Xilinx UG018 FCM Store Instruction

204 www.xilinx.com PowerPC™ 405 Processor Block Reference Guide

1-800-255-7778 UG018 (v2.0) August 20, 2004

Chapter 4: PowerPC 405 APU Controller

Note: Load data can arrive at the same time as the instruction or at a later clock cycle than shown

in Figure 4-8. Also, load data might not be sent back-to-back. Users should look at the valid signal.

FCM Store Instruction

Figure 4-8: APU Controller Decoded a Double Word Load Instruction with LoadWait Example

UG018_04_07_042304

CPMFCMCLK

APUFCMINSTRUCTION

APUFCMINSTRVALID

APUFCMDECODED

APUFCMLOADDATA word0 word1

APUFCMLOADDVALID

FCMAPUDONE

FCMAPULOADWAIT

FCMAPUSLEEPNOTREADY

Figure 4-9: APU Contro ller Decoded Store Instruction

UG018_04_08_032504

CPMFCMCLK

APUFCMINSTRUCTION

APUFCMINSTRVALID

APUFCMDECODED

FCMAPURESULT

FCMAPUDONE

FCMAPUSLEEPNOTREADY

Contents

Main Page Page Page Table of Contents Preface: About This Guide Chapter 1: Introduction to the PowerPC 405 Processor Chapter 2: Input/Output Interfaces Chapter 3: PowerPC 405 OCM Controller Programmers Model Timing Specification for Fixed Latency (Virtex-4 and Virtex-II Pro) Timing Specification for Variable Latency (Virtex-4 DSOCM Controller Only) Application Notes and Reference Designs References Chapter 4: PowerPC 405 APU Controller Introduction FCM Instruction Processing APU Controller Configuration Interface Definition FCM Interface Timing Specification Page Preface About This Guide Guide Contents x Additional Resources x x Conventions Typographical Resource Description/URL Online Document Helvetica bold General Conventions Registers Table 1-1: General Notational Conventions Convention Definition Table 1-2: PowerPC 405 Registers Register Descriptive Name Terms Table 1-2: PowerPC 405 Registers (Continued) Register Descriptive Name Page Page Page Chapter 1 Introduction to the PowerPC 405 Processor x x x x PowerPC Architecture PowerPC Embedded-Environment Architecture x x x x Virtual Environment x Operating Environment Page PowerPC 405 Software Features x Privilege Modes Privileged Mode User Mode Address Translation Modes Real Mode Addressing Modes x x x Data Types Register Set Summary General-Purpose Registers Privileged Registers User Registers Special-Purpose Registers Machine-State Register PowerPC 405 Hardware Organization x x x x Central-Processing Unit Figure 1-2: Pow erPC 405 Organizationa CPUMMU Timers and Debug Cache Units Exception Handling Logic Memory Management Unit x x x x x x Instruction and Data Caches Timer Resources x x x x Programmable Interval Timer Fixed Interval Timer PowerPC 405 Performance Table 1-3: PowerPC 405 Cycles per Instruction Instruction Class Execution Cycles Page Chapter 2 Input/Output Interfaces Signal Naming Conventions x x x x x Table 2-1: Signal Name Prefix Definitions Prefix1 or Prefix2 Definition Location Clock and Power Management Interface x Table 2-1: Signal Name Prefix Definitions (Continued) Prefix1 or Prefix2 Definition Location CPM Interface I/O Signal Summary Figure 2-1: CP M Interface Block Symbol Table 2-2: CPM Interface I/O Signals CPM Interface I/O Signal Descriptions CPMC405CLOCK (Input) PLBCLK (Input) CPMC405CPUCLKEN (Input) CPMC405TIMERCLKEN (Input) CPMC405TIMERTICK (Input) CPMC405SYNCBYPASS (Input, Virtex-4-FX Only) CPMDCRCLK (Input, Virtex-4-FX Only) CPMFCMCLK (Input, Virtex-4-FX Only) C405CPMMSREE (Output) System Design Considerations for Clock Domains x x x x x PLB DCR Virtex-II Pro and ProX Specific Virtex-4 Specific FCM (Virtex-4-FX only) CPU Control Interface CPU Control Interface I/O Signal Summary Figure 2-2: CP U Control Interface Block Symbol Table 2-3: CPU Control Interface I/O Signals CPU Control Interface I/O Signal Descriptions TIEC405MMUEN (Input) TIEC405DETERMINISTICMULT (Input) TIEC405DISOPERANDFWD (Input) C405XXXMACHINECHECK (Output) Reset Interface x Reset Requirements x x x Reset Interface I/O Signal Summary Figure 2-3: Reset Inter face Blo ck Symbo l Table 2-6: Reset Interface I/O Signals Reset Interface I/O Signal Descriptions C405RSTCORERESETREQ (Output) C405RSTCHIPRESETREQ (Output) C405RSTSYSRESETREQ (Output) x RSTC405RESETCORE (Input) RSTC405RESETCHIP (Input) RSTC405RESETSYS (Input) JTGC405TRSTNEG (Input) Instruction-Side Processor Local Bus Interface Instruction-Side PLB Operation x Interaction with the ICU Fill Buffer Prefetch and Address Pipelining x x x x x x xNon-cacheable prefetching is enabled (CCR0[PFNC] 1). x Instruction-Side PLB I/O Signal Table Instruction-Side PLB Interface I/O Signal Descriptions Table 2-7: Instruction-Side PLB Interface Signal Summary C405PLBICUREQUEST (Output) x x x x C405PLBICUABUS[0:29] (Output) C405PLBICUSIZE[2:3] (Output) C405PLBICUCACHEABLE (Output) C405PLBICUU0ATTR (Output) C405PLBICUABORT (Output) C405PLBICUPRIORITY[0:1] (Output) Table 2-8: PLB-Request Priority Encoding Bit 0 Bit 1 Definition PLBC405ICUADDRACK (Input) x x x x PLBC405ICUSSIZE1 (Input) PLBC405ICURDDACK (Input) PLBC405ICURDDBUS[0:63] (Input) x Figure 2-5: Attachment of I SPLB Between 32-Bit Slave and 64-Bit Master PLBC405ICURDWDADDR[1:3] (Input) PLBC405ICUBUSY (Input) Table 2-10: Contents of ICU Read-Data Bus During Line Transfer PLB Slave Size Line Transfer Size Transfer OrderaICU Read-Data Bus [0:31] ICU Read-Data Bus [32:63] PLBC405ICUERR (Input) Instruction-Side PLB Interface Timing Diagrams ISPLB Timing Diagram Assumptions x Page ISPLB Non-Pipelined Cacheable Sequential Fetch (Case 1) Table 2-11: ISPLB Timing Diagram Abbreviations (Continued) AbbreviationaDescription Where Used Figure 2-6: ISPLB Non-Pipelined Cacheable Sequential Fetch (Case 1) ISPLB Non-Pipelined Cacheable Sequential Fetch (Case 2) Figure 2-7: ISPLB Non-Pipelined Cacheable Sequential Fetch (Case 2) ISPLB Pipelined Cacheable Sequential Fetch (Case 1) ISPLB Pipelined Cacheable Sequential Fetch (Case 2) Figure 2-8: IS PLB Pipelined Cacheable Sequential Fetch (Case 1) ISPLB Non-Pipelined Non-Cacheable Sequential Fetch Figure 2-9: IS PLB Pipelined Cacheable Sequential Fetch (Case 2) ISPLB Pipelined Non-Cacheable Sequential Fetch Figure 2-10: ISPLB Non-Pipelined Non-Cacheable Sequential Fetch ISPLB 2:1 Core-to-PLB Line Fetch Figure 2-11: ISPLB Pipelined Non-Cacheable Sequential Fetch PLBC405ICUBUSY ICU rl1 ISPLB 3:1 Core-to-PLB Line Fetch Figure 2-13: ISPLB 3 :1 Core-to-PLB Line Fetch ISPLB Aborted Fetch Request Data-Side Processor Local Bus Interface Figure 2-14: ISPLB Aborted Fetch Request Data-Side PLB Operation Page x Interaction with the DCU Fill Buffer Address Pipelining Unaligned Accesses Guarded Storage Data-Side PLB Interface I/O Signal Table Figure 2-15: Da ta-Side PLB Interface Block Symbol Table 2-12: Data-Side PLB Interface I/O Signal Summary Signal I/O Data-Side PLB Interface I/O Signal Descriptions C405PLBDCUREQUEST (Output) C405PLBDCURNW (Output) C405PLBDCUABUS[0:31] (Output) C405PLBDCUSIZE2 (Output) C405PLBDCUCACHEABLE (Output) C405PLBDCUWRITETHRU (Output) C405PLBDCUU0ATTR (Output) C405PLBDCUGUARDED (Output) C405PLBDCUBE[0:7] (Output) Figure 2-16: Attachment of DSPLB Between 32-Bit Slave and 64-Bit M aster C405PLBDCUPRIORITY[0:1] (Output) C405PLBDCUABORT (Output) Table 2-14: PLB-Request Priority Encoding Bit 0 Bit 1 Definition C405PLBDCUWRDBUS[0:63] (Output) x PLBC405DCUADDRACK (Input) PLBC405DCUSSIZE1 (Input) x PLBC405DCURDDACK (Input) x x x PLBC405DCURDDBUS[0:63] (Input) x PLBC405DCURDWDADDR[1:3] (Input) PLBC405DCUWRDACK (Input) Table 2-16: Contents of DCU Read-Data Bus During Eight-Word Line Transfer PLB-Slave Size Transfer Order DCU Read-Data Bus [0:31] DCU Read-Data Bus PLBC405DCUBUSY (Input) PLBC405DCUERR (Input) Data-Side PLB Interface Timing Diagrams DSPLB Timing Diagram Assumptions x DSPLB Three Consecutive Line Reads Table 2-17: DSPLB Timing Diagram Abbreviations AbbreviationaDescription Where Used DSPLB Line Read/Word Read/Line Read Figure 2-17: DSPLB Three Consecutive Line Reads DSPLB Three Consecutive Word Reads Figure 2-18: DSPLB Li ne Read/Word Read/Line Read DSPLB Three Consecutive Line Writes Figure 2-19: DS PLB Three Consecutive Word Reads DSPLB Line Write/Word Write/Line Write Figure 2-20: DSPLB Three Consecutive Line Writes DSPLB Three Consecutive Word Writes Figure 2-21: DSPLB Line Write/Wor d Write/Line Wri te DSPLB Line Write/Line Read/Word Write Figure 2-22: DSPLB Three Consecutive Word Writes DSPLB Word Write/Word Read/Word Write/Line Read Figure 2-23: DSPLB Line Write/Line Read/Word Write DSPLB Word Write/Line Read/Line Write Figure 2-24: DSPLB Word Write/Word Rea d/Word Write/L ine Read DSPLB 2:1 Core-to-PLB Line Read Figure 2-25: DS PLB Word Write/Line Read/Line Write DSPLB 3:1 Core-to-PLB Line Write Figure 2-26: DSPLB 2:1 Core-to-PLB Line Read DSPLB Aborted Data-Access Request Figure 2-27: DSPLB 3:1 Core-to-PLB Line Write Device-Control Register Interfaces Figure 2-28: DSPLB Aborted Data-Access Request x x x Internal Device Control Register (DCR) Interface Virtex-II Pro and Virtex-II ProX Note: Table 2-18: Virtex-II Pro/ProX DSOCM DCR Address Offset Device Control Register Offset Table 2-19: Virtex-II Pro/ProX ISOCM DCR Address Offset Device Control Register Offset Virtex-4-FX Table 2-20: Virtex-4-FX Internal DCR Address Offset Block Device Control Register Offset External DCR Bus Interface Figure 2-29: Dedica ted EMAC DCR Bus Interface Block Symbol x x x x Note: Figure 2-30: DCR Chain Block Diagram External DCR Bus Interface I/O Signal Summary DCR Slave Figure 2-31: DCR Bus Implementation Virtex-II Pro and Virtex-II ProX Processor Core Virtex-4-FX External DCR Bus Interface I/O Signal Descriptions C405DCRREAD/EXTDCRREAD (Output) C405DCRWRITE/EXTDCRWRITE (Output) C405DCRABUS[0:9]/EXTDCRABUS[0:9] (Output) C405DCRDBUSOUT[0:31]/EXTDCRDBUSOUT[0:31] (Output) External DCR Bus Interface Timing Diagrams DCR Interface 1:1 Clocking, Latched Acknowledge DCR Interface 1:1 Clocking, Combinatorial Acknowledge DCR Interface 2:1 Clocking, Latched Acknowledge Figure 2-34: DCR Interface 1:1 Clocking, Combinatorial Acknowledge Figure 2-35: DCR Inter face 2:1 Clocking, Latched Acknowledge The example in Figure 2-35 assumes the following: read-access or write-access request signal is deasserted. External Interrupt Controller Interface x EIC Interface I/O Signal Summary Figure 2-37: EIC Interface Block Symbol Table 2-23: EIC Interface I/O Signals EIC Interface I/O Signal Descriptions EICC405CRITINPUTIRQ (Input) Figure 2-38: JTAG Interface Block Symbol PPC405 JTAG Debug Port JTAG Interface I/O Signals JTAG Interface I/O Signal Descriptions C405JTGSHIFTDR (Output) C405JTGUPDATEDR (Output) C405JTGPGMOUT (Output) JTAG Instruction Register Figure 2-39: Defau lt Instruction Register Data Path in Virtex with Single PPC405 core Figure 2-40: Instruction Register Data Path for Series PPC405 JTAG Connect ion Table 2-25: PPC405 Instruction Opcodes Instruction Opcode Connecting PPC405 JTAG Logic Directly to Programmable I/O Figure 2-41: Incorrect Wiring of JTAG Chain with Individual PPC405 Connections Page Page Figure 2-44: Correct Wiring of JTAG Chain with Multiplexed PPC405 Connection Connecting PPC405 JTAG Logic in Series with the Dedicated Device JTAG Logic FPGA VHDL and Verilog Instantiation Templates x Page Page Page Page Page Page Debug Interface Debug Interface I/O Signal Summary Figure 2-46: Debug Interface Block Symbol Debug Interface I/O Signal Descriptions DBGC405EXTBUSHOLDACK (Input) DBGC405DEBUGHALT (Input) Table 2-26: Debug Interface I/O Signals DBGC405UNCONDDEBUGEVENT (Input) x x x C405DBGWBFULL (Output) C405DBGWBIAR[0:29] (Output) C405DBGWBCOMPLETE (Output) Trace Interface Trace Interface Signal Summary Trace Interface I/O Signal Descriptions C405TRCTRIGGEREVENTOUT (Output) C405TRCTRIGGEREVENTTYPE[0:10] (Output) Table 2-27: Trace Interface Signals C405TRCCYCLE (Output) C405TRCEVENEXECUTIONSTATUS[0:1] (Output) C405TRCODDEXECUTIONSTATUS[0:1] (Output) Table 2-28: Purpose of C405TRCTRIGGEREVENTTYPE[0:10] Signals Bit Debug Event C405TRCTRACESTATUS[0:3] (Output) Processor Version Register (PVR) Interface (Virtex-4-FX Only) PVR Interface I/O Signal Summary PVR Interface I/O Signal Descriptions Additional FPGA Specific Signals Figure 2-49: FPG A Specific Interface Block Symbol Additional FPGA I/O Signal Descriptions MCBCPUCLKEN (Input) MCBJTAGEN (Input) Page Page Chapter 3 PowerPC 405 OCM Controller Introduction x x x x x x x x Comparison of Virtex-II Pro and Virtex-4 OCM Controllers Functional Features Common Features for DSOCM and ISOCM Features for Data-Side OCM (DSOCM) Features for Instruction-Side OCM (ISOCM) x x x x x Table 3-2: DSOCM and ISOCM Features Feature Data-Side OCM Interface Instruction-Side OCM Interface OCM Controller Operation x Table 3-2: DSOCM and ISOCM Features (Continued) Feature Data-Side OCM Interface Instruction-Side Figure 3-1: OC M Controller Interfaces OCM Interface OCM DCR-Based Control Registers (Accessed Via DCR Instructions) x DSOCM Controller Load/Store Operation Non-Memory Peripherals for DSOCM Execution Re-ordering ISOCM Controller Instruction Fetch Operation x DSOCM Ports Figure 3-3: DSOCM Interface for Virtex-II Pro Figure 3-2: DSOCM Interface for Virtex-4 Data-Side On-Chip Memory (DSOCM) Controller DSOCM Input Ports Table 3-3: DSOCM Input Ports x x Note: Note: DSOCM Input Ports: Attributes x x x x Table 3-4: DSOCM Attributes Attribute Direction Description x x DSOCM Output Ports Table 3-5: DSOCM Output Ports DSOCM-to-BRAM Interfaces Note: Table 3-5: DSOCM Output Ports (Continued) Figure 3-4: DSOCM to BRAM Interface: 8-KByte Example for Virtex-II Pro PORT A PORT B Note: Figure 3-5: DSOCM to BRAM Interface: 8-KByte Example for Virtex-4 PORT A Note: PORT B ISOCM Ports Figure 3-6: DSOC M to Memory-Mapped Slave Peripheral (Virtex-4 Extended Feature) Figure 3-7: ISOC M Interface for Virtex-II Pro ISOCM Input Ports Figure 3-8: ISOC M Interface for Virtex-4 Table 3-6: ISOCM Input Ports x x ISOCM Input Ports, Attributes ISOCM Output Ports Table 3-8: ISOCM Output Ports Table 3-8: ISOCM Output Ports (Continued) Figure 3-9: IS OCM to BRAM Interface: 8 KByte Example in Virtex-II Pro PORT B PORT A Note: Figure 3-10: ISOCM to BRAM Interface: 8 KByte Example in Virtex-4 Programmers Model DCR Registers DSARC/ ISARC Registers PORT B PORT A DSCNTL Registers Table 3-9: DSCNTL Register for Virtex-II Pro ISCNTL Registers Table 3-10: DSCNTL Register for Virtex-4 Table 3-11: ISCNTL Register for Virtex-II Pro Features Introduced in Virtex-4 and Comparison with Virtex-II Pro Table 3-12: ISCNTL Register for Virtex-4 Figure 3-11: DSOCM DCR Registers for Virtex-II Pro Figure 3-12: DSOCM DC R Registers for Virtex-4 Figure 3-13: ISOCM DCR Register s for Virtex-II Pro Figure 3-14: ISOCM DCR Register s for Virtex-4 DCR Write Access Figure 3-15: ISOCM : ISINIT and ISFILL Descriptions (Write Access) for Virtex-II Pro and Virtex-4 DCR Read Access x Figure 3-16: ISOCM: ISINIT and ISFILL Descriptions (Read Access) for Virtex II-Pro Timing Specification for Fixed Latency (Virtex-4 and Virtex-II Pro) Note: Figure 3-17: IS OCM: ISINIT and ISFILL Descriptions (Read Access), for Virtex-4 Single-Cycle Mode Multi-Cycle Mode ISOCM Instruction Fetching Figure 3-18: Instruction Fetch Timing Writing to ISBRAM Figure 3-19: Multi-Cycle Mode (2:1) Instruction Fetch Timing Figure 3-20: S ingle Cycle Mode (1:1) ISOCM Write Timing DSOCM Data Load, Fixed Latency Figure 3-21: Multi C ycle Mode (2:1) ISOCM Write Timing Figure 3-22: Single Cycle Mode (1:1) Data Load Timing Figure 3-23: Multi Cycle Mode (2:1) Data Load Timing DSOCM Store, Fixed Latency Figure 3-24: Single Cycle Mode (1:1) Data Store Timing Timing Specification for Variable Latency (Virtex-4 DSOCM Controller Only) Figure 3-25: Multi Cycle Mode (2:1) Data Store Timing DSOCM Data Load, Variable Latency Note: Figure 3-26: Single Cycle Mode (1:1) DSOCM Read Variable Latency for Virtex-4 Note: DSOCM Data Store, Variable Latency Figure 3-27: Multi Cycle Mode (2:1) DSOCM Read Variable Latency Virtex-4 Note: Figure 3-28: S ingle Cycle Mode (1:1) DSOCM Write Variable Latency Virtex-4 Figure 3-29: Multi C ycle Mode (1:2) DSOCM Write Variable Latency Virtex-4 Application Notes and Reference Designs References Page Chapter 4 PowerPC 405 APU Controller Introduction FCM Instruction Processing Figure 4-1: Pipeline Flow Diagram Enabling the APU Controller Instruction Classes Autonomous Instructions Non-autonomous Instructions Table 4-1: APU Controller-Related MSR Bits Bit(s) in MSR Description Instruction Format Figure 4-2: FCM Instruction Format Table 4-2: APU Op-codes Primary Op-code Extended Op-code Description Instruction Decoding APU Controller Pre-Defined Instruction Decoding Floating Point Instructions Table 4-2: APU Op-codes (Continued) Primary Op-code Extended Op-code Description Note: FCM Load/Store Instructions Note: APU Controller User-Defined Instruction Decoding FCM Pre-Defined Instruction Decoding x Integer Divide Instructions Table 4-3: Load/Store Extended Op-code Field Bit position Description FCM User-Defined Instruction Decoding FCM Exceptions FCM Instruction Flushing x x x APU Controller Configuration General Configuration Register Table 4-4: APU Controller Configuration Register Bit Description UDI Configuration Registers Table 4-4: APU Controller Configuration Register Bit Description (Continued) Table 4-5: UDI Configuration Register Bit Description DCR Access to the Configuration Registers Interface Definition Table 4-5: UDI Configuration Register Bit Description (Continued) APU Controller Input Signals Table 4-6: FCM Interface Input Signals Signal Function Page APU Controller Output Signals Table 4-7: FCM Interface Output Signals Signal Function x x APU Controller Attributes Table 4-8: APU Controller Attributes Attribute Signal Function Page FCM Interface Timing Specification Autonomous Transactions Figure 4-3: APU Controller Decoded Autonomous Transaction Example and operands can be valid on the same FCM clock cycle, or they can be many cycles apart. Note: Figure 4-4: FCM Decoded Autonomous Transaction Example Blocking Transactions Figure 4-5: FCM Decoded Blocking Transaction Example Note: Non-Blocking Transactions Figure 4-6: A PU Controller Decoded Non-Blocking Transaction Example Note: FCM Load Instruction Figure 4-7: APU Controller Decoded Load Instruction Example Note: in Figure 4-7. Figure 4-8: APU Controller Decoded a Double Word Load Instruction with LoadWait Example FCM Store Instruction FCM Exception Figure 4-10: APU Controller Decoded Store Instruction with StoreWBOK=1 Figure 4-11: FCM Exception Note: instruction. FCM Decoding Using Decode Busy Signal Figure 4-12: FCM Decode Asserting DecodeBusy Figure 4-13: FCM Deas ser ti ng DecodeBu s y Appendix A RISCWatch and RISCTrace Interfaces Figure A-1: JTAG-Connector Physical Layout RISCWatch Interface Appendix A: RISCWatch and RISCTrace Interfaces Table A-1: JTAG Connector Signals for RISCWatch Pin RISCWatch Description I/O Signal Name RISCTrace Interface Appendix A: RISCWatch and RISCTrace Interfaces Table A-3: Trace Connector Signals for RISCTrace Pin RISCTrace Description I/O Signal Name Page Page Appendix B Signal Summary Interface Signals Table B-1: PowerPC 405 Interface Signals in A lphabetical Order Page Page Page Page Page Page Page Page Page Appendix C Processor Block Timing Model Figure C-1: PowerP C 405 Processor Block (Simplified) IBM PPC405 Processor Block Timing Parameter Tables and Diagram Table C-2: Parameters Relative to the Core Clock (CPMC405CLOCK) Page Table C-3: Parameters Relative to the DCR Bus Clock (CPMDCRCLK, Virtex-4 Only) Table C-4: Parameters Relative to the FCM Clock (CPMFCMCLK, Virtex-4 Only) Table C-4: Parameters Relative to the FCM Clock (CPMFCMCLK, Virtex-4 Only) (Continued) Table C-5: Parameters Relative to the PLB Clock (PLBCLK) Table C-5: Parameters Relative to the PLB Clock (PLBCLK) (Continued) Table C-6: Parameters Relative to the JTAG Clock (JTAGC405TCK) Table C-7: Parameters Relative to the ISOCM Clock (BRAMISOCMCLK) Table C-7: Parameters Relative to the ISOCM Clock (BRAMISOCMCLK) (Continued) Table C-8: Parameters Relative to the DSOCM Clock (BRAMDSOCMCLK) Figure C-2: Processor Block Timing Relative to Clock Edge Index A B C D G I J L M R S T U V