SEMICONDUCTEURS SPECIFIQUES TSPC603R CINE, on PowerPC 603e RISC MICROPROCESSOR Family AT Beaerone PID7t-603e Specification DESCRIPTION The PID7t-603e implementation of PowerPC603e (after named 603r) is a low-power implementation of reduced instruction set computer (RISC) microprocessors PowerPC family. The 603r implements 32-bit effective addresses, inte- ger data types of 8, 16 and 32 bits, and floating-point datatypes of 32 and 64 bits. The 603r is a low-power 2.5/3.3-volt design and provides four software controllable power-saving modes. The 603r is a superscalar processor capable of issuing and retiring as many as three instructions per clock. Instructions can execute out of order for increased performance ; however, the 603r makes completion appear sequential. The 603r inte- grates five execution units and is able to execute five instruc- tions in parallel. The 603r provides independent on-chip, 16-Kbyte, four-way set-associative, physically addressed caches for instructions and data and on-chip instruction and data memory manage- ment units (MMUs). The MMUs contain 64-entry, two-way set- associative, data and instruction translation lookaside buffers that provide support for demand-paged virtual memory address translation and variable-sized block translation. The 603r has a selectable 32 or 64-bit data bus and a 32-bit address bus. The 603r interface protocol allows multiple mas- ters to complete for system resources through a central exter- nal arbiter. The 603r supports single-beat and burst data trans- fers for memory accesses, and supports memory-mapped I/O. The 603r uses an advanced, 2.5/3.3-V CMOS process techno- logy and maintains full interface compatibility with TTL devi- ces. The 603r integrates in system testability and debugging fea- tures through JTAG boundary-scan capability. MAIN FEATURES 7.4 SPECint95, 6.1 SPECfp95 @ 300 MHz (estimated) Superscalar (3 instructions per clock peak). Dual 16KB caches. Selectable bus clock. 32-bit compatibility PowerPC implementation. On chip debug support. Pp typical = 3.5 Watts (266 MHz), full operating conditions. Nap, doze and sleep modes for power savings. Branch folding. 64-bit data bus (32-bit data bus option). 4-Gbyte direct addressing range. Pipelined single/double precision float unit. IEEE 754 compatible FPU. IEEE P 1149-1 test mode (JTAG/COP). fine Max = 300 MHz. fous Max = 75 MHz. Compatible CMOS input / TTL Output. January 1999 G suffix CBGA 255 Ceramic Ball Grid Array GS suffix CI-CGA 255 Ceramic Ball Grid Array with Solder Column Interposer (SCI) SCREENING / QUALITY / PACKAGING This product is manufactured in full compliance with: m@ CI-CGA 255 : MIL-STD-883 class Q or According to TCS standards (planned) m CBGA 255 : Upscreenings based upon TCS standards m@ Full military temperature range (T, = -55C, To= +125C) Industrial temperature range (T, = -40 C, Te= +110C) m@ Internal // /O Power Supply = 2.545% /3.3V+5%. m@ 255 pin CBGA package and 255 pin CBGA with SCI (CIl- CGA) package. 1/38TSPC603R SUMMARY A. GENERAL DESCRIPTION .............. 3 5.1.3. Condition Register (CR) ............ 22 5.1.4. Floating-Point Status and Control Register 1. INTRODUCTION ............. 000 c eee ee eee 3 (FPSC) ......... 2.00 eee eee 22 5.1.5. Machine State Register (MSR) ...... 22 2. PINASSIGNMENTS ...............0020 ee eeee 4 5.1.6. Segment Registers (SRs) ........... 20 2.1. CBGA package ........... 0... cee eee 4 5.1.7. Special-Purpose Registers (SPRs) ... 22 2.2. Pinoutlisting ..............00 cece eee ee 5 5.2. Instruction set and addressing modes .... 25 5.2.1. PowerPC instruction set and addressing 3. SIGNAL DESCRIPTION .............2..0005: 7 MOdeS ........- 00 0c eee ee eee eee 25 5.2.2. PowerPC 603r microprocessor instruction SOt eee 26 B. DETAILED SPECIFICATIONS .......... 10 5.3. Cache implementation .................. 26 1, SCOPE 20... .....c cece cece c eee eeaee eens 10 5.3.1. PowerPC cache characteristics . .. .. . 26 5.3.2. PowerPC 603r microprocessor cache 2. APPLICABLE DOCUMENTS ................ 10 implementation ..............-..54- 26 3. REQUIREMENTS .............--0000eee eee 10 5.4. Exception model ....................05- 27 5.4.1. PowerPC exception model .......... 27 3.1. General ....... 0.0... eee 10 5.4.2. PowerPC 603r microprocessor exception ; ; model ........... 0c eee eee eee 28 3.2. Design and construction ................ 10 3.2.1. Terminal connections ............... 10 5.5. Memory management .................. 31 3.2.2. Lead material andfinish ............ 10 5.5.1. PowerPC memory management ..... 31 ; ; 5.5.2. PowerPC 603r microprocessor memory 3.3. Absolute maximum ratings .............. 10 management .............. cece eee 31 3.4. Recommended operating conditions ...... 11 5.6. Instruction timing ...................05. 31 3.5. Thermal characteristics ................. 11 6. PREPARATION FOR DELIVERY ..........--. 32 3.6. Power consideration ................... 12 6.1. Packaging .............. ccc cece eee eee 32 3.6.1. Dynamic Power Management ....... 12 a ; 3.6.2. Programmable Power Modes ........ 42 6.2. Certificate of compliance ................ 32 3.6.3. Power Management Modes ......... 12 7. HANDLING ..........0000ceeeeeeeeeeeeeeees 32 3.6.4. Power Management Software Considerations .................... 14 8. PACKAGES MECHANICAL DATA ........... 33 3.6.5. Power dissipation .................. 14 8.1.CBGA package parameters .............. 33 3.7. Marking ........ 00: eee eee eee eee 15 8.2. Mechanical dimensions of the CBGA 4. ELECTRICAL CHARACTERISTICS .......... 15 package ............0 2.0 cee eee eee 33 4.1. General requirements .................. 15 8.3. CI-CGA package parameters ........... 34 4.2. Static characteristics ................... 15 8.4. Mechanical dimensions of the CI-CGA | Package ....... eee eee eee 34 4.3. Dynamic characteristics ................ 16 4.3.1. Clock AC specifications ............. 16 9. CLOCK RELATIONSHIPS CHOICE .......... 35 4.3.2. Input AC specifications ............. 17 4.3.3. Output AC specifications ............ 18 10. SYSTEM DESIGN INFORMATION ........... 36 4.4, STAG AC timing specifications ........... 20 10.1 PLL Power Supply Filtering ...++++++s+0s 36 5. FUNCTIONAL DESCRIPTION ............+-- 29 10.2 Decoupling Recommendations .......... 36 5.1. PowerPC registers and programming 10.3 Connection Recommendations ..........- 36 model .............. 005. Cotte ee sss 22 10.4 Pull-up Resistor Requirements ......... 37 5.1.1. General-Purpose Registers (GPRs) .. 22 5.1.2. Floating-Point Registers (FPRs) ..... 22 11. ORDERING INFORMATION ................ 38 2/38 eS a SEMILONGUCTAUSS SPECIES?TSPC603R A. GENERAL DESCRIPTION 32b 64b address yv__data Figure 1 : Block diagram 1. INTRODUCTION The 603r is a low-power implementation of the PowerPC microprocessor family of reduced instruction set commuter (RISC) micro- processors. The 603r implements the 32-bit portion of the PowerPC architecture, which provides 32-bit effective addresses, integer data types of 8, 16 and 32 bits, and floating-point data types of 32 and 64 bits. For 64-bit PowerPC microprocessors, the PowerPC architecture provides 64-bit integer data types, 64-bit addressing, and other features required to complete the 64-bit architecture. The 603r provides four software controllable power-saving modes. Three of the modes (the nap, doze, and sleep modes) are static in nature, and progressively reduce the amount of power dissipated by the processor. The fourth is a dynamic power management mode that causes the functional units in the 603r to automatically enter a low-power mode when the functional units are idle without affecting operational performance, software execution, or any external hardware. The 603r is a superscalar processor capable of issuing and retiring as many as three instructions per clock. Instructions can execute out of order for increased performance ; however, the 603r makes completion appear sequential. The 603 e integrates five execution units - an integer unit (IU), a floating-point unit (FPU), a branch processing unit (BPU), aload/store unit (LSU) and a system register unit (SRU). The ability to execute five instructions in parallel and the use of simple instructions with rapid execution times yield high efficiency and throughput for 603r-based systems. Most integer instructions execute in one clock cycle. The FPU is pipelined so a single-precision multiply-add instruction can be issued every clock cycle. The 603r provides independent on-chip, 16 Kbyte, four-way set-associative, physically addressed caches for instructions and data and on-chip instruction and data memory management units (MMUs). The MMUs contain 64-entry, two-way set-associative, data and instruction translation lookaside buffers (DTLB and ITLB) that provide support for demand-paged virtual memory address translation and variable-sized block translation. The TLBs and caches use a least recently used (LRU) replacement algorithm. The 603r also supports block address translation through the use of two independent instruction and data block address translation (IBAT and DBAT) arrays of four entries each. Effective addresses are compared simultaneously with all four entries in the BAT array during block translation. In accordance with the PowerPC architecture, if an effective address hits in both the TLB and BAT array, the BAT translation takes priority. The 603r has a selectable 32 - or 64-bit - data bus and a 32-bit address bus. The 603r interface protocol allows multiple masters to compete for system resources through a central external arbiter. The 603r provides a three-state coherency protocol that supports the exclusive, modified, and invalid cache states. This protocol as a compatible subset of the MESI (modified/exclusive/shared/in- valid) four-state protocol and operates coherently in systems that contain four-state caches. The 603r supports single-beat and burst data transfers for memory accesses, and supports memory-mapped I/O. The 603r uses an advanced, 0.29 um 5 metal layer CMOS process technology and maintains full interface compatibility with TTL devices. 3/38TSPC603R 2. PIN ASSIGNMENTS 2.1. CBGA 255 and CI-CGA 255 packages Figure 2 (pin matrix) shows the pinout as viewed from the top of the CBGA and CI-CGA packages. The direction of the top surface view is shown by the side profile of the packages. Pin matrix top view 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 OOQOQOQOQOOOO0O0O000O OOOO QOQOOOO0O0O0O0000 OOQOQOQOQOOO0OOCO0O000O OIOTOTOLOIOTOIOLOLOLOLOIOIOLOle OOQOQOQOQOQOOQOOO0OO000O OOQOQOQOOQOO0O0CO0000O OOQVDQOOOQOOQOOCO0000O OOQOQOOQOQOO0QOOCO0000O OOQOQOQOQOQOOQOOO0O000O OOQOQOQOOOO0O0COO0000O OOQVOQOQOQOQOO0OOCO0000O OOQOQOQOQOOQOOQOQOO0O000O OOOO OQOQOOO0O0O0O0O000 OOQOQOQOQOQOO0OOCO0O000O OOQOQOQOQOQOOOQOOC0O0O000O OIOTOTOTOTOTOLOLOLOLOLOLOLOlOrS Substrate Assembly pia |" CBGA 255 4+HvuvuzerRKertoimmoooe p Encapsulant / meh C I-CG A 255 UUUUUUUUUUUUUUUU Not to scale Figure 2: CBGA 255 and CI-CGA 255 Top view 4/38 a8 Ys Le at RE SPECERIQUESTSPC603R 2.2. Pinout listing Table 1 : Power and ground pins K03, KO5, K12, K14, M05, M07, M10, M12, P07, P10 VDD2 GND PLL(AVDD) | A10 Internal logic | F06, F08, FO9, F11, GO7, G10, H06, HO8, HO9, H11, | C05, C12, E03, E06, E08, E09, E11, E14, F05, FO7, J06, JO8, JO9, J11, KO7, K10, L06, LO8, Log, L11 F10, F12, G06, G08, GO9, G11, H05, HO7, H10, H12, J05, JO7, J10, J12, KO6, KO8, KO9, K11, L05, Output drivers | CO7, E05, E07, E10, E12, G03, G05, G12, G14, LO7, L10, L12, M03, M06, M08, MO9, M11, M14, P05, P12 Table 2 : Signal pinout listing Signal name CBGA Pin number Active 1/0 A[0-31] C16, E04, D13, FO2, D14, GO1, D15, E02, D16, DO4, E13, Go2, E15, HO1, E16, High 1/0 H02, F13, JO1, F14, JO2, F15, HO3, F16, F04, G13, KO1, G15, KO2, H16, M01, J15, PO1 AACK Lo2 Low Input ABB K04 Low 1/0 AP[0-3] C01, B04, BO3, BO2 High 1/0 APE A04 Low Output ARTRY Jo4 Low 1/0 BG Lo1 Low Input BR BOo6 Low Output Cl E01 Low Output CKSTP_IN Dos Low Input CKSTP_OUT AO6 Low Output CLK_OUT DO7 - Output CSE[0-1] BO1, BOS High Output DBB J14 Low 1/0 DBG No1 Low Input DBDIS H15 Low Input DBWO G04 Low Input DH[0-31] P14, T16, R15, 715, R13, R12, P11, N11, R11, T12, T11, R10, PO9, NO9, T10, ROY, | High 1/0 TO9, P08, NO8, RO8, T08, NO7, RO7, TO7, PO6, NO6, ROG, TO6, RO5, NO5, TO5, To4 DL[0-31] K13, K15, K16, L16, L15, L13, L14, M16, M15, M13, N16, N15, N13, N14, P16, High 1/0 P15, R16, R14, 714, N10, P13, N12, T13, P03, NO3, N04, RO3, T01, TO02, P04, TO03, RO4 DP[0-7] M02, LO3, N02, LO4, RO1, P02, M04, R02 High 1/0 DPE AO5 Low Output DRTRY G16 Low Input GBL FO1 Low 1/0 HRESET A07 Low Input INT B15 Low Input L1_TSTCLK! D11 - Input 5/38TSPC603R L2 TSTCLK! D12 - Input CSSD_MODE! B10 Low Input MCP C13 Low Input PLL_CFG[0-3] A08, B09, A09, Dog High Input QACK DO3 Low Input QREQ J03 Low Output RSRV Do1 Low Output SMI A16 Low Input SRESET B14 Low Input SYSCLK cog - Input TA H14 Low Input TBEN C02 High Input TBST A14 Low 1/0 TC[O-1] A02, A03 High Output TCK C11 - Input TDI A11 High Input TDO Al2 High Output TEA H13 Low Input TLBISYNC co4 Low Input TMS B11 High Input TRST C10 Low Input TS J13 Low 1/0 TSIZ[0-2] A13, D10, B12 High 1/0 TT[0-4] B13, A15, B16, C14, C15 High 1/0 WT Do2 Low Output NC BO7, B08, C03, C06, C08, DOS, DOG, FO3, H04, J16 Low Input VOLTDETGND3 FO3 Low Output Notes : 1. These are test signals for factory use only and must be pulled up to OVDD for normal machine operation. 2. OVDD inputs supply power to the I/O drivers and VDD inputs supply power to the processor core. 3. NC (no-connect) in the 603e BGA package; internally tied to GND in the 603r BGA package to indicate to the power supply that a low-voltage processor is present. 6/38 S SPECHIQUESTSPC603R 3. SIGNAL DESCRIPTION Figure 3, Table 3 and Table 4 describe the signals on the TSPC603r and indicate signal functions. The test signals, TRST, TMS, TCK, TDI and TDO, comply with subset P-1149.1 of the IEEE testability bus standard. The 3 signals LSSD_MODE, LI_LTSTCLK and L2_TSTCLK are test signals for factory use only and must be pulled up to VDD for normal machine operations. ADDRESS ARBITRATION ADDRESS START ADDRESS BUS TRANSFER ATTRIBUTE ADDRESS TERMINATION CLOCKS POWER SUPPLY INDICATOR SYSCLK . CLK OUT <n et * PLL_CFG[0-3 > VOLTDETGND 603r Oo MO =| PY = =a 38 ~ = 20 19 40 OD Oo Q) F\a aS DPE INT, SMI MCP HRESET, SRESET RSRV QREQ, QACK TBEN TLBISYNC LSSD_MODE, VDD OVDD GND AVDD Figure 3 : Functional signal groups Table 3 : Address and data bus signal index DH[0-31], DL[0-31 5 DP[0-7 ) |] $$ <<] $a pCKSTP IN, CKSTP OUT Jn = $$ $$ @RST, TCK, TMS, TDI, TRO pb TSTCLK, L2 TSTCLK M$ a) ee EEE EERIE EEEEEET : A a DATA ATTRIBUTION DATA TRANSFER DATA TERMINATION INTERRUPTS CHECKSTOPS RESET PROCESSOR STATUS JTAG/COP INTERFACE LSSD TEST CONTROL POWER SUPPLY Signal name Mnemonic Signal function Signal type Address bus A[0-31] if output, physical address of data to be transferred. VO if input, represents the physical address of a snoop operation. Data bus DH[0-31] Represents the state of data, during a data write operation if output, or VO during a data read operation if input. Data bus DL[0-31] Represents the state of data, during a data write operation if output, or VO during a data read operation if input. 7/38TSPC603R Table 4 : Signal index Signal name Mnemonic Signal function Signal type Address Acknowledge | AACK The address phase of a transaction is complete Input Address Bus Busy ABB If output, the 603r is the address bus master VO If input, the address bus is in use Address Bus Parity AP[0-3] If output, represents odd parity for each of 4 bytes of the physical VO address for a transaction If input, represents odd parity for each of 4 bytes of the physical address for snooping operations Address Parity Error APE Incorrect address bus parity detected on a snoop Output Address retry ARTRY If output, detects a condition in which a snooped address tenure must be | I/O retried If input, must retry the preceding address tenure Bus grant BG May, with the proper qualification, assume mastership of the address Input bus Bus request BR Request mastership of the address bus Output Cache Inhibit cl A single-beat transfer will not be cached Output Test Clock CLK_OUT Provides PLL clock output for PLL testing and monitoring Output Checkstop Input CKSTP_IN Must terminate operation by internally gating off all clocks, and release Input all outputs Checkstop Output CKSTP_OUT Has detected a checkstop condition and has ceased operation Output Cache Set Entry CSE[0-1] Cache replacement set element for the current transaction reloading into | Output or writing out of the cache Data Bus Busy DBB If output, the 603r is the data bus master VO If input, another device is bus master Data Bus Disable DBDIS (For a write transaction) must release data bus and the data bus parity Input to high impedance during the following cycle Data Bus Grant DBG May, with the proper qualification, assume mastership of the data bus Input Data Bus Write Only DBWO May run the data bus tenure Input Data Bus Parity DP[0-7] If output, odd parity for each of 8 bytes of data write transactions VO If input, odd parity for each byte of read data Data Parity Error PE Incorrect data bus parity Output Data Retry DRTRY Must invalidate the data from the previous read operation Input Global GBL If output, a transaction is global VO If input, a transaction must be snooped by the 603r Hard Reset HRESET Initiates a complete hard reset operation Input Interrupt INT Initiates an interrupt if bit EE of MSR register is set Input LSSD_MODE_ |LSSD test control signal for factory use only Input L1_TSTCLK LSSD test control signal for factory use only Input L2_ TSTCLK LSSD test control signal for factory use only Input Machine Check Inter- | MCP Initiates a machine check interrupt operation if the bit ME of MSR regis- | Input rupt ter and bit EMCP of HIDO register are set PLL Configuration PLL_CFG[0-3] | Configures the operation of the PLL and the internal processor clock Input frequency 8/38 % S SPECHIQUESTSPC603R Signal name Mnemonic Signal function Signal type Quiescent QACK All bus activity has terminated and the 603r may enter a quiescent (or Input Acknowledge low power) state Quiescent Request QREQ Is requesting all bus activity normally to enter a quiescent (low power) Output state Reservation RSRV Represents the state of the reservation coherency bitin the reservation | Output address register System Management | SMI Initiates a system management interrupt operation if the bit EE of MSR Input Interrupt register is set Soft Reset SRESET Initiates processing for a reset exception Input System Clock SYSCLK Represents the primary clock input for the 603r, and the bus clock fre- Input quency for 603r bus operation Transfer Acknowledge | TA A single-beat data transfer completed successfully or a data beat in a Input burst transfer completed successfully Timebase Enable TBEN The timebase should continue clocking Input Transfer Burst TBST If output, a burst transfer is in progress VO If input, when snooping for single-beat reads Transfer Code TC[O-1] Special encoding for the transfer in progress Output Test clock TCK Clock signal for the IEEE P1149.1 test access port (TAP) Input Test data input TDI Serial data input for the TAP Input Test data output TDO Serial data output for the TAP Output Transfer Error TE A bus error occurred Input Acknowledge TLBI Syne TLBISYNC Instruction execution should stop after execution of a tlbsyne instruction | Input Test mode select TMS Selects the principal operations of the test-support circuitry Input Test reset TRST Provides an asynchronous reset of the TAP controller Input Transfer Size TSIZ[0-2] For memory accesses, these signals along with TBST indicate the data | I/O transfer size for the current bus operation Transfer start TS If output, begun a memory bus transaction and the address bus and VO transfer attribute signals are valid If input, another master has begun a bus transaction and the address bus and transfer attribute signals are valid for snooping (see GBL) Transfer Type TT[0-4] Type of transfer in progress VO Write-Through WT A single-beat transaction is write-through Output Power supply indicator | VOLTDETGND | Available only on BGA package Output Indicates to the power supply that a low-voltage processor is present. 9/38TSPC603R B. DETAILED SPECIFICATIONS 1. SCOPE This drawing describes the specific requirements for the microprocessor TSPC603r, in compliance with MIL-STD-883 class Bor TCS standard screening. 2. APPLICABLE DOCUMENTS 1) MIL-STD-883 : Test methods and procedures for electronics. 2) MIL-PRF-38535 : General specifications for microcircuits. 3. REQUIREMENTS 3.1. General The microcircuits are in accordance with the applicable documents and as specified herein. 3.2. Design and construction 3.2.1.Terminal connections The terminal connections shall be is shown in Figure 15 ( B. DETAILED SPECIFICATIONS) and Figure 3 ( A. GENERAL DESCRIPTION). 3.2.2.Lead material and finish Lead material and finish shall be as specified in MIL-STD-1835 (see enclosed 8) 3.3. Absolute maximum ratings Absolute maximum ratings are stress rating only and functional operation at the maximum is not guaranteed. Stresses beyond those listed may affect device reliability or cause permanent damage to the device Table 5 : Absolute maximum rating for the 603r Parameter Symbol Min Max Unit Core supply voltage Vdd -0.3 2.75 Vv PLL supply voltage AVdd -0.3 2.75 Vv I/O supply voltage OVag -0.3 3.6 Vv Input voltage Vin -0.3 5.5 Vv Storage temperature range Tstg -55 +150 C Notes: 1. Functional operating conditions are given in AC and DC electrical specifications. Stresses beyond the absolute maximums listed may affect device reliability or cause permanent damage to the device. 2. Caution : Input voltage must not be greater than OVdd by more than 2.5 V at any times, including during power-on reset. 3. Caution : OVdd voltage must not be greater than Vdd/AVdd by more than 1.2 V at any times, including during power-on reset. 4. Caution : Vdd/AVdd voltage must not be greater than OVdd by more than 0.4 V at any times, including during power-on reset. 10/38 S SPECHIQUESTSPC603R 3.4. Recommanded Operating Conditions These are the recommanded and tested operating conditions. Proper device operation outside of these conditions is not guaranteed. Parameter Symbol Min Max Unit Core supply voltage Vdd 2.375 2.625 Vv PLL supply voltage AVdd 2.375 2.625 Vv I/O supply voltage OVag 3.135 3.465 Vv Input voltage Vin GND 5.5 Vv Operating temperature Te -55 +125 C 3.5. Thermal characteristics The data found in this section concerns 603rs packaged in the 255-lead 21 mm multi-layer ceramic (MLC), ceramic BGA package. Data is shown for the case of using the Thermalloy #2328B heat sink. The internal thermal resistance for this package is negligible due to the exposed die design. A thermal interface material is recom- mended at the package lidtoheat sink interface to minimize the thermal contact resistance. Additionally, the CBGA package offers an excellent thermal connection to the card and power planes. Heat generated at the chip is dissipated through the package, the heat sink (when used) and the card. The parallel heat flow paths result in the lowest overall thermal resistance as well as offer significaltly better power dissipation capability if a heat sink is not used. The thermal characteristics for the flip-chip CBGA and CI-CGA packages are as follows : Thermal resistance (junction-to-case) = Rig oF vic = 0.095C/Watt for the 2 packages. Thermal resistance (junction-to-ball) = Rip or vjp = 3.5C/Watt for the CBGA package. Thermal resistance (junction-to-bottom SCI) = Rig or vjg = 3.7C/Watt for the CI-CGA package. The junction temperature can be calculated from the junction to ambient thermal resistance, as follow: Junction temperature: Tj = Tat (Ric + Reg + Rea) * P Where: T, is the ambient temperature in the vicinity of the device Ric is the die junctiontocase thermal resistance of the device Reg is the casetoheat sink thermal resistance of the interface marerial Rea is the heat sink-toambient thermal resistance P is the power dissipated by the device During operation, the diejunction temperatures (Tj) should be maintened less than the value specified in Table 5. The thermal resistance of the thermal interface material (Rg) is typically about 1C/Waitt. Assuming a T, of 85C and a consumption (P) of 3.6 Watts, the junction temperature of the device would be as follow: Tj= 85C + (0.095C/Watt + 1C/Watt + Rega) * 3.5 Watts. For the Thermalloy heat sink #2328B, the heat sink-toambient thermal resistance (Req) versus airflow velocity is shown in Figure 4. Rsa (C/W) 71 Do 2 64 Ss a 54 DO oc a 4, E 34 = KF e 8 YO 3 15 oD < 0 ; 0 1 2 3 Approach air velocity (m/sec) Figure 4 : CBGA thermal management example Assuming an air velocity of 1.0 m/sec, the associated overall thermal resistance and junction temperature, found in Table 6 will result. 11/38TSPC603R Table 6 : Thermal resistance and junction temperature Configuration Rja ((C/W) TCC) With 2328B heat sink 5.0 106 Vendors such as Aavid Engineering Inc., Thermalloy, and Wakefield Engineering can supply heat sinks with a wide range of thermal performance. 3.6. Power consideration The PowerPC603r is a microprocessor specifically designed for low-power operation. As the 603e microprocessor version, the 603r provides both automatic and program-controllable power reduction modes for progressive reduction of power consumption. This chapter describes the hardware support provided by the 603r for power management. 3.6.1.Dynamic Power Management Dynamic power management automatically powers up and down the individual execution units of the 603r, based upon the contents of the instruction stream. For example, if no floating-point instructions are being executed, the floating-point unit is automatically pow- ered down. Power is not actually removed from the execution unit ; instead, each execution unit has an independent clock input, which is automatically controlled on a clock-by- clock basis. Since CMOS circuits consume negligible power when they are not switching, stopping the clock to an execution unit effectively eliminates its power consumption. The operation of DPM is completely transparent to software or any external hardware. Dynamic power management is enabled by setting bit 11 in HIDO on power-up, of following HRESET. 3.6.2.Programmable Power Modes The 603r provides four programmable power states - full power, doze, nap and sleep. Software selects these modes by setting one (and only one) of the three power saving mode bits. Hardware can enable a power management state through external asynchronous interrupts The hardware interrupt causes the transfer of program flow to interrupt handler code. The appropriate mode is then set by the software. The 603r provides a separate interrupt and interrupt vector for power management - the system management interrupt (SMI). The 603r also contains adecrementtimer which allows it to enter the nap or doze mode for a predetermined amountof time and then return to full power operation through the decrementer interrupt (DI). Note that the 603r cannot switch from on power manage- ment mode to another without first returning to fullon mode. The nap and sleep modes disable bus snooping ; therefore, a hardware handshake is provided to ensure coherency before the 603r enters these power management modes. Table 7 summarizes the four power states. Table 7 : Power PC 603r Microprocessor Programmable Power Modes PM Mode Functioning Units Activation Method Full-Power Wake Up Method Full power All units active - - Full power (with DPM) Requested logic by By instruction dispatch - demand Doze - Bus snooping Controlled by SW External asynchronous exceptions - Data cache as needed Decrementer interrupt - Decrementer timer Reset Nap Decrementer timer Controlled by hardware and | External asynchronous exceptions software Decrementer interrupt Reset Sleep None Controlled by hardware and | External asynchronous exceptions software Reset * Exceptions are referred to as interrupts in the architecture specification 3.6.3.Power Management Modes The following sections describe the characteristics of the 603rs power management modes, the requirements for entering and exit- ing the various modes, and the system capabilities provided by the 603r while the power management modes are active. 3.6.3.1. Full-Power Mode with DPM Disabled Full-power mode with DPM disabled power mode is selected when the DPM enable bit (bit 11) in HIDO is cleared. 12/38 S SB SPEER ESTSPC603R - Default state following power-up and HRESET. - All functional units are operating at full processor speed at all times. 3.6.3.2. Full-Power Mode with DPM Enabled Full-power mode with DPM enabled (HIDO[11] = 1) provides on-chip power management without affecting the functionality or perfor- mance of the 603r. - Required functional units are operating at full processor speed. - Functional units are clocked only when needed. - No software or hardware intervention required after mode is set. - Software/hardware and performance transparent. 3.6.3.3. Doze Mode Doze ode disables most functional units but maintains cache coherency by enabling the bus interface unit and snooping. A snoop hit will cause the 603r to enable the data cache, copy the data back to memory, disable the cache, and fully return to the doze state. @ Most functional units disabled. @ Bus snooping and time base/decrementer still enabled. e@ Dose mode sequence : - Set doze bit (HIDO[8) = 1). - 603r enters doze mode after several processor clocks. @ Several methods of returning to full-power mode : - Assert INT, SMI, MCP or decrementer interrupts. - Assert hard reset or soft reset. @ Transition to full-power state takes no more than a few processor cycles. @ PLL running and locked to SYSCLK. 3.6.3.4. Nap Mode The nap mode disables the 603r but still maintains the phase locked loop (PLL) and the time base/decrementer. The time base canbe used to restore the 603r to full-on state after a programmed amount of time. Because bus snooping is disabled for nap and sleep mode, a hardware handshake using the quiesce request (QREQ) and quiesce acknowledge (QACK) signals are requires to maintain data coherency. The 603r will assert the QREQ signal to indicate that it is ready to disable bus snooping. When the system has ensured that snooping is no longer necessary, it will assert QACK and the 603r will enter the sleep or nap mode. @ Time base/decrementer still enabled. Most functional units disabled (including bus snooping). @ All nonessential input receivers disables. e Nap mode sequence : - Set nap bit (HIDO[9] = 1). - 603r asserts quiesce request (QREQ) signal. - System asserts quiesce acknowledge (QACK) signal. - 603r enters sleep mode after several processor clocks. @ Several methods of returning to full-power mode : - Assert INT, SPI, MCP or decrementer interrupts. - Assert hard reset or soft reset. e Transition to full-power takes no more than a few processor cycles. @ PLL running and locked to SYSCLK. 3.6.3.5. Sleep Mode Sleep mode consumes the least amount of power of the four modes since all functional units are disabled. To conserve the maximum amount of power, the PLL may be disabled and the SYSCLK may be removed. Due to the fully static design of the 603r, internal processor state is preserved when no internal clock is present. Because the time base and decrementer are disabled while the 603r is in sleep mode, the 603rs time base contents will have to be updated from an external time base following sleep mode if accurate time-of-day maintenance is required. Before the 603r enters the sleep mode, the 603r will assert the QREQ signal to indicate that it is ready to disable bus snooping. When the system has ensured that snooping is no longer necessary, it willassert QACK and the 603r will enter the sleep mode. @ All functional units disabled (including bus snooping and time base). @ All nonessential input receivers disabled : - Internal clock regenerators disabled. - PLL still running (see below). e@ Sleep mode sequence : - Set sleep bit (HIDO[10] = 1). - 603r asserts quiesce request (QREQ). - System asserts quiesce acknowledge (QACK). - 603r enters sleep mode after several processor clocks. 13/38TSPC603R @ Several methods of returning to full-power mode : - Assert INT, SMI, or MCP interrupts. - Assert hard reset or soft reset. @ PLL may be disabled and SYSCLK may be removed while in sleep mode. e@ Return to full-power mode after PLL and SYSCLK disabled in sleep mode : - Enable SYSCLK. - Reconfigure PLL into desired processor clock mode. - System logic waits for PLL startup and relock time (100 usec). - System logic asserts one of the sleep recovery signals (for example, INT or SMI). 3.6.4.Power Management Software Considerations Since the 603r is a dual issue processor with out -of-order execution capability, care must be taken in how the power management mode is entered. Furthermore, nap and sleep modes require all outstanding bus operations to be completed before the power man- agement mode is entered. Normally during system configuration time, one of the power management modes would be selected by setting the appropriate HIDO mode bit. Later on, the power management mode is invoked by setting the MSR[POW] bit. To provide a clean transition into and out of the power management mode, the stmsr[POW] should be preceded by a sync instruction and fol- lowed by an isyne instruction. 3.6.5.Power dissipation Table 8 : Power dissipation Vdd/AVdd = 2.545% V de, OVdd =3.34+5% Vdc, GND =0 Vdc, 0C < Te < 125C CPU clock Frequency 166 MHz | 200 MHz | 233 MHz | 266 MHz | 300 MHz | Units Full-On Mode (DPM Enabled) Typical 2.1 2.5 3.0 3.5 4.0 WwW Max 3.2 4.0 4.6 5.3 6.0 Ww Doze Mode Typical | 1.5 1.7 1.8 2.0 2.1 Ww Nap Mode Typical | 100 120 140 160 180 mw Sleep Mode Typical | 96 110 123 135 150 mw Sleep Mode-PLL Disabled Typical 60 60 60 60 60 mW Sleep Mode-PLL and SYSCLK Disabled Typical 25 25 25 25 25 mW Maxi- 60 60 60 80 100 mw mum Notes: 1 These values apply for all valid PLL_CFG[0-3] settings and do not include output driver power (OVDD) or analog supply power (AVDD). OVDD power is system dependent but is typically < 10% of VDD. Worst-case AVDD = 15 mW. 2 Typical power is an average value measured at VDD=AVDD=2.5 V, OVV=3.3 V, in a system executing typical applications and benchmark sequences. 3 Maximum power is measured at VDD=2.625 V using a worst-case instruction mix. 4 To calculate the power consumption at low temperature (55 C), use a factor of 1.25. 14/38 i SB SPEER ESTSPC603R 3.7. Marking The document where are defined the marking are identified in the related reference documents. Each microcircuit are legible and permanently marked with the following information as minimum : - Thomson logo, - Manufacturers part number, - Class B identification if applicable, - Date-code of inspection lot, - ESD identifier if available, - Country of manufacturing. 4. ELECTRICAL CHARACTERISTICS 4.1. General requirements Allstatic and dynamic electrical characteristics specified for inspection purposes and the relevant measurement conditions are given below : - Table 9 : Static electrical characteristics for the electrical variants. - Table 10 : Dynamic electrical characteristics for the 603r. These specifications are for 166 MHz to 300 MHz processor core frequencies. The processor core frequency is determined by the bus (SYSCLK) frequency and the settings of the PLL_CFGO to PLL_CFG3 signals. All timings are specified respective to the rise edge of SYSCLK. 4.2. Static characteristics Table 9 : Electrical characteristics Vdd = AVdd = 2.5V+5% ; OVdd =3.3+5 % V de, GND = 0 Vdc, -55C < Te < 125C Characteristics Symbol Min Max Unit Input high voltage (all inputs except SYSCLK) Vin 2.0 5.5 Vv Input low voltage (all inputs except SYSCLK) VIL GND 0.8 Vv SYSCLK input high voltage CVin 2.4 5.5 Vv SYSCLK input low voltage CVIL GND 0.4 Vv Input leakage current Vin = 3.465 VL 3) lin - 30 uA Vin = 5.5 VU1 9) lin - 300 uA Hi-Z (off-state) Vin = 3.465 V1.3) TsI - 30 uA leakage current Vin = 5.5 V1.3) tsI - 300 LA Output high voltage lon =-7 mA Vou 2.4 - Vv Output low voltage lol =+7 mA VoL - 0.4 Vv Capacitance, Vin = 0 V, f = 1 MHz(@) Cin - 10.0 pF (excludes TS, ABB, DBB, and ARTRY) Capacitance, Vin = 0 V, f = 1 MHz(?) Cin - 15.0 pF (for TS, ABB, DBB, and ARTRY) Notes: 1. Excludes test signals (LSSD_MODE, L1_TSTCLK, L2_TSTCLK, and JTAG signals). 2. Capacitance is periodically sampled rather than 100 % tested. 3. Leakage currents are measured for nominal OVdd and Vdd or both OVdd and Vdd. Same variation (for example, both Vdd and OVdd vary by either +5 % or 5 %). 15/38TSPC603R 4.3. Dynamic characteristics 4.3.1.Clock AC specifications Table 10 provides the clock AC timing specifications as defined in Figure 5. Table 10 : Clock AC timing specifications Vdd = AVdd = 2.5V15% ; OVdd =3.345 % V de, GND = 0 Vdc, -55C < Te $ 125C Num Characteristics 166 MHz 200 MHz 233 MHz 266 MHz 300 MHz | Unit | Note Min | Max | Min | Max | Min | Max | Min | Max | Min | Max Processor frequency 150 | 166 | 150 | 200 | 180 | 233 | 180 | 266 | 180 | 300 | MHz 5 VCO frequency 300 | 332 | 300 | 400 | 360 | 466 | 360 | 532 | 360 | 600 | MHz] 5 SYSCLK (bus) frequency 25 | 66.7 | 33.3 | 66.7 | 33.3] 75 | 33.3 | 75 | 33.3} 75 | MHz] 5 1 | SYSCLK cycle time 15 30 | 13.3] 30 | 13.3] 30 | 133) 30 | 13.3] 30 ns 2,3 | SYSCLK rise and fall time - 2.0 - 2.0 - 2.0 - 2.0 - 2.0 ns 1 4 | SYSCLK duty cycle (1.4V mea- 40.0 | 60.0 | 40.0 | 60.0 | 40.0 | 60.0 | 40.0 | 60.0 | 40.0 | 60.0] % 3 sured) SYSCLK jitter {+150} - |[+150} - |+150} - |+150} - |+150] ps 2 603r internal PLL relock time - 100 - 100 - 100 - 100 - 100 | us 3,4 Notes: SYSCLK /\ 1. Rise and fall times for the SYSCLK input are measured from 0.4 V to 2.4 V. 2. Cycle-to-cycle jitter is guaranteed by design. 3. Timing is guaranteed by design and characterization, and is not tested. 4. PLL relock time is the maximum amount of time required for PLL lock after a stable Vdd, OVdd, AVdd and SYSCLK are reached during the power-on reset sequence. This specification also applies when the PLL has been disabled and subsequently re-enabled during sleep mode. Also note that HRESET must be held asserted for a minimum of 255 bus clocks after the PLL relock time (100 us) during the power-on reset sequence. 5. Caution : The SYSCLK frequency and PLL_CFG[0-3] settings must be chosen such that the resulting SYSCLK (bus) frequency, CPU (core) frequency, and PLL (VCO) frequency do not exceed their respective maximum or minimum operating frequencies. Refer to the PLL_CFG[0-3] signal description for valid PLL_CFG[0-3] settings. 16/38 Vy A y CVvih VM \ SJ} ow VM = Midpoint Voltage (1.4V) Figure 5 : SYSCLK input timing diagram SB SPEER ESTSPC603R 4.3.2.Input AC specifications Table 11 provides the input AC timing specifications for the 603r as defined in Figure 6 and Figure 7. Table 11 : Input AC timing specifications Vdd = AVdd =2.5V+5%; OVdd =3.3+5% V de, GND = 0 Vde, -55C < Te < 125C Num Characteristics 166,200 MHz | 233,266 MHz 300 MHz Unit | Note Min | Max | Min | Max | Min | Max 10a _ | Address/data/transfer attribute inputs valid to SYSCLK | 2.5 - 2.5 - 2.5 - ns 2 (input setup) 10b | All other inputs valid to SYSCLK (input setup) 4.0 - 3.5 - 3.5 - ns 3 10c | Mode select inputs valid to HRESET (input setup) (for 8 - 8 - 8 - tsyscik | 4,5,6,7 DRTRY, QACK and TLBISYNC) 11a | SYSCLK to address/data/transfer attribute inputs 1.0 - 1.0 - 1.0 - ns 2 invalid (input hold) 11b | SYSCLK to all other inputs invalid (input hold) 1.0 - 1.0 - 1.0 - ns 3 11c | HRESET to mode select inputs invalid (input hold) (for 0 - 0 - 0 - ns 4,6,7 DRTRY, QACK, and TLBISYNC) Notes : ALL INPUTS f 1. All input specifications are measured from the TTL level (0.8 or 2.0 V) of the signal in question to the 1.4 V of the rising edge of the input SYSCLK. Both input and output timings are measured at the pin. See Figure 7. 2. Address/data/transfer attribute input signals are composed of the following: A[O-31], AP[O3], TT[O-4], TC[O-1], TBST, TSIZ[O-2], GBL, DH[0-31], DL[0-31], DP[9-7]. SRESET, INT, SMI, MCP, TBEN, QACK, TLBISYNC. 4. The setup and hold time is with respect to the rising edge of HRESET. See Figure 7. 5. teyscik is the period of the external clock (SYSCLK) in nanoseconds (ns). The numbers given in the table must be multiplied by the period of SYSCLK to compute the actual time duration (in nanoseconds) of the parameter in question. 6. These values are guaranteed by design, and are not tested. 7. This specification is for configuration mode only. Also note that HRESET must be held asserted for a minimum of 255 bus clocks after the PLL relock time (100 us) during the power-on reset sequence. SYSCLK VM = Midpoint Voltage (1.4V) Figure 6 : Input timing diagram 17/38TSPC603R HRESET S ye AL A Z VM = Midpoint Voltage (1.4 V) MODE PINS Figure 7 : Mode select input timing diagram 4.3.3.Output AC specifications Table 12 provides the output AC timing specifications for the 603r (shown in Figure 8). Table 12 : Output AC timing specifications Vdd = AVdd =2.5V+5%; OVdd =3.345% Vde, GND = 0 Vde, C, = 50 pF, -55C < Te < 125C Num Characteristic 166,200 MHz | 233,266 MHz 300 MHz Unit | Note Min | Max | Min | Max | Min | Max 12 | SYSCLK to output driven (output enable time) 1.0 - 1.0 - 1.0 - ns 13a | SYSCLK to output valid (5.5 V to 0.8 V - TS, ABB, - 9.0 - 9.0 - 9.0 ns 4 ARTRY, DBB) 13b | SYSCLK to output valid (TS, ABB, ARTRY, DBB) - 8.0 - 8.0 - 8.0 ns 6 14a | SYSCLK to output valid (5.5 V to 0.8 V all except TS, - 11.0 - 11.0 - 11.0 ns 4 ABB, ARTRY, DBB) 14b | SYSCLK to output valid (all except - 9.0 - 9.0 - 9.0 ns 6 TS,ABB,ARTRY,DBB) 15 | SYSCLK to output invalid (output hold) 1.0 - 1.0 - 1.0 - ns 3 16 | SYSCLK to output high impedance (all except ARTRY, - 8.5 - 8.0 - 8.0 ns ABB, DBB) 17 | SYSCLK to ABB, DBB, high impedance after precharge - 1.0 - 1.0 - 1.0 | teyscik | 5, 7 18 | SYSCLK to ARTRY high impedance before precharge - 8.0 - 7.5 - 7.5 ns 19 | SYSCLK to ARTRY precharge enable 0.2* - 0.2* - 0.2* - ns 3, 5, teysclk tsysclk tsysclk 8 + 1.0 + 1.0 + 1.0 20 | Maximum dalay to ARTRY precharge - 1.0 - 1.0 - 1.0 | teysck | 5, 8 21 | SYSCLK to ARTRY high impedance after precharge - 2.0 - 2.0 - 2.0 | teysck | 6, 8 Notes:1 All output specifications are measured from the 1.4 V of the rising edge of SYSCLK to the TTL level (0.8 V or 2.0 V) of the signal in question. 18/38 Both input and output timings are measured at the pin. See Figure 8. 2. All maximum timing specifications assume C= 50 pF. 3. This minimum parameter assumes C, = 0 pF. 4. SYSCLK to output valid (5.5 V to 0.8 V) includes the extra delay associated with discharging the external voltage from 5.5 V to 0.8 V instead of from Vdd to 0.8 V (5 V CMOS levels instead of 3.3 V CMOS levels). 5. teyscik is the period of the external bus clock (SYSCLK) in nanoseconds (ns). The numbers given in the table must be multiplied by the period of SYSCLK to compute the actual time duration (in nanoseconds) of the parameter in question. 6. Output signal transitions from GND to 2.0 V or Vdd to 0.8 V. 7. Nominal precharge width for ABB and DBB is 0.5 * teyscik. 8. Nominal precharge width for ARTRY is 1.0 * teyscik- S SPECHIQUESSYSCLK ALL OUTPUTS (Except TS, ABB DBB, ARTRY) 1S > qq Oo o oO TSPC603R @Q@ VM = Midpoint Voltage (1.4 V) Figure 8 : Output timing diagram 19/38TSPC603R 4.4. JTAG AC timing specifications Table 13 : JTAG AC timing specifications (independent of SYSCLK) Vdd = AVdd =2.5V+5%; OVdd =3.345 % Vde, GND = 0 Vde, C_ = 50 pF, -55C < Tce < 125C Num Characteristic Min Max Unit Notes : TCK frequency of operation 0 16 MHz 1 TCK cycle time 62.5 _ ns 2 TCK clock pulse width measured at 1.4 V 25 _ ns 3 TCK rise and fall times 0 3 ns 4 TRST setup time to TCK rising edge 13 _ ns 1 5 TRST assert time 40 ns 6 Boundary scan input data setup time | 6 _ ns 2 7 Boundary scan input data hold time 27 _ ns 2 8 TCK to output data valid 4 .25 ns 3 9 TCK to output high impedance 3 24 ns 3 10 TMS, TDI data setup time 0 _ ns 11 TMS, TDI data hold time 25 _ ns 12 TCK to TDO data valid 4 24 ns 13 TCK to TDO high impedance 3 15 ns Notes: 1. TRST is an asynchronous signal. The setup time is for test purposes only. 2. Non-test signal input timing with respect to TCK. 3. Non-test signal output timing with respect to TCK. < G) > TCK kVM VM KVM VM = Midpoint Voltage (1.4 V) Figure 9 : Clock input timing diagram TCK 20/38 < G) > ~ es Figure 10 : TRST timing diagram & SPREMAQUESTSPC603R TCK * \, (S77) Data Inputs Ny dj Input Data Vaid Ny - Data Outputs x Output Data Valid Ny kr IN Data Outputs NY Data Outputs Ny q Output Data Valid Figure 11 : Boundary-scan timing diagram TCK Ye N, TDI, TMS Ny q Input Data Valid NY TDO x Output Data Valid Ny K TDO ~ Fr TDO NY q Output Data Valid Figure 12 : Test access port timing diagram 21/38TSPC603R 5. FUNCTIONAL DESCRIPTION 5.1. PowerPC registers and programming model The PowerPC architecture defines register-to-register operations for most computational instructions. Source operands for these instructions are accessed from the registers or are provided as immediate values embedded in the instruction opcode. The three-reg- ister instruction format allows specification of a target register distinct from the two source operands. Load and store instructions transfer data between registers and memory. PowerPC processors have two levels of privilege - supervisor mode of operation (typically used by the operating system) and user mode of operation (used by the application software). The programming models incorporate 32 GPRs, 32 FPRs, special-purpose registers (SPRs) and several miscellaneous registers. Each PowerPC microprocessor also has its own unique set of hardware implementation (HID) registers. Having access to privilege instructions, registers, and other resources allows the operating system to control the application environ- ment (providing virtual memory and protecting operating-system and critical machine resources). Instructions that control the state of the processor, the address translation mechanism, and supervisor registers can be executed only when the processor is operating in supervisor mode. The following sections summarize the PowerPC registers that are implemented in the 603r. 5.1.1.General-Purpose Registers (GPRs) The PowerPC architecture defines 32 user-level, general-purpose registers (GPRs). These registers are either 32 bits wide in 32-bit PowerPC microprocessors and 64 bits wide in 64-bit PowerPC microprocessors. The GPRs serve as the data source or destination for all integer instructions. 5.1.2.Floating-Point Registers (FPRs) The PowerPC architecture also defines 32 user-level, 64-bit floating-point registers (FPRs). The FPRs serve as the data source or destination for floating-point instructions. These registers can contain data objects of either single - or double - precision floating-point formats. 5.1.3.Condition Register (CR) The CR is a 32-bit user-level register that consists of eight four-bit fields that reflect the results of certain operations, such as move, integer and floating-point compare, arithmetic, and logical instructions, and provide a mechanism for testing and branching. 5.1.4.Floating-Point Status and Control Register (FPSCR) The floating-point status and control register (FPSCR) is a user-level register that contains all exception signal bits, exception sum- mary bits, exception enable bits, and rounding control bits needed for compliance with the IEEE 754 standard. 5.1.5.Machine State Register (MSR) The machine state register (MSR) is a supervisor-level register that defines the state of the processor. The contents of this register are saved when an exception is taken and restored when the exception handling completes. The 603r implements the MSR as a 32-bit register, 64-bit PowerPC processors implement a 64-bit MSR. 5.1.6.Segment Registers (SRs) For memory management, 32-bit PowerPC microprocessors implement sixteen 32-bit segment registers (SRs). To speed access, the 603r implements the segment registers as two arrays ; a main array (for data memory accesses) and a shadow array (for instruc- tion memory accesses). Loading a segment entry with the Move to Segment Register (stsr) instruction loads both arrays. 5.1.7.Special-Purpose Registers (SPRs) The powerPC operating environment architecture defines numerous special-purpose registers that serve a variety of functions, such as providing controls, indicating status, configuring the processor, and performing special operations. During normal execution, a program can access the registers, shown in Figure 13, depending on the programs access privilege (supervisor or user, determined by the privilege-level (PR) bit in the MSR). Note that register such as the GPRs and FPRs are accessed through operands that are part of the instructions. Access to registers can be explicit (that is, through the use of specific instructions for that purpose such as Move to Special-Purpose Register (mtspr) and Move from Special-Purpose Register (mfspr) instructions) or implicit, as the part of the execution of an instruction. Some registers are accessed both explicitly and implicitly. ll the 603r, all SPRs are 32 bits wide. 5.1.7.1. User-Level SPRs The following 603r SPRs are accessible by user-level software : @ Link register (LR) - The link register can be used to provide the branch target address and to hold the return address after branch and link instructions. The LR is 32 bits wide in 32-bit implementations. Count register (CTR) - The CRT is decremented and tested automatically as a result of branch-and-count instructions. The CTR is 32 bits wide in 32-bit implementations. @ Integer exception register (XER) - The 32-bit XER contains the summary overflow bit, integer carry bit, overflow bit, and a field specifying the number of bytes to be transferred by a Load String Word Indexed (Iswx) or Store String Word Indexed (stswx) instruction. 22/38 & SB SPEER ESTSPC603R 5.1.7.2. Supervisor-Level SPRs The 603r also contains SPRs that can be accessed only by supervisor-level software. These registers consist of the following : The 32-bit DSISR defines the cause of data access and alignment exceptions. The data address register (DAR) is a 32-bit register that holds the address of an access after an alignment or DSI exception. Decrementer register (DEC) is a 32-bit decrementing counter that provides a mechanism for causing a decrementer exception after a programmable delay. The 32-bit SDR1 specifies the page table format used in virtual-to-physical address translation for pages. (Note that physical address is referred to as real address in the architecture specification). The machine status save/restore register 0 (SRRO) is a 32-bit register that is used by the 603r for saving the address of the instruc- tion that caused the exception, and the address to return to when a Return from Interrupt (rfi) instruction is executed. The machine status save/restore register 1 (SRR1) is a 32-bit register used to save machine status on exceptions and to restore machine status when an rfi instruction is executed. @ The 32-bit SPRGO-SPRG3 registers are provided for operating system use. e@ The external access register (EAR) is a 32-bit register that controls access to the external control facility through the External Control In Word Indexed (eciwx) and External Control Out Word Indexed (ecowx) instructions. The time base register (TB) is a 64-bit register that maintains the time of day and operates interval timers. The TB consists of two 32-bit fields - time base upper (TBU) and time base lower (TBL). The processor version register (PVR) is a 32-bit, read-only register that identifies the version (model) and revision level of the PowerPC processor. Block address translation (BAT) arrays - The PowerPC architecture defines 16 BAT registers, divided into four pairs of data BATs (DBATs) and four pairs of instruction BATs (IBATs). See Figure 13 for a list of the SPR numbers for the BAT arrays. The following supervisor-level SPRs are implementation-specific to the 603r : The DMISS and IMISS registers are read-only registers that are loaded automatically upon an instruction or data TLB miss. The HASH1 and HASH2 registers contain the physical addresses of the primary and secondary page table entry groups (PTEGs). The ICMP and DCMP registers contain a duplicate of the first word in the page table entry (PTE) for which the table search is looking. The required physical address (RPA) register is loaded by the processor with the second word of the correct PTE during a page table search. The hardware implementation (HIDO and HID1) registers provide the means for enabling the 603rs checkstops and features, and allows software to read the configuration of the PLL configuration signals. The instruction address breakpoint register (IABR) is loaded with an instruction address that is compared to instruction addresses in the dispatch queue. When an address match occurs, an instruction address breakpoint exception is generated. Figure 13 shows all the 603r registers available at the user and supervisor level. The number to the right of the SPRs indicate the number that is used in the syntax of the instruction operands to access the register. 23/38TSPC603R 7 ee 24/38 USER MODEL General-Purpose Registers GPRO GPRI1 GPR31 Floating-Point Registers FPRO FPRi FPR31 Condition Register CR Floating-Point Status and Control Register FPSCR XE XER a SPR 1 Link Register SPR8 : Count Register SPR 9 Time Base Facility (For Reading) TBL TBR 268 TBU TBR 269 SUPERVISOR MODEL Hardware implementation Machine State Processor Version Registers . Register Register HIDO | SPR1008 SPR 287 HID1 SPR1 009 Memory Management Registers Instruction BAT . Software Tabla Registers Data BAT Registers Search Registers IBATOU | SPR 528 DBATOU | SPR 536 DMISS | SPR976 IBATOL | SPR S29 DBATOL | SPR537 DCMP | SPR977 IBAT1U | SPR 530 DBAT1U | SPR 538 HASH1 | SPR978 IBATIL | SPR S31 DBATIL | SPR539 HASH2 | SPR979 IBAT2U | SPR 532 DBAT2U | SPR S40 iMISS SPR 980 IBAT2L | SPR 33 DBAT2L | SPR 541 ICMP | SPR981 IBAT3U | SPR 534 DBAT3U | SPR 542 RPA SPR 982 IBAT3L | SPR 535 DBAT3L | SPR543 . Segment Registers SDRi SRO : SRI15 Exception Handling Registers Data Address Register DSISR SPR 19 SPRGs SPRGO | SPR 272 SPRG1 | SPR 273 SPRG2 | SPR 274 SPRG3 | SPR 275 Configuration Registers SPR 18 Save and Restore Miscellaneous Registers Time Base Facility (For Writing) TBL SPR 284 TBU SPR 285 Instruction Address Breakpoint Register SPR 1010 SRRO SPR 26 SRR1 SPR 27 Decrementer DEC SPR 22 External Address Register (Optional) SPR 282 (1) These registers are 603r-specific registers. Tey may not be supported by other PowerPC processors. Figure 13 : PowerPC microprocessor programming model - Register \ TS & SPREMAQUESTSPC603R 5.2. Instruction set and addressing modes The following subsections describe the PowerPC instruction set and addressing modes in general. 5.2.1.PowerPC instruction set and addressing modes All PowerPC instructions are encoded as single-word (32-bit) opcodes. Instruction formats are consistent among all instruction types, permitting efficient decoding to occur in parallel with operand accesses. This fixed instruction length and consistent format greatly simplifies instruction pipelining. 5.2.1.1. PowerPC instruction set The PowerPC instructions are divided into the following categories : @ Integer instructions - These include computational and logical instructions. - Integer arithmetic instructions. - Integer compare instructions. - Integer logical instructions. - Integer rotate and shift instructions. e@ Floating-point instructions -These include floating-point computational instructions, as well as instructions that affect the FPSCR. - Floating-point arithmetic instructions. - Floating-point multiply/add instructions. - Floating-point rounding and conversion instructions. - Floating-point compare instructions. - Floating-point status and control instructions. e Load/store instructions - These include integer and floating-point load and store instructions. - Integer load and store instruction. - Integer load and store multiple instructions. - Floating-point load and store. - Primitives used to construct atomic memory operations (lwarx and stwex. instructions). e@ Flow control instructions - These include branching instructions, condition register logical instructions, trap instructions, and other instructions that affect the instruction flow. - Branch and trap instructions. - Condition register logical instructions. e Processor control instructions - These instructions are used for synchronizing memory accesses and management of caches, TLBs, and the segment registers. - Move to/from SPR instructions. - Move to/from MSR. - Synchronize. - Instruction synchronize. @ Memory control instruction - These instructions provide control of caches, TLBs, and segment registers. - Supervisor-level cache management instructions. - User-level cache instructions. - Segment register manipulation instructions. - Translation lookaside buffer management instructions. Note that this grouping of the instructions does not indicate which execution unit executes a particular instruction or group of instruc- tions. Integer instructions operate on byte, half-word, and word operands. Floating-point instructions operate on single-precision (one word) and double-precision (one double word) floating-point operands. The PowerPC architecture uses instructions that are four bytes long and word-aligned. It provides for byte, half-word, and word operand loads and stores between memory and a set of 32 GPRs. It also provides for word and double-word operand loads and stores between memory and a set of 32 floating-point registers (FPRs). Computational instructions do not modify memory. To use a memory operand in acomputation and then modify the same or another memory location, the memory contents must be loaded into a register, modified, and then written back to the target location with distinct instructions. PowerPC processors follow the program flow when they are in the normal execution state. However, the flow of instructions can be interrupted directly by the execution of an instruction or by an asynchronous event. Either kind of exception may cause one of several components of the system software to be invoked. 5.2.1.2. Calculating effective addresses The effective address (EA) is the 32-bit address computed by the processor when executing a memory access or branch instruction or when fetching the next sequential instruction. 25/38TSPC603R The PowerPC architecture supports two simple memory addressing modes : e EA =(RA(0) + offset (including offset = 0) (register indirect with immediate index). @ EA =(RA\0) + rB (register indirect with index). These simple addressing modes allow efficient address generation for memory accesses. Calculation of the effective address for aligned transfers occurs in a single clock cycle. Foramemory access instruction, ifthe sum of the effective address and the operand length exceeds the maximum effective address, the memory operand is considered to wrap around from the maximum effective address to effective address 0. Effective address computations for both data and instruction accesses use 32-bit unsigned binary arithmetic. A carry from bit 0 is ignored in 32-bit implementations. 5.2.2.PowerPC 603r microprocessor instruction set The 603r instruction set is defined as follows : @ The 603r provides hardware support for all 32-bit PowerPC instructions. e@ The 603r provides two implementation-specific instructions used for software table search operations following TLB misses : - Load Data TLB Entry (tlbld). - Load Instruction TLB Entry (tlbli). e@ The 603r implements the following instructions which are defined as optional by the PowerPC architecture : - External Control In Word Indexed (eciwx). - External Control Out Word Indexed (ecowx). - Floating Select (fsed). - Floating Reciprocal Estimate Single-Precision (fres). - Floating Reciprocal Square Root Estimate (frsqrte). - Store Floating-Point as Integer Word (stfiwx). 5.3. Cache implementation The following subsections describe the PowerPC architectures treatment of cache in general, and the 603r specific implementation, respectively. 5.3.1.PowerPC cache characteristics The PowerPC architecture does not define hardware aspects of cache implementations. For example, some PowerPC processors, including the 603r, have separate instruction and data caches (harvare architecture), while others, such as the PowerPC 601 micro- processor, implement a unified cache. PowerPC microprocessor control the following memory access modes on a page or block basis : @ Write-back/write-through mode. @ Cache-inhibited mode. e@ Memory coherency. Note that in the 603r, a cache line is defined as eight words. The VEA defines cache management instructions that provide a means by which the application programmer can affect the cache contents. 5.3.2.PowerPC 603r microprocessor cache implementation The 603r has two 16-Kbyte, four-way set-associative (instruction and data) caches. The caches are physically addressed, and the data cache can operate in either write-back or write-through mode as specified by the PowerPC architecture. The data cache is configured as 128 sets of 4 lines each. Each line consists of 32 bytes, two state bits, and an address tag. The two state bits implement the three-state MEI (modified/exclusive/invalid) protocol. Each line contains eight 32-bit words. Note that the PowerPC architecture defines the term block as the cacheable unit. For the 603r, the block size is equivalent to a cache line. A block diagram of the data cache organization is shown in Figure 14. The instruction cache also consists of 128 sets of 4 lines, and each line consists of 32 bytes, an address tag, and a valid bit. The instruction cache may not be written to except through a line fill operation. The instruction cache is not snooped, and cache coherency must be maintained by software. A fast hardware invalidation capability is provided to support cache maintenance. The organization of the instruction cache is very similar to the data cache shown in Figure 14. Each cache line contains eight contiguous words from memory that are loaded from an 8-word boundary (that is, bits A27-A32 of the effective addresses are zero) ; thus, acache line never crosses a page boundary. Misaligned accesses across a page boundary can incur a performance penalty. The 603s cache lines are loaded in four beats of 64 bits each. The burst load is performed as "critical double word first. The cache that is being loaded is blocked to internal accesses until the load completes. The critical double word is simultaneously written to the cache and forwarded to the requesting unit, thus minimizing stalls due to load delays. To ensure coherency among caches in a multiprocessor (or multiple caching-device) implementation, the 603r implemements the MEI protocol. These three states, modified, exclusive, and invalid, indicate the state of the cache block as follows : Modified - The cache line is modified with respect to system memory ; that is, data for this address is valid only in the cache and not in system memory. 26/38 & SB SPEER ESTSPC603R e@ Exclusive - This cache line holds valid data that is identical to the data at this address in system memory. No other cache has this data. @ Invalid - This cache line does not hold valid data. Cache coherency is enforced by on-chip bus snooping logic. Since the 603rs data cache tags are single ported, a simultaneous load or store and snoop access represent a resource contention. The snoop access is given first access to the tags. The load or store then occurs on the clock following snoop. t q t t i q e 128 Sets |_| * {}}- 9p + ++ 4 e e J J I L i t t t t I t q | Block 0| Address Tag 0 -] State Words 0-7 | Block 1] Address Tag 1 - State Words 0-7 | t i q t 4 q t = Block 2} Address Tag 2 State Words 0-7 |_| Block 3| Address Tag 3 State Words 0-7 LS ~<__ 8 Words/Biock ___> Figure 14 : Data cache organization 5.4. Exception model The following subsections describe the PowerPC exception model and the 603r implementation, respectively. 5.4.1.PowerPC exception model The PowerPC exception mechanism allows the processor to change to supervisor state as a result of external singles, errors, or unusual conditions arising in the execution of instructions, and differ from the arithmetic exceptions defined by the IEEE for floating- point operations. When exceptions occur, information about the state of the processor is saved to certain registers and the processor begins execution at an address (exception vector) predetermined for each exception. Processing of exceptions occurs in supervisor mode. Although multiple exception conditions can map to a single exception vector, a more specific condition may be determined by examin- ing aregister associated with the exception - for example, the DSISR and the FPSCR. Additionally, some exception conditions can be explicitly enable or disabled by software. The PowerPC architecture requires that exceptions be handled in program order ; therefore, although a particular implementation may recognize exception conditions out of order, they are presented strictly in order. When an instruction-caused exception is recog- nized, any unexecuted instructions that appear earlier in the instruction stream, including any that have not yet entered the execute state, are required to complete before the exception is taken. Any exceptions caused by those instructions are handled first. Likewise, exceptions that are asynchronous and precise are recognized when they occur, but are not handled until the instruction currently in the completion state successfully completes execution or generates an exception, and the completed store queue is emptied. Unless a catastrophic causes a system reset or machine check exception, only one exception is handled at atime. If, for example, a single instruction encounters multiple exception conditions, those conditions are encountered sequentially. After the exception hand- ler handles an exception, the instruction execution continues until the next exception condition is encountered. However, in many cases there is no attempt to re-execute the instruction. This method of recognizing and handling exception conditions sequentially guarantees that exceptions are recoverable. Exception handlers should save the information stored in SRRO and SRR1 early to prevent the program state from being lost due toa system reset and machine check exception orto an instruction-caused exception in the exception handler, and before enabling exter- nal interrupts. The PowerPC architecture support four types of exceptions : @ Synchronous, precise - These are causes by instructions. All instruction-caused exceptions are handled precisely ; that is, the machine state at the time the exception occurs is known and can be completely restored. This means that (excluding the trap and system call exceptions) the address of the faulting instruction is provided to the exception handler and that neither the faulting instruction nor subsequent instructions in the code stream will complete execution before the exception is taken. Once the excep- tion is processed, execution resumes atthe address of the faulting instruction (or at an alternate address provided by the exception handler). When an exception is taken due to an trap or system call instruction, execution resumes at an address provided by the handler. @ Synchronous, imprecise - The PowerPC architecture defines two imprecise floating-point exception modes, recoverable and nonrecoverable. Even though the 603r provides a means to enable he imprecise modes, it implements these modes identically to the precise mode (-hat is, all enabled floating-point enabled exceptions are always precise on the 603r). 27/38TSPC603R e@ Asynchronous, maskable - The external, SMI, and decrementer interrupts are maskable asynchronous exceptions. When these exceptions occur, their handling is postponed until the next instruction, and any exceptions associated with that instruction, completes execution. If there are no instructions in the execution units, the exception is taken immediately upon determination of the correct restart address (for loading SRRO). @ Asynchronous, non maskable - There are two non maskable asynchronous exceptions : system reset and the machine check exception. These exceptions may not be recoverable, or may provide a limited degree of recoverability. All exceptions report recoverability through the SMR[RI] bit. 5.4.2.PowerPC 603r microprocessor exception model A specified by the PowerPC architecture, all 603r exceptions can be described as either precise or imprecise and either synchronous or asynchronous. Asynchronous exceptions (some or which are maskable) are caused by events external to the processors execu- tion ; synchronous exceptions, which are all handled precisely by the 603r, are caused by instructions. The 603r exception classes are shown in Table 14. Synchronous/Asynchronous precise/Imprecise Exception type Asynchronous, non maskable Imprecise Machine check System reset Asynchronous, maskable Precise External interrupt Decrementer System management interrupt Synchronous Precise Instruction-caused exceptions Table 15 : PowerPC 603r microprocessor exception classifications Although exceptions have other characteristics as well, such as whether they are maskable or non maskable, the distinctions shown in Table 15 define categories of exceptions that the 603r handles uniquely. Note that Table 15 includes no synchronous imprecise instructions. While the PowerPC architecture supports imprecise handling of floating-point exceptions, the 603r implements these exception modes as precise exceptions. The 603rs exceptions, and conditions that cause them, are listed in Table 16. Exceptions that are specific to the 603r are indicated. 28/38 & SB SPEER ESTSPC603R Table 16 : Exceptions and conditions Exception Type Vector Offset (hex) Causing Conditions Reserved 00000 System reset 00100 Asystem reset is caused by the assertion of either SAESET or HRESET. Machine check 00200 Amachine check is caused by the assertion of the TEA signal during a data bus transaction, assertion of MCP, or an address or data parity error. DSI 00300 The cause of a DSI exception can be determined by the bit settings in the DSISR, listed as follows: 1 Set if the translation of an attempted access is not found in the primary hash table entry group (HTEG), or in the rehashed secondary HTEG, or in the range of a DBAT ragister; otherwise cleared. 4 Set if a memory access is not permitted by the page or DBAT protection mechanism; otherwise cleared. 5 Set by an eciwx or ecowx instruction if the access is to an address that is marked as write-through, or execution of a load/store instruction that accesses a direct-store segment. 6 Set for a store operation and cleared for a load operation. 11 Set if eciwx or ecowx is used and EAR[E] is cleared. ISI 00400 An ISI exception is caused when an instruction fetch cannot be performed for any of the following reasons: * The effective (logical) address cannot be translated. That is, there is a page fault for this portion of the translation, so an ISI exception must be taken to load the PTE (and possibly the page) into memory. * The fetch access violates memory protection. If the key bits (Ks and Kp) in the segment register and the PP bits in the PTE are set to prohibit read access, instructions cannot be fetched from this location. External interrupt 00500 An external interrupt is caused when MSR[EE] = 1 and the INT signal is asserted. Alignment 00600 An alignment exception is caused when the 603e cannot perform a memory access for any of reasons described below: * The operand of a floating-point load or store instruction is not word-aligned. * The operand of imw, stmw, lwarx, and stwex. instructions are not aligned. * The operand of a single-register load or store operation is not aligned, and the 603 is in little-endian mode. * The instruction is imw, stmw, Iswi, Iswx, stswi, stswx and the 6036 is in little endian mode. The operand of debz is in storage that is write-through-required, or caching inhibited. 29/38TSPC603R Exception Vector Offset : sas Type (hex) Causing Conditions Program 00700 Aprogram exception is caused by one of the following exception conditions, which correspond to bit.settings in SRR1 and arise during execution of an instruction: * Floating-point enabled exceptionA floating-point enabled exception condition is generated when the following condition is met: (MSR[FEO} | MSR[FE1)) & FPSCR[FEX] is 1. FPSCR[FEX] is set by the execution of a floating-point instruction that causes an enabled exception or by the execution of one of the move to FPSCR instructions that results in both an exception condition bit and its corresponding enable bit being set in the FPSCR. + Illegal instructionAn illegal instruction program exception is generated when execution of an instruction is attempted with an illegal opcode or illegal combination of opcode and extended opcode fieids (including PowerPC instructions not implemented in the 603e), or when execution of an optional instruction not provided in the 6036 is attempted (these do not include those optional instructions that are treated as no-ops). * Privileged instructionA privileged instruction type program exception is generated when the execution of a privileged instruction is attempted and the MSR register user privilege bit, MSR{PA}, is set. in the 603e, this exception is generated for mtspr or mfspr with an invalid SPR field if SPR[0] = 1 and MSR[PR] = 1. This may not be true for all PowerPC processors. * TrapA trap type prograrn exception is generated when any of the conditions specified in a trap instruction is met. Fioating- 00800 Atloating-point unavailabie exception is caused by an attempt to execute a point floating-point instruction (including floating-point load, store, and move unavailable instructions) when the floating-point available bit is disabled, (MSR[FP] = 0). Dacrementer | 00900 The decrementer exception accurs when the most significant bit of the decrementer (DEC) register transitions from 0 to 1. Must also be enabled with the MSR[EE] bit. Reserved OOAQO-OOBFF | System call oocoo Asystem call exception occurs when a System Call (sc) instruction is executed. Trace ooDoo Atrace exception is taken when MSR[SE] =1 or when the currently completing instruction is a branch and MSR[BE] =1. Reserved oOoE00 The 603e does not generate an exception to this vector. Other PowerPC processors may use this vector for floating-point assist exceptions. Reserved OOE10-OOFFF | Instruction 01000 An instruction translation miss exception is caused when an effective address for translation an instruction fetch cannot be transtated by the ITLB. miss Data load 01100 Adata load translation miss exception is caused when an effective address for a translation data load operation cannot be translated by the DTLB. miss Data store 01200 Adata store translation miss exception is caused when an effective address for a translation data store operation cannot be translated by the DTLB; or where a DTLB hit miss occurs, and the change bit in the PTE must be set due to a data store operation. 30/38 RE SPECIPIQUESTSPC603R Exception Vector Offset Causing Conditions Type (hex) ausing Conditio Instruction 01300 An instruction address breakpoint exception occurs when the address (bits 0- 29) address in the IABR matches the next instruction to complete in the completion unit, and breakpoint the IABR enabie bit (bit 30) is set to 1. System 01400 Asystem management interrupt is caused when MSR[EE] =1 and the SMI input management signal is asserted. interrupt Reserved 01500-02FFF } 5.5. Memory management The following subsections describe the memory management features of the PowerPC architecture, and the 603r implementation, respectively. 5.5.1.PowerPC memory management The primary functions of the MMU are to translate logical (effective) addresses to physical addresses for memory accesses, and to provide access protection on blocks and pages of memory. There are two types of accesses generated by the 603r that require address translation - instruction accesses, and data accesses to memory generated by load and store instructions. The PowerPC MMU and exception model support demand-paged virtual memory. Virtual memory management permits execution of programs larger than the size of physical memory ; demand-paged implies that individual pages are loaded into physical memory from system memory only when they are first accessed by an executing program. The hashed page table is a variable-sized data structure that defines the mapping between virtual page numbers and physical page numbers. The page table size is a power of 2, and its starting address is a multiple of its size. The page table contains a number of page table entry groups (PTEGs). A PTEG contains eight page table entries (PTEs) of eight bytes each ; therefore, each PTEG is 64 bytes long. PTEG addresses are entry points for table search operations. Address translations are enabled by setting bits inthe MSR-MSR[IR] enables instruction address translations and MSR[DR] enables data address translations. 5.5.2.PowerPC 603r microprocessor memory management The instruction and data memory management units in the 603r provide 4 Gbyte of logical address space accessible to supervisor and user programs with a 4-Kbyte page size and 256-Mbyte segment size. Block sizes range from 128 Kbyte to 256Mbyte and are software selectable. In addition, the 603r uses an interim 52-bit virtual address and hashed page tables for generating 32-bit physical addresses. The MMUs in the 603r rely on the exception processing mechanism for the implementation of the paged virtual memory environment and for enforcing protection of designated memory areas. Instruction and data TLBs provide address translation in parallel with the on-chip cache access, incurring no additional time penalty in the event of a TLB hit. A TLB is a cache of the most recently used page table entries. Software is responsible for maintaining the consistency of the TLB with memory. The 603rs TLBs are 64-entry, two-way set-associative caches that contain instruction and data address translations. The 603r provides hardware assist for software table search operations through the ashed page table on TLB misses. Supervisor software can invalidate TLB entries selectively. The 603r also provides independent four-entry BAT arrays for instructions and data that maintain address translations for blocks of memory. These entries define blocks that can vary from 128 Kbyte to 256 Mbyte. The BAT arrays are maintained by system software. As specified by the PowerPC architecture, the hashed page table is a variable-sized data structure that defines the mapping between virtual page numbers and physical page numbers. The page table size is a power of 2, and its starting address is a multiple of its size. Also as specified by the PowerPC architecture, the page table contains a number of page table entry groups (PTEGs). APTEG con- tains eight page table entries (PTEs) of eight bytes each ; therefore, each PTEG is 64 bytes long. PTEG addresses are entry points for table search operations. 5.6. Instruction timing The 603r is a pipelined superscalar processor. A pipelined processor is one in which the processing of an instruction is reduced into discrete stages. Because the processing of an instruction is broken into a series of stages, an instruction does not require the entire resources of an execution unit. For example, after an instruction completes the decode stage, it can pass on to the next stage, while the subsequent instruction can advance into the decode stage. This improves the throughput of the instruction flow. For example, it may take three cycles for a floating-point instruction to complete, but if there are no stalls in the floating-point pipeline, a series of floating-point instructions can have a throughput of one instruction per cycle. 31/38TSPC603R The instruction pipeline in the 603r has four major pipeline stages, described a follows : e@ Thefetch pipeline stage primarily involves retrieving instructions from the memory system and determining the location of the next instruction fetch. Additionally, the BPU decodes branches during the fetch stage and folds out branch instructions before the dis- patch stage if possible. e@ The dispatch pipeline stage is responsible for decoding the instructions supplied by the instruction fetch stage, and determining which of the instructions are eligible to be dispatched in the current cycle. in addition, the source operands of the instructions are read from the appropriate register file and dispatched with the instruction to the execute pipeline stage. At the end of the dispatch pipeline stage, the dispatched instructions and their operands are latched by the appropriate execution unit. @ During the execute pipeline stage each execution unit that has an executable instruction executes the selected instruction (per- haps over multiple cycles), writes the instructions result into the appropriate rename register, and notifies the completion stage that the instruction has finished execution. In the case of an internal exception, the execution unit reports the exception to the completion/writeback pipeline stage and discontinues instruction execution until the exception is handled. The exception is not signaled until that instruction is the next to be completed. Execution of most floating-point instructions is pipelined within the FPU allowing up to three instructions to be executing in the FPU concurrently. The pipeline stages for the floating-point unit are multiply, add, and round-convert. Execution of most load/store instructions is also pipelined. The load/store units has two pipeline stages. The first stage is for effective address calculation and MMU translation and the second stage is for accessing the data in the cache. @ Thecomplete/writeback pipeline stage maintains the correct architectural machine state and transfers the contents of the rename registers to the GPRs and FPRs as instructions are retired. If the completion logic detects an instruction causing an exception, all following instructions are cancelled, their execution results in rename registers are discarded, and instructions are fetched from the correct instruction stream. A superscalar processor is one that issues multiple independentinstructions into multiple pipelines allowing instructions to execute in parallel. The 603r has five independent execution units, one each for integer instructions, floating-point instructions, branch instruc- tions, load/store instructions, and system register instructions. The IU andthe FPU each have dedicated register files for maintaining operands (GPRs and FPRs, respectively), allowing integer calculations and floating-point calculations to occur simultaneously with- out interference. Because the PowerPC architecture can be applied to such a wide variety of implementations, instruction timing among various Pow- erPC processors varies accordingly. 6. PREPARATION FOR DELIVERY 6.1. Packaging Microcircuits are prepared for delivery in accordance with MIL-PRF-38535. 6.2. Certificate of compliance TCS offers a certificate of compliances with each shipment of parts, affirming the products are in compliance either with MIL-STD-883 and guarantying the parameters not tested at temperature extremes for the entire temperature range. 7. HANDLING MOS devices must be handled with certain precautions to avoid damage due to accumulation of static charge. Input protection devi- ces have been designed in the chip to minimize the effect of this static buildup. However, the following handling practices are recom- mended : a) Devices should be handled on benches with conductive and grounded surfaces. b) Ground test equipment, tools and operator. c) Do not handle devices by the leads. d) Store devices in conductive foam or carriers. e) Avoid use of plastic, rubber, or silk in MOS areas. f) Maintain relative humidity above 50 percent if practical. 32/38 & SB SPEER ES8. PACKAGES MECHANICAL DATA The following sections provide the package parameters and mechanical dimensions for the CBGA packages. 8.1. CBGA package parameters TSPC603R The package parameters are as provided in the following list. The package type is 21 mm, 255-lead ceramic ball grid array (CBGA). Package outline ........... 21mmx 21mm Interconnects ............. 255 Pitch ....... ccc cece eee eee 1.27 mm Maximum module height ... 3.00 mm 8.2. Mechanical dimensions of the CBGA package Figure 15 provides the mechanical dimensions and bottom surface nomenclature of the CBGA package. x NOTES: At | 0.200 CORNER a = ASME Y14.5M 1994. (eI MILLIMETERS P 1.270 BSC y 5.000 | 16.000 x 5.000 | 16.000 Qy 0.200 GF N 123 4 5 7 SI] 8 10 1 12 13 14 15 16 OOCOOCOOOOOOO0CCOO |r CODDOOOOOOCOCOOO [a ONDOOOCOQO0CCOCOCOO Ole OCOOCVOONQVIVOOOOCOO IK OOOODOCOTOOOOCO0C00 |m CODD OOCOCTOOCOCOOCCOO [tr CODDDDDC'OOCCO000| KF loo QQoood geo ooooeeel: OOOO OOCTO0O0COCCOO0O Ia F156060000066660500| COODVDOOOOOOOCOOOCO|F COOODDOOOOOCOOOOO]E COOCOVOOD|OOCOCOOOCO Ia 4 [LLOLSOOODOOO909OO} OOOOCOCCOCDOODOOCOOCO|s T ODOOOOCOD'OOODOOOCOO]as el! | ( kK] 255x OD | 0.30|T/E FO | @! 0.150@]T 1. DIMENSIONING AND TOLERANCING PER 2. CONTROLLING DIMENSION: MILLIMETER. INCHES 0.050 BSC 0.197 |0.630 0.197 |0.630 Figure 15 : Mechanical dimensions and bottom surface nomenclature of the CBGA package 33/38TSPC603R 8.3. CI-CGA package parameters The package parameters are as provided in the following list. The package type is 21 mm, 255-lead ceramic ball grid array (CI-CGA). Package outline ........... 21mmx 21mm Interconnects ............. 255 Pitch... 2.2... cee eee eee 1.27 mm Typical module height ...... 3.84 mm 8.4. Mechanical dimensions of the CL-CGA package Figure 16 provides the mechanical dimensions and bottom surface nomenclature of the CBGA package. 2x CORNER | 0.200 NOTES : | 1. DIMENSIONING AND TOLERANCING PER al eS ASME Y14.5M 1994. 7 2. CONTROLLING DIMENSION: MILLIMETER. _ MILLIMETERS DIM MIN MAX A 21.000 BSC B 21.000 BSC p Oy 0.150 | T Cc 3.84 BSC D 0.790 | 0.990 G 1.270 BSC y H 1.545 1.695 K 0.635 BSC x N 5.000 16.000 (J 0200] P 5.000 | 16.000 R 3.02 BSC N U 0.10 BSC - [| Vv 0.25 0.35 123 4 5 7 BS 10 11 12 13 14 15 16 COOQO0N OO VONONOO}t CODOOODO OOD VOODOOO |a COOOVGOGgoOooooo0odod }e Ol OTOlOlSlelele (eleleleleleleren i, COODCOOOCTO COCO COO| COOCDCDOD'OCOCOCOCCOO It CODCOCODTDOOOOOCOO Ix 7 1696690501080056 e013 lel OTelerorele (elolererereleres [) 7 COOOO9O9O0C99COC9OO}s COOCOVVOO ODO OO9N0 O00 |F OOOCOOOQ|OOCOOOOO]e OCOOCDVOODIOOOOCOOOO |} t OOOO 9VO SIO 0999000 {e COOOVVVNOGD'OOOOO9ON |s T OOOOOCOSD'OODOOOO0O|a el | +| HK] 2ssx 2D }| 0.30@|T1 E O/FO | | 0.150 @] T Figure 16 : Mechanical dimensions and bottom surface nomenclature of the CI-CGA package 34/38 & SPREMAQUESTSPC603R 9. CLOCK RELATIONSHIPS CHOICE The 603r microprocessors offer customers numerous clocking options. An internal phase-lock loop synchronizes the processor (CPU) clock to the bus or system clock (SYSCLK) at various ratios. Inside each PowerPC microprocessor is a phase-lock loop circuit. A voltage controlled oscillator (VCO) is precisely controlled in frequency and phase by a frequency/phase detector which compares the input bus frequency (SYSCLK frequency) to a submultiple of the VCO. The ratio of CPU to SYSCLK frequencies is often referred to as the bus mode (for example, 2:1 bus mode). In the Table 17, the horizontal scale represents the bus frequency (SYSCLK) and the vertical scale represents the PLL-CFG[0-3] signals. For a given SYSCLK (bus) frequency, the PLL configuration signals set the internal CPU and VCO frequency of operation. Table 17 : CPU frequencies for common bus frequencies and multipliers PLL_CFG[0-3] CPU Frequency in MHZ (VCO Frequency in MHz) Bus-to | Coreto Bus Bus Bus Bus Bus Bus Bus Core vco 25 MHz 33.33 40 MHz 50 MHz 60 MHz 66.67 75 MHz Multiplier | Multiplier MHz MHz 0100 2x 2x - - - - - - 150 (300) 0101 2x 4x - - - - - - - 0110 2.5x 2x - - - - 150 166 187 (300) (333) (375) 1000 3x 2x - - - 150 180 200 225 (300) (360) (400) (450) 1110 3.5x 2x - - - 175 210 233 263 (350) (420) (466) (525) 1010 4x 2x - - 160 200 240 267 300 (320) (400) (480) (533) (600) 0111 4.5x 2x - 150 180 225 270 300 - (300) (360) (450) (540) (600) 1011 5x 2x - 166 200 250 300 - - (333) (400) (500) (600) 1001 5.5x 2x - 183 220 275 - - - (366) (440) (550) 1101 6x 2x 150 200 240 300 - - - (300) (400) (480) (600) 0011 PLL bypass 1111 Clock off Notes : 1. Some PLL configurations may select bus, CPU or VCO frequencies which are not supported 2. In PLL-bypass mode, the SYSCLK input signal clocks the internal processor directly, the PLL is disabled, and the bus mode is set for 1:1 mode operation. This mode is intended for factory use only. Note : the AC timing specifications given in this document do not apply in PLL-bypass mode. 3. In clockoff mode, no clocking occurs inside the 603e regardless of the SYSCLK input. 35/38TSPC603R 10.SYSTEM DESIGN INFORMATION 10.1.PLL Power Supply Filtering The AVdd power signal is provided on the 603e to provide power to the clock generation phaselocked loop. To ensure stability of the internal clock, the power supplied to the AVdd input signal should be filtered using a circuit similar to the one shown in Figure 17. The circuit should be placed as close to the AVdd pin to ensure it filters out as much noise as possible. The 0.1 uF capacitor should be closest to the AVdd pin, followed by the 10 uF capacitor, and finally the 10 s. resistor to Vdd. These traces should be kept short and direct. Vdd AAA e AVdd du WF 0.1 uF GND Figure 17 : PLL Power Supply Filter Circuit 10.2.Decoupling Recommendations Due to the 603es dynamic power management feature, large address and data buses, and high operating frequencies, the 603e can generate transient power surges and high frequency noise in its power supply, especially while driving large capacitive loads. This noise must be prevented from reaching other components in the 603e system, and the 6036 itself requires a clean, tightly regulated source of power. Therefore, it is recommended that the system designer place at least one decoupling capacitor at each Vdd and OVdd pin of the 603e. It is also recommended that these decoupling capacitors receive their power from separate Vdd, OVdd, and GND power planes in the PCB, utilizing short traces to minimize inductance. These capacitors should vary in value from 220 pF to 10 uF to provide both highand low-frequency filtering, and should be placed as close as possible to their associated Vdd or OVdd pin. Suggested values for the Vdd pins 220 pF (ceramic), 0,01 uF (ceramic) and 0,1 uf (ceramic). Suggested values for the OVdd pins 0,01 wF (ceramic), 0,1 uF (ceramic), and 10 uF (tantalum). Only SMT (surface mount technology) capacitors should be used to minimize lead inductance. In addition, it is recommended that there be several bulk storage capacitors distributed around the PCB, feeding the Vdd and OVdd planes, to enable quick recharging of the smaller chip capacitors. These bulk capacitors should also have a low ESR (equivaleent series resistance) rating to ensure the quick response time necessary. They should also be connected to the power and ground planes through two vias to minimize inductance. Suggested bulk capacitors 100 uF (AVX TPS tantalum) or 330 uf (AVX TPS tanta- lum). 10.3.Connection Recommendations To ensure reliable operation, it is highly recommended to connect unused inputs to an appropriate signal level. Unused active low inputs should be tied to Vdd. Unused active high inputs should be connected to GND. ALL NC (no-connect) signals must remain unconnected. Power and ground connections must be made to all external Vdd, OVdd, and GND pins of the 603e. 10.4.Pullup Resistor Requirements The 603e requires high-resistive (weak : 10 Ks.) pull-up resistors on several control signals of the bus interface to maintain the control signals in the negated state after they have been actively negated and released by the 603e or other bus master. These sig- nals are -TS, ABB, DBB, and ARTRY. In addition, the 603e has three open-drain style outputs that require pull-up resistors (weak or stronger : 4.7 Kx.10 Ks. ) if they are used by the system. These signals are APE, DPE, and CKSTP_OUT. During inactive periods on the bus, the address and transfer attributes on the bus are not driven by any master and may float in the high-impedance state for relatively long periods of time. Since the 603e must continually monitor these signals for snooping, this float condition may cause excessive power draw by the input reveivers on the 603e. It is recommended that these signals be pulled up trough weak (10 Ks. ) pull-up resistors or restored in some manner by the system. The snooped address and transfer attribute inputs are Al 0-3], AP[0-3], TT[0-4], TBST, and GBL. The data bus input receivers are normally turned off when no read operation is in progress and do not require pull-up resistors on the data bus. 36/38 S SPECHIQUESTSPC603R 11. ORDERING INFORMATION TS (X) PC603R M G B/Q 12 L (C) TCS prefix (1) _ Revision level prefix Prototype Bus divider (to be confirmed) Type Lo: Any bus < 75 MHz Temperature range : Tc M: 55, +125 C V: 40, +110C Max internal processor speed) 6 : 166 MHz Package : 8 : 200 MHz G : CBGA 10 : 283 MHz GS : CI-CGA 12 : 266 MHz 14. : 300 MHz Screening leve/@ : : Standard B/Q: MIL-STD-883, class Q B/T : according to MIL-STD-883 U : Upscreening U/T : Upscreening + burn-in (1) THOMSON-CSF SEMICONDUCTEURS SPECIFIQUES (2) For availability of the different versions, contact your TCS sale office 37/38TSPC603R Information furnished is believed to be accurate and reliable. However THOMSON-CSF SEMICONDUCTEURS SPECIFIQUES assumes no responsibility for the consequences of use of such information nor for any infringement of patents or other rights of third parties which may result from its use. No license is granted by implication or otherwise under any patent or patent rights of THOM- SON-CSF SEMICONDUCTEURS SPECIFIQUES. Specifications mentioned in this publication are subject to change without notice. This publication supersedes and replaces all information previously supplied. THOMSON-CSF SEMICONDUCTEURS SPECIFI- QUES products are not authorized for use as critical components in life support devices or systems without express written approval from THOMSON-CSF SEMICONDUCTEURS SPECIFIQUES. The PowerPC names and logo type are trademarks of International Business Machines Corporation, used under licence 1999 THOMSON-CSF SEMICONDUCTEURS SPECIFIQUES - Printed in France - All rights reserved. This product is manufactured and commercialized by THOMSON-CSF SEMICONDUCTEURS SPECIFIQUES - Avenue de Roche- plaine PO Box 123 - 38521 SAINT-EGREVE Cedex - FRANCE. For further information please contact : THOMSON-CSF SEMICONDUCTEURS SPECIFIQUES - Route Dpartementale 128 - PO Box 46 - 91401 ORSAY Cedex - FRANCE - Phone +33 (0)1 69 33 00 00 - Fax +33 (0)1 69 33 03 21 - Telex 616780 F TCS - Email : lafrique@tcs.thomson. fr 38/38 S SPECHIQUES