STS ALT_GP_SMI bit set: _STS - The corresponding bit in the ALT_GPI_SMI_EN register (PMBASE + 38h) is set - The corresponding GPIO must be routed in the GPI_ROUT register to cause an SMI. - The corresponding GPIO must be implemented. All bits are in the resume well. Default for these bits is dependent on the state of the GPIO pins. October 2012 Order Number: 327879-001US Offset Start: 3Ah Offset End: 3Ah Bit Reset Value Bit Access RWC Intel(R) Communications Chipset 89xx Series - Datasheet 333 6.7.3.11 Offset PMBASE +3Ch: UPRWC--USB Per-Port Registers Write Control Table 6-107. Offset PMBASE +3Ch: UPRWC--USB Per-Port Registers Write Control Description: View: PCI Size: 16 bit BAR: PMBASE (IO) Bit Acronym 15 :09 Reserved Reserved Offset Start: 3Ch Offset End: 3Ch Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access Reserved Write Enable Status: 0 = This bit gets set by hardware when the "Per-Port Registers Write Enable" bit is written from 0 to 1 1 = This bit is cleared by software writing a 1b to this bit location The setting condition takes precedence over the clearing condition in the event that both occur at once. When this bit is 1b and bit 0 is 1b, the INTEL_USB2_STS bit is set in the SMI_STS register. 08 RWC Reserved 01 USB Per-Port Registers Write Enable: 0 = Disable 1 = Writes are enabled to the USB Port Disable Override register (Chipset Config - Offset 3518h) 00 Write Enable SMI Enable: 0 = Disable 1 = enables the generation of SMI when the Per-Port Registers Write Enable (bit 1) is written from 0 to 1. Once written to 1b, this bit can not be cleared by software. Intel(R) Communications Chipset 89xx Series - Datasheet 334 B0:D31 :F0 Default: 0000h Bit Range 07 :02 Bus:Device:Function: RW RWO October 2012 Order Number: 327879-001US 6.0 6.7.3.12 Offset PMBASE +42h: GPE_CNTL--General Purpose Control Register Table 6-108. Offset PMBASE +42h: GPE_CNTL--General Purpose Control Register Description: View: PCI Size: 8 bit BAR: PMBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: 00h Bit Range Bit Acronym 07 :03 Reserved Offset Start: 42h Offset End: 42h Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access Reserved 02 This bit controls the polarity of the GPIO27 pin needed to set the GPIO27_STS bit. GPIO27_POL 0 = GPIO27 - 0 will set the GPIO27_STS bit. 1 = GPIO27 - 1 will set the GPIO27_STS bit. RW 01 This bit allows software to control the assertion of SWGPE_STS bit. This bit is used by hardware as the level input signal for the SWGPE_STS bit in the GPE0_STS register. When SWGPE_CTRL is 1, SWGPE_STS will be set to 1, and writes to SWGPE_STS with a value of 1 to clear SWGPE_CTRL SWGPE_STS results in SWGPE_STS being set back to 1 by hardware. When SWGPE_CTRL is 0, writes to SWGPE_STS with a value of 1 will clear SWGPE_STS to 0. This bit is cleared to 0 based on a Power Button Override, CPU Thermal Event as well as by the RSMRST# pin assertion. RW 00 Reserved October 2012 Order Number: 327879-001US Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 335 6.7.3.13 Offset PMBASE +44h: DEVACT_STS--Device Activity Status Register Each bit indicates if an access has occurred to the corresponding device's trap range, or for bits 6:9 if the corresponding PCI interrupt is active. This register is used in conjunction with the Periodic SMI# timer to detect any system activity for legacy power management. The periodic SMI# timer indicates if it is the right time to read the DEVACT_STS register (PMBASE + 44h). Note: Software clears bits that are set in this register by writing a 1 to the bit position. Table 6-109. Offset PMBASE +44h: DEVACT_STS--Device Activity Status Register Description: View: PCI Size: 16 bit Default: 0000h Bit Range Bit Acronym 15 :13 Reserved 12 11 :10 B0:D31 Bus:Device:Function: :F0 BAR: PMBASE (IO) KBC_ACT_STS Reserved Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved KBC (60/64h). 0 = Indicates that there has been no access to this device's I/O range. 1 = This device's I/O range has been accessed. Clear this bit by writing a 1 to the bit location. RWC Reserved 09 PIRQ[D or H]. The corresponding PCI interrupts have not been PIRQDH_ACT_STS active. 1 = At least one of the corresponding PCI interrupts has been active. Clear this bit by writing a 1 to the bit location. RWC 08 PIRQ[C or G]. 0 = The corresponding PCI interrupts have not been active. PIRQCG_ACT_STS 1 = At least one of the corresponding PCI interrupts has been active. Clear this bit by writing a 1 to the bit location. RWC 07 PIRQ[B or F]. 0 = The corresponding PCI interrupts have not been PIRQBF_ACT_STS active. 1 = At least one of the corresponding PCI interrupts has been active. Clear this bit by writing a 1 to the bit location. RWC 06 PIRQ[A or E]. 0 = The corresponding PCI interrupts have not been active. PIRQAE_ACT_STS 1 = At least one of the corresponding PCI interrupts has been active. Clear this bit by writing a 1 to the bit location. RWC 05 :00 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 336 Offset Start: 44h Offset End: 44h October 2012 Order Number: 327879-001US 6.0 6.7.3.14 Offset PMBASE +50h: PM2_CNT--Power Management 2 Control Table 6-110. Offset PMBASE +50h: PM2_CNT--Power Management 2 Control Description: View: PCI BAR: PMBASE (IO) Size: 8 bit Bus:Device:Function: Default: 00h Bit Acronym 07 :01 Reserved Reserved ARB_DIS Arbiter Disable -- This bit is a scratchpad bit for legacy software compatibility. 6.8 Offset Start: 50h Offset End: 50h Power Well: Core Bit Range 00 B0:D31 :F0 Bit Description Sticky Bit Reset Value Bit Access RW System Management TCO Registers (B0:D31:F0) The TCO logic is accessed via registers mapped to the PCI configuration space (Device 31:Function 0) and the system I/O space. For TCO PCI Configuration registers, See LPC Device 31:Function 0 PCI Configuration registers. 6.8.1 TCO Register I/O Map The TCO I/O registers reside in a 32-byte range pointed to by a TCOBASE value, which is, PMBASE + 60h in the PCI config space. The following table shows the mapping of the registers within that 32-byte range. Each register is described in the following sections. Table 6-111. TCO I/O Register Address Map Offset Start 00h Offset End 00h Register ID - Description Default Value "Offset TCOBASE +00h: TCO_RLD--TCO Timer Reload and Current Value Register" 0000h on page 338 02h 02h "Offset TCOBASE +02h: TCO_DAT_IN--TCO Data In Register" on page 338 00h 03h 03h "Offset TCOBASE +03h: TCO_DAT_OUT--TCO Data Out Register" on page 339 00h 04h 04h "Offset TCOBASE +04h: TCO1_STS--TCO1 Status Register" on page 339 2000h 06h 06h "Offset TCOBASE +06h: TCO2_STS--TCO2 Status Register" on page 341 0000h 08h 08h "Offset TCOBASE +08h: TCO1_CNT--TCO1 Control Register" on page 343 0000h 0Ah 0Ah "Offset TCOBASE +0Ah: TCO2_CNT--TCO2 Control Register" on page 344 0008h 0Ch 0Dh "Offset TCOBASE +0Ch and Offset TCOBASE +0Dh: TCO_MESSAGE1 and TCO_MESSAGE2 Registers" on page 345 00h 0Eh 0Eh "Offset TCOBASE + 0Eh: TCO_WDCNT--TCO Watchdog Control Register" on page 345 00h 10h 10h "Offset TCOBASE + 10h: SW_IRQ_GEN--Software IRQ Generation Register" on page 346 03h 12h 12h "Offset TCOBASE + 12h:TCO_TMR--TCO Timer Initial Value Register" on page 346 0004h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 337 6.8.1.1 Offset TCOBASE +00h: TCO_RLD--TCO Timer Reload and Current Value Register Table 6-112. Offset TCOBASE +00h: TCO_RLD--TCO Timer Reload and Current Value Register Description: View: PCI Size: 16 bit BAR: TCOBASE (IO) Bus:Device:Function: Default: 0000h Power Well: Core Bit Range Bit Acronym 15 :10 Reserved Reserved TCO_RLD TCO Timer Value -- Reading this register will return the current count of the TCO timer. Writing any value to this register will reload the timer to prevent the timeout. 09 :00 6.8.1.2 B0:D31 Offset Start: 00h :F0 Offset End: 00h Bit Description Sticky Bit Reset Value Bit Access RW Offset TCOBASE +02h: TCO_DAT_IN--TCO Data In Register Table 6-113. Offset TCOBASE +02h: TCO_DAT_IN--TCO Data In Register Description: View: PCI Size: 8 bit B0:D31 Bus:Device:Function: :F0 BAR: TCOBASE (IO) Default: 00h Power Well: Core Bit Range Bit Acronym Bit Description 07 :00 TCO_DAT_IN TCO Data In Value -- This data register field is used for passing commands from the OS to the SMI handler. Writes to this register will cause an SMI and set the SW_TCO_SMI bit in the TCO1_STS register (B0:D31:F0:04h). Intel(R) Communications Chipset 89xx Series - Datasheet 338 Offset Start: 02h Offset End: 02h Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 6.0 6.8.1.3 Offset TCOBASE +03h: TCO_DAT_OUT--TCO Data Out Register Table 6-114. Offset TCOBASE +03h: TCO_DAT_OUT--TCO Data Out Register Description: View: PCI Size: 8 bit Bit Range 07 :00 6.8.1.4 BAR: TCOBASE (IO) Bus:Device:Function B0:D31 Offset Start: 03h : :F0 Offset End: 03h Default: 00h Bit Acronym Power Well: Core Bit Description Sticky Bit Reset Value TCO Data Out Value -- This data register field is used for passing commands from the SMI handler to the OS. Writes to this register will set TCO_DAT_OUT the TCO_INT_STS bit in the TCO_STS register. It will also cause an interrupt, as selected by the TCO_INT_SEL bits. Bit Access RW Offset TCOBASE +04h: TCO1_STS--TCO1 Status Register Table 6-115. Offset TCOBASE +04h: TCO1_STS--TCO1 Status Register (Sheet 1 of 2) Description: View: PCI Size: 16 bit BAR: TCOBASE (IO) Default: 2000h Bit Range Bit Acronym 15 :14 Reserved 13 12 11 10 09 Bus:Device:Function B0:D31 Offset Start: 04h : :F0 Offset End: 04h TCO_SLVSEL Power Well: Core Bit Description Sticky Bit Access Reserved TCO Slave Select -- This register bit is Read Only by Host and indicates the value of TCO Slave Select Soft Strap. See the PCH Soft Straps section of the SPI Chapter for details. 0 = Software clears this bit by writing a 1 to it. 1 = PCH received a DMI special cycle message DMISERR_STS via DMI indicating that it wants to cause an SERR#. The software must read the Processor to determine the reason for the SERR#. Reserved Bit Reset Value RO RWC Reserved DMISMI_STS 0 = Software clears this bit by writing a 1 to it. 1 = PCH received a DMI special cycle message via DMI indicating that it wants to cause an SMI. The software must read the Processor to determine the reason for the SMI. RWC DMISCI_STS 0 = Software clears this bit by writing a 1 to it. 1 = PCH received a DMI special cycle message via DMI indicating that it wants to cause an SCI. The software must read the Processor to determine the reason for the SCI. RWC October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 339 Table 6-115. Offset TCOBASE +04h: TCO1_STS--TCO1 Status Register (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range 08 Bus:Device:Function B0:D31 Offset Start: 04h : :F0 Offset End: 04h BAR: TCOBASE (IO) Default: 2000h Bit Acronym Power Well: Core Bit Description 0 = Software clears this bit by writing a 1 to it. 1 = PCH sets this bit and generates and SMI# to indicate an invalid attempt to write to the BIOS. This occurs when either: a) The BIOSWP bit is changed from 0 to 1 and BIOSWR_STS the BLD bit is also set, or b) any write is attempted to the BIOS and the BIOSWP bit is also set. Note: On write cycles attempted to the 4 MB lower alias to the BIOS space, the BIOSWR_STS will not be set. Sticky Bit Reset Value Bit Access RWC This bit is in the RTC well. 0 = Cleared by writing a 1 to the bit position or by RTCRST# going active. 1 = This bit is set when the Year byte (RTC I/O space, index offset 09h) rolls over from 99 to 00. Setting this bit will cause an SMI# (but not a wake event). 07 06 :04 03 Note: The NEWCENTURY_STS bit is not valid when the RTC battery is first installed (or when RTC power has not been maintained). Software NEWCENTURY_ can determine if RTC power has not been maintained by checking the RTC_PWR_STS bit STS (B0:D31:F0:A4h, bit 2), or by other means (such as a checksum on RTC RAM). If RTC power is determined to have not been maintained, BIOS should set the time to a valid value and then clear the NEWCENTURY_STS bit. The NEWCENTURY_STS bit may take up to 3 RTC clocks for the bit to be cleared after a 1 is written to the bit to clear it. After writing a 1 to this bit, software should not exit the SMI handler until verifying that the bit has actually been cleared. This will ensure that the SMI is not reentered. Reserved Reserved TIMEOUT 0 = Software clears this bit by writing a 1 to it. 1 = Set by PCH to indicate that the SMI was caused by the TCO timer reaching 0. RWC 02 0 = Software clears this bit by writing a 1 to it. TCO_INT_STS 1 = SMI handler caused the interrupt by writing to the TCO_DAT_OUT register (TCOBASE + 03h). RWC 01 0 = Software clears this bit by writing a 1 to it. SW_TCO_SMI 1 = Software caused an SMI# by writing to the TCO_DAT_IN register (TCOBASE + 02h). RWC 00 0 = Cleared by clearing the associated NMI status bit. NMI2SMI_STS 1 = Set by the PCH when an SMI# occurs because an event occurred that would otherwise have caused an NMI (because NMI2SMI_EN is set). RO Intel(R) Communications Chipset 89xx Series - Datasheet 340 RWC October 2012 Order Number: 327879-001US 6.0 6.8.1.5 Offset TCOBASE +06h: TCO2_STS--TCO2 Status Register Table 6-116. Offset TCOBASE +06h: TCO2_STS--TCO2 Status Register (Sheet 1 of 2) Description: View: PCI Size: 16 bit BAR: TCOBASE (IO) Bit Acronym 15 :05 Reserved 03 02 01 B0:D31 Offset Start: 06h :F0 Offset End: 06h Default: 0000h Bit Range 04 Bus:Device:Function: Bit Description Sticky Resume (Except Bit 0, in RTC) Bit Reset Value Bit Access Reserved SMLink Slave SMI Status -- Allow the software to go directly into pre-determined sleep state. This avoids race conditions. Software clears this bit by SMLINK_SLV_SMI_ writing a 1 to it: STS The bit is reset by RSMRST#, but not due to the PCI Reset associated with exit from S3-S5 states. 1 = PCH sets this bit to 1 when it receives the SMI message on the SMLink's Slave Interface. Reserved Power Well: RWC Reserved 0 = Cleared by PCH based on RSMRST# or by software writing a 1 to this bit. Software should first clear the SECOND_TO_STS bit before writing a 1 to clear the BOOT_STS bit. 1 = Set to 1 when the SECOND_TO_STS bit goes from 0 to 1 and the processor has not fetched the first instruction. If rebooting due to a second TCO timer timeout, and if the BOOT_STS bit is set, the PCH will reboot using the `safe' multiplier (1111). This allows the system to recover from a processor frequency multiplier that is too high, and allows the BIOS to check the BOOT_STS bit at boot. If the bit is set and the frequency multiplier is 1111, then the BIOS knows that the processor has been programmed to an invalid multiplier. RWC 0 = Software clears this bit by writing a 1 to it, or by a RSMRST#. 1 = PCH sets this bit to 1 to indicate that the TIMEOUT bit had been (or is currently) set and a SECOND_TO_STS second timeout occurred before the TCO_RLD register was written. If this bit is set and the NO_REBOOT config bit is 0, then the PCH will reboot the system after the second timeout. The reboot is done by asserting PLTRST#. RWC BOOT_STS October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 341 Table 6-116. Offset TCOBASE +06h: TCO2_STS--TCO2 Status Register (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range BAR: TCOBASE (IO) Bus:Device:Function: B0:D31 Offset Start: 06h :F0 Offset End: 06h Default: 0000h Bit Acronym Bit Description Sticky Power Well: Resume (Except Bit 0, in RTC) Bit Reset Value Bit Access Intruder Detect: 0 = Software clears this bit by writing a 1 to it, or by RTCRST# assertion. 1 = Set by PCH to indicate that an intrusion was detected. This bit is set even if the system is in G3 state. 00 INTRD_DET Notes: 1. This bit has a recovery time. After writing a 1 to this bit position (to clear it), the bit may be read back as a 1 for up 65 microseconds before it is read as a 0. Software must be aware of this recovery time when reading this bit after clearing it. 2. If the INTRUDER# signal is active when the software attempts to clear the INTRD_DET bit, the bit will remain as a 1, and the SMI# will be generated again immediately. The SMI handler can clear the INTRD_SEL bits (TCOBASE + 0Ah, bits 2:1), to avoid further SMIs. However, if the INTRUDER# signals goes inactive and then active again, there will not be further SMI's (because the INTRD_SEL bits would select that no SMI# be generated). 3. If the INTRUDER# signal goes inactive some point after the INTRD_DET bit is written as a 1, then the INTRD_DET signal will go to a 0 when INTRUDER# input signal goes inactive. This is slightly different than a classic sticky bit, since most sticky bits would remain active indefinitely when the signal goes active and would immediately go inactive when a 1 is written to the bit. Intel(R) Communications Chipset 89xx Series - Datasheet 342 RWC October 2012 Order Number: 327879-001US 6.0 6.8.1.6 Offset TCOBASE +08h: TCO1_CNT--TCO1 Control Register Table 6-117. Offset TCOBASE +08h: TCO1_CNT--TCO1 Control Register Description: View: PCI Size: 16 bit BAR: TCOBASE (IO) Bit Acronym 15 :13 Reserved 11 10 B0:D31 :F0 Default: 0000h Bit Range 12 Bus:Device:Function: TCO_LOCK Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved When set to 1, this bit prevents writes from changing the TCO_EN bit (in offset 30h of Power Management I/O space). Once this bit is set to 1, it can not be cleared by software writing a 0 to this bit location. A core-well reset is required to change this bit from 1 to 0. This bit defaults to 0. TCO Timer Halt: 0 = The TCO Timer is enabled to count. 1 = The TCO Timer will halt. It will not count, and TCO_TMR_HLT thus cannot reach a value that will cause an SMI# or set the SECOND_TO_STS bit. When set, this bit will prevent rebooting and prevent Alert On LAN event messages from being transmitted on the SMLink (but not Alert On LAN* heartbeat messages). Reserved Offset Start: 08h Offset End: 08h RWLO RW Reserved 0 = Normal NMI functionality. 1 = Forces all NMIs to instead cause SMIs. The functionality of this bit is dependent upon the settings of the NMI_EN bit and the GBL_SMI_EN bit as detailed in the following table: 09 08 07 :00 NMI2SMI_EN NMI_ EN GBL_S MI_EN Description 0b 0b No SMI# at all because GBL_SMI_EN = 0 0b 1b SMI# will be caused due to NMI events 1b 0b No SMI# at all because GBL_SMI_EN = 0 1b 1b No SMI# due to NMI because NMI_EN = 1 NMI_NOW 0 = Software clears this bit by writing a 1 to it. The NMI handler is expected to clear this bit. Another NMI will not be generated until the bit is cleared. 1 = Writing a 1 to this bit causes an NMI. This allows the BIOS or SMI handler to force an entry to the NMI handler. Reserved Reserved October 2012 Order Number: 327879-001US RW RWC Intel(R) Communications Chipset 89xx Series - Datasheet 343 6.8.1.7 Offset TCOBASE +0Ah: TCO2_CNT--TCO2 Control Register Table 6-118. Offset TCOBASE +0Ah: TCO2_CNT--TCO2 Control Register Description: View: PCI Size: 16 bit BAR: TCOBASE (IO) B0:D31 Offset Start: 0Ah :F0 Offset End: 0Ah Default: 0008h Bit Range Bit Acronym 15 :06 Reserved 05 :04 Bus:Device:Function: OS_POLICY Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access Reserved OS-based software writes to these bits to select the policy that the BIOS will use after the platform resets due the WDT. The following convention is recommended for the BIOS and OS: 00 = Boot normally 01 = Shut down 10 = Don't load OS. Hold in pre-boot state and use LAN to determine next step 11 = Reserved RW These are scratchpad bits. They should not be reset when the TCO logic resets the platform due to Watchdog Timer. 03 02 :01 00 At reset (via RSMRST# asserted) this bit is set and GPIO[11] alerts are disabled. GPIO11_ALERT_ 0 = Enable. DISABLE 1 = Disable GPIO11/MST_SMBALERT# as an alert source for the heartbeats and the SMBus slave. INTRD_SEL Reserved This field selects the action to take if the INTRUDER# signal goes active. 00 = No interrupt or SMI# 01 = Interrupt (as selected by TCO_INT_SEL). 10 = SMI 11 = Reserved RW Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 344 RW October 2012 Order Number: 327879-001US 6.0 6.8.1.8 Offset TCOBASE +0Ch and Offset TCOBASE +0Dh: TCO_MESSAGE1 and TCO_MESSAGE2 Registers I/O Address: TCOBASE +0Ch (Message 1) TCOBASE +0Dh (Message 2) Table 6-119. Offset TCOBASE +0Ch and Offset TCOBASE +0Dh: TCO_MESSAGE1 and TCO_MESSAGE2 Registers Description: View: PCI BAR: TCOBASE (IO) Size: 8 bit Bit Range Bus:Device:Function B0:D31 Offset Start: 0Ch Offset End: 0Dh : :F0 Default: 00h Bit Acronym Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access BIOS can write into these registers to indicate 07 :00 boot progress. The external microcontroller TCO_MESSAGE[n] its can read these registers to monitor the boot RW progress 6.8.1.9 Offset TCOBASE + 0Eh: TCO_WDCNT--TCO Watchdog Control Register Table 6-120. Offset TCOBASE + 0Eh: TCO_WDCNT--TCO Watchdog Control Register Description: View: PCI Size: 8 bit Bit Range 07 :00 BAR: TCOBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: 00h Bit Acronym Power Well: Resume Bit Description Sticky The BIOS or system management software can write into this register to indicate more details on the boot TCO_WDCNT progress. The register will reset to 00h based on a RSMRST# (but not PLTRST#). The external microcontroller can read this register to monitor boot progress. October 2012 Order Number: 327879-001US Offset Start: 0Eh Offset End: 0Eh Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 345 6.8.1.10 Offset TCOBASE + 10h: SW_IRQ_GEN--Software IRQ Generation Register Table 6-121. Offset TCOBASE + 10h: SW_IRQ_GEN--Software IRQ Generation Register Description: View: PCI Size: 8 bit BAR: TCOBASE (IO) Bus:Device:Function: B0:D31 Offset Start: 10h :F0 Offset End: 10h Default: 03h Bit Range Bit Acronym 07 :02 Reserved Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved 01 When software sets this bit to 1, IRQ12 will be IRQ12_CAUSE asserted. When software sets this bit to 0, IRQ12 will be deasserted. RW 00 When software sets this bit to 1, IRQ1 will be IRQ1_CAUSE asserted. When software sets this bit to 0, IRQ1 will be deasserted. RW 6.8.1.11 Offset TCOBASE + 12h:TCO_TMR--TCO Timer Initial Value Register Table 6-122. Offset TCOBASE + 12h:TCO_TMR--TCO Timer Initial Value Register Description: View: PCI Size: 16 bit Default: 0004h Bit Range Bit Acronym 15 :10 Reserved 09 :00 B0:D31 Bus:Device:Function: :F0 BAR: TCOBASE (IO) Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved TCO Timer Initial Value -- Value that is loaded into the timer each time the TCO_RLD register is written. Values of 0000h or 0001h will be ignored and should not be attempted. The timer is clocked at approximately 0.6 seconds, and thus allows timeouts ranging from 1.2 second to 613.8 seconds. The timer has an error of 1 tick (0.6s). The TCO Timer will only count down in the S0 state. Intel(R) Communications Chipset 89xx Series - Datasheet 346 Offset Start: 12h Offset End: 12h RW October 2012 Order Number: 327879-001US 6.0 6.9 General Purpose I/O Registers (B0:D31:F0) 6.9.1 General Purpose I/O Signals The control for the general purpose I/O signals is handled through a 128-byte I/O space. The base offset for this space is selected by the GPIOBASE register. Table 6-123. General Purpose I/O Signals Offset Start 00h Offset End Register ID - Description Default Value 03h "Offset GPIOBASE + 00h: GPIO_USE_SEL--GPIO Use Select Register" on page 348 F96BA1FFh 04h 07h "Offset GPIOBASE + 04h: GP_IO_SEL--GPIO Input/Output Select Register" on page 349 F6FF6EFFh 0Ch 0Fh "Offset GPIOBASE + 0Ch:GP_LVL--GPIO Level for Input or Output Register" on page 349 02FE0100h 18h 1Bh "Offset GPIOBASE + 18h: GPO_BLINK--GPO Blink Enable Register" on page 350 00040000h 1Ch 1Fh "Offset GPIOBASE + 1Ch: GP_SER_BLINK--GP Serial Blink Data" on page 351 00000000h 23h "Offset GPIOBASE + 20h: GP_SB_CMDSTS--GP Serial Blink Command Status" on page 352 00080000h 20h 24h 27h "Offset GPIOBASE + 24h: GP_SB_DATA--GP Serial Blink Data" on page 353 00000000h 28h 29h "Offset GPIOBASE + 28h: GPI_NMI_EN--GPI NMI Enable" on page 353 00000h 2Ah 2Bh "Offset GPIOBASE + 2Ah: GPI_NMI_STS--GPI NMI Status" on page 354 00000h 2Ch 2Fh "Offset GPIOBASE + 2Ch: GPI_INV--GPIO Signal Invert Register" on page 354 00000000h 30h 33h "Offset GPIOBASE + 30h: GPIO_USE_SEL2--GPIO Use Select 2 Register" on page 355 020300FEh 34h 37h "Offset GPIOBASE + 34h: GP_IO_SEL2--GPIO Input/Output Select 2 Register" on page 356 1F57FFF4h 38h 3Bh "Offset GPIOBASE + 38h: GP_LVL2--GPIO Level for Input or Output 2 Register" on A4AA0003h page 356 40h 43h "Offset GPIOBASE + 40h: GPIO_USE_SEL3--GPIO Use Select 3 Register" on page 357 00000000h 44h 47h "Offset GPIOBASE + 44h: GP_IO_SEL3--GPIO Input/Output Select 3 Register" on page 358 00000F00 48h 4Bh "Offset GPIOBASE + 48h: GP_LVL3--GPIO Level for Input or Output 3 Register" on 00000000h page 359 60h 63h "Offset GPIOBASE + 60h: GP_RST_SEL1--GPIO Reset Select" on page 360 01000000h 64h 67h "Offset GPIOBASE + 64h: GP_RST_SEL2--GPIO Reset Select" on page 361 00000000h 68h 6Bh "Offset GPIOBASE + 68h: GP_RST_SEL3--GPIO Reset Select" on page 361 00000000h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 347 6.9.1.1 Offset GPIOBASE + 00h: GPIO_USE_SEL--GPIO Use Select Register Table 6-124. Offset GPIOBASE + 00h: GPIO_USE_SEL--GPIO Use Select Register Description: View: PCI Size: 32 bit Bit Range BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 Offset Start: 00h :F0 Offset End: 03h Default: F96BA1FFh Bit Acronym Bit Description Sticky Power Well: Core for 0:7, 16:23, Resume for 8:15, 24:3 Bit Reset Value Bit Access Enables GPIO[n] (where n is the bit number) to be used as a GPIO, rather than for the native function. 1 = Signal used as GPIO (or unmuxed). 0 = Signal used as native function.(or unimplemented GPIO). Unmuxed GPIOs must report 1b in this register. Unimplemented GPIOs must report 0b in this register. If GPIO vs Native Mode are configured via SPI Soft Strap, the corresponding GPIO_USE_SEL bits for these GPIOs have no effect. 31 :00 Notes: GPIO_USE_ SEL[31:0] 1. The following bits are always 1 because they are always unMultiplexed: 0, 8, 15, 24, 27, and 28. 2. If GPIO[n] does not exist, then, the n-bit in this register will always read as 0 and writes will have no effect. 3. After a full reset (RSMRST#) all multiplexed signals in the resume and core wells are configured as their default function. After only a PLTRST#, the GPIOs in the core well are configured as their default function. 4. When configured to GPIO mode, the muxing logic will present the inactive state to native logic that uses the pin as an input. 5. All GPIOs are reset to the default state by CF9h reset except GPIO24. 6. Bit 26 may be overridden by bit 8 in the GEN_PMCON_3 Register. Intel(R) Communications Chipset 89xx Series - Datasheet 348 RW October 2012 Order Number: 327879-001US 6.0 6.9.1.2 Offset GPIOBASE + 04h: GP_IO_SEL--GPIO Input/Output Select Register Table 6-125. Offset GPIOBASE + 04h: GP_IO_SEL--GPIO Input/Output Select Register Description: View: PCI Size: 16 bit Bit Range 31 :00 6.9.1.3 BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: F6FF6EFFh Bit Acronym Offset Start: 04h Offset End: 07h Power Well: Bit Description Sticky Bit Reset Value Core for 23:16, 07:00 Resume for 31:24, 15:08 Bit Access When configured in native mode (GPIO_USE_SEL[n] is 0), writes to these bits have no effect. The value reported GP_IO_SEL in this register is undefined when programmed as native mode. [31:0] 0 = Output. The corresponding GPIO signal is an output. 1 = Input. The corresponding GPIO signal is an input. RW Offset GPIOBASE + 0Ch:GP_LVL--GPIO Level for Input or Output Register Table 6-126. Offset GPIOBASE + 0Ch:GP_LVL--GPIO Level for Input or Output Register Description: View: PCI Size: 32 bit Bit Range BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 :F0 Core for 23:16, 07:00 Power Well: Resume for 31:24, 15:08 Default 02FE0100h : Bit Acrony m Offset Start: 0Ch Offset End: 0Fh Bit Description Sticky Bit Reset Value Bit Access These registers are implemented as dual read/write with dedicated storage each. Write value will be stored in the write register, while read is coming from the read register which will always reflect the value of the pin. 31 :00 If GPIO[n] is programmed to be an output (via the GP_LVL corresponding bit in the GP_IO_SEL register), then the [31:0] corresponding GP_LVL[n] write register value will drive a high or low value on the output pin. 1 = high, 0 = low. RW When configured in native mode (GPIO_USE_SEL[n] is 0), writes to these bits are stored but have no effect to the pin value. The value reported in this register is undefined when programmed as native mode. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 349 6.9.1.4 Offset GPIOBASE + 18h: GPO_BLINK--GPO Blink Enable Register Table 6-127. Offset GPIOBASE + 18h: GPO_BLINK--GPO Blink Enable Register Description: View: PCI Size: 32 bit Bit Range 31 00 Note: BAR: GPIOBASE (IO) Bus:Device:Function: Default: 00040000h Bit Acronym Offset Start: 18h Offset End: 1Bh Power Well: Bit Description The setting of this bit has no effect if the corresponding GPIO signal is programmed as an input. 0 = The corresponding GPIO will function normally. 1 = If the corresponding GPIO is programmed as an output, the output signal will blink at a rate of approximately once per second. The high and low times GP_BLINK have approximately 0.5 seconds each. The GP_LVL bit is not altered when this bit is set. [31:0] The value of the corresponding GP_LVL bit remains unchanged during the blink process, and does not effect the blink in any way. The GP_LVL bit is not altered when programmed to blink. It will remain at its previous value. These bits correspond to GPIO in the Resume well. These bits revert to the default value based on RSMRST# or a write to the CF9h register (but not just on PLTRST#). Sticky Bit Reset Value Core for 23:16, 07:00 Resume for 31:24, 15:08 Bit Access RW GPIO18 will blink by default immediately after reset. This signal could be connected to an LED to indicate a failed boot (by programming BIOS to clear GP_BLINK18 after successful POST). Intel(R) Communications Chipset 89xx Series - Datasheet 350 B0:D31 :F0 October 2012 Order Number: 327879-001US 6.0 6.9.1.5 Offset GPIOBASE + 1Ch: GP_SER_BLINK--GP Serial Blink Data Table 6-128. Offset GPIOBASE + 1Ch: GP_SER_BLINK--GP Serial Blink Data Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym Power Well: Bit Description Sticky The setting of this bit has no effect if the corresponding GPIO is programmed as an input or if the corresponding GPIO has the GPO_BLINK bit set. When set to a `0', the corresponding GPIO will function normally. When using serial blink, this bit should be set to a 1 while the corresponding GP_IO_SEL bit is set to 1. Setting the GP_IO_SEL bit to 0 after the GP_SER_BLINK GP_SER_BLINK bit ensures PCH will not drive a 1 on the pin as an output. When this corresponding bit is set to a 1 and the [31:0] pin is configured to output mode, the serial blink capability is enabled. The PCH will serialize messages through an open-drain buffer configuration. The value of the corresponding GP_LVL bit remains unchanged and does not impact the serial blink capability in any way. Writes to this register have no effect when the corresponding pin is configured in native mode and the read value returned is undefined. October 2012 Order Number: 327879-001US Offset Start: 1Ch Offset End: 1Fh Bit Reset Value Core for 23:16, 07:00 Resume for 31:24, 15:08 Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 351 6.9.1.6 Offset GPIOBASE + 20h: GP_SB_CMDSTS--GP Serial Blink Command Status Table 6-129. Offset GPIOBASE + 20h: GP_SB_CMDSTS--GP Serial Blink Command Status Description: View: PCI Size: 32 bit BAR: GPIOBASE (IO) Bus:Device:Function: Default: 00080000h Bit Range Bit Acronym 31 :24 Reserved Offset Start: 20h Offset End: 23h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved DLS Data Length Select -- This field determines the number of bytes to serialize on GPIO 00 = Serialize bits 7:0 of GP_SB_DATA (1 byte) 01 = Serialize bits 15:0 of GP_SB_DATA (2 bytes) 10 = Undefined - Software must not write this value 11 = Serialize bits 31:0 of GP_SB_DATA (4 bytes) Software should not modify the value in this register unless the Busy bit is clear. Writes to this register have no effect when the corresponding pin is configured in native mode and the read value returned is undefined. RW 21 :16 DRS Data Rate Select -- This field selects the number of 120ns time intervals to count between Manchester data transitions. The default of 8h results in a 960ns minimum time between transitions. A value of 0h in this register produces undefined behavior. Software should not modify the value in this register unless the Busy bit is clear. RW 15 :09 Reserved 23 :22 07 :01 00 Reserved Busy -- This read-only status bit is the hardware indication that a serialization is in progress. Hardware sets this bit to 1 based on the Go bit being set. Hardware clears this bit when the Go bit is cleared by the hardware. 08 Reserved RO Reserved Go -- This bit is set to 1 by software to start the serialization process. Hardware clears the bit after the serialized data is sent. Writes of 0 to this register have no effect. Software should not write this bit to 1 unless the Busy status bit is cleared. Intel(R) Communications Chipset 89xx Series - Datasheet 352 B0:D31 :F0 RW October 2012 Order Number: 327879-001US 6.0 6.9.1.7 Offset GPIOBASE + 24h: GP_SB_DATA--GP Serial Blink Data Table 6-130. Offset GPIOBASE + 24h: GP_SB_DATA--GP Serial Blink Data Description: View: PCI Size: 32 bit Bit Range 31 :00 6.9.1.8 BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym Offset Start: 24h Offset End: 27h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access This register contains the data serialized out. The GP_SB_DATA number of bits shifted out are selected through the DLS field in the GP_SB_CMDSTS register. This register [31:0] should not be modified by software when the Busy bit is set. RW Offset GPIOBASE + 28h: GPI_NMI_EN--GPI NMI Enable Table 6-131. Offset GPIOBASE + 28h: GPI_NMI_EN--GPI NMI Enable Description: View: PCI Size: 16 bit Bit Range 15 :00 BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 :F0 Core for 07:00 Power Well: Resume for 15:08 Default: 00000h Bit Acronym Bit Description Sticky GPI NMI Enable: This bit only has effect if the corresponding GPIO is used as an input and its GPI_ROUT register is being GPI_NMI_EN programmed to NMI functionality. When set to `1', it [15:0] used to allow active-low and active-high inputs (depends on inversion bit) to cause NMI. October 2012 Order Number: 327879-001US Offset Start: 28h Offset End: 29h Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 353 6.9.1.9 Offset GPIOBASE + 2Ah: GPI_NMI_STS--GPI NMI Status Table 6-132. Offset GPIOBASE + 2Ah: GPI_NMI_STS--GPI NMI Status Description: View: PCI Size: 16 bit Bit Range 15 :00 6.9.1.10 BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 :F0 Core for 07:00 Power Well: Resume for 15:08 Default: 00000h Bit Acronym Offset Start: 2Ah Offset End: 2Bh Bit Description Sticky Bit Reset Value Bit Access GPI NMI Status: This bit is set if the corresponding GPIO is used as an input, and its GPI_ROUT register is being programmed to NMI functionality and also GPI_NMI_EN bit is set GPI_NMI_STS when it detects either: [15:0] 1) active-high edge when its corresponding GPI_INV is configured with value 0. 2) active-low edge when its corresponding GPI_INV is configured with value 1. Writing value of 1 will clear the bit, while writing value of 0 have no effect. RW Offset GPIOBASE + 2Ch: GPI_INV--GPIO Signal Invert Register Table 6-133. Offset GPIOBASE + 2Ch: GPI_INV--GPIO Signal Invert Register Description: View: PCI Size: 32 bit Bit Range 31 :00 B0:D31 Bus:Device:Function: :F0 BAR: GPIOBASE (IO) CPU I/O for 17, Power Well: Core for 16, 07:00 Default: 00000000h Bit Acronym Bit Description GP_INV[n] Input Inversion -- This bit only has effect if the corresponding GPIO is used as an input and used by the GPE logic, where the polarity matters. When set to `1', then the GPI is inverted as it is sent to the GPE logic that is using it. This bit has no effect on the value that is reported in the GP_LVL register. These bits are used to allow both active-low and activehigh inputs to cause SMI# or SCI. In the S0 or S1 state, the input signal must be active for at least two PCI clocks to ensure detection by the PCH. In the S3, S4 or S5 states the input signal must be active for at least 2 RTC clocks to ensure detection. The setting of these bits has no effect if the corresponding GPIO is programmed as an output. These bits correspond to GPI that are in the resume well, and will be reset to their default values by RSMRST# or by a write to the CF9h register. 0 = The corresponding GPI_STS bit is set when the PCH detects the state of the input pin to be high. 1 = The corresponding GPI_STS bit is set when the PCH detects the state of the input pin to be low. Intel(R) Communications Chipset 89xx Series - Datasheet 354 Offset Start: 2Ch Offset End: 2Fh Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 6.0 6.9.1.11 Offset GPIOBASE + 30h: GPIO_USE_SEL2--GPIO Use Select 2 Register Table 6-134. Offset GPIOBASE + 30h: GPIO_USE_SEL2--GPIO Use Select 2 Register Description: View: PCI Size: 32 bit Bit Range BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 Offset Start: 30h :F0 Offset End: 33h Default: 020300FEh Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Core for 23:16, 07:00 Resume for 31:24, 15:08 Bit Access Each bit in this register enables the corresponding GPIO (if it exists) to be used as a GPIO, rather than for the native function. 0 = Signal used as native function. 1 = Signal used as a GPIO. 31 :00 Notes: 1. Bits[3, 25] are always 1 because they are unMultiplexed 2. If GPIO[n] does not exist, then, the (n-32) bit in this register will always read as 0 and writes will have no effect GPIO_USE_SEL2 3. After a full reset RSMRST# all multiplexed signals in [63:32] the resume and core wells are configured as their default function. After only a PLTRST#, the GPIOs in the core well are configured as their default function. 4. When configured to GPIO mode, the muxing logic will present the inactive state to native logic that uses the pin as an input. 5. Bit 26 is ignored, functionality is configured by bits 9:8 of FLMAP0 register. 6. This register corresponds to GPIO[63:32]. Bit 0 corresponds to GPIO32 and bit 31 corresponds to GPIO63. October 2012 Order Number: 327879-001US RW Intel(R) Communications Chipset 89xx Series - Datasheet 355 6.9.1.12 Offset GPIOBASE + 34h: GP_IO_SEL2--GPIO Input/Output Select 2 Register Table 6-135. Offset GPIOBASE + 34h: GP_IO_SEL2--GPIO Input/Output Select 2 Register Description: View: PCI Size: 32 bit Bit Range 31 :00 6.9.1.13 Bus:Device:Function B0:D31 : :F0 BAR: GPIOBASE (IO) Default: 1F57FFF4h Offset Start: 34h Offset End: 37h Power Well: Bit Acronym Bit Description Sticky Bit Reset Value Core for 23:16, 07:00 Resume for 31:24, 15:08 Bit Access 0 = GPIO signal is programmed as an output. 1 = Corresponding GPIO signal (if enabled in the GP_IO_SEL2[63:32] GPIO_USE_SEL2 register) is programmed as an input. This register corresponds to GPIO[63:32]. Bit 0 corresponds to GPIO32. RW Offset GPIOBASE + 38h: GP_LVL2--GPIO Level for Input or Output 2 Register Table 6-136. Offset GPIOBASE + 38h: GP_LVL2--GPIO Level for Input or Output 2 Register Description: View: PCI Bus:Device:Function: B0:D31 :F0 BAR: GPIOBASE (IO) Offset Start: 38h Offset End: 3Bh Core for Size: 32 bit Bit Range 31 :00 Power Well: 23:16, 07:00 Resume for 31:24, 15:08 Default: A4AA0003h Bit Acronym GP_LVL[63:32] Bit Description These registers are implemented as dual read/write with dedicated storage each. Write value will be stored in the write register, while read is coming from the read register which will always reflect the value of the pin. If GPIO[n] is programmed to be an output (via the corresponding bit in the GP_IO_SEL register), then the corresponding GP_LVL[n] write register value will drive a high or low value on the output pin. 1 = high, 0 = low. When configured in native mode (GPIO_USE_SEL[n] is 0), writes to these bits are stored but have no effect to the pin value. The value reported in this register is undefined when programmed as native mode. Sticky Bit Reset Value Bit Access RW This register corresponds to GPIO[63:32]. Bit 0 corresponds to GPIO32. Intel(R) Communications Chipset 89xx Series - Datasheet 356 October 2012 Order Number: 327879-001US 6.0 6.9.1.14 Offset GPIOBASE + 40h: GPIO_USE_SEL3--GPIO Use Select 3 Register Table 6-137. Offset GPIOBASE + 40h: GPIO_USE_SEL3--GPIO Use Select 3 Register Description: View: PCI Size: 32 bit Bit Range BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym 31 :12 Offset Start: 40h Offset End: 43h Power Well: Bit Description Sticky Bit Reset Value Core for 23:16, 07:00 Resume for 31:24, 15:08 Bit Access Always 0. No corresponding GPIO. Each bit in this register enables the corresponding GPIO (if it exists) to be used as a GPIO, rather than for the native function. 0 = Signal used as native function. 1 = Signal used as a GPIO. 11 :08 Notes: 1. The following bit is always 1 because it is always unMultiplexed: 8 2. If GPIO[n] does not exist, then, the (n-32) bit in GPIO_USE_SEL3 this register will always read as 0 and writes will have [75:72] no effect. 3. After a full reset RSMRST# all multiplexed signals in the resume and core wells are configured as their default function. After only a PLTRST#, the GPIOs in the core well are configured as their default function. 4. When configured to GPIO mode, the muxing logic will present the inactive state to native logic that uses the pin as an input. This register corresponds to GPIO[95:64]. Bit 0 corresponds to GPIO64 and bit 32 corresponds to GPIO95. 07 :04 RW Always 0. No corresponding GPIO. Each bit in this register enables the corresponding GPIO (if it exists) to be used as a GPIO, rather than for the native function. 0 = Signal used as native function. 1 = Signal used as a GPIO. 03 :00 Notes: 1. If GPIO[n] does not exist, then, the (n-32) bit in this register will always read as 0 and writes will have GPIO_USE_SEL3 no effect. 2. After a full reset RSMRST# all multiplexed signals in the resume and core wells are configured as their default function. After only a PLTRST#, the GPIOs in the core well are configured as their default function. 3. When configured to GPIO mode, the muxing logic will present the inactive state to native logic that uses the pin as an input. This register corresponds to GPIO[95:64]. Bit 0 corresponds to GPIO64 and bit 32 corresponds to GPIO95. October 2012 Order Number: 327879-001US RW Intel(R) Communications Chipset 89xx Series - Datasheet 357 6.9.1.15 Offset GPIOBASE + 44h: GP_IO_SEL3--GPIO Input/Output Select 3 Register Table 6-138. Offset GPIOBASE + 44h: GP_IO_SEL3--GPIO Input/Output Select 3 Register Description: View: PCI Size: 32 bit Bit Range 31 :12 11 :08 07 :04 03 :00 BAR: GPIOBASE (IO) Bus:Device:Function: Bit Acronym Offset Start: 44h Offset End: 47h Power Well: Default: 00000F00 Bit Description Sticky Bit Reset Value Core for 23:16, 07:00 Resume for 31:24, 15:08 Bit Access Always 0. No corresponding GPIO. 0 = GPIO signal is programmed as an output. 1 = Corresponding GPIO signal (if enabled in the GPIO_IO_SEL3 GPIO_USE_SEL3 register) is [75:72] programmed as an input. This register corresponds to GPIO[95:64]. Bit 0 corresponds to GPIO64. RW Always 0. No corresponding GPIO. 0 = GPIO signal is programmed as an output. 1 = Corresponding GPIO signal (if enabled in the GPIO_USE_SEL3 register) is GPIO_IO_SEL3 programmed as an input. This register corresponds to GPIO[95:64]. Bit 0 corresponds to GPIO64. Intel(R) Communications Chipset 89xx Series - Datasheet 358 B0:D31 :F0 RW October 2012 Order Number: 327879-001US 6.0 6.9.1.16 Offset GPIOBASE + 48h: GP_LVL3--GPIO Level for Input or Output 3 Register Table 6-139. Offset GPIOBASE + 48h: GP_LVL3--GPIO Level for Input or Output 3 Register Description: View: PCI Size: 32 bit Bit Range BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym Power Well: Bit Description Sticky 31 :12 Always 0. No corresponding GPIO. 11 :08 These registers are implemented as dual read/write with dedicated storage each. Write value will be stored in the write register, while read is coming from the read register which will always reflect the value of the pin. If GPIO[n] is programmed to be an output (via the corresponding bit in the GP_IO_SEL register), then the corresponding GP_LVL[n] write register value will drive a high or low value on the output pin. 1 = high, 0 = low. When configured in native mode (GPIO_USE_SEL[n] is 0), writes to these bits are stored but have no effect to the pin value. The value reported in this register is undefined when programmed as native mode. This register corresponds to GPIO[95:64]. Bit 0 corresponds to GPIO64. GP_LVL[75:72] 07 :04 Always 0. No corresponding GPIO. 03 :00 These registers are implemented as dual read/write with dedicated storage each. Write value will be stored in the write register, while read is coming from the read register which will always reflect the value of the pin. If GPIO[n] is programmed to be an output (via the corresponding bit in the GP_IO_SEL register), then the corresponding GP_LVL[n] write register value will drive a high or low value on the output pin. 1 = high, 0 = low. When configured in native mode (GPIO_USE_SEL[n] is 0), writes to these bits are stored but have no effect to the pin value. The value reported in this register is undefined when programmed as native mode. This register corresponds to GPIO[95:64]. Bit 0 corresponds to GPIO64. GP_LVL October 2012 Order Number: 327879-001US Offset Start: 48h Offset End: 4Bh Bit Reset Value Core for 23:16, 07:00 Resume for 31:24, 15:08 Bit Access RW RW Intel(R) Communications Chipset 89xx Series - Datasheet 359 6.9.1.17 Offset GPIOBASE + 60h: GP_RST_SEL1--GPIO Reset Select Table 6-140. Offset GPIOBASE + 60h: GP_RST_SEL1--GPIO Reset Select Description: View: PCI Size: 32 bit Bit Range 31 :24 BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: 01000000h Bit Acronym Offset Start: 60h Offset End: 63h Power Well: Bit Description 0 = Corresponding GPIO registers will be reset by PWROK deassertion, CF9h reset (06h or 0Eh), or SYS_RST# assertion. GP_RST_SEL 1 = Corresponding GPIO registers will be reset by RSMRST# assertion only. [31:24] Sticky Bit Reset Value Core for 23:16, 07:00 Resume for 31:24, 15:08 Bit Access RW GPIO[24] register bits are not cleared by CF9h reset by default. 23 :16 15 :08 07 :00 Reserved Reserved 0 = Corresponding GPIO registers will be reset by PWROK deassertion, CF9h reset (06h or 0Eh), or GP_RST_SEL SYS_RST# assertion. [15:8] 1 = Corresponding GPIO registers will be reset by RSMRST# assertion only. Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 360 RW October 2012 Order Number: 327879-001US 6.0 6.9.1.18 Offset GPIOBASE + 64h: GP_RST_SEL2--GPIO Reset Select Table 6-141. Offset GPIOBASE + 64h: GP_RST_SEL2--GPIO Reset Select Description: View: PCI Size: 32 bit Bit Range 31 :24 23 :16 15 :08 07 :00 6.9.1.19 BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym Offset Start: 64h Offset End: 67h Power Well: Bit Description Sticky Bit Reset Value Core for 23:16, 07:00 Resume for 31:24, 15:08 Bit Access 0 = Corresponding GPIO registers will be reset by PWROK deassertion, CF9h reset (06h or 0Eh), or GP_RST_SEL SYS_RST# assertion. [63:56] 1 = Corresponding GPIO registers will be reset by RSMRST# assertion only. Reserved RW Reserved 0 = Corresponding GPIO registers will be reset by PWROK deassertion, CF9h reset (06h or 0Eh), or GP_RST_SEL SYS_RST# assertion. [47:40] 1 = Corresponding GPIO registers will be reset by RSMRST# assertion only. Reserved RW Reserved Offset GPIOBASE + 68h: GP_RST_SEL3--GPIO Reset Select Table 6-142. Offset GPIOBASE + 68h: GP_RST_SEL3--GPIO Reset Select Description: View: PCI Size: 32 bit Bit Range 31 :12 11 :08 07 :00 BAR: GPIOBASE (IO) Bus:Device:Function: B0:D31 :F0 Core for 23:16, Power Well: 07:00 Resume for 31:24, 15:08 Default: 00000000h Bit Acronym Offset Start: 68h Offset End: 6Bh Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved 0 = Corresponding GPIO registers will be reset by PWROK deassertion, CF9h reset (06h or 0Eh), or SYS_RST# GP_RST_SE assertion. L[75:72] 1 = Corresponding GPIO registers will be reset by RSMRST# assertion only. RW Reserved Reserved October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 361 7.0 Chipset Configuration Registers (B0:D31:F0) This section describes all registers and base functionality that is related to chipset configuration and not a specific interface (such as LPC or PCI Express*). It contains the root complex register block, which describes the behavior of the upstream internal link. This block is mapped into memory space, using the Root Complex Base Address (RCBA) register of the PCI-to-LPC bridge. Accesses in this space must be limited to 32-bit (DW) quantities. Burst accesses are not allowed. All Chipset Configuration Registers are located in the core well unless otherwise indicated. 7.1 Chipset Configuration Registers 7.1.1 Chipset Configuration Register Memory Map (Memory Space) Note: Address locations that are not shown should be treated as Reserved. Table 7-1. Chipset Configuration Registers (Sheet 1 of 3) Offset Start Offset End Default Value Register ID - Description 0014h 0017h "Offset 0014h: V0CTL--Virtual Channel 0 Resource Control Register" on page 365 80000023h 001Ah 001Bh "Offset 001Ah: V0STS--Virtual Channel 0 Resource Status Register" on page 366 0000h 001Ch 001Fh "Offset 001Ch: V1CAP--Virtual Channel 1 Resource Capability Register" on page 366 00008001h 0020h 0023h "Offset 0020h: V1CTL--Virtual Channel 1 Resource Control Register" on page 367 00000000h 0026h 0027h "Offset 0026h: V1STS--Virtual Channel 1 Resource Status Register" on page 367 0000h 0050h 0053h "Offset 0050h: CIR0--Chipset Initialization Register 0" on page 368 00000000h 0088h 008Bh "Offset 0088h: CIR1--Chipset Initialization Register 1" on page 368 00000000h 00ACh 00AFh "Offset 00ACh: REC--Root Error Command Register" on page 369 0000h 01A0h 01A3h "Offset 01A0h: ILCL--Internal Link Capabilities List Register" on page 369 00010006h 01A4h 01A7h "Offset 01A4h: LCAP--Link Capabilities Register" on page 370 00012841h 01A8h 01A9h "Offset 01A8h: LCTL--Link Control Register" on page 370 0000h 01AAh 01ABh "Offset 01AAh: LSTS--Link Status Register" on page 371 0041h 0220h 0223h "Offset 0220h: BCR--Backbone Configuration Register" on page 371 00000000h "Offset 0224h: RPC--Root Port Configuration Register" on page 372 0000000yh (y = 00xxb) 0224h 0227h 0234h 0237h "Offset 0234h: DMIC--DMI Control Register" on page 373 00000000h 0238h 023Ch "Offset 0238h: RPFN--Root Port Function Number and Hide for PCI Express* Root Ports" on page 373 76543210h 0290h 0293h "Offset 0290h: Reserved" on page 374 00000000h Intel(R) Communications Chipset 89xx Series - Datasheet 362 October 2012 Order Number: 327879-001US 7.0 Table 7-1. Offset Start Chipset Configuration Registers (Sheet 2 of 3) Offset End Register ID - Description Default Value 1D40h 1D47h "Offset 1D40H: CIR5--Chipset Initialization Register 5" on page 375 00000000000 000h 1E00h 1E03h "Offset 1E00h: TRSR--Trap Status Register" on page 375 00000000h 1E10h 1E17h "Offset 1E10h: TRCR--Trapped Cycle Register" on page 376 00000000000 00000h 1E18h 1E1Fh "Offset 1E18h: TWDR--Trapped Write Data Register" on page 376 00000000000 00000h 1E80h 1E9Fh "Offset 1E80h: IOTRn--I/O Trap Register (0-3)" on page 377 00000000000 00000h 2010h 2013h "Offset 2010h: DMC--DMI Miscellaneous Control Register" on page 378 00000002h 2024h 2027h "Offset 2024h: CIR6--Chipset Initialization Register 6" on page 378 0B4030C0h 2324h 2327h "Offset 2324h: DMC2--DMI Miscellaneous Control Register 2" on page 379 0FFF0FFFh 2330h 2333h "Offset 60h: SBI Unified AFE Address Register (SATA-B0:D31:F2)" on page 379 00000000h 2334h 2337h "Offset 64h: SBI Unified AFE Data Register (SATA-B0:D31:F2)" on page 380 00000000h 2338h 2339h "Offset 68h: SBI Unified AFE Status Register (SATA-B0:D31:F2)" on page 381 0000h 3000h 3000h "Offset 3000h: TCTL--TCO Configuration Register" on page 382 00h 3100h 3103h "Offset 3100h: D31IP--Device 31 Interrupt Pin Register" on page 383 03243200h 3108h 310Bh "Offset 3108h: D29IP--Device 29 Interrupt Pin Register" on page 384 10004321h 310Ch 310Fh "Offset 310Ch: D28IP--Device 28 Interrupt Pin Register" on page 385 00214321h "Offset 3124h: D22IP--Device 22 Interrupt Pin Register" on page 386 See register description 3124h 3127h 3140h 3127h "Offset 3140h: D31IR--Device 31 Interrupt Route Register" on page 387 3210h 3144h 3145h "Offset 3144h: D29IR--Device 29 Interrupt Route Register" on page 388 3210h 3146h 3147h "Offset 3146h: D28IR--Device 28 Interrupt Route Register" on page 389 3210h 315Ch 315Dh "Offset 315Ch: D22IR--Device 22 Interrupt Route Register" on page 390 3210h 31FEh 31FFh "Offset 31FEh: OIC--Other Interrupt Control Register" on page 391 0000h 3310h 3313h "Offset 3310h: PRSTS--Power and Reset Status" on page 392 02020000h 3314h 3317h "Offset 3314h: CIR7--Chipset Initialization Register 7" on page 393 00000000h 3324h 3327h "Offset 3324h: CIR8--Chipset Initialization Register 8" on page 393 00000000h 3330h 3333h "Offset 3330h: CIR9--Chipset Initialization Register 9" on page 393 00000000h 3340h 3343h "Offset 3340h: CIR10--Chipset Initialization Register 10" on page 394 00000000h 3350h 3353h "Offset 3350h: CIR13--Chipset Initialization Register 13" on page 394 00000000h 3368h 336Bh "Offset 3368h: CIR14--Chipset Initialization Register 14" on page 394 00000000h 3378h 337Bh "Offset 3378h: CIR15--Chipset Initialization Register 15" on page 395 00000000h 3388h 338Bh "Offset 3388h: CIR16--Chipset Initialization Register 16" on page 395 00000000h 33A0h 33A3h "Offset 33A0h: CIR17--Chipset Initialization Register 17" on page 395 00000000h 33A8h 33ABh "Offset 33A8h: CIR18--Chipset Initialization Register 18" on page 396 00000000h 33C0h 33C3h "Offset 33C0h: CIR19--Chipset Initialization Register 19" on page 396 00000000h 33CCh 33CFh "Offset 33CCh: CIR20--Chipset Initialization Register 20" on page 396 00000000h 33D0h 33D3h "Offset 33D0h: CIR21--Chipset Initialization Register 21" on page 397 00000000h 33D4h 33D7h "Offset 33D4h: CIR22--Chipset Initialization Register 22" on page 397 00000000h 3400h 3403h "Offset 3400h: RC--RTC Configuration Register" on page 398 00000000h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 363 Table 7-1. Offset Start 3404h Chipset Configuration Registers (Sheet 3 of 3) Offset End 3407h Default Value Register ID - Description "Offset 3404h: HPTC--High Precision Timer Configuration Register" on page 399 00000000h 00000yy0h (yy = xx0000x0b) 3410h 3413h "Offset 3410h: GCS--General Control and Status Register" on page 400 3414h 3414h "Offset 3414h: BUC--Backed Up Control Register" on page 402 0000000xb "Offset 3418h: FD--Function Disable Register" on page 403 See register description 3418h 341Bh 341Ch 341Fh "Offset 341Ch: CG--Clock Gating" on page 405 00000000h 3420h 3420h "Offset 3420h: FDSW--Function Disable SUS Well" on page 406 00h 3428h 342Bh "Offset 3428h: FD2--Function Disable 2" on page 406 00000000h 3500h 3514h "Offset 3500h: USBIR[0:5]--USB Initialization Register [0-5]" on page 407 20000B4Ah 3564h 3567h "Offset 3564h: USBIRC--USB Initialization Register C" on page 408 0000371Bh 3570h 3573h "Offset 3570h: USBIRA--USB Initialization Register A" on page 408 00000000h 357Ch 357Fh "Offset 357Ch: USBIRB--USB Initialization Register B" on page 409 0001C000h 3590h 3594h "Offset 3590h: MISCCTL--Miscellaneous Control Register" on page 410 00000000h 359Ch 359Fh "Offset 359Ch: PDO--USB Port Disable Override" on page 411 0000h 35A0h 35A3h "Offset 35A0h: USBOCM1--Overcurrent MAP Register 1" on page 412 C0300C03h 35B0h 35B3h "Offset 35B0h: RMHWKCTL--Rate Matching Hub Wake Control Register" on page 413 00000000h Intel(R) Communications Chipset 89xx Series - Datasheet 364 October 2012 Order Number: 327879-001US 7.0 7.1.1.1 Offset 0014h: V0CTL--Virtual Channel 0 Resource Control Register (B0:D31:F0) Table 7-2. Offset 0014h: V0CTL--Virtual Channel 0 Resource Control Register Description: View: PCI Size: 32 bit Bit Range 31 BAR: RCBA Bus:Device:Function: Default: 80000023h Bit Acronym EN 30 :27 Reserved 26 :24 ID 23 :16 Reserved Bit Description Sticky Virtual Channel Enable -- Always set to 1. VC0 is always enabled and cannot be disabled. Virtual Channel Identifier -- Indicates the ID to use for this virtual channel. Bit Access RO RO Reserved ETVN 07 :01 TVM Transaction Class / Virtual Channel Map -- Indicates which transaction classes are mapped to this virtual channel. When a bit is set, this transaction class is mapped to the virtual channel. October 2012 Order Number: 327879-001US Bit Reset Value Reserved 15 :08 Reserved Offset Start: 0014h Offset End: 0017h Power Well: Extended TC/VC Map -- Defines the upper 8-bits of the VC0 16-bit TC/VC mapping registers. These registers use the PCI Express* reserved TC[3] traffic class bit. 00 B0:D31 :F0 RW Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 365 7.1.1.2 Offset 001Ah: V0STS--Virtual Channel 0 Resource Status Register (B0:D31:F0) Table 7-3. Offset 001Ah: V0STS--Virtual Channel 0 Resource Status Register Description: View: PCI Size: 16 bit BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 0000h Bit Range Bit Acronym 15 :02 Reserved Offset Start: 001Ah Offset End: 001Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 01 NP VC Negotiation Pending -- When set, indicates the virtual channel is still being negotiated with ingress ports. RO 00 ATS Port Arbitration Tables Status -- There is no port arbitration table for this VC, so this bit is reserved at 0. RO 7.1.1.3 Offset 001Ch: V1CAP--Virtual Channel 1 Resource Capability Register (B0:D31:F0) Table 7-4. Offset 001Ch: V1CAP--Virtual Channel 1 Resource Capability Register Description: View: PCI Size: 32 bit Default: 00008001h Bit Range Bit Acronym 31 :24 AT 23 Bus:Device:Function: B0:D31 :F0 BAR: RCBA Reserved Power Well: Bit Description Port Arbitration Table Offset -- Indicates the location of the port arbitration table in the root complex. A value of 0h indicates the table is not present Sticky Bit Reset Value Bit Access RO Reserved MTS Maximum Time Slots -- This value is updated by platform BIOS based upon the determination of the number of time slots available in the platform. RWO 15 RTS Reject Snoop Transactions -- All snoopable transactions on VC1 are rejected. RO 14 APS Advanced Packet Switching -- This VC is capable of all transactions, not just advanced packet switching transactions. RO 22 :16 13 :08 Reserved 07 :00 PAC Reserved Port Arbitration Capability -- Indicates the port arbitration capability is hardware-fixed. Intel(R) Communications Chipset 89xx Series - Datasheet 366 Offset Start: 001Ch Offset End: 001Fh October 2012 Order Number: 327879-001US 7.0 7.1.1.4 Offset 0020h: V1CTL--Virtual Channel 1 Resource Control Register (B0:D31:F0) Table 7-5. Offset 0020h: V1CTL--Virtual Channel 1 Resource Control Register Description: View: PCI Size: 32 bit Bit Range 31 BAR: RCBA Bus:Device:Function: Default: 00000000h Bit Acronym Offset Start: 0020h Offset End: 0023h Power Well: Bit Description Sticky Bit Reset Value Bit Access Virtual Channel Enable -- Enables the VC when set. Disables the VC when cleared. EN 30 :28 Reserved 27 :24 ID 23 :16 Reserved RW Reserved Virtual Channel Identifier -- Indicates the ID to use for this virtual channel. 15 :08 ETVN 07 :01 TVM Transaction Class / Virtual Channel Map -- Indicates which transaction classes are mapped to this virtual channel. When a bit is set, this transaction class is mapped to the virtual channel. Reserved RO Reserved Extended TC/VC Map -- Defines the upper 8-bits of the VC0 16-bit TC/VC mapping registers. These registers use the PCI Express* reserved TC[3] traffic class bit. 00 B0:D31 :F0 RW Reserved 7.1.1.5 Offset 0026h: V1STS--Virtual Channel 1 Resource Status Register (B0:D31:F0) Table 7-6. Offset 0026h: V1STS--Virtual Channel 1 Resource Status Register Description: View: PCI Size: 16 bit BAR: RCBA B0:D31 Bus:Device:Function: :F0 Default: 0000h Bit Range Bit Acronym 15 :02 Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 01 NP VC Negotiation Pending -- When set, indicates the virtual channel is still being negotiated with ingress ports. 00 ATS Port Arbitration Tables Status -- Indicates the coherency status of the port arbitration table. This bit is set when LAT (offset 000Ch:bit 0) is written with value 1 and PAS (offset 0014h:bits19:17) has value of 4h. This bit is cleared after the table has been updated. October 2012 Order Number: 327879-001US Offset Start: 0026h Offset End: 0027h RO RO Intel(R) Communications Chipset 89xx Series - Datasheet 367 7.1.1.6 Offset 0050h: CIR0--Chipset Initialization Register 0 (B0:D31:F0) Table 7-7. Offset 0050h: CIR0--Chipset Initialization Register 0 Description: View: PCI Size: 32 bit Bit Range BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym 31 :00 Offset Start: 0050h Offset End: 0053h Power Well: Bit Description Sticky Bit Reset Value CIR0 Field 0 -- BIOS must set this field. Bit Access RW 7.1.1.7 Offset 0088h: CIR1--Chipset Initialization Register 1 (B0:D31:F0) Table 7-8. Offset 0088h: CIR1--Chipset Initialization Register 1 Description: View: PCI Size: 32 bit Default: 00000000h Bit Range Bit Acronym 31 :21 Reserved Reserved Reserved Reserved Reserved Reserved Reserved Reserved 20 19 :16 Bit Description CIR1 Field 2 -- BIOS must set this bit. 12 11 :00 CIR1 Field 1 -- BIOS must set this bit. Intel(R) Communications Chipset 89xx Series - Datasheet 368 Offset Start: 0088h Offset End: 008Bh Power Well: CIR1 Field 3 -- BIOS must set this bit. 15 14 :13 B0:D31 Bus:Device:Function: :F0 BAR: RCBA Sticky Bit Reset Value Bit Access RWO RWO RWO October 2012 Order Number: 327879-001US 7.0 7.1.1.8 Offset 00ACh: REC--Root Error Command Register (B0:D31:F0) Table 7-9. Offset 00ACh: REC--Root Error Command Register Description: View: PCI Size: 32 bit Bit Range 31 30 :00 7.1.1.9 BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 0000h Power Well: Bit Acronym Bit Description DPDP Drop Poisoned Downstream Packets -- Determines how downstream packets on DMI are handled that are received with the EP field set, indicating poisoned data: 1 = This packet and all subsequent packets with data received on DMI for any VC will have their Unsupported Transaction (UT) field set causing them to master Abort downstream. Packets without data such as memory, IO and config read requests are allowed to proceed. 0 = Packets are forwarded downstream without forcing the UT field set. Reserved Offset Start: 00ACh Offset End: 00AFh Sticky Bit Reset Value Bit Access RW Reserved Offset 01A0h: ILCL--Internal Link Capabilities List Register (B0:D31:F0) Table 7-10. Offset 01A0h: ILCL--Internal Link Capabilities List Register Description: View: PCI Size: 32 bit BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 00010006h Offset Start: 01A0h Offset End: 01A3h Power Well: Sticky Bit Access Bit Acronym 31 :20 NEXT Next Capability Offset -- Indicates this is the last item in the list. RO 19 :16 CV Capability Version -- Indicates the version of the capability structure. RO 15 :00 CID Capability ID -- Indicates this is capability for DMI. RO October 2012 Order Number: 327879-001US Bit Description Bit Reset Value Bit Range Intel(R) Communications Chipset 89xx Series - Datasheet 369 7.1.1.10 Offset 01A4h: LCAP--Link Capabilities Register (B0:D31:F0) Table 7-11. Offset 01A4h: LCAP--Link Capabilities Register Description: View: PCI Size: 32 bit Bit Range 31 :18 17 :15 BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 00012841h Bit Acronym Offset Start: 01A4h Offset End: 01A7h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved EL1 L1 Exit Latency -- L1 is supported on DMI. 14 :12 EL0 L0s Exit Latency -- This field indicates that exit latency is 128 ns to less than 256 ns. RWO 11 :10 APMS Active State Link PM Support -- Indicates that L0s and L1 are supported on DMI. RWO 09 :04 MLW Maximum Link Width -- Indicates the maximum link width is 4 ports. RO 03 :00 MLS Maximum Link Speed -- Indicates the link speed is 2.5 Gb/s. RO 7.1.1.11 RO Offset 01A8h: LCTL--Link Control Register (B0:D31:F0) Table 7-12. Offset 01A8h: LCTL--Link Control Register Description: View: PCI Size: 32 bit Default: 0000h Bit Range Bit Acronym 15 :08 Reserved 07 06 :02 01 :00 Bus:Device:Function: B0:D31 :F0 BAR: RCBA ES Reserved ASPM Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Extended Synch -- When set, forces extended transmission of FTS ordered sets when exiting L0s prior to entering L0. RW Reserved Active State Link PM Control -- Indicates whether DMI should enter L0s. 00 = Disabled 01 = L0s entry enabled 10 = Reserved 11 = Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 370 Offset Start: 01A8h Offset End: 01A9h RW October 2012 Order Number: 327879-001US 7.0 7.1.1.12 Offset 01AAh: LSTS--Link Status Register (B0:D31:F0) Table 7-13. Offset 01AAh: LSTS--Link Status Register Description: View: PCI Size: 16 bit BAR: RCBA B0:D31 :F0 Default: 0041h Bit Range Bit Acronym 15 :10 Reserved 09 :04 NLW 03 :00 LS 7.1.1.13 Bus:Device:Function: Offset Start: 01AAh Offset End: 01ABh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Negotiated Link Width -- Negotiated link width is x4 (000100b). RO Link Speed -- Link is 2.5 Gb/s. RO Offset 0220h: BCR--Backbone Configuration Register (B0:D31:F0) Table 7-14. Offset 0220h: BCR--Backbone Configuration Register Description: View: PCI Size: 32 bit BAR: RCBA Default: 00000000h Bit Range Bit Acronym 31 :07 Reserved 06 05 :03 Bus:Device:Function: B0:D31 :F0 Power Well: Bit Description Sticky Reserved October 2012 Order Number: 327879-001US Bit Reset Value Bit Access Reserved BCR Field 2 -- BIOS must set this bit. 02 :00 Offset Start: 0220h Offset End: 0223h RW Reserved BCR Field 1 -- BIOS program this field to 101b RW Intel(R) Communications Chipset 89xx Series - Datasheet 371 7.1.1.14 Offset 0224h: RPC--Root Port Configuration Register (B0:D31:F0) Table 7-15. Offset 0224h: RPC--Root Port Configuration Register Description: View: PCI Size: 32 bit BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 0000000yh (y = 00xxb) Bit Range Bit Acronym 31 :08 Reserved Bit Description Offset Start: 0224h Offset End: 0227h Power Well: Sticky Bit Reset Value Bit Access Reserved HPE High Priority Port Enable: 0 = The high priority path is not enabled. 1 = The port selected by the HPP field in this register is enabled for high priority. It will be arbitrated above all other VC0 (including integrated VC0) devices. RW 06 :04 HPP High Priority Port -- This controls which port is enabled for high priority when the HPE bit in this register is set. 101 = Port 4 010 = Port 3 001 = Port 2 000 = Port 1 RW 03 :02 Reserved 07 Reserved Port Configuration -- This controls how the PCI bridges are organized in various modes of operation for Ports 1-4. For the following mappings, if a port is not shown, it is considered a x1 port with no connection. These bits are set by the PCIEPCS1[1:0] soft strap. 01 :00 PC 11 = 1 x4, Port 1 (x4) 10 = 2x2, Port 1 (x2), Port 3 (x2) 01 = 1x2 and 2x1s, Port 1 (x2), Port 3 (x1) and Port 4 (x1) 00 = 4 x1s, Port 1 (x1), Port 2 (x1), Port 3 (x1) and Port 4 (x1) These bits are in the resume well and are only reset by RSMRST#. Intel(R) Communications Chipset 89xx Series - Datasheet 372 RO October 2012 Order Number: 327879-001US 7.0 7.1.1.15 Offset 0234h: DMIC--DMI Control Register (B0:D31:F0) Table 7-16. Offset 0234h: DMIC--DMI Control Register Description: View: PCI Size: 32 bit BAR: RCBA Bus:Device:Function: Default: 00000000h Offset Start: 0234h Offset End: 0237h Power Well: Bit Range Bit Acronym 31 :02 Reserved Reserved 01 :00 DMICGEN DMI Clock Gate Enable -- BIOS must program this field to 11b. 7.1.1.16 B0:D31 :F0 Bit Description Sticky Bit Reset Value Bit Access RW Offset 0238h: RPFN--Root Port Function Number and Hide for PCI Express* Root Ports (B0:D31:F0) For the PCI Express* root ports, the assignment of a function number to a root port is not fixed. BIOS may re-assign the function numbers on a port by port basis. This capability will allow BIOS to disable/hide any root port and still have functions 0 thru N1 where N is the total number of enabled root ports. Port numbers will remain fixed to a physical root port. The existing root port Function Disable registers operate on physical ports (not functions). Port Configuration (1x4, 4x1, etc.) is not affected by the logical function number assignment and is associated with physical ports. Table 7-17. Offset 0238h: RPFN--Root Port Function Number and Hide for PCI Express* Root Ports Description: View: PCI Size: 32 bit BAR: RCBA Default: 76543210h Bit Range Bit Acronym 31 :16 Reserved 15 14 :12 Bus:Device:Function: B0:D31 :F0 Offset Start: 0238h Offset End: 023Ch Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved RP4CH Root Port 4 Config Hide -- This bit is used to hide the root port and any devices behind it from being discovered by the OS. When set to `1' the root port will not claim any downstream configuration transactions. RWO RP4FN Root Port 4 Function Number -- These bits set the function number for PCI Express* Root Port 4. This root port function number must be a unique value from the other root port function number. RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 373 Table 7-17. Offset 0238h: RPFN--Root Port Function Number and Hide for PCI Express* Root Ports Description: View: PCI Size: 32 bit Bit Range 11 10 :08 07 06 :04 03 02 :00 7.1.1.17 BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 76543210h Bit Acronym RP4CH Offset Start: 0238h Offset End: 023Ch Power Well: Bit Description Sticky Bit Reset Value Root Port 3 Config Hide -- This bit is used to hide the root port and any devices behind it from being discovered by the OS. When set to `1' the root port will not claim any downstream configuration transactions. Bit Access RWO RP3FN Root Port 3 Function Number -- These bits set the function number for PCI Express* Root Port 3. This root port function number must be a unique value from the other root port function number. RW RP3CH Root Port 2 Config Hide -- This bit is used to hide the root port and any devices behind it from being discovered by the OS. When set to `1' the root port will not claim any downstream configuration transactions. RWO RP2FN Root Port 2 Function Number -- These bits set the function number for PCI Express* Root Port 2. This root port function number must be a unique value from the other root port function number. RW RP2CH Root Port 1 Config Hide -- This bit is used to hide the root port and any devices behind it from being discovered by the OS. When set to `1' the root port will not claim any downstream configuration transactions. RWO RP1FN Root Port 1 Function Number -- These bits set the function number for PCI Express* Root Port 1. This root port function number must be a unique value from the other root port function number. RW Offset 0290h: Reserved Register (B0:D31:F0) Table 7-18. Offset 0290h: Reserved Description: View: PCI Size: 32 bit B0:D31 Bus:Device:Function: :F0 BAR: RCBA Default: 00000000h Bit Range Bit Acronym 31 :00 Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 374 Offset Start: 0290h Offset End: 0293h October 2012 Order Number: 327879-001US 7.0 7.1.1.18 Offset 1D40H: CIR5--Chipset Initialization Register 5 (B0:D31:F0) Table 7-19. Offset 1D40H: CIR5--Chipset Initialization Register 5 Description: View: PCI Size: 64 bit BAR: RCBA B0:D31 :F0 Default: 00000000000000h Bit Range Bit Acronym 63 :01 Reserved 00 7.1.1.19 Bus:Device:Function: Offset Start: 1D40h Offset End: 1D47h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved CIR5 Field 1 -- BIOS must program this field to 1b. RW Offset 1E00h: TRSR--Trap Status Register (B0:D31:F0) Table 7-20. Offset 1E00h: TRSR--Trap Status Register Description: View: PCI Size: 32 bit BAR: RCBA Default: 00000000h Bit Range Bit Acronym 31 :04 Reserved 03 :00 Bus:Device:Function: B0:D31 :F0 CTSS October 2012 Order Number: 327879-001US Offset Start: 1E00h Offset End: 1E03h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Cycle Trap SMI# Status -- These bits are set by hardware when the corresponding Cycle Trap register is enabled and a matching cycle is received (and trapped). These bits are OR'ed together to create a single status bit in the Power Management register space. The SMI# and trapping must be enabled in order to set these bits. These bits are set before the completion is generated for the trapped cycle, thereby ensuring that the processor can enter the SMI# handler when the instruction completes. Each status bit is cleared by writing a 1 to the corresponding bit location in this register. RWC Intel(R) Communications Chipset 89xx Series - Datasheet 375 7.1.1.20 Offset 1E10h: TRCR--Trapped Cycle Register (B0:D31:F0) This register saves information about the I/O Cycle that was trapped and generated the SMI# for software to read. Table 7-21. Offset 1E10h: TRCR--Trapped Cycle Register Description: View: PCI Size: 64 bit BAR: RCBA B0:D31 :F0 Default: 0000000000000000h Bit Range Bit Acronym 63 :25 Reserved 24 Bus:Device:Function: Offset Start: 1E10h Offset End: 1E17h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Read/Write#: 0 = Trapped cycle was a write cycle. 1 = Trapped cycle was a read cycle. RWI RO 23 :20 Reserved 19 :16 AHBE Active-high Byte Enables -- This is the dword-aligned byte enables associated with the trapped cycle. A 1 in any bit location indicates that the corresponding byte is enabled in the cycle. RO 15 :02 TIOA Trapped I/O Address -- This is the dword-aligned address of the trapped cycle. RO 01 :00 Reserved 7.1.1.21 Reserved Reserved Offset 1E18h: TWDR--Trapped Write Data Register (B0:D31:F0) This register saves the data from I/O write cycles that are trapped for software to read. Table 7-22. Offset 1E18h: TWDR--Trapped Write Data Register Description: View: PCI Size: 64 bit Bus:Device:Function: B0:D31 :F0 BAR: RCBA Default: 0000000000000000h Bit Range Bit Acronym 63 :32 Reserved 31 :00 TIOD Bit Description Power Well: Sticky Bit Reset Value Bit Access Reserved Trapped I/O Data -- Dword of I/O write data. This field is undefined after trapping a read cycle. Intel(R) Communications Chipset 89xx Series - Datasheet 376 Offset Start: 1E18h Offset End: 1E1Fh RO October 2012 Order Number: 327879-001US 7.0 7.1.1.22 Offset 1E80h: IOTRn--I/O Trap Register (0-3) (B0:D31:F0) Offset Address: 1E80h-1E87h Register 0 1E88h-1E8Fh Register 1 1E90h-1E97h Register 2 1E98h-1E9Fh Register 3 These registers are used to specify the set of I/O cycles to be trapped and to enable this functionality. Table 7-23. Offset 1E80h: IOTRn--I/O Trap Register (0-3) Description: View: PCI Size: 64 bit B0:D31 Bus:Device:Function: :F0 BAR: RCBA Default: 0000000000000000h Bit Range Bit Acronym 63 :50 Reserved Offset Start: 1E80h Offset End: 1E9Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 49 RWM Read/Write Mask: 0 = The cycle must match the type specified in bit 48. 1 = Trapping logic will operate on both read and write cycles. RW 48 RWIO Read/Write#: 0 = Write 1 = Read The value in this field does not matter if bit 49 is set. RW 47 :40 Reserved Reserved 39 :36 BEM Byte Enable Mask -- A 1 in any bit position indicates that any value in the corresponding byte enable bit in a received cycle will be treated as a match. The corresponding bit in the Byte Enables field, below, is ignored. 35 :32 TBE Byte Enables (TBE) -- R/W. Active-high dword-aligned byte enables. 31 :24 Reserved 23 :18 ADMA 17 :16 Reserved 15 :02 IOAD 01 00 Reserved TRSE October 2012 Order Number: 327879-001US RW RW Reserved Address[7:2] Mask -- A 1 in any bit position indicates that any value in the corresponding address bit in a received cycle will be treated as a match. The corresponding bit in the Address field, below, is ignored. The mask is only provided for the lower 6 bits of the dword address, allowing for traps on address ranges up to 256 bytes in size. RW Reserved I/O Address[15:2] -- Dword-aligned address RW Reserved Trap and SMI# Enable: 0 = Trapping and SMI# logic disabled. 1 = The trapping logic specified in this register is enabled. RW Intel(R) Communications Chipset 89xx Series - Datasheet 377 7.1.1.23 Offset 2010h: DMC--DMI Miscellaneous Control Register (B0:D31:F0) Table 7-24. Offset 2010h: DMC--DMI Miscellaneous Control Register Description: View: PCI BAR: RCBA Size: 32 bit B0:D31 :F0 Default: 00000002h Bit Range Bit Acronym 31 :20 Reserved 19 Offset Start: 2010h Offset End: 2013h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved DMI Misc. Control Field 1 -- BIOS shall always program this field as per the BIOS Specification. 0 = Disable DMI Power Savings. 1 = Enable DMI Power Savings. DMC 18 :00 7.1.1.24 Bus:Device:Function: Reserved RW Reserved Offset 2024h: CIR6--Chipset Initialization Register 6 (B0:D31:F0) Table 7-25. Offset 2024h: CIR6--Chipset Initialization Register 6 Description: View: PCI Size: 32 bit Bit Range 31 :24 Default: 0B4030C0h Bit Acronym Reserved 23 :21 Reserved 31 :08 Reserved 07 Bit Description Bit Reset Value Bit Access RW Reserved Reserved RW Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 378 Sticky Reserved CIR6 Field 1 -- BIOS must clear this bit. Reserved Offset Start: 2024h Offset End: 2027h Power Well: CIR6 Field 2 -- BIOS must program this field to 011b. 20 :08 06 :00 Bus:Device:Function: B0:D31 :F0 BAR: RCBA October 2012 Order Number: 327879-001US 7.0 7.1.1.25 Offset 2324h: DMC2--DMI Miscellaneous Control Register 2 (B0:D31:F0) Table 7-26. Offset 2324h: DMC2--DMI Miscellaneous Control Register 2 Description: View: PCI BAR: RCBA Size: 32 bit B0:D31 :F0 Default: 0FFF0FFFh Bit Range Bit Acronym 31 :28 Reserved 27 :16 DMC2 15 :00 Reserved 7.1.1.26 Bus:Device:Function: Offset Start: 2324h Offset End: 2327h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved DMI Misc. Control Field 2 -- BIOS shall always program this field as per the BIOS Specification. RW Reserved Offset 60h: SBI Unified AFE Address Register Table 7-27. Offset 60h: SBI Unified AFE Address Register (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :16 15 :00 BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym Offset Start: 2330h Offset End: 2333h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved IAddress Offset: Register address offset. The contents of this register are sent in the IOSF OFFSET Sideband Message Register Access address[15:0] field. See the MPHY chapters for register address offsets. October 2012 Order Number: 327879-001US 0000h RW Intel(R) Communications Chipset 89xx Series - Datasheet 379 7.1.1.27 Offset 64h: SSBI Unified AFE Data Register Table 7-28. Offset 64h: SBI Unified AFE Data Register (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: RCBA Bus:Device:Function: Default: 00000000h Offset Start: 2334h Offset End: 2337h Power Well: Bit Acronym Bit Description DATA AFE Data: Reads and Writes to this register will trigger an IOSF Sideband access on the private Unified AFE sideband interface. All MPHY accesses are sent as IOSF non-posted transactions. Software needs to ensure the MPHY interface is not busy NOT(MPHYSTAT.BUSY) before writing to this register. A read to this register will trigger a read of the designated MPHY register on the private IOSF sideband interface and when a completion is returned on the sideband interface, a completion can be returned on the IOSF primary interface. Implementation Note: The DMI registers are in memory mapped space. Posted writes to this register will cause a non-posted transaction to occur on the MPHY private sideband interface. The register block must not block accepting of posted writes to the general DMI register space while waiting for a non-posted transaction to complete on the IOSF sideband interface. If software erroneously issues multiple posted writes to this register while MPHYSTAT.BUSY is `1', then the register block is allowed to block accepting of subsequent posted transactions while waiting for the IOSF sideband transaction to complete. Intel(R) Communications Chipset 89xx Series - Datasheet 380 B0:D31 :F0 Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 7.0 7.1.1.28 Offset 68h: SBI Unified AFE Status Register Table 7-29. Offset 68h: SBI Unified AFE Status Register (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range 15 :03 02 :01 00 BAR: RCBA Bus:Device:Function: Default: 0000h Bit Acronym B0:D31 :F0 Offset Start: 2338h Offset End: 2339h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved RESP Response Status (RESPONSE): 00 - Successful 01 - Unsuccessful / Not Supported 10 - Reserved 11 - Powered Down This register reflects the response status for the previously completed transaction. The value of this register is only meaningful if NOT(MPHYSTAT.BUSY). 00b RO BUSY Busy / Ready#: `0': The MPHY IOSF sideband interface is ready for a new transaction `1': The MPHY IOSF sideband interface is busy with the previous transaction. 0b RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 381 7.1.1.29 Offset 3000h: TCTL--TCO Configuration Register (B0:D31:F0) Table 7-30. Offset 3000h: TCTL--TCO Configuration Register Description: View: PCI Size: 8 bit Bit Range 07 06 :03 02 :00 BAR: RCBA Bus:Device:Function: Default: 00h Bit Acronym Bit Description IE TCO IRQ Enable: 0 = TCO IRQ is disabled. 1 = TCO IRQ is enabled, as selected by the TCO_IRQ_SEL field. Reserved IS Offset Start: 3000h Offset End: 3000h Power Well: Sticky Bit Reset Value Bit Access RW Reserved TCO IRQ Select -- Specifies on which IRQ the TCO will internally appear. If not using the APIC, the TCO interrupt must be routed to IRQ9-11, and that interrupt is not sharable with the SERIRQ stream, but is shareable with other PCI interrupts. If using the APIC, the TCO interrupt can also be mapped to IRQ20-23, and can be shared with other interrupt. 000 = IRQ 9 001 = IRQ 10 010 = IRQ 11 011 = Reserved 100 = IRQ 20 (only if APIC enabled) 101 = IRQ 21 (only if APIC enabled) 110 = IRQ 22 (only if APIC enabled) 111 = IRQ 23 (only if APIC enabled) When setting the these bits, the IE bit should be cleared to prevent glitching. When the interrupt is mapped to APIC interrupts 9, 10 or 11, the APIC should be programmed for active-high reception. When the interrupt is mapped to APIC interrupts 20 through 23, the APIC should be programmed for active-low reception. Intel(R) Communications Chipset 89xx Series - Datasheet 382 B0:D31 :F0 RW October 2012 Order Number: 327879-001US 7.0 7.1.1.30 Offset 3100h: D31IP--Device 31 Interrupt Pin Register (B0:D31:F0) Table 7-31. Offset 3100h: D31IP--Device 31 Interrupt Pin Register Description: View: PCI Size: 32 bit BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 03243200h Bit Range Bit Acronym 31 :28 Reserved Offset Start: 3100h Offset End: 3103h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved TSIP Thermal Sensor Pin -- Indicates which pin the Thermal Sensor controller drives as its interrupt 0h = No interrupt 1h = INTA# 2h = INTB# (Default) 3h = INTC# 4h = INTD# 5h-Fh = Reserved RW 23 :20 SIP2 SATA Pin 2 -- Indicates which pin the SATA controller 2 drives as its interrupt. 0h = No interrupt 1h = INTA# 2h = INTB# (Default) 3h = INTC# 4h = INTD# 5h-Fh = Reserved RW 19 :16 Reserved 27 :24 15 :12 SMIP 11 :08 SIP 07 :04 Reserved 03 :00 LIP October 2012 Order Number: 327879-001US Reserved SMBus Pin -- Indicates which pin the SMBus controller drives as its interrupt. 0h = No interrupt 1h = INTA# 2h = INTB# (Default) 3h = INTC# 4h = INTD# 5h-Fh = Reserved RW SATA Pin -- Indicates which pin the SATA controller drives as its interrupt. 0h = No interrupt 1h = INTA# 2h = INTB# (Default) 3h = INTC# 4h = INTD# 5h-Fh = Reserved RW Reserved LPC Bridge Pin -- Currently, the LPC bridge does not generate an interrupt, so this field is read-only and 0. RO Intel(R) Communications Chipset 89xx Series - Datasheet 383 7.1.1.31 Offset 3108h: D29IP--Device 29 Interrupt Pin Register (B0:D31:F0) Table 7-32. Offset 3108h: D29IP--Device 29 Interrupt Pin Register Description: View: PCI Size: 32 bit BAR: RCBA Bit Acronym 31 :04 Reserved E1P Offset Start: 3108h Offset End: 310Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved EHCI #1 -- Indicates which interrupt the EHCI controller #1 drives as its interrupt. 0h = No interrupt 1h = INTA (Default) 2h = INTB 3h = INTC 4h = INTD 5h-7h = Reserved Note: EHCI Controller #1 is mapped to Device 29 Function 0 when RMH is enabled Intel(R) Communications Chipset 89xx Series - Datasheet 384 B0:D31 :F0 Default: 10004321h Bit Range 03 :00 Bus:Device:Function: RW October 2012 Order Number: 327879-001US 7.0 7.1.1.32 Offset 310Ch: D28IP--Device 28 Interrupt Pin Register (B0:D31:F0) Table 7-33. Offset 310Ch: D28IP--Device 28 Interrupt Pin Register Description: View: PCI Size: 32 bit BAR: RCBA Bit Acronym 31 :16 Reserved 11 :08 07 :04 03 :00 B0:D31 :F0 Default: 00214321h Bit Range 15 :12 Bus:Device:Function: Offset Start: 310Ch Offset End: 310Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved P4IP PCI Express* #4 Pin -- Indicates which pin PCI Express* port #4 drives as its interrupt. 0h = No interrupt 1h = INTA 2h = INTB 3h = INTC 4h = INTD (Default) 5h-7h = Reserved RW P3IP PCI Express* #3 Pin -- Indicates which pin PCI Express* port #3 drives as its interrupts. 0h = No interrupt 1h = INTA 2h = INTB 3h = INTC (Default) 4h = INTD 5h-7h = Reserved RW P2IP PCI Express* #2 Pin -- Indicates which pin PCI Express* port #2 drives as its interrupts. 0h = No interrupt 1h = INTA 2h = INTB (Default) 3h = INTC 4h = INTD 5h-7h = Reserved RW P1IP PCI Express* #1 Pin -- Indicates which pin PCI Express* port #1 drives as its interrupt. 0h = No interrupt 1h = INTA (Default) 2h = INTB 3h = INTC 4h = INTD 5h-7h = Reserved RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 385 7.1.1.33 Offset 3124h: D22IP--Device 22 Interrupt Pin Register (B0:D31:F0) Table 7-34. Offset 3124h: D22IP--Device 22 Interrupt Pin Register Description: View: PCI Size: 32 bit BAR: RCBA Bit Acronym 31 :16 Reserved 11 :08 07 :04 03 :00 Bit Description Offset Start: 3124h Offset End: 3127h Power Well: Sticky Bit Reset Value Bit Access Reserved KTIP KT Pin -- Indicates which pin the Keyboard text PCI functionality drives as its interrupt. 0h = No Interrupt 1h = INTA 2h = INTB 3h = INTC 4h = INTD 5h-Fh = Reserved RW IDERIP IDER Pin -- Indicates which pin the IDE Redirect PCI functionality drives as its interrupt. 0h = No Interrupt 1h = INTA 2h = INTB 3h = INTC 4h = INTD 5h-Fh = Reserved RW MEI2IP Intel(R) MEI #2 Pin -- Indicates which pin the Management Engine Interface Controller #2 drives as its interrupt. 0h = No Interrupt 1h = INTA 2h = INTB 3h = INTC 4h = INTD 5h-Fh = Reserved RW MEI1IP Intel(R) MEI #1 Pin -- Indicates which pin the Management Engine Interface Controller #1 drives as its interrupt 0h = No Interrupt 1h = INTA 2h = INTB 3h = INTC 4h = INTD 5h-Fh = Reserved RW Intel(R) Communications Chipset 89xx Series - Datasheet 386 B0:D31 :F0 Default: See register description Bit Range 15 :12 Bus:Device:Function: October 2012 Order Number: 327879-001US 7.0 7.1.1.34 Offset 3140h: D31IR--Device 31 Interrupt Route Register (B0:D31:F0) Table 7-35. Offset 3140h: D31IR--Device 31 Interrupt Route Register Description: View: PCI Size: 16 bit Bit Range 15 14 :12 11 10 :08 07 06 :04 03 02 :00 BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 3210h Bit Acronym Reserved IDR Reserved ICR Reserved IBR Reserved IAR October 2012 Order Number: 327879-001US Offset Start: 3140h Offset End: 3141h Power Well: Bit Description Sticky Reserved Bit Reset Value Bit Access O Interrupt D Pin Route -- Indicates which physical pin on the PCH is connected to the INTD# pin reported for device 31 functions. 0h = PIRQA# 1h = PIRQB# 2h = PIRQC# 3h = PIRQD# (Default) 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Reserved RW O Interrupt C Pin Route -- Indicates which physical pin on the PCH is connected to the INTC# pin reported for device 31 functions. 0h = PIRQA# 1h = PIRQB# 2h = PIRQC# (Default) 3h = PIRQD# 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Reserved RW O Interrupt B Pin Route -- Indicates which physical pin on the PCH is connected to the INTB# pin reported for device 31 functions. 0h = PIRQA# 1h = PIRQB# (Default) 2h = PIRQC# 3h = PIRQD# 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Reserved RW O Interrupt A Pin Route -- Indicates which physical pin on the PCH is connected to the INTA# pin reported for device 31 functions. 0h = PIRQA# (Default) 1h = PIRQB# 2h = PIRQC# 3h = PIRQD# 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# RW Intel(R) Communications Chipset 89xx Series - Datasheet 387 7.1.1.35 Offset 3144h: D29IR--Device 29 Interrupt Route Register (B0:D31:F0) Table 7-36. Offset 3144h: D29IR--Device 29 Interrupt Route Register Description: View: PCI Size: 16 bit Bit Range 15 14 :12 11 10 :08 07 06 :04 03 02 :00 BAR: RCBA Bus:Device:Function: Default: 3210h Bit Acronym Reserved IDR Reserved ICR Reserved IBR Reserved IAR Offset Start: 3144h Offset End: 3145h Power Well: Bit Description Reserved Interrupt D Pin Route -- Indicates which physical pin on the PCH is connected to the INTD# pin reported for device 31 functions. 0h = PIRQA# 1h = PIRQB# 2h = PIRQC# 3h = PIRQD# (Default) 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Reserved Interrupt C Pin Route -- Indicates which physical pin on the PCH is connected to the INTC# pin reported for device 31 functions. 0h = PIRQA# 1h = PIRQB# 2h = PIRQC# (Default) 3h = PIRQD# 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Reserved Interrupt B Pin Route -- Indicates which physical pin on the PCH is connected to the INTB# pin reported for device 31 functions. 0h = PIRQA# 1h = PIRQB# (Default) 2h = PIRQC# 3h = PIRQD# 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Reserved Interrupt A Pin Route -- Indicates which physical pin on the PCH is connected to the INTA# pin reported for device 31 functions. 0h = PIRQA# (Default) 1h = PIRQB# 2h = PIRQC# 3h = PIRQD# 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Intel(R) Communications Chipset 89xx Series - Datasheet 388 B0:D31 :F0 Sticky Bit Reset Value Bit Access O RW O RW O RW O RW October 2012 Order Number: 327879-001US 7.0 7.1.1.36 Offset 3146h: D28IR--Device 28 Interrupt Route Register (B0:D31:F0) Table 7-37. Offset 3146h: D28IR--Device 28 Interrupt Route Register Description: View: PCI Size: 16 bit Bit Range 15 14 :12 11 10 :08 07 06 :04 03 02 :00 BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 3210h Bit Acronym Reserved IDR Reserved ICR Reserved IBR Reserved IAR October 2012 Order Number: 327879-001US Offset Start: 3146h Offset End: 3147h Power Well: Bit Description Sticky Reserved Bit Reset Value Bit Access O Interrupt D Pin Route -- Indicates which physical pin on the PCH is connected to the INTD# pin reported for device 31 functions. 0h = PIRQA# 1h = PIRQB# 2h = PIRQC# 3h = PIRQD# (Default) 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Reserved RW O Interrupt C Pin Route -- Indicates which physical pin on the PCH is connected to the INTC# pin reported for device 31 functions. 0h = PIRQA# 1h = PIRQB# 2h = PIRQC# (Default) 3h = PIRQD# 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Reserved RW O Interrupt B Pin Route -- Indicates which physical pin on the PCH is connected to the INTB# pin reported for device 31 functions. 0h = PIRQA# 1h = PIRQB# (Default) 2h = PIRQC# 3h = PIRQD# 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Reserved RW O Interrupt A Pin Route -- Indicates which physical pin on the PCH is connected to the INTA# pin reported for device 31 functions. 0h = PIRQA# (Default) 1h = PIRQB# 2h = PIRQC# 3h = PIRQD# 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# RW Intel(R) Communications Chipset 89xx Series - Datasheet 389 7.1.1.37 Offset 315Ch: D22IR--Device 22 Interrupt Route Register (B0:D31:F0) Table 7-38. Offset 315Ch: D22IR--Device 22 Interrupt Route Register Description: View: PCI Size: 16 bit Bit Range 15 14 :12 11 10 :08 07 06 :04 03 02 :00 BAR: RCBA Bus:Device:Function: Default: 3210h Bit Acronym Reserved IDR Reserved ICR Reserved IBR Reserved IAR Offset Start: 315Ch Offset End: 315Dh Power Well: Bit Description Reserved Interrupt D Pin Route -- Indicates which physical pin on the PCH is connected to the INTD# pin reported for device 31 functions. 0h = PIRQA# 1h = PIRQB# 2h = PIRQC# 3h = PIRQD# (Default) 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Reserved Interrupt C Pin Route -- Indicates which physical pin on the PCH is connected to the INTC# pin reported for device 31 functions. 0h = PIRQA# 1h = PIRQB# 2h = PIRQC# (Default) 3h = PIRQD# 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Reserved Interrupt B Pin Route -- Indicates which physical pin on the PCH is connected to the INTB# pin reported for device 31 functions. 0h = PIRQA# 1h = PIRQB# (Default) 2h = PIRQC# 3h = PIRQD# 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Reserved Interrupt A Pin Route -- Indicates which physical pin on the PCH is connected to the INTA# pin reported for device 31 functions. 0h = PIRQA# (Default) 1h = PIRQB# 2h = PIRQC# 3h = PIRQD# 4h = PIRQE# 5h = PIRQF# 6h = PIRQG# 7h = PIRQH# Intel(R) Communications Chipset 89xx Series - Datasheet 390 B0:D31 :F0 Sticky Bit Reset Value Bit Access O RW O RW O RW O RW October 2012 Order Number: 327879-001US 7.0 7.1.1.38 Offset 31FEh: OIC--Other Interrupt Control Register (B0:D31:F0) FEC10000h - FEC3FFFFh is allocated to PCIe* when I/OxApic Enable (PAE) bit is set. Table 7-39. Offset 31FEh: OIC--Other Interrupt Control Register Description: View: PCI Size: 16 bit BAR: RCBA Bit Acronym 15 :12 Reserved 11 TPMINTPOL 10 Reserved 08 07 :00 B0:D31 :F0 Default: 0000h Bit Range 09 Bus:Device:Function: Offset Start: 31FEh Offset End: 31FFh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Intel(R) TPM Interrupt Polarity Enable: When this bit is cleared to 0 the Intel TPM IRQ signal will be forced to operate in ActiveHigh mode regardless of the setting of the IOAPIC RTE for the corresponding interrupt. When this bit is set to 1 the Intel TPM IRQ signal's polarity will be set to match the corresponding IOAPIC RTE polarity. If the IOAPIC is programmed for ActiveLow mode the corresponding Intel TPM IRQ will operate in ActiveLow mode. If the IOAPIC RTE is programmed for ActiveHigh mode the corresponding Intel TPM IRQ will operate in ActiveHigh mode. Reserved CEN Coprocessor Error Enable: 0 = FERR# will not generate IRQ13 nor IGNNE#. I1 = f FERR# is low, the PCH generates IRQ13 internally and holds it until an I/O port F0h write. It will also drive IGNNE# active. RW AEN APIC Enable: 0 = The internal IOxAPIC is disabled. 1 = Enables the internal IOxAPIC and its address decode. Note: SW should read this register after modifying APIC enable bit prior to access to the IOxAPIC address range. RW ASEL APIC Range Select -- These bits define address bits 19:12 for the IOxAPIC range. The default value of 00h enables compatibility with prior PCH products as an initial value. This value must not be changed unless the IOxAPIC Enable bit is cleared. RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 391 7.1.1.39 Offset 3310h: PRSTS--Power and Reset Status (B0:D31:F0) Table 7-40. Offset 3310h: PRSTS--Power and Reset Status Description: View: PCI Size: 32 bit BAR: RCBA Default: 02020000h Bit Range Bit Acronym 31 :16 Reserved Reserved Bit Description Sticky Bit Reset Value Bit Access Reserved RWC Reserved VE Watchdog Timer Status -- This bit is set when the VE watchdog timer causes a global reset. 07 Offset Start: 3310h Offset End: 3313h Power Well: Power Management Watchdog Timer -- This bit is set when the Power Management watchdog timer causes a global reset. 15 14 :08 Bus:Device:Function B0:D31 : :F0 RWC Intel(R) Management Engine Watchdog Timer Status -- This bit is set when the Intel(R) Management Engine watchdog timer causes a global reset. 06 05 :03 Reserved Reserved 02 ME_HRST_WARM_STS ME Host Reset Warm Status -- This bit is set when the Intel(R) Management Engine generates a Host reset without power cycling. Software clears this bit by writing a 1 to this bit position. RWC 01 ME_HRST_COLD_STS ME Host Reset Cold Status -- This bit is set when the Intel(R) Management Engine generates a Host reset with power cycling. Software clears this bit by writing a 1 to this bit position. RWC ME_WAKE_STS ME WAKE STATUS -- This bit is set when the Intel(R) Management Engine generates a NonMaskable wake event, and is not affected by any other enable bit. When this bit is set, the Host Power Management logic wakes to S0. RWC 00 Intel(R) Communications Chipset 89xx Series - Datasheet 392 RWC October 2012 Order Number: 327879-001US 7.0 7.1.1.40 Offset 3314h: CIR7--Chipset Initialization Register 7 (B0:D31:F0) Table 7-41. Offset 3314h: CIR7--Chipset Initialization Register 7 Description: View: PCI Size: 32 bit BAR: RCBA B0:D31 :F0 Default: 00000000h Bit Range Bit Acronym 31 :04 Reserved 03 :00 7.1.1.41 Bus:Device:Function: Offset Start: 3314h Offset End: 3317h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved CIR7 Field 1-- BIOS must program this field to 1111b. RW Offset 3324h: CIR8--Chipset Initialization Register 8 (B0:D31:F0) Table 7-42. Offset 3324h: CIR8--Chipset Initialization Register 8 Description: View: PCI Size: 32 bit Bit Range BAR: RCBA Default: 00000000h Bit Acronym Offset Start: 3324h Offset End: 3327h Power Well: Bit Description Sticky Bit Reset Value CIR8 Field 1 -- BIOS must program this field to 04000000h. 31 :00 7.1.1.42 Bus:Device:Function: B0:D31 :F0 Bit Access RW Offset 3330h: CIR9--Chipset Initialization Register 9 (B0:D31:F0) Table 7-43. Offset 3330h: CIR9--Chipset Initialization Register 9 Description: View: PCI Size: 32 bit Bit Range BAR: RCBA B0:D31 Bus:Device:Function: :F0 Default: 00000000h Bit Acronym 31 :00 October 2012 Order Number: 327879-001US Offset Start: 3330h Offset End: 3333h Power Well: Bit Description Sticky CIR9 Field 1 -- BIOS must program this field to 00000000h. Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 393 7.1.1.43 Offset 3340h: CIR10--Chipset Initialization Register 10 (B0:D31:F0) Table 7-44. Offset 3340h: CIR10--Chipset Initialization Register 10 Description: View: PCI Size: 32 bit Bit Range BAR: RCBA B0:D31 :F0 Default: 00000000h Bit Acronym 31 :00 7.1.1.44 Bus:Device:Function: Offset Start: 3340h Offset End: 3343h Power Well: Bit Description Sticky Bit Reset Value CIR10 Field 1 -- BIOS must program this field. Bit Access RW Offset 3350h: CIR13--Chipset Initialization Register 13 (B0:D31:F0) Table 7-45. Offset 3350h: CIR13--Chipset Initialization Register 13 Description: View: PCI Size: 32 bit Bit Range Default: 00000000h Bit Acronym 31 :00 7.1.1.45 B0:D31 Bus:Device:Function: :F0 BAR: RCBA Offset Start: 3350h Offset End: 3353h Power Well: Bit Description Sticky Bit Reset Value CIR13 Field 1 -- BIOS must program this field. Bit Access RW Offset 3368h: CIR14--Chipset Initialization Register 14 (B0:D31:F0) Table 7-46. Offset 3368h: CIR14--Chipset Initialization Register 14 Description: View: PCI Size: 32 bit Bit Range 31 :00 Bus:Device:Function: B0:D31 :F0 BAR: RCBA Default: 00000000h Bit Acronym Power Well: Bit Description CIR14 Field 1 -- BIOS must program this field. Intel(R) Communications Chipset 89xx Series - Datasheet 394 Offset Start: 3368h Offset End: 336Bh Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 7.0 7.1.1.46 Offset 3378h: CIR15--Chipset Initialization Register 15 (B0:D31:F0) Table 7-47. Offset 3378h: CIR15--Chipset Initialization Register 15 Description: View: PCI Size: 32 bit Bit Range BAR: RCBA B0:D31 :F0 Default: 00000000h Bit Acronym 31 :00 7.1.1.47 Bus:Device:Function: Offset Start: 3378h Offset End: 337Bh Power Well: Bit Description Sticky Bit Reset Value CIR15 Field 1 -- BIOS must program this field. Bit Access RW Offset 3388h: CIR16--Chipset Initialization Register 16 (B0:D31:F0) Table 7-48. Offset 3388h: CIR16--Chipset Initialization Register 16 Description: View: PCI Size: 32 bit Bit Range BAR: RCBA Default: 00000000h Bit Acronym 31 :00 7.1.1.48 B0:D31 Bus:Device:Function: :F0 Offset Start: 3388h Offset End: 338Bh Power Well: Bit Description Sticky Bit Reset Value CIR16 Field 1 -- BIOS must program this field. Bit Access RW Offset 33A0h: CIR17--Chipset Initialization Register 17 (B0:D31:F0) Table 7-49. Offset 33A0h: CIR17--Chipset Initialization Register 17 Description: View: PCI Size: 32 bit Bit Range BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym 31 :00 October 2012 Order Number: 327879-001US Offset Start: 33A0h Offset End: 33A3h Power Well: Bit Description Sticky CIR17 Field 1 -- BIOS must program this field. Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 395 7.1.1.49 Offset 33A8h: CIR18--Chipset Initialization Register 18 (B0:D31:F0) Table 7-50. Offset 33A8h: CIR18--Chipset Initialization Register 18 Description: View: PCI Size: 32 bit Bit Range BAR: RCBA B0:D31 :F0 Default: 00000000h Bit Acronym 31 :00 7.1.1.50 Bus:Device:Function: Offset Start: 33A8h Offset End: 33ABh Power Well: Bit Description Sticky Bit Reset Value CIR18 Field 1 -- BIOS must program this field. Bit Access RW Offset 33C0h: CIR19--Chipset Initialization Register 19 (B0:D31:F0) Table 7-51. Offset 33C0h: CIR19--Chipset Initialization Register 19 Description: View: PCI Size: 32 bit Bit Range Default: 00000000h Bit Acronym 31 :00 7.1.1.51 B0:D31 Bus:Device:Function: :F0 BAR: RCBA Offset Start: 33C0h Offset End: 33C3h Power Well: Bit Description Sticky Bit Reset Value CIR19 Field 1 -- BIOS must program this field. Bit Access RW Offset 33CCh: CIR20--Chipset Initialization Register 20 (B0:D31:F0) Table 7-52. Offset 33CCh: CIR20--Chipset Initialization Register 20 Description: View: PCI Size: 32 bit Bit Range 31 :00 Bus:Device:Function: B0:D31 :F0 BAR: RCBA Default: 00000000h Bit Acronym Power Well: Bit Description CIR20 Field 1 -- BIOS must program this field. Intel(R) Communications Chipset 89xx Series - Datasheet 396 Offset Start: 33CCh Offset End: 33CFh Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 7.0 7.1.1.52 Offset 33D0h: CIR21--Chipset Initialization Register 21 (B0:D31:F0) Table 7-53. Offset 33D0h: CIR21--Chipset Initialization Register 21 Description: View: PCI Size: 32 bit Bit Range BAR: RCBA B0:D31 :F0 Default: 00000000h Bit Acronym 31 :00 7.1.1.53 Bus:Device:Function: Offset Start: 33D0h Offset End: 33D3h Power Well: Bit Description Sticky Bit Reset Value CIR21 Field 1 -- BIOS must program this field. Bit Access RW Offset 33D4h: CIR22--Chipset Initialization Register 22 (B0:D31:F0) Table 7-54. Offset 33D4h: CIR22--Chipset Initialization Register 22 Description: View: PCI Size: 32 bit Bit Range BAR: RCBA B0:D31 Bus:Device:Function: :F0 Default: 00000000h Bit Acronym 31 :00 October 2012 Order Number: 327879-001US Offset Start: 33D4h Offset End: 33D7h Power Well: Bit Description Sticky CIR22 Field 1 -- BIOS must program this field. Program this register after all registers in the 3330-33D3 range and B0:D31:F0:A9h are already programmed Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 397 7.1.1.54 Offset 3400h: RC--RTC Configuration Register (B0:D31:F0) Table 7-55. Offset 3400h: RC--RTC Configuration Register Description: View: PCI Size: 32 bit BAR: RCBA Bus:Device:Function: Default: 00000000h Bit Range Bit Acronym 31 :05 Reserved Offset Start: 3400h Offset End: 3403h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved UL Upper 128 Byte Lock: 0 = Bytes not locked. 1 = Bytes 38h-3Fh in the upper 128-byte bank of RTC RAM are locked and cannot be accessed. Writes will be dropped and reads will not return any ensured data. Bit reset on system reset. RWLO 03 LL Lower 128 Byte Lock: 0 = Bytes not locked. 1 = Bytes 38h-3Fh in the lower 128-byte bank of RTC RAM are locked and cannot be accessed. Writes will be dropped and reads will not return any ensured data. Bit reset on system reset. RWLO 02 UE Upper 128 Byte Enable: 0 = Bytes locked. 1 = The upper 128-byte bank of RTC RAM can be accessed. 04 01 :00 Reserved RW Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 398 B0:D31 :F0 October 2012 Order Number: 327879-001US 7.0 7.1.1.55 Offset 3404h: HPTC--High Precision Timer Configuration Register (B0:D31:F0) Table 7-56. Offset 3404h: HPTC--High Precision Timer Configuration Register Description: View: PCI Size: 32 bit BAR: RCBA Bit Acronym 31 :08 Reserved 06 :02 01 :00 B0:D31 :F0 Default: 00000000h Bit Range 07 Bus:Device:Function: AE Reserved AS October 2012 Order Number: 327879-001US Offset Start: 3404h Offset End: 3407h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Address Enable: 0 = Address disabled. 1 = The PCH will decode the High Precision Timer memory address range selected by bits 1:0 below. RW Reserved Address Select -- This 2-bit field selects 1 of 4 possible memory address ranges for the High Precision Timer functionality. The encodings are: 00 = FED0_0000h - FED0_03FFh 01 = FED0_1000h - FED0_13FFh 10 = FED0_2000h - FED0_23FFh 11 = FED0_3000h - FED0_33FFh RW Intel(R) Communications Chipset 89xx Series - Datasheet 399 7.1.1.56 Offset 3410h: GCS--General Control and Status Register (B0:D31:F0) Table 7-57. Offset 3410h: GCS--General Control and Status Register (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR: RCBA Bus:Device:Function: Default: 00000yy0h (yy = xx0000x0b) Bit Range Bit Acronym 31 :13 Reserved Reserved FLRCSSEL Function Level Reset Capability Structure Select: 0 = Function Level Reset (FLR) will utilize the standard capability structure with unique capability ID assigned by PCISIG. 1 = Vendor Specific Capability Structure is selected for FLR. 12 B0:D31 :F0 Bit Description Offset Start: 3410h Offset End: 3413h Power Well: Sticky Bit Reset Value Bit Access RW Boot BIOS Straps -- This field determines the destination of accesses to the BIOS memory range. The default values for these bits represent the strap values of BBS1# / GPIO51 (bit 11) at the rising edge of PWROK and BBS0# (bit 10) at the rising edge of MEPWROK. 11 :10 BBS Bits 11:10 Description 00b RESERVED 01b RESERVED 10b RESERVED 11b SPI RW When SPI is selected, the range that is decoded is further qualified by other configuration bits described in the respective sections. The value in this field can be overwritten by software as long as the BIOS Interface Lock-Down (bit 0) is not set. Note: Booting to PCI is intended for debug/testing only. Boot BIOS Destination Select to LPC/PCI by functional strap or via Boot BIOS Destination Bit will not affect SPI accesses initiated by Intel(R) Management Engine or Integrated GbE LAN. 09 08 :07 06 SERM Reserved FME Server Error Reporting Mode: 0 = The PCH is the final target of all errors. The processor sends a messages to the PCH for the purpose of generating NMI. 1 = The processor is the final target of all errors from PCI Express* and DMI. In this mode, if the PCH detects a fatal, non-fatal, or correctable error on DMI or its downstream ports, it sends a message to the Processor. If the PCH receives an ERR_* message from the downstream port, it sends that message to the Processor. Reserved FERR# MUX Enable -- This bit enables FERR# to be a processor break event indication. 0 = Disabled. 1 = The PCH examines FERR# during a C2, C3, or C4 state as a break event. Intel(R) Communications Chipset 89xx Series - Datasheet 400 RW RW October 2012 Order Number: 327879-001US 7.0 Table 7-57. Offset 3410h: GCS--General Control and Status Register (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 00000yy0h (yy = xx0000x0b) Offset Start: 3410h Offset End: 3413h Power Well: Bit Description NR No Reboot -- This bit is set when the "No Reboot" strap (NRBOOTS pin on the PCH) is sampled high on PWROK. This bit may be set or cleared by software if the strap is sampled low but may not override the strap when it indicates "No Reboot". 0 = System will reboot upon the second timeout of the TCO timer. 1 = The TCO timer will count down and generate the SMI# on the first timeout, but will not reboot on the second timeout. RW 04 AME Alternate Access Mode Enable: 0 = Disabled. 1 = Alternate access read only registers can be written, and write only registers can be read. Before entering a low power state, several registers from powered down parts may need to be saved. In the majority of cases, this is not an issue, as registers have read and write paths. However, several of the ISA compatible registers are either read only or write only. To get data out of write-only registers, and to restore data into read-only registers, the PCH implements an alternate access mode. RW 03 SPS Shutdown Policy Select -- When cleared (default), the PCH will drive INIT# in response to the shutdown Vendor Defined Message (VDM). When set to 1, the PCH will treat the shutdown VDM similar to receiving a CF9h I/O write with data value06h, and will drive PLTRST# active. RW 05 Sticky Bit Reset Value Bit Acronym Bit Access Reserved Page Route -- Determines where to send the reserved page registers. These addresses are sent to PCI or LPC for the purpose of generating POST codes. The I/O addresses modified by this field are: 80h, 84h, 85h, 86h, 88h, 8Ch, 8Dh, and 8Eh. 02 RPR 0 = Writes will be forwarded to LPC, shadowed within the PCH, and reads will be returned from the internal shadow 1 = Reserved RW The aliases for these registers, at 90h, 94h, 95h, 96h, 98h, 9Ch, 9Dh, and 9Eh, are always decoded to LPC. 01 00 Reserved BILD October 2012 Order Number: 327879-001US Reserved BIOS Interface Lock-Down: 0 = Disabled. 1 = Prevents BUC.TS (offset 3414, bit 0) and GCS.BBS (offset 3410h, bits 11:10) from being changed. This bit can only be written from 0 to 1 once. RWLO Intel(R) Communications Chipset 89xx Series - Datasheet 401 7.1.1.57 Offset 3414h: BUC--Backed Up Control Register (B0:D31:F0) All bits in this register are in the RTC well and only cleared by RTCRST#. Table 7-58. Offset 3414h: BUC--Backed Up Control Register Description: View: PCI Size: 8 bit BAR: RCBA Bit Acronym 07 :06 Reserved 04 03 :01 00 B0:D31 :F0 Default: 0000000xb Bit Range 05 Bus:Device:Function: Reserved SDO Reserved TS Offset Start: 3414h Offset End: 3414h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved RW Daylight Savings Override: 0 = Daylight Savings is Enabled. 1 = The DSE bit in RTC Register B is set to Read-only with a value of 0 to disable daylight savings. RW Reserved Top Swap: * When set, PCH will invert either A16, A17, or A18 for cycles going to the BIOS space (but not the feature space) in the FWH. * When cleared, PCH will not invert A16. If booting from LPC (FWH), then the Boot Block size is 64KB and A16 is inverted if Top Swap is enabled. RW If booting from SPI, the BOOT_BLOCK_SIZE soft strap determines if A16, A17, or A18 should be inverted if Top Swap is enabled. If PCH is strapped for Top-Swap (GNT[3]# is low at rising edge of PWROK), this bit cannot be cleared by software. The strap jumper should be removed and the system rebooted. Intel(R) Communications Chipset 89xx Series - Datasheet 402 October 2012 Order Number: 327879-001US 7.0 7.1.1.58 Offset 3418h: FD--Function Disable Register (B0:D31:F0) When disabling a function, only the configuration space is disabled. Software must ensure that all functionality within a controller that is not desired (such as memory spaces, I/O spaces, and DMA engines) is disabled prior to disabling the function. When a function is disabled, software must not attempt to re-enable it. A disabled function can only be re-enabled by a platform reset. Table 7-59. Offset 3418h: FD--Function Disable Register (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: See register description Offset Start: 3418h Offset End: 341Bh Power Well: Bit Reset Value Bit Access Serial ATA Disable 2 -- Default is 0: 0 = The SATA controller #2 (D31:F5) is enabled. 1 = The SATA controller #2 (D31:F5) is disabled. 0 RW Thermal Throttle Disable -- Default is 0: 0 = Thermal Throttle is enabled. 1 = Thermal Throttle is disabled. 0 RW PE4D PCI Express* 4 Disable -- Default is 0. When disabled, the link for this port is put into the "link down" state: 0 = PCI Express* port #4 is enabled. 1 = PCI Express* port #4 is disabled. When disabled, the link for this port is put into the "link down" state. 0 RW PE3D PCI Express* 3 Disable -- Default is 0. When disabled, the link for this port is put into the link down state: 0 = PCI Express* port #3 is enabled. 1 = PCI Express* port #3 is disabled. When disabled, the link for this port is put into the "link down" state.. 0 RW PE2D PCI Express* 2 Disable -- Default is 0. When disabled, the link for this port is put into the "link down" state: 0 = PCI Express* port #2 is enabled. 1 = PCI Express* port #2 is disabled. When disabled, the link for this port is put into the "link down" state. 0 RW 16 PE1D PCI Express* 1 Disable -- Default is 0. When disabled, the link for this port is put into the link down state: 0 = PCI Express* port #1 is enabled. 1 = PCI Express* port #1 is disabled. When disabled, the link for this port is put into the "link down" state. 0 RW 15 U2D USB2 Disable: When set, the USB2 host controller is disabled. 0 RW Bit Range Bit Acronym 31 :26 Reserved 25 SAD2 24 TTD 23 :20 19 18 17 Reserved October 2012 Order Number: 327879-001US Bit Description Sticky Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 403 Table 7-59. Offset 3418h: FD--Function Disable Register (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: See register description Offset Start: 3418h Offset End: 341Bh Power Well: Bit Reset Value Bit Access 0 RW SM Bus Disable: When set, the SM Bus controller is disabled. This only disables the configuration space. 0 RW Serial ATA Disable 1: When set, the serial SATA controller 1 (SATA1, Device 31, Function 2) is disabled. 0 RW Bit Acronym Bit Description Sticky LPC Bridge Disable: When set, the LPC bridge is disabled. Unlike the other disables in this register, the following additional spaces will no longer be decoded by the LPC bridge: 14 LPD * Memory cycles below 16MB (1000000h) * I/O cycles below 64kB (10000h) Memory cycles in the LPC BIOS range below 4GB will still be decodedwhen this bit is set, but the aliases at the top of 1MB (the E and F segment) no longer will be decoded. The decode for the internal I/OxAPIC memory ranges and theEOI message are qualified by the APIC Enable (AEN) bit, but not the LPC Bridge Disable (LPD) bit. 13 :04 Reserved 03 SD 02 SAD1 01 :00 Reserved Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 404 October 2012 Order Number: 327879-001US 7.0 7.1.1.59 Offset 341Ch: CG--Clock Gating (B0:D31:F0) Table 7-60. Offset 341Ch: CG--Clock Gating Description: View: PCI Size: 32 bit BAR: RCBA B0:D31 :F0 Default: 00000000h Bit Range Bit Acronym :31 LPC 30 :20 Bus:Device:Function: Offset Start: 341Ch Offset End: 341Fh Power Well: Bit Description Sticky Legacy Dynamic Clock Gate Enable: 0 = Legacy Dynamic Clock Gating is Disabled 1 = Legacy Dynamic Clock Gating is Enabled Bit Reset Value Bit Access RW Reserved Reserved 19 USB EHCI Static Clock Gate Enable: 0 = USB EHCI Static Clock Gating is Disabled 1 = USB EHCI Static Clock Gating is Enabled RW 18 USB EHCI Dynamic Clock Gate Enable: 0 = USB EHCI Dynamic Clock Gating is Disabled 1 = USB EHCI Dynamic Clock Gating is Enabled RW 17 SATA Port 5 Dynamic Clock Gate Enable: 0 = SATA Port 5 Dynamic Clock Gating is Disabled 1 = SATA Port 5 Dynamic Clock Gating is Enabled RW 16 SATA Port 4 Dynamic Clock Gate Enable: 0 = SATA Port 4 Dynamic Clock Gating is Disabled 1 = SATA Port 4 Dynamic Clock Gating is Enabled RW 15 :06 05 04 :01 Reserved Reserved SMBus Clock Gating EnablE: SMBCGEN 0 = SMBus Clock Gating is Disabled. 1 = SMBus Clock Gating is Enabled. RW Reserved Reserved 00 October 2012 Order Number: 327879-001US PCI Express* Root Port Static Clock Gate Enable: 0 = PCI Express* root port Static Clock Gating is Disabled 1 = PCI Express* root port Static Clock Gating is Enabled RW Intel(R) Communications Chipset 89xx Series - Datasheet 405 7.1.1.60 Offset 3420h: FDSW--Function Disable SUS Well (B0:D31:F0) Table 7-61. Offset 3420h: FDSW--Function Disable SUS Well Description: View: PCI Size: 8 bit Bit Range 07 06 :00 7.1.1.61 BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 00h Bit Acronym FDSWL Reserved Offset Start: 3420h Offset End: 3420h Power Well: Bit Description Sticky Bit Reset Value Function Disable SUS Well Lockdown: 0 = FDSW registers are not locked down 1 = FDSW registers are locked down Bit Access RW Reserved Offset 3428h: FD2--Function Disable 2 (B0:D31:F0) Table 7-62. Offset 3428h: FD2--Function Disable 2 Description: View: PCI Size: 32 bit Bus:Device:Function: B0:D31 :F0 BAR: RCBA Default: 00000000h Bit Range Bit Acronym 31 :05 Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 04 KTD KT Disable -- Default is 0: 0 = Keyboard Text controller (D22:F3) is enabled. 1 = Keyboard Text controller (D22:F3) is Disabled RW 03 IRERD IDE-R Disable -- Default is 0: 0 = IDE Redirect controller (D22:F2) is Enabled. 1 = IDE Redirect controller (D22:F2) is Disabled. RW 02 MEI2D Intel(R) MEI #2 Disable -- Default is 0: 0 = Intel(R) MEI controller #2 (D22:F1) is enabled. 1 = Intel(R) MEI controller #2 (D22:F1) is disabled. RW 01 MEI1D Intel(R) MEI #1 Disable -- Default is 0: 0 = Intel(R) MEI controller #1 (D22:F0) is enabled. 1 = Intel(R) MEI controller #1 (D22:F0) is disabled. RW 00 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 406 Offset Start: 3428h Offset End: 342Bh October 2012 Order Number: 327879-001US 7.0 7.1.1.62 Offset 3500h: USBIR[0:5]--USB Initialization Register [0-5] (B0:D31:F0) Address Port Port Port Port Port Port Offset: 0: 3500h 1: 3504h 2: 3508h 3: 350Ch 4: 3510h 5: 3514h These registers are located in the Suspend well. Table 7-63. Offset 3500h: USBIR[0:5]--USB Initialization Register [0-5] Description: View: PCI Size: 32 bit BAR: RCBA Bus:Device:Function: B0:D31:F0 Default: 20000B4Ah Bit Range Bit Acronym 31 :12 Reserved Offset Start: 3500h Offset End: 3514h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved USB Port[n] Initialization Field 1 -- Configures the USB transmitter of port n. This field must be programmed by BIOS. All other values are reserved. Note: See Platform Design Guide for description of topology definitions. 11 :00 C4Bh Long Topology Setting (10 -14 inches) - This setting should be used for desktop back panel ports, long mobile back panel ports and long mobile docking ports where increased transmit amplitude is required to optimize USB transmit characteristics. Note: This setting may also be used for short topology ports that require an increased transmit amplitude. This setting should not be used for short topology ports that have sufficient transmit amplitude when using the short topology setting. RW A4Bh Short Topology Setting (< 10 inches)- This setting should be used for most desktop front panel ports, short mobile back panel ports and short mobile docking ports for which decreased transmit amplitude is required to optimize USB transmit characteristics. Note: This setting is weaker than the long topology setting and may not be strong enough for use with all short topology ports. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 407 7.1.1.63 Offset 3564h: USBIRC--USB Initialization Register C (B0:D31:F0) Table 7-64. Offset 3564h: USBIRC--USB Initialization Register C Description: View: PCI Size: 32 bit BAR: RCBA Bit Acronym 31 :10 Reserved Offset Start: 3564h Offset End: 3567h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved USB Initialization Register C Field 1. BIOS must program this field to 1b. 08 RW USB Initialization Register C Field 2. BIOS must program this field to 10100b. 07 :03 7.1.1.64 B0:D31 :F0 Default: 0000371Bh Bit Range 02 :00 Bus:Device:Function: Reserved Reserved Offset 3570h: USBIRA--USB Initialization Register A (B0:D31:F0) Table 7-65. Offset 3570h: USBIRA--USB Initialization Register A Description: View: PCI Size: 32 bit Bit Range 31 :00 Bus:Device:Function: B0:D31 :F0 BAR: RCBA Default: 00000000h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access USB Initialization Register A. BIOS must program this field to 0062003h. Intel(R) Communications Chipset 89xx Series - Datasheet 408 Offset Start: 3570h Offset End: 3573h October 2012 Order Number: 327879-001US 7.0 7.1.1.65 Offset 357Ch: USBIRB--USB Initialization Register B (B0:D31:F0) Table 7-66. Offset 357Ch: USBIRB--USB Initialization Register B Description: View: PCI Size: 32 bit BAR: RCBA Bit Acronym 31 :19 Reserved Reserved Bit Description Sticky Reserved Bit Reset Value Bit Access Reserved RW Reserved USB Initialization Register B Field 3. BIOS must program this field to 1b 12 Offset Start: 357Ch Offset End: 357Fh Power Well: USB Initialization Register B Field 2. BIOS must program this field to 11b. 18 :17 11 B0:D31 :F0 Default: 0001C000h Bit Range 16 :13 Bus:Device:Function: RW Reserved USB Initialization Register B Field 1. BIOS must program this field to 1b RW 09 :08 USB Initialization Register B Field 4. BIOS must program this field to 01b. RW 07 :04 USB Initialization Register B Field 5. BIOS must program this field to 0100b. RW 10 03 :00 Reserved October 2012 Order Number: 327879-001US Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 409 7.1.1.66 Offset 3590h: MISCCTL--Miscellaneous Control Register (B0:D31:F0) This register is in the suspend well. This register is not reset on D3-to-D0, HCRESET nor core well reset. Table 7-67. Offset 3590h: MISCCTL--Miscellaneous Control Register Description: View: PCI Size: 32 bit BAR: RCBA Bus:Device:Function: Default: 00000000h Bit Range Bit Acronym 31 :02 Reserved Offset Start: 3590h Offset End: 3594h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 01 EHCI 2 USBR Enable -- When set, this bit enables support for the USB-r redirect device on the EHCI controller in Device 26. SW must complete programming the following registers before this bit is set: 1. UHCI function disables 2. HCCSPARAMS (N_CC, N_Ports) RW 00 EHCI 1 USBR Enable -- When set, this bit enables support for the USB-r redirect device on the EHCI controller in Device 29. SW must complete programming the following registers before this bit is set: 1. UHCI function disables 2. HCCSPARAMS (N_CC, N_Ports) RW Intel(R) Communications Chipset 89xx Series - Datasheet 410 B0:D31 :F0 October 2012 Order Number: 327879-001US 7.0 7.1.1.67 Offset 359Ch: PDO--USB Port Disable Override (B0:D31:F0) The register is implemented in the Suspend well to maintain the policy when the core power goes down. This register is not reset on D3-to-D0, HCRESET nor core well reset. Disabling Port 1 will block the Debug Port functionality for this EHCI. Table 7-68. Offset 359Ch: PDO--USB Port Disable Override Description: View: PCI Size: 32 bit BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 0000h Bit Range Bit Acronym 31 :14 Reserved 13 :00 Offset Start: 359Ch Offset End: 359Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved USB Port Disable -- When RMH is disabled: 0 = Prevents the corresponding USB port from reporting a Device Connection in the EHCI and UHCI Register Space. Attempts to enable the port through the PORTSC register will be ignored by the hardware when this bit is 1. 1 = Allows corresponding USB port to report a Device Connection in the EHCI and UHCI Register Space. When RMH is enabled: 0 = Prevents the corresponding USB port from reporting a Device Connection to the hub. Attempts to enable the port will be ignored by the hardware when this bit is 1. 1 = Allows corresponding USB port to report a Device Connection the hub. RW This register can not be written when the USB Per-Port Registers Write Enable bit (UPRWC register, PMBASE + 3C, bit 1) is 0. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 411 7.1.1.68 Offset 35A0h: USBOCM1--Overcurrent MAP Register 1 (B0:D31:F0) Table 7-69. Offset 35A0h: USBOCM1--Overcurrent MAP Register 1 Description: View: PCI Size: 32 bit Bit Range 31 :24 23 :16 15 :08 07 :00 BAR: RCBA Bus:Device:Function: Default: C0300C03h Offset Start: 35A0h Offset End: 35A3h Power Well: Bit Acronym Bit Description Sticky Bit Reset Value Bit Access OC3 Mapping -- Each bit position maps OC3# to a set of ports as follows: The OC3# pin is ganged to the overcurrent signal of each port that has its corresponding bit set. It is SW`s responsibility to ensure that a given port`s bit map is set only for one OC pin. Bit 31 30 29 28 27 26 25 24 Port X X 5 4 3 2 1 0 OC2 Mapping -- Each bit position maps OC2# to a set of ports as follows: The OC2# pin is ganged to the overcurrent signal of each port that has its corresponding bit set. It is SW`s responsibility to ensure that a given port`s bit map is set only for one OC pin. Bit 23 22 21 20 19 18 17 16 Port X X 5 4 3 2 1 0 OC1 Mapping -- Each bit position maps OC1# to a set of ports as follows: The OC1# pin is ganged to the overcurrent signal of each port that has its corresponding bit set. It is SW`s responsibility to ensure that a given port`s bit map is set only for one OC pin. Bit 15 14 13 12 11 10 9 8 Port X X 5 4 3 2 1 0 OC0 Mapping -- Each bit position maps OC0# to a set of ports as follows: The OC0# pin is ganged to the overcurrent signal of each port that has its corresponding bit set. It is SW`s responsibility to ensure that a given port`s bit map is set only for one OC pin. Bit 7 6 5 4 3 2 1 0 Port X X 5 4 3 2 1 0 Intel(R) Communications Chipset 89xx Series - Datasheet 412 B0:D31 :F0 October 2012 Order Number: 327879-001US 7.0 7.1.1.69 Offset 35B0h: RMHWKCTL--Rate Matching Hub Wake Control Register (B0:D31:F0) Table 7-70. Offset 35B0h: RMHWKCTL--Rate Matching Hub Wake Control Register (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Range Bit Acronym 31 :10 Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 09 RMH 2 Inherit EHCI2 Wake Control Settings -- When this bit is set, the RMH behaves as if bits 6:4 of this register reflect the appropriate bits of EHCI PORTSC0 bits 22:20. 08 RMH 1 Inherit EHCI1 Wake Control Settings -- When this bit is set, the RMH behaves as if bits 2:0 of this register reflect the appropriate bits of EHCI PORTSC0 bits 22:20. 07 RMH 2 Upstream Wake on Device Resume -- This bit governs the hub behavior when globally suspended and the system is in Sx: 0 = Enables the port to be sensitive to device initiated resume events as system wake-up events. i.e, the hub will initiate a resume on its upstream port and cause a wake from Sx when a device resume occurs on an enabled DS port 1 = Device resume event is seen on a downstream port, the hub does not initiate a wake upstream and does not cause a wake from Sx 06 RMH 2 Upstream Wake on OC Disable -- This bit governs the hub behavior when globally suspended and the system is in Sx: 0 = Enables the port to be sensitive to over-current conditions as system wake-up events. i.e, the hub will initiate a resume on its upstream port and cause a wake from Sx when an OC condition occurs on an enabled DS port 1 = Over-current event does not initiate a wake upstream and does not cause a wake from Sx 05 RMH 2 Upstream Wake on Disconnect Disable -- This bit governs the hub behavior when globally suspended and the system is in Sx: 0 = Enables disconnect events on downstream port to be treated as resume events to be propagated upstream. In this case, it is allowed to initiate a wake on its upstream port and cause a system wake from Sx in response to a disconnect event on a downstream port 1 = Downstream disconnect events do not initiate a resume on its upstream port or cause a resume from Sx. 04 RMH 2 Upstream Wake on Connect Enable -- This bit governs the hub behavior when globally suspended and the system is in Sx: 0 = Enables connect events on a downstream port to be treated as resume events to be propagated upstream. As well as waking up the system from Sx. 1 = Downstream connect events do not wake the system from Sx nor does it initiate a resume on its upstream port. October 2012 Order Number: 327879-001US Offset Start: 35B0h Offset End: 35B3h Intel(R) Communications Chipset 89xx Series - Datasheet 413 Table 7-70. Offset 35B0h: RMHWKCTL--Rate Matching Hub Wake Control Register (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: RCBA Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym Offset Start: 35B0h Offset End: 35B3h Power Well: Bit Description 03 RMH 1 Upstream Wake on Device Resume -- This bit governs the hub behavior when globally suspended and the system is in Sx: 0 = Enables the port to be sensitive to device initiated resume events as system wake-up events. i.e, the hub will initiate a resume on its upstream port and cause a wake from Sx when a device resume occurs on an enabled DS port 1 = Device resume event is seen on a downstream port, the hub does not initiate a wake upstream and does not cause a wake from Sx 02 RMH 1 Upstream Wake on OC Disable -- This bit governs the hub behavior when globally suspended and the system is in Sx: 0 = Enables the port to be sensitive to over-current conditions as system wake-up events. i.e, the hub will initiate a resume on its upstream port and cause a wake from Sx when an OC condition occurs on an enabled DS port 1 = Over-current event does not initiate a wake upstream and does not cause a wake from Sx 01 RMH 1 Upstream Wake on Disconnect Disable -- This bit governs the hub behavior when globally suspended and the system is in Sx: 0 = Enables disconnect events on downstream port to be treated as resume events to be propagated upstream. In this case, it is allowed to initiate a wake on its upstream port and cause a system wake from Sx in response to a disconnect event on a downstream port 1 = Downstream disconnect events do not initiate a resume on its upstream port or cause a resume from Sx. 00 RMH 1 Upstream Wake on Connect Enable -- This bit governs the hub behavior when globally suspended and the system is in Sx: 0 = Enables connect events on a downstream port to be treated as resume events to be propagated upstream. As well as waking up the system from Sx. 1 = Downstream connect events do not wake the system from Sx nor does it initiate a resume on its upstream port. Sticky Bit Reset Value Bit Access Intel(R) Communications Chipset 89xx Series - Datasheet 414 October 2012 Order Number: 327879-001US 8.0 8.0 SATA Controller Registers (B0:D31:F2) 8.1 PCI Configuration Registers (SATA-B0:D31:F2) Note: Address locations that are not shown should be treated as Reserved. All of the SATA registers are in the core well. None of the registers can be locked. Note: SATA Controller #1 (SATA - B0:D31:F2) supports two ports (Ports 4 and 5) in the AHCI mode. 8.1.1 SATA Controller PCI Register Address Map Table 8-1. SATA Controller PCI Register Address Map (Sheet 1 of 3) Offset Start Offset End Register ID - Description Default Value 00h 01h "Offset 00h: Vendor Identification Register (SATA-B0:D31:F2)" on page 418 8086h 02h 03h "Offset 02h: Device Identification Register (SATA-B0:D31:F2)" on page 418 2321h 04h 05h "Offset 04h: PCI Command Register (SATA-B0:D31:F2)" on page 418 0000h 06h 07h "Offset 06h: PCI Status Register (SATA-B0:D31:F2)" on page 420 02B0h 08h 08h "Offset 08h: Revision Identification Register (SATA-B0:D31:F2)" on page 421 See bit description 09h 09h "Offset 0Ah: When Sub Class Code Register (B0:D31:F2:Offset 0Ah) = 01h" on page 421 See bit description 0Ah 0Ah "Offset 0Ah: When Sub Class Code Register (B0:D31:F2:Offset 0Ah) = 06h" on page 422 01h 0Ah 0Ah "Offset 0Ah: Sub Class Code Register (SATA-B0:D31:F2)" on page 423 See bit description 0Bh 0Bh "Offset 0Bh: Base Class Code Register (B0:D31:F2)" on page 423 01h 0Dh 0Dh "Offset 0Dh: Primary Master Latency Timer Register (B0:D31:F2)" on page 423 00h 0Eh 0Eh "Offset 0Eh: Header Type (SATA-B0:D31:F2)" on page 424 80h 10h 13h "Offset 10h: Primary Command Block Base Address Register (SATA-B0:D31:F2)" on page 424 00000001h 14h 17h "Offset 14h: Primary Control Block Base Address Register (SATA-B0:D31:F2)" on page 425 00000001h 18h 1Bh "Offset 18h: Secondary Command Block Base Address Register (IDE D31:F1)" on page 425 00000001h 1Ch 1Fh "Offset 1Ch: Secondary Control Block Base Address Register (IDE B0:D31:F2)" on page 426 00000001h 20h 23h "Offset 20h: Legacy Bus Master Base Address Register (SATA-B0:D31:F2)" on page 426 00000001h 24h 27h "Offset 24h: When SCC is Not 01h (SATA-B0:D31:F2)" on page 427 00000000h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 415 Table 8-1. SATA Controller PCI Register Address Map (Sheet 2 of 3) Offset Start Offset End 24h Default Value Register ID - Description 27h "Offset 24h: When SCC is 01h (SATA-B0:D31:F2)" on page 428 00000001h 2Ch 2Dh "Offset 2Ch: Subsystem Vendor Identification Register (SATA-B0:D31:F2)" on page 428 0000h 2Eh 2Fh "Offset 2Eh: Subsystem Identification Register (SATA-B0:D31:F2)" on page 429 0000h 34h 34h "Offset 34h: Capabilities Pointer Register (SATA-B0:D31:F2)" on page 429 80h 3Ch 3Ch "Offset 3Ch: Interrupt Line Register (SATA-B0:D31:F2)" on page 429 00h 3Dh 3Dh "Offset 3Dh: Interrupt Pin Register (SATA-B0:D31:F2)" on page 430 See register description 40h 43h "Offset 40h: IDE Timing Register (SATA-B0:D31:F2)" on page 430 0000h 48h 48h "Offset 48h: Synchronous DMA Control Register (SATA-B0:D31:F2)" on page 431 00h 4Ah 4Bh "Offset 4Ah: Synchronous DMA Timing Register (SATA-B0:D31:F2)" on page 431 0000h 54h 57h "Offset 54h: IDE I/O Configuration Register (SATA-B0:D31:F2)" on page 432 00000000h 70h 71h "Offset 70h: PCI Power Management Capability Identification Register (SATA- B0:D31:F2)" on page 432 XX01h 72h 73h "Offset 72h: PCI Power Management Capabilities Register (SATA-B0:D31:F2)" on page 433 X003h 74h 75h "Offset 74h: PCI Power Management Control and Status Register (SATA-B0:D31:F2)" on page 434 XX08h 80h 81h "Offset 80h: Message Signaled Interrupt Capability Identification (SATA-B0:D31:F2)" on page 435 7005h 82h 83h "Offset 82h: Message Signaled Interrupt Message Control (SATA-B0:D31:F2)" on page 436 0000h 84h 87h "Offset 84h: Message Signaled Interrupt Message Address (SATA-B0:D31:F2)" on page 437 00000000h 88h 89h "Offset 88h: Message Signaled Interrupt Message Data (SATA-B0:D31:F2)" on page 437 0000h 90h 90h "Offset 90h: MAP--Address Map Register (SATA-B0:D31:F2)" on page 438 0000h 92h 93h "Offset 92h: PCS--Port Control and Status Register (SATA-B0:D31:F2)" on page 440 0000h 94h 97h "Offset 94h: SCLKCG--SATA Clock Gating Control Register (SATA-B0:D31:F2)" on page 441 00000000h 9Ch 9Fh "Offset 9Ch: SCLKGC--SATA Clock General Configuration Register (SATA-B0:D31:F2)" 00000000h on page 442 A0h A0h "Offset A0h: SIRI--SATA Indexed Registers Index (SATA-B0:D31:F2)" on page 443 00h A4h A7h "Offset A4h: STRD--SATA Indexed Register Data (SATA-B0:D31:F2)" on page 443 XXXXXXXXh A8h ABh "Offset A8h: SATACR0--SATA Capability Register 0 (SATA-B0:D31:F2)" on page 444 0010B012h ACh AFh "Offset ACh: SATACR1--SATA Capability Register 1 (SATA-B0:D31:F2)" on page 445 00000048h B0h B1h "Offset B0h: FLRCID--FLR Capability ID (SATA-B0:D31:F2)" on page 446 0009h B2h B3h "Offset B2h: FRLCLV--FLR Capability Length and Version (SATA-B0:D31:F2)" on page 446 xx06h B2h B3h "Offset B2h: FLRCLV--FLR Capability Length and Version (SATA-B0:D31:F2)" on page 447 xx06h B4h B5h "Offset B4h: FLRC--FLR Control (SATA-B0:D31:F2)" on page 447 0000h C0h C0h "Offset C0h: ATC--APM Trapping Control Register (SATA-B0:D31:F2)" on page 448 00h C4h C4h "Offset C4h: ATS--APM Trapping Status Register (SATA-B0:D31:F2)" on page 448 00h D0h D0h "Offset D0h: SP--Scratch Pad Register (SATA-B0:D31:F2)" on page 449 00000000h Intel(R) Communications Chipset 89xx Series - Datasheet 416 October 2012 Order Number: 327879-001US 8.0 Table 8-1. SATA Controller PCI Register Address Map (Sheet 3 of 3) Offset Start Offset End Register ID - Description Default Value E0h E3h "Offset E0h: BFCS--BIST FIS Control/Status Register (SATA-B0:D31:F2)" on page 450 00000000h E4h E7h "Offset E4h: BFTD1--BIST FIS Transmit Data1 Register (SATA-B0:D31:F2)" on page 451 00000000h E8h EBh "Offset E8h: BFTD2--BIST FIS Transmit Data2 Register (SATA-B0:D31:F2)" on page 452 00000000h Note: The PCH SATA controller is not arbitrated as a PCI device, therefore it does not need a master latency timer. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 417 8.1.1.1 Offset 00h: Vendor Identification Register (SATA-B0:D31:F2) Table 8-2. Offset 00h: Vendor Identification Register (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D31 :F2 Offset Start: 00h Offset End: 01h Default: 8086h Bit Range Bit Acronym 15 :00 VID Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 8086h RO Vendor Identification: This 16-bit value is assigned to Intel. Intel VID = 8086h 8.1.1.2 Offset 02h: Device Identification Register (SATA-B0:D31:F2) Table 8-3. Offset 02h: Device Identification Register (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bus:Device:Function: B0:D31 :F2 BAR: Configuration Offset Start: 02h Offset End: 03h Default: 2321h Bit Range Bit Acronym 15 :00 DID Power Well: Core Bit Description Sticky Device Identification 8.1.1.3 Offset 04h: PCI Command Register (SATA-B0:D31:F2) Table 8-4. Offset 04h: PCI Command Register (SATA-B0:D31:F2) Bit Reset Value Bit Access 2321h RO Description: View: PCI Size: B0:D31 Bus:Device:Function: :F2 BAR: Configuration Default: 0000h Bit Range Bit Acronym 15 :11 Reserved 10 INTx_Disable 09 FBE Power Well: Core Bit Description Reserved Sticky Bit Reset Value Bit Access 00h Interrupt Disable -- This disables pin-based INTx# interrupts. This bit has no effect on MSI operation. 0 = Internal INTx# messages are generated if there is an interrupt and MSI is not enabled. 1 = Internal INTx# messages will not be generated. 0h RW Fast Back to Back Enable -- Reserved as 0. 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 418 Offset Start: 04h Offset End: 05h October 2012 Order Number: 327879-001US 8.0 Table 8-4. Offset 04h: PCI Command Register (SATA-B0:D31:F2) Description: View: PCI Size: BAR: Configuration Bus:Device:Function: B0:D31 :F2 Default: 0000h Bit Range Bit Acronym Offset Start: 04h Offset End: 05h Power Well: Core Bit Reset Value Bit Access SERR# Enable -- Reserved as 0. 0h RO Bit Description Sticky 08 SERR_EN 07 WCC Wait Cycle Control -- Reserved as 0. 0h RO PER Parity Error Response: 0 = Disabled. SATA controller will not generate PERR# when a data parity error is detected. 1 = Enabled. SATA controller will generate PERR# when a data parity error is detected. 0h RW 05 VPS VGA Palette Snoop --Reserved as 0. 0h RO 04 PMWE Postable Memory Write Enable -- Reserved as 0. 0h RO 03 SCE Special Cycle Enable --Reserved as 0. 0h RO 02 BME Bus Master Enable -- This bit controls the PCH's ability to act as a PCI master for IDE Bus Master transfers. This bit does not impact the generation of completions for split transaction commands. 1 RW 01 MSE Memory Space Enable -- Controls access to the SATA controller's target memory space (for AHCI). This bit is RO `0' when not in AHCI modes. 1 RW/RO IOSE I/O Space Enable -- This bit controls access to the I/O space registers. 0 = Disables access to the Legacy or Native IDE ports (both Primary and Secondary) as well as the Bus Master I/ O registers. 1 = Enable. The Base Address register for the Bus Master registers should be programmed before this bit is set. 1 RW 06 00 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 419 8.1.1.4 Offset 06h: PCI Status Register (SATA-B0:D31:F2) For the writable bits, software must write a 1 to clear bits that are set. Writing a 0 to the bit has no effect. Table 8-5. Offset 06h: PCI Status Register (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: Default: 02B0h Offset Start: 06h Offset End: 07h Power Well: Bit Access Bit Description 15 DPE Detected Parity Error: 0 = Parity error not detected. 1 = Indicates that the PCH detected a parity error on the internal backbone. This bit gets set even if the Parity Error Response bit (B0:D31:D2; Offset 04h bit 6) is not set. RWC 14 SSE Signaled System Error -- Set when the LPC bridge signals a system error to the internal SERR# logic. RWC 13 RMA Received Master Abort: 0 = No master abort received. 1 = Set when the bridge receives a master abort status from the I/O data bus. RWC 12 RTA Reserved as 0 RO 11 STA Signaled Target Abort. RO DEVSEL# Timing Status: 01 = Hardwired; Controls the device select time for the SATA controller's PCI interface. RO 08 07 :05 DEV_STS DPED Data Parity Error Detected -- For PCH, this bit can only be set on read completions received from the bus when there is a parity error. SATA controller, as a master, either detects a parity error or sees the parity error line asserted, and the parity error response bit (bit 6 of the command register) is set. FB2BC Fast Back to Back Capable -- Reserved as 1. 04 UDF 03 66MHZ_CAP 02 :00 CAP_LIST Sticky Bit Reset Value Bit Acronym 10 :09 RWC RO User Definable Features -- Reserved as 0. RO 66MHz Capable -- Reserved as 1. RO Capabilities List -- This bit indicates the presence of a capabilities list. The minimum requirement for the capabilities list must be PCI power management for the SATA controller. RO Intel(R) Communications Chipset 89xx Series - Datasheet 420 B0:D31 :F2 October 2012 Order Number: 327879-001US 8.0 8.1.1.5 Offset 08h: Revision Identification Register (SATA-B0:D31:F2) Table 8-6. Offset 08h: Revision Identification Register (SATA-B0:D31:F2) Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B0:D31 :F2 Default: See bit description Bit Range Bit Acronym 07 :00 RID Offset Start: 08h Offset End: 08h Power Well: Bit Description Sticky Revision ID: 8-bit value indicating the stepping of the SATA Controller hardware Bit Reset Value Bit Access Variable RO 8.1.2 Programming Interface Register (SATA-B0:D31:F2) 8.1.2.1 Offset 0Ah: When Sub Class Code Register (B0:D31:F2:Offset 0Ah) = 01h Table 8-7. Offset 0Ah: When Sub Class Code Register (B0:D31:F2:Offset 0Ah) = 01h (Sheet 1 of 2) Description: View: PCI Size: 8 bit Bit Range Default: See bit description Bit Acronym 03 Reserved SNC October 2012 Order Number: 327879-001US Offset Start: 09h Offset End: 09h Power Well: Bit Description Sticky This read-only bit is a 1 to indicate that the PCH supports bus master operation 07 06 :04 Bus:Device:Function: B0:D31 :F2 BAR: Configuration Bit Reset Value Bit Access RO Reserved. Will always return 0. Secondary Mode Native Capable: 0 = Secondary controller only supports legacy mode. 1 = Secondary controller supports both legacy and native modes. When MAP.MV (B0:D31:F2:Offset 90:bits 1:0) is any value other than 00b, this bit reports as a 0. When MAP.MV is 00b, this bit reports as a 1. RO Intel(R) Communications Chipset 89xx Series - Datasheet 421 Table 8-7. Offset 0Ah: When Sub Class Code Register (B0:D31:F2:Offset 0Ah) = 01h (Sheet 2 of 2) Description: View: PCI Size: 8 bit Bit Range 02 01 00 BAR: Configuration Bus:Device:Function: B0:D31 :F2 Default: See bit description Offset Start: 09h Offset End: 09h Power Well: Sticky Bit Reset Value Bit Acronym Bit Description Bit Access SNE Secondary Mode Native Enable: Determines the mode that the secondary channel is operating in. 0 = Secondary controller operating in legacy (compatibility) mode 1 = Secondary controller operating in native PCI mode. When MAP.MV (B0:D31:F2:Offset 90:bits 1:0) is any value other than 00b, this bit is read-only (RO). When MAP.MV is 00b, this bit is read/write (R/W). If this bit is set by software, then the PNE bit (bit 0 of this register) must also be set by software. While in theory these bits can be programmed separately, such a configuration is not supported by hardware. RW PNC Primary Mode Native Capable: 0 = Primary controller only supports legacy mode. 1 = Primary controller supports both legacy and native modes. When MAP.MV (B0:D31:F2:Offset 90:bits 1:0) is any value other than 00b, this bit reports as a 0. When MAP.MV is 00b, this bit reports as a 1 RO PNE Primary Mode Native Enable: Determines the mode that the primary channel is operating in. 0 = Primary controller operating in legacy (compatibility) mode. 1 = Primary controller operating in native PCI mode. If this bit is set by software, then the SNE bit (bit 2 of this register) must also be set by software simultaneously. RW 8.1.2.2 Offset 0Ah: When Sub Class Code Register (B0:D31:F2:Offset 0Ah) = 06h Table 8-8. Offset 0Ah: When Sub Class Code Register (B0:D31:F2:Offset 0Ah) = 06h Description: View: PCI Size: 8 bit Bus:Device:Function: B0:D31 :F2 BAR: Configuration Default: 01h Bit Range Bit Acronym 07 :00 IF Power Well: Bit Description Interface - Indicates the SATA Controller supports AHCI, Revision 1.2. Intel(R) Communications Chipset 89xx Series - Datasheet 422 Offset Start: 0Ah Offset End: 0Ah Sticky Bit Reset Value Bit Access RO October 2012 Order Number: 327879-001US 8.0 8.1.2.3 Offset 0Ah: Sub Class Code Register (SATA-B0:D31:F2) Table 8-9. Offset 0Ah: Sub Class Code Register (SATA-B0:D31:F2) Description: View: PCI BAR: Configuration Size: 8 bit Bus:Device:Function: B0:D31 :F2 Default: See bit description Power Well: Bit Range Bit Acronym Bit Description 07 :00 SCC Sub Class Code: This field specifies the sub-class code of the controller, per the table below: 8.1.2.4 Offset Start: 0Ah Offset End: 0Ah Sticky Bit Reset Value Bit Access RO Offset 0Bh: Base Class Code Register (B0:D31:F2) Table 8-10. Offset 0Bh: Base Class Code Register (B0:D31:F2) Description: View: PCI Size: 8 bit Default: 01h Bit Range Bit Acronym 07 :00 BCC 8.1.2.5 B0:D31 Bus:Device:Function: :F2 BAR: Configuration Offset Start: 0Bh Offset End: 0Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Base Class Code: 01h = Mass storage device RO Offset 0Dh: Primary Master Latency Timer Register (B0:D31:F2) Table 8-11. Offset 0Dh: Primary Master Latency Timer Register (B0:D31:F2) Description: View: PCI Size: 8 bit BAR: Configuration B0:D31 Bus:Device:Function: :F2 Default: 00h Power Well: Bit Range Bit Acronym Bit Description 07 :00 MLTC Master Latency Timer Count: 00h = Hardwired. The SATA controller is implemented internally, and is not arbitrated as a PCI device, so it does not need a Master Latency Timer. October 2012 Order Number: 327879-001US Offset Start: 0Dh Offset End: 0Dh Sticky Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 423 8.1.2.6 Offset 0Eh: Header Type (SATA-B0:D31:F2) Table 8-12. Offset 0Eh: Header Type (SATA-B0:D31:F2) Description: View: PCI Size: 8 bit Bit Range 07 06 :00 BAR: Configuration Bus:Device:Function: B0:D31 :F2 Default: 80h Offset Start: 0Eh Offset End: 0Eh Power Well: Bit Description MFD Multi-function Device: Indicates this SATA controller is not part of a multifunction device. RO Header Layout: Indicates that the SATA controller uses a target device layout. RO HL Sticky Bit Reset Value Bit Acronym Bit Access 8.1.2.7 Offset 10h: Primary Command Block Base Address Register (SATA- B0:D31:F2) Note: This 8-byte I/O space is used in native mode for the Primary Controller's Command Block. Table 8-13. Offset 10h: Primary Command Block Base Address Register (SATA- B0:D31:F2) Description: View: PCI Size: 32 bit Default: 00000001h Bit Range Bit Acronym 31 :16 Reserved 15 :03 BA 02 :01 Reserved 00 Bus:Device:Function: B0:D31 :F2 BAR: Configuration RTE Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Base Address -- This field provides the base address of the I/O space (8 consecutive I/O locations). RW Reserved Resource Type Indicator -- Hardwired to 1 to indicate a request for I/O space. Intel(R) Communications Chipset 89xx Series - Datasheet 424 Offset Start: 10h Offset End: 13h RO October 2012 Order Number: 327879-001US 8.0 8.1.2.8 Offset 14h: Primary Control Block Base Address Register (SATA- B0:D31:F2) Note: This 4-byte I/O space is used in native mode for the Primary Controller's Command Block. Table 8-14. Offset 14h: Primary Control Block Base Address Register (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit BAR: Configuration B0:D31 :F2 Default: 00000001h Bit Range Bit Acronym 31 :16 Reserved 15 :02 Bus:Device:Function: Offset Start: 14h Offset End: 17h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Base Address -- This field provides the base address of the I/O space (4 consecutive I/O locations). BA 01 Reserved 00 RTE RW Reserved Resource Type Indicator -- Hardwired to 1 to indicate a request for I/O space. RO 8.1.2.9 Offset 18h: Secondary Command Block Base Address Register (IDE D31:F1) Note: This 4-byte I/O space is used in native mode for the Secondary Controller's Command Block. Table 8-15. Offset 18h: Secondary Command Block Base Address Register (IDE D31:F1) Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000001h Bit Range Bit Acronym 31 :16 Reserved 15 :03 BA 02 :01 Reserved 00 B0:D31 Bus:Device:Function: :F2 RTE October 2012 Order Number: 327879-001US Offset Start: 18h Offset End: 1Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Base Address -- This field provides the base address of the I/O space (8 consecutive I/O locations). RW Reserved Resource Type Indicator -- Hardwired to 1 to indicate a request for I/O space. RO Intel(R) Communications Chipset 89xx Series - Datasheet 425 8.1.2.10 Offset 1Ch: Secondary Control Block Base Address Register (IDE B0:D31:F2) Note: This 4-byte I/O space is used in native mode for the Secondary Controller's Command Block. Table 8-16. Offset 1Ch: Secondary Control Block Base Address Register (IDE B0:D31:F2) Description: View: PCI Size: 32 bit BAR: Configuration B0:D31 :F2 Default: 00000001h Bit Range Bit Acronym 31 :16 Reserved 15 :02 BA Offset Start: 1Ch Offset End: 1Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Base Address -- This field provides the base address of the I/O space (4 consecutive I/O locations). 01 Reserved 00 RTE 8.1.2.11 Bus:Device:Function: RW Reserved Resource Type Indicator -- Hardwired to 1 to indicate a request for I/O space. RO Offset 20h: Legacy Bus Master Base Address Register (SATA- B0:D31:F2) The Bus Master IDE interface function uses Base Address register 5 to request a 16byte IO space to provide a software interface to the Bus Master functions. Only 12 bytes are actually used (6 bytes for primary, 6 bytes for secondary). Only bits [15:4] are used to decode the address. Table 8-17. Offset 20h: Legacy Bus Master Base Address Register (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Default: 00000001h Bit Range Bit Acronym 31 :16 Reserved 15 :05 BA 00 Reserved RTE Bit Description Sticky Bit Reset Value Bit Access Reserved Base Address -- This field provides the base address of the I/O space (16 consecutive I/O locations). RW RW/RO Reserved RO Resource Type Indicator -- Hardwired to 1 to indicate a request for I/O space. RO Intel(R) Communications Chipset 89xx Series - Datasheet 426 Offset Start: 20h Offset End: 23h Power Well: Base -- When SCC is 01h, this bit will be R/W resulting in requesting 16B of I/O space. When SCC is not 01h, this bit will be Read Only 0, resulting in requesting 32B of I/O space. 04 03 :01 Bus:Device:Function: B0:D31 :F2 BAR: Configuration October 2012 Order Number: 327879-001US 8.0 8.1.3 AHCI Base Address Register/Serial ATA Index Data Pair Base Address (SATA-B0:D31:F2) When the programming interface is not IDE (i.e. SCC is not 01h), this register is named ABAR. When the programming interface is IDE, this register becomes SIDPBA. Hardware does not clear those BA bits when switching from IDE component to non-IDE component or vice versa. BIOS is responsible for clearing those bits to 0 since the number of writable bits changes after component switching (as indicated by a change in SCC). In the case, this register will then have to be re-programmed to a proper value. 8.1.3.1 Offset 24h: When SCC is Not 01h (SATA-B0:D31:F2) When the programming interface is not IDE, the register represents a memory BAR allocating space for the AHCI memory registers defined in Section 8.4. Note: The ABAR register must be set to a value of 0001_0000h or greater. Table 8-18. Offset 24h: When SCC is Not 01h (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000000h Bit Range Bit Acronym 31 :11 BA 10 :04 Reserved 03 02 :01 00 Bus:Device:Function: B0:D31 :F2 Offset Start: 24h Offset End: 27h Power Well: Bit Description Sticky Base Address -- Base address of register memory space (aligned to 1 KB) Bit Reset Value Bit Access RW Reserved PF Prefetchable -- Indicates that this range is not prefetchable RO TP Type -- Indicates that this range can be mapped anywhere in 32-bit address space. RO Resource Type Indicator -- Hardwired to 0 to indicate a request for register memory space. RO RTE October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 427 8.1.3.2 Offset 24h: When SCC is 01h When the programming interface is IDE, the register becomes an I/O BAR allocating 16 bytes of I/O space for the I/O-mapped registers. Although 16 bytes of locations are allocated, only 8 bytes are used to as SINDX and SDATA registers; with the remaining 8 bytes preserved for future enhancement. Table 8-19. Offset 24h: When SCC is 01h (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit BAR: Configuration Bit Acronym 31 :16 Reserved 15 :04 BA 03 :01 Reserved 8.1.3.3 B0:D31 :F2 Default: 00000001h Bit Range 00 Bus:Device:Function: Offset Start: 24h Offset End: 27h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Base Address -- Base address of the I/O space. RW Reserved Resource Type Indicator -- Indicates a request for I/O space. RTE RO Offset 2Ch: Subsystem Vendor Identification Register (SATA- B0:D31:F2) Table 8-20. Offset 2Ch: Subsystem Vendor Identification Register (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bus:Device:Function: B0:D31 :F2 BAR: Configuration Default: 0000h Bit Range Bit Acronym 15 :00 SVID Power Well: Bit Description Subsystem Vendor ID -- Value is written by BIOS. No hardware action taken on this value. Intel(R) Communications Chipset 89xx Series - Datasheet 428 Offset Start: 2Ch Offset End: 2Dh Sticky Bit Reset Value Bit Access RWO October 2012 Order Number: 327879-001US 8.0 8.1.3.4 Offset 2Eh: Subsystem Identification Register (SATA-B0:D31:F2) Table 8-21. Offset 2Eh: Subsystem Identification Register (SATA-B0:D31:F2) Description: View: PCI BAR: Configuration Size: 16 bit B0:D31 :F2 Offset Start: 2Eh Offset End: 2Fh Default: 0000h Bit Range Bit Acronym 15 :00 SVID 8.1.3.5 Bus:Device:Function: Power Well: Core Bit Description Bit Reset Value Sticky Bit Access Subsystem Vendor ID -- Value is written by BIOS. No hardware action taken on this value. RWO Offset 34h: Capabilities Pointer Register (SATA-B0:D31:F2) Table 8-22. Offset 34h: Capabilities Pointer Register (SATA-B0:D31:F2) Description: View: PCI B0:D31 Bus:Device:Function: :F2 BAR: Configuration Size: 8 bit Default: 80h Power Well: Bit Range Bit Acronym Bit Description 07 :00 CAP_PTR Capabilities Pointer -- Indicates that the first capability pointer offset is 80h. This value changes to 70h if the Sub Class Code (SCC) (Dev 31:F2:0Ah) is configure as IDE mode (value of 01). 8.1.3.6 Offset Start: 34h Offset End: 34h Sticky Bit Reset Value Bit Access RO Offset 3Ch: Interrupt Line Register (SATA-B0:D31:F2) Table 8-23. Offset 3Ch: Interrupt Line Register (SATA-B0:D31:F2) Description: View: PCI Size: 8 bit BAR: Configuration B0:D3 Bus:Device:Function: 1:F2 Default: 00h Power Well: Bit Range Bit Acronym Bit Description 07 :00 IL Interrupt Line -- This field is used to communicate to software the interrupt line that the interrupt pin is connected to. Interrupt Line register is not reset by FLR October 2012 Order Number: 327879-001US Offset Start: 3Ch Offset End: 3Ch Sticky Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 429 8.1.3.7 Offset 3Dh: Interrupt Pin Register (SATA-B0:D31:F2) Table 8-24. Offset 3Dh: Interrupt Pin Register (SATA-B0:D31:F2) Description: View: PCI BAR: Configuration Size: 8 bit Bus:Device:Function: Default: See register description Offset Start: 3Dh Offset End: 3Dh Power Well: Bit Range Bit Acronym Bit Description 07 :00 IP Interrupt Pin -- This reflects the value of D31IP.SIP (Chipset Config Registers:Offset 3100h:bits 11:8). 8.1.3.8 B0:D3 1:F2 Sticky Bit Reset Value Bit Access RO Offset 40h: IDE Timing Register (SATA-B0:D31:F2) Address Offset: Primary: 40h-41h Secondary Offset: 42h-43h Table 8-25. Offset 40h: IDE Timing Register (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range 15 Bus:Device:Function: B0:D3 1:F2 BAR: Configuration Default: 0000h Bit Acronym IDE Offset Start: 40h Offset End: 43h Power Well: Bit Description IDE Decode Enable -- Individually enable/disable the Primary or Secondary decode. 0 = Disable. 1 = Enables the PCH to decode the associated Command Blocks (1F0-1F7h for primary, 170-177h for secondary) and Control Block (3F6h for primary and 376h for secondary). This bit effects the IDE decode ranges for both legacy and native-Mode decoding. Sticky Bit Reset Value Bit Access RW Note: This bit affects SATA operation in both combined and non-combined ATA modes. 14 :00 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 430 October 2012 Order Number: 327879-001US 8.0 8.1.3.9 Offset 48h: Synchronous DMA Control Register (SATA-B0:D31:F2) Address Offset: Primary: 48h Table 8-26. Offset 48h: Synchronous DMA Control Register (SATA-B0:D31:F2) Description: View: PCI BAR: Configuration Size: 8 bit Bus:Device:Function: Default: 00h Bit Acronym Bit Range Offset Start: 48h Offset End: 48h Power Well: Bit Description Sticky Bit Reset Value 07 :04 Reserved IReserved 03 :00 SDMA_CNT Field 1--R/W. This field is R/W to SDMA_C maintain software compatibility. This field has no NT effect on hardware. 8.1.3.10 B0:D3 1:F2 Bit Access RW Offset 4Ah: Synchronous DMA Timing Register (SATA-B0:D31:F2) Address Offset: Primary: 4Ah-4Bh Table 8-27. Offset 4Ah: Synchronous DMA Timing Register (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D3 1:F2 Default: 0000h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value 15 :14 Reserved Reserved 13 :12 SDMA_TIM Field 4--R/W. This field is R/W to SDMA_TI maintain software compatibility. This field has no M4 effect on hardware. 11 :10 Reserved Reserved 09 :08 SDMA_TIM Field 3--R/W. This field is R/W to SDMA_TI maintain software compatibility. This field has no M3 effect on hardware. 07 :06 Reserved Reserved 05 :04 Field 2--R/W. This field is R/W to SDMA_TI SDMA_TIM maintain software compatibility. This field has no M2 effect on hardware. 03 :02 Reserved Reserved 01 :00 Field 1--R/W. This field is R/W to SDMA_TI SDMA_TIM maintain software compatibility. This field has no M1 effect on hardware. October 2012 Order Number: 327879-001US Offset Start: 4Ah Offset End: 4Bh Bit Access RW RW RW RW Intel(R) Communications Chipset 89xx Series - Datasheet 431 8.1.3.11 Offset 54h: IDE I/O Configuration Register (SATA-B0:D31:F2) Address Offset: Primary: 54h-57h Table 8-28. Offset 4Ah: IDE I/O Configuration Register (SATA-B0:D31:F2) Description: View: PCI BAR: Configuration Size: 32 bit Bus:Device:Function: B0:D3 1:F2 Default: 00000000h Bit Acronym Bit Range Offset Start: 54h Offset End: 57h Power Well: Bit Description 31 :24 Reserved Reserved 23 :12 IDE_CONFIG Field 2--R/W. This field is R/W to IDE_CFG maintain software compatibility. This field has no 2 effect on hardware. 11 :08 Reserved Reserved 07 :00 IDE_CONFIG Field 1--R/W. This field is R/W to IDE_CFG maintain software compatibility. This field has no 1 effect on hardware. Sticky Bit Reset Value Bit Access RW RW 8.1.4 PCI Power Management Capabilities 8.1.4.1 Offset 70h: PCI Power Management Capability Identification Register (SATA-B0:D31:F2) Table 8-29. Offset 70h: PCI Power Management Capability Identification Register (SATA- B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range Bus:Device:Function: B0:D3 1:F2 BAR: Configuration Default: XX01h Bit Acronym 15 :08 NEXT 07 :00 CID Power Well: Bit Description Sticky Bit Reset Value Bit Access Next Capability: B0h -- if SCC = 01h (IDE mode) indicating next item is FLR capability pointer. A8h -- for all other values of SCC to point to the next capability structure. RO Capability ID -- Indicates that this pointer is a PCI power management. RO Intel(R) Communications Chipset 89xx Series - Datasheet 432 Offset Start: 70h Offset End: 71h October 2012 Order Number: 327879-001US 8.0 8.1.4.2 Offset 72h: PCI Power Management Capabilities Register (SATA- B0:D31:F2) Table 8-30. Offset 72h: PCI Power Management Capabilities Register (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: Default: X003h Bit Acronym B0:D3 1:F2 Offset Start: 72h Offset End: 73h Power Well: Bit Description Sticky Bit Reset Value Bit Access PME Support: 00000 = If SCC = 01h, indicates no PME support in PME_SUP IDE mode. 01000 = If SCC is not 01h, in a non-IDE mode, indicates PME# can be generated from the D3HOT state in the SATA host controller. RO 10 Support -- Hardwired to 0. The D2 state is not D2_SUP D2 supported RO 09 Support -- Hardwired to 0. The D1 state is not D1_SUP D1 supported RO AUX_CUR Auxiliary Current -- PME# from D3COLD state is not supported, therefore this field is 000b. RO 15 :11 08 :06 05 DSI Device Specific Initialization -- Hardwired to 0 to indicate that no device-specific initialization is required. RO 04 Reserved Reserved 03 Clock -- Hardwired to 0 to indicate that PCI PME_CLK PME clock is not required to generate PME#. RO Version -- Hardwired to 011 to indicates support for Revision 1.2 of the PCI Power Management Specification. RO 02 :00 VER October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 433 8.1.4.3 Offset 74h: PCI Power Management Control and Status Register (SATA-B0:D31:F2) Table 8-31. Offset 74h: PCI Power Management Control and Status Register (SATA- B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range 15 14 :09 08 07 :04 03 02 01 :00 BAR: Configuration Bus:Device:Function: Default: XX08h Offset Start: 74h Offset End: 75h Power Well: Bit Acronym Bit Description PMES PME Status -- Bit is set when a PME event is to be requested, and if this bit and PMEE is set, a PME# will be generated from the SATA controller Note: Whenever SCC = 01h, hardware will automatically change the attribute of this bit to RO `0'. Software is advised to clear PMEE and PMES together prior to changing SCC thru MAP.SMS. This bit is not reset by Function Level Reset. Sticky Bit Reset Value Bit Access RWC Reserved Reserved PMEE PME Enable -- When set, the SATA controller generates WAKE# from D3HOT on a wake event. Note: Whenever SCCSCC = 01h, hardware will automatically change the attribute of this bit to RO `0'. Software is advised to clear PMEE and PMES together prior to changing SCC thru MAP.SMS. This bit is not reset by Function Level Reset. RW Reserved Reserved No Soft Reset -- These bits are used to indicate whether devices transitioning from D3HOT state to D0 state will perform an internal reset. 0 = Device transitioning from D3HOT state to D0 state perform an internal reset. 1 = Device transitioning from D3HOT state to D0 state do not perform an internal reset. Configuration content is preserved. Upon transition NSFRST from the D3 HOT state to D0 state initialized state, no additional operating system intervention is required to preserve configuration context beyond writing to the PowerState bits. Regardless of this bit, the controller transition from D3HOT state to D0 state by a system or bus segment reset will return to the state D0 uninitialized with only PME context preserved if PME is supported and enabled. RO Reserved Reserved PS Power State -- These bits are used both to determine the current power state of the SATA controller and to set a new power state. 00 = D0 state 11 = D3HOT state When in the D3HOT state, the controller's configuration space is available, but the I/O and memory spaces are not. Additionally, interrupts are blocked. Intel(R) Communications Chipset 89xx Series - Datasheet 434 B0:D31 :F2 RW October 2012 Order Number: 327879-001US 8.0 8.1.5 Message Signaled Interrupt Capability The following set of MSI registers are only applicable to AHCI SKUs. There is no support for MSI when the software is operating in legacy (IDE) mode when AHCI is not enabled. Prior to switching from AHCI to IDE mode, software must make sure that MSI is disabled. 8.1.5.1 Offset 80h: Message Signaled Interrupt Capability Identification (SATA-B0:D31:F2) There is no support for MSI when the software is operating in legacy (IDE) mode when AHCI is not enabled. Prior to switching from AHCI to IDE mode, software must make sure that MSI is disabled. Table 8-32. Offset 80h: Message Signaled Interrupt Capability Identification (SATA- B0:D31:F2) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D31 :F2 Default: 7005h Offset Start: 80h Offset End: 81h Power Well: Bit Range Bit Acronym Bit Description 15 :08 NEXT Next Pointer -- Indicates the next item in the list is the PCI power management pointer. RO 07 :00 CID Capability ID -- Capabilities ID indicates MSI. RO October 2012 Order Number: 327879-001US Sticky Bit Reset Value Bit Access Intel(R) Communications Chipset 89xx Series - Datasheet 435 8.1.5.2 Offset 82h: Message Signaled Interrupt Message Control (SATA- B0:D31:F2) There is no support for MSI when the software is operating in legacy (IDE) mode when AHCI is not enabled. Prior to switching from AHCI to IDE mode, software must make sure that MSI is disabled. Table 8-33. Offset 82h: Message Signaled Interrupt Message Control (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range 15 :08 07 06 :04 03 :01 B0:D31 Bus:Device:Function: :F2 BAR: Configuration Offset Start: 82h Offset 83h End: Power Well: Default: 0000h Bit Acronym Bit Description Sticky Bit Reset Bit Access Value Reserved Reserved C64 64 Bit Address Capable: Capable of generating a 32-bit message only. RO MME Multiple Message Enable: When this field is cleared to `000' (and MSIE is set), only a single MSI message will be generated for all SATA ports, and bits[15:0] of the message vector will be driven from MD[15:0]. RO MMC Multiple Message Capable -- Indicates the number of interrupt messages supported by the PCH SATA controller. 000 =1 MSI Capable (When SCC bit is set to 01h. MSI is not supported in IDE mode) 100 = 8 MSI Capable RO MSI Enable -- If set, MSI is enabled and traditional interrupt pins are not used to generate interrupts. This bit is RW when SC.SCC is not 01h and is read-only 0 when SCC is 01h. CMD.ID bit has no effect on MSI. 00 MSIE Note: Software must clear this bit to `0' to disable MSI first before changing the number of messages allocated in the MMC field. Software must also make sure this bit is cleared to `0' when operating in legacy mode (when GHC.AE = 0). Intel(R) Communications Chipset 89xx Series - Datasheet 436 RW/RO October 2012 Order Number: 327879-001US 8.0 8.1.5.3 Offset 84h: Message Signaled Interrupt Message Address (SATA- B0:D31:F2) There is no support for MSI when the software is operating in native (IDE) mode when AHCI is not enabled. Prior to switching from AHCI to IDE mode, software must make sure that MSI is disabled. Table 8-34. Offset 84h: Message Signaled Interrupt Message Address (SATA-B0:D31:F2) Description: View: PCI BAR: Configuration Size: 32 bit B0:D3 1:F2 Default: 00000000h Bit Range Bit Acronym 31 :02 ADDR 01 :00 8.1.5.4 Bus:Device:Function: Offset Start: 84h Offset End: 87h Power Well: Bit Description Sticky Bit Reset Value Bit Access Address -- Lower 32 bits of the system specified message address, always DWORD aligned. RW Reserved Reserved Offset 88h: Message Signaled Interrupt Message Data (SATA- B0:D31:F2) There is no support for MSI when the software is operating in legacy (IDE) mode when AHCI is not enabled. Prior to switching from AHCI to IDE mode, software must make sure that MSI is disabled. Table 8-35. Offset 88h: Message Signaled Interrupt Message Data (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range 15 :00 BAR: Configuration Bus:Device:Function: B0:D31 :F2 Default: 0000h Power Well: Bit Acronym Bit Description DATA Data -- this 16-bit field is programmed by system software if MSI is enabled. Its content is driven onto the lower word of the data bus of the MSI memory write transaction. When the MME field is set to `001' or `010', bit [0] and bits [1:0] respectively of the MSI memory write transaction will be driven based on the source of the interrupt rather than from MD[2:0]. See the description of the MME field. October 2012 Order Number: 327879-001US Offset Start: 88h Offset End: 89h Sticky Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 437 8.1.5.5 Offset 90h: MAP--Address Map Register (SATA-B0:D31:F2) Table 8-36. Offset 90h: MAP--Address Map Register (SATA-B0:D31:F2) (Sheet 1 of 2) Description: View: PCI Size: 16 bit Bit Range 15 BAR: Configuration Bus:Device:Function: B0:D31 :F2 Default: 0000h Bit Acronym Offset Start: 90h Offset End: 90h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved SATA Port Disable (SPD). This register provides a bit corresponding to each of the SATA port on the SATA1 controller. A `1' in a bit position corresponding to the SATA port prevents the SATA port from being enabled via config PCS.PxE. Write of `1' to PCS.PxE has no effect when the corresponding SPD[x] bit is `1' Bit assignment to port: SPD[0] corresponds to SATA port 0; SPD[1] to port 1 and so on. 14 :08 SPD In preventing a port(s) from being enabled, BIOS shall first configures MAP.SPD. And only then BIOS configures the PCS.PxE. RWO In order to securely prevent port 4 and/or 5 from enabled, BIOS shall ensure: * port 4 - D31F2 MAP[12]='1' and D31F5 MAP[8]='1' * port 5 - D31F2 MAP[13]='1' and D31F5 MAP[9]='1' This field is not reset by FLR. SATA Mode Select -- SW programs these bits to control the mode in which the SATA Controller should operate: 00b = IDE mode 01b = AHCI mode 10b = Reserved 11b = Reserved Note: 1. 07 :06 SMS The SATA Function Device ID will change based on the value of this register. 2. When switching from AHCI mode to IDE mode, a 2 port SATA controller (Device 31, Function 5) will be enabled. 3. AHCI mode may only be selected when MV = 00 4. 4. Programming these bits with values that are invalid will result in indeterministic behavior by the HW 5. 5. SW shall not manipulate SMS during runtime operation; i.e. the OS will not do this. The BIOS may choose to switch from one mode to another during POST. These bits are not reset by Function Level Reset. Intel(R) Communications Chipset 89xx Series - Datasheet 438 RW October 2012 Order Number: 327879-001US 8.0 Table 8-36. Offset 90h: MAP--Address Map Register (SATA-B0:D31:F2) (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range 05 BAR: Configuration Bus:Device:Function: B0:D31 :F2 Default: 0000h Bit Acronym SC Offset Start: 90h Offset End: 90h Power Well: Bit Description Sticky SATA Port-to-Controller Configuration -- This bit changes the number of SATA ports available within each SATA Controller. 0 = Up to 4 SATA ports are available for Controller 1 (Device 31 Function 2) with ports [3:0] and up to 2 SATA ports are available for Controller 2 (Device 31 Function 5) with ports [5:4]. 0 = Up to 6 SATA ports are available for Controller 1 (Device 31 Function 2) with ports [5:0] and no SATA ports are available for Controller 2 (Device 31 Function 5). Bit Reset Value Bit Access RW This bit should be set to 1 only in AHCI mode. This bit is not reset by Function Level Reset. 04 :02 01 :00 8.1.5.6 Reserved Reserved MV Map Value -- Reserved Offset 92h: PCS - Port Control and Status Register (SATA-B0:D31:F2) By default, the SATA ports are set to the disabled state (bits [5:0] = `0'). When enabled by software, the ports can transition between the on, partial, and slumber states and can detect devices. When disabled, the port is in the "off" state and cannot detect any devices. If an AHCI-aware enabled operating system is being booted, then system BIOS shall insure that all supported SATA ports are enabled prior to passing control to the OS. Once the AHCI aware OS is booted it becomes the enabling/disabling policy owner for the individual SATA ports. This is accomplished by manipulating a port's PxSCTL and PxCMD fields. Because an AHCI aware OS will typically not have knowledge of the PxE bits and because the PxE bits act as master on/off switches for the ports, pre-boot software must insure that these bits are set to `1' prior to booting the OS, regardless as to whether or not a device is currently on the port. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 439 Table 8-37. Offset 92h: PCS--Port Control and Status Register (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range 15 14 13 12 11 :06 05 04 03 :00 BAR: Configuration Bus:Device:Function: Default: 0000h Offset Start: 92h Offset End: 93h Power Well: Bit Acronym Bit Description ORM OOB Retry Mode: 0 = The SATA controller will not retry after an OOB failure 1 = The SATA controller will continue to retry after an OOB failure until successful (infinite retry) Sticky Bit Reset Value Bit Access RW Reserved Reserved P5P Port 5 Present -- The status of this bit may change at any time. This bit is cleared when the port is disabled via P5E. This bit is not cleared upon surprise removal of a device. 0 = No device detected. 1 = The presence of a device on Port 5 has been detected. RO P4P Port 4 Present -- The status of this bit may change at any time. This bit is cleared when the port is disabled via P4E. This bit is not cleared upon surprise removal of a device. 0 = No device detected. 1 = The presence of a device on Port 4 has been detected. RO Reserved Reserved P5E Port 5 Enabled: 0 = Disabled. The port is in the `off' state and cannot detect any devices. 1 = Enabled. The port can transition between the on, partial, and slumber states and can detect devices. Note: This bit takes precedence over P5CMD.SUD (offset ABAR+298h:bit 1) If MAP.SC is `0', if SCC is `01h' this bit will be read only `0' or if MAP.SPD[5] is `1'. RW P4E Port 4 Enabled: 0 = Disabled. The port is in the `off' state and cannot detect any devices. 1 = Enabled. The port can transition between the on, partial, and slumber states and can detect devices. Note: This bit takes precedence over P4CMD.SUD (offset ABAR+298h:bit 1) If MAP.SC is `0', if SCC is `01h' this bit will be read only `0' or if MAP.SPD[4] is `1'. RW Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 440 B0:D31 :F2 October 2012 Order Number: 327879-001US 8.0 8.1.5.7 Offset 94h: SCLKCG--SATA Clock Gating Control Register (SATA- B0:D31:F2) Table 8-38. Offset 94h: SCLKCG--SATA Clock Gating Control Register (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :30 29 :28 27 :09 BAR: Configuration Bus:Device:Function: Default: 00000000h Bit Acronym B0:D31 :F2 Offset Start: 94h Offset End: 97h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved PCD Port Clock Disable: 0 = All clocks to the associated port logic will operate normally. 1 = The backbone clock driven to the associated port logic is gated and will not toggle. Bit 29: Port 5 Bit 28: Port 4 If a port is not available, software shall set the corresponding bit to 1. Software can also set the corresponding bits to 1 on ports that are disabled. Software cannot set the PCD [port x]='1' if the corresponding PCS.PxE='1' in either Dev31Func2 or Dev31Func5 (dual controller IDE mode). RW Reserved Reserved 08 :00 October 2012 Order Number: 327879-001US SCLKCG Field 1 -- BIOS must program these bits to 183h. RW Intel(R) Communications Chipset 89xx Series - Datasheet 441 8.1.5.8 Offset 9Ch: SCLKGC--SATA Clock General Configuration Register (SATA-B0:D31:F2) Table 8-39. Offset 9Ch: SCLKGC--SATA Clock General Configuration Register (SATA- B0:D31:F2) Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000000h Bit Range Bit Acronym 31 :07 Reserved 00 Offset Start: 9Ch Offset End: 9Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved RW SATA2-port Configuration Indicator: 0 = Normal configuration. 1 = One IDE Controller is implemented supporting only two ports for a Primary Master and a Secondary Master. Note: When set, BIOS must ensure that bit 2 and bit 3 of the AHCI PI registers are zeros. BIOS must also make sure that Port 2 and Port 3 are disabled (via PCS configuration register) and the port clocks are gated (via SCLKCG configuration register). RO SATA4-port All Master Configuration Indicator: 0 = Normal configuration. 1 = Two IDE Controllers are implemented, each supporting two ports for a Primary Master and a SATA4PMIND Secondary Master. Note: When set, BIOS must ensure that bit 2 and bit 3 of the AHCI PI registers are zeros. BIOS must also make sure that Port 2 and Port 3 are disabled (via PCS configuration register) and the port clocks are gated (via SCLKCG configuration register). RO SATA2PIND Intel(R) Communications Chipset 89xx Series - Datasheet 442 B0:D31 :F2 SATA Traffic Monitor: 00000b = Disable. 00011b = Enable. SATA Traffic Monitor allows for aggressive C2 Pop down by monitoring SATA bus mastering activity. When enabled, BIOS must ensure bit 3 and bit 4 of Cx_STATE_CNF (Cx State Configuration Register) are ones. Note: This field is reset by PLTRST# and BIOS is required to reprogram the value after resuming from S3-S5. All other bit combinations are Reserved. 06 :02 01 Bus:Device:Function: October 2012 Order Number: 327879-001US 8.0 8.1.5.9 Offset A0h: SIRI--SATA Indexed Registers Index (SATA-B0:D31:F2) This register provides access to the registers in SATA Indexed Registers. Table 8-40. Offset A0h: SIRI--SATA Indexed Registers Index (SATA-B0:D31:F2) Description: View: PCI BAR: Configuration Size: 8 bit Bus:Device:Function: Default: 00h Bit Acronym Bit Description 07 :02 IDX Index -- This field is a 5-bit index pointer into the SATA Indexed Register space. Data is written into and read from the SIRD register (B0:D31:F2:A4h). 01 :00 Offset Start: A0h Offset End: A0h Power Well: Bit Range 8.1.5.10 B0:D31 :F2 Sticky Bit Reset Value Bit Access RW Reserved Reserved Offset A4h: STRD--SATA Indexed Register Data (SATA-B0:D31:F2) This register is used to read/write the register written in the Offset A0h: SATA Indexed Registers Index (SATA - B0:D31:F2). Table 8-41. Offset A4h: STRD--SATA Indexed Register Data (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B0:D31 :F2 Default: XXXXXXXXh Bit Range Bit Acronym 31 :00 DTA October 2012 Order Number: 327879-001US Offset Start: A4h Offset End: A7h Power Well: Bit Description Sticky Bit Reset Value Data -- 32-bit data value that is written to the register pointed to by SIRI (B0:D31:F2:A0h) or read from the register pointed to by SIRI. Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 443 8.1.5.11 Offset A8h: SATACR0--SATA Capability Register 0 (SATA-B0:D31:F2) This register shall be read-only 0 when SCC is 01h. Table 8-42. Offset A8h: SATACR0--SATA Capability Register 0 (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :24 BAR: Configuration Bus:Device:Function: Default: 0010B012h Bit Acronym Offset Start: A8h Offset End: ABh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved 23 :20 MAJREV Major Revision -- Major revision number of the SATA Capability Pointer implemented. RO 19 :16 MINREV Minor Revision -- Minor revision number of the SATA Capability Pointer implemented. RO 15 :08 NEXT 07 :00 CAP Next Capability Pointer -- Points to the next capability structure. These bits are not reset by Function Level Reset. Capability ID -- This value of 12h has been assigned by the PCI SIG to designate the SATA Capability Structure. Intel(R) Communications Chipset 89xx Series - Datasheet 444 B0:D31 :F2 RWO RO October 2012 Order Number: 327879-001US 8.0 8.1.5.12 Offset ACh: SATACR1--SATA Capability Register 1 (SATA-B0:D31:F2) This register shall be read-only 0 when SCC is 01h. Table 8-43. Offset ACh: SATACR1--SATA Capability Register 1 (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Bus:Device:Function: Default: 00000048h Bit Acronym B0:D31 :F2 Offset Start: ACh Offset End: AFh Power Well: Bit Description Sticky Bit Reset Value Bit Access 31 :16 Reserved Reserved 15 :04 BAR Offset -- Indicates the offset into the BAR where the Index/Data pair are located (in Dword granularity). The Index and Data I/O registers are located at offset 10h within the I/O space defined by LBAR. A value of 004h indicates offset 10h. 000h = 0h offset BAROFST 001h = 4h offset 002h = 8h offset 003h = Bh offset 004h = 10h offset ... FFFh = 3FFFh offset (max 16KB) RO 03 :00 BAR Location -- Indicates the absolute PCI Configuration Register address of the BAR containing the Index/Data pair (in Dword granularity). The Index and Data I/O registers reside within the space defined by LBAR in the SATA controller. A value of 8h indicates offset 20h, which is LBAR. 0000 - 0011b = reserved 0100b = 10h => BAR0 BARLOC 0101b = 14h => BAR1 0110b = 18h => BAR2 0111b = 1Ch => BAR3 1000b = 20h => LBAR 1001b = 24h => BAR5 1010 - 1110b = reserved 1111b = Index/Data pair in PCI Configuration space. This isn't supported in the PCH. RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 445 8.1.5.13 Offset B0h: FLRCID--FLR Capability ID (SATA-B0:D31:F2) Table 8-44. Offset B0h: FLRCID--FLR Capability ID (SATA-B0:D31:F2) Description: View: PCI BAR: Configuration Size: 16 bit Bus:Device:Function: B0:D31 :F2 Default: 0009h Bit Acronym Bit Range Offset Start: B0h Offset End: B1h Power Well: Bit Description Sticky Bit Reset Value Bit Access 15 :08 Next Capability Pointer -- 00h indicates the final item in the capability list. RO 07 :00 Capability ID -- The value of this field depends on the FLRCSSEL bit. 13h = If PFLRCSSEL = 0 09h (Vendor Specific) = If PFLRCSSEL = 1 RO 8.1.5.14 Offset B2h: FLRCLV--FLR Capability Length and Version (SATA- B0:D31:F2) When FLRCSSEL = `0', this register is defined as follows: Table 8-45. Offset B2h: FRLCLV--FLR Capability Length and Version (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range 15 :10 Bus:Device:Function: B0:D31 :F2 BAR: Configuration Default: xx06h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved 09 FLR Capability -- Support for Function Level reset. This bit is not reset by the Function Level Reset. RWO 08 TXP Capability -- Support for Transactions Pending (TXP) bit. TXP must be supported if FLR is supported. RWO 07 :00 Vendor-Specific Capability ID -- This field indicates the # of bytes of this Vendor Specific capability as required by the PCI specification. It has the value of 06h for the FLR capability. Intel(R) Communications Chipset 89xx Series - Datasheet 446 Offset Start: B2h Offset End: B3h RO October 2012 Order Number: 327879-001US 8.0 When FLRCSSEL = `1', this register is defined as follows: Table 8-46. Offset B2h: FLRCLV--FLR Capability Length and Version (SATA-B0:D31:F2) Description: View: PCI BAR: Configuration Size: 16 bit Bus:Device:Function: Offset Start: B2h Offset End: B3h B0:D31 :F2 Default: xx06h Bit Acronym Bit Range Power Well: Bit Description Sticky Bit Reset Value Bit Access 15 :12 Vendor-Specific Capability ID -- A value of 2h identifies this capability as the Function Level Reset (FLR). RO 11 :08 Capability Version -- This field indicates the version of the FLR capability. RO 07 :00 Vendor-Specific Capability ID -- This field indicates the # of bytes of this Vendor Specific capability as required by the PCI specification. It has the value of 06h for the FLR capability. RO 8.1.5.15 Offset B4h: FLRC--FLR Control (SATA-B0:D31:F2) Table 8-47. Offset B4h: FLRC--FLR Control (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range 15 :09 08 07 :01 BAR: Configuration Bus:Device:Function: B0:D31 :F2 Default: 0000h Bit Acronym Offset Start: B4h Offset End: B5h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved TXP Transactions Pending: 0 = Controller has received all non-posted requests. 1 = Controller has issued non-posted requests which has not been completed. RO Reserved Reserved 00 October 2012 Order Number: 327879-001US Initiate FLR -- Used to initiate FLR transition. A write of `1' indicates FLR transition. Since hardware must no t respond to any cycles till FLR completion the value read by software from this bit is `0'. RW Intel(R) Communications Chipset 89xx Series - Datasheet 447 8.1.5.16 Offset C0h: ATC--APM Trapping Control Register (SATA-B0:D31:F2) Table 8-48. Offset C0h: ATC--APM Trapping Control Register (SATA-B0:D31:F2) Description: View: PCI BAR: Configuration Size: 8 bit Bus:Device:Function: B0:D31 :F2 Default: 00h Bit Acronym Bit Range 07 :04 Offset Start: C0h Offset End: C0h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved SST Secondary Slave Trap -- Enables trapping and SMI# assertion on legacy I/O accesses to 170h177h and 376h. The active device on the secondary interface must be device 1 for the trap and/or SMI# to occur. RW 02 SPT Secondary Master Trap -- Enables trapping and SMI# assertion on legacy I/O accesses to 170h177h and 376h. The active device on the secondary interface must be device 0 for the trap and/or SMI# to occur. RW 01 PST Primary Slave Trap -- Enables trapping and SMI# assertion on legacy I/O accesses to 1F0h-1F7h and 3F6h. The active device on the primary interface must be device 1 for the trap and/or SMI# to occur. RW 00 PMT Primary Master Trap -- Enables trapping and SMI# assertion on legacy I/O accesses to 1F0h-1F7h and 3F6h. The active device on the primary interface must be device 0 for the trap and/or SMI# to occur. RW 03 8.1.5.17 Offset C4h: ATS--APM Trapping Status Register (SATA-B0:D31:F2) Table 8-49. Offset C4h: ATS--APM Trapping Status Register (SATA-B0:D31:F2) Description: View: PCI Size: 8 bit Bit Range 07 :04 Bus:Device:Function: B0:D31 :F2 BAR: Configuration Default: 00h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved 03 SST Secondary Slave Trap -- Indicates that a trap occurred to the secondary slave device. RWC 02 SPT Secondary Master Trap -- Indicates that a trap occurred to the secondary master device. RWC 01 PST Primary Slave Trap -- Indicates that a trap occurred to the primary slave device. RWC 00 PMT Primary Master Trap -- Indicates that a trap occurred to the primary master device. RWC Intel(R) Communications Chipset 89xx Series - Datasheet 448 Offset Start: C4h Offset End: C4h October 2012 Order Number: 327879-001US 8.0 8.1.5.18 Offset D0h: SP--Scratch Pad Register (SATA-B0:D31:F2) Table 8-50. Offset D0h: SP--Scratch Pad Register (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: Default: 00000000h Offset Start: D0h Offset End: D0h Power Well: Bit Range Bit Acronym Bit Description 31 :00 DT Data -- This is a read/write register that is available for software to use. No hardware action is taken on this register. October 2012 Order Number: 327879-001US B0:D31 :F2 Sticky Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 449 8.1.5.19 Offset E0h: BFCS--BIST FIS Control/Status Register (SATA- B0:D31:F2) Table 8-51. Offset E0h: BFCS--BIST FIS Control/Status Register (SATA-B0:D31:F2) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 :16 15 14 13 :12 11 10 BAR: Configuration Bus:Device:Function: Default: 00000000h Bit Acronym Offset Start: E0h Offset End: E3h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved P5BFI Port 5 BIST FIS Initiate -- When a rising edge is detected on this bit field, the PCH initiates a BIST FIS to the device on Port 5, using the parameters specified in this register and the data specified in BFTD1 and BFTD2. The BIST FIS will only be initiated if a device on Port 5 is present and ready (not partial/slumber state). After a BIST FIS is successfully completed, software must disable and re-enable the port using the PxE bits at offset 92h prior to attempting additional BIST FISs or to return the PCH to a normal operational mode. If the BIST FIS fails to complete, as indicated by the BFF bit in the register, then software can clear then set the P5BFI bit to initiate another BIST FIS. This can be retried until the BIST FIS eventually completes successfully RW P4BFI Port 4 BIST FIS Initiate -- When a rising edge is detected on this bit field, the PCH initiates a BIST FIS to the device on Port 4, using the parameters specified in this register and the data specified in BFTD1 and BFTD2. The BIST FIS will only be initiated if a device on Port 4 is present and ready (not partial/slumber state). After a BIST FIS is successfully completed, software must disable and re-enable the port using the PxE bits at offset 92h prior to attempting additional BIST FISs or to return the PCH to a normal operational mode. If the BIST FIS fails to complete, as indicated by the BFF bit in the register, then software can clear then set the P4BFI bit to initiate another BIST FIS. This can be retried until the BIST FIS eventually completes successfully RW Reserved Reserved BFS BIST FIS Successful: 0 = Software clears this bit by writing a 1 to it. 1 = This bit is set any time a BIST FIS transmitted by PCH receives an R_OK completion status from the device. Note: This bit must be cleared by software prior to initiating a BIST FIS. RWC BFF BIST FIS Failed: 0 = Software clears this bit by writing a 1 to it. 1 = This bit is set any time a BIST FIS transmitted by PCH receives an R_ERR completion status from the device. Note: This bit must be cleared by software prior to initiating a BIST FIS. RWC Intel(R) Communications Chipset 89xx Series - Datasheet 450 B0:D3 1:F2 October 2012 Order Number: 327879-001US 8.0 Table 8-51. Offset E0h: BFCS--BIST FIS Control/Status Register (SATA-B0:D31:F2) (Sheet 2 of 2) Description: View: PCI BAR: Configuration Size: 32 bit B0:D3 1:F2 Default: 00000000h Bit Acronym Bit Range 09 :08 Offset Start: E0h Offset End: E3h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved 07 :02 BFP 01 :00 8.1.5.20 Bus:Device:Function: BIST FIS Parameters -- These 6 bits form the contents of the upper 6 bits of the BIST FIS Pattern Definition in any BIST FIS transmitted by the PCH. This field is not port specific -- its contents will be used for any BIST FIS initiated on port 0, port 1, port 2 or port 3. The specific bit definitions are: Bit 7: T - Far End Transmit mode Bit 6: A - Align Bypass mode Bit 5: S - Bypass Scrambling Bit 4: L - Far End Retimed Loopback Bit 3: F - Far End Analog Loopback Bit 2: P - Primitive bit for use with Transmit mode RW Reserved Reserved Offset E4h: BFTD1--BIST FIS Transmit Data1 Register (SATA-B0:D31:F2) Table 8-52. Offset E4h: FBTD1--BIST FIS Transmit Data1 Register (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D3 1:F2 Default: 00000000h Bit Acronym 31 :00 October 2012 Order Number: 327879-001US Offset Start: E4h Offset End: E7h Power Well: Bit Description Sticky Bit Reset Value BIST FIS Transmit Data 1 -- The data programmed into this register will form the contents of the second dword of any BIST FIS initiated by the PCH. This register is not port specific -- its contents will be used for BIST FIS initiated on any port. Although the 2nd and 3rd DWs of the BIST FIS are only meaningful when the "T" bit of the BIST FIS is set to indicate "Far-End Transmit mode", this register's contents will be transmitted as the BIST FIS 2nd DW regardless of whether or not the "T" bit is indicated in the BFCS register (B0:D31:F2:E0h). Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 451 8.1.5.21 Offset E8h: BFTD2--BIST FIS Transmit Data2 Register (SATA- B0:D31:F2) Table 8-53. Offset E8h: BFTD2--BIST FIS Transmit Data2 Register (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: Configuration Bus:Device:Function: Default: 00000000h Bit Acronym Offset Start: E8h Offset End: EBh Power Well: Bit Description BIST FIS Transmit Data 2 -- The data programmed into this register will form the contents of the third dword of any BIST FIS initiated by the PCH. This register is not port specific -- its contents will be used for BIST FIS initiated on any port. Although the 2nd and 3rd DWs of the BIST FIS are only meaningful when the "T" bit of the BIST FIS is set to indicate "Far-End Transmit mode", this register's contents will be transmitted as the BIST FIS 3rd DW regardless of whether or not the "T" bit is indicated in the BFCS register (B0:D31:F2:E0h). Intel(R) Communications Chipset 89xx Series - Datasheet 452 B0:D3 1:F2 Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 8.0 8.1.6 SATA Indexed Registers The SATA Indexed Registers are the registers used with the Offset 0Ah: SATA Indexed Registers Index (SATA - B0:D31:F2) register. Table 8-54. SATA Indexed Registers Offset Start Offset End Register ID - Description Default Value 18h 1Bh "Offset 18h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 454 00000000h 1Ch 1Fh "Offset 1Ch: SATA Indexed Registers Index (SATA Test Mode Enable Register) (SATA-B0:D31:F2)" on page 454 00000000h 28h 2Bh "Offset 28h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 455 0B000040h 3Eh 3Fh "Offset 3Eh: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 455 04A4h 54h 57h "Offset 64h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 456 00AAAAAAh 64h 67h "Offset 64h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 456 84848484h 68h 69h "Offset 68h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 457 8484h 78h 7Bh "Offset 78h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 457 33220000h 84h 87h "Offset 84h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 458 00000000h 88h 8Bh "Offset 88h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 458 33332222h 8Ch 8Fh "Offset 8Ch: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 459 55555555h 94h 95h "Offset 94h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 459 5555h A0h A3h "Offset A0h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 460 00000000h A8h ABh "Offset A8h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 460 04040404h C4h C7h "Offset C4h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 461 04040404h C8h CBh "Offset C8h: SATA Indexed Registers Index (SATA Initialization Register) (SATA- B0:D31:F2)" on page 461 October 2012 Order Number: 327879-001US 04040404h Intel(R) Communications Chipset 89xx Series - Datasheet 453 8.1.6.1 Offset 18h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-55. Offset 18h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI BAR: Size: 32 bit B0:D3 1:F2 Default: 00000000h Bit Acronym Bit Range 31 :00 8.1.6.2 Bus:Device:Function: Offset Start: 18h Offset End: 1Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to 001C7000h Offset 1Ch: SATA Indexed Registers Index (SATA Test Mode Enable Register) (SATA-B0:D31:F2) Table 8-56. Offset 1Ch: SATA Indexed Registers Index (SATA Test Mode Enable Register) (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :19 18 B0:D3 Bus:Device:Function: 1:F2 BAR: Default: 00000000h Bit Acronym Offset Start: 1Ch Offset End: 1Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved SATA Test Mode Enable Bit: 0 = Entrance to the PCH SATA test modes are disabled. 1 = This bit allows entrance to the PCH SATA test modes when set. RW This bit should only be used for system board testing. 17 :00 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 454 October 2012 Order Number: 327879-001US 8.0 8.1.6.3 Offset 28h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-57. Offset 28h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI BAR: Size: 32 bit B0:D3 1:F2 Default: 0B000040h Bit Acronym Bit Range 31 :00 8.1.6.4 Bus:Device:Function: Offset Start: 28h Offset End: 2Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to 0A000033h. Offset 3Eh: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-58. Offset 3Eh: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range Bus:Device:Function: B0:D3 1:F2 BAR: Default: 04A4h Bit Acronym 15 :00 October 2012 Order Number: 327879-001US Offset Start: 3Eh Offset End: 3Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to 0464h. Intel(R) Communications Chipset 89xx Series - Datasheet 455 8.1.6.5 Offset 54h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-59. Offset 64h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI BAR: Size: 32 bit B0:D3 1:F2 Default: 00AAAAAAh Bit Acronym Bit Range 31 :24 Offset Start: 54h Offset End: 57h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved 23 :00 8.1.6.6 Bus:Device:Function: BIOS programs this field to 111111h. Offset 64h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-60. Offset 64h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :00 Bus:Device:Function: B0:D3 1:F2 BAR: Default: 84848484h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to CCCCCCCCh Intel(R) Communications Chipset 89xx Series - Datasheet 456 Offset Start: 64h Offset End: 67h October 2012 Order Number: 327879-001US 8.0 8.1.6.7 Offset 68h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-61. Offset 68h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI BAR: Size: 16 bit B0:D3 1:F2 Default: 8484h Bit Acronym Bit Range 15 :00 8.1.6.8 Bus:Device:Function: Offset Start: 68h Offset End: 69h Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to CCCCh Offset 78h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-62. Offset 78h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range Default: 33220000h Bit Acronym 31 :16 15 :00 Bus:Device:Function: B0:D3 1:F2 BAR: Offset Start: 78h Offset End: 7Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to 99AAh Reserved Reserved October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 457 8.1.6.9 Offset 84h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-63. Offset 84h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI BAR: Size: 32 bit B0:D3 1:F2 Default: 00000000h Bit Acronym Bit Range 31 :00 8.1.6.10 Bus:Device:Function: Offset Start: 84h Offset End: 87h Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to 001C7000h Offset 88h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-64. Offset 88h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :00 B0:D3 Bus:Device:Function: 1:F2 BAR: Default: 33332222h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to 9999AAAAh Intel(R) Communications Chipset 89xx Series - Datasheet 458 Offset Start: 88h Offset End: 8Bh October 2012 Order Number: 327879-001US 8.0 8.1.6.11 Offset 8Ch: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-65. Offset 8Ch: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI BAR: Size: 32 bit B0:D3 1:F2 Default: 55555555h Bit Acronym Bit Range 31 :00 8.1.6.12 Bus:Device:Function: Offset Start: 8Ch Offset End: 8Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to 55555555h Offset 94h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-66. Offset 94h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI Size: 16 bit Bit Range B0:D3 Bus:Device:Function: 1:F2 BAR: Default: 5555h Bit Acronym 15 :00 October 2012 Order Number: 327879-001US Offset Start: 94h Offset End: 95h Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to 5555h Intel(R) Communications Chipset 89xx Series - Datasheet 459 8.1.6.13 Offset A0h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-67. Offset A0h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI BAR: Size: 32 bit B0:D3 1:F2 Default: 00000000h Bit Acronym Bit Range 31 :00 8.1.6.14 Bus:Device:Function: Offset Start: A0h Offset End: A3h Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to 001C7000h Offset A8h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-68. Offset A8h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :00 B0:D3 Bus:Device:Function: 1:F2 BAR: Default: 04040404h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to 0C0C0C0Ch Intel(R) Communications Chipset 89xx Series - Datasheet 460 Offset Start: A8h Offset End: ABh October 2012 Order Number: 327879-001US 8.0 8.1.6.15 Offset C4h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-69. Offset C4h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI BAR: Size: 32 bit B0:D3 1:F2 Default: 04040404h Bit Acronym Bit Range 31 :00 8.1.6.16 Bus:Device:Function: Offset Start: C4h Offset End: C7h Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to 0C0C0C0Ch Offset C8h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Table 8-70. Offset C8h: SATA Indexed Registers Index (SATA Initialization Register) (SATA-B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range Bus:Device:Function: B0:D3 1:F2 BAR: Default: 04040404h Bit Acronym 31 :00 October 2012 Order Number: 327879-001US Offset Start: C8h Offset End: CBh Power Well: Bit Description Sticky Bit Reset Value Bit Access BIOS programs this field to 0C0C0C0Ch Intel(R) Communications Chipset 89xx Series - Datasheet 461 8.2 Bus Master IDE I/O Registers (B0:D31:F2) The bus master IDE function uses 16 bytes of I/O space, allocated via the LBAR register, located in Device 31:Function 2 Configuration space, offset 20h. All bus master IDE I/O space registers can be accessed as byte, word, or dword quantities. Reading reserved bits returns an indeterminate, inconsistent value, and writes to reserved bits have no affect (but should not be attempted). These registers are only used for legacy operation. Software must not use these registers when running AHCI. All I/O registers are reset by Function Level Reset. The description of the I/O registers is shown in the following table. 8.2.1 Bus Master IDE I/O Register Address Map Table 8-71. Bus Master IDE I/O Register Address Map Offset Start Offset End Default Value Register ID - Description 00h 00h "Offset 00h: Bus Master IDE Command Register Primary (B0:D31:F2)" on page 463 00h 02h 02h "Offset 02h: Bus Master IDE Status Register Primary (B0:D31:F2)" on page 464 00h 04h 07h "Offset 04h: Bus Master IDE Descriptor Table Pointer Register Primary (B0:D31:F2)" on page 465 00h 08h 08h "Offset 08h: Bus Master IDE Command Register Secondary (B0:D31:F2)" on page 466 00h 0Ah 0Ah "Offset 0Ah: Bus Master IDE Status Register Secondary (B0:D31:F2)" on page 467 00h 0Fh "Offset 0Ch: Bus Master IDE Descriptor Table Pointer Register Secondary (B0:D31:F2)" on page 468 0Ch 00h 10h 10h "Offset 10h: AHCI Index Register (B0:D31:F2)" on page 468 00000000h 14h 14h "Offset 14h: AHCI Index Data Register (B0:D31:F2)" on page 469 00h Intel(R) Communications Chipset 89xx Series - Datasheet 462 October 2012 Order Number: 327879-001US 8.0 8.2.1.1 Offset 00h: Bus Master IDE Command Register Primary (B0:D31:F2) Table 8-72. Offset 00h: Bus Master IDE Command Register Primary (B0:D31:F2) Description: View: PCI Size: 8 bit Bit Range 07 :04 03 02 :01 00 BAR: LBAR (IO) Bus:Device:Function: Default: 00h Bit Acronym B0:D3 1:F2 Offset Start: 00h Offset End: 00h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved. Returns 0. R/WC Read / Write Control -- This bit sets the direction of the bus master transfer: This bit must NOT be changed when the bus master function is active. 0 = Memory reads 1 = Memory writes RW Reserved Reserved. Returns 0. START Start/Stop Bus Master: 0 = All state information is lost when this bit is cleared. Master mode operation cannot be stopped and then resumed. If this bit is reset while bus master operation is still active (i.e., the Bus Master IDE Active bit (B0:D31:F2:BAR + 02h, bit 0) of the Bus Master IDE Status register for that IDE channel is set) and the drive has not yet finished its data transfer (the Interrupt bit in the Bus Master IDE Status register for that IDE channel is not set), the bus master command is said to be aborted and data transferred from the drive may be discarded instead of being written to system memory. 1 = Enables bus master operation of the controller. Bus master operation does not actually start unless the Bus Master Enable bit (D31:F1:04h, bit 2) in PCI configuration space is also set. Bus master operation begins when this bit is detected changing from 0 to 1. The controller will transfer data between the IDE device and memory only when this bit is set. Master operation can be halted by writing a 0 to this bit. Note: This bit is intended to be cleared by software after the data transfer is completed, as indicated by either the Bus Master IDE Active bit being cleared or the Interrupt bit of the Bus Master IDE Status register for that IDE channel being set, or both. Hardware does not clear this bit automatically. If this bit is cleared to 0 prior to the DMA data transfer being initiated by the drive in a device to memory data transfer, then the PCH will not send DMAT to terminate the data transfer. SW intervention (e.g. sending SRST) is required to reset the interface in this condition. October 2012 Order Number: 327879-001US RW Intel(R) Communications Chipset 89xx Series - Datasheet 463 8.2.1.2 Offset 02h: Bus Master IDE Status Register Primary (B0:D31:F2) Table 8-73. Offset 02h: Bus Master IDE Status Register Primary (B0:D31:F2) Description: View: PCI Size: 8 bit Bit Range 07 BAR: LBAR (IO) Bus:Device:Function: Default: 00h Bit Acronym PRDIS Offset Start: 02h Offset End: 02h Power Well: Bit Description PRD Interrupt Status: 0 = Software clears this bit by writing a 1 to it. 1 = This bit is set when the host controller execution of a PRD that has its PRD_INT bit set. Sticky Bit Reset Value Bit Access RWC 06 Drive 1 DMA Capable: 0 = Not Capable. 1 = Capable. Set by device dependent code (BIOS or device driver) to indicate that drive 1 for this channel is capable of DMA transfers, and that the controller has been initialized for optimum performance. The PCH does not use this bit. It is intended for systems that do not attach BMIDE to the PCI bus. RW 05 Drive 0 DMA Capable: 0 = Not Capable 1 = Capable. Set by device dependent code (BIOS or device driver) to indicate that drive 0 for this channel is capable of DMA transfers, and that the controller has been initialized for optimum performance. The PCH does not use this bit. It is intended for systems that do not attach BMIDE to the PCI bus. RW 04 :03 Reserved Reserved. Returns 0. 02 Interrupt: 0 = Software clears this bit by writing a 1 to it. 1 = Set when a device FIS is received with the `I' bit set, provided that software has not disabled interrupts via the IEN bit of the Device Control Register (see chapter 5 of the Serial ATA Specification, Revision 1.0a). RWC 01 Error: 0 = Software clears this bit by writing a 1 to it. 1 = This bit is set when the controller encounters a target abort or master abort when transferring data on PCI. RWC 00 Bus Master IDE Active: 0 = This bit is cleared by the PCH when the last transfer for a region is performed, where EOT for that region is set in the region descriptor. It is also cleared by the PCH when the Start Bus Master bit (B0:D31:F2:BAR+ 00h, bit 0) is cleared in the Command register. When this bit is read as a 0, all data transferred from the drive during the previous bus master command is visible in system memory, unless the bus master command was aborted. 1 = Set by the PCH when the Start bit is written to the Command register. RO ACT Intel(R) Communications Chipset 89xx Series - Datasheet 464 B0:D3 1:F2 October 2012 Order Number: 327879-001US 8.0 8.2.1.3 Offset 04h: Bus Master IDE Descriptor Table Pointer Register Primary (B0:D31:F2) Table 8-74. Offset 04h: Bus Master IDE Descriptor Table Pointer Register Primary (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :02 01 :00 BAR: LBAR (IO) Bus:Device:Function: B0:D3 1:F2 Default: 00h Offset Start: 04h Offset End: 07h Power Well: Bit Acronym Bit Description ADDR Address of Descriptor Table: The bits in this field correspond to bits [31:2] of the memory location of the Physical Region Descriptor (PRD). The Descriptor Table must be Dword-aligned. The Descriptor Table must not cross a 64-K boundary in memory. Sticky Bit Reset Value Bit Access RW Reserved Reserved October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 465 8.2.1.4 Offset 08h: Bus Master IDE Command Register Secondary (B0:D31:F2) Table 8-75. Offset 08h: Bus Master IDE Command Register Secondary (B0:D31:F2) Description: View: PCI Size: 8 bit Bit Range 07 :04 03 02 :01 00 BAR: LBAR (IO) Bus:Device:Function: Default: 00h Bit Acronym Offset Start: 08h Offset End: 08h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved. Returns 0. R/WC Read / Write Control -- This bit sets the direction of the bus master transfer: This bit must NOT be changed when the bus master function is active. 0 = Memory reads 1 = Memory writes RW Reserved Reserved. Returns 0. START Start/Stop Bus Master: 0 = All state information is lost when this bit is cleared. Master mode operation cannot be stopped and then resumed. If this bit is reset while bus master operation is still active (i.e., the Bus Master IDE Active bit (B0:D31:F2:BAR + 02h, bit 0) of the Bus Master IDE Status register for that IDE channel is set) and the drive has not yet finished its data transfer (the Interrupt bit in the Bus Master IDE Status register for that IDE channel is not set), the bus master command is said to be aborted and data transferred from the drive may be discarded instead of being written to system memory. 1 = Enables bus master operation of the controller. Bus master operation does not actually start unless the Bus Master Enable bit (D31:F1:04h, bit 2) in PCI configuration space is also set. Bus master operation begins when this bit is detected changing from 0 to 1. The controller will transfer data between the IDE device and memory only when this bit is set. Master operation can be halted by writing a 0 to this bit. Note: This bit is intended to be cleared by software after the data transfer is completed, as indicated by either the Bus Master IDE Active bit being cleared or the Interrupt bit of the Bus Master IDE Status register for that IDE channel being set, or both. Hardware does not clear this bit automatically. If this bit is cleared to 0 prior to the DMA data transfer being initiated by the drive in a device to memory data transfer, then the PCH will not send DMAT to terminate the data transfer. SW intervention (e.g. sending SRST) is required to reset the interface in this condition. Intel(R) Communications Chipset 89xx Series - Datasheet 466 B0:D3 1:F2 RW October 2012 Order Number: 327879-001US 8.0 8.2.1.5 Offset 0Ah: Bus Master IDE Status Register Secondary (B0:D31:F2) Table 8-76. Offset 0Ah: Bus Master IDE Status Register Secondary (B0:D31:F2) Description: View: PCI Size: 8 bit Bit Range 07 BAR: LBAR (IO) Bus:Device:Function: Default: 00h Bit Acronym PRDIS B0:D3 1:F2 Offset Start: 0Ah Offset End: 0Ah Power Well: Bit Description Sticky Bit Reset Value PRD Interrupt Status: 0 = Software clears this bit by writing a 1 to it. 1 = This bit is set when the host controller execution of a PRD that has its PRD_INT bit set. Bit Access RWC 06 Drive 1 DMA Capable: 0 = Not Capable. 1 = Capable. Set by device dependent code (BIOS or device driver) to indicate that drive 1 for this channel is capable of DMA transfers, and that the controller has been initialized for optimum performance. The PCH does not use this bit. It is intended for systems that do not attach BMIDE to the PCI bus. RW 05 Drive 0 DMA Capable: 0 = Not Capable 1 = Capable. Set by device dependent code (BIOS or device driver) to indicate that drive 0 for this channel is capable of DMA transfers, and that the controller has been initialized for optimum performance. The PCH does not use this bit. It is intended for systems that do not attach BMIDE to the PCI bus. RW 04 :03 Reserved Reserved. Returns 0. 02 Interrupt: 0 = Software clears this bit by writing a 1 to it. 1 = Set when a device FIS is received with the `I' bit set, provided that software has not disabled interrupts via the IEN bit of the Device Control Register (see chapter 5 of the Serial ATA Specification, Revision 1.0a). RWC 01 Error: 0 = Software clears this bit by writing a 1 to it. 1 = This bit is set when the controller encounters a target abort or master abort when transferring data on PCI. RWC 00 Bus Master IDE Active: 0 = This bit is cleared by the PCH when the last transfer for a region is performed, where EOT for that region is set in the region descriptor. It is also cleared by the PCH when the Start Bus Master bit (B0:D31:F2:BAR+ 00h, bit 0) is cleared in the Command register. When this bit is read as a 0, all data transferred from the drive during the previous bus master command is visible in system memory, unless the bus master command was aborted. 1 = Set by the PCH when the Start bit is written to the Command register. RO ACT October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 467 8.2.1.6 Offset 0Ch: Bus Master IDE Descriptor Table Pointer Register Secondary (B0:D31:F2) Table 8-77. Offset 0Ch: Bus Master IDE Descriptor Table Pointer Register Secondary (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :02 01 :00 8.2.1.7 BAR: LBAR (IO) Bus:Device:Function: Offset Start: 0Ch Offset End: 0Fh B0:D3 1:F2 Default: 00h Power Well: Bit Acronym Bit Description ADDR Address of Descriptor Table: The bits in this field correspond to bits [31:2] of the memory location of the Physical Region Descriptor (PRD). The Descriptor Table must be Dword-aligned. The Descriptor Table must not cross a 64-K boundary in memory. Sticky Bit Reset Value Bit Access RW Reserved Reserved Offset 10h: AHCI Index Register (B0:D31:F2) Table 8-78. Offset 10h: AHCI Index Register (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :11 10 :02 01 :00 Bus:Device:Function: B0:D3 1:F2 BAR: LBAR (IO) Default: 00000000h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved INDEX Index -- This Index register is used to select the Dword offset of the Memory Mapped AHCI register to be accessed. A Dword, Word or Byte access is specified by the active byte enables of the I/O access to the Data register. RW Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 468 Offset Start: 10h Offset End: 10h October 2012 Order Number: 327879-001US 8.0 8.2.1.8 Offset 14h: AHCI Index Data Register (B0:D31:F2) This register is available only when SCC is not 01h. Table 8-79. Offset 14h: AHCI Index Data Register (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: LBAR (IO) Bus:Device:Function: Default: 00h Offset Start: 14h Offset End: 14h Power Well: Bit Acronym Bit Description DTA Data -- This Data register is a "window" through which data is read or written to the AHCI memory mapped registers. A read or write to this Data register triggers a corresponding read or write to the memory mapped register pointed to by the Index register. The Index register must be setup prior to the read or write to this Data register. Note: A physical register is not actually implemented as the data is actually stored in the memory mapped registers. Since this is not a physical register, the "default" value is the same as the default value of the register pointed to by Index. October 2012 Order Number: 327879-001US B0:D3 1:F2 Sticky Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 469 8.3 Serial ATA Index/Data Pair Superset Registers All of these I/O registers are in the core well. They are exposed only when SCC is 01h (i.e. IDE programming interface). These are Index/Data Pair registers that are used to access the SerialATA superset registers (SerialATA Status, SerialATA Control and SerialATA Error). The I/O space for these registers is allocated through SIDPBA. Locations with offset from 08h to 0Fh are reserved for future expansion. Software-write operations to the reserved locations will have no effect while software-read operations to the reserved locations will return 0. 8.3.1 Superset Registers Table 8-80. Superset Registers Offset Start Offset End Default Value Register ID - Description 00h 03h "Offset SIDPBA + 00h: Serial ATA Index (B0:D31:F2)" on page 470 00000000h 04h 07h "Offset SIDPBA + 04h: Serial ATA Data (B0:D31:F2)" on page 471 00000000h 8.3.1.1 Offset SIDPBA + 00h: Serial ATA Index (B0:D31:F2) Table 8-81. Offset SIDPBA + 00h: Serial ATA Index (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :16 15 :08 07 :00 Bus:Device:Function: B0:D3 1:F2 BAR: SIDPBA Default: 00000000h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved PIDX Port Index -- This Index field is used to specify the port of the SATA controller at which the portspecific SSTS, SCTL, and SERR registers are located. 00h = Primary Master (Port 0) 01h = Primary Slave (Port 2) 02h = Secondary Master (Port 1) 03h = Secondary Slave (Port 3) All other values are Reserved. RW RIDX Register Index -- This index field is used to specify one out of three registers currently being indexed into. These three registers are the Serial ATA superset SStatus, SControl and SError memory registers and are port specific, hence for this SATA controller, there are four sets of these registers. 00h = SSTS 01h = SCTL 02h = SERR All other values are Reserved. RW Intel(R) Communications Chipset 89xx Series - Datasheet 470 Offset Start: 00h Offset End: 03h October 2012 Order Number: 327879-001US 8.0 8.3.1.2 Offset 04h: Serial ATA Data (B0:D31:F2) Table 8-82. Offset SIDPBA + 04h: Serial ATA Data (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: SIDPBA Bus:Device:Function: B0:D3 1:F2 Default: 00000000h Offset Start: 04h Offset End: 07h Power Well: Bit Acronym Bit Description DTA Data -- This Data register is a "window" through which data is read or written to from the register pointed to by the Serial ATA Index (SINDX) register above. A physical register is not implemented; the data is stored in the memory mapped registers. Since this is not a physical register, the "default" value is the same as the default value of the register pointed to by SINDX.RIDX field. Sticky Bit Reset Value Bit Access RW 8.4 AHCI Registers (B0:D31:F2) Note: These registers are AHCI-specific and available when the PCH is properly configured. The Serial ATA Status, Control, and Error registers are special exceptions and may be accessed on all PCH components if properly configured. The memory mapped registers within the SATA controller exist in non-cacheable memory space. Additionally, locked accesses are not supported. If software attempts to perform locked transactions to the registers, indeterminate results may occur. Register accesses has a maximum size of 64-bits; 64-bit access must not cross an 8-byte alignment boundary. All memory registers are reset by Function Level Reset unless specified otherwise. The registers are broken into two sections - generic host control and port control. The port control registers are the same for all ports, and there are as many registers banks as there are ports. Table 8-83. AHCI Register Address Map ABAR + Offset Mnemonic October 2012 Order Number: 327879-001US Register 00-1Fh GHC 20h-FFh -- Generic Host Control 300h-37Fh P4PCR Port 4 port control registers 380h-3FFh P5PCR Port 5 port control registers Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 471 8.4.1 AHCI Generic Host Control Registers (B0:D31:F2) Table 8-84. AHCI Generic Host Control Registers Offset Start Offset End Default Value Register ID - Description 00h 03h "Offset 00h: Host Capabilities Register (B0:D31:F2)" on page 472 DE127F03h 04h 07h "Offset 04h: Global PCH Control Register (B0:D31:F2)" on page 477 00000000h 08h 0Bh "Offset 08h: Interrupt Status Register (B0:D31:F2)" on page 478 00000000h OCh OFh "Offset 0Ch: Ports Implemented Register (B0:D31:F2)" on page 479 00000000h 10h 13h "Offset 10h: Serial AHCI Version (B0:D31:F2)" on page 479 00010200h 14h 17h "Offset 14h: Serial Command Completion Coalescing Control Register (B0:D31:F2)" on page 480 00000000h 18h 1Bh "Offset 18h: Serial Command Completion Coalescing Ports Register (B0:D31:F2)" on page 481 00000000h 1Ch 1Fh "Offset 1Ch: Serial Enclosure Management Location Register (B0:D31:F2)" on page 481 01600002h 20h 23h "Offset 20h: Serial Enclosure Management Control Register (B0:D31:F2)" on page 482 07010000h 24h 24h "Offset 24h: Serial Extended Host Capabilities (B0:D31:F2)" on page 483 0000000Xh 6Ch 6Ch "Offset SIDPBA + 04h: Serial ATA Index (B0:D31:F2)" on page 484 00000001h 70h 73h "Offset 70h: Serial AHCI Version (B0:D31:F2)" on page 484 00010000h A0h A3h "Offset A0h: Serial Vendor Specific (B0:D31:F2)" on page 485 00000000h 8.4.1.1 Offset 00h: Host Capabilities Register (B0:D31:F2) All bits in this register that are RWO are reset only by PLTRST#. Table 8-85. Offset 00h: Host Capabilities Register (B0:D31:F2) (Sheet 1 of 3) Description: View: PCI Size: 32 bit Bus:Device:Function: B0:D3 1:F2 BAR: ABAR Default: DE127F03h Power Well: Bit Acronym Bit Description S64A Supports 64-bit Addressing -- Indicates that the SATA controller can access 64-bit data structures. The 32-bit upper bits of the port DMA Descriptor, the PRD Base, and each PRD entry are read/write. RO 30 SCQA Supports Command Queue Acceleration -- Hardwired to 1 to indicate that the SATA controller supports SATA command queuing via the DMA Setup FIS. The PCH handles DMA Setup FISes natively, and can handle auto-activate optimization through that FIS. RO 29 SSNTF Supports SNotification Register -- The PCH SATA Controller does not support the SNotification register. RO Bit Range 31 Intel(R) Communications Chipset 89xx Series - Datasheet 472 Offset Start: 00h Offset End: 03h Sticky Bit Reset Value Bit Access October 2012 Order Number: 327879-001US 8.0 Table 8-85. Offset 00h: Host Capabilities Register (B0:D31:F2) (Sheet 2 of 3) Description: View: PCI Size: 32 bit BAR: ABAR Bus:Device:Function: Default: DE127F03h B0:D3 1:F2 Offset Start: 00h Offset End: 03h Power Well: Bit Acronym Bit Description SIS Supports Interlock Switch -- Indicates whether the SATA controller supports interlock switches on its ports for use in Hot Plug operations. This value is loaded by platform BIOS prior to OS initialization. If this bit is set, BIOS must also map the SATAGP pins to the SATA controller through GPIO space. RWO SSS Supports Staggered Spin-up -- Indicates whether the SATA controller supports staggered spin-up on its ports, for use in balancing power spikes. This value is loaded by platform BIOS prior to OS initialization. 0 = Staggered spin-up not supported. 1 = Staggered spin-up supported. RWO 26 SALP Supports Aggressive Link Power Management: 0 = Software shall treat the PxCMD.ALPE and PxCMD.ASP bits as reserved. 1 = The SATA controller supports auto-generating link requests to the partial or slumber states when there are no commands to process. RWO 25 SAL Supports Activity LED -- Indicates that the SATA controller supports a single output pin (SATALED#) which indicates activity. RO SCLO Supports Command List Override -- When set to '1', indicates that the Controller supports the PxCMD.CLO bit and it's associated function. When cleared to '0', The Controller is not capable of clearing the BSY and DRQ bits in the Status register in order to issue a software reset if these bits are still set from a previous operation. RWO ISS Interface Speed Support -- Indicates the maximum speed the SATA controller can support on its ports. 2h =3.0 Gb/s. RWO Bit Range 28 27 24 23 :20 19 18 17 :16 Sticky Bit Reset Value Bit Access SNZO Supports Non-Zero DMA Offsets -- Reserved, as per the AHCI Revision 1.2 specification RO SAM Supports AHCI Mode Only -- The SATA controller may optionally support AHCI access mechanism only. 0 = SATA controller supports both IDE and AHCI Modes 1 = SATA controller supports AHCI Mode Only RO Reserved Reserved 15 PMD PIO Multiple DRQ Block -- The SATA controller supports PIO Multiple DRQ Command Block RO 14 SSC Slumber State Capable -- The SATA controller supports the slumber state. RWO 13 PSC Partial State Capable -- The SATA controller supports the partial state. RWO NCS Number of Command Slots -- Hardwired to 1Fh to indicate support for 32 slots. 12 :08 07 CCCS October 2012 Order Number: 327879-001US Command Completion Coalescing Supported: 0 = Command Completion Coalescing Not Supported 1 = Command Completion Coalescing Supported RO RWO Intel(R) Communications Chipset 89xx Series - Datasheet 473 Table 8-85. Offset 00h: Host Capabilities Register (B0:D31:F2) (Sheet 3 of 3) Description: View: PCI Size: 32 bit Bit Range 06 05 04 :00 BAR: ABAR Bus:Device:Function: Default: DE127F03h Bit Acronym Offset Start: 00h Offset End: 03h Power Well: Bit Description Sticky Bit Reset Value Bit Access EMS Enclosure Management Supported: 0 = Enclosure Management Not Supported 1 = Enclosure Management Supported RWO SXS Supports External SATA: 0 = External SATA is not supported on any ports 1 = External SATA is supported on one or more ports When set, SW can examine each SATA port's Command Register (PxCMD) to determine which port is routed externally. RWO NPS Number of Ports -- Indicates number of supported ports. The number of ports indicated in this field may be more than the number of ports indicated in the PI (ABAR + 0Ch) register. Intel(R) Communications Chipset 89xx Series - Datasheet 474 B0:D3 1:F2 RO October 2012 Order Number: 327879-001US 8.0 8.4.1.2 Offset 04h: Global PCH Control Register (B0:D31:F2) Table 8-86. Offset 04h: Global PCH Control Register (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 30 :03 02 BAR: ABAR Bus:Device:Function: Default: 00000000h B0:D3 1:F2 Offset Start: 04h Offset End: 07h Power Well: Bit Acronym Bit Description AE AHCI Enable: When set, indicates that an AHCI driver is loaded and the controller will be talked to via AHCI mechanisms. This can be used by a PCH that supports both legacy mechanisms (such as SFF-8038i) and AHCI to know when the controller will not be talked to as legacy. 0 = Software will communicate with the PCH using legacy mechanisms. 1 = Software will communicate with the PCH using AHCI. The PCH will not have to allow command processing via both AHCI and legacy mechanisms. Software shall set this bit to 1 before accessing other AHCI registers. Sticky Bit Reset Value Bit Access RW Reserved Reserved MRSM MSI Revert to Single Message: When set to '1' by hardware, indicates that the host controller requested more than one MSI vector but has reverted to using the first vector only. When this bit is cleared to '0', the Controller has not reverted to single MSI mode (i.e. hardware is already in single MSI mode, software has allocated the number of messages requested, or hardware is sharing interrupt vectors if MC.MME < MC.MMC). "MC.MSIE = '1' (MSI is enabled) "MC.MMC > 0 (multiple messages requested) "MC.MME > 0 (more than one message allocated) "MC.MME!= MC.MMC (messages allocated not equal to number requested) RO When this bit is set to '1', single MSI mode operation is in use and software is responsible for clearing bits in the IS register to clear interrupts. This bit shall be cleared to '0' by hardware when any of the four conditions stated is false. This bit is also cleared to '0' when MC.MSIE = '1' and MC.MME = 0h. In this case, the hardware has been programmed to use single MSI mode, and is not "reverting" to that mode. For PCH, the Controller shall always revert to single MSI mode when the number of vectors allocated by the host is less than the number requested. This bit is ignored when GHC.HR = 1. 01 00 IE Interrupt Enable: This global bit enables interrupts from the PCH. 0 = All interrupt sources from all ports are disabled. 1 = Interrupts are allowed from the AHCI controller. RW HR Controller Reset: Resets the PCH AHCI controller. 0 = No effect 1 = When set by SW, this bit causes an internal reset of the PCH AHCI controller. All state machines that relate to data transfers and queuing return to an idle condition, and all ports are re-initialized via COMRESET. For further details, consult section 12.3.3 of the Serial ATA Advanced Host Controller Interface specification. RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 475 8.4.1.3 Offset 08h: Interrupt Status Register (B0:D31:F2) This register indicates which of the ports within the controller have an interrupt pending and require service. Table 8-87. Offset 08h: Interrupt Status Register (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :09 08 07 06 05 04 03 :00 BAR: ABAR Bus:Device:Function: Default: 00000000h Bit Acronym Offset Start: 08h Offset End: 0Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved. Returns 0. CIPS Coalescing Interrupt Pending Status: 0 = No interrupt pending. 1 = A command completion coalescing interrupt has been generated. RWC Reserved Reserved IPS[5] Interrupt Pending Status Port[5]: 0 = No interrupt pending. 1 = Port 5 has an interrupt pending. Software can use this information to determine which ports require service after an interrupt. RWC IPS[4] Interrupt Pending Status Port[4]: 0 = No interrupt pending. 1 = Port 4 has an interrupt pending. Software can use this information to determine which ports require service after an interrupt. RWC Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 476 B0:D31 :F2 October 2012 Order Number: 327879-001US 8.0 8.4.1.4 Offset 0Ch: Ports Implemented Register (B0:D31:F2) This register indicates which ports are exposed to the PCH. It is loaded by platform BIOS. It indicates which ports that the device supports are available for software to use. For ports that are not available, software must not read or write to registers within that port. Table 8-88. Offset 0Ch: Ports Implemented Register (B0:D31:F2) Description: View: PCI BAR: ABAR Size: 32 bit B0:D31 :F2 Default: 00000000h Bit Acronym Bit Range 31 :06 Offset Start: OCh Offset End: OFh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved 05 04 03 :00 8.4.1.5 Bus:Device:Function: PI5 Ports Implemented Port 5: 0 = The port is not implemented. 1 = The port is implemented. This bit is read-only `0' if MAP.SC = `0' or SCC = `01h'. RWO PI4 Ports Implemented Port 4: 0 = The port is not implemented. 1 = The port is implemented. This bit is read-only `0' if MAP.SC = `0' or SCC = `01h'. RWO Reserved Reserved Offset 10h: Serial AHCI Version (B0:D31:F2) This register indicates the major and minor version of the AHCI specification. It is BCD encoded. The upper two bytes represent the major version number, and the lower two bytes represent the minor version number. Example: Version 3.12 would be represented as 00030102h. The current version of the specification is 1.20 (00010200h). Table 8-89. Offset 10h: Serial AHCI Version (B0:D31:F2) Description: View: PCI Size: 32 bit BAR: ABAR Bus:Device:Function: B0:D31 :F2 Default: 00010200h Offset Start: 10h Offset End: 13h Power Well: Bit Range Bit Acronym 31 :16 MJR Major Version Number -- Indicates the major version is 1 RO 15 :00 MNR Minor Version Number -- Indicates the minor version is 20. RO October 2012 Order Number: 327879-001US Bit Description Sticky Bit Reset Value Bit Access Intel(R) Communications Chipset 89xx Series - Datasheet 477 8.4.1.6 Offset 14h: Serial Command Completion Coalescing Control Register (B0:D31:F2) This register is used to configure the command coalescing feature. This register is reserved if command coalescing is not supported (CAP_CCCS = `0'). Table 8-90. Offset 14h: Serial Command Completion Coalescing Control Register (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :16 15 :08 07 :03 BAR: ABAR Bus:Device:Function: Default: 00000000h B0:D31 :F2 Offset Start: 14h Offset End: 17h Power Well: Bit Acronym Bit Description TV Timeout Value -- The timeout value is specified in 10 microsecond intervals. hbaCCC_Timer is loaded with this timeout value. hbaCCC_Timer is only decremented when commands are outstanding on the selected ports. The Controller will signal a CCC interrupt when hbaCCC_Timer has decremented to `0'. The hbaCCC_Timer is reset to the timeout value on the assertion of each CCC interrupt. A timeout value of 0 is invalid. RW CC Command Completions -- Specifies the number of command completions that are necessary to cause a CCC interrupt. The Controller has an internal command completion counter, hbaCCC_CommandsComplete. hbaCCC_CommandsComplete is incremented by one each time a selected port has a command completion. When hbaCCC_CommandsComplete is equal to the command completions value, a CCC interrupt is signaled. The internal command completion counter is reset to `0' on the assertion of each CCC interrupt. RW INT Interrupt -- Specifies the interrupt used by the CCC feature. This interrupt must be marked as unused in the AHCI Ports Implemented memory register by the corresponding bit being set to `0'. Thus, the CCC_interrupt corresponds to the interrupt for an unimplemented port on the controller. When a CCC interrupt occurs, the IS[INT] bit shall be asserted to `1' regardless of whether PIRQ interrupt or MSI is used. Sticky Bit Reset Value Bit Access RO In MSI, CC interrupt may share an interrupt vector with other ports. For example, if the number of message allocated is 4, then CCC interrupt share interrupt vector 3 along with port 3, 4, and 5 but IS[6] shall get set. 02 :01 00 Reserved Reserved EN Enable: 0 = The command completion coalescing feature is disabled and no CCC interrupts are generated 1 = The command completion coalescing feature is enabled and CCC interrupts may be generated based on timeout or command completion conditions. Software shall only change the contents of the TV and CC fields when EN is cleared to '0'. On transition of this bit from '0' to '1', any updated values for the TV and CC fields shall take effect. Intel(R) Communications Chipset 89xx Series - Datasheet 478 RW October 2012 Order Number: 327879-001US 8.0 8.4.1.7 Offset 18h: Serial Command Completion Coalescing Ports Register (B0:D31:F2) This register is used to specify the ports that are coalesced as part of the CCC feature when CCC_CTL.EN = `1'. This register is reserved if command coalescing is not supported (CAP_CCCS = `0'). Table 8-91. Offset 18h: Serial Command Completion Coalescing Ports Register (B0:D31:F2) Description: View: PCI BAR: ABAR Size: 32 bit B0:D3 1:F2 Default: 00000000h Bit Range 31 :00 8.4.1.8 Bus:Device:Function: Offset Start: 18h Offset End: 1Bh Power Well: Bit Acronym Bit Description PRT Ports: 0 = The port is not part of the command completion coalescing feature. 1 = The corresponding port is part of the command completion coalescing feature. Bits set to `1' in this register must also have the corresponding bit set to `1' in the Ports Implemented register. Bits set to '1' in this register must also have the corresponding bit set to '1' in the Ports Implemented register. An updated value for this field shall take effect within one timer increment (1 millisecond). Sticky Bit Reset Value Bit Access RW Offset 1Ch: Serial Enclosure Management Location Register (B0:D31:F2) This register identifies the location and size of the enclosure management message buffer. This register is reserved if enclosure management is not supported (i.e. CAP.EMS = 0). Table 8-92. Offset 1Ch: Serial Enclosure Management Location Register (B0:D31:F2) Description: View: PCI Size: 32 bit BAR: ABAR B0:D3 Bus:Device:Function: 1:F2 Default: 01600002h Bit Range Bit Acronym 31 :16 OFST 15 :00 SZ October 2012 Order Number: 327879-001US Offset Start: 1Ch Offset End: 1Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Offset -- The offset of the message buffer in Dwords from the beginning of the ABAR. RO Buffer Size -- Specifies the size of the transmit message buffer area in Dwords. The PCH SATA controller only supports transmit buffer. A value of `0' is invalid. RO Intel(R) Communications Chipset 89xx Series - Datasheet 479 8.4.1.9 Offset 20h: Serial Enclosure Management Control Register (B0:D31:F2) This register is used to control and obtain status for the enclosure management interface. This register includes information on the attributes of the implementation, enclosure management messages supported, the status of the interface, whether any message are pending, and is used to initiate sending messages. This register is reserved if enclosure management is not supported (CAP_EMS = `0'). Table 8-93. Offset 20h: Serial Enclosure Management Control Register (B0:D31:F2) (Sheet 1 of 2) Description: View: PCI Size: 32 bit B0:D3 Bus:Device:Function: 1:F2 BAR: ABAR Default: 07010000h Bit Range Bit Acronym 31 :27 Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 26 Activity LED Hardware Driven: 1 = The SATA controller drives the activity LED for the LED message type in hardware and does not ATTR.ALHD utilize software for this LED. The host controller does not begin transmitting the hardware based activity signal until after software has written CTL.TM=1 after a reset condition. 25 ATTR.XMT Transmit Only: 0 = The SATA controller supports transmitting and receiving messages. 1 = The SATA controller only supports transmitting messages and does not support receiving messages. RO ATTR.SMB Single Message Buffer: 0 = There are separate receive and transmit buffers such that unsolicited messages could be supported. 1 = The SATA controller has one message buffer that is shared for messages to transmit and messages received. Unsolicited receive messages are not supported and it is software's responsibility to manage access to this buffer. RO Reserved Reserved 24 23 :20 19 RO SES-2 Enclosure Management Messages: 1 = The SATA controller supports the SES-2 message type. RO SUPP.SES2 17 SAF-TE Enclosure Management Messages: SUPP.SAFTE 1 = The SATA controller supports the SAF-TE message type. 15 :10 RWO SGPIO Enclosure Management Messages: SUPP.SGPIO 1 = The SATA controller supports the SGPIO register interface message type. 18 16 SUPP.LED LED Message Types: 1 = The SATA controller supports the LED message type. Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 480 Offset Start: 20h Offset End: 23h RO RO October 2012 Order Number: 327879-001US 8.0 Table 8-93. Offset 20h: Serial Enclosure Management Control Register (B0:D31:F2) (Sheet 2 of 2) Description: View: PCI BAR: ABAR Size: 32 bit Bus:Device:Function: B0:D3 1:F2 Default: 07010000h Bit Range Offset Start: 20h Offset End: 23h Power Well: Bit Acronym Bit Description RST Reset: 1 = A write of `0' to this bit by software will have no effect. 0 = When set by software, The SATA controller shall reset all enclosure management message logic and take all appropriate reset actions to ensure messages can be transmitted / received after the reset. After the SATA controller completes the reset operation, the SATA controller shall set the value to `0'. RW 31 :08 CTL.TM Transmit Message: 0 = A write of `0' to this bit by software will have no effect. 1 = When set by software, The SATA controller shall transmit the message contained in the message buffer. When the message is completely sent, the SATA controller shall set the value to `0'. Software shall not change the contents of the message buffer while CTL.TM is set to '1'. RW 07 :01 Reserved 09 00 8.4.1.10 STS.MR Sticky Bit Reset Value Bit Access Reserved Message Received -- RO. Message Received is not supported in the PCH. RO Offset 24h: Serial Extended Host Capabilities (B0:D31:F2) Table 8-94. Offset 24h: Serial Extended Host Capabilities (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: ABAR Bus:Device:Function: B0:D3 1:F2 Default: 0000000Xh Bit Acronym Offset Start: 24h Offset End: 24h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 481 8.4.1.11 Offset 6Ch: Serial NVMHCI Ports Implemented (B0:D31:F2) Table 8-95. Offset SIDPBA + 04h: Serial ATA Index (B0:D31:F2) Description: View: PCI BAR: ABAR Size: 32 bit B0:D3 1:F2 Default: 00000001h Bit Acronym Bit Range 31 :00 8.4.1.12 Bus:Device:Function: Offset Start: 6Ch Offset End: 6Ch Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved. Returns 0. Offset 70h: Serial AHCI Version (B0:D31:F2) This register indicates the major and minor versions of the NVMHCI specification. It is BCD encoded. The upper two bytes represent the major version number and the lower two bytes represent the minor version number. Example: Version 3.12 would be represented as 00030102h. The current version of the specification is 1.0 (00010000h). Table 8-96. Offset 70h: Serial AHCI Version (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range Bus:Device:Function: B0:D3 1:F2 BAR: ABAR Default: 00010000h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access 31 :16 MJR Major Version Number -- Indicates the major version is 1 RO 15 :00 MNR Minor Version Number -- Indicates the minor version is 0. RO Intel(R) Communications Chipset 89xx Series - Datasheet 482 Offset Start: 70h Offset End: 73h October 2012 Order Number: 327879-001US 8.0 8.4.1.13 Offset A0h: Serial Vendor Specific (B0:D31:F2) Table 8-97. Offset A0h: Serial Vendor Specific (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :01 00 BAR: ABAR Bus:Device:Function: Default: 00000000h Bit Acronym B0:D3 1:F2 Offset Start: A0h Offset End: A3h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved SLPD October 2012 Order Number: 327879-001US Supports Low Power Device Detection -- Indicates whether SATA power management and device hot (un)plug is supported. 0 = Not supported. 1 = Supported. RWO Intel(R) Communications Chipset 89xx Series - Datasheet 483 8.4.2 Port Registers (B0:D31:F2) Ports not available will result in the corresponding Port DMA register space being reserved. The controller shall ignore writes to the reserved space on write cycles and shall return `0' on read cycle accesses to the reserved location. There is a set of Port Registers for each Port 4 and 5. Port 4 registers begin at ABAR + 300h Port 5 registers begin at ABAR + 380h Table 8-98. Port [4:5] DMA Register Address Map Offset Start Offset End Default Value Register ID - Description 300h at 80h 303h at 80h "Offset 300h: Port [4:5] Command List Base Address Register (B0:D31:F2)" on page 487 Undefined 304h at 80h 307h at 80h "Offset 304: Port [4:5] Command List Base Address Upper 32-Bits Register (B0:D31:F2)" on page 487 Undefined 308h + 80h 30Bh + 80h "Offset 308h: Port [4:5] FIS Base Address Register (B0:D31:F2)" on page 488 Undefined 30Fh + 80h "Offset 30Ch: Port [4:5] FIS Base Address Upper 32-Bits Register (B0:D31:F2)" on Undefined page 488 30Ch + 80h 310h + 80h 313h + 80h "Offset 310h: Port [4:5] Interrupt Status Register (B0:D31:F2)" on page 489 0000000h 314h + 80h 317h + 80h "Offset 314h: Port [4:5] Interrupt Enable Register (B0:D31:F2)" on page 491 0000000h 318h + 80h 31Bh + 80h "Offset 318h: Port [4:5] Command Register (B0:D31:F2)" on page 493 0000w00wh 320h + 80h 323h + 80h "Offset 320h: Port [4:5] Task File Data Register (B0:D31:F2)" on page 496 0000007Fh 324h + 80h 327h + 80h "Offset 324h: Port [4:5] Signature Register (B0:D31:F2)" on page 497 FFFFFFFFh 328h + 80h 32Bh + 80h "Offset 328h: Port [4:5] Serial ATA Status Register (B0:D31:F2)" on page 498 00000000h 32Ch + 80h 32Fh + 80h "Offset 32Ch: Port [4:5] Serial ATA Control Register (B0:D31:F2)" on page 500 000000004h 330h + 80h 333h + 80h "Offset 330h: Port [4:5] Serial ATA Error Register (B0:D31:F2)" on page 502 00000000h 334h + 80h 337h + 80h "Offset 334h: Port [4:5] Serial ATA Active (B0:D31:F2)" on page 503 00000000h 338h + 80h 33Bh + 80h "Offset 338h: Port [4:5] Command Issue Register (B0:D31:F2)" on page 504 00000000h Note: Port 4 offset starts at 300h. Port 5 offset starts at 380h. Intel(R) Communications Chipset 89xx Series - Datasheet 484 October 2012 Order Number: 327879-001US 8.0 8.4.2.1 Offset 300h: Port [4:5] Command List Base Address Register (B0:D31:F2) Address Offset: Port 4: ABAR + 300h Port 5: ABAR + 380h Table 8-99. Offset 300h: Port [4:5] Command List Base Address Register (B0:D31:F2) Description: View: PCI BAR: ABAR Size: 32 bit B0:D31 :F2 Default: Undefined Bit Range 31 :10 09 :00 8.4.2.2 Bus:Device:Function: Offset Start: 300h at 80h Offset End: 303h at 80h Power Well: Bit Acronym Bit Description CLB Command List Base Address -- Indicates the 32-bit base for the command list for this port. This base is used when fetching commands to execute. The structure pointed to by this address range is 1 KB in length. This address must be 1-KB aligned as indicated by bits 31:10 being read/write. These bits are not reset on a Controller reset. Sticky Bit Reset Value Bit Access RW Reserved Reserved Offset 304h: Port [4:5] Command List Base Address Upper 32-Bits Register (B0:D31:F2) Address Offset: Port 4: ABAR + 304h Port 5: ABAR + 384h Table 8-100. Offset 304: Port [4:5] Command List Base Address Upper 32-Bits Register (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: ABAR Bus:Device:Function: B0:D31 :F2 Default: Undefined Power Well: Bit Acronym Bit Description CLBU Command List Base Address Upper -- Indicates the upper 32-bits for the command list base address for this port. This base is used when fetching commands to execute. These bits are not reset on a Controller reset. October 2012 Order Number: 327879-001US Offset Start: 304h at 80h Offset End: 307h at 80h Sticky Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 485 8.4.2.3 Offset 308h: Port [4:5] FIS Base Address Register (B0:D31:F2) Address Offset: Port 4: ABAR + 308h Port 5: ABAR + 388h Table 8-101. Offset 308h: Port [4:5] FIS Base Address Register (B0:D31:F2) Description: View: PCI BAR: ABAR Size: 32 bit B0:D31 :F2 Default: Undefined Bit Range 31 :08 07 :00 8.4.2.4 Bus:Device:Function: Offset Start: 308h + 80h Offset End: 30Bh + 80h Power Well: Bit Acronym Bit Description FB FIS Base Address -- Indicates the 32-bit base for received FISes. The structure pointed to by this address range is 256 bytes in length. This address must be 256-byte aligned, as indicated by bits 31:3 being read/write. These bits are not reset on a Controller reset. Sticky Bit Reset Value Bit Access RW Reserved Reserved Offset 30Ch: Port [4:5] FIS Base Address Upper 32-Bits Register (B0:D31:F2) Address Offset: Port 4: ABAR + 30Ch Port 5: ABAR + 38Ch Table 8-102. Offset 30Ch: Port [4:5] FIS Base Address Upper 32-Bits Register (B0:D31:F2) Description: View: PCI Size: 32 bit Bus:Device:Function: B0:D31 :F2 BAR: ABAR Default: Undefined Power Well: Bit Range Bit Acronym Bit Description 31 :00 CLBU Command List Base Address Upper -- Indicates the upper 32-bits for the received FIS base for this port. These bits are not reset on a Controller reset. Intel(R) Communications Chipset 89xx Series - Datasheet 486 Offset Start: 30Ch + 80h Offset End: 30Fh + 80h Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 8.0 8.4.2.5 Offset 310h: Port [4:5] Interrupt Status Register (B0:D31:F2) Address Offset: Port 4: ABAR + 310h Port 5: ABAR + 390h Table 8-103. Offset 310h: Port [4:5] Interrupt Status Register (B0:D31:F2) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: ABAR Bus:Device:Function: Default: 0000000h Bit Acronym B0:D31 :F2 Offset Start: 310h + 80h Offset End: 313h + 80h Power Well: Bit Description Sticky Bit Reset Value Bit Access 31 CPDS Cold Port Detect Status -- Cold presence detect is not supported. RO 30 TFES Task File Error Status -- This bit is set whenever the status register is updated by the device and the error bit (PxTFD.bit 0) is set. RWC 29 HBFS Host Bus Fatal Error Status -- Indicates that the PCH encountered an error that it cannot recover from due to a bad software pointer. In PCI, such an indication would be a target or master abort. RWC 28 HBDS Host Bus Data Error Status -- Indicates that the PCH encountered a data error (uncorrectable ECC / parity) when reading from or writing to system memory. RWC 27 IFS Interface Fatal Error Status -- Indicates that the PCH encountered an error on the SATA interface which caused the transfer to stop. RWC 26 INFS Interface Non-fatal Error Status -- Indicates that the PCH encountered an error on the SATA interface but was able to continue operation. RWC 25 24 23 22 21 :08 07 Reserved Reserved OFS Overflow Status -- indicates that the PCH received more bytes from a device than was specified in the PRD table for the command. RWC Reserved Reserved PRCS PhyRdy Change Status -- When set to one indicates the internal PhyRdy signal changed state. This bit reflects the state of PxSERR.DIAG.N. Unlike most of the other bits in the register, this bit is RO and is only cleared when PxSERR.DIAG.N is cleared. The internal PhyRdy signal also transitions when the port interface enters partial or slumber power management states. Partial and slumber must be disabled when Surprise Removal Notification is desired, otherwise the power management state transitions will appear as false insertion and removal events. RO Reserved Reserved DIS October 2012 Order Number: 327879-001US Device Interlock Status -- When set, indicates that a platform interlock switch has been opened or closed, which may lead to a change in the connection state of the device. This bit is only valid in systems that support an interlock switch (CAP.SIS [ABAR+00:bit 28] set). For systems that do not support an interlock switch, this bit will always be 0. RWC Intel(R) Communications Chipset 89xx Series - Datasheet 487 Table 8-103. Offset 310h: Port [4:5] Interrupt Status Register (B0:D31:F2) (Sheet 2 of 2) Description: View: PCI Size: 32 bit BAR: ABAR Bus:Device:Function: Default: 0000000h Offset Start: 310h + 80h Offset End: 313h + 80h Power Well: Bit Acronym Bit Description 06 PCS Port Connect Change Status -- This bit reflects the state of PxSERR.DIAG.X. (ABAR+130h/1D0h/ 230h/2D0h, bit 26) Unlike other bits in this register, this bit is only cleared when PxSERR.DIAG.X is cleared. 0 = No change in Current Connect Status. 1 = Change in Current Connect Status. RO 05 DPS Descriptor Processed -- A PRD with the I bit set has transferred all its data. RWC 04 UFS Unknown FIS Interrupt -- When set to `1' indicates that an unknown FIS was received and has been copied into system memory. This bit is cleared to `0' by software clearing the PxSERR.DIAG.F bit to `0'. This bit does not directly reflect the PxSERR.DIAG.F bit. PxSERR.DIAG.F is set immediately when an unknown FIS is detected, whereas this bit is set when the FIS is posted to memory. Software should wait to act on an unknown FIS until this bit is set to `1' or the two bits may become out of sync. RO 03 SDBS Set Device Bits Interrupt -- A Set Device Bits FIS has been received with the I bit set and has been copied into system memory. RWC 02 DSS DMA Setup FIS Interrupt -- A DMA Setup FIS has been received with the I bit set and has been copied into system memory. RWC 01 PSS PIO Setup FIS Interrupt -- A PIO Setup FIS has been received with the I bit set, it has been copied into system memory, and the data related to that FIS has been transferred. RWC 00 DHRS Device to Host Register FIS Interrupt -- A D2H Register FIS has been received with the I bit set, and has been copied into system memory. RWC Bit Range Intel(R) Communications Chipset 89xx Series - Datasheet 488 B0:D31 :F2 Sticky Bit Reset Value Bit Access October 2012 Order Number: 327879-001US 8.0 8.4.2.6 Offset 314h: Port [4:5] Interrupt Enable Register (B0:D31:F2) Address Offset: Port 4: ABAR + 314h Port 5: ABAR + 394h This register enables and disables the reporting of the corresponding interrupt to system software. When a bit is set (`1') and the corresponding interrupt condition is active, then an interrupt is generated. Interrupt sources that are disabled (`0') are still reflected in the status registers. Table 8-104. Offset 314h: Port [4:5] Interrupt Enable Register (B0:D31:F2) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: ABAR B0:D31 Bus:Device:Function: :F2 Default: 0000000h Bit Acronym Offset Start: 314h + 80h Offset End: 317h + 80h Power Well: Bit Description Sticky Bit Reset Value Bit Access 31 CPDE Cold Presence Detect Enable -- Cold Presence Detect is not supported. RO 30 TFEE Task File Error Enable -- When set, and GHC.IE and PxTFD.STS.ERR (due to a reception of the error register from a received FIS) are set, the PCH will generate an interrupt. RW 29 HBFE Host Bus Fatal Error Enable -- When set, and GHC.IE and PxS.HBFS are set, the PCH will generate an interrupt. RW 28 HBDE Host Bus Data Error Enable -- When set, and GHC.IE and PxS.HBDS are set, the PCH will generate an interrupt. RW 27 HBDE Host Bus Data Error Enable -- When set, GHC.IE is set, and PxIS.HBDS is set, the PCH will generate an interrupt. RW 26 INFE Interface Non-fatal Error Enable -- When set, GHC.IE is set, and PxIS.INFS is set, the PCH will generate an interrupt. RW 25 24 23 22 21 :08 Reserved Reserved OFE Overflow Error Enable -- When set, and GHC.IE and PxS.OFS are set, the PCH will generate an interrupt. RW Reserved Reserved PRCE PhyRdy Change Interrupt Enable -- When set, and GHC.IE is set, and PxIS.PRCS is set, the PCH shall generate an interrupt. RW Reserved Reserved 07 DIE Device Interlock Enable -- When set, and PxIS.DIS is set, the PCH will generate an interrupt. For systems that do not support an interlock switch, this bit shall be a read-only 0. RW 06 PCE Port Change Interrupt Enable -- When set, and GHC.IE and PxS.PCS are set, the PCH will generate an interrupt. RW 05 DPE Descriptor Processed Interrupt Enable -- When set, and GHC.IE and PxS.DPS are set, the PCH will generate an interrupt RW 04 UFIE Unknown FIS Interrupt Enable -- When set, and GHC.IE is set and an unknown FIS is received, the PCH will generate this interrupt. RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 489 Table 8-104. Offset 314h: Port [4:5] Interrupt Enable Register (B0:D31:F2) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: ABAR Bus:Device:Function: Default: 0000000h Bit Acronym Offset Start: 314h + 80h Offset End: 317h + 80h Power Well: Bit Description Sticky Bit Reset Value Bit Access 03 SDBE Set Device Bits FIS Interrupt Enable -- When set, and GHC.IE and PxS.SDBS are set, the PCH will generate an interrupt. RW 02 DSE DMA Setup FIS Interrupt Enable -- When set, and GHC.IE and PxS.DSS are set, the PCH will generate an interrupt. RW 01 PSE PIO Setup FIS Interrupt Enable -- When set, and GHC.IE and PxS.PSS are set, the PCH will generate an interrupt. RW 00 DHRE Device to Host Register FIS Interrupt Enable -- When set, and GHC.IE and PxS.DHRS are set, the PCH will generate an interrupt. RW Intel(R) Communications Chipset 89xx Series - Datasheet 490 B0:D31 :F2 October 2012 Order Number: 327879-001US 8.0 8.4.2.7 Offset 318h: Port [4:5] Command Register (B0:D31:F2) Address Offset: Port 4: ABAR + 318h Port 5: ABAR + 398h Table 8-105. Offset 318h: Port [4:5] Command Register (B0:D31:F2) (Sheet 1 of 3) Description: View: PCI Size: 32 bit Bit Range B0:D31 Bus:Device:Function: :F2 BAR: ABAR Power Well: Default: 0000w00wh Bit Acronym Offset Start: 318h + 80h Offset 31Bh + 80h End: Bit Description Sticky Bit Reset Value Bit Access Interface Communication Control -- This is a four bit field which can be used to control reset and power states of the interface. Writes to this field will cause actions on the interface, either as primitives or an OOB sequence, and the resulting status of the interface will be reported in the PxSSTS register (Address offset Port 4: ABAR+224h, Port 5: ABAR+2A4h). 31 :28 ICC Value Definition Fh-7h Reserved 6h Slumber: This will cause the PCH to request a transition of the interface to the slumber state. The SATA device may reject the request and the interface will remain in its current state 5h-3h Reserved 2h Partial: This will cause the PCH to request a transition of the interface to the partial state. The SATA device may reject the request and the interface will remain in its current state. 1h Active: This will cause the PCH to request a transition of the interface into the active 0h No-Op / Idle: When software reads this value, it indicates the PCH is not in the process of changing the interface state or sending a device reset, and a new link command may be issued. RW When system software writes a non-reserved value other than NoOp (0h), the PCH will perform the action and update this field back to Idle (0h). If software writes to this field to change the state to a state the link is already in (e.g. interface is in the active state and a request is made to go to the active state), the PCH will take no action and return this field to Idle. When the ALPE bit (bit 26) is set, then this register should not be set to 02h or 06h. 27 ASP Aggressive Slumber / Partial -- When set, and the ALPE bit (bit 26) is set, the PCH shall aggressively enter the slumber state when it clears the PxCI register and the PxSACT register is cleared. When cleared, and the ALPE bit is set, the PCH will aggressively enter the partial state when it clears the PxCI register and the PxSACT register is cleared. If CAP.SALP is cleared to '0', software shall treat this bit as reserved. 26 ALPE Aggressive Link Power Management Enable -- When set, the PCH will aggressively enter a lower link power state (partial or slumber) based upon the setting of the ASP bit (bit 27). RW 25 HBDE Drive LED on ATAPI Enable -- When set, the PCH will drive the LED pin active for ATAPI commands (PxCLB[CHz.A] set) in addition to ATA commands. When cleared, the PCH will only drive the LED pin active for ATA commands. RW October 2012 Order Number: 327879-001US RW Intel(R) Communications Chipset 89xx Series - Datasheet 491 Table 8-105. Offset 318h: Port [4:5] Command Register (B0:D31:F2) (Sheet 2 of 3) Description: View: PCI Size: 32 bit Bit Range 24 Bit Acronym ATAPI Bit Description Device is ATAPI -- When set, the connected device is an ATAPI device. This bit is used by the PCH to control whether or not to generate the desktop LED when commands are active. Reserved Reserved 22 Reserved Reserved 20 19 ESP Offset Start: 318h + 80h Offset 31Bh + 80h End: Power Well: Default: 0000w00wh 23 21 B0:D31 Bus:Device:Function: :F2 BAR: ABAR External SATA Port: 0 = This port supports internal SATA devices only. 1 = This port will be used with an external SATA device. This bit is not reset by Function Level Reset. Sticky Bit Reset Value Bit Access RW RWO Reserved Reserved ISP Interlock Switch Attached to Port -- When interlock switches are supported in the platform (CAP.SIS [ABAR+00h:bit 28] set), this indicates whether this particular port has an interlock switch attached. This bit can be used by system software to enable such features as aggressive power management, as disconnects can always be detected regardless of PHY state with an interlock switch. When this bit is set, it is expected that HPCP (bit 18) in this register is also set. The PCH takes no action on the state of this bit - it is for system software only. For example, if this bit is cleared, and an interlock switch toggles, the PCH still treats it as a proper interlock switch event. RWO This bit is not reset on a Controller reset or by a Function Level Reset. 18 :16 Reserved Reserved 15 CR Controller Running -- When this bit is set, the DMA engines for a port are running. See section 5.2.2 of the Serial ATA AHCI Specification for details on when this bit is set and cleared by the PCH. RO 14 FR FIS Receive Running -- When set, the FIS Receive DMA engine for the port is running. See section 12.2.2 of the Serial ATA AHCI Specification for details on when this bit is set and cleared by the PCH. RO ISS Interlock Switch State -- For systems that support interlock switches (via CAP.SIS [ABAR+00h:bit 28]), if an interlock switch exists on this port (via ISP in this register), this bit indicates the current state of the interlock switch. A 0 indicates the switch is closed, and a 1 indicates the switch is opened. For systems that do not support interlock switches, or if an interlock switch is not attached to this port, this bit reports 0. RO CSS Current Command Slot -- Indicates the current command slot the PCH is processing. This field is valid when the ST bit is set in this register, and is constantly updated by the PCH. This field can be updated as soon as the PCH recognizes an active command slot, or at some point soon after when it begins processing the command. This field is used by software to determine the current command issue location of the PCH. In queued mode, software shall not use this field, as its value does not represent the current command being executed. Software shall only use PxCI and PxSACT when running queued commands. RO 13 12 :08 Intel(R) Communications Chipset 89xx Series - Datasheet 492 October 2012 Order Number: 327879-001US 8.0 Table 8-105. Offset 318h: Port [4:5] Command Register (B0:D31:F2) (Sheet 3 of 3) Description: View: PCI Size: 32 bit Bit Range 07 :05 04 03 B0:D31 Bus:Device:Function: :F2 BAR: ABAR Power Well: Default: 0000w00wh Bit Acronym Offset Start: 318h + 80h Offset 31Bh + 80h End: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved FRE CLO FIS Receive Enable -- When set, the PCH may post received FISes into the FIS receive area pointed to by PxFB (ABAR+108h/188h/ 208h/288h) and PxFBU (ABAR+10Ch/18Ch/20Ch/28Ch). When cleared, received FISes are not accepted by the PCH, except for the first D2H (device-to-host) register FIS after the initialization sequence. System software must not set this bit until PxFB (PxFBU) have been programmed with a valid pointer to the FIS receive area, and if software wishes to move the base, this bit must first be cleared, and software must wait for the FR bit (bit 14) in this register to be cleared. Command List Override -- Setting this bit to '1' causes PxTFD.STS.BSY and PxTFD.STS.DRQ to be cleared to '0'. This allows a software reset to be transmitted to the device regardless of whether the BSY and DRQ bits are still set in the PxTFD.STS register. The Controller sets this bit to '0' when PxTFD.STS.BSY and PxTFD.STS.DRQ have been cleared to '0'. A write to this register with a value of '0' has no effect. RW RW This bit shall only be set to '1' immediately prior to setting the PxCMD.ST bit to '1' from a previous value of '0'. Setting this bit to '1' at any other time is not supported and will result in indeterminate behavior. Software must wait for CLO to be cleared to '0' before setting PxCMD.ST to '1'. 02 01 00 POD Power On Device -- Cold presence detect not supported. Defaults to 1. RO SUD Spin-Up Device -- This bit is R/W and defaults to 0 for systems that support staggered spin-up (R/W when CAP.SSS (ABAR+00h:bit 27) is 1). Bit is RO 1 for systems that do not support staggered spin-up (when CAP.SSS is 0). 0 = No action. 1 = On an edge detect from 0 to 1, the PCH starts a COMRESET initialization sequence to the device. Clearing this bit to '0' does not cause any OOB signal to be sent on the interface. When this bit is cleared to '0' and PxSCTL.DET=0h, the Controller will enter listen mode. RW/RO ST Start -- When set, the PCH may process the command list. When cleared, the PCH may not process the command list. Whenever this bit is changed from a 0 to a 1, the PCH starts processing the command list at entry 0. Whenever this bit is changed from a 1 to a 0, the PxCI register is cleared by the PCH upon the PCH putting the controller into an idle state. See section 12.2.1 of the Serial ATA AHCI Specification for important restrictions on when ST can be set to 1. RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 493 8.4.2.8 Offset 320h: Port [4:5] Task File Data Register (B0:D31:F2) Address Offset: Port 4: ABAR + 320h Port 5: ABAR + 3A0h This is a 32-bit register that copies specific fields of the task file when FISes are received. The FISes that contain this information are: D2H Register FIS,PIO Setup FIS and Set Device Bits FIS. Table 8-106. Offset 320h: Port [4:5] Task File Data Register (B0:D31:F2) Description: View: PCI Size: 32 bit BAR: ABAR Bus:Device:Function: Default: 0000007Fh Bit Range Bit Acronym 31 :16 Reserved 15 :08 ERR B0:D31 :F2 Offset Start: 320h + 80h Offset End: 323h + 80h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Error -- Contains the latest copy of the task file error register. RO Status -- Contains the latest copy of the task file status register. Fields of note in this register that affect AHCI. Bit 07 :00 STS Field Definition 7 BSY Indicates the interface is busy 6:4 N/A Not applicable 3 DRQ Indicates a data transfer is requested 2:1 N/A Not applicable 0 ERR Indicates an error during the transfer Intel(R) Communications Chipset 89xx Series - Datasheet 494 RO October 2012 Order Number: 327879-001US 8.0 8.4.2.9 Offset 324h: Port [4:5] Signature Register (B0:D31:F2) Address Offset: Port 4: ABAR + 324h Port 5: ABAR + 3A4h This is a 32-bit register which contains the initial signature of an attached device when the first D2H Register FIS is received from that device. It is updated once after a reset sequence. Table 8-107. Offset 324h: Port [4:5] Signature Register (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range BAR: ABAR Bus:Device:Function: Default: FFFFFFFFh B0:D31 :F2 Offset Start: 324h + 80h Offset End: 327h + 80h Power Well: Bit Acronym Bit Description Sticky Bit Reset Value Bit Access Signature -- Contains the signature received from a device on the first D2H register FIS. The bit order is as follows: 31 :00 SIG October 2012 Order Number: 327879-001US Bit Field 31:24 LBA High Register 23:16 LBA Mid Register 15:8 LBA Low Register 7:0 Sector Count Register RO Intel(R) Communications Chipset 89xx Series - Datasheet 495 8.4.2.10 Offset 328h: Port [4:5] Serial ATA Status Register (B0:D31:F2) Address Offset: Port 4: ABAR + 328h Port 5: ABAR + 3A8h This is a 32-bit register that conveys the current state of the interface and host. The PCH updates it continuously and asynchronously. When the PCH transmits a COMRESET to the device, this register is updated to its reset values. Table 8-108. Offset 328h: Port [4:5] Serial ATA Status Register (B0:D31:F2) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 :12 BAR: ABAR Bus:Device:Function: Default: 00000000h Bit Acronym B0:D31 :F2 Offset Start: 328h + 80h Offset End: 32Bh + 80h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved Interface Power Management -- Indicates the current interface state: 11 :08 IPM Value Description 0h Device not present or communication not established 1h Interface in active state 2h Interface in PARTIAL power management state 6h Interface in SLUMBER power management state RO All other values reserved. Current Interface Speed -- Indicates the negotiated interface communication speed. 07 :04 SPD Value Description 0h Device not present or communication not established 1h Generation 1 communication rate negotiated 2h Generation 2 communication rate negotiated RO All other values reserved. The PCH supports Gen 1 communication rates (1.5 Gb/s) and Gen 2 rates (3.0 Gb/s). Intel(R) Communications Chipset 89xx Series - Datasheet 496 October 2012 Order Number: 327879-001US 8.0 Table 8-108. Offset 328h: Port [4:5] Serial ATA Status Register (B0:D31:F2) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: ABAR Bus:Device:Function: Default: 00000000h Bit Acronym B0:D31 :F2 Offset Start: 328h + 80h Offset End: 32Bh + 80h Power Well: Bit Description Sticky Bit Reset Value Bit Access Device Detection -- Indicates the interface device detection and Phy state: 03 :00 DET Value Description 0h No device detected and Phy communication not established 1h Device presence detected but Phy communication not established 3h Device presence detected and Phy communication established 4h Phy in offline mode as a result of the interface being disabled or running in a BIST loopback mode RO All other values reserved. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 497 8.4.2.11 Offset 32Ch: Port [4:5] Serial ATA Control Register (B0:D31:F2) Address Offset: Port 4: ABAR + 32Ch Port 5: ABAR + 3ACh This is a 32-bit read-write register by which software controls SATA capabilities. Writes to the SControl register result in an action being taken by the PCH or the interface. Reads from the register return the last value written to it. Table 8-109. Offset 32Ch: Port [4:5] Serial ATA Control Register (B0:D31:F2) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: ABAR Default: 000000004h Bit Acronym Reserved Reserved 19 :16 Reserved Reserved SPM B0:D31 :F2 Offset Start: 32Ch + 80h Offset End: 32Fh + 80h Power Well: Bit Description 31 :20 15 :12 Bus:Device:Function: Select Power Management -- This field is not used by AHCI Sticky Bit Reset Value Bit Access RW Interface Power Management Transitions Allowed -- Indicates which power states the PCH is allowed to transition to: 11 :08 IPM Value Description 0h No interface restrictions 1h Transitions to the PARTIAL state disabled 2h Transitions to the SLUMBER state disabled 3h Transitions to both PARTIAL and SLUMBER states disabled RW All other values reserved Speed Allowed -- Indicates the highest allowable speed of the interface. This speed is limited by the CAP.ISS (ABAR+00h:bit 23:20) field. Value 07 :04 SPD Description 0h No speed negotiation restrictions 1h Limit speed negotiation to Generation 1 communication rate 2h Limit speed negotiation to Generation 2 communication rate RW The PCH Supports Gen 1 communication rates (1.5 Gb/s) and Gen 2 rates (3.0 Gb/s). Intel(R) Communications Chipset 89xx Series - Datasheet 498 October 2012 Order Number: 327879-001US 8.0 Table 8-109. Offset 32Ch: Port [4:5] Serial ATA Control Register (B0:D31:F2) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: ABAR Bus:Device:Function: Default: 000000004h Bit Acronym B0:D31 :F2 Offset Start: 32Ch + 80h Offset End: 32Fh + 80h Power Well: Bit Description Sticky Bit Reset Value Bit Access Device Detection Initialization -- Controls the PCH's device detection and interface initialization. 03 :00 DET Value Description 0h No device detection or initialization action requested 1h Perform interface communication initialization sequence to establish communication. This is functionally equivalent to a hard reset and results in the interface being reset and communications re-initialized 4h Disable the Serial ATA interface and put Phy in offline mode RW All other values reserved. When this field is written to a 1h, the PCH initiates COMRESET and starts the initialization process. When the initialization is complete, this field shall remain 1h until set to another value by software. This field may only be changed to 1h or 4h when PxCMD.ST is 0. Changing this field while the PCH is running results in undefined behavior. Note: It is permissible to implement any of the Serial ATA defined behaviors for transmission of COMRESET when DET=1h. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 499 8.4.2.12 Offset 330h: Port [4:5] Serial ATA Error Register (B0:D31:F2) Address Offset: Port 4: ABAR + 330h Port 5: ABAR + 3B0h Bits 26:16 of this register contains diagnostic error information for use by diagnostic software in validating correct operation or isolating failure modes. Bits 11:0 contain error information used by host software in determining the appropriate response to the error condition. If one or more of bits 11:8 of this register are set, the controller will stop the current transfer. Table 8-110. Offset 330h: Port [4:5] Serial ATA Error Register (B0:D31:F2) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 :27 B0:D31 Bus:Device:Function: :F2 BAR: ABAR Power Well: Default: 00000000h Bit Acronym Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved 26 X Exchanged -- When set to one this bit indicates that a change in device presence has been detected since the last time this bit was cleared. This bit shall always be set to 1 anytime a COMINIT signal is received. This bit is reflected in the P0IS.PCS bit. RWC 25 F Unrecognized FIS Type -- Indicates that one or more FISs were received by the Transport layer with good CRC, but had a type field that was not recognized. RWC 24 T Transport state transition error -- Indicates that an error has occurred in the transition from one state to another within the Transport layer since the last time this bit was cleared. RWC 23 T Transport state transition error -- Indicates that an error has occurred in the transition from one state to another within the Transport layer since the last time this bit was cleared. RWC 22 H Handshake -- Indicates that one or more R_ERR handshake response was received in response to frame transmission. Such errors may be the result of a CRC error detected by the recipient, a disparity or 8b/10b decoding error, or other error condition leading to a negative handshake on a transmitted frame. RWC 21 C CRC Error -- Indicates that one or more CRC errors occurred with the Link Layer. RWC 20 D Disparity Error -- This field is not used by AHCI. RWC 19 B 10b to 8b Decode Error -- Indicates that one or more 10b to 8b decoding errors occurred. RWC 18 W Comm Wake -- Indicates that a Comm Wake signal was detected by the Phy. RWC 17 I Phy Internal Error -- Indicates that the Phy detected some internal error. RWC N PhyRdy Change -- When set to 1 this bit indicates that the internal PhyRdy signal changed state since the last time this bit was cleared. In the PCH, this bit will be set when PhyRdy changes from a 0 -> 1 or a 1 -> 0. The state of this bit is then reflected in the PxIS.PRCS interrupt status bit and an interrupt will be generated if enabled. Software clears this bit by writing a 1 to it. RWC 16 15 :12 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 500 Offset Start: 330h + 80h Offset 333h + 80h End: October 2012 Order Number: 327879-001US 8.0 Table 8-110. Offset 330h: Port [4:5] Serial ATA Error Register (B0:D31:F2) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 11 B0:D31 Bus:Device:Function: :F2 BAR: ABAR Power Well: Default: 00000000h Bit Acronym E 10 Offset Start: 330h + 80h Offset 333h + 80h End: Bit Description Sticky Bit Reset Value Bit Access Internal Error -- The SATA controller failed due to a master or target abort when attempting to access system memory. RWC Protocol Error -- A violation of the Serial ATA protocol was detected. Note: The PCH does not set this bit for all protocol violations that may occur on the SATA link. RWC 09 C Persistent Communication or Data Integrity Error -- A communication error that was not recovered occurred that is expected to be persistent. Persistent communications errors may arise from faulty interconnect with the device, from a device that has been removed or has failed, or a number of other causes. RWC 08 T Transient Data Integrity Error -- A data integrity error occurred that was not recovered by the interface. RWC 07 :02 Reserved Reserved 01 M Recovered Communications Error -- Communications between the device and host was temporarily lost but was re-established. This can arise from a device temporarily being removed, from a temporary loss of Phy synchronization, or from other causes and may be derived from the PhyNRdy signal between the Phy and Link layers. 00 I Recovered Data Integrity Error -- A data integrity error occurred that was recovered by the interface through a retry operation or other recovery action. 8.4.2.13 RWC RWC Offset 334h: Port [4:5] Serial ATA Active (B0:D31:F2) Address Offset: Port 4: ABAR + 334h Port 5: ABAR + 3B4h Table 8-111. Offset 334h: Port [4:5] Serial ATA Active (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: ABAR B0:D31 Bus:Device:Function: :F2 Default: 00000000h Power Well: Bit Acronym Bit Description DS Device Status -- System software sets this bit for SATA queuing operations prior to setting the PxCI.CI bit in the same command slot entry. This field is cleared via the Set Device Bits FIS. This field is also cleared when PxCMD.ST (ABAR+118h/198h/218h/298h:bit 0) is cleared by software, and as a result of a COMRESET or SRST. October 2012 Order Number: 327879-001US Offset Start: 334h + 80h Offset End: 337h + 80h Sticky Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 501 8.4.2.14 Offset 338h: Port [4:5] Command Issue Register (B0:D31:F2) Address Offset: Port 4: ABAR + 338h Port 5: ABAR + 3B8h Table 8-112. Offset 338h: Port [4:5] Command Issue Register (B0:D31:F2) Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: ABAR Bus:Device:Function: Default: 00000000h B0:D31 :F2 Offset Start: 338h + 80h Offset End: 33Bh + 80h Power Well: Bit Acronym Bit Description CI Commands Issued -- This field is set by software to indicate to the PCH that a command has been builtin system memory for a command slot and may be sent to the device. When the PCH receives a FIS which clears the BSY and DRQ bits for the command, it clears the corresponding bit in this register for that command slot. Bits in this field shall only be set to '1' by software when PxCMD.ST is set to '1'. This field is also cleared when PxCMD.ST (ABAR+118h/198h/218h/298h:bit 0) is cleared by software. Sticky Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 502 October 2012 Order Number: 327879-001US 9.0 9.0 SATA Controller Registers (B0:D31:F5) 9.1 PCI Configuration Registers (SATA-B0:D31:F5) Note: Address locations that are not shown should be treated as Reserved. All of the SATA registers are in the core well. None of the registers can be locked. 9.1.1 SATA Controller PCI Register Address Map Table 9-1. SATA Controller PCI Register Address Map (Sheet 1 of 2) Offset Start 00h Offset End 01h Register ID - Description Default Value "Offset 00h: VID: Vendor Identification Register (SATA--B0:D31:F5)" on page 506 8086h 02h 03h "Offset 02h: Device Identification Register (SATA--B0:D31:F5)" on page 507 2326h 04h 05h "Offset 04h: PCI Command Register (SATA--B0:D31:F5)" on page 507 0007h 06h 07h "Offset 06h: PCI Status Register (SATA--B0:D31:F5)" on page 508 02B0h 08h "Offset 08h: RID--Revision Identification Register (SATA--B0:D31:F5)" on page 509 See register description 08h 09h 09h "Offset 09h: Programming Interface Register (SATA--B0:D31:F5)" on page 510 00h 0Ah 0Ah "Offset 0Ah: Sub Class Code Register (SATA--B0:D31:F5)" on page 510 01h 0Bh 0Bh "Offset 0Bh: BCC--Base Class Code Register (SATA-B0:D31:F5)" on page 511 01h 0Dh 0Dh "Offset 0Dh: Primary Master Latency Timer Register (SATA-B0:D31:F5)" on page 511 00h 10h 13h "Offset 10h: Primary Command Block Base Address Register (SATA-B0:D31:F5)" on page 512 00000001h 14h 17h "Offset 14h: Primary Control Block Base Address Register (SATA-B0:D31:F5)" on page 513 00000001h 18h 1Bh "Offset 18h: Secondary Command Block Base Address Register (IDE D31:F1)" on page 513 00000001h 1Ch 1Fh "Offset 1Ch: Secondary Control Block Base Address Register (IDE B0:D31:F5)" on 00000001h page 514 20h 23h "Offset 20h: Legacy Bus Master Base Address Register (SATA-B0:D31:F5)" on page 514 00000001h 24h 27h "Offset 24h: SATA Index/Data Pair Base Address Register (SATA-B0:D31:F5)" on page 515 00000000h 2Ch 2Dh "Offset 2Ch: Subsystem Vendor Identification Register (SATA-B0:D31:F5)" on page 515 0000h 2Eh 2Fh "Offset 2Eh: Subsystem Identification Register (SATA-B0:D31:F5)" on page 516 0000h 34h 34h "Offset 34h: Capabilities Pointer Register (SATA-B0:D31:F5)" on page 516 70h 3Ch 3Ch "Offset 3Ch: Interrupt Line Register (SATA-B0:D31:F5)" on page 516 00h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 505 Table 9-1. SATA Controller PCI Register Address Map (Sheet 2 of 2) Offset Start Offset End Default Value Register ID - Description 3Dh 3Dh "Offset 3Dh: Interrupt Pin Register (SATA-B0:D31:F5)" on page 517 See register description 40h 43h "Offset 40h: IDE Timing Register (SATA-B0:D31:F5)" on page 517 0000h 48h 48h "Offset 48h: Synchronous DMA Control Register (SATA-B0:D31:F5)" on page 518 00h 4Ah 4Bh "Offset 4Ah: Synchronous DMA Timing Register (SATA-B0:D31:F5)" on page 518 0000h 54h 57h "Offset 54h: IDE I/O Configuration Register (SATA-B0:D31:F5)" on page 519 00000000h 70h 71h "Offset 70h: PCI Power Management Capability Identification Register (SATA- B0:D31:F5)" on page 520 B001h 72h 73h "Offset 72h: PC--PCI Power Management Capabilities Register (SATA-B0:D31:F5)" 4003h on page 520 74h 75h "Offset 74h: PCI Power Management Control and Status Register (SATA- B0:D31:F5)" on page 521 0008h 90h 90h "Offset 90h: MAP--Address Map Register (SATA-B0:D31:F5)" on page 522 00h 92h 93h "Offset 92h: Port Control and Status Register (SATA-B0:D31:F5)" on page 523 0000h A8h ABh "Offset A8h: SATA Capability Register 0 (SATA-B0:D31:F5)" on page 524 0010B012h ACh AFh "Offset ACh: SATA Capability Register 1 (SATA-B0:D31:F5)" on page 524 00000048h B0h B1h "Offset B0h: FLR Capability ID (SATA-B0:D31:F5)" on page 525 0009h B2h B3h "Offset B2h: FLR Capability Length and Value (SATA-B0:D31:F5)" on page 525 2006h B2h B3h "Offset B2h: FLR Capability Length and Value (SATA-B0:D31:F5)" on page 526 2006h B4h B5h "Offset B4h: FLR Control (SATA-B0:D31:F5)" on page 526 0000h C0h C0h "Offset C0h: APM Trapping Control Register (SATA-B0:D31:F5)" on page 527 00h C4h C4h "Offset C4h: APM Trapping Control Register (SATA-B0:D31:F5)" on page 527 00h Note: The PCH SATA controller is not arbitrated as a PCI device, therefore it does not need a master latency timer. 9.1.1.1 Offset 00h: Vendor Identification Register (SATA--B0:D31:F5) Table 9-2. Offset 00h: VID: Vendor Identification Register (SATA--B0:D31:F5) Description: View: PCI Size: 16 bit Bus:Device:Function: B0:D31 :F5 BAR: Configuration Default: 8086h Bit Range Bit Acronym 15 :00 VID Power Well: Core Bit Description Vendor Identification: This 16-bit value is assigned to Intel. Intel VID = 8086h Intel(R) Communications Chipset 89xx Series - Datasheet 506 Offset Start: 00h Offset End: 01h Sticky Bit Reset Value Bit Access 8086h RO October 2012 Order Number: 327879-001US 9.0 9.1.1.2 Offset 02h: Device Identification Register (SATA--B0:D31:F5) Table 9-3. Offset 02h: Device Identification Register (SATA--B0:D31:F5) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D31 :F5 Default: 2326h Bit Range Bit Acronym 15 :00 DID Offset Start: 02h Offset End: 03h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 2326h RO Device ID -- This is a 16-bit value assigned to the PCH SATA controller. 9.1.1.3 Offset 04h: PCI Command Register (SATA--B0:D31:F5) Table 9-4. Offset 04h: PCI Command Register (SATA--B0:D31:F5) (Sheet 1 of 2) Description: View: PCI Size: 16 bit B0:D31 Bus:Device:Function: :F5 BAR: Configuration Default: 0007h Bit Range Bit Acronym 15 :11 Reserved Offset Start: 04h Offset End: 05h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00h Interrupt Disable -- This disables pin-based INTx# interrupts. This bit has no effect on MSI operation. 0 = Internal INTx# messages are generated if there is an interrupt and MSI is not enabled. 1 = Internal INTx# messages will not be generated. 0h RW Fast Back to Back Enable -- Reserved as 0. 0h RO 10 INTx_Disable 09 FBE 08 SERR_EN SERR# Enable -- Reserved as 0. 0h RO 07 WCC Wait Cycle Control -- Reserved as 0. 0h RO 06 PER Parity Error Response: 0 = Disabled. SATA controller will not generate PERR# when a data parity error is detected. 1 = Enabled. SATA controller will generate PERR# when a data parity error is detected. 0h RW 05 VPS VGA Palette Snoop -- Reserved as 0. 0h RO 04 PMWE Postable Memory Write Enable -- Reserved as 0. 0h RO 03 SCE Special Cycle Enable -- Reserved as 0. 0h RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 507 Table 9-4. Offset 04h: PCI Command Register (SATA--B0:D31:F5) (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: Default: 0007h Offset Start: 04h Offset End: 05h Power Well: Core Bit Reset Value Bit Access Bus Master Enable -- This bit controls the PCH's ability to act as a PCI master for IDE Bus Master transfers. This bit does not impact the generation of completions for split transaction commands. 1 RW MSE Memory Space Enable -- This controller does not support AHCI, therefore no memory space is required. 1 RO IOSE I/O Space Enable -- This bit controls access to the I/O space registers. 0 = Disables access to the Legacy or Native IDE ports (both Primary and Secondary) as well as the Bus Master I/ O registers. 1 = Enable. The Base Address register for the Bus Master registers should be programmed before this bit is set. 1 RW Bit Acronym Bit Description 02 BME 01 00 B0:D31 :F5 Sticky 9.1.1.4 Offset 06h: PCI Status Register (SATA--B0:D31:F5) Table 9-5. Offset 06h: PCI Status Register (SATA--B0:D31:F5) Description: View: PCI Size: 16 bit Bus:Device:Function: B0:D31 :F5 BAR: Configuration Default: 02B0h Power Well: Bit Reset Value Bit Access DPE Detected Parity Error: 0 = Parity error not detected. 1 = Indicates that the PCH detected a parity error on the internal backbone. This bit gets set even if the Parity Error Response bit (B0:D31:F5: Offset 04h bit 6) is not set. 0h RWC 14 SSE Signaled System Error -- Set when the LPC bridge signals a system error to the internal SERR# logic. 0h RWC 13 RMA Received Master Abort: 0 = No master abort received. 1 = Set when the bridge receives a master abort status from the I/O data bus. 0h RWC 12 Reserved 11 STA 0h RO 00h RO 0h RWC Bit Range Bit Acronym Bit Description 15 10 :09 08 DEV_STS DPED Sticky Reserved Signaled Target Abort -- Reserved as 0. DEVSEL# Timing Status: 01 = Hardwired; Controls the device select time for the SATA controller's PCI interface. Data Parity Error Detected -- For PCH, this bit can only be set on read completions received from SiBUS where there is a parity error. 1 = SATA controller, as a master, either detects a parity error or sees the parity error line asserted, and the parity error response bit (bit 6 of the command register) is set. Intel(R) Communications Chipset 89xx Series - Datasheet 508 Offset Start: 06h Offset End: 07h October 2012 Order Number: 327879-001US 9.0 Table 9-5. Offset 06h: PCI Status Register (SATA--B0:D31:F5) Description: View: PCI Size: 16 bit Bit Range 07 BAR: Configuration Bit Acronym FB2BC UDF 66MHZ_CAP 04 CAP_LIST Bit Description Sticky Reserved Bit Reset Value Bit Access 0h RO User Definable Features -- Reserved as 0. 0h RO 66MHz Capable -- Reserved as 1. 0h RO 1 RO 0h RO Capabilities List -- This bit indicates the presence of a capabilities list. The minimum requirement for the capabilities list must be PCI power management for the SATA controller. Interrupt Status -- Reflects the state of INTx# messages, IRQ14 or IRQ15. 0 = Interrupt is cleared (independent of the state of Interrupt Disable bit in the command register [offset 04h]). 1 = Interrupt is to be asserted INTS Offset Start: 06h Offset End: 07h Power Well: Fast Back to Back Capable -- Reserved as 1. 05 2 :0 B0:D31 :F5 Default: 02B0h 06 03 Bus:Device:Function: Reserved 9.1.1.5 Offset 08h: RID--Revision Identification Register (SATA--B0:D31:F5) Table 9-6. Offset 08h: RID--Revision Identification Register (SATA--B0:D31:F5) Description: View: PCI Size: 8 bit BAR: Configuration B0:D31 Bus:Device:Function: :F5 Default: See bit description Power Well: Bit Range Bit Acronym Bit Description 07 :00 RID Revision ID: This is an 8-bit value indicating the stepping of the SATA controller hardware October 2012 Order Number: 327879-001US Offset Start: 08h Offset End: 08h Sticky Bit Reset Value Bit Access Variable RO Intel(R) Communications Chipset 89xx Series - Datasheet 509 9.1.1.6 Offset 09h: Programming Interface Register (SATA-B0:D31:F5) Table 9-7. Offset 09h: Programming Interface Register (SATA--B0:D31:F5) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration 03 02 01 00 B0:D31 :F5 Default: 00h Bit Acronym Offset Start: 09h Offset End: 09h Power Well: Bit Description Sticky Bit Reset Value Bit Access This read-only bit is a 1 to indicate that the PCH supports bus master operation 07 06 :04 Bus:Device:Function: Reserved RO Reserved SNC Secondary Mode Native Capable -- Indicates whether or not the secondary channel has a fixed mode of operation. 0 = Indicates the mode is fixed and is determined by the (read-only) value of bit 2. This bit will always return `0'. RO SNE Secondary Mode Native Enable: Determines the mode that the secondary channel is operating in. 1 = Secondary controller operating in native PCI mode. This bit will always return `1'. RO PNC Primary Mode Native Capable -- Indicates whether or not the primary channel has a fixed mode of operation. 0 = Indicates the mode is fixed and is determined by the (read-only) value of bit 0. This bit will always return `0'. RO PNE Primary Mode Native Enable: Determines the mode that the primary channel is operating in. 1 = Primary controller operating in native PCI mode. This bit will always return `1'. RO 9.1.1.7 Offset 0Ah: Sub Class Code Register (SATA-B0:D31:F5) Table 9-8. Offset 0Ah: Sub Class Code Register (SATA--B0:D31:F5) Description: View: PCI Size: 8 bit Bus:Device:Function: B0:D31 :F5 BAR: Configuration Default: 01h Bit Range Bit Acronym 07 :00 SCC Power Well: Bit Description Sub Class Code: The value of this field determines whether the controller supports legacy IDE mode. Intel(R) Communications Chipset 89xx Series - Datasheet 510 Offset Start: 0Ah Offset End: 0Ah Sticky Bit Reset Value Bit Access RO October 2012 Order Number: 327879-001US 9.0 9.1.1.8 Offset 0Bh: BCC--Base Class Code Register (SATA-B0:D31:F5) Table 9-9. Offset 0Bh: BCC--Base Class Code Register (SATA-B0:D31:F5) Description: View: PCI Size: 8 bit BAR: Configuration B0:D31 :F5 Default: 01h Bit Range Bit Acronym 07 :00 BCC 9.1.1.9 Bus:Device:Function: Offset Start: 0Bh Offset End: 0Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Base Class Code: 01h = Mass storage device. RO Offset 0Dh: Primary Master Latency Timer Register (SATA-B0:D31:F5) Table 9-10. Offset 0Dh: Primary Master Latency Timer Register (SATA-B0:D31:F5) Description: View: PCI Size: 8 bit BAR: Configuration B0:D31 Bus:Device:Function: :F5 Default: 00h Power Well: Bit Range Bit Acronym Bit Description 07 :00 MLTC Master Latency Timer Count: 00h = Hardwired. The SATA controller is implemented internally, and is not arbitrated as a PCI device, so it does not need a Master Latency Timer. October 2012 Order Number: 327879-001US Offset Start: 0Dh Offset End: 0Dh Sticky Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 511 9.1.1.10 Offset 10h: Primary Command Block Base Address Register (SATA- B0:D31:F5) This 8-byte I/O space is used in native mode for the Primary Controller's Command Block. Table 9-11. Offset 10h: Primary Command Block Base Address Register (SATA- B0:D31:F5) Description: View: PCI Size: 32 bit BAR: Configuration Bit Acronym 31 :16 Reserved 15 :03 BA 02 :01 Reserved RTE Offset Start: 10h Offset End: 13h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Base Address -- This field provides the base address of the I/O space (8 consecutive I/O locations). RW Reserved Resource Type Indicator -- Hardwired to 1 to indicate a request for I/O space. Intel(R) Communications Chipset 89xx Series - Datasheet 512 B0:D31 :F5 Default: 00000001h Bit Range 00 Bus:Device:Function: RO October 2012 Order Number: 327879-001US 9.0 9.1.1.11 Offset 14h: Primary Control Block Base Address Register (SATA- B0:D31:F5) This 4-byte I/O space is used in native mode for the Primary Controller's Command Block. Table 9-12. Offset 14h: Primary Control Block Base Address Register (SATA-B0:D31:F5) Description: View: PCI Size: 32 bit BAR: Configuration Bit Acronym 31 :16 Reserved 15 :03 BA 02 :01 Reserved 9.1.1.12 B0:D31 :F5 Default: 00000001h Bit Range 00 Bus:Device:Function: Offset Start: 14h Offset End: 17h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Base Address -- This field provides the base address of the I/O space (8 consecutive I/O locations). RW Reserved Resource Type Indicator -- Hardwired to 1 to indicate a request for I/O space. RTE RO Offset 18h: SCMD-Secondary Command Block Base Address Register (SATA--B0:D31:F5) This 8-byte I/O space is used in native mode for the Secondary Controller's Command Block. Table 9-13. Offset 18h: Secondary Command Block Base Address Register (IDE D31:F1) Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000001h Bit Range Bit Acronym 31 :16 Reserved 15 :03 BA 02 :01 Reserved 00 Bus:Device:Function: B0:D31 :F5 RTE October 2012 Order Number: 327879-001US Offset Start: 18h Offset End: 1Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Base Address -- This field provides the base address of the I/O space (8 consecutive I/O locations). RW Reserved Resource Type Indicator -- Hardwired to 1 to indicate a request for I/O space. RO Intel(R) Communications Chipset 89xx Series - Datasheet 513 9.1.1.13 Offset 1Ch: SCNL- Secondary Control Block Base Address Register (SATA--B0:D31:F5) This 4-byte I/O space is used in native mode for the Secondary Controller's Command Block. Table 9-14. Offset 1Ch: Secondary Control Block Base Address Register (IDE B0:D31:F5) Description: View: PCI Size: 32 bit BAR: Configuration Bit Acronym 31 :16 Reserved 15 :02 BA 00 9.1.1.14 B0:D31 :F5 Default: 00000001h Bit Range 01 Bus:Device:Function: Offset Start: 1Ch Offset End: 1Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Base Address -- This field provides the base address of the I/O space (8 consecutive I/O locations). Reserved RW Reserved Resource Type Indicator -- Hardwired to 1 to indicate a request for I/O space. RTE RO Offset 20h: Legacy Bus Master Base Address Register (SATA- B0:D31:F5) The Bus Master IDE interface function uses Base Address register 5 to request a 16byte IO space to provide a software interface to the Bus Master functions. Only 12 bytes are actually used (6 bytes for primary, 6 bytes for secondary). Only bits [15:04] are used to decode the address. Table 9-15. Offset 20h: Legacy Bus Master Base Address Register (SATA-B0:D31:F5) Description: View: PCI Size: 32 bit Default: 00000001h Bit Range Bit Acronym 31 :16 Reserved 15 :05 04 03 :01 00 Bus:Device:Function: B0:D31 :F5 BAR: Configuration Power Well: Bit Description Reserved Sticky Bit Reset Value Bit Access O BA Base Address -- This field provides the base address of the I/O space (8 consecutive I/O locations). RW BA4 Base Address 4 -- When SCC is 01h, this bit will be R/W resulting in requesting 16B of I/O space. RW Reserved RTE Reserved Resource Type Indicator -- Hardwired to 1 to indicate a request for I/O space. Intel(R) Communications Chipset 89xx Series - Datasheet 514 Offset Start: 20h Offset End: 23h O RO October 2012 Order Number: 327879-001US 9.0 9.1.1.15 Offset 24h: SATA Index/Data Pair Base Address Register (SATA- B0:D31:F5) Table 9-16. Offset 24h: SATA Index/Data Pair Base Address Register (SATA-B0:D31:F5) Description: View: PCI Size: 32 bit BAR: Configuration Bit Acronym 31 :16 Reserved 15 :04 BA 03 :01 Reserved 9.1.1.16 B0:D31 :F5 Default: 00000000h Bit Range 00 Bus:Device:Function: Offset Start: 24h Offset End: 27h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Base Address -- Base address of register I/O space RW Reserved Resource Type Indicator -- Hardwired to 1 to indicate a request for I/O space. RTE RO Offset 2Ch: Subsystem Vendor Identification Register (SATA- B0:D31:F5) Table 9-17. Offset 2Ch: Subsystem Vendor Identification Register (SATA-B0:D31:F5) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D31 :F5 Default: 0000h Bit Range Bit Acronym 15 :00 SVID October 2012 Order Number: 327879-001US Offset Start: 2Ch Offset End: 2Dh Power Well: Core Bit Description Sticky Subsystem Vendor ID -- Value is written by BIOS. No hardware action taken on this value. Bit Reset Value Bit Access RWO Intel(R) Communications Chipset 89xx Series - Datasheet 515 9.1.1.17 Offset 2Eh: Subsystem Identification Register (SATA-B0:D31:F5) Table 9-18. Offset 2Eh: Subsystem Identification Register (SATA-B0:D31:F5) Description: View: PCI Size: 16 bit BAR: Configuration B0:D31 :F5 Offset Start: 2Eh Offset End: 2Fh Default: 0000h Bit Range Bit Acronym 15 :00 SID 9.1.1.18 Bus:Device:Function: Power Well: Core Bit Description Bit Reset Value Sticky Bit Access Subsystem ID -- Value is written by BIOS. No hardware action taken on this value. RWO Offset 34h: Capabilities Pointer Register (SATA-B0:D31:F5) Table 9-19. Offset 34h: Capabilities Pointer Register (SATA-B0:D31:F5) Description: View: PCI Size: 8 bit Default: 70h Bit Range Bit Acronym 07 :00 CAP_PTR 9.1.1.19 B0:D31 Bus:Device:Function: :F5 BAR: Configuration Offset Start: 34h Offset End: 34h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Capabilities Pointer -- Indicates that the first capability pointer offset is 70h if the Sub Class Code (SCC) (Dev 31:F2:0Ah) is configure as IDE mode (value of 01). RO Offset 3Ch: Interrupt Line Register (SATA-B0:D31:F5) Table 9-20. Offset 3Ch: Interrupt Line Register (SATA-B0:D31:F5) Description: View: PCI Size: 8 bit B0:D31 Bus:Device:Function: :F5 BAR: Configuration Default: 00h Bit Range Bit Acronym 07 :00 IL Power Well: Bit Description Interrupt Line -- This field is used to communicate to software the interrupt line that the interrupt pin is connected to. These bits are not reset by FLR. Intel(R) Communications Chipset 89xx Series - Datasheet 516 Offset Start: 3Ch Offset End: 3Ch Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 9.0 9.1.1.20 Offset 3Dh: Interrupt Pin Register (SATA-B0:D31:F5) Table 9-21. Offset 3Dh: Interrupt Pin Register (SATA-B0:D31:F5) Description: View: PCI BAR: Configuration Size: 8 bit B0:D31 :F5 Default: See register description Bit Range Bit Acronym 07 :00 IP 9.1.1.21 Bus:Device:Function: Offset Start: 3Dh Offset End: 3Dh Power Well: Bit Description Sticky Bit Reset Value Bit Access Interrupt Pin -- This reflects the value of D31IP.SIP1 (Chipset Config Registers:Offset 3100h:bits 11:8). RO Offset 40h: IDE Timing Register (SATA-B0:D31:F5) Address Offset: Primary: 40h-41h Secondary: 42h-43h Table 9-22. Offset 40h: IDE Timing Register (SATA-B0:D31:F5) Description: View: PCI Size: 16 bit Bit Range 15 14 :00 BAR: Configuration Bus:Device:Function: B0:D31 :F5 Default: 0000h Bit Acronym IDE Offset Start: 40h Offset End: 43h Power Well: Bit Description Sticky IDE Decode Enable -- Individually enable/disable the Primary or Secondary decode. 0 = Disable. 1 = Enables the PCH to decode the associated Command Blocks (1F0-1F7h for primary, 170-177h for secondary) and Control Block (3F6h for primary and 376h for secondary). This bit effects the IDE decode ranges for both legacy and native-Mode decoding. This bit affects SATA operation in both combined and noncombined ATA modes. Bit Reset Value Bit Access RW Reserved Reserved October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 517 9.1.1.22 Offset 48h: Synchronous DMA Control Register (SATA-B0:D31:F5) Address Offset: Primary: 48h Table 9-23. Offset 48h: Synchronous DMA Control Register (SATA-B0:D31:F5) Description: View: PCI BAR: Configuration Size: 8 bit Bus:Device:Function: Default: 00h Bit Acronym Bit Range Offset Start: 48h Offset End: 48h Power Well: Bit Description 07 :04 Reserved IReserved 03 :00 SDMA_CNT Field 1--R/W. This field is R/W to SDMA_C maintain software compatibility. This field has no NT effect on hardware. 9.1.1.23 B0:D3 1:F5 Sticky Bit Reset Value Bit Access RW Offset 4Ah: Synchronous DMA Timing Register (SATA-B0:D31:F5) Address Offset: Primary: 4Ah-4Bh Table 9-24. Offset 4Ah: Synchronous DMA Timing Register (SATA-B0:D31:F5) Description: View: PCI Size: 16 bit Bit Range Bus:Device:Function: B0:D3 1:F5 BAR: Configuration Default: 0000h Bit Acronym Power Well: Bit Description 15 :10 Reserved Reserved 09 :08 Field 2--R/W. This field is R/W to SDMA_TI SDMA_TIM maintain software compatibility. This field has no M2 effect on hardware. 07 :02 Reserved Reserved 01 :00 SDMA_TIM Field 1--R/W. This field is R/W to SDMA_TI maintain software compatibility. This field has no M1 effect on hardware. Intel(R) Communications Chipset 89xx Series - Datasheet 518 Offset Start: 4Ah Offset End: 4Bh Sticky Bit Reset Value Bit Access RW RW October 2012 Order Number: 327879-001US 9.0 9.1.1.24 Offset 54h: IDE I/O Configuration Register (SATA-B0:D31:F5) Address Offset: Primary: 54h-57h Table 9-25. Offset 4Ah: IDE I/O Configuration Register (SATA-B0:D31:F5) Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Bus:Device:Function: Default: 00000000h Bit Acronym Offset Start: 54h Offset End: 57h Power Well: Bit Description Sticky Bit Reset Value 31 :24 Reserved Reserved 23 :16 IDE_CONFIG Field 6--R/W. This field is R/W to IDE_CFG maintain software compatibility. This field has no 6 effect on hardware. 15 Reserved Reserved 14 IDE_CONFIG Field 5--R/W. This field is R/W to IDE_CFG maintain software compatibility. This field has no 5 effect on hardware. 13 Reserved Reserved 12 IDE_CONFIG Field 4--R/W. This field is R/W to IDE_CFG maintain software compatibility. This field has no 4 effect on hardware. 11 :08 Reserved Reserved 07 :04 IDE_CONFIG Field 3--R/W. This field is R/W to IDE_CFG maintain software compatibility. This field has no 3 effect on hardware. 03 Reserved Reserved 02 IDE_CONFIG Field 2--R/W. This field is R/W to IDE_CFG maintain software compatibility. This field has no 2 effect on hardware. 01 Reserved Reserved 00 IDE_CONFIG Field 1--R/W. This field is R/W to IDE_CFG maintain software compatibility. This field has no 1 effect on hardware. October 2012 Order Number: 327879-001US B0:D3 1:F5 Bit Access RW RW RW RW RW RW Intel(R) Communications Chipset 89xx Series - Datasheet 519 9.1.1.25 Offset 70h: PCI Power Management Capability Identification Register (SATA-B0:D31:F5) Table 9-26. Offset 70h: PCI Power Management Capability Identification Register (SATA- B0:D31:F5) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D31 :F5 Default: B001h Offset Start: 70h Offset End: 71h Power Well: Bit Acronym Bit Description 15 :08 NEXT Next Capability -- When SCC is 01h, this field will be B0h indicating the next item is FLR Capability Pointer in the list. RO 07 :00 CID Capability ID -- Indicates that this pointer is a PCI power management. RO 9.1.1.26 Sticky Bit Reset Value Bit Range Bit Access Offset 72h: PC--PCI Power Management Capabilities Register (SATA- B0:D31:F5) Table 9-27. Offset 72h: PC--PCI Power Management Capabilities Register (SATA- B0:D31:F5) Description: View: PCI Size: 16 bit Bus:Device:Function: B0:D31 :F5 BAR: Configuration Default: 4003h Bit Range Bit Acronym 15 :11 PME_SUP Power Well: Bit Description Sticky Bit Reset Value Bit Access PME Support -- By default with SCC = 01h, the default value of 00000 indicates no PME support in IDE mode. RO 10 D2_SUP D2 Support -- Hardwired to 0. The D2 state is not supported RO 09 D1_SUP D1 Support -- Hardwired to 0. The D1 state is not supported RO AUX_CUR Auxiliary Current -- PME# from D3COLD state is not supported, therefore this field is 000b. RO Device Specific Initialization -- Hardwired to 0 to indicate that no device-specific initialization is required. RO 08 :06 05 DSI 04 Reserved Reserved PME_CLK PME Clock -- Hardwired to 0 to indicate that PCI clock is not required to generate PME#. RO VER Version -- Hardwired to 011 to indicates support for Revision 1.2 of the PCI Power Management Specification. RO 03 02 :00 Intel(R) Communications Chipset 89xx Series - Datasheet 520 Offset Start: 72h Offset End: 73h October 2012 Order Number: 327879-001US 9.0 9.1.1.27 Offset 74h: PCI Power Management Control and Status Register (SATA-B0:D31:F5) Table 9-28. Offset 74h: PCI Power Management Control and Status Register (SATA- B0:D31:F5) Description: View: PCI Size: 16 bit Bit Range 15 14 :09 08 07 :04 BAR: Configuration B0:D31 :F5 Default: 0008h Bit Acronym PMES Reserved PMEE Reserved 03 NSFRST 02 Reserved 01 :00 Bus:Device:Function: PS October 2012 Order Number: 327879-001US Offset Start: 74h Offset End: 75h Power Well: Bit Description Sticky PME Status -- Bit is set when a PME event is to be requested, and if this bit and PMEE is set, a PME# will be generated from the SATA controller. Note: When SCC=01h this bit will be RO `0'. Software is advised to clear PMEE together with PMES prior to changing SCC through MAP.SMS. This bit is not reset by Function Level Reset. Bit Reset Value Bit Access RWC Reserved PME Enable -- When SCC is not 01h, this bit R/W. When set, the SATA controller generates PME# form D3HOT on a wake event. Note: When SCC=01h this bit will be RO `0'. Software is advised to clear PMEE together with PMES prior to changing SCC through MAP.SMS. This bit is not reset by Function Level Reset. RW Reserved No Soft Reset -- These bits are used to indicate whether devices transitioning from D3HOT state to D0 state will perform an internal reset. 0 = Device transitioning from D3HOT state to D0 state perform an internal reset. 1 = Device transitioning from D3HOT state to D0 state do not perform an internal reset. Configuration content is preserved. Upon transition from the D3HOT state to D0 state initialized state, no additional operating system intervention is required to preserve configuration context beyond writing to the PowerState bits. Regardless of this bit, the controller transition from D3HOT state to D0 state by a system or bus segment reset will return to the state D0 uninitialized with only PME context preserved if PME is supported and enabled. RO Reserved Power State -- These bits are used both to determine the current power state of the SATA controller and to set a new power state. 00 = D0 state 11 = D3HOT state When in the D3HOT state, the controller's configuration space is available, but the I/O and memory spaces are not. Additionally, interrupts are blocked. RW Intel(R) Communications Chipset 89xx Series - Datasheet 521 9.1.1.28 Offset 90h: MAP--Address Map Register (SATA-B0:D31:F5) Table 9-29. Offset 90h: MAP--Address Map Register (SATA-B0:D31:F5) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: Default: 00h Bit Range Bit Acronym 15 :10 Reserved Bit Description Sticky Bit Reset Value Bit Access Reserved SPD 07 :06 SMS SATA Mode Select -- Software programs these bits to control the mode in which the SATA Controller should operate. 00b = IDE Mode All other combinations are reserved. 05 :02 Reserved 01 :00 MV RWO RW Reserved Map Value -- Reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 522 Offset Start: 90h Offset End: 90h Power Well: SATA Port Disable -- SW programs these bits to disable a SATA port on the controller. 1 = Corresponding port is disabled. [Bit 8 = Port 4 and Bit 9 = Port 5] 0 = Corresponding port is enabled. Note: To ensure a port is properly disabled, BIOS shall configure MAP.SPD first and then configure PCS.PxE. In order to securely prevent port 4 and/or port 5 from being enabled, BIOS shall ensure the following: For Port 4 - B0:D31:F2 MAP[12]= 1 and D31F5 MAP[8]= 1 For Port 5 - B0:D31:F2 MAP[13]= 1 and D31F5 MAP[9]= 1 This field is nor reset by Function Level Reset. 09 :08 B0:D31 :F5 October 2012 Order Number: 327879-001US 9.0 9.1.1.29 Offset 92h: Port Control and Status Register (SATA-B0:D31:F5) By default, the SATA ports are set to the disabled state (bits [5:0] = `0'). When enabled by software, the ports can transition between the on, partial, and slumber states and can detect devices. When disabled, the port is in the "off" state and cannot detect any devices. If an AHCI-aware enabled operating system is being booted then system BIOS shall insure that all supported SATA ports are enabled prior to passing control to the OS. Once the AHCI aware OS is booted it becomes the enabling/disabling policy owner for the individual SATA ports. This is accomplished by manipulating a port's PxSCTL and PxCMD fields. Because an AHCI aware OS will typically not have knowledge of the PxE bits and because the PxE bits act as master on/off switches for the ports, pre-boot software must insure that these bits are set to `1' prior to booting the OS, regardless as to whether or not a device is currently on the port. Table 9-30. Offset 92h: Port Control and Status Register (SATA-B0:D31:F5) Description: View: PCI Size: 16 bit BAR: Configuration Default: 0000h Bit Range Bit Acronym 15 :10 Reserved 09 08 07 :02 01 00 Bus:Device:Function: B0:D31 :F5 Offset Start: 92h Offset End: 93h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved P5P Port 5 Present -- The status of this bit may change at any time. This bit is cleared when the port is disabled via P1E. This bit is not cleared upon surprise removal of a device. 0 = No device detected. 1 = The presence of a device on Port 1 has been detected. RO P4P Port 4 Present -- The status of this bit may change at any time. This bit is cleared when the port is disabled via P0E. This bit is not cleared upon surprise removal of a device. 0 = No device detected. 1 = The presence of a device on Port 0 has been detected. RO Reserved Reserved P5E Port 5 Enabled: 0 = Disabled. The port is in the `off' state and cannot detect any devices. 1 = Enabled. The port can transition between the on, partial, and slumber states and can detect devices. This bit is read-only `0' when MAP.SPD[1]= 1. RW P4E Port 4 Enabled: 0 = Disabled. The port is in the `off' state and cannot detect any devices. 1 = Enabled. The port can transition between the on, partial, and slumber states and can detect devices. This bit is read-only `0' when MAP.SPD[0]= 1. RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 523 9.1.1.30 Offset A8h: SATA Capability Register 0 (SATA-B0:D31:F5) Table 9-31. Offset A8h: SATA Capability Register 0 (SATA-B0:D31:F5) Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B0:D31 :F5 Default: 0010B012h Offset Start: A8h Offset End: ABh Power Well: Bit Acronym 31 :24 Reserved 23 :20 MAJREV Major Revision -- Major revision number of the SATA Capability Pointer implemented. RO 19 :16 MINREV Minor Revision -- Minor revision number of the SATA Capability Pointer implemented. RO 15 :08 NEXT 07 :00 CAP 9.1.1.31 Bit Description Sticky Bit Reset Value Bit Range Bit Access Reserved Next Capability Pointer -- Points to the next capability structure. RWO Capability ID -- The value of 12h has been assigned by the PCI SIG to designate the SATA capability pointer. RO Offset ACh: SATA Capability Register 1 (SATA-B0:D31:F5) Table 9-32. Offset ACh: SATA Capability Register 1 (SATA-B0:D31:F5) Description: View: PCI Size: 32 bit Bus:Device:Function: B0:D31 :F5 BAR: Configuration Default: 00000048h Offset Start: ACh Offset End: AFh Power Well: Bit Acronym 31 :16 Reserved Reserved BAROFST BAR Offset -- Indicates the offset into the BAR where the index/Data pair are located (in DWord granularity). The index and Data I/O registers are located at offset 10h within the I/O space defined by LBAR (BAR4). A value of 004h indicates offset 10h. RO BARLOC BAR Location -- Indicates the absolute PCI Configuration Register address of the BAR containing the Index/Data pair (in DWord granularity). The Index and Data I/O registers reside within the space defined by LBAR (BAR4) in the SATA controller. a value of 8h indicates and offset of 20h, which is LBAR (BAR4). RO 15 :04 03 :00 Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 524 Sticky Bit Reset Value Bit Range Bit Access October 2012 Order Number: 327879-001US 9.0 9.1.1.32 Offset B0h: FLR Capability ID (SATA-B0:D31:F5) Table 9-33. Offset B0h: FLR Capability ID (SATA-B0:D31:F5) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D31 :F5 Default: 0009h Bit Acronym Offset Start: B0h Offset End: B1h Power Well: Bit Description Sticky Bit Reset Value Bit Access 15 :08 Next Capability Pointer -- A value of 00h indicates the final item in the Capability List. RO 07 :00 Capability ID -- The value of this field depends on the FLRCSSECL bit. If FLRCSSEL = 0, this field is 13h If FLRCSSEL = 1, this field is 09h, indicating vendor specific capability. RO 9.1.1.33 Offset B2h: FLR Capability Length and Value (SATA-B0:D31:F5) Address Offset: B2h-B3h Default Value: 2006h Function Level Reset:No (Bits 9:8 only) Attribute: Size: RO, RWO 16 bits When FLRCSSEL = `0', this register is defined as follows: Table 9-34. Offset B2h: FLR Capability Length and Value (SATA-B0:D31:F5) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D31 :F5 Default: 2006h Bit Range Bit Acronym 15 :10 Reserved Offset Start: B2h Offset End: B3h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 09 FLR Capability -- This field indicates support for Function Level Reset. RWO 08 TXP Capability -- This field indicates support for the Transactions Pending (TXP) bit. TXP must be supported if FLR is supported. RWO Capability Length -- This field indicates the number of bytes of the Vendor Specific capability as required by the PCI spec. It has the value of 06h for FLR Capability. RO 07 :00 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 525 When FLRCSSEL = `1', this register is defined as follows: Table 9-35. Offset B2h: FLR Capability Length and Value (SATA-B0:D31:F5) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D31 :F5 Default: 2006h Bit Acronym Offset Start: B2h Offset End: B3h Power Well: Bit Description Sticky Bit Reset Value Bit Access 15 :12 Vendor Specific Capability ID -- A value of 02h identifies this capability as a Function Level Reset. RO 11 :08 Capability Version -- This field indicates the version of the FLR capability. RO 07 :00 Capability Length -- This field indicates the number of bytes of the Vendor Specific capability as required by the PCI spec. It has the value of 06h for FLR Capability. RO 9.1.1.34 Offset B4h: FLR Control (SATA-B0:D31:F5) Table 9-36. Offset B4h: FLR Control (SATA-B0:D31:F5) Description: View: PCI Size: 16 bit Default: 0000h Bit Range Bit Acronym 15 :09 Reserved 08 07 :01 00 Bus:Device:Function: B0:D31 :F5 BAR: Configuration TXP Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Transactions Pending: 0 = Completions for all Non-Posted requests have been received by the controller. 1 = Controller has issued Non-Posted request which has not been completed. RO Reserved Initiate FLR -- Used to initiate FLR transition. A write of `1' indicates FLR transition. Intel(R) Communications Chipset 89xx Series - Datasheet 526 Offset Start: B4h Offset End: B5h RW October 2012 Order Number: 327879-001US 9.0 9.1.1.35 Offset C0h: APM Trapping Control Register (SATA-B0:D31:F5) This SATA controller does not support legacy I/O access. Therefore, this register is reserved. Software shall not change the default values of the register; otherwise the result will be undefined. . Table 9-37. Offset C0h: APM Trapping Control Register (SATA-B0:D31:F5) Description: View: PCI Size: 8 bit BAR: Configuration B0:D31 :F5 Default: 00h Bit Range Bit Acronym 07 :00 Reserved 9.1.1.36 Bus:Device:Function: Offset Start: C0h Offset End: C0h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Offset C4h: APM Trapping Control Register (SATA-B0:D31:F5) This SATA controller does not support legacy I/O access. Therefore, this register is reserved. Software shall not change the default values of the register; otherwise the result will be undefined. . Table 9-38. Offset C4h: APM Trapping Control Register (SATA-B0:D31:F5) Description: View: PCI Size: 8 bit BAR: Configuration B0:D31 Bus:Device:Function: :F5 Default: 00h Bit Range Bit Acronym 07 :00 Reserved October 2012 Order Number: 327879-001US Offset Start: C4h Offset End: C4h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 527 9.1.2 Bus Master IDE I/O Registers (B0:D31:F5) The bus master IDE function uses 16 bytes of I/O space, allocated via the LBAR register, located in Bus 0:Device 31:Function 5 Configuration space, offset 20h. All bus master IDE I/O space registers can be accessed as byte, word, or dword quantities. Reading reserved bits returns an indeterminate, inconsistent value, and writes to reserved bits have no affect (but should not be attempted). These registers are only used for legacy (IDE) operation. Software must not use these registers when running AHCI. The description of the I/O registers is shown in the following table. Table 9-39. Bus Master IDE I/O Registers (B0:D31:F5) Offset Start Offset End Default Value Register ID - Description 00h 08h "Offset 00h: Bus Master IDE Command Register (B0:D31:F5)" on page 528 00h 02h 0Ah "Offset 02h: Bus Master IDE Status Register (B0:D31:F5)" on page 530 00h 0Ch 0Fh "Offset 04h: Bus Master IDE Descriptor Table Pointer Register (B0:D31:F5)" on page 531 00h 9.1.2.1 Offset 00h: Bus Master IDE Command Register (B0:D31:F5) Address Offset: Primary: BAR + 00hSecondary: BAR + 08h Table 9-40. Offset 00h: Bus Master IDE Command Register (B0:D31:F5) Description: View: PCI Size: 8 bit Bit Range 07 :00 Bus:Device:Function: B0:D31 :F5 BAR: LBAR (IO) Default: 00h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 528 Offset Start: 00h Offset End: 08h October 2012 Order Number: 327879-001US 9.0 Table 9-40. Offset 00h: Bus Master IDE Command Register (B0:D31:F5) Description: View: PCI Size: 8 bit Bit Range :03 02 :01 :00 BAR: LBAR (IO) Bus:Device:Function: B0:D31 Offset Start: 00h :F5 Offset End: 08h Default: 00h Bit Acronym R/WC Power Well: Bit Description Sticky Read / Write Control -- This bit sets the direction of the bus master transfer: This bit must NOT be changed when the bus master function is active. 0 = Memory reads 1 = Memory writes Bit Reset Value Bit Access RW Reserved Reserved START Start/Stop Bus Master: 0 = All state information is lost when this bit is cleared. Master mode operation cannot be stopped and then resumed. If this bit is reset while bus master operation is still active (i.e., the Bus Master IDE Active bit (B0:D31:F5:BAR + 02h, bit 0) of the Bus Master IDE Status register for that IDE channel is set) and the drive has not yet finished its data transfer (the Interrupt bit in the Bus Master IDE Status register for that IDE channel is not set), the bus master command is said to be aborted and data transferred from the drive may be discarded instead of being written to system memory. 1 = Enables bus master operation of the controller. Bus master operation does not actually start unless the Bus Master Enable bit (D31:F1:04h, bit 2) in PCI configuration space is also set. Bus master operation begins when this bit is detected changing from 0 to 1. The controller will transfer data between the IDE device and memory only when this bit is set. Master operation can be halted by writing a 0 to this bit. RW This bit is intended to be cleared by software after the data transfer is completed, as indicated by either the Bus Master IDE Active bit being cleared or the Interrupt bit of the Bus Master IDE Status register for that IDE channel being set, or both. Hardware does not clear this bit automatically. If this bit is cleared to 0 prior to the DMA data transfer being initiated by the drive in a device to memory data transfer, then the PCH will not send DMAT to terminate the data transfer. SW intervention (e.g. sending SRST) is required to reset the interface in this condition. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 529 9.1.2.2 Offset 02h: Bus Master IDE Status Register (B0:D31:F5) Address Offset: Primary: BAR + 02hSecondary: BAR + 0Ah Table 9-41. Offset 02h: Bus Master IDE Status Register (B0:D31:F5) Description: View: PCI Size: 8 bit Bit Range BAR: LBAR (IO) Bus:Device:Function: Default: 00h Bit Description 07 PRDIS PRD Interrupt Status: 0 = Software clears this bit by writing a 1 to it. 1 = This bit is set when the host controller execution of a PRD that has its PRD_INT bit set. 06 Reserved Reserved Bit Reset Value Bit Access RWC RW Reserved 02 Interrupt: 0 = Software clears this bit by writing a 1 to it. 1 = Set when a device FIS is received with the `I' bit set, provided that software has not disabled interrupts via the IEN bit of the Device Control Register (see chapter 5 of the Serial ATA Specification, Revision 1.0a). RWC 01 Error: 0 = Software clears this bit by writing a 1 to it. 1 = This bit is set when the controller encounters a target abort or master abort when transferring data on PCI. RWC 00 Bus Master IDE Active: 0 = This bit is cleared by the PCH when the last transfer for a region is performed, where EOT for that region is set in the region descriptor. It is also cleared by the PCH when the Start Bus Master bit (B0:D31:F5:BAR+ 00h, bit 0) is cleared in the Command register. When this bit is read as a 0, all data transferred from the drive during the previous bus master command is visible in system memory, unless the bus master command was aborted. 1 = Set by the PCH when the Start bit is written to the Command register. RO ACT Intel(R) Communications Chipset 89xx Series - Datasheet 530 Sticky Reserved Drive 0 DMA Capable: 0 = Not Capable 1 = Capable. Set by device dependent code (BIOS or device driver) to indicate that drive 0 for this channel is capable of DMA transfers, and that the controller has been initialized for optimum performance. The PCH does not use this bit. It is intended for systems that do not attach BMIDE to the PCI bus. 05 Offset Start: 02h Offset End: 0Ah Power Well: Bit Acronym 04 :03 B0:D31 :F5 October 2012 Order Number: 327879-001US 9.0 9.1.2.3 Offset 04h: Bus Master IDE Descriptor Table Pointer Register (B0:D31:F5) Address Offset: Primary: BAR + 04h-07hSecondary: BAR + 0Ch-0Fh Table 9-42. Offset 04h: Bus Master IDE Descriptor Table Pointer Register (B0:D31:F5) Description: View: PCI BAR: LBAR (IO) Size: 32 bit B0:D31 :F5 Default: 00h Bit Range Bit Acronym 31 :02 ADDR 01 :00 Reserved 9.1.3 Bus:Device:Function: Primary 04h Offset Start: 07h Offset End: Secondary Offset Start: 0Ch Offset End: 0Fh Power Well: Bit Description Sticky Bit Reset Value Address of Descriptor Table -- The bits in this field correspond to bits [31:2] of the memory location of the Physical Region Descriptor (PRD). The Descriptor Table must be dword-aligned. The Descriptor Table must not cross a 64-K boundary in memory. Bit Access RW Reserved Serial ATA Index/Data Pair Superset Registers All of these I/O registers are in the core well. They are exposed only when SCC is 01h (i.e. IDE programming interface) and the controller is not in combined mode. These are Index/Data Pair registers that are used to access the SerialATA superset registers (SerialATA Status, SerialATA Control and SerialATA Error). The I/O space for these registers is allocated through SIDPBA. Locations with offset from 08h to 0Fh are reserved for future expansion. Software-write operations to the reserved locations shall have no effect while software-read operations to the reserved locations shall return 0. Table 9-43. Serial ATA Index/Data Pair Superset Registers Offset Start 00h Offset End 03h Register ID - Description "Offset 00h: SINDX--SATA Index Register (B0:D31:F5)" on page 532 Default Value 00000000h 04h 07h "Offset 04h: SDATA--SATA Index Data Register (B0:D31:F5)" on page 532 00000000h 04h 07h "Offset 04h: PxSSTS--Serial ATA Status Register (B0:D31:F5)" on page 533 00000000h 04h 07h "Offset 04h: PxSCTL--Serial ATA Control Register (B0:D31:F5)" on page 535 00000004h 04h 07h "Offset 04h: PxSERR--Serial ATA Error Register (B0:D31:F5)" on page 537 00000000h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 531 9.1.3.1 Offset 00h: SINDX--SATA Index Register (B0:D31:F5) Table 9-44. Offset 00h: SINDX--SATA Index Register (B0:D31:F5) Description: View: PCI Size: 32 bit BAR: SIDPBA Bit Acronym 31 :16 Reserved 07 :00 9.1.3.2 B0:D31 :F5 Default: 00000000h Bit Range 15 :08 Bus:Device:Function: Offset Start: 00h Offset End: 03h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved PIDX Port Index -- This Index field is used to specify the port of the SATA controller at which the port-specific SSTS, SCTL, and SERR registers are located. 00h = Primary Master (Port 4) 02h = Secondary Master (Port 5) All other values are Reserved. RW RIDX Register Index -- This Index field is used to specify one out of three registers currently being indexed into. 00h = SSTS 01h = SCTL 02h = SERR All other values are Reserved RW Offset 04h: SDATA--SATA Index Data Register (B0:D31:F5) Table 9-45. Offset 04h: SDATA--SATA Index Data Register (B0:D31:F5) Description: View: PCI Size: 32 bit Bit Range 31 :00 B0:D31 Bus:Device:Function: :F5 BAR SIDPBA Default: 00000000h Power Well: Bit Acronym Bit Description DATA Data -- This Data register is a "window" through which data is read or written to the memory mapped registers. A read or write to this Data register triggers a corresponding read or write to the memory mapped register pointed to by the Index register. The Index register must be setup prior to the read or write to this Data register. A physical register is not actually implemented as the data is actually stored in the memory mapped registers. Since this is not a physical register, the "default" value is the same as the default value of the register pointed to by Index. Intel(R) Communications Chipset 89xx Series - Datasheet 532 Offset Start: 04h Offset End: 07h Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 9.0 9.1.3.3 Offset 04h: PxSSTS--Serial ATA Status Register (B0:D31:F5) SDATA when SINDX.RIDX is 00h. This is a 32-bit register that conveys the current state of the interface and host. The PCH updates it continuously and asynchronously. When the PCH transmits a COMRESET to the device, this register is updated to its reset values. Table 9-46. Offset 04h: PxSSTS--Serial ATA Status Register (B0:D31:F5) (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR SIDPBA Bus:Device:Function: B0:D31 :F5 Default: 00000000h Bit Range Bit Acronym 31 :12 Reserved Offset Start: 04h Offset End: 07h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Interface Power Management -- Indicates the current interface state: 11 :08 IPM Value Description 0h Device not present or communication not established 1h Interface in active state 2h Interface in PARTIAL power management state 6h Interface in SLUMBER power management state RO All other values reserved. Current Interface Speed -- Indicates the negotiated interface communication speed. 07 :04 SPD Value Description 0h Device not present or communication not established 1h Generation 1 communication rate negotiated 2h Generation 2 communication rate negotiated RO All other values reserved. The PCH Supports Gen 1 communication rates (1.5 Gb/s) and Gen 2 rates (3.0 Gb/s). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 533 Table 9-46. Offset 04h: PxSSTS--Serial ATA Status Register (B0:D31:F5) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR SIDPBA Bus:Device:Function: B0:D31 :F5 Default: 00000000h Bit Acronym Offset Start: 04h Offset End: 07h Power Well: Bit Description Sticky Bit Reset Value Bit Access Device Detection -- Indicates the interface device detection and Phy state: 03 :00 DET Value Description 0h No device detected and Phy communication not established 1h Device presence detected but Phy communication not established 3h Device presence detected and Phy communication established 4h Phy in offline mode as a result of the interface being disabled or running in a BIST loopback mode RO All other values reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 534 October 2012 Order Number: 327879-001US 9.0 9.1.3.4 Offset 04h: PxSCTL--Serial ATA Control Register (B0:D31:F5) SDATA when SINDX.RIDX is 01h. This is a 32-bit read-write register by which software controls SATA capabilities. Writes to the SControl register result in an action being taken by the PCH or the interface. Reads from the register return the last value written to it. Table 9-47. Offset 04h: PxSCTL--Serial ATA Control Register (B0:D31:F5) (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR SIDPBA Bus:Device:Function: B0:D31 :F5 Default: 00000004h Offset Start: 04h Offset End: 07h Power Well: Bit Acronym 31 :20 Reserved 19 :16 PMP Port Multiplier Port -- This field is not used by AHCI. RO SPM Select Power Management -- This field is not used by AHCI. RO 15 :12 Bit Description Sticky Bit Reset Value Bit Range Bit Access Reserved Interface Power Management Transitions Allowed -- Indicates which power states the PCH is allowed to transition to: 11 :08 IPM Value Description 0h No interface restrictions 1h Transitions to the PARTIAL state disabled 2h Transitions to the SLUMBER state disabled 3h Transitions to both PARTIAL and SLUMBER states disabled RW All other values reserved Speed Allowed -- Indicates the highest allowable speed of the interface. This speed is limited by the CAP.ISS (ABAR+00h:bit 23:20) field. Value 07 :04 SPD Description 0h No speed negotiation restrictions 1h Limit speed negotiation to Generation 1 communication rate 2h Limit speed negotiation to Generation 2 communication rate RW All other values reserved. The PCH Supports Gen 1 communication rates (1.5 Gb/s) and Gen 2 rates (3.0 Gb/s). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 535 Table 9-47. Offset 04h: PxSCTL--Serial ATA Control Register (B0:D31:F5) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR SIDPBA Bus:Device:Function: B0:D31 :F5 Default: 00000004h Bit Acronym Offset Start: 04h Offset End: 07h Power Well: Bit Description Sticky Bit Reset Value Bit Access Device Detection Initialization -- Controls the PCH's device detection and interface initialization. Value 03 :00 Description 0h No device detection or initialization action requested 1h Perform interface communication initialization sequence to establish communication. This is functionally equivalent to a hard reset and results in the interface being reset and communications reinitialized DET 4h RW Disable the Serial ATA interface and put Phy in offline mode All other values reserved. When this field is written to a 1h, the PCH initiates COMRESET and starts the initialization process. When the initialization is complete, this field shall remain 1h until set to another value by software. This field may only be changed to 1h or 4h when PxCMD.ST is 0. Changing this field while the PCH is running results in undefined behavior. Intel(R) Communications Chipset 89xx Series - Datasheet 536 October 2012 Order Number: 327879-001US 9.0 9.1.3.5 Offset 04h: PxSERR--Serial ATA Error Register (B0:D31:F5) SDATA when SINDx.RIDX is 02h. Bits 26:16 of this register contains diagnostic error information for use by diagnostic software in validating correct operation or isolating failure modes. Bits 11:0 contain error information used by host software in determining the appropriate response to the error condition. If one or more of bits 11:8 of this register are set, the controller will stop the current transfer. Table 9-48. Offset 04h: PxSERR--Serial ATA Error Register (B0:D31:F5) (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR SIDPBA Bus:Device:Function: B0:D31 :F5 Default: 00000000h Bit Range Bit Acronym 31 :27 Reserved Offset Start: 04h Offset End: 07h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 26 X Exchanged -- When set to one this bit indicates that a change in device presence has been detected since the last time this bit was cleared. This bit shall always be set to 1 anytime a COMINIT signal is received. This bit is reflected in the P0IS.PCS bit. 25 F Unrecognized FIS Type -- Indicates that one or more FISs were received by the Transport layer with good CRC, but had a type field that was not recognized. RWC 24 T Transport state transition error -- Indicates that an error has occurred in the transition from one state to another within the Transport layer since the last time this bit was cleared. RWC 23 T Transport state transition error -- Indicates that an error has occurred in the transition from one state to another within the Transport layer since the last time this bit was cleared. RWC 22 H Handshake -- indicates that one or more R_ERR handshake response was received in response to frame transmission. Such errors may be the result of a CRC error detected by the recipient, a disparity or 8b/10b decoding error, or other error condition leading to a negative handshake on a transmitted frame. RWC 21 C CRC Error -- Indicates that one or more CRC errors occurred with the Link Layer. RWC 20 D Disparity Error -- This field is not used by AHCI. RWC 19 B 10b to 8b Decode Error -- Indicates that one or more 10b to 8b decoding errors occurred. RWC 18 W Comm Wake -- Indicates that a Comm Wake signal was detected by the Phy. RWC 17 I Phy Internal Error -- Indicates that the Phy detected some internal error. RWC N PhyRdy Change -- When set to 1 this bit indicates that the internal PhyRdy signal changed state since the last time this bit was cleared. In the PCH, this bit will be set when PhyRdy changes from a 0 -> 1 or a 1 -> 0. The state of this bit is then reflected in the PxIS.PRCS interrupt status bit and an interrupt will be generated if enabled. Software clears this bit by writing a 1 to it. RWC 16 15 :12 Reserved October 2012 Order Number: 327879-001US RWC Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 537 Table 9-48. Offset 04h: PxSERR--Serial ATA Error Register (B0:D31:F5) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR SIDPBA Bus:Device:Function: B0:D31 :F5 Default: 00000000h Offset Start: 04h Offset End: 07h Power Well: Bit Description 11 E Internal Error -- The SATA controller failed due to a master or target abort when attempting to access system memory. RWC 10 P Protocol Error -- A violation of the Serial ATA protocol was detected. Note: The PCH does not set this bit for all protocol violations that may occur on the SATA link. RWC C Persistent Communication or Data Integrity Error -- A communication error that was not recovered occurred that is expected to be persistent. Persistent communications errors may arise from faulty interconnect with the device, from a device that has been removed or has failed, or a number of other causes. RWC T Transient Data Integrity Error -- A data integrity error occurred that was not recovered by the interface. RWC 09 07 :02 01 00 Reserved M Reserved Sticky Bit Reset Value Bit Acronym Bit Access O Recovered Communications Error -- Communications between the device and host was temporarily lost but was re-established. This can arise from a device temporarily being removed, from a temporary loss of Phy synchronization, or from other causes and may be derived from the PhyNRdy signal between the Phy and Link layers. RWC Intel(R) Communications Chipset 89xx Series - Datasheet 538 October 2012 Order Number: 327879-001US 10.0 10.0 EHCI Controller Registers (B0:D29:F0) 10.1 USB EHCI Configuration Registers (USB EHCI--B0:D29:F0) Register address locations that are not shown in the following table should be treated as Reserved. 10.1.1 USB EHCI PCI Register Address Map Table 10-1. USB EHCI PCI Register Address Map View Offset Start Offset End Register ID: Register Name Default Value BAR BDF PCI Configura tion B0:D29:F0 00h 01h "Offset 00h: VID: Vendor Identification Register (USB EHCI--B0:D29:F0)" on page 542 8086h PCI Configura tion B0:D29:F0 02h 03h "Offset 02h: DID--Device Identification Register (USB EHCI--B0:D29:F0)" on page 542 See register description PCI Configura tion B0:D29:F0 04h 05h "Offset 04h: PCI Command Register (USB EHCI-- B0:D29:F0)" on page 543 0000h PCI Configura tion B0:D29:F0 06h 07h "Offset 06h: PCI Status Register (USB EHCI-- B0:D29:F0)" on page 545 0290h PCI Configura tion B0:D29:F0 08h 08h "Offset 08h: RID--Revision Identification Register (USB EHCI--B0:D29:F0)" on page 546 See register description PCI Configura tion B0:D29:F0 09h 09h "Offset 09h: Programming Interface Register (USB EHCI--B0:D29:F0)" on page 546 20h PCI Configura tion B0:D29:F0 0Ah 0Ah "Offset 0Ah: Sub Class Code Register (USB EHCI-- B0:D29:F0)" on page 546 03h PCI Configura tion B0:D29:F0 0Bh 0Bh "Offset 0Bh: BCC--Base Class Code Register (USB EHCI--B0:D29:F0)" on page 547 0Ch PCI Configura tion B0:D29:F0 0Dh 0Dh "Offset 0Dh: Primary Master Latency Timer Register (USB EHCI--B0:D29:F0)" on page 547 00h PCI Configura tion B0:D29:F0 0Eh 0Eh "Offset 0Eh: Header Type Register (USB EHCI-- B0:D29:F0)" on page 548 80h PCI Configura tion B0:D29:F0 10h 13h "Offset 10h: Memory Base Address Register (USB EHCI--B0:D29:F0)" on page 548 00000000h PCI Configura tion B0:D29:F0 2Ch 2Dh "Offset 2Ch: Subsystem Vendor ID Register (USB EHCI--B0:D29:F0)" on page 549 XXXXh PCI Configura tion B0:D29:F0 2Eh 2Fh "Offset 2Eh: Subsystem ID Register (USB EHCI-- B0:D29:F0)" on page 549 XXXXh PCI Configura tion B0:D29:F0 34h 34h "Offset 34h: Capabilities Pointer Register (USB EHCI--B0:D29:F0)" on page 550 50h PCI Configura tion B0:D29:F0 3Ch 3Ch "Offset 3Ch: Interrupt Line Register (USB EHCI-- B0:D29:F0)" on page 550 00h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 539 Table 10-1. USB EHCI PCI Register Address Map View Offset Start Offset End BDF PCI Configura tion B0:D29:F0 3Dh 3Dh "Offset 3Dh: Interrupt Pin Register (USB EHCI-- B0:D29:F0)" on page 550 See register description PCI Configura tion B0:D29:F0 50h 50h "Offset 50h: PCI Power Management Capability ID Register (USB EHCI--B0:D29:F0)" on page 551 01h PCI Configura tion B0:D29:F0 51h 51h "Offset 51h: Next Item Pointer #1 Register (USB EHCI--B0:D29:F0)" on page 551 58h PCI Configura tion B0:D29:F0 52h 53h "Offset 52h: Power Management Capabilities Register (USB EHCI--B0:D29:F0)" on page 552 C9C2h PCI Configura tion B0:D29:F0 54h 55h "Offset 54h: Power Management Control/Status Register (USB EHCI--B0:D29:F0)" on page 553 0000h PCI Configura tion B0:D29:F0 58h 58h "Offset 58h: Debug Port Capability ID Register (USB EHCI--B0:D29:F0)" on page 554 0Ah PCI Configura tion B0:D29:F0 59h 59h "Offset 59h: Next Item Pointer #2 Register (USB EHCI--B0:D29:F0)" on page 554 98h PCI Configura tion B0:D29:F0 5Ah 5Bh "Offset 5Ah: Debug Port Base Offset Register (USB EHCI--B0:D29:F0)" on page 554 20A0h PCI Configura tion B0:D29:F0 60h 60h "Offset 60h: USB Release Number Register (USB EHCI--B0:D29:F0)" on page 555 20h PCI Configura tion B0:D29:F0 61h 61h "Offset 61h: Frame Length Adjustment Register (USB EHCI--B0:D29:F0)" on page 556 20h PCI Configura tion B0:D29:F0 62h 63h "Offset 62h: Port Wake Capability Register (USB EHCI--B0:D29:F0)" on page 557 01FFh PCI Configura tion B0:D29:F0 68h 6Bh "Offset 68h: Legacy Support Extended Capability Register (USB EHCI--B0:D29:F0)" on page 558 00000001h PCI Configura tion B0:D29:F0 6Ch 6Fh "Offset 6Ch: Legacy Support Extended Control/ Status Register (USB EHCI--B0:D29:F0)" on page 559 00000000h PCI Configura tion B0:D29:F0 70h 73h "Offset 70h: SPECIAL_SMI--Intel Specific USB 2.0 SMI Register (USB EHCI--B0:D29:F0)" on page 562 00000000h PCI Configura tion B0:D29:F0 80h 80h "Offset 80h: ACCESS_CNTL--Access Control Register (USB EHCI--B0:D29:F0)" on page 564 00h PCI Configura tion B0:D29:F0 84h 84h "Offset 84h: EHCIIR1--EHCI Initialization Register 1 (USB EHCI--B0:D29:F0)" on page 564 01h PCI Configura tion B0:D29:F0 98h 98h "Offset 98h: FLR_CID--Function Level Reset Capability ID (USB EHCI--B0:D29:F0)" on page 565 09h PCI Configura tion B0:D29:F0 99h 99h "Offset 99h: FLR_NEXT--Function Level Reset Next Capability Pointer (USB EHCI--B0:D29:F0)" on page 565 00h PCI Configura tion B0:D29:F0 9Ah 9Bh "Offset 9Ah: FLR_CLV--Function Level Reset Capability Length and Version (USB EHCI-- B0:D29:F0)" on page 566 2006h PCI Configura tion B0:D29:F0 9Ah 9Bh "Offset 9Ah: FLR_CLV--Function Level Reset Capability Length and Version (USB EHCI-- B0:D29:F0)" on page 566 2006h PCI Configura tion B0:D29:F0 9Ch 9Ch "Offset 9Ch: FLR_CTRL--Function Level Reset Control Register (USB EHCI--B0:D29:F0)" on page 567 00h PCI Configura tion B0:D29:F0 9Dh 9Dh "Offset 9Dh: FLR_STS--Function Level Reset Status Register (USB EHCI--B0:D29:F0)" on page 567 00h PCI MBAR B0:D29:F0 00h 00h "Offset 00h: CAPLENGTH--Capability Registers Length (USB EHCI--B0:D29:F0)" on page 569 20h Intel(R) Communications Chipset 89xx Series - Datasheet 540 Register ID: Register Name Default Value BAR October 2012 Order Number: 327879-001US 10.0 Table 10-1. USB EHCI PCI Register Address Map View BAR BDF Offset Start Offset End Register ID: Register Name Default Value PCI MBAR B0:D29:F0 02h 03h "Offset 02h: HCIVERSION--Host Controller Interface Version Number Register (USB EHCI-- B0:D29:F0)" on page 569 0100h PCI MBAR B0:D29:F0 04h 07h "Offset 04h: HCSPARAMS--Host Controller Structural Parameters (USB EHCI--B0:D29:F0)" on page 570 00204208h PCI MBAR B0:D29:F0 08h 0Bh "Offset 08h: HCCPARAMS--Host Controller Capability Parameters Register (USB EHCI-- B0:D29:F0)" on page 571 00006881h PCI MBAR B0:D29:F0 20h 23h "Offset 20h: USB2.0_CMD--USB 2.0 Command Register (USB EHCI--B0:D29:F0)" on page 573 00080000h PCI MBAR B0:D29:F0 24h 27h "Offset 24h: USB2.0_STS--USB 2.0 Status Register (USB EHCI--B0:D29:F0)" on page 576 00001000h PCI MBAR B0:D29:F0 28h 2Bh "Offset 28h: USB2.0_INTR--USB 2.0 Interrupt Enable Register (USB EHCI--B0:D29:F0)" on page 579 00000000h PCI MBAR B0:D29:F0 2Ch 2Fh "Offset 2Ch: FRINDEX--Frame Index Register (USB EHCI--B0:D29:F0)" on page 581 00000000h PCI MBAR B0:D29:F0 30h 33h "Offset 30h: CTRLDSSEGMENT--Control Data Structure Segment Register (USB EHCI-- B0:D29:F0)" on page 582 00000000h PCI MBAR B0:D29:F0 34h 37h "Offset 34h: PERIODICLISTBASE--Periodic Frame List Base Address Register (USB EHCI-- B0:D29:F0)" on page 583 00000000h PCI MBAR B0:D29:F0 38h 3Bh "Offset 38h: ASYNCLISTADDR--Current Asynchronous List Address Register (USB EHCI-- B0:D29:F0)" on page 584 00000000h PCI MBAR B0:D29:F0 60h 63h "Offset 60h: CONFIGFLAG--Configure Flag Register (USB EHCI--B0:D29:F0)" on page 585 00000000h PCI MBAR B0:D29:F0 64h 67h "Offset 64h: PORTSC--Port N Status and Control Register (USB EHCI--B0:D29:F0)" on page 586 00003000h PCI MBAR B0:D29:F0 A0h A0h "Offset A0h: CNTL_STS--Control/Status Register (USB EHCI--B0:D29:F0)" on page 592 20h PCI MBAR B0:D29:F0 A4h A7h "Offset A4h: USBPID--USB PIDs Register (USB EHCI--B0:D29:F0)" on page 594 00000000h PCI MBAR B0:D29:F0 A8h AFh "Offset A8h: DATABUF[7:0]--Data Buffer Bytes[7:0] Register (USB EHCI--B0:D29:F0)" on page 595 000000000 0000000h PCI MBAR B0:D29:F0 B0h B3h "Offset B0h: CONFIG--Configuration Register (USB EHCI--B0:D29:F0)" on page 595 00007F01h Note: All configuration registers in this section are in the core well and reset by a core well reset and the D3-to-D0 warm reset, except as noted. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 541 10.1.1.1 Offset 00h: Vendor Identification Register (USB EHCI--B0:D29:F0) Table 10-2. Offset 00h: VID: Vendor Identification Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit BAR: Configuration B0:D29 :F0 Default: 8086h Bit Range Bit Acronym 15 :00 VID 10.1.1.2 Bus:Device:Function: Offset Start: 00h Offset End: 01h Power Well: Core Bit Description Sticky Vendor Identification: This 16-bit value is assigned to Intel. Intel VID = 8086h Bit Reset Value Bit Access 8086h RO Offset 02h: Device Identification Register (USB EHCI--B0:D29:F0) Table 10-3. Offset 02h: DID--Device Identification Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit Bus:Device:Function: B0:D29 :F0 BAR: Configuration Default: See register description Bit Range Bit Acronym 15 :00 DID Bit Description Device ID -- This is a 16-bit value assigned to the PCH USB controller. Intel(R) Communications Chipset 89xx Series - Datasheet 542 Offset Start: 02h Offset End: 03h Power Well: Core Sticky Bit Reset Value Bit Access RO October 2012 Order Number: 327879-001US 10.0 10.1.1.3 Offset 04h: PCI Command Register (USB EHCI--B0:D29:F0) Table 10-4. Offset 04h: PCI Command Register (Sheet 1 of 2) (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: 0000h Bit Range Bit Acronym 15 :11 Reserved 10 INTx_Disable 09 FBE 08 SERR_EN 07 WCC Offset Start: 04h Offset End: 05h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00h Interrupt Disable: 0 = The function is capable of generating interrupts. 1 = The function can not generate its interrupt to the interrupt controller. The corresponding Interrupt Status bit (B0:D29:F0:06h, bit 3) is not affected by the interrupt enable. 0h RW Fast Back to Back Enable -- Hardwired to 0. 0h RO SERR# Enable: 0 = Disables EHC's capability to generate an SERR#. 1 = The Enhanced Host controller (EHC) is capable of generating (internally) SERR# in the following cases: - When it receive a completion status other than "successful" for one of its DMA initiated memory reads on DMI (and subsequently on its internal interface). - When it detects an address or command parity error and the Parity Error Response bit is set. - When it detects a data parity error (when the data is going into the EHC) and the Parity Error Response bit is set. 0h RW Wait Cycle Control -- Hardwired to 0. 0h RO 0h RW 0h RO Parity Error Response: 0 = The EHC is not checking for correct parity (on its internal interface). 1 = The EHC is checking for correct parity (on its internal interface) and halt operation when bad parity is detected during the data phase. 06 PER NOTE: If the EHC detects bad parity on the address or command phases when the bit is set to 1, the host controller does not take the cycle. It halts the host controller (if currently not halted) and sets the Host System Error bit in the USBSTS register. This applies to both requests and completions from the system interface. This bit must be set in order for the parity errors to generate SERR#. 05 VPS 04 PMWE 03 SCE October 2012 Order Number: 327879-001US VGA Palette Snoop -- Hardwired to 0. Postable Memory Write Enable -- Hardwired to 0. 0h RO Special Cycle Enable -- Hardwired to 0. 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 543 Table 10-4. Offset 04h: PCI Command Register (Sheet 2 of 2) (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: Default: 0000h Offset Start: 04h Offset End: 05h Power Well: Core Bit Reset Value Bit Access BME Bus Master Enable: 0 = Disables this functionality. 1 = Enables the PCH to act as a master on the PCI bus for USB transfers. 1 RW 01 MSE Memory Space Enable -- This bit controls access to the USB 2.0 Memory Space registers. 0 = Disables this functionality. 1 = Enables accesses to the USB 2.0 registers. The Base Address register (B0:D29:F0:F0:10h) for USB 2.0 should be programmed before this bit is set. 1 RW 00 IOSE I/O Space Enable -- Hardwired to 0. 1 RO Bit Range 02 Bit Acronym Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 544 B0:D29 :F0 Sticky October 2012 Order Number: 327879-001US 10.0 10.1.1.4 Offset 06h: PCI Status Register (USB EHCI--B0:D29:F0) Table 10-5. Offset 06h: PCI Status Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D29 Offset Start: 06h :F0 Offset End: 07h Default: 0290h Power Well: Bit Reset Value Bit Access DPE Detected Parity Error 0 = Parity error not detected. 1 = Indicates that the PCH detected a parity error on the internal backbone. This bit gets set even if the Parity Error Response bit (B0:D29:F0; Offset 04h bit 6) is not set. 0h RWC 14 SSE Signaled System Error: 0 = No SERR# signaled by the PCH. 1 = This bit is set by the PCH when it signals SERR# (internally). The SER_EN bit (bit 8 of the Command Register) must be 1 for this bit to be set. 0h RWC 13 RMA Received Master Abort: 0 = No master abort received. 1 = Set when the bridge receives a master abort status from the I/ O data bus. 0h RWC 12 RTA Received Target Abort: 0 = No target abort received by EHC on memory access. 1 = This bit is set when EHC, as a master, receives a target abort status on a memory access. This is treated as a Host Error and halts the DMA engines. This event can optionally generate an SERR# by setting the SERR# Enable bit (B0:D29:F0:F0:04h, bit 8). 11 STA Signaled Target Abort -- This bit is used to indicate when the EHCI function responds to a cycle with a target abort. There is no reason for this to happen, so this bit is hardwired to 0. Bit Range 15 10 :09 Bit Acronym DEVT_STS Bit Description Sticky DEVSEL# Timing Status -- This 2-bit field defines the timing for DEVSEL# assertion. 08 DPED Master Data Parity Error Detected: 0 = No data parity error detected on USB2.0 read completion packet. 1 = This bit is set by the PCH when a data parity error is detected on a USB 2.0 read completion packet on the internal interface to the EHCI host controller and bit 6 of the Command register is set to 1. 07 FB2BC Fast Back to Back Capable -- Hardwired to 1. 06 05 04 03 02 :00 UDF User Definable Features -- Hardwired to 0. 66 MHz _CAP 66 MHz Capable -- Hardwired to 0. CAP_LIST IS Reserved Capabilities List -- Hardwired to 1 indicating that offset 34h contains a valid capabilities pointer. Interrupt Status -- This bit reflects the state of this function's interrupt at the input of the enable/disable logic. 0 = This bit will be 0 when the interrupt is deasserted. 1 = This bit is a 1 when the interrupt is asserted. The value reported in this bit is independent of the value in the Interrupt Enable bit. RWC 0h RO 00h RO 0h RWC 0h RO 0h RO 0h RO 1 RO 0h RO Reserved October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 545 10.1.1.5 Offset 08h: RID--Revision Identification Register (USB EHCI-- B0:D29:F0) Table 10-6. Offset 08h: RID--Revision Identification Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: Default: See register description Offset Start: 08h Offset End: 08h Power Well: Bit Range Bit Acronym Bit Description 07 :00 RID Revision ID: This is an 8-bit value indicating the stepping of the USB controller hardware. 10.1.1.6 B0:D29 :F0 Sticky Bit Reset Value Bit Access RO Offset 09h: Programming Interface Register (USB EHCI--B0:D29:F0) Table 10-7. Offset 09h: Programming Interface Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit Bit Range Default: 20h Bit Acronym Offset Start: 09h Offset End: 09h Power Well: Bit Description Sticky Bit Reset Value Bit Access Programming Interface -- A value of 20h indicates that this USB 2.0 host controller conforms to the EHCI Specification. 07 :00 10.1.1.7 Bus:Device:Function: B0:D29 :F0 BAR: Configuration RO Offset 0Ah: Sub Class Code Register (USB EHCI--B0:D29:F0) Table 10-8. Offset 0Ah: Sub Class Code Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit Bus:Device:Function: B0:D29 :F0 BAR: Configuration Default: 03h Bit Range Bit Acronym 07 :00 SCC Power Well: Bit Description Sub Class Code: 03h = Universal serial bus host controller. Intel(R) Communications Chipset 89xx Series - Datasheet 546 Offset Start: 0Ah Offset End: 0Ah Sticky Bit Reset Value Bit Access RO October 2012 Order Number: 327879-001US 10.0 10.1.1.8 Offset 0Bh: BCC--Base Class Code Register (USB EHCI--B0:D29:F0) Table 10-9. Offset 0Bh: BCC--Base Class Code Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit BAR: Configuration B0:D29 :F0 Default: 0Ch Bit Range Bit Acronym 07 :00 BCC 10.1.1.9 Bus:Device:Function: Offset Start: 0Bh Offset End: 0Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Base Class Code: 0Ch = Serial bus controller. RO Offset 0Dh: Primary Master Latency Timer Register (USB EHCI-- B0:D29:F0) Table 10-10. Offset 0Dh: Primary Master Latency Timer Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: 00h Power Well: Bit Range Bit Acronym Bit Description 07 :00 MLTC Master Latency Timer Count: 00h = Hardwired. The SATA controller is implemented internally, and is not arbitrated as a PCI device, so it does not need a Master Latency Timer. October 2012 Order Number: 327879-001US Offset Start: 0Dh Offset End: 0Dh Sticky Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 547 10.1.1.10 Offset 0Eh: Header Type Register (USB EHCI--B0:D29:F0) Table 10-11. Offset 0Eh: Header Type Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: 80h Bit Acronym Offset Start: 0Eh Offset End: 0Eh Power Well: Bit Description Sticky Bit Reset Value Bit Access Multi-Function Device -- When set to `1' indicates this is a multifunction device: 0 = Single-function device 1 = Multi-function device. 07 RO When RMH is enabled this bit defaults to 1, when RMH is disabled this bit defaults to 0. Configuration Layout. Hardwired to 00h, which indicates the standard PCI configuration layout. 06 :00 10.1.1.11 RO Offset 10h: Memory Base Address Register (USB EHCI--B0:D29:F0) Table 10-12. Offset 10h: Memory Base Address Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 32 bit Default: 00000000h Bit Range Bit Acronym 31 :10 BA 09 :04 Reserved 03 02 :01 00 Bus:Device:Function: B0:D29 :F0 BAR: Configuration Power Well: Bit Description Base Address -- Bits [31:10] correspond to memory address signals [31:10], respectively. This gives 1-KB of locatable memory space aligned to 1-KB boundaries. Sticky Bit Reset Value Bit Access RW Reserved PREF Prefetchable -- Hardwired to 0 indicating that this range should not be prefetched. RO TYPE Type -- Hardwired to 00b indicating that this range can be mapped anywhere within 32-bit address space. RO RTE Resource Type Indicator -- Hardwired to 0 indicating that the base address field in this register maps to memory space. RO Intel(R) Communications Chipset 89xx Series - Datasheet 548 Offset Start: 10h Offset End: 13h October 2012 Order Number: 327879-001US 10.0 10.1.1.12 Offset 2Ch: Subsystem Vendor ID Register (USB EHCI--B0:D29:F0) Table 10-13. Offset 2Ch: Subsystem Vendor ID Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit Bit Range 15 :00 BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: XXXXh Bit Acronym Offset Start: 2Ch Offset End: 2Dh Power Well: Bit Description Sticky Bit Reset Value Bit Access Subsystem Vendor ID -- This register, in combination with the USB 2.0 Subsystem ID register, enables the operating system to distinguish each subsystem from the others. SVID RW Writes to this register are enabled when the WRT_RDONLY bit (B0:D29:F0:80h, bit 0) is set to 1. 10.1.1.13 Offset 2Eh: Subsystem ID Register (USB EHCI--B0:D29:F0) Table 10-14. Offset 2Eh: Subsystem ID Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit Bit Range 15 :00 BAR: Configuration B0:D29 Bus:Device:Function: :F0 Default: XXXXh Bit Acronym SID Offset Start: 2Eh Offset End: 2Fh Power Well: Bit Description Sticky Subsystem ID -- BIOS sets the value in this register to identify the Subsystem ID. This register, in combination with the Subsystem Vendor ID register, enables the operating system to distinguish each subsystem from other(s). Bit Reset Value Bit Access RW Writes to this register are enabled when the WRT_RDONLY bit (B0:D29:F0:80h, bit 0) is set to 1. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 549 10.1.1.14 Offset 34h: Capabilities Pointer Register (USB EHCI--B0:D29:F0) Table 10-15. Offset 34h: Capabilities Pointer Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit BAR: Configuration B0:D29 :F0 Default: 50h Bit Range Bit Acronym 07 :00 CAP_PTR 10.1.1.15 Bus:Device:Function: Offset Start: 34h Offset End: 34h Power Well: Bit Description Sticky Bit Reset Value Bit Access Capabilities Pointer -- This register points to the starting offset of the USB 2.0 capabilities ranges. RO Offset 3Ch: Interrupt Line Register (USB EHCI--B0:D29:F0) Table 10-16. Offset 3Ch: Interrupt Line Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit B0:D29 Bus:Device:Function: :F0 BAR: Configuration Default: 00h Power Well: Bit Range Bit Acronym Bit Description 07 :00 INT_LN Interrupt Line -- This data is not used by the PCH. It is used as a scratchpad register to communicate to software the interrupt line to which that the interrupt pin is connected. 10.1.1.16 Offset Start: 3Ch Offset End: 3Ch Sticky Bit Reset Value Bit Access RW Offset 3Dh: Interrupt Pin Register (USB EHCI--B0:D29:F0) Table 10-17. Offset 3Dh: Interrupt Pin Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit Bit Range 07 :00 B0:D29 Bus:Device:Function: :F0 BAR: Configuration Default: See register description Bit Acronym Bit Description Interrupt Pin -- This reflects the value of D29IP.E1IP (Chipset Config Registers:Offset 3108:bits 3:0). Bits 7:4 are always 0h Intel(R) Communications Chipset 89xx Series - Datasheet 550 Offset Start: 3Dh Offset End: 3Dh Power Well: Sticky Bit Reset Value Bit Access RO October 2012 Order Number: 327879-001US 10.0 10.1.1.17 Offset 50h: PCI Power Management Capability ID Register (USB EHCI--B0:D29:F0) Table 10-18. Offset 50h: PCI Power Management Capability ID Register (USB EHCI-- B0:D29:F0) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration B0:D29 :F0 Default: 01h Bit Acronym Offset Start: 50h Offset End: 50h Power Well: Bit Description Sticky Bit Reset Value Bit Access Power Management Capability ID -- A value of 01h indicates that this is a PCI Power Management capabilities field. 07 :00 10.1.1.18 Bus:Device:Function: RO Offset 51h: Next Item Pointer #1 Register (USB EHCI--B0:D29:F0) Table 10-19. Offset 51h: Next Item Pointer #1 Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: 58h Bit Acronym 07 :00 October 2012 Order Number: 327879-001US Offset Start: 51h Offset End: 51h Power Well: Bit Description Sticky Next Item Pointer 1 Value -- (special). This register defaults to 58h, which indicates that the next capability registers begin at configuration offset 58h. This register is writable when the WRT_RDONLY bit (B0:D29:F0:F0:80h, bit 0) is set. This allows BIOS to effectively hide the Debug Port capability registers, if necessary. This register should only be written during system initialization before the plug-and-play software has enabled any master-initiated traffic. Only values of 58h (Debug Port and FLR capabilities visible) and 98h (Debug Port invisible, next capability is FLR) are expected to be programmed in this register. Note: Register not reset by D3-to-D0 warm reset. Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 551 10.1.1.19 Offset 52h: Power Management Capabilities Register (USB EHCI-- B0:D29:F0) Note: Normally, this register is read-only to report capabilities to the power management software. To report different power management capabilities, depending on the system in which the PCH is used, bits 15:11 and 8:6 in this register are writable when the WRT_RDONLY bit (B0:D29:F0:F0:80h, bit 0) is set. The value written to this register does not affect the hardware other than changing the value returned during a read. Note: Reset: core well, but not D3-to-D0 warm reset. Table 10-20. Offset 52h: Power Management Capabilities Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit Bit Range 15 :11 Bus:Device:Function: B0:D29 :F0 BAR: Configuration Default: C9C2h Power Well: Sticky Bit Reset Value Bit Acronym Bit Description Bit Access PME_SUP PME Support -- This 5-bit field indicates the power states in which the function may assert PME#. The PCH EHC does not support the D1 or D2 states. For all other states, the PCH EHC is capable of generating PME#. Software should never need to modify this field. RW 10 D2_SUP D2 Support: 0 = D2 State is not supported RO 09 D1_SUP D1 Support: 0 = D1 State is not supported RO Auxiliary Current -- The PCH EHC reports 375 mA maximum suspend well current required when in the D3COLD state. RW Device Specific Initialization -- The PCH reports 0, indicating that no device-specific initialization is required. RO 08 :06 AUX_CUR 05 DSI 04 Reserved Reserved 03 PME_CLK PME Clock -- The PCH reports 0, indicating that no PCI clock is required to generate PME#. RO Version -- The PCH reports 010b, indicating that it complies with Revision 1.1 of the PCI Power Management Specification. RO 02 :00 VER Intel(R) Communications Chipset 89xx Series - Datasheet 552 Offset Start: 52h Offset End: 53h October 2012 Order Number: 327879-001US 10.0 10.1.1.20 Offset 54h: Power Management Control/Status Register (USB EHCI-- B0:D29:F0) Reset (bits 15, 8): suspend well, and not D3-to-D0 warm reset or core well reset. Table 10-21. Offset 54h: Power Management Control/Status Register (USB EHCI-- B0:D29:F0) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: 0000h Bit Acronym Power Well: Bit Description Sticky PME Status: 0 = Writing a 1 to this bit will clear it and cause the internal PME to deassert (if enabled). 1 = This bit is set when the PCH EHC would normally assert the PME# signal independent of the state of the PME_En bit. 07 :00 Offset Start: 54h Offset End: 55h Bit Reset Value Bit Access RWC This bit must be explicitly cleared by the operating system each time the operating system is loaded. This bit is not reset by Function Level Reset. 14 :13 Data Scale -- Hardwired to 00b indicating it does not support the associated Data register. RO 12 :09 Data Select -- Hardwired to 0000b indicating it does not support the associated Data register. RO PME Enable: 0 = Disable. 1 = Enables the PCH EHC to generate an internal PME signal when PME_Status is 1. 08 RW This bit must be explicitly cleared by the operating system each time it is initially loaded. This bit is not reset by Function Level Reset. 07 :02 Reserved 01 :00 October 2012 Order Number: 327879-001US Reserved Power State -- This 2-bit field is used both to determine the current power state of EHC function and to set a new power state. The definition of the field values are: 00 = D0 state 11 = D3HOT state If software attempts to write a value of 10b or 01b in to this field, the write operation must complete normally; however, the data is discarded and no state change occurs. When in the D3HOT state, the PCH must not accept accesses to the EHC memory range; but the configuration space must still be accessible. When not in the D0 state, the generation of the interrupt output is blocked. Specifically, the PIRQH is not asserted by the PCH when not in the D0 state. When software changes this value from the D3HOT state to the D0 state, an internal warm (soft) reset is generated, and software must re-initialize the function. RW Intel(R) Communications Chipset 89xx Series - Datasheet 553 10.1.1.21 Offset 58h: Debug Port Capability ID Register (USB EHCI--B0:D29:F0) Table 10-22. Offset 58h: Debug Port Capability ID Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration B0:D29 :F0 Default: 0Ah Bit Acronym Offset Start: 58h Offset End: 58h Power Well: Bit Description Sticky Bit Reset Value Bit Access Debug Port Capability ID -- Hardwired to 0Ah indicating that this is the start of a Debug Port Capability structure. 07 :00 10.1.1.22 Bus:Device:Function: RO Offset 59h: Next Item Pointer #2 Register (USB EHCI--B0:D29:F0) Table 10-23. Offset 59h: Next Item Pointer #2 Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit Bit Range Default: 98h Bit Acronym Offset Start: 59h Offset End: 59h Power Well: Bit Description Sticky Bit Reset Value Bit Access Next Item Pointer 2 Capability -- This register points to the next capability in the Function Level Reset capability structure. 07 :00 10.1.1.23 B0:D29 Bus:Device:Function: :F0 BAR: Configuration RO Offset 5Ah: Debug Port Base Offset Register (USB EHCI--B0:D29:F0) Table 10-24. Offset 5Ah: Debug Port Base Offset Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit Bit Range Bus:Device:Function: B0:D29 :F0 BAR: Configuration Default: 20A0h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access 15 :13 BAR Number -- Hardwired to 001b to indicate the memory BAR begins at offset 10h in the EHCI configuration space. RO 12 :00 Debug Port Offset -- Hardwired to 0A0h to indicate that the Debug Port registers begin at offset A0h in the EHCI memory range. RO Intel(R) Communications Chipset 89xx Series - Datasheet 554 Offset Start: 5Ah Offset End: 5Bh October 2012 Order Number: 327879-001US 10.0 10.1.1.24 Offset 60h: USB Release Number Register (USB EHCI--B0:D29:F0) Table 10-25. Offset 60h: USB Release Number Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration B0:D29 :F0 Default: 20h Bit Acronym Offset Start: 60h Offset End: 60h Power Well: Bit Description Sticky Bit Reset Value USB Release Number -- A value of 20h indicates that this controller follows Universal Serial Bus (USB) Specification, Revision 2.0. 07 :00 10.1.1.25 Bus:Device:Function: Bit Access RO Offset 61h: Frame Length Adjustment Register (USB EHCI-- B0:D29:F0) This feature is used to adjust any offset from the clock source that generates the clock that drives the SOF counter. When a new value is written into these six bits, the length of the frame is adjusted. Its initial programmed value is system dependent based on the accuracy of hardware USB clock and is initialized by system BIOS. This register should only be modified when the HChalted bit (B0:D29:F0:F0:CAPLENGTH + 24h, bit 12) in the USB2.0_STS register is a 1. Changing value of this register while the host controller is operating yields undefined results. It should not be reprogrammed by USB system software unless the default or BIOS programmed values are incorrect, or the system is restoring the register while returning from a suspended state. These bits in suspend well and not reset by a D3-to-D0 warm rest or a core well reset. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 555 Table 10-26. Offset 61h: Frame Length Adjustment Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: 20h Bit Range Bit Acronym 07 :06 Reserved Offset Start: 61h Offset End: 61h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved -- These bits are reserved for future use and should read as 00b. Frame Length Timing Value -- Each decimal value change to this register corresponds to 16 high-speed bit times. The SOF cycle time (number of SOF counter clock periods to generate a SOF micro-frame length) is equal to 59488 + value in this field. The default value is decimal 32 (20h), which gives a SOF cycle time of 60000. 05 :00 Frame Length (# 480 MHz Clocks) (decimal) Frame Length Timing Value (this register) (decimal) 59488 0 59504 1 59520 2 -- -- 59984 31 60000 32 -- -- 60480 62 60496 63 Intel(R) Communications Chipset 89xx Series - Datasheet 556 RW October 2012 Order Number: 327879-001US 10.0 10.1.1.26 Offset 62h: Port Wake Capability Register (USB EHCI--B0:D29:F0) This register is in the suspend power well. The intended use of this register is to establish a policy about which ports are to be used for wake events. Bit positions 1- 8(D29) in the mask correspond to a physical port implemented on the current EHCI controller. A 1 in a bit position indicates that a device connected below the port can be enabled as a wake-up device and the port may be enabled for disconnect/connect or overcurrent events as wake-up events. This is an information-only mask register. The bits in this register do not effect the actual operation of the EHCI host controller. The system-specific policy can be established by BIOS initializing this register to a systemspecific value. System software uses the information in this register when enabling devices and ports for remote wake-up. These bits are not reset by a D3-to-D0 warm rest or a core well reset. Table 10-27. Offset 62h: Port Wake Capability Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit BAR: Configuration B0:D29 Bus:Device:Function: :F0 Default: 01FFh Bit Range Bit Acronym 15 :09 Reserved 08 :01 00 October 2012 Order Number: 327879-001US Offset Start: 62h Offset End: 63h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Port Wake Up Capability Mask -- R/W. Bit positions 1 through 8 (Device 29) correspond to a physical port implemented on this host controller. For example, bit position 1 corresponds to port 1, bit position 2 corresponds to port 2, etc. RW Port Wake Implemented -- R/W. A 1 in this bit indicates that this register is implemented to software. RW Intel(R) Communications Chipset 89xx Series - Datasheet 557 10.1.1.27 Offset 68h: Legacy Support Extended Capability Register (USB EHCI-- B0:D29:F0) These bits are not reset by a D3-to-D0 warm rest or a core well reset. Table 10-28. Offset 68h: Legacy Support Extended Capability Register (USB EHCI-- B0:D29:F0) Description: View: PCI Size: 32 bit BAR: Configuration Bit Acronym 31 :25 Reserved Reserved Offset Start: 68h Offset End: 6Bh Power Well: Suspend Bit Description Sticky Bit Reset Value Bit Access Reserved -- Hardwired to 00h HC OS Owned Semaphore -- System software sets this bit to request ownership of the EHCI controller. Ownership is obtained when this bit reads as 1 and the HC BIOS Owned Semaphore bit reads as clear. 24 RW Reserved -- Hardwired to 00h HC BIOS Owned Semaphore -- The BIOS sets this bit to establish ownership of the EHCI controller. System BIOS will clear this bit in response to a request for ownership of the EHCI controller by system software. RW 15 :08 Next EHCI Capability Pointer -- Hardwired to 00h to indicate that there are no EHCI Extended Capability structures in this device. RO 07 :00 Capability ID -- Hardwired to 01h to indicate that this EHCI Extended Capability is the Legacy Support Capability. RO 16 Intel(R) Communications Chipset 89xx Series - Datasheet 558 B0:D29 :F0 Default: 00000001h Bit Range 23 :17 Bus:Device:Function: October 2012 Order Number: 327879-001US 10.0 10.1.1.28 Offset 6Ch: Legacy Support Extended Control/Status Register (USB EHCI--B0:D29:F0) Table 10-29. Offset 6Ch: Legacy Support Extended Control/Status Register (USB EHCI-- B0:D29:F0) (Sheet 1 of 3) Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: 00000000h Bit Acronym Offset Start: 6Ch Offset End: 6Fh Power Well: Suspend Bit Description Sticky Bit Reset Value Bit Access 31 SMI on BAR -- Software clears this bit by writing a 1 to it. 0 = Base Address Register (BAR) not written. 1 = This bit is set to 1 when the Base Address Register (BAR) is written. RWC 30 SMI on PCI Command -- Software clears this bit by writing a 1 to it. 0 = PCI Command (PCICMD) Register Not written. 1 = This bit is set to 1 when the PCI Command (PCICMD) Register is written. RWC 29 SMI on OS Ownership Change -- Software clears this bit by writing a 1 to it. 0 = No HC OS Owned Semaphore bit change. 1 = This bit is set to 1 when the HC OS Owned Semaphore bit in the LEG_EXT_CAP register (B0:D29:F0:F0:68h, bit 24) transitions from 1 to 0 or 0 to 1. RWC 28 :22 Reserved 27 Reserved SMI on Async Advance -- This bit is a shadow bit of the Interrupt on Async Advance bit (B0:D29:F0:F0:CAPLENGTH + 24h, bit 5) in the USB2.0_STS register. RO To clear this bit system software must write a 1 to the Interrupt on Async Advance bit in the USB2.0_STS register. 26 25 SMI on Host System Error -- This bit is a shadow bit of Host System Error bit in the USB2.0_STS register (B0:D29:F0:F0:CAPLENGTH + 24h, bit 4). To clear this bit system software must write a 1 to the Host System Error bit in the USB2.0_STS register.bit in the USB2.0_STS register. SMI on Frame List Rollover -- This bit is a shadow bit of Frame List Rollover bit (B0:D29:F0:F0:CAPLENGTH + 24h, bit 3) in the USB2.0_STS register. RO RO To clear this bit system software must write a 1 to the Frame List Rollover bit in the USB2.0_STS register. 24 SMI on Port Change Detect -- This bit is a shadow bit of Port Change Detect bit (B0:D29:F0:F0:CAPLENGTH + 24h, bit 2) in the USB2.0_STS register. RO To clear this bit system software must write a 1 to the Port Change Detect bit in the USB2.0_STS register. 23 SMI on USB Error -- This bit is a shadow bit of USB Error Interrupt (USBERRINT) bit (B0:D29:F0:F0:CAPLENGTH + 24h, bit 1) in the USB2.0_STS register. RO To clear this bit system software must write a 1 to the USB Error Interrupt bit in the USB2.0_STS register. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 559 Table 10-29. Offset 6Ch: Legacy Support Extended Control/Status Register (USB EHCI-- B0:D29:F0) (Sheet 2 of 3) Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: 00000000h Bit Acronym Power Well: Suspend Bit Description SMI on USB Complete -- This bit is a shadow bit of USB Interrupt (USBINT) bit (B0:D29:F0:F0:CAPLENGTH + 24h, bit 0) in the USB2.0_STS register. 22 Offset Start: 6Ch Offset End: 6Fh Sticky Bit Reset Value Bit Access RO To clear this bit system software must write a 1 to the USB Interrupt bit in the USB2.0_STS register. 21 SMI on BAR Enable: 0 = Disable. 1 = Enable. When this bit is 1 and SMI on BAR (B0:D29:F0:F0:6Ch, bit 31) is 1, then the host controller will issue an SMI. RW 20 SMI on PCI Command Enable: 0 = Disable. 1 = Enable. When this bit is 1 and SMI on PCI Command (B0:D29:F0:F0:6Ch, bit 30) is 1, then the host controller will issue an SMI. RW 19 SMI on OS Ownership Enable: 0 = Disable. 1 = Enable. When this bit is a 1 AND the OS Ownership Change bit (B0:D29:F0:F0:6Ch, bit 29) is 1, the host controller will issue an SMI. RW 18 Reserved Reserved 17 SMI on Async Advance Enable: 0 = Disable. 1 = Enable. When this bit is a 1, and the SMI on Async Advance bit (B0:D29:F0:6Ch, bit 21) is a 1, the host controller will issue an SMI immediately. RW 16 SMI on Host System Error Enable: 0 = Disable. 1 = Enable. When this bit is a 1, and the SMI on Host System Error (B0:D29:F0:6Ch, bit 20) is a 1, the host controller will issue an SMI. RW 15 SMI on Frame List Rollover Enable: 0 = Disable. 1 = Enable. When this bit is a 1, and the SMI on Frame List Rollover bit (B0:D29:F0:6Ch, bit 19) is a 1, the host controller will issue an SMI. RW 14 SMI on Port Change Enable: 0 = Disable. 1 = Enable. When this bit is a 1, and the SMI on Port Change Detect bit (B0:D29:F0:6Ch, bit 18) is a 1, the host controller will issue an SMI. RW 13 SMI on USB Error Enable: 0 = Disable. 1 = Enable. When this bit is a 1, and the SMI on USB Error bit (B0:D29:F0:6Ch, bit 17) is a 1, the host controller will issue an SMI immediately. RW SMI on USB Complete Enable: 0 = Disable. 1 = Enable. When this bit is a 1, and the SMI on USB Complete bit (B0:D29:F0:6Ch, bit 16) is a 1, the host controller will issue an SMI immediately. RW 12 :06 Intel(R) Communications Chipset 89xx Series - Datasheet 560 October 2012 Order Number: 327879-001US 10.0 Table 10-29. Offset 6Ch: Legacy Support Extended Control/Status Register (USB EHCI-- B0:D29:F0) (Sheet 3 of 3) Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: 00000000h Bit Acronym Offset Start: 6Ch Offset End: 6Fh Power Well: Suspend Bit Description Sticky Bit Reset Value Bit Access 05 SMI on BAR -- Software clears this bit by writing a 1 to it. 0 = Base Address Register (BAR) not written. 1 = This bit is set to 1 when the Base Address Register (BAR) is written. RWC 04 SMI on PCI Command -- Software clears this bit by writing a 1 to it. 0 = PCI Command (PCICMD) Register Not written. 1 = This bit is set to 1 when the PCI Command (PCICMD) Register is written. RWC 03 SMI on OS Ownership Change -- Software clears this bit by writing a 1 to it. 0 = No HC OS Owned Semaphore bit change. 1 = This bit is set to 1 when the HC OS Owned Semaphore bit in the LEG_EXT_CAP register (B0:D29:F0:68h, bit 24) transitions from 1 to 0 or 0 to 1. RWC 02 Reserved 01 00 Reserved SMI on Async Advance -- This bit is a shadow bit of the Interrupt on Async Advance bit (D29:F0:CAPLENGTH + 24h, bit 5) in the USB2.0_STS register. To clear this bit system software must write a 1 to the Interrupt on Async Advance bit in the USB2.0_STS register. SMI on Host System Error -- This bit is a shadow bit of Host System Error bit in the USB2.0_STS register (D29:F0:CAPLENGTH + 24h, bit 4). RO RO To clear this bit system software must write a 1 to the Host System Error bit in the USB2.0_STS register. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 561 10.1.1.29 Offset 70h: SPECIAL_SMI--Intel Specific USB 2.0 SMI Register (USB EHCI--B0:D29:F0) Table 10-30. Offset 70h: SPECIAL_SMI--Intel Specific USB 2.0 SMI Register (USB EHCI-- B0:D29:F0) (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: Default: 00000000h Bit Range Bit Acronym 31 :28 Reserved Offset Start: 70h Offset End: 73h Power Well: Suspend Bit Description Sticky Bit Reset Value Bit Access Reserved SMI on PortOwner -- Software clears these bits by writing a 1 to it. 0 = No Port Owner bit change. 1 = Bits 29:22, 27:22 correspond to the Port Owner bits for ports 0 (22) through 5 (27) or 7 (29). These bits are set to 1 when the associated Port Owner bits transition from 0 to 1 or 1 to 0. RWC 21 SMI on PMCSR -- Software clears these bits by writing a 1 to it. 0 = Power State bits Not modified. 1 = Software modified the Power State bits in the Power Management Control/Status (PMCSR) register (B0:D29:F0:54h). RWC 20 SMI on Async -- Software clears these bits by writing a 1 to it. 0 = No Async Schedule Enable bit change 1 = Async Schedule Enable bit transitioned from 1 to 0 or 0 to 1. RWC 19 SMI on Periodic -- Software clears this bit by writing a 1 it. 0 = No Periodic Schedule Enable bit change. 1 = Periodic Schedule Enable bit transitions from 1 to 0 or 0 to 1. RWC 18 SMI on CF -- Software clears this bit by writing a 1 it. 0 = No Configure Flag (CF) change. 1 = Configure Flag (CF) transitions from 1 to 0 or 0 to 1. RWC 17 SMI on HCHalted -- Software clears this bit by writing a 1 it. 0 = HCHalted did Not transition to 1 (as a result of the Run/Stop bit being cleared). 1 = HCHalted transitions to 1 (as a result of the Run/Stop bit being cleared). RWC 16 SMI on HCReset -- Software clears this bit by writing a 1 it. 0 = HCRESET did Not transitioned to 1. 1 = HCRESET transitioned to 1. RWC 29 :22 15 :14 13 :06 05 Reserved Reserved SMI on PortOwner Enable: 0 = Disable. 1 = Enable. When any of these bits are 1 and the corresponding SMI on PortOwner bits are 1, then the host controller will issue an SMI. Unused ports should have their corresponding bits cleared. RW SMI on PMSCR Enable: 0 = Disable. 1 = Enable When this bit is 1 and SMI on PMSCR is 1, then the host controller will issue an SMI. RW Intel(R) Communications Chipset 89xx Series - Datasheet 562 B0:D29 :F0 October 2012 Order Number: 327879-001US 10.0 Table 10-30. Offset 70h: SPECIAL_SMI--Intel Specific USB 2.0 SMI Register (USB EHCI-- B0:D29:F0) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: 00000000h Bit Acronym Offset Start: 70h Offset End: 73h Power Well: Suspend Bit Description Sticky Bit Reset Value Bit Access 04 SMI on Async Enable: 0 = Disable. 1 = Enable When this bit is 1 and SMI on Async is 1, then the host controller will issue an SMI RW 03 SMI on Periodic Enable: 0 = Disable. 1 = Enable When this bit is 1 and SMI on Periodic is 1, then the host controller will issue an SMI. RW 02 SMI on CF Enable: 0 = Disable. 1 = Enable. When this bit is 1 and SMI on CF is 1, then the host controller will issue an SMI. RW 01 SMI on HCHalted Enable: 0 = Disable. 1 = Enable. When this bit is a 1 and SMI on HCHalted is 1, then the host controller will issue an SMI. RW 00 SMI on HCReset Enable: 0 = Disable. 1 = Enable. When this bit is a 1 and SMI on HCReset is 1, then host controller will issue an SMI. RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 563 10.1.1.30 Offset 80h: ACCESS_CNTL--Access Control Register (USB EHCI-- B0:D29:F0) Table 10-31. Offset 80h: ACCESS_CNTL--Access Control Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit BAR: Configuration Bit Acronym 07 :01 Reserved 10.1.1.31 B0:D29 :F0 Default: 00h Bit Range 00 Bus:Device:Function: Offset Start: 80h Offset End: 80h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved When set to 1, this bit enables a select group of normally read-only registers in the EHC function to be written by software. Registers that may only be written when this mode is entered are noted in the summary tables and WRT_RDONLY detailed description as "Read/Write-Special". The registers fall into two categories: 1. System-configured parameters 2. Status bits RW Offset 84h: EHCIIR1--EHCI Initialization Register 1 (USB EHCI-- B0:D29:F0) Table 10-32. Offset 84h: EHCIIR1--EHCI Initialization Register 1 (USB EHCI--B0:D29:F0) Description: View: PCI Size: 32 bit Default: 01h Bit Range Bit Acronym 31 :05 Reserved Reserved Bit Description Sticky Bit Reset Value Bit Access Reserved RW Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 564 Offset Start: 84h Offset End: 87h Power Well: Pre-fetch Based Pause Disable: 0 = Pre-fetch Based Pause is enabled. 1 = Pre-fetch Based Pause is disabled. 04 03 :00 B0:D29 Bus:Device:Function: :F0 BAR: Configuration October 2012 Order Number: 327879-001US 10.0 10.1.1.32 Offset 98h: FLR_CID--Function Level Reset Capability ID (USB EHCI-- B0:D29:F0) Table 10-33. Offset 98h: FLR_CID--Function Level Reset Capability ID (USB EHCI-- B0:D29:F0) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration B0:D29 :F0 Default: 09h Bit Acronym Offset Start: 98h Offset End: 98h Power Well: Bit Description Sticky Bit Reset Value Bit Access Capability ID: 13h = If FLRCSSEL = 0 09h (Vendor Specific Capability) = If FLRCSSEL = 1 07 :00 10.1.1.33 Bus:Device:Function: RO Offset 99h: FLR_NEXT--Function Level Reset Next Capability Pointer (USB EHCI--B0:D29:F0) Table 10-34. Offset 99h: FLR_NEXT--Function Level Reset Next Capability Pointer (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: 00h Bit Acronym 07 :00 October 2012 Order Number: 327879-001US Offset Start: 99h Offset End: 99h Power Well: Bit Description Sticky A value of 00h in this register indicates this is the last capability field Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 565 10.1.1.34 Offset 9Ah: FLR_CLV--Function Level Reset Capability Length and Version (USB EHCI--B0:D29:F0) When FLRCSSEL = `0' this register is defined as follows: Table 10-35. Offset 9Ah: FLR_CLV--Function Level Reset Capability Length and Version (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D29 :F0 Default: 2006h Bit Range Bit Acronym 15 :10 Reserved Offset Start: 9Ah Offset End: 9Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 09 FLR Capability: 1 = Support for Function Level Reset (FLR). RWO 08 TXP Capability: 1 = Support for Transactions Pending (TXP) bit. TXP must be supported if FLR is supported. RWO Capability Length -- This field indicates the # of bytes of this vendor specific capability as required by the PCI specification. It has the value of 06h for the FLR capability. RO 07 :00 When FLRCSSEL = `1' this register is defined as follows: Table 10-36. Offset 9Ah: FLR_CLV--Function Level Reset Capability Length and Version (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit Bit Range Bus:Device:Function: B0:D29 :F0 BAR: Configuration Default: 2006h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access 15 :12 Vendor Specific Capability ID -- A value of 2h in this field identifies this capability as Function Level Reset. RO 11 :08 Capability Version -- This field indicates the version of the FLR capability. RO 07 :00 Capability Length -- This field indicates the # of bytes of this vendor specific capability as required by the PCI specification. It has the value of 06h for the FLR capability. RO Intel(R) Communications Chipset 89xx Series - Datasheet 566 Offset Start: 9Ah Offset End: 9Bh October 2012 Order Number: 327879-001US 10.0 10.1.1.35 Offset 9Ch: FLR_CTRL--Function Level Reset Control Register (USB EHCI--B0:D29:F0) Table 10-37. Offset 9Ch: FLR_CTRL--Function Level Reset Control Register (USB EHCI-- B0:D29:F0) Description: View: PCI Size: 8 bit BAR: Configuration B0:D29 :F0 Default: 00h Bit Range Bit Acronym 07 :01 Reserved Offset Start: 9Ch Offset End: 9Ch Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Initiate FLR -- This bit is used to initiate FLR transition. A write of `1' initiates FLR transition. Since hardware must not respond to any cycles until FLR completion, the value read by software from this bit is always `0'. 00 10.1.1.36 Bus:Device:Function: RW Offset 9Dh: FLR_STS--Function Level Reset Status Register (USB EHCI--B0:D29:F0) Table 10-38. Offset 9Dh: FLR_STS--Function Level Reset Status Register (USB EHCI-- B0:D29:F0) Description: View: PCI Size: 8 bit BAR: Configuration Default: 00h Bit Range Bit Acronym 07 :01 Reserved 00 Bus:Device:Function: B0:D29 :F0 TXP October 2012 Order Number: 327879-001US Offset Start: 9Dh Offset End: 9Dh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Transactions Pending: 0 = Completions for all non-posted requests have been received. 1 = Controller has issued non-posted requests which have no bee completed. RO Intel(R) Communications Chipset 89xx Series - Datasheet 567 10.2 Memory-Mapped I/O Registers The EHCI memory-mapped I/O space is composed of two sets of registers: Capability Registers and Operational Registers. Note: The PCH EHCI controller will not accept memory transactions (neither reads nor writes) as a target that are locked transactions. The locked transactions should not be forwarded to PCI as the address space is known to be allocated to USB. Note: When the EHCI function is in the D3 PCI power state, accesses to the USB 2.0 memory range are ignored and result a master abort. Similarly, if the Memory Space Enable (MSE) bit (B0:D29:F0:04h, bit 1) is not set in the Command register in configuration space, the memory range will not be decoded by the PCH enhanced host controller (EHC). If the MSE bit is not set, then the PCH must default to allowing any memory accesses for the range specified in the BAR to go to PCI. This is because the range may not be valid and, therefore, the cycle must be made available to any other targets that may be currently using that range. 10.2.1 Host Controller Capability Registers These registers specify the limits, restrictions and capabilities of the host controller implementation. Within the host controller capability registers, only the structural parameters register is writable. These registers are implemented in the suspend well and is only reset by the standard suspend-well hardware reset, not by HCRESET or the D3-to-D0 reset. Note: The EHCI controller does not support as a target memory transactions that are locked transactions. Attempting to access the EHCI controller Memory-Mapped I/O space using locked memory transactions will result in undefined behavior. Note: When the USB2 function is in the D3 PCI power state, accesses to the USB2 memory range are ignored and will result in a master abort Similarly, if the Memory Space Enable (MSE) bit is not set in the Command register in configuration space, the memory range will not be decoded by the Enhanced Host Controller (EHC). If the MSE bit is not set, then the EHC will not claim any memory accesses for the range specified in the BAR. Table 10-39. Enhanced Host Controller Capability Registers View BAR BDF Offset Start Offset End Register ID: Register Name Default Value PCI MBAR B:29:0 00h 00h "Offset 00h: CAPLENGTH--Capability Registers Length (USB EHCI--B0:D29:F0)" on page 569 20h PCI MBAR B:29:0 02h 03h "Offset 02h: HCIVERSION--Host Controller Interface Version Number Register (USB EHCI-- B0:D29:F0)" on page 569 0100h PCI MBAR B:29:0 04h 07h "Offset 04h: HCSPARAMS--Host Controller Structural Parameters (USB EHCI--B0:D29:F0)" on page 570 00204208h PCI MBAR B:29:0 08h 0Bh "Offset 08h: HCCPARAMS--Host Controller Capability Parameters Register (USB EHCI-- B0:D29:F0)" on page 571 00006881h Note: "Read/Write Special" means that the register is normally read-only, but may be written when the WRT_RDONLY bit is set. Because these registers are expected to be programmed by BIOS during initialization, their contents must not get modified by HCRESET or D3-to-D0 internal reset. Intel(R) Communications Chipset 89xx Series - Datasheet 568 October 2012 Order Number: 327879-001US 10.0 10.2.1.1 Offset 00h: CAPLENGTH--Capability Registers Length Register (USB EHCI--B0:D29:F0) Table 10-40. Offset 00h: CAPLENGTH--Capability Registers Length (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit Bit Range BAR: MBAR B0:D29 :F0 Default: 20h Bit Acronym Offset Start: 00h Offset End: 00h Power Well: Bit Description Sticky Bit Reset Value Bit Access Capability Register Length Value -- This register is used as an offset to add to the Memory Base Register (B0:D29:F0:Offset 10h) to find the beginning of the Operational Register Space. This field is hardwired to 20h indicating that the Operation Registers begin at offset 20h. 07 :00 10.2.1.2 Bus:Device:Function: RO Offset 02h: HCIVERSION--Host Controller Interface Version Number Register (USB EHCI--B0:D29:F0) Table 10-41. Offset 02h: HCIVERSION--Host Controller Interface Version Number Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 16 bit Bit Range BAR: MBAR B0:D29 Bus:Device:Function: :F0 Default: 0100h Bit Acronym 15 :00 October 2012 Order Number: 327879-001US Offset Start: 02h Offset End: 03h Power Well: Bit Description Sticky Host Controller Interface Version Number -- This is a twobyte register containing a BCD encoding of the version number of interface that this host controller interface conforms. Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 569 10.2.1.3 Offset 04h: HCSPARAMS--Host Controller Structural Parameters (USB EHCI--B0:D29:F0) This register is writable when the WRT_RDONLY bit is set. Table 10-42. Offset 04h: HCSPARAMS--Host Controller Structural Parameters (USB EHCI-- B0:D29:F0) Description: View: PCI Size: 32 bit BAR: MBAR Bus:Device:Function: Default: 00204208h Bit Range Bit Acronym 31 :24 Reserved 23 :20 DP_N 19 :16 Reserved Bit Description Sticky Bit Reset Value Bit Access Reserved Debug Port Number -- Hardwired to 2h indicating that the Debug Port is on the second lowest numbered port on the EHCI: EHCI#1: Port 1 RO Reserved 15 :12 N_CC 11 :08 N_PCC Number of Ports per Companion Controller -- This field indicates the number of ports supported per companion host controller. This field is set to 0h to indicate no other companion support. 07 :04 Reserved N_PORTS Offset Start: 04h Offset End: 07h Power Well: Number of Companion Controllers -- This field indicates the number of companion controllers associated with this USB EHCI host controller. BIOS must program this field to 0b to indicate that companion host controllers are not supported. Port ownership hand-off is not supported. Only high-speed devices are supported on the host controller root ports. 03 :00 B0:D29 :F0 RW RO Reserved. These bits are reserved and default to 0. This field specifies the number of physical downstream ports implemented on this host controller. The value of this field determines how many port registers are addressable in the Operational Register Space. Valid values are in the range of 1h to Fh. A `0' in this field is undefined. RW For Integrated USB 2.0 Rate Matching Hub Enabled: Each EHCI reports 2 ports by default. Port 0 assigned to the RMH and port 1 assigned as the debug port. When the KVM feature is enabled it will show up as Port2 on the EHCI, and BIOS would need to update this field to 3h. Intel(R) Communications Chipset 89xx Series - Datasheet 570 October 2012 Order Number: 327879-001US 10.0 10.2.1.4 Offset 08h: HCCPARAMS--Host Controller Capability Parameters Register (USB EHCI--B0:D29:F0) Table 10-43. Offset 08h: HCCPARAMS--Host Controller Capability Parameters Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 32 bit BAR: MBAR Bus:Device:Function: B0:D29 :F0 Default: 00006881h Bit Range Bit Acronym 31 :18 Reserved Offset Start: 08h Offset End: 0Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 17 ASUC Asynchronous Schedule Update Capability -- There is no functionality associated with this bit. RW 16 PSUC Periodic Schedule Update Capability -- This field is hardwired to 0b to indicate that the EHC hardware supports the Periodic Schedule Update Event Flag in the USB2.0_CMD register. RO EECP EHCI Extended Capabilities Pointer -- This field is hardwired to 68h, indicating that the EHCI capabilities list exists and begins at offset 68h in the PCI configuration space. RO Isochronous Scheduling Threshold -- This field indicates, relative to the current position of the executing host controller, where software can reliably update the isochronous schedule. When bit 7 is 0, the value of the least significant 3 bits indicates the number of microframes a host controller hold a set of isochronous data structures (one or more) before flushing the state. When bit 7 is a 1, then host software assumes the host controller may cache an isochronous data structure for an entire frame. See the EHCI specification for details on how software uses this information for scheduling isochronous transfers. This field is hardwired to 8h. RO 15 :08 07 :04 03 Reserved Reserved 02 Asynchronous Schedule Park Capability -- This bit is hardwired to 0 indicating that the host controller does not support this optional feature RO 01 Programmable Frame List Flag: 0 = System software must use a frame list length of 1024 elements with this host controller. The USB2.0_CMD register (B0:D29:F0:CAPLENGTH + 20h, bits 3:2) Frame List Size field is a read-only register and must be set to 0. 1 = System software can specify and use a smaller frame list and configure the host controller via the USB2.0_CMD register Frame List Size field. The frame list must always be aligned on a 4K page boundary. This requirement ensures that the frame list is always physically contiguous. RO 00 64-bit Addressing Capability -- This field documents the addressing range capability of this implementation. The value of this field determines whether software should use the 32-bit or 64-bit data structures. RO This bit is hardwired to 1. The PCH supports 64 bit addressing only. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 571 10.2.2 Host Controller Operational Registers This section defines the enhanced host controller operational registers. These registers are located after the capabilities registers. The operational register base must be dword-aligned and is calculated by adding the value in the first capabilities register (CAPLENGTH) to the base address of the enhanced host controller register address space (MEM_BASE). Since CAPLENGTH is always 20h, Table 10-44 accounts for this offset. All registers are 32 bits in length. Table 10-44. Enhanced Host Controller Operational Register Address Map View BAR BDF Offset Start Offset End Register ID: Register Name Default Value PCI MBAR B:29:0 20h 23h "Offset 20h: USB2.0_CMD--USB 2.0 Command Register (USB EHCI--B0:D29:F0)" on page 573 00080000h PCI MBAR B:29:0 24h 27h "Offset 24h: USB2.0_STS--USB 2.0 Status Register (USB EHCI--B0:D29:F0)" on page 576 00001000h PCI MBAR B:29:0 28h 2Bh "Offset 28h: USB2.0_INTR--USB 2.0 Interrupt Enable Register (USB EHCI--B0:D29:F0)" on page 579 00000000h PCI MBAR B:29:0 2Ch 2Fh "Offset 2Ch: FRINDEX--Frame Index Register (USB EHCI--B0:D29:F0)" on page 581 00000000h PCI MBAR B:29:0 30h 33h "Offset 30h: CTRLDSSEGMENT--Control Data Structure Segment Register (USB EHCI-- B0:D29:F0)" on page 582 00000000h PCI MBAR B:29:0 34h 37h "Offset 34h: PERIODICLISTBASE--Periodic Frame List Base Address Register (USB EHCI-- B0:D29:F0)" on page 583 00000000h PCI MBAR B:29:0 38h 3Bh "Offset 38h: ASYNCLISTADDR--Current Asynchronous List Address Register (USB EHCI-- B0:D29:F0)" on page 584 00000000h PCI MBAR B:29:0 60h 63h "Offset 60h: CONFIGFLAG--Configure Flag Register (USB EHCI--B0:D29:F0)" on page 585 00000000h PCI MBAR B:29:0 64h 67h "Offset 64h: PORTSC--Port N Status and Control Register (USB EHCI--B0:D29:F0)" on page 586 00003000h PCI MBAR B:29:0 A0h A0h "Offset A0h: CNTL_STS--Control/Status Register (USB EHCI--B0:D29:F0)" on page 592 20h PCI MBAR B:29:0 A4h A7h "Offset A4h: USBPID--USB PIDs Register (USB EHCI--B0:D29:F0)" on page 594 00000000h PCI MBAR B:29:0 A8h AFh "Offset A8h: DATABUF[7:0]--Data Buffer Bytes[7:0] Register (USB EHCI--B0:D29:F0)" on page 595 000000000 0000000h PCI MBAR B:29:0 B0h B3h "Offset B0h: CONFIG--Configuration Register (USB EHCI--B0:D29:F0)" on page 595 00007F01h Note: Software must read and write these registers using only dword accesses.These registers are divided into two sets. The first set at offsets MEM_BASE + 00:3Bh are implemented in the core power well. Unless otherwise noted, the core well registers are reset by the assertion of any of the following: * Core well hardware reset * HCRESET * D3-to-D0 reset Intel(R) Communications Chipset 89xx Series - Datasheet 572 October 2012 Order Number: 327879-001US 10.0 The second set at offsets MEM_BASE + 60h to the end of the implemented register space are implemented in the Suspend power well. Unless otherwise noted, the suspend well registers are reset by the assertion of either of the following: * Suspend well hardware reset * HCRESET 10.2.2.1 Offset 20h: USB2.0_CMD--USB 2.0 Command Register (USB EHCI-- B0:D29:F0) The Command Register indicates the command to be executed by the serial bus host controller. Writing to the register causes a command to be executed. Table 10-45. Offset 20h: USB2.0_CMD--USB 2.0 Command Register (USB EHCI--B0:D29:F0) (Sheet 1 of 3) Description: View PCI : Size: 32 bit B0:D29 Offset Start: 20h Bus:Device:Function: :F0 Offset End: 23h BAR: MBAR Default: 00080000h Bit Range Bit Acronym 31 :24 Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Interrupt Threshold Control -- System software uses this field to select the maximum rate at which the host controller will issue interrupts. The only valid values are defined below. If software writes an invalid value to this register, the results are undefined. 23 :16 15 :14 13 12 11 :08 Reserved ASC Value Maximum Interrupt Interval 00h Reserved 01h 1 micro-frame 02h 2 micro-frames 04h 4 micro-frames 08h 8 micro-frames (default, equates to 1 ms) 10h 16 micro-frames (2 ms) 20h 32 micro-frames (4 ms) 40h 64 micro-frames (8 ms) RW Reserved Asynch Schedule Update -- There is no functionality associated with this bit. RW Periodic Schedule Prefetch Enable -- This bit is used by software to enable the host controller to prefetch the periodic schedule even in C0. 0 = Prefetch based pause enabled only when not in C0. 1 = Prefetch based pause enable in C0. Once software has written a 1b to this bit to enable periodic schedule prefetching, it must disable prefecthing by writing a 0b to this bit whenever periodic schedule updates are about to begin. Software should continue to dynamically disable and re-enable the prefetcher surrounding any updates to the periodic scheduler (i.e. until the host controller has been reset via a HCRESET). RW Unimplemented Asynchronous Park Mode Bits -- Hardwired to 000b indicating the host controller does not support this optional feature. RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 573 Table 10-45. Offset 20h: USB2.0_CMD--USB 2.0 Command Register (USB EHCI--B0:D29:F0) (Sheet 2 of 3) Description: View PCI : Size: 32 bit Bit Range B0:D29 Offset Start: 20h Bus:Device:Function: :F0 Offset End: 23h BAR: MBAR Default: 00080000h Bit Acronym Power Well: Bit Description Light Host Controller Reset -- Hardwired to 0. The PCH does not implement this optional reset. 07 Interrupt on Async Advance Doorbell -- This bit is used as a doorbell by software to tell the host controller to issue an interrupt the next time it advances asynchronous schedule. 0 = The host controller sets this bit to a 0 after it has set the Interrupt on Async Advance status bit (B0:D29:F0:CAPLENGTH + 24h, bit 5) in the USB2.0_STS register to a 1. 1 = Software must write a 1 to this bit to ring the doorbell. When the host controller has evicted all appropriate cached schedule state, it sets the Interrupt on Async Advance status bit in the USB2.0_STS register. If the Interrupt on Async Advance Enable bit in the USB2.0_INTR register (B0:D29:F0:CAPLENGTH + 28h, bit 5) is a 1 then the host controller will assert an interrupt at the next interrupt threshold. See the EHCI specification for operational details. 06 Sticky Bit Reset Value Bit Access RO RW Software should not write a 1 to this bit when the asynchronous schedule is inactive. Doing so will yield undefined results. 05 Asynchronous Schedule Enable -- This bit controls whether the host controller skips processing the Asynchronous Schedule. 0 = Do not process the Asynchronous Schedule 1 = Use the ASYNCLISTADDR register to access the Asynchronous Schedule. RW 04 Periodic Schedule Enable -- This bit controls whether the host controller skips processing the Periodic Schedule. 0 = Do not process the Periodic Schedule 1 = Use the PERIODICLISTBASE register to access the Periodic Schedule. RW Frame List Size -- The PCH hardwires this field to 00b because it only supports the 1024-element frame list size. RO 03 :02 Host Controller Reset -- This control bit used by software to reset the host controller. The effects of this on root hub registers are similar to a Chip Hardware Reset (i.e., RSMRST# assertion and PWROK deassertion on the PCH). When software writes a 1 to this bit, the host controller resets its internal pipelines, timers, counters, state machines, etc. to their initial value. Any transaction currently in progress on USB is immediately terminated. A USB reset is not driven on downstream ports. 01 HCRESET PCI configuration registers and Host controller capability registers are not effected by this reset. All operational registers, including port registers and port state machines are set to their initial values. Port ownership reverts to the companion host controller(s), with the side effects described in the EHCI spec. Software must re-initialize the host controller in order to return the host controller to an operational state. This bit is set to 0 by the host controller when the reset process is complete. Software cannot terminate the reset process early by writing a 0 to this register. Software should not set this bit to a 1 when the HCHalted bit (B0:D29:F0:CAPLENGTH + 24h, bit 12) in the USB2.0_STS register is a 0. Attempting to reset an actively running host controller will result in undefined behavior. This reset me be used to leave EHCI port test modes. Intel(R) Communications Chipset 89xx Series - Datasheet 574 RW October 2012 Order Number: 327879-001US 10.0 Table 10-45. Offset 20h: USB2.0_CMD--USB 2.0 Command Register (USB EHCI--B0:D29:F0) (Sheet 3 of 3) Description: View PCI : Size: 32 bit Bit Range B0:D29 Offset Start: 20h Bus:Device:Function: :F0 Offset End: 23h BAR: MBAR Default: 00080000h Power Well: Bit Acronym Bit Description Sticky Bit Reset Value Bit Access Run/Stop: 0 = Stop (default) 1 = Run. When set to a 1, the Host controller proceeds with execution of the schedule. The Host controller continues execution as long as this bit is set. When this bit is set to 0, the Host controller completes the current transaction on the USB and then halts. The HCHalted bit in the USB2.0_STS register indicates when the Host controller has finished the transaction and has entered the stopped state. Software should not write a 1 to this field unless the host controller is in the Halted state (i.e., HCHalted in the USBSTS register is a 1). The Halted bit is cleared immediately when the Run bit is set. The following table explains how the different combinations of Run and Halted should be interpreted: 00 RS RW Run/ Stop Halted Interpretation 0b 0b In the process of halting 0b 1b Halted 1b 0b Running 1b Invalid - the HCHalted bit clears immediately 1b Memory read cycles initiated by the EHC that receive any status other than Successful will result in this bit being cleared. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 575 10.2.2.2 Offset 24h: USB2.0_STS--USB 2.0 Status Register (USB EHCI-- B0:D29:F0) This register indicates pending interrupts and various states of the Host controller. The status resulting from a transaction on the serial bus is not indicated in this register. See the Interrupts description in section 4 of the EHCI specification for additional information concerning USB 2.0 interrupt conditions. For the writable bits, software must write a 1 to clear bits that are set. Writing a 0 has no effect. Table 10-46. Offset 24h: USB2.0_STS--USB 2.0 Status Register (USB EHCI--B0:D29:F0) (Sheet 1 of 3) Description: View: PCI Size: 32 bit Bus:Device:Function: B0:D29 :F0 BAR: MBAR Default: 00001000h Bit Range Bit Acronym 31 :16 Reserved Offset Start: 24h Offset End: 27h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Asynchronous Schedule Status This bit reports the current real status of the Asynchronous Schedule. 0 = Status of the Asynchronous Schedule is disabled. (Default) 1 = Status of the Asynchronous Schedule is enabled. 15 The Host controller is not required to immediately disable or enable the Asynchronous Schedule when software transitions the Asynchronous Schedule Enable bit (B0:D29:F0:CAPLENGTH + 20h, bit 5) in the USB2.0_CMD register. When this bit and the Asynchronous Schedule Enable bit are the same value, the Asynchronous Schedule is either enabled (1) or disabled (0). RO Periodic Schedule Status This bit reports the current real status of the Periodic Schedule. 0 = Status of the Periodic Schedule is disabled. (Default) 1 = Status of the Periodic Schedule is enabled. 14 The Host controller is not required to immediately disable or enable the Periodic Schedule when software transitions the Periodic Schedule Enable bit (B0:D29:F0:CAPLENGTH + 20h, bit 4) in the USB2.0_CMD register. When this bit and the Periodic Schedule Enable bit are the same value, the Periodic Schedule is either enabled (1) or disabled (0). 13 Reclamation This read-only status bit is used to detect an empty asynchronous schedule. The operational model and valid transitions for this bit are described in Section 4 of the EHCI Specification. RO 12 HCHalted: 0 = This bit is a 0 when the Run/Stop bit is a 1. 1 = The Host controller sets this bit to 1 after it has stopped executing as a result of the Run/Stop bit being set to 0, either by software or by the Host controller hardware (e.g., internal error). (Default) RO 11 :06 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 576 RO October 2012 Order Number: 327879-001US 10.0 Table 10-46. Offset 24h: USB2.0_STS--USB 2.0 Status Register (USB EHCI--B0:D29:F0) (Sheet 2 of 3) Description: View: PCI Size: 32 bit Bit Range BAR: MBAR Bus:Device:Function: B0:D29 :F0 Default: 00001000h Bit Acronym 05 04 Offset Start: 24h Offset End: 27h Power Well: Bit Description Sticky Interrupt on Async Advance -- System software can force the host controller to issue an interrupt the next time the host controller advances the asynchronous schedule by writing a 1 to the Interrupt on Async Advance Doorbell bit (B0:D29:F0:CAPLENGTH + 20h, bit 6) in the USB2.0_CMD register. This bit indicates the assertion of that interrupt source. Host System Error: 0 = No serious error occurred during a host system access involving the Host controller module 1 = The Host controller sets this bit to 1 when a serious error occurs during a host system access involving the Host controller module. A hardware interrupt is generated to the system. Memory read cycles initiated by the EHC that receive any status other than Successful will result in this bit being set. Bit Reset Value Bit Access RWC RWC When this error occurs, the Host controller clears the Run/ Stop bit in the USB2.0_CMDregister (B0:D29:F0:CAPLENGTH + 20h, bit 0) to prevent further execution of the scheduled TDs. A hardware interrupt is generated to the system (if enabled in the Interrupt Enable Register). 03 October 2012 Order Number: 327879-001US Frame List Rollover: 0 = No Frame List Index rollover from its maximum value to 0. 1 = The Host controller sets this bit to a 1 when the Frame List Index rolls over from its maximum value to 0. Since the PCH only supports the 1024-entry Frame List Size, the Frame List Index rolls over every time FRNUM13 toggles. RWC Intel(R) Communications Chipset 89xx Series - Datasheet 577 Table 10-46. Offset 24h: USB2.0_STS--USB 2.0 Status Register (USB EHCI--B0:D29:F0) (Sheet 3 of 3) Description: View: PCI Size: 32 bit Bit Range BAR: MBAR Bus:Device:Function: Default: 00001000h Bit Acronym Offset Start: 24h Offset End: 27h Power Well: Bit Description Sticky Bit Reset Value Bit Access 02 Port Change Detect -- This bit is allowed to be maintained in the Auxiliary power well. Alternatively, it is also acceptable that on a D3 to D0 transition of the EHCI HC device, this bit is loaded with the OR of all of the PORTSC change bits (including: Force port resume, overcurrent change, enable/disable change and connect status change). Regardless of the implementation, when this bit is readable (i.e., in the D0 state), it must provide a valid view of the Port Status registers. 0 = No change bit transition from a 0 to 1 or No Force Port Resume bit transition from 0 to 1 as a result of a J-K transition detected on a suspended port. 1 = The Host controller sets this bit to 1 when any port for which the Port Owner bit is set to 0 has a change bit transition from a 0 to 1 or a Force Port Resume bit transition from 0 to 1 as a result of a J-K transition detected on a suspended port. RWC 01 USBERRINT USB Error Interrupt: 0 = No error condition. 1 = The Host controller sets this bit to 1 when completion of a USB transaction results in an error condition (e.g., error counter underflow). If the TD on which the error interrupt occurred also had its IOC bit set, both this bit and Bit 0 are set. See the EHCI specification for a list of the USB errors that will result in this interrupt being asserted. RWC USBINT USB Interrupt: 0 = No completion of a USB transaction whose Transfer Descriptor had its IOC bit set. No short packet is detected. 1 = The Host controller sets this bit to 1 when the cause of an interrupt is a completion of a USB transaction whose Transfer Descriptor had its IOC bit set. The Host controller also sets this bit to 1 when a short packet is detected (actual number of bytes received was less than the expected number of bytes). RWC 00 Intel(R) Communications Chipset 89xx Series - Datasheet 578 B0:D29 :F0 October 2012 Order Number: 327879-001US 10.0 10.2.2.3 Offset 28h: USB2.0_INTR--USB 2.0 Interrupt Enable Register (USB EHCI--B0:D29:F0) This register enables and disables reporting of the corresponding interrupt to the software. When a bit is set and the corresponding interrupt is active, an interrupt is generated to the host. Interrupt sources that are disabled in this register still appear in the USB2.0_STS Register to allow the software to poll for events. Each interrupt enable bit description indicates whether or not it is dependent on the interrupt threshold mechanism (See Section 4 of the EHCI specification). Table 10-47. Offset 28h: USB2.0_INTR--USB 2.0 Interrupt Enable Register (USB EHCI-- B0:D29:F0) Description: View: PCI Size: 32 bit BAR: MBAR Bus:Device:Function: B0:D29 :F0 Default: 00000000h Bit Range Bit Acronym 31 :06 Reserved Offset Start: 28h Offset End: 2Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 05 Interrupt on Async Advance Enable: 0 = Disable. 1 = Enable. When this bit is a 1, and the Interrupt on Async Advance bit (B0:D29:F0:CAPLENGTH + 24h, bit 5) in the USB2.0_STS register is a 1, the host controller will issue an interrupt at the next interrupt threshold. The interrupt is acknowledged by software clearing the Interrupt on Async Advance bit. RW 04 Host System Error Enable: 0 = Disable. 1 = Enable. When this bit is a 1, and the Host System Error Status bit (B0:D29:F0:CAPLENGTH + 24h, bit 4) in the USB2.0_STS register is a 1, the host controller will issue an interrupt. The interrupt is acknowledged by software clearing the Host System Error bit. RW 03 Frame List Rollover Enable:. 0 = Disable. 1 = Enable. When this bit is a 1, and the Frame List Rollover bit (B0:D29:F0:CAPLENGTH + 24h, bit 3) in the USB2.0_STS register is a 1, the host controller will issue an interrupt. The interrupt is acknowledged by software clearing the Frame List Rollover bit. RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 579 Table 10-47. Offset 28h: USB2.0_INTR--USB 2.0 Interrupt Enable Register (USB EHCI-- B0:D29:F0) Description: View: PCI Size: 32 bit Bit Range BAR: MBAR Bus:Device:Function: Default: 00000000h Bit Acronym Offset Start: 28h Offset End: 2Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access 02 Port Change Interrupt Enable: 0 = Disable. 1 = Enable. When this bit is a 1, and the Port Change Detect bit (B0:D29:F0:CAPLENGTH + 24h, bit 2) in the USB2.0_STS register is a 1, the host controller will issue an interrupt. The interrupt is acknowledged by software clearing the Port Change Detect bit. RW 01 USB Error Interrupt Enable: 0 = Disable. 1 = Enable. When this bit is a 1, and the USBERRINT bit (B0:D29:F0:CAPLENGTH + 24h, bit 1) in the USB2.0_STS register is a 1, the host controller will issue an interrupt at the next interrupt threshold. The interrupt is acknowledged by software by clearing the USBERRINT bit in the USB2.0_STS register. RW 00 USB Interrupt Enable: 0 = Disable. 1 = Enable. When this bit is a 1, and the USBINT bit (B0:D29:F0:CAPLENGTH + 24h, bit 0) in the USB2.0_STS register is a 1, the host controller will issue an interrupt at the next interrupt threshold. The interrupt is acknowledged by software by clearing the USBINT bit in the USB2.0_STS register. RW Intel(R) Communications Chipset 89xx Series - Datasheet 580 B0:D29 :F0 October 2012 Order Number: 327879-001US 10.0 10.2.2.4 Offset 2Ch: FRINDEX--Frame Index Register (USB EHCI--B0:D29:F0) The SOF frame number value for the bus SOF token is derived or alternatively managed from this register. See Section 4 of the EHCI specification for a detailed explanation of the SOF value management requirements on the host controller. The value of FRINDEX must be within 125 s (1 micro-frame) ahead of the SOF token value. The SOF value may be implemented as an 11-bit shadow register. For this discussion, this shadow register is 11 bits and is named SOFV. SOFV updates every 8 micro-frames (1 millisecond). An example implementation to achieve this behavior is to increment SOFV each time the FRINDEX[2:0] increments from 0 to 1. Software must use the value of FRINDEX to derive the current micro-frame number, both for high-speed isochronous scheduling purposes and to provide the get microframe number function required to client drivers. Therefore, the value of FRINDEX and the value of SOFV must be kept consistent if chip is reset or software writes to FRINDEX. Writes to FRINDEX must also write-through FRINDEX[13:3] to SOFV[10:0]. To keep the update as simple as possible, software should never write a FRINDEX value where the three least significant bits are 111b or 000b. Note: sThis register is used by the host controller to index into the periodic frame list. The register updates every 125 microseconds (once each micro-frame). Bits [12:3] are used to select a particular entry in the Periodic Frame List during periodic schedule execution. The number of bits used for the index is fixed at 10 for the PCH since it only supports 1024-entry frame lists. This register must be written as a dword. Word and byte writes produce undefined results. This register cannot be written unless the Host controller is in the Halted state as indicated by the HCHalted bit (B0:D29:F0:CAPLENGTH + 24h, bit 12). A write to this register while the Run/Stop bit (B0:D29:F0:CAPLENGTH + 20h, bit 0) is set to a 1 (USB2.0_CMD register) produces undefined results. Writes to this register also effect the SOF value. See Section 4 of the EHCI specification for details. Table 10-48. Offset 2Ch: FRINDEX--Frame Index Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 32 bit BAR: MBAR Bus:Device:Function: B0:D29 :F0 Default: 00000000h Bit Range Bit Acronym 31 :14 Reserved 13 :00 October 2012 Order Number: 327879-001US Offset Start: 2Ch Offset End: 2Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Frame List Current Index/Frame Number -- The value in this register increments at the end of each time frame (e.g., micro-frame). Bits [12:3] are used for the Frame List current index. This means that each location of the frame list is accessed 8 times (frames or micro-frames) before moving to the next index. RW Intel(R) Communications Chipset 89xx Series - Datasheet 581 10.2.2.5 Offset 30h: CTRLDSSEGMENT--Control Data Structure Segment (USB EHCI--B0:D29:F0) Register This 32-bit register corresponds to the most significant address bits [63:32] for all EHCI data structures. Since the PCH hardwires the 64-bit Addressing Capability field in HCCPARAMS to 1, then this register is used with the link pointers to construct 64-bit addresses to EHCI control data structures. This register is concatenated with the link pointer from either the PERIODICLISTBASE, ASYNCLISTADDR, or any control data structure link field to construct a 64-bit address. This register allows the host software to locate all control data structures within the same 4 GB memory segment. Table 10-49. Offset 30h: CTRLDSSEGMENT--Control Data Structure Segment Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 32 bit Bit Range Bus:Device:Function: B0:D29 :F0 BAR: MBAR Default: 00000000h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access 31 :12 Upper Address[63:44] -- Hardwired to 0s. The PCH EHC is only capable of generating addresses up to 16 terabytes (44 bits of address). RO 11 :00 Upper Address[43:32] -- This 12-bit field corresponds to address bits 43:32 when forming a control data structure address. RW Intel(R) Communications Chipset 89xx Series - Datasheet 582 Offset Start: 30h Offset End: 33h October 2012 Order Number: 327879-001US 10.0 10.2.2.6 Offset 34h: PERIODICLISTBASE--Periodic Frame List Base Address Register (USB EHCI--B0:D29:F0) This 32-bit register contains the beginning address of the Periodic Frame List in the system memory. Since the PCH host controller operates in 64-bit mode (as indicated by the 1 in the 64-bit Addressing Capability field in the HCCSPARAMS register) (offset 08h, bit 0), then the most significant 32 bits of every control data structure address comes from the CTRLDSSEGMENT register. HCD loads this register prior to starting the schedule execution by the host controller. The memory structure referenced by this physical memory pointer is assumed to be 4-Kbyte aligned. The contents of this register are combined with the Frame Index Register (FRINDEX) to enable the Host controller to step through the Periodic Frame List in sequence. Table 10-50. Offset 34h: PERIODICLISTBASE--Periodic Frame List Base Address Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 32 bit BAR: MBAR Bus:Device:Function: B0:D29 :F0 Default: 00000000h Power Well: Bit Range Bit Acronym Bit Description 31 :12 BA (LOW) Base Address -- These bits correspond to memory address signals [31:12], respectively. 11 :00 Reserved Reserved October 2012 Order Number: 327879-001US Offset Start: 34h Offset End: 37h Sticky Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 583 10.2.2.7 Offset 38h: ASYNCLISTADDR--Current Asynchronous List Address Register (USB EHCI--B0:D29:F0) This 32-bit register contains the address of the next asynchronous queue head to be executed. Since the PCH host controller operates in 64-bit mode (as indicated by a 1 in 64-bit Addressing Capability field in the HCCPARAMS register) (offset 08h, bit 0), then the most significant 32 bits of every control data structure address comes from the CTRLDSSEGMENT register (offset 08h). Bits [4:0] of this register cannot be modified by system software and will always return 0s when read. The memory structure referenced by this physical memory pointer is assumed to be 32-byte aligned. Table 10-51. Offset 38h: ASYNCLISTADDR--Current Asynchronous List Address Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 32 bit Bus:Device:Function: B0:D29 :F0 BAR: MBAR Default: 00000000h Bit Range Bit Acronym 31 :05 LPL 04 :00 Reserved Power Well: Bit Description Link Pointer Low -- These bits correspond to memory address signals [31:5], respectively. This field may only reference a Queue Head (QH). Sticky Bit Reset Value Bit Access RW Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 584 Offset Start: 38h Offset End: 3Bh October 2012 Order Number: 327879-001US 10.0 10.2.2.8 Offset 60h: CONFIGFLAG--Configure Flag Register (USB EHCI-- B0:D29:F0) This register is in the suspend power well. It is only reset by hardware when the suspend power is initially applied or in response to a host controller reset. Table 10-52. Offset 60h: CONFIGFLAG--Configure Flag Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 32 bit BAR: MBAR B0:D29 :F0 Default: 00000000h Bit Range Bit Acronym 31 :01 Reserved 00 Bus:Device:Function: CF October 2012 Order Number: 327879-001US Offset Start: 60h Offset End: 63h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Configure Flag -- Host software sets this bit as the last action in its process of configuring the Host controller. This bit controls the default port-routing control logic. Bit values and side-effects are listed below. See section 4 of the EHCI spec for operation details. 0 = Port routing control logic default-routes each port to the UHCIs (default). 1 = Port routing control logic default-routes all ports to this host controller. RW Intel(R) Communications Chipset 89xx Series - Datasheet 585 10.2.2.9 Offset 64h: PORTSC--Port N Status and Control Register (USB EHCI-- B0:D29:F0) When RMH is disabled: Offset: Port Port Port Port Port Port 0: 1: 2: 3: 4: 5: MEM_BASE MEM_BASE MEM_BASE MEM_BASE MEM_BASE MEM_BASE + + + + + + 64h-67h 68h-6Bh 6Ch-6Fh 70h-73h 74h-77h 78h-7Bh When RMH is enabled: Offset: Port 0 RMH: MEM_BASE + 64h-67h Port 1 Debug Port: MEM_BASE + 68-6Bh Port 2 USB redirect (if enabled): MEM_BASE + 6C-6Fh Note: When RMH is enabled this register is associated with the upstream ports of the EHCI controller and does not represent downstream hub ports. USB Hub class commands must be used to determine RMH port status and enable test modes. See Chapter 11 of the USB Specification, Revision 2.0 for more details. Rate Matching Hub wake capabilities can be configured by the RMHWKCTL Register (RCBA+35B0h) located in the Chipset Configuration chapter. A host controller must implement one or more port registers. Software uses the N_Port information from the Structural Parameters Register to determine how many ports need to be serviced. All ports have the structure defined below. Software must not write to unreported Port Status and Control Registers. This register is in the suspend power well. It is only reset by hardware when the suspend power is initially applied or in response to a host controller reset. The initial conditions of a port are: * No device connected * Port disabled. When a device is attached, the port state transitions to the attached state and system software will process this as with any status change notification. See Section 4 of the EHCI specification for operational requirements for how change events interact with port suspend mode. Table 10-53. Offset 64h: PORTSC--Port N Status and Control Register (USB EHCI-- B0:D29:F0) (Sheet 1 of 5) Description: View: PCI Size: 32 bit B0:D29 Offset Start: 64h Bus:Device:Function: :F0 Offset End: 67h BAR: MBAR Default: 00003000h Bit Range Bit Acronym 31 :23 Reserved Power Well: Bit Description Bit Reset Value Bit Access Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 586 Sticky October 2012 Order Number: 327879-001US 10.0 Table 10-53. Offset 64h: PORTSC--Port N Status and Control Register (USB EHCI-- B0:D29:F0) (Sheet 2 of 5) Description: View: PCI Size: 32 bit Bit Range 22 BAR: MBAR Bus:Device:Function: B0:D29 Offset Start: 64h :F0 Offset End: 67h Default: 00003000h Power Well: Bit Acronym Bit Description Sticky WKOC_E Wake on Overcurrent Enable: 0 = Disable. (Default) 1= Enable. Writing this bit to a 1 enables the setting of the PME Status bit in the Power Management Control/Status Register (offset 54, bit 15) when the overcurrent Active bit (bit 4 of this register) is set. Bit Reset Value Bit Access RW 21 Wake on Disconnect Enable: 0 = Disable. (Default) 1 = Enable. Writing this bit to a 1 enables the setting of the WKDSCNNT_E PME Status bit in the Power Management Control/Status Register (offset 54, bit 15) when the Current Connect Status changes from connected to disconnected (i.e., bit 0 of this register changes from 1 to 0). RW 20 Wake on Connect Enable: 0 = Disable. (Default) 1 = Enable. Writing this bit to a 1 enables the setting of the PME Status bit in the Power Management Control/Status Register (offset 54, bit 15) when the Current Connect Status changes from disconnected to connected (i.e., bit 0 of this register changes from 0 to 1). RW WKCNNT_E Port Test Control -- When this field is 0s, the port is NOT operating in a test mode. A non-zero value indicates that it is operating in test mode and the specific test mode is indicated by the specific value. The encoding of the test mode bits are (0110b - 1111b are reserved): Value 19 :16 Maximum Interrupt Interval 0000b Test mode not enabled (default) 0001b Test J_STATE 0010b Test K_STATE 0011b Test SE0_NAK 0100b Test Packet 0101b FORCE_ENABLE RW See USB Specification Revision 2.0, Chapter 7 for details on each test mode. 15 :14 Reserved 13 12 PP October 2012 Order Number: 327879-001US Reserved Port Owner -- This bit unconditionally goes to a 0 when the Configured Flag bit in the USB2.0_CMD register makes a 0 to 1 transition. System software uses this field to release ownership of the port to a selected host controller (in the event that the attached device is not a high-speed device). Software writes a 1 to this bit when the attached device is not a high-speed device. A 1 in this bit means that a companion host controller owns and controls the port. See Section 4 of the EHCI Specification for operational details. RW Port Power -- Read-only with a value of 1. This indicates that the port does have power. RO Intel(R) Communications Chipset 89xx Series - Datasheet 587 Table 10-53. Offset 64h: PORTSC--Port N Status and Control Register (USB EHCI-- B0:D29:F0) (Sheet 3 of 5) Description: View: PCI Size: 32 bit Bit Range BAR: MBAR B0:D29 Offset Start: 64h :F0 Offset End: 67h Default: 00003000h Bit Acronym Power Well: Bit Description Line Status-- These bits reflect the current logical levels of the D+ (bit 11) and D- (bit 10) signal lines. These bits are used for detection of low-speed USB devices prior to the port reset and enable sequence. This field is valid only when the port enable bit is 0 and the current connect status bit is set to a 1. 00 = SE0 10 = J-state 01 = K-state 11 = Undefined 11 :10 09 Bus:Device:Function: Reserved Sticky Bit Reset Value Bit Access RO Reserved Port Reset -- When software writes a 1 to this bit (from a 0), the bus reset sequence as defined in the USB Specification, Revision 2.0 is started. Software writes a 0 to this bit to terminate the bus reset sequence. Software must keep this bit at a 1 long enough to ensure the reset sequence completes as specified in the USB Specification, Revision 2.0. 1 = Port is in Reset. 0 = Port is not in Reset. 08 When software writes a 0 to this bit, there may be a delay before the bit status changes to a 0. The bit status will not read as a 0 until after the reset has completed. If the port is in high-speed mode after reset is complete, the host controller will automatically enable this port (e.g., set the Port Enable bit to a 1). A host controller must terminate the reset and stabilize the state of the port within 2 milliseconds of software transitioning this bit from 0 to 1. RW For example: if the port detects that the attached device is high-speed during reset, then the host controller must have the port in the enabled state within 2 ms of software writing this bit to a 0. The HCHalted bit (B0:D29:F0:CAPLENGTH + 24h, bit 12) in the USB2.0_STS register should be a 0 before software attempts to use this bit. The host controller may hold Port Reset asserted to a 1 when the HCHalted bit is a 1. This bit is 0 if Port Power is 0 System software should not attempt to reset a port if the HCHalted bit in the USB2.0_STS register is a 1. Doing so will result in undefined behavior. Intel(R) Communications Chipset 89xx Series - Datasheet 588 October 2012 Order Number: 327879-001US 10.0 Table 10-53. Offset 64h: PORTSC--Port N Status and Control Register (USB EHCI-- B0:D29:F0) (Sheet 4 of 5) Description: View: PCI Size: 32 bit Bit Range BAR: MBAR Bus:Device:Function: B0:D29 Offset Start: 64h :F0 Offset End: 67h Default: 00003000h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access Suspend: 0 = Port not in suspend state.(Default) 1 = Port in suspend state. Port Enabled Bit and Suspend bit of this register define the port states as follows: 07 Port Enabled Suspend Port State 0 X Disabled 1 0 Enabled 1 1 Suspend RW When in suspend state, downstream propagation of data is blocked on this port, except for port reset. The bit status does not change until the port is suspended and that there may be a delay in suspending a port depending on the activity on the port. The host controller will unconditionally set this bit to a 0 when software sets the Force Port Resume bit to a 0 (from a 1). A write of 0 to this bit is ignored by the host controller. If host software sets this bit to a 1 when the port is not enabled (i.e., Port enabled bit is a 0) the results are undefined. 06 Force Port Resume: 0 = No resume (K-state) detected/driven on port. (Default) 1 = Resume detected/driven on port. Software sets this bit to a 1 to drive resume signaling. The Host controller sets this bit to a 1 if a J-to-K transition is detected while the port is in the Suspend state. When this bit transitions to a 1 because a J-toK transition is detected, the Port Change Detect bit (B0:D29:F0:CAPLENGTH + 24h, bit 2) in the USB2.0_STS register is also set to a 1. If software sets this bit to a 1, the host controller must not set the Port Change Detect bit. RW When the EHCI controller owns the port, the resume sequence follows the defined sequence documented in the USB Specification, Revision 2.0. The resume signaling (Full-speed 'K') is driven on the port as long as this bit remains a 1. Software must appropriately time the Resume and set this bit to a 0 when the appropriate amount of time has elapsed. Writing a 0 (from 1) causes the port to return to high-speed mode (forcing the bus below the port into a high-speed idle). This bit will remain a 1 until the port has switched to the highspeed idle. 05 Overcurrent Change -- The functionality of this bit is not dependent upon the port owner. Software clears this bit by writing a 1 to it. 0 = No change. (Default) 1 = There is a change to Overcurrent Active. 04 Overcurrent Active: 0 = This port does not have an overcurrent condition. (Default) 1 = This port currently has an overcurrent condition. This bit will automatically transition from 1 to 0 when the over current condition is removed. The PCH automatically disables the port when the overcurrent active bit is 1. October 2012 Order Number: 327879-001US RWC RO Intel(R) Communications Chipset 89xx Series - Datasheet 589 Table 10-53. Offset 64h: PORTSC--Port N Status and Control Register (USB EHCI-- B0:D29:F0) (Sheet 5 of 5) Description: View: PCI Size: 32 bit Bit Range BAR: MBAR Bus:Device:Function: Default: 00003000h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access 03 Port Enable/Disable Change -- For the root hub, this bit gets set to a 1 only when a port is disabled due to the appropriate conditions existing at the EOF2 point (See Chapter 11 of the USB Specification for the definition of a port error). This bit is not set due to the Disabled-to-Enabled transition, nor due to a disconnect. Software clears this bit by writing a 1 to it. 0 = No change in status. (Default). 1 = Port enabled/disabled status has changed. RWC 02 Port Enabled/Disabled -- Ports can only be enabled by the host controller as a part of the reset and enable. Software cannot enable a port by writing a 1 to this bit. Ports can be disabled by either a fault condition (disconnect event or other fault condition) or by host software. The bit status does not change until the port state actually changes. There may be a delay in disabling or enabling a port due to other host controller and bus events. 0 = Disable 1 = Enable (Default) RW 01 Connect Status Change -- This bit indicates a change has occurred in the port's Current Connect Status. Software sets this bit to 0 by writing a 1 to it. 0 = No change (Default). 1 = Change in Current Connect Status. The host controller sets this bit for all changes to the port device connect status, even if system software has not cleared an existing connect status change. For example, the insertion status changes twice before system software has cleared the changed condition, hub hardware will be "setting" an already-set bit (i.e., the bit will remain set). RWC 00 Current Connect Status -- This value reflects the current state of the port, and may not correspond directly to the event that caused the Connect Status Change bit (Bit 1) to be set. 0 = No device is present. (Default) 1 = Device is present on port. RO Intel(R) Communications Chipset 89xx Series - Datasheet 590 B0:D29 Offset Start: 64h :F0 Offset End: 67h October 2012 Order Number: 327879-001US 10.0 10.2.3 USB 2.0-Based Debug Port Registers The Debug port's registers are located in the same memory area, defined by the Base Address Register (MEM_BASE), as the standard EHCI registers. The base offset for the debug port registers (A0h) is declared in the Debug Port Base Offset Capability Register at Configuration offset 5Ah (B0:D29:F0:offset 5Ah). The specific EHCI port that supports this debug capability (Port 1 for B0:D29:F0) is indicated by a 4-bit field (bits 20-23) in the HCSPARAMS register of the EHCI controller. The address map of the Debug Port registers is shown in Table 10-54. Table 10-54. Debug Port Register Address Map MEM_BASE + Offset Mnemonic A0-A3h CNTL_STS A4-A7h USBPID A8-AFh DATABUF[7:0] B0-B3h CONFIG Register Name Default Type Control/Status 00000000h R/W, R/WC, RO USB PIDs 00000000h R/W, RO Data Buffer (Bytes 7:0) 00000000 00000000h R/W Configuration 00007F01h R/W Notes: 1. All of these registers are implemented in the core well and reset by PLTRST#, EHC HCRESET, and a EHC D3-to-D0 transition. 2. The hardware associated with this register provides no checks to ensure that software programs the interface correctly. How the hardware behaves when programmed improperly is undefined. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 591 10.2.3.1 Offset A0h: CNTL_STS--Control/Status Register (USB EHCI-- B0:D29:F0) Note: Software should do Read-Modify-Write operations to this register to preserve the contents of bits not being modified. This include Reserved bits. Note: To preserve the usage of Reserved bits in the future, software should always write the same value read from the bit until it is defined. Reserved bits will always return 0 when read. Table 10-55. Offset A0h: CNTL_STS--Control/Status Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit Bit Range 31 BAR: MBAR Default: 20h Bit Acronym Reserved 30 OWNER_CNT 29 Reserved 28 27 :17 16 15 :12 ENABLED_CNT Reserved DONE_STS LINK_ID_STS Offset Start: A0h Offset End: A0h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 0 = Ownership of the debug port is NOT forced to the EHCI controller (Default) 1 = Ownership of the debug port is forced to the EHCI controller (i.e. immediately taken away from the companion Classic USB Host controller) If the port was already owned by the EHCI controller, then setting this bit has no effect. This bit overrides all of the ownershiprelated bits in the standard EHCI registers. RW Reserved 0 = Software can clear this by writing a 0 to it. The hardware clears this bit for the same conditions where the Port Enable/Disable Change bit (in the PORTSC register) is set. (Default) 1 = Debug port is enabled for operation. Software can directly set this bit if the port is already enabled in the associated PORTSC register (this is enforced by the hardware). RW Reserved Software can clear this by writing a 1 to it. 0 = Request Not complete 1 = Set by hardware to indicate that the request is complete. This field identifies the link interface. 0h = Hardwired. Indicates that it is a USB Debug Port. RWC RO 11 Reserved 10 IN_USE_CNT Set by software to indicate that the port is in use. Cleared by software to indicate that the port is free and may be used by other software. This bit is cleared after reset. (This bit has no affect on hardware.) RW EXCEPTION_STS This field indicates the exception when the ERROR_GOOD#_STS bit is set. This field should be ignored if the ERROR_GOOD#_STS bit is 0. 000 =No Error. (Default) Note: This should not be seen since this field should only be checked if there is an error. 001 =Transaction error: Indicates the USB 2.0 transaction had an error (CRC, bad PID, timeout, etc.) 010 =Hardware error. Request was attempted (or in progress) when port was suspended or reset. All Other combinations are reserved RO 09 :07 Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 592 B0:D29 Bus:Device:Function: :F0 October 2012 Order Number: 327879-001US 10.0 Table 10-55. Offset A0h: CNTL_STS--Control/Status Register (USB EHCI--B0:D29:F0) Description: View: PCI BAR: MBAR Size: 8 bit Bit Range 06 05 Bus:Device:Function: B0:D29 :F0 Default: 20h Power Well: Bit Acronym Bit Description ERROR_GOOD#_STS 0 = Hardware clears this bit to 0 after the proper completion of a read or write. (Default) 1 = Error has occurred. Details on the nature of the error are provided in the Exception field. GO_CNT Offset Start: A0h Offset End: A0h 0 = Hardware clears this bit when hardware sets the DONE_STS bit. (Default) 1 = Causes hardware to perform a read or write request. Sticky Bit Reset Value Bit Access RO RW Writing a 1 to this bit when it is already set may result in undefined behavior. 04 03 :00 Software clears this bit to indicate that the current request is a read. Software sets this bit to indicate that the current WRITE_READ#_CNT request is a write. 0 = Read (Default) 1 = Write RW This field is used to indicate the size of the data to be transferred. default = 0h. For write operations, this field is set by software to indicate to the hardware how many bytes of data in Data Buffer are to be transferred to the console. A value of 0h indicates that a zero-length packet should be sent. A value of 1-8 indicates 1-8 bytes are to be transferred. Values 9-Fh are invalid and how hardware behaves if used is undefined. For read operations, this field is set by hardware to indicate to software how many bytes in Data Buffer are valid in response to a read operation. A value of 0h indicates that a zero length packet was returned and the state of Data Buffer is not defined. A value of 1-8 indicates 1-8 bytes were received. Hardware is not allowed to return values 9-Fh. The transferring of data always starts with byte 0 in the data area and moves toward byte 7 until the transfer size is reached. RW DATA_LEN_CNT October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 593 10.2.3.2 Offset A4h: USBPID--USB PIDs Register (USB EHCI--B0:D29:F0) This Dword register is used to communicate PID information between the USB debug driver and the USB debug port. The debug port uses some of these fields to generate USB packets, and uses other fields to return PID information to the USB debug driver. Table 10-56. Offset A4h: USBPID--USB PIDs Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 32 bit BAR: MBAR Bus:Device:Function: Default: 00000000h Bit Range Bit Acronym 31 :24 Reserved Offset Start: A4h Offset End: A7h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 23 :16 Hardware updates this field with the received PID for transactions in either direction. When the controller is writing data, this field is updated with the handshake PID RECEIVED_PID_ that is received from the device. When the host controller STS[23:16] is reading data, this field is updated with the data packet PID (if the device sent data), or the handshake PID (if the device NAKs the request). This field is valid when the hardware clears the GO_DONE#_CNT bit. RO 15 :08 Hardware sends this PID to begin the data packet when SEND_PID_CNT sending data to USB (i.e., WRITE_READ#_CNT is [15:8] asserted). Software typically sets this field to either DATA0 or DATA1 PID values. RW 07 :00 sends this PID as the Token PID for each USB TOKEN_PID_CN Hardware transaction. Software typically sets this field to either IN, T[7:0] OUT, or SETUP PID values. RW Intel(R) Communications Chipset 89xx Series - Datasheet 594 B0:D29 :F0 October 2012 Order Number: 327879-001US 10.0 10.2.3.3 Offset A8h: DATABUF[7:0]--Data Buffer Bytes[7:0] Register (USB EHCI--B0:D29:F0) This register can be accessed as eight separate 8-bit registers or two separate 32-bit registers. Table 10-57. Offset A8h: DATABUF[7:0]--Data Buffer Bytes[7:0] Register (USB EHCI-- B0:D29:F0) Description: View: PCI Size: 64 bit Bit Range BAR: MBAR B0:D29 :F0 Default: 0000000000000000h Bit Acronym Offset Start: A8h Offset End: AFh Power Well: Bit Description Sticky Bit Reset Value Bit Access DATABUFFER[63:0] -- his field is the 8 bytes of the data buffer. Bits 7:0 correspond to least significant byte (byte 0). Bits 63:56 correspond to the most significant byte (byte 7). The bytes in the Data Buffer must be written with data before software initiates a write request. For a read request, the Data Buffer contains valid data when DONE_STS bit (offset A0, bit 16) is cleared by the hardware, ERROR_GOOD#_STS (offset A0, bit 6) is cleared by the hardware, and the DATA_LENGTH_CNT field (offset A0, bits 3:0) indicates the number of bytes that are valid. 63 :00 10.2.3.4 Bus:Device:Function: RW Offset B0h: CONFIG--Configuration Register (USB EHCI--B0:D29:F0) Table 10-58. Offset B0h: CONFIG--Configuration Register (USB EHCI--B0:D29:F0) Description: View: PCI Size: 8 bit BAR: MBAR Default: 00007F01h Bit Range Bit Acronym 31 :15 Reserved 14 :08 07 :04 03 :00 Bus:Device:Function: B0:D29 :F0 Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved This 7-bit field identifies the USB device address USB_ADDRESS_CNF used by the controller for all Token PID generation. (Default = 7Fh) Reserved Offset Start: B0h Offset End: B3h RW Reserved This 4-bit field identifies the EndPoint used by the USB_EndPoint_CNF controller for all Token PID generation. (Default = 1h) RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 595 11.0 SMBus Controller Registers (B0:D31:F3) 11.1 PCI Configuration Registers (SMBus--B0:D31:F3) 11.1.1 SMBus Controller PCI Register Address Map Table 11-1. SMBus Controller PCI Register Address Map Offset Start 00h 02h Offset End 01h 03h Default Value Register ID - Description "Offset 00h: Vendor Identification Register (SMBus--B0:D31:F3)" on page 597 8086h "Offset 02h: Device Identification Register (SMBus--B0:D31:F3)" on page 597 See register description 04h 05h "Offset 04h: PCI Command Register (SMBus--B0:D31:F3)" on page 598 0000h 06h 07h "Offset 06h: PCI Status Register (SMBus--B0:D31:F3)" on page 599 0280h 08h 08h "Offset 08h: Revision Identification Register (SMBus--B0:D31:F3)" on page 600 See register description 09h 09h "Offset 09h: Programming Interface Register (SMBus--B0:D31:F3)" on page 600 00h 0Ah 09h "Offset 0Ah: Sub Class Code Register (SMBus--B0:D31:F3)" on page 600 05h 0Bh 0Bh "Offset 0Bh: Base Class Code Register (SMBus--B0:D31:F3)" on page 601 0Ch 10h 13h "Offset 10h: SMBus Memory Base Address 0 (SMBus--B0:D31:F3)" on page 601 00000004h 14h 17h "Offset 14h: SMBus Memory Base Address 1 (SMBus--B0:D31:F3)" on page 602 00000000h 20h 23h "Offset 20h: SMBus Base Address Register (SMBus--B0:D31:F3)" on page 602 00000001h 2Dh "Offset 2Ch: Subsystem Vendor Identification Register (SMBus--B0:D31:F3)" on page 603 0000h 2Ch 2Eh 2Fh "Offset 2Eh: Subsystem Identification Register (SMBus--B0:D31:F3)" on page 603 0000h 3Ch 3Ch "Offset 3Ch: Interrupt Line Register (SMBus--B0:D31:F3)" on page 604 00h 3Dh 3Dh "Offset 3Dh: Interrupt Pin Register (SMBus--B0:D31:F3)" on page 604 See register description 40h 40h "Offset 40h: Host Configuration Register (SMBus--B0:D31:F3)" on page 605 00h Note: Registers that are not shown should be treated as Reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 596 October 2012 Order Number: 327879-001US 11.0 11.1.1.1 Offset 00h: Vendor Identification Register (SMBus--B0:D31:F3) Table 11-2. Offset 00h: Vendor Identification Register (SMBus--B0:D31:F3) Description: View: PCI Size: 16 bit BAR: Configuration B0:D31 :F3 Default: 8086h Bit Range Bit Acronym 15 :00 VID 11.1.1.2 Bus:Device:Function: Offset Start: 00h Offset End: 01h Power Well: Core Bit Description Sticky Vendor Identification: This 16-bit value is assigned to Intel. Intel VID = 8086h Bit Reset Value Bit Access 8086h RO Offset 02h: Device Identification Register (SMBus--B0:D31:F3) Table 11-3. Offset 02h: Device Identification Register (SMBus--B0:D31:F3) Description: View: PCI Size: 16 bit Bus:Device:Function: B0:D31 :F3 BAR: Configuration Default: See register description Bit Range Bit Acronym 15 :00 DID October 2012 Order Number: 327879-001US Offset Start: 02h Offset End: 03h Power Well: Core Bit Description Sticky Device Identification: Indicates the device number assigned by the SIG. Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 597 11.1.1.3 Offset 04h: PCI Command Register (SMBus--B0:D31:F3) Table 11-4. Offset 04h: PCI Command Register (SMBus--B0:D31:F3) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: Default: 0000h Bit Range Bit Acronym 15 :11 Reserved 10 Offset Start: 04h Offset End: 05h Power Well: Core Bit Description Reserved Sticky Bit Reset Value Bit Access 00h Interrupt Disable -- This disables pin-based INTx# interrupts. This bit has no effect on MSI operation. 0 = Internal INTx# messages are generated if there is an interrupt and MSI is not enabled. 1 = Internal INTx# messages will not be generated. 0h RW Fast Back to Back Enable -- Reserved as 0. 0h RO 09 FBE 08 SERR_EN SERR# Enable -- Reserved as 0. 0h RW 07 WCC Wait Cycle Control -- Reserved as 0. 0h RO 06 PER Parity Error Response: 0 = Disabled. SATA controller will not generate PERR# when a data parity error is detected. 1 = Enabled. SATA controller will generate PERR# when a data parity error is detected. 0h RW 05 VPS VGA Palette Snoop -- Reserved as 0. 0h RO 04 PMWE Postable Memory Write Enable -- Reserved as 0. 0h RO 03 SCE Special Cycle Enable -- Reserved as 0. 0h RO 02 BME Bus Master Enable -- This bit controls the PCH's ability to act as a PCI master for IDE Bus Master transfers. This bit does not impact the generation of completions for split transaction commands. 1 RW 01 MSE Memory Space Enable -- Controls access to the SATA controller's target memory space (for AHCI). This bit is RO `0' when not in AHCI mode. 1 RW/RO IOSE I/O Space Enable -- This bit controls access to the I/O space registers. 0 = Disables access to the Legacy or Native IDE ports (both Primary and Secondary) as well as the Bus Master I/ O registers. 1 = Enable. The Base Address register for the Bus Master registers should be programmed before this bit is set. 1 RW 00 Intel(R) Communications Chipset 89xx Series - Datasheet 598 B0:D31 :F3 October 2012 Order Number: 327879-001US 11.0 11.1.1.4 Offset 06h: PCI Status Register (SMBus--B0:D31:F3) Table 11-5. Offset 06h: PCI Status Register (SMBus--B0:D31:F3) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D31 :F3 Default: 0280h Offset Start: 06h Offset End: 07h Power Well: Bit Description 15 DPE Detected Parity Error: 0 = Parity error not detected. 1 = Indicates that the PCH detected a parity error on the internal backbone. This bit gets set even if the Parity Error Response bit (D30:F0:04 bit 6) is not set. RWC 14 SSE Signaled System Error -- Set when the LPC bridge signals a system error to the internal SERR# logic. RWC 13 RMA Received Master Abort: 0 = No master abort received. 1 = Set when the bridge receives a master abort status from the I/O data bus. RWC 12 RTA Received Target Abort -- Hardwired to 0. RC 11 STA Signaled Target Abort -- Reserved as 0. RO DEVSEL# Timing Status: 01 = Hardwired; Controls the device select time for the SATA controller's PCI interface. RO 10 :09 DEV_STS Sticky Bit Reset Value Bit Acronym Bit Access 08 DPED Data Parity Error Detected -- For PCH, this bit can only be set on read completions received from the bus when there is a parity error. SATA controller, as a master, either detects a parity error or sees the parity error line asserted, and the parity error response bit (bit 6 of the command register) is set. 07 FB2BC Fast Back to Back Capable -- Hardwired to 1. RO 06 UDF User Definable Features -- Hardwired to 0. RO 05 66MHZ_CAP 04 CAP_LIST 03 INTS 02 :00 Reserved October 2012 Order Number: 327879-001US RWC 66 MHz Capable -- Hardwired to 0. RO Capabilities List -- Hardwired to 0 because there are no capability list structures in this function RO Interrupt Status -- This bit indicates that an interrupt is pending. It is independent from the state of the Interrupt Enable bit in the PCI Command register. RO Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 599 11.1.1.5 Offset 08h: Revision Identification Register (SMBus--B0:D31:F3) Table 11-6. Offset 08h: Revision Identification Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit BAR: Configuration B0:D31 :F3 Default: See register description Bit Range Bit Acronym 07 :00 RID 11.1.1.6 Bus:Device:Function: Offset Start: 08h Offset End: 08h Power Well: Bit Description Sticky Revision ID -- This is an 8-bit value indicating the stepping of the SMBus controller hardware. Bit Reset Value Bit Access Variable RO Offset 09h: Programming Interface Register (SMBus--B0:D31:F3) Table 11-7. Offset 09h: Programming Interface Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit Default: 00h Bit Range Bit Acronym 07 :00 Reserved 11.1.1.7 B0:D31 Bus:Device:Function: :F3 BAR: Configuration Offset Start: 09h Offset End: 09h Power Well: Bit Description Sticky Reserved Bit Reset Value Bit Access 00h Offset 0Ah: Sub Class Code Register (SMBus--B0:D31:F3) Table 11-8. Offset 0Ah: Sub Class Code Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit B0:D31 Bus:Device:Function: :F3 BAR: Configuration Default: 05h Bit Range Bit Acronym 07 :00 SCC Power Well: Bit Description Sub Class Code -- SMBus serial controller Intel(R) Communications Chipset 89xx Series - Datasheet 600 Offset Start: 0Ah Offset End: 09h Sticky Bit Reset Value Bit Access 05h RO October 2012 Order Number: 327879-001US 11.0 11.1.1.8 Offset 0Bh: Base Class Code Register (SMBus--B0:D31:F3) Table 11-9. Offset 0Bh: Base Class Code Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit BAR: Configuration B0:D31 :F3 Default: 0Ch Bit Range Bit Acronym 07 :00 BCC 11.1.1.9 Bus:Device:Function: Offset Start: 0Bh Offset End: 0Bh Power Well: Bit Description Sticky Base Class Code -- Serial controller Bit Reset Value Bit Access 0Ch RO Offset 10h: SMBus Memory Base Address 0 (SMBus--B0:D31:F3) Table 11-10. Offset 10h: SMBus Memory Base Address 0 (SMBus--B0:D31:F3) Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000004h Bit Range Bit Acronym 31 :08 BA 07 :04 Reserved 03 02 :01 Bus:Device:Function: B0:D31 :F3 Offset Start: 10h Offset End: 13h Power Well: Bit Description Sticky Base Address -- Provides the 32 byte system memory base address for the PCH SMB logic. Bit Reset Value Bit Access RW Reserved PREF Prefetchable -- Hardwired to 0. Indicates that SMBMBAR is not pre-fetchable. RO ADDRNG Address Range -- Indicates that this SMBMBAR can be located anywhere in 64 bit address space. Hardwired to 10b. RO Memory Space Indicator -- This read-only bit always is 0, indicating that the SMB logic is Memory mapped. RO 00 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 601 11.1.1.10 Offset 14h: SMBus Memory Base Address 1 (SMBus--B0:D31:F3) Table 11-11. Offset 14h: SMBus Memory Base Address 1 (SMBus--B0:D31:F3) Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: Default: 00000000h Offset Start: 14h Offset End: 17h Power Well: Bit Range Bit Acronym Bit Description 31 :00 BA Base Address -- Provides bits 63-32 system memory base address for the PCH SMB logic. 11.1.1.11 B0:D31 :F3 Sticky Bit Reset Value Bit Access RW Offset 20h: SMBus Base Address Register (SMBus--B0:D31:F3) Table 11-12. Offset 20h: SMBus Base Address Register (SMBus--B0:D31:F3) Description: View: PCI Size: 32 bit Default: 00000001h Bit Range Bit Acronym 31 :16 Reserved 15 :05 BA 04 :01 Reserved 00 B0:D31 Bus:Device:Function: :F3 BAR: Configuration Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Base Address -- This field provides the base address of the I/O space (16 consecutive I/O locations). RW Reserved IO Space Indicator -- Hardwired to 1 indicating that the SMB logic is I/O mapped. Intel(R) Communications Chipset 89xx Series - Datasheet 602 Offset Start: 20h Offset End: 23h RO October 2012 Order Number: 327879-001US 11.0 11.1.1.12 Offset 2Ch: Subsystem Vendor Identification Register (SMBus-- B0:D31:F2/F4) Table 11-13. Offset 2Ch: Subsystem Vendor Identification Register (SMBus--B0:D31:F3) Description: View: PCI Size: 16 bit Bit Range 15 :00 11.1.1.13 BAR: Configuration Bus:Device:Function: B0:D31:F 3 Default: 0000h Bit Acronym Offset Start: 2Ch Offset End: 2Dh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Subsystem Vendor ID -- The SVID register, in combination with the Subsystem ID (SID) register, enables the operating system (OS) to distinguish subsystems from each other. The value returned by reads to this register is the same as that which was written by BIOS into the IDE SVID register. Software can write to this register only once per core well reset. Writes should be done as a single 16-bit cycle. SVID RO Offset 2Eh: Subsystem Identification Register (SMBus--B0:D31:F3) Table 11-14. Offset 2Eh: Subsystem Identification Register (SMBus--B0:D31:F3) Description: View: PCI Size: 16 bit Bit Range 15 :00 BAR: Configuration Bus:Device:Function: B0:D31:F 3 Default: 0000h Power Well: Core Bit Acronym Bit Description SID Subsystem ID -- The SID register, in combination with the SVID register, enables the operating system (OS) to distinguish subsystems from each other. The value returned by reads to this register is the same as that which was written by BIOS into the IDE SID register. Software can write to this register only once per core well reset. Writes should be done as a single 16-bit cycle. October 2012 Order Number: 327879-001US Offset Start: 2Eh Offset End: 2Fh Sticky Bit Reset Value Bit Access RWO Intel(R) Communications Chipset 89xx Series - Datasheet 603 11.1.1.14 Offset 3Ch: Interrupt Line Register (SMBus--B0:D31:F3) Table 11-15. Offset 3Ch: Interrupt Line Register (SMBus--B0:D31:F3) Description: View: PCI BAR: Configuration Size: 8 bit B0:D3 1:F3 Default: 00h Bit Acronym Bit Range Offset Start: 3Ch Offset End: 3Ch Power Well: Bit Description Sticky Bit Reset Value Bit Access Interrupt Line -- This data is not used by the PCH. INT_LN It is to communicate to software the interrupt line that the interrupt pin is connected to PIRQB#. 07 :00 11.1.1.15 Bus:Device:Function: RW Offset 3Dh: Interrupt Pin Register (SMBus--B0:D31:F3) Table 11-16. Offset 3Dh: Interrupt Pin Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit Bit Range 07 :00 Bus:Device:Function: B0:D3 1:F3 BAR: Configuration Default: See register description Bit Acronym Bit Description INT_PN Interrupt PIN -- This reflects the value of D31IP.SMIP in chipset configuration space. Intel(R) Communications Chipset 89xx Series - Datasheet 604 Offset Start: 3Dh Offset End: 3Dh Power Well: Sticky Bit Reset Value Bit Access RO October 2012 Order Number: 327879-001US 11.0 11.1.1.16 Offset 40h: Host Configuration Register (SMBus--B0:D31:F3) Table 11-17. Offset 40h: Host Configuration Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B0:D3 1:F3 Default: 00h Offset Start: 40h Offset End: 40h Power Well: Bit Acronym 07 :04 Reserved Reserved 03 SSRESET Soft SMBus Reset: 0 = The HW will reset this bit to 0 when SMBus reset operation is completed. 1 = The SMBus state machine and logic in the PCH is reset. RW 02 I2C_EN 0 = SMBus behavior. 1 = The PCH is enabled to communicate with I2C devices. This will change the formatting of some commands. RW 0 = SMBus interrupts will not generate an SMI# 1 = Any source of an SMB interrupt will instead be routed to generate an SMI#. This bit needs to be set for SMBALERT# to be enabled. RW SMBus Host Enable: 0 = Disable the SMBus Host controller 1 = Enable. The SMB Host controller interface is enabled to execute commands. The INTREN bit (offset SMBASE + 02h, bit 0) needs to be enabled for the SMB Host controller to interrupt or SMI#. The SMB Host controller will not respond to any new requests until all interrupt requests have been cleared. RW 01 00 MB_SMI_EN HST_EN October 2012 Order Number: 327879-001US Bit Description Sticky Bit Reset Value Bit Range Bit Access Intel(R) Communications Chipset 89xx Series - Datasheet 605 11.2 SMBus I/O and Memory Mapped I/O Registers The following SMBus registers can be accessed through I/O BAR or Memory BAR registers in PCI configuration space. The offsets are the same for both I/O and Memory Mapped I/O registers. 11.2.1 SMBus Registers Table 11-18. SMBus Bus Registers Offset Start 00h Offset End 00h Default Value Register ID - Description "Offset 00h: Host Status Register (SMBus--B0:D31:F3)" on page 607 00h 02h 02h "Offset 02h: Host Control Register (SMBus--B0:D31:F3)" on page 609 00h 03h 03h "Offset 03h: Host Command Register (SMBus--B0:D31:F3)" on page 611 00h 04h 04h "Offset 04h: Transmit Slave Address Register (SMBus--B0:D31:F3)" on page 611 00h 05h 05h "Offset 05h: Host Data 0 Register (SMBus--B0:D31:F3)" on page 612 00h 06h 06h "Offset 06h: Host Data 1 Register (SMBus--B0:D31:F3)" on page 612 00h 07h 07h "Offset 07h: Host Block Data Byte Register (SMBus--B0:D31:F3)" on page 613 00h 08h 08h "Offset 08h: Packet Error Check (PEC) Register (SMBus--B0:D31:F3)" on page 614 00h 09h 09h "Offset 09h: Receive Slave Address Register (SMBus--B0:D31:F3)" on page 614 44h 0Ah 0Bh "Offset 0Ah: Receive Slave Data Register (SMBus--B0:D31:F3)" on page 615 0000h 0Ch 0Ch "Offset 0Ch: Auxiliary Status Register (SMBus--B0:D31:F3)" on page 615 00h 0Dh 0Dh "Offset 0Dh: Auxiliary Control Register (SMBus--B0:D31:F3)" on page 616 00h 0Eh 0Eh "Offset 0Eh: SMLink Pin Control Register (SMBus--B0:D31:F3)" on page 617 See register description 0Fh 0Fh "Offset 0Fh: SMBus Pin Control Register (SMBus--B0:D31:F3)" on page 618 See register description 10h 10h "Offset 10h: Slave Status Register (SMBus--B0:D31:F3)" on page 619 00h 11h 11h "Offset 11h: Slave Command Register (SMBus--B0:D31:F3)" on page 620 00h 14h 14h "Offset 14h: Notify Device Address Register (SMBus--B0:D31:F3)" on page 621 00h 16h 16h "Offset 16h: Notify Data Low Byte Register (SMBus--B0:D31:F3)" on page 621 00h 17h 17h "Offset 17h: Notify Data High Byte Register (SMBus--B0:D31:F3)" on page 622 00h Intel(R) Communications Chipset 89xx Series - Datasheet 606 October 2012 Order Number: 327879-001US 11.0 11.2.1.1 Offset 00h: Host Status Register (SMBus--B0:D31:F3) All status bits are set by hardware and cleared by the software writing a one to the particular bit position. Writing a 0 to any bit position has no effect. Table 11-19. Offset 00h: Host Status Register (SMBus--B0:D31:F3) (Sheet 1 of 2) Description: View: PCI Size: 8 bit Bit Range BAR: SMBASE (IO) Bus:Device:Function: B0:D3 1:F3 Default: 00h Bit Acronym Offset Start: 00h Offset End: 00h Power Well: Bit Description Sticky Bit Reset Value Bit Access Byte Done Status: 0 = Software can clear this by writing a 1 to it. 1 = Host controller received a byte (for Block Read commands) or if it has completed transmission of a byte (for Block Write commands) when the 32-byte buffer is not being used. This bit will be set, even on the last byte of the transfer. This bit is not set when transmission is due to the LAN interface heartbeat. This bit has no meaning for block transfers when the 32-byte buffer is enabled. 07 06 05 DS INUSE_STS When the last byte of a block message is received, the host controller will set this bit. However, it will not immediately set the INTR bit (bit 1 in this register). When the interrupt handler clears the DS bit, the message is considered complete, and the host controller will then set the INTR bit (and generate another interrupt). Thus, for a block message of n bytes, the PCH will generate n+1 interrupts. The interrupt handler needs to be implemented to handle these cases. When not using the 32 Byte Buffer, hardware will drive the SMBCLK signal low when the DS bit is set until SW clears the bit. This includes the last byte of a transfer. Software must clear the DS bit before it can clear the BUSY bit. This bit is used as semaphore among various independent software threads that may need to use the PCH's SMBus logic, and has no other effect on hardware. 0 = After a full PCI reset, a read to this bit returns a 0. 1 = After the first read, subsequent reads will return a 1. A write of a 1 to this bit will reset the next read value to 0. Writing a 0 to this bit has no effect. Software can poll this bit until it reads a 0, and will then own the usage of the host controller. 0 = Interrupt or SMI# was not generated by SMBALERT#. Software clears this bit by writing a 1 to it. SMBALERT_STS 1 = The source of the interrupt or SMI# was the SMBALERT# signal. This bit is only cleared by software writing a 1 to the bit position or by RSMRST# going low. If the signal is programmed as a GPIO, then this bit will never be set. 04 FAILED 03 BUS_ERR October 2012 Order Number: 327879-001US RWC RW RWC 0 = Software clears this bit by writing a 1 to it. 1 = The source of the interrupt or SMI# was a failed bus transaction. This bit is set in response to the KILL bit being set to terminate the host transaction. RWC 0 = Software clears this bit by writing a 1 to it. 1 = The source of the interrupt of SMI# was a transaction collision. RWC Intel(R) Communications Chipset 89xx Series - Datasheet 607 Table 11-19. Offset 00h: Host Status Register (SMBus--B0:D31:F3) (Sheet 2 of 2) Description: View: PCI Size: 8 bit Bit Range 02 01 00 BAR: SMBASE (IO) B0:D3 1:F3 Default: 00h Offset Start: 00h Offset End: 00h Power Well: Sticky Bit Reset Value Bit Acronym Bit Description DEV_ERR 0 = Software clears this bit by writing a 1 to it. The PCH will then deassert the interrupt or SMI#. 1 = The source of the interrupt or SMI# was due to one of the following: Invalid Command Field, Unclaimed Cycle (host initiated), Host Device Time-out Error. RWC INTR This bit can only be set by termination of a command. INTR is not dependent on the INTREN bit (offset SMBASE + 02h, bit 0) of the Host controller register (offset 02h). It is only dependent on the termination of the command. If the INTREN bit is not set, then the INTR bit will be set, although the interrupt will not be generated. Software can poll the INTR bit in this noninterrupt case. 0 = Software clears this bit by writing a 1 to it. The PCH then deasserts the interrupt or SMI#. 1 = The source of the interrupt or SMI# was the successful completion of its last command. RWC HOST_BUSY 0 = Cleared by the PCH when the current transaction is completed. 1 = Indicates that the PCH is running a command from the host interface. No SMB registers should be accessed while this bit is set, except the BLOCK DATA BYTE Register. The BLOCK DATA BYTE Register can be accessed when this bit is set only when the SMB_CMD bits in the Host Control Register are programmed for Block command or I2C Read command. This is necessary in order to check the DONE_STS bit. RWC Intel(R) Communications Chipset 89xx Series - Datasheet 608 Bus:Device:Function: Bit Access October 2012 Order Number: 327879-001US 11.0 11.2.1.2 Offset 02h: Host Control Register (SMBus--B0:D31:F3) A read to this register will clear the byte pointer of the 32-byte buffer. Table 11-20. Offset 02h: Host Control Register (SMBus--B0:D31:F3) (Sheet 1 of 2) Description: View: PCI Size: 8 bit Bit Range 07 06 BAR: SMBASE (IO) Bus:Device:Function: B0:D3 1:F3 Default: 00h Offset Start: 02h Offset End: 02h Power Well: Bit Acronym Bit Description PEC_EN 0 = SMBus host controller does not perform the transaction with the PEC phase appended. 1 = Causes the host controller to perform the SMBus transaction with the Packet Error Checking phase appended. For writes, the value of the PEC byte is transferred from the PEC Register. For reads, the PEC byte is loaded in to the PEC Register. This bit must be written prior to the write in which the START bit is set. RW START 0 = This bit will always return 0 on reads. The HOST_BUSY bit in the Host Status register (offset 00h) can be used to identify when the PCH has finished the command. 1 = Writing a 1 to this bit initiates the command described in the SMB_CMD field. All registers should be setup prior to writing a 1 to this bit position. WO Sticky Bit Reset Value Bit Access This bit is used for Block Read commands. 05 LAST_BYTE 1 = Software sets this bit to indicate that the next byte will be the last byte to be received for the block. This causes the PCH to send a NACK (instead of an ACK) after receiving the last byte. WC Once the SECOND_TO_STS bit in TCO2_STS register (B0:D31:F0, TCOBASE+6h, bit 1) is set, the LAST_BYTE bit also gets set. While the SECOND_TO_STS bit is set, the LAST_BYTE bit cannot be cleared. This prevents the PCH from running some of the SMBus commands (Block Read/Write, I2C Read, Block I2C Write). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 609 Table 11-20. Offset 02h: Host Control Register (SMBus--B0:D31:F3) (Sheet 2 of 2) Description: View: PCI Size: 8 bit Bit Range 04 :02 BAR: SMBASE (IO) Bus:Device:Function: Default: 00h Bit Acronym Bit Description The bit encoding below indicates which command the PCH is to perform. If enabled, the PCH will generate an interrupt or SMI# when the command has completed If the value is for a nonsupported or reserved command, the PCH will set the device error (DEV_ERR) status bit (offset SMBASE + 00h, bit 2) and generate an interrupt when the START bit is set. The PCH will perform no command, and will not operate until DEV_ERR is cleared. 000 = Quick: The slave address and read/write value (bit 0) are stored in the transmit slave address register. 001 = Byte: This command uses the transmit slave address and command registers. Bit 0 of the slave address register determines if this is a read or write command. 010 = Byte Data: This command uses the transmit slave address, command, and DATA0 registers. Bit 0 of the slave address register determines if this is a read or write command. If it is a read, the DATA0 register will contain the read data. 011 = Word Data: This command uses the transmit slave address, command, DATA0 and DATA1 registers. Bit 0 of the slave address register determines if this is a read or write command. If it is a read, after the command completes, the DATA0 and DATA1 registers will contain the read data. 100 = Process Call: This command uses the transmit slave address, command, DATA0 and DATA1 registers. Bit 0 of the slave address register determines if this is a read or write command. After the SMB_CMD command completes, the DATA0 and DATA1 registers will contain the read data. 101 = Block: This command uses the transmit slave address, command, DATA0 registers, and the Block Data Byte register. For block write, the count is stored in the DATA0 register and indicates how many bytes of data will be transferred. For block reads, the count is received and stored in the DATA0 register. Bit 0 of the slave address register selects if this is a read or write command. For writes, data is retrieved from the first n (where n is equal to the specified count) addresses of the SRAM array. For reads, the data is stored in the Block Data Byte register. 110 = I2C Read: This command uses the transmit slave address, command, DATA0, DATA1 registers, and the Block Data Byte register. The read data is stored in the Block Data Byte register. The PCH continues reading data until the NAK is received. 111 = Block Process: This command uses the transmit slave address, command, DATA0 and the Block Data Byte register. For block write, the count is stored in the DATA0 register and indicates how many bytes of data will be transferred. For block read, the count is received and stored in the DATA0 register. Bit 0 of the slave address register always indicate a write command. For writes, data is retrieved from the first m (where m is equal to the specified count) addresses of the SRAM array. For reads, the data is stored in the Block Data Byte register. E32B bit in the Auxiliary Control register must be set for this command to work. 01 KILL 00 INTREN Offset Start: 02h Offset End: 02h Power Well: Sticky Bit Reset Value Bit Access RW 0 = Normal SMBus host controller functionality. 1 = Kills the current host transaction taking place, sets the FAILED status bit, and asserts the interrupt (or SMI#). This bit, once set, must be cleared by software to allow the SMBus host controller to function normally. RW 0 = Disable. 1 = Enable the generation of an interrupt or SMI# upon the completion of the command. RW Intel(R) Communications Chipset 89xx Series - Datasheet 610 B0:D3 1:F3 October 2012 Order Number: 327879-001US 11.0 11.2.1.3 Offset 03h: Host Command Register (SMBus--B0:D31:F3) Table 11-21. Offset 03h: Host Command Register (SMBus--B0:D31:F3) Description: View: PCI BAR: SMBASE (IO) Size: 8 bit B0:D3 1:F3 Default: 00h Bit Acronym Bit Range Offset Start: 03h Offset End: 03h Power Well: Bit Description Sticky Bit Reset Value This 8-bit field is transmitted by the host controller in the command field of the SMBus protocol during the execution of any command. 07 :00 11.2.1.4 Bus:Device:Function: Bit Access RW Offset 04h: Transmit Slave Address Register (SMBus--B0:D31:F3) This register is transmitted by the host controller in the slave address field of the SMBus protocol. Table 11-22. Offset 04h: Transmit Slave Address Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit Bit Range 07 :01 00 BAR: SMBASE (IO) B0:D3 Bus:Device:Function: 1:F3 Default: 00h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value This field provides a 7-bit address of the targeted ADDRESS slave. RW October 2012 Order Number: 327879-001US Offset Start: 04h Offset End: 04h Direction of the host transfer. 0 = Write 1 = Read Bit Access RW RW Intel(R) Communications Chipset 89xx Series - Datasheet 611 11.2.1.5 Offset 05h: Host Data 0 Register (SMBus--B0:D31:F3) Table 11-23. Offset 05h: Host Data 0 Register (SMBus--B0:D31:F3) Description: View: PCI BAR: SMBASE (IO) Size: 8 bit B0:D3 1:F3 Default: 00h Bit Range 07 :00 11.2.1.6 Bus:Device:Function: Offset Start: 05h Offset End: 05h Power Well: Bit Acronym Bit Description Data0/ Count This field contains the 8-bit data sent in the DATA0 field of the SMBus protocol. For block write commands, this register reflects the number of bytes to transfer. This register should be programmed to a value between 1 and 32 for block counts. A count of 0 or a count above 32 will result in unpredictable behavior. The host controller does not check or log invalid block counts. Sticky Bit Reset Value Bit Access RW Offset 06h: Host Data 1 Register (SMBus--B0:D31:F3) Table 11-24. Offset 06h: Host Data 1 Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit B0:D3 Bus:Device:Function: 1:F3 BAR: SMBASE (IO) Default: 00h Power Well: Bit Range Bit Acronym Bit Description 07 :00 Data1 This 8-bit register is transmitted in the DATA1 field of the SMBus protocol during the execution of any command. Intel(R) Communications Chipset 89xx Series - Datasheet 612 Offset Start: 06h Offset End: 06h Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 11.0 11.2.1.7 Offset 07h: Host Block Data Byte Register (SMBus--B0:D31:F3) Table 11-25. Offset 07h: Host Block Data Byte Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit Bit Range 07 :00 BAR: SMBASE (IO) Bus:Device:Function: B0:D3 1:F3 Default: 00h Power Well: Bit Acronym Bit Description BDTA This is either a register, or a pointer into a 32-byte block array, depending upon whether the E32B bit is set in the Auxiliary Control register. When the E32B bit (offset SMBASE + 0Dh, bit 1) is cleared, this is a register containing a byte of data to be sent on a block write or read from on a block read, just as it behaved on the ICH3. When the E32B bit is set, reads and writes to this register are used to access the 32-byte block data storage array. An internal index pointer is used to address the array, which is reset to 0 by reading the HCTL register (offset 02h). The index pointer then increments automatically upon each access to this register. The transfer of block data into (read) or out of (write) this storage array during an SMBus transaction always starts at index address 0. When the E2B bit is set, for writes, software will write up to 32bytes to this register as part of the setup for the command. After the Host controller has sent the Address, Command, and Byte Count fields, it will send the bytes in the SRAM pointed to by this register. When the E2B bit is cleared for writes, software will place a single byte in this register. After the host controller has sent the address, command, and byte count fields, it will send the byte in this register. If there is more data to send, software will write the next series of bytes to the SRAM pointed to by this register and clear the DONE_STS bit. The controller will then send the next byte. During the time between the last byte being transmitted to the next byte being transmitted, the controller will insert waitstates on the interface. When the E2B bit is set for reads, after receiving the byte count into the Data0 register, the first series of data bytes go into the SRAM pointed to by this register. If the byte count has been exhausted or the 32-byte SRAM has been filled, the controller will generate an SMI# or interrupt (depending on configuration) and set the DONE_STS bit. Software will then read the data. During the time between when the last byte is read from the SRAM to when the DONE_STS bit is cleared, the controller will insert wait-states on the interface. October 2012 Order Number: 327879-001US Offset Start: 07h Offset End: 07h Sticky Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 613 11.2.1.8 Offset 08h: Packet Error Check (PEC) Register (SMBus--B0:D31:F3) Table 11-26. Offset 08h: Packet Error Check (PEC) Register (SMBus--B0:D31:F3) Description: View: PCI BAR: SMBASE (IO) Size: 8 bit B0:D3 1:F3 Default: 00h Bit Acronym Bit Range Offset Start: 08h Offset End: 08h Power Well: Bit Description Sticky Bit Reset Value This 8-bit register is written with the 8-bit CRC value that is used as the SMBus PEC data prior to a write transaction. For read transactions, the PEC PEC_DATA data is loaded from the SMBus into this register and is then read by software. Software must ensure that the INUSE_STS bit is properly maintained to avoid having this field over-written by a write transaction following a read transaction. 07 :00 11.2.1.9 Bus:Device:Function: Bit Access RW Offset 09h: Receive Slave Address Register (SMBus--B0:D31:F3) Table 11-27. Offset 09h: Receive Slave Address Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit Bit Range 07 06 :00 B0:D3 Bus:Device:Function: 1:F3 BAR: SMBASE (IO) Default: 44h Bit Acronym Reserved Power Well: Resume Bit Description Reserved This field is the slave address that the PCH decodes for read and write cycles. the default is not 0, so SLAVE_ADDR the SMBus Slave Interface can respond even before the processor comes up (or if the processor is dead). This register is cleared by RSMRST#, but not by PLTRST#. Intel(R) Communications Chipset 89xx Series - Datasheet 614 Offset Start: 09h Offset End: 09h Sticky Bit Reset Value Bit Access RW RW October 2012 Order Number: 327879-001US 11.0 11.2.1.10 Offset 0Ah: Receive Slave Data Register (SMBus--B0:D31:F3) This register contains the 16-bit data value written by the external SMBus master. The processor can then read the value from this register. This register is reset by RSMRST#, but not PLTRST# . Table 11-28. Offset 0Ah: Receive Slave Data Register (SMBus--B0:D31:F3) Description: View: PCI BAR: SMBASE (IO) Size: 8 bit Bus:Device:Function: B0:D3 1:F3 Default: 0000h Bit Acronym Bit Range Offset Start: 0Ah Offset End: 0Bh Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access 15 :08 DATA_MSG1 Data Message Byte 1 RO 07 :00 DATA_MSG0 Data Message Byte 0 RO 11.2.1.11 Offset 0Ch: Auxiliary Status Register (SMBus--B0:D31:F3) Table 11-29. Offset 0Ch: Auxiliary Status Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit Bit Range 07 :02 01 00 BAR: SMBASE (IO) Bus:Device:Function: B0:D3 1:F3 Default: 00h Bit Acronym Offset Start: 0Ch Offset End: 0Ch Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved STCO SMBus TCO Mode -- This bit reflects the strap setting of TCO compatible mode vs. Advanced TCO mode. 0 = The PCH is in the compatible TCO mode. 1 = Reserved RO CRCE CRC Error: 0 = Software clears this bit by writing a 1 to it. 1 = This bit is set if a received message contained a CRC error. When this bit is set, the DERR bit of the host status register will also be set. This bit will be set by the controller if a software abort occurs in the middle of the CRC portion of the cycle or an abort happens after the PCH has received the final data bit transmitted by an external slave. RWC October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 615 11.2.1.12 Offset 0Dh: Auxiliary Control Register (SMBus--B0:D31:F3) Table 11-30. Offset 0Dh: Auxiliary Control Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit Bit Range 07 :02 01 00 BAR: SMBASE (IO) Bus:Device:Function: Default: 00h Bit Acronym Offset Start: 0Dh Offset End: 0Dh Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved E32B Enable 32-Byte Buffer: 0 = Disable. 1 = Enable. When set, the Host Block Data register is a pointer into a 32-byte buffer, as opposed to a single register. This enables the block commands to transfer or receive up to 32-bytes before the PCH generates an interrupt. RW AAC Automatically Append CRC: 0 = The PCH will Not automatically append the CRC. 1 = The PCH will automatically append the CRC. This bit must not be changed during SMBus transactions or undetermined behavior will result. It should be programmed only once during the lifetime of the function. RW Intel(R) Communications Chipset 89xx Series - Datasheet 616 B0:D3 1:F3 October 2012 Order Number: 327879-001US 11.0 11.2.1.13 Offset 0Eh: SMLink Pin Control Register (SMBus--B0:D31:F3) This register is in the resume well and is reset by RSMRST#. This register is only applicable in the TCO compatible mode. Table 11-31. Offset 0Eh: SMLink Pin Control Register (SMBus--B0:D31:F3) Description: View: PCI Bus:Device:Functio B0:D3 n: 1:F3 BAR: SMBASE (IO) Size: 8 bit Default: See register description Bit Range Bit Acronym 07 :03 Reserved Offset Start: 0Eh Offset End: 0Eh Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access Reserved 0 = The PCH will drive the SMLINK0 pin low, independent of what the other SMLink logic would otherwise indicate for the SMLINK0 pin. 1 = The SMLINK0 pin is not overdriven low. The other SMLink logic controls the state of the pin. (Default) RW 01 This read-only bit has a default value that is dependent on an external signal level. This pin returns the value on the SMLINK1 pin. This SMLINK1_CUR_STS allows software to read the current state of the pin. 0 = Low 1 = High RO 00 This read-only bit has a default value that is dependent on an external signal level. This pin returns the value on the SMLINK0 pin. This SMLINK0_CUR_STS allows software to read the current state of the pin. 0 = Low 1 = High RO 02 SMLINK_CLK_CTL October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 617 11.2.1.14 Offset 0Fh: SMBus Pin Control Register (SMBus--B0:D31:F3) This register is in the resume well and is reset by RSMRST#. Table 11-32. Offset 0Fh: SMBus Pin Control Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit BAR: SMBASE (IO) Bit Acronym 07 :03 Reserved 01 00 SMBCLK_CTL Bit Description This read-only bit has a default value that is dependent on an external signal level. This pin returns the value on the SMBDATA pin. This allows SMBDATA_CUR_STS software to read the current state of the pin. 0 = Low 1 = High SMBCLK_CUR_STS Offset 0Fh Start: 0Fh Offset End: Power Well: Resume Sticky Bit Reset Value Bit Access Reserved 1 = The SMBCLK pin is not overdriven low. The other SMBus logic controls the state of the pin. 0 = The PCH drives the SMBCLK pin low, independent of what the other SMB logic would otherwise indicate for the SMBCLK pin. (Default) This read-only bit has a default value that is dependent on an external signal level. This pin returns the value on the SMBCLK pin. This allows software to read the current state of the pin. 0 = Low 1 = High Intel(R) Communications Chipset 89xx Series - Datasheet 618 B0:D3 1:F3 Default: See register description Bit Range 02 Bus:Device:Function: RW RO RO October 2012 Order Number: 327879-001US 11.0 11.2.1.15 Offset 10h: Slave Status Register (SMBus--B0:D31:F3) This register is in the resume well and is reset by RSMRST#. All bits in this register are implemented in the 64 kHz clock domain. Therefore, software must poll this register until a write takes effect before assuming that a write has completed internally. Table 11-33. Offset 10h: Slave Status Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit BAR: SMBASE (IO) B0:D3 1:F3 Default: 00h Bit Range Bit Acronym 07 :01 Reserved 00 Bus:Device:Function: Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access Reserved The PCH sets this bit to a 1 when it has completely received a successful Host Notify Command on the SMBus pins. Software reads this bit to determine that the source of the interrupt or SMI# was the reception of the Host Notify Command. Software clears this bit after reading any information needed HOST_NOTIFY_STS from the Notify address and data registers by writing a 1 to this bit. The PCH will allow the Notify Address and Data registers to be over-written once this bit has been cleared. When this bit is 1, the PCH will NACK the first byte (host address) of any new "Host Notify" commands on the SMBus pins. Writing a 0 to this bit has no effect. October 2012 Order Number: 327879-001US Offset Start: 10h Offset End: 10h RWC Intel(R) Communications Chipset 89xx Series - Datasheet 619 11.2.1.16 Offset 11h: Slave Command Register (SMBus--B0:D31:F3) This register is in the resume well and is reset by RSMRST#. Table 11-34. Offset 11h: Slave Command Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit BAR: SMBASE (IO) Bit Acronym 07 :03 Reserved 01 00 Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access Reserved SMBALERT_DIS 0 = Allows the generation of the interrupt or SMI#. 1 = Software sets this bit to block the generation of the interrupt or SMI# due to the SMBALERT# source. This bit is logically inverted and ANDed with the SMBALERT_STS bit (offset SMBASE + 00h, bit 5). The resulting signal is distributed to the SMI# and/or interrupt generation logic. This bit does not effect the wake logic. RW HOST_NOTIFY_WKEN Software sets this bit to 1 to enable the reception of a Host Notify command as a wake event. When enabled this event is "OR'd" in with the other SMBus wake events and is reflected in the SMB_WAK_STS bit of the General Purpose Event 0 Status register. 0 = Disable 1 = Enable RW Software sets this bit to 1 to enable the generation of interrupt or SMI# when HOST_NOTIFY_STS (offset SMBASE + 10h, bit 0) is 1. This enable does not affect the setting of the HOST_NOTIFY_STS bit. When the interrupt is generated, either PIRQB# or SMI# is generated, depending on the value of the HOST_NOTIFY_INTREN SMB_SMI_EN bit (B0:D31:F3:40h, bit 1). If the HOST_NOTIFY_STS bit is set when this bit is written to a 1, then the interrupt (or SMI#) will be generated. The interrupt (or SMI#) is logically generated by AND'ing the STS and INTREN bits. 0 = Disable 1 = Enable RW Intel(R) Communications Chipset 89xx Series - Datasheet 620 B0:D3 Offset Start: 11h 1:F3 Offset End: 11h Default: 00h Bit Range 02 Bus:Device:Function: October 2012 Order Number: 327879-001US 11.0 11.2.1.17 Offset 14h: Notify Device Address Register (SMBus--B0:D31:F3) This register is in the resume well and is reset by RSMRST#. Table 11-35. Offset 14h: Notify Device Address Register (SMBus--B0:D31:F3) Description: View: PCI BAR: SMBASE (IO) Size: 8 bit B0:D3 Offset Start: 14h 1:F3 Offset End: 14h Default: 00h Bit Range Bit Acronym Power Well: Resume Bit Description Sticky Bit Reset Value Bit Access This field contains the 7-bit device address received during the Host Notify protocol of the SMBus 2.0 Specification. Software should only DEVICE_ADDRESS consider this field valid when the HOST_NOTIFY_STS bit (B0:D31:F3:SMBASE +10, bit 0) is set to 1. 07 :01 00 11.2.1.18 Bus:Device:Function: Reserved RO Reserved Offset 16h: Notify Data Low Byte Register (SMBus--B0:D31:F3) This register is in the resume well and is reset by RSMRST#. Table 11-36. Offset 16h: Notify Data Low Byte Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit Bit Range 07 :00 BAR: SMBASE (IO) Bus:Device:Function: B0:D3 1:F3 Default: 00h Bit Acronym Power Well: Resume Bit Description Sticky This field contains the first (low) byte of data received during the Host Notify protocol of the SMBus 2.0 specification. Software should only DATA_LOW_BYTE consider this field valid when the HOST_NOTIFY_STS bit (B0:D31:F3:SMBASE +10, bit 0) is set to 1. October 2012 Order Number: 327879-001US Offset Start: 16h Offset End: 16h Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 621 11.2.1.19 Offset 17h: Notify Data High Byte Register (SMBus--B0:D31:F3) This register is in the resume well and is reset by RSMRST#. Table 11-37. Offset 17h: Notify Data High Byte Register (SMBus--B0:D31:F3) Description: View: PCI Size: 8 bit Bit Range 07 :00 BAR: SMBASE (IO) Bus:Device:Function: B0:D3 Offset Start: 17h 1:F3 Offset End: 17h Default: 00h Bit Acronym DATA_HIGH_BYTE Power Well: Resume Bit Description This field contains the second (high) byte of data received during the Host Notify protocol of the SMBus 2.0 specification. Software should only consider this field valid when the HOST_NOTIFY_STS bit (B0:D31:F3:SMBASE +10, bit 0) is set to 1. Sticky Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 622 October 2012 Order Number: 327879-001US 12.0 12.0 PCI Express* Configuration Registers (B0:D28:F0/1/2/3) 12.1 PCI Express* Configuration Registers (PCI Express* -- B0:D28:F0/F1/F2/F3) Note: Register address locations that are not shown in the following table should be treated as Reserved. 12.1.1 PCI Express* Configuration Registers Address Map Table 12-1. PCI Express* Configuration Registers Address Map (Sheet 1 of 4) Offset Start Offset End Register ID - Description Default Value 00h 01h "Offset 00h: Vendor Identification Register (PCI Express*--B0:D28:F0/F1/F2/F3)" 8086h on page 626 02h 03h "Offset 02h: DID--Device Identification Register (PCI Express*--B0:D28:F0/F1/ F2/F3)" on page 626 See register description 04h 05h "Offset 04h: PCI COMMAND Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 627 0000h 06h 07h "Offset 06h: PCI Status Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 628 0010h 08h 08h "Offset 08h: RID--Revision Identification Register (PCI Express*--B0:D28:F0/F1/ F2/F3)" on page 629 See register description 09h 09h "Offset 09h: Programming Interface Register (PCI Express*--B0:D28:F0/F1/F2/ F3)" on page 630 00h 0Ah 0Ah "Offset 0Ah: Sub Class Code Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 630 04h 0Bh 0Bh "Offset 0Bh: Base Class Code Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 631 06h 0Ch 0Ch "Offset 0Ch: Cache Line Size Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 631 00h 0Dh 0Dh "Offset 0Dh: Primary Latency Timer Register (PCI Express*--B0:D28:F0/F1/F2/ F3)" on page 632 00h 0Eh 0Eh "Offset 0Eh: Header Type Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 632 81h 18h 1Ah "Offset 18h: Bus Number Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 633 000000h 1Bh 1Bh "Offset 1Bh: Secondary Latency Timer (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 633 00h 1Ch 1Dh "Offset 1Ch: IOBL--I/O Base and Limit Register (PCI Express*--B0:D28:F0/F1/F2/ 0000h F3)" on page 634 1Eh 1Fh "Offset 1Eh: Secondary Status Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on 0000h page 635 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 623 Table 12-1. PCI Express* Configuration Registers Address Map (Sheet 2 of 4) Offset Start Offset End Default Value Register ID - Description 20h 23h "Offset 20h: Memory Base and Limit Register (PCI Express*--B0:D28:F0/F1/F2/ F3)" on page 636 00000000h 24h 27h "Offset 24h: Prefetchable Memory Base and Limit Register (PCI Express*-- B0:D28:F0/F1/F2/F3)" on page 637 00010001h 28h 2Bh "Offset 28h: PMBU32--Prefetchable Memory Base Upper 32 Bits Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 637 00000000h 2Ch 2Fh "Offset 28h: PMBU32--Prefetchable Memory Base Upper 32 Bits Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 638 00000000h 34h 40h "Offset 34h: Capabilities List Pointer Register (PCI Express*--B0:D28:F0/F1/F2/ F3)" on page 638 8086h 3Ch 3Dh "Offset 3Ch: Interrupt Information Register (PCI Express*--B0:D28:F0/F1/F2/F3)" See register on page 639 description 3Eh 3fh "Offset 3Eh: Bridge Control Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 639 0000h 40h 41h "Offset 40h: Capabilities List Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 641 8010h 42h 43h "Offset 42h: PCI Express* Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/ 0042h F3)" on page 641 44h 47h "Offset 44h: Device Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on 00008000h page 642 48h 49h "Offset 48h: Device Control Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 643 0000h 4Ah 4Bh "Offset 4Ah: Device Status Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 644 0010h 4Ch 4Fh "Offset 4Ch: Link Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 645 See register description 50h 51h "Offset 50h: Link Control Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 647 0000h 52h 53h "Offset 52h: Link Status Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 648 See register description 54h 57h "Offset 54h: Slot Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 649 00040060h 58h 59h "Offset 58h: Slot Control Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 650 0000h 5Ah 5Bh "Offset 5Ah: Slot Status Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 651 0000h 5Ch 5Dh "Offset 5Ch: Root Control Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 652 0000h 60h 63h "Offset 60h: Root Status Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 653 00000000h 64h 67h "Offset 64h: Device Capabilities 2 Register (PCI Express*--B0:D28:F0/F1/F2/F3)" 00000016h on page 653 68h 69h "Offset 68h: Device Control 2 Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 654 0000h 70h 71h "Offset 70h: Link Control 2 Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 655 0001h 80h 81h "Offset 80h: Message Signaled Interrupt Identifiers Register (PCI Express*-- B0:D28:F0/F1/F2/F3)" on page 655 9005h 82h 83h "Offset 82h: Message Signaled Interrupt Message Control Register (PCI Express*-- 0000h B0:D28:F0/F1/F2/F3)" on page 656 Intel(R) Communications Chipset 89xx Series - Datasheet 624 October 2012 Order Number: 327879-001US 12.0 Table 12-1. PCI Express* Configuration Registers Address Map (Sheet 3 of 4) Offset Start Offset End Register ID - Description Default Value 84h 87h "Offset 84h: Message Signaled Interrupt Message Address Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 657 00000000h 88h 89h "Offset 88h: Message Signaled Interrupt Message Data Register (PCI Express*-- B0:D28:F0/F1/F2/F3)" on page 657 0000h 90h 91h "Offset 90h: Subsystem Vendor Capability Register (PCI Express*--B0:D28:F0/F1/ A00Dh F2/F3)" on page 658 94h 97h "Offset 94h: Subsystem Vendor Identification Register (PCI Express*--B0:D28:F0/ 00000000h F1/F2/F3)" on page 658 A0h A1h "Offset A0h: Power Management Capability Register (PCI Express*--B0:D28:F0/ F1/F2/F3)" on page 659 0001h A2h A3h "Offset A2h: PCI Power Management Capabilities Register (PCI Express*-- B0:D28:F0/F1/F2/F3)" on page 659 C802h A4h A7h "Offset A4h: PCI Power Management Control and Status Register (PCI Express*-- B0:D28:F0/F1/F2/F3)" on page 660 00000000h D4h D7h "Offset D4h: Miscellaneous Port Configuration Register 2 (PCI Express*-- B0:D28:F0/F1/F2/F3)" on page 661 00000000h D8h DBh "Offset D8h: Miscellaneous Port Configuration Register (PCI Express*--B0:D28:F0/ 08110000h F1/F2/F3)" on page 662 DCh DFh "Offset DCh: SMI/SCI Status Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on page 665 00000000h E1h E1h "Offset E1h: Root Port Dynamic Clock Gating Enable (PCI Express*--B0:D28:F0/ F1/F2/F3)" on page 666 00h E8h EBh "Offset E8h: PCI Express* Configuration Register 1 (PCI Express*--B0:D28:F0/F1/ 00000020h F2/F3)" on page 667 104h 107h "Offset 104h: Uncorrectable Error Status Register (PCI Express*--B0:D28:F0/F1/ F2/F3)" on page 667 108h 10Bh "Offset 108h: Uncorrectable Error Mask (PCI Express*--B0:D28:F0/F1/F2/F3)" on 00000000h page 669 10Ch 10Fh "Offset 10Ch: Uncorrectable Error Severity (PCI Express*--B0:D28:F0/F1/F2/F3)" 00060011h on page 671 110h 113h "Offset 110h: Correctable Error Status Register (PCI Express*--B0:D28:F0/F1/F2/ 00000000h F3)" on page 672 114h 117h "Offset 114h: Correctable Error Mask Register (PCI Express*--B0:D28:F0/F1/F2/ F3)" on page 673 00002000h 118h 11Bh "Offset 118h: Advanced Error Capabilities and Control Register (PCI Express*-- B0:D28:F0/F1/F2/F3)" on page 674 00000000h 130h 133h "Offset 130h: Root Error Status Register (PCI Express*--B0:D28:F0/F1/F2/F3)" on 00000000h page 675 180h 183h "Offset 180h: Root Complex Topology Capability List Register (PCI Express*-- B0:D28:F0/F1/F2/F3)" on page 676 184h 187h "Offset 184h: Element Self Description Register (PCI Express*--B0:D28:F0/F1/F2/ See register F3)" on page 676 description 190h 193h "Offset 190h: Upstream Link Description Register (PCI Express*--B0:D28:F0/F1/ F2/F3)" on page 677 8086h 198h 19Fh "Offset 198h: Upstream Link Base Address Register (PCI Express*--B0:D28:F0/ F1/F2/F3)" on page 678 See register description October 2012 Order Number: 327879-001US 00000000000 x0xxx0x0x00 00000x0000b 00010005h Intel(R) Communications Chipset 89xx Series - Datasheet 625 Table 12-1. PCI Express* Configuration Registers Address Map (Sheet 4 of 4) Offset Start Offset End Default Value Register ID - Description 300h 303h "Offset 300h: PCI Express* Configuration Register 2 (PCI Express*--B0:D28:F0/ F1/F2/F3)" on page 678 60005007h 318h 318h "Offset 318h: PCI Express* Extended Test Mode Register (PCI Express*-- B0:D28:F0/F1/F2/F3)" on page 679 See register description 324h 324h "OFfset 324h: PCI Express* Configuration Register 1 (PCI Express*--B0:D28:F0/ F1/F2/F3)" on page 679 14000016h 12.1.1.1 Offset 00h: Vendor Identification Register (PCI Express*--B0:D28:F0/ F1/F2/F3) Table 12-2. Offset 00h: Vendor Identification Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Description: View: PCI Size: 16 bit Default: 8086h Bit Range Bit Acronym 15 :00 VID 12.1.1.2 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Offset Start: 00h Offset End: 01h Power Well: Bit Description Sticky Vendor Identification: This 16-bit value is assigned to Intel. Intel VID = 8086h Bit Reset Value Bit Access 8086h RO Offset 02h: DID--Device Identification Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-3. Offset 02h: DID--Device Identification Register (PCI Express*--B0:D28:F0/ F1/F2/F3) Description: View: PCI Size: 16 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: See register description Bit Range Bit Acronym Bit Description 15 :00 DID Device Identification: This is a 16-bit value assigned to the PCH LPC bridge controller. Intel(R) Communications Chipset 89xx Series - Datasheet 626 Offset Start: 02h Offset End: 03h Power Well: Sticky Bit Reset Value Bit Access RO October 2012 Order Number: 327879-001US 12.0 12.1.1.3 Offset 04h: PCI COMMAND Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-4. Offset 04h: PCI COMMAND Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 0000h Bit Range Bit Acronym 15 :11 Reserved Offset Start: 04h Offset End: 05h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 10 FBE Fast Back to Back Enable -- Reserved per the PCI Express* Base Specification. 09 SEE SERR# Enable: 0 = Disable. 1 = Enables the root port to generate an SERR# message when PSTS.SSE is set. 08 WCC Wait Cycle Control -- Reserved per the PCI Express* Base Specification. 07 PER Parity Error Response: 0 = Disable. 1 = Indicates that the device is capable of reporting parity errors as a master on the backbone. 06 VPS VGA Palette Snoop -- Reserved per the PCI Express* Base Specification. 05 PMWE Postable Memory Write Enable -- Reserved per the PCI Express* Base Specification. 04 SCE Special Cycle Enable -- Reserved per the PCI Express* Base Specification. 03 BME Bus Master Enable: 0 = Disable. All cycles from the device are master aborted 1 = Enable. Allows the root port to forward cycles onto the backbone from a PCI Express* device. RW MSE Memory Space Enable: 0 = Disable. Memory cycles within the range specified by the memory base and limit registers are master aborted on the backbone. 1 = Enable. Allows memory cycles within the range specified by the memory base and limit registers can be forwarded to the PCI Express* device. RW 01 IOSE I/O Space Enable -- This bit controls access to the I/O space registers. 0 = Disable. I/O cycles within the range specified by the I/ O base and limit registers are master aborted on the backbone. 1 = Enable. Allows I/O cycles within the range specified by the I/O base and limit registers can be forwarded to the PCI Express* device. RW 00 Reserved 02 October 2012 Order Number: 327879-001US RW RW Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 627 12.1.1.4 Offset 06h: PCI Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-5. Offset 06h: PCI Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) (Sheet 1 of 2) Description: View: PCI Size: 16 bit Bit Range B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 0010h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access 15 DPE Detected Parity Error: 0 = No parity error detected. 1 = Set when the root port receives a command or data from the backbone with a parity error. This is set even if PCIMD.PER (B0:D28:F0/F1/F2/F3:04, bit 6) is not set. 14 SSE Signaled System Error: 0 = No system error signaled. 1 = Set when the root port signals a system error to the internal SERR# logic. RWC RMA Received Master Abort: 0 = Root port has not received a completion with unsupported request status from the backbone. 1 = Set when the root port receives a completion with unsupported request status from the backbone. RWC RTA Received Target Abort: 0 = Root port has not received a completion with completer abort from the backbone. 1 = Set when the root port receives a completion with completer abort from the backbone. RWC STA Signaled Target Abort: 0 = No target abort received. 1 = Set whenever the root port forwards a target abort received from the downstream device onto the backbone. RWC DPED Master Data Parity Error Detected: 0 = No data parity error received. 1 = Set when the root port receives a completion with a data parity error on the backbone and PCIMD.PER (B0:D28:F0/F1/F2/F3:04, bit 6) is set. RWC 08 FB2BC Fast Back to Back Capable -- Reserved per the PCI Express* Base Specification. 07 Reserved 13 12 11 10 :09 06 05 SC66 RWC Reserved 66 MHz Capable -- Reserved per the PCI Express* Base Specification. Capabilities List -- Hardwired to 1. Indicates the presence of a capabilities list. Intel(R) Communications Chipset 89xx Series - Datasheet 628 Offset Start: 06h Offset End: 07h RO October 2012 Order Number: 327879-001US 12.0 Table 12-5. Offset 06h: PCI Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range Default: 0010h Bit Acronym 02 :00 12.1.1.5 Offset Start: 06h Offset End: 07h Power Well: Bit Description Sticky Bit Reset Value Bit Access Interrupt Status -- Indicates status of Hot-Plug and power management interrupts on the root port that result in INTx# message generation. 0 = Interrupt is deasserted. 1 = Interrupt is asserted. This bit is not set if MSI is enabled. If MSI is not enabled, this bit is set regardless of the state of PCICMD.Interrupt Disable bit (B0:D28:F0/F1/F2/F3:04h:bit 10). 04 03 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Reserved RO Reserved Detected Parity Error:. 0 = No parity error detected. 1 = Set when the root port receives a command or data from the backbone with a parity error. This is set even if PCIMD.PER (B0:D28:F0/F1/F2/F3:04, bit 6) is not set. DPE RWC Offset 08h: RID--Revision Identification Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-6. Offset 08h: RID--Revision Identification Register (PCI Express*--B0:D28:F0/ F1/F2/F3) Description: View: PCI Size: 8 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: See register description Power Well: Bit Range Bit Acronym Bit Description 07 :00 RID Revision ID: This is an 8-bit value indicating the stepping of PCIe controller hardware. October 2012 Order Number: 327879-001US Offset Start: 08h Offset End: 08h Sticky Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 629 12.1.1.6 Offset 09h: Programming Interface Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-7. Offset 09h: Programming Interface Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Description: View: PCI Size: 8 bit Default: 00h Bit Range Bit Acronym 07 :00 PI 12.1.1.7 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Offset Start: 09h Offset End: 09h Power Well: Bit Description Sticky Bit Reset Value Bit Access Programming Interface: 00h = No specific register level programming interface defined. RO Offset 0Ah: Sub Class Code Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-8. Offset 0Ah: Sub Class Code Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 8 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 04h Bit Range Bit Acronym 07 :00 SCC Power Well: Bit Description Sub Class Code -- This field is determined by bit 2 of the MPC register (B0:D28:F0-5:Offset D8h, bit 2). 04h = PCI-to-PCI bridge. 00h = Host Bridge. Intel(R) Communications Chipset 89xx Series - Datasheet 630 Offset Start: 0Ah Offset End: 0Ah Sticky Bit Reset Value Bit Access RO October 2012 Order Number: 327879-001US 12.0 12.1.1.8 Offset 0Bh: Base Class Code Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-9. Offset 0Bh: Base Class Code Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 8 bit BAR: Configuration Default: 06h Bit Range Bit Acronym 07 :00 BCC 12.1.1.9 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Offset Start: 0Bh Offset End: 0Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Base Class Code -- 8-bit value that indicates the type of device for the LPC bridge. 06h = Bridge device. RO Offset 0Ch: Cache Line Size Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-10. Offset 0Ch: Cache Line Size Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 8 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 00h Bit Range Bit Acronym 07 :00 CLS October 2012 Order Number: 327879-001US Offset Start: 0Ch Offset End: 0Ch Power Well: Bit Description Sticky Cache Line Size -- This is read/write but contains no functionality, per the PCI Express* Base Specification. Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 631 12.1.1.10 Offset 0Dh: Primary Latency Timer Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-11. Offset 0Dh: Primary Latency Timer Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Description: View: PCI Size: 8 bit Bit Range Default: 00h Bit Acronym 12.1.1.11 Offset Start: 0Dh Offset End: 0Dh Power Well: Bit Description Sticky Bit Reset Value Bit Access Latency Count. Reserved per the PCI Express* Base Specification. 07 :03 02 :00 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Reserved RO Reserved Offset 0Eh: Header Type Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Table 12-12. Offset 0Eh: Header Type Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 8 bit Bit Range 07 06 :00 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 81h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access Multi-Function Device: 0 = Single-function device. 1 = Multi-function device. RO Configuration Layout -- This field is determined by bit 2 of the MPC register (B0:D28:F0-5:Offset D8h, bit 2). 00h = Indicates a Host Bridge. 01h = Indicates a PCI-to-PCI bridge. RO Intel(R) Communications Chipset 89xx Series - Datasheet 632 Offset Start: 0Eh Offset End: 0Eh October 2012 Order Number: 327879-001US 12.0 12.1.1.12 Offset 18h: Bus Number Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Table 12-13. Offset 18h: Bus Number Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 24 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 000000h Offset Start: 18h Offset End: 1Ah Power Well: Bit Access Bit Acronym Bit Description 23 :16 SBBN Subordinate Bus Number -- Indicates the highest PCI bus number below the bridge. RW 15 :08 SCBN Secondary Bus Number -- Indicates the bus number the port. RW 07 :00 PBN Primary Bus Number -- Indicates the bus number of the backbone. RW 12.1.1.13 Sticky Bit Reset Value Bit Range Offset 1Bh: Secondary Latency Timer (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-14. Offset 1Bh: Secondary Latency Timer (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 8 bit Bit Range BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 00h Bit Acronym 07 :00 October 2012 Order Number: 327879-001US Offset Start: 1Bh Offset End: 1Bh Power Well: Bit Description Sticky Secondary Latency Timer -- Reserved for a Root Port per the PCI Express* Base Specification Bit Reset Value Bit Access RO Intel(R) Communications Chipset 89xx Series - Datasheet 633 12.1.1.14 Offset 1Ch: IOBL--I/O Base and Limit Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-15. Offset 1Ch: IOBL--I/O Base and Limit Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Description: View: PCI Size: 16 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 0000h Offset Start: 1Ch Offset End: 1Dh Power Well: Bit Description Bit Acronym 15 :12 IOLA I/O Limit Address -- I/O Base bits corresponding to address lines 15:12 for 4-KB alignment. Bits 11:0 are assumed to be padded to FFFh. RW 11 :08 IOLC I/O Limit Address Capability -- Indicates that the bridge does not support 32-bit I/O addressing. RO 07 :04 IOBA I/O Base Address -- I/O Base bits corresponding to address lines 15:12 for 4-KB alignment. Bits 11:0 are assumed to be padded to 000h. RW 03 :00 IOBC I/O Base Address Capability -- Indicates that the bridge does not support 32-bit I/O addressing. RO Intel(R) Communications Chipset 89xx Series - Datasheet 634 Sticky Bit Reset Value Bit Range Bit Access October 2012 Order Number: 327879-001US 12.0 12.1.1.15 Offset 1Eh: Secondary Status Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-16. Offset 1Eh: Secondary Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 0000h Bit Acronym Offset Start: 1Eh Offset End: 1Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access DPE Detected Parity Error: 0 = No error. 1 = The port received a poisoned TLP. RWC 14 RSE Received System Error: 0 = No error. 1 = The port received an ERR_FATAL or ERR_NONFATAL message from the device. RWC 13 RMA Received Master Abort: 0 = Unsupported Request not received. 1 = The port received a completion with "Unsupported Request" status from the device. RWC 12 RTA Received Target Abort: 0 = Completion Abort not received. 1 = The port received a completion with "Completion Abort" status from the device. RWC 11 STA Signaled Target Abort: 0 = Completion Abort not sent. 1 = The port generated a completion with "Completion Abort" status to the device. RWC SDTS Secondary DEVSEL# Timing Status: Reserved per PCI Express* Base Specification. 08 DPD Data Parity Error Detected: 0 = Conditions below did not occur. 1 = Set when the BCTRL.PERE (B0:D28:FO/F1/F2/F3:3E: bit 0) is set, and either of the following two conditions occurs: -Port receives completion marked poisoned. -Port poisons a write request to the secondary side. 07 SFBC Secondary Fast Back to Back Capable: Reserved per PCI Express* Base Specification. 16 Reserved 15 10 :09 05 04 :00 SC66 Reserved October 2012 Order Number: 327879-001US RWC Reserved Secondary 66 MHz Capable: Reserved per PCI Express* Base Specification. Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 635 12.1.1.16 Offset 20h: Memory Base and Limit Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Accesses that are within the ranges specified in this register will be sent to the attached device if CMD.MSE (B0:D28:F0/F1/F2/F3:04:bit 1) is set. Accesses from the attached device that are outside the ranges specified will be forwarded to the backbone if CMD.BME (B0:D28:F0/F1/F2/F3:04:bit 2) is set. The comparison performed is MB AD[31:20] ML. Table 12-17. Offset 20h: Memory Base and Limit Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Description: View: PCI Size: 32 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 00000000h Power Well: Bit Range Bit Acronym Bit Description 31 :20 ML Memory Limit -- These bits are compared with bits 31:20 of the incoming address to determine the upper 1-MB aligned value of the range. 19 16 Reserved 15 :04 MB 03 :00 Reserved Sticky Bit Reset Value Bit Access RW Reserved Memory Base -- These bits are compared with bits 31:20 of the incoming address to determine the lower 1-MB aligned value of the range. RW Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 636 Offset Start: 20h Offset End: 23h October 2012 Order Number: 327879-001US 12.0 12.1.1.17 Offset 24h: Prefetchable Memory Base and Limit Register (PCI Express*--B0:D28:F0/F1/F2/F3) Accesses that are within the ranges specified in this register will be sent to the device if CMD.MSE (B0:D28:F0/F1/F2/F3;04, bit 1) is set. Accesses from the device that are outside the ranges specified will be forwarded to the backbone if CMD.BME (B0:D28:F0/F1/F2/F3;04, bit 2) is set. The comparison performed is PMBU32:PMB AD[63:32]:AD[31:20] PMLU32:PML. Table 12-18. Offset 24h: Prefetchable Memory Base and Limit Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 00010001h Offset Start: 24h Offset End: 27h Power Well: Bit Access Bit Acronym Bit Description 31 :20 PML Prefetchable Memory Limit -- These bits are compared with bits 31:20 of the incoming address to determine the upper 1-MB aligned value of the range. RW 19 16 I64L 64-bit Indicator -- Indicates support for 64-bit addressing RO 15 :04 PMB Prefetchable Memory Base -- These bits are compared with bits 31:20 of the incoming address to determine the lower 1-MB aligned value of the range. RW 03 :00 I64B 64-bit Indicator -- Indicates support for 64-bit addressing RO 12.1.1.18 Sticky Bit Reset Value Bit Range Offset 28h: PMBU32--Prefetchable Memory Base Upper 32 Bits Register (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-19. Offset 28h: PMBU32--Prefetchable Memory Base Upper 32 Bits Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 00000000h Power Well: Bit Range Bit Acronym Bit Description 31 :00 PMBU Prefetchable Memory Base Upper Portion -- Upper 32-bits of the prefetchable address base. October 2012 Order Number: 327879-001US Offset Start: 28h Offset End: 2Bh Sticky Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 637 12.1.1.19 Offset 2Ch: Prefetchable Memory Limit Upper 32 Bits Register (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-20. Offset 28h: PMBU32--Prefetchable Memory Base Upper 32 Bits Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 00000000h Power Well: Bit Range Bit Acronym Bit Description 31 :00 PMBU Prefetchable Memory Base Upper Portion -- Upper 32-bits of the prefetchable address base. 12.1.1.20 Offset Start: 2Ch Offset End: 2Fh Sticky Bit Reset Value Bit Access RW Offset 34h: Capabilities List Pointer Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-21. Offset 34h: Capabilities List Pointer Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Description: View: PCI Size: 8 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 8086h Power Well: Bit Range Bit Acronym Bit Description 07 :00 PTR Capabilities Pointer -- Indicates that the pointer for the first entry in the capabilities list is at 40h in configuration space. Intel(R) Communications Chipset 89xx Series - Datasheet 638 Offset Start: 34h Offset End: 40h Sticky Bit Reset Value Bit Access RO October 2012 Order Number: 327879-001US 12.0 12.1.1.21 Offset 3Ch: Interrupt Information Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-22. Offset 3Ch: Interrupt Information Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Description: View: PCI Size: 16 bit Bit Range B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: See register description Bit Acronym Offset Start: 3Ch Offset End: 3Dh Power Well: Bit Description Sticky Bit Reset Value Bit Access Interrupt Pin -- Indicates the interrupt pin driven by the root port. At reset, this register takes on the following values, which reflect the reset state of the B0:D28IP register in config space: 15 :08 IPIN Port Reset Value 1 D28IP.P1IP 2 D28IP.P2IP 3 D28IP.P3IP 4 D28IP.P4IP RO The value that is programmed into B0:D28IP is always reflected in this register. 07 :00 12.1.1.22 Interrupt Line -- Default = 00h. Software written value to indicate which interrupt line (vector) the interrupt is connected to. No hardware action is taken on this register. These bits are not reset by FLR. ILINE RW Offset 3Eh: Bridge Control Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-23. Offset 3Eh: Bridge Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 0000h Bit Range Bit Acronym 15 :12 Reserved 11 DTSE 10 DTS October 2012 Order Number: 327879-001US Offset Start: 3Eh Offset End: 3fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Discard Timer SERR# Enable: Reserved per PCI Express* Base Specification, Revision 1.0a Discard Timer Status: Reserved per PCI Express* Base Specification, Revision 1.0a. Intel(R) Communications Chipset 89xx Series - Datasheet 639 Table 12-23. Offset 3Eh: Bridge Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit Bit Range B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 0000h Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access 09 SDT Secondary Discard Timer: Reserved per PCI Express* Base Specification, Revision 1.0a. 08 PDT Primary Discard Timer -- Reserved per PCI Express* Base Specification, Revision 1.0a. 07 FBE Fast Back to Back Enable -- Reserved per PCI Express* Base Specification, Revision 1.0a. 06 SBR Secondary Bus Reset -- Triggers a hot reset on the PCI Express* port. 05 MAM Master Abort Mode -- Reserved per Express specification. V16 VGA 16-Bit Decode: 0 = VGA range is enabled. 1 = The I/O aliases of the VGA range (See BCTRL:VE definition below), are not enabled, and only the base I/O ranges can be decoded RW VE VGA Enable: 0 = The ranges below will not be claimed off the backbone by the root port. 1 = The following ranges will be claimed off the backbone by the root port: -Memory ranges A0000h-BFFFFh -I/O ranges 3B0h - 3BBh and 3C0h - 3DFh, and all aliases of bits 15:10 in any combination of 1s RW IE ISA Enable -- This bit only applies to I/O addresses that are enabled by the I/O Base and I/O Limit registers and are in the first 64 KB of PCI I/O space. 0 = The root port will not block any forwarding from the backbone as described below. 1 = The root port will block any forwarding from the backbone to the device of I/O transactions addressing the last 768 bytes in each 1-KB block (offsets 100h to 3FFh). RW SE SERR# Enable: 0 = The messages described below are not forwarded to the backbone. 1 = ERR_COR, ERR_NONFATAL, and ERR_FATAL messages received are forwarded to the backbone. RW PERE Parity Error Response Enable -- When set: 0 = Poisoned write TLPs and completions indicating poisoned TLPs will not set the SSTS.DPD (B0:D28:F0/F1/ F2/F3:1E, bit 8). 1 = Poisoned write TLPs and completions indicating poisoned TLPs will set the SSTS.DPD (B0:D28:F0/F1/F2/ F3:1E, bit 8). RW 04 03 02 01 00 Intel(R) Communications Chipset 89xx Series - Datasheet 640 Offset Start: 3Eh Offset End: 3fh RW October 2012 Order Number: 327879-001US 12.0 12.1.1.23 Offset 40h: Capabilities List Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-24. Offset 40h: Capabilities List Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit BAR: Configuration Default: 8010h Bit Range Bit Acronym 15 :08 NEXT 07 :00 CID 12.1.1.24 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Offset Start: 40h Offset End: 41h Power Well: Bit Description Sticky Bit Reset Value Bit Access Next Capability -- Value of 80h indicates the location of the next pointer. RO Capability ID -- Indicates this is a PCI Express* capability. RO Offset 42h: PCI Express* Capabilities Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-25. Offset 42h: PCI Express* Capabilities Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Description: View: PCI Size: 16 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 0042h Bit Range Bit Acronym 15 :14 Reserved 13 :09 IMN Offset Start: 42h Offset End: 43h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Interrupt Message Number -- The PCH does not have multiple MSI interrupt numbers. RO SI Slot Implemented -- Indicates whether the root port is connected to a slot. Slot support is platform specific. BIOS programs this field, and it is maintained until a platform reset. RWO 07 :04 DT Device / Port Type -- Indicates this is a PCI Express* root port. RO 03 :00 CV Capability Version -- Indicates PCI Express* 2.0. RO 08 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 641 12.1.1.25 Offset 44h: Device Capabilities Register (PCI Express*--B0:D28:F0/ F1/F2/F3) Table 12-26. Offset 44h: Device Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Description: View: PCI Size: 32 bit Default: 00008000h Bit Range Bit Acronym 31 :28 Reserved 27 :26 CSPS 25 :18 CSPV 17 :16 Reserved 15 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration RBER Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved RO Captured Slot Power Limit Scale -- Not supported. RO Captured Slot Power Limit Value -- Not supported. RO Reserved Role Based Error Reporting -- Indicates that this device implements the functionality defined in the Error Reporting ECN as required by the PCI Express* 1.1 spec. RO 14 :12 Reserved 11 :09 E1AL EndPoint L1 Acceptable Latency -- This field is reserved with a setting of 000b for devices other than EndPoints, per the PCI Express* 1.1 Spec. RO 08 :06 E0AL EndPoint L0s Acceptable Latency -- This field is reserved with a setting of 000b for devices other than EndPoints, per the PCI Express* 1.1 Spec. RO ETFS Extended Tag Field Supported -- Indicates that 8-bit tag fields are supported. RO 04 :03 PFS Phantom Functions Supported -- No phantom functions supported. RO 02 :00 MPS Max Payload Size Supported -- Indicates the maximum payload size supported is 128B. RO 05 Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 642 Offset Start: 44h Offset End: 47h October 2012 Order Number: 327879-001US 12.0 12.1.1.26 Offset 48h: Device Control Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-27. Offset 48h: Device Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit Bit Range 15 14 :12 BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 0000h Bit Acronym Reserved MRRS 11 ENS 10 APME Offset Start: 48h Offset End: 49h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Max Read Request Size -- Hardwired to 0. RO Enable No Snoop -- Not supported. The root port will never issue non-snoop requests. RO Aux Power PM Enable -- The OS will set this bit to 1 if the device connected has detected aux power. It has no effect on the root port otherwise. RW 09 PFE Phantom Functions Enable -- Not supported. RO 08 ETFE Extended Tag Field Enable -- Not supported. RO MPS Max Payload Size -- The root port only supports 128-B payloads, regardless of the programming of this field. RW ERO Enable Relaxed Ordering -- Not supported. RO URE Unsupported Request Reporting Enable: 0 = The root port will ignore unsupported request errors. 1 = Allows signaling ERR_NONFATAL, ERR_FATAL, or ERR_COR to the Root Control register when detecting an unmasked Unsupported Request (UR). An ERR_COR is signaled when a unmasked Advisory Non-Fatal UR is received. An ERR_FATAL, ERR_or NONFATAL, is sent to the Root Control Register when an uncorrectable non-Advisory UR is received with the severity set by the Uncorrectable Error Severity register. RO FEE Fatal Error Reporting Enable: 0 = The root port will ignore fatal errors. 1 = Enables signaling of ERR_FATAL to the Root Control register due to internally detected errors or error messages received across the link. Other bits also control the full scope of related error reporting. RW Non-Fatal Error Reporting Enable: 0 = The root port will ignore non-fatal errors. 1 = Enables signaling of ERR_NONFATAL to the Root Control register due to internally detected errors or error messages received across the link. Other bits also control the full scope of related error reporting. RW Correctable Error Reporting Enable: 0 =The root port will ignore correctable errors. 1 = Enables signaling of ERR_CORR to the Root Control register due to internally detected errors or error messages received across the link. Other bits also control the full scope of related error reporting. RW 07 :05 04 03 02 01 00 NFE CEE October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 643 12.1.1.27 Offset 4Ah: Device Status Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Table 12-28. Offset 4Ah: Device Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 0010h Bit Range Bit Acronym 15 :06 Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved 05 TDP Transactions Pending -- This bit has no meaning for the root port since only one transaction may be pending to the PCH, so a read of this bit cannot occur until it has already returned to 0. RO 04 APD AUX Power Detected -- The root port contains AUX power for wakeup. RO 03 URD Unsupported Request Detected -- Indicates an unsupported request was detected. FED Fatal Error Detected -- Indicates a fatal error was detected. 0 = Fatal has not occurred. 1 = A fatal error occurred from a data link protocol error, link training error, buffer overflow, or malformed TLP. 02 01 00 NFED CED Non-Fatal Error Detected -- Indicates a non-fatal error was detected. 0 = Non-fatal has not occurred. 1 = A non-fatal error occurred from a poisoned TLP, unexpected completions, unsupported requests, completer abort, or completer timeout. Correctable Error Detected -- Indicates a correctable error was detected. 0 = Correctable has not occurred. 1 = The port received an internal correctable error from receiver errors / framing errors, TLP CRC error, DLLP CRC error, replay num rollover, replay timeout. Intel(R) Communications Chipset 89xx Series - Datasheet 644 Offset Start: 4Ah Offset End: 4Bh RWC RWC RWC RWC October 2012 Order Number: 327879-001US 12.0 12.1.1.28 Offset 4Ch: Link Capabilities Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-29. Offset 4Ch: Link Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/F3) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: See register description Bit Acronym Offset Start: 4Ch Offset End: 4Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Port Number -- Indicates the port number for the root port. This value is different for each implemented port: 31 :24 23 :21 20 PN Reserved LARC 19 :18 Reserved 17 :15 EL1 Function Port # Value of PN Field D28:F0 1 01h D28:F1 2 02h D28:F2 3 03h D28:F3 4 04h RO Reserved Link Active Reporting Capable -- Hardwired to 1 to indicate that this port supports the optional capability of reporting the DL_Active state of the Data Link Control and Management State Machine. RO Reserved L1 Exit Latency -- Set to 010b to indicate an exit latency of 2 s to 4 s. RO L0s Exit Latency -- Indicates as exit latency based upon common-clock configuration. 14 :12 EL0 LCLT.CCC Value of EL0 (these bits) 0 MPC.UCEL (B0:D28:F0/F1/F2/ F3:D8h:bits20:18) 1 MPC.CCEL (B0:D28:F0/F1/F2/ F3:D8h:bits17:15) RO Active State Link PM Support -- Indicates what level of active state link power management is supported on the root port. 11 :10 APMS October 2012 Order Number: 327879-001US Bits Definition 00b Neither L0s nor L1 are supported 01b L0s Entry Supported 10b L1 Entry Supported 11b Both L0s and L1 Entry supported RWO Intel(R) Communications Chipset 89xx Series - Datasheet 645 Table 12-29. Offset 4Ch: Link Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/F3) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: See register description Bit Acronym Offset Start: 4Ch Offset End: 4Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Maximum Link Width -- For the root ports, several values can be taken, based upon the value of the PCH config register field RPC.PC1 (PCH Config Registers:Offset 0224h:bits1:0) for Ports 1-4 and RPC.PC2 (PCH Config Registers:Offset 0224h:bits1:0) for Ports 5 and 6 Value of MLW Field 09 :04 03 :00 MLKW MLS Port # RPC.PC1=00b RPC.PC1=11b 1 01h 04h 2 01h 01h 3 01h 01h 4 01h 01h Maximum Link Speed -- Set to 1h to indicate the link speed is 2.5 Gb/s. Intel(R) Communications Chipset 89xx Series - Datasheet 646 RO RO October 2012 Order Number: 327879-001US 12.0 12.1.1.29 Offset 50h: Link Control Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Table 12-30. Offset 50h: Link Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit BAR: Configuration Default: 0000h Bit Range Bit Acronym 15 :10 Reserved 07 06 Reserved Offset Start: 50h Offset End: 51h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Hardware Autonomous Width Disable - Hardware never attempts to change the link width except when attempting to correct unreliable Link operation. 09 08 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 RO Reserved SE Extended Synch: 0 = Extended synch disabled. 1 = Forces extended transmission of FTS ordered sets in FTS and extra TS2 at exit from L1 prior to entering L0. RW CCC Common Clock Configuration: 0 = The PCH and device are not using a common reference clock. 1 = The PCH and device are operating with a distributed common reference clock. RW RW 05 RL Retrain Link: 0 = This bit always returns 0 when read. 1 = The root port will train its downstream link. Software uses LSTS.LT (B0:D28:F0/F1/F2/F3:52, bit 11) to check the status of training. It is permitted to write 1b to this bit while simultaneously writing modified values to other fields in this register. If the LTSSM is not already in Recovery or Configuration, the resulting Link training must use the modified values. If the LTSSM is already in Recovery or Configuration, the modified values are not required to affect the Link training that is already in progress. 04 LD Link Disable: 0 = Link enabled. 1 = The root port will disable the link. RW 03 RCBC Read Completion Boundary Control) -- Indicates the read completion boundary is 64 bytes. RO 02 Reserved 01 :00 APMC October 2012 Order Number: 327879-001US Reserved Active State Link PM Control -- Indicates whether the root port should enter L0s or L1 or both. Bits Definition 00 Disabled 01 L0s Entry Enabled 10 L1 Entry Enabled 11 L0s and L1 Entry Enabled RW Intel(R) Communications Chipset 89xx Series - Datasheet 647 12.1.1.30 Offset 52h: Link Status Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Table 12-31. Offset 52h: Link Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: See register description Bit Range Bit Acronym 15 :14 Reserved Bit Description Offset Start: 52h Offset End: 53h Power Well: Sticky Bit Reset Value Bit Access Reserved 13 DLLA Data Link Layer Active -- Default value is 0b. 0 = Data Link Control and Management State Machine is not in the DL_Active state 1 = Data Link Control and Management State Machine is in the DL_Active state 12 SCC Slot Clock Configuration -- Set to 1b to indicate that the PCH uses the same reference clock as on the platform and does not generate its own clock. RO 11 LT Link Training -- Default value is 0b. 0 = Link training completed. 1 = Link training is occurring. RO 10 LTE Link Training Error -- Not supported. Set value is 0b. RO RO Negotiated Link Width -- This field indicates the negotiated width of the given PCI Express* link. The contents of this NLW field is undefined if the link has not successfully trained. Port # 09 :04 NLW Possible Values 1 000001b, 000010b, 000100b 2 000001b 3 000001b 4 000001b RO 000001b = x1 link width, 000010b =x2 linkwidth, 000100b = x4 linkwidth 03 :00 LS Link Speed -- This field indicates the negotiated Link speed of the given PCI Express* link. 01h = Link is 2.5 Gb/s. Intel(R) Communications Chipset 89xx Series - Datasheet 648 RO October 2012 Order Number: 327879-001US 12.0 12.1.1.31 Offset 54h: Slot Capabilities Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-32. Offset 54h: Slot Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 00040060h Offset Start: 54h Offset End: 57h Power Well: Sticky Bit Access Bit Acronym 31 :19 PSN 18 :17 Reserved 16 :15 SLS Slot Power Limit Scale -- Specifies the scale used for the slot power limit value. BIOS sets this field and it remains set until a platform reset. RWO 14 :07 SLV Slot Power Limit Value -- Specifies the upper limit (in conjunction with SLS value), on the upper limit on power supplied by the slot. The two values together indicate the amount of power in watts allowed for the slot. BIOS sets this field and it remains set until a platform reset. RWO 06 HPC Hot Plug Capable: 1b = Indicates that Hot-Plug is supported. RWO 05 HPS Hot Plug Surprise: 1b = Indicates the device may be removed from the slot without prior notification. RWO 04 PIP Power Indicator Present: 0b = Indicates that a power indicator LED is not present for this slot. RO 03 AIP Attention Indicator Present: 0b = Indicates that an attention indicator LED is not present for this slot. RO 02 MSP MRL Sensor Present: 0b = Indicates that an MRL sensor is not present. RO 01 PCP Power Controller Present: 0b = Indicates that a power controller is not implemented for this slot. RO 00 ABP Attention Button Present: 0b = Indicates that an attention button is not implemented for this slot. RO October 2012 Order Number: 327879-001US Bit Description Bit Reset Value Bit Range Physical Slot Number -- This is a value that is unique to the slot number. BIOS sets this field and it remains set until a platform reset. RWO Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 649 12.1.1.32 Offset 58h: Slot Control Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Table 12-33. Offset 58h: Slot Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit Default: 0000h Bit Range Bit Acronym 15 :13 Reserved 12 LACE 11 Reserved 10 09 :06 PCC Reserved 05 HPE 04 Reserved 03 02 :00 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration PDE Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Link Active Changed Enable -- When set, this field enables generation of a hot plug interrupt when the Data Link Layer Link Active field (B0:D28:F0/F1/F2/F3:52h:bit 13) is changed. RW Reserved Power Controller Control -- This bit has no meaning for module based Hot-Plug. RO Reserved Hot Plug Interrupt Enable: 0 = Hot plug interrupts based on Hot-Plug events is disabled. 1 = Enables generation of a Hot-Plug interrupt on enabled Hot-Plug events. RW Reserved Presence Detect Changed Enable: 0 = Hot plug interrupts based on presence detect logic changes is disabled. 1 = Enables the generation of a Hot-Plug interrupt or wake message when the presence detect logic changes state. RW Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 650 Offset Start: 58h Offset End: 59h October 2012 Order Number: 327879-001US 12.0 12.1.1.33 Offset 5Ah: Slot Status Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Table 12-34. Offset 5Ah: Slot Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 0000h Bit Range Bit Acronym 15 :09 Reserved 08 LASC 07 Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Link Active State Changed: 1 = This bit is set when the value reported in Data Link Layer Link Active field of the Link Status register (B0:D28:F0/F1/F2/F3:52h:bit 13) is changed. In response to a Data Link Layer State Changed event, software must read Data Link Layer Link Active field of the Link Status register to determine if the link is active before initiating configuration cycles to the hot plugged device. RWC Reserved 06 PDS Presence Detect State -- If XCAP.SI (B0:D28:F0/F1/F2/ F3:42h:bit 8) is set (indicating that this root port spawns a slot), then this bit: 0 = Indicates the slot is empty. 1 = Indicates the slot has a device connected. Otherwise, if XCAP.SI is cleared, this bit is always set (1). 05 MS MRL Sensor State -- Reserved as the MRL sensor is not implemented. 04 Reserved RO Reserved 03 PDC Presence Detect Changed: 0 = No change in the PDS bit. 1 = The PDS bit changed states. 02 MSC MRL Sensor Changed -- Reserved as the MRL sensor is not implemented. 01 PFD Power Fault Detected -- Reserved as a power controller is not implemented. 00 Reserved October 2012 Order Number: 327879-001US Offset Start: 5Ah Offset End: 5Bh RWC Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 651 12.1.1.34 Offset 5Ch: Root Control Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Table 12-35. Offset 5Ch: Root Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit Default: 0000h Bit Range Bit Acronym 15 :04 Reserved 03 02 01 00 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved PIE PME Interrupt Enable: 0 = Interrupt generation disabled. 1 = Interrupt generation enabled when PCISTS.Inerrupt Status (B0:D28:F0/F1/F2/F3:60h, bit 16) is in a set state (either due to a 0 to 1 transition, or due to this bit being set with RSTS.IS already set). RW SFE System Error on Fatal Error Enable: 0 = An SERR# will not be generated. 1 = An SERR# will be generated, assuming CMD. See (B0:D28:F0/F1/F2/F3:04, bit 8) to see if it is set and if a fatal error is reported by any of the devices in the hierarchy of this root port, including fatal errors in this root port. RW SNE System Error on Non-Fatal Error Enable: 0 = An SERR# will not be generated. 1 = An SERR# will be generated, assuming CMD. See (B0:D28:F0/F1/F2/F3:04, bit 8) to see if it is set and if a non-fatal error is reported by any of the devices in the hierarchy of this root port, including non-fatal errors in this root port. RW SCE System Error on Correctable Error Enable: 0 = An SERR# will not be generated. 1 = An SERR# will be generated, assuming CMD.See (B0:D28:F0/F1/F2/F3:04, bit 8) to see if a correctable error is reported by any of the devices in the hierarchy of this root port, including correctable errors in this root port. RW Intel(R) Communications Chipset 89xx Series - Datasheet 652 Offset Start: 5Ch Offset End: 5Dh October 2012 Order Number: 327879-001US 12.0 12.1.1.35 Offset 60h: Root Status Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Table 12-36. Offset 60h: Root Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000000h Bit Range Bit Acronym 31 :18 Reserved 17 16 15 :00 12.1.1.36 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Offset Start: 60h Offset End: 63h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved PP PME Pending: 0 = When the original PME is cleared by software, it will be set again, the requestor ID will be updated, and this bit will be cleared. 1 = Indicates another PME is pending when the PME status bit is set. RO PS PME Status: 0 = PME was not asserted. 1 = Indicates that PME was asserted by the requestor ID in RID. Subsequent PMEs are kept pending until this bit is cleared. RWC RID PME Requestor ID -- Indicates the PCI requestor ID of the last PME requestor. Valid only when PS is set. RO Offset 64h: Device Capabilities 2 Register (PCI Express*--B0:D28:F0/ F1/F2/F3) Table 12-37. Offset 64h: Device Capabilities 2 Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000016h Bit Range Bit Acronym 15 :05 Reserved 04 03 :00 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Offset Start: 64h Offset End: 67h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved CTDS Completion Timeout Disable Supported -- A value of 1b indicates support for the Completion Timeout Disable mechanism. RO CTRS Completion Timeout Ranges Supported - This field indicates device support for the optional Completion Timeout programmability mechanism. This mechanism allows system software to modify the Completion Timeout value. This field is hardwired to support 10 ms to 250 ms and 250 ms to 4 s. RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 653 12.1.1.37 Offset 68h: Device Control 2 Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-38. Offset 68h: Device Control 2 Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 0000h Bit Range Bit Acronym 15 :05 Reserved Offset Start: 68h Offset End: 69h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Completion Timeout Disable -- When set to 1b, this bit disables the Completion Timeout mechanism. 04 CTD If there are outstanding requests when the bit is cleared, it is permitted but not required for hardware to apply the completion timeout mechanism to the outstanding requests. If this is done, it is permitted to base the start time for each request on either the time this bit was cleared or the time each request was issued. RW Completion Timeout Value -- This field allows system software to modify the Completion Timeout value. 03 :00 CTV 0000b Default range: 40-50 ms (spec range 50 us to 50 ms) 0101b: 40-50 ms (spec range is 16 ms to 55 ms) 0110b: 160-170 ms (spec range is 65 ms to 210 ms) 1001b: 400-500 ms (spec range is 260 ms to 900 ms) 1010b: 1.6-1.7 s (spec range is 1 s to 3.5 s) RW All other values are Reserved. Software is permitted to change the value in this field at any time. For requests already pending when the Completion Timeout Value is changed, hardware is permitted to use either the new or the old value for the outstanding requests, and is permitted to base the start time for each request either on when this value was changed or on when each request w as issued. Intel(R) Communications Chipset 89xx Series - Datasheet 654 October 2012 Order Number: 327879-001US 12.0 12.1.1.38 Offset 70h: Link Control 2 Register (PCI Express*--B0:D28:F0/F1/F2/ F3) Table 12-39. Offset 70h: Link Control 2 Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit BAR: Configuration Default: 0001h Bit Range Bit Acronym 15 :04 Reserved 03 :00 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Offset Start: 70h Offset End: 71h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Target Link Speed -- This field sets an upper limit on Link operational speed by restricting the values advertised by the upstream component in its training sequences. TLS RO 0001b: 2.5 GT/s Target Link Speed All other values reserved 12.1.1.39 Offset 80h: Message Signaled Interrupt Identifiers Register (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-40. Offset 80h: Message Signaled Interrupt Identifiers Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 9005h Offset Start: 80h Offset End: 81h Power Well: Bit Acronym Bit Description 15 :08 NEXT Next Pointer -- Indicates the location of the next pointer in the list. RO 07 :00 CID Capability ID -- Capabilities ID indicates MSI. RO October 2012 Order Number: 327879-001US Sticky Bit Reset Value Bit Range Bit Access Intel(R) Communications Chipset 89xx Series - Datasheet 655 12.1.1.40 Offset 82h: Message Signaled Interrupt Message Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-41. Offset 82h: Message Signaled Interrupt Message Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 0000h Bit Range Bit Acronym 15 :08 Reserved Offset Start: 82h Offset End: 83h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved C64 64 Bit Address Capable -- Capable of generating a 32-bit message only. RO 06 :04 MME Multiple Message Enable -- These bits are R/W for software compatibility, but only one message is ever sent by the root port. RW 03 :01 MMC Multiple Message Capable -- Only one message is required. RO 07 00 MSIE MSI Enable: 0 = MSI is disabled. 1 = MSI is enabled and traditional interrupt pins are not used to generate interrupts. RW CMD.BME (B0:D28:F0/F1/F2/F3:04h:bit 2) must be set for an MSI to be generated. If CMD.BME is cleared, and this bit is set, no interrupts (not even pin based) are generated. Intel(R) Communications Chipset 89xx Series - Datasheet 656 October 2012 Order Number: 327879-001US 12.0 12.1.1.41 Offset 84h: Message Signaled Interrupt Message Address Register (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-42. Offset 84h: Message Signaled Interrupt Message Address Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 00000000h Power Well: Bit Range Bit Acronym Bit Description 31 :02 ADDR Address -- Lower 32 bits of the system specified message address, always DW aligned. 01 :00 Reserved 12.1.1.42 Offset Start: 84h Offset End: 87h Sticky Bit Reset Value Bit Access RW Reserved Offset 88h: Message Signaled Interrupt Message Data Register (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-43. Offset 88h: Message Signaled Interrupt Message Data Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 0000h Bit Range Bit Acronym 15 :00 DATA October 2012 Order Number: 327879-001US Offset Start: 88h Offset End: 89h Power Well: Bit Description Sticky Data -- This 16-bit field is programmed by system software if MSI is enabled. Its content is driven onto the lower word (PCI AD[15:0]) during the data phase of the MSI memory write transaction. Bit Reset Value Bit Access RW Intel(R) Communications Chipset 89xx Series - Datasheet 657 12.1.1.43 Offset 90h: Subsystem Vendor Capability Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-44. Offset 90h: Subsystem Vendor Capability Register (PCI Express*--B0:D28:F0/ F1/F2/F3) Description: View: PCI Size: 16 bit Default: A00Dh Bit Range Bit Acronym 15 :08 NEXT 07 :00 CID 12.1.1.44 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Offset Start: 90h Offset End: 91h Power Well: Bit Description Sticky Bit Reset Value Bit Access Next Capability -- Indicates the location of the next pointer in the list. RO Capability Identifier -- Value of 0Dh indicates this is a PCI bridge subsystem vendor capability. RO Offset 94h: Subsystem Vendor Identification Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-45. Offset 94h: Subsystem Vendor Identification Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 00000000h Bit Range Bit Acronym 31 :16 SID 15 :00 SVID Power Well: Bit Description Sticky Bit Reset Value Bit Access Subsystem Identifier -- Indicates the subsystem as identified by the vendor. This field is write once and is locked down until a bridge reset occurs (not the PCI bus reset). RWO Subsystem Vendor Identifier -- Indicates the manufacturer of the subsystem. This field is write once and is locked down until a bridge reset occurs (not the PCI bus reset). RWO Intel(R) Communications Chipset 89xx Series - Datasheet 658 Offset Start: 94h Offset End: 97h October 2012 Order Number: 327879-001US 12.0 12.1.1.45 Offset A0h: Power Management Capability Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-46. Offset A0h: Power Management Capability Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 0001h Offset Start: A0h Offset End: A1h Power Well: Bit Acronym 15 :08 NEXT Next Capability -- Indicates this is the last item in the list. RO CID Capability Identifier -- Value of 01h indicates this is a PCI power management capability. RO 07 :00 12.1.1.46 Bit Description Sticky Bit Reset Value Bit Range Bit Access Offset A2h: PCI Power Management Capabilities Register (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-47. Offset A2h: PCI Power Management Capabilities Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit Default: C802h Bit Range Bit Acronym 15 :11 PMES 10 D25 09 D15 08 :06 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration AC Offset Start: A2h Offset End: A3h Power Well: Bit Description Sticky Bit Reset Value Bit Access PME_Support -- Indicates PME# is supported for states D0, D3HOT and D3COLD. The root port does not generate PME#, but reporting that it does is necessary for some legacy operating systems to enable PME# in devices connected behind this root port. RO D2_Support -- The D2 state is not supported. RO D1_Support -- The D1 state is not supported. RO Aux_Current -- Reports 375 mA maximum suspend well current required when in the D3COLD state. RO Device Specific Initialization: 05 DSI 04 Reserved 1 = Indicates that no device-specific initialization is required. RO Reserved PME Clock: 03 02 :00 PMEC VS October 2012 Order Number: 327879-001US 1 = Indicates that PCI clock is not required to generate PME#. Version -- Indicates support for Revision 1.1 of the PCI Power Management Specification. RO RO Intel(R) Communications Chipset 89xx Series - Datasheet 659 12.1.1.47 Offset A4h: PCI Power Management Control and Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-48. Offset A4h: PCI Power Management Control and Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 00000000h Bit Range Bit Acronym 31 :24 Reserved Power Well: Bit Description BPCE Bus Power / Clock Control Enable -- Reserved per PCI Express* Base Specification, Revision 1.0a. 22 B23S B2/B3 Support -- Reserved per PCI Express* Base Specification, Revision 1.0a. 15 14 :09 08 07 :02 Reserved PMES Reserved PMEE Reserved Sticky Bit Reset Value Bit Access Reserved 23 21 :16 Offset Start: A4h Offset End: A7h Reserved PME Status: 1 = Indicates a PME was received on the downstream link. RO Reserved PME Enable: 1 = Indicates PME is enabled. The root port takes no action on this bit, but it must be R/W for some legacy operating systems to enable PME# on devices connected to this root port. This bit is sticky and resides in the resume well. The reset for this bit is RSMRST# which is not asserted during a warm reset. RW Reserved Power State -- This field is used both to determine the current power state of the root port and to set a new power state. The values are: 01 :00 PS 00 = D0 state 11 = D3HOT state RW When in the D3HOT state, the controller's configuration space is available, but the I/O and memory spaces are not. Type 1 configuration cycles are also not accepted. Interrupts are not required to be blocked as software will disable interrupts prior to placing the port into D3HOT. If software attempts to write a `10' or `01' to these bits, the write will be ignored. Intel(R) Communications Chipset 89xx Series - Datasheet 660 October 2012 Order Number: 327879-001US 12.0 12.1.1.48 Offset D4h: Miscellaneous Port Configuration Register 2 (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-49. Offset D4h: Miscellaneous Port Configuration Register 2 (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000000h Bit Range Bit Acronym 31 :05 Reserved 04 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 ASPMCOEN Offset Start: D4h Offset End: D7h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved ASPM Control Override Enable: 0 = DMI will use the values in the ASPM Control Override registers 1 = DMI will use the ASPM Registers in the Link Control register. RW This register allows BIOS to control the DMI ASPM settings instead of the OS. 03 :02 01 00 ASPMO ASPM Control Override -- Provides BIOS control of whether DMI should enter L0s or L1 or both. Bits Definition 00 Disabled 01 L0s Entry Enabled 10 L1 Entry Enabled 11 L0s and L1 Entry Enabled. RW EOIFD EOI Forwarding Disable -- When set, EOI messages are not claimed on the backbone by this port and will not be forwarded across the PCIe link. 0 = EOI forwarding is enabled. 1 = EOI forwarding is disabled. RW LICTM L1 Completion Timeout Mode: 0 = PCI Express* Specification Compliant. Completion timeout is disabled during software initiated L1, and enabled during ASPM initiate L1. 1 = Completion timeout is enabled during L1, regardless of how L1 entry was initiated. RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 661 12.1.1.49 Offset D8h: Miscellaneous Port Configuration Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-50. Offset D8h: Miscellaneous Port Configuration Register (PCI Express*-- B0:D28:F0/F1/F2/F3) (Sheet 1 of 3) Description: View: PCI Size: 32 bit Bit Range B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 08110000h Offset Start: D8h Offset End: DBh Power Well: Bit Description PMCE Power Management SCI Enable: 0 = SCI generation based on a power management event is disabled. 1 = Enables the root port to generate SCI whenever a power management event is detected. RW 30 HPCE Hot Plug SCI Enable: 0 = SCI generation based on a Hot-Plug event is disabled. 1 = Enables the root port to generate SCI whenever a Hot-Plug event is detected. RW 29 LHO Link Hold Off: 1 = Port will not take any TLP. This is used during loopback mode to fill up the downstream queue. RW ATE Address Translator Enable -- This bit is used to enable address translation via the AT bits in this register during loopback mode. 0 = Disable 1 = Enable RW 31 28 Sticky Bit Reset Value Bit Acronym Bit Access Lane Reversal: 0 = This register reads the setting of the PCIELR1 Soft Strap. 1 = PCI Express* Lanes 0-3 are reversed. No Lane reversal (default). 27 26 25 LR - The port configuration straps must be set such that Port 1 or Port 5 is configured as a x4 port using lanes 0-3, or 47 when Lane Reversal is enabled. x2 lane reversal is not supported. - This register is only valid on port 1 (for ports 1-4) or port 5 (for ports 5-8). IRBNCE Invalid Receive Bus Number Check Enable -- When set, the receive transaction layer will signal an error if the bus number of a Memory request does not fall within the range between SCBN and SBBN. If this check is enabled and the request is a memory write, it is treated as an Unsupported Request. If this check is enabled and the request is a non-posted memory read request, the request is considered a Malformed TLP and a fatal error. Messages, I/O, Config, and Completions are never checked for valid bus number. RW IRRCE Invalid Receive Range Check Enable -- When set, the receive transaction layer will treat the TLP as an Unsupported Request error if the address range of a Memory request does not outside the range between prefetchable and non-prefetchable base and limit. Messages, I/O, Configuration, and Completions are never checked for valid address ranges. RW Intel(R) Communications Chipset 89xx Series - Datasheet 662 RO October 2012 Order Number: 327879-001US 12.0 Table 12-50. Offset D8h: Miscellaneous Port Configuration Register (PCI Express*-- B0:D28:F0/F1/F2/F3) (Sheet 2 of 3) Description: View: PCI Size: 32 bit Bit Range B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 08110000h Offset Start: D8h Offset End: DBh Power Well: Bit Acronym Bit Description Sticky 24 BMERCE BME Receive Check Enable -- When set, the receive transaction layer will treat the TLP as an Unsupported Request error if a memory read or write request is received and the Bus Master Enable bit is not set. Messages, IO, Config, and Completions are never checked for BME. 23 Reserved Reserved Bit Reset Value Bit Access RW FORCEDET Detect Override: 0 = Normal operation. Detected output from AFE is sampled for presence detection. 1 = Override mode. Ignores AFE detect output and link training proceeds as if a device were detected. RW FCDL1E Flow Control During L1 Entr: 0 = No flow control update DLLPs sent during L1 Ack transmission. 1 = Flow control update DLLPs sent during L1 Ack transmission as required to meet the 30 s periodic flow control update. RW UCEL Unique Clock Exit Latency -- This value represents the L0s Exit Latency for unique-clock configurations (LCTL.CCC = 0) (B0:D28:F0/F1/F2/F3:Offset 50h:bit 6). It defaults to 512 ns to less than 1 s, but may be overridden by BIOS. RW 17 :15 CCEL Common Clock Exit Latency -- This value represents the L0s Exit Latency for common-clock configurations (LCTL.CCC = 1) (B0:D28:F0/F1/F2/F3:Offset 50h:bit 6). It defaults to 128 ns to less than 256 ns, but may be overridden by BIOS. RW 14 :08 Reserved 22 21 20 :18 Reserved Port I/OxApic Enable: 0 = Hole is disabled. 1 = A range is opened through the bridge for the following memory addresses: 07 06 :03 PAE Reserved October 2012 Order Number: 327879-001US Port # Address 1 FEC1_0000h - FEC1_7FFFh 2 FEC1_8000h - FEC1_FFFFh 3 FEC2_0000h - FEC2_7FFFh 4 FEC2_8000h - FEC2_FFFFh RW Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 663 Table 12-50. Offset D8h: Miscellaneous Port Configuration Register (PCI Express*-- B0:D28:F0/F1/F2/F3) (Sheet 3 of 3) Description: View: PCI Size: 32 bit Bit Range 02 01 00 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 08110000h Power Well: Sticky Bit Reset Value Bit Acronym Bit Description BT Bridge Type -- This register can be used to modify the Base Class and Header Type fields from the default P2P bridge to a Host Bridge. Having the root port appear as a Host Bridge is useful in some server configurations. 0 = The root port bridge type is a P2P Bridge, Header Sub-Class = 04h, and Header Type = Type 1. 1 = The root port bridge type is a P2P Bridge, Header Sub-Class = 00h, and Header Type = Type 0. RO HPME Hot Plug SMI Enable: 0 = SMI generation based on a Hot-Plug event is disabled. 1 = Enables the root port to generate SMI whenever a Hot-Plug event is detected. RW PMME Power Management SMI Enable: 0 = SMI generation based on a power management event is disabled. 1 = Enables the root port to generate SMI whenever a power management event is detected. RW Intel(R) Communications Chipset 89xx Series - Datasheet 664 Offset Start: D8h Offset End: DBh Bit Access October 2012 Order Number: 327879-001US 12.0 12.1.1.50 Offset DCh: SMI/SCI Status Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Table 12-51. Offset DCh: SMI/SCI Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit Bit Range B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 00000000h Bit Acronym Offset Start: DCh Offset End: DFh Power Well: Bit Description Sticky Bit Reset Value Bit Access Power Management SCI Status: 31 PMCS 30 HPCS 1 = PME control logic needs to generate an interrupt, and this interrupt has been routed to generate an SCI. RWC Hot Plug SCI Status: 29 :05 Reserved 1 = Hot-Plug controller needs to generate an interrupt, and has this interrupt been routed to generate an SCI. RWC Reserved Hot Plug Link Active State Changed SMI Status: 04 HPLAS 1 = SLSTS.LASC (B0:D28:F0/F1/F2/F3:5A, bit 8) transitioned from 0-to-1, and MPC.HPME (B0:D28:F0/F1/ F2/F3:D8, bit 1) is set. When this bit is set, an SMI# will be generated. RWC Hot Plug Command Completed SMI Status: 03 HPCCM 1 = SLSTS.CC (B0:D28:F0/F1/F2/F3:5A, bit 4) transitioned from 0-to-1, and MPC.HPME (B0:D28:F0/F1/ F2/F3:D8, bit 1) is set. When this bit is set, an SMI# will be generated. RWC Hot Plug Attention Button SMI Status: 02 HPABM 1 = SLSTS.ABP (B0:D28:F0/F1/F2/F3:5A, bit 0) transitioned from 0-to-1, and MPC.HPME (B0:D28:F0/F1/ F2/F3:D8, bit 1) is set. When this bit is set, an SMI# will be generated. RWC Hot Plug Presence Detect SMI Status: 01 HPPDM 1 = SLSTS.PDC (B0:D28:F0/F1/F2/F3:5A, bit 3) transitioned from 0-to-1, and MPC.HPME (B0:D28:F0/F1/ F2/F3:D8, bit 1) is set. When this bit is set, an SMI# will be generated. RWC Power Management SMI Status: 00 PMMS October 2012 Order Number: 327879-001US 1 = RSTS.PS (B0:D28:F0/F1/F2/F3:60, bit 16) transitioned from 0-to-1, and MPC.PMME (B0:D28:F0/F1/ F2/F3:D8, bit 1) is set. RWC Intel(R) Communications Chipset 89xx Series - Datasheet 665 12.1.1.51 Offset E1h: Root Port Dynamic Clock Gating Enable (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-52. Offset E1h: Root Port Dynamic Clock Gating Enable (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 8 bit Default: 00h Bit Range Bit Acronym 07 :04 Reserved 03 02 01 00 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved RO SRDLCGEN Shared Resource Dynamic Link Clock Gating Enable: 0 = Disables dynamic clock gating of the shared resource link clock domain. 1 = Enables dynamic clock gating on the root port shared resource link clock domain. Only the value from Port 1 is used for ports 1-4. Only the value from Port 5 is used for ports 5-8. RW SRDBCGEN Shared Resource Dynamic Backbone Clock Gate: 0 = Disables dynamic clock gating of the shared resource backbone clock domain. 1 = Enables dynamic clock gating on the root port shared resource backbone clock domain. Only the value from Port 1 is used for ports 1-4. Only the value from Port 5 is used for ports 5-8. RW RPDLCGEN Root Port Dynamic Link Clock Gate Enable: 0 = Disables dynamic clock gating of the root port link clock domain. 1 = Enables dynamic clock gating on the root port link clock domain. RW RPDBCGEN Root Port Dynamic Backbone Clock Gate Enable: 0 = Disables dynamic clock gating of the root port backbone clock domain. 1 = Enables dynamic clock gating on the root port backbone clock domain. RW Intel(R) Communications Chipset 89xx Series - Datasheet 666 Offset Start: E1h Offset End: E1h October 2012 Order Number: 327879-001US 12.0 12.1.1.52 Offset E8h: PCI Express* Configuration Register 1 (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-53. Offset E8h: PCI Express* Configuration Register 1 (PCI Express*--B0:D28:F0/ F1/F2/F3) Description: View: PCI Size: 32 bit Default: 00000020h Bit Range Bit Acronym 31 :02 Reserved 01 PECR1 00 Reserved 12.1.1.53 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Offset Start: E8h Offset End: EBh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved PECR1 Field 2 -- BIOS may set this bit to 1. RW Reserved Offset 104h: Uncorrectable Error Status Register (PCI Express*-- B0:D28:F0/F1/F2/F3) This register maintains its state through a platform reset. It loses its state upon suspend. Table 12-54. Offset 104h: Uncorrectable Error Status Register (PCI Express*--B0:D28:F0/ F1/F2/F3) (Sheet 1 of 2) Description: View: PCI Size: 32 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 00000000000x0xxx0x0x0000 000x0000b Bit Range Bit Acronym 31 :21 Reserved Offset Start: 104h Offset End: 107h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Unsupported Request Error Status -- Indicates an unsupported request was received. 20 URE 19 EE ECRC Error Status -- ECRC is not supported. 18 MT Malformed TLP Status -- Indicates a malformed TLP was received. RWC 17 RO Receiver Overflow Status -- Indicates a receiver overflow occurred. RWC 16 UC Unexpected Completion Status -- Indicates an unexpected completion was received. RWC 15 CA Completion Abort Status -- Indicates a completer abort was received. RWC 14 CT Completion Timeout Status -- Indicates a completion timed out. This bit is set if Completion Timeout is enabled and a completion is not returned within the time specified by the Completion TImeout Value RWC October 2012 Order Number: 327879-001US RWC RO Intel(R) Communications Chipset 89xx Series - Datasheet 667 Table 12-54. Offset 104h: Uncorrectable Error Status Register (PCI Express*--B0:D28:F0/ F1/F2/F3) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 00000000000x0xxx0x0x0000 000x0000b Bit Acronym Bit Description 13 FCPE Flow Control Protocol Error Status -- Flow Control Protocol Errors not supported. 12 PT 11 :05 04 03 :01 00 Reserved DLPE Reserved TE Poisoned TLP Status -- Indicates a poisoned TLP was received. Power Well: Sticky Bit Reset Value Bit Access RO RWC Reserved Data Link Protocol Error Status -- Indicates a data link protocol error occurred. RWC Reserved Training Error Status -- Training Errors not supported. Intel(R) Communications Chipset 89xx Series - Datasheet 668 Offset Start: 104h Offset End: 107h RO October 2012 Order Number: 327879-001US 12.0 12.1.1.54 Offset 108h: Uncorrectable Error Mask (PCI Express*--B0:D28:F0/F1/ F2/F3) When set, the corresponding error in the UES register is masked, and the logged error will cause no action. When cleared, the corresponding error is enabled. Table 12-55. Offset 108h: Uncorrectable Error Mask (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 00000000h Bit Range Bit Acronym 31 :21 Reserved Offset Start: 108h Offset End: 10Bh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Unsupported Request Error Mask: 0 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is enabled. 1 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is masked. 20 URE 19 EE ECRC Error Mask -- ECRC is not supported. MT Malformed TLP Mask: 0 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is enabled. 1 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is masked. RWO RO Receiver Overflow Mask: 0 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3/:144) is enabled. 1 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is masked. RWO UC Unexpected Completion Mask: 0 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is enabled. 1 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is masked. RWO CA Completion Abort Mask: 0 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is enabled. 1 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is masked. RWO 14 CT Completion Timeout Mask: 0 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is enabled. 1 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is masked. RWO 13 FCPE 18 17 16 15 12 11 :05 PT Reserved October 2012 Order Number: 327879-001US Flow Control Protocol Error Mask -- Flow Control Protocol Errors not supported. Poisoned TLP Mask: 0 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is enabled. 1 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is masked. RWO RO RO RWO Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 669 Table 12-55. Offset 108h: Uncorrectable Error Mask (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit Bit Range 04 03 :01 00 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 00000000h Bit Acronym DLPE Reserved TE Power Well: Bit Description Data Link Protocol Error Mask: 0 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is enabled. 1 = The corresponding error in the UES register (B0:D28:F0/F1/F2/F3:144) is masked. Sticky Bit Reset Value Bit Access RWO Reserved Training Error Mask -- Training Errors not supported Intel(R) Communications Chipset 89xx Series - Datasheet 670 Offset Start: 108h Offset End: 10Bh RO October 2012 Order Number: 327879-001US 12.0 12.1.1.55 Offset 10Ch: Uncorrectable Error Severity (PCI Express*--B0:D28:F0/ F1/F2/F3) Table 12-56. Offset 10Ch: Uncorrectable Error Severity (PCI Express*--B0:D28:F0/F1/F2/ F3) Description: View: PCI Size: 32 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: 00060011h Bit Range Bit Acronym 31 :21 Reserved Offset Start: 10Ch Offset End: 10Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Unsupported Request Error Severity: 0 = Error considered non-fatal. (Default) 1 = Error is fatal. RW 20 URE 19 EE ECRC Error Severity -- ECRC is not supported. RO 18 MT Malformed TLP Severity: 0 = Error considered non-fatal. 1 = Error is fatal. (Default) RW 17 RO Receiver Overflow Severity: 0 = Error considered non-fatal. 1 = Error is fatal. (Default) RO 16 Reserved 15 CA 14 Reserved 13 FCPE 12 PT 11 :05 04 03 :01 00 Reserved DLPE Reserved TE October 2012 Order Number: 327879-001US Reserved Completion Abort Severity: 0 = Error considered non-fatal. (Default) 1 = Error is fatal. RW Reserved Flow Control Protocol Error Severity -- Flow Control Protocol Errors not supported. RO Poisoned TLP Severity: 0 = Error considered non-fatal. (Default) 1 = Error is fatal. RW Reserved Data Link Protocol Error Severity: 0 = Error considered non-fatal. 1 = Error is fatal. (Default) RW Reserved Training Error Severity. TE is not supported. RW Intel(R) Communications Chipset 89xx Series - Datasheet 671 12.1.1.56 Offset 110h: Correctable Error Status Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-57. Offset 110h: Correctable Error Status Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Description: View: PCI Size: 32 bit Default: 00000000h Bit Range Bit Acronym 31 :14 Reserved 13 ANFES 12 RTT 11 :09 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Advisory Non-Fatal Error Status: 0 = Advisory Non-Fatal Error did not occur. 1 = Advisory Non-Fatal Error did occur. RWC Replay Timer Timeout Status -- Indicates the replay timer timed out. RWC Reserved 08 RNR Replay Number Rollover Status -- Indicates the replay number rolled over. RWC 07 BD Bad DLLP Status -- Indicates a bad DLLP was received. RWC Bad TLP Status -- Indicates a bad TLP was received. RWC 06 05 :01 00 BT Reserved RE Reserved Receiver Error Status -- Indicates a receiver error occurred. Intel(R) Communications Chipset 89xx Series - Datasheet 672 Offset Start: 110h Offset End: 113h RWC October 2012 Order Number: 327879-001US 12.0 12.1.1.57 Offset 114h: Correctable Error Mask Register (PCI Express*-- B0:D28:F0/F1/F2/F3) When set, the corresponding error in the CES register is masked, and the logged error will cause no action. When cleared, the corresponding error is enabled. Table 12-58. Offset 114h: Correctable Error Mask Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Description: View: PCI Size: 32 bit BAR: Configuration Default: 00002000h Bit Range Bit Acronym 31 :14 Reserved 13 ANFEM 12 RTT 11 :09 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Reserved Offset Start: 114h Offset End: 117h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Advisory Non-Fatal Error Mask: 0 = Does not mask Advisory Non-Fatal errors. 1 = Masks Advisory Non-Fatal errors from (a) signaling ERR_COR to the device control register and (b) updating the Uncorrectable Error Status register. This register is set by default to enable compatibility with software that does not comprehend Role-Based Error Reporting. The correctable error detected bit in device status register is set whenever the Advisory Non-Fatal error is detected, independent of this mask bit. RWC Replay Timer Timeout Mask -- Mask for replay timer timeout. RWC Reserved Replay Number Rollover Mask -- Mask for replay number rollover. 08 RNR 07 BD Bad DLLP Mask -- Mask for bad DLLP reception. RWC 06 BT Bad TLP Mask -- Mask for bad TLP reception. RWC 05 :01 00 Reserved RE October 2012 Order Number: 327879-001US RWC Reserved Receiver Error Mask -- Mask for receiver errors. RWC Intel(R) Communications Chipset 89xx Series - Datasheet 673 12.1.1.58 Offset 118h: Advanced Error Capabilities and Control Register (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-59. Offset 118h: Advanced Error Capabilities and Control Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit Default: 00000000h Bit Range Bit Acronym 31 :09 Reserved 08 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved ECE ECRC Check Enable -- ECRC is not supported. 07 ECC ECRC Check Capable -- ECRC is not supported. RO 06 EGE ECRC Generation Enable -- ECRC is not supported. RO 05 EGC ECRC Generation Capable -- ECRC is not supported. RO FEP First Error Pointer RO 04 :00 Intel(R) Communications Chipset 89xx Series - Datasheet 674 Offset Start: 118h Offset End: 11Bh RO October 2012 Order Number: 327879-001US 12.0 12.1.1.59 Offset 130h: Root Error Status Register (PCI Express*--B0:D28:F0/ F1/F2/F3) Table 12-60. Offset 130h: Root Error Status Register (PCI Express*--B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 00000000h Bit Range Bit Acronym 31 :27 AEMN 26 :07 Reserved Offset Start: 130h Offset End: 133h Power Well: Bit Description Sticky Advanced Error Interrupt Message Number -- There is only one error interrupt allocated. Bit Reset Value Bit Access RO Reserved 06 FEMR Fatal Error Messages Received -- Set when one or more Fatal Uncorrectable Error Messages have been received. 05 NFEMR Non-Fatal Error Messages Received -- Set when one or more Non-Fatal Uncorrectable error messages have been received RO 04 FUF First Uncorrectable Fatal -- Set when the first Uncorrectable Error message received is for a fatal error. RO 03 MENR Multiple ERR_FATAL/NONFATAL Received -- For the PCH, only one error will be captured. RO 02 ENR ERR_FATAL/NONFATAL Received: 0 = No error message received. 1 = Either a fatal or a non-fatal error message is received. RWC 01 MCR Multiple ERR_COR Received -- For the PCH, only one error will be captured. RO 00 CR October 2012 Order Number: 327879-001US ERR_COR Received: 0 = No error message received. 1 = A correctable error message is received. RO RWC Intel(R) Communications Chipset 89xx Series - Datasheet 675 12.1.1.60 Offset 180h: Root Complex Topology Capability List Register (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-61. Offset 180h: Root Complex Topology Capability List Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit Default: 00010005h Bit Range Bit Acronym 31 :20 NEXT 19 :16 15 :00 12.1.1.61 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Offset Start: 180h Offset End: 183h Power Well: Bit Description Sticky Bit Reset Value Bit Access Next Capability -- Indicates the next item in the list, in this case, end of list. RO CV Capability Version -- Indicates the version of the capability structure. RO CID Capability ID -- Indicates this is a root complex topology capability. RO Offset 184h: Element Self Description Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-62. Offset 184h: Element Self Description Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Description: View: PCI Size: 32 bit Bit Range B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: See register description Bit Acronym Bit Description Offset Start: 184h Offset End: 187h Power Well: Sticky Bit Reset Value Bit Access Port Number -- Indicate the ingress port number for the root port. There is a different value per port: 31 :24 23 :16 PN CID Port # Value 1 01h 2 02h 3 03h 4 04h Component ID -- This field returns the value of the ESD.CID field (PCH Config Space: Offset 0104h:bits 23:16) of the chip configuration section, that is programmed by platform BIOS, since the root port is in the same component as the RCRB. Intel(R) Communications Chipset 89xx Series - Datasheet 676 RO RO October 2012 Order Number: 327879-001US 12.0 Table 12-62. Offset 184h: Element Self Description Register (PCI Express*--B0:D28:F0/F1/ F2/F3) Description: View: PCI Size: 32 bit Default: See register description Bit Range Bit Acronym 15 :08 NLE 07 :04 Reserved 03 :00 12.1.1.62 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Offset Start: 184h Offset End: 187h Power Well: Bit Description Sticky Bit Reset Value Bit Access Number of Link Entries -- The default value of 01h indicates one link entry (corresponding to the RCRB). RO Reserved Element Type -- The default value of 0h indicates that the element type is a root port. ET RO Offset 190h: Upstream Link Description Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-63. Offset 190h: Upstream Link Description Register (PCI Express*--B0:D28:F0/ F1/F2/F3) Description: View: PCI Size: 16 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 8086h Bit Range Bit Acronym 31 :24 PN 23 :16 TCID 15 :02 Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Target Port Number -- Indicates the port number of the RCRB. RO Target Component ID -- This field returns the value of the ESD.CID field (PCH Config Space: Offset 0104h:bits 23:16) of the chip configuration section, that is programmed by platform BIOS, since the root port is in the same component as the RCRB. RO Reserved 01 LT Link Type -- Indicates that the link points to the PCH RCRB. 00 LV Link Valid -- Indicates that this link entry is valid. October 2012 Order Number: 327879-001US Offset Start: 190h Offset End: 193h RO RO Intel(R) Communications Chipset 89xx Series - Datasheet 677 12.1.1.63 Offset 198h: Upstream Link Base Address Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-64. Offset 198h: Upstream Link Base Address Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 64 bit B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Default: See register description Offset Start: 198h Offset End: 19Fh Power Well: Bit Acronym Bit Description 63 :32 BAU Base Address Upper -- The RCRB of the PCH lives in 32-bit space. RO 31 :00 BAL Base Address Lower -- This field matches the RCBA register (B0:D28:F0:Offset F0h) value in the LPC bridge. RO 12.1.1.64 Sticky Bit Reset Value Bit Range Bit Access Offset 300h: PCI Express* Configuration Register 2 (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-65. Offset 300h: PCI Express* Configuration Register 2 (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit Default: 60005007h Bit Range Bit Acronym 31 :22 Reserved 21 20 :00 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration PCR2 Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved PECR2 Field 1 -- BIOS must set this bit to 1b. RW Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 678 Offset Start: 300h Offset End: 303h October 2012 Order Number: 327879-001US 12.0 12.1.1.65 Offset 318h: PCI Express* Extended Test Mode Register (PCI Express*--B0:D28:F0/F1/F2/F3) Table 12-66. Offset 318h: PCI Express* Extended Test Mode Register (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 16 bit Default: See register description Bit Range Bit Acronym 07 :03 Reserved 02 01 :00 12.1.1.66 B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 BAR: Configuration Offset Start: 318h Offset End: 318h Power Well: Bit Description Sticky Bit Reset Value Reserved Scrambler Bypass Mode: 0 = Normal operation. Scrambler and descrambler are used. 1 = Bypasses the data scrambler in the transmit direction and the data de-scrambler in the receive direction. This functionality intended for debug/testing only. If bypassing scrambler with the PCH root port 1 in x4 configuration, each PCH root port must have this bit set. BAU Reserved Bit Access RW Reserved Offset 324h: PCI Express* Configuration Register 1 (PCI Express*-- B0:D28:F0/F1/F2/F3) Table 12-67. OFfset 324h: PCI Express* Configuration Register 1 (PCI Express*-- B0:D28:F0/F1/F2/F3) Description: View: PCI Size: 32 bit BAR: Configuration B0:D28 Bus:Device:Function: :F0/F1/ F2/F3 Default: 14000016h Bit Range Bit Acronym 31 :08 Reserved 07 :00 PEC1 Offset Start: 324h Offset End: 324h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved PEC1 Field 1 -- BIOS must program this field to 40h. RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 679 13.0 High Precision Event Timer Registers The timer registers are memory-mapped in a non-indexed scheme. This allows the processor to directly access each register without having to use an index register. The timer register space is 1024 bytes. The registers are generally aligned on 64-bit boundaries to simplify implementation with IA64 processors. There are four possible memory address ranges beginning at 1) FED0_0000h, 2) FED0_1000h, 3) FED0_2000h, 4) FED0_3000h. The choice of address range will be selected by configuration bits in the High Precision Timer Configuration Register (Chipset Config Registers = RCBA; Offset 3404h). 13.1 High Precision Behavioral Rules Behavioral Rules: * Software must not attempt to read or write across register boundaries. For example, a 32-bit access should be to offset x0h, x4h, x8h, or xCh. 32-bit accesses should not be to 01h, 02h, 03h, 05h, 06h, 07h, 09h, 0Ah, 0Bh, 0Dh, 0Eh, or 0Fh. Any accesses to these offsets will result in an unexpected behavior, and may result in a master abort. However, these accesses should not result in system hangs. 64bit accesses can only be to x0h and must not cross 64-bit boundaries. * Software should not write to read-only registers. * Software should not expect any particular or consistent value when reading reserved registers or bits. 13.1.1 Memory Mapped Registers Table 13-1. Memory Mapped Registers Offset Start Offset End Default Value Register ID - Description 00h 07h "Offset 00h: General Capabilities and Identification Register" on page 681 0429B17F808 6A201h 010h 01Fh "Offset 010h: General Configuration Register" on page 682 00000000 00000000h 020h 027h "Offset 020h: General Interrupt Status Register" on page 683 00000000 00000000h 0F0h 0F7h "Offset 0F0h: Main Counter Value Register" on page 684 N/A 100h 107h "Offset 100h: Timer n Configuration and Capabilities Register" on page 686 N/A 108h 10Fh "Offset 108h: Timer n Comparator Value Register" on page 690 N/A Intel(R) Communications Chipset 89xx Series - Datasheet 680 October 2012 Order Number: 327879-001US 13.0 13.1.1.1 Offset 00h: General Capabilities and Identification Register Table 13-2. Offset 00h: General Capabilities and Identification Register Description: View: IA F Size: 64 bit Bit Range 63 :32 Bus:Device:Function : BAR: HPTC Offset Start: 00h Offset End: 07h Default: 0429B17F8086A201h Bit Acronym Power Well: Bit Description Main Counter Tick Period: * This field indicates the period at which the counter increments in femptoseconds (10^15 seconds). COUNTER_CLK_PER_CAP * This will return 0429B17F when read. * This indicates a period of 69841279 fs (69.841279 ns). Sticky Bit Reset Value Bit Access RO VENDOR_ID_CAP Vendor ID Capability: This is a 16-bit value assigned to Intel (8086h) RO 15 LEG_RT_CAP Legacy Replacement Rout Capable: Hardwired to 1. Legacy Replacement Interrupt Rout option is supported. RO 14 Reserved Reserved. This bit returns 0 when read. 13 COUNT_SIZE_CAP Counter Size Capability: Hardwired to 1. Counter is 64-bit wide. RO Number of Timer Capability: This field indicates the number of timers in this block. 07h = Eight timers. RO Revision Identification: This indicates which revision of the function is implemented. Default value will be 01h. RO 31 :16 12 :08 NUM_TIM_CAP 07 :00 REV_ID October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 681 13.1.1.2 Offset 010h: General Configuration Register Table 13-3. Offset 010h: General Configuration Register Description: View: IA F Size: 64 bit BAR: HPTC Default: 00000000 00000000h Bit Range Bit Acronym 63 :02 Reserved 01 00 LEG_RT_CNF ENABLE_CNF Bit Description Power Well: Sticky Bit Reset Value Bit Access Reserved. These bits return 0 when read. Legacy Replacement Rout: If the ENABLE_CNF bit and the LEG_RT_CNF bit are both set, then the interrupts will be routed as follows: * Timer 0 is routed to IRQ0 in 8259 or IRQ2 in the I/O APIC * Timer 1 is routed to IRQ8 in 8259 or IRQ8 in the I/O APIC * Timer 2-n is routed as per the routing in the timer n config registers. * If the Legacy Replacement Rout bit is set, the individual routing bits for Timers 0 and 1 (APIC) will have no impact. * If the Legacy Replacement Rout bit is not set, the individual routing bits for each of the timers are used. This bit will default to 0. BIOS can set it to 1 to enable the legacy replacement routing, or 0 to disable the legacy replacement routing. Overall Enable: This bit must be set to enable any of the timers to generate interrupts. If this bit is 0, then the main counter will halt (will not increment) and no interrupts will be caused by any of these timers. For level-triggered interrupts, if an interrupt is pending when the ENABLE_CNF bit is changed from 1 to 0, the interrupt status indications (in the various Txx_INT_STS bits) will not be cleared. Software must write to the Txx_INT_STS bits to clear the interrupts. Note: RW RW This bit will default to 0. BIOS can set it to 1 or 0 Intel(R) Communications Chipset 89xx Series - Datasheet 682 Offset Start: 010h Offset End: 017h Bus:Device:Function: October 2012 Order Number: 327879-001US 13.0 13.1.1.3 Offset 020h: General Interrupt Status Register Table 13-4. Offset 020h: General Interrupt Status Register Description: View: IA F Size: 64 bit BAR: HPTC Offset Start: 020h Offset End: 027h Bus:Device:Function: Default: 00000000 00000000h Bit Range Bit Acronym 63 :08 Reserved Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved. These bits return 0 when read. 07 T07_INT_STS Timer 7 Interrupt Active -- Same functionality as Timer 0. RW 06 T06_INT_STS Timer 6 Interrupt Active -- Same functionality as Timer 0. RW 05 T05_INT_STS Timer 5 Interrupt Active -- Same functionality as Timer 0. RW 04 T04_INT_STS Timer 4 Interrupt Active -- Same functionality as Timer 0. RW 03 T03_INT_STS Timer 3 Interrupt Active -- Same functionality as Timer 0. RW 02 T02_INT_STS Timer 2 Interrupt Active -- Same functionality as Timer 0. RW 01 T01_INT_STS Timer 1 Interrupt Active -- Same functionality as Timer 0. RW Timer 0 Interrupt Active -- * The functionality of this bit depends on whether the edge or level-triggered mode is used for this timer. 00 If set to level-triggered mode: * This bit defaults to 0. * This bit will be set by hardware if the corresponding timer interrupt is active. * Once the bit is set, it can be cleared by software by T00_INT_STS writing a 1 to the same bit position. * Writes of 0 to this bit will have no effect. RWC If set to edge-triggered mode: * This bit should be ignored by software. * Software should always write 0 to this bit. Note: October 2012 Order Number: 327879-001US Defaults to 0. In edge triggered mode, this bit will always read as 0 and writes will have no effect. Intel(R) Communications Chipset 89xx Series - Datasheet 683 13.1.1.4 Offset 0F0h: Main Counter Value Register Table 13-5. Offset 0F0h: Main Counter Value Register Description: View: IA F Size: 64 bit Bit Range BAR: HPTC Offset Start: 0F0h Offset End: 0F7h Bus:Device:Function: Default: N/A Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access Counter Value: Reads return the current value of the counter. Writes load the new value to the counter. 63 :00 Notes: * Writes to this register should only be done while the counter is halted. * Reads to this register return the current value of the main counter. * 32-bit counters will always return 0 for the upper 32-bits of this register. COUNTER_VAL[63:0] * If 32-bit software attempts to read a 64-bit counter, it should first halt the counter. Since this delays the interrupts for all of the timers, this should be done only if the consequences are understood. It is strongly recommended that 32bit software only operate the timer in 32-bit mode. * Reads to this register are monotonic. No two consecutive reads return the same value. The second of two reads always returns a larger value (unless the timer has rolled over to 0). Intel(R) Communications Chipset 89xx Series - Datasheet 684 RW October 2012 Order Number: 327879-001US 13.0 13.1.1.5 Offset 100h: Timer n Configuration and Capabilities Register Address Offset: Timer Timer Timer Timer Timer Timer Timer Timer 0: 1: 2: 3: 4: 5: 6: 7: 100-107h, 120-127h, 140-147h, 160-167h, 180-187h, 1A0-1A7h, 1C0-1C7h, 1E0-1E7h, The letter n can be 0, 1, 2, 3, 4, 5, 6, or 7 referring to Timer 0, 1, 2, 3, 4, 5, 6, or 7. Reads or writes to unimplemented timers should not be attempted. Read from any unimplemented registers will return an undetermined value. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 685 Table 13-6. Offset 100h: Timer n Configuration and Capabilities Register (Sheet 1 of 4) Description: View: IA F BAR: HPTC Size: 64 bit Bit Range 63 :56 Offset Start: 100h Offset End: 107h Bus:Device:Function: Default: N/A Power Well: Bit Acronym Reserved Bit Description Sticky Bit Reset Value Bit Access Reserved. These bits will return 0 when read. Timer Interrupt Rout Capability: * In general, a 32-bit field of this register, Bits[63:32] is allocated to indicate which interrupts (IRQ[31:0]) in the 8259 or I/O(x) APIC this timer's interrupt can be routed to. * The Bits are mapped as follows: -- Bit[32] = IRQ[0] -- .. -- Bit[63] = IRQ[31] These bits [55-52, 44, 43) are indicated here because they are the interrupts supported by the Timers. Timer 0, 1 (if LEG_RT_CNF = 0): * Bits 52, 53, 54, and 55 in this field (corresponding to IRQ 20, 21, 22, and 23) have a value of 1 to indicate the supported interrupts by this Timer. Writes will have no effect. 55 :52,44,43 Timer 2: * Bits 43, 52, 53, 54, and 55 in this field (corresponding to IRQ 11, 20, 21, 22, and 23) TIMERn_INT_ROUT_CAP have a value of 1 to indicate the supported interrupts by this Timer. Writes will have no effect. RO Timer 3: * Bits 44, 52, 53, 54, and 55 in this field (corresponding to IRQ 12, 20, 21, 22, and 23) have a value of to indicate the supported interrupts by this Timer1. Writes will have no effect. Timer 4 - 7: * This field is always 0 as interrupts from these timers can only be delivered via direct interrupt messages. Note: Note: If IRQ 11 is used for HPET #2, software should ensure IRQ 11 is not shared with any other devices to ensure the proper operation of HPET #2. If IRQ 12 is used for HPET #3, software should ensure IRQ 12 is not shared with any other devices to ensure the proper operation of HPET #3. 51 :45 Reserved Reserved. 42 :14 Reserved Reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 686 October 2012 Order Number: 327879-001US 13.0 Table 13-6. Offset 100h: Timer n Configuration and Capabilities Register (Sheet 2 of 4) Description: View: IA F BAR: HPTC Size: 64 bit Bit Range Offset Start: 100h Offset End: 107h Bus:Device:Function: Default: N/A Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access Timer Interrupt Rout Configuration: This 5-bit binary field, decodes into a decimal number [0 - 31], to indicate the routing for the interrupt to the 8259 or I/O (x) APIC. Software writes to this field to select which interrupt in the 8259 or I/O (x) will be used for this timer's interrupt. If the value is not supported by this particular timer, then the value read back will not match what is written. The software must only write valid values. Note: For example, (10100b = 20 decimal) to indicate the routing of the interrupt to IRQ[20] Timer 4, 5, 6, 7: This field is Read-only and reads will return 0. 13 :09 Notes: * If the interrupt is handled using the 8259, only interrupts 0-15 are applicable and valid. Software must not program any value other TIMERn_INT_ROUT_CNF than 0-15 in this field. * If the Legacy Replacement Rout bit is set, then Timers 0 and 1 will have a different routing, and this bit field has no effect for those two timers. RW Timer 0,1: * Software is responsible to make sure it programs a valid value (decimal 20, 21, 22, or 23) for this field. The PCH logic does not check the validity of the value written. Timer 2: * Software is responsible to make sure it programs a valid value (decimal 11, 20, 21, 22, or 23) for this field. The PCH logic does not check the validity of the value written. Timer 3: * Software is responsible to make sure it programs a valid value (decimal 12, 20, 21, 22, or 23) for this field. The PCH logic does not check the validity of the value written. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 687 Table 13-6. Offset 100h: Timer n Configuration and Capabilities Register (Sheet 3 of 4) Description: View: IA F BAR: HPTC Size: 64 bit Bit Range Offset Start: 100h Offset End: 107h Bus:Device:Function: Default: N/A Power Well: Bit Acronym Bit Description Sticky Bit Reset Value Bit Access Timer n 32-bit Mode Configuration: Software can set this bit to force a 64-bit timer to behave as a 32-bit timer. Timer 0: * Bit is read/write (default to 0) * 0 = 64 bit; 1 = 32 bit 08 TIMERn_32Mode_CNF Timers 1, 2, 3, 4, 5, 6, 7: * Hardwired to 0. Writes have no effect (since these two timers are 32-bits). Note: 07 Reserved RW/RO When this bit is set to 1, the hardware counter will do a 32-bit operation on comparator match and rollovers; thus, the upper 32-bit of the Timer 0 Comparator Value register is ignored. The upper 32-bit of the main counter is not involved in any rollover from lower 32-bit of the main counter and becomes all zeros Reserved. This bit returns 0 when read. Timer n Value Set: Software uses this bit only for Timer 0 if it has been set to periodic mode. By writing this bit to a 1, the software is then allowed to directly set the timer's accumulator. Software does not have to write this bit back to 1 (it automatically clears). 06 TIMERn_VAL_SET_CNF Software should not write a 1 to this bit position if the timer is set to non-periodic mode. Note: 05 04 TIMERn_SIZE_CAP TIMERn_PER_INT_CAP This bit will return 0 when read. Writes will only have an effect for Timer 0 if it is set to periodic mode. Writes will have no effect for Timers 1, 2, 3, 4, 5, 6, 7. Timer n Size: This read only field indicates the size of the timer. Timer 0: * Value is 1 (64-bits). Timers 1, 2, 3, 4, 5, 6, 7.: * Value is 0 (32-bits). RO Periodic Interrupt Capable: If this bit is 1, the hardware supports a periodic mode for this timer's interrupt. Timer 0: * Hardwired to 1 (supports the periodic interrupt). Timers 1, 2, 3, 4, 5, 6, 7.: * Hardwired to 0 (does not support periodic interrupt). RO Intel(R) Communications Chipset 89xx Series - Datasheet 688 RW October 2012 Order Number: 327879-001US 13.0 Table 13-6. Offset 100h: Timer n Configuration and Capabilities Register (Sheet 4 of 4) Description: View: IA F BAR: HPTC Size: 64 bit Bit Range 03 Default: N/A Bit Acronym TIMERn_TYPE_CNF Power Well: Bit Description Timer n Type: Timer 0: * Bit is read/write. * 0 = Disable timer to generate periodic interrupt; * 1 = Enable timer to generate a periodic interrupt. Timers 1, 2, 3, 4, 5, 6, 7.: * Hardwired to 0. Writes have no affect. Sticky Bit Reset Value Bit Access RW/RO TIMERn_INT_ENB_CNF Timer n Interrupt Enable: This bit must be set to enable timer n to cause an interrupt when it times out. * 0 = Disable (Default). The timer can still count and generate appropriate status bits, but will not cause an interrupt. * 1 = Enable. RW 01 TIMERn_INT_TYPE_CNF Timer Interrupt Type: Timers 1, 2, 3 * 0 = The timer interrupt is edge triggered. This means that an edge-type interrupt is generated. If another interrupt occurs, another edge will be generated. * 1 = The timer interrupt is level triggered. This means that a level-triggered interrupt is generated. The interrupt will be held active until it is cleared by writing to the bit in the General Interrupt Status Register. If another interrupt occurs before the interrupt is cleared, the interrupt will remain active. Timers 4, 5, 6, 7: * This bit is Read-Only, and will return 0 when read RW 00 Reserved 02 Note: Offset Start: 100h Offset End: 107h Bus:Device:Function: Reserved. These bits will return 0 when read. Reads and writes to unimplemented timers should not be attempted. Read from any unimplemented registers will return an undetermined value October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 689 13.1.1.6 Offset 108h: Timer n Comparator Value Register Address Offset: Timer Timer Timer Timer Timer Timer Timer Timer 0: 1: 2: 3: 4: 5: 6: 7: 108h - 10Fh, 128h - 12Fh, 148h - 14Fh, 168h - 16Fh, 188h - 18Fh, 1A8h - 1AFh, 1C8h - 1CFh, 1E8h - 1EFh Table 13-7. Offset 108h: Timer n Comparator Value Register Description: View: IA F Size: 64 bit Bit Range BAR: HPTC Offset Start: 108h Offset End: 10Fh Bus:Device:Function: Default: N/A Bit Acronym Power Well: Bit Description Sticky Bit Reset Value Bit Access Timer Compare Value: Reads to this register return the current value of the comparator. Timers 1, 2, 3 4, 5, 6, 7 (4, 5, 6, 7) are configured to non-periodic mode: * Writes to this register load the value against which the main counter should be compared for this timer. * When the main counter equals the value last written to this register, the corresponding interrupt can be generated (if so enabled). * The value in this register does not change based on the interrupt being generated. 63 :00 Timer 0 is configured to periodic mode: * When the main counter equals the value last written to this register, the corresponding interrupt can be generated (if so enabled). * After the main counter equals the value in this register, the value in this register is increased by the value last written to the register. For example, if the value written to the register is TIMERn_COMP_VAL 00000123h, then: 1. An interrupt will be generated when the main counter reaches 00000123h 2. The value in this register will then be adjusted by the hardware to 00000246h. 3. Another interrupt will be generated when the main counter reaches 00000246h. 4. The value in this register will then be adjusted by the hardware to 00000369h * As each periodic interrupt occurs, the value in this register will increment. When the incremented value is greater than the maximum value possible for this register (FFFFFFFFh for a 32-bit timer or FFFFFFFFFFFFFFFFh for a 64-bit timer), the value will wrap around through 0. For example, if the current value in a 32-bit timer is FFFF0000h and the last value written to this register is 20000, then after the next interrupt the value will change to 00010000h Default value for each timer is all 1s for the bits that are implemented. For example, a 32-bit timer has a default value of 00000000FFFFFFFFh. A 64-bit timer has a default value of FFFFFFFFFFFFFFFFh. RW Intel(R) Communications Chipset 89xx Series - Datasheet 690 October 2012 Order Number: 327879-001US 14.0 14.0 Serial Peripheral Interface (SPI) The Serial Peripheral Interface resides in memory mapped space. This function contains registers that allow for the setup and programming of devices that reside on the SPI interface. Note: All registers in this function (including memory-mapped registers) must be addressable in byte, word, and dword quantities. The software must always make register accesses on natural boundaries (i.e., DWord accesses must be on dword boundaries; word accesses on word boundaries, etc.) In addition, the memory-mapped register space must not be accessed with the LOCK semantic exclusive-access mechanism. If software attempts exclusive-access mechanisms to the SPI memory-mapped space, the results are undefined. 14.1 Serial Peripheral Interface Memory Mapped Configuration Registers The SPI Host Interface registers are memory-mapped in the RCRB (Root Complex Register Block) Chipset Register Space with a base address (SPIBAR) of 3800h and are located within the range of 3800h to 39FFh. The address for RCRB can be found in RCBA Register. The individual registers are then accessible at SPIBAR + Offset as indicated in the following table. These memory mapped registers must be accessed in byte, word, or dword quantities. 14.1.1 Serial Peripheral Interface (SPI) Register Address Map Table 14-1. Serial Peripheral Interface (SPI) Register Address Map (Sheet 1 of 2) Offset Start Offset End Register ID - Description Default Value 00h 03h "Offset 00h: BIOS Flash Primary Region Register (SPI Memory Mapped Configuration Registers)" on page 693 00000000h 04h 05h "Offset 04h: Hardware Sequencing Flash Status Register (SPI Memory Mapped Configuration Registers)" on page 694 0000h 06h 07h "Offset 06h: Hardware Sequencing Flash Control Register SPI Memory Mapped Configuration Registers)" on page 696 0000h 08h 0Bh "Offset 08h: Flash Address Register (SPI Memory Mapped Configuration Registers)" on page 697 00000000h 14h + (4*N) 17h + (4*N) "Offset 14h: Flash Data [N] Register (SPI Memory Mapped Configuration Registers)" on page 698 00000000h 50h 53h "Offset 50h: Flash Regions Access Permissions Register (SPI Memory Mapped Configuration Registers)" on page 699 00000202h 54h 57h "Offset 54h: Flash Region 0 (Flash Descriptor) Register (SPI Memory Mapped Configuration Registers)" on page 700 00000000h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 691 Table 14-1. Serial Peripheral Interface (SPI) Register Address Map (Sheet 2 of 2) Offset Start Offset End Default Value Register ID - Description 58h 5Bh "Offset 58h: Offset 00h: Flash Region 1 (BIOS Descriptor) Register (SPI Memory Mapped Configuration Registers)" on page 700 00000000h 5Ch 5Fh "Offset 5Ch: Flash Region 2 (Intel ME) Register (SPI Memory Mapped Configuration Registers)" on page 701 00000000h 60h 63h "Offset 60h: Flash Region 3 Register (SPI Memory Mapped Configuration Registers)" on page 701 00000000h 64h 67h "Offset 64h: Flash Region 4 (Platform Data) Register (SPI Memory Mapped Configuration Registers)" on page 702 00000000h 74h 77h "Offset 74h: Protected Range 0 Register (SPI Memory Mapped Configuration Registers)" on page 702 00000000h 78h 7Bh "Offset 78h: Protected Range 1 Register (SPI Memory Mapped Configuration Registers)" on page 703 00000000h 7Ch 7Fh "Offset 7Ch: Protected Range 2 Register (SPI Memory Mapped Configuration Registers)" on page 704 00000000h 80h 83h "Offset 80h: Protected Range 3 Register (SPI Memory Mapped Configuration Registers)" on page 705 00000000h 84h 87h "Offset 84h: Protected Range 4 Register (SPI Memory Mapped Configuration Registers)" on page 706 00000000h 90h 90h "Offset 90h: Software Sequencing Flash Status Register (SPI Memory Mapped Configuration Registers)" on page 707 00h 91h 93h "Offset 91h: Software Sequencing Flash Control Register (SPI Memory Mapped Configuration Registers)" on page 708 F80000h 94h 95h "Offset 94h: Prefix Opcode Configuration Register (SPI Memory Mapped Configuration Registers)" on page 710 0000h 96h 97h "Offset 96h: Opcode Type Configuration Register (SPI Memory Mapped Configuration Registers)" on page 711 0000h 98h 9Fh "Offset 98h: Opcode Menu Configuration Register (SPI Memory Mapped Configuration Registers)" on page 712 00000000000 00000h A0h A3h "Offset A0h: BIOS Base Address Configuration Register (SPI Memory Mapped Configuration Registers)" on page 713 00000000h B0h B3h "Offset B0h: Flash Descriptor Observability Control Register (SPI Memory Mapped Configuration Registers)" on page 714 00000000h B4h B7h "Offset B4h: Flash Descriptor Observability Data Register SPI Memory Mapped Configuration Registers)" on page 714 00000000h C0h C3h "Offset C0h: Additional Flash Control Register (SPI Memory Mapped Configuration 00000000h Registers)" on page 715 C4h C7h "Offset C4h: Host Lower Vendor Specific Component Capabilities Register (SPI Memory Mapped Configuration Registers)" on page 716 00000000h C0h C3h "Offset C8h: Host Upper Vendor Specific Component Capabilities Register (SPI Memory Mapped Configuration Registers)" on page 718 00000000h D0h D3h "Offset D0h: Flash Partition Boundary (SPI Memory Mapped Configuration Registers)" on page 720 00000000h Intel(R) Communications Chipset 89xx Series - Datasheet 692 October 2012 Order Number: 327879-001US 14.0 14.1.1.1 Offset 00h: BIOS Flash Primary Region Register (SPI Memory Mapped Configuration Registers) Note: This register is only applicable when SPI device is in descriptor mode. Table 14-2. Offset 00h: BIOS Flash Primary Region Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit BAR: SPIBAR Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Range Bit Acronym 31 :29 Reserved 28 :16 PRL 15 :13 Reserved 12 :00 PRB October 2012 Order Number: 327879-001US Offset Start: 00h Offset End: 03h Power Well: Bit Description Sticky Reserved Bit Reset Value Bit Access O BIOS Flash Primary Region Limit -- This specifies address bits 24:12 for the Primary Region Limit. The value in this register loaded from the contents in the Flash Descriptor.FLREG1.Region Limit Reserved RO O BIOS Flash Primary Region Base -- This specifies address bits 24:12 for the Primary Region Base The value in this register is loaded from the contents in the Flash Descriptor.FLREG1.Region Base RO Intel(R) Communications Chipset 89xx Series - Datasheet 693 14.1.1.2 Offset 04h: Hardware Sequencing Flash Status Register (SPI Memory Mapped Configuration Registers) Table 14-3. Offset 04h: Hardware Sequencing Flash Status Register (SPI Memory Mapped Configuration Registers) (Sheet 1 of 2) Description: View: PCI Size: 16 bit Bit Range 15 14 13 12 :06 05 04 :03 BAR: SPIBAR Bus:Device:Function: Default: 0000h Offset Start: 04h Offset End: 05h Power Well: Sticky Bit Reset Value Bit Acronym Bit Description FLOCKDN Flash Configuration Lock-Down -- When set to 1, those Flash Program Registers that are locked down by this FLOCKDN bit cannot be written. Once set to 1, this bit can only be cleared by a hardware reset due to a global reset or host partition reset in an Intel(R) ME enabled system. RWL FDV Flash Descriptor Valid -- This bit is set to a 1 if the Flash Controller read the correct Flash Descriptor Signature. If the Flash Descriptor Valid bit is not 1, software cannot use the Hardware Sequencing registers, but must use the software sequencing registers. Any attempt to use the Hardware Sequencing registers will result in the FCERR bit being set. RO FDOPSS Flash Descriptor Override Pin-Strap Status -- This register reflects the value the Flash Descriptor Override Pin-Strap. 0 = The Flash Descriptor Override strap is set 1 = No override RO Reserved Reserved Bit Access O SCIP SPI Cycle In Progress -- Hardware sets this bit when software sets the Flash Cycle Go (FGO) bit in the Hardware Sequencing Flash Control register. This bit remains set until the cycle completes on the SPI interface. Hardware automatically sets and clears this bit so that software can determine when read data is valid and/or when it is safe to begin programming the next command. Software must only program the next command when this bit is 0. This field is only applicable when in Descriptor mode and Hardware sequencing is being used. RO BERASE Block/Sector Erase Size -- This field identifies the erasable sector size for all Flash components. Valid Bit Settings: 00 = 256 Byte 01 = 4 K Byte 10 = 8 K Byte 11 = 64 K Byte If the FLA is less than FPBA then this field reflects the value in the LVSCC.LBES register. If the FLA is greater or equal to FPBA then this field reflects the value in the UVSCC.UBES register. This field is only applicable when in Descriptor mode and Hardware sequencing is being used. RO Intel(R) Communications Chipset 89xx Series - Datasheet 694 B0:D31 :F0 October 2012 Order Number: 327879-001US 14.0 Table 14-3. Offset 04h: Hardware Sequencing Flash Status Register (SPI Memory Mapped Configuration Registers) (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range 02 01 00 BAR: SPIBAR Bus:Device:Function: B0:D31 :F0 Default: 0000h Offset Start: 04h Offset End: 05h Power Well: Bit Description AEL Access Error Log -- Hardware sets this bit to a 1 when an attempt was made to access the BIOS region using the direct access method or an access to the BIOS Program Registers that violated the security restrictions. This bit is simply a log of an access security violation. This bit is cleared by software writing a 1. This field is only applicable when in Descriptor mode and Hardware sequencing is being used. RWC FCERR Flash Cycle Error -- Hardware sets this bit to 1 when an program register access is blocked to the FLASH due to one of the protection policies or when any of the programmed cycle registers is written while a programmed access is already in progress. This bit remains asserted until cleared by software writing a 1 or until hardware reset occurs due to a global reset or host partition reset in an Intel(R) ME enabled system. Software must clear this bit before setting the FLASH Cycle GO bit in this register. This field is only applicable when in Descriptor mode and Hardware sequencing is being used. RWC FDONE Flash Cycle Done -- The PCH sets this bit to 1 when the SPI Cycle completes after software previously set the FGO bit. This bit remains asserted until cleared by software writing a 1 or hardware reset due to a global reset or host partition reset in an Intel(R) ME enabled system. When this bit is set and the SPI SMI# Enable bit is set, an internal signal is asserted to the SMI# generation block. Software must make sure this bit is cleared prior to enabling the SPI SMI# assertion for a new programmed access. This field is only applicable when in Descriptor mode and Hardware sequencing is being used. RWC October 2012 Order Number: 327879-001US Sticky Bit Reset Value Bit Acronym Bit Access Intel(R) Communications Chipset 89xx Series - Datasheet 695 14.1.1.3 Offset 06h: Hardware Sequencing Flash Control Register (SPI Memory Mapped Configuration Registers) Table 14-4. Offset 06h: Hardware Sequencing Flash Control Register SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 16 bit Bit Range BAR: SPIBAR Bit Acronym FSMIE 14 Reserved Power Well: Bit Description Flash SPI SMI# Enable -- When set to 1, the SPI asserts an SMI# request whenever the Flash Cycle Done bit is 1. Reserved Sticky Bit Reset Value Bit Access RW O RW Reserved RO FCYCLE Flash Cycle -- This field defines the Flash SPI cycle type generated to the FLASH when the FGO bit is set as defined below: 00 = Read (1 up to 64 bytes by setting FDBC) 01 = Reserved 10 = Write (1 up to 64 bytes by setting FDBC) 11 = Block Erase RW FGO Flash Cycle Go -- A write to this register with a 1 in this bit initiates a request to the Flash SPI Arbiter to start a cycle. This register is cleared by hardware when the cycle is granted by the SPI arbiter to run the cycle on the SPI bus. When the cycle is complete, the FDONE bit is set. Software is forbidden to write to any register in the HSFLCTL register between the FGO bit getting set and the FDONE bit being cleared. Any attempt to violate this rule will be ignored by hardware. Hardware allows other bits in this register to be programmed for the same transaction when writing this bit to 1. This saves an additional memory write. This bit always returns 0 on reads. RWS FDBC 07 :03 Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 696 Offset Start: 06h Offset End: 07h Flash Data Byte Count -- This field specifies the number of bytes to shift in or out during the data portion of the SPI cycle. The contents of this register are 0s based with 0b representing 1 byte and 111111b representing 64 bytes. The number of bytes transferred is the value of this field plus 1. This field is ignored for the Block Erase command. 13 :08 00 B0:D31 :F0 Default: 0000h 15 02 :01 Bus:Device:Function: October 2012 Order Number: 327879-001US 14.0 14.1.1.4 Offset 08h: Flash Address Register (SPI Memory Mapped Configuration Registers) Table 14-5. Offset 08h: Flash Address Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit BAR: SPIBAR B0:D31 :F0 Default: 00000000h Bit Range Bit Acronym 32 :25 Reserved 24 :00 Bus:Device:Function: FLA October 2012 Order Number: 327879-001US Offset Start: 08h Offset End: 0Bh Power Well: Bit Description Sticky Reserved Bit Reset Value Bit Access O Flash Linear Address -- The FLA is the starting byte linear address of a SPI Read or Write cycle or an address within a Block for the Block Erase command. The Flash Linear Address must fall within a region for which BIOS has access permissions. Hardware must convert the FLA into a Flash Physical Address (FPA) before running this cycle on the SPI bus. RW Intel(R) Communications Chipset 89xx Series - Datasheet 697 14.1.1.5 Offset 10h: Flash Data 0 Register (SPI Memory Mapped Configuration Registers) 14.1.1.6 Offset 14h: Flash Data [N] Register (SPI Memory Mapped Configuration Registers) SPIBAR SPIBAR SPIBAR SPIBAR SPIBAR SPIBAR SPIBAR SPIBAR SPIBAR SPIBAR SPIBAR SPIBAR SPIBAR SPIBAR SPIBAR + + + + + + + + + + + + + + + 14h 18h 1Ch 20h 24h 28h 2Ch 30h 34h 38h 3Ch 40h 44h 48h 4Ch Table 14-6. Offset 14h: Flash Data [N] Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Bus:Device:Function: B0:D31 :F0 BAR: SPIBAR Default: 00000000h Bit Range Bit Acronym 31 :00 FD[N] Power Well: Bit Description Flash Data [N] -- Similar definition as Flash Data 0. However, this register does not begin shifting until FD[N1] has completely shifted in/out. Intel(R) Communications Chipset 89xx Series - Datasheet 698 Offset Start: 14h + (4*N) Offset End: 17h + (4*N) Sticky Bit Reset Value Bit Access RW October 2012 Order Number: 327879-001US 14.0 14.1.1.7 Offset 50h: Flash Regions Access Permissions Register (SPI Memory Mapped Configuration Registers) Table 14-7. Offset 50h: Flash Regions Access Permissions Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Bit Range 31 :24 23 :16 15 :08 07 :00 BAR: SPIBAR Bus:Device:Function: B0:D31 :F0 Default: 00000202h Bit Acronym Offset Start: 50h Offset End: 53h Power Well: Bit Description Sticky Bit Reset Value Bit Access BMWAG BIOS Master Write Access Grant -- Each bit [31:29] corresponds to Master[7:0]. BIOS can grant one or more masters write access to the BIOS region 1 overriding the permissions in the Flash Descriptor. Master[1] is Host processor/BIOS, Master[2] is Intel(R) Management Engine, Master[3] is Host processor/GbE. Master[0] and Master[7:4] are reserved. The contents of this register are locked by the FLOCKDN bit. RW BMRAG BIOS Master Read Access Grant -- Each bit [28:16] corresponds to Master[7:0]. BIOS can grant one or more masters read access to the BIOS region 1 overriding the read permissions in the Flash Descriptor. Master[1] is Host processor/BIOS, Master[2] is Intel(R) Management Engine, Master[3] is Host processor/GbE. Master[0] and Master[7:4] are reserved. The contents of this register are locked by the FLOCKDN bit RW BRWA BIOS Region Write Access -- Each bit [15:8] corresponds to Regions [7:0]. If the bit is set, this master can erase and write that particular region through register accesses. The contents of this register are that of the Flash Descriptor. Flash Master 1 Master Region Write Access OR a particular master has granted BIOS write permissions in their Master Write Access Grant register or the Flash Descriptor Security Override strap is set. RO BRRA BIOS Region Read Access -- Each bit [7:0] corresponds to Regions [7:0]. If the bit is set, this master can read that particular region through register accesses. The contents of this register are that of the Flash Descriptor.Flash Master 1.Master Region Write Access OR a particular master has granted BIOS read permissions in their Master Read Access Grant register or the Flash Descriptor Security Override strap is set. RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 699 14.1.1.8 Offset 54h: Flash Region 0 (Flash Descriptor) Register (SPI Memory Mapped Configuration Registers) Table 14-8. Offset 54h: Flash Region 0 (Flash Descriptor) Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit BAR: SPIBAR Offset Start: 54h Offset End: 57h B0:D31 :F0 Default: 00000000h Bit Range Bit Acronym 31 :29 Reserved 28 :16 RL 15 :13 Reserved 12 :00 RB/FDBAR 14.1.1.9 Bus:Device:Function: Power Well: Bit Description Bit Reset Value Sticky Bit Access Reserved O Region Limit -- This specifies address bits 24:12 for the Region 0 Limit. The value in this register is loaded from the contents in the Flash Descriptor.FLREG0.Region Limit RO Reserved O Region Base/Flash Descriptor Base Address Region -- This specifies address bits 24:12 for the Region 0 Base The value in this register is loaded from the contents in the Flash Descriptor.FLREG0.Region Base RO Offset 58h: Offset 00h: Flash Region 1 (BIOS Descriptor) Register (SPI Memory Mapped Configuration Registers) Table 14-9. Offset 58h: Offset 00h: Flash Region 1 (BIOS Descriptor) Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Default: 00000000h Bit Range Bit Acronym 31 :29 Reserved 28 :16 RL 15 :13 Reserved 12 :00 B0:D31 Bus:Device:Function: :F0 BAR: SPIBAR RB Power Well: Bit Description Reserved Region Limit -- This specifies address bits 24:12 for the Region 1 Limit. The value in this register is loaded from the contents in the Flash Descriptor.FLREG1.Region Limit Reserved Region Base -- This specifies address bits 24:12 for the Region 1 Base The value in this register is loaded from the contents in the Flash Descriptor.FLREG1.Region Base Intel(R) Communications Chipset 89xx Series - Datasheet 700 Offset Start: 58h Offset End: 5Bh Sticky Bit Reset Value Bit Access O RO O RO October 2012 Order Number: 327879-001US 14.0 14.1.1.10 Offset 5Ch: Flash Region 2 (Intel ME) Register (SPI Memory Mapped Configuration Registers) Table 14-10. Offset 5Ch: Flash Region 2 (Intel ME) Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit BAR: SPIBAR B0:D31 :F0 Default: 00000000h Bit Range Bit Acronym 31 :29 Reserved 28 :16 RL 15 :13 Reserved 12 :00 RB 14.1.1.11 Bus:Device:Function: Offset Start: 5Ch Offset End: 5Fh Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved O Region Limit -- This specifies address bits 24:12 for the Region 2 Limit. The value in this register is loaded from the contents in the Flash Descriptor.FLREG2.Region Limit RO Reserved O Region Base -- This specifies address bits 24:12 for the Region 2 Base The value in this register is loaded from the contents in the Flash Descriptor.FLREG2.Region Base RO Offset 60h: Flash Region 3 Register (SPI Memory Mapped Configuration Registers) Table 14-11. Offset 60h: Flash Region 3 Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit BAR: SPIBAR B0:D31 Bus:Device:Function: :F0 Default: 00000000h Bit Range Bit Acronym 31 :00 Reserved October 2012 Order Number: 327879-001US Power Well: Bit Description Reserved Offset Start: 60h Offset End: 63h Sticky Bit Reset Value Bit Access O Intel(R) Communications Chipset 89xx Series - Datasheet 701 14.1.1.12 Offset 64h: Flash Region 4 (Platform Data) Register (SPI Memory Mapped Configuration Registers) Table 14-12. Offset 64h: Flash Region 4 (Platform Data) Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit BAR: SPIBAR B0:D31 :F0 Default: 00000000h Bit Range Bit Acronym 31 :00 Reserved 14.1.1.13 Bus:Device:Function: Offset Start: 64h Offset End: 67h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved O Offset 74h: Protected Range 0 Register (SPI Memory Mapped Configuration Registers) Table 14-13. Offset 74h: Protected Range 0 Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Bit Range 31 30 :29 28 :16 15 14 :13 12 :00 B0:D31 Bus:Device:Function: :F0 BAR: SPIBAR Default: 00000000h Power Well: Bit Acronym Bit Description WPE Write Protection Enable -- When set, this bit indicates that the Base and Limit fields in this register are valid and that writes and erases directed to addresses between them (inclusive) must be blocked by hardware. The base and limit fields are ignored when this bit is cleared. Reserved Reserved Sticky Bit Reset Value Bit Access RW O PRL Protected Range Limit -- This field corresponds to FLA address bits 24:12 and specifies the upper limit of the protected range. Address bits 11:0 are assumed to be FFFh for the limit comparison. Any address greater than the value programmed in this field is unaffected by this protected range. RW RPE Read Protection Enable -- When set, this bit indicates that the Base and Limit fields in this register are valid and that read directed to addresses between them (inclusive) must be blocked by hardware. The base and limit fields are ignored when this bit is cleared. RW Reserved PRB Reserved Protected Range Base -- This field corresponds to FLA address bits 24:12 and specifies the lower base of the protected range. Address bits 11:0 are assumed to be 000h for the base comparison. Any address less than the value programmed in this field is unaffected by this protected range. Intel(R) Communications Chipset 89xx Series - Datasheet 702 Offset Start: 74h Offset End: 77h O RW October 2012 Order Number: 327879-001US 14.0 14.1.1.14 Offset 78h: Protected Range 1 Register (SPI Memory Mapped Configuration Registers) Table 14-14. Offset 78h: Protected Range 1 Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Bit Range 31 30 :29 28 :16 15 14 :13 12 :00 BAR: SPIBAR Bus:Device:Function: B0:D31 :F0 Default: 00000000h Power Well: Bit Acronym Bit Description WPE Write Protection Enable -- When set, this bit indicates that the Base and Limit fields in this register are valid and that writes and erases directed to addresses between them (inclusive) must be blocked by hardware. The base and limit fields are ignored when this bit is cleared. Reserved Offset Start: 78h Offset End: 7Bh Sticky Reserved Bit Reset Value Bit Access RW O PRL Protected Range Limit -- This field corresponds to FLA address bits 24:12 and specifies the upper limit of the protected range. Address bits 11:0 are assumed to be FFFh for the limit comparison. Any address greater than the value programmed in this field is unaffected by this protected range. RW RPE Read Protection Enable -- When set, this bit indicates that the Base and Limit fields in this register are valid and that read directed to addresses between them (inclusive) must be blocked by hardware. The base and limit fields are ignored when this bit is cleared. RW Reserved PRB October 2012 Order Number: 327879-001US Reserved O Protected Range Base -- This field corresponds to FLA address bits 24:12 and specifies the lower base of the protected range. Address bits 11:0 are assumed to be 000h for the base comparison. Any address less than the value programmed in this field is unaffected by this protected range. RW Intel(R) Communications Chipset 89xx Series - Datasheet 703 14.1.1.15 Offset 7Ch: Protected Range 2 Register (SPI Memory Mapped Configuration Registers) Table 14-15. Offset 7Ch: Protected Range 2 Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Bit Range 31 30 :29 28 :16 15 14 :13 12 :00 BAR: SPIBAR Bus:Device:Function: Default: 00000000h Bit Acronym Bit Description WPE Write Protection Enable -- When set, this bit indicates that the Base and Limit fields in this register are valid and that writes and erases directed to addresses between them (inclusive) must be blocked by hardware. The base and limit fields are ignored when this bit is cleared. Reserved Offset Start: 7Ch Offset End: 7Fh Power Well: Reserved Sticky Bit Reset Value Bit Access RW O PRL Protected Range Limit -- This field corresponds to FLA address bits 24:12 and specifies the upper limit of the protected range. Address bits 11:0 are assumed to be FFFh for the limit comparison. Any address greater than the value programmed in this field is unaffected by this protected range. RW RPE Read Protection Enable -- When set, this bit indicates that the Base and Limit fields in this register are valid and that read directed to addresses between them (inclusive) must be blocked by hardware. The base and limit fields are ignored when this bit is cleared. RW Reserved PRB Reserved Protected Range Base -- This field corresponds to FLA address bits 24:12 and specifies the lower base of the protected range. Address bits 11:0 are assumed to be 000h for the base comparison. Any address less than the value programmed in this field is unaffected by this protected range. Intel(R) Communications Chipset 89xx Series - Datasheet 704 B0:D31 :F0 O RW October 2012 Order Number: 327879-001US 14.0 14.1.1.16 Offset 80h: Protected Range 3 Register (SPI Memory Mapped Configuration Registers) Table 14-16. Offset 80h: Protected Range 3 Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Bit Range 31 30 :29 28 :16 15 14 :13 12 :00 BAR: SPIBAR Bus:Device:Function: B0:D31 :F0 Default: 00000000h Power Well: Bit Acronym Bit Description WPE Write Protection Enable -- When set, this bit indicates that the Base and Limit fields in this register are valid and that writes and erases directed to addresses between them (inclusive) must be blocked by hardware. The base and limit fields are ignored when this bit is cleared. Reserved Offset Start: 80h Offset End: 83h Sticky Reserved Bit Reset Value Bit Access RW O PRL Protected Range Limit -- This field corresponds to FLA address bits 24:12 and specifies the upper limit of the protected range. Address bits 11:0 are assumed to be FFFh for the limit comparison. Any address greater than the value programmed in this field is unaffected by this protected range. RW RPE Read Protection Enable -- When set, this bit indicates that the Base and Limit fields in this register are valid and that read directed to addresses between them (inclusive) must be blocked by hardware. The base and limit fields are ignored when this bit is cleared. RW Reserved PRB October 2012 Order Number: 327879-001US Reserved O Protected Range Base -- This field corresponds to FLA address bits 24:12 and specifies the lower base of the protected range. Address bits 11:0 are assumed to be 000h for the base comparison. Any address less than the value programmed in this field is unaffected by this protected range. RW Intel(R) Communications Chipset 89xx Series - Datasheet 705 14.1.1.17 Offset 84h: Protected Range 4 Register (SPI Memory Mapped Configuration Registers) Table 14-17. Offset 84h: Protected Range 4 Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Bit Range 31 30 :29 28 :16 15 14 :13 12 :00 BAR: SPIBAR Bus:Device:Function: Default: 00000000h Bit Acronym Bit Description WPE Write Protection Enable -- When set, this bit indicates that the Base and Limit fields in this register are valid and that writes and erases directed to addresses between them (inclusive) must be blocked by hardware. The base and limit fields are ignored when this bit is cleared. Reserved Offset Start: 84h Offset End: 87h Power Well: Reserved Sticky Bit Reset Value Bit Access RW O PRL Protected Range Limit -- This field corresponds to FLA address bits 24:12 and specifies the upper limit of the protected range. Address bits 11:0 are assumed to be FFFh for the limit comparison. Any address greater than the value programmed in this field is unaffected by this protected range. RW RPE Read Protection Enable -- When set, this bit indicates that the Base and Limit fields in this register are valid and that read directed to addresses between them (inclusive) must be blocked by hardware. The base and limit fields are ignored when this bit is cleared. RW Reserved PRB Reserved Protected Range Base -- This field corresponds to FLA address bits 24:12 and specifies the lower base of the protected range. Address bits 11:0 are assumed to be 000h for the base comparison. Any address less than the value programmed in this field is unaffected by this protected range. Intel(R) Communications Chipset 89xx Series - Datasheet 706 B0:D31 :F0 O RW October 2012 Order Number: 327879-001US 14.0 14.1.1.18 Offset 90h: Software Sequencing Flash Status Register (SPI Memory Mapped Configuration Registers) Table 14-18. Offset 90h: Software Sequencing Flash Status Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 8 bit BAR: SPIBAR B0:D31 :F0 Default: 00h Bit Range Bit Acronym 07 :05 Reserved 04 Bus:Device:Function: AEL Offset Start: 90h Offset End: 90h Power Well: Bit Description Sticky Reserved Bit Reset Value Bit Access O Access Error Log -- This bit reflects the value of the Hardware Sequencing Status AEL register. RO FCERR Flash Cycle Error -- Hardware sets this bit to 1 when a programmed access is blocked from running on the SPI interface due to one of the protection policies or when any of the programmed cycle registers is written while a programmed access is already in progress. This bit remains asserted until cleared by software writing a 1 or hardware reset due to a global reset or host partition reset in an Intel(R) ME enabled system. RWC 02 CDS Cycle Done Status -- The PCH sets this bit to 1 when the SPI Cycle completes (i.e., SCIP bit is 0) after software sets the GO bit. This bit remains asserted until cleared by software writing a 1 or hardware reset due to a global reset or host partition reset in an Intel(R) ME enabled system. When this bit is set and the SPI SMI# Enable bit is set, an internal signal is asserted to the SMI# generation block. Software must make sure this bit is cleared prior to enabling the SPI SMI# assertion for a new programmed access. RWC 01 Reserved 03 00 SCIP October 2012 Order Number: 327879-001US Reserved O SPI Cycle In Progress -- Hardware sets this bit when software sets the SPI Cycle Go bit in the Command register. This bit remains set until the cycle completes on the SPI interface. Hardware automatically sets and clears this bit so that software can determine when read data is valid and/or when it is safe to begin programming the next command. Software must only program the next command when this bit is 0. RO Intel(R) Communications Chipset 89xx Series - Datasheet 707 14.1.1.19 Offset 91h: Software Sequencing Flash Control Register (SPI Memory Mapped Configuration Registers) Table 14-19. Offset 91h: Software Sequencing Flash Control Register (SPI Memory Mapped Configuration Registers) (Sheet 1 of 2) Description: View: PCI Size: 24 bit BAR: SPIBAR Bus:Device:Function: Default: F80000h Bit Range Bit Acronym 23 :19 Reserved Offset Start: 91h Offset End: 93h Power Well: Bit Description Reserved - BIOS must set this field to `11111'b Sticky Bit Reset Value Bit Access O SCF SPI Cycle Frequency -- This register sets frequency to use for all SPI software sequencing cycles (write, erase, fast read, read status, etc.) except for the read cycle which always run at 20 MHz. 000 = 20 MHz 001 = 33 MHz 100 = 50 MHz All other values reserved. This register is locked when the SPI Configuration LockDown bit is set. RW 15 SME SPI SMI# Enable -- When set to 1, the SPI asserts an SMI# request whenever the Cycle Done Status bit is 1. RW 14 DS Data Cycle -- When set to 1, there is data that corresponds to this transaction. When 0, no data is delivered for this cycle, and the DBC and data fields themselves are don't cares RW Data Byte Count -- This field specifies the number of bytes to shift in or out during the data portion of the SPI cycle. The valid settings (in decimal) are any value from 0 to 63. The number of bytes transferred is the value of this field plus 1. When this field is 00_0000b, then there is 1 byte to transfer and that 11_1111b means there are 64 bytes to transfer. RW 18 :16 13 :08 07 06 :04 03 DBC Reserved Reserved O COP Cycle Opcode Pointer -- This field selects one of the programmed opcodes in the Opcode Menu to be used as the SPI Command/Opcode. In the case of an Atomic Cycle Sequence, this determines the second command. -- R/W. RW SPOP Sequence Prefix Opcode Pointer -- This field selects one of the two programmed prefix opcodes for use when performing an Atomic Cycle Sequence. A value of 0 points to the opcode in the least significant byte of the Prefix Opcodes register. By making this programmable, the PCH supports flash devices that have different opcodes for enabling writes to the data space vs. status register. RW Intel(R) Communications Chipset 89xx Series - Datasheet 708 B0:D31 :F0 October 2012 Order Number: 327879-001US 14.0 Table 14-19. Offset 91h: Software Sequencing Flash Control Register (SPI Memory Mapped Configuration Registers) (Sheet 2 of 2) Description: View: PCI Size: 24 bit Bit Range BAR: SPIBAR Bus:Device:Function: B0:D31 :F0 Default: F80000h Bit Acronym Offset Start: 91h Offset End: 93h Power Well: Bit Description Sticky Bit Reset Value Bit Access ACS Atomic Cycle Sequence -- When set to 1 along with the SCGO assertion, the PCH will execute a sequence of commands on the SPI interface without allowing the LAN component to arbitrate and interleave cycles. The sequence is composed of: Atomic Sequence Prefix Command (8-bit opcode only) Primary Command specified below by software (can include address and data) Polling the Flash Status Register (opcode 05h) until bit 0 becomes 0b. The SPI Cycle in Progress bit remains set and the Cycle Done Status bit remains unset until the Busy bit in the Flash Status Register returns 0. RW 01 SCGO SPI Cycle Go -- This bit always returns 0 on reads. However, a write to this register with a 1 in this bit starts the SPI cycle defined by the other bits of this register. The "SPI Cycle in Progress" (SCIP) bit gets set by this action. Hardware must ignore writes to this bit while the Cycle In Progress bit is set. Hardware allows other bits in this register to be programmed for the same transaction when writing this bit to 1. This saves an additional memory write. RWS 00 Reserved 02 October 2012 Order Number: 327879-001US Reserved O Intel(R) Communications Chipset 89xx Series - Datasheet 709 14.1.1.20 Offset 94h: Prefix Opcode Configuration Register (SPI Memory Mapped Configuration Registers) This register is not writable when the Flash Configuration Lock-Down bit (SPIBAR + 04h:15) is set. Table 14-20. Offset 94h: Prefix Opcode Configuration Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 16 bit BAR: SPIBAR Bus:Device:Function: B0:D31 :F0 Default: 0000h Offset Start: 94h Offset End: 95h Power Well: Bit Description Bit Acronym 15 :08 PO1 Prefix Opcode 1-- Software programs an SPI opcode into this field that is permitted to run as the first command in an atomic cycle sequence. RW 07 :00 PO0 Prefix Opcode 0 -- Software programs an SPI opcode into this field that is permitted to run as the first command in an atomic cycle sequence. RW Intel(R) Communications Chipset 89xx Series - Datasheet 710 Sticky Bit Reset Value Bit Range Bit Access October 2012 Order Number: 327879-001US 14.0 14.1.1.21 Offset 96h: Opcode Type Configuration Register (SPI Memory Mapped Configuration Registers) Entries in this register correspond to the entries in the Opcode Menu Configuration register. The definition below only provides write protection for opcodes that have addresses associated with them. Therefore, any erase or write opcodes that do not use an address should be avoided (for example, "Chip Erase" and "Auto-Address Increment Byte Program") This register is not writable when the SPI Configuration Lock-Down bit (SPIBAR + 00h:15) is set. Table 14-21. Offset 96h: Opcode Type Configuration Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 16 bit BAR: SPIBAR B0:D31 Bus:Device:Function: :F0 Default: 0000h Offset Start: 96h Offset End: 97h Power Well: 15 :14 OT7 Opcode Type 7 -- See the description for bits 1:0 RW 13 :12 OT6 Opcode Type 6 -- See the description for bits 1:0 RW 11 :10 OT5 Opcode Type 5 -- See the description for bits 1:0 RW 09 :08 OT4 Opcode Type 4 -- See the description for bits 1:0 RW 07 :06 OT3 Opcode Type 3 -- See the description for bits 1:0 RW 05 :04 OT2 Opcode Type 2 -- See the description for bits 1:0 RW 03 :02 OT1 Opcode Type 1 -- See the description for bits 1:0 RW OT0 Opcode Type 0 -- This field specifies information about the corresponding Opcode 0. This information allows the hardware to 1) know whether to use the address field and 2) provide BIOS and Shared Flash protection capabilities. The encoding of the two bits is: 00 = No address associated with this Opcode; Read cycle type 01 = No address associated with this Opcode; Write cycle type 10 = Address required; Read cycle type 11 = Address required; Write cycle type RW October 2012 Order Number: 327879-001US Sticky Bit Access Bit Acronym 01 :00 Bit Description Bit Reset Value Bit Range Intel(R) Communications Chipset 89xx Series - Datasheet 711 14.1.1.22 Offset 98h: Opcode Menu Configuration Register (SPI Memory Mapped Configuration Registers) Eight entries are available in this register to give BIOS a sufficient set of commands for communicating with the flash device, while also restricting what malicious software can do. This keeps the hardware flexible enough to operate with a wide variety of SPI devices. It is recommended that BIOS avoid programming Write Enable opcodes in this menu. Malicious software could then perform writes and erases to the SPI flash without using the atomic cycle mechanism. This could cause functional failures in a shared flash environment. Write Enable opcodes should only be programmed in the Prefix Opcodes. This register is not writable when the SPI Configuration Lock-Down bit (SPIBAR + 00h:15) is set. Table 14-22. Offset 98h: Opcode Menu Configuration Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 64 bit Bus:Device:Function: B0:D31 :F0 BAR: SPIBAR Default: 0000000000000000h Offset Start: 98h Offset End: 9Fh Power Well: Bit Range Bit Acronym 63 :56 AO7 Allowable Opcode 7 -- See the description for bits 7:0 RW 55 :48 AO6 Allowable Opcode 6 -- See the description for bits 7:0 RW 47 :40 AO5 Allowable Opcode 5 -- See the description for bits 7:0 RW 39 :32 AO4 Allowable Opcode 4 -- See the description for bits 7:0 RW 31 :24 AO3 Allowable Opcode 3 -- See the description for bits 7:0 RW 23 :16 AO2 Allowable Opcode 2 -- See the description for bits 7:0 RW 15 :08 AO1 Allowable Opcode 1 -- See the description for bits 7:0 RW 07 :00 AO0 Allowable Opcode 0 -- Software programs an SPI opcode into this field for use when initiating SPI commands through the Control Register. RW Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 712 Sticky Bit Reset Value Bit Access October 2012 Order Number: 327879-001US 14.0 14.1.1.23 Offset A0h: BIOS Base Address Configuration Register (SPI Memory Mapped Configuration Registers) Eight entries are available in this register to give BIOS a sufficient set of commands for communicating with the flash device, while also restricting what malicious software can do. This keeps the hardware flexible enough to operate with a wide variety of SPI devices. Table 14-23. Offset A0h: BIOS Base Address Configuration Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Bit Range 31 :24 23 :08 07 :00 BAR: SPIBAR Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym Offset Start: A0h Offset End: A3h Power Well: Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved BOSF Bottom of System Flash-- This field determines the bottom of the System BIOS. The PCH will not run programmed commands nor memory reads whose address field is less than this value. this field corresponds to bits 23:8 of the 3-byte address; bits 7:0 are assumed to be 00h for this vector when comparing to a potential SPI address. Note: The SPI host controller prevents any programmed cycle using the address register with an address less than the value in this register. Some flash devices specify that the Read ID command must have an address of 0000h or 0001h. If this command must be supported with these device, it must be performed with the BIOS BAR Reserved Reserved October 2012 Order Number: 327879-001US RW O Intel(R) Communications Chipset 89xx Series - Datasheet 713 14.1.1.24 Offset B0h: Flash Descriptor Observability Control Register (SPI Memory Mapped Configuration Registers) This register that can be used to observe the contents of the Flash Descriptor that is stored in the PCH Flash Controller. This register is only applicable when SPI device is in descriptor mode. Table 14-24. Offset B0h: Flash Descriptor Observability Control Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Bit Range 31 :15 BAR: SPIBAR Bus:Device:Function: Default: 00000000h Bit Acronym Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved O 14 :12 FDSS 11 :02 FDSI Flash Descriptor Section Index -- Selects the DW offset within the Flash Descriptor Section to observe. 14.1.1.25 Offset Start: B0h Offset End: B3h Power Well: Flash Descriptor Section Select -- Selects which section within the loaded Flash Descriptor to observe. 000 = Flash Signature and Descriptor Map 001 = Component 010 = Region 011 = Master 111 = Reserved 01 :00 B0:D31 :F0 Reserved Reserved RW RW O Offset B4h: Flash Descriptor Observability Data Register (SPI Memory Mapped Configuration Registers) This register that can be used to observe the contents of the Flash Descriptor that is stored in the PCH Flash Controller. Table 14-25. Offset B4h: Flash Descriptor Observability Data Register SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Bus:Device:Function: B0:D31 :F0 BAR: SPIBAR Default: 00000000h Bit Range Bit Acronym 31 :00 FDSD Power Well: Bit Description Flash Descriptor Section Data -- Returns the DW of data to observe as selected in the Flash Descriptor Observability Control. Intel(R) Communications Chipset 89xx Series - Datasheet 714 Offset Start: B4h Offset End: B7h Sticky Bit Reset Value Bit Access RO October 2012 Order Number: 327879-001US 14.0 14.1.1.26 Offset C0h: Additional Flash Control Register (SPI Memory Mapped Configuration Registers) Table 14-26. Offset C0h: Additional Flash Control Register (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Bit Range 31 :03 02 :01 00 BAR: SPIBAR Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym Offset Start: C0h Offset End: C3h Power Well: Bit Description Sticky Reserved Reserved Bit Reset Value Bit Access O IDCGE Flash Controller Interface Dynamic Clock Gating Enable: 0 = Flash Controller Interface Dynamic Clock Gating is Disabled 1 = Flash Controller Interface Dynamic Clock Gating is Enabled Other configurations are Reserved. RW CDCGE Flash Controller Core Dynamic Clock Gating Enable: 0 = Flash Controller Core Dynamic Clock Gating is Disabled 1 = Flash Controller Core Dynamic Clock Gating is Enabled RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 715 14.1.1.27 Offset C4h: Host Lower Vendor Specific Component Capabilities Register (SPI Memory Mapped Configuration Registers) All attributes described in LVSCC must apply to all flash space below the FPBA, even if it spans between two separate flash parts. This register is only applicable when the SPI device is in descriptor mode. Table 14-27. Offset C4h: Host Lower Vendor Specific Component Capabilities Register (SPI Memory Mapped Configuration Registers) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 :24 23 22 :16 15 :08 07 :05 04 BAR: SPIBAR Bus:Device:Function: Default: 00000000h Bit Acronym Bit Description Reserved Reserved LVCL Vendor Component Lock -- This register locks itself when set. 0 = The lock bit is not set 1 = The Vendor Component Lock bit is set. This bit applies to both UVSCC and LVSCC registers. Reserved Reserved LEO Lower Erase Opcode -- This register is programmed with the Flash erase instruction opcode required by the vendor's Flash component. This register is locked by the Vendor Component Lock (LVCL) bit. Reserved Reserved LWEWS Offset Start: C4h Offset End: C7h Power Well: Write Enable on Write Status -- This register is locked by the Vendor Component Lock (LVCL) bit. 0 = No automatic write of 00h will be made to the SPI flash's status register. 1 = A write of 00h to the SPI flash's status register will be sent on EVERY write and erase to the SPI flash. 06h 01h 00h is the opcode sequence used to unlock the Status register. 1 - This bit should not be set to `1' if there are non volatile bits in the SPI flash's status register. This may lead to premature flash wear out 2 - This is not an atomic sequence. If the SPI component's status register is non-volatile, then BIOS should issue an atomic software sequence cycle to unlock the flash part. 3 - Bit 3 and bit 4 should NOT be both set to `1'. Intel(R) Communications Chipset 89xx Series - Datasheet 716 B0:D31 :F0 Sticky Bit Reset Value Bit Access O RW OO RW O RW October 2012 Order Number: 327879-001US 14.0 Table 14-27. Offset C4h: Host Lower Vendor Specific Component Capabilities Register (SPI Memory Mapped Configuration Registers) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 03 BAR: SPIBAR Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym LWSR Offset Start: C4h Offset End: C7h Power Well: Bit Description Sticky Lower Write Status Required -- This register is locked by the Vendor Component Lock (LVCL) bit. 0 = No automatic write of 00h will be made to the SPI flash's status register) 1 = A write of 00h to the SPI flash's status register will be sent on EVERY write and erase to the SPI flash. 50h 01h 00h is the opcode sequence used to unlock the Status register. 1 - This bit should not be set to `1' if there are non volatile bits in the SPI flash's status register. This may lead to premature flash wear out 2 - This is not an atomic sequence. If the SPI component's status register is non-volatile, then BIOS should issue an atomic software sequence cycle to unlock the flash part. 3 - Bit 3 and bit 4 should NOT be both set to `1'. Bit Reset Value Bit Access RW Lower Write Granularity -- This register is locked by the Vendor Component Lock (LVCL) bit. 0 = 1 Byte 1 = 64 Byte 02 01 :00 LWG LBES 1 - If more than one Flash component exists, this field must be set to the lowest common write granularity of the different Flash components. 2 - If using 64 B write, BIOS must ensure that multiple byte writes do not occur over 256 B boundaries. This will lead to corruption as the write will wrap around the page boundary on the SPI flash part. This is a a feature page writable SPI flash. Lower Block/Sector Erase Size -- This field identifies the erasable sector size for all Flash components. 00 = 256 Byte 01 = 4 KB 10 = 8 KB 11 = 64 KB This register is locked by the Vendor Component Lock (LVCL) bit. Hardware takes no action based on the value of this register. The contents of this register are to be used only by software and can be read in the HSFSTS.BERASE register in both the BIOS and the GbE program registers if FLA is less than FPBA. October 2012 Order Number: 327879-001US RW RW Intel(R) Communications Chipset 89xx Series - Datasheet 717 14.1.1.28 Offset C8h: Host Upper Vendor Specific Component Capabilities Register (SPI Memory Mapped Configuration Registers) All attributes described in UVSCC must apply to all flash space equal to or above the FPBA, even if it spans between two separate flash parts. This register is only applicable when SPI device is in descriptor mode. To prevent this register from being modified you must use LVSCC.VCL bit. Table 14-28. Offset C8h: Host Upper Vendor Specific Component Capabilities Register (SPI Memory Mapped Configuration Registers) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 :16 15 :08 07 :05 04 BAR: SPIBAR Bus:Device:Function: Default: 00000000h Bit Acronym Bit Description Reserved Reserved. UEO Upper Erase Opcode -- This register is programmed with the Flash erase instruction opcode required by the vendor's Flash component. This register is locked by the Vendor Component Lock (UVCL) bit. Reserved Reserved UWEWS Offset Start: C0h Offset End: C3h Power Well: Write Enable on Write Status -- This register is locked by the Vendor Component Lock (UVCL) bit. 0 = No automatic write of 00h will be made to the SPI flash's status register) 1 = A write of 00h to the SPI flash's status register will be sent on EVERY write and erase to the SPI flash. 06h 01h 00h is the opcode sequence used to unlock the Status register. 1 - This bit should not be set to `1' if there are non volatile bits in the SPI flash's status register. This may lead to premature flash wear out 2 - This is not an atomic sequence. If the SPI component's status register is non-volatile, then BIOS should issue an atomic software sequence cycle to unlock the flash part. 3 - Bit 3 and bit 4 should NOT be both set to `1'. Intel(R) Communications Chipset 89xx Series - Datasheet 718 B0:D31 :F0 Sticky Bit Reset Value Bit Access O RW O RW October 2012 Order Number: 327879-001US 14.0 Table 14-28. Offset C8h: Host Upper Vendor Specific Component Capabilities Register (SPI Memory Mapped Configuration Registers) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: SPIBAR Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym Offset Start: C0h Offset End: C3h Power Well: Bit Description Sticky Bit Reset Value Bit Access Upper Write Status Required -- This register is locked by the Vendor Component Lock (UVCL) bit. 0 = No automatic write of 00h will be made to the SPI flash's status register) 1 = A write of 00h to the SPI flash's status register will be sent on EVERY write and erase to the SPI flash. 50h 01h 00h is the opcode sequence used to unlock the Status register. 03 UWSR Notes: 1. This bit should not be set to `1' if there are non volatile bits in the SPI flash's status register. This may lead to premature flash wear out 2. This is not an atomic sequence. If the SPI component's status register is non-volatile, then BIOS should issue an atomic software sequence cycle to unlock the flash part. 3. Bit 3 and bit 4 should NOT be both set to `1'. RW Upper Write Granularity -- This register is locked by the Vendor Component Lock (UVCL) bit. 0 = 1 Byte 1 = 64 Byte 02 01 :00 UWG UBES 1 - If more than one Flash component exists, this field must be set to the lowest common write granularity of the different Flash components. 2 - If using 64 B write, BIOS must ensure that multiple byte writes do not occur over 256 B boundaries. This will lead to corruption as the write will wrap around the page boundary on the SPI flash part. This is a a feature page writable SPI flash. Upper Block/Sector Erase Size -- This field identifies the erasable sector size for all Flash components. Valid Bit Settings: 00 = 256 Byte 01 = 4 KB 10 = 8 KB 11 = 64 KB This register is locked by the Vendor Component Lock (UVCL) bit. Hardware takes no action based on the value of this register. The contents of this register are to be used only by software and can be read in the HSFSTS.BERASE register in both the BIOS and the GbE program registers if FLA is greater or equal to FPBA. October 2012 Order Number: 327879-001US RW RW Intel(R) Communications Chipset 89xx Series - Datasheet 719 14.1.1.29 Offset D0h: Flash Partition Boundary (SPI Memory Mapped Configuration Registers) Table 14-29. Offset D0h: Flash Partition Boundary (SPI Memory Mapped Configuration Registers) Description: View: PCI Size: 32 bit Bit Range 31 :13 12 :00 14.2 BAR: SPIBAR Bus:Device:Function: B0:D31 :F0 Default: 00000000h Bit Acronym Power Well: Bit Description Sticky Reserved Reserved FPBA Offset Start: D0h Offset End: D3h Bit Reset Value Bit Access O Flash Partition Boundary Address -- This register reflects the value of Flash Descriptor Component FPBA field. RO Flash Descriptor Records The following section describes the data structure of the Flash Descriptor on the SPI device. These are not registers within the PCH. 14.2.1 OEM Section Memory Address: F00h Default Value: X Size: 256 Bytes 256 bytes are reserved at the top of the Flash Descriptor for use by the OEM. The information stored by the OEM can only be written during the manufacturing process as the Flash Descriptor read/write permissions must be set to Read Only when the computer leaves the manufacturing floor. The PCH Flash controller does not read this information. FFh is suggested to reduce programming time. Intel(R) Communications Chipset 89xx Series - Datasheet 720 October 2012 Order Number: 327879-001US 15.0 15.0 UART / WDT (SIW) 15.1 Overview The Serial I/O unit and Watchdog Timer (SIW) are similar to currently available Super I/O controllers. It is specifically designed for integration into the ILB. It is connected via the LPC bus and currently consists of two UARTs, a Serial Interrupt Controller, a Watchdog Timer and the LPC interface. 15.2 LPC Logical Devices 4 and 5: Serial Ports (UART1 and UART2) The UARTs are controlled via programmed I/O. The basic programming model is the same for both UARTs with the only difference being the Logical Device Number assigned to each. 15.2.1 UART Register Details There are 12 registers in the UART. These registers share eight address locations in the I/O address space. Table 15-4 shows the registers and their addresses as offsets of a base address. The state of the Divisor Latch Bit (DLAB), which is the MOST significant bit of the Serial Line Control Register, affects the selection of certain of the UART registers. The DLAB bit must be set high by the system software to access the Baud Rate Generator Divisor Latches. Table 15-1. Summary of UART Registers in I/O Space (DLAB=0) Offset Start 00h Offset End 00h Register ID - Description "Offset 00h: RBR--Receive Buffer Register" on page 723 Default Value 00h 00h 00h "Offset 00h: THR--Transmit Holding Register" on page 723 00h 01h 01h "Offset 01h: IER--Interrupt Enable Register" on page 724 00h Table 15-2. Summary of UART Registers in I/O Space (DLAB=1) Offset Start Offset End Register ID - Description Default Value 00h 00h "Offset 00h: DLL--Programmable Baud Rate Generator Divisor Latch Register Low" 02h on page 736 01h 01h "Offset 01h: DLH--Programmable Baud Rate Generator Divisor Latch Register High" on page 737 October 2012 Order Number: 327879-001US 00h Intel(R) Communications Chipset 89xx Series - Datasheet 721 Table 15-3. Summary of UART Timer Registers in I/O Space Offset Start Offset End Default Value Register ID - Description 02h 02h "Offset 02h: IIR--Interrupt Identification Register" on page 725 01h 02h 02h "Offset 02h: FCR--FIFO Control Register" on page 727 00h 03h 03h "Offset 03h: LCR--Line Control Register" on page 728 00h 04h 04h "Offset 04h: MCR--Modem Control Register" on page 730 00h 05h 05h "Offset 05h: LSR--Line Status Register" on page 732 60h 06h 06h "Offset 06h: MSR--Modem Status Register" on page 735 00h 07h 07h "Offset 07h: SCR--Scratchpad Register" on page 736 00h Table 15-4. Internal Register Descriptions Note: UART Register Addresses (Base + offset) DLAB Bit Value Base 0 Receive BUFFER (Read-Only) Base 0 Transmit BUFFER (Write-Only) Base + 01H 0 Interrupt Enable (Read/Write) Base + 02H X Interrupt I.D. (Read-Only) Base + 02H X FIFO Control (Write-Only) Base + 03H X Line Control (Read/Write) Base + 04H X Modem Control (Read/Write) Base + 05H X Line Status (Read-Only) Base + 06H X Modem Status (Read-Only) Base + 07H X Scratch Pad (Read/Write) Base 1 Divisor Latch (Lower Byte, Read/Write) Base + 01H 1 Divisor Latch (Upper Byte, Read/Write) Base Address for the UART registers listed in Table 15-4 is configurable. See Section 15.4.1, "SIW Configuration Register Summary" on page 745 for details. Intel(R) Communications Chipset 89xx Series - Datasheet 722 Register Accessed October 2012 Order Number: 327879-001US 15.0 15.2.1.1 Offset 00h: RBR--Receive Buffer Register In non-FIFO mode, this register holds the character received by the UART's Receive Shift Register. If fewer than eight bits are received, the bits are right-justified and the leading bits are zeroed. Reading the register empties the register and resets the Data Ready (DR) bit in the Line Status Register to zero. Other (error) bits in the Line Status Register are not cleared. In FIFO mode, this register latches the value of the data byte at the top of the FIFO. Table 15-5. Offset 00h: RBR--Receive Buffer Register Description: View: IA F Size: 8 bit Offset Start: 00h Offset End: 00h Base Address: Base (IO) (DLAB = 0) Default: 00h Power Well: Core Bit Range Bit Acronym Bit Description 07 :00 RB_7_0 Data byte received (bits [7:0]), least significant bit first 15.2.1.2 Sticky Bit Reset Value Bit Access 00h RO Offset 00h: THR--Transmit Holding Register This register holds the next data byte to be transmitted. When the Transmit Shift Register becomes empty, the contents of the Transmit Holding Register are loaded into the shift register and the transmit data request (TDRQ) bit in the Line Status Register is set to one. Table 15-6. Offset 00h: THR--Transmit Holding Register Description: View: IA F Size: 8 bit Offset Start: 00h Offset End: 00h Base Address: Base (IO) (DLAB = 0) Default: 00h Power Well: Core Bit Range Bit Acronym Bit Description Sticky 07 :00 TB_7_0 Data byte received (bits [7:0]), least significant bit first Bit Reset Value Bit Access 00h WO In FIFO mode, writing to THR puts data to the top of the FIFO. The data at the bottom of the FIFO is loaded to the shift register when it is empty. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 723 15.2.1.3 Offset 01h: IER--Interrupt Enable Register This register enables five types of interrupts which independently activate the int signal and set a value in the Interrupt Identification Register. Each of the five interrupt types can be disabled by resetting the appropriate bit of the IER register. Similarly, by setting the appropriate bits, selected interrupts can be enabled. Receiver time out interrupt can be configured to be separated from the receive data available interrupt (using the bit 5: COMP) The use of bit 5 to bit 4 is different from the register definition of standard 16550. Table 15-7. Offset 01h: IER--Interrupt Enable Register Description: View: IA F Size: 8 bit Default: 00h Power Well: Core Bit Reset Value Bit Access 00h RO Compatibility Enable: 0 = Bit 0 of this register also controls RTOIE and bit 4 is RSVD. 1 = Bit 4 of this register controls RTOIE. 0h RW Receiver Time Out Interrupt Enable: 0 = Receiver data Time out interrupt disabled 1 = Receiver data Time out interrupt enabled 0h RW Modem Interrupt Enable: 0 = Modem Status interrupt disabled 1 = Modem Status interrupt enabled 0h RW Receiver Line Status Interrupt Enable: 0 = Receiver Line Status interrupt disabled 1 = Receiver Line Status interrupt enabled 0h RW TIE Transmit Data request Interrupt Enable: 0 = Transmit FIFO Data Request interrupt disabled 1 = Transmit FIFO Data Request interrupt enabled 0h RW RAVIE Receiver Data Available Interrupt Enable: When BIT 5 = 1: 0 = Receiver Data Available (Trigger level reached) interrupt disabled 1 = Receiver Data Available (Trigger level reached) interrupt enabled When BIT 5 = 0: 0 = Receiver data Time Out Interrupt also disabled 1 = Receiver data Time Out Interrupt enabled 0h RW Bit Range Bit Acronym 07 :06 RSVD RSVD = 0 05 COMP 04 RTOIE 03 MIE 02 RLSE 01 00 Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 724 Offset Start: 01h Offset End: 01h Base Address: Base (IO) (DLAB = 0) Sticky October 2012 Order Number: 327879-001US 15.0 15.2.1.4 Offset 02h: IIR--Interrupt Identification Register In order to minimize software overhead during data character transfers, the UART prioritizes interrupts into four levels (listed in Table 15-8) and records these in the Interrupt Identification Register. The Interrupt Identification Register (IIR) stores information indicating that a prioritized interrupt is pending and the source of that interrupt. Table 15-8. Interrupt Conditions Priority Level 1 (highest) Interrupt Origin Receiver Line Status. One or more error bits were set. 2 Received Data is available. In FIFO mode, trigger level was reached; in non-FIFO mode, RBR has data. 2 Receiver Time out occurred. It happens in FIFO mode only, when there is data in the receive FIFO but no activity for a time period. 3 Transmitter requests data. In FIFO mode, the transmit FIFO is half or more than half empty; in non-FIFO mode, THR is read already. 4 Modem Status: one or more of the modem input signals has changed state Table 15-9. Offset 02h: IIR--Interrupt Identification Register Description: View: IA F Size: 8 bit Bit Range Base Address: Base (IO) Offset Start: 02h Offset End: 02h Default: 01h Power Well: Core Bit Acronym Bit Description FIFO 00 01 10 11 Sticky Mode Enable Status (bits [1:0]): Non-FIFO mode is selected. Reserved Reserved FIFO mode is selected (TRFIFOE = 1). Bit Reset Value Bit Access 00b RO 07 :06 FIFOES_1_0 05 :04 Reserved Reserved TOD_IID3 Time Out Detected: 0 = No time out interrupt is pending. 1 = Time out interrupt is pending. (FIFO mode only) 0b RO IID_2_1 Interrupt Source Encoded (bits[2:1]): 00 Modem Status (CTS, DSR, RI, DCD modem signals changed state) 01 Transmit FIFO requests data 10 Received Data Available 11 Receive error (Overrun, parity, framing, break, FIFO error) 00b RO 1b RO 03 02 :01 00 P_N October 2012 Order Number: 327879-001US Interrupt Pending: 0 = Interrupt is pending. (Active low) 1 = No interrupt is pending. 00b Intel(R) Communications Chipset 89xx Series - Datasheet 725 Table 15-10. Interrupt Identification Register Decode Interrupt ID bits Interrupt SET/RESET Function 3 2 1 0 Priorit y Type 0 0 0 1 - None 0 1 1 0 Highest 0 1 0 0 Second Highest RESET Control No Interrupt is pending. - Receiver Line Status Overrun Error, Parity Error, Framing Error, Break Interrupt. Reading the Line Status Register. Received Data Available. Non-FIFO mode: Receive Buffer is full. Non-FIFO mode: Reading the Receiver Buffer Register. FIFO mode: Trigger level was reached. FIFO mode: Reading bytes until Receiver FIFO drops below trigger level or setting RESETRF bit in FCR register. 1 1 0 0 Second Highest Character Timeout indication. FIFO Mode only: At least 1 character is in receiver FIFO and there was no activity for a time period. Reading the Receiver FIFO or setting RESETRF bit in FCR register. 0 0 1 0 Third Highest Transmit FIFO Data Request Non-FIFO mode: Transmit Holding Register Empty Reading the IIR Register (if the source of the interrupt) or writing into the Transmit Holding Register. FIFO mode: Transmit FIFO has half or less than half data. Reading the IIR Register (if the source of the interrupt) or writing to the Transmitter FIFO. Clear to Send, Data Set Ready, Ring Indicator, Received Line Signal Detect Reading the modem status register 0 0 0 0 Fourth Highest Modem Status Intel(R) Communications Chipset 89xx Series - Datasheet 726 Source October 2012 Order Number: 327879-001US 15.0 15.2.1.5 Offset 02h: FCR--FIFO Control Register FCR is a write-only register that is located at the same address as the IIR (IIR is a read-only register). FCR enables/disables the transmitter/receiver FIFOs, clears the transmitter/receiver FIFOs, and sets the receiver FIFO trigger level. Table 15-11. Offset 02h: FCR--FIFO Control Register Description: View: IA F Size: 8 bit Bit Range Base Address: Base (IO) Offset Start: 02h Offset End: 02h Default: 00h Power Well: Core Bit Access 00b WO Bit Description 07 :06 ITL_1_0 Interrupt Trigger Level (bits [1:0]): When the number of bytes in the receiver FIFO equals the interrupt trigger level programmed into this field and the Received Data Available Interrupt is enabled (via IER), an interrupt is generated and appropriate bits are set in the IIR. 00 1 byte or more in FIFO causes interrupt 01 RSVD 10 8 bytes or more in FIFO causes interrupt 11 RSVD 05 :03 Reserved Reserved. Must be programmed to 0. RESETTF Reset transmitter FIFO: When RESETTF is set to 1, the transmitter FIFO counter logic is set to 0, effectively clearing all the bytes in the FIFO. The TDRQ bit in LSR are set and IIR shows a transmitter requests data interrupt if the TIE bit in the IER register is set. The transmitter shift register is not cleared; it completes the current transmission. After the FIFO is cleared, RESETTF is automatically reset to 0. 0 = Writing 0 has no effect 1 = The transmitter FIFO is cleared (FIFO counter set to 0). After clearing, bit is automatically reset to 0 0b WO RESETRF Reset Receiver FIFO: When RESETRF is set to 1, the receiver FIFO counter is reset to 0, effectively clearing all the bytes in the FIFO. The DR bit in LSR is reset to 0. All the error bits in the FIFO and the FIFOE bit in LSR are cleared. Any error bits, OE, PE, FE or BI, that had been set in LSR are still set. The receiver shift register is not cleared. If IIR had been set to Received Data Available, it is cleared. After the FIFO is cleared, RESETRF is automatically reset to 0. 0 = Writing 0 has no effect 1 = The receiver FIFO is cleared (FIFO counter set to 0). After clearing, bit is automatically reset to 0 0b WO TRFIFOE Transmit and Receive FIFO Enable: TRFIFOE enables/disables the transmitter and receiver FIFOs. When TRFIFOE = 1, both FIFOs are enabled (FIFO Mode). When TRFIFOE = 0, the FIFOs are both disabled (non-FIFO Mode). Writing a 0 to this bit clears all bytes in both FIFOs. When changing from FIFO mode to nonFIFO mode and vice versa, data is automatically cleared from the FIFOs. This bit must be 1 when other bits in this register are written or the other bits are not programmed. 0 = FIFOs are disabled 1 = FIFOs are enabled 0b WO 02 01 00 October 2012 Order Number: 327879-001US Sticky Bit Reset Value Bit Acronym 000b Intel(R) Communications Chipset 89xx Series - Datasheet 727 15.2.1.6 Offset 03h: LCR--Line Control Register In the Line Control Register (LCR), the system programmer specifies the format of the asynchronous data communications exchange. The serial data format consists of a start bit (logic 0), five to eight data bits, an optional parity bit, and one or two stop bits (logic 1). The LCR has bits for accessing the Divisor Latch and causing a break condition. The programmer can also read the contents of the Line Control Register. The read capability simplifies system programming and eliminates the need for separate storage in system memory. Table 15-12. Offset 03h: LCR--Line Control Register (Sheet 1 of 2) Description: View: IA F Size: 8 bit Bit Range 07 06 Base Address: Base (IO) Offset Start: 03h Offset End: 03h Default: 00h Power Well: Core Bit Reset Value Bit Access DLAB Divisor register access bit: This bit is the Divisor Latch Access Bit. It must be set high (logic 1) to access the Divisor Latches of the Baud Rate Generator during a READ or WRITE operation. It must be set low (logic 0) to access the Receiver Buffer, the Transmit Holding Register, or the Interrupt Enable Register. 0 = Access Transmit Holding register (THR), Receive Buffer Register (RBR) and Interrupt Enable Register. 1 = Access Divisor Latch Registers (DLL and DLH). 0b RW SB Set break: This bit is the set break control bit. It causes a break condition to be transmitted to the receiving UART. When SB is set to a logic 1, the serial output (TXD) is forced to the spacing (logic 0) state and remains there until SB is set to a logic 0. This bit acts only on the TXD pin and has no effect on the transmitter logic. This feature enables the processor to alert a terminal in a computer communications system. If the following sequence is executed, no erroneous characters are transmitted because of the break: Load 00H in the Transmit Holding register in response to a TDRQ interrupt After TDRQ goes high (indicating that 00H is being shifted out), set the break bit before the parity or stop bits reach the TXD pin Wait for the transmitter to be idle (TEMT = 1) and clear the break bit when normal transmission has to be restored During the break, the transmitter can be used as a character timer to accurately establish the break duration. In FIFO mode, wait for the transmitter to be idle (TEMT=1) to set and clear the break bit. 0 = No effect on TXD output 1 = Forces TXD output to 0 (space) 0b RW Bit Acronym Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 728 Sticky October 2012 Order Number: 327879-001US 15.0 Table 15-12. Offset 03h: LCR--Line Control Register (Sheet 2 of 2) Description: View: IA F Size: 8 bit Bit Range 05 04 03 02 01 :00 Base Address: Base (IO) Offset Start: 03h Offset End: 03h Default: 00h Power Well: Core Bit Reset Value Bit Access STKYP Sticky Parity: This bit is the "sticky parity" bit, which can be used in multiprocessor communications. When PEN and STKYP are logic 1, the bit that is transmitted in the parity bit location (the bit just before the stop bit) is the complement of the EPS bit. If EPS is 0, then the bit at the parity bit location is transmitted as a 1. In the receiver, if STKYP and PEN are 1, then the receiver compares the bit that is received in the parity bit location with the complement of the EPS bit. If the values being compared are not equal, the receiver sets the Parity Error bit in LSR and causes an error interrupt if line status interrupts were enabled. For example, if EPS is 0, the receiver expects the bit received at the parity bit location to be 1. If it is not, then the parity error bit is set. By forcing the bit value at the parity bit location, rather than calculating a parity value, a system with a master transmitter and multiple receivers can identify some transmitted characters as receiver addresses and the rest of the characters as data. If PEN = 0, STKYP is ignored. 0 = No effect on parity bit 1 = Forces parity bit to be opposite of EPS bit value 0b RW EPS Even parity Select: This bit is the even parity select bit. When PEN is a logic 1 and EPS is a logic 0, an odd number of logic ones is transmitted or checked in the data word bits and the parity bit. When PEN is a logic 1 and EPS is a logic 1, an even number of logic ones is transmitted or checked in the data word bits and parity bit. If PEN = 0, EPS is ignored. 0 = Sends or checks for odd parity 1 = Sends or checks for even parity 0b RW PEN Parity enable: This is the parity enable bit. When PEN is a logic 1, a parity bit is generated (transmit data) or checked (receive data) between the last data word bit and Stop bit of the serial data. (The parity bit is used to produce an even or odd number of ones when the data word bits and the parity bit are summed.) 0 = No parity function 1 = Allows parity generation and checking 0b RW STB Stop bits: This bit specifies the number of stop bits transmitted and received in each serial character. If STB is a logic 0, one stop bit is generated in the transmitted data. If STB is a logic 1 when a 5-bit word length is selected via bits 0 and 1, then 1 and one half stop bits are generated. If STB is a logic 1 when either a 6, 7, or 8-bit word is selected, then two stop bits are generated. The receiver checks the first stop bit only, regardless of the number of stop bits selected. 0 = 1 stop bit 1 = 2 stop bits, except for 5-bit character then 1-1/2 bits 0b RW 00b RW Bit Acronym Bit Description WLS_1_0 October 2012 Order Number: 327879-001US Sticky Word Length select: The Word Length Select bits specify the number of data bits in each transmitted or received serial character. 00 5-bit character (default) 01 6-bit character 10 7-bit character 11 8-bit character Intel(R) Communications Chipset 89xx Series - Datasheet 729 15.2.1.7 Offset 04h: MCR--Modem Control Register This 8-bit register controls the interface with the modem or data set (or a peripheral device emulating a modem). Table 15-13. Offset 04h: MCR--Modem Control Register (Sheet 1 of 2) Description: View: IA F Size: 8 bit Base Address: Base (IO) Offset Start: 04h Offset End: 04h Default: 00h Power Well: Core Bit Range Bit Acronym 07 :05 Reserved 04 03 Bit Description Reserved Bit Reset Value Bit Access 000b LOOP Loop back test mode: This bit provides a local Loopback feature for diagnostic testing of the UART. When LOOP is set to a logic 1, the following occurs: The transmitter serial output is set to a logic 1 state. The OUT2# signal is forced to a logic 1 state. The receiver serial input is disconnected from the pin. The output of the Transmitter Shift register is "looped back" into the receiver shift register input. The four modem control inputs (CTS#, DSR#, DCD#, and RI#) are disconnected from the pins and the modem control output pins (RTS# and DTR#) are forced to their inactive state. * Coming out of the loopback test mode may result in unpredictable activation of the delta bits (bits 3:0) in the Modem Status Register (MSR). It is recommended that the MSR be read once to clear the delta bits in the MSR. The lower four bits of the Modem Control register are connected to the upper four Modem Status register bits: * DTR = 1 forces DSR to a 1 * RTS = 1 forces CTS to a 1 * OUT1 = 1 forces RI to a 1 * OUT2= 1 forces DCD to a 1 In the diagnostic mode, data that is transmitted is immediately received. This feature allows the processor to verify the transmit and receive data paths of the UART. The transmit, receive and modem control interrupts are operational, except the modem control interrupts are activated by Control register bits, not the modem control inputs. A break signal can also be transferred from the transmitter section to the receiver section in loopback mode. 0 = Normal UART operation 1 = Test mode UART operation 0b RW OUT2 OUT2# signal control: This bit controls the OUT2# output. When the OUT2 bit is set, OUT2# is asserted low. When the OUT2 bit is cleared, OUT2# is deasserted (set high). Outside of the UART module, the OUT2# signal is used to connect the UART's interrupt output to the Interrupt Controller unit. 0 = OUT2# signal is 1, which disables the UART interrupt. 1 = OUT2# signal is 0. 0b RW Intel(R) Communications Chipset 89xx Series - Datasheet 730 Sticky October 2012 Order Number: 327879-001US 15.0 Table 15-13. Offset 04h: MCR--Modem Control Register (Sheet 2 of 2) Description: View: IA F Size: 8 bit Bit Range 02 01 00 15.2.1.8 Base Address: Base (IO) Offset Start: 04h Offset End: 04h Default: 00h Power Well: Core Bit Reset Value Bit Access Test bit: This bit is used only in Loopback test mode. See (LOOP) Above. 0b RW RTS Request to Send: This bit controls the Request to Send (RTS#) output pin. Bit 1 affects the RTS# output in a manner identical to that described below for the DTR bit. 0 = RTS# pin is 1 1 = RTS# pin is 0 0b RW DTR Data Terminal Ready: This bit controls the Data Terminal Ready output. When bit 0 is set to a logic 1, the DTR# output is forced to a logic 0. When bit 0 is reset to a logic 0, the DTR# output pin is forced to a logic 1. * The DTR# output of the UART may be applied to an EIA inverting line driver (such as the DS1488) to obtain the proper polarity input at the succeeding modem or data set. 0 = DTR# pin is 1 1 = DTR# pin is 0 0b RW Bit Acronym OUT1 Bit Description Sticky Offset 05h: LSR--Line Status Register This register provides status information to the processor concerning the data transfers. Bits 5 and 6 show information about the transmitter section. The rest of the bits contain information about the receiver. In non-FIFO mode, three of the LSR register bits, parity error, framing error, and break interrupt, show the error status of the character that has just been received. In FIFO mode, these three bits of status are stored with each received character in the FIFO. LSR shows the status bits of the character at the top of the FIFO. When the character at the top of the FIFO has errors, the LSR error bits are set and are not cleared until software reads LSR, even if the character in the FIFO is read and a new character is now at the top of the FIFO. Bits one through four are the error conditions that produce a receiver line status interrupt when any of the corresponding conditions are detected and the interrupt is enabled. These bits are not cleared by reading the erroneous byte from the FIFO or receive buffer. They are cleared only by reading LSR. In FIFO mode, the line status interrupt occurs only when the erroneous byte is at the top of the FIFO. If the erroneous byte being received is not at the top of the FIFO, an interrupt is generated only after the previous bytes are read and the erroneous byte is moved to the top of the FIFO. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 731 Table 15-14. Offset 05h: LSR--Line Status Register (Sheet 1 of 3) Description: View: IA F Size: 8 bit Bit Range 07 06 05 Base Address: Base (IO) Offset Start: 05h Offset End: 05h Default: 60h Power Well: Core Bit Reset Value Bit Access FIFOE FIFO Error Status: This bit is reset only when all the error bytes have been read from the FIFO. A processor read to the Line Status register does not reset this bit. Non-FIFO mode: 0 = Bit is always "0" indicating no FIFO. FIFO mode: 0 = All error bytes have been read from the FIFO 1 = At least one character in the receiver FIFO contains a parity error, framing error, or break indication. When DMA requests are enabled (IER bit7 is set to 1) and FIFOE is set to 1, no receive DMA request is generated even though the receive FIFO reaches the trigger level and the error interrupt is generated. When DMA requests are not enabled (IER bit7 is set to 0), FIFOE set to 1 does not generate an interrupt. 0b RO TEMT Transmitter Empty: Non-FIFO mode: 0 = Either the Transmit Holding register or the Tansmitter Shift register contain a data character. 1 = The Transmit Holding register and the Transmitter Shift register are both empty. FIFO mode: 0 = Either the transmitter FIFO or the Transmit Shift register contain a data character. 1 = The transmitter FIFO and the Transmit Shift register are both empty. 1b RO TRDQ Transmit Data Request: TDRQ indicates that the UART is ready to accept a new character for transmission. In addition, this bit causes the UART to issue an interrupt to the processor when the transmit data request interrupt enable is set high. Non-FIFO mode: 0 = No character transferred from the Transmit Holding register into the Transmit Shift register. 1 = A character has transferred from the Transmit Holding register into the Transmit Shift register. Note: Bit is reset to logic 0 with the loading of the Transmit Holding register by the processor. FIFO mode: 0 = When at least one byte is written to the transmit FIFO. When more than 16 characters are loaded into the FIFO, the excess characters are lost. 1 = Transmit FIFO is empty or the RESETTF bit in FCR, has been set to 1. 1b RO Bit Acronym Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 732 Sticky October 2012 Order Number: 327879-001US 15.0 Table 15-14. Offset 05h: LSR--Line Status Register (Sheet 2 of 3) Description: View: IA F Size: 8 bit Bit Range 04 03 02 Base Address: Base (IO) Offset Start: 05h Offset End: 05h Default: 60h Power Well: Core Bit Reset Value Bit Access BI Break Interrupt: BI is set to a logic 1 when the received data input is held in the spacing (logic 0) state for longer than a full word transmission time (that is, the total time of Start bit + data bits + parity bit + stop bits).The BI is reset to a logic "0" when the processor reads the Line Status register. 0 = No Break signal has been received. 1 = Break signal occurred. In FIFO mode, only one character (equal to 00H), is loaded into the FIFO regardless of the length of the break condition. BI shows the break condition for the character at the top of the FIFO, not the most recently received character. 0h RO FE Framing Error: FE indicates that the received character did not have a valid stop bit. This bit is reset to a logic "0" when the processor reads the Line Status Register. 0 = No Framing error. 1 = Invalid stop bit has been detected. FE is set to a logic 1 when the bit following the last data bit or parity bit is detected as a logic 0 (spacing level). If the Line Control register had been set for two stop bit mode, the receiver does not check for a valid second stop bit. The FE indicator is reset when the processor reads the Line Status Register. The UART resynchronizes after a framing error. To do this it assumes that the framing error was due to the next start bit, so it samples this "start" bit twice and then takes in the "data". In FIFO mode FE shows a Framing error for the character at the top of the FIFO, not for the most recently received character. 0h RO PE Parity Error: PE indicates that the received data character does not have the correct even or odd parity, as selected by the even parity select bit. The PE is set to logic 1 upon detection of a parity error and is reset to a logic 0 when the processor reads the Line Status register. In FIFO mode, PE shows a parity error for the character at the top of the FIFO, not the most recently received character. 0 = No Parity error. 1 = Parity error has occurred. 0h RO Bit Acronym Bit Description October 2012 Order Number: 327879-001US Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 733 Table 15-14. Offset 05h: LSR--Line Status Register (Sheet 3 of 3) Description: View: IA F Size: 8 bit Bit Range 01 00 Base Address: Base (IO) Offset Start: 05h Offset End: 05h Default: 60h Power Well: Core Bit Reset Value Bit Access OE Overrun Error: In non-FIFO mode, OE indicates that data in the receiver buffer register was not read by the processor before the next character was transferred into the receiver buffer register, thereby destroying the previous character. In FIFO mode, OE indicates that all 16 bytes of the FIFO are full and the most recently received byte has been discarded. The OE indicator is set to a logic "1" upon detection of an overrun condition and reset when the processor reads the Line Status Register. 0 = No data has been lost 1 = Received data has been lost. 0h RO DR Data Ready: DR is set to logic 1 when a complete incoming character has been received and transferred into the Receiver Buffer Register (RBR) or the FIFO. In non-FIFO mode, DR is reset to 0 when the receive buffer is read. In FIFO mode, DR is reset to a logic 0 if the FIFO is empty (last character has been read from Receiver Buffer Register) or the RESETRF bit is set in FCR. 0 = No data has been received 1 = Data available in RBR or the FIFO. 0h RO Bit Acronym Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 734 Sticky October 2012 Order Number: 327879-001US 15.0 15.2.1.9 Offset 06h: MSR--Modem Status Register This 8-bit register provides the current state of the control lines from the modem or data set (or a peripheral device emulating a modem) to the processor. In addition to this current state information, four bits of the Modem Status register provide change information. Bits 03:00 are set to a logic 1 when a control input from the Modem changes state. They are reset to a logic 0 when the processor reads the Modem Status register. When bits 0, 1, 2, or 3 are set to logic 1, a Modem Status interrupt is generated if bit 3 of the Interrupt Enable Register is set. Table 15-15. Offset 06h: MSR--Modem Status Register Description: View: IA F Size: 8 bit Base Address: Base (IO) Offset Start: 06h Offset End: 06h Default: 00h Power Well: Core Bit Reset Value Bit Access DCD Data Carrier Detect: This bit is the complement of the Data Carrier Detect (DCD#) input. This bit is equivalent to bit OUT2 of the Modem Control register if LOOP in the MCR is set to 1. 0 = DCD# pin is 1 1 = DCD# pin is 0 0b RO RI Ring Indicator: This bit is the complement of the ring Indicator (RI#) input. This bit is equivalent to bit OUT1 of the Modem Control register if LOOP in the MCR is set to 1. 0 = RI# pin is 1 1 = RI# pin is 0 0b RO DSR Data Set Ready: This bit is the complement of the Data Set Ready (DSR#) input. This bit is equivalent to bit DTR of the Modem Control register if LOOP in the MCR is set to 1. 0 = DSR# pin is 1 1 = DSR# pin is 0 0b RO 04 CTS Clear to Send: This bit is the complement of the Clear to Send (CTS#) input. This bit is equivalent to bit RTS of the Modem Control register if LOOP in the MCR is set to 1. 0 = CTS# pin is 1 1 = CTS# pin is 0 0b RO 03 DDCD Delta Data Carrier Detect: 0 = No change in DCD# pin since last read of the MSR 1 = DCD# pin has changed state 0b RO 02 TERI Trailing Edge Ring Indicator: 0 = RI# pin has not changed from 0 to 1 since last read of the MSR 1 = RI# pin has changed from 0 to 1 0b RO 01 DDSR Delta Data Set Ready: 0 = No change in DSR# pin since last read of the MSR 1 = DSR# pin has changed state 0b RO 00 DCTS Delta Clear To Send: 0 = No change in CTS# pin since last read of the MSR 1 = CTS# pin has changed state 0b RO Bit Range 07 06 05 Bit Acronym Bit Description October 2012 Order Number: 327879-001US Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 735 15.2.1.10 Offset 07h: SCR--Scratchpad Register This 8-bit read/write register has no effect on the UART. It is intended as a scratchpad register for use by the programmer. Table 15-16. Offset 07h: SCR--Scratchpad Register Description: View: IA F Size: 8 bit Base Address: Base (IO) Offset Start: 07h Offset End: 07h Default: 00h Power Well: Core Bit Range Bit Acronym 07 :00 SP_7_0 15.2.1.11 Bit Description Sticky No effect on UART functionality Bit Reset Value Bit Access 00h RW Offset 00h: DLL--Programmable Baud Rate Generator Divisor Latch Register Low See Section 4.13.4.3, "Programmable Baud Rate Generator" on page 186. Table 15-17. Offset 00h: DLL--Programmable Baud Rate Generator Divisor Latch Register Low Description: View: IA F Size: 8 bit Default: 02h Bit Range Bit Acronym 07 :00 BR_7_0 Power Well: Core Bit Description Low byte compare value to generate baud rate Intel(R) Communications Chipset 89xx Series - Datasheet 736 Offset Start: 00h Offset End: 00h Base Address: Base (IO) (DLAB = 1) Sticky Bit Reset Value Bit Access 02h RW October 2012 Order Number: 327879-001US 15.0 15.2.1.12 Offset 01h: DLH--Programmable Baud Rate Generator Divisor Latch Register High See Section 4.13.4.3, "Programmable Baud Rate Generator" on page 186. Table 15-18. Offset 01h: DLH--Programmable Baud Rate Generator Divisor Latch Register High Description: View: IA F Size: 8 bit Default: 00h Bit Range Bit Acronym 07 :00 BR_15_8 15.3 Offset Start: 01h Offset End: 01h Base Address: Base (IO) (DLAB = 1) Power Well: Core Bit Description Sticky High byte compare value to generate baud rate Bit Reset Value Bit Access 00h RW Logical Device 6: Watchdog Timer The Watchdog timer provides a resolution that ranges from 1 s to 10 minutes. The timer uses a 35-bit down-counter. The counter is loaded with the value from the first Preload register. The timer is then enabled and it starts counting down. The time at which the WDT first starts counting down is called the first stage. If the host fails to reload the WDT before the 35-bit down counter reaches zero the WDT generates an internal interrupt. After the interrupt is generated the WDT loads the value from the second Preload register into the WDT's 35-bit Down-Counter and starts counting down. The WDT is now in the second stage. If the host still fails to reload the WDT before the second timeout, the WDT drives the WDT_TOUT# pin low and sets the timeout bit (WDT_TIMEOUT). This bit indicates that the System has become unstable. The WDT_TOUT# pin is held low until the system is Reset or the WDT times out again (Depends on TOUT_CNF). The process of reloading the WDT involves the following sequence of writes: 1. Write "80" to offset BAR1 + 0Ch 2. Write "86" to offset BAR1 + 0Ch 3. Write `1' to WDT_RELOAD in Reload Register. The same process is used for setting the values in the preload registers. The only difference exists in step 3. Instead of writing a `1' to the WDT_RELOAD, you write the desired preload value into the corresponding Preload register. This value is not loaded into the 35-bit down counter until the next time the WDT reenters the stage. For example, if Preload Value 2 is changed, it is not loaded into the 35-bit down counter until the next time the WDT enters the second stage. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 737 15.3.1 Watchdog Timer Register Details All registers not mentioned are reserved. Table 15-19. Summary of Watchdog Timer Registers in I/O Space Offset Start Offset End Default Value Register ID - Description 00h 00h "Offset 00h: PV1R0--Preload Value 1 Register 0" on page 738 FFh 01h 01h "Offset 01h: PV1R1--Preload Value 1 Register 1" on page 739 FFh 02h 02h "Offset 02h: PV1R2--Preload Value 1 Register 2" on page 739 0Fh 04h 04h "Offset 04h: PV2R0--Preload Value 2 Register 0" on page 740 FFh 05h 05h "Offset 05h: PV2R1--Preload Value 2 Register 1" on page 740 FFh 06h 06h "Offset 06h: PV2R2--Preload Value 2 Register 2" on page 741 0Fh 08h 08h "Offset 08h: GISR--General Interrupt Status Register" on page 741 00h 0Ch 0Ch "Offset 0Ch: RR0--Reload Register 0" on page 742 00h 0Dh 0Dh "Offset 0Dh: RR1--Reload Register 1" on page 742 00h 10h 10h "Offset 10h: WDTCR--WDT Configuration Register" on page 743 00h 18h 18h "Offset 18h: WDTLR--WDT Lock Register" on page 744 00h Note: Base Address for the Watchdog Timer registers, listed in this section, is configurable. See Section 15.4.1, "SIW Configuration Register Summary" on page 745 for details. 15.3.1.1 Offset 00h: PV1R0--Preload Value 1 Register 0 Table 15-20. Offset 00h: PV1R0--Preload Value 1 Register 0 Description: View: IA F Size: 8 bit Bit Range 07 : 00 Base Address: Base (IO) Offset Start: 00h Offset End: 00h Default: FFh Power Well: Core Bit Acronym PLOAD1_7_0 Bit Description Preload_Value_1 [7:0]: This register is used to hold the bits 0 through 7 of the preload value 1 for the WDT Timer. The Value in the Preload Register is automatically transferred into the 35-bit down counter every time the WDT enters the first stage. The value loaded into the preload register needs to be one less than the intended period. This is because the timer makes use of zero-based counting (i.e. zero is counted as part of the decrement). Intel(R) Communications Chipset 89xx Series - Datasheet 738 Sticky Bit Reset Value Bit Access FFh RW October 2012 Order Number: 327879-001US 15.0 15.3.1.2 Offset 01h: PV1R1--Preload Value 1 Register 1 Table 15-21. Offset 01h: PV1R1--Preload Value 1 Register 1 Description: View: IA F Size: 8 bit Bit Range 07 : 00 15.3.1.3 Base Address: Base (IO) Offset Start: 01h Offset End: 01h Default: FFh Power Well: Core Bit Acronym Bit Description Sticky Preload_Value_1 [15:8]: This register is used to hold the bits 8 through 15 of the preload value 1 for the WDT Timer. The Value in the Preload Register is automatically transferred into the 35-bit down counter every time the PLOAD1_15_8 WDT enters the first stage. The value loaded into the preload register needs to be one less than the intended period. This is because the timer makes use of zero-based counting (i.e. zero is counted as part of the decrement). Bit Reset Value Bit Access FFh RW Offset 02h: PV1R2--Preload Value 1 Register 2 Table 15-22. Offset 02h: PV1R2--Preload Value 1 Register 2 Description: View: IA F Size: 8 bit Base Address: Base (IO) Offset Start: 02h Offset End: 02h Default: 0Fh Power Well: Core Bit Range Bit Acronym 07 : 04 Reserved 03 : 00 Bit Description Sticky Reserved Bit Access 0h Preload_Value_1 [19:16]: This register is used to hold the bits 16 through 19 of the preload value 1 for the WDT Timer. The Value in the Preload Register is automatically transferred into the 35-bit down counter every time the WDT enters the first stage. PLOAD_19_16 The value loaded into the preload register needs to be one less than the intended period. This is because the timer makes use of zero-based counting (i.e. zero is counted as part of the decrement). October 2012 Order Number: 327879-001US Bit Reset Value Fh RW Intel(R) Communications Chipset 89xx Series - Datasheet 739 15.3.1.4 Offset 04h: PV2R0--Preload Value 2 Register 0 Table 15-23. Offset 04h: PV2R0--Preload Value 2 Register 0 Description: View: IA F Size: 8 bit Bit Range 07 : 00 15.3.1.5 Base Address: Base (IO) Offset Start: 04h Offset End: 04h Default: FFh Power Well: Core Bit Acronym PLOAD2_7_0 Bit Description Sticky Preload_Value_2 [7:0]: This register is used to hold the bits 0 through 7 of the preload value 2 for the WDT Timer. The Value in the Preload Register is automatically transferred into the 35-bit down counter every time the WDT enters the second stage. The value loaded into the preload register needs to be one less than the intended period. This is because the timer makes use of zero-based counting (i.e., zero is counted as part of the decrement). Bit Reset Value Bit Access FFh RW Offset 05h: PV2R1--Preload Value 2 Register 1 Table 15-24. Offset 05h: PV2R1--Preload Value 2 Register 1 Description: View: IA F Size: 8 bit Bit Range 07 : 00 Base Address: Base (IO) Offset Start: 05h Offset End: 05h Default: FFh Power Well: Core Bit Acronym Bit Description Preload_Value_2 [15:8]: This register is used to hold the bits 8 through 15 of the preload value 2 for the WDT Timer. The Value in the Preload Register is automatically transferred into the 35-bit down counter every time the WDT enters the second stage. PLOAD2_15_8 The value loaded into the preload register needs to be one less than the intended period. This is because the timer makes use of zero-based counting (i.e., zero is counted as part of the decrement). Intel(R) Communications Chipset 89xx Series - Datasheet 740 Sticky Bit Reset Value Bit Access FFh RW October 2012 Order Number: 327879-001US 15.0 15.3.1.6 Offset 06h: PV2R2--Preload Value 2 Register 2 Table 15-25. Offset 06h: PV2R2--Preload Value 2 Register 2 Description: View: IA F Size: 8 bit Base Address: Base (IO) Offset Start: 06h Offset End: 06h Default: 0Fh Power Well: Core Bit Range Bit Acronym 07 : 04 Reserved 03 : 00 15.3.1.7 Bit Description Sticky Reserved Bit Reset Value Bit Access 0h Preload_Value_2 [19:16]: This register is used to hold the bits 16 through 19 of the preload value 2 for the WDT Timer. The Value in the Preload Register is automatically transferred into the 35-bit down counter every time the WDT enters the second stage. PLOAD2_19_16 The value loaded into the preload register needs to be one less than the intended period. This is because the timer makes use of zero-based counting (i.e. zero is counted as part of the decrement). Fh RW Offset 08h: GISR--General Interrupt Status Register Table 15-26. Offset 08h: GISR--General Interrupt Status Register Description: View: IA F Size: 8 bit Base Address: Base (IO) Offset Start: 08h Offset End: 08h Default: 00h Power Well: Core Bit Range Bit Acronym 07 : 01 Reserved 00 SERIRQACT October 2012 Order Number: 327879-001US Bit Description Sticky Reserved Bit Reset Value Bit Access 00h Watchdog Timer SERIRQ Interrupt Active (1st Stage): This bit is set when the first Stage of the 35-bit Down Counter Reaches zero. An SERIRQ interrupt is generated if WDT_INT_TYPE is configured to do so (See WDT Configuration Register). This is a sticky bit and is only cleared by writing a `1'. 0 = No Interrupt 1 = Interrupt Active Note: This bit is not set in free running mode. 0h RWC Intel(R) Communications Chipset 89xx Series - Datasheet 741 15.3.1.8 Offset 0Ch: RR0--Reload Register 0 Table 15-27. Offset 0Ch: RR0--Reload Register 0 Description: View: IA F Size: 8 bit Base Address: Base (IO) Offset Start: 0Ch Offset End: 0Ch Default: 00h Power Well: Core Bit Range Bit Acronym 07 : 00 Reserved 15.3.1.9 Bit Description Sticky Reserved. Must be programmed to 0. Bit Reset Value Bit Access 00h Offset 0Dh: RR1--Reload Register 1 Table 15-28. Offset 0Dh: RR1--Reload Register 1 Description: View: IA F Size: 8 bit Base Address: Base (IO) Offset Start: 0Dh Offset End: 0Dh Default: 00h Power Well: Core Bit Range Bit Acronym 07 : 02 Reserved 01 TOUT 00 RELOAD Bit Description Reserved WDT_TIMEOUT: This bit is located in the RTC Well and it's value is not lost if the host resets the system. It is set to `1' if the host fails to reset the WDT before the 35-bit Down-Counter reaches zero for the second time in a row. This bit is cleared by performing the Register Unlocking Sequence followed by a `1' to this bit. 0 = Normal (Default) 1 = System has become unstable. Note: In free running mode this bit is set every time the down counter reaches zero. WDT_RELOAD: To prevent a timeout the host must perform the Register Unlocking Sequence followed by a `1' to this bit. Intel(R) Communications Chipset 89xx Series - Datasheet 742 Sticky Bit Reset Value Bit Access 00h 0h RW 0h RW October 2012 Order Number: 327879-001US 15.0 15.3.1.10 Offset 10h: WDTCR--WDT Configuration Register Table 15-29. Offset 10h: WDTCR--WDT Configuration Register Description: View: IA F Size: 8 bit Base Address: Base (IO) Offset Start: 10h Offset End: 10h Default: 00h Power Well: Core Bit Range Bit Acronym 07 : 06 Reserved 05 04 : 03 02 01 : 00 Bit Description Sticky Reserved Bit Access 00h WDT Timeout Output Enable: This bit indicates whether or not the WDT toggles the external WDT_TOUT# pin if WDT_TOUT_EN the WDT times out. 0 = Enabled (Default) 1 = Disabled Reserved Bit Reset Value Reserved 0h RW 00h WDT Prescaler Select: The WDT provides two options for prescaling the main Down Counter. The preload values are loaded into the main down counter right justified. The prescaler adjusts the starting point of the 35-bit down counter. 0 = The 20-bit Preload Value is loaded into bits 34:15 of the main down counter. The resulting timer clock is WDT_PRE_SEL the PCI Clock (33 MHz) divided by 215. The approximate clock generated is 1 KHz, (1 ms to 10 min). (Default) 1 = The 20-bit Preload Value is loaded into bits 24:05 of the main down counter. The resulting timer clock is the PCI Clock (33 MHz) divided by 25. The approximate clock generated is 1 MHz, (1 s to 1sec) 0h RW WDT_INT_TYPE: The WDT timer supports programmable routing of interrupts. The set of bits allows the user to choose the type of interrupt desired if the WDT reached the end of the first stage without being reset. The interrupt status is reported in the WDT General Interrupt Status register. WDT_INT_TYPE 00 SERIRQ (Default) 01 reserved 10 reserved 11 Disabled Note: SERIRQ is Active Low 00h RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 743 15.3.1.11 Offset 18h: WDTLR--WDT Lock Register Table 15-30. Offset 18h: WDTLR--WDT Lock Register Description: View: IA F Size: 8 bit Base Address: Base (IO) Offset Start: 18h Offset End: 18h Default: 00h Power Well: Core Bit Range Bit Acronym 07 : 03 Reserved Bit Description Reserved Bit Reset Value Bit Access 0h 02 WDT Timeout Configuration: This register is used to choose the functionality of the timer. 0 = Watchdog Timer Mode: When enabled (i.e. WDT_ENABLE goes from `0' to `1') the timer reloads Preload Value 1 and start decrementing. (Default) Upon reaching the second stage timeout the WDT_TOUT# is driven low once and does not change again until Power is cycled or a hard reset occurs. WDT_TOUT_CN 1 = Free Running Mode: WDT_TOUT# changes from F previous state when the next timeout occurs. The timer ignores the first stage. The timer only uses Preload Value 2. In this mode the timer is restarted whenever WDT_ENABLE goes from a 0 to a 1. This means that the timer reloads Preload Value 2 and start decrementing every time it is enabled. In free running mode it is not necessary to reload the timer as it is done automatically every time the descrementer reaches zero. 0h RW 01 Watchdog Timer Enable: The following bit enables or disables the WDT. 0 = Disabled (Default) 1 = Enabled Note: This bit cannot be modified if WDT_LOCK has been set. Note: In free-running mode Preload Value 2 is reloaded into the down counter every time WDT_ENABLE goes from `0' to `1'. WDT_ENABLE Note: In WDT mode Preload Value 1 is reloaded every time WDT_ENABLE goes from `0' to `1' or the WDT_RELOAD bit is written using the proper sequence of writes (See Register Unlocking Sequence). When the WDT second stage timeout occurs, a reset must happen. Note: Software must guarantee that a timeout is not about to occur before disabling the timer. A reload sequence is suggested. 0h RW 0h RWL 00 WDT_LOCK Watchdog Timer Lock: Setting this bit locks the values of this register until a hard-reset occurs or power is cycled. 0 = Unlocked (Default) 1 = Locked Note: Writing a "0" has no effect on this bit. Write is only allowed from "0" to "1" once. It cannot be changed until either power is cycled or a hardreset occurs. Intel(R) Communications Chipset 89xx Series - Datasheet 744 Sticky October 2012 Order Number: 327879-001US 15.0 15.4 SIW Configuration The configuration of the SIW is flexible and is based on the configuration architecture implemented in typical Plug-and-Play components. The SIW is designed for motherboard applications in which the resources required by their components are known. With its flexible resource allocation architecture, the SIW allows the BIOS to assign resources at POST. 15.4.1 SIW Configuration Register Summary The default values are defined with an h for hex, a bi for binary, or 00 for zero. If no letter follows the default value, assume it is a binary number. Warning: Address locations that are not listed are considered reserved register locations. Reads to reserved registers may return non-zero values. Writes to reserved locations may cause system failure. Reserved bits are Read Only. Table 15-31. Configuration Register Summary (Sheet 1 of 2) Global Configuration Registers Index Type Default Configuration Register 07h RW 00h Logical Device Number 20h R 00h Device ID 21h R 01h Device Rev 28h RW 01h SIW I/F (wait states) 29h RW 02h SIRQ Configuration 2Eh RW 00h Test Mode Configuration Register Logical Device 4 Registers (Serial Port 1) 30h RW 00h Enable 60h RW 00h Base I/O Address MSB 61h RW 00h Base I/O Address LSB 70h RW 00h Primary Interrupt Select 74h RW 04h RSVD 75h RW 04h RSVD F0h RW 00h RSVD Logical Device 5 Registers (Serial Port 2) 30h RW 00h Enable 60h RW 00h Base I/O Address MSB 61h RW 00h Base I/O Address LSB 70h RW 00h Primary Interrupt Select 74h RW 04h RSVD 75h RW 04h RSVD F0h RW 00h RSVD October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 745 Table 15-31. Configuration Register Summary (Sheet 2 of 2) Logical Device 6 Registers (Watchdog Timer) 15.4.1.1 30h RW 00h Enable 60h RW 00h Base I/O Address MSB 61h RW 00h Base I/O Address LSB 70h RW 00h Primary Interrupt Select Global Control/Configuration Registers [00h - 2Fh] The chip-level (global) registers lie in the address range [00h-2Fh]. The design MUST use all eight bits of the ADDRESS Port for register selection. All unimplemented registers and bits ignore writes and return zero when read. The INDEX PORT is used to select a configuration register in the chip. The DATA PORT is then used to access the selected register. These registers are accessible only in the Configuration Mode. Register Logical Device # Default = 00h Address (Type) 07h (RW) Description Logical Device Select: A write to this register selects the current logical device. This allows access to the control and configuration registers for each logical device. Device ID Default = 00h 20h (R) Device ID: A read only register which provides the Device ID. Device Rev Default = 01h 21h (R) Device Rev: A read only register which provides device revision information. SIW Interface Default = 01h 28h Bit 7:1 RSVD = 0 Bit 0 LPC bus wait states 0 = Not Supported 1 = Long wait states (sync 6) SIW Configuration Default = 02h 29h (RW - bit 0, 2, 3) (R - bit 1) Bit 0 SIRQ enable 1 =enabled; enabled logical devices participate in interrupt generation 0 =disabled; serial interrupts disabled Bit 1 IRQ mode (READ ONLY, WRITES IGNORED) 1 =Continuous mode 0 =Quiet mode Bit 3:2 UART_CLK predivide 00 = divide by 1 01 = divide by 8 10 = divide by 26 11 = reserved Bit 7:4 RSVD = 0 Intel(R) Communications Chipset 89xx Series - Datasheet 746 October 2012 Order Number: 327879-001US 15.0 Register SIW Monitor Port Control Register Default = 00h Address (Type) Description 2D (RO - bit 7) (RW - bits 6:0) Bit 0 UART1 Monitor Port Enable (UART1_MONPORTEN): Setting this bit enables the monitor port for UART1. This signal turns on all 8 UART2 monitor ports. Note: wdt_monporten, uart2_monporten and uart1_monporten must be set in a mutually exclusive manner i.e., only one monitor port enable must be set at one time. Bit 1 UART2 Monitor Port Enable (UART2_MONPORTEN): Setting this bit enables the monitor port for UART2. This signal turns on all 8 UART1 monitor ports. Note: wdt_monporten, uart2_monporten and uart1_monporten must be set in a mutually exclusive manner i.e., only one monitor port enable must be set at one time. Bit 2 WDT Monitor Port Enable (WDT_MONPORTEN): Setting this bit enables the monitor port for WDT. This signal turns on all 16 SIW monitor ports. Note: wdt_monporten, uart2_monporten and uart1_monporten must be set in a mutually exclusive manner i.e., only one monitor port enable must be set at one time. Bit 6:3 Monitor Port Slot Select[3:0] (MONPORTSEL): These bits select which Monitor Port Slot is enabled. These bits are used to select up to 16 slots within each UART1 and UART2 source group. Note: WDT has only 1 slot. So, port slot selection is not required. Bit 7 RSVD = 0 Default = 00h 15.4.1.2 2Eh Reserved Logical Device Configuration Registers [30h -- FFh] Used to access the registers that are assigned to each logical unit. This chip supports three logical units and has three sets of logical device registers. The three logical devices are UART1, UART2 and Watchdog Timer. A separate set (bank) of control and configuration registers exists for each logical device and is selected with the Logical Device # Register. The INDEX PORT is used to select a specific logical device register. These registers are then accessed through the DATA PORT. The Logical Device registers are accessible only when the device is in the Configuration State. The logical register addresses are shown in the tables below. The following pseudo code shows how to select and set the LPC Generic I/O Range = to the Watch Dog Timer (WDT) I/O Base Address. Assume LPC Generic I/O Range#4, size=32B, LPC_Address=6000h 1. CFG WR, BDF= 0,31,0, Offset= 90h = 001C6001 LPC Generic I/O Range#s are located at Offset= 84h -- LPC Generic I/O Range#1 Offset= 88h -- LPC Generic I/O Range#2 Offset= 8Ch -- LPC Generic I/O Range#3 Offset= 90h -- LPC Generic I/O Range#4 Enable decoding of I/O locations 4Eh and 4Fh to LPC. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 747 1. CFG WR, BDF= 0,31,0, Offset= 82h, bit 13 Microcontroller Enable #2 (ME2): Enter configuration mode 1. I/O WR, addr = 4Eh, data_len = 1, be_first = 4, be_last = 0, data = 00800000h 2. I/O WR, addr = 4Eh, data_len = 1, be_first = 4, be_last = 0, data = 00860000h Select logical device 6 1. I/O WR, addr = 4Eh, data_len = 1, be_first = 4, be_last = 0, data = 00070000h 2. I/O WR, addr = 4Fh, data_len = 1, be_first = 8, be_last = 0, data = 06000000h Set WDT I/O Base Address = selected LPC Generic I/O Rage LPC_Address 1. I/O WR, addr = 4Eh, data_len = 1, be_first = 4, be_last = 0, data = 00600000h 2. I/O WR, addr = 4Fh, data_len = 1, be_first = 8, be_last = 0, data = 60000000h 3. I/O WR, addr = 4Eh, data_len = 1, be_first = 4, be_last = 0, data = 00610000h 4. I/O WR, addr = 4Fh, data_len = 1, be_first = 8, be_last = 0, data = 00000000h Set WDT enable 1. I/O WR, addr = 4Eh, data_len = 1, be_first = 4, be_last = 0, data = 00300000h 2. I/O WR, addr = 4Fh, data_len = 1, be_first = 8, be_last = 0, data = 01000000h Exit configuration mode 1. I/O WR, addr = 4Eh, data_len = 1, be_first = 4, be_last = 0, data = 00680000h 2. I/O WR, addr = 4Eh, data_len = 1, be_first = 4, be_last = 0, data = 00080000h Table 15-32. Logical Device 4 (Serial Port 1) (Sheet 1 of 2) Logical Device Register Enable Default = 00h I/O Base Address Default = 00h Address 30h (RW) 60h (RW) 61h (Bits 7:3 RW Bits 2:0 RO) Intel(R) Communications Chipset 89xx Series - Datasheet 748 Description Bits[7:1] Reserved, set to zero. Bit[0] 1 = enable the logical device currently selected through the Logical Device # register. 0 = Logical device currently selected is inactive Registers 60h (MSB) and 61h (LSB) set the base address for the device. Note: Decode is on 8 Byte boundaries Comm Decode Ranges 3F8 - 3FF (COM 1) 2F8 - 2FF (COM 2) 220 - 227 228 - 22F 238 - 23F 2E8 - 2EF (COM 4) 338 - 33F 3E8 - 3EF (COM 3) October 2012 Order Number: 327879-001US 15.0 Table 15-32. Logical Device 4 (Serial Port 1) (Sheet 2 of 2) Logical Device Register Primary Interrupt Select Default = 00h Reserved Reserved Reserved Address Description 70h (RW) Bits[3:0] selects which interrupt level is used for the primary Interrupt. 00= no interrupt selected 01= IRQ1 02= IRQ2 03= IRQ3 04= IRQ4 05= IRQ5 06= IRQ6 07= IRQ7 08= IRQ8 09= IRQ9 0A= IRQ10 0B= IRQ11 0C= IRQ12 0D= IRQ13 0E= IRQ14 0F= IRQ15 Bits[7:4] Reserved Note: An Interrupt is activated by enabling this device (offset 30h), setting this register to a non-zero value and setting any combination of bits 0-4 in the corresponding UART IER and the occurrence of the corresponding UART event (i.e. Modem Status Change, Receiver Line Error Condition, Transmit Data Request, Receiver Data Available or Receiver Time Out) and setting the OUT2 bit in the MCR. 74h 75h F0h Bit 7:0 - Reserved Bit 7:0 - Reserved Bit 7:0 - Reserved Table 15-33. Logical Device 5 (Serial Port 2) (Sheet 1 of 2) Logical Device Register Enable Default = 00h I/O Base Address Default = 00h October 2012 Order Number: 327879-001US Address 30h (RW) 60h (RW) 61h (Bits 7:3 RW Bits 2:0 RO) Description Bits[7:1] Reserved, set to zero. Bit[0] 1 = enable the logical device currently selected through the Logical Device # register. 0 = Logical device currently selected is inactive Registers 60h (MSB) and 61h (LSB) set the base address for the device. Note: Decode is on 8 Byte boundaries Comm Decode Ranges 3F8 - 3FF (COM 1) 2F8 - 2FF (COM 2) 220 - 227 228 - 22F 238 - 23F 2E8 - 2EF (COM 4) 338 - 33F 3E8 - 3EF (COM 3) Intel(R) Communications Chipset 89xx Series - Datasheet 749 Table 15-33. Logical Device 5 (Serial Port 2) (Sheet 2 of 2) Logical Device Register Primary Interrupt Select Default = 00h Address Description 70h (RW) Bits[3:0] selects which interrupt level is used for the primary Interrupt. 00= no interrupt selected 01= IRQ1 02= IRQ2 03= IRQ3 04= IRQ4 05= IRQ5 06= IRQ6 07= IRQ7 08= IRQ8 09= IRQ9 0A= IRQ10 0B= IRQ11 0C= IRQ12 0D= IRQ13 0E= IRQ14 0F= IRQ15 Bits[7:4] Reserved Note: An Interrupt is activated by enabling this device (offset 30h), setting this register to a non-zero value and setting any combination of bits 0-4 in the corresponding UART IER and the occurrence of the corresponding UART event (i.e. Modem Status Change, Receiver Line Error Condition, Transmit Data Request, Receiver Data Available or Receiver Time Out) and setting the OUT2 bit in the MCR. Reserved 74h Bit 7:0 - Reserved Reserved 75h Bit 7:0 - Reserved Reserved F0h Bit 7:0 - Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 750 October 2012 Order Number: 327879-001US 15.0 Table 15-34. Logical Device 6 (Watch Dog Timer) Logical Device Register Enable Default = 00h I/O Base Address Default = 00h Primary Interrupt Select Default = 00h Address 30h (RW) Description Bits[7:1] Reserved, set to zero. Bit[0] 1 = enable the logical device currently selected through the Logical Device # register. 0 = Logical device currently selected is inactive 60h (RW) 61h (Bits 7:5 RW Bits 4:0 RO) Registers 60h (MSB) and 61h (LSB) set the base address for the device. Note: Decode is on 32 Byte boundaries. This Base Address must be within the 128 bytes of LG1 Base register. Also the last byte accessed by the WDT must not exceed the LG1 Base Address +128 bytes. See Section 15.4.1.2, "Logical Device Configuration Registers [30h -- FFh]" for example on how to program. 70h (RW) Bits[3:0] selects which interrupt level is used for the primary Interrupt. 00= no interrupt selected 01= IRQ1 02= IRQ2 03= IRQ3 04= IRQ4 05= IRQ5 06= IRQ6 07= IRQ7 08= IRQ8 09= IRQ9 0A= IRQ10 0B= IRQ11 0C= IRQ12 0D= IRQ13 0E= IRQ14 0F= IRQ15 Bits[7:4] Reserved Note: An Interrupt is activated by enabling this device (offset 30h), setting this register to a non-zero value and when the first stage has been allowed to reach zero. An Interrupt is not generated if WDT_TOUT_CNF is set to change output after every timeout (See WDT Lock Register). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 751 16.0 Per Thread WDT (B0:D31:F7) This is the Watchdog Host Controller device to implement per thread reset capability. The controller manifests in the PCH as a PCI Express* legacy EndPoint. This device is intended to be MSI-X compatible and the design/validation should reflect this compatibility. 16.1 PCI Registers This section provides details on list of registers that will be supported. 16.2 Register Attribute Legends The following terminology is used in the definition of the register attribute in this section: Register Attribute Meaning RO Read only register RO/V Read only register / value is variable RW Read-Write register RWS RW with special features WO/V Write only register / value is variable RO/S Read only sticky (only reset by loss of power) Intel(R) Communications Chipset 89xx Series - Datasheet 752 October 2012 Order Number: 327879-001US 16.0 16.3 Configuration Space All of the WDT configuration registers are in the core well. All registers not mentioned are reserved. The configuration space of the PCI registers is accessed through configuration cycle type 0, and it is accessible only as DW (32b) granularity (with byte enables). 16.3.1 PCI Header (B0:D31:F7) Table 16-1. PCI Header (B0:D31:F7) Offset Start Offset End Default Value Register ID - Description 00h 03h "Offset 00h: ID--Identifiers" on page 753 8086h 04h 05h "Offset 04h: CMD--Command" on page 754 0000h 06h 07h "Offset 06h: DSTS--Device Status" on page 755 0011h "Offset 08h: RID--Revision ID" on page 756 See register description 08h 08h 09h 09h "Offset 09h: CC--Class Code" on page 756 088000h 0Ch 0Ch "Offset 0Ch: CLS--Cache Line Size" on page 757 00h 0Dh 0Dh "Offset 0Dh: MLT--Master Latency Timer" on page 757 00h 0Eh 0Eh "Offset 0Eh: HTYPE--Header Type" on page 757 00h 10h 13h "Offset 10h: TBAR--WDT Table Base Address" on page 758 00000000h 14h 17h "Offset 14h: PBAR--WDT PBA Base Address" on page 758 00000000h 18h 1Ch "Offset 18h: CBAR--WDT CFG Base Address" on page 759 00000000h 2Dh 2Dh "Offset 2Ch: SSVID--Subsystem Vendor ID" on page 759 XXh 2Eh 2Eh "Offset 2Eh: SSID Subsystem ID" on page 760 XXh 34h 34h "Offset 34h: CAP--Capabilities Pointer" on page 760 80h 16.3.1.1 Offset 00h: ID--Identifiers Table 16-2. Offset 00h: ID--Identifiers Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B0:D31 :F7 Default: 23608086h Bit Range Bit Acronym 31 :16 DID 15 :00 VID October 2012 Order Number: 327879-001US Offset Start: 00h Offset End: 03h Power Well: Core Bit Reset Value Bit Access Device Identification -- Indicates the device ID assigned to this PCI device. DID = 2360h 2360h RO/V Vendor Identification -- This 16-bit value is assigned to Intel. Intel VID = 8086h 8086h RO Bit Description Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 753 16.3.1.2 Offset 04h: CMD--Command Table 16-3. Offset 04h: CMD--Command Description: View: PCI Size: 16 bit BAR: Configuration Bit Acronym 15 :00 Reserved 10 ID Reserved Offset Start: 04h Offset End: 05h Power Well: Core Bit Description Reserved Sticky Bit Reset Value Bit Access 0b Interrupt Disable -- This bit has no effect. 0b Reserved 0b RW 02 BME Bus Master Enable -- Must be set to `1' for the device to generate PCI Message Signaled Interrupts (MSI). Acts as global interrupt mask. This bit does not affect a device's ability to generate completions as a target. 0b RW 01 MSE Memory Space Enable -- Enables memory space on the device. 0b RW 00 IOSE I/O Space Enable -- Enables IO space on the device. Hardwired to 0 as no I/O space implemented. 0b RO Intel(R) Communications Chipset 89xx Series - Datasheet 754 B0:D31 :F7 Default: 0000h Bit Range 09 :03 Bus:Device:Function: October 2012 Order Number: 327879-001US 16.0 16.3.1.3 Offset 06h: DSTS--Device Status Table 16-4. Offset 06h: DSTS--Device Status Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B0:D31 :F7 Default: 0011h Offset Start: 06h Offset End: 07h Power Well: Core Bit Reset Value Bit Access SFE Signaled Fatal Error -- This register is set when a device sends an ERR_FATAL message. For Bridges this register is set either because the bridge was the detecting agent of a fatal error and fatal error reporting is enabled, or when forwarding a ERR_FATAL message from secondary bus to primary bus. Note: An ERR_FATAL message is sent when the SFE register is set to a `1' (set event) independent of the present state of the SFE register. This register is only reset by a loss of power. Hardwired to 0 as no error reporting is implemented. 0b RO SNE Signaled Non-Fatal Error -- This register is set when a device sends an ERR_NONFATAL message. For Bridges this register is set either because the bridge was the detecting agent of a Non-Fatal error and NonFatal error reporting is enabled, or when forwarding a ERR_NONFATAL message from secondary bus to primary bus. Note: An ERR_NONFATAL message is sent when the SNE register is set to a `1' (set event) independent of the present state of the SNE register. This register is only reset by a loss of power. Hardwired to 0 as no error reporting is implemented. 0b RO 13 SPT Signaled Poisoned TLP -- PCI-M Type 0 Devices: This register is set when a device sends a TLP with the EP bit set. For Bridges this register is set either because the bridge was the detecting agent of a poisoned TLP (independent of any enable bits) targeting the primary bus, or when forwarding a poisoned TLP from secondary bus to primary bus. This register is only reset by a loss of power. Hardwired to 0 as no error reporting is implemented. 0b RO 12 RUR Received Unsupported Request -- This register is set when a device receives a request that it doesn't support. This register is only reset by a loss of power. Hardwired to 0 as no error reporting is implemented. 0b RO RUC Received Unsuccessful Completion -- This register is set when a device receives a completion with unsuccessful completion status. Note: This register is used for tracking purposes only. No error message is generated in response to setting this bit. This register is only reset by a loss of power. 0b RO RPT Received Poisoned TLP -- This register is set when a device receives a downstream TLP with the EP bit set. Note: This register is used for tracking purposes only. No error message is generated in response to setting this bit. This register is only reset by a loss of power. 0b RO/S Bit Range 15 14 11 10 09 :05 Bit Acronym Reserved October 2012 Order Number: 327879-001US Bit Description Reserved Sticky 00h Intel(R) Communications Chipset 89xx Series - Datasheet 755 Table 16-4. Offset 06h: DSTS--Device Status Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B0:D31 :F7 Default: 0011h Bit Acronym Offset Start: 06h Offset End: 07h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 1 RO RO 04 CL Capabilities List -- Indicates presence of capabilities list. 03 IS Interrupt Status -- Reflects the state of INTx# messages. This bit is set when the interrupt is to be asserted. This bit is a 0 after the interrupt is cleared (independent of the state of CMD.ID). 0b Reserved 1b 02 :00 16.3.1.4 Reserved Offset 08h: RID--Revision ID Table 16-5. Offset 08h: RID--Revision ID Description: View: PCI Size: 8 bit Default: See register description Bit Range Bit Acronym 07 :00 RID 16.3.1.5 Bus:Device:Function: B0:D31 :F7 BAR: Configuration Offset Start: 08h Offset End: 08h Power Well: Bit Description Sticky Revision ID -- Indicates stepping of the host controller hardware. Bit Reset Value Bit Access 00h RO Offset 09h: CC--Class Code Table 16-6. Offset 09h: CC--Class Code Description: View: PCI Size: 24 bit Bus:Device:Function: B0:D31 :F7 BAR: Configuration Default: 088000h Bit Range Bit Acronym 23 :16 BCC 15 :08 SCC 07 :00 PI Power Well: Bit Reset Value Bit Access Base Class Code -- Device Base Class Code This field is reserved in this version of the specification. 08h RO Sub-Class Code -- Device Sub-Class Code This field is reserved in this version of the specification. 80h RO Programming Interface -- Device Programming Interface. This field is reserved in this version of the specification. 00h RO Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 756 Offset Start: 09h Offset End: 09h Sticky October 2012 Order Number: 327879-001US 16.0 16.3.1.6 Offset 0Ch: CLS--Cache Line Size Table 16-7. Offset 0Ch: CLS--Cache Line Size Description: View: PCI Size: 8 bit BAR: Configuration B0:D31 :F7 Default: 00h Bit Range Bit Acronym 07 :00 CLS 16.3.1.7 Bus:Device:Function: Offset Start: 0Ch Offset End: 0Ch Power Well: Bit Description Sticky Cache Line Size -- This register has no meaning for this device as it lives on DMI. Bit Reset Value Bit Access 00h RO Offset 0Dh: MLT--Master Latency Timer Table 16-8. Offset 0Dh: MLT--Master Latency Timer Description: View: PCI Size: 8 bit Default: 00h Bit Range Bit Acronym 07 :00 MLT 16.3.1.8 B0:D31 Bus:Device:Function: :F7 BAR: Configuration Offset Start: 0Dh Offset End: 0Dh Power Well: Bit Description Sticky Master Latency Timer -- This register has no meaning as the device lives on DMI. Bit Reset Value Bit Access 00h RO Offset 0Eh: HTYPE--Header Type Table 16-9. Offset 0Eh: HTYPE--Header Type Description: View: PCI Size: 8 bit Bit Range 07 06 :00 BAR: Configuration Bus:Device:Function: B0:D31 :F7 Default: 00h Bit Acronym MFD HL October 2012 Order Number: 327879-001US Offset Start: 0Eh Offset End: 0Eh Power Well: Bit Description Sticky Multi-function Device -- Indicates this controller is not part of a multi-function device. Header Layout -- Indicates that the controller uses a target device layout. Bit Reset Value Bit Access 0b RO 00h RO Intel(R) Communications Chipset 89xx Series - Datasheet 757 16.3.1.9 Offset 10h: TBAR--WDT Table Base Address Table 16-10. Offset 10h: TBAR--WDT Table Base Address Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: Default: 00000000h Bit Acronym Bit Description 31 :11 MBA Memory Base Address -- Software programs this register with the base address of the device's memory region. This allows for 2KB of memory 10 :04 Reserved 00 16.3.1.10 Sticky Reserved Bit Reset Value Bit Access 0000h RW 00h PF PF -- Indicates that this range is not pre-fetchable MS Memory Space -- Set to 0 for Memory Space BAR's. 02 :01 Offset Start: 10h Offset End: 13h Power Well: Core Bit Range :03 B0:D31 :F7 0b Width -- This is a 32 bit register RO 00b RO 0b RO Offset 14h: PBAR--WDT PBA Base Address Table 16-11. Offset 14h: PBAR--WDT PBA Base Address Description: View: PCI Size: 32 bit Bus:Device:Function: B0:D31 :F7 BAR: Configuration Default: 00000000h Power Well: Core Bit Range Bit Acronym Bit Description 31 :07 MBA Memory Base Address -- Software programs this register with the base address of the device's memory region. This allows for 128 Bytes of memory 06 :04 Reserved 03 PF 02 :01 00 Reserved PF -- Indicates that this range is not pre-fetchable Width -- This is a 32 bit register MS Memory Space -- Set to 0 for Memory Space BAR's. Intel(R) Communications Chipset 89xx Series - Datasheet 758 Offset Start: 14h Offset End: 17h Sticky Bit Reset Value Bit Access 0000h RW 00h 0b RO 00b RO 0b RO October 2012 Order Number: 327879-001US 16.0 16.3.1.11 Offset 18h: CBAR--WDT CFG Base Address Table 16-12. Offset 18h: CBAR--WDT CFG Base Address Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: Default: 00000000h Bit Acronym Bit Description 31 :10 MBA Memory Base Address -- Software programs this register with the base address of the device's memory region. This allows for 128 Bytes of memory 09 :04 Reserved PF 00 16.3.1.12 Sticky Reserved Width -- This is a 32 bit register MS Bit Reset Value Bit Access 0000h RW 00h PF -- Indicates that this range is not pre-fetchable 02 :01 Offset Start: 18h Offset End: 1Ch Power Well: Core Bit Range 03 B0:D31 :F7 Memory SpaceSSVID: Set to 0 for Memory Space BAR's. 0b RO 00b RO 0b RO Offset 2Ch: SSVID--Subsystem Vendor ID This register, in combination with the Subsystem ID register, enables the operating system to distinguish each subsystem from the others. There is no reset for this register. Table 16-13. Offset 2Ch: SSVID--Subsystem Vendor ID Description: View: PCI Size: 8 bit B0:D31 Bus:Device:Function: :F7 BAR: Configuration Default: XXh Bit Range Bit Acronym 07 :00 SSVID October 2012 Order Number: 327879-001US Offset Start: 2Dh Offset End: 2Dh Power Well: Core Bit Description Subsystem Vendor ID Sticky Bit Reset Value Bit Access XXh RWO Intel(R) Communications Chipset 89xx Series - Datasheet 759 16.3.1.13 Offset 2Eh: SSID Subsystem ID BIOS sets the value in this register to identify the Subsystem ID. This register, in combination with the Subsystem Vendor ID register, enables the operating system to distinguish each subsystem from other(s). There is no reset for this register. Table 16-14. Offset 2Eh: SSID Subsystem ID Description: View: PCI BAR: Configuration Size: 8 bit Bus:Device:Function: B0:D31 :F7 Default: XXh Bit Range Bit Acronym 07 :00 SSID 16.3.1.14 Offset Start: 2Eh Offset End: 2Eh Power Well: Core Bit Description Sticky Subsystem ID Bit Reset Value Bit Access XXh RWO Offset 34h: CAP--Capabilities Pointer Table 16-15. Offset 34h: CAP--Capabilities Pointer Description: View: PCI Size: 8 bit Default: 80h Bit Range Bit Acronym 07 :00 CP 16.3.2 Bus:Device:Function: B0:D31 :F7 BAR: Configuration Offset Start: 34h Offset End: 34h Power Well: Bit Description Capability Pointer -- Indicates that the first capability pointer offset is offset 80h (the Message Signaled Interrupt capability). Sticky Bit Reset Value Bit Access 80h RO PCI MSI-X Capability Table 16-16. PCI MSI-X Capability Offset Start 80h Offset End 81h Register ID - Description "Offset 80h: MID--MSI-X Capability ID" on page 761 Default Value 0011h 82h 83h "Offset 82h: MC--MSI-X Control & Status" on page 761 007Fh 84h 87h "Offset 84h: MT--MSI-X Table" on page 761 00000000h 88h 8Bh "Offset 88h: MP--MSI-X Message PBA" on page 762 00000001h Intel(R) Communications Chipset 89xx Series - Datasheet 760 October 2012 Order Number: 327879-001US 16.0 16.3.2.1 Offset 80h: MID--MSI-X Capability ID Table 16-17. Offset 80h: MID--MSI-X Capability ID Description: View: PCI Size: 8 bit BAR: Bus:Device:Function: Default: 0011h Offset Start: 80h Offset End: 81h Power Well: Bit Range Bit Acronym Bit Description 15 :08 NEXT Next Capability -- The location of the next capability structure. Hardwired to 00h indicating that this is the last capability header. 07 :00 CID 16.3.2.2 B0:D31 :F7 Sticky Capability ID -- Capabilities ID indicates MSI-X. Bit Reset Value Bit Access 000h RO 11h RO Offset 82h: MC--MSI-X Control & Status Table 16-18. Offset 82h: MC--MSI-X Control & Status Description: View: PCI Size: 16 bit Bit Range B0:D31 Bus:Device:Function: :F7 BAR: Default: 0000007Fh Bit Acronym Offset Start: 82h Offset End: 83h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 15 Enable -- No interrupts sent unless this bit is set 0b RW 14 Mask -- No effect 0b RW Reserved 0h Table Size -- Size of MSI-X Table. Set to 0x7F for 128 interrupts 7Fh 13 :11 Reserved 10 :00 16.3.2.3 RO Offset 84h: MT--MSI-X Table Table 16-19. Offset 84h: MT--MSI-X Table Description: View: PCI Size: 32 bit Bit Range Default: 00000000h Bit Acronym 31 :03 02 :00 Bus:Device:Function: B0:D31 :F7 BAR: Power Well: Core Bit Description Sticky Offset -- location in BAR of the MSI-X Table BIR October 2012 Order Number: 327879-001US Offset Start: 84h Offset End: 87h Indicates BAR 0 at 0x10 Bit Reset Value Bit Access 0000h RO 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 761 16.3.2.4 Offset 88h: MP--MSI-X Message PBA Table 16-20. Offset 88h: MP--MSI-X Message PBA Description: View: PCI Size: 32 bit Bit Range BAR: 16.4 B0:D31 :F7 Default: 00000001h Bit Acronym 31 :03 02 :00 Bus:Device:Function: Power Well: Core Bit Description Offset -- location in BAR of the MSI-X Table BIR Offset Start: 88h Offset End: 8Bh Indicates BAR 0 at 0x14 Sticky Bit Reset Value Bit Access 0000h RO 1h RO MMIO Space The MMIO space of the PCI registers is accessed through memory cycle type, and it is accessible only at DW (32b) granularity (with byte enables). It does not support burst accesses to the register sets. Implementation note: It is expected that design will code the WDT section in a parameterisable manner such that the number of WDTs can be easily varied. The PCH will have a total of 64 WDTs. The configuration of the WDT device has been formatted to fit into an MSI-X compatible device for possible future compatibility with mainstream OS. 16.4.1 Table BAR Range There are eight logical registers for each WDT (four for each interrupt) to comply with the MSI specification. The detail for the first eight CSRs are provided in Table 16-21 below. The register layout for the other WDTs in the BAR is identical. Note: Some registers in this BAR are aliases to each other. This means that the physical register exists at two different offsets and can be read/written at two different addresses. System software only needs to write to one of the addresses in such cases. Other registers are reserved/hardwired to zero - this means they always read as `0x00000000' and write have no effect. The entire Table BAR contents become read only if WDT_GBLCFG.T_LOCK is set, and only a hardware reset will unlock the contents. The BAR has been designed to be largely MSI-X compatible but the aliases and reserved registers violate this - necessary to reduce the impact to floorplan. The descriptions in the expected usage fields hereafter are for situations where interrupt remapping is disabled in the core, which is the expected usage for the PCH. Intel(R) Communications Chipset 89xx Series - Datasheet 762 October 2012 Order Number: 327879-001US 16.0 Table 16-21. Table BAR Range Offset Start Offset End Register ID - Description Default Value 00h 03h "Offset 00h: WDT0MAW--WDT 0 Msg Add WARN" on page 764 FEE00000h 04h 07h "Offset 04h: WDT0MUAW--WDT 0 Msg Upper Add WARN" on page 764 00000000h 08h 0Bh "Offset 08h: WDT0MDW--WDT 0 Msg Data WARN" on page 765 00004000h 0Ch 0Fh "Offset 0Ch: WDT0VCW--WDT 0 Vector Ctrl WARN" on page 765 00000000h 10h 13h "Offset 10h: WDT0MAR--WDT 0 Msg Add RESET" on page 765 See Register Description 14h 17h "Offset 14h: WDT0MUAR--WDT 0 Msg Upper Add RESET" on page 765 00000000h 18h 1Bh "Offset 18h: WDT0MDR--WDT 0 Msg Data RESET" on page 766 00004000h 1Ch 1Fh "Offset 1Ch: WDT0VCR--WDT 0 Vector Ctrl RESET" on page 766 00000000h ... and so on for a total of 64 WDTs 7E0h 7E3h WDT63MAW - WDT 63 Message Address Warn 7E4h 7E7h WDT63MUAW - WDT 63 Message Upper Address Warm 7E8h 7EBh WDT63MDW - WDT 63 Message Data Warn 7ECh 7EFh WDT63VCW - WDT 63 Vector Control Warn 7F0h 7F3h WDT63MAR - WDT 63 Message Address Reset 7F4h 7F7h WDT63MUAR - WDT 63 Message Upper Address Reset 7F8h 7FBh WDT63MDR - WDT 63 Message Data Reset 7FCh 7FFh WDT63VCR - WDT 63 Vector Control Reset October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 763 16.4.1.1 Offset 00h: WDT0MAW--WDT 0 Msg Add WARN Table 16-22. Offset 00h: WDT0MAW--WDT 0 Msg Add WARN Description: View: PCI Size: 32 bit Bit Range BAR: B0:D31 :F7 Default: FEE00000h Bit Acronym 31 :20 19 :12 Bus:Device:Function: Offset Start: 00h Offset End: 03h Power Well: Bit Description Sticky Bit Reset Value Bit Access FEEh RO Address of local APIC area APICID 11 :04 APICID -- for local APIC identification 00h RW Extended APICID -- for local APIC identification 00h RW 03 Should be set to 0 0h RW 02 As WDT0MDW[11] - must be set to same value 0 => physical mode 1 => logical mode 0h RW Reserved 0h 01 :00 16.4.1.2 Reserved Offset 04h: WDT0MUAW--WDT 0 Msg Upper Add WARN This register is hardwired to zero. Intel(R) Communications Chipset 89xx Series - Datasheet 764 October 2012 Order Number: 327879-001US 16.0 16.4.1.3 Offset 08h: WDT0MDW--WDT 0 Msg Data WARN Table 16-23. Offset 08h: WDT0MDW--WDT 0 Msg Data WARN Description: View: PCI Size: 32 bit BAR: Bus:Device:Function: B0:D31 :F7 Default: 00004000h Bit Range Bit Acronym 31 :16 Reserved Offset Start: 08h Offset End: 0Bh Power Well: Bit Description Sticky Bit Reset Value Reserved Bit Access 0000h 15 Should be set to 0 0h RW 14 Should be set to 1 1h RW 13 :12 Should be set to 0 0h RW As WDT0MAW[2] - must be set to same value 0 => physical mode 1 => logical mode 0h RW 10 :08 Delivery Mode for Warning interrupt 0h RW 07 :00 Vector for interrupt 0h RW 11 16.4.1.4 Offset 0Ch: WDT0VCW--WDT 0 Vector Ctrl WARN Table 16-24. Offset 0Ch: WDT0VCW--WDT 0 Vector Ctrl WARN Description: View: PCI Size: 32 bit Default: 00000000h Bit Range Bit Acronym 31 :01 Reserved 00 B0:D31 Bus:Device:Function: :F7 BAR: MASK Offset Start: 0Ch Offset End: 0Fh Power Well: Core Bit Description Sticky Bit Reset Value Reserved Bit Access 0000h Masks the Warning Interrupt. Corresponding PBA bit will still get set but the interrupt will not be delivered. Note: The other WDT (1-63) MSI WARN registers are defined in like manner 16.4.1.5 Offset 10h: WDT0MAR--WDT 0 Msg Add RESET 0h RW This register is an alias of WDT0MAW. 16.4.1.6 Offset 14h: WDT0MUAR--WDT 0 Msg Upper Add RESET This register is hardwired to zero. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 765 16.4.1.7 Offset 18h: WDT0MDR--WDT 0 Msg Data RESET Table 16-25. Offset 18h: WDT0MDR--WDT 0 Msg Data RESET Description: View: PCI Size: 32 bit BAR: Bus:Device:Function: B0:D31 :F7 Default: 00004000h Bit Range Bit Acronym 31 :16 Reserved Offset Start: 18h Offset End: 1Bh Power Well: Bit Description Sticky Reserved Bit Reset Value Bit Access 0000h 15 Should be set to 0 0h RW 14 Should be set to 1 1h RW Should be set to 0 0h RW As WDT0MAW[2] - must be set to same value 0 => physical mode; 1 => logical mode 0h RW 13 :12 11 10 :08 Delivery Mode for RESET interrupt 0h RW 07 :00 Vector for interrupt 0h RW 16.4.1.8 Offset 1Ch: WDT0VCR--WDT 0 Vector Ctrl RESET Table 16-26. Offset 1Ch: WDT0VCR--WDT 0 Vector Ctrl RESET Description: View: PCI Size: 32 bit Default: 00000000h Bit Range Bit Acronym 31 :01 Reserved 00 Note: Bus:Device:Function: B0:D31 :F7 BAR: MASK Power Well: Core Bit Description Reserved Masks the RESET Interrupt. Corresponding PBA bit will still get set but the interrupt will not be delivered. Sticky Bit Reset Value Bit Access 0000h 0h RW The other WDT (1-63) MSI RESET registers are defined in like manner. Intel(R) Communications Chipset 89xx Series - Datasheet 766 Offset Start: 0Ch Offset End: 0Fh October 2012 Order Number: 327879-001US 16.0 16.4.2 PBA BAR Range PBA stands for pending bit array. This BAR complies with the MSI specification. This is a minimum size BAR as the array is small. The pending bits are set to note that RESET/WARN interrupts are pending. The bits will be set when the RESET/WARN event occurs even if the corresponding interrupt is masked. The pending bits will be cleared when the corresponding interrupt is sent or when the WDT for the RESET/WARN pair is reloaded. Setting the RESET bit will clear the corresponding WARN bit. SW cannot write any of the bits. Table 16-27. PBA BAR Range Offset Start Offset End Register ID - Description Default Value 00h 03h "Offset 00h: WDT_PBA0--Pending Bit Array 0" on page 768 00000000h 04h 07h "Offset 04h: WDT_PBA1--Pending Bit Array 1" on page 769 00000000h 08h 0Bh "Offset 08h: WDT_PBA2--Pending Bit Array 2" on page 770 00000000h 0Ch 0Fh "Offset 0Ch: WDT_PBA3--Pending Bit Array 3" on page 771 00000000h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 767 16.4.2.1 Offset 00h: WDT_PBA0--Pending Bit Array 0 Table 16-28. Offset 00h: WDT_PBA0--Pending Bit Array 0 Description: View: PCI Size: 32 bit Bit Range BAR: Bus:Device:Function: Default: 00000000h Bit Acronym Offset Start: 00h Offset End: 03h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0h RO 31 Pending Bit for RESET 15 30 Pending Bit for WARN 15 0h RO 29 Pending Bit for RESET 14 0h RO 28 Pending Bit for WARN 14 0h RO 27 Pending Bit for RESET 13 0h RO 26 Pending Bit for WARN 13 0h RO 25 Pending Bit for RESET 12 0h RO 24 Pending Bit for WARN 12 0h RO 23 Pending Bit for RESET 11 0h RO 22 Pending Bit for WARN 11 0h RO 21 Pending Bit for RESET 10 0h RO 20 Pending Bit for WARN 10 0h RO 19 Pending Bit for RESET 9 0h RO 18 Pending Bit for WARN 9 0h RO 17 Pending Bit for RESET 8 0h RO 16 Pending Bit for WARN 8 0h RO 15 Pending Bit for RESET 7 0h RO 14 Pending Bit for WARN 7 0h RO 13 Pending Bit for RESET 6 0h RO 12 Pending Bit for WARN 6 0h RO 11 Pending Bit for RESET 5 0h RO 10 Pending Bit for WARN 5 0h RO 09 Pending Bit for RESET 4 0h RO 08 Pending Bit for WARN 4 0h RO 07 Pending Bit for RESET 3 0h RO 06 Pending Bit for WARN 3 0h RO 05 Pending Bit for RESET 2 0h RO 04 Pending Bit for WARN 2 0h RO 03 Pending Bit for RESET 1 0h RO 02 Pending Bit for WARN 1 0h RO 01 Pending Bit for RESET 0 0h RO 00 Pending Bit for WARN 0 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 768 B0:D31 :F7 October 2012 Order Number: 327879-001US 16.0 16.4.2.2 Offset 04h: WDT_PBA1--Pending Bit Array 1 Table 16-29. Offset 04h: WDT_PBA1--Pending Bit Array 1 Description: View: PCI Size: 32 bit Bit Range BAR: Bus:Device:Function: B0:D31 :F7 Default: 00000000h Bit Acronym Offset Start: 04h Offset End: 07h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0h RO 31 Pending Bit for RESET 31 30 Pending Bit for WARN 31 0h RO 29 Pending Bit for RESET 30 0h RO 28 Pending Bit for WARN 30 0h RO 27 Pending Bit for RESET 29 0h RO 26 Pending Bit for WARN 29 0h RO 25 Pending Bit for RESET 28 0h RO 24 Pending Bit for WARN 28 0h RO 23 Pending Bit for RESET 27 0h RO 22 Pending Bit for WARN 27 0h RO 21 Pending Bit for RESET 26 0h RO 20 Pending Bit for WARN 26 0h RO 19 Pending Bit for RESET 25 0h RO 18 Pending Bit for WARN 25 0h RO 17 Pending Bit for RESET 24 0h RO 16 Pending Bit for WARN 24 0h RO 15 Pending Bit for RESET 23 0h RO 14 Pending Bit for WARN 23 0h RO 13 Pending Bit for RESET 22 0h RO 12 Pending Bit for WARN 22 0h RO 11 Pending Bit for RESET 21 0h RO 10 Pending Bit for WARN 21 0h RO 09 Pending Bit for RESET 20 0h RO 08 Pending Bit for WARN 20 0h RO 07 Pending Bit for RESET 19 0h RO 06 Pending Bit for WARN 19 0h RO 05 Pending Bit for RESET 18 0h RO 04 Pending Bit for WARN 18 0h RO 03 Pending Bit for RESET 17 0h RO 02 Pending Bit for WARN 17 0h RO 01 Pending Bit for RESET 16 0h RO 00 Pending Bit for WARN 16 0h RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 769 16.4.2.3 Offset 08h: WDT_PBA2--Pending Bit Array 2 Table 16-30. Offset 08h: WDT_PBA2--Pending Bit Array 2 Description: View: PCI Size: 32 bit Bit Range BAR: Bus:Device:Function: Default: 00000000h Bit Acronym Offset Start: 08h Offset End: 0Bh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0h RO 31 Pending Bit for RESET 47 30 Pending Bit for WARN 47 0h RO 29 Pending Bit for RESET 46 0h RO 28 Pending Bit for WARN 46 0h RO 27 Pending Bit for RESET 45 0h RO 26 Pending Bit for WARN 45 0h RO 25 Pending Bit for RESET 44 0h RO 24 Pending Bit for WARN 44 0h RO 23 Pending Bit for RESET 43 0h RO 22 Pending Bit for WARN 43 0h RO 21 Pending Bit for RESET 42 0h RO 20 Pending Bit for WARN 42 0h RO 19 Pending Bit for RESET 41 0h RO 18 Pending Bit for WARN 41 0h RO 17 Pending Bit for RESET 40 0h RO 16 Pending Bit for WARN 40 0h RO 15 Pending Bit for RESET 39 0h RO 14 Pending Bit for WARN 39 0h RO 13 Pending Bit for RESET 38 0h RO 12 Pending Bit for WARN 38 0h RO 11 Pending Bit for RESET 37 0h RO 10 Pending Bit for WARN 37 0h RO 09 Pending Bit for RESET 36 0h RO 08 Pending Bit for WARN 36 0h RO 07 Pending Bit for RESET 35 0h RO 06 Pending Bit for WARN 35 0h RO 05 Pending Bit for RESET 34 0h RO 04 Pending Bit for WARN 34 0h RO 03 Pending Bit for RESET 33 0h RO 02 Pending Bit for WARN 33 0h RO 01 Pending Bit for RESET 32 0h RO 00 Pending Bit for WARN 32 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 770 B0:D31 :F7 October 2012 Order Number: 327879-001US 16.0 16.4.2.4 Offset 0Ch: WDT_PBA3--Pending Bit Array 3 Table 16-31. Offset 0Ch: WDT_PBA3--Pending Bit Array 3 Description: View: PCI Size: 32 bit Bit Range BAR: Bus:Device:Function: B0:D31 :F7 Default: 00000000h Bit Acronym Offset Start: 0Ch Offset End: 0Fh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0h RO 31 Pending Bit for RESET 63 30 Pending Bit for WARN 63 0h RO 29 Pending Bit for RESET 62 0h RO 28 Pending Bit for WARN 62 0h RO 27 Pending Bit for RESET 61 0h RO 26 Pending Bit for WARN 61 0h RO 25 Pending Bit for RESET 60 0h RO 24 Pending Bit for WARN 60 0h RO 23 Pending Bit for RESET 59 0h RO 22 Pending Bit for WARN 59 0h RO 21 Pending Bit for RESET 58 0h RO 20 Pending Bit for WARN 58 0h RO 19 Pending Bit for RESET 57 0h RO 18 Pending Bit for WARN 57 0h RO 17 Pending Bit for RESET 56 0h RO 16 Pending Bit for WARN 56 0h RO 15 Pending Bit for RESET 55 0h RO 14 Pending Bit for WARN 55 0h RO 13 Pending Bit for RESET 54 0h RO 12 Pending Bit for WARN 54 0h RO 11 Pending Bit for RESET 53 0h RO 10 Pending Bit for WARN 53 0h RO 09 Pending Bit for RESET 52 0h RO 08 Pending Bit for WARN 52 0h RO 07 Pending Bit for RESET 51 0h RO 06 Pending Bit for WARN 51 0h RO 05 Pending Bit for RESET 50 0h RO 04 Pending Bit for WARN 50 0h RO 03 Pending Bit for RESET 49 0h RO 02 Pending Bit for WARN 49 0h RO 01 Pending Bit for RESET 48 0h RO 00 Pending Bit for WARN 48 0h RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 771 16.4.3 CFG BAR Range Table 16-32. CFG BAR Range Offset Start Offset End Default Value Register ID - Description 00h 03h "Offset 00h: WDT0COUNT--WDT Count Register" on page 772 00000000h 04h 07h "Offset 04h: WDT0CMD--WDT Cmd/Status Register" on page 773 00000000h .... and so on for a total of 64 WDTs 1F8h 1FBh WDT63COUNT - WDT Count Register 1FCh 1FFh WDT63CMD - WDT Cmd/Stat Register Global Config Section starts at address 0x200 200h 203h "Offset 200h: WDT_PSCALE" on page 773 00000000h 204h 207h "Offset 204h: WDT_GBLCFG" on page 774 00000000h 16.4.3.1 Offset 00h: WDT0COUNT--WDT Count Register Table 16-33. Offset 00h: WDT0COUNT--WDT Count Register Description: View: PCI Size: 32 bit B0:D31 Bus:Device:Function: :F7 BAR: Default: 00000000h Power Well: Core Bit Range Bit Acronym Bit Description 31 :10 COUNT Counter Reload Value for bits [31:10] of counter reloading is disabled if this value is zero. Events (setting PBA bit and interrupts) will not trigger if this value is zero. Value can also be locked setting LOCK bit. 09 :01 Reserved 00 LOCK Reserved - the lower 10 bits of the reload value are always zero Locks bits [31:10] when = 1. Writes of 0 have no effect, requires a reset to unlock the COUNT value. Intel(R) Communications Chipset 89xx Series - Datasheet 772 Offset Start: 00h Offset End: 03h Sticky Bit Reset Value Bit Access 000h RWS 00h 0h RWS October 2012 Order Number: 327879-001US 16.0 16.4.3.2 Offset 04h: WDT0CMD--WDT Cmd/Status Register Any write of any value to this registers will clear its contents and will also reload the WDT timer (if the WDTnCOUNT.COUNT reload value is non-zero). Table 16-34. Offset 04h: WDT0CMD--WDT Cmd/Status Register Description: View: PCI Size: 32 bit BAR: Bus:Device:Function: B0:D31 :F7 Default: 00000000h Bit Range Bit Acronym 31 :02 Reserved Offset Start: 04h Offset End: 07h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00h 01 RT RESET interrupt triggered. This gets set when the RESET interrupt message is issued. SW resets by any write to this register. 0h RWS 00 WT WARN interrupt triggered. This gets set when the WARN interrupt message is issued. SW resets by any write to this register. 0h RWS Note: 63 further pairs of WDT Count/Cmd registers exist at addresses to 0x13C and are all defined in an identical manner. 16.4.3.3 Offset 200h: WDT_PSCALE The prescaler counter runs as a free running counter. It cannot be disabled. The counter rolls over (and decrements the operating WDT counters) when this value is reached. Table 16-35. Offset 200h: WDT_PSCALE Description: View: PCI Size: 32 bit Bit Range 31 30 :10 Bus:Device:Function: B0:D31 :F7 BAR: Default: 00000000h Bit Acronym LOCK Reserved 09 :00 October 2012 Order Number: 327879-001US Offset Start: 200h Offset End: 203h Power Well: Core Bit Description Sticky Locks bits [9:0] when = 1. Writes of 0 have no effect. Requires a reset to unlock the prescaler. Reserved Bit Reset Value Bit Access 0h RWS 00h Prescaler Value applied to all Watchdog timers. Read Only when bit 31 = `1'. 0h RWS Intel(R) Communications Chipset 89xx Series - Datasheet 773 16.4.3.4 Offset 204h: WDT_GBLCFG Table 16-36. Offset 204h: WDT_GBLCFG Description: View: PCI Size: 32 bit BAR: Bus:Device:Function: B0:D31 :F7 Default: 00000000h Bit Range Bit Acronym 31 :02 Reserved 01 T_LOCK 00 GEN Offset Start: 204h Offset End: 207h Power Well: Core Bit Description Reserved Sticky Bit Reset Value Bit Access 00h Makes the entire Table BAR contents read only when set. A write of 1 sets this bit. A write of 0 has no effect. A reset is required to unlock the Table BAR values if BIOS sets this bit. 0h RWS Global Enable - no timers will decrement if this bit is `0'. 0h RWS Intel(R) Communications Chipset 89xx Series - Datasheet 774 October 2012 Order Number: 327879-001US 17.0 17.0 Thermal Sensor Registers (B0:D31:F6) The PCH incorporates two on-die thermal sensors for thermal management.Thermal Sensor 0 (TS0) is the sensor located near the DMI and PCI Express interfaces. Thermal Sensor 1 (TS1) is the sensor located near the SATA and PCIe Ports. The PCH supports all of the capabilities of the four-function mode but does so in analog mode. Therefore, the PCH supports catastrophic, hot, AUX and AUX2 trip points, interrupts, and transitions (low-high and high-low). * Hot On: Calculated value is (itsalertthreshon[7:0] - 50C) * Hot Off: Calculated value is (itsalertthreshoff[7:0] - 50C) * Catastrophic: Calculated value is (itstripthreshon[7:0] - 50C) * AUX: Calculated value is (AUX[7:0] - 50C) * AUX2: Calculated value is (AUX2[7:0] - 50C) 17.1 Thermal Configuration Registers 17.1.1 PCI Configuration Topology The thermal register set is made available to the OS as a PCI Bus device (B0:D31:F6). The bus number is captured for every type 0 configuration write cycles matching the device.function number. The captured bus # is returned as part of the completer ID field of completions. 17.1.2 Configuration Register Access Restrictions All controller registers (including the memory mapped registers) must be addressable as byte, word, and D-word quantities. The software must always make register accesses on natural boundaries; D-word accesses must be on D-word boundaries, word accesses on word boundaries, etc. Note: The thermal memory-mapped register space must not be accessed with the LOCK semantic exclusive-access mechanism. If software attempts exclusive-access mechanisms to the thermal memory-mapped register space, the results are undefined. All of the registers are in the core well unless otherwise indicated. All registers not mentioned are reserved. Reserved registers will always read 00h and writes will have no effect. Software must properly handle reserved bits. Reserved bits may be designated `RsvdP' or `RsvdZ'. Bits marked `RsvdP' must be preserved using readmodify-writes, while `RsvdZ' bits must be written as zeros. This behavior helps to ensure future compatibility. 17.1.3 Temperature Alert The TEMP_ALERT# pin is controlled by MPC firmware. For the PCH, it is set to GP49. See the register descriptions for how to enable this feature. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 775 17.1.4 Register Attribute Legend The following terminology is used in the definition of the register attribute in this specification: Register Attribute Meaning RO Read-Only register RW Read-Write register RWC Read-only status, write-to-clear status register RW1C Read-only status, write-1-to-clear status register ROS Sticky bit- Read-Only register - not affected by reset RWS Sticky bit- Read-Write register - not affected by reset RW1CS Sticky bit- Read-only status, Write-1-to-clear status register HWINIT Hardware initialized; bits are read-only and cannot be reset RsvdP Reserved; software must do a read-modify-write to preserve the value of bits RsvdZ Reserved; software must use zeros for writes to bits RSM Bit(s) are in the `Resume Well', i.e., they maintain their state in any power state from which the Thermal controller may wake the system and are not affected by reset. 17.2 PCI Configuration Registers (Thermal Sensor - B0:D31:F6) 17.2.1 Thermal Sensor PCI Register Address Map Table 17-1. Bus 0, Device 31, Function 6: Summary of Thermal Reporting PCI Configuration Registers (Sheet 1 of 2) Offset Start Offset End Register ID - Description Default Value 00h 01h "Offset 00h: ID: Vendor Identification Register" on page 778 8086h 02h 03h "Offset 02h: DID--Device Identification Register" on page 778 2332h 04h 05h "Offset 04h: ID--Command Register" on page 778 0000h 06h 07h "Offset 06h: ID--STS--Status Register" on page 779 0010h 08h 08h "Offset 08h: RID--Revision Identification Register" on page 780 00h 09h 09h "Offset 09h: PI--Programming Interface Register" on page 781 00h 0Ah 0Ah "Offset 0Ah: SCC--Sub Class Code Register" on page 781 80h 0Bh 0Bh "Offset 0Bh: BCC--Base Class Code Register" on page 781 11h 0Ch 0Ch "Offset 0Ch: Cache Line Size" on page 782 00h 0Dh 0Dh "Offset 0Dh: Latency Timer" on page 782 00h 0Eh 0Eh "Offset 0Eh: Header Type" on page 782 00h 10h 13h "Offset 10h: Thermal Base" on page 783 00000004h 14h 17h "Offset 14h: TBARH--Thermal Base High DWord" on page 783 00000000h 2Ch 2Dh "Offset 2Ch: Subsystem Vendor ID" on page 784 0000h Intel(R) Communications Chipset 89xx Series - Datasheet 776 October 2012 Order Number: 327879-001US 17.0 Table 17-1. Bus 0, Device 31, Function 6: Summary of Thermal Reporting PCI Configuration Registers (Sheet 2 of 2) Offset Start 2Eh Offset End 2Fh Register ID - Description "Offset 2Eh: Subsystem ID" on page 785 Default Value 0000h 34h 34h "Offset 34h: Capabilities Pointer" on page 785 50h 3Ch 3Ch "Offset 3Ch: Interrupt Line" on page 786 00h 3Dh 3Dh "Offset 3Dh: Interrupt Pin" on page 786 See Description 40h 43h "Offset 40h: BIOS Assigned Thermal Base Address" on page 787 00000004h 44h 47h "Offset 44h: BIOS Assigned Thermal Base High DWord" on page 787 00000000h 50h 51h "Offset 50h: PCI Power Management Capability ID" on page 788 8001h 52h 53h "Offset 52h: Power Management Capabilities" on page 788 0023h 54h 57h "Offset 54h: Power Management Control And Status" on page 789 0008h 80h 81h "Offset 80h: MID--Message Signaled Interrupt Identifiers" on page 790 0005h 82h 83h "Offset 82h: MC--Message Signaled Interrupt Message Control" on page 790 0000h 84h 87h "Offset 84h: MA--Message Signaled Interrupt Message Address" on page 791 00000000h 88h 89h "Offset 88h: MD--Message Signaled Interrupt Message Data" on page 791 0000h 94h 94h "Offset 94h: CCTS0RD--Thermal Sensor 0 Remote Diode" on page 791 00h 96h 97h "Offset 96h: CCTS0SSI--Thermal Sensor 0 Sense Stage Inputs" on page 792 0066h 98h 9Bh "Offset 98h: CCTS0DSAC0--Thermal Sensor 0 Delta Sigma ADC Control 0" on page 793 02010083h 9Ch 9Dh "Offset 9Ch: CCTS0DSAC1--Thermal Sensor 0 Delta Sigma ADC Control 1" on page 793 0001h A0h A3h "Offset A0h: CCTS0C--Thermal Sensor 0 Calibration" on page 794 00030016h A4h A7h "Offset A4h: CCTS0TL0--Thermal Sensor 0 Top Logic 0" on page 794 80000000h A8h ABh "Offset A8h: CCTS0TL1--Thermal Sensor 0 Top Logic 1" on page 796 See Description ACh ADh "Offset ACh: CCTS0BP--Thermal Sensor 0 Bandgap Pins" on page 797 001Bh B4h B4h "Offset B4h: CCTS1RD--Thermal Sensor 1 Remote Diode" on page 797 0000h B6h B7h "Offset B6h: CCTS1SSI--Thermal Sensor 1 Sense Stage Inputs" on page 798 0066h B8h BBh "Offset B8h: CCTS1DSAC0--Thermal Sensor 1 Delta Sigma ADC Control 0" on page 798 02010083h BCh BDh "Offset BCh: CCTS1DSAC1--Thermal Sensor 1 Delta Sigma ADC Control 1" on page 799 0001h C0h C3h "Offset C0h: CCTS1C--Thermal Sensor 1 Calibration" on page 799 00030016h C4h C7h "Offset C4h: CCTS1TL0--Thermal Sensor 1 Top Logic 0" on page 800 80999BAAh C8h CBh "Offset C8h: CCTS1TL1--Thermal Sensor 1 Top Logic 1" on page 801 See Description CCh CDh "Offset CCh: CCTS1BP--Thermal Sensor 1 Bandgap Pins" on page 802 001Bh F8h FBh "Offset F8h: MANID--Manufacturing/Process " on page 802 00000F00h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 777 17.2.1.1 Offset 00h: VID--Vendor Identification Table 17-2. Offset 00h: ID: Vendor Identification Register Description: View: PCI Size: 16 bit BAR: Configuration Default: 8086h Bit Range Bit Acronym 15 :00 VID 17.2.1.2 Bus:Device:Function: B:31:6 Offset Start: 00h Offset End: 01h Power Well: Core Bit Description Sticky Vendor Identification: This 16-bit value is assigned to Intel. Bit Reset Value Bit Access 8086h RO Offset 02h: DID--Device Identification Table 17-3. Offset 02h: DID--Device Identification Register Description: View: PCI Size: 16 bit BAR: Configuration Default: 2332h Bit Range Bit Acronym 15 :00 DID 17.2.1.3 Bus:Device:Function: B:31:6 Offset Start: 02h Offset End: 03h Power Well: Core Bit Description Sticky Device Identification: This is a 16-bit value assigned by the PCI SIG. Bit Reset Value Bit Access 2332h RO Offset 04h: CMD--Command Table 17-4. Offset 04h: ID--Command Register Description: View: PCI Size: 16 bit BAR: Configuration Default: 0000h Bit Range Bit Acronym 15 :11 Reserved 10 Bus:Device:Function: B:31:6 ID Power Well: Core Bit Description Reserved Interrupt Disable: Enables the device to assert an INTx#. When set, the Thermal logic's INTx# signal will be deasserted. When cleared AND MSI is not enabled, the INTx# signal may be asserted. NOTE: this bit has no affect on MSI generation. Intel(R) Communications Chipset 89xx Series - Datasheet 778 Offset Start: 04h Offset End: 05h Sticky Bit Reset Value Bit Access 00h 0b RW October 2012 Order Number: 327879-001US 17.0 Table 17-4. Offset 04h: ID--Command Register Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 0000h Offset Start: 04h Offset End: 05h Power Well: Core Bit Reset Value Bit Access Fast Back to Back Enable: Not Implemented. Hardwired to `0'. 0b RO SEN SERR# Enable: Not Implemented. Hardwired to `0'. 0b RO WCC Wait Cycle Control: Not Implemented. Hardwired to `0'. 0b RO Parity Error Response Enable: Not Implemented. Hardwired to `0'. 0b RO VPS VGA Palette Snoop: Not Implemented. Hardwired to `0' 0b RO MWI Memory Write and Invalidate Enable: Not Implemented. Hardwired to `0'. 0b RO 03 SCE Special Cycle Enable: Not Implemented. Hardwired to `0'. 0h RO 02 BME Bus Master Enable: When set, enables. 0b RW 01 MSE Memory Space Enable: When set, enables memory space accesses to the Thermal registers. 0b RW 00 IOS I/O Space Enable: The Thermal logic does not implement IO Space. This bit is Hardwired to `0'. 0b RO Bit Range Bit Acronym Bit Description FBE 08 07 06 PER 05 09 04 17.2.1.4 Sticky Offset 06h: STS--Status Table 17-5. Offset 06h: ID--STS--Status Register (Sheet 1 of 2) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B:31:6 Default: 0010h Power Well: Core Bit Reset Value Bit Access Detected Parity Error: 0 = Parity error not detected. 1 = Indicates that the PCH detected a parity error on the internal backbone. This bit gets set even if the CMD.PER bit "Offset 04h: ID--Command Register" bit 6) is not set. Software clears this bit by writing a `1' to this bit location. 0b RW1C SERR# Status: Not Implemented. Hardwired to `0'. 0b RO RMA Received Master Abort: Not Implemented. Hardwired to `0'. 0b RO RTA Received Target Abort: Not Implemented. Hardwired to `0'. 0b RO 0b RO Bit Acronym Bit Description 15 DPE 14 SERRS 13 12 11 10 :09 08 Offset Start: 06h Offset End: 07h STA Reserved MDPE October 2012 Order Number: 327879-001US Sticky Signaled Target Abort: Not Implemented. Hardwired to `0'. Reserved 00b Master Data Parity Error: Not Implemented. Hardwired to `0'. 0b RO Intel(R) Communications Chipset 89xx Series - Datasheet 779 Table 17-5. Offset 06h: ID--STS--Status Register (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B:31:6 Default: 0010h Offset Start: 06h Offset End: 07h Power Well: Core Bit Reset Value Bit Access Fast Back to Back Capable: Does not apply. Hardwired to `0'. 0b RO Reserved 0b 66 MHz Capable: Does not apply. Hardwired to `0'. 0b RO Bit Acronym Bit Description Sticky 07 FBC 06 Reserved 05 C66 04 CLIST Capabilities List Exists: Indicates the controller contains a capabilities pointer list. The first item is pointed to by looking at configuration offset 34h. 1b RO IS Interrupt Status: Reflects the state of the INTx# signal at the input of the enable/disable circuit. This bit is a 1 when the INTx# is asserted. This bit is a 0 after the interrupt is cleared (independent of the state of the Interrupt Disable bit in the command register). This bit is not set if MSI is enabled. 0b RO 03 2 :0 17.2.1.5 Reserved Reserved 000b Offset 08h: RID--Revision Identification Table 17-6. Offset 08h: RID--Revision Identification Register Description: View: PCI Size: 8 bit Bit Range 07 :00 BAR: Configuration Bus:Device:Function: B:31:6 Default: 00h Power Well: Core Bit Acronym Bit Description RID Revision ID: Indicates the device specific revision identifier. NOTE: The value reported in this register depends on the value of the Revision ID in Device 31, Function 0, Offset 08h. See the LPC section for details. Intel(R) Communications Chipset 89xx Series - Datasheet 780 Offset Start: 08h Offset End: 08h Sticky Bit Reset Value Bit Access 00h RO October 2012 Order Number: 327879-001US 17.0 17.2.1.6 Offset 09h: PI--Programming Interface Table 17-7. Offset 09h: PI--Programming Interface Register Description: View: PCI Size: 8 bit BAR: Configuration Default: 00h Bit Range Bit Acronym 07 :00 PI 17.2.1.7 Bus:Device:Function: B:31:6 Offset Start: 09h Offset End: 09h Power Well: Core Bit Description Sticky Programming Interface -- PCH Thermal logic has no standard programming interface. Bit Reset Value Bit Access 00h RO Offset 0Ah: SCC--Sub Class Code Table 17-8. Offset 0Ah: SCC--Sub Class Code Register Description: View: PCI Size: 8 bit BAR: Configuration Default: 80h Bit Range Bit Acronym 07 :00 SCC 17.2.1.8 Bus:Device:Function: B:31:6 Offset Start: 0Ah Offset End: 0Ah Power Well: Core Bit Description Sticky Sub Class Code -- 8-bit value assigned to PCH Thermal logic. Bit Reset Value Bit Access 80h RO Offset 0Bh: BCC--Base Class Code Table 17-9. Offset 0Bh: BCC--Base Class Code Register Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 11h Bit Range Bit Acronym 07 :00 BCC October 2012 Order Number: 327879-001US Offset Start: 0Bh Offset End: 0Bh Power Well: Core Bit Description Sticky Base Class Code -- 8-bit value assigned to PCH Thermal logic Bit Reset Value Bit Access 11h RO Intel(R) Communications Chipset 89xx Series - Datasheet 781 17.2.1.9 Offset 0Ch: CLS--Cache Line Size Table 17-10. Offset 0Ch: Cache Line Size Description: View: PCI Size: 8 bit BAR: Configuration Default: 00h Bit Range Bit Acronym 07 :00 CLS 17.2.1.10 Bus:Device:Function: B:31:6 Offset Start: 0Ch Offset End: 0Ch Power Well: Core Bit Description Sticky Cache Line Size -- Does not apply to PCI Bus Target-only devices. Bit Reset Value Bit Access 00h RO Offset 0Dh: LT--Latency Timer Table 17-11. Offset 0Dh: Latency Timer Description: View: PCI Size: 8 bit BAR: Configuration Default: 00h Bit Range Bit Acronym 07 :00 LT 17.2.1.11 Bus:Device:Function: B:31:6 Offset Start: 0Dh Offset End: 0Dh Power Well: Core Bit Description Sticky Latency Timer -- Does not apply to PCI Bus Target-only devices. Bit Reset Value Bit Access 00h RO Offset 0Eh: HTYPE--Header Type Table 17-12. Offset 0Eh: Header Type Description: View: PCI Size: 8 bit Bit Range :07 06 :00 BAR: Configuration Bus:Device:Function: B:31:6 Default: 00h Power Well: Core Bit Reset Value Bit Access Multi-Function Device -- This bit is 0 because a multifunction device only needs to be marked as such in Function 0, and the Thermal registers are not in Function 0. 0b RO Header Type -- Implements Type 0 Configuration header. 00h RO Bit Acronym Bit Description MFD HTYPE Intel(R) Communications Chipset 89xx Series - Datasheet 782 Offset Start: 0Eh Offset End: 0Eh Sticky October 2012 Order Number: 327879-001US 17.0 17.2.1.12 Offset 10h: TBAR--Thermal Base This BAR creates 4K bytes of memory space to signify the base address of Thermal memory mapped configuration registers. This memory space is active when the Command (CMD) register Memory Space Enable (MSE) bit is set and either TBAR[31:12] or TBARH are programmed to a non-zero address. This BAR is owned by the operating system, and allows the OS to locate the Thermal registers in system memory space. Table 17-13. Offset 10h: Thermal Base Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000004h Bit Range Bit Acronym 31 :12 TBA 11 :04 Reserved 03 02 :01 00 17.2.1.13 Bus:Device:Function: B:31:6 Power Well: Core Bit Description Sticky Thermal Base Address -- This field provides the base address for the Thermal logic memory mapped configuration registers. 4 KB bytes are requested by hardwiring bits 11:4 to 0s. Reserved ADDRNG SPTYP Bit Reset Value Bit Access 00000h RW 00h Prefetchable -- Indicates that this BAR is NOT prefetchable. PREF Offset Start: 10h Offset End: 13h Address Range -- Indicates that this BAR can be located anywhere in 64 bit address space. Space Type -- Indicates that this BAR is located in memory space. 0b RO 10b RO 0b RO Offset 14h: TBARH--Thermal Base High DWord This BAR extension holds the high 32 bits of the 64 bit TBAR. In conjunction with TBAR, it creates 4 KB of memory space to signify the base address of thermal memory mapped configuration registers. Table 17-14. Offset 14h: TBARH--Thermal Base High DWord Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 00000000h Bit Range Bit Acronym 31 :00 TBAH October 2012 Order Number: 327879-001US Offset Start: 14h Offset End: 17h Power Well: Core Bit Description Sticky Thermal Base Address High -- TBAR bits 61:32. Bit Reset Value Bit Access 00000000h RW Intel(R) Communications Chipset 89xx Series - Datasheet 783 17.2.1.14 Offset 2Ch: SVID--Subsystem Vendor ID This register should be implemented for any function that could be instantiated more than once in a given system,. The SVID register, in combination with the Subsystem ID register, enables the operating environment to distinguish one subsystem from the other(s). Software (BIOS) will write the value to this register. After that, the value can be read, but writes to the register will have no effect. The write to this register should be combined with the write to the SID to create one 32-bit write. This register is not affected by D3HOT to D0 reset. Table 17-15. Offset 2Ch: Subsystem Vendor ID Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 0000h Bit Range Bit Acronym 15 :00 SVID Power Well: Core Bit Description SVID -- These RWO bits have no PCH functionality. Intel(R) Communications Chipset 89xx Series - Datasheet 784 Offset Start: 2Ch Offset End: 2Dh Sticky Bit Reset Value Bit Access 0000h RWO October 2012 Order Number: 327879-001US 17.0 17.2.1.15 Offset 2Eh: SID--Subsystem ID This register should be implemented for any function that could be instantiated more than once in a given system. The SID register, in combination with the Subsystem Vendor ID register make it possible for the operating environment to distinguish one subsystem from the other(s). Software (BIOS) will write the value to this register. After that, the value can be read, but writes to the register will have no effect. The write to this register should be combined with the write to the SVID to create one 32-bit write. This register is not affected by D3HOT to D0 reset. Table 17-16. Offset 2Eh: Subsystem ID Description: View: PCI Size: 16 bit BAR: Configuration Default: 0000h Bit Range Bit Acronym 15 :00 SID 17.2.1.16 Bus:Device:Function: B:31:6 Offset Start: 2Eh Offset End: 2Fh Power Well: Core Bit Description Sticky SID -- These RWO bits have no PCH functionality. Bit Reset Value Bit Access 0000h RWO Offset 34h: CP--Capabilities Pointer Table 17-17. Offset 34h: Capabilities Pointer Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 50h Bit Range Bit Acronym 07 :00 CP October 2012 Order Number: 327879-001US Offset Start: 34h Offset End: 34h Power Well: Core Bit Description Sticky Capability Pointer -- Indicates that the first capability pointer offset is offset 50h (Power Management Capability). Bit Reset Value Bit Access 50h RO Intel(R) Communications Chipset 89xx Series - Datasheet 785 17.2.1.17 Offset 3Ch: INTLN--Interrupt Line Table 17-18. Offset 3Ch: Interrupt Line Description: View: PCI Size: 8 bit BAR: Configuration Default: 00h Bit Range Bit Acronym 07 :00 INTLN 17.2.1.18 Bus:Device:Function: B:31:6 Offset Start: 3Ch Offset End: 3Ch Power Well: Core Bit Description Sticky Interrupt Line -- PCH hardware does not use this field directly. It is used to communicate to software the interrupt line that the interrupt pin is connected to. Bit Reset Value Bit Access 00h RW Offset 3Dh: INTPN--Interrupt Pin Table 17-19. Offset 3Dh: Interrupt Pin Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: See Description Bit Range Bit Acronym 07 :04 Reserved 03 :00 INTPN Power Well: Core Bit Description Reserved Interrupt Pin: This reflects the value of the Device 31 interrupt pin bits 27:24 (TTIP) in chipset configuration space. Intel(R) Communications Chipset 89xx Series - Datasheet 786 Offset Start: 3Dh Offset End: 3Dh Sticky Bit Reset Value Bit Access 0h Desc RO October 2012 Order Number: 327879-001US 17.0 17.2.1.19 Offset 40h: TBARB--BIOS Assigned Thermal Base Address This BAR creates 4K bytes of memory space to signify the base address of Thermal memory mapped configuration registers. This memory space is active when TBARB.SPTYPEN is asserted. This BAR is owned by the BIOS, and allows the BIOS to locate the Thermal registers in system memory space. If both TBAR and TBARB are programmed, then the OS and BIOS each have their own independent "view" of the Thermal registers, and must use the TSIU register to denote Thermal registers ownership/availability. Note: Transactions to TBARB are not affected by BME or MSE settings, only SPTYPEN. In other words, memory reads and writes to TBARB, when SPTYPEN (bit[0]) is `1', are claimed whether MSE is a `1' or `0' or whether the device is in D3 or D0. Table 17-20. Offset 40h: BIOS Assigned Thermal Base Address Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000004h Bit Range Bit Acronym 31 :12 TBA 11 :04 Reserved 03 02 :01 00 17.2.1.20 Bus:Device:Function: B:31:6 Power Well: Core Bit Description Sticky Thermal Base Address -- This field provides the base address for the Thermal logic memory mapped configuration registers. 4K B bytes are requested by hardwiring bits 11:4 to 0s. Reserved Bit Reset Value Bit Access 00000h RW 00h Prefetchable -- Indicates that this BAR is NOT prefetchable. PREF Offset Start: 40h Offset End: 43h ADDRNG Address Range -- Indicates that this BAR can be located anywhere in 64 bit address space. SPTYPEN Space Type Enable: 0 = Disable. 1 = Enable. When set to 1b by software, enables the decode of this memory BAR. 0b RO 10b RO 0b RW Offset 44h: TBARBH--BIOS Assigned Thermal Base High DWord Table 17-21. Offset 44h: BIOS Assigned Thermal Base High DWord Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 00000000h Bit Range Bit Acronym 31 :00 TBAH October 2012 Order Number: 327879-001US Offset Start: 44h Offset End: 47h Power Well: Core Bit Description Sticky Thermal Base Address High: TBAR bits 61:32. Bit Reset Value Bit Access 00000000h RW Intel(R) Communications Chipset 89xx Series - Datasheet 787 17.2.1.21 Offset 50h: PID--PCI Power Management Capability ID Table 17-22. Offset 50h: PCI Power Management Capability ID Description: View: PCI Size: 16 bit BAR: Configuration Default: 8001h Bit Range Bit Acronym 15 :08 NEXT 07 :00 CAP 17.2.1.22 Bus:Device:Function: B:31:6 Offset Start: 50h Offset End: 51h Power Well: Core Bit Reset Value Bit Access Next Capability -- Indicates that the next capability is MSI. 80h RO Capability ID -- Indicates that this pointer is a PCI power management capability 01h RO Bit Description Sticky Offset 52h: PC--Power Management Capabilities Table 17-23. Offset 52h: Power Management Capabilities Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B:31:6 Default: 0023h Bit Acronym 15 :11 Power Well: Core Bit Description PME_Support -- Indicates PME# is not supported Sticky Bit Reset Value Bit Access 00h RO 10 D2_Support -- The D2 state is not supported. 0b RO 09 D1_Support -- The D1 state is not supported. 0b RO 000b RO Device Specific Initialization -- Indicates that devicespecific initialization is required. 1b RO Reserved 0b PME Clock -- Does not apply. Hardwired to 0. 0b RO 011b RO Aux_Current -- PME# from D3COLD state is not supported, therefore this field is 000b. 08 :06 05 DSI 04 Reserved 03 PMEC 02 :00 VS Version -- Indicates support for Revision 1.2 of the PCI Power Management Specification. Intel(R) Communications Chipset 89xx Series - Datasheet 788 Offset Start: 52h Offset End: 53h October 2012 Order Number: 327879-001US 17.0 17.2.1.23 Offset 54h: PCS--Power Management Control And Status Table 17-24. Offset 54h: Power Management Control And Status Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Default: 0008h Bit Acronym 31 :24 BPCCE 22 B23 15 14 :09 08 07 :04 Reserved PMES Reserved PMEE Reserved 03 02 01 :00 Reserved PS October 2012 Order Number: 327879-001US Offset Start: 54h Offset End: 57h Power Well: Core Bit Description Sticky Data -- Does not apply. Hardwired to 0s. 23 21 :16 Bus:Device:Function: B:31:6 Bit Reset Value Bit Access 00h RO Bus Power/Clock Control Enable -- Hardwired to 0. 0b RO B2/B3 Support --Does not apply. Hardwired to 0. 0b RO Reserved 00h PME Status -- This bit is always 0, since this PCI Function does not generate PME# Reserved 0b RO 00h PME Enable -- This bit is always zero, since this PCI Function does not generate PME# 0b Reserved 0h No Soft Reset --When set (`1'), this bit indicates that devices transitioning from D3HOT to D0 because of PowerState commands do not perform an internal reset. Configuration context is preserved. Upon transition from D3HOT to D0 initialized state, no additional operating system intervention is required to preserve Configuration Context beyond writing the PowerState bits. 1b Reserved 0b Power State -- This field is used both to determine the current power state of the Thermal controller and to set a new power state. The values are: 00 = D0 state 11 = D3HOT state If software attempts to write a value of 10b or 01b in to this field, the write operation must complete normally; however, the data is discarded and no state change occurs. When in the D3HOT states, the Thermal controller's configuration space is available, but the I/O and memory spaces are not. Additionally, interrupts are blocked. When software changes this value from the D3HOT state to the D0 state, no internal warm (soft) reset is generated. 00b RO RO RW Intel(R) Communications Chipset 89xx Series - Datasheet 789 17.2.1.24 Offset 80h: MID--Message Signaled Interrupt Identifiers Table 17-25. Offset 80h: MID--Message Signaled Interrupt Identifiers Description: View: PCI Size: 16 bit BAR: Configuration Default: 0005h Bit Range Bit Acronym 15 :08 NEXT 07 :00 CID 17.2.1.25 Bus:Device:Function: B:31:6 Offset Start: 80h Offset End: 81h Power Well: Core Bit Reset Value Bit Access Next Pointer -- Indicates this is the latest pointer 00h RO Capability ID -- Capabilities ID indicates MSI. 05h RO Bit Description Sticky Offset 82h: MC--Message Signaled Interrupt Message Control Table 17-26. Offset 82h: MC--Message Signaled Interrupt Message Control Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 0000h Bit Range Bit Acronym 15 :08 Reserved Power Well: Core Bit Description Reserved Sticky Bit Reset Value Bit Access 00h C64 64 Bit Address Capable -- Capable of generating a 32-bit message only. 0b RO 06 :04 MME Multiple Message Enable -- These bits are RW for software compatibility, but only one message is ever sent by the root port. 0h RW 03 :01 MMC Multiple Message Capable -- Only one message is required. 0h RO MSIE MSI Enable -- If set, MSI is enabled and traditional interrupt pins are not used to generate interrupts. CMD.BME must be set for an MSI to be generated. If CMD.BME is cleared, and this bit is set, no interrupts (not even pin based) are generated. 0b RO 07 00 Intel(R) Communications Chipset 89xx Series - Datasheet 790 Offset Start: 82h Offset End: 83h October 2012 Order Number: 327879-001US 17.0 17.2.1.26 Offset 84h: MA--Message Signaled Interrupt Message Address Table 17-27. Offset 84h: MA--Message Signaled Interrupt Message Address Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 00000000h Power Well: Core Bit Range Bit Acronym Bit Description 31 :02 ADDR Address -- Lower 32 bits of the system specified message address, always DW aligned. 01 :00 Reserved 17.2.1.27 Offset Start: 84h Offset End: 87h Sticky Bit Reset Value Bit Access 0000000h RW Reserved 00b Offset 88h: MD--Message Signaled Interrupt Message Data Table 17-28. Offset 88h: MD--Message Signaled Interrupt Message Data Description: View: PCI Size: 16 bit BAR: Configuration Default: 0000h Bit Range Bit Acronym 15 :00 DATA 17.2.1.28 Bus:Device:Function: B:31:6 Offset Start: 88h Offset End: 89h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0000h RW Data -- This 16-bit field is programmed by system software if MSI is enabled. Its content is driven onto the lower word (PCI AD[15:0]) during the data phase of the MSI memory write transaction. Offset 94h: CCTS0RD--Thermal Sensor 0 Remote Diode Table 17-29. Offset 94h: CCTS0RD--Thermal Sensor 0 Remote Diode Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 00h Bit Range Bit Acronym 07 :01 Reserved 00 October 2012 Order Number: 327879-001US Offset Start: 94h Offset End: 94h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00h Enable remote diodes operation. NOTE: There will be a 2.5ms delay from transition of isnsrmten to delivery of valid data on Offset A8h: CCTS0TL1 - Thermal Sensor 0 Top Logic1 otstemperature[8:0]. 0b RW Intel(R) Communications Chipset 89xx Series - Datasheet 791 17.2.1.29 Offset 96h: CCTS0SSI--Thermal Sensor 0 Sense Stage Inputs Table 17-30. Offset 96h: CCTS0SSI--Thermal Sensor 0 Sense Stage Inputs Description: View: PCI Size: 16 bit Bit Range 15 BAR: Configuration Bus:Device:Function: B:31:6 Default: 0066h Bit Acronym Reserved Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved 0b 14 isnsinternalvrefen -- Enable using an internal reference voltage generated by a resistor divider, instead of using external reference voltage 0b RW 13 isnswastesel -- Select the waste current target: 0 = Current from non-used current mirrors flows to the waste diode 1 = Current from non-used current mirrors flows disable 0b RW isnschopsel[4:0] -- Sense stage current ratio configuration for DEM and chopping (See the Sense stage current ratio configuration) 00h RW 12 :08 07 isnsextcuren -- Enable driving external bias current 0b RW 06 isnsbiasampen -- Sense stage current bias amplifier enable 1b RW isnscurrentsel[2:0] -- Current value select for 4 current sources within the sense stage (See the Sense stage bias current select) 100b RW isnsextresen -- Enable external resistor use for accurate reference current generation 1b RW isnsen (See the Sense stage enable signal) 1b RW Reserved 0b 05 :03 02 01 00 Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 792 Offset Start: 96h Offset End: 97h October 2012 Order Number: 327879-001US 17.0 17.2.1.30 Offset 98h: CCTS0DSAC0--Thermal Sensor 0 Delta Sigma ADC Control 0 Table 17-31. Offset 98h: CCTS0DSAC0--Thermal Sensor 0 Delta Sigma ADC Control 0 Description: View: PCI Size: 32bit Bit Range BAR: Configuration Bus:Device:Function: B:31:6 Default: 02010083h Bit Acronym Offset Start: 98h Offset End: 9Bh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 31 :16 idstiming[15:0] -- Delta Sigma phase timing control (See the Delta Sigma Control and Timing Configuration) 0201h RW 15 :00 idscontrol[15:0] -- Delta Sigma features control (See the Delta Sigma Control and Timing Configuration) 0083 RW 17.2.1.31 Offset 9Ch: CCTS0DSAC 1--Thermal Sensor 0 Delta Sigma ADC Control 1 Table 17-32. Offset 9Ch: CCTS0DSAC1--Thermal Sensor 0 Delta Sigma ADC Control 1 Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 0001h Bit Range Bit Acronym 15 :02 Reserved Offset Start: 9Ch Offset End: 9Dh Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 0000h 01 idshighgain: 0 - default bias current for SD ADC 1 = reduce by 2 bias current for SD ADC.; Increase the amplifier gain, reduce current consumption and bandwidth 0b RW 00 idsen -- DS ADC enable 1b RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 793 17.2.1.32 Offset A0h: CCTS0C--Thermal Sensor 0 Calibration Table 17-33. Offset A0h: CCTS0C--Thermal Sensor 0 Calibration Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 00030016h Bit Range Bit Acronym 31 :18 Reserved Offset Start: A0h Offset End: A3h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 0000h 17 :16 icalconfigsel[1:0] -- Controls for calibration: 00: linearization only 01: do both calibration and linearization calculations 10: do not calibrate and do not linearization 11: calibration calculation only 11b RW 15 :08 isparectrl[7:0] -- Spare bits (See the Spare pin functionality) 00h RW 07 :00 icalcoarsetune[7:0] -- Calibration bits for coarse calibration range selection 16h RW 17.2.1.33 Offset A4h: CCTS0TL0--Thermal Sensor 0 Top Logic 0 Table 17-34. Offset A4h: CCTS0TL0--Thermal Sensor 0 Top Logic 0 Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Default: 80000000h Reserved 27 26 :24 Power Well: Core Bit Access Itsen -- Enable thermal sensor Hard-IP when set to 1 1b RW Reserved 0b Bit Description itsclksel[1:0]: 00 = 100 Mhz, 01 = Reserved 10 = Reserved 11 = 400 Mhz 29 :28 Sticky 00b Reserved Reserved 0b Reserved Reserved 000b Intel(R) Communications Chipset 89xx Series - Datasheet 794 Offset Start: A4h Offset End: A7h Bit Reset Value Bit Acronym 31 30 Bus:Device:Function: B:31:6 RW October 2012 Order Number: 327879-001US 17.0 Table 17-34. Offset A4h: CCTS0TL0--Thermal Sensor 0 Top Logic 0 Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Bus:Device:Function: B:31:6 Default: 80000000h Bit Acronym 23 :16 Power Well: Core Bit Description Sticky itsalertthreshoff[7:0] -- (Hot value) When temperature reaches the value indicated by this bus, alert signal (otstemperaturealrt) is turned off. * Default value has been set for 103C * Calculated value is (itsalertthreshoff [7:0] - 50C) Note: Offset Start: A4h Offset End: A7h Bit Reset Value Bit Access 99h RW 98h RW A2h RW For the PCH these bits are not used and have been replaced with "Offset 0Bh: TS0TTPC-- Thermal Sensor 0 Temperature Trip Point Cont'd" on page 810 bits 7:0 itsalertthreshon[7:0] -- (Hot value) When temperature reaches the value indicated by this bus, alert signal (otstemperaturealrt) is turned on (See the Parallel Temperature Diagram). * Default value has been set for 105C * Calculated value is (itsalertthreshon[7:0] - 50C) 15 :08 Note: Thermal protection logic associated with this field will always default to A2h = 112C whenever this field is programmed to a value greater than A2h due to accidental or malicious programming. Note: For the PCH these bits are not used and have been replaced with "Offset 04h: TS0TTP--Thermal Sensor 0 Temperature Trip Point" on page 808 bits 15:8 itstripthreshon[7:0] -- (Catastrophic value) When temperature reaches the value indicated by this bus, trip signal (otstemperaturetrip) is turned on (See the Parallel Temperature Diagram). * Default value has been set for 112C * Calculated value is (itstripthreshon[7:0] - 50C) 07 :00 October 2012 Order Number: 327879-001US Note: Thermal protection logic associated with this field will always default to A2h = 112C whenever this field is programmed to a value greater than A2h, due to accidental or malicious programming. Note: For the PCH these bits are not used and have been replaced with "Offset 04h: TS0TTP--Thermal Sensor 0 Temperature Trip Point" on page 808 bits 7:0 Intel(R) Communications Chipset 89xx Series - Datasheet 795 17.2.1.34 Offset A8h: CCTS0TL1--Thermal Sensor 0 Top Logic 1 Table 17-35. Offset A8h: CCTS0TL1--Thermal Sensor 0 Top Logic 1 Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Bus:Device:Function: B:31:6 Default: See Description Bit Acronym Offset Start: A8h Offset End: ABh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 00b RW RW 31 :30 itsfilterrate[1:0] -- Choose the CIC filter decimation factor: 00 = Decimate by 64 01 = Decimate by 128 10 = Decimate by 128 11 = Decimate by 256 NOTE: The FIR has a built-in decimation factor of 4, so if the FIR is not bypassed, the total decimation factor is 4 times the value given by itsfilterrate[1:0] 29 :28 itsfiltersel[1:0] -- Select the digital filter mode: 00 = Use CIC followed by FIR 01 = CIC only (bypass the FIR) 10 - Reserved 11 = Reserved 00 Reserved 0b 27 Reserved 26 :24 itsfilterrangesel[2:0] -- Range select for filter output bits according to the gain of the TS analog circuit (See the Range select for filter output) 000b RW 23 :12 itsgammacoeff[11:0] -- Linearization coefficient in Ratiometric mode only 0AFh RW 11 :10 itsoutputbitsel[1:0] -- Resolution selection of temperature outputs (otstemperature[8:0]) from calculated temperature: 00 - default (use the default only) 01 = 1 bit shift left 10 = 1 bit shift right 11 = 2 bit shift left 00b RW 09 Reserved Reserved 0b otstemperature[8:0] -- Temperature reading where 0 = -50oC, resolution of 0.5oC (See the Parallel Temperature Diagram). 08 :00 Note: Note: Default value is X it depends on temperature at power on This temperature at resolution of 1oC can also be read at MMIO location "Offset 03h: TS0TR-- Thermal Sensor 0 Thermometer Read" on page 808 bits 7:0 Intel(R) Communications Chipset 89xx Series - Datasheet 796 XXh RO October 2012 Order Number: 327879-001US 17.0 17.2.1.35 Offset ACh: CCTS0BP--Thermal Sensor 0 Bandgap Pins Table 17-36. Offset ACh: CCTS0BP--Thermal Sensor 0 Bandgap Pins Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 001Bh Bit Range Bit Acronym 15 :05 Reserved Offset Start: ACh Offset End: ADh Power Well: Core Bit Description Sticky Bit Reset Value Reserved Bit Access 000h 04 Ibgen -- Enable signal for the bandgap circuit block. A logic `1' enabled the BG circuit and a logic `0' disables (power-down) the circuit block 1b RW 03 Ibgchopen -- Enable signal for the bandgap circuit block chopping mode. A logic `1' enables the chopping mode and a logic `0' disables the chopping mode. A logic `0' disables all of the internal clock generation circuitry. 1b RW 02 Ibgfilterbypassen -- Select signal for the bandgap circuit block chop clock: 0 - Working with LPF 1 = Bypass LPF 0b RW 11b RW ibgclksource[1:0]: 00 = DS bit stream 01 - reserve 10 = reserve 11 - internal bandgap clock 01 :00 17.2.1.36 Offset B4h: CCTS1RD--Thermal Sensor 1 Remote Diode Table 17-37. Offset B4h: CCTS1RD--Thermal Sensor 1 Remote Diode Description: View: PCI Size: 8 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 0000h Bit Range Bit Acronym 07 :01 Reserved 00 October 2012 Order Number: 327879-001US Offset Start: B4h Offset End: B4h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 0000000b isnsrmten -- Enable remote diodes operation. NOTE: There will be a 2.5msec delay from transition of isnsrmten to delivery of valid data on Offset A8h: CCTS0TL1 - Thermal Sensor 0 Top Logic1 otstemperature[8:0]. 0b RW Intel(R) Communications Chipset 89xx Series - Datasheet 797 17.2.1.37 Offset B6h: CCTS1SSI--Thermal Sensor 1 Sense Stage Inputs Table 17-38. Offset B6h: CCTS1SSI--Thermal Sensor 1 Sense Stage Inputs Description: View: PCI Size: 16 bit Bit Range 15 BAR: Configuration Bus:Device:Function: B:31:6 Default: 0066h Bit Acronym Reserved Offset Start: B6h Offset End: B7h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved 0b 14 isnsinternalvrefen -- Enable using an internal reference voltage generated by a resistor divider instead of using external reference voltage 0b RW 13 isnswastesel -- Select the waste current target: 0 = Current from non-used current mirrors flows to the waste diode 1 = Current from non-used current mirrors flows disable 0b RW isnschopsel[4:0] -- Sense stage current ratio configuration for DEM and chopping (See the Sense stage current ratio configuration) 00000b RW 12 :08 07 isnsextcuren -- Enable driving external bias current 0b RW 06 isnsbiasampen -- Sense stage current bias amplifier enable 1b RW isnscurrentsel[2:0] -- Current value select for 4 current sources within the sense stage (See the Sense stage bias current select) 100b RW 02 isnsextresen -- Enable external resistor use for accurate reference current generation 1b RW 01 isnsen -- Sense stage enable signal 1b RW Reserved 0b 05 :03 00 17.2.1.38 Reserved Offset B8h: CCTS1DSAC0--Thermal Sensor 1 Delta Sigma ADC Control 0 Table 17-39. Offset B8h: CCTS1DSAC0--Thermal Sensor 1 Delta Sigma ADC Control 0 Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Bus:Device:Function: B:31:6 Default: 02010083h Bit Acronym Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 31 :16 idstiming[15:0] -- Delta Sigma phase timing control (See the Delta Sigma Control and Timing Configuration) 0201h RW 15 :00 idscontrol[15:0] -- Delta Sigma features control (See the Delta Sigma Control and Timing Configuration) 0083h RW Intel(R) Communications Chipset 89xx Series - Datasheet 798 Offset Start: B8h Offset End: BBh October 2012 Order Number: 327879-001US 17.0 17.2.1.39 Offset BCh: CCTS1DSAC1--Thermal Sensor 1 Delta Sigma ADC Control 1 Table 17-40. Offset BCh: CCTS1DSAC1--Thermal Sensor 1 Delta Sigma ADC Control 1 Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 0001h Bit Range Bit Acronym 15 :02 Reserved Offset Start: BCh Offset End: BDh Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 0000h 01 idshighgain: 0 = default bias current for SD ADC 1 = reduce by 2 bias current fro SD ADC. Increase the amplifier gain, reduce current consumption and bandwidth 0b RW 00 idsen -- DS ADC enable 1b RW 17.2.1.40 Offset C0h: CCTS1C--Thermal Sensor 1 Calibration Table 17-41. Offset C0h: CCTS1C--Thermal Sensor 1 Calibration Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 00030016h Bit Range Bit Acronym 31 :18 Reserved Offset Start: C0h Offset End: C3h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 0000h 17 :16 icalconfigsel[1:0] -- Controls for calibration: 00: linearization only 01: do both calibration and linearization calculations 10: do not calibrate and do not linearization 11: calibration calculation only 11b RW 15 :08 isparectrl[7:0] -- Spare bits (See the Spare pin functionality) 00h RW 07 :00 icalcoarsetune[7:0] -- Calibration bits for coarse calibration range selection 16h RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 799 17.2.1.41 Offset C4h: CCTS1TL0--Thermal Sensor 1 Top Logic 0 Table 17-42. Offset C4h: CCTS1TL0--Thermal Sensor 1 Top Logic 0 Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Default: 80999BAAh Reserved 27 26 :24 23 :16 Power Well: Core Bit Access Itsen -- Enable thermal sensor Hard-IP when set to 1 1b RW Reserved 0b Bit Description itsclksel[1:0]: 00 = 100 Mhz 01 = Reserved 10 = Reserved 11 = 400 Mhz 29 :28 Offset Start: C4h Offset End: C7h Bit Reset Value Bit Acronym 31 30 Bus:Device:Function: B:31:6 Sticky 00b RW RW Reserved Reserved 0b Reserved Reserved 000b itsalertthreshoff[7:0] -- (Hot value) When temperature reaches the value indicated by this bus, alert signal (otstemperaturealrt) is turned off. * Default value has been set for 103C * Calculated value is (itsalertthreshoff [7:0] - 50C) Note: 99h RW 9Bh RW A2h RW For the PCH these bits are not used and have been replaced with "Offset 4Bh: TS1TTPC-- Thermal Sensor 1 Temperature Trip Point Cont'd" on page 827 bits 7:0 itsalertthreshon[7:0] -- (Hot value) When temperature reaches the value indicated by this bus, alert signal (otstemperaturealrt) is turned on. * Default value has been set for 105C * Calculated value is (itsalertthreshon[7:0] - 50C) 15 :08 Note: Thermal protection logic associated with this field will always default to A2h = 112C whenever this field is programmed to a value greater than A2h due to accidental or malicious programming. Note: For the PCH these bits are not used and have been replaced with "Offset 44h: TS1TTP--Thermal Sensor 1 Temperature Trip Point" on page 824 bits 15:8 itstripthreshon[7:0] -- (Catastrophic value) When temperature reaches the value indicated by this bus, trip signal (otstemperaturetrip) is turned on. * Default value has been set for 112C * Calculated value is (itstripthreshon[7:0] - 50C) 07 :00 Note: Thermal protection logic associated with this field will always default to A2h = 112C whenever this field is programmed to a value greater than A2h, due to accidental or malicious programming. Note: For the PCH these bits are not used and have been replaced with "Offset 44h: TS1TTP--Thermal Sensor 1 Temperature Trip Point" on page 824 bits 7:0 Intel(R) Communications Chipset 89xx Series - Datasheet 800 October 2012 Order Number: 327879-001US 17.0 17.2.1.42 Offset C8h: CCTS1TL1--Thermal Sensor 1 Top Logic 1 Table 17-43. Offset C8h: CCTS1TL1--Thermal Sensor 1 Top Logic 1 Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: See Description Offset Start: C8h Offset End: CBh Power Well: Core Bit Reset Value Bit Access 31 :30 itsfilterrate[1:0] -- Choose the CIC filter decimation factor: 00 = Decimate by 64 01 = Decimate by 128 10 = Decimate by 128 11 = Decimate by 256 NOTE: The FIR has a built-in decimation factor of 4, so if the FIR is not bypassed, the total decimation factor is 4 times the value given by itsfilterrate[1:0] 00b RW 29 :28 itsfiltersel[1:0] -- Select the digital filter mode: 00 = Use CIC followed by FIR 01 = CIC only (bypass the FIR) 10 = Reserved 11 = Reserved 00b RW Bit Range 27 Bit Acronym Reserved Bit Description Sticky Reserved 0b 26 :24 itsfilterrangesel[2:0] -- Range select for filter output bits according to the gain of the TS analog circuit 000b RW 23 :12 itsgammacoeff[11:0] -- Linearization coefficient in Ratiometric mode only 0AFh RW 11 :10 itsoutputbitsel[1:0] -- Resolution selection of temperature outputs (otstemperature[8:0]) from calculated temperature: 00 = default (use the default only) 01 = 1 bit shift left 10 = 1 bit shift right 11 = 2 bit shift left 00b RW 09 Reserved 08 :00 October 2012 Order Number: 327879-001US Reserved 0b otstemperature[8:0] -- Temperature reading, where 0 = 50 oC, resolution of 0.5 oC. NOTE: Default value is X it depends on temperature at power on X RO Intel(R) Communications Chipset 89xx Series - Datasheet 801 17.2.1.43 Offset CCh: CCTS1BP--Thermal Sensor 1 Bandgap Pins Table 17-44. Offset CCh: CCTS1BP--Thermal Sensor 1 Bandgap Pins Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B:31:6 Default: 001Bh Bit Range Bit Acronym 15 :05 Reserved Offset Start: CCh Offset End: CDh Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 000h 04 Ibgen -- Enable signal for the bandgap circuit block. A logic 1 enabled the BG circuit and a logic 0 disables (power-down) the circuit block 1b RW 03 Ibgchopen -- Enable signal for the bandgap circuit block chopping mode. A logic `1' enables the chopping mode and a logic `0' disables the chopping mode. A logic `0' disables all of the internal clock generation circuitry. 1b RW 02 Ibgfilterbypassen -- Select signal for the bandgap circuit block chop clock.: 0 - Working with LPF 1 - Bypass LPF 0b RW 11b RW ibgclksource[1:0]: 00 = DS bit stream 01 = reserve 10 = reserve 11 = internal bandgap clock 01 :00 17.2.1.44 Offset F0h: Reserved 17.2.1.45 Offset F8h: MANID--Manufacturing/Process Table 17-45. Offset F8h: MANID--Manufacturing/Process Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000F00h Bit Range Bit Acronym 31 :24 Reserved 23 :16 15 :00 Bus:Device:Function: B:31:6 Reserved Power Well: Core Bit Description Sticky Bit Reset Value Reserved 00h Stepping ID -- This field is incremented for each stepping of the part. Note: This field can be used by software to differentiate steppings when the Revision ID may not change. Implementation Note: A single Stepping ID can be implemented that is readable from all functions in the chip because all of them are incremented in lock-step. 00h Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 802 Offset Start: F8h Offset End: FBh Bit Access RO 0F00h October 2012 Order Number: 327879-001US 17.0 17.2.2 PCH Thermal Reporting The Thermal Reporting Registers are located in memory space. The base memory address for the registers is specified at TBARB (B0:D31:F6: Offset 40h - 43h). The individual registers are accessible at TBARB + Offset. There are two Thermal Sensors. Registers for Thermal Sensor 0 (TS0) starts at offset 00h and Thermal Sensor 1 (TS1) starts at offset 40h The Manageability Engine (ME) has both read and write access to the thermal sensor offsets 00h-FFh. See the ME C-spec for additional details. The register layout follows with only the lower 9 bits of the offset listed. In the table the new registers have the access requirements for host and ME listed. The first provides host access capabilities, and the second is ME access. 17.2.2.1 DIMM Thermal Reporting Configurations The PCH collects DIMM thermal data via the PECI or Host SMBus depending on the type of CPU on the platform. Platforms based on CPUs with integrated DIMM SMBuses, the PCH collects DIMM thermal data via the PECI bus. An example CPU with integrated DIMM SMBuses is Intel(R) Xeon(R) Processor E5-2600 family. Platforms based on CPUs without integrated DIMM SMBuses, the PCH collects DIMM thermal data via the platform Host SMBus. An example CPU without integrated DIMM SMBuses is Intel(R) Xeon(R) and Intel(R) CoreTM Processors For Communications Infrastructure. In this document, the following references shall be used: * Platforms based on CPUs with integrated DIMM SMBuses shall be referred to as PECI-to-DIMM Processor (PDP) based platform since the PCH uses the PECI bus to access DIMM thermal data. * Platforms based on CPUs without integrated DIMM SMBuses shall be referred to as SMBus-to-DIMM Processor (SDP) based platform since the PCH uses the platform internal Host SMBus to access DIMM thermal data. Note: The PCH ME FW determines the type of platform and the DIMM access port via a SoftStrap setting in the SPI Descriptor. Note: Refer to SMBus-to-PECI Bridge Protocol Application Note (Doc# 460415) for more information. 17.2.3 Thermal Reporting Registers Table 17-46. Bus 0, Device 31, Function 6: Summary of Thermal Reporting Registers Mapped Through TBARB BAR (Sheet 1 of 3) Offset Start Offset End 00h 00h Register ID - Description "Offset 00h: TS0IU--Thermal Sensor 0 In Use" on page 805 Default Value 00h 01h 01h "Offset 01h: TS0C--Thermal Sensor 0 Control" on page 806 00h 02h 02h "Offset 02h: TS0S--Thermal Sensor 0 Status" on page 807 00h 03h 03h "Offset 03h: TS0TR--Thermal Sensor 0 Thermometer Read" on page 808 XXh October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 803 Table 17-46. Bus 0, Device 31, Function 6: Summary of Thermal Reporting Registers Mapped Through TBARB BAR (Sheet 2 of 3) Offset Start Offset End Register ID - Description Default Value 04h 07h "Offset 04h: TS0TTP--Thermal Sensor 0 Temperature Trip Point" on page 808 00000000h 08h 08h "Offset 08h: TS0CO--Thermal Sensor 0 Calibration Offset" on page 810 00h 0Bh 0Bh "Offset 0Bh: TS0TTPC--Thermal Sensor 0 Temperature Trip Point Cont'd" on page 810 00h 0Ch 0Ch "Offset 0Ch: TS0ES--Thermal Sensor 0 Error Status" on page 811 00h 0Dh 0Dh "Offset 0Dh: TS0GPEN--Thermal Sensor 0 General Purpose Event Enables" on page 812 00h 0Eh 0Eh "Offset 0Eh: TS0PC--Thermal Sensor 0 Policy Control" on page 813 00h 10h 11h "Offset 10h: CPEC--CPU Power Error Correction Data" on page 814 00h 12h 13h "Offset 12h: C0TA--CPU0 Temperature Adjust" on page 814 00h 16h 17h "Offset 16h: C1TA--CPU1Temperature Adjust " on page 815 0000h 1Ah 1Bh "Offset 1Ah: MC--MPC Control" on page 815 0000h 22h 23h "Offset 22h: Timestamp" on page 816 0000h 24h 27h "Offset 24h: DIMMEN--DIMM Enable" on page 817 0000h 30h 31h "Offset 30h: C0TV--CPU0 Temperature Value" on page 818 0000h 32h 33h "Offset 32h: C1TV--CPU1 Temperature Value" on page 819 0000h 34h 37h "Offset 34h: C0EV--CPU0 Energy Value" on page 819 00000000h 38h 3Bh "Offset 38h: C1EV--CPU1 Energy Value" on page 819 00000000h 3Fh 3Fh "Offset 3Fh: AE--Alert Enable" on page 820 00h 40h 40h "Offset 40h: TS1IU--Thermal Sensor 1 In Use" on page 821 00h 41h 41h "Offset 41h: TS1C--Thermal Sensor 1 Control" on page 821 00h 42h 42h "Offset 42h: TS1S--Thermal Sensor 1 Status" on page 823 00h 43h 43h "Offset 43h: TS1TR--Thermal Sensor 1 Thermometer Read" on page 824 XXh 44h 47h "Offset 44h: TS1TTP--Thermal Sensor 1 Temperature Trip Point" on page 824 00000000h 48h 48h "Offset 48h: TS1CO--Thermal Sensor 1 Calibration Offset" on page 826 00h 4Bh 4Bh "Offset 4Bh: TS1TTPC--Thermal Sensor 1 Temperature Trip Point Cont'd" on page 827 00h 4Ch 4Ch "Offset 4Ch: TS1ES--Thermal Sensor 1 Error Status" on page 827 00h 4Dh 4Dh "Offset 4Dh: TS1GPEN--Thermal Sensor 1 General Purpose Event Enables" on page 829 00h 4Eh 4Eh "Offset 4Eh: TS1PC--Thermal Sensor 1 Policy Control" on page 830 00h 50h 55h "Offset 50h: HTS--HOST Turbo Status " on page 831 000000000000h 56h 57h "Offset 56h: MTL--MCP Temperature Limit" on page 831 0000h 58h 5Fh "Offset 58h: MTV--MCH Temperature Value" on page 832 00000000000000 00h 60h 61h "Offset 60h: MCPTV--MCP Temperature Value" on page 832 0000h 64h 65h "Offset 64h: MMPC--Max MCH Power Clamp" on page 832 0000h 66h 67h "Offset 66h: MMCPPC--Max MCP Power Clamp" on page 833 0000h 82h 82h "Offset 82h: TS0PIEN--Thermal Sensor 0 PCI Interrupt Event Enables" on page 833 00h 83h 83h "Offset 83h: TS0LOCK--Thermal Sensor 0 Register Lock Controls" on page 834 00h 98h 9Bh "Offset 98h: STS--SMBus Turbo Status" on page 834 00000000h Intel(R) Communications Chipset 89xx Series - Datasheet 804 October 2012 Order Number: 327879-001US 17.0 Table 17-46. Bus 0, Device 31, Function 6: Summary of Thermal Reporting Registers Mapped Through TBARB BAR (Sheet 3 of 3) Offset Start Offset End Register ID - Description Default Value 9Ch 9Ch "Offset 9Ch: SEC--SMBus Event Clear" on page 835 00h A4h A7h "Offset A4h: TC3--Thermal Compares 3" on page 835 00000000h A8h ABh "Offset A8h: TC1--Thermal Compares 1" on page 836 00000000h ACh AFh "Offset ACh: TC2--Thermal Compares 2" on page 836 00000000h B0h B3h "Offset B0h: DIMM0--CPU0 DIMM Data" on page 837 00000000h B4h B7h "Offset B4h: DIMM1--CPU1 DIMM Data" on page 839 00000000h B8h BBh "Offset B8h: DIMMID--DIMM ID" on page 840 00000000h BCh BFh "Offset BCh: ECPCLAMP--EC Power Clamp Data" on page 842 00000000h C2h C2h "Offset C2h: TS1PIEN--Thermal Sensor 1PCI Interrupt Event Enables" on page 842 00h C3h C3h "Offset C3h: TS1LOCK--Thermal Sensor 1 Register Lock Controls" on page 843 00h D8h DBh "Offset D8h: ITV--Internal Temperature Values" on page 843 00000000h Many of the following registers are implemented twice, once for each thermal sensor. 17.2.3.1 Offset 00h: TS0IU--Thermal Sensor 0 In Use Table 17-47. Offset 00h: TS0IU--Thermal Sensor 0 In Use Description: View: PCI Size: 8 bit BAR: TBARB Default: 00h Bit Range Bit Acronym 07 :01 Reserved 00 Bus:Device:Function: B:31:6 TS0IU Offset Start: 00h Offset End: 00h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00h Thermal Sensor In Use -- Software semaphore bit. After a core-well reset, a read to this bit returns a 0. After the first read, subsequent reads will return a 1. A write of a 1 to this bit will reset the next read value to 0. Writing a 0 to this bit has no effect. Software can poll this bit until it reads a 0, and will then own the usage of the thermal sensor. This bit has no other effect on the hardware, and is only used as a semaphore among various independent software threads that may need to use the thermal sensor. Software that reads this register but does not intend to claim exclusive access of the thermal sensor must write a one to this bit if it reads a 0, in order to allow other software threads to claim it. 0b RS/WC NOTE: The ME only reads the thermal interface registers and as such ME reads to TSIU will not set the bit. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 805 17.2.3.2 Offset 01h: TS0C--Thermal Sensor 0 Control This register controls the operation of the thermal sensor. Table 17-48. Offset 01h: TS0C--Thermal Sensor 0 Control (Sheet 1 of 2) Description: View: PCI Size: 8 bit Bit Range BAR: TBARB Default: 00h TS0E 06 Reserved DHA0 Bit Reset Value Bit Access Thermal Sensor enable -- This bit enables thermal reporting. Lockable via TSCO bit 7: 0 -Disabled 1 -Enabled 0b RWL Reserved 0b RW 00b RW 00b RO Bit Description Digital Hysteresis Amount: This bit determines whether no offset, 1.0C, 2.0C, or 3.0C is used for hysteresis for the trip points. 00 = digital hysteresis disabled 01 = enabled, offset is ~1.0C 10 = enabled, offset is ~2.0C 11 = enabled, offset is ~3.0C When set to non-zero value, the hysteresis works as follows: Assume a Aux trip point of 100C. On the cold-to-hot case, where the device is heating up, the Aux trip will occur at 100C. Later as the device cools down, the hotto-cool trip will occur at 97C, assuming a setting of 11. So the hysteresis only affects the hot-to-cool transition. This prevents thrashing on interrupts as the part hovers around 100C. This applies to Aux1, Aux2, Hot, and Cat trip points. Note: 03 :02 SE0 Sticky Not used in PCH Sequencer enable and Rate -- These bits enable the thermal sensor to sequence between four measurement functions and set the sequencer rate. When the sequencer is disabled, the Thermal Sensor Temperature Trip Point Setting [Catastrophic Trip point setting] values will be routed to bank A of the DAC and TSCO 6 lsbs is routed to bank B of the DAC. The Catastrophic trip is fully functional and the Hot trip is functional at fixed offset of ~2.0C below the catastrophic. The Auxiliary trip and HW thermometer are not functional in this mode. When sequencer is enabled, the Catastrophic, Hot, Auxiliary and thermometer circuits will all operate using the programmed trip points and sequencer rate. Note: When disabling the sequencer while thermometer running, the sequencer will finish the current cycle until it starts another one. Lockable via TSCO bit 7. 00: sequencer disabled (i.e, analog sensor mode) 01: enabled, eight clock mode (for testing digital logic) 10: enabled,104 clock mode (normal sequencer operation), provides 4.16 uS settling time @ 25 MHz 11: enabled,820 clock mode (fall back sequencer option), provides 32.8 uS settling time @ 25 MHz. Note: Sequencer is not supported in PCH and has been permenently disabled in this register. Intel(R) Communications Chipset 89xx Series - Datasheet 806 Offset Start: 01h Offset End: 01h Power Well: Core Bit Acronym 07 05 :04 Bus:Device:Function: B:31:6 October 2012 Order Number: 327879-001US 17.0 Table 17-48. Offset 01h: TS0C--Thermal Sensor 0 Control (Sheet 2 of 2) Description: View: PCI Size: 8 bit Bit Range 01 BAR: TBARB Default: 00h TS0OS 17.2.3.3 Reserved Offset Start: 01h Offset End: 01h Power Well: Core Bit Acronym Bit Description Sticky Thermal Sensor output select: This bit muxes between the two comparator outputs. Normally Catastrophic is used. Lockable via TSCO bit 7. 0 = Catastrophic 1 = Hot Note: 00 Bus:Device:Function: B:31:6 Bit Reset Value Bit Access 0b RWL Not used in PCH Reserved 0b Offset 02h: TS0S--Thermal Sensor 0 Status This read only register provides trip point and other status of the thermal sensor. Table 17-49. Offset 02h: TS0S--Thermal Sensor 0 Status Description: View: PCI Size: 8 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Bit Acronym Offset Start: 02h Offset End: 02h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 07 CTI0 Catastrophic Trip Indicator -- a 1 indicates that the temperature is above the catastrophic setting. 0b RO 06 HTI0 Hot Trip Indicator -- a 1 indicates that the temperature is above the Hot setting. 0b RO 05 ATI0 Auxiliary Trip Indicator -- a 1 indicates that the temperature is above the Auxiliary setting. 0b RO 04 Reserved Reserved 0b 03 ATI0 Auxiliary2 Trip Indicator -- a 1 indicates that the temperature is above the Auxiliary2 setting. 0b 02 Reserved Reserved 0b 01 DCCR0 Direct Catastrophic Comparator Read -- This bit reads the output of the Catastrophic comparator directly, without latching via the sequencer circuit. Used for testing. 0b RO 00 DHCR0 Direct Hot comparator read -- This bit reads the output of the Hot comparator directly, without latching via the sequencer circuit. Used for testing. 0b RO October 2012 Order Number: 327879-001US RO Intel(R) Communications Chipset 89xx Series - Datasheet 807 17.2.3.4 Offset 03h: TS0TR--Thermal Sensor 0 Thermometer Read Table 17-50. Offset 03h: TS0TR--Thermal Sensor 0 Thermometer Read Description: View: PCI Size: 8 bit Bit Range 07 :00 17.2.3.5 BAR: TBARB Bus:Device:Function: B:31:6 Default: XXh Offset Start: 03h Offset End: 03h Power Well: Core Bit Acronym Bit Description TR0 Temperature reading, where calculated value is (TR[7:0] - 50 deg C). These register bits reflect "Offset A8h: CCTS0TL1-- Thermal Sensor 0 Top Logic 1" on page 796 register `otstemperature' bits [8:1]. Default value is X it depends on temperature at power on. Sticky Bit Reset Value Bit Access XXh RO Offset 04h: TS0TTP--Thermal Sensor 0 Temperature Trip Point This register sets thermal trip points. This register should be set with setting of "Offset 0Bh: TS0TTPC--Thermal Sensor 0 Temperature Trip Point Cont'd" on page 810 Table 17-51. Offset 04h: TS0TTP--Thermal Sensor 0 Temperature Trip Point (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Bit Acronym Offset Start: 04h Offset End: 07h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access B4h RWL AUX2 Trip point setting -- When temperature reaches the value indicated by this bus, alert signal ("Offset 02h: TS0S--Thermal Sensor 0 Status" on page 807) bit 03 is asserted. * Default value has been set for 130C * Calculated value is (A2TPS[7:0] - 50C) Note: 31 :24 A2TPS0 Note: This register field provides an optional temperature trip pont. By default this trip point has been disabled by setting it to a value greater than the catastrophic trip point to prevent it from unitentionaly firing. If required this field must be programmed to an appropriate value before using in normal operation. AUX2 has a built in hysteresis of 2C. Care must be taken not to program the bits in this field to a value less than 2 or wraparound will occur on the final value used. These bits are lockable via programming the policy-lock down bit (bit 7) of TSPC reg Intel(R) Communications Chipset 89xx Series - Datasheet 808 October 2012 Order Number: 327879-001US 17.0 Table 17-51. Offset 04h: TS0TTP--Thermal Sensor 0 Temperature Trip Point (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Bit Acronym Offset Start: 04h Offset End: 07h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access B4h RWL 9Bh RWL A2h RWL AUX Trip point setting -- When temperature reaches the value indicated by this bus, alert signal ("Offset 02h: TS0S--Thermal Sensor 0 Status" on page 807) bit 05 is asserted. * Default value has been set for 130C * Calculated value is (ATPS[7:0] - 50C) Note: 23 :16 ATPS0 Note: This register field provides an optional temperature trip pont. By default this trip point has been disabled by setting it to a value greater than the catastrophic trip point to prevent it from unitentionaly firing. If required this field must be programmed to an appropriate value before using in normal operation. AUX has a built in hysteresis of 2C. Care must be taken not to program the bits in this field to a value less than 2 or wraparound will occur on the final value used. These bits are lockable via programming the policy-lock down bit (bit 7) of TSPC reg 15 :08 HTPS0 itsalertthreshon[7:0] -- (Hot On value) When temperature reaches the value indicated by this bus, alert signal (otstemperaturealrt) is turned on. * Default value has been set for 105C * Calculated value is (itsalertthreshon[7:0] - 50C) * Lockable via TSLOCK bit 2. itstripthreshon[7:0] -- (Catastrophic value) When temperature reaches the value indicated by this bus, trip signal (otstemperaturetrip) is turned on. 07 :00 CTPS0 * * * October 2012 Order Number: 327879-001US Default value has been set for 112C Calculated value is (itstripthreshon[7:0] - 50C) Lockable via TSCO bit 7. Intel(R) Communications Chipset 89xx Series - Datasheet 809 17.2.3.6 Offset 08h: TS0CO--Thermal Sensor 0 Calibration Offset Table 17-52. Offset 08h: TS0CO--Thermal Sensor 0 Calibration Offset Description: View: PCI Size: 8 bit Bit Range 07 BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Offset Start: 08h Offset End: 08h Power Well: Core Bit Acronym Bit Description LBC0 Lock bit for Catastrophic -- This bit, when written to a 1, locks the Catastrophic programming interface, including bits 6:0 of this register and bits 7:0 of TSTTP, bits 1,2, 3 and 7 of TSC, and bits 0,7 of TSTM. This bit may only be set to a 0 by a Host Partitioned reset (bb_rst). Writing a 0 to this bit has no effect. CF9 warm reset is a Host Partitioned Reset. Sticky Bit Reset Value Bit Access 0b RWLO 00h RWL Calibration Offset -- This field contains the current calibration offset for the Thermal Sensor DAC inputs. The calibration offset is a twos complement signed number which is added to the catastrophic/hot/aux/temp value before going to the DAC. This field is Read/Write and can be modified by Software unless locked by setting bit 7 of this register. 06 :00 CO0 While this is a seven-bit field, the 7th bit is sign extended to 9 bits for TCO operation. The range of 00h to 3fh corresponds to 0 0000 0000 to 0 0011 1111. The range of 41h to 7fh corresponds to 1 1100 0001 (i.e, negative 3fh) to 1 1111 1111 (i.e, negative 1), respectively. Note: 17.2.3.7 Not used in PCH Offset 0Bh: TS0TTPC--Thermal Sensor 0 Temperature Trip Point Cont'd This register sets Hot off thermal trip point. This register should be set with setting of "Offset 04h: TS0TTP--Thermal Sensor 0 Temperature Trip Point" on page 808 Table 17-53. Offset 0Bh: TS0TTPC--Thermal Sensor 0 Temperature Trip Point Cont'd Description: View: PCI Size: 8 bit Bit Range 07 :00 BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Power Well: Core Bit Acronym Bit Description HOTP0 itsalertthreshoff[7:0] -- (Hot off value) When temperature reaches the value indicated by this bus, alert signal (otstemperaturealrt) is turned off. * Default value has been set for 103C * Calculated value is (itsalertthreshoff [7:0] - 50C) Lockable via TSLOCK bit 2. Intel(R) Communications Chipset 89xx Series - Datasheet 810 Offset Start: 0Bh Offset End: 0Bh Sticky Bit Reset Value Bit Access 99h RWL October 2012 Order Number: 327879-001US 17.0 17.2.3.8 Offset 0Ch: TS0ES--Thermal Sensor 0 Error Status Table 17-54. Offset 0Ch: TS0ES--Thermal Sensor 0 Error Status Description: View: PCI Size: 8 bit Bit Range 07 06 05 04 03 02 01 00 BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Offset Start: 0Ch Offset End: 0Ch Power Well: Core Bit Reset Value Bit Access Aux2 High-to-Low Event: 1 = Indicates that an Aux2 Thermal Sensor event based on a higher to lower temperature transition thru the trip point 0 = No trip for this event Software must write a 1 to clear this status bit. 0b RWC CATHL0 Catastrophic High-to-Low Event: 1 = Indicates that a Catastrophic Thermal Sensor event based on a higher to lower temperature transition thru the trip point 0 = No trip for this event Software must write a 1 to clear this status bit. NOTE: This will never be seen. Power cycle will clear this bit. 0b RWC HOTHL0 Hot High-to-Low Event: 1 = Indicates that a Hot Thermal Sensor event based on a higher to lower temperature transition thru the trip point 0 = No trip for this event Software must write a 1 to clear this status bit. 0b RWC AUXHL0 Aux High-to-Low Event: 1 = Indicates that an Aux Thermal Sensor event based on a higher to lower temperature transition thru the trip point 0 = No trip for this event Software must write a 1 to clear this status bit. 0b RWC Aux2 Low-to-High Event: 1 = Indicates that an Aux2 Thermal Sensor trip event occurred based on a lower to higher temperature transition thru the trip point. 0 = No trip for this event Software must write a 1 to clear this status bit. 0b RWC CATLH0 Catastrophic Low-to-High Event: 1 = Indicates that a Catastrophic Thermal Sensor trip event occurred based on a lower to higher temperature transition thru the trip point. 0 = No trip for this event Software must write a 1 to clear this status bit. 0b RWC HOTLH0 Hot Low-to-High Event: 1 = Indicates that a Hot Thermal Sensor trip event occurred based on a lower to higher temperature transition thru the trip point. 0 -No trip for this event Software must write a 1 to clear this status bit. 0b RWC AUXLH0 Aux Low-to-High Event: 1 = Indicates that an Aux Thermal Sensor trip event occurred based on a lower to higher temperature transition thru the trip point. 0 = No trip for this event Software must write a 1 to clear this status bit. 0b RWC Bit Acronym AUX2HL0 AUX2LH0 October 2012 Order Number: 327879-001US Bit Description Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 811 17.2.3.9 Offset 0Dh: TS0GPEN--Thermal Sensor 0 General Purpose Event Enables This register controls the conditions that result in the General Purpose Event (GPE) signal from the Thermal Sensor (TS) logic to assert. When the TS GPE signal asserts, the GPE block reports a `1' in the TCOSCI_STS bit. The MCH thermal sensor logic also sets this bit. Table 17-55. Offset 0Dh: TS0GPEN--Thermal Sensor 0 General Purpose Event Enables Description: View: PCI Size: 8 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Bit Acronym Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 07 Aux2 High-to-Low Enable -- See the description for bit 6. 0b RW 06 Catastrophic High-to-Low Enable: When set to 1, the thermal sensor logic asserts its General Purpose Event signal to the GPE block when the corresponding status bit is set in the Thermal Error Status register. When cleared, the corresponding status bit does not result in the GPE signal assertion. 0b RW 05 Hot High-to-Low Enable -- See the description for bit 6. 0b RW 04 Aux High-to-Low Enable -- See the description for bit 6. 0b RW 03 Aux2 Low-to-High Event -- See the description for bit 6. 0b RW 02 Catastrophic Low-to-High Enable -- See the description for bit 6. 0b RW 01 Hot Low-to-High Event -- See the description for bit 6. 0b RW 00 Aux Low-to-High Event -- See the description for bit 6. 0b RW Intel(R) Communications Chipset 89xx Series - Datasheet 812 Offset Start: 0Dh Offset End: 0Dh October 2012 Order Number: 327879-001US 17.0 17.2.3.10 Offset 0Eh: TS0PC--Thermal Sensor 0 Policy Control This register enables specific events to generate an SMI#. Table 17-56. Offset 0Eh: TS0PC--Thermal Sensor 0 Policy Control Description: View: PCI Size: 8 bit Bit Range BAR: TBARB Default: 00h Bit Acronym Bit Description Sticky Catastrophic Power-Down Enable -- When set to 1, the power management logic transitions to the S5 state when a catastrophic temperature is detected by the sensor. The transition to the S5 state must be unconditional (like the Power Button Override Function). The assertion of the signal controlling the main power well (SLP_S3# or SLP_S4#) must occur within 100 sec of the detection of the catastrophic condition. NOTE: The thermal sensor and response logic is in the core/main power well; therefore, detection of a catastrophic temperature is limited to times when this well is powered and out of reset. NOTE: The sequencer mode can delay the detection of a catastrophic temperature. Assuming the 512-clock sequencer mode is in use, a catastrophic condition may not be detected internally for up to 16.4 secs. Therefore the total delay from the catastrophic condition to the assertion of the SLP_Sx# signals can be 116.4 secs. 06 Reserved Offset Start: 0Eh Offset End: 0Eh Power Well: Core Policy Lock-Down Bit -- When written to 1, this bit prevents any more writes to this register (offset 0Eh) and bits 31:16 of the Thermal Sensor Temperature Trip Point register (offset 04h). The "Lock Bit for Catastrophic" (bit 7 of offset 08h) and the "Lock Bit for Hot" (bit 2 of offset 83h) must also be 1 when this bit is set to 1. When 0, the thermal policies can be programmed and modified. NOTE: For the second thermal sensor, the registers are located 40h above the offsets for the first thermal sensor. This bit is reset by a Host Partitioned Reset (bb_rst). CF9 warm reset is a Host Partitioned Reset. 07 05 :04 Bus:Device:Function: B:31:6 Reserved Bit Reset Value Bit Access 0b RWLO 0 RW 00b 03 SMI Enable on Aux2 Thermal Sensor Trip: 1: Enables SMI# assertions on Aux2 thermal sensor events for either low-to-high or high-to-low events. (Both edges are enabled by this one bit.) 0: Disables SMI# assertions for Aux2 thermal events 0b RW 02 SMI Enable on Catastrophic Thermal Sensor Trip: 1: Enables SMI# assertions on catastrophic thermal sensor events for either low-to-high or high-to-low events. (Both edges are enabled by this one bit.) 0: Disables SMI# assertions for catastrophic thermal events. 0b RW 01 SMI Enable on Hot Thermal Sensor Trip: 1: Enables SMI# assertions on hot thermal sensor events for either low-to-high or high-to-low events. (Both edges are enabled by this one bit.) 0: Disables SMI# assertions for hot thermal events 0b RW 00 SMI Enable on Aux Thermal Sensor Trip: 1: Enables SMI# assertions on aux thermal sensor events for either low-to-high or high-to-low events. (Both edges are enabled by this one bit.) 0: Disables SMI# assertions for aux thermal events 0b RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 813 17.2.3.11 Offset 10h: CPEC--CPU Power Error Correction Data Table 17-57. Offset 10h: CPEC--CPU Power Error Correction Data Description: View: PCI BAR: TBARB Size: 16 bit Default: 00h Bit Range Bit Acronym Offset Start: 10h Offset End: 11h Power Well: Core Bit Description Sticky CPU Power Error Correction Data -- This register is RW to host and has no functionality in PCH. Locked by CTC[7] 15 :00 17.2.3.12 Bus:Device:Function: B:31:6 Bit Reset Value Bit Access 00h RW Offset 12h: C0TA--CPU0 Temperature Adjust 0 Table 17-58. Offset 12h: C0TA--CPU0 Temperature Adjust Description: View: PCI Size: 16 bit Bit Range 15 :00 BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Bit Acronym Power Well: Core Bit Description Host writes the CPU0's TJmax into this register. ME uses the value to create the CPU0's absolute temperature. The value received from the cpu over PECI is a negative offset relative to the C0TA value. Locked by CTC[7]. Intel(R) Communications Chipset 89xx Series - Datasheet 814 Offset Start: 12h Offset End: 13h Sticky Bit Reset Value Bit Access 00h RW October 2012 Order Number: 327879-001US 17.0 17.2.3.13 Offset 16h: C1TA--CPU1 Temperature Adjust Table 17-59. Offset 16h: C1TA--CPU1Temperature Adjust Description: View: PCI Size: 16 bit Bit Range BAR: TBARB Default: 0000h Bit Acronym Offset Start: 16h Offset End: 17h Power Well: Core Bit Description Sticky Host writes the CPU1's TJmax into this register. ME uses the value to create the CPU1's absolute temperature. The value received from the cpu over PECI is a negative offset relative to the C1TA value. Locked by CTC[7]. 15 :00 17.2.3.14 Bus:Device:Function: B:31:6 Bit Reset Value Bit Access 00h RW Offset 1Ah: MC--MPC Control All the bits below are used by ME FW to enable different thermal reporting features. BIOS: Each OEM will program the following differently, based on that OEM's configuration. Table 17-60. Offset 1Ah: MC--MPC Control (Sheet 1 of 2) Description: View: PCI Size: 16 bit Bit Range BAR: TBARB Default: 0000h Reserved 12 11 :08 Reserved Offset Start: 1Ah Offset End: 1Bh Power Well: Core Bit Reset Value Bit Access 2nd CPU Present: 1 = 2nd CPU is present in the system. BIOS sets this bit to `1' in DP systems. When the 2nd CPU is present, all PECI Thermal Reporting commands are executed to both CPUs. 0 = 2nd CPU NOT present. BIOS sets this to `0' in UP systems 0b RW Reserved 0b RW System Thermal Reporting Enable: 1 = Enable System Thermal Reporting 0 = Disable System Thermal Reporting 0b RW Bit Acronym 15 14 :13 Bus:Device:Function: B:31:6 Bit Description Sticky Reserved 0b RW 07 CPU Temperature Read Enable: 1 = Enable PECI Reads of CPU Temp 0 = Disable PECI Reads of CPU Temp 0b RW 06 CPU Energy Read Enable 1 = Enable PECI Reads of CPU Energy 0 = Disable PECI Reads of CPU Energy. 0b RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 815 Table 17-60. Offset 1Ah: MC--MPC Control (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range BAR: TBARB Default: 0000h 17.2.3.15 Reserved Offset Start: 1Ah Offset End: 1Bh Power Well: Core Bit Reset Value Bit Access PCH Temperature Read Enable 1 = Enable PCH Temperature Read 0 = Disable PCH Temperature Read. 0b RW Reserved 0b RW Bit Acronym 05 04 :00 Bus:Device:Function: B:31:6 Bit Description Sticky Offset 22h: Timestamp Table 17-61. Offset 22h: Timestamp Description: View: PCI Size: 16 bit Bit Range BAR: TBARB Default: 0000h Bit Acronym 11 :00 Reserved Offset Start: 22h Offset End: 23h Power Well: Core Bit Description When ME writes a 1 to this bit, bits [11:0] are loaded from the free running 10 msec counter. 15 14 :12 Bus:Device:Function: B:31:6 Reserved Timestamp -- This register provides a RO count of 10 msec pulses for timestamp purposes. SW can read the data and the timestamp and wait several seconds and read the two again. It takes the difference between the 2 samples to calculate power over a large window. SW is responsible for handling counter rollover, which occurs every 40 seconds. Sticky Bit Reset Value Bit Access 0b RO 00b 000h RO NOTE: Between any 2 samples, the difference with respect to the data collection could be off by 1 clock, since the timing on PECI is not precise with the timestamp itself. Intel(R) Communications Chipset 89xx Series - Datasheet 816 October 2012 Order Number: 327879-001US 17.0 17.2.3.16 Offset 24h: DIMMEN--DIMM Enable The bits are configured by BIOS and used by ME FW to identify DIMMs that are present in the system and can be enabled for thermal reporting. This register is used PDP and SDP applications. For SDP platforms, Bits[1:0] are used to enable two DIMMs in CH0; and Bits[5:4] are for two DIMMs in CH1. BIOS sets the bits for the DIMMs that are physically present and wants to enable for thermal polling. Table 17-62. Offset 24h: DIMMEN--DIMM Enable (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 0000h Offset Start: 24h Offset End: 27h Power Well: Core Bit Acronym Bit Description Sticky Bit Reset Value Bit Access 0b RW 0b RW 0b RW 0b RW 0b RW CPU1/CH3 DIMM_Enable Enables reads of DIMM's populated in CPU1/CH3 and have an onboard TS. 31 :28 Bit[28] Bit[29] Bit[30] Bit[31] = = = = CPU1/CH3/DIMM0 CPU1/CH3/DIMM1 CPU1/CH3/DIMM2 Reserved CPU1/CH2 DIMM_Enable Enables reads of DIMM's populated in CPU1/CH2 and have an onboard TS. 27 :24 Bit[24] Bit[25] Bit[26] Bit[27] = = = = CPU1/CH2/DIMM0 CPU1/CH2/DIMM1 CPU1/CH2/DIMM2 Reserved CPU1/CH1 DIMM_Enable Enables reads of DIMM's populated in CPU1/CH1 and have an onboard TS. 23 :20 Bit[20] Bit[21] Bit[22] Bit[23] = = = = CPU1/CH1/DIMM0 CPU1/CH1/DIMM1 CPU1/CH1/DIMM2 Reserved CPU1/CH0 DIMM_Enable Enables reads of DIMM's populated in CPU1/CH0 and have an onboard TS. 19 :16 Bit[16] Bit[17] Bit[18] Bit[19] = = = = CPU1/CH0/DIMM0 CPU1/CH0/DIMM1 CPU1/CH0/DIMM2 Reserved CPU0/CH3 DIMM_Enable Enables reads of DIMM's populated in CPU0/CH3 and have an onboard TS. 15 :12 October 2012 Order Number: 327879-001US Bit[12] Bit[13] Bit[14] Bit[15] = = = = CPU0/CH3/DIMM0 CPU0/CH3/DIMM1 CPU0/CH3/DIMM2 Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 817 Table 17-62. Offset 24h: DIMMEN--DIMM Enable (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 0000h Offset Start: 24h Offset End: 27h Power Well: Core Bit Acronym Bit Description Sticky Bit Reset Value Bit Access 0b RW 0b RW 0b RW CPU0/CH2 DIMM_Enable Enables reads of DIMM's populated in CPU0/CH2 and have an onboard TS. 11 :08 Bit[8] = CPU0/CH2/DIMM0 Bit[9] = CPU0/CH2/DIMM1 Bit[10] = CPU0/CH2/DIMM2 Bit[11] = Reserved CPU0/CH1 DIMM_Enable Enables reads of DIMM's populated in CPU0/CH1 and have an onboard TS. Bit[4] Bit[5] Bit[6] Bit[7] 07 :04 = = = = Note: CPU0/CH1/DIMM0 CPU0/CH1/DIMM1 CPU0/CH1/DIMM2 Reserved Bit[4] and Bit[5] are used to enable two DIMMs in CH1 on SDP platforms CPU0/CH0 DIMM_Enable Enables reads of DIMM's populated in CPU0/CH0 and have an onboard TS. Bit[0] Bit[1] Bit[2] Bit[3] 03 :00 = = = = Note: 17.2.3.17 CPU0/CH0/DIMM0 CPU0/CH0/DIMM1 CPU0/CH0/DIMM2 Reserved Bit[0] and Bit[1] are used to enable two DIMMs in CH0 on SDP platforms Offset 30h: C0TV--CPU0 Temperature Value Table 17-63. Offset 30h: C0TV--CPU0 Temperature Value Description: View: PCI Size: 16 bit Bit Range 15 :00 BAR: TBARB Bus:Device:Function: B:31:6 Default: 0000h Bit Acronym Power Well: Core Bit Description CPU0 Temperature -- This provides the CPU0 temperature. For format, See the spec for EC support. This is byte 3 of the EC data. Intel(R) Communications Chipset 89xx Series - Datasheet 818 Offset Start: 30h Offset End: 31h Sticky Bit Reset Value Bit Access 0000h RO October 2012 Order Number: 327879-001US 17.0 17.2.3.18 Offset 32h: C1TV--CPU1 Temperature Value Table 17-64. Offset 32h: C1TV--CPU1 Temperature Value Description: View: PCI Size: 16 bit Bit Range BAR: TBARB Default: 0000h Bit Acronym Offset Start: 32h Offset End: 33h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access CPU1 Temperature -- This provides the CP1 temperature. for format, See the spec for EC support. This is byte 4 of the EC data. 15 :00 17.2.3.19 Bus:Device:Function: B:31:6 RO Offset 34h: C0EV--CPU0 Energy Value Table 17-65. Offset 34h: C0EV--CPU0 Energy Value Description: View: PCI Size: 32 bit Bit Range BAR: TBARB Default: 00000000h Bit Acronym Offset Start: 34h Offset End: 37h Power Well: Core Bit Description Sticky CPU0 Power -- ME writes the CPU0 power here after collecting it over PECI. This is bytes 14:11 of the EC data. Byte 11 of the EC data is bits [7:0] of this register. 31 :00 17.2.3.20 Bus:Device:Function: B:31:6 Bit Reset Value Bit Access 00000000h RO Offset 38h: C1EV--CPU1 Energy Value Table 17-66. Offset 38h: C1EV--CPU1 Energy Value Description: View: PCI Size: 32 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Bit Acronym 31 :00 October 2012 Order Number: 327879-001US Offset Start: 38h Offset End: 3Bh Power Well: Core Bit Description Sticky CPU1 Power: ME writes the CPU1 power here after collecting it over PECI. This register is used for CPU1 in a DP system. It is not used in a uP system. Bit Reset Value Bit Access 00000000h RO Intel(R) Communications Chipset 89xx Series - Datasheet 819 17.2.3.21 Offset 3Fh: AE--Alert Enable BIOS: Some OEM's may enable the alert feature, whereby the PCH indicates through a pin if a device is outside the thermal bounds. In general BIOS should not program the following register at all. If the OEM wishes to enable the pin-based alert, then BIOS should write a `1' to the bits below that should be enabled for the particular Embedded Controller on that platform. Table 17-67. Offset 3Fh: AE--Alert Enable Description: View: PCI Size: 8 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Bit Acronym Power Well: Core Bit Description Lock Enable -- When `1', this bit (Lock Enable) and CTC[6:0] are locked and the following registers are locked against any writes from the host: -CPU Power Error Correction 0x10 -C0TA 0x12 -PTA 0x14 -C1TA 0x16 07 Offset Start: 3Fh Offset End: 3Fh Sticky Bit Reset Value Bit Access 0b RWL This bit is reset by Host Partitioned Reset (bb_rst). CF9 warm reset is a Host Partitioned Reset. 06 CPU Alert Enable - Locked by CTC[7] 0b RWL 05 MCH Alert Enable - Locked by CTC[7] 0b RWL 04 PCH Alert Enable - Locked by CTC[7] 0b RWL 03 DIMM Alert Enable -- the actual DIMMs that are read and used for comparison are enabled in a different register Locked by CTC[7] 0b RWL 000b RWL 02 :00 Reserved Reserved - Locked by CTC[7] Intel(R) Communications Chipset 89xx Series - Datasheet 820 October 2012 Order Number: 327879-001US 17.0 17.2.3.22 Offset 40h: TS1IU--Thermal Sensor 1 In Use Table 17-68. Offset 40h: TS1IU--Thermal Sensor 1 In Use Description: View: PCI Size: 8 bit BAR: TBARB Default: 00h Bit Range Bit Acronym 07 :01 Reserved 00 Bus:Device:Function: B:31:6 TS1IU Offset Start: 40h Offset End: 40h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00h Thermal Sensor In Use -- Software semaphore bit. After a core-well reset, a read to this bit returns a 0. After the first read, subsequent reads will return a 1. A write of a 1 to this bit will reset the next read value to 0. Writing a 0 to this bit has no effect. Software can poll this bit until it reads a 0, and will then own the usage of the thermal sensor. This bit has no other effect on the hardware, and is only used as a semaphore among various independent software threads that may need to use the thermal sensor. Software that reads this register but does not intend to claim exclusive access of the thermal sensor must write a one to this bit if it reads a 0, in order to allow other software threads to claim it. 0b RS/WC NOTE: The ME only reads the thermal interface registers and as such ME reads to TSIU will not set the bit. 17.2.3.23 Offset 41h: TS1C--Thermal Sensor 1 Control This register controls the operation of the thermal sensor. Table 17-69. Offset 41h: TS1C--Thermal Sensor 1 Control (Sheet 1 of 2) Description: View: PCI Size: 8 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h TS1E 06 Reserved October 2012 Order Number: 327879-001US Power Well: Core Bit Reset Value Bit Access Thermal Sensor enable -- This bit enables thermal reporting. Lockable via TSCO bit 7: 0 -Disabled 1 -Enabled 0b RWL Reserved 0b RW Bit Acronym 07 Offset Start: 41h Offset End: 41h Bit Description Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 821 Table 17-69. Offset 41h: TS1C--Thermal Sensor 1 Control (Sheet 2 of 2) Description: View: PCI Size: 8 bit Bit Range 05 :04 BAR: TBARB Default: 00h DHA1 SE1 Bit Description Digital Hysteresis Amount: This bit determines whether no offset, 1.0C, 2.0C, or 3.0C is used for hysteresis for the trip points. 00 = digital hysteresis disabled 01 = enabled, offset is ~1.0C 10 = enabled, offset is ~2.0C 11 = enabled, offset is ~3.0C When set to non-zero value, the hysteresis works as follows: Assume a Aux trip point of 100C. On the cold-to-hot case, where the device is heating up, the Aux trip will occur at 100C. Later as the device cools down, the hotto-cool trip will occur at 97C, assuming a setting of 11. So the hysteresis only affects the hot-to-cool transition. This prevents thrashing on interrupts as the part hovers around 100C. This applies to Aux1, Aux2, Hot, and Cat trip points. 01 TS1OS 00 Reserved Bit Access 00b RW 00b RO 0b RWL Not used in the PCH Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 822 Bit Reset Value Sequencer is not supported in the PCH and has been permanently disabled in this register. Thermal Sensor output select: This bit muxes between the two comparator outputs. Normally Catastrophic is used. Lockable via TSCO bit 7. 0 = Catastrophic 1 = Hot Note: Sticky Not used in PCH Sequencer enable and Rate -- These bits enable the thermal sensor to sequence between four measurement functions and set the sequencer rate. When the sequencer is disabled, the Thermal Sensor Temperature Trip Point Setting [Catastrophic Trip point setting] values will be routed to bank A of the DAC and TSCO 6 lsbs is routed to bank B of the DAC. The Catastrophic trip is fully functional and the Hot trip is functional at fixed offset of ~2.0C below the catastrophic. The Auxiliary trip and HW thermometer are not functional in this mode. When sequencer is enabled, the Catastrophic, Hot, Auxiliary and thermometer circuits will all operate using the programmed trip points and sequencer rate. Note: When disabling the sequencer while thermometer running, the sequencer will finish the current cycle until it starts another one. Lockable via TSCO bit 7. 00: sequencer disabled (i.e, analog sensor mode) 01: enabled, eight clock mode (for testing digital logic) 10: enabled,104 clock mode (normal sequencer operation), provides 4.16 uS settling time @ 25 MHz 11: enabled,820 clock mode (fall back sequencer option), provides 32.8 uS settling time @ 25 MHz. Note: Offset Start: 41h Offset End: 41h Power Well: Core Bit Acronym Note: 03 :02 Bus:Device:Function: B:31:6 0b October 2012 Order Number: 327879-001US 17.0 17.2.3.24 Offset 42h: TS1S--Thermal Sensor 1 Status This read only register provides trip point and other status of the thermal sensor. Table 17-70. Offset 42h: TS1S--Thermal Sensor 1 Status Description: View: PCI Size: 8 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Bit Acronym Offset Start: 42h Offset End: 42h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 07 CTI1 Catastrophic Trip Indicator -- a 1 indicates that the temperature is above the catastrophic setting. 0b RO 06 HTI1 Hot Trip Indicator -- a 1 indicates that the temperature is above the Hot setting. 0b RO 05 ATI1 Auxiliary Trip Indicator -- a 1 indicates that the temperature is above the Auxiliary setting. 0b RO 04 Reserved Reserved 0b Auxiliary2 Trip Indicator -- a 1 indicates that the temperature is above the Auxiliary2 setting. 0b Reserved 0b 03 ATI1 02 Reserved 01 DCCR1 Direct Catastrophic Comparator Read -- This bit reads the output of the Catastrophic comparator directly, without latching via the sequencer circuit. Used for testing. 0b RO 00 DHCR1 Direct Hot comparator read -- This bit reads the output of the Hot comparator directly, without latching via the sequencer circuit. Used for testing. 0b RO October 2012 Order Number: 327879-001US RO Intel(R) Communications Chipset 89xx Series - Datasheet 823 17.2.3.25 Offset 43h: TS1TR--Thermal Sensor 1 Thermometer Read Table 17-71. Offset 43h: TS1TR--Thermal Sensor 1 Thermometer Read Description: View: PCI Size: 8 bit Bit Range 07 :00 17.2.3.26 BAR: TBARB Bus:Device:Function: B:31:6 Default: XXh Offset Start: 43h Offset End: 43h Power Well: Core Bit Acronym Bit Description TR1 Temperature reading, where calculated value is (TR[7:0] - 50 deg C). These register bits reflect "Offset C8h: CCTS1TL1-- Thermal Sensor 1 Top Logic 1" on page 801 register `otstemperature' bits [8:1]. Default value is X it depends on temperature at power on. Sticky Bit Reset Value Bit Access XXh RO Offset 44h: TS1TTP--Thermal Sensor 1 Temperature Trip Point This register sets thermal trip points. This register should be set with setting of "Offset 4Bh: TS1TTPC--Thermal Sensor 1 Temperature Trip Point Cont'd" on page 827 Table 17-72. Offset 44h: TS1TTP--Thermal Sensor 1 Temperature Trip Point (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Bit Acronym Offset Start: 44h Offset End: 47h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access B4h RWL AUX2 Trip point setting -- When temperature reaches the value indicated by this bus, alert signal ("Offset 42h: TS1S--Thermal Sensor 1 Status" on page 823) bit 03 is asserted. * Default value has been set for 130C * Calculated value is (A2TPS[7:0] - 50C) Note: 31 :24 A2TPS1 Note: This register field provides an optional temperature trip pont. By default this trip point has been disabled by setting it to a value greater than the catastrophic trip point to prevent it from unitentionaly firing. If required this field must be programmed to an appropriate value before using in normal operation. AUX2 has a built in hysteresis of 2C. Care must be taken not to program the bits in this field to a value less than 2 or wraparound will occur on the final value used. These bits are lockable via programming the policy-lock down bit (bit 7) of TSPC reg Intel(R) Communications Chipset 89xx Series - Datasheet 824 October 2012 Order Number: 327879-001US 17.0 Table 17-72. Offset 44h: TS1TTP--Thermal Sensor 1 Temperature Trip Point (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Bit Acronym Offset Start: 44h Offset End: 47h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access B4h RWL 9Bh RWL A2h RWL AUX Trip point setting -- When temperature reaches the value indicated by this bus, alert signal ("Offset 42h: TS1S--Thermal Sensor 1 Status" on page 823) bit 05 is asserted. * Default value has been set for 130C * Calculated value is (ATPS[7:0] - 50C) Note: 23 :16 ATPS1 Note: This register field provides an optional temperature trip pont. By default this trip point has been disabled by setting it to a value greater than the catastrophic trip point to prevent it from unitentionaly firing. If required this field must be programmed to an appropriate value before using in normal operation. AUX has a built in hysteresis of 2C. Care must be taken not to program the bits in this field to a value less than 2 or wraparound will occur on the final value used. These bits are lockable via programming the policy-lock down bit (bit 7) of TSPC reg 15 :08 HTPS1 itsalertthreshon[7:0] -- (Hot On value) When temperature reaches the value indicated by this bus, alert signal (otstemperaturealrt) is turned on. * Default value has been set for 105C * Calculated value is (itsalertthreshon[7:0] - 50C) Lockable via TSLOCK bit 2. 07 :00 CTPS1 itstripthreshon[7:0] -- (Catastrophic value) When temperature reaches the value indicated by this bus, trip signal (otstemperaturetrip) is turned on. * Default value has been set for 112C * Calculated value is (itstripthreshon[7:0] - 50C) Lockable via TSCO bit 7. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 825 17.2.3.27 Offset 48h: TS1CO--Thermal Sensor 1 Calibration Offset Table 17-73. Offset 48h: TS1CO--Thermal Sensor 1 Calibration Offset Description: View: PCI Size: 8 bit Bit Range 07 BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Offset Start: 48h Offset End: 48h Power Well: Core Bit Acronym Bit Description LBC1 Lock bit for Catastrophic -- This bit, when written to a 1, locks the Catastrophic programming interface, including bits 6:0 of this register and bits 7:0 of TSTTP, bits 1,2, 3 and 7 of TSC, and bits 0,7 of TSTM. This bit may only be set to a 0 by a Host Partitioned reset (bb_rst). Writing a 0 to this bit has no effect. CF9 warm reset is a Host Partitioned Reset. Sticky Bit Reset Value Bit Access 0b RWLO 00h RWL Calibration Offset -- This field contains the current calibration offset for the Thermal Sensor DAC inputs. The calibration offset is a twos complement signed number which is added to the catastrophic/hot/aux/temp value before going to the DAC. This field is Read/Write and can be modified by Software unless locked by setting bit 7 of this register. 06 :00 CO1 While this is a seven-bit field, the 7th bit is sign extended to 9 bits for TCO operation. The range of 00h to 3fh corresponds to 0 0000 0000 to 0 0011 1111. The range of 41h to 7fh corresponds to 1 1100 0001 (i.e, negative 3fh) to 1 1111 1111 (i.e, negative 1), respectively. Note: Not used in the PCH Intel(R) Communications Chipset 89xx Series - Datasheet 826 October 2012 Order Number: 327879-001US 17.0 17.2.3.28 Offset 4Bh: TS1TTPC--Thermal Sensor 1 Temperature Trip Point Cont'd This register sets Hot off thermal trip point. This register should be set with setting of "Offset 44h: TS1TTP--Thermal Sensor 1 Temperature Trip Point" on page 824 Table 17-74. Offset 4Bh: TS1TTPC--Thermal Sensor 1 Temperature Trip Point Cont'd Description: View: PCI Size: 8 bit Bit Range 07 :00 17.2.3.29 BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Offset Start: 4Bh Offset End: 4Bh Power Well: Core Bit Acronym Bit Description Sticky HOTP1 itsalertthreshoff[7:0] -- (Hot off value) When temperature reaches the value indicated by this bus, alert signal (otstemperaturealrt) is turned off. * Default value has been set for 103C * Calculated value is (itsalertthreshoff [7:0] - 50C) Lockable via TSLOCK bit 2. Bit Reset Value Bit Access 99h RWL Offset 4Ch: TS1ES--Thermal Sensor 1 Error Status Table 17-75. Offset 4Ch: TS1ES--Thermal Sensor 1 Error Status Description: View: PCI Size: 8 bit Bit Range 07 06 05 BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Offset Start: 4Ch Offset End: 4Ch Power Well: Core Bit Reset Value Bit Access Aux2 High-to-Low Event: 1 = Indicates that an Aux2 Thermal Sensor event based on a higher to lower temperature transition thru the trip point 0 = No trip for this event Software must write a 1 to clear this status bit. 0b RWC CATHL1 Catastrophic High-to-Low Event: 1 = Indicates that a Catastrophic Thermal Sensor event based on a higher to lower temperature transition thru the trip point 0 = No trip for this event Software must write a 1 to clear this status bit. NOTE: This will never be seen. Power cycle will clear this bit. 0b RWC HOTHL1 Hot High-to-Low Event: 1 = Indicates that a Hot Thermal Sensor event based on a higher to lower temperature transition thru the trip point 0 = No trip for this event Software must write a 1 to clear this status bit. 0b RWC Bit Acronym AUX2HL1 October 2012 Order Number: 327879-001US Bit Description Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 827 Table 17-75. Offset 4Ch: TS1ES--Thermal Sensor 1 Error Status Description: View: PCI Size: 8 bit Bit Range 04 03 02 01 00 BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Power Well: Core Bit Reset Value Bit Access Aux High-to-Low Event: 1 = Indicates that an Aux Thermal Sensor event based on a higher to lower temperature transition thru the trip point 0 = No trip for this event Software must write a 1 to clear this status bit. 0b RWC Aux2 Low-to-High Event: 1 = Indicates that an Aux2 Thermal Sensor trip event occurred based on a lower to higher temperature transition thru the trip point. 0 = No trip for this event Software must write a 1 to clear this status bit. 0b RWC CATLH1 Catastrophic Low-to-High Event: 1 = Indicates that a Catastrophic Thermal Sensor trip event occurred based on a lower to higher temperature transition thru the trip point. 0 = No trip for this event Software must write a 1 to clear this status bit. 0b RWC HOTLH1 Hot Low-to-High Event: 1 = Indicates that a Hot Thermal Sensor trip event occurred based on a lower to higher temperature transition thru the trip point. 0 -No trip for this event Software must write a 1 to clear this status bit. 0b RWC AUXLH1 Aux Low-to-High Event: 1 = Indicates that an Aux Thermal Sensor trip event occurred based on a lower to higher temperature transition thru the trip point. 0 = No trip for this event Software must write a 1 to clear this status bit. 0b RWC Bit Acronym Bit Description AUXHL1 AUX2LH1 Intel(R) Communications Chipset 89xx Series - Datasheet 828 Offset Start: 4Ch Offset End: 4Ch Sticky October 2012 Order Number: 327879-001US 17.0 17.2.3.30 Offset 4Dh: TS1GPEN--Thermal Sensor 1 General Purpose Event Enables This register controls the conditions that result in the General Purpose Event (GPE) signal from the Thermal Sensor (TS) logic to assert. When the TS GPE signal asserts, the GPE block reports a `1' in the TCOSCI_STS bit. The MCH thermal sensor logic also sets this bit. Table 17-76. Offset 4Dh: TS1GPEN--Thermal Sensor 1 General Purpose Event Enables Description: View: PCI Size: 8 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Bit Acronym Offset Start: 4Dh Offset End: 4Dh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 07 Aux2 High-to-Low Enable -- See the description for bit 6. 0b RW 06 Catastrophic High-to-Low Enable: When set to 1, the thermal sensor logic asserts its General Purpose Event signal to the GPE block when the corresponding status bit is set in the Thermal Error Status register. When cleared, the corresponding status bit does not result in the GPE signal assertion. 0b RW 05 Hot High-to-Low Enable -- See the description for bit 6. 0b RW 04 Aux High-to-Low Enable -- See the description for bit 6. 0b RW 03 Aux2 Low-to-High Event -- See the description for bit 6. 0b RW 02 Catastrophic Low-to-High Enable -- See the description for bit 6. 0b RW 01 Hot Low-to-High Event -- See the description for bit 6. 0b RW 00 Aux Low-to-High Event -- See the description for bit 6. 0b RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 829 17.2.3.31 Offset 4Eh: TS1PC--Thermal Sensor 1 Policy Control This register enables specific events to generate an SMI#. Table 17-77. Offset 4Eh: TS1PC--Thermal Sensor 1 Policy Control Description: View: PCI Size: 8 bit Bit Range BAR: TBARB Default: 00h Bit Acronym Bit Description Catastrophic Power-Down Enable -- When set to 1, the power management logic transitions to the S5 state when a catastrophic temperature is detected by the sensor. The transition to the S5 state must be unconditional (like the Power Button Override Function). The assertion of the signal controlling the main power well (SLP_S3# or SLP_S4#) must occur within 100 sec of the detection of the catastrophic condition. NOTE: The thermal sensor and response logic is in the core/main power well; therefore, detection of a catastrophic temperature is limited to times when this well is powered and out of reset. NOTE: The sequencer mode can delay the detection of a catastrophic temperature. Assuming the 512-clock sequencer mode is in use, a catastrophic condition may not be detected internally for up to 16.4 secs. Therefore the total delay from the catastrophic condition to the assertion of the SLP_Sx# signals can be 116.4 secs. 06 Reserved Reserved Sticky Bit Reset Value Bit Access 0b RWLO 0 RW 00b 03 SMI Enable on Aux2 Thermal Sensor Trip: 1: Enables SMI# assertions on Aux2 thermal sensor events for either low-to-high or high-to-low events. (Both edges are enabled by this one bit.) 0: Disables SMI# assertions for Aux2 thermal events 0b RW 02 SMI Enable on Catastrophic Thermal Sensor Trip: 1: Enables SMI# assertions on catastrophic thermal sensor events for either low-to-high or high-to-low events. (Both edges are enabled by this one bit.) 0: Disables SMI# assertions for catastrophic thermal events. 0b RW 01 SMI Enable on Hot Thermal Sensor Trip: 1: Enables SMI# assertions on hot thermal sensor events for either low-to-high or high-to-low events. (Both edges are enabled by this one bit.) 0: Disables SMI# assertions for hot thermal events 0b RW 00 SMI Enable on Aux Thermal Sensor Trip: 1: Enables SMI# assertions on aux thermal sensor events for either low-to-high or high-to-low events. (Both edges are enabled by this one bit.) 0: Disables SMI# assertions for aux thermal events 0b RW Intel(R) Communications Chipset 89xx Series - Datasheet 830 Offset Start: 4Eh Offset End: 4Eh Power Well: Core Policy Lock-Down Bit -- When written to 1, this bit prevents any more writes to this register (offset 0Eh) and bits 31:16 of the Thermal Sensor Temperature Trip Point register (offset 04h). The "Lock Bit for Catastrophic" (bit 7 of offset 08h) and the "Lock Bit for Hot" (bit 2 of offset 83h) must also be 1 when this bit is set to 1. When 0, the thermal policies can be programmed and modified. NOTE: For the second thermal sensor, the registers are located 40h above the offsets for the first thermal sensor. This bit is reset by a Host Partitioned Reset (bb_rst). CF9 warm reset is a Host Partitioned Reset. 07 05 :04 Bus:Device:Function: B:31:6 October 2012 Order Number: 327879-001US 17.0 17.2.3.32 Offset 50h: HTS--HOST Turbo Status This register provides the Host view of turbo status. It is used by ME to respond to MCTP messages that request Turbo Status. The SMBus Turbo Status is what an EC can write to. So an EC writes to the SMBus Turbo Status register with its request, which causes an interrupt, and the host updates the Host Turbo Status register so that an EC read of the status gets the Host Turbo Status data and not necessarily what it wrote. Table 17-78. Offset 50h: HTS--HOST Turbo Status Description: View: PCI Size: 48 bit Bit Range BAR: TBARB Default: 000000000000h 00 Offset Start: 50h Offset End: 55h Power Well: Core Bit Reset Value Bit Access See MPC spec for bit definitions -- This represents the data bytes 20:15 provided to the EC when it does a read. Byte 15 is bits [7:0]. 000000000000h RW New Valid Value -- Host sets this bit to indicate the data in this register is valid. Cleared on write of `1' to MEC register. 0b RW Bit Acronym 47 :01 17.2.3.33 Bus:Device:Function: B:31:6 Bit Description Sticky Offset 56h: MTL--MCP Temperature Limit Table 17-79. Offset 56h: MTL--MCP Temperature Limit Description: View: PCI Size: 16 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 0000h Bit Acronym 15 :00 October 2012 Order Number: 327879-001US Offset Start: 56h Offset End: 57h Power Well: Core Bit Description Sticky MCH Temperature Limit -- programmed by BIOS. Scratchpad register for software. Has no functionality in PCH. Bit Reset Value Bit Access 0000h RO Intel(R) Communications Chipset 89xx Series - Datasheet 831 17.2.3.34 Offset 58h: MTV--MCH Temperature Value Table 17-80. Offset 58h: MTV--MCH Temperature Value Description: View: PCI Size: 64 bit Bit Range BAR: TBARB Default: 0000000000000000h Bit Acronym Offset Start: 58h Offset End: 5Fh Power Well: Core Bit Description Sticky MCH Temperature Value -- ME writes the MCH temperature here after collecting it from the DMI PM_Msg. PM_Msg contains 64 bits of data. 63 :00 17.2.3.35 Bus:Device:Function: B:31:6 Bit Reset Value Bit Access 000000000000 0000h RO Offset 60h: MCPTV--MCP Temperature Value Table 17-81. Offset 60h: MCPTV--MCP Temperature Value Description: View: PCI Size: 16 bit Bit Range BAR: TBARB Default: 0000h Bit Acronym Offset Start: 60h Offset End: 61h Power Well: Core Bit Description Sticky MCP Temperature Value -- ME writes the MCP temperature here after reading PECI for CPU temperature. Bits [7:0] represent the max values of the CPU. This is byte 1 of the EC data. Bits[15:8] are reserved. 15 :00 17.2.3.36 Bus:Device:Function: B:31:6 Bit Reset Value Bit Access 0000h RO Offset 64h: MMPC--Max MCH Power Clamp Table 17-82. Offset 64h: MMPC--Max MCH Power Clamp Description: View: PCI Size: 16 bit Bit Range 15 :00 BAR: TBARB Bus:Device:Function: B:31:6 Default: 0000h Bit Acronym Power Well: Core Bit Description Max MCH Power Clamp -- This register is RW to host and has no functionality in PCH. Intel(R) Communications Chipset 89xx Series - Datasheet 832 Offset Start: 64h Offset End: 65h Sticky Bit Reset Value Bit Access 0000h RW October 2012 Order Number: 327879-001US 17.0 17.2.3.37 Offset 66h: MMCPPC--Max MCP Power Clamp Table 17-83. Offset 66h: MMCPPC--Max MCP Power Clamp Description: View: PCI Size: 16 bit Bit Range BAR: TBARB Default: 0000h Bit Acronym Offset Start: 66h Offset End: 67h Power Well: Core Bit Description Sticky Max MMCPPC Power Clamp -- This register is RW to host and has no functionality in PCH. 15 :00 17.2.3.38 Bus:Device:Function: B:31:6 Bit Reset Value Bit Access 0000h RW Offset 82h: TS0PIEN--Thermal Sensor 0 PCI Interrupt Event Enables This register controls the conditions that result in the PCI Interrupt signal from the Thermal Sensor (TS) logic to assert. Note: There is a separate enable register per sensor. Table 17-84. Offset 82h: TS0PIEN--Thermal Sensor 0 PCI Interrupt Event Enables Description: View: PCI Size: 8 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Bit Acronym Offset Start: 82h Offset End: 82h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 07 Aux2 High-to-Low Enable -- See the description for bit 6. 0b RW 06 Catastrophic High-to-Low Enable -- When set to 1, the thermal sensor logic asserts the Thermal logic PCI INTx signal when the corresponding status bit is set in the Thermal Error Status register. When cleared, the corresponding status bit does not result in PCI INTx. 0b RW 05 Hot High-to-Low Enable -- See the description for bit 6. 0b RW 04 Aux High-to-Low Enable -- See the description for bit 6. 0b RW 03 Aux2 Low-to-High Event -- See the description for bit 6. 0b RW 02 Catastrophic Low-to-High Enable -- See the description for bit 6. 0b RW 01 Hot Low-to-High Event -- See the description for bit 6. 0b RW 00 Aux Low-to-High Event -- See the description for bit 6. 0b RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 833 17.2.3.39 Offset 83h: TS0LOCK--Thermal Sensor 0 Register Lock Controls This register permits BIOS to lock the contents of the Catastrophic Trip controls and Hot Trip controls. The Catastrophic Trip controls are locked with TSCO bit 7. Also, bit 0 contains the TBCNTEN bit that controls Thermal Burden Counter enable/disable. Table 17-85. Offset 83h: TS0LOCK--Thermal Sensor 0 Register Lock Controls Description: View: PCI Size: 8 bit BAR: TBARB Default: 00h Bit Range Bit Acronym 07 :03 Reserved 02 01 :00 17.2.3.40 Bus:Device:Function: B:31:6 LBH Reserved Offset Start: 83h Offset End: 83h Power Well: Core Bit Reset Value Bit Access Reserved 0h RW Hot Trip Lock -- This bit, when written to a 1, locks the Hot Trip programming interface, which is bits 15:8 of TSTTP. This bit may only be set to a 0 by a Host Partitioned Reset (bb_rst) reset. Writing a 0 to this bit has no effect. CF9 warm reset is a Host Partitioned Reset. 0b RWL 00b RW Bit Description Sticky Reserved Offset 98h: STS--SMBus Turbo Status Anytime ME writes this register an interrupt must be triggered. This register is updated by ME when it receives an MCTP msg on SMBus from the external controller. When ME moves the data from that msg and writes it to this register, an interrupt, if enabled, needs to be signaled. The write will set a bit in the Turbo Interrupt Status register, which then may cause the interrupt. Table 17-86. Offset 98h: STS--SMBus Turbo Status Description: View: PCI Size: 32 bit Bit Range 31 :01 00 BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Bit Acronym Power Well: Core Bit Reset Value Bit Access STS bits [31:1]. Bits [47:32] of STS are in TC1 register. For SW bit usage, See the MPC spec. This is received from the EC when the EC does the Write STS Preferences Command 00000000h RO New Valid Value. STS[0] ME can set this bit by writing a 1 to it. Writing a 0 has no effect. The bit is cleared when the host writes to the SMBus Event Clear[0]. 0b RO Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 834 Offset Start: 98h Offset End: 9Bh Sticky October 2012 Order Number: 327879-001US 17.0 17.2.3.41 Offset 9Ch: SEC--SMBus Event Clear Table 17-87. Offset 9Ch: SEC--SMBus Event Clear Description: View: PCI Size: 8 bit BAR: TBARB Default: 00h Bit Range Bit Acronym 07 :01 Reserved Offset Start: 9Ch Offset End: 9Ch Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00h When the Host writes a 1 to this bit, it clears bit [0] of the SMBus Turbo Status Register. 00 17.2.3.42 Bus:Device:Function: B:31:6 0b WO Offset A4h: TC3--Thermal Compares 3 Table 17-88. Offset A4h: TC3--Thermal Compares 3 Description: View: PCI Size: 32 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Bit Acronym Power Well: Core Bit Description Sticky 31 :16 CPUTCUL Thermal Compare Upper Limit: -- This is the upper limit used to compare against the CPU's temperature. If the CPU's temperature is greater than this value, then the alert GPIO is asserted. Bits [31:0] of this register are set when the EC does the Write Cpu Temp Limits transaction. 15 :00 CPUTCLL Thermal Compare Lower Limit: -- This is the lower limit used to compare against the CPU's temperature. If the CPU's temperature is lower than this value, then the alert GPIO is asserted. October 2012 Order Number: 327879-001US Offset Start: A4h Offset End: A7h Bit Reset Value Bit Access 0000h RO 0000h RO Intel(R) Communications Chipset 89xx Series - Datasheet 835 17.2.3.43 Offset A8h: TC1--Thermal Compares 1 Table 17-89. Offset A8h: TC1--Thermal Compares 1 Description: View: PCI Size: 32 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Bit Acronym Power Well: Core Bit Description Sticky STS[47:32] -- SMbus Status is a total of 48 bits. This field is bits 47:32. Bits 31:0 are in the STS register. This is received from the EC when the EC does the Write STS Preferences Command 31 :16 Offset Start: A8h Offset End: ABh Bit Reset Value Bit Access 0000h RO 15 :08 Reserved Reserved 00h 07 :00 Reserved Reserved 00h 17.2.3.44 Offset ACh: TC2--Thermal Compares 2 Table 17-90. Offset ACh: TC2--Thermal Compares 2 Description: View: PCI Size: 32 bit BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Power Well: Core Bit Reset Value Bit Access DIMM Thermal Compare Upper Limit: Upper limit used to compare against the DIMM's temperature. If the DIMM's temperature is greater than this value, then the alert GPIO is asserted. 00h RO DIMM Thermal Compare Lower Limit: Lower limit used to compare against the DIMM's temperature. If the DIMM's temperature is lower than this value, then the alert GPIO is asserted. Bits [31:16] of this register are set when the EC does the Write DIMM Temp Limits transaction. 00h RO PCHTCUL PCH Thermal Compare Upper Limit: Upper limit used to compare against the PCH's temperature. If the PCH's temperature is greater than this value, then the alert GPIO is asserted. 00h RO PCHTCLL PCH Thermal Compare Lower Limit: Lower limit used to compare against the PCH's temperature. If the PCH's temperature is lower than this value, then the alert GPIO is asserted. Bits [15:0] of this register are set when the EC does the Write PCH Temp Limits transaction. 00h RO Bit Range Bit Acronym Bit Description 31 :24 DIMMTCUL 23 :16 DIMMTCLL 15 :08 07 :00 Intel(R) Communications Chipset 89xx Series - Datasheet 836 Offset Start: ACh Offset End: AFh Sticky October 2012 Order Number: 327879-001US 17.0 17.2.3.45 Offset B0h: DIMM0--CPU0 DIMM Data This register is used by ME FW to store the temperature values of DIMMs connected to CPU0. Table 17-91. Offset B0h: DIMM0--CPU0 DIMM Data (Sheet 1 of 2) Descri ption: View: PCI Size: 32 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Bit Acronym Offset Start: B0h Offset End: B3h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 00h RO 00h RO PDP Platform: CPU0/CH3_DIMM_TEMP_VAL ME writes the hottest CPU0/CH3 DIMM Data here after collecting it over the PECI Bus. This is in absolute degrees C. Note: 31 :24 This is byte 9 of EC SMBus Data for PDP Platform PCH Mode Read operation SDP Platform: DIMM3 TEMP_VAL ME writes the DIMM3 Data here after collecting it over SMBus. This is in absolute degrees C. Note: This is byte 9 of EC SMBus Data for SDP Platform PCH Mode Read operation PDP Platform: CPU0/CH2_DIMM_TEMP_VAL ME writes the hottest CPU0/CH2 DIMM Data here after collecting it over the PECI Bus. This is in absolute degrees C. Note: 23 :16 This is byte 8 of EC SMBus Data for PDP Platform PCH Mode Read operation SDP Platform: DIMM2 TEMP_VAL ME writes the DIMM2 Data here after collecting it over SMBus. This is in absolute degrees C. Note: October 2012 Order Number: 327879-001US This is byte 8 of EC SMBus Data for SDP Platform PCH Mode Read operation Intel(R) Communications Chipset 89xx Series - Datasheet 837 Table 17-91. Offset B0h: DIMM0--CPU0 DIMM Data (Sheet 2 of 2) Descri ption: View: PCI Size: 32 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Bit Acronym Offset Start: B0h Offset End: B3h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 00h RO 00h RO PDP Platform: CPU0/CH1_DIMM_TEMP_VAL ME writes the hottest CPU0/CH1 DIMM Data here after collecting it over the PECI Bus. This is in absolute degrees C. Note: 15 :08 This is byte 7 of EC SMBus Data for PDP Platform PCH Mode Read operation SDP Platform: DIMM1 TEMP_VAL ME writes the DIMM1 Data here after collecting it over SMBus. This is in absolute degrees C. Note: This is byte 7 of EC SMBus Data for SDP Platform PCH Mode Read operation PDP Platform: CPU0/CH0_DIMM_TEMP_VAL ME writes the hottest CPU0/CH0 DIMM Data here after collecting it over the PECI Bus. This is in absolute degrees C. Note: 07 :00 This is byte 6 of EC SMBus Data for PDP Platform PCH Mode Read operation SDP Platform: DIMM0 TEMP_VAL ME writes the DIMM0 Data here after collecting it over SMBus. This is in absolute degrees C. Note: This is byte 6 of EC SMBus Data for SDP Platform PCH Mode Read operation Intel(R) Communications Chipset 89xx Series - Datasheet 838 October 2012 Order Number: 327879-001US 17.0 17.2.3.46 Offset B4h: DIMM1--CPU1 DIMM Data This register is used by ME FW to store the temperature values of DIMMs connected to CPU1. Table 17-92. Offset B4h: DIMM1--CPU1 DIMM Data Description: View: PCI Size: 32 bit BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Offset Start: B4h Offset End: B7h Power Well: Core Bit Reset Value Bit Access 31 :24 CPU1/CH3_DIMM_TEMP_VAL ME writes the hottest for CPU1/CH3 Data here after collecting it over the PECI Bus. This is in absolute degrees C. This is byte 25 of EC SMBus Data for PDP Platform PCH Mode Read operation 00h RO 23 :16 CPU1/CH2_DIMM_TEMP_VAL ME writes the hottest CPU1/CH2 DIMM Data here after collecting it over the PECI Bus. This is in absolute degrees C. This is byte 24 of EC SMBus Data for PDP Platform PCH Mode Read operation 00h RO 15 :08 CPU1/CH1_DIMM_TEMP_VAL ME writes the hottest CPU1/CH1 DIMM Data here after collecting it over the PECI Bus. This is in absolute degrees C. This is byte 23 of EC SMBus Data for PDP Platform PCH Mode Read operation 00h RO 07 :00 CPU1/CH0_DIMM_TEMP_VAL ME writes the hottest CPU1/CH0 DIMM Data here after collecting it over the PECI Bus. This is in absolute degrees C. This is byte 22 of EC SMBus Data for PDP Platform PCH Mode Read operation 00h RO Bit Range Bit Acronym October 2012 Order Number: 327879-001US Bit Description Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 839 17.2.3.47 Offset B8h: DIMMID--DIMM ID This register is used by ME FW to indicate the hottest DIMMs that have their temperature values stored in DIMM0 (Section 17.2.3.45) and DIMM1 (Section 17.2.3.46) Data registers for CPU0 and CPU1. This register is used in PDP platform applications only. Table 17-93. Offset B8h: DIMMID--DIMM ID (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 :16 BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Bit Acronym Offset Start: B8h Offset End: BBh Power Well: Core Bit Description Reserved Sticky Bit Reset Value Bit Access 0000h RO 0h RO 0h RO 0h RO 0h RO CPU1/CH3_DIMM_ID DIMM ID for the hottest DIMM for CPU1/CH3. This is the ID for the corresponding DIMM in byte 25 of EC SMBus Data for PDP Platform PCH Mode Read operation 15 14 00 = 01 = 10 = 11 = TS. CPU1/CH3/DIMM0 CPU1/CH3/DIMM1 CPU1/CH3/DIMM2 No DIMM populated in CPU1/CH3 or DIMMs have no CPU1/CH2_DIMM_ID DIMM ID for the hottest DIMM for CPU1/CH2. This is the ID for the corresponding DIMM in byte 24 of EC SMBus Data for PDP Platform PCH Mode Read operation 13 :12 00 = 01 = 10 = 11 = TS. CPU1/CH2/DIMM0 CPU1/CH2/DIMM1 CPU1/CH2/DIMM2 No DIMM populated in CPU1/CH2 or DIMMs have no CPU1/CH1_DIMM_ID DIMM ID for the hottest DIMM for CPU1/CH1. This is the ID for the corresponding DIMM in byte 23 of EC SMBus Data for PDP Platform PCH Mode Read operation 11 :10 00 = 01 = 10 = 11 = TS. CPU1/CH1/DIMM0 CPU1/CH1/DIMM1 CPU1/CH1/DIMM2 No DIMM populated in CPU1/CH1 or DIMMs have no CPU1/CH0_DIMM_ID DIMM ID for the hottest DIMM for CPU1/CH0. This is the ID for the corresponding DIMM in byte 22 of EC SMBus Data for PDP Platform PCH Mode Read operation 09 :08 00 = 01 = 10 = 11 = TS. CPU1/CH0/DIMM0 CPU1/CH0/DIMM1 CPU1/CH0/DIMM2 No DIMM populated in CPU1/CH0 or DIMMs have no Intel(R) Communications Chipset 89xx Series - Datasheet 840 October 2012 Order Number: 327879-001US 17.0 Table 17-93. Offset B8h: DIMMID--DIMM ID (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00000000h Bit Acronym Offset Start: B8h Offset End: BBh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0h RO 0h RO 0h RO 0h RO CPU0/CH3_DIMM_ID DIMM ID for the hottest DIMM for CPU0/CH3. This is the ID for the corresponding DIMM in byte 9 of EC SMBus Data for PDP Platform PCH Mode Read operation 07 :06 00 = 01 = 10 = 11 = TS. CPU0/CH3/DIMM0 CPU0/CH3/DIMM1 CPU0/CH3/DIMM2 No DIMM populated in CPU0/CH3 or DIMMs have no CPU0/CH2_DIMM_ID DIMM ID for the hottest DIMM for CPU0/CH2. This is the ID for the corresponding DIMM in byte 8 of EC SMBus Data for PDP Platform PCH Mode Read operation 05 :04 00 = 01 = 10 = 11 = TS. CPU0/CH2/DIMM0 CPU0/CH2/DIMM1 CPU0/CH2/DIMM2 No DIMM populated in CPU0/CH2 or DIMMs have no CPU0/CH1_DIMM_ID DIMM ID for the hottest DIMM for CPU0/CH1. This is the ID for the corresponding DIMM in byte 7 of EC SMBus Data for PDP Platform PCH Mode Read operation 03 :02 00 = 01 = 10 = 11 = TS. CPU0/CH1/DIMM0 CPU0/CH1/DIMM1 CPU0/CH1/DIMM2 No DIMM populated in CPU0/CH1 or DIMMs have no CPU0/CH0_DIMM_ID DIMM ID for the hottest DIMM for CPU0/CH0. This is the ID for the corresponding DIMM in byte 6 of EC SMBus Data for PDP Platform PCH Mode Read operation 01 :00 00 = 01 = 10 = 11 = TS. October 2012 Order Number: 327879-001US CPU0/CH0/DIMM0 CPU0/CH0/DIMM1 CPU0/CH0/DIMM2 No DIMM populated in CPU0/CH0 or DIMMs have no Intel(R) Communications Chipset 89xx Series - Datasheet 841 17.2.3.48 Offset BCh: ECPCLAMP--EC Power Clamp Data Table 17-94. Offset BCh: ECPCLAMP--EC Power Clamp Data Description: View: PCI Size: 32 bit BAR: TBARB Default: 00000000h Bit Range Bit Acronym 31 :16 Reserved 15 :00 17.2.3.49 Bus:Device:Function: B:31:6 Offset Start: BCh Offset End: BFh Power Well: Core Bit Reset Value Bit Access Reserved 0000h RW MPC CPU Power Clamp Data from EC: -- This contains the 16 bits that were received from the EC for the CPU power clamp transaction. This data was used to send a Clamp Power PECI transaction to the cpu. 0000h RW Bit Description Sticky Offset C2h: TS1PIEN--Thermal Sensor 1 PCI Interrupt Event Enables This register controls the conditions that result in the PCI Interrupt signal from the Thermal Sensor (TS) logic to assert. Note: There is a separate enable register per sensor. Table 17-95. Offset C2h: TS1PIEN--Thermal Sensor 1PCI Interrupt Event Enables Description: View: PCI Size: 8 bit Bit Range BAR: TBARB Bus:Device:Function: B:31:6 Default: 00h Bit Acronym Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 07 Aux2 High-to-Low Enable -- See the description for bit 6. 0b RW 06 Catastrophic High-to-Low Enable -- When set to 1, the thermal sensor logic asserts the Thermal logic PCI INTx signal when the corresponding status bit is set in the Thermal Error Status register. When cleared, the corresponding status bit does not result in PCI INTx. 0b RW 05 Hot High-to-Low Enable -- See the description for bit 6. 0b RW 04 Aux High-to-Low Enable -- See the description for bit 6. 0b RW 03 Aux2 Low-to-High Event -- See the description for bit 6. 0b RW 02 Catastrophic Low-to-High Enable -- See the description for bit 6. 0b RW 01 Hot Low-to-High Event -- See the description for bit 6. 0b RW 00 Aux Low-to-High Event -- See the description for bit 6. 0b RW Intel(R) Communications Chipset 89xx Series - Datasheet 842 Offset Start: C2h Offset End: C2h October 2012 Order Number: 327879-001US 17.0 17.2.3.50 Offset C3h: TS1LOCK--Thermal Sensor 1 Register Lock Controls This register permits BIOS to lock the contents of the Catastrophic Trip controls and Hot Trip controls. The Catastrophic Trip controls are locked with TSCO bit 7. Also, bit 0 contains the TBCNTEN bit that controls Thermal Burden Counter enable/disable. Table 17-96. Offset C3h: TS1LOCK--Thermal Sensor 1 Register Lock Controls Description: View: PCI Size: 8 bit BAR: TBARB Default: 00h Bit Range Bit Acronym 07 :03 Reserved 02 01 :00 17.2.3.51 Bus:Device:Function: B:31:6 LBH Reserved Offset Start: C3h Offset End: C3h Power Well: Core Bit Reset Value Bit Access Reserved 0h RW Hot Trip Lock -- This bit, when written to a 1, locks the Hot Trip programming interface, which is bits 15:8 of TSTTP. This bit may only be set to a 0 by a Host Partitioned Reset (bb_rst) reset. Writing a 0 to this bit has no effect. CF9 warm reset is a Host Partitioned Reset. 0b RWL 00b RW Bit Description Sticky Reserved Offset D8h: ITV--Internal Temperature Values Table 17-97. Offset D8h: ITV--Internal Temperature Values Description: View: PCI Size: 32 bit BAR: TBARB Default: 00000000h Bit Range Bit Acronym 31 :24 Reserved 23 :16 15 :08 Bus:Device:Function: B:31:6 Reserved 07 :00 Offset Start: D8h Offset End: DBh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved 00h RW Sequence Number -- Provides a sequence number which can be used by the host to detect if ME has hung. The value will roll over to 0 from 0xF. The count is updated at the MPC rate, which is expected to be 200 msec. Host SW can check this value and if it isn't incremented over a second or so, should assume that the ME MCP code is hung. NOTE: if ME is reset, then this value will not change during the reset. After the reset is done, which may take up to 30 seconds, ME may be on line again and this value will start incrementing; indicating that the thermal values are valid again. This is the value of EC byte 10 00h RO Reserved 00h RW PCH Temperature -- ME writes the PCH temperature here. This is in absolute degrees C. This is byte 2 of the EC data. 00h RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 843 18 Intel(R) Management Engine Interface (MEI - B0:D22:F0/F1) 18.1 First Intel Management Engine Interface (MEI1) Configuration Registers (MEI1 -- B0:D22:F0) / Table 18-1. MEI1 Configuration Registers Address Map (MEI1--B0:D22:F0) (Sheet 1 of 2) Offset Mnemonic & Register Name Type 00h-01h "VID--Vendor Identification Register" 8086h RO 02h-03h "DID--Device Identification Register" See register description RO 04h-05h "PCICMD--PCI Command Register" 0000h R/W, RO 06h-07h "PCISTS--PCI Status Register" 0010h RO See register description RO 0C8000h RO 00h RO 00000000 00000004h R/W, RO 08h 09h-0Bh 0Eh "RID--Revision Identification Register" "CC--Class Code Register" "HTYPE--Header Type Register" 10h-17h "MEI1_MBAR--MEI1 MMIO Base Address Register" 2Ch-2Dh "SVID--Subsystem Vendor ID Register" 0000h R/WO 2Eh-2Fh "SID--Subsystem ID Register" 0000h R/WO 50h RO 34h "CAPP--Capabilities List Pointer Register" 3Ch-3Dh "INTR--Interrupt Information Register" 0000h R/W, RO 3Eh-3Fh "MLMG--Maximum Latency/Minimum Grant Register" 0000h RO 40h-43h "HFS--Host Firmware Status Register" 00000000h RO 44h-47h "ME_UMA--Management Engine UMA Register" 00000000h RO 48-4Bh "GMES--General ME Status" 00000000h RO 4Ch-4Fh "H_GS--Host General Status" 00000000h RO 50h-51h "PID--PCI Power Management Capability ID Register" 6001h RO 52h-53h "PC--PCI Power Management Capabilities Register" C803h RO 54h-55h "PMCS--PCI Power Management Control and Status Register" 0008h R/WC, R/W, RO 8Ch-8Dh "MID--Message Signaled Interrupt Identifiers Register" 0005h RO 8Eh-8Fh "MC--Message Signaled Interrupt Message Control Register" 0080h R/W, RO 90h-93h "MA--Message Signaled Interrupt Message Address Register" 00000000h R/W, RO 94h-97h "MUA--Message Signaled Interrupt Upper Address Register" 00000000h R/W 98h-99h "MD--Message Signaled Interrupt Message Data Register" 0000h R/W Intel(R) Communications Chipset 89xx Series - Datasheet 844 Default October 2012 Order Number: 327879-001US 18 Table 18-1. MEI1 Configuration Registers Address Map (MEI1--B0:D22:F0) (Sheet 2 of 2) Offset Mnemonic & Register Name Default Type A0h "HIDM--MEI Interrupt Delivery Mode" 00h R/W BCh-BFh "HERES--MEI Extend Register Status" 40000000h RO C0h-DFh "HERX--MEI Extend Register DWX" 00000000h RO 18.1.1 VID--Vendor Identification Register Address Offset: 00h-01h Default Value: 8086h Bit 15:0 18.1.2 Attribute: Size: RO 16 bits Description Vendor ID (VID) -- RO. This is a 16-bit value assigned to Intel. DID--Device Identification Register Address Offset: 02h-03h Default Value: See bit description Bit 15:0 Attribute: Size: RO 16 bits Description Device ID (DID) -- RO. This is a 16-bit value assigned to the Intel Management Engine Interface controller. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 845 18.1.3 PCICMD--PCI Command Register Address Offset: 04h-05h Default Value: 0000h Bit 15:11 Attribute: Size: R/W, RO 16 bits Description Reserved 10 Interrupt Disable (ID) -- R/W. Disables this device from generating PCI line based interrupts. This bit does not have any effect on MSI operation. 9:3 Reserved Bus Master Enable (BME)-- R/W.: 2 Controls the Intel MEI host controller's ability to act as a system memory master for data transfers. When this bit is cleared, Intel ME bus master activity stops and any active DMA engines return to an idle condition. This bit is made visible to firmware through the H_PCI_CSR register, and changes to this bit may be configured by the H_PCI_CSR register to generate an ME MSI. When this bit is 0, Intel MEI is blocked from generating MSI to the host CPU. NOTE: This bit does not block Intel MEI accesses to ME-UMA, i.e. writes or reads to the host and ME circular buffers through the read window and write window registers still cause ME backbone transactions to ME-UMA. Memory Space Enable (MSE) -- R/W. Controls access to the Intel ME's memory mapped register space. 1 0 18.1.4 0 = Disable. Memory cycles within the range specified by the memory base and limit registers are master aborted. 1 = Enable. Allows memory cycles within the range specified by the memory base and limit registers accepted. Reserved PCISTS--PCI Status Register Address Offset: 06h-07h Default Value: 0010h Bit 15:5 4 Attribute: Size: RO 16 bits Description Reserved Capabilities List (CL) -- RO. Indicates the presence of a capabilities list, hardwired to 1. Interrupt Status (IS) -- RO. Indicates the interrupt status of the device. 3 2:0 0 = Interrupt is de-asserted. 1 = Interrupt is asserted. Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 846 October 2012 Order Number: 327879-001US 18 18.1.5 RID--Revision Identification Register Offset Address: 08h Default Value: See bit description Bit 7:0 18.1.6 Attribute: Size: Description Revision ID -- RO. 8-bit value indicating the stepping of the SATA Controller hardware CC--Class Code Register Address Offset: 09h-0Bh Default Value: 078000h Attribute: Size: RO 24 bits Bit Description 23:16 Base Class Code (BCC) -- RO. Indicates the base class code of the Intel MEI device. 15:8 7:0 18.1.7 RO 8 bits Sub Class Code (SCC) -- RO. Indicates the sub class code of the Intel MEI device. Programming Interface (PI) -- RO. Indicates the programming interface of the Intel MEI device. HTYPE--Header Type Register Address Offset: 0Eh Default Value: 80h Bit 7 6:0 Attribute: Size: RO 8 bits Description Multi-Function Device (MFD) -- RO. Indicates the Intel MEI host controller is part of a multifunction device. Header Layout (HL) -- RO. Indicates that the Intel MEI uses a target device layout. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 847 18.1.8 MEI1_MBAR--MEI1 MMIO Base Address Register Address Offset: 10h-17h Default Value: 0000000000000004h Attribute: Size: R/W, RO 64 bits This register allocates space for the MEI1 memory mapped registers. Bit 63:4 Base Address (BA) -- R/W. Software programs this field with the base address of this region. 3 Prefetchable Memory (PM) -- RO. Indicates that this range is not pre-fetchable. 2:1 0 18.1.9 Description Type (TP) -- RO. Set to 10b to indicate that this range can be mapped anywhere in 64-bit address space. Resource Type Indicator (RTE) -- RO. Indicates a request for register memory space. SVID--Subsystem Vendor ID Register Address Offset: 2Ch-2Dh Default Value: 0000h 18.1.10 Description 15:0 Subsystem Vendor ID (SSVID) -- R/WO. Indicates the sub-system vendor identifier. This field should be programmed by BIOS during boot-up. Once written, this register becomes Read Only. This field can only be cleared by PLTRST#. SID--Subsystem ID Register Attribute: Size: R/WO 16 bits Bit Description 15:0 Subsystem ID (SSID) -- R/WO. Indicates the sub-system identifier. This field should be programmed by BIOS during boot-up. Once written, this register becomes Read Only. This field can only be cleared by PLTRST#. CAPP--Capabilities List Pointer Register Address Offset: 34h Default Value: 50h Attribute: Size: RO 8 bits Bit Description 7:0 Capabilities Pointer (PTR) -- RO. Indicates that the pointer for the first entry in the capabilities list is at 50h in configuration space. Intel(R) Communications Chipset 89xx Series - Datasheet 848 R/WO 16 bits Bit Address Offset: 2Eh-2Fh Default Value: 0000h 18.1.11 Attribute: Size: October 2012 Order Number: 327879-001US 18 18.1.12 INTR--Interrupt Information Register Address Offset: 3Ch-3Dh Default Value: 0100h Bit 15:8 7:0 18.1.13 R/W, RO 16 bits Description Interrupt Pin (IPIN) -- RO. This indicates the interrupt pin the Intel MEI host controller uses. The value of 01h selects INTA# interrupt pin. Interrupt Line (ILINE) -- R/W. Software written value to indicate which interrupt line (vector) the interrupt is connected to. No hardware action is taken on this register. MLMG--Maximum Latency/Minimum Grant Register Address Offset: 3Eh-3Fh Default Value: 0000h Bit 15:0 18.1.14 Attribute: Size: Attribute: Size: RO 16 bits Description Maximum Latency/Minimum Grant (MLMG)-- RO. Not used. Hardwired to 0000h. HFS--Host Firmware Status Register Address Offset: 40h-43h Default Value: 00000000h Attribute: Size: RO 32 bits Bit Description 31:0 Host Firmware Status (HFS) -- RO. This register field is used by Firmware to reflect the operating environment to the host. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 849 18.1.15 ME_UMA--Management Engine UMA Register Address Offset: 44h-47h Default Value: 80000000h Bit 31 30:7 16 15:6 Attribute: Size: RO 32 bits Description Reserved-- RO. Hardwired to 1. Can be used by host software to discover that this register is valid. Reserved ME UMA Size Valid - RO. This bit indicates that FW has written to the MUSZ field. Reserved ME UMA Size (MUSZ) - RO. This field reflect ME Firmware's desired size of MEUMA memory region. This field is set by ME firmware prior to core power bringup allowing BIOS to initialize memory. 000000b = 0 MB, No memory allocated to MEUMA 5:0 000001b = 1 MB 000010b = 2 MB 000100b = 4 MB 001000b = 8 MB 010000b = 16 MB 100000b = 32 MB 18.1.16 GMES--General ME Status Address Offset: 48h-4Bh Default Value: 00000000h Bit 31:0 18.1.17 RO 32 bits Description General ME Status (ME_GS)-- RO. This field is populated by ME. H_GS--Host General Status Address Offset: 4Ch-4Fh Default Value: 00000000h Bit 31:0 Attribute: Size: RO 32 bits Description Host General Status(H_GS)-- RO. General Status of Host, this field is not used by Hardware Intel(R) Communications Chipset 89xx Series - Datasheet 850 Attribute: Size: October 2012 Order Number: 327879-001US 18 18.1.18 PID--PCI Power Management Capability ID Register Address Offset: 50h-51h Default Value: 6001h RO 16 bits Bit Description 15:8 Next Capability (NEXT) -- RO. Value of 60h indicates the location of the next pointer. 7:0 18.1.19 Attribute: Size: Capability ID (CID) -- RO. Indicates the linked list item is a PCI Power Management Register. PC--PCI Power Management Capabilities Register Address Offset: 52h-53h Default Value: C803h Attribute: Size: RO 16 bits Bit Description 15:11 PME_Support (PSUP) -- RO. This five-bit field indicates the power states in which the function may assert PME#. Intel MEI can assert PME# from any D-state except D1 or D2 which are not supported by Intel MEI. 10:9 8:6 Reserved Aux_Current (AC) -- RO. Reports the maximum Suspend well current required when in the D3cold state. Value of 00b is reported. 5 Device Specific Initialization (DSI) -- RO. Indicates whether device-specific initialization is required. 4 Reserved 3 PME Clock (PMEC) -- RO. Indicates that PCI clock is not required to generate PME#. 2:0 Version (VS) -- RO. Hardwired to 011b to indicate support for Revision 1.2 of the PCI Power Management Specification. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 851 18.1.20 PMCS--PCI Power Management Control and Status Register Address Offset: 54h-55h Default Value: 0008h Bit 15 14:9 8 Attribute: Size: R/WC, R/W, RO 16 bits Description PME Status (PMES) -- R/WC. Bit is set by ME Firmware. Host software clears bit by writing `1' to bit. Reserved PME Enable (PMEE) -- R/W. This bit is read/write and is under the control of host SW. It does not directly have an effect on PME events. However, this bit is shadowed so ME FW can monitor it. ME FW will not cause the PMES bit to transition to 1 while the PMEE bit is 0, indicating that host SW had disabled PME. This bit is reset when PLTRST# asserted. 7:4 3 2 Reserved No_Soft_Reset (NSR) -- RO. This bit indicates that when the Intel MEI host controller is transitioning from D3hot to D0 due to a power state command, it does not perform an internal reset. Configuration context is preserved. Reserved Power State (PS) -- R/W. This field is used both to determine the current power state of the Intel MEI host controller and to set a new power state. The values are: 00 = D0 state (default) 1:0 11 = D3hot state The D1 and D2 states are not supported for the Intel MEI host controller. When in the D3hot state, the Intel ME's configuration space is available, but the register memory spaces are not. Additionally, interrupts are blocked. 18.1.21 MID--Message Signaled Interrupt Identifiers Register Address Offset: 8Ch-8Dh Default Value: 0005h RO 16 bits Bit Description 15:8 Next Pointer (NEXT) -- RO. Value of 00h indicates that this is the last item in the list. 7:0 Capability ID (CID) -- RO. Capabilities ID indicates MSI. Intel(R) Communications Chipset 89xx Series - Datasheet 852 Attribute: Size: October 2012 Order Number: 327879-001US 18 18.1.22 MC--Message Signaled Interrupt Message Control Register Address Offset: 8Eh-8Fh Default Value: 0080h Bit 15:8 7 6:1 0 18.1.23 Reserved. 64 Bit Address Capable (C64) -- RO. Specifies that function is capable of generating 64-bit messages. Reserved MSI Enable (MSIE) -- R/W. If set, MSI is enabled and traditional interrupt pins are not used to generate interrupts. MA--Message Signaled Interrupt Message Address Register Bit 31:2 1:0 Attribute: Size: R/W, RO 32 bits Description Address (ADDR) -- R/W. Lower 32 bits of the system specified message address, always DW aligned. Reserved. MUA--Message Signaled Interrupt Upper Address Register Address Offset: 94h-97h Default Value: 00000000h Bit 31:0 18.1.25 R/W, RO 16 bits Description Address Offset: 90h-93h Default Value: 00000000h 18.1.24 Attribute: Size: Attribute: Size: R/W 32 bits Description Upper Address (UADDR) -- R/W. Upper 32 bits of the system specified message address, always DW aligned. MD--Message Signaled Interrupt Message Data Register Address Offset: 98h-99h Default Value: 0000h Bit 15:0 Attribute: Size: R/W 16 bits Description Data (DATA) -- R/W. This 16-bit field is programmed by system software if MSI is enabled. Its content is driven during the data phase of the MSI memory write transaction. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 853 18.1.26 HIDM--MEI Interrupt Delivery Mode Address Offset: A0h Default Value: 00h Bit 7:2 1 Attribute: Size: R/W 8 bits Description Reserved. MEI Interrupt Delivery Mode (HIDM) -- R/W. These bits control what type of interrupt the Intel MEI will send when ARC writes to set the M_IG bit in AUX space. They are interpreted as follows: 00 = Generate Legacy or MSI interrupt 01 = Generate SCI 10 = Generate SMI 0 18.1.27 Synchronous SMI Occurrence (SSMIO) -- R/WC. This bit is used by firmware to indicate that a synchronous SMI source has been triggered. Host BIOS SMM handler can use this bit as status indication and clear it once processing is completed. A write of 1 from host SW clears this status bit. NOTE: It is possible that an async SMI has occurred prior to sync SMI occurrence and when the BIOS enters the SMM handler, it is possible that both bit 0 and bit 1 of this register could be set. HERES--MEI Extend Register Status Address Offset: BCh-BFh Default Value: 00h Bit Attribute: Size: RO 32 bits Description Extend Register Valid (ERV). 31 Set by firmware after all firmware has been loaded. If ERA field is SHA-1, the result of the extend operation is in HER:5-1. If ERA field is SHA-256, the result of the extend operation is in HER:8-1. Extend Feature Present (EFP). 30 29:4 This bit is hardwired to 1 to allow driver software to easily detect the chipset supports the Extend Register FW measurement feature. Reserved Extend Register Algorithm (ERA). 3:0 This field indicates the hash algorithm used in the FW measurement extend operations. Encodings are: 0h = SHA-1 2h = SHA-256 Other values = Reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 854 October 2012 Order Number: 327879-001US 18 18.1.28 HERX--MEI Extend Register DWX Address Offset: HER1: C0h-C3h HER2: C4h-C7h HER3: C8h-CBh HER4: CCh-CFh HER5: D0h-D3h HER6: D4h-D7h HER7: D8h-DBh HER8: DCh-DFh Default Value: 00000000h Bit Attribute: RO Size: 32 bits Description Extend Register DWX (ERDWX). Nth DWORD result of the extend operation. 31:0 18.2 NOTE: Extend Operation is HER[5:1] if using SHA-1. If using SHA-2 then Extend Operation is HER[8:1] MEI1_MBAR: MEI1 MMIO Registers These MMIO registers are accessible starting at the MEI1 MMIO Base Address (MEI1_MBAR) which gets programmed into B0:D22:F0:Offset 10-17h. These registers are reset by PLTRST# unless otherwise noted. Table 18-2. MEI1 MMIO Register Address Map Offset (BAR= MEI1_MBAR) Mnemonic & Register Name Default Type 00-03h "H_CB_WW--Host Circular Buffer Write Window" 00000000h RO 04h-07h "H_CSR--Host Control Status" 02000000h RO 08h-0Bh "ME_CB_RW--ME Circular Buffer Read Window" 00000000h RO 0Ch-0Fh "ME CSR_HA--ME Control Status Host Access" 02000000h RO 18.2.1 H_CB_WW--Host Circular Buffer Write Window Address Offset: MEI1_MBAR + 00h Default Value: 00000000h Attribute: Size: RO 32 bits Bit Description 31:0 Host Circular Buffer Write Window Field (H_CB_WWF). This bit field is for host to write into its circular buffer. The host's circular buffer is located at the ME subsystem address specified in the Host CB Base Address register. This field is write only, reads will return arbitrary data. Writes to this register will increment the H_CBWP as long as ME_RDY is 1. When ME_RDY is 0, writes to this register have no effect and are not delivered to the H_CB, nor is H_CBWP incriminated. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 855 18.2.2 H_CSR--Host Control Status Address Offset: MEI1_MBAR + 04h Default Value: 02000000h Bit 31:24 Description Host Circular Buffer Depth (H_CBD) -- RO. This field indicates the maximum number of 32 bit entries available in the host circular buffer (H_CB). Host software uses this field along with the H_CBRP and H_CBWP fields to calculate the number of valid entries in the H_CB to read or # of entries available for write. This field is implemented with a "1-hot" scheme. Only one bit will be set to a "1" at a time. Each bit position represents the value n of a buffer depth of (2^n). For example, when bit# 1 is 1, the buffer depth is 2; when bit#2 is 1, the buffer depth is 4, etc. The allowed buffer depth values are 2, 4, 8, 16, 32, 64 and 128. Host CB Write Pointer (H_CBWP). Points to next location in the H_CB for host to write the data. Software uses this field along with H_CBRP and H_CBD fields to calculate the number of valid entries in the H_CB to read or number of entries available for write. 15:8 Host CB Read Pointer (H_CBRP). Points to next location in the H_CB where a valid data is available for embedded controller to read. Software uses this field along with H_CBWR and H_CBD fields to calculate the number of valid entries in the host CB to read or number of entries available for write. Reserved Must be programmed to zero 4 Host Reset (H_RST). Setting this bit to 1 will initiate a Intel MEI reset sequence to get the circular buffers into a known good state for host and ME communication. When this bit transitions from 0 to 1, hardware will clear the H_RDY and ME_RDY bits. 3 Host Ready (H_RDY). This bit indicates that the host is ready to process messages. 2 Host Interrupt Generate (H_IG). Once message(s) are written into its CB, the host sets this bit to one for the HW to set the ME_IS bit in the ME_CSR and to generate an interrupt message to ME. HW will send the interrupt message to ME only if the ME_IE is enabled. HW then clears this bit to 0. 1 Host Interrupt Status (H_IS). Hardware sets this bit to 1 when ME_IG bit is set to 1. Host clears this bit to 0 by writing a 1 to this bit position. H_IE has no effect on this bit. 0 Host Interrupt Enable (H_IE). Host sets this bit to 1 to enable the host interrupt (INTR# or MSI) to be asserted when H_IS is set to 1. ME_CB_RW--ME Circular Buffer Read Window Address Offset: MEI1_MBAR + 08h Default Value: FFFFFFFFh Attribute: Size: RO 32 bits Bit Description 31:0 ME Circular Buffer Read Window Field (ME_CB_RWF). This bit field is for host to read from the ME Circular Buffer. The ME's circular buffer is located at the ME subsystem address specified in the ME CB Base Address register. This field is read only, writes have no effect. Reads to this register will increment the ME_CBRP as long as ME_RDY is 1. When ME_RDY is 0, reads to this register have no effect, all 1s are returned, and ME_CBRP is not incremented. Intel(R) Communications Chipset 89xx Series - Datasheet 856 RO 32 bits 23:16 7:5 18.2.3 Attribute: Size: October 2012 Order Number: 327879-001US 18 18.2.4 ME CSR_HA--ME Control Status Host Access Address Offset: MEI1_MBAR + 0Ch Default Value: 02000000h Bit 31:24 23:16 15:8 7:5 4 3 2 1 0 Attribute: Size: RO 32 bits Description ME Circular Buffer Depth Host Read Access (ME_CBD_HRA). Host read only access to ME_CBD. ME CB Write Pointer Host Read Access (ME_CBWP_HRA). Host read only access to ME_CBWP. ME CB Read Pointer Host Read Access (ME_CBRP_HRA). Host read only access to ME_CBRP. Reserved ME Reset Host Read Access (ME_RST_HRA). Host read access to ME_RST. ME Ready Host Read Access (ME_RDY_HRA): Host read access to ME_RDY. ME Interrupt Generate Host Read Access (ME_IG_HRA). Host read only access to ME_IG. ME Interrupt Status Host Read Access (ME_IS_HRA). Host read only access to ME_IS. ME Interrupt Enable Host Read Access (ME_IE_HRA). Host read only access to ME_IE. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 857 18.3 Second Management Engine Interface (MEI2) Configuration Registers (MEI2--B0:D22:F1) ) Table 18-3. MEI2 Configuration Registers Address Map (MEI2--B0:D22:F1) Offset Default Type "VID--Vendor Identification Register" 8086h RO 02h-03h "DID--Device Identification Register" See register description RO 04h-05h "PCICMD--PCI Command Register" 0000h R/W, RO 06h-07h "PCISTS--PCI Status Register" 0010h RO See register description RO 00h-01h 08h 09h-0Bh 0Eh 10h-17h Mnemonic & Register Name "RID--Revision Identification Register" "CC--Class Code Register" "HTYPE--Header Type Register" "MEI_MBAR--MEI MMIO Base Address Register" RO 00h RO 00000000 00000004h R/W, RO 2Ch-2Dh "SVID--Subsystem Vendor ID Register" 0000h R/WO 2Eh-2Fh "SID--Subsystem ID Register" 0000h R/WO 34h 3Ch-3Dh "CAPP--Capabilities List Pointer Register" "INTR--Interrupt Information Register" 50h RO 0000h R/W, RO 3Eh-3Fh "MLMG--Maximum Latency/Minimum Grant Register" 0000h RO 40h-43h "HFS--Host Firmware Status Register" 00000000h RO RO 48-4Bh "GMES--General ME Status" 00000000h 4Ch-4Fh "H_GS--Host General Status" 00000000h RO 50h-51h "PID--PCI Power Management Capability ID Register" 6001h RO 52h-53h "PC--PCI Power Management Capabilities Register" C803h RO 54h-55h "PMCS--PCI Power Management Control and Status Register" 0008h R/WC, R/W, RO 8Ch-8Dh "MID--Message Signaled Interrupt Identifiers Register" 0005h RO 8Eh-8Fh "MC--Message Signaled Interrupt Message Control Register" 0080h R/W, RO 90h-93h "MA--Message Signaled Interrupt Message Address Register" 00000000h R/W, RO 94h-97h "MUA--Message Signaled Interrupt Upper Address Register" 00000000h R/W 98h-99h "MD--Message Signaled Interrupt Message Data Register" 0000h R/W A0h "HIDM--MEI Interrupt Delivery Mode" 00h R/W BC-BF "HERES--MEI Extend Register Status" 40000000h RO C0-DF "HERX--MEI Extend Register DWX" 00000000h RO 18.3.1 VID--Vendor Identification Register Address Offset: 00h-01h Default Value: 8086h Bit 15:0 Attribute: Size: RO 16 bits Description Vendor ID (VID) -- RO. This is a 16-bit value assigned to Intel. Intel(R) Communications Chipset 89xx Series - Datasheet 858 0C8000h October 2012 Order Number: 327879-001US 18 18.3.2 DID--Device Identification Register Address Offset: 02h-03h Default Value: See bit description Bit 15:0 18.3.3 Attribute: Size: RO 16 bits Description Device ID (DID) -- RO. This is a 16-bit value assigned to the Intel Management Engine Interface controller. PCICMD--PCI Command Register Address Offset: 04h-05h Default Value: 0000h Bit 15:11 10 9:3 2 Attribute: Size: R/W, RO 16 bits Description Reserved Interrupt Disable (ID) -- R/W. Disables this device from generating PCI line based interrupts. This bit does not have any effect on MSI operation. Reserved Bus Master Enable (BME)-- R/W. Controls the Intel MEI host controller's ability to act as a system memory master for data transfers. When this bit is cleared, Intel MEI bus master activity stops and any active DMA engines return to an idle condition. This bit is made visible to firmware through the H_PCI_CSR register, and changes to this bit may be configured by the H_PCI_CSR register to generate an ME MSI. When this bit is 0, Intel MEI is blocked from generating MSI to the host CPU. NOTE: This bit does not block Intel MEI accesses to ME-UMA, i.e. writes or reads to the host and ME circular buffers through the read window and write window registers still cause ME backbone transactions to ME-UMA. Memory Space Enable (MSE) -- R/W. Controls access to the Intel ME's memory mapped register space. 1 0 0 = Disable. Memory cycles within the range specified by the memory base and limit registers are master aborted. 1 = Enable. Allows memory cycles within the range specified by the memory base and limit registers accepted. Reserved October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 859 18.3.4 PCISTS--PCI Status Register Address Offset: 06h-07h Default Value: 0010h Bit 15:5 Attribute: Size: RO 16 bits Description Reserved 4 Capabilities List (CL) -- RO. Indicates the presence of a capabilities list, hardwired to 1. 3 0 = Interrupt is de-asserted. 1 = Interrupt is asserted. Interrupt Status -- RO. Indicates the interrupt status of the device. 2:0 18.3.5 Reserved RID--Revision Identification Register Offset Address: 08h Default Value: See bit description Bit 7:0 18.3.6 Attribute: Size: Description Revision ID -- RO. 8-bit value indicating the stepping of the SATA Controller hardware. CC--Class Code Register Address Offset: 09h-0Bh Default Value: 078000h RO 24 bits Description 23:16 Base Class Code (BCC) -- RO. Indicates the base class code of the Intel MEI device. 7:0 Sub Class Code (SCC) -- RO. Indicates the sub class code of the Intel MEI device. Programming Interface (PI) -- RO. Indicates the programming interface of the Intel MEI device. HTYPE--Header Type Register Address Offset: 0Eh Default Value: 80h Attribute: Size: RO 8 bits Bit Description 7 Multi-Function Device (MFD) -- RO. Indicates the Intel MEI host controller is part of a multifunction device. 6:0 Header Layout (HL) -- RO. Indicates that the Intel MEI uses a target device layout. Intel(R) Communications Chipset 89xx Series - Datasheet 860 Attribute: Size: Bit 15:8 18.3.7 RO 8 bits October 2012 Order Number: 327879-001US 18 18.3.8 MEI_MBAR--MEI MMIO Base Address Register Address Offset: 10h-17h Default Value: 0000000000000004h Attribute: Size: R/W, RO 64 bits This register allocates space for the Intel MEI memory mapped registers. Bit 63:4 Base Address (BA) -- R/W. Software programs this field with the base address of this region. 3 Prefetchable Memory (PM) -- RO. Indicates that this range is not pre-fetchable. 2:1 0 18.3.9 Description Type (TP) -- RO. Set to 10b to indicate that this range can be mapped anywhere in 64-bit address space. Resource Type Indicator (RTE) -- RO. Indicates a request for register memory space. SVID--Subsystem Vendor ID Register Address Offset: 2Ch-2Dh Default Value: 0000h 18.3.10 R/WO 16 bits Bit Description 15:0 Subsystem Vendor ID (SSVID) -- R/WO. Indicates the sub-system vendor identifier. This field should be programmed by BIOS during boot-up. Once written, this register becomes Read Only. This field can only be cleared by PLTRST#. SID--Subsystem ID Register Address Offset: 2Eh-2Fh Default Value: 0000h 18.3.11 Attribute: Size: Attribute: Size: R/WO 16 bits Bit Description 15:0 Subsystem ID (SSID) -- R/WO. Indicates the sub-system identifier. This field should be programmed by BIOS during boot-up. Once written, this register becomes Read Only. This field can only be cleared by PLTRST#. CAPP--Capabilities List Pointer Register Address Offset: 34h Default Value: 50h Attribute: Size: RO 8 bits Bit Description 7:0 Capabilities Pointer (PTR) -- RO. Indicates that the pointer for the first entry in the capabilities list is at 50h in configuration space. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 861 18.3.12 INTR--Interrupt Information Register Address Offset: 3Ch-3Dh Default Value: 0100h Bit 15:8 7:0 18.3.13 Interrupt Pin (IPIN) -- RO. This field indicates the interrupt pin the Intel MEI host controller uses. The value of 01h selects INTA# interrupt pin. Interrupt Line (ILINE) -- R/W. Software written value to indicate which interrupt line (vector) the interrupt is connected to. No hardware action is taken on this register. MLMG--Maximum Latency/Minimum Grant Register Bit 15:0 RO 16 bits Maximum Latency/Minimum Grant (MLMG)-- RO. Not used. Hardwired to 0000h. HFS--Host Firmware Status Register Attribute: Size: RO 32 bits Bit Description 31:0 Host Firmware Status (HFS) -- RO. This register field is used by Firmware to reflect the operating environment to the host. GMES--General ME Status Address Offset: 48h-4Bh Default Value: 00000000h Bit 31:0 Attribute: Size: RO 32 bits Description General ME Status(ME_GS)-- RO. This field is populated by ME. Intel(R) Communications Chipset 89xx Series - Datasheet 862 Attribute: Size: Description Address Offset: 40h-43h Default Value: 00000000h 18.3.15 R/W, RO 16 bits Description Address Offset: 3Eh-3Fh Default Value: 0000h 18.3.14 Attribute: Size: October 2012 Order Number: 327879-001US 18 18.3.16 H_GS--Host General Status Address Offset: 4Ch-4Fh Default Value: 00000000h Bit 31:0 18.3.17 RO 32 bits Description Host General Status(H_GS)-- RO. General Status of Host, this field is not used by Hardware PID--PCI Power Management Capability ID Register Address Offset: 50h-51h Default Value: 6001h Attribute: Size: RO 16 bits Bit Description 15:8 Next Capability (NEXT) -- RO. Value of 60h indicates the location of the next pointer. 7:0 18.3.18 Attribute: Size: Capability ID (CID) -- RO. Indicates the linked list item is a PCI Power Management Register. PC--PCI Power Management Capabilities Register Address Offset: 52h-53h Default Value: C803h Attribute: Size: RO 16 bits Bit Description 15:11 PME_Support (PSUP) -- RO. This five-bit field indicates the power states in which the function may assert PME#. Intel MEI can assert PME# from any D-state except D1 or D2 which are not supported by Intel MEI. 10:9 8:6 Reserved Aux_Current (AC) -- RO. Reports the maximum Suspend well current required when in the D3cold state. Value of 00b is reported. 5 Device Specific Initialization (DSI) -- RO. Indicates whether device-specific initialization is required. 4 Reserved 3 PME Clock (PMEC) -- RO. Indicates that PCI clock is not required to generate PME#. 2:0 Version (VS) -- RO. Hardwired to 011b to indicate support for Revision 1.2 of the PCI Power Management Specification. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 863 18.3.19 PMCS--PCI Power Management Control and Status Register Address Offset: 54h-55h Default Value: 0008h Bit 15 14:9 8 Attribute: Size: R/WC, R/W, RO 16 bits Description PME Status (PMES) -- R/WC. Bit is set by ME Firmware. Host software clears bit by writing 1 to bit. Reserved PME Enable (PMEE) -- R/W. This bit is read/write and is under the control of host SW. It does not directly have an effect on PME events. However, this bit is shadowed so ME FW can monitor it. ME FW will not cause the PMES bit to transition to 1 while the PMEE bit is 0, indicating that host SW had disabled PME. This bit is reset when PLTRST# asserted. 7:4 3 2 Reserved No_Soft_Reset (NSR) -- RO. This bit indicates that when the Intel MEI host controller is transitioning from D3hot to D0 due to a power state command, it does not perform an internal reset. Configuration context is preserved. Reserved Power State (PS) -- R/W. This field is used both to determine the current power state of the Intel MEI host controller and to set a new power state. The values are: 00 = D0 state (default) 1:0 11 = D3hot state The D1 and D2 states are not supported for the Intel MEI host controller. When in the D3hot state, the Intel ME's configuration space is available, but the register memory spaces are not. Additionally, interrupts are blocked. 18.3.20 MID--Message Signaled Interrupt Identifiers Register Address Offset: 8Ch-8Dh Default Value: 0005h RO 16 bits Bit Description 15:8 Next Pointer (NEXT) -- RO. Value of 00h indicates that this is the last item in the list. 7:0 Capability ID (CID) -- RO. Capabilities ID indicates MSI. Intel(R) Communications Chipset 89xx Series - Datasheet 864 Attribute: Size: October 2012 Order Number: 327879-001US 18 18.3.21 MC--Message Signaled Interrupt Message Control Register Address Offset: 8Eh-8Fh Default Value: 0080h Bit 15:8 7 6:1 0 18.3.22 Reserved. 64 Bit Address Capable (C64) -- RO. Specifies that function is capable of generating 64-bit messages. Reserved MSI Enable (MSIE) -- R/W. If set, MSI is enabled and traditional interrupt pins are not used to generate interrupts. MA--Message Signaled Interrupt Message Address Register Bit 31:2 1:0 Attribute: Size: R/W, RO 32 bits Description Address (ADDR) -- R/W. Lower 32 bits of the system specified message address, always DW aligned. Reserved. MUA--Message Signaled Interrupt Upper Address Register Address Offset: 94h-97h Default Value: 00000000h Bit 31:0 18.3.24 R/W, RO 16 bits Description Address Offset: 90h-93h Default Value: 00000000h 18.3.23 Attribute: Size: Attribute: Size: R/W 32 bits Description Upper Address (UADDR) -- R/W. Upper 32 bits of the system specified message address, always DW aligned. MD--Message Signaled Interrupt Message Data Register Address Offset: 98h-99h Default Value: 0000h Bit 15:0 Attribute: Size: R/W 16 bits Description Data (DATA) -- R/W. This 16-bit field is programmed by system software if MSI is enabled. Its content is driven during the data phase of the MSI memory write transaction. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 865 18.3.25 HIDM--MEI Interrupt Delivery Mode Address Offset: A0h Default Value: 00h Bit 7:2 1 Attribute: Size: R/W 8 bits Description Reserved. Intel MEI Interrupt Delivery Mode (HIDM) -- R/W. These bits control what type of interrupt the Intel MEI will send when ARC writes to set the M_IG bit in AUX space. They are interpreted as follows: 00 = Generate Legacy or MSI interrupt 01 = Generate SCI 10 = Generate SMI 0 18.3.26 Synchronous SMI Occurrence (SSMIO) -- R/WC. This bit is used by firmware to indicate that a synchronous SMI source has been triggered. Host BIOS SMM handler can use this bit as status indication and clear it once processing is completed. A write of 1 from host SW clears this status bit. NOTE: It is possible that an async SMI has occurred prior to sync SMI occurrence and when the BIOS enters the SMM handler, it is possible that both bit 0 and bit 1 of this register could be set. HERES--MEI Extend Register Status Address Offset: BCh-BFh Default Value: 00h Attribute: Size: RO 32 bits Bit Description 31 Extend Register Valid (ERV). Set by firmware after all firmware has been loaded. If ERA field is SHA-1, the result of the extend operation is in HER:5-1. If ERA field is SHA256, the result of the extend operation is in HER:8-1. 30 Extend Feature Present (EFP). This bit is hardwired to 1 to allow driver software to easily detect the chipset supports the Extend Register FW measurement feature. 29:4 Reserved Extend Register Algorithm (ERA). This field indicates the hash algorithm used in the FW measurement extend operations. Encodings are: 3:0 0h = SHA-1 2h = SHA-256 Other values = Reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 866 October 2012 Order Number: 327879-001US 18 18.3.27 HERX--MEI Extend Register DWX Address Offset: HER1: C0h-C3h HER2: C4h-C7h HER3: C8h-CBh HER4: CCh-CFh HER5: D0h-D3h HER6: D4h-D7h HER7: D8h-DBh HER8: DCh-DFh Default Value: 00000000h Bit Attribute: RO Size: 32 bits Description Extend Register DWX (ERDWX): 31:0 Xth DWORD result of the extend operation. NOTE: Extend Operation is HER[5:1] if using SHA-1. If using SHA-2, then Extend Operation is HER[8:1] 18.4 MEI2_MBAR--MEI2 MMIO Registers These MMIO registers are accessible starting at the MEI2 MMIO Base Address (MEI2_MBAR) which gets programmed into B0:D22:F1:Offset 10-17h. These registers are reset by PLTRST# unless otherwise noted. Table 18-4. MEI2 MMIO Register Address Map Offset (BAR= MEI2_MBAR) Mnemonic & Register Name Default Type 00-03h "H_CB_WW--Host Circular Buffer Write Window" 00000000h RO 04h-07h "H_CSR--Host Control Status" 02000000h RO 08h-0Bh "ME_CB_RW--ME Circular Buffer Read Window" 00000000h RO 0Ch-0Fh "ME CSR_HA--ME Control Status Host Access" 02000000h RO 18.4.1 H_CB_WW--Host Circular Buffer Write Window Address Offset: MEI2_MBAR + 00h Default Value: 00000000h Attribute: Size: RO 32 bits Bit Description 31:0 Host Circular Buffer Write Window Field (H_CB_WWF). This bit field is for host to write into its circular buffer. The host's circular buffer is located at the ME subsystem address specified in the Host CB Base Address register. This field is write only, reads will return arbitrary data. Writes to this register will increment the H_CBWP as long as ME_RDY is 1. When ME_RDY is 0, writes to this register have no effect and are not delivered to the H_CB, nor is H_CBWP incremented. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 867 18.4.2 H_CSR--Host Control Status Address Offset: MEI2_MBAR + 04h Default Value: 02000000h Bit Attribute: Size: RO 32 bits Description Host Circular Buffer Depth (H_CBD) -- RO. This field indicates the maximum number of 32 bit entries available in the host circular buffer (H_CB). Host software uses this field along with the H_CBRP and H_CBWP fields to calculate the number of valid entries in the H_CB to read or # of entries available for write. 31:24 NOTE: This field is implemented with a "1-hot" scheme. Only one bit will be set to a 1 at a time. Each bit position represents the value n of a buffer depth of (2^n). For example, when bit# 1 is 1, the buffer depth is 2; when bit#2 is 1, the buffer depth is 4, etc. The allowed buffer depth values are 2, 4, 8, 16, 32, 64 and 128. 23:16 Host CB Write Pointer (H_CBWP). Points to next location in the H_CB for host to write the data. Software uses this field along with H_CBRP and H_CBD fields to calculate the number of valid entries in the H_CB to read or number of entries available for write. 15:8 Host CB Read Pointer (H_CBRP). Points to next location in the H_CB where a valid data is available for embedded controller to read. Software uses this field along with H_CBWR and H_CBD fields to calculate the number of valid entries in the host CB to read or number of entries available for write. 7:5 18.4.3 Reserved Must be programmed to zero 4 Host Reset (H_RST). Setting this bit to 1 will initiate a Intel MEI reset sequence to get the circular buffers into a known good state for host and ME communication. When this bit transitions from 0 to 1, hardware will clear the H_RDY and ME_RDY bits. 3 Host Ready (H_RDY). This bit indicates that the host is ready to process messages. 2 Host Interrupt Generate (H_IG). Once message(s) are written into its CB, the host sets this bit to one for the HW to set the ME_IS bit in the ME_CSR and to generate an interrupt message to ME. HW will send the interrupt message to ME only if the ME_IE is enabled. HW then clears this bit to 0. 1 Host Interrupt Status (H_IS). Hardware sets this bit to 1 when ME_IG bit is set to 1. Host clears this bit to 0 by writing a 1 to this bit position. H_IE has no effect on this bit. 0 Host Interrupt Enable (H_IE). Host sets this bit to 1 to enable the host interrupt (INTR# or MSI) to be asserted when H_IS is set to 1. ME_CB_RW--ME Circular Buffer Read Window Address Offset: MEI2_MBAR + 08h Default Value: FFFFFFFFh RO 32 bits Bit Description 31:0 ME Circular Buffer Read Window Field (ME_CB_RWF). This bit field is for host to read from the ME Circular Buffer. The ME's circular buffer is located at the ME subsystem address specified in the ME CB Base Address register. This field is read only, writes have no effect. Reads to this register will increment the ME_CBRP as long as ME_RDY is 1. When ME_RDY is 0, reads to this register have no effect, all 1s are returned, and ME_CBRP is not incremented. Intel(R) Communications Chipset 89xx Series - Datasheet 868 Attribute: Size: October 2012 Order Number: 327879-001US 18 18.4.4 ME CSR_HA--ME Control Status Host Access Address Offset: MEI2_MBAR + 0Ch Default Value: 02000000h Bit 31:24 23:16 15:8 7:5 4 3 2 1 0 Attribute: Size: RO 32 bits Description ME Circular Buffer Depth Host Read Access (ME_CBD_HRA). Host read only access to ME_CBD. ME CB Write Pointer Host Read Access (ME_CBWP_HRA). Host read only access to ME_CBWP. ME CB Read Pointer Host Read Access (ME_CBRP_HRA). Host read only access to ME_CBRP. Reserved ME Reset Host Read Access (ME_RST_HRA). Host read access to ME_RST. ME Ready Host Read Access (ME_RDY_HRA). Host read access to ME_RDY. ME Interrupt Generate Host Read Access (ME_IG_HRA). Host read only access to ME_IG. ME Interrupt Status Host Read Access (ME_IS_HRA). Host read only access to ME_IS. ME Interrupt Enable Host Read Access (ME_IE_HRA). Host read only access to ME_IE. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 869 PCIe EndPoint & GbE - Volume 2 of 4 October 2012 Order Number: 327879-001US 19.0 19.0 PCI Express* EndPoint Introduction 19.1 PCI Express* EndPoint (EP) 19.1.1 Overview The PCIe* EP provides an interface for the system to connect to a PCIe* Root Complex. It implements the EP functionality, including the Type 0 configuration and the PCIe* extended configuration that enable a host CPU to enumerate, configure and use Intel(R) QuickAssist Technology. 19.1.2 Feature List * Type 0 configuration and extended PCIe* configuration registers -- Materializes as a Multi-function PCIe* EndPoint -- Implements Type 0 PCIe* configuration space registers -- Interfaces to Gen 2 PCIe* and is compliant to PCIe* 2.0 Base Specification -- Generates INT[A,B,C,D] messages Collates interrupts from different sources in the EP itself and other agents and generates INTx messages, MSI, or MSI-X. * Single Root I/O Virtualization (SR-IOV for Intel(R) QuickAssist Technology only) Extended Capability Record -- 1 Physical Function (INTx, MSI, MSI-X) -- 16 Virtual Functions (MSI only) October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 871 19.1.3 EP Interface Block Diagram Figure 19-1 shows the block diagram of the EP Interface. Figure 19-1. PCIe* EP Interface Block Diagram PCIe* EndPoint (EP) PCIe G2 EP SMBus ( Slave ) LVIO (x4 , x8 , or x16) Gen2 GbE MAC SFP / I2 C / MDIO SGMII Port 0 Function 1 GbE MAC SFP / I2 C / MDIO SGMII Port 1 Function 2 GbE MAC SFP / I2 C / MDIO SGMII Port 2 Function 3 Intel(R) QuickAssist Technology GbE MAC SFP / I2 C / MDIO Port 3 SGMII Function 4 GbE SMBus ( Master / Slave ) LVIO Function 0 DEVICE 0 Intel(R) Communications Chipset 89xx Series - Datasheet 872 October 2012 Order Number: 327879-001US 19.0 19.1.4 EP Functional Description Figure 19-2 shows the functional description block diagram of the EP Interface. Figure 19-2. EP Functional Description Block Diagram PCIe* Pins Endpoint PCIe* Command Receive, Ordering & Decode PCIe* Command Generation, Ordering, Flow Control PCIe* Request/Data FIFOs PCIe* Config Spaces PCIe* Request/Data FIFOs Address decode & remap Interrupt Handling CSRs Data Transfer Interface Cmd FIFOs Interface Cmd FIFOs EP Agents (QAT, GbE[3:0)]) Transmit Interface (TI) Receive Interface (RI) Interrupts 19.1.4.1 Inbound (EP to IA) Inbound Read Data Return (IA to EP) Outbound (IA to EP) Outbound Read Data Return (EP to IA) Transmit Interface (TI) TI is responsible for converting EP Agents' commands towards system memory and OB completions into the PCIe* transactions. TI connects to the inbound port of the PCIe root complex. 19.1.4.2 Receive Interface (RI) The Receive Interface (RI) is responsible for converting OB PCIe commands from the IA cores into EP Agents' commands. OB read completions for PCIe commands initiated by the TI are also received by the RI. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 873 19.1.4.3 PCIe* EP Functions RI connects to the OB port of the PCIe root complex. Table 19-1 lists the RI supported PCIe transactions. The EP only supports requests that are aligned on DWORD address boundaries. Table 19-1. PCIe* Commands Supported by RI PCIe* Command Description Source EP Attributes MRd Memory Read IA reading EP MMIO region, PCIe* extended Configuration space 4B MWr Memory Write IA writing to EP MMIO region, PCIe* extended configuration space 4-8B IORd IO Read IA reading GbE IO space 4B IOWr IO Write IA writing GbE IO space 4B CfgRd0 Type 0 Configuration Read IA reading EP Configuration Register 1-4B CfgWr0 Type 0 Configuration Write Non-Posted IA writing to EP Configuration Register. 1-4B CmplD Completion with Data Read Completion with Data for a Read Command initiated by TI 4B-256B Msg Messages without Data Root Complex 1. 2. 19.1.4.4 Comments Note 1 Note 2 Power Management Messages IO transactions are supported for GbEs to access the IOADDR/IODATA registers. This is a non-posted transaction. RI will generate a completion without data after the configuration write is completed. The completion will be routed via a RI-TI private bus to the TI. The TI will initiate the PCIe* completion packet. EP Function Mapping The EP implements a multi-function PCIe* device. Table 19-2 lists the mapping of the EP functionality. Table 19-2. PCIe* EP Function Mapping EP Cluster Functional Blocks PCIe* Function Number BARs Interrupt Capability PCIe* Extended/ PCI Capabilities Intel(R) QuickAssist Technology QAT 0 PeQATBAR, PMISCBAR, PETRINGCSRBARa MSI, MSI-X with 17 vectors, INTA MSI, MSI-X, PM, PCIe*, AER, ARI GbE[0] 1 GbE[1] 2 GbE GbE[2] 3 GbE[3] 4 MSI, MSI-X with 10 vectors, INTB GBEPCIBAR0[0:3] GBEPCIBAR1[0:3] GBEPCIBAR2[0:3]b MSI, MSI-X with 10 vectors, INTC MSI, MSI-X with 10 vectors, INTD MSI, MSI-X, PM, PCIe*, AER MSI, MSI-X with 10 vectors, INTB a. PMISCBAR is used for accessing EP, QAT, MSI-X Table, etc. b. Based on GbE. BARs points to CSRs and MSI-X Tables. Intel(R) Communications Chipset 89xx Series - Datasheet 874 October 2012 Order Number: 327879-001US 19.0 19.1.4.5 EP Mapping of BARs to MMIO Table 19-3 shows the mapping of the EP MMIO regions to the BARs. See each PCIe function BAR definition for more details. Table 19-3. EP Mapping of BARs to MMIOs Function # BAR Name (Size) Region Size 0 PeQATBAR, PMISCBAR, PETRINGCSRBAR (128KB) 120KB 0 0 1, 2, 3, 4 (4 GbEs) 1. 19.1.5 4KB Functionality EP, QAT Comments This region is mapped to the EP and QAT CSRs EP MMIO CSRs The EP CSRs are located in the EP. 4KB MSI-X Tables The MSI-X Tables are located in the EP. GBEPCIBAR0[0:3] 128KB GbE internal CSRs See Chapter 29.0, "Function 1-4 (GbE)". This BAR is used to access the GbE Internal CSRs. GBEPCIBAR1[0:3] 32B GbE internal CSRs. See Chapter 29.0, "Function 1-4 (GbE)". This IO BAR is used to access the GbE internal CSRs. GBEPCIBAR2[0:3] 16KB MSI-X Tables The MSI-X Tables are located in the EP. See Chapter 29.0, "Function 1-4 (GbE)". Accesses to MMIO region that are not implemented within the regions defined in this table will return 0's. Accesses to MMIO regions that is not claimed by any of the BARs will return all 1s. PCIe* EP Interrupts The RI is responsible for collating interrupts and errors from the EndPoint and taking the appropriate actions based on the PCIe configurations status. Interrupts from EP can be configured in the RI to be routed to the IA cores. The interrupts can be sent upstream as either an MSI, MSI-X, or by using INTx virtual wire interrupt signaling mechanism (which are messages on PCIe). 19.1.5.1 INTx PCIe supports devices that need to use the PCI compatible legacy interrupt signaling: INT[A,B,C,D]. Instead of dedicated pins for this functionality, PCIe supports inband virtual wires. The EP supports generation of INTx virtual wire interrupt signalling based on internal generated interrupts. The INTx notification is accomplished using PCIe messages. The RI is responsible for collating interrupt sources from the EP Agents and generating the transaction that is translated to a PCIe interrupt message to the IA. Upon detecting an interrupt from an internal source, the RI first determines whether to generate an MSI, MSI-X, or INTx message. 19.1.5.2 MSI-X MSI-X defines a separate optional extension to the basic MSI functionality. Compared to MSI, MSI-X supports a larger number of vectors per function, the ability for the OS to program independent address and data values for each vector. The address and data values for each vector is specified by a table that resides in EP MMIO space. Once an MSI has been generated using MSI-X, its characteristics are similar to an MSI that was generated using an MSI capability record. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 875 MSI-X however, supports per-vector masking that is optional for MSI. MSI-X also supports a Function Mask bit, which when set masks all of the vectors associated with a function. Per-vector masking is managed through a Mask and Pending bit pair per MSI vector or MSI-X Table entry. An MSI vector is masked when its associated Mask bit is set. An MSI-X vector is masked when its associated MSI-X Table entry Mask bit or the MSI-X Function Mask bit is set. While a vector is masked, the function is prohibited from sending the associated message, and the function must set the associated Pending bit whenever the function would otherwise send the message. When software unmasks a vector whose associated Pending bit is set, the function must schedule sending the associated message, and clear the Pending bit as soon as the message has been sent. When MSI-X is enabled, EP generates an interrupt using a PCIe posted memory write transaction. The address and data of that transaction are determined by the system and programmed in PCI MSI-X table entry. 19.1.6 PCIe* EP Errors There are two classes of errors that the EP will encounter: * Errors that occur due to PCIe transactions -- PCIe error management defines the scope of these kinds of errors including how they are reported and handled. * Other errors not directly related to PCIe transactions -- These class of errors are reported via interrupts and do not fall under the scope of PCIe error management mechanisms 19.1.6.1 PCIe* Error Management PCIe error management focuses on errors associated with the PCIe interface and the transactions between the transaction layers of the transmitting and the receiving PCIe ports. 19.1.6.2 PCIe* Error Reporting Mechanisms The PCIe specification provides three mechanisms for reporting errors that occur while servicing to PCIe transactions. 1. PCI a. This mechanism provides backward compatibility with legacy PCI compatible software and is required per the PCIe specification. Legacy compatible PCI software will use this mechanism and control this feature via the PCI configuration Command Register. When this mechanism is enabled, EP will also log error status information in the PCI Configuration Status register. 2. PCIe a. This is the baseline PCIe mechanism for software that understands PCIe devices. The mechanism can be enabled via the PCIe Device Control Register. Error status is logged in the PCIe Device status register. b. Supported error classifications: Unsupported request type, fatal error, non-fatal error, and correctable error. 3. PCIe AER a. The optional advanced error reporting registers can be implemented by PCIe devices. This mechanism is defined using a PCIe extended capability structure. b. All EP Functions will support AER. Intel(R) Communications Chipset 89xx Series - Datasheet 876 October 2012 Order Number: 327879-001US 19.0 19.1.6.3 PCIe* Error Handling and Signalling The EP will classify errors in the following three PCIe specification-defined buckets: 1. Correctable errors: These errors are handled by the HW. 2. Uncorrectable errors (non-fatal): Handled by device-specific software 3. Uncorrectable errors (fatal): Handled by system software EP will use three different mechanisms to signal errors that occur when processing PCIe commands. 1. Completion status 2. Error forwarding or data poisoning 3. Error messages 19.1.6.4 PCIe* Error Sources Table 19-4 defines the standard PCIe defined error sources. The following sections describes each of the PCIe error types, and describes how the EP handles the PCIe errors. Table 19-4. PCIe* Error Sources Type of Error Reported Using Comment ECRC Check Not supported in EP Malformed TLP Uncorrectable Fatal Message RI will perform checks on PCIe transactions received from IA. Completion Timeout Uncorrectable Non-Fatal Message Unsupported Requests Completer Abort (optional) Unexpected Completion Completion Status RI will generate these Split Transaction Errors. Completer Abort is supported in RI. Uncorrectable Non-Fatal Message TL TI will set the EP bit in the header based on data error. Data Corruption/Poisoning Receiver Overflow (Optional) Flow Control Protocol Errors (Optional) PCIe* Header Writes that terminate in RI get dropped and status logged. Writes that target a EP function will be sent with data error and logged in RI. Will be Implemented in RI. Optional FC Errors LCRC Check Failure for TLP/DLLP Sequence Number Check Failure DL Replay Time-out Replay Number Rollover DLL Protocol Errors PL October 2012 Order Number: 327879-001US Receiver/Training Errors (Optional) Intel(R) Communications Chipset 89xx Series - Datasheet 877 19.1.6.4.1 Role Based Error Reporting In earlier versions of the PCI Express Specification, errors were reported by the agent that detected the error. The PCI Express Base Specification, Rev. 1.1 implements a role based error reporting where the response to the errors is based on the components role in the transaction. In general, errors detected in Non Posted transactions are handled by the initial requestor and the completer may optionally send an advisory message to the root complex as an ERR_COR message. Errors in posted transactions are still logged and reported by the target device. Note: If the severity for the error is programmed to fatal in the PCI Express* Uncorrectable Error Severity register, then it is not an Advisory Non-Fatal Error and is signalled with an ERR_FATAL message. A fatal severity overrides all other Advisory Error control bits. The following errors are considered Advisory Non-Fatal Error cases and have different handling depending based on the transaction type. * ECRC Check Failed * Unexpected Completion * Unsupported Request (UR) * Poisoned TLP Received * Completer Abort (CA) * Completion Timeout Table 19-5. Advisory Error Cases Error Type Posted Non Posted Completion ECRC Check Failed EP does not support ECRC checking. Unsupported Request Not Advisory Error Send ERR_NONFATAL Advisory Error Send ERR_COR Signaled via device driver Completer Abort Not Advisory Error Send ERR_NONFATAL Advisory Error Send ERR_COR Signaled via device driver Unexpected Completion N/A N/A Advisory Error Send ERR_COR Poisoned TLP Received Not Advisory Error -Send ERR_NONFATAL N/A Not Advisory Error Send ERR_NONFATAL Completion Timeout N/A N/A Advisory Error - send ERR_COR. The different responses are described in detail in the following sections. 19.1.6.4.2 Malformed Packets The following checks are made to detect malformed TLPs. * Data Payload exceeds the length specified by the value in the Max_Payload_Size field of the Device Control register. * The value in the length field and the actual amount of data received do not match. The value in the length field applies only to data, TLP digest is not included in the length. * A TLP with a 1b in TD field but without a TLP digest or a TLP with a TLP digest but without a 1b in TD field * Address/Length combination which crosses a 4K boundary. * Receipt of Assert_INTx/Deassert_INTx messages. * Packets having undefined Type Field Intel(R) Communications Chipset 89xx Series - Datasheet 878 October 2012 Order Number: 327879-001US 19.0 * IO and Configuration requests are considered malformed when -- TC[2:0] /= 000b -- Attr[1:0] /= 00b -- Length[9:0] /= 0000000001b -- Last DW BE[3:0] /= 0000b When a malformed packet is detected, the packet is dropped and the error is logged. No flow control information is updated for malformed packets. 19.1.6.4.3 ECRC Check Failed EP does not support ECRC checking. 19.1.6.4.4 Unsupported Request Unsupported Requests are detected by the address decode and translation logic. A TLP is treated as unsupported in the following cases: * the TLP fails to match any of the active memory or I/O windows. * a configuration TLP that targets an invalid function number * receipt of a Vendor_Defined Type 0 message * a message request with an undefined or unsupported message code * a poisoned I/O or Configuration request * receipt of a Memory or I/O transaction while in a non-D0 power state * receipt of a Memory Read Lock (MRdLk) No checks are made to for address + length crossing a window boundary. For posted transactions, this is not an Advisory Error and an ERR_NONFATAL is sent to the root complex. For non-posted transactions, this is considered and Advisory Error. An ERR_COR is sent to the root complex and a completion with UR status is returned to the requestor. Note: If the severity setting in the PCI Express Uncorrectable Error Severity register is fatal this is not an Advisory Error and an ERR_FATAL will be sent to the root complex. 19.1.6.4.5 Completer Abort Requests that target abort or master abort on the Internal Bus are treated as a Completer Abort. These requests must have passed the Malformed TLP checks as well as the Unsupported Request checks before they are issued on the internal bus. For posted transactions, an ERR_NONFATAL is sent to the root complex. For non-posted transactions, an ERR_COR is sent to the root complex and a completion with CA status is returned to the requestor. Note: If the severity setting in the PCI Express Uncorrectable Error Severity register is fatal this is not an Advisory Error and an ERR_FATAL will be sent to the root complex. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 879 19.1.6.4.6 Unexpected Completions Unexpected completions occur when a completion transaction ID does not match a outstanding request. If the Requester ID of the completion matches a valid function, the error will get logged in that function. Otherwise the error will get logged against all functions. This is an Advisory Non-Fatal Error and an ERR_COR will be sent to the root complex. Note: If the severity setting in the PCI Express Uncorrectable Error Severity register is fatal this is not an Advisory Error and an ERR_FATAL will be sent to the root complex. 19.1.6.4.7 Poisoned TLP Received Poisoned TLPs can be received for both Inbound Posted (Write/Message) and Inbound Completions. The two TLP types can be handled differently. Poisoned completions are passed through to the target agent with the error bit set. The error is logged in the corresponding function. This is an Advisory Error and an ERR_COR will be issued. Poisoned Memory Writes are passed through to the target agent with the error bit set. The error is logged in the corresponding function. This is a non-fatal error and an ERR_NONFATAL will be issued Note: If the severity setting in the PCI Express Uncorrectable Error Severity register is fatal this is not an non-fatal Error and an ERR_FATAL will be sent to the root complex. Autorecovery is discouraged as the ERR_FATAL message will likely bring down the hierarchy. 19.1.6.4.8 Completion Timeout When an out-bound non-posted request results in a completion timeout, the Advanced Error registers are updated in the corresponding function. This will be treated as an Advisory Error and an ERR_COR will be sent to the root complex. Note: If the severity setting in the PCI Express Uncorrectable Error Severity register is fatal this is not an Advisory Error and an ERR_FATAL will be sent to the root complex. Autorecovery is discouraged as the ERR_FATAL message will likely bring down the hierarchy. 19.1.6.4.9 Non Function Specific Errors The PCI Express specification lists the following errors as non function specific: * All Physical Layer errors * All Data Link Layer errors * These transaction Layer errors -- ECRC Fail -- UR, when caused by no function claiming a TLP -- Receiver Overflow -- Flow Control Protocol Error -- Malformed TLP -- Unexpected Completion, when caused by no Function claiming a Completion On the detection of one of these errors, a multi-function device should generate at most one error reporting message of a given severity, where the message must report the Requester ID of a function of the device that is enabled to report that specific type Intel(R) Communications Chipset 89xx Series - Datasheet 880 October 2012 Order Number: 327879-001US 19.0 of error. If no function is enabled to send a reporting message, the device does not send a reporting message. If all reporting-enabled functions have the same severity level set for the error, only one error message is sent. If all reporting-enabled functions do not have the same severity level set for the error, one error message for each severity level is sent. Software is responsible for scanning all functions in a multifunction device when it detects one of those errors. 19.1.7 Device Specific Error Management EP will report device specific errors using interrupts - INTx or MSI based on the PCIe configuration settings. Table 19-6. Device Specific Errors Function Description Function 0 EP Implements Error source and mask registers. Upon detected an asserted on an unmasked error, RI will generate either INTA or MSI. Function [4:1] 19.1.7.1 Based on GbE. Memory Error Poisoning The TI also tracks data errors for all Transactions that flow through the TI, including accesses to PCIe and CSRs. 19.1.7.1.1 Memory Write Poisoning For a PCIe Memory Write (including Memory Writes triggered by the Ring Controller), the TI pulls data from the EP Master. If the internal bus indicates an error, then the TI logs the error, but it does not abort the Write Transaction. Instead, the TI passes along the Data Error notification along with the Data. On PCIe, the TI will set the EP bit in the header of the packet. 19.1.8 PCIe* Request Generation Table 19-7 shows the supported PCIe packet types generated and accepted by the EP. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 881 Table 19-7. Supported PCIe* Transactions Packet Name TLP Packet Description TI Generate? Comments MRd Memory Read Request MWr Memory Write Request Configuration Type 0 (Rd, Wr) Type 0 Configuration Read & Write Yes No EP will not generate but accept Type 0 Configuration Cycles Msg Message Request without Data No Yes INTx, Error etc. MsgD Message Request With Data Cpl Completion without Data Yes For configuration requests CplD Completion with Data MRdLk Memory Read Request Locked access CplLk Completion without Data for Locked access CplDLk Completion with Data for Locked access IO (Rd, Wr) IO Read & Write Configuration Type 1 (Rd, Wr) Type 1 Configuration Read & Write Ack TLP Acknowledge Nak TLP No Acknowledge No No Yes Read return Not required for a PCIe Device No Only for bridges/root ports Link Layer functionality that is required Power Management Flow Control Yes Required for communicating flow control information between the root port & EP Yes Terminated in Link/Phy Layer. Some functionality will require support in EP. PHY Layer functionality PM_Enter_L23 PM_Active_State_Req uest_L1 Yes Yes Terminated in Link/Phy Layer. Some functionality will require support in EP. PM_Enter_L1 DLLP RI Accept? EP supports L0, L1 PM_Request_Acq InitFC-(P, NP, Cpl) InitFC2)P, NP, CPl) UpdateFC(P, NP, Cpl) TS1 Training Sequence 1 TS2 Training Sequence 2 SKIP Required for compensation of clock frequency variation between transmitter and receiver PLP FTS Fast Training Sequence Required for transition from L0s to L0 IDLE Electrical Idle Ordered Sets The TI generates a PCIe request header for memory based on the EP command and read completions from RI, and messages based on the EP internal interrupts and errors. Intel(R) Communications Chipset 89xx Series - Datasheet 882 October 2012 Order Number: 327879-001US 19.0 19.1.8.1 64-bit Memory Accesses Table 19-8 and Table 19-9 shows the 3 and 4 DW PCIe memory request header format. TI will use the 3 DW header format when upper address bits (63:32) are zero, else it will use the 4 DW header format. Table 19-8. PCIe* Request Header Format for 32-bit Memory Addressing +0 7 6 5 R FMT 4 3 +1 2 1 0 Type 7 6 R 5 4 3 +2 2 TC 1 0 R 7 6 TD EP 5 4 +3 3 Attr Requester ID 2 1 0 7 6 R 5 4 3 2 1 0 Length Last DW BE Tag First DW BE Address[31:2]] R Table 19-9. PCIe* Request Header Format for 64-bit Memory Addressing +0 7 6 5 R FMT 4 3 +1 2 1 0 Type 7 6 R 5 4 3 +2 2 TC 1 0 R 7 6 TD EP 5 4 +3 3 Attr Requester ID 2 1 0 7 6 R 5 4 3 2 1 0 Length Last DW BE Tag First DW BE Address[63:32] Address[31:2]] R TI will generate the header based on the EP command it receives and Table 19-9 shows the mapping of the EP command fields to the PCIe command fields. Some fields are static for all PCIe transactions initiated by the TI while others depend upon the EP command. 19.1.8.2 Completions EP will issue completions for the IA initiated PCIe commands: * IA initiated memory read and configuration read commands will result in a completion with data (CplD) * IA initiated configuration write command will result in a completion without data (Cpl) Table 19-10. PCIe* Completion Header Format +0 7 6 5 R FMT 4 3 +1 2 1 0 Type 7 6 R Completer ID Requester ID October 2012 Order Number: 327879-001US 5 TC 4 3 +2 2 1 R 0 7 6 TD EP 5 4 Attr Completion Status B C M Tag +3 3 2 1 0 7 R 6 5 4 3 2 1 0 Length Byte Count R Lower Address Intel(R) Communications Chipset 89xx Series - Datasheet 883 Table 19-11. PCI Completion Request Generation (Sheet 1 of 2) PCIe* Header Field PCIe* Description PCIe* Field Encoding Length [9:0] Indicates the Length of the data payload in DW 0x0h = 1024DW 0x1h = 1DW . . . 1FFh = 1023DW Attributes [0] No Snoop 1 = No Snoop 0 = Snoop Attributes [1] Relaxed Ordering (RO) 1=RO enabled 0=Strict Ordering EP Poisoned Data 1 = Data is invalid Based on Pushdata Error Digest 1 = Digest field is included EP does not support ECRC so this field is set to 0 for all packets generated by the EP. TC[2:0] Traffic Class 000b = TC0 . . 111b = TC7 Same value as in Request Header Fmt[1:0] Transaction Format 00b = Cpl 10=CplD Generated based on Type of Request that causes the completion header. MRd & CfgRd = CplD CfgWr = Cpl Type[4:0] Type Byte Count Remaining Byte Count until the read request is satisfied Always set to 0. EP read returns will be always < 8DW's and the AC TI will return the data in one completion. Byte Count Modifier Set by PCI-X completers only Always set to 0 Completion Status Status of the completion packet from TD Completer ID TI will support up to 8DWs data payload for completions. Hard-code to 0 EP should never get a request with no snoop or RO bit set. Hardcoded to 01010b for completion packets 000b = Successful Cmpl 001b = Unsupported Cmpl 010b =Config Req Retry 100b = Completer Abort Identifies the Completer ARI Disabled: RID[15:8] = Bus RID[7:3] = Device RID[2:0] = Function ARI Enabled: RID[15:8] = Bus RID[7:0] = Function Intel(R) Communications Chipset 89xx Series - Datasheet 884 Generated Based on? All encodings are possible depending upon the status of the request completion. See Requester ID row at the end of this table for details on how Completer ID is calculated. October 2012 Order Number: 327879-001US 19.0 Table 19-11. PCI Completion Request Generation (Sheet 2 of 2) PCIe* Header Field PCIe* Description PCIe* Field Encoding Generated Based on? Lower Address Lower 7 bits of the first enabled byte of data returned with a read Valid only for Completions with data and refers to the first enabled byte of data being returned by the completer Calculated by RI for Configuration accesses based on the DW aligned address and BEs. All other MMIO accesses are either 4B or 8B aligned. There is no support for BEs for such MMIO accesses. Tag[7:0] Tags to identify requests ARI Disabled: RID[15:8] = Bus RID[7:3] = Device RID[2:0] = Function Bus, Device, Function Number Requester ID[15:0] 19.1.8.3 Same value as in Request Header RID copied from the original request. ARI Enabled: RID[15:8] = Bus RID[7:0] = Function Messages Table 19-12 shows the PCIe message format. The EP generates INTx, error messages, and PME messages. Table 19-12. PCIe* Request Header Format for Messages +0 7 6 5 R FMT 4 3 +1 2 1 0 Type 7 6 R 5 TC 4 +2 3 2 1 0 R Requester ID 7 6 TD EP 5 4 Attr +3 3 2 1 0 R Tag 7 6 5 4 3 2 1 0 Length Message Code Address field not used if Implicit Routing is used. Address field not used if Implicit Routing is used. Table 19-13. PCIe* Message Generation (Sheet 1 of 2) PCIe* Header Field PCIe* Description Field Encoding Comment Length [9:0] Length in DW 0x0h Reserved for messages Attributes [0] No Snoop 0 = Snoop Attributes [1] Relaxed Ordering (RO) 0=Strict Ordering EP Poisoned Data 0 = Data is not poisoned No data packet for messages TD Digest 1 = Digest field is included EP does not support ECRC so this field is set to 0 for all packets generated by the EP. October 2012 Order Number: 327879-001US Fixed for messages Intel(R) Communications Chipset 89xx Series - Datasheet 885 Table 19-13. PCIe* Message Generation (Sheet 2 of 2) PCIe* Header Field PCIe* Description Field Encoding TC[2:0 Traffic Class 000b = TC0 Fmt[1:0] Type[4:0] Transaction Format & Type 01b & 10100b (Msg w/o Data and Terminate at Receiver) Message Code [7:0] 19.2 Assert_INTX 0010 0000b Deassert_INTX 0010 0100b ERR_COR 0011 0000b ERR_NONEATAL 0011 0001b Comment Fixed for INTx & Error messages Based on internal EP interrupt event or error message type. and Error Messages ERR_FATAL 0011 0011b Tag[7:0] Tags to identify requests 00h Posted Message, no tag Requester ID[15:0] Bus, Device, Function Number See the Requester ID row of Table 19-11 for details on how Completer ID is calculated. Fixed for messages Intel(R) QuickAssist Technology With the Intel(R) QuickAssist Integrated Accelerator (QAT) as an essential part of the EP, this allows the IA to efficiently access the QAT services through the QAT APIs. 19.2.1 Features * Symmetric Cryptographic Functions -- Cipher Operations -- Hash/Authenticate Operation -- Cipher-Hash Combined Operation -- Key Derivation Operation -- Random Number Generation * Public Key Functions -- RSA Operation -- Diffie-Helman Operation -- Digital Signature Standard Operation -- Key Derivation Operation -- Elliptic Curve Cryptography: ECDSA* and ECDH* -- Random Number Generation and Prime Number Testing * Compression/Decompression -- Deflate (Lempel-Ziv 77-Stac) Intel(R) Communications Chipset 89xx Series - Datasheet 886 October 2012 Order Number: 327879-001US 19.0 19.2.2 EP Intel(R) QuickAssist Integrated Accelerator (IQIA) The IQIA provides acceleration functions that can be used by the IA cores. The acceleration services can be accessed by a standard PCIe driver. To enhance communication between the IA and the IQIA, the EP implements rings via the Ring Controller to route messages between them. Figure 19-3. EP Block Diagram PCIe* EndPoint (x4, x8, or x16) Gen2 Ring Controller Intel(R) QuickAssist Integrated Accelerator B:D:F - B:0:0 19.2.3 Ring Controller The Ring Controller provides for communication between the IA cores and the Intel(R) QuickAssist Integrated Accelerator. 19.2.3.1 Features * Implements 256 rings for communications between the CPU and the IQIA. * Organized in 16 bundles of 16 rings each. * Each Bundle supports 1 interrupt to the PCIe Gen2 EndPoint Function 0. 19.2.3.2 Functionality The Intel(R) QuickAssist Technology is accessed from the IA CPU via Intel(R) QuickAssist Technology APIs. Intel provides these APIs for the QA IA Services. Interface "A" in Figure 19-4 illustrates an application accessing the IQAT through these APIs. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 887 Interface "A" is not dependent on HW or FW, but Interface "B" is. The interface may change with subsequent releases of the driver, which is maintained by Intel. Figure 19-4. Ring Block Diagram CPU A p p lica tio n A In te l(R) Q u ickA ssist T e ch n o lo g y A P Is P C Ie * E n d P o in t (x4 , x8 , or x 1 6 ) G e n2 B R in g C o n tro lle r In te l (R) Q u ickA ssist In te g ra te d A cce le ra to r B :D :F - B :0 :0 Note: See the Intel(R) QuickAssist Technology Programmer's Guide for information on Intel(R) QuickAssist Technology APIs. Intel(R) Communications Chipset 89xx Series - Datasheet 888 October 2012 Order Number: 327879-001US 20.0 20.0 PCIe Endpoint Function 0 Registers 20.1 EP PF PCI Configuration Space The EP Physical Function implements Type 0 PCIe configuration space that is compliant with PCIe 2.0 Specification. Table 20-1. EP PCI Configuration Registers (Sheet 1 of 3) 3 2 Required PCI Compatible Configuration Space 1 0 Device ID Vendor ID 00h Status Command 04h Class Code BIST Header Type Master Latency Timer Revision ID 08h Cache Line Size 0Ch 10h Base Address Register 1 0 (PeQATBAR) 14h 18h Base Address Register 3 2 (PMISCBAR) 1Ch 20h Base Address Register 5 4 (PETRINGCSRBAR) 24h Cardbus CIS Pointer Subsystem ID 28h Subsystem Vendor ID Expansion ROM Base Address Reserved Min_Gnt Interrupt Pin 30h 34h InterruptLine 3Ch 38h SKU Register 40h RESERVED 44h - 4Bh SKU Register 2 4Ch Reserved 50-5Ch Message Control MSI-X Cap Next Cap Pointer MSI-X CAP_ID MSI-X Table Offset and BIR October 2012 Order Number: 327879-001US Power Management Capabilities Next Cap Pointer Power Management Control and Status 60h 64h MSI-X PBA Offset and BIR PM Cap 2Ch Capabilities Pointer Reserved Max_Lat Offset1 68h PM CAP_ID 6Ch 70h Intel(R) Communications Chipset 89xx Series - Datasheet 889 Table 20-1. EP PCI Configuration Registers (Sheet 2 of 3) 3 1 0 Offset1 Next Cap Pointer PCIe CAP_ID 74h 2 PCI Express Capabilities Register Device Capabilities 78h Device Status Device Control Link Capability 80h Link Status Link Control Slot Capabilities (Reserved) PCIE Cap 88h Slot Status (Reserved) Slot Control (Reserved) 8Ch Root Control (Reserved) 90h Root Status (Reserved) 94h Device Capabilities 2 98h Device Control 2 Link Capabilities 2 (Reserved) Link Status 2 Link Control 2 Slot Status 2 (Reserved) Next Cap Pointer MSI CAP_ID MSI Addr MSI Data BCh MSI Mask C0h C4h Reserved C8h - FCh AER Capability ID 100h Uncorrectable Error Status 104h Uncorrectable Error Mask 108h Uncorrectable Error Severity 10Ch Correctable Error Status 110h Correctable Error Mask 114h Control and Capability 118h Header Log 11Ch Header Log 120h Header Log 124h Header Log 128h Next Capability Pointer/Capability Version ARI Capability ID 138h ARI Control ARI Capability 13Ch Intel(R) Communications Chipset 89xx Series - Datasheet 890 B0h MSI pending Next Capability Pointer/Capability Version AER Configuration Space ACh B4h - BBh Reserved Reserved A4h A8h Slot Control 2 (Reserved) Message Control 9Ch A0h Slot Capabilities 2 (Reserved) ARI Configuration Space 84h Root Capabilities (Reserved) Device Status 2 (Reserved) MSI Cap 7Ch October 2012 Order Number: 327879-001US 20.0 Table 20-1. EP PCI Configuration Registers (Sheet 3 of 3) 3 2 1 Next Capability Pointer/Capability Version 0 Offset1 SRIOV Capability ID 140h SR-IOV Capabilities SR-IOV Status SR-IOV Control 148h Total VF's (RO) Initial VF's (RO) 14Ch Num VF's (RW) 150h VF Stride (RO) First VF Offset (RO) 154h VF Device ID(RO) Reserved Reserved SR-IOV Configuration Space 144h Function Dependency 158h Supported Page Sizes (RO) 15Ch System Page Size (RW) 160h 164h VF BAR0 (RW) 168h 16Ch VF BAR 1 (RW) 170h Reserved Reserved 174h Reserved 178h VF Migration Offset (N/A) 17Ch Reserved 180h - FFCh Notes: 1. Any addresses not shown are Read-only 0. 20.2 Detailed Register Summaries 20.2.1 PCI Views Table 20-2. Bus M, Device 0, Function 0: Summary of PCIe Intel(R) QuickAssist Configuration Registers (Sheet 1 of 3) Offset Start 00h Offset End 01h Register ID - Description "PVID--PF Vendor Identification Register" on page 896 Default Value 8086h 02h 03h "PDID--PF Device Identification Register" on page 896 0434h 04h 05h "PPCICMD--PF Device Command Register" on page 897 0000h 06h 07h "PPCISTS--PF PCI Device Status Register" on page 898 0010h 08h 08h "PRID--PF Revision ID Register" on page 900 10h 09h 0Bh "PCC--PF Class Code Register" on page 900 0B4000h 0Eh 0Eh "PHDR--PF Header Type Register" on page 901 80h 18h 1Bh "PMISCLBAR--PF Miscellaneous Lower Base Address Register" on page 902 00000004h 1Ch 1Fh "PMISCUBAR--PF Miscellaneous Upper Base Address Register" on page 903 00000000h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 891 Table 20-2. Bus M, Device 0, Function 0: Summary of PCIe Intel(R) QuickAssist Configuration Registers (Sheet 2 of 3) Offset Start 20h Offset End 23h Default Value Register ID - Description "PETRINGCSRLBAR--PF Ring CSR Lower Base Address Register" on page 903 00000004h 2Ch 2Dh "PSVID--PF Subsystem Vendor ID Register" on page 904 8086h 2Eh 2Fh "PSID--PF Subsystem ID Register" on page 904 0000h 34h 34h "PCP--PF Capabilities Pointer Register" on page 905 B0h 3Ch 3Ch "PIRQL--PF Interrupt Line Register" on page 905 00h 3Dh 3Dh "PIRQP--PF Interrupt Pin Register" on page 906 01h 60h 60h "PMSI-X--PF Message Signalled Interrupt X Capability ID Register" on page 908 11h 61h 61h "PMSIXNCP--PF MSIX Next Capability Pointer Register" on page 908 6Ch 62h 63h "PMSIXCNTL--PF Message Signalled Interrupt X Control Register" on page 909 0010h 64h 67h "PMSIXTBIR--PF MSI-X Table Offset & Table BIR Register" on page 909 0001B002h 68h 6Bh "PMSIXPBABIR--PF MSI-X Pending Bit Array & BIR Offset Register" on page 910 0001B802h 6Ch 6Ch "PPMCAP--PF Power Management Capabilities ID Register" on page 910 01h 6Dh 6Dh "PPMCP--PF Power Management Next Capability Pointer Register" on page 911 74h 6Eh 6Fh "PPMC--PF Power Management Capabilities Register" on page 911 0023h 70h 73h "PPMCSR--PF Power Management Control and Status Register" on page 912 00000000h 74h 74h "PPCID--PF PCI Express Capability Register" on page 914 10h 75h 75h "PPCP--PF PCI Express Next Capability Pointer Register" on page 914 0h 76h 77h "PPCR--PF PCI Express Capabilities Register" on page 915 0002h 78h 7Bh "PPDCAP--PF PCI Express Device Capabilities Register" on page 915 10008041h 7Ch 7Dh "PPDCNTL--PF PCI Express Device Control Register" on page 917 2810h 7Eh 7Fh "PPDSTAT--PF PCI Express Device Status Register" on page 918 0000h 80h 83h "PLCAPR--PF Link Capabilities Register" on page 919 3B502h 84h 85h "PLCNTLR--PF Link Control Register" on page 921 0000h 86h 87h "PLSR--PF Link Status Register" on page 923 0102h 98h 9Bh "PDCAPR2--PF Device Capabilities 2 Register" on page 924 00000012h 9Ch 9Dh "PDCNTR2--PF Device Control 2 Register" on page 925 0000h A4h A5h "PLCNTLR2--PF Link Control 2 Register" on page 926 0002h A6h A7h "PLSR2--PF Link Status 2 Register" on page 928 0000h B0h B0h "PMSICID--PF Message Signalled Interrupt Capability ID Register" on page 928 05h B1h B1h "PMSINCP--PF Message Signalled Interrupt Next Capability Pointer Register" on page 929 60h B2h B3h "PMSICTL--PF Message Signalled Interrupt Control Register" on page 929 0180h B4h B7h "PMSILADDR--PF Message Signalled Interrupt Lower Address Register" on page 930 00000000h B8h BBh "PMSIUADDR--PF Message Signalled Interrupt Upper Address Register" on page 930 00000000h BCh BDh "PMSIDATA--PF Message Signalled Interrupt Data Register" on page 931 0000h C0h C3h "PMSIMSK - PF Message Signalled Interrupt Mask Register" on page 931 00000000h C4h C7h "PMSIPND--PF Message Signalled Interrupt Pending Register" on page 932 00000000h 100h 103h "PPCIEAERCAPID--PF PCI Express AER Capability ID Register" on page 932 13810001h Intel(R) Communications Chipset 89xx Series - Datasheet 892 October 2012 Order Number: 327879-001US 20.0 Table 20-2. Bus M, Device 0, Function 0: Summary of PCIe Intel(R) QuickAssist Configuration Registers (Sheet 3 of 3) Offset Start Offset End Default Value Register ID - Description "PPAERUCS--PF PCI Express AER Uncorrectable Error Status Register" on page 933 0h 104h 107h 108h 10Bh "PPAERUCM--PF PCI Express AER Uncorrectable Error Mask Register" on page 934 0h 10Fh "PPAERUCSEV--PF PCI Express AER Uncorrectable Error Severity Register" on page 935 10Ch 00062030h 110h 113h "PPAERCS--PF PCI Express AER Correctable Error Register" on page 936 00h 114h 117h "PPAERCM--PF PCI Express AER Correctable Error Mask Register" on page 937 2000h 118h 11Bh "PPAERCTLCAP--PF PCI Express AER Control and Capability Register" on page 938 0h 11Ch 11Fh "PPAERHDRLOG0--PF PCI Express AER Header Log 0 Register" on page 938 0h 120h 123h "PPAERHDRLOG1--PF PCI Express AER Header Log 1 Register" on page 939 0h 124h 127h "PPAERHDRLOG2--PF PCI Express AER Header Log 2 Register" on page 939 0h 128h 12Bh "PPAERHDRLOG3--PF PCI Express AER Header Log 3 Register" on page 940 0h 138h 13Bh "PARIDHDR--PF Alternative Routing ID Capability Header" on page 941 1401000Eh 13Ch 13Dh "PFARICAP--PF ARI Capabilities Register" on page 941 0100h 13Eh 13Fh "PARIDCTL--PF Alternative Routing ID Control Register" on page 942 00000000h 140h 141h "PSRIOVCAPID--PF SR IOV Capability ID Register" on page 942 10h 142h 143h "PSRIOVCVNC--PF SRIOV Capability Version and Next Capability Pointer Register" 1h on page 943 144h 147h "PSRIOVCAP--PF SRIOV Capabilities Register" on page 944 00000000h 148h 14Bh "PSRIOVCS--PF SRIOV Control and Status Register" on page 944 00000000h 14Ch 14Fh "PSRIOVMTOTINI--PF SRIOV Initial and Total VFs Register" on page 945 00100010h 150h 153h "PSRIOVNUMVF--PF SRIOV Number of VFs Register" on page 945 00000000h 154h 155h "PSRIOVFVFO--PF SRIOV First VF Offset Register" on page 946 0008h 156h 157h "PSRIOVVFS--PF SRIOV VF Stride Register" on page 946 0001h 158h 15Bh "PSRIOVFDID--PF SRIOV VF Device ID Register" on page 947 04420000h 15Ch 15Fh "PSRIOVPAGESIZE--PF SRIOV Supported Page Size Register" on page 947 00000553h 160h 163h "PSRIOVSYSPS--PF SRIOV System Page Size Register" on page 948 00000001h 164h 167h "PSRIOVLBAR0--SRIOV Lower BAR0 Register" on page 949 00000004h 168h 16Bh "PSRIOVUBAR0--SRIOV Upper BAR0 Register" on page 950 00000000h 16Ch 16Fh "PSRIOVLBAR1--SRIOV Lower BAR1 Register" on page 951 00000004h 170h 173h "PSRIOVUBAR1--SRIOV Upper BAR1 Register" on page 952 00000000h 17Ch 17Fh "PSRIOVVFMA--PF SRIOV VF Migration Array Register" on page 952 00000000h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 893 Table 20-3. Bus M, Device 0, Function 0: Summary of PCIe Intel(R) QuickAssist Registers Mapped Through PMISCRBAR+1A000h Memory BAR Offset Start 008h Offset End 00Bh Default Value Register ID - Description "ERRSOU2--Error Source Register 2" on page 982 00000000h 018h 01Bh "ERRMSK2--Error Source Mask Register 2" on page 982 00000000h 024h 027h "SINTPF--Signal Raw PF Interrupt Register" on page 983 00000000h 028h 02Bh "SMIAPF--Signal IA PF Interrupt Mask Register" on page 983 00000000h 03Ch 03Fh "GBECFGMMIOV - GBE Configuration and MMIO Valid Register" on page 984 00000000h 040h / at 14h 043h / at 14h "SINTGBE[0:3]--Signal Raw PF Interrupt for GBE Register" on page 985 00000000h 044h / at 14h 047h / at 14h "SMIAGBE[0:3]--Signal IA PF Interrupt Mask GBE Register" on page 985 000003FFh Table 20-4. Bus M, Function 0: Summary of PCIe Intel(R) QuickAssist Registers Mapped Through PMISCSBAR+1A000h Memory BAR Offset Start 03Ch Offset End 03Fh Default Value Register ID - Description "GBECFGMMIOV - GBE Configuration and MMIO Valid Register" on page 984 00000000h Table 20-5. Bus M, Function 8+Index 1: Summary of PCIe Intel(R) QuickAssist Configuration Registers (Sheet 1 of 2) Offset Start Offset End Default Value Register ID - Description 00h 01h "VVID[0:15]--VF Vendor Identification Register" on page 953 FFFFh 02h 03h "VDID[0:15]--VF Device Identification Register" on page 953 FFFFh 04h 05h "VPCICMD[0:15]--VF Device Command Register" on page 954 0000h 06h 07h "VPCISTS[0:15]--VF PCI Device Status Register" on page 955 0010h 08h 08h "VRID[0:15]--VF Revision ID Register" on page 956 10h 09h 0Bh "VCC[0:15]--VF Class Code Register" on page 956 0B4000h 0Eh 0Eh "VHDR[0:15]--VF Header Type Register" on page 956 00h 2Ch 2Dh "VSVID[0:15]--VF Subsystem Vendor ID Register" on page 957 8086h 2Eh 2Fh "VSID[0:15]--VF Subsystem ID Register" on page 957 0000h 34h 34h "VCP[0:15]--VF Capabilities Pointer Register" on page 958 90h 3Ch 3Ch "VIRQL[0:15]--VF Interrupt Line Register" on page 958 00h 3Dh 3Dh "VIRQP[0:15]--VF Interrupt Pin Register" on page 958 00h 50h 50h "VPCID[0:15]--VF PCI Express Capability ID Register" on page 959 10h 51h 51h "VPCP[0:15]--VF PCI Express Next Capability Pointer Register" on page 959 0h 52h 53h "VPCR[0:15]--VF PCI Express Capabilities Register" on page 960 002h 54h 54h "VPDCAP[0:15]--VF PCI Express Device Capabilities Register" on page 960 10008041h Intel(R) Communications Chipset 89xx Series - Datasheet 894 October 2012 Order Number: 327879-001US 20.0 Table 20-5. Bus M, Function 8+Index 1: Summary of PCIe Intel(R) QuickAssist Configuration Registers (Sheet 2 of 2) Offset Start 58h Offset End 59h Default Value Register ID - Description "VPDC[0:15]--VF PCI Express Device Control Register" on page 961 00h 5Ah 5Bh "VPDS[0:15]--VF PCI Express Device Status Register" on page 962 0000h 5Ch 5Fh "VLCR[0:15]--VF Link Capabilities Register" on page 963 3B502h 60h 61h "VLCNTLR[0:15]--VF Link Control Register" on page 965 0000h 62h 63h "VLSR[0:15]--VF Link Status Register" on page 966 0000h 00000012h 74h 77h "DCAPR2[0:15]--Device Capabilities 2 Register" on page 966 90h 90h "VMSICID[0:15]--Message Signalled Interrupt Capability ID Register" on page 967 05h 91h 91h "VMSINCP[0:15]--Message Signalled Interrupt Next Capability Pointer Register" on page 967 00h 92h 93h "VMSICTL[0:15]--Message Signalled Interrupt Control Register" on page 968 0180h 94h 97h "VMSILADDR[0:15]--Message Signalled Interrupt Lower Address Register" on page 968 00000000h 98h 9Bh "VMSIUADDR[0:15]--Message Signalled Interrupt Upper Address Register" on page 969 00000000h 0000h 9Ch 9Dh "VMSIDATA[0:15]--Message Signalled Interrupt Data Register" on page 969 100h 103h "VPCIEAERCAPID[0:15]--VF PCI Express AER Capability ID Register" on page 970 13810001h 104h 107h "VPAERUCS[0:15]--VF PCI Express AER Uncorrectable Error Status Register" on page 971 0h 108h 10Bh "VPAERUCM[0:15]--VF PCI Express AER Uncorrectable Error Mask Register" on page 972 0h 10Ch 10Fh "VPAERUCSEV[0:15]--VF PCI Express AER Uncorrectable Error Severity Register" on page 974 00h 110h 113h "VPAERCS[0:15]--VF PCI Express AER Correctable Error Status Register" on page 975 0h 114h 117h "VPAERCM[0:15]--VF PCI Express AER Correctable Error Mask Register" on page 976 0h 118h 11Bh "VPAERCTLCAP[0:15]--VF PCI Express AER Control and Capability Register" on page 977 0h 11Ch 11Fh "VPAERHDRLOG0[0:15]--VF PCI Express AER Header Log 0 Register" on page 978 0h 120h 123h "VPAERHDRLOG1[0:15]--VF PCI Express AER Header Log 1 Register" on page 979 0h 124h 127h "VPAERHDRLOG2[0:15]--VF PCI Express AER Header Log 2 Register" on page 979 0h 128h 12Bh "VPAERHDRLOG3[0:15]--VF PCI Express AER Header Log 3 Register" on page 980 0h 138h 13Bh "VARIDHDR[0:15]--VF Alternative Routing ID Capability Header" on page 980 1000Eh 13Ch 13Dh "VFARICAP[0:15]--VF ARI Capabilities Register" on page 981 00000000h 13Eh 13Fh "VARIDCTL[0:15]-- VF Alternative Routing ID Control Register" on page 981 00000000h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 895 20.2.2 PCI Standard Header Registers 20.2.2.1 PVID--PF Vendor Identification Register The VID Register contains the vendor identification number. This 16-bit register combined with the Device Identification Register uniquely identifies any PCI device. Writes to this register have no effect. Table 20-6. PVID--PF Vendor Identification Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Offset Start: 00h Offset End: 01h Bus:Function: M:0 Default: 8086h Bit Range Bit Acronym 15 :00 VID 20.2.2.2 Offset Start: 00h Offset End: 01h Bus:Device:Function: B:0:0 Power Well: Core Bit Description Sticky Vendor Identification: This register field contains the PCI standard identification for Intel, 8086h. Bit Reset Value Bit Access 8086h RO PDID--PF Device Identification Register This 16-bit register combined with the Vendor Identification register uniquely identifies any PCI device. Writes to this register have no effect. Table 20-7. PDID--PF Device Identification Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Offset Start: 02h Offset End: 03h Bus:Function: M:0 Default: 0434h Bit Range Bit Acronym 15 :00 DID Power Well: Core Bit Description Device Identification Number: PCH ID = 0434h. Intel(R) Communications Chipset 89xx Series - Datasheet 896 Offset Start: 02h Offset End: 03h Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 0434h RO October 2012 Order Number: 327879-001US 20.0 20.2.2.3 PPCICMD--PF Device Command Register Table 20-8. PPCICMD--PF Device Command Register (Sheet 1 of 2) Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0000h Bit Range Bit Acronym 15 :11 Reserved Offset Start: 04h Offset End: 05h Offset Start: 04h Offset End: 05h Power Well: Core Bit Reset Value Bit Access Reserved 0h RV 0h RW Bit Description Sticky 10 INTD Interrupt Disable: Setting this bit disables generation of INTX messages by the EP.. Default value is 0 which enables the INTX message generation. 09 FBTB Fast Back-to-Back Enable: EP does not implement this functionality and it is not applicable to PCIe devices. The bit is hardwired to 0. 0h RO 08 SER SERR# Enable: When set, this bit enables the non-fatal and fatal errors detected by the EP to be reported to the RC. The error reporting can also be enabled via the PCIe specific bits in the PCIe device control register (Section 20.2.5.5, "PPDCNTL--PF PCI Express Device Control Register") The default value of this bit is 0. 0h RW 07 Reserved Reserved/Does not apply to PCIe. 0h RV 0h RW 06 PER Parity Error Enable: Controls the setting of the Master Data Parity Error bit in the Device Status Register (Section 20.2.5.6, "PPDSTAT--PF PCI Express Device Status Register") The Master Data Parity Error bit is set by the EP if its Parity Error Enable bit is set and either of the following two conditions occurs: * If the EP receives a poisoned Completion from the RC * If the EP poisons a write request. If the Parity Error Enable bit is cleared, the Master Data Parity Error status bit is never set The default value of this bit is 0. 05 VPS VGA Palette Snoop Enable: The device does not implement this functionality/Does not apply to PCIe. The bit is hardwired to 0. 0h RO 04 MWE Memory Write and Invalidate Enable: The device does not implement this functionality/Does not apply to PCIe. The bit is hardwired to 0. 0h RO 03 SS Special Cycle Enable: The device does not implement this functionality/Does not apply to PCIe. The bit is hardwired to 0. 0h RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 897 Table 20-8. PPCICMD--PF Device Command Register (Sheet 2 of 2) Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Offset Start: 04h Offset End: 05h Bus:Device:Function: B:0:0 Offset Start: 04h Offset End: 05h Bus:Function: M:0 Default: 0000h Power Well: Core Bit Reset Value Bit Access BM Bus Master Enable: Controls the ability of EP to issue Memory Read/Write Requests. Clearing (0) this bit prevents EP from issuing any Memory Requests. Because MSIs are in-band memory writes, disabling the bus master enable bit disables MSI as well. PCIe messages are not affected by this bit. See Section 23.3.2, "BME (Bus Master Enable)" for more details on how the EP handles internal commands when this bit is clear. 0h RW 01 MEM Memory Space Enable: Setting this bit enables access to the memory regions the device claims through its BARs. EP will return "unsupported request" completion status & error message in response to memory transactions it receives when this bit is clear. 0h RW 00 IO I/O Space Enable: The device does not implement this functionality since it claims no I/O regions. The bit is hardwired to 0. 0h RO Bit Range 02 20.2.2.4 Bit Acronym Bit Description Sticky PPCISTS--PF Device Status Register Table 20-9. PPCISTS--PF PCI Device Status Register (Sheet 1 of 2) Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range Offset Start: 06h Offset End: 07h Bus:Function: M:0 Default: 0010h Bit Acronym Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 15 DPE Detected Parity Error: This bit is set by EP whenever it receives a Poisoned TLP, regardless of the state the Parity Error Enable bit in the Command register. Default value of this field is 0. 0h RW1C 14 SSE Signaled System Error: This bit is set by the EP when it sends a ERR_FATAL or ERR_NONEATAL message and the SERR bit in the Device Command register bit is set. Default value of this field is 0. 0h RW1C 13 RMA Received Master Abort Status: This bit is set when EP, as a Requestor receives a Completion with Unsupported Request Completion Status. Default value of this field is 0. 0h RW1C 12 RTA Received Target Abort Status: This bit is set when EP, as a Requestor receives a Completion with Completer Abort Completion Status. Default value of this field is 0. 0h RW1C Intel(R) Communications Chipset 89xx Series - Datasheet 898 Offset Start: 06h Offset End: 07h Bus:Device:Function: B:0:0 October 2012 Order Number: 327879-001US 20.0 Table 20-9. PPCISTS--PF PCI Device Status Register (Sheet 2 of 2) Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range 11 10 :09 Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0010h Bit Acronym Offset Start: 06h Offset End: 07h Offset Start: 06h Offset End: 07h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access STA Signaled Target Abort Status: This bit is set when EP completes a Request using Completer Abort Completion Status. Default value of this field is 0. 0h RW1C DST DEVSEL Timing: Does not apply to PCI Express. These bits are hardwired to 0. 00b RO 0h RW1C 08 MDPE Master Data Parity Error Detected: This bit is set by EP, as a Requestor if the Parity Error Enable bit in the Command register is 1b and either of the following two conditions occurs: * Requestor receives a Completion marked poisoned * Requestor detects a parity error in the inbound completion data fifo. * Requestor poisons a write Request If the Parity Error Enable bit is 0b, this bit is never set. Default value of this field is 0. 07 FB2B Fast Back-to-Back Capable: Does not apply to PCI Express. The bit is hardwired to 0. 0h RO 06 Reserved Reserved 0h RV 05 MC66 66 MHz Capable: Does not apply to PCI Express. The bit is hardwired to 0. 0h RO 04 CL Capabilities List: This bit is hardwired to 1 to indicate that EP has a capabilities list. 1h RO 03 IS Interrupt Status: Indicates that the EP has transmitted a INTX message and is awaiting servicing. This bit does not includes MSI's generated by the EP. 0h RO Reserved 0h RV 02 :00 Reserved October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 899 20.2.2.4.1 PRID--PF Revision ID Register The value of this register indicates the chip stepping. It is hardwired on chip and reflects the latest revision. Table 20-10. PRID--PF Revision ID Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 8 bit Bit Range 07 :4 03 :0 20.2.2.5 Offset Start: 08h Offset End: 08h Bus:Device:Function: B:0:0 Offset Start: 08h Offset End: 08h Bus:Function: M:0 Default: 10h Power Well: Core Bit Reset Value Bit Access RIDU Major Revision: Steppings which require all masks to be regenerated. 00b: A stepping 01b: B stepping 10b: C stepping 11b: D stepping 0001b RO RIDL Minor Revision: Incremented for each stepping which does not modify all masks. Reset for each major revision. 00b: x0 stepping 01b: x1 stepping 10b: x2 stepping 11b: x3 stepping 0000b RO Bit Acronym Bit Description Sticky PCC--PF Class Code Register Table 20-11. PCC--PF Class Code Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 24 bit Bit Range 23 :0 Offset Start: 09h Offset End: 0Bh Bus:Function: M:0 Default: 0B4000h Power Well: Core Bit Acronym Bit Description CC Class Code: This value indicates the base class, subclass, and interface. 0B4000h = Base class: Processor, Sub-class Co-processor, no specific register level programming interfaces are defined. Intel(R) Communications Chipset 89xx Series - Datasheet 900 Offset Start: 09h Offset End: 0Bh Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 0B4000h RWOS October 2012 Order Number: 327879-001US 20.0 20.2.2.6 PHDR--PF Header Type Register Table 20-12. PHDR--PF Header Type Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 8 bit Bus:Function: M:0 Default: 80h Bit Range Bit Acronym 07 :00 HDR 20.2.2.7 Bus:Device:Function: B:0:0 Offset Start: 0Eh Offset End: 0Eh Offset Start: 0Eh Offset End: 0Eh Power Well: Core Bit Description Sticky PCI Header Type: The header type of the EP device. 00h = multi-function device with standard header layout. Bit Reset Value Bit Access 80h RO PeQATLBAR--PF QAT Lower Base Address Register This BAR points to the EP Intel(R) QuickAssist Technology. This BAR defined a 512 KB window. Table 20-13. PeSRAMLBAR--PF Memory Lower Base Address Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Device:Function: B:D:0 Bus:Function: M:0 Default: 0000000Ch Bit Reset Value Bit Access Lower Programmable Base Address: These bits are set by BIOS to locate the base address of the region. 0h RW ZERO Lower Bits: Hardwired to 0 (512kB region) - the size of this window is equal to QAT memory capacity 0h RO PREF Prefetchable: Hardwired to 1 to indicate that the region is prefetchable. 0: non-prefetchable 1: prefetchable 1b RO TYP Addressing Type: Hardwired to indicate a 64-bit region. 10b RO MEM Memory Space Indicator: Hardwired to 0 to identify the region as in memory space. 0b RO Bit Acronym Bit Description 31 :19 ADDR 18 :04 02 :01 00 Offset Start: 10h Offset End: 13h Power Well: Core Bit Range 03 Offset Start: 10h Offset End: 13h October 2012 Order Number: 327879-001US Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 901 20.2.2.8 QATUBAR--PF QAT Upper Base Address Register This BAR points to the EP Intel(R) QuickAssist Technology. This BAR defined a 512 KB window. Table 20-14. PeQATUBAR--PF QAT Upper Base Address Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Offset Start: 14h Offset End: 17h Bus:Function: M:0 Default: 00000000h Power Well: Core Bit Range Bit Acronym Bit Description 31 :00 ADDR Upper Programmable Base Address: These bits are set by BIOS to locate the base address of the region. 20.2.2.9 Offset Start: 14h Offset End: 17h Bus:Device:Function: B:D:0 Sticky Bit Reset Value Bit Access 0h RW PMISCLBAR--PF Miscellaneous Lower Base Address Register This BAR is used to access EP and MSI-X Registers. For the EP registers, see Section 20.4, "EP Memory Mapped Registers". The MSI-X tables are also included in the EP register section, starting from Section 20.4.2.5, "GBECFGMMIOV--GBE Configuration and MMIO Valid Register" thru Section 29.4.3.5, "PMSIXPBABIR[0:3]--PF MSI-X Pending Bit Array & BIR Offset Register" Table 20-15. PMISCLBAR--PF Miscellaneous Lower Base Address Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Default: 00000004h Power Well: Core Bit Reset Value Bit Access Lower Programmable Base Address: These bits are set by BIOS to locate the base address of the region. 0h RW ZERO Lower Bits: Hardwired to 0 (128KB region). 0h RO PREF Prefetchable: Hardwired to 0 to indicate that the region is non-prefetchable. 0: non-prefetchable 1: prefetchable 0b RO TYP Addressing Type: Hardwired to indicate a 64-bit region. 10b RO MEM Memory Space Indicator: Hardwired to 0 to identify the region as in memory space. 0b RO Bit Acronym Bit Description 31 :17 ADDR 16 :04 02 :01 00 Intel(R) Communications Chipset 89xx Series - Datasheet 902 Offset Start: 18h Offset End: 1Bh Bus:Function: M:0 Bit Range 03 Offset Start: 18h Offset End: 1Bh Bus:Device:Function: B:0:0 Sticky October 2012 Order Number: 327879-001US 20.0 20.2.2.10 PMISCUBAR--PF Miscellaneous Upper Base Address Register This BAR is used to access the EP MMIO space. Table 20-16. PMISCUBAR--PF Miscellaneous Upper Base Address Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 00000000h Offset Start: 1Ch Offset End: 1Fh Power Well: Core Bit Range Bit Acronym Bit Description 31 :00 ADDR Upper Programmable Base Address: These bits are set by BIOS to locate the base address of the region. 20.2.2.11 Offset Start: 1Ch Offset End: 1Fh Sticky Bit Reset Value Bit Access 0h RW PETRINGCSRLBAR--PF Ring CSR Lower Base Address Register This BAR points to the Ring Controller rings. 20.2.2.12 PETRINGCSRUBAR--PF Ring CSR Upper Base Address Register Table 20-17. PETRINGCSRLBAR--PF Ring CSR Lower Base Address Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 00000004h Bit Reset Value Bit Access Lower Programmable Base Address: These bits are set by BIOS to locate the base address of the region. 0h RW ZERO Lower Bits: Hardwired to 0 (16KB region). 0h RO PREF Prefetchable: Hardwired to 0 to indicate that the region is non-prefetchable. 0: non-prefetchable 1: prefetchable 0b RO TYP Addressing Type: Hardwired to indicate a 64-bit region. 10b RO MEM Memory Space Indicator: Hardwired to 0 to identify the region as in memory space. 0b RO Bit Acronym Bit Description 31 :14 ADDR 13 :04 02 :01 00 Offset Start: 20h Offset End: 23h Power Well: Core Bit Range 03 Offset Start: 20h Offset End: 23h Sticky This BAR points to the Ring Controller rings. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 903 20.2.2.13 PSVID--PF Subsystem Vendor ID Register This register is a write-once register. Once any byte in the register has been written, the register locks against further writes until reset. Table 20-18. PSVID--PF Subsystem Vendor ID Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Offset Start: 2Ch Offset End: 2Dh Bus:Function: M:0 Default: 8086h Bit Range Bit Acronym 15 :00 SVID 20.2.2.14 Offset Start: 2Ch Offset End: 2Dh Bus:Device:Function: B:0:0 Power Well: Core Bit Description Sticky Subsystem Vendor ID: This field is hardwired to ID assigned to Intel. Bit Reset Value Bit Access 8086h RWOS PSID--PF Subsystem ID Register This register is a write-once register. Once any byte in the register has been written, the register locks against further writes until reset. Table 20-19. PSID--PF Subsystem ID Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Offset Start: 2Eh Offset End: 2Fh Bus:Function: M:0 Default: 0000h Power Well: Core Bit Range Bit Acronym Bit Description 15 :00 SID Subsystem ID: Vendor supplied device ID. Default is 0h. Intel(R) Communications Chipset 89xx Series - Datasheet 904 Offset Start: 2Eh Offset End: 2Fh Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 0h RWOS October 2012 Order Number: 327879-001US 20.0 20.2.2.15 PCP--PF Capabilities Pointer Register The Capabilities Pointer Register provides the offset in configuration space to the location where the first set of capabilities registers is located. Table 20-20. PCP--PF Capabilities Pointer Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 8 bit Bus:Function: M:0 Default: B0h Bit Range Bit Acronym 07 :00 CP 20.2.2.16 Bus:Device:Function: B:0:0 Offset Start: 34h Offset End: 34h Offset Start: 34h Offset End: 34h Power Well: Core Bit Description Sticky Pointer to First Capability Structure: Value points to the configuration space offset of the first capability structure (MSI). Bit Reset Value Bit Access B0h RO PIRQL--PF Interrupt Line Register Table 20-21. PIRQL--PF Interrupt Line Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 8 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 00h Offset Start: 3Ch Offset End: 3Ch Power Well: Core Bit Range Bit Acronym Bit Description 07 :00 IRQL Interrupt Line: BIOS writes the interrupt routing information to this register to indicate which input of the interrupt controller this device is connected to. The device itself does not use this information. October 2012 Order Number: 327879-001US Offset Start: 3Ch Offset End: 3Ch Sticky Bit Reset Value Bit Access 0h RW Intel(R) Communications Chipset 89xx Series - Datasheet 905 20.2.2.17 PIRQP--PF Interrupt Pin Register Table 20-22. PIRQP--PF Interrupt Pin Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 8 bit Offset Start: 3Dh Offset End: 3Dh Bus:Function: M:0 Default: 01h Power Well: Core Bit Range Bit Acronym Bit Description 07 :00 IRQP Interrupt Pin: Set to 01h to indicate that EP uses INTA# as its interrupt pin. 20.2.2.18 Offset Start: 3Dh Offset End: 3Dh Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 01h RO SKU: SKU Register This SKU Register (040h) in conjunction with SKU Register 2 (04Ch) determines the PCH SKU. Use the following procedure to determine the SKU: 1. Check the "Bus:Device:Function: M:0:0:08h" register to determine if A0, B0, or C0 stepping. 2. IF A0 then it is SKU4 IF NOT A0 AND bits[9:0] = 3FFh THEN it is SKU1 ELSE go to SKU Register 2 (04Ch). Table 20-23. SKU Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Offset Start: 040h Offset End: 043h Bus:Function: M:0 Default: xxxxxxxxh Bit Range Bit Acronym 31 :10 Reserved 09 :00 SKU Power Well: Core Sticky Bit Reset Value Bit Access Reserved N xh RV SKU Register calculation value N xh RO Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 906 Offset Start: 040h Offset End: 043h Bus:Device:Function: B:D:0 October 2012 Order Number: 327879-001US 20.0 20.2.2.19 SKU Register 2 S This SKU Register 2 (04Ch) in conjunction with SKU Register (040h) determines the PCH SKU. Use the following procedure to determine the SKU (Complete steps 1 & 2 for SKU Register (040h) prior to attempting the following steps): 3. (For non-A0 steppings of the PCH) IF bits[22:19] =0111b it is SKU4 IF bits[22:19] = 0110 AND bits[17:16] = 01 it is SKU3 ELSE it is SKU2 Table 20-24. SKU Register 2 Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Device:Function: B:D:0 Bus:Function: M:0 Default: xxxxxxxxh Bit Range Bit Acronym 31 :23 Reserved 22 :19 SKUR2 Offset Start: 04Ch Offset End: 04Fh Offset Start: 04Ch Offset End: 04Fh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved N xh RV xh RO SKU Register 2A calculation values N Reserved Reserved N xh RV 17 :16 SKUR2B SKU Register 2B calculation value N xh RO 15 :00 Reserved Reserved N xh RV 18 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 907 20.2.3 MSI-X Capability Structure 20.2.3.1 PMSI-X--PF Message Signalled Interrupt X Capability ID Register MSI-X Capability ID Register indicates that the device is capable of generating MSI-X Interrupts. Table 20-25. PMSI-X--PF Message Signalled Interrupt X Capability ID Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 8 bit Offset Start: 60h Offset End: 60h Bus:Function: M:0 Default: 11h Bit Range Bit Acronym 07 :00 MSIX 20.2.3.2 Offset Start: 60h Offset End: 60h Bus:Device:Function: B:0:0 Power Well: Core Bit Description Sticky Capability ID: PCI SIG assigned capability record ID (11h, MSI-X capability). 11h identifies the EP as a device that is MSI-X capable. Bit Reset Value Bit Access 11h RO PMSIXNCP--PF MSIX Next Capability Pointer Register Table 20-26. PMSIXNCP--PF MSIX Next Capability Pointer Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 8 bit Offset Start: 61h Offset End: 61h Bus:Function: M:0 Default: 6Ch Bit Range Bit Acronym 07 :00 MCP Power Well: Core Bit Description Next Capability Pointer: Power Management Capability Intel(R) Communications Chipset 89xx Series - Datasheet 908 Offset Start: 61h Offset End: 61h Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 6Ch RO October 2012 Order Number: 327879-001US 20.0 20.2.3.3 PMSIXCNTL--PF Message Signalled Interrupt X Control Register Table 20-27. PMSIXCNTL--PF Message Signalled Interrupt X Control Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range 15 14 13 :11 10 :00 20.2.3.4 Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0010h FM Reserved Power Well: Core Bit Access MSI-X Enable: This bit enables the EP to generates interrupts using the MSI-X tables instead of the legacy INTx messages. When this bit is set to 1, the EP will not generate INTx messages and must use the MSI-X to signal interrupts. The device driver should not clear this bit to mask interrupts. This bit will be set by the system PCI device manager. 0h RW Function Mask: This bit controls the masking of all vectors implemented in the EP. When this bit is 0, each vectors mask bit determines whether the vector is masked or not. Whent this bit is 1, all vecotrs are masked regardless of the per vector masking bit. 0h RW Always returns 0's when read. 0h RO 10h RO Bit Description Sticky MSI-X Table Size: Number of vectors (encoded as N-1) supported by the EP. The EP supports 17 vectors. TS Offset Start: 62h Offset End: 63h Bit Reset Value Bit Acronym MSIXEN Offset Start: 62h Offset End: 63h PMSIXTBIR--PF MSI-X Table Offset & Table BIR Register Table 20-28. PMSIXTBIR--PF MSI-X Table Offset & Table BIR Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Function: M:0 Default: 0001B002h Bit Range Bit Acronym 31 :3 TO 2 :0 Bus:Device:Function: B:0:0 TBIR October 2012 Order Number: 327879-001US Offset Start: 64h Offset End: 67h Offset Start: 64h Offset End: 67h Power Well: Core Bit Description Sticky Table Offset: Offset to a location in one of the EP's BARs (indicated by TBIR) that points to the location of the base of the MSI-X Table. The EP MSI-X Table maps to an offset of 108KB in PMISCBAR. Table BAR Indicator Register: The BIR points to the EP PMISCBAR register (18h) Bit Reset Value Bit Access 000000000000 000110110000 00000b RO 010b RO Intel(R) Communications Chipset 89xx Series - Datasheet 909 20.2.3.5 PMSIXPBABIR--PF MSI-X Pending Bit Array & BIR Offset Register Table 20-29. PMSIXPBABIR--PF MSI-X Pending Bit Array & BIR Offset Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range 31 :3 2 :00 Offset Start: 68h Offset End: 6Bh Bus:Device:Function: B:0:0 Offset Start: 68h Offset End: 6Bh Bus:Function: M:0 Default: 0001B802h Power Well: Core Bit Acronym Bit Description PBAO Pending Bit Array Offset: Offset to a location in one of the EP BAR's (indicated by PBABIR) that points to the location of the base of the MSI-X PBA. The EP MSI-X PBA maps to an offset of 110KB in PMISCBAR. PBABIR Sticky Pending Bit Array BAR Indicator Register: The PBABIR points to the EP PMISCBAR register (18h) Bit Reset Value Bit Access 000000000000 000110111000 00000000b RO 010b RO 20.2.4 Power Management Capability Structure 20.2.4.1 PPMCAP--PF Power Management Capabilities ID Register EP supports the PCI bus Power Management Interface Specification Rev 1.2 based PCIe Power Management. The specification requires the implementation of the PCI Power Management Capabilities Register. Table 20-30. PPMCAP--PF Power Management Capabilities ID Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 8 bit Offset Start: 6Ch Offset End: 6Ch Bus:Function: M:0 Default: 01h Bit Range Bit Acronym 07 :00 PMID Power Well: Core Bit Description Capability ID: PCI SIG assigned capability record ID (01h, Power Management capability) Intel(R) Communications Chipset 89xx Series - Datasheet 910 Offset Start: 6Ch Offset End: 6Ch Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 01h RO October 2012 Order Number: 327879-001US 20.0 20.2.4.2 PPMCP--PF Power Management Next Capability Pointer Register Table 20-31. PPMCP--PF Power Management Next Capability Pointer Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 8 bit Bus:Function: M:0 Default: 74h Bit Range Bit Acronym 07 :00 PMCP 20.2.4.3 Bus:Device:Function: B:0:0 Offset Start: 6Dh Offset End: 6Dh Offset Start: 6Dh Offset End: 6Dh Power Well: Core Bit Description Sticky Next Capability Pointer: PCI Express Capability. Bit Reset Value Bit Access 74h RO PPMC--PF Power Management Capabilities Register EP supports the PCI bus Power Management Interface Specification Rev 1.2 based PCIe Power Management. The specification requires the implementation of the PCI Power Management Capabilities Register. Table 20-32. PPMC--PF Power Management Capabilities Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0023h Bit Range Bit Acronym 15 :11 PME Offset Start: 6Eh Offset End: 6Fh Power Well: Core Bit Description Sticky PME Support: Set to indicate that the EP will not assert PME# in any state. 10 D2 9 D1 D1 Support. EP does not support D1 state. 8 :6 AC Aux current: Not relevant for EP. 5 DSI 4 Reserved 3 PMC 2 :0 PMV Version. This field is set to 3 to indicate that the EP is compliant with the PCI bus Power Management Interface Specification Rev 1.2. October 2012 Order Number: 327879-001US Offset Start: 6Eh Offset End: 6Fh D2 Support. EP does not support D2 state. Bit Reset Value Bit Access 00000b RO 0b RO 0b RO 000b RO Device specific initialization: A one in this bit indicates that immediately after entry into the D0 Uninitialized state, the function requires additional configuration above and beyond setup of its PCI configuration Header registers before the Class driver can use the function. For the EP this bit is set to 1. 1b RO Reserved 0b RV PME clock. Does not apply to PCIe. 0b RO 011b RO Intel(R) Communications Chipset 89xx Series - Datasheet 911 20.2.4.4 PPMCSR--PF Power Management Control and Status Register EP supports the PCI bus Power Management Interface Specification Rev 1.2 based PCIe Power Management. The specification requires the implementation of the PCI Power Management Capabilities Register. Table 20-33. PPMCSR--PF Power Management Control and Status Register (Sheet 1 of 2) Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Offset Start: 70h Offset End: 73h Bus:Function: M:0 Default: 00000000h Bit Range Bit Acronym 31 :16 Reserved Power Well: Core Bit Description Reserved:Not Supported for EP. Sticky Bit Reset Value Bit Access 0000h RO 0b RO PMS PME Status: Hardwired to 0 since EP will not generate PME. DSC Data Scale: Set to 0 because the EP does not implement the Data register. If the Data register has not been implemented, this field must return "00b" when the PMCSR is read. 00b RO DSEL Data Select: Set to 0 for EP because the EP does not implement the Data register (The Data register is an optional, 8-bit read-only register that provides a mechanism for the function to report state dependent operating data such as power consumed or heat dissipation) If the Data register is not implemented, this field should be read only and return "0000b" when the PMCSR is read. 0000b RO 8 PME PME Enable: Hardwired to 0 since EP does not generate a PME. 0b RO 7 :4 Reserved 0000b RV 15 14 :13 12 :9 Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 912 Offset Start: 70h Offset End: 73h Bus:Device:Function: B:0:0 October 2012 Order Number: 327879-001US 20.0 Table 20-33. PPMCSR--PF Power Management Control and Status Register (Sheet 2 of 2) Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 00000000h Bit Reset Value Bit Access No_Soft_Reset: Set to 0. The EP will not maintain functional context in the D3_hot state. So when the EP is transitioned from D3_hot to D0 it will be in the D0_uniniitialized state and therefore would require software to initialize it before it can transition to the D0_initialized state. When set ("1"), this bit indicates that devices transitioning from D3hot to D0 because of Power State commands do not perform an internal reset. Configuration Context is preserved. Upon transition from the D3_hot to the D0_Initialized state, no additional operating system intervention is required to preserve Configuration Context beyond writing the PowerState bits. When clear ("0"), devices do perform an internal reset upon transitioning from D3_hot to D0 via software control of the Power State bits. Configuration Context is lost when performing the soft reset. Upon transition from the D3_hot to the D0 state, full reinitialization sequence is needed to return the device to D0_initialized state 0b RO Reserved 0b RO 00b RW Bit Description 3 NSR 2 Reserved PS October 2012 Order Number: 327879-001US Offset Start: 70h Offset End: 73h Power Well: Core Bit Acronym 1 :0 Offset Start: 70h Offset End: 73h Sticky Power State. This field is used to determine the current power state of EP and to set a new power state. * 00: D0 * 01: D1 (Not supported by EP, ignore writes with this value) * 10: D2 (Not supported by EP, ignore writes with this value) * 11: D3_hot If software selects a Power state that is not supported by the EP (D2/D1), the writes must complete normally on PCIe, but the write data is discarded and no state change occurs. Intel(R) Communications Chipset 89xx Series - Datasheet 913 20.2.5 PCI Express Capability Structure 20.2.5.1 PPCID--PF PCI Express Capability ID Register The PCI Express Capability List register enumerates the PCI Express Capability structure in the PCI 3.0 configuration space capability list. Table 20-34. PPCID--PF PCI Express Capability Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 8 bit Offset Start: 74h Offset End: 74h Bus:Function: M:0 Default: 10h Bit Range Bit Acronym 07 :00 PCIECID 20.2.5.2 Offset Start: 74h Offset End: 74h Bus:Device:Function: B:0:0 Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 10h RO Capability ID: PCI SIG assigned capability record ID (10h, PCI Express capability) PPCP--PF PCI Express Next Capability Pointer Register Table 20-35. PPCP--PF PCI Express Next Capability Pointer Register Description: View: PCI BAR: Configuration Bus:Device:Function: B:0:0 Offset Start: 75h Offset End: 75h View: PCI PF BAR: Configuration Bus:Device:Function: M:0 Offset Start: 75h Offset End: 75h Size: 8 bit Default: 0h Bit Range Bit Acronym 07 :00 PCIENP Power Well: Core Bit Description Next Capability Pointer: Last Capability. Intel(R) Communications Chipset 89xx Series - Datasheet 914 Sticky Bit Reset Value Bit Access 00h RO October 2012 Order Number: 327879-001US 20.0 20.2.5.3 PPCR--PF PCI Express Capabilities Register Table 20-36. PPCR--PF PCI Express Capabilities Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0002h Bit Range Bit Acronym 15 :14 Reserved Offset Start: 76h Offset End: 77h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00b RO 00000b RO 0b RO Device/Port Type: Indicates the type of PCI Express logical device. Hardwired to 0000b (PCIe Endpoint) 0000b RO Capability Version: Indicates PCI-SIG defined PCI Express capability structure version number. EP is PCIe 2.0 Specification Compliant. 0010b RO 13 :9 IMN Interrupt Message Number: This field indicates which MSI vector is used for the interrupt message generated in association with the status bits in either the Slot Status field of this capability structure (applies to only RC or Switch) Not applicable to EP. 88 SI Slot Implemented: This bit when set indicates that the PCI Express Link associated with this port is connected to a slot. Hardwired to 0 for EP. 7 :4 DPT 3 :0 CV 20.2.5.4 Offset Start: 76h Offset End: 77h PPDCAP--PF PCI Express Device Capabilities Register Table 20-37. PPDCAP--PF PCI Express Device Capabilities Register (Sheet 1 of 2) Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 10008041h Offset Start: 78h Offset End: 7Bh Offset Start: 78h Offset End: 7Bh Power Well: Core Bit Reset Value Bit Access Reserved 0h RO FLR Function level reset: When set indicates that the device supports the FLR feature. 1b RO 27 :26 CSPS Captured Slot Power Limit Scale: Does not apply to EP. 00b RO 25 :18 CSPV Captured Slot Power Limit Value: Does not apply to EP. 0h RO 17 :16 Reserved Reserved 0h RO Role-Based Error Reporting: Indicates that EP conforms to Role based error reporting ECN for PCIe 1.0a and which was subsequently rolled in PCIe 1.1 and future revisions. 1b RO Bit Range Bit Acronym 31 :29 Reserved 28 15 RBEP October 2012 Order Number: 327879-001US Bit Description Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 915 Table 20-37. PPDCAP--PF PCI Express Device Capabilities Register (Sheet 2 of 2) Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Offset Start: 78h Offset End: 7Bh Bus:Function: M:0 Default: 10008041h Power Well: Core Bit Reset Value Bit Access Attention Button/Indicator Present and Power Indicator Present. None of these are implemented in the EP. 000b RO EL1L Endpoint L1 Acceptable Latency: This field indicates the acceptable latency that an Endpoint can withstand due to the transition from L1 state to the L0 state. Hardcoded to the lowest value of 1us. 000b R0 8 :6 EL0L Endpoint L0s Acceptable Latency: This field indicates the acceptable latency that an Endpoint can withstand due to the transition from L1 state to the L0 state. This value is hardcoded to the latency of 64-128ns. 001b R0 5 ETFS Extended Tag Field Supported: This field indicates the maximum supported size of the Tag field as a Requester. When Clear indicates 5-bit Tag field is supported. This limits the EP to maximum of 32 outstanding requests. 0b R0 4 :3 PFS Phantom Functions Supported: This field indicates the support for use of unclaimed function numbers to extend the number of outstanding transactions allowed. EP does not use this capability. 00b RO MPS Max_Payload_Size Supported: This field indicates the maximum payload size that EP can support for TLPs. This value is set to indicate 256B. The defined encodings are: * 000b = 128B max payload size * 001b = 256B max payload size (Max supported) * 010b - 111b = Reserved 001b RO Bit Range Bit Acronym Bit Description 14 :12 ATTN 11 :9 2 :0 Intel(R) Communications Chipset 89xx Series - Datasheet 916 Offset Start: 78h Offset End: 7Bh Bus:Device:Function: B:0:0 Sticky October 2012 Order Number: 327879-001US 20.0 20.2.5.5 PPDCNTL--PF PCI Express Device Control Register Table 20-38. PPDCNTL--PF PCI Express Device Control Register (Sheet 1 of 2) Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 2810h Offset Start: 7Ch Offset End: 7Dh Offset Start: 7Ch Offset End: 7Dh Power Well: Core Bit Reset Value Bit Access Initiate FLR - Used to initiate FLR transition. A write of 1 initiates FLR transition. Since hardware must not respond to any cycles till FLR completion, the value read by software from this bit is 0. 0b RW MRS Max_Read_Request_Size: This field sets the maximum Read Request size for the EP as a Requester. The EP is capable for generating up to 1kB read requests. However requests generated by the EP will be limited by the programmed value in this field. Defined encodings for this field are: 000b = 128B max read request size 001b = 256B max read request size 010b = 512B max read request size (Default) 011b = 1024B max read request size 100b - 111b = Reserved 010b RW 11 ENS Enable No Snoop (NS): If this bit is set to 1, EP is permitted to set the No Snoop bit in the Requester Attributes of transactions it initiates. When clear all transactions will have the No Snoop bit clear. Setting this bit will not cause the EP to set the No Snoop attribute on every memory requests that it initiates. 1b RW 10 APME Auxiliary (AUX) Power PM Enable: This bit when set enables a device to draw AUX power independent of PME AUX power. Does not apply to EP 0b R0 Bit Range 15 14 :12 Bit Acronym Bit Description STARTFLR Sticky 9 PFE Phantom Functions Enable: When set, this bit enables a device to use unclaimed functions as Phantom Functions. Does not apply to EP. 0b RO 8 ETFE Extended Tag Field Enable: When set, this bit enables a device to use an 8-bit Tag field as a requester. Does not to EP. 0b RO MPS Max_Payload_Size (MPS): This field sets maximum TLP payload for EP. As a Receiver, the EP must handle TLPs as large as the set value; as a Transmitter, the EP must not generate TLPs exceeding the set value. The EP is capable of generating up to 256B MPS. However requests generated by the EP will be limited by the programmed value in this field. It is expected that the root complexes that the EP will be paired up with will be limited to 128B and 256B MPS. 000b = 128B (Default) 001b = 256B Others values not supported by EP 000b RW ERO Enable Relaxed Ordering: If this bit is set, EP is permitted to set the Relaxed Ordering bit in the Attributes field of transactions it initiates. Setting this bit does not cause the EP to set the RO on every transaction it issues 1b RW 7 :5 4 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 917 Table 20-38. PPDCNTL--PF PCI Express Device Control Register (Sheet 2 of 2) Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range Offset Start: 7Ch Offset End: 7Dh Bus:Device:Function: B:0:0 Offset Start: 7Ch Offset End: 7Dh Bus:Function: M:0 Default: 2810h Power Well: Core Bit Acronym Bit Description Sticky Bit Reset Value Bit Access 3 URRO Unsupported Request Reporting Enable: This bit, in conjunction with other bits, controls the signaling of Unsupported Requests by sending Error Messages to the root port. When clear it disables sending of error messages. 0b RW 2 FERE Fatal Error Reporting Enable: This bit, in conjunction with other bits, controls sending ERR_FATAL Messages to the root port. When clear disables sending of error messages. 0h RW 1 NERE Non-Fatal Error Reporting Enable: This bit, in conjunction with other bits, controls sending ERR_NONEATAL Messages to the root port. When clear disables sending of error messages. 0h RW 0 CERE Correctable Error Reporting Enable: This bit, in conjunction with other bits, controls sending ERR_COR Messages to the root port. When clear disables sending of error messages. 0h RW 20.2.5.6 PPDSTAT--PF PCI Express Device Status Register Table 20-39. PPDSTAT--PF PCI Express Device Status Register (Sheet 1 of 2) Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Offset Start: 7Eh Offset End: 7Fh Bus:Function: M:0 Default: 0000h Bit Range Bit Acronym 15 :6 Reserved 5 TP 4 APD Power Well: Core Bit Description Transactions Pending: This bit when set indicates that EP has issued Non-Posted Requests which have not been completed either with a completion packet or completion timeout mechanism. AUX Power Detected: Devices that require AUX power report this bit as set if AUX power is detected by the device. Note: Sticky Bit Reset Value Bit Access 0h RO 0h RO 0h RO Does not apply to this EP function. Intel(R) Communications Chipset 89xx Series - Datasheet 918 Offset Start: 7Eh Offset End: 7Fh Bus:Device:Function: B:0:0 October 2012 Order Number: 327879-001US 20.0 Table 20-39. PPDSTAT--PF PCI Express Device Status Register (Sheet 2 of 2) Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range 3 2 1 0 20.2.5.7 Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0000h Offset Start: 7Eh Offset End: 7Fh Offset Start: 7Eh Offset End: 7Fh Power Well: Core Bit Reset Value Bit Access URD Unsupported Request Detected: This bit indicates that EP received an Unsupported Request. A one indicates that an error was detected since the last time this bit was cleared. Errors are updated in this field regardless of whether the error reporting bit is set in the Device Control Register. 0h RW1C FED Fatal Error Detected: This bit indicates status of Fatal errors detected. A one indicates that an error was detected since the last time this bit was cleared. Errors are updated in this field regardless of whether the error reporting bit is set in the Device Control Register. 0h RW1C NED Non-Fatal Error Detected: This bit indicates status of Nonfatal errors detected. A one indicates that an error was detected since the last time this bit was cleared. Errors are updated in this field regardless of whether the error reporting bit is set in the Device Control Register. 0h RW1C CED Correctable Error Detected: This bit indicates status of correctable errors detected. A one indicates that an error was detected since the last time this bit was cleared. Errors are updated in this field regardless of whether the error reporting bit is set in the Device Control Register. 0h RW1C Bit Acronym Bit Description Sticky PLCAPR--PF Link Capabilities Register Table 20-40. PLCAPR--PF Link Capabilities Register (Sheet 1 of 2) Description: Link Capabilities Register View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 3B502h Bit Range Bit Acronym Offset Start: 80h Offset End: 83h Power Well: Core Bit Description Sticky 31 :24 PORTNUM Port Number: Assigned by EP after link training phase. 23 :22 Reserved Reserved 21 :21 LBN 20 :18 Reserved October 2012 Order Number: 327879-001US Offset Start: 80h Offset End: 83h Link Bandwidth Notification Capability: Note: Does not apply to PCIe Endpoints/EP. Reserved Note: Does not apply to EP Bit Reset Value Bit Access 0h RWOS 00b RO 0b RO 000b RO Intel(R) Communications Chipset 89xx Series - Datasheet 919 Table 20-40. PLCAPR--PF Link Capabilities Register (Sheet 2 of 2) Description: Link Capabilities Register View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range 17 :15 11 :10 9 :4 3 :0 Power Well: Core Bit Reset Value Bit Access 111b RO L0EL L0s Exit Latency- Indicates the exit latency from L0s to L0 state. 000b - Less than 64ns 001b - 64ns - 128ns 010b - 128ns - 256ns 011b - 256ns - 512ns 100b - 512ns - 1 is 101b - 1 is - 2 is 110b - 2 is - 4 is 111b - Reserved 011b RO ASLPM Active State Link PM Support - Indicates the level of active state power management supported in EP. Defined encodings are: 00b - Reserved 01b - L0s Entry Supported 10b - Reserved 11b - L0s and L1 Supported 01b RO LINKW Max Link Width- Indicates the max link width. Relevant encoding: 000000b - Reserved 000001b - x1 000010b - x2 000100b - x4 001000b - x8 001100b - x12 010000b - x16 100000b - x32 EP value depends on SKU. However the max link width is x16. 010000b RO 0010b RO Bit Acronym L1EL Bit Description L1 Exit Latency- Indicates the exit latency from L1 to L0 state. 000b - Less than 1 is 001b - 1 is - 2 is 010b - 2 is - 4 is 011b - 4 is - 8 is 100b - 8 is - 16 is 101b - 16 is - 32 is 110b - 32 is - 64 is 111b - L1 transition not supported MAXSPEED Sticky EP does not support ASPM L1 transition Max Link Speed - Indicates Maximum supported Link Speed. Defined encodings are : 0001b - 2.5Gbs Link speed supported (Gen 1) 0010b - 5.0Gbs Link speed supported (Gen 2) EP indicates a max Link Speed of 5.0 Gbs. Intel(R) Communications Chipset 89xx Series - Datasheet 920 Offset Start: 80h Offset End: 83h Bus:Function: M:0 Default: 3B502h Note: 14 :12 Offset Start: 80h Offset End: 83h Bus:Device:Function: B:0:0 October 2012 Order Number: 327879-001US 20.0 20.2.5.8 PLCNTLR--PF Link Control Register Table 20-41. PLCNTLR--PF Link Control Register (Sheet 1 of 2) Description: Link Control Register View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0000h Bit Acronym Bit Description 15 :12 Reserved Reserved LBWINTE Link Autonomous Bandwidth Interrupt Enable - When set to 1b this bit enables the generation of an interrupt to indicate that the Link Autonomous Bandwidth Status bit has been set. LBWMINTE Link Bandwidth Management Interrupt Enable When set to 1b this bit enables the generation of an interrupt to indicate that the Link Bandwidth Management Status bit has been set. Note: 10 Note: 9 WD 8 7 ECLKPM EXTSYNC CCLKCFG 5 RETRAIN October 2012 Order Number: 327879-001US 0000b RO 0b RO 0b RO 0b RO 0b RO 0b RW 0b RW 0b RO EP does not support this feature. This control applies to all functions. Common Clock Configuration - when set indicates that EP and the root port at the other end of the link are operating with a common reference clock. A value of 0 indicates that they operating with an asynchronous clock. This parameter affects the L0s Exit Latencies. After changing the value in this bit in both components on a Link, software must trigger the Link to retrain by writing a 1b to the Retrain Link bit. This control applies to all functions. Retrain Link: A write of 1b to this bit initiates Link retraining by directing the Physical Layer LTSSM to the Recovery state. Note: Bit Access Not applicable to EP. Extended Sync. When set to one, this bit forces the transmission of: - 4096 FTS ordered sets during the L0s state - followed by a single SKP ordered set prior to entering the L0 state, - as well as the transmission of 1024 TS1 ordered sets in the L1 state prior to entering the Recovery state. This mode gives external devices (e.g., logic analyzers) that may be monitoring Link activity time to achieve bit and symbol lock before the Link enters the L0 or Recovery state and resumes communication. Note: Bit Reset Value Not applicable to EP. Enable Clock Power Management - not supported in EP Note: 6 Sticky Hardware Autonomous Width Disable - When set to 1b this bit disables hardware from changing the link width for reasons other than attempting to correct unreliable link operation by reducing link width. Note: Offset Start: 84h Offset End: 85h Power Well: Core Bit Range 11 Offset Start: 84h Offset End: 85h Does not apply to endpoint/EP Intel(R) Communications Chipset 89xx Series - Datasheet 921 Table 20-41. PLCNTLR--PF Link Control Register (Sheet 2 of 2) Description: Link Control Register View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range 4 Power Well: Core Bit Acronym LINKDIS Bit Description Link Disable: This bit disables the Link by directing the LTSSM to the Disabled state when set to 1b. 3 RCB 2 Reserved ASPMC Bit Reset Value Bit Access 0b RO 0b RW 0b RO 00b RW Does not directly impact the EP. Reserved Active State Link PM Control - this field controls the active state PM supported on the link. Link PM functionality is determined by the lowest common denominator of all functions. Defined encodings are: 00b - PM Disabled 01b - L0s Entry Supported 10b - Reserved 11b - L0s and L1 Supported Note: Sticky Does not apply to endpoint/EP Read Completion Boundary (RCB): For PCIe Endpoints this field is set optionally by configuration software to indicate the RCB value of the Root Port upstream from the Endpoint. * 0b = 64B * 1b = 128 byte Note: This control applies to all functions. Intel(R) Communications Chipset 89xx Series - Datasheet 922 Offset Start: 84h Offset End: 85h Bus:Function: M:0 Default: 0000h Note: 1 :0 Offset Start: 84h Offset End: 85h Bus:Device:Function: B:0:0 October 2012 Order Number: 327879-001US 20.0 20.2.5.9 PLSR--PF Link Status Register Table 20-42. PLSR--PF Link Status Register Description: Link Status Register View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range 15 14 Reserved LBWMS DLACT SCLKCFG Bit Description 11 LTINPROG Reserved LTE 9 :4 NLW Bit Access 0b RO 0b RO 0b RO 0b RWOS 0b RO 0b RO 010000b RO 0010b RO This status is reflected in all functions. Link Training - Indicates that link training is in progress. Does not apply to PCIe Endpoints/EP.. This status is reflected in all functions. Link Training Error - Indicates that a link training error has occurred. Does not apply to PCIe Endpoints/EP. This status is reflected in all functions. Negotiated Link Width: Negotiated Link Width Indicates the negotiated width of the link. Relevant encoding for EP are: 000001b - x1 000010b - x2 000100b - x4 001000b - x8 010000b - x16 Note: Bit Reset Value This status is reflected in all functions. Slot Clock Configuration - When set indicates that EP uses the physical reference clock that the platform provides on the connector. This bit must be cleared if EP uses an independent clock. Note: Offset Start: 86h Offset End: 87h This status is reflected in all functions. Data Link Layer Link Active Not supported by PCIe Endpoints/EP. Note: 10 Sticky Link Bandwidth Management Status: Not supported in EP. Note: Offset Start: 86h Offset End: 87h Power Well: Core Bit Acronym Note: 12 Bus:Function: M:0 Default: 0102h Note: 13 Bus:Device:Function: B:0:0 This status is reflected in all functions. Negotiated Link Speed: The negotiated Link Speed. Defined encodings are: 3 :0 NLS 0001b - 2.5Gbs 0010b - 5.0Gbs Note: October 2012 Order Number: 327879-001US This status is reflected in all functions. Intel(R) Communications Chipset 89xx Series - Datasheet 923 20.2.5.10 PDCAPR2--PF Device Capabilities 2 Register Table 20-43. PDCAPR2--PF Device Capabilities 2 Register Description: Device Capabilities 2 Register View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Acronym 31 :5 Reserved 3 :0 Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved. 0b RV CTODS Completion Timeout Disable Supported. A value of 1b indicates support for the completion timeout disable mechanism. 1b RO CTORS Completion Timeout Ranges Supported: This field indicates support for the optional completion timeout programmability mechanism. This mechanism enables system software to modify the completion timeout value. Four time value ranges are defined: * Range A = 50 us to 10 ms * Range B = 10 ms to 250 ms * Range C = 250 ms to 4 s * Range D = 4 s to 64 s Bits are set according to the following table to show the timeout value ranges that are supported. * 0000b = Completion timeout programming not supported. PCH must implement a timeout value in the range 50 us to 50 ms. * 0001b = Range A. * 0010b = Range B. * 0011b = Ranges A & B. * 0110b = Ranges B & C. * 0111b = Ranges A, B & C. * 1110b = Ranges B, C & D. * 1111b = Ranges A, B, C & D. * All other values are reserved. It is strongly recommended that the completion timeout mechanism not expire in less than 10 ms 0010b RO Intel(R) Communications Chipset 89xx Series - Datasheet 924 Offset Start: 98h Offset End: 9Bh Bus:Function: M:0 Default: 00000012h Bit Range 4 Offset Start: 98h Offset End: 9Bh Bus:Device:Function: B:0:0 October 2012 Order Number: 327879-001US 20.0 20.2.5.11 PDCNTR2--PF Device Control 2 Register Table 20-44. PDCNTR2--PF Device Control 2 Register Description: Link Control 2 Register View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range 15 5 4 3 :0 Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0000h Offset Start: 9Ch Offset End: 9Dh Offset Start: 9Ch Offset End: 9Dh Power Well: Core Bit Reset Value Bit Access Reserved. 0b RV CTODIS Completion Timeout Disable: When set to 1b, this bit disables the completion timeout mechanism. Software is permitted to set or clear this bit at any time. When set, the completion timeout detection mechanism is disabled. If there are outstanding requests when the bit is cleared, it is permitted but not required for hardware to apply the completion timeout mechanism to the outstanding requests. If this is done, it is permitted to base the start time for each request on either the time this bit was cleared or the time each request was issued. The default value for this bit is 0b. 0b RW CTOV Completion Timeout Value: In devices that support completion timeout programmability, this field enables system software to modify the completion timeout value. Encoding: * 0000b = Default range: 50 us to 50 ms. It is strongly recommended that the completion timeout mechanism not expire in less than 10 ms. Values available if Range A (50 us to 10 ms) programmability range is supported: * 0001b = 50us to 100 us. * 0010b = 1 ms to 10 ms. Values available if Range B (10 ms to 250 ms) programmability range is supported: * 0101b = 16 ms to 55 ms. * 0110b = 65 ms to 210 ms. Values available if Range C (250 ms to 4 s) programmability range is supported: * 1001b = 260 ms to 900 ms. * 1010b = 1 s to 3.5 s. Values available if the Range D (4 s to 64 s) programmability range is supported: * 1101b = 4 s to 13 s. * 1110b = 17 s to 64 s. Values not defined are reserved. Software is permitted to change the value in this field at any time. For requests already pending when the completion timeout value is changed, hardware is permitted to use either the new or the old value for the outstanding requests and is permitted to base the start time for each request either when this value was changed or when each request was issued. The default value for this field is 0000b. 0000b RW Bit Acronym Reserved October 2012 Order Number: 327879-001US Bit Description Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 925 20.2.5.12 PLCNTLR2--PF Link Control 2 Register Table 20-45. PLCNTLR2--PF Link Control 2 Register (Sheet 1 of 2) Description: Link Control 2 Register View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Acronym 15 :13 Reserved CDE CSOS EMC Compliance De-emphasis - This bit sets the de-emphasis level in Polling.Compliance state if the entry occurred due to the Enter Compliance bit being 1b. Encodings: 1b -3.5 dB 0b -6 dB TMARG SDEM RV Y 0b RWS Y 0b RWS Y 0b RWS Y 0b RWS 0b RO This bit field is valid for function 0 only. Transmit Margin - This field controls the value of the non deemphasized voltage level at the Transmitter pins. Encodings: 000b - Normal operating range 001b - 800-1200 mV for full swing and 400-700 mV for half-swing. 010b - (n-1) - Values must be monotonic with a non-zero slope. The value of n must be greater than 3 and less than 7. At least two of these must be below the normal operating range of n: 200-400 mV for full-swing and 100200 mV for half-swing. other values - reserved This bit field is valid for function 0 only. Selectable De-emphasis This bit is not applicable and reserved for Endpoints. Intel(R) Communications Chipset 89xx Series - Datasheet 926 0h This bit field is valid for function 0 only. Enter Modified Compliance - When this bit is set to 1b, the device transmits modified compliance pattern if the LTSSM enters Polling.Compliance state. Note: 6 Bit Access This bit field is valid for function 0 only. Compliance SOS - When set to 1b, the LTSSM is required to send SOS periodically in between the (modified) compliance patterns. Note: 97 Bit Reset Value Sticky Reserved Note: 10 Power Well: Core Bit Description Note: 11 Offset Start: A4h Offset End: A5h Bus:Function: M:0 Default: 0002h Bit Range 12 Offset Start: A4h Offset End: A5h Bus:Device:Function: B:0:0 October 2012 Order Number: 327879-001US 20.0 Table 20-45. PLCNTLR2--PF Link Control 2 Register (Sheet 2 of 2) Description: Link Control 2 Register View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range 5 Bit Acronym HWAUTOSD ENCOMP TLNKS October 2012 Order Number: 327879-001US Offset Start: A4h Offset End: A5h Bit Reset Value Bit Access 0b RO Y 0b RWS Y 0010b RWS Sticky Hardware Autonomous Speed Disable. When set to 1b, this bit disables hardware from changing the link speed for reasons other than attempting to correct unreliable link operation by reducing link speed. Hard wire to 0b. This bit field is valid for function 0 only. Enter Compliance. Software is permitted to force a link to enter compliance mode at the speed indicated in the Target Link Speed field by setting this bit to 1b in both components on a link and then initiating a hot reset on the link. The default value of this field following a fundamental reset is 0b. This bit field is valid for function 0 only. Target Link Speed. This field is used to set the target compliance mode speed when software is using the Enter Compliance bit to force a link into compliance mode. Defined encodings are: 0001b = 2.5 Gb/s Target Link Speed. 0010b = 5 Gb/s Target Link Speed. All other encodings are reserved. If a value is written to this field that does not correspond to a speed included in the Supported Link Speeds field, the result is undefined. The default value of this field is the highest link speed supported by PCH (as reported in the Supported Link Speeds field of the Link Capabilities register). Note: Offset Start: A4h Offset End: A5h Power Well: Core Bit Description Note: 3 :0 Bus:Function: M:0 Default: 0002h Note: 4 Bus:Device:Function: B:0:0 This bit field is valid for function 0 only. Intel(R) Communications Chipset 89xx Series - Datasheet 927 20.2.5.13 PLSR2--PF Link Status 2 Register Table 20-46. PLSR2--PF Link Status 2 Register Description: Link Status 2 Register View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range 15 1 :0 Offset Start: A6h Offset End: A7h Bus:Device:Function: B:0:0 Offset Start: A6h Offset End: A7h Bus:Function: M:0 Default: 0000h Bit Reset Value Bit Access Reserved 0h RV Current De-emphasis Level - When the Link is operating at 5 GT/s speed, this bit reflects the level of de-emphasis. it is undefined when the Link is operating at 2.5 GT/s speed Encodings: 1b -3.5 dB 0b -6 dB 0b RO Bit Acronym Reserved CDEL Power Well: Core Bit Description Sticky 20.2.6 MSI Capability Structure 20.2.6.1 PMSICID--PF Message Signalled Interrupt Capability ID Register The Message Signalled Interrupt Capability record defines how the device generates PCI MSI messages. It is a PCI SIG-defined capability record and includes the MCID, MCP, MCTL, MADR, and MDATA fields of the configuration header. Table 20-47. PMSICID--PF Message Signalled Interrupt Capability ID Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 8 bit Offset Start: B0h Offset End: B0h Bus:Function: M:0 Default: 05h Bit Range Bit Acronym 07 :00 MCID Power Well: Core Bit Description Capability ID: PCI SIG assigned capability record ID (05h, MSI capability) Intel(R) Communications Chipset 89xx Series - Datasheet 928 Offset Start: B0h Offset End: B0h Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 05h RO October 2012 Order Number: 327879-001US 20.0 20.2.6.2 PMSINCP--PF Message Signalled Interrupt Next Capability Pointer Register Table 20-48. PMSINCP--PF Message Signalled Interrupt Next Capability Pointer Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 8 bit Bus:Function: M:0 Offset Start: B1h Offset End: B1h Offset Start: B1h Offset End: B1h Default: 60h Bit Range Bit Acronym 07 :00 MCP 20.2.6.3 Bus:Device:Function: B:0:0 Power Well: Core Bit Description Sticky Next Capability Pointer: MSI-X capability. Bit Reset Value Bit Access 60h RO PMSICTL--PF Message Signalled Interrupt Control Register Table 20-49. PMSICTL--PF Message Signalled Interrupt Control Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0180h Bit Range Bit Acronym 15 :09 Reserved Offset Start: B2h Offset End: B3h Offset Start: B2h Offset End: B3h Power Well: Core Bit Reset Value Bit Access Reserved 0h RO Bit Description Sticky 08 MC Per-Vector Masking Capable: Per-vector masking capable. 1b RO 07 C64 64 bit Address Capable: Hardwired to 1 to indicate the device generate 64b message addresses. 1b RO 06 :04 MME Multiple Message Enable: System software writes to this field to indicate the number of allocated messages (less than or equal to the number of requested messages in MMC). A value of 0 corresponds to one message. 000h RW 03 :01 MMC Multiple Message Capable: System software reads this field to determine the number of requested messages. Hardwired to 0 to request one message. 000h RO MSIE MSI Enable: System software sets this bit to enable MSI signaling. A device driver is prohibited from writing this bit to mask a device's service request. If 1, the device can use an MSI to request service. If 0, the device cannot use an MSI to request service. 0h RW 00 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 929 20.2.6.4 PMSILADDR--PF Message Signalled Interrupt Lower Address Register Table 20-50. PMSILADDR--PF Message Signalled Interrupt Lower Address Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Offset Start: B4h Offset End: B7h Bus:Function: M:0 Default: 00000000h Power Well: Core Bit Range Bit Acronym Bit Description 31 :02 ADDR Lower Message Address: Written by the system to indicate the lower 30-bits of the address to use for the MSI memory write transaction. 01 :00 Reserved 20.2.6.5 Offset Start: B4h Offset End: B7h Bus:Device:Function: B:0:0 Sticky Reserved Bit Reset Value Bit Access 0h RW 00b RV PMSIUADDR--PF Message Signalled Interrupt Upper Address Register Table 20-51. PMSIUADDR--PF Message Signalled Interrupt Upper Address Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Offset Start: B8h Offset End: BBh Bus:Function: M:0 Default: 00000000h Power Well: Core Bit Range Bit Acronym Bit Description 31 :00 ADDR Upper Message Address: Written by the system to indicate the lower 62-bits of the address to use for the MSI memory write transaction. The lower two bits will always be written as 0. Intel(R) Communications Chipset 89xx Series - Datasheet 930 Offset Start: B8h Offset End: BBh Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 0h RW October 2012 Order Number: 327879-001US 20.0 20.2.6.6 PMSIDATA--PF Message Signalled Interrupt Data Register Table 20-52. PMSIDATA--PF Message Signalled Interrupt Data Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0000h Offset Start: BCh Offset End: BDh Power Well: Core Bit Range Bit Acronym Bit Description 15 :00 DATA Message Data: Written by the system to indicate the lower 16 bits of the data written in the MSI memory write DWORD transaction. The upper 16 bits of the transaction are written as 0. 20.2.6.7 Offset Start: BCh Offset End: BDh Sticky Bit Reset Value Bit Access 0h RW PMSIMSK--PF Message Signalled Interrupt Mask Register Table 20-53. PMSIMSK - PF Message Signalled Interrupt Mask Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 00000000h Bit Range Bit Acronym 31 :01 Reserved 00 :00 MASK0 October 2012 Order Number: 327879-001US Offset Start: C0h Offset End: C3h Offset Start: C0h Offset End: C3h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved. 0h RV Mask Bits: Only on bit defined. See the interrupt section. 0b RW Intel(R) Communications Chipset 89xx Series - Datasheet 931 20.2.6.8 PMSIPND--PF Message Signalled Interrupt Pending Register Table 20-54. PMSIPND--PF Message Signalled Interrupt Pending Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Offset Start: C4h Offset End: C7h Bus:Device:Function: B:0:0 Offset Start: C4h Offset End: C7h Bus:Function: M:0 Default: 00000000h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Bit Range Bit Acronym 31 :01 Reserved Reserved. 0h RV 00 :00 Reserved Pending Bits: Not used. 0b RO 20.2.7 PF Advanced Error Reporting Capability Structure 20.2.7.1 PPCIEAERCAPID--PF PCI Express AER Capability ID Register The PCI Express Capability List register enumerates the PCI Express AER Capability structure in the PCI 3.0 configuration space capability list. Note: Depending on the ACC_VF_ENABLE strap the AER Capability can be the last capability in the list. Table 20-55. PPCIEAERCAPID--PF PCI Express AER Capability ID Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range Offset Start: 100h Offset End: 103h Bus:Function: M:0 Default: 13810001h Power Well: Core Bit Reset Value Bit Access 138h RO Advanced Error Capability Version Number: PCI Express Advanced Error Reporting Extended Capability Version Number. 1h RO Advanced Error Capability ID: PCI Express Extended Capability ID indicating Advanced Error Reporting Capability. 1h RO Bit Acronym Bit Description 31 :20 PCIEAERNCP Next PCI Express Extended Capability Pointer: The Next Capability Pointer default value will be dependent on the ACC_VF_ENABLE strap. 1 - 138h (Default) 0 - 000h (AER is the last extended capability) 19 :16 PCIEAERCVN 15 :00 PCIEAERCID Intel(R) Communications Chipset 89xx Series - Datasheet 932 Offset Start: 100h Offset End: 103h Bus:Device:Function: B:0:0 Sticky October 2012 Order Number: 327879-001US 20.0 20.2.7.2 PPAERUCS--PF PCI Express AER Uncorrectable Error Status Register Table 20-56. PPAERUCS--PF PCI Express AER Uncorrectable Error Status Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0h Bit Range Bit Acronym 31 :21 Reserved Offset Start: 104h Offset End: 107h Offset Start: 104h Offset End: 107h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00b RV Unsupported Request Error Status: As a receiver, Set whenever an unsupported request is detected. The Header is logged. Y 0b RW1CS ECRC Check: As a receiver, set when ECRC check fails. The Header is logged. Not supported in EP Y 0b RW1CS Malformed TLP: As a receiver, set whenever a malformed TLP is detected. The Header is logged. Y 0b RW1CS 20 UR 19 ECRCC 18 MTLP 17 RO Receiver Overflow: Set if PCI Express receive buffers overflow. Y 0b RW1CS 16 EC Unexpected Completion: As a receiver, set whenever a completion is received that does not match the EP requestor ID or outstanding Tag. The Header is logged. Y 0b RW1CS 15 CA Completer Abort: As a completer, set whenever an internal agent signals a data abort. The header is logged. Y 0b RW1CS 14 CT Completion Timeout: As a requester, set whenever an outbound Non Posted Request does not receive a completion within 16-32ms. Y 0b RW1CS 13 FCPES Flow Control Protocol Error Status: Set whenever a flow control protocol error is detected. Not supported. Y 0b RW1CS PTLPR Poisoned TLP Received: As a receiver, set whenever a poisoned TLP is received from PCI Express. The header is logged. Internal queue errors are not covered by this bit, they are logged by the Configuration target of the transaction. Y 0h RW1CS 0000b RV 12 11 :6 Reserved Reserved 5 SDES Surprise Down Error: Not supported. Y 0b RW1CS 4 DLPE Data Link Protocol Error: Set whenever a data link protocol error is detected. Y 0b RW1CS 03 :00 Reserved 0h RV October 2012 Order Number: 327879-001US Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 933 20.2.7.3 PPAERUCM--PF PCI Express AER Uncorrectable Error Mask Register Table 20-57. PPAERUCM--PF PCI Express AER Uncorrectable Error Mask Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Offset Start: 108h Offset End: 10Bh Bus:Function: M:0 Default: 0h Bit Range Bit Acronym 31 :21 Reserved Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 0h RV 20 UR Unsupported Request Error Mask: When `1' error reporting is masked. 19 ECRCC ECRC Check Error Mask: When `1' error reporting is masked. Y 0b RWS 18 MTLP Malformed TLP Error Mask: When `1' error reporting is masked. Y 0b RWS 17 RO Receiver Overflow Error Mask: When `1' error reporting is masked. Y 0b RWS 16 EC Unexpected Completion Error Mask: When `1' error reporting is masked. Y 0b RWS 15 CA Completer Abort Error Mask: When `1' error reporting is masked. Y 0b RWS 14 CT Completion Time Out Error Mask: When `1' error reporting is masked. Y 0b RWS 13 FCPES Flow Control Protocol Error Mask: When `1' error reporting is masked. Not supported. Y 0b RWS 12 PTLPR Poisoned TLP Received Error Mask: When `1' error reporting is masked. Y 0h RWS 0h RV 11 :6 Reserved Y 0b RWS Reserved 5 SDES Surprise Down Error: Not supported. 4 DLPE Data Link Protocol Error Mask: When `1' error reporting is masked. 03 :01 Reserved Reserved 0h RV Reserved Undefined: Reserved 0b RV 00 Intel(R) Communications Chipset 89xx Series - Datasheet 934 Offset Start: 108h Offset End: 10Bh Bus:Device:Function: B:0:0 Y 0b RWS Y 0b RWS October 2012 Order Number: 327879-001US 20.0 20.2.7.4 PPAERUCSEV--PF PCI Express AER Uncorrectable Error Severity Register The Uncorrectable Error Severity register controls whether an individual uncorrectable error is reported as a non-fatal or fatal error. An error is reported as fatal when the corresponding error bit in the severity register is set. If the bit is cleared, the corresponding error is considered non-fatal. Table 20-58. PPAERUCSEV--PF PCI Express AER Uncorrectable Error Severity Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 00062030h Bit Range Bit Acronym 31 :21 Reserved Offset Start: 10Ch Offset End: 10Fh Offset Start: 10Ch Offset End: 10Fh Power Well: Core Bit Description Sticky Bit Reset Value 0h RV Unsupported Request Error Status Severity(URESS): Y 0b RWS ECRC Check Severity: Not supported in EP Y 0b RWS Reserved Bit Access 20 UR 19 ECRCC 18 MTLP Malformed TLP Severity: Y 1b RWS 17 RO Receiver Overflow Severity: Y 1b RWS 16 EC Unexpected Completion Severity: Y 0b RWS 15 CA Completer Abort Severity: Y 0b RWS 14 CT Completion Time Out Severity: Y 0b RWS FCPES Flow Control Protocol Error Severity: Not supported. Y 1b RWS PTLPR Poisoned TLP Received Severity: Y 0h RWS 0h RV 13 12 11 :6 5 Reserved SDES Reserved Surprise Down Error Severity: Not supported. Y 1b RWS Data Link Protocol Error Severity: Y 4 DLPE 1b RWS 03 :01 Reserved Reserved 0h RV Reserved Undefined: Reserved 0b RV 00 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 935 20.2.7.5 PPAERCS--PF PCI Express AER Correctable Error Register Table 20-59. PPAERCS--PF PCI Express AER Correctable Error Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Acronym 31 :14 Reserved 13 ANFES 12 RTTS 11 :09 Reserved 8 RNRS 7 BDLLPS 6 DLPE 05 :01 Reserved RES Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 0h RV Advisory Non-Fatal Error Status Y 0b RW1CS Replay Timer Timeout Status: Set whenever a replay timer timeout occurs. Y 0b RW1CS 0b RV Reserved REPLAY NUM Rollover Status: Set whenever the replay number rolls over from 11 to 00. Y 0h RW1CS Bad DLLP Status: Sets this bit on CRC errors on DLLP. Y 0b RW1CS Bad TLP Status: Sets this bit on CRC errors or sequence number out of range on TLP. Y 0b RW1CS 0h RV 0b RW1CS Reserved Receiver Error Status: Set whenever the physical layer detects a receiver error. Intel(R) Communications Chipset 89xx Series - Datasheet 936 Offset Start: 110h Offset End: 113h Bus:Function: M:0 Default: 00h Bit Range 00 Offset Start: 110h Offset End: 113h Bus:Device:Function: B:0:0 Y October 2012 Order Number: 327879-001US 20.0 20.2.7.6 PPAERCM--PF PCI Express AER Correctable Error Mask Register Table 20-60. PPAERCM--PF PCI Express AER Correctable Error Mask Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Function: M:0 Default: 2000h Bit Range Bit Acronym 31 :14 Reserved 13 ANFES 12 RTTS 11 :09 Reserved 8 RNRS 7 BDLLPS 6 DLPE 05 :01 Reserved 00 Bus:Device:Function: B:0:0 RES October 2012 Order Number: 327879-001US Offset Start: 114h Offset End: 117h Offset Start: 114h Offset End: 117h Power Well: Core Bit Reset Value Bit Access Reserved 0h RV Advisory Non-Fatal Error Mask: this bit is set by default to enable compatibility with software that does not comprehend Role-Based Error Reporting. Y 1b Replay Timer Timeout Mask: Y Bit Description Sticky Reserved 0b RWS RWS 0h RV REPLAY NUM Rollover Mask: Y 0b RWS Bad DLLP Mask: Y 0b RWS Bad TLP Mask: Y 0b RWS 0h RV Y 0b RWS Reserved Receiver Error Mask: Intel(R) Communications Chipset 89xx Series - Datasheet 937 20.2.7.7 PPAERCTLCAP--PF PCI Express AER Control and Capability Register Table 20-61. PPAERCTLCAP--PF PCI Express AER Control and Capability Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Offset Start: 118h Offset End: 11Bh Bus:Device:Function: B:0:0 Offset Start: 118h Offset End: 11Bh Bus:Function: M:0 Default: 0h Power Well: Core Bit Reset Value Bit Access Reserved 0h RV ECRCCE ECRC Check Enable: When set enables ECRC checking. ECRC not supported 0b 7 ECRCCC ECRC Check Capable: Indicates EP is capable of checking ECRC. 6 ECRCGE ECRC Generation Enable: When set enables ECRC generation. ECRC is not supported 5 ECRCGC ECRC Generation Capable: Indicates the EP is not capable of generating ECRC. Bit Range Bit Acronym 31 :09 Reserved 8 04 :00 20.2.7.8 Bit Description Sticky The First Error Pointer: Identifies the bit position of the first error reported in the Section 20-56, "PPAERUCS--PF PCI Express AER Uncorrectable Error Status Register" This register will not update until all bits in the ERRUNC STS register are cleared. TFEP Y RWS 0b 0b Y Y RO RWS 0b RO 00000b ROS PPAERHDRLOG0--PF PCI Express AER Header Log 0 Register Table 20-62. PPAERHDRLOG0--PF PCI Express AER Header Log 0 Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range 31 :00 Offset Start: 11Ch Offset End: 11Fh Bus:Function: M:0 Default: 0h Bit Acronym Power Well: Core Bit Description 1st DWord of the Header for the PCI Express packet in error (HDRLOGDW0): HDRLOGDW0 Once an error is logged in this register, it remains locked for further error logging until the time the software clears the status bit that cause the header log i.e. the error pointer is rearmed to log again. Intel(R) Communications Chipset 89xx Series - Datasheet 938 Offset Start: 11Ch Offset End: 11Fh Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 0h RO October 2012 Order Number: 327879-001US 20.0 20.2.7.9 PPAERHDRLOG1--PF PCI Express AER Header Log 1 Register Table 20-63. PPAERHDRLOG1--PF PCI Express AER Header Log 1 Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range 31 :00 20.2.7.10 Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0h Bit Acronym Offset Start: 120h Offset End: 123h Offset Start: 120h Offset End: 123h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0h RO Second DWord of the Header for the PCI Express packet in error (HDRLOGDW1): HDRLOGDW1 Once an error is logged in this register, it remains locked for further error logging until the time the software clears the status bit that cause the header log i.e. the error pointer is rearmed to log again. PPAERHDRLOG2--PF PCI Express AER Header Log 2 Register Table 20-64. PPAERHDRLOG2--PF PCI Express AER Header Log 2 Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range 31 :00 Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0h Bit Acronym Offset Start: 124h Offset End: 127h Power Well: Core Bit Description Sticky Third DWord of the Header for the PCI Express packet in error (HDRLOGDW2): HDRLOGDW2 Once an error is logged in this register, it remains locked for further error logging until the time the software clears the status bit that cause the header log i.e. the error pointer is rearmed to log again. October 2012 Order Number: 327879-001US Offset Start: 124h Offset End: 127h Bit Reset Value Bit Access 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 939 20.2.7.11 PPAERHDRLOG3--PF PCI Express AER Header Log 3 Register Table 20-65. PPAERHDRLOG3--PF PCI Express AER Header Log 3 Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range 31 :00 20.2.8 Offset Start: 128h Offset End: 12Bh Bus:Device:Function: B:0:0 Offset Start: 128h Offset End: 12Bh Bus:Function: M:0 Default: 0h Bit Acronym Power Well: Core Bit Description Sticky 4th DWord of the Header for the PCI Express packet in error (HDRDWLOG3): HDRDWLOG3 Once an error is logged in this register, it remains locked for further error logging until the time the software clears the status bit that cause the header log i.e. the error pointer is rearmed to log again. Bit Reset Value Bit Access 0h RO PF Alternative Routing-ID Extended Capability Structure This section describes the PCI Express Extended Configuration Space registers that make up the Alternative Routing ID Extended Capability Structure. Some information from the specification is repeated here as an aid to the reader or to describe implementation choice. See the PCI Express* Base Specification 2.0 for the full register descriptions and additional information regarding their operation. 20.2.8.1 PARIDHDR--PF Alternative Routing ID Capability Header This register contains information associated with the Alternative Routing ID capability. This is compliant with the PCI-SIG ECN: Alternatifve Routing-ID Interpretation (ARI), Updated June 4, 2007. Intel(R) Communications Chipset 89xx Series - Datasheet 940 October 2012 Order Number: 327879-001US 20.0 20.2.8.2 PFARICAP--PF ARI Capabilities Register Table 20-66. PARIDHDR--PF Alternative Routing ID Capability Header Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 1401000Eh Offset Start: 138h Offset End: 13Bh Offset Start: 138h Offset End: 13Bh Power Well: Core Bit Range Bit Acronym Bit Description Sticky 31 :20 ARINCO Next Capability Offset - This field contains 140h which points to the next item in the extended capabilities list, the SR-IOV extended capability. 19 :16 ARICV Capability Version - This is set to 1h for the most current version of the specification. 15 :0 ARICV PCI Express Extended Capability ID - The PCI SIG has assigned 000Eh to the ARI extended capability. Bit Reset Value Bit Access 140h RO 1h RO 000EH RO This register contains information associated with the Alternative Routing ID capability. Table 20-67. PFARICAP--PF ARI Capabilities Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range 15 :8 7 :2 Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 0100h Reserved Offset Start: 13Ch Offset End: 13Dh Power Well: Core Bit Reset Value Bit Access 1h RO Reserved. 0h RV Bit Acronym VNFN Offset Start: 13Ch Offset End: 13Dh Bit Description Sticky Next Function Number: The function number of the next highest numbered PF in a multi-function device. The Next Function Number is set based on fuse and EEPROM bits. However, some of the values listed below will never be observed as some fuse and EEPROM bits' combinations are not valid scenarios. 0h - If only Function 0 is enabled. 1h - If Function 1 is enabled. 2h - If Function 1 is disabled, but Function 2 enabled. 3h - If Functions 1-2 are disabled, but Function 3 is enabled. 4h - If Function 1-3 are disabled, but Function 4 is enabled. 1 ACS ACS Functional Groups Capability: Hardwired to Zero . 0b RO 0 MFVC MFVC Functional Groups Capability: Hardwired to Zero 0b RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 941 20.2.8.3 PARIDCTL--PF Alternative Routing ID Control Register This register contains information associated with the Alternative Routing ID capability. Table 20-68. PARIDCTL--PF Alternative Routing ID Control Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Offset Start: 13Eh Offset End: 13Fh Bus:Device:Function: B:0:0 Offset Start: 13Eh Offset End: 13Fh Bus:Function: M:0 Default: 00000000h Bit Range Bit Acronym 15 :7 Reserved 6 :4 FG 3 :2 Reserved 1 ACS 0 MFVC Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved 0h RV Function Group: Hardwired to Zero. 0b RO Reserved 0b RV ACS Functional Groups Enable: Hardwired to Zero. 0b RO MFVC Functional Groups Enable: Hardwired to Zero. 0b RO 20.2.9 PF SR-IOV Extended Capability Structure 20.2.9.1 PSRIOVCAPID--PF SR-IOV Capability ID Register Table 20-69. PSRIOVCAPID--PF SR IOV Capability ID Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Offset Start: 140h Offset End: 141h Bus:Function: M:0 Default: 10h Power Well: Core Bit Range Bit Acronym Bit Description 15 :00 SRIOVOD Capability ID: PCI SIG assigned a capability record ID for SR-IOV per the 1.0 revision Intel(R) Communications Chipset 89xx Series - Datasheet 942 Offset Start: 140h Offset End: 141h Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 0010h RO October 2012 Order Number: 327879-001US 20.0 20.2.9.2 PSRIOVCVNC--PF SRIOV Capability Version and Next Capability Pointer Register Table 20-70. PSRIOVCVNC--PF SRIOV Capability Version and Next Capability Pointer Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bus:Function: M:0 Default: 1h Bit Range Bit Acronym 15 :4 SRIOVNCO 3 :0 Bus:Device:Function: B:0:0 SRIOVCV October 2012 Order Number: 327879-001US Offset Start: 142h Offset End: 143h Offset Start: 142h Offset End: 143h Power Well: Core Bit Description Sticky Next Capability Offset - last extended capability. Capability Version - This is set to 1h for the Single Root I/O Virtualization and Sharing Specification, Revision 0.9. Bit Reset Value Bit Access 00h RO 1h RO Intel(R) Communications Chipset 89xx Series - Datasheet 943 20.2.9.3 PSRIOVCAP--PF SRIOV Capabilities Register 20.2.9.4 PSRIOVCS--PF SRIOV Control and Status Register Table 20-71. PSRIOVCAP--PF SRIOV Capabilities Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Default: 00000000h Power Well: Core Bit Acronym Bit Description 31 :21 VFMINT VF Migration Interrupt Message Number - Indicates the MSI vector used for migration interrupts. The value in this field has no significance if bit 0 of this capability is Clear. Note: :0 Reserved VFMCAP Sticky Bit Reset Value Bit Access 0h RO 0h RO 0b RO N/A for EP Reserved. VF Migration Capable - If Set, the PF is Migration Capable and operating under a Migration Capable MR-PCIM. Note: 20.2.9.5 Offset Start: 144h Offset End: 147h Bus:Function: M:0 Bit Range 20 :1 Offset Start: 144h Offset End: 147h Bus:Device:Function: B:0:0 EP does not support VF migration PSRIOVMTOTINI--PF SRIOV Initial and Total VFs Register Table 20-72. PSRIOVCS--PF SRIOV Control and Status Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Offset Start: 148h Offset End: 14Bh Bus:Function: M:0 Default: 00000000h Bit Range Bit Acronym 31 :5 Reserved Power Well: Core Bit Reset Value Bit Access Reserved 0h RO Bit Description Sticky 4 VFARI ARI Capable Hierarchy - Device may locate VFs in Function numbers 8 to 255 of the captured Bus number. EP supports ARI and will locate the VF1-VF16 in Function 8 through Function 23. 0b RW 3 VFMSE VF MSE - Memory Space Enable for Virtual Functions. 0b RW 2 VFMIE 0b RO 0b RO 0b RW VF Migration Interrupt Enable - Enables / Disables VF Migration State Change Interrupt. Note: 1 VFME 0 VFE EP does not support VF Migration VF Migration Enable - Enables / Disables VF Migration Support. Note: EP does not support VF Migration VF Enable - Enables / Disables VFs. Default value is 0b Intel(R) Communications Chipset 89xx Series - Datasheet 944 Offset Start: 148h Offset End: 14Bh Bus:Device:Function: B:0:0 October 2012 Order Number: 327879-001US 20.0 20.2.9.6 PSRIOVNUMVF--PF SRIOV Number of VFs Register Table 20-73. PSRIOVMTOTINI--PF SRIOV Initial and Total VFs Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 00100010h Offset Start: 14Ch Offset End: 14Fh Power Well: Core Bit Reset Value Bit Access TotalVFs indicates the maximum number of VFs that could be associated with the PF. Since EP does not support VF migration this field must be equal to the INITVF field 10h RO InitialVFs indicates the number of VFs that are initially associated with the PF. Since EP does not support VF migration this field must be equal to the TOTVF field 10h RO Bit Range Bit Acronym Bit Description 31 :16 TOTVF 15 :0 INITVF 20.2.9.7 Offset Start: 14Ch Offset End: 14Fh Sticky PSRIOVFVFO--PF SRIOV First VF Offset Register Table 20-74. PSRIOVNUMVF--PF SRIOV Number of VFs Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Acronym 31 :24 Reserved FUNDEP NUMVF October 2012 Order Number: 327879-001US Offset Start: 150h Offset End: 153h Offset Start: 150h Offset End: 153h Power Well: Core Bit Reset Value Bit Access Reserved 00h RO This field describes dependencies between PFs. VF dependencies are the same as the dependencies of their associated PFs. If a PF is independent from other PFs of a Device, this field shall contain the Function Number of the PF. 00h RO 0 RW Bit Description Note: 15 :0 Bus:Function: M:0 Default: 00000000h Bit Range 23 :16 Bus:Device:Function: B:0:0 Sticky EP implements a single PF and therefore this field is N/A to EP. Num VFs defines the number of VFs software has assigned to the PF. Software sets NumVFs to any value between 1 and TotalVFs (16 for EP) as part of the process of creating VFs. NumVFs VFs shall be visible in the PCI-Express fabric after both NumVFs is set to a valid value and VF Enable is set to one. "Visible in the PCI-Express fabric" means that the VF shall respond to PCI Express transactions targeting the VF, following all other rules defined by this specification and the base specification. Intel(R) Communications Chipset 89xx Series - Datasheet 945 20.2.9.8 PSRIOVVFS--PF SRIOV VF Stride Register Table 20-75. PSRIOVFVFO--PF SRIOV First VF Offset Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range 15 :0 Offset Start: 154h Offset End: 155h Bus:Function: M:0 Default: 0008h Power Well: Core Bit Acronym Bit Description VFOFFSET First VF Offset is a constant and defines the Routing ID (RID) offset of the first VF that is associated with the PF that contains this Capability structure. The first VFs 16-bit RID is calculated by adding the contents of this field to the RID of the PF. Note: 20.2.9.9 Offset Start: 154h Offset End: 155h Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 0008h RO For the EP, the RID of the first VF is at an offset of eight from the RID of the Physical Function (PF). PSRIOVFDID--PF SRIOV VF Device ID Register Table 20-76. PSRIOVVFS--PF SRIOV VF Stride Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 16 bit Bit Range 15 :0 Offset Start: 156h Offset End: 157h Bus:Function: M:0 Default: 0001h Bit Acronym VFSTRIDE Power Well: Core Bit Description VF Stride is a constant and defines the Routing ID (RID) offset from one VF to the next one for all VFs associated with the PF that contains this Capability structure. The next VFs 16-bit RID is calculated by adding the contents of this field to the RID of the current VF. Intel(R) Communications Chipset 89xx Series - Datasheet 946 Offset Start: 156h Offset End: 157h Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 1h RO October 2012 Order Number: 327879-001US 20.0 20.2.9.10 PSRIOVPAGESIZE--PF SRIOV Supported Page Size Register Table 20-77. PSRIOVFDID--PF SRIOV VF Device ID Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 04420000h Offset Start: 158h Offset End: 15Bh Power Well: Core Bit Range Bit Acronym 31 16 VFDID151 VF Device ID - The Device ID that is presented to the OS for every VF. PCH VF DID = 0442h. 15 :00 Reserved Reserved 20.2.9.11 Offset Start: 158h Offset End: 15Bh Bit Description Sticky Bit Reset Value Bit Access 0442h RWOS 0h RV PSRIOVSYSPS--PF SRIOV System Page Size Register Table 20-78. PSRIOVPAGESIZE--PF SRIOV Supported Page Size Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range 31 :00 Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 00000553h Offset Start: 15Ch Offset End: 15Fh Power Well: Core Bit Acronym Bit Description FIRSTVFO This field must define the supported page sizes. This PF supports a page size of 2^(n+12) if bit n is Set. For example, if bit 0 is Set, the PF supports 4k byte page sizes. PFs are required to support 4k, 8k, 64k, 256k, 1M and 4M page sizes. All other page sizes are optional. A page size describes the minimum alignment requirements for VF BAR resources as described in "System Page Size". October 2012 Order Number: 327879-001US Offset Start: 15Ch Offset End: 15Fh Sticky Bit Reset Value Bit Access 00000553h RO Intel(R) Communications Chipset 89xx Series - Datasheet 947 20.2.9.12 PSRIOVLBAR0--SRIOV Lower BAR0 Register Table 20-79. PSRIOVSYSPS--PF SRIOV System Page Size Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range 31 :00 Offset Start: 160h Offset End: 163h Bus:Device:Function: B:0:0 Offset Start: 160h Offset End: 163h Bus:Function: M:0 Default: 00000001h Power Well: Core Bit Acronym Bit Description SYSPAGSIZ This field must define the page size the system will use to map the VFs' memory addresses. Software must set the value of the System Page Size to one of the page sizes set in the Supported Page Sizes field. As with Supported Page Sizes, if bit n is Set in System Page Size, the VFs associated with this PF are required to support a page size of 2^(n+12). For example, if bit 1 is Set, the system is using an 8k byte page size. The results are undefined if more than one bit is set in System Page Size. The results are undefined if a bit is Set in System Page Size that is not Set in Supported Page Sizes. When System Page Size is set, the VF associated with this PF is required to align all BAR resources on a System Page Size boundary. Each VF BARn or VF BARn pair (described below) shall be aligned on a System Page Size boundary. Each VF BARn or VF BARn pair defining a non-zero address space shall be sized to consume an integer multiple of System Page Size bytes. All data structures requiring page size alignment within a VF shall be aligned on a System Page Size boundary. Sticky Bit Reset Value Bit Access 00000001h RW The PSRIOVLBAR0 points to the Ring Controller rings. The size of this BAR is dependent on the System Page Size. However, the programmed value is dependent on the System Page Size and the Number of VFs supported. Each VF is mapped to one bundle (16 rings). Intel(R) Communications Chipset 89xx Series - Datasheet 948 October 2012 Order Number: 327879-001US 20.0 20.2.9.13 PSRIOVUBAR0--SRIOV Upper BAR0 Register Table 20-80. PSRIOVLBAR0--SRIOV Lower BAR0 Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 00000004h Bit Acronym Offset Start: 164h Offset End: 167h Offset Start: 164h Offset End: 167h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0 RW Lower SRIOV Base Address 0: These bits are used to define the actual locations of the Ring bundles when addressed from the PCI bus. PCI Local Bus Specification, Revision 3.0 compliant scanning of SRIOV Base Address 0, at a minimum reveals a 4 KB memory window 0 for PSRIOVBAR0. However, this memory window size will be dictated by the System Page Size as described in Section 20.2.9.11, "PSRIOVSYSPS--PF SRIOV System Page Size Register" 31 :12 VFBAR0 System Page Size 4-KB 8-KB 64-KB 256-KB 1-MB 4-MB 11 :04 Bits Above bit 12 that respond as Reserved Memory None 12:12 15:12 17:12 19:12 21:12 Window Size 64-KB 128-KB 1024-KB 4096-KB 16-MB 64-MB VFZERO Lower Bits: Hardwired to 0 (8KB region). 0h RO 3 VFPREF Prefetchable: Hardwired to 0 to indicate that the region is not prefetchable. 0h RO 02 :01 VFTYP Addressing Type: Hardwired to indicate a 64-bit region. 10b RO VFMEM Memory Space Indicator: Hardwired to 0 to identify the region as in memory space. 0h RO 00 The PSRIOVUBAR0 points to the Ring Controller rings. The size of this BAR is dependent on the System Page Size. However, the programmed value is dependent on the System Page Size and the Number of VFs supported. Each VF is mapped to one bundle (16 rings). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 949 20.2.9.14 PSRIOVLBAR1--SRIOV Lower BAR1 Register Table 20-81. PSRIOVUBAR0--SRIOV Upper BAR0 Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range Offset Start: 168h Offset End: 16Bh Bus:Device:Function: B:0:0 Offset Start: 168h Offset End: 16Bh Bus:Function: M:0 Default: 00000000h Bit Acronym Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0h RW Upper SRIOV Base Address 0: These bits are used to define the actual locations of the Ring bundles when addressed from the PCI bus. PCI Local Bus Specification, Revision 3.0 compliant scanning of SRIOV Base Address 0, at a minimum reveals a 4 KB memory window 0 for PSRIOVBAR0. However, this memory window size will be dictated by the System Page Size as described in Section 20.2.9.11, "PSRIOVSYSPS--PF SRIOV System Page Size Register" 31 :00 VFBAR0 System Page Size 4-KB 8-KB 64-KB 256-KB 1-MB 4-MB Bits Above bit 12 that respond as Reserved Memory None 12:12 15:12 17:12 19:12 21:12 Intel(R) Communications Chipset 89xx Series - Datasheet 950 Window Size 4-KB 8-KB 64-KB 256-KB 1-MB 4-MB October 2012 Order Number: 327879-001US 20.0 20.2.9.15 PSRIOVUBAR1--SRIOV Upper BAR1 Register Table 20-82. PSRIOVLBAR1--SRIOV Lower BAR1 Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range Bus:Device:Function: B:0:0 Bus:Function: M:0 Default: 00000004h Bit Acronym Offset Start: 16Ch Offset End: 16Fh Offset Start: 16Ch Offset End: 16Fh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0h RW Lower SRIOV Base Address 1: These bits are used to define the actual locations of the PF2VF[1-16] - PF to VF[1:16] Doorbell and Interrupt Register when addressed from the PCI bus. PCI Local Bus Specification, Revision 3.0 compliant scanning of SRIOV Base Address 1, at a minimum reveals a 4 KB memory window 1 for PSRIOVBAR1. However, this memory window size will be dictated by the System Page Size as described in Section 20.2.9.11, "PSRIOVSYSPS--PF SRIOV System Page Size Register" 31 :12 VFBAR1 System Page Size 4-KB 8-KB 64-KB 256-KB 1-MB 4-MB 11 :04 Bits Above bit 12 that respond as Reserved Memory None 12:12 15:12 17:12 19:12 21:12 Window Size 4-KB 8-KB 64-KB 256-KB 1-MB 4-MB VFZERO Lower Bits: Hardwired to 0 (4KB region) 0h RO 3 VFPREF Prefetchable: Hardwired to 0 to indicate that the region is not prefetchable. 0h RO 02 :01 VFTYP Addressing Type: Hardwired to indicate a 64-bit region. 10b RO VFMEM Memory Space Indicator: Hardwired to 0 to identify the region as in memory space. 0h RO 00 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 951 20.2.9.16 PSRIOVVFMA--PF SRIOV VF Migration Array Register Table 20-83. PSRIOVUBAR1--SRIOV Upper BAR1 Register Description: View: PCI BAR: Configuration View: PCI PF BAR: Configuration Size: 32 bit Bit Range Offset Start: 170h Offset End: 173h Bus:Device:Function: B:0:0 Offset Start: 170h Offset End: 173h Bus:Function: M:0 Default: 00000000h Bit Acronym Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0h RW Upper SRIOV Base Address 1: These bits are used to define the actual locations of the PF2VF[1-16] - PF to VF[1:16] Doorbell and Interrupt Register when addressed from the PCI bus. PCI Local Bus Specification, Revision 3.0 compliant scanning of SRIOV Base Address 1, at a minimum reveals a 4 KB memory window 1 for PSRIOVBAR1. However, this memory window size will be dictated by the System Page Size as described in Section 20.2.9.11, "PSRIOVSYSPS--PF SRIOV System Page Size Register" 31 :00 VFBAR1 System Page Size 4-KB 8-KB 64-KB 256-KB 1-MB 4-MB Bits Above bit 12 that respond as Reserved Memory None 12:12 15:12 17:12 19:12 21:12 Window Size 4-KB 8-KB 64-KB 256-KB 1-MB 4-MB Table 20-84. PSRIOVVFMA--PF SRIOV VF Migration Array Register Description: View: PCI Size: 32 bit BAR: Configuration Default: 00000000h Bit Range Bit Acronym 31 :00 VFMA Power Well: Core Bit Description N/A Note: Not supported in EP Intel(R) Communications Chipset 89xx Series - Datasheet 952 Offset Start: 17Ch Offset End: 17Fh Bus:Device:Function: B:0:0 Sticky Bit Reset Value Bit Access 00000000h RO October 2012 Order Number: 327879-001US 20.0 20.3 EP VF PCI Configuration Space Highlighted regions of the configurations space are RO for the VF[6:0] configuration space and have the same value as the PF. 20.3.1 PCI Standard Header Registers This section describes the PCI Configuration Space registers that make up the standard Type 0 header. Some information from the specification is repeated here as an aid to the reader or to describe implementation choice. See the PCI Express* Base Specification 2.0 and PCI Local Bus Specification for the full register descriptions and additional information regarding their operation. 20.3.1.1 VVID[0:15]--VF Vendor Identification Register The VID Register contains the vendor identification number. This 16-bit register combined with the Device Identification Register uniquely identifies any PCI device. Writes to this register have no effect. 20.3.1.2 VDID[0:15]--VF Device Identification Register Table 20-85. VVID[0:15]--VF Vendor Identification Register Description: View: PCI VF Size: 16 bit Bus:Function: M:8 + Index1 BAR: Configuration Default: FFFFh Offset Start: 00h Offset End: 01h Power Well: Core Bit Range Bit Acronym Bit Description Sticky 15 :00 VID Vendor ID: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field return FFFFH. Bit Reset Value Bit Access FFFFh RO This 16-bit register combined with the Vendor Identification register uniquely identifies any PCI device. Writes to this register have no effect. Table 20-86. VDID[0:15]--VF Device Identification Register Description: View: PCI VF Size: 16 bit Bit Range 15 :00 + Bus:Function: M:8 Index1 BAR: Configuration Default: FFFFh Bit Acronym DID October 2012 Order Number: 327879-001US Offset Start: 02h Offset End: 03h Power Well: Core Bit Description Sticky Device ID: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field return FFFFH. Software should return the VF Device ID value from the associated PF as the Device ID for the VF. Bit Reset Value Bit Access FFFFh RO Intel(R) Communications Chipset 89xx Series - Datasheet 953 20.3.1.3 VPCICMD[0:15]--VF Device Command Register Table 20-87. VPCICMD[0:15]--VF Device Command Register Description: View: PCI VF Size: 16 bit BAR: Configuration Bus:Function: Default: 0000h Bit Range Bit Acronym 15 :11 Reserved Power Well: Core Bit Reset Value Bit Access Reserved 0h RV Bit Description Sticky 10 INTD Interrupt Disable: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to 0b for all VFs. This bit does not apply to VFs. 0h RO 9 FBTB Fast Back to Back Enable: Does not apply to PCI Express. Hard-wired to 0 0h RO 8 SER SERR# Enable: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to 0b for all VFs. In addition the functionality associated with the setting of this bit in the Section 20.2.2.3, "PPCICMD--PF Device Command Register" will apply to all VFs. 0h RO 7 Reserved Address/Data Stepping Control: Does not apply to PCI Express. Hard-wired to 0. 0h RO 0h RO 6 PER Parity Error Response: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to 0b for all VFs. In addition the functionality associated with the setting of this bit in the Section 20.2.2.3, "PPCICMD--PF Device Command Register" will apply to all VFs. 5 VPS VGA Palette Snoop Enable: Does not apply to PCI Express. Hard-wired to 0. 0h RO 4 MWE Memory Write and Invalidate Enable: Does not apply to PCI Express. Hard-wired to 0. 0h RO 3 SS Special Cycle Enable: Does not apply to PCI Express. Hard-wired to 0. 0h RO 2 BM Bus Master Enable: When cleared, the EP is prevented from issuing any memory or I/O read/write requests. Requests other than memory or I/O requests are not controlled by this bit. The EP will initiate a completion transaction regardless of the setting. 0h RW 1 MEM Memory Enable: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to 0b for all VFs. 0h RO 0 IO I/O Space Enable: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to 0b for all VFs. 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 954 Offset Start: 04h Offset End: 05h M:8 + Index1 October 2012 Order Number: 327879-001US 20.0 20.3.1.4 VPCISTS[0:15]--VF Device Status Register Table 20-88. VPCISTS[0:15]--VF PCI Device Status Register Description: View: PCI VF Size: 16 bit Bit Range BAR: Configuration Bus:Function: M:8 + Index1 Default: 0010h Offset Start: 06h Offset End: 07h Power Well: Core Bit Acronym Bit Description Sticky Bit Reset Value Bit Access 15 DPE Detected Parity Error: set when the EP receives a poisoned TLP regardless of the state of the Parity Error Response in the VPCICMD register. 0h RW1C 14 SSE SERR# Asserted: set when the EP sends an ERR FATAL or ERR NONFATAL message, and the SERR Enable bit in the VPCICMD register is `1'. 0h RW1C 13 RMA Received Master Abort: set when the EP receives a completion with Unsupported Request Completion Status. 0h RW1C 12 RTA Received Target Abort: set when the EP receives a completion with Completer Abort Completion Status. 0h RW1C 11 STA Signaled Target Abort: set when the EP completes a Request using Completer Abort Completion Status 0h RW1C DST DEVSEL# Timing: Does not apply to PCI Express. Hard-wired to 0. 00b RO 0h RW1C 0h RO 0h RV 0h RO 1h RO 0h RO 0h RV 10 :09 8 MDPE 7 FB2B Master Data Parity Error: This bit is set by the EP if its Parity Error Enable bit is set and either of the following two conditions occurs: This bit is set under the following conditions. - EP receives a Poisoned Completion for an Outbound Read Request - EP transmits a Poisoned TLP for an Outbound Write Request. If the Parity Error Response bit is cleared in theSection 20.2.2.3, "PPCICMD--PF Device Command Register", this bit is never set. Fast Back-to-Back: Note: 6 Reserved 5 MC66 Does not apply to PCI Express.Hard-wired to 0. Reserved 66 MHz Capable (C66): Note: 4 CL Capabilities List: All PCI Express devices are required to implement the PCI Express capability structure. Note: 3 IS 02 : 00 Reserved Hard-wired to 1. Interrupt Status: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to 0b for all VFs. Note: October 2012 Order Number: 327879-001US Does not apply to PCI Express. Hard-wired to 0 This bit does not apply to VFs. Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 955 20.3.1.4.1 VRID[0:15]--Revision ID Register The value of this register indicates the chip stepping. It is hardwired on chip and reflects the latest revision. Table 20-89. VRID[0:15]--VF Revision ID Register Description: View: PCI VF Size: 8 bit Bit Range 07 :0 20.3.1.5 BAR: Configuration Bus:Function: Offset Start: 08h Offset End: 08h M:8 + Index1 Default: 10h Power Well: Core Bit Acronym Bit Description Sticky RID EP Revision: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 states that this field should be viewed as a Vendor Defined Extension to the Device ID. This may be different than the PF's Revision ID, but the same value must be reported by all VFs. See the PF RID for a description. Bit Reset Value Bit Access 10h RO VCC[0:15]--VF Class Code Register Table 20-90. VCC[0:15]--VF Class Code Register Description: View: PCI VF Size: 24 bit Default: 0B4000h Bit Range Bit Acronym 23 :0 CC 20.3.1.6 Offset Start: 09h Offset End: 0Bh M:8 + Bus:Function: Index1 BAR: Configuration Power Well: Core Bit Description Sticky Class Code: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field return the same value as the Section 20.2.2.5, "PCC--PF Class Code Register" for all VFs. Bit Reset Value Bit Access 0B4000h RWOS VHDR[0:15]--VF Header Type Register Table 20-91. VHDR[0:15]--VF Header Type Register Description: View: PCI VF Size: 8 bit Default: 00h Bit Range Bit Acronym 07 :00 HDR Power Well: Core Bit Description Header Type: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 states that this field must be set to 00H for VFs. Intel(R) Communications Chipset 89xx Series - Datasheet 956 Offset Start: 0Eh Offset End: 0Eh Bus:Function: M:8 + Index1 BAR: Configuration Sticky Bit Reset Value Bit Access 00h RO October 2012 Order Number: 327879-001US 20.0 20.3.1.7 VSVID[0:15]--VF Subsystem Vendor ID Register Table 20-92. VSVID[0:15]--VF Subsystem Vendor ID Register Description: View: PCI VF Size: 16 bit BAR: Configuration Bus:Function: Default: 8086h Offset Start: 2Ch Offset End: 2Dh Power Well: Core Bit Range Bit Acronym Bit Description 15 :00 SVID Subsystem Vendor ID: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that when read, this read only register must return the same value as the Section 20.2.2.13, "PSVID--PF Subsystem Vendor ID Register" for all VFs. 20.3.1.8 M:8 + Index1 Sticky Bit Reset Value Bit Access 8086h RO VSID[0:15]--VF Subsystem ID Register This register is a write-once register. Once any byte in the register has been written, the register locks against further writes until reset. 20.3.1.9 VCP[0:15]--VF Capabilities Pointer Register Table 20-93. VSID[0:15]--VF Subsystem ID Register Description: View: PCI VF Size: 16 bit Bit Range 15 :00 Bus:Function: M:8 + Index1 BAR: Configuration Default: 0000h Offset Start: 2Eh Offset End: 2Fh Power Well: Core Bit Acronym Bit Description Sticky SID Subsystem ID: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this read only register return the same value for all VFs. This may be a different value than that contained in the PF (e.g.,Section 20.2.2.14, "PSID--PF Subsystem ID Register"). Bit Reset Value Bit Access 0000h RWOS The Capabilities Pointer Register provides the offset in configuration space to the location where the first set of capabilities registers is located. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 957 Table 20-94. VCP[0:15]--VF Capabilities Pointer Register Description: View: PCI VF Size: 8 bit BAR: Configuration Bus:Function: Default: 90h Power Well: Core Bit Range Bit Acronym Bit Description 07 :00 CP Capability List Pointer: This field provides an offset into the EP's configuration space pointing to the first item in the EP capability list which in the EP VF is the MSI extended capabilities header. 20.3.1.10 Offset Start: 34h Offset End: 34h M:8 + Index1 Sticky Bit Reset Value Bit Access 90h RO VIRQL[0:15]--VF Interrupt Line Register Table 20-95. VIRQL[0:15]--VF Interrupt Line Register Description: View: PCI VF Size: 8 bit Default: 00h Power Well: Core Bit Range Bit Acronym Bit Description 07 :00 IRQL Interrupt Assigned: This field does not apply to VFs and is hardwired to Zero. 20.3.1.11 Offset Start: 3Ch Offset End: 3Ch + Bus:Function: M:8 Index1 BAR: Configuration Sticky Bit Reset Value Bit Access 0h RO VIRQP[0:15]--VF Interrupt Pin Register Table 20-96. VIRQP[0:15]--VF Interrupt Pin Register Description: View: PCI VF Size: 8 bit Offset Start: 3Dh Offset End: 3Dh Bus:Function: M:8 + Index1 BAR: Configuration Default: 00h Bit Range Bit Acronym 07 :00 IRQP Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0h RO Interrupt Assigned: Note: This field does not apply to VFs and is hardwired to Zero. Intel(R) Communications Chipset 89xx Series - Datasheet 958 October 2012 Order Number: 327879-001US 20.0 20.3.2 VF PCI Express Capability Structure 20.3.2.1 VPCID[0:15]--VF PCI Express Capability ID Register The PCI Express Capability List register enumerates the PCI Express Capability structure in the PCI 3.0 configuration space capability list. Table 20-97. VPCID[0:15]--VF PCI Express Capability ID Register Description: View: PCI VF Size: 8 bit BAR: Configuration M:8 + Index1 Default: 10h Bit Range Bit Acronym 07 :00 PCIECID 20.3.2.2 Bus:Function: Offset Start: 50h Offset End: 50h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 10h RO Cap Id: This field identifies this item in the linked list of Extended Capability Headers as being the PCI Express capability registers. VPCP[0:15]--VF PCI Express Next Capability Pointer Register Table 20-98. VPCP[0:15]--VF PCI Express Next Capability Pointer Register Description: View: PCI VF Size: 8 bit Bus:Function: M:8 + Index1 BAR: Configuration Default: 0h Bit Range Bit Acronym 07 :00 MCP October 2012 Order Number: 327879-001US Offset Start: 51h Offset End: 51h Power Well: Core Bit Description Sticky Next Capability Pointer: Last capability. Bit Reset Value Bit Access 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 959 20.3.2.3 VPCR[0:15]--VF PCI Express Capabilities Register Table 20-99. VPCR[0:15]--VF PCI Express Capabilities Register Description: View: PCI VF Size: 16 bit BAR: Configuration Bit Acronym 15 :14 Reserved SI 7 :4 DPT 3 :0 CV Power Well: Core Bit Reset Value Bit Access 00b RO 00000b RO 0b RO Device/Port Type: Indicates the type of PCI Express logical device. 0000b - PCI Express Endpoint device 0000b RO Capability Version: Indicates PCI-SIG defined PCI Express capability structure version number EP supports version 2h. 0010b RO Bit Description Sticky Reserved Interrupt Message Number: This only applies to Root Complex and Switch devices. This register is hardcoded to 0. IMN 8 20.3.2.4 Offset Start: 52h Offset End: 53h M:8 + Index1 Default: 002h Bit Range 13 :9 Bus:Function: Slot Implemented: Indicates that the PCI Express Link associated with this port is connected to a slot. Only valid for root complex and switch downstream ports. Hard-wired to 0 VPDCAP[0:15]--VF PCI Express Device Capabilities Register Table 20-100.VPDCAP[0:15]--VF PCI Express Device Capabilities Register (Sheet 1 of 2) Description: View: PCI VF Size: 32 bit Default: 10008041h Bit Range Bit Acronym 31 :29 Reserved 28 FLR Power Well: Core Bit Reset Value Bit Access Reserved 0h RO FLR Cap: Function Level Reset Capability is required for all VFs and PFs according to the Single Root I/O Virtualization and Sharing Specification, Revision 0.9. 1b RO Bit Description Sticky 27 :26 CSPS Captured Slot Power Limit Scale: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 states that this field is undefined for all VFs. 00b RO 25 :18 CSPV Captured Slot Power Limit Value: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 states that this field is undefined for all VFs. 0h RO 17 :16 Reserved Reserved 0h RV Role-Based Error Reporting: this bit is set to indicate that this device implements the Role Base Error Reporting defined in PCI Express Base Specification, Revision 2.0. 1b RO 000b RV 15 14 :12 RBEP Reserved Reserved: Undefined - Treated as Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 960 Offset Start: 54h Offset End: 54h M:8 + Bus:Function: Index1 BAR: Configuration October 2012 Order Number: 327879-001US 20.0 Table 20-100.VPDCAP[0:15]--VF PCI Express Device Capabilities Register (Sheet 2 of 2) Description: View: PCI VF Size: 32 bit BAR: Configuration Bus:Function: M:8 + Index1 Default: 10008041h Offset Start: 54h Offset End: 54h Power Well: Core Bit Reset Value Bit Access Endpoint L1 Acceptable Latency: EP does not support L1 active state power management. 000b RO EL0L Endpoint L0 Acceptable Latency: Total acceptable latency that the EP can withstand due to a transition from L0s to L0 state. 001b RO 5 ETFS Extended Tag Field Supported: Indicates the maximum supported size of the Tag field as a Requester. EP does not generate 8-bit Tags but supports 8-bit Tags as a completer. 0b RO 4 :3 PFS Phantom Functions Supported: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to 00b. 00b RO MPS Max_Payload_Size Supported: This field indicates the maximum payload size that EP can support for TLPs. This value is set to indicate 256B. The defined encodings are: * 000b = 128B max payload size * 001b = 256 bytes max payload size (Max supported) * 010b - 111b = Reserved 001b RO Bit Range Bit Acronym 11 :9 EL1L 8 :6 2 :0 20.3.2.5 Bit Description Sticky VPDC[0:15]--VF PCI Express Device Control Register Table 20-101.VPDC[0:15]--VF PCI Express Device Control Register (Sheet 1 of 2) Description: View: PCI VF Size: 16 bit Bit Range 15 M:8 + Bus:Function: Index1 BAR: Configuration Default: 00h Offset Start: 58h Offset End: 59h Power Well: Core Bit Acronym Bit Description Sticky STARTFLR Initiate Function Level Reset: A write of 1b to this bit initiates Function Level Reset to the EP. The value is always read as 0b. Bit Reset Value Bit Access 0b RW 000b RO MRS Max Read Request Size: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. 11 ENS Enable No Snoop: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. 0b RO 10 APME Aux Power PM Enable: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. 0b R0 14 :12 9 PFE Phantom Functions Enable: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. 0b RO 8 ETFE Extended Tag Field Enable: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. 0b RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 961 Table 20-101.VPDC[0:15]--VF PCI Express Device Control Register (Sheet 2 of 2) Description: View: PCI VF Size: 16 bit BAR: Configuration Bus:Function: Default: 00h Power Well: Core Bit Reset Value Bit Access 000b RO Enable Relaxed Ordering: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. 0b RO URRO Unsupported Request Reporting Enable (URRE): The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. In addition the functionality associated with this bit setting is controlled by the corresponding bit in "PPDCNTL--PF PCI Express Device Control Register", and will apply to all VFs. 0b RO FERE Fatal Error Reporting Enable: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. In addition the functionality associated with this bit setting is controlled by the corresponding bit in "PPDCNTL--PF PCI Express Device Control Register", and will apply to all VFs. 0h RO NERE Non-Fatal Error Reporting Enable: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. In addition the functionality associated with this bit setting is controlled by the corresponding bit in "PPDCNTL--PF PCI Express Device Control Register", and will apply to all VFs. 0h RO CERE Correctable Error Reporting Enable: The Single Root I/ O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. In addition the functionality associated with this bit setting is controlled by the corresponding bit in "PPDCNTL--PF PCI Express Device Control Register", and will apply to all VFs. 0h RO Bit Range Bit Acronym Bit Description 7 :5 MPS Max Payload Size: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. 4 ERO 3 2 1 0 20.3.2.6 Offset Start: 58h Offset End: 59h M:8 + Index1 Sticky VPDS[0:15]--VF PCI Express Device Status Register Table 20-102.VPDS[0:15]--VF PCI Express Device Status Register (Sheet 1 of 2) Description: View: PCI VF Size: 16 bit Default: 0000h Bit Range Bit Acronym 15 :6 Reserved Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0h RV 5 TP Transactions Pending: This bit when set indicates that a device has issued Non-Posted Requests which have not been completed. A device reports this bit cleared only when all Completions for any outstanding Non-Posted Requests have been received. 0h RW1C 4 APD AUX Power Detected: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 962 Offset Start: 5Ah Offset End: 5Bh + Bus:Function: M:8 Index1 BAR: Configuration October 2012 Order Number: 327879-001US 20.0 Table 20-102.VPDS[0:15]--VF PCI Express Device Status Register (Sheet 2 of 2) Description: View: PCI VF Size: 16 bit Bit Range 3 2 1 0 20.3.2.7 BAR: Configuration Bus:Function: M:8 + Index1 Default: 0000h Offset Start: 5Ah Offset End: 5Bh Power Well: Core Bit Reset Value Bit Access URD Unsupported Request Detected: This bit indicates that the device received an Unsupported Request. Errors are logged in this register regardless of whether error reporting is enabled or not in the Device Control Register. For a multi-function device, each function indicates status of errors as perceived by the respective function. 0h RW1C FED Fatal Error Detected: This bit indicates status of Fatal errors detected. A one indicates that an error was detected since the last time this bit was cleared. Errors are updated in this field regardless of whether the error reporting bit is set in the Device Control Register. For a multi-function device, each function indicates status of errors as perceived by the respective function. 0h RW1C NED Non-Fatal Error Detected: This bit indicates status of Nonfatal errors detected. A one indicates that an error was detected since the last time this bit was cleared. Errors are updated in this field regardless of whether the error reporting bit is set in the Device Control Register. For a multi-function device, each function indicates status of errors as perceived by the respective function. 0h RW1C CED Correctable Error Detected: This bit indicates status of correctable errors detected. A one indicates that an error was detected since the last time this bit was cleared. Errors are updated in this field regardless of whether the error reporting bit is set in the Device Control Register. For a multi-function device, each function indicates status of errors as perceived by the respective function. 0h RW1C Bit Acronym Bit Description Sticky VLCR[0:15]--VF Link Capabilities Register Table 20-103.VLCR[0:15]--VF Link Capabilities Register (Sheet 1 of 2) Description: Link Capabilities Register View: PCI VF Size: 32 bit Bus:Function: M:8 + Index1 BAR: Configuration Default: 3B502h Bit Range Bit Acronym Power Well: Core Bit Description Sticky 31 :24 PORTNUM Port Number: Assigned by EP after link training phase. 23 :22 Reserved Reserved 21 :21 Reserved 20 :18 Reserved October 2012 Order Number: 327879-001US Offset Start: 5Ch Offset End: 5Fh Bit Reset Value Bit Access 0h RO 00b RO Reserved 0b RO Reserved 000b RO Intel(R) Communications Chipset 89xx Series - Datasheet 963 Table 20-103.VLCR[0:15]--VF Link Capabilities Register (Sheet 2 of 2) Description: Link Capabilities Register View: PCI VF Size: 32 bit Bit Range 17 :15 14 :12 11 :10 9 :4 BAR: Configuration Bus:Function: Offset Start: 5Ch Offset End: 5Fh M:8 + Index1 Default: 3B502h Power Well: Core Bit Reset Value Bit Access L1EL L1 Exit Latency- Indicates the exit latency from L1 to L0 state. EP does not support L1 transition 000b - Less than 1 is 001b - 1 is - 2 is 010b - 2 is - 4 is 011b - 4 is - 8 is 100b - 8 is - 16 is 101b - 16 is - 32 is 110b - 32 is - 64 is 111b - L1 transition not supported 111b RO L0EL L0s Exit Latency- Indicates the exit latency from L0s to L0 state. 000b - Less than 64ns 001b - 64ns - 128ns 010b - 128ns - 256ns 011b - 256ns - 512ns 100b - 512ns - 1 is 101b - 1 is - 2 is 110b - 2 is - 4 is 111b - Reserved 011b RO ASLPM Active State Link PM Support - Indicates the level of active state power management supported in EP. Defined encodings are: 00b - Reserved 01b - L0s Entry Supported 10b - Reserved 11b - L0s and L1 Supported 01b RO 010000b RO 0010b RO Bit Acronym Bit Description LINKW Max Link Width- Indicates the max link width. Relevant encoding: 000000b - Reserved 000001b - x1 000010b - x2 000100b - x4 001000b - x8 001100b - x12 010000b - x16 100000b - x32 Sticky EP value depends on SKU. However the max link width is x16. Max Link Speed - Indicates Maximum supported Link Speed. Defined encodings are : 3 :0 MAXSPEED 0001b - 2.5Gbs Link speed supported (Gen 1) 0010b - 5.0Gbs Link speed supported (Gen 2) EP indicates a max Link Speed of 5.0 Gbs. Intel(R) Communications Chipset 89xx Series - Datasheet 964 October 2012 Order Number: 327879-001US 20.0 20.3.2.8 VLCNTLR[0:15]--VF Link Control Register Table 20-104.VLCNTLR[0:15]--VF Link Control Register Description: Link Control Register View: PCI VF Size: 16 bit BAR: Configuration Bus:Function: M:8 + Index1 Default: 0000h Offset Start: 60h Offset End: 61h Power Well: Core Bit Reset Value Bit Access Reserved 0000b RV Reserved Reserved 0b RO Reserved Reserved 0b RO 9 Reserved Reserved: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. 0b RO 8 ECLKPM The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. 0b RO 7 EXTSYNC Extended Synch: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. 0b RO 6 CCLKCFG Common Clock Configuration: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros.r 0b RO 5 RETRAIN Retrain Link: Not Applicable to endpoints. Hard-wired to 0 0b RO 4 LINKDIS Link Disable: Not Applicable to endpoints. Hard-wired to 0 0b RO Read Completion Boundary (RCB) Control:The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. 0b RO Reserved 0b RO 00b RO Bit Range Bit Acronym 15 :12 Reserved 11 10 3 RCB 2 Reserved 1 :0 ASPMC 20.3.2.9 Bit Description Sticky Active State PM Control: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field is hardwired to all zeros. VLSR[0:15]--VF Link Status Register For the VF, all settings in this field are reserved and the PF setting in the Section 20.2.5.9, "PLSR--PF Link Status Register" applies to all of the VFs. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 965 20.3.2.10 VDCAPR2[0:15]--VF Device Capabilities 2 Register Table 20-105.VLSR[0:15]--VF Link Status Register Description: Link Status Register View: PCI VF Size: 16 bit BAR: Configuration Bus:Function: Offset Start: 62h Offset End: 63h M:8 + Index1 Default: 0000h Bit Range Bit Acronym 15 :00 Reserved Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 0H RV Table 20-106.DCAPR2[0:15]--Device Capabilities 2 Register Description: Device Capabilities 2 Register View: PCI VF Size: 32 bit Default: 00000012h Bit Range Bit Acronym 31 :5 Reserved 4 3 :0 Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved. 0b RV CTODS Completion Timeout Disable Supported. A value of 1b indicates support for the completion timeout disable mechanism. The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 states that this field value must be identical as the PF value. 1b RO CTORS Completion Timeout Ranges Supported: This field indicates support for the optional completion timeout programmability mechanism. This mechanism enables system software to modify the completion timeout value. Four time value ranges are defined: * Range A = 50 us to 10 ms * Range B = 10 ms to 250 ms * Range C = 250 ms to 4 s * Range D = 4 s to 64 s Bits are set according to the following table to show the timeout value ranges that are supported. * 0000b = Completion timeout programming not supported. PCH must implement a timeout value in the range 50 us to 50 ms. * 0001b = Range A. * 0010b = Range B. * 0011b = Ranges A & B. * 0110b = Ranges B & C. * 0111b = Ranges A, B & C. * 1110b = Ranges B, C & D. * 1111b = Ranges A, B, C & D. * All other values are reserved. It is strongly recommended that the completion timeout mechanism not expire in less than 10 ms. The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 states that this field value must be identical as the PF value. 0010b RO Intel(R) Communications Chipset 89xx Series - Datasheet 966 Offset Start: 74h Offset End: 77h Bus:Device:Function: M:8 + Index1 BAR: Configuration October 2012 Order Number: 327879-001US 20.0 20.3.3 VF MSI Capability Structure 20.3.3.1 VMSICID[0:15]--Message Signalled Interrupt Capability ID Register The Message Signalled Interrupt Capability record defines how the device generates PCI MSI messages. It is an 10B PCI SIG-defined capability record and includes the MCID, MCP, MCTL, MADR, and MDATA fields of the configuration header. Table 20-107.VMSICID[0:15]--Message Signalled Interrupt Capability ID Register Description: View: PCI VF Size: 8 bit Default: 05h Bit Range Bit Acronym 07 :00 MCID 20.3.3.2 M:8 + Bus:Function: Index1 BAR: Configuration Offset Start: 90h Offset End: 90h Power Well: Core Bit Description Sticky Capability ID: PCI SIG assigned capability record ID (05h, MSI capability) Bit Reset Value Bit Access 05h RO VMSINCP[0:15]--Message Signalled Interrupt Next Capability Pointer Register Table 20-108.VMSINCP[0:15]--Message Signalled Interrupt Next Capability Pointer Register Description: View: PCI VF Size: 8 bit Bus:Function: M:8 + Index1 BAR: Configuration Default: 00h Power Well: Core Bit Range Bit Acronym Bit Description 07 :00 MCP Next Capability Pointer: Next Capability is PCI Express. October 2012 Order Number: 327879-001US Offset Start: 91h Offset End: 91h Sticky Bit Reset Value Bit Access 00h RO Intel(R) Communications Chipset 89xx Series - Datasheet 967 20.3.3.3 VMSICTL[0:15]--Message Signalled Interrupt Control Register Table 20-109.VMSICTL[0:15]--Message Signalled Interrupt Control Register Description: View: PCI VF Size: 16 bit BAR: Configuration Bus:Function: Offset Start: 92h Offset End: 93h M:8 + Index1 Default: 0180h Bit Range Bit Acronym 15 :09 Reserved Power Well: Core Bit Reset Value Bit Access Reserved 0h RO Bit Description Sticky 08 MC Per-Vector Masking Capable: Per-vector masking capable. 1b RO 07 C64 64 bit Address Capable: Hardwired to 0 to indicate the device does not generate 64b message addresses. 1h RO 06 :04 MME Multiple Message Enable: System software writes to this field to indicate the number of allocated messages (less than or equal to the number of requested messages in MMC). A value of 0 corresponds to one message. 000h RW 03 :01 MMC Multiple Message Capable: System software reads this field to determine the number of requested messages. Hardwired to 0 to request one message. 000h RO MSIE MSI Enable: System software sets this bit to enable MSI signaling. A device driver is prohibited from writing this bit to mask a device's service request. If 1, the device can use an MSI to request service. If 0, the device cannot use an MSI to request service. 0h RW 00 20.3.3.4 VMSILADDR[0:15]--Message Signalled Interrupt Lower Address Register Table 20-110.VMSILADDR[0:15]--Message Signalled Interrupt Lower Address Register Description: View: PCI VF Size: 32 bit Default: 00000000h Power Well: Core Bit Range Bit Acronym Bit Description 31 :02 ADDR Message Address: Written by the system to indicate the lower 30-bits of the address to use for the MSI memory write transaction. 01 :00 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 968 Offset Start: 94h Offset End: 97h + Bus:Function: M:8 Index1 BAR: Configuration Sticky Bit Reset Value Bit Access 0h RW 00b RV October 2012 Order Number: 327879-001US 20.0 20.3.3.5 VMSIUADDR[0:15]--Message Signalled Interrupt Upper Address Register Table 20-111.VMSIUADDR[0:15]--Message Signalled Interrupt Upper Address Register Description: View: PCI VF Size: 32 bit BAR: Configuration Bus:Function: Default: 00000000h Offset Start: 98h Offset End: 9Bh Power Well: Core Bit Range Bit Acronym Bit Description 31 :00 ADDR Message Address: Written by the system to indicate the lower 32-bits of the address to use for the MSI memory write transaction. The lower two bits will always be written as 0. 20.3.3.6 M:8 + Index1 Sticky Bit Reset Value Bit Access 0h RW VMSIDATA[0:15]--Message Signalled Interrupt Data Register Table 20-112.VMSIDATA[0:15]--Message Signalled Interrupt Data Register Description: View: PCI VF Size: 16 bit M:8 + Bus:Function: Index1 BAR: Configuration Default: 0000h Power Well: Core Bit Range Bit Acronym Bit Description 15 :00 DATA Message Data: Written by the system to indicate the lower 16 bits of the data written in the MSI memory write DWORD transaction. The upper 16 bits of the transaction are written as 0. 20.3.3.7 Offset Start: 9Ch Offset End: 9Dh Sticky Bit Reset Value Bit Access 0h RW VMSIMSK--VF Message Signalled Interrupt Mask Register Table 20-113.VMSIMSK--VF Message Signalled Interrupt Mask Register Description: View: PCI Size: 32 bit BAR: Configuration + Bus:Device:Function: M:8 Index1 Default: 00000000h Bit Range Bit Acronym 31 :01 Reserved 00 :00 MASK0 October 2012 Order Number: 327879-001US Offset Start: A0h Offset End: A3h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved. 0h RV Mask Bits: Only one bit defined. See the interrupt section. 0b RW Intel(R) Communications Chipset 89xx Series - Datasheet 969 20.3.3.8 VMSIPND--VF Message Signalled Interrupt Pending Register Table 20-114.VMSIPND--VF Message Signalled Interrupt Pending Register Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: Offset Start: A4h Offset End: A7h M:8 + Index1 Default: 00000000h Bit Range Bit Acronym 31 :01 Reserved 00 :00 Reserved Power Well: Core Bit Reset Value Bit Access Reserved. 0h RV Pending Bits: Only one bit defined. See the interrupt section. 0b RO Bit Description Sticky 20.3.4 VF Advanced Error Reporting Capability Structure 20.3.4.1 VPCIEAERCAPID[0:15]--VF PCI Express AER Capability ID Register The PCI Express Capability List register enumerates the PCI Express AER Capability structure in the PCI 3.0 configuration space capability list. Table 20-115.VPCIEAERCAPID[0:15]--VF PCI Express AER Capability ID Register Description: View: PCI VF Size: 32 bit Default: 13810001h Power Well: Core Bit Reset Value Bit Access Next PCI Express Extended Capability Pointer: This is hardwired to 138H to point to the Alternative Routing ID extended capability. 138h RO PCIEAERCVN Advanced Error Capability Version Number: PCI Express Advanced Error Reporting Extended Capability Version Number. 1h RO PCIEAERCID Advanced Error Capability ID: PCI Express Extended Capability ID indicating Advanced Error Reporting Capability. 1h RO Bit Range Bit Acronym 31 20 PCIEAERNCP 19 :16 15 :00 Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 970 Offset Start: 100h Offset End: 103h M:8 + Bus:Function: Index1 BAR: Configuration Sticky October 2012 Order Number: 327879-001US 20.0 20.3.4.2 VPAERUCS[0:15]--VF PCI Express AER Uncorrectable Error Status Register Table 20-116.VPAERUCS[0:15]--VF PCI Express AER Uncorrectable Error Status Register Description: View: PCI VF Size: 32 bit BAR: Configuration Bus:Function: M:8 + Index1 Default: 0h Bit Range Bit Acronym 31 :21 Reserved Offset Start: 104h Offset End: 107h Power Well: Core Bit Description Sticky Reserved Unsupported Request Error Status: As a receiver, Set whenever an unsupported request is detected. The Header is logged. Bit Access 00b RV 0b RW1CS 20 UR 19 ECRCC ECRC Check: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 0b RV 18 MTLP Malformed TLP: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 0b RV 17 RO Receiver Overflow: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 0b RV 16 EC Unexpected Completion: As a receiver, set whenever a completion is received that does not match the EP requestor ID or outstanding Tag. The Header is logged. Y 0b RW1CS 15 CA Completer Abort: As a completer, set whenever an internal agent signals a data abort. The header is logged. Y 0b RW1CS 14 CT Completion Timeout: As a requester, set whenever an outbound Non Posted Request does not receive a completion within 16-32ms. Y 0b RW1CS 0b RV 0h RW1CS 0000b RV 13 12 11 :6 FCPES Flow Control Protocol Error Status: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. Not supported. PTLPR Poisoned TLP Received: As a receiver, set whenever a poisoned TLP is received from PCI Express. The header is logged. Internal queue errors are not covered by this bit, they are logged by the Configuration target of the transaction. Reserved October 2012 Order Number: 327879-001US Reserved. Y Bit Reset Value Y Intel(R) Communications Chipset 89xx Series - Datasheet 971 Table 20-116.VPAERUCS[0:15]--VF PCI Express AER Uncorrectable Error Status Register Description: View: PCI VF Size: 32 bit Bit Range BAR: Configuration Bus:Function: Default: 0h Bit Acronym Power Well: Core Bit Description Sticky 05 SDES Surprise Down Error: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. Not supported. 04 DLPE Data Link Protocol Error: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 03 00 20.3.4.3 Offset Start: 104h Offset End: 107h M:8 + Index1 Reserved Reserved. Bit Reset Value Bit Access 0b RV 0000b RV 0b RV VPAERUCM[0:15]--VF PCI Express AER Uncorrectable Error Mask Register Table 20-117.VPAERUCM[0:15]--VF PCI Express AER Uncorrectable Error Mask Register (Sheet 1 of 2) Description: View: PCI VF Size: 32 bit Default: 0h Power Well: Core Bit Reset Value Bit Access Reserved 00b RV Reserved ACS Violation Error Mask: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this mask in the Section 20.2.7.3, "PPAERUCM--PF PCI Express AER Uncorrectable Error Mask Register" applies to all of the VFs. 00b RV 20 UR Unsupported Request Error Mask: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this mask in the Section 20.2.7.3, "PPAERUCM--PF PCI Express AER Uncorrectable Error Mask Register" applies to all of the VFs. 0b RV 19 ECRCC ECRC Check Error Mask: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 0b RV 18 MTLP Malformed TLP Error Mask: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 0b RV Bit Range Bit Acronym 31 :22 Reserved 21 Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 972 Offset Start: 108h Offset End: 10Bh + Bus:Function: M:8 Index1 BAR: Configuration Sticky October 2012 Order Number: 327879-001US 20.0 Table 20-117.VPAERUCM[0:15]--VF PCI Express AER Uncorrectable Error Mask Register (Sheet 2 of 2) Description: View: PCI VF Size: 32 bit Bit Range 17 16 15 14 13 12 11 :6 BAR: Configuration Bus:Function: Default: 0h Bit Acronym Bit Description Sticky Bit Reset Value Bit Access RO Receiver Overflow Error Mask: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 0b RV EC Unexpected Completion Error Mask: The Single Root I/ O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this mask in the Section 20.2.7.3, "PPAERUCM--PF PCI Express AER Uncorrectable Error Mask Register" applies to all of the VFs. 0b RV CA Completer Abort Error Mask: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this mask in the Section 20.2.7.3, "PPAERUCM--PF PCI Express AER Uncorrectable Error Mask Register" applies to all of the VFs. 0b RV CT Completion Time Out Error Mask: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this mask in the Section 20.2.7.3, "PPAERUCM--PF PCI Express AER Uncorrectable Error Mask Register" applies to all of the VFs. 0b RV FCPES Flow Control Protocol Error Mask: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. Not supported. 0b RV PTLPR Poisoned TLP Received Error Mask: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this mask in the Section 20.2.7.3, "PPAERUCM--PF PCI Express AER Uncorrectable Error Mask Register" applies to all of the VFs. 0h RV 0000b RV 0b RV 0b RV Reserved Reserved. 05 SDES 04 DLPE Data Link Protocol Error Mask: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 00 Offset Start: 108h Offset End: 10Bh Power Well: Core Surprise Down Error Mask: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. Not supported. 03 :01 M:8 + Index1 Reserved Reserved 0000b RV Reserved Reserved 0b RV October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 973 20.3.4.4 VPAERUCSEV[0:15]--VF PCI Express AER Uncorrectable Error Severity Register Table 20-118.VPAERUCSEV[0:15]--VF PCI Express AER Uncorrectable Error Severity Register (Sheet 1 of 2) Description: View: PCI VF Size: 32 bit BAR: Configuration Bus:Function: Default: 00h Power Well: Core Bit Reset Value Bit Access 00b RV AVES ACS Violation Error Severity: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this field in the Section 20.2.7.4, "PPAERUCSEV--PF PCI Express AER Uncorrectable Error Severity Register"applies to all of the VFs. 0b RV 20 URES Unsupported Request Error Severity: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this field in the Section 20.2.7.4, "PPAERUCSEV--PF PCI Express AER Uncorrectable Error Severity Register" applies to all of the VFs. 0b RV 19 ECRCC ECRC Check Error Severity: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 0b RV 18 MTLP Malformed TLP Error Severity: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 0b RV 17 RO Receiver Overflow Error Severity: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 0b RV EC Unexpected Completion Error Severity: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this field in the Section 20.2.7.4, "PPAERUCSEV--PF PCI Express AER Uncorrectable Error Severity Register" applies to all of the VFs. 0b RV CA Completer Abort Error Severity: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this field in the Section 20.2.7.4, "PPAERUCSEV--PF PCI Express AER Uncorrectable Error Severity Register" applies to all of the VFs. 0b RV CT Completion Time Out Error Severity: The Single Root I/ O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this field in the Section 20.2.7.4, "PPAERUCSEV--PF PCI Express AER Uncorrectable Error Severity Register" applies to all of the VFs. 0b RV Bit Range Bit Acronym 31 :22 Reserved 21 16 15 14 Bit Description Preserved Intel(R) Communications Chipset 89xx Series - Datasheet 974 Offset Start: 10Ch Offset End: 10Fh M:8 + Index1 Preserved Sticky October 2012 Order Number: 327879-001US 20.0 Table 20-118.VPAERUCSEV[0:15]--VF PCI Express AER Uncorrectable Error Severity Register (Sheet 2 of 2) Description: View: PCI VF Size: 32 bit Bit Range 13 12 11 :6 BAR: Configuration Bus:Function: Default: 00h Bit Reset Value Bit Access FCPES Flow Control Protocol Error Severity: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. Not supported. 0b RV PTLPR Poisoned TLP Received Error Field: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this field in the Section 20.2.7.4, "PPAERUCSEV--PF PCI Express AER Uncorrectable Error Severity Register" applies to all of the VFs. 0h RV Preserved 0000b RV 0000b RV 0b RV 0000b RV 0b RV Bit Acronym Bit Description Reserved Sticky 5 SDES 04 DLPE Data Link Protocol Error Severity: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 00 20.3.4.5 Offset Start: 10Ch Offset End: 10Fh Power Well: Core Surprise Down Error Severity: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. Not supported. 03 :01 M:8 + Index1 Reserved Preserved Reserved Undefined: Preserved VPAERCS[0:15]--VF PCI Express AER Correctable Error Status Register Table 20-119.VPAERCS[0:15]--VF PCI Express AER Correctable Error Status Register Description: View: PCI VF Size: 32 bit Default: 0h Bit Range Bit Acronym 31 :14 Reserved 13 ANFES 12 RTTS 11 :09 + Bus:Function: M:8 Index1 BAR: Configuration Reserved October 2012 Order Number: 327879-001US Offset Start: 110h Offset End: 113h Power Well: Core Bit Description Sticky Reserved. Bit Reset Value Bit Access 00b RV Advisory Non-Fatal Error Status: 0b RV Replay Timer Timeout Status: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 0b RV Reserved. 0b RV Intel(R) Communications Chipset 89xx Series - Datasheet 975 Table 20-119.VPAERCS[0:15]--VF PCI Express AER Correctable Error Status Register Description: View: PCI VF Size: 32 bit Bit Range BAR: Configuration Bus:Function: Offset Start: 110h Offset End: 113h M:8 + Index1 Default: 0h Bit Acronym Power Well: Core Bit Description Sticky REPLAY NUM Rollover Status: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. Bit Reset Value Bit Access 0h RV 08 RNRS 07 BDLLPS Bad DLLP Status: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 0000b RV 06 DLPE Bad TLP Status: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. 0b RV 00000b RV 0b RV 05 :01 00 20.3.4.6 Reserved Reserved. Receiver Error Status: As a non-function specific error, the Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros. RES VPAERCM[0:15]--VF PCI Express AER Correctable Error Mask Register Table 20-120.VPAERCM[0:15]--VF PCI Express AER Correctable Error Mask Register Description: View: PCI VF Size: 32 bit Default: 0h Bit Range Bit Acronym 31 :14 Reserved 13 12 11 :09 Power Well: Core Bit Description Reserved Sticky Bit Reset Value Bit Access 00b RV 0b ANFES Advisory Non-Fatal Error Mask: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be treated as Preserved and that the setting of the corresponding bit in the Section 20.2.7.6, "PPAERCM--PF PCI Express AER Correctable Error Mask Register" will apply to the VFs. 0b RTTS Replay Timer Timeout Mask: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be treated as Preserved and that the setting of the corresponding bit in the Section 20.2.7.6, "PPAERCM--PF PCI Express AER Correctable Error Mask Register" will apply to the VFs. Reserved Reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 976 Offset Start: 114h Offset End: 117h M:8 + Bus:Function: Index1 BAR: Configuration RV RV 000b RV October 2012 Order Number: 327879-001US 20.0 Table 20-120.VPAERCM[0:15]--VF PCI Express AER Correctable Error Mask Register Description: View: PCI VF Size: 32 bit Bit Range 08 07 06 05 :01 00 20.3.4.7 BAR: Configuration Bus:Function: M:8 + Index1 Default: 0h Offset Start: 114h Offset End: 117h Power Well: Core Bit Description 0b RNRS REPLAY NUM Rollover Mask: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be treated as Preserved and that the setting of the corresponding bit in the Section 20.2.7.6, "PPAERCM--PF PCI Express AER Correctable Error Mask Register" will apply to the VFs. 0b BDLLPS Bad DLLP Mask: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be treated as Preserved and that the setting of the corresponding bit in the Section 20.2.7.6, "PPAERCM--PF PCI Express AER Correctable Error Mask Register" will apply to the VFs. 0b DLPE Bad TLP Mask: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be treated as Preserved and that the setting of the corresponding bit in the Section 20.2.7.6, "PPAERCM--PF PCI Express AER Correctable Error Mask Register" will apply to the VFs. Reserved Sticky Bit Reset Value Bit Acronym RV RV RV Reserved. 00h Receiver Error Mask: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be treated as Preserved and that the setting of the corresponding bit in the Section 20.2.7.6, "PPAERCM--PF PCI Express AER Correctable Error Mask Register" will apply to the VFs. RES Bit Access RV 0b RV VPAERCTLCAP[0:15]--VF PCI Express AER Control and Capability Register Table 20-121.VPAERCTLCAP[0:15]--VF PCI Express AER Control and Capability Register Description: View: PCI VF Size: 32 bit Bus:Function: M:8 + Index1 BAR: Configuration Default: 0h Offset Start: 118h Offset End: 11Bh Power Well: Core Sticky Bit Access 0 RV Bit Acronym 31 :09 Reserved Reserved 0b 08 ECRCCE ECRC Check Enable: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this field in the Section 20.2.7.1, "PPCIEAERCAPID--PF PCI Express AER Capability ID Register" applies to all of the VFs. ECRC is not supported 07 ECRCCC ECRC Check Capable: Indicates the EP is not capable of checking ECRC. 0b October 2012 Order Number: 327879-001US Bit Description Bit Reset Value Bit Range RV RO Intel(R) Communications Chipset 89xx Series - Datasheet 977 Table 20-121.VPAERCTLCAP[0:15]--VF PCI Express AER Control and Capability Register Description: View: PCI VF Size: 32 bit Bit Range BAR: Configuration Bus:Function: Default: 0h Bit Acronym Power Well: Core Bit Description Sticky 06 ECRCGE ECRC Generation Enable: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 requires that this field be hardwired to all zeros and that the setting of this field in the Section 20.2.7.1, "PPCIEAERCAPID--PF PCI Express AER Capability ID Register" applies to all of the VFs. ECRC is not supported 05 ECRCGC ECRC Generation Capable: Indicates the EP is not capable of generating ECRC. 04 :00 20.3.4.8 Offset Start: 118h Offset End: 11Bh M:8 + Index1 The First Error Pointer: Identifies the bit position of the first error reported in the Section 20-116, "VPAERUCS[0:15]--VF PCI Express AER Uncorrectable Error Status Register" register. Note: This register will not update until all bits in the ERRUNC STS register are cleared. TFEP Bit Reset Value Bit Access 0b RV 00000b Y 0 RO ROS-V VPAERHDRLOG0[0:15]--VF PCI Express AER Header Log 0 Register Table 20-122.VPAERHDRLOG0[0:15]--VF PCI Express AER Header Log 0 Register Description: View: PCI VF Size: 32 bit Bit Range 31 :00 Default: 0h Bit Acronym Power Well: Core Bit Description First DWord of the Header for the PCI Express packet in error (HDRLOGDW0): Once an error is logged in this register, it remains locked HDRLOGDW0 for further error logging until the time the software clears the status bit that cause the header log i.e. the error pointer is rearmed to log again. Intel(R) Communications Chipset 89xx Series - Datasheet 978 Offset Start: 11Ch Offset End: 11Fh Bus:Function: M:8 + Index1 BAR: Configuration Sticky Bit Reset Value Bit Access 0h RO October 2012 Order Number: 327879-001US 20.0 20.3.4.9 VPAERHDRLOG1[0:15]--VF PCI Express AER Header Log 1 Register Table 20-123.VPAERHDRLOG1[0:15]--VF PCI Express AER Header Log 1 Register Description: View: PCI VF Size: 32 bit Bit Range 31 :00 20.3.4.10 BAR: Configuration Bus:Function: M:8 + Index1 Default: 0h Bit Acronym Offset Start: 120h Offset End: 123h Power Well: Core Bit Description Sticky 2nd DWord of the Header for the PCI Express packet in error (HDRLOGDW1): Once an error is logged in this register, it remains locked HDRLOGDW1 for further error logging until the time the software clears the status bit that cause the header log i.e. the error pointer is rearmed to log again. Bit Reset Value Bit Access 0h RO VPAERHDRLOG2[0:15]--VF PCI Express AER Header Log 2 Register Table 20-124.VPAERHDRLOG2[0:15]--VF PCI Express AER Header Log 2 Register Description: View: PCI VF Size: 32 bit Bit Range 31 :00 M:8 + Bus:Function: Index1 BAR: Configuration Default: 0h Bit Acronym Power Well: Core Bit Description Sticky 3rd DWord of the Header for the PCI Express packet in error (HDRLOGDW2): Once an error is logged in this register, it remains locked HDRLOGDW2 for further error logging until the time the software clears the status bit that cause the header log i.e. the error pointer is rearmed to log again. October 2012 Order Number: 327879-001US Offset Start: 124h Offset End: 127h Bit Reset Value Bit Access 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 979 20.3.4.11 VPAERHDRLOG3[0:15]--VF PCI Express AER Header Log 3 Register Table 20-125.VPAERHDRLOG3[0:15]--VF PCI Express AER Header Log 3 Register Description: View: PCI VF Size: 32 bit Bit Range 31 :00 20.3.5 BAR: Configuration Bus:Function: Offset Start: 128h Offset End: 12Bh M:8 + Index1 Default: 0h Bit Acronym Power Well: Core Bit Description Sticky 4th DWord of the Header for the PCI Express packet in error (HDRDWLOG3): Once an error is logged in this register, it remains locked HDRDWLOG3 for further error logging until the time the software clears the status bit that cause the header log i.e. the error pointer is rearmed to log again. Bit Reset Value Bit Access 0h RO VF Alternative Routing ID Extended Capability Structure This section describes the PCI Express Extended Configuration Space registers that make up the Alternative Routing ID Extended Capability Structure. Some information from the specification is repeated here as an aid to the reader or to describe implementation choice. See the PCI Express* Base Specification 2.0 for the full register descriptions and additional information regarding their operation. 20.3.5.1 VARIDHDR[0:15]--VF Alternative Routing ID Capability Header This register contains information associated with the Alternative Routing ID capability . This is compliant with the PCI-SIG ECN: Alternatifve Routing-ID Interpretation (ARI), Updated June 4, 2007. Table 20-126.VARIDHDR[0:15]--VF Alternative Routing ID Capability Header Description: View: PCI VF Size: 32 bit Default: 1000Eh Bit Range Bit Acronym 31 :20 ARINCO 19 :16 ARICV 15 :0 ARICV Power Well: Core Bit Reset Value Bit Access 00h RO Capability Version: This is set to 1h for the most current version of the specification. 1h RO PCI Express Extended Capability ID: The PCI SIG has assigned 000Eh to the ARI extended capability. 000EH RO Bit Description Next Capability Offset: This field contains 000h indicating the end of the VF Extended capability list. Intel(R) Communications Chipset 89xx Series - Datasheet 980 Offset Start: 138h Offset End: 13Bh + Bus:Function: M:8 Index1 BAR: Configuration Sticky October 2012 Order Number: 327879-001US 20.0 20.3.5.2 VFARICAP[0:15]--VF ARI Capabilities Register This register contains information associated with the Alternative Routing ID capability. Table 20-127.VFARICAP[0:15]--VF ARI Capabilities Register Description: View: PCI VF Size: 16 bit BAR: Configuration M:8 + Index1 Default: 00000000h Bit Range Bit Acronym 15 :8 VNFN 7 :2 Bus:Function: Reserved Offset Start: 13Ch Offset End: 13Dh Power Well: Core Bit Reset Value Bit Access Next Function Number: The Single Root I/O Virtualization and Sharing Specification, Revision 0.9 states that this field is undefined for VFs. 0h RV Reserved. 0h RV Bit Description Sticky 1 ACS ACS Functional Groups Capability: not supported. 0b RO 0 MFVC MFVC Functional Groups Capability: not supported. 0b RO 20.3.5.3 VARIDCTL[0:15]--VF Alternative Routing ID Control Register This register contains information associated with the Alternative Routing ID capability. Table 20-128.VARIDCTL[0:15]-- VF Alternative Routing ID Control Register Description: View: PCI VF Size: 16 bit Default: 00000000h Bit Range Bit Acronym 15 :7 Reserved 6 :4 FG 3 :2 Reserved 1 ACS 0 MFVC 20.4 + Bus:Function: M:8 Index1 BAR: Configuration Offset Start: 13Eh Offset End: 13Fh Power Well: Core Bit Reset Value Bit Access Reserved 0h RV Function Group: Hardwired to Zero. 0b RO Reserved 0b RV Bit Description Sticky ACS Functional Groups Enable: Hardwired to Zero. 0b RO MFVC Functional Groups Enable: Hardwired to Zero. 0b RO EP Memory Mapped Registers This section describes the MMIO registers that are located in the EP. These registers are exposed from PCI. 20.4.1 Detailed Register Summary 20.4.2 CSRs This section describes the CSRs that are mapped in PMISCBAR MMIO space. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 981 20.4.2.1 ERRSOU2--Error Source Register 2 This register captures the error sources that are generated by the Ring Controller. Table 20-129.ERRSOU2--Error Source Register 2 Description: EP Error Sources View: PCI Size: 32 bits BAR: PMISCBAR+1A000h Default: 00000000h Power Well: Core Bit Range Bit Acronym 31 :16 Reserved Reserved 15 :0 BUN150 Bit[n] maps to Interrupt from Bundle[n], where n varies from 0 to 15. 20.4.2.2 Offset Start: 008h Offset End: 00Bh Bus:Device:Function: B:D:0 Bit Description Sticky Bit Reset Value Bit Access N 0b RO ERRMSK2--Error Source Mask Register 2 ERRMSK2 provides mask bits for ERRSOU2 interrupts. This register can be used to enable/disable interrupts pending in ERRSOU2. Software can poll ERRSOU2 even if the interrupts are masked. Table 20-130.ERRMSK2--Error Source Mask Register 2 Description: Error Source Mask 2 View: PCI Size: 32 bits BAR: PMISCBAR+1A000h Default: 00000000h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved Reserved BUN150 Mask bits: Bit[n] maps to Interrupt from Bundle[n], where n varies from 0 to 15. N 0b RW Bit Range Bit Acronym 31 :16 15 :0 Intel(R) Communications Chipset 89xx Series - Datasheet 982 Offset Start: 018h Offset End: 01Bh Bus:Device:Function: B:D:0 October 2012 Order Number: 327879-001US 20.0 20.4.2.3 SINTPF--Signal Raw PF Interrupt Register This register reflects the EP interrupts for the PF. These interrupts can either be routed to the IA based on values programmed in SMIAPF register. See Section 20.4.2.4, "SMIAPF--Signal IA PF Interrupt Mask Register". The interrupt is sourced from the ERRSOU[2] register. Table 20-131.SINTPF--Signal Raw PF Interrupt Register Description: SDATA View: PCI BAR: PMISCBAR+1A000h Size: 32 bits Bus:Device:Function: B:D:0 Default: 00000000h Power Well: Core Bit Range Bit Acronym 31: 16 Reserved Reserved BUN0 15 Bundle15 0 Interrupts. See Table 20-129, "ERRSOU2-- Error Source Register 2" for interrupts. 15 :00 20.4.2.4 Offset Start: 024h Offset End: 027h Bit Description Sticky Bit Reset Value Bit Access 0h RV 0000h RO SMIAPF--Signal IA PF Interrupt Mask Register This register controls the ability for the PF to generate an IA interrupt, based on interrupt reflected in the SINTPF register. Each mask bit corresponds to a bit in the SINTPF register. See Table 20-131, "SINTPF--Signal Raw PF Interrupt Register". This register is used to route interrupts reflected in SINTPF to the IA. Software should program SMIAPF register once during system initialization and should not re-program them during normal operation as this register is not meant to be used as interrupt mask bits. In addition, software should program this register appropriately so interrupts are not routed to the IA. Table 20-132.SMIAPF--Signal IA PF Interrupt Mask Register Description: SDATA View: PCI Size: 32 bits BAR: PMISCBAR+1A000h Bus:Device:Function: B:D:0 Default: 00000000h Power Well: Core Bit Range Bit Acronym 31 :16 Reserved Reserved 15 :00 BUN0 15 Bundle15 0 Interrupts Mask bit: If set to 1h, an interrupt is sent to the IA as a either an INTx, MSI, or MSI-X. See Table 20-131, "SINTPF--Signal Raw PF Interrupt Register". October 2012 Order Number: 327879-001US Offset Start: 028h Offset End: 02Bh Bit Description Sticky Bit Reset Value Bit Access 0h RW 0000h RW Intel(R) Communications Chipset 89xx Series - Datasheet 983 20.4.2.5 GBECFGMMIOV--GBE Configuration and MMIO Valid Register Since the Configuration and MMIO CSRs are loaded from EEPROM. This register contains four read-only signals, one per GbE port, that indicate when the GbE Configuration and MMIO CSRs are valid. The EP will use these signals and respond with a Configuration Request Retry Status if the Configuration spaces are not valid when the Host tries to access the GbE configuration spaces. Table 20-133.GBECFGMMIOV - GBE Configuration and MMIO Valid Register Description: GbE CFG and MMIO Valid. View: PCI BAR: PMISCBAR+1A000h View: PCI PF BAR: PMISCBAR+1A000h Size: 32 bits Offset Start: 03Ch Offset End: 03Fh Bus:Function: M:0 Default: 00000000h Bit Range Bit Acronym 31 :04 Reserved Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved :03 GBE3 GbE3:GbE CSRs are loaded from EEPROM this signal indicates when the GbE config and MMIO CSRs are valid. 0 - Valid 1 - Not valid 0b RO :02 GBE2 GbE2:GbE CSRs are loaded from EEPROM this signal indicates when the GbE config and MMIO CSRs are valid. 0 - Valid 1 - Not valid 0b RO :01 GBE1 GbE1:GbE CSRs are loaded from EEPROM this signal indicates when the GbE config and MMIO CSRs are valid. 0 - Valid 1 - Not valid 0b RO GBE0 GbE0:GbE CSRs are loaded from EEPROM this signal indicates when the GbE config and MMIO CSRs are valid. 0 - Valid 1 - Not valid 0b RO :00 Intel(R) Communications Chipset 89xx Series - Datasheet 984 Offset Start: 03Ch Offset End: 03Fh Bus:Device:Function: B:D:0 October 2012 Order Number: 327879-001US 20.0 20.4.2.6 SINTGBE[0:3]--Signal Raw PF Interrupt GBE Register This register reflects the GbE interrupts. These interrupts can be routed to the IA based on values programmed in SMMEGBE and SMIAGBE registers. See Section 20.4.2.7, "SMIAGBE[0:3]--Signal IA Interrupt Mask GBE Register". These interrupts are sourced from the GbEs. Table 20-134.SINTGBE[0:3]--Signal Raw PF Interrupt for GBE Register Description: SDATA View: PCI BAR: PMISCBAR+1A000h Size: 32 bits Default: 00000000h Bit Range Bit Acronym 31: 10 Reserved :09 MISCINT :08 TCPTIMER 07 :00 20.4.2.7 Bus:Device:Function: B:D:0 RXTX07 040h / at Offset Start: 14h Offset End: 043h / at 14h Power Well: Core Bit Reset Value Bit Access Reserved 0h RV Other causes -- Summarizes legacy interrupts into one extended cause. 0h RO Bit Description Sticky TCP Timer 8 Rx/Tx Queue Interrupts 0h RO 00h RO SMIAGBE[0:3]--Signal IA Interrupt Mask GBE Register This register controls the ability for the GBE to generate an IA interrupt, based on pending interrupt the SINTGBE register. Each mask bit corresponds to a bit in the SINTGBE register. See Table 20-134, "SINTGBE[0:3]--Signal Raw PF Interrupt for GBE Register". This register is used to route interrupts reflected in SINTGBE to either IA. Software should program SMIAGBE register once during system initialization and should not re-program this register during normal operation as thesethis register is not meant to be used as interrupt mask bits. In addition, software should program this register appropriately so interrupts are not routed to the IA. . Table 20-135.SMIAGBE[0:3]--Signal IA PF Interrupt Mask GBE Register Description: SDATA View: PCI Size: 32 bits BAR: PMISCBAR+1A000h Bus:Device:Function: B:D:0 Default: 000003FFh Bit Range Bit Acronym 31 :10 Reserved October 2012 Order Number: 327879-001US Power Well: Core Bit Description Reserved 044h / at Offset Start: 14h Offset End: 047h / at 14h Sticky Bit Reset Value Bit Access 0h RW Intel(R) Communications Chipset 89xx Series - Datasheet 985 Table 20-135.SMIAGBE[0:3]--Signal IA PF Interrupt Mask GBE Register Description: SDATA View: PCI Size: 32 bits Bit Range BAR: PMISCBAR+1A000h 044h / at Offset Start: 14h Offset End: 047h / at 14h Bus:Device:Function: B:D:0 Default: 000003FFh Bit Acronym Power Well: Core Bit Description Sticky Bit Reset Value Bit Access :09 MISCINT MISCINT from SINTGBE. When this bit is set, an interrupt is sent to the IA as either an INTx, MSI, or MSI-X. Default interrupt steered to IA. 1b RW :08 TCPTIMER TCP Timer from SINTGBE. When this bit is set, an interrupt is sent to the IA as either an INTx, MSI, or MSI-X. Default interrupt steered to IA. 1b RW RXTX07 8 Rx/Tx Queue Interrupts. 1-1 mapping to SINTGBE bits. When a bit is set, an interrupt is sent to the IA as either an INTx, MSI, or MSI-X. Default interrupts steered to IA. FFh RW 07 :00 Each VINTSOUx is an interrupt pending register which corresponds to a given VFx. The register can be used by software to identify the source of the VFx interrupt. A VINTSOUx register is mapped in both PF and its corresponding VFx address space. Access to the PF is provided only for testing or debugging purposes. Interrupts in the VINTSOUx can be masked using the its corresponding VINTMSKx register. Each VINTMSKx corresponds to a VINTSOUx register. Each VINTMSKx corresponds a given VFx. Software can use the VINTMSKx to enable and disable interrupts pending in VINSOUx. Intel(R) Communications Chipset 89xx Series - Datasheet 986 October 2012 Order Number: 327879-001US 21.0 21.0 GbE Controller Overview The PCH Gigabit Ethernet (GbE) Controller provides four fully-integrated Gigabit Ethernet Media Access Control (MAC) controllers with four SGMII/SerDes ports that can be connected to an external PHY. The GbE Controller can support up to four SerDes ports LAN on Motherboard (LOM) design or mezzanine card. The GbE Controller can also be used in embedded applications such as add-on card switches and network appliances. Note: The PCH GbE Ethernet Interface and the integrated four GbE MAC Controllers, are collectively referred to as the "GbE Controller" or the "Controller" throughout this document. 21.1 Feature Summary External interfaces provided: * Serializer-Deserializer (SerDes) to support 1000Base-SX/LX (optical fiber). * Serializer-Deserializer (SerDes) to support 1000BASE-KX and 1000BASE-BX for Gigabit backplane applications. * SGMII interface for SFP/external PHY1 connections * SMBus for Manageability connection to BMC Intel(R) I/O Acceleration Technology v3.0 is supported * Stateless offloads (Header split, RSS) * Direct Cache Access Power saving features: * Advanced Configuration and Power Interface (ACPI) power management states and wake-up capability * Advanced Power Management (APM) wake-up functionality Additional product details: * Support for PCI 3.0 vital product data * Eight TX and eight RX queues * IPMI BMC pass-thru * DMTF Management Component Transport Protocol (MCTP) over SMBus/ I2C and MCTP over PCIe Vendor Defined Messaging * IEEE 1588 Precision Time Protocol support * Per-Packet timestamp 1. See the "Supported Ethernet PHY Devices for the Intel(R) Communications Chipset 89xx Series" Application Note. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 987 21.2 Scope This document provides the external architecture (including device operation, pin descriptions, register definitions, etc.) for the GbE interface in. This document is a reference for software device driver developers, board designers, test engineers, and others who may need specific technical or programming information. 21.3 Terminology and Acronyms Table 21-1. Glossary Term Definition 1000BASE-BX 1000BASE-BX is the PICMG 3.1 electrical specification for transmission of 1 Gb/s Ethernet or 1 Gb/s fibre channel encoded data over the backplane. 1000BASE-KX 1000BASE-KX is the IEEE802.3ap electrical specification for transmission of 1 Gb/s Ethernet over the backplane. 1000BASE-CX 1000BASE-X over specialty shielded 150 balanced copper jumper cable assemblies as specified in IEEE 802.3 Clause 39. 1000BASE-T 1000BASE-T is the specification for 1 Gb/s Ethernet over category 5e twisted pair cables as defined in IEEE 802.3 clause 40. b/w Bandwidth. BIOS Basic Input/Output System. BMC Baseboard Management Controller. BT Bit Time. CGB Control Interface for the GbE Controller on the EP Bus DCA Direct Cache Access. DFT Design for Testability. DQ Descriptor Queue. DW Double word (4 bytes). EEPROM Electrically Erasable Programmable Memory. A non-volatile memory located on the LAN controller that is directly accessible from the host. EOP End of Packet. FC Flow Control. Firmware (FW) Embedded code on the LAN controller that is responsible for the implementation of the pass through functionality. Host Interface RAM on the LAN controller that is shared between the firmware and the host. RAM is used to pass commands from the host to firmware and responses from the firmware to the host. HPC High - Performance Computing. IPC Inter Processor Communication. IPG Inter Packet Gap. LAN (auxiliary Power-Up) The event of connecting the LAN controller to a power source (occurs even before system power-up). LOM LAN on Motherboard. LTR Latency Tolerance Reporting (PCIe protocol) LSO Large Send Offload. MAC Media Access Control. Intel(R) Communications Chipset 89xx Series - Datasheet 988 October 2012 Order Number: 327879-001US 21.0 Table 21-1. Glossary (Continued) Term 21.3.1 Definition MDIO Management Data Input/Output Interface over MDC/MDIO lines. MIFS/MIPG Minimum Inter Frame Spacing/Minimum Inter Packet Gap. MMW Maximum Memory Window. MSS Maximum Segment Size. Largest amount of data, in a packet (without headers) that can be transmitted. Specified in Bytes. MPS Maximum Payload Size in PCIe specification. MTU Maximum Transmit Unit. Largest packet size (headers and data) that can be transmitted. Specified in bytes. NIC Network Interface Controller. OBFF Opportunistic Buffer Flush/Fill (PCIe protocol). PBA Printed Board Assembly PCS Physical Coding Sub layer. PHY Physical Layer Device. PMA Physical Medium Attachment. PMD Physical Medium Dependent. RMII Reduced Media Independent Interface (Reduced MII). SA (In MAC context) Source Address. SDP Software Defined Pins. SerDes Serializer and De-Serializer Circuit. SFD Start Frame Delimiter. SGMII Serialized Gigabit Media Independent Interface. SMBus System Management Bus. A bus that carries various manageability components, including the LAN controller, BIOS, sensors and remote-control devices. TCO Total Cost of Ownership (TCO) System Management. TLP Transaction Layer Packet in the PCI Express specification. TPH TLP Process Hints (PCIe protocol). TSO Transmit Segmentation offload - A mode in which a large TCP/UDP I/O is handled to the device and the device segments it to L2 packets according to the requested MSS. VPD Vital Product Data (PCI protocol). External Specification and Documents The GbE Controller implements features from the following specifications. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 989 21.3.1.1 Network Interface Documents * IEEE standard 802.3, 2006 Edition (Ethernet). Incorporates various IEEE Standards previously published separately. Institute of Electrical and Electronic Engineers (IEEE). * IEEE standard 1149.1, 2001 Edition (JTAG). Institute of Electrical and Electronics Engineers (IEEE) * IEEE Std 1149.6-2003, IEEE Standard for Boundary-Scan Testing of Advanced Digital Networks, IEEE, 2003. * IEEE standard 802.1Q for VLAN * PICMG3.1 Ethernet/Fibre Channel Over PICMG 3.0 Draft Specification, January 14, 2003, Version D1.0 * Serial-GMII Specification, Cisco Systems document ENG-46158, Revision 1.7 * INF-8074i Specification for SFP (Small Form factor Pluggable) Transceiver (ftp://ftp.seagate.com/sff) * IEEE Std 802.3ap-2007 * IEEE 1588* Standard for a Precision Clock Synchronization Protocol for Networked Measurement and Control Systems, November 8 2002 * IEEE 802.1AS Timing and Synchronization for Time- Sensitive Applications in Bridged Local Area Networks Draft 2.0, February 22, 2008 21.3.1.2 Host Interface Documents * PCI-Express 2.0 Base specification, Revision 1.0 * PCI Specification, version 3.0 * PCI Bus Power Management Interface Specification, Rev. 1.2, March 2004 * Advanced Configuration and Power Interface Specification, Rev 2.0b, October 2002 21.3.1.3 Networking Protocol Documents * IPv4 specification (RFC 791) * IPv6 specification (RFC 2460) * TCP/UDP specification (RFC 793/768) * SCTP specification (RFC 2960) * ARP specification (RFC 826) * EUI-64 specification, http://standards.ieee.org/regauth/oui/tutorials/EUI64.html. 21.3.1.4 Manageability Document System Management Bus (SMBus) Specification, SBS Implementers Forum, Ver. 2.0, August 2000 21.4 Product Overview The Controller supports four SerDes or SGMII ports for MAC to external PHY connections. Intel(R) Communications Chipset 89xx Series - Datasheet 990 October 2012 Order Number: 327879-001US 21.0 21.5 External Interface Figure 21-1 shows the high-level connectivity of the GbE interface in a PCH implementation. Figure 21-1. High-Level View of GbE Interface Connectivity in PCH Implementation 21.5.1 PCIe Interface (Connected to PCI Express Through PCH PCIe End Point) The PCIe Gen2 Interface, provided by PCH PCIe Endpoint implementation, is used by PCH as a host interface. The four GbE ports of the Controller behave like four separate PCI functions in End Point Configuration. 21.5.2 Network Interfaces Four independent interfaces are used to connect the four Controller ports to external devices. The following protocols are supported: * SerDes interface to connect over a backplane to another SerDes compliant device or to an Optical module. The Controller supports both 1000BASE-BX and 1000BASE-KX (Without IEEE802.3ap Backplane Auto-Negotiation). * SGMII interface to attach to an external PHY, either on board or via an SFP module. The SGMII interface shares the same pins as the SerDes. 21.5.3 EEPROM Interface The EEPROM stores information needed for the GbE MACs and other PCH functionality. This other functionality mandates an EEPROM necessary for PCH even if the GbE MACs are not being used. However, if all GbE MACs are disabled via the EEPROM, then the IA core will not be able to access the EEPROM. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 991 The Controller uses an EEPROM device for storing product configuration information. Several words of the EEPROM are accessed automatically by the Controller after reset in order to provide pre-boot configuration data that must be available to PCH before it is accessed by host software. The remainder of the stored information is accessed by various software modules used to report product configuration, serial number, etc. The Controller is intended for use with a SPI (4-wire) serial EEPROM device such as an AT25040AN or compatible EEPROM device. 21.5.4 SMBus Interface SMBus is an optional interface for pass-through and/or configuration traffic between a BMC and the Controller. The Controller's SMBus interface can be configured to support both slow and fast timing modes. 21.5.5 MDIO/I2C Two-Wire Interfaces The Controller implements four management Interfaces for control of an external PHY. Each interface can be either a 2-wire Standard-mode I2C interface used to control an SFP module or an MII Management Interface (also known as the Management Data Input/Output or MDIO Interface) for control plane connection between the MAC and PHY devices (master side). This interface provides the MAC and software with the ability to monitor and control the state of the external PHY. The Controller supports the data formats defined in IEEE 802.3 clause 22. The Controller supports shared MDIO operation for all ports or separate MDIO connection per port. When configured via the MDICNFG register to separate MDIO operation each MDIO interface should be connected to the relevant PHY. When configured via the MDICNFG register to shared MDIO operation the MDC/MDIO interface of LAN port 0 can be shared by all ports to support connection to a multi-port PHY with a single MDC/MDIO interface. 21.5.6 Software-Definable Pins (SDP) Interface (General-Purpose I/O) The Controller has two software-defined pins (SDP pins) per port that can be used for miscellaneous hardware or software-control purposes. These pins can be individually configurable to act as either input or output pins. The default direction of each pin is configurable via the EEPROM (see Section 24.3.13 and Section 28.1.4.2), as well as the default value of all pins configured as outputs. To avoid signal contention, all pins are set as input pins until the EEPROM configuration is loaded. All SDP pins can be configured for use as general-purpose interrupt (GPI) inputs. To act as GPI pins, the desired pins must be configured as inputs. A corresponding GPI interrupt-detection enable bit is then used to enable rising-edge detection of the input pin (rising-edge detection occurs by comparing values sampled at the internal clock rate, as opposed to an edge-detection circuit). When detected, a corresponding GPI interrupt is indicated in the Interrupt Cause register. The use, direction, and values of SDP pins are controlled and accessed using fields in the Device Control (CTRL) register and Extended Device Control (CTRL_EXT) register. Intel(R) Communications Chipset 89xx Series - Datasheet 992 October 2012 Order Number: 327879-001US 21.0 21.5.7 LED Interface The Controller implements one output driver per port intended for driving external LED circuits. LED outputs can be individually configured to select the particular event, state, or activity, which is indicated on that output. In addition, each LED can be individually configured for output polarity as well as for blinking versus non-blinking (steady-state) indication. The configuration for LED outputs is specified via the LEDCTL register. Furthermore, the hardware-default configuration for all LED outputs can be specified via EEPROM fields (see Section 24.3.12 and Section 24.3.12), thereby supporting LED displays configurable to a particular OEM preference. See Section 26.5 for more detailed description of LED behavior. 21.6 GbE Controller Features and Capabilities Table 21-2 through Table 21-7 list the Controller's features and capabilities. Table 21-2. GbE Controller Features GbE Controller Feature Serial FLASH interface N 4-wire SPI EEPROM interface Y Configurable LED operation for software or OEM custom-tailoring of LED displays Y Protected EEPROM space for private configuration N Watchdog timer Y Table 21-3. GbE Controller Network Features GbE Controller Feature Half duplex at 10/100 Mb/s operation and full duplex operation at all supported speeds Y Jumbo frames supported Y Size of jumbo frames supported 9.5 KB Flow control support: send/receive PAUSE frames and receive FIFO thresholds Y Statistics for management and RMON Y 802.1q VLAN support Y SerDes interface for external PHY connection or system interconnect 1000BASE-KX interface for Blade Server Backplane connections 802.3ap Backplane Auto-negotiation 4 ports Y N SGMII interface for external 1000BASE-T PHY connection 4 ports SerDes support of non-Auto-Negotiation partner Y SerDes signal detect Y External PHY control I/F MDC/MDIO 2 wire I/F I2C clock stretching October 2012 Order Number: 327879-001US Shared or per function Per function Y Intel(R) Communications Chipset 89xx Series - Datasheet 993 Table 21-4. GbE Controller Host Interface Features Feature GbE Controller 64-bit address support for systems using more than 4 GB of physical memory Y 24 Per port and for all ports Outstanding requests for Tx buffers per port Outstanding requests for Tx descriptors per port 4 Per port and for all ports Outstanding requests for Rx descriptors per port 4 Per port and for all ports Max payload size supported 256 Bytes Max request size supported 1 KB Vital Product Data (VPD) Y 64-bit CSR access using 64-bit operationa N CSR access via Configuration space Y a. 64-bit CSRs should be access using 2 seperate 32-bit read/write operations. Table 21-5. GbE Controller LAN Functions Features Feature Programmable host memory receive buffers Y Descriptor ring management hardware for transmit and receive Y ACPI register set and power down functionality supporting D0 & D3 states Y IEEE802.3az (Energy Efficient Ethernet) N Software controlled global reset bit (resets everything except the configuration registers) Y Software Definable Pins (SDP) - per port 2 SDP pins can be configured as general purpose interrupts Y Wake up Y Flexible wake-up filters 8 Flexible filters for queue assignment in normal operation 8 IPv6 wake-up filters Default configuration by the EEPROM for all LEDs for pre-driver functionality Y 1 LED LAN function disable capability Y Programmable memory transmit buffers Y Double VLAN Y IEEE 1588 Y Per-Packet Timestamp Y Intel(R) Communications Chipset 89xx Series - Datasheet 994 GbE Controller October 2012 Order Number: 327879-001US 21.0 Table 21-6. GbE Controller LAN Performance Features Feature GbE Controller TCP segmentation offload Up to 256 KB Y IPv6 support for IP/TCP and IP/UDP receive checksum offload Y Fragmented UDP checksum offload for packet reassembly Y Message Signaled Interrupts (MSI) Y Message Signaled Interrupts (MSI-X) number of vectors per port 10 Packet interrupt coalescing timers (packet timers) and absolute-delay interrupt timers for both transmit and receive operation Y Interrupt throttling control to limit maximum interrupt rate and improve CPU utilization Y Rx packet split header Y Receive Side Scaling (RSS) number of queues per port Up to 8 Total number of Rx queues per port 8 Total number of TX queues per port 8 RX header replication Low latency interrupt DCA support (Legacy DCA support only) TCP timer interrupts No snoop Relax ordering Yes to all TSO interleaving for reduced latency Y Receive side coalescing N SCTP receive and transmit checksum offload Y UDP TSO Y IPSec offload N Table 21-7. GbE Controller Virtualization Features Feature GbE Controller Support for Virtual Machines Device queues (VMDq) per port L2 MAC address filters (unicast and multicast) L2 VLAN filters PCI-SIG SR-IOV Multicast/Broadcast Packet replication 8 pools (single queue) 24 Per pool N Y on Receive VM to VM Packet forwarding (Packet Loopback) N RSS replication N Traffic shaping N MAC and VLAN anti-spoofing N Malicious driver detection Y Per-pool statistics Y October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 995 Table 21-7. GbE Controller Virtualization Features (Continued) Feature GbE Controller Per-pool off loads Y Per-pool jumbo support Y Mirroring rules 4 21.6.1 Network Interface 21.6.1.1 Quad Port The Controller supports four LAN ports (10/100/1000BASE-T on SerDes) and associated MAC and DMA queues enabling reduction of BOM cost and board space. 21.6.1.2 Shared MDIO Support The Controller enables sharing of the MDIO interface on LAN port 0 by all ports connected to an external 1000BASE-T PHY to support connection to an external multi-port PHY device. To abstract actual MDIO interface configuration from Software driver, MDIO setup (PHY Address and Shared MDIO) is loaded into the MDICNFG register from the EEPROM following reset. Further information can be found in Section 22.5.2.2 and Section 28.1.4.6. 21.6.1.3 I2C Clock Stretching There are situations where an I2C slave can not support the clock speed or needs to delay an access initiated by the master. Delaying the access is done by a mechanism referred to as clock stretching. An I2C slave is allowed to hold down the clock if it needs to delay an access. The master is required to read back the clock signal after releasing it to a high-z state and wait until the line has actually gone high as a result of an external Pull-up resistor. When configured to operate in I2C mode, the Controller supports clock stretching on the I2C_CLK pin. When accessing PHY registers via the I2C interface the PHY can delay the access or reduce clock speed if required. 21.6.1.4 1000BASE-KX Backplane Ethernet Interface The Controller can interface up to four 1Gbps 1000BASE-KX lanes for Blade Server backplane interconnect without need for additional glue logic. The Controller supports only parallel detection of 1000BASE-KX signaling and does not support the full auto-negotiation for Backplane Ethernet protocol. 21.6.2 HOST Interface 21.6.2.1 PCIe Connectivity through EndPoint Interface The Controller connects to PCIe through a resident End Point as four functions of that EndPoint. 21.6.2.2 64-bit BAR Support The Controller supports 64-bit BAR. Intel(R) Communications Chipset 89xx Series - Datasheet 996 October 2012 Order Number: 327879-001US 21.0 21.6.3 Performance Features 21.6.4 Receive and Transmit Queues The number of Receive and Transmit queues supported by the Controller for Receive Side Scaling (RSS) filtering and Virtualization is 8 per port. 21.6.4.1 Unused Receive and Transmit Ports Buffer Sharing The Controller supports up to four Gigabit Ethernet ports. When device is configured to work only as a single port or dual port device, available internal receive and transmit buffer memory for the remaining active ports, can be increased by four for the single port case and by two for the dual port case. See Section 26.1.3.2 and Section 26.2.1.2 for more information. 21.6.5 Virtualization The Controller does not support the following virtualization features: * PCIe SR-IOV * VM to VM packet forwarding * Anti-Spoof protection The Controller supports 8 queues per port that can be shared between eight virtual machines (VMDq2). By associating receive and transmit packets to separate queues dedicated to a VM, performance of the VMM (Virtual Machine Monitor) is improved. The Controller also supports the following hardware Virtualization features: * Per VM interrupt assignment * Enabling read of statistic counters by VMs without initiating a clear by read * Copy of received multicast and broadcast packets to multiple queues * Per VM Offload * Per Pool statistics * Mirroring * Storm Control 21.6.5.1 Malicious Driver Detection Malicious behavior exhibited by the driver can be a result of incorrect activation of the network controller or a virus on a certain VM. The Controller contains internal circuitry to protect from an attack on one virtual machine disrupting operation of other virtual machines. When malicious driver behavior is detected on a certain queue, PCH disables activity of the queue and sends notification to the VMM. See Section 26.8.3.7.2 for details. 21.6.5.2 2-tuple Filtering The Controller supports only 2-tuple filtering compared to previous Intel products that supported 5-tuple filtering. The Controller enables queuing, generating immediate interrupts and time stamping according to TCP/UDP port destination address and L4 protocol fields. See Chapter 26.0, "GbE Inline Functions" for more information. 21.6.6 Security Offload The Controller does not support the IPSec offload features. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 997 21.6.6.1 Transmit Queue Prioritization The Controller supports strict Transmit queue prioritization to enable transmission of high priority real-time packets ahead of low priority packets. A Transmit queue is either a high priority queue or low priority queue, with the high priority queues always served ahead of the low priority queues. See Section 26.2.2.5 for information. 21.6.7 Manageability 21.6.7.1 SMBus Interface to External BMC The Controller supports four managed ports on a single SMBus interface to external BMC. 21.6.7.1.1 Teaming and Fail-over Support by BMC The LAN ports act independently of each other and each port has a separate SMBus address. As a result, fail-over where manageability traffic is routed to an active port if any of the ports fail, is not supported internally. The BMC is responsible for teaming and fail-over. 21.6.7.1.2 Auto-ARP Reply on SMBus The Controller cannot be programmed for auto-ARP reply on reception of ARP request packets and does not support sending of gratuitous ARP packets to reduce the traffic over the BMC interconnect. 21.6.7.2 Exclusive Management Filtering In the Controller, decisions regarding forwarding of packets to the host and to the BMC are separate and are configured through two sets of registers. However, the BMC may define some types of traffic as exclusive. This traffic will be forwarded only to the BMC, even if it passes the filtering process of the host. These types of traffic are defined using the MNGONLY register. Further information can be found in Section 26.1.2.3 and Section 27.4.1. 21.6.8 Time SYNC (IEEE1588 and IEEE 802.1AS) 21.6.8.1 Per Packet Timestamp The Controller supports adding an optional tailored header before the MAC header of the packet in the receive buffer. The 128 bit tailored header contains a 64 bit timestamp (MSB bits) and 64 reserved bits (LSB bits). The timestamp is sampled on the MAC/PHY interface, on detection of a received packet that meets certain criteria. Time stamping the received packet on the PHY/MAC interface avoids incurring additional delays due to PCIe latencies. The timestamp is placed in the receive buffer ahead of the actual received packet. See "Inline Functions/Receive Packet Timestamp in Buffer" section for more details. 21.6.8.2 Improved System Time Accuracy System time accuracy has been increased in the Controller. System time corrections with a resolution of 2-32 ns can be entered using the TIMINCA.Incvalue field. 21.6.8.3 SYSTIM Synchronized Pulse Generation on SDP Pins The Controller has the capability to generate a pulse synchronized to system time. Intel(R) Communications Chipset 89xx Series - Datasheet 998 October 2012 Order Number: 327879-001US 21.0 21.6.8.4 Time SYNC Interrupts The Controller has the additional capability to generate interrupts on occurrence of Time Sync events. 21.7 Device Data Flows 21.7.1 Transmit Data Flow Tx data flow provides a high level description of all data/control transformation steps needed for sending Ethernet packets to the line. Table 21-8. Transmit Data Flow Step Description 1 The host creates a descriptor ring and configures one of PCH's transmit queues with the address location, length, head and tail pointers of the ring (one of 16 available Tx queues). 2 The host is requested by the TCP/IP stack to transmit a packet, it gets the packet data within one or more data buffers. 3 The host initializes descriptor(s) that point to the data buffer(s) and have additional control parameters that describe the needed hardware functionality. The host places that descriptor in the correct location at the appropriate Tx ring. 4 The host updates the appropriate queue tail pointer (TDT) 5 PCH's DMA senses a change of a specific TDT and as a result sends a request to fetch the descriptor(s) from host memory. 6 The descriptor(s) content is received in a PCIe read completion and is written to the appropriate location in the descriptor queue internal cache. 7 The DMA fetches the next descriptor from the internal cache and processes its content. As a result, the DMA sends requests to fetch the packet data from system memory. 8 The packet data is received and passes through the transmit DMA that performs all programmed data manipulations (various CPU off loading tasks as checksum off load, TSO off load, etc.) on the packet data on the fly. 9 While the packet is passing through the DMA, it is stored into the transmit FIFO. After the entire packet is stored in the transmit FIFO, it is forwarded to the transmit switch module. 10 The transmit switch arbitrates between host and management packets and eventually forwards the packet to the MAC. 11 The MAC appends the L2 CRC to the packet and sends the packet to the line using a pre-configured interface. 12 When all the completions for a given packet are done, the DMA updates the appropriate descriptor(s). 13 After enough descriptors are gathered for write back or the interrupt moderation timer expires, the descriptors are written back to host memory using writes transactions (through EP interface). Alternatively, the head pointer can only be written back. 14 After the interrupt moderation timer expires, an interrupt is generated to notify the host device driver that the specific packet has been read to PCH and the driver can release the buffers. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 999 21.7.2 Rx Data Flow Rx data flow provides a high level description of all data/control transformation steps needed for receiving Ethernet packets. Table 21-9. Receive Data Flow Step Description 1 The host creates a descriptor ring and configures one of PCH's receive queues with the address location, length, head, and tail pointers of the ring (one of 16 available Rx queues). 2 The host initializes descriptors that point to empty data buffers. The host places these descriptors in the correct location at the appropriate Rx ring. 3 The host updates the appropriate queue tail pointer (RDT). 4 PCH's DMA senses a change of a specific RDT and as a result sends a request to fetch the descriptors from host memory. 5 The descriptors content is received in a read completion (using EP interface) and is written to the appropriate location in the descriptor queue internal cache. 6 As packet enters the Rx MAC. The RX MAC checks the CRC of the packet. 7 The MAC forwards the packet to an Rx filter 8 If the packet matches the pre-programmed criteria of the Rx filtering, it is forwarded to the Rx FIFO. VLAN and CRC are optionally stripped from the packet and L3/L4 checksum are checked and the destination queue is fixed. 9 The receive DMA fetches the next descriptor from the internal cache of the appropriate queue to be used for the next received packet. 10 After the entire packet is placed into the Rx FIFO, the receive DMA posts the packet data to the location indicated by the descriptor through the EP interface. If the packet size is greater than the buffer size, more descriptors are fetched and their buffers are used for the received packet. 11 When the packet is placed into host memory, the receive DMA updates all the descriptor(s) that were used by packet data. 12 After enough descriptors are gathered for write back or the interrupt moderation timer expires or the packet requires immediate forwarding, the receive DMA writes back the descriptor content along with status bits that indicate the packet information including what off loads were done on that packet. 13 After the interrupt moderation timer completes or an immediate packet is received, PCH initiates an interrupt to the host to indicate that a new received packet is already in host memory. 14 Host reads the packet data and sends it to the TCP/IP stack for further processing. The host releases the associated buffers and descriptors once they are no longer in use. Intel(R) Communications Chipset 89xx Series - Datasheet 1000 October 2012 Order Number: 327879-001US 22.0 22.0 GbE Interconnects 22.1 PCIe Interface The GbE interface connects to PCIe through the EndPoint, therefore all PCIe related transaction details as well as error handling are explained in the PCIe EndPoint section of this document. 22.1.1 Special GbE Controller Features * Packet sizes/formats: -- Maximum upstream (write) PCIe payload size of 256 bytes -- Maximum downstream (read) PCIe payload size of 1K bytes * Baseline messaging: -- In-band messaging of formerly side-band legacy signals (such as interrupts, etc.) -- System-level power management supported via messages * Power management: -- Full support for PCI-PM -- Wake capability from D3cold state -- Compliant with ACPI, PCI-PM software model -- Active state power management 22.1.2 Relaxed Ordering The controller takes advantage of the relaxed ordering rules in PCIe. By setting the relaxed ordering bit in the packet header, the controller enables the system to optimize performance in the following cases: * Relaxed ordering for descriptor and data reads: When the controller emits a read transaction, its split completion has no ordering relationship with the writes from the CPUs (same direction). It should be allowed to bypass the writes from the CPUs. * Relaxed ordering for receiving data writes: When the controller issues receive DMA data writes, it also enables them to bypass each other in the path to system memory because software does not process this data until their associated descriptor writes complete. * The controller cannot relax ordering for descriptor writes, MSI/MSI-X writes or PCIe messages. Relaxed ordering can be used in conjunction with the no-snoop attribute to enable the memory controller to advance non-snoop writes ahead of earlier snooped writes. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1001 Relaxed ordering is enabled in the controller by clearing the RO_DIS bit in the CTRL_EXT register. Actual setting of relaxed ordering is done for LAN traffic by the host through the DCA registers. 22.1.3 Snoop Not Required The controller sets the Snoop Not Required attribute bit for master data writes. System logic might provide a separate path into system memory for non-coherent traffic. The non-coherent path to system memory provides higher, more uniform, bandwidth for write requests. Note: The Snoop Not Required attribute does not alter transaction ordering. Therefore, to achieve maximum benefit from Snoop Not Required transactions, it is advisable to set the relaxed ordering attribute as well (assuming that system logic supports both attributes). In fact, some chipsets require that relaxed ordering is set for no-snoop to take effect. Global no-snoop support is enabled in the controller by clearing the NS_DIS bit in the CTRL_EXT register. Actual setting of no snoop is done for LAN traffic by the host through the DCA registers. 22.1.4 No Snoop and Relaxed Ordering for LAN Traffic Software might configure non-snoop and relax order attributes for each queue and each type of transaction by setting the respective bits in the RXCTRL and TXCTRL registers. Table 22-1 lists the default behavior for the No-Snoop and Relaxed Ordering bits for LAN traffic when I/OAT 2 is enabled. Table 22-1. LAN Traffic Attributes No-Snoop Default Relaxed Ordering Default Rx Descriptor Read N Y Rx Descriptor Write-Back N N Relaxed ordering must never be used for this traffic. Rx Data Write Y Y See the note below. Transaction Rx Replicated Header N Y Tx Descriptor Read N Y Tx Descriptor Write-Back N Y Tx TSO Header Read N Y Tx Data Read N Y Note: Rx payload no-snoop is also conditioned by the NSE bit in the receive descriptor. See Section 22.1.4.1. Intel(R) Communications Chipset 89xx Series - Datasheet 1002 Comments October 2012 Order Number: 327879-001US 22.0 22.1.4.1 No-Snoop Option for Payload Under certain conditions, which occur when I/OAT is enabled, software knows that it is safe to transfer (DMA) a new packet into a certain buffer without snooping. This scenario typically occurs when software is posting a receive buffer to hardware that the CPU has not accessed since the last time it was owned by hardware. This might happen if the data was transferred to an application buffer by the I/OAT DMA engine. In this case, software should be able to set a bit in the receive descriptor indicating that the controller should perform a no-snoop DMA transfer when it eventually writes a packet to this buffer. When a non-snoop transaction is activated, the TLP header has a non-snoop attribute in the Transaction Descriptor field. This is triggered by the NSE bit in the receive descriptor. 22.1.5 PCIe Power Management Described in Chapter 25.0, "GbE Power Management". 22.2 Management Interfaces The controller contains one SMBus interface to an external BMC for manageability. * SMBus Since the manageability sideband throughput is lower than the network link throughput, the controller allocates a 2 KB internal buffer for incoming network packets prior to being sent over the sideband interface. The controller also allocates a 2 KB internal buffer for outgoing network packets prior to being sent over the Ethernet link. Allocated buffer size is a function of number of active LAN ports. If only 2 LAN ports are enabled, internal Buffer size is 4 KB. If only a single LAN port is enabled, internal buffer size is 8 KB. Enabled LAN ports are defined in the Manageability Capability / Manageability Enable EEPROM word (Word 0x54). 22.2.1 SMBus SMBus is an optional interface for pass-through and/or configuration traffic between an external BMC and the controller. The SMBus commands used to configure or read status from the controller are described in Chapter 27.0, "GbE Platform Manageability". 22.2.1.1 Channel Behavior 22.2.1.1.1 SMBus Addressing The SMBus is presented as four SMBus devices on the SMBus were each device has a different SMBus addresses. All pass-through functionality is duplicated per SMBus address, where each SMBus address is connected to a different LAN port. Note: DO NOT configure ports to the same address. When a LAN function is disabled, the corresponding SMBus address is not presented to the external BMC. The SMBus addresses are set by SMBus Address 0, SMBus Address 1, SMBus Address 2 and SMBus Address 3 words in the EEPROM. The SMBus addresses (those that are enabled from the EEPROM) can be re-assigned using the SMBus ARP protocol. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1003 Besides the SMBus address values, all the previously stated parameters of the SMBus (SMBus channel selection, address mode, address enable) can be set only through EEPROM configuration. The EEPROM is read by the controller at power-up, resets, and other cases described in Section 23.2, "Reset Operation". All SMBus addresses should be in Network Byte Order (NBO); most significant byte first. 22.2.1.1.2 SMBus Notification Methods The controller supports three methods of informing the external BMC that it has information that is needed to be read by an external BMC: * SMBus alert. * Asynchronous notify. * Direct receive. The notification method that is used by the controller can be configured from the SMBus using the Receive Enable command. The default method is set from the EEPROM in the Notification Method field. The following events cause the controller to send a notification event to the external BMC: * Receiving a LAN packet that was designated to the BMC. * Receiving a request status command from the BMC initiates a status response (see Chapter 27.0, "GbE Platform Manageability"). * Status change has occurred and the controller is configured to notify the external BMC upon one of the status changes. The following event triggers a notification to the BMC: -- A change in any of the Status Data 1 bits of the Read Status command (see Chapter 27.0, "GbE Platform Manageability" for description of this command). There might be cases where the external BMC is hung and is unable to respond to the SMBus notification. The controller has a time-out value defined in the EEPROM (see Section 24.0, "Non-Volatile Memory Map - EEPROM") to avoid hanging while waiting for the notification response. If the BMC does not respond before the timeout expires, the notification is de-asserted. 22.2.1.1.2.1 SMBus Alert and Alert Response Method The SMBus Alert# signal acts as an asynchronous interrupt signal to an external SMBus master. The controller asserts this signal each time it has a message that it needs the external BMC to read and if the chosen notification method is the SMBus-alert method. The SMBus alert is an open-drain signal, which means that other devices besides the controller can be connected on the same alert pin and the external BMC needs a mechanism to distinguish between the alert sources as described: The external BMC can respond to the alert by issuing an ARA (Alert Response Address) cycle (see Figure 22-2) to detect the alert source device. The controller responds to the ARA cycle (if it was the SMBus alert source) and de-asserts the alert when the ARA cycle completes. Following the ARA cycle, the external BMC issues a Read command to retrieve the controller message. Some BMCs do not implement ARA cycle transactions. These BMCs respond to an alert by issuing a Read command to all active controller ports (0xC0/0xD0 or 0xDE). The controller always responds to a Read command, even if it is not the source of the Intel(R) Communications Chipset 89xx Series - Datasheet 1004 October 2012 Order Number: 327879-001US 22.0 notification. The default response is a status transaction. If the controller is the source of the SMBus alert, it replies to the read transaction and de-asserts the alert after the command byte of the read transaction. The ARA cycle is an SMBus receive byte transaction to SMBus Address 0001-100b. The ARA transaction does not support PEC. The ARA transaction format is as follows: Table 22-2. SMBus ARA Cycle Format 1 7 1 1 7 S Alert Response Address Rd A Slave Device Address 0 Manageability Slave SMBus Address 0001 100 Note: 1 1 0 1 1 A P 1 Since the master-receiver (BMC receiver) is involved in the transaction it must signal the end of data to the PCH by generating a NACK (a `1' in the ACK bit position) on the Slave Device Address byte that was clocked out by the PCH. The PCH releases the data line to allow the master to generate a STOP condition. 22.2.1.1.2.2 Asynchronous Notify Method When configured to asynchronous notify method, the controller acts as SMBus master and notifies the external BMC by issuing a modified form of the write word transaction. The asynchronous notify transaction SMBus address and data payload is configured using the Receive Enable command or using the EEPROM defaults. The asynchronous notify method is not protected by a PEC byte. Table 22-3. Asynchronous Notify Command Format 1 7 1 1 S Target Address Wr A Sending Device Address 0 Manageability Slave SMBus Address BMC Slave Address 0 7 8 1 8 1 1 Data Byte Low A Data Byte High A P Interface 0 Alert Value 0 1 1 A 0 *** 0 The target address and data byte low/high is taken from the Receive Enable command (see Chapter 27.0, "GbE Platform Manageability") or EEPROM configuration (See Section 24.0, "Non-Volatile Memory Map - EEPROM"). 22.2.1.1.2.3 Direct Receive Method If configured, the controller has the capability to send the message it needs to transfer to the external BMC as a master over the SMBus, instead of alerting the BMC, and waiting for it to read the message. Table 22-4 shows the message format when receiving a LAN packet that was designated to the BMC. The "F", "L" and command fields in the message are the same as the op-code returned by the PCH in response to a BMC Receive TCO Packet Block October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1005 read command (See Chapter 27.0, "GbE Platform Manageability"). The rules for the "F" and "L" flags are also the same as used in the Receive TCO Packet Block Read command. Table 22-4. Direct Receive - LAN Packet Receive Transaction Format 1 7 1 1 1 1 6 1 S Target Address Wr A F L Command A BMC Slave Address 0 0 First Flag Last Flag Receive TCO Command 01 0000b 0 8 1 8 1 Byte Count A Data Byte 1 A N 0 *** 0 1 8 1 1 A Data Byte N A P 0 *** 0 The following table shows the message format when a status change has occurred and the controller is configured to notify the external BMC upon a status change. The op-code and Status data fields returned by the controller are the same as in response to a BMC Read Status command (see Chapter 27.0, "GbE Platform Manageability"). Table 22-5. Direct Receive - Status Change Transaction Format 1 7 1 1 S Target Address Wr BMC Slave Address 22.2.1.1.3 0 8 1 A Command A 0 Read status op-code 0xDD 0 8 1 8 1 8 1 1 Byte Count A Data 2 (Status Data 1) A Data 3 (Status Data 2) A P 0x02 0 0 *** 0 Receive TCO Flow The controller is used as a channel for receiving packets from the network link and passing them to the external BMC. The BMC can configure the controller to pass specific packets to the BMC. Once a full packet is received from the link and identified as a manageability packet that should be transferred to the BMC, the controller starts the receive TCO transaction flow to the BMC. The maximum SMBus fragment length is defined in the EERPOM. The controller uses the SMBus notification method to notify the BMC that it has data to deliver. The packet is divided into fragments, where the controller uses the maximum fragment size allowed in each fragment. The last fragment of the packet transfer is always the status of the packet. As a result, the packet is transferred in at least two fragments. The data of the packet is transferred in the Receive TCO LAN packet transaction. Intel(R) Communications Chipset 89xx Series - Datasheet 1006 October 2012 Order Number: 327879-001US 22.0 When SMBus alert is selected as the BMC notification method, the controller notifies the BMC on each fragment of a multi-fragment packet. When asynchronous notify is selected as the BMC notification method, the controller notifies the BMC only on the first fragment of a received packet. It is BMC's responsibility to read the full packet including all the fragments. Any timeout on SMBus notification results in discarding the entire packet. Any NACK by the BMC on one of the controller receive bytes also causes the packet to be silently discarded. If a SMBus time-out occurs during reception of a packet from the network to the BMC, the controller silently discards the packet. The maximum size of the received packet is limited by the controller hardware to 1536 bytes. Packets larger then 1536 bytes are silently discarded. Any packet smaller than 1536 bytes is processed by the controller. Note: When the RCV_EN bit is cleared, all receive TCO functionality is disabled, not just the packets that are directed to the BMC. 22.2.1.1.4 Transmit TCO Flow The controller is used as a channel for transmitting packets from the external BMC to the network link. The network packet is transferred from the external BMC over the SMBus, and then, when fully received by the controller, is transmitted over the network link. Each SMBus address is connected to a different LAN port. When a packet is received in SMBus transactions using SMBus Address 0, SMBus Address 1, SMBus Address 2 or SMBus Address 3 it is transmitted to the network using LAN port 0, LAN port 1, LAN port 2 or LAN port 3 respectively. The controller supports packets up to the Ethernet packet length (1536 bytes). SMBus transactions can be up to 240 bytes in length, which means that packets can be transferred over the SMBus in more than one fragment. In each command byte there are the F and L bits. When the F bit is set, it means that this is the first fragment of the packet; L means that it is the last fragment of the packet. Note: When both flags are set, the entire packet is in one fragment. The packet is sent over the network link, only after all its fragments are received correctly over the SMBus. The controller calculates the L2 CRC on the transmitted packet and adds its four bytes at the end of the packet. Any other packet field (such as XSUM) must be calculated and inserted by the external BMC (The controller does not change any field in the transmitted packet, besides adding padding and CRC bytes). Note: If the packet sent by the BMC is larger than 1536 bytes, then the packet is discarded by the controller and Abort is asserted. The minimum packet length defined by the 802.3 specification is 64 bytes. The controller pads packets that are less than 64 bytes to meet the specification requirements. There is one exception, when the packet sent over the SMBus is less than 32 bytes, the external BMC must pad it for at least 32 bytes. The passing bytes value should be zero. Note: Packets that are smaller than 32 bytes (including padding), are discarded by the controller and Abort is asserted. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1007 If the network link goes down at anytime while the controller is receiving a packet from the BMC for transmission on the network, it silently discards the packet. Any link down event during the transfer of a packet to the BMC over the SMBus (after it is fully received from the network), does not stop the operation. Note: If a SMBus time-out occurs during transmission of a packet from the BMC to the network the controller silently discards the packet. The transmit SMBus transactions are described in Chapter 27.0, "GbE Platform Manageability". 22.2.1.1.4.4 Transmit Errors in Sequence Handling Once a packet is transferred over the SMBus from the BMC to the controller, the F and L flags should follow specific rules. The F flag defines that this is the first fragment of the packet; The L flag defines that the transaction contains the last fragment of the packet. The following table lists the options regarding the flags in transmit packet transactions: Table 22-6. Flags in Transmit Packet Transactions Previous Current Action/Notes Last First Accepts both. Last Not First Error for current transaction. Current transaction is discarded and an abort status is asserted. Not Last First Error for previous transaction. Previous transaction (until previous first) is discarded. Current packet is processed. No abort status is asserted. Not Last Not First Processes the current transaction. Since every other Block Write command in the TCO protocol has both F and L flags off, they cause flushing any pending transmit fragments that were previously received. In other words, when running the TCO transmit flow, no other block write transactions are allowed in between the fragments. 22.2.1.1.5 TCO Command Aborted Flow Bit 6 in first byte of the status returned from the controller to the external BMC indicates that there was a problem with previous SMBus transactions or with the completion of the operation requested in previous transaction. An abort can be asserted for any of the following reasons: * Any error in the SMBus protocol (NACK, SMBus timeout). * If the BMC does not respond until the notification timeout (programmed in the EEPROM) expires. * Any error in compatibility between required protocols to specific functionality (Receive Enable command with byte count not 1/14 as defined in the command specification). * If the controller does not have space to store the transmit packet from the BMC (in its internal buffer before sending it to the link). In this case, the entire transaction completes, but the packet is discarded and the BMC is notified about it through the Abort bit. * Error in the F/L bit sequence during multi-fragment transactions. * If the packet sent by the BMC is larger than 1536 bytes. Intel(R) Communications Chipset 89xx Series - Datasheet 1008 October 2012 Order Number: 327879-001US 22.0 * If the packet sent by the BMC is smaller than 32 bytes (including padding). * An internal reset to the controller manageability unit occurred. * Following an unsuccessful internal register access when the MCSR_TO_RETRY EEPROM bit is cleared. Note: An abort in the status does not always imply that the last transaction of the sequence was incorrect. There is a gap between the time the status is read from the controller and the time the transaction occurred. 22.2.1.1.6 Concurrent SMBus Transactions Concurrent SMBus transactions (receive, transmit and configuration read/write) are allowed between the four addresses supported by the controller. Transmit fragments can be sent between receive fragments and configuration Read/Write commands can also be issued between receive and transmit fragments. 22.2.1.1.7 SMBus ARP Functionality The controller supports SMBus ARP protocol as defined in the SMBus 2.0 specification. The controller is a persistent slave address device meaning that its SMBus address is valid after power-up and loaded from the EEPROM. The controller supports all SMBus ARP commands defined in the SMBus specification, both general and directed. Note: SMBus ARP can be disabled through EEPROM configuration (See Chapter 27.0, "GbE Platform Manageability"). SMBus-ARP transactions are described in Chapter 27.0, "GbE Platform Manageability". 22.2.1.1.7.5 SMBus ARP Response Behavior The controller responds as four SMBus devices, meaning that it has four sets of AR/AV flags (one for each port). The controller responds four times to the SMBus-ARP master, one time for each port. All SMBus addresses are taken from the SMBus ARP addresses word of the EEPROM. The UDID is different between the four ports in the Vendor Specific ID field, which represent the MAC address, which is different between the four ports. The controller answers first as port 0, and only when the address is assigned, starts answering as port 1, 2, and 3 to the Get UDID command. 22.2.1.1.7.6 SMBus ARP Flow SMBus-ARP flow is based on the status of two flags: * AV - Address Valid - This flag is set when the controller has a valid SMBus address. * AR - Address Resolved - This flag is set when the controller's SMBus address is resolved (SMBus address was assigned by the SMBus-ARP process). Note: These flags are internal the controller flags and not shown to external SMBus devices. Since the controller is a Persistent SMBus Address (PSA) device, the AV flag is always set, while the AR flag is cleared after power-up until the SMBus-ARP process completes. Since the AV flag is always set, the PCH always has a valid SMBus address. When the SMBus master needs to start an SMBus-ARP process, it resets (In terms of ARP functionality) all the devices on the SMBus by issuing either Prepare to ARP or Reset Device commands. When the controller accepts one of these commands, it clears its AR flag (if set from previous SMBus-ARP process), but not its AV flag (the current SMBus address remains valid until the end of the SMBus ARP process). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1009 The meaning of an AR flag cleared is that the controller answers the following SMBus ARP transactions that are issued by the master. The SMBus master then issues a Get UDID command (general or directed), to identify the devices on the SMBus. The controller responds to the directed command all the time and to the general command only if its AR flag is not set. After the Get UDID command, the master assigns the controller's SMBus address by issuing an Assign Address command. The controller checks whether the UDID matches its own UDID, and if they match, it switches its SMBus address to the address assigned by the command (byte 17). After accepting the Assign Address command, the AR flag is set and from this point on (as long as the AR flag is set), the controller does not respond to the Get UDID general command, while all other commands should be processed even if the AR flag is set. The controller stores the SMBus address that was assigned in the SMBus-ARP process in its EEPROM, so after the next power-up, it returns to its assigned SMBus address. Figure 22-7 shows the SMBus-ARP behavior of the controller. Intel(R) Communications Chipset 89xx Series - Datasheet 1010 October 2012 Order Number: 327879-001US 22.0 Figure 22-7. SMBus ARP Flow Power-Up reset Set AV flag; Clear AR flag Load SMB address from EPROM SMB packet received NO Yes Process regular command NO SMB ARP address match Yes Prepare to ARP ? YES ACK the comamd and clear AR flag YES ACK the comamd and clear AR flag NO Reset device NO Assign Address command YES UDID match YES AR flag set NO Get UDID command directed YES YES NACK packet ACK packet Set slave address Set AR flag. YES NO Get UDID command general NO NO Return UDID NACK packet Return UDID NO Illegal command handling October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1011 22.2.1.1.7.8 SMBus ARP UDID Content The Unique Device Identifier (UDID) provides a mechanism to isolate each device for the purpose of address assignment. Each device has a unique identifier. The 128-bit number is comprised of the following fields: Table 22-7. Unique Device Identifier (UDID) 1 Byte 1 Byte 2 Bytes 2 Bytes 2 Bytes 2 Bytes 2 Bytes 4 Bytes Device Capabilities Version / Revision Vendor ID Device ID Interface Sub-system Vendor ID Sub- system Device ID Vendor Specific ID See below See below 0x8086 PCIe Dev ID 0x0004 0x0000 0x0000 See below MSB LSB Where: * Vendor ID - The device manufacturer's ID as assigned by the SBS Implementers' Forum or the PCI SIG - Constant value: 0x8086. * Device ID - The device ID as assigned by the device manufacturer (identified by the Vendor ID field) - Constant value * Interface - Identifies the protocol layer interfaces supported over the SMBus connection by the device - In this case, SMBus Version 2.0 - Constant value: 0x0004. * Sub-system Fields - These fields are not supported and return zeros. Device Capabilities: Dynamic and Persistent Address, PEC Support bit: Table 22-8. Dynamic and Persistent Address, PEC Support Bit 7 6 5 Address Type 0b 1b 4 3 2 1 Reserved (0) Reserved (0) Reserved (0) Reserved (0) Reserved (0) 0b 0b 0b 0b 0b 0 PEC Supported 0b MSB LSB Version/Revision: UDID Version 1, Silicon Revision: Table 22-9. Version/Revision: UDID Version 1, Silicon Revision 7 6 Reserved (0) Reserved (0) 5 UDID Version 4 Silicon Revision ID 0b 0b 001b See below MSB 3 2 1 0 LSB Table 22-10. Silicon Revision ID Silicon version Revision ID A0 000b B0 001b C0 010b C1 011b Intel(R) Communications Chipset 89xx Series - Datasheet 1012 October 2012 Order Number: 327879-001US 22.0 Vendor Specific ID - Four LSB bytes of the controller Ethernet MAC address. The controller Ethernet address is taken from offsets 0 to 1 from start of the relevant sections in the EEPROM. In the PCH there are four MAC addresses (one for each port). Table 22-11. Vendor Specific ID 1 Byte 1 Byte 1 Byte 1 Byte MAC Address, byte 3 MAC Address, byte 2 MAC Address, byte 1 MAC Address, byte 0 MSB LSB 22.3 EEPROM 22.3.1 EEPROM Interface 22.3.1.1 General Overview The PCH uses an EEPROM device for storing product configuration information. The EEPROM is divided into three general regions: * Hardware accessed - Loaded by the controller after power-up, PCI reset de-assertion, (D3 ->D0 transition, or a software-commanded EEPROM read (CTRL_EXT.EE_RST). * Manageability firmware accessed - Loaded by the controller in pass-through mode after power-up or firmware reset. * Software accessed - Used only by software. The meaning of these registers, as listed here, is a convention for software only and is ignored by the controller. Table 22-12 lists the structure of the EEPROM image in the PCH . Table 22-12. EEPROM Structure Address Content 0x0 - 0x9 LAN 0 MAC address and software area 0xA - 0x2F LAN 0 and Common hardware area 0x30 - 0x3E PXE area 0x3F Software Checksum, for Words 0x00 - 0x3F 0x40 - 0x4F Software area 0x50 - 0x7F FW pointers 0x80 -0xBF LAN 1 hardware area (with SW checksum in 0xBF) 0xC0 - 0xFF LAN 2 hardware area (with SW checksum in 0xFF) 0x100 - 0x13F LAN 3 hardware area (with SW checksum in 0x13F) The EEPROM mapping is described in Section 24.0, "Non-Volatile Memory Map EEPROM". October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1013 22.3.1.2 EEPROM Device The EEPROM interface supports an SPI interface and expects the EEPROM to be capable of 2 MHz operation. The controller is compatible with various sizes of 4-wire serial EEPROM devices. If pass-through mode functionality is desired, up to 256 Kbits serial SPI compatible EEPROM can be used. If no manageability mode is desired, a 8192-bit serial SPI compatible EEPROM can be used. All EEPROM's are accessed in 16-bit words although the EEPROM is designed to also accept 8-bit data accesses. The controller automatically determines the address size to be used with the SPI EEPROM it is connected to and sets the EEPROM address size field of the EEPROM Control and Data register (EEC.EE_ADDR_SIZE) field appropriately. Software can use this size to determine how to access the EEPROM. The exact size of the EEPROM is determined within one of the EEPROM words. Note: The different EEPROM sizes have two differing numbers of address bits (8 bits or 16 bits), and therefore must be accessed with a slightly different serial protocol. Software must be aware of this if it accesses the EEPROM using direct access. 22.3.1.3 HW Initial Load Process Upon power on reset or PCIe reset, the controller reads the global device parameters from the EEPROM including all the parameters impacting the content of the PCIe configuration space. Upon a software reset to one of the ports (CTRL.RST set to 1), a partial load is done of the parameters relevant to the port where the software reset occurred. Upon a software reset to all ports (CTRL.DEV_RST = 1) a partial load is done of the parameters relevant to all ports. Table 22-13 lists the words read in each EEPROM auto-read sequence. During full load after power-on all hardware related EEPROM words are loaded. Following a software reset only a subset of the hardware related EEPROM words are loaded. For details of the content of each word - see Section 24.0, "Non-Volatile Memory Map - EEPROM". LANx_start parameter in Table 22-13 relates to start of LAN related EEPROM section where: -- LAN0_start = 0x0 -- LAN1_start = 0x80 -- LAN2_start = 0xC0 -- LAN3_start = 0x100 Table 22-13. EEPROM Auto-Load Sequence (Sheet 1 of 4) EEPROM Word EEPROM Word Address EEPROM sizing 012 Y CSR Auto Configuration Power-Up LAN0 027 Y CSR Auto Configuration Power-Up LAN1 LAN1_start + 027 Y CSR Auto Configuration Power-Up LAN2 LAN2_start + 027 Y Intel(R) Communications Chipset 89xx Series - Datasheet 1014 Full Load (Power-up) Full Load No MGMT (PCI RST) Y SWa reset port 0 Load Y SWa reset port 1 Load Y SWa reset port 2 Load Y SWa reset port 3 Load Y October 2012 Order Number: 327879-001US 22.0 Table 22-13. EEPROM Auto-Load Sequence (Sheet 2 of 4) EEPROM Word EEPROM Word Address Full Load (Power-up) CSR Auto Configuration Power-Up LAN3 LAN3_start + 027 Y Init Control 1 00A Y Full Load No MGMT (PCI RST) Y Device Rev ID 01E Y Y Subsystem IDb 00B Y Y Subsystem Vendor IDb 00C Y Y Device ID - LAN 0c 00D Y Y c LAN1_start + 00D Y Y Device ID - LAN 2c LAN2_start + 00D Y Y 3c LAN3_start + 00D Y Y 00E Y Y 016 Y Y LAN1_start + 016 Y Y LAN2_start + 016 Y Y LAN3_start + 016 Y Y LAN power consumption 022 Y Y CSR Auto Configuration Pointer and CSR Auto Configuration structures LAN0 017 Y Y CSR Auto Configuration Pointer and CSR Auto Configuration structures LAN1 LAN1_start + 017 Y Y CSR Auto Configuration Pointer and CSR Auto Configuration structures LAN2 LAN2_start + 017 Y Y CSR Auto Configuration Pointer and CSR Auto Configuration structures LAN3 LAN3_start + 017 Y Y VPD Pointer to table 02F Y Y VPD table entry ID TAG ID STRING Y Y VPD read or write area TAG VPD TAG 1 Y Y Device ID - LAN 1 Device ID - LAN Vendor ID - LAN 0c Phy_Rst_Control0 Phy_Rst_Control1 Phy_Rst_Control2 Phy_Rst_Control3 VPD read or write area length VPD TAG 1 LENGTH Y Y VPD read or write area TAG VPD TAG 2 Y Y VPD read or write area length VPD TAG 2 LENGTH Y Y VPD end TAG VPD END Y Y Init Control 3 LAN 0 024 Y Y Init Control 3 LAN 1 LAN1_start + 024 Y Y October 2012 Order Number: 327879-001US SWa reset port 0 Load Y SWa reset port 1 Load Y SWa reset port 2 Load Y SWa reset port 3 Load Y Y Y Y Y Intel(R) Communications Chipset 89xx Series - Datasheet 1015 Table 22-13. EEPROM Auto-Load Sequence (Sheet 3 of 4) Full Load (Power-up) Full Load No MGMT (PCI RST) EEPROM Word EEPROM Word Address Init Control 3 LAN 2 LAN2_start + 024 Y Y Init Control 3 LAN 3 LAN3_start + 024 Y Y SWa reset port 0 Load SWa reset port 1 Load SWa reset port 2 Load SWa reset port 3 Load Init Control 4 LAN 0 013 Y Y Init Control 4 LAN 1 LAN1_start + 013 Y Y Y Init Control 4 LAN 2 LAN2_start + 013 Y Y Init Control 4 LAN 3 LAN3_start + 013 Y Y LEDCTL 0 default LAN 0 01F Y Y LEDCTL 0 default LAN 1 LAN1_start + 01F Y Y LEDCTL 0 2 default LAN 2 LAN2_start + 01F Y Y LEDCTL 0 2 default LAN 3 LAN3_start + 01F Y Y Init Control 2 00F Y Y Y Ethernet address byte 2-1 LAN 0 000 Y Y Y Ethernet address byte 4-3 LAN 0 001 Y Y Y Ethernet address byte 6-5 LAN 0 002 Y Y Y Ethernet address byte 2-1 LAN 1 LAN1_start + 000 Y Y Y Ethernet address byte 4-3 LAN 1 LAN1_start + 001 Y Y Y Ethernet address byte 6-5 LAN 1 LAN1_start + 002 Y Y Y Ethernet address byte 2-1 LAN 2 LAN2_start + 000 Y Y Y Ethernet address byte 4-3 LAN 2 LAN2_start + 001 Y Y Y Ethernet address byte 6-5 LAN 2 LAN2_start + 002 Y Y Y Ethernet address byte 2-1 LAN 3 LAN3_start + 000 Y Y Y Ethernet address byte 4-3 LAN 3 LAN3_start + 001 Y Y Y Ethernet address byte 6-5 LAN 3 LAN3_start + 002 Y Y Y Software defined pins control - LAN0 020 Y Y Software defined pins control - LAN1 LAN1_start + 020 Y Y Software defined pins control - LAN2 LAN2_start + 020 Y Y Software defined pins control - LAN3 LAN3_start + 020 Y Y Y Y Y Y Y Y Y Y Y Y Management Sectiond Pass Through LAN Configuration Pointer LAN0 011 Intel(R) Communications Chipset 89xx Series - Datasheet 1016 Y October 2012 Order Number: 327879-001US 22.0 Table 22-13. EEPROM Auto-Load Sequence (Sheet 4 of 4) EEPROM Word EEPROM Word Address Full Load (Power-up) Pass Through LAN Configuration Pointer LAN1 LAN1_start + 011 Y Pass Through LAN Configuration Pointer LAN2 LAN2_start + 011 Y Pass Through LAN Configuration Pointer LAN3 LAN3_start + 011 Y Management Hardware Config Control 023 Y MNG Capabilities 054 Y Sideband Configuration Pointer 057 Y Firmware patch Pointer 051 Y Watchdog configuration 02Ee Full Load No MGMT (PCI RST) SWa reset port 0 Load SWa reset port 1 Load SWa reset port 2 Load SWa reset port 3 Load HW Section Continued Y Y Y Y Y Y a. Upon assertion of CTRL.DEV_RST by software partial load of parameters relevant to all ports is done. Assertion of CTRL_EXT.EE_RST causes load of per port parameters similar to CTRL.RST. b. Loaded only if load subsystem ID bit is set c. Loaded only if load device ID bit is set d. EEPROM words listed under Management Section are also loaded following Firmware Reset e. Word also loaded following Firmware reset. 22.3.1.4 Software Accesses The controller provides two methods for software access to the EEPROM. It can either use the built-in controller to read the EEPROM or access the EEPROM directly using the EEPROM's 4-wire interface. Software can use the EEPROM Read (EERD) register to cause the controller to read a word from the EEPROM that the software can then use. To do this, software writes the address to read to the Read Address (EERD.ADDR) field and simultaneously writes a 1b to the Start Read bit (EERD.START). The controller reads the word from the EEPROM, sets the Read Done bit (EERD.DONE), and places the data in the Read Data field (EERD.DATA). Software can poll the EEPROM Read register until it sees the Read Done bit set and then uses the data from the Read Data field. Any words read this way are not written to the PCH's internal registers. Software can also directly access the EEPROM's 4-wire interface through the EEPROM Control and Data (EEC) register. It can use this for reads, writes, or other EEPROM operations. To directly access the EEPROM, software should follow these steps: 1. Write a 1b to the EEPROM Request bit (EEC.EE_REQ). 2. Read the EEPROM Grant bit (EEC.EE_GNT) until it becomes 1b. It remains 0b as long as the hardware is accessing the EEPROM. 3. Write or read the EEPROM using the direct access to the 4-wire interface as defined in the EEPROM Control and Data (EEC) register. The exact protocol used depends on the EEPROM placed on the board and can be found in the appropriate datasheet. 4. Write a 0b to the EEPROM Request bit (EEC.EE_REQ) to enable EEPROM access by other drivers. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1017 Finally, software can cause the controller to re-read the per-function hardware accessed fields of the EEPROM (setting the controller's internal registers appropriately similar to software reset) by writing a 1b to the EEPROM Reset bit of the Extended Device Control register (CTRL_EXT.EE_RST). Note: If the EEPROM does not contain a valid signature (see Section 22.3.1.5), the controller assumes 16-bit addressing. In order to access an EEPROM that requires 8-bit addressing, software must use the direct access mode. 22.3.1.5 Signature Field The controller determines if an EEPROM is present by attempting to read it. The controller first reads the EEPROM Sizing word at address 0x12. It checks the signature value for bits 15 and 14. If bit 15 is 0b and bit 14 is 1b, it considers the EEPROM to be present and valid and reads additional EEPROM words and then programs its internal registers based on the values read. Otherwise, it ignores the values it reads from that location and does not read any other words as part of the auto-read process. However, the EEPROM is still accessible to software. In most applications, initial EEPROM programming is done directly on the EEPROM pins. Nevertheless, it is desired to enable existing software utilities (accessing the EEPROM via the host interface) to initially program the entire EEPROM. Following a power-up sequence, the controller reads the hardware initialization words in the EEPROM. If the signature in word 0x12 does not equal 01b, the EEPROM is assumed as non-programmed. There are two effects of a non-valid signature: * The controller does not read any further EEPROM data. * The controller enables access to any location in the EEPROM via the EEPROM CSR registers. 22.3.1.6 EEPROM Recovery The PCH requires EEPROM to boot and therefore must be provided with valid EEPROM before power up. 22.3.2 Shared EEPROM The controller uses a single EEPROM device to configure hardware default parameters for all LAN devices, including Ethernet Individual Addresses (IA), LED behaviors, receive packet filters for manageability, wake-up capability, etc. Certain EEPROM words are used to specify hardware parameters that are LAN device-independent (such as those that affect circuit behavior). Other EEPROM words are associated with a specific LAN device. All LAN devices access the EEPROM to obtain their respective configuration settings. 22.3.2.1 EEPROM Deadlock Avoidance The EEPROM is a shared resource between the following clients: * Hardware auto-read. * Port 0 LAN driver accesses. * Port 1 LAN driver accesses. * Port 2 LAN driver accesses. * Port 3 LAN driver accesses. * Firmware accesses. Intel(R) Communications Chipset 89xx Series - Datasheet 1018 October 2012 Order Number: 327879-001US 22.0 All clients can access the EEPROM using parallel access, where hardware implements the actual access to the EEPROM. Hardware can schedule these accesses so that all clients get served without starvation. However, software and hardware clients can access the EEPROM using bit banging. In this case, there is a request/grant mechanism that locks the EEPROM to the exclusive usage of one client. If this client is stuck (without releasing the lock), the other clients are not able to access the EEPROM. In order to avoid this, the controller implements a timeout mechanism, which releases the grant from a client that didn't toggle the EEPROM bit-bang interface for more than two seconds. The EEPROM deadlock avoidance mechanism is enabled when the Deadlock Timeout Enable bit in the Initialization Control Word 1 EEPROM word is set to 1. Note: If an agent that was granted access to the EEPROM for bit-bang access didn't toggle the bit bang interface for 500 ms, it should check if it still owns the interface before continuing the bit-banging. 22.3.2.2 EEPROM Map Shared Words The EEPROM map in Chapter 24.0, "Non-Volatile Memory Map - EEPROM" identifies those words configuring either LAN devices or the entire controller Quad Port Interface as "all". Those words configuring a specific LAN device parameter are identified by their LAN number. The following EEPROM words warrant additional notes specifically related to quad-LAN support: Table 22-14. Notes on EEPROM Words Initialization Control 1, Initialization Control 2 (shared between LANs) These EEPROM words specify hardware-default values for parameters that apply a single value to all LAN devices, such as link configuration parameters required for auto-negotiation, wake-up settings, PCIe bus advertised capabilities, etc. Initialization Control 3, Initialization Control 4 (unique to each LAN) This EEPROM word configures default values associated with each LAN device's hardware connections, including which link mode (SGMII, SerDes, 1000BASE-KX) is used with this LAN device. Because a separate EEPROM word configures the defaults for each LAN, extra care must be taken to ensure that the EEPROM image does not specify a resource conflict. 22.3.3 Vital Product Data (VPD) Support The EEPROM image might contain an area for VPD. This area is managed by the OEM vendor and doesn't influence the behavior of hardware. Word 0x2F of the EEPROM image contains a pointer to the VPD area in the EEPROM. A value of 0xFFFF means VPD is not supported and the VPD capability doesn't appear in the configuration space. The VPD area should be aligned to a Dword boundary in the EEPROM. The maximum area size is 256 bytes but can be smaller. The VPD block is built from a list of resources. A resource can be either large or small. The structure of these resources is listed in the following tables. Table 22-15. Small Resource Structure Offset Content October 2012 Order Number: 327879-001US 0 1-n Tag = 0xxx, xyyyb (Type = Small(0), Item Name = xxxx, length = yyy bytes) Data Intel(R) Communications Chipset 89xx Series - Datasheet 1019 Table 22-16. Large Resource Structure Offset Content 0 1-2 3-n Tag = 1xxx, xxxxb (Type = Large(1), Item Name = xxxxxxx) Length Data The controller parses the VPD structure during the auto-load process (power up and PCIe reset or warm reset) in order to detect the read-only and read/write area boundaries. The controller assumes the following VPD structure: Table 22-17. VPD Structure Tag Structure Type Length (Bytes) Data 0x82 Large Length of identifier string Identifier Identifier string. 0x90 Large Length of RO area RO data VPD-R list containing one or more VPD keywords This part is optional and might not appear. 0x91 Large Length of R/W area RW data VPD-W list containing one or more VPD keywords. This part is optional and might not appear. 0x78 Small N/A N/A Note: Resource Description End tag. The VPD-R and VPD-W structures can be in any order. If the controller doesn't detect a value of 0x82 in the first byte of the VPD area, or the structure doesn't follow the description listed in Table 22-17, it assumes the area is not programmed and the entire 256 bytes area is read only. If a VPD-W tag is found after the VPD-R tag, the area defined by it's size is writable via the VPD structure. See the PCI 3.0 specification (Appendix I) for details of the different tags. In any case, the VPD area is accessible for read and write via the regular EEPROM mechanisms. The VPD area can be accessed through the PCIe configuration space VPD capability structure. Write accesses to a read-only area or any access outside of the VPD area via this structure are ignored. Note: Write access to Dwords, which are only partially in the read/write area, are ignored. It is responsibility of VPD software to make the right alignment to enable a write to the entire area. 22.4 Configurable I/O Pins 22.4.1 General-Purpose I/O (Software-Definable Pins) The controller has two software-defined pins (SDP pins) per port that can be used for miscellaneous hardware or software-controllable purposes. These pins and their function are bound to a specific LAN device. For example, eight SDP pins cannot be associated with a single LAN device. These pins can each be individually configurable to act as either input or output pins. The default direction of each of the four pins is configurable via the EEPROM as well as the default value of any pins configured as outputs. To avoid signal contention, both pins are set as input pins until after the EEPROM configuration has been loaded. In addition to both pins being individually configurable as inputs or outputs, they can be configured for use as General-Purpose Interrupt (GPI) inputs. To act as GPI pins, the desired pins must be configured as inputs. A separate GPI interrupt-detection enable is Intel(R) Communications Chipset 89xx Series - Datasheet 1020 October 2012 Order Number: 327879-001US 22.0 then used to enable rising-edge detection of the input pin (rising-edge detection occurs by comparing values sampled at the internal clock rate as opposed to an edge-detection circuit). When detected, a corresponding GPI interrupt is indicated in the Interrupt Cause register. The use, direction, and values of SDP pins are controlled and accessed using fields in the Device Control (CTRL) register and Extended Device Control (CTRL_EXT) register. The SDPs can be used for special purpose mechanism such as watchdog indication (see Section 22.4.2 for details) or IEEE 1588 support. 22.4.2 Software Watchdog In some situations it might be useful to give an indication to manageability firmware or to external devices that the controller hardware or the software device driver is not functional. For example, in a pass-through NIC, the controller might be bypassed if it is not functional. In order to provide this functionality, a watchdog mechanism is used. This mechanism can be enabled by default, according to EEPROM configuration. Once the host driver is up and it determines that hardware is functional, it might reset the watchdog timer to indicate that the controller is functional. The software device driver then should re-arm the timer periodically. If the timer is not re-armed after pre-programmed timeout (such as the timer runs out before it is reset), an interrupt is given to firmware and a pre-programmed SDP (either SDP0[0] or SDP1[0]) is raised. The SDP indication is shared between the ports. The register controlling this feature is WDSTP. This register enables the setting of a time-out period and the activation of this mode. Both get their default from the EEPROM. The re-arming of the timer is done by setting the WDSWSTS.Dev_functional bit. If software needs to trigger the watchdog immediately because it suspects hardware is stuck, it can set the WDSWSTS.Force_WD bit. It can also give firmware an indication of the watchdog reason for using the WDSWSTS.stuck_reason field. The SDP on which the watchdog indication is indicated, is set using the CTRL.SDP0_WDE bit. In this mode the CTRL.SDP0_IODIR should be set to output. The CTRL.SDP0_DATA bit indicates the polarity of the indication. Setting this bit in one of the cores causes the watchdog indications of all ports to be indicated on this SDP. 22.4.2.1 Watchdog Rearm After a watchdog indication was received, in order to rearm the mechanism the following flow should be used: 1. Clear WD_enable bit in the WDSTP register. 2. Clear SDP0_WDE bit in CTRL register. 3. Set SDP0_WDE bit in CTRL register. 4. Set WD_enable in the WDSTP register. 22.4.3 LEDs The controller provides one LED per port that can be used to indicate different statuses of the traffic. The default setup of the LEDs is done via EEPROM word offsets 0x1C and 0x1F from start of relevant LAN port section (LaN port 0, 1, 2 and 3). This setup is reflected in the LEDCTL register of each port. Each software device driver can change its setup individually. For each of the LEDs the following parameters can be defined: October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1021 * Mode: Defines which information is reflected by this LED. The encoding is described in the LEDCTL register. * Polarity: Defines the polarity of the LED. * Blink mode: Determines whether or not the LED should blink or be stable. In addition, the blink rate of all LEDs can be defined. The possible rates are 200 ms or 83 ms for each phase. There is one rate for all the LEDs of a port. 22.5 Network Interfaces 22.5.1 Overview The controller MAC provides a complete CSMA/CD function supporting IEEE 802.3 (10 Mb/s), 802.3u (100 Mb/s), 802.3z and 802.3ab (1000 Mb/s) implementations. The controller performs all of the functions required for transmission, reception, and collision handling called out in the standards. Each controller MAC can be configured to use a different media interface. The controller supports the following potential configurations: * External SerDes device such as an optical SerDes (SFP or on board) or backplane (1000BASE-BX or 1000BASE-KX) connections. * External SGMII device. This mode is used for connections to external 10/100/1000 BASE-T PHYs that support the SGMII MAC interface. Selection between the various configurations is programmable via each MAC's Extended Device Control register (CTRL_EXT.LINK_MODE bits) and default is set via EEPROM settings. Table 22-18 lists the encoding on the LINK_MODE field for each of the modes. Table 22-18. Encoding on LINK_MODE Field for Each Mode Link Mode 000b GBE Controller Mode reserved 001b 1000BASE-KX 010b SGMII 011b SerDes (1000BASE-BX) The GMII/MII interface, used to communicate between the SGMII PCS, supports 10/100/1000 Mb/s operation, with both half- and full-duplex operation at 10/100 Mb/s, and only full-duplex operation at 1000 Mb/s. The SerDes function can be used to implement a fiber-optics-based solution or backplane connection without requiring an external TBI mode transceiver/SerDes. The SerDes interface can be used to connect to SFP modules. As such, this SerDes interface has the following limitations: * No Tx clock * AC coupling only Intel(R) Communications Chipset 89xx Series - Datasheet 1022 October 2012 Order Number: 327879-001US 22.0 22.5.2 MAC Functionality 22.5.2.1 Internal GMII/MII Interface The controller's MAC and PCS communicate through an internal MII interface that can be configured for 10/100 Mb/s (MII) mode of operation. For proper network operation, both the MAC and PHY must be properly configured (either explicitly via software or via hardware auto-negotiation) to identical speed and duplex settings. All MAC configuration is performed using Device Control registers mapped into system memory or I/O space. In addition an external MDIO/MDC interface is available to configure external PHY's that are connected to the controller via the SGMII interface. 22.5.2.2 MDIO/MDC PHY Management Interface The controller implements an IEEE 802.3 MII Management Interface (also known as the Management Data Input/Output or MDIO Interface) between the MAC and a PHY. This interface provides the MAC and software the ability to monitor and control the state of the PHY. The MDIO interface defines a physical connection, a special protocol that runs across the connection, and an internal set of addressable registers. The interface consists of a data line (MDIO) and clock line (MDC), which are accessible by software via the MAC register space. * MDC (management data clock): This signal is used by the PHY as a clock timing reference for information transfer on the MDIO signal. The MDC is not required to be a continuous signal and can be frozen when no management data is transferred. The MDC signal has a maximum operating frequency of 2.5 MHz. * MDIO (management data I/O): This used to transfer control and status information to and from the PHY (to read and write the PHY management registers). Software can use MDIO accesses to read or write registers of an external SGMII PHY, by accessing MDIC register. MDIO configuration setup (External PHY, PHY Address and Shared MDIO) is defined in the MDICNFG register. When working in SGMII/SerDes mode, the external PHY (if it exists) can be accessed either through MDC/MDIO as previously described, or via a two wire I2C interface bus using the I2CCMD register (see Chapter 28.0, "GbE Programming Interface"). The two wire interface bus or the MDC/MDIO bus are connected via the same pins, and thus are mutually exclusive. In order to be able to control an external device, either by I2C or MDC/MDIO, the 2 wires SFP Enable bit in Initialization Control 3 EEPROM word, that's loaded into the CTRL_EXT.I2C Enabled register bit, should be set. Each port has its own MDC/MDIO or two wire interface bus. However, the MDC/MDIO bus of LAN port 0 may be shared by all ports configured to external PHY operation (set to 1), to allow control of a multi PHY chip with a single MDC/MDIO bus. MDIO operation using a shared bus or a separate bus is controlled by the MDICNFG.Com_MDIO bit that's loaded from Initialization Control 3 EEPROM word following reset. The external port PHY Address is written in the MDICNFG.PHYADD register field and is loaded from the Initialization Control 4 EEPROM word following reset. 22.5.2.2.1 Detection of External I2C or MDIO Connection When the CTRL_EXT.I2C Enabled bit is set to 1, Software can recognize type of external PHY control bus (MDIO or I2C) connection according to the values loaded from the EEPROM to the MDICNFG.Destination bit and the CTRL_EXT.LINK_MODE field in the following manner: October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1023 * External I2C operating mode - MDICNFG.Destination equals 0 and CTRL_EXT.LINK_MODE is not equal to 0. * External MDIO Operating mode - MDICNFG.Destination equals 1 and CTRL_EXT.LINK_MODE is not equal to 0. 22.5.2.2.2 MDIC and MDICNFG Register Usage For a MDIO read cycle, the sequence of events is as follows: 1. If default MDICNFG register values loaded from EEPROM need to be updated. The processor performs a PCIe write access to the MDICNFG register to define the: -- PHYADD = Address of external PHY. -- Destination = External PHY. -- Com_MDIO = Shared or separate MDIO external PHY connection. 2. The processor performs a PCIe write cycle to the MDIC register with: -- Ready = 0b -- Interrupt Enable set to 1b or 0b -- Opcode = 10b (read) -- REGADD = Register address of the specific register to be accessed (0 through 31). 3. The MAC applies the following sequence on the MDIO signal to the PHY: <PREAMBLE><01><10><PHYADD><REGADD><Z> where Z stands for the MAC tri-stating the MDIO signal. 4. The PHY returns the following sequence on the MDIO signal: <0><DATA><IDLE>. 5. The MAC discards the leading bit and places the following 16 data bits in the MII register. 6. The controller asserts an interrupt indicating MDIO "Done" if the Interrupt Enable bit was set. 7. The controller sets the Ready bit in the MDIC register indicating the Read is complete. 8. The processor might read the data from the MDIC register and issue a new MDIO command. For a MDIO write cycle, the sequence of events is as follows: 1. If default MDICNFG register values loaded from EEPROM need to be updated. The processor performs a PCIe write cycle to the MDICNFG register to define the: -- PHYADD = Address of external PHY. -- Destination = External PHY. -- Com_MDIO = Shared or separate MDIO external PHY connection. 2. The processor performs a PCIe write cycle to the MDIC register with: -- Ready = 0b. -- Interrupt Enable set to 1b or 0b. -- Opcode = 01b (write). -- REGADD = Register address of the register to be accessed (0 through 31). -- Data = Specific data for desired control of the PHY. Intel(R) Communications Chipset 89xx Series - Datasheet 1024 October 2012 Order Number: 327879-001US 22.0 3. The MAC applies the following sequence on the MDIO signal to the PHY: <PREAMBLE><01><01><PHYADD><REGADD><10><DATA><IDLE> 4. The controller asserts an interrupt indicating MDIO "Done" if the Interrupt Enable bit was set. 5. The controller sets the Ready bit in the MDIC register to indicate that the write operation completed. 6. The CPU might issue a new MDIO command. Note: A MDIO read or write might take as long as 64 s from the processor write to the Ready bit assertion. When a shared MDC/MDIO bus is used, each transaction can take up to 256 s to complete if other ports are using the bus concurrently. If an invalid opcode is written by software, the MAC does not execute any accesses to the PHY registers. If the PHY does not generate a 0b as the second bit of the turn-around cycle for reads, the MAC aborts the access, sets the E (error) bit, writes 0xFFFF to the data field to indicate an error condition, and sets the Ready bit. Note: After a PHY reset, access through the MDIC register should not be attempted for 300 sec. 22.5.3 SerDes, SGMII and 1000BASE-KX Support The controller can be configured to follow either SGMII, Serial-SerDes or 1000BASE-KX standards. When in SGMII mode, the controller can be configured to operate in 1 Gb/s, 100 Mb/s or 10 Mb/s speeds. When in the 10/100 Mb/s speed, the controller can be configured to half-duplex mode of operation. When configured for Serial-SerDes or 1000BASE-KX operation, the port supports only 1 Gb/s, full-duplex operation. Since the serial interfaces are defined as differential signals, internally the hardware has analog and digital blocks. Following is the initialization/configuration sequence for the analog and digital blocks. 22.5.3.1 SerDes, SGMII and 1000BASE-KX PCS Block The link setup for SerDes, 1000BASE-KX and SGMII are described in sections 22.5.4.1, 22.5.4.2 and 22.5.4.3 respectively. 22.5.3.2 GbE Physical Coding Sub-Layer (PCS) The controller integrates the 802.3z PCS function on-chip. The packet encapsulation is based on the Fiber Channel (FC0/FC1) physical layer and uses the same coding scheme to maintain transition density and DC balance. The physical layer device is the SerDes and is used for 1000BASE-SX, -L-, or -CX configurations. 22.5.3.2.1 8B10B Encoding/Decoding The GbE PCS circuitry uses the same transmission-coding scheme used in the fiber channel physical layer specification. The 8B10B-coding scheme was chosen by the standards committee in order to provide a balanced, continuous stream with sufficient transition density to allow for clock recovery at the receiving station. There is a 25% overhead for this transmission code, which accounts for the data-signaling rate of 1250 Mb/s with 1000 Mb/s of actual data. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1025 22.5.3.2.2 Code Groups and Ordered Sets Code group and ordered set definitions are defined in clause 36 of the IEEE 802.3z standard. These represent special symbols used in the encapsulation of GbE packets. The following table contains a brief description of defined ordered sets and included for informational purposes only. See clause 36 of the IEEE 802.3z specification for more details. Table 22-19. Ordered Sets Code Ordered_Set # of Code Groups Usage /C/ Configuration 4 General reference to configuration ordered sets, either /C1/ or /C2/, which is used during auto-negotiation to advertise and negotiate link operation information between link partners. Last 2 code groups contain configuration base and next page registers. /C1/ Configuration 1 4 See /C/. Differs from /C2/ in 2nd code group for maintaining proper signaling disparitya. /C2/ Configuration 2 4 See /C/. Differs from /C1/ in 2nd code group for maintaining proper signaling disparity1. /I/ IDLE 2 General reference to idle ordered sets. Idle characters are continually transmitted by the end stations and are replaced by encapsulated packet data. The transitions in the idle stream enable the SerDes to maintain clock and symbol synchronization between link partners. /I1/ IDLE 1 2 See /I/. Differs from /I2/ in 2nd code group for maintaining proper signaling disparity1. /I2/ IDLE 2 2 See /I/. Differs from /I1/ in 2nd code group for maintaining proper signaling disparity1. /R/ Carrier_Extend 1 This ordered set is used to indicate carrier extension to the receiving PCS. It is also used as part of the end_of_packet encapsulation delimiter as well as IPG for packets in a burst of packets. /S/ Start_of_Packet 1 The SPD (start_of_packet delimiter) ordered set is used to indicate the starting boundary of a packet transmission. This symbol replaces the last byte of the preamble received from the MAC layer. /T/ End_of_Packet 1 The EPD (end_of_packet delimiter) is comprised of three ordered sets. The /T/ symbol is always the first of these and indicates the ending boundary of a packet. /V/ Error_Propagation 1 The /V/ ordered set is used by the PCS to indicate error propagation between stations. This is normally intended to be used by repeaters to indicate collisions. a. The concept of running disparity is defined in the standard. In summary, this refers to the 1-0 and 0-1 transitions within 8B10B code groups. 22.5.4 Auto-Negotiation and Link Setup Features The method for configuring the link between two link partners is highly dependent on the mode of operation as well as the functionality provided by the specific physical layer device (SerDes). In SerDes mode, the PCH provides the complete PCS and Auto-negotiation functionality as defined in IEEE802.3 clause 36 and clause 37.In SGMII mode, the PCH supports the SGMII link auto-negotiation process, whereas the link auto-negotiation, as defined in is done by the external PHY. In 1000BASE-KX mode, the PCH supports only parallel detect of 1000BASE-KX signaling and does not support the full Auto-Negotiation for Backplane Ethernet protocol as defined in IEEE802.3ap clause 73. Configuring the link can be accomplished by several methods ranging from software forcing link settings, software-controlled negotiation, MAC-controlled auto-negotiation, to auto-negotiation initiated by a PHY. The following sections describe processes of bringing the link up including configuration of the PCH and the transceiver, as well as the various methods of determining duplex and speed configuration. Intel(R) Communications Chipset 89xx Series - Datasheet 1026 October 2012 Order Number: 327879-001US 22.0 The process of determining link configuration differs slightly based on the specific link mode (external SerDes, SGMII or 1000BASE-KX) being used. When operating in a SerDes mode, the PCS layer performs auto-negotiation per clause 37 of the 802.3z standard. The transceiver used in this mode does not participate in the auto-negotiation process as all aspects of auto-negotiation are controlled by the PCH. When operating in SGMII mode, the PCS layer performs SGMII auto-negotiation per the SGMII specification. The external PHY is responsible for the Ethernet auto-negotiation process. When operating in 1000BASE-KX mode the PCH performs parallel detect of 1000BASE-KX operation but does not implement the full Auto-Negotiation for Backplane Ethernet sequence as defined in IEEE802.3ap clause 73. 22.5.4.1 SerDes Link Configuration When using SerDes link mode, link mode configuration can be performed using the PCS function in the PCH. The hardware supports both hardware and software auto-negotiation methods for determining the link configuration, as well as allowing for a manual configuration to force the link. Hardware auto-negotiation is the preferred method. 22.5.4.1.1 Signal Detect Indication The SRDS_0/1/2/3_SIG_DET pins can be connected to a Signal Detect or loss-of-signal (LOS) output that indicates when no laser light is being received when the PCH is used in a 1000BASE-SX or -LX implementation (SerDes operation). It prevents false carrier cases occurring when transmission by a non connected port couples in to the input. Unfortunately, there is no standard polarity for this signal coming from different manufacturers. The CTRL.ILOS bit provides for inversion of the signal from different external optical module vendors, and should be set when the external optical module provides a negative-true loss-of-signal. 22.5.4.1.2 MAC Link Speed SerDes operation is only defined for 1000 Mb/s operation. Other link speeds are not supported. When configured for the SerDes interface, the MAC speed-determination function is disabled and the Device Status register bits (STATUS.SPEED) indicate a value of 10b for 1000 Mb/s. 22.5.4.1.3 SerDes Mode Auto-Negotiation In SerDes mode, after power up or reset via PCIE_EP_RST#, the controller initiates IEEE802.3 clause 37 auto-negotiation based on the default settings in the device control and transmit configuration or PCS Link Control Word registers, as well as settings read from the EEPROM. If enabled in the EEPROM, the PCH immediately performs auto-negotiation. TBI mode auto-negotiation, as defined in clause 37 of the IEEE 802.3z standard, provides a protocol for two devices to advertise and negotiate a common operational mode across a GbE link. The controller fully supports the IEEE 802.3z auto-negotiation function when using the on-chip PCS and internal SerDes. TBI mode auto-negotiation is used to determine the following information: * Duplex resolution (even though the PCH MAC only supports full-duplex in SerDes mode). * Flow control configuration. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1027 Note: Since speed for SerDes modes is fixed at 1000 Mb/s, speed settings in the Device Control register are unaffected by the auto-negotiation process. Note: Auto-negotiation can be initiated at power up or by asserting PCIE_EP_RST# and enabling specific bits in the EEPROM. The auto-negotiation process is accomplished by the exchange of /C/ ordered sets that contain the capabilities defined in the PCS_ANADV register in the 3rd and 4th symbols of the ordered sets. Next page are supported using the PCS_NPTX_AN register. Bits FD and LU in the Device Status (STATUS) register, and bits in the PCS_LSTS register provide status information regarding the negotiated link. Auto-negotiation can be initiated by the following: * PCS_LCMD.AN_ENABLE transition from 0b to 1b * Receipt of /C/ ordered set during normal operation * Receipt of a different value of the /C/ ordered set during the negotiation process * Transition from loss of synchronization to synchronized state (if AN_ENABLE is set). * PCS_LCMD.AN_RESTART transition from 0b to 1b Resolution of the negotiated link determines device operation with respect to flow control capability and duplex settings. These negotiated capabilities override advertised and software-controlled device configuration. Software must configure the PCS_ANADV fields to the desired advertised base page. The bits in the Device Control register are not mapped to the txConfigWord field in hardware until after auto-negotiation completes. Table 22-20 lists the mapping of the PCS_ANADV fields to the Config_reg Base Page encoding per clause 37 of the standard. Table 22-20. 802.3z Advertised Base Page Mapping 15 14 13:12 11:9 8:7 6 5 4:0 Nextp Ack RFLT rsv ASM Hd Fd rsv The partner advertisement can be seen in the PCS_ LPAB and PCS_ LPABNP registers. 22.5.4.1.4 Forcing Link-up in SerDes Mode Forcing link can be accomplished by software by writing a 1b to CTRL.SLU, which forces the MAC PCS logic into a link-up state (enables listening to incoming characters when SRDS_[n]_SIG_DET is asserted by the external optical module. Note: The PCS_LCMD.AN_ENABLE bit must be set to a logic zero to enable forcing link. Note: When link is forced via the CTRL.SLU bit, the link does not come up unless the SRDS_[n]_SIG_DET signal is asserted or an internal energy indication is received from the SerDes receiver, implying that there is a valid signal being received by the optical module or SerDes circuitry. The source of the signal detect is defined by the ENRGSRC bit in the CONNSW register. 22.5.4.1.5 HW Detection of Non-Auto-Negotiation Partner Hardware can detect a SerDes link partner that sends idle code groups continuously, but does not initiate or answer an auto-negotiation process. In this case, hardware initiates an auto-negotiation process, and if it fails after some timeout, a link up is Intel(R) Communications Chipset 89xx Series - Datasheet 1028 October 2012 Order Number: 327879-001US 22.0 assumed. To enable this functionality the PCS_LCTL.AN_TIMEOUT_EN bit should be set. This mode can be used instead of the force link mode as a way to support a partner that do not support auto-negotiation. 22.5.4.2 1000BASE-KX Link Configuration When using 1000BASE-KX link mode, link mode configuration is forced manually by software since the PCH does not support IEEE802.3 clause 73 backplane auto-negotiation. 22.5.4.2.1 MAC Link Speed 1000BASE-KX operation is only defined for 1000 Mb/s operation. Other link speeds are not supported. When configured for the 1000BASE-KX interface, the MAC speed-determination function is disabled and the Device Status register bits (STATUS.SPEED) indicate a value of 10b for 1000 Mb/s. 22.5.4.2.2 1000BASE-KX Auto-Negotiation The controller only supports parallel detection of the 1000BASE-KX link and does not support the full IEEE802.3ap clause 73 backplane auto-negotiation protocol. 22.5.4.2.3 Forcing Link-up in 1000BASE-KX Mode In 1000BASE-KX mode (EXT_CTRL.LINK_MODE = 001b) the controller always operates in force link mode. The MAC PCS logic is placed in a link-up state once energy indication, implying that a valid signal is being received by the 1000BASE-KX circuitry, is received. When in the link-up state PCS logic can lock on incoming characters. Note: The PCS_LCMD.AN_ENABLE bit has no effect when in 1000BASE-KX mode. Auto-negotiation is always disabled on ports operating in 1000BASE-KX mode. The source of the signal detect in 1000BASE-KX is internal and the CONNSW.ENRGSRC bit has no effect. In 1000BASE-KX mode the PCH always operates in force link mode and the value of the CTRL.SLU bit has no effect on link-up detection 22.5.4.2.4 1000BASE-KX HW Detection of Link Partner In 1000BASE-KX mode, hardware detects a 1000BASE-KX link partner that sends idle or none idle code groups continuously. In 1000BASE-KX operation force link-up mode is used. 22.5.4.3 SGMII Link Configuration When working in SGMII mode, the actual link setting is done by the external PHY and is dependent on the settings of this PHY. The SGMII auto-negotiation process described in the sections that follow is only used to establish the MAC/PHY connection. 22.5.4.3.1 SGMII Auto-Negotiation This auto-negotiation process is not dependent on the SRDS_[n]_SIG_DET signal, as this signal indicates optical module signal detection and is not relevant in SGMII mode. The outcome of this auto-negotiation process includes the following information: * Link status * Speed * Duplex October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1029 This information is used by hardware to configure the MAC, when operating in SGMII mode. Bits FD and LU of the Device Status (STATUS) register and bits in the PCS_LSTS register provide status information regarding the negotiated link. Auto-negotiation can be initiated by the following: * PCS_LCMD.AN_ENABLE transition from 0b to 1b. * Receipt of /C/ ordered set during normal operation. * Receipt of different value of the /C/ ordered set during the negotiation process. * Transition from loss of synchronization to a synchronized state (if AN_ENABLE is set). * PCS_LCMD.AN_RESTART transition from 0b to 1b. Auto-negotiation determines the controller operation with respect to speed and duplex settings. These negotiated capabilities override advertised and software controlled device configuration. When working in SGMII mode, there is no need to set the PCAS_ANADV register, as the MAC advertisement word is fixed. The result of the SGMII level auto-negotiation can be read from the PCS_LPAB register. 22.5.4.3.2 Forcing Link in SGMII Mode In SGMII, forcing of the link cannot be done at the PCS level, only in the external PHY. The forced speed and duplex settings are reflected by the SGMII auto-negotiation process; the MAC settings are automatically done according to this functionality. 22.5.4.3.3 MAC Speed Resolution The MAC speed and duplex settings are always set according to the SGMII auto-negotiation process. 22.5.4.4 Loss of Signal/Link Status Indication For all modes of operation, an LOS/LINK signal provides an indication of physical link status to the MAC. When the MAC is configured for optical SerDes mode, the input reflects loss-of-signal connection from the optics. In backplane mode, where there is no LOS external indication, an internal indication from the SerDes receiver can be used. In SFP systems the LOS indication from the SFP can be used. Assuming that the MAC has been configured with CTRL.SLU=1b, the MAC status bit STATUS.LU, when read, generally reflects whether SerDes has link. When loss-of-signal asserted from the SerDes is asserted, the MAC considers this to be a transition to a link-down situation (such as cable unplugged, loss of link partner, etc.). If the Link Status Change (LSC) interrupt is enabled, the MAC generates an interrupt to be serviced by the software device driver. 22.5.5 Ethernet Flow Control (FC) The controller supports flow control as defined in 802.3x as well as the specific operation of asymmetrical flow control defined by 802.3z. Flow control is implemented as a means of reducing the possibility of receive buffer overflows, which result in the dropping of received packets, and allows for local controlling of network congestion levels. This can be accomplished by sending an indication to a transmitting station of a nearly full receive buffer condition at a receiving station. Intel(R) Communications Chipset 89xx Series - Datasheet 1030 October 2012 Order Number: 327879-001US 22.0 The implementation of asymmetric flow control allows for one link partner to send flow control packets while being allowed to ignore their reception. For example, not required to respond to PAUSE frames. The following registers are defined for the implementation of flow control: * CTRL.RFCE field is used to enable reception of legacy flow control packets and reaction to them. * CTRL.TFCE field is used to enable transmission of legacy flow control packets. * Flow Control Address Low, High (FCAL/H) - 6-byte flow control multicast address * Flow Control Type (FCT) 16-bit field to indicate flow control type * Flow Control bits in Device Control (CTRL) register - Enables flow control modes. * Discard PAUSE Frames (DPF) and Pass MAC Control Frames (PMCF) in RCTL controls the forwarding of control packets to the host. * Flow Control Receive Threshold High (FCRTH) - A 13-bit high watermark indicating receive buffer fullness. A single watermark is used in link FC mode. * Flow Control Receive Threshold Low (FCRTL) - A 13-bit low watermark indicating receive buffer emptiness. A single watermark is used in link FC mode. * Flow Control Transmit Timer Value (FCTTV) - a set of 16-bit timer values to include in transmitted PAUSE frame. A single timer is used in Link FC mode. * Flow Control Refresh Threshold Value (FCRTV) - 16-bit PAUSE refresh threshold value October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1031 22.5.5.1 MAC Control Frames and Receiving Flow Control Packets 22.5.5.1.1 Structure of 802.3X FC Packets Three comparisons are used to determine the validity of a flow control frame: 1. A match on the 6-byte multicast address for MAC control frames or to the station address of the controller (Receive Address Register 0). 2. A match on the type field. 3. A comparison of the MAC Control Op-Code field. The 802.3x standard defines the MAC control frame multicast address as 01-80-C2-00-00-01. The Type field in the FC packet is compared against an IEEE reserved value of 0x8808. The final check for a valid PAUSE frame is the MAC control op-code. At this time only the PAUSE control frame op-code is defined. It has a value of 0x0001. Frame-based flow control differentiates XOFF from XON based on the value of the PAUSE timer field. Non-zero values constitute XOFF frames while a value of zero constitutes an XON frame. Values in the Timer field are in units of pause quantum (slot time). A pause quantum lasts 64 byte times, which is converted in absolute time duration according to the line speed. Note: XON frame signals the cancellation of the pause from initiated by an XOFF frame pause for zero pause quantum. Table 22-21 lists the structure of a 802.3X FC packet. Table 22-21. 802.3X Packet Format DA 01_80_C2_00_00_01 (6 bytes) SA Port MAC address (6 bytes) Type 0x8808 (2 bytes) Op-code 0x0001 (2 bytes) Time XXXX (2 bytes) Pad 42 bytes CRC 4 bytes 22.5.5.1.2 Operation and Rules The controller operates in Link FC. * Link FC is enabled by the RFCE bit in the CTRL Register. Note: Link flow control capability is negotiated between link partners via the auto negotiation process. It is the software device driver responsibility to reconfigure the link flow control configuration after the capabilities to be used where negotiated as it might modify the value of these bits based on the resolved capability between the local device and the link partner. Once the receiver has validated receiving an XOFF, or PAUSE frame, the PCH performs the following: * Increments the appropriate statistics register(s) * Sets the Flow_Control State bit in the FCSTS0 register. Intel(R) Communications Chipset 89xx Series - Datasheet 1032 October 2012 Order Number: 327879-001US 22.0 * Initializes the pause timer based on the packet's PAUSE timer field (overwriting any current timer's value) * Disables packet transmission or schedules the disabling of transmission after the current packet completes. Resumption of transmission might occur under the following conditions: * Expiration of the PAUSE timer * Reception of an XON frame (a frame with its PAUSE timer set to 0b) Both conditions clear the relevant Flow_Control State bit in the relevant FCSTS0 register and transmission can resume. Hardware records the number of received XON frames. 22.5.5.1.3 Timing Considerations When operating at 1 Gb/s line speed, the controller must not begin to transmit a (new) frame more than two pause-quantum-bit times after receiving a valid link XOFF frame, as measured at the wires. A pause quantum is 512-bit times. When operating in full duplex at 100 Mb/s speeds, the controller must not begin to transmit a (new) frame more than 576-bit times after receiving a valid link XOFF frame, as measured at the wire. 22.5.5.2 PAUSE and MAC Control Frames Forwarding Two bits in the Receive Control register, control forwarding of PAUSE and MAC control frames to the host. These bits are Discard PAUSE Frames (DPF) and Pass MAC Control Frames (PMCF): * The DPF bit controls forwarding of PAUSE packets to the host. * The PMCF bit controls forwarding of non-PAUSE packets to the host. Note: When flow control reception is disabled (CTRL.RFCE = 0), legacy flow control packets are not recognized and are parsed as regular packets. Table 22-22 lists the behavior of the DPF bit. Table 22-22. Forwarding of PAUSE Packet to Host (DPF Bit) RFCE DPF Are FC Packets Forwarded to Host? 0 X Yes. Packets needs to pass the L2 filters (see ).a 1 0 Yes. Packets needs to pass the L2 filters a. The flow control multicast address is not part of the L2 filtering unless explicitly required. Table 22-23 defines the behavior of the PMCF bit. Table 22-23. Transfer of Non-PAUSE Control Packets to Host (PMCF Bit) RFCE PMCF 0 X Yes. Packets needs to pass the L2 filters X 0 Yes. Packets needs to pass the L2 filters October 2012 Order Number: 327879-001US Are Non-FC MAC Control Packets Forwarded to Host? Intel(R) Communications Chipset 89xx Series - Datasheet 1033 22.5.5.3 Transmission of PAUSE Frames The controller generates PAUSE packets to ensure there is enough space in its receive packet buffers to avoid packet drop. The controller monitors the fullness of its receive packet buffers and compares it with the contents of a programmable threshold. When the threshold is reached, the controller sends a PAUSE frame. The controller supports the sending of link Flow Control (FC). Note: Similar to receiving link flow control packets previously mentioned, link XOFF packets can be transmitted only if this configuration has been negotiated between the link partners via the auto-negotiation process or some higher level protocol. The setting of this bit by the software device driver indicates the desired configuration. Note: The transmission of flow control frames should only be enabled in full-duplex mode per the IEEE 802.3 standard. Software should ensure that the transmission of flow control packets is disabled when the PCH is operating in half-duplex mode. 22.5.5.3.1 Operation and Rules Transmission of link PAUSE frames is enabled by software writing a 1b to the TFCE bit in the Device Control register. The content of the Flow Control Receive Threshold High (FCRTH) register determines at what point the controller first transmits a PAUSE frame. The controller monitors the fullness of the receive packet buffer and compares it with the contents of FCRTH. When the threshold is reached, the PCH sends a PAUSE frame with its pause time field equal to FCTTV. At this time, the controller starts counting an internal shadow counter (reflecting the pause timeout counter at the partner end) from zero. When the counter reaches the value indicated in FCRTV register, then, if the PAUSE condition is still valid (meaning that the buffer fullness is still above the high watermark), a XOFF message is sent again. Once the receive buffer fullness reaches the low water mark, the controller sends a XON message (a PAUSE frame with a timer value of zero). Software enables this capability with the XONE field of the FCRTL. The controller sends an additional PAUSE frame if it has previously sent one and the packet buffer overflows. This is intended to minimize the amount of packets dropped if the first PAUSE frame did not reach its target. 22.5.5.3.2 Software Initiated PAUSE Frame Transmission The controller has the added capability to transmit an XOFF frame via software. This is accomplished by software writing a 1b to the SWXOFF bit of the Transmit Control register. Once this bit is set, hardware initiates the transmission of a PAUSE frame in a manner similar to that automatically generated by hardware. The SWXOFF bit is self-clearing after the PAUSE frame has been transmitted. Note: The Flow Control Refresh Threshold mechanism does not work in the case of software-initiated flow control. Therefore, it is the software's responsibility to re-generate PAUSE frames before expiration of the pause counter at the other partner's end. The state of the CTRL.TFCE bit or the negotiated flow control configuration does not affect software generated PAUSE frame transmission. Intel(R) Communications Chipset 89xx Series - Datasheet 1034 October 2012 Order Number: 327879-001US 22.0 Note: Software sends an XON frame by programming a 0b in the PAUSE timer field of the FCTTV register. Software generation of XON packet is not allowed while the hardware flow control mechanism is active, as both use the FCTTV registers for different purposes. Note: XOFF transmission is not supported in 802.3x for half-duplex links. Software should not initiate an XOFF or XON transmission if the PCH is configured for half-duplex operation. Note: When flow control is disabled, pause packets (XON, XOFF, and other FC) are not detected as flow control packets and can be counted in a variety of counters (such as multicast). 22.5.5.4 IPG Control and Pacing The controller supports the following modes of controlling IPG duration: * Fixed IPG - the IPG is extended by a fixed duration 22.5.5.4.1 Fixed IPG Extension The controller allows controlling of the IPG duration. The IPGT configuration field enables an extension of IPG in 4-byte increments. One possible use of this capability is to allow the insertion of bytes into the transmit packet after it has been transmitted by the controller without violating the minimum IPG requirements. For example, a security device connected in series to the controller might add security headers to transmit packets before the packets are transmitted on the network. 22.5.6 Loopback Support 22.5.6.1 General The controller supports the following types of internal loopback in the LAN interfaces: * MAC Loopback (Point 1) * SerDes, SGMII or 1000BASE-KX Loopback (Point 2) By setting the device to loopback mode, packets that are transmitted towards the line will be looped back to the host. The controller is fully functional in these modes, just not transmitting data over the lines. The following figure shows the points of loopback. Figure 22-9. GbE Controller Loopback Modes 1 PCIe October 2012 Order Number: 327879-001US Packet Buffer and DMA 2 GMII MAC SerDes Interface SerDes SGMII Intel(R) Communications Chipset 89xx Series - Datasheet 1035 22.5.6.2 MAC Loopback In MAC loopback, the PHY and SerDes blocks are not functional and data is looped back before these blocks. MAC loopback is operational only when working in PHY mode (CTRL_EXT.LINK_MODE = 000b). 22.5.6.2.1 Setting GbE Controller to MAC Loopback Mode The following procedure puts the controller in MAC loopback mode: * Set RCTL.LBM to 2'b01 (bits 7:6) * Set CTRL.SLU (bit 6, should be set by default) * Set CTRL.FRCSPD & FRCDPLX (bits 11&12) * Set CTRL.SPEED to 2'b10 (1G) and CTRL.FD * Set CTRL.ILOS Filter configuration and other TX/RX processes are the same as in normal mode. 22.5.6.3 SerDes, SGMII and 1000BASE-KX Loopback In SerDes, SGMII, or 1000BASE-KX loopback mode, data is looped back at the end of the relevant functionality. All of the design that is functional in SerDes/SGMII or 1000BASE-KX mode is involved in the loopback. Note: SerDes loopback is functional only if the SerDes link is up. 22.5.6.3.1 Setting GbE Controller to SerDes, SGMII, or 1000BASE-KX Loopback Mode The following procedure places the GbE controller in SerDes loopback mode: * Set MPHY to support SerDes Near End Digital Loopback: -- GBE_MPHY_ADDR_CTRL[15:0] (CSR 0x0024 BITS 15:0) = 16'h0004 -- GBE_MPHY_DATA.DATA (CSR 0x0E10 BITS 31:0) = 32'h0000_0070 -- GBE_MPHY_ADDR_CTRL[15:0] (CSR 0x0024 BITS 15:0) = 16'h0008 -- GBE_MPHY_DATA.DATA (CSR 0x0E10 BITS 31:0) = 32'h0000_0800 -- GBE_MPHY_CTRL.MISC_CTRL (CSR 0x0E08 BITS 30:28) = 3'b010 -- MANC.RCV_TCO_EN (CSR 0x5820 BIT 17) = 0 -- RCTL.LBM (CSR 0x0100 BITS 7:6) = 3 * Set Link mode to either SerDes, SGMII or 1000BASE-KX by: -- 1000BASE-KX: CTRL_EXT.LINK_MODE (CSR 0x18 BITS 24:22) = 001b -- SGMII: CTRL_EXT.LINK_MODE (CSR 0x18 BITS 24:22) = 010b -- SerDes: CTRL_EXT.LINK_MODE (CSR 0x18 BITS 24:22) = 011b * Configure SERDES to loopback: RCTL.LBM = 11b * Move to Force mode by setting the following bits: -- CTRL.FD (CSR 0x0 bit 0) = 1 -- CTRL.SLU (CSR 0x0 bit 6) = 1 -- CTRL.RFCE (CSR 0x0 bit 27) = 0 -- CTRL.TFCE (CSR 0x0 bit 28) = 0 -- PCS_LCTL.FORCE_LINK (CSR 0X4208 bit 5) = 1 Intel(R) Communications Chipset 89xx Series - Datasheet 1036 October 2012 Order Number: 327879-001US 22.0 -- PCS_LCTL.FSD (CSR 0X4208 bit 4) = 1 -- PCS_LCTL.FDV (CSR 0X4208 bit 3) = 1 -- PCS_LCTL.FLV (CSR 0X4208 bit 0) = 1 -- PCS_LCTL.AN_ENABLE (CSR 0X4208 bit 16) = 0 -- CONNSW.ENRGSRC (CSR 0X0034 bit 2) = 0 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1037 23.0 GbE Initialization 23.1 Power Up 23.1.1 Power-Up Sequence The following figure shows the power-up sequence from power ramp up and to when the GbE Controller is ready to accept host commands. Figure 23-1. Power-Up - General Flow Vcc power on HW operation FW operation LAN_PW R_GO OD reset Load EEPROM PE_RST_n reset Initialize FW Reset MAC Load EEPROM Configure MAC Initialize PCI-E (For CC, it done by EP) Yes Initialize Managem ent Link D 0 m ode Run Manageability FW Dr mode No Platform powered ? Intel(R) Communications Chipset 89xx Series - Datasheet 1038 October 2012 Order Number: 327879-001US 23.0 23.1.2 Power-Up Timing Diagram Figure 23-2. Power-Up Timing Diagram Power txog 1 Xosc tppg Power-On-Reset (internal) 2 tPVPGL PCI Express reference clock 7 tPW RGD-CLK 6 PERST# 8 3 tee Auto Read NVM Load tee Auto Read Ext. Conf. tpgtrn 4 PHY State Ext. Conf. 9 Powered-down tpgcfg Active / Down 10 tpgres 11 PCI Express Link up 12 13 L0 5 Manageability / Wake D-State Dr D0u D0a Table 23-1. Notes to Power-Up Timing Diagram Note Description 1 Xosc is stable txog after the power is stable 2 Internal Reset is released after all power supplies are good and tppg after Xosc is stable. 3 An NVM read starts on the rising edge of the internal Reset or GBE_AUX_PWR_OK. 4 After reading the NVM, PHY might exit power down mode. 5 APM Wakeup and/or manageability might be enabled based on NVM contents. 6 The PCIe reference clock is valid tPE_RST-CLK before the de-assertion of PCIE_EP_RST# (according to PCIe spec). 7 PCIE_EP_RST# is de-asserted tPVPGL after power is stable (according to PCIe spec). 8 De-assertion of PCIE_EP_RST# causes the NVM to be re-read, and disables Wake Up. 9 After reading the NVM, PHY exits power-down mode. 10 Link training starts after tpgtrn from PCIE_EP_RST# de-assertion. 11 A first PCIe configuration access might arrive after tpgcfg from PCIE_EP_RST# de-assertion. 12 A first PCI configuration response can be sent after tpgres from PCIE_EP_RST# de-assertion 13 Writing a 1 to the Memory Access Enable bit in the PCI Command Register transitions the device from D0u to D0 state. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1039 23.2 Reset Operation 23.2.1 Hardware-Based Reset Sources The Controller reset sources are described below. 23.2.1.1 GBE_AUX_PWR_OK This reset comes directly from the GbE Cluster GBE_AUX_PWR_OK pin. This reset will cause all sticky and non-sticky registers to be cleared and FSM/queues to reset to their initial value. This reset acts as a master reset of the entire Controller function. It is level sensitive, and while it is zero holds all of the registers in reset. 23.2.1.2 PCIE_EP_RST# The de-assertion of PCIE_EP_RST# indicates that both the power and the PCIe clock sources are stable. This pin asserts an internal reset also after a D3cold exit. Most units are reset on the rising edge of PCIE_EP_RST#. 23.2.1.3 Function Level Reset (FLR) A FLR reset to a function is issued, by setting bit 15 in the Device Control configuration register, is equivalent to a D0 D3 D0 transition. The only difference is that this reset does not require driver intervention in order to stop the master transactions of this function. This reset is per function and resets only the function without effecting activity of other functions or Lan ports. The EEPROM is partially reloaded after an FLR reset. The words read from EEPROM at FLR are the same as read following a full software reset. A list of these words can be found in Chapter 22.0, "GbE Interconnects". A FLR reset to a function resets all the queues, interrupts, and statistics registers attached to the function. It also resets PCIe R/W configuration bits as well as disables transmit and receive flows for the queues allocated to the function. All pending read requests are dropped and PCIe read completions to the function might be completed as unexpected completions and silently discarded (following update of flow control credits) without logging or signaling as an error. Note: If software initiates a FLR when the Transactions Pending bit in the Device Status configuration register is set to 1b, then software must not initialize the function until allowing time for any associated completions to arrive. The Transactions Pending bit is cleared upon completion of the FLR. Intel(R) Communications Chipset 89xx Series - Datasheet 1040 October 2012 Order Number: 327879-001US 23.0 23.3 Software Based Reset Sources 23.3.1 Full Software Reset (DEV_RST and RST) Software can reset the Controller by either setting the Port Software Reset (CTRL.RST) or the Device Reset bit (CTRL.DEV_RST) in the Device Control Register. The Port Software Reset (CTRL.RST) is per function and resets only the function that received the software reset. The Device Reset (CTRL.DEV_RST) resets all functions and common logic. In both cases, PCI configuration space (configuration and mapping) of the device is unaffected. Prior to issuing software reset the driver needs to operate the Bus Master Enable algorithm as defined in Section 23.3.2.2. Port Software Reset (CTRL.RST) can be used to reset the respective port. The RST bit is provided primarily to recover from an indeterminate or suspected Port hung hardware state. Most registers (receive, transmit, interrupt, statistics, etc.) and state machines in the port are set to their power-on reset values, approximating the state following a power-on or PCIe reset. However, PCIe configuration registers and logic common to all ports is not reset, leaving the device mapped into system memory space and accessible by a driver. Note: To ensure that software reset has fully completed and that the Controller responds correctly to subsequent accesses, wait at least 3 milliseconds after setting CTRL.RST before attempting to check if the bit was cleared or before attempting to access any other register. Device Reset (CTRL.DEV_RST) can be used to globally reset the entire component, if the DEV_RST_EN bit in Initialization Control 4 EEPROM word is set. This bit is provided as a last-ditch software mechanism to recover from an indeterminate or suspected hardware hung state that could not be resolved by setting the CTRL.RST bit. When setting CTRL.DEV_RST, most registers (receive, transmit, interrupt, statistics, etc.) and state machines on ports are set to power-on reset values, approximating the state following a power-on or PCI reset. However, PCIe configuration registers are not reset, leaving the device mapped into system memory space and accessible by a driver. When CTRL.DEV_RST is asserted by software on a LAN port, all LAN ports (including LAN ports that didn't initiate the reset) are placed in a reset state. To notify software device drivers on all ports that CTRL.DEV_RST has been asserted, an interrupt is generated and the ICR.DRSTA bit is set on all ports. Following Device Reset assertion or reception of Device Reset interrupt (ICR.DRSTA) software should initiate the following steps to re-initialize the port: 1. Wait for the GCR.DEV_RST in progress bit to be cleared. 2. Issue a port reset by setting the CTL.RST bit, before re-initializing the port. The additional port software reset is required to verify that any Transmit descriptor tail (TDT) updates programmed between device reset and device reset interrupt reception are cleared. 3. Re-initialize the port. When asserting the CTRL.RST software reset bit, only some EEPROM bits related to the specific function are re-read. When setting the CTRL.DEV_RST software reset bit, EEPROM bits related to all ports are re-read (see Section 23.3.2.1). Bits re-read from EEPROM are reset to default values. Fields controlled by the LED, SDP and Init3 words of the EEPROM are not reset and not re-read after a software reset. For the list of words read from EEPROM at full software reset, see Section 23.3.2.1. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1041 23.3.2 BME (Bus Master Enable) Disabling Bus Master activity of a function by clearing the Configuration Command register.BME bit to 0, resets all DMA activities and MSI/MSIx operations related to the port. The Master disable is per function and resets only the DMA activities related to this function without effecting activity of other functions or LAN ports. Configuration accesses and target accesses to the function are still enabled and BMC can still transmit and receive packets on the port. A Master Disable to a function resets all the queues and DMA related interrupts attached to this function. It also disables the transmit and receive flows for the queues allocated to this function. All pending read requests are dropped and PCIe read completions to this function might be completed as unexpected completions and silently discarded (following update of flow control credits) without logging or signaling it as an error. Note: Prior to issuing a master disable the Driver needs to implement the master disable algorithm as defined in Section 25.2.3.2. After Master Enable is set back to 1 driver should re-initialize the transmit and receive queues. The GbE interface implementation recommends using GIO Master Disable algorithm before disabling BME. Flow when GIO Master Disable is used is shown in "Transaction Flow for BME when GIO Master Disable Algorithm is Not Used". Flow when GIO Master Disable is not used is shown in "Transaction Flow for BME when GIO Master Disable Algorithm is Used". 23.3.2.1 Transaction Flow for BME when GIO Master Disable Algorithm is Used 1. Software sets GIO_MASTER_DISABLE. 2. Software waits for some time and checks if GIO_MASTER_ENABLE_Status is Clear. a. Status does not get clear. Go to step 3 b. Status is clear (No problem exists. Perform BME_Disable). If error was received by Gasket in any of the completed transactions, Interrupt will be asserted by Gasket to EP. 3. Software disables BME to the function. a. In response to BME deassertion, EP sends dummy completions for all pending transactions for the affected port. 4. Once GbE receives these dummy completions, it will send completions with bogus data to DMA block for BME-Disabled function. 5. Once all completions for the function have been returned to DMA, GbE Gasket generates GbE_FLR_ACK for the function to EP. This signal GbE_FLR_ACK will play the ACK role for both FLR assertion and BME De-assertion. 6. In future, whenever BME is enabled again for this function, EP will send the pending Interrupt which it might have collected when BME was asserted. 23.3.2.2 Transaction Flow for BME when GIO Master Disable Algorithm is Not Used 1. Software can disable the BME to the function without following GIO_MASTER_DISABLE flow. This is because BME is controlled by OS which may not know about device specific GIO_MASTER_DISABLE register. Same logic applies to FLR where GIO_Master_Disable may not be used. a. In response to BME deassertion, EP sends dummy completions for all pending transactions for the affected port. Intel(R) Communications Chipset 89xx Series - Datasheet 1042 October 2012 Order Number: 327879-001US 23.0 2. Once GbE receives these dummy completions, it will send completions with bogus data to DMA block for BME-Disabled function. 3. Once all completions for the function have been returned to DMA, GbE Gasket generates GbE_FLR_ACK for the function to EP. This signal GbE_FLR_ACK will play the ACK role for both FLR assertion and BME De-assertion. 4. In future, whenever BME is enabled again for this function, EP will send the pending Interrupt which it might have collected when BME was asserted. 23.3.3 Force TCO This reset is generated when manageability logic is enabled and BMC detects that the Controller does not receive or transmit data correctly. Force TCO reset is enabled if the Reset on Force TCO bit in the Management Control EEPROM word is set to 1. Force TCO reset is generated in pass through mode when BMC issues a Force TCO command with bit 1 set and the above conditions exist. 23.3.4 Firmware Reset This reset is activated by writing a 1 to the FWR bit in the HOST Interface Control Register (HICR) in CSR address 0x8F00. 23.3.5 EEPROM Reset Writing a 1 to the EEPROM Reset bit of the Extended Device Control Register (CTRL_EXT.EE_RST) causes the Controller to re-read the per-function configuration from the EEPROM, setting the appropriate bits in the registers loaded by the EEPROM. 23.3.6 External Phy Reset To provide the possibility of using Management Unit to program external PHY on power-on and on PHY reset events, a timer is provided that delays the PHY reset completion signal received by Management Unit. This is to allow enough time between power-on, PHY reset and PHY programming through MDIO/I2C interface. The value of the delay is being programmed in PHY_rst_cmpl_dly CSR through EEPROM on power-on and optionally adjusted by the driver afterwards. The timer counter is halted until the "programming_done" bit in PHY_rst_cmpl_dly is set. This is to prevent signaling MNG until the desired value is set in the CSR. It is required from the Platform Design that external PHY hard reset be derived from the same Power Good signal that is the source for GBE_AUX_POWER_GOOD. This is to provide a common starting point for both external PHY reset and aforementioned timer. For SW initiated PHY reset the driver sets the PHY_RST bit in Device Control CSR, resets the external PHY through MDIO interface and then clears the PHY_RST bit in Device Control CSR. 23.3.7 CSR Access Flow Following Power-on and Reset Events Following PCIE_EP_RST#, In-band, FLR and ACPI(D3->D0) resets, HOST will not be able to access GbE Config Space CSRs until the CSR data is valid. In the response to HOST PCIe CFG requests, PCIe EP will send config retry response until the above signal is deasserted - per function. MMIO CSRs in the GbE can be only accessesed when the ACPI FSM moves to D0a state - as the effect of setting the appropriate CFG space bits in EP CFG Command Register. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1043 Any agent that wants to access GbE CSR following power-up or PCIE_EP_RST#, In-band, FLR and ACPI resets needs to poll the GBECFGMMIOV register, see Section 20.4.2.5, "GBECFGMMIOV--GBE Configuration and MMIO Valid Register", that reflects the validity of the CSR data. Once the CSR data is validated for a given port the agent is free to access GbE CSRs. After issuing a software device reset and software port reset any agent needs to wait for at least 3 milliseconds before it can access the GbE MMIO CSRs. 23.3.8 Reset Effects The resets affect the following registers and logic: Table 23-2. GbE Controller Reset Effects - Common Resets GBE_AUX_ PWR_OK PCIE_EP_ RST# LTSSM (PCIe back to detect/polling) X X PCIe Link data path X X Reset Activation Read EEPROM (Per Function) SW CTRL. DEV_RST In-Band PCIe Reset FW Reset Notes X X X 17. Read EEPROM (Complete Load) X X X PCI Configuration Registers- non sticky X X X 2. PCI Configuration Registers sticky X X X 3. PCIe local registers X X X 4. Data path X X X X On-die memories X X X X MAC, PCS, Auto Negotiation, and other port related logic X X X X 14. DMA X X X X Functions queue enable X X X X Function interrupt & statistics registers X X X X Wake Up (PM) Context X 1. Wake Up Control Register X 7. 6. Wake Up Status Registers X 8. Manageability Control Registers X 9. MMS Unit X Wake-Up Management Registers X X X X 2.,10. Memory Configuration Registers X X X X 2. EEPROM request X PHY/SERDES PHY X X X Strapping Pins X X X Intel(R) Communications Chipset 89xx Series - Datasheet 1044 X X 15. 2. October 2012 Order Number: 327879-001US 23.0 Table 23-3. GbE Controller Reset Effects - Per Function Resets Reset Activation D3hot -> D0 FLR Port SW Reset (CTRL.RST) Force TCO EE Reset X X X X X Read EEPROM (Per Function) PCI Configuration Registers RO Notes 2. PCI Configuration Registers MSI-X X X 5. Data path X X X X On-die memories X X X X MAC, PCS, Auto Negotiation, and other port related logic X X X X DMA X X PCI Configuration Registers RW PCIe local registers 4. 14. 16. Wake Up (PM) Context 6. Wake Up Control Register 7. Wake Up Status Registers 8. Manageability Control Registers 9. Function queue enable X X X Function interrupt & statistics registers X X X Wake-Up Management Registers X X X X 2.,10. Memory Configuration Registers X X X X 2. EEPROM request X X X 15. Strapping Pins Notes: 1. If GBE_AUX_PWR_AVAIL = 0b the Wakeup Context is reset (PME_Status and PME_En bits should be 0b at reset if the Controller does not support PME from D3cold). 2. The following register fields do not follow the general rules above: a. "CTRL.SDP0_IODIR, CTRL.SDP1_IODIR, CTRL_EXT.SDP2_IODIR, CTRL_EXT.SDP3_IODIR, CONNSW.ENRGSRC field, CTRL_EXT.SFP_Enable, CTRL_EXT.LINK_MODE, CTRL_EXT.EXT_VLAN and LED configuration registers are reset on GBE_AUX_PWR_OK only. Any EEPROM read resets these fields to the values in the EEPROM. b. The Aux Power Detected bit in the PCIe Device Status register is reset on GBE_AUX_PWR_OK and PCIE_EP_RST# (PCIe reset) assertion only. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1045 c. FLA - reset on GBE_AUX_PWR_OK only. d. The bits mentioned in the next note. 3. The following registers are part of this group: a. VPD registers b. Max payload size field in PCIe Capability Control register (offset 0xA8). c. Active State Link PM Control field, Common Clock Configuration field and Extended Synch field in PCIe Capability Link Control register (Offset 0xB0). d. Read Completion Boundary in the PCIe Link Control register (Offset 0xB0). 4. The following registers are part of this group: a. SWSM b. GCR (only part of the bits - see register description for details) c. FUNCTAG d. GSCL_1/2/3/4 e. GSCN_0/1/2/3 f. SW_FW_SYNC - only part of the bits - see register description for details. 5. The following registers are part of this group: a. MSIX control register, MSIX PBA and MSIX per vector mask. 6. The Wake Up Context is defined in the PCI Bus Power Management Interface Specification (Sticky bits). It includes: -- PME_En bit of the Power Management Control/Status Register (PMCSR). -- PME_Status bit of the Power Management Control/Status Register (PMCSR). -- Aux_En in the PCIe registers -- The device Requester ID (since it is required for the PM_PME TLP). The shadow copies of these bits in the Wakeup Control Register are treated identically. 7. Refers to bits in the Wake Up Control Register that are not part of the Wake-Up Context (the PME_En and PME_Status bits). 8. Wake Up Status registers: a. Wake Up Status Register b. Wake Up Packet Length. c. Wake Up Packet Memory. 9. Manageability control registers: a. MANC 0x5820 b. MFUTP0-7 0x5030 - 0x504C c. MNGONLY 0x5864 d. MAVTV0-7 0x5010 - 0x502C e. MDEF0-7 0x5890 - 0x58AC f. MDEF_EXT 0x5930 - 0x594C g. METF0-3 0x5060 - 0x506C h. MIPAF0-15 0x58B0 - 0x58EC i. MMAH/MMAL0-1 0x5910 - 0x591C j. FWSM Intel(R) Communications Chipset 89xx Series - Datasheet 1046 October 2012 Order Number: 327879-001US 23.0 10. Wake-up Management registers: a. Wake Up Filter Control b. IP Address Valid c. IPv4 Address Table d. IPv6 Address Table e. Flexible Filter Length Table f. Flexible Filter Mask Table 11. Other configuration registers: a. General Registers b. Interrupt Registers c. Receive Registers d. Transmit Registers e. Statistics Registers f. Diagnostic Registers Of these registers, MTA[n], VFTA[n], WUPM[n], FTFT[n], FHFT[n], FHFT_EXT[n], TDBAH/TDBAL, and RDBAH/RDVAL registers have no default value. If the functions associated with the registers are enabled they must be programmed by software. Once programmed, their value is preserved through all resets as long as power is applied to the Controller. Note: In situations where the device is reset using the software reset CTRL.RST or CTRL.DEV_RST the transmit data lines are forced to all zeros. This causes a substantial number of symbol errors to be detected by the link partner. In TBI mode, if the duration is long enough, the link partner might restart the Auto-Negotiation process by sending "break-link" (/C/ codes with the configuration register value set to all zeros). 12. These registers include: a. MSI/MSI-X enable bits b. BME c. Error indications 13. These registers include: a. RXDCTL.Enable 14. The contents of the following memories are cleared to support the requirements of PCIe FLR: a. The Tx packet buffers b. The Rx packet buffers 15. Includes EEC.REQ, EEC.GNT, FLA.REQ and FLA.GNT fields. 16. The following DMA Registers are cleared only by GBE_AUX_PWR_OK, PCIe Reset or CTRL.DEV_RST: DMCTLX, DTPARS, DRPARS and DDPARS. 17. CTRL.DEV_RST assertion causes read of function related sections for all ports 23.4 Software Initialization and Diagnostics This chapter discusses general software notes for the Controller, especially initialization steps. This includes general hardware, power-up state, basic device configuration, initialization of transmit and receive operation, link configuration, software reset capability, statistics, and diagnostic hints. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1047 23.4.1 Power Up State When the Controller powers up, it reads the EEPROM. The EEPROM contains sufficient information to bring the link up and configure the Controller for manageability and/or APM wakeup. However, software initialization is required for normal operation. The power-up sequence, as well as transitions between power states, are described in section 23.1.1. The detailed timing is given in Section 25.3. The next section gives more details on configuration requirements. 23.4.2 Initialization Sequence The following sequence of commands is typically issued to the device by the software device driver in order to initialize the Controller to normal operation. The major initialization steps are: * Disable Interrupts - see Interrupts during initialization. * Issue Global Reset and perform General Configuration - see Global Reset and General Configuration. * Setup the PHY and the link - see Link Setup Mechanisms and Control/Status Bit Summary. * Initialize all statistical counters - see Initialization of Statistics. * Initialize Receive - see Receive Initialization. * Initialize Transmit - see Transmit Initialization. * Enable Interrupts - see Interrupts during initialization. 23.4.3 Interrupts During Initialization * Most drivers disable interrupts during initialization to prevent re-entering the interrupt routine. Interrupts are disabled by writing to the EIMC (Extended Interrupt Mask Clear) register. The interrupts need to be disabled also after issuing a global reset, so a typical driver initialization flow is: * Disable interrupts * Issue a Global Reset * Disable interrupts (again) * ... After the initialization is done, a typical driver enables the desired interrupts by writing to the EIMS (Extended Interrupt Mask Set) register. Intel(R) Communications Chipset 89xx Series - Datasheet 1048 October 2012 Order Number: 327879-001US 23.0 23.4.4 Global Reset and General Configuration Device initialization typically starts with a global reset that places the device into a known state and enables the device driver to continue the initialization sequence. Several values in the Device Control Register (CTRL) need to be set, upon power up, or after a device reset for normal operation. * FD bit should be set per interface negotiation (if done in software), or is set by the hardware if the interface is Auto-Negotiating. This is reflected in the Device Status Register in the Auto-Negotiation case. * Speed is determined via Auto-Negotiation by the PHY, Auto-Negotiation by the PCS layer in SGMII/SerDes mode, or forced by software if the link is forced. Status information for speed is also readable in the STATUS register. * ILOS bit should normally be set to 0. 23.4.5 Flow Control Setup If flow control is enabled, program the FCRTL0, FCRTH0, FCTTV and FCRTV registers. In order to avoid packet losses, FCRTH should be set to a value equal to at least two max size packet below the receive buffer size. E.g. Assuming a packet buffer size of 32 KB and expected max size packet of 9.5K, the FCRTH0 value should be set to 32 - 2 * 9.5 = 17KB i.e. FCRTH0.RTH should be set to 0x440. 23.4.6 Link Setup Mechanisms and Control/Status Bit Summary The CTRL_EXT.LINK_MODE value should be set to the desired mode prior to the setting of the other fields in the link setup procedures. 23.4.6.1 MAC/SERDES Link Setup (CTRL_EXT.LINK_MODE = 011b) Link setup procedures using an external SERDES interface mode: 23.4.6.1.1 Hardware Auto-Negotiation Enabled (PCS_LCTL. AN ENABLE = 1b) CTRL.FD Ignored; duplex is set by priority resolution of PCS_ANDV and PCS_LPAB. CTRL.SLU Must be set to 1 by software to enable communications to the SerDes. CTRL.RFCE Set by Hardware according to auto negotiation resolution1. CTRL.TFCE Set by Hardware according to auto negotiation resolution1. CTRL.SPEED Ignored; speed always 1000Mb/s when using SerDes mode communications. STATUS.FD Reflects hardware-negotiated priority resolution. STATUS.LU Reflects PCS_LSTS.AN COMPLETE (Auto-Negotiation complete). STATUS.SPEED Reflects 1000Mb/s speed, reporting fixed value of (10)b. PCS_LCTL.FSD Must be zero. PCS_LCTL.Force Flow Control Must be zero1. 1. If PCS_LCTL.Force Flow Control is set, the auto negotiation result is not reflected in the CTRL.RFCE and CTRL.TFCE registers. In This case, the software must set these fields after reading flow control resolution from PCS registers. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1049 23.4.6.1.2 PCS_LCTL.FSV Must be set to 10b. Only 1000 Mb/s is supported in SerDes mode. PCS_LCTL.FDV Ignored; duplex is set by priority resolution of PCS_ANDV and PCS_LPAB. Auto-Negotiation Skipped (PCS_LCTL. AN ENABLE = 0b) CTRL.FD Must be set to 1b. - only full duplex is supported in SerDes mode. CTRL.SLU Must be set to 1 by software to enable communications to the SerDes. CTRL.RFCE Set by software for the desired mode of operation. CTRL.TFCE Set by software for the desired mode of operation. CTRL.SPEED Must be set to 10b. Only 1000 Mb/s is supported in SerDes mode. STATUS.FD Reflects the value written by software to CTRL.FD. STATUS.LU Reflects whether the PCS detected comma symbols, qualified with CTRL.SLU (set to 1). STATUS.SPEED Reflects 1000Mb/s speed, reporting fixed value of (10)b. PCS_LCTL.FSD Must be set to 1 by software to enable communications to the SerDes. PCS_LCTL.Force Flow Control Must be set to 1. 23.4.6.2 PCS_LCTL.FSV Must be set to 10b. Only 1000 Mb/s is supported in SerDes mode. PCS_LCTL.FDV Must be set to 1b - only full duplex is supported in SerDes mode. MAC/SGMII Link Setup (CTRL_EXT.LINK_MODE = 010b) Link setup procedures using an external SGMII interface mode: 23.4.6.2.1 Hardware Auto-Negotiation Enabled (PCS_LCTL. AN ENABLE = 1b, CTRL.FRCDPLX = 0b, CTRL.FRCSPD = 0b) CTRL.FD Ignored; duplex is set by priority resolution of PCS_ANDV and PCS_LPAB. CTRL.SLU Must be set to 1 by software to enable communications to the SerDes. CTRL.RFCE Must be set by software after reading flow control resolution from PCS registers. CTRL.TFCE Must be set by software after reading flow control resolution from PCS registers. CTRL.SPEED Ignored; speed setting is established from SGMII's internal indication to the MAC after SGMII PHY has auto-negotiated a successful link-up. STATUS.FD Reflects hardware-negotiated priority resolution. STATUS.LU Reflects PCS_LSTS.Link OK STATUS.SPEED Reflects actual speed setting negotiated by the SGMII and indicated to the MAC. Intel(R) Communications Chipset 89xx Series - Datasheet 1050 October 2012 Order Number: 327879-001US 23.0 PCS_LCTL.Force Flow Control Ignored. PCS_LCTL.FSD Should be set to zero. PCS_LCTL.FSV Ignored; speed is set by priority resolution of PCS_ANDV and PCS_LPAB. PCS_LCTL.FDV Ignored; duplex is set by priority resolution of PCS_ANDV and PCS_LPAB. 23.4.6.3 MAC/1000BASE-KX Link Setup (CTRL_EXT.LINK_MODE = 001b) 23.4.6.3.1 Auto-Negotiation Skipped (PCS_LCTL. AN ENABLE = 0b; CTRL.FRCSPD = 1b; CTRL.FRCDPLX = 1) Link setup procedures using an external 1000BASE-KX Server Backplane interface mode: CTRL.FD Must be set to 1. 1000BASE-KX always in full duplex mode. CTRL.SLU Must be set to 1 by software to enable communications to the SerDes. CTRL.RFCE Must be set by software for the desired mode of operation. CTRL.TFCE Must be set by software for the desired mode of operation. CTRL.SPEED Must be set to 10b. Only 1000 Mb/s is supported in 1000BASE-KX mode. STATUS.FD Reflects the value written by software to CTRL.FD. STATUS.LU Reflects whether the PCS detected comma symbols, qualified with CTRL.SLU (set to 1). STATUS.SPEED Reflects 1000Mb/s speed, reporting fixed value of (10b). PCS_LCTL.FSD Must be set to 1 by software to enable communications to the 1000BASE-KX SerDes. PCS_LCTL.Force Flow Control 23.4.7 Must be set to 1. PCS_LCTL.FSV Must be set to 10b. Only 1000 Mb/s is supported in 1000BASE-KX mode. PCS_LCTL.FDV Must be set to 1b - only full duplex is supported in 1000BASE-KX mode. Initialization of Statistics Statistics registers are hardware-initialized to values as detailed in each particular register's description. The initialization of these registers begins upon transition to D0 active power state (when internal registers become accessible, as enabled by setting the Memory Access Enable of the PCIe Command register), and is guaranteed to be completed within 1 sec of this transition. Access to statistics registers prior to this interval might return indeterminate values. All of the statistical counters are cleared on read and a typical device driver reads them (thus making them zero) as a part of the initialization sequence. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1051 23.4.8 Receive Initialization Program the Receive address register(s) per the station address. This can come from the EEPROM or by any other means (for example, on some machines, this comes from the system PROM not the EEPROM on the adapter card). Set up the MTA (Multicast Table Array) by software. This means zeroing all entries initially and adding in entries as requested. Program RCTL with appropriate values. If initializing it at this stage, it is best to leave the receive logic disabled (EN = 0b) until after the receive descriptor rings have been initialized. If VLANs are not used, software should clear VFE. Then there is no need to initialize the VFTA. Select the receive descriptor type. The following should be done once per receive queue needed: * Allocate a region of memory for the receive descriptor list. * Receive buffers of appropriate size should be allocated and pointers to these buffers should be stored in the descriptor ring. * Program the descriptor base address with the address of the region. * Set the length register to the size of the descriptor ring. * Program SRRCTL of the queue according to the size of the buffers, the required header handling and the drop policy. * If header split or header replication is required for this queue, program the PSRTYPE register according to the required headers. * Enable the queue by setting RXDCTL.ENABLE. In the case of queue zero, the enable bit is set by default - so the ring parameters should be set before RCTL.RXEN is set. * Poll the RXDCTL register until the ENABLE bit is set. The tail should not be bumped before this bit was read as one. * Program the direction of packets to this queue according to the mode selected in the MRQC register. Packets directed to a disabled queue are dropped. Note: The tail register of the queue (RDT[n]) should not be bumped until the queue is enabled. 23.4.8.1 Initialize the Receive Control Register To properly receive packets the receiver should be enabled by setting RCTL.RXEN. This should be done only after all other setup is accomplished. If software uses the Receive Descriptor Minimum Threshold Interrupt, that value should be set. 23.4.8.2 Dynamic Enabling and Disabling of Receive Queues Receive queues can be dynamically enabled or disabled given the following procedure is followed: Enabling: * Follow the per queue initialization described in the previous section. Disabling: * Disable the direction of packets to this queue. * Disable the queue by clearing RXDCTL.ENABLE. The Controller stops fetching and writing back descriptors from this queue immediately. The Controller eventually completes the storage of one buffer allocated to this queue. Any further packet Intel(R) Communications Chipset 89xx Series - Datasheet 1052 October 2012 Order Number: 327879-001US 23.0 directed to this queue is dropped. If the currently processed packet is spread over more than one buffer, all subsequent buffers are not written. * The Controller clears RXDCTL.ENABLE only after all pending memory accesses to the descriptor ring or to the buffers are done. The driver should poll this bit before releasing the memory allocated to this queue. The RX path might be disabled only after all Rx queues are disabled. 23.4.9 Transmit Initialization * Program the TCTL register according to the MAC behavior needed. If work in half duplex mode is expected, program the TCTL_EXT.COLD field. a value reflecting the Controller and the PHY SGMII delays should be used. A suggested value for a typical PHY is 0x46 for 10 Mbps and 0x4C for 100 Mbps. The following should be done once per transmit queue: * Allocate a region of memory for the transmit descriptor list. * Program the descriptor base address with the address of the region. * Set the length register to the size of the descriptor ring. * Program the TXDCTL register with the desired TX descriptor write back policy. Suggested values are: -- WTHRESH = 1b -- All other fields 0b. * If needed, set the TDWBAL/TWDBAH to enable head write back * Enable the queue using TXDCTL.ENABLE (queue zero is enabled by default). * Poll the TXDCTL register until the ENABLE bit is set. Note: The tail register of the queue (TDT[n]) should not be bumped until the queue is enabled. Enable transmit path by setting TCTL.EN. This should be done only after all other settings are done. 23.4.9.1 Dynamic Queue Enabling and Disabling Transmit queues can be dynamically enabled or disabled given the following procedure is followed: Enabling: * Follow the per queue initialization described in the previous section. Disabling: * Stop storing packets for transmission in this queue. * Wait until the head of the queue (TDH) is equal to the tail (TDT), i.e. the queue is empty. * Disable the queue by clearing TXDCTL.ENABLE. The Tx path might be disabled only after all Tx queues are disabled. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1053 23.4.10 Virtualization Initialization Flow 23.4.10.1 VMDq Mode 23.4.10.1.1 Global Filtering and Offload Capabilities * Select the VMDQ pooling method - MAC/VLAN filtering for pool selection. MRQC.Multiple Receive Queues Enable = 011b. * Set the RPLOLR and RPLPSRTYPE registers to define the behavior of replicated packets. * Configure VT_CTL.DEF_PL to define the default pool. If packets with no pools should be dropped, set VT_CTL.Dis_def_Pool field. * If needed, enable padding of small packets via the RCTL.PSP 23.4.10.1.2 Per Pool Settings As soon as a pool of queues is associated to a VM the software should set the following parameters: 1. Address filtering: a. The unicast MAC address of the VM by enabling the pool in the RAH/RAL registers. b. If all the MAC addresses are used, the unicast hash table (UTA) can be used. Pools servicing VMs whose address is in the hash table should be declared as so by setting the VMOLR.ROPE. Packets received according to this method didn't pass perfect filtering and are indicated as such. c. Enable the pool in all the RAH/RAL registers representing the multicast MAC addresses this VM belongs to. d. If all the MAC addresses are used, the multicast hash table (MTA) can be used. Pools servicing VMs using multicast addresses in the hash table should be declared as so by setting the VMOLR.ROMPE. Packets received according to this method didn't pass perfect filtering and are indicated as such. e. Define whether this VM should get all multicast/broadcast packets in the same VLAN via the VMOLR.MPE and VMOLR.BAM fields f. Enable the pool in each VLVF register representing a VLAN this VM belongs to. g. Define whether the pool belongs to the default VLAN and should accept untagged packets via the VMOLR.AUPE field 2. Offloads a. Define whether VLAN header should be stripped from the packet (defined by VMOLR.strvlan). b. Set which header split is required via the PSRTYPE register. c. Set whether larger than standard packet are allowed by the VM and what is the largest packet allowed (jumbo packets support) via VMOLR.RLPML & VMOLR.RLE. 3. Queues a. Enable Rx & Tx queues as described in Section 23.4.8 & Section 23.4.9 b. For each Rx queue a drop/no drop flag can be set in SRRCTL.DROP_EN, controlling the behavior in cases no receive buffers are available in the queue to receive packets. The usual behavior is to allow drops in order to avoid head of line blocking, unless a no-drop behavior is needed for some type of traffic (e.g. storage). Intel(R) Communications Chipset 89xx Series - Datasheet 1054 October 2012 Order Number: 327879-001US 23.0 23.4.10.1.3 Security Features: Storm Control The driver may set limits to the broadcast or multicast traffic it can receive. 1. It should set the how many 64 byte chunks of Broadcast and Multicast traffic are acceptable per interval via the BSCTRH and MSCTRH respectively. 2. It should then set the interval to be used via the SCCRL.Interval field and which action to take when the broadcast or multicast traffic crosses the programmed threshold via the SCCRL.BDIPW, SCCRL.BDICW, SCCRL.MDIPW, and SCCRL.MDICW fields. * The driver may be notified of storm control events through the ICR.SCE interrupt cause. 23.5 Access to Shared Resources Part of the resources for the GbE interface in the PCH are shared between several software entities - namely the drivers of the four ports and the internal firmware. In order to avoid contentions, a driver that needs to access one of these resources should use the flow described in Section 23.5.1 in order to acquire ownership of this resource and use the flow described in Section 23.5.2 in order to relinquish ownership of this resource. The shared resources are: 1. EEPROM. 2. All SerDes ports. 3. CSRs accessed by the internal firmware after the initialization process. Currently there are no such CSRs. 4. Software to Software Mailbox Note: Any other software tool that accesses the Controller register set directly should also follow the flow described below. 23.5.1 Acquiring Ownership Over a Shared Resource The following flow should be used to acquire a shared resource: a. If SWSM.SMBI is read as zero, the semaphore was taken. b. Otherwise, repeat steps. SW_FW_SYNC a. Set the bit. b. Read If was successfully set - the semaphore was acquired - otherwise, go back to step a. SW_FW_SYNC Software reads the Software-Firmware Synchronization Register (SW_FW_SYNC) and checks both bits in the pair of bits that control the resource it wishes to own. a. If both bits are cleared (both firmware and other software does not own the resource), software sets the software bit in the pair of bits that control the resource it wishes to own. b. If one of the bits is set (firmware or other software owns the resource), software tries again later. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1055 At this stage, the shared resources is owned by the driver and it may access it. The SWSM and SW_FW_SYNC registers can now be used to take ownership of another shared resources. Note: Bit should not exceed 100 mS. If software takes ownership for a longer duration, Firmware may implement a timeout mechanism and take ownership. Software ownership of bits in SW_FW_SYNC register should not exceed 1 second. If Software takes ownership for a longer duration, Firmware may implement a timeout mechanism and take ownership of the relevant SW_FW_SYNC bits. 23.5.2 Releasing Ownership Over a Shared Resource The following flow should be used to release a shared resource: a. If SWSM.SMBI is read as zero, the semaphore was taken. b. Otherwise, go back to step a. SW_FW_SYNC a. Set the bit. b. Read If was successfully set - the semaphore was acquired - otherwise, go back to step a. SW_FW_SYNC Clear the bit in SW_FW_SYNC that controls the software ownership of the resource to indicate this resource is free. At this stage, the shared resource are released by the driver and it may not access it. The SWSM and SW_FW_SYNC registers can now be used to take ownership of another shared resource. Intel(R) Communications Chipset 89xx Series - Datasheet 1056 October 2012 Order Number: 327879-001US 24.0 24.0 Non-Volatile Memory Map - EEPROM 24.1 EEPROM General Map The GbE controller EEPROM is partitioned into four main blocks followed by Firmware and PXE structures. The first 128 word section is allocated to words common to all LAN ports, Firmware, Software, PXE and LAN port 0. The following 64 word sections are allocated to LAN port 1, LAN port 2 and LAN port 3 as shown in Table 24-1. On a software reset (CTRL.RST or CTRL.DEV_RST) only the words defined in the auto load sequence in Section 22.3.1.3 are re-loaded by HW. On power-up, Hardware reset or CTRL_EXT.EE_RST assertion all words are loaded from EEPROM. Table 24-1. EEPROM Top Level Partitioning EEPROM Word Address Partition 0x0 to 0x7F Common Words (PCIe, PXE, SW and FW) and LAN port 0 words 0x80 to 0xBF LAN Port 1 words 0xC0 to 0xFF LAN Port 2 words 0x100 to 0x13F LAN Port 3 words 0x140 to ... PXE and FW structures Table 24-2 lists the GbE controller EEPROM word map for Common words and Lan Port 0. Note: Reserved EEPROM entries can contain any value, except where noted. Table 24-2. Common and Lan Port 0 EEPROM Map (Sheet 1 of 3) EEPROM Word Address Used By/In High Byte Low Byte Which LAN 0 HW Ethernet Address Byte 2 Ethernet Address Byte 1 LAN 0 01 HW Ethernet Address Byte 4 Ethernet Address Byte 3 LAN 0 02 HW Ethernet Address Byte 6 Ethernet Address Byte 5 LAN 0 03 SW Compatibility High Compatibility Low All 04 Reserved Reserved 05 Compatibility High Compatibility Low Reserved Reserved 06 - 07 08 SW Reserved (must default to FAFAh) All All 09 SW Pointer to PBA Structure All 0A HW Init Control 1 All 0B HW Subsystem ID All 0C HW Subsystem Vendor ID All October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1057 Table 24-2. Common and Lan Port 0 EEPROM Map (Sheet 2 of 3) EEPROM Word Address Used By/In 0D HW Device ID LAN 0 0E HW Vendor ID All 0F HW Init Control 2 All High Byte 10 Which LAN Reserved (Value should be FFFFh) 11 HW (MNG HW) Pass Through LAN Configuration Pointer LAN 0 (MNG HW) 12 HW EEPROM Sizing All 13 HW Init Control 4 LAN 0 14:15 Reserved 16 HW 17 HW GbE Controller: (PHY_RST_Control) * 15:4 --> Reserved * 3 --> Phy_Rst_Cmpl_Prog_Done * 2:0 --> Phy_Rst_Cmpl_Dly 18:1D LAN 0 CSR Auto Configuration Pointer LAN 0 Reserved 1E HW Device REV ID All 1F HW LEDCTL 0 Default LAN 0 20 HW Software Defined Pins Control LAN 0 21 HW CGB Functions Control All 22 Reserved 23 HW Management Hardware Config Control LAN 0 (MNG HW) 24 HW Init Control 3 LAN 0 Reserved All 25-2D 2E HW Watchdog Configuration All 2F OEM VPD Pointer All 30 PXE PXE main setup options LAN0 31 PXE PXE Configuration Customization options LAN0 32 Reserved (Value should be FFFFh) 33 Reserved (Value should be FFFFh) 34 PXE PXE main setup options LAN1 35 PXE PXE Configuration Customization options LAN1 36 Reserved (Value should be FFFFh) 37 Pointer to Alternate MAC address All 38 PXE PXE main setup options LAN2 39 PXE PXE Configuration Customization options LAN2 3A PXE PXE main setup options LAN3 3B PXE PXE Configuration Customization options LAN3 3C Reserved (Value should be FFFFh) 3D Reserved (Value should be FFFFh) 3E 3F Reserved (Value should be FFFFh) SW Intel(R) Communications Chipset 89xx Series - Datasheet 1058 Low Byte Software Checksum, Words 00h - 3Eh October 2012 Order Number: 327879-001US 24.0 Table 24-2. Common and Lan Port 0 EEPROM Map (Sheet 3 of 3) EEPROM Word Address Used By/In 40:41 SW Reserved 42 SW Image Unique ID Low 43 SW Image Unique ID High High Byte Low Byte 44:4F Which LAN Reserved 50 FW PHY Configuration Pointer (Word 0x50) All MNG 51 FW Firmware patch Pointer MNG Code Reserved MNG FW MNG Capabilities MNG HW Reserved MNG FW Sideband Configuration Pointer MNG HW 58:5D Reserved MNG 5E:7F Reserved MNG 80:BF LAN Port 1 words C0:FF LAN Port2 words 52:53 54 55:56 57 100:13F LAN Port3 words 140 FW/PXE Firmware and PXE Structures Table 24-3 maps the GbE controller EEPROM words that can hold different content for LAN Ports 0, 1, 2 and 3. Addresses listed in the table are an offset from the LAN Base address of the relevant EEPROM LAN section. EEPROM LAN Base addresses of the LAN ports are as follows: * LAN Port 0 EEPROM section Base Address - 0x0 * LAN Port 1 EEPROM section Base Address - 0x80 * LAN Port 2 EEPROM section Base Address - 0xC0 * LAN Port 3 EEPROM section Base Address - 0x100 Note: Reserved EEPROM entries can contain any value, except where noted. Table 24-3. LAN Ports 1, 2, and 3 EEPROM Map (Sheet 1 of 2) EEPROM Word Offset Used By/In High Byte Low Byte 0 HW Ethernet Address Byte 2 Ethernet Address Byte 1 01 HW Ethernet Address Byte 4 Ethernet Address Byte 3 02 HW Ethernet Address Byte 6 Ethernet Address Byte 5 03:0C HW Reserved 0D HW Device ID 0E:10 HW Reserved 11 HW (MNG HW) Pass Through LAN Configuration Pointer 12 HW Reserved 13 HW Init Control 4 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1059 Table 24-3. LAN Ports 1, 2, and 3 EEPROM Map (Sheet 2 of 2) EEPROM Word Offset 24.2 Used By/In High Byte Low Byte 14: 16 HW Reserved 17 HW CSR Auto Configuration Pointer 18:1B HW Reserved 1C HW Reserved 1D:1E HW Reserved 1F HW LEDCTL 0 Default 20 HW Software Defined Pins Control 21 HW CGB Functions Control 22:23 HW Reserved 24 HW Init Control 3 25:26 HW Reserved 27 HW CSR Auto Configuration Power-Up Pointer 28:29 HW Reserved 2A:3E HW Reserved 3F SW Software Checksum, Words 00h - 3Eh EEPROM Configuration Notes Setup and configuration of the PCH GbE MAC has the following dependencies which must be considered when programming the EEPROM. * The Intel Network Boot drivers and Intel GbE drivers assumes the EEPROM contains the following MAC configuration details: -- Connection `Type': The configuration details associated with the MAC connection type (SGMII/SerDes) must be programmed to the EEPROM. -- Auto-Negotiation: Auto-Negotiation must be programmed correctly for the MAC configuration via the Auto-Negotiation Enable (ANE) bit. ANE is contained within Initialization Control Word 2 which is one of the common words as listed in Figure 24-2, "Common and Lan Port 0 EEPROM Map" on page 1057. The common words are common across all four MACs therefore auto-negotiation is either enabled or disabled across all MACs. -- PCS parallel detect: PCS parallel detect must be correctly programmed for the MAC configuration via the PCS parallel detect bit. PCS parallel detect is contained within Initialization Control Word 2 which is one of the common words as listed in Figure 24-2, "Common and Lan Port 0 EEPROM Map" on page 1057. The common words are common across all four MACs therefore PCS parallel detect is either enabled or disabled across all MACs. Since the common words are shared across the four MACs, the EEPROM configuration of the four MACs is not entirely independent. The supported connection types, as configured from the EEPROM, are listed in Figure 24-4, "Supported MAC Connection Types" on page 1061. Intel(R) Communications Chipset 89xx Series - Datasheet 1060 October 2012 Order Number: 327879-001US 24.0 Table 24-4. Supported MAC Connection Types Link Mode 24.3 ANE PCS Comment All enabled MACs in SGMII 1 0 All enabled MACs in SerDes KX 0 1 No ANE All enabled MACs in SerDes BX 1 0 ANE All enabled MACs in SerDes BX 0 1 No ANE All enabled MACs in SGMII and SerDes BX ANE 1 0 ANE. The MACs are mixed between SGMII and SerDes. Hardware Accessed Words This section describes the EEPROM words that are loaded by the GbE controller hardware. Most of these bits are located in configuration registers. The words are only read and used if the signature field in the EEPROM Sizing EEPROM word (word 0x12) is valid. Note: When Word is mentioned before an EEPROM address, address is the absolute address in the EEPROM. When Offset is mentioned before an EEPROM address, the address is relative to the start of the relevant EEPROM section. 24.3.1 Ethernet Address (LAN Base Address + Offsets 0x00-0x02) The Ethernet Individual Address (IA) is a 6-byte field that must be unique for each NIC, and thus unique for each copy of the EEPROM image. The first three bytes are vendor specific. For example, the IA is equal to [00 AA 00] or [00 A0 C9] for Intel products. The value from this field is loaded into the Receive Address Register 0 (RAL0/RAH0). For the purpose of this specification, the IA byte numbering convention is indicated as follows: IA Byte / Value Vendor 1 2 3 4 5 6 Intel Original 00 Intel New 00 AA 00 variable variable variable A0 C9 variable variable variable The Ethernet address is loaded for LAN0 from Addresses 0x0 to 0x02 and for LAN 1, 2 and 3 from offsets 0x0 to 0x2 at the start of the relevant sections. 24.3.2 Initialization Control Word 1 (Word 0x0A) The Initialization Control Word 1 in the Common section contains initialization values that: * Set defaults for some internal registers * Enable/disable specific features * Determine which PCI configuration space values are loaded from the EEPROM October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1061 Bit Name Default Value in EEPROM Description 15:14 Reserved 0x0 Reserved 13 Reserved 0b Reserved. 12 Reserved 0b Reserved 11 FRCSPD 0b Default setting for the Force Speed bit in the Device Control register (CTRL[11]). See Section 28.1.4.2 . Bit must always be 0b.. This bit sets the FRCDPLX (bit 12) of the "Device Control Register CTRL [0:3] (0x00000; R/W)": 10 FD 0b 0x0 - All enabled Ports are in SGMII mode. 0x1 - Any enabled Port is in SerDes mode Note: If mixing SerDes and SGMII modes set FD to 1'b1 in the EEPROM word Init_CTRL_1. For all ports in SGMII mode must use EEPROM "CSR Auto Config Ptr (LANx + 0x17)" to set FRCDPLX to 0 in "Device Control Register - CTRL [0:3] (0x00000; R/W)". Reserved. 9:7 Reserved 11b 6 SDP_IDDQ_EN 0b When set, SDP IOs keep their value and direction when the controller enters dynamic IDDQ mode either due to PCIe entering Dr state. Otherwise, SDP IOs moves to HighZ + pull-up mode in dynamic IDDQ mode. Reflected in EEDIAG (See Section 28.3.1.3). 5 Deadlock Timeout Enable 1b If set, a device granted access to the EEPROM that does not toggle the interface for more than 2 seconds will have the grant revoked. See Section 22.3.2.1. 4 Reserved 0b Reserved. 3 Power Management 1b 0b = Reserved 1b = Full support for power management (For normal operation, this bit must be set to 1b). 2 Reserved 0b Reserved 1 Load Subsystem IDs 1b When this bit is set to 1b the controller loads its PCIe Subsystem ID and Subsystem Vendor ID from the EEPROM (Subsystem ID and Subsystem Vendor ID EEPROM words). 0 Load Vendor/Device IDs 1b When set to 1b the controller loads its PCIe Device IDs from the EEPROM (Device ID EEPROM words) and the PCIe Vendor ID from the EEPROM. 24.3.3 Subsystem ID (Word 0x0B) If the Load Subsystem IDs in Initialization Control Word 1 EEPROM word is set, the Subsystem ID word in the Common section is read in to initialize the PCIe Subsystem ID. Default value is 0x0. 24.3.4 Subsystem Vendor ID (Word 0x0C) If the Load Subsystem IDs bit in Initialization Control Word 1 EEPROM word is set, the Subsystem Vendor ID word in the Common section is read in to initialize the PCIe Subsystem Vendor ID. The default value is 0x8086. Intel(R) Communications Chipset 89xx Series - Datasheet 1062 October 2012 Order Number: 327879-001US 24.0 24.3.5 Device ID (LAN Base Address + Offset 0x0D) If the Load Vendor/Device IDs bit in Initialization Control Word 1 is set, the Device ID EEPROM word is read in from the Common, LAN 1, LAN 2 and LAN 3 sections to initialize the Device ID of LAN0, LAN1, LAN2 and LAN3 functions, respectively. The default value is 0x10C9. 24.3.6 Vendor ID (Word 0x0E) If the Load Vendor/Device IDs bit in Initialization Control Word 1 EEPROM word is set, this word is read in to initialize the PCIe Vendor ID. The default value is 0x8086. Note: If a value of 0xFFFF is placed in the Vendor ID EEPROM word, the value in the PCIe Vendor ID register will return to the default 0x8086 value. This functionality is implemented to avoid a system hang situation. 24.3.7 Initialization Control Word 2 (Word 0x0F) The Initialization Control Word 2 read by the GbE controller, contains additional initialization values that: * Set defaults for some internal registers * Enable/disable specific features Bit 15 Default Value in EEPROM Name APM PME# Enable 0b Description Initial value of the Assert PME On APM Wakeup bit in the Wake Up Control (WUC.APMPME) register. See Section 28.20.1.1 . Controls PCS parallel detection: 14 PCS parallel detect 1b SGMII = All enabled media interfaces are in SGMII mode set to 0x0. SerDes & SGMII = Both SerDes and SGMII media interfaces are active set to 0x0. SerDes = All enabled media interfaces are in SerDes mode set to 0x0 or 0x1 depending on your system. Mapped to PCS_LCTL.AN TIMEOUT EN bit. PCS Link Control - PCS_LCTL [0:3] (0x4208; RW). See Section 28.17.1.2 . 13:12 Pause Capability 11b Desired pause capability for advertised configuration base page. Mapped to PCS_ANADV.ASM. See Section 28.17.1.4 . 11 ANE 0b Auto-Negotiation Enable Mapped to PCS_LCTL.AN_ENABLE. See Section 28.17.1.2 10:8 Reserved 7 MAC clock gating enable 1b Enables MAC clock gating power saving mode. Mapped to STATUS[31]. This bit is relevant only if the Enable Dynamic MAC Clock Gating bit is set. See Section 28.1.4.3. Note: Bit impacts clock gating of both regular and Wake-up MAC clocks. 6 Reserved 0b Reserved and RO as "0" 5 Reserved 4 Reserved 0b Reserved and RO as "0" 3 Enable Dynamic MAC Clock Gating 0b When set, enables dynamic MAC clock gating mechanism. See Section 28.1.4.4. 2 SerDes Low Power Enable 0b October 2012 Order Number: 327879-001US Reserved Reserved When set, enables the SerDes to enter a low power state when the function is in Dr state. See Chapter 25.0 and Section 28.1.4.4. Intel(R) Communications Chipset 89xx Series - Datasheet 1063 Bit Default Value in EEPROM Name Reserved. 1 0 Reserved 24.3.8 0b 0b Description Reserved. Reserved EEPROM Sizing (Word 0x12) Provides indication on EEPROM size. Bit Name Description Signature The Signature field indicates to the controller that there is a valid EEPROM present. If the signature field is 01b, EEPROM read is performed, otherwise the other bits in this word are ignored, no further EEPROM read is performed, and default values are used for the configuration space IDs. 13:10 EEPROM Size These bits indicate the EEPROM's actual size. Mapped to EEC[14:11]. 0000b = Reserved 0001b = Reserved 0010b = Reserved 0011b = Reserved 0100b = Reserved 0101b = Reserved 0110b = Reserved 0111b = 16 KB 1000b = 32 KB 1001b = Reserved 1001b = Reserved 1011b = Reserved See Section 28.3.1.1 9:0 Reserved Reserved 15:14 Intel(R) Communications Chipset 89xx Series - Datasheet 1064 October 2012 Order Number: 327879-001US 24.0 24.3.9 Initialization Control 4 (LAN Base Address + Offset 0x13) These words control general initialization values of LAN 0, LAN 1, LAN 2, and LAN 3 ports. Bit 15:12 Default Value in EEPROM Name TXPbsize Description 0x8 Transmit internal buffer size: 0x0 - 20 KB 0x1 - 40 KB 0x2 - 80 KB 0x3 - 1 KB 0x4 - 2 KB 0x5 - 4 KB 0x6 - 8 KB 0x7 - 16 KB 0x8 - 19 KB 0x9 - 38 KB 0xA - 76 KB 0xB:0XF reserved. Note: When 4 ports are enabled maximum buffer size is 20KB. When 2 ports are enabled maximum buffer size is 40KB. When only a single port is enabled maximum buffer size is 80KB. Sets value of ITPBS.TXPbsize. See Section 28.2. 11:8 RXPbsize 0x8 Receive internal buffer size: 0x0 - 36 KB 0x1 - 72 KB 0x2 - 144 KB 0x3 - 1 KB 0x4 - 2 KB 0x5 - 4 KB 0x6 - 8 KB 0x7 - 16 KB 0x8 - 35 KB 0x9 - 70 KB 0xA - 140 KB Note: When 4 ports are enabled maximum buffer size is 36 KB. When 2 ports are enabled maximum buffer size is 72 KB. When only a single port is enabled maximum buffer size is 144 KB. Sets value of IRPBS.RXPbsize. See Section 28.2. 7 Reserved 0b Reserved 6 LPLU 1b Low Power Link Up Enables a decrease in link speed in non-D0a states when power policy and power management states dictate it. 5:1 Phy_Add 0x00 0x01 0x02 0x03 Phy Address. Value loaded to MDICNFG.PHY_ADD field. 0 DEV_RST_EN 1b Enable software reset (CTRL.DEV_RST) generation to all ports (See Section 23.3). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1065 24.3.10 PHY_RST_CONTROL (LAN Base Address + Offset 0x16) These words control Phy reset for LAN0, 1, 2, 3. . Bit Default Value in EEPROM Name Description 15:4 Reserved 0x0 Reserved 3 PHY_Rst_Cmpl_Prog _Done 0x1 This bit loads CGB_PHY_RST_CMPL_DLY CSR bit[3] when EEPROM load is done. 2:0 PHY_Rst_Cmpl_Dly 0x3 This field loads the initial value to CGB_PHY_RST_CMPL_DLY CSR bit[2:0] when EEPROM load is done. 24.3.11 Bit Device Rev ID (Word 0x1E) Default Value in EEPROM Name Description 15:8 Reserved 0x0 Reserved 7:0 DEVREVID 0x0 Device Revision ID The actual device revision ID is the EEPROM value XORed with the hardware default of Rev ID. For the Controller A0, the default value is zero. 24.3.12 LED 0 Configuration Defaults (LAN Base Address + Offset 0x1F) These EEPROM words specify the hardware defaults for the LEDCTL register fields controlling the LED0 (LINK_UP) output behaviors. Words control LEDs behavior of LAN 0, LAN 1, LAN 2, and LAN 3 ports. . Bit Name Default Value in EEPROM Description Initial value of the LED0_MODE field specifying what event/state/pattern is displayed on LED0 (LINK_UP) output. A value of 0010b (0x2) indicates the LINK_UP state. See Section 28.1.4.2. 3:0 LED0 Mode 0010b 4 Reserved 0b Reserved. Set to 0b. 5 Global Blink Mode 0b Global Blink Mode 0b = Blink at 200 ms on and 200ms off. 1b = Blink at 83 ms on and 83 ms off. See Section 28.1.4.2. 6 LED0 Invert 0b Initial value of LED0_IVRT field. 0b = Active-low output. See Section 28.1.4.2. 7 LED0 Blink 0b Initial value of LED0_BLINK field. 0b = Non-blinking. See Section 28.1.4.2. 15:8 Reserved 0x0 Reserved A value of 0x0002 is used to configure default hardware LED behavior (LED0=LINK_UP). Intel(R) Communications Chipset 89xx Series - Datasheet 1066 October 2012 Order Number: 327879-001US 24.0 24.3.13 Software Defined Pins Control (LAN Base Address + Offset 0x20) These words at offset 0x20 from start of relevant EEPROM section are used to configure initial settings of software defined pins (SDPs) for LAN 0, LAN 1, LAN 2 and LAN 3. Bit 15:12 11 Name Reserved LAN_DIS Default Value in EEPROM Description 0x0 Reserved 0b LAN Disable In LAN ports 1, 2 and 3, when set to 1b, the appropriate LAN is disabled (both PCIe function and LAN access for manageability are disabled). Note: Has not effect for LAN port 0 - always enabled. Note: If LAN_DIS is set to 1b for a port then PHY CFG pointer (offset 0x50) must not have a PHY write command for that port set. See Section 24.6, "PHY Configuration Structure" on page 1075. 10 LAN_PCI_DIS 0b LAN PCIe Function Disable In LAN ports 0, 1, 2 and 3, when set to 1b, the appropriate LAN PCI function is disabled. For example, in the case where the LAN is functional for manageability operation but is not connected to the host through the PCIe interface. Reflected in EEDIAG (See Section 28.3.1.3). Note: Has not effect for LAN port 0 - always enabled. 9 SDPDIR[1] 0b SDP1 Pin - Initial Direction This bit configures the initial hardware value of the SDP1_IODIR bit in the Device Control (CTRL) register following power up. See Section 28.1.4.2 . 8 SDPDIR[0] 0b SDP0 Pin - Initial Direction This bit configures the initial hardware value of the SDP0_IODIR bit in the Device Control (CTRL) register following power up. See Section 28.1.4.2 . 7:6 Reserved 00b Reserved WD_SDP0 0b When set, SDP[0] is used as a watchdog timeout indication. When reset, it is used as an SDP (as defined in bits 8 and 0). See Section 28.1.4.2 . 4:3 Reserved 00b Reserved 2 D3COLD_WAKEU P_ADVEN 1b Configures the initial hardware default value of the ADVD3WUC bit in the Device Control (CTRL) register following power up. See Section 28.1.4.2 . 1 SDPVAL[1] 0b SDP1 Pin - Initial Output Value This bit configures the initial power-on value output on SDP1 (when configured as an output) by configuring the initial hardware value of the SDP1_DATA bit in the Device Control (CTRL) register after power up. See See Section 28.1.4.2 . 0 SDPVAL[0] 0b SDP0 Pin - Initial Output Value This bit configures the initial power-on value output on SDP0 (when configured as an output) by configuring the initial hardware value of the SDP0_DATA bit in the Device Control (CTRL) register after power up. See Section 28.1.4.2 . 5 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1067 24.3.14 CGB Functions Control (Word 0x21) Bits 15:11 10 Name Reserved BAR32 Default Value in EEPROM Description 0x0 Reserved 0x1 Bit (loaded to the BARCTRL register) preserves the legacy 32 bit BAR mode when BAR32 is set. When cleared to 0b 64 bit BAR addressing mode is selected. Note: If PREFBAR is set the BAR32 bit should always be 0 (64 bit BAR addressing mode). Refer to bits[29:20] of Section 29.3.2, "Base Address Registers (0x10...0x27; R/W)". 9 PREFBAR 0x0 0 = BARs are marked as non prefetchable 1 = BARs are marked as prefetchable Note: PREFBAR should be set when 64 bit BARs are used. Bit should not be set in 32 bit BAR addressing mode. Refer to bits[29:20] of Section 29.3.2, "Base Address Registers (0x10...0x27; R/W)". 8:0 Reserved 0x0 Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1068 October 2012 Order Number: 327879-001US 24.0 24.3.15 Initialization Control 3 (LAN Base Address + Offset 0x24) These words control general initialization values of LAN 0, LAN 1, LAN 2 and LAN 3 ports. Bit 15 Default Value in EEPROM Name SerDes Energy Source 0b Description SerDes Energy Source Detection When set to 0b, source is internal SerDes Rx circuitry for electrical idle or link-up indication. When set to 1b, source is external SRDS_[n]_SIG_DET signal for electrical idle or Link-up indication. This bit also indicates the source of the signal detect while establishing a link in SerDes mode. This bit sets the default value of the CONNSW.ENRGSRC bit. 2 wires SFP interface enable - bit is used to enable interfacing an external PHY either VIA the MDIO or I2C interface 0b = Disabled. When disabled, the 2 wires I/F pads are isolated. 1b = Enabled. Used to set the default value of CTRL_EXT.I2C Enabled. See Section 28.1.4.4 . Note: Common MDIO is selected per LAN by setting Com_MDIO <bit 3> of Initialization Control 3 - If Common MDIO mode this bit must be set to 1b for LAN 0 - If LAN 1 is in Common MDIO mode set this bit for LAN 1 to 0b - If LAN 2 is in Common MDIO mode set this bit for LAN 2 to 0b - If LAN 3 is in Common MDIO mode set this bit for LAN 3 to 0b See note in Com_MDIO <bit 3> and MDIO_I2C <bit 2> of Initialization Control 3 14 2 wires SFP Enable 0b 13 ILOS 0b Default setting for the loss-of-signal polarity bit (CTRL[7]). See Section 28.1.4.2 0x0 Reserved 12:11 10 APM Enable 0b Initial value of Advanced Power Management Wake Up Enable bit in the Wake Up Control (WUC.APME) register. Mapped to CTRL[6] and to WUC[0]. See Section 28.1.4.2 and See Section 28.20.1.1 . 9 Reserved 0b Reserved 8 ACBYP 0b Bypass on-chip AC coupling in RX input buffers ACBYP = 0 -Normal mode; on-chip AC coupling present. ACBYP = 1 - On-chip AC coupling bypassed. 7 reserved 1b reserved. 6 reserved 0b reserved. 00b Initial value of Link Mode bits of the Extended Device Control (CTRL_EXT.LINK_MODE) register, specifying which link interface and protocol is used by the MAC. 00b = Reserved Note: This is the lowest power mode. This is default for the PCH to keep MAC in lowest power mode by default. 01b = MAC and SerDes I/F operate in 1000BASE-KX mode. 10b = MAC and SerDes operate in SGMII mode. 11b = MAC and SerDes I/F operate in SerDes (1000BASE-BX) mode. See Section 28.1.4.4 . 5:4 Link Mode October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1069 Bit 3 Default Value in EEPROM Name Com_MDIO 2 MDIO_I2C 1 Ext_VLAN 01 Keep_PHY_Link_U p_En Description When interfacing an external SGMII phy or SerDes, bit defines if MDIO access is routed to a shared MDIO port on LAN 0, to support multi port external PHYs or to the dedicated per function MDIO port. 0b - MDIO access routed to the LAN port's MDIO interface. 1b - MDIO accesses on this LAN port routed to LAN port 0 MDIO interface. Note: If Common MDIO <this bit> is set for LAN 1, LAN 2 or LAN 3 then the following bits must be set to 1b for LAN 0 - This bit Com_MDIO <bit 3> - 2 wire SFP Enable <bit 14> Initialization Control 3 - MDIO_I2C <bit 2> Initialization Control 3 Used to set the default value of MDICNFG.Com_MDIO bit. See Section 28.1.4.6. 0b MDIO/I2C Interface Selection: 0b = I2C 1b = MDIO Note: This bit must be set to 1b for each LAN that has Com_MDIO <bit 3> set. See note in Com_MDIO <bit 3>. 0b Sets the default for CTRL_EXT[26] bit. Indicates that additional VLAN is expected in this system. See Section 28.1.4.4 . 0b Enables No PHY Reset when the Baseboard Management Controller (BMC) indicates that the PHY should be kept on. When asserted, this bit prevents the PHY reset signal and the power changes reflected to the PHY according to the MANC.Keep_PHY_Link_Up value. Note: This EEPROM bit should be set to the same value for all LAN ports. 1. BMC uses this bit to keep MAC interface alive even if PCIe function is off. 24.3.16 PCIe Control 3 (Word 0x29) This word is used for programming PCIe functionality. Bits 15:6 Default Value in EEPROM Name Reserved Description 0x0 Reserved 5 Wake_pin_enable 0b Enables the use of the wake pin for a PME event in all non L2 power states. WAKE_N pin will be asserted in the following condition: 1) wake enable from LAN port filter state machine 2) no PCIe power i.e pe_rst_n asserted If the pe_rst_n is deasserted then this bit from EEPROM can disable the above gating i.e. wake pin will be always asserted when LAN port filter indicates pme. 4:0 Reserved 0x0 Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1070 October 2012 Order Number: 327879-001US 24.0 24.3.17 Bit Watchdog Configuration (Word 0x2E) Default Value in EEPROM Name Description 15 Watchdog Enable 0b Enable watchdog interrupt. See Section 28.15.1.1. 14:11 Watchdog Timeout 0x2 Watchdog timeout period (in seconds). See Section 28.15.1.1. 10:0 Reserved 0x0 Reserved 24.3.18 VPD Pointer (Word 0x2F) This word points to the Vital Product Data (VPD) structure. This structure is available for the NIC vendor to store it's own data. A value of 0xFFFF indicates that the structure is not available. Bit 15:0 Name Description Offset to VPD structure in words. Bit 15 must be set to 0b (VPD area must be in the first 32 Kbytes of the EEPROM). VPD offset 24.4 CSR Auto Configuration Pointer (LAN Base Address + Offset 0x17) Words points to the CSR auto configuration structures of LAN 0, LAN 1, LAN 2 and LAN3. Sections are loaded during HW auto-load as described in Section 22.3.1.3. If no CSR autoload is required for the specific LAN port, the word shall be set to 0xFFFF. The CSR Auto Configuration structure format is listed in the Table 24-5. For LAN 0, 1, 2, or 3 this structure may contain Generic LANx CSR load values. Table 24-5. CSR Auto Configuration Structure Format Offset High Byte[15:8] Low Byte[7:0] Section 0x0 Section Length = 3*n (n - number of CSRs to configure) Section 24.4.1 0x1 Block CRC8 Section 24.4.2 0x2 CSR Address Section 24.4.3 0x3 Data LSB Section 24.4.4 Data MSB Section 24.4.5 0x4 ... 3*n - 1 CSR Address Section 24.4.3 3*n Data LSB Section 24.4.4 3*n + 1 Data MSB Section 24.4.5 The CSR Auto Configuration structure format is listed in the following tables. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1071 24.4.1 CSR Configuration Section Length - Offset 0x0 The section length word contains the length of the section in words. Section length count does not include the section length word and Block CRC8 word. Bits 15:0 Name Default Section_length 24.4.2 Section length in words. Block CRC8 (Offset 0x1) Bit Name 15:8 Reserved 7:0 CRC8 24.4.3 Description Description CSR Address - Offset 0x2 The CSR Address field is used to access one of the MMIO Registers listed in the "Programming Interface" chapter. However, in this structure the CSR address must only specify the aligned DWORD address of the given register, for example, only bits [15:2] of the register address as specified in the Programming Interface chapter must be used. Bits 15:0 24.4.4 Bits 15:0 24.4.5 Bits 15:0 Name Default CSR_ADDR CSR Address in Double Words (4 bytes) CSR Data LSB - Offset 0x3 Name Default CSR_Data_LSB Description CSR Data LSB CSR Data MSB - Offset 0x4 Name Default CSR_Data_MSB Intel(R) Communications Chipset 89xx Series - Datasheet 1072 Description Description CSR Data MSB October 2012 Order Number: 327879-001US 24.0 24.5 CSR Auto Configuration Power-Up Pointer (LAN Base Address + Offset 0x27) Words points to the CSR auto configuration Power-Up structures of LAN 0, LAN 1, LAN 2 and LAN3 that are read only following power-up. Sections are loaded during HW auto-load as described in Section 22.3.1.3. If no CSR autoload is required for the specific LAN port, the word shall be set to 0xFFFF. The CSR Auto Configuration Power-Up structure format is listed Table 24-6. For LAN 0 the structure may contain three sub-sections in the following order: PCIe MPHY, LAN 0 MPHY, and Generic LAN 0 CSR load values. For LAN 1, 2, or 3 the structure may contain 2 sub-sections in the following order: LANx MPHY, and Generic LANx CSR load values. There is only a Section Length and Block CRC8 per structure block regardless of the number of sub-sections in each block. The hardware simply writes the MSB/LSB Data pair to the CSR Address provided. Table 24-6. CSR Auto Configuration Power-Up Structure Format Offset High Byte[15:8] Low Byte[7:0] Section 0x0 Section Length = 3*n (n - number of CSRs to configure) Section 24.5.1 0x1 Block CRC8 Section 24.5.2 0x2 CSR Address Section 24.5.3 0x3 Data LSB Section 24.5.4 Data MSB Section 24.5.5 0x4 ... 3*n - 1 CSR Address Section 24.5.3 3*n Data LSB Section 24.5.4 3*n + 1 Data MSB Section 24.5.5 24.5.1 CSR Configuration Power-Up Section Length - Offset 0x0 The section length word contains the length of the section in words. Section length count does not include the section length word and Block CRC8 word. Bits 15:0 Name Default Section_length 24.5.2 Description Section length in words. Block CRC8 (Offset 0x1) Bit Name 15:8 Reserved 7:0 CRC8 October 2012 Order Number: 327879-001US Description Intel(R) Communications Chipset 89xx Series - Datasheet 1073 24.5.3 CSR Address - Offset 0x2 The CSR Address field is used to access one of the MMIO Registers listed in the "Programming Interface" chapter. However, in this structure the CSR address must only specify the aligned DWORD address of the given register, for example, only bits [15:2] of the register address as specified in the Programming Interface chapter must be used. Bits 15:0 24.5.4 Bits 15:0 24.5.5 Bits 15:0 24.5.6 Name Default CSR_ADDR Description CSR Address in Double Words (4 bytes) CSR Data LSB - Offset 0x3 Name Default CSR_Data_LSB Description CSR Data LSB CSR Data MSB - Offset 0x4 Name Default CSR_Data_MSB Description CSR Data MSB EICT Information Structure The EEPROM space prior to the beginning of the CSR Auto Configuration Power-Up Structure for each LANx when be preserved to accommodate the EICT revision history. Table 24-7 describes the structure entries. Table 24-7. EICT Information Structure Format Variable Name Description Size U8 solution[8] Read from mphy file - bin min trace length in inches, bin max trace length in inches. 8 Bytes U8 type[8] File type PCIe, SerD SFP, SerD Bkp, SGMI, SGMI Bkp. 8 Bytes U16 version Version number between 00 - 99 read from file. 2 Bytes U16 size Size in words, calculated by EICT. 2 Bytes U16 strucVer This would be set/used by EICT to version this structure. For this structure the value would be 0x1 set by EICT. 2 Bytes U16 Signature Always 0x55AA set by EICT. 2 Bytes Intel(R) Communications Chipset 89xx Series - Datasheet 1074 October 2012 Order Number: 327879-001US 24.0 24.6 PHY Configuration Structure This section describes the PHY auto configuration structure used to configure PHY related circuitry. The programming in this section is applied after each PHY reset. After the registers in this section are written to the PHY, the relevant EEMNGCTL.CFG_DONE bit is set. The PHY Configuration Pointer (Word 0x50) points to the start (offset 0x0) of this type of structure, to configure PHY registers (External PHYs). If pointer is 0xFFFF then no structure exists. Structure is loaded during HW EEPROM auto-load. Table 24-8. PHYAuto Configuration Structure Format Offset High Byte[15:8] Low Byte[7:0] Section 0x0 Section Length = 2*n (n - number of registers to configure) 0x1 Block CRC8 Table 24-10 0x2 PHY number and PHY address Table 24-11 0x3 PHY data (MDIC[15:0] or I2CCMD[15:0]) Table 24-12 ... 2*n PHY number and PHY address Table 24-11 2*n + 1 PHY data (MDIC[15:0] or I2CCMD[15:0]) Table 24-12 24.6.1 PHY Configuration Section Length - Offset 0x0 The section length word contains the length of the section in words. Section length count does not include the section length word and Block CRC8 word. Table 24-9. HY Configuration Section Length Bits Name 15 Reserved 14:0 Section_length 24.6.2 Default Description Section length in words. Block CRC8 (Offset 0x1) Table 24-10. Block CRC8 structure Bit Name 15:8 Reserved 7:0 CRC8 October 2012 Order Number: 327879-001US Description Intel(R) Communications Chipset 89xx Series - Datasheet 1075 24.6.3 PHY Number and PHY Address - (Offset 2*n; [n = 1... Section Length]) Table 24-11. PHY Number and PHY Address Bits Name Default Description 15:12 Reserved Reserved 11 Apply to port 3 If set, apply to programming when the PHY of port three is reset. Note: Must be disabled/cleared when Port 3 is disabled via LAN_DIS. 10 Apply to port 2 If set, apply to programming when the PHY of port two is reset. Note: Must be disabled/cleared when Port 2 is disabled via LAN_DIS. 9 Apply to port 1 If set, apply to programming when the PHY of port one is reset. Note: Must be disabled/cleared when Port 1 is disabled via LAN_DIS. 8 Apply to port 0 If set, apply to programming when the PHY of port zero is reset. Note: Must be disabled/cleared when Port 0 is disabled via LAN_DIS. 7:0 PHY address PHY address to which the data is written 24.6.4 PHY Data (Offset 2*n + 1; [n = 1... Section Length]) Table 24-12. PHY Data Bits 15:0 Name Default Reg_Data 24.7 Description MDIC[15:0]/I2CCMD[15:0] value (Data). Pointers and Control Words Used By Firmware Word 0x51 is used to point to the load Firmware patch structure. Word 0x11 in each LAN port EEPROM section are used to point to structures specific to Pass through operation. Words 0x54 and 0x23 control some aspects of the Firmware functionality. A value of 0xFFFF for a pointer indicates that the relevant structure is not present in the EEPROM. 24.7.1 Pass Through LAN Configuration Pointer (LAN Base Address + Offset 0x11) Bit 15:0 Name PHY Configuration Pointer (Word 0x50) Bit 15:0 Pointer to the PT LAN Configuration Structure. See Section 24.9 for details of the structure. Pointer 24.7.2 Name Pointer Intel(R) Communications Chipset 89xx Series - Datasheet 1076 Description Description Pointer to PHY configuration structure. A value of 0xFFFF means the pointer is invalid. October 2012 Order Number: 327879-001US 24.0 24.7.3 Firmware Patch Pointer (Word 0x51) Bit 15:0 Name Pointer 24.7.4 24.7.5 Pointer to loader patch structure. See Section 24.8 for details of the structure. Sideband Configuration Pointer (Word 0x57) Bit 15:0 Description Name Description Pointer to the Sideband configuration pointer structure. See Section 24.10 for details of the structure. Pointer Manageability Capability/Manageability Enable (Word 0x54) Bit Name Description 15 Reserved Reserved 14 Redirection Sideband Interface 0b = SMBus. (For PCH, it needs to be set to 0) 13:12 Reserved Reserved. 11 MCSR_TO_RETRY 0b - On CSR access Timeout return failed access to BMC. 1b - On CSR access Timeout retry access internally and don't return failed access. Default value - 1b 10:8 Manageability Mode 0x0 = None. 0x1 = Reserved. 0x2 = Pass Through (PT) mode. 0x3 = Reserved. 0x4 = Host interface enable only. 0x5:0x7 = Reserved. 7 Port3 Manageability Capable 0b = Not capable 1b = Bit 3 is applicable to port 3. Note: Must be set if bit 3 is set. 6 Port2 Manageability Capable 0b = Not capable 1b = Bit 3 is applicable to port 2. Note: Must be set if bit 3 is set. 5 Port1 Manageability Capable 0b = Not capable 1b = Bit 3 is applicable to port 1. Note: Must be set if bit 3 is set. 4 Port0 Manageability Capable 0b = Not capable 1b = Bit 3 is applicable to port 0. Note: Must be set if bit 3 is set. 3 Pass Through Capable 0b = Disable. 1b = Enable. Must be set if bits[10:8] (Manageability Mode) = 0x2 (PT). 2:0 Reserved Reserved October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1077 24.7.6 Management HW Config Control (Word 0x23) This word contains bits that configure special firmware behavior. Bit Name Description 15 LAN3_FTCO_RST_DIS LAN3 force TCO reset disable (1b disable; 0b enable). 14 LAN2_FTCO_RST_DIS LAN2 force TCO reset disable (1b disable; 0b enable). 13 LAN1_FTCO_RST_DIS LAN1 force TCO reset disable (1b disable; 0b enable). 12 LAN0_FTCO_RST_DIS LAN0 force TCO reset disable (1b disable; 0b enable). 11 Reserved Reserved 10 PARITY_ERR_RST_EN When set enables reset of Management logic and generation of internal Firmware reset as a result of Parity Error detected in Management memories. 9 Enable All Phys in D3 0 - Only ports activated for BMC activity will stay active in D3: In SMBUS mode - according to EEPROM port enable bit. 1- All phys will stay active in D3. 8:4 Reserved Reserved 3 LAN3_FTCO_ISOL_DIS LAN3 force TCO Isolate disable (1b disable; 0b enable). 2 LAN2_FTCO_ISOL_DIS LAN2 force TCO Isolate disable (1b disable; 0b enable). 1 LAN1_FTCO_ISOL_DIS LAN1 force TCO Isolate disable (1b disable; 0b enable). 0 LAN0_FTCO_ISOL_DIS LAN0 force TCO Isolate disable (1b disable; 0b enable). 24.8 Firmware Patch Structure This structure is used for all FW patches. Firmware patch Pointer (Word 0x51) points to the start (offset 0x0) of this kind of structure. If pointer is 0xFFFF then no structure exists. Structure is loaded during HW EEPROM auto-load as described in Chapter 22.0, "GbE Interconnects". 24.8.1 Firmware Patch Data Size (Offset 0x0) The Data Size word contains the length of the section in words. Data Size count does not include the section length word and Block CRC8 word. Bit 15:0 Name Data Size (Words) 24.8.2 Block CRC8 (Offset 0x1) Bit Name 15:8 Reserved 7:0 CRC8 Intel(R) Communications Chipset 89xx Series - Datasheet 1078 Description Description October 2012 Order Number: 327879-001US 24.0 24.8.3 Patch Ram Address Word (Offset 0x2) Bit 15:0 Name Ram address 24.8.4 Ram Address to load patch. Patch Version 1 Word (Offset 0x3) Bit Name 15:8 Patch Generation Hour 7:0 Patch Generation Minutes 24.8.5 Name 15:8 Patch Generation Month 7:0 Patch Generation Day Name 15:8 0x00 0x01 Patch Silicon Version Compatibility 0x10 0x11 7:0 Patch Generation Year Name 15:8 Patch Major Number 7:0 Patch Minor Number = = = = A0. A1. B0. B1. Description Patch Data Words (Offset 0x7, Block Length) Bit 15:0 Description Patch Version 4 Word (Offset 0x6) Bit 24.8.8 Description Patch Version 3 Word (Offset 0x5) Bit 24.8.7 Description Patch Version 2 Word (Offset 0x4) Bit 24.8.6 Description Name Description Patch Firmware Data October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1079 24.9 PT LAN Configuration Structure The Pass Through LAN Configuration Pointer (LAN Base Address + Offset 0x11) points to the start (offset 0x0) of this type of structure, to configure manageability filters. If pointer is 0x FFF then no structure exists. Structure is loaded during HW EEPROM auto-load. When interface to BMC is via SMBus the PT LAN Configuration Structure must be used to configure the required filters. The PT LAN Configuration Structure format is listed in the following tables. Table 24-13. PT LAN Configuration Structure Format Offset High Byte[15:8] 0x0 Low Byte[7:0] Section Section Length = 3*n (n - number of CSRs to configure) Section 24.9.1 0x1 Block CRC8 Section 24.9.2 0x2 CSR Address Section 24.9.3 0x3 Data LSB Section 24.9.4 0x4 Data MSB Section 24.9.5 3*n - 1 CSR Address Section 24.9.3 3*n Data LSB Section 24.9.4 3*n + 1 Data MSB Section 24.9.5 ... 24.9.1 PT LAN Configuration Structure Section Length - Offset 0x0 The section length word contains the length of the section in words. Section length count does not include the section length word and Block CRC8 word. Bits Name 15 Reserved 14:0 Section_length 24.9.2 Section length in words. Name 15:8 Reserved 7:0 CRC8 Description CSR Address - (Offset 2*n; [n = 1... Section Length]) Bits Name 15 Reserved 14:0 CSR_ADDR Default Intel(R) Communications Chipset 89xx Series - Datasheet 1080 Description Block CRC8 (Offset 0x1) Bit 24.9.3 Default Description CSR Address in Double Words (4 bytes) October 2012 Order Number: 327879-001US 24.0 24.9.4 Bits 15:0 24.9.5 Bits 15:0 24.9.6 CSR Data LSB - (Offset 0x1 + 2*n; [n = 1... Section Length]) Name Default CSR_Data_LSB Description CSR Data LSB CSR Data MSB - (Offset 0x2 + 2*n; [n = 1... Section Length]) Name Default CSR_Data_MSB Description CSR Data MSB Manageability Filters The following table lists registers that can be programmed via the PT LAN Configuration structure. Name Description Section MAVTV Management VLAN TAG Value Section 28.21.1.1, "Management VLAN TAG Value--MAVTV [0:3][0:7] (0x5010 +4*n [n=0...7]; RW)" MFUTP Management Flex UDP/TCP Ports Section 28.21.1.2, "Management Flex UDP/TCP Ports--MFUTP[0:3][0:3] (0x5030 + 4*n [n=0...3]; RW)" METF Management Ethernet Type Filters Section 28.21.1.3, "Management Ethernet Type Filters--METF [0:3][0:3] (0x5060 + 4*n [n=0...3]; RW)" MNGONLY Management Only Traffic Register Section 28.21.1.4, "Management Control Register--MANC [0:3] (0x5820; RW)" MDEF Manageability Decision Filters Section 28.21.1.6, "Manageability Decision Filters--MDEF [0:3][0:7] (0x5890 + 4*n [n=0...7]; RW)" MDEF_EXT Manageability Decision Filters Extension Section 28.21.1.7, "Manageability Decision Filters--MDEF_EXT [0:3][0:7] (0x5930 + 4*n[n=0...7]; RW)" MIPAF Manageability IP Address Filter registers (IPv4 or IPV6) Note: Can be used to filter IPV4 Address of ARP packets. Section 28.21.1.8, "Manageability IP Address Filter--MIPAF [0:3][0:15] (0x58B0 + 4*n [n=0...15]; RW)" MMAL Manageability MAC Address Low Registers Section 28.21.1.9, "Manageability MAC Address Low--MMAL [0:3] (0x5910 + 8*n [n= 0...1]; RW)" MMAH Manageability MAC Address High Registers Section 28.21.1.10, "Manageability MAC Address High--MMAH [0:3][0:1] (0x5914 + 8*n [n=0...1]; RW)" FTFT Flexible TCO Filter Table registers Section 28.21.1.11, "Flexible TCO Filter Table Registers--FTFT [0:3] (0x9400-0x94FC; RW)" October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1081 24.10 Sideband Configuration Structure The Sideband Configuration Pointer (Word 0x57) points to the start (offset 0x0) of this structure. If pointer is 0xFFFF then no structure exists. 24.10.1 Section Length (Offset 0x0) The section length word contains the length of the section in words. Section length count does not include the section length word and Block CRC8 word. Bits 15:0 Name Default Section_length 24.10.2 Section length in words. Block CRC8 (Offset 0x1) Bit Name 15:8 Reserved 7:0 CRC8 24.10.3 0 Description Description SMBus Max Fragment Size (Offset 0x2) Bit Name Description 15:8 Reserved 7:0 Maximum SMBus Fragment Size. Value should be in the 32 to 240 Byte SMBus Max Fragment Size (Bytes) range. Note: Fragment size should be a multiple of 4 + 1 (e.g. 33, 37...). 24.10.4 SMBus Notification Timeout (Offset 0x3) Bit 15:0 Reserved Name SMBus Notification Timeout (ms) 24.10.5 Description Timeout until the discarding of a packet not read by the external BMC completes. 0b - No discard. SMBus Slave Address 0 1 (Offset 0x4) Bit Name Description 15:9 SMBus 1 Slave Address SMBus Slave Address for port 1. 8 Reserved Reserved. 7:1 SMBus 0 Slave Address SMBus Slave Address for port 0. 0 Reserved Reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 1082 October 2012 Order Number: 327879-001US 24.0 24.10.6 SMBus Slave Address 2 3 (Offset 0x5) Bit Name Description 15:9 SMBus 3 Slave Address SMBus Slave Address for port 3. 8 Reserved Reserved. 7:1 SMBus 2 Slave Address SMBus Slave Address for port 2. 0 Reserved Reserved. 24.10.7 Reserved (Offset 0x6) Bit 15:0 Name Reserved 24.10.8 Description Reserved LAN Receive Enable 1 (Offset 0x7) Bit Name Description 15:9 Receive Enable Byte 12 BMC SMBus slave address. 8 Reserved Reserved 7 Reserved Reserved 6 Reserved Reserved 5:4 Notification Method 00b 01b 10b 11b 3 Enable ARP Response 0 - Disable ARP Response 1 - Enable ARP Response 2 Enable Status Reporting 1:0 Reserved 24.10.9 = = = = SMB alert. Asynchronous notify. Direct receive. Reserved. Reserved LAN Receive Enable 2 (Offset 0x8) Bit Name Description 15:8 Receive Enable Byte 14 Alert value. 7:0 Receive Enable Byte 13 Interface value. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1083 24.10.10 SMBus Flags (Offset 0x9) Bit 15:6 Name Description Reserved Reserved 5 SMBus Block Read Command 0b = Block read command is C0. 1b = Block read command is D0. 4:3 Reserved Reserved 2 Disable SMBus ARP Functionality 1 SMBus ARP PEC 0 Reserved 24.10.11 LAN Receive Enable 3 (Offset 0xA) Bit 15:4 3 2 1 0 Name Description Reserved Reserved Enable BMC Dedicated MAC port 3 Configure BMC dedicated MAC Address on Port 3. 0b = The PCH will share MAC address for MNG traffic with host MAC address on Port 3. Host MAC address is specified in EEPROM words located at LAN Base Address + offset 0x0-0x2 per port. 1b = The PCH will use the BMC dedicated MAC address on port 3 (MAC address (MMAL/H)as filter for incoming receive packets. Note: MMAL/H registers can be programmed from the EEPROM using the PT LAN Configuration structure. Enable BMC Dedicated MAC port 2 Configure BMC dedicated MAC Address on Port 2. 0b = The PCH will share MAC address for MNG traffic with host MAC address on Port 2. Host MAC address is specified in EEPROM words located at LAN Base Address + offset 0x0-0x2 per port . 1b = The PCH will use the BMC dedicated MAC address on port 2(MAC address (MMAL/H)as filter for incoming receive packets. Note: MMAL/H registers can be programmed from the EEPROM using the PT LAN Configuration structure. Enable BMC Dedicated MAC port 1 Configure BMC dedicated MAC Address on Port 1. 0b = The PCH will share MAC address for MNG traffic with host MAC address on Port 1. Host MAC address is specified in EEPROM words located at LAN Base Address + offset 0x0-0x2 per port. 1b = The PCH will use the BMC dedicated MAC address on port 1(MAC address (MMAL/H) as filter for incoming receive packets. Note: MMAL/H registers can be programmed from the EEPROM using the PT LAN Configuration structure. Enable BMC Dedicated MAC port 0 Configure BMC dedicated MAC Address on Port 0. 0b = The PCH will share MAC address for MNG traffic with host MAC address on Port 0. Host MAC address is specified in EEPROM words located at LAN Base Address + offset 0x0-0x2 per port. 1b = The PCH will use the BMC dedicated MAC address on port 3 (MAC address (MMAL/H) as filter for incoming receive packets. Note: MMAL/H registers can be programmed from the EEPROM using the PT LAN Configuration structure. Intel(R) Communications Chipset 89xx Series - Datasheet 1084 October 2012 Order Number: 327879-001US 24.0 This value will be stored in the LSB Portion of the LAN0 Management Control (MANC) register. Bit Name Description 15 TCO_Isolate TCO Isolate command. Note: Value placed in this field is not written to MANC register. 14 FW_RESET FW Reset occurred. Note: Value placed in this field is not written to MANC register. 13:0 Reserved Reserved. This value will be stored in the MSB Portion of the LAN0 Management Control (MANC) register. Bit 15:11 Name Description Reserved Reserved. 10 NET_TYPE NET TYPE: 0b = pass only un-tagged packets. 1b = pass only VLAN tagged packets. Valid only if FIXED_NET_TYPE is set. 9 FIXED_NET_TYPE Fixed net type: If set, only packets matching the net type defined by the NET_TYPE field passes to manageability. Otherwise, both tagged and un-tagged packets can be forwarded to the manageability engine. 8 EN_IPv4_FILTER Enable IPv4 address Filters - when set, the last 128 bits of the MIPAF register are used to store 4 IPv4 addresses for IPv4 filtering. When cleared, these bits store a single IPv6 filter. 7 EN_XSUM_FILTER Enable checksum filtering to MNG When this bit is set, only packets that pass L3, L4 checksum are sent to the MNG block. 6:4 Reserved Reserved should be 0 2 KEEP_PHY_LINK_UP Block PHY reset and power state changes. When this bit is set the PHY reset and power state changes do not reach to the PHY (PHY is not reset), This bit can not be written to, unless the Keep_PHY_Link_Up_En EEPROM bit is set. 1 RCV_TCO_EN Enable BMC to network and network to BMC traffic 0b - The BMC can not communicate with the network. 1b - The BMC can communicate with the network When cleared the BMC traffic is not forwarded to the network and the network traffic is not forwarded to the BMC even if the decision filters indicates it should. This bit does not impact the OS to BMC traffic. 0 TCO_RESET TCO Reset occurred Note: Value placed in this field is not written to MANC register. This value will be stored in the LSB Portion of the LAN1 Management Control (MANC) register. Bit Name Description 15 TCO_Isolate TCO Isolate command. Note: Value placed in this field is not written to MANC register. 14 FW_RESET FW Reset occurred. Note: Value placed in this field is not written to MANC register. 13:0 Reserved Reserved. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1085 This value will be stored in the MSB Portion of the LAN1 Management Control (MANC) register. Bit Name Description 15:11 Reserved Reserved. 10 NET_TYPE NET TYPE: 0b = pass only un-tagged packets. 1b = pass only VLAN tagged packets. Valid only if FIXED_NET_TYPE is set. 9 FIXED_NET_TYPE Fixed net type: If set, only packets matching the net type defined by the NET_TYPE field passes to manageability. Otherwise, both tagged and un-tagged packets can be forwarded to the manageability engine. 8 EN_IPv4_FILTER Enable IPv4 address Filters - when set, the last 128 bits of the MIPAF register are used to store 4 IPv4 addresses for IPv4 filtering. When cleared, these bits store a single IPv6 filter. 7 EN_XSUM_FILTER Enable checksum filtering to MNG When this bit is set, only packets that pass L3, L4 checksum are sent to the MNG block. 6:4 Reserved Reserved should be 0 KEEP_PHY_LINK_UP Block PHY reset and power state changes. When this bit is set the PHY reset and power state changes do not reach to the PHY (PHY is not reset), This bit can not be written to, unless the Keep_PHY_Link_Up_En EEPROM bit is set. 1 RCV_TCO_EN Enable BMC to network and network to BMC traffic The BMC can not communicate with the network. The BMC can communicate with the network When cleared the BMC traffic is not forwarded to the network and the network traffic is not forwarded to the BMC even if the decision filters indicates it should. This bit does not impact the OS to BMC traffic. 0 TCO_RESET TCO Reset occurred Note: Value placed in this field is not written to MANC register. 2 This value will be stored in the LSB Portion of the LAN2 Management Control (MANC) register. Bit Name Description 15 TCO_Isolate TCO Isolate command. Note: Value placed in this field is not written to MANC register. 14 FW_RESET FW Reset occurred. Note: Value placed in this field is not written to MANC register. 13:0 Reserved Reserved. This value will be stored in the MSB Portion of the LAN2 Management Control (MANC) register. Bit Name Description 15:11 Reserved Reserved. 10 NET_TYPE NET TYPE: 0b = pass only un-tagged packets. 1b = pass only VLAN tagged packets. Valid only if FIXED_NET_TYPE is set. 9 FIXED_NET_TYPE Fixed net type: If set, only packets matching the net type defined by the NET_TYPE field passes to manageability. Otherwise, both tagged and un-tagged packets can be forwarded to the manageability engine. Intel(R) Communications Chipset 89xx Series - Datasheet 1086 October 2012 Order Number: 327879-001US 24.0 Bit Name Description 8 EN_IPv4_FILTER Enable IPv4 address Filters - when set, the last 128 bits of the MIPAF register are used to store 4 IPv4 addresses for IPv4 filtering. When cleared, these bits store a single IPv6 filter. 7 EN_XSUM_FILTER Enable checksum filtering to MNG When this bit is set, only packets that pass L3, L4 checksum are sent to the MNG block. 6:4 Reserved Reserved should be 0 2 KEEP_PHY_LINK_UP Block PHY reset and power state changes. When this bit is set the PHY reset and power state changes do not reach to the PHY (PHY is not reset), This bit can not be written to, unless the Keep_PHY_Link_Up_En EEPROM bit is set. 1 RCV_TCO_EN Enable BMC to network and network to BMC traffic The BMC can not communicate with the network. The BMC can communicate with the network When cleared the BMC traffic is not forwarded to the network and the network traffic is not forwarded to the BMC even if the decision filters indicates it should. This bit does not impact the OS to BMC traffic. 0 TCO_RESET TCO Reset occurred Note: Value placed in this field is not written to MANC register. This value will be stored in the LSB Portion of the LAN3 Management Control (MANC) register. Bit Name Description 15 TCO_Isolate TCO Isolate command. Note: Value placed in this field is not written to MANC register. 14 FW_RESET FW Reset occurred. Note: Value placed in this field is not written to MANC register. 13:0 Reserved Reserved. This value will be stored in the MSB Portion of the LAN3 Management Control (MANC) register. Bit 15:11 Name Description Reserved Reserved. 10 NET_TYPE NET TYPE: 0b = pass only un-tagged packets. 1b = pass only VLAN tagged packets. Valid only if FIXED_NET_TYPE is set. 9 FIXED_NET_TYPE Fixed net type: If set, only packets matching the net type defined by the NET_TYPE field passes to manageability. Otherwise, both tagged and un-tagged packets can be forwarded to the manageability engine. 8 EN_IPv4_FILTER Enable IPv4 address Filters - when set, the last 128 bits of the MIPAF register are used to store 4 IPv4 addresses for IPv4 filtering. When cleared, these bits store a single IPv6 filter. 7 EN_XSUM_FILTER Enable checksum filtering to MNG When this bit is set, only packets that pass L3, L4 checksum are sent to the MNG block. 6:4 Reserved Reserved should be 0 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1087 Bit Name Description KEEP_PHY_LINK_UP This bit gates the ability for MNG to be able to modify bit 18 (KEEP_PHY_LINK_UP) in Management Control Register - MANC [0:3] (0x5820; RW). If this bit is set, MNG can modify bit 18 in MANCx, otherwise, bit 18 of MANCx cannot be modified by MNG. 1 RCV_TCO_EN Enable BMC to network and network to BMC traffic The BMC can not communicate with the network. The BMC can communicate with the network When cleared the BMC traffic is not forwarded to the network and the network traffic is not forwarded to the BMC even if the decision filters indicates it should. This bit does not impact the OS to BMC traffic. 0 TCO_RESET TCO Reset occurred Note: Value placed in this field is not written to MANC register. 2 24.11 Software Accessed Words Words 0x03 to 0x07 in the EEPROM image are reserved for compatibility information. New bits within these fields will be defined as the need arises for determining software compatibility between various hardware revisions. Words 0x8 and 0x09 are used to indicate the Printed Board Assembly (PBA) number and words 0x42 and 0x43 identifies the EEPROM image. Words 0x30 to 0x3E have been reserved for configuration and version values to be used by PXE code. The only exception is word 0x37 used for Pointer to Alternate MAC address. 24.11.1 Compatibility (Word 0x03) Bit Description 15:12 Reserved (set to 0000b). 11 LOM/Not a LOM. 0b = NIC. 1b = LOM. 10 Server/Not a Server NIC. 0b = Client. 1b = Server. 9 Client/Not a Client NIC. 0b = Server. 1b = Client. 8 Retail/OEM. 0b = Retail. 1b = OEM. 7:6 Reserved (set to 00b). 5 Reserved (set to 1b). 4 SMBus Connected. 0b = Not connected. 1b = Connected. Intel(R) Communications Chipset 89xx Series - Datasheet 1088 October 2012 Order Number: 327879-001US 24.0 Bit 24.11.2 Description 3 Reserved (set to 0b). 2 Reserved (set to 0b). 1:0 Reserved (set to 00b) Port Identification LED Blinking (Word 0x04) Driver Software provides a method to identify an external port on a system through a command that causes the LED to blink. Based on the setting in word 0x4 the LED should blink between STATE1 and STATE2 when a port identification command is issued. STATE1 and STATE2 are software driver states between which software toggles and the table shown below indicates how the LED is driven by software within each state. When word 0x4 is equal to 0xFFFF the blinking behavior revert to a default. Bit 15:12 Description Control for LED 3. 0001b = Default in STATE1 + Default in STATE2. 0010b = Default in STATE1 + LED is ON in STATE2. 0011b = Default in STATE1 + LED is OFF in STATE2. 0100b = LED is ON in STATE1 + Default in STATE2. 0101b = LED is ON in STATE1 + LED is ON in STATE2. 0110b = LED is ON in STATE1 + LED is OFF in STATE2. 0111b = LED is OFF in STATE1 + Default in STATE2. 1000b = LED is OFF in STATE1 + LED is ON in STATE2. 1001b = LED is OFF in STATE1 + LED is OFF in STATE2. Note: Default is the LED behavior in STATE1 or STATE2 based on the value indicated in LED Control Register (LEDCTL) bits[3:0]. 24.11.3 11:8 Control for LED 2 - same encoding as for LED 3. 7:4 Control for LED 1 - same encoding as for LED 3. 3:0 Control for LED 0 - same encoding as for LED 3. OEM Specific (Word 0x06, 0x07) These words are available for OEM use. 24.11.4 EEPROM Image Revision (Word 0x05) This word is valid only for device starter images and indicates the ID and version of the EEPROM image. This word is used for tracking the EEPROM image revision and does not have any effect on the hardware functionality. Bit 15:12 Description EEPROM major version. 11:4 EEPROM minor version. 3:0 EEPROM image ID. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1089 24.11.5 PBA Number (Word 0x09) Word 0x09 contains the pointer to the PBA structure. Bit 15:00 Description Pointer to PBA Structure 24.11.5.1 PBA Structure The Printed Board Assembly (PBA) number that are stored in a multi-byte field. The purpose of this information is to allow customer support (or any user) to identify the exact revision level of a product. Network driver software should not rely on this field to identify the product or its capabilities. Note: This PBA number is not related to the MSI-X Pending Bit Array (PBA). 24.11.6 PXE Main Setup Options (Word 0x30/0x34/0x38/0x3A) The configuration of the PXE software is controlled by EEPROM on the adapter. The main setup options are stored in word 0x30 an word 0x34. These options are those that can be changed by the user via the Control-S setup menu. The default value for this word is 0x0100. The signature bits in Word 0x31 will be checked by software to determine whether the contents of this word are valid. Word 0x30 is used to configure the LAN0 port, word 0x34 is used to configure the LAN1 port, word 0x38 is used to configure the LAN2 port, and word 0x3A is used to configure the LAN3 port. These Words are only mapped in GbE0 space (Function 1). Bit Description 15:13 Reserved. Must be 0. 12:10 Bits 12-10 control forcing speed and duplex during driver operation. Valid values are: 000b - Auto negotiate 001b - 10Mbps Half Duplex 010b - 100Mbps Half Duplex 011b - Not valid (treated as 000b) 100b - 10Mbps Full Duplex 101b - 100Mbps Full Duplex 111b - 1000Mbps Full Duplex Only applicable for copper-based adapters. Default value is 000b. 9 Reserved 8 Display Setup Message. If the bit is set to 1, the "Press Control-S" message is displayed after the title message. Default value is 1. 7:6 Prompt Time. These bits control how long the "Press Control-S" setup prompt message is displayed, if enabled by bit 8 00 = 2 seconds (default) 01 = 3 seconds 10 = 5 seconds 11 = 0 seconds Note: The Ctrl-S message is not displayed if 0 seconds prompt time is selected. 5 Deprecated. Must be 0. Intel(R) Communications Chipset 89xx Series - Datasheet 1090 October 2012 Order Number: 327879-001US 24.0 Bit 24.11.7 Description 4:3 Default Boot Selection. These bits select which device is the default boot device. These bits are only used if the agent detects that the BIOS does not support boot order selection or if the MODE field of word 31h is set to .Legacy mode 00 = Network boot, then local boot (default) 01 = Local boot, then network boot 10 = Network boot only 11 = Local boot only 2 Deprecated. Must be 0. 1:0 Protocol Select. These bits select the active boot protocol: 00 = PXE (default value) 01 = Reserved 10 = Reserved 11 = Reserved Only the default value of 00b should be initially programmed into the adapter; other values should only be set by configuration utilities. PXE Configuration Customization Options (Word 0x31/0x35/0x39/0x3B) Word 0x31 and word 0x35 of the EEPROM contains settings that can be programmed by an OEM or network administrator to customize the operation of the software. These settings cannot be changed from within the Control-S setup menu. The lower byte contains settings that would typically be configured by a network administrator using an external utility; these settings generally control which setup menu options are changeable. The upper byte is generally settings that would be used by an OEM to control the operation of the agent in a LOM environment, although there is nothing in the agent to prevent their use on a NIC implementation. The default value for this word is 0x4000. Word 0x31 is used to configure the primary port and word 0x35 is used to configure the secondary port, word 0x39 is used to configure the LAN2 port, and word 0x3B is used to configure the LAN3 port. These Words are only mapped in GbE0 space (Function 1). Bit Description 15:14 Signature. Must be set to 01 to indicate that this word has been programmed by the agent or other configuration software. 13:12 Reserved - must be 0 11 Selects Continuous Retry operation. If this bit is set, IBA will NOT transfer control back to the BIOS if it fails to boot due to a network error (such as failure to receive DHCP replies). Instead, it will restart the PXE boot process again. If this bit is set, the only way to cancel PXE boot is for the user to press ESC on the keyboard. Retry will not be attempted due to hardware conditions such as an invalid EEPROM checksum or failing to establish link. Default value is 0. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1091 Bit 24.11.8 Description 10:8 Selects the agent's boot order setup mode. This field changes the agent's default behavior in order to make it compatible with systems that do not completely support the BBS and PnP Expansion ROM standards. Valid values and their meanings are: 000b Normal behavior. The agent will attempt to detect BBS and PnP Expansion ROM support as it normally does. (default) 001b Force Legacy mode. The agent will not attempt to detect BBS or PnP Expansion ROM supports in the BIOS and will assume the BIOS is not compliant. The user can change the BIOS boot order in the Setup Menu. 010b Force BBS mode. The agent will assume the BIOS is BBS-compliant, even though it may not be detected as such by the agent's detection code. The user can NOT change the BIOS boot order in the Setup Menu. 011b Force PnP Int18 mode. The agent will assume the BIOS allows boot order setup for PnP Expansion ROMs and will hook interrupt 18h (to inform the BIOS that the agent is a bootable device) in addition to registering as a BBS IPL device. The user can NOT change the BIOS boot order in the Setup Menu. 100b Force PnP Int19 mode. The agent will assume the BIOS allows boot order setup for PnP Expansion ROMs and will hook interrupt 19h (to inform the BIOS that the agent is a bootable device) in addition to registering as a BBS IPL device. The user can NOT change the BIOS boot order in the Setup Menu. 101b - 111b Reserved for future use. If specified, is treated as a value of 000b. 7:6 Reserved - must be 0 5 Reserved 4 Disable Legacy Wakeup Support. If this bit is set to 1, the user is not allowed to change the Legacy OS Wakeup Support menu option. Default value is 0. 3 Disable Boot Selection. If this bit is set to 1, the user is not allowed to change the boot order menu option. Default value is 0. 2 Disable Protocol Select. If set to 1, the user is not allowed to change the boot protocol. Default value is 0. 1 Disable Title Message. If this bit is set to 1, the title message displaying the version of the Boot Agent is suppressed; the Control-S message is also suppressed. This is for OEMs who do not wish the boot agent to display any messages at system boot. Default value is 0. 0 Disable Setup Menu. If this bit is set to 1, the user is not allowed to invoke the setup menu by pressing Control-S. In this case, the EEPROM may only be changed via an external program. Default value is 0. PXE Boot Agent Version Number (Word 0x32) Word 32h of the EEPROM is used to store the version of the boot agent that is stored in the flash image. Some older diagnostic tools read this word in order to report the version of the Boot Agent in the flash. Note: PCH does not support Flash so this word is not valid. 24.11.9 IBA Capabilities (Word 0x33) Word 33h of the EEPROM is used to enumerate the boot technologies that have been programmed into the flash. This is updated by flash configuration tools and is not updated or read by IBA. Note: PCH does not support Flash so this word is not valid. Intel(R) Communications Chipset 89xx Series - Datasheet 1092 October 2012 Order Number: 327879-001US 24.0 24.11.10 PXE Reserved Words (Words 0x36, 0x3C, 0x3E) These words are reserved for other PXE usages. 24.11.11 Alternate MAC Address Pointer (Word 0x37) This word may point to a location in the EEPROM containing additional MAC addresses used by system management functions. If the additional MAC addresses are not supported, the word shall be set to 0xFFFF. The structure of the alternate MAC address block can be found in Table 24-14. Table 24-14. Alternate MAC Address Block Description1 Word Offset 0x0 ... 0x2 Alternate MAC Address for LAN port assigned to PCI function 1 0x3 ... 0x5 Alternate MAC Address for LAN port assigned to PCI function 2 0x6 ... 0x8 Alternate MAC Address for LAN port assigned to PCI function 3 0x9 ... 0xB Alternate MAC Address for LAN port assigned to PCI function 4 1. An alternate MAC Address value of 0xFFFF-FFFF-FFFF means that no alternate MAC address is present for the port. 24.11.12 Checksum Word (Offset 0x3F) The checksum words (Offset 0x3F from start of Common, LAN 1, LAN 2 and LAN 3 sections) are used to ensure that the base EEPROM image is a valid image. The value of this word should be calculated such that after adding all the words (0x00:0x3E), including the checksum word itself, the sum should be 0xBABA. The initial value in the 16-bit summing register should be 0x0000 and the carry bit should be ignored after each addition. Note: Hardware does not calculate the checksum word during EEPROM write; it must be calculated by software independently and included in the EEPROM write data. Hardware does not compute a checksum over words 0x00:0x3F during EEPROM reads in order to determine validity of the EEPROM image; this field is provided strictly for software verification of EEPROM validity. All hardware configurations based on word 0x00:0x3F content is based on the validity of the Signature field of the EEPROM Sizing EEPROM word (Signature must be 01b). 24.11.13 Image Unique ID (Word 0x42, 0x43) These words contain a unique 32-bit ID for each image generated by Intel to enable tracking of images and comparison to the original image if testing a customer EEPROM image. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1093 25.0 GbE Power Management This section describes how power management is implemented in the GbE Controller. The Controller supports the Advanced Configuration and Power Interface (ACPI) specification as well as Advanced Power Management (APM). Note: Power management can be disabled via bits in the Initialization Control Words, which is loaded from the EEPROM during power-up reset. Even when disabled, the power management register set is still present. Power management support is required by the PCIe specification. 25.1 General Power State Information 25.2 Power States The Controller supports D0 and D3 (D3 Hot and D3 Cold) power states defined in the PCI Power Management and PCIe specifications. D0 is divided into two sub-states: D0u (D0 un-initialized), and D0a (D0 active). In addition, the Controller supports a Dr state that is entered when PCIE_EP_RST# is asserted (including the D3cold state). Figure 25-1 shows the power states and transitions between them. Figure 25-1. Power Management State Diagram Inte rn a l P o w e r O n R e se t a sse rtio n H o t (in -b a n d ) R e se t P E R S T # de a ssertio n & E E P R O M re a d done Dr D0u PERST# a sse rtio n W rite 1 1 b to P o w e r S ta te E n a b le m a ste r o r sla ve a cce ss PERST# a sse rtio n W rite 0 0 b to P o w e r S ta te D3 Intel(R) Communications Chipset 89xx Series - Datasheet 1094 PERST# a sse rtio n W rite 1 1 b to P ow e r S ta te D0a October 2012 Order Number: 327879-001US 25.0 25.2.1 D0 Uninitialized (D0u) State The D0u state is a low-power state used after GBE_AUX_PWR_OK (PCIE_EP_RST#) is de-asserted following power-up (cold or warm), on hot reset (in-band reset through PCIe physical layer message), or on D3 exit. When entering D0u, the Controller disables wake ups and asserts a reset to the PHY while the EEPROM is being read. If the APM Mode bit in the EEPROM's Initialization Control Word 2 is set, then APM wake up is enabled. 25.2.1.1 Entry Into D0u State D0u is reached from either the Dr state (on de-assertion of PCIE_EP_RST#) or the D3hot state (by configuration software writing a value of 00b to the Power State field of the PCI PM registers). De-asserting PCIE_EP_RST# means that the entire state of the Controller is cleared, other than sticky bits. State is loaded from the EEPROM, followed by establishment of the PCIe link. Once this is done, configuration software can access the Controller. On a transition from D3 to D0u state, the Controller PCI configuration space is not reset. However, the Controller requires that software perform a full re-initialization of the function including its PCI configuration space. 25.2.2 D0 Active (D0a) State Once memory space is enabled, the Controller enters an active state. It can transmit and receive packets if properly configured by the software device driver. The PHY is enabled or re-enabled by the software device driver to operate/auto-negotiate to full line speed/power if not already operating at full capability. Any APM wake up previously active remains active. The software device driver can deactivate APM wake up by writing to the Wake Up Control (WUC) register or activate other wake-up filters by writing to the Wake Up Filter Control (WUFC) register. 25.2.2.1 Entry to D0a State D0a is entered from the D0u state by writing a 1b to the Memory Access Enable or the I/O Access Enable bit of the PCI Command register. The DMA and MAC of the appropriate LAN function are also enabled. 25.2.3 D3 State (PCI-PM D3hot) The Controller transitions to D3 when the system writes a 11b to the Power State field of the Power Management Control/Status Register (PMCSR). Any wake-up filter settings that were enabled before entering this reset state are maintained. Upon completion or during the transition to D3 state, the Controller clears the Memory Access Enable and I/O Access Enable bits of the PCI Command register, which disables memory access decode. In D3, the Controller only responds to PCI configuration accesses and does not generate master cycles. Configuration and message requests are the only TLPs accepted by a function in the D3hot state. All other received requests must be handled as unsupported requests, and all received completions can optionally be handled as unexpected completions. If an error caused by a received TLP (such as an unsupported request) is detected while in D3hot, and reporting is enabled, the link must be returned to L0 if it is not already in L0 and an error message must be sent. See section 5.3.1.4.1 in the PCIe Base Specification October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1095 A D3 state is followed by either a D0u state (in preparation for a D0a state) or by a transition to Dr state (PCI-PM D3cold state). To transition back to D0u, the system writes a 00b to the Power State field of the Power Management Control/Status Register (PMCSR). Transition to Dr state is through PCIE_EP_RST# assertion. 25.2.3.1 Entry to D3 State Transition to D3 state is through a configuration write to the Power State field of the PCI-PM registers. Prior to transition from D0 to the D3 state, the software device driver disables scheduling of further tasks to the Controller; it masks all interrupts and does not write to the Transmit Descriptor Tail (TDT) register or to the Receive Descriptor Tail (RDT) register and operates the master disable algorithm as defined in Section 25.2.3.2. If wake up capability is needed, the software device driver should set up the appropriate wake up registers and then the system writes a 1b to the PME_En bit of the Power Management Control / Status Register (PMCSR) or to the Auxiliary (AUX) Power PM Enable bit of the PCIe Device Control register prior to the transition to D3. As a response to being programmed into D3 state, the Controller suspends scheduling new transactions. The PCIe EP discards all the completions destined for the Controller and sends dummy completions to the Controller for active read requests. The EP also clears Memory Access Enable and I/O Access Enable bits of the PCIe Command register which disables memory access decode. Any Receive packets that have not been put in system memory are kept and discarded upon D3 exit. Any transmit packet not yet transmitted can be transmitted or dropped depending on their progress made in the Controller. 25.2.3.2 Master Disable Via CTRL Register System software can disable master accesses on the PCIe link by either clearing the PCI Bus Master bit or by bringing the function into a D3 state. From that time on, the Controller must not issue master accesses for this function. Due to the full-duplex nature of PCIe, and the pipelined design in the Controller, it might happen that multiple requests from several functions are pending when the master disable request arrives. The protocol described in this section insures that a function does not issue master requests to the PCIe link after its Master Enable bit is cleared (or after entry to D3 state). Two configuration bits are provided for the handshake between the Controller function and its software device driver: * GIO Master Disable bit in the Device Control (CTRL) register - When the GIO Master Disable bit is set, the Controller blocks new master requests by this function. The Controller then proceeds to issue any pending requests by this function. This bit is cleared on master reset (LAN_PWR_GOOD, PCIe reset and software reset) to enable master accesses. * GIO Master Enable Status bit in the Device Status (STATUS) register - Cleared by the Controller when the GIO Master Disable bit is set and no master requests are pending by the relevant function and is set otherwise. Indicates that no master requests are issued by this function as long as the GIO Master Disable bit is set. The following activities must end before the Controller clears the GIO Master Enable Status bit: -- Master requests by the transmit and receive engines (for both data and MSI/MSIx interrupts). -- All pending completions to the Controller are received. Intel(R) Communications Chipset 89xx Series - Datasheet 1096 October 2012 Order Number: 327879-001US 25.0 In the event of a PCIe Master disable (Configuration Command register.BME set to 0) on a certain function or LAN port or if the function is moved into D3 state during a DMA access, the Controller generates an internal reset to the function and stops all port DMA accesses and interrupts related to the function. Following move to normal operating mode software driver should re-initialize the receive and transmit queues of the relevant port. Notes: The software device driver sets the GIO Master Disable bit when notified of a pending master disable (or D3 entry). The Controller then blocks new requests and proceeds to issue any pending requests by this function. The software device driver then polls the GIO Master Enable Status bit. Once the bit is cleared, it is guaranteed that no requests are pending from this function. The software device driver might time out if the GIO Master Enable Status bit is not cleared within a given time. The GIO Master Disable bit must be cleared to enable a master request to the PCIe link. This can be done either through reset or by the software device driver. 25.2.4 Dr State (D3cold) Transition to Dr state is initiated on several occasions: * On system power up - Dr state begins with the assertion of the internal power detection circuit and ends with de-assertion of PCIE_EP_RST#. * On transition from a D0a state - During operation the system might assert PCIE_EP_RST# at any time. In an ACPI system, a system transition to the G2/S5 state causes a transition from D0a to Dr state. * On transition from a D3 state - The system transitions the Controller into the Dr state by asserting PCIe PCIE_EP_RST#. Any wake-up filter settings that were enabled before entering this reset state are maintained. The system might maintain PCIE_EP_RST# asserted for an arbitrary time. The de-assertion (rising edge) of PCIE_EP_RST# causes a transition to D0u state. While in Dr state, the Controller might enter one of several modes with different levels of functionality and power consumption. The lower-power modes are achieved when the Controller is not required to maintain any functionality (see Section 25.2.4.1). 25.2.4.1 Dr Disable Mode The Controller enters a Dr disable mode on transition to D3cold state when it does not need to maintain any functionality. The conditions to enter either state are: * The Controller (all PCI functions) is in Dr state * APM WOL is inactive for all LAN functions * Pass-through manageability is disabled * ACPI PME is disabled for all PCI functions * The Controller Power Down En EEPROM bit (word 0x1E, bit 15) is set (default hardware value is disabled). Entering Dr disable mode is usually done by asserting PCIe PCIE_EP_RST#. It might also be possible to enter Dr disable mode by reading the EEPROM while already in Dr state. The usage model for this later case is on system power up, assuming that manageability and wake up are not required. Once the Controller enters Dr state on power-up, the EEPROM is read. If the EEPROM contents determine that the conditions October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1097 to enter Dr disable mode are met, the Controller then enters this mode (assuming that PCIe PCIE_EP_RST# is still asserted). Note: Exiting Dr disable mode can be done by de-asserting PCIE_EP_RST#. 25.2.4.2 Entry to Dr State Dr entry on platform power-up begins with the assertion of AUX_PWR_GOOD. The EEPROM is read and determines the Controller configuration. If the APM Enable bit in the EEPROM's Initialization Control Word 3 is set, then APM wake up is enabled. Entering Dr state from D0a state can be done by asserting PCIE_EP_RST#. An ACPI transition to the G2/S5 state is reflected in the the Controller transition from D0a to Dr state. The transition can be orderly (such as, user selecting the shut down option), in which case the software device driver might have a chance to intervene. Or, it might be an emergency transition (such as power button override), in which case, the software device driver is not notified. Note: Transition from D3 (hot) state to Dr state is done by asserting PCIE_EP_RST#. 25.2.4.3 Auxiliary Power Usage If the D3COLD_WAKEUP_ADVEN bit in the Software Defined Pins Control EEPROM word is set to 1, the Controller uses the AUX_PWR indication that auxiliary power is available to it, and therefore advertises D3cold wake up support. If D3cold is supported, the PME_En and PME_Status bits of the Power Management Control/Status Register (PMCSR), as well as their shadow bits in the Wake Up Control (WUC) register are reset only by the power up reset (detection of power rising). The only effect of setting AUX_PWR to 1b is advertising D3cold wake-up support and changing the functionality in reset of PME_En and PME_Status. 25.2.5 Device Power-Down State The Controller enters a global power-down state if all of the following conditions are met: * The Controller Power Down Enable EEPROM bit (word 0x1E bit 15) was set (default hardware value is disabled). * The Controller is in Dr state. * The link connections of all ports (SerDes) are in power down mode. 25.3 Timing of Power-State Transitions Note: All reference to PHY in this section also apply to SGMII/SERDES interface. The Controller in the PCH does not have an integrated PHY but all the signals mentioned in this section also apply to SERDES/SGMII. The following sections give detailed timing for the state transitions. In the diagrams the dotted connecting lines represent the Controller requirements, while the solid connecting lines represent the Controller guarantees. The timing diagrams are not to scale. The clocks edges are shown to indicate running clocks only and are not to be used to indicate the actual number of cycles for any operation. Intel(R) Communications Chipset 89xx Series - Datasheet 1098 October 2012 Order Number: 327879-001US 25.0 25.3.1 Power Up (Off to Dup to D0u to D0a) Figure 25-2. Power Up (Off to Dup to D0u to D0a) power txog 1 Xosc Internal Power On Reset tppg 2 tPVPG L 6 PCIe reference clock tPWRGD-CLK 7 PE_RSTn tppg-clkint tclkpr 9 Internal PwrGd (PLL) tee 3 Reading EEPROM Read 3GIO tee 10 Read rest Read 3GIO Read rest 11 tpgcfg tpgtrn 12 14 15 L0 Internal Bus Link up Wakeup Enabled tpgres 13 5 APM/SMBus Wakeup DState 5 APM/SMBus Wakeup Dr D0u D0a Table 25-1. Power Up (Off to Dup to D0u to D0a) Note 1 Description Xosc is stable txog after power is stable. 2 LAN_PWR_GOOD is asserted after all power supplies are good and tppg after Xosc is stable. 3 An EEPROM read starts on the rising edge of LAN_PWR_GOOD. 5 APM wake-up mode can be enabled based on what is read from the EEPROM. 7 PCIE_EP_RST# is de-asserted tPVPGL after power is stable (according to PCIe specification). 9 The PCIe internal PWRGD signal is asserted tclkpr after the external PCIE_EP_RST# signal. 10 Asserting internal PCIe PWRGD causes the EEPROM to be re-read, asserts PHY reset, and disables wake up. 11 After reading the EEPROM, PHY reset is de-asserted. 12 Link training starts after tpgtrn from PCIE_EP_RST# de-assertion. 13 A first PCIe configuration access might arrive after tpgcfg from PCIE_EP_RST# de-assertion. 14 A first PCI configuration response can be sent after tpgres from PCIE_EP_RST# de-assertion. 15 Writing a 1b to the Memory Access Enable bit in the PCI Command Register transitions the Controller from D0u to D0. state. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1099 25.3.2 Transition From D0a to D3 and Back Without PCIE_EP_RST# Figure 25-3. Transition From D0a to D3 and Back Without PCIE_EP_RST# PCIe Reference Clock PE_RSTn D0 Write 3 2 PHY Reset PCIe link td0mem tee Reading EEPROM Memory Access Enable 6 Read EEPROM 5 D3 write 7 1 L0 L0 L1 4 Any mode Wakeup Enabled PHY Power State DState full D0a APM / SMBus power-managed D3 power-managed D0u full D0 Table 25-2. Transition From D0a to D3 and Back Without PCIE_EP_RST# Note Description 1 Writing 11b to the Power State field of the Power Management Control/Status Register (PMCSR) transitions the Controller to D3. 2 The system can keep the Controller in D3 state for an arbitrary amount of time. 3 To exit D3 state, the system writes 00b to the Power State field of the PMCSR. 4 APM wake-up or SMBus mode might be enabled based on what is read in the EEPROM. 5 After reading the EEPROM, reset to the PHY is de-asserted. The PHY operates at reduced-speed if APM wake up or SMBus is enabled, else powered-down. 6 The system can delay an arbitrary time before enabling memory access. 7 Writing a 1b to the Memory Access Enable bit or to the I/O Access Enable bit in the PCI Command Register transitions the Controller from D0u to D0 state and returns the PHY to full-power/speed operation. Intel(R) Communications Chipset 89xx Series - Datasheet 1100 October 2012 Order Number: 327879-001US 25.0 25.3.3 Transition From D0a to D3 and Back With PCIE_EP_RST# Figure 25-4. Transition From D0a to D3 and Back With PCIE_EP_RST# 4a PCIe reference clock tclkpg PE_RSTn 4b tPWRGD-CLK tl2clk 6 tpgdl 3 tl2pg tppg-clkint 7 Internal PCIe clock) 2.5Ghz( 8 tclkpr Internal PwrGd (PLL) 9 tee 5 Reading EEPROM Read EEPROM 11 Reset to PHY (active low) D3 write PCIe Link L0 tpgtrn 2 1 L1 tpgcfg 12 L0 tpgres 14 13 L/2L3 15 L0 10 Wakeup Enabled Any mode PHY Power State full DState D0a APM/SMBus power-managed D3 full D0u Dr D0a Table 25-3. Transition From D0a to D3 and Back With PCIE_EP_RST# Note Description 1 Writing 11b to the Power State field of the PMCSR transitions the Controller to D3. PCIe link transitions to L1 state. 2 The system can delay an arbitrary amount of time between setting D3 mode and transitioning the link to an L2 or L3 state. 3 Following link transition, PCIE_EP_RST# is asserted. 4 The system must assert PCIE_EP_RST# before stopping the PCIe reference clock. It must also wait tl2clk after link transition to L2/L3 before stopping the reference clock. 5 On assertion of PCIE_EP_RST#, the Controller transitions to Dr state. 6 The system starts the PCIe reference clock tPE_RST-CLK before de-assertion PCIE_EP_RST#. 7 The internal PCIe clock is valid and stable tppg-clkint from PCIE_EP_RST# de-assertion. 8 The PCIe internal PWRGD signal is asserted tclkpr after the external PCIE_EP_RST# signal. 9 Asserting internal PCIe PWRGD causes the EEPROM to be re-read, asserts PHY reset, and disables wake up. 10 APM wake-up mode might be enabled based on what is read from the EEPROM. 11 After reading the EEPROM, PHY reset is de-asserted. 12 Link training starts after tpgtrn from PCIE_EP_RST# de-assertion. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1101 Table 25-3. Transition From D0a to D3 and Back With PCIE_EP_RST# (Continued) Note Description 13 A first PCIe configuration access might arrive after tpgcfg from PCIE_EP_RST# de-assertion. 14 A first PCI configuration response can be sent after tpgres from PCIE_EP_RST# de-assertion. 15 Writing a 1b to the Memory Access Enable bit in the PCI Command Register transitions the Controller from D0u to D0 state. 25.3.4 Transition From D0a to Dr and Back Without Transition to D3 Figure 25-5. Transition From D0a to Dr and Back Without Transition to D3 1 PCIe Reference Clock tclkpg tPWRGD-CLK 3 PE_RSTn tpgdl tppg-clkint 4 Internal PCIe Clock ( 2.5 GHz) 5 tclkpr Internal PwrGd (PLL) 6 tee 2 Reading EEPROM Read EEPROM 8 Reset to PHY (active low) tpgtrn 9 PCIe Link tpgcfg tpgres 11 10 L0 12 L0 7 Wakeup Enabled Any mode PHY Power State DState full APM/SMBus power-managed D0a full D0u Dr D0a Table 25-4. Transition From D0a to Dr and Back Without Transition to D3 Note Description 1 The system must assert PCIE_EP_RST# before stopping the PCIe reference clock. It must also wait tl2clk after link transition to L2/L3 before stopping the reference clock. 2 On assertion of PCIE_EP_RST#, the Controller transitions to Dr state and the PCIe link transition to electrical idle. 3 The system starts the PCIe reference clock tPE_RST-CLK before de-assertion PCIE_EP_RST#. 4 The internal PCIe clock is valid and stable tppg-clkint from PCIE_EP_RST# de-assertion. 5 The PCIe internal PWRGD signal is asserted tclkpr after the external PCIE_EP_RST# signal. 6 Asserting internal PCIe PWRGD causes the EEPROM to be re-read, asserts PHY reset, and disables wake up. 7 APM wake-up mode might be enabled based on what is read from the EEPROM. Intel(R) Communications Chipset 89xx Series - Datasheet 1102 October 2012 Order Number: 327879-001US 25.0 Table 25-4. Transition From D0a to Dr and Back Without Transition to D3 (Continued) Note Description 8 After reading the EEPROM, PHY reset is de-asserted. 9 Link training starts after tpgtrn from PCIE_EP_RST# de-assertion. 10 A first PCIe configuration access might arrive after tpgcfg from PCIE_EP_RST# de-assertion. 11 A first PCI configuration response can be sent after tpgres from PCIE_EP_RST# de-assertion. 12 Writing a 1b to the Memory Access Enable bit in the PCI Command Register transitions the Controller from D0u to D0 state. 25.4 Wake Up The Controller supports two modes of wake-up management: 1. Advanced Power Management (APM) wake up 2. ACPI/PCIe defined wake up The usual model is to activate one mode at a time but not both modes together. If both modes are activated, the Controller might wake up the system in unexpected events. For example, if APM is enabled together with PCIe PME, a magic packet might wake up the system even if APMPME is disabled. Alternatively, if APM is enabled together with some PCIe filters, packets matching these filters might wake up the system even if PCIe PME is disabled. 25.4.1 Advanced Power Management Wake Up Advanced Power Management Wake Up or APM Wakeup (also known as Wake on LAN) is a feature that existed in earlier 10/100 Mb/s NICs. This functionality was designed to receive a broadcast or unicast packet with an explicit data pattern, and then assert a subsequent signal to wake up the system. This was accomplished by using a special signal that ran across a cable to a defined connector on the motherboard. The NIC would assert the signal for approximately 50 ms to signal a wake up. The Controller now uses (if configured) an in-band PM_PME message for this functionality. On power up, the Controller reads the APM Enable bits from the EEPROM Initialization Control Word 3 into the APM Enable (APME) bits of the Wakeup Control (WUC) register. These bits control enabling of APM wake up. When APM wake up is enabled, the Controller checks all incoming packets for Magic Packets. See Section 25.4.3.1.4 for a definition of Magic Packets. Once the Controller receives a matching magic packet, and if the Assert PME On APM Wakeup (WUC.APMPME) bit is set in the Wake Up Control (WUC) register, it: * Sets the PME_Status bit in the PMCSR register and issues a PM_PME message (in most cases, this might require asserting the PE_WAKE_N signal first to resume power and clock to the PCIe interface). * Stores the first 128 bytes of the packet in the Wake Up Packet Memory (WUPM) register. * Sets the Magic Packet Received bit in the Wake Up Status (WUS) register. * Sets the packet length in the Wake Up Packet Length (WUPL) register. The Controller maintains the first Magic Packet received in the Wake Up Packet Memory (WUPM) register until the software device driver writes a 1b to the Magic Packet Received MAG bit in the Wake Up Status (WUS) register. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1103 APM wake up is supported in all power states and only disabled if a subsequent EEPROM read results in the APM Wake Up bit being cleared or software explicitly writes a 0b to the APM Wake Up (APME) bit of the WUC register. Note: When WUC.APMPME is set PE_WAKE_N is asserted and a PM_PME message is issued even if PMCSR.PME_En is cleared. To enable disabling of system Wake-up when PMCSR.PME_En is cleared, Software driver should clear the WUC.APMPME bit after power-up or PCIe reset. 25.4.2 PCIe Power Management Wake Up The Controller supports PCIe power management based wake ups. It can generate system wake-up events from three sources: * Reception of a Magic Packet. * Reception of a network wakeup packet. * Detection of a link change of state. Activating PCIe power management wake up requires the following: * The software device driver programs the Wake Up Filter Control (WUFC) register to indicate the packets it needs to wake up and supplies the necessary data to the IPv4/v6 Address Table (IP4AT, IP6AT) and the Flexible Host Filter Table (FHFT). It can also set the Link Status Change Wake Up Enable (LNKC) bit in the Wake Up Filter Control (WUFC) register to cause wake up when the link changes state. * The operating system (at configuration time) writes a 1b to the PME_En bit of the Power Management Control/Status (PMCSR.8) register. Normally, after enabling wake up, the operating system writes 11b to the lower two bits of the PMCSR register to place the Controller into low-power mode. Once wake up is enabled, the Controller monitors incoming packets, first filtering them according to its standard address filtering method, then filtering them with all of the enabled wakeup filters. If a packet passes both the standard address filtering and at least one of the enabled wakeup filters, the Controller: * Sets the PME_Status bit in the PMCSR. * Asserts PE_WAKE_N (if the PME_En bit in the PMCSR is set). * Stores the first 128 bytes of the packet in the Wakeup Packet Memory (WUPM) register. * Sets one or more of the received bits in the Wake Up Status (WUS) register. The Controller sets more than one bit if a packet matches more than one filter. * Sets the packet length in the Wake Up Packet Length (WUPL) register. If enabled, a link state change wake up causes similar results, setting PME_Status, asserting PE_WAKE_N and setting the Link Status Changed (LNKC) bit in the Wake Up Status (WUS) register when the link goes up or down. The Controller supports the following change described in the PCIe Base Specification, Rev. 1.1RD (section 5.3.3.4) - On receiving a PME_Turn_Off message, the Controller must block the transmission of PM_PME messages and transmit a PME_TO_Ack message upstream. The Controller is permitted to send a PM_PME message after the Link is returned to an L0 state through LDn. PE_WAKE_N remains asserted until the operating system either writes a 1b to the PME_Status bit of the PMCSR register or writes a 0b to the PME_En bit. Intel(R) Communications Chipset 89xx Series - Datasheet 1104 October 2012 Order Number: 327879-001US 25.0 After receiving a wake-up packet, the Controller ignores any subsequent wake-up packets until the software device driver clears all of the received bits in the Wake Up Status (WUS) register. It also ignores link change events until the software device driver clears the Link Status Changed (LNKC) bit in the Wake Up Status (WUS) register. Note: A wake on link change is not supported when configured to SerDes or 1000BASE-KX mode. 25.4.3 Wake-Up Packets The Controller supports various wake-up packets using two types of filters: * Pre-defined filters * Flexible filters Each of these filters are enabled if the corresponding bit in the Wake Up Filter Control (WUFC) register is set to 1b. 25.4.3.1 Pre-Defined Filters The following packets are supported by the Controller's pre-defined filters: * Directed packet (including exact, multicast indexed, and broadcast) * Magic Packet * ARP/IPv4 request packet * Directed IPv4 packet * Directed IPv6 packet Each of these filters are enabled if the corresponding bit in the Wakeup Filter Control (WUFC) register is set to 1b. The explanation of each filter includes a table showing which bytes at which offsets are compared to determine if the packet passes the filter. Note: Both VLAN frames and LLC/SNAP can increase the given offsets if they are present. 25.4.3.1.1 Directed Exact Packet The Controller generates a wake-up event after receiving any packet whose destination address matches one of the 24 valid programmed receive addresses, if the Directed Exact Wake Up Enable bit is set in the Wake Up Filter Control (WUFC.EX) register. 25.4.3.1.2 Directed Multicast Packet For multicast packets, the upper bits of the incoming packet's destination address index a bit vector, the Multicast Table Array (MTA) that indicates whether to accept the packet. If the Directed Multicast Wake Up Enable bit set in the Wake Up Filter Control (WUFC.MC) register and the indexed bit in the vector is one, then the Controller generates a wake-up event. The exact bits used in the comparison are programmed by software in the Multicast Offset field of the Receive Control (RCTL.MO) register. 25.4.3.1.3 Broadcast If the Broadcast Wake Up Enable bit in the Wake Up Filter Control (WUFC.BC) register is set, the Controller generates a wake-up event when it receives a broadcast packet. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1105 25.4.3.1.4 Magic Packet Magic packets are defined in: http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/20213 .pdf as: "Once the LAN controller has been put into the Magic Packet mode, it scans all incoming frames addressed to the node for a specific data sequence.This sequence indicates to the controller that this is a Magic Packet frame. A Magic Packet frame must also meet the basic requirements for the LAN technology chosen, such as SOURCE ADDRESS, DESTINATION ADDRESS (which may be the receiving station's IEEE address or a MULTICAST address which includes the BROADCAST address), and CRC. The specific data sequence consists of 16 repetitions of the IEEE address of this node, with no breaks or interruptions. This sequence can be located anywhere within the packet, but must be preceded by a synchronization stream. The synchronization stream allows the scanning state machine to be much simpler. The synchronization stream is defined as 6 bytes of 0xFF. The device will also accept a BROADCAST frame, as long as the 16 repetitions of the IEEE address match the address of the machine to be awakened." The Controller expects the destination address to either: * Be the broadcast address (FF.FF.FF.FF.FF.FF) * Match the value in Receive Address 0 (RAH0, RAL0) register. This is initially loaded from the EEPROM but can be changed by the software device driver. * Match any other address filtering (RAH[n], RAL[n]) enabled by the software device driver. The Controller searches for the contents of Receive Address 0 (RAH0, RAL0) register as the embedded IEEE address. It considers any non-0xFF byte after a series of at least 6 0xFFs to be the start of the IEEE address for comparison purposes. For example, it catches the case of 7 0xFFs followed by the IEEE address). As soon as one of the first 96 bytes after a string of 0xFFs don't match, it continues to search for another set of at least 6 0xFFs followed by the 16 copies of the IEEE address later in the packet. This definition precludes the first byte of the destination address from being FF. A Magic Packet's destination address must match the address filtering enabled in the configuration registers with the exception that broadcast packets are considered to match even if the Broadcast Accept bit of the Receive Control (RCTL.BAM) register is 0b. If APM wake up (wake up by a Magic Packet) is enabled in the EEPROM, the Controller starts up with the Receive Address 0 (RAH0, RAL0) register loaded from the EEPROM. This enables the Controller to accept packets with the matching IEEE address before the software device driver loads. Table 25-5. Magic Packet Structure Offset # of bytes Field Value Action 0 6 Destination Address Compare 6 6 Source Address Skip 12 8 Possible LLC/SNAP Header Skip 12 4 Possible VLAN Tag Skip 12 4 Type Skip Any 6 Synchronizing Stream FF*6+ Compare any+6 96 16 copies of Node Address A*16 Compare Intel(R) Communications Chipset 89xx Series - Datasheet 1106 Comment MAC header - processed by main address filter. Compared to Receive Address 0 (RAH0, RAL0) register. October 2012 Order Number: 327879-001US 25.0 25.4.3.1.5 ARP/IPv4 Request Packet The Controller supports receiving ARP request packets for wake up if the ARP bit is set in the Wake Up Filter Control (WUFC) register. Four IPv4 addresses are supported, which are programmed in the IPv4 Address Table (IP4AT). A successfully matched packet must contain a broadcast MAC address, a protocol type of 0x0806, an ARP op-code of 0x01, and one of the four programmed IPv4 addresses. The Controller also handles ARP request packets that have VLAN tagging on both Ethernet II and Ethernet SNAP types. Table 25-6. ARP Packet Structure and Processing Offset # of bytes Field Value Action 0 6 Destination Address Compare 6 6 Source Address Skip 12 8 Possible LLC/SNAP Header Skip 12 4 Possible VLAN Tag 12 2 Type 14 2 HW Type 0x0001 Compare 16 2 Protocol Type 0x0800 Compare 18 1 Hardware Size 0x06 Compare 19 1 Protocol Address Length 0x04 Compare 20 2 Operation 0x0001 Compare 22 6 Sender HW Address - Ignore 28 4 Sender IP Address - Ignore 32 6 Target HW Address - Ignore 38 4 Target IP Address IP4AT Compare 25.4.3.1.6 Comment MAC header - processed by main address filter. Skip 0x0806 Compare ARP May match any of four values in IP4AT. Directed Ipv4 Packet The Controller supports receiving directed IPv4 packets for wake up if the IPV4 bit is set in the Wake Up Filter Control (WUFC) register. Four IPv4 addresses are supported, which are programmed in the IPv4 Address Table (IP4AT). A successfully matched packet must contain the station's MAC address, a protocol type of 0x0800, and one of the four programmed IPv4 addresses. The Controller also handles directed IPv4 packets that have VLAN tagging on both Ethernet II and Ethernet SNAP types. Table 25-7. IPv4 Packet Structure and Processing Offset # of bytes Field Value Action 0 6 Destination Address Compare 6 6 Source Address Skip Comment MAC header - processed by main address filter. 12 8 Possible LLC/SNAP Header Skip 12 4 Possible VLAN Tag Skip 12 2 Type 0x0800 Compare IP 14 1 Version/ HDR length 0x4X Compare Check IPv4 15 1 Type of Service - Ignore 16 2 Packet Length - Ignore October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1107 Table 25-7. IPv4 Packet Structure and Processing (Continued) 18 2 Identification - Ignore 20 2 Fragment Info - Ignore 22 1 Time to live - Ignore 23 1 Protocol - Ignore 24 2 Header Checksum - Ignore 26 4 Source IP Address - Ignore 30 4 Destination IP Address IP4AT Compare 25.4.3.1.7 May match any of four values in IP4AT. Directed IPv6 Packet The Controller supports receiving directed IPv6 packets for wake up if the IPV6 bit is set in the Wake Up Filter Control (WUFC) register. One IPv6 address is supported and is programmed in the IPv6 Address Table (IP6AT). A successfully matched packet must contain the station's MAC address, a protocol type of 0x86DD, and the programmed IPv6 address. In addition, the IPAV.V60 bit should be set. The Controller also handles directed IPv6 packets that have VLAN tagging on both Ethernet II and Ethernet SNAP types. Table 25-8. IPv6 Packet Structure and Processing Offset # of bytes Field Value Action 0 6 Destination Address Compare 6 6 Source Address Skip Comment MAC header - processed by main address filter. 12 8 Possible LLC/SNAP Header Skip 12 4 Possible VLAN Tag Skip 12 2 Type 0x86DD Compare IP 14 1 Version/ Priority 0x6X Compare Check IPv6 15 3 Flow Label - Ignore 18 2 Payload Length - Ignore 20 1 Next Header - Ignore 21 1 Hop Limit - Ignore 22 16 Source IP Address - Ignore 38 16 Destination IP Address IP6AT Compare 25.4.3.2 Match value in IP6AT. Flexible Filters The Controller supports a total of 8 flexible filters. Each filter can be configured to recognize any arbitrary pattern within the first 128 bytes of the packet. To configure the flexible filters, software programs the mask values (required values and the minimum packet length), into the Flexible Host Filter Table (FHFT and FHFT_EXT). These 8 flexible filters contain separate values for each filter. Software must also enable the filters in the Wake Up Filter Control (WUFC) register, and enable the overall wake up functionality. The overall wake up functionality must be enabled by setting PME_En in the PMCSR or the Wake Up Control (WUC) register. Once enabled, the flexible filters scan incoming packets for a match. If the filter encounters any byte in the packet where the mask bit is one and the byte doesn't match the value programmed in the Flexible Host Filter Table (FHFT or FHFT_EXT), then Intel(R) Communications Chipset 89xx Series - Datasheet 1108 October 2012 Order Number: 327879-001US 25.0 the filter fails that packet. If the filter reaches the required length without failing the packet, it passes the packet and generates a wake-up event. It ignores any mask bits set to one beyond the required length. Note: The flex filters are temporarily disabled when read from or written to by the host. Any packet received during a read or write operation is dropped. Filter operation resumes once the read or write access completes. The following packets are listed for reference purposes only. The flexible filter could be used to filter these packets. 25.4.3.2.1 IPX Diagnostic Responder Request Packet An IPX diagnostic responder request packet must contain a valid MAC address, a protocol type of 0x8137, and an IPX diagnostic socket of 0x0456. It might include LLC/SNAP headers and VLAN tags. Since filtering this packet relies on the flexible filters, which use offsets specified by the operating system directly, the operating system must account for the extra offset LLC/SNAP headers and VLAN tags. Table 25-9. IPX Diagnostic Responder Request Packet Structure and Processing Offset # of bytes Field Value Action 0 6 Destination Address Compare 6 6 Source Address Skip 12 8 Possible LLC/SNAP Header Skip 12 4 Possible VLAN Tag Skip 12 2 Type 0x8137 Compare 14 16 Some IPX Stuff - Ignore 30 2 IPX Diagnostic Socket 0x0456 Compare 25.4.3.2.2 Comment IPX Directed IPX Packet A valid directed IPX packet contains the station's MAC address, a protocol type of 0x8137, and an IPX node address that is equal to the station's MAC address. It might include LLC/SNAP headers and VLAN tags. Since filtering this packet relies on the flexible filters, which use offsets specified by the operating system directly, the operating system must account for the extra offset LLC/SNAP headers and VLAN tags. Table 25-10. IPX Packet Structure and Processing Offset # of bytes Field Value Action 0 6 Destination Address Compare 6 6 Source Address Skip 12 8 Possible LLC/SNAP Header Skip 12 4 Possible VLAN Tag Skip 12 2 Type 0x8137 Compare 14 10 Some IPX Info - Ignore 24 6 IPX Node Address Receive Address 0 Compare October 2012 Order Number: 327879-001US Comment MAC header - processed by main address filter. IPX Must match receive address 0. Intel(R) Communications Chipset 89xx Series - Datasheet 1109 25.4.3.2.3 IPv6 Neighbor Discovery Filter In IPv6, a neighbor discovery packet is used for address resolution. A flexible filter can be used to check for a neighborhood discovery packet. 25.4.3.2.4 Utilizing Flex Wake-Up Filters In Normal Operation The Controller enables utilizing the WoL Flex filters in normal operation, when in D0 power management state, for queuing decisions. Further information is in Section 26.1.1.6. 25.4.3.3 Wake Up Packet Storage The Controller saves the first 128 bytes of the wake-up packet in its internal buffer, which can be read through the Wake Up Packet Memory (WUPM) register after the system wakes up. Intel(R) Communications Chipset 89xx Series - Datasheet 1110 October 2012 Order Number: 327879-001US 26.0 26.0 GbE Inline Functions 26.1 Receive Functionality 26.1.1 Receive Queues Assignment A received packet goes through three stages of filtering as shown in Figure 26-1. Figure 26-1 describes a switch-like structure that is used in virtualization mode to route packets between the network port (top of drawing) and one or more virtual ports (bottom of figure), where each virtual port can be associated with a virtual machine, a VMM or any other software entity. The first step in queue assignment is to verify that the packet is destined to the port. This is done by a set of L2 filters as described in Section 26.1.2. The second stage is specific to virtualization environments and defines the virtual ports (called pools) that are the targets for the Rx packet. A packet can be associated with any number of ports/pools using a selection process described in Section 26.1.1.2. Figure 26-1. Stages in Packet Filtering L2 Filters Queue/Pool Select In the third stage, a received packet that successfully passed the Rx filters is associated with one or more receive descriptor queues as described in this section. The following filter mechanisms determines the destination of a receive packet. These are described briefly in this section and in full details in separate sections: October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1111 * Virtualization -- In a virtualized environment, DMA resources are shared between more than one software entity (operating system and/or software device driver). This is done by allocating receive descriptor queues/pools to virtual partitions (VMM or VMs). Virtualization assigns to each received packet one or more pool indices. Packets are routed to a pool based on their pool index and other considerations. See Section 26.1.1.2 for details on routing for virtualization. * RSS -- Receive Side Scaling distributes packet processing between several processor cores by assigning packets into different descriptor queues. RSS assigns to each received packet an RSS index. Packets are routed to a queue out of a set of Rx queues based on their RSS index and other considerations. See Section 26.1.1.8 for details on RSS. * L2 Ethertype filters -- These filters identify packets by their L2 Ether type and assign them to receive queues. Examples of possible uses are LLDP packets and 802.1X packets. See Section 26.1.1.4 for mode details. The Controller incorporates 8 Ether-type filters per port. * 2-tuple filters -- These filters identify packets with specific TCP/UDP destination port and/or L4 protocol. Each filter consists of a 2-tuple (protocol and destination TCP/UDP port) and routes packets into one of the Rx queues. The Controller has 8 such filters per port. See Section 26.1.1.5 for details. * TCP SYN filters -- The Controller might route TCP packets with their SYN flag set into a separate queue. SYN packets are often used in SYN attacks to load the system with numerous requests for new connections. By filtering such packets to a separate queue, security software can monitor and act on SYN attacks. The Controller has one such filter per port. See Section 26.1.1.7 for more details. * Flex Filters - These filters can be either used as WoL filters when the Controller is in D3 state or for queueing in normal operating mode (D0 state). Filters enable queueing according to a match of any 128 Byte sequence at the beginning of a packet. Each one of the 128 bytes can be either compared or masked using a dedicated mask field. The Controller has 8 such filters per port. See Section 26.1.1.6 for details. Typically, packet reception consists of recognizing the presence of a packet on the wire, performing address filtering, storing the packet in the receive data FIFO, transferring the data to one of the 16 receive queues in host memory, and updating the state of a receive descriptor. Note: Maximum supported received-packet size is 9.5 KB (9728 bytes). A received packet is allocated to a queue based on the previous criteria and the following order: 1. Queue by L2 Ether-type filters (if a match) 2. If RFCTL.SYNQFP is 0b (2-tuple filter and Flex filter have priority), then: a. Queue by Flex filter (if a match) b. Queue by 2-tuple filter c. Queue by SYN filter (if a match) 3. If RFCTL.SYNQFP is 1b (SYN filter has priority), then: a. Queue by SYN filter (if a match) b. Queue by Flex filter (if a match) c. Queue by 2-tuple filter (if a match) 4. Define a pool (if virtualization enabled) 5. Queue by RSS (if RSS enabled). Intel(R) Communications Chipset 89xx Series - Datasheet 1112 October 2012 Order Number: 327879-001US 26.0 26.1.1.1 Queuing in a Non-Virtualized Environment When the MRQC.Multiple Receive Queues Enable field equals 010b (Multiple receive queues as defined by filters and RSS for 8 queues) the received packet is assigned to a queue in the following manner: * L2 Ether-type filters -- Each filter identifies one of 16 receive queues. * SYN filter -- Identifies one of 16 receive queues. * Flex Filter - Each filter identifies one of 16 receive queues. * 2-tuple filters -- Each filter identifies one of 16 receive queues. * RSS filters - Identifies one of 2 x 16 queues through the RSS index. The following modes are supported: -- No RSS -- The default queue as defined in MRQC.DEF_Q is used for packets that do not meet any of the previous conditions. -- RSS only -- A set of 16 queues is allocated for RSS. The queue is identified through the RSS index. It is possible to use a subset of the 16 queues. When the MRQC.Multiple Receive Queues Enable field equals 000b (Multiple receive queues as defined by filters) the received packet is assigned to a queue in the following manner: * L2 Ether-type filters -- Each filter identifies one of 16 receive queues. * SYN filter -- Identifies one of 16 receive queues. * Flex Filter - Each filter identifies one of 16 receive queues. * 2-tuple filters -- Each filter identifies one of 16 receive queues. 26.1.1.2 Receive Queuing in a Virtualized Environment In VMDq mode, system software allocates the pools to the VMM, an IOVM, or to VMs. When the MRQC.Multiple Receive Queues Enable field equals 011b (Multiple receive queues as defined by VMDq), the received packets are allocated to the 16 receive queues/pools in the following manner: Incoming packets are associated with pools/ queues based on their L2 characteristics as described in Section 26.8.3. This section describes the following stage, where an Rx queue is assigned to each replication of the Rx packet as determined by its pool's association. A received packet is assigned to a queue/pool in the following manner: * L2 Ether-type filters -- Each filter identifies a specific receive queue (the queue is usually allocated to the VMM or a service operating system). * SYN filter -- Not supported in VT modes. * 2-tuple filters - Not supported Flex filters --Not supported in VT mode. * RSS filters - Not supported in VT mode October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1113 Figure 26-2. Receive Queuing Flow (Virtualization) S ta rt P a c k e t m a tc h e s L 2 E th e rT y p e F ilte r ? Yes T h e L 2 E t h e r T y p e filt e r d e f in e s th e R x Q u e u e No C r e a te T a r g e t V M L is t A p p ly m ir o r r in g r u le s T a r g e t V M L is t E m p ty ? Yes No U s e V T _ C T L .D E F _ P L a n d T a r g e t V M L is t t o s e t q u e u e n u m b e r.s D is c a r d P a c k e t END 26.1.1.3 Queue Configuration Registers Configuration registers (CSRs) that control queue operation are replicated per queue (total of 16 copies of each register per port). Each of the replicated registers correspond to a queue such that the queue index equals the serial number of the register (such as register 0 corresponds to queue 0, etc.). Registers included in this category are: * RDBAL and RDBAH -- Rx Descriptor Base * RDLEN -- RX Descriptor Length * RDH -- RX Descriptor Head * RDT -- RX Descriptor Tail * RXDCTL -- Receive Descriptor Control * RXCTL -- Rx DCA Control * SRRCTL -- Split and Replication Receive Control * PSRTYPE -- Packet Split Receive type Intel(R) Communications Chipset 89xx Series - Datasheet 1114 October 2012 Order Number: 327879-001US 26.0 26.1.1.4 L2 Ether-Type Filters These filters identify packets by L2 Ether-type and assign them to a receive queue. The following usages have been identified: * IEEE 802.1X packets -- Extensible Authentication Protocol over LAN (EAPOL). * Time sync packets (such as IEEE 1588) -- Identifies Sync or Delay_Req packets * IEEE802.1AB LLDP (Link Layer Discovery Protocol) packets. The Controller incorporates 8 Ether-type filters. The Packet Type field in the Rx descriptor captures the filter number that matched the L2 Ether-type. See Section 26.1.5 for decoding of the Packet Type field. The Ether-type filters are configured via the ETQF register as follows: * The EType field contains the 16-bit Ether-type compared against all L2 type fields in the Rx packet. * The Filter Enable bit enables identification of Rx packets by Ether-type according to this filter. If this bit is cleared, the filter is ignored for all purposes. * The Rx Queue field contains the absolute destination queue for the packet. * The 1588 Time Stamp field indicates that the packet should be time stamped according to the IEEE 1588 specification. * The Queue Enable field enables forwarding Rx packets based on the Ether-type defined in this register. Note: Software should not assign the same Ether-type value to different ETQF filters with different Rx Queue assignments. Special considerations for Virtualization modes: * Packets that match an Ether-type filter are diverted from their original pool (the pool identified by the L2 filters) to the pool used as the pool to which the queue in the Queue field belongs. In other words, The L2 filters are ignored in determining the pool for such packets. * The same applies for multi-cast packets. A single copy is posted to the pool defined by the filter. * Mirroring rules: -- If a pool is being mirrored, the pool to which the queue belongs to is used to determine if a packet that matches the filter should be mirrored. -- The queue inside the pool is used for both the original pool and the mirroring pool. 26.1.1.5 2-Tuple Filters These filters identify specific packets destined to a certain TCP/UDP port and implement a specific protocol. Each filter consists of a 2-tuple (protocol and destination TCP/UDP port) and forwards packets into one of the receive queues. The Controller incorporates 8 such filters. The 2-tuple filters are configured via the TTQF (See Section 28.1.3), IMIR (See Section 28.10.1.1) and IMIR_EXT (See Section 28.10.1.2) registers as follows (per filter): * Protocol -- Identifies the IP protocol, part of the 2-tuple queue filters. Enabled by a bit in the TTQF.Mask field. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1115 * Destination port -- Identifies the TCP/UDP destination port, part of the 2-tuple queue filters. Enabled by the IMIR.PORT_BP bit. * Size threshold -- Identifies the length of the packet that should trigger the filter. This is the length as received by the host, not including any part of the packet removed by hardware. Enabled by the IMIREXT.Size_BP field. * Control Bits -- Identify TCP flags that might be part of the filtering process. Enabled by the IMIREXT.CtrlBit_BP field. * Rx queue -- Determines the Rx queue for packets that match this filter: -- The TTQF.Rx Queue field contains the queue serial number. * Queue enable -- Enables forwarding a packet that uses this filter to the queue defined in the TTQF.Rx Queue field. * Mask -- A 1-bit field that masks the L4 protocol check. The filter is a logical AND of the non-masked 2-tuple fields. If all 2-tuple fields are masked, the filter is not used for queue forwarding. Notes: 26.1.1.6 * If more than one 2-tuple filter with the same priority is matched by the packet, the first filter (lowest ordinal number) is used in order to define the queue destination of this packet. * The immediate interrupt and 1588 actions are defined by the OR of all the matching filters. Flex Filters The Controller supports a total of 8 flexible filters. Each filter can be configured to recognize any arbitrary pattern within the first 128 bytes of the packet. To configure the flexible filters, software programs the mask values (required values and the minimum packet length), into the Flexible Host Filter Table (FHFT and FHFT_EXT, See Section 28.20.2 and Section 28.20.3). These 8 flexible filters can be used as for wake-up when in D3 state or for queueing when in D0 state. Software must enable the filters in the Wake Up Filter Control (WUFC See Section 28.20.1.2) register for operation in D3 or D0 mode. In D0 mode these filters enable forwarding of packets that match up to 128 Bytes defined in the filter to one of the receive queues. In D3 mode these filters can be used for Wake-on-Lan as described in Section 25.4.3.2. Once enabled, the flexible filters scan incoming packets for a match. If the filter encounters any byte in the packet where the mask bit is one and the byte doesn't match the value programmed in the Flexible Host Filter Table (FHFT or FHFT_EXT), then the filter fails that packet. If the filter reaches the required length without failing the packet, it forwards the packet to the appropriate receive queue. It ignores any mask bits set to one beyond the required length (defined in the Length field in the FHFT or FHFT_EXT registers). Note: The flex filters are temporarily disabled when read from or written to by the host. Any packet received during a read or write operation is dropped. Filter operation resumes once the read or write access completes. The flex fitters are configured in D0 state via the WUFC, FHFT and FHFT_EXT registers as follows (per filter): * Byte Sequence to be compared -- Program 128 Byte sequence, mask bits and Length field in FHFT and FHFT_EXT registers. * Filter Priority -- Program filter priority in queueing field in FHFT and FHFT_EXT registers. Intel(R) Communications Chipset 89xx Series - Datasheet 1116 October 2012 Order Number: 327879-001US 26.0 * Receive queue -- Program receive queue to forward packet in queueing field in FHFT and FHFT_EXT registers. * Filter actions -- Program immediate interrupt requirement in queueing field in FHFT and FHFT_EXT registers. * Filter enable -- Set WUFC.FLEX_HQ bit to 1 to enable flex filter operation in D0 state. Set appropriate WUFC.FLX[n] bit to 1 to enable specific flex filter. Before entering D3 state software device driver programs the FHFT and FHFT_EXT filters for appropriate wake events and enables relevant filters by setting the WUFC.FLX[n] bit to 1. Following move to D0 state the software device driver programs the FHFT and FHFT_EXT filters for appropriate queueing decisions enables relevant filters by setting the WUFC.FLX[n] bit to 1 and the WUFC.FLEX_HQ bit to 1. Notes: If more than one flex filter with the same priority is matched by the packet, the first filter (lowest address) is used in order to define the queue destination of this packet. The immediate interrupt action is defined by the OR of all the matching filters. 26.1.1.7 SYN Packet Filters The Controller might forward TCP packets whose SYN flag is set into a separate queue. SYN packets are often used in SYN attacks to load the system with numerous requests for new connections. By filtering such packets to a separate queue, security software can monitor and act on SYN attacks. SYN filters are configured via the SYNQF registers as follows: * Queue En -- Enables forwarding of SYN packets to a specific queue. * Rx Queue field -- Contains the destination queue for the packet. This filter is not to be used in a virtualized environment. 26.1.1.8 Receive-Side Scaling (RSS) RSS is a mechanism to distribute received packets into several descriptor queues. Software then assigns each queue to a different processor, sharing the load of packet processing among several processors. The Controller uses RSS as one ingredient in its packet assignment policy (the others are the various filters and virtualization). The RSS output is a RSS index. The Controller's global assignment uses these bits (or only some of the LSB bits) as part of the queue number. RSS is enabled in the MRQC register. The RSS Status field in the descriptor write-back is enabled when the RXCSUM.PCSD bit is set (fragment checksum is disabled). RSS is therefore mutually exclusive with UDP fragmentation. Also, support for RSS is not provided when legacy receive descriptor format is used. When RSS is enabled, the Controller provides software with the following information as required by Microsoft* RSS specification or for device driver assistance: * A Dword result of the Microsoft* RSS hash function, to be used by the stack for flow classification, is written into the receive packet descriptor (required by Microsoft* RSS). * A 4-bit RSS Type field conveys the hash function used for the specific packet (required by Microsoft* RSS). Figure 26-3 shows the process of computing an RSS output: October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1117 1. The receive packet is parsed into the header fields used by the hash operation (such as IP addresses, TCP port, etc.). 2. A hash calculation is performed. The Controller supports a single hash function, as defined by Microsoft* RSS. The Controller does not indicate to the software device driver which hash function is used. The 32-bit result is fed into the packet receive descriptor. 3. The seven LSB bits of the hash result are used as an index into a 128-entry indirection table. Each entry provides a 3-bit RSS output index. When RSS is disabled, packets are assigned an RSS output index = zero. System software might enable or disable RSS at any time. While disabled, system software might update the contents of any of the RSS-related registers. When multiple requests queues are enabled in RSS mode, un-decodable packets are assigned an RSS output index = zero. The 32-bit tag (normally a result of the hash function) equals zero. Figure 26-3. RSS Block Diagram Parsed Receive Packet RSS Hash Indirection Table 128 x 3 7 LS 32 Packet Descriptor 0 3 RSS Disable or (RSS And Not Decodable) 3 RSS Output Index Intel(R) Communications Chipset 89xx Series - Datasheet 1118 October 2012 Order Number: 327879-001US 26.0 26.1.1.8.1 RSS Hash Function Section 26.1.1.8.1 provides a verification suite used to validate that the hash function is computed according to Microsoft nomenclature. The Controller hash function follows Microsoft definition. A single hash function is defined with several variations for the following cases: * TcpIPv4 -- The Controller parses the packet to identify an IPv4 packet containing a TCP segment per the criteria described later in this section. If the packet is not an IPv4 packet containing a TCP segment, RSS is not done for the packet. * IPv4 -- The Controller parses the packet to identify an IPv4 packet. If the packet is not an IPv4 packet, RSS is not done for the packet. * TcpIPv6 -- The Controller parses the packet to identify an IPv6 packet containing a TCP segment per the criteria described later in this section. If the packet is not an IPv6 packet containing a TCP segment, RSS is not done for the packet. * TcpIPv6Ex -- The Controller parses the packet to identify an IPv6 packet containing a TCP segment with extensions per the criteria described later in this section. If the packet is not an IPv6 packet containing a TCP segment, RSS is not done for the packet. Extension headers should be parsed for a Home-Address-Option field (for source address) or the Routing-Header-Type-2 field (for destination address). * IPv6Ex -- The Controller parses the packet to identify an IPv6 packet. Extension headers should be parsed for a Home-Address-Option field (for source address) or the Routing-Header-Type-2 field (for destination address). The packet is not required to contain any of these extension headers to be hashed by this function. In this case, the IPv6 hash is used. If the packet is not an IPv6 packet, RSS is not done for the packet. * IPv6 -- The Controller parses the packet to identify an IPv6 packet. If the packet is not an IPv6 packet, receive-side-scaling is not done for the packet. The following additional cases are not part of the Microsoft RSS specification: * UdpIPV4 -- The Controller parses the packet to identify a packet with UDP over IPv4. * UdpIPV6 -- The Controller parses the packet to identify a packet with UDP over IPv6. * UdpIPV6Ex -- The Controller parses the packet to identify a packet with UDP over IPv6 with extensions. A packet is identified as containing a TCP segment if all of the following conditions are met: * The transport layer protocol is TCP (not UDP, ICMP, IGMP, etc.). * The TCP segment can be parsed (such as IP options can be parsed, packet not encrypted). * The packet is not fragmented (even if the fragment contains a complete TCP header). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1119 Bits[31:16] of the Multiple Receive Queues Command (MRQC) register enable each of the above hash function variations (several can be set at a given time). If several functions are enabled at the same time, priority is defined as follows (skip functions that are not enabled): IPv4 packet: 1. Try using the TcpIPv4 function. 2. Try using IPV4_UDP function. 3. Try using the IPv4 function. IPv6 packet: 1. If TcpIPv6Ex is enabled, try using the TcpIPv6Ex function; else if TcpIPv6 is enabled try using the TcpIPv6 function. 2. If UdpIPv6Ex is enabled, try using UdpIPv6Ex function; else if UpdIPv6 is enabled try using UdpIPv6 function. 3. If IPv6Ex is enabled, try using the IPv6Ex function, else if IPv6 is enabled, try using the IPv6 function. The following combinations are currently supported: * Any combination of IPv4, TcpIPv4, and UdpIPv4. * And/or. * Any combination of either IPv6, TcpIPv6, and UdpIPv6 or IPv6Ex, TcpIPv6Ex, and UdpIPv6Ex. When a packet cannot be parsed by the previously mentioned rules, it is assigned an RSS output index = zero. The 32-bit tag (normally a result of the hash function) equals zero. The 32-bit result of the hash computation is written into the packet descriptor and also provides an index into the indirection table. The following notation is used to describe the hash functions: * Ordering is little endian in both bytes and bits. For example, the IP address 161.142.100.80 translates into 0xa18e6450 in the signature. * A "^ "denotes bit-wise XOR operation of same-width vectors. * @x-y denotes bytes x through y (including both of them) of the incoming packet, where byte 0 is the first byte of the IP header. In other words, it is considered that all byte-offsets as offsets into a packet where the framing layer header has been stripped out. Therefore, the source IPv4 address is referred to as @12-15, while the destination v4 address is referred to as @16-19. * @x-y, @v-w denotes concatenation of bytes x-y, followed by bytes v-w, preserving the order in which they occurred in the packet. All hash function variations (IPv4 and IPv6) follow the same general structure. Specific details for each variation are described in the following section. The hash uses a random secret key length of 320 bits (40 bytes); the key is typically supplied through the RSS Random Key Register (RSSRK). The algorithm works by examining each bit of the hash input from left to right. Intel's nomenclature defines left and right for a byte-array as follows: Given an array K with k bytes, Intel's nomenclature assumes that the array is laid out as shown: K[0] K[1] K[2] ... K[k-1] Intel(R) Communications Chipset 89xx Series - Datasheet 1120 October 2012 Order Number: 327879-001US 26.0 K[0] is the left-most byte, and the MSB of K[0] is the left-most bit. K[k-1] is the right-most byte, and the LSB of K[k-1] is the right-most bit. ComputeHash(input[], N) For hash-input input[] of length N bytes (8N bits) and a random secret key K of 320 bits Result = 0; For each bit b in input[] { if (b == 1) then Result ^= (left-most 32 bits of K); shift K left 1 bit position; } return Result; The following four pseudo-code examples are intended to help clarify exactly how the hash is to be performed in four cases, IPv4 with and without ability to parse the TCP header and IPv6 with an without a TCP header. 26.1.1.8.1.4 Hash for IPv4 with TCP Concatenate SourceAddress, DestinationAddress, SourcePort, DestinationPort into one single byte-array, preserving the order in which they occurred in the packet: Input[12] = @12-15, @16-19, @20-21, @22-23. Result = ComputeHash(Input, 12); 26.1.1.8.1.5 Hash for IPv4 with UDP Concatenate SourceAddress, DestinationAddress, SourcePort, DestinationPort into one single byte-array, preserving the order in which they occurred in the packet: Input[12] = @12-15, @16-19, @20-21, @22-23. Result = ComputeHash(Input, 12); 26.1.1.8.1.6 Hash for IPv4 without TCP Concatenate SourceAddress and DestinationAddress into one single byte-array Input[8] = @12-15, @16-19 Result = ComputeHash(Input, 8) 26.1.1.8.1.7 Hash for IPv6 with TCP Similar to above: Input[36] = @8-23, @24-39, @40-41, @42-43 Result = ComputeHash(Input, 36) 26.1.1.8.1.8 Hash for IPv6 with UDP Similar to above: Input[36] = @8-23, @24-39, @40-41, @42-43 Result = ComputeHash(Input, 36) October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1121 26.1.1.8.1.9 Hash for IPv6 without TCP Input[32] = @8-23, @24-39 Result = ComputeHash(Input, 32) 26.1.1.8.2 Indirection Table The RETA indirection table is a 128-entry structure, indexed by the seven LSB bits of the hash function output. Each entry of the table contains the following: * Bits [2:0] - RSS index Note: In RSS only mode, all 3 bits are used. In VMDq mode RSS is not supported. System software might update the indirection table during run time. Such updates of the table are not synchronized with the arrival time of received packets. Therefore, it is not guaranteed that a table update takes effect on a specific packet boundary. 26.1.1.8.3 RSS Verification Suite Assume that the random key byte-stream is: 0x6d, 0x5a, 0x56, 0xda, 0x25, 0x5b, 0x0e, 0xc2, 0x41, 0xd0, 0x77, 0x6a, 0x67, 0xca, 0xcb, 0x42, 0x25, 0x2b, 0x2d, 0xb7, 0x3d, 0xcb, 0xa3, 0x3b, 0x43, 0xae, 0x80, 0xbe, 0xa3, 0x7b, 0x30, 0xac, 0x8f, 0x30, 0xf2, 0x01, 0xb0, 0xb4, 0x0c, 0xfa 26.1.1.8.3.10 IPv4 Table 26-1. IPv4 Destination Address/Port Source Address/Port IPv4 Only IPv4 With TCP 161.142.100.80:1766 66.9.149.187:2794 0x323e8fc2 0x51ccc178 65.69.140.83:4739 199.92.111.2:14230 0xd718262a 0xc626b0ea 12.22.207.184:38024 24.19.198.95:12898 0xd2d0a5de 0x5c2b394a 209.142.163.6:2217 38.27.205.30:48228 0x82989176 0xafc7327f 202.188.127.2:1303 153.39.163.191:44251 0x5d1809c5 0x10e828a2 26.1.1.8.3.11 IPv6 The IPv6 address tuples are only for verification purposes and might not make sense as a tuple. Table 26-2. IPv6 Destination Address/Port Source Address/Port IPv6 Only IPv6 With TCP 3ffe:2501:200:3::1 (1766) 3ffe:2501:200:1fff::7 (2794) 0x2cc18cd5 0x40207d3d ff02::1 (4739) 3ffe:501:8::260:97ff:fe40:efab (14230) 0x0f0c461c 0xdde51bbf fe80::200:f8ff:fe21:67cf (38024) 3ffe:1900:4545:3:200:f8ff:fe21:67cf (44251) 0x4b61e985 0x02d1feef Intel(R) Communications Chipset 89xx Series - Datasheet 1122 October 2012 Order Number: 327879-001US 26.0 26.1.1.8.4 Association Through MAC Address Each of the 24 MAC address filters can be associated with a VM. The POOLSEL field in the Receive Address High (RAH) register determines the target VM. Packets that do not match any of the MAC filters (such as promiscuous) are assigned with the default VM as defined in the VT_CTL.DEF_PL field. Software can program different values to the MAC filters (any bits in RAH or RAL) at any time. The Controller would respond to the change on a packet boundary but does not guarantee the change to take place at some precise time. 26.1.2 L2 Packet Filtering The receive packet filtering role is to determine which of the incoming packets are allowed to pass to the local system and which of the incoming packets should be dropped since they are not targeted to the local system. Received packets can be destined to the host, to a manageability controller (BMC), or to both. This section describes how host filtering is done, and the interaction with management filtering. As shown in Figure 26-12, host filtering has three stages: 1. Packets are filtered by L2 filters (MAC address, unicast/multicast/broadcast). See Section 26.1.2.1 for details. 2. Packets are then filtered by VLAN if a VLAN tag is present. See Section 26.1.2.2 for details. 3. Packets are filtered by the manageability filters (port, IP, flex, other). See Section 27.4 for details. A packet is not forwarded to the host if any of the following takes place: 1. The packet does not pass MAC address filters as described later in this section. * The packet does not pass VLAN filtering as described later in this section. * The packet passes manageability filtering and then the manageability filters determine if the packet should be sent only to the BMC (see Section 27.4 and the MNGONLY register). A packet that passes receive filtering as previously described might still be dropped due to other reasons. Normally, only good packets are received. These are defined as those packets with no Under Size Error, Over Size Error, Packet Error, Length Error and CRC Error are detected. However, if the store-bad-packet bit is set (FCTRL.SBP), then bad packets that don't pass the filter function are stored in host memory. Packet errors are indicated by error bits in the receive descriptor (RDESC.ERRORS). It is possible to receive all packets, regardless of whether they are bad, by setting the promiscuous enabled (Unicast and Multicast) and the store-bad-packet bits in the RCTL register. If there is insufficient space in the receive FIFO, hardware drops the packet and indicates the missed packet in the appropriate statistics registers. When the packet is routed to a queue with the SRRCTL.Drop_En bit set to 1, receive packets are dropped when insufficient receive descriptors exist to write the packet into system memory. Note: CRC errors before the SFD are ignored. Any packet must have a valid SFD in order to be recognized by the Controller (even bad packets). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1123 Figure 26-12.Receive Filtering Flow Chart MNG Filter FAIL MAC Address Filter Packet Arrived FAIL VLAN Filter PASS MNG only? Pass No Discard packet 26.1.2.1 To Host To MNG MAC Address Filtering Figure 26-13 shows the MAC address filtering. A packet passes successfully through the MAC address filtering if any of the following conditions are met: 1. It is a unicast packet and promiscuous unicast filtering is enabled. 2. It is a multicast packet and promiscuous multicast filtering is enabled. 3. It is a unicast packet and it matches one of the unicast MAC filters. 4. It is a multicast packet and it matches one of the multicast filters. 5. It is a broadcast packet and Broadcast Accept Mode (RCTL.BAM) is enabled. Intel(R) Communications Chipset 89xx Series - Datasheet 1124 October 2012 Order Number: 327879-001US 26.0 Figure 26-13.Host MAC Address Receive Filtering Flow Chart Start Yes Promiscuous Unicast enable Unicast Packet type No Yes Unicast pass Broadcast Yes Broadcast accept mode Multicast no No Promiscuous multicast enable Yes no VLAN filtering yes Multicast pass No Discard packet October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1125 26.1.2.1.1 Unicast Filter The entire MAC address is checked against the 24 host unicast addresses. The 24 host unicast addresses are controlled by the host interface (the BMC must not change them). The other 2 addresses are dedicated to management functions and are only accessed by the BMC. The destination address of incoming packet must exactly match one of the pre-configured host address filters. These addresses can be unicast or multicast. Those filters are configured through RAL, and RAH registers. Promiscuous Unicast -- Receive all unicasts. Promiscuous unicast mode in the RCTL register can be set/cleared only through the host interface (not by the BMC). This mode is usually used when the Controller is used as a sniffer. Unicast Hash Table -- Destination address matching the Unicast Hash Table (UTA). 26.1.2.1.2 Multicast Filter (Partial) The 12-bit portion of incoming packet multicast address must exactly match Multicast Filter Address (MFA) in order to pass multicast filtering. Those 12 bits out of 48 bits of the destination address can be selected by the MO field of RCTL (Section 28.9.1.1). These entries can be configured only by the host interface and cannot be controlled by the BMC. Packets received according to this mode have the PIF bit in the descriptor set to indicate imperfect filtering that should be validated by the software device driver. Promiscuous Multicast -- Receive all multicast. Promiscuous multicast mode can be set/cleared in the RCTL register only through the host interface (not by the BMC) and it is usually used when the Controller is used as a sniffer. Note: When the promiscuous bit is set and a multicast packet is received, the PIF bit of the packet status is not set. 26.1.2.2 VLAN Filtering A receive packet that successfully passed MAC address filtering is then subjected to VLAN header filtering. 1. If the packet does not have a VLAN header, it passes to the next filtering stage. Note: If extended VLAN is enabled (CTRL_EXT.EXTENDED_VLAN is set), it is assumed that the first VLAN tag is an extended VLAN and it is skipped. All next stages refer to the second VLAN. 2. If VLAN filtering is disabled (RCTL.VFE bit is cleared), the packet is forwarded to the next filtering stage. 3. If the packet has a VLAN header, and it matches an enabled host VLAN filter (relevant bit in VFTA table is set), the packet is forwarded to the next filtering stage. 4. Otherwise, the packet is dropped. Figure 26-14 shows the VLAN filtering flow. Intel(R) Communications Chipset 89xx Series - Datasheet 1126 October 2012 Order Number: 327879-001US 26.0 Figure 26-14.VLAN Filtering MAC address filtering No Packet has VLAN Header Yes No Host VLAN Filters enable Yes Yes PASS L2 No Discard packet MNG Only Check 26.1.2.3 host VLAN Filters pass Platform Manageability Filtering Platform manageability filtering is described in Section 27.4. Figure 26-15 shows the manageability portion of the packet filtering and it is brought here to make the receive packet filtering functionality description complete. Note: The manageability engine (i.e. BMC) might decide to block part of the received packets from also being sent to the Host, according to the external BMC instructions and the EEPROM settings. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1127 Figure 26-15.Platform Manageability Filtering Packet received from LAN Yes RCV EN No RCV All Pass decision filters No Yes No Yes Fixed Net type No Yes Yes EN XSUM Yes Match Net type No XSUM error Yes No No MNGONLY Send packet to MNG Not considered For MNG Yes Block packet from Host 26.1.3 Receive Data Storage 26.1.3.1 Host Buffers Each descriptor points to one or more memory buffers that are designated by the software device driver to store packet data. The size of the buffer can be set using the per queue SRRCTL[n].BSIZEPACKET field. In addition, for advanced descriptor usage the SRRCTL.BSIZEHEADER field is used to define the size of the buffers allocated to headers. Intel(R) Communications Chipset 89xx Series - Datasheet 1128 October 2012 Order Number: 327879-001US 26.0 The Controller places no alignment restrictions on receive memory buffer addresses. This is desirable in situations where the receive buffer was allocated by higher layers in the networking software stack, as these higher layers might have no knowledge of a specific device's buffer alignment requirements. Note: When the No-Snoop Enable bit is used in advanced descriptors, the buffer address is 16-bit (2-byte) aligned. 26.1.3.2 On-Chip Receive Buffers The Controller allocates by default a 32 KB on-chip packet buffer per port. The buffer can be used to store packets until they are forwarded to the host. The Controller utilizes a single common ram structure for the on-chip receive buffers allocated to the various ports. If a port is disabled, so that it can't be accessed by host and management, by: * Setting EEPROM bit LAN_DIS or the LAN_PCI_DIS in the "Software Defined Pins Control" word for the relevant port to 1. The freed buffer space can be allocated to the active ports via the "Initialization Control 4" EEPROM word. Actual on-chip receive buffer allocated to the port can be read in the IRPBS register. 26.1.3.3 On-Chip Descriptor Buffers The Controller contains a 16 descriptor cache for each receive queue used to reduce the latency of packet processing and to optimize the usage of PCIe bandwidth by fetching and writing back descriptors in bursts. The fetch and writeback algorithm are described in Section 26.1.6 and Section 26.1.7. 26.1.4 Legacy Receive Descriptor Format A receive descriptor is a data structure that contains the receive data buffer address and fields for hardware to store packet information. If SRRCTL[n].DESCTYPE = 000b, the Controller uses the legacy Receive descriptor as shown in Table 26-3. Note: Legacy descriptors should not be used when advanced features such as Virtualization are activated. Table 26-3. Legacy Receive Descriptor (RDESC) Layout 63 48 47 0 8 40 39 0 32 31 16 15 Buffer Address [63:0] VLAN Tag Errors Status Fragment Checksum Length After receiving a packet for the Controller, hardware stores the packet data into the indicated buffer and writes the length, packet checksum, status, errors, and status fields. Packet Buffer Address (64) - Physical address of the packet buffer. Length Field (16) Length covers the data written to a receive buffer including CRC bytes (if any). Software must read multiple descriptors to determine the complete length for a packet that spans multiple receive buffers. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1129 Fragment Checksum (16) This field is used to provide the fragment checksum value. This field equals to the unadjusted 16-bit ones complement of the packet. Checksum calculation starts at the L4 layer (after the IP header) until the end of the packet excluding the CRC bytes. In order to use the fragment checksum assist to offload L4 checksum verification, software might need to back out some of the bytes in the packet. For more details see Section 26.1.11.2 Status Field (8) Status information indicates whether the descriptor has been used and whether the referenced buffer is the last one for the packet. See Table 26-4 for the layout of the Status field. Error status information is shown in Figure 26-8. Table 26-4. Receive Status (RDESC.STATUS) Layout 7 6 5 4 3 2 1 0 PIF IPCS L4CS UDPCS VP Rsv EOP DD * PIF (bit 7) - Passed in-exact filter * IPCS (bit 6) - Ipv4 checksum calculated on packet * L4CS (bit 5) - L4 (UDP or TCP) checksum calculated on packet * UDPCS (bit 4) - UDP checksum calculated on packet * VP (bit 3) - Packet is 802.1q (matched VET); indicates strip VLAN in 802.1q packet * RSV (bit 2) - Reserved * EOP (bit 1) - End of packet * DD (bit 0) - Descriptor done EOP and DD The following table lists the meaning of these bits: Table 26-5. Receive Status Bits DD EOP Description 0b 0b Software setting of the descriptor when it hands it off to the hardware. 0b 1b Reserved (invalid option). 1b 0b A completion status indication for a non-last descriptor of a packet that spans across multiple descriptors. In a single packet case, DD indicates that the hardware is done with the descriptor and its buffers. Only the Length fields are valid on this descriptor. 1b 1b A completion status indication of the entire packet. Software might take ownership of its descriptors. All fields in the descriptor are valid (reported by the hardware). VP Field The VP field indicates whether the incoming packet's type matches the VLAN Ethernet Type programmed in the VET Register. For example, if the packet is a VLAN (802.1q) type, it is set if the packet type matches VET and CTRL.VME is set (Vlan mode enabled). It also indicates that VLAN has been stripped from the 802.1q packet. For more details, see Section 26.4. IPCS (Ipv4 Checksum), L4CS (L4 Checksum), and UDPCS (UDP Checksum) The meaning of these bits is shown in the table below: Intel(R) Communications Chipset 89xx Series - Datasheet 1130 October 2012 Order Number: 327879-001US 26.0 Table 26-6. IPCS, L4CS, and UDPCS L4CS UDPCS IPCS Functionality 0b 0b 0b Hardware does not provide checksum offload. Special case: Hardware does not provide UDP checksum offload for IPV4 packet with UDP checksum = 0b 1b 0b 1b / 0b Hardware provides IPv4 checksum offload if IPCS is active and TCP checksum is offload. A pass/fail indication is provided in the Error field - IPE and L4E. 0b 1b 1b / 0b Hardware provides IPv4 checksum offload if IPCS is active and UDP checksum is offload. A pass/fail indication is provided in the Error field - IPE and L4E. See Table 26-18 for a description of supported packet types for receive checksum offloading. Unsupported packet types do not have the IPCS or L4CS bits set. IPv6 packets do not have the IPCS bit set, but might have the L4CS bit set if the Controller recognized the TCP or UDP packet. PIF Hardware supplies the PIF field to expedite software processing of packets. Software must examine any packet with PIF set to determine whether to accept the packet. If PIF is clear, then the packet is known to be for this station, so software need not look at the packet contents. Multicast packets passing only the Multicast Vector (MTA) or unicast packets passing only the Unicast Hash Table (UTA) but not any of the MAC address exact filters (RAH, RAL) have PIF set. In addition, the following condition causes PIF to be cleared: * The DA of the packet is a multicast address and promiscuous multicast is set (RCTL.MPE = 1b). * The DA of the packet is a broadcast address and accept broadcast mode is set (RCTL.BAM = 1b) A MAC control frame forwarded to the host (RCTL.PMCF = 0b) that does not match any of the exact filters, has the PIF bit set. Error Field (8) Most error information appears only when the store-bad-packet bit (RCTL.SBP) is set and a bad packet is received. See Table 26-7 for a definition of the possible errors and their bit positions. Table 26-7. RXE, IPE and L4E 7 6 5 RXE IPE L4E 4 3 2 1 0 Reserved * RXE (bit 7) - RX Data Error * IPE (bit 6) - Ipv4 Checksum Error * L4E (bit 5) - TCP/UDP Checksum Error * Reserved (bit 4:0) IPE/L4E The IP and TCP/UDP checksum error bits from Table 26-7 are valid only when the IPv4 or TCP/UDP checksum(s) is performed on the received packet as indicated via IPCS and L4CS. These, along with the other error bits, are valid only when the EOP and DD bits are set in the descriptor. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1131 Note: Receive checksum errors have no affect on packet filtering. If receive checksum offloading is disabled (RXCSUM.IPOFL and RXCSUM.TUOFL), the IPE and L4E bits are 0b. RXE The RXE error bit is asserted in one of two cases (software might distinguish between these errors by monitoring the respective statistics registers): 1. CRC error is detected. CRC can be a result of reception of /V/ symbol on the TBI interface or assertion of RxERR on the MII/GMII interface or bad EOP or lose of sync during packet reception. Packets with a CRC error are posted to host memory only when store-bad-packet bit (RCTL.SBP) is set. VLAN Tag Field (16) Hardware stores additional information in the receive descriptor for 802.1q packets. If the packet type is 802.1q (determined when a packet matches VET and RCTL.VME = 1b), then the VLAN Tag field records the VLAN information and the four-byte VLAN information is stripped from the packet data storage. Otherwise, the VLAN Tag field contains 0x0000. The rule for VLAN tag is to use network ordering (also called big endian). It appears in the following manner in the descriptor: Table 26-8. VLAN Tag Field Layout (for 802.1q Packet) 15 PRI 13 12 11 CFI 0 VLAN 26.1.5 Advanced Receive Descriptors 26.1.5.1 Advanced Receive Descriptors (RDESC) - Read Format Table 26-9 shows the receive descriptor. This is the format that software writes to the descriptor queue and hardware reads from the descriptor queue in host memory. Hardware writes back the descriptor in a different format, shown in Table 26-10. Table 26-9. RDESC Descriptor Read Format 63 1 0 0 Packet Buffer Address [63:1] A0/NSE 8 Header Buffer Address [63:1] DD Packet Buffer Address (64) - Physical address of the packet buffer. The lowest bit is either A0 (LSB of address) or NSE (No-Snoop Enable), depending on bit RXCTL.RXdataWriteNSEn of the relevant queue. See Section 28.12.1.1. Header Buffer Address (64) - Physical address of the header buffer. The lowest bit is DD. Note: The Controller does not support null descriptors (a descriptor with a packet or header address that is always equal to zero). When software sets the NSE bit in the receive descriptor, the Controller places the received packet associated with this descriptor in memory at the packet buffer address with NSE set in the PCIe attribute fields. NSE does not affect the data written to the header buffer address. Intel(R) Communications Chipset 89xx Series - Datasheet 1132 October 2012 Order Number: 327879-001US 26.0 When a packet spans more than one descriptor, the header buffer address is not used for the second, third, etc. descriptors; only the packet buffer address is used in this case. NSE is enabled for packet buffers that the software device driver knows have not been touched by the processor since the last time they were used, so the data cannot be in the processor cache and snoop is always a miss. Avoiding these snoop misses improves system performance. No-snoop is particularly useful when the DMA engine is moving the data from the packet buffer into application buffers, and the software device driver is using the information in the header buffer for its work with the packet. Note: When No-Snoop Enable is used, relaxed ordering should also be enabled with CTRL_EXT.RO_DIS. 26.1.5.2 Advanced Receive Descriptors (RDESC) - Writeback Format When the Controller writes back the descriptors, it uses the descriptor format shown in Table 26-10. Note: SRRCTL[n]. DESCTYPE must be set to a value other than 000b for the Controller to write back the special descriptors. Table 26-10. RDESC Descriptor Write-Back Format 63 0 48 47 35 34 32 RSS Hash Value/Fragment Checksum and IP identification 8 VLAN Tag PKT_LEN 31 SPH 30 21 20 HDR_LEN Extended Error 19 RSV 17 16 4 Packet Type 3 0 RSS Type Extended Status RSS Type (4) Table 26-11. RSS Type Packet Type Description 0x0 No hash computation done for this packet. 0x1 HASH_TCP_IPV4 0x2 HASH_IPV4 0x3 HASH_TCP_IPV6 0x4 HASH_IPV6_EX 0x5 HASH_IPV6 0x6 HASH_TCP_IPV6_EX 0x7 HASH_UDP_IPV4 0x8 HASH_UDP_IPV6 0x9 HASH_UDP_IPV6_EX 0xA:0xF Reserved The Controller must identify the packet type and then choose the appropriate RSS hash function to be used on the packet. The RSS type reports the packet type that was used for the RSS hash function. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1133 Packet Type (13) * VPKT (bit 12) - VLAN Packet indication The 12 LSB bits of the packet type reports the packet type identified by the hardware as follows: Table 26-12. Packet Type LSB Bits (11:10) Bit Index Bit 11 = 0b Bit 11 = 1b (L2 packet) 0 IPV4 - IPv4 header present 1 IPV4E - IPv4 Header includes extensions 2 IPV6 - IPv6 header present 3 IPV6E- IPv6 Header includes extensions 4 TCP - TCP header present 5 UDP - UDP header present Reserved 6 SCTP - SCTP header present Reserved 7 NFS - NFS header present Reserved EtherType - ETQF register index that matches the packet. Special types might be defined for 1588, 802.1X, LLDP or any other requested type. Reserved 10:8 Reserved Reserved RSV(5): Reserved. HDR_LEN (10) - The length (bytes) of the header as parsed by the Controller. In split mode when HBO (Header Buffer Overflow) is set in the Extended error field, the HDR_LEN can be greater then zero though nothing is written to the header buffer. In header replication mode, the HDR_LEN field does not reflect the size of the data actually stored in the header buffer because the Controller fills the buffer up to the size configured by SRRCTL[n].BSIZEHEADER, which might be larger than the header size reported here. This field is only valid in the first descriptor of a packet and should be ignored in all subsequent descriptors. Note: When the packet is time stamped and the time stamp is placed at the beginning of the buffer the RDESC.HDR_LEN field is updated with the additional time stamp bytes (16 bytes). For further information see Section 26.1.10. Other packet types are posted sequentially in the host packet buffer. Each line in the following table has an enable bit in the PSRTYPE register. When one of the bits is set, the corresponding packet type is split. If the bit is not set, a packet matching the header layout is not split. Header split and replication is described in Section 26.1.9 while the packet types for this functionality are enabled by the PSRTYPE[n] registers (Section 28.9.1.3). Note: The header of a fragmented IPv6 packet is defined before the fragmented extension header. SPH (1) - Split Header - When set, indicates that the HDR_LEN field reflects the length of the header found by hardware. If cleared, the HDR_LEN field should be ignored, unless SRRCTL[n].DESCTYPE is set to Split - always use header buffer mode and PKT_LEN = 0. In this case, the HDR_LEN reflects the size of the packet, even if SPH bit is cleared. Intel(R) Communications Chipset 89xx Series - Datasheet 1134 October 2012 Order Number: 327879-001US 26.0 In the case were SRRCTL[n].DESCTYPE is set to Header replication mode, SPH bit is set but the HDR_LEN field does not reflect the size of the data actually stored in the header buffer, because the Controller fills the buffer up to the size configured by SRRCTL[n].BSIZEHEADER. RSS Hash / Fragment Checksum (32) This field has multiplexed functionality according to the received packet type (reported on the Packet Type field in this descriptor) and device setting. Fragment Checksum (16-Bit; 63:48) The fragment checksum word contains the unadjusted one's complement checksum of the IP payload and is used to offload checksum verification for fragmented UDP packets as described in Section 26.1.11.2. This field is mutually exclusive with the RSS hash. It is enabled when the RXCSUM.PCSD bit is cleared and the RXCSUM.IPPCSE bit is set. IP identification (16-Bit; 47:32) The IP identification word identifies the IP packet to whom this fragment belongs and is used to offload checksum verification for fragmented UDP packets as described in Section 26.1.11.2. This field is mutually exclusive with the RSS hash. It is enabled when the RXCSUM.PCSD bit is cleared and the RXCSUM.IPPCSE bit is set. RSS Hash Value (32) The RSS hash value is required for RSS functionality as described in Section 26.1.1.8. This bit is mutually exclusive with the fragment checksum. It is enabled when the RXCSUM.PCSD bit is set. Extended Status (20) Status information indicates whether the descriptor has been used and whether the referenced buffer is the last one for the packet. Table 26-13 lists the extended status word in the last descriptor of a packet (EOP is set). Table 26-14 lists the extended status word in any descriptor but the last one of a packet (EOP is cleared). Table 26-13. Receive Status (RDESC.STATUS) Layout of Last Descriptor 19 18 17 Rsv 16 15 TS TSIP 14 13 Reserved 12 11 10 Strip CRC LLINT UDPV VEXT Rsv PIF IPCS L4I UDPCS VP Rsv EOP DD 9 8 7 6 5 4 3 2 1 0 Table 26-14. Receive Status (RDESC.STATUS) Layout of Non-Last Descriptor 19 .................................2 Reserved 1 0 EOP = 0b DD TS (16) - Time Stamped Packet (Time Sync). The Time Stamp bit is set to indicate that the device recognized a Time Sync packet and time stamped it in the RXSTMPL/H time stamp registers (See Section 26.9.2.2 and Section 26.9.1.1). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1135 TSIP (15) - Timestamp in packet. The Timestamp In Packet bit is set to indicate that the received packet arrival time was captured by the hardware and the timestamp was placed in the receive buffer. For further details see Section 26.1.10. Reserved (2, 8, 14:13, 18) - Reserved at zero. PIF (7), IPCS(6), UDPCS(4), VP(3), EOP (1), DD (0) - These bits are described in the legacy descriptor format in Section 26.1.4. L4I (5) - This bit indicates that an L4 integrity check was done on the packet, either TCP checksum, UDP checksum or SCTP CRC checksum. This bit is valid only for the last descriptor of the packet. An error in the integrity check is indicated by the L4E bit in the error field. The type of check done can be induced from the packet type bits 4, 5 and 6. If bit 4 is set, a TCP checksum was done. If bit 5 is set a UDP checksum was done, and if bit 6 is set, a SCTP CRC checksum was done. VEXT (9) - First VLAN is found on a double VLAN packet. This bit is valid only when CTRL_EXT.EXTENDED_VLAN is set. For more details see Section 26.4.5. UDPV (10) - This bit indicates that the incoming packet contains a valid (non-zero value) checksum field in an incoming fragmented UDP Ipv4 packet. This means that the Fragment Checksum field in the receive descriptor contains the UDP checksum as described in Section 26.1.11.2. When this field is cleared in the first fragment that contains the UDP header, means that the packet does not contain a valid UDP checksum and the checksum field in the Rx descriptor should be ignored. This field is always cleared in incoming fragments that do not contain the UDP header. LLINT (11) - This bit indicates that the packet caused an immediate interrupt via the low latency interrupt mechanism. Strip CRC (12) - This bit indicates that Ethernet CRC has been stripped from incoming packet. Strip CRC operation is defined by the RCTL.SECRC bit. Extended Error (12) Table 26-15 and the text that follows describes the possible errors reported by hardware . Table 26-15. Receive Errors (RDESC.ERRORS) Layout 11 10 9 RXE IPE L4E 8 7 6 4 Reserved 3 HBO 2 0 Reserved RXE (bit 11) - RXE is described in the legacy descriptor format in Section 26.1.4. IPE (bit 10) - The IPE error indication is described in the legacy descriptor format in Section 26.1.4. L4E (bit 9) - L4 error indication - When set, indicates that hardware attempted to do an L4 integrity check as described in the L4I bit, but the check failed. Reserved (bits 6:4) HBO (bit 3) - Header Buffer Overflow Note: The HBO bit is relevant only if SPH is set. 1. In both header replication modes, HBO is set if the header size (as calculated by hardware) is bigger than the allocated buffer size (SRRCTL.BSIZEHEADER) but the replication still takes place up to the header buffer size. Hardware sets this bit in order to indicate to software that it needs to allocate bigger buffers for the headers. Intel(R) Communications Chipset 89xx Series - Datasheet 1136 October 2012 Order Number: 327879-001US 26.0 2. In header split mode, when SRRCTL[n] BSIZEHEADER is smaller than HDR_LEN, then HBO is set to 1b, In this case, the header is not split. Instead, the header resides within the host packet buffer. The HDR_LEN field is still valid and equal to the calculated size of the header. However, the header is not copied into the header buffer. 3. In header split mode, always use header buffer mode, when SRRCTL[n] BSIZEHEADER is smaller than HDR_LEN, then HBO is set to 1b. In this case, the header buffer is used as part of the data buffers and contains the first BSIZEHEADER bytes of the packet. The HDR_LEN field is still valid and equal to the calculated size of the header. Note: Most error information appears only when the store-bad-packet bit (RCTL.SBP) is set and a bad packet is received. Reserved (bits 2:0) - Reserved PKT_LEN (16) - Number of bytes existing in the host packet buffer The length covers the data written to a receive buffer including CRC bytes (if any). Software must read multiple descriptors to determine the complete length for packets that span multiple receive buffers. If SRRCTL.DESC_TYPE = 4 (advanced descriptor header replication large packet only) and the total packet length is smaller than the size of the header buffer (no replication is done), this field continues to reflect the size of the packet, although no data is written to the packet buffer. Otherwise, if the buffer is not split because the header is bigger than the allocated header buffer, this field reflects the size of the data written to the first packet buffer (header and data). Note: When the packet is time stamped and the time stamp is placed at the beginning of the buffer, the RDESC.PKT_LEN field is updated with the additional time stamp bytes (16 bytes). For further information see Section 26.1.10. VLAN Tag (16) These bits are described in the legacy descriptor format in Section 26.1.4. 26.1.6 Receive Descriptor Fetching The fetching algorithm attempts to make the best use of PCIe bandwidth by fetching a cache-line (or more) descriptor with each burst. The following paragraphs briefly describe the descriptor fetch algorithm and the software control provided. When the on-chip buffer is empty, a fetch happens as soon as any descriptors are made available (host writes to the tail pointer). When the on-chip buffer is nearly empty (RXDCTL.PTHRESH), a prefetch is performed each time enough valid descriptors (RXDCTL.HTHRESH) are available in host memory. When the number of descriptors in host memory is greater than the available on-chip descriptor cache, the Controller might elect to perform a fetch that is not a multiple of cache-line size. Hardware performs this non-aligned fetch if doing so results in the next descriptor fetch being aligned on a cache-line boundary. This enables the descriptor fetch mechanism to be most efficient in the cases where it has fallen behind software. All fetch decisions are based on the number of descriptors available and do not take into account any split of the transaction due to bus access limitations. Note: The Controller NEVER fetches descriptors beyond the descriptor tail pointer. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1137 26.1.7 Receive Descriptor Write-Back Processors have cache-line sizes that are larger than the receive descriptor size (16 bytes). Consequently, writing back descriptor information for each received packet would cause expensive partial cache-line updates. A receive descriptor packing mechanism minimizes the occurrence of partial line write-backs. To maximize memory efficiency, receive descriptors are packed together and written as a cache-line whenever possible. Descriptors write-backs accumulate and are opportunistically written out in cache line-oriented chunks, under the following scenarios: * RXDCTL.WTHRESH descriptors have been used (the specified maximum threshold of unwritten used descriptors has been reached). * The receive timer expires (EITR) - in this case all descriptors are flushed ignoring any cache-line boundaries. * Explicit software flush (RXDCTL.SWFLS). * Dynamic packets - if at least one of the descriptors that are waiting for write-back are classified as packets requiring immediate notification the entire queue is flushed out. When the number of descriptors specified by RXDCTL.WTHRESH have been used, they are written back regardless of cache-line alignment. It is therefore recommended that WTHRESH be a multiple of cache-line size. When the receive timer (EITR) expires, all used descriptors are forced to be written back prior to initiating the interrupt, for consistency. Software might explicitly flush accumulated descriptors by writing the RXDCTLn register with the SWFLS bit set. When the Controller does a partial cache-line write-back, it attempts to recover to cache-line alignment on the next write-back. For applications where the latency of received packets is more important that the bus efficiency and the CPU utilization, an EITR value of zero may be used. In this case, each receive descriptor will be written to the host immediately. If RXDCTL.WTHRESH equals zero, then each descriptor will be written back separately, otherwise, write back of descriptors may be coalesced if descriptor accumulates in the internal descriptor ring due to bandwidth constraints. All write-back decisions are based on the number of descriptors available and do not take into account any split of the transaction due to bus access limitations. 26.1.8 Receive Descriptor Ring Structure Figure 26-16 shows the structure of each of the 16 receive descriptor rings. Hardware maintains 16 circular queues of descriptors and writes back used descriptors just prior to advancing the head pointer(s). Head and tail pointers wrap back to base when size descriptors have been processed. Intel(R) Communications Chipset 89xx Series - Datasheet 1138 October 2012 Order Number: 327879-001US 26.0 Figure 26-16.Receive Descriptor Ring Structure Circular Buffer Queues Base Head Receive Queue Tail Base + Size Software inserts receive descriptors by advancing the tail pointer(s) to refer to the address of the entry just beyond the last valid descriptor. This is accomplished by writing the descriptor tail register(s) with the offset of the entry beyond the last valid descriptor. The hardware adjusts its internal tail pointer(s) accordingly. As packets arrive, they are stored in memory and the head pointer(s) is incremented by hardware. When the head pointer(s) is equal to the tail pointer(s), the queue(s) is empty. Hardware stops storing packets in system memory until software advances the tail pointer(s), making more receive buffers available. The receive descriptor head and tail pointers reference to 16-byte blocks of memory. Shaded boxes in Figure 26-16 represent descriptors that have stored incoming packets but have not yet been recognized by software. Software can determine if a receive buffer is valid by reading the descriptors in memory. Any descriptor with a non-zero DD value has been processed by the hardware and is ready to be handled by the software. Note: The head pointer points to the next descriptor that is written back. After the descriptor write-back operation completes, this pointer is incremented by the number of descriptors written back. Hardware owns all descriptors between [head... tail]. Any descriptor not in this range is owned by software. The receive descriptor rings are described by the following registers: * Receive Descriptor Base Address (RDBA7 to RDBA0) register: This register indicates the start of the descriptor ring buffer. This 64-bit address is aligned on a 16-byte boundary and is stored in two consecutive 32-bit registers. Hardware ignores the lower 4 bits. * Receive Descriptor Length (RDLEN7 to RDLEN0) registers: This register determines the number of bytes allocated to the circular buffer. This value must be a multiple of 128 (the maximum cache-line size). Since each October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1139 descriptor is 16 bytes in length, the total number of receive descriptors is always a multiple of eight. * Receive Descriptor Head (RDH7 to RDH0) registers: This register holds a value that is an offset from the base and indicates the in-progress descriptor. There can be up to 64 KB, 8 KB descriptors in the circular buffer. Hardware maintains a shadow copy that includes those descriptors completed but not yet stored in memory. * Receive Descriptor Tail (RDT7 to RDT0) registers: This register holds a value that is an offset from the base and identifies the location beyond the last descriptor hardware can process. This is the location where software writes the first new descriptor. If software statically allocates buffers, uses legacy receive descriptors, and uses memory read to check for completed descriptors, it has to zero the status byte in the descriptor before bumping the tail pointer to make it ready for reuse by hardware. Zeroing the status byte is not a hardware requirement but is necessary for performing an in-memory scan. All the registers controlling the descriptor rings behavior should be set before receive is enabled, apart from the tail registers that are used during the regular flow of data. 26.1.8.1 Low Receive Descriptors Threshold As described above, the size of the receive queues is measured by the number of receive descriptor. During run time the software processes completed descriptors and then increments the Receive Descriptor Tail registers (RDT). At the same time, the hardware may post new packets received from the LAN incrementing the Receive Descriptor Head registers (RDH) for each used descriptor. The number of usable (free) descriptors for the hardware is the distance between Tail and Head registers. When the Tail reaches the Head, there are no free descriptors and further packets may be either dropped or block the receive FIFO. In order to avoid this behavior, the Controller may generate a low latency interrupt (associated with the relevant receive queue) once the amount of free descriptors is less or equal than the threshold. The threshold is defined in 16 descriptors granularity per queue in the SRRCTL[n].RDMTS field. 26.1.9 Header Splitting and Replication 26.1.9.1 Purpose This feature consists of splitting or replicating packet's header to a different memory space. This helps the host to fetch headers only for processing: headers are replicated through a regular snoop transaction in order to be processed by the host CPU. It is recommended to perform this transaction with the DCA feature enabled (see Section 28.12) or in conjunction with a software-prefetch. The packet (header and payload) is stored in memory through a (optionally) non-snoop transaction. Later, a transaction moves the payload from the software device driver buffer to application memory or it is moved using a normal memory copy operation. The Controller supports header splitting in several modes: * Legacy mode: legacy descriptors are used; headers and payloads are not split. * Advanced mode, no split: advanced descriptors are in use; header and payload are not split. Intel(R) Communications Chipset 89xx Series - Datasheet 1140 October 2012 Order Number: 327879-001US 26.0 * Advanced mode, split: advanced descriptors are in use; header and payload are split to different buffers. If the packet cannot be split, only the packet buffer is used. * Advanced mode, replication: advanced descriptors are in use; header is replicated in a separate buffer and also in a payload buffer. * Advanced mode, replication, conditioned by packet size: advanced descriptors are in use; replication is performed only if the packet is larger than the header buffer size. * Advanced mode, split, always use header buffer: advanced descriptors are in use; header and payload are split to different buffers. If no split is done, the first part of the packet is stored in the header buffer. 26.1.9.2 Description In Figure 26-17.Header Splitting 63 32 31 0 0 Packet Buffer Address 8 Header Buffer Address Header Header Buffer 1 Payload Payload Buffer 0 Host Memory October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1141 Figure 26-18.Header Replication 63 32 31 0 Packet Buffer Address 8 Header Buffer Address 0 Header and First Part of Payload Header Header Buffer Header and Payload Packet Buffer Payload Host Memory The physical address of each buffer is written in the Buffer Addresses fields. The sizes of these buffers are statically defined by BSIZEPACKET and BSIZEHEADER fields in the SRRCTL[n] registers. The packet buffer address includes the address of the buffer assigned to the replicated packet, including header and data payload portions of the received packet. In the case of a split header, only the payload is included. The header buffer address includes the address of the buffer that contains the header information. The receive DMA module stores the header portion of the received packets into this buffer. The Controller uses the packet replication or splitting feature when the SRRCTL[n].DESCTYPE is larger than one. The software device driver must also program the buffer sizes in the SRRCTL[n] registers. When header split is selected, the packet is split only on selected types of packets. A bit exists for each option in PSRTYPE[n] registers so several options can be used in conjunction with them. If one or more bits are set, the splitting is performed for the corresponding packet type. The following table lists the behavior of the Controller in the different modes Table 26-16. GBE Controller Split/Replicated Header Behavior DESCTYPE Split Condition SPH HBO HDR_LEN Header and Payload DMA 1. Header can't be decoded 0b 0b Min(Packet length, Buffer size) N/A Header + Payload Packet buffer 2. Header <= BSIZEHEADER 1b 0b Min(Payload length, Buffer size)a Header size Header Header buffer Payload Packet buffer 3. Header > BSIZEHEADER 1b 1b Min(Packet length, Buffer size) Header sizeb Header + Payload Packet buffer Intel(R) Communications Chipset 89xx Series - Datasheet 1142 PKT_LEN October 2012 Order Number: 327879-001US 26.0 Table 26-16. GBE Controller Split/Replicated Header Behavior Split - always use header buffer Replicate Large Packet only 1. Packet length <= BSIZEHEADER 0b 0b Zero Packet length Header + Payload Header buffer 2. Header can't be Decoded and Packet length > BSIZEHEADER 0b 0b Min(Packet length - BSIZEHEADER, Data Buffer size) BSIZEHEADER Header + Payload Header + Packet buffersc 3. Header <= BSIZEHEADER and Packet length >= BSIZEHEADER 1b 0b Min(Payload length, Data Buffer size) Header Size Header Header buffer Payload Packet buffer 4. Header > BSIZEHEADER 1b 1b Min(Packet length - BSIZEHEADER, Data Buffer size) Header Size2 Header + Payload Header + Packet bufferc 1. Header + Payload <= BSIZEHEADER 0b/1b4 0b Packet length Header size, N/A4 Header + Payload Header buffer Min(Packet length, Buffer size) Header size, N/A4 (Header + Payload)(partial6) Header buffer Header + Payload Packet buffer 2. Header + Payload > BSIZEHEADER 4 0b/1b 0b/1b 5 a. In a header only packet (such as TCP ACK packet), the PKT_LEN is zero. b. The HDR_LEN doesn't reflect the actual data size stored in the Header buffer. It reflects the header size determined by the parser. When timestamp in packet is enabled header size reflects the additional 16 bytes of the timestamp. c. If the packet spans more than one descriptor, only the header buffer of the first descriptor is used. The header buffer is used for the first part of the packet until it is filled up, and then the first packet buffer is used for the continuation of the packet. Software Notes: * If SRRCTL[n].NSE is set, all buffers' addresses in a packet descriptor must be word aligned. * Packet header can't span across buffers, therefore, the size of the header buffer must be larger than any expected header size. Otherwise, only the part of the header fitting the header buffer is replicated. In the case of header split mode (SRRCTL[n].DESCTYPE = 010b), a packet with a header larger than the header buffer is not split. 26.1.10 Receive Packet Timestamp in Buffer The Controller supports adding an optional tailored header before the MAC header of the packet in the receive buffer. The 64 MSB bits of the 128 bit tailored header include a timestamp composed of the packet reception time measured in the SYSTIML (Low DW) and SYSTIMH (High DW) registers (See Section 26.9.2.1 for further information on SYSTIML/H operation). The 64 LSB bits of the tailored header are reserved. The timestamp information is placed in Networking order (Big Endian) format as can be seen in Table 26-17. Table 26-17. Timestamp Layout in Buffer 0 3 4 SYTIMH 7 STIML 8 11 Reserved (0x0) 12 15 Reserved (0x0) 16... Received Packet When SRRCTL[n].Timestamp is set to 1, packets received to the queue will be time stamped if they meet one of the following conditions: -- Meet the criteria defined in the TSYNCRXCTL.Type field (See Section 28.16.1.1 and Section 28.16.1.26). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1143 -- Match the value defined in one of the ETQF registers with the 1588 time stamp bit set (See Section 26.1.1.4). -- Match a 2-tuple filter with the TTQF.1588 time stamp set (See Section 26.1.1.5). When detecting a receive packet that should be time stamped, the Controller will: * Place a 64 bit timestamp, indicating the time a packet was received by the MAC, at the beginning of the receive buffer before the received packet. * Set the TSIP bit in the RDESC.STATUS field of the last receive descriptor. * Update the RDESC.Packet Type field in the last receive descriptor. Value in this field enables identifying that this is a PTP (Precision Time Protocol) packet (this indication is only relevant for L2 packets). * Update the RDESC.HDR_LEN and RDESC.PKT_LEN values to include size of timestamp. Software driver should take into account the additional size of the timestamp when preparing the receive descriptors for the relevant queue. 26.1.11 Receive Packet Checksum Off Loading The Controller supports the off loading of three receive checksum calculations: the packet checksum, the IPv4 header checksum, and the TCP/UDP checksum. The packet checksum is the one's complement over the receive packet, starting from the byte indicated by RXCSUM.PCSS (zero corresponds to the first byte of the packet), after stripping. For packets with a VLAN header, the packet checksum includes the header if VLAN striping is not enabled by the CTRL.VME. If a VLAN header strip is enabled, the packet checksum and the starting offset of the packet checksum exclude the VLAN header due to masking of VLAN header. For example, for an Ethernet II frame encapsulated as an 802.3ac VLAN packet and CTRL.VME is set and with RXCSUM.PCSS set to 14, the packet checksum would include the entire encapsulated frame, excluding the 14-byte Ethernet header (DA, SA, type/length) and the 4-byte q-tag. The packet checksum does not include the Ethernet CRC if the RCTL.SECRC bit is set. Software must make the required offsetting computation (to remove the bytes that should not have been included and to include the pseudo-header) prior to comparing the packet checksum against the TCP checksum stored in the packet. For supported packet/frame types, the entire checksum calculation can be off loaded to the Controller. If RXCSUM.IPOFL is set to 1b, the Controller calculates the IPv4 checksum and indicates a pass/fail indication to software via the IPv4 Checksum Error bit (RDESC.IPE) in the Error field of the receive descriptor. Similarly, if RXCSUM.TUOFL is set to 1b, the Controller calculates the TCP or UDP checksum and indicates a pass/fail condition to software via the TCP/UDP Checksum Error bit (RDESC.L4E). These error bits are valid when the respective status bits indicate the checksum was calculated for the packet (RDESC.IPCS and RDESC.L4CS, respectively). Similarly, if RFCTL.Ipv6_DIS and RFCTL.IP6Xsum_DIS are cleared to 0b and RXCSUM.TUOFL is set to 1b, the Controller calculates the TCP or UDP checksum for IPv6 packets. It then indicates a pass/fail condition in the TCP/UDP Checksum Error bit (RDESC.L4E). If neither RXCSUM.IPOFL nor RXCSUM.TUOFL are set, the Checksum Error bits (IPE and L4E) are 0b for all packets. Supported frame types: * Ethernet II * Ethernet SNAP Intel(R) Communications Chipset 89xx Series - Datasheet 1144 October 2012 Order Number: 327879-001US 26.0 Table 26-18. Supported Receive Checksum Capabilities Hardware IP Checksum Calculation Packet Type Hardware TCP/UDP Checksum Calculation IPv4 packets. Yes Yes IPv6 packets. No (n/a) Yes IPv6 packet with next header options: * Hop-by-hop options * Destinations options (without Home option) * Destinations options (with Home option) * Routing (with Segments Left zero) * Routing (with Segments Left > zero) * Fragment No No No No No No Yes Yes No Yes No No IPv4 tunnels: * IPv4 packet in an IPv4 tunnel. * IPv6 packet in an IPv4 tunnel. No Yes (IPv4) No Yesa IPv6 tunnels: * IPv4 packet in an IPv6 tunnel. * IPv6 packet in an IPv6 tunnel. No No No No Packet is an IPv4 fragment. Yes Nob Packet is greater than 1518, 1522 or 1526 bytes (LPE=1b)c. Yes Yes Packet has 802.3ac tag. Yes Yes IPv4 packet has IP options (IP header is longer than 20 bytes). Yes Yes Packet has TCP or UDP options. Yes Yes IP header's protocol field contains a protocol number other than TCP or UDP. Yes No (n/a) (n/a) (n/a) (n/a) (n/a) (n/a) a. The IPv6 header portion can include supported extension headers as described in the IPv6 filter section. b. UDP checksum of first fragment is supported. c. Depends on number of VLAN tags. 26.1.11.1 Filters Details The previous table lists general details about what packets are processed. In more detail, the packets are passed through a series of filters to determine if a receive checksum is calculated: 26.1.11.1.1 MAC Address Filter This filter checks the MAC destination address to be sure it is valid (such as IA match, broadcast, multicast, etc.). The receive configuration settings determine which MAC addresses are accepted. See the various receive control configuration registers such as RCTL (RCTL.UPE, RCTL.MPE, RCTL.BAM), MTA, RAL, and RAH. 26.1.11.1.2 SNAP/VLAN Filter This filter checks the next headers looking for an IP header. It is capable of decoding Ethernet II, Ethernet SNAP, and IEEE 802.3ac headers. It skips past any of these intermediate headers and looks for the IP header. The receive configuration settings determine which next headers are accepted. See the various receive control configuration registers such as RCTL (RCTL.VFE), VET, and VFTA. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1145 26.1.11.1.3 IPv4 Filter This filter checks for valid IPv4 headers. The version field is checked for a correct value (4). IPv4 headers are accepted if they are any size greater than or equal to five (Dwords). If the IPv4 header is properly decoded, the IP checksum is checked for validity. The RXCSUM.IPOFL bit must be set for this filter to pass. 26.1.11.1.4 IPv6 Filter This filter checks for valid IPpv6 headers, which are a fixed size and have no checksum. The IPv6 extension headers accepted are: hop-by-hop, destination options, and routing. The maximum size next header accepted is 16 Dwords (64 bytes). 26.1.11.1.5 IPv6 Extension Headers IPv4 and TCP provide header lengths, which enable hardware to easily navigate through these headers on packet reception for calculating checksum and CRCs, etc. For receiving IPv6 packets; however, there is no IP header length to help hardware find the packet's ULP (such as TCP or UDP) header. One or more IPv6 extension headers might exist in a packet between the basic IPv6 header and the ULP header. The hardware must skip over these extension headers to calculate the TCP or UDP checksum for received packets. The IPv6 header length without extensions is 40 bytes. The IPv6 field Next Header Type indicates what type of header follows the IPv6 header at offset 40. It might be an upper layer protocol header such as TCP or UDP (Next Header Type of 6 or 17, respectively), or it might indicate that an extension header follows. The final extension header indicates with its Next Header Type field the type of ULP header for the packet. IPv6 extension headers have a specified order. However, destinations must be able to process these headers in any order. Also, IPv6 (or IPv4) might be tunneled using IPv6, and thus another IPv6 (or IPv4) header and potentially its extension headers might be found after the extension headers. The IPv4 Next Header Type is at byte offset nine. In IPv6, the first Next Header Type is at byte offset six. All IPv6 extension headers have the Next Header Type in their first eight bits. Most have the length in the second eight bits (Offset Byte[1]) as shown: Table 26-19. Typical IPv6 Extended Header Format (Traditional Representation) 0 1 2 34 5 6 Next Header Type 7 8 9 0 12 1 3 4 5 6 7 8 9 0 2 1 2 3 4 5 67 8 9 0 3 1 Length The following table lists the encoding of the Next Header Type field and information on determining each header type's length. The IPv6 extension headers are not otherwise processed by the Controller so their details are not covered here. Intel(R) Communications Chipset 89xx Series - Datasheet 1146 October 2012 Order Number: 327879-001US 26.0 Table 26-20. Header Type Encoding and Lengths Header Next Header Type Header Length (Units are Bytes Unless Otherwise Specified) IPv6 6 Always 40 bytes IPv4 4 Offset Bits[7:4] Unit = 4 bytes TCP 6 Offset Byte[12].Bits[7:4] Unit = 4 bytes UDP 17 Always 8 bytes Hop by Hop Options 0 (Note 1) 8+Offset Byte[1] Destination Options 60 8+Offset Byte[1] Routing 43 8+Offset Byte[1] Fragment 44 Always 8 bytes Authentication 51 8+4*(Offset Byte[1]) Encapsulating Security Payload 50 Note 3 No Next Header 59 Note 2 Notes: 1. Hop-by-hop options header is only found in the first Next Header Type of an IPv6 header. 2. When a No Next Header type is encountered, the rest of the packet should not be processed. 3. Encapsulated security payload - the Controller cannot offload packets with this header type. The Controller hardware acceleration does not support all IPv6 extension header types (see Table 26-18). The RFCTL.Ipv6_DIS bit must be cleared for this filter to pass. 26.1.11.1.6 UDP/TCP Filter This filter checks for a valid UDP or TCP header. The prototype next header values are 0x11 and 0x06, respectively. The RXCSUM.TUOFL bit must be set for this filter to pass. 26.1.11.2 Receive UDP Fragmentation Checksum The Controller might provide receive fragmented UDP checksum offload. The Controller should be configured in the following manner to enable this mode: The RXCSUM.PCSD bit should be cleared. The Packet Checksum and IP Identification fields are mutually exclusive with the RSS hash. When the RXCSUM.PCSD bit is cleared, Packet Checksum and IP Identification are active instead of RSS hash. The RXCSUM.IPPCSE bit should be set. This field enables the IP payload checksum enable that is designed for the fragmented UDP checksum. The RXCSUM.PCSS field must be zero. The packet checksum start should be zero to enable auto-start of the checksum calculation. The following table lists the exact description of the checksum calculation. The following table also lists the outcome descriptor fields for the following incoming packets types: October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1147 Table 26-21. Descriptor Fields Incoming Packet Type Fragment Checksum UDPV UDPCS / L4CS Non IP Packet 0b 0b 0b / 0b Ipv6 Packet 0b 0b Depends on transport header. Non fragmented Ipv4 packet 0b 0b Depends on transport header. Fragmented Ipv4, when not first fragment The unadjusted one's complement checksum of the IP payload. 0b 1b / 0b Fragmented Ipv4, for the first fragment Same as above 1 if the UDP header checksum is valid (not zero) 1b / 0b Note: When the software device driver computes the 16-bit ones complement, the sum on the incoming packets of the UDP fragments, it should expect a value of 0xFFFF. See Section 26.1.11 for supported packet formats. 26.1.12 SCTP Offload If a receive packet is identified as SCTP, the Controller checks the CRC32 checksum of this packet and identifies this packet as SCTP. Software is notified of the CRC check via the L4I bit in the Extended Status field of the Rx descriptor. The detection of a SCTP packet is indicated via the SCTP bit in the packet Type field of the Rx descriptor. The checker assumes the following SCTP packet format: Table 26-22. SCTP Header 0 1 2 3 4 5 6 7 8 1 9 0 12 3 4 5 6 7 Source Port 8 2 9 0 1 2 3 4 5 67 8 3 9 0 1 Destination Port Verification Tag Checksum Chunks 1...n 26.2 Transmit Functionality 26.2.1 Packet Transmission Output packets are made up of pointer-length pairs constituting a descriptor chain (descriptor based transmission). Software forms transmit packets by assembling the list of pointer-length pairs, storing this information in the transmit descriptor, and then updating the on-chip transmit tail pointer to the descriptor. The transmit descriptor and buffers are stored in host memory. Hardware typically transmits the packet only after it has completely fetched all the L2 packet data from host memory and deposited it into the on-chip transmit FIFO. This permits TCP or UDP checksum computation and avoids problems with PCIe under-runs. Another transmit feature of the Controller is TCP/UDP segmentation. The hardware has the capability to perform packet segmentation on large data buffers offloaded from the Network Operating System (NOS). This feature is discussed in detail in Section 26.2.4. In addition, the Controller supports SCTP offloading for transmit requests. See section Section 26.2.5.3 for details about SCTP. Intel(R) Communications Chipset 89xx Series - Datasheet 1148 October 2012 Order Number: 327879-001US 26.0 26.2.1.1 Transmit Data Storage Data is stored in buffers pointed to by the descriptors. Alignment of data is on an arbitrary byte boundary with the maximum size per descriptor limited only to the maximum allowed packet size (9728 bytes). A packet typically consists of two (or more) buffers, one (or more) for the header and one for the actual data. Each buffer is referenced by a different descriptor. Some software implementations copy the header(s) and packet data into one buffer and use only one descriptor per transmitted packet. 26.2.1.2 On-Chip Transmit Buffers The Controller allocates by default a 20KB on-chip packet buffer per port. The buffer can be used to store packets until they are transmitted on the line. The Controller utilizes a single common ram structure for the on-chip transmit buffers allocated to the various ports. If a port is disabled, so that it can't be accessed by host and management, by setting EEPROM bit LAN_DIS in the "Software Defined Pins Control" word for the relevant port to 1. The freed buffer space can be allocated to the active ports via the "Initialization Control 4" EEPROM word. Actual on-chip transmit buffer allocated to the port can be read in the ITPBS register. 26.2.1.3 On-Chip Descriptor Buffers The Controller contains a 24 descriptor cache for each transmit queue used to reduce the latency of packet processing and to optimize the usage of the PCIe bandwidth by fetching and writing back descriptors in bursts. The fetch and writeback algorithm are described in Section 26.2.2.5 and Section 26.2.2.6. 26.2.1.4 Transmit Contexts The Controller provides hardware checksum offload and TCP/UDP segmentation facilities. These features enable TCP and UDP packet types to be handled more efficiently by performing additional work in hardware, thus reducing the software overhead associated with preparing these packets for transmission. Part of the parameters used by these features is handled though contexts. A context refers to a set of device registers loaded or accessed as a group to provide a particular function. The Controller supports 2x8 context register sets (two per queue) per port on-chip. The transmit queues can contain transmit data descriptors (similar to the receive queue) as well as transmit context descriptors. The contexts are queue specific and one context cannot be reused from one queue to another. This differs from the method used in previous devices that supported a pool of contexts to be shared between queues. A transmit context descriptor differs from a data descriptor as it does not point to packet data. Instead, this descriptor provides the ability to write to the on-chip context register sets that support the transmit checksum offloading and the segmentation features of the Controller. The Controller supports one type of transmit context. This on-chip context is written with a transmit context descriptor DTYP=2 and is always used as context for transmit data descriptor DTYP=3. The IDX field contains an index to one of the two queue contexts. Software must track what context is stored in each IDX location. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1149 Each advanced data descriptor that uses any of the advanced offloading features must refer to a context. Contexts can be initialized with a transmit context descriptor and then used for a series of related transmit data descriptors. The context, for example, defines the checksum and offload capabilities for a given type of TCP/IP flow. All packets of this type can be sent using this context. Software is responsible for ensuring that a context is only overwritten when it is no longer needed. Hardware does not include any logic to manage the on-chip contexts; it is completely up to software to populate and then use the on-chip context table. Note: Software should not queue more than 2 context descriptors in sequence without an intervening data descriptor, to achieve adequate performance. Each context defines information about the packet sent including the total size of the MAC header (TDESC.MACHDR), the maximum amount of payload data that should be included in each packet (TDESC.MSS), TCP header length (TDES.TCPHDR), IP header length (TDESC.IPHDR), and information about what type of protocol (TCP, IP, etc.) is used. Other than TCP, IP (TDESC.TUCMD), most information is specific to the segmentation capability. Because there are dedicated on-chip resources for contexts, they remain constant until they are modified by another context descriptor. This means that a context can be used for multiple packets (or multiple segmentation blocks) unless a new context is loaded prior to each new packet. Depending on the environment, it might be unnecessary to load a new context for each packet. For example, if most traffic generated from a given node is standard TCP frames, this context could be setup once and used for many frames. Only when some other frame type is required would a new context need to be loaded by software. This new context could use a different index or the same index. This same logic can also be applied to the TCP/UDP segmentation scenario, though the environment is a more restrictive one. In this scenario, the host is commonly asked to send messages of the same type, TCP/IP for instance, and these messages also have the same Maximum Segment Size (MSS). In this instance, the same context could be used for multiple TCP messages that require hardware segmentation. 26.2.2 Transmit Descriptors The PCH supports legacy descriptors and advanced descriptors. Legacy descriptors are intended to support legacy drivers to enable fast platform power up and to facilitate debug. These descriptors should not be used when advanced features such as virtualization are used. The Legacy descriptors are recognized as such based on the DEXT bit as discussed later in this section. In addition, the Controller supports two types of advanced transmit descriptors: 1. Advanced Transmit Context Descriptor, DTYP = 0010b. 2. Advanced Transmit Data Descriptor, DTYP = 0011b. Note: DTYP values 0000b and 0001b are reserved. The transmit data descriptor (both legacy and advanced) points to a block of packet data to be transmitted. The advanced transmit context descriptor does not point to packet data. It contains control/context information that is loaded into on-chip registers that affect the processing of packets for transmission. The following sections describe the descriptor formats. Intel(R) Communications Chipset 89xx Series - Datasheet 1150 October 2012 Order Number: 327879-001US 26.0 26.2.2.1 Legacy Transmit Descriptor Format Legacy descriptors are identified by having bit 29 of the descriptor (TDESC.DEXT) set to 0b. In this case, the descriptor format is defined as shown in Table 26-23. The address and length must be supplied by software. Also, bits in the command byte are optional, as are the CSO, and CSS fields. Table 26-23. Transmit Descriptor (TDESC) Fetch Layout - Legacy Mode 63 48 47 40 39 36 0 8 35 32 31 24 23 16 15 0 Buffer Address [63:0] VLAN CSS ExtCMD STA CMD CSO Length Table 26-24. Transmit Descriptor (TDESC) Write-Back Layout - Legacy Mode 63 48 0 8 47 40 39 36 35 32 31 24 23 16 Reserved VLAN CSS 15 0 Reserved Reserved STA CMD CSO Length Note: For frames that span multiple descriptors, the VLAN, CSS, CSO, CMD.VLE, CMD.IC, and CMD.IFCS are valid only in the first descriptors and are ignored in the subsequent ones. 26.2.2.1.1 Length Length (TDESC.LENGTH) specifies the length in bytes to be fetched from the buffer address provided. The maximum length associated with any single legacy descriptor is 9728 bytes. Descriptor length(s) might be limited by the size of the transmit FIFO. All buffers comprising a single packet must be able to be stored simultaneously in the transmit FIFO. For any individual packet, the sum of the individual descriptors' lengths must be below 9728 bytes. Note: The maximum allowable packet size for transmits can change, based on the value written to the DMA TX Max Allowable packet size (DTXMXPKTSZ) register. Note: Descriptors with zero length (null descriptors) transfer no data. Null descriptors can only appear between packets and must have their EOP bits set. Note: If the TCTL.PSP bit is set, the total length of the packet transmitted, not including FCS should be at least 17 bytes. 26.2.2.1.2 Checksum Offset and Start - CSO and CSS A Checksum Offset (TDESC.CSO) field indicates where, relative to the start of the packet, to insert a TCP checksum if this mode is enabled. A Checksum Start (TDESC.CSS) field indicates where to begin computing the checksum. Both CSO and CSS are in units of bytes and must be in the range of data provided in the descriptors. For short packets that are not padded by software, CSS and CSO must be in the range of the unpadded data length, not the eventual padded length (64 bytes). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1151 CSO must be set to the location of TCP checksum in the packet. CSS must be set to the beginning of the IP header or the L4 (TCP) header. Checksum calculation is not done if CSO or CSS are out of range. This occurs if (CSS > length) OR (CSO > length - 1). In the case of an 802.1Q header, the offset values depend on the VLAN insertion enable (VLE) bit. If they are not set (VLAN tagging included in the packet buffers), the offset values should include the VLAN tagging. If these bits are set (VLAN tagging is taken from the packet descriptor), the offset values should exclude the VLAN tagging. Note: Software must compute an offsetting entry to back out the bytes of the header that are not part of the IP pseudo header and should not be included in the TCP checksum and store it in the position where the hardware computed checksum is to be inserted. Hardware does not add the 802.1Q Ethertype or the VLAN field following the 802.1Q Ethertype to the checksum. So for VLAN packets, software can compute the values to back out only on the encapsulated packet rather than on the added fields. Note: UDP checksum calculation is not supported by the legacy descriptors. When using legacy descriptors, the Controller is not aware of the L4 type of the packet and thus, does not support the translation of a checksum result of 0x0000 to 0xFFFF needed to differentiate between an UDP packet with a checksum of zero and an UDP packet without checksum. Because the CSO field is eight bits wide, it puts a limit on the location of the checksum to 255 bytes from the beginning of the packet. Hardware adds the checksum to the field at the offset indicated by the CSO field. Checksum calculations are for the entire packet starting at the byte indicated by the CSS field. A value of zero corresponds to the first byte in the packet. CSS must be set in the first descriptor for a packet. Table 26-25. Transmit Command (TDESC.CMD) Layout 7 6 5 4 3 2 1 0 RSV VLE DEXT Rsv RS IC IFCS EOP 26.2.2.1.3 Command Byte - CMD The CMD byte stores the applicable command and has the fields shown in Figure 26-25. * RSV (bit 7) - Reserved * VLE (bit 6) - VLAN Packet Enable * DEXT (bit 5) - Descriptor Extension (0 for legacy mode) * Reserved (bit 4) - Reserved * RS (bit 3) - Report Status * IC (bit 2) - Insert Checksum * IFCS (bit 1) - Insert FCS * EOP (bit 0) - End of Packet VLE: Indicates that the packet is a VLAN packet. For example, hardware should add the VLAN Ethertype and an 802.1q VLAN tag to the packet. Intel(R) Communications Chipset 89xx Series - Datasheet 1152 October 2012 Order Number: 327879-001US 26.0 Table 26-26. VLAN Tag Insertion Decision Table VLE Action 0b Send generic Ethernet packet unless relevant bit in VMVIR registers is set. 1b Send 802.1Q packet; the Ethernet Type field comes from the VET register and the VLAN data comes from the VLAN field of the TX descriptor. RS: Signals the hardware to report the status information. This is used by software that does in-memory checks of the transmit descriptors to determine which ones are done. For example, if software queues up 10 packets to transmit, it can set the RS bit in the last descriptor of the last packet. If software maintains a list of descriptors with the RS bit set, it can look at them to determine if all packets up to (and including) the one with the RS bit set have been buffered in the output FIFO. Looking at the status byte and checking the Descriptor Done (DD) bit enables this operation. If DD is set, the descriptor has been processed. See Table 26-27 for the layout of the status field. IC: If set, requests hardware to add the checksum of the data from CSS to the end of the packet at the offset indicated by the CSO field. IFCS: When set, hardware appends the MAC FCS at the end of the packet. When cleared, software should calculate the FCS for proper CRC check. There are several cases in which software must set IFCS: * Transmitting a short packet while padding is enabled by the TCTL.PSP bit. * Checksum offload is enabled by the IC bit in the TDESC.CMD. * VLAN header insertion enabled by the VLE bit in the TDESC.CMD or by the VMVIR registers. EOP: When set, indicates this is the last descriptor making up the packet. More than one descriptor can be used to form a packet. Note: As opposed to , VLE, IFCS, CSO, and IC must be set correctly only in the first descriptor of each packet. In previous silicon generations, some of these bits were required to be set in the last descriptor of a packet. 26.2.2.1.4 Status - STA Table 26-27. Transmit Status (TDESC.STA) Layout 3 2 1 Reserved 26.2.2.1.5 0 DD DD (Bit 0) - Descriptor Done Status The DD bit provides the transmit status, when RS is set in the command: DD indicates that the descriptor is done and is written back after the descriptor has been processed. Note: When head write back is enabled (TDWBAL[n].Head_WB_En = 1), the write-back of the DD bit to the descriptor is not executed. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1153 26.2.2.1.6 VLAN The VLAN field is used to provide the 802.1q/802.1ac tagging information. The VLAN field is valid only on the first descriptor of each packet when the VLE bit is set. The rule for VLAN tag is to use network ordering. The VLAN field is placed in the transmit descriptor in the following manner: Table 26-28. VLAN Field (TDESC.VLAN) Layout 15 13 12 PRI 11 0 CFI VLAN ID * VLAN ID - the 12-bit tag indicating the VLAN group of the packet. * Canonical Form Indication (CFI) - Set to zero for Ethernet packets. * PRI - indicates the priority of the packet. Note: The VLAN tag should be sent in network order (also called big endian). 26.2.2.2 Advanced Transmit Context Descriptor Table 26-29. Transmit Context Descriptor (TDESC) Layout - (Type = 0010b) 63 0 39 Reserved 63 8 40 48 MSS 32 31 Reserved 47 40 L4LEN 39 RSV1 38 36 IDX 16 15 VLAN 35 30 Reserved 9 8 0 MACLEN 29 28 24 23 20 DEXT RSVa DTYP IPLEN 19 9 TUCMD 8 0 Reserved a. RSV - Reserved 26.2.2.2.1 IPLEN (9) IP header length. If an offload is requested, IPLEN must be greater than or equal to 20 and less than or equal to 511. 26.2.2.2.2 MACLEN (7) This field indicates the length of the MAC header. When an offload is requested (either TSE or IXSM or TXSM is set), MACHDR must be larger than or equal to 14 and less than or equal to 127. This field should include only the part of the L2 header supplied by the software device driver and not the parts added by hardware. The following table lists the value of MACLEN in the different cases. Table 26-30. MACLEN Values SNAP Regular VLAN Extended VLAN MACLEN No By hardware or no VLAN No 14 No By hardware or no VLAN Yes 18 No By software No 18 No By software Yes 22 Yes By hardware or no VLAN No 22 Intel(R) Communications Chipset 89xx Series - Datasheet 1154 October 2012 Order Number: 327879-001US 26.0 Table 26-30. MACLEN Values SNAP Regular VLAN Yes Extended VLAN By hardware or no VLAN Yes MACLEN 26 Yes By software No 26 Yes By software Yes 30 VLAN (16) - 802.1Q VLAN tag to be inserted in the packet during transmission. This VLAN tag is inserted and needed only when a packet using this context has its DCMD.VLE bit set. This field should include the entire 16-bit VLAN field including the CFI and Priority fields as shown in Table 26-28. Note: The VLAN tag should be sent in network order. 26.2.2.2.3 TUCMD (11) Table 26-31. Transmit Command (TDESC.TUCMD) Layout 10 6 Reserved 5 4 Reserved Reserved 3 2 L4T 1 0 IPV4 SNAP * RSV (bit 10:6) - Reserved * RSV (bit 5:4) - Reserved * L4T (bit 3:2) - L4 Packet TYPE (00b: UDP; 01b: TCP; 10b: SCTP; 11b: Reserved) * IPV4 (bit 1) - IP Packet Type: When 1b, Ipv4; when 0b, Ipv6 * SNAP (bit 0) - SNAP indication 26.2.2.2.4 DTYP(4) Always 0010b for this type of descriptor. 26.2.2.2.5 DEXT(1) Descriptor Extension (1b for advanced mode). 26.2.2.2.6 IDX (3) Index into the hardware context table where this context is stored. In the Controller, the 2 available register context sets per queue are accessed using the LSB bit and the two MSB bits are reserved and should always be 0. 26.2.2.2.7 L4LEN (8) Layer 4 header length. If TSE is set in the data descriptor pointing to this context, this field must be greater than or equal to 12 and less than or equal to 255. Otherwise, this field is ignored. 26.2.2.2.8 MSS (16) Controls the Maximum Segment Size (MSS). This specifies the maximum TCP payload segment sent per frame, not including any header or trailer. The total length of each frame (or section) sent by the TCP/UDP segmentation mechanism (excluding Ethernet CRC) as follows: Total length is equal to: October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1155 MACLEN + 4(if VLE set) + 4 or 8(if CMTGI is set or if also RLTTGI is set assuming BCNTLEN is clear) + IPLEN + L4LEN + MSS The one exception is the last packet of a TCP/UDP segmentation, which is typically shorter. MSS is ignored when DCMD.TSE is not set. Note: The headers lengths must meet the following: MACLEN + IPLEN + L4LEN <= 512 Note: The MSS value should be larger than 0 and the maximum MSS value should not exceed 9216 bytes (9KB) length. The context descriptor requires valid data only in the fields used by the specific offload options. The following table lists the required valid fields according to the different offload options. Table 26-32. Valid Field in Context vs. Required Offload Required Offload TSE a TXSM Valid Fields in Context IXSM b VLAN L4LEN IPLEN MACLEN MSS L4T IPV4 1b 1b X VLE Yes Yes Yes Yes Yes Yes 0b 1b X2 VLE No Yes Yes No Yes Yes 0b 0b 1b VLE No Yes Yes No No Yes 0b 0b 0b No context required unless VLE is set. a. If TSE is set, TXSM must be set to 1. b. X - don't care 26.2.2.3 Advanced Transmit Data Descriptor Table 26-33. Advanced Transmit Data Descriptor (TDESD) Layout - (Type = 0011b) 0 Address[63:0] 8 PAYLEN 63 POPTS 46 45 40 RSV1 39 IDX 38 36 STA 35 32 DCMD 31 24 DTYP MAC RSVa 23 20 19 18 17 16 DTALEN 15 0 a. RSV - Reserved Table 26-34. Advanced Tx Descriptor Write-back Format RSVa 0 8 Reserved 63 STA 36 35 32 Reserved 31 0 a. RSV - Reserved Note: For frames that span multiple descriptors, all fields apart from DCMD.EOP, DCMD.RS, DCMD.DEXT, DTALEN, Address and DTYP are valid only in the first descriptor and are ignored in the subsequent ones. Intel(R) Communications Chipset 89xx Series - Datasheet 1156 October 2012 Order Number: 327879-001US 26.0 26.2.2.3.1 Address (64) Physical address of a data buffer in host memory that contains a portion of a transmit packet. 26.2.2.3.2 DTALEN (16) Length in bytes of data buffer at the address pointed to by this specific descriptor. Note: If the TCTL.PSP bit is set, the total length of the packet transmitted, not including FCS, should be at least 17 bytes. Note: The maximum allowable packet size for transmits is based on the value written to the DMA TX Max Allowable packet size (DTXMXPKTSZ) register. Default value is 9728 bytes. 26.2.2.3.3 MAC (2) Table 26-35. Transmit Data (TDESD.MAC) Layout bit 1 bit 0 1588 Reserved * Reserved (bit 0) * 1588 (bit 1) - IEEE1588 Timestamp packet. 26.2.2.3.4 DTYP (4) 0011b is the value for this descriptor type. 26.2.2.3.5 DCMD (8) Table 26-36. Transmit Data (TDESD.DCMD) Layout 7 6 5 4 3 2 1 0 TSE VLE DEXT Reserved RS Reserved IFCS EOP * TSE (bit 7) - TCP/UDP Segmentation Enable * VLE (bit 6) - VLAN Packet Enable * DEXT (bit 5) - Descriptor Extension (1b for advanced mode) * Reserved (bit 4) * RS (bit 3) - Report Status * Reserved (bit 2) * IFCS (bit 1) - Insert FCS * EOP (bit 0) - End Of Packet TSE indicates a TCP/UDP segmentation request. When TSE is set in the first descriptor of a TCP packet, hardware must use the corresponding context descriptor in order to perform TCP segmentation. The type of segmentation applied is defined according to the TUCMD.L4T field in the context descriptor. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1157 Note: It is recommended that TCTL.PSP be enabled when TSE is used since the last frame can be shorter than 60 bytes - resulting in a bad frame if TCTL.PSP is disabled. VLE indicates that the packet is a VLAN packet and hardware must add the VLAN Ethertype and an 802.1q VLAN tag to the packet. DEXT must be 1b to indicate advanced descriptor format (as opposed to legacy). RS signals hardware to report the status information. This is used by software that does in-memory checks of the transmit descriptors to determine which ones are done. For example, if software queues up 10 packets to transmit, it can set the RS bit in the last descriptor of the last packet. If software maintains a list of descriptors with the RS bit set, it can look at them to determine if all packets up to (and including) the one with the RS bit set have been buffered in the output FIFO. Looking at the status byte and checking the DD bit do this. If DD is set, the descriptor has been processed. See the sections that follow for the layout of the status field. Note: Descriptors with zero length transfer no data. IFCS, when set, hardware appends the MAC FCS at the end of the packet. When cleared, software should calculate the FCS for proper CRC check. There are several cases in which software must set IFCS: * Transmitting a short packet while padding is enabled by the TCTL.PSP bit. * Checksum offload is enabled by the either the TXSM or IXSM bits in the TDESD.POPTS field. * VLAN header insertion enabled by the VLE bit in the TDESD.DCMD. * TCP/UDP segmentation offload enabled by TSE bit in the TDESD.DCMD. EOP indicates whether this is the last buffer for an incoming packet. 26.2.2.3.6 STA (4) * Rsv (bits 1-3) - Reserved * DD (bit 0) - Descriptor Done 26.2.2.3.7 IDX (3) Index into the hardware context table to indicate which context should be used for this request. If no offload is required, this field is not relevant and no context needs to be initiated before the packet is sent. See Table 26-32 for details in which packets require a context reference. 26.2.2.3.8 POPTS (6) Table 26-37. Transmit Data (TDESD.POPTS) Layout 5 3 Reserved 2 1 0 Reserved TXSM IXSM * Reserved (bits 5:3) * Reserved (bit 2) * TXSM (bit 1) - Insert L4 Checksum * IXSM (bit 0) - Insert IP Checksum Intel(R) Communications Chipset 89xx Series - Datasheet 1158 October 2012 Order Number: 327879-001US 26.0 TXSM, when set to 1b, L4 checksum must be inserted. In this case, TUCMD.L4T in the context descriptor indicates whether the checksum is TCP, UDP, or SCTP. When DCMD.TSE in TDESD is set, TXSM must be set to 1b. If this bit is set, the packet should at least contain a TCP header. IXSM, when set to 1b, indicates that IP checksum must be inserted. For IPv6 packets this bit must be cleared. If the DCMD.TSE bit is set in data descriptor, and TUCMD.IPV4 is set in context descriptor, POPTS.IXSM must be set to 1b as well. If this bit is set, the packet should at least contain an IP header. 26.2.2.3.9 PAYLEN (18) PAYLEN indicates the size (in byte units) of the data buffer(s) in host memory for transmission. In a single send packet, PAYLEN defines the entire packet size fetched from host memory. It does not include the fields that hardware adds such as: optional VLAN tagging, Ethernet CRC or Ethernet padding. In a large send case (regardless if it is transmitted on a single or multiple packets), PAYLEN defines the protocol payload size fetched from host memory. In TCP or UDP segmentation offload, PAYLEN defines the TCP/UDP payload size. Note: When a packet spreads over multiple descriptors, all the descriptor fields are only valid in the first descriptor of the packet, except for RS, which is always checked, and EOP, which is always set at last descriptor of the series. 26.2.2.4 Transmit Descriptor Ring Structure The transmit descriptor ring structure is shown in Figure 26-19. A pair of hardware registers maintains each transmit descriptor ring in the host memory. New descriptors are added to the queue by software by writing descriptors into the circular buffer memory region and moving the tail pointer associated with that queue. The tail pointer points to one entry beyond the last hardware owned descriptor. Transmission continues up to the descriptor where head equals tail at which point the queue is empty. Descriptors passed to hardware should not be manipulated by software until the head pointer has advanced past them. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1159 Figure 26-19.Transmit Descriptor Ring Structure Circular Buffer Base Head Transmit Queue Tail Base + Size The shaded boxes in the figure represent descriptors that are not currently owned by hardware that software can modify. The transmit descriptor ring is described by the following registers: * Transmit Descriptor Base Address register (TDBA 0-7): This register indicates the start address of the descriptor ring buffer in the host memory; this 64-bit address is aligned on a 16-byte boundary and is stored in two consecutive 32-bit registers. Hardware ignores the lower four bits. * Transmit Descriptor Length register (TDLEN 0-7): This register determines the number of bytes allocated to the circular buffer. This value must be zero modulo 128. * Transmit Descriptor Head register (TDH 0-7): This register holds a value that is an offset from the base and indicates the in-progress descriptor. There can be up to 64 KB descriptors in the circular buffer. Reading this register returns the value of head corresponding to descriptors already loaded in the output FIFO. This register reflects the internal head of the hardware write-back process including the descriptor in the posted write pipe and might point further ahead than the last descriptor actually written back to the memory. * Transmit Descriptor Tail register (TDT 0-7): This register holds a value, which is an offset from the base, and indicates the location beyond the last descriptor hardware can process. This is the location where software writes the first new descriptor. The driver should not handle to the Controller descriptors that describe a partial packet. Consequently, the number of descriptors used to describe a packet can not be larger than the ring size. The base register indicates the start of the circular descriptor queue and the length register indicates the maximum size of the descriptor ring. The lower seven bits of length are hard wired to 0b. Byte addresses within the descriptor buffer are computed as follows: address = base + (ptr * 16), where ptr is the value in the hardware head or tail register. Intel(R) Communications Chipset 89xx Series - Datasheet 1160 October 2012 Order Number: 327879-001US 26.0 The size chosen for the head and tail registers permit a maximum of 65528 (64 KB) descriptors, or approximately 16 KB packets for the transmit queue given an average of four descriptors per packet. Once activated, hardware fetches the descriptor indicated by the hardware head register. The hardware tail register points one descriptor beyond the last valid descriptor. Software can read and detect which packets have already been processed by hardware as follows: * Read the head register to determine which packets (those logically before the head) have been transferred to the on-chip FIFO or transmitted. This method is not recommended because races between the internal update of the head register and the actual write-back of descriptors might occur. * Read the value of the head as stored at the address pointed by the TDBAH/TDBAL pair. * Track the DD bits in the descriptor ring. All the registers controlling the descriptor rings behavior should be set before transmit is enabled, apart from the tail registers which are used during the regular flow of data. Note: Software can determine if a packet has been sent by either of three methods: setting the RS bit in the transmit descriptor command field or by performing a PIO read of the transmit head register, or by reading the head value written by the Controller to the address pointed by the TDWBAL and TDWBAH registers (see Section 26.2.3 for details). Checking the transmit descriptor DD bit or head value in memory eliminates a potential race condition. All descriptor data is written to the I/O bus prior to incrementing the head register, but a read of the head register could pass the data write in systems performing I/O write buffering. Updates to transmit descriptors use the same I/O write path and follow all data writes. Consequently, they are not subject to the race. In general, hardware prefetches packet data prior to transmission. Hardware typically updates the value of the head pointer after storing data in the transmit FIFO. 26.2.2.5 Transmit Descriptor Fetching The descriptor processing strategy for transmit descriptors is essentially the same as for receive descriptors except that a different set of thresholds are used. In addition the Controller enables assigning 2 priority levels to each transmit descriptor queue using the TXDCTL[n].Priority bit. A transmit descriptor belonging to a queue with the TXDCTL[n].Priority set to 1 will always be fetched prior to a transmit descriptor belonging to a queue with the TXDCTL[n].Priority bit reset to 0. The number of on-chip transmit descriptors per queue is 24. When there is an on-chip descriptor buffer empty, a fetch happens as soon as any descriptors are made available (host writes to the tail pointer). If several on-chip descriptor queues are in this situation at the same time, the queue with the highest priority as defined by the TXDCTL[n].Priority bit, is fetched. If several on-chip transmit descriptor queues with the same priority need to fetch descriptors, descriptors from queues that are more starved are fetched. If a number of queues have a similar priority and starvation level, highest indexed queue is served first and so forth, down to the lowest indexed queue. Note: The starvation level of a queue corresponds to the number of descriptors above the prefetch threshold (TXDCTL[n].PTHRESH) that are already in the internal queue. The queue is more starved if there are less descriptors in the internal transmit descriptor cache. Comparing starvation level might be done roughly, not at the single descriptor level of resolution. A queue is considered empty for the transmit descriptor fetch algorithm as long as: October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1161 * There is still no complete packet (single or large send) in its corresponding internal queue. * There is no descriptor already in its way from system memory to the internal cache. * The internal corresponding internal descriptor cache is not full. Each time a descriptor fetch request is sent for an empty queue, the maximum available number of descriptor is requested, regardless of cache alignment issues. When the on-chip buffer is nearly empty (below TXDCTL[n].PTHRESH), a prefetch is performed each time enough valid descriptors (TXDCTL[n].HTHRESH) are available in host memory and no other DMA activity of greater priority is pending (descriptor fetches and write-backs or packet data transfers). When the number of descriptors in host memory is greater than the available on-chip descriptor storage, the Controller might elect to perform a fetch that is not a multiple of cache-line size. Hardware performs this non-aligned fetch if doing so results in the next descriptor fetch being aligned on a cache-line boundary. This enables the descriptor fetch mechanism to be more efficient in the cases where it has fallen behind software. Note: The Controller NEVER fetches descriptors beyond the descriptor tail pointer. 26.2.2.6 Transmit Descriptor Write-Back The descriptor write-back policy for transmit descriptors is similar to that of the receive descriptors when the TXDCTL[n].WTHRESH value is not 0x0. In this case, all descriptors are written back regardless of the value of their RS bit. When the TXDCTL[n].WTHRESH value is 0x0, since transmit descriptor write-backs do not happen for every descriptor, only transmit descriptors that have the RS bit set are written back. Any descriptor write-back includes the full 16 bytes of the descriptor. Since the benefit of delaying and then bursting transmit descriptor write-backs is small at best, it is likely that the threshold is left at the default value (0b) to force immediate write-back of transmit descriptors with their RS bit set and to preserve backward compatibility. Descriptors are written back in one of three cases: * TXDCTL[n].WTHRESH = 0x0 and a descriptor which has RS set is ready to be written back. * The corresponding EITR counter has reached zero. * TXDCTL[n].WTHRESH > 0x0 and TXDCTL[n].WTHRESH descriptors have accumulated. For the first condition, write-backs are immediate. This is the default operation and is backward compatible with implementation. The other two conditions are only valid if descriptor bursting is enabled (Section 28.11.1.15). In the second condition, the EITR counter is used to force timely write-back of descriptors. The first packet after timer initialization starts the timer. Timer expiration flushes any accumulated descriptors and sets an interrupt event (TXDW). For the final condition, if TXDCTL[n].WTHRESH descriptors are ready for write-back, the write-back is performed. Intel(R) Communications Chipset 89xx Series - Datasheet 1162 October 2012 Order Number: 327879-001US 26.0 An additional mode in which transmit descriptors are not written back at all and the head pointer of the descriptor ring is written instead as described in Section 26.2.3. Note: When transmit ring is smaller than internal cache size (24 descriptors) then at list one full packet should be placed in the ring and TXDCTL[n].WTHRESH value should be less than ring size. If TXDCTL[n].WTHRESH is 0x0 (transmit RS mode) then at least one descriptor should have the RS ring set inside the ring. 26.2.3 Transmit Completions Head Write Back In legacy hardware, transmit requests are completed by writing the DD bit to the transmit descriptor ring. This causes cache thrash since both the software device driver and hardware are writing to the descriptor ring in host memory. Instead of writing the DD bits to signal that a transmit request completed, hardware can write the contents of the descriptor queue head to host memory. The software device driver reads that memory location to determine which transmit requests are complete. In order to improve the performance of this feature, the software device driver needs to program DCA registers to configure which CPU is processing each TX queue. 26.2.3.1 Description The head counter is reflected in a memory location that is allocated by software, for each queue. Head write back occurs if TDWBAL[n].Head_WB_En is set for this queue and the RS bit is set in the Tx descriptor, following corresponding data upload into packet buffer. If the head write-back feature is enabled, the Controller ignores WTRESH and takes in account only descriptors with the RS bit set (as if the WTRESH was set to 0b). In addition, the head write-back occurs upon EITR expiration for queues where the WB_on_EITR bit in TDWBAL is set. Software can also enable coalescing of the head write-back operations to reduce traffic on the PCIe bus, by programming the TXDCTL.HWBTHRESH field to a value greater than 0. In this case head write-back operation will occur only after the internal pending write-back count is greater than the TXDCTL.HWBTHRESH value. The software device driver has control on this feature through Tx queue 0-7 head write-back address, low (TDWBAL[n]) and high (TDWBAH[n]) registers thus supporting 64-bit address access. See registers description in Section 28.11.1.16 and Section 28.11.1.17. The 2 low register's LSB bits of the TDWBAL[n] register hold the control bits. 1. The Head_WB_En bit enables activation of the head write back feature. When TDWBAL[n].Head_WB_En is set to 1 no TX descriptor write-back is executed for this queue. 2. The WB_on_EITR bit enables head write upon EITR expiration. When Head write back operation is enabled (TDWBAL[n].Head_WB_En = 1) setting the TDWBAL[n].WB_on_EITR bit to 1 enables placing an upper limit on delay of head write-back operation. The 30 upper bits of the TDWBAL[n] register hold the lowest 32 bits of the head write-back address, assuming that the two last bits are zero. The TDWBAH[n] register holds the high part of the 64-bit address. Note: Hardware writes a full Dword when writing this value, so software should reserve enough space for each head value. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1163 Note: If software enables Head Write-Back, it must also disable PCI Express Relaxed Ordering on the write-back transactions. This is done by disabling bit 11 in the TXCTL register for each active transmit queue. See Section 28.12.1.2. Note: The Controller might update the Head with values that are larger then the last Head pointer which holds a descriptor with RS bit set, but still the value will always point to a free descriptor (descriptor that is not owned by the the Controller anymore). 26.2.4 TCP/UDP Segmentation Hardware TCP segmentation is one of the offloading options supported by the Windows* and Linux* TCP/IP stack. This is often referred to as TCP Segmentation Offloading or TSO. This feature enables the TCP/IP stack to pass to the network device driver a message to be transmitted that is bigger than the Maximum Transmission Unit (MTU) of medium. It is then the responsibility of the software device driver and hardware to divide the TCP message into MTU size frames that have appropriate layer 2 (Ethernet), 3 (IP), and 4 (TCP) headers. These headers must include sequence number, checksum fields, options and flag values as required. Some of these values (such as the checksum values) are unique for each packet of the TCP message and other fields such as the source IP address are constant for all packets associated with the TCP message. The Controller supports also UDP segmentation for embedded applications, although this offload is not supported by the regular Windows* and Linux* stacks. Any reference in this section to TCP segmentation, should be considered as referring to both TCP and UDP segmentation. Padding (TCTL.PSP) must be enabled in TCP segmentation mode, since the last frame might be shorter than 60 bytes, resulting in a bad frame if PSP is disabled. The offloading of these mechanisms from the software device driver to the Controller saves significant CPU cycles. The software device driver shares the additional tasks to support these options. 26.2.4.1 Assumptions The following assumptions apply to the TCP segmentation implementation in the Controller: * The RS bit operation is not changed. * Interrupts are set after data in buffers pointed to by individual descriptors is transferred (DMA'd) to hardware. 26.2.4.2 Transmission Process The transmission process for regular (non-TCP segmentation packets) involves: * The protocol stack receives from an application a block of data that is to be transmitted. * The protocol stack calculates the number of packets required to transmit this block based on the MTU size of the media and required packet headers. For each packet of the data block: * Ethernet, IP and TCP/UDP headers are prepared by the stack. * The stack interfaces with the software device driver and commands it to send the individual packet. * The software device driver gets the frame and interfaces with the hardware. * The hardware reads the packet from host memory (via DMA transfers). Intel(R) Communications Chipset 89xx Series - Datasheet 1164 October 2012 Order Number: 327879-001US 26.0 * The software device driver returns ownership of the packet to the Network Operating System (NOS) when hardware has completed the DMA transfer of the frame (indicated by an interrupt). The transmission process for the Controller TCP segmentation offload implementation involves: * The protocol stack receives from an application a block of data that is to be transmitted. * The stack interfaces to the software device driver and passes the block down with the appropriate header information. * The software device driver sets up the interface to the hardware (via descriptors) for the TCP segmentation context. Hardware DMA's (transfers) the packet data and performs the Ethernet packet segmentation and transmission based on offset and payload length parameters in the TCP/IP context descriptor including: * Packet encapsulation * Header generation and field updates including IPv4, IPV6, and TCP/UDP checksum generation * The software device driver returns ownership of the block of data to the NOS when hardware has completed the DMA transfer of the entire data block (indicated by an interrupt). 26.2.4.2.1 TCP Segmentation Data Fetch Control To perform TCP Segmentation in the Controller, the DMA must be able to fit at least one packet of the segmented payload into available space in the on-chip Packet Buffer. The DMA does various comparisons between the remaining payload and the Packet Buffer available space, fetching additional payload and sending additional packets as space permits. To support interleaving between descriptor queues at Ethernet frame resolution inside TSO requests, the frame header pointed to by the so called header descriptors are reread from system memory by hardware for every LSO segment. the Controller stores in an internal cache only the header's descriptors instead of the header's content. To limit the internal cache size software should not spread the L3/L4 header (TCP, UDP, IPV4 or IPV6) on more than 4 descriptors. In the last header buffer it's allowed to mix header and data. This limitation stands for up to Layer4 header included, and for IPv4 or IPv6 indifferently. 26.2.4.2.2 TCP Segmentation Write-Back Modes Since the TCP segmentation mode uses the buffers that contains the L3/L4 header multiple times, there are some limitations on the usage of different combinations of writeback and buffer release methods in order to guarantee the header buffer's availability until the entire packet is processed. These limitations are described in Table 26-38 below. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1165 Table 26-38. Write Back Options For Large Send WTHRESH RS HEAD Write Back Enable Hardware Behavior Software Expected Behavior for TSO packets. 0 Set in EOP descriptors only Disable Hardware writes back descriptors with RS bit set one at a time. Software can retake ownership of all descriptors up to last descriptor with DD bit set. 0 Set in any descriptors Disable Hardware writes back descriptors with RS bit set one at a time. Software can retake ownership of entire packets (EOP bit set) up to last descriptor with DD bit set. 0 Not set at all Disable Hardware does not write back any descriptor (since RS bit is not set) Software should poll the TDH register. The TDH register reflects the last descriptor that software can take ownership of.a Enable Hardware writes back the head pointer only at EITR expire event reflecting the last descriptor that software can take ownership of. Software may poll the TDH register or use the head value written back at EITR expire event. The TDH register reflects the last descriptor that software can take ownership of. Software can retake ownership of entire packets up to last descriptor with both DD and EOP bits set. Note: The TDH register reflects the last descriptor that software can take ownership ofa. 0 Not set at all >0 Don't care Disable Hardware writes back all the descriptors in bursts and set all the DD bits. Don't care Set in EOP descriptors only Enable Hardware writes back the Head pointer per each descriptor with RS bit set.b Software can retake ownership of all descriptors up to the descriptor pointed by the head pointer read from system memory (by interrupt or polling). Don't care Set in any descriptors Enable Hardware writes back the Head pointer per each descriptor with RS bit set. This mode is illegal since software won't access the descriptor, it cannot tell when the pointer passed the EOP descriptor. a. Polling of the TDH register is a valid method only when the RS bit is never set, otherwise race conditions between software and hardware accesses to the descriptor ring can occur. b. At EITR expire event, the Hardware writes back the head pointer reflecting the last descriptor that software can take ownership of. 26.2.4.3 TCP Segmentation Performance Performance improvements for a hardware implementation of TCP Segmentation off-load include: * The stack does not need to partition the block to fit the MTU size, saving CPU cycles. * The stack only computes one Ethernet, IP, and TCP header per segment, saving CPU cycles. * The Stack interfaces with the device driver only once per block transfer, instead of once per frame. * Larger PCIe bursts are used which improves bus efficiency (such as lowering transaction overhead). * Interrupts are easily reduced to one per TCP message instead of one per packet. * Fewer I/O accesses are required to command the hardware. Intel(R) Communications Chipset 89xx Series - Datasheet 1166 October 2012 Order Number: 327879-001US 26.0 26.2.4.4 Packet Format Typical TCP/IP transmit window size is 8760 bytes (about 6 full size frames). Today the average size on corporate Intranets is 12-14KB, and normally the maximum window size allowed is 64KB (unless Windows Scaling - RFC 1323 is specified). A TCP message can be as large as 256 KB and is generally fragmented across multiple pages in host memory. The Controller partitions the data packet into standard Ethernet frames prior to transmission. The Controller supports calculating the Ethernet, IP, TCP, and UDP headers, including checksum, on a frame-by-frame basis. Table 26-39. TCP/IP or UDP/IP Packet Format Sent by Host L2/L3/L4 Header Ethernet IPv4/IPv6 Data TCP/UDP DATA (full TCP message) Table 26-40. TCP/IP or UDP/IP Packet Format Sent by the Controller Pseudo Header (updated) Data (first MSS) FCS ... Pseudo Header (updated) Data (Next MSS) FCS ... Frame formats supported by the Controller include: * Ethernet 802.3 * IEEE 802.1Q VLAN (Ethernet 802.3ac) * Ethernet Type 2 * Ethernet SNAP * IPv4 headers with options * IPv6 headers with extensions * TCP with options * UDP with options. VLAN tag insertion might be handled by hardware Note: UDP (unlike TCP) is not a "reliable protocol", and fragmentation is not supported at the UDP level. UDP messages that are larger than the MTU size of the given network medium are normally fragmented at the IP layer. This is different from TCP, where large TCP messages can be fragmented at either the IP or TCP layers depending on the software implementation. The Controller has the ability to segment UDP traffic (in addition to TCP traffic), however, because UDP packets are generally fragmented at the IP layer, the Controller's "TCP Segmentation" feature is not normally useful to handle UDP traffic. 26.2.4.5 TCP/UDP Segmentation Indication Software indicates a TCP/UDP Segmentation transmission context to the hardware by setting up a TCP/IP Context Transmit Descriptor (see Section 26.2.2). The purpose of this descriptor is to provide information to the hardware to be used during the TCP segmentation off-load process. Setting the TSE bit in the TDESD.DCMD field to 1b indicates that this descriptor refers to the TCP Segmentation context (as opposed to the normal checksum off loading context). This causes the checksum off loading, packet length, header length, and maximum segment size parameters to be loaded from the Context descriptor into the device. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1167 The TCP Segmentation prototype header is taken from the packet data itself. Software must identity the type of packet that is being sent (IPv4/IPv6, TCP/UDP, other), calculate appropriate checksum off loading values for the desired checksum, and calculate the length of the header which is pre-appended. The header might be up to 240 bytes in length. Once the TCP Segmentation context has been set, the next descriptor provides the initial data to transfer. This first descriptor(s) must point to a packet of the type indicated. Furthermore, the data it points to might need to be modified by software as it serves as the prototype header for all packets within the TCP Segmentation context. The following sections describe the supported packet types and the various updates which are performed by hardware. This should be used as a guide to determine what must be modified in the original packet header to make it a suitable prototype header. The following summarizes the fields considered by the driver for modification in constructing the prototype header. IP Header For IPv4 headers: * Identification Field should be set as appropriate for first packet of send (if not already) * Header Checksum should be zeroed out unless some adjustment is needed by the driver TCP Header * Sequence Number should be set as appropriate for first packet of send (if not already) * PSH, and FIN flags should be set as appropriate for LAST packet of send * TCP Checksum should be set to the partial pseudo-header sum as follows (there is a more detailed discussion of this is Section 26.2.4.6): Table 26-41. TCP Partial Pseudo-Header Sum for IPv4 IP Source Address IP Destination Address Zero Layer 4 Protocol ID Zero Table 26-42. TCP Partial Pseudo-Header Sum for IPv6 IPv6 Source Address IPv6 Final Destination Address Zero Zero Next Header UDP Header * Checksum should be set as in TCP header, above The following sections describe the updating process performed by the hardware for each frame sent using the TCP Segmentation capability. Intel(R) Communications Chipset 89xx Series - Datasheet 1168 October 2012 Order Number: 327879-001US 26.0 26.2.4.6 Transmit Checksum Offloading with TCP/UD Segmentation The Controller supports checksum off-loading as a component of the TCP Segmentation off-load feature and as a standalone capability. Section 26.2.5 describes the interface for controlling the checksum off-loading feature. This section describes the feature as it relates to TCP Segmentation. The Controller supports IP and TCP header options in the checksum computation for packets that are derived from the TCP Segmentation feature. Note: The Controller is capable of computing one level of IP header checksum and one TCP/UDP header and payload checksum. In case of multiple IP headers, the driver needs to compute all but one IP header checksum. The Controller calculates check sums on the fly on a frame-by-frame basis and inserts the result in the IP/TCP/UDP headers of each frame. TCP and UDP checksum are a result of performing the checksum on all bytes of the payload and the pseudo header. Three specific types of checksum are supported by the hardware in the context of the TCP Segmentation off-load feature: * IPv4 checksum * TCP checksum Each packet that is sent via the TCP segmentation off-load feature optionally includes the IPv4 checksum and either the TCP checksum. All checksum calculations use a 16-bit wide one's complement checksum. The checksum word is calculated on the outgoing data. Table 26-43. Supported Transmit Checksum Capabilities Packet Type Hardware IP Checksum Calculation Hardware TCP/UDP Checksum Calculation IP v4 packets Yes Yes IP v6 packets (no I checksum in Ipv6) NA Yes Packet is greater than 1518, 1522 or 1526 bytes (LPE=1b)a Yes Yes Packet has 802.3ac tag Yes Yes Packet has IP options (IP header is longer than 20 bytes) Yes Yes Packet has TCP or UDP options Yes Yes IP header's protocol field contains a protocol # other than TCP or UDP. Yes No a. Depends on number of VLAN tags. The table below summarizes the conditions of when checksum off loading can/should be calculated. Table 26-44. Conditions for Checksum Off Loading Packet Type Non TSO TSO IPv4 TCP/UDP Reason IP Raw packet (non TCP/UDP protocol) Yes No Yes Yes TCP segment or UDP datagram with checksum off-load No No Non-IP packet or checksum not offloaded Yes Yes For TSO, checksum off-load must be done October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1169 26.2.4.7 TCP/UDP/IP Header Update IP/TCP or IP/UDP header is updated for each outgoing frame based on the IP/TCP header prototype which hardware DMA's from the first descriptor(s). The checksum fields and other header information are later updated on a frame-by-frame basis. The updating process is performed concurrently with the packet data fetch. The following sections define what fields are modified by hardware during the TCP Segmentation process by the Controller. Note: Software must make PAYLEN and HDRLEN value of Context descriptors correct. Otherwise, the failure of Large Send due to either under-run or over-run might cause hardware to send bad packets or even cause TX hardware to hang. The indication of Large Send failure can be checked in the TSCTFC statistic register. 26.2.4.7.1 TCP/UDP/IP Header for the First Frames The hardware makes the following changes to the headers of the first packet that is derived from each TCP segmentation context. MAC Header (for SNAP) * Type/Len field = MSS + MACLEN + IPLEN + L4LEN - 14 -4 (if VLAN added by Software) Ipv4 Header * IP Identification: Value in the IPv4 header of the prototype header in the packet data itself * IP Total Length = MSS + L4LEN + IPLEN * IP Checksum Ipv6 Header * Payload Length = MSS + L4LEN + IPV6_HDR_extension1 TCP Header * Sequence Number: The value is the Sequence Number of the first TCP byte in this frame. * The flag values of the first frame are set by ANDing the flag word in the pseudo header with the DTXTCPFLGL.TCP_flg_first_seg register field. The default value of the DTXTCPFLGL.TCP_flg_first_seg are set so that the FIN flag and the PSH flag are cleared in the first frame. * TCP Checksum UDP Header * UDP Length = MSS + L4LEN * UDP Checksum 26.2.4.7.2 TCP/IP Header for the Subsequent Frames The hardware makes the following changes to the headers for subsequent packets that are derived as part of a TCP segmentation context: Number of bytes left for transmission = PAYLEN - (N * MSS). Where N is the number of frames that have been transmitted. 1. IPV6_HDR_extension is calculated as IPLEN - 40 bytes. Intel(R) Communications Chipset 89xx Series - Datasheet 1170 October 2012 Order Number: 327879-001US 26.0 MAC Header (for SNAP Packets) Type/Len field = MSS + MACLEN + IPLEN + L4LEN - 14 - 4 (if VLAN added by Software) Ipv4 Header * IP Identification: incremented from last value (wrap around based on 16 bit-width)) * IP Total Length = MSS + L4LEN + IPLEN * IP Checksum Ipv6 Header * Payload Length = MSS + L4LEN + IPV6_HDR_extension1 TCP Header * Sequence Number update: Add previous TCP payload size to the previous sequence number value. This is equivalent to adding the MSS to the previous sequence number. * The flag values of the subsequent frames are set by ANDing the flag word in the pseudo header with the DTXTCPFLGL.TCP_Flg_mid_seg register field. The default value of the DTXTCPFLGL.TCP_Flg_mid_seg are set so that if the FIN flag and the PSH flag are cleared in these frames. * TCP Checksum TCP Header * UDP Length = MSS + L4LEN * UDP Checksum 26.2.4.7.3 TCP/IP Header for the Last Frame The hardware makes the following changes to the headers for the last frame of a TCP segmentation context: Last frame payload bytes = PAYLEN - (N * MSS) MAC Header (for SNAP Packets) * Type/Len field = Last frame payload bytes + MACLEN + IPLEN + L4LEN - 14 - 4 (if VLAN added by Software) Ipv4 Header * IP Total Length = last frame payload bytes + L4LEN + IPLEN * IP Identification: incremented from last value (wrap around based on 16 bit-width) * IP Checksum Ipv6 Header * Payload Length = last frame payload bytes + L4LEN + IPV6_HDR_extension1 TCP Header * Sequence Number update: Add previous TCP payload size to the previous sequence number value. This is equivalent to adding the MSS to the previous sequence number. * The flag values of the last frames are set by ANDing the flag word in the pseudo header with the DTXTCPFLGH.TCP_Flg_lst_seg register field. The default value of 1. IPV6_HDR_extension is calculated as IPLEN - 40 bytes. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1171 the DTXTCPFLGH.TCP_Flg_lst_seg are set so that if the FIN flag and the PSH flag are set in the last frame. * TCP Checksum TCP Header * UDP Length = Last frame payload bytes + L4LEN * UDP Checksum 26.2.4.8 IP/TCP/UDP Checksum Offloading The Controller performs checksum off loading as part of the TCP segmentation off-load feature. These specific checksum are supported under TCP segmentation: * IPv4 checksum * TCP checksum See Section 26.2.5 for description of checksum off loading of a single-send packet. 26.2.4.9 Data Flow The flow used by the Controller to do TCP segmentation is as follows: 1. Get a descriptor with a request for a TSO off-load of a TCP packet. 2. First Segment processing: a. Fetch all the buffers containing the header as calculated by the MACLEN, IPLEN and L4LEN fields. Save the addresses and lengths of the buffers containing the header (up to 4 buffers). The header content is not saved. b. Fetch data up to the MSS from subsequent buffers & calculate the adequate checksum(s). c. Update the Header accordingly and update internal state of the packet (next data to fetch and TCP SN). d. Send the packet to the network. e. If total packet was sent, go to step 4. else continue. 3. Next segments a. Wait for next arbitration of this queue. b. Fetch all the buffers containing the header from the saved addresses. Subsequent reads of the header might be done with a no snoop attribute. c. Fetch data up to the MSS or end of packet from subsequent buffers & calculate the adequate checksum(s). d. Update the Header accordingly and update internal state of the packet (next data to fetch and TCP SN). e. If total packet was sent, request is done, else restart from step 3. 4. Release all buffers (update head pointer). Note: Descriptors are fetched in a parallel process according to the consumption of the buffers. Intel(R) Communications Chipset 89xx Series - Datasheet 1172 October 2012 Order Number: 327879-001US 26.0 26.2.5 Checksum Offloading in Non-Segmentation Mode The previous section on TCP Segmentation off-load describes the IP/TCP/UDP checksum off loading mechanism used in conjunction with TCP Segmentation. The same underlying mechanism can also be applied as a standalone feature. The main difference in normal packet mode (non-TCP Segmentation) is that only the checksum fields in the IP/TCP/UDP headers need to be updated. Before taking advantage of the Controller's enhanced checksum off-load capability, a checksum context must be initialized. For the normal transmit checksum off-load feature this is performed by providing the device with a Descriptor with TSE=0b in the TDESD.DCMD field and setting either the TXSM or IXSM bits in the TDESD.POPTS field. Setting TSE=0b indicates that the normal checksum context is being set, as opposed to the segmentation context. For additional details on contexts, see Section 26.2.2.4. Note: Enabling the checksum off loading capability without first initializing the appropriate checksum context leads to unpredictable results. CRC appending (TDESC.CMD.IFCS) must be enabled in TCP/IP checksum mode, since CRC must be inserted by hardware after the checksum has been calculated. As mentioned in Section 26.2.2, it is not necessary to set a new context for each new packet. In many cases, the same checksum context can be used for a majority of the packet stream. In this case, some performance can be gained by only changing the context on an as needed basis or electing to use the off-load feature only for a particular traffic type, thereby avoiding the need to read all context descriptors except for the initial one. Each checksum operates independently. Insertion of the IP and TCP checksum for each packet are enabled through the Transmit Data Descriptor POPTS.TSXM and POPTS.IXSM fields, respectively. 26.2.5.1 IP Checksum Three fields in the Transmit Context Descriptor set the context of the IP checksum off loading feature: * TUCMD.IPv4 * IPLEN * MACLEN TUCMD.IPv4=1b specifies that the packet type for this context is IPv4, and that the IP header checksum should be inserted. TUCMD.IPv4=0b indicates that the packet type is IPv6 (or some other protocol) and that the IP header checksum should not be inserted. MACLEN specifies the byte offset from the start of the DMA'd data to the first byte to be included in the checksum, the start of the IP header. The minimal allowed value for this field is 12. The maximum value for this field is 127. This is adequate for typical applications. Note: The MACLEN+IPLEN value needs to be less than the total DMA length for a packet. If this is not the case, the results are unpredictable. IPLEN specifies the IP header length. Maximum allowed value for this field is 511 Bytes. MACLEN+IPLEN specify where the IP checksum should stop. This is limited to the first 127+511 bytes of the packet and must be less than or equal to the total length of a given packet. If this is not the case, the result is unpredictable. The 16-bit IPv4 Header Checksum is placed at the two bytes starting at MACLEN+10. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1173 As mentioned in Section 26.2.2.2, Transmit Contexts, it is not necessary to set a new context for each new packet. In many cases, the same checksum context can be used for a majority of the packet stream. In this case, some performance can be gained by only changing the context on an as needed basis or electing to use the off-load feature only for a particular traffic type, thereby avoiding all context descriptor reads except for the initial one. 26.2.5.2 TCP Checksum Three fields in the Transmit Context Descriptor set the context of the TCP checksum off loading feature: * MACLEN * IPLEN * TUCMD.L4T TUCMD.L4T=01b specifies that the packet type is TCP, and that the 16-bit TCP header checksum should be inserted at byte offset MACLEN+IPLEN+16. TUCMD.L4T=00b indicates that the packet is UDP and that the 16-bit checksum should be inserted starting at byte offset MACLEN+IPLEN+6. IPLEN+MACLEN specifies the byte offset from the start of the DMA'd data to the first byte to be included in the checksum, the start of the TCP header. The minimal allowed value for this sum is 32/42 for UDP or TCP respectively. Note: The IPLEN+MACLEN+L4LEN value needs to be less than the total DMA length for a packet. If this is not the case, the results are unpredictable. The TCP/UDP checksum always continues to the last byte of the DMA data. Note: For non-TSO, software still needs to calculate a full checksum for the TCP/UDP pseudo-header. This checksum of the pseudo-header should be placed in the packet data buffer at the appropriate offset for the checksum calculation. 26.2.5.3 SCTP CRC Offloading For SCTP packets, a CRC32 checksum offload is provided. Three fields in the Transmit Context Descriptor set the context of the STCP checksum off loading feature: * MACLEN * IPLEN * TUCMD.L4T TUCMD.L4T=10b specifies that the packet type is SCTP, and that the 32-bit STCP CRC should be inserted at byte offset MACLEN+IPLEN+8. IPLEN+MACLEN specifies the byte offset from the start of the DMA'd data to the first byte to be included in the checksum, the start of the STCP header. The minimal allowed value for this sum is 26. The SCTP CRC calculation always continues to the last byte of the DMA data. The SCTP total L3 payload size (PAYLEN - IPLEN - MACLEN) should be a multiple of 4 bytes (SCTP padding not supported). 1. TSO is not available for SCTP packets. Intel(R) Communications Chipset 89xx Series - Datasheet 1174 October 2012 Order Number: 327879-001US 26.0 2. The CRC field of the SCTP header must be set to zero prior to requesting a CRC calculation offload. 26.2.5.4 Checksum Supported Per Packet Types The following table summarizes which checksum is supported per packet type. Note: TSO is not supported for packet types for which IP checksum & TCP checksum can not be calculated. Table 26-45. Checksum Per Packet Type Hardware IP Checksum Calculation Packet Type Hardware TCP/UDP/SCTP Checksum Calculation Ipv4 packets Yes Yes Ipv6 packets No (n/a) Yes Ipv6 packet with next header options: * Hop-by-Hop options * Destinations options * Routing (w len 0b) * Routing (w len >0b) * Fragment * Home option No No No No No No Yes Yes Yes No No No Ipv4 tunnels: * Ipv4 packet in an Ipv4 tunnel * Ipv6 packet in an Ipv4 tunnel Either IP or TCP/SCTP1 Either IP or TCP/SCTP1 Either IP or TCP/SCTP a Either IP or TCP/SCTP1 Ipv6 tunnels: * Ipv4 packet in an Ipv6 tunnel * Ipv6 packet in an Ipv6 tunnel No No Yes Yes Packet is an Ipv4 fragment Yes No (n/a) (n/a) (n/a) (n/a) (n/a) (n/a) Packet is greater than 1518, 1522 or 1526 bytes; (LPE=1b) Yes Yes Packet has 802.3ac tag Yes Yes Packet has TCP or UDP options Yes Yes IP header's protocol field contains protocol # other than TCP or UDP. Yes No a. For the tunneled case, the driver might do only the TCP checksum or Ipv4 checksum. If TCP checksum is desired, the driver should define the IP header length as the combined length of both IP headers in the packet. If an IPv4 checksum is required, the IP header length should be set to the Ipv4 header length. 26.2.6 Multiple Transmit Queues The number of transmit queues is 16, to match the expected number of CPU cores on server processors and to support virtualization mode. If there are more CPUs cores than queues, then one queue might be used to service more than one CPU. For transmission process, each thread might place a queue in the host memory of the CPU it is tied to. Th PCH supports assigning either high or low priority to each transmit queue. High priority is assigned to by setting the TXDCTL[n].priority bit to 1. When high priority is assigned to a specific transmit queue, the PCH will always prioritize DMA accesses such as transmit descriptor read, transmit descriptor writeback and transmit data fetch, before servicing lower priority transmit queues on a specific Physical Function. Note: Throughput of low priority transmit queues can be significantly impacted if high priority queues utilize the DMA resources fully. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1175 26.3 Interrupts 26.3.1 Mapping of Interrupt Causes The Controller supports the following interrupt modes: * PCI legacy interrupts or MSI - selected when GPIE.Multiple_MSIX is 0b * MSI-X when GPIE.Multiple_MSIX is 1b. Note: If only one MSI-X vector is allocated by the operating system, then the driver might use the non MSI-X mapping method even in MSI-X mode. Mapping of interrupts causes is different in each of the above modes and is described below. 26.3.1.1 Legacy and MSI Interrupt Modes In legacy and MSI modes, an interrupt cause is reflected by setting a bit in the EICR register. This section describes the mapping of interrupt causes (a specific Rx queue event or a Link Status Change event) to bits in the EICR. Mapping of queue-related causes is accomplished through the IVAR register. Each possible queue interrupt cause (each Rx or Tx queue) is allocated an entry in the IVAR, and each entry in the IVAR identifies one bit in the EICR register among the bits allocated to queue interrupt causes. It is possible to map multiple interrupt causes into the same EICR bit. In this mode, different queue related interrupt causes can be mapped to the first 16 bits of the EICR register. The following configuration and parameters are involved: * The IVAR[3:0] entries map 8 Tx queues and 16 Rx queues into the EICR[7:0] bits. * The IVAR_MISC that maps non-queue causes is not used. * The EICR[30] bit is allocated to the TCP timer interrupt cause. * The EICR[31] bit is allocated to the other interrupt causes summarized in the ICR register. * A single interrupt vector is provided. Intel(R) Communications Chipset 89xx Series - Datasheet 1176 October 2012 Order Number: 327879-001US 26.0 Figure 26-20.Cause Mapping in Legacy Mode 0 Cause 15 EICR (reflect causes) Queue Related causes 0 IVAR . . . Cause 0 3 IVAR_Misc 30 31 Other causes Single Vector TCP timer Other The table below maps the different interrupt causes into the IVAR registers. Table 26-46. Cause Allocation in the IVAR Registers - MSI and Legacy Mode Interrupt Entry Description Rx_i i*2 (i= 0...7) Receive queues i - Associates an interrupt occurring in the Rx queues i with a corresponding bit in the EICR register. Tx_i i*2+1 (i= 0...7) Transmit queues i- Associates an interrupt occurring in the Tx queues I with a corresponding bit in the EICR register. 26.3.1.2 MSI-X Mode - VMDq Mode In a VMDq setup, the Controller can request up to 25 Vectors. In MSI-X mode, an interrupt cause is mapped into an MSI-X vector. This section describes the mapping of interrupt causes (a specific Rx queue event or other events) to MSI-X vectors. Mapping is accomplished through the IVAR register. Each possible cause for an interrupt is allocated an entry in the IVAR, and each entry in the IVAR identifies one MSI-X vector. It is possible to map multiple interrupt causes into the MSI-X vector. The EICR also reflects interrupt vectors. The EICR bits allocated for queue causes reflect the MSI-X vector (bit 2 is set when MSI-X vector 2 is used). The following configuration and parameters are involved: * The IVAR[3:0] registers map 16 Tx queues, 16 Rx queues, events to up to 8 interrupt vectors. * The IVAR_MISC register maps a TCP timer and other events to 2 MSI-X vectors. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1177 Figure 26-21.Cause Mapping in MSI-X Mode 0 IVAR Interrupt causes (queues) MSI-X Vector 3 IVAR_Misc MSI-X Vector Interrupt causes (Other) 0 EICR 24 30 31 TCP timer Other Table 26-47 below defines which interrupt cause is represented by each entry in the MSI-X Allocation registers. The software has access to 18 mapping entries to map each cause to one of the 25 MSI-x vectors. Table 26-47. Cause Allocation in the IVAR Registers Interrupt Entry Description Rx_i i*2 (i= 0...7) Receive queues i - Associates an interrupt occurring in the Rx queues i with a corresponding entry in the MSI-X Allocation registers. Tx_i i*2+1 (i= 0...7) Transmit queues i- Associates an interrupt occurring in the Tx queues I with a corresponding entry in the MSI-X Allocation registers. TCP timer 16 TCP Timer - Associates an interrupt issued by the TCP timer with a corresponding entry in the MSI-X Allocation registers Other cause 17 Other causes - Associates an interrupt issued by the "other causes" with a corresponding entry in the MSI-X Allocation registers 26.3.2 Legacy Interrupt Registers The interrupt logic consists of the registers listed in the tables below, plus the registers associated with MSI/MSI-X signaling. The first table describes the use of the registers in legacy mode and the second one the use of the register when using the extended interrupts functionality. Table 26-48. Interrupt Registers - Legacy Mode Register Acronym Function Interrupt Cause ICR Records interrupt conditions. Interrupt Cause Set ICS Allows software to set bits in the ICR. Intel(R) Communications Chipset 89xx Series - Datasheet 1178 October 2012 Order Number: 327879-001US 26.0 Table 26-48. Interrupt Registers - Legacy Mode Interrupt Mask Set/Read IMS Interrupt Mask Clear IMC Clears bits in the interrupt mask. IAM Under some conditions, the content of this register is copied to the mask register following read or write of ICR. Interrupt Acknowledge auto-mask Sets or reads bits in the interrupt mask. Table 26-49. Interrupt Registers - Extended Mode Register Acronym Function Extended Interrupt Cause EICR Records interrupt causes from receive and transmit queues. An interrupt is signaled when unmasked bits in this register are set. Extended Interrupt Cause Set EICS Allows software to set bits in the Interrupt Cause register. Extended Interrupt Mask Set/Read EIMS Sets or read bits in the interrupt mask. Extended Interrupt Mask Clear EIMC Clears bits in the interrupt mask. EIAC Allows bits in the EICR to be cleared automatically following an MSI-X interrupt without a read or write of the EICR. Extended Interrupt Acknowledge auto-mask EIAM This register is used to decide which masks are cleared in the extended mask register following read or write of EICR or which masks are set following a write to EICS. In MSI-X mode, this register also controls which bits in EIMC are cleared automatically following an MSI-X interrupt. Interrupt Cause ICR Records interrupt conditions for special conditions - a single interrupt from all the conditions of ICR is reflected in the "other" field of the EICR. Extended Interrupt Auto Clear Interrupt Cause Set ICS Allows software to set bits in the ICR. Interrupt Mask Set/Read IMS Sets or reads bits in the other interrupt mask. Interrupt Mask Clear IMC Clears bits in the Other interrupt mask. Interrupt Acknowledge auto-mask IAM Under some conditions, the content of this register is copied to the mask register following read or write of ICR. General Purpose Interrupt Enable GPIE Controls different behaviors of the interrupt mechanism. 26.3.2.1 Interrupt Cause Register (ICR) 26.3.2.1.1 Legacy Mode In Legacy mode, ICR is used as the sole interrupt cause register. Upon reception of an interrupt, the interrupt handling routine can read this register in order to find out what are the causes of this interrupt. 26.3.2.1.2 Advanced Mode In advanced mode, this register captures the interrupt causes not directly captured by the EICR. These are infrequent management interrupts and error conditions. When EICR is used in advanced mode, the RX /TX related bits in ICR should be masked. ICR bits are cleared on register read. If GPIE.NSICR = 0b, then the clear on read occurs only if no bit is set in the IMS or at least one bit is set in the IMS and there is a true interrupt as reflected in ICR.INTA. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1179 26.3.2.2 Interrupt Cause Set Register (ICS) This registers allows setting the bits of ICR by software, by writing a 1b in the corresponding bits in ICS. Used usually to rearm interrupts the software didn't have time to handle in the current interrupt routine. 26.3.2.3 Interrupt Mask Set/Read Register (IMS) An interrupt is enabled if its corresponding mask bit in this register is set to 1b, and disabled if its corresponding mask bit is set to 0b. A PCIe interrupt is generated whenever one of the bits in this register is set, and the corresponding interrupt condition occurs. The occurrence of an interrupt condition is reflected by having a bit set in the Interrupt Cause Register. Reading this register returns which bits have an interrupt mask set. A particular interrupt might be enabled by writing a 1b to the corresponding mask bit in this register. Any bits written with a 0b are unchanged. Thus, if software desires to disable a particular interrupt condition that had been previously enabled, it must write to the Interrupt Mask Clear Register (see below), rather than writing a 0b to a bit in this register. 26.3.2.4 Interrupt Mask Clear Register (IMC) Software blocks interrupts by clearing the corresponding mask bit. This is accomplished by writing a 1b to the corresponding bit in this register. Bits written with 0b are unchanged (their mask status does not change). 26.3.2.5 Interrupt Acknowledge Auto-Mask Register (IAM) An ICR read or write has the side effect of writing the contents of this register to the mask register. If GPIE.NSICR = 0b, then the copy of this register to the mask register occurs only if at least one bit is set in the mask register and there is a true interrupt as reflected in ICR.INTA. 26.3.2.6 Extended Interrupt Cause Registers (EICR) 26.3.2.6.1 MSI/INT-A Mode (GPIE.Multiple_MSIX = 0) This register records the interrupts causes to provide to the software information on the interrupt source. The interrupt causes include: 1. The Receive and Transmit queues -- Each queue (either Tx or Rx) can be mapped to one of the 16 interrupt causes bits (RxTxQ) available in this register according to the mapping in the IVAR registers 2. Indication for the TCP timer interrupt. -- Writing a 1b clears the corresponding bit in this register. Reading this register auto-clears all bits. 26.3.2.6.2 MSI-X Mode (GPIE.Multiple_MSIX = 1) This register records the interrupt vectors currently emitted. In this mode only the first 25 bits are valid. For all the subsequent registers, in MSI-X mode, each bit controls the behavior of one vector. Intel(R) Communications Chipset 89xx Series - Datasheet 1180 October 2012 Order Number: 327879-001US 26.0 Bits in this register can be configured to auto-clear when the MSI-X interrupt message is sent, in order to minimize driver overhead when using MSI-X interrupt signaling. Writing a 1b clears the corresponding bit in this register. Reading this register does not clear any bits. 26.3.2.7 Extended Interrupt Cause Set Register (EICS) This registers allows to set the bits of EICR by software, by writing a 1b in the corresponding bits in EICS. Used usually to rearm interrupts that the software didn't have time to handle in the current interrupt routine. 26.3.2.8 Extended Interrupt Mask Set and Read Register (EIMS) & Extended Interrupt Mask Clear Register (EIMC) Interrupts appear on PCIe only if the interrupt cause bit is a one and the corresponding interrupt mask bit is a one. Software blocks assertion of an interrupt by clearing the corresponding bit in the mask register. The cause bit stores the interrupt event regardless of the state of the mask bit. Different Clear (EIMC) and set (EIMS) registers make this register more "thread safe" by avoiding a read-modify-write operation on the mask register. The mask bit is set for each bit written to a one in the set register (EIMS) and cleared for each bit written in the clear register (EIMC). Reading the set register (EIMS) returns the current mask register value. 26.3.2.9 Extended Interrupt Auto Clear Enable Register (EIAC) Each bit in this register enables clearing of the corresponding bit in EICR following interrupt generation. When a bit is set, the corresponding bit in EICR are automatically cleared following an interrupt. This feature should only be used in MSI-X mode. When used in conjunction with MSI-X interrupt vector, this feature allows interrupt cause recognition, and selective interrupt cause, without requiring software to read or write the EICR register; therefore, the penalty related to a PCIe read or write transaction is avoided. See section 26.3.5 for additional information on the interrupt cause reset process. 26.3.2.10 Extended Interrupt Auto Mask Enable Register (EIAM) Each bit set in this register enables clearing of the corresponding bit in the extended mask register following read or write-to-clear to EICR. It also enables setting of the corresponding bit in the extended mask register following a write-to-set to EICS. This mode is provided in case MSI-X is not used, and therefore auto-clear through EIAC register is not available. In MSI-X mode, the driver software might set the bits of this register to select mask bits that must be reset during interrupt processing. In this mode, each bit in this register enables clearing of the corresponding bit in EIMC following interrupt generation. 26.3.2.11 GPIE Register There are a few bits in the GPIE register that define the behavior of the interrupt mechanism. The setting of these bits is different in each mode of operation. The following table describes the recommended setting of these bits in the different modes: October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1181 Table 26-50. Settings for Different Interrupt Modes Field Multiple_ MSIX EIAME PBA_ support 26.3.3 Bit(s) 4 30 31 Initial Value Description INT-x/ MSI + Legacy INT-x/ MSI + Extend MSI-X Multi vector 0b Multiple_MSIX - multiple vectors: 0b = non-MSIX or MSI-X with 1 vector IVAR maps Rx/Tx causes to 16 EICR bits, but MSIX[0] is asserted for all. 1b = MSIX mode, IVAR maps Rx/Tx causes to 25 EICR bits. When set, the EICR register is not clear on read. 0b 0b 1b 0b 0b EIAME: When set, upon firing of an MSI-X message, mask bits set in EIAM associated with this message are cleared. Otherwise, EIAM is used only upon read or write of EICR/EICS registers. 0b 0b 1b 1b 0b PBA support: When set, setting one of the extended interrupts masks via EIMS causes the PBA bit of the associated MSI-X vector to be cleared. Otherwise, the Controller behaves in a way that supports legacy INT-x interrupts. Should be cleared when working in INT-x or MSI mode and set in MSI-X mode. 0b 0b 1b 1b MSI-X Single vector MSI-X and Vectors MSI-X defines a separate optional extension to basic MSI functionality. Compared to MSI, MSI-X supports a larger maximum number of vectors per function, the ability for software to control aliasing when fewer vectors are allocated than requested, plus the ability for each vector to use an independent address and data value, specified by a table that resides in Memory Space. However, most of the other characteristics of MSI-X are identical to those of MSI. For more information on MSI-X, see the PCI Local Bus Specification, Revision 3.0. MSI-X maps each of the the Controller interrupt causes into an interrupt vector that is conveyed by the Controller as a posted-write PCIe transaction. Mapping of an interrupt cause into an MSI-X vector is determined by system software (a device driver) through a translation table stored in the MSI-X Allocation registers. Each entry of the allocation registers defines the vector for a single interrupt cause. There are 18 extended interrupt causes that exist in the Controller: 1. traffic causes -- 16 Tx, 16 Rx. 2. TCP timer 26.3.4 Interrupt Moderation An interrupt is generated upon receiving of incoming packets, as throttled by the EITR registers (see Section 28.7.1.22). There is an EITR register per MSI-X vector. In MSI-X mode, each active bit in EICR can trigger the interrupt vector it is allocated to. Following the allocation, the EITR corresponding to the MSI-X vector is tied to one or more bits in EICR. When MSI-X is not activated, the interrupt moderation is controlled by register EITR[0]. Intel(R) Communications Chipset 89xx Series - Datasheet 1182 October 2012 Order Number: 327879-001US 26.0 Software can use EITR to limit the rate of delivery of interrupts to the host CPU. This register provides a guaranteed inter-interrupt delay between interrupts asserted by the network controller, regardless of network traffic conditions. The following formula converts the inter-interrupt interval value to the common 'interrupts/sec.' performance metric: interrupts/sec = (2 * 10-6sec x interval)-1 Note: In the Controller the interval granularity is 2 sec so some of the LSB bits of the interval are used for the low latency interrupt moderation. For example, if the interval is programmed to 125d, the network controller guarantees the CPU is not interrupted by the network controller for at least 250 s from the last interrupt. In this case, the maximum observable interrupt rate from the adapter should not exceed 4000 interrupts/sec. Inversely, inter-interrupt interval value can be calculated as: inter-interrupt interval = (2 * 10-6 sec x interrupt/sec)-1 The optimal performance setting for this register is system and configuration specific. The Extended Interrupt Throttle Register should default to zero upon initialization and reset. It loads in the value programmed by the software after software initializes the device. When software wants to force an immediate interrupt, for example after setting a bit in the EICR with the Extended Interrupt Cause Set register, a value of 0 can be written to the Counter to generate an interrupt immediately. This write should include re-writing the Interval field with the desired constant, as it is used to reload the Counter immediately for the next throttling interval. The Controller implements interrupt moderation to reduce the number of interrupts software processes. The moderation scheme is based on the EITR (Interrupt Throttle Register). Whenever an interrupt event happens, the corresponding bit in the EICR is activated. However, an interrupt message is not sent out on the PCIe interface until the EITR counter assigned to that EICR bit has counted down to zero. As soon as the interrupt is issued, the EITR counter is reloaded with its initial value and the process repeats again. The interrupt flow should follow the diagram below: October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1183 Figure 26-22.Interrupt Throttle Flow Diagram Load counter with interval Start count down No Counter = 0 ? Yes Yes No Interrupt active ? Counter written to 0 Yes Assert Interrupt For cases where the Controller is connected to a small number of clients, it is desirable to fire off the interrupt as soon as possible with minimum latency. For these cases, when the EITR counter counts down to zero and no interrupt event has happened, then the EITR counter is not re-armed but stays at zero. Thus, the next interrupt event triggers an interrupt immediately. That scenario is illustrated as "Case B" below. Figure 26-23.Case A: Heavy Load, Interrupts Moderated EICR clear Intr EICR clear Intr ITR delay Pkt Pkt Pkt Intel(R) Communications Chipset 89xx Series - Datasheet 1184 EICR clear Intr ITR delay Pkt Pkt Pkt Pkt Pkt October 2012 Order Number: 327879-001US 26.0 Figure 26-24.Light Load, Interrupts Immediately on Packet Receive Intr EICR clear Intr ITR delay Pkt 26.3.5 Pkt Clearing Interrupt Causes The Controller has three methods available for to clear EICR bits: Autoclear, clear-on-write, and clear-on-read. ICR bits might only be cleared with clear-on-write or clear-on-read. 26.3.5.1 Auto-Clear In systems that support MSI-X, the interrupt vector allows the interrupt service routine to know the interrupt cause without reading the EICR. With interrupt moderation active, software load from spurious interrupts is minimized. In this case, the software overhead of a I/O read or write can be avoided by setting appropriate EICR bits to autoclear mode by setting the corresponding bits in the Extended Interrupt Auto-clear Enable Register (EIAC). When auto-clear is enabled for an interrupt cause, the EICR bit is set when a cause event mapped to this vector occurs. When the EITR Counter reaches zero, the MSI-X message is sent on PCIe. Then the EICR bit is cleared and enabled to be set by a new cause event. The vector in the MSI-X message signals software the cause of the interrupt to be serviced. It is possible that in the time after the EICR bit is cleared and the interrupt service routine services the cause, for example checking the transmit and receive queues, that another cause event occurs that is then serviced by this ISR call, yet the EICR bit remains set. This results in a "spurious interrupt". Software can detect this case, for example if there are no entries that require service in the transmit and receive queues, and exit knowing that the interrupt has been automatically cleared. The use of interrupt moderations through the EITR register limits the extra software overhead that can be caused by these spurious interrupts. 26.3.5.2 Write to Clear In the case where the driver wishes to configure itself in MSI-X mode to not use the "auto-clear" feature, it might clear the EICR bits by writing to the EICR register. Any bits written with a 1b is cleared. Any bits written with a 0b remain unchanged. 26.3.5.3 Read to Clear The EICR and ICR registers are cleared on a read. The driver should never do a read-to-clear of the EICR when in MSI-X mode, since this might clear interrupt cause events which are processed by a different interrupt handler (assuming multiple vectors). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1185 26.3.6 Rate Controlled Low Latency Interrupts (LLI) There are some types of network traffic for which latency is a critical issue. For these types of traffic, interrupt moderation hurts performance by increasing latency between the time a packet is received by hardware and the time it is handled to the host operating system. This traffic can be identified by the 2-tuple value, in conjunction with Control Bits and specific size. In addition packets with specific ethernet types, TCP flag or specific VLAN priority might generate an immediate interrupt. Low latency interrupts shares the filters used by the queueing mechanism described in Section 26.1.1. Each of these filters, in addition to the queueing action might also indicate matching packets might generate immediate interrupt. If a received packet matches one of these filters, hardware should interrupt immediately, overriding the interrupt moderation by the EITR counter. Each time a Low Latency Interrupt is fired, the EITR interval is loaded and down-counting starts again. The logic of the low latency interrupt mechanism is as follows: * There are 8 2-tuple filters. The content of each filter is described in Section 26.1.1.5. The immediate interrupt action of each filter can be enabled or disabled. If one of the filters detects an adequate packet, an immediate interrupt is issued. * There are 8 flex filters. The content of each filter is described in Section 26.1.1.6. The immediate interrupt action of each filter can be enabled or disabled. If one of the filters detects an adequate packet, an immediate interrupt is issued. * When VLAN priority filtering is enabled, VLAN packets must trigger an immediate interrupt when the VLAN Priority is equal to or above the VLAN priority threshold. This is regardless of the status of the 2-tuple or Flex filters. * The SYN packets filter defined in Section 26.1.1.7 and the ethernet type filters defined in section Section 26.1.1.4 might also be used to indicate low latency interrupt conditions. Note: EITR is re-armed to the `EITR.Moderation Counter' value following a low latency interrupt. Note: Immediate interrupts are available only when using advanced receive descriptors and not for legacy descriptors. Note: Packets that are dropped or have errors do not cause a Low Latency Interrupt. 26.3.6.1 Rate Control Mechanism In a network with lots of latency sensitive traffics the Low Latency Interrupt can eliminate the Interrupt throttling capability by flooding the Host with too many interrupts (more than the Host can handle). In order to mitigate the above, the Controller supports a credit base mechanism to control the rate of the Low Latency Interrupts. Rules: * The default value of each counter is 0b (no moderation). This also preserves backward compatibility. * The counter increments at a configurable rate, and saturates at the maximum value (31d). Intel(R) Communications Chipset 89xx Series - Datasheet 1186 October 2012 Order Number: 327879-001US 26.0 -- The configurable rate granularity is 4 s (250K interrupt/sec. down to 250K/32 ~ 8K interrupts per sec.). * A LLI might be issued as long as the counter value is strictly positive (> zero). -- The credit counter allows bursts of low latency interrupts but the interrupt average are not more than the configured rate. * Each time a Low Latency Interrupt is fired the credit counter decrements by one. * Once the counter reaches zero, a low latency interrupt cannot be fired -- Must wait for the next ITR expired or for the next incrementing of this counter (if the EITR expired happened first the counter does not decrement). The EITR and GPIE registers manage rate control of LLI: * The LL Interval field in the GPIE register controls the rate of credits * The 5-bit LL Counter field in the EITR register contains the credits 26.3.7 TCP Timer Interrupt 26.3.7.1 Introduction In order to implement TCP timers for IOAT, software needs to take action periodically (every 10 milliseconds). Today, the driver must rely on software-based timers, whose granularity can change from platform to platform. This software timer generates a software NIC interrupt, which then allows the driver to perform timer functions as part of its usual DPC, avoiding cache thrash and enabling parallelization. The timer interval is system-specific. It would be more accurate and more efficient for this periodic timer to be implemented in hardware. The driver would program a timeout value (usual value of 10 ms), and each time the timer expires, hardware sets a specific bit in the EICR. When an interrupt occurs (due to normal interrupt moderation schemes), software reads the EICR and discovers that it needs to process timer events during that DPC. The timeout should be programmable by the driver, and the driver should be able to disable the timer interrupt if it is not needed. 26.3.7.2 Description A stand-alone down-counter is implemented. An interrupt is issued each time the value of the counter is zero. The software is responsible for setting initial value for the timer in the TCPTIMER.Duration field. Kick-starting is done by writing a 1b to the TCPTIMER.KickStart bit. Following the kick-start, an internal counter is set to the value defined by the TCPTIMER.Duration field. Then during the count operation, the counter is decreased by one each millisecond. When the counter reaches zero, an interrupt is issued (see EICR register Section 28.7.1.1). The counter re-starts counting from its initial value if the TCPTIMER.Loop field is set. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1187 26.3.8 Special Cases About Interrupts 26.3.8.1 Ordering Issue Between CSR Rd/Wr Received With Error by the Controller in the PCH and its Generated Interrupt When a CSR Rd/Wr request is received by the Controller in the PCH with error, interrupt is generated to EP before the request has been serviced by GbE i.e. read data returned to host or write data written to its destination. This breaks the PCIE ordering rules as interrupt messaging may actually pass actual CSR read/write completion. 26.4 802.1q VLAN Support The Controller provides several specific mechanisms to support 802.1q VLANs: * Optional adding (for transmits) and stripping (for receives) of IEEE 802.1q VLAN tags. * Optional ability to filter packets belonging to certain 802.1q VLANs. 26.4.1 802.1q VLAN Packet Format The following diagram compares an untagged 802.3 Ethernet packet with an 802.1q VLAN tagged packet: Table 26-51. Comparing Packets 802.3 Packet DA #Octets 6 802.1q VLAN Packet #Octets DA 6 6 SA 6 SA Type/Length 2 802.1q Tag 4 Data 46-1500 Type/Length 2 CRC 4 Data 46-1500 CRC* 4 Note: The CRC for the 802.1q tagged frame is re-computed, so that it covers the entire tagged frame including the 802.1q tag header. Also, max frame size for an 802.1q VLAN packet is 1522 octets as opposed to 1518 octets for a normal 802.3z Ethernet packet. 26.4.2 802.1q Tagged Frames For 802.1q, the Tag Header field consists of four octets comprised of the Tag Protocol Identifier (TPID) and Tag Control Information (TCI); each taking 2 octets. The first 16 bits of the tag header makes up the TPID. It contains the "protocol type" which identifies the packet as a valid 802.1q tagged packet. The two octets making up the TCI contain three fields: * User Priority (UP) * Canonical Form Indicator (CFI). Should be 0b for transmits. For receives, the device has the capability to filter out packets that have this bit set. See the CFIEN and CFI bits in the RCTL described in Section 28.9.1.1. * VLAN Identifier (VID) The bit ordering is shown below: Intel(R) Communications Chipset 89xx Series - Datasheet 1188 October 2012 Order Number: 327879-001US 26.0 Table 26-52. TCI Bit Ordering Octet 1 UP Octet 2 VID 26.4.3 Transmitting and Receiving 802.1q Packets 26.4.3.1 Adding 802.1q Tags on Transmits Software might command the Controller to insert an 802.1q VLAN tag on a per packet or per flow basis. If CTRL.VME is set to 1b, and the VLE bit in the transmit descriptor is set to 1b, then the Controller inserts a VLAN tag into the packet that it transmits over the wire. The Tag Protocol Identifier (TPID) field of the 802.1q tag comes from the VET register. 8021.Q tag insertion is done in different ways for legacy and advanced Tx descriptors: * Legacy Transmit Descriptors:, The Tag Control Information (TCI) of the 802.1q tag comes from the VLAN field (see Figure 26-8) of the descriptor. See Table 26-26 for more information regarding hardware insertion of tags for transmits. * Advanced Transmit Descriptor: The Tag Control Information (TCI) of the 802.1q tag comes from the VLAN Tag field (see Table 26.2.2.2.1) of the advanced context descriptor. The IDX field of the advanced Tx descriptor should be set to the adequate context. 26.4.3.2 Stripping 802.1q Tags on Receives Software might instruct the Controller to strip 802.1q VLAN tags from received packets. If the CTRL.VME bit is set to 1b, and the incoming packet is an 802.1q VLAN packet (its Ethernet Type field matched the VET), then the Controller strips the 4 byte VLAN tag from the packet, and stores the TCI in the VLAN Tag field (see Figure 26-4 and See "Receive UDP Fragmentation Checksum) of the receive descriptor. The Controller also sets the VP bit in the receive descriptor to indicate that the packet had a VLAN tag that was stripped. If the CTRL.VME bit is not set, the 802.1Q packets can still be received if they pass the receive filter, but the VLAN tag is not stripped and the VP bit is not set. See Table 26-25 for more information regarding receive packet filtering. 26.4.4 802.1q VLAN Packet Filtering VLAN filtering is enabled by setting the RCTL.VFE bit to 1b. If enabled, hardware compares the type field of the incoming packet to a 16-bit field in the VLAN Ether Type (VET) register. If the VLAN type field in the incoming packet matches the VET register, the packet is then compared against the VLAN Filter Table Array (VFTA[127:0]) for acceptance. The Controller provides exact VLAN filtering for VLAN tags for host traffic and VLAN tags for manageability traffic. The Virtual LAN ID field indexes a 4096 bit vector. If the indexed bit in the vector is one; there is a Virtual LAN match. Software might set the entire bit vector to ones if the node does not implement 802.1q filtering. The register description of the VLAN Filter Table Array is described in detail in Section 28.9.1.20. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1189 In summary, the 4096-bit vector is comprised of 128, 32-bit registers. The VLAN Identifier (VID) field consists of 12 bits. The upper 7 bits of this field are decoded to determine the 32-bit register in the VLAN Filter Table Array to address and the lower 5 bits determine which of the 32 bits in the register to evaluate for matching. The BMC configures the Controller with eight different manageability VIDs via the Management VLAN TAG Value [7:0] - MAVTV[7:0] registers and enables each filter in the MDEF register. Two other bits in the Receive Control register (see Section 28.9.1.1), CFIEN and CFI, are also used in conjunction with 802.1q VLAN filtering operations. CFIEN enables the comparison of the value of the CFI bit in the 802.1q packet to the Receive Control register CFI bit as acceptance criteria for the packet. Note: The VFE bit does not effect whether the VLAN tag is stripped. It only effects whether the VLAN packet passes the receive filter. The following table lists reception actions per control bit settings. Figure 26-25.Packet Reception Decision Table Is packet 802.1q? CTRL. VME RCTL. VFE Action No Xa X1 Normal packet reception Yes 0b 0b Receive a VLAN packet if it passes the standard MAC address filters (only). Leave the packet as received in the data buffer. VP bit in receive descriptor is cleared. Yes 0b 1b Receive a VLAN packet if it passes the standard filters and the VLAN filter table. Leave the packet as received in the data buffer (the VLAN tag would not be stripped). VP bit in receive descriptor is cleared. Yes 1b 0b Receive a VLAN packet if it passes the standard filters (only). Strip off the VLAN information (four bytes) from the incoming packet and store in the descriptor. Sets VP bit in receive descriptor. Yes 1b 1b Receive a VLAN packet if it passes the standard filters and the VLAN filter table. Strip off the VLAN information (four bytes) from the incoming packet and store in the descriptor. Sets VP bit in receive descriptor. a. X - Don't care Note: A packet is defined as a VLAN/802.1q packet if its type field matches the VET. 26.4.5 Double VLAN Support The Controller supports a mode where all received and sent packet have at least one VLAN tag in addition to the regular tagging which might optionally be added. This mode is used for systems where the switches add an additional tag containing switching information. This mode is activated by setting CTRL_EXT.EXTENDED_VLAN bit. The default value of this bit is set according to the Ext_VLAN (bit 1) in the Initialization Control 3 EEPROM word for ports 0 to 3. See Section 24.3.15 for more information. The type of the VLAN tag used for the additional VLAN is defined in the VET.VET_EXT field. Intel(R) Communications Chipset 89xx Series - Datasheet 1190 October 2012 Order Number: 327879-001US 26.0 26.4.5.1 Transmit Behavior With External VLAN It is expected that the driver include the external VLAN header as part of the transmit data structure. The software may post the internal VLAN header as part of the transmit data structure or embedded in the transmit descriptor (see Section 26.2.2 for details). the Controller does not relate to the external VLAN header other than the capability of "skipping" it for parsing of inner fields. Notes: * The VLAN header in a packet that carries a single VLAN header is treated as the external VLAN. * The Controller expects that any transmitted packet has at least the external VLAN added by the software. For those packets where an external VLAN is not present, any offload that relates to inner fields to the EtherType may not be provided. 26.4.5.2 Receive Behavior With External VLAN When a port of the Controller is working in this mode, the Controller assumes that all packets received by this port have at least one VLAN, including packet received or sent on the manageability interface. One exception to this rule are flow control PAUSE packets which are not expected to have any VLAN. Other packets may contain no VLAN, however a received packet that does not contain the first VLAN is forwarded to the host but filtering and offloads are not applied to this packet. See the table below for the supported receive processing when the device is set to "Double VLAN" mode. Stripping of VLAN is done on the second VLAN if it exists. All the filtering functions of the Controller ignore the first VLAN in this mode. The presence of a first VLAN tag is indicated it in the RDESC.STATUS.VEXT bit. Queue assignment of the Rx packets is not affected by the extended VLAN header. It may depend on the internal VLAN, MAC address or any upper layer content as described in Section 26.1.1. Table 26-53. Receive Processing in Double VLAN Mode VLAN Headers Status.VEXT Status.VP Packet Parsing Rx offload functions Extended and internal 1 1 + + Internal Only Not supported V-Ext 1 0 + + Nonea 0 0 + (flow control only) - a. A few examples for packets that may not carry any VLAN header may be: Flow control and Priority Flow Control; LACP; LLDP; GMRP; 802.1x packets October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1191 26.5 Configurable LED Outputs The Controller implements 4 output drivers (one per port) intended for driving external LED circuits per port. Each LAN device provides an independent LED output - the pin and its function are bound to a specific LAN device. The configuration for LED outputs is specified via the LEDCTL register. Furthermore, the hardware-default configuration for all the LED outputs, can be specified via EEPROM fields, thereby supporting LED displays configurable to a particular OEM preference. LED's can be configured to use one of a variety of sources for output indication. The MODE bits control the LED source as described in Table 26-54. The IVRT bits allow the LED source to be inverted before being output or observed by the blink-control logic. LED outputs are assumed to normally be connected to the negative side (cathode) of an external LED. The BLINK bits control whether the LED should be blinked (on for 200ms, then off for 200ms) while the LED source is asserted. The blink control might be especially useful for ensuring that certain events which are sufficiently visible by a human eye. Note: When LED Blink mode is enabled the appropriate LED Invert bit should be set to 0b. 26.5.1 MODE Encoding for LED Outputs Table 26-54 lists the MODE encoding for LED outputs used to select the desired LED signal source for each LED output. Table 26-54. Mode Encoding for LED Outputs a Mode 0010b Selected Mode Source Indication LINK_UP Asserted when any speed link is established and maintained. 0011b FILTER_ACTIVITY Asserted when link is established and packets are being transmitted or received that passed MAC filtering. 0100b LINK/ACTIVITY Asserted when link is established and when there is no transmit or receive activity. 1001b FULL_DUPLEX Asserted when the link is configured for full duplex operation (de-asserted in half-duplex). 1010b COLLISION Asserted when a collision is observed. 1011b ACTIVITY Asserted when link is established and packets are being transmitted or received. 1101b PAUSED Asserted when the PCH's transmitter is flow controlled. 1110b LED_ON Always high (Asserted) 1111b LED_OFF Always low (De-asserted) a. Encoding not shown are reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 1192 October 2012 Order Number: 327879-001US 26.0 26.6 Error Correction and Detection The Controller's main internal memories are protected by error correcting code or parity bits. The larger memories or critical memories are protected by an error correcting code (ECC). The smaller memories are protected either with an error correcting code (ECC for critical memories) or by parity. 26.7 CPU affinity Features 26.7.1 Direct Cache Access (DCA) 26.7.1.1 DCA Description Direct Cache Access (DCA) is a method to improve network I/O performance by placing some posted inbound writes indirectly within CPU cache. DCA requires that memory writes go to host memory and then the processor prefetch the cache lines specified by the memory write. Through research and experiments, DCA has been shown to reduce CPU Cache miss rates significantly. Figure 26-26.Diagram of DCA Implementation on FSB System CPU CPU demand read Cache BIL-DCA MCH DCA triggered HW Prefetch Memory Memory Write DMA Write NIC As shown in Figure 26-26, DCA provides a mechanism where the posted write data from an I/O device, such as an Ethernet NIC, can be placed into CPU cache with a hardware pre-fetch. This mechanism is initialized upon a power good reset. A software device driver for the I/O device configures the I/O device for DCA and sets up the appropriate DCA target ID for the device to send data. DCA implementation is controlled by separated registers (RXCTL and TXCTL) for each receive and transmit queue. In addition, a DCA Enable bit can be found in the DCA_CTRL register, and a DCA_ID register can be found for each port, in order to make visible the function, device, and bus numbers to the driver. The RXCTL and TXCTL registers can be written by software on the fly and can be changed at any time. When software changes the register contents, hardware applies changes only after all the previous packets in progress for DCA have been completed. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1193 There are 2 modes for DCA implementation: 1. Legacy DCA: The DCA target ID is derived from CPU ID. 2. DCA: The DCA target ID is derived from APIC ID. The software driver selects one of these modes through the DCA_mode register. The details of both modes are described below. 26.7.1.2 Details of Implementation 26.7.1.2.1 PCIe Message Format for DCA Figure 26-27 shows the format of the PCIe message for DCA. Figure 26-27.PCIe Message Format for DCA +0 6 R 5 Fmt= 11 4 3 +1 2 1 0 Type=00000b 7 6 R 5 4 3 TC +2 2 1 0 Rsv 7 6 T D E P Requester ID 5 4 Attr +3 3 2 1 R DCA preferences 0 7 6 5 4 3 2 1 0 Length Last DW BE First DW BE Address [63:32] Address [32:2] R Length specific data Length specific data TLP digest The DCA preferences field has the following formats. Table 26-55. Legacy DCA Systems Bits Name Description 0 DCA indication 0b: DCA disabled 1b: DCA enabled 4:1 DCA Target ID The DCA Target ID specifies the target cache for the data. 7:5 Reserved Reserved Table 26-56. DCA Systems Bits Name Description 7:0 DCA target ID 0000.0000b: DCA is disabled Other: Target Core Id derived from APIC Id. The method for this is described in DCA Platform Architecture Specification, section 7.3.1 Intel(R) Communications Chipset 89xx Series - Datasheet 1194 October 2012 Order Number: 327879-001US 26.0 Note: All functions within the Controller have to adhere to the "tag encoding" rules for DCA writes. Even if a given function is not capable of DCA, but other functions are capable of DCA, memory writes from the non-DCA function must set the Tag field to "00000000". 26.8 Virtualization I/O virtualization is a mechanism to share I/O resources among several consumers. For example, in a virtual system, multiple operating systems are loaded and each executes as though the whole system's resources were at its disposal. However, for the limited number of I/O devices, this presents a problem because each operating system might be in a separate memory domain and all the data movement and device management has to be done by a VMM (Virtual Machine Monitor). VMM access adds latency and delay to I/O accesses and degrades I/O performance. Virtualized devices are designed to reduce the burden of VMM by making certain functions of an I/O device shared and thus can be accessed directly from each guest operating system or Virtual Machine (VM). Two modes to support operation in a Virtualized environment were implemented in previous products: 1. Direct assignment of part of the port resources to different guest OSes using the PCI sig SR-IOV standard. Also known as "Native mode" or pass through mode. This mode is referenced as IOV mode through this chapter 2. Central management of the networking resources by an IOVM or by the VMM. This mode is referenced as VMDq mode. Two VMDq modes exist, VMDq1 and a more advanced version named VMDq2. In a virtualized environment, the Controller serves up to 8 virtual machines (VMs). The Controller supports fully VMDq1 mode and partially VMDq2 mode and does not support SR-IOV. In a virtualized environment, the Controller serves up to 8 virtual machines (VMs) per port. The Controller's 16 queues can be accessed by 8 different VMs if configured properly. When the Controller is enabled for multiple queue direct access for VMs, it becomes a VMDq device. Note: Most configuration and resources are shared across queues. System software must resolve any conflicts in configuration between the VMs. The following section describes the support the Controller provides for VMDq. 26.8.1 Assignment of MSI-X Vectors to VM MSI-X vectors are used for three purposes: 1. Differentiation of interrupt causes that avoids the need to read an interrupt cause register. 2. Assignment of different interrupt handling to different CPUs. 3. The implementation of interrupts in the Controller adds another use of allowing different interrupt moderation rates. MSI-X vectors are allocated for cause differentiation and interrupt rate differentiation. The interrupts that the VM driver needs to handle are Tx packet sent per queue, Rx packet received per queue, TCP timer and some special events. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1195 26.8.2 VM Resource Summary The Controller supports 8 VMs per port, each VM can utilize 2 queue pair (Tx/Rx) per VM and 3 MSI-X vectors per VM. If the amount of VMs supported is less than 8, the available resources can be distributed between the active VMs. 26.8.3 Packet Switching (VMDq) Model 26.8.3.1 VMDq Assumptions To support receive packet replication it is assumed that Broadcast and Multicast traffic volume is relatively low. In the case of large Broadcast and Multicast traffic volume that causes congestion on the PCIe interface, the Storm Control circuitry can be programmed to drop Broadcast and Multicast packets or flow control can be activated to avoid internal receive buffer overflow. 26.8.3.2 VM Selection The VM selection is done by MAC address and VLAN tag. Broadcast and Multicast packets are forwarded according to the individual setting of each VM and might be replicated to multiple VMs. 26.8.3.2.1 Filtering Capabilities The following capabilities exists in to decide what is the final destination of each packet in addition to the regular L2 filtering capabilities: * MAC addresses filters (RAH/RAL registers) for both unicast and multicast filtering. These are shared with L2 filtering. For example, the same MAC addresses are used to determine if a packet is received by the switch and to determine the forwarding destination. * Shared VLAN filters (VLVF registers) - each VM can be made member of each VLAN. * Multicast exact filtering using the existing remaining RAH/RAL registers otherwise an imperfect multicast table is shared between VMs. * 256 hash filtering of multicast addresses shared between the VMs (MTA table). * Promiscuous unicast, multicast & enable broadcast per VM. Note: Packets for which no queueing decision was done and still accepted by the L2 filtering, is directed to the queue pool of the default VM or dropped. 26.8.3.3 L2 Filtering L2 filtering is the 1st stage of 3 stages that determine the destination of a received packet. The 3 stages are defined in Section 26.1.1. All received packets pass the same filtering as in the non virtualized case; regular VLAN filtering using the global VLAN table (VFTA) and filtering according to the RAH/RAL registers and according to the various promiscuous bits. Note: Every VLAN tag set in the VLVF registers should be asserted also in the VFTA table. Note: The RCTL.UPE bit (Promiscuous unicast) is not available per VM and might be modified by the IOVM or VMM. Intel(R) Communications Chipset 89xx Series - Datasheet 1196 October 2012 Order Number: 327879-001US 26.0 26.8.3.4 VMDq Receive Packets Switching Receive packet switching is the 2nd stage of 3 stages that determine the destination of a received packet. The 3 stages are defined in Section 26.1.1. In this stage the VM is identified by the "pool list" as described in Section 26.8.3.4.1 and Section 26.8.3.4.2. When working in a virtualized environment, a single receive queue can still be determined by the Ethertype filters. If these filters don't match, then a pool list should be found. When working in replication mode, broadcast and multicast packets can be forwarded to more than one VM, and can be replicated to more than one receive queue. Replication is enabled by the Rpl_En bit in the VT_CTL register. In virtualization mode, the pool list is a list of one or more VMs to which the packet should be forwarded. The pool list is used in choosing the target queue list except for cases in which high priority filters take precedence. There is a difference in the way the pool list is generated when replication mode is enabled or disabled. 26.8.3.4.1 VMDq Replication Mode Enabled When replication mode is enabled (VT_CTL.Rpl_En = 1), each broadcast/multicast packet can go to more than one pool. Finding the pool list should be done according to the following steps: 1. Exact unicast or multicast match - If there is a match in one of the exact filters (RAL/RAH), for unicast or multicast packets, take the RAH.POOLSEL[7:0] field as a candidate for the pool list. 2. Broadcast - If the packet is a broadcast packet, add pools for which their VMOLR.BAM bit (Broadcast Accept Mode) is set. 3. Unicast hash - If the packet is a unicast packet, and the prior steps yielded no pools, check it against the Unicast hash table (UTA). If there is a match, add pools for which their VMOLR.ROPE bit (Receive Overflow packet enable) is set. 4. Multicast hash - If the packet is a multicast packet and the prior steps yielded no pools, check it against the multicast hash table (MTA). If there is a match, add pools for which their VMOLR.ROMPE bit (Receive Multicast packet enable) is set. 5. Multicast Promiscuous - If the packet is a multicast packet, take the candidate list from prior steps and add pools for which their VMOLR.MPE bit (Multicast Promiscuous Enable) is set. 6. Ignore MAC (VLAN only filtering) - If VT_CTL.IGMAC bit is set, then the previous steps are ignored and a full pool list is assumed for the next step. 7. VLAN groups - This step is relevant only if the RCTL.VFE bit is set, otherwise it is skipped. Packets should be sent only to VMs that belong to the packet's VLAN group. Note: a. Tagged packets: enable only pools in the packet's VLAN group as defined by the VLAN filters - VLVF[n].VLAN_id and their pool list - VLVF[n].POOLSEL[7:0]. b. Untagged packets: enable only pools with their VMOLR.AUPE bit set c. If there is no match, the pool list should be empty. In a VLAN network, untagged packets are not expected. Such packets received by the switch should be dropped, unless their destination is a virtual port set to receive these packets. The setting is done through the VMOLR.AUPE bit. It is assumed that VMs for which this bit is set are members of a default VLAN and thus only MAC queuing is done on these packets. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1197 8. Default Pool - If the pool list is empty at this stage and the VT_CTL.Dis_Def_Pool bit is not set, then set the default pool bit in the target pool list (from VT_CTL.Def_PL). 9. Ethertype filters - If the one of the Ethertype filters (ETQF) matches the packet and queuing action is requested, the VM list is set to the pool pointed by the filter. 10. Mirroring - For each of the 4 mirroring rules add the destination (mirroring) pool (VMRCTL.MP) to the pool list according to the following rules: a. Pool mirroring - if VMRCTL.VPME is set and one of the bits in the pool list matches one of the bits in the VMRVM register. b. VLAN port mirroring - if VMRCTL.VLME is set and the index of the VLAN of the packet in the VLVF table matches one of the bits in the VMRVLAN register. c. Uplink port mirroring - if VMRCTL.UPME is set and the pool list is not empty and the packet came from the LAN. 11. Length Limit: If the packet is longer than a legal Ethernet packet, remove from the pool list all the pools for which the VMOLR.LPE bit is not set or for which the packet length is larger than the value in the VMOLR.RLPML field. The above process, up to stage 9. can be logically described by the following scheme: Intel(R) Communications Chipset 89xx Series - Datasheet 1198 October 2012 Order Number: 327879-001US 26.0 Figure 26-28.Pool List Selection - Replication Enabled Exact match RAH.POOLSEL[7:0] Broadcast packet & VMOLR[7:0].BAM Multicast packet & MTA match Unicast packet & VMOLR[7..0].GPE & UTA match 0xFF/0x00 Bitwise OR Bitwise OR 8 Multicast packet & VMOLR[7..0]..MPE 8 0 1 VMD_CTL.IGMAC * 8 VFE && untagged packet && VMOLR[7:0].AUPE 8 ~VFE || match VLVF.POOLSEL[7:0] Bitwise OR 8 8 Bitwise AND Default Pool Dis_def_Pool 8 8 0 1 Pool from ETQF 8 Target Pool List[7:0] 26.8.3.4.2 VMDq Replication Mode Disabled When replication mode is disabled (VT_CTL.Rpl_En = 0), the software should take care of multicast and broadcast packets and check which of the VMs should get them. In this mode the pool list always contains one pool only according to the following steps: 1. Exact unicast or multicast match - If the packet DA matches one of the exact filters (RAL/RAH), take the RAH.POOLSEL[7:0] field as a candidate for the pool list. 2. Ignore MAC (VLAN only filtering) - If VT_CTL.IGMAC bit is set, then the previous step is ignored and a full pool list is assumed for the next step. 3. Unicast hash - If the packet is a unicast packet, and the prior steps yielded no pools, check it against the Unicast hash table (UTA). If there is a match, add the October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1199 pool for which the VMOLR.ROPE bit (Receive Overflow packet enable) is set. (See software limitation no 3. below). 4. VLAN groups - This step is relevant only if the RCTL.VFE bit is set, otherwise it is skipped. Packets should be sent only to VMs that belong to the packet's VLAN group. a. Tagged packets: enable only pools in the packet's VLAN group as defined by the VLAN filters - VLVF[n].VLAN_id and their pool list - VLVF[n].POOLSEL[7:0]. b. Untagged packets: enable only pools with their VMOLR.AUPE bit set c. If there is no match, the pool list should be empty. 5. Default pool- If the packet is a unicast packet and no pool was chosen and the VT_CTL.Dis_Def_Pool bit is not set, then set the default pool bit in the pool list (from VT_CTL.Def_PL). 6. Broadcast or Multicast - If the packet is a Multicast or Broadcast packet and was not forwarded in step 1 and 2, set the default pool bit in the pool list (from VT_CTL.Def_PL). 7. Length Limit: If the packet is longer than a legal Ethernet packet, remove from the pool list all the pools for which the VMOLR.LPE bit is not set or for which the packet length is larger than the value in the VMOLR.RLPML field. The above process can be logically described by the following scheme: Figure 26-29.Pool List Selection - Replication Disabled Unicast packet & VMOLR7..0.ROPE & UTA match VFE && untagged packet && VMOLR[7:0].AUPE Exact match RAH.POOLSEL[7:0] ~VFE || match VLVF.POOLSEL[7:0] VMD_CTL.IGMAC 8 8 8 Unicast DIS_DEF_POOL Default Pool 8 8 0 1 8 Target pool List[7:0] Intel(R) Communications Chipset 89xx Series - Datasheet 1200 October 2012 Order Number: 327879-001US 26.0 The following limitations applies when replication is disabled: 1. It is the software responsibility to not set more than one bit in the bitmaps of the exact filters. Multiple bits might be set in an RAH register as long as it is guaranteed that the packet is sent to only one queue by other means (VLAN) 2. The software must not set per-VM promiscuous bits (multicast or broadcast). 3. The software must not set the Rope bit in more than one VMOLR register. 4. If VT_CTL.IGMAC bit is set, the software must not set the VMOLR.AUPE in more than one VMOLR register and must not set more than one bit in each of the VLVF.POOLSEL bitmaps. 5. The software must not activate mirroring. 6. The software should take care not to set the Rope bit in more than one VMOLR register. 26.8.3.5 Mirroring Support The Controller supports 4 mirroring rules. Each rule can be of one of 3 types. Only Egress mirroring is supported and not ingress. For example, the mirroring is done on the receive path and mirrored packets reflect all the changes that occur to the received packet (e.g. Vlan stripping). Mirroring is supported only to virtual ports and not to the uplink (i.e. a mirrored packet can not be sent back to the Network). Mirroring should be activated only when one of the VMDq queueing modes is used. The following types of rules are supported: 1. Virtual port mirroring - reflects all the packets sent to a set of given VMs. 2. Uplink port mirroring - reflects all the traffic received from the network. 3. VLAN mirroring - reflects all the traffic received in a set of given VLANs. All the modes can be accumulated into a single rule. This new mirroring mode is controlled by a set of rule control registers: * VMRCTL - controls the rules to be applied and the destination port. * VMRVLAN - controls the VLAN ports as listed in the VLVF table taking part in the VLAN mirror rule. * VMRVM - controls the VMs ports taking part of the Virtual port mirror rule. Mirroring is supported only when replication is enabled. The exact flow of mirroring is described in step 10. in Section 26.8.3.4.1. 26.8.3.6 VMDq Offload Support In case of packets directed to one VM only, the offloads are determined by this specific VM setting. However, the Controller can not apply different offloads (VLAN & CRC strip + decision of size of header for split/replication offload) to different replication of the same packet. The following sections describes the rules used to decide which offloads to apply in case of replicated packets. If replication is disabled, the offloads are determined by the unique destination of the packet. Note: In a virtualization environment (MRQC.Multiple Receive Queues Enable = 011b), the global VLAN strip and CRC strip bits (CTRL.VME & RCTL.SECRC) are ignored and the VM specific bits in VMOLR & RPLOLR are used instead. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1201 26.8.3.6.1 Replication by Exact MAC Address As mentioned above, the same MAC address can be assigned to more than one VM. This is used for the following cases: * Multicast address - In this case, the different VMs might be part of the same VLAN. The offloads applied to packets matching this address are defined in the replicated packets offloads register (RPLOLR register). * Unicast - Same MAC different VLAN - In this case, each VM should belong to different VLAN(s). The applied offloads is according to the pool selected by the MAC/VLAN pair. One exception is the VLAN strip decision which is done according to the first pool parameters. This means that all the pools sharing a MAC address should use a common VLAN strip policy. 26.8.3.6.2 Replication by Promiscuous Modes A packet might be replicated to multiple VMs because part of the VMs are set to receive all multicast or broadcast packets or because of a packet matching one of the hash tables (UTA or MTA). The offloads applied to packet are defined in the replicated packets offloads registers (RPLOLR register). In case of unicast packet, the offloads are applied according to the first pool selected to receive the packet. 26.8.3.6.3 Replication by Mirroring Offloads of mirrored packets are determined according to the original pool. 26.8.3.6.4 VLAN Only Filtering If VT_CTL.IGMAC bit is set, the pool is defined according to the VLAN only. In this mode, only a uniform VLAN strip policy is supported. This means that the VMOLR.STRVLAN bit should be set to the same value for all VMs. 26.8.3.6.5 Small Packets Padding In Virtualized systems, the driver receiving the packet in the VM might not be aware of all the hardware offloads applied to the packet. Thus, in case of stripping actions by the hardware (VLAN strip), it might receive packets which are smaller than a legal packet. The Controller provides an option to pad small packets in such cases so that all packets have a legal size. This option can be enabled only if the CRC is stripped. In these cases, all packets are padded to 60 bytes (legal packet - 4 bytes CRC). The padding is done with zero data. This function is enabled via the RCTL.PSP bit. 26.8.3.7 Security Features The Controller allows some security checks on the inbound and outbound traffic of the switch. 26.8.3.7.1 Inbound Security Each incoming packet from the LAN is filtered according to the VLAN tag so that packets from one VLAN can not be received by VMs that are not members of that VLAN. Intel(R) Communications Chipset 89xx Series - Datasheet 1202 October 2012 Order Number: 327879-001US 26.0 26.8.3.7.2 Interrupt on Misbehavior of VM (Malicious Driver Detection) The hardware can be programmed to take some action as a result of some misbehavior of a VM. These actions might hint to the fact that some VM is malicious and the VMM should remedy the situation. In order to inform the VMM of this fact, an interrupt bit exists in the ICR register (ICR.MDDET bit) to indicate the occurrence of such behavior. The LVMMC register contains information on which queue (LVMMC.Last_Q) and port (LVMMC.Mal_PF) the malicious behavior was detected. The LVMMC register is clear by read. Malicious driver behavior detection is enabled by setting the DTXCTL.MDP_EN bit to 1. In addition to fully enable the Malicious driver detection functionality the DTXCTL.OutOfSyncEnable bit should also be set to 1b. On detection of a malicious driver event the Controller stops activity of the offending queue and generates an interrupt by asserting the ICR.MDDET bit. Cause of Malicious driver activation is reported in the LVMMC register. To re-activate offending queue driver should write a 0 to the relevant bit in the MDFB register. 26.8.3.7.3 Storm Control As there is no separate path for multicast & broadcast packets, too much replicated packets might cause congestions in the data path. In order to avoid such scenarios, broadcast and multicast storm control rate limiters are added. The rate controllers define windows and the maximal allowed number of multicast or broadcast bytes/packets per window. Once the threshold is crossed different types of policies can be applied. 26.8.3.7.3.30 Assumptions * Only one interval size and interval counter is used for both broadcast & multicast storm control mechanisms. * The threshold and actions for each mechanism are separate. * The traffic used to calculate the broadcast & multicast rate is all the traffic with a local destination. * The storm control does not block traffic to the network. * The basic unit of traffic counted is 64 bytes of data. 26.8.3.7.3.31 Storm Control Functionality The time interval over which Broadcast Storm control is performed is controlled by three factors. * SCBI register * Port speed. * The value in SCCRL.INTERVAL The first two factors determine the Unit time interval as described in Table 26-57. The interval is automatically chosen internal to hardware based on port speed. The third factor (Interval field) determines how many of such unit intervals are considered for one Storm Control Interval. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1203 Table 26-57. Storm Control Interval by Speed Port Speed MIN Time Interval MAX Time Interval 1 Gb/s 100 s 100 ms 100 Mb/s 1 ms 1s 10 Mb/s 10 ms 10 s The number of 64 bytes chunk of Broadcast or Multicast packets that are allowed in a given interval is determined by setting the BSCTRH or MSCTRH register respectively. The Controller supports two modes of reactions to storm event: 1. Block all Multicast or Broadcast packets from the moment the threshold is crossed until the end of the interval. The block is removed at the end of the interval until the threshold is crossed again. This mode is set by asserting SCCRL.MDICW (for multicast) or SCCRL.BDICW (for broadcasts). This mode is used as a rate limiter. 2. Block all Multicast or Broadcast packets from the moment the threshold is crossed until a full interval without threshold crossing is registered. The block is removed at the end of the interval until the threshold is crossed again. This mode is set by asserting SCCRL.MDICW and SCCRL.MDIPW (for multicast) or SCCRL.BDICW and SCCRL.BDIPW (for broadcasts). This mode is used for storm blocking. The Controller can be programmed to add all packets for which a queue was not found for storm control calculation. For example, packets that passed the 1st stage of L2 filtering but didn't pass the 2nd stage of pooling, or where sent to the default pool, as broadcast packets. This mode is activated by setting the SCCRL.BIDU field. Any change in the storm control state (block or pass of multicast or broadcast packets) is indicated to the software via the ICR.SCE interrupt cause. The current state is reflected in the SCSTS register. For diagnostic purpose only, the storm control timer and counters can be read via the SCTC, MSCCNT & BSCCNT registers. 26.8.3.8 External Switch Loopback Support One of the long term solutions for the switching issue is a mode where an external switch would do the loopback of VM to VM traffic and the NIC is responsible for the replication of multicast packets only. In order to support this mode the received packets SA should be compared to the exact MAC addresses to check if the packet originated from a local source, so that the packet is not forwarded to the VM originator. This mode is enabled by the VT_CTL.FLP bit. Intel(R) Communications Chipset 89xx Series - Datasheet 1204 October 2012 Order Number: 327879-001US 26.0 26.8.3.9 Switch Control The VMM/IOVM driver has some control of the switch logic. The following registers are available to the VMM/IOVM for this purpose: VLVF: VLAN queuing table: A set of 32 VLAN entries with an associated per VM bit map allowing allocation of each VM to each of the 32 VLAN tags. VT_CTL: VT Control register - contains the following fields: * Replication enable - allows replication of multicast & broadcast packets. If this bit is cleared, Rx multicast & broadcast packets are sent to the default VM. * Default pool - defines where to send packets that passed L2 filtering but didn't pass any of the queueing mechanisms. * Default pool disable- defines whether to drop packets that passed L2 filtering but didn't pass any of the queueing mechanisms. VMOLR/RPMOLR: Defines the offloads and pool selection options for each VM and for replicated packets. In addition the storm control mechanism is programmed as described in Section 26.8.3.7.3.31. 26.8.4 Virtualization of the Hardware This section describes additional features used in VMDq mode. 26.8.4.1 Per Pool Statistics Part of the statistics are by definition shared and can not be allocated to a specific VM. For example, CRC error count can not be allocated to a specific VM, as the destination of such a packet is not known if the CRC is wrong. All the non specific statistics is handled by the VMM in the same way as it is done in non virtualized systems. A VM might require a statistic from the VMM but might not access it directly. The conceptual model used to gather statistics in a virtualization context is that each queue pool is considered as a virtual link and the Ethernet link is considered as the uplink of the switch. Thus any packet sent by a VM is counted in the Tx statistics, even if it was dropped by the MAC from some reason. In the same way, a replicated packet is counted in each of the VMs receiving it. The following statistics are be provided per VM: 1. Good Packet received count (VFGPRC). 2. Good Packet transmitted count (VFGPTC). 3. Good octets received count (VFGORC). 4. Good octets transmitted count (VFGOTC). 5. Multicast Packets Received Count (VFMPRC). Note: All the per VM statistics are read only (RO) and wrap around after reaching their maximal value. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1205 26.9 Time SYNC (IEEE1588 and IEEE 802.1AS) IEEE 1588 addresses the clock synchronization requirements of measurement and control systems. The protocol supports system-wide synchronization accuracy in the sub-microsecond range with minimal network and local clock computing resources. The protocol is spatially localized and allows simple systems to be installed and operated without requiring the administrative attention of users. The IEEE802.1AS standard specifies the protocol used to ensure that synchronization requirements are met for time sensitive applications, such as audio and video, across Bridged and Virtual Bridged Local Area Networks consisting of LAN media where the transmission delays are fixed and symmetrical; for example, IEEE 802.3 full duplex links. This includes the maintenance of synchronized time during normal operation and following addition, removal, or failure of network components and network reconfiguration. It specifies the use of IEEE 1588 specifications where applicable. Activation of the the Controller Time Sync mechanism is possible in full duplex mode only. No limitations on wire speed exist, although wire speed might affect the accuracy. Time Sync protocol is tolerant of dropping packets as well as missing timestamps. 26.9.1 Flow and Hardware/Software Responsibilities The operation of a PTP (Precision Time Protocol) enabled network is divided into two stages, initialization and time synchronization. At the initialization stage every master enabled node starts by sending Sync packets that include the clock parameters of its clock. Upon reception of a Sync packet a node compares the received clock parameters to its own. If the received clock parameters of a peer are better, the node moves to Slave state and stops sending Sync packets. When in slave state the node continuously compares the incoming packet clock parameters to its currently chosen master. If the new clock parameters are better then the current master selection, it changes master clock source. Eventually the best master clock source is chosen. Every node has a defined Sync packet time-out interval. If no Sync packet is received from its chosen master clock source during the interval it moves back to master state and starts sending Sync packets until a new Best Master Clock (BMC) is chosen. The time synchronization stage is different for master and slave nodes. If a node is in master state it should periodically send a Sync packet which is time stamped by hardware on the transmit path (as close as possible to the PHY). After the Sync packet a Follow_up packet is sent which includes the value of the timestamp kept from the Sync packet. In addition the master should timestamp Delay_Req packets on its RX path and return to the slave that sent it the timestamp value using a Delay_Response packet. A node in Slave state should timestamp every incoming Sync packet that is received from its selected master, software uses this value for time offset calculation. In addition it should periodically send Delay_Req packets in order to calculate the path delay from its master. Every sent Delay_Req packet sent by the slave is time stamped and kept. Using the value received from the master Delay_Response packet the slave can now calculate the path delay from the master to the slave. The synchronization protocol flow and the offset calculation are described in the following figure. Intel(R) Communications Chipset 89xx Series - Datasheet 1206 October 2012 Order Number: 327879-001US 26.0 Figure 26-32.Sync Flow and Offset Calculation T Sync Toffset = [(T2-T1)-(T3-T4)]/2 Timestamp T2 Timestamp Follow_up (T1) Master Delay_req T4 Timestamp Slave T3 Timestamp Delay_response (T4) The hardware's responsibilities are: 1. Identify the packets that require time stamping. 2. Time stamp the packets on both receive and transmit paths. 3. Store the time stamp value for software. 4. Keep the system time in hardware and give a time adjustment service to the software. 5. Maintain auxiliary features related to the system time. The software's responsibilities are: 1. BMC protocol execution which means defining the node state (master or slave) and selection of the master clock if in slave state. 2. Generate PTP packets, consume PTP packets. 3. Calculate the time offset and adjust the system time using the hardware mechanism. 4. Enable configuration and usage of the auxiliary features. Table 26-58. Chronological Order of Events for Sync and Path Delay Action Responsibility Node Role Generate a Sync packet with timestamp notification in descriptor Software Master Timestamp the packet and store the value in registers (T1) Hardware Master Timestamp incoming Sync packet, store the value in register and store the sourceID and sequenceID in registers (T2) Hardware Slave Read the timestamp from register, prepare a Follow_Up packet and send Software Master Once Follow_Up packet is received, load T2 from registers and T1 from Follow_up packet Software Slave Generate a Delay_Req packet with timestamp notification in descriptor Software Slave Timestamp the packet and store the value in registers (T3) Hardware Slave Timestamp incoming Delay_Req packet, store the value in register and store the sourceID and sequenceID in registers (T4) Hardware Master Read the timestamp from register and send back to Slave using a Delay_Response packet Software Master Once Delay_Response packet is received, calculate offset using T1, T2, T3 and T4 values Slave October 2012 Order Number: 327879-001US Software Intel(R) Communications Chipset 89xx Series - Datasheet 1207 26.9.1.1 TimeSync Indications in Receive and Transmit Packet Descriptors Certain indications are transferred between software and hardware regarding PTP packets. On the transmit path the software should set the 1588 bit in the transmit packet descriptor (MAC field bit 1). To indicate that the transmit packet time stamp should be taken and placed in the TXSTMPH and TXSTMPL time stamp registers. On the receive path the hardware transfers three indications to software in the receive descriptor: 1. An indication in RDESC.Packet Type that this packet is a PTP packet (no matter if timestamp is sampled or not). This indication is used also by PTP packets required for protocol management. Note: This indication is only relevant for L2 type packets (the PTP packet is identified according to its Ethertype). PTP packets have the L2Type bit in the Packet Type field set (bit 11) and the Etype matches the filter number set by the software to filter PTP packets. UDP type PTP packets don't require such an indication since the port number (319 for event and 320 for all other PTP packets) directs the packets toward the time sync application. 2. A second indication in the RDESC.STATUS.TS bit to indicate to the software that time stamp was taken for this packet and placed in the RXSTMPH and RXSTMPL time stamp registers. Software needs to access the time stamp registers to get the time stamp values. 3. A third indication in the RDESC.STATUS.TSIP bit to indicate that a time stamp was taken for this packet and placed at the start of the receive buffer (For further information see Section 26.1.10). 26.9.2 Hardware Time Sync Elements All time sync hardware elements are reset to their initial values as defined in the registers section upon MAC reset. 26.9.2.1 System Time Structure and Mode of Operation The time sync logic contains the SYSTIM counter to maintain the system time value. This is a 72 bit counter that is built of the SYSTIMR, SYSTIML and SYSTIMH registers. When in Master state the SYSTIMH, SYSTIML and SYSTIMR registers should be set once by the software according to general system requirements, when in slave state software should update the system time on every sync event as described in Section 26.9.2.3. Setting the system time is done by direct write to the SYSTIMH register and fine tuning the setting of the SYSTIM register, using the adjustment mechanism described in Section 26.9.2.3. Read access to the SYSTIMH, SYSTIML and SYSTIMR registers should be executed in the following order: 1. Software reads register SYSTIMR. Note: At this stage the hardware latches the value of SYSTIMH and SYSTIML registers. 2. Software reads register SYSTIML. 3. Software reads register SYSTIMH. Note: The latched SYSTIMH and SYSTIML values (from last read of SYSTIMR register) should be returned by hardware when reading the SYSTIML and SYSTIMH registers. Intel(R) Communications Chipset 89xx Series - Datasheet 1208 October 2012 Order Number: 327879-001US 26.0 The SYSTIM timer value in the SYSTIMH, SYSTIML and SYSTIMR registers, is updated periodically each 8 nS clock cycle according to the following formula: New SYSTIM = Old SySTIM + 8 nS +/- TIMINCA.Incvalue * 2-32 nS Where subtraction or addition of the TIMINCA.Incvalue value is defined according to the TIMINCA.ISGN value (0 - Add, 1 - Subtract). For the TIMINCA register description see section 28.16.1.12. 26.9.2.2 Time Stamp Mechanism The time stamp logic is located on transmit and receive paths at a location as close as possible to the PHY, to reduce delay uncertainties originating from implementation differences. The time stamp logic operation is slightly different on transmit and on receive paths. The transmit logic decides to timestamp a packet if the transmit timestamp is enabled (TSYNCTXCTL.EN = 1) and the time stamp bit in the packet descriptor (TDESD.MAC.1588 = 1) is set. On the transmit side only the time is captured in the TXSTMPL and TXSTMPH registers. The receive logic parses the received frame and timestamps the receive packet according to the conditions defined in the following fields: 1. TSYNCRXCTL.Type field that defines type of packets to be sampled. 2. TSYNCRXCFG.CTRLT field that defines message type criteria for timestamping V1 type packets when TSYNCRXCTL.Type register field equals 001b. 3. TSYNCRXCFG.MSGT field that defines message type criteria for timestamping V2 type packets when TSYNCRXCTL.Type register field equals 000b or 010b. When conditions to timestamp a receive packet are met, the timestamp is latched in the RXSTMPL and RXSTMPH registers and the packet's sourceId and sequenceId are latched in the RXSATRL and RXSATRH timestamp registers. Three indications are placed in the receive descriptor to support TimeSync operation: 1. RDESC.Packet Type - Value in this field identifies that this is a PTP packet (this indication is only for L2 packets since on the UDP packets the port number directs the packet to the application). 2. RDESC.STATUS.TS - Bit identifies that a time stamp was taken for this packet and latched in RXSTMPL and RXSTMPH registers and the packet's sourceId and sequenceId are latched in the RXSATRL and RXSATRH timestamp registers. 3. RDESC.STATUS.TSIP - Bit identifies that a time stamp was taken for this packet and placed at the start of the receive buffer (For further information see Section 26.1.10). For more details see the timestamp registers in Section 28.16. Figure 26-33 defines the exact point where the time value is captured. On both transmit and receive sides the timestamp values are locked in registers until software reads the TXSTMPH register to unlock Transmit timestamp registers or reads the RXSTMPH register to unlock the receive timestamp registers. As a result, if a new PTP packet that needs to be time stamped arrives before software accesses the timestamp registers, it is not time stamped. In some cases on the receive path a packet that was timestamped might be lost and not reach the host. To avoid a deadlock condition on the time stamp registers the software should keep a watch dog timer to clear locking of the time stamp register. The interval counted by such a timer should be higher then the expected interval between two Sync or Delay_Req packets depends on the node state (Master or Slave). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1209 Note: When TSYNCRXCTL.Type value is 100b, the Receive timestamp registers are not locked after a timestamp event. Figure 26-33.Time Stamp Point 26.9.2.3 Time Adjustment Mode of Operation A Node in a Time Sync network can be in one of two states Master or Slave. When a Time Sync entity is in the Master state it should synchronize other entities to its System Clock. In this case no time adjustments are needed. When the entity is in slave state it should adjust its system clock by using the data arriving in the Follow_Up and Delay_Response packets and the time stamp values of the Sync and Delay_Req packets. When all the values are available the software in the slave entity can calculate its offset in the following manner: Toffset = [(T2-T1) - (T3-T4)]/2 T1 T2 T3 T4 - Timing data in Follow_Up packet Sync Time Stamp Delay_Req Time Stamp Timing data in Delay_Response packet After offset calculation the system time register should be updated. This is done by writing the calculated offset to the TIMADJL and TIMADJH registers. The order should be as follows: 1. Write the low portion of the offset to the TIMADJL.TADJL field. 2. Write the high portion of the offset to the TIMADJH.TADJH field with the sign, to define if the offset should be added or subtracted, to the TIMADJH.Sign bit. After the writing to the TIMADJH register, the value of TIMADJH and TIMADJL is added once to the system time (SYSTIML and SYSTIMH). 26.9.3 Time Sync Related Auxiliary Elements The time sync logic implements three types of auxiliary elements using the precise system timer (SYSTIML and SYSTIMH). 26.9.3.1 Target Time The two target time registers TRGTTIML/H0 and TRGTTIML/H1 enable generating a time triggered event to external hardware using one of the SDP pins according to the setup defined in the TSSDP and TSAUXC registers (See Section 28.16.1.15 and Section 28.16.1.27). Each target time register is structured the same as the system Intel(R) Communications Chipset 89xx Series - Datasheet 1210 October 2012 Order Number: 327879-001US 26.0 time register. If the value of the system time is equal or has passed the value written to one of the target time registers, a change in level or a pulse is generated on the programmed SDP outputs. To generate a level change or pulse on one of the SDP pins driver should: 1. Select SDP pin using the appropriate TSSDP.TS_SDPx_SEL field. 2. Program TRGTTIML/H0 and TRGTTIML/H1 registers to define time of level change and pulse duration. 3. Program TSAUXC.PLSGx bit to define if level change or pulse is driven at Target Time on the selected SDP pin. -- If pulse generation is selected (TSAUXC.PLSGx = 1) program the TSAUXC.PLSNegx bit to define if generated pulse is positive or negative. 4. Enable level change or pulse generation by setting the TSAUXC.EN_TTx bit. Each target time register has an enable bit located in the auxiliary control register (TSAUXC.EN_TTx). When the SDP level has changed (if TSAUXC.PLSGx = 0) or when a pulse is generated (if TSAUXC.PLSGx = 1) on the selected SDP pin, the enable bit is cleared and needs to be set again by software to get another target time event. The Target Time registers (TRGTTIML/H0 or TRGTTIML/H1) can also be used for synchronizing the configurable clock output to the system time (SYSTIM). Setting the appropriate TSAUXC.STx bit to 1 enables start of clock generation only after Target Time is reached. 26.9.3.2 Configurable Frequency Clock This feature enables to generate up to 2 programmable clocks on the appropriate SDP pins by configuring the SDP pins using the TSSDP register and by programming appropriate values to the Frequency out Control registers (FREQOUT0 and FREQOUT1). The output clocks are synchronized to the global System (SYSTIM) clock and are affected by System time corrections programmed in the TIMINCA register and the TIMADJL/H registers. When clock generation is enabled, the error correction programmed in the TIMADJL/H registers is compensated from the clock output gradually, at a rate of 1 ns per 8 nS internal clock cycle. The gradual compensation is done to avoid large duty cycle variations in the output clock. Before updating the TIMADJL/H registers, software should verify that the appropriate TSICR.TADJ0/1 register bit was set to indicate that previous one time adjustment has completed. To generate either Clock 0 or Clock 1 on one of the SDP pins, the following steps should be taken: 1. Define half period of the clock by programming the CHCT field in the relevant FREQOUT0/1 register. Value placed in CHCT field is multiplied by 8 nS to get half clock cycle time. 2. Define SDPx pin that drives clock by programming the appropriate TSSDP.TS_SDPx_SEL field and setting the TSSDP.TS_SDPx_EN bit to 1 (where x is 0,1,2 or 3). 3. If clock start needs to be aligned to the system time (SYSTIM), program start of clock toggle in the appropriate Target Time (TRGTTIML0/1 and TRGTTIMH0/1) registers and set the relevant TSAUXC.ST0/1 field to 1. 4. To start clock operation, set the relevant TSAUXC.EN_CLK0/1 bit to 1. Clock out drives initially a logical 0. Clock value toggles each time a System Time duration of FREQOUT0/1 is reached or passed. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1211 Note: Clock output mechanism should be activated only after SYSTIMH/L timer is aligned to global system clock and SYSTIM timer error correction entered using the TIMADJL/H registers is below 64 s. 26.9.3.3 Time Stamp Events Upon a change in the input level of one of the SDP pins that was configured to detect Time stamp events using the TSSDP register, a time stamp of the system time is captured into one of the two auxiliary time stamp registers (AUXSTMPL/H0 or AUXSTMPL/H1). For example to define timestamping of events in the AUXSTMPL0 and AUXSTMPH0 registers, Software should: 1. Set the TSSDP.AUX0_SDP_SEL field to select the SDP pin that detects the level change and set the TSSDP.AUX0_TS_SDP_EN bit to 1. 2. Set the TSAUXC.EN_TS0 bit to 1 to enable timestamping. 26.9.4 Time SYNC Interrupts Time Sync related interrupts can be generated by programming the TSICR, TSIM and TSIS registers. The TSICR register logs the interrupt cause, the TSIM register enables masking specific TSICR bits and the TSIS register enables Software generated Time Sync interrupts. Detailed description of the Time Sync interrupt registers can be found in Section 28.16.2. Occurrence of a Time Sync interrupt sets the ICR.Time_Sync interrupt bit. 26.9.5 PTP Packet Structure The time sync implementation supports both the 1588 V1 and V2 PTP frame formats. The V1 structure can come only as UDP payload over IPv4 while the V2 can come over L2 with its Ethertype or as a UDP payload over IPv4 or IPv6. The 802.1AS uses only the layer 2 V2 format. Table 26-59. V1 and V2 PTP Message Structure Offset in Bytes V1 Fields Bits 0 1 2 3 V2 Fields 76543210 versionPTP version Network Intel(R) Communications Chipset 89xx Series - Datasheet 1212 76543210 transport Specifica messageType Reserved versionPTP message Length October 2012 Order Number: 327879-001US 26.0 Table 26-59. V1 and V2 PTP Message Structure (Continued) Offset in Bytes V1 Fields Bits V2 Fields 76543210 76543210 4 domain Number 5 Reserved 6 flags 7 8 9 10 11 Subdomain 12 correctionField 13 14 15 16 17 reserved 18 19 20 message Type 21 Source communication technology 22 23 24 Sourceuuid 25 Source Port Identity 26 27 28 source port id 29 30 sequenceId 31 sequenceId 32 control control 33 reserved Log Message Interval flags N/A 34 35 a. Should be all zero. Note: Only the fields with the bold italic format colored red are of interest to the hardware. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1213 Table 26-60. PTP Message Over Layer 2 Ethernet (L2) VLAN (Optional) PTP Ethertype PTP message UDP PTP message Table 26-61. PTP Message Over Layer 4 Ethernet (L2) IP (L3) When a PTP packet is recognized (by Ethertype or UDP port address) on the receive side then if the version is V1 then the Control field at offset 32 should be compared to the TSYNCRXCFG.CTRLT message field (see Section 28.16.1.26) otherwise the byte at offset zero should be used for comparison to the TSYNCRXCFG.MSGT field. The rest of the required fields are at the same location and size for both V1 and V2. Table 26-62. Message Decoding for V1 (Control Field at Offset 32) Enumeration Value PTP_SYNC_MESSAGE 0 PTP_DELAY_REQ_MESSAGE 1 PTP_FOLLOWUP_MESSAGE 2 PTP_DELAY_RESP_MESSAGE 3 PTP_MANAGEMENT_MESSAGE 4 reserved 5-255 Table 26-63. Message Decoding for V2 (Message Id Field at Offset 0) MessageId Message Type Value (hex) PTP_SYNC_MESSAGE Event 0 PTP_DELAY_REQ_MESSAGE Event 1 PTP_PATH_DELAY_REQ_MESSAGE Event 2 PTP_PATH_DELAY_RESP_MESSAGE Event 3 Unused Event 4-7 PTP_FOLLOWUP_MESSAGE General PTP_DELAY_RESP_MESSAGE General 9 PTP_PATH_DELAY_FOLLOWUP_MESSAGE General A PTP_ANNOUNCE_MESSAGE General B PTP_SIGNALLING_MESSAGE General C 8 PTP_MANAGEMENT_MESSAGE General D Unused General E-F If V2 mode is configured in the TSYNCRXCTL.Type field (see Section 28.16.1.1) then the time stamp should be taken on PTP_PATH_DELAY_REQ_MESSAGE and PTP_PATH_DELAY_RESP_MESSAGE according to the value in the TSYNCRXCFG.MSGT message field described in Section 28.16.1.26. Intel(R) Communications Chipset 89xx Series - Datasheet 1214 October 2012 Order Number: 327879-001US 26.0 26.10 Statistic Counters The Controller supports different statistic counters as described in Section 28.18. The statistic counters can be used to create statistic reports as required by different standards. The Controller statistic counters allow support for the following standards: * IEEE 802.3 clause 30 management - DTE section. * NDIS 6.0 OID_GEN_STATISTICS. * RFC 2819 - RMON Ethernet statistics group. * Linux Kernel (version 2.6) net_device_stats The following section describes the match between the internal GBE Controller statistic counters and the counters requested by the different standards. 26.10.1 IEEE 802.3 Clause 30 Management The Controller supports the Basic and Mandatory Packages defined in clause 30 of the IEEE 802.3 spec. The following table describes the matching between the internal statistics and the counters requested by these packages. Table 26-64. IEEE 802.3 Mandatory Package Statistics Mandatory package capability GBE Controller counter FramesTransmittedOK GPTC SingleCollisionFrames SCC MultipleCollisionFrames MCC FramesReceivedOK GPRC FrameCheckSequenceErrors CRCERRS AlignmentErrors ALGNERRC Notes and limitations The GBE Controller doesn't include flow control packets. The GBE Controller doesn't include flow control packets. In addition, part of the recommended package is also implemented as described in the following table Table 26-65. IEEE 802.3 Recommended Package Statistics Recommended package capability OctetsTransmittedOK GBE Controller counter GOTCH/GOTCL Notes and limitations The Controller counts also the DA/SA/LT/CRC as part of the octets. The Controller doesn't count Flow control packets. FramesWithDeferredXmissions DC LateCollisions LATECOL FramesAbortedDueToXSColls ECOL FramesLostDueToIntMACXmitError HTDMPC The Controller counts the excessive collisions in this counter, while 802.3 increments no other counters, while this counter is incremented CarrierSenseErrors TNCRS The Controller doesn't count cases of CRS de-assertion in the middle of the packet. OctetsReceivedOK TORL+TORH The Controller counts also the DA/SA/LT/CRC as part of the octets. Doesn't count Flow control packets. FramesLostDueToIntMACRcvError RNBC SQETestErrors N/A MACControlFramesTransmitted N/A October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1215 Table 26-65. IEEE 802.3 Recommended Package Statistics (Continued) MACControlFramesReceived N/A UnsupportedOpcodesReceived FCURC PAUSEMACCtrlFramesTransmitted XONTXC + XOFFTXC PAUSEMACCtrlFramesReceived XONRXC + XOFFRXC Part of the optional package is also implemented as described in the following table. Table 26-66. IEEE 802.3 Optional Package Statistics Optional package capability GBE Controller counter Notes MulticastFramesXmittedOK MPTC The Controller doesn't count FC packets BroadcastFramesXmittedOK BPTC MulticastFramesReceivedOK MPRC BroadcastFramesReceivedOK BPRC InRangeLengthErrors LENERRS OutOfRangeLengthField N/A FrameTooLongErrors ROC + RJC 26.10.2 The Controller doesn't count FC packets Packets parsed as Ethernet II packets OID_GEN_STATISTICS The Controller supports the part of the OID_GEN_STATISTICS as defined by Microsoft* NDIS 6.0 spec. The following table describes the matching between the internal statistics and the counters requested by this structure. Table 26-67. Microsoft* OID_GEN_STATISTICS (Sheet 1 of 2) OID entry GbE Controller counters ifInDiscards; CRCERRS + RLEC + RXERRC + MPC + RNBC + ALGNERRC ifInErrors; CRCERRS + RLEC + RXERRC + ALGNERRC ifHCInOctets; GORCL/GOTCL ifHCInUcastPkts; GPRC - MPRC - BPRC ifHCInMulticastPkts; MPRC ifHCInBroadcastPkts; BPRC ifHCOutOctets; GOTCL/GOTCH ifHCOutUcastPkts; GPTC - MPTC - BPTC ifHCOutMulticastPkts; MPTC ifHCOutBroadcastPkts; BPTC ifOutErrors; ECOL + LATECOL ifOutDiscards; ECOL ifHCInUcastOctets; N/A ifHCInMulticastOctets; N/A ifHCInBroadcastOctets; N/A Intel(R) Communications Chipset 89xx Series - Datasheet 1216 Notes October 2012 Order Number: 327879-001US 26.0 Table 26-67. Microsoft* OID_GEN_STATISTICS (Sheet 2 of 2) ifHCOutUcastOctets; N/A ifHCOutMulticastOctets; N/A ifHCOutBroadcastOctets; N/A 26.10.3 RMON The Controller supports the part of the RMON Ethernet statistics group as defined by IETF RFC 2819. The following table describes the matching between the internal statistics and the counters requested by this group. Table 26-68. RMON Statistics RMON statistic GBE Controller counters etherStatsDropEvents MPC + RNBC etherStatsOctets TOTL + TOTH etherStatsPkts TPR etherStatsBroadcastPkts BPRC etherStatsMulticastPkts MPRC etherStatsCRCAlignErrors CRCERRS + ALGNERRC etherStatsUndersizePkts Notes The Controller doesn't count FC packets RUC etherStatsOversizePkts ROC etherStatsFragments RFC Should count bad aligned fragments as well etherStatsJabbers RJC Should count bad aligned jabbers as well etherStatsCollisions COLC etherStatsPkts64Octets PRC64 RMON counts bad packets as well etherStatsPkts65to127Octets PRC127 RMON counts bad packets as well etherStatsPkts128to255Octets PRC255 RMON counts bad packets as well etherStatsPkts256to511Octets PRC511 RMON counts bad packets as well etherStatsPkts512to1023Octets PRC1023 RMON counts bad packets as well etherStatsPkts1024to1518Octets PRC1522 RMON counts bad packets as well October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1217 26.10.4 Linux net_device_stats The Controller supports part of the net_device_stats as defined by Linux Kernel version 2.6 (defined in <linux/netdevice.h>). The following table describes the matching between the internal statistics and the counters requested by this structure. / Table 26-69. Linux net_device_stats net_device_stats field GBE Controller counters Notes rx_packets GPRC The Controller doesn't count flow controls - can be accounted for by using the XONRXC and XOFFRXC counters tx_packets GPTC The Controller doesn't count flow controls - can be accounted for by using the XONTXC and XOFFTXC counters rx_bytes GORCL + GORCH tx_bytes GOTCL + GOTCH rx_errors CRCERRS + RLEC + RXERRC + ALGNERRC tx_errors ECOL + LATECOL rx_dropped N/A tx_dropped N/A multicast MPTC collisions COLC rx_length_errors RLEC rx_over_errors N/A rx_crc_errors CRCERRS rx_frame_errors ALGNERRC rx_fifo_errors HRMPC rx_missed_errors MPC tx_aborted_errors ECOL tx_carrier_errors N/A tx_fifo_errors N/A tx_heartbeat_errors N/A tx_window_errors LATECOL rx_compressed N/A tx_compressed N/A Intel(R) Communications Chipset 89xx Series - Datasheet 1218 October 2012 Order Number: 327879-001US 26.0 Figure 26-34. Total Packet Received TPR L2 filtering Error packets SYMERRS SCVPC ALGNERRC CRCERRS LENERRS Errors Flow control detection Flow control packets XONRXC XOFFRXC M issed packet MPC Packet buffer full Good packets received GPRC M anageability filtering VM DQ switch Pool/Storm control Queues M anageability only packets M NGPRC Packets dropped by Switch SDPC Drop of queues S(RQDPC[15:0]) Packets received by host RPTHC October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1219 27.0 GbE Platform Manageability 27.1 Platform Configurations This section describes the hardware configurations for platform management. It describes the partitioning of platform manageability among system components and the functionality provided by the GbE Controller in each of the platform configurations. The Controller interfaces with an on-board BMC component for basic manageability functionality. The link between the Controller and the BMC is SMBus. Note: Link configuration is loaded from the "Redirection LAN Interface" EEPROM bits in the Manageability Capability/Manageability Enable word (word 0x54 - see Section 24.7.5, "Manageability Capability/Manageability Enable (Word 0x54)"). 27.1.1 On-Board BMC Configurations The following figure shows a SMBus-only connection between the Controller and the BMC. All communication between the Controller and the BMC (pass-through traffic, configuration, and status) pass through a single interface. The protocol details for this configuration are according to the SMBus commands described in Section 27.2.1. Figure 27-1. Controller to BMC Connectivity Through a SMBus-only Connection CPU X 1 P C Ie P C Ie BMC SM B us In te g ra te d GbE C o n tro lle r See following sections for description of traffic types that use the SMBus interfaces. Intel(R) Communications Chipset 89xx Series - Datasheet 1220 October 2012 Order Number: 327879-001US 27.0 27.1.2 GbE Controller The GbE Controller on a PCIe I/F may be remotely configured through the SMBus interface. 27.2 Pass Through Functionality The Controller supports traffic pass through to an external BMC. The usable bandwidth for either direction is up to 400 Kbps in SMBus mode. The following list describes the different flows that can make up pass through traffic: * BMC management traffic * Keyboard or mouse traffic for KVM (low data rate) * Video traffic for KVM (low average rate of 150 Kbytes/s to 200 Kbytes/s) - transmit only * USB 2.0 redirect (up to 50 Mb/s) * IDE redirect for remote CD/floppy (rate - priority 1 - CDx7 = 1.05 Mb/s. Priority 2 CDx24 = 64 Mb/s * Serial Over LAN (SoL) - 300 Kbytes/s Pass through traffic is carried through a SMBus (legacy devices) based on the configuration loaded from the EEPROM. The management interface does not change dynamically and is loaded only during a the Controller power up. 27.2.1 SMBus Pass Through (PT) Functionality In addition to carrying pass through traffic, in SMBus mode the Controller enables loading default configuration of filters by EEPROM. When working in SMBus mode the default values of the manageability receive filters can be set according to the PT LAN structure defined in Chapter 24.0, "Non-Volatile Memory Map - EEPROM". 27.2.1.1 Pass Through (PT) Modes In pass through mode the manageability traffic is handled by the BMC. The Controller is presented on the manageability link as four different devices (via different SMBus addresses) where each device is connected to a different LAN port. There is no logical connection between the devices. Fail-over if required between the LAN ports is done by the external BMC (by sending/receiving packets through different ports). The status reports to the BMC, ARP handling, DHCP and other pass through functionality are unique for each port). 27.3 Programming Interfaces 27.3.1 SMBus Programming This section describes the SMBus transactions supported in Advanced Pass Through (APT) mode. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1221 27.3.1.1 Write SMBus Transactions (BMC to GbE Controller) Table 27-1 summarizes the SMBus write transactions supported by the Controller. Table 27-1. SMBus Write Transactions TCO Command 27.3.1.1.1 Transaction Command First: Middle: Last: Transmit Packet Block Write Transmit Packet Block Write Single: Receive Enable Block Write Single: Management Control Block Write Single: Update MNG RCV filter parameters Block Write 0x84 0x04 0x44 Fragmentation Section Multiple 27.3.1.1.1 0xC4 Single 27.3.1.1.1 0xCA Single 27.3.1.2 0xC1 Single 27.3.1.2.2 Single: 0xCC Single 27.3.1.2.3 Force TCO Block Write Single: 0xCF Single 27.3.1.2.1 Request Status Block Write Single: 0xDD Single 27.3.1.1.2 Transmit Packet Command The Transmit Packet command behavior is detailed in section 22.2.1.1.4. The Transmit Packet fragments have the following format. Function Command Transmit first fragment 0x84 Transmit middle fragment 0x04 Transmit last fragment 0x44 Transmit single fragment 0xC4 Byte Count N Data 1 Packet data MSB ... ... Data N Packet data LSB The payload length is limited to the maximum payload length set in the EEPROM. If the overall packet length is larger than 1536 bytes or smaller than 32 bytes (including padding), then the packet is discarded by the Controller and Abort is asserted. 27.3.1.1.2 Request Status Command The TCO controller can initiate a request to read the Controller manageability status by sending this command. When it receives this command, the Controller initiates a notification to the external controller (when it is ready with the status), and then the external controller will be able to read the status, by issuing a read status command (see section 27.3.1.3.3). Request Status Command format: Function Command Byte Count Data 1 Request status 0xDD 1 0 Intel(R) Communications Chipset 89xx Series - Datasheet 1222 October 2012 Order Number: 327879-001US 27.0 27.3.1.2 Receive Enable Command The Receive Enable command is a single fragment command that is used to configure the Controller. This command has two formats: * Short: 1-byte legacy format (providing backward compatibility with previous components). * Long: 14-byte advanced format (allowing greater configuration capabilities). Note: If the receive enable command is short and thus does not include all the parameters, then the parameters will be taken from most recent previous configuration (either the most recent long Receive-Enable command in which the particular value was set, or the EEPROM if there was no such previous long Receive-Enable command). Func. Legacy receive enable Advance d receive enable Byte Coun t Cmd 1 0xCA 14 0x0E Data 1 Receive control byte Data 2 ... MAC addr. MSB ... Data 7 Data 8 - - MAC addr. LSB ... Data 11 Data 12 Data 13 Data 14 - - - - RSV BMC SMBus addr. Interf. data byte Alert value byte ... RSV Where: * Receive control byte (data byte 1): Field Bit(s) Description Receive TCO enable. 0b = Disable Receive and Transmit of TCO packets 1b = Enable Receive and Transmit of TCO packets. Setting this bit will enable BMC to network and network to BMC traffic and all manageability receive filtering operations (Will set bit MANC.RCV_TCO_EN). The enable of the specific filtering is done through a special configuration command (see Section 27.3.1.2.3). Notes: 1. When this bit is cleared all receive TCO functionality is disabled, not just the packets that are directed to the BMC. 2. Bit can also be set using the Enable TCO to Network bit in the Update MNG RCV Filter Parameters command (Parameter Filter Enable). RCV_EN 0 Reserved 1 Reserved EN_STA 2 Enable Status reporting to the BMC, on status change. 0b = Disable status reporting 1b = Enable status reporting Reserved 3 Reserved, must be 0b. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1223 Field Bit(s) Description NM 4-5 Notification Method. Define the notification method the Controller will use. 00b = SMBus Alert 01b = Asynchronous Notify 10b = Direct Receive 11b = Not Supported. Note: In multiple SMBus address mode, all SMBus addresses must be configured to the same notification method. Reserved 6 Reserved, must be 1b. 7 Configure BMC dedicated MAC Address. Note: This bit should be zero when the RCV_EN bit (bit 0) is not set. 0b = The Controller will share MAC address for MNG traffic with host MAC address which is specified in EEPROM words 0x0-0x2. 1b = The Controller will use the BMC dedicated MAC address as filter for incoming receive packets. The BMC MAC address is set in bytes 2-7 in this command. If short version of the command is used, the Controller will use MAC address configured in the most recent long version of the command in which the CBDM bit was set. If no such previous long command exists, then the Controller will use MAC address 1 (Mac address programmed in MMAL/H1 registers). When Dedicated MAC address feature is activated, the Controller will use the following registers to filter in all the traffic addressed to the BMC MAC. BMC should not modify these registers: Manageability Decision Filter - MDEF7 (and corresponding bit 7 in Management Only traffic Register - MNGONLY) Manageability MAC Address Low - MMAL1 Manageability MAC Address High - MMAH1 CBDM * MNG MAC address (data bytes 2-7) Ignored if CBDM bit is not set. This MAC address will be used for configuration of the dedicated MAC address. This MAC address will continue to be used when CBDM bit is set in subsequent short versions of this command. * Reserved (data bytes 8-11). * Asynchronous Notification SMBus address (data byte 12) This address will be used for the Asynchronous Notification SMBus transaction and for Direct Receive. * Interface data (data byte 13) Interface data byte to be used in Asynchronous Notification. * Alert data (data byte 14). Alert Value data byte to be used in the Asynchronous Notification. 27.3.1.2.1 Force TCO Command Depending on the bit set in the TCO mode field this command will cause the Controller to perform either: 1. TCO Reset, if Force TCO reset is enabled in the EEPROM. The Force TCO reset will clear the data-path (RX/TX) of the Controller to enable the BMC to transmit/receive packets through the Controller. -- The Controller will consider Force TCO reset command as an indication that the OS is hung and will clear the DRV_LOAD flag. If TCO reset is disabled in EEPROM, the Controller clears the CTRL_EXT.DRV_LOAD bit but does not reset the data-path. -- Following TCO reset management sets MANC.TCO_RESET to 1. Intel(R) Communications Chipset 89xx Series - Datasheet 1224 October 2012 Order Number: 327879-001US 27.0 2. TCO Isolate, If TCO Isolate is enabled in the EEPROM. The TCO Isolate command will disable PCIe write operations to the LAN port. -- If TCO Isolate is disabled in the EEPROM, the controller will does not execute the command but sends a response to the BMC with a successful completion. -- Following TCO Isolate management sets MANC.TCO_Isolate to 1. 3. Firmware Reset. This command will cause re-initialization of all the manageability functions and re-load of manageability related EEPROM words (e.g. Firmware patch code). -- On reception of Firmware reset command management sets MANC.FW_RESET to 1 and increments the FWSM.Reset_Cnt field. Note: Force TCO reset will effect only the port related to the SMBus address the command was issued to. Following firmware reset, BMC will need to re-initialize all ports. Force TCO Command format: Function Command Byte Count Data 1 Force TCO 0xCF 1 TCO mode Where TCO Mode field controls the following operations: Field Bit(s) Description Do TCO reset. 0b = Do Nothing 1b = Perform TCO Reset Note: Setting this bit generates a one time LAN port reset event. DO_TCO_RSTa 0 Reserved 1 Reserved Firmware Reset 2 Reset Manageability and re-load Manageability related EEPROM words. 0b = Do Nothing 1b = Issue Firmware Reset to manageability. Note: Setting this bit generates a one time Firmware reset event. Following Firmware Reset Management related data from EEPROM is loaded. Reserved 3-7 Reserved (Set to 0x00). a. TCO Reset operation enabled in EEPROM. 27.3.1.2.2 Management Control This command is used to set generic manageability parameters. The parameters list is shown in the table below. The command is C1h, which states that it is a management control command. The first data byte is the parameter number and the data after-words (length and content) are parameter specific as shown in the table. Note: If in the update configuration, the parameter that the BMC sets is not supported by the Controller, the Controller will not NACK the transaction. After the transaction ends, the Controller will discard the data and will assert a transaction abort status. This is the format of the Management control command: October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1225 27.3.1.2.3 Function Command Byte Count Data 1 Management control 0xC1 N Parameter number Data 2 ... Data N Parameter dependent Update MNG RCV Filter Parameters This command is used to set the manageability receive filters parameters. The parameters list is shown in the table below. The command is 0xCC, which states that it is a parameter update. The first data byte is the parameter number and the data after-words (length and content) are parameter specific as shown in the table. If in the update configuration, the parameter that the BMC sets is not supported by the Controller, it will not NACK the transaction. After the transaction ends, the Controller will discard the data and will assert a transaction abort status. Detailed description of receive filtering capabilities and configuration is in Section 27.4. This is the format of the Update MNG RCV filter parameters command: Function Command Byte Count Data 1 Update MNG RCV filter parameters 0xCC N Parameter number Data 2 ... Data N Parameter dependent The table below shows the different parameters and their contents. Parameter # Parameter Data Filters Enable 0x1 Defines generic filters configuration. The structure of this parameter is 4 bytes. Where Data 2 is the MSB and Data 5 is the LSB. The parameter updates the MANC register. The general filter enable is in the receive enable command, which enables receive filtering. This parameter specifies which filters should be enabled. ARP filtering and dedicated MAC address can also be enabled through the receive enable command (see Section 27.3.1.2). MNGONLY configuration 0xF This parameter defines which of the packets types identified as manageability packets in the receive path will never be directed to the host memory. Data 2:5: MNGONLY register bytes - Data 2 is the MSB 0x10 Flex filter 0 mask. This parameter defines the Mask bits in the FTFT filter. Data 2:17 - MASK. Bit 0 in data 2 is the first bit of the MASK Data 18:19 - Reserved. Should be zero. Data 20 - Flexible Filter length. Must be greater than or equal one. 0x11 Data 2 - Group of Flex filter's bytes. This parameter defines the Data pattern bytes in the FTFT filter: 0x0 = bytes 0-29 0x1 = bytes 30-59 0x2 = bytes 60-89 0x3 = bytes 90-119 0x4 = bytes 120-127 Data 3:32 - Flex filter data bytes. Data 3 is LSB. Group's length is not mandatory 30 bytes; it may vary according to filter's length and must NOT be padded by zeros. 0x62 Three bytes to load the VLAN tag filters (MAVTV) Data 2 - VLAN Filter number Data 3 - MSB of VLAN Filter Data 4 - LSB of VLAN Filter Flex filter 0 enable MASK and length Flex filter 0 data VLAN Filters Intel(R) Communications Chipset 89xx Series - Datasheet 1226 October 2012 Order Number: 327879-001US 27.0 Parameter Flex Ports Filters IPv4 Filters IPv6 Filters MAC Filters Ethertype Filters Extended Decision Filter 27.3.1.3 # Parameter Data 0x63 Three bytes to load the manageability flex port filters (MFUTP). Data 2 - Flex Port Filter number Data 3 - MSB of flex port filter Data 4 - LSB of flex port filter 0x64 Five bytes to load the IPv4 address filter (MIPAF, DW 15:12) Data 2 - IPv4 address filter number (0-3) Data 3 - LSB of IPv4 address filter ... Data 6 - MSB of IPv4 address filter 0x65 17 bytes to load IPv6 address filter (MIPAF) Data 2 - IPv6 address filter number (0-3) Data 3 - LSB of IPv6 address filter ... Data 18 - MSB of IPv6 address filter 0x66 Seven bytes to load MAC address filters (MMAL, MMAH) Data 2 - MAC address filters pair number (0-1) Data 3 - MSB of MAC address ... Data 8 - LSB of MAC address 0x67 Five bytes to load Ethertype filters (METF) Data 2 - METF filter index (valid values are 0,) Data 3 - MSB of METF ... Data 6 - LSB of METF 0x68 Nine bytes to load the extended decision filters (MDEF_EXT & MDEF) Data 2 - MDEF filter index (valid values are 0...6) Data 3 - MSB of MDEF_EXT (DecisionFilter1) .... Data 6 - LSB of MDEF_EXT (DecisionFilter1) Data 7 - MSB of MDEF (DecisionFilter0) .... Data 10 - LSB of MDEF (DecisionFilter0) The command shall overwrite any previously stored value. Read SMBus Transactions (Controller to BMC) The table below summarizes the different SMBus read transactions supported by the Controller. The read transactions are compatible with the SMBus Read Block Protocol format. TCO Command Transaction Command Op-Code Fragmentation Section Receive TCO Packet Block Read 0xC0 or 0xD0 First: Middle: Lasta 0x90 0x10 0x50 Multiple 27.3.1.3.1 Read configuration Block Read 0xC6 + 0xD2 Single: 0xD2 Single 27.3.1.3.5 Read Receive Enable configuration Block Read 0xDA Single: 0xDA Single 27.3.1.3.7 Read the Controller Status Block Read 0xC0 or 0xD0 or 0xDE Single: 0xDD Single 27.3.1.3.3 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1227 TCO Command Transaction Command Op-Code Fragmentation Section Read Management parameters Block Read 0xD1 Single: 0xD1 Single 27.3.1.3.5 Read MNG RCV filter parameters Block Read 0xCD Single: 0xCD Single 27.3.1.3.6 Get system MAC address Block Read 0xD4 Single 0xD4 Single 27.3.1.3.4 a. Last fragment of the receive TCO packet is the packet status. Notes: 27.3.1.3.1 0xC0/0xD0 commands are used for more than one payload. If the BMC issues these read commands and the Controller has no pending data to transfer, it will always return as default opcode 0xDD with the Controller status, and will not NACK the transaction. Receive TCO LAN Packet Transaction The BMC uses this command to read the packet received on the LAN and its status. When the Controller has a packet to deliver to the BMC, it asserts the SMBus notification, for the BMC to read the data. Upon receiving notification of the arrival of LAN receive packet, the BMC should begin issuing a Receive TCO packet command using the block read protocol. The packet can be delivered in more than one SMBus fragment (at least two - one for the packet, and the other one for the status), and the BMC should follow the "F" and "L" bits (2 MSB bits of the op-code). The op-code can have these values: * 0x90 - First Fragment * 0x10 - Middle Fragment. * 0x50 - Packet status (last fragment) as described below in Section 27.3.1.3.2. The Receive packet will be silently discarded if following packet reception and notification no valid BMC transaction on the SMBus is detected (no receive fragment is read by the BMC, no Status Read operation is executed by the BMC) within a timeout period. The timeout period is set according to the SMBus notification timeout EEPROM parameter. Function Command Receive TCO packet 0xC0 or 0xD0 Data returned from the Controller: Function Byte Count Receive TCO First Fragment Receive TCO Middle Fragment Data 2 ... Data N 90 N Receive TCO Last Fragment (TCO status, see Section 27.3.1.3.2) Intel(R) Communications Chipset 89xx Series - Datasheet 1228 Data 1 (Op-Code) 10 Packet Data Byte ... Packet Data Byte 50 October 2012 Order Number: 327879-001US 27.0 27.3.1.3.2 Receive TCO LAN Status Payload Transaction This is the last transaction that the Controller issues when a packet that was received from the LAN is transferred to the BMC. The transaction contains the status of the received packet. This is the format of the status transaction: Function Receive TCO long status Byte Count Data 1 (Op-Code) Data 2 - Data 17 (Status data) 0x50 See below 17 (0x11) The status is 16 bytes where byte 0 (bits 0-7) is set in Data 2 of the status and byte 15 is set in Data 17 of the status. The table below describes the content of the status data: Table 27-2. TCO LAN Packet Status Data Name Bits Description Packet Length 13:0 Packet length including CRC, only 14 LSB bits. Packet status 35:14 See Table 27-3 Reserved 42:36 Reserved Error 47:43 See Table 27-4 VLAN 63:48 The two bytes of the VLAN header tag. Reserved 67:64 Reserved Packet type 80:68 See Table 27-6 Reserved 84:81 Reserved MNG status 127:85 See Table 27-7. This field should be ignored if Receive TCO is not enabled, The meaning of the bits inside of each field is in Section 26.1.5. Table 27-3. Packet Status Info Field Bit(s) Description LAN# 21:20 Indicates the source port of the packet VP 19 VLAN Stripped -insertion of VLAN TAG is needed. VEXT 18 Additional VLAN present in packet Reserved 17:15 Reserved Reserved 13:12 Reserved CRC stripped 11 Insertion of CRC is needed. Reserved 10:6 Reserved UDPV 5 UDP checksum valid Reserved 4:3 Reserved IPCS 2 Ipv4 Checksum Calculated on packet L4I 1 L4 (TCP/UDP) Checksum calculated on packet UDPCS 0 UDP checksum calculated on packet October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1229 Table 27-4. Error Status Info Field Bit(s) Description RXE 4 RX Data Error IPE 3 Ipv4 Checksum Error L4E 2 L4 (TCP/UDP) Checksum Error Reserved 1:0 Reserved Table 27-6. Packet Type Bit Index Bit 11 = 0b 12 Bit 11 = 1b (L2 packet) VLAN packet indication 11 Packet matched one of the ETQF filters. 9:8 Reserved Reserved 7 NFS - NFS header present 6 SCTP - SCTP header present 5 UDP - UDP header present 4 TCP - TCP header present Reserved 3 IPV6E- IPv6 Header includes extensions 2 IPV6 - IPv6 header present 1 IPV4E - IPv4 Header includes extensions 0 IPV4 - IPv4 header present EtherType - ETQF register index that matches the packet. Special types might be defined for 1588, 802.1X, or any other requested type. Table 27-7. MNG Status Name Decision Filter match Description 42:35 Set when there is a match to one of the Decision filters IPv4/IPv6 match 34 Set when there is an IPv4 match or IPv6 match. This bit is valid only if the bit 30 (IP match bit) or bit 4 (ARP match bit) are set. IP address match 33 Set when there is a match to any of the IP address filters 32:31 Set when there is a match to the IP filter number. (IPv4 or IPv6) IP address Index Flex TCO filter match 30 Set when there is a match to the Flex port filter Reserved 29:27 Reserved L4 port match 26 Set when there is a match to any of the UDP / TCP port filters L4 port Filter Index 25:19 Indicate the flex filter number Unicast Address match 18 Set when there is a match to any of the 2 Unicast MAC addresses. Unicast Address Index 17:15 Indicates which of the 2 Unicast MAC addresses match the packet. Valid only if the Unicast Address match is set. MNG VLAN Address Match 14 Set when the MNG packet matches one of the MNG VLAN filters Pass MNG VLAN Filter Index 13:11 Indicates which of the Vlan filters match the packet. Reserved 10:8 Reserved 7 Set when the MNG packet is an ARP response/request packet Pass ARP req / ARP resp Intel(R) Communications Chipset 89xx Series - Datasheet 1230 Bits October 2012 Order Number: 327879-001US 27.0 Table 27-7. MNG Status (Continued) Name 27.3.1.3.3 Bits Description Pass MNG neighbor 6 Set when the MNG packet is a neighbor discovery packet. Pass MNG broadcast 5 Set when the MNG packet is a broadcast packet Pass RMCP 0x0298 4 Set when the UDP/TCP port of the MNG packet is 0x298 Pass RMCP 0x026F 3 Set when the UDP/TCP port of the MNG packet is 0x26F Manageability Ethertype filter passed 2 Indicates that one of the METF filters matched manageability Ethertype filter index 1:0 Indicates which of the METF filters matched Read Status Command The BMC can read the Controller status. The Controller will assert an alert prior to the BMC reading the status bytes. There can be two reasons for the Controller to send status to the BMC: * The external BMC asserts a request for reading the Controller status (Section 22.2.1.1.2). * The Controller detects a change in one of the "Status Data 1" bits, and was set to send status to the BMC, on status change, in the receive enable command (Section 22.2.1.1.2). Commands C0h/D0h are for backward compatibility. 0xD0/0xC0 can be used for other payloads the Controller will define in the op-code, which payload this transaction will be. When 0xDE command is set, the Controller will always return opcode 0xDD with the Controller status. The BMC will read the event causing the notification, using the read status command shown below: Function Read status Command 0xC0 or 0xD0 or 0xDE Function Byte Count Data 1 (Op-Code) Receive TCO partial status 3 0xDD Data 2 (Status Data 1) Data 3 (Status Data 2) See below The table below shows Status Data 1 byte: Bit 7 6 Name LAN Port Lsb TCO command aborted October 2012 Order Number: 327879-001US Description When Lan Port Msb is 0b = Status came 1b = Status came When Lan Port Msb is 0b = Status came 1b = Status came 0: from from 1: from from LAN port 0. LAN port 1. LAN port 2. LAN port 3. 0b = A TCO command abort event has not occurred since the last Read Status cycle. 1b = A TCO command abort event has occurred since the last Read Status cycle. See Section 22.2.1.1.5 for command abort flow. Intel(R) Communications Chipset 89xx Series - Datasheet 1231 5 Link Status indication 4 Reserved 3 Initialization indication 0b = LAN link down 1b = LAN link is up 0b = An EEPROM reload event has not occurred since the last Read Status cycle. 1b = An EEPROM reload event has occurred since the last Read Status cycle. See note 1. Defines together with LAN Port Lsb the port that sent the Status: 2 1:0 LAN Port Msb Lan Port Msb 0 0 1 1 00b 01b 10b 11b Power state = = = = Lan Port Lsb 0 1 0 1 Status Status Status Status came came came came from from from from LAN LAN LAN LAN port port port port 0. 1. 2. 3. Dr state D0u state D0 state D3 state Notes: 1. This indication will be asserted when the Controller manageability block will reload EEPROM and its internal data-base will be updated to EEPROM default values. This is an indication that the external BMC should re-configure the Controller, if other values beside the EEPROM default should be configured. Status data byte 2 is used by the BMC as an indication whether the LAN driver is alive and running. The driver valid indication is a bit that is set by the driver when it is coming up, and cleared when it goes down to Dx state or cleared by the HW on PCI reset. Bits 2 and 1 indicate that the LAN driver is not stuck. Bit 2 indicates whether the interrupt line of the LAN function is asserted, and bit 1 indicates whether driver between the last Read Status Cycle dealt the interrupt line. The table below shows Status Data 2 byte: Bit Name 7 Reserved Reserved 6 Reserved Reserved 5 Reserved Reserved 4 Reserved Reserved 3 Driver Valid indication 0b = LAN driver is not alive 1b = LAN driver is alive 2 Interrupt pending indication 0b = LAN interrupt is not asserted. 1b = LAN interrupt is asserted. 1 ICR register read/write 0b = ICR register was not read since the last Read Status Cycle. 1b = ICR register was read since the last Read Status cycle. Reading the ICR means that the driver has dealt with the interrupt that was asserted 0 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1232 Description October 2012 Order Number: 327879-001US 27.0 The table below shows the possible values of bits 2, 1 and what the BMC can assume based on their value: Previous Current Don't care 00b Interrupt is not pending - O.K 00b 01b New interrupt is asserted - O.K 10b 01b New interrupt is asserted - O.K 11b 01b Interrupt is waiting for reading - O.K 01b 01b Interrupt is waiting for reading by driver more than one read status Cycle - Not OK (possible driver hang state). Don't Care 11b Previous interrupt was read and current interrupt is pending - O.K Don't Care 10b Interrupt is not pending - O.K Note: The BMC reads should consider the time it takes for the driver to deal with the interrupt (few micro-seconds), too frequent reads will give false indications. 27.3.1.3.4 Get System MAC Address The Get System MAC Address will return system MAC Address over the SMBus. This command is a single fragment Read Block transaction, which will return the following database: The Get System MAC Address returns the system MAC Address (RAL0, RAH0). Get System MAC Address format: Function Command Get system MAC address 0xD4 Data returned from the Controller: 27.3.1.3.5 Function Byte Count Data 1 (Op-Code) Data 2 ... Data 7 Get system MAC address 7 0xD4 MAC address MSB ... MAC address LSB Read Management Parameters To read the management parameters the BMC should execute two SMBus transactions: a block write that sets the parameter that the BMC wants to read, and then a block read that reads the parameter. This is the block write transaction: Function Management control Request October 2012 Order Number: 327879-001US Command 0xC1 Byte Count 1 Data 1 Parameter number Intel(R) Communications Chipset 89xx Series - Datasheet 1233 Following the block write the BMC should issue a block read that will read the parameter that was set in the block write command: Function Command Read management parameter 0xD1 Data returned from the Controller: Function Byte Count Data 1 (Op-Code) Data 2 Read management parameter N 0xD1 Parameter number Data 3 ... Data N Parameter dependent The returned data is as follows: Parameter # Parameter Data Reserved 0x00 A single byte parameter: Data 3 Bit [7:0] Reserved Wrong parameter request 0xFE Returned by the Controller only. This parameter is returned on read transaction, if in the previous read command the BMC sets a parameter that is not supported by the Controller. Controller ot ready 0xFF Returned by the Controller only, on read parameters command when the data that should have been read is not ready. This parameter has no data. The BMC should retry the read transaction. Notes: The parameter that is returned might not be the parameter requested by the BMC. The BMC should verify the parameter number (default parameter to be returned is 1h). It is BMC's responsibility to follow the procedure defined above. If the BMC sends a block read command (as described above) that is not preceded by a block write command with bytecount=1, the Controller will set the Parameter Number in the read block transaction to 0xFE. 27.3.1.3.6 Read MNG RCV Filter Parameters To read the MNG RCV filter parameters, the BMC should execute two SMBus transactions: a block write that sets the parameter that the BMC wants to read, and then a block read that reads the parameter. This is the block write transaction: Function Command Byte Count Data 1 Data 2 MNG RCV filter parameters to read 0xCC 1 or 2 Parameter number Parameter data Intel(R) Communications Chipset 89xx Series - Datasheet 1234 October 2012 Order Number: 327879-001US 27.0 The table below shows the different parameters and their contents: Parameter # Parameter Data Filters Enable 0x1 None MNGONLY configuration 0xF None Flex filter 0 enable MASK and length 0x10 None Flex filter 0 data 0x11 Data 2 - Group of Flex filter's bytes: 0x0 = bytes 0-29 0x1 = bytes 30-59 0x2 = bytes 60-89 0x3 = bytes 90-119 0x4 = bytes 120-127 VLAN Filters 0x62 One byte to define the accessed VLAN tag filter (MAVTV) Data 2 - VLAN Filter number Flex Ports Filters 0x63 One byte to define the accessed manageability flex port filter (MFUTP). Data 2 - Flex Port Filter number IPv4 Filter 0x64 One byte to define the accessed IPv4 address filter (MIPAF) Data 2 - IPv4 address filter number IPv6 Filters 0x65 One byte to define the accessed IPv6 address filter (MIPAF) Data 2 - IPv6 address filter number MAC Filters 0x66 One byte to define the accessed MAC address filters pair (MMAL, MMAH) Data 2 - MAC address filters pair number (0-1) Ethertype Filters 0x67 One byte to define Ethertype filters (METF) Data 2 - METF filter index (valid values are 0,1,2 and 3.) Extended Decision Filter 0x68 One byte to define the extended decision filters (MDEF_EXT & MDEF) Data 2 - MDEF filter index (valid values are 0...6) Wrong parameter request 0xFE Returned by the Controller only. This parameter is returned on read transaction, if in the previous read command the BMC sets a parameter that is not supported by the Controller. Controller is not ready 0xFF Returned by the Controller only, on read parameters command when the data that should have been read is not ready. This parameter has no data. Following the block write the BMC should issue a block read that will read the parameter that was defined in the block write command: Function Command Request MNG RCV filter parameters 0xCD Data returned from the Controller: Function Byte Count Data 1 (Op-Code) Data 2 Read MNG RCV filter parameters N 0xCD Parameter number Data 3 ... Data N Parameter dependent The returned data is in the same format of the "update" command 1. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1235 Note: If the returned parameter is not the parameter that the BMC requested, then the BMC should verify the parameter number (default parameter to be returned is 1h). Note: If the parameter number is 0xFF, it means that the data that the Controller should supply is not yet ready. The BMC should retry the read transaction. Note: It is BMC's responsibility to follow the above procedure. If the BMC sends a block read command (as described above) that is not preceded by a block write command, then the Controller will set Parameter Number in the read block transaction to be 0xFE. 27.3.1.3.7 Read Receive Enable Configuration The BMC uses this command to read the receive configuration data. This data can be configured in receive enable command or through EPROM loading upon power-up. Read Receive Enable Configuration command format (SMBus Read Block Protocol): Function Command Read receive enable 0xDA Data returned from the Controller. Function Byte Coun t Data 1 (Op-C ode) Data 2 Data 3 ... Data 8 Data 9 ... Data 12 Data 13 Data 14 Data 15 Read receive enable 15 (0x0F ) 0xDA Receiv e contro l byte MAC addres s MSB ... MAC addres s LSB IP addres s MSB ... IP addres s LSB BMC SMBus address Interfac e data byte Alert value byte The detailed description of each field is specified in the receive enable command description in Section 27.3.1.2. 27.3.1.4 SMBus ARP Transactions All SMBus-ARP transactions include PEC byte. 27.3.1.4.1 Prepare to ARP This command will clear the Address Resolved flag (set to false). It will not affect the status or validity of the dynamic SMBus Address (will not clear the address Valid flag). It is used to inform all devices that the ARP Master is starting the ARP process: 27.3.1.4.2 1 7 1 1 8 1 8 1 1 S Slave Address Wr A Command A PEC A P 1100 001 0 0 0000 0001 0 [Data dependent value] 0 Reset Device (General) This command will clear the Address Resolved flag (set to false). It will not affect the status or validity of the dynamic SMBus Address (will not clear the address Valid flag). Intel(R) Communications Chipset 89xx Series - Datasheet 1236 October 2012 Order Number: 327879-001US 27.0 27.3.1.4.3 1 7 1 1 8 1 S Slave Address Wr 1100 001 0 8 1 1 A Command 0 0000 0010 A PEC A P 0 [Data dependent value] 0 Reset Device (Directed) The Command field is NACK-ed if the bits 7 through 1 do not match the current the Controller SMBus address. It clears the Address Resolved flag (set to false). It will not affect the status or validity of the dynamic SMBus Address (will not clear the address Valid flag). 27.3.1.4.4 1 7 1 1 8 1 8 1 1 S Slave Address Wr A Command A PEC A P 1100 001 0 0 Targeted slave address | 0 0 [Data dependent value] 0 Assign Address This command assigns the Controller SMBus address. The address and command bytes are always acknowledged. The transaction is aborted immediately (NACK-ed) if any of the UDID bytes differ from the Controller UDID bytes as defined in Section 22.2.1.1.7. If successful, the MNG will update the SMBus address internally. This command will also set the Address Resolved flag to true. 1 S 7 1 1 8 1 8 1 Slave Address Wr A Command A Byte Count A 1100 001 0 0 0000 0100 0 0001 0001 0 *** 8 1 8 1 8 1 8 1 Data-1 A Data-2 A Data-3 A Data-4 A UDID byte 15 (MSB) 0 UDID byte 14 0 UDID byte 13 0 UDID byte 12 0 8 1 8 1 8 1 8 1 Data-5 A Data-6 A Data-7 A Data-8 A UDID byte 11 0 UDID byte 10 0 UDID byte 9 0 UDID byte 8 0 8 1 8 1 8 1 Data-9 A Data-10 A Data-11 A UDID byte 7 0 UDID byte 6 0 UDID byte 5 0 October 2012 Order Number: 327879-001US *** *** *** Intel(R) Communications Chipset 89xx Series - Datasheet 1237 27.3.1.4.5 8 1 8 1 Data-12 UDID byte 4 8 1 A Data-13 0 UDID byte 3 8 1 A Data-14 0 UDID byte 2 A Data-15 A 0 UDID byte 1 0 *** 8 1 8 1 8 1 1 Data-16 A Data-17 A PEC A P UDID byte 0 (LSB) 0 Assigned Address 0 [Data dependent value] 0 Get UDID (General and Directed) The Get UDID command depends on whether this is a directed or general command. The General Get UDID SMBus transaction supports a constant command value of 0x03. The Directed Get UDID SMBus transaction supports a dynamic command value equal to the dynamic SMBus address with the LSB bit set. Note: Bit 0 (LSB) of Data byte 17 will always be 1b. If the SMBus Address has been resolved (Address Resolved flag is true), then for a general command the MNG will not acknowledge (NACK) this transaction, or for a directed command the MNG will always acknowledge (ACK) this transaction. This command does not affect the status or validity of the dynamic SMBus Address (will not clear the address Valid flag) nor of the Address Resolved flag. The command returns the UDID bytes as defined in Section 22.2.1.1.7. S Slave Address Wr A Command A 1100 001 0 0 See below 0 7 1 1 8 1 Slave Address Rd A Byte Count A 1100 001 1 0 0001 0001 0 *** *** 8 1 8 1 8 1 8 1 Data-1 A Data-2 A Data-3 A Data-4 A UDID byte 15 (MSB) 0 UDID byte 14 0 UDID byte 13 0 UDID byte 12 0 8 1 8 1 Data-5 UDID byte 11 A Data-6 0 UDID byte 10 Intel(R) Communications Chipset 89xx Series - Datasheet 1238 S 8 1 8 1 A Data-7 0 UDID byte 9 A Data-8 A 0 UDID byte 8 0 *** *** October 2012 Order Number: 327879-001US 27.0 8 1 8 1 Data-9 UDID byte 7 8 1 A Data-10 0 UDID byte 6 A Data-11 A 0 UDID byte 5 0 *** 8 1 8 1 8 1 8 1 Data-12 A Data-13 A Data-14 A Data-15 A UDID byte 4 0 UDID byte 3 0 UDID byte 2 0 UDID byte 1 0 8 1 8 1 Data-16 UDID byte 0 (LSB) A Data-17 0 Device Slave Address 27.4 Manageability Receive Filtering 27.4.1 Overview and General Structure *** 8 1 1 A PEC ~A P 0 [Data dependent value] 1 This section describes the manageability receive packet filtering flow. The description applies to each of the Controller LAN ports. A packet that is received by the Controller can have one of the following results: * Discarded * Sent to host memory * Sent to the external BMC * Sent to both the BMC and host memory The decisions regarding forwarding of packets to the host and to the BMC are separate and are configured through two sets of registers. However, the BMC may define some types of traffic as exclusive. This traffic will be forwarded only to the BMC, even if it passes the filtering process of the host. These types of traffic are defined using the MNGONLY register. The BMC controls the types of packets that it receives by programming the receive manageability filters. The software device driver can not write to the manageability filter registers. The table below lists the registers that are only written by the BMC. Table 27-8. Registers Written by the BMC Register Functionality When Reset MANC General configuration of the manageability filters. LAN_PWR_GOOD and Firmware Reset MNGONLY Enables routing of packets exclusively to the manageability. LAN_PWR_GOOD and Firmware Reset MDEF[7:0] MDEF_EXT[7:0] Configuration of manageability decision filters LAN_PWR_GOOD and Firmware Reset MMAH[1:0], MMAL[1:0] Two Unicast MAC manageability addresses LAN_PWR_GOOD MAVTV[7:0] Eight VLAN tag values LAN_PWR_GOOD MFUTP[3:0] Eight UDP/TCP destination port values LAN_PWR_GOOD October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1239 Table 27-8. Registers Written by the BMC Register Functionality When Reset FTFT Values and mask and length for the 1 flex TCO filter LAN_PWR_GOOD MIPAF IP address for manageability filtering LAN_PWR_GOOD METF L2 EtherType values LAN_PWR_GOOD These registers are reset only on LAN_PWR_GOOD. In SMBus PT mode, registers that enable filters or functionality are loaded from the EEPROM following a firmware reset. See Section 24.10 for description of their location in the EEPROM map. The high-level structure of manageability filtering is done using two steps. 1. The packet is parsed and fields in the header are compared to programmed filters. 2. A set of decision filters are applied to the result of the first step. General rules: * Fragmented packets are passed to manageability but not parsed beyond the IP header. * Packets with L2 errors (CRC, alignment, etc.) are never forwarded to manageability. The following sections describe the manageability filtering, followed by the final filtering rules. The filtering rules are created by programming the decision filters as described in Section 27.4.4. 27.4.2 L2 Filters 27.4.2.1 MAC and VLAN filters The manageability MAC filters allow comparison of the Destination MAC address to one of two filters defined in the MMAH and MMAL registers. The VLAN filters allow comparison of the 12-bit VLAN tag to one of eight filters defined in the MAVTV registers. 27.4.2.2 EtherType Filters The manageability L2 EtherType filters allow filtering of receive packets based on the Layer 2 EtherType field. The L2 type field of incoming packets is compared against the EtherType filters programmed in the METF.EType (up to 4 filters) and the result is incorporated to the decision filters. Each of the manageability EtherType filters can be configured as pass ("positive") or reject ("negative") polarity. When there is no match on its EtherType, and the polarity bit is set for a filter, that filter is enabled so as to participate in the decision process. Note: For the reverse polarity mode to be effective and block certain type of packets, the ethertype filter should be part of all the enabled decision filters. Intel(R) Communications Chipset 89xx Series - Datasheet 1240 October 2012 Order Number: 327879-001US 27.0 27.4.3 L3 and L4 Filters 27.4.3.1 ARP Filtering The Controller supports filtering of both ARP request packets (initiated externally) and ARP responses (to requests initiated by the BMC or the Controller). 27.4.3.2 Neighbor Discovery Filtering The Controller supports filtering of neighbor Discovery packets. neighbor Discovery uses the IPV6 destination address filters defined in the MIPAF registers (all enabled IPv6 addresses are matched for neighbor Discovery). 27.4.3.3 RMCP Port Filtering The Controller supports reception of RMCP packets by enabling filtering of the fixed destination ports numbers, port 0x26F and port 0x298. 27.4.3.4 Flex Port Filtering The Controller implements 8 flex UDP/TCP destination port filters. The Controller directs packets whose L4 destination port matches the value of the respective word in the MFUTP registers. The BMC must ensure that only valid entries are enabled in the decision filters below. 27.4.3.5 The Controller IP Address Filtering The Controller supports filtering by IP address through IPv4 and IPv6 address filters, dedicated to manageability. Two modes are possible, depending on the value of the MANC. EN_IPv4_FILTER bit: * EN_IPv4_FILTER = 0b: The Controller provides four IPv6 address filters. * EN_IPv4_FILTER = 1b: The Controller provides three IPv6 address filters and four IPv4 address filters. 27.4.3.6 Checksum Filter If bit MANC.EN_XSUM_FILTER is set, the Controller directs packets to the BMC only if they match all other filters previously described as well as pass L3/L4 checksum (if it exists). 27.4.4 Manageability Decision Filters The manageability decision filters are a set of eight filters with the same structure. The filtering rule for each decision filter is programmed by the BMC. The filtering rule defines which of the manageability filters participate in the decision. A packet that passes at least one rule is directed to manageability. The manageability filters are controlled by the BMC only and not by the LAN driver. To filter network traffic according to a filtering rule, the BMC should define the filtering rule by programming the MDEF and MDEF_EXT registers and set the MDEF_EXT.apply_to_network_traffic bit to 1, to enable the decision filter. The inputs to each decision filter are: * Packet passed one of the management L2 unicast address filter. * Packet is a broadcast packet. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1241 * Packet has a VLAN header and it passed one of the manageability VLAN filters. * Packet matched one of the IPv4 or IPv6 manageability address filters. * Packet is a multicast packet. * Packet passed ARP filtering (request or response). * Packet passed neighbor Discovery filtering. * Packet passed 0x298/0x26F port filter. * Packet passed a valid flex port filter. * Packet passed a valid flex TCO filter. * Packet passed or failed an L2 EtherType filter. The structure of each of the decision filters is shown in Section 27-2, "Manageability Decision Filters". A boxed "x.y" number indicates that the input is conditioned on a mask bit "y" defined in the MDEF and MDEF_EXT registers for this rule (where x=0 denotes MDEF and x=1 denotes MDEF_EXT). The decision filter rules are as follows: * All enabled AND filters must match for the decision filter to match. An AND filter not enabled in the MDEF/MDEF_EXT registers is ignored. In a set of AND filters from the same type, it is enough that one of the enabled filter match. The AND filter types are: -- L2 address - includes the Unicast AND, Broadcast AND and Multicast AND filters. -- VLAN - includes all the VLAN AND filters. -- IPv4 - includes all the IPv4 AND filters. -- IPv6 - includes all the IPv6 AND filters. * If at least one OR filter is enabled in the respective MDEF and MDEF_EXT registers, then at least one of the enabled OR filters must match for the decision filter to match. Otherwise, the OR filters are ignored in the decision (the filter might still match). A decision filter (for any of the eight filters) defines which of the above inputs is enabled as part of the rule. The BMC programs two 32-bit registers per rule (MDEF[7:0] and MDEF_EXT[7:0]) with the settings as described in Chapter 28.0, "GbE Programming Interface". A set bit allows its corresponding filter to participate in the filtering decision. Intel(R) Communications Chipset 89xx Series - Datasheet 1242 October 2012 Order Number: 327879-001US 27.0 Figure 27-2. Manageability Decision Filters L2 EtherType 3 1.11 L2 EtherType 0 1.8 Flex TCO 1.24 Flex Port 7 1.23 Flex Port 0 1.16 Port 0x26F 0.31 Port 0x298 0.30 Neighbor Discovery 0.29 ARP Request 0.28 ARP Response 0.27 Broadcast 0.25 L2 unicast address 0 0.21 L2 unicast address 1 0.22 L2 EtherType 1.28 L2 EtherType 0 1.0 L2 EtherType 3 1.3 IPv4 address 0 0.13 IPv4 address 3 0.16 IPv6 address 0 0.17 IPv6 address 3 0.20 VLAN 0 0.5 VLAN 7 0.12 Broadcast 0.4 L2 unicast address 1 0.1 L2 unicast address 0 Multicast October 2012 Order Number: 327879-001US 0.0 0.26 1.29 Flow Control L2 EtherType Intel(R) Communications Chipset 89xx Series - Datasheet 1243 27.4.4.1 Exclusive Traffic The decisions regarding forwarding of packets to the host for LAN traffic or to the LAN for host traffic are independent from the management decision filters. However, the BMC may define some types of traffic as exclusive. The behavior for such traffic is defined by the using the bits corresponding to the decision filter in the MNGONLY register (one bit per each of the eight decision rules) and the MDEF_EXT.apply_to_network_traffic bit. Table 27-9 describes the behavior in each case. If one or more filters match the traffic and at least one of the filters is set as exclusive, the traffic is treated as exclusive. Table 27-9. Exclusive Traffic Behavior Filter match MNGONLY = 0 Traffic is forwarded to the manageability. Traffic is forwarded to the host according to host filtering 27.4.5 Filter doesn't match MNGONLY = 1 Traffic is forwarded only to manageability. N/A Traffic is forwarded to the host according to host filtering Possible Configurations This section describes ways of using management filters. Actual usage may vary. 27.4.5.1 Dedicated MAC Packet Filtering * Select one of the eight rules for dedicated MAC filtering. * Load Host MAC address to one of the management MAC address filters and set the appropriate bit in field 1:0 of the MDEF register. * Set other bits to qualify which packets are allowed to pass through. For example: -- Load one or more management VLAN filters and set the appropriate bits in field 12:5 of the MDEF register to qualify the relevant manageability VLANs. -- Set relevant bits in field 20:13 of the MDEF register to qualify with a match to one of the IP addresses. -- Set any L3/L4 bits (bits 31:27 in the MDEF register and bits 23:16 in the MDEF_EXT register) to filter using any set of L3/L4 filters. 27.4.5.2 Broadcast Packet Filtering * Select one of the eight rules for broadcast filtering. * Set bit 25 in the MDEF register of the decision rule to enforce broadcast filtering. * Set other bits to qualify which broadcast packets are allowed to pass through. For example: -- Set bit 5 in the MDEF register to filter with the first manageability VLAN. -- Set relevant bits in field 20:13 of the MDEF register to qualify with a match to one of the IP addresses. -- Set any L3/L4 bits (bits 31:27 in the MDEF register and bits 23:16 in the MDEF_EXT register) to filter with any set of L3/L4 filters. Intel(R) Communications Chipset 89xx Series - Datasheet 1244 October 2012 Order Number: 327879-001US 27.0 27.4.5.3 VLAN Packet Filtering * Select one of the eight rules for VLAN filtering. Set bit 2 of the decision rule to enforce VLAN filtering. * Load one or more management VLAN filters and set the appropriate bits in field 12:5 of the MDEF register to qualify the relevant manageability VLANs. * Set other bits to qualify which VLAN packets are allowed to pass through. For example: Set any L3/L4 bits (bits 31:27 in the MDEF register and bits 23:16 in the MDEF_EXT register) to filter using appropriate L3/L4 filter set. 27.4.5.4 IPv6 Filtering IPv6 filtering is done using the following IPv6-specific filters: * IP Unicast filtering -- requires filtering for Link Local address and a Global address. Filtering setup might depend on whether or not the MAC address is shared with the Host or dedicated to manageability: -- Dedicated MAC address (for example, dynamic address allocation with DHCP does not support multiple IP addresses for one MAC address). In this case, filtering can be done at L2 using two dedicated unicast MAC filters. -- Shared MAC address (for example, static address allocation sharing addresses with Host). In this case, filtering needs to be done at L3, requiring two IPv6 address filters, one per address. * A neighbor Discovery filter -- Supports IPv6 neighbor Discovery protocol. Since the protocol relies on multicast packets, Supports filtering of these packets. IPv6 multicast addresses are translated into corresponding Ethernet multicast addresses in the form of 33-33-xxxx-xx-xx, where the last 32 bits of address are taken from the last 32 bits of the IPv6 multicast address. As a result, two direct MAC filters can be used to filter IPv6 solicited-node multicast packets and IPv6 all node multicast packets. 27.4.5.5 Receive Filtering with Shared IP - CPMP When using the SMBus interface, it is possible to share the Host MAC and IP address with the BMC. When the BMC shares the MAC and IP address with the Host, receive filtering is based on identifying specific flows through port allocation. The following setting might be used: * Select one of the eight rules for Dedicated MAC filtering. * Load Host MAC address to one of the management MAC address filters and set the appropriate bit in field 1:0 of the MDEF register to enforce MAC address filtering using the MAC address. * If VLAN is used for management, load one or more management VLAN filters and set the appropriate bits in field 12:5 of the MDEF register to qualify the relevant manageability VLANs. * ARP filter/Neighbor Discovery filter is enabled when the BMC is responsible for handling the ARP protocol. Set bit 27 or bit 28 in the MDEF register for this functionality. * Set other bits to qualify which packets are allowed to pass through. For example: Set any L3/L4 bits (bits 31:27 in the MDEF register and bits 23:16 in the MDEF_EXT register) to filter using the appropriate L3/L4 filters. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1245 28.0 GbE Programming Interface 28.1 Introduction This chapter details the programmer visible state inside the GbE Controller. In some cases, it describes hardware structures invisible to software to clarify a concept. The controllers's address space is mapped into four regions with PCI Base Address Registers. These regions are listed in Table 28-1. Table 28-1. Address Space Regions Addressable Content How Mapped Size of Region Internal registers and memories ("Memory BAR") Direct memory-mapped 128K Internal registers and memories I/O Window mapped 32 bytes (3) MSI-X (optional) Direct memory-mapped 16K The internal registers and memories can be accessed though I/O space indirectly. The internal register/memory space is described in the following sections. The PHY registers are accessed through the MDIO interface. 28.1.1 Memory, I/O Address, and Configuration Decoding 28.1.1.1 Memory-Mapped Access to Internal Registers and Memories The internal registers and memories might be accessed as direct memory-mapped offsets from the base address register (BAR0 or BAR 0/1). See Section 28.1.3 for the appropriate offset for each specific internal register. 28.1.1.2 Memory-Mapped Access to MSI-X Tables The MSI-X tables may be accessed as direct memory-mapped offsets from the base address register (BAR3). See Section 28.1.3 for the appropriate offset for each specific internal MSIX register. 28.1.1.3 I/O-Mapped Access to Internal Registers and Memories To support pre-boot operation (prior to the allocation of physical memory base addresses), all internal registers and memories can be accessed using I/O operations. I/O accesses are supported only if an I/O Base Address is allocated and mapped (BAR2), the BAR contains a valid (non-zero value), and I/O address decoding is enabled in the PCIe configuration. When an I/O BAR is mapped, the I/O address range allocated opens a 32-byte "window" in the system I/O address map. Within this window, two I/O addressable registers are implemented: IOADDR and IODATA. The IOADDR register is used to specify a reference to an internal register or memory, and then the IODATA register is used as a "window" to the register or memory address specified by IOADDR: Intel(R) Communications Chipset 89xx Series - Datasheet 1246 October 2012 Order Number: 327879-001US 28.0 Table 28-2. IOADDR and IODATA in I/O Address Space Offset Abbreviation Name RW Size 0x00 IOADDR Internal Register or Internal Memory location address. 0x00000-0x1FFFF - Internal Registers and Memories 0x20000-0xFFFFFFFF - Undefined RW 4 bytes 0x04 IODATA Data field for reads or writes to the Internal Register or Internal Memory Location as identified by the current value in IOADDR. All 32 bits of this register are read/write-able. RW 4 bytes 0x08 - 0x1F Reserved Reserved RO 4 bytes 28.1.1.3.1 IOADDR (I/O Offset 0x00) The IOADDR register must be written as a DWORD access. Writes of less than 32 bits are ignored. Reads of any size return a DWORD of data; however, the chipset or CPU might only return a subset of that DWORD. For software programmers, the IN and OUT instructions must be used to cause I/O cycles to be used on the PCIe bus. Because writes must be to a 32-bit quantity, the source register of the OUT instruction must be EAX (the only 32-bit register supported by the OUT command). For reads, the IN instruction can have any size target register, but it is recommended that the 32-bit EAX register be used. Because only a particular range is addressable, the upper register bits are hard coded to zero. Bits 30 through 20 are not write-able and always read back as 0b. Set Bit 31 (IOADDR.Configuration IO Access Enable) of the IOADDR register to 1. At hardware reset (LAN_PWR_GOOD) or PCI Reset, this register value resets to 0x00000000. Once written, the value is retained until the next write or reset. 28.1.1.3.2 IODATA (I/O Offset 0x04) The IODATA register must be written as a DWORD access when the IOADDR register contains a value for the Internal Register and Memories (for example, 0x000000x1FFFC). In this case, writes that are less than 32 bits are ignored. Reads to IODATA of any size returns a DWORD of data. However, the chipset or CPU might only return a subset of that DWORD. For software programmers, the IN and OUT instructions must be used to cause I/O cycles to be used on the PCIe bus. Where 32-bit quantities are required on writes, the source register of the OUT instruction must be EAX (the only 32-bit register supported by the OUT command). Writes and reads to IODATA when the IOADDR register value is in an undefined range (0x20000-0xFFFFFFFC) should not be performed. Results cannot be determined. Notes: There are no special software timing requirements on accesses to IOADDR or IODATA. All accesses are immediate, except when data is not readily available or acceptable. In this case, the Controller delays the results through normal bus methods (for example, split transaction or transaction retry). Because a register/memory read or write takes two IO cycles to complete, software must provide a guarantee that the two IO cycles occur as an atomic operation. Otherwise, results can be non-deterministic from the software viewpoint. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1247 28.1.1.3.3 Undefined I/O Offsets I/O offsets 0x08 through 0x1F are considered to be reserved offsets with the I/O window. Dword reads from these addresses returns 0xFFFF; writes to these addresses are discarded. 28.1.1.4 Configuration Access to Internal Registers and Memories To support 'legacy' pre-boot 16-bit operating environments without requiring IO address space, the Controller enables accessing CSRs via configuration address space by mapping the IOADDR and IODATA registers into configuration address space. If the CSR_conf_en bit in the PCIe Init Configuration 2 EEPROM word (Word 0x19) is set to 1, access to CSRs via Configuration address space is enabled. The registers mapping in this case is shown in Table 28-3. Table 28-3. IOADDR and IODATA in Configuration Address Space Configuration Address Abbreviation Name RW Size 0x98 IOADDR Internal register or internal memory location address. 0x00000-0x1FFFF - Internal Registers and Memories 0x20000-0x7FFFFF - Undefined RW 4 bytes 0x9C IODATA Data field for reads or writes to the internal register or internal memory location as identified by the current value in IOADDR. All 32 bits of this register are read/write-able. RW 4 bytes Software writes data to an internal CSR via configuration space in the following manner: 1. CSR address is written to the IOADDR register where: a. Bit 31 (IOADDR.Configuration IO Access Enable) of the IOADDR register should be set to 1. b. Bits 30:0 of IOADDR should hold the actual address of the internal register or memory being written to. 2. Data to be written is written into the IODATA register. -- The IODATA register is used as a "window" to the register or memory address specified by IOADDR register. As a result the data written to the IODATA register is written into the CSR pointed to by bits 30:0 of the IOADDR register. 3. IOADDR.Configuration IO Access Enable is cleared, to avoid un-intentional CSR read operations (that may cause clear by read) by other applications scanning the configuration space. Software reads data from an internal CSR via Configuration space in the following manner: 1. CSR address is written to the IOADDR register where: a. Bit 31 (IOADDR.Configuration IO Access Enable) of the The IOADDR register should be set to 1. b. Bits 30:0 of IOADDR should hold the actual address of the internal register or memory being read. 2. CSR value is read from the IODATA register. a. The IODATA register is used as a "window" to the register or memory address specified by IOADDR register. As a result the data read from the IODATA register is the data of the CSR pointed to by bits 30:0 of the IOADDR register Intel(R) Communications Chipset 89xx Series - Datasheet 1248 October 2012 Order Number: 327879-001US 28.0 3. IOADDR.Configuration IO Access Enable is cleared, to avoid un-intentional CSR read operations (that may cause clear by read) by other applications scanning the configuration space. Notes: * If the CSR_conf_en bit in the PCIe Init Configuration 2 EEPROM word is cleared, then accesses to the IOADDR and IODATA registers via the configuration address space are ignored and have no effect on the register and the CSRs referenced by the IOADDR register. * When Function is in D3 state Software should not attempt to access CSRs via the IOADDR and IODATA Configuration registers. * To enable CSR access via configuration space, Software should set to 1 bit 31 (IOADDR.Configuration IO Access Enable) of the IOADDR register. Software should clear bit 31 of the IOADDR register after completing CSR access to avoid an unintentional "clear by read" operation, by another application scanning the configuration address space. * Bit 31 of the IOADDR register (IOADDR.Configuration IO Access Enable) has no effect when initiating access via IO Address space. 28.1.2 Register Conventions All registers in the Controller are defined to be 32 bits and should be accessed as 32-bit double-words, with the exceptions of register pairs where two 32 bit registers make up a larger logical size. Some registers contain bits that are marked as "reserved". Software should never set these bits to a value of "1." Reads from registers containing reserved bits might return indeterminate values in the reserved bit-positions unless read values are explicitly stated. When read, software should ignore these reserved bits. Any register address not explicitly declared in this specification should be considered as reserved, and should not be written to. Writing to reserved or undefined register addresses might cause indeterminate behavior. Reads from reserved or undefined configuration register addresses might return indeterminate values unless read values are explicitly stated for specific addresses. Most registers define the initial hardware values prior to being programmed. In some cases, hardware initial values are undefined and is listed as such via the text "undefined", "unknown", or "X". Such configuration values might need to be set via EEPROM configuration or via software in order for proper operation to occur; this need is dependent on the function of the bit. Other registers might cite a hardware default which is overridden by a higher-precedence operation. Operations which might supersede hardware defaults might include a valid EEPROM load, completion of a hardware operation (such as hardware auto-negotiation), or writing of a different register whose value is then reflected in another bit. For registers that should be accessed as 32-bit double words, partial writes (less than a 32-bit double word) does not take effect. The write is ignored. Partial reads returns all 32 bits of data regardless of the byte enables. Note: Partial reads to read-on-clear registers (ICR) can have unexpected results since all 32 bits are read regardless of the byte enables. Partial reads should not be done. Note: All statistics registers are implemented as 32-bit registers. Though some logical statistics registers represent counters in excess of 32 bits in width, registers must be accessed using 32-bit operations (for example, independent access to each 32-bit field). When reading 64 bits statistics registers the least significant 32-bit register should be read first. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1249 See the special notes for VLAN Filter Table, Multicast Table Arrays and Packet Buffer Memory that are in the specific register definitions. The Controller register fields are assigned one of the attributes described in Table 28-4. Table 28-4. GbE Controller Register Field Attributes Attribute Description RW Read-Write field: Register bits are read-write and can be either set or cleared by software to the desired state. RWS Read-Write Status field: Register bits are read-write and can be either set or cleared by software to the desired state. However, the value of this field might be changed by the hardware to reflect a status change. RO Read-only register: Register bits are read-only and should not be altered by software. Register bits might be initialized by hardware mechanisms such as serial EEPROM or reflect a status of the hardware state. R/W1C Read-only status, Write-1-to-clear status register: Register bits indicate status when read, a set bit indicating a status event can be cleared by writing a 1b. Writing a 0b to R/W1C bit has no effect. Rsv Reserved. Write 0 to these fields and ignore read. RC Read-only status, Read-to-clear status register: Register bits indicate status when read, a set bit indicating a status event is cleared by reading it. SC Self Clear field: a command field that is self clearing. These field are always read as zero. WO Write only field: a command field that can not be read, These field read values are undefined. RC/W Read-Write status, Read-to-clear status register: Read-to-clear status register. Register bits indicate status when read. Register bits are read-write and can be either set or cleared by software to the desired state. RC/W1C Read-only status, Write-1-to-clear status register: Read-to-clear status register. Register bits indicate status when read, a set bit indicating a status event can be cleared by writing a 1b or by reading the register. Writing a 0b to RC/W1C bit has no effect. RS Read Set - This is the attribute used for Semaphore bits. These bits are set by read in case the previous values were zero. In this case the read value is zero; otherwise the read value is one. Cleared by write zero. Note: For all binary equations appearing in the register map, the symbol "|" is equivalent to a binary OR operation. 28.1.2.1 Registers Byte Ordering This section defines the structure of registers that contain fields carried over the network. Some examples are L2, L3 and L4 fields. The following example is used to describe byte ordering over the wire (hex notation): Last First ...,06, 05, 04, 03, 02, 01, 00 Each byte is sent with the LSbit first. That is, the bit order over the wire for this example is Last First ..., 0000 0011, 0000 0010, 0000 0001, 0000 0000 The general rule for register ordering is to use Host Ordering (also called little endian). Using the above example, a 6-byte fields (MAC address) is stored in a CSR in the following manner: Intel(R) Communications Chipset 89xx Series - Datasheet 1250 October 2012 Order Number: 327879-001US 28.0 Byte 3 Byte 2 Byte 1 Byte0 DW address (N) 0x03 DW address (N+4) ... 0x02 ... 0x01 0x05 0x00 0x04 The exceptions listed below use network ordering (also called big endian). Using the above example, a 16-bit field (EtherType) is stored in a CSR in the following manner: (DW aligned) or (Word aligned) Byte 3 Byte 2 Byte 1 Byte0 ... ... 0x00 0x01 0x00 0x01 ... ... The following exceptions use network ordering: All ETherType fields. For example, the VET EXT field in the VET register, the EType field in the ETQF register, the EType field in the METF register. Note: The "normal" notation as it appears in text books, etc. is to use network ordering. Example: Suppose a MAC address of 00-A0-C9-00-00-00. The order on the network is 00, then A0, then C9, etc. However, the host ordering presentation would be: Byte 3 Byte 2 Byte 1 Byte0 DW address (N) 00 C9 A0 00 DW address (N+4) ... ... 00 00 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1251 28.1.3 Detailed Register Summary 28.1.3.1 PCI Views Table 28-5. Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR0[03]B:0:1+Index 1 Registers (Sheet 1 of 10) Offset Start Offset End Default Value Register ID - Description 0000h 0003h "Device Control Register--CTRL [0:3] (0x00000; R/W)" on page 1264 0x08100201 00008h 0000Bh "Device Status Register-- STATUS[0:3] (0x0008; R)" on page 1268 0x80080400 00018h 0001Bh "Extended Device Control Register--CTRL_EXT [0:3] (0x0018; R/W)" on page 1269 0x00500000 00020h 00023h "MDI Control Register--MDIC[0:3] (0x0020; R/W)" on page 1272 0x10000000 00E04h 00E07h "MDC/MDIO Configuration Register--MDICNFG [0:3] (0x0E04; R/W)" on page 1273 0x0 0034h 0037h "Copper/Fiber Switch Control--CONNSW[0:3] (0x0034; R/W)" on page 1274 0x0 0038h 003Bh "VLAN Ether Type--VET [0:3] (0x0038; R/W)" on page 1274 0x81008100 E00h E03h "LED Control--LEDCTL [0:3](0x0E00; RW)" on page 1275 0x07568302 2404h 2407h "Internal Receive Packet Buffer Size--IRPBS [0:3] (0x2404; RO)" on page 1276 0x0 3404h 3407h "Internal Transmit Packet Buffer Size--ITPBS [0:3] (0x3404; RO)" on page 1277 0x8 0010h 0013h "EEPROM Control Register--EEC [0:3] (0x0010; R/W)" on page 1278 Xh 0014h 0017h "EEPROM Read Register--EERD [0:3] (0x0014; RW)" on page 1279 0x0 1038h 103Bh "EEPROM Diagnostic--EEDIAG [0:3] (0x1038; RO)" on page 1280 0x0 1060h 1063h "VPD Diagnostic Register--VPDDIAG [0:3] (0x1060; RO)" on page 1281 0x0 1010h 1013h "Management EEPROM Control Register--EEMNGCTL [0:3] (0x1010; RO)" on page 1282 0x80000000 1014h 1017h "Management EEPROM Read/Write Data--EEMNGDATA [0:3] (0x1014; RO)" on page 1283 0x0 0028h 002Bh "Flow Control Address Low--FCAL [0:3] (0x0028; RO)" on page 1284 0x00C28001 002Ch 002Fh "Flow Control Address High--FCAH [0:3] (0x002C; RO)" on page 1285 0x00000100 0030h 0033h "Flow Control Type--FCT [0:3] (0x0030; R/W)" on page 1285 0x00008808 0170h 0173h "Flow Control Transmit Timer Value--FCTTV [0:3] (0x0170; R/W)" on page 1286 0x0 2160h 2163h "Flow Control Receive Threshold Low--FCRTL0 [0:3] (0x2160; R/W)" on page 1287 0x0 2168h 216Bh "Flow Control Receive Threshold High--FCRTH0 [0:3] (0x2168; R/W)" on page 1288 0x0 2460h 2463h "Flow Control Refresh Threshold Value--FCRTV[0:3](0x2460; R/W)" on page 1289 0x0 2464h 2467h "Flow Control Status--FCSTS0 [0:3] (0x2464; RO)" on page 1290 0x0 5B30h 5B33h "Function Active and Power State to MNG--FACTPS[0:3] (0x5B30; RO)" on page 1290 0x0 5B64h 5B67h "Mirrored Revision ID--MREVID[0:3] (0x5B64; R/W)" on page 1292 0x0 5B6Ch 5B6Fh "PCIE Control Extended Register--GCR_EXT[0:3] (0x5B6C; R/W)" on page 1293 0x0 5B50h 5B53h "Software Semaphore--SWSM[0:3] (0x5B50; R/W)" on page 1294 0x0 5B54h 5B57h "Firmware Semaphore--FWSM[0:3] (0x5B54; R/WS)" on page 1295 0x0 Intel(R) Communications Chipset 89xx Series - Datasheet 1252 October 2012 Order Number: 327879-001US 28.0 Table 28-5. Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR0[03]B:0:1+Index 1 Registers (Sheet 2 of 10) Offset Start Offset End Default Value Register ID - Description 5B5Ch 5B5Fh "Firmware Synchronization--SW_FW_SYNC[0:3] (0x5B5C; RWS)" on page 1297 0x0 5B04h 5B07h "Software Mailbox Write--SWMBWR[0:3] (0x5B04; R/W)" on page 1298 0x0 5B08h 5B0Bh "Software Mailbox 0--SWMB0[0:3] (0x5B08; RO)" on page 1298 0x0 5B0Ch 5B0Fh "Software Mailbox 1--SWMB1[0:3] (0x5B0C; RO)" on page 1298 0x0 5B18h 5B1Bh "Software Mailbox 2--SWMB2[0:3] (0x5B18; RO)" on page 1299 0x0 5B1Ch 5B1Fh "Software Mailbox 3--SWMB3[0:3] (0x5B1C; RO)" on page 1299 0x0 1580h 1583h "EICR Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0)" on page 1300 0x0 1580h 1583h "EICR Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1)" on page 1301 0x0 1520h 1523h "EICS Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0)" on page 1301 0x0 1520h 1523h "EICS Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1)" on page 1302 0x0 1524h 1527h "EIMS Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0)" on page 1302 0x80000000 1524h 1527h "EIMS Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1)" on page 1303 Xh 1528h 152Bh "EIMC Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0)" on page 1304 0x80000000 1528h 152Bh "EIMC Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1)" on page 1304 0x0 152Ch 152Fh "Extended Interrupt Auto Clear--EIAC [0:3] (0x152C; R/W)" on page 1305 0x0 1530h 1533h "EIAM Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0)" on page 1306 0x0 1530h 1533h "EIAM Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1)" on page 1306 0x0 1500h 1503h "Interrupt Cause Read Register--ICR [0:3] (0x1500; RC/W1C)" on page 1307 0x0 1504h 1507h "Interrupt Cause Set Register--ICS [0:3] (0x1504; WO)" on page 1309 0x0 1508h 150Bh "Interrupt Mask Set/Read Register--IMS [0:3] (0x1508; R/W)" on page 1311 0x0 150Ch 150Fh "Interrupt Mask Clear Register--IMC [0:3] (0x150C; WO)" on page 1313 0x0 1510h 1513h "Interrupt Acknowledge Auto Mask Register--IAM [0:3] (0x1510; R/W)" on page 1315 0x0 1680h at 4 1683h at 4 "Interrupt Throttle--EITR [0:3][0:9] (0x1680 + 4*n [n = 0...9]; R/W)" on page 1316 0x0 1700h at 4 1703h at 4 "Interrupt Vector Allocation Registers--IVAR [0:3][0:3](0x1700 + 4*n [n=0...3]; RW)" on page 1317 0x0 1740h 1743h "Interrupt Vector Allocation Registers--MISC IVAR_MISC[0:3] (0x1740; RW)" on page 1318 0x0 1514h 1517h "General Purpose Interrupt Enable--GPIE [0:3] (0x1514; RW)" on page 1319 0x0 5B68h 5B6Bh "MSIX PBA Clear--PBACL [0:3] (0x5B68; R/W1C)" on page 1323 0X0 0100h 0103h "Receive Control Register--RCTL[0:3] (0x0100; R/W)" on page 1324 0x0 C00Ch at 0x4 C00Fh at 0x4 "Split and Replication Receive Control--SRRCTL [0:3][0:7] (0xC00C + 0x40*n [n=0...7]; R/W)" on page 1327 0x00000400 C04Ch at 0x40 C04Fh at 0x40 "Split and Replication Receive Control--SRRCTL [0:3][1:7](0xC00C + 0x40*n [n=0...7]; R/W)" on page 1328 0x80000400 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1253 Table 28-5. Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR0[03]B:0:1+Index 1 Registers (Sheet 3 of 10) Default Value Offset Start Offset End Register ID - Description 5480h at 0x4 5483h at 0x4 "Packet Split Receive Type--PSRTYPE [0:3][0:7] (0x5480 + 4*n [n=0...7]; R/W)" on page 1330 0x0007FFFE 54C0h 54C3h "Replicated Packet Split Receive Type--RPLPSRTYPE [0:3] (0x54C0; R/W)" on page 1331 0x0007FFFE C000h at 0x40 C000h at 0x40 "Receive Descriptor Base Address Low--RDBAL [0:3][0:7] (0xC000 + 0x40*n [n=0...7]; R/W)" on page 1332 0x0 C004h at 0x40 C007h at 0x40 "Receive Descriptor Base Address High--RDBAH [0:3][0:7] (0xC004 + 0x40*n [n=0...7]; R/W)" on page 1332 0x0 C008h at 0x40 C00Bh at 0x40 "Receive Descriptor Ring Length--RDLEN [0:3][0:7](0xC008 + 0x40*n [n=0...7]; 0x0 R/W)" on page 1333 C010h at 0x40 C013h at 0x40 "Receive Descriptor Head--RDH [0:3][0:7] (0xC010 + 0x40*n [n=0...7]; RO)" on 0x0 page 1333 C018h at 0x40 C01Bh at 0x40 "Receive Descriptor Tail--RDT [0:3][0:7] (0xC018 + 0x40*n [n=0...7]; R/W)" on page 1334 C028h at 0x40 C02Bh at 0x40 "Receive Descriptor Control--RXDCTL [0:3][0:7] (0xC028 + 0x40*n [n=0...7]; R/ 0x0000140C W)" on page 1334 C030h at 0x40 C033h at 0x40 "Receive Queue Drop Packet Count--RQDPC [0:3][0:7] (0xC030 + 0x40*n [n=0...7]; RC/W)" on page 1336 0x0 5000h 5003h "Receive Checksum Control--RXCSUM [0:3] (0x5000; R/W)" on page 1337 0x00000300 5004h 5007h "Receive Long Packet Maximum Length--RLPML [0:3] (0x5004; R/W)" on page 1339 0x00002600 5008h 500Bh "Receive Filter Control Register--RFCTL [0:3] (0x5008; R/W)" on page 1339 0x1 5200h at 0x4 5203h at 0x4 "Multicast Table Array--MTA [0:3][0:127] (0x5200 + 4*n [n=0...127]; R/W)" on page 1340 0x0 5400h at 0x8 5403h at 0x8 "Receive Address Low 0--RAL0 [0:3][0:15] (0x5400 + 8*n [n=0...15]; 0x54E0 + 0x0 8*n [n=0...7]; R/W)" on page 1341 54E0h at 0x8 54E3h at 0x8 "Receive Address Low 1--RAL1 [0:3][0:7] (0x54E0 + 8*n [n=0...7]; R/W)" on page 1342 5404h at 0x8 5407h at 0x8 "Receive Address High 0--RAH0 [0:3][0:15] (0x5404 + 8*n [n=0...15]; 0x54E04 0x0 + 8*n [n=0...7]; R/W)" on page 1343 54E4h at 0x8 54E7h at 0x8 "Receive Address High 1--RAH1 [0:3][0:7] (0x54E4 + 8*n [n=0...7]; R/W)" on page 1344 0x0 5600h at 0x4 5603h at 0x4 "VLAN Filter Table Array--VFTA [0:3][0:127] (0x5600 + 4*n [n=0...127]; R/W)" on page 1345 0x0 5818h 581Bh "Multiple Receive Queues Command Register--MRQC [0:3](0x5818; R/W)" on page 1346 0x0 5C80h at 0x4 5C83h at 0x4 "RSS Random Key Register--RSSRK [0:3][0:9] (0x5C80 + 4*n [n=0...9]; R/W)" on page 1348 0x0 5C00h at 0x4 5C03h at 0x4 "Redirection Table--RETA [0:3][0:31] (0x5C00 + 4*n [n=0...31]; R/W)" on page 1348 0x0 5A80h at 0x4 5A83h at 0x4 "Immediate Interrupt RX--IMIR [0:3][0:7] (0x5A80 + 4*n [n=0...7]; R/W)" on page 1350 0x0 5AA0h at 0x4 5AA3h at 0x4 "Immediate Interrupt Rx Ext.--IMIREXT [0:3][0:7] (0x5AA0 + 4*n [n=0...7]; R/ W)" on page 1351 0x0 59E0h at 0x4 59E3h at 0x4 "2tuples Queue Filter--TTQF[0:3][0:7] (0x59E0 + 4*n[n=0..7]; R/W)" on page 1352 0x0 5AC0h 5AC3h "Immediate Interrupt Rx VLAN Priority--IMIRVP [0:3] (0x5AC0; R/W)" on page 1352 0x0 55FCh 55FFh "SYN Packet Queue Filter--SYNQF [0:3] (0x55FC; RW)" on page 1353 0x0 Intel(R) Communications Chipset 89xx Series - Datasheet 1254 0x0 0x0 October 2012 Order Number: 327879-001US 28.0 Table 28-5. Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR0[03]B:0:1+Index 1 Registers (Sheet 4 of 10) Register ID - Description Default Value Offset Start Offset End 5CB0h at 0x4 5CB3h at 0x4 0400h 0403h "Transmit Control Register--TCTL [0:3] (0x0400; R/W)" on page 1355 0x000400F8 0404h 0407h "Transmit Control Extended--TCTL_EXT [0:3] (0x0404; R/W)" on page 1356 0x00010840 "EType Queue Filter--ETQF [0:3][0:7] (0x5CB0 + 4*n[n=0...7]; RW)" on page 1354 0x0 0410h 0413h "Transmit IPG Register--TIPG [0:3] (0x0410; R/W)" on page 1357 0x00601008 041Ch 041Fh "Retry Buffer Control--RETX_CTL [0:3] (0x041C; RW)" on page 1358 0x00000003 0x00400004 3590h 3593h "DMA Tx Control--DTXCTL [0:3] (0x3590; R/W)" on page 1358 359Ch 359Fh "DMA TX TCP Flags Control Low--DTXTCPFLGL [0:3] (0x359C; RW)" on page 1359 0x00400004 35A0h 35A3h "DMA TX TCP Flags Control High--DTXTCPFLGH [0:3] (0x35A0; RW)" on page 1360 0x00000F7F 3540h 3543h "DMA TX Max Total Allow Size Requests--DTXMXSZRQ [0:3] (0x3540; RW)" on page 1360 0x00000010 355Ch 355Fh "DMA TX Maximum Packet Size--DTXMXPKTSZ [0:3] (0x355C; RW)" on page 1361 0x00000098 E000h E003h "Transmit Descriptor Base Address Low--TDBAL [0:3][0:7] (0xE000 + 0x40*n [n=0...7]; R/W)" on page 1361 0x0 E004h at 0x40 E007h at 0x40 "Transmit Descriptor Base Address High--TDBAH [0:3][0:7] (0xE004 + 0x40*n [n=0...7]; R/W)" on page 1362 0x0 E008h at 0x40 E00Bh at 0x40 "Transmit Descriptor Ring Length--TDLEN [0:3][0:7] (0xE008 + 0x40*n [n=0...7]; R/W)" on page 1362 0x0 E010h at 0x40 E013h at 0x40 "Transmit Descriptor Head--TDH [0:3] [0:7] (0xE010 + 0x40*n [n=0...7]; RO)" on 0x0 page 1363 E018h at 0x40 E01Bh at 0x40 "Transmit Descriptor Tail--TDT [0:3][0:7] (0xE018 + 0x40*n [n=0...7]; R/W)" on page 1363 E028h at 0x40 E02Bh at 0x40 "Transmit Descriptor Control--TXDCTL [0:3][0:7] (0xE028 + 0x40*n [n=0...7]; R/ 0x0 W)" on page 1364 E038h at 0x40 E03Bh at 0x40 "Tx Descriptor Completion Write-Back Address Low--TDWBAL [0:3][0:7] (0xE038 + 0x40*n [n=0...7]; R/W)" on page 1366 E03Ch at 0x40 E03Fh at 0x40 "Tx Descriptor Completion Write-Back Address High - TDWBAH [0:3][0:7] (0xE03C 0x0 + 0x40*n [n=0...7];R/W)" on page 1366 C014h at 0x40 C017h at 0x40 "Rx DCA Control Registers--RXCTL [0:3][0:7] (0xC014 + 0x40*n [n=0...7]; R/W)" 0x0000A200 on page 1367 E014h at 0x40 E017h at 0x40 "Tx DCA Control Registers--TXCTL [0:3][0:7] (0xE014 + 0x40*n [n=0...7]; R/W)" 0x00002A00 on page 1369 5B70h 5B73h "DCA Requester ID Information--DCA_ID [0:3] (0x5B70; RO)" on page 1370 0x0 5B74h 5B77h "DCA Control--DCA_CTRL [0:3] (0x5B74; R/W)" on page 1370 0x1 581Ch 581Fh "VMDq Control Register--VT_CTL [0:3] (0x581C; R/W)" on page 1371 0x0 0x0 0x0 3558h 355Bh "Malicious Driver Free Block--MDFB [0:3] (0x3558; RWS)" on page 1372 0x0 3548h 354Bh "Last VM Misbehavior Cause--LVMMC[0:3] (0x3548; RC)" on page 1372 0x0 5AD0h at 0x4 5AD3h at 0x4 "VM Offload Register--VMOLR [0:3][0:7] (0x5AD0 + 4*n [n=0...7]; RW)" on page 1373 0x80002600 5AF0h "Replication Offload Register--RPLOLR [0:3] (0x5AF0; RW" on page 1374 0x80000000 5DD0h at 0x4 5DD3h at 0x4 "VLAN VM Filter--VLVF [0:3][0:31] (0x5D00 + 4*n [n=0...31]; RW)" on page 1375 0x0 A000h at 0x4 A003h at 0x4 "Unicast Table Array--UTA [0:3][0:127] (0xA000 + 4*n [n=0...127]; R/W)" on page 1376 0x0 5DB0h 5DB3h "Storm Control Control Register--SCCRL [0:3] (0x5DB0; RW)" on page 1376 0x00000800 5AF3h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1255 Table 28-5. Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR0[03]B:0:1+Index 1 Registers (Sheet 5 of 10) Offset Start 5DB4h Offset End 5DB7h Default Value Register ID - Description "Storm Control Status--SCSTS [0:3] (0x5DB4;RO)" on page 1377 0x0 5DB8h 5DBBh "Broadcast Storm Control Threshold--BSCTRH [0:3] (0x5DB8;RW)" on page 1378 0x0 5DBCh 5DBFh "Multicast Storm Control Threshold--MSCTRH [0:3] (0x5DBC; RW)" on page 1378 0x0 5DC0h 5DC3h "Broadcast Storm Control Current Count - BSCCNT [0:3] (0x5DC0;RO)" on page 1378 0x0 5DC4h 5DC7h "Multicast Storm Control Current Count--MSCCNT [0:3] (0x5DC4;RO)" on page 1379 0x0 5DC8h 5DCBh "Storm Control Time Counter--SCTC [0:3] (0x5DC8; RO)" on page 1379 0x0 5DCCh 5DCFh "Storm Control Basic Interval--SCBI [0:3] (0x5DCC; RW)" on page 1380 0x0 5D80h at 0x4 5D83h at 0x4 "Virtual Mirror Rule Control--VMRCTL [0:3][0:3] (0x5D80 + 0x4*n [n= 0...3]; RW)" on page 1380 0x0 5D90h at 0x4 5D93h at 0x4 "Virtual Mirror Rule VLAN--VMRVLAN [0:3][0:3] (0x5D90 + 0x4*n [n= 0...3]; RW)" on page 1381 0x0 5DA0h at 0x4 5DA3h at 0x4 "Virtual Mirror Rule VM--VMRVM [0:3][0:3] (0x5DA0 + 0x4*n [n= 0...3]; RW)" on 0x0 page 1381 2500h 2503h "DMA Receive Power Saving Register--DMARPS [0:3] (0x2500; R/W)" on page 1382 0x0000010A 2504h 2507h "DMA Transmit Power Saving Register--DMATPS [0:3] (0x2504; R/W)" on page 1382 0x0000010A 1040h 1043h "Watchdog Setup--WDSTP [0:3] (0x1040; R/W)" on page 1384 0x01000000 1044h 1047h "Watchdog Software Device Status--WDSWSTS [0:3] (0x1044; R/W)" on page 1385 0x0 1048h 104Bh "Free Running Timer--FRTIMER [0:3] (0x1048; RWS)" on page 1385 0x0 0x0 104Ch 104Fh "TCP Timer--TCPTIMER [0:3] (0x104C; R/W)" on page 1386 B620h B623h "RX Time Sync Control Register--TSYNCRXCTL [0:3] (0xB620;RW)" on page 1387 0x0 B624h B627h "RX timestamp Low--RXSTMPL [0:3] (0xB624; RO)" on page 1388 0x0 B628h B62Bh "RX Timestamp High--RXSTMPH [0:3] (0xB628; RO)" on page 1388 0x0 B62Ch B62Fh "RX Timestamp Attributes Low--RXSATRL[0:3] (0xB62C; RO)" on page 1388 0x0 B630h B633h "RX timestamp Attributes High--RXSATRH [0:3] (0xB630; RO)" on page 1389 0x0 B614h B617h "TX Time Sync Control Register--TSYNCTXCTL [0:3] (0xB614; RW)" on page 1389 0x0 B618h B61Bh "TX Timestamp Value Low--TXSTMPL [0:3] (0xB618;RO)" on page 1390 B61Ch B61Fh "TX Timestamp Value High--TXSTMPH[0:3] (0xB61C; RO)" on page 1390 0x0 B6F8h B6FBh "System Time Register Residue--SYSTIMR [0:3] (0xB6F8; RW)" on page 1390 0x0 0x0 B600h B603h "System Time Register Low--SYSTIML [0:3] (0xB600; RW)" on page 1391 0x0 B604h B607h "System Time Register High--SYSTIMH [0:3] (0xB604; RW)" on page 1391 0x0 B60Ch B60Fh "Time Adjustment Offset Register Low--TIMADJL [0:3] (0xB60C; RW)" on page 1392 0x0 B610h B613h "Time Adjustment Offset Register High--TIMADJH [0:3] (0xB610;RW)" on page 1392 0x0 B640h B643h "TimeSync Auxiliary Control Register--TSAUXC [0:3] (0xB640; RW)" on page 1393 0x80000000 B644h B647h "Target Time Register 0 Low--TRGTTIML0 [0:3] (0xB644; RW)" on page 1395 0x0 B648h B64Bh "Target Time Register 0 High--TRGTTIMH0 [0:3] (0xB648; RW)" on page 1395 0x0 B64Ch B64Fh "Target Time Register 1 Low--TRGTTIML1 [0:3] (0xB64C; RW)" on page 1395 0x0 Intel(R) Communications Chipset 89xx Series - Datasheet 1256 October 2012 Order Number: 327879-001US 28.0 Table 28-5. Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR0[03]B:0:1+Index 1 Registers (Sheet 6 of 10) Offset Start Offset End Register ID - Description Default Value B650h B653h "Target Time Register 1 High--TRGTTIMH1 [0:3] (0xB650; RW)" on page 1396 0x0 B654h B657h "Frequency Out 0 Control Register--FREQOUT0 [0:3] (0xB654; RW)" on page 1396 0x0 B658h B65Bh "Frequency Out 1 Control Register--FREQOUT1 [0:3] (0xB658; RW)" on page 1397 0x0 B65Ch B65Fh "Auxiliary Time Stamp 0 Register Low--AUXSTMPL0 [0:3] (0xB65C; RO)" on page 1397 0x0 B660h B663h "Auxiliary Time Stamp 0 Register High--AUXSTMPH0 [0:3] (0xB660; RO)" on page 1398 0x0 B664h B667h "Auxiliary Time Stamp 1 Register Low--AUXSTMPL1 [0:3] (0xB664; RO)" on page 1398 0x0 B668h B66Bh "Auxiliary Time Stamp 1 Register High--AUXSTMPH1 [0:3] (0xB668; RO)" on page 1399 0x0 5F50h 5F53h "Time Sync RX Configuration--TSYNCRXCFG [0:3] (0x5F50; R/W)" on page 1399 0x0 003Ch 003Fh "Time Sync SDP Configuration Register--TSSDP [0:3] (0x003C; R/W)" on page 1400 0x0 B66Ch B66Fh "Time Sync Interrupt Cause Register--TSICR [0:3] (0xB66C; RC/W1C)" on page 1401 0x0 B674h B677h "Time Sync Interrupt Mask Register--TSIM [0:3] (0xB674; RW)" on page 1402 0x0 B670h B673h "Time Sync Interrupt Set Register--TSIS [0:3] (0xB670; WO)" on page 1403 0x0 4200h 4203h "PCS Configuration--PCS_CFG [0:3] (0x4200; R/W)" on page 1404 0x00000008 4208h 420Bh "PCS Link Control--PCS_LCTL [0:3] (0x4208; RW)" on page 1404 0x0204000C 420Ch 420Fh "PCS Link Status--PCS_LSTS [0:3] (0x420C; RO)" on page 1406 0x0000000C 4218h 421Bh "AN Advertisement--PCS_ANADV [0:3] (0x4218; R/W)" on page 1408 0x00000020 421Ch 421Fh "Link Partner Ability--PCS_LPAB [0:3] (0x421C; RO)" on page 1409 0x0 4220h 4223h "Next Page Transmit--PCS_NPTX [0:3] (0x4220; RW)" on page 1410 0x0 4224h 4227h "Link Partner Ability Next Page--PCS_LPABNP [0:3] (0x4224; RO)" on page 1411 0x0 0102Ch 0102Fh "SFP I2C Parameters--I2CPARAMS [0:3] (0x102C; R/W)" on page 1412 0x00000046 04000h 04003h "CRC Error Count--CRCERRS [0:3] (0x4000; RC)" on page 1413 0x0 04004h 04007h "Alignment Error Count--ALGNERRC [0:3] (0x4004; RC)" on page 1414 0x0 04008h 0400Bh "Symbol Error Count--SYMERRS [0:3] (0x4008; RC)" on page 1414 0x0 04010h 04013h "Missed Packets Count--MPC [0:3] (0x4010; RC)" on page 1415 0x0 04014h 04017h "Single Collision Count--SCC [0:3] (0x4014; RC)" on page 1415 0x0 04018h 0401Bh "Excessive Collisions Count--ECOL [0:3] (0x4018; RC)" on page 1416 0x0 0401Ch 0401Fh "Multiple Collision Count--MCC [0:3] (0x401C; RC)" on page 1416 0x0 04020h 04023h "Late Collisions Count--LATECOL [0:3] (0x4020; RC)" on page 1417 0x0 04028h 0402Bh "Collision Count--COLC [0:3] (0x4028; RC)" on page 1417 0x0 04030h 04033h "Defer Count--DC [0:3] (0x4030; RC)" on page 1418 0x0 0403Fh "Host Transmit Discarded Packets by MAC Count--HTDPMC [0:3] (0x403C; RC)" on 0x0 page 1418 0403Ch 04040h 04043h "Receive Length Error Count--RLEC [0:3] (0x4040; RC)" on page 1419 0x0 04048h 0404Bh "XON Received Count--XONRXC [0:3] (0x4048; RC)" on page 1419 0x0 0404Ch 0404Fh "XON Transmitted Count--XONTXC [0:3] (0x404C; RC)" on page 1420 0x0 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1257 Table 28-5. Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR0[03]B:0:1+Index 1 Registers (Sheet 7 of 10) Offset Start 04050h Offset End 04053h Default Value Register ID - Description "XOFF Received Count--XOFFRXC [0:3] (0x4050; RC)" on page 1420 0x0 04054h 04057h "XOFF Transmitted Count--XOFFTXC [0:3] (0x4054; RC)" on page 1421 0x0 04058h 0405Bh "FC Received Unsupported Count--FCRUC [0:3] (0x4058; RC)" on page 1421 0x0 0405Ch 0405Fh "Packets Received [64 Bytes] Count--PRC64 [0:3] (0x405C; RC)" on page 1422 0x0 04060h 04063h "Packets Received [65-127 Bytes] Count--PRC127 [0:3] (0x4060; RC)" on page 1422 0 04064h 04067h "Packets Received [128-255 Bytes] Count--PRC255 [0:3] (0x4064; RC)" on page 1423 0x0 04068h 0406Bh "Packets Received [256-511 Bytes] Count--PRC511 [0:3] (0x4068; RC)" on page 1424 0x0 0406Ch 0406Fh "Packets Received [512-1023 Bytes] Count--PRC1023 [0:3] (0x406C; RC)" on page 1424 0x0 04070h 04073h "Packets Received [1024 to Max Bytes] Count--PRC1522 [0:3] (0x4070; RC)" on page 1425 0x0 04074h 04077h "Good Packets Received Count--GPRC [0:3] (0x4074; RC)" on page 1425 0x0 04078h 0407Bh "Broadcast Packets Received Count - BPRC [0:3] (0x4078; RC)" on page 1426 0x0 0407Ch 0407Fh "Multicast Packets Received Count--MPRC [0:3] (0x407C; RC)" on page 1426 0x0 04080h 04083h "Good Packets Transmitted Count--GPTC [0:3] (0x4080; RC)" on page 1427 0h 04088h 0408Bh "Good Octets Received Count--GORCL [0:3] (0x4088; RC)" on page 1427 0x0 0408Ch 0408Fh "Good Octets Received Count--GORCH [0:3] (0x408C; RC)" on page 1428 0x0 04090h 04093h "Good Octets Transmitted Count--GOTCL [0:3] (0x4090; RC)" on page 1428 0x0 04094h 04097h "Good Octets Transmitted Count--GOTCH [0:3] (4094; RC)" on page 1429 0x0 040A0h 040A3h "Receive No Buffers Count--RNBC [0:3] (0x40A0; RC)" on page 1429 0x0 040A4h 040A7h "Receive Undersize Count--RUC [0:3] (0x40A4; RC)" on page 1430 0x0 040A8h 040ABh "Receive Fragment Count--RFC [0:3] (0x40A8; RC)" on page 1430 0x0 040ACh 040AFh "Receive Oversize Count--ROC [0:3] (0x40AC; RC)" on page 1431 0x0 040B0h 040B3h "Receive Jabber Count--RJC [0:3] (0x40B0; RC)" on page 1431 0x0 040B4h 040B7h "Management Packets Received Count--MNGPRC [0:3] (0x40B4; RC)" on page 1432 0x0 040B8h 040BBh "Management Packets Dropped Count--MPDC [0:3] (0x40B8; RC)" on page 1432 0x0 040BFh "Management Packets Transmitted Count--MNGPTC [0:3] (0x40BC; RC)" on page 1433 0x0 040BCh 040C0h 040C3h "Total Octets Received--TORL [0:3] (0x40C0; RC)" on page 1433 0x0 040C4h 040C7h "Total Octets Received--TORH [0:3] (0x40C4; RC)" on page 1434 0x0 040C8h 040CBh "Total Octets Transmitted--TOTL [0:3] (0x40C8; RC)" on page 1434 0x0 040CCh 040CFh "Total Octets Transmitted--TOTH [0:3] (0x40CC; RC)" on page 1435 0x0 040D0h 040D3h "Total Packets Received--TPR [0:3] (0x40D0; RC)" on page 1435 0x0 040D4h 040D7h "Total Packets Transmitted--TPT [0:3] (0x40D4; RC" on page 1436 0x0 040D8h 040DBh "Packets Transmitted [64 Bytes] Count--PTC64 [0:3] (0x40D8; RC)" on page 1436 0x0 040DCh 040DFh "Packets Transmitted [65-127 Bytes] Count--PTC127 [0:3] (0x40DC; RC)" on page 1437 0x0 040E0h 040E3h "Packets Transmitted [128-255 Bytes] Count--PTC255 [0:3] (0x40E0; RC)" on page 1437 0x0 Intel(R) Communications Chipset 89xx Series - Datasheet 1258 October 2012 Order Number: 327879-001US 28.0 Table 28-5. Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR0[03]B:0:1+Index 1 Registers (Sheet 8 of 10) Offset Start Offset End Register ID - Description Default Value 040E4h 040E7h "Packets Transmitted [256-511 Bytes] Count--PTC511 [0:3] (0x40E4; RC)" on page 1438 0x0 040E8h 040EBh "Packets Transmitted [512-1023 Bytes] Count--PTC1023 [0:3] (0x40E8; RC)" on page 1438 0x0 040ECh 040EFh "Packets Transmitted [1024 Bytes or Greater] Count--PTC1522 [0:3] (0x40EC; RC)" on page 1439 0x0 040F0h 040F3h "Multicast Packets Transmitted Count--MPTC [0:3] (0x40F0; RC)" on page 1439 0x0 040F4h 040F7h "Broadcast Packets Transmitted Count--BPTC [0:3] (0x40F4; RC)" on page 1440 0x0 040F8h 040FBh "TCP Segmentation Context Transmitted Count--TSCTC [0:3] (0x40F8; RC)" on page 1440 0x0 04100h 04103h "Interrupt Assertion Count--IAC [0:3] (0x4100; RC)" on page 1441 0x0 04104h 04107h "Rx Packets to Host Count--RPTHC [0:3] (0x4104; RC)" on page 1441 0x0 04108h 0410Bh "Debug Counter 1--DBGC1 [0:3] (0x4108; RC)" on page 1441 0x0 0410Ch 0410Fh "Debug Counter 2--DBGC2 [0:3] (0x410C; RC)" on page 1442 0x0 04110h 04113h "Debug Counter 3--DBGC3 [0:3] (0x4110; RC)" on page 1443 0x0 0x0 0411Ch 0411Fh "Debug Counter 4--DBGC4 [0:3] (0x411C; RC)" on page 1443 04118h 0411Bh "Host Good Packets Transmitted Count--HGPTC [0:3] (0x4118; RC)" on page 1444 0x0 04120h 04123h "Receive Descriptor Minimum Threshold Count--RXDMTC [0:3] (0x4120; RC)" on page 1444 0x0 04128h 0412Bh "Host Good Octets Received Count--HGORCL [0:3] (0x4128; RC)" on page 1445 0x0 0412Ch 0412Fh "Host Good Octets Received Count--HGORCH [0:3] (0x412C; RC)" on page 1445 0x0 04130h 04133h "Host Good Octets Transmitted Count--HGOTCL [0:3] (0x4130; RC)" on page 1446 0x0 04134h 04137h "Host Good Octets Transmitted Count - HGOTCH [0:3] (0x4134; RC)" on page 1446 0x0 04138h 0413Bh "Length Error Count--LENERRS [0:3] (0x4138; RC)" on page 1447 0x0 04228h 0422Bh "SerDes/SGMII/KX Code Violation Packet Count--SCVPC [0:3] (0x4228; RW)" on page 1447 0x0 041A4h 041A7h "Switch Drop Packet Count--SDPC [0:3] (0x41A4; RC)" on page 1448 0x0 10010h at 0x100 10013h at 0x100 "Per Queue Good Packets Received Count--VFGPRC [0:3][0:7](0x10010 + n*0x100 [n=0...7]; RO)" on page 1448 0x0 10014h at 0x100 10017h at 0x100 "Per Queue Good Packets Transmitted Count--VFGPTC [0:3][0:7](0x10014 + n*0x100 [n=0...7]; RO)" on page 1449 0x0 10018h at 0x100 1001Bh at 0x100 "Per Queue Good Octets Received Count--VFGORC [0:3][0:7] (0x10018 + n*0x100 [n=0...7]; RO)" on page 1450 0x0 10034h at 0x100 10037h at 0x100 "Per Queue Good Octets Transmitted Count--VFGOTC [0:3][0:7] (0x10034 + n*0x100 [n=0...7]; RO)" on page 1451 0x0 10038h at 0x100 1003Bh at 0x100 "Per Queue Multicast Packets Received Count--VFMPRC [0:3][0:7] (0x10038 + n*0x100 [n=0...7]; RO)" on page 1451 0x0 04140h 04143h "BMC Management Packets Dropped Count--BMPDC [0:3] (0x4140; RC)" on page 1452 0x0 4144h 4147h "BMC Management Packets Transmitted Count--BMNGPTC [0:3] (0x4144; RC)" on 0x0 page 1452 413Ch 413Fh "BMC Management Packets Received Count--BMNGPRC [0:3] (0x413C; RC)" on page 1453 0x0 0x0 4400h 4403h "BMC Total Unicast Packets Received--BUPRC [0:3] (0x4400; RC)" on page 1453 4404h 4407h "BMC Total Multicast Packets Received--BMPRC [0:3] (0x4404; RC)" on page 1454 0x0 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1259 Table 28-5. Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR0[03]B:0:1+Index 1 Registers (Sheet 9 of 10) Offset Start Offset End Default Value Register ID - Description 4408h 440Bh "BMC Total Broadcast Packets Received--BBPRC [0:3] (0x4408; RC)" on page 1454 0x0 440Ch 440Fh "BMC Total Unicast Packets Transmitted--BUPTC [0:3] (0x440C; RC)" on page 1454 0x0 4410h 4413h "BMC Total Multicast Packets Transmitted--BMPTC [0:3] (0x4410; RC)" on page 1455 0x0 4414h 4417h "BMC Total Broadcast Packets Transmitted--BBPTC [0:3] (0x4414; RC)" on page 1455 0x0 4418h 441Bh "BMC FCS Receive Errors--BCRCERRS [0:3] (0x4418; RC)" on page 1456 0x0 0x0 441Ch 441Fh "BMC Alignment Errors--BALGNERRC [0:3] (0x441C; RC)" on page 1456 4420h 4423h "BMC Pause XON Frames Received--BXONRXC [0:3] (0x4420; RC)" on page 1456 0x0 4424h 4427h "BMC Pause XOFF Frames Received--BXOFFRXC [0:3] (0x4424; RC)" on page 1457 0x0 4428h 442Bh "BMC Pause XON Frames Transmitted--BXONTXC [0:3] (0x4428; RC)" on page 1457 0x0 442Ch 442Fh "BMC Pause XOFF Frames Transmitted--BXOFFTXC [0:3] (0x442C; RC)" on page 1458 0x0 4430h 4433h "BMC Single Collision Transmit Frames--BSCC [0:3] (0x4430; RC)" on page 1458 0x0 4434h 4437h "BMC Multiple Collision Transmit Frames--BMCC [0:3] (0x4434; RC)" on page 1458 0x0 5800h 5803h "Wakeup Control Register--WUC [0:3] (0x5800; R/W)" on page 1459 0x0 5808h 580Bh "Wakeup Filter Control Register--WUFC [0:3] (0x5808; R/W)" on page 1460 0x0 5810h 5813h "Wakeup Status Register--WUS [0:3] (0x5810; R/W1C)" on page 1461 0x0 5900h 5903h "Wakeup Packet Length--WUPL [0:3] (0x5900; RO)" on page 1462 0x0 5A00h at 4 5A03h at 4 "Wakeup Packet Memory--WUPM [0:3][0:31] (0x5A00 + 4*n [n=0...31]; RO)" on 0x0 page 1462 5838h 583Bh "IP Address Valid--IPAV [0:3] (0x5838; R/W)" on page 1463 0x0 5840h at 8 5843h at 8 "IPv4 Address Table--IP4AT [0:3][0:3] (0x5840 + 8*n [n=0...3]; R/W)" on page 1463 0x0 5880h at 4 5883h at 4 "IPv6 Address Table--IP6AT [0:3][0:3] (0x5880 + 4*n [n=0...3]; R/W)" on page 1464 0x0 9000h at 100h at /10h 9003h at 100h at /10h "Flex Filter Even Data Register Fields--FEDR [0:3][0:3][0:15] (0x9000 +16*n[0..15]; RW)" on page 1465 0x0 9004h at 100h at /10h 9007h at 100h at /10h "Flex Filter Odd Data Register Fields--FODR [0:3][0:3][0:15] (0x9000 +16*n[0..15];RW)" on page 1466 0x0 9008h at 100h at /10h 900Ch at 100h at /10h "Flex Filter Mask Field Register--FMFR [0:3][0:3][0:15] (0x9008 +16*n[0..15];RW)" on page 1466 0x0 90FCh at 100h 90FFh at 100h "Flex Filter Queueing Field--FQFR [0:3][0:3][0:15] (0x90FC + 16*n[n=0..15] ;RW)" on page 1467 0x0 9A00h at 100h at /10h 9A03h at 100h at /10h "Flex Filter Even Data Register Extended--FHFT_EXT_FEDR [0:3][0:3][0:15] (0x9A00 +16*n[0..15]; RW)" on page 1468 0x0 9A04h at 100h at /10h 9A07h at 100h at /10h "Flex Filter Odd Data Register Extended--FHFT_EXT_FODR [0:3][0:3][0:15] (0x9A00 +16*n[0..15];RW)" on page 1469 0x0 9A08h at 100h at /10h 9A0Ch at 100h at /10h "Flex Filter Mask Field Extended--FHFT_EXT_FMFR [0:3][0:3][0:15] (0x9A08 +16*n[0..15];RW)" on page 1469 0x0 9AFCh at 100h 9AFFh at 100h "Flex Filter Queueing Extended--FHFT_EXT_FQFR [0:3][0:3] (0x9AFC;RW)" on page 1470 0x0 Intel(R) Communications Chipset 89xx Series - Datasheet 1260 October 2012 Order Number: 327879-001US 28.0 Table 28-5. Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR0[03]B:0:1+Index 1 Registers (Sheet 10 of 10) Offset Start Offset End Register ID - Description Default Value 5010h at 4 5013h at 4 "Management VLAN TAG Value--MAVTV [0:3] [0:7] (0x5010 +4*n [n=0...7]; RW)" on page 1471 0x0 5030h at 4 5033h at 4 "Management Flex UDP/TCP Ports--MFUTP [0:3][0:3] (0x5030 + 4*n [n=0...3]; RW)" on page 1471 0x0 5060h at 4 5063h at 4 "Management Ethernet Type Filters--METF [0:3][0:3] (0x5060 + 4*n [n=0...3]; RW)" on page 1472 0x0 0x0 5820h 5823h "Management Control Register--MANC [0:3] (0x5820; RW)" on page 1472 5864h 5867h "Management Only Traffic Register--MNGONLY [0:3] (0x5864; RW)" on page 1474 0x0 5890h at 4 5893h at 4 "Manageability Decision Filters--MDEF [0:3][0:7] (0x5890 + 4*n [n=0...7]; RW)" 0x0 on page 1474 5930h at 4 5933h at 4 "Manageability Decision Filters--MDEF_EXT [0:3][0:7] (0x5930 + 4*n[n=0...7]; RW)" on page 1476 0x20000000 58B0h at 4 58B3h at 4 "Manageability IP Address Filter--MIPAF [0:3][0:15] (0x58B0 + 4*n [n=0...15]; RW)" on page 1479 0x0 5910h at 8 5913h at 8 "Manageability MAC Address Low--MMAL [0:3][0:1] (0x5910 + 8*n [n= 0...1]; RW)" on page 1479 0x0 5914h at 8 5917h at 8 "Manageability MAC Address High--MMAH [0:3][0:1] (0x5914 + 8*n [n=0...1]; RW)" on page 1480 0x0 1084h 1087h "Parity and ECC Error Indication--PEIND [0:3] (0x1084; RC)" on page 1482 0x0 1088h 108Bh "Parity and ECC Indication Mask--PEINDM [0:3] (0x1088; RW)" on page 1483 0xF 3510h 3513h "DMA Transmit Descriptor Parity Status--DTPARS [0:3] (0x3510; RW1C)" on page 1483 0x0 3514h 3517h "DMA Receive Descriptor Parity Status--DRPARS [0:3] (0x3514; RW1C)" on page 1484 0x0 3518h 351Bh "Dhost Parity Status--DDPARS [0:3] (0x3518; RW1C)" on page 1484 0x0 345Ch 345Fh "Tx Packet Buffer ECC Status--TPBECCSTS [0:3] (0x345C; RW)" on page 1485 0x00010000 5F54h 5F57h "LAN Port Parity Error Control Register--LANPERRCTL [0:3] (0x5F54; RW)" on page 1485 0x0 5F58h 5F5Bh "LAN Port Parity Error Status Register--LANPERRSTS [0:3] (0x5F58; RO)" on page 1486 0x0 Table 28-6. Bus B, Device 0, Function 1 + Index 1: Summary of PCI GBEPCIBAR3[03]B:0:1+Index 1 Registers Offset Start Offset End Register ID - Description Default Value 0000h at 0x10 0003h at 0x10 "MSIX Table Entry Lower Address--MSIXTADD [0:3][0:9] (BAR3: 0x0000 + 0x10*n [n=0...9]; R/W)" on page 1320 0x0 0004h at 0x10 0007h at 0x10 "MSIX Table Entry Upper Address--MSIXTUADD [0:3][0:9] (BAR3: 0x0004 + 0x10*n [n=0...9]; R/W)" on page 1321 0x0 0008h at 0x10 000Bh at 0x10 "MSIX Table Entry Message--MSIXTMSG [0:3][0:9] (BAR3: 0x0008 + 0x10*n [n=0...9]; R/W)" on page 1321 0x0 000Ch at 0x10 000Fh at 0x10 "MSIX Table Entry Vector Control--MSIXTVCTRL [0:3][0:9] (BAR3: 0x000C + 0x10*n [n=0...9]; R/W)" on page 1322 0x00000001 2000h 2003h "MSIXPBA Bit Description--MSIXPBA [0:3] (BAR3: 0x2000; RO)" on page 1322 0x0 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1261 28.1.4 Alias Addresses Certain registers maintain an alias address designed for backward compatibility with software written previously. For these registers, the alias address is shown in Table 28-7. Those registers can be accessed by software at either the new offset or the alias offset. It is recommended that software that is written solely for the PCH use the new address offset. Table 28-7. Register Summary Offset Alias Offset Abbreviation Name RW General Interrupts 0x1510 0x00E0 IAM Interrupt Acknowledge Auto Mask RW Receive 0xC000 0x0110, 0x2800 RDBAL[0] RX Descriptor Base Low queue 0 RW 0xC004 0x0114, 0x2804 RDBAH[0] RX Descriptor Base High queue 0 RW 0xC008 0x0118, 0x2808 RDLEN[0] RX Descriptor Ring Length queue 0 RW 0xC010 0x0120, 0x2810 RDH[0] RX Descriptor Head queue 0 RO 0xC018 0x0128, 0x2818 RDT[0] RX Descriptor Tail queue 0 RW 0xC028 0x02828 RXDCTL[0] Receive Descriptor Control queue 0 RW 0xC040 + 0x40 * (n-1) 0x2900+ 0x100 * (n1) RDBAL[1 - 3] RX Descriptor Base Low queue 1 - 3 RW 0xC044 + 0x40 * (n-1) 0x2904 + 0x100 * (n1) RDBAH[1 - 3] RX Descriptor Base High queue 1 - 3 RW 0xC048 + 0x40 * (n-1) 0x2908 + 0x100 * (n1) RDLEN[1 - 3] RX Descriptor Ring Length queue 1 - 3 RW 0xC050 + 0x40 * (n-1) 0x2910 + 0x100 * (n1) RDH[1 - 3] RX Descriptor Head queue 1 - 3 RO 0xC058 + 0x40 * (n-1) 0x2918 + 0x100 * (n1) RDT[1 - 3] RX Descriptor Tail queue 1 - 3 RW 0xC068 + 0x40 * (n-1) 0x2928 + 0x100 * (n1) RXDCTL[1 - 3] Receive Descriptor Control queue 1 - 3 RW 0xC100 + 0x40 * (n- 4) N/A RDBAL[4-7] RX Descriptor Base Low queue 4 - 7 RW 0x5200- 0x53FC 0x02000x03FC MTA[127:0] Multicast Table Array (n) RW 0x5400 + 8*n 0x0040 + 8*n RAL[0-15] Receive Address Low (15:0) RW 0x5404 + 8 *n 0x0044 + 8 *n RAH[0-15] Receive Address High (15:0) RW 0x5600-0x57FC 0x06000x07FC VFTA[127:0] VLAN Filter Table Array (n) RW Intel(R) Communications Chipset 89xx Series - Datasheet 1262 October 2012 Order Number: 327879-001US 28.0 Table 28-7. Register Summary (Continued) Offset Alias Offset Abbreviation Name RW Transmit 0xE000 0x0420, 0x3800 TDBAL[0] TX Descriptor Base Low 0 RW 0xE004 0x0424, 0x3804 TDBAH[0] TX Descriptor Base High 0 RW 0xE008 0x0428, 0x3808 TDLEN[0] TX Descriptor Ring Length 0 RW 0xE010 0x0430, 0x3810 TDH[0] TX Descriptor Head 0 RO 0xE018 0x0438, 0x3818 TDT[0] TX Descriptor Tail 0 RW 0xE038 0x3838 TDWBAL[0] Transmit Descriptor WB Address Low queue 0 RW 0xE03C 0x383C TDWBAH[0] Transmit Descriptor WB Address High queue 0 RW 0xE040 + 0x40 * (n-1) 0x3900 + 0x100 * (n1) TDBAL[1-3] TX Descriptor Base Low queue 1 - 3 RW 0xE044 + 0x40 * (n-1) 0x3904 + 0x100 * (n1) TDBAH[1-3] TX Descriptor Base High queue 1 - 3 RW 0xE048 + 0x40 * (n-1) 0x3908 + 0x100 * (n1) TDLEN[1-3] TX Descriptor Ring Length queue 1 - 3 RW 0xE050 + 0x40 * (n-1) 0x3910 + 0x100 * (n1) TDH[1-3] TX Descriptor Head queue 1 - 3 RO 0xE058 + 0x40 * (n-1) 0x3918 + 0x100 * (n1) TDT[1-3] TX Descriptor Tail queue 1 - 3 RW 0xE078 + 0x40 * (n-1) 0x3938 + 0x100 * (n1) TDWBAL[1-3] Transmit Descriptor WB Address Low queue 1 - 3 RW 0xE07C + 0x40 * (n-1) 0x393C + 0x100 * (n1) TDWBAH[1-3] Transmit Descriptor WB Address High queue 1 - 3 RW 0xE1B8 + 0x40 * (n - 4) N/A TDWBAL[4 - ] Transmit Descriptor WB Address Low queue 4 - RW Filters VMDq Statistics Statistics October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1263 28.1.4.1 MSI-X BAR Register Summary Table 28-8. MSI-X Register Summary Category Offset MSI-X Table 0x0000 + n*0x10 [n=0...9] MSIXTADD MSI-X Table Entry Lower Address RW page 1320 MSI-X Table 0x0004 + n*0x10 [n=0...9] MSIXTUADD MSI-X Table Entry Upper Address RW page 1321 MSI-X Table 0x0008 + n*0x10 [n=0...9] MSIXTMSG MSI-X Table Entry Message R/W page 1321 MSI-X Table 0x000C + n*0x10 [n=0...9] MSIXTVCTRL MSI-X Table Entry Vector Control R/W page 1322 MSI-X Table 0x02000 MSIXPBA MSIXPBA Bit Description RO page 1322 28.1.4.2 Abbreviation Name RW Page Device Control Register--CTRL [0:3] (0x00000; R/W) This register, as well as the Extended Device Control register (CTRL_EXT), controls the major operational modes for the device. While software write to this register to control device settings, several bits (such as FD and SPEED) can be overridden depending on other bit settings and the resultant link configuration determined by the PHY's AutoNegotiation resolution. See Section 23.4.6 for details on the setup of these registers in the different link modes. Note: This register is also aliased at address 0x0004. Table 28-9. Device Control Register--CTRL [0:3] (0x00000; R/W) (Sheet 1 of 4) Description: View: PCI Size: 32 bit Bit Range Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x08100201 Bit Acronym 31 PHY_RST 30 VME Power Well: GBEAUX Bit Reset Value Bit Access PHY Reset For SW initiated PHY reset the driver sets this bit, resets the external PHY through MDIO interface and then clears the PHY_RST bit. 0h R/W VLAN Mode Enable When set to 1b, VLAN information is stripped from all received 802.1Q packets. 0h R/W Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 1264 Offset Start: 0000h Offset End: 0003h Sticky October 2012 Order Number: 327879-001US 28.0 Table 28-9. Device Control Register--CTRL [0:3] (0x00000; R/W) (Sheet 2 of 4) Description: View: PCI Size: 32 bit Bit Range 29 28 27 26 25 :24 23 22 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x08100201 Offset Start: 0000h Offset End: 0003h Power Well: GBEAUX Bit Reset Value Bit Access 0h R/W TFCE Transmit Flow Control Enable When set, indicates that the Controller transmits flow control packets (XON and XOFF frames) based on the receiver fullness. If Auto-Negotiation is enabled, this bit is set to the negotiated duplex value. In SerDes mode the resolution is done by the hardware. In SGMII or 1000BASE-KX modes it should be done by the software. 0h R/W RFCE Receive Flow Control Enable When set, indicates that the Controller responds to the reception of flow control packets. If Auto-Negotiation is enabled, this bit is set to the negotiated flow control value. In SerDes mode the resolution is done by the hardware. In SGMII or 1000BASE-KX modes it should be done by the software. 1h R/W RST Port Software Reset This bit performs reset to the respective port, resulting in a state nearly approximating the state following a powerup reset or internal PCIe reset, except for system PCI configuration and logic used by all ports. 0b = Normal. 1b = Reset. This bit is self clearing and is referred to as software reset or global reset. 0h R/W SDP1_IODIR SDP1 Pin Direction Controls whether software-controllable pin SDP1 is configured as an input or output (0b = input, 1b = output). Initial value is EEPROM-configurable. This bit is not affected by software or system reset, only by initial power-on or direct software writes. 0ha R/W SDP0_IODIR SDP0 Pin Direction Controls whether software-controllable pin SDP0 is configured as an input or output (0b = input, 1b = output). Initial value is EEPROM-configurable. This bit is not affected by software or system reset, only by initial power-on or direct software writes. 0ha R/W Bit Acronym Bit Description Sticky Device Reset This bit performs a reset of the entire controller device, resulting in a state nearly approximating the state following a power-up reset or internal PCIe reset, except for system PCI configuration. 0b = Normal. 1b = Reset. This bit is self clearing. DEV_RST (SC) Notes: 1. Assertion of DEV_RST generates an interrupt on all ports via the ICR.DRSTA interrupt bit. 2. Device Reset (CTRL.DEV_RST) can be used to globally reset the entire component if the DEV_RST_EN bit in Initialization Control 4 EEPROM word is set. 3. Assertion of DEV_RST sets on all ports the STATUS.DEV_RST_SET bit. For additional information see Section 28.2.1.2. Reserved October 2012 Order Number: 327879-001US Reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 1265 Table 28-9. Device Control Register--CTRL [0:3] (0x00000; R/W) (Sheet 3 of 4) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x08100201 Offset Start: 0000h Offset End: 0003h Power Well: GBEAUX Bit Reset Value Bit Access SDP0_WDE SDP0 used for Watchdog indication When set, SDP0 is used as a watchdog indication. When set, the SDP0_DATA bit indicates the polarity of the watchdog indication. In this mode, SDP0_IODIR must be set to an output. 0ha R/W ADVD3WUC D3Cold Wake up Capability Advertisement Enable When set, D3Cold wake up capability is advertised based on whether AUX_PWR advertises presence of auxiliary power (yes if AUX_PWR is indicated, no otherwise). When 0b, however, D3Cold wake up capability is not advertised even if AUX_PWR presence is indicated. The initial value is EEPROM configurable. If full 1Gb/sec. operation in D3 state is desired but the system's power requirements in this mode would exceed the D3Cold Wake up-Enabled specification limit (375mA at 3.3V), this bit can be used to prevent the capability from being advertised to the system. See bit 2 of Software Defined Pins Control (LAN Base Address + Offset 0x20) in EEPROM. a SDP1 DATA (RWS) SDP1 Data Value Used to read or write the value of software-controlled IO pin SDP1. If SDP1 is configured as an output (SDP1_IODIR = 1b), this bit controls the value driven on the pin (initial value EEPROM-configurable). If SDP1 is configured as an input, reads return the current value of the pin. 0ha R/W 18 SDP0 DATA (RWS) SDP0 Data Value Used to read or write the value of software-controlled IO pin SDP0. If SDP0 is configured as an output (SDP0_IODIR = 1b), this bit controls the value driven on the pin (initial value EEPROM-configurable). If SDP0 is configured as an input, reads return the current value of the pin. When the SDP0_WDE bit is set, this field indicates the polarity of the watchdog indication. 0ha R/W 17 SDP1_GPIEN General Purpose Interrupt Detection Enable for SDP1 If software-controlled IO pin SDP1 is configured as an input, this bit (when 1b) enables the use for GPI interrupt detection. 0h R/W 16 SDP0_GPIEN General Purpose Interrupt Detection Enable for SDP0 If software-controlled IO pin SDP0 is configured as an input, this bit (when 1b) enables the use for GPI interrupt detection. 0h R/W 0h R/W Bit Range 21 20 19 15 :13 12 Bit Acronym Bit Description Reserved Reserved Write 0, ignore on read. FRCDPLX Force Duplex When set to 1b, software can override the duplex indication from the PHY. Otherwise, in 10/100/1000Base-T link mode, the duplex setting is sampled from the PHY into the MAC on the asserting edge of the PHY LINK signal. When asserted, the CTRL.FD bit sets duplex. Intel(R) Communications Chipset 89xx Series - Datasheet 1266 B:0:1+ Index1 Sticky October 2012 Order Number: 327879-001US 28.0 Table 28-9. Device Control Register--CTRL [0:3] (0x00000; R/W) (Sheet 4 of 4) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x08100201 Bit Reset Value Bit Access 0ha R/W SPEED Speed selection. These bits determine the speed configuration and are written by software after reading the PHY configuration through the MDIO interface. These signals are ignored when Auto-Speed Detection is enabled. 00b = 10 Mb/s. 01b = 100 Mb/s. 10b = 1000 Mb/s. 11b = not used. 2h R/W ILOS Invert Loss-of-Signal (LOS/LINK) Signal Bit controls the polarity of the SRDS_[n]_SIG_DET signal or internal Link up signal depending on the CONNSW.ENRGSRC bit. 0b = Do not invert (active high input signal). 1b = Invert signal (active low input signal). Should be set to zero when working in 1000BASE-KX mode. Note: ILOS bit value is updated to Initial Value only after LAN_PWR_GOOD or PCIe reset. 0ha R/W SLU Set Link Up. See Section 22.5.4 for more information about AutoNegotiation and link configuration in the various modes. The "Set Link Up" is normally initialized to 0. However, if the APM Enable bit is set in the EEPROM then it is initialized to 1b. In SerDes and 1000Base-KX modes Link up can be forced by setting this bit as described in Section 22.5.4.1.4. 0ha R/W 0h R/W 1ha R/W Bit Description 11 FRCSPD Force Speed This bit is set when software needs to manually configure the MAC speed settings according to the SPEED bits. MAC and PHY must resolve to the same speed configuration or software must manually set the PHY to the same speed as the MAC. Software must clear this bit to enable the PHY or ASD function to control the MAC speed setting. This bit is superseded by the CTRL_EXT.SPD_BYPS bit which has a similar function. 10 Reserved Reserved Write as 0b to ensure future compatibility. 07 06 05 :03 Reserved Sticky Reserved Write 0 ignore on read. When set to 1b, the function of this bit blocks new master requests including manageability requests. If no master requests are pending by this function, the STATUS.GIO Master Enable Status bit is set. See Section 25.2.3.2 for further information. 02 GIO Master Disable 01 Reserved This bit is reserved and should be set to 0b for future compatibility. FD Full-Duplex Controls the MAC duplex setting when explicitly set by software. 0b = half duplex. 1b = full duplex. 00 Offset Start: 0000h Offset End: 0003h Power Well: GBEAUX Bit Acronym 09 :08 B:0:1+ Index1 a. These bits are loaded from EEPROM. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1267 28.1.4.3 Device Status Register--STATUS [0:3] (0x0008; R) Table 28-10. Device Status Register-- STATUS[0:3] (0x0008; R) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 30 :21 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x80080400 Bit Acronym MAC clock gating Enable Reserved Offset Start: 00008h Offset End: 0000Bh Power Well: GBEAUX Bit Description MAC clock gating Enable bit loaded from the EEPROMindicates the device support gating of the MAC clock. Sticky Bit Reset Value Bit Access 1ha R Reserved 20 Device Reset Set DEV_RST_SET When set indicates that a device reset (CTRL.DEV_RST) (R/W1C) was initiated by one of the software drivers. Note: Bit cleared by write 1. 0h R 19 Cleared by the Controller when the CTRL.GIO Master Disable bit is set and no master requests are pending by GIO Master this function and is set otherwise. Indicates that no Enable Status master requests are issued by this function as long as the CTRL.GIO Master Disable bit is set. 1h R X R TXOFF Transmission Paused This bit indicates the state of the transmit function when symmetrical flow control has been enabled and negotiated with the link partner. This bit is set to 1b when transmission is paused due to the reception of an XOFF frame. It is cleared (0b) upon expiration of the pause timer or the receipt of an XON frame. X R LAN ID LAN ID Provides software a mechanism to determine the LAN identifier for the MAC. 00b = LAN 0. 01b = LAN 1. 10b = LAN 2. 11b = LAN 3. X R 18 :10 Reserved 09 :08 ASDV 07 :05 Reserved 04 03 :02 Reserved Auto-Speed Detection Value Speed result sensed by the Controller'sMAC autodetection function. These bits are provided for diagnostics purposes only. The ASD calculation can be initiated by software writing a logic 1b to the CTRL_EXT.ASDCHK bit. The resultant speed detection is reflected in these bits. See Section 28.1.4.4 for details. Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1268 B:0:1+ Index1 October 2012 Order Number: 327879-001US 28.0 Table 28-10. Device Status Register-- STATUS[0:3] (0x0008; R) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 01 00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x80080400 Offset Start: 00008h Offset End: 0000Bh Power Well: GBEAUX Bit Reset Value Bit Access LU Link Up. 0 = no link established; 1 = link established. For this to be valid, the Set Link Up bit of the Device Control Register (CTRL.SLU) must be set. Link up provides a useful indication of whether something is attached to the port. Successful negotiation of features/ link parameters results in link activity. The link startup process (and consequently the duration for this activity after reset) can be several 100's of ms. When the SerDes, SGMII or 1000BASE-KX interface is used, this indicates loss-of-signal; if Auto-Negotiation is also enabled, this can also indicate successful Auto-Negotiation. See Section 22.5.4 for more details. Note: In SerDes mode bit is always 0. X R FD Full Duplex. 0 = Half duplex (HD). 1= Full duplex (FD). Reflects duplex setting of the MAC and/or link. FD reflects the actual MAC duplex configuration. This normally reflects the duplex setting for the entire link, as it normally reflects the duplex configuration negotiated between the PHY and link partner (copper link) or MAC and link partner (fiber link). X R Bit Acronym Bit Description Sticky a. If the signature bits of the EEPROM's Initialization Control Word 1 match (01b), this bit is read from the EEPROM. 28.1.4.4 Extended Device Control Register--CTRL_EXT [0:3] (0x0018; R/W) This register provides extended control of the Controller's functionality beyond that provided by the Device Control register (CTRL). Table 28-11. Extended Device Control Register--CTRL_EXT [0:3] (0x0018; R/W) (Sheet 1 of 3) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x00500000 Bit Range Bit Acronym 31 :29 Reserved 28 DRV_LOAD 27 Reserved October 2012 Order Number: 327879-001US Offset Start: 00018h Offset End: 0001Bh Power Well: GBEAUX Bit Reset Value Bit Access Driver Loaded This bit should be set by the driver after it is loaded. This bit should be cleared when the driver unloads or after a PCIe soft reset. The MNG controller reads this bit to indicate to the manageability controller (BMC) that the driver has loaded. 0h R/W Reserved 0h R/W Bit Description Sticky Reserved Write 0, Ignore on read. Intel(R) Communications Chipset 89xx Series - Datasheet 1269 Table 28-11. Extended Device Control Register--CTRL_EXT [0:3] (0x0018; R/W) (Sheet 2 of 3) Description: View: PCI Size: 32 bit Bit Range 26 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x00500000 Offset Start: 00018h Offset End: 0001Bh Power Well: GBEAUX Bit Reset Value Bit Access Extended VLAN When set, all incoming Rx packets are expected to have at least one VLAN with the Ether type as defined in VET.EXT_VET that should be ignored. The packets can have a second VLAN that should be used for all filtering purposes. All Tx packets are expected to have at least one VLAN added to them by the host. In the case of an additional VLAN request (VLE - VLAN Enable is set in transmit descriptor) the second VLAN is added after the first VLAN is added by the host. This bit is reset only by a power up reset or by an EEPROM full auto load and should only be changed while Tx and Rx processes are stopped. 0h1 R/W Enable I2C This bit enables the SFPx_I2C pins that can be used to access external SFP modules or an external 1000BASE-T PHY via the MDIO interface. If cleared, the SFPx_I2C pads are isolated and accesses to the SFPx_I2C pins through the I2CCMD register or the MDIC register are ignored. 0h1 R/W 0x00 R/W 0h1 R/W 0h1 R/W RO_DIS Relaxed Ordering Disabled When set to 1b, the Controller does not request any relaxed ordering transactions on the PCIe interface regardless of the state of bit 4 in the PCIe Device Control register. When this bit is cleared and bit 4 of the PCIe Device Control register is set, the Controller requests relaxed ordering transactions as specified by registers RXCTL and TXCTL (per queue and per flow). 0h R/W NS_DIS No Snoop Disable When set to 1b, the Controller does not set the no snoop attribute in any PCIe packet, independent of PCIe configuration and the setting of individual no snoop enable bits. When set to 0b, behavior of no snoop is determined by PCIe configuration and the setting of individual no snoop enable bits. 0h R/W Bit Acronym Bit Description Extended VLAN 25 I2C Enabled 24 Reserved Sticky Reserved. Write 0, ignore on read. Link Mode Controls interface on the link. 23 :22 LINK_MODE 00b = Default Note: This is the lowest power mode. This is default to keep MAC in lowest power mode by default. 01b = 1000BASE-KX. 10b = SGMII. 11b = SerDes interface. 21 Reserved Reserved. Should be set to 0b. 20 Reserved Reserved 19 Dynamic MAC Clock Gating 18 17 16 When set, enables Dynamic MAC Clock Gating. SerDes Low Power When set, allows the SerDes to enter a low power Enable state when the function is in Dr state. Intel(R) Communications Chipset 89xx Series - Datasheet 1270 October 2012 Order Number: 327879-001US 28.0 Table 28-11. Extended Device Control Register--CTRL_EXT [0:3] (0x0018; R/W) (Sheet 3 of 3) Description: View: PCI BAR: GBEPCIBAR0[0:3] Size: 32 bit Bit Range Bus:Device:Function: Default: 0x00500000 Bit Reset Value Bit Access 0h R/W EE_RST EEPROM Reset When set, initiates a reset-like event to the EEPROM function. This causes the EEPROM to be read as if a RST# assertion had occurred. All Controller functions should be disabled prior to setting this bit. This bit is self-clearing. 0h RW1C ASDCHK ASD Check Initiates an Auto-Speed-Detection (ASD) sequence to sense the frequency of the PHY receive clock (RX_CLK). The results are reflected in STATUS.ASDV. This bit is self-clearing. 0h RW1C Reserved Reserved. Should be written as 0b to ensure future compatibility. 0h R/W Bit Description 15 SPD_BYPS Speed Select Bypass When set to 1b, all speed detection mechanisms are bypassed, and the Controller is immediately set to the speed indicated by CTRL.SPEED. This provides a method for software to have full control of the speed settings of the Controller and when the change takes place, by overriding the hardware clock switching circuitry. 14 Reserved 12 11 :00 Offset Start: 00018h Offset End: 0001Bh Power Well: GBEAUX Bit Acronym 13 B:0:1+ Index1 Sticky Reserved The PCH allows up to two externally controlled interrupts. All software-definable pins, these can be mapped for use as GPI interrupt bits. Mappings are enabled by the SDPx_GPIEN bits only when these signals are also configured as inputs via SDPx_IODIR. When configured to function as external interrupt pins, a GPI interrupt is generated when the corresponding pin is sampled in an active-high state. The bit mappings are shown in the table bellow for clarity. Table 28-12. Mappings for SDI Pins Used as GPI SDP Pin Used as GPI Resulting ICR Bit (GPI) CTRL_EXT Field Settings Direction Enable as GPI interrupt 1 SDP1_IODIR SDP1_GPIEN 12 0 SDP0_IODIR SDP0_GPIEN 11 Note: If software uses the EE_RST function and desires to retain current configuration information, the contents of the control registers should be read and stored by software. Control register values are changed by a read of the EEPROM which occurs upon assertion of the EE_RST bit. Note: The EEPROM reset function can read configuration information out of the EEPROM which affects the configuration of PCIe space BAR settings. The changes to the BARs are not visible unless the system reboots and the BIOS is allowed to re-map them. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1271 Note: The SPD_BYPS bit performs a similar function to the CTRL.FRCSPD bit in that the Controller's speed settings are determined by the value software writes to the CRTL.SPEED bits. However, with the SPD_BYPS bit asserted, the settings in CTRL.SPEED take effect immediately rather than waiting until after the Controller's clock switching circuitry performs the change. 28.1.4.5 MDI Control Register--MDIC [0:3] (0x0020; R/W) Software uses this register to read or write Management Data Interface (MDI) registers in an external SGMII PHY. See Section 22.5.2.2.2 for details on usage of this register. Table 28-13. MDI Control Register--MDIC[0:3] (0x0020; R/W) Description: View: PCI Size: 32 bit Bit Range 31 Default: 0x10000000 Power Well: GBEAUX Bit Reset Value Bit Access 0h R/W Error This bit is set to 1b by hardware when it fails to complete an MDI read. Software should make sure this bit is clear (0h) before issuing an MDI read or write command. Note: bit is valid only when the Ready bit is set. 0h R/W Bit Acronym Reserved Offset Start: 00020h Offset End: 00023h Bit Description Reserved. Sticky 30 MDI_ERR (RWS) 29 MDI_IE Interrupt Enable When set to 1 an Interrupt is generated at the end of an MDI cycle to indicate an end of a read or write operation to the PHY. 0h R/W R (RWS) Ready Bit Set to 1b by the Controller at the end of the MDI transaction (for example, indication of a Read or Write completion). It should be reset to 0b by software at the same time the command is written. 1h R/W Opcode 01b = MDI Write 10b = MDI Read All other values are reserved. 0x0 R/W 0x0 R/W X R/W 28 27 :26 OP 25 :21 Reserved 20 :16 REGADD 15 :00 DATA Reserved. PHY Register Address: Reg. 0, 1, 2,...31 Data In a Write command, software places the data bits and the MAC shifts them out to the PHY. In a Read command, the MAC reads these bits serially from the PHY and software can read them from this location. Intel(R) Communications Chipset 89xx Series - Datasheet 1272 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] October 2012 Order Number: 327879-001US 28.0 28.1.4.6 MDC/MDIO Configuration Register--MDICNFG [0:3] (0x0E04; R/W) This register is used to configure the MDIO connection that is accessed via the MDIC register. See Section 22.5.2.2.2 for details on usage of this register. Table 28-14. MDC/MDIO Configuration Register--MDICNFG [0:3] (0x0E04; R/W) Description: View: PCI Size: 32 bit Bit Range 31 30 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 00E04h Offset End: 00E07h Power Well: GBEAUX Bit Reset Value Bit Access Destinationa Destination 0b = Reserved. 1b = The MDIO transaction is directed to the external MDIO pins (I2C Interface). Note: * When PHY registers access is initiated via the I2CCMD interface, access is always via the external I2C Interface. In this case the destination field should always be 0. 0h R/W Com_MDIOb When interfacing an external SGMII PHY bit defines if MDIO access is routed to the common MDIO port on LAN 0, to support multi port external PHYs, or to the dedicated per function MDIO port. 0b - MDIO access routed to the LAN port's MDIO interface. 1b - MDIO accesses on this LAN port routed to LAN port 0 MDIO interface 0h R/W Bit Acronym Bit Description 29 :26 Reserved Reserved. Write 0, ignore on read. 25 :21 PHYADDc External PHY Address 20 :00 Reserved Reserved. Write 0, ignore on read. Sticky 0x00 0x01 0x02 0x03 - LAN LAN LAN LAN 0h 1h 2h 3h R/W a. Destination Loaded from EEPROM Initialization Control 3 word. When external PHY supports a MDIO interface bit is 1, otherwise bit is 0. b. Common MDIO usage configuration bit is loaded from Initialization Control 3 EEPROM word. c. PHYADD Loaded from Initialization Control 4 EEPROM word to allocate per port address when using common MDIO port. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1273 28.1.4.7 Copper/Fiber Switch Control--CONNSW [0:3] (0x0034; R/W) Table 28-15. Copper/Fiber Switch Control--CONNSW[0:3] (0x0034; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bit Acronym 31 :10 Reserved 08 :04 03 02 01 00 B:0:1+ Index1 Default: 0x0 Bit Range 09 Bus:Device:Function: SerDesD (RO) Reserved ASCLR_DIS ENRGSRC Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0h RO Reserved N SerDes Signal Detect Indication Indicates the SerDes signal detect value according to the selected source. Valid only if LINK_MODE is SerDes, 1000BASE-KX or SGMII. N Reserved N Auto sense clear disable N 0h R/W SerDes Energy Detect Source If set, the OMED interrupt cause is set after asserting the external signal detect pin. If cleared, the OMED interrupt cause is set after exiting from electrical idle of the SerDes receiver. This bit also defines the source of the signal detect indication used to set link up while is SerDes mode. N 0ha R/W AUTOSENSE_CONF Auto Sense Configuration Mode AUTOSENSE_EN Offset Start: 0034h Offset End: 0037h Auto Sense Enable N 0h R/W N 0h R/W a. The default value of the ENRGSRC bit in this register is defined in the Initialization Control 3 (Offset 0x24) EEPROM word (bit 15). 28.1.4.8 VLAN Ether Type--VET [0:3] (0x0038; R/W) This register contains the type field hardware matches against to recognize an 802.1Q (VLAN) Ethernet packet and uses when add and transmit VLAN Ethernet packets. To be compliant with the 802.3ac standard, this register should be programmed with the value 0x8100. For VLAN transmission the upper byte is first on the wire (VET[15:8]). Table 28-16. VLAN Ether Type--VET [0:3] (0x0038; R/W) Description: View: PCI Size: 32 bit Default: 0x81008100 Bit Range Bit Acronym 31 :16 VET EXT 15 :00 VET Offset Start: 0038h Offset End: 003Bh Power Well: GBEAUX Bit Reset Value Bit Access External VLAN Ether Type. 0x8100 R/W VLAN EtherType Should be programmed with 0x8100. 0x8100 RO Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 1274 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Sticky October 2012 Order Number: 327879-001US 28.0 28.1.4.9 LED Control--LEDCTL [0:3] (0x0E00; RW) This register controls the setup of the LEDs. See Section 26.5.1 for details of the MODE fields encoding. Table 28-17. LED Control--LEDCTL [0:3](0x0E00; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] LED0_BLINK LED0/LINK# Blink This field specifies whether to apply blink logic to the (possibly inverted) LED control source prior to the LED output. 0b = Do not blink asserted LED output. 1b = Blink asserted LED output. 0h1 RW LED0_IVRT LED0/LINK# Invert This field specifies the polarity/ inversion of the LED source prior to output or blink control. 0b = Do not invert LED source. 1b = Invert LED source. 0h1 RW 0h1 RW 2ha RW Reserved 04 03 :00 Power Well: GBEAUX Bit Access 31 :08 05 Offset Start: E00h Offset End: E03h Bit Reset Value Bit Acronym 06 B:0:1+ Index1 Default: 0x07568302 Bit Range 07 Bus:Device:Function: Bit Description Sticky Reserved Write as 0 ignore on read. Global Blink Mode GLOBAL_BLINK_M This field specifies the blink mode of all the LEDs. ODE 0b = Blink at 200 ms on and 200 ms off. 1b = Blink at 83 ms on and 83 ms off. Reserved LED0_MODE Reserved LED0/LINK# Mode This field specifies the control source for the LED0 output. An initial value of 0010b selects LINK_UP# indication. a. These bits are read from the EEPROM. 28.2 Internal Packet Buffer Size Registers The following registers define the size of the on-chip receive and transmit buffers used to receive and transmit packets. The overall available internal buffer size in the Controller for all ports is 144 KB for receive buffers and 80 KB for transmit Buffers. Disabled ports memory can be shared between active ports and sharing can be asymmetric. The default buffer size for each port is loaded from the EEPROM on initialization. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1275 28.2.1 Detailed Register Descriptions 28.2.1.1 Internal Receive Packet Buffer Size--IRPBS [0:3](0x2404; RO) Table 28-18. Internal Receive Packet Buffer Size--IRPBS [0:3] (0x2404; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :04 Reserved 03 :00 Bus:Device:Function: RXPbsizea Offset Start: 2404h Offset End: 2407h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RO Reserved Receive internal buffer size: 0x0 - 36 KB 0x1 - 72 KB 0x2 - 144 KB 0x3 - 1 KB 0x4 - 2 KB 0x5 - 4 KB 0x6 - 8 KB 0x7 - 16 KB 0x8 - 35 KB 0x9 - 70 KB 0xA - 140 KB 0xB:0xF - reserved Notes: 1. When 4 ports are active maximum buffer size can be 36 KB. When 2 ports are active maximum buffer size can be 72 KB. When only a single port is active maximum buffer size can be 144 KB. For further information see Section 26.1.3.2. 2. Values bellow 35 KB should be used for diagnostic purposes only. 3. When port is disabled for both PCIe and Management access, the buffer size allocated to the port is 0 Bytes. Available internal memory can be used by other ports. 4. When BMC to Host traffic is enabled maximum available receive buffer for all ports is 140 KB. a. Value loaded from Initialization Control 4 EEPROM word. Intel(R) Communications Chipset 89xx Series - Datasheet 1276 October 2012 Order Number: 327879-001US 28.0 28.2.1.2 Internal Transmit Packet Buffer Size--ITPBS [0:3] (0x3404; RO) Table 28-19. Internal Transmit Packet Buffer Size--ITPBS [0:3] (0x3404; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x8 Offset Start: 3404h Offset End: 3407h Power Well: GBEAUX Bit Range Bit Acronym 31 :04 Reserved Reserved TXPbsize Transmit internal buffer sizea 0x0 - 20 KB 0x1 - 40 KB 0x2 - 80 KB 0x3 - 1 KB 0x4 - 2 KB 0x5 - 4 KB 0x6 - 8 KB 0x7 - 16 KB 0x8 - 19 KB 0x9 - 38 KB 0xA - 76 KB 0xB:0xF - reserved Notes: 1. When 4 ports are active maximum buffer size can be 20KB. When only 2 ports are active maximum buffer size is 40KB. When only a single port is active maximum buffer size is 80KB. 2. Values bellow 20 KB should be used for diagnostic purposes only. 3. When port is disabled for both PCIe and management access, the buffer size allocated to the port is 0 Bytes. Available internal memory can be used by other ports. 03 :00 B:0:1+ Index1 Bit Description Sticky Bit Reset Value Bit Access 0x8 RO a. Value loaded from Initialization Control 4 EEPROM word. 28.3 EEPROM Registers 28.3.1 Detailed Register Descriptions 28.3.1.1 EEPROM Control and Data Register--EEC [0:3] (0x0010; R/W) This register provides software direct access to the EEPROM. Software can control the EEPROM by successive writes to this register. Data and address information is clocked into the EEPROM by software toggling the EE_SK and EE_DI bits (0 and 2) of this register with EE_CS set to 0b. Data output from the EEPROM is latched into the EE_DO bit (bit 3) via the internal 62.5 MHz clock and can be accessed by software via reads of this register. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1277 Table 28-20. EEPROM Control Register--EEC [0:3] (0x0010; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: Xh Offset Start: 0010h Offset End: 0013h Power Well: GBEAUX Bit Reset Value Bit Access 0x0 R/W 2h R/W EE_ADDR_SIZE EEPROM Address Size This field defines the address size of the EEPROM. This bit is set by the EEPROM size auto-detect mechanism. If no EEPROM is present or the signature is not valid, a 16-bit address is assumed. 0b = 8- and 9-bit. 1b = 16-bit. 0h R/W Auto_RD (RO) EEPROM Auto Read Done When set to 1b, this bit indicates that the auto read by hardware from the EEPROM is done. This bit is also set when the EEPROM is not present or when its signature is not valid. 0h R/W 08 EE_PRES (RO) EEPROM Present This bit indicates that an EEPROM is present by monitoring the EE_DO input for an active-low acknowledge by the serial EEPROM during initial EEPROM scan. 1b = EEPROM present. X R/W 07 EE_GNT Grant EEPROM Access When this bit is 1b the software can access the EEPROM using the SK, CS, DI, and DO bits. 0h R/W 06 EE_REQ Request EEPROM Access The software must write a 1b to this bit to get direct EEPROM access. It has access when EE_GNT is 1b. When the software completes the access it must write a 0b. 0h R/W Data output bit from the EEPROMb The EE_DO input signal is mapped directly to this bit in the register and contains the EEPROM data output. This bit is RO from a software perspective; writes to this bit have no effect. X R/W Data input to the EEPROM When EE_GNT = 1b, the EE_DI output signal is mapped directly to this bit. Software provides data input to the EEPROM via writes to this bit. 0h R/W Bit Range Bit Acronym 31 :16 Reserved 15 Bit Description Sticky Reserved Write 0 ignore on read. EPROM access blocked EEPROM access blocked - Bit is set by HW when bit EE_BLOCKED (R/ banging transactions are blocked due to write to read-only sections or any access to hidden area. W1C) Note: Bit is cleared by write one. EEPROM Size This field defines the size of the EEPROM: 14 :11a 10 09 05 :04 EE_SIZE (RO) FWE 03 EE_DO (RO) 02 EE_DI Field Value EEPROM Size 0111b 16 Kbytes 1000b 32 Kbytes 1011b - 1111b Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1278 EEPROM Address Size 2 bytes 2 bytes October 2012 Order Number: 327879-001US 28.0 Table 28-20. EEPROM Control Register--EEC [0:3] (0x0010; R/W) Description: View: PCI Size: 32 bit Bit Range 01 00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: Xh Offset Start: 0010h Offset End: 0013h Power Well: GBEAUX Bit Reset Value Bit Access EE_CS Chip select input to the EEPROM When EE_GNT = 1b, the EE_CS output signal is mapped to the chip select of the EEPROM device. Software enables the EEPROM by writing a 1b to this bit. 0h R/W EE_SK Clock input to the EEPROM When EE_GNT = 1b, the EE_SK output signal is mapped to this bit and provides the serial clock input to the EEPROM. Software clocks the EEPROM via toggling this bit with successive writes. 0h R/W Bit Acronym Bit Description Sticky a. These bits are read from the EEPROM. b. Value depends on voltage level on EE_DO pin following initialization 28.3.1.2 EEPROM Read Register - EERD [0:3] (0x0014; RW) This register is used by software to cause the Controller to read individual words in the EEPROM. To read a word, software writes the address to the Read Address field and simultaneously writes a 1b to the Start Read field. The Controller reads the word from the EEPROM and places it in the Read Data field, setting the Read Done field to 1b. Software can poll this register, looking for a 1b in the Read Done field, and then using the value in the Read Data field. When this register is used to read a word from the EEPROM, that word does not influence any of the Controller's internal registers even if it is normally part of the autoread sequence. Table 28-21. EEPROM Read Register--EERD [0:3] (0x0014; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :16 DATA (RO) 15 :02 ADDR October 2012 Order Number: 327879-001US Offset Start: 0014h Offset End: 0017h Power Well: GBEAUX Bit Description Sticky Read Data. Data returned from the EEPROM read. Read Address This field is written by software along with Start Read to indicate the word to read. Bit Reset Value Bit Access X RW 0x0 RW Intel(R) Communications Chipset 89xx Series - Datasheet 1279 Table 28-21. EEPROM Read Register--EERD [0:3] (0x0014; RW) Description: View: PCI Size: 32 bit Bit Range 01 00 28.3.1.3 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 0014h Offset End: 0017h Power Well: GBEAUX Bit Reset Value Bit Access DONE (RO) Read Done Set to 1b when the EEPROM read completes. Set to 0b when the EEPROM read is not completed. Writes by software are ignored. Reset by setting the START bit. 0h RW START Start Read Writing a 1b to this bit causes the EEPROM to read a (16-bit) word at the address stored in the EE_ADDR field and then storing the result in the EE_DATA field. This bit is self-clearing. 0h RW1C Bit Acronym Bit Description Sticky EEPROM Diagnostic--EEDIAG [0:3] (0x1038; RO) This register reflects the values of EEPROM bits influencing the hardware that are not reflected otherwise. Table 28-22. EEPROM Diagnostic--EEDIAG [0:3] (0x1038; RO) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 30 :29 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Acronym Power Well: GBEAUX Bit Description a deadlock condition was detected in the Deadlock Release Indicates EEPROM and the current grant was released. Reserved Offset Start: 1038h Offset End: 103Bh Sticky Bit Reset Value Bit Access X RO Reserved 28 Reflects bit 10 in the Software Defined Pins Control LAN3 PCI Disable (Offset 0x20) EEPROM word controlling the disabling of LAN3 as PCIe function. 0h RO 27 Reflects bit 10 in the Software Defined Pins Control LAN2 PCI Disable (Offset 0x20) EEPROM word controlling the disabling of LAN2 as PCIe function. 0h RO 26 :25 Reserved Reserved 24 LAN3 Disable Reflects bit 11 in the Software Defined Pins Control (Offset 0x20) EEPROM word controlling the disabling of LAN3. 0h RO 23 LAN2 Disable Reflects bit 11 in the Software Defined Pins Control (Offset 0x20) EEPROM word controlling the disabling of LAN2. 0h RO X RO 22 :16 Reserved 15 :14 EEPROM Serial State Reserved State of the EEPROM serial access arbitration state machine. Intel(R) Communications Chipset 89xx Series - Datasheet 1280 B:0:1+ Bus:Device:Function: Index1 October 2012 Order Number: 327879-001US 28.0 Table 28-22. EEPROM Diagnostic--EEDIAG [0:3] (0x1038; RO) (Sheet 2 of 2) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Bit Acronym Bit Description 13 :12 EEPROM Parallel State State of the EEPROM parallel access arbitration state machine. Reserved 06 PLL Shutdown Enable 05 Reserved Offset Start: 1038h Offset End: 103Bh Power Well: GBEAUX Bit Range 11 :07 B:0:1+ Index1 Sticky Bit Reset Value Bit Access X RO 0h RO Reserved Reflects bit 4 in EEPROM 0x0F controlling the PLL shutdown enable control. Reserved 04 EEPROM Deadlock Reflects bit 5 in EEPROM word 0x0A controlling the Release Enable EEPROM deadlock release enable. 0h RO 03 Reflects bit 10 in the Software Defined Pins Control LAN1 PCI Disable (Offset 0x20) EEPROM word 0x10 controlling the disabling of the LAN1 PCIe function. 0h RO 0h RO 02 01 00 28.3.1.4 LAN1 Disable Reserved Reflects bit 11 in the Software Defined Pins Control (Offset 0x20) EEPROM word controlling the disabling of LAN1. Reserved VPD Diagnostic Register--VPDDIAG [0:3] (0x1060; RO) This register stores the VPD parameters as parsed by the auto-load process. This register is used for debug only. Table 28-23. VPD Diagnostic Register--VPDDIAG [0:3] (0x1060; RO) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Offset Start: 1060h Offset End: 1063h Power Well: GBEAUX Bit Reset Value Bit Access Offset of the End tag in VPD relative to the start of VPD (in bytes). X RO WR Tag Offset of the Write tag in VPD relative to the start of VPD (in bytes). X RO 13 :05 RD Tag Offset of the Read tag in VPD relative to the start of VPD (in bytes). X RO 04 :01 Reserved X RO Bit Range Bit Acronym 31 :23 End Tag 22 :14 00 Valid October 2012 Order Number: 327879-001US Bit Description Sticky Reserved VPD structure valid Intel(R) Communications Chipset 89xx Series - Datasheet 1281 28.3.1.5 Management--EEPROM CSR I/F The following registers are reserved for Firmware access to the EEPROM and are not writable by the host. 28.3.1.6 Management EEPROM Control Register--EEMNGCTL [0:3] (0x1010; RO) Table 28-24. Management EEPROM Control Register--EEMNGCTL [0:3] (0x1010; RO) (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x80000000 Offset Start: 1010h Offset End: 1013h Power Well: GBEAUX Bit Reset Value Bit Access 1h RO CFG_DONE 3 Configuration cycle is done for port 3a - This bit indicates that the configuration cycle (configuration of SerDes, PCIe and PLLs) is done for port 3. This bit is set to 1b to indicate configuration done, and cleared by hardware on any of the reset sources that cause initialization of the PHY. Note: Port 3 driver should not try to access the PHY for configuration before this bit is set. 0h RO CFG_DONE 2 Configuration cycle is done for port 2a - This bit indicates that the configuration cycle (configuration of SerDes, PCIe and PLLs) is done for port 2. This bit is set to 1b to indicate configuration done, and cleared by hardware on any of the reset sources that cause initialization of the PHY. Note: Port 2 driver should not try to access the PHY for configuration before this bit is set. 0h RO CFG_DONE 1 Configuration cycle is done for port 1a - This bit indicates that configuration cycle (configuration of SerDes, PCIe and PLLs) is done for port 1. This bit is set to 1b to indicate configuration done, and cleared by hardware on any of the reset sources that cause initialization of the PHY. Note: Port 1 driver should not try to access the PHY for configuration before this bit is set. 0h RO 18 CFG_DONE 0 Configuration cycle is done for port 0a - This bit indicates that configuration cycle (configuration of SerDes, PHY, PCIe and PLLs) is done for port 0. This bit is set to 1b to indicate configuration done, and cleared by hardware on any of the reset sources that causes initialization of the PHY. Note: Port 0 driver should not try to access the PHY for configuration before this bit is set. 0h RO 17 EEBUSY EEPROM Busy - This bit indicates that the EEPROM is busy processing an EEPROM transaction and shouldn't be accessed. 0h RO 16 WRITE Write - This bit tells the EEPROM if the current operation is read or write: 0b = read 1b = write 0h RO Bit Range 31 30 :22 21 20 19 Bit Acronym Bit Description DONE Transaction Done - This bit is cleared after Start Write or Start Read bit is set by the MNG and is set back again when the EEPROM write or read transaction is done. Reserved Sticky Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1282 Bus:Device:Function: October 2012 Order Number: 327879-001US 28.0 Table 28-24. Management EEPROM Control Register--EEMNGCTL [0:3] (0x1010; RO) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 15 14 :00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x80000000 Offset Start: 1010h Offset End: 1013h Power Well: GBEAUX Bit Reset Value Bit Access Start - Writing a 1b to this bit causes the EEPROM to start the read or write operation according to the write bit. 0h RO Address - This field is written by MNG along with Start Read or Start write to indicate the EEPROM address to read or write. 0x0 RO Bit Acronym Bit Description START ADDR Sticky a. Bit relates to physical port. If LAN Function Swap (FACTPS.LAN Function Sel = 1) is done, Software should poll CFG_DONE bit of original port to detect end of PHY configuration operation. 28.3.1.7 Management EEPROM Read/Write Data--EEMNGDATA [0:3] (0x1014; RO) Table 28-25. Management EEPROM Read/Write Data--EEMNGDATA [0:3] (0x1014; RO) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym 31 :16 RDDATA Read Data Data returned from the EEPROM read. 15 :00 WRDATA Write Data Data to be written to the EEPROM. October 2012 Order Number: 327879-001US Offset Start: 1014h Offset End: 1017h Bit Description Sticky Bit Reset Value Bit Access X RO 0x0 RO Intel(R) Communications Chipset 89xx Series - Datasheet 1283 28.4 Flow Control Registers 28.4.1 Detailed Register Descriptions 28.4.1.1 Flow Control Address Low--FCAL [0:3] (0x0028; RO) Flow control packets are defined by 802.3X to be either a unique multicast address or the station address with the Ether Type field indicating PAUSE. The FCA registers provide the value hardware uses to compare incoming packets against, to determine that it should PAUSE its output. The FCAL register contains the lower bits of the internal 48-bit Flow Control Ethernet address. All 32 bits are valid. Software need to access the High and Low registers as separate 32-bit read operations. The complete flow control multicast address is: 0x01_80_C2_00_00_01; where 0x01 is the first byte on the wire, 0x80 is the second, etc. Note: Any packet matching the contents of {FCAH, FCAL, FCT} when CTRL.RFCE is set is acted on by the Controller. Whether flow control packets are passed to the host (software) depends on the state of the RCTL.DPF bit and whether the packet matches any of the normal filters. Table 28-26. Flow Control Address Low--FCAL [0:3] (0x0028; RO) Description: View: PCI Size: 32 bit Default: 0x00C28001 Bit Range Bit Acronym 31 :00 FCAL Offset Start: 0028h Offset End: 002Bh Power Well: GBEAUX Bit Description Flow Control Address Low Intel(R) Communications Chipset 89xx Series - Datasheet 1284 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Sticky Bit Reset Value Bit Access 0x00C28001 RO October 2012 Order Number: 327879-001US 28.0 28.4.1.2 Flow Control Address High--FCAH [0:3] (0x002C; RO) This register contains the upper bits of the 48-bit Flow Control Ethernet address. Only the lower 16 bits of this register have meaning. The complete Flow Control address is {FCAH, FCAL}. The complete flow control multicast address is: 0x01_80_C2_00_00_01; where 0x01 is the first byte on the wire, 0x80 is the second, etc. Table 28-27. Flow Control Address High--FCAH [0:3] (0x002C; RO) Description: View: PCI Size: 32 bit Default: 0x00000100 Bit Range Bit Acronym 31 :16 Reserved 15 :00 FCAH 28.4.1.3 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Offset Start: 002Ch Offset End: 002Fh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0100h RO Reserved Write 0 ignore on read. Flow Control Address High Should be programmed with 0x01_00. Flow Control Type--FCT [0:3] (0x0030; R/W) This register contains the type field that hardware matches to recognize a flow control packet. Only the lower 16 bits of this register have meaning. This register should be programmed with 0x88_08. The upper byte is first on the wire FCT[15:8]. Table 28-28. Flow Control Type--FCT [0:3] (0x0030; R/W) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x00008808 Bit Range Bit Acronym 31 :16 Reserved 15 :00 FCT October 2012 Order Number: 327879-001US Offset Start: 0030h Offset End: 0033h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x8808 R/W Reserved Write 0 ignore on read. Flow Control Type Intel(R) Communications Chipset 89xx Series - Datasheet 1285 28.4.1.4 Flow Control Transmit Timer Value--FCTTV [0:3] (0x0170; R/W) The 16-bit value in the TTV field is inserted into a transmitted frame (either XOFF frames or any PAUSE frame value in any software transmitted packets). It counts in units of slot time of 64 bytes. If software needs to send an XON frame, it must set TTV to 0b prior to initiating the PAUSE frame. Table 28-29. Flow Control Transmit Timer Value--FCTTV [0:3] (0x0170; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :16 Reserved 15 :00 TTV Offset Start: 0170h Offset End: 0173h Power Well: GBEAUX Bit Reset Value Bit Access Reserved Write 0 ignore on read. 0h R/W Transmit Timer Value X R/W Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 1286 Bus:Device:Function: Sticky October 2012 Order Number: 327879-001US 28.0 28.4.1.5 Flow Control Receive Threshold Low--FCRTL0 [0:3] (0x2160; R/W) This register contains the receive threshold used to determine when to send an XON packet The complete register reflects the threshold in units of bytes. The lower 4 bits must be programmed to 0b (16 byte granularity). Software must set XONE to enable the transmission of XON frames. Each time hardware crosses the receive-high threshold (becoming more full), and then crosses the receive-low threshold and XONE is enabled (1b), hardware transmits an XON frame. When XONE is set, the RTL field should be programmed to at least 1b (at least 16 bytes). Flow control reception/transmission are negotiated capabilities by the Auto-Negotiation process. When the Controller is manually configured, flow control operation is determined by the CTRL.RFCE and CTRL.TFCE bits. Table 28-30. Flow Control Receive Threshold Low--FCRTL0 [0:3] (0x2160; R/W) Description: View: PCI Size: 32 bit Bit Range 31 30 :17 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Acronym XONE Reserved 16 :04 RTL 03 :00 Reserved October 2012 Order Number: 327879-001US Offset Start: 2160h Offset End: 2163h Power Well: GBEAUX Bit Description Sticky XON Enable 0b = Disabled. 1b = Enabled. Bit Reset Value Bit Access 0h R/W 0x0 R/W Reserved Write 0 ignore on read. Receive Threshold Low. FIFO low water mark for flow control transmission. An XON packet is sent if the occupied space in the packet buffer is smaller or equal than this watermark. This field is in 16 bytes granularity. Reserved Write 0 ignore on read. Intel(R) Communications Chipset 89xx Series - Datasheet 1287 28.4.1.6 Flow Control Receive Threshold High--FCRTH0 [0:3] (0x2168; R/W) This register contains the receive threshold used to determine when to send an XOFF packet. The complete register reflects the threshold in units of bytes. This value must be at maximum 48 bytes less than the maximum number of bytes allocated to the Receive Packet Buffer (IRPBS.RXPbsize), and the lower 4 bits must be programmed to 0b (16 byte granularity). The value of RTH should also be bigger than FCRTL.RTL. Each time the receive FIFO reaches the fullness indicated by RTH, hardware transmits a PAUSE frame if the transmission of flow control frames is enabled. Flow control reception/transmission are negotiated capabilities by the Auto-Negotiation process. When the Controller is manually configured, flow control operation is determined by the CTRL.RFCE and CTRL.TFCE bits. Table 28-31. Flow Control Receive Threshold High--FCRTH0 [0:3] (0x2168; R/W) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :18 Reserved 17 :04 RTH 03 :00 Reserved Offset Start: 2168h Offset End: 216Bh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 R/W Reserved Write 0 ignore on read. Receive Threshold High FIFO high water mark for flow control transmission. An XOFF packet is sent if the occupied space in the packet buffer is bigger or equal than this watermark. This field is in 16 bytes granularity. Note: When in DMA coalescing operation threshold high value defined in FCRTC.RTH_Coal is used instead of RTH value to allow increase of Receive Threshold High value by maximum supported Jumbo frame size. Increase is possible since during DMA coalescing operation transmit buffer is always empty. Reserved Write 0 ignore on read. Intel(R) Communications Chipset 89xx Series - Datasheet 1288 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] October 2012 Order Number: 327879-001US 28.0 28.4.1.7 Flow Control Refresh Threshold Value--FCRTV[0:3] (0x2460; R/W) Table 28-32. Flow Control Refresh Threshold Value--FCRTV[0:3](0x2460; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :16 Reserved 15 :00 Bus:Device:Function: FC_refresh_th October 2012 Order Number: 327879-001US Offset Start: 2460h Offset End: 2463h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 R/W Reserved Flow Control Refresh Threshold This value indicates the threshold value of the flow control shadow counter; when the counter reaches this value, and the conditions for PAUSE state are still valid (buffer fullness above low threshold value), a PAUSE (XOFF) frame is sent to link partner. If this field contains zero value, the Flow Control Refresh is disabled. Intel(R) Communications Chipset 89xx Series - Datasheet 1289 28.4.1.8 Flow Control Status--FCSTS0 [0:3] (0x2464; RO) This register describes the status of the flow control machine. Table 28-33. Flow Control Status--FCSTS0 [0:3] (0x2464; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Offset Start: 2464h Offset End: 2467h Power Well: GBEAUX Bit Reset Value Bit Access 0x0 RO The size of data in the memory is below the low threshold X RO The size of data in the memory is above the high threshold X RO 0h RO Bit Range Bit Acronym 31 :16 Refresh counter 15 :03 Reserved Reserved 02 Below low 01 Above high 00 Bus:Device:Function: Bit Description Sticky Flow control refresh counter Flow control state machine signal Flow_control state 0b = XON 1b = XOFF 28.5 GbE PCIe Registers 28.5.1 Detailed Register Description 28.5.1.1 Function Active and Power State to MNG--FACTPS[0:3] (0x5B30; RO) Firmware uses this register for configuration Table 28-34. Function Active and Power State to MNG--FACTPS[0:3] (0x5B30; RO) (Sheet 1 of 3) Description: View: PCI Size: 32 bit Bit Range 31 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Acronym Reserved 29 MNGCG 17 Reserved Func4 Aux_En Bit Description Sticky Bit Reset Value Bit Access 0h RO 0h RO 0h RO Reserved as 0 MNG Clock Gated When set, indicates that the manageability clock is gated. Reserved Function 4 Auxiliary (AUX) Power PM Enable bit shadow from the configuration space. Intel(R) Communications Chipset 89xx Series - Datasheet 1290 Offset Start: 5B30h Offset End: 5B33h Power Well: CGBAhUX Indication that one or more of the functions power states had changed. This bit is also a signal to the PM State Changed MNG unit to create an interrupt. This bit is cleared on read, and is not set for at least 8 cycles after it was cleared. 30 28 :18 B:0:1+ Bus:Device:Function: Index1 October 2012 Order Number: 327879-001US 28.0 Table 28-34. Function Active and Power State to MNG--FACTPS[0:3] (0x5B30; RO) (Sheet 2 of 3) Description: View: PCI Size: 32 bit Bit Range 16 15 :14 13 12 11 :10 09 08 07 :06 05 :04 03 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Power Well: CGBAhUX Bit Reset Value Bit Access 0h RO 0h RO Function 3 Auxiliary (AUX) Power PM Enable bit shadow from the configuration space. 0h RO LAN 2 Enable When set to 0b, it indicates that the LAN 2 function is disabled. When the function is enabled, the bit is set to 1b. The LAN 2 enable bit is set by the LAN2_DIS_N strapping pin. The PCH does not support the strap pin. 0h RO 0h RO Function 2 Auxiliary (AUX) Power PM Enable bit shadow from the configuration space. 0h RO LAN 1 Enable When set to 0b, it indicates that the LAN 1 function is disabled. When the function is enabled, the bit is set to 1b. The LAN 1 enable bit is set by the LAN1_DIS_N strapping pin. The PCH does not support the strap pin. 0h RO 0h RO 0h RW Bit Acronym Bit Description LAN3 Valid LAN 3 Enable When set to 0b, it indicates that the LAN 3 function is disabled. When the function is enabled, the bit is set to 1b. The LAN 3 enable bit is set by the LAN3_DIS_N strapping pin. The PCH does not support the strap pin. Sticky Power state indication of Function 4 DR 00b D0u Func4 Power State 01b 10b D0a 11b D3 Func3 Aux_En LAN2 Valid Power state indication of Function 3 DR 00b D0u Func3 Power State 01b 10b D0a 11b D3 Func2 Aux_En LAN1 Valid Power state indication of Function 2 DR 00b D0u Func2 Power State 01b 10b D0a 11b D3 Reserved Func1_Aux_En October 2012 Order Number: 327879-001US Offset Start: 5B30h Offset End: 5B33h Reserved. Function 1Auxiliary (AUX) Power PM Enable bit shadow from the configuration space. Intel(R) Communications Chipset 89xx Series - Datasheet 1291 Table 28-34. Function Active and Power State to MNG--FACTPS[0:3] (0x5B30; RO) (Sheet 3 of 3) Description: View: PCI Size: 32 bit Bit Range 02 01 :00 28.5.1.2 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Offset Start: 5B30h Offset End: 5B33h B:0:1+ Index1 Default: 0x0 Power Well: CGBAhUX Bit Acronym Bit Description LAN0 Valid LAN 0 Enable When set to 0b, it indicates that the LAN 0 function is disabled. When the function is enabled, the bit is set to 1b. The LAN 0 enable bit is set by the LAN0_DIS_N strapping pin. The PCH does not support the strap pin. Sticky Power state indication of Function 1 DR 00b D0u Func1 Power State 01b 10b D0a 11b D3 Bit Reset Value Bit Access 0h RO 0h RO Mirrored Revision ID--MREVID[0:3] (0x5B64; R/W) Table 28-35. Mirrored Revision ID--MREVID[0:3] (0x5B64; R/W) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Default: 0x0 31 :16 Reserved 15 :08 Step REV ID 07 :00 EEPROM RevID Offset Start: 5B64h Offset End: 5B67h Power Well: GBEAUX Bit Reset Value Bit Access Revision ID . 0x0 R/W Mirroring of Revision ID loaded from the EEPROM in PCIe configuration space (from word Device Rev ID word, address 0x1E). 0x0 R/W Bit Acronym Bit Description Sticky Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1292 B:0:1+ Bus:Device:Function: Index1 October 2012 Order Number: 327879-001US 28.0 28.5.1.3 PCIE Control Extended Register--GCR_EXT[0:3] (0x5B6C; R/W) Table 28-36. PCIE Control Extended Register--GCR_EXT[0:3] (0x5B6C; R/W) Description: View: PCI Size: 32 bit Bit Range 31 30 29 :05 04 03 :00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 5B6Ch Offset End: 5B6Fh Power Well: GBEAUX Bit Reset Value Bit Access Reserved 0x0 RW Reserved Reserved 0x0 RO Reserved Reserved 0x0 RW 0x0 RW Bit Acronym Bit Description Reserved APBACD Reserved October 2012 Order Number: 327879-001US Sticky Auto PBA Clear Disable. When set to 1, Software can clear the PBA only by direct write to clear access to the PBA bit. When set to 0, any active PBA entry is cleared on the falling edge of the appropriate interrupt request to the PCIe block. The appropriate interrupt request is cleared when software sets the associated interrupt mask bit in the EIMS (reenabling the interrupt) or by direct write to clear to the PBA. Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1293 28.6 Semaphore Registers This section contains registers common to all ports used to coordinate between all functions. The usage of these registers is described in Section 23.5. 28.6.1 Detailed Register Descriptions 28.6.1.1 Software Semaphore--SWSM[0:3] (0x5B50; R/W) Table 28-37. Software Semaphore--SWSM[0:3] (0x5B50; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Default: 0x0 SWESMBI Software/Firmware Semaphore bit This bit should be set only by the device driver (read only to firmware). The bit is not set if bit 0 in the FWSM register is set. The device driver should set this bit and than read it to verify that it was set. If it was set, it means that the device driver can access the SW_FW_SYNC register. The device driver should clear this bit after modifying the SW_FW_SYNC register. Notes: * If Software takes ownership of the SWSM.SWESMBI bit for a duration longer than 100 mS, Firmware may take ownership of the bit. Hardware clears this bit on PCIe reset. 0x0 R/W SMBI (RS) Software/Software Semaphore Bit This bit is set by hardware when this register is read by the device driver and cleared when the HOST driver writes a 0b to it. The first time this register is read, the value is 0b. In the next read the value is 1b (hardware mechanism). The value remains 1b until the software device driver clears it. This bit can be used as a semaphore between all the device's drivers in the Controller. This bit is cleared on PCIe reset. 0x0 R/W 31 :02 Reserved Reserved Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 1294 Power Well: GBEAUX Bit Access Bit Acronym 00 Offset Start: 5B50h Offset End: 5B53h Bit Reset Value Bit Range 01 B:0:1+ Bus:Device:Function: Index1 Sticky October 2012 Order Number: 327879-001US 28.0 28.6.1.2 Firmware Semaphore--FWSM[0:3] (0x5B54; R/WS) Table 28-38. Firmware Semaphore--FWSM[0:3] (0x5B54; R/WS) (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 5B54h Offset End: 5B57h Power Well: GBEAUX Bit Reset Value Bit Access 30 SerDes3 configuration error indication SERDES3_Config_ Set to 1b by firmware when it fails to configure LAN3 SerDes. Err_Ind Cleared by firmware upon successful configuration of LAN3 SerDes. 0h R/WS 29 SerDes2 configuration error indication SERDES2_Config_ Set to 1b by firmware when it fails to configure LAN2 SerDes. Err_Ind Cleared by firmware upon successful configuration of LAN2 SerDes. 0h R/WS 27 PHY/SerDes1 configuration error indication PHY_SERDES1_Co Set to 1b by firmware when it fails to configure LAN1 nfig_ PHY/SerDes. Err_Ind Cleared by firmware upon successful configuration of LAN1 PHY/SerDes. 0h R/WS 26 PHY/SerDes0 configuration error indication PHY_SERDES0_Co Set to 1b by firmware when it fails to configure LAN0 nfig_ PHY/SerDes. Err_Ind Cleared by firmware upon successful configuration of LAN0 PHY/SerDes. 0h R/WS 0x0 R/WS 0h R/WS Bit Range 31 28 25 Bit Acronym1 Reserved Reserved Reserved 24 :19 Ext_Err_Ind 18 :16 Reset_Cnt October 2012 Order Number: 327879-001US Bit Description Sticky Reserved. Reserved. Reserved External error indication Firmware writes here the reason that the firmware operation has stopped. For example, EEPROM CRC error, etc. Possible values: 0x00: No Error 0x01 to 0x04: Reserved. 0x05: EEPROM CRC error in SB section. 0x06: EEPROM CRC error in PT-LAN/CSRs sections. 0x07: EEPROM CRC error in FW Code section. 0x08: CRC error in PHY section. 0x09: Reserved. 0x0A: No Manageability (No EEPROM) 0x0F: Management memory Parity error. 0x10 to 0x03F: Reserved Note: When management error is detected, ICR.MGMT is set and an interrupts is sent to the Host. Ext_Err_ind values of 0x00 or 0x0A do not cause interrupt generation. Reset Counter Firmware increments the count on every Firmware reset. After 7 Firmware reset events counter stays stuck at 7 and does not wrap around. Intel(R) Communications Chipset 89xx Series - Datasheet 1295 Table 28-38. Firmware Semaphore--FWSM[0:3] (0x5B54; R/WS) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 15 14 :07 06 05 :04 03 :01 00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Power Well: GBEAUX Bit Acronym1 Bit Description FW_Val_Bit Firmware Valid Bit Hardware clears this bit in reset de-assertion so software can know firmware mode (bits 1-3) bits are invalid. Firmware should set this bit to 1b when it is ready (end of boot sequence). Reserved EEP_Reload_Ind Offset Start: 5B54h Offset End: 5B57h Sticky Bit Reset Value Bit Access 0h R/WS 0h R/WS 0x0 R/WS 0h R/WS Reserved EEPROM reloaded indication Set to 1b after firmware reloads the EEPROM. Cleared by firmware once the "Clear Bit" host command is received from host software. Reserved Reserved FW_Mode Firmware Mode Indicates the firmware mode as follows: 000b = No MNG. 001b = Reserved. 010b = PT mode. 011b = Reserved. 100b = Host Interface enable only. Software/Firmware Semaphore Firmware should set this bit to 1b before accessing the SW_FW_SYNC register. If the software is using the SWSM register and does not lock the EEP_FW_Semaphor SW_FW_SYNC, firmware is able to set this bit to 1b. e Firmware should set this bit back to 0b after modifying the SW_FW_SYNC register. Note: If Software takes ownership of the SWSM.SWESMBI bit for a duration longer than 100 mS, Firmware may take ownership of the bit. Notes: 1. This register should be written only by the manageability firmware. The device driver should only read this register. 2. Firmware ignores the EEPROM semaphore in operating system hung states. 3. Bits 15:00 are cleared on firmware reset. Intel(R) Communications Chipset 89xx Series - Datasheet 1296 October 2012 Order Number: 327879-001US 28.0 28.6.1.3 Software-Firmware Synchronization--SW_FW_SYNC[0:3] (0x5B5C; RWS) Check with LAD for Phy Related Bits This register is intended to synchronize between software and firmware. This register is common to all ports. Note: If software takes ownership of bits in the SW_FW_SYNC register for a duration longer than 1 second, then firmware may take ownership of the bit. Table 28-39. Firmware Synchronization--SW_FW_SYNC[0:3] (0x5B5C; RWS) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bit Acronym 31 :23 Reserved Reserved 22 :20 Reserved Reserved 18 :17 16 15 :09 B:0:1+ Index1 Default: 0x0 Bit Range 19 Bus:Device:Function: FW_MAC_CSR_SM Reserved FW_EEP_SM Reserved Offset Start: 5B5Ch Offset End: 5B5Fh Power Well: GBEAUX Bit Description Sticky When set to 1b, firmware owns access to shared CSRs Bit Reset Value Bit Access 0h RWS 0h RWS 0h RWS Reserved When set to 1b, EEPROM access is owned by firmware Reserved When Set to 1b, SWMBWR mailbox write register, is owned by software driver. 08 SW_MB_SM 07 Reserved 06 SW_PHY_SM3 When set to 1b, SerDes/PHY 3 access is owned by software 0h RWS 05 SW_PHY_SM2 When set to 1b, SerDes/PHY 2 access is owned by software 0h RWS 04 Reserved 0h RWS Reserved. Write 0, ignore on read. Reserved When set to 1b, software owns access to shared CSRs 03 SW_MAC_CSR_SM 02 SW_PHY_SM1 When set to 1b, SerDes/PHY 1 access is owned by software 0h RWS 01 SW_PHY_SM0 When set to 1b, SerDes/PHY 0 access is owned by software 0h RWS 00 SW_EEP_SM When set to 1b, EEPROM access is owned by software 0h RWS Reset conditions: * The software-controlled bits 15:0 are reset as any other CSR on global resets, D3hot exit and Forced TCO. Software is expected to clear the bits on entry to D3 state. * The firmware controlled bits (bits 31:16) are reset on LAN_PWR_GOOD (power-up) and firmware reset. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1297 28.6.1.4 Software Mailbox Write--SWMBWR[0:3] (0x5B04; R/W) Table 28-40. Software Mailbox Write--SWMBWR[0:3] (0x5B04; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 Mailbox 28.6.1.5 Bus:Device:Function: Offset Start: 5B04h Offset End: 5B07h Power Well: GBEAUX Bit Description Sticky Message sent from driver to the other drivers. The interpretation of this field is defined by the drivers. This register is reset only by power on reset. Bit Reset Value Bit Access 0x0 R/W Software Mailbox 0--SWMB0[0:3] (0x5B08; RO) Table 28-41. Software Mailbox 0--SWMB0[0:3] (0x5B08; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 Mailbox 28.6.1.6 Bus:Device:Function: B:0:1+ Index1 Offset Start: 5B08h Offset End: 5B0Bh Power Well: GBEAUX Bit Description Sticky Message sent from the driver of port 0. The interpretation of this field is defined by the drivers. This register is reset only by power on reset. Bit Reset Value Bit Access 0x0 RO Software Mailbox 1--SWMB1[0:3] (0x5B0C; RO) Table 28-42. Software Mailbox 1--SWMB1[0:3] (0x5B0C; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 Mailbox Offset Start: 5B0Ch Offset End: 5B0Fh Power Well: GBEAUX Bit Description Message sent from the driver of port 1. The interpretation of this field is defined by the drivers. This register is reset only by power on reset. Intel(R) Communications Chipset 89xx Series - Datasheet 1298 Bus:Device:Function: B:0:1+ Index1 Sticky Bit Reset Value Bit Access 0x0 RO October 2012 Order Number: 327879-001US 28.0 28.6.1.7 Software Mailbox 2--SWMB2[0:3] (0x5B18; RO) Table 28-43. Software Mailbox 2--SWMB2[0:3] (0x5B18; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 Mailbox 28.6.1.8 Bus:Device:Function: Offset Start: 5B18h Offset End: 5B1Bh Power Well: GBEAUX Bit Description Sticky Message sent from the driver of port 2. The interpretation of this field is defined by the drivers. This register is reset only by power on reset. Bit Reset Value Bit Access 0x0 RO Software Mailbox 3--SWMB3[0:3] (0x5B1C; RO) Table 28-44. Software Mailbox 3--SWMB3[0:3] (0x5B1C; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Bus:Device:Function: Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 Mailbox Offset Start: 5B1Ch Offset End: 5B1Fh Power Well: GBEAUX Bit Description Sticky Message sent from the driver of port 3.The interpretation of this field is defined by the drivers. This register is reset only by power on reset. Bit Reset Value Bit Access 0x0 RO 28.7 Interrupt Register Descriptions 28.7.1 Detailed Register Descriptions 28.7.1.1 Extended Interrupt Cause--EICR[0:3] (0x1580; RC/W1C) This register contains the frequent interrupt conditions for the Controller. Each time an interrupt causing event occurs, the corresponding interrupt bit is set in this register. An interrupt is generated each time one of the bits in this register is set and the corresponding interrupt is enabled via the Interrupt Mask Set/Read register. The interrupt might be delayed by the selected Interrupt Throttling register. The software device driver cannot determine from the RxTxQ bits what was the cause of the interrupt. The possible causes for asserting these bits are: * Receive Descriptor Write Back, Receive Descriptor Minimum Threshold hit, low latency interrupt for Rx, Transmit Descriptor Write Back. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1299 Writing a 1b to any bit in the register clears that bit. Writing a 0b to any bit has no effect on that bit. Register bits are cleared on register read. Auto clear can be enabled for any or all of the bits in this register. Note: Bits are not reset by Device Reset (CTRL.DEV_RST). 28.7.1.2 EICR Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) Table 28-45. EICR Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Acronym Bit Description Sticky Bit Reset Value TCP Timer TCP Timer Expired Activated when the TCP timer reaches its terminal count. 0h Reserved Reserved 30 RxTxQ Receive/Transmit Queue Interrupts One bit per queue or a bundle of queues, activated on receive/transmit queue events for the corresponding bit, such as: Receive Descriptor Write Back, Receive Descriptor Minimum Threshold hit Transmit Descriptor Write Back. The mapping of actual queue to the appropriate RxTxQ bit is according to the IVAR registers. Intel(R) Communications Chipset 89xx Series - Datasheet 1300 Power Well: GBEAUX 0h Other Cause 07 :00 Offset Start: 1580h Offset End: 1583h Interrupt Cause Active Activated when any bit in the ICR register is set. 31 29 :08 Bus:Device:Function: B:0:1+ Index1 Bit Access 0x0 October 2012 Order Number: 327879-001US 28.0 28.7.1.3 EICR Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) Table 28-46. EICR Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :10 Reserved 09 :00 MSIX 28.7.1.4 Bus:Device:Function: Offset Start: 1580h Offset End: 1583h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved Indicates an interrupt cause mapped to MSI-X vectors 9:0 Note: Bits are not reset by Device Reset (CTRL.DEV_RST). 0x0 Extended Interrupt Cause Set--EICS [0:3] (0x1520; WO) Software uses this register to set an interrupt condition. Any bit written with a 1b sets the corresponding bit in the Extended Interrupt Cause Read register. An interrupt is then generated if one of the bits in this register is set and the corresponding interrupt is enabled via the Extended Interrupt Mask Set/Read register. Bits written with 0b are unchanged. 28.7.1.5 EICS Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) Table 28-47. EICS Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) Description: View: PCI Size: 32 bit Bit Range 31 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Acronym Power Well: GBEAUX Bit Description Sticky Bit Reset Value Set when any bit in Other Causes is set 0h TCP Timer Sets the corresponding EICR TCP Timer interrupt condition. 0h 29 :08 Reserved Reserved 07 :00 RxTxQ 30 Note: Other Causes Offset Start: 1520h Offset End: 1523h Sets to corresponding EICR RxTXQ interrupt condition. Bit Access 0x0 To set bit 31 of the EICR (Other Causes), the ICS and IMS registers should be used in order to enable one of the legacy causes. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1301 28.7.1.6 EICS Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) Table 28-48. SEICS Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) Description: View: PCI BAR: GBEPCIBAR0[0:3] Size: 32 bit B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :10 Reserved 09 :00 MSIX 28.7.1.7 Bus:Device:Function: Offset Start: 1520h Offset End: 1523h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved Sets the corresponding EICR bit of MSI-X vectors 9:0 0x0 Extended Interrupt Mask Set/Read--EIMS [0:3] (0x1524; RWS) Reading of this register returns which bits have an interrupt mask set. An interrupt in EICR is enabled if its corresponding mask bit is set to 1b and disabled if its corresponding mask bit is set to 0b. A PCI interrupt is generated each time one of the bits in this register is set and the corresponding interrupt condition occurs (subject to throttling). The occurrence of an interrupt condition is reflected by having a bit set in the Extended Interrupt Cause Read register. An interrupt might be enabled by writing a 1b to the corresponding mask bit location (as defined in the EICR register) in this register. Any bits written with a 0b are unchanged. As a result, if software needs to disable an interrupt condition that had been previously enabled, it must write to the Extended Interrupt Mask Clear register rather than writing a 0b to a bit in this register. 28.7.1.8 EIMS Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) Table 28-49. EIMS Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) Description: View: PCI BAR: GBEPCIBAR0[0:3] Size: 32 bit Bit Range Default: 0x80000000 Bit Acronym Offset Start: 1524h Offset End: 1527h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Set Mask bit for the corresponding EICR other cause interrupt condition. 1h TCP Timer Set Mask bit for the corresponding EICR TCP timer interrupt condition. 0h 29 :08 Reserved Reserved 07 :00 RxTxQ Note: 31 Other Cause 30 Set Mask bit for the corresponding EICR RxTXQ interrupt. Bit Access 0x0 Bits are not reset by Device Reset (CTRL.DEV_RST). Intel(R) Communications Chipset 89xx Series - Datasheet 1302 B:0:1+ Bus:Device:Function: Index1 October 2012 Order Number: 327879-001US 28.0 28.7.1.9 EIMS Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) Table 28-50. EIMS Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: Xh Bit Range Bit Acronym 31 :10 Reserved 09 :00 MSIX 28.7.1.10 Bus:Device:Function: Offset Start: 1524h Offset End: 1527h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved Set Mask bit for the corresponding EICR bit of MSI-X vectors 9:0. Note: Bits are not reset by Device Reset (CTRL.DEV_RST). 0x0 Extended Interrupt Mask Clear - EIMC [0:3] (0x1528; WO) This register provides software a way to disable certain or all interrupts. Software disables a given interrupt by writing a 1b to the corresponding bit in this register. On interrupt handling, the software device driver should set all the bits in this register related to the current interrupt request even though the interrupt was triggered by part of the causes that were allocated to this vector. Interrupts are presented to the bus interface only when the mask bit is set to 1b and the cause bit is set to 1b. The status of the mask bit is reflected in the Extended Interrupt Mask Set/Read register and the status of the cause bit is reflected in the Interrupt Cause Read register. Software blocks interrupts by clearing the corresponding mask bit. This is accomplished by writing a 1b to the corresponding bit location (as defined in the EICR register) of that interrupt in this register. Bits written with 0b are unchanged (their mask status does not change). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1303 28.7.1.11 EIMC Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) Table 28-51. EIMC Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x80000000 Offset Start: 1528h Offset End: 152Bh Power Well: GBEAUX Bit Reset Value Bit Access Clear Mask bit for the corresponding EICR other cause interrupt. 1h WO TCP Timer Clear Mask bit for the corresponding EICR TCP timer interrupt. 0h WO 29 :08 Reserved Reserved 07 :00 RxTxQ 0x0 WO Bit Range Bit Acronym Bit Description 31 Other Cause 30 28.7.1.12 Sticky Clear Mask bit for the corresponding EICR RxTXQ interrupt. EIMC Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) Table 28-52. EIMC Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :10 Reserved 09 :00 MSIX Offset Start: 1528h Offset End: 152Bh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RW Reserved clear Mask bit for the corresponding EICR bit of MSI-X vectors 9:0 Intel(R) Communications Chipset 89xx Series - Datasheet 1304 Bus:Device:Function: B:0:1+ Index1 October 2012 Order Number: 327879-001US 28.0 28.7.1.13 Extended Interrupt Auto Clear--EIAC [0:3] (0x152C; R/W) This register is mapped like the EICS, EIMS, and EIMC registers, with each bit mapped to the corresponding MSI-X vector. This register is relevant to MSI-X mode only, where read-to-clear can not be used, as it might erase causes tied to other vectors. If any bits are set in EIAC, the EICR register should not be read. Bits without auto clear set, need to be cleared with write-to-clear. Note: EICR bits that have auto clear set are cleared by the internal emission of the corresponding MSI-X message even if this vector is disabled by the operating system. The MSI-X message can be delayed by EITR moderation from the time the EICR bit is activated. Table 28-53. Extended Interrupt Auto Clear--EIAC [0:3] (0x152C; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :10 Reserved 09 :00 Bus:Device:Function: B:0:1+ Index1 MSIX October 2012 Order Number: 327879-001US Offset Start: 152Ch Offset End: 152Fh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 R/W Reserved Auto clear bit for the corresponding EICR bit of MSI-X vectors 9:0. Notes: * Bits are not reset by Device Reset (CTRL.DEV_RST). * When GPIE.Multiple_MSIX = 0 (Non MSI-X mode) bits 8 and 9 are read only and should be ignored. Intel(R) Communications Chipset 89xx Series - Datasheet 1305 28.7.1.14 Extended Interrupt Auto Mask Enable--EIAM [0:3] (0x1530; R/W) Each bit in this register enables clearing of the corresponding bit in EIMS following read- or write-to-clear to EICR or setting of the corresponding bit in EIMS following a write-to-set to EICS. In MSI-X mode, this register controls which of the bits in EIMC to clear upon interrupt generation. 28.7.1.15 EIAM Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) Table 28-54. IAM Register Bit Description--Non MSI-X Mode (GPIE.Multiple_MSIX = 0) Description: View: PCI BAR: GBEPCIBAR0[0:3] Size: 32 bit Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 1530h Offset End: 1533h Power Well: GBEAUX Bit Reset Value Bit Access Auto mask bit for the corresponding EICR other cause interrupt condition. 0h R/W TCP Timer Auto mask bit for the corresponding EICR TCP timer interrupt condition. 0h R/W 29 :08 Reserved Reserved 07 :00 RxTxQ 0x0 RW Bit Range Bit Acronym Bit Description 31 Other Cause 30 Note: Sticky Auto Mask bit for the corresponding EICR RxTxQ interrupt. Bits are not reset by Device Reset (CTRL.DEV_RST). 28.7.1.16 EIAM Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) Table 28-55. EIAM Register Bit Description--MSI-X Mode (GPIE.Multiple_MSIX = 1) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :10 Reserved 09 :00 MSIX Offset Start: 1530h Offset End: 1533h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RW Reserved Auto Mask bit for the corresponding EICR bit of MSI-X vectors 9:0. Note: Bits are not reset by Device Reset (CTRL.DEV_RST). Intel(R) Communications Chipset 89xx Series - Datasheet 1306 B:0:1+ Bus:Device:Function: Index1 October 2012 Order Number: 327879-001US 28.0 28.7.1.17 Interrupt Cause Read Register--ICR [0:3] (0x1500; RC/W1C) This register contains the interrupt conditions for the Controller that are not present directly in the EICR. Each time an ICR interrupt causing event occurs, the corresponding interrupt bit is set in this register. The EICR.Other bit reflects the setting of interrupt causes from ICR as masked by the Interrupt Mask Set/Read register. Each time all un-masked causes in ICR are cleared, the EICR.Other bit is also cleared. ICR bits are cleared on register read. Clear-on-read can be enabled/disabled through a general configuration register bit. Auto clear is not available for the bits in this register. To prevent unwanted LSC (Link Status Change) interrupts during initialization, software should disable this interrupt until the end of initialization. Table 28-56. Interrupt Cause Read Register--ICR [0:3] (0x1500; RC/W1C) (Sheet 1 of 3) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Offset Start: 1500h Offset End: 1503h Power Well: GBEAUX Bit Reset Value Bit Access INTA Interrupt Asserted: Indicates that the INT line is asserted. Can be used by driver in shared interrupt scenario to decide if the received interrupt was emitted by the Controller. This bit is not valid in MSI/ MSI-X environments 0h RC/W1C 30 DRSTA Device Reset Asserted Indicates the CTRL.DEV_RST was asserted on another port or on this port. When device reset occurs all ports should re-initialize registers and descriptor rings. Note: Bit is not reset by Device Reset (CTRL.DEV_RST). 0h RC/W1C 29 TCP timer TCP timer interrupt 0h RC/W1C Detected Malicious driver behavior Occurs when one of the queues used malformed descriptors. In virtualized systems, might indicate a malicious or buggy driver. Note: This bit should never rise during normal operation. 0h RC/W1C Software Watchdog This bit is set after a software watchdog timer times out. 0h RC/W1C Storm Control Event This bit is set when multicast or broadcast storm control mechanism is activated or de-activated. 0h RC/W1C CGB Exception Occurred Set when any error as reported in CGB_ICAUSE register, happens. Only [24:31[and [4] are relevant for the PCH. 0h RC/W1C Reserved Write 0, ignore on read. 0h RC/W1C Fatal Error This bit is set when a fatal error is detected in one of the memories 0h RC/W1C Bit Range 31 Bit Acronym Bit Description 28 MDDET 27 Reserved 26 Software WD 25 SCE 24 CGB Exception 23 Reserved 22 FER October 2012 Order Number: 327879-001US Sticky Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1307 Table 28-56. Interrupt Cause Read Register--ICR [0:3] (0x1500; RC/W1C) (Sheet 2 of 3) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Offset Start: 1500h Offset End: 1503h Power Well: GBEAUX Bit Reset Value Bit Access Other Media Energy Detect When in SerDes/SGMII mode, indicates that link status has changed on the external media. 0h RC/W1C Bit Acronym Bit Description Sticky 21 Reserved 20 OMED 19 Time_Sync Time_Sync Interrupt This interrupt cause is set if Interrupt is generated by the Time Sync Interrupt registers (TSICR, TSIM and TSIS). 0h RC/W1C MNG Manageability Event Detected Indicates that a manageability event happened. When the Controller is at power down mode, the IPMI can generate a PME for the same events that would cause an interrupt when the Controller is in the D0 state. Note: Further information on manageability event can be found in FWSM register. 0h RC/W1C 18 17 :13 Reserved Reserved Reserved 12 GPI_SDP1 General Purpose Interrupt on SDP1 If GPI interrupt detection is enabled on this pin (via CTRL_EXT), this interrupt cause is set when the SDP1 is sampled high. 0h RC/W1C 11 GPI_SDP0 General Purpose Interrupt on SDP0 If GPI interrupt detection is enabled on this pin (via CTRL_EXT), this interrupt cause is set when the SDP0 is sampled high. 0h RC/W1C 10 Reserved 0h RC/W1C Reserved 09 MDAC MDIO Access Complete. Set when a MDIO access or a SFP I2C transaction completes. 08 SWMB Set when one of the drivers wrote a message using the SWMBWR mailbox register.. 0h RC/W1C 07 RXDW Receiver Descriptor Write Back Set when the Controller writes back an Rx descriptor to memory. 0h RC/W1C 06 Rx Miss Missed packet interrupt is activated for each received packet that overflows the Rx packet buffer (overrun). The packet is dropped and also increments the associated MPC counter. Note: Could be caused by no available receive buffers or because PCIe receive bandwidth is inadequate. 0h RC/W1C 05 Reserved Reserved 04 RXDMT0 Receive Descriptor Minimum Threshold Reached Indicates that the minimum number of receive descriptors are available and software should load more receive descriptors. 0h RC/W1C 03 Reserved Reserved Write 0, ignore on read. 02 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1308 Bus:Device:Function: October 2012 Order Number: 327879-001US 28.0 Table 28-56. Interrupt Cause Read Register--ICR [0:3] (0x1500; RC/W1C) (Sheet 3 of 3) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Acronym 01 Reserved 00 TXDW 28.7.1.18 Bus:Device:Function: Offset Start: 1500h Offset End: 1503h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0h RC/W1C Reserved Write 0, ignore on read. Transmit Descriptor Written Back Set when the Controller writes back a Tx descriptor to memory. Interrupt Cause Set Register--ICS [0:3] (0x1504; WO) Software uses this register to set an interrupt condition. Any bit written with a 1b sets the corresponding interrupt. This results in the corresponding bit being set in the Interrupt Cause Read Register (see Section 28.7.1.17). A PCIe interrupt is generated if one of the bits in this register is set and the corresponding interrupt is enabled through the Interrupt Mask Set/Read Register (see Section 28.7.1.19). Bits written with 0 are unchanged. Table 28-57. Interrupt Cause Set Register--ICS [0:3] (0x1504; WO) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 30 DRSTA 29 TCP timer 28 MDDET Power Well: GBEAUX Bit Reset Value Bit Access Sets the Device Reset Asserted Interrupt. When setting this bit a DRSTA interrupt is generated on this port only. 0h WO Sets the TCP timer interrupt. 0h WO Sets the Detected Malicious driver behavior Interrupt. 0h WO Bit Acronym Reserved Offset Start: 1504h Offset End: 1507h Bit Description Sticky Reserved. 27 Reserved 26 Software WD Sets the Software Watchdog Interrupt. 0h WO 25 SCE Set the Storm Control Event Interrupt 0h WO 24 Reserved Reserved 23 Reserved Reserved 22 FER 0h WO 21 Reserved 20 OMED Sets the Other Media Energy Detected Interrupt. 0h WO 19 Time_Sync Sets the Time_Sync interrupt. 0h WO October 2012 Order Number: 327879-001US Reserved. Sets the Fatal Error Interrupt. Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1309 Table 28-57. Interrupt Cause Set Register--ICS [0:3] (0x1504; WO) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 18 17 :14 BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Acronym MNG Offset Start: 1504h Offset End: 1507h Power Well: GBEAUX Bit Description Sets the Management Event Interrupt. Sticky Bit Reset Value Bit Access 0h WO Reserved Reserved 13 Reserved Reserved 12 GPI_SDP1 Sets the General Purpose Interrupt, related to SDP1 pin. 0h WO 11 GPI_SDP0 Sets the General Purpose Interrupt, related to SDP0 pin. 0h WO Reserved Reserved 10 :09 08 SWMB Sets the SWMB mailbox interrupt. 0h WO 07 RXDW Sets the Receiver Descriptor Write Back Interrupt. 0h WO Sets the Rx Miss Interrupt. 0h WO 06 Rx Miss 05 Reserved Reserved 04 RXDMT0 Sets the Receive Descriptor Minimum Threshold Hit Interrupt. 0h WO 03 Reserved Reserved Write 0, ignore on read. 0h WO 02 Reserved Reserved 01 Reserved Reserved Write 0, ignore on read. 0h WO 00 TXDW Sets the Transmit Descriptor Written Back Interrupt. 0h WO Intel(R) Communications Chipset 89xx Series - Datasheet 1310 Bus:Device:Function: October 2012 Order Number: 327879-001US 28.0 28.7.1.19 Interrupt Mask Set/Read Register--IMS [0:3] (0x1508; R/W) Reading this register returns bits that have an interrupt mask set. An interrupt is enabled if its corresponding mask bit is set to 1b and disabled if its corresponding mask bit is set to 0b. A PCIe interrupt is generated each time one of the bits in this register is set and the corresponding interrupt condition occurs. The occurrence of an interrupt condition is reflected by having a bit set in the Interrupt Cause Read Register (see Section 28.7.1.17). A particular interrupt can be enabled by writing a 1b to the corresponding mask bit in this register. Any bits written with a 0b are unchanged. As a result, if software desires to disable a particular interrupt condition that had been previously enabled, it must write to the Interrupt Mask Clear Register (see Section 28.7.1.20) rather than writing a 0b to a bit in this register. Table 28-58. Interrupt Mask Set/Read Register--IMS [0:3] (0x1508; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Offset Start: 1508h Offset End: 150Bh Power Well: GBEAUX Bit Reset Value Bit Access Sets/Reads the mask for Device Reset Asserted Interrupt. Note: Bit is not reset by Device Reset (CTRL.DEV_RST). 0h R/W Sets/Reads the mask for TCP timer interrupt. 0h R/W Sets/Reads the mask for the Detected Malicious driver behavior Interrupt. 0h R/W Bit Acronym Bit Description Sticky 31 Reserved 30 DRSTA 29 TCP timer 28 MDDET 27 Reserved 26 Software WD Sets/Reads the mask for the Software Watchdog Interrupt. 0h R/W 25 SCE Sets/Reads the mask for the Storm Control Event Interrupt. 0h R/W 24 Reserved Reserved 23 Reserved Reserved Write 0, ignore on read. 22 FER 0h R/W 21 Reserved Sets/Reads the mask for Other Media Energy Detected Interrupt. 0h R/W Sets/Reads the mask for Time_Sync Interrupt. 0h R/W Sets/Reads the mask for Management Event Interrupt. 0h R/W 20 OMED 19 Time_Sync 18 MNG 17 :13 Reserved. Reserved Sets/Reads the mask for the Fatal Error Interrupt. Reserved Reserved Reserved 12 GPI_SDP1 Sets/Reads the mask for General Purpose Interrupt, related to SDP1 pin. 0h R/W 11 GPI_SDP0 Sets/Reads the mask for General Purpose Interrupt, related to SDP0 pin. 0h R/W 10 :09 Reserved Reserved 08 SWMB Sets/Reads the mask for Mailbox interrupt. 0h R/W 07 RXDW Sets/Reads the mask for Receiver Descriptor Write Back Interrupt. 0h R/W October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1311 Table 28-58. Interrupt Mask Set/Read Register--IMS [0:3] (0x1508; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Bit Acronym Offset Start: 1508h Offset End: 150Bh Power Well: GBEAUX Bit Description 06 Rx Miss Sets/Reads the mask for the Rx Miss Interrupt. 05 Reserved Reserved 04 RXDMT0 Sets/Reads the mask for Receive Descriptor Minimum Threshold Hit Interrupt. 03 Reserved Reserved Write 0, ignore on read. 02 Reserved Reserved 01 Reserved Reserved Write 0, ignore on read. 00 TXDW Sets/Reads the mask for Transmit Descriptor Written Back Interrupt. Intel(R) Communications Chipset 89xx Series - Datasheet 1312 B:0:1+ Index1 Sticky Bit Reset Value Bit Access 0h R/W 0h R/W 0h R/W October 2012 Order Number: 327879-001US 28.0 28.7.1.20 Interrupt Mask Clear Register--IMC [0:3] (0x150C; WO) Software uses this register to disable an interrupt. Interrupts are presented to the bus interface only when the mask bit is set to 1b and the cause bit set to 1b. The status of the mask bit is reflected in the Interrupt Mask Set/Read Register (see Section 28.7.1.19), and the status of the cause bit is reflected in the Interrupt Cause Read Register (see Section 28.7.1.17). Reading this register returns the value of the IMS register. Software blocks interrupts by clearing the corresponding mask bit. This is accomplished by writing a 1b to the corresponding bit in this register. Bits written with 0b are unchanged (their mask status does not change). In interrupt handling, the software device driver should set all the bits in this register related to the current interrupt request, even though the interrupt was triggered by part of the causes that were allocated to this vector. Table 28-59. Interrupt Mask Clear Register--IMC [0:3] (0x150C; WO) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range Default: 0x0 Reserved 30 DRSTA 29 TCP timer MDDET 27 Reserved 26 Software WD Power Well: GBEAUX Bit Access Clears the mask for Device Reset Asserted Interrupt. 0h WO Clears the mask for TCP timer interrupt. 0h WO Clears the mask for the Detected Malicious driver behavior Interrupt. 0h WO Reserved. 0h WO Clears the mask for Software Watchdog Interrupt. 0h WO Clears the mask for the Storm Control Event Interrupt. 0h WO 0h WO 0h WO Bit Description SCE 24 Reserved Reserved 23 Reserved Reserved Write 0, ignore on read. 22 FER 21 Reserved OMED 19 Time_Sync 18 MNG 17 :13 Sticky Reserved 25 20 Offset Start: 150Ch Offset End: 150Fh Bit Reset Value Bit Acronym 31 28 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Clears the mask for the Fatal Error Interrupt. Reserved Clears the mask for the Other Media Energy Detected Interrupt. Clears the mask for the Time_Sync Interrupt. 0h WO Clears the mask for the Management Event Interrupt. 0h WO Reserved Reserved 12 GPI_SDP1 Clears the mask for the General Purpose Interrupt, related to SDP1 pin. 0h WO 11 GPI_SDP0 Clears the mask for the General Purpose Interrupt, related to SDP0 pin. 0h WO 10 :09 Reserved Reserved 08 SWMB Clears the mask for the Software Mailbox Interrupt. 0h WO 07 RXDW Clears the mask for the Receiver Descriptor Write Back Interrupt. 0h WO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1313 Table 28-59. Interrupt Mask Clear Register--IMC [0:3] (0x150C; WO) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Bit Acronym Offset Start: 150Ch Offset End: 150Fh Power Well: GBEAUX Bit Description Clears the mask for the Rx Miss Interrupt. Sticky Bit Reset Value Bit Access 0h WO 0h WO 06 Rx Miss 05 Reserved Reserved 04 RXDMT0 Clears the mask for Receive Descriptor Minimum Threshold Hit Interrupt. 03 Reserved Reserved Write 0, ignore on read. 02 Reserved Reserved 01 Reserved Reserved Write 0, ignore on read. 0h WO 00 TXDW Clears the mask for Transmit Descriptor Written Back Interrupt. 0h WO Intel(R) Communications Chipset 89xx Series - Datasheet 1314 B:0:1+ Index1 October 2012 Order Number: 327879-001US 28.0 28.7.1.21 Interrupt Acknowledge Auto Mask Register--IAM [0:3] (0x1510; R/ W) Table 28-60. Interrupt Acknowledge Auto Mask Register--IAM [0:3] (0x1510; R/W) Description: View: PCI BAR: GBEPCIBAR0[0:3] Size: 32 bit Bit Range 31 :0 28.7.1.22 Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 1510h Offset End: 1513h Power Well: GBEAUX Bit Acronym Bit Description Sticky IAM_VALUE An ICR read or write will have the side effect of writing the contents of this register to the IMC register. If GPIE.NSICR = 0, then the copy of this register to IMS will occur only if at least one bit is set in the IMS and there is a true interrupt as reflected in ICR.INTA. Note: Bit 30 of this register is not reset by Device Reset (CTRL.DEV_RST). Bit Reset Value Bit Access 0h R/W Interrupt Throttle--EITR [0:3] (0x1680 + 4*n [n = 0...9]; R/W) Each EITR is responsible for an interrupt cause (RxTxQ, TCP timer and Other Cause). The allocation of EITR-to-interrupt cause is through the IVAR registers. Software uses this register to pace (or even out) the delivery of interrupts to the host processor. This register provides a guaranteed inter-interrupt delay between interrupts asserted by the Controller, regardless of network traffic conditions. To independently validate configuration settings, software can use the following algorithm to convert the inter-interrupt interval value to the common interrupts/sec. performance metric: interrupts/sec = (8 * 10-6sec x interval)-1 A counter counts in units of 8*10-6 sec. After counting "interval "number of units, an interrupt is sent to the software. The above equation gives the number of interrupts per second. The equation below time in seconds between consecutive interrupts. For example, if the interval is programmed to 125 (decimal), the Controller guarantees the processor does not receive an interrupt for 1 ms from the last interrupt. The maximum observable interrupt rate from the Controller should never exceed 7813 interrupts/sec. Inversely, inter-interrupt interval value can be calculated as: inter-interrupt interval = (8 * 10-6sec x interrupt/sec)-1 The optimal performance setting for this register is very system and configuration specific. An initial suggested range is 2 to 175 (2 - AF). Note: Setting EITR to a non zero value can cause an interrupt cause Rx/Tx statistics miscount. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1315 Table 28-61. Interrupt Throttle--EITR [0:3][0:9] (0x1680 + 4*n [n = 0...9]; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Offset Start: 1680h at 4 Offset End: 1683h at 4 Power Well: GBEAUX Bit Reset Value Bit Access 0h R/W Down counter, exposes only the 10 most significant bits of the real 12-bit counter. Loaded with Interval value whenever the associated interrupt is signaled. Counts down to 0 and stops. The associated interrupt is signaled Moderation whenever this counter is zero and an associated (via the Counter (RWS) Interrupt Select register) EICR bit is set. If the CNT_INGR is not set this counter can be directly written by software at any time to alter the throttles performance. 0x0 R/W Reflects the current credits for that EITR for LL interrupts. If the CNT_INGR is not set this counter can be directly written by software at any time to alter the throttles performance 0x0 R/W LLI_EN LLI moderation enable. 0h R/W 14 :02 Interval Minimum inter-interrupt interval. The interval is specified in 1 s increments. A zero disables interrupt throttling logic. 0x0 R/W 01 :00 Reserved Bit Range 31 30 :21 20 :16 15 Note: Bit Acronym CNT_INGR (WO) LL Counter (RWS) Bit Description When set the hardware does not override the counters fields (ITR counter and LLI credit counter), so they keep their previous value. Relevant for the current write only and is always read as zero Sticky Reserved EITR register and interrupt mechanism is not reset by Device Reset (CTRL.DEV_RST). Occurrence of Device Reset interrupt causes immediate generation of all pending interrupts. Intel(R) Communications Chipset 89xx Series - Datasheet 1316 B:0:1+ Index1 October 2012 Order Number: 327879-001US 28.0 28.7.1.23 Interrupt Vector Allocation Registers--IVAR [0:3] (0x1700 + 4*n [n=0...3]; RW) These registers have two modes of operation: 1. In MSI-X mode these registers define the allocation of the different interrupt causes as defined in Section 28-62 to one of the MSI-X vectors. Each INT_Alloc[i] (i=0...15) field is a byte indexing an entry in the MSI-X Table Structure and MSI-X PBA Structure. 2. In non MSI-X mode these registers define the allocation of the Rx and Tx queues interrupt causes to one of the RxTxQ bits in the EICR. Each INT_Alloc[i] (i=0...15) field is a byte indexing the appropriate RxTxQ bit as defined in Section 28-63. Entries are mapped as follows: a. Queues RX0, TX0, RX1, TX1 are mapped in IVAR[0] Register. b. Queues RX2, TX2, RX3, TX3 are mapped in IVAR[1] Register. c. Queues RX4, TX4, RX5, TX5 are mapped in IVAR[2] Register. d. Queues RX6, TX6, RX7, TX7 in are mapped IVAR[3] Register. Table 28-62. Interrupt Vector Allocation Registers--IVAR [0:3][0:3](0x1700 + 4*n [n=0...3]; RW) Description: View: PCI Size: 32 bit Bit Range 31 Default: 0x0 Bit Acronym Reserved 27 :24 INT_Alloc[3] 22 :20 19 :16 15 INT_Alloc[2] Reserved 11 :08 INT_Alloc[1] 06 :04 03 :00 Sticky Defines the MSI-X vector assigned to the interrupt cause associated with this entry, as defined in Section 28-63. Valid values are 0 to 9 for MSI-X mode and 0 to 7 in non MSI-X mode. INT_Alloc[0] October 2012 Order Number: 327879-001US Bit Access 0h RW 0x0 RW 0h RW 0x0 RW 0h RW 0x0 RW 0h RW 0x0 RW Reserved Defines the MSI-X vector assigned to the interrupt cause associated with this entry, as defined in Section 28-63. Valid values are 0 to 9 for MSI-X mode and 0 to 7 in non MSI-X mode. Reserved Defines the MSI-X vector assigned to the interrupt cause associated with this entry, as defined in Section 28-63. Valid values are 0 to 9 for MSI-X mode and 0 to 7 in non MSI-X mode. INT_Alloc_val[0 Valid bit for INT_Alloc[0] ] Reserved Bit Reset Value Reserved INT_Alloc_val[1 Valid bit for INT_Alloc[1] ] 14 :12 07 Bit Description INT_Alloc_val[2 Valid bit for INT_Alloc[2] ] Reserved Offset Start: 1700h at 4 Offset End: 1703h at 4 Power Well: GBEAUX INT_Alloc_val[3 Valid bit for INT_Alloc[3] ] 30 :28 23 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Reserved Defines the MSI-X vector assigned to the interrupt cause associated with this entry, as defined in Section 28-63. Valid values are 0 to 9 for MSI-X mode and 0 to 7 in non MSI-X mode. Intel(R) Communications Chipset 89xx Series - Datasheet 1317 Note: If invalid values are written to the INT_Alloc fields the result is unexpected. DW 0 31 24 INT_ALLOC[3] 23 16 INT_ALLOC[2] 1 2 15 8 7 0 INT_ALLOC[1] INT_ALLOC[0] ... ... INT_ALLOC[13] INT_ALLOC[12] ... 3 INT_ALLOC[15] 28.7.1.24 INT_ALLOC[14] Interrupt Vector Allocation Registers--MISC IVAR_MISC [0:3] (0x1740; RW) This register is used only in MSI-X mode. This register defines the allocation of the Other Cause and TCP Timer interrupts to one of the MSI-X vectors. Table 28-63. Interrupt Vector Allocation Registers--MISC IVAR_MISC[0:3] (0x1740; RW) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :16 Reserved 15 14 :12 11 :08 07 06 :04 03 :00 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Power Well: GBEAUX Bit Description INT_Alloc_val[1 Valid bit for INT_Alloc[16] 6] Bit Access 0h RW 0x0 RW 0h RW 0x0 RW Reserved the MSI-X vector assigned to the TCP timer INT_Alloc[16] Defines interrupt cause. Valid values are 0 to 9. Intel(R) Communications Chipset 89xx Series - Datasheet 1318 Bit Reset Value Reserved INT_Alloc[17] Defines the MSI-X vector assigned to the "Other Cause" interrupt. Valid values are 0 to 9. Reserved Sticky Reserved INT_Alloc_val[1 Valid bit for INT_Alloc[17] 7] Reserved Offset Start: 1740h Offset End: 1743h October 2012 Order Number: 327879-001US 28.0 28.7.1.25 General Purpose Interrupt Enable--GPIE [0:3] (0x1514; RW) Table 28-64. General Purpose Interrupt Enable--GPIE [0:3] (0x1514; RW) Description: View: PCI BAR: GBEPCIBAR0[0:3] Size: 32 bit Bit Range 30 EIAME 29 :12 Reserved 11 :07 LL Interval 06 :05 Reserved Bit Reset Value Bit Access PBA Support: When set, setting one of the extended interrupts masks via EIMS causes the PBA bit of the associated MSI-X vector to be cleared. Otherwise, the Controller behaves in a way that supports legacy INT-x interrupts. Note: Should be cleared when working in INT-x or MSI mode and set in MSI-X mode. 0h RW Extended Interrupt Auto Mask enable: When set (usually in MSI-X mode); upon firing of an MSI-X message, bits set in EIAM associated with this message are cleared. Otherwise, EIAM is used only upon read or write of EICR/ EICS registers. 0h RW 0x0 RW 0h RW 0h RW Bit Description Sticky Reserved Low latency credits increment rate. The interval is specified in 4 s increments. A value of 0x0 disables moderation of LLI for all interrupt vectors. Reserved 0 = on-MSI mode, or MSI-X with single vector, IVAR maps Rx/Tx causes, to 16 EICR bits, but MSIX[0] is asserted for all. Multiple MSIX 1 = MSIX mode, IVAR maps Rx/Tx causes, TCP Timer and "Other Cause" interrupts to 10 MSI-x vectors reflected in 10 EICR bits. 03 :01 Reserved 00 Reserved. Non Selective Interrupt clear on read: When set, every read of ICR clears it. When this bit is cleared, an ICR read causes it to be cleared only if an actual interrupt was asserted or IMS = 0b. NSICR Offset Start: 1514h Offset End: 1517h Power Well: GBEAUX Bit Acronym PBA_support 28.8 B:0:1+ Index1 Default: 0x0 31 04 Bus:Device:Function: MSI-X Table Register Descriptions These registers are used to configure the MSI-X mechanism. The message address and message upper address registers sets the address for each of the vectors. The message register sets the data sent to the relevant address. The vector control registers are used to enable specific vectors. The pending bit array register indicates which vectors have pending interrupts. The structure is listed in Table 28-65. Table 28-65. MSI-X Table Structure DWORD3 MSIXTVCTRL DWORD2 MSIXTMSG DWORD1 MSIXTUADD DWORD0 MSIXTADD Entry Number BAR 3 - Offset Vector Control Msg Data Msg Upper Addr Msg Addr Entry 0 Base (0x0000) Vector Control Msg Data Msg Upper Addr Msg Addr Entry 1 Base + 1*16 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1319 Table 28-65. MSI-X Table Structure Vector Control Msg Data ... Vector Control Note: Msg Upper Addr Msg Addr Entry 2 Base + 2*16 ... ... ... ... Msg Data Msg Upper Addr Msg Addr Entry (N-1) Base + (N-1) *16 N = 25. Table 28-66. MSI-X PBA Structure MSIXPBA[63:0] QWORD Number BAR 3 - Offset Pending Bits 0 through 63 QWORD0 Base (0x2000) Pending Bits 64 through 127 QWORD1 Base+1*8 ... ... ... Pending Bits ((N-1) div 64)*64 through N-1 QWORD((N-1) div 64) BASE + ((N-1) div 64)*8 Note: N = 25. As a result, only QWORD0 is implemented. 28.8.1 Detailed Register Descriptions 28.8.1.1 MSIX Table Entry Lower Address--MSIXTADD [0:3][0:9] (BAR3: 0x0000 + 0x10*n [n=0...9]; R/W) Table 28-67. MSIX Table Entry Lower Address--MSIXTADD [0:3][0:9] (BAR3: 0x0000 + 0x10*n [n=0...9]; R/W) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR3[0:3] Default: 0x0 Bit Range Bit Acronym 31 :02 Message Address 01 :00 Message Address LSB (RO) Power Well: GBEAUX Bit Reset Value Bit Access System-Specific Message Lower Address For MSI-X messages, the contents of this field from an MSI-X table entry specifies the lower portion of the DWORD-aligned address for the memory write transaction. 0x0 R/W For proper DWORD alignment, software must always write 0b's to these two bits. Otherwise, the result is undefined. 0x0 R/W Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 1320 0000h at Offset Start: 0x10 Offset End: 0003h at 0x10 Sticky October 2012 Order Number: 327879-001US 28.0 28.8.1.2 MSIX Table Entry Upper Address--MSIXTUADD [0:3][0:9] (BAR3: 0x0004 + 0x10*n [n=0...9]; R/W) Table 28-68. MSIX Table Entry Upper Address--MSIXTUADD [0:3][0:9] (BAR3: 0x0004 + 0x10*n [n=0...9]; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR3[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 Message Address 28.8.1.3 Bus:Device:Function: 0004h at Offset Start: 0x10 Offset End: 0007h at 0x10 Power Well: GBEAUX Bit Description Sticky System-Specific Message Upper Address Bit Reset Value Bit Access 0x0 R/W MSIX Table Entry Message--MSIXTMSG [0:3][0:9] (BAR3: 0x0008 + 0x10*n [n=0...9]; R/W) Table 28-69. MSIX Table Entry Message--MSIXTMSG [0:3][0:9] (BAR3: 0x0008 + 0x10*n [n=0...9]; R/W) Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: GBEPCIBAR3[0:3] B:0:1+ Bus:Device:Function: Index1 Default: 0x0 Bit Acronym Power Well: GBEAUX Bit Description Sticky System-Specific Message Data For MSI-X messages, the contents of this field from an MSI-X table entry specifies the data written during the Message Data memory write transaction. In contrast to message data used for MSI messages, the low-order message data bits in MSI-X messages are not modified by the function. October 2012 Order Number: 327879-001US 0008h at Offset Start: 0x10 Offset End: 000Bh at 0x10 Bit Reset Value Bit Access 0x0 R/W Intel(R) Communications Chipset 89xx Series - Datasheet 1321 28.8.1.4 MSIX Table Entry Vector Control--MSIXTVCTRL [0:3][0:9] (BAR3: 0x000C + 0x10*n [n=0...9]; R/W) Table 28-70. MSIX Table Entry Vector Control--MSIXTVCTRL [0:3][0:9] (BAR3: 0x000C + 0x10*n [n=0...9]; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR3[0:3] Bit Acronym 31 :01 Reserved 28.8.1.5 B:0:1+ Index1 Default: 0x00000001 Bit Range 00 Bus:Device:Function: Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 1h R/W Reserved When this bit is set, the function is prohibited from sending a message using this MSI-X table entry. However, any other MSI-X table entries programmed with the same vector are still capable of sending an equivalent message unless they are also masked. Mask 000Ch at Offset Start: 0x10 Offset End: 000Fh at 0x10 MSIXPBA Bit Description--MSIXPBA [0:3] (BAR3: 0x2000; RO) Table 28-71. MSIXPBA Bit Description--MSIXPBA [0:3] (BAR3: 0x2000; RO) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR3[0:3] Default: 0x0 Bit Range Bit Acronym 31 :10 Reserved 09 :00 Pending Bits Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RO Reserved For each pending bit that is set, the function has a pending message for the associated MSI-X Table entry. Pending bits that have no associated MSI-X table entry are reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 1322 Offset Start: 2000h Offset End: 2003h October 2012 Order Number: 327879-001US 28.0 28.8.1.6 MSIX PBA Clear--PBACL [0:3] (0x5B68; R/W1C) Table 28-72. MSIX PBA Clear--PBACL [0:3] (0x5B68; R/W1C) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0X0 Bit Range Bit Acronym 31 :10 Reserved 09 :00 Bus:Device:Function: PENBITCLR October 2012 Order Number: 327879-001US Offset Start: 5B68h Offset End: 5B6Bh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 R/W1C Reserved MSI-X Pending bits Clear Writing a 1b to any bit clears the corresponding MSIXPBA bit; writing a 0b has no effect. Note: Bits are set for a single PCIe clock cycle and than cleared. Intel(R) Communications Chipset 89xx Series - Datasheet 1323 28.9 Receive Registers 28.9.1 Detailed Register Descriptions 28.9.1.1 Receive Control Register--RCTL [0:3] (0x0100; R/W) This register controls all Controller receiver functions. Table 28-73. Receive Control Register--RCTL[0:3] (0x0100; R/W) (Sheet 1 of 3) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Offset Start: 0100h Offset End: 0103h Power Well: GBEAUX Bit Reset Value Bit Access 0h R/W PMCF Pass MAC Control Frames Filters out unrecognized pause and other control frames. 0b = Pass/forward pause frames. 1b = Filter pause frames (default). PMCF controls the DMA function of MAC control frames (other than flow control). A MAC control frame in this context must be addressed to either the MAC control frame multicast address or the station address, match the type field, and NOT match the PAUSE opcode of 0x0001. If PMCF = 1b then frames meeting this criteria are transferred to host memory. 0h R/W 22 DPF Discard Pause Frames with Station MAC Address Controls whether pause frames directly addressed to this station are forwarded to the host. 0b = incoming pause frames with station MAC address are forwarded to the host. 1b = incoming pause frames with station MAC address are discarded. Note: Pause frames with other MAC addresses (multicast address) are always discarded unless the specific address is added to the accepted MAC addresses (either multicast or unicast). 0h R/W 21 PSP Pad Small Receive packets. If this field is set, RCTL.SECRC should be set also. 0h R/W 20 CFI Canonical Form Indicator bit value 0b = 802.1Q packets with CFI equal to this field are accepted. 1b = 802.1Q packet is discarded. 0h R/W Bit Range Bit Acronym 31 :27 Reserved Reserved Should be written with 0b to ensure future compatibility. SECRC Strip Ethernet CRC from incoming packet Causes the CRC to be stripped from all packets. 0b = Does not strip CRC 1b = Strips CRC. This bit controls whether the hardware strips the Ethernet CRC from the received packet. This stripping occurs prior to any checksum calculations. The stripped CRC is not transferred to host memory and is not included in the length reported in the descriptor. Reserved Reserved Should be written with 0b to ensure future compatibility. 26 25 :24 23 Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 1324 B:0:1+ Index1 Sticky October 2012 Order Number: 327879-001US 28.0 Table 28-73. Receive Control Register--RCTL[0:3] (0x0100; R/W) (Sheet 2 of 3) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 0100h Offset End: 0103h Power Well: GBEAUX Bit Reset Value Bit Access CFIEN Canonical Form Indicator Enable 0b = Disabled (CFI bit found in received 802.1Q packet's tag is not compared to decide packet acceptance). 1b = Enabled (CFI bit found in received 802.1Q packet's tag must match RCTL.CFI to accept 802.1Q type packet. 0h R/W VFE VLAN Filter Enable 0b = Disabled (filter table does not decide packet acceptance). 1b = Enabled (filter table decides packet acceptance for 802.1Q packets). Three bits [20:18] control the VLAN filter table. The first determines whether the table participates in the packet acceptance criteria. The next two are used to decide whether the CFI bit found in the 802.1Q packet should be used as part of the acceptance criteria. 0h R/W BSIZE Receive Buffer Size BSIZE controls the size of the receive buffers and permits software to trade-off descriptor performance versus required storage space. Buffers that are 2048 bytes require only one descriptor per receive packet maximizing descriptor efficiency. 00b = 2048 Bytes. 01b = 1024 Bytes. 10b = 512 Bytes. 11b = 256 Bytes. Notes: 1. BSIZE not modified when RXEN is set to 1b. Set RXEN =0 when modifying the buffer size by changing this field. 2. BSIZE value only defines receive buffer size of queues with a SRRCTL.BSIZEPACKET value of 0. 0h R/W 15 BAM Broadcast Accept Mode. 0b = Ignore broadcast (unless it matches through exact or imperfect filters). 1b = Accept broadcast packets. 0h R/W 14 Reserved 0h R/W Bit Range 19 18 17 :16 Bit Acronym Bit Description Reserved Multicast Offset Determines which bits of the incoming multicast address are used in looking up the bit vector. 00b = bits [47:36] of received destination multicast address. 01b = bits [46:35] of received destination multicast address. 10b = bits [45:34] of received destination multicast address. 11b = bits [43:32] of received destination multicast address. 13 :12 MO 11 :10 Reserved Reserved Set to 0b for compatibility. 09 :08 Reserved Reserved October 2012 Order Number: 327879-001US Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 1325 Table 28-73. Receive Control Register--RCTL[0:3] (0x0100; R/W) (Sheet 3 of 3) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Offset Start: 0100h Offset End: 0103h Power Well: GBEAUX Bit Reset Value Bit Access LBM Loopback mode. Controls the loopback mode of the Controller. 00b = Normal operation 01b = MAC loopback (test mode). 10b = Undefined. 11b = Loopback via internal SerDes (SerDes/SGMII/KX mode only). Note: PHY devices require programming for loopback operation using MDIO accesses. 0h R/W 05 LPE Long Packet Reception Enable 0b = Disabled. 1b = Enabled. LPE controls whether long packet reception is permitted. Hardware discards long packets if LPE is 0b. A long packet is one longer than 1522 bytes. If LPE is 1b, the maximum packet size that the Controller can receive is 9.5Kbytes. 0h R/W 04 MPE Multicast Promiscuous Enabled 0b = Disabled. 1b = Enabled. 0h R/W 03 UPE Unicast Promiscuous Enabled 0b = Disabled. 1b = Enabled. 0h R/W SBP Store Bad Packets 0b = do not store. 1b = store bad packets. This bit controls the MAC receive behavior. A packet is required to pass the address (or normal) filtering before the SBP bit becomes effective. If SBP = 0b, then all packets with layer 1 or 2 errors are rejected. The appropriate statistic would be incremented. If SBP = 1b, then these packets are received (and transferred to host memory). The receive descriptor error field (RDESC.ERRORS) should have the corresponding bit(s) set to signal the software device driver that the packet is erred. In some operating systems the software device driver passes this information to the protocol stack. In either case, if a packet only has layer 3+ errors, such as IP or TCP checksum errors, and passes other filters, the packet is always received (layer 3+ errors are not used as a packet filter). Note: symbol errors before the SFD are ignored. Any packet must have a valid SFD (RX_DV with no RX_ER in 10/100/1000BASE-T mode) in order to be recognized by the Controller (even bad packets). Also, erred packets are not routed to the MNG even if this bit is set. 0h R/W 01 RXEN Receiver Enable The receiver is enabled when this bit is set to 1b. Writing this bit to 0b stops reception after receipt of any in progress packet. All subsequent packets are then immediately dropped until this bit is set to 1b. 0h R/W 00 Reserved Bit Range 07 :06 02 Bit Acronym Bit Description Sticky Reserved Write to 0b for future compatibility. Intel(R) Communications Chipset 89xx Series - Datasheet 1326 B:0:1+ Index1 October 2012 Order Number: 327879-001US 28.0 28.9.1.2 Split and Replication Receive Control--SRRCTL [0:3][0:7](0xC00C + 0x40*n [n=0...7]; R/W) Table 28-74. Split and Replication Receive Control--SRRCTL [0:3][0:7] (0xC00C + 0x40*n [n=0...7]; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 30 29 :28 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x00000400 Bit Reset Value Bit Access Drop Enabled If set, packets received to the queue when no descriptors are available to store them are dropped. The packet is dropped only if there are not enough free descriptors in the host descriptor ring to store the packet. If there are enough descriptors in the host, but they are not yet fetched by the Controller, then the packet is not dropped and there are no release of packets until the descriptors are fetched. Default is 0b for queue 0 and 1b for the other queues. 0h R/W Timestamp Received packet 0 - Do not place timestamp at beginning of receive buffer. 1- Place timestamp at beginning of receive buffer. Timestamp is placed only in buffers of received packets that meet the criteria defined in the TSYNCRXCTL.Type field, 2-tuple filters or ETQF registers. When set a 40 bit time stamp generated from the value in SYSTIMH and SYSTIML registers is placed in the receive buffer before the MAC header of the packets defined in the TSYNCRXCTL.Type field. 0h R/W 0h R/W 0x0 R/W 4h R/W Bit Acronym Bit Description Drop_En Timestamp Reserved Sticky Reserved. Write 0 ignore on read. 27 :25 DESCTYPE 24 :20 RDMTS Receive Descriptor Minimum Threshold Size A low latency interrupt (LLI) associated with this queue is asserted whenever the number of free descriptors becomes equal to RDMTS multiplied by 16. 19 :14 Reserved 13 :12 Reserved Reserved. Write 0 ignore on read. Reserved Must be set to 00b. Receive Buffer Size for Header Buffer The value is in 64 bytes resolution. Valid value can be from 64 bytes to 960 bytes. Default buffer size is 256 BSIZEHEADER bytes. This field must be greater than 0 if the value of DESCTYPE is greater or equal to 2. Note: When SRRCTL.Timestamp is set to 1 and the value of DESCTYPE is greater or equal to 2, BSIZEHEADER size should be equal or greater than 2 (128 bytes). October 2012 Order Number: 327879-001US C00Ch at Offset Start: 0x4 Offset End: C00Fh at 0x4 Power Well: GBEAUX Defines the descriptor in Rx 000b = Legacy. 001b = Advanced descriptor one buffer. 010b = Advanced descriptor header splitting. 011b = Advanced descriptor header replication - replicate always. 100b = Advanced descriptor header replication large packet only (larger than header buffer size). 111b = Reserved. 11 :08 B:0:1+ Index1 Intel(R) Communications Chipset 89xx Series - Datasheet 1327 Table 28-74. Split and Replication Receive Control--SRRCTL [0:3][0:7] (0xC00C + 0x40*n [n=0...7]; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 07 06 :00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x00000400 Bit Acronym Reserved C00Ch at Offset Start: 0x4 Offset End: C00Fh at 0x4 Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0h R/W Reserved Receive Buffer Size for Packet Buffer The value is in 1 KB resolution. Valid values can be from 1 BSIZEPACKET KB to 127 KB. Default buffer size is 0 KB. If this field is equal 0x0, then RCTL.BSIZE determines the packet buffer size. Table 28-75. Split and Replication Receive Control--SRRCTL [0:3][1:7](0xC00C + 0x40*n [n=0...7]; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 30 29 :28 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x80000400 Power Well: GBEAUX Bit Reset Value Bit Access Drop_En Drop Enabled If set, packets received to the queue when no descriptors are available to store them are dropped. The packet is dropped only if there are not enough free descriptors in the host descriptor ring to store the packet. If there are enough descriptors in the host, but they are not yet fetched by the Controller, then the packet is not dropped and there are no release of packets until the descriptors are fetched. Default is 0b for queue 0 and 1b for the other queues. 1h R/W Timestamp Timestamp Received packet 0 - Do not place timestamp at beginning of receive buffer. 1- Place timestamp at beginning of receive buffer. Timestamp is placed only in buffers of received packets that meet the criteria defined in the TSYNCRXCTL.Type field, 2-tuple filters or ETQF registers. When set a 40 bit time stamp generated from the value in SYSTIMH and SYSTIML registers is placed in the receive buffer before the MAC header of the packets defined in the TSYNCRXCTL.Type field. 0h R/W Bit Acronym Bit Description Reserved Sticky Reserved. Write 0 ignore on read. Intel(R) Communications Chipset 89xx Series - Datasheet 1328 C04Ch at Offset Start: 0x40 Offset End: C04Fh at 0x40 October 2012 Order Number: 327879-001US 28.0 Table 28-75. Split and Replication Receive Control--SRRCTL [0:3][1:7](0xC00C + 0x40*n [n=0...7]; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 27 :25 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x80000400 Bit Acronym DESCTYPE C04Ch at Offset Start: 0x40 Offset End: C04Fh at 0x40 Power Well: GBEAUX Bit Description Sticky Defines the descriptor in Rx 000b = Legacy. 001b = Advanced descriptor one buffer. 010b = Advanced descriptor header splitting. 011b = Advanced descriptor header replication - replicate always. 100b = Advanced descriptor header replication large packet only (larger than header buffer size). Bit Reset Value Bit Access 0h R/W 0x0 R/W 4h R/W 0h R/W 101b = Reserved. 111b = Reserved. 24 :20 RDMTS 19 :14 Reserved 13 :12 Reserved 11 :08 07 06 :00 Receive Descriptor Minimum Threshold Size A low latency interrupt (LLI) associated with this queue is asserted whenever the number of free descriptors becomes equal to RDMTS multiplied by 16. Reserved. Write 0 ignore on read. Reserved Must be set to 00b. Receive Buffer Size for Header Buffer The value is in 64 bytes resolution. Valid value can be from 64 bytes to 1024 bytes. Default buffer size is 256 BSIZEHEADER bytes. This field must be greater than 0 if the value of DESCTYPE is greater or equal to 2. Note: When SRRCTL.Timestamp is set to 1 and the value of DESCTYPE is greater or equal to 2, BSIZEHEADER size should be equal or greater than 2 (128 bytes). Reserved Reserved Receive Buffer Size for Packet Buffer The value is in 1 KB resolution. Valid values can be from 1 BSIZEPACKET KB to 127 KB. Default buffer size is 0 KB. If this field is equal 0x0, then RCTL.BSIZE determines the packet buffer size. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1329 28.9.1.3 Packet Split Receive Type--PSRTYPE [0:3][0:7] (0x5480 + 4*n [n=0...7]; R/W) This register enables or disables each type of header that needs to be split. Each register controls the behavior of 2 queue. * Packet Split Receive Type Register (queue 0) - PSRTYPE0 (0x5480) * Packet Split Receive Type Register (queue 1) - PSRTYPE1 (0x5484) * Packet Split Receive Type Register (queue 2) - PSRTYPE2 (0x5488) * Packet Split Receive Type Register (queue 3) - PSRTYPE3 (0x548C) * Packet Split Receive Type Register (queue 4) - PSRTYPE3 (0x5490) * Packet Split Receive Type Register (queue 5) - PSRTYPE3 (0x5494) * Packet Split Receive Type Register (queue 6) - PSRTYPE3 (0x5498) * Packet Split Receive Type Register (queue 7) - PSRTYPE3 (0x549C) Table 28-76. Packet Split Receive Type--PSRTYPE [0:3][0:7] (0x5480 + 4*n [n=0...7]; R/ W) Description: View: PCI Size: 32 bit Bit Range 31 :19 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0007FFFE Power Well: GBEAUX Bit Reset Value Bit Access Header includes MAC, (VLAN/SNAP) IPv6, UDP, NFS only 1h R/W Header includes MAC, (VLAN/SNAP) IPv6, TCP, NFS only 1h R/W Bit Acronym Reserved Bit Description Sticky Reserved 18 PSR_type18 17 PSR_type17 16 Reserved 15 PSR_type15 Header includes MAC, (VLAN/SNAP) IPv4, IPv6, UDP, NFS only 1h R/W 14 PSR_type14 Header includes MAC, (VLAN/SNAP) IPv4, IPv6, TCP, NFS only 1h R/W 13 Reserved 12 PSR_type12 Header includes MAC, (VLAN/SNAP) IPv4, UDP, NFS only 1h R/W 11 PSR_type11 Header includes MAC, (VLAN/SNAP) IPv4, TCP, NFS only 1h R/W Reserved Reserved 10 Reserved 09 PSR_type9 Reserved Header includes MAC, (VLAN/SNAP) IPv6, UDP only 1h R/W 08 PSR_type8 Header includes MAC, (VLAN/SNAP) IPv6, TCP only 1h R/W 07 PSR_type7 Header includes MAC, (VLAN/SNAP) IPv6 only 1h R/W 06 PSR_type6 Header includes MAC, (VLAN/SNAP) IPv4, IPv6, UDP only 1h R/W 05 PSR_type5 Header includes MAC, (VLAN/SNAP) IPv4, IPv6, TCP only 1h R/W 04 PSR_type4 Header includes MAC, (VLAN/SNAP) IPv4, IPv6 only 1h R/W 03 PSR_type3 Header includes MAC, (VLAN/SNAP) IPv4, UDP only 1h R/W 02 PSR_type2 Header includes MAC, (VLAN/SNAP) IPv4, TCP only 1h R/W 01 PSR_type1 Header includes MAC, (VLAN/SNAP) IPv4 only 1h R/W 00 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1330 5480h at Offset Start: 0x4 Offset End: 5483h at 0x4 October 2012 Order Number: 327879-001US 28.0 28.9.1.4 Replicated Packet Split Receive Type--RPLPSRTYPE [0:3] (0x54C0; R/ W) This register enables or disables each type of header that needs to be split. This register controls the behavior of replicated packets. Table 28-77. Replicated Packet Split Receive Type--RPLPSRTYPE [0:3] (0x54C0; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0007FFFE Bit Range Bit Acronym 31 :19 Reserved Offset Start: 54C0h Offset End: 54C3h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved 18 PSR_type18 Header includes MAC, (VLAN/SNAP) IPv6, UDP, NFS only 1h R/W 17 PSR_type17 Header includes MAC, (VLAN/SNAP) IPv6, TCP, NFS only 1h R/W 16 Reserved 1h R/W 1h R/W Header includes MAC, (VLAN/SNAP) IPv4, UDP, NFS only 1h R/W Header includes MAC, (VLAN/SNAP) IPv4, TCP, NFS only 1h R/W Reserved 15 PSR_type15 Header includes MAC, (VLAN/SNAP) IPv4, IPv6, UDP, NFS only 14 PSR_type14 Header includes MAC, (VLAN/SNAP) IPv4, IPv6, TCP, NFS only 13 Reserved 12 PSR_type12 Reserved 11 PSR_type11 10 Reserved 09 PSR_type9 Header includes MAC, (VLAN/SNAP) IPv6, UDP only 1h R/W 08 PSR_type8 Header includes MAC, (VLAN/SNAP) IPv6, TCP only 1h R/W Reserved 07 PSR_type7 Header includes MAC, (VLAN/SNAP) IPv6 only 1h R/W 06 PSR_type6 Header includes MAC, (VLAN/SNAP) IPv4, IPv6, UDP only 1h R/W 05 PSR_type5 Header includes MAC, (VLAN/SNAP) IPv4, IPv6, TCP only 1h R/W 04 PSR_type4 Header includes MAC, (VLAN/SNAP) IPv4, IPv6 only 1h R/W 03 PSR_type3 Header includes MAC, (VLAN/SNAP) IPv4, UDP only 1h R/W 02 PSR_type2 Header includes MAC, (VLAN/SNAP) IPv4, TCP only 1h R/W 01 PSR_type1 Header includes MAC, (VLAN/SNAP) IPv4 only 1h R/W 00 Reserved October 2012 Order Number: 327879-001US Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1331 28.9.1.5 Receive Descriptor Base Address Low--RDBAL [0:3] [0:7] (0xC000 + 0x40*n [n=0...7]; R/W) This register contains the lower bits of the 64-bit descriptor base address. The lower four bits are always ignored. The Receive Descriptor Base Address must point to a 128 byte-aligned block of data. Note: In order to keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x2800, 0x2900, 0x2A00 & 0x2B00 respectively. Table 28-78. Receive Descriptor Base Address Low--RDBAL [0:3][0:7] (0xC000 + 0x40*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :07 RDBAL 06 :00 Lower_0 28.9.1.6 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] C000h at Offset Start: 0x40 Offset End: C000h at 0x40 Power Well: GBEAUX Bit Description Sticky Receive Descriptor Base Address Low Ignored on writes. Returns 0x0 on reads. Bit Reset Value Bit Access X R/W 0x0 R/W Receive Descriptor Base Address High--RDBAH [0:3] [0:7] (0xC004 + 0x40*n [n=0...7]; R/W) This register contains the upper 32 bits of the 64-bit descriptor base address. Table 28-79. Receive Descriptor Base Address High--RDBAH [0:3][0:7] (0xC004 + 0x40*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :00 RDBAH Note: Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Power Well: GBEAUX Bit Description Receive Descriptor Base Address [63:32] Sticky Bit Reset Value Bit Access X R/W In order to keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x2804, 0x2904, 0x2A04 & 0x2B04 respectively. Intel(R) Communications Chipset 89xx Series - Datasheet 1332 C004h at Offset Start: 0x40 Offset End: C007h at 0x40 October 2012 Order Number: 327879-001US 28.0 28.9.1.7 Receive Descriptor Ring Length--RDLEN [0:3][0:7] (0xC008 + 0x40*n [n=0...7]; R/W) This register sets the number of bytes allocated for descriptors in the circular descriptor buffer. It must be 128-byte aligned. Table 28-80. Receive Descriptor Ring Length--RDLEN [0:3][0:7](0xC008 + 0x40*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :20 Reserved 19 :07 LEN 06 :00 Bus:Device:Function: 0 C008h at Offset Start: 0x40 Offset End: C00Bh at 0x40 Power Well: GBEAUX Bit Reset Value Bit Access Descriptor Ring Length (number of 8 descriptor sets) 0x0 R/W Ignore on writes. Bits 6:0 must be set to zero. Bits 4:0 always read as zero. 0x0 R/W Bit Description Sticky Reserved Reads as 0b. Should be written to 0b for future compatibility. Note: To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x2808, 0x2908, 0x2A08 & 0x2B08 respectively. 28.9.1.8 Receive Descriptor Head--RDH [0:3][0:7] (0xC010 + 0x40*n [n=0...7]; RO) The value in this register might point to descriptors that are still not in host memory. As a result, the host cannot rely on this value in order to determine which descriptor to process. Table 28-81. Receive Descriptor Head--RDH [0:3][0:7] (0xC010 + 0x40*n [n=0...7]; RO) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym 31 :16 Reserved Reserved Should be written to 0b. 15 :00 RDH Receive Descriptor Head Note: C010h at Offset Start: 0x40 Offset End: C013h at 0x40 Bit Description Sticky Bit Reset Value Bit Access 0x0 RO To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x2810, 0x2910, 0x2A10 & 0x2B10 respectively. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1333 28.9.1.9 Receive Descriptor Tail--RDT [0:3][0:7] (0xC018 + 0x40*n [n=0...7]; R/W) This register contains the tail pointers for the receive descriptor buffer. The register points to a 16-byte datum. Software writes the tail register to add receive descriptors to the hardware free list for the ring. Note: Writing the RDT register while the corresponding queue is disabled is ignored by the PCH. Note: To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x2818, 0x2918, 0x2A18& 0x2B18 respectively. Table 28-82. Receive Descriptor Tail--RDT [0:3][0:7] (0xC018 + 0x40*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :16 Reserved 15 :00 RDT 28.9.1.10 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] C018h at Offset Start: 0x40 Offset End: C01Bh at 0x40 Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 R/W Reserved. Ignore on read, write 0 for future compatibility. Receive Descriptor Tail Receive Descriptor Control--RXDCTL [0:3][0:7] (0xC028 + 0x40*n [n=0...7]; R/W) This register controls the fetching and write-back of receive descriptors. The three threshold values are used to determine when descriptors are read from and written to host memory. The values are in units of descriptors (each descriptor is 16 bytes). Table 28-83. Receive Descriptor Control--RXDCTL [0:3][0:7] (0xC028 + 0x40*n [n=0...7]; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Default: 0x0000140C Bit Range Bit Acronym 31 :27 Reserved 26 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0h R/W Reserved Receive Software Flush Enables software to trigger receive descriptor write-back SWFLUSH (WC) flushing, independently of other conditions. This bit is cleared by hardware after write-back flush is triggered (may take a number of cycles). Intel(R) Communications Chipset 89xx Series - Datasheet 1334 C028h at Offset Start: 0x40 Offset End: C02Bh at 0x40 October 2012 Order Number: 327879-001US 28.0 Table 28-83. Receive Descriptor Control--RXDCTL [0:3][0:7] (0xC028 + 0x40*n [n=0...7]; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 25 24 :21 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0000140C Bit Acronym Bit Description Sticky ENABLE Receive Queue Enable When set, the Enable bit enables the operation of the specific receive queue. 1b =Enables queue. 0b =Disables queue. Setting this bit initializes Head and Tail registers (RDH[n] and RDT[n]) of the specific queue. Until then, the state of the queue is kept and can be used for debug purposes. When disabling a queue, this bit is cleared only after all activity in the queue has stopped. Note: If queue is enabled and RCTL.RXEN is cleared, receive activity on the queue will not commence. Reserved 20 :16 WTHRESH Write-Back Threshold WTHRESH controls the write-back of processed receive descriptors. This threshold refers to the number of receive descriptors in the on-chip buffer that are ready to be written back to host memory. In the absence of external events (explicit flushes), the write-back occurs only after at least WTHRESH descriptors are available for write-back. Possible values for this field are 0 to 15. Note: Since the default value for write-back threshold is 1b, the descriptors are normally written back as soon as one cache line is available. WTHRESH must contain a nonzero value to take advantage of the write-back bursting capabilities of the Controller. Note: It's recommended not to place a value above 0xC in the WTHRESH field. 15 :13 Reserved Reserved 12 :08 HTHRESH Host Threshold Field defines when receive descriptor prefetch is performed. Each time enough valid descriptors, as defined in the HTHRESH field, are available in host memory a prefetch is performed. Possible values for this field are 0 to 16. 07 :05 Reserved Reserved PTHRESH Prefetch Threshold PTHRESH is used to control when a prefetch of descriptors is considered. This threshold refers to the number of valid, unprocessed receive descriptors the Controller has in its on-chip buffer. If this number drops below PTHRESH, the algorithm considers pre-fetching descriptors from host memory. This fetch does not happen unless there are at least HTHRESH valid descriptors in host memory to fetch. Note: HTHRESH should be given a non zero value each time PTHRESH is used. Possible values for this field are 0 to 16. Note: C028h at Offset Start: 0x40 Offset End: C02Bh at 0x40 Power Well: GBEAUX Reserved 04 :00 B:0:1+ Index1 Bit Reset Value Bit Access 0h R/W 1h R/W Ah R/W Ch R/W To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x2828, 0x2928, 0x2A28 & 0x2B28 respectively. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1335 28.9.1.11 Receive Queue Drop Packet Count--RQDPC [0:3][0:7] (0xC030 + 0x40*n [n=0...7]; RC/W) Table 28-84. Receive Queue Drop Packet Count--RQDPC [0:3][0:7] (0xC030 + 0x40*n [n=0...7]; RC/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 C030h at Offset Start: 0x40 Offset End: C033h at 0x40 Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 RQDPC Receive Queue drop packet count - counts the number of packets dropped by a queue due to lack of descriptors available. Note: Counter does not wrap around when reaching a value of 0xFFFFFFFF. Sticky Bit Reset Value Bit Access 0x0 RC/W Note: To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x2830, 0x2930, 0x2A30 & 0x2B30 respectively. Note: Packets dropped due to the queue being disabled may not be counted by this register. Intel(R) Communications Chipset 89xx Series - Datasheet 1336 October 2012 Order Number: 327879-001US 28.0 28.9.1.12 Receive Checksum Control--RXCSUM [0:3] (0x5000; R/W) The Receive Checksum Control register controls the receive checksum off loading features of the Controller. The Controller supports the off loading of three receive checksum calculations: the Packet Checksum, the IP Header Checksum, and the TCP/ UDP Checksum. Note: This register should only be initialized (written) when the receiver is not enabled (only write this register when RCTL.RXEN = 0b). Table 28-85. Receive Checksum Control--RXCSUM [0:3] (0x5000; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x00000300 Bit Range Bit Acronym 31 :14 Reserved Offset Start: 5000h Offset End: 5003h Power Well: GBEAUX Bit Reset Value Bit Access Packet Checksum Disable The packet checksum and IP identification fields are mutually exclusive with the RSS hash. Only one of the two options is reported in the Rx descriptor. RXCSUM.PCSD Legacy Rx Descriptor (SRRCTL.DESCTYPE = 000b): 0b (checksum enable) - Packet checksum is reported in the Rx descriptor. 1b (checksum disable) - Not supported. RXCSUM.PCSD Extended or Header Split Rx Descriptor (SRRCTL.DESCTYPE not equal 000b): 0b (checksum enable) - checksum and IP identification are reported in the Rx descriptor. 1b (checksum disable) - RSS Hash value is reported in the Rx descriptor. 0h R/W Bit Description Sticky Reserved 13 PCSD 12 IPPCSE IP Payload Checksum Enable See PCSS description. 0h R/W 11 CRCOFL CRC32 Offload Enable Enables the SCTP CRC32 checksum off-loading feature. If RXCSUM.CRCOFL is set to 1b, the Controller calculates the CRC32 checksum and indicates a pass/fail indication to software via the CRC32 Checksum Valid bit (RDESC.L4I) in the Extended Status field of the receive descriptor. In non I/OAT, this bit is read only as 0b. 0h R/W 10 Reserved Reserved 09 TUOFLD TCP/UDP Checksum Off-load Enable 1h R/W October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1337 Table 28-85. Receive Checksum Control--RXCSUM [0:3] (0x5000; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 08 07 :00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x00000300 Offset Start: 5000h Offset End: 5003h Power Well: GBEAUX Bit Reset Value Bit Access IPOFLD IP Checksum Off-load Enable RXCSUM.IPOFLD is used to enable the IP Checksum offloading feature. If RXCSUM.IPOFLD is set to 1b, the Controller calculates the IP checksum and indicates a pass/fail indication to software via the IP Checksum Error bit (IPE) in the Error field of the receive descriptor. Similarly, if RXCSUM.TUOFLD is set to 1b, the Controller calculates the TCP or UDP checksum and indicates a pass/ fail indication to software via the TCP/UDP Checksum Error bit (L4E). Similarly, if RFCTL.IPv6_DIS and RFCTL.IP6Xsum_DIS are cleared to 0b and RXCSUM.TUOFLD is set to 1b, the Controller calculates the TCP or UDP checksum for IPv6 packets. It then indicates a pass/fail condition in the TCP/UDP Checksum Error bit (RDESC.L4E). This applies to checksum off loading only. Supported frame types: Ethernet II Ethernet SNAP 1h R/W PCSS Packet Checksum Start Controls the packet checksum calculation. The packet checksum shares the same location as the RSS field and is reported in the receive descriptor when the RXCSUM.PCSD bit is cleared. If the RXCSUM.IPPCSE is set, the Packet checksum is aimed to accelerate checksum calculation of fragmented UDP packets. See section Section 26.1.11.2 for a detailed explanation. If RXCSUM.IPPCSE is cleared (the default value), the checksum calculation that is reported in the Rx Packet checksum field is the unadjusted 16-bit ones complement of the packet. The packet checksum starts from the byte indicated by RXCSUM.PCSS (0b corresponds to the first byte of the packet), after VLAN stripping if enabled by the CTRL.VME. For example, for an Ethernet II frame encapsulated as an 802.3ac VLAN packet and with RXCSUM.PCSS set to 14, the packet checksum would include the entire encapsulated frame, excluding the 14-byte Ethernet header (DA, SA, Type/Length) and the 4-byte VLAN tag. The packet checksum does not include the Ethernet CRC if the RCTL.SECRC bit is set. Software must make the required offsetting computation (to back out the bytes that should not have been included and to include the pseudo-header) prior to comparing the packet checksum against the L4 checksum stored in the packet checksum. The partial checksum in the descriptor is aimed to accelerate checksum calculation of fragmented UDP packets. Note: The PCSS value should not exceed a pointer to the IP header start. If exceeded, the IP header checksum or TCP/UDP checksum is not calculated correctly. 0x0 R/W Bit Acronym Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 1338 B:0:1+ Index1 Sticky October 2012 Order Number: 327879-001US 28.0 28.9.1.13 Receive Long Packet Maximum Length--RLPML [0:3] (0x5004; R/W) Table 28-86. Receive Long Packet Maximum Length--RLPML [0:3] (0x5004; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x00002600 Bit Range Bit Acronym 31 :14 Reserved 13 :00 RLPML 28.9.1.14 Bus:Device:Function: Offset Start: 5004h Offset End: 5007h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 2600h R/W Reserved Maximum allowed long packet length. This length is the global length of the packet including all the potential headers of suffixes in the packet. Receive Filter Control Register--RFCTL [0:3] (0x5008; R/W) Table 28-87. Receive Filter Control Register--RFCTL [0:3] (0x5008; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x1 Offset Start: 5008h Offset End: 500Bh Power Well: GBEAUX Bit Reset Value Bit Access Defines the priority between SYNQF & 2 tuple filter 0b = 2-tuple filter priority 1b = SYN filter priority. 0h R/W Forward Length Error Packet 0b = packet with length error are dropped. 1b = packets with length error are forwarded to the host. 0h R/W Reserved Reserved Must be set to 00b. 0h R/W 15 Reserved Reserved 14 IPFRSP_DIS 0h R/W 13 Reserved Reserved 12 Reserved Reserved 11 IPv6XSUM_DIS IPv6 XSUM Disable Disables XSUM on IPv6 packets. 0h R/W 10 IPv6_DIS IPv6 Disable Disables IPv6 packet filtering. Any received IPv6 packet is parsed only as an L2 packet. 0h R/W Bit Range Bit Acronym 31 :20 Reserved Reserved Write 0 ignore on read. 19 SYNQFP 18 LEF 17 :16 October 2012 Order Number: 327879-001US Bit Description Sticky IP Fragment Split Disable When this bit is set, the header of IP fragmented packets are not set. Intel(R) Communications Chipset 89xx Series - Datasheet 1339 Table 28-87. Receive Filter Control Register--RFCTL [0:3] (0x5008; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x1 Offset Start: 5008h Offset End: 500Bh Power Well: GBEAUX Bit Reset Value Bit Access NFS_VER NFS Version 00b = NFS version 2. 01b = NFS version 3. 10b = NFS version 4. 11b = Reserved for future use. 0h R/W 07 NFSR_DIS NFS Read Disable Disables filtering of NFS read reply headers. 0h R/W 06 NFSW_DIS NFS Write Disable Disables filtering of NFS write request headers. 0h R/W Bit Range 09 :08 05 :00 28.9.1.15 Bit Acronym Reserved Bit Description Sticky Reserved Multicast Table Array--MTA [0:3][0:127] (0x5200 + 4*n [n=0...127]; R/W) There is one register per 32 bits of the Multicast Address Table for a total of 128 registers. Software must mask to the desired bit on reads and supply a 32-bit word on writes. The first bit of the address used to access the table is set according to the RX_CTRL.MO field. Note: All accesses to this table must be 32 bit. Table 28-88. Multicast Table Array--MTA [0:3][0:127] (0x5200 + 4*n [n=0...127]; R/W) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 Bit Vector 5200h at Offset Start: 0x4 Offset End: 5203h at 0x4 Power Well: GBEAUX Bit Description Word wide bit vector specifying 32 bits in the multicast address filter table. Sticky Bit Reset Value Bit Access X R/W Figure 28-1 shows the multicast lookup algorithm. The destination address shown represents the internally stored ordering of the received DA. Bit 0 indicated in this diagram is the first on the wire. Intel(R) Communications Chipset 89xx Series - Datasheet 1340 October 2012 Order Number: 327879-001US 28.0 Figure 28-1. Multicast Table Array Destination Address 47:40 39:32 31:24 23:16 15:8 7:0 RCTL.MO[1:0] Multicast Table Array 32 x 128 (4096 bit vector) word ? pointer[11:5] ... ... bit pointer[4:0] 28.9.1.16 Receive Address Low 0--RAL0 [0:3][0:15] (0x5400 + 8*n [n=0...15]; R/W) While "n" is the exact unicast/multicast address entry and it is equal to 0,1,...15. These registers contain the lower bits of the 48 bit Ethernet address. All 32 bits are valid. These registers are reset by a software reset or platform reset. The first register (RAL0) is loaded from the EEPROM after a software or platform reset. RAL0 and RAL1 together makeup the RAL - Receive Address Low registers. Table 28-89. Receive Address Low 0--RAL0 [0:3][0:15] (0x5400 + 8*n [n=0...15]; 0x54E0 + 8*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 RAL Note: Bus:Device:Function: B:0:1+ Index1 5400h at Offset Start: 0x8 Offset End: 5403h at 0x8 Power Well: GBEAUX Bit Description Sticky Receive address low Contains the lower 32-bit of the 48-bit Ethernet address. Bit Reset Value Bit Access X R/W The RAL field should be written in network order. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1341 28.9.1.17 Receive Address Low 1--RAL1 [0:3][0:7] (0x54E0 + 8*n [n=0...7]; R/ W) While "n" is the exact unicast/multicast address entry and it is equal to 0,1,...7. These registers contain the lower bits of the 48 bit Ethernet address. All 32 bits are valid. These registers are reset by a software reset or platform reset. The first register (RAL0) is loaded from the EEPROM after a software or platform reset. RAL0 and RAL1 together makeup the RAL - Receive Address Low registers. Table 28-90. Receive Address Low 1--RAL1 [0:3][0:7] (0x54E0 + 8*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 RAL 54E0h at Offset Start: 0x8 Offset End: 54E3h at 0x8 Power Well: GBEAUX Bit Description Receive address low Contains the lower 32-bit of the 48-bit Ethernet address. Sticky Bit Reset Value Bit Access X R/W Note: The RAL field should be written in network order. 28.9.1.18 Receive Address High 0--RAH0 [0:3][0:15] (0x5404 + 8*n [n=0...15]; R/W) These registers contain the upper bits of the 48 bit Ethernet address. The complete address is [RAH, RAL]. AV determines whether this address is compared against the incoming packet and is cleared by a master reset. ASEL enables the Controller to perform special filtering on receive packets. After reset, The first register (Receive Address Register 0) is loaded from the IA field in the EEPROM with its Address Select field set to 00b and its Address Valid field set to 1b. Note: The RAH field should be written in network order. Note: The first receive address register (RAH0) is also used for exact match pause frame checking (DA matches the first register). As a result, RAH0 should always be used to store the individual Ethernet MAC address of the Controller. RAH0 and RAH1 together makeup the RAH - Receive Address High registers. Intel(R) Communications Chipset 89xx Series - Datasheet 1342 October 2012 Order Number: 327879-001US 28.0 Table 28-91. Receive Address High 0--RAH0 [0:3][0:15] (0x5404 + 8*n [n=0...15]; 0x54E04 + 8*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit Bit Range 31 30 :26 25 :18 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Power Well: GBEAUX Bit Acronym Bit Description Addr_Valid Address Valid Cleared after master reset. The Address Valid field of the Receive Address Register 0 is set to 1b after a software or PCI reset or EEPROM read. In entries 0-15 this bit is cleared by master reset. Sticky Reserved Reserved Write 0 Ignore on reads. POOLSEL Pool Select In virtualization modes (MRQC.Multiple Receive Queues Enable = 011b) indicates which Pool should get the packets matching this MAC address. This field is a bit map (bit per VM) where more than one bit can be set according to the limitations defined in Section 26.8.3.4. If all the bits are zero, this address is used only for L2 filtering and is not used as part of the queueing decision. 17 :16 ASEL Address Select Selects how the address is to be used in the address filtering. 00b = Destination address (required for normal mode) 01b = Source address. This mode should not be used in virtualization mode. 10b = Reserved 11b = Reserved 15 :00 RAH Receive address High Contains the upper 16 bits of the 48-bit Ethernet address. October 2012 Order Number: 327879-001US 5404h at Offset Start: 0x8 Offset End: 5407h at 0x8 Bit Reset Value Bit Access 0x0 R/W 0x0 R/W X R/W X R/W Intel(R) Communications Chipset 89xx Series - Datasheet 1343 28.9.1.19 Receive Address High 1--RAH1 [0:3][0:7] (0x54E4 + 8*n [n=0...7]; R/W) RAH0 and RAH1 together makeup the RAH - Receive Address High registers. Table 28-92. Receive Address High 1--RAH1 [0:3][0:7] (0x54E4 + 8*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit Bit Range 31 30 :26 25 :18 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Power Well: GBEAUX Bit Acronym Bit Description Addr_Valid Address Valid Cleared after master reset. If an EEPROM is present, the Address Valid field of the Receive Address Register 0 is set to 1b after a software or PCI reset or EEPROM read. In entries 0-15 this bit is cleared by master reset. Reserved Reserved Write 0 Ignore on reads. POOLSEL Pool Select In virtualization modes (MRQC.Multiple Receive Queues Enable = 011b) indicates which Pool should get the packets matching this MAC address. This field is a bit map (bit per VM) where more than one bit can be set according to the limitations defined in Section 26.8.3.4. If all the bits are zero, this address is used only for L2 filtering and is not used as part of the queueing decision. 17 :16 ASEL Address Select Selects how the address is to be used in the address filtering. 00b = Destination address (required for normal mode) 01b = Source address. This mode should not be used in virtualization mode. 10b = Reserved 11b = Reserved 15 :00 RAH Receive address High Contains the upper 16 bits of the 48-bit Ethernet address. Intel(R) Communications Chipset 89xx Series - Datasheet 1344 54E4h at Offset Start: 0x8 Offset End: 54E7h at 0x8 Sticky Bit Reset Value Bit Access 0x0 R/W 0x0 R/W X R/W X R/W October 2012 Order Number: 327879-001US 28.0 28.9.1.20 VLAN Filter Table Array--VFTA [0:3][0:127] (0x5600 + 4*n [n=0...127]; R/W) There is one register per 32 bits of the VLAN Filter Table. The size of the word array depends on the number of bits implemented in the VLAN Filter Table. Software must mask to the desired bit on reads and supply a 32-bit word on writes. Note: All accesses to this table must be 32 bit. The algorithm for VLAN filtering using the VFTA is identical to that used for the Multicast Table Array. See Section 28.9.1.15 for a block diagram of the algorithm. If VLANs are not used, there is no need to initialize the VFTA. Table 28-93. VLAN Filter Table Array--VFTA [0:3][0:127] (0x5600 + 4*n [n=0...127]; R/ W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 Bit Vector Double-word wide bit vector specifying 32 bits in the VLAN Filter table. October 2012 Order Number: 327879-001US 5600h at Offset Start: 0x4 Offset End: 5603h at 0x4 Sticky Bit Reset Value Bit Access X R/W Intel(R) Communications Chipset 89xx Series - Datasheet 1345 28.9.1.21 Multiple Receive Queues Command Register--MRQC [0:3] (0x5818; R/ W) The MRQC.Multiple Receive Queues Enable field is used to enable/disable RSS hashing and also to enable multiple receive queues. Disabling this feature is not recommended. Model usage is to reset the Controller after disabling the RSS. Table 28-94. Multiple Receive Queues Command Register--MRQC [0:3](0x5818; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Offset Start: 5818h Offset End: 581Bh Power Well: GBEAUX Bit Acronym Bit Description 31 :16 RSS Field Enable Each bit, when set, enables a specific field selection to be used by the hash function. Several bits can be set at the same time. Bit[16] = Enable TcpIPv4 hash function Bit[17] = Enable IPv4 hash function Bit[18] = Enable TcpIPv6Ex hash function Bit[19] = Enable IPv6Ex hash function Bit[20] = Enable IPv6 hash function Bit[21] = Enable TCPIPv6 hash function Bit[22] = Enable UDPIPv4 Bit[23] = Enable UDPIPv6 Bit[24] = Enable UDPIPv6Ext Bit[25] = Reserved Bits[31:26] - Reserved 15 :06 Reserved Reserved. Intel(R) Communications Chipset 89xx Series - Datasheet 1346 B:0:1+ Index1 Sticky Bit Reset Value Bit Access 0x0 R/W October 2012 Order Number: 327879-001US 28.0 Table 28-94. Multiple Receive Queues Command Register--MRQC [0:3](0x5818; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 05 :03 02 :00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 5818h Offset End: 581Bh Power Well: GBEAUX Bit Reset Value Bit Access Defines the default queue in non VMDq mode according to value of the Multiple Receive Queues Enable field. If Multiple Receive Queues Enable = 000b: Def_Q defines the destination of all packets not forwarded by filters. 001b: Def_Q field is ignored 010b: Def_Q defines the destination of all packets not forwarded by RSS or filters. 011b - Def_Q field is ignored. Queueing decision of all packets not forwarded by MAC address and Ether-type filters is according to VT_CTL.DEF_PL field. 100-101b: Def_Q field is ignored. 110b: Def_Q field is ignored. Note: In VMDq mode (Multiple Receive Queues Enable = 011b) the default queue is set according to the VT_CTL.DEF_PL if packet passes MAC Address filtering of a filter with RAH.POOLSEL = 0x0 or is a broadcast or multicast packet and does not match Ether-type queuing decision filters. 0x0 R/W Multiple Receive Queues Enable Enables support for Multiple Receive Queues and defines the mechanism that controls queue allocation. 000b = Multiple receive queues as defined by filters (2tuple filters, L2 Ether-type filters, SYN filter and Flex Filters). 001b = Reserved 010b = Multiple receive queues as defined by filters and RSS for 8 queuesa. 011b = Multiple receive queues as defined by VMDq based Multiple Receive on packet destination MAC address (RAH.POOLSEL) and Queues Enable Ether-type queuing decision filters. 100b = Reserved 101b = Reserved 110b = Reserved 111b = Reserved. If VT is not supported, the only allowed values for this field are 000b and 010b. Writing any other value is ignored. Allowed values for this field are 000b, 010b and 011b. Any other value is ignored. 0x0 R/W Bit Acronym Bit Description Def_Q Sticky a. The RXCSUM.PCSD bit should be set to enable reception of the RSS hash value in the receive descriptor. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1347 RSS Random Key Register--RSSRK [0:3][0:9] (0x5C80 + 4*n [n=0...9]; R/W) 28.9.1.22 Table 28-95. RSS Random Key Register--RSSRK [0:3][0:9] (0x5C80 + 4*n [n=0...9]; R/W) Description: View: PCI B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Size: 32 bit Default: 0x0 5C80h at Offset Start: 0x4 Offset End: 5C83h at 0x4 Power Well: GBEAUX Bit Reset Value Bit Access Byte n*4+3 of the RSS random key (n=0,1,...9). 0x0 R/W K2 Byte n*4+2 of the RSS random key (n=0,1,...9). 0x0 R/W K1 Byte n*4+1 of the RSS random key (n=0,1,...9). 0x0 R/W K0 Byte n*4 of the RSS random key (n=0,1,...9). 0x0 R/W Bit Range Bit Acronym 31 :24 K3 23 :16 15 :08 07 :00 Bit Description Sticky The RSS Random Key Register stores a 40 byte key used by the RSS hash function. 31 24 23 K[3] 16 15 8 7 0 K[2] K[1] ... ... ... ... K[39] ... ... K[36] 28.9.1.23 K[0] Redirection Table--RETA [0:3][0:31] (0x5C00 + 4*n [n=0...31]; R/W) The redirection table is a 128-entry table with each entry being eight bits wide. Only 1 to 3 bits of each entry are used to store the queue index. The table is configured through the following R/W registers. Table 28-96. Redirection Table--RETA [0:3][0:31] (0x5C00 + 4*n [n=0...31]; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :24 Entry 3 23 :16 Entry 2 Power Well: GBEAUX Bit Description Determines the tag value and physical queue for index 4*n+3 (n=0...31). Determines the tag value and physical queue for index 4*n+2 (n=0...31). Intel(R) Communications Chipset 89xx Series - Datasheet 1348 5C00h at Offset Start: 0x4 Offset End: 5C03h at 0x4 Sticky Bit Reset Value Bit Access 0x0 R/W 0x0 R/W October 2012 Order Number: 327879-001US 28.0 Table 28-96. Redirection Table--RETA [0:3][0:31] (0x5C00 + 4*n [n=0...31]; R/W) (Sheet 2 of 2) Description: View: PCI BAR: GBEPCIBAR0[0:3] Size: 32 bit Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 15 :08 Entry 1 07 :00 Entry 0 31 24 Tag 3 5C00h at Offset Start: 0x4 Offset End: 5C03h at 0x4 Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 R/W 0x0 R/W Determines the tag value and physical queue for index 4*n+1 (n=0...31). Determines the tag value and physical queue for index 4*n+0 (n=0...31). 23 16 Tag 2 15 8 Tag 1 7 0 Tag 0 ... ... ... ... Tag 127 ... ... ... Each entry (byte) of the redirection table contains the following: 7:3 2:0 Reserved Queue index * Bits [7:3] - Reserved * Bits [2:0] - Queue index for all pools or in regular RSS. In RSS only mode, all bits are used. The contents of the redirection table are not defined following reset of the Memory Configuration Registers. System software must initialize the Table prior to enabling multiple receive queues. It can also update the redirection table during run time. Such updates of the table are not synchronized with the arrival time of received packets. Therefore, it is not guaranteed that a table update takes effect on a specific packet boundary. Note: In case the operating system provides a redirection Table whose size is smaller than 128 bytes, the software usually replicates the operating system-provided redirection table to span the whole 128 bytes of the hardware's redirection table. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1349 28.10 Filtering Register 28.10.1 Detailed Register Descriptions 28.10.1.1 Immediate Interrupt RX--IMIR [0:3][0:7] (0x5A80 + 4*n [n=0...7]; R/W) This register defines the filtering that corrects which packet triggers low latency interrupt. The following IMIREXT register includes a size threshold and a control bits bitmap to trigger an immediate interrupt. Note: The Port field should be written in network order. Note: If one of the actions for this filter is set, then at least one of the PORT_BP, Size_BP, one of the Mask bits or CtrlBit_BP bits should be cleared. Table 28-97. Immediate Interrupt RX--IMIR [0:3][0:7] (0x5A80 + 4*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit Bit Range Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Power Well: GBEAUX Bit Reset Value Bit Access 0h R/W PORT_BP Port Bypass When set to 1b, the TCP port check is bypassed and only other conditions are checked. When set to 0b, the TCP port is checked to fit the port field. X R/W Immediate Interrupt Enables issuing an immediate interrupt when the following conditions are met: * The 2-tuple filter associated with this register matches * The Length filter associated with this filter matches * The TCP flags filter associated with this filter matches 0h R/W 0x0 R/W Bit Acronym Bit Description 31 :29 Filter Priority Defines the priority of the filter assuming two filters with same priority don't match. If two filters with the same priority match the incoming packet, the first filter (lowest ordinal number) is used in order to define the queue destination of this packet. 28 :18 Reserved Reserved 17 16 15 :00 Destination TCP port Destination Port This field is compared with the Destination TCP port in incoming packets. Intel(R) Communications Chipset 89xx Series - Datasheet 1350 5A80h at Offset Start: 0x4 Offset End: 5A83h at 0x4 Sticky October 2012 Order Number: 327879-001US 28.0 28.10.1.2 Immediate Interrupt Rx Ext.--IMIREXT [0:3][0:7] (0x5AA0 + 4*n [n=0...7]; R/W) Table 28-98. Immediate Interrupt Rx Ext.--IMIREXT [0:3][0:7] (0x5AA0 + 4*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] CtrlBit_BP Control Bits Bypass When set to 1b, the control bits check is bypassed. When set to 0b, the control bits check is performed. X R/W CtrlBit Control Bit When a bit in this field equals 1b, an interrupt is immediately issued after receiving a packet with the corresponding TCP control bits turned on. Bit 13 (URG): Urgent pointer field significant Bit 14 (ACK): Acknowledgment field Bit 15 (PSH): Push function Bit 16 (RST): Reset the connection Bit 17 (SYN): Synchronize sequence numbers Bit 18 (FIN): No more data from sender X R/W Size Bypass When 1b, the size check is bypassed. When 0b, the size check is performed. X R/W Size Threshold These 12 bits define a size threshold; a packet with a length below this threshold triggers an interrupt. Enabled by Size_Thresh_en. X R/W Reserved Reserved 11 :00 Power Well: GBEAUX Bit Access 31 :20 12 5AA0h at Offset Start: 0x4 Offset End: 5AA3h at 0x4 Bit Reset Value Bit Acronym 18 :13 B:0:1+ Index1 Default: 0x0 Bit Range 19 Bus:Device:Function: Size_BP Size_Thresh Bit Description Sticky Note: The size used for this comparison is the size of the packet as forwarded to the host and does not include any of the fields stripped by the MAC (VLAN or CRC). As a result, setting the RCTL.SECRC & CTRL.VME bits should be taken into account while calculating the size threshold. Note: The value of the IMIR and IMIREXT registers after reset is unknown (apart from the IMIR.PORT_IM_EN bit which is guaranteed to be cleared). Therefore, both registers should be programmed before IMIR.PORT_IM_EN is set for a given flow. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1351 28.10.1.3 2tuples Queue Filter--TTQF [0:3][0:7] (0x59E0 + 4*n[n=0..7];RW) Table 28-99. 2tuples Queue Filter--TTQF[0:3][0:7] (0x59E0 + 4*n[n=0..7]; R/W) Description: View: PCI Size: 32 bit Bit Range 31 :28 27 BAR: GBEPCIBAR0[0:3] Bit Acronym Mask 1588 Time Stamp Reserved 18 :16 Rx Queue Reserved_1 Bit Description Bit Access When set, packets that match this filter are time stamped according to the IEEE 1588 specification 0x0 R/W Reserved 0h R/W Identifies the Rx queue associated with this 2-tuple filter 0h R/W 0h R/W 0h R/W Reserved Write 1 ignore on read. 11 :09 Reserved Reserved Write 0 ignore on read. Queue Enable When set, enables filtering of Rx packets by the 2-tuples defined in this filter to the queue indicated in this register. Protocol Bit Reset Value 0xF Reserved 28.10.1.4 Sticky Mask bits for the 2-tuple fields . The corresponding field participates in the match if the bit below is cleared: Bit 28 - Mask protocol comparison Bits 31 - 29 Reserved Reserved 07 :00 59E0h at Offset Start: 0x4 Offset End: 59E3h at 0x4 Power Well: GBEAUX 14 :12 08 B:0:1+ Index1 Default: 0x0 26 :19 15 Bus:Device:Function: IP L4 Protocol, part of the 2-tuple queue filters. Immediate Interrupt Rx VLAN Priority--IMIRVP [0:3] (0x5AC0; R/W) Table 28-100.Immediate Interrupt Rx VLAN Priority--IMIRVP [0:3] (0x5AC0; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :04 Reserved Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1352 Offset Start: 5AC0h Offset End: 5AC3h October 2012 Order Number: 327879-001US 28.0 Table 28-100.Immediate Interrupt Rx VLAN Priority--IMIRVP [0:3] (0x5AC0; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 03 02 :00 28.10.1.5 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 5AC0h Offset End: 5AC3h Power Well: GBEAUX Bit Reset Value Bit Access Vlan_pri_en VLAN Priority Enable When set to 1b, an incoming packet with VLAN tag with a priority equal or higher to Vlan_Pri triggers an immediate interrupt, regardless of the EITR moderation. When set to 0b, the interrupt is moderated by EITR. 0h R/W Vlan_Pri VLAN Priority This field includes the VLAN priority threshold. When Vlan_pri_en is set to 1b, then an incoming packet with a VLAN tag with a priority field equal or higher to VlanPri triggers an immediate interrupt, regardless of the EITR moderation. 0h R/W Bit Acronym Bit Description Sticky SYN Packet Queue Filter--SYNQF [0:3] (0x55FC; RW) Table 28-101.SYN Packet Queue Filter--SYNQF [0:3] (0x55FC; RW) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Acronym Power Well: GBEAUX Bit Description Sticky 31 :04 Reserved Reserved 03 :01 Rx Queue Identifies an Rx queue associated with SYN packets. 00 Queue Enable When set, enables forwarding of Rx packets to the queue indicated in this register. October 2012 Order Number: 327879-001US Offset Start: 55FCh Offset End: 55FFh Bit Reset Value Bit Access 0x0 R/W 0h R/W Intel(R) Communications Chipset 89xx Series - Datasheet 1353 28.10.1.6 EType Queue Filter--ETQF [0:3][0:7] (0x5CB0 + 4*n[n=0...7]; RW) Table 28-102.EType Queue Filter--ETQF [0:3][0:7] (0x5CB0 + 4*n[n=0...7]; RW) Description: View: PCI Size: 32 bit Bit Range 31 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Bit Acronym 5CB0h at Offset Start: 0x4 Offset End: 5CB3h at 0x4 Power Well: GBEAUX Bit Description When set, enables filtering of Rx packets by the EType Queue Enable defined in this register to the queue indicated in this register. Sticky Bit Reset Value Bit Access 0h RW 30 1588 time stamp When set, packets with this EType are time stamped according to the IEEE 1588 specification. 0h RW 29 Immediate Interrupt When set, packets that match this filter generate an immediate interrupt. 0x0 RW 28 Reserved Reserved 27 Reserved Reserved 26 Filter enable 0h RW When set, this filter is valid. Any of the actions controlled by the following fields are gated by this field. 25 :19 Reserved Reserved 18 :16 Rx Queue Identifies the receive queue associated with this EType. 0x0 RW Identifies the protocol running on top of IEEE 802. Used to forward Rx packets containing this EType to a specific Rx queue. 0x0 RW 15 :00 EType Intel(R) Communications Chipset 89xx Series - Datasheet 1354 B:0:1+ Index1 October 2012 Order Number: 327879-001US 28.0 28.11 Transmit Registers 28.11.1 Detailed Register Descriptions 28.11.1.1 Transmit Control Register--TCTL [0:3] (0x0400; R/W) This register controls all transmit functions for the Controller. Table 28-103.Transmit Control Register--TCTL [0:3] (0x0400; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x000400F8 Bit Range Bit Acronym 31 :25 Reserved Offset Start: 0400h Offset End: 0403h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0h R/W Reserved Re-transmit on Late Collision When set, enables the Controller to re-transmit on a late collision event. Note: RTLC configures the Controller to perform retransmission of packets when a late collision is detected. The collision window is speed dependent: 64 bytes for 10/ 100 Mb/s and 512 bytes for 1000 Mb/s operation. If a late collision is detected when this bit is disabled, the transmit function assumes the packet has successfully transmitted. This bit is ignored in full-duplex mode. 24 RTLC 23 Reserved Reserved SWXOFF Software XOFF Transmission When set to 1b, the Controller schedules the transmission of an XOFF (PAUSE) frame using the current value of the PAUSE timer (FCTTV.TTV). This bit self-clears upon transmission of the XOFF frame. Note: While 802.3x flow control is only defined during full duplex operation, the sending of PAUSE frames via the SWXOFF bit is not gated by the duplex settings within the Controller. Software should not write a 1b to this bit while the Controller is configured for half-duplex operation. 0h R/W BST Back-Off Slot Time This value determines the back-off slot time value in byte time. 40h R/W CT Collision Threshold This determines the number of attempts at retransmission prior to giving up on the packet (not including the first transmission attempt). While this can be varied, it should be set to a value of 15 in order to comply with the IEEE specification requiring a total of 16 attempts. The Ethernet back-off algorithm is implemented and clamps to the maximum number of slot-times after 10 retries. This field only has meaning when in half-duplex operation. Note: Software can choose to abort packet transmission in less than the Ethernet mandated 16 collisions. For this reason, hardware provides CT support. Fh R/W 03 PSP Pad Short Packets 0b = Do not pad. 1b = Pad. Padding makes the packet 64 bytes long. This is not the same as the minimum collision distance. If padding of short packets is allowed, the total length of a packet not including FCS should be not less than 17 bytes. 1h R/W 02 Reserved 22 21 :12 11 :04 October 2012 Order Number: 327879-001US Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1355 Table 28-103.Transmit Control Register--TCTL [0:3] (0x0400; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x000400F8 Offset Start: 0400h Offset End: 0403h Power Well: GBEAUX Bit Acronym Bit Description 01 EN Transmit Enable The transmitter is enabled when this bit is set to 1b. Writing 0b to this bit stops transmission after any in progress packets are sent. Data remains in the transmit FIFO until the device is re-enabled. Software should combine this operation with reset if the packets in the TX FIFO should be flushed. 00 Reserved 28.11.1.2 B:0:1+ Index1 Sticky Bit Reset Value Bit Access 0h R/W Reserved Write as 0b for future compatibility. Transmit Control Extended--TCTL_EXT [0:3] (0x0404; R/W) This register controls late collision detection. The COLD field is used to determine the latest time in which a collision indication is considered as a valid collision and not a late collision. When using an SGMII connected external PHY, the SGMII interface adds some delay on top of the time budget allowed by the specification (collisions in valid network topographies even after 512-bit time can be expected). In order to accommodate this condition, COLD should be updated to take the SGMII inbound and outbound delays. Table 28-104.Transmit Control Extended--TCTL_EXT [0:3] (0x0404; R/W) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x00010840 Bit Range Bit Acronym 31 :20 Reserved 19 :10 COLD 09 :00 Reserved Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x42 R/W Reserved. Collision Distance Used to determine the latest time in which a collision indication is considered as a valid collision and not a late collision. Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1356 Offset Start: 0404h Offset End: 0407h October 2012 Order Number: 327879-001US 28.0 28.11.1.3 Transmit IPG Register--TIPG (0x0410; R/W) This register controls the Inter Packet Gap (IPG) timer. Table 28-105.Transmit IPG Register--TIPG [0:3] (0x0410; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Power Well: GBEAUX Bit Access IPGR IPG After Deferral Specifies the total IPG time for non back-to-back transmissions (transmission following deferral) in half duplex. Measured in increments of the MAC clock: 8 ns MAC clock when operating @ 1 Gb/s. 80 ns MAC clock when operating @ 100 Mb/s 800 ns MAC clock when operating @ 10 Mb/s. An offset of 5-byte times must be added to the programmed value to determine the total IPG after a defer event. A value of 7 is recommended to achieve a 12-byte effective IPG. The IPGR must never be set to a value greater than IPGT. If IPGR is set to a value equal to or larger that IPGT, it overrides the IPGT IPG setting in half duplex resulting in inter-packet gaps that are larger then intended by IPGT. In this case, full duplex is unaffected and always relies on IPGT. 0x06 R/W IPGR1 IPG Part 1 Specifies the portion of the IPG in which the transmitter defers to receive events. IPGR1 should be set to 2/3 of the total effective IPG (8). Measured in increments of the MAC clock: 8 ns MAC clock when operating @ 1 Gb/s. 80 ns MAC clock when operating @ 100 Mb/s 800 ns MAC clock when operating @ 10 Mb/s. 0x04 R/W IPGT IPG Back to Back Specifies the IPG length for back to back transmissions in both full and half duplex. Measured in increments of the MAC clock: 8 ns MAC clock when operating @ 1 Gb/s. 80 ns MAC clock when operating @ 100 Mb/s. 800 ns MAC clock when operating @ 10 Mb/s. IPGT specifies the IPG length for back-to-back transmissions in both full duplex and half duplex. An offset of 4 byte times is added to the programmed value to determine the total IPG. As a result, a value of 8 is recommended to achieve a 12 byte time IPG. 0x08 R/W 31 :30 Reserved 09 :00 Offset Start: 0410h Offset End: 0413h Bit Reset Value Bit Acronym 19 :10 B:0:1+ Index1 Default: 0x00601008 Bit Range 29 :20 Bus:Device:Function: October 2012 Order Number: 327879-001US Bit Description Sticky Reserved Read as 0b. Should be written with 0b for future compatibility. Intel(R) Communications Chipset 89xx Series - Datasheet 1357 28.11.1.4 Retry Buffer Control--RETX_CTL [0:3] (0x041C; RW) This register controls the collision retry buffer. Table 28-106.Retry Buffer Control--RETX_CTL [0:3] (0x041C; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x00000003 Bit Range Bit Acronym 31 :04 Reserved 03 :00 Water Mark 28.11.1.5 Bus:Device:Function: Offset Start: 041Ch Offset End: 041Fh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x3 RW Reserved Retry buffer water mark. This parameters defines the minimal number of QWords that should be present in the retry buffer before transmission is started. DMA Tx Control--DTXCTL [0:3] (0x3590; R/W) This register is used to set some parameters controlling the DMA Tx behavior. Table 28-107.DMA Tx Control--DTXCTL [0:3] (0x3590; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x00400004 Bit Range Bit Acronym 31 :26 Reserved Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x4 R/W Reserved Context switch threshold. Defines the amount of back to back transmit context descriptors above which the driver will be considered as malicious. Maximum value should be less than 19. 25 :20 Cswthresh 19 :06 Reserved Reserved 05 MDP_EN Malicious driver protection enable 0b = mechanism is disabled. 1b = mechanism is enabled. 0h R/W 04 OutOfSyncEnable 0b = Out Of sync mechanism is disabled. 1b = Out Of sync mechanism is enabled. 0h R/W 03 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1358 Offset Start: 3590h Offset End: 3593h October 2012 Order Number: 327879-001US 28.0 Table 28-107.DMA Tx Control--DTXCTL [0:3] (0x3590; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 02 01 :00 28.11.1.6 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x00400004 Power Well: GBEAUX Bit Acronym Bit Description 8023LL 802.3 length location 1b = The location of the 802.3 length field in 802.3+SNAP packets, is assumed to be 8 bytes before the end of the MAC header. 0b = The location of the 802.3 length field in 802.3+SNAP packets, is calculated from the beginning of the MAC header assuming no VLAN present in the packet sent by the software. This bit is used only in case of large send (TSO) with SNAP mode. Reserved Offset Start: 3590h Offset End: 3593h Sticky Bit Reset Value Bit Access 1h R/W Reserved. DMA TX TCP Flags Control Low--DTXTCPFLGL [0:3] (0x359C; RW) This register holds the buses that "AND" the control flags in TCP header for the first and middle segments of a TSO packet. Table 28-108.DMA TX TCP Flags Control Low--DTXTCPFLGL [0:3] (0x359C; RW) Description: View: PCI Size: 32 bit Bit Range 31 :28 27 :16 15 :12 11 :00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x00400004 Bit Acronym Reserved Power Well: GBEAUX Bit Description Sticky Bit Access 0xF76 RW 0xFF6 RW Reserved TCP_flg_first_se TCP Flags first segment. Bits that are used to execute an g AND operation with the TCP flags in the TCP header in the first segment October 2012 Order Number: 327879-001US Bit Reset Value Reserved TCP_Flg_mid_se TCP Flags middle segments. Bits that are used to execute g an AND operation with the TCP flags in the TCP header in the middle segments Reserved Offset Start: 359Ch Offset End: 359Fh Intel(R) Communications Chipset 89xx Series - Datasheet 1359 28.11.1.7 DMA TX TCP Flags Control High--DTXTCPFLGH [0:3] (0x35A0; RW) This register holds the buses that "AND" the control flags in TCP header for the last segment of a TSO packet. See Section 26.2.4.7.3 for details of use of this register. Table 28-109.DMA TX TCP Flags Control High--DTXTCPFLGH [0:3] (0x35A0; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bit Acronym 31 :12 Reserved 28.11.1.8 B:0:1+ Index1 Default: 0x00000F7F Bit Range 11 :00 Bus:Device:Function: Offset Start: 35A0h Offset End: 35A3h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0xF7F RW Reserved. TCP Flags last segment. Bits that are used to execute an TCP_Flg_lst_seg AND operation with the TCP flags at TCP header in the last segment DMA TX Max Total Allow Size Requests--DTXMXSZRQ [0:3] (0x3540; RW) This register limits the total number of data bytes that might be in outstanding PCIe requests from the host memory. This allows requests to send low latency packets to be serviced in a timely manner, as this request is serviced right after the current outstanding requests are completed. Table 28-110.DMA TX Max Total Allow Size Requests--DTXMXSZRQ [0:3] (0x3540; RW) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x00000010 Bit Range Bit Acronym 31 :12 Reserved Offset Start: 3540h Offset End: 3543h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x10 rW Reserved Maximum allowable size of concurrent TX outstanding requests on PCIe. 11 :00 Max_bytes_num Field defines maximum size in 256 byte resolution of _req outstanding TX requests to be sent on PCIe. If total amount of outstanding TX requests is higher than defined in this field, no further TX outstanding requests are sent. Intel(R) Communications Chipset 89xx Series - Datasheet 1360 October 2012 Order Number: 327879-001US 28.0 28.11.1.9 DMA TX Maximum Packet Size--DTXMXPKTSZ [0:3] (0x355C; RW) This register limits the total number of data bytes that might be transmitted in a single frame. Reducing packet size enables better utilization of transmit buffer. Table 28-111.DMA TX Maximum Packet Size--DTXMXPKTSZ [0:3] (0x355C; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x00000098 Bit Range Bit Acronym 31 :09 Reserved 08 :00 Bus:Device:Function: Offset Start: 355Ch Offset End: 355Fh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x98 RW Reserved Maximum transmit packet size that is allowed to be transmitted by the driver. Value entered is in 64 Bytes MAX_TPKT_SIZ resolution. E Default value enables transmission of maximum sized 9728 Byte Jumbo frames. 28.11.1.10 Transmit Descriptor Base Address Low--TDBAL [0:3][0:7] (0xE000 + 0x40*n [n=0...7]; R/W) These registers contain the lower 32 bits of the 64-bit descriptor base address. The lower 7 bits are ignored. The Transmit Descriptor Base Address must point to a 128byte aligned block of data. Table 28-112.Transmit Descriptor Base Address Low--TDBAL [0:3][0:7] (0xE000 + 0x40*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :07 TDBAL 06 :00 Lower_0 Note: Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Offset Start: E000h Offset End: E003h Power Well: GBEAUX Bit Description Transmit Descriptor Base Address Low Ignored on writes. Returns 0x0 on reads. Sticky Bit Reset Value Bit Access X R/W 0x0 R/W To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x3800, 0x3900, 0x3A00 & 0x3B00 respectively. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1361 28.11.1.11 Transmit Descriptor Base Address High--TDBAH [0:3][0:7] (0xE004 + 0x40*n [n=0...7]; R/W) These registers contain the upper 32 bits of the 64-bit descriptor base address. Table 28-113.Transmit Descriptor Base Address High--TDBAH [0:3][0:7] (0xE004 + 0x40*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :00 TDBAH Note: B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] E004h at Offset Start: 0x40 Offset End: E007h at 0x40 Power Well: GBEAUX Bit Description Sticky Transmit Descriptor Base Address [63:32] Bit Reset Value Bit Access X R/W To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x3804, 0x3904, 0x3A04 & 0x3B04 respectively. 28.11.1.12 Transmit Descriptor Ring Length--TDLEN [0:3][0:7] (0xE008 + 0x40*n [n=0...7]; R/W) These registers contain the descriptor ring length. The registers indicates the length in bytes and must be 128-byte aligned. Table 28-114.Transmit Descriptor Ring Length--TDLEN [0:3][0:7] (0xE008 + 0x40*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :20 Reserved 19 :07 LEN 06 :00 0 Note: B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Power Well: GBEAUX Bit Reset Value Bit Access Descriptor Ring Length (number of 8 descriptor sets) 0x0 R/W Ignore on writes. Reads back as 0b. 0x0 R/W Bit Description Sticky Reserved Reads as 0b. Should be written to 0b. To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x3808, 0x3908, 0x3A08 & 0x3B08 respectively. Intel(R) Communications Chipset 89xx Series - Datasheet 1362 E008h at Offset Start: 0x40 Offset End: E00Bh at 0x40 October 2012 Order Number: 327879-001US 28.0 28.11.1.13 Transmit Descriptor Head--TDH [0:3][0:7] (0xE010 + 0x40*n [n=0...7]; RO) These registers contain the head pointer for the transmit descriptor ring. It points to a 16-byte datum. Hardware controls this pointer. Note: The values in these registers might point to descriptors that are still not in host memory. As a result, the host cannot rely on these values in order to determine which descriptor to release. Table 28-115.Transmit Descriptor Head--TDH [0:3] [0:7] (0xE010 + 0x40*n [n=0...7]; RO) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym 31 :16 Reserved Reserved Should be written to 0b. 15 :00 TDH Transmit Descriptor Head Note: E010h at Offset Start: 0x40 Offset End: E013h at 0x40 Bit Description Sticky Bit Reset Value Bit Access 0x0 RO To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x3810, 0x3910, 0x3A10 & 0x3B10 respectively. 28.11.1.14 Transmit Descriptor Tail--TDT [0:3][0:7] (0xE018 + 0x40*n [n=0...7]; R/W) These registers contain the tail pointer for the transmit descriptor ring and points to a 16-byte datum. Software writes the tail pointer to add more descriptors to the transmit ready queue. Hardware attempts to transmit all packets referenced by descriptors between head and tail. Table 28-116.Transmit Descriptor Tail--TDT [0:3][0:7] (0xE018 + 0x40*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :16 Reserved 15 :00 TDT Note: Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] E018h at Offset Start: 0x40 Offset End: E01Bh at 0x40 Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 R/W Reserved Reads as 0b. Should be written to 0b for future compatibility. Transmit Descriptor Tail To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x3818, 0x3918, 0x3A18 & 0x3B18 respectively. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1363 28.11.1.15 Transmit Descriptor Control--TXDCTL [0:3][0:7] (0xE028 + 0x40*n [n=0...7]; R/W) These registers control the fetching and write-back operations of transmit descriptors. The three threshold values are used to determine when descriptors are read from and written to host memory. The values are in units of descriptors (each descriptor is 16 bytes). Since write-back of transmit descriptors is optional (under the control of RS bit in the descriptor), not all processed descriptors are counted with respect to WTHRESH. Descriptors start accumulating after a descriptor with RS set is processed. In addition, with transmit descriptor bursting enabled, some descriptors are written back that did not have RS set in their respective descriptors. Note: When WTHRESH = 0b, only descriptors with the RS bit set are written back Table 28-117.Transmit Descriptor Control--TXDCTL [0:3][0:7] (0xE028 + 0x40*n [n=0...7]; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 :28 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Acronym HWBTHRESH E028h at Offset Start: 0x40 Offset End: E02Bh at 0x40 Power Well: GBEAUX Bit Description Transmit Head writeback threshold If value of field is greater than 0x0, head writeback to host will occur only when the amount of internal pending write backs exceeds this threshold. See Section 26.2.3 for additional information. Sticky Bit Reset Value Bit Access 0x0 R/W 0h R/W 0h R/W Note: When activating this mode the WB_on_EITR bit in the TDWBAL register should be set to guarantee a write back after a timeout even if the threshold has not been reached. 27 26 Priority Transmit queue priority 0 - Low priority 1 - High priority When set, Transmit DMA resources are always allocated to the queue before low priority queues. Arbitration between transmit queues with same priority is done in a Round Robin fashion. Transmit Software Flush This bit enables software to trigger descriptor write-back flushing, independently of other conditions. This bit is self cleared by hardware. Bit will clear after write-back flush is triggered (may take a number of SWFLSH (WC) cycles). Note: When working in head write-back mode (TDWBAL.Head_WB_En = 1) TDWBAL.WB_on_EITR bit should be set for transmit descriptor flush to occur. Intel(R) Communications Chipset 89xx Series - Datasheet 1364 October 2012 Order Number: 327879-001US 28.0 Table 28-117.Transmit Descriptor Control--TXDCTL [0:3][0:7] (0xE028 + 0x40*n [n=0...7]; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 25 24 :21 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Bit Acronym Bit Description Sticky ENABLE Transmit Queue Enable When set, this bit enables the operation of a specific transmit queue. Setting this bit initializes the Tail and Head registers (TDT[n] and TDH[n]) of a specific queue. Until then, the state of the queue is kept and can be used for debug purposes. When disabling a queue, this bit is cleared only after all transmit activity on this queue is stopped. Note: When transmit queue is enabled and descriptors exist, descriptors and data are fetched immediately. Actual transmit activity on port starts only if the TCTL.EN bit is set. Reserved 20 :16 WTHRESH Write-Back Threshold Controls the write-back of processed transmit descriptors. This threshold refers to the number of transmit descriptors in the on-chip buffer that are ready to be written back to host memory. In the absence of external events (explicit flushes), the write-back occurs only after at least WTHRESH descriptors are available for write-back. Possible values for this field are 0 to 23. Note: Since the default value for write-back threshold is 0b, descriptors are normally written back as soon as they are processed. WTHRESH must be written to a non-zero value to take advantage of the write-back bursting capabilities of the Controller. 15 :13 Reserved Reserved 12 :08 HTHRESH Host Threshold Prefetch of transmit descriptors is considered when number of valid transmit descriptors in host memory is at least HTHRESH. Note: HTHRESH should be given a non zero value each time PTHRESH is used. 07 :05 Reserved Reserved PTHRESH Prefetch Threshold Controls when a prefetch of descriptors is considered. This threshold refers to the number of valid, unprocessed transmit descriptors the Controller has in its on-chip buffer. If this number drops below PTHRESH, the algorithm considers pre-fetching descriptors from host memory. However, this fetch does not happen unless there are at least HTHRESH valid descriptors in host memory to fetch. Note: When PTHRESH is 0x0 a Transmit descriptor fetch operation is done when any valid descriptors are available in Host memory and space is available in internal buffer. Note: E028h at Offset Start: 0x40 Offset End: E02Bh at 0x40 Power Well: GBEAUX Reserved 04 :00 B:0:1+ Index1 Bit Reset Value Bit Access 0h R/W 0x0 R/W 0x0 R/W 0x0 R/W To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x3828, 0x3928, 0x3A28 and 0x3B28 respectively. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1365 28.11.1.16 Tx Descriptor Completion Write-Back Address Low--TDWBAL [0:3][0:7] (0xE038 + 0x40*n [n=0...7]; R/W) Table 28-118.Tx Descriptor Completion Write-Back Address Low--TDWBAL [0:3][0:7] (0xE038 + 0x40*n [n=0...7]; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Bit Reset Value Bit Access Lowest 32 bits of the head write-back memory location (DWORD aligned). Last 2 bits of this field are ignored and are always read as 0.0, meaning that the actual address is QWORD aligned. 0x0 R/W WB_on_EITR When set, a head write back is done upon EITR expiration. 0h R/W Head_WB_En Head Write-Back Enable 1b = Head write-back is enabled. 0b = Head write-back is disabled. When head_WB_en is set, SN_WB_en is ignored and no descriptor write-back is executed. 0h R/W Bit Acronym Bit Description 31 :02 HeadWB_Low 00 Note: E038h at Offset Start: 0x40 Offset End: E03Bh at 0x40 Power Well: GBEAUX Bit Range 01 B:0:1+ Index1 Sticky To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x3838, 0x3938, 0x3A38 & 0x3B38 respectively. 28.11.1.17 Tx Descriptor Completion Write-Back Address High - TDWBAH [0:3][0:7] (0xE03C + 0x40*n [n=0...7];R/W) Table 28-119.Tx Descriptor Completion Write-Back Address High - TDWBAH [0:3][0:7] (0xE03C + 0x40*n [n=0...7];R/W) Description: View: PCI Size: 32 bit Bit Range 31 :00 Note: Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Acronym Power Well: GBEAUX Bit Description HeadWB_High Highest 32 bits of the head write-back memory location. Sticky Bit Reset Value Bit Access 0x0 R/W To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x383C, 0x393C, 0x3A3C & 0x3B3C respectively. Intel(R) Communications Chipset 89xx Series - Datasheet 1366 E03Ch at Offset Start: 0x40 Offset End: E03Fh at 0x40 October 2012 Order Number: 327879-001US 28.0 28.12 DCA and TPH Registers 28.12.1 Detailed Register Descriptions 28.12.1.1 Rx DCA Control Registers--RXCTL [0:3][0:7] (0xC014 + 0x40*n [n=0...7]; R/W) Note: RX data write no-snoop is activated when the NSE bit is set in the receive descriptor. Table 28-120.Rx DCA Control Registers--RXCTL [0:3][0:7] (0xC014 + 0x40*n [n=0...7]; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0000A200 Bit Acronym 31 :24 CPUID 23 :16 Reserved C014h at Offset Start: 0x40 Offset End: C017h at 0x40 Power Well: GBEAUX Bit Description Sticky Physical ID Legacy DCA capable platforms - the device driver, upon discovery of the physical CPU ID and CPU Bus ID, programs the CPUID field with the Physical CPU and Bus ID associated with this Rx queue. Bit Reset Value Bit Access 0x0 R/W Reserved 15 RxRepHeader ROEn Receive Replicated/Split Header Relax Order Enable 1h R/W 14 RxRepHeader NSEn Receive Replicated/Split Header No Snoop Enable This bit must be reset to 0b to ensure correct functionality of header write to host memory. Note: When TPH is enabled No Snoop bit should be 0. 0h R/W 13 RXdataWrite ROEn Receive Data Write Relax Order Enable (header replication: header and data) 1h R/W 12 RXdataWrite NSEn Receive Data Write No Snoop Enable (header replication: header and data) When set to 0b, the last bit of the Packet Buffer Address field in the advanced receive descriptor is used as the LSB of the packet buffer address (A0), thus enabling Byte alignment of the buffer. When set to 1b, the last bit of the Packet Buffer Address field in advanced receive descriptor is used as the NoSnoop Enabling (NSE) bit (buffer is Word aligned). If also set to 1b, the NSE bit determines whether the data buffer is snooped or not. Note: When TPH is enabled No Snoop bit should be 0. 0h R/W 11 RXdescWBROen (RO) Receive Descriptor Write-Back Relax Order Enable This bit must be reset to 0b to ensure correct functionality of descriptor write-back. 0h R/W 10 Receive Descriptor Write-Back No Snoop Enable RXdescWBNSen This bit must be reset to 0b to ensure correct functionality of descriptor write-back. Note: When TPH is enabled No Snoop bit should be 0. 0h R/W 09 RXdescReadROE Receive Descriptor Read Relax Order Enable n 1h R/W October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1367 Table 28-120.Rx DCA Control Registers--RXCTL [0:3][0:7] (0xC014 + 0x40*n [n=0...7]; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bus:Device:Function: Default: 0x0000A200 C014h at Offset Start: 0x40 Offset End: C017h at 0x40 Power Well: GBEAUX Bit Access 08 Receive Descriptor Read No Snoop Enable This bit must be reset to 0b to ensure correct functionality RXdescReadNSE (Except if the software driver can guarantee the data is n present in the main memory before the DMA process occurs). Note: When TPH is enabled No Snoop bit should be 0. 0h R/W 07 Receive Payload DCA Enable Rx Payload DCA When set, hardware enables DCA for all Ethernet payloads written into memory. When cleared, hardware does not EN enable DCA for Ethernet payloads. This bit is cleared as a default. 0h R/W 06 Receive Header DCA Enable Rx Header DCA When set, hardware enables DCA for all received header EN buffers. When cleared, hardware does not enable DCA for Rx headers. This bit is cleared as a default. 0h R/W 05 Descriptor DCA Enable RX Descriptor When set, hardware enables DCA for all Rx descriptors written back into memory. When cleared, hardware does DCA EN not enable DCA for descriptor write-backs. This bit is cleared as a default. 0h R/W 0h R/W 04 03 :00 Bit Acronym Bit Description Reserved Reserved Reserved Reserved Sticky To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x2814, 0x2914, 0x2A14 & 0x2B14 respectively. Intel(R) Communications Chipset 89xx Series - Datasheet 1368 B:0:1+ Index1 Bit Reset Value Bit Range Note: BAR: GBEPCIBAR0[0:3] October 2012 Order Number: 327879-001US 28.0 28.12.1.2 Tx DCA Control Registers--TXCTL [0:3][0:7] (0xE014 + 0x40*n [n=0...7]; R/W) Table 28-121.Tx DCA Control Registers--TXCTL [0:3][0:7] (0xE014 + 0x40*n [n=0...7]; R/ W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x00002A00 Bit Range Bit Acronym 31 :24 CPUID 23 :14 Reserved E014h at Offset Start: 0x40 Offset End: E017h at 0x40 Power Well: GBEAUX Bit Description Sticky Physical ID Legacy DCA capable platforms - the device driver, upon discovery of the physical CPU ID and CPU Bus ID, programs the CPUID field with the Physical CPU and Bus ID associated with this Tx queue. Bit Reset Value Bit Access 0x0 R/W Reserved Write 0 ignore on read. 13 TXDataReadROE Tx Data Read Relax Order Enable n 1h R/W 12 TXDataReadNSE Tx Data Read No Snoop Enable n 0h R/W 11 TXdescWBROen 1h R/W 10 Tx Descriptor Write-Back No Snoop Enable TXdescWBNSen This bit must be reset to 0b to ensure correct functionality of descriptor write-back. Also applies to head write-back, when enabled. 0h R/W 09 TXdescRDROEn Tx Descriptor Read Relax Order Enable 1h R/W 08 Tx Descriptor Read No Snoop Enable This bit must be reset to 0b to ensure correct functionality TXdescRDNSen (unless the software device driver has written this bit with a write-through instruction). Note: When TPH is enabled No Snoop bit should be 0. 0h R/W 0h R/W 07 :06 05 04 :00 Note: Bus:Device:Function: Reserved TX Descriptor DCA EN Reserved Tx Descriptor Write-Back Relax Order Enable Applies to head write-back, when enabled. Reserved Descriptor DCA Enable When set, hardware enables DCA for all Tx descriptors written back into memory. When cleared, hardware does not enable DCA for descriptor write-backs. This bit is cleared as a default and also applies to head write-back when enabled. Reserved To keep compatibility with the 82575, for queues 0-3, these registers are aliased to addresses 0x3814, 0x3914, 0x3A14 & 0x3B14 respectively. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1369 28.12.1.3 DCA Requester ID Information--DCA_ID [0:3] (0x5B70; RO) Note: The DCA requester ID field, composed of Device ID, Bus #, and Function # is set up in MMIO space for software to program the DCA Requester ID Authentication register. Table 28-122.DCA Requester ID Information--DCA_ID [0:3] (0x5B70; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bit Acronym 31 :16 Reserved 15 :08 Bus Number 02 :00 28.12.1.4 B:0:1+ Index1 Default: 0x0 Bit Range 07 :03 Bus:Device:Function: Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RO 0x0 RO 0h RO Reserved Bus Number Bus number assigned to the function based on BIOS/ operating system enumeration. Device Number Device Number Device number assigned to the function based on BIOS/ operating system enumeration. Function Number Offset Start: 5B70h Offset End: 5B73h Function Number Function number assigned to the function based on BIOS/ operating system enumeration. DCA Control--DCA_CTRL [0:3] (0x5B74; R/W) This CSR is common to all functions. Table 28-123.DCA Control--DCA_CTRL [0:3] (0x5B74; R/W) Description: View: PCI Size: 32 bit Default: 0x1 Bit Range Bit Acronym 31 :05 Reserved 04 :01 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] DCA_MODE Offset Start: 5B74h Offset End: 5B77h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 R/W 1h R/W Reserved DCA Mode 000b = Legacy DCA is supported. The TAG field in the TLP header is based on the following coding: bit 0 is DCA enable; bits 3:1 are CPU ID). 001b - 111b = Reserved 00 DCA_DIS DCA Disable 0b = DCA tagging is enabled. 1b = DCA tagging is disabled. Intel(R) Communications Chipset 89xx Series - Datasheet 1370 October 2012 Order Number: 327879-001US 28.0 28.13 Virtualization Registers 28.13.1 Detailed Register Descriptions 28.13.1.1 VMDq Control Register--VT_CTL [0:3] (0x581C; R/W) Table 28-124.VMDq Control Register--VT_CTL [0:3] (0x581C; R/W) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 581Ch Offset End: 581Fh Power Well: GBEAUX Bit Reset Value Bit Access Replication Enable 0h R/W Bit Acronym Bit Description Sticky 31 Reserved 30 Rpl_En 29 Dis_Def_Pool Drop if no pool is found. If this bit is asserted, then in a RX switching virtualized environment, if there is no destination pool, the packet is discarded and not sent to the default pool. Otherwise, it is sent to the pool defined by the DEF_PL field. 0h R/W 28 IGMAC If set, MAC address is ignored during pool decision. Pooling is based on VLAN only. If this bit is set, then the VMOLR.strvlan should be set to the same value for all pools, 0h R/W FLP Filter local packets - filter incoming packets whose MAC source address matches one of the LAN port DA MAC addresses. If the SA of the received packet matches one of the DA in the RAH/RAL registers, then the VM tied to this DA does not receive the packet. Other VMs can still receive it. 0h R/W 000h R/W 27 26 :10 Reserved 09 :07 DEF_PL 06 :00 Reserved October 2012 Order Number: 327879-001US Reserved Reserved Default pool - used to queue packets that did not pass any VM queuing decision. Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1371 28.13.1.2 Malicious Driver Free Block--MDFB [0:3] (0x3558; RWS) Table 28-125.Malicious Driver Free Block--MDFB [0:3] (0x3558; RWS) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 Block Queue 28.13.1.3 Bus:Device:Function: Offset Start: 3558h Offset End: 355Bh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RWS Reserved Indicates queue that was blocked due to malicious behavior. When bit is set, to commence activity on offending queue, Software should initiate a software reset and re-initialize all queues on the port. Last VM Misbehavior Cause--LVMMC [0:3] (0x3548; RC) Bits in LVMMC register define the cause for blocking the malicious queue that was reported in the MDFB.Block Queue field. Table 28-126. Last VM Misbehavior Cause--LVMMC[0:3] (0x3548; RC) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :29 Last_Q 28 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Mal_PF Power Well: GBEAUX Bit Reset Value Bit Access Last queue that detected malicious behavior. 0x0 RC Malicious Driver behavior detected on current PF 0h RC 0h RC Bit Description Reserved Reserved Ignore on read write 0. 24 ILL_DBU Illegal DBU configuration. 23 Reserved Reserved 27 :25 Sticky 22 No EOP Packet without EOP (i.e. bigger than the ring size) 0h RC 21 Wrong_null Null without EOP 0h RC 20 DESC_TYPE Wrong descriptor type (other than 2,3) 0h RC 19 Reserved Reserved 18 SSO_TCP Wrong parameter of headers for TCP SSO 0h RC 17 SSO_UDP Wrong parameter of headers for UDP SSO 0h RC Reserved Reserved 0h RC 16 :12 11 SCTP_aligned 10 Off_Ill 09 Adv_Size CRC request of non 4 byte aligned data. Illegal offload request. 0h RC Illegal advanced descriptor size. 0h RC Intel(R) Communications Chipset 89xx Series - Datasheet 1372 Offset Start: 3548h Offset End: 354Bh October 2012 Order Number: 327879-001US 28.0 Table 28-126. Last VM Misbehavior Cause--LVMMC[0:3] (0x3548; RC) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Bit Reset Value Bit Access Illegal legacy descriptor size. 0h RC Illegal SCTP header was detected in a single send operation. 0h RC 0h RC Bit Acronym Bit Description Sticky Leg_Size 07 SCTP SSO 06 UDP LSO Illegal UDP header was detected in a large send operation. 05 Reserved Reserved 04 TCP LSO Offset Start: 3548h Offset End: 354Bh Power Well: GBEAUX 08 03 B:0:1+ Index1 Illegal TCP header was detected in a large send operation. Wrong MAC_IP Wrong MAC+IP header size 0h RC 0h RC 02 IPV6 Header Illegal IPV6 header size. 0h RC 01 IPV4 Header Illegal IPV4 header size. 0h RC 00 Mac Header Illegal MAC header size. 0h RC 28.13.1.4 VM Offload Register--VMOLR [0:3][0:7] (0x5AD0 + 4*n [n=0...7]; RW) This register controls the offload and queueing options applied to each pool (VM). Table 28-127.VM Offload Register--VMOLR [0:3][0:7] (0x5AD0 + 4*n [n=0...7]; RW) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x80002600 5AD0h at Offset Start: 0x4 Offset End: 5AD3h at 0x4 Power Well: GBEAUX Bit Reset Value Bit Access 0h RW multicast promiscuous 0h RW Bit Acronym Bit Description Sticky 31 Reserved 30 strvlan 29 Reserved 28 mpe 27 bam Broadcast accept 0h RW 26 rope receive overflow packets - accept packets that match the UTA table. 0h RW 25 rompe receive overflow multicast packets - accept packets that match the MTA table. 0h RW October 2012 Order Number: 327879-001US Reserved - must be set to one. VLAN strip Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1373 Table 28-127.VM Offload Register--VMOLR [0:3][0:7] (0x5AD0 + 4*n [n=0...7]; RW) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 24 23: 17 16 BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x80002600 Bit Acronym Reserved lpe Bit Description Sticky Reserved rlpml Bit Reset Value Bit Access 0h RW 0h RW 2600h RW Reserved Long packet enable 13 :00 5AD0h at Offset Start: 0x4 Offset End: 5AD3h at 0x4 Power Well: GBEAUX Accept Untagged packets enable. When set, packets without VLAN tag can be forwarded to this queue, assuming they pass the MAC address queueing mechanism. aupe 15 :14 28.13.1.5 Bus:Device:Function: Reserved Long packet size (9k default) Replication Offload Register--RPLOLR [0:3] (0x5AF0; RW) This register describes the off loads applied to multicast packets. Table 28-128.Replication Offload Register--RPLOLR [0:3] (0x5AF0; RW Description: View: PCI Size: 32 bit Bit Range 31 30 29 :00 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x80000000 Bit Acronym Reserved strvlan Reserved Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0h RW Reserved - must be set to one. VLAN strip Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1374 Offset Start: 5AF0h Offset End: 5AF3h October 2012 Order Number: 327879-001US 28.0 28.13.1.6 VLAN VM Filter--VLVF [0:3][0:31] (0x5D00 + 4*n [n=0...31]; RW) This register set describes which VLANs the local VMs are part of. Each of the 32 registers contains a VLAN tag and a list of the VMs which are part of it. Only packets with a VLAN matching one of the VLAN tags of which the VM is member of are forwarded to this VM Table 28-129.VLAN VM Filter--VLVF [0:3][0:31] (0x5D00 + 4*n [n=0...31]; RW) Description: View: PCI Size: 32 bit Bit Range 31 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Acronym VI_En 5DD0h at Offset Start: 0x4 Offset End: 5DD3h at 0x4 Power Well: GBEAUX Bit Description Sticky VLAN Id Enable - this filter is valid. Note: If RCTL.VFE is 0 all VLVF filters are disabled. Bit Reset Value Bit Access 0h RW 30 :20 Reserved Reserved Write 0, ignore on read. 19 :12 POOLSEL Pool Select (bitmap) Field defines to which VMs a packet with the VLAN_Id should be forwarded to. A bit is allocated to each of the 8 VMs, enabling forwarding the packet with the VLAN_Id to multiple VMs. 0x0 RW 11 :00 VLAN_Id Defines a VLAN tag, to which each VM whose bit is set in the POOLSEL field, belongs to. 0x0 RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1375 28.13.1.7 Unicast Table Array--UTA [0:3][0:127] (0xA000 + 4*n [n=0...127]; R/W) There is one register per 32 bits of the Unicast Address Table for a total of 128 registers (the UTA[127:0] designation). Software must mask to the desired bit on reads and supply a 32-bit word on writes. The first bit of the address used to access the table is set according to the RCTL.MO field. Note: All accesses to this table must be 32 bit. The lookup algorithm is the same one used for the MTA table. This table should be zeroed by software before start of work. Table 28-130.Unicast Table Array--UTA [0:3][0:127] (0xA000 + 4*n [n=0...127]; R/W) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :00 Bit Vector 28.13.1.8 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] A000h at Offset Start: 0x4 Offset End: A003h at 0x4 Power Well: GBEAUX Bit Description Sticky Word wide bit vector specifying 32 bits in the unicast destination address filter table. Bit Reset Value Bit Access X R/W Storm Control Control Register--SCCRL [0:3] (0x5DB0; RW) Table 28-131.Storm Control Control Register--SCCRL [0:3] (0x5DB0; RW) (Sheet 1 of 2) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x00000800 Power Well: GBEAUX Bit Range Bit Acronym 31 :18 Reserved Reserved 17 :08 INTERVAL BSC/MSC Time-interval-specification: The interval size for applying Ingress Broadcast or Multicast Storm Control. Interrupt decisions are made at the end of each interval (and most flags are also set at interval end). Setting this field resets the counter. 07 :05 Reserved Reserved Bit Description Sticky Bit Reset Value Bit Access 0x8 RW 04 BIDU BSC Includes Destination Unresolved packets: If bit is set, unicast received packets with no destination pool and sent to the default pool are included in IBSC 0h RW 03 BDICW Drop broadcast packets (excluding flow control and manageability packets) if broadcast threshold is exceeded in current window 0h RW Intel(R) Communications Chipset 89xx Series - Datasheet 1376 Offset Start: 5DB0h Offset End: 5DB3h October 2012 Order Number: 327879-001US 28.0 Table 28-131.Storm Control Control Register--SCCRL [0:3] (0x5DB0; RW) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x00000800 Offset Start: 5DB0h Offset End: 5DB3h Power Well: GBEAUX Bit Reset Value Bit Access Drop broadcast packets (excluding flow control and manageability packets) if broadcast threshold is exceeded in previous window 0h RW MDICW Drop multicast packets (excluding flow control and manageability packets) if multicast threshold is exceeded in current window 0h RW MDIPW Drop multicast packets (excluding flow control and manageability packets) if multicast threshold is exceeded in previous window 0h RW Bit Acronym Bit Description 02 BDIPW 01 00 28.13.1.9 B:0:1+ Index1 Sticky Storm Control Status--SCSTS [0:3] (0x5DB4;RO) Table 28-132.Storm Control Status--SCSTS [0:3] (0x5DB4;RO) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :04 Reserved 03 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] MSCAP Offset Start: 5DB4h Offset End: 5DB7h Power Well: GBEAUX Bit Reset Value Bit Access Multicast storm control active in previous window 0h RO Bit Description Sticky Reserved 02 MSCA Multicast storm control active 0h RO 01 BSCAP Broadcast storm control active in previous window 0h RO 00 BSCA Broadcast storm control active 0h RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1377 28.13.1.10 Broadcast Storm Control Threshold--BSCTRH [0:3] (0x5DB8;RW) Table 28-133.Broadcast Storm Control Threshold--BSCTRH [0:3] (0x5DB8;RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 5DB8h Offset End: 5DBBh Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :19 Reserved Reserved 18 :00 UTRESH Traffic Upper Threshold-size: Represents the upper threshold for broadcast storm control. Sticky Bit Reset Value Bit Access 0x0 RW 28.13.1.11 Multicast Storm Control Threshold--MSCTRH [0:3] (0x5DBC; RW) Table 28-134.Multicast Storm Control Threshold--MSCTRH [0:3] (0x5DBC; RW) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Offset Start: 5DBCh Offset End: 5DBFh Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :19 Reserved Reserved 18 :00 UTRESH Traffic Upper Threshold-size: Represents the upper threshold for multicast storm control. Sticky Bit Reset Value Bit Access 0x0 RW 28.13.1.12 Broadcast Storm Control Current Count - BSCCNT [0:3] (0x5DC0;RO) Table 28-135.Broadcast Storm Control Current Count - BSCCNT [0:3] (0x5DC0;RO) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym 31 :25 Reserved Reserved 24 :00 CCOUNT IBSC Traffic Current Count: Represents the count of broadcast traffic received in the current time interval in units of 64-byte segments. Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 1378 Offset Start: 5DC0h Offset End: 5DC3h Sticky Bit Reset Value Bit Access 0x0 RO October 2012 Order Number: 327879-001US 28.0 28.13.1.13 Multicast Storm Control Current Count--MSCCNT [0:3] (0x5DC4;RO) Table 28-136.Multicast Storm Control Current Count--MSCCNT [0:3] (0x5DC4;RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 5DC4h Offset End: 5DC7h Power Well: GBEAUX Bit Range Bit Acronym Bit Description Sticky 31 :25 Reserved Reserved. 24 :00 CCOUNT IMSC Traffic Current Count: Represents the count of multicast traffic received in the current time interval in units of f 64-byte segments. Bit Reset Value Bit Access 0x0 RO 28.13.1.14 Storm Control Time Counter--SCTC [0:3] (0x5DC8; RO) Table 28-137.Storm Control Time Counter--SCTC [0:3] (0x5DC8; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Bus:Device:Function: Index1 Default: 0x0 Bit Range Bit Acronym 31 :10 Reserved 09 :00 COUNT Offset Start: 5DC8h Offset End: 5DCBh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RO Reserved SC Time Counter: The counter for number of time units elapsed since the end of the last time interval. This register keeps track of the number of time units elapsed since the end of last time interval. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1379 28.13.1.15 Storm Control Basic Interval--SCBI [0:3] (0x5DCC; RW) This register defines the basic interval used as the base for the SCCRL.Interval counting in 10 Mb/s speed. This register is defined in 16 ns clock cycles. The interval in 1000/ 100 is 100 or 10 time smaller respectively. Table 28-138.Storm Control Basic Interval--SCBI [0:3] (0x5DCC; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :25 Reserved 24 :00 BI Offset Start: 5DCCh Offset End: 5DCFh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x5F5E10 RW Reserved Basic interval 28.13.1.16 Virtual Mirror Rule Control--VMRCTL [0:3][0:3] (0x5D80 + 0x4*n [n= 0...3]; RW) This register controls the rules to be applied and the destination port. Table 28-139.Virtual Mirror Rule Control--VMRCTL [0:3][0:3] (0x5D80 + 0x4*n [n= 0...3]; RW) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :11 Reserved 10 :08 MP 07 :04 Reserved Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RW 0h RW Reserved VM Mirror port destination. Packets destined to certain VLAN groups, are mirrored to the queue defined by the MP field, according to the VMRVLAN register. Packets destined to certain VMs, are mirrored to the queue defined by the MP field, according to the VMRVM register. Note: If the VMRCTL.UPME bit is set to 1 all packets received on the port will be forwarded to the queue defined in the MP field. Reserved VLAN mirroring enable. Reflects all the traffic received in a set of given VLANs. bit enables mirroring operation based 03 VLME 02 Reserved 01 UPME Uplink port mirroring enable. Reflects all the traffic received from the network. 0h RW 00 VPME Virtual pool mirroring enable. Reflects all the packets sent to a set of given VMs. 0h RW Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1380 5D80h at Offset Start: 0x4 Offset End: 5D83h at 0x4 October 2012 Order Number: 327879-001US 28.0 28.13.1.17 Virtual Mirror Rule VLAN--VMRVLAN [0:3][0:3] (0x5D90 + 0x4*n [n= 0...3]; RW) This register controls the VLAN ports as listed in the VLVF table taking part in the VLAN mirror rule. Table 28-140.Virtual Mirror Rule VLAN--VMRVLAN [0:3][0:3](0x5D90 + 0x4*n [n= 0...3]; RW) Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 5D90h at Offset Start: 0x4 Offset End: 5D93h at 0x4 Power Well: GBEAUX Bit Acronym Bit Description Sticky VLAN Bitmap listing that defines which of the 32 VLANs defined in the VLVF registers participate in the mirror rule. Packets that have a matching Vlan_ID as defined by the VLVF registers will also be forwarded (mirrored) to the queue defined in the VMRCTL.MP field, if the VMRCTL.VLME bit is set to 1. Bit Reset Value Bit Access 0x0 RW 28.13.1.18 Virtual Mirror Rule VM--VMRVM [0:3][0:3] (0x5DA0 + 0x4*n [n= 0...3]; RW) This register controls the VLAN ports as listed in the VLVF table taking part in the VLAN mirror rule. Table 28-141.Virtual Mirror Rule VM--VMRVM [0:3][0:3] (0x5DA0 + 0x4*n [n= 0...3]; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 Bus:Device:Function: B:0:1+ Index1 VM October 2012 Order Number: 327879-001US 5DA0h at Offset Start: 0x4 Offset End: 5DA3h at 0x4 Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RW Reserved Bitmap listing of VMs participating in the mirror rule. Packets that are forwarded to the queues defined in the VMRVM.VM field, will also be forwarded (mirrored) to the queue defined in the VMRCTL.MP field, if the VMRCTL.VPME bit is set to 1. Intel(R) Communications Chipset 89xx Series - Datasheet 1381 28.14 Power Management Registers The following registers enable control of various DMA power saving features. 28.14.1 Detailed Register Descriptions 28.14.1.1 DMA Receive Power Saving Register--DMARPS [0:3] (0x2500; R/W) Table 28-142.DMA Receive Power Saving Register--DMARPS [0:3] (0x2500; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0000010A Bit Reset Value Bit Access RXLPKBL Receive low power low Kilobyte threshold This value indicates maximum receive buffer full threshold that enables DMA to move to low power mode. Value in 1KB blocks. Maximum allowed value depends on Receive buffer size (IRPBS.RXPbsize). Note: RXBLPL<= RXBLPH value. 0x1 R/W RXLPKBH Receive low power high Kilobyte threshold This value indicates minimum receive buffer full threshold that causes DMA to move to high power mode. Value in 1KB blocks. Maximum allowed value depends on Receive buffer size (IRPBS.RXPbsize) Note: When RXBLPH Value is 0x0, receive Kilobyte threshold is not used to decide DMA move between high and low power modes. 0xA R/W Bit Acronym 31 :16 Reserved Reserved Write 0 ignore on read. 07 :00 28.14.1.2 Offset Start: 2500h Offset End: 2503h Power Well: GBEAUX Bit Range 15 :08 B:0:1+ Index1 Bit Description Sticky DMA Transmit Power Saving Register--DMATPS [0:3] (0x2504; R/W) Table 28-143.DMA Transmit Power Saving Register--DMATPS [0:3] (0x2504; R/W) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 30 :15 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0000010A Power Well: GBEAUX Bit Acronym Bit Description DMAPS_EN DMA Power Saving Enable. Bit enables DMA power saving as a function of utilization. 0 - Disable DMA power saving 1 - Enable DMA power saving Reserved Sticky Bit Reset Value Bit Access 0h R/W Reserved Write 0 ignore on read. Intel(R) Communications Chipset 89xx Series - Datasheet 1382 Offset Start: 2504h Offset End: 2507h October 2012 Order Number: 327879-001US 28.0 Table 28-143.DMA Transmit Power Saving Register--DMATPS [0:3] (0x2504; R/W) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 14 :08 07 06 :00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0000010A Bit Acronym Offset Start: 2504h Offset End: 2507h Power Well: GBEAUX Bit Description Sticky TXLPKBL Transmit low power low Kilobyte threshold This value indicates maximum transmit buffer full threshold that causes DMA to move to high power mode. Value in 1KB blocks. Maximum allowed value depends on Transmit buffer size (ITPBS.TXPbsize). Note: TXBLPL<= TXBLPH value. Reserved Reserved Write 0 ignore on read. TXLPKBHa Transmit low power high Kilobyte threshold This value indicates minimum transmit buffer full threshold that enables DMA to move to low power mode. Value in 1KB blocks. Maximum allowed value depends on Transmit buffer size (ITPBS.TXPbsize). Note: When TXBLPH Value is 0x0, transmit Kilobyte threshold is not used to decide DMA move between high and low power modes. Bit Reset Value Bit Access 0x1 R/W 0xA R/W a. Value of TXLPKBH should be greater than maximum packet size to minimize latency. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1383 28.15 Timer Registers 28.15.1 Detailed Register Descriptions 28.15.1.1 Watchdog Setup--WDSTP [0:3] (0x1040; R/W) Table 28-144.Watchdog Setup--WDSTP [0:3] (0x1040; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x01000000 Offset Start: 1040h Offset End: 1043h Power Well: GBEAUX Bit Reset Value Bit Access WD_Timeout Defines the number of seconds until the watchdog expires. The granularity of this timer is 1 sec. The minimal value allowed for this register when the watchdog mechanism is enabled is two. Setting this field to 1b might cause the watchdog to expire immediately. Note: Only 4 LSB bits loaded from EEPROM. Initial value of 4 MSB bits is 0000b. 0x01 R/W 23 :16 WD_Timer (RWS) Indicates the current value of the timer. Resets to the timeout value each time the Controller functional bit in Software Device Status register is set. If this timer expires, the WD interrupt to the firmware and the WD SDP is asserted. As a result, this timer is stuck at zero until it is re-armed. Note: Writing to this field is only for DFX purposes. X R/W 15 :02 Reserved Bit Range 31 :24 Bit Acronym Bit Description Sticky Reserved 01 WD_Timer_ Load_enable (SC) Enables the load of the watchdog timer by writing to WD_Timer field. If this bit is not set, the WD_Timer field is loaded by the value of WD_Timeout. Note: Writing to this field is only for DFX purposes. 0h R/W 00 WD_Enable Enable Watchdog Timer 0ha R/W a. Value read from the EEPROM. Intel(R) Communications Chipset 89xx Series - Datasheet 1384 October 2012 Order Number: 327879-001US 28.0 28.15.1.2 Watchdog Software Device Status--WDSWSTS [0:3] (0x1044; R/W) Table 28-145.Watchdog Software Device Status--WDSWSTS [0:3] (0x1044; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 1044h Offset End: 1047h Power Well: GBEAUX Bit Reset Value Bit Access 0x0 R/W 01 Setting this bit causes the WD timer to expire immediately. The WD_timer field is set to 0b. It can be Force_WD (SC) used by software in order to indicate some fatal error detected in the software or in the hardware. This bit is self clearing. 0h R/W 00 Dev_Functional Each time this bit is set, the watchdog timer is re-armed. (SC) This bit is self clearing 0h R/W Bit Range Bit Acronym Bit Description 31 :24 Stuck Reason This field can be used by software to indicate to the firmware the reason the Controller is malfunctioning. The encoding of this field is software/firmware dependent. A value of 0 indicates a functional Controller. 23 :02 Reserved 28.15.1.3 Sticky Reserved Free Running Timer--FRTIMER [0:3] (0x1048; RWS) This register reflects the value of a free running timer that can be used for various timeout indications. The register is reset by a PCI reset and/or software reset. Note: Writing to this register is for DFX purposes only. Table 28-146.Free Running Timer--FRTIMER [0:3] (0x1048; RWS) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :20 Seconds 19 :10 Millisecond 09 :00 Microsecond October 2012 Order Number: 327879-001US Offset Start: 1048h Offset End: 104Bh Power Well: GBEAUX Bit Reset Value Bit Access Number of seconds from the timer start (up to 4095 seconds). X RWS Number of milliseconds in the current second. X RWS Number of microseconds in the current millisecond. X RWS Bit Description Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 1385 28.15.1.4 TCP Timer--TCPTIMER [0:3] (0x104C; R/W) Table 28-147.TCP Timer--TCPTIMER [0:3] (0x104C; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Offset Start: 104Ch Offset End: 104Fh Power Well: GBEAUX Bit Reset Value Bit Access TCP Loop When set to 1b, the TCP counter reloads duration each time it reaches zero, and continues down-counting from this point without kick-starting. When set to 0b, the TCP counter stops at a zero value and does not re-start until KickStart is activated. Note: Setting this bit alone is not enough to start the timer activity. The KickStart bit should also be set. 0h R/W 10 TCP Count Finish This bit enables software to trigger a TCP timer interrupt, regardless of the internal state. TCPCountFinish Writing a 1b to this bit triggers an interrupt and resets the (WS) internal counter to its initial value. Down-count does not restart until either KickStart is activated or Loop is set. Writing a 0b has no effect. 0h R/W 09 TCP Count Enable 1b = TCP timer counting enabled. 0b = TCP timer counting disabled. Once enabled, the TCP counter counts from its internal state. If the internal state is equal to 0b, the down-count does not restart until KickStart is activated. If the internal state is not 0b, the down-count continues from internal state. This enables a pause in the counting for debug purpose. 0h R/W Counter Kick-Start KickStart (WS) Writing a 1b to this bit kick-starts the counter down-count from the initial value defined in the Duration field. Writing a 0b has no effect. 0h R/W 0x0 R/W Bit Range Bit Acronym 31 :12 Reserved 11 08 07 :00 Loop TCPCountEn Duration Bit Description Sticky Reserved Duration Duration of the TCP interrupt interval in msec. Intel(R) Communications Chipset 89xx Series - Datasheet 1386 B:0:1+ Index1 October 2012 Order Number: 327879-001US 28.0 28.16 Time Sync Register Descriptions 28.16.1 Detailed Register Descriptions 28.16.1.1 RX Time Sync Control Register--TSYNCRXCTL [0:3] (0xB620;RW) Table 28-148.RX Time Sync Control Register--TSYNCRXCTL [0:3] (0xB620;RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Power Well: GBEAUX Bit Access 0h RW Type Type of packets to timestamp 000b - time stamp L2 (V2) packets only (Sync or Delay_req depends on message type in Section 28.16.1.26 and packets with Pdelay_Req and Pdelay_Resp message ID values) 001b - time stamp L4 (V1) packets only (Sync or Delay_req depends on message type in Section 28.16.1.26) 010b - time stamp V2 (L2 and L4) packets (Sync or Delay_req depends on message type in Section 28.16.1.26 and packets with Pdelay_Req and Pdelay_Resp message ID values) 100b - time stamp all packets. In this mode no locking is done to the timestamp value in the RXSTMPL/H timestamp registers, the RDESC.STATUS.TS bit in the receive descriptor stays 0, while the RDESC.STATUS.TSIP bit in the receive descriptor is always 1 if placing timestamp in receive buffer is enabled (See Section 26.1.10). 101b - Time stamp all packets which have a Message Type bit 3 zero, which means timestamp all event packets. This is applicable for V2 packets only. 011b, 110b and 111b - reserved Note: Field is also used for defining packets that have timestamp captured in receive buffer (See Section 26.1.10). 0x0 RW RXTT(RO/V) Rx timestamp valid (='1' when a valid value for Rx timestamp is captured in the Rx timestamp register, clear by read of Rx timestamp register RXSATRH) 0x0 RW 31 :05 Reserved 00 Offset Start: B620h Offset End: B623h Bit Reset Value Bit Acronym 03 :01 B:0:1+ Index1 Default: 0x0 Bit Range 04 Bus:Device:Function: En October 2012 Order Number: 327879-001US Bit Description Sticky Reserved Enable RX timestamp 0 = time stamping disabled. 1 = time stamping enabled. Intel(R) Communications Chipset 89xx Series - Datasheet 1387 28.16.1.2 RX timestamp Low--RXSTMPL [0:3] (0xB624; RO) Table 28-149.RX timestamp Low--RXSTMPL [0:3] (0xB624; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 RTSL 28.16.1.3 Bus:Device:Function: Offset Start: B624h Offset End: B627h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RO Rx timestamp LSB value Value in 1 nS resolution. RX Timestamp High--RXSTMPH [0:3] (0xB628; RO) Table 28-150.RX Timestamp High--RXSTMPH [0:3] (0xB628; RO) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 RTSH 28.16.1.4 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Offset Start: B628h Offset End: B62Bh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RO Reserved. Rx timestamp MSB value Value in 232 nS resolution. RX Timestamp Attributes Low--RXSATRL[0:3] (0xB62C; RO) Table 28-151.RX Timestamp Attributes Low--RXSATRL[0:3] (0xB62C; RO) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 SourceIDL Sourceuuid low Intel(R) Communications Chipset 89xx Series - Datasheet 1388 Offset Start: B62Ch Offset End: B62Fh Sticky Bit Reset Value Bit Access 0x0 RO October 2012 Order Number: 327879-001US 28.0 28.16.1.5 RX timestamp Attributes High--RXSATRH [0:3] (0xB630; RO) Table 28-152.RX timestamp Attributes High--RXSATRH [0:3] (0xB630; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :16 SequenceID 15 :00 SourceIDH 28.16.1.6 Bus:Device:Function: Offset Start: B630h Offset End: B633h Power Well: GBEAUX Bit Reset Value Bit Access SequenceId 0x0 RO Sourceuuid high 0x0 RO Bit Description Sticky TX Time Sync Control Register--TSYNCTXCTL [0:3] (0xB614; RW) Table 28-153.TX Time Sync Control Register--TSYNCTXCTL [0:3] (0xB614; RW) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :05 Reserved 04 03 :01 00 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] EN Reserved TXTT(RO/V) October 2012 Order Number: 327879-001US Offset Start: B614h Offset End: B617h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0h RW 0h RW Reserved Enable Transmit timestamp 0b = time stamping disabled. 1b = time stamping enabled. Reserved Transmit timestamp valid (equals 1b when a valid value for Tx timestamp is captured in the Tx timestamp register, clear by read of Tx timestamp register TXSTMPH) Intel(R) Communications Chipset 89xx Series - Datasheet 1389 28.16.1.7 TX Timestamp Value Low--TXSTMPL [0:3] (0xB618;RO) Table 28-154.TX Timestamp Value Low--TXSTMPL [0:3] (0xB618;RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 TTSL 28.16.1.8 Bus:Device:Function: Offset Start: B618h Offset End: B61Bh Power Well: GBEAUX Bit Description Sticky Transmit timestamp LSB value Value in 1 nS resolution. Bit Reset Value Bit Access 0x0 RO TX Timestamp Value High--TXSTMPH[0:3] (0xB61C; RO) Table 28-155.TX Timestamp Value High--TXSTMPH[0:3] (0xB61C; RO) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 TTSH 28.16.1.9 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Offset Start: B61Ch Offset End: B61Fh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RO Reserved Write 0 ignore on read. Transmit timestamp MSB value Value in 232 nS resolution. System Time Register Residue--SYSTIMR [0:3] (0xB6F8; RW) Table 28-156.System Time Register Residue--SYSTIMR [0:3] (0xB6F8; RW) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 STR Power Well: GBEAUX Bit Description System time Residue value. Value in 2-32 nS resolution. Intel(R) Communications Chipset 89xx Series - Datasheet 1390 Offset Start: B6F8h Offset End: B6FBh Sticky Bit Reset Value Bit Access 0x0 RW October 2012 Order Number: 327879-001US 28.0 28.16.1.10 System Time Register Low--SYSTIML [0:3] (0xB600; RW) Table 28-157.System Time Register Low--SYSTIML [0:3] (0xB600; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 STL Offset Start: B600h Offset End: B603h Power Well: GBEAUX Bit Description Sticky System time LSB value Value in 1 nS resolution. Bit Reset Value Bit Access 0x0 RW 28.16.1.11 System Time Register High--SYSTIMH [0:3] (0xB604; RW) Table 28-158.System Time Register High--SYSTIMH [0:3] (0xB604; RW) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 STH October 2012 Order Number: 327879-001US Offset Start: B604h Offset End: B607h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RW Reserved Write 0 ignore on read. System time MSB value Value in 232 nS resolution. Intel(R) Communications Chipset 89xx Series - Datasheet 1391 28.16.1.12 Increment Attributes Register--TIMINCA [0:3] (0xB608; RW) 28.16.1.13 Time Adjustment Offset Register Low--TIMADJL [0:3] (0xB60C; RW) Table 28-159.Time Adjustment Offset Register Low--TIMADJL [0:3] (0xB60C; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 TADJL Offset Start: B60Ch Offset End: B60Fh Power Well: GBEAUX Bit Description Sticky Time adjustment value - Low Value in 1 nS resolution. Bit Reset Value Bit Access 0x0 RW 28.16.1.14 Time Adjustment Offset Register High--TIMADJH [0:3] (0xB610;RW) Table 28-160.Time Adjustment Offset Register High--TIMADJH [0:3] (0xB610;RW) Description: View: PCI Size: 32 bit Bit Range 31 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Acronym Power Well: GBEAUX Bit Description Sign Sign (0b="+", 1b ="-") 30 :08 Reserved Reserved. Write 0 ignore on read. 07 :00 TADJH Time adjustment value - High Value in 232 resolution. Intel(R) Communications Chipset 89xx Series - Datasheet 1392 Offset Start: B610h Offset End: B613h Sticky Bit Reset Value Bit Access 0h RW 0x0 RW October 2012 Order Number: 327879-001US 28.0 28.16.1.15 TimeSync Auxiliary Control Register--TSAUXC [0:3] (0xB640; RW) Table 28-161.TimeSync Auxiliary Control Register--TSAUXC [0:3] (0xB640; RW) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range 31 30 :21 20 19 18 17 16 :12 11 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x80000000 Bit Acronym Disable systime Offset Start: B640h Offset End: B643h Power Well: GBEAUX Bit Description Sticky Disable SYSTIM count operation 0b - SYSTIM timer activated 1b - SYSTIM timer disabled. Value of SYSTIMH, SYSTIML and SYSTIMR remains constant. Bit Reset Value Bit Access 1h RW Reserved Reserved PLSNeg1 Generate Negative pulse on Target Time 1 when PLSG1 is 1. 0 - Generate positive pulse on Target Time 1when PLSG1 is 1. 1 - Generate Negative pulse on target time 1 when PLSG1 is 1. Note: If PLSNeg1 = 1, at start the selected SDP pin is set to 1. 0h RW PLSG1 Use Target Time 1 to generate start of pulse and Target Time 0 to generate end of pulse. SDP pin selected to drive pulse or level change is set according to the TSSDP.TS_SDPx_SEL field with a value of 01b and TSSDP.TS_SDPx_EN bit with a value of 1b. 0 - Target Time 1 generates change in SDP level. 1 - Target time 1 generates start of pulse on SDP pin. Note: Pulse or level change is generated when TSAUXC.EN_TT1 is set to 1. 0h RW PLSNeg0 Generate Negative pulse on Target Time 0 when PLSG0 is 1. 0 - Generate positive pulse on Target Time 0 when PLSG0 is 1. 1 - Generate Negative pulse on target time 0 when PLSG0 is 1. Note: If PLSNeg0 = 1, at start the selected SDP pin is set to 1. 0h RW PLSG0 Use Target Time 0 to generate start of pulse and Target Time 1 to generate end of pulse. SDP pin selected to drive pulse or level change is set according to the TSSDP.TS_SDPx_SEL field with a value of 00h and TSSDP.TS_SDPx_EN bit with a value of 1b. 0 - Target Time 0 generates change in SDP level. 1 - Target time 0 generates start of pulse on SDP pin. Note: Pulse or level change is generated when TSAUXC.EN_TT0 is set to 1. 0h RW Auxiliary timestamp taken - cleared when read from auxiliary timestamp 1 occurred 0h RW Enable hardware time stamp 1 Enable Time stamping occurrence of change in SDP pin into the AUXSTMPL1 and AUXSTMPH1 registers. SDP pin (0 to 1) is selected for time stamping, if the SDP pin is selected via the TSSDP.AUX1_SDP_SEL field and the TSSDP.AUX1_TS_SDP_EN bit is set to 1b. 0h RW Auxiliary timestamp taken - cleared when read from auxiliary timestamp 0 occurred 0h RW Reserved AUTT1 10 EN_TS1 09 AUTT0 October 2012 Order Number: 327879-001US Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1393 Table 28-161.TimeSync Auxiliary Control Register--TSAUXC [0:3] (0xB640; RW) (Sheet 2 of 2) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x80000000 Offset Start: B640h Offset End: B643h Power Well: GBEAUX Bit Reset Value Bit Access EN_TS0 Enable hardware time stamp 0 Enable Time stamping occurrence of change in SDP pin into the AUXSTMPL0 and AUXSTMPH0 registers. SDP pin (0 to 1) is selected for time stamping, if the SDP pin is selected via the TSSDP.AUX0_SDP_SEL field and the TSSDP.AUX0_TS_SDP_EN bit is set to 1b. 0h RW 07 ST1 Start Clock 1 Toggle on Target Time 1 Enable Clock 1 toggle only after Target Time 1, that's defined in the TRGTTIML1 and TRGTTIMH1 registers, has passed. The clock output is initially 1 and toggles with a frequency defined in the FREQOUT1 register 0h RW 06 Reserved Reserved 05 EN_CLK1 Enable Configurable Frequency Clock 1 Clock is generated according to frequency defined in the FREQOUT1 register on the SDP pin (0 to 1) that has both: 1. TSSDP.TS_SDPx_SEL field with a value of 11b. 2. TSSDP.TS_SDPx_EN value of 1b. 0h RW 04 ST0 Start Clock 0 Toggle on Target Time 0 Enable Clock 0 toggle only after target time 0, that's defined in the TRGTTIML0 and TRGTTIMH0 registers, has passed. The clock output is initially 0 and toggles with a frequency defined in the FREQOUT0 register. 0h RW 03 Reserved Reserved EN_CLK0 Enable Configurable Frequency Clock 0 Clock is generated according to frequency defined in the FREQOUT0 register on the SDP pin (0 to 1) that has both: 1. TSSDP.TS_SDPx_SEL field with a value of 10b. 2. TSSDP.TS_SDPx_EN value of 1b. 0h RW EN_TT1 Enable target time 1. Enable bit is set by software to 1b, to enable pulse or level change generation as a function of the TSAUXC.PLSG1 and TSAUXC.PLSNeg1 bits. The bit is cleared by hardware when the target time is hit and Pulse or level change occurs. 0h RW EN_TT0 Enable target time 0. Enable bit is set by software to 1b, to enable pulse or level change generation as a function of the TSAUXC.PLSG0 and TSAUXC.PLSNeg0 bits. The bit is cleared by hardware when the target time is hit and Pulse or level change occurs. 0h RW Bit Range 08 02 01 00 Bit Acronym Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 1394 B:0:1+ Index1 Sticky October 2012 Order Number: 327879-001US 28.0 28.16.1.16 Target Time Register 0 Low--TRGTTIML0 [0:3] (0xB644; RW) Table 28-162.Target Time Register 0 Low--TRGTTIML0 [0:3] (0xB644; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 TTL Offset Start: B644h Offset End: B647h Power Well: GBEAUX Bit Description Sticky Target Time 0 LSB register Value in 1 nS resolution. Bit Reset Value Bit Access 0x0 RW 28.16.1.17 Target Time Register 0 High--TRGTTIMH0 [0:3] (0xB648; RW) Table 28-163.Target Time Register 0 High--TRGTTIMH0 [0:3] (0xB648; RW) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 TTH Offset Start: B648h Offset End: B64Bh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RW Reserved Write 0 ignore on read. Target Time 0 MSB register Value in 232 nS resolution. 28.16.1.18 Target Time Register 1 Low--TRGTTIML1 [0:3] (0xB64C; RW) Table 28-164.Target Time Register 1 Low--TRGTTIML1 [0:3] (0xB64C; RW) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 TTL October 2012 Order Number: 327879-001US Offset Start: B64Ch Offset End: B64Fh Power Well: GBEAUX Bit Description Target Time 1 LSB register Value in 1 nS resolution. Sticky Bit Reset Value Bit Access 0x0 RW Intel(R) Communications Chipset 89xx Series - Datasheet 1395 28.16.1.19 Target Time Register 1 High--TRGTTIMH1 [0:3] (0xB650; RW) Table 28-165.Target Time Register 1 High--TRGTTIMH1 [0:3] (0xB650; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 TTH Offset Start: B650h Offset End: B653h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RW Reserved Write 0 ignore on read. Target Time 1 MSB register Value in 232 nS resolution. 28.16.1.20 Frequency Out 0 Control Register--FREQOUT0 [0:3] (0xB654; RW) Table 28-166.Frequency Out 0 Control Register--FREQOUT0 [0:3] (0xB654; RW) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] CHCT Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RW Reserved Write 0 ignore on read. Clock Out Half Cycle Time Half Cycle time of Clock 0 in 8 nS resolution. Clock is generated on SDP pin when TSAUXC.EN_CLK0 is set to 1. SDP pin (0 to 1) that drives Clock 0 is selected according to the TSSDP.TS_SDPx_SEL field that has a value of 10b and a TSSDP.TS_SDPx_EN value of 1b. If TSAUXC.ST0 is set to 1, start of clock toggle is defined by Target Time 0 (TRGTTIML0 and TRGTTIMH0) registers. Notes: 1. Setting this register to zero while using the frequency out feature, is illegal. 2. Clock 0 generation should not be enabled so long as absolute value of SYSTIM error correction using the TRGTTIMH0 and TRGTTIML0 registers is greater then 2 sec. Intel(R) Communications Chipset 89xx Series - Datasheet 1396 Offset Start: B654h Offset End: B657h October 2012 Order Number: 327879-001US 28.0 28.16.1.21 Frequency Out 1 Control Register--FREQOUT1 [0:3] (0xB658; RW) Table 28-167.Frequency Out 1 Control Register--FREQOUT1 [0:3] (0xB658; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 Bus:Device:Function: Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RW Reserved Write 0 ignore on read. Clock Out Half Cycle Time Half Cycle time of Clock 1 in 8 nS resolution. Clock is generated on SDP pin when TSAUXC.EN_CLK1 is set to 1. SDP pin (0 to 1) that drives Clock 1 is selected according to the TSSDP.TS_SDPx_SEL field that has a value of 11b and a TSSDP.TS_SDPx_EN value of 1b. If TSAUXC.ST1 is set to 1, start of clock toggle is defined by Target Time 1 (TRGTTIML1 and TRGTTIMH1) registers. Notes: 1. Setting this register to zero while using the frequency out feature, is illegal. 2. Clock 1 generation should not be enabled so long as absolute value of SYSTIM error correction using the TRGTTIMH1 and TRGTTIML1 registers is greater then 2 sec. CHCT Offset Start: B658h Offset End: B65Bh 28.16.1.22 Auxiliary Time Stamp 0 Register Low--AUXSTMPL0 [0:3] (0xB65C; RO) Table 28-168.Auxiliary Time Stamp 0 Register Low--AUXSTMPL0 [0:3] (0xB65C; RO) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 TSTL October 2012 Order Number: 327879-001US Offset Start: B65Ch Offset End: B65Fh Power Well: GBEAUX Bit Description Auxiliary Time Stamp 0 LSB value Value in 1 nS resolution. Sticky Bit Reset Value Bit Access 0x0 RO Intel(R) Communications Chipset 89xx Series - Datasheet 1397 28.16.1.23 Auxiliary Time Stamp 0 Register High--AUXSTMPH0 [0:3] (0xB660; RO) Reading this register will release the value stored in AUXSTMPH/L0 and will allow time stamping of the next value. Table 28-169.Auxiliary Time Stamp 0 Register High--AUXSTMPH0 [0:3] (0xB660; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 TSTH Offset Start: B660h Offset End: B663h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RO Reserved Write 0 ignore on read. Auxiliary Time Stamp 0 MSB value Value in 232 nS resolution. 28.16.1.24 Auxiliary Time Stamp 1 Register Low--AUXSTMPL1 [0:3] (0xB664; RO) Table 28-170.Auxiliary Time Stamp 1 Register Low--AUXSTMPL1 [0:3] (0xB664; RO) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 TSTL Power Well: GBEAUX Bit Description Auxiliary Time Stamp 1 LSB value Value in 1 nS resolution. Intel(R) Communications Chipset 89xx Series - Datasheet 1398 Offset Start: B664h Offset End: B667h Sticky Bit Reset Value Bit Access 0x0 RO October 2012 Order Number: 327879-001US 28.0 28.16.1.25 Auxiliary Time Stamp 1 Register High--AUXSTMPH1 [0:3] (0xB668; RO) Reading this register will release the value stored in AUXSTMPH/L1 and will allow stamping of the next value. Table 28-171.Auxiliary Time Stamp 1 Register High--AUXSTMPH1 [0:3] (0xB668; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 TSTH Offset Start: B668h Offset End: B66Bh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RO Reserved Write 0 ignore on read. Auxiliary Time Stamp 1 MSB value Value in 232 nS resolution. 28.16.1.26 Time Sync RX Configuration--TSYNCRXCFG [0:3] (0x5F50; R/W) Table 28-172.Time Sync RX Configuration--TSYNCRXCFG [0:3] (0x5F50; R/W) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :16 Reserved Offset Start: 5F50h Offset End: 5F53h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved 15 :08 MSGT V2 Message Type to timestamp 0x0 R/W 07 :00 CTRLT V1 control to timestamp 0x0 R/W October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1399 28.16.1.27 Time Sync SDP Configuration Register--TSSDP [0:3] (0x003C; R/W) This register defines the assignment of SDP pins to the Time sync auxiliary capabilities. Table 28-173.Time Sync SDP Configuration Register--TSSDP [0:3] (0x003C; R/W) Description: View: PCI Size: 32 bit Bit Range 31 :12 11 10 :09 08 07 :06 05 04 :03 02 01 :00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Offset Start: 003Ch Offset End: 003Fh Power Well: GBEAUX Bit Reset Value Bit Access 0h R/W SDP1 allocation to Tsync event - when TS_SDP1_EN is set, these bits select the Tsync event that is routed to SDP1. TS_SDP1_SEL 00b = Target Time 0 is output on SDP1 01b = Target Time 1 is output on SDP1 10b = Freq Clock 0 is output on SDP1 11b = Freq Clock 1 is output on SDP1 0h R/W TS_SDP0_EN 0h R/W 0h R/W When set indicates that one of the SDPs can be used as an AUX1_TS_SDP_ external trigger to Aux timestamp 1 (if this bit is set to EN one of the SDP pins, the corresponding pin should be configured to input mode using SPD_DIR) 0h R/W Select one of the SDPs to serve as the trigger for auxiliary time stamp 1 (in AUXSTMPL1 and AUXSTMPH1 registers) AUX1_SDP_SEL 00b = SDP0 is assigned 01b = SDP1 is assigned 0h R/W When set indicates that one of the SDPs can be used as an AUX0_TS_SDP_ external trigger to Aux timestamp 0 (if this bit is set to EN one of the SDP pins, the corresponding pin should be configured to input mode using SPD_DIR) 0h R/W Select one of the SDPs to serve as the trigger for auxiliary time stamp 0 (AUXSTMPL0 and AUXSTMPH0 registers) AUX0_SDP_SEL 00b = SDP0 is assigned 01b = SDP1 is assigned 0h R/W Bit Acronym Reserved TS_SDP1_EN Bit Description Sticky Reserved When set indicates that SDP1 is assigned to Tsync. When set indicates that SDP0 is assigned to Tsync. SDP0 allocation to Tsync event - when TS_SDP0_EN is set, these bits select the Tsync event that is routed to SDP0. TS_SDP0_SEL 00b = Target Time 0 is output on SDP0 01b = Target Time 1 is output on SDP0 10b = Freq Clock 0 is output on SDP0 11b = Freq Clock 1 is output on SDP0 Intel(R) Communications Chipset 89xx Series - Datasheet 1400 B:0:1+ Index1 October 2012 Order Number: 327879-001US 28.0 28.16.2 Time Sync Interrupt Registers 28.16.2.1 Time Sync Interrupt Cause Register--TSICR [0:3] (0xB66C; RC/W1C) Value of register is always read as 0x0. Once ICR.Time_Sync is set, internal value of Register should be cleared by write 1 to all bits or cleared by read to enable reception of an additional ICR.Time_Sync interrupt. Table 28-174.Time Sync Interrupt Cause Register--TSICR [0:3] (0xB66C; RC/W1C) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved Offset Start: B66Ch Offset End: B66Fh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0h RC/W1C Reserved 07 TADJ Time Adjust 0 done Set when Time Adjust to clock out 0 or 1 completed 06 AUTT1 Auxiliary timestamp 1 taken. Set when new timestamp is loaded into AUXSTMP 1 (auxiliary timestamp 1) register. 0h RC/W1C 05 AUTT0 Auxiliary timestamp 0 taken. Set when new timestamp is loaded into AUXSTMP 0 (auxiliary timestamp 0) register. 0h RC/W1C 04 TT1 Target time 1 trigger. Set when Target Time 1 (TRGTTIML/H1) trigger occurs. 0h RC/W1C 03 TT0 Target time 0 trigger. Set when Target Time 0 (TRGTTIML/H0) trigger occurs. 0h RC/W1C 02 RXTS Receive Time Stamp Set when new timestamp is loaded into RXSTMP register 0h RC/W1C 01 TXTS Transmit Time Stamp Set when new timestamp is loaded into TXSTMP register 0h RC/W1C 00 SYS WARP SYSTIM Warp around. Set when SYSTIM Warp Around occurs. Warp around occurrence can be used by Software to update software time sync time. 0h RC/W1C October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1401 28.16.2.2 Time Sync Interrupt Mask Register--TSIM [0:3] (0xB674; RW) Table 28-175.Time Sync Interrupt Mask Register--TSIM [0:3] (0xB674; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved Offset Start: B674h Offset End: B677h Power Well: GBEAUX Bit Reset Value Bit Access Time Adjust 0 done mask 0 - No Interrupt generated when TSICR.TADJ is set. 1 - Interrupt generated when TSICR.TADJ is set. 0h RW Bit Description Sticky Reserved. Write 0, ignore on read. 07 TADJ 06 AUTT1 Auxiliary timestamp 1 taken Mask 0 - No Interrupt generated when TSICR.AUTT1 is set. 1 - Interrupt generated when TSICR.AUTT1 is set. 0h RW 05 AUTT0 Auxiliary timestamp 0 taken Mask 0 - No Interrupt generated when TSICR.AUTT0 is set. 1 - Interrupt generated when TSICR.AUTT0 is set. 0h RW 04 TT1 Target time 1 Trigger Mask 0 - No Interrupt generated when TSICR.TT1 is set. 1 - Interrupt generated when TSICR.TT1 is set. 0h RW 03 TT0 Target time 0 Trigger Mask 0 - No Interrupt generated when TSICR.TT0 is set. 1 - Interrupt generated when TSICR.TT0 is set. 0h RW 02 RXTS Receive Time Stamp Mask 0 - No Interrupt generated when TSICR.RXTS is set. 1 - Interrupt generated when TSICR.RXTS is set. 0h RW 01 TXTS Reserved. Write 0, ignore on read. 00 SYS WARP 0h RW Time Adjust 0 done mask 0 - No Interrupt generated when TSICR.TADJ is set. 1 - Interrupt generated when TSICR.TADJ is set. Intel(R) Communications Chipset 89xx Series - Datasheet 1402 B:0:1+ Index1 October 2012 Order Number: 327879-001US 28.0 28.16.2.3 Time Sync Interrupt Set Register--TSIS [0:3] (0xB670; WO) TSIS register is Write Only. Writing 1 to a bit sets respective interrupt bit in TSICR register. Write operation causes a single interrupt event and bit is cleared internally. Table 28-176.Time Sync Interrupt Set Register--TSIS [0:3] (0xB670; WO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved Offset Start: B670h Offset End: B673h Power Well: GBEAUX Bit Reset Value Bit Access Set Time Adjust 0 done Interrupt 0 - No TSICR.TADJ interrupt set. 1 - TSICR.TADJ interrupt set. 0h WO Bit Description Sticky Reserved. Write 0, ignore on read. 07 TADJ 06 AUTT1 Set Auxiliary Timestamp 1 Taken Interrupt 0 - No TSICR.AUTT1 Interrupt set. 1 - TSICR.AUTT1 Interrupt set. 0h WO 05 AUTT0 Set Auxiliary Timestamp 0 Taken Interrupt 0 - No TSICR.AUTT0 Interrupt set. 1 - TSICR.AUTT0 Interrupt set. 0h WO 04 TT1 Set Target Time 1 Trigger Interrupt 0 - No TSICR.TT1 Interrupt set. 1 - TSICR.TT1 Interrupt set. 0h WO 03 TT0 Set Target Time 0 Trigger Interrupt 0 - No TSICR.TT0 Interrupt set. 1 - TSICR.TT0 Interrupt set. 0h WO 02 RXTS Set Receive Time Stamp Interrupt 0 - No TSICR.RXTS Interrupt set. 1 -TSICR.RXTS Interrupt set. 0h WO 01 TXTS Set Transmit Time Stamp Interrupt 0 - No TSICR.TXTS Interrupt set. 1 -TSICR.TXTS interrupt set. 0h WO 00 SYS WARP Set SYSTIM Warp Around Interrupt 0 - No TSICR.SWARP Interrupt set. 1 -TSICR.SWARP interrupt set. 0h WO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1403 28.17 PCS Registers These registers are used to configure the SerDes, SGMII and 1000BASE-KX PCS logic. Usage of these registers is described in Section 22.5.4.1 & Section 22.5.4.3. 28.17.1 Detailed Register Descriptions 28.17.1.1 PCS Configuration--PCS_CFG [0:3] (0x4200; R/W) Table 28-177.PCS Configuration--PCS_CFG [0:3] (0x4200; R/W) Description: View: PCI Size: 32 bit Bit Range B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x00000008 Bit Acronym Power Well: GBEAUX Bit Description 31 SRESET Soft Reset Setting this bit puts all modules within the MAC in reset except the Host Interface. The Host Interface is reset via HRST. This bit is NOT self clearing; GMAC is in a reset state until this bit is cleared. 30 PCS Isolate PCS Isolate Setting this bit isolates the PCS logic from the MAC's data path. PCS control codes are still sent and received. 29 :04 03 02 :00 28.17.1.2 Reserved PCS Enable Reserved Offset Start: 4200h Offset End: 4203h Sticky Bit Reset Value Bit Access 0h R/W 0h R/W 1h R/W Reserved PCS Enable Enables the PCS logic of the MAC. Should be set in SGMII, 1000BASE-KX and SerDes mode for normal operation. Clearing this bit disables RX/TX of both data and control codes. Use this to force link down at the far end. Reserved PCS Link Control--PCS_LCTL [0:3] (0x4208; RW) Table 28-178.PCS Link Control--PCS_LCTL [0:3] (0x4208; RW) (Sheet 1 of 3) Description: View: PCI Size: 32 bit Default: 0x0204000C Bit Range Bit Acronym 31 :26 Reserved 25 24 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Power Well: GBEAUX Bit Description Fast Link Timer AN timer is reduced if this bit is set. Intel(R) Communications Chipset 89xx Series - Datasheet 1404 Sticky Bit Reset Value Bit Access 1h RW 0h RW Reserved Link OK Fix Enable LINK OK FIX EN Control for enabling/disabling LinkOK/SyncOK fix. Should be set for normal operation. FAST LINK TIMER Offset Start: 4208h Offset End: 420Bh October 2012 Order Number: 327879-001US 28.0 Table 28-178.PCS Link Control--PCS_LCTL [0:3] (0x4208; RW) (Sheet 2 of 3) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0204000C Offset Start: 4208h Offset End: 420Bh Power Well: GBEAUX Bit Reset Value Bit Access Reserved 0h RW 20 AN SGMII TRIGGER AN SGMII Trigger If this bit is cleared, then AN is not automatically triggered in SGMII mode even if SYNC fails. AN is triggered only in response to PHY messages or by a manual setting like changing the AN Enable/Restart bits. 0h RW 19 AN SGMII BYPASS AN SGMII Bypass If this bit is set, then IDLE detect state is bypassed during AN in SGMII mode. This reduces the acknowledge time in SGMII mode. 0h RW AN Timeout Enable This bit enables the AN Timeout feature. During AN, if the link partner does not respond with AN pages, but AN TIMEOUT EN continues to send good IDLE symbols, then LINK UP is assumed. (This enables LINK UP condition when link partner is not AN-capable and does not affect otherwise). This bit should not be set in SGMII mode. 1h RW Bit Range Bit Acronym 23 :21 Reserved 18 17 16 15 :08 Bit Description Sticky AN RESTART (SC) AN Restart Used to reset/restart the link auto-negotiation process when using SerDes mode. Setting this bit restarts the Auto-negotiation process. This bit is self clearing. 0h RW AN_ENABLE AN Enable Setting this bit enables the AN process in SerDes operating mode. Note: When link-up is forced (CTRL.SLU=1) the AN_ENABLE bit should be 0. 0ha RW 0 RW Flow control mode is set according to the AN process by following Table 37-4 in the IEEE 802.3 spec. Flow control is set according to FC_TX_EN / FC_RX_EN bits in CTRL register. 0h RW Reserved Reserved 07 Force Flow Control 06 LINK LATCH LOW (LL) Link Latch Low Enable If this bit is set, then link OK going LOW (negative edge) is latched until a processor read. Afterwards, link OK is continuously updated until link OK again goes LOW (negative edge is seen). 0h RW Force Link Force Link If this bit is set, then the internal LINK_OK variable is forced to forced link value (bit 0 of this register). Otherwise, LINK_OK is decided by internal AN/SYNC state machines. 0h RW FSD Force Speed and Duplex If this bit is set, then speed and duplex mode is forced to forced speed value and forced duplex value, respectively. Otherwise, speed and duplex mode are decided by internal AN/SYNC state machines. 0h RW FDV Forced Duplex Value This bit denotes the duplex mode when force speed and duplex is set. This value is also used when AN is disabled or when in SerDes mode. 1b = Full duplex (SerDes/SGMII/1000BASE-KX). 0b = Half duplex (SGMII). 1h RW 05 04 03 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1405 Table 28-178.PCS Link Control--PCS_LCTL [0:3] (0x4208; RW) (Sheet 3 of 3) Description: View: PCI Size: 32 bit Bit Range 02 :01 00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0204000C Offset Start: 4208h Offset End: 420Bh Power Well: GBEAUX Bit Reset Value Bit Access FSV Forced Speed Value These bits denote the speed when force speed and duplex is set. This value is also used when AN is disabled or when in SerDes mode. 00b = 10 Mb/s (SGMII). 01b = 100 Mb/s (SGMII). 10b = 1000 Mb/s (SerDes/SGMII/1000BASE-KX). 11b = Reserved. 2h RW FLV Forced Link Value This bit denotes the link condition when force link is set. 0b = Forced link down. 1b = Forced link up. 0h RW Bit Acronym Bit Description Sticky a. Read from EEPROM word 0x0F, bit 11. 28.17.1.3 PCS Link Status--PCS_LSTS [0:3] (0x420C; RO) Table 28-179.PCS Link Status--PCS_LSTS [0:3] (0x420C; RO) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Default: 0x0000000C Bit Range Bit Acronym 31 :21 Reserved 20 AN ERROR (RWS) 19 AN REMOTE FAULT 18 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Power Well: GBEAUX Bit Reset Value Bit Access AN Error This bit indicates that a AN error condition was detected in SerDes/SGMII mode. Valid after the AN Complete bit is set. AN error conditions: SerDes mode: Both nodes not Full Duplex SGMII mode: PHY is set to 1000 Mb/s Half Duplex mode. Software can also force a AN error condition by writing to this bit (or can clear a existing AN error condition). This bit is cleared at the start of AN. 0h RO AN Remote Fault This bit indicates that an AN page was received with a remote fault indication during an AN process. This bit cleared on reads. 0h RO 0h RO Bit Description Sticky Reserved AN Timed Out AN TIMEDOUT This bit indicates an AN process was timed out. Valid after the AN Complete bit is set. Intel(R) Communications Chipset 89xx Series - Datasheet 1406 Offset Start: 420Ch Offset End: 420Fh October 2012 Order Number: 327879-001US 28.0 Table 28-179.PCS Link Status--PCS_LSTS [0:3] (0x420C; RO) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range 17 16 15 :05 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0000000C Bit Acronym AN PAGE RECEIVED Offset Start: 420Ch Offset End: 420Fh Power Well: GBEAUX Bit Description Sticky AN Page Received This bit indicates that a link partner's page was received during an AN process. This bit is cleared on reads. AN Complete This bit indicates that the AN process has completed.This AN COMPLETE bit is set when the AN process reached the Link OK state. It is reset upon AN restart or reset. It is set even if the AN negotiation failed and no common capabilities where found. Bit Reset Value Bit Access 0h RO 0h RO Reserved Reserved 04 SYNC OK Sync OK This bit indicates the current value of Sync OK from the PCS Sync state machine. 0h RO 03 DUPLEX Duplex This bit denotes the current duplex mode. 1b = Full duplex. 0b = Half duplex. 1h RO Speed This bit denotes the current operating Speed. 00b = 10 Mb/s. 01b = 100 Mb/s. 10b = 1000 Mb/s. 11b = Reserved. 2h RO Link OK This bit denotes the current link ok status. 0b = Link down. 1b = Link up/OK. 0h RO 02 :01 00 SPEED LINK OK October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1407 28.17.1.4 AN Advertisement--PCS_ANADV [0:3] (0x4218; R/W) Table 28-180.AN Advertisement--PCS_ANADV [0:3] (0x4218; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x00000020 Offset Start: 4218h Offset End: 421Bh Power Well: GBEAUX Bit Reset Value Bit Access 0h R/W 0h R/W ASM Local PAUSE Capabilities the Controller'sPAUSE capability is encoded in this field. 00b = No PAUSE. 01b = Symmetric PAUSE. 10b = Asymmetric PAUSE to link partner. 11b = Both symmetric and asymmetric PAUSE to the Controller. 0ha R/W 06 HDCAP (RO) Half Duplex This bit indicates that the Controller is capable of half duplex operation. This bit is tied to 0b because the Controller does not support half duplex in SerDes mode. 0h R/W 05 FDCAP Full Duplex Setting this bit indicates that the Controller is capable of full duplex operation. This bit should be set to 1b for normal operation. 1h R/W Bit Range Bit Acronym 31 :16 Reserved 15 NEXTP 14 Reserved 13 :12 RFLT 11 :09 Reserved 08 :07 04 :00 Reserved Bit Description Sticky Reserved Next Page Capable The Controller asserts this bit to request a next page transmission. The Controller clears this bit when no subsequent next pages are requested. Reserved Remote Fault The Controller's remote fault condition is encoded in this field. The Controller might indicate a fault by setting a non-zero remote fault encoding and re-negotiating. 00b = No error, link OK. 01b = Link failure. 10b = Offline. 11b = Auto-negotiation error. Reserved Reserved a. Loaded from EEPROM word 0x0F, bits 13:12. Intel(R) Communications Chipset 89xx Series - Datasheet 1408 October 2012 Order Number: 327879-001US 28.0 28.17.1.5 Link Partner Ability--PCS_LPAB [0:3] (0x421C; RO) Table 28-181.Link Partner Ability--PCS_LPAB [0:3] (0x421C; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] 0h RO 0h RO 0h RO 0h RO LPASM LP ASMDR/LP PAUSE (SerDes) The link partner's PAUSE capability is encoded in this field. 00b = No PAUSE. 01b = Symmetric PAUSE. 10b = Asymmetric PAUSE to link partner. 11b = Both symmetric and asymmetric PAUSE to the Controller. These bits are reserved while in SGMII mode. 0h RO LPHD LP Half Duplex (SerDes) When set to 1b, the link partner is capable of half duplex operation. When set to 0b, the link partner is not capable of half duplex mode. This bit is reserved while in SGMII mode. 0h RO LPFD LP Full Duplex (SerDes) When set to 1b, the link partner is capable of full duplex operation. When set to 0b, the link partner is not capable of full duplex mode. This bit is reserved while in SGMII mode. 0h RO Reserved 15 LPNEXTP 14 ACK 13 :12 PRF 11 :10 SGMII SPEED 05 04 :00 Power Well: GBEAUX Bit Access 31 :16 06 Offset Start: 421Ch Offset End: 421Fh Bit Reset Value Bit Acronym 08 :07 B:0:1+ Index1 Default: 0x0 Bit Range 09 Bus:Device:Function: Reserved Reserved October 2012 Order Number: 327879-001US Bit Description Sticky Reserved LP Next Page Capable (SerDes) The link partner asserts this bit to indicate its ability to accept next pages. Acknowledge (SerDes) The link partner has acknowledge page reception. SGMII: Reserved. LP Remote Fault (SerDes) The link partner's remote fault condition is encoded in this field. 00b = No error, link ok. 10b = Link failure. 01b = Offline. 11b = Auto-negotiation error. SGMII [13]: Reserved SerDes: reserved. Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1409 28.17.1.6 Next Page Transmit--PCS_NPTX [0:3] (0x4220; RW) Table 28-182.Next Page Transmit--PCS_NPTX [0:3] (0x4220; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Bit Range Bit Acronym 31 :16 Reserved 15 NXTPG 14 Reserved Offset Start: 4220h Offset End: 4223h Power Well: GBEAUX Bit Reset Value Bit Access 0h RW Message/Unformatted Page This bit is used to differentiate a Message Page from an Unformatted Page. The encoding is: 0b = Unformatted page. 1b = Message page. 0h RW Bit Description Sticky Reserved Next Page Used to indicate whether or not this is the last Next Page to be transmitted. The encoding is: 0b = Last page. 1b = Additional Next Pages follow. Reserved 13 PGTYPE 12 ACK2 Acknowledge 2 Used to indicate that a device has successfully received its Link Partners' Link Code Word. 0h RW TOGGLE Toggle This bit is used to ensure synchronization with the Link Partner during Next Page exchange. This bit always takes the opposite value of the Toggle bit in the previously exchanged Link Code Word. The initial value of the Toggle bit in the first Next Page transmitted is the inverse of bit 11 in the base Link Code Word and, therefore, can assume a value of 0b or 1b. The Toggle bit is set as follows: 0b = Previous value of the transmitted Link Code Word when 1b 1b = Previous value of the transmitted Link Code Word when 0b. 0h RW CODE Message/Unformatted Code Field The Message Field is an 11-bit wide field that encodes 2048 possible messages. Unformatted Code Field is an 11bit wide field that might contain an arbitrary value. 0x0 RW 11 10 :00 Intel(R) Communications Chipset 89xx Series - Datasheet 1410 B:0:1+ Index1 October 2012 Order Number: 327879-001US 28.0 28.17.1.7 Link Partner Ability Next Page--PCS_LPABNP [0:3] (0x4224; RO) Table 28-183.Link Partner Ability Next Page--PCS_LPABNP [0:3] (0x4224; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :16 Reserved Offset Start: 4224h Offset End: 4227h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0 RO Reserved 15 NXTPG Next Page Used to indicate whether or not this is the last Next Page to be transmitted. The encoding is: 0b = Last page. 1b = Additional Next Pages follow. 14 ACK Acknowledge The Link Partner has acknowledged Next Page reception. 0 RO Message Page This bit is used to differentiate a Message Page from an Unformatted Page. The encoding is: 0b = Unformatted page. 1b = Message page. 0 RO 13 MSGPG 12 ACK2 Acknowledge 2 Used to indicate that a device has successfully received its Link Partners' Link Code Word. 0 RO TOGGLE Toggle This bit is used to ensure synchronization with the Link Partner during Next Page exchange. This bit always takes the opposite value of the Toggle bit in the previously exchanged Link Code Word. The initial value of the Toggle bit in the first Next Page transmitted is the inverse of bit 11 in the base Link Code Word and, therefore, can assume a value of 0b or 1b. The Toggle bit is set as follows: 0b = Previous value of the transmitted Link Code Word when 1b 1b = Previous value of the transmitted Link Code Word when 0b. 0 RO CODE Message/Unformatted Code Field The Message Field is an 11-bit wide field that encodes 2048 possible messages. Unformatted Code Field is an 11bit wide field that might contain an arbitrary value. 0 RO 11 10 :00 28.17.1.8 SFP I2C Command--I2CCMD [0:3] (0x1028; R/W) This register is used by software to read or write to the configuration registers in an SFP module when the CTRL_EXT.I2C Enabled bit is set to 1. Note: According to the SFP specification, only reads are allowed from this interface; however, SFP vendors also provide a writable register through this interface (for example, PHY registers). As a result, write capability is also supported. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1411 28.17.1.9 SFP I2C Parameters--I2CPARAMS [0:3] (0x102C; R/W) This register is used to set the parameters for the I2C access to the SFP module and to allow bit banging access to the I2C interface. Table 28-184.SFP I2C Parameters--I2CPARAMS [0:3] (0x102C; R/W) Description: View: PCI Size: 32 bit Bit Range 31 30 :16 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x00000046 Bit Acronym I2C Data order Reserved Offset Start: 0102Ch Offset End: 0102Fh Power Well: GBEAUX Bit Description Sticky 0b - I2CCMD.DATA field read in Byte order. 1b - I2CCMD.DATA field read in Word order. Bit Reset Value Bit Access 0x0 R/W Reserved R/W I2C 15 clk_stretch_dis 0b - Enable slave clock stretching support in access. 1b - Disable clock stretching support in I2C access. 0h R/W 14 CLK_IN (RO) I2C Clock In Value Reflects the value of the I2C_CLK pad. While in bit bang mode when the CLK_OE_N field is zero, this field reflects the value set in the CLK_OUT field. X R/W I2C Clock Output Enable While in bit bang mode, controls the direction of the I2C_CLK pad of this port. 0b = Pad is output. 1b = Pad is input. 0h R/W X R/W I2C_DATA_OE_N While in bit bang mode, controls the direction of the I2C_DATA pad of this port. 0b = Pad is output. 1b = Pad is input. 0h R/W 0h R/W 13 12 11 CLK_OE_N I2C_DATA_IN DATA_IN (RO) Reflects the value of the I2C_DATA pad. While in bit bang mode when the DATA_OE_N field is zero, this field reflects the value set in the DATA_OUT field. DATA_OE_N 10 DATA_OUT I2C_DATA While in bit bang mode and when the DATA_OE_N field is zero, controls the value driven on the I2C_DATA pad of this port. 09 CLK I2C Clock While in bit bang mode, controls the value driven on the I2C_CLK pad of this port. 0h R/W I2CBB_EN I2C Bit Bang Enable If set, the I2C_CLK and I2C_DATA lines are controlled via the CLK, DATA and DATA_OE_N fields of this register. Otherwise, they are controlled by the hardware machine activated via the I2CCMD or MDIC registers. 0h R/W 07 :05 Read Time Read Time Defines the delay between a read access and the next access. The value is in microseconds. A value of Zero is not valid 2h R/W 04 :00 Write Time Write Time Defines the delay between a write access and the next access. The value is in milliseconds. A value of zero is not valid. 06h R/W 08 Intel(R) Communications Chipset 89xx Series - Datasheet 1412 B:0:1+ Index1 October 2012 Order Number: 327879-001US 28.0 28.18 Statistics Register Descriptions All Statistics registers reset when read. In addition, they stick at 0xFFFF_FFFF when the maximum value is reached. For the receive statistics it should be noted that a packet is indicated as received if it passes the Controller's filters and is placed into the packet buffer memory. A packet does not have to be transferred to host memory in order to be counted as received. Due to divergent paths between interrupt-generation and logging of relevant statistics counts, it might be possible to generate an interrupt to the system for a noteworthy event prior to the associated statistics count actually being incremented. This is extremely unlikely due to expected delays associated with the system interruptcollection and ISR delay, but might be observed as an interrupt for which statistics values do not quite make sense. Hardware guarantees that any event noteworthy of inclusion in a statistics count is reflected in the appropriate count within 1 s; a small time-delay prior to a read of statistics might be necessary to avoid the potential for receiving an interrupt and observing an inconsistent statistics count as part of the ISR. 28.18.1 Detailed Register Descriptions 28.18.1.1 CRC Error Count--CRCERRS [0:3] (0x4000; RC) Counts the number of receive packets with CRC errors. In order for a packet to be counted in this register, it must pass address filtering and must be 64 bytes or greater (from <Destination Address> through <CRC>, inclusively) in length. If receives are not enabled, then this register does not increment. Table 28-185.CRC Error Count--CRCERRS [0:3] (0x4000; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 CEC October 2012 Order Number: 327879-001US Power Well: GBEAUX Bit Description CRC error count Offset Start: 04000h Offset End: 04003h Sticky Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1413 28.18.1.2 Alignment Error Count--ALGNERRC [0:3] (0x4004; RC) Counts the number of receive packets with alignment errors (the packet is not an integer number of bytes in length). In order for a packet to be counted in this register, it must pass address filtering and must be 64 bytes or greater (from <Destination Address> through <CRC>, inclusive) in length. If receives are not enabled, then this register does not increment. This register is valid only in SGMII mode during 10/100 Mb/s operation. Table 28-186.Alignment Error Count--ALGNERRC [0:3] (0x4004; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 AEC 28.18.1.3 Bus:Device:Function: Offset Start: 04004h Offset End: 04007h Power Well: GBEAUX Bit Description Sticky Alignment error count Bit Reset Value Bit Access 0x0 RC Symbol Error Count--SYMERRS [0:3] (0x4008; RC) Counts the number of symbol errors between reads. The count increases for every bad symbol received, whether or not a packet is currently being received and whether or not the link is up. When working in SerDes/SGMII/1000BASE-KX mode these statistics can be read from the SCVPC register. Table 28-187.Symbol Error Count--SYMERRS [0:3] (0x4008; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 SYMERRS Power Well: GBEAUX Bit Description Symbol Error Count Intel(R) Communications Chipset 89xx Series - Datasheet 1414 Offset Start: 04008h Offset End: 0400Bh Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.18.1.4 Missed Packets Count--MPC [0:3] (0x4010; RC) Counts the number of missed packets. Packets are missed when the receive FIFO has insufficient space to store the incoming packet. This can be caused because of too few buffers allocated, or because there is insufficient bandwidth on the PCI bus. Events setting this counter causes ICR.Rx Miss, the Receiver Overrun Interrupt, to be set. This register does not increment if receives are not enabled. These packets are also counted in the Total Packets Received register as well as in Total Octets Received. Table 28-188.Missed Packets Count--MPC [0:3] (0x4010; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 MPC 28.18.1.5 Bus:Device:Function: Offset Start: 04010h Offset End: 04013h Power Well: GBEAUX Bit Description Sticky Missed Packets Count Bit Reset Value Bit Access 0x0 RC Single Collision Count--SCC [0:3] (0x4014; RC) This register counts the number of times that a successfully transmitted packet encountered a single collision. This register only increments if transmits are enabled (TCTL.EN is set) and the Controller is in half-duplex mode. Table 28-189.Single Collision Count--SCC [0:3] (0x4014; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 SCC Number of times a transmit encountered a single collision. October 2012 Order Number: 327879-001US Offset Start: 04014h Offset End: 04017h Sticky Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1415 28.18.1.6 Excessive Collisions Count--ECOL [0:3] (0x4018; RC) When 16 or more collisions have occurred on a packet, this register increments, regardless of the value of collision threshold. If collision threshold is set below 16, this counter won't increment. This register only increments if transmits are enabled (TCTL.EN is set) and the Controller is in half-duplex mode. Table 28-190.Excessive Collisions Count--ECOL [0:3] (0x4018; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 ECC 28.18.1.7 Bus:Device:Function: Offset Start: 04018h Offset End: 0401Bh Power Well: GBEAUX Bit Description Sticky Number of packets with more than 16 collisions. Bit Reset Value Bit Access 0x0 RC Multiple Collision Count--MCC [0:3] (0x401C; RC) This register counts the number of times that a transmit encountered more than one collision but less than 16. This register only increments if transmits are enabled (TCTL.EN is set) and the Controller is in half-duplex mode. Table 28-191.Multiple Collision Count--MCC [0:3] (0x401C; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 MCC Power Well: GBEAUX Bit Description Number of times a successful transmit encountered multiple collisions. Intel(R) Communications Chipset 89xx Series - Datasheet 1416 Offset Start: 0401Ch Offset End: 0401Fh Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.18.1.8 Late Collisions Count--LATECOL [0:3] (0x4020; RC) Late collisions are collisions that occur after one slot time. This register only increments if transmits are enabled (TCTL.EN is set) and the Controller is in half-duplex mode. Table 28-192.Late Collisions Count--LATECOL [0:3] (0x4020; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 LCC 28.18.1.9 Bus:Device:Function: Offset Start: 04020h Offset End: 04023h Power Well: GBEAUX Bit Description Sticky Number of packets with late collisions. Bit Reset Value Bit Access 0x0 RC Collision Count--COLC [0:3] (0x4028; RC) This register counts the total number of collisions seen by the transmitter. This register only increments if transmits are enabled (TCTL.EN is set) and the Controller is in halfduplex mode. This register applies to clear as well as secure traffic. Table 28-193.Collision Count--COLC [0:3] (0x4028; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Bus:Device:Function: Index1 Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 CCC Total number of collisions experienced by the transmitter. October 2012 Order Number: 327879-001US Offset Start: 04028h Offset End: 0402Bh Sticky Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1417 28.18.1.10 Defer Count--DC [0:3] (0x4030; RC) This register counts defer events. A defer event occurs when the transmitter cannot immediately send a packet due to the medium being busy either because another device is transmitting, the IPG timer has not expired, half-duplex deferral events, reception of XOFF frames, or the link is not up. This register only increments if transmits are enabled (TCTL.EN is set). This counter does not increment for streaming transmits that are deferred due to TX IPG. Table 28-194.Defer Count--DC [0:3] (0x4030; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 CDC Offset Start: 04030h Offset End: 04033h Power Well: GBEAUX Bit Description Sticky Number of defer events. Bit Reset Value Bit Access 0x0 RC 28.18.1.11 Host Transmit Discarded Packets by MAC Count--HTDPMC [0:3] (0x403C; RC) Table 28-195.Host Transmit Discarded Packets by MAC Count--HTDPMC [0:3] (0x403C; RC) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 HTDPMC Offset Start: 0403Ch Offset End: 0403Fh Power Well: GBEAUX Bit Description Number of packets sent by the host but discarded by the MAC Sticky Bit Reset Value Bit Access 0x0 RC This register counts the number of packets sent by the host (and not the manageability engine) that are dropped by the MAC. This can include packets dropped because of excessive collisions or link fail events. Intel(R) Communications Chipset 89xx Series - Datasheet 1418 October 2012 Order Number: 327879-001US 28.0 28.18.1.12 Receive Length Error Count--RLEC [0:3] (0x4040; RC) This register counts receive length error events. A length error occurs if an incoming packet passes the filter criteria but is undersized or oversized. Packets less than 64 bytes are undersized. Packets over 1518/1522/1526 bytes (according to the number of VLAN tags present) are oversized if Long Packet Enable (LPE) is 0b. If LPE is 1b, then an incoming, packet is considered oversized if it exceeds the size defined in RLPML.RLPML field. If receives are not enabled, this register does not increment. These lengths are based on bytes in the received packet from <Destination Address> through <CRC>, inclusive. Note: Runt packets smaller than 25 bytes may not be counted by this counter. Table 28-196.Receive Length Error Count--RLEC [0:3] (0x4040; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 RLEC Offset Start: 04040h Offset End: 04043h Power Well: GBEAUX Bit Description Sticky Number of packets with receive length errors. Bit Reset Value Bit Access 0x0 RC 28.18.1.13 XON Received Count--XONRXC [0:3] (0x4048; RC) This register counts the number of valid XON packets received. XON packets can use the global address, or the station address. This register only increments if receives are enabled (RCTL.RXEN is set). Table 28-197.XON Received Count--XONRXC [0:3] (0x4048; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 XONRXC October 2012 Order Number: 327879-001US Offset Start: 04048h Offset End: 0404Bh Power Well: GBEAUX Bit Description Number of XON packets received. Sticky Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1419 28.18.1.14 XON Transmitted Count--XONTXC [0:3] (0x404C; RC) This register counts the number of XON packets transmitted. These can be either due to a full queue or due to software initiated action (using TCTL.SWXOFF). This register only increments if transmits are enabled (TCTL.EN is set). Table 28-198.XON Transmitted Count--XONTXC [0:3] (0x404C; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 XONTXC Offset Start: 0404Ch Offset End: 0404Fh Power Well: GBEAUX Bit Description Sticky Number of XON packets transmitted. Bit Reset Value Bit Access 0x0 RC 28.18.1.15 XOFF Received Count--XOFFRXC [0:3] (0x4050; RC) This register counts the number of valid XOFF packets received. XOFF packets can use the global address or the station address. This register only increments if receives are enabled (RCTL.RXEN is set). Table 28-199.XOFF Received Count--XOFFRXC [0:3] (0x4050; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 XOFFRXC Power Well: GBEAUX Bit Description Number of XOFF packets received. Intel(R) Communications Chipset 89xx Series - Datasheet 1420 Offset Start: 04050h Offset End: 04053h Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.18.1.16 XOFF Transmitted Count--XOFFTXC [0:3] (0x4054; RC) This register counts the number of XOFF packets transmitted. These can be either due to a full queue or due to software initiated action (using TCTL.SWXOFF). This register only increments if transmits are enabled (TCTL.EN is set). Table 28-200.XOFF Transmitted Count--XOFFTXC [0:3] (0x4054; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 XOFFTXC Offset Start: 04054h Offset End: 04057h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RC Number of XOFF packets transmitted. 28.18.1.17 FC Received Unsupported Count--FCRUC [0:3] (0x4058; RC) This register counts the number of unsupported flow control frames that are received. The FCRUC counter increments when a flow control packet is received that matches either the reserved flow control multicast address (in FCAH/L) or the MAC station address, and has a matching flow control type field match (to the value in FCT), but has an incorrect opcode field. This register only increments if receives are enabled (RCTL.RXEN is set). Table 28-201.FC Received Unsupported Count--FCRUC [0:3] (0x4058; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 FCRUC October 2012 Order Number: 327879-001US Offset Start: 04058h Offset End: 0405Bh Power Well: GBEAUX Bit Description Sticky Number of unsupported flow control frames received. Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1421 28.18.1.18 Packets Received [64 Bytes] Count--PRC64 [0:3] (0x405C; RC) This register counts the number of good packets received that are exactly 64 bytes (from <Destination Address> through <CRC>, inclusive) in length. Packets that are counted in the Missed Packet Count register are not counted in this register. Packets sent to the manageability engine are included in this counter. This register does not include received flow control packets and increments only if receives are enabled (RCTL.RXEN is set). Table 28-202.Packets Received [64 Bytes] Count--PRC64 [0:3] (0x405C; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 PRC64 Offset Start: 0405Ch Offset End: 0405Fh Power Well: GBEAUX Bit Description Sticky Number of packets received that are 64 bytes in length. Bit Reset Value Bit Access 0x0 RC 28.18.1.19 Packets Received [65-127 Bytes] Count--PRC127 [0:3] (0x4060; RC) This register counts the number of good packets received that are 65-127 bytes (from <Destination Address> through <CRC>, inclusive) in length. Packets that are counted in the Missed Packet Count register are not counted in this register. Packets sent to the manageability engine are included in this counter. This register does not include received flow control packets and increments only if receives are enabled (RCTL.RXEN is set). Table 28-203.Packets Received [65-127 Bytes] Count--PRC127 [0:3] (0x4060; RC) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0 Bit Range Bit Acronym 31 :00 PRC127 Power Well: GBEAUX Bit Description Number of packets received that are 65-127 bytes in length. Intel(R) Communications Chipset 89xx Series - Datasheet 1422 Offset Start: 04060h Offset End: 04063h Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.18.1.20 Packets Received [128-255 Bytes] Count--PRC255 [0:3] (0x4064; RC) This register counts the number of good packets received that are 128-255 bytes (from <Destination Address> through <CRC>, inclusive) in length. Packets that are counted in the Missed Packet Count register are not counted in this register. Packets sent to the manageability engine are included in this counter. This register does not include received flow control packets and increments only if receives are enabled (RCTL.RXEN is set). Table 28-204.Packets Received [128-255 Bytes] Count--PRC255 [0:3] (0x4064; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 PRC255 October 2012 Order Number: 327879-001US Offset Start: 04064h Offset End: 04067h Power Well: GBEAUX Bit Description Sticky Number of packets received that are 128-255 bytes in length. Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1423 28.18.1.21 Packets Received [256-511 Bytes] Count--PRC511 [0:3] (0x4068; RC) This register counts the number of good packets received that are 256-511 bytes (from <Destination Address> through <CRC>, inclusive) in length. Packets that are counted in the Missed Packet Count register are not counted in this register. Packets sent to the manageability engine are included in this counter. This register does not include received flow control packets and increments only if receives are enabled (RCTL.RXEN is set). Table 28-205.Packets Received [256-511 Bytes] Count--PRC511 [0:3] (0x4068; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 PRC511 Offset Start: 04068h Offset End: 0406Bh Power Well: GBEAUX Bit Description Sticky Number of packets received that are 256-511 bytes in length. Bit Reset Value Bit Access 0x0 RC 28.18.1.22 Packets Received [512-1023 Bytes] Count--PRC1023 [0:3] (0x406C; RC) This register counts the number of good packets received that are 512-1023 bytes (from <Destination Address> through <CRC>, inclusive) in length. Packets that are counted in the Missed Packet Count register are not counted in this register. Packets sent to the manageability engine are included in this counter. This register does not include received flow control packets and increments only if receives are enabled (RCTL.RXEN is set). Table 28-206.Packets Received [512-1023 Bytes] Count--PRC1023 [0:3] (0x406C; RC) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 PRC1023 Power Well: GBEAUX Bit Description Number of packets received that are 512-1023 bytes in length. Intel(R) Communications Chipset 89xx Series - Datasheet 1424 Offset Start: 0406Ch Offset End: 0406Fh Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.18.1.23 Packets Received [1024 to Max Bytes] Count--PRC1522 [0:3] (0x4070; RC) This register counts the number of good packets received that are from 1024 bytes to the maximum (from <Destination Address> through <CRC>, inclusive) in length. The maximum is dependent on the current receiver configuration (for example, LPE, etc.) and the type of packet being received. If a packet is counted in Receive Oversized Count, it is not counted in this register (see Section 28.18.1.35). This register does not include received flow control packets and only increments if the packet has passed address filtering and receives are enabled (RCTL.RXEN is set). Packets sent to the manageability engine are included in this counter. Due to changes in the standard for maximum frame size for VLAN tagged frames in 802.3, the Controller accepts packets that have a maximum length of 1522 bytes. The RMON statistics associated with this range has been extended to count 1522 byte long packets. If CTRL.Extended_VLAN is set, packets up to 1526 bytes are counted by this counter. Table 28-207.Packets Received [1024 to Max Bytes] Count--PRC1522 [0:3] (0x4070; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Bus:Device:Function: Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 PRC1522 Offset Start: 04070h Offset End: 04073h Power Well: GBEAUX Bit Description Sticky Number of packets received that are 1024-Max bytes in length. Bit Reset Value Bit Access 0x0 RC 28.18.1.24 Good Packets Received Count--GPRC [0:3] (0x4074; RC) This register counts the number of good packets received of any legal length. The legal length for the received packet is defined by the value of Long Packet Enable (CTRL.LPE) (see Section 28.18.1.35). This register does not include received flow control packets and only counts packets that pass filtering. This register only increments if receives are enabled (RCTL.RXEN is set). This register does not count packets counted by the Missed Packet Count (MPC) register. Packets sent to the manageability engine are included in this counter. Note: GPRC can count packets interrupted by a link disconnect although they have a CRC error. Table 28-208.Good Packets Received Count--GPRC [0:3] (0x4074; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 GPRC October 2012 Order Number: 327879-001US Offset Start: 04074h Offset End: 04077h Power Well: GBEAUX Bit Description Sticky Number of good packets received (of any length). Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1425 28.18.1.25 Broadcast Packets Received Count--BPRC [0:3] (0x4078; RC) This register counts the number of good (no errors) broadcast packets received. This register does not count broadcast packets received when the broadcast address filter is disabled. This register only increments if receives are enabled (RCTL.RXEN is set). This register does not count packets counted by the Missed Packet Count (MPC) register. Packets sent to the manageability engine are included in this counter. Table 28-209.Broadcast Packets Received Count - BPRC [0:3] (0x4078; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 BPRC Offset Start: 04078h Offset End: 0407Bh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RC Number of broadcast packets received. 28.18.1.26 Multicast Packets Received Count--MPRC [0:3] (0x407C; RC) This register counts the number of good (no errors) multicast packets received. This register does not count multicast packets received that fail to pass address filtering nor does it count received flow control packets. This register only increments if receives are enabled (RCTL.RXEN is set). This register does not count packets counted by the Missed Packet Count (MPC) register. Packets sent to the manageability engine are included in this counter. Table 28-210.Multicast Packets Received Count--MPRC [0:3] (0x407C; RC) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 MPRC Power Well: GBEAUX Bit Description Number of multicast packets received. Intel(R) Communications Chipset 89xx Series - Datasheet 1426 Offset Start: 0407Ch Offset End: 0407Fh Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.18.1.27 Good Packets Transmitted Count--GPTC [0:3] (0x4080; RC) This register counts the number of good (no errors) packets transmitted. A good transmit packet is considered one that is 64 or more bytes in length (from <Destination Address> through <CRC>, inclusively) in length. This does not include transmitted flow control packets. This register only increments if transmits are enabled (TCTL.EN is set). The register counts clear as well as secure packets. Table 28-211.Good Packets Transmitted Count--GPTC [0:3] (0x4080; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0h Bit Range Bit Acronym 31 :00 GPTC Offset Start: 04080h Offset End: 04083h Power Well: GBEAUX Bit Description Bit Reset Value Sticky Number of good packets transmitted. 0x0 Bit Access RC 28.18.1.28 Good Octets Received Count--GORCL [0:3] (0x4088; RC) These registers make up a 64-bit register that counts the number of good (no errors) octets received. This register includes bytes received in a packet from the <Destination Address> field through the <CRC> field, inclusive; GORCL must be read before GORCH. In addition, it sticks at 0xFFFF_FFFF_FFFF_FFFF when the maximum value is reached. Only octets of packets that pass address filtering are counted in this register. This register does not count octets of packets counted by the Missed Packet Count (MPC) register. Octets of packets sent to the manageability engine are included in this counter. This register only increments if receives are enabled (RCTL.RXEN is set). These octets do not include octets of received flow control packets. Table 28-212.Good Octets Received Count--GORCL [0:3] (0x4088; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Bus:Device:Function: Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 GORCL October 2012 Order Number: 327879-001US Offset Start: 04088h Offset End: 0408Bh Power Well: GBEAUX Bit Description Sticky Number of good octets received - lower 4 bytes. Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1427 28.18.1.29 Good Octets Received Count--GORCH [0:3] (0x408C; RC) Table 28-213.Good Octets Received Count--GORCH [0:3] (0x408C; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 GORCH Offset Start: 0408Ch Offset End: 0408Fh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RC Number of good octets received - upper 4 bytes. 28.18.1.30 Good Octets Transmitted Count--GOTCL [0:3] (0x4090; RC) These registers make up a 64-bit register that counts the number of good (no errors) packets transmitted. This register must be accessed using two independent 32-bit accesses; GOTCL must be read before GOTCH. In addition, it sticks at 0xFFFF_FFFF_FFFF_FFFF when the maximum value is reached. This register includes bytes transmitted in a packet from the <Destination Address> field through the <CRC> field, inclusive. This register counts octets in successfully transmitted packets that are 64 or more bytes in length. This register only increments if transmits are enabled (TCTL.EN is set). The register counts clear as well as secure octets. These octets do not include octets in transmitted flow control packets. Table 28-214.Good Octets Transmitted Count--GOTCL [0:3] (0x4090; RC) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 GOTCL Power Well: GBEAUX Bit Description Number of good octets transmitted - lower 4 bytes. Intel(R) Communications Chipset 89xx Series - Datasheet 1428 Offset Start: 04090h Offset End: 04093h Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.18.1.31 Good Octets Transmitted Count--GOTCH [0:3] (4094; RC) Table 28-215.Good Octets Transmitted Count--GOTCH [0:3] (4094; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Offset Start: 04094h Offset End: 04097h B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 GOTCH Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RC Number of good octets transmitted - upper 4 bytes. 28.18.1.32 Receive No Buffers Count--RNBC [0:3] (0x40A0; RC) This register counts the number of times that data was not placed to host because there were no available buffers to store those frames (receive descriptor head and tail pointers were equal). The packet is still received if there is space in the FIFO. This register only increments if receives are enabled (RCTL.RXEN is set). This register does not increment when flow control packets are received. Table 28-216.Receive No Buffers Count--RNBC [0:3] (0x40A0; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 RNBC October 2012 Order Number: 327879-001US Offset Start: 040A0h Offset End: 040A3h Power Well: GBEAUX Bit Description Number of receive no buffer conditions Sticky Bit Reset Value 0x0 Bit Access RC Intel(R) Communications Chipset 89xx Series - Datasheet 1429 28.18.1.33 Receive Undersize Count--RUC [0:3] (0x40A4; RC) This register counts the number of received frames that passed address filtering, and were less than minimum size (64 bytes from <Destination Address> through <CRC>, inclusive), and had a valid CRC. This register only increments if receives are enabled (RCTL.RXEN is set). Table 28-217.Receive Undersize Count--RUC [0:3] (0x40A4; RC) Description: View: PCI BAR: GBEPCIBAR0[0:3] Size: 32 bit Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 RUC Offset Start: 040A4h Offset End: 040A7h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RC Number of receive undersize errors. 28.18.1.34 Receive Fragment Count--RFC [0:3] (0x40A8; RC) This register counts the number of received frames that passed address filtering, and were less than minimum size (64 bytes from <Destination Address> through <CRC>, inclusive), but had a bad CRC (this is slightly different from the Receive Undersize Count register). This register only increments if receives are enabled (RCTL.RXEN is set). Table 28-218.Receive Fragment Count--RFC [0:3] (0x40A8; RC) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :00 RFC Note: Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RC Number of receive fragment errors. Runt packets smaller than 25 bytes may not be counted by this counter. Intel(R) Communications Chipset 89xx Series - Datasheet 1430 Offset Start: 040A8h Offset End: 040ABh October 2012 Order Number: 327879-001US 28.0 28.18.1.35 Receive Oversize Count--ROC [0:3] (0x40AC; RC) This register counts the number of received frames with valid CRC field that passed address filtering, and were greater than maximum size. Packets over 1522 bytes are oversized if LongPacketEnable (RCTL.LPE) is 0b. If LongPacketEnable is 1b, then an incoming packet is considered oversized if it exceeds the value set in the RLPML register. In VMDq mode, a packet is counted only if it is bigger than the VOMLR.RLPML value for all the VMs that where supposed to receive the packet. If receives are not enabled, this register does not increment. These lengths are based on bytes in the received packet from <Destination Address> through <CRC>, inclusive. Note: The maximum size of a packet when LPE is 0b is fixed according to the CTRL_EXT.Extended_VLAN bit and the detection of a VLAN tag in the packet. Table 28-219.Receive Oversize Count--ROC [0:3] (0x40AC; RC) Description: View: PCI Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :00 ROC Offset Start: 040ACh Offset End: 040AFh Power Well: GBEAUX Bit Description Sticky Number of receive oversize errors. Bit Reset Value Bit Access 0x0 RC 28.18.1.36 Receive Jabber Count--RJC [0:3] (0x40B0; RC) This register counts the number of received frames that passed address filtering, and were greater than maximum size and had a bad CRC (this is slightly different from the Receive Oversize Count register). Packets over 1518/1522/1526 bytes are oversized if LPE is 0b. If LPE is 1b, then an incoming packet is considered oversized if it exceeds RLPML.LPML bytes. If receives are not enabled, this register does not increment. These lengths are based on bytes in the received packet from <Destination Address> through <CRC>, inclusive. Note: The maximum size of a packet when LPE is 0b is fixed according to the CTRL_EXT.Extended_VLAN bit and the detection of a VLAN tag in the packet. Table 28-220.Receive Jabber Count--RJC [0:3] (0x40B0; RC) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 RJC October 2012 Order Number: 327879-001US Offset Start: 040B0h Offset End: 040B3h Power Well: GBEAUX Bit Description Number of receive jabber errors. Sticky Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1431 28.18.1.37 Management Packets Received Count--MNGPRC [0:3] (0x40B4; RC) This register counts the total number of packets received that pass the management filters as described in the Total Cost of Ownership (TCO) System Management Bus Interface Application Note. Any packets with errors are not counted, except packets that are dropped because the management receive FIFO is full. Packets sent to both the host and the management interface are not counted by this counter. Table 28-221.Management Packets Received Count--MNGPRC [0:3] (0x40B4; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 MNGPRC Offset Start: 040B4h Offset End: 040B7h Power Well: GBEAUX Bit Description Sticky Number of management packets received. Bit Reset Value Bit Access 0x0 RC 28.18.1.38 Management Packets Dropped Count--MPDC [0:3] (0x40B8; RC) This register counts the total number of packets received that pass the management filters as described in the Total Cost of Ownership (TCO) System Management Bus Interface Application Note, that are dropped because the management receive FIFO is full. Management packets include any packet directed to the manageability console (for example, BMC and ARP packets). Table 28-222.Management Packets Dropped Count--MPDC [0:3] (0x40B8; RC) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 MPDC Power Well: GBEAUX Bit Description Number of management packets dropped. Intel(R) Communications Chipset 89xx Series - Datasheet 1432 Offset Start: 040B8h Offset End: 040BBh Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.18.1.39 Management Packets Transmitted Count--MNGPTC [0:3] (0x40BC; RC) This register counts the total number of transmitted packets originating from the manageability path. Table 28-223.Management Packets Transmitted Count--MNGPTC [0:3] (0x40BC; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 MPTC Offset Start: 040BCh Offset End: 040BFh Power Well: GBEAUX Bit Description Sticky Number of management packets transmitted. Bit Reset Value Bit Access 0x0 RC 28.18.1.40 Total Octets Received--TORL [0:3] (0x40C0; RC) These registers make up a logical 64-bit register which counts the total number of octets received. This register must be accessed using two independent 32-bit accesses; TORL must be read before TORH. This register sticks at 0xFFFF_FFFF_FFFF_FFFF when the maximum value is reached. All packets received have their octets summed into this register, regardless of their length, whether they are erred, or whether they are flow control packets. This register includes bytes received in a packet from the <Destination Address> field through the <CRC> field, inclusive. This register only increments if receives are enabled (RCTL.RXEN is set). Note: Broadcast rejected packets are counted in this counter (as opposed to all other rejected packets that are not counted). Table 28-224.Total Octets Received--TORL [0:3] (0x40C0; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 TORL October 2012 Order Number: 327879-001US Offset Start: 040C0h Offset End: 040C3h Power Well: GBEAUX Bit Description Sticky Number of total octets received - lower 4 bytes. Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1433 28.18.1.41 Total Octets Received--TORH [0:3] (0x40C4; RC) Table 28-225.Total Octets Received--TORH [0:3] (0x40C4; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 TORH Offset Start: 040C4h Offset End: 040C7h Power Well: GBEAUX Bit Description Sticky Number of total octets received - upper 4 bytes. Bit Reset Value Bit Access 0x0 RC 28.18.1.42 Total Octets Transmitted--TOTL [0:3] (0x40C8; RC) These registers make up a 64-bit register that counts the total number of octets transmitted. This register must be accessed using two independent 32-bit accesses; TOTL must be read before TOTH. This register sticks at 0xFFFF_FFFF_FFFF_FFFF when the maximum value is reached. All transmitted packets have their octets summed into this register, regardless of their length or whether they are flow control packets. This register includes bytes transmitted in a packet from the <Destination Address> field through the <CRC> field, inclusive. Octets transmitted as part of partial packet transmissions (for example, collisions in half-duplex mode) are not included in this register. This register only increments if transmits are enabled (TCTL.EN is set). Table 28-226.Total Octets Transmitted--TOTL [0:3] (0x40C8; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 TOTL Number of total octets transmitted - lower 4 bytes. Intel(R) Communications Chipset 89xx Series - Datasheet 1434 Offset Start: 040C8h Offset End: 040CBh Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.18.1.43 Total Octets Transmitted--TOTH [0:3] (0x40CC; RC) Table 28-227.Total Octets Transmitted--TOTH [0:3] (0x40CC; RC) Description: View: PCI BAR: GBEPCIBAR0[0:3] Size: 32 bit Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 TOTH Offset Start: 040CCh Offset End: 040CFh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RC Number of total octets transmitted - upper 4 bytes. 28.18.1.44 Total Packets Received--TPR [0:3] (0x40D0; RC) This register counts the total number of all packets received. All packets received are counted in this register, regardless of their length, whether they have errors, or whether they are flow control packets. This register only increments if receives are enabled (RCTL.RXEN is set). Note: Broadcast rejected packets are counted in this counter (as opposed to all other rejected packets that are not counted). Note: Runt packets smaller than 25 bytes may not be counted by this counter. TPR can count packets interrupted by a link disconnect although they have a CRC error. Table 28-228.Total Packets Received--TPR [0:3] (0x40D0; RC) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 TPR October 2012 Order Number: 327879-001US Offset Start: 040D0h Offset End: 040D3h Power Well: GBEAUX Bit Description Number of all packets received. Sticky Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1435 28.18.1.45 Total Packets Transmitted--TPT [0:3] (0x40D4; RC) This register counts the total number of all packets transmitted. All packets transmitted are counted in this register, regardless of their length, or whether they are flow control packets. Partial packet transmissions (collisions in half-duplex mode) are not included in this register. This register only increments if transmits are enabled (TCTL.EN is set). This register counts all packets, including standard packets, secure packets, packets received over the SMBus, and packets generated by the PT function. Table 28-229.Total Packets Transmitted--TPT [0:3] (0x40D4; RC Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 TPT Offset Start: 040D4h Offset End: 040D7h Power Well: GBEAUX Bit Description Sticky Number of all packets transmitted. Bit Reset Value Bit Access 0x0 RC 28.18.1.46 Packets Transmitted [64 Bytes] Count--PTC64 [0:3] (0x40D8; RC) This register counts the number of packets transmitted that are exactly 64 bytes (from <Destination Address> through <CRC>, inclusive) in length. Partial packet transmissions (collisions in half-duplex mode) are not included in this register. This register does not include transmitted flow control packets (which are 64 bytes in length). This register only increments if transmits are enabled (TCTL.EN is set). This register counts all packets, including standard packets, secure packets, packets received over the SMBus, and packets generated by the PT function. Table 28-230.Packets Transmitted [64 Bytes] Count--PTC64 [0:3] (0x40D8; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 PTC64 Power Well: GBEAUX Bit Description Number of packets transmitted that are 64 bytes in length. Intel(R) Communications Chipset 89xx Series - Datasheet 1436 Offset Start: 040D8h Offset End: 040DBh Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.18.1.47 Packets Transmitted [65-127 Bytes] Count--PTC127 [0:3] (0x40DC; RC) This register counts the number of packets transmitted that are 65-127 bytes (from <Destination Address> through <CRC>, inclusive) in length. Partial packet transmissions (for example, collisions in half-duplex mode) are not included in this register. This register only increments if transmits are enabled (TCTL.EN is set). This register counts all packets, including standard packets, secure packets, packets received over the SMBus, and packets generated by the PT function. Table 28-231.Packets Transmitted [65-127 Bytes] Count--PTC127 [0:3] (0x40DC; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 PTC127 Offset Start: 040DCh Offset End: 040DFh Power Well: GBEAUX Bit Description Sticky Number of packets transmitted that are 65-127 bytes in length. Bit Reset Value Bit Access 0x0 RC 28.18.1.48 Packets Transmitted [128-255 Bytes] Count--PTC255 [0:3] (0x40E0; RC) This register counts the number of packets transmitted that are 128-255 bytes (from <Destination Address> through <CRC>, inclusive) in length. Partial packet transmissions (collisions in half-duplex mode) are not included in this register. This register only increments if transmits are enabled (TCTL.EN is set). This register counts all packets, including standard packets, secure packets, packets received over the SMBus, and packets generated by the PT function. Table 28-232.Packets Transmitted [128-255 Bytes] Count--PTC255 [0:3] (0x40E0; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 PTC255 Number of packets transmitted that are 128-255 bytes in length. October 2012 Order Number: 327879-001US Offset Start: 040E0h Offset End: 040E3h Sticky Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1437 28.18.1.49 Packets Transmitted [256-511 Bytes] Count--PTC511 [0:3] (0x40E4; RC) This register counts the number of packets transmitted that are 256-511 bytes (from <Destination Address> through <CRC>, inclusive) in length. Partial packet transmissions (for example, collisions in half-duplex mode) are not included in this register. This register only increments if transmits are enabled (TCTL.EN is set). This register counts all packets, including standard and secure packets. Management packets must never be more than 200 bytes. Table 28-233.Packets Transmitted [256-511 Bytes] Count--PTC511 [0:3] (0x40E4; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 040E4h Offset End: 040E7h Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 PTC511 Number of packets transmitted that are 256-511 bytes in length. Sticky Bit Reset Value Bit Access 0x0 RC 28.18.1.50 Packets Transmitted [512-1023 Bytes] Count--PTC1023 [0:3] (0x40E8; RC) This register counts the number of packets transmitted that are 512-1023 bytes (from <Destination Address> through <CRC>, inclusive) in length. Partial packet transmissions (for example, collisions in half-duplex mode) are not included in this register. This register only increments if transmits are enabled (TCTL.EN is set). This register counts all packets, including standard and secure packets. Management packets must never be more than 200 bytes. Table 28-234.Packets Transmitted [512-1023 Bytes] Count--PTC1023 [0:3] (0x40E8; RC) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 PTC1023 Number of packets transmitted that are 512-1023 bytes in length. Intel(R) Communications Chipset 89xx Series - Datasheet 1438 Offset Start: 040E8h Offset End: 040EBh Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.18.1.51 Packets Transmitted [1024 Bytes or Greater] Count--PTC1522 [0:3] (0x40EC; RC) This register counts the number of packets transmitted that are 1024 or more bytes (from <Destination Address> through <CRC>, inclusive) in length. Partial packet transmissions (for example, collisions in half-duplex mode) are not included in this register. This register only increments if transmits are enabled (TCTL.EN is set). Due to changes in the standard for maximum frame size for VLAN tagged frames in 802.3, the Controller transmits packets that have a maximum length of 1522 bytes. The RMON statistics associated with this range has been extended to count 1522 byte long packets. This register counts all packets, including standard and secure packets (management packets must never be more than 200 bytes). If CTRL.Extended_VLAN is set, packets up to 1526 bytes are counted by this counter. Table 28-235.Packets Transmitted [1024 Bytes or Greater] Count--PTC1522 [0:3] (0x40EC; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 PTC1522 Offset Start: 040ECh Offset End: 040EFh Power Well: GBEAUX Bit Description Sticky Number of packets transmitted that are 1024 or more bytes in length. Bit Reset Value Bit Access 0x0 RC 28.18.1.52 Multicast Packets Transmitted Count--MPTC [0:3] (0x40F0; RC) This register counts the number of multicast packets transmitted. This register does not include flow control packets and increments only if transmits are enabled (TCTL.EN is set). Counts clear as well as secure traffic. Table 28-236.Multicast Packets Transmitted Count--MPTC [0:3] (0x40F0; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 MPTC October 2012 Order Number: 327879-001US Offset Start: 040F0h Offset End: 040F3h Power Well: GBEAUX Bit Description Sticky Number of multicast packets transmitted. Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1439 28.18.1.53 Broadcast Packets Transmitted Count--BPTC [0:3] (0x40F4; RC) This register counts the number of broadcast packets transmitted. This register only increments if transmits are enabled (TCTL.EN is set). This register counts all packets, including standard and secure packets (management packets must never be more than 200 bytes). Table 28-237.Broadcast Packets Transmitted Count--BPTC [0:3] (0x40F4; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 BPTC Offset Start: 040F4h Offset End: 040F7h Power Well: GBEAUX Bit Description Sticky Number of broadcast packets transmitted count. Bit Reset Value Bit Access 0x0 RC 28.18.1.54 TCP Segmentation Context Transmitted Count--TSCTC [0:3] (0x40F8; RC) This register counts the number of TCP segmentation offload transmissions and increments once the last portion of the TCP segmentation context payload is segmented and loaded as a packet into the on-chip transmit buffer. It is not a measurement of the number of packets sent out (covered by other registers). This register only increments if transmits and TCP segmentation offload are enabled. This counter only counts pure TSO transmissions. Table 28-238.TCP Segmentation Context Transmitted Count--TSCTC [0:3] (0x40F8; RC) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 TSCTC Number of TCP Segmentation contexts transmitted count. Intel(R) Communications Chipset 89xx Series - Datasheet 1440 Offset Start: 040F8h Offset End: 040FBh Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.18.1.55 Interrupt Assertion Count--IAC [0:3] (0x4100; RC) This counter counts the total number of LAN interrupts generated in the system. In case of MSI-X systems, this counter reflects the total number of MSI-X messages that are emitted. Table 28-239.Interrupt Assertion Count--IAC [0:3] (0x4100; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Offset Start: 04100h Offset End: 04103h Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym Bit Description Sticky 31 :00 IAC This is a count of all the LAN interrupt assertions that have occurred. Bit Reset Value Bit Access 0x0 RC 28.18.1.56 Rx Packets to Host Count--RPTHC [0:3] (0x4104; RC) Table 28-240.Rx Packets to Host Count--RPTHC [0:3] (0x4104; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 04104h Offset End: 04107h Power Well: GBEAUX Bit Range Bit Acronym Bit Description Sticky 31 :00 RPTHC This is a count of all the received packets sent to the host. Bit Reset Value Bit Access 0x0 RC 28.18.1.57 Debug Counter 1--DBGC1 [0:3] (0x4108; RC) Table 28-241.Debug Counter 1--DBGC1 [0:3] (0x4108; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Bus:Device:Function: Index1 Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 DBGC1 This field counts the events according to the value of the PBDIAG.STAT_SEL field. The list of possible values for this counter are described in Table 28-242: October 2012 Order Number: 327879-001US Offset Start: 04108h Offset End: 0410Bh Sticky Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1441 Table 28-242.DBGC1 Values Stat Sel Counter 1 content 0 Reserved 1 The number of Tx descriptor WB transactions performed for Q0 2 The number of Rx descriptor WB transactions performed for Q0 3 The number of Rx descriptor immediate WB transactions performed for Q0 4 The number of Tx host descriptors read by the Descriptor Handler 5 The number of Rx host descriptors read by the Descriptor Processor 6 The number of Tx data read requests done by the Dhost 7 The number of TX packets sent to DBU 28.18.1.58 Debug Counter 2--DBGC2 [0:3] (0x410C; RC) Table 28-243.Debug Counter 2--DBGC2 [0:3] (0x410C; RC) Description: View: PCI B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Size: 32 bit Default: 0x0 Offset Start: 0410Ch Offset End: 0410Fh Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 DBGC2 This field counts the events according to the value of the PBDIAG.STAT_SEL field. The list of possible values for this counter are described in Table 28-244. Sticky Bit Reset Value Bit Access 0x0 RC Table 28-244.DBGC2 Values Stat sel Counter 2 content 0 Number of Rx filter packets read from DBU 1 The number of Tx descriptor WB transactions performed for Q1 2 The number of Rx descriptor WB transactions performed for Q1 3 The number of Rx descriptor immediate WB transactions performed for Q1 4 The number of Tx host descriptors written back to host 5 The number of Rx host descriptors written back to host 6 The number of Rx data write requests done by the Dhost 7 Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1442 October 2012 Order Number: 327879-001US 28.0 28.18.1.59 Debug Counter 3--DBGC3 [0:3] (0x4110; RC) Table 28-245.Debug Counter 3--DBGC3 [0:3] (0x4110; RC) Description: View: PCI BAR: GBEPCIBAR0[0:3] Size: 32 bit Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 04110h Offset End: 04113h Power Well: GBEAUX Bit Range Bit Acronym Bit Description Sticky 31 :00 DBGC3 This field counts the events according to the value of the PBDIAG.STAT_SEL field. The list of possible values for this counter are described in Table 28-246: Bit Reset Value Bit Access 0x0 RC Table 28-246.DBGC3 Values Stat sel Counter 3 content 0 Reserved 1 The number of Tx descriptor WB transactions performed for Q2 2 The number of Rx descriptor WB transactions performed for Q2 3 The number of Rx descriptor immediate WB transactions performed for Q2 4 The number of dropped Tx packets 5 The number of Rx packets written to the DBU 6 The number of PCIe write requests done by the DMA 7 The total amount of single send packets 28.18.1.60 Debug Counter 4--DBGC4 [0:3] (0x411C; RC) Table 28-247.Debug Counter 4--DBGC4 [0:3] (0x411C; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 DBGC4 This field counts the events according to the value of the PBDIAG.STAT_SEL field. The list of possible values for this counter are described in Table 28-248. October 2012 Order Number: 327879-001US Offset Start: 0411Ch Offset End: 0411Fh Sticky Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1443 Table 28-248.DBGC4 Values Stat sel Counter 4 content 0 Reserved 1 The number of Tx descriptor WB transactions performed for Q3 2 The number of Rx descriptor WB transactions performed for Q3 3 The number of Rx descriptor immediate WB transactions performed for Q3 4 The number of Tx packets read from the DBU 5 The number of Rx packets read from the DBU 6 Number of PCIe read/write requests done by the Dhost 7 The total amount of large send packets 28.18.1.61 Host Good Packets Transmitted Count--HGPTC [0:3] (0x4118; RC) Table 28-249.Host Good Packets Transmitted Count--HGPTC [0:3] (0x4118; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 HGPTC Offset Start: 04118h Offset End: 0411Bh Power Well: GBEAUX Bit Description Sticky Number of good packets transmitted by the host. Bit Reset Value Bit Access 0x0 RC This register counts the number of good (non-erred) packets transmitted sent by the host. A good transmit packet is considered one that is 64 or more bytes in length (from <Destination Address> through <CRC>, inclusively) in length. This does not include transmitted flow control packets or packets sent by the manageability engine. This register only increments if transmits are enabled (TCTL.EN is set). 28.18.1.62 Receive Descriptor Minimum Threshold Count--RXDMTC [0:3] (0x4120; RC) Table 28-250.Receive Descriptor Minimum Threshold Count--RXDMTC [0:3] (0x4120; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 RXDMTC Offset Start: 04120h Offset End: 04123h Power Well: GBEAUX Bit Description This is a count of the receive descriptor minimum threshold events Sticky Bit Reset Value Bit Access 0x0 RC This register counts the number of events where the number of descriptors in one of the Rx queues was lower than the threshold defined for this queue. Intel(R) Communications Chipset 89xx Series - Datasheet 1444 October 2012 Order Number: 327879-001US 28.0 28.18.1.63 Host Good Octets Received Count--HGORCL [0:3] (0x4128; RC) Table 28-251.Host Good Octets Received Count--HGORCL [0:3] (0x4128; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 HGORCL Offset Start: 04128h Offset End: 0412Bh Power Well: GBEAUX Bit Description Sticky Number of good octets received by host - lower 4 bytes Bit Reset Value Bit Access 0x0 RC 28.18.1.64 Host Good Octets Received Count--HGORCH [0:3] (0x412C; RC) Table 28-252.Host Good Octets Received Count--HGORCH [0:3] (0x412C; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 HGORCH Offset Start: 0412Ch Offset End: 0412Fh Power Well: GBEAUX Bit Description Sticky Number of good octets received by host - upper 4 bytes Bit Reset Value Bit Access 0x0 RC These registers make up a logical 64-bit register which counts the number of good (non-erred) octets received. This register includes bytes received in a packet from the <Destination Address> field through the <CRC> field, inclusive. This register must be accessed using two independent 32-bit accesses.; HGORCL must be read before HGORCH. In addition, it sticks at 0xFFFF_FFFF_FFFF_FFFF when the maximum value is reached. Only packets that pass address filtering are counted in this register. This register counts only octets of packets that reached the host. The only exception is packets dropped by the DMA because of lack of descriptors in one of the queues. These packets are included in this counter. This register only increments if receives are enabled (RCTL.RXEN is set). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1445 28.18.1.65 Host Good Octets Transmitted Count--HGOTCL [0:3] (0x4130; RC) Table 28-253.Host Good Octets Transmitted Count--HGOTCL [0:3] (0x4130; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 HGOTCL Offset Start: 04130h Offset End: 04133h Power Well: GBEAUX Bit Description Sticky Number of good octets transmitted by host - lower 4 bytes Bit Reset Value Bit Access 0x0 RC 28.18.1.66 Host Good Octets Transmitted Count - HGOTCH [0:3] (0x4134; RC) Table 28-254.Host Good Octets Transmitted Count - HGOTCH [0:3] (0x4134; RC) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 HGOTCH Offset Start: 04134h Offset End: 04137h Power Well: GBEAUX Bit Description Number of good octets transmitted by host - upper 4 bytes Sticky Bit Reset Value Bit Access 0x0 RC These registers make up a logical 64-bit register which counts the number of good (non-erred) packets transmitted. This register must be accessed using two independent 32-bit accesses. This register resets whenever the upper 32 bits are read (HGOTCH). In addition, it sticks at 0xFFFF_FFFF_FFFF_FFFF when the maximum value is reached. This register includes bytes transmitted in a packet from the <Destination Address> field through the <CRC> field, inclusive. This register counts octets in successfully transmitted packets which are 64 or more bytes in length. This register only increments if transmits are enabled (TCTL.EN is set). The register counts clear as well as secure octets. These octets do not include octets in transmitted flow control packets or manageability packets. Intel(R) Communications Chipset 89xx Series - Datasheet 1446 October 2012 Order Number: 327879-001US 28.0 28.18.1.67 Length Error Count--LENERRS [0:3] (0x4138; RC) Table 28-255.Length Error Count--LENERRS [0:3] (0x4138; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 LENERRS Offset Start: 04138h Offset End: 0413Bh Power Well: GBEAUX Bit Description Sticky Length error count. Bit Reset Value Bit Access 0x0 RC Counts the number of receive packets with Length errors. For example, valid packets (no CRC error) with a length/Type field with a value smaller or equal to 1500 greater than the frame size. In order for a packet to be counted in this register, it must pass address filtering and must be 64 bytes or greater (from <Destination Address> through <CRC>, inclusive) in length. If receives are not enabled, then this register does not increment. 28.18.1.68 SerDes/SGMII/KX Code Violation Packet Count--SCVPC [0:3] (0x4228; RW) This register contains the number of code violation packets received. Code violation is defined as an invalid received code in the middle of a packet. Table 28-256.SerDes/SGMII/KX Code Violation Packet Count--SCVPC [0:3] (0x4228; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Bus:Device:Function: Index1 Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 CODEVIO Code Violation Packet Count: At any point of time this field specifies number of unknown protocol packets received. Valid only in SGMII/SerDes/1000BASE-KX modes. October 2012 Order Number: 327879-001US Offset Start: 04228h Offset End: 0422Bh Sticky Bit Reset Value Bit Access 0x0 RW Intel(R) Communications Chipset 89xx Series - Datasheet 1447 28.18.1.69 Switch Drop Packet Count--SDPC [0:3] (0x41A4; RC) Table 28-257.Switch Drop Packet Count--SDPC [0:3] (0x41A4; RC) Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 041A4h Offset End: 041A7h Power Well: GBEAUX Bit Acronym Bit Description SDPC Switch Drop Packet Count: This register counts Rx packets dropped at the pool selection stage of the switch or by the storm control mechanism. For example, packets that where not routed to any of the pools and the VT_CTL.Dis_Def_Pool is set. Valid only in VMDq mode. Sticky Bit Reset Value Bit Access 0x0 RC 28.18.1.70 Virtualization Statistical Counters The Controller supports 5 statistical counters per queue to reduce processing overhead in virtualization operating mode. 28.18.1.71 Per Queue Good Packets Received Count--VFGPRC [0:3][0:7] (0x10010 + n*0x100 [n=0...7]; RO) This register counts the number of legal length good packets received in queue[n]. The legal length for the received packet is defined by the value of Long Packet Enable (CTRL.LPE) (see Section 28.18.1.35). This register does not include received flow control packets and only counts packets that pass filtering. This register only increments if receive is enabled. Note: VFGPRC may count packets interrupted by a link disconnect although they have a CRC error. Unlike some other statistics registers that are not allocated per VM, this register is not cleared on read. Furthermore, the register wraps around back to 0x0000 on the next increment when reaching a value of 0xFFFF and then continues normal count operation. Table 28-258.Per Queue Good Packets Received Count--VFGPRC [0:3][0:7](0x10010 + n*0x100 [n=0...7]; RO) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 GPRC Power Well: GBEAUX Bit Description Number of good packets received (of any length). Intel(R) Communications Chipset 89xx Series - Datasheet 1448 10010h at Offset Start: 0x100 Offset End: 10013h at 0x100 Sticky Bit Reset Value Bit Access 0x0 RO October 2012 Order Number: 327879-001US 28.0 28.18.1.72 Per Queue Good Packets Transmitted Count--VFGPTC [0:3][0:7] (0x10014 + n*0x100 [n=0...7]; RO) This register counts the number of good (no errors) packets transmitted on queue[n]. A good transmit packet is considered one that is 64 or more bytes in length (from <Destination Address> through <CRC>, inclusively) in length. This does not include transmitted flow control packets. This register only increments if transmits are enabled (TCTL.EN is set). The register counts clear as well as secure packets. This counter includes loopback packets or packets later dropped by the MAC. Note: Unlike some other statistic registers that are not allocated per VM, this register is not cleared on read. Furthermore, the register wraps around back to 0x0000 on the next increment when reaching a value of 0xFFFF and then continues normal count operation. Table 28-259.Per Queue Good Packets Transmitted Count--VFGPTC [0:3][0:7](0x10014 + n*0x100 [n=0...7]; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 GPTC October 2012 Order Number: 327879-001US 10014h at Offset Start: 0x100 Offset End: 10017h at 0x100 Power Well: GBEAUX Bit Description Number of good packets transmitted. Sticky Bit Reset Value Bit Access 0x0 RO Intel(R) Communications Chipset 89xx Series - Datasheet 1449 28.18.1.73 Per Queue Good Octets Received Count--VFGORC [0:3][0:7] (0x10018 + n*0x100 [n=0...7]; RO) This register counts the number of good (no errors) octets received on queue[n]. This register includes bytes received in a packet from the <Destination Address> field through the <CRC> field, inclusive. Only octets of packets that pass address filtering are counted in this register. This register only increments if receive is enabled. Note: Unlike some other statistic registers that are not allocated per VM, this register is not cleared on read. Furthermore, the register wraps around back to 0x0000 on the next increment when reaching a value of 0xFFFF and then continues normal count operation. Table 28-260.Per Queue Good Octets Received Count--VFGORC [0:3][0:7] (0x10018 + n*0x100 [n=0...7]; RO) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 GORC Power Well: GBEAUX Bit Description Number of good octets received. Intel(R) Communications Chipset 89xx Series - Datasheet 1450 10018h at Offset Start: 0x100 Offset End: 1001Bh at 0x100 Sticky Bit Reset Value Bit Access 0x0 RO October 2012 Order Number: 327879-001US 28.0 28.18.1.74 Per Queue Good Octets Transmitted Count--VFGOTC [0:3][0:7] (0x10034 + n*0x100 [n=0...7]; RO) This register counts the number of good (no errors) packets transmitted on queue[n]. This register includes bytes transmitted in a packet from the <Destination Address> field through the <CRC> field, inclusive. Register also counts any padding and VLAN tag that were added by the hardware. This register counts octets in successfully transmitted packets that are 64 or more bytes in length. Octets counted do not include octets in transmitted flow control packets. This register only increments if transmit is enabled. Note: Unlike some other statistic registers that are not allocated per VM, this register is not cleared on read. Furthermore, the register wraps around back to 0x0000 on the next increment when reaching a value of 0xFFFF and then continues normal count operation. Table 28-261.Per Queue Good Octets Transmitted Count--VFGOTC [0:3][0:7] (0x10034 + n*0x100 [n=0...7]; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 GOTC 10034h at Offset Start: 0x100 Offset End: 10037h at 0x100 Power Well: GBEAUX Bit Description Sticky Number of good octets transmitted - lower 4 bytes. Bit Reset Value Bit Access 0x0 RO 28.18.1.75 Per Queue Multicast Packets Received Count--VFMPRC [0:3][0:7] (0x10038 + n*0x100 [n=0...7]; RO) This register counts the number of good (no errors) multicast packets received on queue[n]. This register does not count multicast packets received that fail to pass address filtering nor does it count received flow control packets. This register only increments if receive is enabled. Note: Unlike some other statistic registers that are not allocated per VM, this register is not cleared on read. Furthermore, the register wraps around back to 0x0000 on the next increment when reaching a value of 0xFFFF and then continues normal count operation. Table 28-262.Per Queue Multicast Packets Received Count--VFMPRC [0:3][0:7] (0x10038 + n*0x100 [n=0...7]; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Bus:Device:Function: Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 MPRC October 2012 Order Number: 327879-001US 10038h at Offset Start: 0x100 Offset End: 1003Bh at 0x100 Power Well: GBEAUX Bit Description Number of multicast packets received. Sticky Bit Reset Value Bit Access 0x0 RO Intel(R) Communications Chipset 89xx Series - Datasheet 1451 28.19 Manageability Statistics This section describes a set of statistics counters used by the SMBUS interface and are not accessible to the host driver. 28.19.1 Detailed Register Descriptions 28.19.1.1 BMC Management Packets Dropped Count--BMPDC [0:3] (0x4140; RC) This register counts the total number of packets received that pass the management filters as described in the Total Cost of Ownership (TCO) System Management Bus Interface Application Note, that are dropped because the management receive FIFO is full. Management packets include any packet directed to the manageability console (for example, BMC and ARP packets). This register is available to the firmware only. Table 28-263.BMC Management Packets Dropped Count--BMPDC [0:3] (0x4140; RC) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :00 BMPDC 28.19.1.2 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Offset Start: 04140h Offset End: 04143h Power Well: GBEAUX Bit Description Sticky Number of management packets dropped. Bit Reset Value Bit Access 0x0 RC BMC Management Packets Transmitted Count--BMNGPTC [0:3] (0x4144; RC) This register counts the total number of transmitted packets originating from the manageability path. This register is available to the firmware only. Table 28-264.BMC Management Packets Transmitted Count--BMNGPTC [0:3] (0x4144; RC) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 BMNGPTC Power Well: GBEAUX Bit Description Number of management packets transmitted. Intel(R) Communications Chipset 89xx Series - Datasheet 1452 Offset Start: 4144h Offset End: 4147h Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.19.1.3 BMC Management Packets Received Count--BMNGPRC [0:3] (0x413C; RC) This register counts the total number of packets received that pass the management filters as described in the Total Cost of Ownership (TCO) System Management Bus Interface Application Note. Any packets with errors are not counted, except packets that are dropped because the management receive FIFO is full. This register is available to the firmware only. Table 28-265.BMC Management Packets Received Count--BMNGPRC [0:3] (0x413C; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 BMNGPRC 28.19.1.4 Bus:Device:Function: Offset Start: 413Ch Offset End: 413Fh Power Well: GBEAUX Bit Description Sticky Number of management packets received. Bit Reset Value Bit Access 0x0 RC BMC Total Unicast Packets Received--BUPRC [0:3] (0x4400; RC) This register counts the number of good (no errors) unicast packets received. This register does not count unicast packets received that fail to pass address filtering. This register only increments if receives are enabled (RCTL.RXEN is set). This register does not count packets counted by the Missed Packet Count (MPC) register. Packets sent to the manageability engine are included in this counter. This register is available to the firmware only. Table 28-266.BMC Total Unicast Packets Received--BUPRC [0:3] (0x4400; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 BUPRC October 2012 Order Number: 327879-001US Offset Start: 4400h Offset End: 4403h Power Well: GBEAUX Bit Description Number of Unicast packets received. Sticky Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1453 28.19.1.5 BMC Total Multicast Packets Received--BMPRC [0:3] (0x4404; RC) This register counts the same events as the MPRC register (Section 28.18.1.26) for the BMC usage.This register is available to the firmware only. Table 28-267.BMC Total Multicast Packets Received--BMPRC [0:3] (0x4404; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 BMPRC 28.19.1.6 Bus:Device:Function: Offset Start: 4404h Offset End: 4407h Power Well: GBEAUX Bit Description Sticky Number of multicast packets received. Bit Reset Value Bit Access 0x0 RC BMC Total Broadcast Packets Received--BBPRC [0:3] (0x4408; RC) This register counts the same events as the BPRC register (Section 28.18.1.25) for the BMC usage.This register is available to the firmware only. Table 28-268.BMC Total Broadcast Packets Received--BBPRC [0:3] (0x4408; RC) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :00 BBPRC 28.19.1.7 Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Offset Start: 4408h Offset End: 440Bh Power Well: GBEAUX Bit Description Sticky Number of broadcast packets received. Bit Reset Value Bit Access 0x0 RC BMC Total Unicast Packets Transmitted--BUPTC [0:3] (0x440C; RC) This register counts the number of unicast packets transmitted. This register increments only if transmits are enabled (TCTL.EN is set).This register is available to the firmware only Table 28-269.BMC Total Unicast Packets Transmitted--BUPTC [0:3] (0x440C; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 BUPTC Power Well: GBEAUX Bit Description Number of unicast packets transmitted. Intel(R) Communications Chipset 89xx Series - Datasheet 1454 Offset Start: 440Ch Offset End: 440Fh Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.19.1.8 BMC Total Multicast Packets Transmitted--BMPTC [0:3] (0x4410; RC) This register counts the same events as the MPTC register (Section 28.18.1.52) for the BMC usage.This register is available to the firmware only. Table 28-270.BMC Total Multicast Packets Transmitted--BMPTC [0:3] (0x4410; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 BMPTC 28.19.1.9 Bus:Device:Function: Offset Start: 4410h Offset End: 4413h Power Well: GBEAUX Bit Description Sticky Number of multicast packets transmitted. Bit Reset Value Bit Access 0x0 RC BMC Total Broadcast Packets Transmitted--BBPTC [0:3] (0x4414; RC) This register counts the same events as the BPTC register (Section 28.18.1.53) for the BMC usage.This register is available to the firmware only. Table 28-271.BMC Total Broadcast Packets Transmitted--BBPTC [0:3] (0x4414; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 BBPTC October 2012 Order Number: 327879-001US Offset Start: 4414h Offset End: 4417h Power Well: GBEAUX Bit Description Sticky Number of broadcast packets transmitted count. Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1455 28.19.1.10 BMC FCS Receive Errors--BCRCERRS [0:3] (0x4418; RC) This register counts the same events as the CRCERRS register (Section 28.18.1.1) for the BMC usage.This register is available to the firmware only. Table 28-272.BMC FCS Receive Errors--BCRCERRS [0:3] (0x4418; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 BCRCERRS Offset Start: 4418h Offset End: 441Bh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RC CRC error count 28.19.1.11 BMC Alignment Errors--BALGNERRC [0:3] (0x441C; RC) This register counts the same events as the ALGNERRC register (Section 28.18.1.2) for the BMC usage.This register is available to the firmware only. Table 28-273.BMC Alignment Errors--BALGNERRC [0:3] (0x441C; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 AEC Offset Start: 441Ch Offset End: 441Fh Power Well: GBEAUX Bit Description Sticky Alignment error count Bit Reset Value Bit Access 0x0 RC 28.19.1.12 BMC Pause XON Frames Received--BXONRXC [0:3] (0x4420; RC) This register counts the same events as the XONRXC register (Section 28.18.1.13) for the BMC usage.This register is available to the firmware only. Table 28-274.BMC Pause XON Frames Received--BXONRXC [0:3] (0x4420; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 BXONRXC Power Well: GBEAUX Bit Description Number of XON packets received. Intel(R) Communications Chipset 89xx Series - Datasheet 1456 Offset Start: 4420h Offset End: 4423h Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.19.1.13 BMC Pause XOFF Frames Received--BXOFFRXC [0:3] (0x4424; RC) This register counts the same events as the XOFFRXC register (Section 28.18.1.15) for the BMC usage.This register is available to the firmware only. Table 28-275.BMC Pause XOFF Frames Received--BXOFFRXC [0:3] (0x4424; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 BXOFFRXC Offset Start: 4424h Offset End: 4427h Power Well: GBEAUX Bit Description Sticky Number of XOFF packets received. Bit Reset Value Bit Access 0x0 RC 28.19.1.14 BMC Pause XON Frames Transmitted--BXONTXC [0:3] (0x4428; RC) This register counts the same events as the XONTXC register (Section 28.18.1.14) for the BMC usage.This register is available to the firmware only. Table 28-276.BMC Pause XON Frames Transmitted--BXONTXC [0:3] (0x4428; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 BXONTXC October 2012 Order Number: 327879-001US Offset Start: 4428h Offset End: 442Bh Power Well: GBEAUX Bit Description Number of XON packets transmitted. Sticky Bit Reset Value Bit Access 0x0 RC Intel(R) Communications Chipset 89xx Series - Datasheet 1457 28.19.1.15 BMC Pause XOFF Frames Transmitted--BXOFFTXC [0:3] (0x442C; RC) This register counts the same events as the XOFFTXC register (Section 28.18.1.16) for the BMC usage.This register is available to the firmware only. Table 28-277.BMC Pause XOFF Frames Transmitted--BXOFFTXC [0:3] (0x442C; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 BXOFFTXC Offset Start: 442Ch Offset End: 442Fh Power Well: GBEAUX Bit Description Sticky Number of XOFF packets transmitted. Bit Reset Value Bit Access 0x0 RC 28.19.1.16 BMC Single Collision Transmit Frames--BSCC [0:3] (0x4430; RC) This register counts the same events as the SCC register (Section 28.18.1.5) for the BMC usage.This register is available to the firmware only. Table 28-278.BMC Single Collision Transmit Frames--BSCC [0:3] (0x4430; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Offset Start: 4430h Offset End: 4433h Power Well: GBEAUX Bit Range Bit Acronym Bit Description 31 :00 BSCC Number of times a transmit encountered a single collision. Sticky Bit Reset Value Bit Access 0x0 RC 28.19.1.17 BMC Multiple Collision Transmit Frames--BMCC [0:3] (0x4434; RC) This register counts the same events as the MCC register (Section 28.18.1.7) for the BMC usage.This register is available to the firmware only. Table 28-279.BMC Multiple Collision Transmit Frames--BMCC [0:3] (0x4434; RC) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 BMCC Power Well: GBEAUX Bit Description Number of times a successful transmit encountered multiple collisions. Intel(R) Communications Chipset 89xx Series - Datasheet 1458 Offset Start: 4434h Offset End: 4437h Sticky Bit Reset Value Bit Access 0x0 RC October 2012 Order Number: 327879-001US 28.0 28.20 Wake Up Controls Registers 28.20.1 Detailed Register Descriptions 28.20.1.1 Wakeup Control Register--WUC [0:3] (0x5800; R/W) The PME_En and PME_Status bits of this register are reset when LAN_PWR_GOOD is 0b. When AUX_PWR = 0b, this register is also reset by de-asserting PE_RST_N and during a D3 to D0 transition. The other bits are reset using the standard internal resets. Table 28-280.Wakeup Control Register--WUC [0:3] (0x5800; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Bus:Device:Function: Index1 Default: 0x0 Power Well: GBEAUX Bit Reset Value Bit Access 0h1 R/W 0h R/W PME_En PME_En This read/write bit is used by the software device driver to enable generation of a PME event without writing to the Power Management Control / Status Register (PMCSR) in the PCIe configuration space. Note: Bit reflects value of PMCSR.PME_En bit when the bit in the PMCSR register is modified. However when value of WUC.PME_En bit is modified by software device driver, value is not reflected in the PMCSR.PME_En bit. 0h R/W APME Advance Power Management Enable If set to 1b, APM Wakeup is enabled. If this bit is set and the APMPME bit is cleared, reception of a magic packet asserts the WUS.MAG bit but does not assert a PME. 0ha R/W Bit Range Bit Acronym 31 :04 Reserved Reserved APMPME Assert PME On APM Wakeup If set to 1b, the Controller sets the PME_Status bit in the Power Management Control / Status Register (PMCSR) and asserts PE_WAKE_N and sends a PM_PME PCIe message when APM Wakeup is enabled (WUC.APME = 1) and the Controller receives a matching Magic Packet. Note: When WUC.APMPME is set PE_WAKE_N is asserted and a PM_PME message is sent even if PMCSR.PME_En is cleared. 03 02 01 00 Offset Start: 5800h Offset End: 5803h Bit Description Sticky PME_Status PME_Status (R/ This bit is set when the Controller receives a wakeup event. It is the same as the PME_Status bit in the Power W1C) Management Control / Status Register (PMCSR). Writing a 1b to this bit clears also the PME_Status bit in the PMCSR. a. Loaded from the EEPROM. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1459 28.20.1.2 Wakeup Filter Control Register--WUFC [0:3] (0x5808; R/W) This register is used to enable each of the pre-defined and flexible filters for wakeup support. A value of 1b means the filter is turned on.; A value of 0b means the filter is turned off. If the NoTCO bit is set, then any packet that passes the manageability packet filtering described in Section 27.4, "Manageability Receive Filtering", does not cause a Wake Up event even if it passes one of the Wake Up Filters. This bit is set at initialization and during any EEPROM read if the SMBus Enable bit of the EEPROM's Management Control word is 1b. Otherwise its initial value is 0b. Table 28-281.Wakeup Filter Control Register--WUFC [0:3] (0x5808; R/W) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :24 Reserved Offset Start: 5808h Offset End: 580Bh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved 23 FLX7 Flexible Filter 7 Enable. 0h R/W 22 FLX6 Flexible Filter 6 Enable. 0h R/W 21 FLX5 Flexible Filter 5 Enable. 0h R/W 20 FLX4 Flexible Filter 4 Enable. 0h R/W 19 FLX3 Flexible Filter 3 Enable. 0h R/W 18 FLX2 Flexible Filter 2 Enable. 0h R/W 17 FLX1 Flexible Filter 1 Enable. 0h R/W 16 FLX0 Flexible Filter 0 Enable. 0h R/W 15 NoTCO Ignore TCO/management packets for wake up. 0h R/W FLEX_HQ Flex filters Host Queuing 0 - Do not use Flex filters for queueing decisions in D0 state. 1 - Use flex filters in queuing decisions in D0 state. Note: Should be enabled only when multi queueing is enabled (MRQC.Multiple Receive Queues = 010b or 000b). 0h R/W Reserved Reserved. Set these bits to 0b. 14 13 :08 07 IPv6 Directed IPv6 Packet Wakeup Enable. 0h R/W 06 IPv4 Directed IPv4 Packet Wakeup Enable. 0h R/W 05 ARP ARP Request Packet Wakeup Enable. 0h R/W 04 BC Broadcast Wakeup Enable. 0h R/W 03 MC Directed Multicast Wakeup Enable. 0h R/W 02 EX Directed Exact Wakeup Enable.a 0h R/W Magic Packet Wakeup Enable. 0h R/W 01 MAG 00 Reserved Reserved a. If the RCTL.UPE is set, and the EX bit is set also, any unicast packet wakes up the system. Intel(R) Communications Chipset 89xx Series - Datasheet 1460 October 2012 Order Number: 327879-001US 28.0 28.20.1.3 Wakeup Status Register--WUS [0:3] (0x5810; R/W1C) This register is used to record statistics about all wakeup packets received. If a packet matches multiple criteria then multiple bits could be set. Writing a 1b to any bit clears that bit. This register is not cleared when RST# is asserted. It is only cleared when LAN_PWR_GOOD is de-asserted or when cleared by the software device driver. Note: If additional packets are received that matches one of the wakeup filters, after the original wakeup packet is received, the WUS register is updated with the matching filters accordingly. Table 28-282.Wakeup Status Register--WUS [0:3] (0x5810; R/W1C) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :24 Reserved Offset Start: 5810h Offset End: 5813h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved 23 FLX7 Flexible Filter 7 Matcha. 0h R/W1C 22 FLX6 Flexible Filter 6 Matcha. 0h R/W1C a 21 FLX5 Flexible Filter 5 Match . 0h R/W1C 20 FLX4 Flexible Filter 4 Matcha. 0h R/W1C a 19 FLX3 Flexible Filter 3 Match . 0h R/W1C 18 FLX2 Flexible Filter 2 Matcha. 0h R/W1C 17 FLX1 Flexible Filter 1 Matcha. 0h R/W1C 16 FLX0 Flexible Filter 0 Matcha. 0h R/W1C 0h R/W1C 15 :09 Reserved Reserved 08 MNG Indicates that a manageability event that should cause a PME happened. 07 IPv6 Directed IPv6 Packet Received. 0h R/W1C 06 IPv4 Directed IPv4 Packet Received. 0h R/W1C 05 ARP ARP Request Packet Received. 0h R/W1C 04 BC Broadcast Packet Received. 0h R/W1C 03 MC Directed Multicast Packet Received The packet was a multicast packet hashed to a value that corresponded to a 1 bit in the Multicast Table Array. 0h R/W1C 02 EX Directed Exact Packet Received The packet's address matched one of the 16 preprogrammed exact values in the Receive Address registers. 0h R/W1C 01 MAG Magic Packet Received. 0h R/W1C 00 Reserved Reserved a. Bit is set only when flex filter match is detected and WUFC.FLEX_HQ is 0. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1461 28.20.1.4 Wakeup Packet Length--WUPL [0:3] (0x5900; RO) This register indicates the length of the first wakeup packet received. It is valid if one of the bits in the Wakeup Status register (WUS) is set. It is not cleared by any reset. Table 28-283.Wakeup Packet Length--WUPL [0:3] (0x5900; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :12 Reserved 11 :00 LEN 28.20.1.5 Bus:Device:Function: Offset Start: 5900h Offset End: 5903h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access X RO Reserved Length of wakeup packet. (If jumbo frames are enabled and the packet is longer than 2047 bytes then this field is 2047.) Wakeup Packet Memory--WUPM [0:3][0:31] (0x5A00 + 4*n [n=0...31]; RO) This register is read-only and it is used to store the first 128 bytes of the wakeup packet for software retrieval after system wakeup. It is not cleared by any reset. Table 28-284.Wakeup Packet Memory--WUPM [0:3][0:31] (0x5A00 + 4*n [n=0...31]; RO) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 WUPD Power Well: GBEAUX Bit Description Wakeup Packet Data Intel(R) Communications Chipset 89xx Series - Datasheet 1462 Offset Start: 5A00h at 4 Offset End: 5A03h at 4 Sticky Bit Reset Value Bit Access X RO October 2012 Order Number: 327879-001US 28.0 28.20.1.6 IP Address Valid--IPAV [0:3] (0x5838; R/W) The IP Address Valid indicates whether the IP addresses in the IP Address Table are valid. Table 28-285.IP Address Valid--IPAV [0:3] (0x5838; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bit Acronym 31 :17 Reserved 16 B:0:1+ Index1 Default: 0x0 Bit Range 15 :04 Bus:Device:Function: V60 Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0h R/W Reserved IPv6 Address 0 Valid. Reserved Offset Start: 5838h Offset End: 583Bh Reserved 03 V43 IPv4 Address 3 Valid. 0h R/W 02 V42 IPv4 Address 2 Valid. 0h R/W 01 V41 IPv4 Address 1 Valid. 0h R/W 00 V40 IPv4 Address 0 Valid. 0h R/W 28.20.1.7 IPv4 Address Table--IP4AT [0:3][0:3] (0x5840 + 8*n [n=0...3]; R/ W) The IPv4 Address Table is used to store the four IPv4 addresses for the ARP/IPv4 Request packet and Directed IP packet wakeup. Note: This table is not cleared by any reset. Table 28-286.IPv4 Address Table--IP4AT [0:3][0:3] (0x5840 + 8*n [n=0...3]; R/W) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] B:0:1+ Bus:Device:Function: Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 IP Address October 2012 Order Number: 327879-001US Offset Start: 5840h at 8 Offset End: 5843h at 8 Power Well: GBEAUX Bit Description Sticky IPv4 Address n Note: These registers are written in Big Endian order (LS byte is first on the wire and is the MS byte of the IPV4 Address). Bit Reset Value Bit Access X R/W Intel(R) Communications Chipset 89xx Series - Datasheet 1463 28.20.1.8 IPv6 Address Table--IP6AT [0:3][0:3] (0x5880 + 4*n [n=0...3]; R/ W) The IPv6 Address Table is used to store the IPv6 addresses for neighbor Discovery packet filtering and Directed IP packet wakeup. Note: This table is not cleared by any reset. Table 28-287.IPv6 Address Table--IP6AT [0:3][0:3] (0x5880 + 4*n [n=0...3]; R/W) Description: View: PCI BAR: GBEPCIBAR0[0:3] Size: 32 bit Bus:Device:Function: Default: 0x0 Bit Acronym Bit Description 31 :00 IP Address IPv6 Address bytes 4*n+1:4*n +4 Note: These registers appear in Big Endian order (LS byte, LS address is first on the wire and is the MS byte of the IPV6 Address). Field Dword # 28.20.2 Offset Start: 5880h at 4 Offset End: 5883h at 4 Power Well: GBEAUX Bit Range IPV6ADDR0 B:0:1+ Index1 Address Bit(s) Initial Value Sticky Bit Reset Value Bit Access 0x0 R/W Description 0 0x5880 31:0 X IPv6 Address 0, bytes 1-4 1 0x5884 31:0 X IPv6 Address 0, bytes 5-8 2 0x5888 31:0 X IPv6 Address 0, bytes 9-12 3 0x588C 31:0 X IPv6 Address 0, bytes 16-13 Flexible Host Filter Table Registers--FHFT [0:3] (0x9000 0x93FC; RW) Each of the 4 Flexible Host Filters Table registers (FHFT) contains a 128 byte pattern and a corresponding 128-bit mask array. If enabled, the first 128 bytes of the received packet are compared against the non-masked bytes in the FHFT register. Each 128 byte filter is composed of 32 DW entries, where each 2 DWs are accompanied by an 8-bit mask, one bit per filter byte. When a bit in the 8-bit mask field is set the corresponding Byte in the filter is compared. The 8 lsb bits of the last DW of each filter contains a length field defining the number of bytes from the beginning of the packet compared by this filter, the length field should be 8 bytes aligned value. If actual packet length is less than (length - 8) (length is the value specified by the length field), the filter fails. Otherwise, it depends on the result of actual byte comparison. The value should not be greater than 128. Note: The length field must be 8 bytes aligned. For filtering packets shorter than 8 bytes aligned the values should be rounded up to the next 8 bytes aligned value, the hardware implementation compares 8 bytes at a time so it should get extra zero masks (if needed) until the end of the length value. Intel(R) Communications Chipset 89xx Series - Datasheet 1464 October 2012 Order Number: 327879-001US 28.0 Bits 31:8 of the last DW of each filter also includes a Queueing field. When the Controller is in D0 state, the WUFC.FLEX_HQ bit is set to 1, MRQC.Multiple Receive Queues = 010b or 000b and the packet matches the flex filter, the Queueing field defines the receive queue for the packet, priority of the filter and actions to be initiated. 31 0 31 8 7 0 31 0 31 0 Reserved Reserved Mask [7:0] DW 1 DW 0 Reserved Reserved Mask [15:8] DW 3 DW 2 Reserved Reserved Mask [23:16] DW 5 DW 4 Reserved Reserved Mask [31:24] DW 7 DW 6 .... 31 8 7 0 31 8 7 0 31 0 31 0 Reserved Reserved Reserved Mask [127:120] DW 29 DW 28 Queueing Length Reserved Mask [127:120] DW 31 DW 30 Accessing the FHFT registers during filter operation can result in a packet being misclassified if the write operation collides with packet reception. It is therefore advised that the flex filters are disabled prior to changing their setup. 28.20.2.1 Flex Filter Even Data Register Fields--FEDR[0:3][0:3][0:15](0x9000 + 0x10*n[n=0..15]) Table 28-288.Flex Filter Even Data Register Fields--FEDR [0:3][0:3][0:15] (0x9000 +16*n[0..15]; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :00 FEDR October 2012 Order Number: 327879-001US 9000h at 100h at / Offset Start: 10h Offset End: 9003h at 100h at / 10h Power Well: GBEAUX Bit Description Flex Filter Even Data Register Fields Sticky Bit Reset Value Bit Access X RW Intel(R) Communications Chipset 89xx Series - Datasheet 1465 28.20.2.2 Flex Filter Odd Data Register Fields--FODR[0:3][0:3][0:15] (0x9004 + 16*n[n=0..15]) Table 28-289.Flex Filter Odd Data Register Fields--FODR [0:3][0:3][0:15] (0x9000 +16*n[0..15];RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 FODR 28.20.2.3 B:0:1+ Bus:Device:Function: Index1 9004h at 100h at / Offset Start: 10h Offset End: 9007h at 100h at / 10h Power Well: GBEAUX Bit Description Sticky Flex Filter Even Data Register Fields Bit Reset Value Bit Access X RW Flex Filter Mask Fields--FMFR[0:3][0:3][0:15] (0x9008 + 16*n[n=0..15]) Table 28-290.Flex Filter Mask Field Register--FMFR [0:3][0:3][0:15] (0x9008 +16*n[0..15];RW) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 FMFR Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access X RW Reserved Flex Filter Mask Field Register Intel(R) Communications Chipset 89xx Series - Datasheet 1466 9008h at 100h at / Offset Start: 10h Offset End: 900Ch at 100h at / 10h October 2012 Order Number: 327879-001US 28.0 28.20.2.4 Flex Filter Queueing Field--FQFR[0:3][0:3][0:15] (0x90FC + 16*n[n=0..15]) Queueing field resides in bits 31:8 of last DW (DW 63) of flex filter. The queueing field defines the receive queue to forward the packet (RQUEUE), the filter priority (FLEX_PRIO) and additional filter actions. Operations defined in queueing field are enabled when the Controller is in D0 state, MRQC.Multiple Receive Queues = 010b or 000b, WUFC.FLEX_HQ is 1 and relevant WUFC.FLX[n] bit is set. Table 28-291.Flex Filter Queueing Field--FQFR [0:3][0:3][0:15] (0x90FC + 16*n[n=0..15] ;RW) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :25 Reserved 24 23 :19 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Immediate Interrupt Reserved 90FCh at Offset Start: 100h Offset End: 90FFh at 100h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access X RW X RW Reserved Write 0, ignore on read. Enables issuing an immediate interrupt when the Flex filter matches incoming packet. Reserved Write 0, ignore on read. Flex filter Priority Defines the priority of the filter assuming two filters with same priority don't match. If two filters with the same priority match the incoming packet, the first filter (lowest address) is used in order to define the queue destination of this packet. 18 :16 FLEX_PRIO 15 :11 Reserved Reserved Write 0, ignore on read. 10 :08 RQUEUE Receive Queue Defines receive queue associated with this Flex filter. When match occurs in D0 state, packet is forwarded to the receive queue. X RW 07 :00 Length Length Filter Length in bytes. Should be 8 bytes aligned and not greater then 128 bytes. X RW October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1467 28.20.2.5 Flex Filter 0--Example Field Dword Filter 0 DW0 0 Address 0x9000 Bit(s) Initial Value 31:0 X Filter 0 DW1 1 0x9004 31:0 X Filter 0 Mask[7:0] 2 0x9008 7:0 X Reserved 3 0x900C 31:0 X Filter 0 DW2 4 0x9010 31:0 X Filter 0 DW30 60 0x90F0 31:0 X ... Filter 0 DW31 61 0x90F4 31:0 X Filter 0 Mask[127:120] 62 0x90F8 7:0 X Length 63 0x90FC 7:0 X Filter 0 Queueing 63 0x90FC 31:8 X 28.20.3 Flexible Host Filter Table Extended Registers--FHFT_EXT [0:3] (0x9A00 - 0x9DFC; RW) Each of the 4 additional Flexible Host Filters Table extended registers (FHFT_EXT) contains a 128 Byte pattern and a corresponding 128-bit mask array. If enabled, the first 128 Bytes of the received packet are compared against the non-masked bytes in the FHFT_EXT register. The layout and access rules of this table are the same as in FHFT. 28.20.3.1 Flex Filter Even Data Register Extended--FHFT_EXT_FEDR (0x9A00 + 0x10*n[n=0..15]) Table 28-292.Flex Filter Even Data Register Extended--FHFT_EXT_FEDR [0:3][0:3][0:15] (0x9A00 +16*n[0..15]; RW) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 FHFT_FEDR Power Well: GBEAUX Bit Description Flex Filter Even Data Register Fields Intel(R) Communications Chipset 89xx Series - Datasheet 1468 9A00h at 100h at / Offset Start: 10h Offset End: 9A03h at 100h at / 10h Sticky Bit Reset Value Bit Access X RW October 2012 Order Number: 327879-001US 28.0 28.20.3.2 Flex Filter Odd Data Register Fields--FHFT_EXT_FODR [0:3][0:3][0:15] (0x9A04 + 16*n[n=0..15]) Table 28-293.Flex Filter Odd Data Register Extended--FHFT_EXT_FODR [0:3][0:3][0:15] (0x9A00 +16*n[0..15];RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :00 FHFT_FODR 28.20.3.3 B:0:1+ Bus:Device:Function: Index1 9A04h at 100h at / Offset Start: 10h Offset End: 9A07h at 100h at / 10h Power Well: GBEAUX Bit Description Sticky Flex Filter Even Data Register Fields Bit Reset Value Bit Access X RW Flex Filter Mask Fields--FHFT_EXT_FMFR [0:3][0:3][0:15] (0x9A08 +16*n[0..15];RW) Table 28-294.Flex Filter Mask Field Extended--FHFT_EXT_FMFR [0:3][0:3][0:15] (0x9A08 +16*n[0..15];RW) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 FHFT_FMFR October 2012 Order Number: 327879-001US 9A08h at 100h at / Offset Start: 10h Offset End: 9A0Ch at 100h at / 10h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access X RW Reserved Flex Filter Mask Field Register Intel(R) Communications Chipset 89xx Series - Datasheet 1469 28.20.3.4 Flex Filter Queueing Field--FHFT_EXT_FQFR [0:3][0:3] (0x9AFC;RW) Queueing field resides in bits 31:8 of last DW (DW 63) of flex filter. The queueing field defines the receive queue to forward the packet (RQUEUE), the filter priority (FLEX_PRIO) and additional filter actions. Operations defined in queueing field are enabled when the Controller is in D0 state, MRQC.Multiple Receive Queues = 010b or 000b, WUFC.FLEX_HQ is 1 and relevant WUFC.FLX[n] bit is set. Table 28-295.Flex Filter Queueing Extended--FHFT_EXT_FQFR [0:3][0:3] (0x9AFC;RW) Description: View: PCI Size: 32 bit Default: 0x0 Bit Range Bit Acronym 31 :25 Reserved 24 23 :19 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Immediate Interrupt Reserved Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access X RW X RW Reserved Write 0, ignore on read. Enables issuing an immediate interrupt when the Flex filter matches incoming packet. Reserved Write 0, ignore on read. Flex filter Priority Defines the priority of the filter assuming two filters with same priority don't match. If two filters with the same priority match the incoming packet, the first filter (lowest address) is used in order to define the queue destination of this packet. 18 :16 FLEX_PRIO 15 :11 Reserved Reserved Write 0, ignore on read. 10 :08 RQUEUE Receive Queue Defines receive queue associated with this Flex filter. When match occurs in D0 state, packet is forwarded to the receive queue. X RW 07 :00 Length Length Filter Length in bytes. Should be 8 bytes aligned and not greater then 128 bytes. X RW Intel(R) Communications Chipset 89xx Series - Datasheet 1470 9AFCh at Offset Start: 100h Offset End: 9AFFh at 100h October 2012 Order Number: 327879-001US 28.0 28.21 Management Registers All management registers are controlled by the remote BMC for both read and write. Host accesses to the management registers are blocked for write. The attributes for the fields in this chapter refer to the BMC access rights. 28.21.1 Detailed Register Descriptions 28.21.1.1 Management VLAN TAG Value--MAVTV [0:3][0:7] (0x5010 +4*n [n=0...7]; RW) Where "n" is the VLAN filter serial number, equal to 0,1,...7. Table 28-296.Management VLAN TAG Value--MAVTV [0:3] [0:7] (0x5010 +4*n [n=0...7]; RW) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :12 Reserved 11 :00 VID Offset Start: 5010h at 4 Offset End: 5013h at 4 Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RW Reserved Write 0, ignore on read Contains the VLAN ID that should be compared with the incoming packet if the corresponding bit in MDEF is set. The MAVTV registers are written by the BMC and are not accessible to the host for writing. The registers are used to filter manageability packets as described in the Management chapter. 28.21.1.2 Management Flex UDP/TCP Ports--MFUTP[0:3][0:3] (0x5030 + 4*n [n=0...3]; RW) Where each 32-bit register (n=0,...,3) refers to two UDP/TCP port filters (register at address offset n=0 refers to UDP/TCP ports 0 and 1, register at address offset n=1 refers to UDP/TCP ports 2 and 3, etc.). Table 28-297.Management Flex UDP/TCP Ports--MFUTP [0:3][0:3] (0x5030 + 4*n [n=0...3]; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :16 MFUTP_odd 15 :00 MFUTP_even October 2012 Order Number: 327879-001US Offset Start: 5030h at 4 Offset End: 5033h at 4 Power Well: GBEAUX Bit Reset Value Bit Access 2*n+1 Management Flex UDP/TCP port 0x0 RW 2*n Management Flex UDP/TCP port 0x0 RW Bit Description Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 1471 The MFUTP registers are written by the BMC and are not accessible to the host for writing. The registers are used to filter manageability packets. See section 27.4 . Reset - The MFUTP registers are cleared on LAN_PWR_GOOD only. The initial values for this register can be loaded from the EEPROM after power-up reset. Note: The MFUTP_even and MFUTP_odd fields should be written in network order. 28.21.1.3 Management Ethernet Type Filters--METF [0:3][0:3] (0x5060 + 4*n [n=0...3]; RW) Table 28-298.Management Ethernet Type Filters--METF [0:3][0:3] (0x5060 + 4*n [n=0...3]; RW) Description: View: PCI Size: 32 bit Bit Range 31 30 B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Acronym Reserved Polarity 29 :16 Reserved 15 :00 METF Offset Start: 5060h at 4 Offset End: 5063h at 4 Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0h RW 0x0 RW Reserved 0b = Positive filter - forward packets matching this filter to the manageability block. 1b = Negative filter - block packets matching this filter from the manageability block. Reserved EtherType value to be compared against the L2 EtherType field in the Rx packet. The METF registers are written by the BMC and are not accessible to the host for writing. The registers are used to filter manageability packets. See section 27.4 . Reset - The METF registers are cleared on LAN_PWR_GOOD only. The initial values for this register might be loaded from the EEPROM after power-up reset. 28.21.1.4 Management Control Register--MANC [0:3] (0x5820; RW) Table 28-299.Management Control Register--MANC [0:3] (0x5820; RW) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Power Well: GBEAUX Bit Range Bit Acronym 31 :27 Reserved Reserved NET_TYPE NET TYPE: 0b = pass only un-tagged packets. 1b = pass only VLAN tagged packets. Valid only if FIXED_NET_TYPE is set. 26 Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 1472 Offset Start: 5820h Offset End: 5823h Sticky Bit Reset Value Bit Access 0h RW October 2012 Order Number: 327879-001US 28.0 Table 28-299.Management Control Register--MANC [0:3] (0x5820; RW) (Sheet 2 of 2) Description: View: PCI Size: 32 bit Bit Range BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Acronym Offset Start: 5820h Offset End: 5823h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 25 Fixed net type: If set, only packets matching the net type FIXED_NET_TYP defined by the NET_TYPE field passes to manageability. E Otherwise, both tagged and un-tagged packets can be forwarded to the manageability engine 0h RW 24 Enable IPv4 address Filters - when set, the last 128 bits of EN_IPv4_FILTE the MIPAF register are used to store 4 IPv4 addresses for R IPv4 filtering. When cleared, these bits store a single IPv6 filter. 0h RW 23 Enable IPv4 address Filters - when set, the last 128 bits of EN_XSUM_FILT the MIPAF register are used to store 4 IPv4 addresses for ER IPv4 filtering. When cleared, these bits store a single IPv6 filter. 0h RW 0h RW 0h RW 0h RW 0h RW 22 :20 19 18 :17 16 Reserved Reserved. Write 0 ignore on read. RCV_ALL Receive All Enable (promiscuous mode) When set, all the traffic received from the wire is forwarded to the manageability. This bit should be set only for debug purposes. Reserved TCO_RESET 15 TCO_Isolate (RO) 14 FW_RESET 13 :00 Reserved October 2012 Order Number: 327879-001US Reserved TCO Reset occurred. Set to 1b on a TCO Reset, to reset LAN port by BMC. Set to 1 on a TCO Isolate command. When the "TCO_Isolate" bit is set. Host write cycles are completed successfully on the PCIe but silently ignored by internal logic. When FW initiates the TCO Isolate command it also initiates a FW interrupt via the ICR.MNG bit to the host and writes a value of 0x0E to the FWSM.Ext_Err_Ind field. This bit is Read Only and mirrors the value of the Isolate bit in the internal Management registers. FW Reset occurred. Set to 1b on a TCO Firmware Reset. Cleared following a BMC read operation. Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1473 28.21.1.5 Management Only Traffic Register--MNGONLY [0:3] (0x5864; RW) The MNGONLY register allows exclusive filtering of certain type of traffic to the BMC. Exclusive filtering enables the BMC to define certain packets that are forwarded to the BMC but not to the host. The packets will not be forwarded to the host even if they pass the host L2 filtering process. Each manageability decision filter (MDEF and MDEF_EXT) has a corresponding bit in the MNGONLY register. When a manageability decision filter (MDEF and MDEF_EXT) forwards a packet to manageability, it may also block the packet from being forwarded to the host if the corresponding MNGONLY bit is set. Table 28-300.Management Only Traffic Register--MNGONLY [0:3] (0x5864; RW) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :08 Reserved 07 :00 Exclusive to MNG Offset Start: 5864h Offset End: 5867h Power Well: GBEAUX Bit Description Sticky Bit Reset Valuea Bit Access 0x0 RW Reserved Exclusive to MNG - when set, indicates that packets forwarded by the manageability filters to manageability are not sent to the host. Bits 0...7 correspond to decision rules defined in registers MDEF[0...7] and MDEF_EXT[0...7]. a. Reset - The MNGONLY register is cleared on LAN_PWR_GOOD and firmware reset. The initial values for this register can be loaded from the EEPROM after power-up reset or firmware reset. 28.21.1.6 Manageability Decision Filters--MDEF [0:3][0:7] (0x5890 + 4*n [n=0...7]; RW) Where "n" is the decision filter. Table 28-301.Manageability Decision Filters--MDEF [0:3][0:7] (0x5890 + 4*n [n=0...7]; RW) (Sheet 1 of 2) Description: View: PCI Size: 32 bit Bit Range Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Acronym Power Well: GBEAUX Bit Description Sticky Bit Reset Valuea Bit Access 31 Port 0x26F Port 0x26F - Controls the inclusion of Port 0x26F filtering in the manageability filter decision (OR section). 0h RW 30 Port 0x298 Port 0x298 - Controls the inclusion of Port 0x298 filtering in the manageability filter decision (OR section). 0h RW 29 Neighbor Discovery neighbor Discovery - Controls the inclusion of neighbor Discovery filtering in the manageability filter decision (OR section). The neighbor types accepted by this filter are types 0x86, 0x87, 0x88 and 0x89. 0h RW 0h RW 28 ARP Response - Controls the inclusion of ARP Response ARP Response filtering in the manageability filter decision (OR section). Intel(R) Communications Chipset 89xx Series - Datasheet 1474 Offset Start: 5890h at 4 Offset End: 5893h at 4 October 2012 Order Number: 327879-001US 28.0 Table 28-301.Manageability Decision Filters--MDEF [0:3][0:7] (0x5890 + 4*n [n=0...7]; RW) (Sheet 2 of 2) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 5890h at 4 Offset End: 5893h at 4 Power Well: GBEAUX Bit Reset Valuea Bit Access 0h RW 0h RW 0h RW Unicast OR Unicast - Controls the inclusion of unicast address 0 to 1 respectively in the manageability filter decision (OR section). 0x0 RW IPv6 Address IPv6 Address - Controls the inclusion of IPV6 address 0 o 3 respectively in the manageability filter decision (AND section). Bit 17 corresponds to IPV6 address 0, etc Note: This field is relevant only if MANC.EN_IPv4_FILTER is cleared. 0h RW 16 :13 IPv4 Address IPv4 Address - Controls the inclusion of IPV4 address 0 to 3 respectively in the manageability filter decision (AND section). Bit 13 corresponds to IPV4 address 0, etc. Note: This field is relevant only if MANC.EN_IPv4_FILTER is set. 0x0 RW 12 :05 VLAN AND VLAN - Controls the inclusion of VLAN tag 0 to 7 respectively in the manageability filter decision (AND section). Bit 5 corresponds to VLAN tag 0, etc. 0x0 RW 0h RW 0x0 RW Bit Range 27 Bit Acronym ARP Request Bit Description Sticky ARP Request - Controls the inclusion of ARP Request filtering in the manageability filter decision (OR section). 26 Multicast - Controls the inclusion of Multicast address filtering in the manageability filter decision (AND section). Broadcast packets are not included by this bit. The packet Multicast AND must pass some L2 filtering to be included by this bit - either by the MANC.MCST_PASS_L2 or by some dedicated MAC address. 25 Broadcast OR 24 :23 22 :21 20 :17 04 Reserved Broadcast - Controls the inclusion of broadcast address filtering in the manageability filter decision (OR section). Reserved Broadcast - Controls the inclusion of broadcast address Broadcast AND filtering in the manageability filter decision (AND section). 03 :02 Reserved 01 :00 Unicast AND Reserved Unicast - Controls the inclusion of unicast address 0 to 1 in the manageability filter decision (AND section). Bit 0 corresponds to unicast address 0, etc. a. Default values are read from EEPROM. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1475 28.21.1.7 Manageability Decision Filters--MDEF_EXT [0:3][0:7] (0x5930 + 4*n[n=0...7]; RW) Table 28-302.Manageability Decision Filters--MDEF_EXT [0:3][0:7] (0x5930 + 4*n[n=0...7]; RW) Description: View: PCI Size: 32 bit Bit Range 31 30 29 28 :25 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x20000000 Bit Acronym Reserved Power Well: GBEAUX Bit Description Sticky Bit Reset Valuea Bit Access 0h RW 1h RW Reserved This decision filter does not apply to traffic received from Apply_to_netwo the network. rk_traffic This decision filter applies to traffic received from the network. Flow Control Discard Offset Start: 5930h at 4 Offset End: 5933h at 4 0 = Apply filtering rules to packets with Flow Control EtherType. 1 = Discard packets with Flow Control EtherType. Note: Flow Control EtherType is 0x8808 Reserved Reserved Flex TCO Flex TCO - Controls the inclusion of Flex TCO filtering in the manageability filter decision (OR section). Bit 24 corresponds to Flex TCO filter 0. 0h RW 23 :16 Flex port Flex port - Controls the inclusion of Flex port filtering in the manageability filter decision (OR section). Bit 16 corresponds to flex port 0, etc. 0x0 RW 15 :12 Reserved Reserved for additional L2 EtherType OR filters. 0x0 RW 11 :08 L2 EtherType OR L2 EtherType - Controls the inclusion of L2 EtherType filtering in the manageability filter decision (OR section). 0x0 RW 07 :04 Reserved 03 :00 L2 EtherType AND 0x0 RW 24 Reserved L2 EtherType - Controls the inclusion of L2 EtherType filtering in the manageability filter decision (AND section). a. Default values are read from EEPROM. 28.21.1.8 Manageability IP Address Filter--MIPAF [0:3][0:15] (0x58B0 + 4*n [n=0...15]; RW) The Manageability IP Address Filter register stores IP addresses for manageability filtering. The MIPAF register can be used in two configurations, depending on the value of the MANC. EN_IPv4_FILTER bit: * EN_IPv4_FILTER = 0: the last 128 bits of the register store a single IPv6 address (IPV6ADDR3) * EN_IPv4_FILTER = 1: the last 128 bits of the register store 4 IPv4 addresses (IPV4ADDR[3:0]) MANC.EN_IPv4_FILTER = 0: Intel(R) Communications Chipset 89xx Series - Datasheet 1476 October 2012 Order Number: 327879-001US 28.0 DWORD# Address 0 0x58B0 1 0x58B4 2 0x58B8 3 0x58BC 4 0x58C0 5 0x58C4 6 0x58C8 7 0x58CC 8 0x58D0 9 0x58D4 10 0x58D8 11 0x58DC 12 0x58E0 13 0x58E4 14 0x58E8 15 0x58EC 31 0 IPV6ADDR0 IPV6ADDR1 IPV6ADDR2 IPV6ADDR3 Field definitions for 0 setting: Field IPV6ADDR0 IPV6ADDR1 IPV6ADDR2 IPV6ADDR3 Dword # Address Bit(s) Initial Value Description 0 0x58B0 31:0 X* IPv6 Address 0, bytes 1-4 (LS byte is first on the wire) 1 0x58B4 31:0 X* IPv6 Address 0, bytes 5-8 2 0x58B8 31:0 X* IPv6 Address 0, bytes 9-12 3 0x58BC 31:0 X* IPv6 Address 0, bytes 13-16 X* IPv6 Address 1, bytes 1-4 (LS byte is first on the wire) 0 0x58C0 31:0 1 0x58C4 31:0 X* IPv6 Address 1, bytes 5-8 2 0x58C8 31:0 X* IPv6 Address 1, bytes 9-12 3 0x58CC 31:0 X* IPv6 Address 1, bytes 13-16 0 0x58D0 31:0 X* IPv6 Address 2, bytes 1-4 (LS byte is first on the wire) 1 0x58D4 31:0 X* IPv6 Address 2, bytes 5-8 2 0x58D8 31:0 X* IPv6 Address 2, bytes 9-12 3 0x58DC 31:0 X* IPv6 Address 2, bytes 13-16 0 0x58E0 31:0 X* IPv6 Address 3, bytes 1-4 (LS byte is first on the wire) 1 0x58E4 31:0 X* IPv6 Address 3, bytes 5-8 2 0x58E8 31:0 X* IPv6 Address 3, bytes 9-12 3 0x58EC 31:0 X* IPv6 Address 3, bytes 13-16 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1477 MANC.EN_IPv4_FILTER = 1: DWORD# Address 0 0x58B0 1 0x58B4 2 0x58B8 3 0x58BC 4 0x58C0 5 0x58C4 6 0x58C8 7 0x58CC 8 0x58D0 31 0 IPV6ADDR0 IPV6ADDR1 9 0x58D4 10 0x58D8 11 0x58DC 12 0x58E0 IPV4ADDR0 13 0x58E4 IPV4ADDR1 14 0x58E8 IPV4ADDR2 15 0x58EC IPV4ADDR3 IPV6ADDR2 Field definitions for 1 Setting: Field IPV6ADDR0 IPV6ADDR1 IPV6ADDR2 Dword # Address Bit(s) Initial Valuea 0 0x58B0 31:0 X IPv6 Address 0, bytes 1-4 (LS byte is first on the wire) 1 0x58B4 31:0 X IPv6 Address 0, bytes 5-8 Description 2 0x58B8 31:0 X IPv6 Address 0, bytes 9-12 3 0x58BC 31:0 X IPv6 Address 0, bytes 16-13 0 0x58C0 31:0 X IPv6 Address 1, bytes 1-4 (LS byte is first on the wire) 1 0x58C4 31:0 X IPv6 Address 1, bytes 5-8 2 0x58C8 31:0 X IPv6 Address 1, bytes 9-12 3 0x58CC 31:0 X IPv6 Address 1, bytes 16-13 0 0x58D0 31:0 X IPv6 Address 2, bytes 1-4 (LS byte is first on the wire) 1 0x58D4 31:0 X IPv6 Address 2, bytes 5-8 2 0x58D8 31:0 X IPv6 Address 2, bytes 9-12 3 0x58DC 31:0 X IPv6 Address 2, bytes 16-13 IPV4ADDR0 0 0x58E0 31:0 X IPv4 Address 0 (LS byte is first on the wire) IPV4ADDR1 1 0x58E4 31:0 X IPv4 Address 1 (LS byte is first on the wire) IPV4ADDR2 2 0x58E8 31:0 X IPv4 Address 2 (LS byte is first on the wire) IPV4ADDR3 3 0x58EC 31:0 X IPv4 Address 3 (LS byte is first on the wire) a. The initial values for these registers can be loaded from the EEPROM after power-up reset. The registers are written by the BMC and not accessible to the host for writing. Intel(R) Communications Chipset 89xx Series - Datasheet 1478 October 2012 Order Number: 327879-001US 28.0 Initial value: Table 28-303.Manageability IP Address Filter--MIPAF [0:3][0:15] (0x58B0 + 4*n [n=0...15]; RW) Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Acronym IP_ADDR 4 bytes Offset Start: 58B0h at 4 Offset End: 58B3h at 4 Power Well: GBEAUX Bit Description Sticky Bit Reset Valuea Bit Access X RW 4 bytes of IP (v6 or v4) address. i mod 4 = 0 to bytes 1 - 4 i mod 4 = 1 to bytes 5 - 8 i mod 4 = 0 to bytes 9 - 12 i mod 4 = 0 to bytes 13 - 16 where i div 4 is the index of IP address (0...3) a. Reset - The MNGONLY register is cleared on LAN_PWR_GOOD and firmware reset. The initial values for this register can be loaded from the EEPROM after power-up reset or firmware reset. Reset - The registers are cleared on LAN_PWR_GOOD only. Note: These registers should be written in network order. 28.21.1.9 Manageability MAC Address Low--MMAL [0:3] (0x5910 + 8*n [n= 0...1]; RW) Where "n" is the exact unicast/Multicast address entry, equal to 0...1. Table 28-304.Manageability MAC Address Low--MMAL [0:3][0:1] (0x5910 + 8*n [n= 0...1]; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Offset Start: 5910h at 8 Offset End: 5913h at 8 Power Well: GBEAUX Bit Range Bit Acronym Bit Description Sticky 31 :00 MMAL Manageability MAC Address Low. The lower 32 bits of the 48 bit Ethernet address. Bit Reset Valuea Bit Access X RW a. The initial values for these registers can be loaded from the EEPROM after power-up reset. The registers are written by the BMC and not accessible to the host for writing. These registers contain the lower bits of the 48 bit Ethernet address. The MMAL registers are written by the BMC and are not accessible to the host for writing. The registers are used to filter manageability packets. See section 27.4 . Reset - The MMAL registers are cleared on LAN_PWR_GOOD only. The initial values for this register can be loaded from the EEPROM after power-up reset. Note: The MMAL.MMAL field should be written in network order. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1479 28.21.1.10 Manageability MAC Address High--MMAH [0:3][0:1] (0x5914 + 8*n [n=0...1]; RW) Where "n" is the exact unicast/Multicast address entry, equal to 0...1. Table 28-305.Manageability MAC Address High--MMAH [0:3][0:1] (0x5914 + 8*n [n=0...1]; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :16 Reserved 15 :00 MMAH Offset Start: 5914h at 8 Offset End: 5917h at 8 Power Well: GBEAUX Bit Reset Valuea Bit Access Reserved. Reads as 0. Ignored on write. 0x0 RW Manageability MAC Address High. The upper 16 bits of the 48 bit Ethernet address. 0x0 RW Bit Description Sticky a. The initial values for these registers can be loaded from the EEPROM after power-up reset. The registers are written by the BMC and not accessible to the host for writing. These registers contain the upper bits of the 48 bit Ethernet address. The complete address is {MMAH, MMAL}. The MMAH registers are written by the BMC and are not accessible to the host for writing. The registers are used to filter manageability packets. See section 27.4 . Reset - The MMAL registers are cleared on LAN_PWR_GOOD only. The initial values for this register can be loaded from the EEPROM after power-up reset or firmware reset. Note: The MMAH.MMAH field should be written in network order. 28.21.1.11 Flexible TCO Filter Table Registers--FTFT [0:3] (0x9400-0x94FC; RW) The Flexible TCO Filter Table registers (FTFT) contains a 128 byte pattern and a corresponding 128-bit mask array. If enabled, the first 128 bytes of the received packet are compared against the non-masked bytes in the FTFT register. The 128 byte filter is composed of 32 DW entries, where each 2 DWs are accompanied by an 8-bit mask, one bit per filter byte. The bytes in each 2 DWs are written in network order i.e. byte0 written to bits [7:0], byte1 to bits [15:8] etc. The mask field is set so that bit0 in the mask masks byte0, bit 1 masks byte 1 etc. A value of 1 in the mask field means that the appropriate byte in the filter should be compared to the appropriate byte in the incoming packet. Note: The mask field must be 8 bytes aligned even if the length field is not 8 bytes aligned, as the hardware implementation compares 8 bytes at a time so it should get extra masks until the end of the next quad word. Any mask bit that is located after the length should be set to 0 indicating no comparison should be done. Note: In case the actual length which is defined by the length field register and the mask bits is not 8 bytes aligned there may be a case that a packet which is shorter then the actual required length passes the flexible filter. This may occur due to comparison of up to 7 bytes that come after the packet, but are not a real part of the packet. Intel(R) Communications Chipset 89xx Series - Datasheet 1480 October 2012 Order Number: 327879-001US 28.0 The last DW of the filter contains a length field defining the number of bytes from the beginning of the packet compared by this filter. If actual packet length is less than length specified by this field, the filter fails. Otherwise, it depends on the result of actual byte comparison. The value should not be greater than 128. 31 0 31 8 Reserved Reserved 7 0 31 0 Mask [7:0] 31 0 DW 1 DW 0 Reserved Reserved Mask [15:8] DW 3 DW 2 Reserved Reserved Mask [23:16] DW 5 DW 4 Reserved Reserved Mask [31:24] DW 7 DW 6 31 8 7 0 31 8 7 0 31 0 31 0 Reserved Reserved Reserved Mask [127:120] DW 29 DW 28 Reserved Length Reserved Mask [127:120] DW 31 DW 30 Field definitions for Filter Table Registers: Field Dword Filter 0 DW0 0 Address 0x9400 Bit(s) 31:0 Initial Value X Filter 0 DW1 1 0x9404 31:0 X Filter 0 Mask[7:0] 2 0x9408 7:0 X Reserved 3 0x940C Filter 0 DW2 4 0x9410 31:0 X 60 0x94F0 31:0 X X ... Filter 0 DW30 Filter 0 DW31 61 0x94F4 31:0 X Filter 0 Mask[127:120] 62 0x94F8 7:0 X Length 63 0x94FC 6:0 X The initial values for the FTFT registers can be loaded from the EEPROM after power-up reset. The FTFT registers are written by the BMC and are not accessible to the host for writing. The registers are used to filter manageability packets. Reset - The FTFT registers are cleared on LAN_PWR_GOOD only. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1481 28.22 Memory Error Registers Description The Controller main internal memories are protected by error correcting code (ECC) or parity bits. The Controller contains several registers that enable and report detection of internal memory errors. 28.22.1 Parity and ECC Error Indication--PEIND [0:3] (0x1084; RC) Table 28-306.Parity and ECC Error Indication--PEIND [0:3] (0x1084; RC) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: Default: 0x0 Bit Range Bit Acronym 31 :04 Reserved Offset Start: 1084h Offset End: 1087h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved Write 0 ignore on read. 03 dma_parity_fat Fatal Error detected in DMA Memory al_ind (LH) Bit is latched high and cleared on read. 0h RC 02 pcie_parity_fata Fatal Error detected in PCIe Memory. l_ind (LH) Bit is latched high and cleared on read. 0h RC 01 mng_parity_fat Fatal Error detected in Management Memory. al_ind (LH) Bit is latched high and cleared on read. 0h RC 00 lanport_parity_f Fatal Error detected in LAN port Memory. atal_ind (LH) Bit is latched high and cleared on read. 0h RC Intel(R) Communications Chipset 89xx Series - Datasheet 1482 B:0:1+ Index1 October 2012 Order Number: 327879-001US 28.0 28.22.2 Parity and ECC Indication Mask--PEINDM [0:3] (0x1088; RW) Table 28-307.Parity and ECC Indication Mask--PEINDM [0:3] (0x1088; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0xF Bit Range Bit Acronym 31 :04 Reserved Offset Start: 1088h Offset End: 108Bh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Reserved Write 0 ignore on read. 03 dma_parity_fat When set and PEIND.dma_parity_fatal_ind is set, enable al_ind interrupt generation by setting the ICR.FER bit. 1h 02 pcie_parity_fata When set and PEIND.pcie_parity_fatal_ind is set, enable l_ind interrupt generation by setting the ICR.FER bit. 1h 01 mng_parity_fat When set and PEIND.mng_parity_fatal_ind is set, enable al_ind interrupt generation by setting the ICR.FER bit. 1h 00 lanport_parity_f When set and PEIND.lanport_parity_fatal_ind is set, atal_ind enable interrupt generation by setting the ICR.FER bit. 1h 28.22.3 Bit Access DMA Transmit Descriptor Parity Status--DTPARS [0:3] (0x3510; RW1C) Table 28-308.DMA Transmit Descriptor Parity Status--DTPARS [0:3] (0x3510; RW1C) Description: View: PCI Size: 32 bit B:0:1+ Bus:Device:Function: Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :07 Reserved Offset Start: 3510h Offset End: 3513h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved. Write 0, ignore on read. 06 par_ind_dtx_ica dtx_icache memory parity error indication che 0h RW1C 05 par_ind_dtx_dh dtx_dh_rf memory parity error indication _rf 0h RW1C 04 par_ind_dtx_dp dtx_dpt2 memory parity error indication t2 0h RW1C 03 par_ind_dtx_te dtx_temp memory parity error indication mp 0h RW1C 02 par_ind_dtx_hd dtx_hdr memory parity error indication r 0h RW1C 01 par_ind_dtx_cnt dtx_cntxt memory parity error indication xt 0h RW1C 00 par_ind_dtx_lso dtx_lso memory parity error indication 0h RW1C October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1483 28.22.4 DMA Receive Descriptor Parity Status--DRPARS [0:3] (0x3514; RW1C) Table 28-309.DMA Receive Descriptor Parity Status--DRPARS [0:3] (0x3514; RW1C) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :04 Reserved Offset Start: 3514h Offset End: 3517h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved. Write 0, ignore on read. 03 par_ind_drx_srr drx_srrctl memory parity error indication. ctl 0h RW1C 02 par_ind_drx_ica drx_icache memory parity error indication. che 0h RW1C 01 par_ind_drx_dh drx_dh_rf memory parity error indication. _rf 0h RW1C 00 par_ind_drx_de drx_desc memory parity error indication. sc 0h RW1C 28.22.5 Dhost Parity Status--DDPARS [0:3] (0x3518; RW1C) Table 28-310.Dhost Parity Status--DDPARS [0:3] (0x3518; RW1C) Description: View: PCI Size: 32 bit Bus:Device:Function: B:0:1+ Index1 BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :04 Reserved Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved. Write 0, ignore on read. 03 par_ind_dma_st dma_stat memory parity error indication. at 0x0 RW1C 02 par_ind_dhost_r dhost_rx_desc memory parity error indication. x_desc 0h RW1C 01 par_ind_dhost_t dhost_tx_data memory parity error indication. x_data 0h RW1C 00 par_ind_dhost_t dhost_tx_desc memory parity error indication. x_desc 0h RW1C Intel(R) Communications Chipset 89xx Series - Datasheet 1484 Offset Start: 3518h Offset End: 351Bh October 2012 Order Number: 327879-001US 28.0 28.22.6 Tx Packet Buffer ECC Status--TPBECCSTS [0:3] (0x345C; RW) Table 28-311.Tx Packet Buffer ECC Status--TPBECCSTS [0:3] (0x345C; RW) Description: View: PCI Size: 32 bit Bit Range 31 :27 26 16 BAR: GBEPCIBAR0[0:3] B:0:1+ Index1 Default: 0x00010000 Bit Acronym Reserved ECC Enable Reserved 07 :00 Corr_err_cnt (RC) Offset Start: 345Ch Offset End: 345Fh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0h RW 1h RW 0x0 RW Reserved Pb_cor_err_sta Status of PB Correctable Error. (RC) This bit is cleared by a read. 15 :08 28.22.7 Bus:Device:Function: ECC Enable for Packet Buffer. Reserved Write 0 ignore on read Correctable Error Count This counter is incremented every time a correctable error is detected; the counter stops after reaching 0xFF. This field is cleared by read. LAN Port Parity Error Control Register--LANPERRCTL [0:3] (0x5F54; RW) Table 28-312.LAN Port Parity Error Control Register--LANPERRCTL [0:3] (0x5F54; RW) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Default: 0x0 Bit Range Bit Acronym 31 :09 Reserved 08 :00 B:0:1+ Bus:Device:Function: Index1 mrx_flx_en October 2012 Order Number: 327879-001US Offset Start: 5F54h Offset End: 5F57h Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access 0x0 RW Reserved Enable mrx_flx parity error indication Note: Software should program the FHFT, FHFT_EXT, FTFT memories before enabling parity check, to avoid incorrect parity error detection. Intel(R) Communications Chipset 89xx Series - Datasheet 1485 28.22.8 LAN Port Parity Error Status Register--LANPERRSTS [0:3] (0x5F58; RO) Parity indication bits cleared by issuing reset. Table 28-313.LAN Port Parity Error Status Register--LANPERRSTS [0:3] (0x5F58; RO) Description: View: PCI Size: 32 bit BAR: GBEPCIBAR0[0:3] Bus:Device:Function: B:0:1+ Index1 Default: 0x0 Bit Range Bit Acronym 31 :15 Reserved Offset Start: 5F58h Offset End: 5F5Bh Power Well: GBEAUX Bit Description Sticky Bit Reset Value Bit Access Reserved 14 f_vlan f_vlan parity error indication 0h RO 13 f_mulc f_mulc parity error indication 0h RO 12 f_rss f_rss parity error indication 0h RO 11 f_uta f_uta parity error indication 0h RO 10 stat_regs stat_regs parity error indication 0h RO 09 retx_buf retx_buf parity error indication 0h RO mrx_flx mrx_flx parity error indication 0x0 RO 08 :00 Intel(R) Communications Chipset 89xx Series - Datasheet 1486 October 2012 Order Number: 327879-001US 29.0 29.0 Function 1-4 (GbE) 29.1 Introduction This chapter outlines the GbE Controller CSRs and other topics. The Controller is a multi-function device with the following functions: * LAN 0 * LAN 1 * LAN 2 * LAN 3 All functions contain the following regions of the PCI configuration space: * Mandatory PCI configuration registers * Power management capabilities * MSI and MSI-X capabilities * PCIe* extended capabilities The Controller Physical Function implements Type 0 PCIe configuration space that is compliant with PCIe 2.0 Specification. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1487 ) Figure 29-1. GbE Controller PCI Configuration Registers (Sheet 1 of 2) 3 2 1 0 Device ID Vendor ID 00h Status Command 04h Class Code BIST Header Type Required PCI Compatible Configuration Space Master Latency Timer Revision ID 08h Cache Line Size 0Ch Base Address Register 0 (GbEPCIBAR0) 10h Base Address Register 1 (GbEPCIBAR1) 14h Base Address Register 2 (GbEPCIBAR2) 18h Base Address Register 3 (GbEPCIBAR3) 1Ch Base Address Register 4 (GbEPCIBAR4) 20h Base Address Register 5 (GbEPCIBAR5) 24h Cardbus CIS Pointer Subsystem ID 28h Subsystem Vendor ID Expansion ROM Base Address 30h 34h Interrupt Pin Interrupt Line 3Ch Next Cap Pointer PM CAP_ID 40h Reserved Max_Lat Min_Gnt Power Management Capabilities 38h Power Management Control and Status Message Control MSI Cap Next Cap Pointer 44h MSI CAP_ID MSI Addr Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1488 5Ch Mask Bits 60h Pending Bits 64h Reserved 68h Reserved MSI-X Cap 50h 54h - 5Bh MSI Data Message Control 2Ch Capabilities Pointer Reserved PM Cap Offset1 Next Cap Pointer 6Ch MSI-X CAP_ID 70h MSI-X Table Offset and BIR 74h MSI-X PBA Offset and BIR 78h IOADDR 98h IODATA 9Ch October 2012 Order Number: 327879-001US 29.0 Figure 29-1. GbE Controller PCI Configuration Registers (Sheet 2 of 2) 3 2 PCI Express Capabilities Register 1 0 Next Cap Pointer PCIe CAP_ID Device Capabilities Device Control Link Capability Link Control Slot Capabilities (Reserved) B4h Slot Status (Reserved) Slot Control (Reserved) B8h Root Control (Reserved) BCh Root Status (Reserved) C0h Device Capabilities 2 C4h Device Control 2 Link Capabilities 2 (Reserved) Link Status 2 Link Control 2 Slot Status 2 (Reserved) Reserved Next Cap Pointer VPD CAP_ID E0h E4h Reserved E8h - FCh AER Capability ID Uncorrectable Error Status Reserved D8h VPD Data Next Capability Pointer/Capability Version AER Configuration Space D0h D4h Slot Control 2 (Reserved) VPD Address C8h CCh Slot Capabilities 2 (Reserved) ARI Configuration Space B0h Root Capabilities (Reserved) Device Status 2 (Reserved) Serial ID Space A8h ACh Link Status VPD A0h A4h Device Status PCIE Cap Offset1 100h 104h Uncorrectable Error Mask 108h Uncorrectable Error Severity 10Ch Correctable Error Status 110h Correctable Error Mask 114h Control and Capability 118h Header Log 11Ch Header Log 120h Header Log 124h Header Log 128h Next Capability Pointer/Capability Version Serial ID Capability ID 140h Serial Number Register (Lower DWORD) 144h Serial Number Register (Upper DWORD) 148h Next Capability Pointer/Capability Version ARI Capability ID 150 ARI Control ARI Capability 154h Reserved 158h - FFCh Notes: 1. Any addresses not shown are Read-only 0. A description of the registers is provided in the following sections. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1489 29.2 Detailed Register Summary 29.2.1 PCI Views Table 29-1. Bus B, Device D, Function 1: Summary of PCIe GigE Configuration Registers Offset Start 02h Offset End 03h Register ID - Description "PDID0--PF Device Identification Register (GbE0 )" on page 1494 Default Value 0436h Table 29-2. Bus B, Device D, Function 2: Summary of PCIe GigE Configuration Registers Offset Start 02h Offset End 03h Register ID - Description "PDID1--PF Device Identification Register (GbE1 )" on page 1494 Default Value 0436h Table 29-3. Bus B, Device D, Function 3: Summary of PCIe GigE Configuration Registers Offset Start 02h Offset End 03h Register ID - Description "PDID2--PF Device Identification Register (GbE2 )" on page 1495 Default Value 0436h Table 29-4. Bus B, Device D, Function 4: Summary of PCIe GigE Configuration Registers Offset Start 02h Offset End 03h Register ID - Description "PDID3--PF Device Identification Register (GbE3 )" on page 1495 Default Value 0436h Table 29-5. Bus B, Device D, Function 1+Index1: Summary of PCIe GigE Configuration Registers (Sheet 1 of 3) Offset Start Offset End Register ID - Description Default Value 00h 01h "PVID[0:3]--PF Vendor Identification Register" on page 1493 8086h 04h 05h "PPCICMD[0:3]--PF Device Command Register" on page 1496 0000h 06h 07h "PPCISTS[0:3]--PF PCI Device Status Register" on page 1497 0010h 08h 08h "PRID[0:3]--PF Revision ID Register" on page 1499 10h 09h 0Bh "PCC[0:3]--PF Class Code Register" on page 1499 020000h 0Eh 0Eh "PHDR[0:3]--PF Header Type Register" on page 1500 80h 10h 13h "GbEPCIBAR0[0:3]--BAR0 Base Address Register" on page 1501 00000000h 14h 17h "GbEPCIBAR1[0:3]--BAR1 Base Address Register" on page 1502 00000000h 18h 1Bh "GbEPCIBAR2[0:3]--BAR2 Base Address Register" on page 1502 00000001h 1Ch 1Fh "GbEPCIBAR3[0:3]--BAR3 Base Address Register" on page 1503 00000000h Intel(R) Communications Chipset 89xx Series - Datasheet 1490 October 2012 Order Number: 327879-001US 29.0 Table 29-5. Bus B, Device D, Function 1+Index1: Summary of PCIe GigE Configuration Registers (Sheet 2 of 3) Offset Start 20h Offset End 23h Default Value Register ID - Description "GbEPCIBAR4[0:3]--BAR4 Base Address Register" on page 1504 00000000h 24h 27h "GbEPCIBAR5[0:3]--BAR5 Base Address Register" on page 1505 00000000h 2Ch 2Dh "PSVID[0:3]--PF Subsystem Vendor ID Register" on page 1505 8086h 2Eh 2Fh "PSID[0:3]--PF Subsystem ID Register" on page 1506 0000h 34h 34h "PCP[0:3]--PF Capabilities Pointer Register" on page 1506 40h 3Ch 3Ch "PIRQL[0:3]--PF Interrupt Line Register" on page 1507 00h 3Dh 3Dh "PIRQP0--PF Interrupt Pin Register 0" on page 1507 02h 3Dh 3Dh "PIRQP1--PF Interrupt Pin Register 1" on page 1508 03h 3Dh 3Dh "PIRQP2--PF Interrupt Pin Register 2" on page 1508 04h 3Dh 3Dh "PIRQP3--PF Interrupt Pin Register 3" on page 1509 02h 40h 40h "PPMCAP[0:3]--PF Power Management Capabilities ID Register" on page 1510 01h 41h 41h "PPMCP[0:3]--PF Power Management Next Capability Pointer Register" on page 1510 50h 42h 43h "PPMC[0:3]--PF Power Management Capabilities Register" on page 1511 C823hh 44h 47h "PPMCSR[0:3]--PF Power Management Control and Status Register" on page 1512 00000000h 50h 50h "PMSICID[0:3] - Message Signalled Interrupt Capability ID Register" on page 1513 05h 51h 51h "PMSINCP[0:3]--Message Signalled Interrupt Next Capability Pointer Register" on page 1513 70h 52h 53h "PMSICTL[0:3]--Message Signalled Interrupt Control Register" on page 1514 0180h 54h 57h "PMSILADDR[0:3]--Message Signalled Interrupt Lower Address Register" on page 1514 00000000h 58h 5Bh "PMSIUADDR[0:3]--Message Signalled Interrupt Upper Address Register" on page 1515 00000000h 5Ch 5Dh "PMSIDATA[0:3]--Message Signalled Interrupt Data Register" on page 1515 0000h 60h 63h "PMSIMSK[0:3]--Message Signalled Interrupt Mask Register" on page 1515 0000h 64h 67h "PMSIPND[0:3]--Message Signalled Interrupt Pending Register" on page 1516 0000h 70h 70h "PMSI-X[0:3]--PF Message Signalled Interrupt X Capability ID Register" on page 1517 11h 71h 71h "PMSIXNCP[0:3]--PF MSIX Next Capability Pointer Register" on page 1518 A0h 72h 73h "PMSIXCNTL[0:3]--PF Message Signalled Interrupt X Control Register" on page 1518 0009h 74h 77h "PMSIXTBIR[0:3]--PF MSI-X Table Offset & Table BIR Register" on page 1519 00000003h 7Bh "PMSIXPBABIR[0:3]--PF MSI-X Pending Bit Array & BIR Offset Register" on page 1519 00002003h 78h 98h 9Bh "IOADDR[0:3]--IOADDR Register" on page 1520 0h 9Ch 9Fh "IODATA[0:3]--IODATA Register" on page 1520 0h A0h A0h "PPCID[0:3]--PF PCI Express Capability Register" on page 1521 10h A1h A1h "PPCP[0:3]--PF PCI Express Next Capability Pointer Register" on page 1522 00h A2h A3h "PPCR[0:3]--PF PCI Express Capabilities Register" on page 1522 0002h A4h A7h "PPDCAP[0:3]--PF PCI Express Device Capabilities Register" on page 1523 10008041h A8h A9h "PPDCNTL[0:3]--PF PCI Express Device Control Register" on page 1524 2810h AAh ABh "PPDSTAT[0:3]--PF PCI Express Device Status Register" on page 1525 0010h October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1491 Table 29-5. Bus B, Device D, Function 1+Index1: Summary of PCIe GigE Configuration Registers (Sheet 3 of 3) Offset Start ACh Offset End AFh Default Value Register ID - Description "PLCAPR[0:3]--PF Link Capabilities Register" on page 1526 3B502h B0h B1h "PLCNTLR[0:3]--PF Link Control Register" on page 1528 0000h B2h B3h "PLSR[0:3]--PF Link Status Register" on page 1529 0102h C4h C7h "PDCAPR2[0:3]--PF Device Capabilities 2 Register" on page 1531 00000012h C8h C9h "PDCNTR2[0:3]--PF Device Control 2 Register" on page 1532 0000h D0h D1h "PLCNTLR2[0:3]--PF Link Control 2 Register" on page 1533 0000h D2h D3h "PLSR2[0:3]--PF Link Status 2 Register" on page 1535 0000h E0h E0h "VPDCID[0:3]--VPD Capability ID Register" on page 1536 3h E1h E1h "VPDNCP[0:3]--VPD Next Capability Pointer Register" on page 1536 0h E2h E3h "VPDADDR[0:3]--VPD Address Register" on page 1537 0h E4h E7h "VPDDATA[0:3]--VPD Data" on page 1537 0h 100h 103h "PPCIEAERCAPID[0:3]--PF PCI Express AER Capability ID Register" on page 1539 13810001h 104h 107h "PPAERUCS[0:3]--PF PCI Express AER Uncorrectable Error Status Register" on page 1539 0h 108h 10Bh "PPAERUCM[0:3]--PF PCI Express AER Uncorrectable Error Mask Register" on page 1540 0h 10Ch 10Fh "PPAERUCSEV[0:3]--PF PCI Express AER Uncorrectable Error Severity Register" on 00062030h page 1541 110h 113h "PPAERCS[0:3]--PF PCI Express AER Correctable Error Register" on page 1542 00h 114h 117h "PPAERCM[0:3]--PF PCI Express AER Correctable Error Mask Register" on page 1543 2000h 118h 11Bh "PPAERCTLCAP[0:3]--PF PCI Express AER Control and Capability Register" on page 1543 0h 11Ch 11Fh "PPAERHDRLOG0[0:3]--PF PCI Express AER Header Log 0 Register" on page 1544 0h 120h 123h "PPAERHDRLOG1[0:3]--PF PCI Express AER Header Log 1 Register" on page 1544 0h 124h 127h "PPAERHDRLOG2[0:3]--PF PCI Express AER Header Log 2 Register" on page 1545 0h 128h 12Bh "PPAERHDRLOG3[0:3]--PF PCI Express AER Header Log 3 Register" on page 1545 0h 138h 13Bh "PARIDHDR[0:3]--PF Alternative Routing ID Capability Header" on page 1546 0001000Eh 13Ch 13Dh "PFARICAP0--PF ARI Capabilities Register" on page 1547 0200h 13Eh 13Fh "PARIDCTL[0:3]--PF Alternative Routing ID Control Register" on page 1550 00000000h Table 29-6. Bus B, Device D, Function 1+Index 1: Summary of GigE Registers Mapped Through GbEPCI Memory BAR Offset Start Offset End Default Value Register ID - Description 00h 03h "IOADDR[0:3]--IOADDR Register" on page 1520 0h 04h 07h "IODATA[0:3]--IODATA Register" on page 1520 0h Intel(R) Communications Chipset 89xx Series - Datasheet 1492 October 2012 Order Number: 327879-001US 29.0 29.3 Mandatory PCI Configuration Registers 29.3.1 Detailed Register Descriptions 29.3.1.1 PVID[0:3]--PF Vendor Identification Register The VID Register contains the vendor identification number. This 16-bit register combined with the Device Identification Register uniquely identifies any PCI device. Writes to this register have no effect. Table 29-7. PVID[0:3]--PF Vendor Identification Register Description: View: PCI Size: 16 bit Bit Range 15 : 00 Note: BAR: Configuration B:D:1+ Bus:Device:Function: Index1 Default: 8086h Bit Acronym VID Offset Start: 00h Offset End: 01h Power Well: Core Bit Description Sticky Vendor Identification: This register field contains the PCI standard identification for Intel, 8086h. Bit Reset Value Bit Access 8086h RO Loaded from EEPROM offset 0x0E (Vendor ID). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1493 29.3.1.2 PDID0--PF Device Identification Register (GbE0) This 16-bit register combined with the Vendor Identification register uniquely identifies any PCI device. Writes to this register have no effect. This is a read-only register. This field identifies individual GbE Controller functions. It has the same default value for all LAN functions and is loaded from EEPROM on power up. Note: Loaded from EEPROM offset 0x0D. Table 29-8. PDID0--PF Device Identification Register (GbE0 ) Description: View: PCI 1 Size: 16 bit Bit Range 15 : 00 29.3.1.3 BAR: Configuration Offset Start: 02h Offset End: 03h Bus:Device:Function: B:D:1 Default: 0436h Bit Acronym DID15_0 Power Well: Core Bit Description Sticky Device Identification Number: This is a 16-bit value assigned to the Controller Device (including bit 0 defined by DID0). Bit 0 is set via fuse bit: Default DID = 0436h Bit Reset Value Bit Access 0436h RO PDID1--PF Device Identification Register (GbE1) This 16-bit register combined with the Vendor Identification register uniquely identifies any PCI device. Writes to this register have no effect. This is a read-only register. This field identifies individual GbE Controller functions. It has the same default value for all LAN functions and is loaded from EEPROM on power up. Note: Loaded from EEPROM offset (LAN Port 1 EEPROM section Base Address+ 0x0D). Table 29-9. PDID1--PF Device Identification Register (GbE1 ) Description: View: PCI 2 Size: 16 bit Bit Range 15 : 00 BAR: Configuration Default: 0436h Bit Acronym DID15_0 Power Well: Core Bit Description Device Identification Number: This is a 16-bit value assigned to the Controller Device. Bit 0 is set via fuse bit: Default DID = 0436h Intel(R) Communications Chipset 89xx Series - Datasheet 1494 Offset Start: 02h Offset End: 03h Bus:Device:Function: B:D:2 Sticky Bit Reset Value Bit Access 0436h RO October 2012 Order Number: 327879-001US 29.0 29.3.1.4 PDID2--PF Device Identification Register (GbE2) This 16-bit register combined with the Vendor Identification register uniquely identifies any PCI device. Writes to this register have no effect. This is a read-only register. This field identifies individual GbE Controller functions. It has the same default value for all LAN functions and is loaded from EEPROM on power up. The following table describes the possible values according to the SKU and functionality of each function. Note: Loaded from EEPROM offset (LAN Port 2EEPROM section Base Address+ 0x0D). Table 29-10. PDID2--PF Device Identification Register (GbE2 ) Description: View: PCI 3 Size: 16 bit Bit Range 15 : 00 29.3.1.5 BAR: Configuration Bus:Device:Function: B:D:3 Default: 0436h Bit Acronym DID15_0 Offset Start: 02h Offset End: 03h Power Well: Core Bit Description Sticky Device Identification Number: This is a 16-bit value assigned to the Controller Device. Bit 0 is set via fuse bit: Default DID = 0436h Bit Reset Value Bit Access 0436h RO PDID3--PF Device Identification Register (GbE3) This 16-bit register combined with the Vendor Identification register uniquely identifies any PCI device. Writes to this register have no effect. This is a read-only register. This field identifies individual GbE Controller functions. It has the same default value for all LAN functions and is loaded from EEPROM on power up but can be written by BIOS during initialization with a different value for each port. The following table describes the possible values according to the SKU and functionality of each function. Note: Loaded from EEPROM offset (LAN Port 3EEPROM section Base Address+ 0x0D). Table 29-11. PDID3--PF Device Identification Register (GbE3 ) Description: View: PCI 4 Size: 16 bit Bit Range 15 : 00 BAR: Configuration Bus:Device:Function: B:D:4 Default: 0436h Bit Acronym DID15_0 October 2012 Order Number: 327879-001US Offset Start: 02h Offset End: 03h Power Well: Core Bit Description Sticky Device Identification Number: This is a 16-bit value assigned to the Controller Device. Bit 0 is set via fuse bit: Default DID = 0436h Bit Reset Value Bit Access 0436h RO Intel(R) Communications Chipset 89xx Series - Datasheet 1495 29.3.1.6 PPCICMD[0:3]--PF Device Command Register This is a read/write register. Each function has its own command register. Unless explicitly specified, functionality is the same in all functions. Table 29-12. PPCICMD[0:3]--PF Device Command Register (Sheet 1 of 2) Description: View: PCI Size: 16 bit Bit Range 15 : 11 BAR: Configuration Bus:Device:Function: Default: 0000h Bit Reset Value Bit Access Reserved 0h RV Bit Acronym Reserved Offset Start: 04h Offset End: 05h Power Well: Core Bit Description Sticky 10 INTD Interrupt Disable: Setting this bit disables generation of INTX messages by the Controller. Default value is 0 which enables the INTX message generation. 0h RW 09 FBTB Fast Back-to-Back Enable: The Controller does not implement this functionality and it is not applicable to PCIe devices. The bit is hardwired to 0. 0h RO 08 SER SERR# Enable: When set, this bit enables the non-fatal and fatal errors detected by the Controller to be reported to the RC. The error reporting can also be enabled via the PCIe specific bits in the PCIe device control register (Section 29.4.6.5, "PPDCNTL[0:3]--PF PCI Express Device Control Register") The default value of this bit is 0. 0h RW 07 Reserved Reserved/Does not apply to PCIe. 0h RV 0h RW 06 PER Parity Error Enable: Controls the setting of the Master Data Parity Error bit in the Device Status Register (Section 29.4.6.6, "PPDSTAT[0:3]--PF PCI Express Device Status Register") The Master Data Parity Error bit is set by the EP if its Parity Error Enable bit is set and either of the following two conditions occurs: * If the EP receives a poisoned Completion from the RC * If the EP poisons a write request. If the Parity Error Enable bit is cleared, the Master Data Parity Error status bit is never set The default value of this bit is 0. 05 VPS VGA Palette Snoop Enable: The device does not implement this functionality/Does not apply to PCIe. The bit is hardwired to 0. 0h RO 04 MWE Memory Write and Invalidate Enable: The device does not implement this functionality/Does not apply to PCIe. The bit is hardwired to 0. 0h RO 03 SS Special Cycle Enable: The device does not implement this functionality/Does not apply to PCIe. The bit is hardwired to 0. 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 1496 B:D:1+ Index1 October 2012 Order Number: 327879-001US 29.0 Table 29-12. PPCICMD[0:3]--PF Device Command Register (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range 02 01 00 29.3.1.7 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0000h Offset Start: 04h Offset End: 05h Power Well: Core Bit Reset Value Bit Access BM Bus Master Enable: Controls the ability of the Controller to issue Memory Read/Write Requests. Clearing (0) this bit prevents the Controller from issuing any Memory Requests. Because MSIs are in-band memory writes, disabling the bus master enable bit disables MSI as well. PCIe messages are not affected by this bit. 0h RW MEM Memory Space Enable: Setting this bit enables access to the memory regions the device claims through its BARs. EP will return "unsupported request" completion status & error message in response to memory transactions it receives when this bit is clear. 0h RW IO I/O Space Enable: Setting this bit enables I/O space. EP will return "unsupported request" completion status & error message in response to memory transactions it receives when this bit is clear. Note: Bit 0 should be loaded from EEPROM - PCIe Init Configuration 2 EEPROM Word (0x19). This register bit is RW when bit 14 of the EEPROM word is a set (default value) or in EEPROM-less mode. And it is a RO when bit 14 of the EEPROM word is cleared. The default value is always 0b. 0h RW Bit Acronym Bit Description Sticky PPCISTS[0:3]--PF Device Status Register Each function has its own status register. Unless explicitly specified, functionality is the same in all functions. Table 29-13. PPCISTS[0:3]--PF PCI Device Status Register (Sheet 1 of 2) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0010h Bit Acronym Offset Start: 06h Offset End: 07h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 15 DPE Detected Parity Error: This bit is set by EP whenever it receives a Poisoned TLP, regardless of the state the Parity Error Enable bit in the Command register. Default value of this field is 0. 0h RW1C 14 SSE Signaled System Error: This bit is set by the EP when it sends a ERR_FATAL or ERR_NONEATAL message and the SERR bit in the Device Command register bit is set. Default value of this field is 0. 0h RW1C 13 RMA Received Master Abort Status: This bit is set when EP, as a Requestor receives a Completion with Unsupported Request Completion Status. Default value of this field is 0. 0h RW1C October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1497 Table 29-13. PPCISTS[0:3]--PF PCI Device Status Register (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: Default: 0010h Bit Reset Value Bit Access Received Target Abort Status: This bit is set when EP, as a Requestor receives a Completion with Completer Abort Completion Status. Default value of this field is 0. 0h RW1C STA Signaled Target Abort Status: This bit is set when EP completes a Request using Completer Abort Completion Status. Default value of this field is 0. 0h RW1C DST DEVSEL Timing: Does not apply to PCI Express. These bits are hardwired to 0. 00b RO 0h RW1C Bit Description 12 RTA 11 Sticky 08 MDPE Master Data Parity Error Detected: This bit is set by EP, as a Requestor if the Parity Error Enable bit in the Command register is 1b and either of the following two conditions occurs: * Requestor receives a Completion marked poisoned * Requestor detects a parity error in the inbound completion data fifo. * Requestor poisons a write Request If the Parity Error Enable bit is 0b, this bit is never set. Default value of this field is 0. 07 FB2B Fast Back-to-Back Capable: Does not apply to PCI Express. The bit is hardwired to 0. 0h RO 06 Reserved Reserved 0h RV 05 MC66 66 MHz Capable: Does not apply to PCI Express. The bit is hardwired to 0. 0h RO 04 CL Capabilities List: This bit is hardwired to 1 to indicate that the Controller has a capabilities list. 1h RO 03 IS Interrupt Status: Indicates that the EP has transmitted a INTX message and is awaiting servicing. This bit does not includes MSI's generated by the EP. 0h RO Reserved 0h RV 02 : 00 Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1498 Offset Start: 06h Offset End: 07h Power Well: Core Bit Acronym 10 : 09 B:D:1+ Index1 October 2012 Order Number: 327879-001US 29.0 29.3.1.8 PRID[0:3]--PF Revision ID Register The default revision ID of the Controller is 0x00. The value of the rev ID is a logic XOR between the default value and the value read from EEPROM. All LAN functions have the same revision ID. Note: Loaded from EEPROM offset 0x1E. Table 29-14. PRID[0:3]--PF Revision ID Register Description: View: PCI Size: 8 bit Bit Range 07 :4 03 :0 29.3.1.9 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 10h Offset Start: 08h Offset End: 08h Power Well: Core Bit Reset Value Bit Access RIDU Major Revision: Steppings which require all masks to be regenerated. 00b: A stepping 01b: B stepping 10b: C stepping 11b: D stepping 0001b RO RIDL Minor Revision: Incremented for each stepping which does not modify all masks. Reset for each major revision. 00b: x0 stepping 01b: x1 stepping 10b: x2 stepping 11b: x3 stepping 0000b RO Bit Acronym Bit Description Sticky PCC[0:3]--PF Class Code Register The class code is a RO hard coded value that identifies the Controller's functionality. * LAN 0...LAN3 - 0x020000/0x010000 - Ethernet/SCSI Adapter1 Note: Loaded from EEPROM Table 29-15. PCC[0:3]--PF Class Code Register Description: View: PCI Size: 24 bit BAR: Configuration B:D:1+ Bus:Device:Function: Index1 Default: 020000h Bit Range Bit Acronym 23 : 0 CC Offset Start: 09h Offset End: 0Bh Power Well: Core Bit Description Sticky Class Code: 0x020000/0x010000 - Ethernet/SCSI Adaptor Bit Reset Value Bit Access 020000h RO 1. Selected according to bit 11, 12, 13 or 14 in Device Rev ID EEPROM word for LAN0, LAN 1, LAN2 or LAN3 respectively. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1499 29.3.1.10 PHDR[0:3]--PF Header Type Register This indicates if a device is single function or multifunction. EP always is shown as MF device. Table 29-16. PHDR[0:3]--PF Header Type Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 29.3.2 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 80h Power Well: Core Bit Acronym HDR Offset Start: 0Eh Offset End: 0Eh Bit Description Sticky Bit Reset Value Bit Access 80h RO PCI Header Type: The header type of the EP device. 80h = multi-function device with standard header layout. Base Address Registers (0x10...0x27; R/W) The Base Address registers (BARs) are used to map the Controller register space of the various functions. The Controller has a memory BAR, IO BAR, and MSI-X BAR described in the following table below. The BARs location and sizes are described in Table 29-18 and Table 29-19. The fields within each BAR are then described in Table 29-18. 32-bit addresses may be used in one register for each memory mapping window or 64-bit addresses with 2 registers for each memory mapping window depending on the BARCTRL.BAR32 bit. Table 29-17. GbE Controller Base Address Registers Description - LAN 0...3 29.3.2.1 Mapping Windows Mapping Description Memory BAR The internal MMIO registers are accessed as direct memory mapped offsets from the Base Address register. Software can access a Dword or 64 bits. IO BAR All internal MMIO registers can be accessed using I/O operations. There are two 4byte registers in the IO mapping window: Addr Reg and Data Reg accessible as Dword entities. IO BAR support depends on the IO_Sup bit in the EEPROM "PCIe Init Configuration 2" word. MSI-X BAR The MSI-X vectors and Pending bit array (PBA) structures are accessed as direct memory mapped offsets from the MSI-X BAR. Software can access Dword entities. Base Address Register Fields All base address registers have the following fields: Table 29-18. Base Address Registers' Fields Field Bits Mem / IO Space Indication 0 RO 2:1 RO Memory Type R/W Description 0b = Indicates memory space. 1b = Indicates I/O. 00b = 32-bit BAR written by BIOS 10b = 64-bit BAR written by BIOS Intel(R) Communications Chipset 89xx Series - Datasheet 1500 October 2012 Order Number: 327879-001US 29.0 Field Bits R/W Description 0b = Non-prefetchable space. 1b = Prefetchable space. Prefetch Memory 3 R The Controller implements Non-prefetchable space in 32-bit memory BAR mode, since it has read-side effects. This bit is written by BIOS. It is required to be set to 1 for 64-bit memory BARs. The length of the RW bits and RO 0b bits depend on the mapping window sizes. Init value of the RW fields is 0x0. Address Space (Low register for 64bit Memory BARs) 29.3.2.2 31:4 R/W Mapping Window RO bits Memory CSR 16:4 for 128KB Other values are reserved MSI-X space is 16KB 13:4 I/O spaces size is 32 bytes 4:0 GbEPCIBAR0[0:3]--BAR0 Base Address Register See Table 29-18 and Table 29-19. This register is used as either a 32-bit BAR or the lower 32-bit of a 64-bit BAR. The Prefetchable and Addressing Type fields are dependent on the BAR width. Table 29-19. GbEPCIBAR0[0:3]--BAR0 Base Address Register Description: View: PCI Size: 32 bit Bit Range B:D:1+ Bus:Device:Function: Index1 BAR: Configuration Default: 00000000h Bit Acronym Offset Start: 10h Offset End: 13h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 31 : 17 ADDR Programmable Base Address: These bits are set by BIOS to locate the base address of the region. 0h RW 16 : 04 ZERO Lower Bits: Hardwired to 0 (128KB region). 0h RO 0b RO 00b RO 0b RO Prefetchable: 0: non-prefetchable (default for 32-bit) 1: prefetchable (default for 64-bit) 03 02 : 01 PREF TYP Note: Addressing Type: 00 - 32-bit (default for 32-bit) 10 - 64-bit (default for 64-bit) Note: 00 MEM October 2012 Order Number: 327879-001US BARCTRL.BAR32 controls reset value. When BARCTRL.BAR32=1 then PREF=0. When BARCTRL.BAR32=0 then PREF=BARCTRL.PREFBAR. BARCTRL.BAR32 controls reset value. Memory Space Indicator: Hardwired to 0 to identify the region as in memory space. Intel(R) Communications Chipset 89xx Series - Datasheet 1501 29.3.2.3 GbEPCIBAR1[0:3]--BAR1 Base Address Register See Table 29-18 and Table 29-19. This register is reserved when BAR0 is setup as 32bit. It is has RW attributes when BAR0 is setup as 64-bit. BARCTRL.BAR32 is used as the qualifier signal. Table 29-20. GbEPCIBAR1[0:3]--BAR1 Base Address Register Description: View: PCI Size: 32 bit Bit Range 31 : 00 BAR: Configuration B:D:1+ Index1 Default: 00000000h Bit Acronym ADDRRES Offset Start: 14h Offset End: 17h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 0h RV - Reserved for 32-bit (RV for 32-bit BARs) - Upper Address for 64-bit (RW for 64-bit BARs) Note: 29.3.2.4 Bus:Device:Function: BARCTRL.BAR32 controls these bits attributes. GbEPCIBAR2[0:3]--BAR2 Base Address Register See Table 29-18 and Table 29-19. This is always fixed as an IO BAR. Table 29-21. GbEPCIBAR2[0:3]--BAR2 Base Address Register Description: View: PCI Size: 32 bit Bus:Device:Function: B:D:1+ Index1 BAR: Configuration Default: 00000001h Power Well: Core Bit Range Bit Acronym 31 :07 ADDR Programmable Base Address: These bits are set by BIOS to locate the base address of the region. 06 :02 ZERO Lower Bits: Hardwired to 0 (128 Bytes) 01 00 Bit Description Sticky Bit Reset Value Bit Access 0h RW 000b RO RES Reserved 0b RO MEM Memory Space Indicator: Hardwired to 1 for I/O region 1b RO Intel(R) Communications Chipset 89xx Series - Datasheet 1502 Offset Start: 18h Offset End: 1Bh October 2012 Order Number: 327879-001US 29.0 29.3.2.5 GbEPCIBAR3[0:3]--BAR3 Base Address Register See Table 29-18 and Table 29-19. This register is used as either a 32-bit BAR or is reserved. For 64-bit BAR support BAR4-5 are used for this function. Table 29-22. GbEPCIBAR3[0:3]--BAR3 Base Address Register Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Bit Acronym ADDR 13 : 04 RES PREF TYP MEM Prefetchable: 0: non-prefetchable (default for 32-bit) October 2012 Order Number: 327879-001US Bit Access 0h RW 0h RO 0b RO 00b RO 0b RO hardwired to 0. Addressing Type: 00 - 32-bit (default for 32-bit) hardwired to 0. Memory Space Indicator: Hardwired to 0 to identify the region as in memory space. Note: Bit Reset Value BARCTRL.BAR32 controls these bits attributes. Hardwired to 0 (16KB region). Note: 00 Sticky Programmable Base Address: These bits are set by BIOS to locate the base address of the region. RW - for 32-bit BARs. Note: Offset Start: 1Ch Offset End: 1Fh Power Well: Core Bit Description Note: 02 : 01 B:D:1+ Index1 Default: 00000000h 31 : 14 03 Bus:Device:Function: hardwired to 0. Intel(R) Communications Chipset 89xx Series - Datasheet 1503 29.3.2.6 GbEPCIBAR4[3:0]--BAR4 Base Address Register See Table 29-18 and Table 29-19. This register is the lower 32-bit of a 64-bit BAR. This register is reserved when GbE is setup to use 32-bit BARs. BARCTRL.BAR32 is used as the qualifier signal. Table 29-23. GbEPCIBAR4[0:3]--BAR4 Base Address Register Description: View: PCI Size: 32 bit Bit Range 31 : 14 BAR: Configuration 03 Bit Acronym ADDR RES PREF TYP Programmable Base Address: These bits are set by BIOS to locate the base address of the region. - RW for 64-bit BARs. - RV for 32-bit BARs. 00 MEM Hardwired to 0 (16KB region). Bit Reset Value Bit Access 0h RV 0h RO 0 RO 00b RO 0b RO BARCTRL.BAR32 controls reset value. When BARCTRL.BAR32=0 then PREF=BARCTRL.PREFBAR otherwise PREF=0. Addressing Type: 00 - 32-bit (default for 32-bit)- not applicable as this register is not used in 32-bit mode. 10 - 64-bit (default for 64-bit) BARCTRL.BAR32 controls reset value. Memory Space Indicator: Hardwired to 0 to identify the region as in memory space. Intel(R) Communications Chipset 89xx Series - Datasheet 1504 Sticky BARCTRL.BAR32 controls these bits attributes. Prefetchable: 0: non-prefetchable (default for 32-bit) - not applicable as this register is not used in 32-bit mode. 1: prefetchable (default for 64-bit) Note: Offset Start: 20h Offset End: 23h Power Well: Core Bit Description Note: 02 : 01 B:D:1+ Index1 Default: 00000000h Note: 13 : 04 Bus:Device:Function: October 2012 Order Number: 327879-001US 29.0 29.3.2.7 GbEPCIBAR5[0:3]--BAR5 Base Address Register See Table 29-18 and Table 29-19. This register is the upper 32-bit of a 64-bit BAR. This register is reserved when GbE is setup to use 32-bit BARs. BARCTRL.BAR32 is used as the qualifier signal. Table 29-24. GbEPCIBAR5[0:3]--BAR5 Base Address Register Description: View: PCI Size: 32 bit Bit Range 31 : 00 BAR: Configuration B:D:1+ Index1 Default: 00000000h Bit Acronym ADDRRES Offset Start: 24h Offset End: 27h Power Well: Core Bit Description Sticky - Reserved for 32-bit (RV for 32-bit BARs) - Upper Address for 64-bit (RW for 64-bit BARs) Note: 29.3.2.8 Bus:Device:Function: Bit Reset Value Bit Access 0h RV BARCTRL.BAR32 controls these bits attributes. PSVID[0:3]--PF Subsystem Vendor ID Register This value can be loaded automatically from EEPROM address 0x0C at power up or reset. A value of 0x8086 is the default for this field at power up if the EEPROM does not respond or is not programmed. All functions are initialized to the same value. Table 29-25. PSVID[0:3]--PF Subsystem Vendor ID Register Description: View: PCI Size: 16 bit Bit Range 15 : 00 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 8086h Bit Acronym SVID October 2012 Order Number: 327879-001US Offset Start: 2Ch Offset End: 2Dh Power Well: Core Bit Description Sticky Subsystem Vendor ID: This value can be loaded automatically from EEPROM address 0x0C at power up or reset. A value of 0x8086 is the default for this field at power up if the EEPROM does not respond or is not programmed. Bit Reset Value Bit Access 8086h RO Intel(R) Communications Chipset 89xx Series - Datasheet 1505 29.3.2.9 PSID[0:3]--PF Subsystem ID Register Default to 0h. Value can be loaded fro EEPROM offset 0x0B. Table 29-26. PSID[0:3]--PF Subsystem ID Register Description: View: PCI Size: 16 bit Bit Range 15 : 00 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0000h Offset Start: 2Eh Offset End: 2Fh Power Well: Core Bit Acronym Bit Description SID Subsystem ID: Vendor supplied device ID. Default is 0h and identifies the device as Intel.23`. Loaded from EEPROM in reset. Sticky Bit Reset Value Bit Access 0h RO 29.3.2.10 Expansion ROM Base Address (0x30; RO) (Flash Not Implemented) 29.3.2.11 PCP[0:3]--PF Capabilities Pointer Register Table 29-27. PCP[0:3]--PF Capabilities Pointer Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 Bus:Device:Function: B:D:1+ Index1 BAR: Configuration Default: 40h Bit Acronym CP Offset Start: 34h Offset End: 34h Power Well: Core Bit Description Pointer to First Capability Structure: Value points to the configuration space offset of the first capability structure (MSI). Sticky Bit Reset Value Bit Access 40h RO The Capabilities Pointer field (Cap_Ptr) is an 8-bit field that provides an offset in the device's PCI configuration space for the location of the first item in the Capabilities Linked List (CLL). The Controller sets this bit and implements a capabilities list to indicate that it supports PCI power management, Message Signaled Interrupts (MSIs), and PCIe extended capabilities. Its value is 0x40, which is the address of the first entry: PCI power management. Intel(R) Communications Chipset 89xx Series - Datasheet 1506 October 2012 Order Number: 327879-001US 29.0 29.3.2.12 PIRQL[0:3]--PF Interrupt Line Register Table 29-28. PIRQL[0:3]--PF Interrupt Line Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 29.3.2.13 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 00h Offset Start: 3Ch Offset End: 3Ch Power Well: Core Bit Acronym Bit Description Sticky IRQL Interrupt Line: BIOS writes the interrupt routing information to this register to indicate which input of the interrupt controller this device is connected to. The device itself does not use this information. Bit Reset Value Bit Access 0h RW PIRQP0--PF Interrupt Pin Register 0 Read only register. For PCH implementation, interrupt mapping is fixed to LAN0 = INTB#, LAN1 = INTC#, LAN2 = INTD#, and LAN3 = INTB#. Table 29-29. PIRQP0--PF Interrupt Pin Register 0 Description: View: PCI Size: 8 bit Bit Range 07 : 00 Bus:Device:Function: B:D:1+ Index1 BAR: Configuration Default: 02h Bit Acronym IRQP October 2012 Order Number: 327879-001US Offset Start: 3Dh Offset End: 3Dh Power Well: Core Bit Description Interrupt Pin: Set as follows: PIRQ[0]/LAN0 = 02h (INTB#) PIRQ[1]/LAN1 = 03h (INTC#) PIRQ[2]/LAN2 = 04h (INTD#) PIRQ[3]/LAN3 = 02h (INTB#) Sticky Bit Reset Value Bit Access 02h RO Intel(R) Communications Chipset 89xx Series - Datasheet 1507 29.3.2.14 PIRQP1--PF Interrupt Pin Register 1 Read only register. For PCH implementation, interrupt mapping is fixed to LAN0 = INTB#, LAN1 = INTC#, LAN2 = INTD#, and LAN3 = INTB#. Table 29-30. PIRQP1--PF Interrupt Pin Register 1 Description: View: PCI Size: 8 bit Bit Range 07 : 00 29.3.2.15 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 03h Bit Acronym Power Well: Core Bit Description Sticky Interrupt Pin: Set as follows: PIRQ[0]/LAN0 = 02h (INTB#) PIRQ[1]/LAN1 = 03h (INTC#) PIRQ[2]/LAN2 = 04h (INTD#) PIRQ[3]/LAN3 = 02h (INTB#) IRQP Offset Start: 3Dh Offset End: 3Dh Bit Reset Value Bit Access 03h RO PIRQP2--PF Interrupt Pin Register 2 Read only register. For PCH implementation, interrupt mapping is fixed to LAN0 = INTB#, LAN1 = INTC#, LAN2 = INTD#, and LAN3 = INTB#. Table 29-31. PIRQP2--PF Interrupt Pin Register 2 Description: View: PCI Size: 8 bit Bit Range 07 : 00 Bus:Device:Function: B:D:1+ Index1 BAR: Configuration Default: 04h Bit Acronym IRQP Power Well: Core Bit Description Interrupt Pin: Set as follows: PIRQ[0]/LAN0 = 02h (INTB#) PIRQ[1]/LAN1 = 03h (INTC#) PIRQ[2]/LAN2 = 04h (INTD#) PIRQ[3]/LAN3 = 02h (INTB#) Intel(R) Communications Chipset 89xx Series - Datasheet 1508 Offset Start: 3Dh Offset End: 3Dh Sticky Bit Reset Value Bit Access 04h RO October 2012 Order Number: 327879-001US 29.0 29.3.2.16 PIRQP3--PF Interrupt Pin Register 3 Read only register. For PCH implementation, interrupt mapping is fixed to LAN0 = INTB#, LAN1 = INTC#, LAN2 = INTD#, and LAN3 = INTB#. Table 29-32. PIRQP3--PF Interrupt Pin Register 3 Description: View: PCI Size: 8 bit Bit Range 07 : 00 29.4 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 02h Power Well: Core Bit Acronym IRQP Offset Start: 3Dh Offset End: 3Dh Bit Description Sticky Interrupt Pin: Set as follows: PIRQ[0]/LAN0 = 02h (INTB#) PIRQ[1]/LAN1 = 03h (INTC#) PIRQ[2]/LAN2 = 04h (INTD#) PIRQ[3]/LAN3 = 02h (INTB#) Bit Reset Value Bit Access 02h RO PCI Capabilities The first entry of the PCI capabilities link list is pointed by the Cap_Ptr register. The following table describes the capabilities supported by the Controller. Address Item Next Pointer 0x40-47 PCI Power Management 0x50 0x50-67 Message Signaled Interrupt 0x70 0x70-8B Extended Message Signaled Interrupt 0xA0 0xA0-DB PCIe Capabilities 0xE0/0x001 0xE0-0xE7 Vital Product Data Capability 0x00 1. Next pointer is 0x00 if the VPD area in the EEPROM does not exist. In EEPROM-less mode, the PCIe capability is the last capabilities section. 29.4.1 Power Management Capability Structure All fields are reset on full power-up. All of the fields except PME_En and PME_Status are reset on exit from D3cold state. If aux power is not supplied, the PME_En and PME_Status fields also reset on exit from D3cold state. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1509 29.4.1.1 PPMCAP[0:3]--PF Power Management Capabilities ID Register EP supports the PCI bus Power Management Interface Specification Rev 1.2 based PCIe Power Management. The specification requires the implementation of the PCI Power Management Capabilities Register. Table 29-33. PPMCAP[0:3]--PF Power Management Capabilities ID Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 29.4.1.2 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 01h Bit Acronym PMID Offset Start: 40h Offset End: 40h Power Well: Core Bit Description Sticky Capability ID: PCI SIG assigned capability record ID (01h, Power Management capability) Bit Reset Value Bit Access 01h RO PPMCP[0:3]--PF Power Management Next Capability Pointer Register Table 29-34. PPMCP[0:3]--PF Power Management Next Capability Pointer Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 B:D:1+ Bus:Device:Function: Index1 BAR: Configuration Default: 50h Bit Acronym PMCP Power Well: Core Bit Description Next Capability Pointer: PCI Express Capability. Intel(R) Communications Chipset 89xx Series - Datasheet 1510 Offset Start: 41h Offset End: 41h Sticky Bit Reset Value Bit Access 50h RO October 2012 Order Number: 327879-001US 29.0 29.4.1.3 PPMC[0:3]--PF Power Management Capabilities Register This field describes the Controller's functionality at the power management states as described in the following table. Each device function has its own register. Table 29-35. PPMC[0:3]--PF Power Management Capabilities Register Description: View: PCI Size: 16 bit Bit Range 15 :11 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: C823hh Offset Start: 42h Offset End: 43h Power Well: Core Bit Acronym Bit Description Sticky PME PME_Support - This 5-bit field indicates the power states in which the function may assert PME#. A value of 0b for any bit indicates that the function is not capable of asserting the PME# signal while in that power state. bit(11) X XXX1b - PME# can be asserted from D0 bit(12) X XX1Xb - PME# can be asserted from D1 bit(13) X X1XXb - PME# can be asserted from D2 bit(14) X 1XXXb - PME# can be asserted from D3hot bit(15) 1 XXXXb - PME# can be asserted from D3cold Bit Reset Value Bit Access 11001b RO Value of this field is a function of Aux Pwr availability and Power Management (PM Ena) bit in Initialization Control Word 1 (word 0x0A) EEPROM word. Conditionaaaaaaaaa FunctionalityaaaaValue PM Dis in EEProm No PME at all states 00000b PM Ena & NoAux Pwr PME at D0 and D3hot a01001b PM Ena & Aux Pwr PME at D0, D3hot and D3cold 11001b 10 D2 D2_Support The Controller does not support D2 state. 0b RO 9 D1 D1_Support The Controller does not support D1 state. 0b RO 8 :6 AC AUX Current - Required current defined in the Data Register. Not supported by CC. 000b RO 5 DSI DSI The Controller requires its device driver to be executed following transition to the D0 uninitialized state. 1b RO 4 Reserved Reserved 0b RO 3 PMC PME_Clock Disabled. Hardwired to 0b. 0b RO 2 :0 PMV Version The Controller complies with the PCI PM specification, revision 1.2. 11b RO October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1511 29.4.1.4 PPMCSR--PF Power Management Control and Status Register This register is used to control and monitor power management events in the Controller. Each device function has its own PMCSR. Table 29-36. PPMCSR[0:3]--PF Power Management Control and Status Register Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: Default: 00000000h Bit Range Bit Acronym 31 :16 Reserved Offset Start: 44h Offset End: 47h Power Well: Core Bit Description Bit Reset Value Bit Access 0000h RO PMS 0b RW1C DSC Data Scale: Set to 0 because the EP does not implement the Data register. If the Data register has not been implemented, this field must return "00b" when the PMCSR is read. 00b RO 12 :9 DSEL Data Select: Set to 0 for EP because the EP does not implement the Data register (The Data register is an optional, 8-bit read-only register that provides a mechanism for the function to report state dependent operating data such as power consumed or heat dissipation) If the Data register is not implemented, this field should be read only and return "0000b" when the PMCSR is read. 0000b RO 88 PME PME Enable: This is driven through sideband signal. EP is required to generate PME. 0b RW 7 :4 Reserved 14 :13 3 NSR 2 Reserved 1 :0 PS Reserved/NA/Not Supported for EP. Sticky PME Status: This is driven through sideband signal. EP is required to generate PME. This bit is set to 1b when the function detects a wake-up event independent of the state of the PME_En bit. Writing a 1b clears this bit. 15 0000b RV No_Soft_Reset: This bit is always set to 0b to indicate that GbE performs an internal reset after a transition from D3hot to D0 via software control of the PowerState bits. Configuration context is lost when performing the soft reset. After transitioning from the D3hot to the D0 state, full re-initialization sequence is needed to return GbE to D0 Initialized. Reserved 0b RO Reserved 0b RO 00b RW Power State. This field is used to determine the current power state of EP and to set a new power state. * 00: D0 * 01: D1 (Not supported by EP, ignore writes with this value) * 10: D2 (Not supported by EP, ignore writes with this value) * 11: D3_hot If software selects a Power state that is not supported by the EP (D2/D1), the writes must complete normally on PCIe, but the write data is discarded and no state change occurs. The GbE also supports D3_cold. Intel(R) Communications Chipset 89xx Series - Datasheet 1512 B:D:1+ Index1 October 2012 Order Number: 327879-001US 29.0 29.4.2 MSI Capability Structure This structure is required for PCIe devices. Byte Offset 0x50 Byte 3 Message Control (0x0080) Byte 1 Byte 0 Next Pointer (0x70) Capability ID (0x05) 0x54 Message Address 0x58 Message Upper Address 0x5C 29.4.2.1 Byte 2 Reserved Message Data 0x60 Mask bits 0x64 Pending bits PMSICID[0:3]--Message Signalled Interrupt Capability ID Register The Message Signalled Interrupt Capability record defines how the device generates PCI MSI messages. It is an 10B PCI SIG-defined capability record and includes the MCID, MCP, MCTL, MADR, and MDATA fields of the configuration header. Table 29-37. PMSICID[0:3] - Message Signalled Interrupt Capability ID Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 29.4.2.2 BAR: Configuration B:D:1+ Bus:Device:Function: Index1 Default: 05h Bit Acronym MCID Offset Start: 50h Offset End: 50h Power Well: Core Bit Description Sticky Capability ID: PCI SIG assigned capability record ID (05h, MSI capability) Bit Reset Value Bit Access 05h RO PMSINCP[0:3]--Message Signalled Interrupt Next Capability Pointer Register Table 29-38. PMSINCP[0:3]--Message Signalled Interrupt Next Capability Pointer Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 70h Bit Acronym MCP October 2012 Order Number: 327879-001US Offset Start: 51h Offset End: 51h Power Well: Core Bit Description Sticky Next Capability Pointer: MSI-X capability. Bit Reset Value Bit Access 70h RO Intel(R) Communications Chipset 89xx Series - Datasheet 1513 29.4.2.3 PMSICTL[0:3]--Message Signalled Interrupt Control Register Table 29-39. PMSICTL[0:3]--Message Signalled Interrupt Control Register Description: View: PCI Size: 16 bit Bit Range 15 : 09 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0180h Power Well: Core Bit Reset Value Bit Access Reserved 0h RO Bit Acronym Reserved Offset Start: 52h Offset End: 53h Bit Description Sticky 08 MC Per-Vector Masking Capable: Hard-wired to 1. A value of 1b indicates that GbE is capable of per-vector masking. 1b RO 07 C64 64 bit Address Capable: Hardwired to 1 to indicate the device generate 64b message addresses. 1b RO 06 : 04 MME Multiple Message Enable: System software writes to this field to indicate the number of allocated messages (less than or equal to the number of requested messages in MMC). A value of 0 corresponds to one message. 000h RW 03 : 01 MMC Multiple Message Capable: System software reads this field to determine the number of requested messages. Hardwired to 0 to request one message. 000h RO MSIE MSI Enable: System software sets this bit to enable MSI signaling. A device driver is prohibited from writing this bit to mask a device's service request. If 1, the device can use an MSI to request service. If 0, the device cannot use an MSI to request service. 0b RW 00 29.4.2.4 PMSILADDR[0:3]--Message Signalled Interrupt Lower Address Register Table 29-40. PMSILADDR[0:3]--Message Signalled Interrupt Lower Address Register Description: View: PCI Size: 32 bit Bus:Device:Function: B:D:1+ Index1 BAR: Configuration Default: 00000000h Power Well: Core Bit Range Bit Acronym Bit Description 31 :02 ADDR Lower Message Address: Written by the system to indicate the lower 30-bits of the address to use for the MSI memory write transaction. The lower two bits will always be written as 0. 01 :00 Reserved Reserved Intel(R) Communications Chipset 89xx Series - Datasheet 1514 Offset Start: 54h Offset End: 57h Sticky Bit Reset Value Bit Access 0h RW 00b RV October 2012 Order Number: 327879-001US 29.0 29.4.2.5 PMSIUADDR[0:3]--Message Signalled Interrupt Upper Address Register Table 29-41. PMSIUADDR[0:3]--Message Signalled Interrupt Upper Address Register Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: Default: 00000000h Offset Start: 58h Offset End: 5Bh Power Well: Core Bit Range Bit Acronym Bit Description 31 :00 ADDR Upper Message Address: Written by the system to indicate the upper 30-bits of the address to use for the MSI memory write transaction. 29.4.2.6 B:D:1+ Index1 Sticky Bit Reset Value Bit Access 0h RW PMSIDATA[0:3]--Message Signalled Interrupt Data Register Table 29-42. PMSIDATA[0:3]--Message Signalled Interrupt Data Register Description: View: PCI Size: 16 bit Bit Range 15 : 00 29.4.2.7 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0000h Offset Start: 5Ch Offset End: 5Dh Power Well: Core Bit Acronym Bit Description Sticky DATA Message Data: Written by the system to indicate the lower 16 bits of the data written in the MSI memory write DWORD transaction. The upper 16 bits of the transaction are written as 0. Bit Reset Value Bit Access 0h RW PMSIMSK[0:3]--Message Signalled Interrupt Mask Register The Mask Bits and Pending Bits registers enable software to disable or defer message sending on a per-vector basis. As the Controller supports only one message, only bit 0 of these register is implemented. Table 29-43. PMSIMSK[0:3]--Message Signalled Interrupt Mask Register Description: View: PCI Size: 16 bit Bit Range 15 : 01 00 : 00 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0000h MSK0 October 2012 Order Number: 327879-001US Power Well: Core Bit Reset Value Bit Access Reserved 0h RV MSI Vector 0 Mask If set, the Controller is prohibited from sending MSI messages. 0h RW Bit Acronym Reserved Offset Start: 60h Offset End: 63h Bit Description Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 1515 29.4.2.8 PMSIPND[0:3]--Message Signalled Interrupt Mask Pending Register Table 29-44. PMSIPND[0:3]--Message Signalled Interrupt Pending Register Description: View: PCI Size: 16 bit Bit Range BAR: Configuration B:D:1+ Index1 Default: 0000h Reserved 00 : 00 PND0 Offset Start: 64h Offset End: 67h Power Well: Core Bit Reset Value Bit Access Reserved 0h RV If set, the Controller has a pending MSI message. 0h RO Bit Acronym 15 : 01 29.4.3 Bus:Device:Function: Bit Description Sticky MSI-X Configuration More than one MSI-X capability structure per function is prohibited, but a function is permitted to have both an MSI and an MSI-X capability structure. In contrast to the MSI capability structure, which directly contains all of the control/ status information for the function's vectors, the MSI-X capability structure instead points to an MSI-X table structure and a MSI-X Pending Bit Array (PBA) structure, each residing in memory space. Each structure is mapped by a Base Address Register (BAR) belonging to the function, located beginning at 0x10 in configuration space. A BAR Indicator Register (BIR) indicates which BAR, and a Qword-aligned offset indicates where the structure begins relative to the base address associated with the BAR. The BAR is permitted to be either 32-bit or 64-bit, but must map to memory space. A function is permitted to map both structures with the same BAR, or to map each structure with a different BAR. The MSI-X table structure, typically contains multiple entries, each consisting of several fields: message address, message upper address, message data, and vector control. Each entry is capable of specifying a unique vector. The PBA structure, described in the same section, contains the function's pending bits, one per Table entry, organized as a packed array of bits within Qwords. The last Qword might not be fully populated. To request service using a given MSI-X table entry, a function performs a Dword memory write transaction using: * The contents of the Message Data field entry for data. * The contents of the Message Upper Address field for the upper 32 bits of the address. * The contents of the Message Address field entry for the lower 32 bits of the address. A memory read transaction from the address targeted by the MSI-X message produces undefined results. The MSI-X table and MSI-X PBA are permitted to co-reside within a naturally aligned 4 KB address range, though they must not overlap with each other. Intel(R) Communications Chipset 89xx Series - Datasheet 1516 October 2012 Order Number: 327879-001US 29.0 MSI-X table entries and Pending bits are each numbered 0 through N-1, where N-1 is indicated by the Table Size field in the MSI-X Message Control register. For a given arbitrary MSI-X table entry K, its starting address can be calculated with the formula: Entry starting address = Table base + K*16 For the associated Pending bit K, its address for Qword access and bit number within that Qword can be calculated with the formulas: Qword address = PBA base + (K div 64)*8 Qword bit# = K mod 64 Software that chooses to read Pending bit K with Dword accesses can use these formulas: Dword address = PBA base + (K div 32)*4 Dword bit# = K mod 32 The Controller also supports the table-less MSI-X mode, where a single interrupt vector is provided. The MSI-X table and MSI-X PBA are not used. Instead, the capability structure includes several additional fields (Message Address, Message Address Upper, and Message Data) for vector configuration. The Controller embeds the number of the original MSI-X vectors (i.e. the vectors supported if the number of vectors was not limited to 1) in the LSB bits of the Message Data field. Table 29-45. MSI-X Capability Structure Byte Offset 0x70 29.4.3.1 Byte 3 Byte 2 Message Control (0x00090) Byte 1 Byte 0 Next Pointer (0xA0) Capability ID (0x11) 0x74 Table Offset 0x78 PBA offset PMSI-X[0:3]--PF Message Signalled Interrupt X Capability ID Register MSI-X Capability ID Register indicates that the device is capable of generating MSI-X Interrupts. Table 29-46. PMSI-X[0:3]--PF Message Signalled Interrupt X Capability ID Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 11h Bit Acronym MSIX October 2012 Order Number: 327879-001US Offset Start: 70h Offset End: 70h Power Well: Core Bit Description Sticky Capability ID: PCI SIG assigned capability record ID (11h, MSI-X capability). 11h identifies the EP as a device that is MSI-X capable. Bit Reset Value Bit Access 11h RO Intel(R) Communications Chipset 89xx Series - Datasheet 1517 29.4.3.2 PMSIXNCP[0:3]--PF MSIX Next Capability Pointer Register Table 29-47. PMSIXNCP[0:3]--PF MSIX Next Capability Pointer Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 29.4.3.3 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: A0h Bit Acronym MCP Offset Start: 71h Offset End: 71h Power Well: Core Bit Description Sticky Next Capability Pointer: Power Management Capability Bit Reset Value Bit Access A0h RO PMSIXCNTL[0:3]--PF Message Signalled Interrupt X Control Register Table 29-48. PMSIXCNTL[0:3]--PF Message Signalled Interrupt X Control Register Description: View: PCI Size: 16 bit Bit Range 15 14 Bus:Device:Function: B:D:1+ Index1 BAR: Configuration Default: 0009h Bit Acronym MSIXEN FM 13 :11 Reserved 10 :00 TS Power Well: Core Bit Reset Value Bit Access MSI-X Enable: This bit enables the EP to generates interrupts using the MSI-X tables instead of the legacy INTx messages. When this bit is set to 1, the EP will not generate INTx messages and must use the MSI-X to signal interrupts. The device driver should not clear this bit to mask interrupts. This bit will be set by the system PCI device manager. 0h RW Function Mask: This bit controls the masking of all vectors implemented in the EP. When this bit is 0, each vectors mask bit determines whether the vector is masked or not. When this bit is 1, all vectors are masked regardless of the per vector masking bit. 0h RW Always returns 0's when read. 0h RO MSI-X Table Size: Number of vectors (encoded as N-1) supported by the EP. The EP supports 10 vectors. 9h RO Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 1518 Offset Start: 72h Offset End: 73h Sticky October 2012 Order Number: 327879-001US 29.0 29.4.3.4 PMSIXTBIR[0:3]--PF MSI-X Table Offset & Table BIR Register Table 29-49. PMSIXTBIR[0:3]--PF MSI-X Table Offset & Table BIR Register Description: View: PCI Size: 32 bit Bit Range 31 :3 2 :0 BAR: Configuration B:D:1+ Index1 Default: 00000003h Bit Description TO Table Offset: Offset to a location in one of the EP's BAR's (indicated by TBIR) that points to the location of the base of the MSI-X Table. The MSI-X Table is mapped to offset 0KB of the GbE BAR3 or BAR4. Sticky Table BAR Indicator Register: The BIR is mapped in the GbE BAR3 or BAR4. TBIR Offset Start: 74h Offset End: 77h Power Well: Core Bit Acronym Note: 29.4.3.5 Bus:Device:Function: Bit Reset Value Bit Access 0h RO 11b RO Reset value controls by EEPROM. PMSIXPBABIR[0:3]--PF MSI-X Pending Bit Array & BIR Offset Register Table 29-50. PMSIXPBABIR[0:3]--PF MSI-X Pending Bit Array & BIR Offset Register Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 00002003h Power Well: Core Bit Range Bit Acronym Bit Description 31 :3 PBAO Pending Bit Array Offset: Offset to a location in one of the EP BAR's (indicated by PBABIR) that points to the location of the base of the MSI-X PBA. The MSI-X PBA maps to offset 8KB of GbE BAR3 or BAR4. 2 : 00 PBABIR October 2012 Order Number: 327879-001US Sticky Pending Bit Array BAR Indicator Register: The MSI-X PBA is mapped in GbE BAR3 or BAR4. Note: Offset Start: 78h Offset End: 7Bh Bit Reset Value Bit Access 000000000000 000000100000 00000b RO 11b RO Reset value controls by EEPROM. Intel(R) Communications Chipset 89xx Series - Datasheet 1519 29.4.4 CSR Access Via Configuration Address Space 29.4.4.1 IOADDR[0:3]--IOADDR Register Read/write register. Each function has its own IOADDR register. Functionality is the same in all functions. Register is cleared at Power-up (LAN_PWR_GOOD) or PCIe reset. Table 29-51. IOADDR[0:3]--IOADDR Register Description: View: PCI BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Offset Start: 98h Offset End: 9Bh View: PCI BAR: GbEPCIBAR2 Bus:Device:Function: B:D:1+ Index1 Offset Start: 00h Offset End: 03h Size: 32 bit Bit Range 31 30 :0 Default: 0h Power Well: Core Bit Reset Value Bit Access Configuration IO Access Enable. 0b - CSR configuration read or write disabled. 1b - CSR Configuration read or write enabled When bit is set accesses to the IODATA register actually generate transactions to the device. Otherwise, accesses to the IODATA register are don't-cares (write are discarded silently, reads return arbitrary results). 0b RW1 Internal Register or Internal Memory location Address. 0x00000-0x1FFFF - Internal Registers and Memories 0x20000-0x7FFFFFFF - Undefined 0h RW1 Bit Acronym Bit Description EN ADDR Sticky 1. In the event that the CSR_conf_en bit in the PCIe Init Configuration 2 EEPROM word is cleared, accesses to the IOADDR register via configuration address space is ignored and has no effect on the register and the CSRs referenced by the IOADDR register. 29.4.4.2 IODATA[0:3]--IODATA Register This is a read/write register. Each function has its own IODATA register. Functionality is the same in all functions. Register is cleared at Power-up (LAN_PWR_GOOD) or PCIe reset. Table 29-52. IODATA[0:3]--IODATA Register Description: View: PCI BAR: Configuration B:D:1+ Bus:Device:Function: Index1 Offset Start: 9Ch Offset End: 9Fh View: PCI BAR: GbEPCIBAR2 Bus:Device:Function: B:D:1+ Index1 Offset Start: 04h Offset End: 07h Size: 32 bit Default: 0h Bit Range Bit Acronym 31 :0 DATA Power Well: Core Bit Description Data field for reads or writes to the Internal Register location as identified by the current value in IOADDR. All 32 bits of this register are read/write-able. Sticky Bit Reset Value Bit Access 0h RW1 1. In the event that the CSR_conf_en bit in the PCIe Init Configuration 2 EEPROM word is cleared, access to the IODATA register via configuration address space is ignored and has no effect on the register and the CSRs referenced by the IOADDR register. Intel(R) Communications Chipset 89xx Series - Datasheet 1520 October 2012 Order Number: 327879-001US 29.0 29.4.5 PCIe Configuration Registers PCIe provides two mechanisms to support native features: * PCIe defines a PCI capability pointer indicating support for PCIe. * PCIe extends the configuration space beyond the 256 bytes available for PCI to 4096 bytes. The Controller implements the PCIe capability structure for endpoint devices as follows: Byte Offset Byte 3 Byte 2 PCI Express Capability Register (0x0002) 0xA0 0xA4 Byte 1 Byte 0 Next Pointer (0xE0) Capability ID (0x10) Device Capability 0xA8 Device Status Device Control 0xAC Link Capability 0xB0 Link Status Link Control 0xB4 Reserved 0xB8 Reserved Reserved 0xBC Reserved 0xC0 Reserved Reserved 0xC4 Device Capabilities 2 0xC8 Reserved Device Control 2 0xCC Reserved 0xD0 Link Status 2 Link Control 2 0xD4 Reserved 0xD8 Reserved Reserved 29.4.6 PCI Express Capability Structure 29.4.6.1 PPCID[0:3]--PF PCI Express Capability ID Register The PCI Express Capability List register enumerates the PCI Express Capability structure in the PCI 3.0 configuration space capability list. Table 29-53. PPCID[0:3]--PF PCI Express Capability Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 10h Bit Acronym PCIECID October 2012 Order Number: 327879-001US Offset Start: A0h Offset End: A0h Power Well: Core Bit Description Sticky Capability ID: PCI SIG assigned capability record ID (10h, PCI Express capability) Bit Reset Value Bit Access 10h RO Intel(R) Communications Chipset 89xx Series - Datasheet 1521 29.4.6.2 PPCP[0:3]--PF PCI Express Next Capability Pointer Register Table 29-54. PPCP[0:3]--PF PCI Express Next Capability Pointer Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 29.4.6.3 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 00h Offset Start: A1h Offset End: A1h Power Well: Core Bit Acronym Bit Description PCIENP Next Capability Pointer: Offset to the next capability item in the capability list. Its value of 0xE0 points to the VPD structure. If VPD is disabled, a value of 0x00 value indicates that it is the last item in the capability-linked list. VPD enable is a sideband signal from GbE to EP. Value loaded by EEPROM. Sticky Bit Reset Value Bit Access 00h RO PPCR[0:3]--PF PCI Express Capabilities Register Table 29-55. PPCR[0:3]--PF PCI Express Capabilities Register Description: View: PCI Size: 16 bit Bit Range 15 : 14 Bus:Device:Function: B:D:1+ Index1 BAR: Configuration Default: 0002h Bit Acronym Power Well: Core Bit Description Reserved Bit Reset Value Bit Access 00b RO IMN 00000b RO 8 SI Slot Implemented: This bit when set indicates that the PCI Express Link associated with this port is connected to a slot. Hardwired to 0 for EP. 0b RO 7 :4 DPT Device/Port Type: Indicates the type of PCI Express logical device. Hardwired to 0000b (PCIe Endpoint) 0000b RO 3 :0 CV Capability Version: Indicates PCI-SIG defined PCI Express capability structure version number. EP is PCIe 2.0 Specification Compliant. 0010b RO Intel(R) Communications Chipset 89xx Series - Datasheet 1522 Sticky Interrupt Message Number: This field indicates which MSI vector is used for the interrupt message generated in association with the status bits in either the Slot Status field of this capability structure (applies to only RC or Switch) Not applicable to EP. 13 :9 Reserved Offset Start: A2h Offset End: A3h October 2012 Order Number: 327879-001US 29.0 29.4.6.4 PPDCAP[0:3]--PF PCI Express Device Capabilities Register Table 29-56. PPDCAP[0:3]--PF PCI Express Device Capabilities Register Description: View: PCI Size: 32 bit BAR: Configuration B:D:1+ Index1 Default: 10008041h Bit Range Bit Acronym 31 :29 Reserved 28 Bus:Device:Function: FLR Offset Start: A4h Offset End: A7h Power Well: Core Bit Reset Value Bit Access Reserved 0h RO Function level reset: When set indicates that the device supports the FLR feature. When SR-IOV is enabled both PF & VF have to support FLR Note: Loaded from Bit 10 (FLR capability enable) in PCIE Control2 (offset 0x28) 1b RO Bit Description Sticky 27 :26 CSPS Captured Slot Power Limit Scale: Does not apply to EP. 00b RO 25 :18 CSPV Captured Slot Power Limit Value: Does not apply to EP. 0h RO 17 :16 Reserved Reserved 0h RO RBEP Role-Based Error Reporting: Indicates that EP conforms to Role based error reporting ECN for PCIe 1.0a and which was subsequently rolled in PCIe 1.1 and future revisions. 1b RO 14 :12 ATTN Attention Button/Indicator Present and Power Indicator Present. None of these are implemented in the EP. 000b RO 11 :9 EL1L Endpoint L1 Acceptable Latency: This field indicates the acceptable latency that an Endpoint can withstand due to the transition from L1 state to the L0 state. Hardcoded to the lowest value of 1us. 000b R0 8 :6 EL0L Endpoint L0s Acceptable Latency: This field indicates the acceptable latency that an Endpoint can withstand due to the transition from L1 state to the L0 state. This value is hardcoded to the latency of 64-128ns. 001b R0 5 ETFS Extended Tag Field Supported: This field indicates the maximum supported size of the Tag field as a Requester. When Clear indicates 5-bit Tag field is supported. This limits the EP to maximum of 32 outstanding requests. 0b R0 4 :3 PFS Phantom Functions Supported: This field indicates the support for use of unclaimed function numbers to extend the number of outstanding transactions allowed. EP does not use this capability. 00b RO MPS Max_Payload_Size Supported: This field indicates the maximum payload size that EP can support for TLPs. This value is set to indicate 256B. The defined encodings are: * 000b = 128B max payload size * 001b = 256 bytes max payload size (Max supported) * 010b = 512 bytes max payload size * 011b = 1KB max payload size * 100b = 2KB max payload size * 101b = 4KB max payload size 001b RO 15 2 :0 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1523 29.4.6.5 PPDCNTL[0:3]--PF PCI Express Device Control Register Table 29-57. PPDCNTL[0:3]--PF PCI Express Device Control Register (Sheet 1 of 2) Description: View: PCI Size: 16 bit Bit Range 15 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 2810h Offset Start: A8h Offset End: A9h Power Well: Core Bit Acronym Bit Description STARTFLR Initiate FLR - Used to initiate FLR transition. A write of 1 initiates FLR transition. Since hardware must not respond to any cycles till FLR completion, the value read by software from this bit is 0. Sticky Bit Reset Value Bit Access 0b RW 10b RW 1b RW Max_Read_Request_Size: This field sets the maximum Read Request size for the EP as a Requester. The EP is capable for generating up to 1kB read requests. However requests generated by the EP will be limited by the programmed value in this field. 14 :12 MRS Defined encodings for this field are: 000b 128B max read request size 001b 256B max read request size 010b 512B max read request size (Default) 011b 1024B max read request size 100b 2048B max read request size 101b 4096B max read request size 11 ENS Enable No Snoop (NS): If this bit is set to 1, EP is permitted to set the No Snoop bit in the Requester Attributes of transactions it initiates. When clear all transactions will have the No Snoop bit clear. Setting this bit will not cause the EP to set the No Snoop attribute on every memory requests that it initiates. 10 APME Auxiliary (AUX) Power PM Enable: This bit when set enables a device to draw AUX power independent of PME AUX power. 0b RWS 9 PFE Phantom Functions Enable: When set, this bit enables a device to use unclaimed functions as Phantom Functions. Does not apply to EP. 0b RO 8 ETFE Extended Tag Field Enable: When set, this bit enables a device to use an 8-bit Tag field as a requester. Does not to EP. 0b RO 000b RW Max_Payload_Size: This field sets maximum TLP payload for EP. As a Receiver, the EP must handle TLPs as large as the set value; as a Transmitter, the EP must not generate TLPs exceeding the set value. The EP is capable of generating up to 265B MPS. However requests generated by the EP will be limited by the programmed value in this field. 7 :5 MPS It is expected that the root complexes that the EP will be paired up with will be limited to 128B and 256B MPS. 000b = 128B (Default) 001b = 256B 010b = 512B 011b = 1kB Others values not supported by EP Intel(R) Communications Chipset 89xx Series - Datasheet 1524 October 2012 Order Number: 327879-001US 29.0 Table 29-57. PPDCNTL[0:3]--PF PCI Express Device Control Register (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range 4 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 2810h Offset Start: A8h Offset End: A9h Power Well: Core Bit Reset Value Bit Access Enable Relaxed Ordering: If this bit is set, EP is permitted to set the Relaxed Ordering bit in the Attributes field of transactions it initiates. Setting this bit does not cause the EP to set the RO on every transaction it issues 1b RW 0b RW Bit Acronym Bit Description ERO Sticky 3 URRO Unsupported Request Reporting Enable: This bit, in conjunction with other bits, controls the signaling of Unsupported Requests by sending Error Messages to the root port. When clear it disables sending of error messages. 2 FERE Fatal Error Reporting Enable: This bit, in conjunction with other bits, controls sending ERR_FATAL Messages to the root port. When clear disables sending of error messages. 0h RW 1 NERE Non-Fatal Error Reporting Enable: This bit, in conjunction with other bits, controls sending ERR_NONEATAL Messages to the root port. When clear disables sending of error messages. 0h RW 0 CERE Correctable Error Reporting Enable: This bit, in conjunction with other bits, controls sending ERR_COR Messages to the root port. When clear disables sending of error messages. 0h RW 29.4.6.6 PPDSTAT[0:3]--PF PCI Express Device Status Register Table 29-58. PPDSTAT[0:3]--PF PCI Express Device Status Register (Sheet 1 of 2) Description: View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0010h Bit Range Bit Acronym 15 :6 Reserved Offset Start: AAh Offset End: ABh Power Well: Core Bit Reset Value Bit Access 0h RO Transactions Pending: This bit when set indicates that EP has issued Non-Posted Requests which have not been completed either with a completion packet or completion timeout mechanism. 0h RO Bit Description Sticky 5 TP 4 APD AUX Power Detected: Devices that require AUX power report this bit as set if AUX power is detected by the device. Default value is set based on an external pin strap. In addition an EEPROM bit can gate this advertisement on a per function basis. 1b RO URD Unsupported Request Detected: This bit indicates that EP received an Unsupported Request. A one indicates that an error was detected since the last time this bit was cleared. Errors are updated in this field regardless of whether the error reporting bit is set in the Device Control Register. 0h RW1C 3 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1525 Table 29-58. PPDSTAT[0:3]--PF PCI Express Device Status Register (Sheet 2 of 2) Description: View: PCI Size: 16 bit Bit Range 2 1 0 29.4.6.7 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0010h Offset Start: AAh Offset End: ABh Power Well: Core Bit Reset Value Bit Access FED Fatal Error Detected: This bit indicates status of Fatal errors detected. A one indicates that an error was detected since the last time this bit was cleared. Errors are updated in this field regardless of whether the error reporting bit is set in the Device Control Register. 0h RW1C NED Non-Fatal Error Detected: This bit indicates status of Nonfatal errors detected. A one indicates that an error was detected since the last time this bit was cleared. Errors are updated in this field regardless of whether the error reporting bit is set in the Device Control Register. 0h RW1C CED Correctable Error Detected: This bit indicates status of correctable errors detected. A one indicates that an error was detected since the last time this bit was cleared. Errors are updated in this field regardless of whether the error reporting bit is set in the Device Control Register. 0h RW1C Bit Acronym Bit Description Sticky PLCAPR[0:3]--PF Link Capabilities Register Table 29-59. PLCAPR[0:3]--PF Link Capabilities Register (Sheet 1 of 2) Description: Link Capabilities Register View: PCI Size: 32 bit B:D:1+ Bus:Device:Function: Index1 BAR: Configuration Default: 3B502h Power Well: Core Bit Reset Value Bit Access 0h RWOS 00b RO 0b RO Does not apply to EP 000b RO L1 Exit Latency- Indicates the exit latency from L1 to L0 state. EP does not support L1 transition 000b - Less than 1 is 001b - 1 is - 2 is 010b - 2 is - 4 is 011b - 4 is - 8 is 100b - 8 is - 16 is 101b - 16 is - 32 is 110b - 32 is - 64 is 111b - L1 transition not supported 111b RO Bit Range Bit Acronym 31 :24 PORTNUM Port Number: Assigned by EP after link training phase. 23 :22 Reserved Reserved 21 :21 LBN 20 :18 NA 17 :15 L1EL Bit Description Link Bandwidth Notification Capability: Does not apply to endpoints/EP Intel(R) Communications Chipset 89xx Series - Datasheet 1526 Offset Start: ACh Offset End: AFh Sticky October 2012 Order Number: 327879-001US 29.0 Table 29-59. PLCAPR[0:3]--PF Link Capabilities Register (Sheet 2 of 2) Description: Link Capabilities Register View: PCI Size: 32 bit Bit Range 14 :12 11 :10 9 :4 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 3B502h Offset Start: ACh Offset End: AFh Power Well: Core Bit Reset Value Bit Access L0EL L0s Exit Latency- Indicates the exit latency from L0s to L0 state. 000b - Less than 64ns 001b - 64ns - 128ns 010b - 128ns - 256ns 011b - 256ns - 512ns 100b - 512ns - 1 is 101b - 1 is - 2 is 110b - 2 is - 4 is 111b - Reserved 011b RO ASLPM Active State Link PM Support - Indicates the level of active state power management supported in EP. Defined encodings are: 00b - Reserved 01b - L0s Entry Supported 10b - Reserved 11b - L0s and L1 Supported 01b RO 010000b RO 0010b RO Bit Acronym LINKW Bit Description Sticky Max Link Width- Indicates the max link width. Relevant encoding: 000000b - Reserved 000001b - x1 000010b - x2 000100b - x4 001000b - x8 001100b - x12 010000b - x16 100000b - x32 EP value depends on SKU. However the max link width is x16. Max Link Speed - Indicates Maximum supported Link Speed. Defined encodings are: 3 :0 MAXSPEED 0001b - 2.5Gbs Link speed supported (Gen 1) 0010b - 5.0Gbs Link speed supported (Gen 2) EP indicates a max Link Speed of 5.0 Gbs. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1527 29.4.6.8 PLCNTLR[0:3]--PF Link Control Register Table 29-60. PLCNTLR[0:3]--PF Link Control Register (Sheet 1 of 2) Description: Link Control Register View: PCI Size: 16 bit BAR: Configuration Bus:Device:Function: Default: 0000h Offset Start: B0h Offset End: B1h Power Well: Core Bit Reset Value Bit Access 0000b RO LBWINTE Link Autonomous Bandwidth Interrupt Enable - When set to 1b this bit enables the generation of an interrupt to indicate that the Link Autonomous Bandwidth Status bit has been set. Not applicable to EP. 0b RO LBWMINTE Link Bandwidth Management Interrupt Enable - When set to 1b this bit enables the generation of an interrupt to indicate that the Link Bandwidth Management Status bit has been set. Not applicable to EP 0b RO 9 WD Hardware Autonomous Width Disable - When set to 1b this bit disables hardware from changing the link width for reasons other than attempting to correct unreliable link operation by reducing link width. EP will not support this feature. 0b RO 8 ECLKPM Enable Clock Power Management - not supported in EP 0b RO 0b RW 0b RW Bit Range Bit Acronym 15 :12 Reserved Reserved 11 10 7 EXTSYNC Bit Description Extended Sync. When set to one, this bit forces the transmission of: - 4096 FTS ordered sets during the L0s state - followed by a single SKP ordered set prior to entering the L0 state, - as well as the transmission of 1024 TS1 ordered sets in the L1 state prior to entering the Recovery state. This mode gives external devices (e.g., logic analyzers) that may be monitoring Link activity time to achieve bit and symbol lock before the Link enters the L0 or Recovery state and resumes communication. Note: 6 CCLKCFG Sticky This control applies to all functions. Common Clock Configuration - when set indicates that EP and the root port at the other end of the link are operating with a common reference clock. A value of 0 indicates that they operating with an asynchronous clock. This parameter affects the L0s Exit Latencies. After changing the value in this bit in both components on a Link, software must trigger the Link to retrain by writing a 1b to the Retrain Link bit. Note: This control applies to all functions. 5 RETRAIN Retrain Link: A write of 1b to this bit initiates Link retraining by directing the Physical Layer LTSSM to the Recovery state. Does not apply to endpoint/EP 0b RO 4 LINKDIS Link Disable: This bit disables the Link by directing the LTSSM to the Disabled state when set to 1b. Does not apply to endpoint/EP 0b RO Intel(R) Communications Chipset 89xx Series - Datasheet 1528 B:D:1+ Index1 October 2012 Order Number: 327879-001US 29.0 Table 29-60. PLCNTLR[0:3]--PF Link Control Register (Sheet 2 of 2) Description: Link Control Register View: PCI Size: 16 bit Bit Range BAR: Configuration Bus:Device:Function: Default: 0000h Bit Reset Value Bit Access Read Completion Boundary (RCB): For PCIe Endpoints this field is set optionally by configuration software to indicate the RCB value of the Root Port upstream from the Endpoint. - 0b = 64B - 1b = 128 byte Does not directly impact the EP. 0b RW Reserved 0b RO 00b RW Bit Description 3 RCB 2 Reserved ASPMC Sticky Active State Link PM Control - this field controls the active state PM supported on the link. Link PM functionality is determined by the lowest common denominator of all functions. Defined encodings are: 00b - PM Disabled 01b - L0s Entry Supported 10b - Reserved 11b - L0s and L1 Supported Note: 29.4.6.9 Offset Start: B0h Offset End: B1h Power Well: Core Bit Acronym 1 :0 B:D:1+ Index1 This control applies to all functions. PLSR[0:3]--PF Link Status Register Table 29-61. PLSR[0:3]--PF Link Status Register (Sheet 1 of 2) Description: Link Status Register View: PCI Size: 16 bit Bit Range BAR: Configuration Default: 0102h Reserved 14 LBWMS Bit Description 12 DLACT SCLKCFG Link Bandwidth Management Status - not supported in EP. Note: LTINPROG October 2012 Order Number: 327879-001US This status is reflected in all functions. Slot Clock Configuration - When set indicates that EP uses the physical reference clock that the platform provides on the connector. This bit must be cleared if EP uses an independent clock. Bit Access 0b RO 0b RO 0b RO 0b RWOS 0b RO This status is reflected in all functions. Link Training - Indicates that link training is in progress. Does not apply to EP/Endpoint. Note: Bit Reset Value This status is reflected in all functions. Data Link Layer Link Active - not supported in EP. Note: 11 Sticky Reserved Note: Offset Start: B2h Offset End: B3h Power Well: Core Bit Acronym 15 13 Bus:Device:Function: B:D:1+ Index1 This status is reflected in all functions. Intel(R) Communications Chipset 89xx Series - Datasheet 1529 Table 29-61. PLSR[0:3]--PF Link Status Register (Sheet 2 of 2) Description: Link Status Register View: PCI Size: 16 bit Bit Range 10 BAR: Configuration B:D:1+ Index1 Default: 0102h LTE NLW Bit Description Link Training Error - Indicates that a link training error has occurred. Does not apply to EP. Sticky Bit Reset Value Bit Access 0b RO 010000b RO 0010b RO This status is reflected in all functions. Negotiated Link Width: Negotiated Link Width Indicates the negotiated width of the link. Relevant encoding for EP are: 000001b - x1 000010b - x2 000100b - x4 001000b - x8 010000b - x16 Note: Offset Start: B2h Offset End: B3h Power Well: Core Bit Acronym Note: 9 :4 Bus:Device:Function: This status is reflected in all functions. Negotiated Link Speed: The negotiated Link Speed. Defined encodings are: 3 :0 NLS 0001b - 2.5Gbs 0010b - 5.0Gbs Note: This status is reflected in all functions. Intel(R) Communications Chipset 89xx Series - Datasheet 1530 October 2012 Order Number: 327879-001US 29.0 29.4.6.10 PDCAPR2[0:3]--PF Device Capabilities 2 Register Table 29-62. PDCAPR2[0:3]--PF Device Capabilities 2 Register Description: Device Capabilities 2 Register View: PCI Size: 32 bit BAR: Configuration Offset Start: C4h Offset End: C7h Power Well: Core Bit Reset Value Bit Access Reserved. 0b RV CTODS Completion Timeout Disable Supported. A value of 1b indicates support for the completion timeout disable mechanism. 1b RO CTORS Completion Timeout Ranges Supported: This field indicates support for the optional completion timeout programmability mechanism. This mechanism enables system software to modify the completion timeout value. Four time value ranges are defined: * Range A = 50 us to 10 ms * Range B = 10 ms to 250 ms * Range C = 250 ms to 4 s * Range D = 4 s to 64 s Bits are set according to the following table to show the timeout value ranges that are supported. * 0000b = Completion timeout programming not supported. * 0001b = Range A. * 0010b = Range B. * 0011b = Ranges A & B. * 0110b = Ranges B & C. * 0111b = Ranges A, B & C. * 1110b = Ranges B, C & D. * 1111b = Ranges A, B, C & D. * All other values are reserved. It is strongly recommended that the completion timeout mechanism not expire in less than 10 ms 0010b RO Bit Acronym 31 :5 Reserved 3 :0 B:D:1+ Index1 Default: 00000012h Bit Range 4 Bus:Device:Function: October 2012 Order Number: 327879-001US Bit Description Sticky Intel(R) Communications Chipset 89xx Series - Datasheet 1531 29.4.6.11 PDCNTR2[0:3]--PF Device Control 2 Register Table 29-63. PDCNTR2[0:3]--PF Device Control 2 Register Description: Link Control 2 Register View: PCI Size: 16 bit BAR: Configuration Power Well: Core Bit Access Reserved. 0b RV CTODIS Completion Timeout Disable: When set to 1b, this bit disables the completion timeout mechanism. Software is permitted to set or clear this bit at any time. When set, the completion timeout detection mechanism is disabled. If there are outstanding requests when the bit is cleared, it is permitted but not required for hardware to apply the completion timeout mechanism to the outstanding requests. If this is done, it is permitted to base the start time for each request on either the time this bit was cleared or the time each request was issued. The default value for this bit is 0b. 0b RW CTOV Completion Timeout Value: In devices that support completion timeout programmability, this field enables system software to modify the completion timeout value. Encoding: * 0000b = Default range: 50 us to 50 ms. It is strongly recommended that the completion timeout mechanism not expire in less than 10 ms. Values available if Range A (50 us to 10 ms) programmability range is supported: * 0001b = 50us to 100 us. * 0010b = 1 ms to 10 ms. Values available if Range B (10 ms to 250 ms) programmability range is supported: * 0101b = 16 ms to 55 ms. * 0110b = 65 ms to 210 ms. Values available if Range C (250 ms to 4 s) programmability range is supported: * 1001b = 260 ms to 900 ms. * 1010b = 1 s to 3.5 s. Values available if the Range D (4 s to 64 s) programmability range is supported: * 1101b = 4 s to 13 s. * 1110b = 17 s to 64 s. Values not defined are reserved. Software is permitted to change the value in this field at any time. For requests already pending when the completion timeout value is changed, hardware is permitted to use either the new or the old value for the outstanding requests and is permitted to base the start time for each request either when this value was changed or when each request was issued. The default value for this field is 0000b. 0000b RW 15 :5 Reserved Bit Description Intel(R) Communications Chipset 89xx Series - Datasheet 1532 Offset Start: C8h Offset End: C9h Bit Reset Value Bit Acronym 3 :0 B:D:1+ Index1 Default: 0000h Bit Range 4 Bus:Device:Function: Sticky October 2012 Order Number: 327879-001US 29.0 29.4.6.12 PLCNTLR2[0:3]--PF Link Control 2 Register Table 29-64. PLCNTLR2[0:3]--PF Link Control 2 Register (Sheet 1 of 2) Description: Link Control 2 Register View: PCI Size: 16 bit BAR: Configuration Bit Acronym 15 :13 Reserved CDE CSOS 10 EMC Compliance De-emphasis - This bit sets the de-emphasis level in Polling.Compliance state if the entry occurred due to the Enter Compliance bit being 1b. Encodings: 1b -3.5 dB 0b -6 dB TMARG Bit Access 0h RV 0b RV 0b RV 0b RV 0b RV 0b RO This Bit field is reserved for non-0 function. Enter Modified Compliance - When this bit is set to 1b, the device transmits modified compliance pattern if the LTSSM enters Polling.Compliance state. This bit field is reserved for non-0 function. Transmit Margin - This field controls the value of the non deemphasized voltage level at the Transmitter pins. Encodings: 000b - Normal operating range 001b - 800-1200 mV for full swing and 400-700 mV for half-swing. 010b - (n-1) - Values must be monotonic with a non-zero slope. The value of n must be greater than 3 and less than 7. At least two of these must be below the normal operating range of n: 200-400 mV for full-swing and 100200 mV for half-swing. other values - reserved Note: Bit Reset Value This bit field is reserved for non-0 function. Compliance SOS - When set to 1b, the LTSSM is required to send SOS periodically in between the (modified) compliance patterns. Note: 9 :7 Sticky Reserved Note: Offset Start: D0h Offset End: D1h Power Well: Core Bit Description Note: 11 B:D:1+ Index1 Default: 0000h Bit Range 12 Bus:Device:Function: This bit field is reserved for non-0 function. Selectable De-emphasis. 6 SDEM October 2012 Order Number: 327879-001US Note: This bit is not applicable and reserved for Endpoints. Intel(R) Communications Chipset 89xx Series - Datasheet 1533 Table 29-64. PLCNTLR2[0:3]--PF Link Control 2 Register (Sheet 2 of 2) Description: Link Control 2 Register View: PCI Size: 16 bit Bit Range 5 BAR: Configuration Bit Acronym HWAUTOSD ENCOMP TLNKS Hardware Autonomous Speed Disable. When set to 1b, this bit disables hardware from changing the link speed for reasons other than attempting to correct unreliable link operation by reducing link speed. Hard wire to 0b. Bit Reset Value Bit Access 0b RV 0b RV 0 RV This bit field is reserved for non-0 function. Target Link Speed. This field is used to set the target compliance mode speed when software is using the Enter Compliance bit to force a link into compliance mode. Defined encodings are: 0001b = 2.5 Gb/s Target Link Speed. 0010b = 5 Gb/s Target Link Speed. All other encodings are reserved. If a value is written to this field that does not correspond to a speed included in the Supported Link Speeds field, the result is undefined. The default value of this field is the highest link speed supported (as reported in the Supported Link Speeds field of the Link Capabilities register). This bit field is reserved for non-0 function. Intel(R) Communications Chipset 89xx Series - Datasheet 1534 Sticky This bit field is reserved for non-0 function. Enter Compliance. Software is permitted to force a link to enter compliance mode at the speed indicated in the Target Link Speed field by setting this bit to 1b in both components on a link and then initiating a hot reset on the link. The default value of this field following a fundamental reset is 0b. Note: Offset Start: D0h Offset End: D1h Power Well: Core Bit Description Note: 3 :0 B:D:1+ Index1 Default: 0000h Note: 4 Bus:Device:Function: October 2012 Order Number: 327879-001US 29.0 29.4.6.13 PLSR2[0:3]--PF Link Status 2 Register Table 29-65. PLSR2[0:3]--PF Link Status 2 Register Description: Link Status 2 Register View: PCI Size: 16 bit BAR: Configuration Bit Acronym 15 :1 Reserved CDEL Offset Start: D2h Offset End: D3h Power Well: Core Bit Description Sticky Reserved Current De-emphasis Level - When the Link is operating at 5 GT/s speed, this bit reflects the level of de-emphasis. it is undefined when the Link is operating at 2.5 GT/s speed Encodings: 1b -3.5 dB 0b -6 dB Note: 29.4.7 B:D:1+ Index1 Default: 0000h Bit Range 0 Bus:Device:Function: Bit Reset Value Bit Access 0h RV 0b RO This bit field is reserved and not defined for this function. Vital Product Data (VPD) Registers The Controller supports access to a VPD structure stored in the EEPROM using the following set of registers. Note: The VPD structure is available through all port functions. As the interface is common to all functions, accessing the VPD structure of one function while an access to the EEPROM is in process on another function can yield unexpected results. Byte Offset 0xE0 0xE4 October 2012 Order Number: 327879-001US Byte 3 Byte 2 VPD address Byte 1 Byte 0 Next Pointer (0x00) Capability ID (0x03) VPD data Intel(R) Communications Chipset 89xx Series - Datasheet 1535 29.4.7.1 VPDCID[0:3]--VPD Capability ID Register This field equals 0x3 indicating the linked list item as being the VPD registers. Table 29-66. VPDCID[0:3]--VPD Capability ID Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 29.4.7.2 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 3h Bit Acronym PCIECID Offset Start: E0h Offset End: E0h Power Well: Core Bit Description Sticky Capability ID: VPD Bit Reset Value Bit Access 3h RO VPDNCP[0:3]--VPD Next Capability Pointer Register Table 29-67. VPDNCP[0:3]--VPD Next Capability Pointer Register Description: View: PCI Size: 8 bit Bit Range 07 : 00 B:D:1+ Bus:Device:Function: Index1 BAR: Configuration Default: 0h Bit Acronym PCIENP Power Well: Core Bit Description Next Capability Pointer: Last Capability. Intel(R) Communications Chipset 89xx Series - Datasheet 1536 Offset Start: E1h Offset End: E1h Sticky Bit Reset Value Bit Access 0h RO October 2012 Order Number: 327879-001US 29.0 29.4.7.3 VPDADDR[0:3]--VPD Address Register Dword-aligned byte address of the VPD area in the EEPROM to be accessed. The register is read/write with the initial value at power-up indeterminate. Table 29-68. VPDADDR[0:3]--VPD Address Register Description: View: PCI Size: 16 bit Bit Range 15 14 : 00 29.4.7.4 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0h Offset Start: E2h Offset End: E3h Power Well: Core Bit Reset Value Bit Access FLAG Flag: A flag used to indicate when the transfer of data between the VPD Data register and the storage component completes. The Flag register is written when the VPD Address register is written. 0b = Read. Set by hardware when data is valid. 1b = Write. Cleared by hardware when data is written to the EEPROM. The VPD address and data should not be modified before the action completes. 0h RW VADDR VPD Address. Dword-aligned byte address of the VPD area in the EEPROM to be accessed. The register is read/write with the initial value at power-up indeterminate. The two LSBs are RO as zero. This is the address relative to the start of the VPD area. As the maximal size supported is 256 bytes, bits 14:8 should always be zero. 0h RW Bit Acronym Bit Description Sticky VPDDATA[0:3]--VPD Data This register contains the VPD read/write data. Table 29-69. VPDDATA[0:3]--VPD Data Description: View: PCI Size: 32 bit 31 :0 29.5 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0h VPDDATA Offset Start: E4h Offset End: E7h Power Well: Core VPD Data VPD data can be read or written through this register. The LSB of this register (at offset four in this capability structure) corresponds to the byte of VPD at the address specified by the VPD Address register. The data read from or written to this register uses the normal PCI byte transfer capabilities. Four bytes are always transferred between this register and the VPD storage component. Reading or writing data outside of the VPD space in the storage component is not allowed. In a write access, the data should be set before the address and the flag is set. 0h RW PCIe Extended Configuration Space PCIe extended configuration space is located in a flat memory-mapped address space. PCIe extends the configuration space beyond the 256 bytes available for PCI to 4096 bytes. The Controller decodes an additional 4-bits (bits 27:24) to provide the additional October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1537 configuration space as shown in Table 29-70. PCIe reserves the remaining four bits (bits 31:28) for future expansion of the configuration space beyond 4096 bytes. The configuration address for a PCIe device is computed using a PCI-compatible bus, device, and function numbers as follows. Table 29-70. PCIe Extended Configuration Space 31 28 0000b 27 20 Bus # 19 15 14 Device # 12 11 Fun # 2 Register Address (offset) 1 0 00b PCIe extended configuration space is allocated using a linked list of optional or required PCIe extended capabilities following a format resembling PCI capability structures. The first PCIe extended capability is located at offset 0x100 in the device configuration space. The first Dword of the capability structure identifies the capability/version and points to the next capability. The Controller supports the following PCIe extended capabilities. Table 29-71. PCIe Extended Capability Structure Capability Offset Next Header Advanced Error Reporting Capability 0x100 0x140/0x1A01 Serial Number2 0x140 0x1A0 1. Depends on EEPROM settings enabling the Serial Number in PCIe Init Configuration 2 EEPROM Word. 2. Not available in EEPROM-less systems. 29.5.1 Advanced Error Reporting (AER) Capability The PCIe AER capability is an optional extended capability to support advanced error reporting. The following table lists the PCIe AER extended capability structure for PCIe devices. Byte Offset Byte 3 Byte 2 0x100 Next Capability Ptr. (0x140/ 0x1A01) Version (0x1) 0x104 Byte 1 Byte 0 AER Capability ID (0x0001) Uncorrectable Error Status 0x108 Uncorrectable Error Mask 0x10C Uncorrectable Error Severity 0x110 Correctable Error Status 0x114 Correctable Error Mask 0x118 Advanced Error Capabilities and Control Register 0x11C... 0x128 Header Log 1. Depends on EEPROM settings enabling the Serial Number structure in PCIe Init Configuration 2 EEPROM Word. Intel(R) Communications Chipset 89xx Series - Datasheet 1538 October 2012 Order Number: 327879-001US 29.0 29.5.2 PF Advanced Error Reporting Capability Structure 29.5.2.1 PPCIEAERCAPID[0:3]--PF PCI Express AER Capability ID Register The PCI Express Capability List register enumerates the PCI Express AER Capability structure in the PCI 3.0 configuration space capability list. Note: Depending on the ACC_VF_ENABLE strap the AER Capability can be the last capability in the list. Table 29-72. PPCIEAERCAPID[0:3]--PF PCI Express AER Capability ID Register Description: View: PCI Size: 32 bit Bit Range BAR: Configuration B:D:1+ Bus:Device:Function: Index1 Default: 13810001h Power Well: Core Bit Reset Value Bit Access 138h RO Advanced Error Capability Version Number: PCI Express Advanced Error Reporting Extended Capability Version Number. 1h RO Advanced Error Capability ID: PCI Express Extended Capability ID indicating Advanced Error Reporting Capability. 1h RO Bit Acronym Bit Description 31 : 20 PCIEAERNCP Next PCI Express Extended Capability Pointer: The Next Capability Pointer default value will be dependent on the ACC_VF_ENABLE strap. 1 - 138h (Default) 0 - 000h (AER is the last extended capability) 19 :16 PCIEAERCVN 15 : 00 PCIEAERCID 29.5.2.2 Offset Start: 100h Offset End: 103h Sticky PPAERUCS[0:3]--PF PCI Express AER Uncorrectable Error Status Register Table 29-73. PPAERUCS[0:3]--PF PCI Express AER Uncorrectable Error Status Register Description: View: PCI Size: 32 bit BAR: Configuration B:D:1+ Bus:Device:Function: Index1 Default: 0h Bit Range Bit Acronym 31 :21 Reserved 20 UR 19 ECRCC 18 MTLP 17 RO October 2012 Order Number: 327879-001US Offset Start: 104h Offset End: 107h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00b RV Unsupported Request Error Status: As a receiver, Set whenever an unsupported request is detected. The Header is logged. Y 0b RW1CS ECRC Check: As a receiver, set when ECRC check fails. The Header is logged. Y 0b RW1CS Malformed TLP: As a receiver, set whenever a malformed TLP is detected. The Header is logged. Y 0b RW1CS Receiver Overflow: Set if PCI Express receive buffers overflow. Y 0b RW1CS Intel(R) Communications Chipset 89xx Series - Datasheet 1539 Table 29-73. PPAERUCS[0:3]--PF PCI Express AER Uncorrectable Error Status Register Description: View: PCI Size: 32 bit Bit Range BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0h Bit Acronym Offset Start: 104h Offset End: 107h Power Well: Core Bit Description Sticky Bit Reset Value Bit Access 16 EC Unexpected Completion: As a receiver, set whenever a completion is received that does not match the GbE requestor ID or outstanding Tag. The Header is logged. Y 0b RW1CS 15 CA Completer Abort: As a completer, set whenever an internal agent signals a data abort. The header is logged. Y 0b RW1CS 14 CT Completion Timeout: As a requester, set whenever an outbound Non Posted Request does not receive a completion within 16-32ms. Y 0b RW1CS 13 FCPES Flow Control Protocol Error Status: Set whenever a flow control protocol error is detected. Not supported. Y 0b RW1CS PTLPR Poisoned TLP Received: As a receiver, set whenever a poisoned TLP is received from PCI Express. The header is logged. Internal queue errors are not covered by this bit, they are logged by the Configuration target of the transaction. Y 0h RW1CS 0000b RV 12 11 :6 Reserved Reserved 5 SDES Surprise Down Error: Not supported. Y 0b RW1CS 4 DLPE Data Link Protocol Error: Set whenever a data link protocol error is detected. Y 0b RW1CS 3 :00 Reserved 0h RV 29.5.2.3 Reserved PPAERUCM[0:3]--PF PCI Express AER Uncorrectable Error Mask Register Table 29-74. PPAERUCM[0:3]--PF PCI Express AER Uncorrectable Error Mask Register Description: View: PCI Size: 32 bit Bus:Device:Function: B:D:1+ Index1 BAR: Configuration Default: 0h Bit Range Bit Acronym 31 :21 Reserved Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00b RV 20 UR Unsupported Request Error Mask: When `1' error reporting is masked. 19 ECRCC ECRC Check Error Mask: When `1' error reporting is masked. Y 0b RWS 18 MTLP Malformed TLP Error Mask: When `1' error reporting is masked. Y 0b RWS Intel(R) Communications Chipset 89xx Series - Datasheet 1540 Offset Start: 108h Offset End: 10Bh Y 0b RWS October 2012 Order Number: 327879-001US 29.0 Table 29-74. PPAERUCM[0:3]--PF PCI Express AER Uncorrectable Error Mask Register Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0h Offset Start: 108h Offset End: 10Bh Power Well: Core Bit Acronym Bit Description Sticky Bit Reset Value Bit Access 17 RO Receiver Overflow Error Mask: When `1' error reporting is masked. Y 0b RWS 16 EC Unexpected Completion Error Mask: When `1' error reporting is masked. Y 0b RWS 15 CA Completer Abort Error Mask: When `1' error reporting is masked. Y 0b RWS 14 CT Completion Time Out Error Mask: When `1' error reporting is masked. Y 0b RWS 13 FCPES Flow Control Protocol Error Mask: When `1' error reporting is masked. Not supported. Y 0b RWS 12 PTLPR Poisoned TLP Received Error Mask: When `1' error reporting is masked. Y 0h RWS 0000b RV Bit Range 11 :6 Reserved Reserved 5 SDES Surprise Down Error: Not supported. 4 DLPE Data Link Protocol Error Mask: When `1' error reporting is masked. 03 :01 Reserved Reserved 0 Reserved Undefined: Reserved 29.5.2.4 Y 0b RWS Y 0b RWS 0000b RV 0b RV PPAERUCSEV[0:3]--PF PCI Express AER Uncorrectable Error Severity Register The Uncorrectable Error Severity register controls whether an individual uncorrectable error is reported as a non-fatal or fatal error. An error is reported as fatal when the corresponding error bit in the severity register is set. If the bit is cleared, the corresponding error is considered non-fatal. Table 29-75. PPAERUCSEV[0:3]--PF PCI Express AER Uncorrectable Error Severity Register Description: View: PCI Size: 32 bit Bus:Device:Function: B:D:1+ Index1 BAR: Configuration Default: 00062030h Bit Range Bit Acronym 31 :21 Reserved 20 UR 19 ECRCC 18 MTLP October 2012 Order Number: 327879-001US Offset Start: 10Ch Offset End: 10Fh Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00b RV Unsupported Request Error Status Severity(URESS): Y 0b RWS ECRC Check Severity: Y 0b RWS Malformed TLP Severity: Y 1b RWS Intel(R) Communications Chipset 89xx Series - Datasheet 1541 Table 29-75. PPAERUCSEV[0:3]--PF PCI Express AER Uncorrectable Error Severity Register Description: View: PCI Size: 32 bit Bit Range 17 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 00062030h Bit Acronym Offset Start: 10Ch Offset End: 10Fh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access RO Receiver Overflow Severity: Y 1b RWS 16 EC Unexpected Completion Severity: Y 0b RWS 15 CA Completer Abort Severity: Y 0b RWS 14 CT Completion Time Out Severity: Y 0b RWS Y 1b RWS Y 0h RWS 0000b RV 13 FCPES Flow Control Protocol Error Severity: Not supported. 12 PTLPR Poisoned TLP Received Severity: 11 :6 Reserved Reserved 5 SDES Surprise Down Error Severity: Not supported. 04 DLPE Data Link Protocol Error Severity: 3 :01 Reserved Reserved 0 Reserved Undefined: Reserved 29.5.2.5 Y 1b RWS Y 1b RWS 0000b RV 0b RV PPAERCS[0:3]--PF PCI Express AER Correctable Error Register Table 29-76. PPAERCS[0:3]--PF PCI Express AER Correctable Error Register Description: View: PCI Size: 32 bit Default: 00h Bit Range Bit Acronym 31 :14 Reserved 13 12 11 :09 ANFES RTTS Reserved 08 RNRS 07 BDLLPS 06 05 :01 00 B:D:1+ Bus:Device:Function: Index1 BAR: Configuration DLPE Reserved RES Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00b RV Advisory Non-Fatal Error Status Y 0b RW1CS Replay Timer Timeout Status: Set whenever a replay timer timeout occurs. Y 0b RW1CS 0b RV Reserved REPLAY NUM Rollover Status: Set whenever the replay number rolls over from 11 to 00. Y 0h RW1CS Bad DLLP Status: Sets this bit on CRC errors on DLLP. Y 0000b RW1CS Bad TLP Status: Sets this bit on CRC errors or sequence number out of range on TLP. Y 0b RW1CS 00000b RV 0b RW1CS Reserved Receiver Error Status: Set whenever the physical layer detects a receiver error. Intel(R) Communications Chipset 89xx Series - Datasheet 1542 Offset Start: 110h Offset End: 113h Y October 2012 Order Number: 327879-001US 29.0 29.5.2.6 PPAERCM[0:3]--PF PCI Express AER Correctable Error Mask Register Table 29-77. PPAERCM[0:3]--PF PCI Express AER Correctable Error Mask Register Description: View: PCI Size: 32 bit BAR: Configuration Bit Acronym 31 :14 Reserved 13 ANFES 12 RTTS Reserved 08 RNRS 07 BDLLPS 06 DLPE 05 :01 00 29.5.2.7 B:D:1+ Index1 Default: 2000h Bit Range 11 :09 Bus:Device:Function: Reserved RES Offset Start: 114h Offset End: 117h Power Well: Core Bit Description Sticky Reserved Bit Reset Value Bit Access 00b RV Advisory Non-Fatal Error Mask: this bit is set by default to enable compatibility with software that does not comprehend Role-Based Error Reporting. Y 1b RWS Replay Timer Timeout Mask: Y 0b RWS Reserved 000b RV REPLAY NUM Rollover Mask: Y 0b RWS Bad DLLP Mask: Y 0b RWS Bad TLP Mask: Y 0b RWS Reserved Receiver Error Mask: Y 00h RV 0b RWS PPAERCTLCAP[0:3]--PF PCI Express AER Control and Capability Register Table 29-78. PPAERCTLCAP[0:3]--PF PCI Express AER Control and Capability Register Description: View: PCI Size: 32 bit BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0h Power Well: Core Bit Reset Value Bit Access 0 RV 0b RWS 0b RO 0b RWS ECRC Generation Capable: Indicates the EP is capable of generating ECRC. 0b RO The First Error Pointer: Identifies the bit position of the first error reported in the Section 29-73, "PPAERUCS[0:3]--PF PCI Express AER Uncorrectable Error Status Register" Note: This register will not update until all bits in the ERRUNC STS register are cleared. 0 ROS Bit Range Bit Acronym 31 :09 Reserved Reserved ECRCCE ECRC Check Enable: When set enables ECRC checking. 07 ECRCCC ECRC Check Capable: Indicates EP is capable of checking ECRC. 06 ECRCGE ECRC Generation Enable: When set enables ECRC generation. 05 ECRCGC TFEP 08 04 :00 October 2012 Order Number: 327879-001US Offset Start: 118h Offset End: 11Bh Bit Description Sticky Y Y Intel(R) Communications Chipset 89xx Series - Datasheet 1543 29.5.2.8 PPAERHDRLOG0[0:3]--PF PCI Express AER Header Log 0 Register Table 29-79. PPAERHDRLOG0[0:3]--PF PCI Express AER Header Log 0 Register Description: View: PCI Size: 32 bit Bit Range 31 :00 29.5.2.9 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0h Bit Acronym Offset Start: 11Ch Offset End: 11Fh Power Well: Core Bit Description Sticky 1st DWord of the Header for the PCI Express packet in error (HDRLOGDW0): Once an error is logged in this register, it remains locked HDRLOGDW0 for further error logging until the time the software clears the status bit that cause the header log i.e. the error pointer is rearmed to log again. Bit Reset Value Bit Access 0h RO PPAERHDRLOG1[0:3]--PF PCI Express AER Header Log 1 Register Table 29-80. PPAERHDRLOG1[0:3]--PF PCI Express AER Header Log 1 Register Description: View: PCI Size: 32 bit Bit Range 31 :00 Bus:Device:Function: B:D:1+ Index1 BAR: Configuration Default: 0h Bit Acronym Power Well: Core Bit Description 2nd DWord of the Header for the PCI Express packet in error (HDRLOGDW1): Once an error is logged in this register, it remains locked HDRLOGDW1 for further error logging until the time the software clears the status bit that cause the header log i.e. the error pointer is rearmed to log again. Intel(R) Communications Chipset 89xx Series - Datasheet 1544 Offset Start: 120h Offset End: 123h Sticky Bit Reset Value Bit Access 0h RO October 2012 Order Number: 327879-001US 29.0 29.5.2.10 PPAERHDRLOG2[0:3]--PF PCI Express AER Header Log 2 Register Table 29-81. PPAERHDRLOG2[0:3]--PF PCI Express AER Header Log 2 Register Description: View: PCI Size: 32 bit Bit Range 31 :00 29.5.2.11 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0h Bit Acronym Offset Start: 124h Offset End: 127h Power Well: Core Bit Description Sticky 3rd DWord of the Header for the PCI Express packet in error (HDRLOGDW2): Once an error is logged in this register, it remains locked HDRLOGDW2 for further error logging until the time the software clears the status bit that cause the header log i.e. the error pointer is rearmed to log again. Bit Reset Value Bit Access 0h RO PPAERHDRLOG3[0:3]--PF PCI Express AER Header Log 3 Register Table 29-82. PPAERHDRLOG3[0:3]--PF PCI Express AER Header Log 3 Register Description: View: PCI Size: 32 bit Bit Range 31 :00 BAR: Configuration Bus:Device:Function: B:D:1+ Index1 Default: 0h Bit Acronym Power Well: Core Bit Description Sticky 4th DWord of the Header for the PCI Express packet in error (HDRDWLOG3): Once an error is logged in this register, it remains locked HDRDWLOG3 for further error logging until the time the software clears the status bit that cause the header log i.e. the error pointer is rearmed to log again. October 2012 Order Number: 327879-001US Offset Start: 128h Offset End: 12Bh Bit Reset Value Bit Access 0h RO Intel(R) Communications Chipset 89xx Series - Datasheet 1545 29.5.3 PF Alternative Routing-ID Extended Capability Structure This section describes the PCI Express Extended Configuration Space registers that make up the Alternative Routing ID Extended Capability Structure. Some information from the specification is repeated here as an aid to the reader or to describe implementation choice. See the PCI Express* Base Specification 2.0 for the full register descriptions and additional information regarding their operation. 29.5.3.1 PARIDHDR[0:3]--PF Alternative Routing ID Capability Header This register contains information associated with the Alternative Routing ID capability that is implemented in the GbE. This is compliant with the PCI-SIG ECN: Alternatifve Routing-ID Interpretation (ARI), Updated June4,2007. Table 29-83. PARIDHDR[0:3]--PF Alternative Routing ID Capability Header Description: View: PCI Size: 32 bit B:D:1+ Bus:Device:Function: Index1 BAR: Configuration Default: 0001000Eh Power Well: Core Bit Range Bit Acronym Bit Description 31 :20 ARINCO Next Capability Offset - This is the last extended capability. 19 :16 ARICV Capability Version - This is set to 1h for the most current version of the specification. 15 :00 ARICV PCI Express Extended Capability ID - The PCI SIG has assigned 000Eh to the ARI extended capability. Intel(R) Communications Chipset 89xx Series - Datasheet 1546 Offset Start: 138h Offset End: 13Bh Sticky Bit Reset Value Bit Access 000h RO 1h RO 000EH RO October 2012 Order Number: 327879-001US 29.0 29.5.3.2 PFARICAP0--PF ARI Capabilities Register This register contains information associated with the Alternative Routing ID capability that is implemented in the GbE. Table 29-84. PFARICAP0--PF ARI Capabilities Register Description: View: PCI Size: 16 bit Bit Range 15 :8 7 :2 BAR: Configuration Bus:Device:Function: B:D:1 Default: 0200h Bit Acronym VNFN Reserved 1 ACS 0 MFVC October 2012 Order Number: 327879-001US Offset Start: 13Ch Offset End: 13Dh Power Well: Core Bit Description Sticky Next Function Number: The function number of the next highest numbered PF in a multi-function device. The Next Function Number is set based on fuse and EEPROM bits. However, some of the values listed below will never be observed as some fuse and EEPROM bits' combinations are not valid scenarios. 0h - If Function 1 is the highest function. 2h - If Function 2 is enabled. 3h - If Function 2 is disabled, but Function 3 is enabled. 4h - If Function 2-3 are disabled, but Function 4 is enabled. Bit Reset Value Bit Access 2h RO Reserved. 0h RV ACS Functional Groups Capability: GbE does not support. 0b RO MFVC Functional Groups Capability: GbE does not support. 0b RO Intel(R) Communications Chipset 89xx Series - Datasheet 1547 29.5.3.3 PFARICAP1--PF ARI Capabilities Register This register contains information associated with the Alternative Routing ID capability that is implemented in the GbE. Table 29-85. PFARICAP1--PF ARI Capabilities Register Description: View: PCI Size: 16 bit Bit Range 15 :8 BAR: Configuration Default: 0300h Bit Reset Value Bit Access 3h RO Reserved. 0h RV ACS Functional Groups Capability: GbE does not support. 0b RO MFVC Functional Groups Capability: GbE does not support. 0b RO Bit Acronym VNFN 7 :2 Reserved 1 ACS 0 MFVC Power Well: Core Bit Description Next Function Number: The function number of the next highest numbered PF in a multi-function device. The Next Function Number is set based on fuse and EEPROM bits. However, some of the values listed below will never be observed as some fuse and EEPROM bits' combinations are not valid scenarios. 0h - If Function 2 is the highest function. 3h - If Function 3 is enabled. 4h - If Function 3 is disabled, but Function 4 is enabled. Intel(R) Communications Chipset 89xx Series - Datasheet 1548 Offset Start: 13Ch Offset End: 13Dh Bus:Device:Function: B:D:2 Sticky October 2012 Order Number: 327879-001US 29.0 29.5.3.4 PFARICAP2--PF ARI Capabilities Register This register contains information associated with the Alternative Routing ID capability that is implemented in the GbE. Table 29-86. PFARICAP2--PF ARI Capabilities Register Description: View: PCI Size: 16 bit Bit Range 15 :8 7 :2 BAR: Configuration Default: 0400h Bit Acronym VNFN Reserved 1 ACS 0 MFVC 29.5.3.5 Bus:Device:Function: B:D:3 Offset Start: 13Ch Offset End: 13Dh Power Well: Core Bit Description Sticky Next Function Number: The function number of the next highest numbered PF in a multi-function device. The Next Function Number is set based on fuse and EEPROM bits. However, some of the values listed below will never be observed as some fuse and EEPROM bits' combinations are not valid scenarios. 0h - If Function 3 is the highest function. 4h - If Function 4 is enabled. Bit Reset Value Bit Access 4h RO Reserved. 0h RV ACS Functional Groups Capability: GbE does not support. 0b RO MFVC Functional Groups Capability: GbE does not support. 0b RO PFARICAP3--PF ARI Capabilities Register This register contains information associated with the Alternative Routing ID capability that is implemented in the GbE. Table 29-87. PFARICAP3--PF ARI Capabilities Register (Sheet 1 of 2) Description: View: PCI Size: 16 bit Bit Range 15 :8 BAR: Configuration Bus:Device:Function: B:D:4 Default: 0500h Bit Acronym VNFN October 2012 Order Number: 327879-001US Offset Start: 13Ch Offset End: 13Dh Power Well: Core Bit Description Sticky Next Function Number: The function number of the next highest numbered PF in a multi-function device. The Next Function Number is set based on fuse and EEPROM bits. However, some of the values listed below will never be observed as some fuse and EEPROM bits' combinations are not valid scenarios. 0h - If Function 4 is the highest function. Bit Reset Value Bit Access 5h RO Intel(R) Communications Chipset 89xx Series - Datasheet 1549 Table 29-87. PFARICAP3--PF ARI Capabilities Register (Sheet 2 of 2) Description: View: PCI Size: 16 bit BAR: Configuration Default: 0500h Bit Range Bit Acronym 7 :2 Reserved 1 ACS 0 MFVC 29.5.3.6 Offset Start: 13Ch Offset End: 13Dh Bus:Device:Function: B:D:4 Power Well: Core Bit Reset Value Bit Access Reserved. 0h RV ACS Functional Groups Capability: GbE does not support. 0b RO MFVC Functional Groups Capability: GbE does not support. 0b RO Bit Description Sticky PARIDCTL[0:3]--PF Alternative Routing ID Control Register This register contains information associated with the Alternative Routing ID capability that is implemented in the GbE. Table 29-88. PARIDCTL[0:3]--PF Alternative Routing ID Control Register Description: View: PCI Size: 16 bit Default: 00000000h Bit Range Bit Acronym 15 :7 Reserved 6 :4 FG 3 :2 Reserved 1 ACS 0 MFVC 29.5.4 B:D:1+ Bus:Device:Function: Index1 BAR: Configuration Offset Start: 13Eh Offset End: 13Fh Power Well: Core Bit Description Sticky Bit Reset Value Bit Access Reserved 0h RV Function Group: Hardwired to Zero as GbE does not support Function Groups. 0b RO Reserved 0b RV ACS Functional Groups Enable: Hardwired to Zero as GbE does not support. 0b RO MFVC Functional Groups Enable: Hardwired to Zero as GbE does not support. 0b RO TLP Processing Hint Requester (TPH) Capability The PCIe TPH Requester capability is an optional extended capability to support TLP Processing Hints. This is not supported by the GbE Controller MAC. 29.5.4.1 TPH CAP ID (0x1A0; RO) Not supported. 29.5.4.2 TPH Requester Capabilities (0x1A4; RO) Not Supported Intel(R) Communications Chipset 89xx Series - Datasheet 1550 October 2012 Order Number: 327879-001US 29.0 29.5.4.3 TPH Requester Control (0x1A8; R/W) Not Supported. 29.5.4.4 TPH Steering table (0x1AC - 0x1B8; R/W) Not Supported. 29.5.4.5 Latency Tolerance Requirement Reporting (LTR) Capability Not Supported. 29.5.4.6 LTR CAP ID (0x1C0; RO) Not Supported 29.5.4.7 LTR Capabilities (0x1C4; RW) Not Supported 29.5.5 I/O Space The I/O Space is 32 bytes and consists of only two 32-bit registers (IOADDR and IODATA). Table 29-89. IOBAR Register Map Name IOBAR Offset IOADDR 00h IODATA 04h Reserved 08h-1Ch October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1551 Test Features - Volume 3 of 4 October 2012 Order Number: 327879-001US 30.0 30.0 PCH Global Test Features 30.1 JTAG JTAG refers to the JTAG/IEEE 1149.1 Test Access Port, and is one of the primary debug and test mechanisms used on the PCH. The following subsections describe the purpose and use of the JTAG port on the PCH. 30.1.1 JTAG Functions Overview The PCH supports the IEEE 1149.1-2001 (Standard Test Access Port and BoundaryScan Architecture) specification. The TAP provides instructions and logic to support IEEE 1149.6-2003 (Standard for Boundary-Scan Testing of Advanced Digital Networks). The Hi-speed IOs (PCIe, DMI,SATA,SGMII) all support 1149.6 with the rest supporting 1149.1. 30.1.2 PCH TAP Interfaces The PCH has two four-pin JTAG interfaces, one for the Legacy I/O (DMI) pins and the other for the pins associated with the PCI Express Endpoint. Each JTAG interface can be used independently or cascaded together as indicated in the following figure. Table 30-1. PCH Die-level Legacy I/O (DMI)TAP Pin Interface Pin Name Type Function Operation and Properties JTCK Input JTAG Test Clock * Clock input for the TAP controller, instruction register and test data register. JTMS Input Test Mode Select * * * Changes the state of the TAP controller. Pulled high when not being driven. Is sampled on the rising edge of TCK. * * * Serial data input to the instruction register and test data registers. Pulled high when not being driven. Is sampled on the rising edge of TCK. * * * Serial data output. Inactive (high-Z) during non-shift operations. Data is clocked out on the falling edge of TCK. JTDI Input JTDO Output October 2012 Order Number: 327879-001US Test Data In Test Data Out Intel(R) Communications Chipset 89xx Series - Datasheet 1553 Table 30-2. PCH Die-Level Endpoint TAP Pin Interface Pin Name Type Function EP_JTCK Input JTAG Test Clock EP_JTMS Input Test Mode Select EP_JTDI Input Test Data In EP_JTRSTB Input Jtag Test Reset EP_JTDO Output Operation and Properties * Clock input for the TAP controller, instruction register and test data register. * * * Changes the state of the TAP controller. Pulled high using Internal Pull-Up when not being driven. Is sampled on the rising edge of TCK. * Serial data input to the instruction register and test data registers. Pulled high using Internal Pull-Up when not being driven. Is sampled on the rising edge of TCK. * * Test Data Out * * * Jtag Reset (TRSTB) Pulled high using Internal Pull-Up when not being driven * * Serial data output. Inactive (high-Z) during non-shift operations. OpenDrain. Data is clocked out on the falling edge of TCK. * Figure 30-1. TAP Connectivity PCH Package PCH Die Endpoint TAP EP_JTCK EP_JTMS EP_JTDI EP_JTDI TCK TMS EP_JTDO EP_JTDO TDI TDO TCK TMS Legacy I/O (DMI) TAP JTCK JTMS JTDI 30.1.3 JTDI TCK TMS JTDO TDI JTDO TDO TAP Controller Operation and State Diagram The TAP controller contains a state machine that is the heart of the JTAG TAP (Figure 30-2). The TAP controller is asynchronously reset via the hardware TAP reset. Once reset is de-asserted, the TAP controller samples the TMS pin at the rising edge of TCLK pin, and it sequences through the states under the control of the TMS pin. Holding TMS high for five or more TCLK cycles will take the TAP to the Test-Logic-Reset state regardless of what state it is in. It is recommended that TMS be held high at the de-assertion of reset to ensure deterministic operation of the TAP. The TDO pin is output-enabled only during Shift-DR or Shift-IR states. Intel(R) Communications Chipset 89xx Series - Datasheet 1554 October 2012 Order Number: 327879-001US 30.0 Figure 30-2. TAP Controller State Diagram October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1555 Technical Specifications - Volume 4 of 4 October 2012 Order Number: 327879-001US 31.0 31.0 System Clocks 31.1 External Clock Requirements Table 31-1 shows all the input clocks to the PCH. There are three modes of operation: * Normal - all clocks are required (all clocks listed in Table 31-1). * EndPoint mode - only EndPoint clocks are required. * Non-EndPoint Mode - only Non-EndPoint clocks are required. Table 31-1. Platform External Clock Interface Note: Signal Name Clock Domain Frequency Destination CRU_CLK100P/N Clock Resource Unit 100 MHz EndPoint GBE_CLK100P/N GbE Ethernet 100 MHz EndPoint PCIE_EP_CLK100P/N PCI Express* EndPoint 100 MHz EndPoint DMI_CLK100P/N Direct Media Interface 100 MHz Non-EndPoint PCICLK Low Pin Count 33 MHz Non-EndPoint REF_CLK14 Reference Clock 14.31818 MHz Non-EndPoint RTCX1 RTCX2 RTC 32.768 KHz Non-EndPoint SATA_CLK100P/N Serial ATA 100 MHz Non-EndPoint UART_CLK UART 48/14.7456 MHz Non-EndPoint USB_CLK96P/N USB 96 MHz Non-EndPoint Depending on the PCH mode used for your design, all clocks may not be required. For example, in Normal mode, all interfaces are active (all EndPoint and Non-EndPoint clocks are required). The other two modes - EndPoint Mode or Non-EndPoint Mode only require partial clock inputs. See Table 31-1 to determine which clocks are required for your mode of operation. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1557 32.0 Signal Descriptions This chapter provides a detailed description of each signal. The signals are arranged in functional groups according to their associated interface. The "#" symbol at the end of the signal name indicates that the active, or asserted state occurs when the signal is at a low voltage level. When "#" is not present, the signal is asserted when at the high voltage level. 32.1 Name Convention Table 32-1 provides the legend for interpreting the Type field that appears throughout the tables in this section. Table 32-1. Signal Type Definitions Type Description # Active low signal I Input pin O Output pin I/OD Bi-directional Input/Open Drain output pin I/O Bi-directional Input /Output pin. OD Open Drain output pin T/S Tri-State pin NC No Connection to pin RSVD Reserved Pin. This signal must be connected as described in signal description. CMOS CMOS buffers The "Type" for each signal is indicative of the functional operating mode of the signal. Unless otherwise noted, a signal is considered to be in the functional operating mode after RTCRST# deasserts for signals in the RTC well, after RSMRST# deasserts for signals in the suspend well, after PWROK asserts for signals in the core well, and after PCIE_EP_RST# deasserts for signals in the EndPoint wells. Intel(R) Communications Chipset 89xx Series - Datasheet 1558 October 2012 Order Number: 327879-001US 32.0 Figure 32-1. Interface Signals Block Diagram P C IE _E P _ T X p[15 :0] P C IE _E P _ T X n[15 :0] P C IE _E P _ R X p [15 :0] P C IE _E P _ R X n [15 :0] P C IE _E P _R S T # P C IE _E P _IC O M P I P C IE _E P _V R E F 1P 8 EP_SM BDAT E P _S M B C LK E P _S M B A LR T # G BE_SM BDAT G B E _S M B C LK G B E _S M B A LR T # G B E _ E E _D O G B E _ E E _D I G B E _ E E _S K G B E _E E _C S # C R U _C LK 100 P /N P C IE _E P _C LK 100 P /N G B E _C LK 100 P /N G B E _W A K E # G B E _A U X _P W R _A V A IL G B E _ A U X _P W R _ O K E P _M A IN _P W R _ O K E P _ JT C K E P _ JTR S T # E P _JT M S E P _JT D I E P _JT D O E ndP oint P C I-E E n d P o in t In te rfa ce G bE MAC 0 S R D S O _ 0_P , S R D S O _0 _N S R D S I_ 0_P , S R D S I_0 _N S F P 0_I2C _D A T A S F P 0_I2C _C LK G B E 0_S W D P [1:0] S R D S 0_S D G B E 0_LE D SM Bus E n d P o in t (S la ve) SM Bus G bE (M a ste r/S la ve) G bE MAC 1 S R D S O _ 1_P , S R D S O _1 _N S R D S I_ 1_P , S R D S I_1 _N S F P 1_I2C _D A T A S F P 1_I2C _C LK G B E 1_S W D P [1:0] S R D S 1_S D G B E 1_LE D G bE MAC 2 S R D S O _ 2_P , S R D S O _2 _N S R D S I_ 2_P , S R D S I_2 _N S F P 2_I2C _D A T A S F P 2_I2C _C LK G B E 2_S W D P [1:0] S R D S 2_S D G B E 2_LE D G bE MAC 3 S R D S O _ 3_P , S R D S O _3 _N S R D S I_ 3_P , S R D S I_3 _N S F P 3_I2C _D A T A S F P 3_I2C _C LK G B E 3_S W D P [1:0] S R D S 3_S D G B E 3_LE D SPI EEPROM C lock Inputs P ow er C ontrol JT A G E ndP oint D M I_T X p[3:0], D M I_T X n[3:0 ] D M I_ R X p[3:0], D M I_ R X n[3:0 ] D M I_IR C O M P D M I_V R E F D irect M edia Interface P C IE _R C _T X p[3:0] P C IE _R C _T X n[3:0] P C IE _R C _R X p [3:0] P C IE _R C _R X n [3:0] P C I-E R oot C om plex Interface S IU [1:0]_T X D S IU [1:0]_ R X D S IU [1 :0 ]_ R T S # S IU [1:0]_ D T R # S IU [1:0]_D S R # S IU [1 :0 ]_ D C D # S IU [1 :0 ]_ C T S # S IU [1 :0 ]_ R I# DUART Interface S A T A 4_ T X p, S A T A 4_T X n S A T A 4_ R X p, S A T A 4 _R X n S A T A 5_ T X p, S A T A 5_T X n S A T A 5_ R X p, S A T A 5 _R X n S A T A _LE D # S erial A T A Interface USB W D T_ T O U T# LP C Interface R C IN # A 20G A T E # CPUPW RG D S E R IR Q CRU _EN B M B U S Y # / G P IO 0 G P IO RTCX1 RTCX2 RTCRST# SRTCRST# IV C C _R T C PW ROK RSMRST# D M I_C LK 100 P /N S A T A _C LK 100 P /N U S B _ C LK 96 P /N P C IC LK U A R T _C LK R E F _ C LK 14 JT C K JT M S JT D I JT D O October 2012 Order Number: 327879-001US P rocessor Interface M isc. S ignals 67 G eneral P urpose I/O S M B us H ost (M aster) S M B us S M LIN K (S lave ) S A T A _IC O M P O S A T A _V R E F 1 P 8 S A T A 4_ G P / G P IO 16 S A T A 5_ G P / T E M P _A LE R T # / G P IO 49 S C LO C K / G P IO 22 S LO A D / G P IO 38 S D A T A O U T 0 / G P IO 39 S D A T A O U T 1 / G P IO 48 U S B [5 :0 ]_ D p; U S B [5:0]_D n O C 0 # / G P IO 59 O C 1 # / G P IO 40 O C 2 # / G P IO 41 O C 3 # / G P IO 42 U S B R B IA S p U S B R B IA S n LA D [3:0] LF R A M E # LD R Q 0# LD R Q 1# / G P IO 23 IN IT 3_3 V # M S T _S M B D A T M S T _S M B C LK M S T _S M B A LE R T # /G P IO 11 S M L1 D A T / G P IO 75 S M L1 C LK / G P IO 58 S M L1 A LE R T # / G P IO 74 IN T R U D E R # SPI F lash S P I_M IS O S P I_M O S I S P I_C LK S P I_C S 0 # S P I_C S 1 # T H R M T R IP # S Y S _R E S E T # S LP _ S 3# S LP _ S 4# S LP _ S 5# / G P IO 63 RTC PECI C lock Inputs JT A G P ow er M gnt. M EPW ROK S Y S _P W R O K DRAM PW RGD PW RBTN# R I# W AKE# P LT R S T # PM SYNC# S U S _ C LK / G P IO 62 S U S _ S T A T # / G P IO 61 Intel(R) Communications Chipset 89xx Series - Datasheet 1559 32.2 JTAG Boundary Scan Chain (BSC) 1149.1 and 1149.6 Chain There is a total of two boundary scan chains: one is dedicated to the EndPoint and the other to the Non-EndPoint section of the chip. The PCH implements a consistent, IEEE 1149.1 compliant JTAG Boundary Scan Chain (BSC) for most interfaces enabling a lowcost manufacturing test for boards. The BSC cell is defined in the Boundary Scan Description Language (BSDL) documentation and the boundary scan is initiated by the JTAG entry. The PCI Express* (EndPoint device) is implemented in JTAG boundary scan IEEE 1149.6. Both protocols 1149.1 and 1149.6 are implemented in a single chain, however the new features are used to validate connectivity of high speed IO pins. By toggling the pins individually or in combination, the connectivity of the pins are validated at the output pin. 32.3 Direct Media Interface Table 32-2. Direct Media Interface Signals Signal Name I/O Type Technology Ball Count DMI_RXn0 DMI_RXp0 I LV Diff DMI_RXn1 DMI_RXp1 I DMI_RXn2 DMI_RXp2 BScan Support Description 2 1149.1 Direct Media Interface Differential Receive Pair 0. LV Diff 2 1149.1 Direct Media Interface Differential Receive Pair 1. I LV Diff 2 1149.1 Direct Media Interface Differential Receive Pair 2. DMI_RXn3 DMI_RXp3 I LV Diff 2 1149.1 Direct Media Interface Differential Receive Pair 3. DMI_TXn0 DMI_TXp0 O LV Diff 2 1149.1 Direct Media Interface Differential Transmit Pair 0. DMI_TXn1 DMI_TXp1 O LV Diff 2 1149.1 Direct Media Interface Differential Transmit Pair 1. DMI_TXn2 DMI_TXp2 O LV Diff 2 1149.1 Direct Media Interface Differential Transmit Pair 2. DMI_TXn3 DMI_TXp3 O LV Diff 2 1149.1 Direct Media Interface Differential Transmit Pair 3. DMI_IRCOMP I Analog 1 DMI_VREF I IO Supply 1 TOTAL External Pull-up 10.5K 1% Direct Media Interface Impedance/Current Compensation Output. Determines DMI output impedance and bias current. Connect pull-up to VCCA1P8_DMI, 1.8V voltage supply. Direct Media Interface Reference Voltage. Connect to VCCA1P8_DMI, 1.8V voltage supply. 18 Intel(R) Communications Chipset 89xx Series - Datasheet 1560 Internal/ External Resistor Pull-Up/ Down October 2012 Order Number: 327879-001US 32.0 32.4 PCI Express* EndPoint Table 32-3. PCI Express* Endpoint Interface Signals (Sheet 1 of 3)1 Signal Name I/O Type Technology Ball Count Internal/ External Resistor Pull-Up/ Down BScan Support Description PCIE_EP_RXn0 PCIE_EP_RXp0 I LV Diff 2 1149.1 & 1149.6 PCI Express* EndPoint Receive Data Pair Lane 0. A serial differential input pair running at a bit rate of 2.5 GT/s or 5 GT/s. The signals for this 1x16 interface can be trained to 1x4, 1x8 or 1x16 ports. (This is skew dependent; check the skew you are designing for.) These port configurations map to signals as follows (y is either "n" or "p"): TRAINING: 1X16 Interface Configuration: * PCIE_EP_RXy[15:0] 1X8 Interface Configuration: * PCIE_EP_RXy[7:0] 1X4 Interface Configuration: * PCIE_EP_RXy[3:0] PCIE_EP_RXn1 PCIE_EP_RXp1 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 1 PCIE_EP_RXn2 PCIE_EP_RXp2 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 2 PCIE_EP_RXn3 PCIE_EP_RXp3 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 3 PCIE_EP_RXn4 PCIE_EP_RXp4 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 4 PCIE_EP_RXn5 PCIE_EP_RXp5 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 5 PCIE_EP_RXn6 PCIE_EP_RXp6 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 6 PCIE_EP_RXn7 PCIE_EP_RXp7 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 7 PCIE_EP_RXn8 PCIE_EP_RXp8 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 8 PCIE_EP_RXn9 PCIE_EP_RXp9 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 9 PCIE_EP_RXn10 PCIE_EP_RXp10 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 10 PCIE_EP_RXn11 PCIE_EP_RXp11 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 11 PCIE_EP_RXn12 PCIE_EP_RXp12 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 12 PCIE_EP_RXn13 PCIE_EP_RXp13 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 13 PCIE_EP_RXn14 PCIE_EP_RXp14 I LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 14 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1561 Table 32-3. PCI Express* Endpoint Interface Signals (Sheet 2 of 3)1 Signal Name I/O Type Technology Ball Count PCIE_EP_RXn15 PCIE_EP_RXp15 I LV Diff 2 Internal/ External Resistor Pull-Up/ Down BScan Support Description 1149.1 & 1149.6 PCI Express EndPoint Receive Data Pair Lane 15 PCIE_EP_TXn0 PCIE_EP_TXp0 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 0. A serial differential output pair running at a bit rate of 2.5 GT/s or 5 GT/s. The signals for this 1x16 interface can be trained to 1x4, 1x8 or 1x16 ports. (This is skew dependent; check the skew you are designing for.) These port configurations map to signals as follows (y is either "n" or "p"): TRAINING: 1X16 Interface Configuration: * PCIE_EP_TXy[15:0] 1X8 Interface Configuration: * PCIE_EP_TXy[7:0] 1X4 Interface Configuration: * PCIE_EP_TXy[3:0] PCIE_EP_TXn1 PCIE_EP_TXp1 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 1 PCIE_EP_TXn2 PCIE_EP_TXp2 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 2 PCIE_EP_TXn3 PCIE_EP_TXp3 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 3 PCIE_EP_TXn4 PCIE_EP_TXp4 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 4 PCIE_EP_TXn5 PCIE_EP_TXp5 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 5 PCIE_EP_TXn6 PCIE_EP_TXp6 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 6 PCIE_EP_TXn7 PCIE_EP_TXp7 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 7 PCIE_EP_TXn8 PCIE_EP_TXp8 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 8 PCIE_EP_TXn9 PCIE_EP_TXp9 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 9 PCIE_EP_TXn10 PCIE_EP_TXp10 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 10 PCIE_EP_TXn11 PCIE_EP_TXp11 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 11 PCIE_EP_TXn12 PCIE_EP_TXp12 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 12 PCIE_EP_TXn13 PCIE_EP_TXp13 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 13 PCIE_EP_TXn14 PCIE_EP_TXp14 O LV Diff 2 1149.1 & 1149.6 PCI Express EndPoint Transmit Data Pair Lane 14 Intel(R) Communications Chipset 89xx Series - Datasheet 1562 October 2012 Order Number: 327879-001US 32.0 Table 32-3. PCI Express* Endpoint Interface Signals (Sheet 3 of 3)1 Signal Name I/O Type Technology Ball Count PCIE_EP_TXn15 PCIE_EP_TXp15 O LV Diff 2 PCIE_EP_ICOMPI I Analog 1 Internal/ External Resistor Pull-Up/ Down BScan Support 1149.1 & 1149.6 External Pull-up 10.5K 1% Description PCI Express EndPoint Transmit Data Pair Lane 15 PCI Express EndPoint Impedance/Current Compensation Output. Connect pull-up to VCCAEP1P8_PE, 1.8V voltage supply. PCIE_EP_RST# I LVTTL 1 PCI Express Fundamental Reset. This is the warm reset for the EndPoint. On the platform it will be tied to either the PCH output PLTRST# or the PCH input SYS_RESET#. PCIE_EP_VREF1P8 I IO Supply 1 PCI Express EndPoint Reference Voltage. Connect to VCCAEP1P8_PE, 1.8V voltage supply. TOTAL 67 1. All transmit and receive signals require AC coupling capacitors. Consult the PDG for capacitor values and routing guidelines. 32.5 . Gigabit Ethernet Interface See the Supported Ethernet PHY Devices for the Intel(R) Communications Chipset 89xx Series Application Note for more information on the PHY devices supported. Table 32-4. SerDes/SGMII Interface Signals (Sheet 1 of 2)1 Signal Name I/O Type Technology Ball Count Internal/ External Resistor Pull-Up/ Down BScan Support Description SRDSI_0_N SRDSI_0_P I LV Diff 2 1149.1 & 1149.6 SerDes/SGMII Serial Data input pair Port 0: Differential SerDes Receive interface. A Serial differential input pair with a data rate of 1.25 Gb/s, and an embedded clock operating at 625 MHz. The embedded clock present in this input is recovered along with the data. SRDSI_1_N SRDSI_1_P I LV Diff 2 1149.1 & 1149.6 SerDes/SGMII Serial Data input pair Port 1. SRDSI_2_N SRDSI_2_P I LV Diff 2 1149.1 & 1149.6 SerDes/SGMII Serial Data input pair Port 2. SRDSI_3_N SRDSI_3_P I LV Diff 2 1149.1 & 1149.6 SerDes/SGMII Serial Data input pair Port 3. 1149.1 & 1149.6 SerDes/SGMII Serial Data output pair Port 0: Differential SerDes Transmit interface. A serial differential output pair with a data rate of 1.25 Gb/s, and a clock operating at 625 MHz. This output carries both data and an embedded clock that is recovered along with data at the receiving end. SRDSO_0_N SRDSO_0_P O October 2012 Order Number: 327879-001US LV Diff 2 Intel(R) Communications Chipset 89xx Series - Datasheet 1563 Table 32-4. SerDes/SGMII Interface Signals (Sheet 2 of 2)1 Signal Name I/O Type Technology Ball Count SRDSO_1_N SRDSO_1_P O LV Diff SRDSO_2_N SRDSO_2_P O SRDSO_3_N SRDSO_3_P SRDS0_SD Internal/ External Resistor Pull-Up/ Down BScan Support Description 2 1149.1 & 1149.6 SerDes/SGMII Serial Data output pair Port 1. LV Diff 2 1149.1 & 1149.6 SerDes/SGMII Serial Data output pair Port 2. O LV Diff 2 1149.1 & 1149.6 SerDes/SGMII Serial Data output pair Port 3. I LVTTL 1 1149.1 Signal Detect Port 0. Indicates that signal (light) is detected from the fiber. High for signal detect. Low otherwise. SRDS1_SD I LVTTL 1 1149.1 Signal Detect Port 1. Indicates that signal (light) is detected from the fiber. High for signal detect. Low otherwise. SRDS2_SD I LVTTL 1 1149.1 Signal Detect Port 2. Indicates that signal (light) is detected from the fiber. High for signal detect Low otherwise. 1149.1 Signal Detect Port 3. Indicates that signal (light) is detected from the fiber. High for signal detect Low otherwise. SRDS3_SD I LVTTL 1 GBE_IRCOMP I Analog 1 GBE_VREF1P8 I IO Supply 1 TOTAL External Pull-up 10.5K 1% GbE Impedance/Current Compensation Output. Connect pull-up to VCCAEP1P8AUX, 1.8V EndPoint core supply. GbE Reference Voltage. Connect to VCCAEP1P8AUX, 1.8V voltage supply. 22 1. All Serial Data Input signals require AC coupling capacitors. Consult the PDG for capacitor values and routing guidelines. Signal Detect Port signal must be pull down with a 10K resistor if not being used. Intel(R) Communications Chipset 89xx Series - Datasheet 1564 October 2012 Order Number: 327879-001US 32.0 Table 32-5. SFP Interface Signals Signal Name SFP0_I2C_CLK SFP0_I2C_DATA SFP1_I2C_CLK SFP1_I2C_DATA SFP2_I2C_CLK SFP2_I2C_DATA SFP3_I2C_CLK SFP3_I2C_DATA I/O Type O,OD I/OD O,OD I/OD O,OD I/OD O,OD I/OD TOTAL Technology LVTTL LVTTL LVTTL LVTTL LVTTL LVTTL LVTTL LVTTL Ball Count Internal/ External Resistor Pull-Up/ Down 1 External pull-up required1 1 External pull-up required1 1 External pull-up required1 1 External pull-up required1 1 External pull-up required1 1 External pull-up required1 1 External pull-up required1 1 External pull-up required1 BScan Support Description 1149.1 Small Form-Factor Pluggable Port 0 I2C Clock. Two Wire Serial Interface (TWSI) Clock Signal. Connects to Mod-Def1 input of SFP (OD). Can also be used as MDC pin (Out). Only one of this modes can be selected at boot with EEPROM. 1149.1 Small Form-Factor Pluggable Port 0 I2C Data. Two Wire Serial Interface (TWSI) Data Signal. Connects to Mod-Def2 input of SFP (I/OD). Can also be used as MDIO pin (I/OD). Only one of this modes can be selected at boot with EEPROM. 1149.1 Small Form-Factor Pluggable Port 1 I2C Clock. Two Wire Serial Interface (TWSI) Clock Signal. Connects to Mod-Def1 input of SFP (OD). Can also be used as MDC pin (Out). Only one of this modes can be selected at boot with EEPROM. 1149.1 Small Form-Factor Pluggable Port 1 I2C Data. Two Wire Serial Interface (TWSI) Data Signal. Connects to Mod-Def2 input of SFP (I/OD). Can also be used as MDIO pin (I/OD). Only one of this modes can be selected at boot with EEPROM. 1149.1 Small Form-Factor Pluggable Port 2 I2C Clock. Two Wire Serial Interface (TWSI) Clock Signal. Connects to Mod-Def1 input of SFP (OD). Can also be used as MDC pin (Out). Only one of this modes can be selected at boot with EEPROM. 1149.1 Small Form-Factor Pluggable Port 2 I2C Data. Two Wire Serial Interface (TWSI) Data Signal. Connects to Mod-Def2 input of SFP (I/OD). Can also be used as MDIO pin (I/OD). Only one of this modes can be selected at boot with EEPROM. 1149.1 Small Form-Factor Pluggable Port 3 I2C Clock. Two Wire Serial Interface (TWSI) Clock Signal. Connects to Mod-Def1 input of SFP (OD). Can also be used as MDC pin (Out). Only one of this modes can be selected at boot with EEPROM. 1149.1 Small Form-Factor Pluggable Port 3 I2C Data. Two Wire Serial Interface (TWSI) Data Signal. Connects to Mod-Def2 input of SFP (I/OD). Can also be used as MDIO pin (I/OD). Only one of this modes can be selected at boot with EEPROM. 8 1. Consult the Platform Design Guide for resistor values. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1565 Table 32-6. GbE EEPROM SPI Interface Signals Internal/ External Resistor Pull-Up/ Down Signal Name I/O Type Technology Ball Count GBE_EE_CS# O LVTTL 1 1149.1 GbE EEPROM Chip Select. Chip Select output to EEPROM. GBE_EE_DI O LVTTL 1 1149.1 GbE EEPROM Data Input. Data is output to EEPROM. GBE_EE_DO I LVTTL 1 1149.1 GbE EEPROM Data Output. Data is input from the EEPROM. GBE_EE_SK O LVTTL 1 1149.1 GbE EEPROM Serial Clock. Serial clock to EEPROM operates at ~2 MHz TOTAL BScan Support Description 4 Table 32-7. GbE SMBus Interface Signals (Master/Slave) Signal Name GBE_SMBCLK GBE_SMBDAT GBE_SMBALRT# I/O Type I/OD I/OD I/OD TOTAL Technology LVTTL LVTTL LVTTL Ball Count 1 Internal/ External Resistor Pull-Up/ Down External pull-up required1 1 External pull-up required1 1 External pull-up required1 BScan Support Description 1149.1 GbE SMBus Clock. One clock pulse is generated for each data bit transferred. External pull-up resistor to VCCPEP3P3AUX is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. 1149.1 GbE SMBus Data. Stable during the high period of the clock (unless it is a start or stop condition). External pull-up resistor to VCCPEP3P3AUX is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. 1149.1 GbE SMBus Alert. Acts as an interrupt of a slave device on SMBus. External pull-up resistor to VCCPEP3P3AUX is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. 3 1. Consult the Platform Design Guide for resistor values. Intel(R) Communications Chipset 89xx Series - Datasheet 1566 October 2012 Order Number: 327879-001US 32.0 Table 32-8. LED, Software Defined, Miscellaneous Signals (Sheet 1 of 3) Internal/ External Resistor Pull-Up/ Down Signal Name I/O Type Technology Ball Count EP_MAIN_PWR_OK I LVTTL 1 EndPoint Main Power OK. Indicates that Main Power Wells are up. This signal can be connected to PWROK. Auxiliary Power Available. This pin is a strapping option pin, latched at the rising edge of PCIE_EP_RST# or InBand PCIe Reset. If this pin is driven high during init time, it indicates that Auxiliary Power is available and the device should support D3COLD power state if enabled to do so. This pin value should be established before GBE_AUX_POWER_OK goes high. GBE_AUX_PWR_AVAIL I LVTTL 1 GBE_AUX_PWR_OK I LVTTL 1 GBE_WAKE# OD LVTTL 1 GBE0_LED GBE1_LED O O LVTTL LVTTL 1 1 BScan Support External pullup required1 1149.1 Description Auxiliary Power OK. Power good reset for AUX well. Indicates power to AUX well is stable. External pullup required1 Internal Pull-up (Active for 200us after assertion of AUX_PWROK) Internal Pull-up (Active for 200us after assertion of AUX_PWROK) 1149.1 Wake on LAN. Wake Output signal for wake on LAN. 1149.1 Port 0 LED. Programmable LED, mode encoding set by GbE EEPROM. On power-up this signal becomes an input to Strap SMBus Slave Address bit 2 The EP SMBus slave is accessed with address[7:1] = 1110_XX0. Sampling occurs during the first 200ns after assertion of AUX_PWROK. Thereafter the pin functions as an LED output. 1149.1 Port 1 LED. Programmable LED, mode encoding set by GbE EEPROM. On power-up this signal becomes an input to Strap SMBus Slave Address bit 3. The EP SMBus slave is accessed with address[7:1] = 1110_XX0. Sampling occurs during the first 200ns after assertion of AUX_PWROK. Thereafter the pin functions as an LED output. Port 2 LED. Programmable LED, mode encoding set by GbE EEPROM. GBE2_LED O October 2012 Order Number: 327879-001US LVTTL 1 Internal Pull-up (Active for 200us after assertion of AUX_PWROK) 1149.1 This signal can be use as a strapping function to enable or disable PCIe* SRIOV GBE2_LED: 0 Disable SRIOV (Pull-Down Required) 1 Enable SRIOV (default) Sampling occurs during the first 200ns after assertion of AUX_PWROK. Thereafter the pin functions as an LED output. Intel(R) Communications Chipset 89xx Series - Datasheet 1567 Table 32-8. LED, Software Defined, Miscellaneous Signals (Sheet 2 of 3) Signal Name GBE3_LED GBE0_SWDP[1:0] GBE1_SWDP[1:0] I/O Type O I/O I/O Technology LVTTL LVTTL LVTTL Ball Count 1 2 2 Intel(R) Communications Chipset 89xx Series - Datasheet 1568 Internal/ External Resistor Pull-Up/ Down Internal Pull-up (Active for 200us after assertion of AUX_PWROK) BScan Support Description 1149.1 Port 3 LED. Programmable LED, mode encoding set by GbE EEPROM. This signal can be use as a strapping function to control reset sequence. 0 Silicon A0 Compatible Reset Sequence (Pull-Down Required) 1 PCIe* CEM 2.0 Compliant Reset sequence (default) Sampling occurs during the first 200ns after assertion of AUX_PWROK. Thereafter the pin functions as a LED output. 1149.1 Software Defined Pins for function 0. These pins are software programmable w/with input/output capability. These default to inputs upon power up, but may have their direction and output values defined in the EEPROM. The SDP bits may be mapped to the General Purpose Interrupt bits when configured as inputs. The GBE0_SWDP0 pin can be used as a watchdog output indication. All the SDP pins can be used as SFP sideband signals (TxDisable, present & TxFault). This product does not use these signals; it is available for SW control over SFP. 1149.1 Software Defined Pins for function 1. These pins are software programmable w/with input/output capability. These default to inputs upon power up, but may have their direction and output values defined in the EEPROM. The SDP bits may be mapped to the General Purpose Interrupt bits when configured as inputs. The GBE1_SDP0 pin can be used as a watchdog output indication. All the SDP pins can be used as SFP sideband signals (TxDisable, present & TxFault). This product does not use these signals; it is available for SW control over SFP. October 2012 Order Number: 327879-001US 32.0 Table 32-8. LED, Software Defined, Miscellaneous Signals (Sheet 3 of 3) Signal Name GBE2_SWDP[1:0] GBE3_SWDP[1:0] I/O Type I/O I/O TOTAL Technology LVTTL LVTTL Ball Count 2 2 Internal/ External Resistor Pull-Up/ Down BScan Support Description 1149.1 Software Defined Pins for function 2. These pins are software programmable w/with input/output capability. These default to inputs upon power up, but may have their direction and output values defined in the EEPROM. The SDP bits may be mapped to the General Purpose Interrupt bits when configured as inputs. The GBE2_SDP0 pin can be used as a watchdog output indication. All the SDP pins can be used as SFP sideband signals (TxDisable, present & TxFault). This product does not use these signals; it is available for SW control over SFP. 1149.1 Software Defined Pins for function 3. These pins are software programmable w/with input/output capability. These default to inputs upon power up, but may have their direction and output values defined in the EEPROM. The SDP bits may be mapped to the General Purpose Interrupt bits when configured as inputs. The GBE3_SDP0 pin can be used as a watchdog output indication. All the SDP pins can be used as SFP sideband signals (TxDisable, present & TxFault). This product does not use these signals; it is available for SW control over SFP. 16 1. Consult the Platform Design Guide for resistor values. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1569 32.6 EndPoint Management SMBus Interface (Slave) Table 32-9. EndPoint Management SMBus Interface Signals (Slave) Signal Name I/O Type EP_SMBCLK EP_SMBDAT EP_SMBALRT# I I/OD OD Technology LVTTL LVTTL LVTTL TOTAL Ball Count 1 1 1 Internal/ External Resistor Pull-Up/ Down External pull-up required1 External pull-up required1 External pull-up required1 BScan Support Description 1149.1 Management SMBus Clock. One clock pulse is received for each data bit transferred. External pull-up resistor to VCCPEP3P3AUX is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. 1149.1 Management SMBus Data. Stable during the high period of the clock (unless it is a start or stop condition). External pull-up resistor to VCCPEP3P3AUX is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. 1149.1 Management SMBus Alert. Acts as an interrupt of a slave device on SMBus. External pull-up resistor to VCCPEP3P3AUX is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. 3 1. Consult the Platform Design Guide for resistor values. 32.7 PCI Express* Root Complex Table 32-10. PCI Express* Root Complex Interface Signals (Sheet 1 of 2)1 Signal Name Technology Ball Count PCIE_RC_RXp0 PCIE_RC_Rxn0 I LV Diff 2 1149.1 PCI Express Root Complex Receive Data Pair Lane 0 PCIE_RC_RXp1 PCIE_RC_Rxn1 I LV Diff 2 1149.1 PCI Express Root Complex Receive Data Pair Lane 1 PCIE_RC_RXp2 PCIE_RC_RXn2 I LV Diff 2 1149.1 PCI Express Root Complex Receive Data Pair Lane 2 PCIE_RC_RXp3 PCIE_RC_RXn3 I LV Diff 2 1149.1 PCI Express Root Complex Receive Data Pair Lane 3 PCIE_RC_TXp0 PCIE_RC_TXn0 O LV Diff 2 1149.1 PCI Express Root Complex Transmit Data Pair Lane 0 Intel(R) Communications Chipset 89xx Series - Datasheet 1570 Internal/ External Resistor Pull-Up/ Down I/O Type BScan Support Description October 2012 Order Number: 327879-001US 32.0 Table 32-10. PCI Express* Root Complex Interface Signals (Sheet 2 of 2)1 Internal/ External Resistor Pull-Up/ Down Signal Name I/O Type Technology Ball Count PCIE_RC_TXp1 PCIE_RC_TXn1 O LV Diff 2 1149.1 PCI Express Root Complex Transmit Data Pair Lane 1 PCIE_RC_TXp2 PCIE_RC_TXn2 O LV Diff 2 1149.1 PCI Express Root Complex Transmit Data Pair Lane 2 PCIE_RC_TXp3 PCIE_RC_TXn3 O LV Diff 2 1149.1 PCI Express Root Complex Transmit Data Pair Lane 3 TOTAL BScan Support Description 16 1. All Transmit signals require AC coupling capacitors. Consult the PDG for capacitor values and routing guidelines. 32.8 Serial ATA Interface The PCH supports only two SATA ports (ports 4 & 5). Table 32-11. SATA Interface Signals (Sheet 1 of 3)1 Signal Name SATA4_RXN SATA4_RXP SATA4_TXN SATA4_TXP SATA5_GP / TEMP_ALERT# / GPIO49 I/O Type I O I/O October 2012 Order Number: 327879-001US Technology LV Diff LV Diff LVTTL Ball Count 2 2 1 Internal/ External Resistor Pull-Up/ Down BScan Support Description 1149.1 Serial ATA 4 Differential Receive Pair. These are inbound high-speed differential signals from Port 4. In compatible mode (IDE Mode), SATA Port 4 is the primary master of SATA Controller. Note: Port4 corresponds to signal SATA4 1149.1 Serial ATA 4 Differential Transmit Pairs. These are outbound high-speed differential signals to Port 4. In compatible mode (IDE Mode), SATA Port 4 is the primary master of SATA Controller. Note: Port4 corresponds to signal SATA4 1149.1 Serial ATA 5 General Purpose. This is an input pin which can be configured as an interlock switch corresponding to SATA Port 5. When used as an interlock switch status indication, this signal should be drive to `0' to indicate that the switch is closed and to `1' to indicate that the switch is open. Temperature Alert. Used as an alert (active low) to indicate to the external controller (such as, EC or SIO) that temperatures are out of range for the PCH or Memory Controller or the processor core. If interlock switches or Temp Alert are not required, this pin can be configured as GPIO Port 49. Intel(R) Communications Chipset 89xx Series - Datasheet 1571 Table 32-11. SATA Interface Signals (Sheet 2 of 3)1 Signal Name SATA5_RXN SATA5_RXP SATA5_TXN SATA5_TXP SDATAOUT0 / GPIO39 SDATAOUT1 / GPIO48 SCLOCK / GPIO22 SLOAD / GPIO38 I/O Type I O I/O I/O I/O I/O Technology LV Diff LV Diff LVTTL LVTTL LVTTL LVTTL Ball Count 2 2 1 1 1 1 SATA_LED# OD LVTTL 1 SATA_VREF1P8 I IO Supply 1 Intel(R) Communications Chipset 89xx Series - Datasheet 1572 Internal/ External Resistor Pull-Up/ Down External pull-up required2 BScan Support Description 1149.1 Serial ATA 5 Differential Receive Pair. These are inbound high-speed differential signals from Port 5. In compatible mode (IDE Mode), SATA Port 5 is the secondary master of SATA Controller. Note: Port5 corresponds to signal SATA5 1149.1 Serial ATA 5 Differential Transmit Pair. These are outbound high-speed differential signals to Port 5. In compatible mode (IDE Mode), SATA Port 5 is the secondary master of SATA Controller. Note: Port5 corresponds to signal SATA5 1149.1 SATA Serial GPIO Data Out 0. Driven by the controller to indicate the drive status in the following sequence: drive 4, 5...4,5... If SDATAOUT0 interface is not used, the signals can be used as GPIO Port 39. 1149.1 SATA Serial GPIO Data Out 1. Driven by the controller to indicate the drive status in the following sequence: drive 4, 5...4,5... If SDATAOUT1 interface is not used, the signals can be used as GPIO Port 48. 1149.1 SGPIO Reference Clock. The SATA controller uses rising edges of this clock to transmit serial data, and the target uses the falling edge of this clock to latch data. If SCLOCK interface is not used, this signal can be used as a GPIO Port 22. 1149.1 SATA Serial GPIO Load. The controller drives a `1' at the rising edge of SCLOCK to indicate either the start or end of a bit stream. A 4-bit vendor specific pattern will be transmitted right after the signal assertion. If SLOAD interface is not used, this signal can be used as a GPIO Port 38. 1149.1 Serial ATA LED. This signal is an opendrain output pin driven during SATA command activity. It is to be connected to external circuitry that can provide the current to drive a platform LED. When active, the LED is on. When tri-stated, the LED is off. An external pull-up resistor to VCC3P3 is required. Serial ATA Reference Voltage. Connect to VCCA1P8_SATA, 1.8V Core reference voltage. October 2012 Order Number: 327879-001US 32.0 Table 32-11. SATA Interface Signals (Sheet 3 of 3)1 I/O Type Signal Name Technology Ball Count SATA4_GP / GPIO16 I/O LVTTL 1 SATA_ICOMPO I Analog 1 TOTAL Internal/ External Resistor Pull-Up/ Down BScan Support Description 1149.1 Serial ATA 4 General Purpose. This is an input pin which can be configured as an interlock switch corresponding to SATA Port 4. When used as an interlock switch status indication, this signal should be drive to `0' to indicate that the switch is closed and to `1' to indicate that the switch is open. If interlock switches are not required, this pin can be configured as GPIO Port 16. External Pull-up 10.5K 1% Serial ATA Compensation Output. Connected pull-up resistor to VCCA1P8_SATA. 17 1. All Transmit and Receive signals (SATA4/5_TXN/P and SATA4/5_RXN/P) require AC coupling capacitors. Consult the PDG for capacitor values and routing guidelines. 2. Consult the Platform Design Guide for resistor values. 32.9 LPC Interface Processor booting is done via the SPI Boot Interface, NOT the LPC Interface. Table 32-12. LPC Interface Signals Internal/ External Resistor Pull-Up/ Down Signal Name I/O Type Technology Ball Count INIT3_3V# O LVTTL 1 LAD[3:0] I/O LVTTL 4 Internal pull-up 1 Internal pull-up LDRQ0# I LVTTL LDRQ1# / GPIO23 I/O LVTTL 1 LFRAME# O LVTTL 1 TOTAL October 2012 Order Number: 327879-001US Internal pull-up BScan Support Description 1149.1 Initialization 3.3 V. INIT3_3V# is asserted driven out for 16 PCI clocks to reset the processor. 1149.1 Low Pin Count Address Data Bus. Data/ Address signals for the LPC bus. 1149.1 LPC Serial DMA/Master Request Input Bit 0. Used by LPC devices, such as Super I/O chips, to request DMA or bus master access. This signal is typically connected to external Super I/O device. 1149.1 LPC Serial DMA/Master Request Input Bit 1. Used by LPC devices, such as Super I/O chips, to request DMA or bus master access. This signal is typically connected to external Super I/O device. An internal pull-up resistor is provided on these signals. If LDRQ1# interface is not used, this signal can be used as a GPIO Port 23. 1149.1 LPC Frame. This signal is multiplexed with the LPC LFRAME# signal. 8 Intel(R) Communications Chipset 89xx Series - Datasheet 1573 32.10 USB Interface Table 32-13. USB Interface Signals (Sheet 1 of 2) Signal Name OC0# OC1# OC2# OC3# / / / / GPIO59 GPIO40 GPIO41 GPIO42 USB0_Dn USB0_Dp USB1_Dn USB1_Dp USB2_Dn USB2_Dp USB3_Dn USB3_Dp USB4_Dn USB4_Dp USB5_Dn USB5_Dp I/O Type I/O I/O I/O I/O Technology LVTTL LV Diff LV Diff LV Diff Ball Count 4 4 4 4 Intel(R) Communications Chipset 89xx Series - Datasheet 1574 Internal/ External Resistor Pull-Up/ Down Internal pull-down Internal pull-down Internal pull-down BScan Support Description 1149.1 Overcurrent Indicators. These signals set corresponding bits in the USB controllers to indicate that an overcurrent condition has occurred. OC[3:0]# may optionally be used as GPIO Ports [42,41,40,59]. Notes: 1. OC# pins are 3.3V and NOT 5 V tolerant. 2. OC# pins must be shared between ports. 3. OC#[3:0] can only be used for EHCI controller #1. 1149.1 Universal Serial Bus Port [1:0] Differential. These differential pairs are used to transmit Data/Address/Command signals for ports 0 and 1. These ports can be routed to UHCI controller #1 or EHCI controller #1. Note: No external resistors are required on these signals. The driver contains an integrated 15 k pulldown and provides an output driver impedance of 45 which requires no external series resistor. 1149.1 Universal Serial Bus Port [3:2] Differential. These differential pairs are used to transmit data/address/command signals for ports 2 and 3. These ports can be routed to UHCI controller #2 or the EHCI controller #1. Note: No external resistors are required on these signals. The driver contains an integrated 15 k pulldown and provides an output driver impedance of 45 which requires no external series resistor. 1149.1 Universal Serial Bus Port [5:4] Differential. These differential pairs are used to transmit Data/Address/Command signals for ports 4 and 5. These ports can be routed to UHCI controller #3 or the EHCI controller #1. Note: No external resistors are required on these signals. The driver contains an integrated 15 k pulldown and provides an output driver impedance of 45 which requires no external series resistor. October 2012 Order Number: 327879-001US 32.0 Table 32-13. USB Interface Signals (Sheet 2 of 2) Technology Ball Count Internal/ External Resistor Pull-Up/ Down I Analog 1 External pull-down required1 USB Resistor Bias Complement. Analog connection point for an external resistor. Used to set transmit currents and internal load resistors. O Analog 1 External pull-down required1 USB Resistor Bias. Analog connection point for an external resistor. Used to set transmit currents and internal load resistors. Signal Name I/O Type USB_RBIASn USB_RBIASp TOTAL BScan Support Description 18 1. Consult the Platform Design Guide for resistor values. 32.11 UART Interface Table 32-14. UART Interface Signals (Sheet 1 of 3) Signal Name I/O Type Technology Ball Count Internal/ External Resistor Pull-Up/ Down BScan Support Description SIU0_CTS# I LVTTL 1 1149.1 UART Port 0 Clear to Send: Active low, this pin indicates that data can be exchanged between UART port 0 and the external interface. This pin has no effect on the transmitter. SIU0_DCD# I LVTTL 1 1149.1 UART Port 0 Data Carrier Detect: Active low, this pin indicates that data carrier has been detected by the external agent for UART port 0. 1149.1 UART Port 0 Data Set Ready: Active low, this pin indicates that the external agent is ready to communicate with UART port 0. This pin has no effect on the transmitter. 1149.1 UART Port 0 Data Terminal Ready: When low, this pin informs the modem or data set that the UART port 0 is ready to establish a communication link. SIW Configuration Port Address Select Strap: this strap selects the IO address for the SIW configuration port. Sampling occurs on the rising edge of PWROK. The straps are defined as follows: 0 = IO Addresses 02Eh and 02Fh (Pulldown required) 1 = IO Addresses 04Eh and 04Fh (default) SIU0_DSR# SIU0_DTR# I O October 2012 Order Number: 327879-001US LVTTL LVTTL 1 1 Weak Internal pull-up Intel(R) Communications Chipset 89xx Series - Datasheet 1575 Table 32-14. UART Interface Signals (Sheet 2 of 3) Signal Name SIU0_RI# I/O Type I Technology LVTTL Ball Count 1 BScan Support Description 1149.1 UART Port 0 Ring Indicator: Active low, this pin indicates that a telephone ringing signal has been received by the external agent for UART port 0. Note: This pin is Modem Status Input whose condition can be tested by the processor by reading bit 6 (RI) of the MSR. Bit 6 is the complement of the RI# signal. Bit 2 (TERI) of the MSR indicates whether the RI# input has transitioned back to an inactive state. When the RI bit of the MSR changes from a 1 to 0 an interrupt is generated if the Modem Status Interrupt is enabled. SIU0_RTS# O LVTTL 1 1149.1 UART Port 0 Request To Send: When low this pin informs the modem or data set that UART port 0 is wants to send data on an established communication link. The RTS# output signal can be set to an active low by programming the RTS (bit 1) of the Modem Control Register to a logic `1'. A Reset operation sets this signal to its inactive state (logic `1'). LOOP mode operation holds this signal in its inactive state. SIU0_RXD I LVTTL 1 1149.1 UART Port 0 Serial Data Input: Serial data input form device pin to the receive port for UART port 0. 1149.1 UART Port 0 Serial Data Output: Serial data output to the communication peripheral/modem or data set for UART port 0. Upon reset, the TXD pins will be set to MARKING condition (logic `1' state). SIU0_TXD O LVTTL 1 SIU1_CTS# I LVTTL 1 1149.1 UART Port 1 Clear to Send: Active low, this pin indicates that data can be exchanged between UART port 1 and the external interface. This pin has no effect on the transmitter. SIU1_DCD# I LVTTL 1 1149.1 UART Port 1 Data Carrier Detect: Active low, this pin indicates that data carrier has been detected by the external agent for UART port 1. SIU1_DSR# I LVTTL 1 1149.1 UART Port 1 Data Set Ready: Active low, this pin indicates that the external agent is ready to communicate with UART port 1. This pin has no effect on the transmitter. SIU1_DTR# O LVTTL 1 1149.1 UART Port 1 Data Terminal Ready: When low, this pin informs the modem or data set that the UART port 1 is ready to establish a communication link. Intel(R) Communications Chipset 89xx Series - Datasheet 1576 Internal/ External Resistor Pull-Up/ Down October 2012 Order Number: 327879-001US 32.0 Table 32-14. UART Interface Signals (Sheet 3 of 3) Signal Name I/O Type SIU1_RI# I Technology LVTTL Ball Count 1 Internal/ External Resistor Pull-Up/ Down BScan Support Description 1149.1 UART Port 1 Ring Indicator: Active low, this pin indicates that a telephone ringing signal has been received by the external agent for UART port 1. Note: This pin is Modem Status Input whose condition can be tested by the processor by reading bit 6 (RI) of the MSR. Bit 6 is the complement of the RI# signal. Bit 2 (TERI) of the MSR indicates whether the RI# input has transitioned back to an inactive state. When the RI bit of the MSR changes from a 1 to 0 an interrupt is generated if the Modem Status Interrupt is enabled. SIU1_RTS# O LVTTL 1 1149.1 UART Port 1 Request To Send: When low this pin informs the modem or data set that UART port 1 is wants to send data on an established communication link. The RTS# output signal can be set to an active low by programming the RTS (bit 1) of the Modem Control Register to a logic `1'. A Reset operation sets this signal to its inactive state (logic `1'). LOOP mode operation holds this signal in its inactive state. SIU1_RXD I LVTTL 1 1149.1 UART Port 1 Serial Data Input: Serial data input form device pin to the receive port for UART port 1. 1149.1 UART Port 1 Serial Data Output: Serial data output to the communication peripheral/modem or data set for UART port 1. Upon reset, the TXD pins will be set to MARKING condition (logic `1' state). 1149.1 The WatchDog Timer output. The signal is driven low when the main 35-bit down counter reaches zero during the second stage. For use by board management controller. SIU1_TXD O WDT_TOUT# O TOTAL October 2012 Order Number: 327879-001US LVTTL LVTTL 1 1 17 Intel(R) Communications Chipset 89xx Series - Datasheet 1577 32.12 Host SMBus (Master) Interface Table 32-15. Host SMBus (Master) Interface Signals I/O Type Signal Name MST_SMBCLK I/OD MST_SMBDAT I/OD MST_SMBALERT# / GPIO11 SML1CLK / GPIO58 SML1DAT / GPIO75 SML1ALERT# / GPIO74 I/O I/O I/O I/O TOTAL Technology LVTTL LVTTL LVTTL LVTTL LVTTL LVTTL Ball Count 1 1 1 1 1 1 Internal/ External Resistor Pull-Up/ Down External pull-up required1 External pull-up required1 External pull-up required1 When SMBUS: External pull-up required1 When SMBUS: External pull-up required1 When SMBUS: External pull-up required1 BScan Support Description 1149.1 Master SMBus Clock. One clock pulse is generated for each data bit transferred. External pull-up resistor to VCCSUS3P3 is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. 1149.1 Master SMBus Data. Stable during the high period of the clock (unless it is a start or stop condition). External pull-up resistor to VCCSUS3P3 is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. 1149.1 SMBus Alert. This signal is used to wake the system or generate SMI#. External pull-up resistor to VCCSUS3P3 is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. This signal can also be configured to GPIO Port 11. 1149.1 System Management Link 1 Clock. SMBus link to external BMC. External pull-up resistor to VCCSUS3P3 is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. If SML1CLK interface is not used, the signals can be used as GPIO Port 58. 1149.1 System Management Link 1 Data. SMBus link to external BMC. External pull-up required to VCCSUS3P3 is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. If SML1DAT interface is not used, the signals can be used as GPIO Port 75. 1149.1 System Management Link Alert 1. This signal can be connected to an external BMC. External pull-up resistor to VCCSUS3P3 is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. If SML1ALERT# interface is not used, the signals can be used as GPIO Port 74. 6 1. Consult the Platform Design Guide for resistor values. Intel(R) Communications Chipset 89xx Series - Datasheet 1578 October 2012 Order Number: 327879-001US 32.0 32.13 Serial Peripheral Interface Boot Interface Note: Use SPI Flash to boot the system. Table 32-16. SPI Boot Interface Signals Signal Name I/O Type Technology Ball Count SPI_CLK O LVTTL SPI_CS0# O SPI_CS1# Internal/ External Resistor Pull-Up/ Down BScan Support Description 1 1149.1 SPI Clock. SPI clock signal, during idle the bus owner will drive the clock signal low. LVTTL 1 1149.1 SPI Chip Select 0. Used as the SPI bus request signal. O LVTTL 1 Weak Internal pull-up 1149.1 SPI Chip Select 1. Used as the SPI bus request signal. SPI_MISO I LVTTL 1 Weak Internal pull-up 1149.1 SPI Master IN Slave OUT. Data input pin. SPI_MOSI O LVTTL 1 1149.1 SPI Master OUT Slave IN. Data output pin. Description TOTAL 32.14 5 Interrupt Interface Table 32-17. Interrupt Interface Signals Signal Name I/O Type SERIRQ IO TOTAL Technology Ball Count Internal/ External Resistor Pull-Up/ Down BScan Support LVTTL 1 External pull-up required1 1149.1 Serial Interrupt Request. This pin implements the serial interrupt protocol. 1 1. Consult the Platform Design Guide for resistor values. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1579 32.15 Processor Interface Table 32-18. Processor Interface Signals Signal Name I/O Type A20GATE I CPUPWRGD O RCIN# I TOTAL Technology LVTTL LVCMOS LVTTL Ball Count Internal/ External Resistor Pull-Up/ Down 1 Depends on driver 1 1 Description 1149.1 A20 Gate. A20GATE is from the keyboard controller. The signal acts as an alternative method to force the A20M# signal active. It saves the external OR gate needed with various other chipsets. 1149.1 CPU Power Good. This signal should be connected to the processor's VCCPWRGOOD_1 and VCCPWRGOOD_0 input to indicate when the processor power is valid. Logic level is determined by the voltage value of VCCPCPU. 1149.1 Keyboard Controller Reset CPU. The keyboard controller can generate INIT# to the processor. This saves the external OR gate with the PCH's other sources of INIT#. When the PCH detects the assertion of this signal, INIT# is generated for 16 PCI clocks. Note: The PCH will ignore RCIN# assertion during transitions to the S1, S3, S4, and S5 states. 3 Intel(R) Communications Chipset 89xx Series - Datasheet 1580 Depends on driver BScan Support October 2012 Order Number: 327879-001US 32.0 32.16 Power Management Interface Table 32-19. Power Management Interface Signals (Sheet 1 of 3) Signal Name I/O Type Technology Ball Count Internal/ External Resistor Pull-Up/ Down BScan Support DRAM Power OK. This signal should connect to the processor's SM_DRAMPWROK pin. The PCH asserts this pin to indicate when DRAM power is on. DRAMPWRGD O LVTTL 1 MEPWROK I LVTTL 1 Management Engine Power OK. When asserted, indicates that power to the ME subsystem is stable. Connect to PWROK on platform. 1 1149.1 Platform Environment Control Interface: Single-wire, serial bus. Connect to corresponding pin of the processor for accessing processor digital thermometer. Logic level is determined by the value of VCCPCPU. PECI I/O SSTL 1149.1 Description PLTRST# O LVTTL 1 1149.1 Platform Reset. The PCH asserts PLTRST# to reset devices on the platform (e.g., SIO, LAN, Processor, etc.). The PCH asserts PLTRST# during power-up and when S/W initiates a hard reset sequence through the Reset Control register (I/O Register CF9h). The PCH drives PLTRST# inactive a minimum of 1 ms after both PWROK and SYS_PWROK are driven high. The PCH drives PLTRST# active a minimum of 1 ms when initiated through the Reset Control register (I/O Register CF9h). Note: PLTRST# is in the VCCSUS3P3 well. PMSYNC O LVCMOS 1 1149.1 Power Management Sync. Provides state information from the PCH to the Processor relevant to C-state transitions. Logic level is determined by the value of VCCPCPU. PWRBTN# I LVTTL 1 Internal pull-up 1149.1 Power Button. The Power Button will cause SMI# or SCI to indicate a system request to go to a sleep state. If the system is already in a sleep state, this signal will cause a wake event. If PWRBTN# is pressed for more than 4 seconds, this will cause an unconditional transition (power button override) to the S5 state. Override will occur even if the system is in the S1-S4 states. This signal has an internal pull-up resistor and has an internal 16 ms de-bounce on the input. RI# I LVTTL 1 Internal pull-up 1149.1 Ring Indicate. This signal is an input from a modem. It can be enabled as a wake event, and this is preserved across power failures. 1149.1 S3 Sleep Control. SLP_S3# is for power plane control. This signal shuts off power to all non-critical systems when in S3 (Suspend To RAM), S4 (Suspend to Disk), or S5 (Soft Off) states. SLP_S3# O October 2012 Order Number: 327879-001US LVTTL 1 Intel(R) Communications Chipset 89xx Series - Datasheet 1581 Table 32-19. Power Management Interface Signals (Sheet 2 of 3) Signal Name SLP_S4# SLP_S5# / GPIO63 SUS_CLK / GPIO62 SUS_STAT# / GPIO61 SYS_PWROK SYS_RESET# I/O Type O I/O I/O I/O I I Technology LVTTL LVTTL LVTTL LVTTL LVTTL LVTTL Ball Count 1 1 1 1 BScan Support 1 Description 1149.1 S4 Sleep Control. SLP_S4# is for power plane control. This signal shuts power to all non-critical systems when in the S4 (Suspend to Disk) or S5 (Soft Off) state. Note: This pin must be used to control the DRAM power in order to use the PCH's DRAM power-cycling feature. 1149.1 S5 Sleep Control. SLP_S5# is for power plane control. This signal is used to shut power off to all non-critical systems when in the S5 (Soft Off) states. If SLP_S5# interface is not used, the signals can be used as GPIO Port 63. 1149.1 Suspend Clock. This clock is an output of the RTC generator circuit. It is used by other chips for refresh clock. If SUS_CLK interface is not used, the signals can be used as GPIO Port 62. 1149.1 Suspend Status: This signal is asserted by the PCH to indicate that the system will be entering a low power state soon. This can be monitored by devices with memory that need to switch from normal refresh to suspend refresh mode. It can also be used by other peripherals as an indication that they should isolate their outputs that may be going to powered-off planes. If SUS_STAT# interface is not used, this signal can be used as a GPIO Port 61. System Power OK. This generic power good input to the PCH is driven and utilized in a platform-specific manner. While PWROK always indicates that the CORE well of the PCH is stable, SYS_PWROK is used to inform the PCH that power is stable to some other system component(s) and the system is ready to start the exit from reset. The particular component(s) associated with SYS_PWROK can vary across platform types supported by the same generation of the PCH. Depending on the platform, the PCH may expect (and wait) for SYS_PWROK at different stages of the boot flow before continuing. Note: Consult the platform design guide for instructions on how to connect SYS_PWROK for that platform. 1 Intel(R) Communications Chipset 89xx Series - Datasheet 1582 Internal/ External Resistor Pull-Up/ Down External pull-up required1 1149.1 System Reset. This pin forces an internal reset after being debounced. The PCH will reset immediately if the SMBus is idle; otherwise, it will wait up to 25 ms 2 ms for the SMBus to idle before forcing a reset on the system. October 2012 Order Number: 327879-001US 32.0 Table 32-19. Power Management Interface Signals (Sheet 3 of 3) Signal Name I/O Type Technology Ball Count THRMTRIP# I LVCMOS 1 WAKE# I LVTTL 1 TOTAL Internal/ External Resistor Pull-Up/ Down External pull-up required1 BScan Support Description 1149.1 Thermal Trip. When low, this signal indicates that a thermal trip from the processor occurred, and the PCH will immediately transition to a S5 state. The PCH will not wait for the processor stop grant cycle since the processor has overheated. Logic level is determined by the value of VCCPCPU. 1149.1 PCI Express Wake Event. Sideband wake signal on PCI Express asserted by components requesting wake up. 16 1. Consult the Platform Design Guide for resistor values. 32.17 Thermal Sensor Current Reference Table 32-20. Thermal Sensor Current Reference Technology Ball Count Internal/ External Resistor Pull-Up/ Down I Analog 1 External pull-down 8.06K 1% to VSS Thermal Sensor 0 Current Reference. Connect to an external pull down resistor to ground. I Analog 1 External pull-down 8.06K 1% to VSS Thermal Sensor 1 Current Reference. Connect to an external pull down resistor to ground. Signal Name I/O Type TS0_IREF_N TS1_IREF_N TOTAL October 2012 Order Number: 327879-001US BScan Support Description 2 Intel(R) Communications Chipset 89xx Series - Datasheet 1583 32.18 Miscellaneous Interface Table 32-21. Miscellaneous Interface Signals Signal Name I/O Type Technology Ball Count Internal/ External Resistor Pull-Up/ Down BScan Support Description BMBUSY# / GPIO0 I/O LVTTL 1 1149.1 Bus Master Busy. This signal is used to support the C3 state. It indicates that a bus master device is busy. When this signal is asserted, the BM_STS bit will be set. If this signal goes active in a C3 state, it is treated as a break event. Note: This signal is internally synchronized using the PCICLK and a two-stage synchronizer. It does not need to meet any particular setup or hold time. This signal can also be used as GPIO Port 0. EP_CRU_EN I LVTTL 1 1149.1 EndPoint CRU (Clock Resource Unit) Enable. Enables CRU Clock inputs CRU_CLK100[P/N] TOTAL 32.19 2 General Purpose Input/Output Interface There are 67 GPIO pins (multiplexed or dedicated). Table 32-22. General Purpose I/O Interface Signals (Sheet 1 of 8) Signal Name I/O Type Default Mode Ball Count Default Direction and Logic State Power Well Internal/ External Resistor Pull-Up/ Down Description BMBUSY#/GPI0 I/O GPI 1 I CORE Bus Master Busy. This signal is used to support the C3 state. It indicates that a bus master device is busy. When this signal is asserted, the BM_STS bit will be set. If this signal goes active in a C3 state, it is treated as a break event. Note: This signal is internally synchronized using the PCICLK and a two-stage synchronizer. It does not need to meet any particular setup or hold time. This signal can also be used as GPIO Port 0. GPIO1 I/O GPI 1 I CORE General Purpose I/O Port 1. Not Multiplexed. GPIO2 1 I/O GPI 1 I CORE General Purpose I/O Port 2. Not Multiplexed. GPIO3 1 I/O GPI 1 I CORE General Purpose I/O Port 3. Not Multiplexed. GPIO4 1 I/O GPI 1 I CORE General Purpose I/O Port 4. Not Multiplexed. Intel(R) Communications Chipset 89xx Series - Datasheet 1584 October 2012 Order Number: 327879-001US 32.0 Table 32-22. General Purpose I/O Interface Signals (Sheet 2 of 8) Default Mode Ball Count Default Direction and Logic State Power Well I/O GPI 1 I CORE General Purpose I/O Port 5. Not Multiplexed. GPIO6 I/O GPI 1 I CORE General Purpose I/O Port 6. Not Multiplexed. GPIO7 I/O GPI 1 I CORE General Purpose I/O Port 7. Not Multiplexed. GPIO8 I/O GPO 1 O (High) SUS GPIO92 I/O Native 1 I SUS General Purpose I/O Port 9. Not Multiplexed.3 GPIO102 I/O Native 1 I SUS General Purpose I/O Port 10. Not Multiplexed.3 SMBus Alert. This signal is used to wake the system or generate SMI#. External pull-up resistor to VCCSUS3P3 is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. This signal can also be configured to GPIO Port 11. Signal Name I/O Type GPIO51 Internal/ External Resistor Pull-Up/ Down Weak Internal pull-up for strap. External pull-up required Description General Purpose I/O Port 8. Not Multiplexed. This signal has a weak internal pull-up and must not be pulled low at boot up. MST_SMBALERT#/ GPIO112 I/O MST_SMBALERT# 1 I SUS GPIO12 I/O Native 1 O (Low) SUS GPIO134 I/O GPI 1 I SUS GPIO14 I/O Native 1 I SUS General Purpose I/O Port 14. Not Multiplexed.3 GPIO15 I/O GPO 1 O (Low) SUS General Purpose I/O Port 15. Not Multiplexed. General Purpose I/O Port 12. Not Multiplexed.3 Weak internal pull-down General Purpose I/O Port 13. Not Multiplexed. Serial ATA 4 General Purpose. This is an input pin which can be configured as an interlock switch corresponding to SATA Port 4. When used as an interlock switch status indication, this signal should be drive to `0' to indicate that the switch is closed and to `1' to indicate that the switch is open. If interlock switches are not required, this pin can be configured as GPIO Port 16. SATA4_GP/GPIO16 I/O GPI 1 I CORE GPIO17 I/O GPI 1 I CORE GPIO18 I/O Native 1 I CORE General Purpose I/O Port 18. Not Multiplexed.3 GPIO19 I/O GPI 1 I CORE General Purpose I/O Port 19. Not Multiplexed. GPIO20 I/O Native 1 I CORE General Purpose I/O Port 20. Not Multiplexed.3 October 2012 Order Number: 327879-001US Platform Dependent General Purpose I/O Port 17. Not Multiplexed. This signal strapping sets the DMI termination voltage. See PDG for additional information. Intel(R) Communications Chipset 89xx Series - Datasheet 1585 Table 32-22. General Purpose I/O Interface Signals (Sheet 3 of 8) Signal Name I/O Type GPIO21 I/O SCLOCK/GPIO22 I/O Default Mode Ball Count Default Direction and Logic State Power Well GPI 1 I CORE General Purpose I/O Port 21. Not Multiplexed. CORE SGPIO Reference Clock. The SATA controller uses rising edges of this clock to transmit serial data, and the target uses the falling edge of this clock to latch data. If SCLOCK interface is not used, this signal can be used as a GPIO Port 22. LPC Serial DMA/Master Request Input Bit 1. Used by LPC devices, such as Super I/O chips, to request DMA or bus master access. This signal is typically connected to external Super I/O device. An internal pull-up resistor is provided on these signals. If LDRQ1# interface is not used, this signal can be used as a GPIO Port 23. GPI 1 I Internal/ External Resistor Pull-Up/ Down Description LDRQ1#/GPIO23 I/O LDRQ1# 1 I CORE GPIO24 I/O GPO 1 O (Low) SUS General Purpose I/O Port 24. Not Multiplexed. GPIO25 I/O Native 1 I SUS General Purpose I/O Port 25. Not Multiplexed.3 GPIO26 I/O Native 1 I SUS General Purpose I/O Port 26. Not Multiplexed.3 GPIO27 I/O GPO 1 O (Low) SUS General Purpose I/O Port 27. Not Multiplexed. GPIO28 I/O GPI 1 I SUS General Purpose I/O Port 28. Not Multiplexed. GPIO30 I/O GPI 1 I SUS GPIO315 I/O GPI 1 I SUS General Purpose I/O Port 31. Not Multiplexed. GPIO324 I/O Native 1 O (High) CORE General Purpose I/O Port 32. Not Multiplexed.3 CORE General Purpose I/O Port 33. Not Multiplexed. Flash Descriptor Security Overwrite. This signal is used to set the security override strap on the PCH. If sampled low, the Flash Descriptor Security will be overridden. If high, the security measures defined in the Flash Descriptor will be in effect. This strap should only be enabled (pulled low) in manufacturing environments using an external pull-down resistor. GPIO33 0 Enable (Pull-Down Required) 1 Disable (Default) When the Security Overwrite is enable, it allows permission to every master to read and write to the entire Flash Components including areas outside the defined regions. GPIO334 I/O GPO 1 Intel(R) Communications Chipset 89xx Series - Datasheet 1586 O (High) Internal pull-down Weak Internal pull-up General Purpose I/O Port 30. Not Multiplexed. October 2012 Order Number: 327879-001US 32.0 Table 32-22. General Purpose I/O Interface Signals (Sheet 4 of 8) Default Mode Ball Count Default Direction and Logic State Power Well I/O GPI 1 I CORE General Purpose I/O Port 34. Not Multiplexed. GPIO35 I/O GPO 1 O (Low) CORE General Purpose I/O Port 35. Not Multiplexed. GPIO36 I/O GPI 1 I CORE General Purpose I/O Port 36. Not Multiplexed. CORE General Purpose I/O Port 37. Can be used for ADR (Asynchronous DRAM Refresh) trigger on platform. Only supported if processor supports ADR. HARDWARE activation mechanism that triggers memory controller of CPU to put SDRAM into self-refresh mode. Activation of ADR flushes contents of some write data buffers to the DIMM before self refresh entry. CORE SATA Serial GPIO Load. The controller drives a `1' at the rising edge of SCLOCK to indicate either the start or end of a bit stream. A 4bit vendor specific pattern will be transmitted right after the signal assertion. If SLOAD interface is not used, this signal can be used as a GPIO Port 38. CORE SATA Serial GPIO Data Out 0. Driven by the controller to indicate the drive status in the following sequence: drive 4, 5...4,5... If SDATAOUT0 interface is not used, the signals can be used as GPIO Port 39. SUS Overcurrent Indicators. These signals set corresponding bits in the USB controllers to indicate that an overcurrent condition has occurred. OC[3:0]# may optionally be used as GPIO Ports [42,41,40,59]. Notes: 1. OC# pins are 3.3V and NOT 5 V tolerant. 2. OC# pins must be shared between ports 3. OC#[3:0] can only be used for EHCI controller #1 SUS Overcurrent Indicators. These signals set corresponding bits in the USB controllers to indicate that an overcurrent condition has occurred. OC[3:0]# may optionally be used as GPIO Ports [42,41,40,59]. Notes: 1. OC# pins are 3.3V and NOT 5 V tolerant. 2. OC# pins must be shared between ports 3. OC#[3:0] can only be used for EHCI controller #1 Signal Name I/O Type GPIO34 ADR/GPIO37 SLOAD/GPIO38 SDATAOUT0/ GPIO39 OC1#/GPIO402 OC2#/GPIO412 I/O I/O I/O I/O I/O October 2012 Order Number: 327879-001US GPI GPI GPI OC1# OC2# 1 1 1 1 1 I I I I I Internal/ External Resistor Pull-Up/ Down Description Intel(R) Communications Chipset 89xx Series - Datasheet 1587 Table 32-22. General Purpose I/O Interface Signals (Sheet 5 of 8) Signal Name I/O Type Default Mode Ball Count Default Direction and Logic State Power Well Internal/ External Resistor Pull-Up/ Down Description OC3#/GPIO422 I/O OC3# 1 I SUS Overcurrent Indicators. These signals set corresponding bits in the USB controllers to indicate that an overcurrent condition has occurred. OC[3:0]# may optionally be used as GPIO Ports [42,41,40,59]. Notes: 1. OC# pins are 3.3V and NOT 5 V tolerant. 2. OC# pins must be shared between ports 3. OC#[3:0] can only be used for EHCI controller #1 GPIO432 I/O Native 1 I SUS General Purpose I/O Port 43. Not Multiplexed.3 GPIO44 I/O Native 1 I SUS General Purpose I/O Port 44. Not Multiplexed.3 GPIO45 I/O Native 1 I SUS General Purpose I/O Port 45. Not Multiplexed.3 GPIO46 I/O Native 1 I SUS General Purpose I/O Port 46. Not Multiplexed.3 GPIO47 I/O Native 1 I SUS General Purpose I/O Port 47. Not Multiplexed.3 SDATAOUT1/ GPIO48 I/O GPI 1 I CORE SATA Serial GPIO Data Out 1. Driven by the controller to indicate the drive status in the following sequence: drive 4, 5...4,5... If SDATAOUT1 interface is not used, the signals can be used as GPIO Port 48. SATA5_GP/ TEMP_ALERT#/ GPIO49 I/O GPI 1 I CORE Serial ATA 5 General Purpose. This is an input pin which can be configured as an interlock switch corresponding to SATA Port 5. When used as an interlock switch status indication, this signal should be drive to `0' to indicate that the switch is closed and to `1' to indicate that the switch is open. Temperature Alert. Used as an alert (active low) to indicate to the external controller (such as, EC or SIO) that temperatures are out of range for the PCH or Memory Controller or the processor core. If interlock switches or Temp Alert are not required, this pin can be configured as GPIO Port 49. GPIO502 I/O Native 1 I CORE General Purpose I/O Port 50. Not Multiplexed.3 Intel(R) Communications Chipset 89xx Series - Datasheet 1588 October 2012 Order Number: 327879-001US 32.0 Table 32-22. General Purpose I/O Interface Signals (Sheet 6 of 8) Signal Name I/O Type Default Mode Ball Count Default Direction and Logic State Power Well Internal/ External Resistor Pull-Up/ Down Description Weak Internal pull-up BIOS Boot Strap 1. BBS1 BBS0 0 0 NOT VALID 0 1 NOT VALID 1 0 NOT VALID 1 1 SPI BOOT If BBS1 interface is not used, the signals can be used as GPIO Port 51. BBS1/GPIO51 I/O BBS1 1 O (High) CORE GPIO522 I/O Native 1 I CORE General Purpose I/O Port 52. Not Multiplexed.3 General Purpose I/O Port 53. Not Multiplexed.3 External platform dependent. DMI Coupling Strap. GPIO53 0 AC Coupling (Pull-Down Required) 1 DC Coupling (Default) See PDG for additional information. GPIO53 I/O Native 1 O (High) CORE GPIO542 I/O Native 1 I CORE Platform Dependent. Weak internal pull-up General Purpose I/O Port 54. Not Multiplexed.3 General Purpose I/O Port 55. Not Multiplexed.3 Strap for BIOS Boot-Block Update Scheme. This mode allows the PCH to swap the Top-Block in the SPI (the boot block) with another location. GPIO55 0 Enable Top-Block Swap (Pull-down Required) 1 Disable Top-Block Swap (Default) Weak Internal pull-up Note: The internal pull-up is disabled after PLTRST# deasserts. If the signal is sampled low, this indicates that the system is strapped to the "Top-Block Swap" mode (the PCH inverts A16 for all cycles targeting BIOS space). The status of this strap is readable via the Top Swap bit (Chipset Config Registers:Offset 3414h:bit 0). Software will not be able to clear the Top-Swap bit until the system is rebooted without GPIO55 being pulled down. GPIO55 I/O Native 1 O (High) CORE GPIO56 I/O Native 1 I SUS General Purpose I/O Port 56. Not Multiplexed.3 GPIO57 I/O GPI 1 I SUS General Purpose I/O Port 57. Not Multiplexed. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1589 Table 32-22. General Purpose I/O Interface Signals (Sheet 7 of 8) Signal Name SML1CLK/GPIO58 I/O Type I/O Default Mode SML1CLK Ball Count 1 Default Direction and Logic State I Power Well SUS Internal/ External Resistor Pull-Up/ Down Description When SMBUS: External pull-up required6 System Management Link 1 Clock. SMBus link to external BMC. External pull-up resistor to VCCSUS3P3 is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. If SML1CLK interface is not used, the signals can be used as GPIO Port 58. OC0#/GPIO592 I/O OC0# 1 I SUS Overcurrent Indicators. These signals set corresponding bits in the USB controllers to indicate that an overcurrent condition has occurred. OC[3:0]# may optionally be used as GPIO Ports [42,41,40,59]. Notes: 1. OC# pins are 3.3V and NOT 5 V tolerant. 2. OC# pins must be shared between ports 3. OC#[3:0] can only be used for EHCI controller #1 GPIO60 I/O Native 1 I SUS General Purpose I/O Port 60. Not Multiplexed.3 SUS Suspend Status: This signal is asserted by the PCH to indicate that the system will be entering a low power state soon. This can be monitored by devices with memory that need to switch from normal refresh to suspend refresh mode. It can also be used by other peripherals as an indication that they should isolate their outputs that may be going to powered-off planes. If SUS_STAT# interface is not used, this signal can be used as a GPIO Port 61. SUS Suspend Clock. This clock is an output of the RTC generator circuit. It is used by other chips for refresh clock. If SUS_CLK interface is not used, the signals can be used as GPIO Port 62. SUS_STAT#/ GPIO61 SUS_CLK/GPIO62 I/O I/O SUS_STAT# SUS_CLK 1 1 O (High) O (Low) SLP_S5#/GPIO63 I/O SLP_S5# 1 O (High) SUS S5 Sleep Control. SLP_S5# is for power plane control. This signal is used to shut power off to all noncritical systems when in the S5 (Soft Off) states. If SLP_S5# interface is not used, the signals can be used as GPIO Port 63. GPIO724 I/O Native 1 I SUS General Purpose I/O Port 72. Not Multiplexed.3 GPIO73 I/O Native 1 I SUS General Purpose I/O Port 73. Not Multiplexed.3 Intel(R) Communications Chipset 89xx Series - Datasheet 1590 October 2012 Order Number: 327879-001US 32.0 Table 32-22. General Purpose I/O Interface Signals (Sheet 8 of 8) I/O Type Signal Name SML1ALERT#/ GPIO742 SML1DAT/GPIO75 I/O 2 I/O TOTAL Default Mode SML1ALERT# SML1DAT Ball Count 1 1 Default Direction and Logic State I I Power Well SUS SUS Internal/ External Resistor Pull-Up/ Down Description When SMBUS: External pull-up required6 System Management Link Alert 1. This signal can be connected to an external BMC. External pull-up resistor to VCCSUS3P3 is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. If SML1ALERT# interface is not used, the signals can be used as GPIO Port 74. When SMBUS: External pull-up required6 System Management Link 1 Data. SMBus link to external BMC. External pull-up required to VCCSUS3P3 is required. Resistor value should be calculated based on the bus load, see the Platform Design Guide. If SML1DAT interface is not used, the signals can be used as GPIO Port 75. 67 1. When this signal is configured as GPO, the output stage is an open drain. 2. When the multiplexed GPIO is used as GPIO functionality, care should be taken to ensure the signal is stable in its inactive state of the native functionality, immediately after reset until it is initialized to GPIO functionality. Multiplexed signals is visible or Intel reserved. 3. For GPIOs where GPIO versus Native Mode is configured using SPI Soft Strap, the corresponding GPIO_USE_SEL bits for these GPIOs have no effect. The GPIO_USE_SEL bits for these GPIOs may change to reflect the Soft-Strap configuration even though GPIO Lockdown Enable (GLE) bit is set. For signals that have Native defaults, these signals must be configured by using SPI Soft Strap. 4. The functionality that is multiplexed with the GPIO may not be used in desktop configuration. 5. In a ME disabled system, GPIO31 may be used as ACPRESENT from the EC. 6. Consult the Platform Design Guide for resistor values. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1591 32.20 Real Time Clock (RTC) Interface Table 32-23. Real Time Clock (RTC) Interface Signals (Sheet 1 of 2) Signal Name Technology Ball Count RTCX1 I Analog 1 Crystal Input 1. This signal is connected to the 32.768 kHz crystal. If no external crystal is used, then RTCX1 can be driven with the desired clock rate. RTCX2 O Analog 1 Crystal Input 2. This signal is connected to the 32.768 kHz crystal. If no external crystal is used, then RTCX2 should be left floating. 1 External RC Circuit required1 RTC Reset. When asserted, this signal resets register bits in the RTC well. Notes: 1. Unless CMOS is being cleared (only to be done in the G3 power state), the RTCRST# input must always be high when all other RTC power planes are on. 2. In the case where the RTC battery is not working or missing on the platform, the RTCRST# pin must rise before the RSMRST# pin. 3. Requires an external RC circuit. see Reference Board Schematic or PDG. External RC Circuit required1 Secondary RTC Reset: This signal resets the manageability register bits in the RTC well when the RTC battery is removed. Notes: 1. The SRTCRST# input must always be high when all other RTC power planes are on. 2. In the case where the RTC battery is not working or missing on the platform, the SRTCRST# pin must rise before the RSMRST# pin. See Reference board schematic or PDG RTCRST# SRTCRST# I I LVTTL LVTTL 1 BScan Support Description INTRUDER# I LVTTL 1 External pull-up 1M Intruder Detect. This signal can be set to disable the system, when the system enclosure is open. This signal's status is readable, so it can be used like a GPI if the Intruder Detection is not needed. This pin requires an external pull-up pin of 1M to VCC3P3_RTC. IVCC_RTC I Analog 1 External Cap to VSS required Internal RTC Voltage Regulator. Generates internally a 1.5V voltage regulator, which requires an external 0.1uF decoupling capacitor. Intel(R) Communications Chipset 89xx Series - Datasheet 1592 Internal/ External Resistor Pull-Up/ Down I/O Type October 2012 Order Number: 327879-001US 32.0 Table 32-23. Real Time Clock (RTC) Interface Signals (Sheet 2 of 2) Signal Name I/O Type PWROK I RSMRST# I TOTAL Technology LVTTL LVTTL Ball Count Internal/ External Resistor Pull-Up/ Down BScan Support Description 1 Power OK. When asserted, PWROK is an indication to the PCH that all of its core power rails have been stable for 10 ms. PWROK can be driven asynchronously. When PWROK is negated, the PCH asserts PLTRST#. Note: It is required that the power rails associated with PCIe typically the 3.3V, 5V, and 12V core well rails) have been valid for 99ms prior to PWROK assertion in order to comply with the 100ms PCIe 1.1 specification on PLTRST# deassertion. PWROK must not glitch, even if RSMRST# is low. 1 Resume Well Reset. This signal is used for resetting the resume power plane logic. This signal must be asserted for at least 10 ms after the suspend power wells are valid. When deasserted, this signal is an indication that the suspend power wells are stable. External pull-down required2 8 1. Consult the Platform Design Guide for resistor and capacitor values. 2. Consult the Platform Design Guide for resistor values. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1593 32.21 System Input Clock Table 32-24. Clock Input Interface Signals Internal/ External Resistor Pull-Up/ Down Signal Name I/O Type Technology Ball Count CRU_CLK100N CRU_CLK100P I LV Diff 2 Clock and Reset Unit 100 MHz Differential input Clock. DMI_CLK100N DMI_CLK100P I LV Diff 2 Direct Media Interface 100 MHz Differential Clock. These signals are used to run the Direct Media Interface. Runs at 100 MHz. GbE 100 MHZ Differential Clock with 50 ppm max jitter. External SerDes/SGMII differential 100MHz reference clock from an external generator. This clock must be running from the auxiliary power. GBE_AUX_PWR_OK is asserted only when this clock is stable. GBE_CLK100N GBE_CLK100P I LV Diff 2 PCICLK I LVTTL 1 BScan Support 1149.1 Description LPC clock. PCI clock used for the LPC bus (up to 33 MHz). PCIE_EP_CLK100N PCIE_EP_CLK100P I LV Diff 2 PCI Express End Point Clock (Differential): A 100 MHz differential clock input. This clock is used as the reference clock for the PCIe Tx/Rx circuitry and by the PCIe core PLL to generate clocks for the PCIe core logic. REF_CLK14 I LVTTL 1 Reference Clock 14.31818 MHz Input. Single-ended 14.31818 MHz reference clock driven by a clock chip. Serial ATA 100 MHz Differential Clock. Reference clock 100 MHz Differential signal from a clock chip, provided separately from CLKIN_DMI, for use only as a 100 MHz source for SATA. SATA_CLK100N SATA_CLK100P I LV Diff 2 UART_CLK I LVTTL 1 USB_CLK96N USB_CLK96P I LV Diff 2 TOTAL UART Clock: This clock is passed to the baud clock generation logic for the UART. The clock can run at either 14.7456, or 48 MHz. USB Clock 96 MHz. Differential reference input clock from an external clock chip. 15 Intel(R) Communications Chipset 89xx Series - Datasheet 1594 1149.1 October 2012 Order Number: 327879-001US 32.0 32.22 JTAG Interface A hardware implementation is required if the customer wants to enable Boundary Scan for testing during manufacturing. Details on enabling JTAG are published in the PCH BSDL package. Contact your Intel Representative to request this package. This only applies to the JTAG chain in Table 32-25, "PCH JTAG Interface Signals". Table 32-25. PCH JTAG Interface Signals Signal Name I/O Type Technology Ball Count Internal/ External Resistor Pull-Up/ Down External pull-down 51 1% to VSS required JTCK I LVCMOS 1 JTDI I LVCMOS 1 JTDO OD JTMS I LVCMOS LVCMOS TOTAL BScan Support Description JTAG Test Clock for the JTAG controller. JTAG Test Data In. Sampled with the rising edge of JTCK. 1 External pull-up 51 1% to VCCSUS required JTAG Test Data Out. 1 External pull-up 51 1% to VCCSUS required JTAG Test Mode Select. Selects the state of the JTAG controller. Sampled with the rising edge of JTCK. 4 Table 32-26. EndPoint JTAG Interface Signals (Sheet 1 of 2) Signal Name I/O Type Technology Ball Count Internal/ External Resistor Pull-Up/ Down EP_JTCK I LVCMOS 1 External pull-down 51 1% to VSS required EP_JTDI I LVCMOS 1 Internal pull-up 1 External pull-up 51 1% to VCCEPAUX required EP_JTDO OD October 2012 Order Number: 327879-001US LVCMOS BScan Support Description EndPoint JTAG Test Clock for the JTAG controller. EndPoint JTAG Test Data In. Sampled with the rising edge of EP_JTCK. EndPoint JTAG Test Data Out. Intel(R) Communications Chipset 89xx Series - Datasheet 1595 Table 32-26. EndPoint JTAG Interface Signals (Sheet 2 of 2) Signal Name I/O Type EP_JTMS I EP_JTRST# I TOTAL Technology LVCMOS LVCMOS Ball Count Internal/ External Resistor Pull-Up/ Down 1 External pull-up 51 1% to VCCEPAUX required EndPoint JTAG Test Mode Select. Selects the state of the JTAG controller. Sampled with the rising edge of EP_JTCK. 1 Internal pull-up See Description EndPoint JTAG Reset. Resets JTAG controller when asserted. The signal has an internal pull-up resistor to comply with 1149.1. An external 51 1% pull-down resistor is required to disable JTAG and keep TAP in safe mode. Description 5 Intel(R) Communications Chipset 89xx Series - Datasheet 1596 BScan Support October 2012 Order Number: 327879-001US 32.0 32.23 Strapping Signals Table 32-27. Strap Signals (Sheet 1 of 4) Signal Name GBE_AUX_PWR_AVAIL GPIO8 GPIO17 GPIO33 I/O Type I I/O I/O I/O October 2012 Order Number: 327879-001US Technology LVTTL LVTTL LVTTL LVTTL Ball Count Internal/ External Resistor Pull-Up/ Down Strap Qualifier 1 External pull-up required1 Rise Edge PCIE_EP_RST# 1 Weak Internal pullup for strap. Rise Edge RSMRST# 1 1 Platform Dependent Weak Internal pullup Rise Edge PWROK Rise Edge PWROK when PLTRST# Asserted BScan Support Description 1149.1 Auxiliary Power Available. This pin is a strapping option pin, latched at the rising edge of PCIE_EP_RST# or In-Band PCIe Reset. If this pin is driven high during init time, it indicates that auxiliary power is available and the device should support D3COLD power state if enabled to do so. This pin value should be established before GBE_AUX_POWER_OK goes high. 1149.1 General Purpose I/O Port 8. Not Multiplexed. This signal has a weak internal pull-up and must not be pulled low at boot up. 1149.1 General Purpose I/O Port 17. Not Multiplexed. This signal strapping sets the DMI termination voltage. See PDG for additional information. 1149.1 General Purpose I/O Port 33. Not Multiplexed. Flash Descriptor Security Overwrite. This signal is used to set the security override strap on the PCH. If sampled low, the Flash Descriptor Security will be overridden. If high, the security measures defined in the Flash Descriptor will be in effect. This strap should only be enabled (pulled low) in manufacturing environments using an external pull-down resistor. GPIO33 0 Enable (Pull-Down Required) 1 Disable (Default) When the Security Overwrite is enable, it allows permission to every master to read and write to the entire Flash Components including areas outside the defined regions. Intel(R) Communications Chipset 89xx Series - Datasheet 1597 Table 32-27. Strap Signals (Sheet 2 of 4) Signal Name GPIO53 GPIO55 I/O Type I/O I/O Technology LVTTL LVTTL Ball Count 1 1 Intel(R) Communications Chipset 89xx Series - Datasheet 1598 Internal/ External Resistor Pull-Up/ Down Platform Dependent. Weak internal pullup Weak Internal pullup Strap Qualifier Rise Edge PWROK when PLTRST# Asserted Rise Edge PWROK when PLTRST# Asserted BScan Support Description 1149.1 General Purpose I/O Port 53. Not Multiplexed. External platform dependent. DMI Coupling Strap. GPIO53 0 AC Coupling (Pull-Down Required) 1 DC Coupling (Default) See PDG for additional information. 1149.1 General Purpose I/O Port 55. Not Multiplexed.3 Strap for BIOS Boot-Block Update Scheme. This mode allows the PCH to swap the Top-Block in the SPI (the boot block) with another location. GPIO55 0 Enable Top-Block Swap (Pull-down Required) 1 Disable Top-Block Swap (Default) Note: The internal pullup is disabled after PLTRST# deasserts. If the signal is sampled low, this indicates that the system is strapped to the "Top-Block Swap" mode (the PCH inverts A16 for all cycles targeting BIOS space). The status of this strap is readable via the Top Swap bit (Chipset Config Registers:Offset 3414h:bit 0). Software will not be able to clear the Top-Swap bit until the system is rebooted without GPIO55 being pulled down. October 2012 Order Number: 327879-001US 32.0 Table 32-27. Strap Signals (Sheet 3 of 4) Signal Name SIU0_DTR# BBS1 / GPIO51 GBE0_LED GBE1_LED I/O Type O I/O O O October 2012 Order Number: 327879-001US Technology LVTTL LVTTL LVTTL LVTTL Ball Count 1 1 1 1 Internal/ External Resistor Pull-Up/ Down Weak Internal pullup Weak Internal pullup Internal Pull-up active for 200us after assertion of AUX_PWROK Internal Pull-up active for 200us after assertion of AUX_PWROK Strap Qualifier Rise Edge PWROK when PLTRST# Asserted Rise Edge PWROK when PLTRST# Asserted After assertion of AUX_PWROK After assertion of AUX_PWROK BScan Support Description 1149.1 UART Port 0 Data Terminal Ready: When low, this pin informs the modem or data set that the UART port 0 is ready to establish a communication link. SIW Configuration Port Address Select Strap: this strap selects the IO address for the SIW configuration port. Sampling occurs on the rising edge of PWROK. The straps are defined as follows: 0 = IO Addresses 02Eh and 02Fh (Pull-Down Required) 1 = IO Addresses 04Eh and 04Fh (default) 1149.1 BIOS Boot Strap 1. BBS1 BBS0 0 0 NOT VALID 0 1 NOT VALID 1 0 NOT VALID 1 1 SPI BOOT If BBS1 interface is not used, the signals can be used as GPIO Port 51. 1149.1 Port 0 LED. Programmable LED, mode encoding set by GbE EEPROM. On power-up this signal becomes an input to Strap SMBus Slave Address bit 2 The EP SMBus slave is accessed with address[7:1] = 1110_XX0. Sampling occurs during the first 200s after assertion of AUX_PWROK. Thereafter the pin functions as an LED output. 1149.1 Port 1 LED. Programmable LED, mode encoding set by GbE EEPROM. On power-up this signal becomes an input to Strap SMBus Slave Address bit 3. The EP SMBus slave is accessed with address[7:1] = 1110_XX0. Sampling occurs during the first 200s after assertion of AUX_PWROK. Thereafter the pin functions as an LED output. Intel(R) Communications Chipset 89xx Series - Datasheet 1599 Table 32-27. Strap Signals (Sheet 4 of 4) Signal Name I/O Type Technology Ball Count Internal/ External Resistor Pull-Up/ Down Strap Qualifier BScan Support Description Port 2 LED. Programmable LED, mode encoding set by GbE EEPROM. GBE2_LED O GBE3_LED O NRBOOTS O BBS0 O TOTAL LVTTL LVTTL LVTTL LVTTL 1 Internal Pull-up active for 200us after assertion of AUX_PWROK After assertion of AUX_PWROK 1 Internal Pull-up Active for 200us after assertion of AUX_PWROK After assertion of AUX_PWROK 1 Weak Internal pulldown Rise Edge PWROK when PLTRST# Asserted 1 Weak Internal pullup Rise Edge PWROK when PLTRST# Asserted 1149.1 This signal can be use as a strapping function to enable or disable PCIe* SRIOV GBE2_LED: 0 Disable SRIOV (PullDown Required) 1 Enable SRIOV (default) Sampling occurs during the first 200ns after assertion of AUX_PWROK. Thereafter the pin functions as an LED output. 1149.1 Port 3 LED. Programmable LED, mode encoding set by GbE EEPROM. This signal can be use as a strapping function to control reset sequence. 0 A0 Compatible Reset Sequence (Pull-Down Required) 1 PCIe* CEM 2.0 Compliant Reset sequence (default) Sampling occurs during the first 200ns after assertion of AUX_PWROK. Thereafter the pin functions as a LED output. 1149.1 No Re-Boot Second Try Strap. See PDG for additional information. 1149.1 BIOS Boot Strap 0. BBS1 BBS0 0 0 NOT VALID 0 1 NOT VALID 1 0 NOT VALID 1 1 SPI BOOT 14 1. Consult the Platform Design Guide for resistor values. Intel(R) Communications Chipset 89xx Series - Datasheet 1600 October 2012 Order Number: 327879-001US 32.0 32.24 Reserved Signals Table 32-28. Reserved Signals (Sheet 1 of 3) Internal/External Resistor Pull-Up/ Down Signal Name Ball Count RSVD0 1 1149.1 Reserved Signal 0. RSVD1 1 1149.1 Reserved Signal 1. RSVD2 1 1149.1 Reserved Signal 2. RSVD3 1 1149.1 Reserved Signal 3. RSVD4 1 1149.1 Reserved Signal 4. RSVD5 1 1149.1 Reserved Signal 5. RSVD6 1 1149.1 Reserved Signal 6. RSVD7 1 1149.1 Reserved Signal 7. RSVD8 1 1149.1 Reserved Signal 8. BScan Support Description RSVD9 1 1149.1 Reserved Signal 9. RSVD10 1 1149.1 Reserved Signal 10. RSVD11 1 1149.1 Reserved Signal 11. RSVD12 1 1149.1 Reserved Signal 12. RSVD13 1 1149.1 Reserved Signal 13. RSVD14 1 1149.1 Reserved Signal 14. RSVD15 1 1149.1 Reserved Signal 15. RSVD16 1 1149.1 Reserved Signal 16. RSVD17 1 1149.1 Reserved Signal 17. RSVD18 1 1149.1 Reserved Signal 18. RSVD19 1 1149.1 Reserved Signal 19. RSVD20 1 1149.1 Reserved Signal 20. RSVD21 1 1149.1 Reserved Signal 21. RSVD22 1 1149.1 Reserved Signal 22. RSVD23 1 1149.1 Reserved Signal 23. RSVD24 1 1149.1 Reserved Signal 24. RSVD25 1 1149.1 Reserved Signal 25. RSVD26 1 1149.1 Reserved Signal 26. RSVD27 1 1149.1 Reserved Signal 27. RSVD28 1 1149.1 Reserved Signal 28. RSVD29 1 1149.1 Reserved Signal 29. RSVD30 1 1149.1 Reserved Signal 30. RSVD31 1 1149.1 Reserved Signal 31. RSVD32 1 1149.1 Reserved Signal 32. RSVD33 1 1149.1 Reserved Signal 33. RSVD34 1 1149.1 Reserved Signal 34. RSVD35 1 1149.1 Reserved Signal 35. RSVD36 1 October 2012 Order Number: 327879-001US Reserved Signal 36 Intel(R) Communications Chipset 89xx Series - Datasheet 1601 Table 32-28. Reserved Signals (Sheet 2 of 3) Signal Name Ball Count RSVD37 1 Internal/External Resistor Pull-Up/ Down Description Reserved Signal 37 RSVD38 1 External pull-up required1 RSVD39 1 External pull-down required 1K to VSS RSVD41 1 Reserved Signal 38. This signal needs a pull-up to VCCSUS. Reserved Signal 39. Reserved Signal 41. RSVD42 1 Reserved Signal 42. RSVD43 1 Reserved Signal 43. RSVD44 1 Reserved Signal 44. RSVD45 1 Reserved Signal 45. RSVD46 1 RSVD47 1 1149.1 Reserved Signal 53. RSVD48 1 1149.1 Reserved Signal 48. RSVD49 1 1149.1 Reserved Signal 49. RSVD50 1 1149.1 Reserved Signal 50. RSVD51 1 RSVD52 1 RSVD53 1 Reserved Signal 46. Reserved Signal 51. Reserved Signal 52. Weak Internal pulldown 1149.1 Reserved Signal 53. RSVD54 1 Reserved Signal 54. RSVD55 1 Reserved Signal 55. RSVD56 1 Reserved Signal 56. RSVD57 1 Reserved Signal 57. RSVD59 1 Reserved Signal 59. RSVD60 1 Reserved Signal 60. RSVD61 1 Reserved Signal 61. RSVD62 1 Reserved Signal 62. RSVD63 1 Reserved Signal 63. RSVD64 1 Reserved Signal 64. RSVD65 1 Reserved Signal 65. RSVD66 1 Reserved Signal 66. RSVD67 1 Reserved Signal 67. RSVD68 1 Reserved Signal 68. RSVD69 1 Reserved Signal 69. RSVD70 1 Reserved Signal 70. RSVD71 1 External 100pF cap to ground. Reserved Signal 71. Connect to 100pF capacitor tied to ground. RSVD72 1 External Pull-down required <1K to VSS Reserved Signal 72. This pin must be tied to ground or pull to ground through a resistor. RSVD73 1 1149.1 Reserved Signal 73. RSVD74 1 1149.1 Reserved Signal 74. Intel(R) Communications Chipset 89xx Series - Datasheet 1602 BScan Support October 2012 Order Number: 327879-001US 32.0 Table 32-28. Reserved Signals (Sheet 3 of 3) Internal/External Resistor Pull-Up/ Down Signal Name Ball Count RSVD75 1 1149.1 Reserved Signal 75. RSVD76 1 1149.1 Reserved Signal 76. BScan Support Description RSVD77 1 RSVD78 1 Reserved Signal 77. RSVD79 1 Reserved Signal 79 RSVD80 1 Reserved Signal 80 RSVD81 1 Reserved Signal 81 RSVD82 1 Reserved Signal 82 RSVD83 1 Reserved Signal 83 RSVD84 1 Reserved Signal 84 RSVD85 1 Reserved Signal 85 RSVD86 1 Reserved Signal 86 RSVD87 1 Reserved Signal 87 RSVD88 1 Reserved Signal 88 RSVD89 1 Reserved Signal 89 RSVD90 1 Reserved Signal 90 TOTAL 89 1149.1 Reserved Signal 78. 1. Consult the Platform Design Guide for resistor values. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1603 32.25 Power and Ground Signals Table 32-29. Power and Ground Signals (Sheet 1 of 2) Signal Name Ball Count Supply VCC 8 1.0 V VCC1P05 4 1.05 V Power supply for internal reference. This voltage can not be changed from the recommended value. VCC1P8 12 1.8 V Power supply for core well I/O buffers. This power will be off in S3/S4/S5 or G3 VCC3P3 27 3.3 V Power supply for core well I/O buffers. This power will be off in S3/S4/S5 or G3 2 3.3 V RTC power supply. Can drop to 2.0V if all other planes shut down. This power is not expected to be shut down unless the RTC battery is removed or drained. A suspend well regulator drives in Sx states. Note: Implementations should not attempt to clear CMOS by using a jumper to pull VCC3P3_RTC low. Clearing CMOS in a PCH-platform can be done by using a jumper on RTCRST#. VCCA_DMI 21 1.0 V 1.05 V Analog Power supply for DMI. This voltage can be 1.0V or 1.05V based on the processor used. See the respective processor documentation to find the appropriate voltage level. This power will be shut down in S3, S4, and S5 states. This requires a LC filter. VCCA_SATA 9 1.0 V Analog Power supply for SATA I/O interface. VCC3P3_RTC Description Power supply for core logic. Will be off in S3/S4/S5 VCCA_USB 9 1.0 V Analog Power supply for USB level shifter. VCCA1P5_TS0 2 1.5 V Analog Power supply for Thermal Sensor 0. Analog Supply. VCCA1P5_TS1 2 1.5 V Analog Power supply for Thermal Sensor 1. Analog Supply. VCCA1P8_DMI 2 1.8 V Analog Power supply for DMI PLL and PHY Compensation Logic. No LC needed. This power is supplied with core well. VCCA1P8_SATA 2 1.8 V Analog Power supply for the SATA PLL and PHY Compensation Logic. No LC needed. VCCA1P8_TS0 2 1.8 V Analog Power supply for Thermal Sensor 0. Analog Supply. VCCA1P8_TS1 2 1.8 V Analog Power supply for Thermal Sensor 1. Analog Supply. VCCA1P8_USB 3 1.8 V Analog Power supply for USB PLL and PHY Compensation Logic. VCCAEP_PE 36 1.0 V Analog Power supply EndPoint for PCI Express I/O interface. Analog power. VCCAEP1P8_CRU 3 1.8 V Analog Power supply EndPoint Unit for CRU circuitry (CRU PLL and REFCLK inputs). VCCAEP1P8_PE 2 1.8 V Analog Power supply EndPoint for PCI Express PLL and PHY Compensation Logic. VCCAEP1P8AUX 6 1.8 V Analog Auxiliary Power supply EndPoint Unit for I/O Logic, Gigabit Ethernet PLL and PHY Compensation Logic. This power is not expected to be shut off unless the system is unplugged. VCCAEPAUX 21 1.05 V VCCASUS1P8_USB 1 1.8 V VCCEP 8 1.0 V Power supply EndPoint Unit for core logic. Will be off in S3/S4/S5 Analog Auxiliary Power supply EndPoint Unit. Supply voltage for Gigabit Ethernet Core, and I/O interface EndPoint Unit. Analog Sustain Power supply for USB band gap. VCCEP1P0_CRU 4 1.0 V Digital Power supply EndPoint Unit for CRU circuitry (CRU PLL and REFCLK inputs). VCCEPAUX 7 1.0 V Auxiliary Power supply EndPoint Unit for core logic. Intel(R) Communications Chipset 89xx Series - Datasheet 1604 October 2012 Order Number: 327879-001US 32.0 Table 32-29. Power and Ground Signals (Sheet 2 of 2) Signal Name Ball Count Supply Description VCCPCPU 1 1.0 V Power supply for CPU I/O Signals: Powered by the same supply as the processor I/O voltage. This supply is used to drive the processor interface signals. Note: See the respective processor documentation for the appropriate voltage level. VCCPEP3P3AUX 8 3.3 V Auxiliary Power supply EndPoint Unit for I/O Logic. This power is not expected to be shut off unless the system is unplugged. VCCSUS 1 1.0 V Sustain power supply for core logic. Will be off in G3 only VCCSUS_USB 9 1.0 V Sustain power supply for USB2 High Speed TX/RX interface and control Logic. VCCSUS1P8 6 1.8 V Sustain power supply for I/O logic. This power is not expected to be shut off unless the system is in G3 or unplugged. VCCSUS1P8_USB 7 1.8 V Sustain Power supply for USB TX/RX interface and control Logic. VCCSUS3P3 10 3.3 V Sustain power supply for I/O logic. This power is not expected to be shut off unless the system is in G3 or unplugged. VCCSUS3P3_RTC 2 3.3 V Sustain power supply for RTC Block: Suspend well power for part of the RTC logic. Can be connected to VCCSUS3P3 on the board. VCCSUS3P3_USB 4 3.3 V Sustain Power supply for USB classic TX output interface. VSS 48 GND Ground VSSA_USB 1 GND Quiet Ground for USB TOTAL 292 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1605 33.0 Electrical Characteristics This chapter contains DC and AC characteristics. AC timing diagrams are included. Note: Icc values in this section are pre-silicon estimates. 33.1 Absolute Maximum and Minimum Ratings Table 33-1 specifies the absolute maximum and minimum ratings of the PCH. At conditions outside of the functional operating condition limits, but within the absolute maximum and minimum ratings, neither functionality nor long-term reliability can be expected. If a device is returned to conditions within the functional operating limits after having been subjected to conditions outside these limits (but within the absolute maximum and minimum ratings) the device may be functional, but with its lifetime degraded depending on exposure to conditions exceeding the functional operating condition limits. At conditions exceeding the absolute maximum and minimum ratings, neither functionality nor long-term reliability can be expected. Moreover, if a device is subjected to these conditions for any length of time, it will either not function or its reliability will be severely degraded when returned to conditions within the functional operating condition limits. Although the PCH contains protective circuitry to resist damage from Electro-Static Discharge (ESD), precautions should always be taken to avoid high static voltages or electric fields Table 33-1. PCH Absolute Maximum and Minimum Ratings Symbol Tj max Tstorage Parameter Junction Temperature Storage Temperature Range Maximum Limits Units 0 103 C -65 140 C Supply Voltage (3.3V Nominal) Vss-0.5 3.70 V Supply Voltage (1.8V Nominal) Vss-0.3 1.98 V Supply Voltage (1.5V Nominal) Vss-0.2 1.65 V Supply Voltage (1.05V Nominal) Vss-0.2 1.30 V Supply Voltage (1.0V Nominal) Vss-0.2 1.30 V Intel(R) Communications Chipset 89xx Series - Datasheet 1606 Minimum Limits October 2012 Order Number: 327879-001US 33.0 33.2 Recommended Power Supply Range Table 33-2. Operating Conditions Power Supply Rails Symbol Tolerance Min Nominal Max Unit VCC +/-5% 0.95 1.00 1.05 V VCC1P05 +/-5% 1.00 1.05 1.10 V VCC1P8 +/-5% 1.71 1.80 1.89 V VCC3P3 +/-5% 3.13 3.30 3.46 V VCC3P3_RTC +/-5% 3.13 3.30 3.46 V V VCC3P3_RTC (Battery) 2.00 3.00 3.47 0.95 1.00 1.05 1.00 1.05 1.10 1.00 1.05 VCCA_DMI1 +/-5% VCCA_SATA +/-5% 0.95 VCCA_USB +/-5% 0.95 1.00 1.05 V VCCA1P5_TS0 +/-5% 1.42 1.50 1.57 V V V VCCA1P5_TS1 +/-5% 1.42 1.50 1.57 V VCCA1P8_DMI +/-5% 1.71 1.80 1.89 V VCCA1P8_SATA +/-5% 1.71 1.80 1.89 V VCCA1P8_TS0 +/-5% 1.71 1.80 1.89 V VCCA1P8_TS1 +/-5% 1.71 1.80 1.89 V VCCA1P8_USB +/-5% 1.71 1.80 1.89 V VCCAEP_PE +/-5% 0.95 1.00 1.05 V VCCAEP1P8_CRU +/-5% 1.71 1.80 1.89 V VCCAEP1P8_PE +/-5% 1.71 1.80 1.89 V VCCAEP1P8AUX +/-5% 1.71 1.80 1.89 V VCCAEPAUX +/-5% 0.99 1.05 1.10 V VCCASUS1P8_USB +/-5% 1.71 1.80 1.89 V VCCEP +/-5% 0.95 1.00 1.05 V VCCEP1P0_CRU +/-5% 0.95 1.00 1.05 V VCCEPAUX +/-5% 0.95 1.00 1.05 V VCCPCPU +/-5% 0.95 1.00 1.05 V VCCPEP3P3AUX +/-5% 3.13 3.30 3.46 V VCCSUS +/-5% 0.95 1.00 1.05 V VCCSUS_USB +/-5% 0.95 1.00 1.05 V VCCSUS1P8 +/-5% 1.71 1.80 1.89 V VCCSUS1P8_USB +/-5% 1.71 1.80 1.89 V VCCSUS3P3 +/-5% 3.13 3.30 3.46 V VCCSUS3P3_RTC +/-5% 3.13 3.30 3.46 V VCCSUS3P3_USB +/-5% 3.13 3.30 3.46 V VSSA_USB +/-5% GND V 1. Analog Power supply for DMI. This voltage can be 1.0V or 1.05V based on the processor used. See the respective processor documentation for the appropriate voltage level. This power will be shut down in S3, S4, and S5 states. This requires a LC filter. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1607 33.3 Maximum ICC Supply Current Table 33-3. Maximum Supply Current Symbol A0-Stepping ICC (max) B0/C0 Stepping ICC (max)1. 1.00 V VCC 2000 mA 2000 mA 1.05 V VCC1P05 35 mA 35 mA 1.80 V VCC1P8 50 mA 50 mA 3.30 V VCC3P3 150 mA 150 mA 3.30 V VCC3P3_RTC 3 mA 3 mA 3.00 V VCC3P3_RTC (Battery) 6 A 6 A 1.05 V VCCA_DMI 800 mA 800 mA 1.00 V VCCA_SATA 300 mA 300 mA 1.00 V VCCA_USB 120 mA 120 mA 1.50 V VCCA1P5_TS0 10 mA 10 mA 1.50 V VCCA1P5_TS1 10 mA 10 mA 1.80 V VCCA1P8_DMI 55 mA 55 mA 1.80 V VCCA1P8_SATA 55 mA 55 mA 1.80 V VCCA1P8_TS0 30 mA 30 mA 1.80 V VCCA1P8_TS1 30 mA 30 mA 1.80 V VCCA1P8_USB 45 mA 45 mA 1.00 V VCCAEP_PE 1600 mA 1600 mA 1.80 V VCCEP1P8_CRU 90 mA 90 mA Well/ Voltage Rail 1.80 V VCCAEP1P8_PE 55 mA 55 mA 1.80 V VCCAEP1P8AUX 100 mA 100 mA 1.05 V VCCAEPAUX 550 mA 550 mA 1.80 V VCCASUS1P8_USB 15 mA 15 mA 1.00 V VCCEP 12.5 A 10.0 A 1.00 V VCCEP1P0_CRU 100 mA 100 mA 1.00 V VCCEPAUX1 1750 mA 1500 mA 1.00 V VCCPCPU 50 mA 50 mA 3.30 V VCCPEP3P3AUX 150 mA 150 mA 1.00 V VCCSUS 520 mA 520 mA 1.00 V VCCSUS_USB 50 mA 50 mA 1.80 V VCCSUS1P8 50 mA 50 mA 1.80 V VCCSUS1P8_USB 250 mA 250 mA 3.30 V VCCSUS3P3 150 mA 150 mA 3.30 V VCCSUS3P3_RTC 3 mA 3 mA 3.30 V VCCSUS3P3_USB 90 mA 90 mA 1. Pre-silicon estimates and subject to change. Intel(R) Communications Chipset 89xx Series - Datasheet 1608 October 2012 Order Number: 327879-001US 33.0 33.4 Power Supply Pin Groupings The following tables are suggestions of how to group power pins. These groupings help to minimize the number of power regulators required to implement power requirements. Table 33-4, "Power for Standby Support Designs" is for designs that require standby functionality for systems that go into power saving modes. Table 33-5, "Power for Non-Standby Support Designs" is for system designs that are meant to be either ON or OFF and do not go into power saving modes. Table 33-4. Power for Standby Support Designs (Sheet 1 of 2) Board Regulator Grouping1 Domain VCCAEP1P8_PE 1 Core EndPoint VCCAEP1P8_CRU 1 Core EndPoint VCCA1P8_DMI 1 Core Non-EndPoint Customer Name Normal Mode VCCA1P8_SATA 1 Core Non-EndPoint VCCA1P8_USB 1 Core Non-EndPoint VCCA1P8_TS0 1 Core Non-EndPoint VCCA1P8_TS1 1 Core Non-EndPoint VCC1P8 1 Core Non-EndPoint VCCAEP_PE 2 Core EndPoint VCCEP1P0_CRU 2 Core EndPoint VCCA_SATA 2 Core Non-EndPoint VCCA_USB 2 Core Non-EndPoint VCC 2 Core Non-EndPoint VCCEP 2 Core EndPoint VCCAEP1P8AUX 3 Aux EndPoint VCCASUS1P8_USB 3 Sus Non-EndPoint VCCSUS1P8_USB 3 Sus Non-EndPoint VCCSUS1P8 3 Sus Non-EndPoint VCCAEPAUX 4 Aux EndPoint VCCPEP3P3AUX 5 Aux EndPoint VCCSUS3P3_USB 5 Sus Non-EndPoint VCCSUS3P3_RTC 5 Sus Non-EndPoint VCCSUS3P3 5 Sus Non-EndPoint VCCA_DMI 6 Core Non-EndPoint2 VCC1P05 7 Core Non-EndPoint. VCCPCPU 8 Core Non-EndPoint. Connect to CPU IO rail VCCEPAUX 9 Aux EndPoint VCCSUS_USB 9 Sus Non-EndPoint VCCSUS 9 Sus Non-EndPoint VCCA1P5_TS0 10 Core Non-EndPoint VCCA1P5_TS1 10 Core Non-EndPoint October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1609 Table 33-4. Power for Standby Support Designs (Sheet 2 of 2) Customer Name VCC3P3 VCC3P3_RTC VSS Board Regulator Grouping1 Domain Normal Mode 11 Core Non-EndPoint Battery RTC Non-EndPoint G GND EndPoint and Non-EndPoint 1. Board Regulator Grouping refers to the minimum number of regulators that can be used for adequate operation. 2. Voltage level depends on processor use in your design. Table 33-5. Power for Non-Standby Support Designs (Sheet 1 of 2)1 Board Regulator Grouping2 Domain VCCAEP1P8_PE 1 Main EndPoint VCCAEP1P8_CRU 1 Main EndPoint VCCA1P8_DMI 1 Main Non-EndPoint VCCA1P8_SATA 1 Main Non-EndPoint VCCA1P8_USB 1 Main Non-EndPoint VCCA1P8_TS0 1 Main Non-EndPoint VCCA1P8_TS1 1 Main Non-EndPoint VCC1P8 1 Main Non-EndPoint VCCAEP1P8AUX 1 Main EndPoint VCCASUS1P8_USB 1 Main Non-EndPoint VCCSUS1P8_USB 1 Main Non-EndPoint VCCSUS1P8 1 Main Non-EndPoint VCCAEP_PE 2 Main EndPoint Pin Name Normal Mode VCCEP1P0_CRU 2 Main EndPoint VCCA_SATA 2 Main Non-EndPoint VCCA_USB 2 Main Non-EndPoint VCC 2 Main Non-EndPoint VCCEPAUX 2 Main EndPoint VCCSUS_USB 2 Main Non-EndPoint VCCSUS 2 Main Non-EndPoint VCCEP 2 Main EndPoint VCCAEPAUX 3 Main EndPoint VCCA_DMI 4 Main Non-EndPoint 3 VCC1P05 5 Main Non-EndPoint VCCPCPU 6 Main Non-EndPoint. Connect to CPU IO rail VCCPEP3P3AUX 7 Main EndPoint VCCSUS3P3_USB 7 Main Non-EndPoint VCCSUS3P3_RTC 7 Main Non-EndPoint VCCSUS3P3 7 Main Non-EndPoint VCC3P3 7 Main Non-EndPoint VCC3P3_RTC 7 Main Non-EndPoint Intel(R) Communications Chipset 89xx Series - Datasheet 1610 October 2012 Order Number: 327879-001US 33.0 Table 33-5. Power for Non-Standby Support Designs (Sheet 2 of 2)1 Pin Name Board Regulator Grouping2 Domain Normal Mode VCCA1P5_TS0 8 Main Non-EndPoint VCCA1P5_TS1 8 Main Non-EndPoint VSS G GND EndPoint and Non-EndPoint 1. For non-standby support, you can connect Suspend [SUS] and Auxiliary [AUX] supplies together. 2. Board Regulator Grouping refers to the minimum number of regulators that can be used for adequate operation. 3. Voltage level depends on processor use in your design. 33.5 General DC Characteristics Table 33-6. DC Input Characteristic Signal Association (Sheet 1 of 2) Symbol Associated Signals VIH1/VIL1 SMBus Signals: EP_SMBCLK, EP_SMBDAT SMBus GbE Signals: GBE_SMBCLK, GBE_SMBDAT, GBE_SMBALRT# GbE Signals: SFP[3:0]_I2C_DATA, SRDS[3:0]_SD, GBE[3:0]_LED, GBE[3:0]_SWDP[1:0], GBE_AUX_PWR_OK, GBE_EE_DO PCIe* Signals: PCIE_EP_RST Power Management Signals: EP_CRU_EN, EP_MAIN_PWR_OK, GBE_AUX_PWR_AVAIL, GBE_AUX_PWR_OK VIH2/VIL2 Clock Signals: REF_CLK14, PCICLK, UART_CLK Processor Signals: A20GATE, RCIN# Interrupt Signals: SERIRQ Power Management Signals: BMBUSY# UART Signals: SIU[1:0]_RXD, SIU[1:0]_DSR#, SIU[1:0]_DCD#, SIU[1:0]_CTS#, SIU[1:0]_RI# SATA Signals: SATA4_GP, SATA5_GP SPI Signals: SPI_MISO LPC/Firmware Hub Signals: LAD[3:0], LDRQ0#, LDRQ1# GPIO Signals: GPIO[55, 54, 53, 52, 51, 50, 49, 48, 39, 38, 37, 36, 35, 34, 33, 32, 23, 22, 21, 20, 19, 18, 17,16, 7, 6, 5, 4, 3, 2, 1, 0] Strap Signals: SPI_MOSI, BBS[1:0]# VIH3/VIL3 SMBus Signals: MST_SMBCLK, MST_SMBDAT System Management Link Signals: SML1CLK, SML1DAT Power Management Signals: PWRBTN#, RI#, SYS_RESET#, WAKE#, SYS_PWROK, MEPWROK GPIO Signals: GPIO[75, 74, 73, 72, 63, 62, 61, 60, 59, 58, 57, 56, 47, 46, 45, 44, 43, 42, 41, 40, 31, 30, 28, 27, 26, 25, 24, 15, 14, 13, 12, 11, 10, 9, 8] USB Signals: OC[3:0]# VIH4/VIL4 JTAG Signals: EP_JTDI, EP_JTMS, EP_JTRST#, EP_JTCK VIH5/VIL5 JTAG Signals: JTDI, JTMS, JTCK VIH6/VIL6 Processor Signals: THRMTRIP# VIMIN7/VIMAX7 VIH8/VIL8 VIMIN9 - Gen1i, m VIMAX9 - Gen1i. m VIMIN9 - Gen2i, m VIMAX9 - Gen2i. m October 2012 Order Number: 327879-001US PCI Express* Data RX Signals: PCIE_EP_RX[p,n][15:0], PCIE_RC_RX[p,n][3:0], DMI_RX[p,n][3:0] Real Time Clock Signals: RTCX1 (Input from Clock Generator) SATA RX Signals: SATA[5:4]_RX[P,N] Intel(R) Communications Chipset 89xx Series - Datasheet 1611 Table 33-6. DC Input Characteristic Signal Association (Sheet 2 of 2) Symbol Associated Signals Rin-Diff10 GbE Signals: SRDSI_[3:0]_[P,N] VIH11 (Absolute Maximum) VIL11 (Absolute Minimum) Vclk_in_cross (abs) Clock Signals: CRU_CLK100[P,N], DMI_CLK100[P,N], USB_CLK96[P,N], SATA_CLK100[P,N], PCIE_EP_CLK100[P,N], GBE_CLK100[P,N] VIH12/VIL12 Power Management Signals: PWROK, RSMRST# System Management Signals: INTRUDER# Reset Signals: SRTCRST# VIH13/VIL13 Reset Signals: RTCRST# VIH_PECI/VIL_PECI Thermal Reporting Signals: PECI USB Signals: USB[5:0]D[p,n] (Low-speed and Full-speed)1 VDI / VCM / VSE VHSSQ / VHSDSC / VHSCM USB Signals: USB[5:0]D[p,n] (in High-speed Mode)1 1. USB not 5V tolerant Table 33-7. DC Input Characteristics (Sheet 1 of 2) Symbol Parameter Min Max Unit Notes VIL1 Input Low Voltage -0.3 0.8 V VIH1 Input High Voltage 2.0 VCCPEP3P3AUX+0.3 V VIL2 Input Low Voltage -0.3 0.8 V VIH2 Input High Voltage 2.0 VCC3P3+0.3 V Note 8 VIL3 Input Low Voltage -0.3 0.8 V Note 8 VIH3 Input High Voltage 2.0 VCCSUS3P3+0.3 V Note 8 VIL4 Input Low Voltage -0.1 0.35 V VIH4 Input High Voltage 0.75 VCCEPAUX+0.1 V VIL5 Input Low Voltage -0.1 0.35 V VIH5 Input High Voltage 0.75 VCCSUS+0.1 V VIL6 Input Low Voltage -0.1 0.35 V VIH6 Input High Voltage 0.75 VCCPCPU+0.1 V VIMIN7 Minimum Input Voltage 175 -- mVdiff p-p Note 4 VIMAX7 Maximum Input Voltage -- 1200 mVdiff p-p Note 4 Note 8 VIL8 Input Low Voltage -0.5 0.10 V VIH8 Input High Voltage 0.40 1.2 V VIMIN9 Gen1i Minimum Input Voltage - 1.5 Gb/s internal SATA 325 -- mVdiff p-p Note 5 VIMAX9 Gen1i Maximum Input Voltage - 1.5 Gb/s internal SATA -- 600 mVdiff p-p Note 5 Intel(R) Communications Chipset 89xx Series - Datasheet 1612 October 2012 Order Number: 327879-001US 33.0 Table 33-7. DC Input Characteristics (Sheet 2 of 2) Symbol Parameter Min Max Unit Notes VIMIN9 Gen2i Minimum Input Voltage - 3.0 Gb/s internal SATA 275 -- mVdiff p-p Note 5 VIMAX9 Gen2i Maximum Input Voltage - 3.0 Gb/s internal SATA -- 750 mVdiff p-p Note 5 Rin-Diff10 Received Differential Input Impedance GbE 80 120 VIL11 (Absolute Minimum) Input Low Voltage -0.3 -- V VIH11 (Absolute Maximum) Input High Voltage -- 1.150 V VIL12 Input Low Voltage -0.5 0.78 V VIH12 Input High Voltage 2.0 VCC3P3_RTC+0.5 V VIL13 Input Low Voltage -0.5 0.78 V VIH13 Input High Voltage 2.3 VCC3P3_RTC+0.5 V VIL_PECI Input Low Voltage -0.15 0.275(VCCPCPU) V VIH_PECI Input High Voltage VCCPCPU+0.15 V Note 6 Note 6 VDI Differential Input Sensitivity 0.2 -- V Note 1,3 VCM Differential Common Mode Range 0.8 2.5 V Note 2,3 VSE Single-Ended Receiver Threshold 0.8 2.0 V Note 3 VHSSQ HS Squelch Detection Threshold 100 150 mV Note 7 HS Disconnect Detection Threshold 525 625 mV Note 7 HS Data Signaling Common Mode Voltage Range -50 500 mV Note 7 VHSDSC VHSCM Notes: 1. 2. 3. 4. 5. 6. 7. 8. VDI = | USB[x]D[p] - USB[x]D[n] Includes VDI range Applies to Low-Speed/High-Speed USB PCI Express mVdiff p-p = 2*|PCIE_EP_RX[p][x] - PCIE_EP_RX[n][x]| SATA Vdiff, RX (VIMAX9/MIN9) is measured at the SATA connector on the receiver side (generally, the motherboard connector), where SATA mVdiff p-p = 2*|SATA[x]RXp - SATA[x]RXn| VCC3P3_RTC is the voltage applied to the VCC3P3_RTC well of the PCH. When the system is in a G3 state, this is generally supplied by the coin cell battery, but for S5 and greater, this is generally VCCSUS3P3. Applies to High-Speed USB 2.0. 3.3V refers to VCCSUS3P3 for signals in the suspend well and to VCC3P3 for signals in the core well and to VCC3P3 for signals in the ME well. See Section 32.0 for signal and power well association. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1613 Table 33-8. DC Output Characteristic Signal Association Symbol Associated Signals VOH1/VOL1 Processor Signal: PMSYNC; CPUPWRGD VOH2/VOL2 SPI EEPROM: GBE_EE_CS#, GBE_EE_DI, GBE_EE_SK GbE: GBE[3:0]_SWDP[1:0], GBE[3:0]_LED VOL3 GbE: GBE_WAKE# I2C GbE Signals: SFP[3:0]_I2C_CLK, SFP[3:0]_I2C_DATA SMBus GbE Signals: GBE_SMBCLK, GBE_SMBDAT, GBE_SMBALRT# SPI EEPROM: GBE_EE_CS#, GBE_EE_DI, GBE_EE_SK SMBus Signals: EP_SMBCLK, EP_SMBDAT, EP_SMBALRT# VOH4/VOL4 LPC/Firmware Hub Signals: LAD[3:0], LFRAME#, INIT3_3V# UART Signals: SIU[1:0]_TXD, SIU[1:0]_RTS#, SIU[1:0]_DTR#, WDT_TOUT# SATA Signals: SATA_LED#, SCLOCK, SLOAD, SDATAOUT0, SDATAOUT1 GPIO Signals: GPIO [55, 54, 53, 52, 51, 50, 49, 48, 39, 38, 37, 36, 35, 34, 33, 32, 23, 22, 21, 20, 19, 18, 17,16, 7, 6, 5, 4, 3, 2, 1, 0] SPI Signals: SPI_CS0#, SPI_CS1#, SPI_MOSI, SPI_CLK Interrupt Signals: SERIRQ Miscellaneous: BBS0 VOH5/VOL5 Power Management Signals: SLP_S3#, SLP_S4#, SLP_S5#, SUS_CLK, SUS_STAT#, DRAMPWRGD System Management Link Signals: SML1CLK, SML1DAT, SML1ALERT# GPIO Signals: GPIO [75, 74, 73, 72, 63, 62, 61, 60, 59, 58, 57, 56, 47, 46, 45, 44, 43, 42, 41, 40, 31, 30, 28, 27, 26, 25, 24, 15, 14, 13, 12, 11, 10, 9, 8] VOL6 SMBus Signals: MST_SMBCLK, MST_SMBDAT, MST_SMBALERT# (open drain suspend signals) VOL7 JTAG Signals: JTDO VOL8 JTAG Signals: EP_JTDO VOH9/VOL9 Reset Signals: PLTRST# VOH10/VOL10 USB Signals: USB[5:0]D[p,n] in Low-speed and Full-speed Modes VOMIN11/VOMAX11 SATA Data TX Signals: SATA[5:4]_TX[P,N] VOMIN12/VOMAX12 PCI Express* Data TX Signals: PCIE_EP_TX[p,n][15:0], PCIE_RC_TX[p,n][3:0], DMI_TX[p,n][3:0] VOMIN13/VOMAX13 GbE: SRDSO_[3:0]_P, SRDSO_[3:0]_N VOH_PECI/VOL_PECI PECI Signal: PECI VHSOI VHSOH VHSOL VCHIRPJ VCHIRPK USB Signals: USB[5:0]D[p,n] in High-speed Mode Intel(R) Communications Chipset 89xx Series - Datasheet 1614 October 2012 Order Number: 327879-001US 33.0 Table 33-9. DC Output Characteristics (Sheet 1 of 2) Symbol Parameter Min Max Unit IOL / IOH Notes Max VOL1 Output Low Voltage -- 0.255 V 3 mA VOH1 Output High Voltage VCCPCPU - 0.3 VCCPCPU V -3 mA VOL2 Output Low Voltage -- 0.4 V 10 mA VOL2 Output Low Voltage -- 0.2 V 100 uA VOH2 Output High Voltage 2.4 -- V -10 mA VOH2 Output High Voltage VCCEP3P3AUX-0.2 -- V -100 uA VOL3 Output Low Voltage -- 0.4 V 10 mA Note 1,4 VOL3 Output Low Voltage -- 0.2 V 100 uA Note 1,4 VOL4 Output Low Voltage -- 0.4 V 6 mA Note 1 VOH4 Output High Voltage VCC3P3- 0.5 VCC3P3 V -2 mA Note 1 VOL5 Output Low Voltage -- 0.4 V 6 mA VOH5 Output High Voltage VCCSUS3P3- 0.5 VCCSUS3P3 V -2 mA VOL6 Output High Voltage -- 0.4 V 6 mA VOL7 Output High Voltage -- 0.1(VCCEPAUX) V 1.5 mA VOL8 Output High Voltage -- 0.1(VCCSUS) V 1.5 mA VOL9 Output Low Voltage -- 0.1(VCCSUS3P3) V 1.5 mA VOH9 Output High Voltage 0.9(VCCSUS3P3) VCCSUS3P3 V -2.0 mA VOL10 Output Low Voltage -- 0.4 V 6 mA Output High Voltage VCCSUS3P3_USB0.5 VCCSUS3P3_USB V -2 mA VOH10 Note 1, 5 VOMIN11Gen1i,m Minimum Output Voltage 400 -- mVdiff p-p Note 3 VOMAX11Gen1i,m Maximum Output Voltage -- 600 mVdiff p-p Note 3 VOMIN11Gen2i,m Minimum Output Voltage 400 -- mVdiff p-p Note 3 VOMAX11Gen2i,m Maximum Output Voltage -- 700 mVdiff p-p Note 3 VOMIN12 Output Low Voltage 400 -- mVdiff p-p Note 2 VOMAX12 Output High Voltage -- 600 mVdiff p-p Note 2 VOMIN13 SerDes/SGMII Output Low Voltage 800 -- mVdiff p-p VOMAX13 SerDes/SGMII Output High Voltage -- 1200 mVdiff p-p VOL_PECI Output Low Voltage 0.25 (VCCPCPU) V 0.5 mA VOH_PECI Output High Voltage 0.75 (VCCPCPU) VCCPCPU V -6 mA -10.0 10.0 mV VHSOI HS Idle Level October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1615 Table 33-9. DC Output Characteristics (Sheet 2 of 2) Symbol Parameter Min Max Unit IOL / IOH Notes Max VHSOH HS Data Signaling High 360 440 mV VHSOL HS Data Signaling Low -10.0 10.0 mV VCHIRPJ Chirp J Level 700 1100 mV VCHIRPK Chirp K Level -900 -500 mV Notes: 1. The GBE_SMBDAT, GBE_SMBCLK, GBE_SMBALRT#, EP_SMBDAT, EP_SMBCLK, EP_SMBALRT#, MST_SMBDAT, MST_SMBCLK, and MST_SMBALRT# have open drain driver, and the VOH does not Apply. SATA_LED# has an open-collector driver, and the VOH spec does not apply. These signals must have external pull up resistor. 2. PCI Express mVdiff p-p = 2*|PCIE_EP_T[p][x]- PCIE_EP_T[n][x]|, PCI Express mVdiff p-p = 2*|PCIE_RC_T[p][x] - PCIE_RC_T[n][x]| 3. SATA Vdiff, tx (VOMIN7/VOMAX7) is measured at the SATA connector on the transmit side (generally, the motherboard connector), where SATA mVdiff p-p = 2*|SATA[x]TXP - SATA[x]TXN| 4. Reference supply or power well VCCEP3P3AUX. 5. Reference supply or power well VCCSUS3P3. 33.6 PCI Express* Root Complex and DMI Gen1 DC/AC Characteristics Table 33-10. PCI Express* and DMI Gen1 TX Specification Interface Timings Symbol UI VTX-DIFFpp TTX-DE-RATIO TTX-EYE Parameter Unit Interval Differential Peak to Peak Output Voltage Min Nom Max Unit 399.88 400 400.12 ps 1.2 V -4.0 dB 0.800 De-Emphasized Differential Output Voltage (Ration) -3.0 Minimum TX Eye Width 0.75 -3.5 Maximum time between the jitter MEDIAN-to-MAX- median and maximum deviation from the median. JITER FALL D+/D- TX Output Rise/Fall Time 0.125 RMS AC Peak Common Mode Output Voltage VTX-CM-DCLINE-DELTA Absolute Delta of DC Common Mode Voltage between D+ and D- VTX-DC-CM The TX DC Common Mode Voltage ITX-SHORT TX Short Circuit Current Limit RLTX-DIFF Differential Return Loss 10 DC Differential TX Impedance 80 ZTX-DIFF-DC 1. Note: UI 20 mV 0 25 mV 0 3.6 V 90 mA dB 100 120 DMI DC coupling uses half signal level (1/2 of the MIN and MAX) For additional information, see the PCI Express Protocol, Rev1.1 Intel(R) Communications Chipset 89xx Series - Datasheet 1616 UI 0.125 VTX-CM-ACp Note 1 UI TTX-EYE- TTX-RISE, TTX- Notes October 2012 Order Number: 327879-001US 33.0 Figure 33-1. PCI Express Gen1 Transmitter Eye Table 33-11. PCI Express* and DMI Gen1 RX Specification Interface Timings Symbol UI vRX-DIFFpp TRX-EYE TRX-EYE-MEDIAN- Parameter Unit Interval Differential Input Peak to Peak Voltage Minimum Receiver Eye Width Min Nom Max Unit 399.88 400 400.12 ps 0.175 1.2 0.4 V UI to-MAX-JITER Maximum time between the jitter median and maximum deviation from the median. 0.3 UI VRX-CM-ACp AC Peak Common Mode Input Voltage 150 mV RLRX-DIFF RLRX-CM ZRX-DIFF-DC ZRX-DC Note: Differential Return Loss Common Mode Return Loss 10 dB 6 dB DC Differential Input Impedance 80 100 120 DC Input Impedance 40 50 60 For additional information, see the PCI Express Protocol, Rev1.1 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1617 Figure 33-2. PCI Express Gen1 Receiver Eye Intel(R) Communications Chipset 89xx Series - Datasheet 1618 October 2012 Order Number: 327879-001US 33.0 33.7 PCIe* Gen2 EndPoint DC/AC Specification Table 33-12. PCI Express* Gen2 TX Specification Interface Timings Symbol 2.5 GT/S 5.0 GT/S Unit 399.88(min) 400.12(max) 199.94(min) 200.06(max) ps Differential Peak to Peak Output Voltage 0.8(min) 1.2(max) 0.8(min) 1.2(max) V VTX-DIFF-PP-LOW Low Power Differential Peak to Peak Output Voltage Swing 0.4(min) 1.2(max) 0.4(min) 1.2(max) V VTX-DE-RATIO-3.5dB De-Emphasized Differential Output Voltage (Ration) 3.0(min) 4.0(max) 3.0(min) 4.0(max) dB VTX-DE-RATIO-6dB De-Emphasized Differential Output Voltage (Ration) N/A 5.5(min) 6.5(max) dB Not Specified 0.9(min) UI 0.75(min) 0.75(min) UI 0.125(max) Not Specified UI Not Specified 0.15 (max) UI UI VTX-DIFF-PP TMIN-PULSE TTX-EYE TTX-EYE-MEDIAN-toMAX-JITER TTX-HF-DJ-DD Parameter Unit Interval Instantaneous LonePulse Width Minimum TX Eye Width Maximum time between the jitter median and maximum deviation from the median. TX Deterministic Jitter >1.5 MHz TTX-LF-RMS TX RMS Jitter <1.5 MHz Not Specified 3.0 ps RMS TTX-RISE, TTX-FALL TX Output Rise/Fall Time 0.125 (min) 0.15 (max) UI TRF-MISMATCH TX Output Rise/Fall Mismatch Not Specified 0.1 (max) UI BWTX-PLL Maximum TX PLL Bandwidth 22 (max) 16 (max) MHz BWTX-PLL-LO-3dB Minimum TX PLL Bandwidth for 3dB Peaking 1.5 (min) 8 (min) MHz BWTX-PLL-LO-1dB Minimum TX PLL Bandwidth for 1dB Peaking Not Specified 5 (min) MHz PKGTX-PLL1 TX PLL Peaking with 8MHz min BW Not Specified 3.0 (max) dB PKGTX-PLL2 TX PLL Peaking with 5MHz min BW Not Specified 1.0 (max) dB 10 (min) 10 (min) for 0.05-1.25 GHz 8 (min) for 1.25-2.5 GHz dB 6 (min) 6 (min) dB 80 (min) 120 (max) 120 (max) RLTX-DIFF RLTX-CM TX Package plus Si Differential Common Mode Return Loss ZTX-DIFF-DC DC Differential TX Impedance VTX-CM-AC-PP TX AC Peak Common Mode Output Voltage (5.0 GT/s) Not Specified 100 (max) mVpp VTX-CM-AC-PP TX AC Peak Common Mode Output Voltage (2.5 GT/s) Not Specified 100 (max) mVpp 90(max) 90 (max) mA ITX-SHORT TX Short Circuit Current Limit VTX-DC-CM The TX DC Common Mode Voltage 0 (min) 3.6 (max) 0 (min) 3.6 (max) V VTX-CM-DC-LINE- Absolute Delta of DC Common Mode Voltage between D+ and D- 0 (min) 25 (max) 0 (min) 25 (max) mV 75 (min)1 200 (max) 75 (min)1 200 (max) nF DELTA CTX 1. TX Package plus Si Differential Return Loss AC Coupling Capacitor For additional information, see the PCI Express Protocol, Rev2.0. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1619 Figure 33-3. PCI Express Transmitter Eye Table 33-13. PCI Express* Gen2 RX Specification Interface Timings (Sheet 1 of 2) Symbol UI vRX-DIFF-PP-CC TRX-EYE Parameter Unit Interval Differential Input Peak to Peak Voltage for Common Refclk RX Architecture 5.0 GT/S Unit 399.88 (min) 400.12 (max) 199.94 (min) 200.06 (max) ps 0.175 (min) 1.2 (max) 0.120 (min) 1.2 (max) V Minimum Receiver Eye Width N/A 0.40 (min) UI TRX-TJ-CC Max Receiver Inherent Timing Error N/A 0.40 (max) UI TRX-DJ-DD-CC Max Receiver Inherent Deterministic Timing Error N/A 0.30 (max) UI 0.3 (max) Not Specified UI Minimum Width Pulse at RX Not Specified 0.6 (min) UI Min/Max Pulse Voltage on Consecutive UI Not Specified 5 (max) - Maximum RX PLL Bandwidth 22 (max) 16 (max) MHz BWRX-PLL-LO-3dB Minimum RX PLL Bandwidth for 3dB peaking 1.5 (min) 8 (min) MHz BWRX-PLL-LO-1dB Minimum RX PLL Bandwidth for 1dB peaking Not Specified 5 (min) MHz RX PLL peaking with 8 MHz minimum Bandwidth Not Specified 3.0 dB TRX-EYE-MEDIAN-to-MAXJITER TRX-MIN-PULSE VRX-MAX-MIN-RATIO BWRX-PLL-HI PKGRX-PLL1 Maximum time between the jitter median and maximum deviation from the median. Intel(R) Communications Chipset 89xx Series - Datasheet 1620 2.5 GT/S October 2012 Order Number: 327879-001US 33.0 Table 33-13. PCI Express* Gen2 RX Specification Interface Timings (Sheet 2 of 2) Symbol Parameter RX PLL peaking with 5 MHz minimum Bandwidth PKGRX-PLL2 RLRX-DIFF Rx Package plus Si Differential Return Loss Common Mode Return Loss RLRX-CM ZRX-DIFF-DC ZRX-DC DC Input Impedance vRX-CM-AC-P Note: DC Differential Input Impedance RX AC Common Mode Voltage 2.5 GT/S 5.0 GT/S Unit Not Specified 1.0 dB 10 (min) 10 (min) for 0.05-1.5GHz 8 (min) for 1.252.5GHz dB 6 (min) 6 (min) dB 80 (min) 120 (max) Not Specified 40 (min) 60 (max) 40 (min) 60 (max) 150 (max) 150 (max) mVP For additional information, see the PCI Express Protocol, Rev 2.0. Figure 33-4. PCI Express Receiver Eye VTS-Diff = 0mV D+/D- Crossing point VRS-Diffp-p-Min>175mV .4 UI =TRX-EYE min 33.7.1 PCIe* Specification - Input Clock The input clock for PCIe* must be a differential input clock in frequency of 100 MHz. For full specifications, see the PCI-Express Card Electromechanical Specifications (refclk specifications). October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1621 33.8 SATA Gen1/Gen2 DC/AC Characteristics Table 33-14. SATA Specification Interface Timings (Sheet 1 of 2) Symbol UI Parameter Gen1 Unit Interval Gen2 Units Min Max Min Max 666.43 670.23 333.21 335.11 85 115 Notes ps Receiver Parameter ZdiffRX RX Pair Differential Impedance TJ-Con-DD-5 TJ at Connector, Data-Data, 5 0.43 UI DJ-Con-DD-5 DJ at Connector, Data-Data, 5 0.35 UI TJ-Con-DD-250 TJ at Connector, Data-Data, 250 0.60 UI DJ-Con-DD-250 DJ at Connector, Data-Data, 250 0.42 TTJ-Con-CD-10 TJ at Connector, Clk-Data, fBAUD/10 TDJ-Con-CD-10 TTJ-Con-CD-500 TDJ-Con-CD-500 DJ at Connector, Clk-Data, fBAUD/500 Tjcon CD-1667 Tj at Connector Clk-Data, fbaud/1667 .65 Dj con CD-1667 Dj at Connector Clk-Data, fbaud/1667 .35 RLDD11,RX RLCC11,RX VdiffRX 85 115 UI 0.46 UI DJ at Connector, Clk-Data, fBAUD/10 0.35 UI TJ at Connector, Clk-Data, fBAUD/500 0.60 UI 0.42 UI .65 UI .35 UI Min Differential Mode return loss (limits 150 - 300 MHz) 18 dB Min Differential Mode return loss (limits 1.2 - 2.4 GHz) 8 dB Min Differential Mode return loss (limits 2.4 - 3.0 GHz) 3 dB RX Common Mode Return Loss (limits 150 - 600 MHz) 5 dB RX Common Mode Return Loss (limits 1.2 - 2.4 GHz) 2 dB RX Common Mode Return Loss (limits 3.0 - 5.0 GHz) 1 dB RX Differential Input Voltage 240 600 240 750 mV Transmitter Parameter ZdiffTX RX Pair Differential Impedance 85 115 85 115 vTX-DIFF-PP VdiffTX, TX Differential Input Voltage 400 600 400 700 mvPP TTX-RISE TX Rise Time 100 273 67 136 ps Note 1 TTX-FALL TX Fall Time 100 273 67 136 ps Note 1 - 20 - TTX-SKEW TX Differential Skew VCM-AC TX AC Common Mode DJ-Con-DD-250 DJ at Connector, Data-Data, 250 TTJ-Con-CD-500 TJ at Connector, Clk-Data, fBAUD/500 TDJ-Con-CD-500 DJ at Connector, Clk-Data, fBAUD/500 Tjcon CD-1667 Tj at Connector Clk-Data, fbaud/1667 .65 Dj con CD-1667 Dj at Connector Clk-Data, fbaud/1667 .35 Intel(R) Communications Chipset 89xx Series - Datasheet 1622 20 ps 50 mv 0.220 UI 0.37 UI 0.19 UI .65 UI .35 UI October 2012 Order Number: 327879-001US 33.0 Table 33-14. SATA Specification Interface Timings (Sheet 2 of 2) Symbol RLDD11,TX RLCC11,TX Parameter Gen2 Units Min Differential Mode return loss (limits 150 - 300 MHz) 14 dB Min Differential Mode return loss (limits 1.2 - 2.4 GHz) 6 dB Min Differential Mode return loss (limits 2.4 - 3.0 GHz) 3 dB RX Common Mode Return Loss (limits 150 - 600 MHz) 5-8 dB RX Common Mode Return Loss (limits 1.2 - 2.4 GHz) 2 dB RX Common Mode Return Loss (limits 3.0 - 5.0 GHz) 1 dB 1. Gen1 Notes 20% - 80% Note: All parameters measured at Rload = 100 10% load. Note: For a detailed description of the symbols, see the IEEE1596.3-1996 Standard. 33.9 Gigabit Ethernet SGMII DC/AC Characteristics Table 33-15. GbE SGMII Driver DC Specification Interface Symbol1 Parameter2 vOut-High Output voltage high vOut-Low Output voltage low vRing Output ringing Min Nom Max Unit 1200 mV 800 mV 10 % |vOut-Diff| Output Differential Voltage 400 600 mV ROut Output impedance (single ended) 40 60 |vOut-Diff| Change in vOut-Diff between "0" and "1" 25 mV IShort-a, IShort-b Output current on Short to GND 40 mA 1. For a detailed description of the symbols, see the IEEE1596.3-1996 standard 2. All parameters measured at Rload = 100 10% load Table 33-16. GbE SGMII Receiver DC Specification Interface (AC Coupled) Symbol1 Parameter2 Min Max Unit VI Input voltage range a or b 675 1725 mV Vidth Input differential threshold -50 +50 mV Input differential hystersis 25 Received differential input impedance 80 RIn-Diff mV 120 1. For a detailed description of the symbols see the IEEE1596.3-1996 standard 2. All parameters measured at Rload = 100 10% load October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1623 Table 33-17. GbE SGMII Driver AC Specification Interface Symbol1 Parameter Signal Speed, per lane 1.25 Gb/Sec TFall vOut-Diff Fall time (20%-80%) 130 pSec TRise vOut-Diff Rise time (20%-80%) 130 pSec Min Nom Max Unit 1. For a detailed description of the symbols, see the IEEE1596.3-1996 standard 33.10 SerDes DC/AC Specification The SerDes interface supports the following standards: 1. PICMG 3.1 specification Rev 1.0 1000BASE-BX. 2. 1000BASE-KX electrical specification defined IEEE802.3ap clause 70. 3. SGMII on 1000BASE-BX or 1000BASE-KX compliant electrical interface (AC coupling with internal clock recovery). 4. SFP (Small Form-Factor Pluggable) Transceiver Rev 1.0 The specifications define the interface for the back-plane board connection, Interface to external 1000BASE-T PHY and the interface to fiber or SFP module. Table 33-19. SerDes GbE 1000 Base RX Specifications Symbol Parameter Min Nom Signal Speed, per lane 1.25 BER Target Bit Error Rate 10-12 VRX-Diffpp Differential Peak to Peak Input Voltage Max Unit Gb/Sec 1600 mV Table 33-20. SerDes GbE 1000 Base TX Specifications Symbol Parameter Min Signal Speed, per lane VTX-Diffpp Differential Peak to Peak Output Voltage VTX-Diffpp-DISABLE Differential Peak to Peak Output Voltage TX Disabled VTX-CM Common Mode Voltage Limits Nom Max 1.25 800 -0.15 Unit Gb/Sec 1200 mV 30 mV 1.2 V VTX-DIFF TEMP Differential Output Template TTRAN-TIME Transition Time (20%-80%) 130 PS TTX-RANDOM-JITTER Random Jitter Peak to Peak 0.27 UI TTX-DETERMIN-JITTER Deterministic Jitter Peak to Peak 0.17 UI TTX-TOTAL-JITTER Total Jitter Peak to Peak 0.35 UI Intel(R) Communications Chipset 89xx Series - Datasheet 1624 October 2012 Order Number: 327879-001US 33.0 33.11 CRU Clock DC/AC Specification Table 33-21. CRU Differential Input Clock DC Specifications Symbol Parameter Min Typical Max Units Figure CRU Clock (CRU_CLK100[P,N]) from a clock chip Notes 1 VIL Voltage Input Low -0.150 0.00 0.150 V 33-5 1 VIH Voltage Input High 0.660 0.710 0.850 V 33-5 1 VISE Voltage Input Single Ended 1.15 V 33-5 1 VCROSS(abs) Absolute Crossing Point 0.250 N/A 0.550 V 33-5, 33-6 1, 2, 3, 6, 9 VCROSS(rel) Relative Crossing Point 0.250+0.5 (VHavg-0.71) N/A 0.550+0.5 (VHavg-0.71) V 33-5, 33-6 1, 2, 3, 6, 7, 9 VCROSS Range of Crossing Point N/A N/A 0.140 V 33-5, 33-6 1, 2, 8, 9 VRBM Ringback Margin 0.200 N/A N/A V 33-5 1, 4, 9 VTM Threshold Margin VCROSS-0.10 N/A VCROSS+0.10 V 33-5 1, 5, 9 VTM Threshold Margin VCROSS-0.10 N/A VCROSS+0.10 V 33-5 1, 5, 9 -0.1 Notes: 1. Unless otherwise noted, all specifications in this table apply to all processor frequencies. 2. Crossing voltage is defined as the instantaneous voltage value when the rising edge of CRU_CLK100P equals the falling edge of CRU_CLK100N. 3. VHavg is the statistical average of the VH measured by the oscilloscope. 4. Ringback Margin is defined as the absolute voltage difference between the maximum Rising Edge Ringback and the maximum Falling Edge Ringback. 5. Threshold Region is defined as a region entered around the crossing point voltage in which the differential receiver switches. It includes input threshold hysteresis. 6. The crossing point must meet the absolute and relative crossing point specifications simultaneously. 7. VHavg can be measured directly using "Vtop" on Agilent* scopes and "High" on Tektronix* scopes. 8. VCROSS is defined as the total variation of all crossing voltages as defined in note 2. 9. Guaranteed by design. These values are typical values seen for this process, but not measured during production testing. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1625 Figure 33-5. CRU Differential Clock Waveform T3 Overshoot CRU_CLK100N VH Crossing Voltage Threshold Region Crossing Voltage Ringback Margin Rising Edge Ringback Falling Edge Ringback CRU_CLK100P VL Undershoot T4 T1 T1: CRU_CLK100[P, N] period T2: CRU_CLK100[P, N] period stability (not shown) T3: CRU_CLK100[P, N] pulse high time T4: CRU_CLK100[P, N] pulse low time T5: CRU_CLK100[P, N] rise time through the threshold region (not shown) T6: CRU_CLK100[P, N] fall time through the threshold region (not shown) B4965-01 Figure 33-6. CRU Differential Clock Cross-Point Specification 650 Crossing Point (mV) 600 550 550 mV 500 550 + 0.5 (VHavg - 710) 450 400 250 + 0.5 (VHavg - 710) 350 300 250 250 mV 200 660 670 680 690 700 710 720 730 740 750 760 770 780 790 800 810 820 830 840 850 VHavg (mV) B6569-01 Intel(R) Communications Chipset 89xx Series - Datasheet 1626 October 2012 Order Number: 327879-001US 33.0 Table 33-22. CRU Differential Input Clock Timing Specifications Symbol Parameter Min Nom Max Unit Figures Notes MHz - - CRU Clock (CRU_CLK100[P,N]) from a clock chip F CRU_CLK100 Frequency Duty Cycle - Duty Cycle T1 CRU_CLK100 Period T2 45 - 55 % - - 10.00 - 10.20 ns 33-5 1, 6 - 200.00 ps - 2, 6 350 ppm - - CRU_CLK100 Period Stability N/A PPM Parts per million Frequency Tolerance -350 Tjitter Cycle to cycle Jitter T3 100 - 50 150 ps - - TPH CRU_CLK100 Pulse High Time 3.94 5.00 6.12 ns 33-5 4, 6 T4 TPL CRU_CLK100 Pulse Low Time 3.94 5.00 6.12 ns 33-5 4, 6 T5 CRU_CLK100 Rise Time 175 - 700 ps 33-5 3, 5, 6 T6 CRU_CLK100 Fall Time 175 - 700 ps 33-5 3, 5, 6 Notes: 1. The period specified here is the average period. A given period may vary from this specification as governed by the period Stability specification (T2). 2. In this context, period stability is defined as the worst case timing difference between successive crossover voltages. In other words, the largest absolute difference between adjacent clock periods must be less than the period stability. 3. Slew rate is measured between the 35% and 65% points of the clock swing (VL to VH). 4. Combining the longest clock-high and clock-low times would violate the max clock period, and that combining the shortest clock-high and clock-low times would violate the minimum clock period. The clock-high and clock-low times specify the most extreme allowable combination of clock period and duty cycle. 5. Slew rate specifications apply to both rising and falling edges. 6. Guaranteed by design. These values are typical values seen for this process, but not measured during production testing. 33.12 SATA, DMI, and PCIe* Clocks DC/AC Specification The specifications that follow are from the CK505 Clock Synthesizer/Driver Specification. More details can be obtained in the CK505 specification. These specifications are valid at the silicon pad only. Ledges on the clock signals will be observed if measuring the clocks at the vias on the bottom of the component, since the clocks rely on reflective switching. The Frequency and Period numbers in Table 33-23 are taken directly from the "AbsPerMin" and "AbsPerMax" clock-clock period variation in the CK505 Clock Synthesizer/Driver Specification, with SSC enabled (except for the Display clock). Table 33-23. SATA, DMI, and PCIe* Differential Input Clock DC Specification (Sheet 1 of 2) Symbol Parameter Min Nom Max Unit -0.15 V VIL Voltage Input Low VIH Voltage Input High VISE Voltage Input Single Ended -0.1 1.15 V VCross Absolute Crossing Point 0.25 0.55 V VCross-delta Vcross Variation 0.14 V October 2012 Order Number: 327879-001US 0.15 V Intel(R) Communications Chipset 89xx Series - Datasheet 1627 Table 33-23. SATA, DMI, and PCIe* Differential Input Clock DC Specification (Sheet 2 of 2) Symbol Parameter Vmax (Absolute Overshoot) Single-ended max voltage Vmin (Absolute Undershoot) Single-ended min voltage Vrb-diff Differential ringback voltage threshold Min Nom Max 1.15 -0.1 Unit V V -0.10 .10 V Max Unit Table 33-24. SATA, DMI, and PCIe* Differential Input Clock AC Specification Symbol Parameter Min Nom SATA, DMI, and PCIe* Clocks (SATA_CLK100[P,N], DMI_CLK100[P,N], PCIE_EP_CLK100[P,N]) from a clock chip F Frequency Tperiod-abs Absolute Period (including Jitter and Spread Spectrum) Duty Cycle Duty Cycle of reference clock PPM Parts per million Frequency Tolerance ERRefclk-diffRrise ERRefclk-diffFall 100.0 9.847 MHz 10.203 ns 40 60 % -300 300 % Differential Rising and Falling edge rates 0.6 4 V/ns TStable Allowed time before ringback 500 Tjitter Cycle to cycle jitter -- ps 86 ps Table 33-24, "SATA, DMI, and PCIe* Differential Input Clock AC Specification" includes specs for common clock platforms. For a design that has separate clock platforms, see the PCIe* Rev2 specification for the separate refclock architecture specifications. Intel(R) Communications Chipset 89xx Series - Datasheet 1628 October 2012 Order Number: 327879-001US 33.0 Figure 33-7. SATA, DMI, and PCIe* Clocks Differential Clock Waveform T3 Overshoot CLK100N VH Crossing Voltage Threshold Region Crossing Voltage Ringback Margin Rising Edge Ringback Falling Edge Ringback CLK100P VL Undershoot T4 T1 T1: CLK100[P, N] period T2: CLK100[P, N] period stability (not shown) T3: CLK100[P, N] pulse high time T4: CLK100[P, N] pulse low time T5: CLK100[P, N] rise time through the threshold region (not shown) T6: CLK100[P, N] fall time through the threshold region (not shown) B4965-01 Figure 33-8. Single-Ended Measurement Point for Absolute Cross Point and Swing CLK100N CLK100P -0.10 V Figure 33-9. Single-Ended Clock Measurement Points for Delta Cross Point October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1629 Figure 33-10.Differential Clock Cross Point Specification 650 Crossing Point (mV) 600 550 550 mV 500 450 550 + 0.5 (VHavg - 700) 400 250 + 0.5 (VHavg - 700) 350 300 250 250 mV 200 660 670 680 690 700 710 720 730 740 750 760 770 780 790 800 810 820 830 840 850 VHavg (mV) Figure 33-11.Differential Measurement Point for Duty Cycle and Period Figure 33-12.Differential Measurement Point for Rise and Fall Time Intel(R) Communications Chipset 89xx Series - Datasheet 1630 October 2012 Order Number: 327879-001US 33.0 Figure 33-13.Differential Measurement Point for Ringback VIH = +150mV VRB = +100mV 0mV VRB = -100mV VIL = -150mV 33.13 AC Characteristics Table 33-25. Digital 3.3V I/O Timing Characteristics Symbol Parameter Min Nom Max Unit Cload TOUT-RISE Output Rise Time 0.53 0.89 1.27 ns 20pF TOUT-FALL Output Fall Time 0.56 0.93 1.27 ns 20pF Output Delay Rise 1.15 1.96 3.11 ns 20pF Output Delay Fall 1.22 2.11 3.33 ns 20pF TIN-RISE Input Rise Time 0.03 0.1 ns 0.2pF TIN-FALL Input Fall Time 0.03 0.1 ns 0.2pF Input Delay Rise 0.3 1.5 ns 0.2pF Input Delay Fall 0.3 1.5 ns 0.2pF TOUT-DELAYRISE TOUT-DELAYFALL TIN-DELAYRISE TIN-DELAYFALL Notes: 1. Input delay test conditions: Maximum input level = VIN = 2.7V; Input rise/fall time TPAD-RISE/TPADFALL (0.2VIN to 0.8VIN) = 1ns (Slew Rate ~ 1.5ns). 2. This table can be used for all 3.3V I/O timing not specified anywhere in this document. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1631 Figure 33-14.Digital I/O Output Timing Diagram TOUT-DELAY-FALL TOUT-DELAY-RISE 0.5VCC 0.8VCC3P3 TOUT-FALL TOUT-RISE 0.5VCC3P3 0.2VCC3P3 PAD Figure 33-15.Digital I/O Input Timing Diagram VIN=2.7 0.8VIN TPAD-RISE TPAD-FALL 0.5VIN TPAD-RISE = TPAD-FALL = 1n Sec 0.2VIN PAD TIN-RISE TIN-FALL 0.8VCC 0.5VCC 0.2VCC TIN-DELAY-RISE Intel(R) Communications Chipset 89xx Series - Datasheet 1632 TIN-DELAY-FALL October 2012 Order Number: 327879-001US 33.0 Table 33-26. Clock Timings (Sheet 1 of 2) Symbol Parameter Min Max Unit Notes Figure RTC Clock (RTCX1, RTCX2 32.768KHz) Cin1, Cin2 Tolerance -20 20 Duty Cycle 45 55 High time 10 Low time 10 1 Pin Input Capacitance ppm % s 3 33-22 -- s 3 33-22 1.2 pF UART Clock (UART_CLK 14.7456MHz) Period 67 70 ns High time 20 -- ns 33-22 Low time 20 -- ns 33-22 Duty Cycle 45 55 % Rise time 1 4.0 V/ns 3 33-22 Fall time 1 4.0 V/ns 3 33-22 Jitter cycle to cycle -- 350 ps 3 33-22 29.566 30.584 ns LPC Clock (PCICLK 33.00MHz) t1 Period t2 High Time 10.826 17.850 ns t3 Low Time 10.426 17.651 ns 33-22 Duty Cycle 45 55 % 33-22 t4 Rising Edge Rate 1.0 4 V/ns 3 t5 Falling Edge Rate 1.0 4 V/ns 3 33-22 -- 500 ps 3 33-22 Jitter cycle to cycle 33-22 Timer (8254) Clock (REF_CLK14 14.31818MHz) t6 Period 69.820 69.862 ns t7 High Time 29.975 38.467 ns 33-22 t8 Low Time 29.575 38.267 ns 33-22 Duty Cycle PPM Frequency Stability 45 55 % -500 +500 PPM 3 33-22 5,6,7 t4 Rising Edge Rate 1.0 4 V/ns 3 33-22 t5 Falling Edge Rate 1.0 4 V/ns 3 33-22 -- 1000 ps 3 33-22 20.831 20.835 ns Jitter cycle to cycle UART Clock (UART_CLK 48MHz) t9 Period 1 t10 High Time 8.217 11.152 ns t11 Low Time 7.817 10.952 ns 33-22 45 55 % 33-22 Duty Cycle October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1633 Table 33-26. Clock Timings (Sheet 2 of 2) Symbol Parameter Min Max Unit Notes Figure t4 Rising Edge Rate 1.0 4 V/ns 3 33-22 t5 Falling Edge Rate 1.0 4 V/ns 3 33-22 -100 +100 PPM -- 350 ps 3 33-22 Frequency Stability (see the Clock Chip Specification) Jitter cycle to cycle SMBus/SMLink Clock (MST_SMBCLK, EP_SMBCLK, GBE_MST_SMBCLK, SML0CLK) fsmb Operating Frequency 10 100 KHz t22 High time 4.0 50 s t23 Low time 4.7 -- s t24 Rise time -- 1000 ns 3 33-31 t25 Fall time -- 300 ns 3 33-31 2 33-31 33-31 GbE Clock (GBE_CLK100[P,N]) from a clock chip t36 Duty Cycle Period 9.997 10.0533 45 55 1 8 V/ns -50 +50 PPM -5 +5 PPM per Year 10 nF Voltage Swing (HCSL) 0.55 0.85 Vpp Voltage Input Single Ended -0.1 1.15 V 80 ps Duty Cycle Slew rate PPM Frequency Stability (see the Clock Chip Specification) Aging Coupling Capacitor VISE Jitter Cycle to cycle ns 100MHz 5,6,7 USB Clock (USB_CLK96[P,N]) from a clock chip t36 Period Slew rate VISE 10.066 10.768 ns 1 8 V/ns 4 Frequency Stability (see the Clock Chip Specification) -100 +100 PPM 5,6,7 Voltage Swing (HCSL) 0.55 0.85 Vpp Voltage Input Single Ended -0.1 1.15 V -- 250 ps Jitter cycle to cycle 5,6,7 Suspend Clock (SUSCLK) fsusclk kHz 2 t39 Operating Frequency High Time 10 32.768 -- s 2 t39a Low Time 10 -- s 2 SPI Clock (SPI_CLK) Slew_Rise Output Rise Slew Rate (0.2VCC3P3 0.6VCC3P3) 1 4 V/ns 1 Slew_Fall Output Fall Slew Rate (0.6VCC3P3 0.2VCC3P3) 1 4 V/ns 1 1. The UART_CLK @48MHz expects a 40/60% duty cycle. Intel(R) Communications Chipset 89xx Series - Datasheet 1634 October 2012 Order Number: 327879-001US 33.0 2. 3. 4. 5. 6. 7. 8. SUSCLK duty cycle can range from 30% minimum to 70% maximum. Rise and Fall time is measured 20% and 80% of signal edge. See CK420 Clock Synthesizer Specification for measurement procedure. Jitter is specified as cycle to cycle measured in pico seconds. Period min and max includes cycle to cycle jitter. On all jitter measurements care should be taken to set the zero crossing voltage (for rising edge) of the clock to be the point where the edge rate is the fastest. Using a Math function = Average (Derivative (Ch1)) and set the averages to 64, place the cursors where the slope is the highest on the rising edge usually this lower half of the rising edge. The reason this is defined is for users trying to measure in a system it is impossible to get the probe exactly at the end of the Transmission line with large Flip Chip components, this results in a reflection induced ledge in the middle of the rising edge and will significantly increase measured jitter. Phase jitter requirement: The designated Gen2 outputs will meet the reference clock jitter requirements from the PCI Express Gen2 Base Spec. The test is to be performed on a component test board under quiet conditions with all clock outputs on. Jitter analysis is performed using a standardized tool provided by the PCI SIG. Measurement methodology is defined in Intel document "PCI Express Reference Clock Jitter Measurements". Testing condition: 1K pull up to VCC3P3, 1K pull down and 10pF pull down and 1/2 inch trace Table 33-27. Universal Serial Bus (USB) Timing Sym Parameter Min Max Units Notes Fig High-speed Source (Note 7) t100 USB[5:0]Dp, USB[5:0]Dn Driver Rise Time 0.8 1.2 ns 1, CL = 10 pF t101 USB[5:0]Dp, USB[5:0]Dn Driver Fall Time 0.8 1.2 ns 1, CL = 10 pF Full-speed Source (Note 8) t102 USB[5:0]Dp, USB[5:0]Dn Driver Rise Time 4 20 ns 1, CL = 50 pF t103 USB[5:0]Dp, USB[5:0]Dn Driver Fall Time 4 20 ns 1, CL = 50 pF t104 Source Differential Driver Jitter - To Next Transition - For Paired Transitions -3.5 -4 3.5 4 ns ns 2, 3 33-29 t105 Source SE0 interval of EOP 160 175 ns 4 33-30 t108 EOP Width: Must accept as EOP 82 -- ns 2 33-30 Low-speed Source (Note 9) t110 USB[5:0]Dp, USB[5:0]Dn Driver Rise Time 75 300 ns 1, 6 CL = 50 pF CL = 350 pF t111 USB[5:0]Dp, USB[5:0]Dn Driver Fall Time 75 300 ns 1,6 CL = 50 pF CL = 350 pF t112 Source Differential Driver Jitter To Next Transition For Paired Transitions -25 -14 25 14 ns ns 2, 3 t114 Source Jitter for Differential Transition to SE0 Transition -40 100 ns 5 t116 EOP Width: Must accept as EOP 670 -- ns 2 Notes: 1. 2. 3. 4. 5. 6. 7. 8. 9. 33-29 33-30 Driver output resistance under steady state drive is spec'd at 28 at minimum and 43 at maximum. Timing difference between the differential data signals. Measured at crossover point of differential data signals. Measured at 50% swing point of data signals. Measured from last crossover point to 50% swing point of data line at leading edge of EOP. Measured from 10% to 90% of the data signal. High-speed Data Rate has minimum of 479.760 Mb/s and maximum of 480.240 Mb/s. Full-speed Data Rate has minimum of 11.97 Mb/s and maximum of 12.03 Mb/s. Low-speed Data Rate has a minimum of 1.48 Mb/s and a maximum of 1.52 Mb/s. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1635 Table 33-28. SMBus Timing Sym Parameter Min Max Units Notes Fig t131 Hold Time after (repeated) Start Condition. After this period, the first clock is generated. 4.0 -- s 33-31 t132 Repeated Start Condition Setup Time 4.7 -- s 33-31 4.0 -- s 0 -- ns 250 -- ns t133 Stop Condition Setup Time t134 Data Hold Time t135 Data Setup Time 1. 33-31 1 33-31 33-31 t134 has a minimum timing for I2C of 0 ns, while the minimum timing for SMBus is 300 ns. Table 33-29. LPC Timing Sym t150 Parameter Min Max Units 2 13 ns 33-23 ns 33-27 28 ns 33-25 LAD[3:0] Valid Delay from PCICLK Rising t151 LAD[3:0] Output Enable Delay from PCICLK Rising 2 t152 LAD[3:0] Float Delay from PCICLK Rising -- Notes Fig t153 LAD[3:0] Setup Time to PCICLK Rising 7 -- ns 33-24 t154 LAD[3:0] Hold Time from PCICLK Rising 0 -- ns 33-24 t155 LDRQ[1:0]# Setup Time to PCICLK Rising 12 -- ns 33-24 t156 LDRQ[1:0]# Hold Time from PCICLK Rising 0 -- ns 33-24 t157 LFRAME# Valid Delay from PCICLK Rising 2 13 ns 33-23 Table 33-30. SPI Timings (20 MHz) Sym t180a Parameter Serial Clock Frequency - 20M Hz Operation Min Max Units Notes 17.2 18.4 MHz 1 Fig t182a SPI Clock Duty cycle at the host 43 57 % 33-33 t183a Tco of SPI_MOSI with respect to serial clock falling edge at the host -5 13 ns 33-33 t184a Setup of SPI_MISO with respect to serial clock falling edge at the host 16 -- ns 33-33 t185a Hold of SPI_MISO with respect to serial clock falling edge at the host 0 -- ns 33-33 t186a Setup of SPI_CS[1:0]# assertion with respect to serial clock rising at the host 30 -- ns 33-33 t187a Hold of SPI_CS[1:0]# deassertion with respect to serial clock falling at the host 30 -- ns 33-33 t188a SPI_CLK high time 23.84 -- ns 2 33-33 t189a SPI_CLK low time 31.84 -- ns 2 33-33 Note: 1. 2. The typical clock frequency driven by the PCH is 17.86 MHz. Measurement point for low time and high time is taken at 0.5 (VCC3P3) Intel(R) Communications Chipset 89xx Series - Datasheet 1636 October 2012 Order Number: 327879-001US 33.0 Table 33-31. SPI Timings (33 MHz) Sym Parameter Min Max Units Notes 1 Fig t180b Serial Clock Frequency - 33 MHz Operation 30.3 32.19 MHz t182b SPI Clock Duty cycle at the host 50 50 % 33-33 t183b Tco of SPI_MOSI with respect to serial clock falling edge at the host -5 5 ns 33-33 t184b Setup of SPI_MISO with respect to serial clock falling edge at the host 8 -- ns 33-33 t185b Hold of SPI_MISO with respect to serial clock falling edge at the host 0 -- ns 33-33 t186b Setup of SPI_CS[1:0]# assertion with respect to serial clock rising at the host 30 -- ns 33-33 t187b Hold of SPI_CS[1:0]# deassertion with respect to serial clock falling at the host 30 -- ns 33-33 t188b SPI_CLK high time 14.88 -- ns 2 33-33 t189b SPI_CLK low time 15.18 -- ns 2 33-33 Note: 1. 2. The typical clock frequency driven by the PCH is 31.25 MHz. Measurement point for low time and high time is taken at 0.5 (VCC3P3) Table 33-32. SPI Timings (50 MHz) Sym Parameter Min Max Units Notes Fig 1 33-33 t180c Serial Clock Frequency - 50MHz Operation 46.99 53.40 MHz t182c SPI Clock Duty cycle at the host 40 60 % 33-33 t183c Tco of SPI_MOSI with respect to serial clock falling edge at the host -3 3 ns 33-33 t184c Setup of SPI_MISO with respect to serial clock falling edge at the host 8 - ns 33-33 t185c Hold of SPI_MISO with respect to serial clock falling edge at the host 0 - ns 33-33 t186c Setup of SPI_CS[1:0]# assertion with respect to serial clock rising edge at the host 30 - ns 33-33 t187c Hold of SPI_CS[1:0]# assertion with respect to serial clock rising edge at the host 30 - ns 33-33 t188c SPI_CLK High time 7.1 - ns 2, 3, 4 33-33 t189c SPI_CLK Low time 11.17 - ns 2, 3 33-33 1. 2. 3. 4. Typical clock frequency driven by the PCH is 50 MHz. This frequency is not available for ES1 samples. When using 50 MHz mode ensure target flash component can meet t188c and t189c specifications. Measurement point for low time and high time is taken at 0.5(VCC3P3) When operating at 50MHz, the PCH uses a divide down from 125MHz clock. Due to the 40% duty cycle, the PCH SPI Flash Controller cannot meet the minimum high timing requirements of a 50MHz SPI Flash component and a 66MHz rate or faster SPI Flash component must be used. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1637 33.13.1 I2C/SFP AC Specification Table 33-33 indicates the timing of the SFPn_I2C_CLK and SFPn_I2C_DATA pins when operating in I2C mode. Table 33-33. I2C Timing Parameters Symbol Parameter Min Typ Max Units THD:DAT Data hold time 0.3 s TSU:DAT Data setup time 0.25 s Figure 33-16.I2C I/F Timing Diagram Intel(R) Communications Chipset 89xx Series - Datasheet 1638 October 2012 Order Number: 327879-001US 33.0 33.13.2 MDIO DC Specification Table 33-34, "DC Input Characteristics: MDIO Mode of Operation" and Table 33-35, "DC Output Characteristics: MDIO Mode of Operation" indicate the DC and timing of the SFPn_I2C_CLK and SFPn_I2C_DATA pins when operating in MDIO mode. Table 33-34. DC Input Characteristics: MDIO Mode of Operation Symbol Parameter Conditions Min Nominal Max Units Notes VIH Input high voltage VIH > VIH_Min VCC = Min 2.0 - 3.6 V - VIL Input low voltage VIH > VIL_Max VCC = Min - - 0.8 V - IIH Input high current VCC = Max VIN = 2.5 V - - 15 uA - IIL Input low current VCC = Max VIN = 0.4 V -15 - - uA - leak Input Leakage Current 0<VIN< VCCPEP3P3AUX -10 - 10 A - CIN Input pin cap - - 8 pF 1 Notes: 1. Guaranteed by design. These values are typical values seen for this process, but not measured during production testing. Table 33-35. DC Output Characteristics: MDIO Mode of Operation Symbol Parameter Conditions Min Typical Max Units Notes VOH Output high voltage VCC = 3.3V IOH = -1mA 2.4 - VCCPEP3P3AUX V 1 VOH Output high voltage VCC = 3.3V IOH = -4mA 2.4 - VCCPEP3P3AUX V 1 VOL Output low voltage VCC = Min IOL = 1mA - - 0.4 V Notes: 1. The MDIO buffer is powered from the 3.3V Auxiliary power rail. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1639 33.13.3 MDIO AC Specification The PCH is designed to support the MDIO specifications defined in IEEE 802.3 clause 22. Table 33-36, "MDIO I/F Timing Parameters" contains specifications applicable over a recommended operating range at VCC3P3 = 3.3V and Cload = 16pF (unless otherwise noted). For MDIO I/F timing specification, see Table 33-36, Figure 33-17, and Figure 33-18. Table 33-36. MDIO I/F Timing Parameters Symbol Parameter Min Max Units Note 500 nS 1, 2 0 nS 1, 2 10 nS 1, 2 tMCLK MDC clock period (2 MHz) tMHI MDIO input hold time after the rising edge of MDC clock tMHO MDIO output hold time after the rising edge of MDC clock tMSU MDIO setup time 10 tMPR MDIO propagation Delay 10 Typ TMCLK/2 +10 nS 1, 2 nS 1, 2 Notes: 1. The table above applies to MDIO0, MDC0, MDIO1, MDC1, MDIO2, MDC2, MDIO3, and MDC3. 2. Timing measured relative to MDC reference voltage of 2.0V (Vih). Figure 33-17.MDIO Input AC Timing Diagram TMCLK TMSU TMHI MDC MDIO Intel(R) Communications Chipset 89xx Series - Datasheet 1640 October 2012 Order Number: 327879-001US 33.0 Figure 33-18.MDIO Output AC Timing Diagram TMCLK TMHO TMPR MDC MDIO 33.13.4 EEPROM AC Specification The PCH is designed to support a standard serial EEPROM. The following timing specifications are applicable over a recommended operating range at VCC3P3 = 3.3V and Cload = 16pF unless otherwise noted. For EEPROM I/F timing specification, see Table 33-37 and Figure 33-19. Table 33-37. EEPROM I/F Timing Parameters Symbol Parameter Min Typ Max Units Note tSCK GBE_EE_SK clock frequency 0 2 2.1 MHz 1 2 s 2 s tRI Input rise time tFI Input fall time tWH GBE_EE_SK high time 200 250 tWL GBE_EE_SK low time 200 250 tCS GBE_EE_CS# high time 250 ns ns ns tCSS GBE_EE_CS# setup time 250 ns tCSH GBE_EE_CS# hold time 250 ns tSU Data-in setup time 50 ns tH Data-in hold time 50 ns tV Output valid 0 tHO Output hold time 0 tDIS Output disable time 2 200 ns ns 250 ns Notes: 1. Clock is 2MHz 2. 50% duty cycle. October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1641 Figure 33-19.EEPROM Timing Diagram tCS VIH GBE_EE_CS# t CSS t CSH t WH VIH GBE_EE_SK VIL t SU t WL tH VIH GBE_EE_DI VIL VOH GBE_EE_DO VOL VALID IN t val t HO t DIS HI-Z HI-Z Table 33-38. PECI Timings Sym Parameter Min Max Units Notes 0.495 0.495 500 250 s s 1 tBIT Bit time (overall time evident on PECI) Bit time driven by an originator tBIT,jitter Bit time jitter between adjacent bits in an PECI message header or data bytes after timing has been negotiated -- -- % tBIT,drift Change in bit time across a PECI address or PECI message bits as driven by the originator. This limit only applies across tBIT-A bit drift and tBIT-M drift. -- -- % 4 tH1 High level time for logic '1' 0.6 0.8 x tBIT 2 tH0 High level time for logic '0' 0.2 0.4 x tBIT tPECIR Rise time (measured from VOL to VIH,min, Vtt(nom) -5%) -- 30 + 5 ns/node 3 tPECIF Fall time (measured from VOH to VIL,max, Vtt(nom) +5%) -- 30 ns/node 3 Notes: 1. The originator must drive a more restrictive time to allow for quantized sampling errors by a client yet still attain the minimum time less than 500 s. tBIT limits apply equally to tBIT-A and tBIT-M. PCH is targeted on 2 MHz which is 500 ns bit time. 2. The minimum and maximum bit times are relative to tBIT defined in the Timing Negotiation pulse. 3. Extended trace lengths may appear as additional nodes. 4. tBIT-A is the negotiated address bit time and tBIT-M is the negotiated message bit time. Intel(R) Communications Chipset 89xx Series - Datasheet 1642 October 2012 Order Number: 327879-001US 33.0 33.13.5 UART AC Specification This section describes the Universal Asynchronous Receiver/Transmitter (UART) serial port used for the two UARTs that are integrated into the Serial I/O unit and Watchdog Timer (SIW). The UART can be controlled via programmed I/O. The basic programming model is the same for both UARTs, with the only difference being the Logical Device Number assigned to each. The serial port consists of a UART that supports all the functions of a standard 16550 UART, including hardware flow control interface. Table 33-39. UART Received Timing Symbol Parameter Min tSetUp Set Up time from the rising edge of UART_CLK tHold Hold time from the rising edge of UART_CLK Typ Max Units Note 7 nS 1 0 nS 1 Notes: 1. Applies to SIU[1:0]_RX, SIU[1:0]_CTS#, SIU[1:0]]_DSR#, SIU[1:0]_DCD#, SIU[1:0]_SI# Table 33-40. UART Transmit Timing Symbol Parameter Min tValid1-Delay1 Valid Output Delay 1 from the rising edge of UART_CLK 2 tValid-Delay2 Valid Output Delay 2 from the rising edge of UART_CLK 2 Typ Max Units Note 11 nS 1 nS 2 Notes: 1. Applies to SIU[1:0]_TX 2. Applies to SIU[1:0]DTR#, SIU[1:0]_RST# October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1643 33.13.6 JTAG AC Specification The PCH is designed to support the IEEE 1149.1 standard. For JTAG I/F timing specification, see Table 33-41 and Figure 33-20. Table 33-41. JTAG I/F Timing Parameters Symbol Parameter Min Typ Max tJCLK JTCK/EP_JTCK clock frequency tJH JTMS/EP_TMS and JTDI/EP_TDI hold time 10 nS tJSU JTMS/EP_TMS and JTDI/EP_TDI setup time 10 nS tJPR JTDO/EP_TDO propagation Delay 1. 10 Units 15 MHz Note 1 nS Timing measured relative to JTCK reference voltage of VCC3P3/2. Note: Table 33-41 applies to JTCK, JTMS, JTDI and JTDO. Figure 33-20.JTAG AC Timing Diagram Tjclk Tjsu Tjh JTCK JTMS JTDI JTDO Tjpr 33.14 AC Timing Diagrams Figure 33-21.Clock Cycle Time Intel(R) Communications Chipset 89xx Series - Datasheet 1644 October 2012 Order Number: 327879-001US 33.0 Figure 33-22.Clock Timing Period High Time 2.0V 0.8V Low Time Fall Time Rise Time Figure 33-23.Valid Delay from Rising Clock Edge Clock 1.5V Valid Delay Output VT Figure 33-24.Setup and Hold Times 1.5V Clock Setup Time Input Hold Time VT VT Figure 33-25.Float Delay Input VT Float Delay Output October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1645 Figure 33-26.Pulse Width Pulse Width VT VT Figure 33-27.Output Enable Delay 1.5V Clock Output Enable Delay Output VT Figure 33-28.USB Rise and Fall Times Rise Time 90% CL Fall Time 90% Differential Data Lines 10% 10% CL tR tF Low-speed: 75 ns at CL = 50 pF, 300 ns at C L = 350 pF Full-speed: 4 to 20 ns at C L = 50 pF High-speed: 0.8 to 1.2 ns at C L = 10 pF Intel(R) Communications Chipset 89xx Series - Datasheet 1646 October 2012 Order Number: 327879-001US 33.0 Figure 33-29.USB Jitter T period Crossover Points Differential Data Lines Jitter Consecutive Transitions Paired Transitions Figure 33-30.USB EOP Width Tperiod Data Crossover Level Differential Data Lines EOP Width Figure 33-31.SMBus Transaction t19 t20 t21 SMBCLK t135 t131 t134 t133 t132 t18 SMBDATA t130 October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1647 Figure 33-32.SMBus Timeout Start Stop t137 CLKack t138 CLKack t138 SMBCLK SMBDATA Figure 33-33.SPI Timings t188 t189 t182 t182 SPI_CLK t183 SPI_MOSI t184 t185 SPI_MISO t186 t187 SPI_CS[1:0]# Intel(R) Communications Chipset 89xx Series - Datasheet 1648 October 2012 Order Number: 327879-001US 33.0 Figure 33-34.Measurement Points for Differential Waveforms Differential Clock - Single Ended Measurements V max = 1.15V V max = 1.15V Clock# Vcross max = 550mV Vcross max = 550mV Vcross min = 300 mV Vcross min = 300 mV Clock V min = -0.30V V min = -0.30V Clock# Vcross delta = 140mV Vcross delta = 140mV Clock Clock# Vcross median ll fa Vcross median T Vcross median +75mV Tr i se Clock# Vcross median -75mV Clock Clock Differential Clock - Differential Measurements Clock Period (Differential ) Positive Duty Cycle (Differential ) Negative Duty Cycle (Differential ) 0.0V Clock-Clock# Rise Edge Rate Fall Edge Rate Vih_min = +150 mV 0.0V Vil_max = -150 mV Clock-Clock# October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1649 34.0 Thermal Specifications and Design Considerations For thermal specifications and design guidelines, see the Intel(R) Communications Chipset 89xx Series Thermal Mechanical Design Guide. Intel(R) Communications Chipset 89xx Series - Datasheet 1650 October 2012 Order Number: 327879-001US 35.0 35.0 Packaging Information 35.1 Package Introduction The product line is offered with the following specifications: * Flip-Chip Ball Grid Array (FCBGA) * Package Size: 27mm x 27mm * Ball Count: 942 * Ball Pitch: 0.7mm 35.2 Ball Map Information This section contains the following ball map information: * Table 35-1, "Alphabetical Ball Listing" * Table 35-2, "Alphabetical Signal Listing" * Table 35-3, "Ball Map Color Code" * Table 35-4, "Bottom View Left" * Table 35-5, "Bottom View Right" * Table 35-6, "Top View Left" * Table 35-7, "Top View Right" October 2012 Order Number: 327879-001US Intel(R) Communications Chipset 89xx Series - Datasheet 1651 Packaging Information 35.2.1 Ball Map Pin Lists Table 35-1. Alphabetical Ball Listing Ball Signal A4 VSS C4 A6 RSVD8 A9 A12 Signal Ball Signal RSVD6 E17 RSVD26 C6 SRDS1_SD E19 VSS VSS C9 RSVD79 E21 RSVD34 EP_JTCK C12 EP_SMBALRT# E27 PCIE_EP_RXn15 A16 VSS C16 RSVD13 E29 PCIE_EP_RXn14 A19 VSS C19 RSVD28 E31 PCIE_EP_RXn13 A22 RSVD27 C22 VSS E37 PCIE_EP_RXp9 A26 VSS C26 VSS E39 PCIE_EP_RXp8 A29 VSS C29 PCIE_EP_RXp14 E41 PCIE_EP_RXn8 A32 VSS C32 VSS E49 VSS A36 VSS C36 PCIE_EP_RXp10 E52 VSS A39 VSS C39 PCIE_EP_RXn9 E54 VSS A42 VSS C42 PCIE_EP_RXn7 E56 RSVD54 A46 VSS C46 PCIE_EP_RXp4 F1 RSVD3 A49 VSS C49 VSS F3 SRDS2_SD A52 VSS C52 VSS F5 RSVD85 A54 VSS C54 RSVD52 F8 GBE_EE_DO B2 RSVD2 D1 RSVD4 F10 PCIE_EP_RST# B5 RSVD1 D3 VSS F12 VSS B7 GBE_AUX_PWR_AVAIL D5 GBE_EE_DI F13 RSVD16 B11 RSVD81 D7 VSS F15 VSS B13 EP_SMBCLK D13 EP_JTRST# F16 VSS B15 RSVD5 D15 RSVD22 F18 RSVD9 B17 RSVD11 D23 RSVD23 F20 RSVD30 B21 RSVD21 D25 RSVD31 F22 RSVD18 B23 RSVD25 D33 PCIE_EP_RXn12 F23 RSVD19 B25 RSVD29 D35 PCIE_EP_RXp11 F25 VSS B27 PCIE_EP_RXp15 D43 PCIE_EP_RXn6 F26 VSS B31 PCIE_EP_RXp13 D45 PCIE_EP_RXp5 F28 VSS B33 PCIE_EP_RXp12 D47 VSS F30 VSS B35 PCIE_EP_RXn11 D51 PCIE_EP_RXp3 F32 VSS B37 PCIE_EP_RXn10 D53 TS0_IREF_N F33 VSS B41 PCIE_EP_RXp7 D55 RSVD51 F35 VSS B43 PCIE_EP_RXp6 D57 RSVD55 F36 VSS B45 PCIE_EP_RXn5 E2 RSVD0 F38 VSS B47 PCIE_EP_RXn4 E4 GBE2_LED F40 VSS B51 PCIE_EP_RXn3 E6 VSS F42 VSS B53 VSS E9 RSVD82 F43 VSS VSS E11 RSVD80 F45 VSS B56 Intel(R) Communications Chipset 89xx Series - Datasheet 1652 Ball October 2012 Order Number: 327879-001US Packaging Information Ball Signal Signal Ball Signal F46 VSS J31 PCIE_EP_TXp14 M12 EP_SMBDAT F48 VSS J33 PCIE_EP_TXp12 M15 EP_JTDO F50 VSS J36 PCIE_EP_TXp10 M17 RSVD10 F53 VSS J38 PCIE_EP_TXp9 M20 RSVD15 F55 VSS J41 PCIE_EP_TXn7 M22 RSVD33 F57 VSS J43 PCIE_EP_TXp6 M25 RSVD70 G2 VSS J46 PCIE_EP_TXn5 M27 VSS G4 GBE2_SWDP1 J48 VSS M31 PCIE_EP_TXn15 G54 PCIE_EP_RXn2 J53 VSS M33 VSS G56 PCIE_EP_RXp2 J55 VSS M36 PCIE_EP_TXn11 H6 RSVD84 J57 VSS M38 PCIE_EP_TXp8 H8 RSVD86 K6 VSS M41 RSVD61 H10 GBE3_SWDP1 K8 GBE_AUX_PWR_OK M43 VSS H12 GBE_WAKE# K9 VSS M46 PCIE_EP_TXp1 H15 VSS K46 VSS M48 PCIE_EP_TXp2 H17 RSVD7 K49 PCIE_EP_TXp3 M49 PCIE_EP_TXn2 H20 VSS K50 PCIE_EP_TXn3 M50 VSS H22 RSVD17 K52 VSS M52 VSS H25 RSVD35 L2 SFP3_I2C_CLK M55 PCIE_EP_RXn1 H27 RSVD32 L5 EP_MAIN_PWR_OK M57 VSS H31 PCIE_EP_TXn14 L12 EP_CRU_EN N2 SFP1_I2C_CLK H33 VSS L15 RSVD90 N4 SFP3_I2C_DATA H36 PCIE_EP_TXn10 L17 VSS N6 GBE0_LED H38 PCIE_EP_TXn9 L20 RSVD24 N15 VSS H41 PCIE_EP_TXp7 L22 VSS N52 VSS H43 VSS L25 RSVD69 N54 PCIE_EP_RXn0 H46 PCIE_EP_TXp5 L27 RSVD72 N56 PCIE_EP_RXp0 H48 PCIE_EP_TXp4 L31 VSS P17 RSVD12 H50 PCIE_EP_TXn4 L33 PCIE_EP_TXn12 P20 VSS H52 VSS L36 VSS P22 VCCAEPAUX J1 RSVD83 L38 VSS P25 RSVD71 J3 GBE1_SWDP1 L41 VSS P27 VSS J5 VSS L43 PCIE_EP_TXn6 P31 PCIE_EP_TXp15 J10 GBE_EE_SK L53 PCIE_EP_RXp1 P33 PCIE_EP_TXn13 J12 VSS L56 VSS P36 PCIE_EP_TXp11 J15 RSVD89 M1 VSS P38 PCIE_EP_TXn8 J17 EP_JTMS M3 SRDS0_SD P41 VSS J20 RSVD14 M6 VSS P43 RSVD62 J22 CRU_CLK100N M8 GBE0_SWDP1 R2 GBE1_SWDP0 J25 CRU_CLK100P M9 SRDS3_SD R4 GBE3_SWDP0 J27 VSS M11 VSS R6 GBE2_SWDP0 Intel(R) Communications Chipset 89xx Series - Datasheet 1653 Ball October 2012 Order Number: 327879-001US Packaging Information Ball Signal Signal Ball Signal R8 SFP2_I2C_CLK U22 VCCAEP1P8_CRU Y20 VSS R9 VSS U25 VCCEPAUX Y22 VCCEPAUX R11 GBE1_LED U27 VSS Y24 VSS R12 GBE_EE_CS# U31 VCCEP Y26 VSS R14 RSVD88 U33 VSS Y28 VCCEP R15 RSVD36 U36 VCCEP Y30 VCCEP R17 VSS U38 PCIE_EP_ICOMPI Y32 VSS R20 RSVD20 U41 VSS Y34 VCCEP R22 VCCAEP1P8_CRU U43 PCIE_EP_CLK100N Y36 VSS R25 VSS U44 VSS Y38 VCCAEP_PE R27 VCCAEPAUX U46 VSS Y40 VSS R31 VSS U47 DMI_TXn0 Y41 VCCAEP_PE R33 PCIE_EP_TXp13 U49 DMI_TXp0 Y43 VCCAEP1P8_PE R36 VSS U50 VSS Y44 PCIE_EP_VREF1P8 R38 VSS U53 VSS Y46 VSS R41 PCIE_EP_CLK100P U56 DMI_RXn0 Y47 DMI_TXn1 R43 VSS V6 VSS Y49 DMI_TXp1 R44 VSS V20 VSS Y50 VSS R46 PCIE_EP_TXn1 V22 VCCEPAUX Y52 VSS R47 VSS V25 VSS AA2 SRDSO_1_P R49 PCIE_EP_TXn0 V27 VCCEP AA5 SRDSO_0_N R50 PCIE_EP_TXp0 V31 VSS AA20 VSS R52 VSS V33 VCCEP AA22 VCCEPAUX R54 VSS V36 VSS AA24 VSS R56 VSS V38 VSS AA26 VSS T1 VSS V52 VSS AA28 VCCEP T3 SFP0_I2C_CLK W1 VSS AA30 VCCEP T6 VSS W3 SRDSO_0_P AA32 VSS T52 VSS W5 VSS AA34 VCCEP T55 DMI_RXp0 W53 DMI_RXp1 AA36 VSS T57 VSS W55 DMI_RXn1 AA38 VCCAEP_PE U2 SFP0_I2C_DATA W57 VSS AA53 VSS U5 VSS Y6 VSS AA56 DMI_RXp2 U8 SFP2_I2C_DATA Y8 SFP1_I2C_DATA AB1 VSS U9 GBE3_LED Y9 GBE0_SWDP0 AB3 SRDSO_1_N U11 GBE_SMBDAT Y11 VSS AB6 VSS U12 GBE_SMBALRT# Y12 GBE_SMBCLK AB8 RSVD64 U14 VSS Y14 RSVD87 AB9 VSS U15 RSVD37 Y15 EP_JTDI AB11 SRDSO_3_P U17 VCCAEPAUX Y17 VSS AB12 SRDSO_3_N U20 VCCAEPAUX Y18 VSS AB14 VSS Intel(R) Communications Chipset 89xx Series - Datasheet 1654 Ball October 2012 Order Number: 327879-001US Packaging Information Ball Signal Ball AB15 VCCPEP3P3AUX AE20 VSS AG28 VSS AB17 VCCEPAUX AE22 VCCEPAUX AG30 VCCEP AB18 VSS AE24 VCCEPAUX AG32 VSS AB40 VSS AE26 VSS AG34 VCCEP AB41 VCCAEP_PE AE28 VCCEP AG36 VSS AB43 VCCAEP_PE AE30 VSS AG38 VCCAEP_PE AB44 VSS AE32 VCCEP AG40 VSS AB46 RSVD42 AE34 VSS AG41 VCCAEP_PE AB47 VSS AE36 VCCAEP_PE AG43 VCCAEP_PE AB49 DMI_TXp2 AE38 VSS AG44 VCCA1P8_DMI AB50 DMI_TXn2 AE40 VCCAEP_PE AG46 DMI_VREF AB52 VSS AE41 VSS AG47 VSS AB55 DMI_RXn2 AE43 VCCAEP_PE AG49 PCIE_RC_TXp0 AB57 VSS AE44 VSS AG50 PCIE_RC_TXn0 AC2 SRDSO_2_P AE46 RSVD41 AG53 PCIE_RC_RXp0 AC4 SRDSO_2_N AE47 DMI_IRCOMP AG56 PCIE_RC_Rxn0 AC6 VSS AE49 VSS AH6 VSS AC20 VCCEPAUX AE50 DMI_TXp3 AH52 VSS AC22 VSS AE52 VSS AJ1 VSS AC24 VCCEP1P0_CRU AE54 DMI_CLK100P AJ3 SRDSI_1_P AC26 VCCEP1P0_CRU AE56 DMI_CLK100N AJ5 SRDSI_0_N AC28 VSS AF1 GBE_CLK100P AJ20 VSS AC30 VCCEP AF3 GBE_CLK100N AJ22 VCCEPAUX AC32 VSS AF6 VSS AJ24 VCCEPAUX AC34 VCCEP AF52 DMI_TXn3 AJ26 VSS AC36 VSS AF55 VSS AJ28 VCCEP AC38 VCCAEP_PE AF57 VSS AJ30 VSS AC52 VSS AG2 VSS AJ32 VCCEP AC54 DMI_RXn3 AG5 SRDSI_0_P AJ34 VSS AC56 DMI_RXp3 AG8 GBE_IRCOMP AJ36 VCCEP AE2 VSS AG9 VCCAEP1P8AUX AJ38 VSS AE4 VSS AG11 VCCAEP1P8AUX AJ53 PCIE_RC_RXp1 AE6 VSS AG12 VSS AJ55 PCIE_RC_Rxn1 AE8 RSVD63 AG14 VCCAEPAUX AJ57 VSS AE9 VSS AG15 VSS AK6 VSS AE11 GBE_VREF1P8 AG17 VCCPEP3P3AUX AK52 VSS AE12 VSS AG18 VSS AL2 SRDSI_2_P AE14 VCCPEP3P3AUX AG20 VCCEPAUX AL5 SRDSI_1_N AE15 VSS AG22 VSS AL8 VCCAEP1P8AUX AE17 VCCPEP3P3AUX AG24 VCCEPAUX AL9 VSS AE18 VSS AG26 VCCEP AL11 VCCAEPAUX Intel(R) Communications Chipset 89xx Series - Datasheet 1655 Signal Ball Signal October 2012 Order Number: 327879-001US Packaging Information Ball Signal Signal Ball Signal AL12 VCCAEPAUX AN18 VSS AT8 VCCSUS3P3_USB AL14 VSS AN20 VSS AT9 VSSA_USB AL15 VCCAEPAUX AN22 VSS AT11 VCCASUS1P8_USB AL17 VSS AN24 VCC AT12 VSS AL18 VCCAEPAUX AN26 VSS AT14 VCCSUS1P8_USB AL20 VCCEPAUX AN28 VCC AT15 VCCSUS1P8_USB AL22 VSS AN30 VSS AT17 VCCSUS1P8_USB AL24 VSS AN32 VCC AT18 VSS AL26 VCCEP AN34 VCCA1P8_TS0 AT40 VCCA_DMI AL28 VSS AN36 VSS AT41 VSS AL30 VCCEP AN38 VCCA_DMI AT43 VCCA_DMI AL32 VSS AN40 VSS AT44 SATA_ICOMPO AL34 VCCEP AN41 VCCA_DMI AT46 VSS AL36 VSS AN43 VSS AT47 SATA_VREF1P8 AL38 VCCA_DMI AN44 VCCA_DMI AT49 VSS AL40 VSS AN46 PCIE_RC_TXp2 AT50 PCIE_RC_TXn3 AL41 VCCA_DMI AN47 PCIE_RC_TXn2 AT52 VSS AL43 VSS AN49 VSS AT55 SATA5_RXP AL44 VCCA_DMI AN50 PCIE_RC_TXp3 AT57 SATA4_RXN AL46 VSS AN52 VSS AU2 USB_RBIASp AL47 PCIE_RC_TXp1 AN54 PCIE_RC_RXn3 AU5 USB_RBIASn AL49 PCIE_RC_TXn1 AN56 PCIE_RC_RXp3 AU20 VSS AL50 VSS AR2 RSVD46 AU22 VCCSUS AL53 VSS AR4 VSS AU24 VSS AL56 PCIE_RC_RXp2 AR6 VSS AU26 VCC AM1 VSS AR20 VSS AU28 VSS AM3 SRDSI_2_N AR22 VCCSUS AU30 VCC AM6 VSS AR24 VSS AU32 VSS AM52 VSS AR26 VCC AU34 VCC AM55 PCIE_RC_RXn2 AR28 VSS AU36 VSS AM57 VSS AR30 VCC AU38 VSS AN2 SRDSI_3_P AR32 VCC AU53 SATA5_RXN AN4 SRDSI_3_N AR34 VCC AU56 VSS AN6 VSS AR36 VCC AV6 VSS AN8 VCC1P05 AR38 VCCA_DMI AV8 VCCSUS_USB AN9 VCC1P05 AR52 VSS AV9 VCCSUS3P3_USB AN11 VCCA1P8_TS1 AR54 VSS AV11 VCCSUS3P3_USB AN12 VCCSUS1P8_USB AR56 SATA4_RXP AV12 VCCSUS3P3_USB AN14 VSS AT1 VSS AV14 VCCSUS1P8 AN15 VCCA1P5_TS1 AT3 RSVD45 AV15 VSS AN17 VCCA1P5_TS0 AT6 VSS AV17 VCCSUS1P8 Intel(R) Communications Chipset 89xx Series - Datasheet 1656 Ball October 2012 Order Number: 327879-001US Packaging Information Ball Signal Signal Ball Signal AV18 VSS BA12 VCCSUS3P3_USB BC31 VSS AV20 VCCSUS BA14 VSS BC33 VCC1P8 AV22 VSS BA15 VCCSUS3P3 BC36 VCC1P8 AV24 VCCSUS BA17 VCCSUS1P8 BC38 VSS AV26 VCC3P3 BA20 VSS BC41 VCCPCPU AV28 VSS BA22 VCCSUS BC43 VCCA_SATA AV30 VCC3P3 BA25 VCC3P3 BC44 VSS AV32 VSS BA27 VSS BC46 AV34 VCC3P3 BA31 VCC3P3 SATA5_GP/ TEMP_ALERT#/GPIO49 AV36 VCC3P3 BA33 VSS BC47 SATA4_GP/GPIO16 AV38 VCC3P3 BA36 VCC3P3 BC49 RSVD47 AV40 VSS BA38 VCC3P3 BC50 GPIO20 AV41 VCCA_SATA BA41 VSS BC52 VSS AV43 VSS BA43 VCCA_SATA AV44 VCCA_SATA BA44 RSVD44 AV46 VCCA1P8_SATA BA46 RSVD43 AV47 SATA4_TXP BA47 VSS AV49 SATA4_TXN BA49 SATA5_TXP AV50 VSS BA50 SATA5_TXN AV52 VSS BA53 PMSYNC AW1 VSS BA56 VSS AW3 USB_CLK96P BB1 VSS AW5 USB_CLK96N BB3 USB1_Dn AW53 SATA_CLK100N BB6 VSS AW55 SATA_CLK100P BB52 VSS AW57 VSS BB55 PECI AY6 VSS BB57 CPUPWRGD AY20 VSS BC2 USB1_Dp AY22 VSS BC4 USB2_Dn AY25 VSS BC6 VSS AY27 VCC3P3 BC8 VCCA1P8_USB AY31 VSS BC9 VSS AY33 VCC3P3 BC11 VCCA_USB AY36 VSS BC12 VCCA_USB AY38 VSS BC14 VCCSUS3P3 AY52 VSS BC15 VCCSUS3P3 BA2 USB0_Dn BC17 VSS BA5 USB0_Dp BC20 VCCSUS BA8 VCCSUS_USB BC22 VSS BA9 VCCA1P8_USB BC25 VCC3P3 BA11 VSS BC27 VCC1P8 Intel(R) Communications Chipset 89xx Series - Datasheet 1657 Ball BC54 SPI_MOSI BC56 SPI_CS1# BD15 VSS BD17 VSS BD20 VCCSUS BD22 VCCSUS BD25 VSS BD27 VSS BD31 VCC1P8 BD33 VSS BD36 VSS BD38 SATA_LED# BD41 SPI_CLK BD43 VSS BE2 USB3_Dn BE4 USB2_Dp BE6 VSS BE52 VSS BE54 RSVD49 BE56 RSVD48 BF1 VSS BF3 USB3_Dp BF6 VSS BF8 USB5_Dn BF9 USB5_Dp BF11 VSS BF12 VSS BF15 GPIO12 BF17 VCCSUS3P3 October 2012 Order Number: 327879-001US Packaging Information Ball Signal Signal Ball Signal BF20 SLP_S5#/GPIO63 BJ1 USB4_Dp BK52 VSS BF22 VSS BJ3 VSS BL2 MST_SMBCLK BF25 VCCSUS3P3_RTC BJ5 VSS BL4 MST_SMBDAT BF27 VCC3P3_RTC BJ10 GPIO15 BL54 A20GATE BF31 VSS BJ12 VSS BL56 SERIRQ BF33 LFRAME# BJ15 RSVD78 BM1 RSVD66 BF36 GPIO52 BJ17 VSS BM3 VSS BF38 GPIO17 BJ20 SML1CLK/GPIO58 BM5 RSVD73 BF41 VSS BJ22 VSS BM8 VSS BF43 GPIO19 BJ25 GPIO13 BM10 GPIO30 BF46 GPIO35 BJ27 RTCRST# BM12 VSS BF47 RCIN# BJ31 SIU0_RXD BM13 SUS_STAT#/GPIO61 BF49 GPIO21 BJ33 LAD1 BM15 VSS BF50 VSS BJ36 LDRQ0# BM16 VSS BF52 THRMTRIP# BJ38 PCICLK BM18 OC3#/GPIO42 BF55 RSVD50 BJ41 BBS0 BM20 GPIO46 BF57 VSS BJ43 GPIO4 BM22 VSS BG2 USB4_Dn BJ46 RSVD75 BM23 VSS BG5 VSS BJ48 SYS_RESET# BM25 INTRUDER# BG12 PLTRST# BJ53 REF_CLK14 BM26 GPIO26 BG15 MST_SMBALERT#/ GPIO11 BJ55 SPI_CS0# BM28 VSS BJ57 VSS BM30 JTDI BK6 VSS BM32 VSS BK8 VSS BM33 SIU0_RTS# BK10 OC2#/GPIO41 BM35 VSS BK12 OC0#/GPIO59 BM36 SIU1_RXD BK15 GPIO31 BM38 BBS1/GPIO51 BK17 RI# BM40 VSS BK20 SML1DAT/GPIO75 BM42 GPIO53 BK22 GPIO43 BM43 VSS BK25 GPIO25 BM45 GPIO5 BK27 PWROK BM46 VSS BK31 SRTCRST# BM48 VSS BK33 SIU0_RI# BM50 VSS BK36 SIU1_DCD# BM53 NRBOOTS BK38 VSS BM55 VSS BK41 SIU1_TXD BM57 VSS BK43 GPIO2 BN2 RSVD68 BK46 RSVD76 BN4 OC1#/GPIO40 BK48 VSS BN6 VSS BK50 INIT3_3V# BN9 VSS BG17 GPIO28 BG20 PWRBTN# BG22 GPIO10 BG25 RSVD39 BG27 VSS BG31 SIU0_DCD# BG33 LDRQ1#/GPIO23 BG36 VSS BG38 GPIO32 BG41 VSS BG43 GPIO50 BG46 VSS BG53 ADR/GPIO37 BG56 GPIO18 BH6 VSS BH8 RSVD74 BH9 TS1_IREF_N BH49 VSS BH50 SPI_MISO BH52 VSS Intel(R) Communications Chipset 89xx Series - Datasheet 1658 Ball October 2012 Order Number: 327879-001US Packaging Information Ball Signal Signal Ball Signal BN11 GPIO57 BR36 VSS BU42 VSS BN17 SLP_S4# BR39 LAD0 BU46 VSS BN19 SLP_S3# BR42 GPIO3 BU49 SLOAD/GPIO38 BN21 GPIO9 BR46 GPIO54 BU52 VSS BN27 RSMRST# BR49 SCLOCK/GPIO22 BU54 VSS BN29 VSS BR52 SDATAOUT0/GPIO39 BN31 JTMS BR54 VSS BN37 SIU1_RI# BT2 VSS BN39 UART_CLK BT5 RSVD57 BN41 GPIO33 BT7 VSS BN47 GPIO1 BT11 VSS BN49 GPIO34 BT13 GPIO72 BN52 SIU1_DTR# BT15 VSS BN54 GPIO36 BT17 DRAMPWRGD BN56 VSS BT21 GPIO45 BP1 RSVD65 BT23 GPIO44 BP3 RSVD67 BT25 RSVD53 BP5 RSVD60 BT27 VSS BP7 GPIO24 BT31 RSVD38 BP13 SUS_CLK/GPIO62 BT33 RSVD77 BP15 SYS_PWROK BT35 SIU1_CTS# BP23 GPIO14 BT37 WDT_TOUT# BP25 IVCC_RTC BT41 LAD2 BP33 SIU0_CTS# BT43 SIU0_TXD BP35 SIU1_DSR# BT45 BMBUSY#/GPIO0 BP43 SIU0_DTR# BT47 GPIO7 BP45 GPIO55 BT51 SDATAOUT1/GPIO48 BP51 VSS BT53 VSS BP53 GPIO6 BT56 VSS BP55 VSS BU4 RSVD59 BP57 VSS BU6 RSVD56 BR4 VSS BU9 GPIO47 BR6 GPIO27 BU12 SML1ALERT#/GPIO74 BR9 GPIO56 BU16 GPIO8 BR12 GPIO60 BU19 WAKE# BR16 MEPWROK BU22 VSS BR19 VSS BU26 RTCX2 BR22 GPIO73 BU29 JTDO BR26 RTCX1 BU32 SIU0_DSR# BR29 JTCK BU36 SIU1_RTS# BR32 VSS BU39 LAD3 Intel(R) Communications Chipset 89xx Series - Datasheet 1659 Ball October 2012 Order Number: 327879-001US Packaging Information Table 35-2. Alphabetical Signal Listing Signal Ball Signal Ball A20GATE BL54 GBE_CLK100N AF3 GPIO21 BF49 GPIO4 BJ43 GBE_CLK100P AF1 GPIO24 BP7 BBS0 BJ41 GBE_EE_CS# R12 GPIO25 BK25 BBS1/GPIO51 BM38 GBE_EE_DI D5 GPIO26 BM26 BMBUSY#/GPIO0 BT45 GBE_EE_DO F8 GPIO27 BR6 CPUPWRGD BB57 GBE_EE_SK J10 GPIO28 BG17 CRU_CLK100N J22 GBE_IRCOMP AG8 GPIO30 BM10 CRU_CLK100P J25 GBE_SMBALRT# U12 GPIO31 BK15 DMI_CLK100N AE56 GBE_SMBCLK Y12 GPIO32 BG38 DMI_CLK100P AE54 GBE_SMBDAT U11 GPIO33 BN41 DMI_IRCOMP AE47 GBE_VREF1P8 AE11 GPIO34 BN49 DMI_RXn0 U56 GBE_WAKE# H12 GPIO35 BF46 DMI_RXn1 W55 GBE0_LED N6 GPIO36 BN54 DMI_RXn2 AB55 GBE0_SWDP0 Y9 ADR/GPIO37 BG53 DMI_RXn3 AC54 GBE0_SWDP1 M8 GPIO43 BK22 DMI_RXp0 T55 GBE1_LED R11 GPIO44 BT23 DMI_RXp1 W53 GBE1_SWDP0 R2 GPIO45 BT21 DMI_RXp2 AA56 GBE1_SWDP1 J3 GPIO46 BM20 DMI_RXp3 AC56 GBE2_LED E4 GPIO47 BU9 DMI_TXn0 U47 GBE2_SWDP0 R6 GPIO50 BG43 DMI_TXn1 Y47 GBE2_SWDP1 G4 GPIO52 BF36 DMI_TXn2 AB50 GBE3_LED U9 GPIO53 BM42 DMI_TXn3 AF52 GBE3_SWDP0 R4 GPIO54 BR46 DMI_TXp0 U49 GBE3_SWDP1 H10 GPIO55 BP45 DMI_TXp1 Y49 GPIO1 BN47 GPIO56 BR9 DMI_TXp2 AB49 GPIO2 BK43 GPIO57 BN11 DMI_TXp3 AE50 GPIO3 BR42 GPIO60 BR12 DMI_VREF AG46 GPIO5 BM45 GPIO72 BT13 DRAMPWRGD BT17 GPIO6 BP53 GPIO73 BR22 EP_CRU_EN L12 GPIO7 BT47 INIT3_3V# BK50 EP_JTCK A12 GPIO8 BU16 INTRUDER# BM25 EP_JTDI Y15 GPIO9 BN21 IVCC_RTC BP25 EP_JTDO M15 GPIO10 BG22 JTCK BR29 EP_JTMS J17 GPIO12 BF15 JTDI BM30 EP_JTRST# D13 GPIO13 BJ25 JTDO BU29 EP_MAIN_PWR_OK L5 GPIO14 BP23 JTMS BN31 EP_SMBALRT# C12 GPIO15 BJ10 LAD0 BR39 EP_SMBCLK B13 GPIO17 BF38 LAD1 BJ33 EP_SMBDAT M12 GPIO18 BG56 LAD2 BT41 GBE_AUX_PWR_AVAIL B7 GPIO19 BF43 LAD3 BU39 K8 GPIO20 BC50 LDRQ0# BJ36 GBE_AUX_PWR_OK Intel(R) Communications Chipset 89xx Series - Datasheet 1660 Signal Ball October 2012 Order Number: 327879-001US Packaging Information Signal Ball Ball Signal Ball LDRQ1#/GPIO23 BG33 PCIE_EP_RXp9 E37 PCIE_RC_Rxn1 AJ55 LFRAME# BF33 PCIE_EP_RXp10 C36 PCIE_RC_RXn2 AM55 MEPWROK BR16 PCIE_EP_RXp11 D35 PCIE_RC_RXn3 AN54 PCIE_EP_RXp12 B33 PCIE_RC_RXp0 AG53 PCIE_EP_RXp13 B31 PCIE_RC_RXp1 AJ53 PCIE_EP_RXp14 C29 PCIE_RC_RXp2 AL56 PCIE_EP_RXp15 B27 PCIE_RC_RXp3 AN56 PCIE_EP_TXn0 R49 PCIE_RC_TXn0 AG50 PCIE_EP_TXn1 R46 PCIE_RC_TXn1 AL49 PCIE_EP_TXn2 M49 PCIE_RC_TXn2 AN47 PCIE_EP_TXn3 K50 PCIE_RC_TXn3 AT50 PCIE_EP_TXn4 H50 PCIE_RC_TXp0 AG49 PCIE_EP_TXn5 J46 PCIE_RC_TXp1 AL47 PCIE_EP_TXn6 L43 PCIE_RC_TXp2 AN46 PCIE_EP_TXn7 J41 PCIE_RC_TXp3 AN50 PCIE_EP_TXn8 P38 PECI BB55 PCIE_EP_TXn9 H38 PLTRST# BG12 PCIE_EP_TXn10 H36 PMSYNC BA53 PCIE_EP_TXn11 M36 PWRBTN# BG20 PCIE_EP_TXn12 L33 PWROK BK27 PCIE_EP_TXn13 P33 RCIN# BF47 PCIE_EP_TXn14 H31 REF_CLK14 BJ53 PCIE_EP_TXn15 M31 RI# BK17 PCIE_EP_TXp0 R50 RSMRST# BN27 PCIE_EP_TXp1 M46 RSVD0 E2 PCIE_EP_TXp2 M48 RSVD1 B5 PCIE_EP_TXp3 K49 RSVD2 B2 PCIE_EP_TXp4 H48 RSVD3 F1 PCIE_EP_TXp5 H46 RSVD4 D1 PCIE_EP_TXp6 J43 RSVD5 B15 PCIE_EP_TXp7 H41 RSVD6 C4 PCIE_EP_TXp8 M38 RSVD7 H17 PCIE_EP_TXp9 J38 RSVD8 A6 PCIE_EP_TXp10 J36 RSVD9 F18 PCIE_EP_TXp11 P36 RSVD10 M17 PCIE_EP_TXp12 J33 RSVD11 B17 PCIE_EP_TXp13 R33 RSVD12 P17 PCIE_EP_TXp14 J31 RSVD13 C16 PCIE_EP_TXp15 P31 RSVD14 J20 PCIE_EP_VREF1P8 Y44 RSVD15 M20 PCIE_RC_Rxn0 AG56 RSVD16 F13 MST_SMBALERT#/ GPIO11 BG15 MST_SMBCLK BL2 MST_SMBDAT BL4 NRBOOTS BM53 OC0#/GPIO59 BK12 OC1#/GPIO40 BN4 OC2#/GPIO41 BK10 OC3#/GPIO42 BM18 PCICLK BJ38 PCIE_EP_CLK100N U43 PCIE_EP_CLK100P R41 PCIE_EP_ICOMPI U38 PCIE_EP_RST# F10 PCIE_EP_RXn0 N54 PCIE_EP_RXn1 M55 PCIE_EP_RXn2 G54 PCIE_EP_RXn3 B51 PCIE_EP_RXn4 B47 PCIE_EP_RXn5 B45 PCIE_EP_RXn6 D43 PCIE_EP_RXn7 C42 PCIE_EP_RXn8 E41 PCIE_EP_RXn9 C39 PCIE_EP_RXn10 B37 PCIE_EP_RXn11 B35 PCIE_EP_RXn12 D33 PCIE_EP_RXn13 E31 PCIE_EP_RXn14 E29 PCIE_EP_RXn15 E27 PCIE_EP_RXp0 N56 PCIE_EP_RXp1 L53 PCIE_EP_RXp2 G56 PCIE_EP_RXp3 D51 PCIE_EP_RXp4 C46 PCIE_EP_RXp5 D45 PCIE_EP_RXp6 B43 PCIE_EP_RXp7 B41 PCIE_EP_RXp8 E39 Intel(R) Communications Chipset 89xx Series - Datasheet 1661 Signal October 2012 Order Number: 327879-001US Packaging Information Signal Ball Signal Ball Ball RSVD17 H22 RSVD60 BP5 SATA4_RXP AR56 RSVD18 F22 RSVD61 M41 SATA4_TXN AV49 RSVD19 F23 RSVD62 P43 SATA4_TXP AV47 RSVD20 R20 RSVD63 AE8 BC46 RSVD21 B21 RSVD64 AB8 SATA5_GP/ TEMP_ALERT#/GPIO49 RSVD22 D15 RSVD65 BP1 SATA5_RXN AU53 RSVD23 D23 RSVD66 BM1 SATA5_RXP AT55 RSVD24 L20 RSVD67 BP3 SATA5_TXN BA50 RSVD25 B23 RSVD68 BN2 SATA5_TXP BA49 RSVD26 E17 RSVD69 L25 SCLOCK/GPIO22 BR49 RSVD27 A22 RSVD70 M25 SDATAOUT0/GPIO39 BR52 RSVD28 C19 RSVD71 P25 SDATAOUT1/GPIO48 BT51 RSVD29 B25 RSVD72 L27 SERIRQ BL56 RSVD30 F20 RSVD73 BM5 RSVD31 D25 RSVD74 BH8 RSVD32 H27 RSVD75 BJ46 RSVD33 M22 RSVD76 BK46 RSVD34 E21 RSVD77 BT33 RSVD35 H25 RSVD78 BJ15 RSVD36 R15 RSVD79 C9 RSVD37 U15 RSVD80 E11 RSVD38 BT31 RSVD81 B11 RSVD39 BG25 RSVD82 E9 RSVD41 AE46 RSVD83 J1 RSVD42 AB46 RSVD84 H6 RSVD43 BA46 RSVD85 F5 RSVD44 BA44 RSVD86 H8 RSVD45 AT3 RSVD87 Y14 RSVD46 AR2 RSVD88 R14 RSVD47 BC49 RSVD89 J15 RSVD48 BE56 RSVD90 L15 RSVD49 BE54 RTCRST# BJ27 RSVD50 BF55 RTCX1 BR26 RSVD51 D55 RTCX2 BU26 RSVD52 C54 SATA_CLK100N AW53 RSVD53 BT25 SATA_CLK100P AW55 RSVD54 E56 SATA_ICOMPO AT44 RSVD55 D57 SATA_LED# BD38 RSVD56 BU6 SATA_VREF1P8 AT47 RSVD57 BT5 SATA4_GP/GPIO16 BC47 RSVD59 BU4 SATA4_RXN AT57 Intel(R) Communications Chipset 89xx Series - Datasheet 1662 Signal SFP0_I2C_CLK T3 SFP0_I2C_DATA U2 SFP1_I2C_CLK N2 SFP1_I2C_DATA Y8 SFP2_I2C_CLK R8 SFP2_I2C_DATA U8 SFP3_I2C_CLK L2 SFP3_I2C_DATA N4 SIU0_CTS# BP33 SIU0_DCD# BG31 SIU0_DSR# BU32 SIU0_DTR# BP43 SIU0_RI# BK33 SIU0_RTS# BM33 SIU0_RXD BJ31 SIU0_TXD BT43 SIU1_CTS# BT35 SIU1_DCD# BK36 SIU1_DSR# BP35 SIU1_DTR# BN52 SIU1_RI# BN37 SIU1_RTS# BU36 SIU1_RXD BM36 SIU1_TXD BK41 SLOAD/GPIO38 BU49 SLP_S3# BN19 SLP_S4# BN17 SLP_S5#/GPIO63 BF20 SML1ALERT#/GPIO74 BU12 October 2012 Order Number: 327879-001US Packaging Information Signal Ball Ball Signal Ball SML1CLK/GPIO58 BJ20 USB0_Dp BA5 VCCA_DMI AL38 SML1DAT/GPIO75 BK20 USB1_Dn BB3 VCCA_DMI AL41 SPI_CLK BD41 USB1_Dp BC2 VCCA_DMI AL44 SPI_CS0# BJ55 USB2_Dn BC4 VCCA_DMI AN38 SPI_CS1# BC56 USB2_Dp BE4 VCCA_DMI AN41 SPI_MISO BH50 USB3_Dn BE2 VCCA_DMI AN44 SPI_MOSI BC54 USB3_Dp BF3 VCCA_DMI AR38 SRDS0_SD M3 USB4_Dn BG2 VCCA_DMI AT40 SRDS1_SD C6 USB4_Dp BJ1 VCCA_DMI AT43 SRDS2_SD F3 USB5_Dn BF8 VCCA_SATA AV41 SRDS3_SD M9 USB5_Dp BF9 VCCA_SATA AV44 SRDSI_0_N AJ5 VCC AN24 VCCA_SATA BA43 SRDSI_0_P AG5 VCC AN28 VCCA_SATA BC43 SRDSI_1_N AL5 VCC AN32 VCCA_USB BC11 SRDSI_1_P AJ3 VCC AR26 VCCA_USB BC12 SRDSI_2_N AM3 VCC AR30 VCCA1P5_TS0 AN17 SRDSI_2_P AL2 VCC AR32 VCCA1P5_TS1 AN15 SRDSI_3_N AN4 VCC AR34 VCCA1P8_DMI AG44 SRDSI_3_P AN2 VCC AR36 VCCA1P8_SATA AV46 SRDSO_0_N AA5 VCC AU26 VCCA1P8_TS0 AN34 SRDSO_0_P W3 VCC AU30 VCCA1P8_TS1 AN11 SRDSO_1_N AB3 VCC AU34 VCCA1P8_USB BA9 SRDSO_1_P AA2 VCC1P05 AN8 VCCA1P8_USB BC8 SRDSO_2_N AC4 VCC1P05 AN9 VCCAEP_PE Y38 SRDSO_2_P AC2 VCC1P8 BC27 VCCAEP_PE Y41 SRDSO_3_N AB12 VCC1P8 BC33 VCCAEP_PE AA38 SRDSO_3_P AB11 VCC1P8 BC36 VCCAEP_PE AB41 SRTCRST# BK31 VCC1P8 BD31 VCCAEP_PE AB43 SUS_CLK/GPIO62 BP13 VCC3P3 AV26 VCCAEP_PE AC38 SUS_STAT#/GPIO61 BM13 VCC3P3 AV30 VCCAEP_PE AE36 SYS_PWROK BP15 VCC3P3 AV34 VCCAEP_PE AE40 SYS_RESET# BJ48 VCC3P3 AV36 VCCAEP_PE AE43 THRMTRIP# BF52 VCC3P3 AV38 VCCAEP_PE AG38 TS0_IREF_N D53 VCC3P3 AY27 VCCAEP_PE AG41 TS1_IREF_N BH9 VCC3P3 AY33 VCCAEP_PE AG43 UART_CLK BN39 VCC3P3 BA25 VCCAEP1P8_CRU R22 USB_CLK96N AW5 VCC3P3 BA31 VCCAEP1P8_CRU U22 USB_CLK96P AW3 VCC3P3 BA36 VCCAEP1P8_PE Y43 USB_RBIASn AU5 VCC3P3 BA38 VCCAEP1P8AUX AG9 USB_RBIASp AU2 VCC3P3 BC25 VCCAEP1P8AUX AG11 USB0_Dn BA2 VCC3P3_RTC BF27 VCCAEP1P8AUX AL8 Intel(R) Communications Chipset 89xx Series - Datasheet 1663 Signal October 2012 Order Number: 327879-001US Packaging Information Signal Ball Ball Signal Ball VCCAEPAUX P22 VCCEPAUX AE22 VSS A16 VCCAEPAUX R27 VCCEPAUX AE24 VSS A19 VCCAEPAUX U17 VCCEPAUX AG20 VSS A26 VCCAEPAUX U20 VCCEPAUX AG24 VSS A29 VCCAEPAUX AG14 VCCEPAUX AJ22 VSS A32 VCCAEPAUX AL11 VCCEPAUX AJ24 VSS A36 VCCAEPAUX AL12 VCCEPAUX AL20 VSS A39 VCCAEPAUX AL15 VCCPCPU BC41 VSS A42 VCCAEPAUX AL18 VCCPEP3P3AUX AB15 VSS A46 VCCASUS1P8_USB AT11 VCCPEP3P3AUX AE14 VSS A49 VCCEP U31 VCCPEP3P3AUX AE17 VSS A52 VCCEP U36 VCCPEP3P3AUX AG17 VSS A54 VCCEP V27 VCCSUS AR22 VSS B53 VCCEP V33 VCCSUS AU22 VSS B56 VCCEP Y28 VCCSUS AV20 VSS C22 VCCEP Y30 VCCSUS AV24 VSS C26 VCCEP Y34 VCCSUS BA22 VSS C32 VCCEP AA28 VCCSUS BC20 VSS C49 VCCEP AA30 VCCSUS BD20 VSS C52 VCCEP AA34 VCCSUS BD22 VSS D3 VCCEP AC30 VCCSUS_USB AV8 VSS D7 VCCEP AC34 VCCSUS_USB BA8 VSS D47 VCCEP AE28 VCCSUS1P8 AV14 VSS E6 VCCEP AE32 VCCSUS1P8 AV17 VSS E19 VCCEP AG26 VCCSUS1P8 BA17 VSS E49 VCCEP AG30 VCCSUS1P8_USB AN12 VSS E52 VCCEP AG34 VCCSUS1P8_USB AT14 VSS E54 VCCEP AJ28 VCCSUS1P8_USB AT15 VSS F12 VCCEP AJ32 VCCSUS1P8_USB AT17 VSS F15 VCCEP AJ36 VCCSUS3P3 BA15 VSS F16 VCCEP AL26 VCCSUS3P3 BC14 VSS F25 VCCEP AL30 VCCSUS3P3 BC15 VSS F26 VCCEP AL34 VCCSUS3P3 BF17 VSS F28 VCCEP1P0_CRU AC24 VCCSUS3P3_RTC BF25 VSS F30 VCCEP1P0_CRU AC26 VCCSUS3P3_USB AT8 VSS F32 VCCEPAUX U25 VCCSUS3P3_USB AV9 VSS F33 VCCEPAUX V22 VCCSUS3P3_USB AV11 VSS F35 VCCEPAUX Y22 VCCSUS3P3_USB AV12 VSS F36 VCCEPAUX AA22 VCCSUS3P3_USB BA12 VSS F38 VCCEPAUX AB17 VSS A4 VSS F40 VCCEPAUX AC20 VSS A9 VSS F42 Intel(R) Communications Chipset 89xx Series - Datasheet 1664 Signal October 2012 Order Number: 327879-001US Packaging Information Signal Ball Ball Signal Ball VSS F43 VSS N15 VSS Y11 VSS F45 VSS N52 VSS Y17 VSS F46 VSS P20 VSS Y18 VSS F48 VSS P27 VSS Y20 VSS F50 VSS P41 VSS Y24 VSS F53 VSS R9 VSS Y26 VSS F55 VSS R17 VSS Y32 VSS F57 VSS R25 VSS Y36 VSS G2 VSS R31 VSS Y40 VSS H15 VSS R36 VSS Y46 VSS H20 VSS R38 VSS Y50 VSS H33 VSS R43 VSS Y52 VSS H43 VSS R44 VSS AA20 VSS H52 VSS R47 VSS AA24 VSS J5 VSS R52 VSS AA26 VSS J12 VSS R54 VSS AA32 VSS J27 VSS R56 VSS AA36 VSS J48 VSS T1 VSS AA53 VSS J53 VSS T6 VSS AB1 VSS J55 VSS T52 VSS AB6 VSS J57 VSS T57 VSS AB9 VSS K6 VSS U5 VSS AB14 VSS K9 VSS U14 VSS AB18 VSS K46 VSS U27 VSS AB40 VSS K52 VSS U33 VSS AB44 VSS L17 VSS U41 VSS AB47 VSS L22 VSS U44 VSS AB52 VSS L31 VSS U46 VSS AB57 VSS L36 VSS U50 VSS AC6 VSS L38 VSS U53 VSS AC22 VSS L41 VSS V6 VSS AC28 VSS L56 VSS V20 VSS AC32 VSS M1 VSS V25 VSS AC36 VSS M6 VSS V31 VSS AC52 VSS M11 VSS V36 VSS AE2 VSS M27 VSS V38 VSS AE4 VSS M33 VSS V52 VSS AE6 VSS M43 VSS W1 VSS AE9 VSS M50 VSS W5 VSS AE12 VSS M52 VSS W57 VSS AE15 VSS M57 VSS Y6 VSS AE18 Intel(R) Communications Chipset 89xx Series - Datasheet 1665 Signal October 2012 Order Number: 327879-001US Packaging Information Signal Ball Ball Signal Ball VSS AE20 VSS AL40 VSS AU38 VSS AE26 VSS AL43 VSS AU56 VSS AE30 VSS AL46 VSS AV6 VSS AE34 VSS AL50 VSS AV15 VSS AE38 VSS AL53 VSS AV18 VSS AE41 VSS AM1 VSS AV22 VSS AE44 VSS AM6 VSS AV28 VSS AE49 VSS AM52 VSS AV32 VSS AE52 VSS AM57 VSS AV40 VSS AF6 VSS AN6 VSS AV43 VSS AF55 VSS AN14 VSS AV50 VSS AF57 VSS AN18 VSS AV52 VSS AG2 VSS AN20 VSS AW1 VSS AG12 VSS AN22 VSS AW57 VSS AG15 VSS AN26 VSS AY6 VSS AG18 VSS AN30 VSS AY20 VSS AG22 VSS AN36 VSS AY22 VSS AG28 VSS AN40 VSS AY25 VSS AG32 VSS AN43 VSS AY31 VSS AG36 VSS AN49 VSS AY36 VSS AG40 VSS AN52 VSS AY38 VSS AG47 VSS AR4 VSS AY52 VSS AH6 VSS AR6 VSS BA11 VSS AH52 VSS AR20 VSS BA14 VSS AJ1 VSS AR24 VSS BA20 VSS AJ20 VSS AR28 VSS BA27 VSS AJ26 VSS AR52 VSS BA33 VSS AJ30 VSS AR54 VSS BA41 VSS AJ34 VSS AT1 VSS BA47 VSS AJ38 VSS AT6 VSS BA56 VSS AJ57 VSS AT12 VSS BB1 VSS AK6 VSS AT18 VSS BB6 VSS AK52 VSS AT41 VSS BB52 VSS AL9 VSS AT46 VSS BC6 VSS AL14 VSS AT49 VSS BC9 VSS AL17 VSS AT52 VSS BC17 VSS AL22 VSS AU20 VSS BC22 VSS AL24 VSS AU24 VSS BC31 VSS AL28 VSS AU28 VSS BC38 VSS AL32 VSS AU32 VSS BC44 VSS AL36 VSS AU36 VSS BC52 Intel(R) Communications Chipset 89xx Series - Datasheet 1666 Signal October 2012 Order Number: 327879-001US Packaging Information Signal Ball Ball VSS BD15 VSS BM16 VSS BD17 VSS BM22 VSS BD25 VSS BM23 VSS BD27 VSS BM28 VSS BD33 VSS BM32 VSS BD36 VSS BM35 VSS BD43 VSS BM40 VSS BE6 VSS BM43 VSS BE52 VSS BM46 VSS BF1 VSS BM48 VSS BF6 VSS BM50 VSS BF11 VSS BM55 VSS BF12 VSS BM57 VSS BF22 VSS BN6 VSS BF31 VSS BN9 VSS BF41 VSS BN29 VSS BF50 VSS BN56 VSS BF57 VSS BP51 VSS BG5 VSS BP55 VSS BG27 VSS BP57 VSS BG36 VSS BR4 VSS BG41 VSS BR19 VSS BG46 VSS BR32 VSS BH6 VSS BR36 VSS BH49 VSS BR54 VSS BH52 VSS BT2 VSS BJ3 VSS BT7 VSS BJ5 VSS BT11 VSS BJ12 VSS BT15 VSS BJ17 VSS BT27 VSS BJ22 VSS BT53 VSS BJ57 VSS BT56 VSS BK6 VSS BU22 VSS BK8 VSS BU42 VSS BK38 VSS BU46 VSS BK48 VSS BU52 VSS BK52 VSS BU54 VSS BM3 VSSA_USB AT9 VSS BM8 WAKE# BU19 VSS BM12 WDT_TOUT# BT37 VSS BM15 Intel(R) Communications Chipset 89xx Series - Datasheet 1667 Signal October 2012 Order Number: 327879-001US 35.2.2 Ball Map Illustrations Table 35-3, "Ball Map Color Code" contains the signal color codes used in Table 35-4 through Table 35-7. Table 35-3. Ball Map Color Code Color Signals USB; SATA; SIU, PCIE_EP, SFP VSSA; VSS RSVD PCIE_RC GbE; SRD GPIO SPI DMI VCCEP; VCCAEP; VCCPEP VCC; VCCA_; VCC3P3; VCC1P8; VCCA1 VCCSUS Table 35-4. Bottom View Left (Sheet 1 of 4) A B C RSVD 2 2 RSVD 6 RSVD 1 5 RSVD 8 L M VSS P R GBE1 _SWD P1 T V VSS SFP1_ I2C_C LK GBE1 _SWD P0 SFP3_ I2C_D ATA Y AA AB SRDS O_1_ P SRDS O_0_ P GBE2 _SWD P0 VSS VSS RSVD 86 GBE_ AUX_ PWR_ OK GBE0 _SWD P1 SFP2_ I2C_C LK VSS SRDS 3_SD VSS VSS AG RSVD 82 Intel(R) Communications Chipset 89xx Series - Datasheet AH VSS GBE_ CLK10 0N SRDSI _1_P VSS SRDSI _0_P VSS AJ VSS SRDS O_0_ N VSS VSS AF VSS SRDS O_1_ N SRDS O_2_ N VSS AE GBE_ CLK10 0P SRDS O_2_ P GBE3 _SWD P0 RSVD 84 AC VSS SFP0_ I2C_D ATA EP_M AIN_P WR_O K GBE0 _LED W VSS SFP0_ I2C_C LK SRDS 0_SD VSS U VSS VSS VSS VSS SFP2_ I2C_D ATA SFP1_ I2C_D ATA RSVD 64 RSVD 63 GBE_I RCOM P GBE3 _LED GBE0 _SWD P0 VSS VSS VCCA EP1P8 AUX SRDSI _0_N VSS VSS RSVD 79 1668 N VSS GBE2 _SWD P1 GBE_ EE_D O VSS K SFP3_ I2C_C LK RSVD 85 8 9 J RSVD 83 SRDS 2_SD GBE_ EE_DI GBE_ AUX_ PWR_ AVAIL 7 H VSS GBE2 _LED SRDS 1_SD G RSVD 3 VSS VSS F RSVD 0 3 6 E RSVD 4 1 4 D October 2012 Order Number: 327879-001US 35.0 Table 35-4. Bottom View Left (Sheet 2 of 4) A B C D E G PCIE_ EP_RS T# 10 RSVD 81 11 12 F EP_S MBCL K 13 J K L N P VSS VSS EP_JT RST# M GBE_ WAKE # VSS EP_S EP_CR MBDA U_EN T RSVD 5 16 RSVD 22 RSVD 26 RSVD 28 VSS VSS RSVD 89 RSVD EP_JT 7 MS RSVD 30 RSVD 21 21 RSVD 27 26 RSVD 18 VSS RSVD 23 RSVD 29 27 VSS AB AC AE AF AG AH AJ GBE_ SMBD AT VSS SRDS O_3_ P GBE_ VREF1 P8 VCCA EP1P8 AUX GBE_ EE_CS # GBE_ SMBA LRT# GBE_ SMBC LK SRDS O_3_ N VSS VSS RSVD 88 VSS RSVD 87 VSS VCCP EP3P3 AUX VCCA EPAU X RSVD EP_JT 90 DO VSS RSVD 36 RSVD 37 EP_JT DI VCCP EP3P3 AUX VSS VSS VSS RSVD 10 RSVD 12 VSS VCCA EPAU X VSS VCCE PAUX VCCP EP3P3 AUX VCCP EP3P3 AUX VSS VSS VSS VSS VSS RSVD 14 RSVD RSVD 24 15 VSS VCCE PAUX VSS VCCE PAUX VSS VSS VCCE PAUX VSS VCCE PAUX RSVD 20 VCCA EPAU X VSS VSS VSS CRU_ RSVD CLK10 17 0N VSS RSVD 33 VCCA VCCA EPAU EP1P8 X _CRU VCCA VCCE EP1P8 PAUX _CRU VCCE VCCE PAUX PAUX RSVD 19 RSVD 31 VSS VSS PCIE_ EP_RX p15 AA GBE1 _LED 24 25 Y RSVD 34 RSVD 25 23 W VSS 20 22 V RSVD 9 18 19 U VSS RSVD 11 17 VSS RSVD 13 VSS T RSVD 16 14 15 R GBE3 GBE_ _SWD EE_SK P1 RSVD 80 EP_S MBAL RT# EP_JT CK H CRU_ RSVD CLK10 35 0P RSVD RSVD 69 70 RSVD 71 VSS VCCE PAUX October 2012 Order Number: 327879-001US RSVD 32 VSS RSVD 72 VSS VSS VCCA EPAU X VSS VSS VCCE VCCE P1P0_ PAUX CRU VCCE PAUX VCCE PAUX VSS VSS VCCE P1P0_ CRU VCCE P VSS VSS VSS PCIE_ EP_RX n15 VSS VSS VCCE P Intel(R) Communications Chipset 89xx Series - Datasheet 1669 Table 35-4. Bottom View Left (Sheet 3 of 4) A B C D E 28 29 PCIE_ EP_RX p14 VSS 32 PCIE_ EP_RX p13 VSS H J K L M N P R T U V VSS PCIE_ EP_RX p11 PCIE_ EP_RX n10 PCIE_ EP_RX p7 VSS VCCE VCCE P P VCCE P VSS VCCE P VSS VSS VCCE P VSS VCCE P VCCE P VSS VCCE P VSS VSS VCCA EP_PE VSS VCCE P VCCA EP_PE VSS VCCA EP_PE VSS PCIE_ PCIE_ EP_TX EP_TX n14 p14 VSS PCIE_ EP_TX n15 PCIE_ EP_TX p15 VSS VCCE P VSS VSS VSS PCIE_ EP_TX p12 PCIE_ EP_TX n12 VSS PCIE_ PCIE_ EP_TX EP_TX n13 p13 VSS VSS VCCE P VSS PCIE_ PCIE_ EP_TX EP_TX n10 p10 VSS PCIE_ EP_TX n11 PCIE_ EP_TX p11 VSS VCCE P VSS VSS PCIE_ PCIE_ EP_TX EP_TX n9 p9 VSS PCIE_ EP_TX p8 PCIE_ EP_TX n8 VSS PCIE_ EP_IC OMPI VSS PCIE_ PCIE_ EP_TX EP_TX p7 n7 PCIE_ EP_RX n5 VSS VCCA VCCA EP_PE EP_PE VSS VSS VCCA EP_PE VSS RSVD 61 VSS PCIE_ EP_CL K100P VSS VCCA EP_PE VCCA EP_PE VSS VCCA EP_PE PCIE_ EP_TX n6 VSS RSVD 62 VSS PCIE_ EP_CL K100 N VCCA EP1P8 _PE VCCA EP_PE VCCA EP_PE VCCA EP_PE VSS VSS PCIE_ EP_VR EF1P8 VSS VSS VCCA 1P8_D MI VSS PCIE_ EP_RX n6 VSS VSS PCIE_ EP_TX p6 PCIE_ EP_RX p5 VSS Intel(R) Communications Chipset 89xx Series - Datasheet 1670 VSS VSS 44 45 AJ VCCE P PCIE_ EP_RX n8 PCIE_ EP_RX n7 PCIE_ EP_RX p6 AH VSS VSS VSS AG PCIE_ EP_RX p8 40 41 AF PCIE_ EP_RX p9 PCIE_ EP_RX n9 VSS AE VSS PCIE_ EP_RX p10 VSS AC VCCE P VSS 38 43 AB VSS VSS PCIE_ EP_RX n12 PCIE_ EP_RX n11 37 42 AA VCCE VCCE P P 35 39 Y VCCE VCCE P P 34 36 W VSS PCIE_ EP_RX n13 PCIE_ EP_RX p12 33 G PCIE_ EP_RX n14 30 31 F October 2012 Order Number: 327879-001US 35.0 Table 35-4. Bottom View Left (Sheet 4 of 4) A 46 B C D PCIE_ EP_RX p4 VSS VSS VSS PCIE_ EP_RX n3 VSS 53 54 PCIE_ EP_TX p4 L M N P PCIE_ EP_TX p1 VSS R T U V W Y AA AB AC AE AF AG PCIE_ EP_TX n1 VSS VSS RSVD 42 RSVD 41 DMI_ VREF VSS DMI_T Xn0 DMI_T Xn1 VSS DMI_I RCOM P VSS AH AJ PCIE_ EP_TX p2 VSS PCIE_ EP_TX p3 PCIE_ EP_TX n2 PCIE_ EP_TX n0 DMI_T Xp0 DMI_T Xp1 DMI_T Xp2 VSS PCIE_ RC_T Xp0 PCIE_ EP_TX n4 PCIE_ EP_TX n3 VSS PCIE_ EP_TX p0 VSS VSS DMI_T Xn2 DMI_T Xp3 PCIE_ RC_T Xn0 VSS VSS VSS VSS VSS VSS TS0_I REF_N VSS RSVD 52 VSS RSVD 55 57 VSS VSS VSS PCIE_ EP_RX p2 DMI_ RXp1 VSS VSS DMI_ RXn1 DMI_T Xn3 VSS PCIE_ RC_R Xp0 DMI_ RXn2 DMI_ RXn0 VSS VSS VSS PCIE_ RC_R Xp1 DMI_ DMI_ CLK10 RXn3 0P DMI_ RXp0 PCIE_ EP_RX p0 VSS VSS VSS VSS PCIE_ EP_RX n1 VSS VSS VSS PCIE_ EP_RX p1 PCIE_ EP_RX n0 VSS RSVD 54 VSS VSS PCIE_ EP_RX n2 VSS RSVD 51 55 56 K PCIE_ EP_RX p3 VSS VSS J VSS VSS 51 H PCIE_ PCIE_ EP_TX EP_TX p5 n5 VSS 50 52 G VSS 48 49 F VSS PCIE_ EP_RX n4 47 E DMI_ DMI_ CLK10 RXp3 0N DMI_ RXp2 VSS VSS PCIE_ RC_R xn1 VSS VSS PCIE_ RC_R xn0 VSS VSS Table 35-5. Bottom View Right (Sheet 1 of 4) AK AL 1 2 3 4 AM AN AR VSS SRDSI _2_P AT AU VSS AY BA VSS RSVD 45 SRDSI _3_N AW USB_ RBIAS p SRDSI RSVD _3_P 46 SRDSI _2_N AV VSS October 2012 Order Number: 327879-001US BB BD BE VSS USB0_ Dn USB_ CLK96 P BC BF BG USB3_ Dn USB1_ Dn USB2_ Dp BK BL BM BN RSVD 66 MST_ SMBC LK USB4_ Dn USB3_ Dp USB2_ Dn BJ USB4_ Dp VSS USB1_ Dp BH VSS BR BT BU RSVD 65 RSVD 68 VSS RSVD 67 VSS MST_ SMBD AT BP OC1# / GPIO4 0 VSS Intel(R) Communications Chipset 89xx Series - Datasheet 1671 RSVD 59 Table 35-5. Bottom View Right (Sheet 2 of 4) AK AM AN AR AT VSS AU AV USB_ RBIAS n SRDSI _1_N 5 6 AL VSS VSS VSS VSS AW AY USB_ CLK96 N VSS BA BB BC BD BE BF USB0_ Dp VSS BG BH VSS VSS VSS VSS VSS BJ BK BM BN RSVD 73 VSS VSS BL VSS VCCAE P1P8A UX VCC1P 05 VCCS US3P3 _USB VCCS US_U SB VCCS US_U SB VCCA1 P8_US B USB5_ Dn RSVD 74 9 VSS VCC1P 05 VSSA_ USB VCCS US3P3 _USB VCCA1 P8_US B VSS USB5_ Dp TS1_I REF_N VSS GPIO2 7 VSS 11 VCCAE PAUX VCCA1 P8_TS 1 VCCA SUS1P 8_USB VCCS US3P3 _USB VSS VCCA_ USB VSS 12 VCCAE PAUX VCCS US1P8 _USB VSS VCCS US3P3 _USB VCCS US3P3 _USB VCCA_ USB VSS VSS OC0# / GPIO5 9 13 14 VSS VSS VCCS US1P8 _USB VCCS US1P8 VSS VCCS US3P3 15 VCCAE PAUX VCCA1 P5_TS 1 VCCS US1P8 _USB VSS VCCS US3P3 VCCS US3P3 VSS MST_ GPIO1 SMBA LERT# 2 / GPIO1 RSVD GPIO3 78 1 16 VSS VCCA1 P5_TS 0 VCCS US1P8 _USB VCCS US1P8 18 VCCAE PAUX VSS VSS VSS VCCS US1P8 VSS VSS VCCS GPIO2 US3P3 8 VSS VSS VSS VCCS US SUS_S TAT# / GPIO6 1 SUS_ CLK / GPIO6 2 GPIO7 2 VSS SYS_P WROK VSS MEPW ROK OC3# / GPIO4 2 VSS VSS VCCS US VCCS US SLP_S 5# / PWRB GPIO6 TN# 3 SML1C SML1 LK / DAT / GPIO5 GPIO7 8 5 VCCS US VCCS US VSS VSS VCCS US Intel(R) Communications Chipset 89xx Series - Datasheet 1672 VSS VCCS US VSS GPIO1 0 VSS GPIO4 3 WAKE # VSS GPIO4 6 GPIO4 5 GPIO9 VSS GPIO8 DRAM PWRG D SLP_S 4# RI# 21 VSS SML1A LERT# / GPIO7 4 GPIO6 0 SLP_S 3# VSS GPIO4 7 VSS VSS 19 VCCEP AUX GPIO5 6 GPIO3 0 VSS 17 RSVD 56 VSS GPIO5 7 PLTRS T# BU VSS VSS OC2# GPIO1 / 5 GPIO4 1 10 BT RSVD 57 GPIO2 4 8 22 BR RSVD 60 7 20 BP VSS GPIO7 3 October 2012 Order Number: 327879-001US VSS 35.0 Table 35-5. Bottom View Right (Sheet 3 of 4) AK AL AM AN AR AT AU AV AW AY BA BB BC BD BE BF BG BH BJ BK 23 24 VSS VCC VSS VSS VSS VCCEP VSS VCC VCC VCC VSS VSS VCC3P 3 VSS VCCS RSVD US3P3 39 _RTC GPIO1 GPIO2 3 5 VSS VCC1P 8 VSS VCC3P 3_RTC VSS VSS VCC VCC INTRU DER# IVCC_ RTC RSVD 53 VCC VCC VSS VCC3P 3 RSMR ST# VCCA1 P8_TS 0 VCC VCC VSS VSS VCC VSS VSS VCC1P 8 VSS SIU0_ DCD# SIU0_ SRTCR RXD ST# VSS VCC1P 8 VSS LDRQ LFRAM 1# / E# GPIO2 3 LAD1 SIU0_ RI# RSVD 38 VCCA_ VCCA_ DMI DMI VSS SIU0_ CTS# RSVD 77 VSS SIU1_ DSR# SIU1_ CTS# VCC3P 3 VCC3P 3 VSS VCC3P 3 VCC1P 8 VSS GPIO5 2 VSS LDRQ SIU1_ 0# DCD# SIU1_ RXD VCC3P 3 VSS VCC3P 3 VSS SATA_ LED# GPIO1 GPIO3 7 2 PCICL K VSS VSS VCCA_ DMI October 2012 Order Number: 327879-001US VSS SIU1_ RTS# VSS WDT_ TOUT # BBS1 / GPIO5 1 UART_ CLK VSS SIU0_ DSR# VSS SIU1_ RI# VCCA_ DMI JTDO SIU0_ RTS# 39 40 JTCK JTMS 37 38 RTCX2 VSS VSS VCC3P 3 VCCEP VCC3P 3 35 36 RTCX1 JTDI VSS 33 34 BU VSS VSS VSS BT GPIO4 4 VSS VCCEP BR GPIO1 4 RTCRS PWRO T# K VSS 31 32 BP VSS 29 30 BN GPIO2 6 VCC3P 3 VSS VCC3P 3 VCC3P 3 27 28 BM VCCS US 25 26 BL LAD0 VSS Intel(R) Communications Chipset 89xx Series - Datasheet 1673 LAD3 Table 35-5. Bottom View Right (Sheet 4 of 4) AK AL AM VCCA_ DMI 41 AN AR VCCA_ DMI AT AU AV AW AY VCCA_ SATA VSS BA BB BC BD BE VCCPC SPI_C PU LK VSS BF BG VSS VSS BH BJ BK BBS0 SIU1_ TXD BL BM BN VSS VSS VCCA_ DMI VSS VCCA_ SATA VCCA_ SATA 44 VCCA_ DMI VCCA_ DMI SATA_ ICOMP O VCCA_ SATA RSVD 44 VSS GPIO1 GPIO5 9 0 VSS GPIO4 GPIO2 45 46 VSS PCIE_ RC_TX p2 VSS VCCA1 P8_SA TA RSVD 43 SATA5 _GP / TEMP_ ALERT #/ GPIO3 5 47 PCIE_ RC_TX p1 PCIE_ RC_TX n2 SATA_ VREF1 P8 SATA4 _TXP VSS SATA4 _GP / GPIO1 6 RCIN# RSVD RSVD 75 76 VSS GPIO3 SYS_R ESET# SIU0_ DTR# SIU0_ TXD GPIO5 GPIO5 5 BMBU SY# / GPIO0 GPIO5 4 VSS 49 PCIE_ RC_TX n1 VSS VSS SATA4 _TXN SATA5 _TXP RSVD 47 GPIO2 1 VSS 50 VSS PCIE_ RC_TX p3 PCIE_ RC_TX n3 VSS SATA5 _TXN GPIO2 0 VSS SPI_M ISO VSS 53 VSS VSS VSS PCIE_ RC_RX n3 PCIE_ RC_RX p2 57 SATA_ CLK10 0N VSS SATA_ CLK10 0P SATA5 _RXP VSS SATA4 _RXN THRM TRIP# SCLOC K/ GPIO2 2 GPIO3 4 INIT3 _3V# VSS ADR/ GPIO3 7 SPI_C S1# CPUP WRGD RSVD 48 REF_C LK14 NRBO OTS SPI_C S0# GPIO1 8 VSS GPIO6 GPIO3 6 VSS SERIR Q VSS SDATA OUT0 / GPIO3 9 SIU1_ DTR# A20GA TE RSVD 50 SLOA D/ GPIO3 8 VSS VSS RSVD 49 PECI VSS VSS VSS PMSY NC Intel(R) Communications Chipset 89xx Series - Datasheet 1674 VSS SPI_M OSI PCIE_ SATA4 RC_RX _RXP p3 VSS VSS VSS PCIE_ RC_RX n2 55 VSS SATA5 _RXN VSS 54 56 GPIO7 VSS VSS VSS VSS 51 VSS BU VSS VSS GPIO1 48 BT LAD2 GPIO5 3 43 BR GPIO3 3 42 52 BP VSS VSS VSS VSS VSS VSS VSS SDATA OUT1 / GPIO4 8 VSS VSS October 2012 Order Number: 327879-001US 35.0 Table 35-6. Top View Left (Sheet 1 of 4) BU BT BR 2 RSVD 57 5 RSVD 56 7 GPIO5 6 GPIO6 0 VSS VSS 17 18 SYS_P WROK VSS VSS MEPW ROK GPIO8 DRAM PWRG D BA AY VSS USB1 _Dp AW AV AU VSS AT USB_ RBIAS p USB_ CLK96 P RSVD 45 USB_ CLK96 N USB0 _Dp VSS VSS VSS VSS VSS RSVD 74 USB5 _Dn VCCA 1P8_U SB TS1_I REF_N USB5 _Dp VSS VSS AM AL SRDSI _2_P SRDSI _2_N SRDSI _1_P SRDSI _3_N USB_ RBIAS n SRDSI _1_N VSS VSS PLTRS T# GPIO3 RSVD 1 78 VSS VSS VSS VSS VCCS US_U SB VCCS US_U SB VCCS US3P3 _USB VCC1 P05 VCCA EP1P8 AUX VSS VCCA 1P8_U SB VCCS US3P3 _USB VSSA _USB VCC1 P05 VSS VSS VCCA _USB VSS VCCS US3P3 _USB VCCA SUS1 P8_US B VCCA 1P8_T S1 VCCA EPAU X VSS VCCA _USB VCCS US3P3 _USB VCCS US3P3 _USB VSS VCCS US1P8 _USB VCCA EPAU X VCCS US3P3 VSS VCCS US1P8 VCCS US1P8 _USB VSS VSS RI# OC3# / GPIO4 2 October 2012 Order Number: 327879-001US VSS SRDSI _0_N VSS MST_ SMBA GPIO1 LERT# 2 / GPIO1 VSS VCCS US3P3 VCCS US3P3 VSS VCCS US1P8 _USB VCCA 1P5_T S1 VCCA EPAU X GPIO2 VCCS 8 US3P3 VSS VSS VCCS US1P8 VCCS US1P8 VCCS US1P8 _USB VCCA 1P5_T S0 VSS VSS VSS VSS VCCA EPAU X VSS SLP_S 4# AJ VSS OC2# GPIO1 / 5 GPIO4 1 OC0# / GPIO5 9 AK VSS VSS VSS AN RSVD SRDSI 46 _3_P USB2 _Dn VSS AR VSS USB0 _Dn USB1 _Dn 14 16 BB SUS_ STAT #/ GPIO6 1 SUS_ CLK / GPIO6 2 GPIO7 2 15 BC USB3 _Dp GPIO5 7 VSS 13 BD USB2 _Dp VSS GPIO3 0 SML1 ALERT #/ GPIO7 4 BE USB3 _Dn MST_ SMBD AT VSS 10 12 BF VSS VSS VSS 11 BG GPIO2 4 VSS GPIO4 7 BH USB4 _Dn RSVD 73 8 9 BJ USB4 _Dp VSS RSVD 60 GPIO2 7 BK MST_ SMBC LK OC1# / GPIO4 0 VSS BL RSVD 66 RSVD 67 RSVD 59 BM RSVD 68 VSS 3 6 BN RSVD 65 1 4 BP Intel(R) Communications Chipset 89xx Series - Datasheet 1675 Table 35-6. Top View Left (Sheet 2 of 4) BU 19 BT WAKE # BR BP BN AR AN VSS VSS VSS VCCE PAUX VSS VSS VCCS US VCCS US VSS VSS VCCE PAUX VCCS US VSS VSS VCC VSS VCCE PAUX VCC3 P3 VCC VCC VSS VCCE P VSS VSS VSS VSS VSS VCC VSS VCCE P JTDI VCC3 P3 VCC VCC VSS VCCE P VSS VSS VSS VCC VCC VSS VCCE P VCC3 P3 VCC VCC VCCA 1P8_T S0 VCCE P VSS VCC3 P3 VSS VCC VSS VSS VCCE P GPIO4 6 GPIO4 5 21 GPIO4 4 23 BK BJ BH BG BF BE BD BC BB BA AY VSS VCCS US AV AU SML1 SML1 DAT / CLK / GPIO7 GPIO5 5 8 SLP_S PWRB 5# / TN# GPIO6 3 VCCS VCCS US US VSS VCCS US VSS GPIO4 3 GPIO1 0 VCCS US GPIO1 4 VSS VSS VSS VSS RSVD 53 26 RTCX2 27 IVCC_ RTC INTRU DER# RSMR ST# VSS JTDO JTCK 32 RSVD 38 SIU0_ DSR# PWRO RTCR K ST# SRTC SIU0_ RST# RXD JTMS VSS RSVD 77 33 VCCS RSVD US3P3 39 _RTC VSS VCC3 P3 VCC3 P3 36 VSS VCC3 P3_RT C VSS VCC1 P8 VSS SIU0_ DCD# VSS SIU0_ RTS# SIU0_ LAD1 RI# LDRQ 1# / LFRA GPIO2 ME# 3 SIU1_ DSR# AJ VCC1 P8 VSS VCC3 P3 VSS VSS VCC1 P8 VSS VCC3 P3 VSS VSS SIU1_ RXD SIU1_ LDRQ DCD# 0# VSS GPIO5 2 Intel(R) Communications Chipset 89xx Series - Datasheet 1676 AK VCC3 P3 VSS SIU0_ CTS# SIU1_ CTS# SIU1_ RTS# AL VSS 34 35 AM VSS 30 31 GPIO2 GPIO1 5 3 GPIO2 6 RTCX1 28 29 AT VSS 24 25 AW GPIO9 GPIO7 3 VSS BL SLP_S 3# VSS 20 22 BM VSS VCC1 P8 VCC3 P3 VSS October 2012 Order Number: 327879-001US 35.0 Table 35-6. Top View Left (Sheet 3 of 4) BU BT BR BP WDT_ TOUT # 37 BN BBS1 / GPIO5 1 LAD3 BK BJ VSS PCICL K BH BG BF BE GPIO3 3 LAD2 VSS GPIO3 GPIO1 2 7 SIU0_ TXD SIU0_ DTR# BMBU SY# / GPIO0 45 AY VCC3 P3 VSS AW GPIO5 5 AL VCCA VCCA _DMI _DMI AK VCCA _DMI VSS VCCA _DMI VCCA _DMI VCCA _SATA VCCA _SATA VSS VCCA _DMI VSS VSS VSS RSVD 44 VCCA _SATA SATA_ ICOM PO VCCA _DMI VCCA _DMI GPIO3 5 SATA5 _GP / TEMP _ALER T# / RSVD 43 VCCA 1P8_S ATA VSS PCIE_ RC_T Xp2 VSS RCIN # SATA4 _GP / GPIO1 6 VSS SATA4 _TXP SATA_ VREF1 P8 PCIE_ RC_T Xn2 PCIE_ RC_T Xp1 VSS GPIO2 1 RSVD 47 SATA5 _TXP SATA4 _TXN VSS VSS PCIE_ RC_T Xn1 INIT3 _3V# SPI_M ISO VSS GPIO2 0 SATA5 _TXN VSS PCIE_ RC_T Xn3 PCIE_ RC_T Xp3 VSS VSS VSS THRM TRIP# VSS VSS VSS VSS AJ VSS SPI_C VCCP LK CPU VSS GPIO5 GPIO1 0 9 VSS VSS SYS_ RESET # VSS SDAT AOUT 0/ GPIO3 9 VSS VSS AM VCCA _SATA GPIO3 4 50 53 AN VSS VSS GPIO1 SCLO CK / GPIO2 2 VSS VSS AR VSS RSVD RSVD 76 75 VSS SDAT AOUT 1/ GPIO4 8 VCC3 P3 AT GPIO5 48 51 AU VSS GPIO2 GPIO4 VSS GPIO7 SLOA D/ GPIO3 8 AV VCCA _DMI SIU1_ BBS0 TXD VSS GPIO5 4 VSS 47 54 VSS BA VSS 44 52 SATA_ LED# BB GPIO5 3 GPIO3 43 49 BC VSS 41 46 BD UART _CLK LAD0 40 42 BL SIU1_ RI# 38 39 BM SIU1_ DTR# NRBO OTS GPIO6 VSS GPIO3 6 REF_C LK14 A20G ATE October 2012 Order Number: 327879-001US VSS VSS ADR/ GPIO3 7 VSS VSS PMSY NC RSVD 49 SPI_M OSI SATA_ CLK10 0N VSS VSS SATA5 _RXN VSS VSS VSS VSS PCIE_ RC_R Xn3 Intel(R) Communications Chipset 89xx Series - Datasheet 1677 PCIE_ RC_R Xp1 Table 35-6. Top View Left (Sheet 4 of 4) BU BT BR 55 BP BN VSS 56 VSS BM BK VSS BJ BH BG BF SPI_C S0# VSS BE GPIO1 8 VSS BD BC RSVD 50 SERIR Q VSS 57 BL VSS BB BA AY SPI_C S1# AU AT AR AN PCIE_ SATA4 RC_R _RXP Xp3 VSS SATA4 _RXN VSS AM AL AK PCIE_ RC_R Xn2 SATA5 _RXP VSS CPUP WRGD VSS AV SATA_ CLK10 0P PECI RSVD 48 AW AJ PCIE_ RC_R xn1 PCIE_ RC_R Xp2 VSS VSS Table 35-7. Top View Right (Sheet 1 of 4) AH AG AF AC GBE_C LK100 P 1 2 VSS VSS 4 AB SRDS O_2_P VSS V U T VSS VSS VSS N M SFP1_ I2C_C LK SFP0_ I2C_C LK SRDS O_0_P K GBE2_ SWDP 0 H F GBE2_ SWDP 1 RSVD 85 RSVD 6 GBE_E E_DI RSVD 63 RSVD 64 SFP1_ I2C_D ATA SFP2_ I2C_D ATA SFP2_ I2C_C LK GBE0_ SWDP 1 GBE_A UX_P WR_O K 9 VCCAE P1P8A UX VSS VSS GBE0_ SWDP 0 GBE3_ LED VSS SRDS 3_SD VSS RSVD 86 SRDS 1_SD VSS 11 VCCAE P1P8A UX GBE_V REF1P 8 SRDS O_3_P VSS GBE_S MBDA T GBE1_ LED 12 VSS VSS SRDS O_3_ N GBE_S MBCL K GBE_S MBAL RT# GBE_E E_CS # GBE_A UX_P WR_A VAIL RSVD 80 VSS GBE_ WAKE # Intel(R) Communications Chipset 89xx Series - Datasheet 1678 VSS RSVD 81 EP_SM BALRT # VSS RSVD 16 13 RSVD 79 PCIE_ EP_RS T# VSS EP_SM EP_CR BDAT U_EN RSVD 8 GBE_E E_DO RSVD 82 GBE3_ GBE_E SWDP E_SK 1 10 VSS RSVD 1 VSS GBE_I RCOM P A VSS 7 8 B RSVD 2 GBE2_ LED RSVD 84 VSS C RSVD 4 SRDS 2_SD VSS D RSVD 0 VSS GBE1_ SWDP 1 VSS E RSVD 3 SFP3_ I2C_D ATA GBE0_ LED G RSVD 83 EP_MA IN_PW R_OK VSS J SFP3_ I2C_C LK VSS VSS L SRDS 0_SD GBE3_ SWDP 0 VSS P VSS GBE1_ SWDP 0 SRDS O_2_ N VSS R VSS SFP0_ I2C_D ATA SRDS O_0_ N VSS W SRDS O_1_P SRDSI _0_P VSS Y SRDS O_1_ N VSS 5 AA VSS GBE_C LK100 N 3 6 AE EP_JT RST# October 2012 Order Number: 327879-001US EP_JT CK EP_SM BCLK 35.0 Table 35-7. Top View Right (Sheet 2 of 4) AH AG AF AE AC AB AA Y W V U T R 14 VCCAE PAUX VCCPE P3P3A UX VSS RSVD 87 VSS RSVD 88 15 VSS VSS VCCPE P3P3A UX EP_JT DI RSVD 37 RSVD 36 P N VSS M L EP_JT RSVD DO 90 K J H RSVD 89 VSS 16 G F VCCPE P3P3A UX VCCPE P3P3A UX VCCEP AUX VSS 18 VSS VSS VSS VSS VCCAE PAUX VSS RSVD 12 RSVD 10 VSS EP_JT RSVD MS 7 VSS VCCEP AUX VSS VSS VSS VCCEP VCCEP AUX AUX RSVD 9 VSS VCCAE PAUX RSVD 20 VSS RSVD RSVD 15 24 RSVD 14 VSS VCCAE VCCEP P1P8_ AUX CRU VCCAE VCCAE P1P8_ PAUX CRU RSVD 33 VSS CRU_ RSVD CLK10 17 0N 23 24 VCCEP AUX VCCEP VCCEP 1P0_C AUX RU VSS VSS VCCEP VSS VCCEP 1P0_C RU VSS VCCEP VSS VSS VCCEP VSS RSVD 71 RSVD RSVD 70 69 CRU_ RSVD CLK10 35 0P VSS VCCAE PAUX VSS VSS RSVD 72 VSS RSVD 19 RSVD 23 RSVD 25 VSS RSVD 31 RSVD 29 VSS PCIE_ EP_RX n15 RSVD 32 VCCEP VCCEP PCIE_ EP_RX p15 PCIE_ EP_RX p14 VSS VSS VSS October 2012 Order Number: 327879-001US VSS VSS VCCEP VCCEP 31 RSVD 27 VSS PCIE_ EP_RX n14 VCCEP RSVD 21 RSVD 18 29 30 VSS RSVD 30 VSS VCCEP VSS VCCEP AUX VSS 27 28 RSVD 28 VSS 25 26 VSS RSVD 11 RSVD 34 VCCEP AUX A RSVD 5 RSVD 26 21 VSS B RSVD 13 VSS VCCEP AUX C RSVD 22 VSS 19 22 D VSS 17 20 E VCCEP VSS PCIE_ EP_TX p15 PCIE_ EP_TX n15 VSS PCIE_ PCIE_ EP_TX EP_TX p14 n14 PCIE_ EP_RX n13 PCIE_ EP_RX p13 Intel(R) Communications Chipset 89xx Series - Datasheet 1679 Table 35-7. Top View Right (Sheet 3 of 4) AH 32 AG VSS AF AE AC VCCEP VSS AB AA Y VSS VSS 33 34 W V VSS VCCEP T R P N M L K J H VSS PCIE_ PCIE_ EP_TX EP_TX p13 n13 VSS PCIE_ EP_TX n12 PCIE_ EP_TX p12 VSS VSS VCCAE P_PE F C VSS VSS VSS PCIE_ EP_RX p11 PCIE_ EP_RX n11 VSS VSS VSS VSS VCCEP VSS PCIE_ EP_TX p11 PCIE_ EP_TX n11 PCIE_ PCIE_ EP_TX EP_TX p10 n10 VSS PCIE_ EP_RX p10 VSS VCCAE P_PE VCCAE VCCAE P_PE P_PE VSS PCIE_ EP_IC OMPI VSS PCIE_ EP_TX n8 PCIE_ EP_TX p8 PCIE_ PCIE_ EP_TX EP_TX p9 n9 VSS 40 VSS VCCAE P_PE VSS VSS 41 VCCAE P_PE VSS VCCAE P_PE VCCAE P_PE VSS PCIE_ EP_RX n9 VSS PCIE_ EP_CL K100P VSS RSVD 61 PCIE_ PCIE_ EP_TX EP_TX n7 p7 VSS PCIE_ EP_RX n8 PCIE_ EP_RX p7 PCIE_ EP_RX n7 VSS 43 VCCAE P_PE VCCAE P_PE VCCAE P_PE VCCAE P1P8_ PE PCIE_ EP_CL K100N VSS 44 VCCA1 P8_D MI VSS VSS PCIE_ EP_VR EF1P8 VSS VSS RSVD 62 VSS PCIE_ EP_TX n6 PCIE_ EP_TX p6 VSS 45 46 DMI_V REF RSVD 41 RSVD 42 VSS VSS PCIE_ EP_TX n1 47 VSS DMI_I RCOM P VSS DMI_T Xn1 DMI_T Xn0 VSS PCIE_ EP_TX p1 VSS DMI_T Xp1 DMI_T Xp0 Intel(R) Communications Chipset 89xx Series - Datasheet 1680 PCIE_ EP_TX n0 PCIE_ PCIE_ EP_TX EP_TX n5 p5 PCIE_ EP_TX n2 VSS PCIE_ EP_TX p3 PCIE_ EP_TX p4 VSS VSS PCIE_ EP_RX n6 PCIE_ EP_RX p6 VSS PCIE_ EP_RX p5 PCIE_ EP_RX n5 PCIE_ EP_RX p4 VSS VSS PCIE_ EP_RX n4 VSS PCIE_ EP_TX p2 48 DMI_T Xp2 VSS VSS 42 VSS VSS PCIE_ EP_RX n10 PCIE_ EP_RX p8 PCIE_ RC_TX p0 A VSS PCIE_ EP_RX p12 PCIE_ EP_RX p9 VCCAE P_PE B PCIE_ EP_RX n12 39 49 D VSS 37 38 E VCCEP VCCEP 35 36 G VSS VCCEP VCCEP U VSS VSS VSS October 2012 Order Number: 327879-001US VSS 35.0 Table 35-7. Top View Right (Sheet 4 of 4) AH AG AF PCIE_ RC_TX n0 50 AE AC DMI_T Xp3 AB AA DMI_T Xn2 Y W V VSS U T R P N PCIE_ EP_TX p0 VSS M L K J PCIE_ EP_TX n3 VSS H G PCIE_ EP_TX n4 F E 53 PCIE_ EP_RX p3 DMI_T Xn3 VSS VSS VSS VSS VSS VSS DMI_R Xp1 VSS PCIE_ RC_Rx n0 DMI_C DMI_R LK100 Xp3 N VSS DMI_R Xn1 DMI_R Xp2 VSS October 2012 Order Number: 327879-001US VSS VSS VSS VSS VSS PCIE_ EP_RX p0 VSS VSS VSS VSS A VSS VSS RSVD 51 VSS RSVD 54 VSS VSS RSVD 52 VSS PCIE_ EP_RX p2 VSS B PCIE_ EP_RX n3 TS0_I REF_N VSS PCIE_ EP_RX n2 PCIE_ EP_RX n1 VSS VSS PCIE_ EP_RX n0 DMI_R Xp0 DMI_R Xn0 VSS VSS PCIE_ EP_RX p1 VSS DMI_R Xn2 VSS VSS VSS DMI_C DMI_R LK100 Xn3 P 55 57 VSS PCIE_ RC_RX p0 54 56 C VSS 51 52 D VSS RSVD 55 Intel(R) Communications Chipset 89xx Series - Datasheet 1681 35.3 Package Mechanical Information Figure 35-1. Mechanical Package Intel(R) Communications Chipset 89xx Series - Datasheet 1682 October 2012 Order Number: 327879-001US