Title Page
PPC440x5 CPU Core
User’s Manual
Preliminary
SA14-2613-02
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Contents
Figures ............................................................................................................................ 15
About This Book ............................................................................................................ 23
1.2 The PPC440x5 as a PowerPC Implementation .............................................................................. 29
1.3.1 Superscalar Instruction Unit .................................................................................................. 30
1.3.2 Execution Pipelines ............................................................................................................... 31
1.3.3.1 Instruction Cache Controller (ICC) ................................................................................. 31
1.3.4 Memory Management Unit (MMU) ........................................................................................ 32
1.4.1 Processor Local Bus (PLB) ................................................................................................... 36
1.4.2 Device Control Register (DCR) Interface .............................................................................. 36
2. Programming Model ................................................................................................. 39
2.1.1 Storage Operands ................................................................................................................. 39
2.1.2.1 Data Storage Addressing Modes ................................................................................... 41
2.1.2.2 Instruction Storage Addressing Modes .......................................................................... 41
2.1.3 Byte Ordering ........................................................................................................................ 42
2.1.3.1 Structure Mapping Examples ......................................................................................... 43
2.1.3.4 Byte-Reverse Instructions .............................................................................................. 46
2.2.1 Register Types ...................................................................................................................... 52
ppc440x5TOC.fm.
September 12, 2002
Page 3 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
2.4 Implemented Instruction Set Summary ........................................................................................... 56
2.4.1 Integer Instructions ................................................................................................................ 57
2.4.1.4 Integer Compare Instructions ......................................................................................... 59
2.4.1.6 Integer Rotate Instructions ............................................................................................. 59
2.4.2 Branch Instructions ................................................................................................................ 60
2.4.3.1 Condition Register Logical Instructions .......................................................................... 61
2.4.3.4 Processor Synchronization Instruction ........................................................................... 61
2.4.4 Storage Control Instructions .................................................................................................. 62
2.5 Branch Processing .......................................................................................................................... 64
2.5.5.3 Condition Register (CR) ................................................................................................. 67
2.6 Integer Processing .......................................................................................................................... 71
2.7.1 Special Purpose Registers General (USPRG0, SPRG0–SPRG7) ........................................ 75
2.7.2 Processor Version Register (PVR) ........................................................................................ 75
2.7.4 Core Configuration Register 0 (CCR0) .................................................................................. 76
2.7.5 Core Configuration Register 1 (CCR1) .................................................................................. 78
2.9 Speculative Accesses ..................................................................................................................... 81
ppc440x5TOC.fm.
September 12, 2002
Page 4 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
2.10.2 Execution Synchronization .................................................................................................. 83
2.10.3 Storage Ordering and Synchronization ............................................................................... 84
3. Initialization ............................................................................................................... 85
3.2 Reset Types .................................................................................................................................. 89
4. Instruction and Data Caches ................................................................................... 95
4.1.2 Cache Locking and Transient Mechanism ............................................................................ 99
4.2.2 Speculative Prefetch Mechanism ........................................................................................ 105
4.2.3 Instruction Cache Coherency .............................................................................................. 106
4.2.4 Instruction Cache Control and Debug ................................................................................. 108
4.2.4.7 Simulating Instruction Cache Parity Errors for Software Testing ................................. 114
4.3.1 DCC Operations .................................................................................................................. 116
4.3.1.3 Store Operations .......................................................................................................... 119
4.3.1.4 Line Flush Operations .................................................................................................. 121
4.3.1.5 Data Read PLB Interface Requests ............................................................................. 122
4.3.1.6 Data Write PLB Interface Requests ............................................................................. 123
4.3.1.7 Storage Access Ordering ............................................................................................. 124
4.3.2 Data Cache Coherency ....................................................................................................... 124
4.3.3 Data Cache Control and Debug .......................................................................................... 125
4.3.3.4 dcbt and dcbtst Operation ............................................................................................ 126
4.3.3.7 Simulating Data Cache Parity Errors for Software Testing .......................................... 130
5. Memory Management ............................................................................................. 133
ppc440x5TOC.fm.
September 12, 2002
Page 5 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
5.1.1 Support for PowerPC Book-E MMU Architecture ................................................................ 133
5.3.1 Virtual Address Formation ................................................................................................... 138
5.4 Address Translation ...................................................................................................................... 140
5.5 Access Control .............................................................................................................................. 142
5.5.1 Execute Access ................................................................................................................... 142
5.6.1 Write-Through (W) ............................................................................................................... 145
5.6.3 Memory Coherence Required (M) ....................................................................................... 146
5.6.6 User-Definable (U0–U3) ...................................................................................................... 147
5.6.7 Supported Storage Attribute Combinations ......................................................................... 147
5.7.1 Memory Management Unit Control Register (MMUCR) ...................................................... 148
5.8 Shadow TLB Arrays ...................................................................................................................... 151
5.9 TLB Management Instructions ...................................................................................................... 152
5.9.1 TLB Search Instruction (tlbsx[.]) .......................................................................................... 153
5.9.2 TLB Read/Write Instructions (tlbre/tlbwe) ............................................................................ 153
5.9.3 TLB Sync Instruction (tlbsync) ............................................................................................. 154
5.11.1 Reading TLB Parity Bits with tlbre ..................................................................................... 155
5.11.2 Simulating TLB Parity Errors for Software Testing ............................................................ 156
6. Interrupts and Exceptions ..................................................................................... 159
6.1 Overview ....................................................................................................................................... 159
6.2 Interrupt Classes ........................................................................................................................... 159
6.2.2.1 Synchronous, Precise Interrupts .................................................................................. 160
6.3 Interrupt Processing ...................................................................................................................... 162
6.4.1 Machine State Register (MSR) ............................................................................................ 165
ppc440x5TOC.fm.
September 12, 2002
Page 6 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
6.4.8 Data Exception Address Register (DEAR) .......................................................................... 170
6.4.9 Interrupt Vector Offset Registers (IVOR0–IVOR15) ........................................................... 170
6.4.10 Interrupt Vector Prefix Register (IVPR) ............................................................................. 171
6.4.11 Exception Syndrome Register (ESR) ................................................................................ 172
6.4.12 Machine Check Status Register (MCSR) .......................................................................... 174
6.5.1 Critical Input Interrupt .......................................................................................................... 178
6.5.2 Machine Check Interrupt ..................................................................................................... 178
6.5.6 Alignment Interrupt .............................................................................................................. 185
6.5.7 Program Interrupt ................................................................................................................ 187
6.5.8 Floating-Point Unavailable Interrupt .................................................................................... 190
6.5.11 Decrementer Interrupt ....................................................................................................... 191
6.5.12 Fixed-Interval Timer Interrupt ............................................................................................ 192
6.6 Interrupt Ordering and Masking .................................................................................................... 199
6.7 Exception Priorities ....................................................................................................................... 202
6.7.2 Exception Priorities for Floating-Point Load and Store Instructions .................................... 203
6.7.6 Exception Priorities for Privileged Instructions .................................................................... 205
6.7.13 Exception Priorities for All Other Instructions .................................................................... 208
7.1.2 Writing the Time Base ......................................................................................................... 210
7.2 Decrementer (DEC) ...................................................................................................................... 211
ppc440x5TOC.fm.
September 12, 2002
Page 7 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
7.5 Timer Control Register (TCR) ....................................................................................................... 215
7.7 Freezing the Timer Facilities ......................................................................................................... 217
8. Debug Facilities ...................................................................................................... 219
8.2.1 Internal Debug Mode ........................................................................................................... 219
8.2.2 External Debug Mode .......................................................................................................... 220
8.2.3 Debug Wait Mode ................................................................................................................ 220
8.2.4 Trace Debug Mode .............................................................................................................. 221
8.3.1 Instruction Address Compare (IAC) Debug Event ............................................................... 222
8.3.1.1 IAC Debug Event Fields ............................................................................................... 222
8.3.2.2 DAC Debug Event Processing ..................................................................................... 229
8.3.2.4 DAC Debug Events Applied to Various Instruction Types ........................................... 230
8.3.3.2 DVC Debug Event Processing ..................................................................................... 233
8.3.3.4 DVC Debug Events Applied to Various Instruction Types ........................................... 233
8.3.4 Branch Taken (BRT) Debug Event ...................................................................................... 234
8.3.8 Interrupt (IRPT) Debug Event .............................................................................................. 236
8.3.9 Unconditional Debug Event (UDE) ...................................................................................... 237
8.4 Debug Reset ................................................................................................................................. 238
8.5 Debug Timer Freeze ..................................................................................................................... 238
8.6.4 Debug Status Register (DBSR) .......................................................................................... 244
9. Instruction Set ........................................................................................................ 249
ppc440x5TOC.fm.
September 12, 2002
Page 8 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
9.3 Pseudocode .................................................................................................................................. 251
9.3.1 Operator Precedence .......................................................................................................... 253
9.4 Register Usage ............................................................................................................................. 253
add ............................................................................................................................................ 255
adde .......................................................................................................................................... 257
addme ....................................................................................................................................... 262
and ............................................................................................................................................ 264
b ................................................................................................................................................ 268
cmpli.......................................................................................................................................... 285
cntlzw ........................................................................................................................................ 286
crand ......................................................................................................................................... 287
crnand ....................................................................................................................................... 290
crnor .......................................................................................................................................... 291
cror ............................................................................................................................................ 292
crorc .......................................................................................................................................... 293
crxor .......................................................................................................................................... 294
dccci .......................................................................................................................................... 304
dcread ....................................................................................................................................... 305
divw ........................................................................................................................................... 307
divwu ......................................................................................................................................... 308
icbi............................................................................................................................................. 313
ppc440x5TOC.fm.
September 12, 2002
Page 9 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
icbt............................................................................................................................................. 314
icread......................................................................................................................................... 317
lbz.............................................................................................................................................. 321
lbzu............................................................................................................................................ 322
lha.............................................................................................................................................. 325
lhau............................................................................................................................................ 326
lhaux.......................................................................................................................................... 327
lhax............................................................................................................................................ 328
lhbrx........................................................................................................................................... 329
lhz.............................................................................................................................................. 330
lhzu............................................................................................................................................ 331
ppc440x5TOC.fm.
September 12, 2002
Page 10 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
mtspr ......................................................................................................................................... 370
mulchw ...................................................................................................................................... 373
mulchwu .................................................................................................................................... 374
mulhhw...................................................................................................................................... 375
mulhhwu.................................................................................................................................... 376
mulhw........................................................................................................................................ 377
mulhwu...................................................................................................................................... 378
mullhw ....................................................................................................................................... 379
mullhwu ..................................................................................................................................... 380
mulli........................................................................................................................................... 381
mullw ......................................................................................................................................... 382
nand .......................................................................................................................................... 383
neg ............................................................................................................................................ 384
nmacchw ................................................................................................................................... 385
nmacchws ................................................................................................................................. 386
nmachhw................................................................................................................................... 387
nmachhws ................................................................................................................................. 388
nmaclhw .................................................................................................................................... 389
nmaclhws .................................................................................................................................. 390
nor ............................................................................................................................................. 391
or ............................................................................................................................................... 392
orc ............................................................................................................................................. 393
ori .............................................................................................................................................. 394
rfi ............................................................................................................................................... 397
rlwimi ......................................................................................................................................... 399
rlwinm........................................................................................................................................ 400
slw ............................................................................................................................................. 405
srawi.......................................................................................................................................... 407
stmw.......................................................................................................................................... 418
stswi .......................................................................................................................................... 419
stswx ......................................................................................................................................... 421
stw............................................................................................................................................. 422
stwbrx........................................................................................................................................ 423
stwcx. ........................................................................................................................................ 424
ppc440x5TOC.fm.
September 12, 2002
Page 11 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
subf............................................................................................................................................ 429
subfc.......................................................................................................................................... 430
subfe.......................................................................................................................................... 431
subfme....................................................................................................................................... 433
subfze........................................................................................................................................ 434
tw............................................................................................................................................... 440
wrtee.......................................................................................................................................... 446
wrteei......................................................................................................................................... 447
xori............................................................................................................................................. 449
xoris........................................................................................................................................... 450
10. Register Summary ............................................................................................... 451
10.2 Reserved Fields .......................................................................................................................... 457
CCR0......................................................................................................................................... 460
CCR1......................................................................................................................................... 462
DCDBTRH................................................................................................................................. 478
DEC........................................................................................................................................... 481
DECAR...................................................................................................................................... 482
GPR0–GPR31........................................................................................................................... 489
ppc440x5TOC.fm.
September 12, 2002
Page 12 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
ICDBTRH .................................................................................................................................. 492
ICDBTRL................................................................................................................................... 493
INV0–INV3 ................................................................................................................................ 494
ITV0–ITV3 ................................................................................................................................. 495
IVLIM......................................................................................................................................... 496
IVPR.......................................................................................................................................... 498
MCSRR0 ................................................................................................................................... 501
MCSRR1 ................................................................................................................................... 502
MSR .......................................................................................................................................... 504
PID ............................................................................................................................................ 506
PIR ............................................................................................................................................ 507
PVR........................................................................................................................................... 508
SPRG0–SPRG7........................................................................................................................ 510
TBU ........................................................................................................................................... 514
TCR........................................................................................................................................... 515
TSR ........................................................................................................................................... 516
XER........................................................................................................................................... 518
A.1.2.1 I-Form .......................................................................................................................... 522
A.1.2.3 SC-Form ...................................................................................................................... 522
A.1.2.7 XFX-Form .................................................................................................................... 524
A.1.2.8 XO-Form ...................................................................................................................... 524
A.1.2.9 M-Form ........................................................................................................................ 524
A.4 Preserved Instruction Opcodes .................................................................................................... 557
A.5 Reserved Instruction Opcodes ..................................................................................................... 558
ppc440x5TOC.fm.
September 12, 2002
Page 13 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Index ............................................................................................................................. 571
ppc440x5TOC.fm.
September 12, 2002
Page 14 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Figures
Figure 1-1. PPC440 Core Block Diagram .................................................................................................30
Figure 2-1. User Programming Model Registers ......................................................................................48
Figure 2-2. Supervisor Programming Model Registers ............................................................................49
Figure 2-3. Link Register (LR) ..................................................................................................................67
Figure 2-4. Count Register (CTR) ............................................................................................................67
Figure 2-5. Condition Register (CR) .........................................................................................................68
Figure 2-6. General Purpose Registers (R0-R31) ....................................................................................71
Figure 2-7. Integer Exception Register (XER) ..........................................................................................72
Figure 2-8. Special Purpose Registers General (USPRG0, SPRG0–SPRG7) ........................................75
Figure 2-9. Processor Version Register (PVR) .........................................................................................76
Figure 2-10. Processor Identification Register (PIR) ..................................................................................76
Figure 2-11. Core Configuration Register 0 (CCR0) ..................................................................................77
Figure 2-12. Core Configuration Register 1 (CCR1) ..................................................................................78
Figure 2-13. Reset Configuration ...............................................................................................................79
Figure 4-1. Instruction Cache Normal Victim Registers (INV0–INV3) ......................................................97
Figure 4-1. Instruction Cache Transient Victim Registers (ITV0–ITV3) ....................................................97
Figure 4-1. Data Cache Normal Victim Registers (DNV0–DNV3) ............................................................97
Figure 4-1. Data Cache Transient Victim Registers (DTV0–DTV3) .........................................................97
Figure 4-2. Instruction Cache Victim Limit (IVLIM) ...................................................................................99
Figure 4-2. Data Cache Victim Limit (DVLIM) ..........................................................................................99
Figure 4-3. Cache Locking and Transient Mechanism (Example 1)1 .....................................................102
Figure 4-4. Cache Locking and Transient Mechanism (Example 2) .......................................................103
Figure 4-5. Core Configuration Register 0 (CCR0) ................................................................................109
Figure 4-6. Core Configuration Register 1 (CCR1) ................................................................................110
Figure 4-7. Instruction Cache Debug Data Register (ICDBDR) .............................................................113
Figure 4-8. Instruction Cache Debug Tag Register High (ICDBTRH) ....................................................113
Figure 4-9. Instruction Cache Debug Tag Register Low (ICDBTRL) ......................................................113
Figure 4-10. Data Cache Debug Tag Register High (DCDBTRH) ............................................................128
Figure 4-11. Data Cache Debug Tag Register Low (DCDBTRL) .............................................................128
Figure 5-1. Virtual Address to TLB Entry Match Process .......................................................................140
Figure 5-2. Effective-to-Real Address Translation Flow .........................................................................141
Figure 5-3. Memory Management Unit Control Register (MMUCR) .......................................................148
Figure 5-4. Process ID (PID) ..................................................................................................................151
Figure 5-5. TLB Entry Word Definitions ..................................................................................................154
Figure 6-1. Machine State Register (MSR) ............................................................................................165
Figure 6-2. Save/Restore Register 0 (SRR0) .........................................................................................167
ppc440x5LOF.fm.
September 12, 2002
Page 15 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Figure 6-3. Save/Restore Register 1 (SRR1) .........................................................................................168
Figure 6-4. Critical Save/Restore Register 0 (CSRR0) ...........................................................................168
Figure 6-5. Critical Save/Restore Register 1 (CSRR1) ...........................................................................169
Figure 6-6. Machine Check Save/Restore Register 0 (MCSRR0) ..........................................................169
Figure 0-1. Machine Check Save/Restore Register 1 (MCSRR1) ..........................................................170
Figure 6-7. Data Exception Address Register (DEAR) ...........................................................................170
Figure 6-8. Interrupt Vector Offset Registers (IVOR0–IVOR15) ............................................................171
Figure 6-9. Interrupt Vector Prefix Register (IVPR) ................................................................................172
Figure 6-10. Exception Syndrome Register (ESR) ...................................................................................172
Figure 6-11. Machine Check Status Register (MCSR) .............................................................................174
Figure 7-1. Relationship of Timer Facilities to the Time Base ................................................................209
Figure 7-2. Time Base Lower (TBL) ........................................................................................................210
Figure 7-3. Time Base Upper (TBU) .......................................................................................................210
Figure 7-4. Decrementer (DEC) ..............................................................................................................211
Figure 7-5. Decrementer Auto-Reload (DECAR) ....................................................................................212
Figure 7-6. Watchdog State Machine .....................................................................................................215
Figure 7-7. Timer Control Register (TCR) ...............................................................................................216
Figure 7-8. Timer Status Register (TSR) ................................................................................................217
Figure 8-1. Debug Control Register 0 (DBCR0) .....................................................................................239
Figure 8-2. Debug Control Register 1 (DBCR1) .....................................................................................240
Figure 8-3. Debug Control Register 2 (DBCR2) .....................................................................................243
Figure 8-4. Debug Status Register (DBSR) ............................................................................................244
Figure 8-5. Instruction Address Compare Registers (IAC1–IAC4) .........................................................246
Figure 8-6. Data Address Compare Registers (DAC1–DAC2) ...............................................................246
Figure 8-7. Data Value Compare Registers (DVC1–DVC2) ...................................................................246
Figure 8-8. Debug Data Register (DBDR) ..............................................................................................247
Figure 10-1. Core Configuration Register 0 (CCR0) .................................................................................460
Figure 10-2. Core Configuration Register 1 (CCR1) .................................................................................462
Figure 10-3. Condition Register (CR) .......................................................................................................464
Figure 10-4. Critical Save/Restore Register 0 (CSRR0) ...........................................................................465
Figure 10-5. Critical Save/Restore Register 1 (CSRR1) ...........................................................................466
Figure 10-6. Count Register (CTR) ...........................................................................................................467
Figure 10-7. Data Address Compare Registers (DAC1–DAC2) ...............................................................468
Figure 10-8. Debug Control Register 0 (DBCR0) .....................................................................................469
Figure 10-9. Debug Control Register 1 (DBCR1) .....................................................................................471
Figure 10-10. Debug Control Register 2 (DBCR2) .....................................................................................473
Figure 10-11. Debug Data Register (DBDR) ..............................................................................................475
ppc440x5LOF.fm.
September 12, 2002
Page 16 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Figure 10-14. Data Cache Debug Tag Register Low (DCDBTRL) .............................................................479
Figure 10-15. Data Exception Address Register (DEAR) ...........................................................................480
Figure 10-16. Decrementer (DEC) .............................................................................................................481
Figure 10-17. Decrementer Auto-Reload (DECAR) ...................................................................................482
Figure 10-18. Data Cache Normal Victim Registers (DNV0–DNV3) ..........................................................483
Figure 10-19. Data Cache Transient Victim Registers (DTV0–DTV3) .......................................................484
Figure 10-20. Data Value Compare Registers (DVC1–DVC2) ...................................................................485
Figure 10-21. Data Cache Victim Limit (DVLIM) ........................................................................................486
Figure 10-22. Exception Syndrome Register (ESR) ...................................................................................487
Figure 10-23. General Purpose Registers (R0-R31) ..................................................................................489
Figure 10-24. Instruction Address Compare Registers (IAC1–IAC4) .........................................................490
Figure 10-25. Instruction Cache Debug Data Register (ICDBDR) .............................................................491
Figure 10-26. Instruction Cache Debug Tag Register High (ICDBTRH) ....................................................492
Figure 10-27. Instruction Cache Debug Tag Register Low (ICDBTRL) ......................................................493
Figure 10-28. Instruction Cache Normal Victim Registers (INV0–INV3) ....................................................494
Figure 10-29. Instruction Cache Transient Victim Registers (ITV0–ITV3) ..................................................495
Figure 10-30. Instruction Cache Victim Limit (IVLIM) .................................................................................496
Figure 10-31. Interrupt Vector Offset Registers (IVOR0–IVOR15) ............................................................497
Figure 10-32. Interrupt Vector Prefix Register (IVPR) ................................................................................498
Figure 10-33. Link Register (LR) ................................................................................................................499
Figure 10-34. Machine Check Status Register (MCSR) .............................................................................500
Figure 10-35. Machine Check Save/Restore Register 0 (MCSRR0) ..........................................................501
Figure 0-2. Machine Check Save/Restore Register 1 (MCSRR1) ..........................................................502
Figure 10-36. Memory Management Unit Control Register (MMUCR) .......................................................503
Figure 10-37. Machine State Register (MSR) ............................................................................................504
Figure 10-38. Process ID (PID) ..................................................................................................................506
Figure 10-39. Processor Identification Register (PIR) ................................................................................507
Figure 10-40. Processor Version Register (PVR) .......................................................................................508
Figure 10-41. Reset Configuration .............................................................................................................509
Figure 10-42. Special Purpose Registers General (SPRG0–SPRG7) .......................................................510
Figure 10-43. Save/Restore Register 0 (SRR0) .........................................................................................511
Figure 10-44. Save/Restore Register 1 (SRR1) .........................................................................................512
Figure 10-45. Time Base Lower (TBL) .......................................................................................................513
Figure 10-46. Time Base Upper (TBU) .......................................................................................................514
Figure 10-47. Timer Control Register (TCR) ..............................................................................................515
Figure 10-48. Timer Status Register (TSR) ................................................................................................516
Figure 10-49. User Special Purpose Register General (USPRG0) ............................................................517
Figure 10-50. Integer Exception Register (XER) ........................................................................................518
Figure A-1. I Instruction Format ..............................................................................................................522
ppc440x5LOF.fm.
September 12, 2002
Page 17 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Figure A-2. B Instruction Format .............................................................................................................522
Figure A-3. SC Instruction Format ...........................................................................................................522
Figure A-4. D Instruction Format .............................................................................................................522
Figure A-5. X Instruction Format .............................................................................................................523
Figure A-6. XL Instruction Format ...........................................................................................................524
Figure A-7. XFX Instruction Format .........................................................................................................524
Figure A-8. XO Instruction Format ..........................................................................................................524
Figure A-9. M Instruction Format .............................................................................................................524
ppc440x5LOF.fm.
September 12, 2002
Page 18 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Data Operand Definitions .......................................................................................................40
Alignment Effects for Storage Access Instructions ................................................................40
Register Categories ...............................................................................................................50
Instruction Categories ............................................................................................................57
Integer Storage Access Instructions ......................................................................................58
Integer Arithmetic Instructions ................................................................................................58
Integer Logical Instructions ....................................................................................................59
Integer Compare Instructions .................................................................................................59
Integer Trap Instructions .......................................................................................................59
Table 2-10. Integer Rotate Instructions .....................................................................................................59
Table 2-11. Integer Shift Instructions ........................................................................................................60
Table 2-12. Integer Select Instruction .......................................................................................................60
Table 2-13. Branch Instructions ................................................................................................................60
Table 2-14. Condition Register Logical Instructions ..................................................................................61
Table 2-15. Register Management Instructions ........................................................................................61
Table 2-16. System Linkage Instructions ..................................................................................................61
Table 2-17. Processor Synchronization Instruction ...................................................................................62
Table 2-18. Cache Management Instructions ...........................................................................................62
Table 2-19. TLB Management Instructions ...............................................................................................62
Table 2-20. Storage Synchronization Instructions .....................................................................................63
Table 2-21. Allocated Instructions .............................................................................................................63
Table 2-22. BO Field Definition .................................................................................................................65
Table 2-23. BO Field Examples ................................................................................................................65
Table 2-24. CR Updating Instructions .......................................................................................................69
Table 2-25. XER[SO,OV] Updating Instructions ........................................................................................73
Table 2-26. XER[CA] Updating Instructions ..............................................................................................73
Table 2-27. Privileged Instructions ............................................................................................................80
Reset Values of Registers and Other PPC440x5 Facilities ...................................................86
Instruction and Data Cache Array Organization .....................................................................96
Cache Sizes and Parameters ................................................................................................96
Victim Index Field Selection ...................................................................................................98
Icread and dcread Cache Line Selection .............................................................................112
Data Cache Behavior on Store Accesses ............................................................................121
TLB Entry Fields ...................................................................................................................135
Page Size and Effective Address to EPN Comparison ........................................................140
Page Size and Real Address Formation ..............................................................................142
ppc440x5LOT.fm.
September 12, 2002
Page 19 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Watchdog Timer Exception Behavior ...................................................................................214
Debug Event Summary ........................................................................................................237
Allocated Instructions ...........................................................................................................250
Operator Precedence ...........................................................................................................253
Extended Mnemonics for addic ............................................................................................259
Extended Mnemonics for addic. ...........................................................................................260
Extended Mnemonics for addis ............................................................................................261
Extended Mnemonics for bcctr, bcctrl ..................................................................................275
Table 9-10. Extended Mnemonics for bclr, bclrl ......................................................................................279
Table 9-11. Extended Mnemonics for cmp ..............................................................................................282
Table 9-12. Extended Mnemonics for cmpi .............................................................................................283
Table 9-13. Extended Mnemonics for cmpl .............................................................................................284
Table 9-14. Extended Mnemonics for cmpli ............................................................................................285
Table 9-15. Extended Mnemonics for creqv ............................................................................................289
Table 9-16. Extended Mnemonics for crnor .............................................................................................291
Table 9-17. Extended Mnemonics for cror ...............................................................................................292
Table 9-18. Extended Mnemonics for crxor .............................................................................................294
Table 9-19. Extended Mnemonics for mfspr ............................................................................................364
Table 9-20. FXM Bit Field Correspondence ............................................................................................367
Table 9-21. Extended Mnemonics for mtcrf .............................................................................................367
Table 9-22. Extended Mnemonics for mtspr ............................................................................................371
Table 9-23. Extended Mnemonics for nor, nor. .......................................................................................391
Table 9-24. Extended Mnemonics for or, or. ...........................................................................................392
Table 9-25. Extended Mnemonics for ori .................................................................................................394
Table 9-26. Extended Mnemonics for rlwimi, rlwimi. ..............................................................................399
Table 9-27. Extended Mnemonics for rlwinm, rlwinm. .............................................................................400
Table 9-28. Extended Mnemonics for rlwnm, rlwnm. ..............................................................................403
Table 9-29. Extended Mnemonics for subf, subf., subfo, subfo. ..............................................................429
Table 9-30. Extended Mnemonics for subfc, subfc., subfco, subfco. ......................................................430
ppc440x5LOT.fm.
September 12, 2002
Page 20 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 9-31. Extended Mnemonics for tw .................................................................................................441
Table 9-32. Extended Mnemonics for twi ................................................................................................444
Table 10-1. Register Categories .............................................................................................................452
Table 10-2. Special Purpose Registers Sorted by SPR Number ............................................................454
Table 10-3. Interrupt Types Associated with each IVOR ........................................................................497
PPC440x5 Instruction Syntax Summary ..............................................................................525
Allocated Opcodes ...............................................................................................................557
Preserved Opcodes .............................................................................................................558
Reserved-nop Opcodes .......................................................................................................558
PPC440x5 Instructions by Opcode ......................................................................................559
ppc440x5LOT.fm.
September 12, 2002
Page 21 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
ppc440x5LOT.fm.
September 12, 2002
Page 22 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
About This Book
This user’s manual provides the architectural overview, programming model, and detailed information about
the instruction set, registers, and other facilities of the IBM™ Book-E Enhanced PowerPC™ 440x5
(PPC440x5™) 32-bit embedded controller core.
The PPC440x5 embedded controller core features:
• Book-E Enhanced PowerPC Architecture™
• Dual-issue superscalar pipeline with dynamic branch prediction
• Separate, configurable (up to 32KB each) instruction and data caches, with cache line locking
• DSP acceleration with 24 new integer multiply-accumulate (MAC) instructions
• Memory Management Unit (MMU) with 64-entry TLB and support for page sizes of 1KB–256MB
• 64GB (36-bit) physical address capability
• 128-bit PLB interface, part of the IBM CoreConnect™ on-chip system bus architecture
• JTAG debug interface with extensive integrated debug facilities, including real-time trace
Who Should Use This Book
This book is for system hardware and software developers, and for application developers who need to
understand the PPC440x5. The audience should understand embedded system design, operating systems,
RISC microprocessing, and computer organization and architecture.
How to Use This Book
This book describes the PPC440x5 device architecture, programming model, registers, and instruction set.
This book contains the following chapters:
Chapter 1.
Chapter 2.
Chapter 3.
Chapter 4.
Chapter 5.
Chapter 6.
Chapter 7.
Chapter 8.
Chapter 9.
Overview
Programming Model
Initialization
Instruction and Data Caches
Memory Management
Interrupts and Exceptions
Timer Facilities
Debug Facilities
Instruction Set
Chapter 10. Register Summary
This book contains the following appendixes:
Appendix A. Instruction Summary
Appendix B. PPC440 Core Compiler Optimizations
To help readers find material in these chapters, this book contains:
preface.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Contents, on page v.
Figures, on page xi.
Tables, on page xiii.
Index, on page 571.
Notation
The manual uses the following notational conventions:
• Active low signals are shown with an overbar (Active_Low)
• All numbers are decimal unless specified in some special way.
• 0bnnnn means a number expressed in binary format.
• 0xnnnn means a number expressed in hexadecimal format.
Underscores may be used between digits.
• RA refers to General Purpose Register (GPR) RA.
• (RA) refers to the contents of GPR RA.
• (RA|0) refers to the contents of GPR RA, or to the value 0 if the RA field is 0.
• Bits in registers, instructions, and fields are specified as follows.
• Bits are numbered most-significant bit to least-significant bit, starting with bit 0.
Note: This document differs from the Book-E architecture specification in the use of bit numbering
for architected registers. Book-E defines the full, 64-bit instruction set architecture, and all
registers are shown as having bit numbers from 0 to 63, with bit 63 being the least
significant. This manual describes a 32-bit subset implementation of the architecture.
Architected registers are described as being 32 bits long, with bits numbered from 0 to 31,
and with bit 31 being the least significant. When this document refers to register bits 0 to 31,
they actually correspond to bits 32 to 63 of the same register in the Book-E architecture
specification.
• X means bit p of register, instruction, or field X
p
• X means bits p through q of register, instruction, or field X
p:q
• X
means bits p, q,... of register, instruction, or field X
p,q,...
• X[p] means a named field p of register X.
• X[p:q] means named fields p through q of register X.
• X[p,q,...] means named fields p, q,... of register X.
...
• ¬X means the ones complement of the contents of X.
• A period (.) as the last character of an instruction mnemonic means that the instruction records status
information in certain fields of the Condition Register as a side effect of execution, as described in
• The symbol || is used to describe the concatenation of two values. For example, 0b010 || 0b111 is the
same as 0b010111.
n
• x means x raised to the n power.
preface.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
n
n
n
•
x means the replication of x, n times (that is, x concatenated to itself n – 1 times). 0 and 1 are special
cases:
n
5
•
•
0 means a field of n bits with each bit equal to 0. Thus 0 is equivalent to 0b00000.
n
5
1 means a field of n bits with each bit equal to 1. Thus 1 is equivalent to 0b11111.
• /, //, ///, ... denotes a reserved field in an instruction or in a register.
• ? denotes an allocated bit in a register.
• A shaded field denotes a field that is reserved or allocated in an instruction or in a register.
Related Publications
The following book describes the Book-E Enhanced PowerPC Architecture:
• Book E: PowerPC Architecture Enhanced for Embedded Applications
(www.chips.ibm.com/techlib/products/powerpc/manuals/)
The following CD-ROM contains publications describing the IBM PowerPC 400 family of embedded control-
lers, including this manual PowerPC PPC440x5 User’s Manual, and application and technical notes.
• IBM PowerPC Embedded Processor Solutions (Order Number SC09-3032)
preface.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
preface.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
1. Overview
The IBM™ PowerPC™ 440x5 32-bit embedded processor core, referred to as the PPC440x5 core, imple-
ments the Book-E Enhanced PowerPC Architecture.
This chapter describes:
• PPC440x5 core features
• The PPC440x5 core as an implementation of the Book-E Enhanced PowerPC Architecture
• The organization of the PPC440x5 core, including a block diagram and descriptions of the functional units
• PPC440x5 core interfaces
1.1 PPC440x5 Features
The PPC440x5 core is a high-performance, low-power engine that implements the flexible and powerful
Book-E Enhanced PowerPC Architecture.
The PPC440x5 contains a dual-issue, superscalar, pipelined processing unit, along with other functional
elements required by embedded ASIC product specifications. These other functions include memory
management, cache control, timers, and debug facilities. Interfaces for custom co-processors and floating
point functions are provided, along with separate instruction and data cache array interfaces which can be
configured to various sizes (optimized for 32KB). The processor local bus (PLB) system interface has been
extended to 128 bitsand is fully compatible with the IBM CoreConnect on-chip system architecture, providing
the framework to efficiently support system-on-a-chip (SOC) designs.
In addition, the PPC440x5 core is a member of the PowerPC 400 Series of advanced embedded processors
cores, which is supported by the PowerPC Embedded Tools Program. In this program, IBM and many third-
party vendors offer a full range of robust development tools for embedded applications. Among these are
compilers, debuggers, real-time operating systems, and logic analyzers.
PPC440x5 features include:
• High performance, dual-issue, superscalar 32-bit RISC CPU
• Superscalar implementation of the full 32-bit Book-E Enhanced PowerPC Architecture
• Seven stage, highly-pipelined micro-architecture
• Dual instruction fetch, decode, and out-of-order issue
• Out-of-order dispatch, execution, and completion
• High-accuracy dynamic branch prediction using a Branch History Table (BHT)
• Reduced branch latency using Branch Target Address Cache (BTAC)
• Three independent pipelines
• Combined complex integer, system, and branch pipeline
• Simple integer pipeline
• Load/store pipeline
• Single cycle multiply
• Single cycle multiply-accumulate (DSP instruction set extensions)
overview.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
• 9-port (6-read, 3-write) 32x32-bit General Purpose Register (GPR) file
• Hardware support for all CPU misaligned accesses
• Full support for both big and little endian byte ordering
• Extensive power management designed into core for maximum performance/power efficiency
• Primary caches
• Independently configurable instruction and data cache arrays
• Array size offerings: 32KB, 16KB, and 8KB
• Single-cycle access
• 32-byte (eight word) line size
• Highly-associative (64-way for 32KB/16KB, 32-way for 8KB)
• Write-back and write-through operation
• Control over whether stores will allocate or write-through on cache miss
• Extensive load/store queues and multiple line fill/flush buffers
• Non-blocking with up to four outstanding load misses
• Cache line locking supported
• Caches can be partitioned to provide separate regions for “transient” instructions and data
• High associativity permits efficient allocation of cache memory
• Critical word first data access and forwarding
• Cache tags and data are parity protected against soft errors.
• Memory Management Unit
• Separate instruction and data shadow TLBs
• 64-entry, fully-associative unified TLB array
• Variable page sizes (1KB-256MB), simultaneously resident in TLB
• 4-bit extended real address for 36-bit (64 GB) addressability
• Flexible TLB management with software page table search
• Storage attibute controls for write-through, caching inhibited, guarded, and byte order (endianness)
• Four user-definable storage attribute controls (for controlling CodePack™ code compression and
transient data, for example)
• TLB tags and data are parity protected against soft errors.
• Debug facilities
• Extensive hardware debug facilities incorporated into the IEEE 1149.1 JTAG port
• Multiple instruction and data address breakpoints (including range)
• Data value compare
• Single-step, branch, trap, and other debug events
• Non-invasive real-time software trace interface
• Timer facilities
– 64-bit time base
overview.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
– Decrementer with auto-reload capability
– Fixed Interval Timer (FIT)
– Watchdog Timer with critical interrupt and/or auto-reset
• Multiple core Interfaces defined by the IBM CoreConnect on-chip system architecture
• PLB interfaces
• Three independent 128-bit interfaces for instruction reads, data reads, and data writes
• Glueless attachment to 32-, 64-, or 128-bit CoreConnect system environments
• Multiple CPU:PLB frequency ratios supported (N:1, N:2, N:3)
• 6.4 GB/sec maximum data rate to CPU
• On-chip memory (OCM) integration capability over the PLB interface
• Auxiliary Processor Unit (APU) Port
• Provides functional extensions to the processor pipelines, including GPR file operations
• 128-bit load/store interface (direct access between APU and the primary data cache)
• Interface can support APU execution of all PowerPC floating point instructions
• Attachment capability for DSP co-processing such as accumulators and SIMD computation
• Enables customer-specific instruction enhancements for multimedia applications
• Device Control Register (DCR) interface for independent access to on-chip control registers
• Avoids contention for high-bandwidth PLB system bus
• Clock and power management interface
• JTAG debug interface
1.2 The PPC440x5 as a PowerPC Implementation
The PPC440x5 core implements the full, 32-bit fixed-point subset of the Book-E Enhanced PowerPC Archi-
tecture. The PPC440x5 core fully complies with these architectural specifications. The 64-bit operations of
the architecture are not supported, and the core does not implement the floating point operations, although a
floating point unit (FPU) may be attached (using the APU interface). Within the core, the 64-bit operations and
the floating point operations are trapped, and the floating point operations can be emulated using software.
See Appendix A of the Book-E Enhanced PowerPC Architecture specification for more information on 32-bit
subset implementations of the architecture.
Note: This document differs from the Book-E architecture specification in the use of bit numbering for
architected registers. Specifically, Book-E defines the full, 64-bit instruction set architecture, and
thus all registers are shown as having bit numbers from 0 to 63, with bit 63 being the least
significant. On the other hand, this document describes the PPC440x5 core, which is a 32-bit
subset implementation of the architecture. Accordingly, all architected registers are described as
being 32 bits in length, with the bits numbered from 0 to 31, and with bit 31 being the least
significant. Therefore, when this document makes reference to register bit numbers from 0 to 31,
they actually correspond to bits 32 to 63 of the same register in the Book-E architecture
specification.
overview.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
1.3 PPC440x5 Organization
The PPC440x5 core includes a seven-stage pipelined PowerPC core, which consists of a three stage, dual-
issue instruction fetch and decode unit with attached branch unit, together with three independent, 4-stage
pipelines for complex integer, simple integer, and load/store operations, respectively. The PPC440x5 core
also includes a memory management unit (MMU); separate instruction and data cache units; JTAG, debug,
and trace logic; and timer facilities.
128-bit
PLB
128-bit
PLB
Instruction Cache
(Size Configurable)
Data Cache
(Size Configurable)
MMU
64-entry
I-Cache Controller
Load/Store Queues
D-Cache Controller
Instruction
Unit
Branch
Unit
4KB
BHT
Target
Addr
Cache
Issue
Issue
0
1
Complex
Integer
Pipe
Simple
Integer
Pipe
Load
Store
Pipe
GPR
File
GPR
File
MAC
Figure 1-1. PPC440 Core Block Diagram
1.3.1 Superscalar Instruction Unit
The instruction unit of the PPC440x5 core fetches, decodes, and issues two instructions per cycle to any
combination of the three execution pipelines and/or the APU interface (see “Execution Pipelines” below, and
Auxiliary Processor Unit (APU) Port on page 36). The instruction unit includes a branch unit which provides
dynamic branch prediction using a branch history table (BHT), as well as a branch target address cache
(BTAC). These mechanisms greatly improve the branch prediction accuracy and reduce the latency of taken
branches, such that the target of a branch can usually be executed immediately after the branch itself, with no
penalty.
overview.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
1.3.2 Execution Pipelines
The PPC440x5 core contains three execution pipelines: complex integer, simple integer, and load/store.
Each pipeline consists of four stages and can access the nine-ported (six read, three write) GPR file. In order
to improve performance and avoid contention for the GPR file, there are two identical copies of it. One is dedi-
cated to the complex integer pipeline, while the other is shared by the simple integer and the load/store pipe-
lines.
The complex integer pipeline handles all arithmetic, logical, branch, and system management instructions
(such as interrupt and TLB management, move to/from system registers, and so on). This pipeline also
handles multiply and divide operations, and 24 DSP instructions that perform a variety of multiply-accumulate
operations. The complex integer pipeline multiply unit can perform 32-bit × 32-bit multiply operations with
single-cycle throughput and three-cycle latency;16-bit × 32-bit multiply operations have only two-cycle
latency. Divide operations take 33 cycles.
The simple integer pipeline can handle most arithmetic and logical operations which do not update the Condi-
tion Register (CR).
The load/store pipeline handles all load, store, and cache management instructions. All misaligned opera-
tions are handled in hardware, with no penalty on any operation which is contained within an aligned 16-byte
region. The load/store pipeline supports all operations to both big endian and little endian data regions.
and performance implications in the PPC440x5 core.
1.3.3 Instruction and Data Cache Controllers
The PPC440x5 core provides separate instruction and data cache controllers and arrays, which allow concur-
rent access and minimize pipeline stalls. The storage capacity of the cache arrays, which can range from
8KB–32KB each, depends upon the implementation. Both cache controllers have 32-byte lines, and both are
highly-associative, with 64-way set-associativity for 32KB and 16KB sizes, and 32-way set-associativity for
the 8KB size. Both caches support parity checking on the tags and data in the memory arrays, to protect
against soft errors. If a parity error is detected, the CPU will cause a machine check exception.
The PowerPC instruction set provides a rich set of cache management instructions for software-enforced
coherency. The PPC440x5 implementation also provides special debug instructions that can directly read the
tag and data arrays. See Chapter 4, “Instruction and Data Caches,” for detailed information about the instruc-
tion and data cache controllers.
The cache controllers connect to the PLB for connection to the IBM CoreConnect system-on-a-chip environ-
ment.
1.3.3.1 Instruction Cache Controller (ICC)
The ICC delivers two instructions per cycle to the instruction unit of the PPC440x5 core. The ICC also
handles the execution of the PowerPC instruction cache management instructions for coherency. The ICC
includes a speculative pre-fetch mechanism which can be configured to automatically pre-fetch a burst of up
to three additional lines upon any fetch request which misses in the instruction cache. These speculative pre-
fetches can be abandoned if the instruction execution branches away from the original instruction stream.
overview.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
The ICC supports cache line locking, at either an 8-line or 16-line granularity, depending on cache size (16-
line for 32KB, 8-line for 8KB and 16KB). In addition, the notion of a “transient” portion of the cache is
supported, in which the cache can be configured such that only a limited portion is used for instruction cache
lines from memory pages that are designated by a storage attribute from the MMU as being transient in
nature. Such memory pages would contain code which is unlikely to be reused once the processor moves on
to the next series of instruction lines, and thus performance may be improved by preventing each series of
instruction lines from overwriting all of the “regular” code in the instruction cache.
1.3.3.2 Data Cache Controller (DCC)
The DCC handles all load and store data accesses, as well as the PowerPC data cache management instruc-
tions. All misaligned accesses are handled in hardware, with those accesses that are contained within a half-
line (16 bytes) being handled as a single request. Load and store accesses which cross a 16-byte boundary
are broken into two separate accesses by the hardware.
The DCC interfaces to the APU port to provide direct load/store access to the data cache for APU load and
store operations. Such APU load and store instructions can access up to 16 bytes (one quadword) in a single
cycle.
The data cache can be operated in a store-in (copy-back) or write-through manner, according to the write-
through storage attribute specified for the memory page by the MMU. The DCC also supports both “store-
with-allocate” and “store-without-allocate” operations, such that store operations that miss in the data cache
can either “allocate” the line in the cache by reading it in and storing the new data into the cache, or alterna-
tively bypassing the cache on a miss and simply storing the data to memory. This characteristic can also be
specified on a page-by-page basis by a storage attribute in the MMU.
The DCC also supports cache line locking and “transient” data, in the same manner as the ICC (see Instruc-
The DCC provides extensive load, store, and flush queues, such that up to three outstanding line fills and up
to four outstanding load misses can be pending, and the DCC can continue servicing subsequent load and
store hits in an out-of-order fashion. Store-gathering can also be performed on caching inhibited, write-
through, and “without-allocate” store operations, for up to 16 contiguous bytes. Finally, each cache line has
four separate “dirty” bits (one per doubleword), so that the amount of data flushed on cache line replacement
can be minimized.
1.3.4 Memory Management Unit (MMU)
The PPC440x5 supports a flat, 36-bit (64GB) real (physical) address space. This 36-bit real address is gener-
ated by the MMU, as part of the translation process from the 32-bit effective address, which is calculated by
the processor core as an instruction fetch or load/store address.
The MMU provides address translation, access protection, and storage attribute control for embedded appli-
cations. The MMU supports demand paged virtual memory and other management schemes that require
precise control of logical to physical address mapping and flexible memory protection. Working with appro-
priate system level software, the MMU provides the following functions:
• Translation of the 32-bit effective address space into the 36-bit real address space
• Page level read, write, and execute access control
• Storage attributes for cache policy, byte order (endianness), and speculative memory access
• Software control of page replacement strategy
overview.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
The translation lookaside buffer (TLB) is the primary hardware resource involved in the control of translation,
protection, and storage attributes. It consists of 64 entries, each specifying the various attributes of a given
page of the address space. The TLB is fully-associative; the entry for a given page can be placed anywhere
in the TLB. The TLB tag and data memory arrays are parity protected against soft errors; if a parity error is
detected, the CPU will cause a machine check exception.
Software manages the establishment and replacement of TLB entries. This gives system software significant
flexibility in implementing a custom page replacement strategy. For example, to reduce TLB thrashing or
translation delays, software can reserve several TLB entries for globally accessible static mappings. The
instruction set provides several instructions for managing TLB entries. These instructions are privileged and
the processor must be in supervisor state in order for them to be executed.
The first step in the address translation process is to expand the effective address into a virtual address. This
is done by taking the 32-bit effective address and appending to it an 8-bit Process ID (PID), as well as a 1-bit
Management”). The AS identifier is provided by the Machine State Register (MSR, see Chapter 6, “Interrupts
and Exceptions,” which contains separate bits for the instruction fetch address space (MSR[IS]) and the data
access address space (MSR[DS]). Together, the 32-bit effective address, the 8-bit PID, and the 1-bit AS form
a 41-bit virtual address. This 41-bit virtual address is then translated into the 36-bit real address using the
TLB.
The MMU divides the address space (whether effective, virtual, or real) into pages. Eight page sizes (1KB,
4KB, 16KB, 64KB, 256KB, 1MB, 16MB, 256MB) are simultaneously supported, such that at any given time
the TLB can contain entries for any combination of page sizes. In order for an address translation to occur, a
valid entry for the page containing the virtual address must be in the TLB. An attempt to access an address
for which no TLB entry exists causes an Instruction (for fetches) or Data (for load/store accesses) TLB Error
exception.
To improve performance, both the instruction cache and the data cache maintain separate “shadow” TLBs.
The instruction shadow TLB (ITLB) contains four entries, while the data shadow TLB (DTLB) contains eight.
These shadow arrays minimize TLB contention between instruction fetch and data load/store operations. The
instruction fetch and data access mechanisms only access the main 64-entry unified TLB when a miss occurs
in the respective shadow TLB. The penalty for a miss in either of the shadow TLBs is three cycles. Hardware
manages the replacement and invalidation of both the ITLB and DTLB; no system software action is required.
Each TLB entry provides separate user state and supervisor state read, write, and execute permission
controls for the memory page associated with the entry. If software attempts to access a page for which it
does not have the necessary permission, an Instruction (for fetches) or Data (for load/store accesses)
Storage exception will occur.
Each TLB entry also provides a collection of storage attributes for the associated page. These attributes
control cache policy (such as cachability and write-through as opposed to copy-back behavior), byte order
(big endian as opposed to little endian), and enabling of speculative access for the page. In addition, a set of
four, user-definable storage attributes are provided. These attributes can be used to control various system-
level behaviors, such as instruction compression using IBM CodePack technology. They can also be config-
ured to control whether data cache lines are allocated upon a store miss, and whether accesses to a given
overview.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
1.3.5 Timers
The PPC440x5 contains a Time Base and three timers: a Decrementer (DEC), a Fixed Interval Timer (FIT),
and a Watchdog Timer. The Time Base is a 64-bit counter which gets incremented at a frequency either
equal to the processor core clock rate or as controlled by a separate asynchronous timer clock input to the
core. No interrupt is generated as a result of the Time Base wrapping back to zero.
The DEC is a 32-bit register that is decremented at the same rate at which the Time Base is incremented.
The user loads the DEC register with a value to create the desired interval. When the register is decremented
to zero, a number of actions occur: the DEC stops decrementing, a status bit is set in the Timer Status
Register (TSR), and a Decrementer exception is reported to the interrupt mechanism of the PPC440x5 core.
Optionally, the DEC can be programmed to reload automatically the value contained in the Decrementer
Auto-Reload register (DECAR), after which the DEC resumes decrementing. The Timer Control Register
(TCR) contains the interrupt enable for the Decrementer interrupt.
The FIT generates periodic interrupts based on the transition of a selected bit from the Time Base. Users can
select one of four intervals for the FIT period by setting a control field in the TCR to select the appropriate bit
from the Time Base. When the selected Time Base bit transitions from 0 to 1, a status bit is set in the TSR
and a Fixed Interval Timer exception is reported to the interrupt mechanism of the PPC440x5 core. The FIT
interrupt enable is contained in the TCR.
Similar to the FIT, the Watchdog Timer also generates a periodic interrupt based on the transition of a
selected bit from the Time Base. Users can select one of four intervals for the watchdog period, again by
setting a control field in the TCR to select the appropriate bit from the Time Base. Upon the first transition
from 0 to 1 of the selected Time Base bit, a status bit is set in the TSR and a Watchdog Timer exception is
reported to the interrupt mechanism of the PPC440x5 core. The Watchdog Timer can also be configured to
initiate a hardware reset if a second transition of the selected Time Base bit occurs prior to the first Watchdog
exception being serviced. This capability provides an extra measure of recoverability from potential system
lock-ups.
1.3.6 Debug Facilities
The PPC440x5 debug facilities include debug modes for the various types of debugging used during hard-
ware and software development. Also included are debug events that allow developers to control the debug
process. Debug modes and debug events are controlled using debug registers in the chip. The debug regis-
ters are accessed either through software running on the processor, or through the JTAG port.
The debug modes, events, controls, and interfaces provide a powerful combination of debug facilities for
hardware development tools, such as the RISCWatch™ debugger from IBM.
1.3.6.1 Debug Modes
The PPC440x5 core supports four debug modes: internal, external, real-time-trace, and debug wait. Each
mode supports a different type of debug tool used in embedded systems development. Internal debug mode
supports software-based ROM monitors, and external debug mode supports a hardware emulator type of
debug. Real-time-trace mode uses the debug facilities to indicate events within a trace of processor execu-
overview.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
tion in real time. Debug wait mode enables the processor to continue to service real-time critical interrupts
while instruction execution is otherwise stopped for hardware debug. The debug modes are controlled by
Debug Control Register 0 (DBCR0) and the setting of bits in the Machine State Register (MSR).
Internal debug mode supports accessing architected processor resources, setting hardware and software
breakpoints, and monitoring processor status. In internal debug mode, debug events can generate debug
exceptions, which can interrupt normal program flow so that monitor software can collect processor status
and alter processor resources.
Internal debug mode relies on exception-handling software—running on the processor—along with an
external communications path to debug software problems. This mode is used while the processor continues
executing instructions and enables debugging of problems in application or operating system code. Access to
debugger software executing in the processor while in internal debug mode is through a communications port
on the processor board, such as a serial port or ethernet connection.
External debug mode supports stopping, starting, and single-stepping the processor, accessing architected
processor resources, setting hardware and software breakpoints, and monitoring processor status. In
external debug mode, debug events can architecturally “freeze” the processor. While the processor is frozen,
normal instruction execution stops, and the architected processor resources can be accessed and altered
using a debug tool (such as RISCWatch) attached through the JTAG port. This mode is useful for debugging
hardware and low-level control software problems.
1.3.6.2 Development Tool Support
The PPC440x5 provides powerful debug support for a wide range of hardware and software development
tools.
The OS Open real-time operating system debugger is an example of an operating system-aware debugger,
implemented using software traps.
RISCWatch is an example of a development tool that uses the external debug mode, debug events, and the
JTAG port to support hardware and software development and debugging.
The RISCTrace™ feature of RISCWatch is an example of a development tool that uses the real-time trace
capability of the PPC440x5.
1.4 Core Interfaces
Several interfaces to the PPC440x5 core support the IBM CoreConnect on-chip system architecture, which
simplifies the attachment of on-chip devices. These interfaces include:
• Processor local bus (PLB)
• Device configuration register (DCR) interface
• Auxiliary processor unit (APU) port
• JTAG, debug, and trace ports
• Interrupt interface
• Clock and power management interface
Several of these interfaces are described briefly in the sections below.
overview.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
1.4.1 Processor Local Bus (PLB)
There are three independent 128-bit PLB interfaces to the PPC440x5 core. Each of these interfaces includes
a 36-bit address bus and a 128-bit data bus. One PLB interface supports instruction cache reads, while the
other two support data cache reads and writes, respectively. The frequency of each PLB interface can be
independently specified, allowing an IBM CoreConnect system in which the interfaces are not all connected
as part of the same PLB and in which each PLB subsystem operates at its own frequency. Each PLB inter-
face frequency can be configured to any value such that the ratio of the processor core frequency to the PLB
(core:PLB) is n:1, n:2, or n:3, where n is any integer greater than or equal to the denominator of the ratio.
Each of the PLB interfaces supports connection to a PLB subsystem of either 32, 64, or 128 bits. The instruc-
tion and data cache controllers handle any dynamic data bus resizing which is required when the subsystem
data width is less than the 128 bits of the PPC440x5 core PLB interfaces.
The data cache PLB interfaces make requests for 32-byte lines, as well as for 1 - 15 bytes within a 16-byte
(quadword) aligned region. A 16-byte line request is used for quadword APU load operations to caching
inhibited pages, and for quadword APU store operations to caching inhibited, write-through, or “without allo-
cate” pages.
The instruction cache controller makes 32-byte line read requests, and also presents quadword burst read
requests for up to three 32-byte lines (six quadwords), as part of its speculative line fill mechanism.
Each of the PLB interfaces fully supports the address pipelining capabilities of the PLB, and in fact can go
beyond the pipeline depth and minimum latency which the PLB supports. Specifically, each interface
supports up to three pipelined request/acknowledge sequences prior to performing the data transfers associ-
ated with the first request. For the data cache, if each of the requests must themselves be broken into three
separate transactions (for example, for a misaligned doubleword request to a 32-bit PLB slave), then the
interface actually supports up to nine outstanding request/acknowledge sequences prior to the first data
transfer. Furthermore, each PLB interface tolerates a zero-cycle latency between the request and the
address and data acknowledge (that is, the request, address acknowledge, and data acknowledge may all
occur in the same cycle).
1.4.2 Device Control Register (DCR) Interface
The DCR interface provides a mechanism for the PPC440x5 core to setup other on-chip facilities. For
example, programmable resources in an external bus interface unit may be configured for usage with various
memory devices according to their transfer characteristics and address assignments. DCRs are accessed
through the use of the PowerPC mfdcr and mtdcr instructions.
The interface is interlocked with control signals such that it may be connected to peripheral units that may be
clocked at different frequencies from the processor core. The design allows for future expansion of the non-
core facilities without changing the I/O on either the PPC440x5 core or the ASIC peripherals.
The DCR interface also allows the PPC440x5 core to communicate with peripheral devices without using the
PLB interface, thereby avoiding the impact to the primary system bus bandwidth, and without additional
segmentation of the useable address map.
1.4.3 Auxiliary Processor Unit (APU) Port
This interface provides the PPC440x5 core with the flexibility for attaching a tightly-coupled coprocessor-type
macro incorporating instructions which go beyond those provided within the processor core itself. The APU
port provides sufficient functionality for attachment of various coprocessor functions such as a fully-compliant
overview.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
PowerPC floating point unit (single or double precision), multimedia engine, DSP, or other custom function
implementing algorithms appropriate for specific system applications. The APU interface supports dual-issue
pipeline designs, and can be used with macros that contain their own register files, or with simpler macros
which use the CPU GPR file for source and/or target operands. APU load and store instructions can directly
access the PPC440x5 data cache, with operands of up to a quadword (16 bytes) in length.
The APU interface provides the capability for a coprocessor to execute concurrently with the PPC440x5 core
instructions that are not part of the PowerPC instruction set. Accordingly, areas have been reserved within
the architected instruction space to allow for these customer-specific or application-specific APU instruction
set extensions.
1.4.4 JTAG Port
The PPC440x5 JTAG port is enhanced to support the attachment of a debug tool such as the RISCWatch
product from IBM. Through the JTAG test access port, and using the debug facilities designed into the
PPC440x5 core, a debug workstation can single-step the processor and interrogate internal processor state
to facilitate hardware and software debugging. The enhancements comply with the IEEE 1149.1 specification
for vendor-specific extensions, and are therefore compatible with standard JTAG hardware for boundary-
scan system testing.
overview.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
overview.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
2. Programming Model
The programming model of the PPC440x5 core describes how the following features and operations of the
core appear to programmers:
2.1 Storage Addressing
As a 32-bit implementation of the Book-E Enhanced PowerPC Architecture, the PPC440x5 core implements
a uniform 32-bit effective address (EA) space. Effective addresses are expanded into virtual addresses and
then translated to 36-bit (64GB) real addresses by the memory management unit (see Memory Management
on page 133 for more information on the translation process). The organization of the real address space into
a physical address space is system-dependent, and is described in the user’s manuals for chip-level products
that incorporate a PPC440x5 core.
The PPC440x5 generates an effective address whenever it executes a storage access, branch, cache
management, or translation lookaside buffer (TLB) management instruction, or when it fetches the next
sequential instruction.
2.1.1 Storage Operands
Bytes in storage are numbered consecutively starting with 0. Each number is the address of the corre-
sponding byte.
Data storage operands accessed by the integer load/store instructions may be bytes, halfwords, words, or—
for load/store multiple and string instructions—a sequence of words or bytes, respectively. Data storage oper-
ands accessed by auxiliary processor (AP) load/store instructions can be bytes, halfwords, words, double-
words, or quadwords. The address of a storage operand is the address of its first byte (that is, of its lowest-
numbered byte). Byte ordering can be either big endian or little endian, as controlled by the endian storage
attribute (see Byte Ordering on page 42; also see Endian (E) on page 146 for more information on the endian
storage attribute).
Operand length is implicit for each scalar storage access instruction type (that is, each storage access
instruction type other than the load/store multiple and string instructions). The operand of such a scalar
storage access instruction has a “natural” alignment boundary equal to the operand length. In other words,
the ‘natural’ address of an operand is an integral multiple of the operand length. A storage operand is said to
be aligned if it is aligned at its natural boundary: otherwise it is said to be unaligned.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Data storage operands for storage access instructions have the following characteristics.
Table 2-1. Data Operand Definitions
Storage Access Instruction
Type
Operand
Length
Addr[28:31] if aligned
0bxxxx
Byte (or String)
8 bits
Halfword
2 bytes
4 bytes
8 bytes
16 bytes
0bxxx0
0bxx00
0bx000
0b0000
Word (or Multiple)
Doubleword (AP only)
Quadword (AP only)
Note: An “x” in an address bit position indicates that the bit can be 0
or 1 independent of the state of other bits in the address.
The alignment of the operand effective address of some storage access instructions may affect performance,
and in some cases may cause an Alignment exception to occur. For such storage access instructions, the
best performance is obtained when the storage operands are aligned. Table 2-2 summarizes the effects of
alignment on those storage access instruction types for which such effects exist. If an instruction type is not
shown in the table, then there are no alignment effects for that instruction type.
Table 2-2. Alignment Effects for Storage Access Instructions
Storage Access
Alignment Effects
InstructionType
Broken into two byte accesses if crosses 16-byte boundary (EA[28:31] = 0b1111); otherwise
no effect
Integer load/store halfword
Broken into two accesses if crosses 16-byte boundary (EA[28:31] > 0b1100); otherwise no
effect
Integer load/store word
Broken into a series of 4-byte accesses until the last byte is accessed or a 16-byte boundary is
reached, whichever occurs first. If bytes remain past a 16-byte boundary, resume accessing 4
bytes at a time until the last byte is accessed or the next 16-byte boundary is reached, which-
Integer load/store multiple or string
ever occurs first; repeat.
Alignment exception if crosses 16-byte boundary (EA[28:31] = 0b1111); otherwise no effect
(see note)
AP load/store halfword
Alignment exception if crosses 16-byte boundary (EA[28:31] > 0b1100); otherwise no effect
(see note)
AP load/store word
Alignment exception if crosses 16-byte boundary (EA[28:31] > 0b1000); otherwise no effect
(see note)
AP load/store doubleword
Alignment exception if crosses 16-byte boundary (EA[28:31] ≠ 0b0000); otherwise no effect
AP load/store quadword
Note: An auxiliary processor can specify that the EA for a given AP load/store instruction must be aligned at the
operand-size boundary, or alternatively, at a word boundary. If the AP so indicates this requirement and the
calculated EA fails to meet it, the PPC440x5 core generates an Alignment exception. Alternatively, an
auxiliary processor can specify that the EA for a given AP load/store instruction should be “forced” to be
aligned, by ignoring the appropriate number of low-order EA bits and processing the AP load/store as if
those bits were 0. Byte, halfword, word, doubleword, and quadword AP load/store instructions would ignore
0, 1, 2, 3, and 4 low-order EA bits, respectively.
Cache management instructions access cache block operands, and for the PPC440x5 core the cache block
size is 32 bytes. However, the effective addresses calculated by cache management instructions are not
required to be aligned on cache block boundaries. Instead, the architecture specifies that the associated low-
order effective address bits (bits 27:31 for PPC440x5) are ignored during the execution of these instructions.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Similarly, the TLB management instructions access page operands, and—as determined by the page size—
the associated low-order effective address bits are ignored during the execution of these instructions.
Instruction storage operands, on the other hand, are always four bytes long, and the effective addresses
calculated by Branch instructions are therefore always word-aligned.
2.1.2 Effective Address Calculation
For a storage access instruction, if the sum of the effective address and the operand length exceeds the
32
maximum effective address of 2 –1 (that is, the storage operand itself crosses the maximum address
boundary), the result of the operation is undefined, as specified by the architecture. The PPC440x5 core
performs the operation as if the storage operand wrapped around from the maximum effective address to
effective address 0. Software, however, should not depend upon this behavior, so that it may be ported to
other implementations that do not handle this scenario in the same fashion. Accordingly, software should
ensure that no data storage operands cross the maximum address boundary.
Note that since instructions are words and since the effective addresses of instructions are always implicitly
on word boundaries, it is not possible for an instruction storage operand to cross any word boundary,
including the maximum address boundary.
Effective address arithmetic, which calculates the starting address for storage operands, wraps around from
the maximum address to address 0, for all effective address computations except next sequential instruction
fetching. See Instruction Storage Addressing Modes on page 41 for more information on next sequential
instruction fetching at the maximum address boundary.
2.1.2.1 Data Storage Addressing Modes
There are two data storage addressing modes supported by the PPC440x5 core:
• Base + displacement (D-mode) addressing mode:
The 16-bit D field is sign-extended and added to the contents of the GPR designated by RA or to zero if
RA = 0; the low-order 32 bits of the sum form the effective address of the data storage operand.
• Base + index (X-mode) addressing mode:
The contents of the GPR designated by RB (or the value 0 for lswi and stswi) are added to the contents
of the GPR designated by RA, or to 0 if RA = 0; the low-order 32 bits of the sum form the effective
address of the data storage operand.
2.1.2.2 Instruction Storage Addressing Modes
There are four instruction storage addressing modes supported by the PPC440x5 core:
• I-form branch instructions (unconditional):
The 24-bit LI field is concatenated on the right with 0b00, sign-extended, and then added to either the
address of the branch instruction if AA=0, or to 0 if AA=1; the low-order 32 bits of the sum form the effec-
tive address of the next instruction.
• Taken B-form branch instructions:
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
The 14-bit BD field is concatenated on the right with 0b00, sign-extended, and then added to either the
address of the branch instruction if AA=0, or to 0 if AA=1; the low-order 32 bits of the sum form the effec-
tive address of the next instruction.
• Taken XL-form branch instructions:
The contents of bits 0:29 of the Link Register (LR) or the Count Register (CTR) are concatenated on the
right with 0b00 to form the 32-bit effective address of the next instruction.
• Next sequential instruction fetching (including non-taken branch instructions):
The value 4 is added to the address of the current instruction to form the 32-bit effective address of the
next instruction. If the address of the current instruction is 0xFFFFFFFC, the PPC440x5 core wraps the
next sequential instruction address back to address 0. This behavior is not required by the architecture,
which specifies that the next sequential instruction address is undefined under these circumstances.
Therefore, software should not depend upon this behavior, so that it may be ported to other implementa-
tions that do not handle this scenario in the same fashion. Accordingly, if software wishes to execute
across this maximum address boundary and wrap back to address 0, it should place an unconditional
branch at the boundary, with a displacement of 4.
In addition to the above four instruction storage addressing modes, the following behavior applies to
branch instructions:
• Any branch instruction with LK=1:
The value 4 is added to the address of the current instruction and the low-order 32 bits of the result are
placed into the LR. As for the similar scenario for next sequential instruction fetching, if the address of the
branch instruction is 0xFFFF FFFC, the result placed into the LR is architecturally undefined, although
once again the PPC440x5 core wraps the LR update value back to address 0. Again, however, software
should not depend on this behavior, in order that it may be ported to implementations which do not han-
dle this scenario in the same fashion.
2.1.3 Byte Ordering
If scalars (individual data items and instructions) were indivisible, there would be no such concept as “byte
ordering.” It is meaningless to consider the order of bits or groups of bits within the smallest addressable unit
of storage, because nothing can be observed about such order. Only when scalars, which the programmer
and processor regard as indivisible quantities, can comprise more than one addressable unit of storage does
the question of order arise.
For a machine in which the smallest addressable unit of storage is the 64-bit doubleword, there is no question
of the ordering of bytes within doublewords. All transfers of individual scalars between registers and storage
are of doublewords, and the address of the byte containing the high-order eight bits of a scalar is no different
from the address of a byte containing any other part of the scalar.
For the Book-E Enhanced PowerPC Architecture, as for most current computer architectures, the smallest
addressable unit of storage is the 8-bit byte. Many scalars are halfwords, words, or doublewords, which
consist of groups of bytes. When a word-length scalar is moved from a register to storage, the scalar occu-
pies four consecutive byte addresses. It thus becomes meaningful to discuss the order of the byte addresses
with respect to the value of the scalar: which byte contains the highest-order eight bits of the scalar, which
byte contains the next-highest-order eight bits, and so on.
Given a scalar that contains multiple bytes, the choice of byte ordering is essentially arbitrary. There are 4! =
24 ways to specify the ordering of four bytes within a word, but only two of these orderings are sensible:
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
• The ordering that assigns the lowest address to the highest-order (“left-most”) eight bits of the scalar, the
next sequential address to the next-highest-order eight bits, and so on.
This ordering is called big endian because the “big end” (most-significant end) of the scalar, considered
as a binary number, comes first in storage. IBM RISC System/6000, IBM System/390, and Motorola
680x0 are examples of computer architectures using this byte ordering.
• The ordering that assigns the lowest address to the lowest-order (“right-most”) eight bits of the scalar, the
next sequential address to the next-lowest-order eight bits, and so on.
This ordering is called little endian because the “little end” (least-significant end) of the scalar, considered
as a binary number, comes first in storage. The Intel x86 is an example of a processor architecture using
this byte ordering.
PowerPC Book-E supports both big endian and little endian byte ordering, for both instruction and data
storage accesses. Which byte ordering is used is controlled on a memory page basis by the endian (E)
storage attribute, which is a field within the TLB entry for the page. The endian storage attribute is set to 0 for
a big endian page, and is set to 1 for a little endian page. See Memory Management on page 133 for more
information on memory pages, the TLB, and storage attributes, including the endian storage attribute.
2.1.3.1 Structure Mapping Examples
The following C language structure, s, contains an assortment of scalars and a character string. The
comments show the value assumed to be in each structure element; these values show how the bytes
comprising each structure element are mapped into storage.
struct {
int a;
/* 0x1112_1314 word */
long long b;
char *c;
char d[7];
short e;
int f;
/* 0x2122_2324_2526_2728 doubleword */
/* 0x3132_3334 word */
/* 'A','B','C','D','E','F','G' array of bytes */
/* 0x5152 halfword */
/* 0x6162_6364 word */
} s;
C structure mapping rules permit the use of padding (skipped bytes) to align scalars on desirable boundaries.
The structure mapping examples below show each scalar aligned at its natural boundary. This alignment
introduces padding of four bytes between a and b, one byte between d and e, and two bytes between e and
f. The same amount of padding is present in both big endian and little endian mappings.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Big Endian Mapping
The big endian mapping of structure s follows (the data is highlighted in the structure mappings). Addresses,
in hexadecimal, are below the data stored at the address. The contents of each byte, as defined in structure
s, is shown as a (hexadecimal) number or character (for the string elements). The shaded cells correspond to
padded bytes.
11
0x00
21
12
0x01
22
13
0x02
23
14
0x03
24
0x04
25
0x05
26
0x06
0x07
27
28
0x08
0x09
0x0A 0x0B
0x0C
0x0D
0x0E 0x0F
31
32
33
0x12
'G'
34
'A'
'B'
'C'
'D'
0x10
'E'
0x11
'F'
0x13
0x14
51
0x15
52
0x16
0x17
0x18
0x19
0x1A 0x1B
0x1C
0x1D
0x1E 0x1F
61
62
63
64
0x20
0x21
0x22
0x23
0x24
0x25
0x26
0x27
Little Endian Mapping
Structure s is shown mapped little endian.
14
0x00
28
13
0x01
27
12
0x02
26
11
0x03
25
0x04
24
0x05
23
0x06
0x07
22
21
0x08
0x09
0x0A 0x0B
0x0C
0x0D
0x0E 0x0F
34
33
32
0x12
'G'
31
'A'
'B'
'C'
'D'
0x10
'E'
0x11
'F'
0x13
0x14
52
0x15
51
0x16
0x17
0x18
64
0x19
63
0x1A 0x1B
0x1C
0x1D
0x1E 0x1F
62
61
0x20
0x21
0x22
0x23
0x24
0x25
0x26
0x27
2.1.3.2 Instruction Byte Ordering
PowerPC Book-E defines instructions as aligned words (four bytes) in memory. As such, instructions in a big
endian program image are arranged with the most-significant byte (MSB) of the instruction word at the
lowest-numbered address.
Consider the big endian mapping of instruction p at address 0x00, where, for example, p = add r7, r7, r4:
MSB
0x00
LSB
0x01
0x02
0x03
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
On the other hand, in a little endian mapping the same instruction is arranged with the least-significant byte
(LSB) of the instruction word at the lowest-numbered address:
LSB
MSB
0x03
0x00
0x01
0x02
By the definition of PowerPC Book-E bit numbering, the most-significant byte of an instruction is the byte
containing bits 0:7 of the instruction. As depicted in the instruction format diagrams (see Instruction Formats
this difference in byte orderings, the processor must perform whatever byte reversal is required (depending
on the particular byte ordering in use) in order to correctly deliver the opcode field to the instruction decoder.
In the PPC440x5, this reversal is performed between the memory interface and the instruction cache,
according to the value of the endian storage attribute for each memory page, such that the bytes in the
instruction cache are always correctly arranged for delivery directly to the instruction decoder.
If the endian storage attribute for a memory page is reprogrammed from one byte ordering to the other, the
contents of the memory page must be reloaded with program and data structures that are in the appropriate
byte ordering. Furthermore, anytime the contents of instruction memory change, the instruction cache must
be made coherent with the updates by invalidating the instruction cache and refetching the updated memory
contents with the new byte ordering.
2.1.3.3 Data Byte Ordering
Unlike instruction fetches, data accesses cannot be byte-reversed between memory and the data cache.
Data byte ordering in memory depends upon the data type (byte, halfword, word, and so on) of a specific data
item. It is only when moving a data item of a specific type from or to an architected register (as directed by the
execution of a particular storage access instruction) that it becomes known what kind of byte reversal may be
required due to the byte ordering of the memory page containing the data item. Therefore, byte reversal
during load or store accesses is performed between data cache (or memory, on a data cache miss, for
example) and the load register target or store register source, depending on the specific type of load or store
instruction (that is, byte, halfword, word, and so on).
Comparing the big endian and little endian mappings of structure s, as shown in Structure Mapping Examples
size of the particular data item. For example (again referring to the big endian and little endian mappings of
structure s):
• The word a has its four bytes reversed within the word spanning addresses 0x00 – 0x03.
• The halfword e has its two bytes reversed within the halfword spanning addresses 0x1C – 0x1D.
Note that the array of bytes d, where each data item is a byte, is not reversed when the big endian and little
endian mappings are compared. For example, the character 'A' is located at address 0x14 in both the big
endian and little endian mappings.
The size of the data item being loaded or stored must be known before the processor can decide whether,
and if so, how to reorder the bytes when moving them between a register and the data cache (or memory).
• For byte loads and stores, including strings, no reordering of bytes occurs, regardless of byte ordering.
• For halfword loads and stores, bytes are reversed within the halfword, for one byte order with respect to
the other.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
• For word loads and stores (including load/store multiple), bytes are reversed within the word, for one byte
order with respect to the other.
• For doubleword loads and stores (AP loads/stores only), bytes are reversed within the doubleword, for
one byte order with respect to the other.
• For quadword loads and stores (AP loads/stores only), bytes are reversed within the quadword, for one
byte order with respect to the other.
Note that this mechanism applies independent of the alignment of data. In other words, when loading a multi-
byte data operand with a scalar load instruction, bytes are accessed from the data cache (or memory) starting
with the byte at the calculated effective address and continuing with consecutively higher-numbered bytes
until the required number of bytes have been retrieved. Then, the bytes are arranged such that either the byte
from the highest-numbered address (for big endian storage regions) or the lowest-numbered address (for
little endian storage regions) is placed into the least-significant byte of the register. The rest of the register is
filled in corresponding order with the rest of the accessed bytes. An analogous procedure is followed for
scalar store instructions.
For load/store multiple instructions, each group of four bytes is transferred between memory and the register
according to the procedure for a scalar load word instruction.
For load/store string instructions, the most-significant byte of the first register is transferred to or from memory
at the starting (lowest-numbered) effective address, regardless of byte ordering. Subsequent register bytes
(from most-significant to least-significant, and then moving into the next register, starting with the most-signif-
icant byte, and so on) are transferred to or from memory at sequentially higher-numbered addresses. This
behavior for byte strings ensures that if two strings are loaded into registers and then compared, the first
bytes of the strings are treated as most significant with respect to the comparison.
2.1.3.4 Byte-Reverse Instructions
PowerPC Book-E defines load/store byte-reverse instructions which can access storage which is specified as
being of one byte ordering in the same manner that a regular (that is, non-byte-reverse) load/store instruction
would access storage which is specified as being of the opposite byte ordering. In other words, a load/store
byte-reverse instruction to a big endian memory page transfers data between the data cache (or memory)
and the register in the same manner that a normal load/store would transfer the data to or from a little endian
memory page. Similarly, a load/store byte-reverse instruction to a little endian memory page transfers data
between the data cache (or memory) and the register in the same manner that a normal load/store would
transfer the data to or from a big endian memory page.
The function of the load/store byte-reverse instructions is useful when a particular memory page contains a
combination of data with both big endian and little endian byte ordering. In such an environment, the Endian
storage attribute for the memory page would be set according to the predominant byte ordering for the page,
and the normal load/store instructions would be used to access data operands which used this predominant
byte ordering. Conversely, the load/store byte-reverse instructions would be used to access the data oper-
ands which were of the other (less prevalent) byte ordering.
Software compilers cannot typically make general use of the load/store byte-reverse instructions, so they are
ordinarily used only in special, hand-coded device drivers.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
2.2 Registers
This section provides an overview of the register categories and types provided by the PPC440x5. Detailed
descriptions of each of the registers are provided within the chapters covering the functions with which they
are associated (for example, the cache control and cache debug registers are described in Instruction and
Data Caches on page 95). An alphabetical summary of all registers, including bit definitions, is provided in
All registers in the PPC440x5 core are architected as 32 bits wide, although certain bits in some registers are
reserved and thus not necessarily implemented. For all registers with fields marked as reserved, these
reserved fields should be written as 0 and read as undefined. The recommended coding practice is to
perform the initial write to a register with reserved fields set to 0, and to perform all subsequent writes to the
register using a read-modify-write strategy: read the register; use logical instructions to alter defined fields,
leaving reserved fields unmodified; and write the register.
All of the registers are grouped into categories according to the processor functions with which they are asso-
ciated. In addition, each register is classified as being of a particular type, as characterized by the specific
instructions which are used to read and write registers of that type. Finally, most of the registers contained
within the PPC440x5 core are defined by the Book-E Enhanced PowerPC Architecture, although some regis-
ters are implementation-specific and unique to the PPC440x5.
Figure 2-1 on page 48 illustrates the PPC440x5 registers contained in the user programming model, that is,
those registers to which access is non-privileged and which are available to both user and supervisor
programs. Figure 2-2 on page 49 illustrates the PPC440x5 registers contained in the supervisor program-
ming model, to which access is privileged and which are available to supervisor programs only. See User and
Supervisor Modes on page 80 for more information on privileged instructions and register access, and the
user and supervisor programming models.
Table 2-3 on page 50, lists each register category and the registers that belong to each category, along with
their types and a cross-reference to the section of this document which describes them more fully. Registers
that are not part of PowerPC Book-E, and are thus specific to the PPC440x5, are shown in italics in
Table 2-3. Unless otherwise indicated, all registers have read/write access.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Integer Processing
Branch Control
General Purpose
Condition Register
CR
Count Register
CTR
GPR0
GPR1
GPR2
•
•
•
Link Register
LR
GPR31
Processor Control
Integer Exception Register
XER
SPR General 4–7
SPRG4
SPRG5
Timer
Time Base
SPRG5
TBL
TBU
SPRG7
User SPR General 0
USPRG0
Figure 2-1. User Programming Model Registers
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
Processor Control
PPC440x5 CPU Core
Timer
Time Base
Storage Control
Machine State Register
Process ID
MSR
TBU
PID
Processor Version Register
PVR
TBL
MMU Control Register
MMUCR
Timer Control Register
Processor ID Register
TCR
Timer Status Register
TSR
Debug
Debug Status Register
PIR
Core Configuration Registers
CCR0
DBSR
Debug Data Register
DBDR
Decrementer
CCR1
DEC
Decrementer Auto-Reload
DECAR
Reset Configuration
RSTCFG
Debug Control Registers
DBCR0
SPR General
SPRG0
Cache Control
Instruction Cache Victim Limit
DBCR1
DBCR2
•
•
•
IVLIM
Data Address Compares
Instruction Cache Normal Victim
SPRG7
DAC1
DAC2
INV0
INV1
INV2
Interrupt Processing
Exception Syndrome Register
Data Value Compares
DVC1
ESR
Machine Check Syndrome Register
MCSR
INV3
Instruction Cache Transient Victim
ITV0
DVC2
Data Exception Address Register
DEAR
Instruction Address Compares
IAC1
IAC2
IAC3
IAC4
ITV1
ITV2
Save/Restore Registers
SRR0
ITV3
SRR1
Data Cache Victim Limit
Critical Save/Restore Registers
CSRR0
DVLIM
Cache Debug
Data Cache Normal Victim
Instruction Cache Debug Data Register
CSRR1
DNV0
DNV1
DNV2
ICDBDR
Machine Check Save/Restore Registers
MCSRR0
Instruction Cache Debug Tag Registers
ICDBTRH
ICDBTRL
MCSRR1
Interrupt Vector Prefix Register
IVPR
DNV3
Data Cache Transient Victim
Data Cache Debug Tag Registers
DTV0
DTV1
DTV2
DTV3
Interrupt Vector Offset Registers
IVOR0
DCDBTRH
DCDBTRL
•
•
•
IVOR15
Figure 2-2. Supervisor Programming Model Registers
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table 2-3. Register Categories
Register Category
Register(s)
Model and Access
Type Page
CR
User
User
User
CR
67
67
CTR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
DCR
GPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
MSR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
Branch Control
LR
66
DNV0–DNV3
DTV0–DTV3
DVLIM
Supervisor
97
Supervisor
97
Supervisor
99
Cache Control
Cache Debug
Debug
INV0–INV3
ITV0–ITV3
IVLIM
Supervisor
97
Supervisor
97
Supervisor
99
DCDBTRH, DCDBTRL
ICDBDR, ICDBTRH, ICDBTRL
DAC1–DAC2
DBCR0–DBCR2
DBDR
Supervisor, read-only
Supervisor, read-only
Supervisor
127
112
246
239
247
244
246
245
53
Supervisor
Supervisor
DBSR
Supervisor
DVC1–DVC2
IAC1–IAC4
Implemented outside core
GPR0–GPR31
XER
Supervisor
Supervisor
Supervisor
Device Control
User
71
Integer Processing
User
72
CSRR0–CSRR1
DEAR
Supervisor
168
170
172
170
171
174
169
167
108
108
165
75
Supervisor
ESR
Supervisor
IVOR0–IVOR15
IVPR
Supervisor
Interrupt Processing
Supervisor
MCSR
Supervisor
MCSRR0-MCSRR1
SRR0–SRR1
CCR0
Supervisor
Supervisor
Supervisor
CCR1
Supervisor
MSR
Supervisor
PIR, PVR
Supervisor, read-only
Supervisor, read-only
Supervisor
Processor Control
Storage Control
RSTCFG
79
SPRG0–SPRG3
SPRG4–SPRG7
USPRG0
75
User, read-only; Supervisor
User
75
75
MMUCR
Supervisor
148
151
PID
Supervisor
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 2-3. Register Categories
Register Category
Register(s)
Model and Access
Supervisor
Type Page
DEC
SPR
SPR
SPR
SPR
SPR
211
211
209
215
216
DECAR
TBL, TBU
TCR
Supervisor, write-only
User read, Supervisor write
Supervisor
Timer
TSR
Supervisor
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
2.2.1 Register Types
There are five register types contained within and/or supported by the PPC440x5 core. Each register type is
characterized by the instructions which are used to read and write the registers of that type. The following
subsections provide an overview of each of the register types and the instructions associated with them.
2.2.1.1 General Purpose Registers
The PPC440x5 core contains 32 integer general purpose registers (GPRs); each contains 32 bits. Data from
the data cache or memory can be loaded into GPRs using integer load instructions; the contents of GPRs can
be stored to the data cache or memory using integer store instructions. Most of the integer instructions refer-
ence GPRs. The GPRs are also used as targets and sources for most of the instructions which read and write
the other register types.
2.2.1.2 Special Purpose Registers
Special Purpose Registers (SPRs) are directly accessed using the mtspr and mfspr instructions. In addi-
tion, certain SPRs may be updated as a side-effect of the execution of various instructions. For example, the
Integer Exception Register (XER) (see Integer Exception Register (XER) on page 72) is an SPR which is
updated with arithmetic status (such as carry and overflow) upon execution of certain forms of integer arith-
metic instructions.
SPRs control the use of the debug facilities, timers, interrupts, memory management, caches, and other
architected processor resources. Table 10-2 on page 454 shows the mnemonic, name, and number for each
SPR, in order by SPR number. Each of the SPRs is described in more detail within the section or chapter
covering the function with which it is associated. See Table 2-3 on page 50 for a cross-reference to the asso-
ciated document section for each register.
2.2.1.3 Condition Register
The Condition Register (CR) is a 32-bit register of its own unique type and is divided up into eight, indepen-
dent 4-bit fields (CR0–CR7). The CR may be used to record certain conditional results of various arithmetic
and logical operations. Subsequently, conditional branch instructions may designate a bit of the CR as one of
the branch conditions (see Branch Processing on page 64). Instructions are also provided for performing
logical bit operations and for moving fields within the CR.
See Condition Register (CR) on page 67 for more information on the various instructions which can update
the CR.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
2.2.1.4 Machine State Register
The Machine State Register (MSR) is a register of its own unique type that controls important chip functions,
such as the enabling or disabling of various interrupt types.
The MSR can be written from a GPR using the mtmsr instruction. The contents of the MSR can be read into
a GPR using the mfmsr instruction. The MSR[EE] bit can be set or cleared atomically using the wrtee or
wrteei instructions. The MSR contents are also automatically saved, altered, and restored by the interrupt-
handling mechanism. See Machine State Register (MSR) on page 165 for more detailed information on the
MSR and the function of each of its bits.
2.2.1.5 Device Control Registers
Device Control Registers (DCRs) are on-chip registers that exist architecturally and physically outside the
PPC440x5 core, and thus are not specified by the Book-E Enhanced PowerPC Architecture, nor by this
user’s manual for the PPC440x5 core. Rather, PowerPC Book-E simply defines the existence of the DCR
address space and the instructions that access the DCRs, and does not define any particular DCRs. The
DCR access instructions are mtdcr (move to device control register) and mfdcr (move from device control
register), which move data between GPRs and the DCRs.
DCRs may be used to control various on-chip system functions, such as the operation of on-chip buses,
peripherals, and certain processor core behaviors.
2.3 Instruction Classes
PowerPC Book-E architecture defines all instructions as falling into exactly one of the following four classes,
as determined by the primary opcode (and the extended opcode, if any):
1. Defined
2. Allocated
3. Preserved
4. Reserved (-illegal or -nop)
2.3.1 Defined Instruction Class
This class of instructions consists of all the instructions defined in PowerPC Book-E. In general, defined
instructions are guaranteed to be supported within a PowerPC Book-E system as specified by the architec-
ture, either within the processor implementation itself or within emulation software supported by the system
operating software.
One exception to this is that, for implementations (such as the PPC440x5) that only provide the 32-bit subset
of PowerPC Book-E, it is not expected (and likely not even possible) that emulation of the 64-bit behavior of
the defined instructions will be provided by the system.
As defined by PowerPC Book-E, any attempt to execute a defined instruction will:
• cause an Illegal Instruction exception type Program interrupt, if the instruction is not recognized by the
implementation; or
• cause an Unimplemented Instruction exception type Program interrupt, if the instruction is recognized by
the implementation and is not a floating-point instruction, but is not supported by the implementation; or
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
• cause a Floating-Point Unavailable interrupt if the instruction is recognized as a floating-point instruction,
but floating-point processing is disabled; or
• cause an Unimplemented Instruction exception type Program interrupt, if the instruction is recognized as
a floating-point instruction and floating-point processing is enabled, but the instruction is not supported by
the implementation; or
• perform the actions described in the rest of this document, if the instruction is recognized and supported
by the implementation. The architected behavior may cause other exceptions.
The PPC440x5 core recognizes and fully supports all of the instructions in the defined class, with a few
exceptions. First, because the PPC440x5 is a 32-bit implementation, those operations which are defined
specifically for 64-bit operation are not supported at all, and will always cause an Illegal Instruction exception
type Program interrupt.
Second, instructions that are defined for floating-point processing are not supported within the PPC440x5
core, but may be implemented within an auxiliary processor and attached to the core using the AP interface.
If no such auxiliary processor is attached, attempting to execute any floating-point instructions will cause an
Illegal Instruction exception type Program interrupt. If an auxiliary processor which supports the floating-point
instructions is attached, the behavior of these instructions is as defined above and as determined by the
implementation details of the floating-point auxiliary processor.
Finally, there are two other defined instructions which are not supported within the PPC440x5 core. One is a
TLB management instruction (tlbiva, TLB Invalidate Virtual Address) that is specifically intended for
coherent multiprocessor systems. The other is mfapidi (Move From Auxiliary Processor ID Indirect), which
is a special instruction intended to assist with identification of the auxiliary processors which may be attached
to a particular processor implementation. Since the PPC440x5 core does not support mfapidi, the means of
identifying the auxiliary processors in a PPC440x5 core-based system are implementation-dependent.
Execution of either tlbiva or mfapidi will cause an Illegal Instruction exception type Program interrupt.
2.3.2 Allocated Instruction Class
This class of instructions contains a set of primary opcodes, as well as extended opcodes for certain primary
Allocated instructions are provided for purposes that are outside the scope of PowerPC Book-E, and are for
implementation-dependent and application-specific use.
PowerPC Book-E declares that any attempt to execute an allocated instruction results in one of the following
effects:
• Causes an Illegal Instruction exception type Program interrupt, if the instruction is not recognized by the
implementation
• Causes an Auxiliary Processor Unavailable interrupt if the instruction is recognized by the implementa-
tion, but allocated instruction processing is disabled
• Causes an Unimplemented Instruction exception type Program interrupt, if the instruction is recognized
and allocated instruction processing is enabled, but the instruction is not supported by the implementa-
tion
• Perform the actions described for the particular implementation of the allocated instruction. The imple-
mentation-dependent behavior may cause other exceptions.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
In addition to supporting the defined instructions of PowerPC Book-E, the PPC440x5 also implements a
number of instructions which use the allocated instruction opcodes, and thus are not part of the PowerPC
Book-E architecture. Table 2-21 on page 63 identifies the allocated instructions that are implemented within
the PPC440x5 core. All of these instructions are always enabled and supported, and thus they always
perform the functions defined for them within this document, and never cause Illegal Instruction, Auxiliary
Processor Unavailable, nor Unimplemented Instruction exceptions.
The PPC440x5 also supports the use of any of the allocated opcodes by an attached auxiliary processor,
except for those allocated opcodes which have been implemented within the PPC440x5 core, as mentioned
above. Also, there is one other allocated opcode (primary opcode 31, secondary opcode 262) that has been
implemented within the PPC440x5 core and is thus not available for use by an attached auxiliary processor.
This is the opcode which was used on previous PowerPC 400 Series embedded controllers for the icbt
(Instruction Cache Block Touch) instruction. The icbt instruction is now part of the defined instruction class
for PowerPC Book-E, and uses a new opcode (primary opcode 31, secondary opcode 22). The PPC440x5
implements the new defined opcode, but also continues to support the previous opcode, in order to support
legacy software written for earlier PowerPC 400 Series implementations. The icbt instruction description in
Instruction Set on page 249 only identifies the defined opcode, although Appendix A, “Instruction Summary,”
includes both the defined and the allocated opcode in the table which lists all the instructions by opcode. In
order to ensure portability between the PPC440x5 and future PowerPC Book-E implementations, software
should take care to only use the defined opcode for icbt, and avoid usage of the previous opcode which is
now in the allocated class.
2.3.3 Preserved Instruction Class
The preserved instruction class is provided to support backward compatibility with the PowerPC Architecture,
and/or earlier versions of the PowerPC Book-E architecture. This instruction class includes opcodes which
were defined for these previous architectures, but which are no longer defined for PowerPC Book-E.
Any attempt to execute a preserved instruction results in one of the following effects:
• Performs the actions described in the previous version of the architecture, if the instruction is recognized
by the implementation
• Causes an Illegal Instruction exception type Program interrupt, if the instruction is not recognized by the
implementation.
The only preserved instruction recognized and supported by the PPC440x5 is the mftb (Move From Time
Base) opcode. This instruction was used in the the PowerPC Architecture to read the Time Base Upper
(TBU) and Time Base Lower (TBL) registers. PowerPC Book-E architecture instead defines TBU and TBL as
Special Purpose Registers (SPRs), and thus the mfspr (Move From Special Purpose Register) instruction is
used to read them. In order to enable legacy time base management software to be run on the PPC440x5,
the core also supports the preserved opcode of mftb. However, the mftb instruction is not included in the
various sections of this document that describe the implemented instructions, and software should take care
to use the currently architected mechanism of mfspr to read the time base registers, in order to guarantee
portability between the PPC440x5 and future implementations of PowerPC Book-E.
On the other hand, Appendix A, “Instruction Summary,” does identify the mftb instruction as an implemented
preserved opcode in the table which lists all the instructions by opcode.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
2.3.4 Reserved Instruction Class
This class of instructions consists of all instruction primary opcodes (and associated extended opcodes, if
applicable) which do not belong to either the defined, allocated, or preserved instruction classes.
Reserved instructions are available for future versions of PowerPC Book-E architecture. That is, future
versions of PowerPC Book-E may define any of these instructions to perform new functions or make them
available for implementation-dependent use as allocated instructions. There are two types of reserved
instructions: reserved-illegal and reserved-nop.
Any attempt to execute a reserved-illegal instruction will cause an Illegal Instruction exception type Program
interrupt on implementations (such as the PPC440x5) that conform to the current version of PowerPC Book-
E. Reserved-illegal instructions are, therefore, available for future extensions to PowerPC Book-E that would
affect architected state. Such extensions might include new forms of integer or floating-point arithmetic
instructions, or new forms of load or store instructions that affect architected registers or the contents of
memory.
Any attempt to execute a reserved-nop instruction, on the other hand, either has no effect (that is, is treated
as a no-operation instruction), or causes an Illegal Instruction exception type Program interrupt, on imple-
mentations (such as the PPC440x5) that conform to the current version of PowerPC Book-E. Because imple-
mentations are typically expected to treat reserved-nop instructions as true no-ops, these instruction opcodes
are thus available for future extensions to PowerPC Book-E which have no effect on architected state. Such
extensions might include performance-enhancing hints, such as new forms of cache touch instructions. Soft-
ware would be able to take advantage of the functionality offered by the new instructions, and still remain
backwards-compatible with implementations of previous versions of PowerPC Book-E.
The PPC440x5 implements all of the reserved-nop instruction opcodes as true no-ops. The specific reserved-
2.4 Implemented Instruction Set Summary
This section provides an overview of the various types and categories of instructions implemented within the
PPC440x5. In addition, Instruction Set on page 249 provides a complete alphabetical listing of every imple-
mented instruction, including its register transfer language (RTL) and a detailed description of its operation.
along with a short-form description and its extended mnemonic(s).
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 2-4 summarizes the PPC440x5 instruction set by category. Instructions within each category are
described in subsequent sections.
Table 2-4. Instruction Categories
Category
Subcategory
Integer Storage Access
Integer Arithmetic
Instruction Types
load, store
add, subtract, multiply, divide, negate
and, andc, or, orc, xor, nand, nor, xnor, extend sign, count
leading zeros
Integer Logical
Integer Compare
Integer Select
Integer Trap
compare, compare logical
Integer
Branch
select operand
trap
Integer Rotate
Integer Shift
rotate and insert, rotate and mask
shift left, shift right, shift right algebraic
branch, branch conditional, branch to link, branch to count
crand, crandc, cror, crorc, crnand, crnor, crxor, crxnor
Condition Register Logical
Register Management
move to/from SPR, move to/from DCR, move to/from MSR,
write to external interrupt enable bit, move to/from CR
system call, return from interrupt, return from critical interrupt,
return from machine check interrupt
Processor Control
Storage Control
System Linkage
Processor Synchronization
Cache Management
instruction synchronize
data allocate, data invalidate, data touch, data zero, data flush,
data store, instruction invalidate, instruction touch
TLB Management
read, write, search, synchronize
Storage Synchronization
memory synchronize, memory barrier
multiply-accumulate, negative multiply-accumulate, multiply
halfword
Allocated Arithmetic
Allocated Logical
detect left-most zero byte
Allocated
data congruence-class invalidate, instruction congruence-class
invalidate
Allocated Cache Management
Allocated Cache Debug
data read, instruction read
2.4.1 Integer Instructions
Integer instructions transfer data between memory and the GPRs, and perform various operations on the
GPRs. This category of instructions is further divided into seven sub-categories, described below.
2.4.1.1 Integer Storage Access Instructions
Integer storage access instructions load and store data between memory and the GPRs. These instructions
operate on bytes, halfwords, and words. Integer storage access instructions also support loading and storing
multiple registers, character strings, and byte-reversed data, and loading data with sign-extension.
Table 2-5 shows the integer storage access instructions in the PPC440x5. In the table, the syntax “[u]” indi-
cates that the instruction has both an “update” form (in which the RA addressing register is updated with the
calculated address) and a “non-update” form. Similarly, the syntax “[x]” indicates that the instruction has both
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
an “indexed” form (in which the address is formed by adding the contents of the RA and RB GPRs) and a
“base + displacement” form (in which the address is formed by adding a 16-bit signed immediate value (spec-
ified as part of the instruction) to the contents of GPR RA. See the detailed instruction descriptions in Instruc-
Table 2-5. Integer Storage Access Instructions
Loads
Word
Stores
Word
Byte
Halfword
Multiple/String
Byte
Halfword
Multiple/String
stw[u][x]
stwbrx
stwcx.
lmw
lswi
lswx
stmw
stswi
stswx
lha[u][x]
lhbrx
lhz[u][x]
lwarx
lwbrx
lwz[u][x]
sth[u][x]
sthbrx
lbz[u][x]
stb[u][x]
2.4.1.2 Integer Arithmetic Instructions
Arithmetic operations are performed on integer or ordinal operands stored in registers. Instructions that
perform operations on two operands are defined in a three-operand format; an operation is performed on the
operands, which are stored in two registers. The result is placed in a third register. Instructions that perform
operations on one operand are defined in a two-operand format; the operation is performed on the operand in
a register and the result is placed in another register. Several instructions also have immediate formats in
which one of the source operands is a field in the instruction.
Most integer arithmetic instructions have versions that can update CR[CR0] and/or XER[SO, OV] (Summary
Overflow, Overflow), based on the result of the instruction. Some integer arithmetic instructions also update
XER[CA] (Carry) implicitly. See Integer Processing on page 71 for more information on how these instruc-
tions update the CR and/or the XER.
Table 2-6 lists the integer arithmetic instructions in the PPC440x5. In the table, the syntax “[o]” indicates that
the instruction has both an “o” form (which updates the XER[SO,OV] fields) and a “non-o” form. Similarly, the
syntax “[.]” indicates that the instruction has both a “record” form (which updates CR[CR0]) and a “non-
record” form.
Table 2-6. Integer Arithmetic Instructions
Add
Subtract
Multiply
Divide
Negate
add[o][.]
addc[o][.]
adde[o][.]
addi
subf[o][.]
subfc[o][.]
subfe[o][.]
subfic
mulhw[.]
mulhwu[.]
mulli
divw[o][.]
divwu[o][.]
neg[o][.]
addic[.]
subfme[o][.] mullw[o][.]
subfze[o][.]
addis
addme[o][.]
addze[o][.]
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
2.4.1.3 Integer Logical Instructions
Table 2-7 lists the integer logical instructions in the PPC440x5. See Integer Arithmetic Instructions on
Table 2-7. Integer Logical Instructions
Count
leading
zeros
And with
complement
Or with
complement
And
Nand
Or
Nor
Xor
Equivalence
Extend sign
and[.]
andi.
andis.
or[.]
ori
oris
xor[.]
xori
xoris
extsb[.]
extsh[.]
andc[.]
nand[.]
orc[.]
nor[.]
eqv[.]
cntlzw[.]
2.4.1.4 Integer Compare Instructions
These instructions perform arithmetic or logical comparisons between two operands and update the CR with
the result of the comparison.
Table 2-8 lists the integer compare instructions in the PPC440x5.
Table 2-8. Integer Compare Instructions
Arithmetic
Logical
cmp
cmpi
cmpl
cmpli
2.4.1.5 Integer Trap Instructions
Table 2-9 lists the integer trap instructions in the PPC440x5.
Table 2-9. Integer Trap Instructions
Trap
tw
twi
2.4.1.6 Integer Rotate Instructions
These instructions rotate operands stored in the GPRs. Rotate instructions can also mask rotated operands.
Table 2-10 lists the rotate instructions in the PPC440x5. See Integer Arithmetic Instructions on page 58 for an
explanation of the “[.]” syntax.
Table 2-10. Integer Rotate Instructions
Rotate and Insert
Rotate and Mask
rlwinm[.]
rlwnm[.]
rlwimi[.]
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
2.4.1.7 Integer Shift Instructions
Table 2-11 lists the integer shift instructions in the PPC440x5. Note that the shift right algebraic insructions
implicitly update the XER[CA] field. See Integer Arithmetic Instructions on page 58 for an explanation of the
“[.]” syntax.
Table 2-11. Integer Shift Instructions
Shift Right
Algebraic
Shift Left
Shift Right
sraw[.]
srawi[.]
slw[.]
srw[.]
2.4.1.8 Integer Select Instruction
Table 2-12 lists the integer select instruction in the PPC440x5. The RA operand is 0 if the RA field of the
instruction is 0, or is the contents of GPR[RA] otherwise.
Table 2-12. Integer Select Instruction
Integer Select
isel
2.4.2 Branch Instructions
These instructions unconditionally or conditionally branch to an address. Conditional branch instructions can
test condition codes set in the CR by a previous instruction and branch accordingly. Conditional branch
instructions can also decrement and test the Count Register (CTR) as part of branch determination, and can
save the return address in the Link Register (LR).The target address for a branch can be a displacement from
the current instruction address or an absolute address, or contained in the LR or CTR.
Table 2-13 lists the branch instructions in the PPC440x5. In the table, the syntax “[l]” indicates that the
instruction has both a “link update” form (which updates LR with the address of the instruction after the
branch) and a “non-link update” form. Similarly, the syntax “[a]” indicates that the instruction has both an
“absolute address” form (in which the target address is formed directly using the immediate field specified as
part of the instruction) and a “relative” form (in which the target address is formed by adding the specified
immediate field to the address of the branch instruction).
Table 2-13. Branch Instructions
Branch
b[l][a]
bc[l][a]
bcctr[l]
bclr[l]
2.4.3 Processor Control Instructions
Processor control instructions manipulate system registers, perform system software linkage, and synchro-
nize processor operations. The instructions in these three sub-categories of processor control instructions are
described below.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
2.4.3.1 Condition Register Logical Instructions
These instructions perform logical operations on a specified pair of bits in the CR, placing the result in
another specified bit. The benefit of these instructions is that they can logically combine the results of several
comparison operations without incurring the overhead of conditional branching between each one. Software
performance can significantly improve if multiple conditions are tested at once as part of a branch decision.
Table 2-14 lists the condition register logical instructions in the PPC440x5.
Table 2-14. Condition Register Logical Instructions
crand
crandc
creqv
crnor
cror
crorc
crxor
crnand
2.4.3.2 Register Management Instructions
These instructions move data between the GPRs and control registers in the PPC440x5.
Table 2-15 lists the register management instructions in the PPC440x5.
Table 2-15. Register Management Instructions
CR
DCR
MSR
SPR
mcrf
mcrxr
mfcr
mfmsr
mtmsr
wrtee
mfdcr
mtdcr
mfspr
mtspr
mtcrf
wrteei
2.4.3.3 System Linkage Instructions
These instructions invoke supervisor software level for system services, and return from interrupts.
Table 2-16 lists the system linkage instructions in the PPC440x5.
Table 2-16. System Linkage Instructions
rfi
rfci
rfmci
sc
2.4.3.4 Processor Synchronization Instruction
Tne processor synchronization instruction, isync, forces the processor to complete all instructions preceding
the isync before allowing any context changes as a result of any instructions that follow the isync. Addition-
ally, all instructions that follow the isync will execute within the context established by the completion of all
synchronizing effect of isync.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table 2-17 shows the processor synchronization instruction in the PPC440x5.
Table 2-17. Processor Synchronization Instruction
isync
2.4.4 Storage Control Instructions
These instructions manage the instruction and data caches and the TLB of the PPC440x5 core. Instructions
are also provided to synchronize and order storage accesses. The instructions in these three sub-categories
of storage control instructions are described below.
2.4.4.1 Cache Management Instructions
These instructions control the operation of the data and instruction caches. Instructions are provided to fill,
flush, invalidate, or zero data cache blocks, where a block is defined as a 32-byte cache line. instructions are
also provided to fill or invalidate instruction cache blocks.
Table 2-18 lists the cache management instructions in the PPC440x5.
Table 2-18. Cache Management Instructions
Data Cache
Instruction Cache
dcba
dcbf
dcbi
dcbst
dcbt
icbi
icbt
dcbtst
dcbz
2.4.4.2 TLB Management Instructions
The TLB management instructions read and write entries of the TLB array, and search the TLB array for an
entry which will translate a given virtual address. There is also an instruction for synchronizing TLB updates
with other processors, but since the PPC440x5 core is intended for use in uni-processor environments, this
instruction performs no operation on the PPC440x5.
Table 2-19 lists the TLB management instructions in the PPC440x5. See Integer Arithmetic Instructions on
Table 2-19. TLB Management Instructions
tlbre
tlbsx[.]
tlbsync
tlbwe
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
2.4.4.3 Storage Synchronization Instructions
The storage synchronization instructions allow software to enforce ordering amongst the storage accesses
caused by load and store instructions, which by default are “weakly-ordered” by the processor. “Weakly-
ordered” means that the processor is architecturally permitted to perform loads and stores generally out-of-
order with respect to their sequence within the instruction stream, with some exceptions. However, if a
storage synchronization instruction is executed, then all storage accesses prompted by instructions
preceding the synchronizing instruction must be performed before any storage accesses prompted by
instructions which come after the synchronizing instruction. See Synchronization on page 82 for more infor-
mation on storage synchronization.
Table 2-17 shows the storage synchronization instructions in the PPC440x5.
Table 2-20. Storage Synchronization Instructions
msync
mbar
2.4.5 Allocated Instructions
These instructions are not part of the PowerPC Book-E architecture, but they are included as part of the
PPC440x5 core. Architecturally, they are considered allocated instructions, as they use opcodes which are
within the allocated class of instructions, which the PowerPC Book-E architecture identifies as being available
for implementation-dependent and/or application-specific purposes. However, all of the allocated instructions
which are implemented within the PPC440x5 core are “standard” for IBM’s family of PowerPC embedded
controllers, and are not unique to the PPC440x5.
The allocated instructions implemented within the PPC440x5 are divided into four sub-categories, and are
shown in Table 2-21. See Integer Arithmetic Instructions on page 58 for an explanation of the “[.]” and “[o]”
syntax.
Table 2-21. Allocated Instructions
Cache
Management
Cache
Debug
Arithmetic
Logical
Negative
Multiply-Accumulate
Multiply-Accumulate
Multiply Halfword
macchw[o][.]
macchws[o][.]
macchwsu[o][.]
macchwu[o][.]
machhw[o][.]
machhws[o][.]
machhwsu[o][.]
machhwu[o][.]
maclhw[o][.]
nmacchw[o][.]
nmacchws[o][.]
nmachhw[o][.]
nmachhws[o][.]
nmaclhw[o][.]
nmaclhws[o][.]
mulchw[.]
mulchwu[.]
mulhhw[.]
mulhhwu[.]
mullhw[.]
dccci
iccci
dcread
icread
dlmzb[.]
mullhwu[.]
maclhws[o][.]
maclhwsu[o][.]
maclhwu[o][.]
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
2.5 Branch Processing
The four branch instructions provided by PPC440x5 are summarized in Table 2.4.2 on Page 60. In addition,
each of these instructions is described in detail in Instruction Set on page 249. The following sections provide
additional information on branch addressing, instruction fields, prediction, and registers.
2.5.1 Branch Addressing
The branch instruction (b[l][a]) specifies the displacement of the branch target address as a 26-bit value (the
24-bit LI field right-extended with 0b00). This displacement is regarded as a signed 26-bit number covering
an address range of ±32MB. Similarly, the branch conditional instruction (bc[l][a]) specifies the displacement
as a 16-bit value (the 14-bit BD field right-extended with 0b00). This displacement covers an address range of
±32KB.
For the relative form of the branch and branch conditional instructions (b[l] and bc[l], with instruction field
AA = 0), the target address is the address of the branch instruction itself (the Current Instruction Address, or
CIA) plus the signed displacement. This address calculation is defined to “wrap around” from the maximum
effective address (0xFFFFFFFF) to 0x0000 0000, and vice-versa.
For the absolute form of the branch and branch conditional instructions (ba[l] and bca[l], with instruction field
AA = 1), the target address is the sign-extended displacement. This means that with absolute forms of the
branch and branch conditional instructions, the branch target can be within the first or last 32MB or 32KB of
the address space, respectively.
The other two branch instructions, bclr (branch conditional to LR) and bcctr (branch conditional to CTR), do
not use absolute nor relative addressing. Instead, they use indirect addressing, in which the target of the
branch is specified indirectly as the contents of the LR or CTR.
2.5.2 Branch Instruction BI Field
Conditional branch instructions can optionally test one bit of the CR, as indicated by instruction field BO[0]
(see BO field description below). The value of instruction field BI specifies the CR bit to be tested (0-31). The
BI field is ignored if BO[0] = 1. The branch (b[l][a]) instruction is by definition unconditional, and hence does
not have a BI instruction field. Instead, the position of this field is part of the LI displacement field.
2.5.3 Branch Instruction BO Field
The BO field specifies the condition under which a conditional branch is taken, and whether the branch decre-
ments the CTR. The branch (b[l][a]) instruction is by definition unconditional, and hence does not have a BO
instruction field. Instead, the position of this field is part of the LI displacement field.
Conditional branch instructions can optionally test one bit in the CR. This option is selected when BO[0] = 0; if
BO[0] = 1, the CR does not participate in the branch condition test. If the CR condition option is selected, the
condition is satisfied (branch can occur) if the CR bit selected by the BI instruction field matches BO[1].
Conditional branch instructions can also optionally decrement the CTR by one, and test whether the decre-
mented value is 0. This option is selected when BO[2] = 0; if BO[2] = 1, the CTR is not decremented and
does not participate in the branch condition test. If CTR decrement option is selected, BO[3] specifies the
condition that must be satisfied to allow the branch to be taken. If BO[3] = 0, CTR ≠ 0 is required for the
branch to occur. If BO[3] = 1, CTR = 0 is required for the branch to occur.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 2-22 summarizes the usage of the bits of the BO field. BO[4] is further discussed in Branch Prediction
Table 2-22. BO Field Definition
BO Bit
Description
CR Test Control
0 Test CR bit specified by BI field for value specified by BO[1]
1 Do not test CR
BO[0]
CR Test Value
0 If BO[0] = 0, test for CR[BI] = 0.
1 If BO[0] = 0, test for CR[BI] = 1.
BO[1]
BO[2]
CTR Decrement and Test Control
0 Decrement CTR by one and test whether the decremented
CTR satisfies the condition specified by BO[3].
1 Do not decrement CTR, do not test CTR.
CTR Test Value
0 If BO[2] = 0, test for decremented CTR ≠ 0.
BO[3]
BO[4]
1 If BO[2] = 0, test for decremented CTR = 0.
Branch Prediction Reversal
0 Apply standard branch prediction.
1 Reverse the standard branch prediction.
Table 2-23 lists specific BO field contents, and the resulting actions; z represents a mandatory value of zero,
Table 2-23. BO Field Examples
BO Value
0000y
0001y
001zy
0100y
0101y
011zy
1z00y
1z01y
1z1zz
Description
Decrement the CTR, then branch if the decremented CTR ≠ 0 and CR[BI]=0.
Decrement the CTR, then branch if the decremented CTR = 0 and CR[BI] = 0.
Branch if CR[BI] = 0.
Decrement the CTR, then branch if the decremented CTR ≠ 0 and CR[BI] = 1.
Decrement the CTR, then branch if the decremented CTR=0 and CR[BI] = 1.
Branch if CR[BI] = 1.
Decrement the CTR, then branch if the decremented CTR ≠ 0.
Decrement the CTR, then branch if the decremented CTR = 0.
Branch always.
2.5.4 Branch Prediction
Conditional branches might be taken or not taken; if taken, instruction fetching is re-directed to the target
address. If the branch is not taken, instruction fetching simply falls through to the next sequential instruction.
The PPC440x5 core attempts to predict whether or not a branch is taken before all information necessary to
determine the branch direction is available. This action is called branch prediction. The core can then
prefetch instructions down the predicted path. If the prediction is correct, performance is improved because
the branch target instruction is available immediately, instead of having to wait until the branch conditions are
resolved. If the prediction is incorrect, then the prefetched instructions (which were fetched from addresses
down the “wrong” path of the branch) must be discarded, and new instructions fetched from the correct path.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
The PPC440x5 core combines the static prediction mechanism defined by PowerPC Book-E, together with a
dynamic branch prediction mechanism, in order to provide correct branch prediction as often as possible. The
dynamic branch prediction mechanism is an implementation optimization, and is not part of the architecture,
additional information on the dynamic branch prediction mechanism.
The static branch prediction mechanism enables software to designate the “preferred” branch prediction via
bits in the instruction encoding. The “default” static branch prediction for conditional branches is as follows:
Predict that the branch is to be taken if ((BO[0] ∧ BO[2]) ∨ s) = 1
where s is bit 16 of the instruction (the sign bit of the displacement for all bc forms, and zero for all bclr and
bcctr forms). In other words, conditional branches are predicted taken if they are “unconditional” (i.e., they do
not test the CR nor the CTR decrement, and are always taken), or if their branch displacement is “negative”
(i.e., the branch is branching “backwards” from the current instruction address). The standard prediction for
this case derives from considering the relative form of bc, often used at the end of loops to control the
number of times that a loop is executed. The branch is taken each time the loop is executed except the last,
so it is best if the branch is predicted taken. The branch target is the beginning of the loop, so the branch
displacement is negative and s = 1. Because this situation is most common, a branch is taken if s = 1.
If branch displacements are positive, s = 0, then the branch is predicted not taken. Also, if the branch instruc-
tion is any form of bclr or bcctr except the “unconditional” form, then s = 0, and the branch is predicted not
taken.
There is a peculiar consequence of this prediction algorithm for the absolute forms of bc (bca and bcla). As
described in Branch Addressing on page 64, if s = 1, the branch target is in high memory. If s = 0, the branch
target is in low memory. Because these are absolute-addressing forms, there is no reason to treat high and
low memory differently. Nevertheless, for the high memory case the standard prediction is taken, and for the
low memory case the standard prediction is not taken.
Another bit in the BO field allows software further control over branch prediction. Specifically, BO[4] is the
prediction reversal bit. If BO[4] = 0, the default prediction is applied. If BO[4] = 1, the reverse of the default
prediction is applied. For the cases in Table 2-23 where BO[4] = y, software can reverse the default predic-
tion by setting y to 1. This should only be done when the default prediction is likely to be wrong. Note that for
the “branch always” condition, reversal of the default prediction is not allowed, as BO[4] is designated as z for
this case, meaning the bit must be set to 0 or the instruction form is invalid.
2.5.5 Branch Control Registers
There are three registers in the PPC440x5 which are associated with branch processing, and they are
described in the following sections.
2.5.5.1 Link Register (LR)
The LR is written from a GPR using mtspr, and can be read into a GPR using mfspr. The LR can also be
updated by the “link update” form of branch instructions (instruction field LK = 1). Such branch instructions
load the LR with the address of the instruction following the branch instruction (4 + address of the branch
instruction). Thus, the LR contents can be used as a return address for a subroutine that was entered using a
link update form of branch. The bclr instruction uses the LR in this fashion, enabling indirect branching to any
address.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
When being used as a return address by a bclr instruction, bits 30:31 of the LR are ignored, since all instruc-
tion addresses are on word boundaries.
Access to the LR is non-privileged.
0
31
Figure 2-3. Link Register (LR)
0:31
Link Register contents
Target address of bclr instruction
2.5.5.2 Count Register (CTR)
The CTR is written from a GPR using mtspr, and can be read into a GPR using mfspr. The CTR contents
can be used as a loop count that gets decremented and tested by conditional branch instructions that specify
count decrement as one of their branch conditions (instruction field BO[2] = 0). Alternatively, the CTR
contents can specify a target address for the bcctr instruction, enabling indirect branching to any address.
Access to the CTR is non-privileged.
0
31
Figure 2-4. Count Register (CTR)
Used as count for branch conditional with decre-
0:31
Count
ment instructions, or as target address for bcctr
instructions
2.5.5.3 Condition Register (CR)
The CR is used to record certain information (“conditions”) related to the results of the various instructions
which are enabled to update the CR. A bit in the CR may also be selected to be tested as part of the condition
of a conditional branch instruction.
The CR is organized into eight 4-bit fields (CR0–CR7), as shown in Figure 2-5. Table 2-24 lists the instruc-
tions which update the CR.
Access to the CR is non-privileged.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
CR2
CR6
CR0
CR4
0
3
4
7
8
11 12
15 16
19 20
23 24
27 28
31
CR5
CR1
CR3
CR7
Figure 2-5. Condition Register (CR)
0:3
CR0
CR1
CR2
CR3
CR4
CR5
CR6
CR7
Condition Register Field 0
Condition Register Field 1
Condition Register Field 2
Condition Register Field 3
Condition Register Field 4
Condition Register Field 5
Condition Register Field 6
Condition Register Field 7
4:7
8:11
12:15
16:19
20:23
24:27
28:31
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 2-24. CR Updating Instructions
Processor
Control
Storage
Control
Auxiliary
Processor
Integer
CR-Logical
and Register
Management
Storage
Access
TLB
Mgmt.
Arithmetic
and Logical
Arithmetic
Logical
Compare
Rotate
Shift
macchw.[o]
macchws.[o]
macchwsu.[o]
macchwu.[o]
machhw.[o]
machhws.[o]
machhwsu.[o]
machhwu.[o]
maclhw.[o]
and.
andi.
andis.
add.[o]
addc.[o]
adde.[o]
addic.
addme.[o]
addze.[o]
andc.
nand.
crand
crandc
creqv
crnand
crnor
cror
maclhws.[o]
maclhwsu.[o]
maclhwu.[o]
subf.[o]
slw.
cmp
cmpi
or.
orc.
subfc.[o]
subfe.[o]
subfme.[o]
subfze.[o]
rlwimi.
nmacchw.[o]
nmacchws.[o]
nmachhw.[o]
nmachhws.[o]
nmaclhw.[o]
nmaclhws.[o]
srw.
stwcx.
tlbsx.
rlwinm.
rlwnm.
crorc
crxor
cmpl
cmpli
nor.
xor.
eqv.
sraw.
srawi.
mulhw.
mulhwu.
mullw.[o]
mcrf
mcrxr
mtcrf
mulchw.
mulchwu.
mulhhw.
mulhhwu.
mullhw.
divw.[o]
divwu.[o]
extsb.
extsh.
neg.[o]
cntlzw.
mullhwu.
dlmzb.
Instruction Set on page 249, provides detailed information on how each of these instructions updates the CR.
To summarize, the CR can be accessed in any of the following ways:
• mfcr reads the CR into a GPR. Note that this instruction does not update the CR and is therefore not
• Conditional branch instructions can designate a CR bit to be used as a branch condition. Note that these
• mtcrf sets specified CR fields by writing to the CR from a GPR, under control of a mask field specified as
part of the instruction.
• mcrf updates a specified CR field by copying another specified CR field into it.
• mcrxr copies certain bits of the XER into a specified CR field, and clears the corresponding XER bits.
• Integer compare instructions update a specified CR field.
• CR-logical instructions update a specified CR bit with the result of any one of eight logical operations on
a specified pair of CR bits.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
• Certain forms of various integer instructions (the “.” forms) implicitly update CR[CR0], as do certain forms
of the auxiliary processor instructions implemented within the PPC440x5 core.
• Auxiliary processor instructions may in general update a specified CR field in an implementation-speci-
fied manner. In addition, if an auxiliary processor implements the floating-point operations specified by
PowerPC Book-E, then those instructions update the CR in the manner defined by the architecture. See
Book E: PowerPC Architecture Enhanced for Embedded Applications for details.
CR[CR0] Implicit Update By Integer Instructions
various “dot-form” instructions, indicated by a “.” in the instruction mnemonic. Most of these instructions
update CR[CR0] according to an arithmetic comparison of 0 with the 32-bit result which the instruction writes
to the GPR file. That is, after performing the operation defined for the instruction, the 32-bit result which is
written to the GPR file is compared to 0 using a signed comparison, independent of whether the actual oper-
ation being performed by the instruction is considered “signed” or not. For example, logical instructions such
as and., or., and nor. update CR[CR0] according to this signed comparison to 0, even though the result of
such a logical operation is not typically interpreted as a signed value. For each of these dot-form instructions,
the individual bits in CR[CR0] are updated as follows:
CR[CR0] — LT
Less than 0; set if the most-significant bit of the 32-bit result is 1.
0
CR[CR0] — GT
Greater than 0; set if the 32-bit result is non-zero and the most-
significant bit of the result is 0.
1
CR[CR0] — EQ
Equal to 0; set if the 32-bit result is 0.
2
CR[CR0] — SO
Summary overflow; a copy of XER[SO] at the completion of the
instruction (including any XER[SO] update being performed the
instruction itself.
3
Note that if an arithmetic overflow occurs, the “sign” of an instruction result indicated in CR[CR0] might not
represent the “true” (infinitely precise) algebraic result of the instruction that set CR0. For example, if an add.
instruction adds two large positive numbers and the magnitude of the result cannot be represented as a twos-
complement number in a 32-bit register, an overflow occurs and CR[CR0] is set, even though the infinitely
0
precise result of the add is positive.
Similarly, adding the largest 32-bit twos-complement negative number (0x80000000) to itself results in an
arithmetic overflow and 0x0000 0000 is recorded in the target register. CR[CR0] is set, indicating a result of
2
0, but the infinitely precise result is negative.
CR[CR0] is a copy of XER[SO] at the completion of the instruction, whether or not the instruction which is
3
updating CR[CR0] is also updating XER[SO]. Note that if an instruction causes an arithmetic overflow but is
not of the form which actually updates XER[SO], then the value placed in CR[CR0] does not reflect the arith-
3
metic overflow which occurred on the instruction (it is merely a copy of the value of XER[SO] which was
already in the XER before the execution of the instruction updating CR[CR0]).
There are a few dot-form instructions which do not update CR[CR0] in the fashion described above. These
instructions are: stwcx., tlbsx., and dlmzb. See the instruction descriptions in Instruction Set on page 249 for
details on how these instructions update CR[CR0].
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
CR Update By Integer Compare Instructions
Integer compare instructions update a specified CR field with the result of a comparison of two 32-bit
numbers, the first of which is from a GPR and the second of which is either an immediate value or from
another GPR. There are two types of integer compare instructions, arithmetic and logical, and they are distin-
guished by the interpretation given to the 32-bit numbers being compared. For arithmetic compares, the
numbers are considered to be signed, whereas for logical compares, the numbers are considered to be
unsigned. As an example, consider the comparison of 0 with 0xFFFFFFFF. In an arithmetic compare, 0 is
larger; in a logical compare, 0xFFFFFFFF is larger.
A compare instruction can direct its result to any CR field. The BF field (bits 6:8) of the instruction specifies
the CR field to be updated. After a compare, the specified CR field is interpreted as follows:
CR[(BF)] — LT
The first operand is less than the second operand.
The first operand is greater than the second operand.
The first operand is equal to the second operand.
Summary overflow; a copy of XER[SO].
0
CR[(BF)] — GT
1
CR[(BF)] — EQ
2
CR[(BF)] — SO
3
2.6 Integer Processing
Integer processing includes loading and storing data between memory and GPRs, as well as performing
various operations on the values in GPRs and other registers (the categories of integer instructions are
summarized in Table 2-4 on page 57). The sections which follow describe the registers which are used for
integer processing, and how they are updated by various instructions. In addition, Condition Register (CR) on
page 67 provides more information on the CR updates caused by integer instructions. Finally, Instruction Set
on page 249 also provides details on the various register updates performed by integer instructions.
2.6.1 General Purpose Registers (GPRs)
The PPC440x5 contains 32 GPRs. The contents of these registers can be transferred to and from memory
using integer storage access instructions. Operations are performed on GPRs by most other instructions.
Access to the GPRs is non-privileged.
0
31
Figure 2-6. General Purpose Registers (R0-R31)
0:31
General Purpose Register data
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
2.6.2 Integer Exception Register (XER)
The XER records overflow and carry indications from integer arithmetic and shift instructions. It also provides
a byte count for string indexed integer storage access instructions (lswx and stswx). Note that the term
exception in the name of this register does not refer to exceptions as they relate to interrupts, but rather to the
arithmetic exceptions of carry and overflow.
Figure 2-7 illustrates the fields of the XER, while Tables 2-25 and 2-26 list the instructions which update
XER[SO,OV] and the XER[CA] fields, respectively. The sections which follow the figure and tables describe
the fields of the XER in more detail.
Access to the XER is non-privileged.
TBC
CA
2
SO
0
1
3
24 25
31
OV
Figure 2-7. Integer Exception Register (XER)
Summary Overflow
Can be set by mtspr or by integer or auxiliary
processor instructions with the [o] option; can be
reset by mtspr or by mcrxr.
0 No overflow has occurred.
1 Overflow has occurred.
0
SO
Can be set by mtspr or by integer or allocated
instructions with the [o] option; can be reset by
mtspr, by mcrxr, or by integer or allocated
instructions with the [o] option.
Overflow
0 No overflow has occurred.
1 Overflow has occurred.
1
OV
Can be set by mtspr or by certain integer arith-
metic and shift instructions; can be reset by
mtspr, by mcrxr, or by certain integer arithmetic
and shift instructions.
Carry
0 Carry has not occurred.
1 Carry has occurred.
2
CA
3:24
Reserved
Used as a byte count by lswx and stswx; written
by dlmzb[.] and by mtspr.
25:31
TBC
Transfer Byte Count
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 2-25. XER[SO,OV] Updating Instructions
Processor Con-
trol
Integer Arithmetic
Auxiliary Processor
Multiply-Accumu- Negative Multi-
Register Man-
agement
Add
Subtract
Multiply
Divide
Negate
late
ply- Accumulate
macchwo[.]
macchwso[.]
macchwsuo[.]
macchwuo[.]
machhwo[.]
machhwso[.]
machhwsuo[.] nmachhwso[.]
machhwuo[.]
maclhwo[.]
nmacchwo[.]
nmacchwso[.]
nmachhwo[.]
addo[.]
addco[.]
addeo[.]
addmeo[.] subfmeo[.]
addzeo[.] subfzeo[.]
subfo[.]
subfco[.]
subfeo[.]
mtspr
mcrxr
divwo[.]
divwuo[.]
mullwo[.]
nego[.]
nmaclhwo[.]
nmaclhwso[.]
maclhwso[.]
maclhwsuo[.]
maclhwuo[.]
Table 2-26. XER[CA] Updating Instructions
Integer
Shift
Processor
Control
Integer Arithmetic
Shift
Right
Algebraic
Register
Management
Add
Subtract
subfc[o][.]
subfe[o][.]
subfic
subfme[o][.]
subfze[o][.]
addc[o][.]
adde[o][.]
addic[.]
addme[o][.]
addze[o][.]
mtspr
mcrxr
sraw[.]
srawi[.]
2.6.2.1 Summary Overflow (SO) Field
This field is set to 1 when an instruction is executed that causes XER[OV] to be set to 1, except for the case
of mtspr(XER), which writes XER[SO,OV] with the values in (RS) , respectively. Once set, XER[SO] is not
0:1
reset until either an mtspr(XER) is executed with data that explicitly writes 0 to XER[SO], or until an mcrxr
instruction is executed. The mcrxr instruction sets XER[SO] (as well as XER[OV,CA]) to 0 after copying all
three fields into CR[CR0] (and setting CR[CR0] to 0).
0:2
3
Given this behavior, XER[SO] does not necessarily indicate that an overflow occurred on the most recent
integer arithmetic operation, but rather that one occurred at some time subsequent to the last clearing of
XER[SO] by mtspr(XER) or mcrxr.
XER[SO] is read (along with the rest of the XER) into a GPR by mfspr(XER). In addition, various integer
3
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
2.6.2.2 Overflow (OV) Field
This field is updated by certain integer arithmetic instructions to indicate whether the infinitely precise result of
the operation can be represented in 32 bits. For those integer arithmetic instructions that update XER[OV]
31
31
and produce signed results, XER[OV] = 1 if the result is greater than 2 – 1 or less than –2 ; otherwise,
XER[OV] = 0. For those integer arithmetic instructions that update XER[OV] and produce unsigned results
(certain integer divide instructions and multiply-accumulate auxiliary processor instructions), XER[OV] = 1 if
32
the result is greater than 2 –1; otherwise, XER[OV] = 0. See the instruction descriptions in Instruction Set on
The mtspr(XER) and mcrxr instructions also update XER[OV]. Specifically, mcrxr sets XER[OV] (and
XER[SO,CA]) to 0 after copying all three fields into CR[CR0] (and setting CR[CR0] to 0), while
0:2
3
mtspr(XER) writes XER[OV] with the value in (RS) .
1
XER[OV] is read (along with the rest of the XER) into a GPR by mfspr(XER).
2.6.2.3 Carry (CA) Field
This field is updated by certain integer arithmetic instructions (the “carrying” and “extended” versions of add
and subract) to indicate whether or not there is a carry-out of the most-significant bit of the 32-bit result.
XER[CA] = 1 indicates a carry. The integer shift right algebraic instructions update XER[CA] to indicate
whether or not any 1-bits were shifted out of the least significant bit of the result, if the source operand was
The mtspr(XER) and mcrxr instructions also update XER[CA]. Specifically, mcrxr sets XER[CA] (as well as
XER[SO,OV]) to 0 after copying all three fields into CR[CR0] (and setting CR[CR0] to 0), while
0:2
3
mtspr(XER) writes XER[CA] with the value in (RS) .
2
XER[CA] is read (along with the rest of the XER) into a GPR by mfspr(XER). In addition, the “extended”
versions of the add and subtract integer arithmetic instructions use XER[CA] as a source operand for their
arithmetic operations.
Transfer Byte Count (TBC) Field
The TBC field is used by the string indexed integer storage access instructions (lswx and stswx) as a byte
count. The TBC field is updated by the dlmzb[.] instruction with a value indicating the number of bytes up to
and including the zero byte detected by the instruction (see the instruction description for dlmzb in Instruction
Set on page 249 for more details). The TBC field is also written by mtspr(XER) with the value in (RS)
.
25:31
XER[TBC] is read (along with the rest of the XER) into a GPR by mfspr(XER).
2.7 Processor Control
The PPC440x5 core provides several registers for general processor control and status. These include:
• Machine State Register (MSR)
Controls interrupts and other processor functions
• Special Purpose Registers General (SPRGs)
SPRs for general purpose software use
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
• Processor Version Register (PVR)
Indicates the specific implementation of a processor
• Processor Identification Register (PIR)
Indicates the specific instance of a processor in a multi-processor system
• Core Configuration Register 0 (CCR0)
Controls specific processor functions, such as instruction prefetch
• Reset Configuration (RSTCFG)
Reports the values of certain fields of the TLB as supplied at reset
Except for the MSR, each of these registers is described in more detail in the following sections. The MSR is
2.7.1 Special Purpose Registers General (USPRG0, SPRG0–SPRG7)
USPRG0 and SPRG0–SPRG7 are provided for general purpose, system-dependent software use. One
common system usage of these registers is as temporary storage locations. For example, a routine might
save the contents of a GPR to an SPRG, and later restore the GPR from it. This is faster than a save/restore
to a memory location. These registers are written using mtspr and read using mfspr.
Access to USPRG0 is non-privileged for both read and write.
Access to SPRG4–SPRG7 is non-privileged for read but privileged for write, and hence different SPR
numbers are used for reading than for writing.
Access to SPRG0–SPRG3 is privileged for both read and write.
0
31
Figure 2-8. Special Purpose Registers General (USPRG0, SPRG0–SPRG7)
0:31
General data
Software value; no hardware usage.
2.7.2 Processor Version Register (PVR)
The PVR is a read-only register typically used to identify a specific processor core and chip implementation.
Software can read the PVR to determine processor core and chip hardware features. The PVR can be read
into a GPR using mfspr.
Refer to PowerPC 440x5 Embedded Processor Data Sheet for the PVR value.
Access to the PVR is privileged.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
OWN
0
11 12
31
PVN
Figure 2-9. Processor Version Register (PVR)
0:11
OWN
Owner Identifier
Identifies the owner of a core.
Implementation-specific value identifying the spe-
cific version and use of a processor core within a
chip.
12:31
PVN
Processor Version Number
2.7.3 Processor Identification Register (PIR)
The PIR is a read-only register that uniquely identifies a specific instance of a processor core, within a multi-
processor configuration, enabling software to determine exactly which processor it is running on. This capa-
bility is important for operating system software within multiprocessor configurations. The PIR can be read
into a GPR using mfspr.
Because the PPC440x5 is a uniprocessor, PIR[PIN] = 0b0000.
Access to the PIR is privileged.
PIN
0
27 28
31
Figure 2-10. Processor Identification Register (PIR)
0:27
28:31
Reserved
PIN
Processor Identification Number (PIN)
2.7.4 Core Configuration Register 0 (CCR0)
The CCR0 controls a number of special chip functions, including data cache and auxiliary processor opera-
tion, speculative instruction fetching, trace, and the operation of the cache block touch instructions. The
CCR0 is written from a GPR using mtspr, and can be read into a GPR using mfspr. Figure 2-11 on page 77
illustrates the fields of the CCR0, and gives a brief description of their functions. A cross reference after the
bit-field description indicates the section of this document which describes each field in more detail.
Access to the CCR0 is privileged.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
PRE
DSTG
DTB GDCBT
ICSLC
0
1
2
3
4
5
9
10 11 12
DAPUIB
15 16 17 18 19
GICBT
22 23 24
FLSTA
27 28 29 30 31
ICSLT
CRPE
Figure 2-11. Core Configuration Register 0 (CCR0)
0
Reserved
Parity Recoverability Enable
0 Semi-recoverable parity mode enabled
for data cache
1 Fully recoverable parity mode enabled
for data cache
Must be set to 1 to guarantee full recoverability
from MMU and data cache parity errors.
1
PRE
2:3
4
Reserved
Cache Read Parity Enable
0 Disable parity information reads
1 Enable parity information reads
When enabled, execution of icread, dcread, or
tlbre loads parity information into the ICDBTRH,
DCDBTRL, or target GPR, respectively.
CRPE
5:9
Reserved
Disable Store Gathering
0 Enabled; stores to contiguous addresses
may be gathered into a single transfer
1 Disabled; all stores to memory will be
performed independently
10
DSTG
See Store Gathering on page 119.
Disable APU Instruction Broadcast
0 Enabled.
1 Disabled; instructions not broadcast to
APU for decoding
This mechanism is provided as a means of reduc-
ing power consumption when an auxilliary pro-
cessor is not attached and/or is not being used.
11
DAPUIB
See Initialization on page 85.
12:15
16
Reserved
Disable Trace Broadcast
0 Enabled.
1 Disabled; no trace information is
broadcast.
This mechanism is provided as a means of reduc-
ing power consumption when instruction tracing is
not needed.
DTB
See Initialization on page 85.
Guaranteed Instruction Cache Block Touch
0 icbt may be abandoned without having
filled cache line if instruction pipeline
stalls.
17
GICBT
See icbt Operation on page 111.
1 icbt is guaranteed to fill cache line even
if instruction pipeline stalls.
Guaranteed Data Cache Block Touch
0 dcbt/dcbtst may be abandoned without
having filled cache line if load/store
pipeline stalls.
See Data Cache Control and Debug on
page 125.
18
GDCBT
1 dcbt/dcbtst are guaranteed to fill cache
line even if load/store pipeline stalls.
19:22
Reserved
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Force Load/Store Alignment
0 No Alignment exception on integer
storage access instructions, regardless
of alignment
1 An alignment exception occurs on
integer storage access instructions if
data address is not on an operand
boundary.
23
FLSTA
See Load and Store Alignment on page 117.
24:27
28:29
Reserved
Number of additional lines (0–3) to fill on instruc-
tion fetch miss.
ICSLC
ICSLT
Instruction Cache Speculative Line Count
See Speculative Prefetch Mechanism on
page 105.
Number of doublewords that must have already
been filled in order that the current speculative
line fill is not abandoned on a redirection of the
instruction stream.
30:31
Instruction Cache Speculative Line Threshold
See Speculative Prefetch Mechanism on
page 105.
2.7.5 Core Configuration Register 1 (CCR1)
Bits 0:19 of CCR1 can cause all possible parity error exceptions to verify correct machine check exception
handler operation. Other CCR1 bits can force a full-line data cache flush, or select a CPU timer clock input
other than CPUClock. The CCR1 is written from a GPR using mtspr, and can be read into a GPR using
mfspr. Figure 2-12 illustrates the fields of the CCR1, and gives a brief description of their functions.
Access to the CCR1 is privileged.
DCTPEI
FCOM
DCUPEI
ICDPEI
FFF
0
7
8
9
10 11 12 13 14 15 16
19 20 21
23 24 25
TCS
31
ICTPEI DCDPEI DCMPEI
MMUPEI
Figure 2-12. Core Configuration Register 1 (CCR1)
Instruction Cache Data Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bits recorded when the
instruction cache is filled. Each of the 8 bits corre-
sponds to one of the instruction words in the line.
0:7
ICDPEI
ICTPEI
DCTPEI
DCDPEI
Instruction Cache Tag Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bits recorded for the tag
field in the instruction cache.
8:9
Data Cache Tag Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bits recorded for the tag
field in the data cache.
10:11
12
Controls inversion of parity bits recorded for the
data field in the data cache.
Data Cache Data Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Controls inversion of parity bit recorded for the U
fields in the data cache.
Data Cache U-bit Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
13
14
DCUPEI
Data Cache Modified-bit Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bits recorded for the
modified (dirty) field in the data cache.
DCMPEI
FCOM
Force icbt , dcbt, dcbtst, dcbst, dcbf, dcbi, and
dcbz to appear to miss the caches. The intended
use is with icbt and dcbt only, which will fill a dupli-
cate line and allow testing of multi-hit parity errors.
See Section 4.2.4.7 Simulating Instruction Cache
Parity Errors for Software Testing on page 114 and
Figure 4.3.3.7 on page 130.
Force Cache Operation Miss
0 normal operation
1 cache ops appear to miss the cache
15
Controls inversion of parity bits recorded for the tag
field in the MMU.
Memory Management Unit Parity Error Insert
0 record even parity (normal)
16:19
MMUPEI
1 record odd parity (simulate parity error)
When flushing 32-byte (8-word) lines from the data
cache, normal operation is to write nothing, a dou-
ble word, quad word, or the entire 8-word block to
the memory as required by the dirty bits. This bit
ensures that none or all dirty bits are set so that
either nothing or the entire 8-word block is written
to memory when flushing a line from the data
cache. Refer to Section 4.3.1.4 Line Flush Opera-
tions on page 121.
Force Full-line Flush
0 flush only as much data as necessary.
1 always flush entire cache lines
20
FFF
21:23
24
Reserved
Timer Clock Select
0 CPU timer advances by one at each rising edge
of the CPU input clock (CPMC440CLOCK).
1 CPU timer advances by one for each rising edge
of the CPU timer clock
When TCS = 1, CPU timer clock input can toggle
at up to half of the CPU clock frequency.
TCS
(CPMC440TIMERCLOCK).
25:31
Reserved
2.7.6 Reset Configuration (RSTCFG)
The read-only RSTCFG register reports the values of certain fields of TLB as supplied at reset.
Access to RSTCFG is privileged.
U0
15 16 17 18 19 20
U1 U3
U2
E
0
23 24 25
27 28
31
ERPN
Figure 2-13. Reset Configuration
0:15
Reserved
U0 Storage Attribute
0 U0 storage attribute is disabled
16
U0
See Table 5-1 on page 135.
1 U0 storage attribute is enabled
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
U1 Storage Attribute
0 U1 storage attribute is disabled
1 U1 storage attribute is enabled
17
18
U1
U2
U3
See Table 5-1 on page 135.
See Table 5-1 on page 135.
See Table 5-1 on page 135.
U2 Storage Attribute
0 U2 storage attribute is disabled
1 U2 storage attribute is enabled
U3 Storage Attribute
0 U3 storage attribute is disabled
19
1 U3 storage attribute is enabled
20:23
24
Reserved
E Storage Attribute
0 Accesses to the page are big endian.
E
1 Accesses to the page are little endian.
25:27
Reserved
This TLB field is prepended to the translated
address to form a 36-bit real address. See Table
5.4 Address Translation on page 140 and Table
5-3 Page Size and Real Address Formation on
page 142.
28:31
ERPN
Extended Real Page Number
2.8 User and Supervisor Modes
PowerPC Book-E architecture defines two operating “states” or “modes,” supervisor (privileged), and user
(non-privileged). Which mode the processor is operating in is controlled by MSR[PR]. When MSR[PR] is 0,
the processor is in supervisor mode, and can execute all instructions and access all registers, including privi-
leged ones. When MSR[PR] is 1, the processor is in user mode, and can only execute non-privileged instruc-
tions and access non-privileged registers. An attempt to execute a privileged instruction or to access a
privileged register while in user mode causes a Privileged Instruction exception type Program interrupt to
occur.
Note that the name “PR” for the MSR field refers to an historical alternative name for user mode, which is
“problem state.” Hence the value 1 in the field indicates “problem state,” and not “privileged” as one might
expect.
2.8.1 Privileged Instructions
The following instructions are privileged and cannot be executed in user mode:
Table 2-27. Privileged Instructions
dcbi
dccci
dcread
iccci
icread
mfdcr
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 2-27. Privileged Instructions (continued)
mfmsr
mfspr
mtdcr
mtmsr
mtspr
rfci
5
5
rfi
rfmci
tlbre
tlbsx
tlbsync
tlbwe
wrtee
wrteei
2.8.2 Privileged SPRs
Most SPRs are privileged. The only defined non-privileged SPRs are the LR, CTR, XER, USPRG0, SPRG4–
7 (read access only), TBU (read access only), and TBL (read access only). The PPC440x5 core also treats
all SPR numbers with a 1 in bit 5 of the SPRN field as privileged, whether the particular SPR number is
defined or not. Thus the core causes a Privileged Instruction exception type Program interrupt on any attempt
to access such an SPR number while in user mode. In addition, the core causes an Illegal Instruction excep-
tion type Program interrupt on any attempt to access while in user mode an undefined SPR number with a 0
in SPRN . On the other hand, the result of attempting to access an undefined SPR number in supervisor
5
mode is undefined, regardless of the value in SPRN .
5
2.9 Speculative Accesses
The PowerPC Book-E Architecture permits implementations to perform speculative accesses to memory,
either for instruction fetching, or for data loads. A speculative access is defined as any access that is not
required by the sequential execution model (SEM).
For example, the PPC440x5 speculatively prefetches instructions down the predicted path of a conditional
branch; if the branch is later determined to not go in the predicted direction, the fetching of the instructions
from the predicted path is not required by the SEM and thus is speculative. Similarly, the PPC440x5 executes
load instructions out-of-order, and may read data from memory for a load instruction that is past an undeter-
mined branch.
Sometimes speculative accesses are inappropriate, however. For example, attempting to access data at
addresses to which I/O devices are mapped can cause problems. If the I/O device is a serial port, reading it
speculatively could cause data to be lost.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
The architecture provides two mechanisms for protecting against errant accesses to such “non-well-behaved”
memory addresses. The first is the guarded (G) storage attribute, and protects against speculative data
accesses. The second is the execute permission mechanism, and protects against speculative instruction
2.10 Synchronization
The PPC440x5 supports the synchronization operations of the PowerPC Book-E architecture. There are
three kinds of synchronization defined by the architecture, each of which is described in the following
sections.
2.10.1 Context Synchronization
The context of a program is the environment in which the program executes. For example, the mode (user or
supervisor) is part of the context, as are the address translation space and storage attributes of the memory
pages being accessed by the program. Context is controlled by the contents of certain registers and other
resources, such as the MSR and the translation lookaside buffer (TLB).
Under certain circumstances, it is necessary for the hardware or software to force the synchronization of a
program’s context. Context synchronizing operations include all interrupts except Machine Check, as well as
the isync, sc, rfi, rfci, and rfmci instructions. Context synchronizing operations satisfy the following require-
ments:
1. The operation is not initiated until all instructions preceding the operation have completed to the point at
which they have reported any and all exceptions that they will cause.
2. All instructions preceding the operation must complete in the context in which they were initiated. That is,
they must not be affected by any context changes caused by the context synchronizing operation, or any
instructions after the context synchronizing operation.
3. If the operation is the sc instruction (which causes a System Call interrupt) or is itself an interrupt, then
the operation is not initiated until no higher priority interrupt is pending (see Interrupts and Exceptions on
4. All instructions that follow the operation must be re-fetched and executed in the context that is established
by the completion of the context synchronizing operation and all of the instructions which preceded it.
Note that context synchronizing operations do not force the completion of storage accesses, nor do they
enforce any ordering amongst accesses before and/or after the context synchronizing operation. If such
behavior is required, then a storage synchronizing instruction must be used (see Storage Ordering and
Also note that architecturally Machine Check interrupts are not context synchronizing. Therefore, an instruc-
tion that precedes a context synchronizing operation can cause a Machine Check interrupt after the context
synchronizing operation occurs and additional instructions have completed. For the PPC440x5 core, this can
only occur with Data Machine Check exceptions, and not Instruction Machine Check exceptions.
The following scenarios use pseudocode examples to illustrate the effects of context synchronization. Subse-
quent text explains how software can further guarantee “storage ordering.”
1. Consider the following self-modifying code instruction sequence:
stw XYZ Store to caching inhibited address XYZ
isync
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
XYZ
fetch and execute the instruction at address XYZ
In this sequence, the isync instruction does not guarantee that the XYZ instruction is fetched after the
store has occurred to memory. There is no guarantee which XYZ instruction will execute; either the old
version or the new (stored) version might.
2. Now consider the required self-modifying code sequence:
stw
Write new instruction to data cache
dcbst
msync
icbi
Push the new instruction from the data cache to memory
Guarantee that dcbst completes before subsequent instructions begin
invalidate old copy of instruction in instruction cache
msync
isync
Guarantee that icbi completes before subsequent instructions begin
force context synchronization, discard ed instructions and re-fetch, fetch of
OWN
PVN
System-dependent PVR[OWN] value (after reset and otherwise) is specified by core input signals
System-dependent PVR[PVN] value (after reset and otherwise) is specified by core input signals
PVR
stored instruction guaranteed to get new value
3. This final example illustrates the use of isync with context changes to the debug facilities
mtdbcr0 Enable the instruction address compare (IAC) debug event
isync
XYZ
Wait for the new Debug Control Register 0 (DBCR0) context to be established
This instruction is at the IAC address; an isync is necessary to guarantee that the
IAC event is recognized on the execution of this instruction; without the isync, the
XYZ instruction may be prefetched and dispatched to execution before
recognizing that the IAC event has been enabled.
2.10.2 Execution Synchronization
Execution synchronization is a subset of context synchronization. An execution synchronizing operation
satisfies the first two requirements of context synchronizing operations, but not the latter two. That is, execu-
tion synchronizing operations guarantee that preceding instructions execute in the “old” context, but do not
guarantee that subsequent instructions operate in the “new” context. An example of a scenario requiring
execution synchronization would be just before the execution of a TLB-updating instructions (such as tlbwe).
An execution synchronizing instruction should be executed to guarantee that all preceding storage access
instructions have performed their address translations before executing tlbwe to invalidate an entry which
might be used by those preceding instructions.
There are four execution synchronizing instructions: mtmsr, wrtee, wrteei, and msync. Of course, all
context synchronizing instruction are also implicitly execution synchronizing, since context synchronization is
a superset of execution synchronization.
Note that PowerPC Book-E imposes additional requirements on updates to MSR[EE] (the external interrupt
enable bit). Specifically, if a mtmsr, wrtee, or wrteei instruction sets MSR[EE] = 1, and an External Input,
Decrementer, or Fixed Interval Timer exception is pending, the interrupt must be taken before the instruction
that follows the MSR[EE]-updating is executed. In this sense, these MSR[EE]-updating instructions can be
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
thought of as being context synchronizing with respect to the MSR[EE] bit, in that it guarantees that subse-
quent instructions execute (or are prevented from executing and an interrupt taken) according to the new
context of MSR[EE].
2.10.3 Storage Ordering and Synchronization
Storage synchronization enforces ordering between storage access instructions executed by the PPC440x5
core. There are two storage synchronizing instructions: msync and mbar. PowerPC Book-E architecture
defines different ordering requirements for these two instructions, but the PPC440x5 core implements them in
an identical fashion. Architecturally, msync is the “stronger” of the two, and is also execution synchronizing,
whereas mbar is not.
mbar acts as a “barrier” between all storage access instructions executed before the mbar and all those
executed after the mbar. That is, mbar ensures that all of the storage accesses initiated by instructions
before the mbar are performed with respect to the memory subsystem before any of the accesses initiated by
instructions after the mbar. However, mbar does not prevent subsequent instructions from executing (nor
even from completing) before the completion of the storage accesses initiated by instructions before the
mbar.
msync, on the other hand, does guarantee that all preceding storage accesses have actually been
performed with respect to the memory subsystem before the execution of any instruction after the msync.
Note that this requirement goes beyond the requirements of mere execution synchronization, in that execu-
tion synchronization doesn’t require the completion of preceding storage accesses.
The following two examples illustrate the distinctive use of mbar vs. msync.
stw
Store data to an I/O device
Wait for store to actually complete
Reconfigure the I/O device
msync
mtdcr
In this example, the mtdcr is reconfiguring the I/O device in a manner which would cause the preceding store
instruction to fail, were the mtdcr to change the device before the completion of the store. Since mtdcr is not
a storage access instruction, the use of mbar instead of msync would not guarantee that the store is
performed before letting the mtdcr reconfigure the device. It only guarantees that subsequent storage
accesses are not performed to memory or any device before the earlier store.
Now consider this next example:
stb X
mbar
Store data to an I/O device at address X, causing a status bit at address Y to be reset
Guarantee preceding store is performed to the device before any subequent
storage accesses are performed
lbz Y
Load status from the I/O device at address Y
Here, mbar is appropriate instead of msync, because all that is required is that the store to the I/O device
happens before the load does, but not that other instructions subsequent to the mbar won’t get executed
before the store.
prgmodel.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
3. Initialization
This chapter describes the initial state of the PPC440x5 core after a hardware reset, and contains a descrip-
tion of the initialization software required to complete initialization so that the PPC440x5 core can begin
executing application code. Initialization of other on-chip and/or off-chip system components may also be
needed, in addition to the processor core initialization described in this chapter.
3.1 PPC440x5 Core State After Reset
In general, the contents of registers and other facilities within the PPC440x5 core are undefined after a hard-
ware reset. Reset is defined to initialize only the minimal resources required such that instructions can be
fetched and executed from the initial program memory page, and so that repeatable, deterministic behavior
can be guaranteed provided that the proper software initialization sequence is followed. System software
must fully configure the rest of the PPC440x5 core resources, as well as the other facilities within the chip
and/or system.
The following list summarizes the requirements of the Book-E Enhanced PowerPC Architecture with regards
to the processor state after reset, prior to any additional initialization by software.
• All fields of the MSR are set to 0, disabling all asynchronous interrupts, placing the processor in supervi-
sor mode, and specifying that instruction and data accesses are to the system (as opposed to applica-
tion) address space.
• DBCR0[RST] is set to 0, thereby ending any previous software-initiated reset operation.
• DBSR[MRR] records the type of the just ended reset operation (core, chip, or system; see Reset Types
• TCR[WRC] is set to 0, thereby disabling the Watchdog timer reset operation.
• TSR[WRS] records the type of the just ended reset operation, if the reset was initiated by the Watchdog
Timer (otherwise this field is unchanged from its pre-reset value).
• The PVR is defined, after reset and otherwise, to contain a value that indicates the specific processor
implementation.
• The program counter (PC) is set to 0xFFFFFFFC, the effective address (EA) of the last word of the
address space.
The memory management resources are set to values such that the processor is able to successfully fetch
and execute instructions and read (but not write) data within the 4KB program memory page located at the
end of the 32-bit effective address space. Exactly how this is accomplished is implementation-dependent. For
example, it may or may not be the case that a TLB entry is established in a manner which is visible to soft-
ware using the TLB management instructions. Regardless of how the implementation enables access to the
initial program memory page, instruction execution starts at the effective adddress of 0xFFFFFFFC, the last
word of the effective address space. The instruction at this address must be an unconditional branch back-
wards to the start of the initialization sequence, which must lie somewhere within the initial 4KB program
memory page. The real address to which the initial effective address will be translated is also implementation-
or system-dependent, as are the various storage attributes of the initial program memory page such as the
caching inhibited and endian attributes.
Note: In the PPC440x5 core, a single entry is established in the instruction shadow TLB (ITLB) and data
ware insert an entry into the UTLB to cover this same memory region before performing any context synchro-
init.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
nizing operation (including causing any exceptions which would lead to an interrupt), since a context
synchronizing operation will invalidate the shadow TLB entries.
Initialization software should consider all other resources within the PPC440x5 core to be undefined after
reset, in order for the initialization sequence to be compatible with other PowerPC implementations. There
are, however, additional core resources which are initialized by reset, in order to guarantee correct and deter-
ministic operation of the processor during the initialization sequence. Table 3-1 shows the reset state of all
PPC440x5 core resources which are defined to be initialized by reset. While certain other register fields and
other facilities within the PPC440x5 core may be affected by reset, this is not an architectural nor hardware
requirement, and software must treat those resources as undefined. Likewise, even those resources which
are included in Table 3-1 but which are not identified in the previous list as being architecturally required,
should be treated as undefined by the initialization software.
During chip initialization, some chip control registers must be initialized to ensure proper chip operation.
Peripheral devices can also be initialized as appropriate for the system design.
Table 3-1. Reset Values of Registers and Other PPC440x5 Facilities
Resource
Field
DAPUIB
DTB
Reset Value
Comment
0
0
Enable broadcast of instruction data to auxiliary processor interface
Enable broadcast of trace information
CCR0
ICDPEI
ICTPEI
DCTPEI
DCDPEI
DCUPEI
DCMPEI
FCOM
MMUPEI
FFF
0
0
0
Disable Parity Error Insertion (enabled only for s/w testing)
0
CCR1
0
0
0
Do not force cache ops to miss.
0
Disable Parity Error Insertion (enabled only for s/w testing)
Flush only as much data from dirty lines as needed.
External Debug mode disabled
0
EDM
0
RST
0b00
0
Software-initiated debug reset disabled
Instruction completion debug events disabled
Branch taken debug events disabled
Instruction Address Compare 1 (IAC1) debug events disabled
IAC2 debug events disabled
ICMP
BRT
0
DBCR0
IAC1
0
IAC2
0
IAC3
0
IAC3 debug events disabled
IAC4
0
IAC4 debug events disabled
init.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 3-1. Reset Values of Registers and Other PPC440x5 Facilities
Resource
Field
UDE
Reset Value
0
Comment
Unconditional debug event has not occurred
Indicates most recent type of reset as follows:
00 No reset has occurred since this field last cleared by software
01 Core reset
10 Chip reset
MRR
Reset-dependent
11 System reset
ICMP
BRT
IRPT
TRAP
IAC1
IAC2
IAC3
IAC4
DAC1R
DAC1W
DAC2R
DAC2W
RET
MCI
0
Instruction completion debug event has not occurred
Branch taken debug event has not occurred
Interrupt debug event has not occurred
0
0
0
Trap debug event has not occurred
DBSR
0
IAC1 debug event has not occurred
0
IAC2 debug event has not occurred
0
IAC3 debug event has not occurred
0
IAC4 debug event has not occurred
0
Data address compare 1 (DAC1) read debug event has not occurred
DAC1 write debug event has not occurred
DAC2 read debug event has not occurred
DAC2 write debug event has not occurred
Return debug event has not occurred
0
0
0
0
ESR
0
Synchronous Instruction Machine Check exception has not occurred
Asynchronous Instruction Machine Check exception has not occurred
Wait state disabled
MCSR
MCS
WE
0
0
CE
0
Asynchronous critical interrupts disabled
Asynchronous non-critical interrupts disabled
Processor in supervisor mode
EE
0
PR
0
FP
0
Floating-point Unavailable interrupts disabledStorage
Machine Check interrupts disabled
ME
0
MSR
FE0
0
Floating-point Enabled interrupts disabled
Debug Wait mode disabled
DWE
DE
0
0
Debug interrupts disabled
FE1
0
Floating-point Enabled interrupts disabled
Instruction fetch access is to system-level virtual address space
Data access is to system level virtual address space
Initial reset instruction fetched from last word of effective addess space
IS
0
DS
0
PC
0xFFFFFFFC
OWN
PVN
System-dependent PVR[OWN] value (after reset and otherwise) is specified by core input signals
System-dependent PVR[PVN] value (after reset and otherwise) is specified by core input signals
PVR
init.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table 3-1. Reset Values of Registers and Other PPC440x5 Facilities
Resource
RSTCFG
TCR
Field
Reset Value
System-dependent
System-dependent
System-dependent
System-dependent
System-dependent
System-dependent
0b00
Comment
All RSTCFG fields are specified by core input signals
Watchdog Timer reset disabled
U0
U1
U2
U3
E
EPRN
WRC
Match EA of initial reset instruction (EPN
compared to the EA because the page size is 4KB).
are undefined, as they are not
20:21
EPN
0xFFFFF
0:19
V
1
0
Translation table entry for the initial program memory page is valid.
TS
Initial program memory page is in system-level virtual address space.
Initial program memory page size is 4KB.
SIZE
TID
0b0001
Initial program memory page is globally shared; no match required against PID
register.
0x00
RPN
0xFFFFF || 0b00
System-dependent
Initial program memory page mapped effective=real.
0:21
Extended real page number of the initial program memory page is specified by
core input signals.
ERPN
Reset value of user-definable storage attributes are specified by core input sig-
nals
U0–U3
System-dependent
W
I
0
1
0
1
Write-through storage attribute disabled.
Caching inhibited storage attribute enabled.
Memory coherent storage attribute disabled.
Guarded storage attribute enabled.
1
TLBentry
M
G
E
System-dependent Reset value of endian storage attribute is specified by a core input signal.
SX
SW
SR
1
0
1
Supervisor mode execution access enabled.
Supervisor mode write access disabled.
Supervisor mode read access enabled.
Copy of TCR[WRC] If reset caused by Watchdog Timer
TSR
WRS
Unchanged
Undefined
If reset not caused by Watchdog Timer
After power-up
Note 1: “TLBentry” refers to an entry in the shadow instruction and data TLB arrays that is automatically
configured by the PPC440x5 core to enable fetching and reading (but not writing) from the initial
program memory page. This entry is not architecturally visible to software, and is invalidated upon any
context synchronizing operation. Software must initialize a corresponding entry in the main unified
TLB array before executing any operation which could lead to a context synchronization. See
Initialization Software Requirements on page 89 for more information.
init.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
3.2 Reset Types
The PPC440x5 core supports three types of reset: core, chip, and system. The type of reset is indicated by a
set of core input signals. For each type of reset, the core resources are initialized as indicated in Table 3-1 on
chip logic. The chip reset operation is intended to reset the entire chip, but off-chip hardware in the system is
not informed of the reset operation. System reset is intended to reset the entire chip, and also to signal the
rest of the off-chip system that the chip is being reset.
3.3 Reset Sources
A reset operation can be initiated on the PPC440x5 core through the use of any of four separate mecha-
nisms. The first is a set of three input signals to the core, one for each of the three reset types. These signals
can be asserted asynchronously by hardware outside the core to initiate a reset operation. The second reset
source is the TCR[WRC] field, which can be setup by software to initiate a reset operation upon certain
Watchdog Timer expiration events. The third reset source is the DBCR0[RST] field, which can be written by
software to immediately initiate a reset operation. The fourth reset source is the JTAG interface, which can be
used by a JTAG-attached debug tool to initiate a reset operation asynchronously to program execution on the
PPC440x5 core.
3.4 Initialization Software Requirements
After a reset operation occurs, the PPC440x5 core is initialized to a minimum configuration to enable the
fetching and execution of the software initialization code, and to guarantee deterministic behavior of the core
during the execution of this code. Initialization software is necessary to complete the configuration of the
processor core and the rest of the on-chip and off-chip system.
The system must provide non-volatile memory (or memory initialized by some mechanism other than the
PPC440x5 core) at the real address corresponding to effective address 0xFFFFFFFC, and at the rest of the
initial program memory page. The instruction at the initial address must be an unconditional branch back-
wards to the beginning of the initialization software sequence.
The initialization software functions described in this section perform the configuration tasks required to
prepare the PPC440x5 core to boot an operating system and subsequently execute an application program.
The initialization software must also perform functions associated with hardware resources that are outside
the PPC440x5 core, and hence that are beyond the scope of this manual. This section makes reference to
some of these functions, but their full scope is described in the user’s manual for the specific chip and/or
system implementation.
Initialization software should perform the following tasks in order to fully configure the PPC440x5 core. For
more information on the various functions referenced in the initialization sequence, see the corresponding
chapters of this document.
1. Branch backwards from effective address 0xFFFFFFFC to the start of the initialization sequence
init.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
2. Invalidate the instruction cache (iccci)
3. Invalidate the data cache (dccci)
4. Synchronize memory accesses (msync)
This step forces any data PLB operations that may have been in progress prior to the reset operation to
complete, thereby allowing subsequent data accesses to be initiated and completed properly.
5. Clear DBCR0 register (disable all debug events)
Although the PPC440x5 core is defined to reset some of the debug event enables during the reset oper-
ation (as specified in Table 3-1 on page 86), this is not required by the architecture and hence the initial-
ization software should not assume this behavior. Software should disable all debug events in order to
prevent non-deterministic behavior on the trace interface to the core.
6. Clear DBSR register (initialize all debug event status)
Although the PPC440x5 core is defined to reset the DBSR debug event status bits during the reset oper-
ation (as specified in Table 3-1 on page 86), this is not required by the architecture and hence the initial-
ization software should not assume this behavior. Software should clear all such status in order to
prevent non-deterministic behavior on the JTAG interface to the core.
7. Initialize CCR0 register
1. Enable/disable broadcast of instructions to auxiliary processor (save power if no AP attached)
2. Enable/disable broadcast of trace information (save power if not tracing)
3. Enable/configure or disable speculative instruction cache line prefetching
4. Specify behavior for icbt and dcbt/dcbtst instructions
5. Enable/disable gathering of separate store accesses
6. Enable/disable hardware support for misaligned data accesses
7. Enable/disable parity error recoverability (recoverability lowers load/store performance marginally.)
8. Enable/disable cache read of parity bits depending on s/w compatibility requirements
8. Initialize CCR1 register
1. enable/disable full-line flushes as desired.
2. disable force cache-op miss (FCOM) and various parity error insertion (xxxPEI).
3. Users may wish to initialize CCR1[TCS] here, or in the timer facilities section.
9. Configure instruction and data cache regions
These steps must be performed prior to enabling the caches by setting the caching inhibited storage
attribute of the corresponding TLB entry to 0.
1. Clear the instruction and data cache normal victim index registers (INV0–INV3, DNV0–DNV3)
2. Clear the instruction and data cache transient victim index registers (ITV0–ITV3, DTV0–DTV3)
3. Set the instruction and data cache victim limit registers (IVLIM and DVLIM) according to the desired
size of the normal, locked, and transient regions of each cache
10. Setup TLB entry to cover initial program memory page
Since the PPC440x5 core only initializes an architecturally-invisible shadow TLB entry during the reset
operation, and since all shadow TLB entries are invalidated upon any context synchronization, special
init.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
care must be taken during the initialization sequence to prevent any such context synchronizing opera-
tions (such as interrupts and the isync instruction) until after this step is completed, and an architected
TLB entry has been established in the TLB. Particular care should be taken to avoid store operations,
since write permission is disabled upon reset, and an attempt to execute any store operation would result
in a Data Storage interrupt, thereby invalidating the shadow TLB entry.
1. Initialize MMUCR
- Specify TID field to be written to TLB entries
- Specify TS field to be used for TLB searches
- Specify store miss allocation behavior
- Enable/disable transient cache mechanism
- Enable/disable cache locking exceptions
2. Write TLB entry for initial program memory page
- Specify EPN, RPN, ERPN, and SIZE as appropriate for system
- Set valid bit
- Specify TID = 0 (disable comparison to PID) or else initialize PID register to matching value
- Specify TS = 0 (system address space) or else MSR[IS,DS] must be set to correspond to TS=1
- Specify storage attributes (W, I, M, G, E, U0–U3) as appropriate for system
- Enable supervisor mode fetch, read, and write access (SX, SR, SW)
3. Initialize PID register to match TID field of TLB entry (unless using TID = 0)
4. Setup for subsequent MSR[IS,DS] initialization to correspond to TS field of TLB entry
Only necessary if TS field of TLB entry being set to 1 (MSR[IS,DS] already reset to 0)
- Write new MSR value into SRR1
- Write address from which to continue execution into SRR0
5. Setup for subsequent change in instruction fetch address
Only necessary if EPN field of TLB entry changed from the initial value (EPN
- Write initial/new MSR value into SRR1
≠ 0xFFFFF)
0:19
- Write address from which to continue execution into SRR0
6. Initialize or invalidate all other TLB entries as desired
7. Context synchronize to invalidate shadow TLB contents and cause new TLB contents to take effect
- Use isync if not changing MSR contents and not changing the effective address of the rest of
the initialization sequence
- Use rfi if changing MSR to match new TS field of TLB entry (SRR1 will be copied into MSR, and
program execution will resume at value in SRR0)
- Use rfi if changing next instruction fetch address to correspond to new EPN field of TLB entry
(SRR1 will be copied into MSR, and program execution will resume at value in SRR0)
Instruction and data caches will now begin to be used, if the corresponding TLB entry has been setup
with the caching inhibited storage attribute set to 0. Initialization software can now branch outside of
the initial 4KB memory region as controlled by the address and size of the new TLB entry and/or any
other TLB entries which have been setup.
init.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
11. Initialize interrupt resources
1. Initialize IVPR to specify high-order address of the interrupt handling routines
Make sure that the corresponding address region is covered by a TLB entry (or entries)
2. Initialize IVOR0–IVOR15 registers (individual interrupt vector addresses)
Make sure that the corresponding addresses are covered by a TLB entry (or entries)
Because the low order four bits of IVOR0–IVOR15 are reserved, the values written to those bits are
ignored when the registers are written, and are read as zero when the registers are used. Therefore,
all interrupt vector offsets are implicitly aligned on quadword boundaries. Software must take care to
assure that all interrupt handlers are quadword-aligned.
3. Setup corresponding memory contents with the interrupt handling routines
more information on the instruction sequence necessary to synchronize changes to program memory
prior to executing the new instructions.)
12. Configure debug facilities as desired
1. Write DBCR1 and DBCR2 to specify IAC and DAC event conditions
2. Clear DBSR to initialize IAC auto-toggle status
3. Initialize IAC1–IAC4, DAC1–DAC2, DVC1–DVC2 registers to desired values
4. Write MSR[DWE] to enable Debug Wait mode (if desired)
5. Write DBCR0 to enable desired debug mode(s) and event(s)
6. Context synchronize to establish new debug facility context (isync)
13. Configure timer facilities as desired
1. Write DEC to 0 to prevent Decrementer exception after TSR is cleared
2. Write TBL to 0 to prevent Fixed Interval Timer and Watchdog Timer exceptions after TSR is cleared,
and to prevent increment into TBH prior to full initialization
3. CCR1[TCS] (Timer Clock Select) can be initialized here, or earlier with the rest of the CCR1.
4. Clear TSR to clear all timer exception status
5. Write TCR to configure and enable timers as desired
Software must take care with respect to the enabling of the Watchdog Timer reset function, as once
this function is enabled, it cannot be disabled except by reset itself
6. Initialize TBH value as desired
7. Initialize TBL value as desired
8. Initialize DECAR to desired value (if enabling the auto-reload function)
9. Initialize DEC to desired value
14. Initialize facilities outside the processor core which are possible sources of asynchronous interrupt
requests (including DCRs and/or other memory-mapped resources)
This must be done prior to enabling asynchronous interrupts in the MSR
15. Initialize the MSR to enable interrupts as desired
init.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
1. Set MSR[CE] to enable/disable Critical Input and Watchdog Timer interrupts
2. Set MSR[EE] to enable/disable External Input, Decrementer, and Fixed Interval Timer interrupts
3. Set MSR[DE] to enable/disable Debug interrupts
4. Set MSR[ME] to enable/disable Machine Check interrupts
Software should first check the status of the ESR[MCI] field and MCSR[MCS] field to determine
whether any Machine Check exceptions have occurred after these fields were cleared by reset and
before Machine Check interrupts were enabled (by this step). Any such exceptions would have set
ESR[MCI] or MCSR[MCS] to 1, and this status can only be cleared explicitly by software. After the
MCSR[MCS] field is known to be clear, the MCSR status bits (MCSR[1:8]) should be cleared by soft-
ware to avoid possible confusion upon later service of a machine check interrupt. Once MSR[ME]
has been set to 1, subsequent Machine Check exceptions will result in a Machine Check interrupt.
5. Context synchronize to establish new MSR context (isync)
16. Initialize any other processor core resources as required by the system (GPRs, SPRGs, and so on)
17. Initialize any other facilities outside the processor core as required by the system
18. Initialize system memory as required by the system software
information on the instruction sequence necessary to synchronize changes to program memory prior to
executing the new instructions)
19. Start the system software
System software is generally responsible for initializing and/or managing the rest of the MSR fields,
including:
1. MSR[FP] to enable or disable the execution of floating-point instructions
2. MSR[FE0,FE1] to enable/disable Floating-Point Enabled exception type Program interrupts
3. MSR[PR] to specify user mode or supervisor mode
4. MSR[IS,DS] to specify application address space or system address space for instructions and data
5. MSR[WE] to place the processor into Wait State (halt execution pending an interrupt)
init.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
init.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
4. Instruction and Data Caches
The PPC440x5 core provides separate instruction and data cache controllers and arrays, which allow concur-
rent access and minimize pipeline stalls. The storage capacity of the cache arrays, which can range from
8KB–32KB each, depends upon the implementation. Both cache controllers have 32-byte lines, and both are
highly associative, having 64-way set-associativity for 32KB and 16KB sizes, and 32-way set-associativity for
the 8KB size. The PowerPC instruction set provides a rich set of cache management instructions for soft-
ware-enforced coherency. The PPC440x5 implementation also provides special debug instructions that can
directly read the tag and data arrays. The cache controllers interface to the processor local bus (PLB) for
connection to the IBM CoreConnect system-on-a-chip environment.
Both the data and instruction caches are parity protected against soft errors. If such errors are detected, the
CPU will vector to the machine check interrupt handler, where software can take appropriate action. The
details of suggested interrupt handling are described below in section 4.2, “Instruction Cache Controller,” and
The rest of this chapter provides more detailed information about the operation of the instruction and data
cache controllers and arrays.
4.1 Cache Array Organization and Operation
The instruction and data cache arrays are organized identically, although the fields of the tag and data
portions of the arrays are slightly different because the functions of the arrays differ, and because the instruc-
tion cache is virtually tagged while the data cache has real tags.
The associativity of each cache varies according to its size: the 32KB and 16KB cache sizes are 64-way set-
associative, while the 8KB cache size is 32-way set-associative. Accordingly, the number of “sets” in each
cache varies according to its size: the 32KB cache has 16 sets, while the 16KB and 8KB caches have 8 sets.
Regardless of cache array size, the cache line size is always 32 bytes.
The organization of the cache into “ways” and “sets” is as follows. Using the 32KB cache as an example,
there are 64 ways in each set, with a set consisting of all 64 lines (one line from each way) at which a given
memory location can reside. Conversely, and again using the 32KB cache as an example, there are 16 sets
in each way, with a way consisting of 16 lines (one from each set).
Table 4-1 on page -96 illustrates generically the ways and sets of the cache arrays, for any cache size, while
Table 4-2 on page -96 provides specific values for the parameters used in Table 4-1, for the different cache
sizes. As shown in Table 4-2, the tag field for each line in each way holds the high-order address bits associ-
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
ated with the line that currently resides in that way. The middle-order address bits form an index to select a
specific set of the cache, while the five lowest-order address bits form a byte-offset to choose a specific byte
(or bytes, depending on the size of the operation) from the 32-byte cache line.
Table 4-1. Instruction and Data Cache Array Organization
Way 0
Line 0
Line 1
Way 1
Line n
Way w – 2
Line (w – 2)n
Way w – 1
Line (w – 1)n
• • •
• • •
• • •
Set 0
Set 1
Line n + 1
Line (w – 2)n + 1
Line (w – 1)n + 1
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Set n – 2
Line n – 2
Line 2n – 2
Line (w – 1)n – 2
Line wn – 2
• • •
• • •
Set n – 1
Line n – 1
Line 2n – 1
Line (w – 1)n – 1
Line wn – 1
Table 4-2. Cache Sizes and Parameters
Tag
Address Bits
Set
Address Bits
Byte Offset
Address Bits
Array Size
w (Ways)
n (Sets)
1
8KB
16KB
32KB
32
64
64
8
8
A
A
A
0:23
24:26
27:31
A
A
A
0:23
24:26
27:31
16
A
A
A
0:22
23:26
27:31
Note 1: The tag address bits shown in the table refer to the effective address bits,
and are for illustrative purposes only. Because the instruction cache is
tagged with the virtual address, and the data cache is tagged with the real
address, the actual tag address bits contained within each array are
different. See Figure 4-8 and Figure 4-9 on page 113 for instruction cache
tag information, and Figure 4-10 and Figure 4-11 on page 128 for data
cache tag information. Also, see “Instruction Cache Synonyms” on
page -107 for details on instruction cache synonyms associated with the use
of virtual tags for the instruction cache.
4.1.1 Cache Line Replacement Policy
Memory addresses are specified as being cacheable or caching inhibited on a page basis, using the caching
inhibited (I) storage attribute (see Caching Inhibited (I) on page 145). When a program references a cache-
able memory location and that location is not already in the cache (a cache miss), the line may be brought
into the cache (a cache line fill operation) and placed into any one of the ways within the set selected by the
middle portion of the address (the specific address bits that select the set are specified in Table 4-2). If the
particular way within the set already contains a valid line from some other address, the existing line is
removed and replaced by the newly referenced line from memory. The line being replaced is referred to as
the victim.
The way selected to be the victim for replacement is controlled by a field within a Special Purpose Register
(SPR). There is a separate “victim index field” for each set within the cache. The registers controlling the
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
VNDXA
VNDXC
0
7
8
15 16
23 24
31
VNDXB
VNDXD
Figure 4-1. Instruction Cache Normal Victim Registers (INV0–INV3) Instruction Cache Transient Victim
Registers (ITV0–ITV3) Data Cache Normal Victim Registers (DNV0–DNV3) Data Cache Transient
Victim Registers (DTV0–DTV3)
For all victim index fields, the number of bits used
to select the cache way for replacement depends
on the implemented cache size. See Table 4-3,” on
page -98
Victim Index A (for cache lines with EA[25:26] =
0b00)
0:7
VNDXA
for more information.
Victim Index B (for cache lines with EA[25:26] =
0b01)
8:15
VNDXB
VNDXC
VNDXD
Victim Index C (for cache lines with EA[25:26] =
0b10)
16:23
24:31
Victim Index D (for cache lines with EA[25:26] =
0b11)
Each of the 16 SPRs illustrated in Figure 4-1 can be written from a GPR using mtspr, and can be read into a
GPR using mfspr. In general, however, these registers are initialized by software once at startup, and then
are managed automatically by hardware after that. Specifically, every time a new cache line is placed into the
cache, the appropriate victim index field (as controlled by the type of access and the particular cache set
being updated) is first referenced to determine which way within that set should be replaced. Then, that same
field is incremented such that the ways within that set are replaced in a round-robin fashion as each new line
is brought into that set. When the victim index field value reaches the index of the last way (according to the
size of the cache and the type of access being performed), the value is wrapped back to the index of the first
way for that type of access. The first and last ways for the different types of accesses are controlled by fields
in a pair of victim limit SPRs, one for each cache (see Cache Locking and Transient Mechanism on page 99
for more information).
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
The size of the victim index fields varies according to the size of the respective cache. Also, which field is
used varies according to the type of access, the size of the cache, and the address of the cache line.
Table 4-3 describes the correlation between the victim index fields and different access types, cache sizes,
and addresses.
Table 4-3. Victim Index Field Selection
1,2
Victim Index Field
3
Address
23:26
8KB Cache
16KB Cache
32KB Cache
0
1
xxV0[VNDXA]
xxV0[VNDXA]
xxV0[VNDXA]
3:7
2:7
2:7
xxV0[VNDXB]
xxV0[VNDXC]
xxV0[VNDXD]
xxV1[VNDXA]
xxV1[VNDXB]
xxV1[VNDXC]
xxV1[VNDXD]
xxV0[VNDXA]
xxV0[VNDXB]
xxV0[VNDXC]
xxV0[VNDXD]
xxV1[VNDXA]
xxV1[VNDXB]
xxV1[VNDXC]
xxV1[VNDXD]
xxV0[VNDXB]
xxV0[VNDXC]
xxV0[VNDXD]
xxV1[VNDXA]
xxV1[VNDXB]
xxV1[VNDXC]
xxV1[VNDXD]
xxV0[VNDXA]
xxV0[VNDXB]
xxV0[VNDXC]
xxV0[VNDXD]
xxV1[VNDXA]
xxV1[VNDXB]
xxV1[VNDXC]
xxV1[VNDXD]
xxV0[VNDXB]
xxV0[VNDXC]
xxV0[VNDXD]
xxV1[VNDXA]
xxV1[VNDXB]
xxV1[VNDXC]
xxV1[VNDXD]
xxV2[VNDXA]
xxV2[VNDXB]
xxV2[VNDXC]
xxV2[VNDXD]
xxV3[VNDXA]
xxV3[VNDXB]
xxV3[VNDXC]
xxV3[VNDXD]
3:7
3:7
3:7
3:7
3:7
3:7
3:7
3:7
3:7
3:7
3:7
3:7
3:7
3:7
3:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2:7
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Note 1: In the victim index field columns, the “xx” in the SPR name refers
to one of “IN”, “IT”, “DN”, or “DT”, depending on whether the
access is to the instruction or data cache, and whether it is a
Note 2: As shown in the table, the 8KB cache size only uses bits 3:7 of
the victim index fields to select a way, since there are only 32
ways. Similarly, the 16KB and 32KB cache sizes uses bits 2:7 of
the victim index fields, since those cache sizes have 64 ways. In
all cases, the unused bits of the victim index fields are reserved.
The size of the fields of the victim limit registers (IVLIM, DVLIM)
are similarly affected by the number of sets in the cache (See
Note 3: Since the 8KB and 16KB cache sizes only have 8 sets, they only
use Address
to select the set and the victim index field, and
24:26
thus they do not use the xxV2 and xxV3 SPRs.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
4.1.2 Cache Locking and Transient Mechanism
Both caches support locking, at a “way” granularity. Any number of ways can be locked, from 0 ways to one
less than the total number of ways (64 ways for 32KB and 16KB cache sizes, 32 ways for the 8KB cache
size). At least one way must always be left unlocked, for use by cacheable line fills. Each way contains one
line from each set; that is, either 16 lines (512 bytes), for the 32KB cache size, or 8 lines (256 bytes), for the
16KB and 8KB cache sizes.
In addition, a portion of each cache can be designated as a “transient” region, by specifying that only a limited
number of ways are used for cache lines from memory pages that are identified as being transient in nature
such memory pages can be used for code sequences that are unlikely to be reused once the processor
moves on to the next series of instruction lines. Thus, performance may be improved by preventing each
series of instruction lines from overwriting the rest of the “regular” code in the instruction cache. Similarly, for
the data cache, transient pages can be used for large “streaming” data structures, such as multimedia data.
As each piece of the data stream is processed and written back to memory, the next piece can be brought in,
overwriting the previous (now obsolete) cache lines instead of displacing other areas of the cache, which may
contain other data that should remain in the cache.
A set of fields in a pair of victim limit registers specifies which ways of the cache are used for normal
accesses and/or transient accesses, as well as which ways are locked. These registers, Instruction Cache
from a GPR using mtspr, and can be read into a GPR using mfspr.
TFLOOR
NFLOOR
0
1
2
9
10
12 13
20 21
23 24
31
TCEILING
Figure 4-2. Instruction Cache Victim Limit (IVLIM) Data Cache Victim Limit (DVLIM)
0:1
Reserved
The number of bits in the TFLOOR field varies,
depending on the implemented cache size. See
Table 4-3,” on page -98 for more information.
2:9
TFLOOR
Transient Floor
Reserved
10:12
13:20
21:23
24:31
The number of bits in the TCEILING field varies,
depending on the implemented cache size. See
Table 4-3,” on page -98 for more information.
TCEILING
Transient Ceiling
Reserved
The number of bits in the NFLOOR field varies,
depending on the implemented cache size. See
Table 4-3,” on page -98 for more information.
NFLOOR
Normal Floor
When a cache line fill occurs as the result of a normal memory access (that is, one not marked as transient
using the U1 storage attribute from the MMU; see Memory Management on page 133), the cache line to be
replaced is selected by the corresponding victim index field from one of the normal victim index registers
(INV0–INV3 for instruction cache lines, DNV0–DNV3 for data cache lines). As the processor increments any
of these normal victim index fields according to the round-robin mechanism described in Cache Line
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Replacement Policy on page 96, the values of the fields are constrained to lie within the range specified by
the NFLOOR field of the corresponding victim limit register, and the last way of the cache (way 31 for the 8KB
cache size, way 63 for the 16KB or 32KB cache size). That is, when one of the normal victim index fields is
incremented past the last way of the cache, it wraps back to the value of the NFLOOR field of the associated
victim limit register.
Similarly, when a cache line fill occurs as the result of a transient memory access, the cache line to be
replaced is selected by the corresponding victim index field from one of the transient victim index registers
(ITV0–ITV3 for instruction cache lines, DTV0–DTV3 for data cache lines). As the processor increments any of
these transient victim index fields according to the round-robin replacement mechanism, the values of the
fields are constrained to lie within the range specified by the TFLOOR and the TCEILING fields of the corre-
sponding victim limit register. That is, when one of the transient victim index fields is incremented past the
TCEILING value of the associated victim limit register, it wraps back to the value of the TFLOOR field of that
victim limit register.
Given the operation of this mechanism, if both the NFLOOR and TFLOOR fields are set to 0, and the
TCEILING is set to the index of the last way of the cache, then all cache line fills—both normal and tran-
sient—are permitted to use the entire cache, and nothing is locked. Alternatively, if both the NFLOOR and
TFLOOR fields are set to values greater than 0, the lines in those ways of the cache whose indexes are
between 0 and the lower of the two floor values are effectively locked, as no cache line fills (neither normal
nor transient) will be allowed to replace the lines in those ways. Yet another example is when the TFLOOR is
lower than the NFLOOR, and the TCEILING is lower than the last way of the cache. In this scenario, the ways
between the TFLOOR and the NFLOOR contain only transient lines, while the ways between the NFLOOR
and the TCEILING may contain either normal or transient lines, and the ways from the TCEILING to the last
way of the cache contain only normal lines.
Programming Note: It is a programming error for software to program the TCEILING field to a
value lower than that of the TFLOOR field. Furthermore, software must
initialize each of the normal and transient victim index fields to values that
are between the ranges designated by the respective victim limit fields,
prior to performing any cacheable accesses intended to utilize these
ranges.
In order to setup a locked area within the data cache, software must perform the following steps (the proce-
dure for the instruction cache is similar, with icbt instructions substituting for dcbt instructions):
1. Execute msync and then isync to guarantee all previous cache operation have completed.
2. Mark all TLB entries associated with memory pages which are being used to perform the locking function
as caching-inhibited. Leave the TLB entries associated with the memory pages containing the data which
is to be locked into the data cache marked as cacheable, however.
3. Execute msync and then isync again, to cause the new TLB entry values to take effect.
4. Set both the NFLOOR and the TFLOOR values to the index of the first way which should be locked, and
set the TCEILING value to the last way of the cache.
5. Set each of the normal and transient victim index fields to the same value as the NFLOOR and TFLOOR.
6. Execute dcbt instructions to the cache lines within the cacheable memory pages which contain the data
which is to be locked in the data cache. The number of dcbt instructions executed to any given set should
not exceed the number of ways which will exist in the locked region (otherwise not all of the lines will be
able to be simultaneously locked in the data cache). Remember that when a series of dcbt instructions
are executed to sequentially increasing addresses (with the address increment being the size of a cache
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
block -- 32 bytes), it takes sixteen such dcbt operations (one for each set) before the next way of the ini-
tial set will be targeted again.
7. Execute msync and then isync again, to guarantee that all of the dcbt operations have completed and
updated the corresponding victim index fields.
8. Set the NFLOOR, TFLOOR, and TCEILING values to the desired indices for the operating normal and
transient regions of the cache. Both the NFLOOR and the TFLOOR values should be set higher than the
highest locked way of the data cache; otherwise, subsequent normal and/or transient accesses could
overwrite a way containing a line which was to be locked.
9. Set each of the normal and transient victim index fields to the value of the NFLOOR and TFLOOR,
respectively.
10. Restore the cacheability of the memory pages which were used to perform the locking function to the
desired operating values, by clearing the caching-inhibited attribute of the TLB entries which were
updated in step 2.
11. Execute msync and then isync again, to cause the new TLB entry values to take effect.
The ways of the data cache whose indices are below the lower of the NFLOOR and TFLOOR values will now
be locked.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
nisms. Other configurations are possible, given the ability to program each of the victim limit fields to different
relative values, although some configurations are not necessarily useful or practical.
1
Figure 4-3. Cache Locking and Transient Mechanism (Example 1)
1
Cache Set n
2
Way w
NORMAL LINES
Way NFLOOR
Way TCEILING
TRANSIENT LINES
Way TFLOOR
Way TFLOOR – 1
LOCKED LINES
Way 0
Note 1: This example illustrates partitioning of the cache
into locked, transient, and normal regions with no
overlap. The figure illustrates a single set, but all
sets of the cache are partitioned according to the
same victim limit values.
Note 2: w = 31 for 8KB cache, 63 for 16KB and 32KB
cache.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Figure 4-4. Cache Locking and Transient Mechanism (Example 2)
1
Cache Set n
2
Way w
NORMAL LINES
Way TCEILING+1
Way TCEILING
NORMAL/TRANSIENT LINES
Way NFLOOR
Way NFLOOR-1
TRANSIENT LINES
Way TFLOOR
Way TFLOOR-1
LOCKED LINES
Way 0
Note 1: This example illustrates partitioning of the cache
into locked, transient, and normal regions where
the transient and normal regions partially overlap.
The figure illustrates a single set, but all sets of the
cache are partitioned according to the same victim
limit values.
Note 2: w = 31 for 8KB cache, 63 for 16KB and 32KB
cache.
4.2 Instruction Cache Controller
The instruction cache controller (ICC) delivers two instructions per cycle to the instruction unit of the
PPC440x5 core. The ICC interfaces to the PLB using a 128-bit read interface, although it supports direct
attachment to 32-bit and 64-bit PLB subsystems, as well as 128-bit PLB subsystems. The ICC handles
frequency synchronization between the PPC440x5 core and the PLB, and can operate at any ratio of n:1, n:2,
and n:3, where n is an integer greater than the corresponding denominator.
The ICC provides a speculative prefetch mechanism which can be configured to automatically prefetch a
burst of up to three additional lines upon any fetch request which misses in the instruction cache.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
The ICC also handles the execution of the PowerPC instruction cache management instructions, for touching
(prefetching) or invalidating cache lines, or for flash invalidation of the entire cache. Resources for controlling
and debugging the instruction cache operation are also provided.
The rest of this section describes each of these functions in more detail.
4.2.1 ICC Operations
When the ICC receives an instruction fetch request from the instruction unit of the PPC440x5 core, the ICC
simultaneously searches the instruction cache array for the cache line associated with the virtual address of
page 133 for information about address translation). If the requested cache line is found in the array (a cache
hit), the pair of instructions at the requested address are returned to the instruction unit. If the requested
cache line is not found in the array (a cache miss), the ICC sends a request for the entire cache line (32
bytes) to the instruction PLB interface, using the real address. Note that the entire 32-bytecache line is
requested, even if the caching inhibited (I) storage attribute is set for the memory page containing that cache
line (see Caching Inhibited (I) on page 145). Also note that the request to the instruction PLB interface is sent
using the specific instruction address requested by the instruction unit, so that the memory subsystem may
read the cache line target word first (if it supports such operation) and supply the requested instructions
before retrieving the rest of the cache line.
As the ICC receives each portion of the cache line from the instruction PLB interface, it is placed into the
instruction cache line fill data (ICLFD) buffer. Instructions from this buffer may be bypassed to the instruction
unit as requested, without waiting for the entire cache line to be filled. Once the entire cache line has been
filled into the buffer, and assuming that the memory page containing that line is cacheable, it is written into the
instruction cache. If the memory page containing the line is caching inhibited, the line will remain in the ICLFD
until it is displaced by a subsequent request for another cache line (either cachable or caching inhibited).
If a memory subsystem error (such as an address time-out, invalid address, or some other type of hardware
error external to the PPC440x5 core) occurs during the filling of the cache line, the line will not be written into
the instruction cache, although instructions from the line may still be forwarded to the instruction unit from the
ICLFD. Later, if execution of an instruction from that line is attempted, an Instruction Machine Check excep-
tion will be reported, and a Machine Check interrupt (if enabled) will result. See Machine Check Interrupt on
Once a request for a cache line read has been requested on the instruction PLB interface, the entire line read
will be performed and the line will be written into the instruction cache (assuming no error occurs on the read),
regardless of whether or not the instruction stream branches (or is interrupted) away from the line being read.
This behavior is due to the nature of the PLB architecture, and the fact that once started, a cache line read
request type cannot be abandoned. This does not mean, however, that the ICC will wait for this cache line
read to complete before responding to a new request from the instruction unit (due, perhaps, to a branch redi-
rection, or an interrupt). Instead, the ICC will immediately access the cache to determine if the cache line at
the new address requested by the instruction unit is already in the cache. If so, the requested pair of instruc-
tions from this line will immediately be forwarded to the instruction unit, while the ICC in parallel continues to
fill the previously requested cache line. In other words, the instruction cache is completely non-blocking.
If the newly requested cache line is instead a miss in the instruction cache, the ICC will immediately attempt
to cancel the previous cache line read request. If the previous cache line read request has not yet been
requested on the PLB bus, the old request will be cancelled and the new request will be made. If the previous
cache line read request has already been requested, then as previously stated it cannot be abandoned, but
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
the ICC will immediately present the request for the new cache line, such that it may be serviced immediately
after the previous cache line read is completed. The ICC never aborts any PLB request once it has been
made, except when a processor reset occurs while the PLB request is being made.
Programming Note:
It is a programming error for an instruction fetch request to reference a valid cache line in the
instruction cache if the caching inhibited storage attribute is set for the memory page containing
the cache line. The result of attempting to execute an instruction from such an access is
undefined. After processor reset, hardware automatically sets the caching inhibited storage
attribute for the memory page containing the reset address, and also automatically flash
invalidates the instruction cache. Subsequently, lines will not be placed into the instruction cache
unless they are accessed by reference to a memory page for which the caching inhibited attribute
has been turned off. If software subsequently turns on the caching inhibited storage attribute for
such a page, software must make sure that no lines from that page remain valid in the instruction
cache, before attempting to fetch and execute instructions from the (now caching inhibited) page.
4.2.2 Speculative Prefetch Mechanism
The ICC can be configured to automatically prefetch up to three more cache lines upon (in addition to the line
being requested by the instruction unit) in response to a cache miss. This speculative prefetch only occurs on
requests for lines from cacheable memory pages, and then only if enabled by the setting of certain fields in
the Core Configuration Register 0 (CCR0) (see Figure 4-5 on page 109).
CCR0[ICSLC] specifies the number of additional cache lines (from 0 to 3) to speculatively prefetch upon an
instruction cache miss. If this field is non-zero, upon an instruction cache miss, the ICC will first check the
cache to see whether the additional lines are themselves already in the cache. If not, then the ICC will
present a fixed-length burst request to the instruction PLB interface, requesting the additional cache line(s).
The burst request is presented after the cache line request for the initial cache line requested by the instruc-
tion unit is presented and acknowledged on the PLB.
The speculative line fill mechanism will not request lines past the end of the minimum memory page size,
which is 1KB. That is, if the line requested by the instruction unit is at or near the end of an aligned 1KB
boundary, the speculative prefetch mechanism will only request those additional lines specified by the
CCR0[ICSLC] field that are also within the same 1KB page of memory. This allows the speculative prefetch
mechanism to operate without having to access the Memory Management Unit (MMU) for a translation for the
next page address.
Another field in the CCR0 register, CCR0[ICSLT], specifies a threshold value that is used to determine
whether the speculative burst request should be abandoned prior to completion, as a result of a change in
direction in the instruction stream (such as a branch or interrupt). If the instruction unit requests a new cache
line and the new request is a hit in the instruction cache, both the original line fill request and any speculative
burst request associated with it will be unaffected. Furthermore, if the new cache line requested by the
instruction unit is a miss in the instruction cache, any prior request which has not yet been requested on the
PLB interface will be cancelled, regardless of the value of CCR0[ICSLT]. However, if a prior speculative burst
request has already been requested on the PLB interface, the value of CCR0[ICSLT] determines if and when
the speculative burst request will be abandoned. CCR0[ICSLT] specifies the number of doublewords (8-byte
units) of the current cache line which must already have been received by the ICC, in order that the filling of
the current cache line will not be abandoned (note that in this context, the term “current” refers to the cache
line with which the next PLB data transfer is associated, at the time that the ICC determines that it needs to
request a new line). That is, if the ICC has already received the number of doublewords indicated by
CCR0[ICSLT], the ICC will not terminate the burst until it has received that entire cache line. All additional
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
lines beyond the one in progress at the time that the ICC determines that it needs to request a new line will be
abandoned. For example, if CCR0[ICSLC] is set to 3, and the ICC is in the middle of receiving the data for the
first of the three speculative lines at the time that the new instruction cache miss request is received from the
instruction unit, the second and third lines of the speculative burst will be abandoned, and whether the first of
the speculative lines is abandoned is controlled by CCR0[ICSLT].
Since cache lines contain 32 bytes, there are four doublewords in each cache line. Thus, CCR0[ICSLT] can
be set to a value from 0 to 3. If CCR0[ICSLT] = 0, the current line fill will be completed regardless of how
many doublewords have already been received. Similarly, if CCR0[ICSLT] = 3, the current line fill will be
abandoned if only two or fewer doublewords have been received by the ICC.
If at the time that the ICC determines that it needs to request a new line and abandon a speculative burst
request, the ICC has still not received all of the data associated with the initial cache line request which
prompted the speculative burst request, then this initial cache line is considered the “current” line, and the
speculative burst request will be abandoned without filling any of the speculative lines, regardless of the
setting of CCR0[ICSLT]. The filling of the initial cache line will be completed, however, as the PLB protocol
does not provide for the abandonment of the cache line (non-burst) request type.
Regardless of the value of CCR0[ICSLT], any time that a cache line fill is abandoned such that all of the data
for that cache line is not received, the line may still be used to bypass instructions to the instruction unit, but it
will not be written into the instruction cache, and it will be overwritten in the ICLFD buffer as soon as instruc-
tions for a new line begin arriving from the PLB.
4.2.3 Instruction Cache Coherency
In general, the PPC440x5 does not automatically enforce coherency between the instruction cache, data
cache, and memory. If the contents of memory location are changed, either within the data cache or within
memory itself, and whether by the PPC440x5 core through the execution of store instructions or by some
other mechanism in the system writing to memory, software must use cache management instructions to
ensure that the instruction cache is made coherent with these changes. This involves invalidating any obso-
lete copies of these memory locations within the instruction cache, so that they will be reread from memory
the next time they are referenced by program execution.
4.2.3.1 Self-Modifying Code
To illustrate the use of the cache management instructions to enforce instruction cache coherency, consider
the example of self-modifying code, whereby the program executing on the PPC440x5 core stores new data
to memory, with the intention of later branching to and executing this new “data,” which are actually instruc-
tions.
The following code example illustrates the required sequence for software to use when writing self-modifying
code. This example assumes that addr1 references a cacheable memory page.
stw
regN, addr1
addr1
# store the data (an instruction) in regN to addr1 in the data cache
# write the new instruction from the data cache to memory
# wait until the data actually reaches the memory
# invalidate addr1 in the instruction cache if it exists
# wait for the instruction cache invalidation to take effect
# flush any prefetched instructions within the ICC and instruction
# unit and re-fetch them (an older copy of the instruction at addr1
# may have already been fetched)
dcbst
msync
icbi
msync
isync
addr1
addr1
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
At this point, software may begin executing the instruction at addr1 and be guaranteed that the new instruc-
tion will be recognized.
4.2.3.2 Instruction Cache Synonyms
A synonym is a cache line that is associated with the same real address as another cache line that is in the
cache array at the same time. Such synonyms can occur when different virtual addresses are mapped to the
same real address, and the virtual address is used either as an index to the cache array (a virtually-indexed
cache) or as the cache line tag (a virtually-tagged cache).
The instruction cache on the PPC440x5 is real-indexed but virtually-tagged and thus it is possible for
synonyms to exist in the cache. (The data cache on the other hand is both real-indexed and real-tagged, and
thus cannot have any synonyms.) Because of this, special care must be taken when managing instruction
cache coherency and attempting to invalidate lines in the cache.
As explained in Memory Management on page 133, the virtual address (VA) consists of the 32-bit effective
address (EA; for instruction fetches, this is the address calculated by the instruction unit and sent to the ICC)
combined with the 8-bit Process ID (PID) and the 1-bit address space (MSR[IS] for instruction fetches). As
chooses the byte
27:31
offset within the cache line, while VA
is used as the index to select a set, and then the rest of the virtual
23:26
address is used as the tag. The tag thus consists of EA
, the PID, and MSR[IS] (for instruction fetches; for
0:22
cache management instructions such as icbi, MSR[DS] is used to specify the address space; see the instruc-
tion descriptions for the instruction cache management instructions for more information). The tag portion of
the VA is compared against the corresponding tag fields of each cache line within the way selected by
VA
.
23:26
Note that the address translation architecture of PowerPC Book-E is such that the low-order address bits
22:31 are always the same for the EA, VA, and real address (RA), because these bits are never translated
due to the minimum page size being 1KB (these low-order 10 bits are always used for the byte offset within
the page). As the page size increases, more and more low-order bits are used for the byte offset within the
page, and thus fewer and fewer bits are translated between the VA and the RA (see Table 5-3 on page 142).
Synonyms only become possible when the system-level memory management software establishes multiple
mappings to the same real page, which by definition involves different virtual addresses (either through differ-
ences in the higher-order EA bits which make up the VA, or through different process IDs, or different
address spaces, or some combination of these three portions of the VA).
A further requirement for synonyms to exist in the instruction cache is for more than one of the virtual pages
which map to a given real page to have execute permission, and for these pages to be cacheable (cache
lines associated with pages without execute permission, or for which the caching inhibited storage attribute is
set, cannot be placed in the instruction cache).
If the system-level memory management software permits instruction cache synonyms to be created, then
extra care must be taken when attempting to invalidate instruction cache lines associated with a particular
address. If software desires to invalidate only the cache line which is associated with a specific VA, then only
a single icbi instruction need be executed, specifying that VA. If, however, software wishes to invalidate all
instruction cache lines which are associated with a particular RA, then software must issue an icbi instruction
for each VA which has a mapping to that particular RA and for which a line might exist in the instruction
cache. In order to do this, the memory management software must keep track of which mappings to a given
RA exist (or ever existed, if a mapping has been removed but cache lines associated with it might still exist),
so that icbi instructions can be executed using the necessary VAs.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Alternatively, software can execute an iccci instruction, which flash invalidates the entire instruction cache
without regard to the addresses with which the cache lines are associated.
4.2.4 Instruction Cache Control and Debug
The PPC440x5 core provides various registers and instructions to control instruction cache operation and to
help debug instruction cache problems.
4.2.4.1 Instruction Cache Management and Debug Instruction Summary
For detailed descriptions of the instructions summarized in this section, see Instruction Set on page 249 Also,
see Instruction Cache Coherency on page 106 for more information on how these instructions are used to
manage coherency in the instruction cache.
In the instruction descriptions, the term “block” describes the unit of storage operated on by the cache block
instructions. For the PPC440x5 core, this is the same as a cache line.
The following instructions are used by software to manage the instruction cache:
icbi
icbt
Instruction Cache Block Invalidate
Invalidates a cache block.
Instruction Cache Block Touch
Initiates a block fill, enabling a program to begin a cache block fetch before the program
needs an instruction in the block. The program can subsequently branch to the
instruction address and fetch the instruction without incurring a cache miss.
iccci
Instruction Cache Congruence Class Invalidate
Flash invalidates the entire instruction cache. Execution of this instruction is privileged.
Instruction Cache Read
icread
Reads a cache line (tag and data) from a specified index of the instruction cache, into a
set of SPRs. Execution of this instruction is privileged.
4.2.4.2 Core Configuration Register 0 (CCR0)
The CCR0 register controls the speculative prefetch mechanism and the behavior of the icbt instruction. The
CCR0 register also controls various other functions within the PPC440x5 core that are unrelated to the
instruction cache. Each of the these functions is discussed in more detail in the related sections of this
manual.
Figure 4-5 illustrates the fields of the CCR0 register.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
PRE
DSTG
DTB GDCBT
ICSLC
0
1
2
3
4
5
9
10 11 12
DAPUIB
15 16 17 18 19
GICBT
22 23 24
FLSTA
27 28 29 30 31
ICSLT
CRPE
Figure 4-5. Core Configuration Register 0 (CCR0)
0
Reserved
Parity Recoverability Enable
0 Semi-recoverable parity mode enabled
for data cache
1 Fully recoverable parity mode enabled
for data cache
Must be set to 1 to guarantee full recoverability
from MMU and data cache parity errors.
1
PRE
2:3
4
Reserved
Cache Read Parity Enable
0 Disable parity information reads
1 Enable parity information reads
When enabled, execution of icread, dcread, or
tlbre loads parity information into the ICDBTRH,
DCDBTRL, or target GPR, respectively.
CRPE
5:9
Reserved
Disable Store Gathering
0 Enabled; stores to contiguous addresses
may be gathered into a single transfer
1 Disabled; all stores to memory will be
performed independently
10
DSTG
See Store Gathering on page 119.
Disable APU Instruction Broadcast
0 Enabled.
1 Disabled; instructions not broadcast to
APU for decoding
This mechanism is provided as a means of reduc-
ing power consumption when an auxilliary pro-
cessor is not attached and/or is not being used.
11
DAPUIB
See Initialization on page 85.
12:15
16
Reserved
Disable Trace Broadcast
0 Enabled.
1 Disabled; no trace information is
broadcast.
This mechanism is provided as a means of reduc-
ing power consumption when instruction tracing is
not needed.
DTB
See Initialization on page 85.
Guaranteed Instruction Cache Block Touch
0 icbt may be abandoned without having
filled cache line if instruction pipeline
stalls.
17
GICBT
See icbt Operation on page 111.
1 icbt is guaranteed to fill cache line even
if instruction pipeline stalls.
Guaranteed Data Cache Block Touch
0 dcbt/dcbtst may be abandoned without
having filled cache line if load/store
pipeline stalls.
See Data Cache Control and Debug on
page 125.
18
GDCBT
1 dcbt/dcbtst are guaranteed to fill cache
line even if load/store pipeline stalls.
19:22
Reserved
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Force Load/Store Alignment
0 No Alignment exception on integer
storage access instructions, regardless
of alignment
1 An alignment exception occurs on
integer storage access instructions if
data address is not on an operand
boundary.
23
FLSTA
See Load and Store Alignment on page 117.
24:27
28:29
Reserved
Number of additional lines (0–3) to fill on instruc-
tion fetch miss.
ICSLC
ICSLT
Instruction Cache Speculative Line Count
See Speculative Prefetch Mechanism on
page 105.
Number of doublewords that must have already
been filled in order that the current speculative
line fill is not abandoned on a redirection of the
instruction stream.
30:31
Instruction Cache Speculative Line Threshold
See Speculative Prefetch Mechanism on
page 105.
4.2.4.3 Core Configuration Register 1 (CCR1)
The CCR1 register controls parity error insertion for software testing, one option for line flush behavior in the
D-cache, and a control bit that selects the timer input clock. Each of the these functions is discussed in more
detail in the related sections of this manual.
Figure 4-6 illustrates the fields of the CCR1 register.
DCTPEI
FCOM
DCUPEI
ICDPEI
FFF
0
7
8
9
10 11 12 13 14 15 16
19 20 21
23 24 25
TCS
31
ICTPEI DCDPEI DCMPEI
MMUPEI
Figure 4-6. Core Configuration Register 1 (CCR1)
Instruction Cache Data Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bits recorded when the
instruction cache is filled. Each of the 8 bits corre-
sponds to one of the instruction words in the line.
0:7
ICDPEI
ICTPEI
DCTPEI
DCDPEI
Instruction Cache Tag Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bits recorded for the tag
field in the instruction cache.
8:9
Data Cache Tag Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bits recorded for the tag
field in the data cache.
10:11
12
Data Cache Data Parity Error Insert
0 record even parity (normal)
Controls inversion of parity bits recorded for the
data field in the data cache.
1 record odd parity (simulate parity error)
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Data Cache U-bit Parity Error Insert
0 record even parity (normal)
Controls inversion of parity bit recorded for the U
fields in the data cache.
13
14
DCUPEI
1 record odd parity (simulate parity error)
Data Cache Modified-bit Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bits recorded for the
modified (dirty) field in the data cache.
DCMPEI
FCOM
Force icbt , dcbt, dcbtst, dcbst, dcbf, dcbi, and
dcbz to appear to miss the caches. The intended
use is with icbt and dcbt only, which will fill a dupli-
cate line and allow testing of multi-hit parity errors.
See Section 4.2.4.7 Simulating Instruction Cache
Parity Errors for Software Testing on page 114 and
Figure 4.3.3.7 on page 130.
Force Cache Operation Miss
0 normal operation
1 cache ops appear to miss the cache
15
Memory Management Unit Parity Error Insert
0 record even parity (normal)
Controls inversion of parity bits recorded for the tag
field in the MMU.
16:19
MMUPEI
1 record odd parity (simulate parity error)
When flushing 32-byte (8-word) lines from the data
cache, normal operation is to write nothing, a dou-
ble word, quad word, or the entire 8-word block to
the memory as required by the dirty bits. This bit
ensures that none or all dirty bits are set so that
either nothing or the entire 8-word block is written
to memory when flushing a line from the data
cache. Refer to Section 4.3.1.4 Line Flush Opera-
tions on page 121.
Force Full-line Flush
0 flush only as much data as necessary.
1 always flush entire cache lines
20
FFF
21:23
24
Reserved
Timer Clock Select
0 CPU timer advances by one at each
rising edge of the CPU input clock
(CPMC440CLOCK).
1 CPU timer advances by one for each
rising edge of the CPU timer clock
(CPMC440TIMERCLOCK).
When TCS = 1, CPU timer clock input can toggle
at up to half of the CPU clock frequency.
TCS
25:31
Reserved
4.2.4.4 icbt Operation
The icbt instruction is typically used as a “hint” to the processor that a particular block of instructions is likely
to be executed in the near future. Thus the processor can begin filling that block into the instruction cache, so
that when the executing program eventually branches there the instructions will already be present in the
cache, thereby improving performance.
Of course, it would not typically be advantageous if the filling of the cache line requested by the icbt itself
caused a delay in the fetching of instructions needed by the currently executing program. For this reason, the
default behavior of the icbt instruction is for it to have the lowest priority for sending a request to the PLB. If a
subsequent instruction cache miss occurs due to a request from the instruction unit, then the line fill for the
icbt will be abandoned (if it has not already been acknowledged on the PLB).
On the other hand, the icbt instruction can also be used as a convenient mechanism for setting up a fixed,
known environment within the instruction cache. This is useful for establishing contents for cache line locking,
or for deterministic performance on a particular sequence of code, or even for debugging of low-level hard-
ware and software problems.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
When being used for these latter purposes, it is important that the icbt instruction deliver a deterministic
result, namely the guaranteed establishment in the cache of the specified line. Accordingly, the PPC440x5
core provides a field in the CCR0 register that can be used to cause the icbt instruction to operate in this
manner. Specifically, when the CCR0 [GICBT] field is set, the execution of icbt is guaranteed to establish the
specified cache line in the instruction cache (assuming that a TLB entry for the referenced memory page
exists and has both read and execute permission, and that the caching inhibited storage attribute is not set).
The cache line fill associated with such a guaranteed icbt will not be abandoned due to subsequent instruc-
tion cache misses.
Operation of the icbt instruction is affected by the CCR1[FCOM] bit, which forces the icbt to appear to miss
the cache, even if it should really be a hit. This causes two copies of the line to be established in the cache,
4.2.4.5 icread Operation
The icread instruction can be used to directly read both the tag and instruction information of a specified
word in a specified entry of the instruction cache. The instruction information is read into the Instruction
Cache Debug Data Register (ICDBDR), while the tag information is read into a pair of SPRs, the Instruction
Cache Debug Tag Register High (ICDBTRH) and Instruction Cache Debug Tag Register Low (ICDBTRL).
From there, the information can subsequently be moved into GPRs using mfspr instructions.
The execution of the icread instruction generates the equivalent of an EA, which is then broken down
according to the size of the cache and used to select a specific instruction word from a specific cache line, as
Table 4-4. Icread and dcread Cache Line Selection
Way
Selection
Set
Selection
Word
Selection
Array Size
Ignored
8KB
16KB
32KB
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
0:18
0:17
0:16
19:23
18:23
17:22
24:26
24:26
23:26
27:29
27:29
27:29
The EA generated by the icread instruction must be word-aligned (that is, EA
must be 0); otherwise, it is
30:31
a programming error and the result is undefined.
If the CCR0[CRPE] bit is set, execution of the icread instruction also loads parity information into the
ICBDTRH.
Execution of the icread instruction is privileged, and is intended for use for debugging purposes only.
Programming Note:
The PPC440x5 does not automatically synchronize context between an icread instruction and
the subsequent mfspr instructions which read the results of the icread instruction into GPRs. In
order to guarantee that the mfspr instructions obtain the results of the icread instruction, a
sequence such as the following must be used:
icread
regA,regB
# read cache information (the contents of GPR A and GPR B are
# added and the result used to specify a cache line index to be read)
# ensure icread completes before attempting to read results
isync
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
mficdbdr regC
# move instruction information into GPR C
# move high portion of tag into GPR D
# move low portion of tag into GPR E
mficdbtrh regD
mficdbtrl regE
The following figures illustrate the ICDBDR, ICDBTRH, and ICDBTRL.
0
31
Figure 4-7. Instruction Cache Debug Data Register (ICDBDR)
0:31
Instruction machine code from instruction cache
TPAR
TEA
0
23 24 25 26 27 28
DAPAR
31
V
Figure 4-8. Instruction Cache Debug Tag Register High (ICDBTRH)
Bits 0:23 of the 32-bit effective address associated
with the cache line read by icread.
0:23
Tag Effective Address
Cache Line Valid
0 Cache line is not valid.
1 Cache line is valid.
The valid indicator for the cache line read by
icread.
24
V
The parity bits for the address tag for the cache
line read by icread, if CCR0[CRPE] is set.
25:26
TPAR
Tag Parity
The parity bit for the instruction word at the 32-bit
effective address specified in the icread instruc-
tion, if CCR0[CRPE] is set.
27
DAPAR
Instruction Data parity
Reserved
28:31
TID
TS
0
21 22 23 24
TD
31
Figure 4-9. Instruction Cache Debug Tag Register Low (ICDBTRL)
0:21
Reserved
The address space portion of the virtual address
associated with the cache line read by icread.
22
TS
Translation Space
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Translation ID (TID) Disable
0 TID enable
1 TID disable
TID Disable field for the memory page associated
with the cache line read by icread.
23
TD
TID field portion of the virtual address associated
with the cache line read by icread.
24:31
TID
Translation ID
4.2.4.6 Instruction Cache Parity Operations
The instruction cache contains parity bits and multi-hit detection hardware to protect against soft data errors.
Both the instruction tags and data are protected. Instruction cache lines consist of a tag field, 256 bits of data,
and 10 parity bits. The tag field is stored in CAM (Content Addressible Memory) cells, while the data and
parity bits are stored in normal RAM cells. The instruction cache is virtually tagged and indexed, so the tag
field contains a TID field that is compared to the PID value, a TD bit that can be set to disable the TID
comparison for shared pages, and the effective address bits to be compared to the fetch request. The exact
number of effective address bits depends on the specific cache size.
Two types of errors may be detected by the instruction cache parity logic. In the first type, the parity bits
stored in the RAM array are checked against the appropriate data in the instruction cache line when the RAM
line is read for an instruction fetch. Note that a parity error will not be signaled as a result of an icread instruc-
tion.
The second type of parity error that may be detected is a multi-hit, sometimes referred to as an MHIT. This
type of error may occur when a tag address bit is corrupted, leaving two tags in the instruction cache array
that match the same input address. Multi-hit errors may be deteced on any instruction fetch. No parity errors
of any kind are detected on speculative fetch lookups or icbt lookups, Rather, such lookups are treated as
cache hits and cause no further action until an instruction fetch lookup at the offending address causes an
error to be detected.
If a parity error is detected, and the MSR[ME] is asserted, (i.e. Machine Check interrupts are enabled), the
processor vectors to the Machine Check interrupt handler. As is the case for any Machine Check interrupt,
after vectoring to the machine check handler, the MCSRR0 contains the value of the oldest “uncommitted”
instruction in the pipeline at the time of the exception and MCSRR1 contains the old Machine Status Register
(MSR) context. The interrupt handler is able to query Machine Check Status Register (MCSR) to find out that
it was called due to a instruction cache parity error, and is then expected to invalidate the I-cache (using
iccci). The handler returns to the interrupted process using the rfmci instruction.
As long as parity checking and machine check interrupts are enabled, instruction cache parity errors are
always recoverable. That is, they are detected and cause a machine check interrupt before the parity error
can cause the machine to update the architectural state with corrupt data. Also note that the machine check
interrupt is asynchronous; that is, the return address in the MCSRR0 does not point at the instruction address
that contains the parity error. Rather, the Machine Check interrupt is taken as soon as the parity error is
detected, and some instructions in progress will get flushed and re-excuted after the interrupt, just as if the
machine were responding to an external interrupt.
4.2.4.7 Simulating Instruction Cache Parity Errors for Software Testing
Because parity errors occur in the cache infrequently and unpredictably, it is desirable to provide users with a
way to simulate the effect of an instruction cache parity error so that interrupt handling software may be exer-
cised. This is exactly the purpose of the CCR1[ICDPEI], CCR1[ICTPEI], and CCR1[FCOM] fields.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
There are 10 parity bits stored in the RAM cells of each instruction cache line. Two of those bits hold the
parity for the tag information, and the remaining 8 bits hold the parity for each of the 8 32-bit instruction words
in the line. (There are two parity bits for the tag data because the parity is calculated for alternating bits of the
tag field, to guard against a single particle strike event that upsets two adjacent bits. The instruction data bits
are physically interleaved in such a way as to allow the use of a single parity bit per instruction word.) The
parity bits are calculated and stored as the line is filled into the cache. Usually parity is calculated as the even
parity for each set of bits to be protected, which the checking hardware expects. However, if any of the the
CCR1[ICTPEI] bits are set, the calculated parity for the corresponding bits of the tag are inverted and stored
as odd parity. Similarly, if any of the CCR1[ICDPEI] bits are set, the parity for the corresponding instruction
word is set to odd parity. Then, when the instructions stored with odd parity are fetched, they will cause a
Parity exception type Machine Check interrupt and exercise the interrupt handling software. The following
pseudo-code is an example of how to use the CCR1[ICDPEI] field to simulate a parity error on word 0 of a
target cache line:
; make sure all this code in the cache before execution
icbi <target line address>
msync
; get the target line out of the cache
; wait for the icbi
mtspr CCR1, 0x80000000
isync
icbt <target line address>
msync
; Set CCR1[ICDPEI ]
0
; wait for the CCR1 context to update
; this line fills and sets odd parity for word 0
; wait for the fill to finish
mtspr CCR1, 0x0
isync
; Reset CCR1[ICDPEI ]
; wait for the CCR1 context to update
0
br <word 0 of target line>
; fetching the target of the branch causes interrupt
Note that any instruction lines filled while bits are set in the CCR1[ICDPEI] or CCR1[ICTPEI] field will be
affected, so users must code carefully to affect only the intended addresses.
The CCR1[FCOM] (Force Cache Operation Miss) bit enables the simulation of a multi-hit parity error. When
set, it will cause an icbt to appear to be a miss, initiating a line fill, even if the line is really already in the
cache. Thus, this bit allows the same line to be filled to the cache multiple times, which will generate a multi-
hit parity error when an attempt is made to fetch an instruction from those cache lines. The following pseudo-
code is an example of how to use the CCR1[FCOM] field to simulate a multi-hit parity error in the instruction
cache:
; make sure all this code is cached and the “target line” is also
; in the cache before execution (use icbt as necessary)
mtspr CCR1, 0x00010000
isync
icbt <target line address>
msync
; Set CCR1[FCOM]
; wait for the CCR1 context to update
; this line fills a second copy of the target line
; wait for the fill to finish
mtspr CCR1, 0x0
isync
br <word 0 of target line>
; Reset CCR1[FCOM]
; wait for the CCR1 context to update
; fetching the target of the branch causes interrupt
4.3 Data Cache Controller
The data cache controller (DCC) handles the execution of the storage access instructions, moving data
between memory, the data cache, and the PPC440x5 core GPR file. The DCC interfaces to the PLB using
two independent 128-bit interfaces, one for read operations and one for writes. Both of these PLB interfaces
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
support direct attachment to 32-bit and 64-bit PLB subsystems, as well as 128-bit PLB subsystems. The DCC
handles frequency synchronization between the PPC440x5 core and the PLB, and can operate at any ratio of
n:1, n:2, and n:3, where n is an integer greater than the corresponding denominator.
The DCC also handles the execution of the PowerPC data cache management instructions, for touching
(prefetching), flushing, invalidating, or zeroing cache lines, or for flash invalidation of the entire cache.
Resources for controlling and debugging the data cache operation are also provided.
The DCC interfaces to the Auxiliary Processor (AP) port to provide direct load/store access to the data cache
for AP load and store operations, as well as for floating-point load and store instructions. AP load and store
instructions can access up to 16 bytes (one quadword) in a single cycle.
Extensive load, store, and flush queues are also provided, such that up to three outstanding line fills, up to
four outstanding load misses, and up to two outstanding line flushes can be pending, with the DCC continuing
to service subsequent load and store hits in an out-of-order fashion.
The rest of this section describes each of these functions in more detail.
4.3.1 DCC Operations
When the DCC executes a load, store, or data cache management instruction, the DCC first translates the
effective address specified by the instruction into a real address (see Memory Management on page 133 for
more information on address translation). Next, the DCC searches the data cache array for the cache line
associated with the real address of the requested data. If the cache line is found in the array (a cache hit),
that cache line is used to satisfy the request, according to the type of operation (load, store, and so on).
If the cache line is not found in the array (a cache miss), the next action depends upon the type of instruction
being executed, as well as the storage attributes of the memory page containing the data being accessed.
For most operations, and assuming the memory page is cacheable (see Caching Inhibited (I) on page 145),
the DCC will send a request for the entire cache line (32 bytes) to the data read PLB interface. The request to
the data read PLB interface is sent using the specific byte address requested by the instruction, so that the
memory subsystem may read the cache line target word first (if it supports such operation) and supply the
specific byte[s] requested before retrieving the rest of the cache line.
While the DCC is waiting for a cache line read to complete, it can continue to process subsequent instruc-
tions, and handle those accesses that hit in the data cache. That is, the data cache is completely non-
blocking.
As the DCC receives each portion of the cache line from the data read PLB interface, it is placed into one of
three data cache line fill data (DCLFD) buffers. Data from these buffers may be bypassed to the GPR file to
satisfy load instructions, without waiting for the entire cache line to be filled. Once the entire cache line has
been filled into the buffer, it will be written into the data cache at the first opportunity (either when the data
cache is otherwise idle, or when subsequent operations require that the DCLFD buffer be written to the data
cache).
If a memory subsystem error (such as an address time-out, invalid address, or some other type of hardware
error external to the PPC440x5 core) occurs during the filling of the cache line, the line will still be written into
the data cache, and data from the line may still be delivered to the GPR file for load instructions. However,
the DCC will also report a Data Machine Check exception to the instruction unit of the PPC440x5 core, and a
Machine Check interrupt (if enabled) will result. See Machine Check Interrupt on page 178 for more informa-
tion on Machine Check interrupts.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Once a data cache line read request has been made, the entire line read will be performed and the line will be
written into the data cache, regardless of whether or not the instruction stream branches (or is interrupted)
away from the instruction which prompted the initial line read request. That is, if a data cache line read is initi-
ated speculatively, before knowing whether or not a given instruction execution is really required (for
example, on a load instruction which is after an unresolved branch), that line read will be completed, even if it
is later determined that the cache line is not really needed. The DCC never aborts any PLB request once it
has been made, except when a processor reset occurs while the PLB request is being made.
In general, the DCC will initiate memory read requests without waiting to determine whether the access is
actually required by the sequential execution model (SEM). That is, the request will be initiated speculatively,
even if the instruction causing the request might be abandoned due to a branch, interrupt, or other change in
the instruction flow. Of course, write requests to memory cannot be initiated speculatively, although a line fill
request in response to a cacheable store access which misses in the data cache could be.
On the other hand, if the guarded storage attribute is set for the memory page being accessed, then the
memory request will not be initiated until it is guaranteed that the access is required by the SEM. Once initi-
ated, the access will not be abandoned, and the instruction is guaranteed to complete, prior to any change in
the instruction stream. That is, if the instruction stream is interrupted, then upon return the instruction execu-
tion will resume after the instruction which accessed guarded storage, such that the guarded storage access
will not be re-executed.
Programming Note:
It is a programming error for a load, store, or dcbz instruction to reference a valid cache line in
the data cache if the caching inhibited storage attribute is set for the memory page containing the
cache line. The result of such an access is undefined. After processor reset, hardware
automatically sets the caching inhibited storage attribute for the memory page containing the
dcbz instructions. Subsequently, lines will not be placed into the data cache unless they are
accessed by reference to a memory page for which the caching inhibited attribute has been
turned off. If software subsequently turns on the caching inhibited storage attribute for such a
page, software must make sure that no lines from that page remain valid in the data cache
(typically by using the dcbf instruction), before attempting to access the (now caching inhibited)
page with load, store, or dcbz instructions.
The only instructions that are permitted to reference a caching inhibited line which is a hit in the
data cache are the cache management instructions dcbst, dcbf, dcbi, dccci, and dcread. The
dcbt and dcbtst instructions have no effect if they reference a caching inhibited address,
regardless of whether the line exists in the data cache.
4.3.1.1 Load and Store Alignment
The DCC implements all of the integer load and store instructions defined for 32-bit implementations by the
PowerPC Book-E architecture. These include byte, halfword, and word loads and stores, as well as load and
store string (0 to 127 bytes) and load and store multiple (1 to 32 registers) instructions. Integer byte, halfword,
and word loads and stores are performed with a single access to memory if the entire data operand is
contained within an aligned 16-byte (quadword) block of memory, regardless of the actual operand alignment
within that block. If the data operand crosses a quadword boundary, the load or store is performed using two
accesses to memory.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
The load and store string and multiple instructions are performed using one memory access for each four
bytes, unless and until an access would cross an aligned quadword boundary. The access that would cross
the boundary is shortened to access just the number of bytes left within the current quadword block, and then
the accesses are resumed with four bytes per access, starting at the beginning of the next quadword block,
until the end of the load or store string or multiple is reached.
The DCC handles all misaligned integer load and store accesses in hardware, without causing an Alignment
exception. However, the control bit CCR0[FLSTA] can be set to force all misaligned storage access instruc-
tions to cause an Alignment exception (see Figure 4-5 on page 109). When this bit is set, all integer storage
accesses must be aligned on an operand-size boundary, or an Alignment exception will result. Load and
store multiple instructions must be aligned on a 4-byte boundary, while load and store string instructions can
be aligned on any boundary (these instructions are considered to reference byte strings, and hence the
operand size is a byte).
The DCC also supports load and store operations over the AP interface. These can include floating-point load
and store instructions (as defined by PowerPC Book-E), as well as AP load and store instructions for auxiliary
processors. While floating-point loads and stores can access either four or eight bytes, AP loads and store
can access up to a sixteen bytes.
The DCC handles all misaligned floating-point and AP loads and stores with a single memory access, as long
as they do not cross a quadword boundary. If such an access crosses a quadword boundary, the DCC will
signal an Alignment exception and an interrupt will result.
The AP interface also supports other options with regards to the handling of misaligned AP and floating-point
loads and stores. The AP interface can specify that the DCC should signal an Alignment exception on any AP
or floating-point load or store access which is not aligned on either an operand-size boundary or a word
boundary. Alternatively, the AP interface can specify that the DCC should force the storage access to be
aligned on an operand-size boundary by zeroing the appropriate number of low-order address bits.
Floating-point and AP loads and stores are also subject to the function of CCR0[FLSTA].
4.3.1.2 Load Operations
Load instructions that reference cacheable memory pages and miss in the data cache result in cache line
read requests being presented to the data read PLB interface. Load operations to caching inhibited memory
pages, however, will only access the bytes specifically requested, according to the type of load instruction.
This behavior (of only accessing the requested bytes) is only architecturally required when the guarded
storage attribute is also set, but the DCC will enforce this requirement on any load to a caching inhibited
memory page. Subsequent load operations to the same caching inhibited locations will cause new requests
to be sent to the data read PLB interface (data from caching inhibited locations will not be reused from the
DCLFD buffer).
The DCC includes three DCLFD buffers, such that a total of three independent data cache line fill requests
can be in progress at one time. The DCC can continue to process subsequent load and store accesses while
these line fills are in progress.
The DCC also includes a 4-entry load miss queue (LMQ), which holds up to four outstanding load instructions
that have either missed in the data cache or access caching inhibited memory pages. Collectively, any LMQ
entries which reference cacheable memory pages can reference no more than three different cache lines,
since there are only three DCLFD buffers. A load instruction in the LMQ remains there until the requested
data arrives in the DCLFD buffer, at which time the data is delivered to the register file and the instruction is
removed from the LMQ.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
4.3.1.3 Store Operations
The processing of store instructions in the DCC is affected by several factors, including the caching inhibited
(I), write-through (W), and guarded (G) storage attributes, as well as whether or not the allocation of data
cache lines is enabled for cacheable store misses. There are three different behaviors to consider:
• Whether a data cache line is allocated (if the line is not already in the data cache)
• Whether the data is written directly to memory or only into the data cache
• Whether the store data can be gathered with store data from previous or subsequent store instructions
before being written to memory
Allocation of Data Cache Line on Store Miss
Of course, if the caching inhibited attribute is set for the memory page being referenced by the store instruc-
tion, no data cache line will be allocated. For cacheable store accesses, allocation is controlled by one of two
mechanisms: either by a “global” control bit in the Memory Management Unit Control Register (MMUCR),
which is applied to all cacheable store accesses regardless of address; or by the U2 storage attribute for the
more information on how store miss cache line allocation is controlled.
Regardless of which mechanism is controlling the allocation, if the corresponding bit is set, the cacheable
store miss is handled as a store without allocate (SWOA). That is, if SWOA is indicated, then if the access
misses in the data cache, then the line will not be allocated (read from memory), and instead the byte[s] being
stored will be written directly to memory. Of course, if the cache line has already been allocated and is being
read into a DCLFD buffer (due perhaps to a previous cacheable load access), then the SWOA indication is
ignored and the access is treated as if it were a store with allocate. Similarly, if SWOA is not indicated, the
cache line will be allocated and the cacheable store miss will result in the cache line being read from memory.
Direct Write to Memory
Of course, if the caching inhibited attribute is set for the memory page being referenced by the store instruc-
tion, the data must be written directly to memory. For cacheable store accesses that are also write-through,
the store data will also be written directly to memory, regardless of whether the access hits in the data cache,
and independent of the SWOA mechanism. For cacheable store accesses that are not write through, whether
the data is written directly to memory depends on both whether the access hits or misses in the data cache,
and the SWOA mechanism. If the access is either a hit in the data cache, or if SWOA is not indicated, then
the data will only be written to the data cache, and not to memory. Conversely, if the cacheable store access
is both a miss in the data cache and SWOA is indicated, the access will be treated as if it were caching inhib-
ited and the data will be written directly to memory and not to the data cache (since the data cache line is
neither there already nor will it be allocated).
Store Gathering
In general, memory write operations caused by separate store instructions that specify locations in either
write-through or caching inhibited storage may be gathered into one simultaneous access to memory. Simi-
larly, store accesses that are handled as if they were caching inhibited (due to their being both a miss in the
data cache and being indicated as SWOA) may be gathered. Store accesses that are only written into the
data cache do not need to be gathered, because there is no performance penalty associated with the sepa-
rate accesses to the array.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
A given sequence of two store operations may only be gathered together if the targeted bytes are contained
within the same aligned quadword of memory, and if they are contiguous with respect to each other. Subse-
quent store operations may continue to be gathered with the previously gathered sequence, subject to the
same two rules (same aligned quadword and contiguous with the collection of previously gathered bytes). For
example, a sequence of three store word operations to addresses 4, 8, and 0 may all be gathered together,
as the first two are contiguous with each other, and the third (store word to address 0) is contiguous with the
gathered combination of the previous two.
An additional requirement for store gathering applies to stores which target caching inhibited memory pages.
Specifically, a given store to a caching inhibited page can only be gathered with previous store operations if
the bytes targeted by the given store do not overlap with any of the previously gathered bytes. In other words,
a store to a caching inhibited page must be both contiguous and non-overlapping with the previous store
operation(s) with which it is being gathered. This ensures that the multiple write operations associated with a
sequence of store instructions which each target a common caching inhibited location will each be performed
independently on that target location.
Finally, a given store operation will not be gathered with an earlier store operation if it is separated from the
earlier store operation by an msync or an mbar instruction, or if either of the two store operations reference a
memory page which is both guarded and caching inhibited, or if store gathering is disabled altogether by
CCR0[DSTG] (see Figure 4-5 on page 109).
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 4-5 summarizes how the various storage attributes and other circumstances affect the DCC behavior
on store accesses.
Table 4-5. Data Cache Behavior on Store Accesses
Store Access Attributes
Write
DCC Actions
Caching
Inhibited
(I)
Guarded
(G)
Write
Cache?
Write
Memory?
1
Hit/Miss
SWOA
through
(W)
Gather?
0
0
0
0
0
0
0
0
1
1
Hit
Hit
—
—
—
0
0
1
—
0
Yes
Yes
Yes
Yes
Yes
Yes
No
No
N/A
Yes
No
Yes
Yes
No
Hit
1
1
Miss
Miss
Miss
Miss
Miss
0
—
0
N/A
Yes
No
0
1
Yes
Yes
Yes
Yes
Yes
Yes
0
1
1
1
—
—
0
Yes
No
1
1
No
2
3
4
—
—
—
—
0
No
Yes
2
3
—
—
1
No
No
Note 1: If store gathering is disabled altogether (by setting CCR0[DSTG] to 1), then
such gathering will not occur, regardless of the indication in this table.
Furthermore, where this table indicates that store gathering may occur it is
presumed that the operations being gathered are targeting the same aligned
quadword of memory, and are contiguous with respect to each other.
Note 2: It is a programming error for a data cache hit to occur on a store access to a
caching inhibited page. The result of such an access is undefined.
Note 3: It is programming error for the write-through storage attribute to be set for a
page which also has the caching inhibited storage attribute set. The result of an
access to such a page is undefined.
Note 4: Stores to caching inhibited memory locations may only be gathered with
previous store operations if none of the targeted bytes overlap with the bytes
targeted by the previous store operations.
4.3.1.4 Line Flush Operations
When a store operation (or the dcbz instruction) writes data into the data cache without also writing the data
to main memory, the cache line is said to become dirty, meaning that the data in the cache is the current
value, whereas the value in memory is obsolete. Of course, when such a dirty cache line is replaced (due, for
example, to a new cache line fill overwriting the existing line in the cache), the data in the cache line must be
copied to memory. Otherwise, the results of the previous store operation[s] that caused the cache line to be
marked as dirty would be lost. The operation of copying a dirty cache line to memory is referred to as a cache
line flush. Cache lines are flushed either due to being replaced when a new cache line is filled, or in response
to an explicit software flush request associated with the execution of a dcbst or dcbf instruction.
The DCC implements four dirty bits per cache line, one for each aligned doubleword within the cache line.
Whenever any byte of a given doubleword is stored into a data cache line without also writing that same byte
to memory, the corresponding dirty bit for that cache line is set (if CCR1[FFF] is set, then all four dirty bits are
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
set instead of just the one corresponding dirty bit). When a data cache line is flushed, the type of request
made to the data write PLB interface depends upon which dirty bits associated with the line are set, and the
state of the CCR1[FFF] bit. If the CCR1[FFF] bit is set, the request will always be for an entire 32-byte line.
Most users will leave the CCR1[FFF] reset to zero, in which case the controller minimizes the size of the
transfer by the following algorithm. If only one dirty bit is set, the request type will be for a single doubleword
write. If only two dirty bits are set, and they are in the same quadword, then the request type will be for a 16-
byte line write. If two or more dirty bits are set, and they are in different quadwords, the request type will be for
an entire 32-byte line write. Regardless of the type of request generated by a cache line flush, the address is
always specified as the first byte of the request.
If a store access occurs to a cache line in a memory page for which the write-through storage attribute is set,
the dirty bits for that cache line do not get updated, since such a store access will be written directly to
memory (and into the data cache as well, if the access is either a hit or if the cache line is allocated upon a
miss).
On the other hand, it is permissible for there to exist multiple TLB entries that map to the same real memory
page, but specify different values for the write-through storage attribute. In this case, it is possible for a store
operation to a virtual page which is marked as non-write-through to have caused the cache line to be marked
as dirty, so that a subsequent store operation to a different virtual page mapped to the same real page but
marked as write-through encounters a dirty line in the data cache. If this happens, the store to the write-
through page will write the data for the store to both the data cache and to memory, but it will not modify the
dirty bits for the cache line.
4.3.1.5 Data Read PLB Interface Requests
When a PLB read request results from an access to a cacheable memory location, the request is always for a
32-byte line read, regardless of the type and size of the access that prompted the request. The address
presented will be for the first byte of the target of the access.
On the other hand, when a PLB read request results from an access to a caching-inhibited memory location,
only the byte[s] specifically accessed will be requested from the PLB, according to the type of instruction
prompting the access. Based on the type of storage access instructions (including integer, floating-point, and
AP), and based on the mechanism for handling misaligned accesses which cross a quadword boundary (see
caching inhibited requests:
• 1-byte read (any byte address 0–15 within a quadword)
• 2-byte read (any byte address 0–14 within a quadword)
• 3-byte read (any byte address 0–13 within a quadword)
• 4-byte read (any byte address 0–12 within a quadword)
• 8-byte read (any byte address 0–8 within a quadword)
This request can only occur due to a doubleword floating-point or AP load instruction
• 16-byte line fill (must be for byte address 0 of a quadword)
This request can only occur due to a quadword AP load instruction
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
4.3.1.6 Data Write PLB Interface Requests
When a PLB write request results from a data cache line flush, the specific type and size of the request is as
When a PLB write request results from store operations to caching-inhibited, write-through, and/or store
without allocate (SWOA) memory locations, the type and size of the request can be any one of the following
• 1-byte write request (any byte address 0–15 within a quadword)
• 2-byte write request (any byte address 0–14 within a quadword)
• 3-byte write request (any byte address 0–13 within a quadword)
• 4-byte write request (any byte address 0–12 within a quadword)
• 5-byte write request (any byte address 0–11 within a quadword)
Only possible due to store gathering
• 6-byte write request (any byte address 0–10 within a quadword)
Only possible due to store gathering
• 7-byte write request (any byte address 0–9 within a quadword)
Only possible due to store gathering
• 8-byte write request (any byte address 0–8 within a quadword)
Only possible due to store gathering, or due to a floating-point or AP doubleword store
• 9-byte write request (any byte address 0–7 within a quadword)
Only possible due to store gathering
• 10-byte write request (any byte address 0–6 within a quadword)
Only possible due to store gathering
• 11-byte write request (any byte address 0–5 within a quadword)
Only possible due to store gathering
• 12-byte write request (any byte address 0–4 within a quadword)
Only possible due to store gathering
• 13-byte write request (any byte address 0–3 within a quadword)
Only possible due to store gathering
• 14-byte write request (any byte address 0–2 within a quadword)
Only possible due to store gathering
• 15-byte write request (any byte address 0–1 within a quadword)
Only possible due to store gathering
• 16-byte line write request (must be to byte address 0 of a quadword)
Only possible due to store gathering, or due to an AP quadword store
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
4.3.1.7 Storage Access Ordering
In general, the DCC can perform load and store operations out-of-order with respect to the instruction stream.
That is, the memory accesses associated with a sequence of load and store instructions may be performed in
memory in an order different from that implied by the order of the instructions. For example, loads can be
processed ahead of earlier stores, or stores can be processed ahead of earlier loads. Also, later loads and
stores that hit in the data cache may be processed before earlier loads and stores that miss in the data
cache.
The DCC does enforce the requirements of the SEM, such that the net result of a sequence of load and store
operations is the same as that implied by the order of the instructions. This means, for example, that if a later
load reads the same address written by an earlier store, the DCC guarantees that the load will use the data
written by the store, and not the older “pre-store” data. But the memory subsystem could still see a read
access associated with an even later load before it sees the write access associated with the earlier store.
If the DCC needs to make a read request to the data read PLB interface, and this request conflicts with (that
is, references one or more of the same bytes as) an earlier write request which is being made to the data
write PLB interface, the DCC will withhold the read request from the data read PLB interface until the write
request has been acknowledged on the data write PLB interface. Once the earlier write request has been
acknowledged, the read request will be presented, and it is the responsibility of the PLB subsystem to ensure
that the data returned for the read request reflects the value of the data written by the write operation.
Conversely, if a write request conflicts with an earlier read request, the DCC will withhold the write request
until the read request has been acknowledged, at which point it is the responsibility of the PLB subsystem to
ensure that the data returned for the read request does not reflect the newer data being written by the write
request.
The PPC440x5 core provides storage synchronization instructions to enable software to control the order in
which the memory accesses associated with a sequence of instructions are performed. See Storage
4.3.2 Data Cache Coherency
The PPC440x5 core does not enforce the coherency of the data cache with respect to alterations of memory
performed by entities other than the PPC440x5 core. Similarly, if entities other than the PPC440x5 core
attempt to read memory locations which currently exist within the PPC440x5 core data cache and in a modi-
fied state, the PPC440x5 core does not recognize such accesses and thus will not respond to such accesses
with the modified data from the cache. In other words, the data cache on the PPC440x5 core is not a
snooping data cache, and there is no hardware enforcement of data cache coherency with memory with
respect to other entities in the system which access memory.
It is the responsibility of software to manage this coherency through the appropriate use of the caching inhib-
ited storage attribute, the write-through storage attribute, and/or the data cache management instructions.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
4.3.3 Data Cache Control and Debug
The PPC440x5 core provides various registers and instructions to control data cache operation and to help
debug data cache problems.
4.3.3.1 Data Cache Management and Debug Instruction Summary
For detailed descriptions of the instructions summarized in this section, see Instruction Set on page 249
In the instruction descriptions, the term “block” describes the unit of storage operated on by the cache block
instructions. For the PPC440x5 core, this is the same as a cache line.
The following instructions are used by software to manage the data cache.
dcba
dcbf
Data Cache Block Allocate
This instruction is implemented as a nop on the PPC440x5 core.
Data Cache Block Flush
Writes a cache block to memory (if the block has been modified) and then invalidates the
block.
dcbi
Data Cache Block Invalidate
Invalidates a cache block. Any modified data is discarded and not flushed to memory.
Execution of this instruction is privileged.
Data Cache Block Store
dcbst
dcbt
Writes a cache block to memory (if the block has been modified) and leaves the block
valid but marked as unmodified.
Data Cache Block Touch
Initiates a cache block fill, enabling the fill to begin prior to the executing program
requiring any data in the block. The program can subsequently access the data in the
block without incurring a cache miss.
dcbtst
dcbz
Data Cache Block Touch for Store
Implemented identically to the dcbt instruction.
Data Cache Block Set to Zero
Establishes a cache line in the data cache and sets the line to all zeros, without first
reading the previous contents of the cache block from memory, thereby improving
performance. All four doublewords in the line are marked as dirty.
dccci
Data Cache Congruence Class Invalidate
Flash invalidates the entire data cache. Execution of this instruction is privileged.
Data Cache Read
dcread
Reads a cache line (tag and data) from a specified index of the data cache, into a GPR
and a pair of SPRs. Execution of this instruction is privileged.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
4.3.3.2 Core Configuration Register 0 (CCR0)
The CCR0 register controls the behavior of the dcbt instruction, the handling of misaligned memory
accesses, and the store gathering mechanism. The CCR0 register also controls various other functions within
the PPC440x5 core that are unrelated to the data cache. Each of these functions is discussed in more detail
in the related sections of this manual.
4.3.3.3 Core Configuration Register 1 (CCR1)
The CCR1 register controls the behavior of the line flushes in response to cast-outs or dcbf or dcbst instruc-
tions. It also contains bits to control the artificial injection of parity errors for software testing purposes. Some
of those bits affect the data cache, while other control the MMU or instruction cache. Each of these functions
is discussed in more detail in the related sections of this manual.
4.3.3.4 dcbt and dcbtst Operation
The dcbt instruction is typically used as a “hint” to the processor that a particular block of data is likely to be
referenced by the executing program in the near future. Thus the processor can begin filling that block into
the data cache, so that when the executing program eventually performs a load from the block it will already
be present in the cache, thereby improving performance.
The dcbtst instruction is typically used for a similar purpose, but specifically for cases where the executing
program is likely to store to the referenced block in the near future. The differentiation in the purpose of the
dcbtst instruction relative to the dcbt instruction is only relevant within shared-memory systems with hard-
ware-enforced support for cache coherency. In such systems, the dcbtst instruction would attempt to estab-
lish the block within the data cache in such a fashion that the processor would most readily be able to
subsequently write to the block (for example, in a processor with a MESI-protocol cache subsystem, the block
might be obtained in Exclusive state). However, because the PPC440x5 core does not provide support for
hardware-enforced cache coherency, the dcbtst instruction is handled in an identical fashion to the dcbt
instruction. The rest of this section thus makes reference only to the dcbt instruction, but in all cases the
information applies to dcbtst as well.
Of course, it would not typically be advantageous if the filling of the cache line requested by the dcbt itself
caused a delay in the reading of data needed by the currently executing program. For this reason, the default
behavior of the dcbt instruction is for it to be ignored if the filling of the requested cache block cannot be
immediately commenced and waiting for such commencement would result in the DCC execution pipeline
being stalled. For example, the dcbt instruction will be ignored if all three DCLFD buffers are already in use,
and execution of subsequent storage access instructions is pending.
On the other hand, the dcbt instruction can also be used as a convenient mechanism for setting up a fixed,
known environment within the data cache. This is useful for establishing contents for cache line locking, or for
deterministic performance on a particular sequence of code, or even for debugging of low-level hardware and
software problems.
When being used for these latter purposes, it is important that the dcbt instruction deliver a deterministic
result, namely the guaranteed establishment in the cache of the specified line. Accordingly, the PPC440x5
core provides a field in the CCR0 register which can be used to cause the dcbt instruction to operate in this
manner. Specifically, when the CCR0 [GDCBT] field is set, the execution of dcbt is guaranteed to establish
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
the specified cache line in the data cache (assuming that a TLB entry for the referenced memory page exists
and has read permission, and that the caching inhibited storage attribute is not set). The cache line fill associ-
ated with such a guaranteed dcbt will occur regardless of any potential instruction execution-stalling circum-
stances within the DCC.
Operation of the dcbt instruction is affected by the CCR1[FCOM] bit, which forces the dcbt to appear to miss
the cache, even if it should really be a hit. This causes two copies of the line to be established in the cache,
4.3.3.5 dcread Operation
The dcread instruction can be used to directly read both the tag information and a specified data word in a
specified entry of the data cache. The data word is read into the target GPR specified in the instruction
encoding, while the tag information is read into a pair of SPRs, Data Cache Debug Tag Register High
(DCDBTRH) and Data Cache Debug Tag Register Low (DCDBTRL). The tag information can subsequently
be moved into GPRs using mfspr instructions.
The execution of the dcread instruction generates the equivalent of an EA, which is then broken down
according to the size of the cache and used to select a specific data word from a specific cache line, as
The EA generated by the dcread instruction must be word-aligned (that is, EA
must be 0); otherwise, it is
30:31
a programming error and the result is undefined.
If the CCR0[CRPE] bit is set, execution of the dcread instruction also loads parity information into the
DCDBTRL. Note that the DCDBTRL[DPAR] field, unlike all the other parity fields, loads the check values of
the parity, instead of the raw parity values. That is, the DPAR field will always load with zeros unless a parity
error has occurred, or been inserted intentionally using the appropriate bits in the CCR1. This behavior is an
artifact of the hardware design of the parity checking logic.
Execution of the dcread instruction is privileged, and is intended for use for debugging purposes only.
Programming Note:
The PPC440x5 core does not support the use of the dcread instruction when the DCC is still in
the process of performing cache operations associated with previously executed instructions
(such as line fills and line flushes). Also, the PPC440x5 core does not automatically synchronize
context between a dcread instruction and the subsequent mfspr instructions that read the
results of the dcread instruction into GPRs. In order to guarantee that the dcread instruction
operates correctly, and that the mfspr instructions obtain the results of the dcread instruction, a
sequence such as the following must be used:
msync
dcread
# ensure that all previous cache operations have completed
regT,regA,regB# read cache information; the contents of GPR A and GPR B are
# added and the result used to specify a cache line index to be read;
# the data word is moved into GPR T and the tag information is read
# into DCDBTRH and DCDBTRL
isync
# ensure dcread completes before attempting to read results
mfdcdbtrh regD
mfdcdbtrl regE
# move high portion of tag into GPR D
# move low portion of tag into GPR E
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
The following figures illustrate the DCDBTRH and DCDBTRL.
TERA
TRA
0
23 24 25
V
27 28
31
Figure 4-10. Data Cache Debug Tag Register High (DCDBTRH)
Bits 0:23 of the lower 32 bits of the 36-bit real
address associated with the cache line read by
dcread.
0:23
TRA
Tag Real Address
Cache Line Valid
0 Cache line is not valid.
1 Cache line is valid.
The valid indicator for the cache line read by
dcread.
24
V
25:27
28:31
Reserved
Upper 4 bits of the 36-bit real address associated
with the cache line read by dcread.
TERA
Tag Extended Real Address
DPAR
UPAR
D
U1
27 28 29 30 31
U0
U3
0
12 13 14 15 16
TPAR
19 20
23 24
U2
MPAR
Figure 4-11. Data Cache Debug Tag Register Low (DCDBTRL)
0:12
Reserved
The parity for the U0-U3 bits in the cache line read
by dcread if CCR0[CRPE] = 1, otherwise 0.
13
UPAR
TPAR
U bit parity
The parity for the tag bits in the cache line read by
dcread if CCR0[CRPE] = 1, otherwise 0.
14:15
16:19
Tag parity
Data parity
The parity check values for the data bytes in the
word read by dcread if CCR0[CRPE] = 1, other-
wise 0.
DPAR
MPAR
The parity for the modified (dirty) indicators for
each of the four doublewords in the cache line read
by dcread if CCR0[CRPE] = 1, otherwise 0.
20:23
Modified (dirty) parity
The “dirty” (modified) indicators for each of the four
doublewords in the cache line read by dcread.
24:27
28
D
Dirty Indicators
The U0 storage attribute for the memory page
associated with this cache line read by dcread.
U0
U1
U2
U3
U0 Storage Attribute
U1 Storage Attribute
U2 Storage Attribute
U3 Storage Attribute
The U1 storage attribute for the memory page
associated with this cache line read by dcread.
29
The U2 storage attribute for the memory page
associated with this cache line read by dcread.
30
The U3 storage attribute for the memory page
associated with this cache line read by dcread.
31
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
4.3.3.6 Data Cache Parity Operations
The data cache contains parity bits and multi-hit detection hardware to protect against soft data errors. Both
the data cache tags and data are protected. Data cache lines consist of a tag field, 256 bits of data, 4 modi-
fied (dirty) bits, 4 user attribute (U) bits, and 39 parity bits. The tag field is stored in CAM (Content Addressible
Memory) cells, while the data and parity bits are stored in normal RAM cells. The data cache is physically
tagged and indexed, so the tag field contains a real address that is compared to the real address produced by
the translation hardware when a load, store , or other cache operation is executed. The exact number of
effective address bits depends on the specific cache size.
Two types of errors are detected by the data cache parity logic. In the first type, the parity bits stored in the
RAM array are checked against the appropriate data in the RAM line any time the RAM line is read. The RAM
data may be read by an indexed operation such as a reload dump (RLD), or by a CAM lookup that matches
the tag address, such as a load, dcbf, dcbi, or dcbst. If a line is to be cast out of the cache due to replace-
ment or in response to a dcbf, dcbi, or dcbst, and is determined to have a parity error of this type, no effort
is made to prevent the erroneous data from being written onto the PLB. However, the write data on the PLB
interface is accompanied by a signal indicating that the data has a parity error.
The second type of parity error that may be detected is a multi-hit, also referred to as an MHIT. This type of
error may occur when a tag address bit is corrupted, leaving two tags in the memory array that match the
same input. This type of error may be detected on any CAM lookup cycle, such as for stores, loads, dcbf,
dcbi, dcbst, dcbt, dcbtst, or dcbz instructions. Note that a parity error will not be signaled as a result of an
dcread instruction.
If a parity error is detected and the MSR[ME] is asserted, (i.e. Machine Check interrupts are enabled), the
processor vectors to the Machine Check interrupt handler. As is the case for any machine check interrupt,
after vectoring to the machine check handler, the MCSRR0 contains the value of the oldest “uncommitted”
instruction in the pipeline at the time of the exception and MCSRR1 contains the old (MSR) context. The
interrupt handler is able to query Machine Check Status Register (MCSR) to find out that it was called due to
a D-cache parity error, and is then expected to either invalidate the data cache (using dccci), or to invoke the
OS to abort the process or reset the processor, as appropriate. The handler returns to the interrupted
process using the rfmci instruction.
If the interrupt handler is executed before a parity error is allowed to corrupt the state of the machine, the
executing process is recoverable, and the interrupt handler can just invalidate the data cache and resume
the process. In order to guarantee that all parity errors are recoverable, user code must have two characteris-
tics: first, it must mark all cacheable data pages as “write-through” instead of “copy-back.” Second, the soft-
ware-settable bit (CCR0[PRE]) must be set. This bit forces all load instructions to stall in the last stage of the
load/store pipeline for one cycle, but only if needed to ensure that parity errors are recoverable. The pipeline
stall guarantees that any parity error is detected and the resulting Machine Check interrupt taken before the
load instruction completes and the target GPR is corrupted. Setting CCR0[PRE] degrades overall application
performance. However, if the state of the load/store pipeline is such that a load instruction stalls in the last
stage for some reason unrelated to parity recoverability, then CCR0[PRE] does not cause an additional cycle
stall.
Note that the Parity exception type Machine Check interrupt is asynchronous; that is, the return address in
the MCSRR0 does not necessarily point at the instruction address that detected the parity error in the data
cache. Rather, the Machine Check interrupt is taken as soon as the parity error is detected, and some
instructions in progress may get flushed and re-excuted after the interrupt, just as if the machine were
responding to an external interrupt.
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
MCSR[DCSP] and MCSR[DCFP] indicate what type of data cache operation caused a parity exception. One
of the two bits will be set if a parity error is detected in the data cache, along with MCSR[MCS]. See Machine
MCSR[DCSP] is set if a parity error is detected during these search operations:
1. Multi-hit parity errors on any instruction that does a CAM lookup
2. Tag or data parity errors on load instructions
3. Tag parity errors on dcbf, dcbi, or dcbst instructions
MCSR[DCFP] is set if a parity error is detected during these flush operations:
1. Data, dirty, or user parity errors on dcbf or dcbst instructions
2. Tag, data, dirty, or user parity errors on a line that is cast out for replacement
4.3.3.7 Simulating Data Cache Parity Errors for Software Testing
Because parity errors occur in the cache infrequently and unpredictably, it is desirable to provide users with a
way to simulate the effect of an data cache parity error so that interrupt handling software may be exercised.
This is exactly the purpose of the CCR1[DCDPEI], CCR1[DCTPEI], CCR1[DCUPEI], CCR1[DCMPEI],and
CCR1[FCOM] fields.
The 39 data cache parity bits in each cache line contain one parity bit per data byte (i.e. 32 parity bits per 32
byte line), plus 2 parity bits for the address tag (note that the valid (V) bit, is not included in the parity bit calcu-
lation for the tag), plus 1 parity bit for the 4-bit U field on the line, plus a parity bit for each of the 4 modified
(dirty) bits on the line. (There are two parity bits for the tag data because the parity is calculated for alternating
bits of the tag field, to guard against a single particle strike event that upsets two adjacent bits. The other data
bits are physically interleaved in such a way as to allow the use of a single parity bit per data byte or other
field.) All parity bits are calculated and stored as the line is initially filled into the cache. In addition, the data
and modified (dirty) parity bits (but not the tag and user parity bits) are updated as the line is updated as the
result of executing a store instruction or dcbz.
Usually parity is calculated as the even parity for each set of bits to be protected, which the checking hard-
ware expects. However, if any of the the CCR1[DCTPEI] bits are set, the calculated parity for the corre-
sponding bits of the tag are inverted and stored as odd parity. Likewise, if the CCR1[DCUPEI] bit is set, the
calculated parity for the user bits is inverted and stored as odd parity. Similarly, if the CCR1[DCDPEI] bit is
set, the parity for any data bytes that are written, either during the process of a line fill or by execution of a
store instruction, is set to odd parity. Then, when the data stored with odd parity is subsequently loaded, it will
cause a Parity exception type Machine Check interrupt and exercise the interrupt handling software. The
following pseudocode is an example of how to use the CCR1[DCDPEI] field to simulate a parity error on byte
0 of a target cache line:
dcbt <target line address>
msync
; get the target line into the cache
; wait for the dcbt
mtspr CCR1, Rx
isync
stb <target byte address>
msync
; Set CCR1[DCDPEI]
; wait for the CCR1 context to update
; store some data at byte 0 of the target line
; wait for the store to finish
mtspr CCR1, Rz
isync
lb <byte 0 of target line>
; Reset CCR1[ICDPEI ]
; wait for the CCR1 context to update
; load byte causes interrupt
0
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
If the CCR1[DCMPEI] bit is set, the parity for any modified (dirty) bits that are written, either during the
process of a line fill or by execution of a store instruction or dcbz, is set to odd parity. If the CCR1[FFF] bit is
also set in addition to CCR1[DCMPEI], then the parity for all four modified (dirty) bits is set to odd parity. Store
access to a cache line that is already in the cache and in a memory page for which the write-through storage
attribute is set does not update the modified (dirty bits) nor the modified (dirty) parity bits, so for these
accesses the CCR1[DCMPEI] setting has no effect.
The CCR1[FCOM] (Force Cache Operation Miss) bit enables the simulation of a multi-hit parity error. When
set, it will cause an dcbt to appear to be a miss, initiating a line fill, even if the line is really already in the
cache. Thus, this bit allows the same line to be filled to the cache multiple times, which will generate a multi-
hit parity error when an attempt is made to read data from those cache lines. The following pseudocode is an
example of how to use the CCR1[FCOM] field to simulate a multi-hit parity error in the data cache:
mtspr CCR0, Rx
dcbt <target line address>
msync
; set CCR0[GDCBT]
; this dcbt fills a first copy of the target line, if necessary
; wait for the fill to finish
mtspr CCR1, Ry
isync
dcbt <target line address>
msync
; set CCR1[FCOM]
; wait for the CCR1 context to update
; fill a second copy of the target line
; wait for the fill to finish
mtspr CCR1, Rz
isync
br <byte 0 of target line>
; reset CCR1[FCOM]
; wait for the CCR1 context to update
; load byte causes interrupt
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
cache.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
5. Memory Management
The PPC440x5 supports a uniform, 4 gigabyte (GB) effective address (EA) space, and a 64GB (36-bit) real
address (RA) space. The PPC440x5 memory management unit (MMU) performs address translation
between effective and real addresses, as well as protection functions. With appropriate system software, the
MMU supports:
• Translation of effective addresses into real addresses
• Software control of the page replacement strategy
• Page-level access control for instruction and data accesses
• Page-level storage attribute control
5.1 MMU Overview
The PPC440x5 generates effective addresses for instruction fetches and data accesses. An effective
address is a 32-bit address formed by adding an index or displacement to a base address (see Effective
Address Calculation on page 41). Instruction effective addresses are for sequential instruction fetches, and
for fetches caused by changes in program flow (branches and interrupts). Data effective addresses are for
load, store and cache management instructions. The MMU expands effective addresses into virtual
addresses (VAs) and then translates them into real addresses (RAs); the instruction and data caches use
real addresses to access memory.
The PPC440x5 MMU supports demand-paged virtual memory and other management schemes that depend
on precise control of effective to real address mapping and flexible memory protection. Translation misses
and protection faults cause precise interrupts. The hardware provides sufficient information to correct the
fault and restart the faulting instruction.
The MMU divides storage into pages. The page represents the granularity of address translation, access
control, and storage attribute control. PowerPC Book-E architecture defines sixteen page sizes, of which the
PPC440x5 MMU supports eight. These eight page sizes (1KB, 4KB, 16KB, 64KB, 256KB, 1MB, 16MB, and
256MB) are simultaneously supported. A valid entry for a page referenced by an effective address must be in
the translation lookaside buffer (TLB) in order for the address to be accessed. An attempt to access an
address for which no TLB entry exists causes an Instruction or Data TLB Error interrupt, depending on the
type of access (instruction or data). See Interrupts and Exceptions on page 159 for more information on these
and other interrupt types.
The TLB is parity protected against soft errors. If such errors are detected, the CPU can be configured to
vector to the machine check interrupt handler, where software can take appropriate action. The details of
parity checking and suggested interrupt handling are described below.
5.1.1 Support for PowerPC Book-E MMU Architecture
The Book-E Enhanced PowerPC Architecture defines specific requirements for MMU implementations, but
also leaves the details of several features implementation-dependent. The PPC440x5 core is fully compliant
with the required MMU mechanisms defined by PowerPC Book-E, but a few optional mechanisms are not
supported. These are:
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
• Memory coherence required (M) storage attribute
Because the PPC440x5 does not provide hardware support for multiprocessor coherence, the memory
coherence required storage attribute has no effect. If a TLB entry is created with M=1, then any memory
transactions for the page associated with that TLB entry will be indicated as being memory coherence
required via a corresponding transfer attribute interface signal, but the setting will have no effect on the
operation within the PPC440x5 core.
• TLB Invalidate virtual address (tlbiva)instruction
The tlbiva instruction is used to support the invalidation of TLB entries in a multiprocessor environment
with hardware-enforced coherency, which is not supported by the PPC440x5. Consequently, the
attempted execution of this instruction will cause an Illegal Instruction exception type Program interrupt.
The tlbwe instruction may be used to invalidate TLB entries in a uniprocessor environment.
• TLB Synchronize (tlbsync) instruction
The tlbsync instruction is used to synchronize software TLB management operations in a multiprocessor
environment with hardware-enforced coherency, which is not supported by the PPC440x5. Conse-
quently, this instruction is treated as a no-op.
• Page Sizes
PowerPC Book-E defines sixteen different page sizes, but does not require that an implementation sup-
port all of them. Furthermore, some of the page sizes are only applicable to 64-bit implementations, as
they are larger than a 32-bit effective address space can support (4GB). Accordingly, the PPC440x5 sup-
ports eight of the sixteen page sizes, from 1KB up to 256MB, as mentioned above and as listed in Table
• Address Space
Since the PPC440x5 is a 32-bit implementation of the 64-bit PowerPC Book-E architecture, there are dif-
ferences in the sizes of some of the TLB fields. First, the Effective Page Number (EPN) field varies from
4 to 22 bits, depending on page size. Second, the page number portion of the real address is made up of
a concatenation of two TLB fields, rather than a single Real Page Number (RPN) field as described in
PowerPC Book-E. These fields are the RPN field (which can vary from 4 to 22 bits, depending on page
size), and the Extended Real Page Number (ERPN) field, which is 4 bits, for a total of 36 bits of real
5.2 Translation Lookaside Buffer
The Translation Lookaside Buffer (TLB) is the hardware resource that controls translation, protection, and
storage attributes. A single unified 64-entry, fully-associative TLB is used for both instruction and data
accesses. In addition, the PPC440x5 implements two separate, smaller “shadow” TLB arrays, one for instruc-
tion fetch accesses and one for data accesses. These shadow TLBs improve performance by lowering the
latency for address translation, and by reducing contention for the main unified TLB between instruction
fetching and data storage accesses. See Shadow TLB Arrays on page 151 for additional information on the
operation of the shadow TLB arrays.
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Maintenance of TLB entries is under software control. System software determines the TLB entry replace-
ment strategy and the format and use of any page table information. A TLB entry contains all of the informa-
tion required to identify the page, to specify the address translation, to control the access permissions, and to
designate the storage attributes.
A TLB entry is written by copying information from a GPR and the MMUCR[STID] field, using a series of three
tlbwe instructions. A TLB entry is read by copying the information into a GPR and the MMUCR[STID] field,
using a series of three tlbre instructions. Software can also search for specific TLB entries using the tlbsx[.]
instruction. See TLB Management Instructions on page 152 for more information on these instructions.
Each TLB entry identifies a page and defines its translation, access controls, and storage attributes. Accord-
ingly, fields in the TLB entry fall into four categories:
• Page identification fields (information required to identify the page to the hardware translation mecha-
nism).
• Address translation fields
• Access control fields
• Storage attribute fields
Table 5-1 summarizes the TLB entry fields for each of the categories.
Table 5-1. TLB Entry Fields
TLB
Word
Bit
Field
Description
Page Identification Fields
Effective Page Number (variable size, from 4 - 22 bits)
Bits 0:n–1 of the EPN field are compared to bits 0:n–1 of the effective address (EA) of the storage
0
0
0:21
22
EPN
V
access (where n = 32–log (page size in bytes) and page size is specified by the SIZE field of the
2
Valid (1 bit)
This bit indicates that this TLB entry is valid and may be used for translation. The Valid bit for a
given entry can be set or cleared with a tlbwe instruction.
Translation Address Space (1 bit)
This bit indicates the address space this TLB entry is associated with. For instruction fetch
accesses, MSR[IS] must match the value of TS in the TLB entry for that TLB entry to provide the
translation. Likewise, for data storage accesses (including instruction cache management opera-
tions), MSR[DS] must match the value of TS in the TLB entry. For the tlbsx[.] instruction, the
MMUCR[STS] field must match the value of TS.
0
23
TS
Page Size (4 bits)
SIZE
0
0
0
24:27
28:31
32:39
SIZE
The SIZE field specifies the size of the page associated with the TLB entry as 4
KB, where
Tag Parity (4 bits)
The TPAR field reads the parity bits associated with TLB word 0. These bits will be loaded into a
GPR as a result of a tlbre, but are ignored when executing a tlbwe, since the parity to be writ-
ten is calculated by the processor hardware.
TPAR
Translation ID (8 bits)
TID
Field used to identify a globally shared page (TID=0) or the process ID of the owner of a private
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table 5-1. TLB Entry Fields (continued)
TLB
Word
Bit
Field
Description
Address Translation Fields
Real Page Number (variable size, from 4 - 22 bits)
Bits 0:n–1 of the RPN field are used to replace bits 0:n–1 of the effective address to produce a
portion of the real address for the storage access (where n = 32–log (page size in bytes) and
2
1
0:21
RPN
page size is specified by the SIZE field of the TLB entry). Software must set unused low-order
Parity for TLB word 1 (2 bits)
The PAR1 field reads the parity bits associated with TLB word 1. These bits will be loaded into a
GPR as a result of a tlbre, but are ignored when executing a tlbwe, since the parity to be writ-
ten is calculated by the processor hardware.
1
1
22:23
28:31
PAR1
Extended Real Page Number (4 bits)
ERPN
The 4-bit ERPN field are prepended to the rest of the translated address to form a total of a 36-bit
Storage Attribute Fields
Parity for TLB word 2 (2 bits)
The PAR2 field reads the parity bits associated with TLB word 2. These bits will be loaded into a
GPR as a result of a tlbre, but are ignored when executing a tlbwe, since the parity to be writ-
ten is calculated by the processor hardware..
2
2
0:1
16
PAR2
U0
U1
Specifies the U0 storage attribute for the page associated with the TLB entry. The function of this
storage attribute is system-dependent, and has no effect within the PPC440x5 core.
Specifies the U1 storage attribute for the page associated with the TLB entry. The function of this
storage attribute is system-dependent, but the PPC440x5 core can be programmed to use this
attribute to designate a memory page as containing transient data and/or instructions (see
2
2
17
18
Specifies the U2 storage attribute for the page associated with the TLB entry. The function of this
storage attribute is system-dependent, but the PPC440x5 core can be programmed to use this
attribute to specify whether or not stores that miss in the data cache should allocate the line in the
U2
2
2
19
20
U3
W
Specifies the U3 storage attribute for the page associated with the TLB entry. The function of this
storage attribute is system-dependent, and has no effect within the PPC440x5 core.
0
1
The page is not write-through (that is, the page is copy-back).
The page is write-through.
2
21
I
0
1
The page is not caching inhibited (that is, the page is cacheable).
The page is caching inhibited.
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 5-1. TLB Entry Fields (continued)
TLB
Word
Bit
Field
Description
0
1
The page is not memory coherence required.
The page is memory coherence required.
2
22
M
Note that the PPC440x5 does not support multiprocessing, and thus all storage accesses will
behave as if M=0. Setting M=1 in a TLB entry has no effect other than to cause any system inter-
face transactions to the corresponding page to be indicated as memory coherence required via
the “transfer attributes” interface signals.
2
2
23
24
G
E
0
1
The page is not guarded.
The page is guarded.
All accesses to the page are performed with big-endian byte ordering, which means that
the byte at the effective address is considered the most-significant byte of a multi-byte sca-
0
1
All accesses to the page are performed with little-endian byte ordering, which means that
the byte at the effective address is considered the least-significant byte of a multi-byte sca-
Access Control Fields
Instruction fetch is not permitted from this page while MSR[PR]=1 and the attempt to exe-
2
2
26
27
UX
0
1
cute an instruction from this page while MSR[PR] =1 will cause an Execute Access Control
exception type Instruction Storage interrupt.
Instruction fetch and execution is permitted from this page while MSR[PR]=1.
Store operations and the dcbz instruction are not permitted to this page when MSR[PR]=1
and will cause a Write Access Control exception type Data Storage interrupt.
UW
0
1
Store operations and the dcbz instruction are permitted to this page when MSR[PR]=1.
Load operations and the dcbt, dcbtst, dcbst, dcbf, icbt, and icbi instructions are not per-
mitted from this page when MSR[PR]=1 and will cause a Read Access Control exception.
Except for the dcbt, dcbtst, and icbt instructions, a Data Storage interrupt will occur (see
0
2
2
28
UR
Load operations and the dcbt, dcbtst, dcbst, dcbf, icbt, and icbi instructions are permit-
ted from this page when MSR[PR]=1.
1
Instruction fetch is not permitted from this page while MSR[PR]=0 and the attempt to exe-
cute an instruction from this page while MSR[PR] =0 will cause an Execute Access Control
exception type Instruction Storage interrupt.
29
30
SX
0
1
Instruction fetch and execution is permitted from this page while MSR[PR]=0.
Store operations and the dcbz and dcbi instructions are not permitted to this page when
MSR[PR]=0 and will cause a Write Access Control exception type Data Storage interrupt.
0
2
SW
Store operations and the dcbz and dcbi instructions are permitted to this page when
MSR[PR]=0.
1
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table 5-1. TLB Entry Fields (continued)
TLB
Word
Bit
31
Field
Description
Load operations and the dcbt, dcbtst, dcbst, dcbf, icbt, and icbi instructions are not per-
mitted from this page when MSR[PR]=0 and will cause a Read Access Control exception.
Except for the dcbt, dcbtst, and icbt instructions, a Data Storage interrupt will occur (see
0
1
2
SR
Load operations and the dcbt, dcbtst, dcbst, dcbf, icbt, and icbi instructions are permit-
ted from this page when MSR[PR]=0.
5.3 Page Identification
The Valid (V), Effective Page Number (EPN), Translation Space Identifier (TS), Page Size (SIZE), and Trans-
lation ID (TID) fields of a particular TLB entry identify the page associated with that TLB entry. Except as
noted, all comparisons must succeed to validate this entry for subsequent translation and access control
processing. Failure to locate a matching TLB entry based on this criteria for instruction fetches will result in a
TLB Miss exception type Instruction TLB Error interrupt. Failure to locate a matching TLB entry based on this
criteria for data storage accesses will result in a TLB Miss exception which may result in a Data TLB Error
interrupt, depending on the type of data storage access (certain cache management instructions do not result
in an interrupt if they cause an exception; they simply no-op).
5.3.1 Virtual Address Formation
The first step in page identification is the expansion of the effective address into a virtual address. Again, the
effective address is the 32-bit address calculated by a load, store, or cache management instruction, or as
part of an instruction fetch. The virtual address is formed by prepending the effective address with a 1-bit
address space identifier and an 8-bit process identifier. The process identifier is contained in the Process ID
(PID) register. The address space identifier is provided by MSR[IS] for instruction fetches, and by MSR[DS]
for data storage accesses and cache management operations, including instruction cache management
operations. The resulting 41-bit value forms the virtual address, which is then compared to the virtual
addresses contained in the TLB entries.
Note that the tlbsx[.] instruction also forms a virtual address, for software controlled search of the TLB. This
instruction calculates the effective address in the same manner as a data access instruction, but the process
identifier and address space identifier are provided by fields in the MMUCR, rather than by the PID and MSR,
5.3.2 Address Space Identifier Convention
The address space identifier differentiates between two distinct virtual address spaces, one generally associ-
ated with interrupt-handling and other system-level code and/or data, and the other generally associated with
application-level code and/or data.
Typically, user mode programs will run with MSR[IS,DS] both set to 1, allowing access to application-level
code and data memory pages. Then, on an interrupt, MSR[IS,DS] are both automatically cleared to 0, so that
the interrupt handler code and data areas may be accessed using system-level TLB entries (that is, TLB
entries with the TS field = 0). It is also possible that an operating system could set up certain system-level
code and data areas (and corresponding TLB entries with the TS field = 1) in the application-level address
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
space, allowing user mode programs running with MSR[IS,DS] set to 1 to access them (system library
routines, for example, which may be shared by multiple user mode and/or supervisor mode programs).
System-level code wishing to use these areas would have to first set the corresponding MSR[IS,DS] field in
order to use the application-level TLB entries, or there would have to be alternative system-level TLB entries
set up.
The net of this is that the notion of application-level code running with MSR[IS,DS] set to 1 and using corre-
sponding TLB entries with the TS=1, and conversely system-level code running with MSR[IS,DS] set to 0 and
using corresponding TLB entries with TS=0, is by convention. It is possible to run in user mode with
MSR[IS,DS] set to 0, and conversely to run in supervisor mode with MSR[IS,DS] set to 1, with the corre-
sponding TLB entries being used. The only fixed requirement in this regard is the fact that MSR[IS,DS] are
cleared on an interrupt, and thus there must be a TLB entry for the system-level interrupt handler code with
TS=0 in order to be able to fetch and execute the interrupt handler itself. Whether or not other system-level
code switches MSR[IS,DS] and creates corresponding system-level TLB entries depends upon the operating
system environment.
Programming Note: Software must ensure that there is always a valid TLB entry with TS=0
and with supervisor mode execute access permission (SX=1)
corresponding to the effective address of the interrupt handlers.
Otherwise, an Instruction TLB Error interrupt could result upon the fetch
of the interrupt handler for some other interrupt type, and the registers
holding the state of the routine which was executing at the time of the
original interrupt (SRR0/SRR1) could be corrupted. See Interrupts and
5.3.3 TLB Match Process
This virtual address is used to locate the associated entry in the TLB. The address space identifier, the
process identifier, and a portion of the effective address of the storage access are compared to the TS, TID,
and EPN fields, respectively, of each TLB entry.
The virtual address matches a TLB entry if:
• The valid (V) field of the TLB entry is 1, and
• The value of the address space identifier is equal to the value of the TS field of the TLB entry, and
• Either the value of the process identifier is equal to the value of the TID field of the TLB entry (private
page), or the value of the TID field is 0 (globally shared page), and
• The value of bits 0:n–1 of the effective address is equal to the value of bits 0:n-1 of the EPN field of the
TLB entry (where n = 32–log (page size in bytes) and page size is specified by the value of the SIZE
2
field of the TLB entry). See Table 5-2 Page Size and Effective Address to EPN Comparison on page 140.
A TLB Miss exception occurs if there is no matching entry in the TLB for the page specified by the virtual
address (except for the tlbsx[.] instruction, which simply returns an undefined value to the GPR file and (for
2
Programming Note: Although it is possible for software to create multiple TLB entries that
match the same virtual address, doing so is a programming error and the
results are undefined.
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Figure 5-1 illustrates the criteria for a virtual address to match a specific TLB entry, while Table 5-2 defines
the page sizes associated with each SIZE field value, and the associated comparison of the effective address
to the EPN field.
TLBentry[V]
TLB entry matches virtual address
TLBentry[TS]
=?
AS
Process ID
private page
=?
shared page
Legend:
TLBentry[TID]
=0?
=?
MSR[IS] for instruction fetches, or
MSR[DS] for data storage accesses, or
MMUCR[STS] for tlbsx[.]
AS
{
TLBentry[EPN]
0:N-1
EA
effective address
EA
0:N-1
PID register for storage accesses
Process ID
N-1
{
MMUCR[STID] for tlbsx[.]
31 – log (page size)
2
Figure 5-1. Virtual Address to TLB Entry Match Process
Table 5-2. Page Size and Effective Address to EPN Comparison
SIZE
Page Size
EA to EPN Comparison
0b0000
0b0001
0b0010
0b0011
0b0100
0b0101
0b0110
0b0111
0b1000
0b1001
0b1010
0b1011
0b1100
0b1101
0b1110
0b1111
1KB
4KB
16KB
64KB
EPN
EPN
EPN
EPN
EPN
EPN
not supported
EPN =? EA
not supported
EPN =? EA
=? EA
=? EA
=? EA
=? EA
=? EA
=? EA
0:21
0:19
0:17
0:15
0:13
0:11
0:21
0:19
0:17
0:15
0:13
0:11
256KB
1MB
not supported
16MB
not supported
256MB
not supported
not supported
not supported
not supported
not supported
not supported
0:7
0:7
0:3
0:3
not supported
not supported
not supported
not supported
not supported
not supported
5.4 Address Translation
Once a TLB entry is found which matches the virtual address associated with a given storage access, as
described in Page Identification on page 138, the virtual address is translated to a real address according to
the procedures described in this section.
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
The Real Page Number (RPN) and Extended Real Page Number (ERPN) fields of the matching TLB entry
provide the page number portion of the real address. Let n=32–log (page size in bytes) where page size is
2
specified by the SIZE field of the matching TLB entry. Bits n:31 of the effective address (the “page offset”) are
appended to bits 0:n–1 of the RPN field, and bits 0:3 of the ERPN field are prepended to this value to produce
the 36-bit real address (that is, RA = ERPN || RPN
|| EA
).
0:3
0:n–1
n:31
Figure 5-2 illustrates the address translation process, while Table 5-3 defines the relationship between the
different page sizes and the real address formation.
MSR[DS] for data storage accesses
MSR[IS] for instruction fetch
32-bit Effective Address
PID
Effective Page Number (EPN)
Offset
0
0
n–1
n
31
7
41-bit Virtual Address
64-entry TLB
RPN
0:n-1
ERPN
0:3
Extended
RPN
Offset
Real Page Number (RPN)
0
3
0
n–1
n
31
36-bit Real Address
NOTE: n = 32–log (page size)
2
Figure 5-2. Effective-to-Real Address Translation Flow
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table 5-3. Page Size and Real Address Formation
SIZE
Page Size
RPN bits required to be 0
none
Real Address
0b0000
0b0001
0b0010
0b0011
0b0100
0b0101
0b0110
0b0111
0b1000
0b1001
0b1010
0b1011
0b1100
0b1101
0b1110
0b1111
1KB
4KB
16KB
64KB
RPN
RPN
RPN
RPN
RPN
RPN
|| EA
|| EA
|| EA
|| EA
|| EA
|| EA
0:21
0:19
0:17
0:15
0:13
0:11
22:31
20:31
18:31
16:31
14:31
12:31
RPN
RPN
RPN
RPN
RPN
=0
=0
=0
=0
=0
20:21
18:21
16:21
14:21
12:21
256KB
1MB
not supported
16MB
not supported
256MB
not supported
not supported
not supported
not supported
not supported
not supported
not supported
RPN =0
not supported
RPN || EA
8:31
8:21
0:7
not supported
RPN =0
not supported
RPN || EA
4:31
4:21
0:3
not supported
not supported
not supported
not supported
not supported
not supported
not supported
not supported
not supported
not supported
not supported
not supported
5.5 Access Control
Once a matching TLB entry has been identified and the address has been translated, the access control
mechanism determines whether the program has execute, read, and/or write access to the page referenced
by the address, as described in the following sections.
5.5.1 Execute Access
The UX or SX bit of a TLB entry controls execute access to a page of storage, depending on the operating
mode (user or supervisor) of the processor.
• User mode (MSR[PR] = 1)
Instructions may be fetched and executed from a page in storage while in supervisor mode if the SX access
control bit for that page is equal to 1. If the SX access control bit is equal to 0, then instructions from that page
will not be fetched, and will not be placed into any cache as the result of a fetch request to that page while in
supervisor mode.
Furthermore, if the sequential execution model calls for the execution in supervisor mode of an instruction
from a page that is not enabled for execution in supervisor mode (that is, SX=0 when MSR[PR]=0), an
Execute Access Control exception type Instruction Storage interrupt is taken (See Interrupts and Exceptions
5.5.2 Write Access
The UW or SW bit of aTLB entry controls write access to a page, depending on the operating mode (user or
supervisor) of the processor.
• User mode (MSR[PR] = 1)
Store operations (including the store-class cache management instruction dcbz) are permitted to a page
in storage while in user mode if the UW access control bit for that page is equal to 1. If execution of a
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
store operation is attempted in user mode to a page for which the UW access control bit is 0, then a Write
Access Control exception occurs. If the instruction is an stswx with string length 0, then no interrupt is
interrupt is taken.
Note that although the dcbi cache management instruction is a store-class instruction, its execution is
privileged and thus will not cause a Data Storage interrupt if execution of it is attempted in user mode (a
Privileged Instruction exception type Program interrupt will occur instead).
• Supervisor mode (MSR[PR] = 0)
Store operations (including the store-class cache management instructions dcbz and dcbi) are permitted
to a page in storage while in supervisor mode if the SW access control bit for that page is equal to 1. If
execution of a store operation is attemped in supervisor mode to a page for which the SW access control
bit is 0, then a Write Access Control exception occurs. If the instruction is an stswx with string length 0,
then no interrupt is taken and no operation is performed (see Access Control Applied to Cache Manage-
ment Instructions on page 143). For all other store operations, execution of the instruction is suppressed
and a Data Storage interrupt is taken.
5.5.3 Read Access
The UR or SR bit of a TLB entry controls read access to a page, depending on the operating mode (user or
supervisor) of the processor.
• User mode (MSR[PR] = 1)
Load operations (including the load-class cache management instructions dcbst, dcbf, dcbt, dcbtst,
icbi, and icbt) are permitted from a page in storage while in user mode if the UR access control bit for
that page is equal to 1. If execution of a load operation is attempted in user mode to a page for which the
UR access control bit is 0, then a Read Access Control exception occurs. If the instruction is a load (not
including lswx with string length 0) or is a dcbst, dcbf, or icbi, then execution of the instruction is sup-
pressed and a Data Storage interrupt is taken. On the other hand, if the instruction is an lswx with string
length 0, or is a dcbt, dcbtst, or icbt, then no interrupt is taken and no operation is performed (see
• Supervisor mode (MSR[PR] = 0)
Load operations (including the load-class cache management instructions dcbst, dcbf, dcbt, dcbtst,
icbi, and icbt) are permitted from a page in storage while in supervisor mode if the SR access control bit
for that page is equal to 1. If execution of a load operation is attempted in supervisor mode to a page for
which the SR access control bit is 0, then a Read Access Control exception occurs. If the instruction is a
load (not including lswx with string length 0) or is a dcbst, dcbf, or icbi, then execution of the instruction
is suppressed and a Data Storage interrupt is taken. On the other hand, if the instruction is an lswx with
string length 0, or is a dcbt, dcbtst, or icbt, then no interrupt is taken and no operation is performed (see
5.5.4 Access Control Applied to Cache Management Instructions
This section summarizes how each of the cache management instructions is affected by the access control
mechanism.
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
• dcbz instructions are treated as stores with respect to access control since they actually change the data
in a cache block. As such, they can cause Write Access Control exception type Data Storage interrupts.
• dcbi instructions are treated as stores with respect to access control since they can change the value of
a storage location by invalidating the “current” copy of the location in the data cache, effectively “restor-
ing” the value of the location to the “former” value which is contained in memory. As such, they can cause
Write Access Control exception type Data Storage interrupts.
• dcba instructions are treated as no-ops by the PPC440x5 under all circumstances, and thus can not
cause any form of Data Storage interrupt.
• icbi instructions are treated as loads with respect to access control. As such, they can cause Read
Access Control exception type Data Storage interrupts. Note that this instruction may cause a Data Stor-
age interrupt (and not an Instruction Storage interrupt), even though it otherwise would perform its opera-
tion on the instruction cache. Instruction storage interrupts are associated with exceptions which occur
upon the fetch of an instruction, whereas Data storage interrupts are associated with exceptions which
occur upon the execution of a storage access or cache management instruction.
• dcbt, dcbtst, and icbt instructions are treated as loads with respect to access control. As such, they can
cause Read Access Control exceptions. However, because these instructions are intended to act merely
as “hints” that the specified cache block will likely be accessed by the processor in the near future, such
exceptions will not result in a Data Storage interrupt. Instead, if a Read Access Control exception occurs,
the instruction is treated as a no-op.
• dcbf and dcbst instructions are treated as loads with respect to access control. As such, they can cause
Read Access Control exception type Data Storage interrupts. Flushing or storing a dirty line from the
cache is not considered a store since an earlier store operation has already updated the cache line, and
the dcbf or dcbst instruction is simply causing the results of that earlier store operation to be propagated
to memory.
• dccci and iccci instructions do not even generate an address, nor are they affected by the access control
mechanism. They are privileged instructions, and if executed in supervisor mode they will flash invalidate
the entire associated cache.
Table 5-4 summarizes the effect of access control on each of the cache management instructions.
Table 5-4. Access Control Applied to Cache Management Instructions
Read
Write
Protection
Violation
Protection
Violation
Instruction
Exception?
Exception?
No
Yes
No
No
No
Yes
No
No
No
Yes
No
No
dcba
dcbf
dcbi
Yes
dcbst
dcbt
1
Yes
1
Yes
dcbtst
dcbz
dccci
icbi
No
No
Yes
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 5-4. Access Control Applied to Cache Management Instructions
Read
Write
Protection
Violation
Protection
Violation
Instruction
Exception?
Exception?
1
Yes
No
No
icbt
iccci
No
1.
dcbt, dcbtst, or icbt may cause a Read
Access Control exception but will not result in a Data
Storage interrupt
5.6 Storage Attributes
Each TLB entry specifies a number of storage attributes for the memory page with which it is associated.
Storage attributes affect the manner in which storage accesses to a given page are performed. The storage
attributes (and their corresponding TLB entry fields) are:
• Write-through (W)
• Caching inhibited (I)
• Memory coherence required (M)
• Guarded (G)
• Endianness (E)
• User-definable (U0, U1, U2, U3)
All combinations of these attributes are supported except combinations which simultaneously specify a region
as write-through and caching inhibited.
5.6.1 Write-Through (W)
If a memory page is marked as write-through (W=1), then the data for all store operations to that page are
written to memory, as opposed to only being written into the data cache. If the referenced line also exists in
the data cache (that is, the store operation is a “hit”), then the data will also be written into the data cache,
although the cache line will not be marked as having been modified (that is, the “dirty” bit(s) will not be set).
See Instruction and Data Caches on page 95 for more information on the handling of accesses to write-
through storage.
5.6.2 Caching Inhibited (I)
If a memory page is marked as caching inhibited (I=1), then all load, store, and instruction fetch operations
perform their access in memory, as opposed to in the respective cache. If I=0, then the page is cacheable
and the operations may be performed in the cache.
It is a programming error for the target location of a load, store, dcbz, or fetch access to caching inhibited
storage to be in the respective cache; the results of such an access are undefined. It is not a programming
error for the target locations of the other cache management instructions to be in the cache when the caching
inhibited storage attribute is set. The behavior of these instructions is defined for both I=0 and I=1 storage.
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
See Instruction and Data Caches on page 95 for more information on the handling of accesses to caching
inhibited storage.
5.6.3 Memory Coherence Required (M)
The memory coherence required (M) storage attribute is defined by the architecture to support cache and
memory coherency within multiprocessor shared memory systems. Because the PPC440x5 does not provide
hardware support for multiprocessor coherence, the memory coherence required storage attribute has no
effect. If a TLB entry is created with M = 1, any storage accesses to the page associated with that TLB entry
are indicated, using the corresponding transfer attribute interface signal, as being memory coherence
required, but the setting has no effect on the operation within the PPC440x5 core.
5.6.4 Guarded (G)
The guarded storage attribute is provided to control “speculative” access to “non-well-behaved” memory loca-
tions. Storage is said to be “well-behaved” if the corresponding real storage exists and is not defective, and if
the effects of a single access to it are indistinguishable from the effects of multiple identical accesses to it. As
such, data and instructions can be fetched out-of-order from well-behaved storage without causing undesired
side effects.
In general, storage that is not well-behaved should be marked as guarded. Because such storage may repre-
sent a control register on an I/O device or may include locations that do not exist, an out-of-order access to
such storage may cause an I/O device to perform unintended operations or may result in a Machine Check
exception. For example, if the input buffer of a serial I/O device is memory-mapped, then an out-of-order or
speculative access to that location could result in the loss of an item of data from the input buffer, if the
instruction execution is interrupted and later re-attempted.
A data access to a guarded storage location is performed only if either the access is caused by an instruction
that is known to be required by the sequential execution model, or the access is a load and the storage loca-
tion is already in the data cache. Once a guarded data storage access is initiated, if the storage is also
caching inhibited then only the bytes specifically requested are accessed in memory, according to the
operand size for the instruction type. Data storage accesses to guarded storage which is marked as cache-
able may access the entire cache block, either in the cache itself or in memory.
Instruction fetch is not affected by guarded storage. While the architecture does not prohibit instruction
fetching from guarded storage, system software should generally prevent such instruction fetching by
marking all guarded pages as “no-execute” (UX/SX = 0). Then, if an instruction fetch is attempted from such
a page, the memory access will not occur and an Execute Access Control exception type Instruction Storage
interrupt will result if and when execution is attempted for an instruction at any address within the page.
See Section 4 Instruction and Data Caches on page 95 for more information on the handling of accesses to
guarded storage. Also see Partially Executed Instructions on page 164 for information on the relationship
between the guarded storage attribute and instruction restart and partially executed instructions.
5.6.5 Endian (E)
The endian (E) storage attribute controls the byte ordering with which load, store, and fetch operations are
performed. Byte ordering refers to the order in which the individual bytes of a multiple-byte scalar operand are
arranged in memory. The operands in a memory page with E=0 are arranged with big-endian byte ordering,
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
which means that the bytes are arranged with the most-significant byte at the lowest-numbered memory
address. The operands in a memory page with E=1 are arranged with little-endian byte ordering, which
means that the bytes are arranged with the least-significant byte at the lowest-numbered address.
See Byte Ordering on page 42 for a more detailed explanation of big-endian and little-endian byte ordering.
5.6.6 User-Definable (U0–U3)
The PPC440x5 core provides four user-definable (U0–U3) storage attributes which can be used to control
system-dependent behavior of the storage system. By default, these storage attributes do not have any effect
on the operation of the PPC440x5 core, although all storage accesses indicate to the memory subsystem the
values of U0–U3 using the corresponding transfer attribute interface signals. The specific system design may
then take advantage of these attributes to control some system-level behaviors. As an example, one of the
user-definable storage attributes could be used to enable code compession using the IBM CodePack core, if
this function is included within a specific implementation incorporating the PPC440x5 core.
On the other hand, the PPC440x5 core can be programmed to make specific use of two of the four user-
definable storage attributes. Specifically, by enabling the function using a control bit in the MMUCR (see
Memory Management Unit Control Register (MMUCR) on page 148), the U1 storage attribute can be used to
designate whether storage accesses to the associated memory page should use the “normal” or “transient”
region of the respective cache. Similarly, another control bit in the MMUCR can be set to enable the U2
storage attribute to be used to control whether or not store accesses to the associated memory page which
miss in the data cache should allocate the line in the cache. The U1 or U2 storage attributes do not affect
PPC440x5 core operation unless they are enabled using the MMUCR to perform these specific functions.
See Instruction and Data Caches on page 95 for more information on the mechanisms that can be controlled
by the U1 and U2 storage attributes.
The U0 and U3 storage attributes have no such mechanism that enables them to control any specific function
within the PPC440x5 core.
5.6.7 Supported Storage Attribute Combinations
Storage modes where both W = 1 and I = 1 (which would represent write-through but caching inhibited
storage) are not supported. For all supported combinations of the W and I storage attributes, the G, E, and
U0-U3 storage attributes may used in any combination.
5.7 Storage Control Registers
In addition to the two registers described below, the MSR[IS,DS] bits specify which of the two address spaces
the respective instruction or data storage accesses are directed towards. Also, the MSR[PR] bit is used by
the access control mechanism. See Machine State Register (MSR) on page 165 for more detailed informa-
tion on the MSR and the function of each of its bits.
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
5.7.1 Memory Management Unit Control Register (MMUCR)
The MMUCR is written from a GPR using mtspr, and can be read into a GPR using mfspr. In addition, the
MMUCR[STID] is updated with the TID field of the selected TLB entry when a tlbre instruction is executed.
Conversely, the TID field of the selected TLB entry is updated with the value of the MMUCR[STID] field when
a tlbwe instruction is executed. Other functions associated with the STID and other fields of the MMUCR are
described in more detail in the sections that follow.
The following figure illustrates the MMUCR.
STID
SWOA
U2SWOAE IULXE
0
6
7
8
9
10 11 12 13 14 15 16
DULXE
23 24
31
STS
U1TE
Figure 5-3. Memory Management Unit Control Register (MMUCR)
0:6
Reserved
Store Without Allocate
0 Cacheable store misses allocate a line in the
data cache.
7
SWOA
If MMUCR[U2SWOAE] = 1, this field is ignored.
1 Cacheable store misses do not allocate a line
in the data cache.
8
Reserved
U1 Transient Enable
0 Disable U1 storage attribute as transient
storage attribute.
9
U1TE
1 Enable U1 storage attribute as transient
storage attribute.
U2 Store without Allocate Enable
0 Disable U2 storage attribute control of store
without allocate.
1 Enable U2 storage attribute control of store
without allocate.
If MMUCR[U2SWOAE] = 1, the U2 storage
attribute overrides MMUCR[SWOA].
10
U2SWOAE
11
12
Reserved
Data Cache Unlock Exception Enable
0 Data cache unlock exception is disabled.
1 Data cache unlock exception is enabled.
dcbf in user mode will cause Cache Locking
exception type Data Storage interrupt when
MMUCR[DULXE] is 1.
DULXE
IULXE
Instruction Cache Unlock Exception Enable
0 Instruction cache unlock exception is disabled.
1 Instruction cache unlock exception is enabled.
icbi in user mode will cause Cache Locking
exception type Data Storage interrupt when
MMUCR[IULXE] is 1.
13
14
Reserved
Specifies the value of the translation space (TS)
field for the tlbsx[.] instruction
15
STS
Search Translation Space
Reserved
16:23
Specifies the value of the process identifier to be
compared against the TLB entry’s TID field for
the tlbsx[.] instruction; also used to transfer a
TLB entry’s TID value for the tlbre and tlbwe
instructions.
24:31
STID
Search Translation ID
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Store Without Allocate (SWOA) Field
Performance for certain applications can be affected by the allocation of cache lines on store misses. If the
store accesses for a particular application are distributed sparsely in memory, and if the data is typically not
re-used after having been stored, then performance may be improved by avoiding the latency and bus band-
width associated with filling the entire cache line containing the bytes being stored. On the other hand, if an
application typically stores to contiguous locations, or tends to store repeatedly to the same locations or to re-
access data after it has been stored, then performance would likely be improved by allocating the line in the
cache upon the first miss so that subsequent accesses will hit in the cache.
The SWOA field is one of two MMUCR fields which can control the allocation of cache lines upon store
misses. The other is the U2SWOAE field, and if U2SWOAE is 1 then the U2 storage attribute controls the
U2SWOAE field is 0, then the SWOA field controls cache line allocation for all cacheable store misses.
Specifically, if a cacheable store access misses in the data cache, then if SWOA is 0, then the cache line will
be filled into the data cache, and the store data will be written into the cache (as well as to memory if the
associated memory page is also marked as write-through; see Write-Through (W) on page 145). Conversely,
if SWOA is 1, then cacheable store misses will not allocate the line in the data cache, and the store data will
be written to memory only, whether or not the write-through attribute is set.
See Instruction and Data Caches on page 95 for more information on cache line allocation on store misses.
U1 Transient Enable (U1TE) Field
When U1TE is 1, then the U1 storage attribute is enabled to control the transient mechanism of the instruc-
tion and data caches (see User-Definable (U0–U3) on page 147). If the U1 field of the TLB entry for the
memory page being accessed is 0, then the access will use the normal portion of the cache. If the U1 field is
1, then the transient portion of cache will be used.
If the U1TE field is 0, then the transient cache mechanism is disabled and all accesses use the normal portion
of the cache.
U2 Store Without Allocate Enable (U2SWOAE) Field
An explanation of the allocation of cache lines on store misses is provided in the section on the SWOA field
above. The U2SWOAE field is the other mechanism which can control such allocation. If U2SWOAE is 0,
then the SWOA field determines whether or not a cache line is allocated on a store miss.
When U2SWOAE is 1, then the U2 storage attribute is enabled to control the allocation on a memory page
basis, and the SWOA field is ignored (see User-Definable (U0–U3) on page 147). If the U2 field of the TLB
entry for the memory page containing the bytes being stored is 0, then the cache line will be allocated in the
data cache on a store miss. If the U2 field is 0, then the cache line will not be allocated.
Data Cache Unlock Exception Enable (DULXE) Field
The DULXE field can be used to force a Cache Locking exception type Data Storage interrupt to occur if a
dcbf instruction is executed in user mode (MSR[PR]=1). Since dcbf can be executed in user mode and since
it causes a cache line to be flushed from the data cache, it has the potential for allowing an application
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
program to remove a locked line from the cache. The locking and unlocking of cache lines is generally a
supervisor mode function, as the supervisor has access to the various mechanisms which control the cache
locking mechanism (e.g., the Data Cache Victim Limit (DVLIM) and Instruction Cache Victim Limit (IVLIM)
registers, and the MMUCR). Therefore, the DULXE field provides a means to prevent any dcbf instructions
executed while in user mode from flushing any cache lines.
Note that with the PPC440x5 core, the Cache Locking exception occurs independent of whether the target
line is truly locked or not. This behavior is necessary because the instruction execution pipeline is such that
the exception determination must be made before it is determined whether or not the target line is actually
locked (or whether it is even a hit).
Software at the Data Storage interrupt handler can determine whether the target line is locked, and if so
whether or not the application should be allowed to unlock it.
If DULXE is 0, or if dcbf is executed while in supervisor mode, then the instruction execution is allowed to
proceed and flush the target line, independent of whether it is locked or not.
Instruction Cache Unlock Exception Enable (IULXE) Field
The IULXE field can be used to force a Cache Locking exception type Data Storage interrupt to occur if an
icbi instruction is executed in user mode (MSR[PR]=1). Since icbi can be executed in user mode and since it
causes a cache line to be removed from the instruction cache, it has the potential for allowing an application
program to remove a locked line from the cache. The locking and unlocking of cache lines is generally a
supervisor mode function, as the supervisor has access to the various mechanisms which control the cache
locking mechanism (e.g., the DVLIM and IVLIM registers, and the MMUCR). Therefore, the IULXE field
provides a means to prevent any icbi instructions executed while in user mode from flushing any cache lines.
Note that with the PPC440x5 core, the Cache Locking exception occurs independent of whether the target
line is truly locked or not. This behavior is necessary because the instruction execution pipeline is such that
the exception determination must be made before it is determined whether or not the target line is actually
locked (or whether it is even a hit).
Software at the Data Storage interrupt handler can determine whether the target line is locked, and if so
whether or not the application should be allowed to unlock it.
If IULXE is 0, or if icbi is executed while in supervisor mode, then the instruction execution is allowed to
proceed and flush the target line, independent of whether it is locked or not.
Search Translation Space (STS) Field
The STS field is used by the tlbsx[.] instruction to designate the value against which the TS field of the TLB
entries is to be matched. For instruction fetch and data storage accesses, the TS field of the TLB entries is
compared with the MSR[IS] bit or the MSR[DS] bit, respectively. For tlbsx[.] however, the MMUCR[STS] field
is used, allowing the TLB to be searched for entries with a TS field which is references an address space
other than the one being used by the currently executing process.
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Search Translation ID (STID) Field
The STID field is used by the tlbsx[.] instruction to designate the process identifier value to be compared with
the TID field of the TLB entries. For instruction fetch and data storage accesses and cache management
operations, the TID field of the TLB entries is compared with the value in the PID register (see Process ID
(PID) on page 151). For tlbsx[.] however, the MMUCR[STID] field is used, allowing the TLB to be searched
for entries with a TID field which does not match the Process ID of the currently executing process.
The MMUCR[STID] field is also used to transfer the TLB entry’s TID field on tlbre and tlbwe instructions
which target TLB word 0, as there are not enough bits in the GPR used for transferring the other fields such
that it could hold this field as well.
matching process. Also see TLB Read/Write Instructions (tlbre/tlbwe) on page 153 for more information on
how the MMUCR[STID] field is used by these instructions.
5.7.2 Process ID (PID)
The Process ID (PID) is a 32-bit register, although only the lower 8 bits are defined in the PPC440x5 core.
The 8-bit PID value is used as a portion of the virtual address for accessing storage (see Virtual Address
whether or not the entry corresponds to a given virtual address. If an entry’s TID field is 0 (signifying that the
entry defines a “global” as opposed to “private” page), then the PID value is ignored when determining
whether the entry corresponds to a given virtual address. See TLB Match Process on page 139 for a more
detailed description of the use of the PID value in the TLB match process.
The PID is written from a GPR using mtspr, and can be read into a GPR using mfspr. The following figure
illustrates the PID.
0
23 24
31
Figure 5-4. Process ID (PID)
0:23
24:31
Reserved
Process ID
5.8 Shadow TLB Arrays
The PPC440x5 core implements two shadow TLB arrays, one for instruction fetches and one for data
accesses. These arrays “shadow” the value of a subset of the entries in the main, unified TLB (the UTLB in
the context of this discussion). The purpose of the shadow TLB arrays is to reduce the latency of the address
translation operation, and to avoid contention for the UTLB array between instruction fetches and data
accesses.
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
The instruction shadow TLB (ITLB) contains four entries, while the data shadow TLB (DTLB) contains eight.
There is no latency associated with accessing the shadow TLB arrays, and instruction execution continues in
a pipelined fashion as long as the requested address is found in the shadow TLB. If the requested address is
not found in the shadow TLB, the instruction fetch or data storage access is automatically stalled while the
address is looked up in the UTLB. If the address is found in the UTLB, the penalty associated with the miss in
the shadow array is three cycles. If the address is also a miss in the UTLB, then an Instruction or Data TLB
Miss exception is reported.
The replacement of entries in the shadow TLB’s is managed by hardware, in a round-robin fashion. Upon a
shadow TLB miss which leads to a UTLB hit, the hardware will automatically cast-out the oldest entry in the
shadow TLB and replace it with the new translation.
The hardware will also automatically invalidate all of the entries in both of the shadow TLB’s upon any context
synchronization (see Context Synchronization on page 82). Context synchronizing operations include the
following:
• Any interrupt (including Machine Check)
• Execution of isync
• Execution of rfi, rfci, or rfmci
• Execution of sc
Note that there are other “context changing” operations which do not cause automatic context synchroniza-
tion in the hardware. For example, execution of a tlbwe instruction changes the UTLB contents but does not
cause a context synchronization and thus does not invalidate or otherwise update the shadow TLB entries. In
order for changes to the entries in the UTLB (or to other address-related resources such as the PID) to be
reflected in the shadow TLB’s, software must ensure that a context synchronizing operation occurs prior to
any attempt to use any address associated with the updated UTLB entries (either the old or new contents of
those entries). By invalidating the shadow TLB arrays, a context synchronizing operation forces the hardware
to refresh the shadow TLB entries with the updated information in the UTLB as each memory page is
accessed.
Note: Of the items in the preceding list of shadow TLB invalidating operations, the Machine Check interrupt
is not architecturally required to be context synchronizing, and thus is not guaranteed to cause invalidation of
any shadow TLB arrays on implementations other than the PPC440x5. Consequently, software which is
intended to be portable to other implementations should not depend on this behavior, and should insert the
appropriate architecturally-defined context synchronizing operation as necessary for desired operation.
5.9 TLB Management Instructions
The processor does not imply any format for the page tables or the page table entries. Software has signifi-
cant flexibility in organizing the size, location, and format of the page table, and in implementing a custom
TLB entry replacement strategy. For example, software can “lock” TLB entries that correspond to frequently
used storage, so that those entries are never cast out of the TLB, and TLB Miss exceptions to those pages
never occur.
In order to enable software to manage the TLB, a set of TLB management instructions is implemented within
the PPC440x5 core. These instructions are described briefly in the sections which follow, and in detail in
handling of TLB-related exceptions. One such resource is Save/Restore Register 0 (SRR0), which provides
the exception-causing address for Instruction TLB Error and Instruction Storage interrupts. Another resource
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
is the Data Exception Address Register (DEAR), which provides the exception-causing address for Data TLB
Error and Data Storage interrupts. Finally, the Exception Syndrome Register (ESR) provides bits to differen-
All of the TLB management instructions are privileged, in order to prevent user mode programs from affecting
the address translation and access control mechanisms.
5.9.1 TLB Search Instruction (tlbsx[.])
The tlbsx[.] instruction can be used to locate an entry in the TLB which is associated with a particular virtual
address. This instruction forms an effective address for which the TLB is to be searched, in the same manner
by which data storage access instructions perform their address calculation, by adding the contents of regis-
ters RA (or the value 0 if RA=0) and RB together. The MMUCR[STID] and MMUCR[STS] fields then provide
the process ID and address space portions of the virtual address, respectively. Next, the TLB is searched for
this virtual address, with the searching process including the notion of disabling the comparison to the
process ID if the TID field of a given TLB entry is 0 (see TLB Match Process on page 139). Finally, the TLB
index of the matching entry is written into the target register (RT). This index value can then serve as the
source value for a subsequent tlbre or tlbwe instruction, to read or update the entry. If no matching entry is
found, then the target register contents are undefined.
The “record form” of the instruction (tlbsx.) updates CR[CR0] with the result of the search: if a match is
2
found, then CR[CR0] is set to 1; otherwise it is set to 0.
2
When the TLB is searched using a tlbsx instruction, if a matching entry is found, the parity calculated for the
tag is compared to the parity stored in the TPAR field. A mismatch causes a parity error exception. Parity
errors in words 1 and 2 of the entry will not cause parity error exceptions when executing a tlbsx instruction.
5.9.2 TLB Read/Write Instructions (tlbre/tlbwe)
TLB entries can be read and written by the tlbre and tlbwe instructions, respectively. Since a TLB entry
contains more than 32 bits, multiple tlbre/tlbwe instructions must be executed in order to transfer all of the
TLB entry information. A TLB entry is divided into three portions, TLB word 0, TLB word 1, and TLB word 2.
The RA field of the tlbre and tlbwe instructions designates a GPR from which the low-order six bits are used
to specify the TLB index of the TLB entry to be read or written. An immediate field (WS) designates which
word of the TLB entry is to be transferred (that is, WS=0 specifies TLB word 0, and so on). Finally, the
contents of the selected TLB word are transferred to or from a designated target or source GPR (and the
MMUCR[STID] field, for TLB word 0; see below), respectively.
The fields in each TLB word are illustrated in Figure 5-5. The bit numbers indicate which bits of the
target/source GPR correspond to each TLB field. Note that the TID field of TLB word 0 is transferred to/from
the MMUCR[STID] field, rather than to/from the target/source GPR.
When executing a tlbre, the parity fields (TPAR, PAR1, and PAR2) are loaded if and only if the CCR0[CRPE]
bit is set. Otherwise those fields are loaded with zeros. When the tlbre is executed, If the parity bits stored for
the particular word that is read by the tlbre indicate a parity error, the parity error exception will be generated
regardless of the state of the CCR0[CRPE] bit.
When executing a tlbwe, bits in the source GPR that correspond to the parity fields are ignored, as the hard-
ware calculates the parity to be recorded in those fields of the entry.
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
TLB Word 0 (WS=0)
0
21 22 23 24
27 28 31 32
TPAR
39
EPN
RPN
V
TS SIZE
TID
TLB Word 1 (WS=1)
24
0
21 22 23
PAR1
27 28
31
ERPN
TLB Word 2 (WS=2)
0
1
2
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
U0 U1 U2 U3
31
14
0
PAR2
W
I
M
G
E
UX UW UR SX SW SR
Figure 5-5. TLB Entry Word Definitions
5.9.3 TLB Sync Instruction (tlbsync)
The tlbsync instruction is used to synchronize software TLB management operations in a multiprocessor
environment with hardware-enforced coherency, which is not supported by the PPC440x5. Consequently,
this instruction is treated as a no-op. It is provided in support of software compatibility between PowerPC-
based systems.
5.10 Page Reference and Change Status Management
When performing page management, it is useful to know whether a given memory page has been refer-
enced, and whether its contents have been modified. Note that this may be more involved than determining
whether a given TLB entry has been used to reference or change memory, since multiple TLB entries may
translate to the same memory page. If it is necessary to replace the contents of some memory page with
other contents, a page which has been referenced (accessed for any purpose) is more likely to be maintained
than a page which has never been referenced. If the contents of a given memory page are to be replaced and
the contents of that page have been changed, then the current contents of that page must be written to
backup physical storage (such as a hard disk) before replacement.
Similarly, when performing TLB management, it is useful to know whether a given TLB entry has been refer-
enced. When making a decision about which entry of the TLB to replace in order to make room for a new
entry, an entry which has never been referenced is a more likely candidate to be replaced.
The PPC440x5 does not automatically record references or changes to a page or TLB entry. Instead, the
interrupt mechanism may be used by system software to maintain reference and change information for TLB
entries and their associated pages, respectively.
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Execute, Read and Write Access Control exceptions may be used to allow software to maintain reference
and change information for a TLB entry and for its associated memory page. The following description
explains one way in which system software can maintain such reference and change information.
The TLB entry is originally written into the TLB with its access control bits (UX, SX, UR, SR, UW, and SW) off.
The first attempt of application code to use the page will therefore cause an Access Control exception and a
corresponding Instruction or Data Storage interrupt. The interrupt handler records the reference to the TLB
entry and to the associated memory page in a software table, and then turns on the appropriate access
control bit, thereby indicating that the particular TLB entry has been referenced. An initial read from the page
is handled by only turning on the appropriate UR or SR access control bit, leaving the page “read-only”.
Subsequent read accesses to the page via that TLB entry will proceed normally.
If a write access is later attempted, a Write Access Control exception type Data Storage interrupt will occur.
The interrupt handler records the change status to the memory page in a software table, and then turns on
the appropriate UW or SW access control bit, thereby indicating that the memory page associated with the
particular TLB entry has been changed. Subsequent write accesses to the page via that TLB entry will
proceed normally.
5.11 TLB Parity Operations
The TLB is parity protected against soft errors in the TLB memory array that are caused by alpha particle
impacts. If such errors are detected, the CPU can be configured to vector to the machine check interrupt
handler, which can restore the corrupted state of the TLB from the page tables in system memory.
The TLB is a 64-entry CAM/RAM with 40 tag bits, 41 data bits, and 8 parity bits per entry. Tag and data bits
are parity protected with four parity bits for the 40-bit tag, two parity bits for 26 bits of data (i.e. those read and
written as word 1 by the tlbre and tlbwe instructions), and two more parity bits for the remaining 15 bits of
data (i.e. word 2). The parity bits are stored in the TLB entries in fields named TPAR, PAR1, and PAR2,
respectively. See Figure 5-5 TLB Entry Word Definitions
Unlike the instruction and data cache CAM/RAMs, the TLB does not detect multiple hits due to parity errors in
the tags. The TLB is a relatively small memory array, and the reduction in Soft Error Rate (SER) provided by
adding multi-hit detection to the circuit is small, and so, not worth the expense of the feature.
TLB parity bits are set any time the TLB is updated, which is always done via a tlbwe instruction. TLB parity
is checked each time the TLB is searched or read, whether to re-fill the ITLB or DTLB, or as a result of a
tlbsx or tlbre instruction. When executing an ITLB or DTLB refill, parity is checked for the tag and both data
words. When executing a tlbsx, data output is not enabled for the translation and protection outputs of the
TLB, so only the tag parity is checked. When executing a tlbre, parity is checked only for the word specified
in the WS field of the tlbre instruction. Detection of a parity error causes a machine check exception. If
MSR[ME] is set (which is the usual case), the processor takes a machine check interrupt.
5.11.1 Reading TLB Parity Bits with tlbre
If CCR0[CRPE] is set, execution of a tlbre instruction updates the target register with parity values as well as
the tag or other data from the TLB. However, since a tlbre that detects a parity error will cause a machine
check exception, the target register can only be updated with a “bad” parity value if the MSR[ME] bit is
cleared, preventing the machine check interrupt. Thus the usual flow of code that detects a parity error in the
TLB and then finds out which entry is erroneous would proceed as:
1. A tlbre instruction is executed from normal OS code, resulting in a parity exception. The exception sets
MCSR[TLBE] and MCSR[MCS].
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
2. MSR[ME] = 1, so the CPU vectors to the machine check handler (i.e takes the machine check interrupt)
and resets the MSR[ME] bit. Note that even though the parity error causes an asynchronous interrupt,
that interrupt is guaranteed to be taken before the tlbre instruction completes if the CCR0[PRE] (Parity
Recoverability Enable) is set, and so the target register (RT) of the tlbre will not be updated.
3. The Machine Check handler code includes a series of tlbre instructions to query the state of the TLB and
find the erroneous entry. When a tlbre encounters an erroneous entry and MSR[ME] = 0, the parity
exception still happens, setting the MCSR[MCS] and MCSR[TLBE] bits. Additionally, since MSR[ME] = 0,
MCSR[IMCE] is set, indicating that an imprecise machine check was detected. Finally, the instruction
completes, (since no interrupt is taken because MSR[ME} = 0), updating the target register with data from
the TLB, including the parity information.
Note that the tlbre causes an exception when it detects a parity error, but the icread and dcread instruc-
tions do not. This inconsistency is explained because OS code commonly uses a sequence of tlbsx and
tlbre instructions to update the “changed” bit in the page table entries. (See section 5.10, “Page Reference
and Change Status Management.”) Forcing the software to check the parity manually for each tlbre would be
a performance limitation. No such functional use exists for the icread and dcread instructions; they are used
only in debugging contexts with no significant performance requirements.
As is the case for any machine check interrupt, after vectoring to the machine check handler, the MCSRR0
contains the value of the oldest “uncommitted” instruction in the pipeline at the time of the exception and
MCSRR1 contains the old (MSR) context. The interrupt handler is able to query Machine Check Status
Register (MCSR) to find out that it was called due to a TLB parity exception, and then use tlbre instructions
to find the error in the TLB and restore it from a known good copy in main memory.
Note: A parity error on the TLB entry which maps the machine check exception handler code
prevents recovery. In effect, one of the 64 TLB entries is unprotected, in that the machine
cannot recover from an error in that entry. It is possible to add logic to get around this problem,
but the reduction in SER achieved by protecting 63 out of 64 TLB entries is sufficient. Further,
the software technique of simply dedicating a TLB entry to the page that contains the machine
check handler and periodically refreshing that entry from a known good copy can reduce the
probability that the entry will be used with a parity error to near zero.
As mentioned above, any tlbre or tlbsx instruction that causes a machine check interrupt will be flushed
from the pipeline before it completes. Further, any instruction that causes a DTLB or ITLB refill which causes
a TLB parity error will be flushed before it completes.
5.11.2 Simulating TLB Parity Errors for Software Testing
Because parity errors occur in the TLB infrequently and unpredictably, it is desirable to provide users with a
way to simulate the effect of a TLB parity error so that interrupt handling software may be exercised. This is
exactly the purpose of the 4-bit CCR1[MMUPEI] field.
Usually, parity is calculated as the even parity for each set of bits to be protected, which the checking hard-
ware expects. This calculation is done as the TLB data is stored with a tlbwe instruction. However, if any of
the the CCR1[MMUPEI] bits are set, the calculated parity for the corresponding bits of the data being stored
are inverted and stored as odd parity. Then, when the data stored with odd parity is subsequently used to
refill the DTLB or ITLB, or by a tlbsx or tlbre instruction, it will cause a Parity exception type Machine Check
interrupt and exercise the interrupt handling software. The following pseudo-code is an example of how to
use the CCR1[MMUPEI] field to simulate a parity error on a TLB entry:
mtspr CCR1, Rx
isync
tlbwe Rs,Ra,0
tlbwe Rs,Ra,1
; Set some CCR1[MMUPEI] bits
; wait for the CCR1 context to update
; write some data to the TLB with bad parity
; write some data to the TLB with bad parity
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
tlbwe Rs,Ra,2
isync
mtspr CCR1, Rz
; write some data to the TLB with bad parity
; wait for the tlbwe(s) to finish
; Reset CCR1[MMUPEI]
isync
tlbre RT,RA,WS
; wait for the CCR1 context to update
; tlbre with bad parity causes interrupt
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
mmu.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
6. Interrupts and Exceptions
Interrupt Processing on page 162 explains in general how interrupts are processed, including the require-
ments for partial execution of instructions.
Several registers support interrupt handling and control. Interrupt Processing Registers on page 165
describes these registers.
Table 6-2 Interrupt and Exception Types on page 175 lists the interrupts and exceptions handled by the
PPC440x5, in the order of Interrupt Vector Offset Register (IVOR) usage. Detailed descriptions of each inter-
rupt type follow, in the same order.
Finally, Interrupt Ordering and Masking on page 199 and Exception Priorities on page 202 define the priority
order for the processing of simultaneous interrupts and exceptions.
6.1 Overview
An interrupt is the action in which the processor saves its old context (Machine State Register (MSR) and
next instruction address) and begins execution at a pre-determined interrupt-handler address, with a modified
MSR. Exceptions are the events that may cause the processor to take an interrupt, if the corresponding inter-
rupt type is enabled.
Exceptions may be generated by the execution of instructions, or by signals from devices external to the
PPC440x5 core, the internal timer facilities, debug events, or error conditions.
6.2 Interrupt Classes
All interrupts, except for Machine Check, can be categorized according to two independent characteristics of
the interrupt:
• Asynchronous or synchronous
• Critical or non-critical
6.2.1 Asynchronous Interrupts
Asynchronous interrupts are caused by events that are independent of instruction execution. For asynchro-
nous interrupts, the address reported to the interrupt handling routine is the address of the instruction that
would have executed next, had the asynchronous interrupt not occurred.
6.2.2 Synchronous Interrupts
Synchronous interrupts are those that are caused directly by the execution (or attempted execution) of
instructions, and are further divided into two classes, precise and imprecise.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Synchronous, precise interrupts are those that precisely indicate the address of the instruction causing the
exception that generated the interrupt; or, for certain synchronous, precise interrupt types, the address of the
immediately following instruction.
Synchronous, imprecise interrupts are those that may indicate the address of the instruction which caused
the exception that generated the interrupt, or the address of some instruction after the one which caused the
exception.
6.2.2.1 Synchronous, Precise Interrupts
When the execution or attempted execution of an instruction causes a synchronous, precise interrupt, the
following conditions exist when the associated interrupt handler begins execution:
• SRR0 (see Save/Restore Register 0 (SRR0) on page 167) or CSRR0 (see Critical Save/Restore Register
0 (CSRR0) on page 168) addresses either the instruction which caused the exception that generated the
interrupt, or the instruction immediately following this instruction. Which instruction is addressed can be
determined from a combination of the interrupt type and the setting of certain fields of the ESR (see
• The interrupt is generated such that all instructions preceding the instruction which caused the exception
appear to have completed with respect to the executing processor. However, some storage accesses
associated with these preceding instructions may not have been performed with respect to other proces-
sors and mechanisms.
• The instruction which caused the exception may appear not to have begun execution (except for having
caused the exception), may have been partially executed, or may have completed, depending on the
• Architecturally, no instruction beyond the one which caused the exception has executed.
6.2.2.2 Synchronous, Imprecise Interrupts
When the execution or attempted execution of an instruction causes a synchronous, imprecise interrupt, the
following conditions exist when the associated interrupt handler begins execution:
• SRR0 or CSRR0 addresses either the instruction which caused the exception that generated the inter-
rupt, or some instruction following this instruction.
• The interrupt is generated such that all instructions preceding the instruction addressed by SRR0 or
CSRR0 appear to have completed with respect to the executing processor.
• If the imprecise interrupt is forced by the context synchronizing mechanism, due to an instruction that
causes another exception that generates an interrupt (for example, Alignment, Data Storage), then SRR0
addresses the interrupt-forcing instruction, and the interrupt-forcing instruction may have been partially
• If the imprecise interrupt is forced by the execution synchronizing mechanism, due to executing an execu-
tion synchronizing instruction other than msync or isync, then SRR0 or CSRR0 addresses the interrupt-
forcing instruction, and the interrupt-forcing instruction appears not to have begun execution (except for
its forcing the imprecise interrupt). If the imprecise interrupt is forced by an msync or isync instruction,
then SRR0 or CSRR0 may address either the msync or isync instruction, or the following instruction.
• If the imprecise interrupt is not forced by either the context synchronizing mechanism or the execution
synchronizing mechanism, then the instruction addressed by SRR0 or CSRR0 may have been partially
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
• No instruction following the instruction addressed by SRR0 or CSRR0 has executed.
The only synchronous, imprecise interrupts in the PPC440x5 core are the “special cases” of “delayed” inter-
rupts, which can result when certain kinds of exceptions occur while the corresponding interrupt type is
disabled. The first of these is the Floating-Point Enabled exception type Program interrupt. For this type of
interrupt to occur, a floating-point unit must be attached to the auxiliary processor interface of the PPC440x5
core, and the Floating-point Enabled Exception Summary bit of the Floating-Point Status and Control
Register (FPSCR[FEX]) must be set while Floating-point Enabled exception type Program interrupts are
disabled due to MSR[FE0,FE1] both being 0. If and when such interrupts are subsequently enabled, by
setting one or the other (or both) of MSR[FE0,FE1] to 1 while FPSCR[FEX] is still 1, then a synchronous,
imprecise form of Floating-Point Enabled exception type Program interrupt will occur, and SRR0 will be set to
the address of the instruction which would have executed next (that is, the instruction after the one which
updated MSR[FE0,FE1]). If the MSR was updated by an rfi, rfci, or rfmci instruction, then SRR0 will be set to
the address to which the rfi, rfci, or rfmci was returning, and not to the instruction address which is sequen-
tially after the rfi, rfci, or rfmci.
The second type of delayed interrupt which may be handled as a synchronous, imprecise interrupt is the
Debug interrupt. Similar to the Floating-Point Enabled exception type Program interrupt, the Debug interrupt
can be temporarily disabled by an MSR bit, MSR[DE]. Accordingly, certain kinds of Debug exceptions may
occur and be recorded in the DBSR while MSR[DE] is 0, and later lead to a delayed Debug interrupt if
MSR[DE] is set to 1 while a Debug exception is still set in the DBSR. If and when this occurs, the interrupt will
either be synchronous and imprecise, or it will be asynchronous, depending on the type of Debug exception
causing the interrupt. In either case, CSRR0 is set to the address of the instruction which would have
executed next (that is, the instruction after the one which set MSR[DE] to 1). If MSR[DE] is set to 1 by rfi, rfci,
or rfmci, then CSRR0 is set to the address to which the rfi, rfci, or rfmci was returning, and not to the
address of the instruction which was sequentially after the rfi, rfci, or rfmci.
Besides these special cases of Program and Debug interrupts, all other synchronous interrupts are handled
precisely by the PPC440x5 core, including FP Enabled exception type Program interrupts even when the
processor is operating in one of the architecturally-defined imprecise modes (MSR[FE0,FE1] = 0b01 or
0b10).
See Program Interrupt on page 187 and Debug Interrupt on page 195 for a more detailed description of these
interrupt types, including both the precise and imprecise cases.
6.2.3 Critical and Non-Critical Interrupts
Interrupts can also be classified as critical or noncritical interrupts. Certain interrupt types demand immediate
attention, even if other interrupt types are currently being processed and have not yet had the opportunity to
save the state of the machine (that is, return address and captured state of the MSR). To enable taking a crit-
ical interrupt immediately after a non-critical interrupt has occurred (that is, before the state of the machine
has been saved), two sets of Save/Restore Register pairs are provided. Critical interrupts use the
Save/Restore Register pair CSRR0/CSRR1. Non-Critical interrupts use Save/Restore Register pair
SRR0/SRR1.
6.2.4 Machine Check Interrupts
Machine Check interrupts are a special case. They are typically caused by some kind of hardware or storage
subsystem failure, or by an attempt to access an invalid address. A Machine Check may be caused indirectly
by the execution of an instruction, but not be recognized and/or reported until long after the processor has
executed past the instruction that caused the Machine Check. As such, Machine Check interrupts cannot
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
properly be classified as either synchronous or asynchronous, nor as precise or imprecise. They also do not
belong to either the critical or the non-critical interrupt class, but instead have associated with them a unique
pair of save/restore registers, Machine Check Save/Restore Registers 0/1 (MCSRR0/1).
Architecturally, the following general rules apply for Machine Check interrupts:
1. No instruction after the one whose address is reported to the Machine Check interrupt handler in
MCSRR0 has begun execution.
2. The instruction whose address is reported to the Machine Check interrupt handler in MCSRR0, and all
prior instructions, may or may not have completed successfully. All those instructions that are ever going
to complete appear to have done so already, and have done so within the context existing prior to the
Machine Check interrupt. No further interrupt (other than possible additional Machine Check interrupts)
will occur as a result of those instructions.
With the PPC440x5, Machine Check interrupts can be caused by Machine Check exceptions on a memory
access for an instruction fetch, for a data access, or for a TLB access. Some of the interrupts generated
behave as synchronous, precise interrupts, while other are handled in an asynchronous fashion.
In the case of an Instruction Synchronous Machine Check exception, the PPC440x5 will handle the interrupt
as a synchronous, precise interrupt, assuming Machine Check interrupts are enabled (MSR[ME] = 1). That is,
if a Machine Check exception is detected during an instruction fetch, the exception will not be reported to the
interrupt mechanism unless and until execution is attempted for the instruction address at which the Machine
Check exception occurred. If, for example, the direction of the instruction stream is changed (perhaps due to
a branch instruction), such that the instruction at the address associated with the Machine Check exception
will not be executed, then the exception will not be reported and no interrupt will occur. If and when an
Instruction Machine Check exception is reported, and if Machine Check interrupts are enabled at the time of
the reporting of the exception, then the interrupt will be synchronous and precise and MCSRR0 will be set to
the instruction address which led to the exception. If Machine Check interrupts are not enabled at the time of
the reporting of an Instruction Machine Check exception, then a Machine Check interrupt will not be gener-
ated (ever, even if and when MSR[ME] is subsequently set to 1), although the ESR[MCI] field will be set to 1
to indicate that the exception has occurred and that the instruction associated with the exception has been
executed.
Instruction Asynchronous Machine Check, Data Asynchronous Machine Check, and TLB Asynchronous
Machine Check exceptions, on the other hand, are handled in an “asynchronous” fashion. That is,the address
reported in MCSRR0 may not be related to the instruction which prompted the access which led , directly or
indirectly, to the Machine Check exception. The address may be that of an instruction before or after the
exception-causing instruction, or it may reference the exception causing instruction, depending on the nature
of the access, the type of error encountered , and the circumstances of the instruction’s execution within the
processor pipeline. If MSR[ME] is 0 at the time of a Machine Check exception that is handled in this asyn-
chronous way, a Machine Check interrupt will subsequently occur if and when MSR[ME] is set to 1.
6.3 Interrupt Processing
Associated with each kind of interrupt is an interrupt vector, that is, the address of the initial instruction that is
executed when the corresponding interrupt occurs.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Interrupt processing consists of saving a small part of the processor state in certain registers, identifying the
cause of the interrupt in another register, and continuing execution at the corresponding interrupt vector loca-
tion. When an exception exists and the corresponding interrupt type is enabled, the following actions are
performed, in order:
1. SRR0 (for non-critical class interrupts) or CSRR0 (for critical class interrupts) or MCSRR0 (for Machine
Check interrupts) is loaded with an instruction address that depends on the type of interrupt; see the spe-
cific interrupt description for details.
2. The ESR is loaded with information specific to the exception type. Note that many interrupt types can only
be caused by a single type of exception, and thus do not need nor use an ESR setting to indicate the
cause of the interrupt. Machine Check interrupts load the MCSR
3. SRR1 (for non-critical class interrupts) or CSRR1 (for critical class interrupts) or MCSRR1 (for Machine
Check interrupts) is loaded with a copy of the contents of the MSR.
4. The MSR is updated as described below. The new values take effect beginning with the first instruction
following the interrupt.
• MSR[WE,EE,PR,FP,FE0,DWE,FE1,IS,DS] are set to 0 by all interrupts.
• MSR[CE,DE] are set to 0 by all critical class interrupts and left unchanged by all non-critical class
interrupts.
• MSR[ME] is set to 0 by Machine Check interrupts and left unchanged by all other interrupts.
5. Instruction fetching and execution resumes, using the new MSR value, at the interrupt vector address,
which is specific to the interrupt type, and is determined as follows:
IVPR
|| IVORn
|| 0b0000
0:15
16:27
where n specifies the IVOR register to be used for a particular interrupt type (see Interrupt Vector Offset
At the end of a non-critical interrupt handling routine, execution of an rfi causes the MSR to be restored from
the contents of SRR1 and instruction execution to resume at the address contained in SRR0. Likewise,
execution of an rfci performs the same function at the end of a critical interrupt handling routine, using
CSRR0 instead of SRR0 and CSRR1 instead of SRR1. rfmci uses MCSRR0 and MCSRR1 in the same
manner.
Programming Note: In general, at process switch, due to possible process interlocks and
possible data availability requirements, the operating system needs to
consider executing the following instructions.
• stwcx., to clear the reservation if one is outstanding, to ensure that a lwarx
in the “old” process is not paired with a stwcx. in the “new” process. See
Set” for more information on storage reservations.
• msync, to ensure that all storage operations of an interrupted process are
complete with respect to other processors before that process begins
executing on another processor.
• isync, rfi, rfci, or rfmci, to ensure that the instructions in the “new” process
execute in the “new” context.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
6.3.1 Partially Executed Instructions
In general, the architecture permits load and store instructions to be partially executed, interrupted, and then
to be restarted from the beginning upon return from the interrupt. In order to guarantee that a particular load
or store instruction will complete without being interrupted and restarted, software must mark the storage
being referred to as Guarded, and must use an elementary (not a string or multiple) load or store that is
aligned on an operand-sized boundary.
In order to guarantee that load and store instructions can, in general, be restarted and completed correctly
without software intervention, the following rules apply when an instruction is partially executed and then
interrupted:
• For an elementary load, no part of the target register (GPR(RT), FPR(FRT), or auxiliary processor regis-
ter) will have been altered.
• For the “update” forms of load and store instructions, the update register, GPR(RA), will not have been
altered.
On the other hand, the following effects are permissible when certain instructions are partially executed and
then restarted:
• For any store instruction, some of the bytes at the addressed storage location may have been accessed
and/or updated (if write access to that page in which bytes were altered is permitted by the access control
mechanism). In addition, for the stwcx. instruction, if the address is not aligned on a word boundary, then
the value in CR[CR0] is undefined, as is whether or not the reservation (if one existed) has been cleared.
• For any load, some of the bytes at the addressed storage location may have been accessed (if read
access to that page in which bytes were accessed is permitted by the access control mechanism). In
addition, for the lwarx instruction, if the address is not aligned on a word boundary, it is undefined
whether or not a reservation has been set.
• For load multiple and load string instructions, some of the registers in the range to be loaded may have
been altered. Including the addressing registers (GPR(RA), and possibly GPR(RB)) in the range to be
loaded is an invalid form of these instructions (and a programming error), and thus the rules for partial
execution do not protect against overwriting of these registers. Such possible overwriting of the address-
ing registers makes these invalid forms of load multiple and load strings inherently non-restartable.
In no case will access control be violated.
As previously stated, the only load or store instructions that are guaranteed to not be interrupted after being
partially executed are elementary, aligned, guarded loads and stores. All others may be interrupted after
being partially executed. The following list identifies the specific instruction types for which interruption after
partial execution may occur, as well as the specific interrupt types that could cause the interruption:
1. Any load or store (except elementary, aligned, guarded):
Critical Input
Machine Check
External Input
Program (Imprecise Mode Floating-Point Enabled)
Note that this type of interrupt can lead to partial execution of a load or store instruction under the
architectural definition only; the PPC440x5 core handles the imprecise modes of the Floating-Point
Enabled exceptions precisely, and hence this type of interrupt will not lead to partial execution.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Decrementer
Fixed-Interval Timer
Watchdog Timer
Debug (Unconditional Debug Event)
2. Unaligned elementary load or store, or any load or store multiple or string:
All of the above listed under item 1, plus the following:
Alignment
Data Storage (if the access crosses a memory page boundary)
Debug (Data Address Compare, Data Value Compare)
6.4 Interrupt Processing Registers
The interrupt processing registers include the Save/Restore Registers (SRR0–SRR1), Critical Save/Restore
Registers (CSRR0–CSRR1), Data Exception Address Register (DEAR), Interrupt Vector Offset Registers
(IVOR0–IVOR15), Interrupt Vector Prefix Register (IVPR), and Exception Syndrome Register (ESR). Also
described in this section is the Machine State Register (MSR), which belongs to the category of processor
control registers.
6.4.1 Machine State Register (MSR)
The MSR is a register of its own unique type that controls important chip functions, such as the enabling or
disabling of various interrupt types.
The MSR can be written from a GPR using the mtmsr instruction. The contents of the MSR can be read into
a GPR using the mfmsr instruction. The MSR[EE] bit can be set or cleared atomically using the wrtee or
wrteei instructions. The MSR contents are also automatically saved, altered, and restored by the interrupt-
handling mechanism.
EE
FP
WE
FE0
DE
IS
0
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
31
FE1
CE
ME
DWE
DS
PR
Figure 6-1. Machine State Register (MSR)
0:12
Reserved
Wait State Enable
0 The processor is not in the wait state.
1 The processor is in the wait state.
If MSR[WE] = 1, the processor remains in the wait
state until an interrupt is taken, a reset occurs, or
an external debug tool clears WE.
13
WE
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Critical Interrupt Enable
0 Critical Input and Watchdog Timer interrupts are
disabled.
14
15
16
CE
1 Critical Input and Watchdog Timer interrupts are
enabled.
Reserved
External Interrupt Enable
0 External Input, Decrementer, and Fixed Interval
Timer interrupts are disabled.
1 External Input, Decrementer, and Fixed Interval
Timer interrupts are enabled.
EE
PR
Problem State
0 Supervisor state (privileged instructions can be
executed)
17
1 Problem state (privileged instructions can not be
executed)
Floating Point Available
0 The processor cannot execute floating-point
instructions
1 The processor can execute floating-point
instructions
18
19
20
FP
Machine Check Enable
0 Machine Check interrupts are disabled
1 Machine Check interrupts are enabled.
ME
FE0
Floating-point exception mode 0
0 If MSR[FE1] = 0, ignore exceptions mode; if
MSR[FE1] = 1, imprecise nonrecoverable mode
1 If MSR[FE1] = 0, imprecise recoverable mode; if
MSR[FE1] = 1, precise mode
Debug Wait Enable
0 Disable debug wait mode.
1 Enable debug wait mode.
21
22
DWE
DE
Debug interrupt Enable
0 Debug interrupts are disabled.
1 Debug interrupts are enabled.
Floating-point exception mode 1
0 If MSR[FE0] = 0, ignore exceptions mode; if
MSR[FE0] = 1, imprecise recoverable mode
1 If MSR[FE0] = 0, imprecise non-recoverable
mode; if MSR[FE0] = 1, precise mode
23
FE1
24:25
Reserved
Instruction Address Space
0 All instruction storage accesses are directed to
address space 0 (TS = 0 in the relevant TLB
entry).
1 All instruction storage accesses are directed to
address space 1 (TS = 1 in the relevant TLB
entry).
26
27
IS
Data Address Space
0 All data storage accesses are directed to
address space 0 (TS = 0 in the relevant TLB
entry).
DS
1 All data storage accesses are directed to
address space 1 (TS = 1 in the relevant TLB
entry).
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
28:31
Reserved
6.4.2 Save/Restore Register 0 (SRR0)
SRR0 is an SPR that is used to save machine state on non-critical interrupts, and to restore machine state
when an rfi is executed. When a non-critical interrupt occurs, SRR0 is set to an address associated with the
process which was executing at the time. When rfi is executed, instruction execution returns to the address in
SRR0.
In general, SRR0 contains the address of the instruction that caused the non-critical interrupt, or the address
of the instruction to return to after a non-critical interrupt is serviced. See the individual descriptions under
Interrupt Definitions on page 175 for an explanation of the precise address recorded in SRR0 for each non-
critical interrupt type.
SRR0 can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
0
29 30 31
Figure 6-2. Save/Restore Register 0 (SRR0)
0:29
30:31
Return address for non-critical interrupts
Reserved
6.4.3 Save/Restore Register 1 (SRR1)
SRR1 is an SPR that is used to save machine state on non-critical interrupts, and to restore machine state
when an rfi is executed. When a non-critical interrupt is taken, the contents of the MSR (prior to the MSR
being cleared by the interrupt) are placed into SRR1. When rfi is executed, the MSR is restored with the
contents of SRR1.
Bits of SRR1 that correspond to reserved bits in the MSR are also reserved.
Programming Note: An MSR bit that is reserved may be altered by rfi, consistent with the
value being restored from SRR1.
SRR1 can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
IS
EE
DE
FP
WE
FE0
0
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
FE1
31
CE
ME
DWE
DS
PR
Figure 6-3. Save/Restore Register 1 (SRR1)
Copy of the MSR at the time of a non-critical inter-
rupt.
0:31
6.4.4 Critical Save/Restore Register 0 (CSRR0)
CSRR0 is an SPR that is used to save machine state on critical interrupts, and to restore machine state when
an rfci is executed. When a critical interrupt occurs, CSRR0 is set to an address associated with the process
which was executing at the time. When rfci is executed, instruction execution returns to the address in
CSRR0.
In general, CSRR0 contains the address of the instruction that caused the critical interrupt, or the address of
the instruction to return to after a critical interrupt is serviced. See the individual descriptions under Interrupt
Definitions on page 175 for an explanation of the precise address recorded in CSRR0 for each critical inter-
rupt type.
CSRR0 can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
0
29 30 31
Figure 6-4. Critical Save/Restore Register 0 (CSRR0)
0:29
30:31
Return address for critical interrupts
Reserved
6.4.5 Critical Save/Restore Register 1 (CSRR1)
CSRR1 is an SPR that is used to save machine state on critical interrupts, and to restore machine state when
an rfci is executed. When a critical interrupt is taken, the contents of the MSR (prior to the MSR being
cleared by the interrupt) are placed into CSRR1. When rfci is executed, the MSR is restored with the
contents of CSRR1.
Bits of CSRR1 that correspond to reserved bits in the MSR are also reserved.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Programming Note: An MSR bit that is reserved may be altered by rfci, consistent with the
value being restored from CSRR1.
CSRR1 can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
EE
FP
WE
FE0
DE
IS
0
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
FE1
31
CE
ME
DWE
DS
PR
Figure 6-5. Critical Save/Restore Register 1 (CSRR1)
0:31
Copy of the MSR when a critical interrupt is taken
6.4.6 Machine Check Save/Restore Register 0 (MCSRR0)
MCSRR0 is an SPR that is used to save machine state on Machine Check interrupts, and to restore machine
state when an rfmci is executed. When a machine check interrupt occurs, MCSRR0 is set to an address
associated with the process which was executing at the time. When rfmci is executed, instruction execution
returns to the address in MCSRR0.
In general, MCSRR0 contains the address of the instruction that caused the Machine Check interrupt, or the
address of the instruction to return to after a machine check interrupt is serviced. See the individual descrip-
tions under Interrupt Definitions on page 175 for an explanation of the precise address recorded in MCSRR0
for each Machine Check interrupt type.
MCSRR0 can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
0
29 30 31
Figure 6-6. Machine Check Save/Restore Register 0 (MCSRR0)
0:29
Return address for machine check interrupts
Reserved
30:31
6.4.7 Machine Check Save/Restore Register 1 (MCSRR1)
MCSRR1 is an SPR that is used to save machine state on Machine Check interrupts, and to restore machine
state when an rfmci is executed. When a machine check interrupt is taken, the contents of the MSR (prior to
the MSR being cleared by the interrupt) are placed into MCSRR1. When rfmci is executed, the MSR is
restored with the contents of MCSRR1.
Bits of MCSRR1 that correspond to reserved bits in the MSR are also reserved.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Programming Note: An MSR bit that is reserved may be altered by rfmci, consistent with
the value being restored from MCSRR1.
MCSRR1 can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
IS
EE
DE
FP
WE
FE0
0
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
31
FE1
CE
ME
DWE
DS
PR
Figure 0-1. Machine Check Save/Restore Register 1 (MCSRR1)
0:31
Copy of the MSR at the time of a machine check interrupt.
6.4.8 Data Exception Address Register (DEAR)
The DEAR contains the address that was referenced by a load, store, or cache management instruction that
caused an Alignment, Data TLB Miss, or Data Storage exception.
The DEAR can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
0
31
Figure 6-7. Data Exception Address Register (DEAR)
Address of data exception for Data Storage, Align-
ment, and Data TLB Error interrupts
0:31
6.4.9 Interrupt Vector Offset Registers (IVOR0–IVOR15)
An IVOR specifies the quadword (16 byte)-aligned interrupt vector offset from the base address provided by
the IVPR (see Interrupt Vector Prefix Register (IVPR) on page 171) for its respective interrupt type. IVOR0–
IVOR15 are provided for the defined interrupt types. The interrupt vector effective address is formed as
follows:
IVPR
|| IVORn
|| 0b0000
0:15
16:27
where n specifies the IVOR register to be used for the particular interrupt type.
Any IVOR can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Figure 6-8 shows the IVOR field definitions, while Table 6-1 identifies the specfic IVOR register associated
with each interrupt type.
IVO
0
15 16
27 28
31
Figure 6-8. Interrupt Vector Offset Registers (IVOR0–IVOR15)
0:15
Reserved
16:27
28:31
IVO
Interrupt Vector Offset
Reserved
Table 6-1. Interrupt Types Associated with each IVOR
IVOR
IVOR0
Interrupt Type
Critical Input
IVOR1
IVOR2
IVOR3
IVOR4
IVOR5
IVOR6
IVOR7
IVOR8
IVOR9
IVOR10
IVOR11
IVOR12
IVOR13
IVOR14
IVOR15
Machine Check
Data Storage
Instruction Storage
External Input
Alignment
Program
Floating Point Unavailable
System Call
Auxiliary Processor Unavailable
Decrementer
Fixed Interval Timer
Watchdog Timer
Data TLB Error
Instruction TLB Error
Debug
6.4.10 Interrupt Vector Prefix Register (IVPR)
The IVPR provides the high-order 16 bits of the effective address of the interrupt vectors, for all interrupt
types. The interrupt vector effective address is formed as follows:
IVPR
|| IVORn
|| 0b0000
0:15
16:27
where n specifies the IVOR register to be used for the particular interrupt type.
The IVPR can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
IVP
0
15 16
31
Figure 6-9. Interrupt Vector Prefix Register (IVPR)
0:15
IVP
Interrupt Vector Prefix
Reserved
16:31
6.4.11 Exception Syndrome Register (ESR)
The ESR provides a syndrome to differentiate between the different kinds of exceptions that can generate the
same interrupt type. Upon the generation of one of these types of interrupt, the bit or bits corresponding to the
specific exception that generated the interrupt is set, and all other ESR bits are cleared. Other interrupt types
do not affect the contents of the ESR. Figure 6-10 on page 172 provides a summary of the fields of the ESR
along with their definitions. See the individual interrupt descriptions under Interrupt Definitions on page 175
for an explanation of the ESR settings for each interrupt type, as well as a more detailed explanation of the
function of certain ESR fields.
The ESR can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
MCI
PIL
BO
PCRE
PCRF
PTR
6
ST
8
AP
0
1
3
4
5
7
9
10 11 12 13 14 15 16
26 27 28 29
PCMP
31
PPR
FP
DLK
PUO PIE
Figure 6-10. Exception Syndrome Register (ESR)
Machine Check—Instruction Fetch Exception
This is an implementation-dependent field of the
ESR and is not part of the PowerPC Book-E Archi-
tecture.
0
Instruction Machine Check exception did not
0
MCI
occur.
1 Instruction Machine Check exception occurred.
Reserved
1:3
4
Program Interrupt—Illegal Instruction Exception
0 Illegal Instruction exception did not occur.
1 Illegal Instruction exception occurred.
PIL
Program Interrupt—Privileged Instruction Excep-
tion
5
6
PPR
PTR
0
Privileged Instruction exception did not occur.
1
Privileged Instruction exception occurred.
Program Interrupt—Trap Exception
0 Trap exception did not occur.
1 Trap exception occurred.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Floating Point Operation
0 Exception was not caused by a floating point
instruction.
1 Exception was caused by a floating point
instruction.
7
FP
ST
Store Operation
0 Exception was not caused by a store-type
storage access or cache management
instruction.
8
1 Exception was caused by a store-type storage
access or cache management instruction.
9
Reserved
Data Storage Interrupt—Locking Exception
00 Locking exception did not occur.
01 Locking exception was caused by dcbf.
10 Locking exception was caused by icbi.
11 Reserved
10:11
DLK
AP
AP Operation
0 Exception was not caused by an auxiliary
processor instruction.
12
1 Exception was caused by an auxiliary processor
instruction.
Program Interrupt—Unimplemented Operation
Exception
0 Unimplemented Operation exception did not
occur.
1 Unimplemented Operation exception occurred.
13
14
PUO
BO
Byte Ordering Exception
0 Byte Ordering exception did not occur.
1 Byte Ordering exception occurred.
Program Interrupt—Imprecise Exception
0 Exception occurred precisely; SRR0 contains
the address of the instruction that caused the
exception.
1 Exception occurred imprecisely; SRR0 contains
the address of an instruction after the one which
caused the exception.
This field is only set for a Floating-Point Enabled
exception type Program interrupt, and then only
when the interrupt occurs imprecisely due to
MSR[FE0,FE1] being set to a non-zero value when
an attached floating-point unit is already signaling
the Floating-Point Enabled exception (that is,
FPSCR[FEX] is already 1).
15
PIE
16:26
27
Reserved
This is an implementation-dependent field of the
ESR and is not part of the PowerPC Book-E Archi-
tecture.
Program Interrupt—Condition Register Enable
0 Instruction which caused the exception is not a
floating-point CR-updating instruction.
1 Instruction which caused the exception is a
floating-point CR-updating instruction.
PCRE
PCMP
This field is only defined for a Floating-Point
Enabled exception type Program interrupt, and
then only when ESR[PIE] is 0.
This is an implementation-dependent field of the
ESR and is not part of the PowerPC Book-E Archi-
tecture.
Program Interrupt—Compare
0 Instruction which caused the exception is not a
floating-point compare type instruction
1 Instruction which caused the exception is a
floating-point compare type instruction.
28
This field is only defined for a Floating-Point
Enabled exception type Program interrupt, and
then only when ESR[PIE] is 0.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
This is an implementation-dependent field of the
ESR and is not part of the PowerPC Book-E Archi-
tecture.
Program Interrupt—Condition Register Field
If ESR[PCRE]=1, this field indicates which CR field
was to be updated by the floating-point instruction
which caused the exception.
29:31
PCRF
This field is only defined for a Floating-Point
Enabled exception type Program interrupt, and
then only when ESR[PIE] is 0.
6.4.12 Machine Check Status Register (MCSR)
The MCSR contains status to allow the Machine Check interrupt handler software to determine the cause of a
machine check exception. Any Machine Check exception that is handled as an asynchronous interrupt sets
MCSR[MCS] and other appropriate bits of the MCSR. If MSR[ME] and MCSR[MCS] are both set, the
The MCSR is read into a GPR using mfspr. Clearing the MCSR is performed using mtspr by placing a 1 in
the GPR source register in all bit positions which are to be cleared in the MCSR, and a 0 in all other bit posi-
tions. The data written from the GPR to the MCSR is not direct data, but a mask. A 1 clears the bit and a 0
leaves the corresponding MCSR bit unchanged.
DRB
2
TLBP
IMPE
8
MCS
0
DCSP
6
1
3
4
5
7
9
31
IB
ICP
DCFP
DWB
Figure 6-11. Machine Check Status Register (MCSR)
Set when a machine check exception occurs
that is handled in the asynchronous fashion.
One of MCSR bits 1:7 will be set simulta-
neously to indicate the exception type. When
MSR[ME] and this bit are both set, Machine
Check interrupt is taken.
Machine Check Summary
0 No async machine check exception pending
1 Async machine check exception pending
0
MCS
Instruction PLB Error
0 Exception not caused by Instruction Read PLB
interrupt request (IRQ)
1 Exception caused by Instruction Read PLB interrupt
request (IRQ)
1
2
IB
Data Read PLB Error
0 Exception not caused by Data Read PLB interrupt
request (IRQ)
DRB
1 Exception caused by Data Read PLB interrupt
request (IRQ)
Data Write PLB Error
0 Exception not caused by Data Write PLB interrupt
request (IRQ)
1 Exception caused by Data Write PLB interrupt
request (IRQ)
3
4
DWB
TLBP
Translation Lookaside Buffer Parity Error
0 Exception not caused by TLB parity error
1 Exception caused by TLB parity error
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Instruction Cache Parity Error
0 Exception not caused by I-cache parity error
5
ICP
1 Exception caused by I-cache parity error
Set if and only If the DCU parity error was dis-
covered during a DCU Search operation.
Data Cache Search Parity Error
0 Exception not caused by DCU Search parity error
1 Exception caused by DCU Search parity error
6
7
DCSP
See Data Cache Parity Operations on
page 129.
Set if and only If the DCU parity error was dis-
covered during a DCU Flush operation.
Data Cache Flush Parity Error
0 Exception not caused by DCU Flush parity error
1 Exception caused by DCU Flush parity error
DCFP
IMPE
See Data Cache Parity Operations on
page 129.
Imprecise Machine Check Exception
0 No imprecise machine check exception occurred.
1 Imprecise machine check exception occurred.
Set if a machine check exception occurs that
sets MCSR[MCS] (or would if it were not
already set) and MSR[ME] = 0.
8
9:31
Reserved
6.5 Interrupt Definitions
register. The table also summarizes the various exception types that may cause that interrupt type; the clas-
sification of the interrupt; which ESR bit(s) can be set, if any; and which mask bit(s) can mask the interrupt
type, if any.
Detailed descrptions of each of the interrupt types follow the table..
Table 6-2. Interrupt and Exception Types
ESR
IVOR
Interrupt Type
Exception Type
IVOR0
Critical Input
Critical Input
x
x
CE
ME
ME
ME
2
2
Instruction Machine Check
Data Machine Check
TLB Machine Check
Read Access Control
Write Access Control
Cache Locking
[MCI]
IVOR1
Machine Check
x
x
x
x
x
x
[FP,AP]
ST,[FP,AP]
IVOR2
IVOR3
Data Storage
{DLK ,DLK }
0
1
Byte Ordering
BO,[ST],[FP,AP]
Execute Access Control
Byte Ordering
Instruction Storage
BO
IVOR4
IVOR5
External Input
Alignment
External Input
x
EE
Alignment
x
[ST],[FP,AP]
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table 6-2. Interrupt and Exception Types
ESR
(See Note 4)
IVOR
Interrupt Type
Exception Type
Illegal Instruction
x
x
x
PIL
Privileged Instruction
Trap
PPR,[AP]
PTR
IVOR6
Program
FP,[PIE],[PCRE]
{PCMP,PCRF}
FE0
FE1
FP Enabled
x
x
AP Enabled
x
x
x
x
x
AP
Unimplemented Op
FP Unavailable
System Call
PUO,[FP,AP]
IVOR7
FP Unavailable
System Call
IVOR8
IVOR9
AP Unavailable
Decrementer
AP Unavailable
Decrementer
IVOR10
IVOR11
IVOR12
IVOR13
IVOR14
x
x
x
EE
EE
CE
DIE
FIE
Fixed Interval Timer
Watchdog Timer
Data TLB Error
Instruction TLB Error
Fixed Interval Timer
Watchdog Timer
Data TLB Miss
Instruction TLB Miss
Trap
x
WIE
x
x
x
x
x
x
x
x
x
[ST],[FP,AP]
x
x
x
x
x
x
x
x
x
x
x
x
x
x
DE
DE
DE
DE
DE
DE
DE
DE
DE
IDM
IDM
IDM
IDM
IDM
IDM
IDM
IDM
IDM
Instruction Address Compare
Data Address Compare
Data Value Compare
Instruction Complete
Branch Taken
x
x
IVOR15
Debug
Return
x
Interrupt
x
x
Unconditional
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 6-2. Interrupt and Exception Types
ESR
(See Note 4)
IVOR
Interrupt Type
Exception Type
Table Notes
1. Although it is not specified as part of Book E, it is common for system implementations to provide, as part of the
interrupt controller, independent mask and status bits for the various sources of Critical Input and External Input
interrupts.
2. Machine Check interrupts are not classified as asynchronous nor synchronous. They are also not classified as
critical or non-critical, because they use their own unique set ot Save/Restore Registers, MCSRR0/1. See
3. Debug exceptions have special rules regarding their interrupt classification (synchronous or asynchronous,
and precise or imprecise), depending on the particular debug mode being used and other conditions (see
4. In general, when an interrupt causes a particular ESR bit or bits to be set as indicated in the table, it also
causes all other ESR bits to be cleared. Special rules apply to the ESR[MCI] field; see Machine Check
Interrupt on page 178. If no ESR setting is indicated for any of the exception types within a given interrupt type,
then the ESR is unchanged for that interrupt type.
The syntax for the ESR setting indication is as follows:
[xxx] means ESR[xxx] may be set
[xxx,yyy,zzz] means any one (or none) of ESR[xxx] or ESR[yyy] or ESR[zzz] may be set, but never more than one
{xxx,yyy,zzz} means that any combination of ESR[xxx], ESR[yyy], and ESR[zzz] may be set, including all or none
xxx means ESR[xxx] will be set
5. Byte Ordering exception type Data Storage interrupts can only occur when the PPC440x5 core is connected to
a floating-point unit or auxiliary processor, and then only when executing FP or AP load or store instructions.
6. Byte Ordering exception type Instruction Storage interrupts are defined by the PowerPC Book-E architecture,
but cannot occur within the PPC440x5 core. The core is capable of executing instructions from both big endian
and little endian code pages.
7. Unimplemented Operation exception type Program interrupts can only occur when the PPC440x5 core is
connected to a floating-point unit or auxiliary processor, and then only when executing instruction opcodes
which are recognized by the floating-point unit or auxiliary processor but are not implemented within the
hardware.
8. Floating-Point Unavailable and Auxiliary Processor Unavailable interrupts, as well as Floating-Point Enabled
and Auxiliary Processor Enabled exception type Program interrupts, can only occur when the PPC440x5 core
is connected to a floating-point unit or auxiliary processor, and then only when executing instruction opcodes
which are recognized by the floating-point unit or auxiliary processor, respectively.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
6.5.1 Critical Input Interrupt
A Critical Input interrupt occurs when no higher priority exception exists, a Critical Input exception is
presented to the interrupt mechanism, and MSR[CE] = 1. A Critical Input exception is caused by the activa-
tion of an asynchronous input to the PPC440x5 core. Although the only mask for this interrupt type within the
core is the MSR[CE] bit, system implementations typically provide an alternative means for independently
masking the interrupt requests from the various devices which collectively may activate the PPC440x5 core
Critical Input interrupt request input.
Note: MSR[CE] also enables the Watchdog Timer interrupt.
When a Critical Input interrupt occurs, the interrupt processing registers are updated as indicated below (all
registers not listed are unchanged), and instruction execution resumes at address IVPR[IVP] || IVOR0[IVO] ||
0b0000.
Critical Save/Restore Register 0 (CSRR0)
Set to the effective address of the next instruction to be executed.
Critical Save/Restore Register 1 (CSRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
ME
Unchanged.
All other MSR bits set to 0.
Programming Note: Software is responsible for taking any action(s) that are required by
the implementation in order to clear any Critical Input exception status
(such that the Critical Input interrupt request input signal is
deasserted) before reenabling MSR[CE], in order to avoid another,
redundant Critical Input interrupt.
6.5.2 Machine Check Interrupt
A Machine Check interrupt occurs when no higher priority exception exists, a Machine Check exception is
presented to the interrupt mechanism, and MSR[ME] = 1. The PowerPC architecture specifies Machine
Check interrupts as neither synchronous nor asynchronous, and indeed the exact causes and details of
handling such interrupts are implementation dependent. Regardless, for this particular processor core, it is
useful to describe the handling of interrupts caused by various types of Machine Check exceptions in those
terms. The PPC440x5 core includes four types of Machine Check exceptions. They are:
Instruction Synchronous Machine Check exception
A Instruction Synchronous Machine Check exception is caused when timeout or read error is
signaled on the instruction read PLB interface during an instruction fetch operation.
Such an exception is not presented to the interrupt handling mechanism, however, unless
and until such time as the execution is attempted of an instruction at an address associated
with the instruction fetch for which the Instruction Machine Check exception was asserted.
When the exception is presented, the ESR[MCI] bit will be set to indicated the type of excep-
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
tion, regardless of the state of the MSR[ME] bit.
If MSR[ME] is 1 when the Instruction Machine Check exception is presented to the interrupt
mechanism, then execution of the instruction associated with the exception will be sup-
pressed, a Machine Check interrupt will occur, and the interrupt processing registers will be
ated with the exception will be processed as though the exception did not exist and a
Machine Check interrupt will not occur (ever, even if and when MSR[ME] is subsequently set
to 1), although the ESR will still be updated as described on Page 179.
Instruction Asynchronous Machine Check exception
An Instruction Asynchronous Machine Check exception is caused when either:
• an instruction cache parity error is detected
• the read interrupt request is asserted on the instruction read PLB interface.
Data Asynchronous Machine Check exception
A Data Asynchronous Machine Check exception is caused when one of the following occurs:
• a timeout, read error, or read interrupt request is signaled on the data read PLB interface, during a
data read operation
• a timeout, write error, or write interrupt request is signaled on the data write PLB interface, during a
data write operation
• a parity error is detected on an access to the data cache.
TLB Asynchronous Machine Check exception
A TLB Asynchronous Machine Check exception is caused when a parity error is detected on an access to
the TLB.
When any Machine Check exception which is handled as an asynchronous interrupt occurs, it is immediately
presented to the interrupt handling mechanism. MCSR[MCS] is set, as are other bits of the MCSR as appro-
priate. A Machine Check interrupt will occur immediately if MSR[ME] is 1, and the interrupt processing regis-
ters will be updated as described below. If MSR[ME] is 0, however, then the exception will be “recorded” by
the setting of the MCSR[MCS] bit, and deferred until such time as MSR[ME] is subsequently set to 1. Any
time the MCSR[MCS] and MSR[ME] are both set to 1, the Machine Check interrupt will be taken. Therefore,
MCSR[MCS] must be cleared by software in the Machine Check interrupt handler before executing an rfmci
to return to processing with MSR[ME] set to 1.
When a Machine Check interrupt occurs, the interrupt processing registers are updated as indicated below
(all registers not listed are unchanged), and instruction execution resumes at address IVPR[IVP] ||
IVOR1[IVO] || 0b0000.
Machine Check Save/Restore Register 0 (MCSRR0)
For an Instruction Synchronous Machine Check exception, set to the effective address of the
instruction presenting the exception. For an Instruction Asynchronous Machine Check, Data
Asynchronous Machine Check, or TLB Asynchronous Machine Check exception, set to the
effective address of the next instruction to be executed.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Machine Check Save/Restore Register 1 (MCSRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
All MSR bits set to 0.
Exception Syndrome Register (ESR)
MCI
Set to 1 for an Instruction Machine Check exception; otherwise left unchanged.
All other defined ESR bits are set to 0 for an Instruction Machine Check exception; otherwise they are left
unchanged.
Programming Note: If an Instruction Synchronous Machine Check exception is associated
with an instruction, and execution of that instruction is attempted while
MSR[ME] is 0, then no Machine Check interrupt will occur, but
ESR[MCI] will still be set to 1 when the instruction actually executes.
Once set, ESR[MCI] cannot be cleared except by software, using the
mtspr instruction. When processing a Machine Check interrupt
handler, software should query ESR[MCI] to determine the type of
Machine Check exception, and then clear ESR[MCI]. Then, prior to re-
enabling Machine Check interrupts by setting MSR[ME] to 1, software
should query the status of ESR[MCI] again to determine whether any
additional Instruction Machine Check exceptions have occurrred while
MSR[ME] was disabled.
Machine Check Status Register (MCSR)
The MCSR collects status for the Machine Check exceptions that are handled as asynchronous interrupts.
MCSR[MCS] is set by any Instruction Asynchronous Machine Check exception, Data Asynchronous Machine
Check exception, or TLB Asynchronous Machine Check exception. Other bits in the MCSR are set to indicate
the exact type of Machine Check exception.
MCS Set to 1.
IB
Set to 1 if Instruction Read PLB Interrupt Request (IRQ) is asserted; otherwise set to
0.
DRB
Set to 1 if Data Read PLB Interrupt Request (IRQ) is asserted; otherwise set to 0.
DWB Set to 1 if Data Write PLB Interrupt Request (IRQ) is asserted; otherwise set to 0.
TLBP Set to 1 if the exception is a TLB parity error; otherwise set to 0.
ICP
Set to 1 if the exception is an instruction cache parity error; otherwise set to 0.
DCSP Set to 1 if the exception is a data cache parity error that resulted during a DCU
DCFP Set to 1 if the exception is a data cache parity error that resulted during a DCU Flush
IMPE Set to 1 if MCSR[MCS] is set (or would be, if it were not already set) and MSR[ME]
= 0; otherwise set to 0. When set, this bit indicates that a Machine Check exception
happened while Machine Check interrupts were disabled.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
See Machine Check Interrupts on page 161 for more information on the handling of Machine Check interrupts
within the PPC440x5 core.
Programming Note: If a Instruction Synchronous Machine Check exception occurs (i.e.
an error occurs on the PLB transfer that is intended to fill a line in the
instruction cache, any data associated with the exception will not be
placed into the instruction cache. On the other hand, if a Data
Asynchronous Machine Check exception occurs due to a PLB error
during a cacheable read operation, the data associated with the
exception will be placed into the data cache, and could subsequently
be loaded into a register. Similarly, if a Data Asynchronous Machine
Check exception due to a PLB error occurs during a caching inhibited
read operation, the data associated with the exception will be read into
a register. Data Asynchronous Machine Check exceptions resulting
from parity errors may or may not corrupt a GPR value, depending on
Programming Note: Since a dcbz instruction establishes a real address in the data cache
without actually reading the block of data from memory, it is possible
for a delayed Data Machine Check exception to occur if and when a
line established by a dcbz instruction is cast-out of the data cache and
written to memory, if the address of the cache line is not valid within
the system implementation.
6.5.3 Data Storage Interrupt
A Data Storage interrupt may occur when no higher priority exception exists and a Data Storage exception is
presented to the interrupt mechanism. The PPC440x5 core includes four types of Data Storage exception.
They are:
Read Access Control exception
A Read Access Control exception is caused by one of the following:
• While in user mode (MSR[PR] = 1), a load, icbi, icbt, dcbst, dcbf, dcbt, or dcbtst instruction
attempts to access a location in storage that is not enabled for read access in user mode (that is, the
TLB entry associated with the memory page being accessed has UR=0).
• While in supervisor mode (MSR[PR] = 0), a load, icbi, icbt, dcbst, dcbf, dcbt, or dcbtst instruction
attempts to access a location in storage that is not enabled for read access in supervisor mode (that
is, the TLB entry associated with the memory page being accessed has SR=0).
Programming Note: The instruction cache management instructions icbi and icbt are
treated as “loads” from the addressed byte with respect to address
translation and protection. These instruction cache management
instructions use MSR[DS] rather than MSR[IS] to determine
translation for their target effective address. Similarly, they use the
read access control field (UR or SR) rather than the execute access
control field (UX or SX) of the TLB entry to determine whether a Data
Storage exception should occur. Instruction Storage exceptions and
Instruction TLB Miss exceptions are associated with the fetching of
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
instructions not with the execution of instructions. Data Storage
exceptions and Data TLB Miss exceptions are associated with the
execution of instruction cache management instructions, as well as
with the execution of load, store, and data cache management
instructions.
Write Access Control exception
A Write Access Control exception is caused by one of the following:
• While in user mode (MSR[PR] = 1), a store, dcbz, or dcbi instruction attempts to access a location in
storage that is not enabled for write access in user mode (that is, the TLB entry associated with the
memory page being accessed has UW=0).
• While in supervisor mode (MSR[PR] = 0), a store, dcbz, or dcbi instruction attempts to access a
location in storage that is not enabled for write access in supervisor mode (that is, the TLB entry
associated with the memory page being accessed has SW=0).
Byte Ordering exception
A Byte Ordering exception will occur when a floating-point unit or auxiliary processor is
attached to the PPC440x5 core, and a floating-point or auxiliary processor load or store
instruction attempts to access a memory page with a byte order which is not supported by
the attached processor. Whether or not a given load or store instruction type is supported for
a given byte order is dependent on the implementation of the floating-point or auxiliary pro-
cessor. All integer load and store instructions supported by the PPC440x5 core are
supported for both big endian and little endian memory pages.
Cache Locking exception
A Cache Locking exception is caused by one of the following:
• While in user mode (MSR[PR] = 1) with MMUCR[IULXE]=1, execution of an icbi instruction is
attempted. The exception occurs whether or not the cache line targeted by the icbi instruction is
actually locked in the instruction cache.
• While in user mode (MSR[PR] = 1) with MMUCR[DULXE]=1, execution of a dcbf instruction is
attempted. The exception occurs whether or not the cache line targeted by the dcbf instruction is
actually locked in the data cache.
exceptions, respectively.
If a stwcx. instruction causes a Write Access Control exception, but the processor does not have the reserva-
tion from a lwarx instruction, then a Data Storage interrupt does not occur and the instruction completes,
updating CR[CR0] to indicate the failure of the store due to the lost reservation.
If a Data Storage exception occurs on any of the following instructions, then the instruction is treated as
a no-op, and a Data Storage interrupt does not occur.
• lswx or stswx with a length of zero (although the target register of lswx will still be undefined, as it is
whether or not a Data Storage exception occurs)
• icbt
• dcbt
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
• dcbtst
For all other instructions, if a Data Storage exception occurs, then execution of the instruction causing the
exception is suppressed, a Data Storage interrupt is generated, the interrupt processing registers are
updated as indicated below (all registers not listed are unchanged), and instruction execution resumes at
address IVPR[IVP] || IVOR2[IVO] || 0b0000.
Save/Restore Register 0 (SRR0)
Set to the effective address of the instruction causing the Data Storage interrupt.
Save/Restore Register 1 (SRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
CE, ME, DE Unchanged.
All other MSR bits set to 0.
Data Exception Address Register (DEAR)
If the instruction causing the Data Storage exception does so with respect to the memory
page targeted by the initial effective address calculated by the instruction, then the DEAR is
set to this calculated effective address. On the other hand, if the Data Storage exception only
occurs due to the instruction causing the exception crossing a memory page boundary, in
that the exception is with respect to the attributes of the page accessed after crossing the
boundary, then the DEAR is set to the address of the first byte within that page.
For example, consider a misaligned load word instruction that targets effective address
0x00000FFF, and that the page containing that address is a 4KB page. The load word will
thus cross the page boundary, and access the next page starting at address 0x00001000. If
a Read Access Control exception exists within the first page (because the Read Access
Control field for that page is 0), the DEAR will be set to 0x00000FFF. On the other hand, if
the Read Access Control field of the first page is 1, but the same field is 0 for the next page,
then the Read Access Control exception exists only for the second page and the DEAR will
be set to 0x00001000. Furthermore, the load word instruction in this latter scenario will have
Exception Syndrome Register (ESR)
FP
Set to 1 if the instruction causing the interrupt is a floating-point load or store;
otherwise set to 0.
ST
Set to 1 if the instruction causing the interrupt is a store, dcbz, or dcbi instruction;
otherwise set to 0.
DLK
Set to 0b10 if an icbi instruction caused a Cache Locking exception; set to 0b01 if a
dcbf instruction caused a Cache Locking exception; otherwise set to 0b00. Note that
a Read Access Control exception may occur in combination with a Cache Locking
exception, in which case software would need to examine the TLB entry associated
with the address reported in the DEAR to determine whether both exceptions had
occurred, or just a Cache Locking exception.
0:1
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
AP
BO
Set to 1 if the instruction causing the interrupt is an auxiliary processor load or store;
otherwise set to 0.
Set to 1 if the instruction caused a Byte Ordering exception; otherwise set to 0. Note
that a Read or Write Access Control exception may occur in combination with a Byte
Ordering exception, in which case software would need to examine the TLB entry
associated with the address reported in the DEAR to determine whether both
exceptions had occurred, or just a Byte Ordering exception.
MCI
Unchanged.
All other defined ESR bits are set to 0.
6.5.4 Instruction Storage Interrupt
An Instruction Storage interrupt occurs when no higher priority exception exists and an Instruction Storage
exception is presented to the interrupt mechanism. Note that although an Instruction Storage exception may
occur during an attempt to fetch an instruction, such an exception is not actually presented to the interrupt
mechanism until an attempt is made to execute that instruction. The PPC440x5 core includes one type of
Instruction Storage exception. That is:
Execute Access Control exception
An Execute Access Control exception is caused by one of the following:
• While in user mode (MSR[PR] = 1), an instruction fetch attempts to access a location in storage that
is not enabled for execute access in user mode (that is, the TLB entry associated with the memory
page being accessed has UX = 0).
• While in supervisor mode (MSR[PR] = 0), an instruction fetch attempts to access a location in storage
that is not enabled for execute access in supervisor mode (that is, the TLB entry associated with the
memory page being accessed has SX = 0).
Architecture Note: The PowerPC Book-E architecture defines an additional Instruction
Storage exception -- the Byte Ordering exception. This exception is
defined to assist implementations that cannot support dynamically
switching byte ordering between consecutive instruction fetches
and/or cannot support a given byte order at all. The PPC440x5 core
however supports instruction fetching from both big endian and little
endian memory pages, so this exception cannot occur.
When an Instruction Storage interrupt occurs, the processor suppresses the execution of the instruction
causing the Instruction Storage exception, the interrupt processing registers are updated as indicated below
(all registers not listed are unchanged), and instruction execution resumes at address IVPR[IVP] ||
IVOR3[IVO] || 0b0000.
Save/Restore Register 0 (SRR0)
Set to the effective address of the instruction causing the Instruction Storage interrupt.
Save/Restore Register 1 (SRR1)
Set to the contents of the MSR at the time of the interrupt.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Machine State Register (MSR)
CE, ME, DE Unchanged.
All other MSR bits set to 0.
Exception Syndrome Register (ESR)
BO
Set to 0.
MCI
Unchanged.
All other defined ESR bits are set to 0.
6.5.5 External Input Interrupt
An External Input interrupt occurs when no higher priority exception exists, an External Input exception is
presented to the interrupt mechanism, and MSR[EE] = 1. An External Input exception is caused by the activa-
tion of an asynchronous input to the PPC440x5 core. Although the only mask for this interrupt type within the
core is the MSR[EE] bit, system implementations typically provide an alternative means for independently
masking the interrupt requests from the various devices which collectively may activate the core’s External
Input interrupt request input.
Note: MSR[EE] also enables the External Input and Fixed-Interval Timer interrupts.
When an External Input interrupt occurs, the interrupt processing registers are updated as indicated below
(all registers not listed are unchanged), and instruction execution resumes at address IVPR[IVP] ||
IVOR4[IVO] || 0b0000.
Save/Restore Register 0 (SRR0)
Set to the effective address of the next instruction to be executed.
Save/Restore Register 1 (SRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
CE, ME, DE Unchanged.
All other MSR bits set to 0.
Programming Note: Software is responsible for taking any action(s) that are required by
the implementation in order to clear any External Input exception
status (such that the External Input interrupt request input signal is
deasserted) before reenabling MSR[EE], in order to avoid another,
redundant External Input interrupt.
6.5.6 Alignment Interrupt
An Alignment interrupt occurs when no higher priority exception exists and an Alignment exception is
presented to the interrupt mechanism. An Alignment exception occurs if execution of any of the following is
attempted:
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
• An integer load or store instruction that references a data storage operand that is not aligned on an oper-
and-sized boundary, when CCR0[FLSTA] is 1. Load and store multiple instructions are considered to ref-
erence word operands, and hence word-alignment is required for the target address of these instructions
when CCR0[FLSTA] is 1. Load and store string instructions are considered to reference byte operands,
and hence they cannot cause an Alignment exception due to CCR0[FLSTA] being 1, regardless of the tar-
get address alignment.
• A floating-point or auxiliary processor load or store instruction that references a data storage operand
that crosses a quadword (16 byte) boundary.
• A floating-point or auxiliary processor load or store instruction that references a data storage operand
that is not aligned on an operand-sized boundary, when the attached processing unit indicates to the
PPC440x5 core that the instruction requires operand-alignment.
• A floating-point or auxiliary processor load or store instruction that references a data storage operand
that is not aligned on a word boundary, when the attached processing unit indicates to the PPC440x5
core that the instruction requires word-alignment.
• A dcbz instruction that targets a memory page that is either write-through required or caching inhibited.
If a stwcx. instruction causes an Alignment exception, and the processor does not have the reservation from
a lwarx instruction, then an Alignment interrupt still occurs.
Programming Note: The architecture does not support the use of an unaligned effective
address by the lwarx and stwcx. instructions. If an Alignment
interrupt occurs due to the attempted execution of one of these
instructions, the Alignment interrupt handler must not attempt to
emulate the instruction, but instead should treat the instruction as a
programming error.
When an Alignment interrupt occurs, the processor suppresses the execution of the instruction causing the
Alignment exception, the interrupt processing registers are updated as indicated below (all registers not listed
are unchanged), and instruction execution resumes at address IVPR[IVP] || IVOR5[IVO] || 0b0000.
Save/Restore Register 0 (SRR0)
Set to the effective address of the instruction causing the Alignment interrupt.
Save/Restore Register 1 (SRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
CE, ME, DE Unchanged.
All other MSR bits set to 0.
Data Exception Address Register (DEAR)
Set to the effective address of the target data operand as calculated by the instruction caus-
ing the Alignment exception. Note that for dcbz, this effective address is not necessarily the
address of the first byte of the targeted cache block, but could be the address of any byte
within the block (it will be the address calculated by the dcbz instruction).
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Exception Syndrome Register (ESR)
FP
ST
AP
Set to 1 if the instruction causing the interrupt is a floating-point load or store;
otherwise set to 0.
Set to 1 if the instruction causing the interrupt is a store, dcbz, or dcbi instruction;
otherwise set to 0.
Set to 1 if the instruction causing the interrupt is an auxiliary processor load or store;
otherwise set to 0.
All other defined ESR bits are set to 0.
6.5.7 Program Interrupt
A Program interrupt occurs when no higher priority exception exists, a Program exception is presented to the
interrupt mechanism, and -- for the Floating-Point Enabled form of Program exception only -- MSR[FE0,FE1]
is non-zero. The PPC440x5 core includes six types of Program exception. They are:
Illegal Instruction exception
An Illegal Instruction exception occurs when execution is attempted of any of the following
kinds of instructions:
• a reserved-illegal instruction
• when MSR[PR] = 1 (user mode), an mtspr or mfspr that specifies an SPRN value with SPRN = 0
5
(user-mode accessible) that represents an unimplemented Special Purpose Register. For mtspr, this
includes any SPR number other than the XER, LR, CTR, or USPRG0. For mfspr, this includes any
SPR number other than the ones listed for mtspr, plus SPRG4-7, TBH, and TBL.
• a defined instruction which is not implemented within the PPC440x5 core, and which is not a floating-
point instruction. This includes all instructions that are defined for 64-bit implementations only, as well
as tlbiva and mfapidi (see the PowerPC Book-E specification)
• a defined floating-point instruction that is not recognized by an attached floating-point unit (or when
no such floating-point unit is attached)
• an allocated instruction that is not implemented within the PPC440x5 core and which is not recog-
nized by an attached auxiliary processor (or when no such auxiliary processor is attached)
defined and allocated instructions.
Privileged Instruction exception
A Privileged Instruction exception occurs when MSR[PR] = 1 and execution is attempted of
any of the following kinds of instructions:
• a privileged instruction
• an mtspr or mfspr instruction that specifies an SPRN value with SPRN = 1 (a Privileged Instruction
5
exception occurs regardless of whether or not the SPR referenced by the SPRN value is defined)
Trap exception
A Trap exception occurs when any of the conditions specified in a tw or twi instruction are
met. However, if Trap debug events are enabled (DBCR0[TRAP]=1), internal debug mode is
enabled (DBCR0[IDM]=1), and Debug interrupts are enabled (MSR[DE]=1), then a Trap
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
exception will cause a Debug interrupt to occur, rather than a Program interrupt.
Unimplemented Operation exception
An Unimplemented Operation exception occurs when execution is attempted of any of the
following kinds of instructions:
• a defined floating-point instruction that is recognized but not supported by an attached floating-point
unit, when floating-point instruction processing is enabled (MSR[FP]=1).
• an allocated instruction that is not implemented within the PPC440x5 core, and is recognized but not
supported by an attached auxiliary processor, when auxiliary processor instruction processing is
enabled. The enabling of auxiliary processor instruction processing is implementation-dependent.
Floating-Point Enabled exception
A Floating-Point Enabled exception occurs when the execution or attempted execution of a
defined floating-point instruction causes FPSCR[FEX] to be set to 1, in an attached floating-
point unit. FPSCR[FEX] is the Floating-Point Status and Control Register Floating-Point Ena-
bed Exception Summary bit (see the user’s manual for the floating-point unit implementation
for more details).
If MSR[FE0,FE1] is non-zero when the Floating-Point Enabled exception is presented to the
interrupt mechanism, then a Program interrupt will occur, and the interrupt processing regis-
ters will be updated as described below. If MSR[FE0,FE1] are both 0, however, then a
Program interrupt will not occur and the instruction associated with the exception will exe-
cute according to the definition of the floating-point unit (see the user’s manual for the
floating-point unit implementation). If and when MSR[FE0,FE1] are subsequently set to a
non-zero value, and the Floating-Point Enabled exception is still being presented to the inter-
rupt mechanism (that is, FPSCR[FEX] is still set), then a “delayed” Program interrupt will
occur, updating the interrupt processing registers as described below.
form of “delayed” Floating-Point Enabled exception.
Auxiliary Processor Enabled exception
An Auxiliary Processor Enabled exception may occur due to the execution or attempted exe-
cution of an allocated instruction that is not implemented within the PPC440x5 core, but is
recognized and supported by an attached auxiliary processor. The cause of such an excep-
tion is implementation-dependent. See the user’s manual for the auxiliary processor
implementation for more details.
When a Program interrupt occurs, the processor suppresses the execution of the instruction causing the
Program exception (for all cases except the “delayed” form of Floating-Point Enabled exception described
above), the interrupt processing registers are updated as indicated below (all registers not listed are
unchanged), and instruction execution resumes at address IVPR[IVP] || IVOR6[IVO] || 0b0000.
Save/Restore Register 0 (SRR0)
Set to the effective address of the instruction causing the Program interrupt, for all cases
except the “delayed” form of Floating-Point Enabled exception described above.
For the special case of the delayed Floating-Point Enabled exception, where the exception
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
was already being presented to the interrupt mechanism at the time MSR[FE0,FE1] was
changed from 0 to a non-zero value, SRR0 is set to the address of the instruction that would
have executed after the MSR-changing instruction. If the instruction which set
MSR[FE0,FE1] was rfi, rfci, or rfmci, then CSRR0 is set to the address to which the rfi, rfci,
or rfmci was returning, and not to the address of the instruction which was sequentially after
the rfi, rfci, or rfmci.
Save/Restore Register 1 (SRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
CE, ME, DE Unchanged.
All other MSR bits set to 0.
Exception Syndrome Register (ESR)
PIL
Set to 1 for an Illegal Instruction exception; otherwise set to 0
Set to 1 for a Privileged Instruction exception; otherwise set to 0
Set to 1 for a Trap exception; otherwise set to 0
PPR
PTR
PUO
FP
Set to 1 for an Unimplemented Operation exception; otherwise set to 0
Set to 1 if the instruction causing the interrupt is a floating-point instruction;
otherwise set to 0.
AP
Set to 1 if the instruction causing the interrupt is an auxiliary processor instruction;
otherwise set to 0.
PIE
Set to 1 if a “delayed” form of Floating-point Enabled exception type Program
interrupt; otherwise set to 0. The setting of ESR[PIE] to 1 indicates to the Program
interrupt handler that the interrupt was imprecise due to being caused by the
changing of MSR[FE0,FE1] and not directly by the execution of the floating-point
instruction which caused the exception by setting FPSCR[FEX]. Thus the Program
interrupt handler can recognize that SRR0 contains the address of the instruction
after the MSR-changing instruction, and not the address of the instruction that
caused the Floating-Point Enabled exception.
PCRE Set to 1 if a Floating-Point Enabled exception and the floating-point instruction which
caused the exception was a CR-updating instruction. Note that ESR[PCRE] is
undefined if ESR[PIE] is 1.
PCMP Set to 1 if a Floating-Point Enabled exception and the instruction which caused the
exception was a floating-point compare instruction. Note that ESR[PCMP] is
undefined if ESR[PIE] is 1.
PCRF Set to the number of the CR field (0 - 7) which was to be updated, if a Floating-Point
Enabled exception and the floating-point instruction which caused the exception was
a CR-updating instruction. Note that ESR[PCRF] is undefined if ESR[PIE] is 1.
All other defined ESR bits are set to 0.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Programming Note: The ESR[PCRE,PCMP,PCRF] fields are provided to assist the
Program interrupt handler with the emulation of part of the function of
the various floating-point CR-updating instructions, when any of these
instructions cause a precise (“non-delayed”) Floating-Point Enabled
exception type Program interrupt. The PowerPC Book-E floating-point
architecture defines that when such exceptions occur, the CR is to be
updated even though the rest of the instruction execution may be
suppressed. The PPC440x5 core, however, does not support such CR
updates when the instruction which is supposed to cause the update
is being suppressed due to the occurrence of a synchronous, precise
interrupt. Instead, the PPC440x5 core records in the
ESR[PCRE,PCMP,PCRF] fields information about the instruction
causing the interrupt, to assist the Program interrupt handler software
in performing the appropriate CR update manually.
6.5.8 Floating-Point Unavailable Interrupt
A Floating-Point Unavailable interrupt occurs when no higher priority exception exists, an attempt is made to
execute a floating-point instruction which is recognized by an attached floating-point unit, and MSR[FP]=0.
When a Floating-Point Unavailable interrupt occurs, the processor suppresses the execution of the instruc-
tion causing the Floating-Point Unavailable exception, the interrupt processing registers are updated as indi-
cated below (all registers not listed are unchanged), and instruction execution resumes at address IVPR[IVP]
|| IVOR7[IVO] || 0b0000.
Save/Restore Register 0 (SRR0)
Set to the effective address of the next instruction causing the Floating-Point Unavailable
interrupt.
Save/Restore Register 1 (SRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
CE, ME, DE Unchanged.
All other MSR bits set to 0.
6.5.9 System Call Interrupt
A System Call interrupt occurs when no higher priority exception exists and a system call (sc) instruction is
executed.
When a System Call interrupt occurs, the interrupt processing registers are updated as indicated below (all
registers not listed are unchanged), and instruction execution resumes at address IVPR[IVP] || IVOR8[IVO] ||
0b0000.
Save/Restore Register 0 (SRR0)
Set to the effective address of the instruction after the system call instruction.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Save/Restore Register 1 (SRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
CE, ME, DE Unchanged.
All other MSR bits set to 0.
6.5.10 Auxiliary Processor Unavailable Interrupt
An Auxiliary Processor Unavailable interrupt occurs when no higher priority exception exists, an attempt is
made to execute an auxiliary processor instruction which is not implemented within the PPC440x5 core but
which is recognized by an attached auxiliary processor, and auxiliary processor instruction processing is not
enabled. The enabling of auxiliary processor instruction processing is implementation-dependent. See the
user’s manual for the attached auxiliary processor.
When an Auxiliary Processor Unavailable interrupt occurs, the processor suppresses the execution of the
instruction causing the Auxiliary Processor Unavailable exception, the interrupt processing registers are
updated as indicated below (all registers not listed are unchanged), and instruction execution resumes at
address IVPR[IVP] || IVOR9[IVO] || 0b0000.
Save/Restore Register 0 (SRR0)
Set to the effective address of the next instruction causing the Auxiliary Processor Unavail-
able interrupt.
Save/Restore Register 1 (SRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
CE, ME, DE Unchanged.
All other MSR bits set to 0.
6.5.11 Decrementer Interrupt
A Decrementer interrupt occurs when no higher priority exception exists, a Decrementer exception exists
(TSR[DIS] = 1), and the interrupt is enabled (TCR[DIE] = 1 and MSR[EE] = 1). See Chapter 7, “Timer Facili-
Note: MSR[EE] also enables the External Input and Fixed-Interval Timer interrupts.
When a Decrementer interrupt occurs, the interrupt processing registers are updated as indicated below (all
registers not listed are unchanged), and instruction execution resumes at address IVPR[IVP] || IVOR10[IVO]
|| 0b0000.
Save/Restore Register 0 (SRR0)
Set to the effective address of the next instruction to be executed.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Save/Restore Register 1 (SRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
CE, ME, DE Unchanged.
All other MSR bits set to 0.
Programming Note: Software is responsible for clearing the Decrementer exception status
by writing to TSR[DIS], prior to reenabling MSR[EE], in order to avoid
another, redundant Decrementer interrupt.
6.5.12 Fixed-Interval Timer Interrupt
A Fixed-Interval Timer interrupt occurs when no higher priority exception exists, a Fixed-Interval Timer excep-
tion exists (TSR[FIS] = 1), and the interrupt is enabled (TCR[FIE] = 1 and MSR[EE]=1). See Chapter 7,
Note: MSR[EE] also enables the External Input and Decrementer interrupts.
When a Fixed interval Timer interrupt occurs, the interrupt processing registers are updated as indicated
below (all registers not listed are unchanged), and instruction execution resumes at address IVPR[IVP] ||
IVOR11[IVO] || 0b0000.
Save/Restore Register 0 (SRR0)
Set to the effective address of the next instruction to be executed.
Save/Restore Register 1 (SRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
CE, ME, DE Unchanged.
All other MSR bits set to 0.
Programming Note: Software is responsible for clearing the Fixed Interval Timer exception
status by writing to TSR[FIS], prior to reenabling MSR[EE], in order to
avoid another, redundant Fixed Interval Timer interrupt.
6.5.13 Watchdog Timer Interrupt
A Watchdog Timer interrupt occurs when no higher priority exception exists, a Watchdog Timer exception
exists (TSR[WIS] = 1), and the interrupt is enabled (TCR[WIE] = 1 and MSR[CE] = 1). See Chapter 7, “Timer
Note: MSR[CE] also enables the Critical Input interrupt.
When a Watchdog Timer interrupt occurs, the interrupt processing registers are updated as indicated below
(all registers not listed are unchanged), and instruction execution resumes at address IVPR[IVP] ||
IVOR12[IVO] || 0b0000.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Critical Save/Restore Register 0 (CSRR0)
Set to the effective address of the next instruction to be executed.
Critical Save/Restore Register 1 (CSRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
ME
Unchanged.
All other MSR bits set to 0.
Programming Note: Software is responsible for clearing the Watchdog Timer exception
status by writing to TSR[WIS], prior to reenabling MSR[CE], in order to
avoid another, redundant Watchdog Timer interrupt.
6.5.14 Data TLB Error Interrupt
A Data TLB Error interrupt may occur when no higher priority exception exists and a Data TLB Miss exception
is presented to the interrupt mechanism. A Data TLB Miss exception occurs when a load, store, icbi, icbt,
dcbst, dcbf, dcbz, dcbi, dcbt, or dcbtst instruction attempts to access a virtual address for which a valid
Programming Note: The instruction cache management instructions icbi and icbt are
treated as “loads” from the addressed byte with respect to address
translation and protection, and therefore use MSR[DS] rather than
MSR[IS] as part of the calculated virtual address when searching the
TLB to determine translation for their target storage address.
Instruction TLB Miss exceptions are associated with the fetching of
instructions not with the execution of instructions. Data TLB Miss
exceptions are associated with the execution of instruction cache
management instructions, as well as with the execution of load, store,
and data cache management instructions.
If a stwcx. instruction causes a Data TLB Miss exception, and the processor does not have the reservation
from a lwarx instruction, then a Data TLB Error interrupt still occurs.
If a Data TLB Miss exception occurs on any of the following instructions, then the instruction is treated
as a no-op, and a Data TLB Error interrupt does not occur.
• lswx or stswx with a length of zero (although the target register of lswx will be undefined)
• icbt
• dcbt
• dcbtst
For all other instructions, if a Data TLB Miss exception occurs, then execution of the instruction causing the
exception is suppressed, a Data TLB Error interrupt is generated, the interrupt processing registers are
updated as indicated below (all registers not listed are unchanged), and instruction execution resumes at
address IVPR[IVP] || IVOR13[IVO] || 0b0000.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Save/Restore Register 0 (SRR0)
Set to the effective address of the instruction causing the Data TLB Error interrupt.
Save/Restore Register 1 (SRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
CE, ME, DE Unchanged.
All other MSR bits set to 0.
Data Exception Address Register (DEAR)
If the instruction causing the Data TLB Miss exception does so with respect to the memory
page targeted by the initial effective address calculated by the instruction, then the DEAR is
set to this calculated effective address. On the other hand, if the Data TLB Miss exception
only occurs due to the instruction causing the exception crossing a memory page boundary,
in that the missing TLB entry is for the page accessed after crossing the boundary, then the
DEAR is set to the address of the first byte within that page.
As an example, consider a misaligned load word instruction that targets effective address
0x00000FFF, and that the page containing that address is a 4KB page. The load word will
thus cross the page boundary, and attempt to access the next page starting at address
0x00001000. If a valid TLB entry does not exist for the first page, then the DEAR will be set
to 0x00000FFF. On the other hand, if a valid TLB entry does exist for the first page, but not
for the second, then the DEAR will be set to 0x00001000. Furthermore, the load word
instruction in this latter scenario will have been partially executed (see Partially Executed
Exception Syndrome Register (ESR)
FP
Set to 1 if the instruction causing the interrupt is a floating-point load or store;
otherwise set to 0.
ST
Set to 1 if the instruction causing the interrupt is a store, dcbz, or dcbi instruction;
otherwise set to 0.
AP
MCI
Set to 1 if the instruction causing the interrupt is an auxiliary processor load or store;
otherwise set to 0.
Unchanged.
All other defined ESR bits are set to 0.
6.5.15 Instruction TLB Error Interrupt
An Instruction TLB Error interrupt occurs when no higher priority exception exists and an Instruction TLB Miss
exception is presented to the interrupt mechanism. Note that although an Instruction TLB Miss exception may
occur during an attempt to fetch an instruction, such an exception is not actually presented to the interrupt
mechanism until an attempt is made to execute that instruction. An Instruction TLB Miss exception occurs
when an instruction fetch attempts to access a virtual address for which a valid TLB entry does not exist. See
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
When an Instruction TLB Error interrupt occurs, the processor suppresses the execution of the instruction
causing the Instruction TLB Miss exception, the interrupt processing registers are updated as indicated below
(all registers not listed are unchanged), and instruction execution resumes at address IVPR[IVP] ||
IVOR14[IVO] || 0b0000.
Save/Restore Register 0 (SRR0)
Set to the effective address of the instruction causing the Instruction TLB Error interrupt.
Save/Restore Register 1 (SRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
CE, ME, DE Unchanged.
All other MSR bits set to 0.
6.5.16 Debug Interrupt
A Debug interrupt occurs when no higher priority exception exists, a Debug exception exists in the Debug
Status Register (DBSR), the processor is in internal debug mode (DBCR0[IDM]=1), and Debug interrupts are
enabled (MSR[DE] = 1). A Debug exception occurs when a debug event causes a corresponding bit in the
DBSR to be set.
There are several types of Debug exception, as follows:
Instruction Address Compare (IAC) exception
An IAC Debug exception occurs when execution is attempted of an instruction whose
address matches the IAC conditions specified by the various debug facility registers. This
exception can occur regardless of debug mode, and regardless of the value of MSR[DE].
Data Address Compare (DAC) exception
A DAC Debug exception occurs when the DVC mechanism is not enabled, and execution is
attempted of a load, store, icbi, icbt, dcbst, dcbf, dcbz, dcbi, dcbt, or dcbtst instruction
whose target storage operand address matches the DAC conditions specified by the various
debug facility registers. This exception can occur regardless of debug mode, and regardless
of the value of MSR[DE].
Programming Note: The instruction cache management instructions icbi and icbt are
treated as “loads” from the addressed byte with respect to Debug
exceptions. IAC Debug exceptions are associated with the fetching of
instructions not with the execution of instructions. DAC Debug
exceptions are associated with the execution of instruction cache
management instructions, as well as with the execution of load, store,
and data cache management instructions.
Data Value Compare (DVC) exception
A DVC Debug exception occurs when execution is attempted of a load, store, or dcbz
instruction whose target storage operand address matches the DAC and DVC conditions
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
specified by the various debug facility registers. This exception can occur regardless of
debug mode, and regardless of the value of MSR[DE].
Branch Taken (BRT) exception
A BRT Debug exception occurs when BRT debug events are enabled (DBCR0[BRT]=1) and
execution is attempted of a branch instruction for which the branch conditions are met. This
exception cannot occur in internal debug mode when MSR[DE]=0, unless external debug
mode or debug wait mode is also enabled.
Trap (TRAP) exception
A TRAP Debug exception occurs when TRAP debug events are enabled (DBCR0[TRAP]=1)
and execution is attempted of a tw or twi instruction that matches any of the specified trap
conditions. This exception can occur regardless of debug mode, and regardless of the value
of MSR[DE].
Return (RET) exception
A RET Debug exception occurs when RET debug events are enabled (DBCR0[RET]=1) and
execution is attempted of an rfi, rfci, or rfmci instruction. For rfi, the RET Debug exception
can occur regardless of debug mode and regardless of the value of MSR[DE]. For rfci or
rfmci, the RET Debug exception cannot occur in internal debug mode when MSR[DE]=0,
unless external debug mode or debug wait mode is also enabled.
Instruction Complete (ICMP) exception
An ICMP Debug exception occurs when ICMP debug events are enabled (DBCR0[ICMP]=1)
and execution of any instruction is completed. This exception cannot occur in internal debug
mode when MSR[DE]=0, unless external debug mode or debug wait mode is also enabled.
Interrupt (IRPT) exception
An IRPT Debug exception occurs when IRPT debug events are enabled (DBCR0[IRPT]=1)
and an interrupt occurs. For non-critical class interrupt types, the IRPT Debug exception can
occur regardless of debug mode and regardless of the value of MSR[DE]. For critical class
interrupt types, the IRPT Debug exception cannot occur in internal debug mode (regardless
of the value of MSR[DE]), unless external debug mode or debug wait mode is also enabled.
Unconditional Debug Event (UDE) exception
A UDE Debug exception occurs when an Unconditional Debug Event is signaled over the
JTAG interface to the PPC440x5 core. This exception can occur regardless of debug mode,
and regardless of the value of MSR[DE].
There are four debug modes supported by the PPC440x5 core. They are: internal debug mode, external
debug mode, debug wait mode, and trace mode. Debug exceptions and interrupts are affected by the debug
mode(s) which are enabled at the time of the Debug exception. Debug interrupts occur only when internal
debug mode is enabled, although it is possible for external debug mode and/or debug wait mode to be
enabled as well. The remainder of this section assumes that internal debug mode is enabled and that
external debug mode and debug wait mode are not enabled, at the time of a Debug exception.
each of the Debug exception types when operating in each of the modes.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Programming Note: It is a programming error for software to enable internal debug mode
(by setting DBCR0[IDM] to 1) while Debug exceptions are already
present in the DBSR. Software must first clear all DBSR Debug
exception status (that is, all fields except IDE, MRR, IAC12ATS, and
IAC34ATS) before setting DBCR0[IDM] to 1.
If a stwcx. instruction causes a DAC or DVC Debug exception, but the processor does not have the reserva-
tion from a lwarx instruction, then the Debug exception is not recorded in the DBSR and a Debug interrupt
does not occur. Instead, the instruction completes and updates CR[CR0] to indicate the failure of the store
due to the lost reservation.
If a DAC exception occurs on an lswx or stswx with a length of zero, then the instruction is treated as a
no-op, the Debug exception is not recorded in the DBSR, and a Debug interrupt does not occur.
If a DAC exception occurs on an icbt, dcbt, or dcbtst instruction which is being no-op’ed due to some
other reason (either the referenced cache block is in a caching inhibited memory page, or a Data
Storage or Data TLB Miss exception occurs), then the Debug exception is not recorded in the DBSR and
a Debug interrupt does not occur. On the other hand, if the icbt, dcbt, or dcbtst instruction is not being
no-op’ed for one of these other reasons, the DAC Debug exception does occur and is handled in the
same fashion as other DAC Debug exceptions (see below).
For all other cases, when a Debug exception occurs, it is immediately presented to the interrupt handling
mechanism. A Debug interrupt will occur immediately if MSR[DE] is 1, and the interrupt processing registers
will be updated as described below. If MSR[DE] is 0, however, then the exception condition remains set in the
DBSR. If and when MSR[DE] is subsequently set to 1, and the exception condition is still present in the
DBSR, a “delayed” Debug interrupt will then occur either as a synchronous, imprecise interrupt, or as an
asynchronous interrupt, depending on the type of Debug exception.
When a Debug interrupt occurs, the interrupt processing registers are updated as indicated below (all regis-
ters not listed are unchanged) and instruction execution resumes at address IVPR[IVP] || IVOR15[IVO] ||
0b0000.
Critical Save/Restore Register 0 (CSRR0)
For Debug exceptions that occur while Debug interrupts are enabled (MSR[DE] = 1), CSRR0
is set as follows:
• For IAC, BRT, TRAP, and RET Debug exceptions, set to the address of the instruction causing the
Debug interrupt. Execution of the instruction causing the Debug exception is suppressed, and the
interrupt is synchronous and precise.
• For DAC and DVC Debug exceptions, if DBCR2[DAC12A] is 0, set to the address of the instruction
causing the Debug interrupt. Execution of the instruction causing the Debug exception is suppressed,
and the interrupt is synchronous and precise.
If DBCR2[DAC12A] is 1, however, then DAC and DVC Debug exceptions are handled asynchro-
nously, and CSRR0 is set to the address of the instruction that would have executed next had the
Debug interrupt not occurred. This could either be the address of the instruction causing the DAC or
DVC Debug exception, or the address of a subsequent instruction.
• For ICMP Debug exceptions, set to the address of the next instruction to be executed (the instruction
after the one whose completion caused the ICMP Debug exception). The interrupt is synchronous
and precise.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Since the ICMP Debug exception does not suppress the execution of the instruction causing the
exception, but rather allows it to complete before causing the interrupt, the behavior of the interrupt is
different in the special case where the instruction causing the ICMP Debug exception is itself setting
MSR[DE] to 0. In this case, the interrupt will be delayed and will occur if and when MSR[DE] is again
set to 1, assuming DBSR[ICMP] is still set. If the Debug interrupt occurs in this fashion, it will be syn-
chronous and imprecise, and CSRR0 will be set to the address of the instruction after the one which
set MSR[DE] to 1 (not the one which originally caused the ICMP Debug exception and in so doing set
MSR[DE] to 0). If the instruction which set MSR[DE] to 1 was rfi, rfci, or rfmci, then CSRR0 is set to
the address to which the rfi, rfci, or rfmci was returning, and not to the address of the instruction
which was sequentially after the rfi, rfci, or rfmci.
• For IRPT Debug exceptions, set to the address of the first instruction in the interrupt handler associ-
ated with the interrupt type that caused the IRPT Debug exception. The interrupt is asynchronous.
• For UDE Debug exceptions, set to the address of the instruction that would have executed next if the
Debug interrupt had not occurred. The interrupt is asynchronous.
For all Debug exceptions that occur while Debug interrupts are disabled (MSR[DE] = 0), the Debug inter-
rupt will be delayed and will occur if and when MSR[DE] is again set to 1, assuming the Debug exception
status is still set in the DBSR. If the Debug interrupt occurs in this fashion, CSRR0 is set to the address of
the instruction after the one which set MSR[DE]. If the instruction which set MSR[DE] was rfi, rfci, or
rfmci, then CSRR0 is set to the address to which the rfi, rfci, or rfmci was returning, and not to the
address of the instruction which was sequentially after the rfi, rfci, or rfmci. The interrupt is either syn-
chronous and imprecise, or asynchronous, depending on the type of Debug exception, as follows:
• For IAC and RET Debug exceptions, the interrupt is synchronous and imprecise.
• For BRT Debug exceptions, this scenario cannot occur. BRT Debug exceptions are not recognized
when MSR[DE]=0 if operating in internal debug mode.
• For TRAP Debug exceptions, the Debug interrupt is synchronous and imprecise. However, under
these conditions (TRAP Debug exception occurring while MSR[DE] is 0), the attempted execution of
the trap instruction for which one or more of the trap conditions is met will itself lead to a Trap excep-
tion type Program interrupt. The corresponding Debug interrupt which will occur later if and when
Debug interrupts are enabled will be in addition to the Program interrupt.
• For DAC and DVC Debug exceptions, if DBCR2[DAC12A] is 0, then the interrupt is synchronous and
imprecise. If DBCR2[DAC12A] is 1, then the interrupt is asynchronous.
• For ICMP Debug exceptions, this scenario cannot occur in this fashion. ICMP Debug exceptions are
not recognized when MSR[DE]=0 if operating in internal debug mode. However, a similar scenario
can occur when MSR[DE] is 1 at the time of the ICMP Debug exception, but the instruction whose
completion is causing the exception is itself setting MSR[DE] to 0. This scenario is described above
in the subsection on the ICMP Debug exception for which MSR[DE] is 1 at the time of the exception.
In that scenario, the interrupt is synchronous and imprecise.
• For IRPT and UDE Debug exceptions, the interrupt is asynchronous.
Critical Save/Restore Register 1 (CSRR1)
Set to the contents of the MSR at the time of the interrupt.
Machine State Register (MSR)
ME
Unchanged.
All other MSR bits set to 0.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
6.6 Interrupt Ordering and Masking
It is possible for multiple exceptions to exist simultaneously, each of which could cause the generation of an
interrupt. Furthermore, the PowerPC Book-E architecture does not provide for the generation of more than
one interrupt of the same class (critical or non-critical) at a time. Therefore, the architecture defines that inter-
rupts are ordered with respect to each other, and provides a masking mechanism for certain persistent inter-
rupt types.
When an interrupt type is masked (disabled), and an event causes an exception that would normally generate
an interrupt of that type, the exception persists as a status bit in a register (which register depends upon the
exception type). However, no interrupt is generated. Later, if the interrupt type is enabled (unmasked), and
the exception status has not been cleared by software, the interrupt due to the original exception event will
then finally be generated.
All asynchronous interrupt types can be masked. Machine Check interrupts can be masked, as well. In addi-
tion, certain synchronous interrupt types can be masked. The two synchronous interrupt types which can be
masked are the Floating-Point Enabled exception type Program interrupt (masked by MSR[FE0,FE1), and
the IAC, DAC, DVC, RET, and ICMP exception type Debug interrupts (masked by MSR[DE]).
Architecture Note: When an otherwise synchronous, precise interrupt type is “delayed” in
this fashion via masking, and the interrupt type is later enabled, the
interrupt that is then generated due to the exception event that
occurred while the interrupt type was disabled is then considered a
synchronous, imprecise class of interrupt.
In order to prevent a subsequent interrupt from causing the state information (saved in SRR0/SRR1,
CSRR0/CSRR1, or MCSRR0/MCSRR1) from a previous interrupt to be overwritten and lost, the PPC440x5
core performs certain functions. As a first step, upon any non-critical class interrupt, the processor automati-
cally disables any further asynchronous, non-critical class interrupts (External Input, Decrementer, and Fixed
Interval Timer) by clearing MSR[EE]. Likewise, upon any critical class interrupt, hardware automatically
disables any further asynchronous interrupts of either class (critical and non-critical) by clearing MSR[CE]
and MSR[DE], in addition to MSR[EE]. The additional interrupt types that are disabled by the clearing of
MSR[CE,DE] are the Critical Input, Watchdog Timer, and Debug interrupts. For machine check interrupts, the
processor automatically disables all maskable interrupts by clearing MSR[ME] as well as MSR[EE,CE,DE].
This first step of clearing MSR[EE] (and MSR[CE,DE] for critical class interrupts, and MSR[ME] for machine
checks) prevents any subsequent asynchronous interrupts from overwriting the relevant save/restore regis-
ters (SRR0/SRR1, CSRR0/CSRR1, or MCSRR0/MCSRR1), prior to software being able to save their
contents. The processor also automatically clears, on any interrupt, MSR[WE,PR,FP,FE0,FE1,IS,DS]. The
clearing of these bits assists in the avoidance of subsequent interrupts of certain other types. However, guar-
anteeing that these interrupt types do not occur and thus do not overwrite the save/restore registers also
requires the cooperation of system software. Specifically, system software must avoid the execution of
instructions that could cause (or enable) a subsequent interrupt, if the contents of the save/restore registers
have not yet been saved.
6.6.1 Interrupt Ordering Software Requirements
The following list identifies the actions that system software must avoid, prior to having saved the
save/restore registers’ contents:
• Reenabling of MSR[EE] (or MSR[CE,DE] in critical class interrupt handlers)
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
This prevents any asynchronous interrupts, as well as (in the case of MSR[DE]) any Debug interrupts
(which include both synchronous and asynchronous types).
• Branching (or sequential execution) to addresses not mapped by the TLB, or mapped without execute
access permission
This prevents Instruction Storage and Instruction TLB Error interrupts.
• Load, store, or cache management instructions to addresses not mapped by the TLB or not having the
necessary access permission (read or write)
This prevents Data Storage and Data TLB Error interrupts.
• Execution of system call (sc) or trap (tw, twi) instructions
This prevents System Call and Trap exception type Program interrupts.
• Execution of any floating-point instructions
This prevents Floating-Point Unavailable interrupts. Note that this interrupt would occur upon the execu-
tion of any floating-point instruction, due to the automatic clearing of MSR[FP]. However, even if software
were to re-enable MSR[FP], floating-point instructions must still be avoided in order to prevent Program
interrupts due to the possibility of Floating-Point Enabled and/or Unimplemented Operation exceptions.
• Reenabling of MSR[PR]
This prevents Privileged Instruction exception type Program interrupts. Alternatively, software could re-
enable MSR[PR], but avoid the execution of any privileged instructions.
• Execution of any Auxiliary Processor instructions that are not implemented in the PPC440x5 core
This prevents Auxiliary Processor Unavailable interrupts, as well as Auxiliary Processor Enabled and
Unimplemented Operation exception type Program interrupts. Note that the auxiliary processor instruc-
tions that are implemented within the PPC440x5 core do not cause any of these types of exceptions, and
can therefore be executed prior to software having saved the save/restore registers’ contents.
• Execution of any illegal instructions, or any defined instructions not implemented within the PPC440x5
core (64-bit instructions, tlbiva, mfapidi)
This prevents Illegal Instruction exception type Program interrupts.
• Execution of any instruction that could cause an Alignment interrupt
This prevents Alignment interrupts. See Alignment Interrupt on page 185 for a complete list of instruc-
tions that may cause Alignment interrupts.
• In the Machine Check handler, use of the caches and TLBs until any detected parity errors have been
corrected.
This will avoid additional parity errors.
It is not necessary for hardware or software to avoid critical class interrupts from within non-critical class inter-
rupt handlers (and hence the processor does not automatically clear MSR[CE,ME,DE] upon a non-critical
interrupt), since the two classes of interrupts use different pairs of save/restore registers to save the instruc-
tion address and MSR. The converse, however, is not true. That is, hardware and software must cooperate in
the avoidance of both critical and non-critical class interrupts from within critical class interrupt handlers, even
though the two classes of interrupts use different save/restore register pairs. This is because the critical class
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
interrupt may have occurred from within a non-critical class interrupt handler, prior to the non-critical class
interrupt handler having saved SRR0 and SRR1. Therefore, within the critical class interrupt handler, both
pairs of save/restore registers may contain data that is necessary to the system software.
Similarly, the Machine Check handler must avoid further machine checks, as well as both critical and non-crit-
ical interrupts, since the machine check handler may have been called from within a critical or non-critical
interrupt handler.
6.6.2 Interrupt Order
The following is a prioritized listing of the various enabled interrupt types for which exceptions might exist
simultaneously:
1. Synchronous (non-debug) interrupts:
1. Data Storage
2. Instruction Storage
3. Alignment
4. Program
5. Floating-Point Unavailable
6. System Call
7. Auxiliary Processor Unavailable
8. Data TLB Error
9. Instruction TLB Error
Only one of the above types of synchronous interrupts may have an existing exception generating it at
architecture.
2. Machine Check
3. Debug
4. Critical Input
5. Watchdog Timer
6. External Input
7. Fixed-Interval Timer
8. Decrementer
Even though, as indicated above, the non-critical, synchronous exception types listed under item 1 are gener-
ated with higher priority than the critical interrupt types listed in items 2-5, the fact is that these non-critical
interrupts will immediately be followed by the highest priority existing critical interrupt type, without executing
any instructions at the non-critical interrupt handler. This is because the non-critical interrupt types do not
automatically clear MSR[ME,DE,CE] and hence do not automatically disable the critical interrupt types. In all
other cases, a particular interrupt type from the above list will automatically disable any subsequent interrupts
of the same type, as well as all other interrupt types that are listed below it in the priority order.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
6.7 Exception Priorities
PowerPC Book-E requires all synchronous (precise and imprecise) interrupts to be reported in program
order, as implied by the sequential execution model. The one exception to this rule is the case of multiple
synchronous imprecise interrupts. Upon a synchronizing event, all previously executed instructions are
required to report any synchronous imprecise interrupt-generating exceptions, and the interrupt(s) will then
be generated according to the general interrupt ordering rules outlined in Interrupt Order on page 201. For
example, if a mtmsr instruction causes MSR[FE0,FE1,DE] to all be set, it is possible that a previous Floating-
Point Enabled exception and a previous Debug exception both are still being presented (in the FPSCR and
DBSR, respectively). In such a scenario, a Floating-Point Enabled exception type Program interrupt will occur
first, followed immediately by a Debug interrupt.
For any single instruction attempting to cause multiple exceptions for which the corresponding synchronous
interrupt types are enabled, this section defines the priority order by which the instruction will be permitted to
cause a single enabled exception, thus generating a particular synchronous interrupt. Note that it is this
exception priority mechanism, along with the requirement that synchronous interrupts be generated in
program order, that guarantees that at any given time there exists for consideration only one of the synchro-
nous interrupt types listed in item 1 of Interrupt Order on page 201. The exception priority mechanism also
prevents certain debug exceptions from existing in combination with certain other synchronous interrupt-
generating exceptions.
This section does not define the permitted setting of multiple exceptions for which the corresponding interrupt
types are disabled. The generation of exceptions for which the corresponding interrupt types are disabled will
have no effect on the generation of other exceptions for which the corresponding interrupt types are enabled.
Conversely, if a particular exception for which the corresponding interrupt type is enabled is shown in the
following sections to be of a higher priority than another exception, the occurrence of that enabled higher
priority exception will prevent the setting of the other exception, independent of whether that other exception’s
corresponding interrupt type is enabled or disabled.
Except as specifically noted below, only one of the exception types listed for a given instruction type will be
permitted to be generated at any given time, assuming the corresponding interrupt type is enabled. The
priority of the exception types are listed in the following sections ranging from highest to lowest, within each
instruction type.
Finally, note that Machine Check exceptions are defined by the PowerPC architecture to be neither synchro-
nous nor asynchronous. As such, Machine Check exceptions are not considered in the remainder of this
section, which is specifically addressing the priority of synchronous interrupts.
6.7.1 Exception Priorities for Integer Load, Store, and Cache Management Instructions
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of any integer load, store, or cache management instruction.
Included in this category is the former opcode for the icbt instruction, which is an allocated opcode still
supported by the PPC440x5 core.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
4. Program (Illegal Instruction exception)
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Only applies to the defined 64-bit load, store, and cache management instructions, which are not recog-
nized by the PPC440x5 core.
5. Program (Privileged Instruction)
Only applies to the dcbi instruction, and only occurs if MSR[PR]=1.
6. Data TLB Error (Data TLB Miss exception)
7. Data Storage (all exception types except Byte Ordering exception)
8. Alignment (Alignment exception)
9. Debug (DAC or DVC exception)
10. Debug (ICMP exception)
6.7.2 Exception Priorities for Floating-Point Load and Store Instructions
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of any floating-point load or store instruction.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
4. Program (Illegal Instruction exception)
This exception will occur if no floating-point unit is attached to the PPC440x5 core, or if the particular
floating-point load or store instruction is not recognized by the attached floating-point unit.
5. Floating-Point Unavailable (Floating-Point Unavailable exception)
This exception will occur if an attached floating-point unit recognizes the instruction, but floating-point
instruction processing is disabled (MSR[FP]=0).
6. Program (Unimplemented Operation exception)
This exception will occur if an attached floating-point unit recognizes but does not support the instruction,
and floating-point instruction processing is enabled (MSR[FP]=1).
7. Data TLB Error (Data TLB Miss exception)
8. Data Storage (all exception types except Cache Locking exception)
9. Alignment (Alignment exception)
10. Debug (DAC or DVC exception)
11. Debug (ICMP exception)
6.7.3 Exception Priorities for Allocated Load and Store Instructions
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of any allocated load or store instruction.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
4. Program (Illegal Instruction exception)
This exception will occur if no auxiliary processor unit is attached to the PPC440x5 core, or if the particu-
lar allocated load or store instruction is not recognized by the attached auxiliary processor.
5. Program (Privileged Instruction exception)
This exception will occur if an attached auxiliary processor unit recognizes the instruction and indicates
that the instruction is privileged, but MSR[PR]=1.
6. Auxiliary Processor Unavailable (Auxiliary Processor Unavailable exception)
This exception will occur if an attached auxiliary processor recognizes the instruction, but indicates that
auxiliary processor instruction processing is disabled (whether or not auxiliary processor instruction pro-
cessing is enabled is implementation-dependent).
7. Program (Unimplemented Operation exception)
This exception will occur if an attached auxiliary processor recognizes but does not support the instruc-
tion, and also indicates that auxiliary processor instruction processing is enabled (whether or not auxiliary
processor instruction processing is enabled is implementation-dependent).
8. Data TLB Error (Data TLB Miss exception)
9. Data Storage (all exception types except Cache Locking exception)
10. Alignment (Alignment exception)
11. Debug (DAC or DVC exception)
12. Debug (ICMP exception)
6.7.4 Exception Priorities for Floating-Point Instructions (Other)
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of any floating-point instruction other than a load or store.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
4. Program (Illegal Instruction exception)
This exception will occur if no floating-point unit is attached to the PPC440x5 core, or if the particular
floating-point instruction is not recognized by the attached floating-point unit.
5. Floating-Point Unavailable (Floating-Point Unavailable exception)
This exception will occur if an attached floating-point unit recognizes the instruction, but floating-point
instruction processing is disabled (MSR[FP]=0).
6. Program (Unimplemented Operation exception)
This exception will occur if an attached floating-point unit recognizes but does not support the instruction,
and floating-point instruction processing is enabled (MSR[FP]=1).
7. Program (Floating-Point Enabled exception)
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
This exception will occur if an attached floating-point unit recognizes and supports the instruction, float-
ing-point instruction processing is enabled (MSR[FP]=1), and the instruction sets FPSCR[FEX] to 1.
8. Debug (ICMP exception)
6.7.5 Exception Priorities for Allocated Instructions (Other)
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of any allocated instruction other than a load or store, and which is
not one of the allocated instructions implemented within the PPC440x5 core.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
4. Program (Illegal Instruction exception)
This exception will occur if no auxiliary processor unit is attached to the PPC440x5 core, or if the particu-
lar allocated instruction is not recognized by the attached auxiliary processor and is not one of the allo-
cated instructions implemented within the PPC440x5 core.
5. Program (Privileged Instruction exception)
This exception will occur if an attached auxiliary processor unit recognizes the instruction and indicates
that the instruction is privileged, but MSR[PR]=1.
6. Auxiliary Processor Unavailable (Auxiliary Processor Unavailable exception)
This exception will occur if an attached auxiliary processor recognizes the instruction, but indicates that
auxiliary processor instruction processing is disabled (whether or not auxiliary processor instruction pro-
cessing is enabled is implementation-dependent).
7. Program (Unimplemented Operation exception)
This exception will occur if an attached auxiliary processor recognizes but does not support the instruc-
tion, and also indicates that auxiliary processor instruction processing is enabled (whether or not auxiliary
processor instruction processing is enabled is implementation-dependent).
8. Program (Auxiliary Processor Enabled exception)
This exception will occur if an attached auxiliary processor recognizes and supports the instruction, indi-
cates that auxiliary processor instruction processing is enabled, and the instruction execution results in
an Auxiliary Processor Enabled exception. Whether or not auxiliary processor instruction processing is
enabled is implementation-dependent, as is whether or not a given auxiliary processor instruction results
in an Auxiliary Processor Enabled exception.
9. Debug (ICMP exception)
6.7.6 Exception Priorities for Privileged Instructions
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of any privileged instruction other than dcbi, rfi, rfci, rfmci, or any
allocated instruction not implemented within the PPC440x5 core (all of which are covered elsewhere). This
list does cover, however, the dccci, dcread, iccci, and icread instructions, which are privileged, allocated
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
instructions that are implemented within the PPC440x5 core. This list also covers the defined 64-bit privileged
instructions, the tlbiva instruction, and the mfapidi instruction, all of which are not implemented by the
PPC440x5 core.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
4. Program (Illegal Instruction exception)
Only applies to the defined 64-bit privileged instructions, the tlbiva instruction, and the mfapidi instruc-
tion.
5. Program (Privileged Instruction exception)
Does not apply to the defined 64-bit privileged instructions, the tlbiva instruction, nor the mfapidi instruc-
tion.
6. Debug (ICMP exception)
Does not apply to the defined 64-bit privileged instructions, the tlbiva instruction, nor the mfapidi instruc-
tion.
6.7.7 Exception Priorities for Trap Instructions
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of a trap (tw, twi) instruction.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
4. Debug (TRAP exception)
5. Program (Trap exception)
6. Debug (ICMP exception)
6.7.8 Exception Priorities for System Call Instruction
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of a system call (sc) instruction.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
4. System Call (System Call exception)
5. Debug (ICMP exception)
Since the System Call exception does not suppress the execution of the sc instruction, but rather the excep-
tion occurs once the instruction has completed, it is possible for an sc instruction to cause both a System Call
exception and an ICMP Debug exception at the same time. In such a case, the associated interrupts will
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
6.7.9 Exception Priorities for Branch Instructions
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of a branch instruction.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
4. Debug (BRT exception)
5. Debug (ICMP exception)
6.7.10 Exception Priorities for Return From Interrupt Instructions
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of an rfi, rfci, or rfmci instruction.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
4. Debug (RET exception)
5. Debug (ICMP exception)
6.7.11 Exception Priorities for Preserved Instructions
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of a preserved instruction.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
4. Program (Illegal Instruction exception)
Applies to all preserved instructions except the mftb instruction, which is the only preserved class
instruction implemented within the PPC440x5 core.
5. Debug (ICMP exception)
Only applies to the mftb instruction, which is the only preserved class instruction implemented within the
PPC440x5 core.
6.7.12 Exception Priorities for Reserved Instructions
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of a reserved instruction.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
4. Program (Illegal Instruction exception)
Applies to all reserved instruction opcodes except the reserved-nop instruction opcodes.
5. Debug (ICMP exception)
Only applies to the reserved-nop instruction opcodes.
6.7.13 Exception Priorities for All Other Instructions
The following list identifies the priority order of the exception types that may occur within the PPC440x5 core
as the result of the attempted execution of all other instructions (that is, those not covered by one of the
sections 6.7.1 through 6.7.12). This includes both defined instructions and allocated instructions implemented
within the PPC440x5 core.
1. Debug (IAC exception)
2. Instruction TLB Error (Instruction TLB Miss exception)
3. Instruction Storage (Execute Access Control exception)
4. Program (Illegal Instruction exception)
Applies only to the defined 64-bit instructions, as these are not implemented within the PPC440x5 core.
5. Debug (ICMP exception)
Does not apply to the defined 64-bit instructions, as these are not implemented by the PPC440x5 core.
intrupts.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
7. Timer Facilities
The PPC440x5 provides four timer facilities: a time base, a Decrementer (DEC), a Fixed Interval Timer (FIT),
and a Watchdog Timer. These facilities, which share the same source clock frequency, can support:
• Time-of-day functions
• General software timing functions
• Peripherals requiring periodic service
• General system maintenance
• System error recoverability
Time Base (Incrementer)
TBU (32 bits)
31
TBL (32 bits)
Reference
Clock
31
0
0
Source
33
TBU (2 clocks)
31
29
TBL (2 clocks)
3
Watchdog Timer
Period
25
TBL (2 clocks)
7
21
TBL (2 clocks)
11
25
TBL (2 clocks)
7
21
TBL (2 clocks)
11
Fixed Interval Timer
Period
17
TBL (2 clocks)
15
13
TBL (2 clocks)
19
DEC (Decrementer)
(32 bits)
0
31
Zero Detect
Decrementer exception
Figure 7-1. Relationship of Timer Facilities to the Time Base
7.1 Time Base
The time base is a 64-bit register which increments once during each period of the source clock, and provides
a time reference. Access to the time base is via two Special Purpose Registers (SPRs). The Time Base
Upper (TBU) SPR contains the high-order 32 bits of the time base, while the Time Base Lower (TBL) SPR
contains the low-order 32 bits.
timers.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Software access to TBU and TBL is non-privileged for read but privileged for write, and hence different SPR
numbers are used for reading than for writing. TBU and TBL are written using mtspr and read using mfspr.
The period of the 64-bit time base is approximately 1462 years for a 400 MHz clock source. The time base
value itself does not generate any exceptions, even when it wraps. For most applications, the time base is set
once at system reset and only read thereafter. Note that Fixed Interval Timer and Watchdog Timer excep-
tions (discussed below) are caused by 0→1 transitions of selected bits from the time base. Transitions of
these bits caused by software alteration of the time base have the same effect as transitions caused by
normal incrementing of the time base.
Figure 7-2 illustrates the TBL.
0
31
Figure 7-2. Time Base Lower (TBL)
0:31
Time Base Lower
Low-order 32 bits of time base.
Figure 7-3 illustrates the TBU.
0
31
Figure 7-3. Time Base Upper (TBU)
0:31
Time Base Upper
High-order 32 bits of time base.
7.1.1 Reading the Time Base
The following code provides an example of reading the time base.
loop:
mfspr Rx,TBU
mfspr Ry,TBL
mfspr Rz,TBU
cmpw Rz, Rx
bne loop
# read TBU into GPR Rx
# read TBL into GPR Ry
# read TBU again, this time into GPR Rz
# see if old = new
# loop/reread if rollover occurred
The comparison and loop ensure that a consistent pair of values is obtained.
7.1.2 Writing the Time Base
The following code provides an example of writing the time base.
lwz
Rx, upper
# load 64-bit time base value into GPRs Rx and Ry
timers.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
lwz
li
Ry, lower
Rz, 0
# set GPR Rz to 0
mtspr TBL,Rz
mtspr TBU,Rx
mtspr TBL,Ry
# force TBL to 0 (thereby preventing wrap into TBU)
# set TBU to initial value
# set TBL to initial value
7.2 Decrementer (DEC)
The DEC is a 32-bit privileged SPR that decrements at the same rate that the time base increments. The
DEC is read and written using mfspr and mtspr, respectively. When a non-zero value is written to the DEC,
it begins to decrement with the next time base clock. A Decrementer exception is signalled when a decrement
occurs on a DEC count of 1, and the Decrementer Interrupt Status field of the Timer Status Register
Interrupts and Exceptions on page 159 provides more information on the handling of Decrementer interrupts.
The Decrementer interrupt handler software should clear TSR[DIS] before re-enabling MSR[EE], in order to
avoid another Decrementer interrupt due to the same exception (unless TCR[DIE] is cleared instead).
The behavior of the DEC itself upon a decrement from a DEC value of 1 depends on which of two modes it is
operating in -- normal, or auto-reload. The mode is controlled by the Auto-Reload Enable (ARE) field of the
TCR. When operating in normal mode (TCR[ARE]=0), the DEC simply decrements to the value 0 and then
stops decrementing until it is re-initialized by software.
When operating in auto-reload mode (TCR[ARE]=1), however, instead of decrementing to the value 0, the
DEC is reloaded with the value in the Decrementer Auto-Reload (DECAR) register (see Figure 7-5 on
page 212), and continues to decrement with the next time base clock (assuming the DECAR value was non-
zero). The DECAR register is a 32-bit privileged, write-only SPR, and is written using mtspr.
The auto-reload feature of the DEC is disabled upon reset, and must be enabled by software.
Figure 7-4 illustrates the DEC.
0
31
Figure 7-4. Decrementer (DEC)
0:31
Decrement value
Figure 7-5 illustrates the DECAR.
timers.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
0
31
Figure 7-5. Decrementer Auto-Reload (DECAR)
Copied to DEC at next time base clock when
0:31
Decrementer auto-reload value
DEC = 1 and auto-reload is enabled
(TCR[ARE] = 1).
Using mtspr to force the DEC to 0 does not cause a Decrementer exception and thus does not cause
TSR[DIS] to be set. However, if a time base clock causes a decrement from a DEC value of 1 to occur simul-
taneously with the writing of the DEC by a mtspr instruction, then the Decrementer exception will occur,
TSR[DIS] will be set, and the DEC will be written with the value from the mtspr.
In order for software to quiesce the activity of the DEC and eliminate all DEC exceptions, the following proce-
dure should be followed:
1. Write 0 to TCR[DIE]. This prevents a Decrementer exception from causing a Decrementer interrupt.
2. Write 0 to TCR[ARE]. This disables the DEC auto-reload feature.
3. Write 0 to the DEC to halt decrementing. Although this action does not itself cause a Decrementer excep-
tion, it is possible that a decrement from a DEC value of 1 has occurred since the last time that TSR[DIS]
was cleared.
4. Write 1 to TSR[DIS] (DEC Interrupt Status bit). This clears the Decrementer exception by setting
TSR[DIS] to 0. Because the DEC is no longer decrementing (due to having been written with 0 in step 3),
no further Decrementer exceptions are possible.
7.3 Fixed Interval Timer (FIT)
The FIT provides a mechanism for causing periodic exceptions with a regular period. The FIT would typically
be used by system software to invoke a periodic system maintenance function, executed by the Fixed Interval
Timer interrupt handler.
A Fixed Interval Timer exception occurs on a 0→1 transition of a selected bit from the time base. Note that a
Fixed Interval Timer exception will also occur if the selected time base bit transitions from 0→1 due to a
mtspr instruction that writes 1 to that time base bit when its previous value was 0.
The Fixed Interval Timer Period (FP) field of the TCR selects one of four bits from the time base, as shown in
Table 7-1. Fixed Interval Timer Period Selection
Period
(Time Base Clocks)
Period
(400 Mhz Clock)
TCR[FP]
Time Base Bit
13
0b00
0b01
0b10
TBL
2
2
2
clocks
clocks
clocks
20.48 µs
327.68 µs
5.2 ms
19
17
21
TBL
15
TBL
11
timers.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 7-1. Fixed Interval Timer Period Selection (continued)
Period
(Time Base Clocks)
Period
(400 Mhz Clock)
TCR[FP]
0b11
Time Base Bit
TBL
25
2
clocks
83.9 ms
7
When a Fixed Interval Timer exception occurs, the exception status is recorded by setting the Fixed interval
Timer Interrupt Status (FIS) field of the TSR to 1. A Fixed Interval Timer interrupt will occur if it is enabled by
both the Fixed Interval Timer Interrupt Enable (FIE) field of the TCR and by MSR[EE]. Fixed-Interval Timer
The Fixed Interval Timer interrupt handler software should clear TSR[FIS] before re-enabling MSR[EE], in
order to avoid another Fixed Interval Timer interrupt due to the same exception (unless TCR[FIE] is cleared
instead).
7.4 Watchdog Timer
The Watchdog Timer provides a mechanism for system error recovery in the event that the program running
on the PPC440x5 core has stalled and cannot be interrupted by the normal interrupt mechanism. The
Watchdog Timer can be configured to cause a critical-class Watchdog Timer interrupt upon the expiration of
a single period of the Watchdog Timer. It can also be configured to invoke a processor-initiated reset upon
the expiration of a second period of the Watchdog Timer.
A Watchdog Timer exception occurs on a 0→1 transition of a selected bit from the time base. Note that a
Watchdog Timer exception will also occur if the selected time base bit transitions from 0→1 due to a mtspr
instruction that writes 1 to that time base bit when its previous value was 0.
The Watchdog Timer Period (WP) field of the TCR selects one of four bits from the time base, as shown in
Table 7-2. Watchdog Timer Period Selection
Period
(Time Base Clocks)
Period
(400 MHz Clock)
TCR[WP]
Time Base Bit
TBL
21
0b00
0b01
0b10
0b11
2
2
2
2
clocks
clocks
clocks
clocks
5.2 ms
11
25
29
33
TBL
TBL
83.9 ms
1.34 s
7
3
TBU
21.47 s
31
The action taken upon a Watchdog Timer exception depends upon the status of the Enable Next Watchdog
(ENW) and Watchdog Timer Interrupt Status (WIS) fields of the TSR at the time of the exception. When
TSR[ENW] = 0, the next Watchdog Timer exception is “disabled”, and the only action to be taken upon the
exception is to set TSR[ENW] to 1. By clearing TSR[ENW], software can guarantee that the time until the
next enabled Watchdog Timer exception will be at least one full Watchdog Timer period (and a maximum of
two full Watchdog Timer periods).
When TSR[ENW] = 1, the next Watchdog Timer exception is enabled, and the action to be taken upon the
exception depends on the value of TSR[WIS] at the time of the exception. If TSR[WIS] = 0, then the action is
to set TSR[WIS] to 1, at which time a Watchdog Timer interrupt will occur if enabled by both the Watchdog
Timer Interrupt Enable (WIE) field of the TCR and by the Critical Interrupt Enable (CE) field of the MSR. The
Watchdog Timer interrupt handler software should clear TSR[WIS] before re-enabling MSR[CE], in order to
timers.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
avoid another Watchdog Timer interrupt due to the same exception (unless TCR[WIE] is cleared instead).
Watchdog Timer Interrupt on page 192 provides more information on the handling of Watchdog Timer inter-
rupts.
If TSR[WIS] is already 1 at the time of the next Watchdog Timer exception, then the action to take depends
on the value of the Watchdog Reset Control (TRC) field of the TCR. If TCR[WRC] is non-zero, then the value
of the TCR[WRC] field will be copied into TSR[WRS], TCR[WRC] will be cleared, and a core reset will occur
Note that once software has set TCR[WRC] to a non-zero value, it cannot be reset by software; this feature
prevents errant software from disabling the Watchdog Timer reset capability.
Table 7-3 summarizes the action to be taken upon a Watchdog Timer exception according to the values of
TSR[ENW] and TSR[WIS].
Table 7-3. Watchdog Timer Exception Behavior
TSR[ENW]
TSR[WIS]
Action upon Watchdog Timer exception
0
0
0
1
Set TSR[ENW] to 1
Set TSR[ENW] to 1
Set TSR[WIS] to 1. If Watchdog Timer interrupts are enabled (TCR[WIE]=1 and
MSR[CE]=1), then interrupt.
1
1
0
1
Cause Watchdog Timer reset action specified by TCR[WRC].
Reset will copy pre-reset TCR[WRC] into TSR[WRS], then clear TCR[WRC].
A typical system usage of the Watchdog Timer function is to enable the Watchdog Timer interrupt and the
Watchdog Timer reset function in the TCR (and MSR), and to start out with both TSR[ENW] and TSR[WIS]
clear. Then, a recurring software loop of reliable duration (or perhaps the interrupt handler for a periodic inter-
rupt such as the Fixed Interval Timer interrup) performs a periodic check of system integrity. Upon successful
completion of the system check, software clears TSR[ENW], thereby ensuring that a minimum of one full
Watchdog Timer period and a maximum of two full Watchdog Timer periods must expire before an enabled
Watchdog Timer exception will occur.
If for some reason the recurring software loop is not successfully completed (and TSR[ENW] thus not
cleared) during this period of time, then an enabled Watchdog Timer exeption will occur. This will set
TSR[WIS] and a Watchdog Timer interrupt will occur (if enabled by both TCR[WIE] and MSR[CE]). The
occurrence of a Watchdog Timer interrupt in this kind of system is interpreted as a “system error”, insofar as
the system was for some reason unable to complete the periodic system integrity check in time to avoid the
enabled Watchdog Timer exception. The action taken by the Watchdog Timer interrupt handler is of course
system-dependent, but typically the software will attempt to determine the nature of the problem and correct it
if possible. If and when the system attempts to resume operation, the software would typically clear both
TSR[WIS] and TSR[ENW], thus providing a minimum of another full Watchdog Timer period for a new system
integrity check to occur.
Finally, if for some reason the Watchdog Timer interrupt is disabled, and/or the Watchdog Timer interrupt
handler is unsuccessful in clearing TSR[WIS] and TSR[ENW] prior to another Watchdog Timer exception,
then the next exception will cause a processor reset operation to occur, according to the value of TCR[WRC].
timers.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Figure 7-6 illustrates the sequence of Watchdog Timer events which occurs according to this typical system
usage.
Watchdog Timer exception disabled;
next exception sets TSR[ENW] so
subsequent exception will set TSR[WIS].
Watchdog Timer exception
Exception
enabled; next exception sets
TSR[WIS] and causes
interrupt if enabled by
SW Loop
TSR[ENW,WIS] = 0b10
TSR[ENW,WIS] = 0b00
TCR[WIE] and MSR[CE].
Exception
Watchdog Timer
interrupt handler
If TCR[WRC] ≠ 0b00,
Exception
Exception
TSR[ENW,WIS} = 0b01
TSR[ENW,WIS] = 0b11
then RESET
;
else nothing.
Watchdog Timer exception disabled
but TSR[WIS] already set; this state
should not occur.
Watchdog Timer exception enabled and first
exception status still set; next exception
will cause RESET if enabled by TCR[WRC].
Figure 7-6. Watchdog State Machine
7.5 Timer Control Register (TCR)
The TCR is a privileged SPR that controls DEC, FIT, and Watchdog Timer operation. The TCR is read into a
GPR using mfspr, and is written from a GPR using mtspr.
The Watchdog Timer Reset Control (WRC) field of the TCR is cleared to 0 by processor reset (see Chapter 3,
“Initialization”). Each bit of this 2-bit field is set only by software and is cleared only by hardware. For each bit
of the field, once software has written it to 1, that bit remains 1 until processor reset occurs. This is to prevent
errant code from disabling the Watchdog Timer reset function.
The Auto-Reload Enable (ARE) field of the TCR is also cleared to zero by processor reset. This disables
the auto-reload feature of the DEC.
timers.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
WP
WIE
4
FP FIE
0
1
2
3
5
6
7
8
9
10
31
WRC
DIE
ARE
Figure 7-7. Timer Control Register (TCR)
Watchdog Timer Period
21
00 2 time base clocks
25
01 2 time base clocks
0:1
WP
29
10 2 time base clocks
33
11 2 time base clocks
TCR[WRC] resets to 0b00.
Watchdog Timer Reset Control
00 No Watchdog Timer reset will occur.
01 Core reset
10 Chip reset
11 System reset
Type of reset to cause upon Watchdog Timer excep-
tion with TSR[ENW,WIS]=0b11.
2:3
WRC
This field can be set by software, but cannot be
cleared by software, except by a software-induced
reset.
Watchdog Timer Interrupt Enable
0 Disable Watchdog Timer interrupt.
1 Enable Watchdog Timer interrupt.
4
5
WIE
Decrementer Interrupt Enable
0 Disable Decrementer interrupt.
1 Enable Decrementer interrupt.
DIE
Fixed Interval Timer (FIT) Period
13
00 2 time base clocks
17
01 2 time base clocks
6:7
8
FP
21
10 2 time base clocks
25
11 2 time base clocks
FIT Interrupt Enable
0 Disable Fixed Interval Timer interrupt.
FIE
1 Enable Fixed Interval Timer interrupt.
Auto-Reload Enable
0 Disable auto reload.
1 Enable auto reload.
9
ARE
TCR[ARE] resets to 0b0.
10:31
Reserved
7.6 Timer Status Register (TSR)
The TSR is a privileged SPR that records the status of DEC, FIT, and Watchdog Timer events. The fields of
the TSR are generally set to 1 only by hardware and cleared to 0 only by software. Hardware cannot clear
any fields in the TSR, nor can software set any fields. Software can read the TSR into a GPR using mfspr.
Clearing the TSR is performed using mtspr by placing a 1 in the GPR source register in all bit positions which
are to be cleared in the TSR, and a 0 in all other bit positions. The data written from the GPR to the TSR is not
direct data, but a mask. A 1 clears the bit and a 0 leaves the corresponding TSR bit unchanged.
timers.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
ENW WRS
FIS
5
0
1
2
3
4
6
31
WIS
DIS
Figure 7-8. Timer Status Register (TSR)
Enable Next Watchdog Timer Exception
0 Action on next Watchdog Timer exception is to set
TSR[ENW] = 1.
1 Action on next Watchdog Timer exception is governed
by TSR[WIS].
0
ENW
WIS
Watchdog Timer Interrupt Status
0 Watchdog Timer exception has not occurred.
1
1 Watchdog Timer exception has occurred.
Watchdog Timer Reset Status
00 No Watchdog Timer reset has occurred.
01 Core reset was forced by Watchdog Timer.
10 Chip reset was forced by Watchdog Timer.
11 System reset was forced by Watchdog Timer.
2:3
4
WRS
Decrementer Interrupt Status
0 Decrementer exception has not occurred.
1 Decrementer exception has occurred.
DIS
FIS
Fixed Interval Timer (FIT) Interrupt Status
0 Fixed Interval Timer exception has not occurred.
5
1 Fixed Interval Timer exception has occurred.
6:31
Reserved
7.7 Freezing the Timer Facilities
The debug mechanism provides a means for temporarily “freezing” the timers upon a debug exception.
Specifically, the time base and Decrementer can be prevented from incrementing and decrementing, respec-
tively, whenever a debug exception is recorded in the Debug Status Register (DBSR). This allows a
debugger to simulate the appearance of “real time”, even though the application has been temporarily
“halted” to service the debug event.
7.8 Selection of the Timer Clock Source
The source clock of the timers is selected by the Timer Clock Select (TCS) field of the Core Configuration
Register 1 (CCR1). When set to zero, CCR1[TCS] selects the CPU clock. This is the highest frequency timer
clock source.
When set to one, CCR1[TCS] selects an input to the CPU core as the timer clock. The input is sampled by a
latch clocked by the CPU clock, and so cannot cycle any faster than 1/2 the frequency of the CPU clock.
timers.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
timers.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
8. Debug Facilities
The debug facilities of the PPC440x5 include support for several debug modes for debugging during hard-
ware and software development, as well as debug events that allow developers to control the debug process.
Debug registers control these debug modes and debug events. The debug registers may be accessed either
through software running on the processor or through the JTAG debug port of the PPC440x5 core. Access to
the debug facilities through the JTAG debug port is typically provided by a debug tool such as the RISC-
Watch™ development tool from IBM. A trace port, which enables the tracing of code running in real time, is
also provided.
8.1 Support for Development Tools
The RISCWatch product from IBM is an example of a development tool that uses external debug mode,
debug events, and the JTAG debug port to implement a hardware and software development tool. The RISC-
Trace™ feature of RISCWatch is an example of a development tool that uses the real-time instruction trace
capability of the PPC440x5 core.
8.2 Debug Modes
The following sections describe the various debug modes supported by the PPC440x5. Each of these debug
modes supports a particular type of debug tool or debug task commonly used in embedded systems develop-
ment. For all debug modes, the various debug event types are enabled by the setting of corresponding fields
in Debug Control Register 0 (DBCR0), and upon their occurrence are recorded in the Debug Status Register
(DBSR).
There are four debug modes:
• Internal debug mode
• External debug mode
• Debug wait mode
• Trace debug mode
The PowerPC Book-E architecture specification deals only with internal debug mode, and the relationship of
Debug interrupts to the rest of the interrupt architecture. Internal debug mode is the mode which involves
debug software running on the processor itself, typically in the form of the Debug interrupt handler. The other
debug modes, on the other hand, are outside the scope of the architecture, and involve special-purpose
debug hardware external to the PPC440x5 core, connected either to the JTAG interface (for external debug
mode and debug wait mode) or the trace interface (for trace debug mode). Details of these interfaces and
their operation are beyond the scope of this manual.
8.2.1 Internal Debug Mode
Internal debug mode provides access to architected processor resources and supports setting hardware and
software breakpoints and monitoring processor status. In this mode, debug events are considered excep-
tions, which, in addition to recording their status in the DBSR, generate Debug interrupts if and when such
interrupts are enabled (Machine State Register (MSR) DE field is 1; see Interrupts and Exceptions on
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
debugger software at the interrupt handler can check processor status and other conditions related to the
debug event, as well as alter processor resources using all of the instructions defined for the PPC440x5.
Internal debug mode relies on this interrupt handling software at the Debug interrupt vector to debug software
problems. This mode, used while the processor executes instructions, enables debugging of both application
programs and operating system software, including all of the non-critical class interrupt handlers.
In this mode, the debugger software can communicate with the outside world through a communications port,
such as a serial port, external to the processor core.
To enable internal debug mode, the IDM field of DBCR0 must be set to 1 (DBCR0[IDM] = 1). This mode can
be enabled in combination with external debug mode (see External Debug Mode below) and/or debug wait
8.2.2 External Debug Mode
External debug mode provides access to architected processor resources and supports stopping, starting,
and stepping the processor, setting hardware and software breakpoints, and monitoring processor status. In
this mode, debug events record their status in the DBSR and then cause the processor to enter the stop
state, in which normal instruction execution stops and architected processor resources and memory can be
accessed and altered via the JTAG interface. While in the stop state, interrupts are temporarily disabled.
Storage access control by a memory management unit (MMU) remains in effect while in external debug
mode; the debugger may need to modify MSR or TLB values to access protected memory.
External debug mode relies only on internal processor resources, and no Debug interrupt handling software,
so it can be used to debug both system hardware and software problems. This mode can also be used for
software development on systems without a control program, or to debug control program problems,
including problems within the Debug interrupt handler itself, or within any other critical class interrupt
handlers.
External debug mode is enabled by setting DBCR0[EDM] to 1. This mode can be enabled in combination with
Mode below). External debug mode takes precedence over internal debug mode however, in that debug
events will first cause the processor to enter stop state rather than generating a Debug interrupt, although a
Debug interrupt may be pending while the processor is in the stop state.
8.2.3 Debug Wait Mode
Debug wait mode is similar to external debug mode in that debug events cause the processor to enter the
stop state. However, interrupts are still enabled while in debug wait mode, such that if and when an exception
occurs for which the associated interrupt type is enabled, the processor will leave the stop state and generate
the interrupt. This mode is useful for real-time hardware environments which cannot tolerate interrupts being
disabled for an extended period of time. In such environments, if external debug mode were to be used,
various I/O devices could operate incorrectly due to not being serviced in a timely fashion when they assert
an interrupt request to the processor, if the processor happened to be in stop state at the time of the interrupt
request.
When in debug wait mode, as with external debug mode, access to the architected processor resources and
memory is via the JTAG interface.
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Debug wait mode is enabled by setting both MSR[DWE] and the debug wait mode enable within the JTAG
controller to 1. Since MSR[DWE] is automatically cleared upon any interrupt, debug wait mode is temporarily
disabled upon an interrupt, and then can be automatically re-enabled when returning from the interrupt due to
the restoration of the MSR value upon the execution of an rfi, rfci, or rfmci instruction.
While debug wait mode can be enabled in combination with external debug mode, external debug mode
takes precedence and interrupts are temporarily disabled, thereby effectively nullifying the effect of debug
wait mode. Similarly, debug wait mode can be enabled in combination with internal debug mode. However, if
Debug interrupts are enabled (MSR[DE] is 1), then any debug event will lead to an exception and a corre-
sponding Debug interrupt, which takes precedence over the stop state associated with debug wait mode. On
the other hand, if Debug interrupts are disabled (MSR[DE] is 0), then debug wait mode will take effect and a
debug event will cause the processor to enter stop state.
8.2.4 Trace Debug Mode
Trace debug mode is simply the absence of each of the other modes. That is, if internal debug mode, external
debug mode, and debug wait mode are all disabled, then the processor is in trace debug mode. While in trace
debug mode, all debug events are simply recorded in the DBSR, and are indicated over the trace interface
from the PPC440x5 core. The processor does not enter the stop state, nor does a Debug interrupt occur.
8.3 Debug Events
There are several different kinds of debug events, each of which is enabled by a field in DBCR0 (except for
the Unconditional debug event) and recorded in the DBSR. Debug Modes on page 219 describes the opera-
tion that results when a debug event occurs while operating in any of the debug modes.
event type follow the table.
Table 8-1. Debug Events
Event
Description
Caused by the attempted execution of an instruction for which the address matches the
conditions specified by DBCR0, DBCR1, and the IAC1–IAC4 registers.
Instruction Address Compare (IAC)
Caused by the attempted execution of a load, store, or cache management instruction
for which the data storage address matches the conditions specified by DBCR0,
DBCR2, and the DAC1–DAC2 registers.
Data Address Compare (DAC)
Data Value Compare (DVC)
Branch Taken (BRT)
Caused by the attempted execution of a load, store, or cache management instruction
for which the data storage address matches the conditions specified by DBCR0,
DBCR2, and the DAC1–DAC2 registers, and for which the referenced data matches
the value specified by the DVC1–DVC2 registers.
Caused by the attempted execution of a branch instruction for which the branch condi-
tions are met (that is, for a branch instruction that results in the re-direction of the
instruction stream).
Caused by the attempted execution of a tw or twi instruction for which the trap condi-
tions are met.
Trap (TRAP)
Return (RET)
Caused by the attempted execution of an rfi, rfci, or rfmci instruction.
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table 8-1. Debug Events
Event
Description
Instruction Complete (ICMP)
Interrupt (IRPT)
Caused by the successful completion of the execution of any instruction.
Caused by the generation of an interrupt.
Caused by the assertion of an unconditional debug event request from the JTAG inter-
face to the PPC440x5 core.
Unconditional (UDE)
8.3.1 Instruction Address Compare (IAC) Debug Event
IAC debug events occur when execution is attempted of an instruction for which the instruction address and
other parameters match the IAC conditions specified by DBCR0, DBCR1, and the IAC registers. There are
four IAC registers on the PPC440x5, IAC1–IAC4. Depending on the IAC mode specified by DBCR1, these
IAC registers can be used to specify four independent, exact IAC addresses, or they can be configured in
pairs (IAC1/IAC2 and IAC3/IAC4) in order to specify ranges of instruction addresses for which IAC debug
events should occur.
8.3.1.1 IAC Debug Event Fields
There are several fields in DBCR0 and DBCR1 which are used to specify the IAC conditions, as follows:
IAC Event Enable Field
DBCR0[IAC1, IAC2, IAC3, IAC4] are the individual IAC event enables for each of the four IAC
events: IAC1, IAC2, IAC3, and IAC4. For a given IAC event to occur, the corresponding IAC
event enable bit in DBCR0 must be set. When a given IAC event occurs, the corresponding
DBSR[IAC1, IAC2, IAC3, IAC4] bit is set.
IAC Mode Field
DBCR1[IAC12M, IAC34M] control the comparison mode for the IAC1/IAC2 and IAC3/IAC4
events, respectively. There are three comparison modes supported by the PPC440x5:
• Exact comparison mode (DBCR1[IAC12M/IAC34M] = 0b00)
In this mode, the instruction address is compared to the value in the corresponding IAC register, and
the IAC event occurs only if the comparison is an exact match.
• Range inclusive comparison mode (DBCR1[IAC12M/IAC34M] = 0b10)
In this mode, the IAC1 or IAC2 event occurs only if the instruction address is within the range defined
by the IAC1 and IAC2 register values, as follows: IAC1 ≤ address < IAC2. Similarly, the IAC3 or IAC4
event occurs only if the instruction address is within the range defined by the IAC3 and IAC4 register
values, as follows: IAC3 ≤ address < IAC4.
For a given IAC1/IAC2 or IAC3/IAC4 pair, when the instruction address falls within the specified
range, either one or both of the corresponding IAC debug event bits will be set in the DBSR, as deter-
mined by which of the two corresponding IAC event enable bits are set in DBCR0. For example,
when the IAC1/IAC2 pair are set to range inclusive comparison mode, and the instruction address
falls within the defined range, then DBCR1[IAC1, IAC2] will determine whether one or the other or
both of DBSR[IAC1, IAC2] are set. It is a programming error to set either of the IAC pairs to a range
comparison mode (either inclusive or exclusive) without also enabling at least one of the correspond-
ing IAC event enable bits in DBCR0.
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Note that the IAC range auto-toggle mechanism can “switch” the IAC range mode from inclusive to
• Range exclusive comparison mode (DBCR1[IAC12M/IAC34M] = 0b11)
In this mode, the IAC1 or IAC2 event occurs only if the instruction address is outside the range
defined by the IAC1 and IAC2 register values, as follows: address < IAC1 or address ≥ IAC2. Simi-
larly, the IAC3 or IAC4 event occurs only if the instruction address is outside the range defined by the
IAC3 and IAC4 register values, as follows: address < IAC3 or address ≥ IAC4.
For a given IAC1/IAC2 or IAC3/IAC4 pair, when the instruction address falls outside the specified
range, either one or both of the corresponding IAC debug event bits will be set in the DBSR, as deter-
mined by which of the two corresponding IAC event enable bits are set in DBCR0. For example,
when the IAC1/IAC2 pair are set to range exclusive comparison mode, and the instruction address
falls outside the defined range, then DBCR1[IAC1, IAC2] will determine whether one or the other or
both of DBSR[IAC1, IAC2] are set. It is a programming error to set either of the IAC pairs to a range
comparison mode (either inclusive or exclusive) without also enabling at least one of the correspond-
ing IAC event enable bits in DBCR0.
Note that the IAC range auto-toggle mechanism can “switch” the IAC range mode from inclusive to
The PowerPC Book-E architecture defines DBCR1[IAC12M/IAC34M] = 0b01 as IAC address bit
mask mode, but that mode is not supported by the PPC440x5, and that value of the
IAC12M/IAC34M fields is reserved.
IAC User/Supervisor Field
DBCR1[IAC1US, IAC2US, IAC3US, IAC4US] are the individual IAC user/supervisor fields for
each of the four IAC events. The IAC user/supervisor fields specify what operating mode the
processor must be in order for the corresponding IAC event to occur. The operating mode is
determined by the Problem State field of the Machine State Register (MSR[PR]; see User and
does not matter; the IAC debug event may occur independent of the state of MSR[PR]. When this
field is 0b10, the processor must be operating in supervisor mode (MSR[PR] = 0). When this field
is 0b11, the processor must be operating in user mode (MSR[PR] = 1). The IAC user/supervisor
field value of 0b01 is reserved.
If a pair of IAC events (IAC1/IAC2 or IAC3/IAC4) are operating in range inclusive or range
exclusive mode, it is a programming error (and the results of any instruction address comparison
are undefined) if the corresponding pair of IAC user/supervisor fields are not set to the same
value. For example, if IAC1/IAC2 are operating in one of the range modes, then both
DBCR1[IAC1US] and DBCR1[IAC2US] must be set to the same value.
IAC Effective/Real Address Field
DBCR1[IAC1ER, IAC2ER, IAC3ER, IAC4ER] are the individual IAC effective/real address fields
for each of the four IAC events. The IAC effective/real address fields specify whether the
instruction address comparison should be performed using the effective, virtual, or real address
addresses). When the IAC effective/real address field is 0b00, the comparison is performed using
the effective address only—the IAC debug event may occur independent of the instruction
address space (MSR[IS]). When this field is 0b10, the IAC debug event occurs only if the
effective address matches the IAC conditions and is in virtual address space 0 (MSR[IS] = 0).
Similarly, when this field is 0b11, the IAC debug event occurs only if the effective address
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
matches the IAC conditions and is in virtual address space 1 (MSR[IS] = 1). Note that in these
latter two modes, in which the virtual address space of the instruction is considered, it is not the
entire virtual address which is considered. The Process ID, which forms the final part of the virtual
address, is not considered. Finally, the IAC effective/real address field value of 0b01 is reserved,
and corresponds to the PowerPC Book-E architected real address comparison mode, which is
not supported by the PPC440x5.
If a pair of IAC events (IAC1/IAC2 or IAC3/IAC4) are operating in range inclusive or range
exclusive mode, it is a programming error (and the results of any instruction address comparison
are undefined) if the corresponding pair of IAC effective/real address fields are not set to the
same value. For example, if IAC1/IAC2 are operating in one of the range modes, then both
DBCR1[IAC1ER] and DBCR1[IAC2ER] must be set to the same value.
IAC Range Mode Auto-Toggle Field
DBCR1[IAC12AT, IAC34AT] control the auto-toggle mechanism for the IAC1/IAC2 and
IAC3/IAC4 events, respectively. When the IAC mode for one of the pairs of IAC debug events is
set to one of the range modes (either range inclusive or range exclusive), then the IAC range
mode auto-toggle field corresponding to that pair of IAC debug events controls whether or not the
range mode will automatically “toggle” from inclusive to exclusive, and vice-versa. When the IAC
range mode auto-toggle field is set to 1, this automatic toggling is enabled; otherwise it is
disabled. It is a programming error (and the results of any instruction address comparison are
undefined) if an IAC range mode auto-toggle field is set to 1 without the corresponding IAC mode
field being set to one of the range modes.
When auto-toggle is enabled for a pair of IAC debug events, then upon each occurrence of an
IAC debug event within that pair the value of the corresponding auto-toggle status field in the
DBSR (DBSR[IAC12ATS, IAC34ATS]) is reversed. That is, if the auto-toggle status field is 0
before the occurrence of the IAC debug event, then it will be changed to 1 at the same time that
the IAC debug event is recorded in the DBSR. Conversely, if the auto-toggle status field is 1
before the occurrence of the IAC debug event, then it will be changed to 0 at the same time that
the IAC debug event is recorded in the DBSR.
Furthermore, when auto-toggle is enabled, the auto-toggle status field of the DBSR affects the
interpretation of the IAC mode field of DBCR1. If the auto-toggle status field is 0, then the IAC
is, the IAC mode field value of 0b10 selects range inclusive mode, whereas the value of 0b11
selects range exclusive mode. On the other hand, when the auto-toggle status field is 1, then the
interpretation of the IAC mode field is “reversed”. That is, the IAC mode field value of 0b10
selects range exclusive mode, whereas the value of 0b11 selects range inclusive mode.
The relationship of the IAC mode, IAC range mode auto-toggle, and IAC range mode auto-toggle
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 8-2. IAC Range Mode Auto-Toggle Summary
DBCR1
IAC12M/IAC34M
0b10
DBCR1
DBSR
IAC Mode
IAC12AT/IAC34AT
IAC12ATS/IAC34ATS
0
1
1
0
1
1
—
0
Range Inclusive
Range Inclusive
Range Exclusive
Range Exclusive
Range Exclusive
Range Inclusive
0b10
0b10
1
0b11
—
0
0b11
0b11
1
The affect of the auto-toggle mechanism is to cause the IAC mode to switch back and forth
between range inclusive mode and range exclusive mode, as each IAC range mode debug event
occurs. For example, if the IAC mode is set to range inclusive, and auto-toggle is enabled, and
the auto-toggle status field is 0, then the first IAC debug event will be a range inclusive event.
Upon that event, the DBSR auto-toggle status field will be set to 1, and the next IAC debug event
will then be a range exclusive event. Upon this next event, the DBSR auto-toggle status field will
be set back to 0, such that the next IAC debug event will again be a range inclusive event.
This auto-toggling between range inclusive and range exclusive IAC modes is particularly helpful
when enabling IAC range mode debug events in trace debug mode. A common debug operation
is to detect when the instruction stream enters a particular region of the instruction address space
(range inclusive mode). Once having entered the region of interest (a range inclusive event), it is
common for the debugger to then want to be informed when that region is exited (a range
exclusive event). By automatically toggling to range exclusive mode upon the occurrence of the
range inclusive IAC debug event, this particular debug operation is facilitated. Furthermore, by
not remaining in range inclusive mode upon entry to the region of interest, the debugger avoids a
continuous stream of range inclusive IAC debug events while the processor continues to execute
instructions within that region, which can often be for a very long series of instructions.
8.3.1.2 IAC Debug Event Processing
When operating in external debug mode or debug wait mode, the occurrence of an IAC debug event is
recorded in the corresponding bit of the DBSR and causes the instruction execution to be suppressed. The
processor then enters the stop state and ceases the processing of instructions. The program counter will
contain the address of the instruction which caused the IAC debug event. Similarly, when operating in internal
debug mode with Debug interrupts enabled (MSR[DE] = 1), the occurrence of an IAC debug event is
recorded in the DBSR and causes the instruction execution to be suppressed. A Debug interrupt then occurs
with Critical Save/Restore Register 0 (CSRR0) set to the address of the instruction which caused the IAC
debug event.
When operating in internal debug mode (and not also in external debug mode nor debug wait mode) with
Debug interrupts disabled (MSR[DE] = 0), the behavior of IAC debug events depends on the IAC mode. If the
IAC mode is set to exact comparison, then an IAC debug event can occur and will set the corresponding IAC
field of the DBSR, along with the Imprecise Debug Event (IDE) field of the DBSR. The instruction execution is
not suppressed, as no Debug interrupt will occur immediately. Instead, instruction execution continues, and a
Debug interrupt will occur if and when MSR[DE] is set to 1, thereby enabling Debug interrupts, assuming soft-
ware has not cleared the IAC debug event status from the DBSR in the meantime. Upon such a “delayed”
interrupt, the Debug interrupt handler software may query the DBSR[IDE] field to determine that the Debug
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
interrupt has occurred imprecisely. On the other hand, if the IAC mode is set to either range inclusive or range
exclusive mode, then IAC debug events cannot occur when operating in internal debug mode with
MSR[DE] = 0, unless external debug mode and/or debug wait mode is also enabled.
When operating in trace mode, the occurrence of an IAC debug event simply sets the corresponding IAC field
of the DBSR and is indicated over the trace interface, and instruction execution continues.
8.3.2 Data Address Compare (DAC) Debug Event
DAC debug events occur when execution is attempted of a load, store, or cache management instruction for
which the data storage address and other parameters match the DAC conditions specified by DBCR0,
DBCR2, and the DAC registers. There are two DAC registers on the PPC440x5, DAC1 and DAC2.
Depending on the DAC mode specified by DBCR2, these DAC registers can be used to specify two indepen-
dent, exact DAC addresses, or they can be configured to operate as a pair. When operating as a pair, then
can specify either a range of data storage addresses for which DAC debug events should occur, or a combi-
nation of an address and an address bit mask for selective comparison with the data storage address.
Note that for integer load and store instructions, and for cache management instructions, the address that is
used in the DAC comparison is the starting data address calculated as part of the instruction execution. As
explained in the instruction definitions for the cache management instructions, the target operand of these
instructions is an aligned cache block, which on the PPC440x5 is 32 bytes. Therefore, the storage reference
for these instructions effectively ignores the low-order five bits of the calculated data address, and the entire
aligned 32-byte cache block—which starts at the calculated data address as modified with the low-order five
bits set to 0b00000—is accessed. However, the DAC comparison does not take into account this implicit 32-
byte alignment of the storage reference of a cache management instruction, and instead the DAC compar-
ison considers the entire data address, as calculated according to the instruction definition.
On the other hand, for auxiliary processor load and store instructions, the AP interface can specify that the
PPC440x5 should force the storage access to be aligned on an operand-size boundary, by zeroing the
appropriate number of low-order address bits. In such a case, the DAC comparison is performed against this
modified, alignment-forced address, rather than the original address as calculated according to the instruction
definition.
8.3.2.1 DAC Debug Event Fields
There are several fields in DBCR0 and DBCR2 which are used to specify the DAC conditions, as follows:
DAC Event Enable Field
DBCR0[DAC1R, DAC1W, DAC2R, DAC2W] are the individual DAC event enables for the two
DAC events DAC1 and DAC2. For each of the two DAC events, there is one enable for DAC read
events, and another for DAC write events. Load, dcbt, dcbtst, icbi, and icbt instructions may
cause DAC read events, while store, dcbst, dcbf, dcbi, and dcbz instructions may cause DAC
information on these instructions and the types of DAC debug events they may cause). For a
given DAC event to occur, the corresponding DAC event enable bit in DBCR0 for the particular
operation type must be set. When a DAC event occurs, the corresponding DBSR[DAC1R,
DAC1W, DAC2R, DAC2W] bit is set. These same DBSR bits are shared by DVC debug events
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
DAC Mode Field
DBCR2[DAC12M] controls the comparison mode for the DAC1 and DAC2 events. There are four
comparison modes supported by the PPC440x5:
• Exact comparison mode (DBCR2[DAC12M] = 0b00)
In this mode, the data address is compared to the value in the corresponding DAC register, and the
DAC event occurs only if the comparison is an exact match.
• Address bit mask mode (DBCR2[DAC12M] = 0b01)
In this mode, the DAC1 or DAC2 event occurs only if the data address matches the value in the
DAC1 register, as masked by the value in the DAC2 register. That is, the DAC1 register specifies an
address value, and the DAC2 register specifies an address bit mask which determines which bit of
the data address should participate in the comparison to the DAC1 value. For every bit set to 1 in the
DAC2 register, the corresponding data address bit must match the value of the same bit position in
the DAC1 register. For every bit set to 0 in the DAC2 register, the corresponding address bit compar-
ison does not affect the result of the DAC event determination.
This comparison mode is useful for detecting accesses to a particular byte address, when the
accesses may be of various sizes. For example, if the debugger is interested in detecting accesses to
byte address 0x00000003, then these accesses may occur due to a byte access to that specific
address, or due to a halfword access to address 0x00000002, or due to a word access to address
0x00000000. By using address bit mask mode and specifying that the low-order two bits of the
address should be ignored (that is, setting the address bit mask in DAC2 to 0xFFFFFFFC), the
debugger can detect each of these types of access to byte address 0x00000003.
When the data address matches the address bit mask mode conditions, either one or both of the
DAC debug event bits corresponding to the operation type (read or write) will be set in the DBSR, as
determined by which of the corresponding two DAC event enable bits are set in DBCR0. That is,
when an address bit mask mode DAC debug event occurs, the setting of DBCR2[DAC1R, DAC1W,
DAC2R, DAC2W] will determine whether one or the other or both of the DBSR[DAC1R, DAC1W,
DAC2R, DAC2W] bits corresponding to the operation type are set. It is a programming error to set
the DAC mode field to address bit mask mode without also enabling at least one of the four DAC
event enable bits in DBCR0.
• Range inclusive comparison mode (DBCR2[DAC12M] = 0b10)
In this mode, the DAC1 or DAC2 event occurs only if the data address is within the range defined by
the DAC1 and DAC2 register values, as follows: DAC1 ≤ address < DAC2.
When the data address falls within the specified range, either one or both of the DAC debug event
bits corresponding to the operation type (read or write) will be set in the DBSR, as determined by
which of the corresponding two DAC event enable bits are set in DBCR0. That is, when a range inclu-
sive mode DAC debug event occurs, the setting of DBCR2[DAC1R, DAC1W, DAC2R, DAC2W] will
determine whether one or the other or both of the DBSR[DAC1R, DAC1W, DAC2R, DAC2W] bits
corresponding to the operation type are set. It is a programming error to set the DAC mode field to a
range comparison mode (either inclusive or exclusive) without also enabling at least one of the four
DAC event enable bits in DBCR0.
• Range exclusive comparison mode (DBCR2[DAC12M] = 0b11)
In this mode, the DAC1 or DAC2 event occurs only if the data address is outside the range defined by
the DAC1 and DAC2 register values, as follows: address < DAC1 or address ≥ DAC2.
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
When the data address falls outside the specified range, either one or both of the DAC debug event
bits corresponding to the operation type (read or write) will be set in the DBSR, as determined by
which of the corresponding two DAC event enable bits are set in DBCR0. That is, when a range
exclusive mode DAC debug event occurs, the setting of DBCR2[DAC1R, DAC1W, DAC2R, DAC2W]
will determine whether one or the other or both of the DBSR[DAC1R, DAC1W, DAC2R, DAC2W] bits
corresponding to the operation type are set. It is a programming error to set the DAC mode field to a
range comparison mode (either inclusive or exclusive) without also enabling at least one of the four
DAC event enable bits in DBCR0.
DAC User/Supervisor Field
DBCR2[DAC1US, DAC2US] are the individual DAC user/supervisor fields for the two DAC
events. The DAC user/supervisor fields specify what operating mode the processor must be in
order for the corresponding DAC event to occur. The operating mode is determined by the
DAC debug event may occur independent of the state of MSR[PR]. When this field is 0b10, the
processor must be operating in supervisor mode (MSR[PR] = 0). When this field is 0b11, the
processor must be operating in user mode (MSR[PR] = 1). The DAC user/supervisor field value of
0b01 is reserved.
If the DAC mode is set to one of the “paired” modes (address bit mask mode, or one of the two
range modes), it is a programming error (and the results of any data address comparison are
undefined) if DBCR2[DAC1US] and DBCR2[DAC2US] are not set to the same value.
DAC Effective/Real Address Field
DBCR2[DAC1ER, DAC2ER] are the individual DAC effective/real address fields for the two DAC
events. The DAC effective/real address fields specify whether the instruction address comparison
address field is 0b00, the comparison is performed using the effective address only; the DAC
debug event may occur independent of the data address space (MSR[DS]). When this field is
0b10, the DAC debug event occurs only if the effective address matches the DAC conditions and
is in virtual address space 0 (MSR[DS] = 0). Similarly, when this field is 0b11, the DAC debug
event occurs only if the effective address matches the DAC conditions and is in virtual address
space 1 (MSR[DS] = 1). Note that in these latter two modes, in which the virtual address space of
the data is considered, it is not the entire virtual address which is considered. The Process ID,
which forms the final part of the virtual address, is not considered. Finally, the DAC effective/real
address field value of 0b01 is reserved, and corresponds to the PowerPC Book-E architected real
address comparison mode, which is not supported by the PPC440x5.
If the DAC mode is set to one of the “paired” modes (address bit mask mode, or one of the two
range modes), it is a programming error (and the results of any data address comparison are
undefined) if DBCR2[DAC1ER] and DBCR2[DAC2ER] are not set to the same value.
DVC Byte Enable Field
DBCR2[DVC1BE, DVC2BE] are the individual data value compare (DVC) byte enable fields for
the two DVC events. These fields must be disabled (by being set to 4b0000) in order for the
corresponding DAC debug event to be enabled. In other words, when any of the DVC byte enable
field bits for a given DVC event are set to 1, the corresponding DAC event is disabled, and the
various DAC field conditions are used in conjunction with the DVC field conditions to determine
for more information on DVC events.
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
8.3.2.2 DAC Debug Event Processing
The behavior of the PPC440x5 upon a DAC debug event depends on the setting of DBCR2[DAC12A]. This
field of DBCR2 controls whether DAC debug events are processed in a synchronous (DBCR2[DAC12A] = 0)
or an asynchronous (DBCR2[DAC12A] = 1) fashion.
DBCR2[DAC12A] = 0 (Synchronous Mode)
When operating in external debug mode or debug wait mode, the occurrence of a DAC debug
event is recorded in the corresponding bit of the DBSR and causes the instruction execution to be
suppressed. The processor then enters the stop state and ceases the processing of instructions.
The program counter will contain the address of the instruction which caused the DAC debug
event. Similarly, when operating in internal debug mode with Debug interrupts enabled
(MSR[DE] = 1), the occurrence of a DAC debug event is recorded in the DBSR and causes the
instruction execution to be suppressed. A Debug interrupt will occur with CSRR0 set to the
address of the instruction which caused the DAC debug event.
When operating in internal debug mode (and not also in external debug mode nor debug wait
mode) with Debug interrupts disabled (MSR[DE] = 0), then a DAC debug event will set the
corresponding DAC field of the DBSR, along with the Imprecise Debug Event (IDE) field of the
DBSR. The instruction execution is not suppressed, as no Debug interrupt will occur immediately.
Instead, instruction execution continues, and a Debug interrupt will occur if and when MSR[DE] is
set to 1, thereby enabling Debug interrupts, assuming software has not cleared the DAC debug
event status from the DBSR in the meantime. Upon such a “delayed” interrupt, the Debug
interrupt handler software may query the DBSR[IDE] field to determine that the Debug interrupt
has occurred imprecisely.
When operating in trace mode, the occurrence of a DAC debug event simply sets the
corresponding DAC field of the DBSR and is indicated over the trace interface, and instruction
execution continues. DBCR2[DAC12A] does not affect the processing of DAC debug events
when operating in trace mode.
Engineering Note: When DAC debug events are enabled in any debug mode other than
trace mode, and DBCR2[DAC12A] is set to 0 (synchronous mode), in
order for the PPC440x5 to deal with a DAC-related Debug interrupt in a
synchronous fashion, the processing of all potential DAC debug event-
causing instructions (loads, stores, and cache management instructions)
is impacted by one processor cycle. This one cycle impact occurs
whether or not the instruction is actually causing a DAC debug event.
Overall processor performance is thus significantly impacted if
synchronous mode DAC debug events are enabled. In order to maintain
normal processor performance while DAC debug events are enabled and
in the absence of any actual DAC debug events, software should set
DBCR2[DAC12A] to 1.
DBCR2[DAC12A] = 1 (Asynchronous Mode)
When operating in external debug mode or debug wait mode, the occurrence of a DAC debug
event is recorded in the corresponding bit of the DBSR and causes the processor to enter stop
state and cease processing instructions. However, the determination and processing of the DAC
debug event is not handled synchronously with respect to the instruction execution. That is, the
processor may process the DAC debug event and enter the stop state either before or after the
completion of the instruction causing the event. If the DAC debug event is processed before the
completion of the instruction causing the event, then upon entering the stop state the program
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
counter will contain the address of that instruction, and that instruction’s execution will have been
suppressed. Conversely, if the DAC debug event is processed after the completion of the
instruction causing the event, then the program counter will contain the address of some
instruction after the one which caused the event. Whether or not the DAC debug event
processing occurs before or after the completion of the instruction depends on the particular
circumstances surrounding the instruction’s execution, the details of which are generally beyond
the scope of this document.
Similarly, when operating in internal debug mode with Debug interrupts enabled (MSR[DE] = 1),
the occurrence of a DAC debug event is recorded in the DBSR and will generate a Debug
interrupt with CSRR0 set to the address of the instruction which caused the DAC debug event, or
to the address of some subsequent instruction, depending upon whether the event is processed
before or after the instruction completes.
When operating in internal debug mode (and not also in external debug mode nor debug wait
mode) with Debug interrupts disabled (MSR[DE] = 0), then a DAC debug event will set the
corresponding DAC field of the DBSR, along with the Imprecise Debug Event (IDE) field of the
DBSR. Instruction execution continues, and a Debug interrupt will occur if and when MSR[DE] is
set to 1, thereby enabling Debug interrupts, assuming software has not cleared the DAC debug
event status from the DBSR in the meantime. Upon such a “delayed” interrupt, the Debug
interrupt handler software may query the DBSR[IDE] field to determine that the Debug interrupt
has occurred imprecisely.
When operating in trace mode, the occurrence of a DAC debug event simply sets the
corresponding DAC field of the DBSR and is indicated over the trace interface, and instruction
execution continues. DBCR2[DAC12A] does not affect the processing of DAC debug events
when operating in trace mode.
8.3.2.3 DAC Debug Events Applied to Instructions that Result in Multiple Storage Accesses
Certain misaligned load and store instructions are handled by making multiple, independent storage
accesses. Similarly, load and store multiple and string instructions which access more than one register result
in more than one storage access. Load and Store Alignment on page 117 provides a detailed description of
the circumstances that lead to such multiple storage accesses being made as the result of the execution of a
single instruction.
Whenever the execution of a given instruction results in multiple storage accesses, the data address of each
access is independently considered for whether or not it will cause a DAC debug event.
8.3.2.4 DAC Debug Events Applied to Various Instruction Types
Various special cases apply to the cache management instructions, the store word conditional indexed
(stwcx.) instruction, and the load and store string indexed (lswx, stswx) instructions, with regards to DAC
debug events. These special cases are as follows:
dcbz, dcbi
The dcbz and dcbi instructions are considered “stores” with respect to both storage access
control and DAC debug events. The dcbz instruction directly changes the contents of a given
storage location, whereas the dcbi instruction can indirectly change the contents of a given
storage location by invalidating data which has been modified within the data cache, thereby
“restoring” the value of the location to the “old” contents of memory. As “store” operations, they
may cause DAC write debug events.
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
dcbst, dcbf
The dcbst and dcbf instructions are considered “loads” with respect to storage access control,
since they do not change the contents of a given storage location. They may merely cause the
data at that storage location to be moved from the data cache out to memory. However, in a
debug environment, the fact that these instructions may lead to write operations on the external
interface is typically the event of interest. Therefore, these instructions are considered “stores”
with respect to DAC debug events, and may cause DAC write debug events.
dcbt, dcbtst, icbt
The touch instructions are considered “loads” with respect to both storage access control
and DAC debug events. However, these instructions are treated as no-ops if they reference
caching inhibited storage locations, or if they cause Data Storage or Data TLB Miss
exceptions. Consequently, if a touch instruction is being treated as a no-op for one of these
reasons, then it does not cause a DAC read debug event. However, if a touch instruction is
not being treated as a no-op for one of these reasons, it may cause a DAC read debug
event.
dcba
The dcba instruction is treated as a no-op by the PPC440x5, and thus will not cause a DAC
debug event.
icbi
The icbi instruction is considered a “load” with respect to both storage access control and DAC
debug events, and thus may cause a DAC read debug event.
dccci, dcread, iccci, icread
The dccci and iccci instructions do not generate an address, but rather they affect the entire
data and instruction cache, respectively. Similarly, the dcread and icread instructions do not
generate an address, but rather an “index” which is used to select a particular location in the
respective cache, without regard to the storage address represented by that location.
Therefore, none of these instructions cause DAC debug events.
stwcx.
If the execution of a stwcx. instruction would otherwise have caused a DAC write debug event,
but the processor does not have the reservation from a lwarx instruction, then the DAC write
debug event does not occur since the storage location does not get written.
lswx, stswx
DAC debug events do not occur for lswx or stswx instructions with a length of 0
(XER[TBC] = 0), since these instructions do not actually access storage.
8.3.3 Data Value Compare (DVC) Debug Event
DVC debug events occur when execution is attempted of a load, store, or dcbz instruction for which the data
storage address and other parameters match the DAC conditions specified by DBCR0, DBCR2, and the DAC
registers, and for which the data accessed matches the DVC conditions specified by DBCR2 and the DVC
registers. In other words, in order for a DVC debug event to occur, the conditions for a DAC debug event must
first be met, and then the data must also match the DVC conditions. Data Address Compare (DAC) Debug
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Event on page 226 describes the DAC conditions. In addition to the DAC conditions, there are two DVC regis-
ters on the PPC440x5, DVC1 and DVC2. The DVC registers can be used to specify two independent, 4-byte
data values, which are selectively compared against the data being accessed by a given load, store, or cache
management instruction.
When a DVC event occurs, the corresponding DBSR[DAC1R, DAC1W, DAC2R, DAC2W] bit is set. These
same DBSR bits are shared by DAC debug events.
8.3.3.1 DVC Debug Event Fields
In addition to the DAC debug event fields described in DAC Debug Event Fields on page 226, and the DVC
registers themselves, there are two fields in DBCR2 which are used to specify the DVC conditions, as follows:
DVC Byte Enable Field
DBCR2[DVC1BE, DVC2BE] are the individual DVC byte enable fields for the two DVC events.
When one or the other (or both) of these fields is disabled (by being set to 4b0000), the
corresponding DVC debug event is disabled (the corresponding DAC debug event may still be
enabled, as determined by the DAC debug event enable field of DBCR0). When either one or
both of these fields is enabled (by being set to a non-zero value), then the corresponding DVC
debug event is enabled.
Each bit of a given DVC byte enable field corresponds to a byte position within an aligned word of
memory. For a given aligned word of memory, the byte offsets (or “byte lanes”) within that word
are numbered 0, 1, 2, and 3, starting from the left-most (most significant) byte of the word.
Accordingly, bits 0:3 of a given DVC byte enable field correspond to bytes 0:3 of an aligned word
of memory being accessed.
For an access to “match” the DVC conditions for a given byte, the access must be actually
transferring data on that given byte position and the data must match the corresponding byte
value within the DVC register.
For each storage access, the DVC comparison is made against the bytes that are being
accessed within the aligned word of memory containing the starting byte of the transfer. For
example, consider a load word instruction with a starting data address of x01. The four bytes from
memory are located at addresses 0x01–0x04, but the aligned word of memory containing the
starting byte consists of addresses 0x00–0x03. Thus the only bytes being accessed within the
aligned word of memory containing the starting byte are the bytes at addresses 0x01–0x03, and
only these bytes are considered in the DVC comparison. The byte transferred from address 0x04
is not considered.
DVC Mode Field
DBCR2[DVC1M, DVC2M] are the individual DVC mode fields for the two DVC events. Each one
of these fields specifies the particular data value comparison mode for the corresponding DVC
debug event. There are three comparison modes supported by the PPC440x5:
• AND comparison mode (DBCR2[DVC1M, DVC2M] = 0b01)
In this mode, all data byte lanes enabled by a DVC byte enable field must be being accessed and
must match the corresponding byte data value in the corresponding DVC1 or DVC2 register.
• OR comparison mode (DBCR2[DVC1M, DVC2M] = 0b10)
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
In this mode, at least one data byte lane that is enabled by a DVC byte enable field must be being
accessed and must match the corresponding byte data value in the corresponding DVC1 or DVC2
register.
• AND-OR comparison mode (DBCR2[DVC1M, DVC2M] = 0b11)
In this mode, the four byte lanes of an aligned word are divided into two pairs, with byte lanes 0 and 1
being in one pair, and byte lanes 2 and 3 in the other pair. The DVC comparison mode for each pair
of byte lanes operates in AND mode, and then the results of these two AND mode comparisons are
ORed together to determine whether a DVC debug event occurs. In other words, a DVC debug event
occurs if either one or both of the pairs of byte lanes satisfy the AND mode comparison requirements.
This mode may be used to cause a DVC debug event upon an access of a particular halfword data
value in either of the two halfwords of a word in memory.
8.3.3.2 DVC Debug Event Processing
The behavior of the PPC440x5 upon a DVC debug event depends on the setting of DBCR2[DAC12A]. This
field of DBCR2 controls whether DVC debug events are processed in a synchronous (DBCR2[DAC12A] = 0)
or an asynchronous (DBCR2[DAC12A] = 1) fashion. The processing of DVC debug events is the same as it is
8.3.3.3 DVC Debug Events Applied to Instructions that Result in Multiple Storage Accesses
Certain misaligned load and store instructions are handled by making multiple, independent storage
accesses. Similarly, load and store multiple and string instructions which access more than one register result
in more than one storage access. Load and Store Alignment on page 117 provides a detailed description of
the circumstances that lead to such multiple storage accesses being made as the result of the execution of a
single instruction.
Whenever the execution of a given instruction results in multiple storage accesses, the address and data of
each access is independently considered for whether or not it will cause a DVC debug event.
8.3.3.4 DVC Debug Events Applied to Various Instruction Types
Various special cases apply to the cache management instructions, the store word conditional indexed
(stwcx.) instruction, and the load and store string indexed (lswx, stswx) instructions, with regards to DVC
debug events. These special cases are as follows:
dcbz
The dcbz instruction is the only cache management instruction which can cause a DVC debug
event. dcbz is the only such instruction which actually writes new data to a storage location (in
this case, an entire 32-byte data cache line is written to zeroes).
stwcx.
If the execution of a stwcx. instruction would otherwise have caused a DVC write debug event,
but the processor does not have the reservation from a lwarx instruction, then the DVC write
debug event does not occur since the storage location does not get written.
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
lswx, stswx
DVC debug events do not occur for lswx or stswx instructions with a length of 0
(XER[TBC] = 0), since these instructions do not actually access storage.
8.3.4 Branch Taken (BRT) Debug Event
BRT debug events occur when BRT debug events are enabled (DBCR0[BRT] = 1) and execution is
attempted of a branch instruction for which the branch condition(s) are satisfied, such that the instruction
stream will be redirected to the target address of the branch.
When operating in external debug mode or debug wait mode, the occurrence of a BRT debug event is
recorded in DBSR[BRT] and causes the instruction execution to be suppressed. The processor then enters
the stop state and ceases the processing of instructions. The program counter will contain the address of the
branch instruction which caused the BRT debug event. Similarly, when operating in internal debug mode with
Debug interrupts enabled (MSR[DE] = 1), the occurrence of a BRT debug event is recorded in DBSR[BRT]
and causes the instruction execution to be suppressed. A Debug interrupt will occur with CSRR0 set to the
address of the branch instruction which caused the BRT debug event.
When operating in internal debug mode (and not also in external debug mode nor debug wait mode) with
Debug interrupts disabled (MSR[DE] = 0), then BRT debug events cannot occur. Since taken branches are a
very common operation and thus likely to be frequently executed within the critical class interrupt handlers
(which typically have MSR[DE] set to 0), allowing BRT debug events under these conditions would lead to an
undesirable number of delayed (and hence imprecise) Debug interrupts.
When operating in trace mode, the occurrence of a BRT debug event is simply recorded in DBSR[BRT] and is
indicated over the trace interface, and instruction execution continues.
8.3.5 Trap (TRAP) Debug Event
TRAP debug events occur when TRAP debug events are enabled (DBCR0[TRAP] = 1) and execution is
attempted of a trap (tw, twi) instruction for which the trap condition is satisfied.
When operating in external debug mode or debug wait mode, the occurrence of a TRAP debug event is
recorded in DBSR[TRAP] and causes the instruction execution to be suppressed. The processor then enters
the stop state and ceases the processing of instructions. The program counter will contain the address of the
trap instruction which caused the TRAP debug event. Similarly, when operating in internal debug mode with
Debug interrupts enabled (MSR[DE] = 1), the occurrence of a TRAP debug event is recorded in
DBSR[TRAP] and causes the instruction execution to be suppressed. A Debug interrupt will occur with
CSRR0 set to the address of the trap instruction which caused the TRAP debug event.
When operating in internal debug mode (and not also in external debug mode nor debug wait mode) with
Debug interrupts disabled (MSR[DE] = 0), the occurrence of a TRAP debug event will set DBSR[TRAP],
along with the Imprecise Debug Event (IDE) field of the DBSR. Although a Debug interrupt will not occur
immediately, the instruction execution is suppressed as a Trap exception type Program interrupt will occur
instead. A Debug interrupt will also occur later, if and when MSR[DE] is set to 1, thereby enabling Debug
interrupts, assuming software has not cleared the TRAP debug event status from the DBSR in the meantime.
Upon such a “delayed” interrupt, the Debug interrupt handler software may query the DBSR[IDE] field to
determine that the Debug interrupt has occurred imprecisely.
When operating in trace mode, the occurrence of a TRAP debug event is simply recorded in DBSR[TRAP]
and is indicated over the trace interface, and instruction execution continues.
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
8.3.6 Return (RET) Debug Event
RET debug events occur when RET debug events are enabled (DBCR0[RET] = 1) and execution is
attempted of a return (rfi, rfci, or rfmci) instruction.
When operating in external debug mode or debug wait mode, the occurrence of a RET debug event is
recorded in DBSR[RET] and causes the instruction execution to be suppressed. The processor then enters
the stop state and ceases the processing of instructions. The program counter will contain the address of the
return instruction which caused the RET debug event. Similarly, when operating in internal debug mode with
Debug interrupts enabled (MSR[DE] = 1), the occurrence of a RET debug event is recorded in DBSR[RET]
and causes the instruction execution to be suppressed. A Debug interrupt will occur with CSRR0 set to the
address of the return instruction which caused the RET debug event.
When operating in internal debug mode (and not also in external debug mode nor debug wait mode) with
Debug interrupts disabled (MSR[DE] = 0), then RET debug events can occur only for rfi instructions, and not
for rfci or rfmci instructions. Since the rfci or rfmci instruction is typically used to return from a critical class
interrupt handler (including the Debug interrupt itself), and MSR[DE] is typically 0 at the time of the return, the
rfci or rfmci must not be allowed to cause a RET debug event under these conditions, or else it would not be
possible to return from the critical class interrupts.
For the rfi instruction only, if a RET debug event occurs under these conditions (internal debug mode
enabled, external debug mode and debug wait mode disabled, and MSR[DE] = 0), then DBSR[RET] is set,
along with the Imprecise Debug Event (IDE) field of the DBSR. The instruction execution is not suppressed,
as no Debug interrupt will occur immediately. Instead, instruction execution continues, and a Debug interrupt
will occur if and when MSR[DE] is set to 1, thereby enabling Debug interrupts, assuming software has not
cleared the RET debug event status from the DBSR in the meantime. Upon such a “delayed” interrupt, the
Debug interrupt handler software may query the DBSR[IDE] field to determine that the Debug interrupt has
occurred imprecisely.
When operating in trace mode, the occurrence of a RET debug event is simply recorded in DBSR[RET] and is
indicated over the trace interface, and instruction execution continues.
8.3.7 Instruction Complete (ICMP) Debug Event
ICMP debug events occur when ICMP debug events are enabled (DBCR0[ICMP] = 1) and the PPC440x5
completes the execution of any instruction.
When operating in external debug mode or debug wait mode, the occurrence of an ICMP debug event is
recorded in DBSR[ICMP] and causes the processor to enter the stop state and cease processing instructions.
The program counter will contain the address of the instruction which would have executed next, had the
ICMP debug event not occurred. Note that if the instruction whose completion caused the ICMP debug event
was a branch instruction (and the branch conditions were satisfied), then upon entering the stop state the
program counter will contain the target of the branch, and not the address of the instruction that is sequen-
tially after the branch. Similarly, if the ICMP debug event is caused by the execution of a return (rfi, rfci, or
rfmci) instruction, then upon entering the stop state the program counter will contain the address being
returned to, and not the address of the instruction which is sequentially after the return instruction.
When operating in internal debug mode with Debug interrupts enabled (MSR[DE] = 1), the occurrence of an
ICMP debug event is recorded in DBSR[ICMP] and a Debug interrupt will occur with CSRR0 set to the
address of the instruction which would have executed next, had the ICMP debug event not occurred. Note
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
that there is a special case of MSR[DE] = 1 at the time of the execution of the instruction causing the ICMP
debug event, but that instruction itself sets MSR[DE] to 0. This special case is described in more detail in
When operating in internal debug mode (and not also in external debug mode nor debug wait mode) with
Debug interrupts disabled (MSR[DE] = 0), then ICMP debug events cannot occur. Since the code at the
beginning of the critical class interrupt handlers (including the Debug interrupt itself) must execute at least
temporarily with MSR[DE] = 0, there would be no way to avoid causing additional ICMP debug events and
setting DBSR[IDE], if ICMP debug events were allowed to occur under these conditions.
The PPC440x5 does not support the use of the ICMP debug event when operating in trace mode. Software
must not enable ICMP debug events unless one of the other debug modes is enabled as well.
8.3.8 Interrupt (IRPT) Debug Event
IRPT debug events occur when IRPT debug events are enabled (DBCR0[IRPT] = 1) and an interrupt occurs.
When operating in external debug mode or debug wait mode, the occurrence of an IRPT debug event is
recorded in DBSR[IRPT] and causes the processor to enter the stop state and cease processing instructions.
The program counter will contain the address of the instruction which would have executed next, had the
IRPT debug event not occurred. Since the IRPT debug event is caused by the occurrence of an interrupt, by
definition this address is that of the first instruction of the interrupt handler for the interrupt type which caused
the IRPT debug event.
When operating in internal debug mode with external debug mode and debug wait mode both disabled (and
regardless of the value of MSR[DE]), an IRPT debug event can only occur due to a non-critical class interrupt.
Critical class interrupts (Machine Check, Critical Input, Watchdog Timer, and Debug interrupts) cannot cause
IRPT debug events in internal debug mode (unless also in external debug mode or debug wait mode), as
otherwise the Debug interrupt which would occur as the result of the IRPT debug event would by necessity
always be imprecise, since the critical class interrupt which would be causing the IRPT debug event would
itself be causing MSR[DE] to be set to 0.
For a non-critical class interrupt which is causing an IRPT debug event while internal debug mode is enabled
and external debug mode and debug wait mode are both disabled, the occurrence of the IRPT debug event is
recorded in DBSR[IRPT]. If MSR[DE] is 1 at the time of the IRPT debug event, then a Debug interrupt occurs
with CSRR0 set to the address of the instruction which would have executed next, had the IRPT debug event
not occurred. Since the IRPT debug event is caused by the occurrence of some other interrupt, by definition
this address is that of the first instruction of the interrupt handler for the interrupt type which caused the IRPT
debug event. If MSR[DE] is 0 at the time of the IRPT debug event, then the Imprecise Debug Event (IDE) field
of the DBSR is also set and a Debug interrupt does not occur immediately. Instead, instruction execution
continues, and a Debug interrupt will occur if and when MSR[DE] is set to 1, thereby enabling Debug inter-
rupts, assuming software has not cleared the IRPT debug event status from the DBSR in the meantime.
Upon such a “delayed” interrupt, the Debug interrupt handler software may query the DBSR[IDE] field to
determine that the Debug interrupt has occurred imprecisely.
When operating in trace mode, the occurrence of an IRPT debug event is simply recorded in DBSR[IRPT]
and is indicated over the trace interface, and instruction execution continues.
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
8.3.9 Unconditional Debug Event (UDE)
UDE debug events occur when a debug tool asserts the unconditional debug event request via the JTAG
interface. The UDE debug event is the only event which does not have a corresponding enable field in
DBCR0.
When operating in external debug mode or debug wait mode, the occurrence of a UDE debug event is
recorded in DBSR[UDE] and causes the processor to enter the stop state and cease processing instructions.
The program counter will contain the address of the instruction which would have executed next, had the
UDE debug event not occurred. Similarly, when operating in internal debug mode with Debug interrupts
enabled (MSR[DE] = 1), the occurrence of a UDE debug event is recorded in DBSR[UDE] and a Debug inter-
rupt will occur with CSRR0 set to the address of the instruction which would have executed next, had the
UDE debug event not occurred.
When operating in internal debug mode (and not also in external debug mode nor debug wait mode) with
Debug interrupts disabled (MSR[DE] = 0), the occurrence of a UDE debug event will set DBSR[UDE], along
with the Imprecise Debug Event (IDE) field of the DBSR. The Debug interrupt will not occur immediately.
Instead, instruction execution continues, and a Debug interrupt will occur if and when MSR[DE] is set to 1,
thereby enabling Debug interrupts, assuming software has not cleared the UDE debug event status from the
DBSR in the meantime. Upon such a “delayed” interrupt, the Debug interrupt handler software may query the
DBSR[IDE] field to determine that the Debug interrupt has occurred imprecisely.
When operating in trace mode, the occurrence of a UDE debug event simply sets DBSR[UDE] and is indi-
cated over the trace interface, and instruction execution continues.
8.3.10 Debug Event Summary
occurrence.
Table 8-3. Debug Event Summary
External
Debug
Mode
Debug
Wait
Mode
Internal
Debug
Mode
Debug Events
MSR
DE
IAC
Yes
Yes
Yes
DAC
Yes
Yes
Yes
DVC
Yes
Yes
Yes
BRT
Yes
Yes
Yes
TRAP
Yes
RET
ICMP
Yes
IRPT
Yes
UDE
Yes
Yes
Yes
Enabled
—
—
—
—
—
—
1
Yes
Yes
Yes
Enabled
Disabled
Yes
Yes
Yes
Disabled
Enabled
Yes
Yes
Note 1
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table 8-3. Debug Event Summary (continued)
External
Debug
Mode
Debug
Wait
Mode
Internal
Debug
Mode
Debug Events
MSR
DE
IAC
Note 2
Yes
DAC
Yes
Yes
DVC
Yes
Yes
BRT
No
TRAP
Yes
RET
ICMP
No
IRPT
Note 1
yes
UDE
Yes
Yes
Disabled
Disabled
Disabled
Disabled
Enabled
Disabled
0
Note 3
Yes
—
Yes
Yes
Note 4
Table Notes
1. IRPT debug events may only occur for non-critical class interrupts when operating in internal debug mode with
external debug mode and debug wait mode both disabled.
2. IAC debug events may not occur in internal debug mode with MSR[DE] = 0 and with external debug mode
and debug wait mode both disabled, and the IAC mode set to range inclusive or range exclusive. They may
occur if the IAC mode is set to exact.
3. RET debug events may not occur for rfci or rfmci instructions when operating in internal debug mode with
MSR[DE] = 0 and with external debug mode and debug wait mode both disabled. They may only occur in this
mode for the rfi instruction.
4. ICMP debug events are not permitted when operating in trace debug mode. Software must not enable ICMP
debug events unless one of the other debug modes is enabled.
8.4 Debug Reset
Software can initiate an immediate reset operation by setting DBCR0[RST] to a non-zero value. The results of
a reset operation within the PPC440x5 core are described in Initialization on page 85. The results of a reset
operation on the rest of the chip and/or system is dependent on the particular type of reset operation (core,
chip, or system reset), and on the particular chip and system implementation. See the chip user’s manual for
details.
8.5 Debug Timer Freeze
In order to maintain the semblance of “real time” operation while a system is being debugged, DBCR0[FT]
can be set to 1, which will cause all of the timers within the PPC440x5 core to stop incrementing or decre-
menting for as long as a debug event bit is set in the DBSR, or until DBCR0[FT] is set to 0. See Timer Facili-
8.6 Debug Registers
Various Special Purpose Registers (SPRs) are used to enable the debug modes, to configure and record
debug events, and to communicate with debug tool hardware and software. These debug registers may be
accessed either through software running on the processor or through the JTAG debug port of the
PPC440x5.
Programming Note: It is the responsibility of software to synchronize the context of any
changes to the debug facility registers. Specifically, when changing the
contents of any of the debug facility registers, software must execute an
isync instruction both before and after the changes to these registers, to
ensure that all preceding instructions use the old values of the registers,
and that all succeeding instructions use the new values. In addition, when
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
changing any of the debug facility register fields related to the DAC and/or
DVC debug events, software must execute an msync instruction before
making the changes, to ensure that all storage accesses complete using
the old context of these register fields.
8.6.1 Debug Control Register 0 (DBCR0)
DBCR0 is an SPR that is used to enable debug modes and events, reset the processor, and control timer
operation when debugging. DBCR0 can be written from a GPR using mtspr, and can be read into a GPR
using mfspr.
EDM
0
RST
BRT TRAP IAC2 IAC4 DAC1WDAC2W
10 11 12 13 14 15 16 17
ICMP IRPT IAC1 IAC3 DAC1R DAC2R RET
FT
1
2
3
4
5
6
7
8
9
30 31
IDM
Figure 8-1. Debug Control Register 0 (DBCR0)
External Debug Mode
0 Disable external debug mode.
1 Enable external debug mode.
0
1
EDM
IDM
Internal Debug Mode
0 Disable internal debug mode.
1 Enable internal debug mode.
Reset
00 No action
01 Core reset
10 Chip reset
2:3
RST
11 System reset
Attention: Writing 01, 10, or 11 to this field causes a processor reset to occur.
Instruction completions do not cause
instruction completion debug events if
MSR[DE] = 0 in internal debug mode,
unless also in external debug mode or
debug wait mode.
Instruction Completion Debug Event
0 Disable instruction completion debug event.
1 Enable instruction completion debug event.
4
ICMP
Taken branches do not cause branch
taken debug events if MSR[DE] = 0 in
internal debug mode, unless also in
external debug mode or debug wait
mode.
Branch Taken Debug Event
0 Disable branch taken debug event.
1 Enable branch taken debug event.
5
6
BRT
Critical interrupts do not cause interrupt
debug events in internal debug mode,
unless also in external debug mode or
debug wait mode.
Interrupt Debug Event
0 Disable interrupt debug event.
1 Enable interrupt debug event.
IRPT
Trap Debug Event
0 Disable trap debug event.
1 Enable trap debug event.
7
8
TRAP
IAC1
Instruction Address Compare (IAC) 1 Debug Event
0 Disable IAC 1 debug event.
1 Enable IAC 1 debug event.
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
IAC 2 Debug Event
0 Disable IAC 2 debug event.
9
IAC2
1 Enable IAC 2 debug event.
IAC 3 Debug Event
0 Disable IAC 3 debug event.
1 Enable IAC 3 debug event.
10
11
12
13
14
15
IAC3
IAC 4 Debug Event
0 Disable IAC 4 debug event.
1 Enable IAC 4 debug event.
IAC4
Data Address Compare (DAC) 1 Read Debug Event
0 Disable DAC 1 read debug event.
1 Enable DAC 1 read debug event.
DAC1R
DAC1W
DAC2R
DAC2W
DAC 1 Write Debug Event
0 Disable DAC 1 write debug event.
1 Enable DAC 1 write debug event.
DAC 2 Read Debug Event
0 Disable DAC 2 read debug event.
1 Enable DAC 2 read debug event.
DAC 2 Write Debug Event
0 Disable DAC 2 write debug event.
1 Enable DAC 2 write debug event.
rfci/rfmci does not cause a return
Return Debug Event
0 Disable return (rfi/rfci/rfmci) debug event.
1 Enable return (rfi/rfci/rfmci) debug event.
debug event if MSR[DE] = 0 in internal
debug mode, unless also in external
debug mode or debug wait mode.
16
RET
FT
17:30
31
Reserved
Freeze timers on debug event
0 Timers are not frozen.
1 Freeze timers if a DBSR field associated with a debug event
is set.
8.6.2 Debug Control Register 1 (DBCR1)
DBCR1 is an SPR that is used to configure IAC debug events. DBCR1 can be written from a GPR using
mtspr, and can be read into a GPR using mfspr.
IAC4US
IAC1US
IAC2US
IAC12M
IAC3US
14 15 16 17 18 19 20 21 22 23 24 25 26
IAC3ER
IAC34M
0
1
2
3
4
5
6
7
8
9
10
30 31
IAC12AT
IAC1ER
IAC4ER
IAC2ER
IAC34AT
Figure 8-2. Debug Control Register 1 (DBCR1)
Instruction Address Compare (IAC) 1 User/Super-
visor
00 Both
01 Reserved
0:1
IAC1US
10 Supervisor only (MSR[PR] = 0)
11 User only (MSR[PR] = 1)
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
IAC 1 Effective/Real
00 Effective (MSR[IS] = don’t care)
01 Reserved
10 Virtual (MSR[IS] = 0)
11 Virtual (MSR[IS] = 1)
2:3
IAC1ER
IAC2US
IAC2ER
IAC 2 User/Supervisor
00 Both
01 Reserved
10 Supervisor only (MSR[PR] = 0)
11 User only (MSR[PR] = 1)
4:5
6:7
IAC 2 Effective/Real
00 Effective (MSR[IS] = don’t care)
01 Reserved
10 Virtual (MSR[IS] = 0)
11 Virtual (MSR[IS] = 1)
IAC 1/2 Mode
00 Exact match
Match if address[0:29] = IAC 1/2[0:29]; two inde-
pendent compares
8:9
IAC12M
01 Reserved
10 Range inclusive
11 Range exclusive
Match if IAC1 ≤ address < IAC2
Match if address < IAC1 OR address ≥ IAC2
10:14
15
Reserved
IAC 1/2 Auto-Toggle Enable
0 Disable IAC 1/2 auto-toggle
1 Enable IAC 1/2 auto-toggle
IAC12AT
IAC3US
IAC 3 User/Supervisor
00 Both
01 Reserved
10 Supervisor only (MSR[PR] = 0)
11 User only (MSR[PR] = 1)
16:17
18:19
20:21
22:23
IAC 3 Effective/Real
00 Effective (MSR[IS] = don’t care)
01 Reserved
10 Virtual (MSR[IS] = 0)
IAC3ER
IAC4US
IAC4ER
11 Virtual (MSR[IS] = 1)
IAC 4 User/Supervisor
00 Both
01 Reserved
10 Supervisor only (MSR[PR] = 0)
11 User only (MSR[PR] = 1)
IAC 4 Effective/Real
00 Effective (MSR[IS] = don’t care)
01 Reserved
10 Virtual (MSR[IS] = 0)
11 Virtual (MSR[IS] = 1)
IAC 3/4 Mode
00 Exact match
Match if address[0:29] = IAC 3/4[0:29]; two inde-
pendent compares
24:25
26:30
IAC34M
01 Reserved
10 Range inclusive
11 Range exclusive
Match if IAC3 ≤ address < IAC4
Match if address < IAC3 OR address ≥ IAC4
Reserved
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
IAC3/4 Auto-Toggle Enable
0 Disable IAC 3/4 auto-toggle
1 Enable IAC 3/4 auto-toggle
31
IAC34AT
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
8.6.3 Debug Control Register 2 (DBCR2)
DBCR2 is an SPR that is used to configure DAC and DVC debug events. DBCR2 can be written from a GPR
using mtspr, and can be read into a GPR using mfspr.
DAC2US
DAC12M
DVC1BE
DAC1US
DVC1M
0
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16
19 20
23 24
27 28
31
DAC1ER
DAC2ER DAC12A
DVC2M
DVC2BE
Figure 8-3. Debug Control Register 2 (DBCR2)
Data Address Compare (DAC) 1 User/Supervisor
00 Both
01 Reserved
10 Supervisor only (MSR[PR] = 0)
11 User only (MSR[PR] = 1)
0:1
2:3
4:5
6:7
DAC1US
DAC1ER
DAC2US
DAC2ER
DAC 1 Effective/Real
00 Effective (MSR[DS] = don’t care)
01 Reserved
10 Virtual (MSR[DS] = 0)
11 Virtual (MSR[DS] = 1)
DAC 2 User/Supervisor
00 Both
01 Reserved
10 Supervisor only (MSR[PR] = 0)
11 User only (MSR[PR] = 1)
DAC 2 Effective/Real
00 Effective (MSR[DS] = don’t care)
01 Reserved
10 Virtual (MSR[DS] = 0)
11 Virtual (MSR[DS] = 1)
DAC 1/2 Mode
00 Exact match
Match if address[0:31] = DAC 1/2[0:31]; two inde-
pendent compares
01 Address bit mask
8:9
DAC12M
DAC12A
Match if address = DAC1; only compare bits cor-
responding to 1 bits in DAC2
Match if DAC1 ≤ address < DAC2
Match if address < DAC1 OR address ≥ DAC2
10 Range inclusive
11 Range exclusive
DAC 1/2 Asynchronous
0 Debug interrupt caused by DAC1/2 exception
will be synchronous
1 Debug interrupt caused by DAC1/2 exception
will be asynchronous
10
11
Reserved
Data Value Compare (DVC) 1 Mode
00 Reserved
01 AND all bytes enabled by DVC1BE
10 OR all bytes enabled by DVC1BE
12:13
DVC1M
11 AND-OR pairs of bytes enabled by DVC1BE
(0 AND 1) OR (2 AND 3)
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
DVC 2 Mode
00 Reserved
01 AND all bytes enabled by DVC2BE
10 OR all bytes enabled by DVC2BE
14:15
DVC2M
11 AND-OR pairs of bytes enabled by DVC2BE
(0 AND 1) OR (2 AND 3)
16:19
20:23
24:27
28:31
Reserved
DVC1BE
DVC2BE
DVC 1 Byte Enables 0:3
Reserved
DVC 2 Byte Enables 0:3
8.6.4 Debug Status Register (DBSR)
The DBSR contains status on debug events as well as information on the type of the most recent reset. The
status bits are set by the occurrence of debug events, while the reset type information is updated upon the
occurrence of any of the three reset types.
The DBSR is read into a GPR using mfspr. Clearing the DBSR is performed using mtspr by placing a 1 in
the GPR source register in all bit positions which are to be cleared in the DBSR, and a 0 in all other bit posi-
tions. The data written from the GPR to the DBSR is not direct data, but a mask. A 1 clears the bit and a 0
leaves the corresponding DBSR bit unchanged.
IDE
0
MRR BRT TRAP IAC2 IAC4
DAC2W
DAC1W
IAC34ATS
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17
29 30 31
UDE
ICMP IRPT IAC1 IAC3 DAC1R DAC2R
IAC12ATS
RET
Figure 8-4. Debug Status Register (DBSR)
For synchronous debug events in internal debug
mode, this field indicates whether the correspond-
ing Debug interrupt occurs precisely or impre-
cisely
Imprecise Debug Event
0 Debug event ocurred while MSR[DE] = 1
1 Debug event occurred while MSR[DE] = 0
0
1
IDE
Unconditional Debug Event
0 Event didn’t occur
1 Event occurred
UDE
Most Recent Reset
00 No reset has occurred since this field was last
cleared by software.
01 Core reset
10 Chip reset
This field is set upon any processor reset to a
value indicating the type of reset.
2:3
MRR
11 System reset
Instruction Completion Debug Event
0 Event didn’t occur
1 Event occurred
4
5
ICMP
BRT
Branch Taken Debug Event
0 Event didn’t occur
1 Event occurred
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Interrupt Debug Event
0 Event didn’t occur
1 Event occurred
6
IRPT
Trap Debug Event
0 Event didn’t occur
1 Event occurred
7
TRAP
IAC1
IAC 1 Debug Event
0 Event didn’t occur
1 Event occurred
8
IAC 2 Debug Event
0 Event didn’t occur
1 Event occurred
9
IAC2
IAC 3 Debug Event
0 Event didn’t occur
1 Event occurred
10
11
12
13
14
15
IAC3
IAC 4 Debug Event
0 Event didn’t occur
1 Event occurred
IAC4
DAC 1 Read Debug Event
0 Event didn’t occur
1 Event occurred
DAC1R
DAC1W
DAC2R
DAC2W
RET
DAC 1 Write Debug Event
0 Event didn’t occur
1 Event occurred
DAC 2 Read Debug Event
0 Event didn’t occur
1 Event occurred
DAC 2 Write Debug Event
0 Event didn’t occur
1 Event occurred
Return Debug Event
0 Event didn’t occur
1 Event occurred
16
17:29
Reserved
IAC 1/2 Auto-Toggle Status
0 Range is not reversed from value specified in
DBCR1[IAC12M]
1 Range is reversed from value specified in
DBCR1[IAC12M]
30
31
IAC12ATS
IAC34ATS
IAC 3/4 Auto-Toggle Status
0 Range is not reversed from value specified in
DBCR1[IAC34M]
1 Range is reversed from value specified in
DBCR1[IAC34M]
8.6.5 Instruction Address Compare Registers (IAC1–IAC4)
The four IAC registers specify the addresses upon which IAC debug events should occur. Each of the IAC
registers can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
0
29 30 31
Figure 8-5. Instruction Address Compare Registers (IAC1–IAC4)
0:29
Instruction Address Compare (IAC) word address
Reserved
30:31
8.6.6 Data Address Compare Registers (DAC1–DAC2)
The two DAC registers specify the addresses upon which DAC (and/or DVC) debug events should occur.
Each of the DAC registers can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
0
31
Figure 8-6. Data Address Compare Registers (DAC1–DAC2)
0:31
Data Address Compare (DAC) byte address
8.6.7 Data Value Compare Registers (DVC1–DVC2)
The DVC registers specify the data values upon which DVC debug events should occur. Each of the DVC
registers can be written from a GPR using mtspr, and can be read into a GPR using mfspr.
0
31
Figure 8-7. Data Value Compare Registers (DVC1–DVC2)
0:31
Data value to compare
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
8.6.8 Debug Data Register (DBDR)
The DBDR can be used for communication between software running on the processor and debug tool hard-
ware and software. The DBDR can be written from a GPR using mtspr, and can be read into a GPR using
mfspr.
0
31
Figure 8-8. Debug Data Register (DBDR)
0:31
Debug Data
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
debug.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
Preliminary
PPC440x5 CPU Core User’s Manual
9. Instruction Set
Descriptions of the PPC440x5 instructions follow. Each description contains the following elements:
• Instruction names (mnemonic and full)
• Instruction syntax
• Instruction format diagram
• Pseudocode description
• Prose description
• Registers altered
Where appropriate, instruction descriptions list invalid instruction forms and exceptions, and provide
programming notes.
Table 9-1 summarizes the PPC440x5 instruction set by category.
Table 9-1. Instruction Categories
Category
Sub-Category
Integer Storage Access
Integer Arithmetic
Instruction Types
load, store
add, subtract, multiply, divide, negate
and, andc, or, orc, xor, nand, nor, xnor, extend sign, count lead-
ing zeros
Integer Logical
Integer
Integer Compare
Integer Trap
compare, compare logical
trap
Integer Rotate
Integer Shift
rotate and insert, rotate and mask
shift left, shift right, shift right algebraic
select operand
Integer Select
branch, branch conditional, branch to link, branch to count
crand, crandc, cror, crorc, crnand, crnor, crxor, crxnor
Branch
Condition Register Logical
Register Management
Processor Control
move to/from SPR, move to/from DCR, move to/from MSR,
write to external interrupt enable bit, move to/from CR
system call, return from interrupt, return from critical interrupt,
return from machine check interrupt
System Linkage
Processor Synchronization
Cache Management
instruction synchronize
data allocate, data invalidate, data touch, data zero, data flush,
data store, instruction invalidate, instruction touch
Storage Control
TLB Management
read, write, search, synchronize
Storage Synchronization
memory synchronize, memory barrier
multiply-accumulate, negative multiply-accumulate, multiply
halfword
Allocated Arithmetic
Allocated Logical
detect left-most zero byte
Allocated
data congruence-class invalidate, instruction congruence-class
invalidate
Allocated Cache Management
Allocated Cache Debug
data read, instruction read
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
PPC440x5 CPU Core User’s Manual
Preliminary
9.1 Instruction Set Portability
To support embedded real-time applications, the PPC440x5 core implements the defined instruction set of
the Book-E Enhanced PowerPC Architecture, with the exception of those operations which are defined for
64-bit implementations only, and those which are defined as floating-point operations. Support for the
floating-point operations is provided via the auxiliary processor interface, while the 64-bit operations are not
supported at all. See Instruction Classes on page 53 for more information on the support for defined instruc-
tions within the PPC440x5.
The PPC440x5 core also implements a number of instructions that are not part of PowerPC Book-E architec-
ture, but are included as part of the PPC440x5. Architecturally, they are considered allocated instructions, as
they use opcodes which are within the allocated class of instructions, which the PowerPC Book-E architec-
ture identifies as being available for implementation-dependent and/or application-specific purposes.
However, all of the allocated instructions which are implemented within the PPC440x5 core are “standard” for
IBM PowerPC 400 Series family of embedded controllers, and are not unique to the PPC440x5.
The allocated instructions implemented within the PPC440x5 are divided into four sub-categories, and are
shown in Table 9-2. Programs using these instructions may not be portable to other PowerPC Book-E imple-
mentations.
Table 9-2. Allocated Instructions
Cache
Management
Cache
Debug
Arithmetic
Logical
Negative
Multiply-Accumulate
Multiply-Accumulate
Multiply Halfword
macchw[o][.]
macchws[o][.]
macchwsu[o][.]
macchwu[o][.]
machhw[o][.]
machhws[o][.]
machhwsu[o][.]
machhwu[o][.]
maclhw[o][.]
nmacchw[o][.]
nmacchws[o][.]
nmachhw[o][.]
nmachhws[o][.]
nmaclhw[o][.]
nmaclhws[o][.]
mulchw[.]
mulchwu[.]
mulhhw[.]
mulhhwu[.]
mullhw[.]
dccci
iccci
dcread
icread
dlmzb[.]
mullhwu[.]
maclhws[o][.]
maclhwsu[o][.]
maclhwu[o][.]
9.2 Instruction Formats
For more detailed information about instruction formats, including a summary of instruction field usage and
Instructions are four bytes long. Instruction addresses are always word-aligned.
Instruction bits 0 through 5 always contain the primary opcode. Many instructions have an extended opcode
field as well. The remaining instruction bits contain additional fields. All instruction fields belong to one of the
following categories:
• Defined
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
Preliminary
PPC440x5 CPU Core User’s Manual
These instruction fields contain values, such as opcodes, that cannot be altered. The instruction format
diagrams specify the values of defined fields.
• Variable
These fields contain operands, such as general purpose register specifiers and immediate values, each
of which may contain any one of a number of values. The instruction format diagrams specify the field
names of variable fields.
• Reserved
Bits in a reserved field should be set to 0. In the instruction format diagrams, reserved fields are shaded.
If any bit in a defined field does not contain the specified value, the instruction is illegal and an Illegal Instruc-
tion exception type Program interrupt occurs. If any bit in a reserved field does not contain 0, the instruction
form is invalid and its result is architecturally undefined. Unless otherwise noted, the PPC440x5 core will
execute all invalid instruction forms without causing an Illegal Instruction exception.
9.3 Pseudocode
The pseudocode that appears in the instruction descriptions provides a semi-formal language for describing
instruction operations.
The pseudocode uses the following notation:
+
Twos complement addition
%
Remainder of an integer division; (33 % 32) = 1.
Unsigned comparison relations
The contents of GPR r, where 0 ≤ r ≤ 31.
The contents of the register RA or 0, if the RA field is 0.
The contents of a GPR, where x is A, B, S, or T
A binary number
u
>
u
<
,
(GPR(r))
(RA|0)
(Rx)
0bn
0xn
A hexadecimal number
<, >
Signed comparison relations
Assignment
=
=, ≠
Equal, not equal relations
CEIL(x)
CIA
Least integer ≥ x.
Current instruction address; the 32-bit address of the instruction being
described by a sequence of pseudocode. This address is used to set the
next instruction address (NIA). Does not correspond to any architected
register.
DCR(DCRN)
A Device Control Register (DCR) specified by the DCRF field in an
mfdcr or mtdcr instruction
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
PPC440x5 CPU Core User’s Manual
Preliminary
EA
Effective address; the 32-bit address, derived by applying indexing or
indirect addressing rules to the specified operand, that specifies an
location in main storage.
EXTS(x)
FLD
The result of extending x on the left with sign bits.
An instruction or register field
FLD
FLD
FLD
A bit in a named instruction or register field
A list of bits, by number or name, in a named instruction or register field
A range of bits in a named instruction or register field
General Purpose Register (GPR) r, where 0 ≤ r ≤ 31.
b
b,b, . . .
b:b
GPR(r)
GPRs
RA, RB,
. . .
MASK(MB,ME)
Mask having 1s in positions MB through ME (wrapping if MB > ME) and
0s elsewhere.
MS(addr, n)
NIA
The number of bytes represented by n at the location in main storage
represented by addr.
Next instruction address; the 32-bit address of the next instruction to be
executed. In pseudocode, a successful branch is indicated by assigning
a value to NIA. For instructions that do not branch, the NIA is CIA +4.
PC
Program counter.
REG[FLD, FLD
REG[FLD:FLD]
REG[FLD]
]
A list of fields in a named register
A range of fields in a named register
A field in a named register
. . .
REG
REG
REG
A bit in a named register
b
A list of bits, by number or name, in a named register
A range of bits in a named register
b,b, . . .
b:b
RESERVE
Reserve bit; indicates whether a process has reserved a block of
storage.
ROTL((RS),n)
SPR(SPRN)
Rotate left; the contents of RS are shifted left the number of bits
specified by n.
A Special Purpose Register (SPR) specified by the SPRF field in an
mfspr or mtspr instruction
c
A four-bit object used to store condition results in compare instructions.
0:3
do
Do loop. “to” and “by” clauses specify incrementing an iteration variable;
“while” and “until” clauses specify terminating conditions. Indenting
indicates the scope of a loop.
if...then...else...
Conditional execution; if condition then a else b, where a and b represent
one or more pseudocode statements. Indenting indicates the ranges of a
and b. If b is null, the else does not appear.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
Preliminary
PPC440x5 CPU Core User’s Manual
instruction(EA)
An instruction operating on a data or instruction cache block associated
with an EA.
leave
n
Leave innermost do loop or do loop specified in a leave statement.
A decimal number
n
b
The bit or bit value b is replicated n times.
Bit positions which are don’t-cares.
Concatenation
xx
||
×
÷
⊕
–
Multiplication
Division yielding a quotient
Exclusive-OR (XOR) logical operator
Twos complement subtraction, unary minus
NOT logical operator
¬
∧
∨
AND logical operator
OR logical operator
9.3.1 Operator Precedence
Table 9-3. Operator Precedence
Operators
Associativity
Left to right
REG , REG[FLD], function evaluation
b
n
b
Right to left
Right to left
Left to right
Left to right
Left to right
Left to right
Left to right
Left to right
None
¬, – (unary minus)
×, ÷
+, –
||
u
>
u
<
=, ≠, <, >,
∧, ⊕
∨
,
←
9.4 Register Usage
Each instruction description lists the registers altered by the instruction. Some register changes are explicitly
detailed in the instruction description (for example, the target register of a load instruction). Some instructions
also change other registers, but the details of the changes are not included in the instruction descriptions.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
PPC440x5 CPU Core User’s Manual
Preliminary
Common examples of these kinds of register changes include the Condition Register (CR) and the Integer
Exception Register (XER). For discussion of the CR, see Condition Register (CR) on page 67. For discussion
9.5 Alphabetical Instruction Listing
The following pages list the instructions, both defined and allocated, which are implemented within the
PPC440x5 core.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
add
Add
Preliminary
PPC440x5 CPU Core User’s Manual
add
Add
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
add
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
add.
addo
addo.
31
RT
RA
RB
OE
266
Rc
0
6
11
16
21 22
31
(RT) ← (RA) + (RB)
The sum of the contents of register RA and the contents of register RB is placed into register RT.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
addc
Add Carrying
PPC440x5 CPU Core User’s Manual
Preliminary
addc
Add Carrying
addc
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
addc.
addco
addco.
31
RT
RA
RB
OE
10
Rc
0
6
11
16
21 22
31
(RT) ← (RA) + (RB)
u
>
32
if (RA) + (RB)
2
– 1 then
XER[CA] ← 1
else
XER[CA] ← 0
The sum of the contents of register RA and register RB is placed into register RT.
XER[CA] is set to a value determined by the unsigned magnitude of the result of the add operation.
Registers Altered
• RT
• XER[CA]
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
adde
Add Extended
Preliminary
PPC440x5 CPU Core User’s Manual
adde
Add Extended
adde
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
adde.
addeo
addeo.
31
RT
RA
RB
OE
138
Rc
0
6
11
16
21 22
31
(RT) ← (RA) + (RB) + XER[CA]
u
>
32
if (RA) + (RB) + XER[CA]
XER[CA] ← 1
else
2
– 1 then
XER[CA] ← 0
The sum of the contents of register RA, register RB, and XER[CA] is placed into register RT.
XER[CA] is set to a value determined by the unsigned magnitude of the result of the add operation.
Registers Altered
• RT
• XER[CA]
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
addi
Add Immediate
PPC440x5 CPU Core User’s Manual
Preliminary
addi
Add Immediate
addi
RT, RA, IM
14
RT
RA
IM
0
6
11
16
31
(RT) ← (RA|0) + EXTS(IM)
If the RA field is 0, the IM field, sign-extended to 32 bits, is placed into register RT.
If the RA field is nonzero, the sum of the contents of register RA and the contents of the IM field, sign-
extended to 32 bits, is placed into register RT.
Registers Altered
• RT
Programming Note
To place an immediate, sign-extended value into the GPR specified by RT, set RA = 0.
Table 9-4. Extended Mnemonics for addi
Other Registers
Mnemonic
Operands
Function
Altered
Load address (RA ≠ 0); D is an offset from a base address that
is assumed to be (RA).
(RT) ← (RA) + EXTS(D)
RT, D(RA)
la
Extended mnemonic for
addiRT,RA,D
Load immediate.
(RT) ← EXTS(IM)
RT, IM
li
Extended mnemonic for
addi RT,0,IM
Subtract EXTS(IM) from (RA|0).
Place result in RT.
Extended mnemonic for
RT, RA, IM
subi
addi RT,RA,−IM
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
addic
Add Immediate Carrying
Preliminary
PPC440x5 CPU Core User’s Manual
addic
Add Immediate Carrying
addic
RT, RA, IM
12
RT
RA
IM
0
6
11
16
31
(RT) ← (RA) + EXTS(IM)
u
>
32
if (RA) + EXTS(IM)
XER[CA] ← 1
else
2
– 1 then
XER[CA] ← 0
The sum of the contents of register RA and the contents of the IM field, sign-extended to 32 bits, is placed
into register RT.
XER[CA] is set to a value determined by the unsigned magnitude of the result of the add operation.
Registers Altered
• RT
• XER[CA]
Table 9-5. Extended Mnemonics for addic
Other Registers
Mnemonic
Operands
Function
Subtract EXTS(IM) from (RA)
Altered
Place result in RT; place carry-out in XER[CA].
Extended mnemonic for
RT, RA, IM
subic
addic RT,RA,−IM
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
addic.
Add Immediate Carrying and Record
PPC440x5 CPU Core User’s Manual
Preliminary
addic.
Add Immediate Carrying and Record
addic.
RT, RA, IM
13
RT
RA
IM
0
6
11
16
31
(RT) ← (RA) + EXTS(IM)
u
>
32
if (RA) + EXTS(IM)
XER[CA] ← 1
else
2
– 1 then
XER[CA] ← 0
The sum of the contents of register RA and the contents of the IM field, sign-extended to 32 bits, is placed
into register RT.
XER[CA] is set to a value determined by the unsigned magnitude of the result of the add operation.
Registers Altered
• RT
• XER[CA]
• CR[CR0]
Programming Note
addic. is one of three instructions that implicitly update CR[CR0] without having an RC field. The other
instructions are andi. and andis..
Table 9-6. Extended Mnemonics for addic.
Other Registers
Mnemonic
Operands
Function
Subtract EXTS(IM) from (RA).
Altered
Place result in RT; place carry-out in XER[CA].
Extended mnemonic for
RT, RA, IM
CR[CR0]
subic.
addic. RT,RA,−IM
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
addis
Add Immediate Shifted
Preliminary
PPC440x5 CPU Core User’s Manual
addis
Add Immediate Shifted
addis
RT, RA, IM
15
RT
RA
IM
0
6
11
16
31
16
(RT) ← (RA|0) + (IM || 0)
If the RA field is 0, the IM field is concatenated on its right with sixteen 0-bits and placed into register RT.
If the RA field is nonzero, the contents of register RA are added to the contents of the extended IM field. The
sum is stored into register RT.
Registers Altered
• RT
Programming Note
An addi instruction stores a sign-extended 16-bit value in a GPR. An addis instruction followed by an ori
instruction stores an arbitrary 32-bit value in a GPR, as shown in the following example:
addis
ori
RT, 0, high 16 bits of value
RT, RT, low 16 bits of value
Table 9-7. Extended Mnemonics for addis
Other Registers
Altered
Mnemonic
Operands
Function
Load immediate shifted.
16
(RT) ← (IM ||
0)
RT, IM
lis
Extended mnemonic for
addis RT,0,IM
16
Subtract (IM ||
0) from (RA|0).
Place result in RT.
RT, RA, IM
subis
Extended mnemonic for
addis RT,RA,−IM
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
addme
Add to Minus One Extended
PPC440x5 CPU Core User’s Manual
Preliminary
addme
Add to Minus One Extended
addme
RT, RA
RT, RA
RT, RA
RT, RA
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
addme.
addmeo
addmeo.
31
RT
RA
OE
234
Rc
0
6
11
16
21 22
31
(RT) ← (RA) + XER[CA] + (–1)
if (RA) + XER[CA] + 0xFFFF FFFF
XER[CA] ← 1
u
>
32
2
– 1 then
else
XER[CA] ← 0
The sum of the contents of register RA, XER[CA], and –1 is placed into register RT.
XER[CA] is set to a value determined by the unsigned magnitude of the result of the add operation.
Registers Altered
• RT
• XER[CA]
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
addze
Add to Zero Extended
Preliminary
PPC440x5 CPU Core User’s Manual
addze
Add to Zero Extended
addze
RT, RA
RT, RA
RT, RA
RT, RA
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
addze.
addzeo
addzeo.
31
RT
RA
OE
202
Rc
0
6
11
16
21 22
31
(RT) ← (RA) + XER[CA]
u
>
32
if (RA) + XER[CA]
XER[CA] ← 1
else
2
– 1 then
XER[CA] ← 0
The sum of the contents of register RA and XER[CA] is placed into register RT.
XER[CA] is set to a value determined by the unsigned magnitude of the result of the add operation.
Registers Altered
• RT
• XER[CA]
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
and
AND
PPC440x5 CPU Core User’s Manual
Preliminary
and
AND
and
and.
RA, RS, RB
RA, RS, RB
Rc=0
Rc=1
31
RS
RA
RB
28
Rc
0
6
11
16
21
31
(RA) ← (RS) ∧ (RB)
The contents of register RS are ANDed with the contents of register RB; the result is placed into register RA.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
andc
AND with Complement
Preliminary
PPC440x5 CPU Core User’s Manual
andc
AND with Complement
andc
andc.
RA,RS,RB
RA,RS,RB
Rc=0
Rc=1
31
RS
RA
RB
60
Rc
0
6
11
16
21
2
31
(RA) ← (RS) ∧ ¬(RB)
The contents of register RS are ANDed with the ones complement of the contents of register RB; the result is
placed into register RA.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
andi.
AND Immediate
PPC440x5 CPU Core User’s Manual
Preliminary
andi.
AND Immediate
andi.
RA, RS, IM
28
RS
RA
IM
0
6
11
16
31
16
(RA) ← (RS) ∧ ( 0 || IM)
The IM field is extended to 32 bits by concatenating 16 0-bits on its left. The contents of register RS is ANDed
with the extended IM field; the result is placed into register RA.
Registers Altered
• RA
• CR[CR0]
Programming Note
The andi. instruction can test whether any of the 16 least-significant bits in a GPR are 1-bits.
andi. is one of three instructions that implicitly update CR[CR0] without having an Rc field. The other instruc-
tions are addic. and andis..
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
andis.
AND Immediate Shifted
Preliminary
PPC440x5 CPU Core User’s Manual
andis.
AND Immediate Shifted
andis.
RA, RS, IM
29
RS
RA
IM
0
6
11
16
31
16
(RA) ← (RS) ∧ (IM || 0)
The IM field is extended to 32 bits by concatenating 16 0-bits on its right. The contents of register RS are
ANDed with the extended IM field; the result is placed into register RA.
Registers Altered
• RA
• CR[CR0]
Programming Note
The andis. instruction can test whether any of the 16 most-significant bits in a GPR are 1-bits.
andis. is one of three instructions that implicitly update CR[CR0] without having an Rc field. The other
instructions are addic. and andi..
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
b
Branch
PPC440x5 CPU Core User’s Manual
Preliminary
b
Branch
b
ba
bl
target
target
target
target
AA=0, LK=0
AA=1, LK=0
AA=0, LK=1
AA=1, LK=1
bla
18
LI
AA LK
0
6
30 31
If AA = 1 then
LI ← target
6:29
2
NIA ← EXTS(LI || 0)
else
LI ← (target – CIA)
6:29
2
NIA ← CIA + EXTS(LI || 0)
if LK = 1 then
(LR) ← CIA + 4
PC ← NIA
The next instruction address (NIA) is the effective address of the branch target. The NIA is formed by adding
a displacement to a base address. The displacement is obtained by concatenating two 0-bits to the right of
the LI field and sign-extending the result to 32 bits.
If the AA field contains 0, the base address is the address of the branch instruction, which is the current
instruction address (CIA). If the AA field contains 1, the base address is 0.
Instruction execution resumes with the instruction at the NIA.
If the LK field contains 1, then (CIA + 4) is placed into the LR.
Registers Altered
• LR if LK contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bc
Branch Conditional
Preliminary
PPC440x5 CPU Core User’s Manual
bc
Branch Conditional
bc
bca
bcl
BO, BI, target
BO, BI, target
BO, BI, target
BO, BI, target
AA=0, LK=0
AA=1, LK=0
AA=0, LK=1
AA=1, LK=1
bcla
16
BO
BI
BD
AA LK
30 31
0
6
11
16
if BO = 0 then
2
CTR ← CTR – 1
if (BO = 1 ∨ ((CTR = 0) = BO )) ∧ (BO = 1 ∨ (CR = BO )) then
2
3
0
BI
1
if AA = 1 then
BD ← target
16:29
2
NIA ← EXTS(BD || 0)
else
BD ← (target – CIA)
16:29
2
NIA ← CIA + EXTS(BD || 0)
else
NIA ← CIA + 4
if LK = 1 then
(LR) ← CIA + 4
PC ← NIA
If BO contains 0, the CTR decrements, and the decremented value is tested for 0 as part of the branch
2
condition. In this case, BO indicates whether the test for 0 must be true or false in order for the branch to be
3
taken. If BO contains 1, then the CTR is neither decremented nor tested as part of the branch condition.
2
If BO contains 0, then the CR bit specified by the BI field is compared to BO as part of the branch condition.
0
1
If BO contains 1, then the CR is not tested as part of the branch condition, and the BI field is ignored.
0
The next instruction address (NIA) is either the effective address of the branch target, or the address of the
instruction after the branch, depending on whether the branch is taken or not. The branch target address is
formed by adding a displacement to a base address. The displacement is obtained by concatenating two 0-
bits to the right of the BD field and sign-extending the result to 32 bits.
If the AA field contains 0, the base address is the address of the branch instruction, which is the current
instruction address (CIA). If the AA field contains 1, the base address is 0.
BO affects branch prediction, a performance-improvement feature. See Branch Prediction on page 65 for a
4
complete discussion.
Instruction execution resumes with the instruction at the NIA.
If the LK field contains 1, then (CIA + 4) is placed into the LR.
Registers Altered
• CTR if BO contains 0
2
• LR if LK contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bc
Branch Conditional
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-8. Extended Mnemonics for bc, bca, bcl, bcla
Other Registers
Mnemonic
Operands
Function
Altered
Decrement CTR; branch if CTR ≠ 0.
Extended mnemonic for
bdnz
bc 16,0,target
Extended mnemonic for
bca 16,0,target
bdnza
bdnzl
target
Extended mnemonic for
bcl 16,0,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 16,0,target
bdnzla
Decrement CTR.
Branch if CTR ≠ 0 AND CR
= 0.
cr_bit
bdnzf
Extended mnemonic for
bc 0,cr_bit,target
Extended mnemonic for
bca 0,cr_bit,target
bdnzfa
bdnzfl
bdnzfla
cr_bit, target
Extended mnemonic for
bcl 0,cr_bit,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 0,cr_bit,target
Decrement CTR.
Branch if CTR ≠ 0 AND CR
= 1.
cr_bit
bdnzt
Extended mnemonic for
bc 8,cr_bit,target
Extended mnemonic for
bca 8,cr_bit,target
bdnzta
bdnztl
bdnztla
cr_bit, target
Extended mnemonic for
bcl 8,cr_bit,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 8,cr_bit,target
Decrement CTR; branch if CTR = 0.
Extended mnemonic for
bdz
bc 18,0,target
Extended mnemonic for
bca 18,0,target
bdza
bdzl
target
Extended mnemonic for
bcl 18,0,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 18,0,target
bdzla
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bc
Branch Conditional
Preliminary
PPC440x5 CPU Core User’s Manual
Table 9-8. Extended Mnemonics for bc, bca, bcl, bcla (continued)
Other Registers
Altered
Mnemonic
Operands
Function
Decrement CTR
Branch if CTR = 0 AND CR
Extended mnemonic for
= 0.
cr_bit
bdzf
bc 2,cr_bit,target
Extended mnemonic for
bca 2,cr_bit,target
bdzfa
bdzfl
cr_bit, target
Extended mnemonic for
bcl 2,cr_bit,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 2,cr_bit,target
bdzfla
Decrement CTR.
Branch if CTR = 0 AND CR
Extended mnemonic for
= 1.
cr_bit
bdzt
bc 10,cr_bit,target
Extended mnemonic for
bca 10,cr_bit,target
bdzta
bdztl
cr_bit, target
Extended mnemonic for
bcl 10,cr_bit,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 10,cr_bit,target
bdztla
Branch if equal.
Use CR0 if cr_field is omitted.
Extended mnemonic for
beq
bc 12,4∗cr_field+2,target
Extended mnemonic for
bca 12,4∗cr_field+2,target
beqa
beql
[cr_field,] target
Extended mnemonic for
bcl 12,4∗cr_field+2,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 12,4∗cr_field+2,target
beqla
Branch if CR
= 0.
Extendedcrm_bnitemonic for
bf
bc 4,cr_bit,target
Extended mnemonic for
bca 4,cr_bit,target
bfa
bfl
cr_bit, target
Extended mnemonic for
bcl 4,cr_bit,target
LR
LR
Extended mnemonic for
bcla 4,cr_bit,target
bfla
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bc
Branch Conditional
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-8. Extended Mnemonics for bc, bca, bcl, bcla (continued)
Other Registers
Altered
Mnemonic
Operands
Function
Branch if greater than or equal.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bge
bc 4,4∗cr_field+0,target
Extended mnemonic for
bca 4,4∗cr_field+0,target
bgea
bgel
[cr_field,] target
Extended mnemonic for
bcl 4,4∗cr_field+0,target
LR
LR
Extended mnemonic for
bcla 4,4∗cr_field+0,target
bgela
Branch if greater than.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bgt
bc 12,4∗cr_field+1,target
Extended mnemonic for
bca 12,4∗cr_field+1,target
bgta
bgtl
[cr_field,] target
[cr_field,] target
[cr_field,] target
Extended mnemonic for
bcl 12,4∗cr_field+1,target
LR
LR
Extended mnemonic for
bcla 12,4∗cr_field+1,target
bgtla
Branch if less than or equal.
Use CR0 if cr_field is omitted.
Extended mnemonic for
ble
bc 4,4∗cr_field+1,target
Extended mnemonic for
bca 4,4∗cr_field+1,target
blea
blel
Extended mnemonic for
bcl 4,4∗cr_field+1,target
LR
LR
Extended mnemonic for
bcla 4,4∗cr_field+1,target
blela
Branch if less than
Use CR0 if cr_field is omitted.
Extended mnemonic for
blt
bc 12,4∗cr_field+0,target
Extended mnemonic for
bca 12,4∗cr_field+0,target
blta
bltl
Extended mnemonic for
bcl 12,4∗cr_field+0,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 12,4∗cr_field+0,target
bltla
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bc
Branch Conditional
Preliminary
PPC440x5 CPU Core User’s Manual
Table 9-8. Extended Mnemonics for bc, bca, bcl, bcla (continued)
Other Registers
Altered
Mnemonic
Operands
Function
Branch if not equal.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bne
bc 4,4∗cr_field+2,target
Extended mnemonic for
bca 4,4∗cr_field+2,target
bnea
bnel
[cr_field,] target
Extended mnemonic for
bcl 4,4∗cr_field+2,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 4,4∗cr_field+2,target
bnela
Branch if not greater than.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bng
bc 4,4∗cr_field+1,target
Extended mnemonic for
bca 4,4∗cr_field+1,target
bnga
bngl
[cr_field,] target
[cr_field,] target
[cr_field,] target
Extended mnemonic for
bcl 4,4∗cr_field+1,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 4,4∗cr_field+1,target
bngla
Branch if not less than; use CR0 if cr_field is omitted.
Extended mnemonic for
bc 4,4∗cr_field+0,target
bnl
Extended mnemonic for
bca 4,4∗cr_field+0,target
bnla
bnll
Extended mnemonic for
bcl 4,4∗cr_field+0,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 4,4∗cr_field+0,target
bnlla
Branch if not summary overflow.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bns
bc 4,4∗cr_field+3,target
Extended mnemonic for
bca 4,4∗cr_field+3,target
bnsa
bnsl
Extended mnemonic for
bcl 4,4∗cr_field+3,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 4,4∗cr_field+3,target
bnsla
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bc
Branch Conditional
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-8. Extended Mnemonics for bc, bca, bcl, bcla (continued)
Other Registers
Mnemonic
Operands
Function
Altered
Branch if not unordered.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bnu
bc 4,4∗cr_field+3,target
Extended mnemonic for
bca 4,4∗cr_field+3,target
bnua
bnul
[cr_field,] target
Extended mnemonic for
bcl 4,4∗cr_field+3,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 4,4∗cr_field+3,target
bnula
Branch if summary overflow.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bso
bc 12,4∗cr_field+3,target
Extended mnemonic for
bca 12,4∗cr_field+3,target
bsoa
bsol
[cr_field,] target
Extended mnemonic for
bcl 12,4∗cr_field+3,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 12,4∗cr_field+3,target
bsola
Branch if CR
= 1.
Extendedcrm_bnitemonic for
bt
bc 12,cr_bit,target
Extended mnemonic for
bca 12,cr_bit,target
bta
btl
cr_bit, target
Extended mnemonic for
bcl 12,cr_bit,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 12,cr_bit,target
btla
Branch if unordered.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bun
bc 12,4∗cr_field+3,target
Extended mnemonic for
bca 12,4∗cr_field+3,target
buna
bunl
[cr_field], target
Extended mnemonic for
bcl 12,4∗cr_field+3,target
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Extended mnemonic for
bcla 12,4∗cr_field+3,target
bunla
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bcctr
Branch Conditional to Count Register
PPC440x5 CPU Core User’s Manual
Preliminary
bcctr
Branch Conditional to Count Register
bcctr
bcctrl
BO, BI
BO, BI
LK=0
LK=1
19
BO
BI
528
LK
0
6
11
16
21
31
if (BO = 1 ∨ (CR = BO )) then
0
BI
|| 0
1
2
NIA ← CTR
0:29
else
NIA ← CIA + 4
if LK = 1 then
(LR) ← CIA + 4
PC ← NIA
If BO contains 0, then the CR bit specified by the BI field is compared to BO as part of the branch condition.
0
1
If BO contains 1, then the CR is not tested as part of the branch condition, and the BI field is ignored.
0
The next instruction address (NIA) is either the effective address of the branch target, or the address of the
instruction after the branch, depending on whether the branch is taken or not. The branch target address is
formed by concatenating two 0-bits to the right of the 30 most significant bits of the CTR.
BO affects branch prediction, a performance-improvement feature. See Branch Prediction on page 65 for a
4
complete discussion.
Instruction execution resumes with the instruction at the NIA.
If the LK field contains 1, then (CIA + 4) is placed into the LR.
Registers Altered
• LR if LK contains 1
Invalid Instruction Forms
• Reserved fields
• If BO contains 0, the instruction form is invalid, and the result of the instruction (in particular, the branch
2
target address and whether or not the branch is taken) is undefined. The architecture does not permit the
combination of decrementing the CTR as part of the branch condition, together with using the CTR as the
branch target address.
Table 9-9. Extended Mnemonics for bcctr, bcctrl
Other Registers
Mnemonic
Operands
Function
Altered
Branch unconditionally to address in CTR.
Extended mnemonic for
bctr
bcctr 20,0
Extended mnemonic for
bcctrl 20,0
(LR) ← CIA + 4.
bctrl
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bcctr
Branch Conditional to Count Register
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-9. Extended Mnemonics for bcctr, bcctrl (continued)
Other Registers
Mnemonic
Operands
Function
Altered
Branch, if equal, to address in CTR
Use CR0 if cr_field is omitted.
Extended mnemonic for
beqctr
[cr_field]
bcctr 12,4∗cr_field+2
Extended mnemonic for
bcctrl 12,4∗cr_field+2
(LR) ← CIA + 4.
beqctrl
bfctr
Branch, if CR
= 0, to address in CTR.
Extendedcmr_bnitemonic for
bcctr 4,cr_bit
cr_bit
Extended mnemonic for
bcctrl 4,cr_bit
(LR) ← CIA + 4.
bfctrl
Branch, if greater than or equal, to address in CTR. Use CR0 if
cr_field is omitted.
Extended mnemonic for
bcctr 4,4∗cr_field+0
bgectr
bgectrl
bgtctr
bgtctrl
blectr
blectrl
bltctr
[cr_field]
Extended mnemonic for
bcctrl 4,4∗cr_field+0
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Branch, if greater than, to address in CTR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
[cr_field]
[cr_field]
[cr_field]
[cr_field]
bcctr 12,4∗cr_field+1
Extended mnemonic for
bcctrl 12,4∗cr_field+1
Branch, if less than or equal, to address in CTR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bcctr 4,4∗cr_field+1
Extended mnemonic for
bcctrl 4,4∗cr_field+1
Branch, if less than, to address in CTR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bcctr 12,4∗cr_field+0
Extended mnemonic for
bcctrl 12,4∗cr_field+0
bltctrl
bnectr
bnectrl
Branch, if not equal, to address in CTR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bcctr 4,4∗cr_field+2
Extended mnemonic for
bcctrl 4,4∗cr_field+2
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bcctr
Branch Conditional to Count Register
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-9. Extended Mnemonics for bcctr, bcctrl (continued)
Other Registers
Altered
Mnemonic
Operands
Function
Branch, if not greater than, to address in CTR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bngctr
bngctrl
bnlctr
[cr_field]
bcctr 4,4∗cr_field+1
Extended mnemonic for
bcctrl 4,4∗cr_field+1
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Branch, if not less than, to address in CTR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
[cr_field]
[cr_field]
[cr_field]
bcctr 4,4∗cr_field+0
Extended mnemonic for
bcctrl 4,4∗cr_field+0
bnlctrl
bnsctr
bnsctrl
bnuctr
bnuctrl
bsoctr
Branch, if not summary overflow, to address in CTR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bcctr 4,4∗cr_field+3
Extended mnemonic for
bcctrl 4,4∗cr_field+3
Branch, if not unordered, to address in CTR; use CR0 if cr_field
is omitted.
Extended mnemonic for
bcctr 4,4∗cr_field+3
Extended mnemonic for
bcctrl 4,4∗cr_field+3
(LR) ← CIA + 4.
Branch, if summary overflow, to address in CTR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
[cr_field]
bcctr 12,4∗cr_field+3
Extended mnemonic for
bcctrl 12,4∗cr_field+3
(LR) ← CIA + 4.
bsoctrl
btctr
Branch if CR
= 1 to address in CTR.
Extendedcrm_bnitemonic for
bcctr 12,cr_bit
cr_bit
Extended mnemonic for
bcctrl 12,cr_bit
(LR) ← CIA + 4.
btctrl
Branch if unordered to address in CTR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bunctr
bunctrl
[cr_field]
bcctr 12,4∗cr_field+3
Extended mnemonic for
bcctrl 12,4∗cr_field+3
(LR) ← CIA + 4.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bclr
Branch Conditional to Link Register
PPC440x5 CPU Core User’s Manual
Preliminary
bclr
Branch Conditional to Link Register
bclr
bclrl
BO, BI
BO, BI
LK=0
LK =1
19
BO
BI
16
LK
0
6
11
16
21
31
if BO = 0 then
2
CTR ← CTR – 1
if (BO = 1 ∨ ((CTR = 0) = BO )) ∧ (BO = 1 ∨ (CR = BO )) then
2
3
0
BI
1
2
NIA ← LR
|| 0
0:29
else
NIA ← CIA + 4
if LK = 1 then
(LR) ← CIA + 4
PC ← NIA
If BO contains 0, the CTR decrements, and the decremented value is tested for 0 as part of the branch
2
condition. In this case, BO indicates whether the test for 0 must be true or false in order for the branch to be
3
taken. If BO contains 1, then the CTR is neither decremented nor tested as part of the branch condition.
2
If BO contains 0, then the CR bit specified by the BI field is compared to BO as part of the branch condition.
0
1
If BO contains 1, then the CR is not tested as part of the branch condition, and the BI field is ignored.
0
The next instruction address (NIA) is either the effective address of the branch target, or the address of the
instruction after the branch, depending on whether the branch is taken or not. The branch target address is
formed by concatenating two 0-bits to the right of the 30 most significant bits of the LR.
BO affects branch prediction, a performance-improvement feature. See Branch Prediction on page 65 for a
4
complete discussion.
Instruction execution resumes with the instruction at the NIA.
If the LK field contains 1, then (CIA + 4) is placed into the LR.
Registers Altered
• CTR if BO contains 0
2
• LR if LK contains 1
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bclr
Branch Conditional to Link Register
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-10. Extended Mnemonics for bclr, bclrl
Other Registers
Altered
Mnemonic
Operands
Function
Branch unconditionally to address in LR.
Extended mnemonic for
blr
bclr 20,0
Extended mnemonic for
bclrl 20,0
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
blrl
Decrement CTR.
Branch if CTR ≠ 0 to address in LR.
Extended mnemonic for
bdnzlr
bdnzlrl
bdnzflr
bdnzflrl
bdnztlr
bdnztlrl
bdzlr
bclr 16,0
Extended mnemonic for
bclrl 16,0
Decrement CTR.
Branch if CTR ≠ 0 AND CR
= 0 to address in LR.
cr_bit
Extended mnemonic for
cr_bit
bclr 0,cr_bit
Extended mnemonic for
bclrl 0,cr_bit
Decrement CTR.
Branch if CTR ≠ 0 AND CR
= 1 to address in LR.
cr_bit
Extended mnemonic for
cr_bit
bclr 8,cr_bit
Extended mnemonic for
bclrl 8,cr_bit
Decrement CTR.
Branch if CTR = 0 to address in LR.
Extended mnemonic for
bclr 18,0
Extended mnemonic for
bclrl 18,0
bdzlrl
Decrement CTR.
Branch if CTR = 0 AND CR
= 0 to address in LR.
cr_bit
bdzflr
Extended mnemonic for
bclr 2,cr_bit
cr_bit
cr_bit
Extended mnemonic for
bclrl 2,cr_bit
bdzflrl
bdztlr
Decrement CTR.
Branch if CTR = 0 AND CR
= 1 to address in LR.
cr_bit
Extended mnemonic for
bclr 10,cr_bit
Extended mnemonic for
bclrl 10,cr_bit
(LR) ← CIA + 4.
bdztlrl
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bclr
Branch Conditional to Link Register
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-10. Extended Mnemonics for bclr, bclrl (continued)
Other Registers
Mnemonic
Operands
Function
Branch if equal to address in LR.
Altered
Use CR0 if cr_field is omitted.
Extended mnemonic for
beqlr
[cr_field]
bclr 12,4∗cr_field+2
Extended mnemonic for
bclrl 12,4∗cr_field+2
(LR) ← CIA + 4.
beqlrl
bflr
Branch if CR
= 0 to address in LR.
Extendedcrm_bnitemonic for
bclr 4,cr_bit
cr_bit
Extended mnemonic for
bclrl 4,cr_bit
(LR) ← CIA + 4.
bflrl
Branch, if greater than or equal, to address in LR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bgelr
bgelrl
bgtlr
bgtlrl
blelr
[cr_field]
bclr 4,4∗cr_field+0
Extended mnemonic for
bclrl 4,4∗cr_field+0
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Branch, if greater than, to address in LR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
[cr_field]
[cr_field]
[cr_field]
[cr_field]
bclr 12,4∗cr_field+1
Extended mnemonic for
bclrl 12,4∗cr_field+1
Branch, if less than or equal, to address in LR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bclr 4,4∗cr_field+1
Extended mnemonic for
bclrl 4,4∗cr_field+1
blelrl
bltlr
Branch, if less than, to address in LR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bclr 12,4∗cr_field+0
Extended mnemonic for
bclrl 12,4∗cr_field+0
bltlrl
bnelr
bnelrl
Branch, if not equal, to address in LR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bclr 4,4∗cr_field+2
Extended mnemonic for
bclrl 4,4∗cr_field+2
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
bclr
Branch Conditional to Link Register
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-10. Extended Mnemonics for bclr, bclrl (continued)
Other Registers
Altered
Mnemonic
Operands
Function
Branch, if not greater than, to address in LR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bnglr
bnglrl
bnllr
[cr_field]
bclr 4,4∗cr_field+1
Extended mnemonic for
bclrl 4,4∗cr_field+1
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
(LR) ← CIA + 4.
Branch, if not less than, to address in LR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
[cr_field]
[cr_field]
[cr_field]
bclr 4,4∗cr_field+0
Extended mnemonic for
bclrl 4,4∗cr_field+0
bnllrl
bnslr
bnslrl
bnulr
bnulrl
bsolr
Branch if not summary overflow to address in LR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bclr 4,4∗cr_field+3
Extended mnemonic for
bclrl 4,4∗cr_field+3
Branch if not unordered to address in LR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bclr 4,4∗cr_field+3
Extended mnemonic for
bclrl 4,4∗cr_field+3
Branch if summary overflow to address in LR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
[cr_field]
bclr 12,4∗cr_field+3
Extended mnemonic for
bclrl 12,4∗cr_field+3
(LR) ← CIA + 4.
bsolrl
btlr
Branch if CR
= 1 to address in LR.
Extendedcrm_bnitemonic for
bclr 12,cr_bit
cr_bit
Extended mnemonic for
bclrl 12,cr_bit
(LR) ← CIA + 4.
btlrl
Branch if unordered to address in LR.
Use CR0 if cr_field is omitted.
Extended mnemonic for
bunlr
[cr_field]
bclr 12,4∗cr_field+3
Extended mnemonic for
bclrl 12,4∗cr_field+3
(LR) ← CIA + 4.
bunlrl
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
cmp
Compare
PPC440x5 CPU Core User’s Manual
Preliminary
cmp
Compare
cmp
BF, 0, RA, RB
31
BF
RA
RB
0
0
6
9
11
16
21
31
4
c
←
0
0:3
if (RA) < (RB) then c ← 1
0
if (RA) > (RB) then c ← 1
1
if (RA) = (RB) then c ← 1
2
c ← XER[SO]
3
n ← BF
CR[CRn] ← c
0:3
The contents of register RA are compared with the contents of register RB using a 32-bit signed compare.
The CR field specified by the BF field is updated to reflect the results of the compare and the value of
XER[SO] is placed into the same CR field.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• CR[CRn] where n is specified by the BF field
Invalid Instruction Forms
• Reserved fields
Programming Note
PowerPC Book-E architecture defines this instruction as cmp BF,L,RA,RB, where L selects operand size for
64-bit implementations. For all 32-bit implementations, L = 0 is required (L = 1 is an invalid form); hence for
the PPC440x5 core, use of the extended mnemonic cmpw BF,RA,RB is recommended.
Table 9-11. Extended Mnemonics for cmp
Other Registers
Mnemonic
Operands
Function
Altered
Compare Word; use CR0 if BF is omitted.
Extended mnemonic for
[BF,] RA, RB
cmpw
cmp BF,0,RA,RB
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
cmpi
Compare Immediate
Preliminary
PPC440x5 CPU Core User’s Manual
cmpi
Compare Immediate
cmpi
BF, 0, RA, IM
11
BF
6
RA
IM
0
9
11
16
31
4
c
←
0
0:3
if (RA) < EXTS(IM) then c ← 1
0
if (RA) > EXTS(IM) then c ← 1
1
if (RA) = EXTS(IM) then c ← 1
2
c3 ← XER[SO]
n ← BF
CR[CRn] ← c0:3
The IM field is sign-extended to 32 bits. The contents of register RA are compared with the extended IM field,
using a 32-bit signed compare.
The CR field specified by the BF field is updated to reflect the results of the compare and the value of
XER[SO] is placed into the same CR field.
Registers Altered
• CR[CRn] where n is specified by the BF field
Invalid Instruction Forms
• Reserved fields
Programming Note
PowerPC Book-E Architecture defines this instruction as cmpi BF,L,RA,IM, where L selects operand size for
64-bit implementations. For all 32-bit implementations, L = 0 is required (L = 1 is an invalid form); hence for
the PPC440x5 core, use of the extended mnemonic cmpwi BF,RA,IM is recommended.
Table 9-12. Extended Mnemonics for cmpi
Other Registers
Mnemonic
Operands
Function
Altered
Compare Word Immediate.
Use CR0 if BF is omitted.
Extended mnemonic for
[BF,] RA, IM
cmpwi
cmpi BF,0,RA,IM
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
cmpl
Compare Logical
PPC440x5 CPU Core User’s Manual
Preliminary
cmpl
Compare Logical
cmpl
BF, 0, RA, RB
31
BF
RA
RB
32
0
6
9
11
16
21
31
4
c
←
0
0:3
u
<
u
>
if (RA)
if (RA)
(RB) then c ← 1
0
(RB) then c ← 1
1
if (RA) = (RB) then c ← 1
2
c ← XER[SO]
3
n ← BF
CR[CRn] ← c
0:3
The contents of register RA are compared with the contents of register RB, using a 32-bit unsigned compare.
The CR field specified by the BF field is updated to reflect the results of the compare and the value of
XER[SO] is placed into the same CR field.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• CR[CRn] where n is specified by the BF field
Invalid Instruction Forms
• Reserved fields
Programming Notes
PowerPC Book-E Architecture defines this instruction as cmpl BF,L,RA,RB, where L selects operand size
for 64-bit implementations. For all 32-bit implementations, L = 0 is required (L = 1 is an invalid form); hence
for PPC440x5 core, use of the extended mnemonic cmplw BF,RA,RB is recommended.
Table 9-13. Extended Mnemonics for cmpl
Other Registers
Mnemonic
Operands
Function
Altered
Compare Logical Word.
Use CR0 if BF is omitted.
Extended mnemonic for
[BF,] RA, RB
cmplw
cmpl BF,0,RA,RB
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
cmpli
Compare Logical Immediate
Preliminary
PPC440x5 CPU Core User’s Manual
cmpli
Compare Logical Immediate
cmpli
BF, 0, RA, IM
10
BF
RA
IM
0
6
9
11
16
31
4
u
c
←
0
(
(
0:3
16
16
16
if (RA)
if (RA)
0 || IM) then c ← 1
<
0
u
0 || IM) then c ← 1
>
1
if (RA) = ( 0 || IM) then c ← 1
2
c ← XER[SO]
3
n ← BF
CR[CRn] ← c
0:3
The IM field is extended to 32 bits by concatenating 16 0-bits to its left. The contents of register RA are
compared with IM using a 32-bit unsigned compare.
The CR field specified by the BF field is updated to reflect the results of the compare and the value of
XER[SO] is placed into the same CR field.
Registers Altered
• CR[CRn] where n is specified by the BF field
Invalid Instruction Forms
• Reserved fields
Programming Note
PowerPC Book-E Architecture defines this instruction as cmpli BF,L,RA,IM, where L selects operand size
for 64-bit implementations. For all 32-bit implementations, L = 0 is required (L = 1 is an invalid form); hence
for the PPC440x5 core, use of the extended mnemonic cmplwi BF,RA,IM is recommended.
Table 9-14. Extended Mnemonics for cmpli
Other Registers
Mnemonic
Operands
Function
Changed
Compare Logical Word Immediate.
Use CR0 if BF is omitted.
Extended mnemonic for
[BF,] RA, IM
cmplwi
cmpli BF,0,RA,IM
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
cntlzw
Count Leading Zeros Word
PPC440x5 CPU Core User’s Manual
Preliminary
cntlzw
Count Leading Zeros Word
cntlzw
cntlzw.
RA, RS
RA, RS
Rc=0
Rc=1
31
RS
RA
26
Rc
0
6
11
16
21
31
n ← 0
do while n < 32
if (RS) = 1 then leave
n
n ← n + 1
(RA) ← n
The consecutive leading 0 bits in register RS are counted; the count is placed into register RA.
The count ranges from 0 through 32, inclusive.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
crand
Condition Register AND
Preliminary
PPC440x5 CPU Core User’s Manual
crand
Condition Register AND
crand
BT, BA, BB
19
BT
BA
BB
257
0
6
11
16
21
31
CR
← CR
∧ CR
BB
BT
BA
The CR bit specified by the BA field is ANDed with the CR bit specified by the BB field; the result is placed
into the CR bit specified by the BT field.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• CR
BT
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
crandc
Condition Register AND with Complement
PPC440x5 CPU Core User’s Manual
Preliminary
crandc
Condition Register AND with Complement
crandc
BT, BA, BB
19
BT
BA
BB
129
0
6
11
16
21
31
CR
BT
← CR
∧ ¬CR
BA
BB
The CR bit specified by the BA field is ANDed with the ones complement of the CR bit specified by the BB
field; the result is placed into the CR bit specified by the BT field.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• CR
BT
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
creqv
Condition Register Equivalent
Preliminary
PPC440x5 CPU Core User’s Manual
creqv
Condition Register Equivalent
creqv
BT, BA, BB
19
BT
BA
BB
289
0
6
11
16
21
31
CR
← ¬(CR
⊕ CR
)
BB
BT
BA
The CR bit specified by the BA field is XORed with the CR bit specified by the BB field; the ones complement
of the result is placed into the CR bit specified by the BT field.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• CR
BT
Invalid Instruction Forms
• Reserved fields
Table 9-15. Extended Mnemonics for creqv
Other Registers
Altered
Mnemonic
Operands
Function
CR set.
Extended mnemonic for
crset
bx
creqv bx,bx,bx
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
crnand
Condition Register NAND
PPC440x5 CPU Core User’s Manual
Preliminary
crnand
Condition Register NAND
crnand
BT, BA, BB
19
BT
BA
BB
225
0
6
11
16
21
31
CR
BT
← ¬(CR
∧ CR
)
BA
BB
The CR bit specified by the BA field is ANDed with the CR bit specified by the BB field; the ones complement
of the result is placed into the CR bit specified by the BT field.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• CR
BT
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
crnor
Condition Register NOR
Preliminary
PPC440x5 CPU Core User’s Manual
crnor
Condition Register NOR
crnor
BT, BA, BB
19
BT
BA
BB
33
0
6
11
16
21
31
CR
← ¬(CR
∨ CR
BB
)
BT
BA
The CR bit specified by the BA field is ORed with the CR bit specified by the BB field; the ones complement
of the result is placed into the CR bit specified by the BT field.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• CR
BT
Invalid Instruction Forms
• Reserved fields
Table 9-16. Extended Mnemonics for crnor
Other Registers
Altered
Mnemonic
Operands
Function
CR not.
Extended mnemonic for
bx, by
crnot
crnor bx,by,by
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
cror
Condition Register OR
PPC440x5 CPU Core User’s Manual
Preliminary
cror
Condition Register OR
cror
BT, BA, BB
19
BT
BA
BB
449
0
6
11
16
21
31
CR
← CR
∨ CR
BB
BT
BA
The CR bit specified by the BA field is ORed with the CR bit specified by the BB field; the result is placed into
the CR bit specified by the BT field.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• CR
BT
Invalid Instruction Forms
• Reserved fields
Table 9-17. Extended Mnemonics for cror
Other Registers
Altered
Mnemonic
Operands
Function
CR move.
Extended mnemonic for
bx, by
crmove
cror bx,by,by
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
crorc
Condition Register OR with Complement
PPC440x5 CPU Core User’s Manual
Preliminary
crorc
Condition Register OR with Complement
crorc
BT, BA, BB
19
BT
BA
BB
417
0
6
11
16
21
31
CR
← CR
∨ ¬CR
BB
BT
BA
The condition register (CR) bit specified by the BA field is ORed with the ones complement of the CR bit
specified by the BB field; the result is placed into the CR bit specified by the BT field.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• CR
BT
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
crxor
Condition Register XOR
PPC440x5 CPU Core User’s Manual
Preliminary
crxor
Condition Register XOR
crxor
BT, BA, BB
19
BT
BA
BB
193
0
6
11
16
21
31
CR
← CR
⊕ CR
BB
BT
BA
The CR bit specified by the BA field is XORed with the CR bit specified by the BB field; the result is placed
into the CR bit specified by the BT field.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• CR
BT
Invalid Instruction Forms
• Reserved fields
Table 9-18. Extended Mnemonics for crxor
Other Registers
Altered
Mnemonic
Operands
Function
Condition register clear.
Extended mnemonic for
bx
crclr
crxor bx,bx,bx
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
dcbf
Data Cache Block Flush
PPC440x5 CPU Core User’s Manual
Preliminary
dcbf
Data Cache Block Flush
dcbf
RA, RB
31
RA
RB
86
0
6
11
16
21
31
EA ← (RA|0) + (RB)
DCBF(EA)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
If the data block corresponding to the EA is in the data cache and marked as modified (stored into), the data
block is copied back to main storage and then marked invalid in the data cache. If the data block is not
marked as modified, it is simply marked invalid in the data cache. The operation is performed whether or not
the memory page referenced by the EA is marked as cacheable.
If the data block at the EA is not in the data cache, no operation is performed.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Exceptions
This instruction is considered a “load” with respect to Data Storage exceptions. See Data Storage Interrupt on
This instruction is considered a “store” with respect to data address compare (DAC) Debug exceptions. See
This instruction may cause a Cache Locking type of Data Storage exception. See Data Storage Interrupt on
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
dcbi
Data Cache Block Invalidate
Preliminary
PPC440x5 CPU Core User’s Manual
dcbi
Data Cache Block Invalidate
dcbi
RA, RB
31
RA
RB
470
0
6
11
16
21
31
EA ← (RA|0) + (RB)
DCBI(EA)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
If the data block at the EA is in the data cache, the data block is marked invalid, regardless of whether or not
the memory page referenced by the EA is marked as cacheable. If modified data existed in the data block
prior to the operation of this instruction, that data is lost.
If the data block at the EA is not in the data cache, no operation is performed.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Notes
Execution of this instruction is privileged.
Exceptions
This instruction is considered a “store” with respect to Data Storage exceptions. See Data Storage Interrupt
This instruction is considered a “store” with respect to data address compare (DAC) Debug exceptions. See
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
dcbst
Data Cache Block Store
PPC440x5 CPU Core User’s Manual
Preliminary
dcbst
Data Cache Block Store
dcbst
RA, RB
31
RA
RB
54
0
6
11
16
21
31
EA ← (RA|0) + (RB)
DCBST(EA)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0, and is the contents of register RA otherwise.
If the data block at the EA is in the data cache and marked as modified, the data block is copied back to main
storage and marked as unmodified in the data cache.
If the data block at the EA is in the data cache, and is not marked as modified, or if the data block at the EA is
not in the data cache, no operation is performed.
The operation specified by this instruction is performed whether or not the memory page referenced by the
EA is marked as cacheable.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Exceptions
This instruction is considered a “load” with respect to Data Storage exceptions. See Data Storage Interrupt on
This instruction is considered a “store” with respect to data address compare (DAC) Debug exceptions. See
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
dcbt
Data Cache Block Touch
Preliminary
PPC440x5 CPU Core User’s Manual
dcbt
Data Cache Block Touch
dcbt
RA, RB
31
RA
RB
278
0
6
11
16
21
31
EA ← (RA|0) + (RB)
DCBT(EA)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 when the RA field is 0, and is the contents of register RA otherwise.
If the data block at the EA is not in the data cache and the memory page referenced by the EA is marked as
cacheable, the block is read from main storage into the data cache.
If the data block at the EA is in the data cache, or if the memory page referenced by the EA is marked as
caching inhibited, no operation is performed.
This instruction is not allowed to cause Data Storage interrupts nor Data TLB Error interrupts. If execution of
the instruction causes either of these types of exception, then no operation is performed, and no interrupt
occurs.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Notes
The dcbt instruction allows a program to begin a cache block fetch from main storage before the program
needs the data. The program can later load data from the cache into registers without incurring the latency of
a cache miss.
Exceptions
This instruction is considered a “load” with respect to Data Storage exceptions. See Data Storage Interrupt on
This instruction is considered a “load” with respect to data address compare (DAC) Debug exceptions. See
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
dcbtst
Data Cache Block Touch for Store
PPC440x5 CPU Core User’s Manual
Preliminary
dcbtst
Data Cache Block Touch for Store
dcbtst
RA, RB
31
RA
RB
246
0
6
11
16
21
31
EA ← (RA|0) + (RB)
DCBTST(EA)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
If the data block at the EA is not in the data cache and the memory page referenced by the EA address is
marked as cacheable, the data block is loaded into the data cache.
If the data block at the EA is in the data cache, or if the memory page referenced by the EA is marked as
caching inhibited, no operation is performed.
This instruction is not allowed to cause Data Storage interrupts nor Data TLB Error interrupts. If execution of
the instruction causes either of these types of exception, then no operation is performed, and no interrupt
occurs.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Notes
The dcbtst instruction allows a program to begin a cache block fetch from main storage before the program
needs the data. The program can later store data from GPRs into the cache block, without incurring the
latency of a cache miss.
Architecturally, dcbtst is intended to bring a cache block into the data cache in a manner which will permit
future instructions to store to that block efficiently. For example, in an implementation which supports the
“MESI” cache coherency protocol, the block would be brought into the cache in “Exclusive” mode, allowing
the block to be stored to without having to broadcast any coherency operations on the system bus. However,
since the PPC440x5 core does not support hardware-enforcement of multiprocessor coherency, there is no
distinction between a block being brought in for a read or a write, and hence the implementation of the dcbtst
instruction is identical to the implementation of the dcbt instruction.
Exceptions
This instruction is considered a “load” with respect to Data Storage exceptions. See Data Storage Interrupt on
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
dcbtst
Data Cache Block Touch for Store
PPC440x5 CPU Core User’s Manual
Preliminary
This instruction is considered a “load” with respect to data address compare (DAC) Debug exceptions. See
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
dcbz
Data Cache Block Set to Zero
PPC440x5 CPU Core User’s Manual
Preliminary
dcbz
Data Cache Block Set to Zero
dcbz
RA, RB
31
RA
RB
1014
0
6
11
16
21
31
EA ← (RA|0) + (RB)
DCBZ(EA)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
If the data block at the EA is in the data cache and the memory page referenced by the EA is marked as
cacheable and non-write-through, the data in the cache block is set to 0 and marked as dirty (modified).
If the data block at the EA is not in the data cache and the memory page referenced by the EA is marked as
cacheable and non-write-through, a cache block is established and set to 0 and marked as dirty. Note that
nothing is read from main storage, as described in the programming note.
If the memory page referenced by the EA is marked as either write-through or as caching inhibited, an Align-
ment exception occurs.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Notes
Because dcbz can establish an address in the data cache without copying the contents of that address from
main storage, the address established may be invalid with respect to the storage subsystem. A subsequent
operation may cause the address to be copied back to main storage, for example, to make room for a new
cache block; a Data Machine Check exception could occur under these circumstances.
If dcbz is attempted to an EA in a memory page which is marked as caching inhibited or as write-through, the
software alignment exception handler should emulate the instruction by storing zeros to the block referenced
by the EA. The store instructions in the emulation software will cause main storage to be updated (and
possibly the cache, if the EA is in a page marked as write-through).
Exceptions
An alignment exception occurs if the EA is marked as caching inhibited or as write-through.
This instruction is considered a “store” with respect to Data Storage exceptions. See Data Storage Interrupt
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
dcbz
Data Cache Block Set to Zero
Preliminary
PPC440x5 CPU Core User’s Manual
This instruction is considered a “store” with respect to data address compare (DAC) Debug exceptions. See
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
dccci
Data Cache Congruence Class Invalidate
PPC440x5 CPU Core User’s Manual
Preliminary
dccci
Data Cache Congruence Class Invalidate
dccci
RA, RB
31
RA
RB
454
0
6
11
16
21
31
DCCCI
This instruction flash invalidates the entire data cache array. The RA and RB operands are not used; previous
implementations used these operands to calculate an effective address (EA) which specified the particular
block or blocks to be invalidated. The instruction form (including the specification of RA and RB operands) is
maintained for software and tool compatibility.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Notes
Execution of this instruction is privileged.
This instruction is intended for use in the power-on reset routine to invalidate the entire data cache array
before caching is enabled.
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
dcread
Data Cache Read
Preliminary
PPC440x5 CPU Core User’s Manual
dcread
Data Cache Read
dcread
RT, RA, RB
31
RT
RA
RB
486
0
6
11
16
21
31
EA ← (RA|0) + (RB)
INDEX ← EA
17:26
WORD ← EA
27:29
(RT) ← (data cache data)[INDEX,WORD]
DCDBTRH ← (data cache tag high)[INDEX]
DCDBTRL ← (data cache tag low)[INDEX]
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
EA
selects a line of tag and data from the data cache. EA
selects a word from the 8-word data
17:26
27:29
portion of the selected cache line, and this word is read into register RT. EA
value placed in register RT is undefined.
must be 0b00; if not, the
30:31
The tag portion of the selected cache line is read into the DCDBTRH and DCDBTRL registers, as follows:
Tag
Field
Register[bit(s)]
Name
Bits 0:23 of the lower 32 bits of the 36-bit real
address associated with this cache line
DCDBTRH[0:23]
TRA
Tag Real Address
The valid indicator for the cache line (1 indi-
cates valid)
DCDBTRH[24]
V
Valid
DCDBTRH[25:27]
DCDBTRH[28:31]
reserved
Reserved fields are read as 0s
Upper 4 bits of the 36-bit real address associ-
ated with this cache line
TERA
Tag Extended Real Address
DCDBTRL[0:23]
DCDBTRL[24:27]
reserved
Reserved fields are read as 0s
The “dirty” (modified) indicators for each of the
four doublewords in the cache line
D
Dirty Indicators
The U0 storage attribute for the memory page
associated with this cache line
DCDBTRL[28]
DCDBTRL[29]
DCDBTRL[30]
DCDBTRL[31]
U0
U1
U2
U3
U0 Storage Attribute
U1 Storage Attribute
U2 Storage Attribute
U3 Storage Attribute
The U0 storage attribute for the memory page
associated with this cache line
The U0 storage attribute for the memory page
associated with this cache line
The U0 storage attribute for the memory page
associated with this cache line
This instruction can be used by a debug tool to determine the contents of the data cache, without knowing the
specific addresses of the lines which are currently contained within the cache.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
dcread
Data Cache Read
PPC440x5 CPU Core User’s Manual
Preliminary
Registers Altered
• RT
• DCDBTRH
• DCDBTRL
Invalid Instruction Forms
• Reserved fields
Programming Note
Execution of this instruction is privileged.
The PPC440x5 core does not support the use of the dcread instruction when the data cache controller is still
in the process of performing cache operations associated with previously executed instructions (such as line
fills and line flushes). Also, the PPC440x5 core does not automatically synchronize context between a
dcread instruction and the subsequent mfspr instructions that read the results of the dcread instruction into
GPRs. In order to guarantee that the dcread instruction operates correctly, and that the mfspr instructions
obtain the results of the dcread instruction, a sequence such as the following must be used:
msync
dcread
# ensure that all previous cache operations have completed
regT,regA,regB# read cache information; the contents of GPR A and GPR B are
# added and the result used to specify a cache line index to be read;
# the data word is moved into GPR T and the tag information is read
# into DCDBTRH and DCDBTRL
isync
# ensure dcread completes before attempting to read results
mfdcdbtrh regD
mfdcdbtrl regE
# move high portion of tag into GPR D
# move low portion of tag into GPR E
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
divw
Divide Word
Preliminary
PPC440x5 CPU Core User’s Manual
divw
Divide Word
divw
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
divw.
divwo
divwo.
31
RT
RA
RB
OE
491
Rc
0
6
11
16
21 22
31
(RT) ← (RA) ÷ (RB)
The contents of register RA are divided by the contents of register RB. The quotient is placed into register RT.
Both the dividend and the divisor are interpreted as signed integers. The quotient is the unique signed integer
that satisfies:
dividend = (quotient × divisor) + remainder
where the remainder has the same sign as the dividend and its magnitude is less than that of the divisor.
If an attempt is made to perform (0x8000 0000 ÷ –1) or (n ÷ 0), the contents of register RT are undefined; if
the Rc field also contains 1, the contents of CR[CR0] are undefined. Either invalid division operation sets
0:2
XER[OV, SO] (and CR[CR0] if Rc contains 1) to 1 if the OE field contains 1.
3
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[OV, SO] if OE contains 1
Programming Note
The 32-bit remainder can be calculated using the following sequence of instructions:
divw
mullw
subf
RT,RA,RB
RT,RT,RB
RT,RT,RA
# RT = quotient
# RT = quotient × divisor
# RT = remainder
The sequence does not calculate correct results for the invalid divide operations.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
divwu
Divide Word Unsigned
PPC440x5 CPU Core User’s Manual
Preliminary
divwu
Divide Word Unsigned
divwu
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
divwu.
divwuo
divwuo.
31
RT
RA
RB
OE
459
Rc
0
6
11
16
21 22
31
(RT) ← (RA) ÷ (RB)
The contents of register RA are divided by the contents of register RB. The quotient is placed into register RT.
The dividend and the divisor are interpreted as unsigned integers. The quotient is the unique unsigned
integer that satisfies:
dividend = (quotient × divisor) + remainder
If an attempt is made to perform (n ÷ 0), the contents of register RT are undefined; if the Rc also contains 1,
the contents of CR[CR0] are also undefined. The invalid division operation also sets XER[OV, SO] (and
0:2
CR[CR0] if Rc contains 1) to 1 if the OE field contains 1.
3
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[OV, SO] if OE contains 1
Programming Note
The 32-bit remainder can be calculated using the following sequence of instructions
divwu
mullw
subf
RT,RA,RB
RT,RT,RB
RT,RT,RA
# RT = quotient
# RT = quotient × divisor
# RT = remainder
This sequence does not calculate the correct result if the divisor is 0.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
dlmzb
Determine Leftmost Zero Byte
Preliminary
PPC440x5 CPU Core User’s Manual
dlmzb
ddetelrmmine lezft mbost zero byte
dlmzb.
RA, RS, RB
RA, RS, RB
Rc=0
Rc=1
31
RS
RA
RB
78
Rc
0
6
11
16
21
31
d ← (RS) || (RB)
i, x, y ← 0
do while (x < 8) ∧ (y = 0)
x ← x + 1
if d
= 0 then
i:i + 7
y ← 1
else
i ← i + 8
(RA) ← x
XER[TBC] ← x
if Rc = 1 then
CR[CR0] ←XER[SO]
3
if y = 1 then
if x < 5 then
CR[CR0]
else
CR[CR0]
← 0b010
0:2
0:2
← 0b100
else
CR[CR0]
← 0b001
0:2
The contents of registers RS and RB are concatenated to form an 8-byte operand. The operand is searched
for the leftmost byte in which each bit is 0 (a 0-byte).
Bytes in the operand are numbered from left to right starting with 1. If a 0-byte is found, its byte number is
placed into XER[TBC] and register RA. Otherwise, the number 8 is placed into XER[TBC] and register RA.
If the Rc field contains 1, XER[SO] is copied to CR[CR0] and CR[CR0] are updated as follows:
3
0:2
• If no 0-byte is found, CR[CR0] is set to 0b001.
0:2
• If the leftmost 0-byte is in the first 4 bytes (in the RS register), CR[CR0] is set to 0b010.
0:2
• If the leftmost 0-byte is in the last 4 bytes (in the RB register), CR[CR0] is set to 0b100.
0:2
Registers Altered
• XER[TBC]
• RA
• CR[CR0] if Rc contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
eqv
Equivalent
PPC440x5 CPU Core User’s Manual
Preliminary
eqv
Equivalent
eqv
eqv.
RA, RS, RB
RA, RS, RB
Rc=0
Rc=1
31
RS
RA
RB
284
Rc
0
6
11
16
21
31
(RA) ← ¬((RS) ⊕ (RB))
The contents of register RS are XORed with the contents of register RB; the ones complement of the result is
placed into register RA.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
extsb
Extend Sign Byte
Preliminary
PPC440x5 CPU Core User’s Manual
extsb
Extend Sign Byte
extsb
extsb.
RA, RS
RA, RS
Rc=0
Rc=1
31
RS
RA
954
Rc
0
6
11
16
21
31
(RA) ← EXTS(RS)
24:31
The least significant byte of register RS is sign-extended to 32 bits by replicating bit 24 of the register into bits
0 through 23 of the result. The result is placed into register RA.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
extsh
Extend Sign Halfword
PPC440x5 CPU Core User’s Manual
Preliminary
extsh
Extend Sign Halfword
extsh
extsh.
RA, RS
RA, RS
Rc=0
Rc=1
31
RS
RA
922
Rc
0
6
11
16
21
31
(RA) ← EXTS(RS)
16:31
The least significant halfword of register RS is sign-extended to 32 bits by replicating bit 16 of the register into
bits 0 through 15 of the result. The result is placed into register RA.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
icbi
Instruction Cache Block Invalidate
PPC440x5 CPU Core User’s Manual
Preliminary
icbi
Instruction Cache Block Invalidate
icbi
RA, RB
31
RA
RB
982
0
6
11
16
21
31
EA ← (RA|0) + (RB)
ICBI(EA)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
If the instruction block at the EA is in the instruction cache, the cache block is marked invalid.
If the instruction block at the EA is not in the instruction cache, no additional operation is performed.
The operation specified by this instruction is performed whether or not the EA is marked as cacheable.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Note
Instruction cache management instructions use MSR[DS], not MSR[IS], as part of the virtual address. Also,
the instruction cache on the PPC440x5 is “virtually-tagged”, which means that the EA is converted to a virtual
Exceptions
Instruction Storage interrupts and Instruction TLB Error interrupts are associated with exceptions which occur
during instruction fetching, not during instruction execution. Execution of instruction cache management
instructions may cause Data Storage or Data TLB Error exceptions.
This instruction is considered a “load” with respect to Data Storage exceptions. See Data Storage Interrupt on
This instruction is considered a “load” with respect to data address compare (DAC) Debug exceptions. See
This instruction may cause a Cache Locking type of Data Storage exception. See Data Storage Interrupt on
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
icbt
Instruction Cache Block Touch
PPC440x5 CPU Core User’s Manual
Preliminary
icbt
Instruction Cache Block Touch
icbt
RA, RB
31
RA
RB
22
0
6
11
16
21
31
EA← (RA|0) + (RB)
ICBT(EA)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
If the instruction block at the EA is not in the instruction cache and the memory page referenced by the EA is
marked as cacheable, the instruction block is fetched into the instruction cache.
If the instruction block at the EA is in the instruction cache, or if the memory page referenced by the EA is
marked as caching inhibited, no operation is performed.
If the memory page referenced by the EA is marked as “no-execute” for the current operating mode (user
mode or supervisor mode, as specified by MSR[PR]), no operation is performed.
This instruction is not allowed to cause Data Storage interrupts nor Data TLB Error interrupts. If execution of
the instruction causes either of these types of exception, then no operation is performed, and no interrupt
occurs.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Notes
This instruction allows a program to begin a cache block fetch from main storage before the program needs
the instruction. The program can later branch to the instruction address and fetch the instruction from the
cache without incurring the latency of a cache miss.
Instruction cache management instructions use MSR[DS], not MSR[IS], as part of the virtual address. Also,
the instruction cache on the PPC440x5 is “virtually-tagged”, which means that the EA is converted to a virtual
Exceptions
Instruction Storage interrupts and Instruction TLB Error interrupts are associated with exceptions which occur
during instruction fetching, not during instruction execution. Execution of instruction cache management
instructions may cause Data Storage or Data TLB Error exceptions, but are not allowed to cause the associ-
ated interrupt. Instead, if such an exception occurs, then no operation is performed.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
icbt
Instruction Cache Block Touch
Preliminary
PPC440x5 CPU Core User’s Manual
This instruction is considered a “load” with respect to Data Storage exceptions. See Data Storage Interrupt on
This instruction is considered a “load” with respect to data address compare (DAC) Debug exceptions. See
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
iccci
Instruction Cache Congruence Class Invalidate
PPC440x5 CPU Core User’s Manual
Preliminary
iccci
Instruction Cache Congruence Class Invalidate
iccci
RA, RB
31
RA
RB
966
0
6
11
16
21
31
ICCCI
This instruction flash invalidates the entire instruction cache array. The RA and RB operands are not used;
previous implementations used these operands to calculate an effective address (EA) which specified the
particular block or blocks to be invalidated. The instruction form (including the specification of RA and RB
operands) is maintained for software and tool compatibility.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Notes
Execution of this instruction is privileged.
This instruction is intended for use in the power-on reset routine to invalidate the entire instruction cache
array before caching is enabled.
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
icread
Instruction Cache Read
Preliminary
PPC440x5 CPU Core User’s Manual
icread
Instruction Cache Read
icread
RA, RB
31
RA
RB
998
0
6
11
16
21
31
EA ← (RA|0) + (RB)
INDEX ← EA
17:26
WORD ← EA
27:29
ICDBDR ← (instruction cache data)[INDEX,WORD]
ICDBTRH ← (instruction cache tag high)[INDEX]
ICDBTRL ← (instruction cache tag low)[INDEX]
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
EA
selects a line of tag and data (instructions) from the instruction cache. EA
selects a 32-bit
17:26
27:29
instruction from the 8-instruction data portion of the selected cache line, and this instruction is read into the
ICDBDR. EA are ignored, as are EA
.
0:16
30:31
The tag portion of the selected cache line is read into the ICDBTRH and ICDBTRL registers, as follows:
Tag
Field
Register[bit(s)]
ICDBTRH[0:23]
Name
Bits 0:23 of the 32-bit effective address associ-
ated with this cache line
TEA
Tag Effective Address
The valid indicator for the cache line (1 indi-
cates valid)
ICDBTRH[24]
V
Valid
ICDBTRH[25:31]
reserved
Reserved fields are read as 0s
ICDBTRL[0:21]
ICDBTRL[22]
reserved
Reserved fields are read as 0s
The address space portion of the virtual
address associated with this cache line.
TS
Translation Space
TID Disable field for the memory page associ-
ated with this cache line
ICDBTRL[23]
TD
TID
Translation ID (TID) Disable
Translation ID
TID field portion of the virtual address associ-
ated with this cache line
ICDBTRL[24:31]
The instruction cache on PPC440x5 is “virtually-tagged”, which means that the tag field contains the virtual
information on the function of the TS, TD, and TID fields.
This instruction can be used by a debug tool to determine the contents of the instruction cache, without
knowing the specific addresses of the lines which are currently contained within the cache.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
icread
Instruction Cache Read
PPC440x5 CPU Core User’s Manual
Preliminary
Registers Altered
• ICDBDR
• ICDBTRH
• ICDBTRL
Invalid Instruction Forms
• Reserved fields
Programming Note
Execution of this instruction is privileged.
The PPC440x5 does not automatically synchronize context between an icread instruction and the subse-
quent mfspr instructions which read the results of the icread instruction into GPRs. In order to guarantee that
the mfspr instructions obtain the results of the icread instruction, a sequence such as the following must be
used:
icread
regA,regB
# read cache information (the contents of GPR A and GPR B are
# added and the result used to specify a cache line index to be read)
# ensure icread completes before attempting to read results
# move instruction information into GPR C
isync
mficdbdr regC
mficdbtrh regD
mficdbtrl regE
# move high portion of tag into GPR D
# move low portion of tag into GPR E
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
isel
Integer Select
Preliminary
PPC440x5 CPU Core User’s Manual
isel
Add Immediate
isel
RT, RA, RB, CRb
31
RT
RA
RB
CRb
15
0
6
11
16
21
26
31
if CR[CRb] = 1 then
(RT) ← (RA|0)
else
(RT) ← (RB)
If CR[CRb] = 0, register RT is written with the contents of register RB.
If CR[CRb] = 1 and RA ≠ 0, register RT is written with the contents of register RA.
If CR[CRb] = 1 and RA = 0, register RT is written with 0.
Registers Altered
• RT
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
isync
Instruction Synchronize
PPC440x5 CPU Core User’s Manual
Preliminary
isync
Instruction Synchronize
isync
19
150
0
6
21
31
The isync instruction is a context synchronizing instruction.
isync provides an ordering function for the effects of all instructions executed by the processor. Executing
isync insures that all instructions preceding the isync instruction execute before isync completes, except
that storage accesses caused by those instructions need not have completed. Furthermore, all instructions
preceding the isync are guaranteed to be unaffected by any context changes initiated by instructions after
the isync.
No subsequent instructions are initiated by the processor until isync completes. Finally, execution of isync
causes the processor to discard any prefetched instructions (prefetched from the cache, not instructions that
are in the cache or on their way into the cache), with the effect that subsequent instructions are fetched and
executed in the context established by the instructions preceding isync.
isync causes any caching inhibited instruction fetches from memory to be aborted and any data associated
with them to be discarded. Cacheable instruction fetches from memory are not aborted however, as these
should be handled by the icbi instructions which must precede the isync if software wishes to invalidate any
cached instructions.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Note
The following code example illustrates the necessary steps for self-modifying code. This example assumes
that addr1 is both data and instruction cacheable.
stw
regN, addr1
addr1
# data in regN is to become an instruction at addr1
# forces data from the data cache to memory
# wait until the data actually reaches the memory
# invalidate the instruction if it is in the cache (or in the #
dcbst
msync
icbi
addr1
process of being fetched into the cache)
msync
isync
# wait until the icbi completes
# discard and refetch any instructions (including
# possibly the instruction at addr1) which may have
# already been fetched from the cache and be in the
# pipeline after the isync
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lbz
Load Byte and Zero
Preliminary
PPC440x5 CPU Core User’s Manual
lbz
Load Byte and Zero
lbz
RT, D(RA)
34
RT
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
24
(RT) ←
0 || MS(EA,1)
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 if the RA field is 0 and is the
contents of register RA otherwise.
The byte at the EA is extended to 32 bits by concatenating 24 0-bits to its left. The result is placed into
register RT.
Registers Altered
• RT
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lbzu
Load Byte and Zero with Update
PPC440x5 CPU Core User’s Manual
Preliminary
lbzu
Load Byte and Zero with Update
lbzu
RT, D(RA)
35
RT
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
(RA) ← EA
24
(RT) ←
0 || MS(EA,1)
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 if the RA field is 0 and is the
contents of register RA otherwise. The EA is placed into register RA.
The byte at the EA is extended to 32 bits by concatenating 24 0-bits to its left. The result is placed into
register RT.
Registers Altered
• RA
• RT
Invalid Instruction Forms
• RA = RT
• RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lbzux
Load Byte and Zero with Update Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
lbzux
Load Byte and Zero with Update Indexed
lbzux
RT, RA, RB
31
RT
RA
RB
119
0
6
11
16
21
31
EA ← (RA|0) + (RB)
(RA) ← EA
24
(RT) ←
0 || MS(EA,1)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise. The EA is
placed into register RA.
The byte at the EA is extended to 32 bits by concatenating 24 0-bits to its left. The result is placed into
register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RA
• RT
Invalid Instruction Forms
• Reserved fields
• RA = RT
• RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lbzx
Load Byte and Zero Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
lbzx
Load Byte and Zero Indexed
lbzx
RT,RA, RB
31
RT
RA
RB
87
0
6
11
16
21
31
EA ← (RA|0) + (RB)
24
(RT) ←
0 || MS(EA,1)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
The byte at the EA is extended to 32 bits by concatenating 24 0-bits to its left. The result is placed into
register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RT
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lha
Load Halfword Algebraic
Preliminary
PPC440x5 CPU Core User’s Manual
lha
Load Halfword Algebraic
lha
RT, D(RA)
42
RT
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
(RT) ← EXTS(MS(EA,2))
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 if the RA field is 0 and is the
contents of register RA otherwise.
The halfword at the EA is sign-extended to 32 bits and placed into register RT.
Registers Altered
• RT
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lhau
Load Halfword Algebraic with Update
PPC440x5 CPU Core User’s Manual
Preliminary
lhau
Load Halfword Algebraic with Update
lhau
RT, D(RA)
43
RT
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
(RA) ← EA
(RT) ← EXTS(MS(EA,2))
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 when the RA field is 0 and is
the contents of register RA otherwise. The EA is placed into register RA.
The halfword at the EA is sign-extended to 32 bits and placed into register RT.
Registers Altered
• RA
• RT
Invalid Instruction Forms
• RA = RT
• RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lhaux
Load Halfword Algebraic with Update Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
lhaux
Load Halfword Algebraic with Update Indexed
lhaux
RT, RA, RB
31
RT
RA
RB
375
0
6
11
16
21
31
EA ← (RA|0) + (RB)
(RA) ← EA
(RT) ← EXTS(MS(EA,2))
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise. The EA is
placed into register RA.
The halfword at the EA is sign-extended to 32 bits and placed into register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RA
• RT
Invalid Instruction Forms
• Reserved fields
• RA = RT
• RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lhax
Load Halfword Algebraic Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
lhax
Load Halfword Algebraic Indexed
lhax
RT, RA, RB
31
RT
RA
RB
343
0
6
11
16
21
31
EA ← (RA|0) + (RB)
(RT) ← EXTS(MS(EA,2))
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
The halfword at the EA is sign-extended to 32 bits and placed into register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RT
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lhbrx
Load Halfword Byte-Reverse Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
lhbrx
Load Halfword Byte-Reverse Indexed
lhbrx
RT, RA, RB
31
RT
RA
RB
790
0
6
11
16
21
31
EA ← (RA|0) + (RB)
16
(RT) ←
0 || BYTE_REVERSE(MS(EA,2))
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
The halfword at the EA is byte-reversed from the default byte ordering for the memory page referenced by the
EA. The resulting halfword is extended to 32 bits by concatenating 16 0-bits to its left. The result is placed into
register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RT
Invalid Instruction Forms
• Reserved fields
Programming Note
page using the opposite byte ordering from that specified by the Endian storage attribute.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lhz
Load Halfword and Zero
PPC440x5 CPU Core User’s Manual
Preliminary
lhz
Load Halfword and Zero
lhz
RT, D(RA)
40
RT
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
16
(RT) ←
0 || MS(EA,2)
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 if the RA field is 0 and is the
contents of register RA otherwise.
The halfword at the EA is extended to 32 bits by concatenating 16 0-bits to its left. The result is placed into
register RT.
Registers Altered
• RT
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lhzu
Load Halfword and Zero with Update
PPC440x5 CPU Core User’s Manual
Preliminary
lhzu
Load Halfword and Zero with Update
lhzu
RT, D(RA)
41
RT
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
(RA) ← EA
16
(RT) ←
0 || MS(EA,2)
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 if the RA field is 0 and is the
contents of register RA otherwise. The EA is placed into register RA.
The halfword at the EA is extended to 32 bits by concatenating 16 0-bits to its left. The result is placed into
register RT.
Registers Altered
• RA
• RT
Invalid Instruction Forms
• RA = RT
• RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lhzux
Load Halfword and Zero with Update Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
lhzux
Load Halfword and Zero with Update Indexed
lhzux
RT, RA, RB
31
RT
RA
RB
311
0
6
11
16
21
31
EA ← (RA|0) + (RB)
(RA) ← EA
16
(RT) ←
0 || MS(EA,2)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise. The EA is
placed into register RA.
The halfword at the EA is extended to 32 bits by concatenating 16 0-bits to its left. The result is placed into
register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RA
• RT
Invalid Instruction Forms
• Reserved fields
• RA = RT
• RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lhzx
Load Halfword and Zero Indexed
Preliminary
PPC440x5 CPU Core User’s Manual
lhzx
Load Halfword and Zero Indexed
lhzx
RT, RA, RB
31
RT
RA
RB
279
0
6
11
16
21
31
EA ← (RA|0) + (RB)
16
(RT) ←
0 || MS(EA,2)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
The halfword at the EA is extended to 32 bits by concatenating 16 0-bits to its left. The result is placed into
register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RT
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lmw
Load Multiple Word
PPC440x5 CPU Core User’s Manual
Preliminary
lmw
Load Multiple Word
lmw
RT, D(RA)
46
RT
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
r ← RT
do while r ≤ 31
GPR(r)) ← MS(EA,4)
r ← r + 1
EA ← EA + 4
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field in the instruction to 32 bits. The base address is 0 if the RA field
is 0 and is the contents of register RA otherwise.
A series of consecutive words starting at the EA are loaded into a set of consecutive GPRs, starting with
register RT and continuing to and including GPR(31).
Registers Altered
• RT through GPR(31).
Invalid Instruction Forms
• RA is in the range of registers to be loaded, including the case RA = RT = 0.
Programming Note
This instruction can be restarted, meaning that it could be interrupted after having already updated some of
the target registers, and then re-executed from the beginning (after returning from the interrupt), in which
case the registers which had already been loaded prior to the interrupt will be loaded a second time. Note that
if RA is in the range of registers to be loaded (an invalid form; see above) and is also one of the registers
which is loaded prior to the interrupt, then when the instruction is restarted the re-calculated EA will be incor-
rect, since RA will no longer contain the original base address. Hence the definition of this as an invalid form
which software must avoid.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lswi
Load String Word Immediate
Preliminary
PPC440x5 CPU Core User’s Manual
lswi
Load String Word Immediate
lswi
RT, RA, NB
31
RT
RA
NB
597
0
6
11
16
21
31
EA ← (RA|0)
if NB = 0 then
CNT ← 32
else
CNT ← NB
n ← CNT
RFINAL ← ((RT + CEIL(CNT/4) – 1) % 32)
r ← RT – 1
i ← 0
do while n > 0
if i = 0 then
r ← r + 1
if r = 32 then
r ← 0
GPR(r)) ← 0
GPR(r)
) ← MS(EA,1)
i:i+7
i ← i + 8
if i = 32 then
i ← 0
EA ← EA + 1
n ← n – 1
An effective address (EA) is determined by the RA field. If the RA field contains 0, the EA is 0. Otherwise, the
EA is the contents of register RA.
The NB field specifies the byte count CNT. If the NB field contains 0, the byte count is CNT = 32. Otherwise,
the byte count is CNT = NB.
A series of CNT consecutive bytes in main storage, starting at the EA, are loaded into CEIL(CNT/4) consecu-
tive GPRs, four bytes per GPR, until the byte count is exhausted. Bytes are loaded into GPRs; the byte at the
lowest address is loaded into the most significant byte. Bits to the right of the last byte loaded into the last
GPR are set to 0.
The set of loaded GPRs starts at register RT, continues consecutively through GPR(31), and wraps to
register 0, loading until the byte count is exhausted, which occurs in register R
.
FINAL
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RT and subsequent GPRs as described above.
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lswi
Load String Word Immediate
PPC440x5 CPU Core User’s Manual
Preliminary
• RA is in the range of registers to be loaded
• RA = RT = 0
Programming Note
This instruction can be restarted, meaning that it could be interrupted after having already updated some of
the target registers, and then re-executed from the beginning (after returning from the interrupt), in which
case the registers which had already been loaded prior to the interrupt will be loaded a second time. Note that
if RA is in the range of registers to be loaded (an invalid form; see above) and is also one of the registers
which is loaded prior to the interrupt, then when the instruction is restarted the re-calculated EA will be incor-
rect, since RA will no longer contain the original base address. Hence the definition of this as an invalid form
which software must avoid.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lswx
Load String Word Indexed
Preliminary
PPC440x5 CPU Core User’s Manual
lswx
Load String Word Indexed
lswx
RT, RA, RB
31
RT
RA
RB
533
0
6
11
16
21
31
EA ← (RA|0) + (RB)
CNT ← XER[TBC]
n ← CNT
RFINAL ← ((RT + CEIL(CNT/4) – 1) % 32)
r ← RT – 1
i ← 0
do while n > 0
if i = 0 then
r ← r + 1
if r = 32 then
r ← 0
GPR(r)) ← 0
GPR(r)
) ← MS(EA,1)
i:i+7
i ← i + 8
if i = 32 then
i ← 0
EA ← EA + 1
n ← n – 1
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
A byte count CNT is obtained from XER[TBC].
A series of CNT consecutive bytes in main storage, starting at the EA, are loaded into CEIL(CNT/4) consecu-
tive GPRs, four bytes per GPR, until the byte count is exhausted. Bytes are loaded into GPRs; the byte
having the lowest address is loaded into the most significant byte. Bits to the right of the last byte loaded in
the last GPR used are set to 0.
The set of consecutive GPRs loaded starts at register RT, continues through GPR(31), and wraps to register
0, loading until the byte count is exhausted, which occurs in register R
.
FINAL
If XER[TBC] is 0, the byte count is 0 and the contents of register RT are undefined.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RT and subsequent GPRs as described above.
Invalid Instruction Forms
• Reserved fields
• RA or RB is in the range of registers to be loaded.
• RA = RT = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lswx
Load String Word Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
Programming Note
This instruction can be restarted, meaning that it could be interrupted after having already updated some of
the target registers, and then re-executed from the beginning (after returning from the interrupt), in which
case the registers which had already been loaded prior to the interrupt will be loaded a second time. Note that
if RA or RB is in the range of registers to be loaded (an invalid form; see above) and is also one of the regis-
ters which is loaded prior to the interrupt, then when the instruction is restarted the re-calculated EA will be
incorrect, since the affected register will no longer contain the original base address or index. Hence the defi-
nition of these as invalid forms which software must avoid.
If XER[TBC] = 0, the contents of register RT are undefined and lswx is treated as a no-op. Furthermore, if the
EA is such that a Data Storage, Data TLB Error, or Data Address Compare Debug exception occurs, lswx is
treated as a no-op and no interrupt occurs as a result of the exception.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lwarx
Load Word and Reserve Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
lwarx
Load Word and Reserve Indexed
lwarx
RT, RA, RB
31
RT
RA
RB
20
0
6
11
16
21
31
EA ← (RA|0) + (RB)
RESERVE ← 1
(RT) ← MS(EA,4)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
The word at the EA is placed into register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Execution of the lwarx instruction sets the reservation bit.
Registers Altered
• RT
Invalid Instruction Forms
• Reserved fields
Programming Note
The lwarx and stwcx. instructions are typically paired in a loop, as shown in the following example, to create
the effect of an atomic operation to a memory area used as a semaphore between multiple processes. Only
lwarx can set the reservation bit to 1. stwcx. sets the reservation bit to 0 upon its completion, whether or not
stwcx. actually stored (RS) to memory. CR[CR0] must be examined to determine whether (RS) was sent to
2
memory.
loop: lwarx
“alter”
# read the semaphore from memory; set reservation
# change the semaphore bits in the register as required
# attempt to store the semaphore; reset reservation
# some other process intervened and cleared the reservation prior to the above
# stwcx.; try again
stwcx.
bne loop
The PowerPC Book-E architecture specifies that the EA for the lwarx instruction must be word-aligned (that
is, a multiple of 4 bytes); otherwise, the result is undefined. Although the PPC440x5 will execute lwarx
regardless of the EA alignment, in order for the operation of the pairing of lwarx and stwcx. to produce the
desired result, software must ensure that the EA for both instructions is word-aligned. This requirement is due
to the manner in which misaligned storage accesses may be broken up into separate, aligned accesses by
the PPC440x5.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lwbrx
Load Word Byte-Reverse Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
lwbrx
Load Word Byte-Reverse Indexed
lwbrx
RT, RA, RB
31
RT
RA
RB
534
0
6
11
16
21
31
EA ← (RA|0) + (RB)
(RT) ← BYTE_REVERSE(MS(EA,4))
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
The word at the EA is byte-reversed from the default byte ordering for the memory page referenced by the
EA. The result is placed into register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RT
Invalid Instruction Forms
• Reserved fields
Programming Note
page using the opposite byte ordering from that specified by the Endian storage attribute.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lwz
Load Word and Zero
Preliminary
PPC440x5 CPU Core User’s Manual
lwz
Load Word and Zero
lwz
RT, D(RA)
32
RT
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
(RT) ← MS(EA,4)
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 if the RA field is 0 and is the
contents of register RA otherwise.
The word at the EA is placed into register RT.
Registers Altered
• RT
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lwzu
Load Word and Zero with Update
PPC440x5 CPU Core User’s Manual
Preliminary
lwzu
Load Word and Zero with Update
lwzu
RT, D(RA)
33
RT
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
(RA) ← EA
(RT) ← MS(EA,4)
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 if the RA field is 0 and is the
contents of register RA otherwise. The EA is placed into register RA.
The word at the EA is placed into register RT.
Registers Altered
• RA
• RT
Invalid Instruction Forms
• RA = RT
• RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lwzux
Load Word and Zero with Update Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
lwzux
Load Word and Zero with Update Indexed
lwzux
RT, RA, RB
31
RT
RA
RB
55
0
6
11
16
21
31
EA ← (RA|0) + (RB)
(RA) ← EA
(RT) ← MS(EA,4)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise. The EA is
placed into register RA.
The word at the EA is placed into register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RA
• RT
Invalid Instruction Forms
• Reserved fields
• RA = RT
• RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
lwzx
Load Word and Zero Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
lwzx
Load Word and Zero Indexed
lwzx
RT, RA, RB
31
RT
RA
RB
23
0
6
11
16
21
31
EA ← (RA|0) + (RB)
(RT) ← MS(EA,4)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
The word at the EA is placed into register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RT
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
macchw
Multiply Accumulate Cross Halfword to Word Modulo Signed
PPC440x5 CPU Core User’s Manual
Preliminary
macchw
Multiply Accumulate Cross Halfword to Word Modulo Signed
macchw
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
macchw.
macchwo
macchwo.
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
4
RT
RA
RB
OE
172
Rc
0
6
11
16
21 22
31
prod0:31 ← (RA)16:31 × (RB)
signed
0:15
temp0:32 ← prod
+ (RT)
0:31
(RT) ← temp
1:32
The low-order halfword of RA is multiplied by the high-order halfword of RB. The signed product is summed
with the contents of RT and RT is updated with the low-order 32 bits of the signed sum.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
macchws
Multiply Accumulate Cross Halfword to Word Saturate Signed
PPC440x5 CPU Core User’s Manual
Preliminary
macchws
Multiply Accumulate Cross Halfword to Word Saturate Signed
macchws
macchws.
macchwso
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
macchwso. RT, RA, RB
4
RT
RA
RB
OE
236
Rc
0
6
11
16
21 22
31
prod0:31 ← (RA)16:31 × (RB)
signed
0:15
temp0:32 ← prod
+ (RT)
0:31
if ((prod = RT ) ∧ (RT ≠ temp )) then (RT) ← (RT || 31(¬RT ))
0
0
0
1
0
0
else (RT) ← temp
1:32
The low-order halfword of RA is multiplied by the high-order halfword of RB. The signed product is summed
with the contents of RT.
If the signed sum can be represented in 32 bits, then RT is updated with the low-order 32 bits of the signed
sum.
If the signed sum cannot be represented in 32 bits, then RT is updated with a value which is “saturated” to the
31
31
nearest representable value. That is, if the signed sum is less than –2 , then RT is updated with –2 . Like-
31
31
wise, if the signed sum is greater than 2 – 1, then RT is updated with 2 – 1.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
macchwsu
Multiply Accumulate Cross Halfword to Word Saturate Unsigned
PPC440x5 CPU Core User’s Manual
Preliminary
macchwsu
Multiply Accumulate Cross Halfword to Word Saturate Unsigned
macchwsu
RT, RA, RB
OE=0, Rc=0
macchwsu. RT, RA, RB
macchwsuo RT, RA, RB
macchwsuo. RT, RA, RB
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
4
RT
RA
RB
OE
204
Rc
0
6
11
16
21 22
31
prod0:31 ← (RA)16:31 × (RB)
unsigned
0:15
temp0:32 ← prod
+ (RT)
0:31
32
(RT) ← (temp
∨ temp )
1:32
0
The low-order halfword of RA is multiplied by the high-order halfword of RB. The unsigned product is
summed with the contents of RT.
If the unsigned sum can be represented in 32 bits, then RT is updated with the low-order 32 bits of the
unsigned sum.
If the unsigned sum cannot be represented in 32 bits, then RT is updated with a value which is “saturated” to
32
the maximum representable value of 2 – 1.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
macchwu
Multiply Accumulate Cross Halfword to Word Modulo Unsigned
PPC440x5 CPU Core User’s Manual
Preliminary
macchwu
Multiply Accumulate Cross Halfword to Word Modulo Unsigned
macchwu
macchwu.
macchwuo
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
macchwuo. RT, RA, RB
4
RT
RA
RB
OE
140
Rc
0
6
11
16
21 22
31
prod0:31 ← (RA)16:31 × (RB)
unsigned
0:15
temp0:32 ← prod
+ (RT)
0:31
(RT) ← temp
1:32
The low-order halfword of RA is multiplied by the high-order halfword of RB. The unsigned product is
summed with the contents of RT and RT is updated with the low-order 32 bits of the unsigned sum.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
machhw
Multiply Accumulate High Halfword to Word Modulo Signed
PPC440x5 CPU Core User’s Manual
Preliminary
machhw
Multiply Accumulate High Halfword to Word Modulo Signed
machhw
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
machhw.
machhwo
machhwo.
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
4
RT
RA
RB
OE
44
Rc
0
6
11
16
21 22
31
prod0:31 ← (RA)0:15 × (RB)
signed
0:15
temp0:32 ← prod
+ (RT)
0:31
(RT) ← temp
1:32
The high-order halfword of RA is multiplied by the high-order halfword of RB. The signed product is summed
with the contents of RT and RT is updated with the low-order 32 bits of the signed sum.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
machhws
Multiply Accumulate High Halfword to Word Saturate Signed
PPC440x5 CPU Core User’s Manual
Preliminary
machhws
Multiply Accumulate High Halfword to Word Saturate Signed
machhws
machhws.
machhwso
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
machhwso. RT, RA, RB
4
RT
RA
RB
OE
108
Rc
0
6
11
16
21 22
31
prod0:31 ← (RA)0:15 × (RB)
signed
0:15
temp0:32 ← prod
+ (RT)
0:31
if ((prod = RT ) ∧ (RT ≠ temp )) then (RT) ← (RT || 31(¬RT ))
0
0
0
1
0
0
else (RT) ← temp
1:32
The high-order halfword of RA is multiplied by the high-order halfword of RB. The signed product is summed
with the contents of RT.
If the signed sum can be represented in 32 bits, then RT is updated with the low-order 32 bits of the signed
sum.
If the signed sum cannot be represented in 32 bits, then RT is updated with a value which is “saturated” to the
31
31
nearest representable value. That is, if the signed sum is less than –2 , then RT is updated with –2 . Like-
31
31
wise, if the signed sum is greater than 2 – 1, then RT is updated with 2 – 1.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
machhwsu
Multiply Accumulate High Halfword to Word Saturate Unsigned
PPC440x5 CPU Core User’s Manual
Preliminary
machhwsu
Multiply Accumulate High Halfword to Word Saturate Unsigned
machhwsu
RT, RA, RB
OE=0, Rc=0
machhwsu. RT, RA, RB
machhwsuo RT, RA, RB
machhwsuo. RT, RA, RB
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
4
RT
RA
RB
OE
76
Rc
0
6
11
16
21 22
31
prod0:31 ← (RA)0:15 × (RB)
unsigned
0:15
temp0:32 ← prod
+ (RT)
0:31
32
(RT) ← (temp
∨ temp )
1:32
0
The high-order halfword of RA is multiplied by the high-order halfword of RB. The unsigned product is
summed with the contents of RT.
If the unsigned sum can be represented in 32 bits, then RT is updated with the low-order 32 bits of the
unsigned sum.
If the unsigned sum cannot be represented in 32 bits, then RT is updated with a value which is “saturated” to
32
the maximum representable value of 2 – 1.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
machhwu
Multiply Accumulate High Halfword to Word Modulo Unsigned
PPC440x5 CPU Core User’s Manual
Preliminary
machhwu
Multiply Accumulate High Halfword to Word Modulo Unsigned
machhwu
machhwu.
machhwuo
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
machhwuo. RT, RA, RB
4
RT
RA
RB
OE
12
Rc
0
6
11
16
21 22
31
prod0:31 ← (RA)0:15 × (RB)
unsigned
0:15
temp0:32 ← prod
+ (RT)
0:31
(RT) ← temp
1:32
The high-order halfword of RA is multiplied by the high-order halfword of RB. The unsigned product is
summed with the contents of RT and RT is updated with the low-order 32 bits of the unsigned sum.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
maclhw
Multiply Accumulate Low Halfword to Word Modulo Signed
PPC440x5 CPU Core User’s Manual
Preliminary
maclhw
Multiply Accumulate Low Halfword to Word Modulo Signed
maclhw
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
maclhw.
maclhwo
maclhwo.
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
4
RT
RA
RB
OE
428
Rc
0
6
11
16
21 22
31
prod0:31 ← (RA)16:31 × (RB)
signed
16:31
temp0:32 ← prod
+ (RT)
0:31
(RT) ← temp
1:32
The low-order halfword of RA is multiplied by the low-order halfword of RB. The signed product is summed
with the contents of RT and RT is updated with the low-order 32 bits of the signed sum.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
maclhws
Multiply Accumulate Low Halfword to Word Saturate Signed
PPC440x5 CPU Core User’s Manual
Preliminary
maclhws
Multiply Accumulate Low Halfword to Word Saturate Signed
maclhws
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
maclhws.
maclhwso
maclhwso.
4
RT
RA
RB
OE
492
Rc
0
6
11
16
21 22
31
prod0:31 ← (RA)16:31 × (RB)
signed
16:31
temp0:32 ← prod
+ (RT)
0:31
if ((prod = RT ) ∧ (RT ≠ temp )) then (RT) ← (RT || 31(¬RT ))
0
0
0
1
0
0
else (RT) ← temp
1:32
The low-order halfword of RA is multiplied by the low-order halfword of RB. The signed product is summed
with the contents of RT.
If the signed sum can be represented in 32 bits, then RT is updated with the low-order 32 bits of the signed
sum.
If the signed sum cannot be represented in 32 bits, then RT is updated with a value which is “saturated” to the
31
31
nearest representable value. That is, if the signed sum is less than –2 , then RT is updated with –2 . Like-
31
31
wise, if the signed sum is greater than 2 – 1, then RT is updated with 2 – 1.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
maclhwsu
Multiply Accumulate Low Halfword to Word Saturate Unsigned
PPC440x5 CPU Core User’s Manual
Preliminary
maclhwsu
Multiply Accumulate Low Halfword to Word Saturate Unsigned
maclhwsu
maclhwsu.
maclhwsuo RT, RA, RB
maclhwsuo. RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
4
RT
RA
RB
OE
460
Rc
0
6
11
16
21 22
31
prod0:31 ← (RA)16:31 × (RB)
unsigned
16:31
temp0:32 ← prod
+ (RT)
0:31
32
(RT) ← (temp
∨ temp )
1:32
0
The low-order halfword of RA is multiplied by the low-order halfword of RB. The unsigned product is summed
with the contents of RT.
If the unsigned sum can be represented in 32 bits, then RT is updated with the low-order 32 bits of the
unsigned sum.
If the unsigned sum cannot be represented in 32 bits, then RT is updated with a value which is “saturated” to
32
the maximum representable value of 2 – 1.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
maclhwu
Multiply Accumulate Low Halfword to Word Modulo Unsigned
PPC440x5 CPU Core User’s Manual
Preliminary
maclhwu
Multiply Accumulate Low Halfword to Word Modulo Unsigned
maclhwu
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
maclhwu.
maclhwuo
maclhwuo.
4
RT
RA
RB
OE
396
Rc
0
6
11
16
21 22
31
prod0:31 ← (RA)16:31 × (RB)
unsigned
16:31
temp0:32 ← prod
+ (RT)
0:31
(RT) ← temp
1:32
The low-order halfword of RA is multiplied by the low-order halfword of RB. The unsigned product is summed
with the contents of RT and RT is updated with the low-order 32 bits of the unsigned sum.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mbar
Memory Barrier
Preliminary
PPC440x5 CPU Core User’s Manual
mbar
Memory Barrier
mbar
31
854
0
6
21
31
The mbar instruction ensures that all loads and stores preceding mbar complete with respect to main
storage before any loads and stores following mbar access main storage.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Note
Architecturally, mbar merely orders storage accesses, and does not perform execution nor context synchro-
nization (see Synchronization on page 82). Therefore, non-storage access instructions after mbar could
complete before the storage access instructions which were executed prior to mbar have actually completed
their storage accesses. The msync instruction, on the other hand, is execution synchronizing, and does
guarantee that all storage accesses initiated by instructions executed prior to the msync have completed
before any instructions after the msync begin execution. However, the PPC440x5 core implements the mbar
instruction identically to the msync instruction, and thus both are execution synchronizing.
Software should nevertheless use the correct instruction (mbar or msync) as called for by the specific
ordering and synchronizing requirements of the application, in order to guarantee portability to other imple-
mentations.
See Storage Ordering and Synchronization on page 84 for additional information on the use of the msync
and mbar instructions.
Architecture Note
mbar replaces the PowerPC eieio instruction. mbar uses the same opcode as eieio; PowerPC applications
which used eieio will get the function of mbar when executed on a PowerPC Book-E implementation. mbar
is architecturally “stronger” than eieio, in that eieio forced separate ordering amongst different categories of
storage accesses, while mbar forces such ordering amongst all storage accesses as a single category.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mcrf
Move Condition Register Field
PPC440x5 CPU Core User’s Manual
Preliminary
mcrf
Move Condition Register Field
mcrf
BF, BFA
19
BF
BFA
0
0
6
9
11
14
21
31
m ← BFA
n ← BF
(CR[CRn]) ← (CR[CRm])
The contents of the CR field specified by the BFA field are placed into the CR field specified by the BF field.
Registers Altered
• CR[CRn] where n is specified by the BF field.
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mcrxr
Move to Condition Register from XER
PPC440x5 CPU Core User’s Manual
Preliminary
mcrxr
Move to Condition Register from XER
mcrxr
BF
31
BF
512
0
6
9
21
31
n ← BF
(CR[CRn]) ← XER
0:3
4
XER
←
0
0:3
The contents of XER are placed into the CR field specified by the BF field. XER are then set to 0.
0:3
0:3
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• CR[CRn] where n is specified by the BF field.
• XER[SO, OV, CA]
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mfcr
Move From Condition Register
PPC440x5 CPU Core User’s Manual
Preliminary
mfcr
Move From Condition Register
mfcr
RT
31
RT
19
0
6
11
21
31
(RT) ← (CR)
The contents of the CR are placed into register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RT
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mfdcr
Move from Device Control Register
PPC440x5 CPU Core User’s Manual
Preliminary
mfdcr
Move from Device Control Register
mfdcr
RT, DCRN
31
RT
DCRF
323
0
6
11
21
31
DCRN ← DCRF
|| DCRF
0:4
5:9
(RT) ← (DCR(DCRN))
The contents of the DCR specified by the DCRF field are placed into register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RT
Invalid Instruction Forms
• Reserved fields
Programming Notes
Execution of this instruction is privileged.
The DCR number (DCRN) specified in the assembler language coding of the mfdcr instruction refers to a
DCR number. The assembler handles the unusual register number encoding to generate the DCRF field.
Architecture Note
The specific numbers and definitions of any DCRs are outside the scope of both the PowerPC Book-E archi-
tecture and the PPC440x5 core. Any DCRs are defined as part of the chip-level product incorporating the
PPC440x5 core.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mfmsr
Move From Machine State Register
PPC440x5 CPU Core User’s Manual
Preliminary
mfmsr
Move From Machine State Register
mfmsr
RT
31
RT
83
0
6
11
21
31
(RT) ← (MSR)
The contents of the MSR are placed into register RT.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RT
Invalid Instruction Forms
• Reserved fields
Programming Note
Execution of this instruction is privileged.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mfspr
Move From Special Purpose Register
PPC440x5 CPU Core User’s Manual
Preliminary
mfspr
Move From Special Purpose Register
mfspr
RT, SPRN
31
RT
SPRF
339
0
6
11
21
31
SPRN ← SPRF
|| SPRF
0:4
5:9
(RT) ← (SPR(SPRN))
The contents of the SPR specified by the SPRF field are placed into register RT. See Special Purpose Regis-
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RT
Invalid Instruction Forms
• Reserved fields
• Invalid SPRF values
Programming Note
Execution of this instruction is privileged if instruction bit 11 contains 1. See Privileged SPRs on page 81 for a
list of privileged SPRs.
The SPR number (SPRN) specified in the assembler language coding of the mfspr instruction refers to an
SPR number. The assembler handles the unusual register number encoding to generate the SPRF field.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mfspr
Move From Special Purpose Register
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-19. Extended Mnemonics for mfspr
Mnemonic
Operands
Function
mfccr0
mfccr1
mfcsrr0
mfcsrr1
mfctr
mfdac1
mfdac2
mfdbcr0
mfdbcr1
mfdbcr2
mfdbdr
mfdbsr
mfdcdbtrh
mfdcdbtrl
mfdear
mfdec
mfdnv0
mfdnv1
mfdnv2
mfdnv3
mfdtv0
mfdtv1
mfdtv2
mfdtv3
mfdvc1
mfdvc2
mfdvlim
mfesr
Move from special purpose register SPRN.
Extended mnemonic for
mfspr RT,SPRN
RT
mfiac1
mfiac2
mfiac3
mfiac4
mficdbdr
mficdbtrh
mficdbtrl
mfinv0
mfinv1
mfinv2
mfinv3
mfitv0
mfitv1
mfitv2
mfitv3
mfivlim
mfivor0
mfivor1
mfivor2
mfivor3
mfivor4
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mfspr
Move From Special Purpose Register
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-19. Extended Mnemonics for mfspr (continued)
Mnemonic
Operands
Function
mfivor5
mfivor6
mfivor7
mfivor8
mfivor9
mfivor10
mfivor11
mfivor12
mfivor13
mfivor14
mfivor15
mfivpr
mflr
mfmcsr
mfmcsrr0
mfmcsrr1
mfmmucr
mfpid
mfpir
mfpvr
mfsprg0
mfsprg1
mfsprg2
mfsprg3
mfsprg4
mfsprg5
mfsprg6
mfsprg7
mfsrr0
mfsrr1
mftbl
mftbu
mftcr
mftsr
mfusprg0
mfxer
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
msync
Memory Synchronize
PPC440x5 CPU Core User’s Manual
Preliminary
msync
Memory Synchronize
msync
31
598
0
6
21
31
The msync instruction guarantees that all instructions initiated by the processor preceding msync will
complete before msync completes, and that no subsequent instructions will be initiated by the processor
until after msync completes. msync also will not complete until all storage accesses associated with instruc-
tions preceding msync have completed.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None.
Invalid Instruction Forms
• Reserved fields
Programming Notes
The msync instruction is execution synchronizing (see Execution Synchronization on page 83), and guaran-
tees that all storage accesses initiated by instructions executed prior to the msync have completed before
any instructions after the msync begin execution. On the other hand, architecturally the mbar instruction
merely orders storage accesses, and does not perform execution synchronization. Therefore, non-storage
access instructions after mbar could complete before the storage access instructions which were executed
prior to mbar have actually completed their storage accesses. However, the PPC440x5 core implements the
mbar instruction identically to the msync instruction, and thus both are execution synchronizing.
Software should nevertheless use the correct instruction (mbar or msync) as called for by the specific
ordering and synchronizing requirements of the application, in order to guarantee portability to other imple-
mentations.
See Storage Ordering and Synchronization on page 84 for additional information on the use of the msync
and mbar instructions.
Architecture Note
mbar replaces the PowerPC eieio instruction. mbar uses the same opcode as eieio; PowerPC applications
which used eieio will get the function of mbar when executed on a PowerPC Book-E implementation. mbar
is architecturally “stronger” than eieio, in that eieio forced separate ordering amongst different categories of
storage accesses, while mbar forces such ordering amongst all storage accesses as a single category.
msync replaces the PowerPC sync instruction. msync uses the same opcode as sync; PowerPC applica-
tions which used sync get the function of msync when executed on a PowerPC Book-E implementation.
msync is architecturally identical to the version of sync specified by an earlier version of the PowerPC archi-
tecture.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mtcrf
Move to Condition Register Fields
PPC440x5 CPU Core User’s Manual
Preliminary
mtcrf
Move to Condition Register Fields
mtcrf
FXM, RS
31
RS
FXM
144
0
6
4
11 12
4
20 21
31
4
4
mask ← (FXM ) || (FXM ) || ... || (FXM ) || (FXM )
0
1
6
7
(CR) ← ((RS) ∧ mask) ∨ ((CR) ∧ ¬mask)
Some or all of the contents of register RS are placed into the CR as specified by the FXM field.
Each bit in the FXM field controls the copying of 4 bits in register RS into the corresponding bits in the CR.
The correspondence between the bits in the FXM field and the bit copying operation is shown in the following
table:
Table 9-20. FXM Bit Field Correspondence
FXM Bit Number
CR Bits Affected
0:3
0
1
2
3
4
5
6
7
4:7
8:11
12:15
16:19
20:23
24:27
28:31
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• CR
Invalid Instruction Forms
• Reserved fields
Table 9-21. Extended Mnemonics for mtcrf
Mnemonic
Operands
Function
Move to CR.
Extended mnemonic for
RS
mtcr
mtcrf 0xFF,RS
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mtdcr
Move To Device Control Register
PPC440x5 CPU Core User’s Manual
Preliminary
mtdcr
Move To Device Control Register
mtdcr
DCRN, RS
31
RS
DCRF
451
0
6
11
21
31
DCRN ← DCRF
|| DCRF
(DCR(DCRN)) ← (RS)
5:9
0:4
The contents of register RS are placed into the DCR specified by the DCRF field.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• DCR(DCRN)
Invalid Instruction Forms
• Reserved fields
Programming Note
Execution of this instruction is privileged.
The DCR number (DCRN) specified in the assembler language coding of the mtdcr instruction refers to a
DCR number. The assembler handles the unusual register number encoding to generate the DCRF field.
Architecture Note
The specific numbers and definitions of any DCRs are outside the scope of both the PowerPC Book-E archi-
tecture and the PPC440x5 core. Any DCRs are defined as part of the chip-level product incorporating the
PPC440x5 core.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mtmsr
Move To Machine State Register
Preliminary
PPC440x5 CPU Core User’s Manual
mtmsr
Move To Machine State Register
mtmsr
RS
31
RS
146
0
6
11
21
31
(MSR) ← (RS)
The contents of register RS are placed into the MSR.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• MSR
Invalid Instruction Forms
• Reserved fields
Programming Note
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mtspr
Move To Special Purpose Register
PPC440x5 CPU Core User’s Manual
Preliminary
mtspr
Move To Special Purpose Register
mtspr
SPRN, RS
31
RS
SPRF
467
0
6
11
21
31
SPRN ← SPRF
|| SPRF
5:9
0:4
(SPR(SPRN)) ← (RS)
The contents of register RS are placed into register RT. See Special Purpose Registers Sorted by SPR
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• SPR(SPRN)
Invalid Instruction Forms
• Reserved fields
• Invalid SPRF values
Programming Note
Execution of this instruction is privileged if instruction bit 11 contains 1. See Privileged SPRs on page 81 for a
list of privileged SPRs.
The SPR number (SPRN) specified in the assembler language coding of the mtspr instruction refers to an
SPR number. The assembler handles the unusual register number encoding to generate the SPRF field.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mtspr
Move To Special Purpose Register
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-22. Extended Mnemonics for mtspr
Mnemonic
Operands
Function
mtccr0
mtccr1
mtcsrr0
mtcsrr1
mtctr
mtdac1
mtdac2
mtdbcr0
mtdbcr1
mtdbcr2
mtdbdr
mtdbsr
mtdear
mtdec
mtdecar
mtdnv0
mtdnv1
mtdnv2
mtdnv3
mtdtv0
mtdtv1
mtdtv2
mtdtv3
mtdvc1
mtdvc2
mtdvlim
mtesr
Move to special purpose register SPRN.
Extended mnemonic for
mtspr RT,SPRN
RT
mtiac1
mtiac2
mtiac3
mtiac4
mtinv0
mtinv1
mtinv2
mtinv3
mtitv0
mtitv1
mtitv2
mtitv3
mtivlim
mtivor0
mtivor1
mtivor2
mtivor3
mtivor4
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mtspr
Move To Special Purpose Register
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-22. Extended Mnemonics for mtspr (continued)
Mnemonic
Operands
Function
mtivor5
mtivor6
mtivor7
mtivor8
mtivor9
mtivor10
mtivor11
mtivor12
mtivor13
mtivor14
mtivor15
mtivpr
mtlr
mtmcsr
mtmcsrr0
mtmcsrr1
mtmmucr
mtpid
mtsprg0
mtsprg1
mtsprg2
mtsprg3
mtsprg4
mtsprg5
mtsprg6
mtsprg7
mtsrr0
mtsrr1
mttbl
mttbu
mttcr
mttsr
mtusprg0
mtxer
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mulchw
Multiply Cross Halfword to Word Signed
PPC440x5 CPU Core User’s Manual
Preliminary
mulchw
Multiply Cross Halfword to Word Signed
mulchw
mulchw.
RT, RA, RB
RT, RA, RB
Rc=0
Rc=1
4
RT
RA
RB
168
Rc
0
6
11
16
21
31
(RT) :31 ← (RA)16:31 × (RB)
signed
0:15
0
The low-order halfword of RA is multiplied by the high-order halfword of RB, considering both source oper-
ands as signed integers. The 32-bit result is placed into register RT.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mulchwu
Multiply Cross Halfword to Word Unsigned
PPC440x5 CPU Core User’s Manual
Preliminary
mulchwu
Multiply Cross Halfword to Word Unsigned
mulchwu
mulchwu.
RT, RA, RB
RT, RA, RB
Rc=0
Rc=1
4
RT
RA
RB
136
Rc
0
6
11
16
21
31
(RT)
← (RA)16:31 × (RB)
unsigned
:
0:15
0 31
The low-order halfword of RA is multiplied by the high-order halfword of RB, considering both source oper-
ands as unsigned integers. The 32-bit result is placed into register RT.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mulhhw
Multiply High Halfword to Word Signed
PPC440x5 CPU Core User’s Manual
Preliminary
mulhhw
Multiply High Halfword to Word Signed
mulhhw
mulhhw.
RT, RA, RB
RT, RA, RB
Rc=0
Rc=1
4
RT
RA
RB
40
Rc
0
6
11
16
21
31
(RT)
← (RA)0:15 × (RB)
signed
0:15
:
0 31
The high-order halfword of RA is multiplied by the high-order halfword of RB, considering both source oper-
ands as signed integers. The 32-bit result is placed into register RT.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mulhhwu
Multiply High Halfword to Word Unsigned
PPC440x5 CPU Core User’s Manual
Preliminary
mulhhwu
Multiply High Halfword to Word Unsigned
mulhhwu
mulhhwu.
RT, RA, RB
RT, RA, RB
Rc=0
Rc=1
4
RT
RA
RB
8
Rc
0
6
11
16
21
31
(RT)
← (RA)0:15 × (RB)
unsigned
:
0:15
0 31
The high-order halfword of RA is multiplied by the high-order halfword of RB, considering both source oper-
ands as unsigned integers. The 32-bit result is placed into register RT.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mulhw
Multiply High Word
Preliminary
PPC440x5 CPU Core User’s Manual
mulhw
Multiply High Word
mulhw
mulhw.
RT, RA, RB
RT, RA, RB
Rc=0
Rc=1
31
RT
RA
RB
75
Rc
0
6
11
16
21 22
31
prod
(RT) ← prod
← (RA) × (RB) signed
0:63
0:31
The 64-bit signed product of registers RA and RB is formed. The most significant 32 bits of the result is
placed into register RT.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
Programming Note
The most significant 32 bits of the product, unlike the least significant 32 bits, may differ depending on
whether the registers RA and RB are interpreted as signed or unsigned quantities. mulhw generates the
correct result when these operands are interpreted as signed quantities. mulhwu generates the correct
result when these operands are interpreted as unsigned quantities.
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mulhwu
Multiply High Word Unsigned
PPC440x5 CPU Core User’s Manual
Preliminary
mulhwu
Multiply High Word Unsigned
mulhwu
mulhwu.
RT, RA, RB
RT, RA, RB
Rc=0
Rc=1
31
RT
RA
RB
11
Rc
0
6
11
16
21 22
31
prod
(RT) ← prod
← (RA) × (RB) unsigned
0:63
0:31
The 64-bit unsigned product of registers RA and RB is formed. The most significant 32 bits of the result are
placed into register RT.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
Programming Note
The most significant 32 bits of the product, unlike the least significant 32 bits, may differ depending on
whether the registers RA and RB are interpreted as signed or unsigned quantities. The mulhw instruction
generates the correct result when these operands are interpreted as signed quantities. The mulhwu instruc-
tion generates the correct result when these operands are interpreted as unsigned quantities.
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mullhw
Multiply Low Halfword to Word Signed
PPC440x5 CPU Core User’s Manual
Preliminary
mullhw
Multiply High Halfword to Word Signed
mullhw
mullhw.
RT, RA, RB
RT, RA, RB
Rc=0
Rc=1
4
RT
RA
RB
424
Rc
0
6
11
16
21
31
(RT)
← (RA)16:31 × (RB)
signed
:
16:31
0 31
The low-order halfword of RA is multiplied by the low-order halfword of RB, considering both source operands
as signed integers. The 32-bit result is placed into register RT.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mullhwu
Multiply Low Halfword to Word Unsigned
PPC440x5 CPU Core User’s Manual
Preliminary
mullhwu
Multiply High Halfword to Word Unsigned
mullhwu
mullhwu.
RT, RA, RB
RT, RA, RB
Rc=0
Rc=1
4
RT
RA
RB
392
Rc
0
6
11
16
21
31
(RT)
← (RA)16:31 × (RB)
unsigned
:
16:31
0 31
The low-order halfword of RA is multiplied by the low-order halfword of RB, considering both source operands
as unsigned integers. The 32-bit result is placed into register RT.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mulli
Multiply Low Immediate
Preliminary
PPC440x5 CPU Core User’s Manual
mulli
Multiply Low Immediate
mulli
RT, RA, IM
7
RT
RA
IM
0
6
11
16
31
prod
(RT) ← prod
← (RA) × IM
0:47
16:47
The 48-bit product of register RA and the 16-bit IM field is formed. The least significant 32 bits of the product
are placed into register RT.
Registers Altered
• RT
Programming Note
The least significant 32 bits of the product are correct, regardless of whether register RA and field IM are
interpreted as signed or unsigned numbers.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
mullw
Multiply Low Word
PPC440x5 CPU Core User’s Manual
Preliminary
mullw
Multiply Low Word
mullw
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
mullw.
mullwo
mullwo.
31
RT
RA
RB
OE
235
Rc
0
6
11
16
21 22
31
prod
(RT) ← prod
← (RA) × (RB) signed
0:63
32:63
The 64-bit signed product of register RA and register RB is formed. The least significant 32 bits of the result is
placed into register RT.
If the signed product cannot be represented in 32 bits and OE=1, XER[SO, OV] are set to 1.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE=1
Programming Note
The least significant 32 bits of the product are correct, regardless of whether register RA and register RB are
interpreted as signed or unsigned numbers. The overflow indication, however, is calculated specifically for a
64-bit signed product, and is dependent upon interpretation of the source operands as signed numbers.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
nand
NAND
Preliminary
PPC440x5 CPU Core User’s Manual
nand
NAND
nand
nand.
RA, RS, RB
RA, RS, RB
Rc=0
Rc=1
31
RT
RA
RB
476
Rc
0
6
11
16
21
31
(RA) ← ¬((RS) ∧ (RB))
The contents of register RS is ANDed with the contents of register RB; the ones complement of the result is
placed into register RA.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
neg
Negate
PPC440x5 CPU Core User’s Manual
Preliminary
neg
Negate
neg
RT, RA
RT, RA
RT, RA
RT, RA
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
neg.
nego
nego.
31
RT
RA
OE
104
Rc
0
6
11
16
21 22
31
(RT) ← ¬(RA) + 1
The twos complement of the contents of register RA are placed into register RT.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE=1
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
nmacchw
Negative Multiply Accumulate Cross Halfword to Word Modulo Signed
PPC440x5 CPU Core User’s Manual
Preliminary
nmacchw
Negative Multiply Accumulate Cross Halfword to Word Modulo Signed
nmacchw
nmacchw.
nmacchwo
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
nmacchwo. RT, RA, RB
4
RT
RA
RB
OE
174
Rc
0
6
11
16
21 22
31
nprod0:31 ← –((RA)16:31 × (RB)
) signed
0:15
temp0:32 ← nprod
+ (RT)
0:31
(RT) ← temp
1:32
The low-order halfword of RA is multiplied by the high-order halfword of RB. The signed product is subtracted
from the contents of RT and RT is updated with the low-order 32 bits of the result.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
nmacchws
Negative Multiply Accumulate Cross Halfword to Word Saturate Signed
PPC440x5 CPU Core User’s Manual
Preliminary
nmacchws
Negative Multiply Accumulate High Halfword to Word Saturate Signed
nmacchws
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
nmacchws. RT, RA, RB
nmacchwso RT, RA, RB
nmacchwso. RT, RA, RB
4
RT
RA
RB
OE
238
Rc
0
6
11
16
21 22
31
nprod0:31 ← –((RA)16:31 × (RB)
signed
0:15
temp0:32 ← nprod
+ (RT)
0:31
if ((nprod = RT ) ∧ (RT ≠ temp )) then (RT) ← (RT || 31(¬RT ))
0
0
0
1
0
0
else (RT) ← temp
1:32
The low-order halfword of RA is multiplied by the high-order halfword of RB. The signed product is subtracted
from the contents of RT.
If the result of the subtraction can be represented in 32 bits, then RT is updated with the low-order 32 bits of
the result.
If the result of the subtraction cannot be represented in 32 bits, then RT is updated with a value which is
31
“saturated” to the nearest representable value. That is, if the result is less than –2 , then RT is updated with
31
31
31
–2 . Likewise, if the result is greater than 2 – 1, then RT is updated with 2 – 1.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
nmachhw
Negative Multiply Accumulate High Halfword to Word Modulo Signed
PPC440x5 CPU Core User’s Manual
Preliminary
nmachhw
Negative Multiply Accumulate High Halfword to Word Modulo Signed
nmachhw
nmachhw.
nmachhwo
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
nmachhwo. RT, RA, RB
4
RT
RA
RB
OE
46
Rc
0
6
11
16
21 22
31
nprod0:31 ← –((RA)0:15 × (RB)
) signed
0:15
temp0:32 ← nprod
+ (RT)
0:31
(RT) ← temp
1:32
The high-order halfword of RA is multiplied by the high-order halfword of RB. The signed product is
subtracted from the contents of RT and RT is updated with the low-order 32 bits of the result.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
nmachhws
Negative Multiply Accumulate High Halfword to Word Saturate Signed
PPC440x5 CPU Core User’s Manual
Preliminary
nmachhws
Negative Multiply Accumulate High Halfword to Word Saturate Signed
nmachhws
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
nmachhws. RT, RA, RB
nmachhwso RT, RA, RB
nmachhwso. RT, RA, RB
4
RT
RA
RB
OE
110
Rc
0
6
11
16
21 22
31
nprod0:31 ← –((RA)0:15 × (RB)
) signed
0:15
temp0:32 ← nprod
+ (RT)
0:31
if ((nprod = RT ) ∧ (RT ≠ temp )) then (RT) ← (RT || 31(¬RT ))
0
0
0
1
0
0
else (RT) ← temp
1:32
The high-order halfword of RA is multiplied by the high-order halfword of RB. The signed product is
subtracted from the contents of RT.
If the result of the subtraction can be represented in 32 bits, then RT is updated with the low-order 32 bits of
the result.
If the result of the subtraction cannot be represented in 32 bits, then RT is updated with a value which is
31
“saturated” to the nearest representable value. That is, if the result is less than –2 , then RT is updated with
31
31
31
–2 . Likewise, if the result is greater than 2 – 1, then RT is updated with 2 – 1.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
nmaclhw
Negative Multiply Accumulate Low Halfword to Word Modulo Signed
PPC440x5 CPU Core User’s Manual
Preliminary
nmaclhw
Negative Multiply Accumulate Low Halfword to Word Modulo Signed
nmaclhw
nmaclhw.
nmaclhwo
nmaclhwo.
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
4
RT
RA
RB
OE
430
Rc
0
6
11
16
21 22
31
nprod0:31 ← –((RA)16:31 × (RB)
) signed
16:31
temp0:32 ← nprod
+ (RT)
0:31
(RT) ← temp
1:32
The low-order halfword of RA is multiplied by the low-order halfword of RB. The signed product is subtracted
from the contents of RT and RT is updated with the low-order 32 bits of the result.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
nmaclhws
Negative Multiply Accumulate High Halfword to Word Saturate Signed
PPC440x5 CPU Core User’s Manual
Preliminary
nmaclhws
Negative Multiply Accumulate Low Halfword to Word Saturate Signed
nmaclhws
nmaclhws.
nmaclhwso RT, RA, RB
nmaclhwso. RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
4
RT
RA
RB
OE
494
Rc
0
6
11
16
21 22
31
nprod0:31 ← –((RA)16:31 × (RB)
) signed
16:31
temp0:32 ← nprod
+ (RT)
0:31
if ((nprod = RT ) ∧ (RT ≠ temp )) then (RT) ← (RT || 31(¬RT ))
0
0
0
1
0
0
else (RT) ← temp
1:32
The low-order halfword of RA is multiplied by the low-order halfword of RB. The signed product is subtracted
from the contents of RT.
If the result of the subtraction can be represented in 32 bits, then RT is updated with the low-order 32 bits of
the result.
If the result of the subtraction cannot be represented in 32 bits, then RT is updated with a value which is
31
“saturated” to the nearest representable value. That is, if the result is less than –2 , then RT is updated with
31
31
31
–2 . Likewise, if the result is greater than 2 – 1, then RT is updated with 2 – 1.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Architecture Note
This instruction is implementation-specific and programs which use this instruction may not be portable to
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
nor
NOR
Preliminary
PPC440x5 CPU Core User’s Manual
nor
NOR
nor
nor.
RA, RS, RB
RA, RS, RB
Rc=0
Rc=1
31
RT
RA
RB
124
Rc
0
6
11
16
21
31
(RA) ← ¬((RS) ∨ (RB))
The contents of register RS is ORed with the contents of register RB; the ones complement of the result is
placed into register RA.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
Table 9-23. Extended Mnemonics for nor, nor.
Other Registers
Mnemonic
Operands
Function
Altered
Complement register.
(RA) ← ¬(RS)
not
Extended mnemonic for
RA, RS
nor RA,RS,RS
Extended mnemonic for
nor. RA,RS,RS
CR[CR0]
not.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
or
OR
PPC440x5 CPU Core User’s Manual
Preliminary
or
OR
or
or.
RA, RS, RB
RA, RS, RB
Rc=0
Rc=1
31
RS
RA
RB
444
Rc
0
6
11
16
21
31
(RA) ← (RS) ∨ (RB)
The contents of register RS is ORed with the contents of register RB; the result is placed into register RA.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
Table 9-24. Extended Mnemonics for or, or.
Other Registers
Mnemonic
Operands
Function
Altered
Move register.
(RT) ← (RS)
mr
Extended mnemonic for
RT, RS
or RT,RS,RS
Extended mnemonic for
or. RT,RS,RS
CR[CR0]
mr.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
orc
OR with Complement
Preliminary
PPC440x5 CPU Core User’s Manual
orc
OR with Complement
orc
orc.
RA, RS, RB
RA, RS, RB
Rc=0
Rc=1
31
RT
RA
RB
412
Rc
0
6
11
16
21
31
(RA) ← (RS) ∨ ¬(RB)
The contents of register RS is ORed with the ones complement of the contents of register RB; the result is
placed into register RA.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
ori
OR Immediate
PPC440x5 CPU Core User’s Manual
Preliminary
ori
OR Immediate
ori
RA, RS, IM
24
RS
RA
IM
0
6
11
16
31
16
(RA) ← (RS) ∨ ( 0 || IM)
The IM field is extended to 32 bits by concatenating 16 0-bits on the left. Register RS is ORed with the
extended IM field; the result is placed into register RA.
Registers Altered
• RA
Table 9-25. Extended Mnemonics for ori
Other Registers
Mnemonic
Operands
Function
Changed
Preferred no-op; triggers optimizations based on no-ops.
Extended mnemonic for
nop
ori 0,0,0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
oris
OR Immediate Shifted
Preliminary
PPC440x5 CPU Core User’s Manual
oris
OR Immediate Shifted
oris
RA, RS, IM
25
RS
RA
IM
0
6
11
16
31
16
(RA) ← (RS) ∨ (IM || 0)
The IM Field is extended to 32 bits by concatenating 16 0-bits on the right. Register RS is ORed with the
extended IM field and the result is placed into register RA.
Registers Altered
• RA
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
rfci
Return From Critical Interrupt
PPC440x5 CPU Core User’s Manual
Preliminary
rfci
Return From Critical Interrupt
rfci
19
51
0
6
21
31
(PC) ← (CSRR0)
(MSR) ← (CSRR1)
This instruction is used to return from a critical interrupt.
The program counter (PC) is restored with the contents of CSRR0 and the MSR is restored with the contents
of CSRR1.
Instruction execution returns to the address contained in the PC.
Registers Altered
• MSR
Programming Note
Execution of this instruction is privileged and context-synchronizing (see Context Synchronization on
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
rfi
Return From Interrupt
Preliminary
PPC440x5 CPU Core User’s Manual
rfi
Return From Interrupt
rfi
19
50
0
6
21
31
(PC) ← (SRR0)
(MSR) ← (SRR1)
This instruction is used to return from a non-critical interrupt.
The program counter (PC) is restored with the contents of SRR0 and the MSR is restored with the contents of
SRR1.
Instruction execution returns to the address contained in the PC.
Registers Altered
• MSR
Invalid Instruction Forms
• Reserved fields
Programming Note
Execution of this instruction is privileged and context-synchronizing (see Context Synchronization on
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
rfmci
Return From Machine Check Interrupt
PPC440x5 CPU Core User’s Manual
Preliminary
rfmci
Return From Critical Interrupt
rfmci
19
38
0
6
21
31
(PC) ← (MCSRR0)
(MSR) ← (MCSRR1)
This instruction is used to return from a machine check interrupt.
The program counter (PC) is restored with the contents of MCSRR0 and the MSR is restored with the
contents of MCSRR1.
Instruction execution returns to the address contained in the PC.
Registers Altered
• MSR
Programming Note
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
rlwimi
Rotate Left Word Immediate then Mask Insert
PPC440x5 CPU Core User’s Manual
Preliminary
rlwimi
Rotate Left Word Immediate then Mask Insert
rlwimi
rlwimi.
RA, RS, SH, MB, ME
RA, RS, SH, MB, ME
Rc=0
Rc=1
20
RS
RA
SH
MB
ME
Rc
0
6
11
16
21
26
31
r ← ROTL((RS), SH)
m ← MASK(MB, ME)
(RA) ← (r ∧ m) ∨ ((RA) ∧ ¬m)
The contents of register RS are rotated left by the number of bit positions specified in the SH field. A mask is
generated, having 1-bits starting at the bit position specified in the MB field and ending in the bit position
specified by the ME field, with 0-bits elsewhere.
If the starting point of the mask is at a higher bit position than the ending point, the 1-bits portion of the mask
wraps from the highest bit position back around to the lowest. The rotated data is inserted into register RA, in
positions corresponding to the bit positions in the mask that contain a 1-bit.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
Table 9-26. Extended Mnemonics for rlwimi, rlwimi.
Other Registers
Mnemonic
Operands
Function
Altered
Insert from left immediate (n > 0).
(RA)
Extended mnemonic for
← (RS)
b:b+n-1
0:n-1
inslwi
RA, RS, n, b
rlwimi RA,RS,32−b,b,b+n−1
Extended mnemonic for
rlwimi. RA,RS,32−b,b,b+n−1
CR[CR0]
inslwi.
insrwi
insrwi.
I(nRsAe)rt from rig←ht im(RmSe)diate. (n > 0)
b:b+n-1
32-n:31
Extended mnemonic for
RA, RS, n, b
rlwimi RA,RS,32−b−n,b,b+n−1
Extended mnemonic for
rlwimi. RA,RS,32−b−n,b,b+n−1
CR[CR0]
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
rlwinm
Rotate Left Word Immediate then AND with Mask
PPC440x5 CPU Core User’s Manual
Preliminary
rlwinm
Rotate Left Word Immediate then AND with Mask
rlwinm
rlwinm.
RA, RS, SH, MB, ME
RA, RS, SH, MB, ME
Rc=0
Rc=1
21
RS
RA
SH
MB
ME
Rc
0
6
11
16
21
26
31
r ← ROTL((RS), SH)
m ← MASK(MB, ME)
(RA) ← r ∧ m
The contents of register RS are rotated left by the number of bit positions specified in the SH field. A mask is
generated, having 1-bits starting at the bit position specified in the MB field and ending in the bit position
specified by the ME field with 0-bits elsewhere.
If the starting point of the mask is at a higher bit position than the ending point, the 1-bits portion of the mask
wraps from the highest bit position back around to the lowest. The rotated data is ANDed with the generated
mask; the result is placed into register RA.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
Table 9-27. Extended Mnemonics for rlwinm, rlwinm.
Other Registers
Mnemonic
Operands
Function
Altered
Clear left immediate. (n < 32)
n
(RA)
←
0
0:n-1
clrlwi
Extended mnemonic for
RA, RS, n
rlwinm RA,RS,0,n,31
Extended mnemonic for
rlwinm. RA,RS,0,n,31
CR[CR0]
clrlwi.
Clear left and shift left immediate.
(n ≤ b < 32)
(RA)
(RA)
(RA)
← (RS)
b:31
b-n:31-n
32-n:31
0:b-n-1
n
←
0
clrlslwi
clrlslwi.
b-n
←
0
RA, RS, b, n
Extended mnemonic for
rlwinm RA,RS,n,b−n,31−n
Extended mnemonic for
rlwinm. RA,RS,n,b−n,31−n
CR[CR0]
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
rlwinm
Rotate Left Word Immediate then AND with Mask
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-27. Extended Mnemonics for rlwinm, rlwinm. (continued)
Other Registers
Altered
Mnemonic
Operands
Function
Clear right immediate. (n < 32)
n
(RA)
←
0
32-n:31
clrrwi
Extended mnemonic for
RA, RS, n
rlwinm RA,RS,0,0,31−n
Extended mnemonic for
rlwinm. RA,RS,0,0,31−n
CR[CR0]
clrrwi.
E(RxAtr)act and←left(jRusSt)ify immediate. (n > 0)
0:n-1
n:31
b:b+n-1
0
32-n
(RA)
←
extlwi
extlwi.
extrwi
Extended mnemonic for
rlwinm RA,RS,b,0,n−1
RA, RS, n, b
Extended mnemonic for
rlwinm. RA,RS,b,0,n−1
CR[CR0]
Extract and right justify immediate. (n > 0)
(RA)
← (RS)
32-n:31
0:31-n
b:b+n-1
32-n
(RA)
←
0
Extended mnemonic for
RA, RS, n, b
rlwinm RA,RS,b+n,32−n,31
Extended mnemonic for
rlwinm. RA,RS,b+n,32−n,31
CR[CR0]
CR[CR0]
CR[CR0]
extrwi.
rotlwi
Rotate left immediate.
(RA) ← ROTL((RS), n)
Extended mnemonic for
RA, RS, n
rlwinm RA,RS,n,0,31
Extended mnemonic for
rlwinm. RA,RS,n,0,31
rotlwi.
rotrwi
rotrwi.
Rotate right immediate.
(RA) ← ROTL((RS), 32−n)
Extended mnemonic for
RA, RS, n
rlwinm RA,RS,32−n,0,31
Extended mnemonic for
rlwinm. RA,RS,32−n,0,31
S(RhAift) left imm←edia(RteS. )(n < 32)
32-n:31
Extended mnemonic for
0:31-n
n:31
n
(RA)
←
0
slwi
RA, RS, n
rlwinm RA,RS,n,0,31−n
Extended mnemonic for
rlwinm. RA,RS,n,0,31−n
CR[CR0]
slwi.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
rlwinm
Rotate Left Word Immediate then AND with Mask
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-27. Extended Mnemonics for rlwinm, rlwinm. (continued)
Other Registers
Altered
Mnemonic
Operands
Function
Shift right immediate. (n < 32)
(RA)
(RA)
← (RS)
0:31-n
n:31
0:n-1
n
←
0
srwi
Extended mnemonic for
rlwinm RA,RS,32−n,n,31
RA, RS, n
Extended mnemonic for
rlwinm. RA,RS,32−n,n,31
CR[CR0]
srwi.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
rlwnm
Rotate Left Word then AND with Mask
PPC440x5 CPU Core User’s Manual
Preliminary
rlwnm
Rotate Left Word then AND with Mask
rlwnm
rlwnm.
RA, RS, RB, MB, ME
RA, RS, RB, MB, ME
Rc=0
Rc=1
23
RS
RA
RB
MB
ME
Rc
0
6
11
16
21
26
31
r ← ROTL((RS), (RB)
m ← MASK(MB, ME)
(RA) ← r ∧ m
)
27:31
The contents of register RS are rotated left by the number of bit positions specified by the contents of register
RB . A mask is generated, having 1-bits starting at the bit position specified in the MB field and ending in
27:31
the bit position specified by the ME field with 0-bits elsewhere.
If the starting point of the mask is at a higher bit position than the ending point, the ones portion of the mask
wraps from the highest bit position back to the lowest. The rotated data is ANDed with the generated mask
and the result is placed into register RA.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
Table 9-28. Extended Mnemonics for rlwnm, rlwnm.
Other Registers
Mnemonic
Operands
Function
Altered
Rotate left.
(RA) ← ROTL((RS), (RB)
)
27:31
rotlw
Extended mnemonic for
RA, RS, RB
rlwnm RA,RS,RB,0,31
Extended mnemonic for
rlwnm. RA,RS,RB,0,31
CR[CR0]
rotlw.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
sc
System Call
PPC440x5 CPU Core User’s Manual
Preliminary
sc
System Call
sc
17
1
0
6
30 31
SRR1 ← MSR
SRR0 ← 4 + address of sc instruction
4
PC ← IVPR
|| IVOR8
|| 0
0:15
16:27
9
MSR[WE, EE, PR, FP, FE0, FE1, DWE, DS, IS] ← 0
A System Call exception is generated, and a System Call interrupt occurs (see System Call Interrupt on
and (4 + address of sc instruction) is placed into SRR0.
The program counter (PC) is then loaded with the interrupt vector address. The interrupt vector address is
formed by concatenating the high halfword of the Interrupt Vector Prefix Register (IVPR), bits 16:27 of the
Interrupt Vector Offset Register 8 (IVOR8), and 0b0000.
The MSR[WE, EE, PR, FP, FE0, FE1, DWE, DS, IS] bits are set to 0.
Program execution continues at the new address in the PC.
Registers Altered
• SRR0
• SRR1
• MSR[WE, EE, PR, FP, FE0, FE1, DWE, DS, IS]
Invalid Instruction Forms
• Reserved fields
Programming Note
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
slw
Shift Left Word
Preliminary
PPC440x5 CPU Core User’s Manual
slw
Shift Left Word
slw
slw.
RA, RS, RB
RA, RS, RB
Rc=0
Rc=1
31
RS
RA
RB
24
Rc
0
6
11
16
21
31
n ← (RB)
26:31
r ← ROTL((RS), n)
if n < 32 then
m ← MASK(0, 31 – n)
else
m ←
(RA) ← r ∧ m
32
0
The contents of register RS are shifted left by the number of bits specified by the contents of register RB
.
26:31
Bits shifted left out of the most significant bit are lost, and 0-bits fill vacated bit positions on the right. The
result is placed into register RA.
Note that if RB = 1, then the shift amount is 32 bits or more, and thus all bits are shifted out such that
26
register RA is set to zero.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
sraw
Shift Right Algebraic Word
PPC440x5 CPU Core User’s Manual
Preliminary
sraw
Shift Right Algebraic Word
sraw
sraw.
RA, RS, RB
RA, RS, RB
Rc=0
Rc=1
31
RS
RA
RB
792
Rc
0
6
11
16
21
31
n ← (RB)
26:31
r ← ROTL((RS), 32 – n)
if n < 32 then
m ← MASK(n, 31)
else
32
m ←
0
s ← (RS)
0
32
(RA) ← (r ∧ m) ∨ ( s ∧ ¬m)
XER[CA] ← s ∧ ((r ∧ ¬m) ≠ 0)
The contents of register RS are shifted right by the number of bits specified the contents of register RB
.
26:31
Bits shifted out of the least significant bit are lost. Bit 0 of Register RS is replicated to fill the vacated positions
on the left. The result is placed into register RA.
If register RS contains a negative number and any 1-bits were shifted out of the least significant bit position,
XER[CA] is set to 1; otherwise, it is set to 0.
Note that if RB = 1, then the shift amount is 32 bits or more, and thus all bits are shifted out such that
26
register RA and XER[CA] are set to bit 0 of register RS.
Registers Altered
• RA
• XER[CA]
• CR[CR0] if Rc contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
srawi
Shift Right Algebraic Word Immediate
PPC440x5 CPU Core User’s Manual
Preliminary
srawi
Shift Right Algebraic Word Immediate
srawi
srawi.
RA, RS, SH
RA, RS, SH
Rc=0
Rc=1
31
RS
RA
SH
824
Rc
0
6
11
16
21
31
n ← SH
r ← ROTL((RS), 32 – n)
m ← MASK(n, 31)
s ← (RS)
0
32
(RA) ← (r ∧ m) ∨ ( s ∧ ¬m)
XER[CA] ← s ∧ ((r ∧ ¬m)≠0)
The contents of register RS are shifted right by the number of bits specified in the SH field. Bits shifted out of
the least significant bit are lost. Bit RS is replicated to fill the vacated positions on the left. The result is
0
placed into register RA.
If register RS contains a negative number and any 1-bits were shifted out of the least significant bit position,
XER[CA] is set to 1; otherwise, it is set to 0.
Registers Altered
• RA
• XER[CA]
• CR[CR0] if Rc contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
srw
Shift Right Word
PPC440x5 CPU Core User’s Manual
Preliminary
srw
Shift Right Word
srw
srw.
RA, RS, RB
RA, RS, RB
Rc=0
Rc=1
31
RS
RA
RB
536
Rc
0
6
11
16
21
31
n ← (RB)
26:31
r ← ROTL((RS), 32 – n)
if n < 32 then
m ← MASK(n, 31)
else
32
m ←
0
(RA) ← r ∧ m
The contents of register RS are shifted right by the number of bits specified the contents of register RB
.
26:31
Bits shifted right out of the least significant bit are lost, and 0-bits fill the vacated bit positions on the left. The
result is placed into register RA.
Note that if RB = 1, then the shift amount is 32 bits or more, and thus all bits are shifted out such that
26
register RA is set to zero.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stb
Store Byte
Preliminary
PPC440x5 CPU Core User’s Manual
stb
Store Byte
stb
RS, D(RA)
38
RS
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
MS(EA, 1) ← (RS)
24:31
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 when the RA field is 0, and is
the contents of register RA otherwise.
The least significant byte of register RS is stored into the byte at the EA.
Registers Altered
• None
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stbu
Store Byte with Update
PPC440x5 CPU Core User’s Manual
Preliminary
stbu
Store Byte with Update
stbu
RS, D(RA)
39
RS
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
MS(EA, 1) ← (RS)
24:31
(RA) ← EA
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 when the RA field is 0, and is
the contents of register RA otherwise.
The least significant byte of register RS is stored into the byte at the EA.
The EA is placed into register RA.
Registers Altered
• RA
Invalid Instruction Forms
RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stbux
Store Byte with Update Indexed
Preliminary
PPC440x5 CPU Core User’s Manual
stbux
Store Byte with Update Indexed
stbux
RS, RA, RB
31
RS
RA
RB
247
0
6
11
16
21
31
EA ← (RA|0) + (RB)
MS(EA, 1) ← (RS)
24:31
(RA) ← EA
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 when the RA field is 0, and is the contents of register RA otherwise.
The least significant byte of register RS is stored into the byte at the EA.
The EA is placed into register RA.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RA
Invalid Instruction Forms
• Reserved fields
RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stbx
Store Byte Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
stbx
Store Byte Indexed
stbx
RS, RA, RB
31
RS
RA
RB
215
0
6
11
16
21
31
EA ← (RA|0) + (RB)
MS(EA, 1) ← (RS)
24:31
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 when the RA field is 0, and is the contents of register RA otherwise.
The least significant byte of register RS is stored into the byte at the EA.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
sth
Store Halfword
Preliminary
PPC440x5 CPU Core User’s Manual
sth
Store Halfword
sth
RS, D(RA)
44
RS
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
MS(EA, 2) ← (RS)
16:31
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 when the RA field is 0 and is
the contents of register RA otherwise.
The least significant halfword of register RS is stored into the halfword at the EA in main storage.
Registers Altered
• None
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
sthbrx
Store Halfword Byte-Reverse Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
sthbrx
Store Halfword Byte-Reverse Indexed
sthbrx
RS, RA, RB
31
RS
RA
RB
918
0
6
11
16
21
31
EA ← (RA|0) + (RB)
MS(EA, 2) ← BYTE_REVERSE((RS)
)
16:31
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 when the RA field is 0 and is the contents of register RA otherwise.
The least significant halfword of register RS is byte-reversed from the default byte ordering for the memory
page referenced by the EA. The resulting halfword is stored at the EA.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Note
page 133). The store byte reverse instructions provide a mechanism for data to be stored to a memory page
using the opposite byte ordering from that specified by the Endian storage attribute.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
sthu
Store Halfword with Update
Preliminary
PPC440x5 CPU Core User’s Manual
sthu
Store Halfword with Update
sthu
RS, D(RA)
45
RS
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
MS(EA, 2) ← (RS)
16:31
(RA) ← EA
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 when the RA field is 0, and is
the contents of register RA otherwise.
The least significant halfword of register RS is stored into the halfword at the EA.
The EA is placed into register RA.
Registers Altered
• RA
Invalid Instruction Forms
RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
sthux
Store Halfword with Update Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
sthux
Store Halfword with Update Indexed
sthux
RS, RA, RB
31
RS
RA
RB
439
0
6
11
16
21
31
EA ← (RA|0) + (RB)
MS(EA, 2) ← (RS)
16:31
(RA) ← EA
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 when the RA field is 0, and is the contents of register RA otherwise.
The least significant halfword of register RS is stored into the halfword at the EA.
The EA is placed into register RA.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RA
Invalid Instruction Forms
• Reserved fields
• RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
sthx
Store Halfword Indexed
Preliminary
PPC440x5 CPU Core User’s Manual
sthx
Store Halfword Indexed
sthx
RS, RA, RB
31
RS
RA
RB
407
0
6
11
16
21
31
EA ← (RA|0) + (RB)
MS(EA, 2) ← (RS)
16:31
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 when the RA field is 0, and is the contents of register RA otherwise.
The least significant halfword of register RS is stored into the halfword at the EA.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stmw
Store Multiple Word
PPC440x5 CPU Core User’s Manual
Preliminary
stmw
Store Multiple Word
stmw
RS, D(RA)
47
RS
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
r ← RS
do while r ≤ 31
MS(EA, 4) ← (GPR(r))
r ← r + 1
EA ← EA + 4
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 when the RA field is 0, and is
the contents of register RA otherwise.
The contents of a series of consecutive registers, starting with register RS and continuing through GPR(31),
are stored into consecutive words starting at the EA.
Registers Altered
• None
Programming Note
This instruction can be restarted, meaning that it could be interrupted after having already stored some of the
register values to memory, and then re-executed from the beginning (after returning from the interrupt), in
which case the registers which had already been stored prior to the interrupt will be stored a second time.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stswi
Store String Word Immediate
Preliminary
PPC440x5 CPU Core User’s Manual
stswi
Store String Word Immediate
stswi
RS, RA, NB
31
RS
RA
NB
725
0
6
11
16
21
31
EA ← (RA|0)
if NB = 0 then
n ← 32
else
n ← NB
r ← RS – 1
i ← 0
do while n > 0
if i = 0 then
r ← r + 1
if r = 32 then
r ← 0
MS(EA,1) ← (GPR(r)
i ← i + 8
)
i:i+7
if i = 32 then
i ← 0
EA ← EA + 1
n ← n – 1
An effective address (EA) is determined by the RA field. If the RA field contains 0, the EA is 0; otherwise, the
EA is the contents of register RA.
A byte count is determined by the NB field. If the NB field contains 0, the byte count is 32; otherwise, the byte
count is the contents of the NB field.
The contents of a series of consecutive GPRs (starting with register RS, continuing through GPR(31) and
wrapping to GPR(0) as necessary, and continuing to the final byte count) are stored, starting at the EA. The
bytes in each GPR are accessed starting with the most significant byte. The byte count determines the
number of transferred bytes.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stswi
Store String Word Immediate
PPC440x5 CPU Core User’s Manual
Preliminary
Programming Note
This instruction can be restarted, meaning that it could be interrupted after having already stored some of the
register values to memory, and then re-executed from the beginning (after returning from the interrupt), in
which case the registers which had already been stored prior to the interrupt will be stored a second time.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stswx
Store String Word Indexed
Preliminary
PPC440x5 CPU Core User’s Manual
stswx
Store String Word Indexed
stswx
RS, RA, RB
31
RS
RA
RB
661
0
6
11
16
21
31
EA ← (RA|0) + (RB)
n ← XER[TBC]
r ← RS – 1
i ← 0
do while n > 0
if i = 0 then
r ← r + 1
if r = 32 then
r ← 0
MS(EA, 1) ← (GPR(r)
i ← i + 8
)
i:i+7
if i = 32 then
i ← 0
EA ← EA + 1
n ← n – 1
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 when the RA field is 0, and is the contents of register RA otherwise.
A byte count is contained in XER[TBC].
The contents of a series of consecutive GPRs (starting with register RS, continuing through GPR(31) and
wrapping to GPR(0) as necessary, and continuing to the final byte count) are stored, starting at the EA. The
bytes in each GPR are accessed starting with the most significant byte. The byte count determines the
number of transferred bytes.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Note
This instruction can be restarted, meaning that it could be interrupted after having already stored some of the
register values to memory, and then re-executed from the beginning (after returning from the interrupt), in
which case the registers which had already been stored prior to the interrupt will be stored a second time.
If XER[TBC] = 0, no GPRs are stored to memory, and stswx is treated as a no-op. Furthermore, if the EA is
such that a Data Storage, Data TLB Error, or Data Address Compare Debug exception occurs, stswx is
treated as a no-op and no interrupt occurs as a result of the exception.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stw
Store Word
PPC440x5 CPU Core User’s Manual
Preliminary
stw
Store Word
stw
RS, D(RA)
36
RS
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
MS(EA, 4) ← (RS)
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 when the RA field is 0, and is
the contents of register RA otherwise.
The contents of register RS are stored at the EA.
Registers Altered
• None
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stwbrx
Store Word Byte-Reverse Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
stwbrx
Store Word Byte-Reverse Indexed
stwbrx
RS, RA, RB
31
RS
RA
RB
662
0
6
11
16
21
31
EA ← (RA|0) + (RB)
MS(EA, 4) ← BYTE_REVERSE((RS)
)
0:31
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 when the RA field is 0 and is the contents of register RA otherwise.
The word in register RS is byte-reversed from the default byte ordering for the memory page referenced by
the EA. The resulting word is stored at the EA.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Note
page 133). The store byte reverse instructions provide a mechanism for data to be stored to a memory page
using the opposite byte ordering from that specified by the Endian storage attribute.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stwcx.
Store Word Conditional Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
stwcx.
Store Word Conditional Indexed
stwcx.
RS, RA, RB
31
RS
RA
RB
150
1
0
6
11
16
21
31
EA ← (RA|0) + (RB)
if RESERVE = 1 then
MS(EA, 4) ← (RS)
RESERVE ← 0
2
(CR[CR0]) ← 0 || 1 || XER[SO]
else
2
(CR[CR0]) ← 0 || 0 || XER[SO]
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 when the RA field is 0, and is the contents of register RA otherwise.
If the reservation bit contains 1 when the instruction is executed, the contents of register RS are stored into
the word at the EA and the reservation bit is cleared. If the reservation bit contains 0 when the instruction is
executed, no store operation is performed.
CR[CR0] is set as follows:
• CR[CR0] are cleared
0:1
• CR[CR0] is set to indicate whether or not the store was performed (1 indicates that it was)
2
• CR[CR0] is set to the contents of the XER[SO] bit
3
Registers Altered
• CR[CR0]
Programming Notes
The lwarx and stwcx. instructions are typically paired in a loop, as shown in the following example, to create
the effect of an atomic operation to a memory area used as a semaphore between multiple processes. Only
lwarx can set the reservation bit to 1. stwcx. sets the reservation bit to 0 upon its completion, whether or not
stwcx. actually stored (RS) to memory. CR[CR0] must be examined to determine whether (RS) was sent to
2
memory.
loop: lwarx
“alter”
# read the semaphore from memory; set reservation
# change the semaphore bits in the register as required
# attempt to store the semaphore; reset reservation
# some other process intervened and cleared the reservation prior to the above
# stwcx.; try again
stwcx.
bne loop
The PowerPC Book-E architecture specifies that the EA for the lwarx instruction must be word-aligned (that
is, a multiple of 4 bytes); otherwise, the result is undefined. Although the PPC440x5 will execute stwcx.
regardless of the EA alignment, in order for the operation of the pairing of lwarx and stwcx. to produce the
desired result, software must ensure that the EA for both instructions is word-aligned. This requirement is due
to the manner in which misaligned storage accesses may be broken up into separate, aligned accesses by
the PPC440x5.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stwcx.
Store Word Conditional Indexed
Preliminary
PPC440x5 CPU Core User’s Manual
The PowerPC Book-E architecture also specifies that it is implementation-dependent as to whether a Data
Storage, Data TLB Error, Alignment, or Debug interrupt occurs when the reservation bit is off at the time of
execution of an stwcx. instruction, and when the conditions are such that a non-stwcx. store-type storage
access instruction would have resulted in such an interrupt. The PPC440x5 implements stwcx. such that
Data Storage and Debug (DAC and/or DVC exception type) interrupts do not occur when the reservation bit is
off at the time of execution of the stwcx. Instead, the stwcx. instruction completes without causing the inter-
rupt and without storing to memory, and CR[CR0] is updated to indicate the failure of the stwcx.
On the other hand, the PPC440x5 causes a Data TLB Error interrupt if a Data TLB Miss exception occurs
during due to the execution of a stwcx. instruction, regardless of the state of the reservation. Similarly, the
PPC440x5 causes an Alignment interrupt if the EA of the stwcx. operand is not word-aligned when
CCR0[FLSTA] is 1, regardless of the state of the reservation (see Core Configuration Register 0 (CCR0) on
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stwu
Store Word with Update
PPC440x5 CPU Core User’s Manual
Preliminary
stwu
Store Word with Update
stwu
RS, D(RA)
37
RS
RA
D
0
6
11
16
31
EA ← (RA|0) + EXTS(D)
MS(EA, 4) ← (RS)
(RA) ← EA
An effective address (EA) is formed by adding a displacement to a base address. The displacement is
obtained by sign-extending the 16-bit D field to 32 bits. The base address is 0 when the RA field is 0, and is
the contents of register RA otherwise.
The contents of register RS are stored into the word at the EA.
The EA is placed into register RA.
Registers Altered
• RA
Invalid Instruction Forms
RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stwux
Store Word with Update Indexed
Preliminary
PPC440x5 CPU Core User’s Manual
stwux
Store Word with Update Indexed
stwux
RS, RA, RB
31
RS
RA
RB
183
0
6
11
16
21
31
EA ← (RA|0) + (RB)
MS(EA, 4) ← (RS)
(RA) ← EA
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 when the RA field is 0, and is the contents of register RA otherwise.
The contents of register RS are stored into the word at the EA.
The EA is placed into register RA.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• RA
Invalid Instruction Forms
• Reserved fields
• RA = 0
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
stwx
Store Word Indexed
PPC440x5 CPU Core User’s Manual
Preliminary
stwx
Store Word Indexed
stwx
RS, RA, RB
31
RS
RA
RB
151
0
6
11
16
21
31
EA ← (RA|0) + (RB)
MS(EA,4) ← (RS)
An effective address (EA) is formed by adding an index to a base address. The index is the contents of
register RB. The base address is 0 when the RA field is 0, and is the contents of register RA otherwise.
The contents of register RS are stored into the word at the EA.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
subf
Subtract From
Preliminary
PPC440x5 CPU Core User’s Manual
subf
Subtract From
subf
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
subf.
subfo
subfo.
31
RT
RA
RB
OE
40
Rc
0
6
11
16
21 22
31
(RT) ← ¬(RA) + (RB) + 1
The sum of the ones complement of register RA, register RB, and 1 is stored into register RT.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Table 9-29. Extended Mnemonics for subf, subf., subfo, subfo.
Other Registers
Altered
Mnemonic
Operands
Function
Subtract (RB) from (RA).
(RT) ← ¬(RB) + (RA) + 1.
sub
Extended mnemonic for
subf RT,RB,RA
Extended mnemonic for
subf. RT,RB,RA
CR[CR0]
sub.
RT, RA, RB
Extended mnemonic for
subfo RT,RB,RA
XER[SO, OV]
subo
subo.
Extended mnemonic for
subfo. RT,RB,RA
CR[CR0]
XER[SO, OV]
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
subfc
Subtract From Carrying
PPC440x5 CPU Core User’s Manual
Preliminary
subfc
Subtract From Carrying
subfc
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
subfc.
subfco
subfco.
31
RT
RA
RB
OE
8
Rc
0
6
11
16
21 22
31
(RT) ← ¬(RA) + (RB) + 1
u
>
32
if ¬(RA) + (RB) + 1
XER[CA] ← 1
else
2
– 1 then
XER[CA] ← 0
The sum of the ones complement of register RA, register RB, and 1 is stored into register RT.
XER[CA] is set to a value determined by the unsigned magnitude of the result of the subtract operation.
Registers Altered
• RT
• XER[CA]
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Table 9-30. Extended Mnemonics for subfc, subfc., subfco, subfco.
Other Registers
Mnemonic
Operands
Function
Altered
Subtract (RB) from (RA).
(RT) ← ¬(RB) + (RA) + 1.
Place carry-out in XER[CA].
Extended mnemonic for
subc
subfc RT,RB,RA
Extended mnemonic for
subfc. RT,RB,RA
RT, RA, RB
CR[CR0]
subc.
Extended mnemonic for
subfco RT,RB,RA
XER[SO, OV]
subco
subco.
Extended mnemonic for
subfco. RT,RB,RA
CR[CR0]
XER[SO, OV]
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
subfe
Subtract From Extended
Preliminary
PPC440x5 CPU Core User’s Manual
subfe
Subtract From Extended
subfe
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
subfe.
subfeo
subfeo.
31
RT
RA
RB
OE
136
Rc
0
6
11
16
21 22
31
(RT) ← ¬(RA) + (RB) + XER[CA]
u
>
32
if ¬(RA) + (RB) + XER[CA]
XER[CA] ← 1
else
2
– 1 then
XER[CA] ← 0
The sum of the ones complement of register RA, register RB, and XER[CA] is placed into register RT.
XER[CA] is set to a value determined by the unsigned magnitude of the result of the subtract operation.
Registers Altered
• RT
• XER[CA]
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
subfic
Subtract From Immediate Carrying
PPC440x5 CPU Core User’s Manual
Preliminary
subfic
Subtract From Immediate Carrying
subfic
RT, RA, IM
8
RT
RA
IM
0
6
11
16
31
(RT) ← ¬(RA) + EXTS(IM) + 1
u
>
32
if ¬(RA) + EXTS(IM) + 1
XER[CA] ← 1
else
2
– 1 then
XER[CA] ← 0
The sum of the ones complement of RA, the IM field sign-extended to 32 bits, and 1 is placed into register
RT.
XER[CA] is set to a value determined by the unsigned magnitude of the result of the subtract operation.
Registers Altered
• RT
• XER[CA]
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
subfme
Subtract from Minus One Extended
PPC440x5 CPU Core User’s Manual
Preliminary
subfme
Subtract from Minus One Extended
subfme
RT, RA
RT, RA
RT, RA
RT, RA
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
subfme.
subfmeo
subfmeo.
31
RT
RA
OE
232
Rc
0
6
11
16
21 22
31
(RT) ← ¬(RA) – 1 + XER[CA]
if ¬(RA) + 0xFFFF FFFF + XER[CA]
XER[CA] ← 1
u
>
32
2
– 1 then
else
XER[CA] ← 0
The sum of the ones complement of register RA, –1, and XER[CA] is placed into register RT.
XER[CA] is set to a value determined by the unsigned magnitude of the result of the subtract operation.
Registers Altered
• RT
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
• XER[CA]
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
subfze
Subtract from Zero Extended
PPC440x5 CPU Core User’s Manual
Preliminary
subfze
Subtract from Zero Extended
subfze
RT, RA
RT, RA
RT, RA
RT, RA
OE=0, Rc=0
OE=0, Rc=1
OE=1, Rc=0
OE=1, Rc=1
subfze.
subfzeo
subfzeo.
31
RT
RA
OE
200
Rc
0
6
11
16
21 22
31
(RT) ← ¬(RA) + XER[CA]
u
>
32
if ¬(RA) + XER[CA]
XER[CA] ← 1
else
2
– 1 then
XER[CA] ← 0
The sum of the ones complement of register RA and XER[CA] is stored into register RT.
XER[CA] is set to a value determined by the unsigned magnitude of the result of the subtract operation.
Registers Altered
• RT
• XER[CA]
• CR[CR0] if Rc contains 1
• XER[SO, OV] if OE contains 1
Invalid Instruction Forms
• Reserved fields
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
tlbre
TLB Read Entry
Preliminary
PPC440x5 CPU Core User’s Manual
tlbre
TLB Read Entry
tlbre
RT, RA, WS
31
RT
RA
WS
946
0
6
11
16
21
31
tlbentry ← TLB[(RA)
]
26:31
if WS =0
(RT)0:27 ← tlbentry[EPN,V,TS,SIZE]
if CCR0[CRPE] = 0
4
(RT)28:31 ← 0
else
(RT)28:31 ← TPAR
MMUCR[STID] ← tlbentry[TID]
else if WS = 1
(RT)0:21 ← tlbentry[RPN]
if CCR0[CRPE] = 0
2
(RT)22:23 ← 0
else
(RT)22:23 ← PAR1
4
(RT)24:27 ← 0
(RT)28:31 ← tlbentry[ERPN]
else if WS = 2
if CCR0[CRPE] = 0
2
(RT)0:1 ← 0
else
(RT)0:1 ← PAR2
14
(RT)
←
0
2:15
(RT)16:24 ← tlbentry[U0,U1,U2,U3,W,I,M,G,E]
(RT)25 ← 0
(RT)26:31 ← tlbentry[UX,UW,UR,SX,SW,SR]
else (RT), MMUCR[STID] ← undefined
The contents of the specified portion of the selected TLB entry are placed into register RT (and also
MMUCR[STID] if WS = 0).
The parity bits in the TLB entry (TPAR, PAR1, and PAR2) are placed into the register RT if and only if the
Cache Read Parity Enable bit, CCR0[CRPE], is set to 1.
The contents of RA are used as an index into the TLB. If this value is greater than the index of the highest
numbered TLB entry (63), the results are undefined.
The WS field specifies which portion of the TLB entry is placed into RT. If WS = 0, the TID field of the selected
entry fields.
If the value of the WS field is greater than 2, the instruction form is invalid and the result is undefined.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
tlbre
TLB Read Entry
PPC440x5 CPU Core User’s Manual
Preliminary
Registers Altered
• RT
• MMUCR[STID] (if WS = 0)
Invalid Instruction Forms
• Reserved fields
• Invalid WS value
Programming Notes
Execution of this instruction is privileged.
The PPC440x5 core does not automatically synchronize the context of the MMUCR[STID] field between a
tlbre instruction which updates the field, and a tlbsx[.] instruction which uses it as a source operand. There-
fore, software must execute an isync instruction between the execution of a tlbre instruction and a subse-
quent tlbsx[.] instruction to ensure that the tlbsx[.] instruction will use the new value of MMUCR[STID]. On
the other hand, the PPC440x5 core does automatically synchronize the context of MMUCR[STID] between
tlbre and tlbwe, as well as between tlbre and mfspr which specifies the MMUCR as the source SPR, so no
isync is required in these cases.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
tlbsx
TLB Search Indexed
Preliminary
PPC440x5 CPU Core User’s Manual
tlbsx
TLB Search Indexed
tlbsx
tlbsx.
RT, RA, RB
RT, RA, RB
Rc=0
Rc=1
31
RT
RA
RB
914
Rc
0
6
11
16
21
31
EA ← (RA|0) + (RB)
if Rc = 1
CR[CR0] ← 0
0
CR[CR0] ← 0
1
CR[CR0] ← XER[SO}
3
if Valid TLB entry matching EA and MMUCR[STID,STS] is in the TLB then
(RT) ← Index of matching TLB Entry
if Rc = 1
CR[CR0] ← 1
2
else
(RT) ← Undefined
if Rc = 1
CR[CR0] ← 0
2
An effective address is formed by adding an index to a base address. The index is the contents of register
RB. The base address is 0 if the RA field is 0 and is the contents of register RA otherwise.
The TLB is searched for a valid entry which translates EA and MMUCR[STID,STS]. See Memory Manage-
ment on page 133 for descriptions of the TLB fields and how they participate in the determination of a match.
If a matching entry is found, its index (0 - 63) is placed into bits 26:31 of RT, and bits 0:25 are set to 0. If no
match is found, the contents of RT are undefined.
The record bit (Rc) specifies whether the results of the search will affect CR[CR0] as shown above, such that
CR[CR0] can be tested if there is a possibility that the search may fail.
2
Registers Altered
• CR[CR0] if Rc contains 1
Invalid Instruction Forms
• None
Programming Notes
Execution of this instruction is privileged.
The PPC440x5 core does not automatically synchronize the context of the MMUCR[STID] field between a
tlbre instruction which updates the field, and a tlbsx[.] instruction which uses it as a source operand. There-
fore, software must execute an isync instruction between the execution of a tlbre instruction and a subse-
quent tlbsx[.] instruction to ensure that the tlbsx[.] instruction will use the new value of MMUCR[STID]. On
the other hand, the PPC440x5 core does automatically synchronize the context of MMUCR[STID] between
tlbre and tlbwe, as well as between tlbre and mfspr which specifies the MMUCR as the source SPR, so no
isync is required in these cases.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
tlbsync
TLB Synchronize
PPC440x5 CPU Core User’s Manual
Preliminary
tlbsync
TLB Synchronize
tlbsync
31
566
0
6
21
31
The tlbsync instruction is provided by the PowerPC Book-E architecture to support synchronization of TLB
operations between processors in a coherent multi-processor system. Since the PPC440x5 core does not
support coherent multi-processing, this instruction performs no operation, and is provided only to facilitate
code portability.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
Programming Note
This instruction is privileged. Translation is not required to be active during the execution of this instruction.
Since the PPC440x5 core does not support tightly-coupled multiprocessor systems, tlbsync performs no
operation.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
tlbwe
TLB Write Entry
Preliminary
PPC440x5 CPU Core User’s Manual
tlbwe
TLB Write Entry
tlbwe
RS, RA, WS
31
RS
RA
WS
978
0
6
11
16
21
31
tlbentry ← TLB[(RA)
]
26:31
if WS = 0
tlbentry[EPN,V,TS,SIZE] ← (RS)
tlbentry[TID] ← MMUCR[STID]
else if WS = 1
0:27
tlbentry[RPN] ← (RS)
0:21
tlbentry[ERPN] ← (RS)
28:31
else if WS = 2
tlbentry[U0,U1,U2,U3,W,I,M,G,E] ← (RS)
16:24
tlbentry[UX,UW,UR,SX,SW,SR] ← (RS)
26:31
else tlbentry ← undefined
The contents of the specified portion of the selected TLB entry are replaced with the contents of register RS
(and also MMUCR[STID] if WS = 0).
Parity check bits are automatically calculated and stored in the TLB entry as the tlbwe is executed. The
contents of the RS register in the TPAR, PAR1, and PAR2 fields (for WS=0,1,or 2, respectively) is ignored by
tlbwe; the parity is calculated from the other data bits being written to the TLB entry.
The contents of RA are used as an index into the TLB. If this value is greater than the index of the highest
numbered TLB entry (63), the results are undefined.
The WS field specifies which portion of the TLB entry is replaced by the contents of RS. If WS = 0, the TID
If the value of the WS field is greater than 2, the instruction form is invalid and the result is undefined.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• None
Invalid Instruction Forms
• Reserved fields
• Invalid WS value
Programming Note
Execution of this instruction is privileged.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
tw
Trap Word
PPC440x5 CPU Core User’s Manual
Preliminary
tw
Trap Word
tw
TO, RA, RB
31
TO
RA
RB
4
0
6
11
16
21
31
if ( ((RA)
((RA)
(RB) ∧ TO = 1)
∨
∨
∨
∨
)
<
>
0
(RB) ∧ TO = 1)
1
((RA) = (RB) ∧ TO = 1)
((RA)
((RA)
2
u
(RB) ∧ TO = 1)
<
3
u
(RB) ∧ TO = 1)
>
4
SRR0 ← address of tw instruction
SRR1 ← MSR
ESR[PTR] ← 1 (other bits cleared)
9
MSR[WE, EE, PR, FP, FE0, FE1, DWE, DS, IS]) ← 0
4
PC ← IVPR
|| IVOR6
|| 0
0:15
16:27
else no operation
Register RA is compared with register RB. If any comparison condition enabled by the TO field is true, a Trap
exception type Program interrupt occurs as follows (see Program Interrupt on page 187 for more information
on Program interrupts). The contents of the MSR are copied into SRR1 and the address of the tw instruction)
is placed into SRR0. ESR[PTR] is set to 1 and the other bits ESR bits cleared to indicate the type of exception
causing the Program interrupt.
The program counter (PC) is then loaded with the interrupt vector address. The interrupt vector address is
formed by concatenating the high halfword of the Interrupt Vector Prefix Register (IVPR), bits 16:27 of the
Interrupt Vector Offset Register 6 (IVOR6), and 0b0000.
MSR[WE, EE, PR, FP, FE0, FE1, DWE, DS, IS] are set to 0.
Program execution continues at the new address in the PC.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• SRR0 (if trap condition is met)
• SRR1 (if trap condition is met)
• MSR[WE, EE, PR, FP, FE0, FE1, DWE, DS, IS] (if trap condition is met)
• ESR (if trap condition is met)
Invalid Instruction Forms
• Reserved fields
Programming Notes
This instruction can be inserted into the execution stream by a debugger to implement breakpoints, and is not
typically used by application code.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
tw
Trap Word
Preliminary
PPC440x5 CPU Core User’s Manual
The enabling of trap debug events may affect the interrupt type caused by the execution of tw instruction.
Specifically, trap instructions may be enabled to cause Debug interrupts instead of Program interrupts. See
Table 9-31. Extended Mnemonics for tw
Other Registers
Mnemonic
Operands
Function
Altered
Trap unconditionally.
Extended mnemonic for
trap
tw 31,0,0
Trap if (RA) equal to (RB).
Extended mnemonic for
tw 4,RA,RB
RA, RB
tweq
twge
twgt
twle
Trap if (RA) greater than or equal to (RB).
Extended mnemonic for
tw 12,RA,RB
RA, RB
RA, RB
RA, RB
RA, RB
RA, RB
RA, RB
RA, RB
RA, RB
RA, RB
RA, RB
RA, RB
Trap if (RA) greater than (RB).
Extended mnemonic for
tw 8,RA,RB
Trap if (RA) less than or equal to (RB).
Extended mnemonic for
tw 20,RA,RB
Trap if (RA) logically greater than or equal to (RB).
Extended mnemonic for
tw 5,RA,RB
twlge
twlgt
twlle
twllt
Trap if (RA) logically greater than (RB).
Extended mnemonic for
tw 1,RA,RB
Trap if (RA) logically less than or equal to (RB).
Extended mnemonic for
tw 6,RA,RB
Trap if (RA) logically less than (RB).
Extended mnemonic for
tw 2,RA,RB
Trap if (RA) logically not greater than (RB).
Extended mnemonic for
tw 6,RA,RB
twlng
twlnl
twlt
Trap if (RA) logically not less than (RB).
Extended mnemonic for
tw 5,RA,RB
Trap if (RA) less than (RB).
Extended mnemonic for
tw 16,RA,RB
Trap if (RA) not equal to (RB).
Extended mnemonic for
twne
tw 24,RA,RB
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
tw
Trap Word
PPC440x5 CPU Core User’s Manual
Preliminary
Table 9-31. Extended Mnemonics for tw (continued)
Other Registers
Altered
Mnemonic
Operands
Function
Trap if (RA) not greater than (RB).
Extended mnemonic for
RA, RB
twng
tw 20,RA,RB
Trap if (RA) not less than (RB).
Extended mnemonic for
RA, RB
twnl
tw 12,RA,RB
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
twi
Trap Word Immediate
Preliminary
PPC440x5 CPU Core User’s Manual
twi
Trap Word Immediate
twi
TO, RA, IM
3
TO
RA
IM
0
6
11
16
31
if ( ((RA)
((RA)
EXTS(IM) ∧ TO = 1)
∨
<
>
0
EXTS(IM) ∧ TO = 1)
∨
∨
∨
)
1
((RA) = EXTS(IM) ∧ TO = 1)
2
u
((RA)
((RA)
EXTS(IM) ∧ TO = 1)
<
3
u
EXTS(IM) ∧ TO = 1)
>
4
SRR0 ← address of twi instruction
SRR1 ← MSR
ESR[PTR] ← 1 (other bits cleared)
9
MSR[WE, EE, PR, FP, FE0, FE1, DWE, DS, IS]) ← 0
4
PC ← IVPR
|| IVOR6
|| 0
0:15
16:27
else no operation
Register RA is compared with the sign-extended IM field. If any comparison condition selected by the TO field
is true, a Trap exception type Program interrupt occurs as follows (see Program Interrupt on page 187 for
more information on Program interrupts). The contents of the MSR are copied into SRR1 and the address of
the twi instruction) is placed into SRR0. ESR[PTR] is set to 1 and the other bits ESR bits cleared to indicate
the type of exception causing the Program interrupt.
The program counter (PC) is then loaded with the interrupt vector address. The interrupt vector address is
formed by concatenating the high halfword of the Interrupt Vector Prefix Register (IVPR), bits 16:27 of the
Interrupt Vector Offset Register 6 (IVOR6), and 0b0000.
MSR[WE, EE, PR, FP, FE0, FE1, DWE, DS, IS] are set to 0.
Program execution continues at the new address in the PC.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• SRR0 (if trap condition is met)
• SRR1 (if trap condition is met)
• MSR[WE, EE, PR, FP, FE0, FE1, DWE, DS, IS] (if trap condition is met)
• ESR (if trap condition is met)
Invalid Instruction Forms
• Reserved fields
Programming Notes
This instruction can be inserted into the execution stream by a debugger to implement breakpoints, and is not
typically used by application code.
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
twi
Trap Word Immediate
PPC440x5 CPU Core User’s Manual
Preliminary
The enabling of trap debug events may affect the interrupt type caused by the execution of tw instruction.
Specifically, trap instructions may be enabled to cause Debug interrupts instead of Program interrupts. See
Table 9-32. Extended Mnemonics for twi
Other Registers
Mnemonic
Operands
Function
Trap if (RA) equal to EXTS(IM).
Altered
Extended mnemonic for
RA, IM
tweqi
twi 4,RA,IM
Trap if (RA) greater than or equal to EXTS(IM).
Extended mnemonic for
twi 12,RA,IM
RA, IM
RA, IM
RA, IM
RA, IM
RA, IM
RA, IM
RA, IM
RA, IM
RA, IM
RA, IM
RA, IM
RA, IM
twgei
twgti
twlei
Trap if (RA) greater than EXTS(IM).
Extended mnemonic for
twi 8,RA,IM
Trap if (RA) less than or equal to EXTS(IM).
Extended mnemonic for
twi 20,RA,IM
Trap if (RA) logically greater than or equal to EXTS(IM).
Extended mnemonic for
twi 5,RA,IM
twlgei
twlgti
twllei
twllti
twlngi
twlnli
twlti
Trap if (RA) logically greater than EXTS(IM).
Extended mnemonic for
twi 1,RA,IM
Trap if (RA) logically less than or equal to EXTS(IM).
Extended mnemonic for
twi 6,RA,IM
Trap if (RA) logically less than EXTS(IM).
Extended mnemonic for
twi 2,RA,IM
Trap if (RA) logically not greater than EXTS(IM).
Extended mnemonic for
twi 6,RA,IM
Trap if (RA) logically not less than EXTS(IM).
Extended mnemonic for
twi 5,RA,IM
Trap if (RA) less than EXTS(IM).
Extended mnemonic for
twi 16,RA,IM
Trap if (RA) not equal to EXTS(IM).
Extended mnemonic for
twi 24,RA,IM
twnei
twngi
Trap if (RA) not greater than EXTS(IM).
Extended mnemonic for
twi 20,RA,IM
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
twi
Trap Word Immediate
Preliminary
PPC440x5 CPU Core User’s Manual
Table 9-32. Extended Mnemonics for twi (continued)
Other Registers
Altered
Mnemonic
Operands
Function
Trap if (RA) not less than EXTS(IM).
Extended mnemonic for
RA, IM
twnli
twi 12,RA,IM
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
wrtee
Write External Enable
PPC440x5 CPU Core User’s Manual
Preliminary
wrtee
Write External Enable
wrtee
RS
31
RS
131
0
6
11
21
31
MSR[EE] ← (RS)
16
MSR[EE] is set to the value specified by bit 16 of register RS.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• MSR[EE]
Invalid Instruction Forms:
• Reserved fields
Programming Notes
Execution of this instruction is privileged.
This instruction is typically used as part of a code sequence which can provide the equivalent of an atomic
read-modify-write of the MSR, as follows:
mfmsr Rn #save EE in Rn[16]
wrteei 0
#Turn off EE (leaving other bits unchanged)
#Code with EE disabled
•
•
•
wrtee Rn
#restore EE without affecting any MSR changes that occurred in the disabled code
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
wrteei
Write External Enable Immediate
Preliminary
PPC440x5 CPU Core User’s Manual
wrteei
Write External Enable Immediate
wrteei
E
31
E
163
0
6
16 17
21
31
MSR[EE] ← E
MSR[EE] is set to the value specified by the E field.
If instruction bit 31 contains 1, the contents of CR[CR0] are undefined.
Registers Altered
• MSR[EE]
Invalid Instruction Forms:
• Reserved fields
Programming Notes
Execution of this instruction is privileged.
This instruction is typically used as part of a code sequence which can provide the equivalent of an atomic
read-modify-write of the MSR, as follows:
mfmsr Rn #save EE in Rn[16]
wrteei 0
#Turn off EE (leaving other bits unchanged)
#Code with EE disabled
•
•
•
wrtee Rn
#restore EE without affecting any MSR changes that occurred in the disabled code
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
xor
XOR
PPC440x5 CPU Core User’s Manual
Preliminary
xor
XOR
xor
xor.
RA, RS, RB
RA, RS, RB
Rc=0
Rc=1
31
RS
RA
RB
316
Rc
0
6
11
16
21
31
(RA) ← (RS) ⊕ (RB)
The contents of register RS are XORed with the contents of register RB; the result is placed into register RA.
Registers Altered
• RA
• CR[CR0] if Rc contains 1
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
xori
XOR Immediate
Preliminary
PPC440x5 CPU Core User’s Manual
xori
XOR Immediate
xori
RA, RS, IM
26
RS
RA
IM
0
6
11
16
31
16
(RA) ← (RS) ⊕ ( 0 || IM)
The IM field is extended to 32 bits by concatenating 16 0-bits on the left. The contents of register RS are
XORed with the extended IM field; the result is placed into register RA.
Registers Altered
• RA
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
xoris
XOR Immediate Shifted
PPC440x5 CPU Core User’s Manual
Preliminary
xoris
XOR Immediate Shifted
xoris
RA, RS, IM
27
RS
RA
IM
0
6
11
16
31
16
(RA) ← (RS) ⊕ (IM || 0)
The IM field is extended to 32 bits by concatenating 16 0-bits on the right. The contents of register RS are
XORed with the extended IM field; the result is placed into register RA.
Registers Altered
• RA
instrset.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
10. Register Summary
This chapter provides an alphabetical listing of and bit definitions for the registers contained in the PPC440x5
core.
The registers, of five types, are grouped into several functional categories according to the processor func-
tions with which they are associated. More information about the registers and register categories is provided
in Section 2.2 Registers on page 47, and in the chapters describing the processor functions with which each
register category is associated.
10.1 Register Categories
Table 10-1 summarizes the register categories and the registers contained in each category. Italicized
register names are implementation-specific. All other registers are defined by the Book-E Enhanced
PowerPC Architecture.
regsummIntro.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table 10-1. Register Categories
Register Category
Register(s)
Model and Access
Type Page
CR
User
User
User
CR
67
67
CTR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
DCR
GPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
MSR
SPR
SPR
SPR
SPR
SPR
SPR
SPR
Branch Control
LR
66
DNV0–DNV3
DTV0–DTV3
DVLIM
Supervisor
97
Supervisor
97
Supervisor
99
Cache Control
Cache Debug
Debug
INV0–INV3
ITV0–ITV3
IVLIM
Supervisor
97
Supervisor
97
Supervisor
99
DCDBTRH, DCDBTRL
ICDBDR, ICDBTRH, ICDBTRL
DAC1–DAC2
DBCR0–DBCR2
DBDR
Supervisor, read-only
Supervisor, read-only
Supervisor
127
112
246
239
247
244
246
245
53
Supervisor
Supervisor
DBSR
Supervisor
DVC1–DVC2
IAC1–IAC4
Implemented outside core
GPR0–GPR31
XER
Supervisor
Supervisor
Supervisor
Device Control
User
71
Integer Processing
User
72
CSRR0–CSRR1
DEAR
Supervisor
168
170
172
170
171
174
169
167
108
108
165
75
Supervisor
ESR
Supervisor
IVOR0–IVOR15
IVPR
Supervisor
Interrupt Processing
Supervisor
MCSR
Supervisor
MCSRR0-MCSRR1
SRR0–SRR1
CCR0
Supervisor
Supervisor
Supervisor
CCR1
Supervisor
MSR
Supervisor
PIR, PVR
Supervisor, read-only
Supervisor, read-only
Supervisor
Processor Control
Storage Control
RSTCFG
79
SPRG0–SPRG3
SPRG4–SPRG7
USPRG0
75
User, read-only; Supervisor
User
75
75
MMUCR
Supervisor
148
151
PID
Supervisor
regsummIntro.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 10-1. Register Categories
Register Category
Register(s)
Model and Access
Supervisor
Type Page
DEC
SPR
SPR
SPR
SPR
SPR
211
211
209
215
216
DECAR
TBL, TBU
TCR
Supervisor, write-only
User read, Supervisor write
Supervisor
Timer
TSR
Supervisor
regsummIntro.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table 10-2 Special Purpose Registers Sorted by SPR Number on page 454, lists the Special Purpose Regis-
ters (SPRs) in order by SPR number (SPRN). The table provides mnemonics, names, SPRN, model (user or
supervisor), and access. All SPR numbers not listed are reserved, and should be neither read nor written.
Note that three registers, DBSR, MCSR, and TSR, are indicated as having the access type of read/clear.
These three registers are status registers, and as such behave differently than other SPRs when written. The
term “read/clear” does not mean that these registers are automatically cleared upon being read. Rather, the
“clear” refers to their behavior when being written. Instead of simply overwriting the SPR with the data in the
source GPR, the status SPR is updated by zeroing those bit positions corresponding to 1 values in the source
GPR, with those bit positions corresponding to 0 values in the source GPR being left unchanged. In this
fashion, it is possible for software to read one of these status SPRs, and then write to it using the same data
which was read. Any bits which were read as 1 will then be cleared, and any bits which were not yet set at the
time the SPR was read will be left unchanged. If any of these previously clear bits happen to be set between
the time the SPR is read and when it is written, then when the SPR is later read again, software will observe
any newly set bit[s]. If it were not for this behavior, then software could erroneously clear bits which it had not
yet observed as having been set, and overlook the occurrence of certain exceptions.
Table 10-2. Special Purpose Registers Sorted by SPR Number
Mnemonic
XER
Register Name
Integer Exception Register
SPRN
0x001
Model
Access
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Write-only
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read-only
Read-only
Read-only
Read-only
Read-only
Read-only
Read/Write
Read/Write
Read/Write
User
User
User
LR
Link Register
0x008
0x009
0x016
0x01A
0x01B
0x030
0x036
0x03A
0x03B
0x03D
0x03E
0x03F
0x100
0x104
0x105
0x106
0x107
0x10C
0x10D
0x110
0x111
0x112
CTR
Count Register
DEC
Decrementer
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
User
SRR0
SRR1
PID
Save/Restore Register 0
Save/Restore Register 1
Process ID
DECAR
CSRR0
CSRR1
DEAR
ESR
Decrementer Auto-Reload
Critical Save/Restore Register 0
Critical Save/Restore Register 1
Data Exception Address Register
Exception Syndrome Register
Interrupt Vector Prefix Register
User Special Purpose Register General 0
Special Purpose Register General 4
Special Purpose Register General 5
Special Purpose Register General 6
Special Purpose Register General 7
Time Base Lower
IVPR
USPRG0
SPRG4
SPRG5
SPRG6
SPRG7
TBL
User
User
User
User
User
TBU
Time Base Upper
User
SPRG0
SPRG1
SPRG2
Special Purpose Register General 0
Special Purpose Register General 1
Special Purpose Register General 2
Supervisor
Supervisor
Supervisor
regsummIntro.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table 10-2. Special Purpose Registers Sorted by SPR Number
Mnemonic
SPRG3
Register Name
Special Purpose Register General 3
Special Purpose Register General 4
Special Purpose Register General 5
Special Purpose Register General 6
Special Purpose Register General 7
Time Base Lower
SPRN
0x113
Model
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Access
Read/Write
Write-only
Write-only
Write-only
Write-only
Write-only
Write-only
Read-only
Read-only
Read/Clear
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Clear
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
SPRG4
SPRG5
SPRG6
SPRG7
TBL
0x114
0x115
0x116
0x117
0x11C
0x11D
0x11E
0x11F
0x130
0x134
0x135
0x136
0x138
0x139
0x13A
0x13B
0x13C
0x13D
0x13E
0x13F
0x150
0x154
0x190
0x191
0x192
0x193
0x194
0x195
0x196
0x197
0x198
0x199
0x19A
0x19B
0x19C
0x19D
TBU
Time Base Upper
PIR
Processor ID Register
PVR
Processor Version Register
Debug Status Register
DBSR
DBCR0
DBCR1
DBCR2
IAC1
Debug Control Register 0
Debug Control Register 1
Debug Control Register 2
Instruction Address Compare 1
Instruction Address Compare 2
Instruction Address Compare 3
Instruction Address Compare 4
Data Address Compare 1
IAC2
IAC3
IAC4
DAC1
DAC2
DVC1
DVC2
TSR
Data Address Compare 2
Data Value Compare 1
Data Value Compare 2
Timer Status Register
TCR
Timer Control Register
IVOR0
IVOR1
IVOR2
IVOR3
IVOR4
IVOR5
IVOR6
IVOR7
IVOR8
IVOR9
IVOR10
IVOR11
IVOR12
IVOR13
Interrupt Vector Offset Register 0
Interrupt Vector Offset Register 1
Interrupt Vector Offset Register 2
Interrupt Vector Offset Register 3
Interrupt Vector Offset Register 4
Interrupt Vector Offset Register 5
Interrupt Vector Offset Register 6
Interrupt Vector Offset Register 7
Interrupt Vector Offset Register 8
Interrupt Vector Offset Register 9
Interrupt Vector Offset Register 10
Interrupt Vector Offset Register 11
Interrupt Vector Offset Register 12
Interrupt Vector Offset Register 13
regsummIntro.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table 10-2. Special Purpose Registers Sorted by SPR Number
Mnemonic
IVOR14
Register Name
Interrupt Vector Offset Register 14
Interrupt Vector Offset Register 15
Machine Check Save Restore Register 0
Machine Check Save Restore Register 1
Machine Check Status Register
Instruction Cache Normal Victim 0
Instruction Cache Normal Victim 1
Instruction Cache Normal Victim 2
Instruction Cache Normal Victim 3
Instruction Cache Transient Victim 0
Instruction Cache Transient Victim 1
Instruction Cache Transient Victim 2
Instruction Cache Transient Victim 3
Core Configuration Register 1
SPRN
0x19E
Model
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Supervisor
Access
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read/Write
Read-only
Read-only
Read-only
Read-only
Read-only
Read/Write
Read/Write
Read-only
Read/Write
IVOR15
MCSRR0
MCSRR1
MCSR
INV0
0x19F
0x23A
0x23B
0x23C
0x370
0x371
0x372
0x373
0x374
0x375
0x376
0x377
0x378
0x390
0x391
0x392
0x393
0x394
0x395
0x396
0x397
0x398
0x399
0x39B
0x39C
0x39D
0x39E
0x39F
0x3B2
0x3B3
0x3D3
0x3F3
INV1
INV2
INV3
ITV0
ITV1
ITV2
ITV3
CCR1
DNV0
Data Cache Normal Victim 0
DNV1
Data Cache Normal Victim 1
DNV2
Data Cache Normal Victim 2
DNV3
Data Cache Normal Victim 3
DTV0
Data Cache Transient Victim 0
Data Cache Transient Victim 1
Data Cache Transient Victim 2
Data Cache Transient Victim 3
Data Cache Victim Limit
DTV1
DTV2
DTV3
DVLIM
IVLIM
Instruction Cache Victim Limit
RSTCFG
DCDBTRL
DCDBTRH
ICDBTRL
ICDBTRH
MMUCR
CCR0
Reset Configuration
Data Cache Debug Tag Register Low
Data Cache Debug Tag Register High
Instruction Cache Debug Tag Register Low
Instruction Cache Debug Tag Register High
Memory Management Unit Control Register
Core Configuration Register 0
ICDBDR
DBDR
Instruction Cache Debug Data Register
Debug Data Register
regsummIntro.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
10.2 Reserved Fields
For all registers with fields marked as reserved, the reserved fields should be written as zero and read as
undefined. That is, when writing to a reserved field, write a zero to that field. When reading from a reserved
field, ignore that field.
The recommended coding practice is to perform the initial write to a register with reserved fields as described
in the preceding paragraph, and to perform all subsequent writes to the register using a read-modify-write
strategy: read the register, alter desired fields with logical instructions, and then write the register.
10.3 Device Control Registers
Device Control Registers (DCRs), which are architecturally outside of the processor core, are accessed using
the mfdcr and mtdcr instructions. DCRs are used to control, configure, and hold status for various functional
units that are not part of the RISC processor core. Although the PPC440x5 core does not contain DCRs, the
mfdcr and mtdcr instructions are provided.
The mfdcr and mtdcr instructions are privileged, for all DCR numbers. Therefore, all DCR accesses are
privileged. All DCR numbers are reserved, and should be neither read nor written, unless they are part of a
Core+ASIC implementation.
regsummIntro.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
regsummIntro.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
Preliminary
PPC440x5 CPU Core User’s Manual
0.Register Summary
10.4 Alphabetical Register Listing
The following pages list the registers available in the PPC440x5 core. For each register, the following infor-
mation is supplied:
• Register mnemonic and name
• Cross reference to detailed register information
• Register type (if SPR); the types of the other registers are the same as the register names (CR, GPR,
MSR)
• Register number (address)
• Register programming model (user or supervisor) and access (read-clear, read-only, read/write (R/W),
write-only)
• A diagram illustrating the register fields (all register fields have mnemonics, unless there is only one field)
• A table describing the register fields, giving field mnemonics, field bit locations, field names, and the func-
tions associated with the various field values
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
CCR0
Core Configuration Register 0
PPC440x5 CPU Core User’s Manual
Preliminary
CCR0
SPR 0x3B3 Supervisor R/W
PRE
DSTG
DTB GDCBT
ICSLC
0
1
2
3
4
5
9
10 11 12
DAPUIB
15 16 17 18 19
GICBT
22 23 24
FLSTA
27 28 29 30 31
ICSLT
CRPE
Figure 10-1. Core Configuration Register 0 (CCR0)
0
Reserved
Parity Recoverability Enable
0 Semi-recoverable parity mode enabled for data
cache
1 Fully recoverable parity mode enabled for data
cache
Must be set to 1 to guarantee full recoverability
from MMU and data cache parity errors.
1
PRE
2:3
4
Reserved
Cache Read Parity Enable
0 Disable parity information reads
1 Enable parity information reads
When enabled, execution of icread, dcread, or
tlbre loads parity information into the ICDBTRH,
DCDBTRL, or target GPR, respectively.
CRPE
5:9
Reserved
Disable Store Gathering
0 Enabled; stores to contiguous addresses may be
gathered into a single transfer
10
DSTG
See Store Gathering on page 119.
1 Disabled; all stores to memory will be performed
independently
Disable APU Instruction Broadcast
0 Enabled.
1 Disabled; instructions not broadcast to APU for
decoding
This mechanism is provided as a means of reduc-
ing power consumption when an auxilliary pro-
cessor is not attached and/or is not being used.
11
DAPUIB
See Initialization on page 85.
12:15
16
Reserved
This mechanism is provided as a means of reduc-
ing power consumption when instruction tracing is
not needed.
Disable Trace Broadcast
0 Enabled.
1 Disabled; no trace information is broadcast.
DTB
See Initialization on page 85.
Guaranteed Instruction Cache Block Touch
0 icbt may be abandoned without having filled
cache line if instruction pipeline stalls.
1 icbt is guaranteed to fill cache line even if
instruction pipeline stalls.
17
GICBT
See icbt Operation on page 111.
Guaranteed Data Cache Block Touch
0 dcbt/dcbtst may be abandoned without
having filled cache line if load/store pipeline
stalls.
See Data Cache Control and Debug on
page 125.
18
GDCBT
1 dcbt/dcbtst are guaranteed to fill cache line
even if load/store pipeline stalls.
19:22
Reserved
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
CCR0 (cont.)
Core Configuration Register 0
Preliminary
PPC440x5 CPU Core User’s Manual
Force Load/Store Alignment
0 No Alignment exception on integer storage
access instructions, regardless of alignment
1 An alignment exception occurs on integer
storage access instructions if data address is not
on an operand boundary.
23
FLSTA
See Load and Store Alignment on page 117.
24:27
28:29
Reserved
Number of additional lines (0–3) to fill on instruc-
tion fetch miss.
ICSLC
ICSLT
Instruction Cache Speculative Line Count
See Speculative Prefetch Mechanism on
page 105.
Number of doublewords that must have already
been filled in order that the current speculative
line fill is not abandoned on a redirection of the
instruction stream.
30:31
Instruction Cache Speculative Line Threshold
See Speculative Prefetch Mechanism on
page 105.
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
CCR1
Core Configuration Register 1
PPC440x5 CPU Core User’s Manual
Preliminary
CCR1
SPR 0x378 Supervisor R/W
DCTPEI
FCOM
DCUPEI
ICDPEI
FFF
0
7
8
9
10 11 12 13 14 15 16
19 20 21
23 24 25
TCS
31
ICTPEI DCDPEI DCMPEI
MMUPEI
Figure 10-2. Core Configuration Register 1 (CCR1)
Instruction Cache Data Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bits recorded when the
instruction cache is filled. Each of the 8 bits corre-
sponds to one of the instruction words in the line.
0:7
8:9
10:11
12
ICDPEI
ICTPEI
Instruction Cache Tag Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bits recorded for the tag
field in the instruction cache.
Data Cache Tag Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bits recorded for the tag
field in the data cache.
DCTPEI
DCDPEI
DCUPEI
DCMPEI
Controls inversion of parity bits recorded for the
data field in the data cache.
Data Cache Data Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bit recorded for the U
fields in the data cache.
Data Cache U-bit Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
13
Data Cache Modified-bit Parity Error Insert
0 record even parity (normal)
1 record odd parity (simulate parity error)
Controls inversion of parity bits recorded for the
modified (dirty) field in the data cache.
14
Force icbt , dcbt, dcbtst, dcbst, dcbf, dcbi, and
dcbz to appear to miss the caches. The intended
use is with icbt and dcbt only, which will fill a dupli-
cate line and allow testing of multi-hit parity errors.
See Section 4.2.4.7 Simulating Instruction Cache
Parity Errors for Software Testing on page 114 and
Figure 4.3.3.7 on page 130.
Force Cache Operation Miss
0 normal operation
1 cache ops appear to miss the cache
15
FCOM
Controls inversion of parity bits recorded for the tag
field in the MMU.
Memory Management Unit Parity Error Insert
0 record even parity (normal)
16:19
MMUPEI
1 record odd parity (simulate parity error)
When flushing 32-byte (8-word) lines from the data
cache, normal operation is to write nothing, a dou-
ble word, quad word, or the entire 8-word block to
the memory as required by the dirty bits. This bit
ensures that none or all dirty bits are set so that
either nothing or the entire 8-word block is written
to memory when flushing a line from the data
cache. Refer to Section 4.3.1.4 Line Flush Opera-
tions on page 121.
Force Full-line Flush
0 flush only as much data as necessary.
1 always flush entire cache lines
20
FFF
21:23
Reserved
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
CCR1 (cont.)
Core Configuration Register 1
Preliminary
PPC440x5 CPU Core User’s Manual
Timer Clock Select
0 CPU timer advances by one at each rising edge
of the CPU input clock (CPMC440CLOCK).
1 CPU timer advances by one for each rising edge
of the CPU timer clock
When TCS = 1, CPU timer clock input can toggle
at up to half of the CPU clock frequency.
24
TCS
(CPMC440TIMERCLOCK).
25:31
Reserved
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
CR
Condition Register
PPC440x5 CPU Core User’s Manual
Preliminary
CR
User Read/Write
CR2
CR6
CR0
CR4
0
3
4
7
8
11 12
15 16
19 20
23 24
27 28
31
CR5
CR1
CR3
CR7
Figure 10-3. Condition Register (CR)
0:3
CR0
CR1
CR2
CR3
CR4
CR5
CR6
CR7
Condition Register Field 0
Condition Register Field 1
Condition Register Field 2
Condition Register Field 3
Condition Register Field 4
Condition Register Field 5
Condition Register Field 6
Condition Register Field 7
4:7
8:11
12:15
16:19
20:23
24:27
28:31
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
CSRR0
Critical Save/Restore Register 0
Preliminary
PPC440x5 CPU Core User’s Manual
CSRR0
SPR 0x03A Supervisor R/W
0
29 30 31
Figure 10-4. Critical Save/Restore Register 0 (CSRR0)
0:29
Return address for critical interrupts
Reserved
30:31
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
CSRR1
Critical Save/Restore Register 1
PPC440x5 CPU Core User’s Manual
Preliminary
CSRR1
SPR 0x03B Supervisor R/W
EE
FP
WE
FE0
DE
IS
0
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
FE1
31
CE
ME
DWE
DS
PR
Figure 10-5. Critical Save/Restore Register 1 (CSRR1)
0:31
Copy of the MSR when a critical interrupt is taken
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
CTR
Count Register
Preliminary
PPC440x5 CPU Core User’s Manual
CTR
SPR 0x009 User R/W
0
31
Figure 10-6. Count Register (CTR)
Used as count for branch conditional with decre-
ment instructions, or as target address for bcctr
instructions
0:31
Count
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DAC1–DAC2
Data Address Compare Registers
PPC440x5 CPU Core User’s Manual
Preliminary
DAC1–DAC2
SPR 0x13C–0x13D Supervisor R/W
0
31
Figure 10-7. Data Address Compare Registers (DAC1–DAC2)
0:31
Data Address Compare (DAC) byte address
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DBCR0
Debug Control Register 0
Preliminary
PPC440x5 CPU Core User’s Manual
DBCR0
SPR 0x134 Supervisor R/W
EDM
0
RST
BRT TRAP IAC2 IAC4 DAC1WDAC2W
10 11 12 13 14 15 16 17
ICMP IRPT IAC1 IAC3 DAC1R DAC2R RET
FT
1
2
3
4
5
6
7
8
9
30 31
IDM
Figure 10-8. Debug Control Register 0 (DBCR0)
External Debug Mode
0 Disable external debug mode.
1 Enable external debug mode.
0
1
EDM
IDM
Internal Debug Mode
0 Disable internal debug mode.
1 Enable internal debug mode.
Reset
00 No action
01 Core reset
10 Chip reset
2:3
RST
11 System reset
Attention: Writing 01, 10, or 11 to this field causes a processor reset to occur.
Instruction completions do not cause
Instruction Completion Debug Event
0 Disable instruction completion debug event.
1 Enable instruction completion debug event.
instruction completion debug events if
MSR[DE] = 0 in internal debug mode,
unless also in external debug mode or
debug wait mode.
4
ICMP
Taken branches do not cause branch
taken debug events if MSR[DE] = 0 in
internal debug mode, unless also in
external debug mode or debug wait
mode.
Branch Taken Debug Event
0 Disable branch taken debug event.
1 Enable branch taken debug event.
5
6
BRT
Critical interrupts do not cause interrupt
debug events in internal debug mode,
unless also in external debug mode or
debug wait mode.
Interrupt Debug Event
0 Disable interrupt debug event.
1 Enable interrupt debug event.
IRPT
Trap Debug Event
0 Disable trap debug event.
1 Enable trap debug event.
7
TRAP
IAC1
IAC2
IAC3
IAC4
Instruction Address Compare (IAC) 1 Debug Event
0 Disable IAC 1 debug event.
1 Enable IAC 1 debug event.
8
IAC 2 Debug Event
0 Disable IAC 2 debug event.
1 Enable IAC 2 debug event.
9
IAC 3 Debug Event
0 Disable IAC 3 debug event.
1 Enable IAC 3 debug event.
10
11
IAC 4 Debug Event
0 Disable IAC 4 debug event.
1 Enable IAC 4 debug event.
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DBCR0 (cont.)
Debug Control Register 0
PPC440x5 CPU Core User’s Manual
Preliminary
Data Address Compare (DAC) 1 Read Debug Event
0 Disable DAC 1 read debug event.
1 Enable DAC 1 read debug event.
12
13
14
15
DAC1R
DAC1W
DAC2R
DAC2W
DAC 1 Write Debug Event
0 Disable DAC 1 write debug event.
1 Enable DAC 1 write debug event.
DAC 2 Read Debug Event
0 Disable DAC 2 read debug event.
1 Enable DAC 2 read debug event.
DAC 2 Write Debug Event
0 Disable DAC 2 write debug event.
1 Enable DAC 2 write debug event.
rfci/rfmci does not cause a return
Return Debug Event
0 Disable return (rfi/rfci/rfmci) debug event.
1 Enable return (rfi/rfci/rfmci) debug event.
debug event if MSR[DE] = 0 in internal
debug mode, unless also in external
debug mode or debug wait mode.
16
RET
FT
17:30
31
Reserved
Freeze timers on debug event
0 Timers are not frozen.
1 Freeze timers if a DBSR field associated with a debug event
is set.
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DBCR1
Debug Control Register 1
Preliminary
PPC440x5 CPU Core User’s Manual
DBCR1
SPR 0x135 Supervisor R/W
IAC4US
IAC1US
IAC2US
IAC12M
IAC3US
IAC34M
0
1
2
3
4
5
6
7
8
9
10
14 15 16 17 18 19 20 21 22 23 24 25 26
IAC3ER
30 31
IAC12AT
IAC1ER
IAC4ER
IAC2ER
IAC34AT
Figure 10-9. Debug Control Register 1 (DBCR1)
Instruction Address Compare (IAC) 1 User/Super-
visor
00 Both
01 Reserved
0:1
IAC1US
10 Supervisor only (MSR[PR] = 0)
11 User only (MSR[PR] = 1)
IAC 1 Effective/Real
00 Effective (MSR[IS] = don’t care)
01 Reserved
10 Virtual (MSR[IS] = 0)
11 Virtual (MSR[IS] = 1)
2:3
4:5
6:7
IAC1ER
IAC2US
IAC2ER
IAC 2 User/Supervisor
00 Both
01 Reserved
10 Supervisor only (MSR[PR] = 0)
11 User only (MSR[PR] = 1)
IAC 2 Effective/Real
00 Effective (MSR[IS] = don’t care)
01 Reserved
10 Virtual (MSR[IS] = 0)
11 Virtual (MSR[IS] = 1)
IAC 1/2 Mode
00 Exact match
Match if address[0:29] = IAC 1/2[0:29]; two inde-
pendent compares
8:9
IAC12M
01 Reserved
10 Range inclusive
11 Range exclusive
Match if IAC1 ≤ address < IAC2
Match if address < IAC1 OR address ≥ IAC2
10:14
15
Reserved
IAC 1/2 Auto-Toggle Enable
0 Disable IAC 1/2 auto-toggle
1 Enable IAC 1/2 auto-toggle
IAC12AT
IAC3US
IAC 3 User/Supervisor
00 Both
01 Reserved
10 Supervisor only (MSR[PR] = 0)
11 User only (MSR[PR] = 1)
16:17
18:19
IAC 3 Effective/Real
00 Effective (MSR[IS] = don’t care)
01 Reserved
IAC3ER
10 Virtual (MSR[IS] = 0)
11 Virtual (MSR[IS] = 1)
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DBCR1 (cont.)
Debug Control Register 1
PPC440x5 CPU Core User’s Manual
Preliminary
IAC 4 User/Supervisor
00 Both
01 Reserved
10 Supervisor only (MSR[PR] = 0)
11 User only (MSR[PR] = 1)
20:21
22:23
IAC4US
IAC4ER
IAC 4 Effective/Real
00 Effective (MSR[IS] = don’t care)
01 Reserved
10 Virtual (MSR[IS] = 0)
11 Virtual (MSR[IS] = 1)
IAC 3/4 Mode
00 Exact match
Match if address[0:29] = IAC 3/4[0:29]; two inde-
pendent compares
24:25
IAC34M
01 Reserved
10 Range inclusive
11 Range exclusive
Match if IAC3 ≤ address < IAC4
Match if address < IAC3 OR address ≥ IAC4
26:30
31
Reserved
IAC3/4 Auto-Toggle Enable
0 Disable IAC 3/4 auto-toggle
1 Enable IAC 3/4 auto-toggle
IAC34AT
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DBCR2
Debug Control Register 2
Preliminary
PPC440x5 CPU Core User’s Manual
DBCR2
SPR 0x136 Supervisor R/W
DAC2US
DAC12M
DVC1BE
DAC1US
DVC1M
0
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16
19 20
23 24
27 28
31
DAC1ER
DAC2ER DAC12A
DVC2M
DVC2BE
Figure 10-10. Debug Control Register 2 (DBCR2)
Data Address Compare (DAC) 1 User/Supervisor
00 Both
01 Reserved
10 Supervisor only (MSR[PR] = 0)
11 User only (MSR[PR] = 1)
0:1
2:3
4:5
6:7
DAC1US
DAC1ER
DAC2US
DAC2ER
DAC 1 Effective/Real
00 Effective (MSR[DS] = don’t care)
01 Reserved
10 Virtual (MSR[DS] = 0)
11 Virtual (MSR[DS] = 1)
DAC 2 User/Supervisor
00 Both
01 Reserved
10 Supervisor only (MSR[PR] = 0)
11 User only (MSR[PR] = 1)
DAC 2 Effective/Real
00 Effective (MSR[DS] = don’t care)
01 Reserved
10 Virtual (MSR[DS] = 0)
11 Virtual (MSR[DS] = 1)
DAC 1/2 Mode
00 Exact match
Match if address[0:31] = DAC 1/2[0:31]; two inde-
pendent compares
01 Address bit mask
8:9
DAC12M
DAC12A
Match if address = DAC1; only compare bits cor-
responding to 1 bits in DAC2
Match if DAC1 ≤ address < DAC2
Match if address < DAC1 OR address ≥ DAC2
10 Range inclusive
11 Range exclusive
DAC 1/2 Asynchronous
0 Debug interrupt caused by DAC1/2 exception
will be synchronous
1 Debug interrupt caused by DAC1/2 exception
will be asynchronous
10
11
Reserved
Data Value Compare (DVC) 1 Mode
00 Reserved
01 AND all bytes enabled by DVC1BE
10 OR all bytes enabled by DVC1BE
12:13
DVC1M
11 AND-OR pairs of bytes enabled by DVC1BE
(0 AND 1) OR (2 AND 3)
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DBCR2 (cont.)
Debug Control Register 2
PPC440x5 CPU Core User’s Manual
Preliminary
DVC 2 Mode
00 Reserved
01 AND all bytes enabled by DVC2BE
10 OR all bytes enabled by DVC2BE
14:15
DVC2M
11 AND-OR pairs of bytes enabled by DVC2BE
(0 AND 1) OR (2 AND 3)
16:19
20:23
24:27
28:31
Reserved
DVC1BE
DVC2BE
DVC 1 Byte Enables 0:3
Reserved
DVC 2 Byte Enables 0:3
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DBSR
Debug Status Register
PPC440x5 CPU Core User’s Manual
Preliminary
DBSR
SPR 0x130 Supervisor Read/Clear
IDE
0
MRR BRT TRAP IAC2 IAC4
DAC2W
DAC1W
IAC34ATS
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17
29 30 31
UDE
ICMP IRPT IAC1 IAC3 DAC1R DAC2R
IAC12ATS
RET
Figure 10-12. Debug Status Register (DBSR)
For synchronous debug events in internal debug
mode, this field indicates whether the correspond-
ing Debug interrupt occurs precisely or impre-
cisely
Imprecise Debug Event
0 Debug event ocurred while MSR[DE] = 1
1 Debug event occurred while MSR[DE] = 0
0
1
IDE
Unconditional Debug Event
0 Event didn’t occur
1 Event occurred
UDE
Most Recent Reset
00 No reset has occurred since this field was last
cleared by software.
01 Core reset
10 Chip reset
This field is set upon any processor reset to a
value indicating the type of reset.
2:3
MRR
11 System reset
Instruction Completion Debug Event
0 Event didn’t occur
1 Event occurred
4
ICMP
BRT
Branch Taken Debug Event
0 Event didn’t occur
1 Event occurred
5
Interrupt Debug Event
0 Event didn’t occur
1 Event occurred
6
IRPT
TRAP
IAC1
IAC2
IAC3
IAC4
DAC1R
Trap Debug Event
0 Event didn’t occur
1 Event occurred
7
IAC 1 Debug Event
0 Event didn’t occur
1 Event occurred
8
IAC 2 Debug Event
0 Event didn’t occur
1 Event occurred
9
IAC 3 Debug Event
0 Event didn’t occur
1 Event occurred
10
11
12
IAC 4 Debug Event
0 Event didn’t occur
1 Event occurred
DAC 1 Read Debug Event
0 Event didn’t occur
1 Event occurred
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DBSR (cont.)
Debug Status Register
Preliminary
PPC440x5 CPU Core User’s Manual
DAC 1 Write Debug Event
0 Event didn’t occur
1 Event occurred
13
14
15
DAC1W
DAC2R
DAC2W
RET
DAC 2 Read Debug Event
0 Event didn’t occur
1 Event occurred
DAC 2 Write Debug Event
0 Event didn’t occur
1 Event occurred
Return Debug Event
0 Event didn’t occur
1 Event occurred
16
17:29
Reserved
IAC 1/2 Auto-Toggle Status
0 Range is not reversed from value specified in
DBCR1[IAC12M]
1 Range is reversed from value specified in
DBCR1[IAC12M]
30
31
IAC12ATS
IAC34ATS
IAC 3/4 Auto-Toggle Status
0 Range is not reversed from value specified in
DBCR1[IAC34M]
1 Range is reversed from value specified in
DBCR1[IAC34M]
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DCDBTRH
Data Cache Debug Tag Register High
PPC440x5 CPU Core User’s Manual
Preliminary
DCDBTRH
SPR 0x39D Supervisor Read-Only
TERA
TRA
0
23 24 25
V
27 28
31
Figure 10-13. Data Cache Debug Tag Register High (DCDBTRH)
Bits 0:23 of the lower 32 bits of the 36-bit real
address associated with the cache line read by
dcread.
0:23
24
TRA
V
Tag Real Address
Cache Line Valid
0 Cache line is not valid.
1 Cache line is valid.
The valid indicator for the cache line read by
dcread.
25:27
28:31
Reserved
Upper 4 bits of the 36-bit real address associated
with the cache line read by dcread.
TERA
Tag Extended Real Address
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DCDBTRL
Data Cache Debug Tag Register Low
PPC440x5 CPU Core User’s Manual
Preliminary
DCDBTRL
SPR 0x39C Supervisor Read-Only
DPAR
UPAR
D
U1
27 28 29 30 31
U0
U3
0
12 13 14 15 16
TPAR
19 20
23 24
U2
MPAR
Figure 10-14. Data Cache Debug Tag Register Low (DCDBTRL)
0:12
Reserved
The parity for the U0-U3 bits in the cache line read
by dcread if CCR0[CRPE] = 1, otherwise 0.
13
UPAR
TPAR
U bit parity
The parity for the tag bits in the cache line read by
dcread if CCR0[CRPE] = 1, otherwise 0.
14:15
16:19
Tag parity
Data parity
The parity check values for the data bytes in the
word read by dcread if CCR0[CRPE] = 1, other-
wise 0.
DPAR
MPAR
The parity for the modified (dirty) indicators for
each of the four doublewords in the cache line read
by dcread if CCR0[CRPE] = 1, otherwise 0.
20:23
Modified (dirty) parity
The “dirty” (modified) indicators for each of the four
doublewords in the cache line read by dcread.
24:27
28
D
Dirty Indicators
The U0 storage attribute for the memory page
associated with this cache line read by dcread.
U0
U1
U2
U3
U0 Storage Attribute
U1 Storage Attribute
U2 Storage Attribute
U3 Storage Attribute
The U1 storage attribute for the memory page
associated with this cache line read by dcread.
29
The U2 storage attribute for the memory page
associated with this cache line read by dcread.
30
The U3 storage attribute for the memory page
associated with this cache line read by dcread.
31
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DEAR
Data Exception Address Register
PPC440x5 CPU Core User’s Manual
Preliminary
DEAR
SPR 0x03D Supervisor R/W
0
31
Figure 10-15. Data Exception Address Register (DEAR)
Address of data exception for Data Storage, Align-
ment, and Data TLB Error interrupts
0:31
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DECAR
Decrementer Auto-Reload
PPC440x5 CPU Core User’s Manual
Preliminary
DECAR
SPR 0x036 Supervisor Write-Only
0
31
Figure 10-17. Decrementer Auto-Reload (DECAR)
Copied to DEC at next time base clock when
0:31
Decrementer auto-reload value
DEC = 1 and auto-reload is enabled
(TCR[ARE] = 1).
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DNV0–DNV3
Data Cache Normal Victim 0–3
Preliminary
PPC440x5 CPU Core User’s Manual
DNV0–DNV3
SPR 0x390–0x393 Supervisor R/W
VNDXA
VNDXC
0
7
8
15 16
23 24
31
VNDXB
VNDXD
Figure 10-18. Data Cache Normal Victim Registers (DNV0–DNV3)
For all victim index fields, the number of bits used
to select the cache way for replacement depends
on the implemented cache size. See Table 4-3, on
page -98
Victim Index A (for cache lines with EA[25:26] =
0b00)
0:7
VNDXA
VNDXB
for more information.
Victim Index B (for cache lines with EA[25:26] =
0b01)
8:15
Victim Index C (for cache lines with EA[25:26] =
0b10)
16:23
24:31
VNDXC
Victim Index D (for cache lines with EA[25:26] =
0b11)
VNDXD
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DTV0–DTV3
Data Cache Transient Victim 0–3
PPC440x5 CPU Core User’s Manual
Preliminary
DTV0–DTV3
SPR 0x394–0x397 Supervisor R/W
VNDXA
VNDXC
0
7
8
15 16
23 24
31
VNDXB
VNDXD
Figure 10-19. Data Cache Transient Victim Registers (DTV0–DTV3)
For all victim index fields, the number of bits used
to select the cache way for replacement depends
on the implemented cache size. See Table 4-3, on
page -98
Victim Index A (for cache lines with EA[25:26] =
0b00)
0:7
VNDXA
VNDXB
for more information.
Victim Index B (for cache lines with EA[25:26] =
0b01)
8:15
Victim Index C (for cache lines with EA[25:26] =
0b10)
16:23
24:31
VNDXC
Victim Index D (for cache lines with EA[25:26] =
0b11)
VNDXD
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DVC1–DVC2
Data Value Compare Registers
Preliminary
PPC440x5 CPU Core User’s Manual
DVC1–DVC2
SPR 0x13E–0x13F Supervisor R/W
0
31
Figure 10-20. Data Value Compare Registers (DVC1–DVC2)
0:31
Data value to compare
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
DVLIM
Data Cache Victim Limit
PPC440x5 CPU Core User’s Manual
Preliminary
DVLIM
SPR 0x398 Supervisor R/W
TFLOOR
NFLOOR
0
1
2
9
10
12 13
20 21
23 24
31
TCEILING
Figure 10-21. Data Cache Victim Limit (DVLIM)
0:1
Reserved
The number of bits in the TFLOOR field varies,
depending on the implemented cache size. See
Table 4-3, on page -98 for more information.
2:9
TFLOOR
Transient Floor
Reserved
10:12
13:20
21:23
24:31
The number of bits in the TCEILING field varies,
depending on the implemented cache size. See
Table 4-3, on page -98 for more information.
TCEILING
Transient Ceiling
Reserved
The number of bits in the NFLOOR field varies,
depending on the implemented cache size. See
Table 4-3, on page -98 for more information.
NFLOOR
Normal Floor
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
ESR
Exception Syndrome Register
Preliminary
PPC440x5 CPU Core User’s Manual
ESR
SPR 0x03E Supervisor R/W
MCI
0
PIL
4
BO
PCRE
PCRF
PTR
6
ST
8
AP
1
3
5
7
9
10 11 12 13 14 15 16
26 27 28 29
PCMP
31
PPR
FP
DLK
PUO PIE
Figure 10-22. Exception Syndrome Register (ESR)
Machine Check—Instruction Fetch Exception
This is an implementation-dependent field of the
ESR and is not part of the PowerPC Book-E Archi-
tecture.
0
Instruction Machine Check exception did not
0
MCI
occur.
1 Instruction Machine Check exception occurred.
Reserved
1:3
4
Program Interrupt—Illegal Instruction Exception
0 Illegal Instruction exception did not occur.
1 Illegal Instruction exception occurred.
PIL
Program Interrupt—Privileged Instruction Excep-
tion
5
6
PPR
PTR
0
Privileged Instruction exception did not occur.
1
Privileged Instruction exception occurred.
Program Interrupt—Trap Exception
0 Trap exception did not occur.
1 Trap exception occurred.
Floating Point Operation
0 Exception was not caused by a floating point
instruction.
1 Exception was caused by a floating point
instruction.
7
8
FP
ST
Store Operation
0 Exception was not caused by a store-type
storage access or cache management
instruction.
1 Exception was caused by a store-type storage
access or cache management instruction.
9
Reserved
Data Storage Interrupt—Locking Exception
00 Locking exception did not occur.
01 Locking exception was caused by dcbf.
10 Locking exception was caused by icbi.
11 Reserved
10:11
DLK
AP Operation
0 Exception was not caused by an auxiliary
processor instruction.
1 Exception was caused by an auxiliary processor
instruction.
12
13
AP
Program Interrupt—Unimplemented Operation
Exception
0 Unimplemented Operation exception did not
PUO
occur.
1 Unimplemented Operation exception occurred.
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
ESR (cont.)
Exception Syndrome Register
PPC440x5 CPU Core User’s Manual
Preliminary
Byte Ordering Exception
0 Byte Ordering exception did not occur.
14
BO
1 Byte Ordering exception occurred.
Program Interrupt—Imprecise Exception
0 Exception occurred precisely; SRR0 contains
the address of the instruction that caused the
exception.
1 Exception occurred imprecisely; SRR0 contains
the address of an instruction after the one which
caused the exception.
This field is only set for a Floating-Point Enabled
exception type Program interrupt, and then only
when the interrupt occurs imprecisely due to
MSR[FE0,FE1] being set to a non-zero value when
an attached floating-point unit is already signaling
the Floating-Point Enabled exception (that is,
FPSCR[FEX] is already 1).
15
PIE
16:26
27
Reserved
This is an implementation-dependent field of the
ESR and is not part of the PowerPC Book-E Archi-
tecture.
Program Interrupt—Condition Register Enable
0 Instruction which caused the exception is not a
floating-point CR-updating instruction.
1 Instruction which caused the exception is a
floating-point CR-updating instruction.
PCRE
PCMP
PCRF
This field is only defined for a Floating-Point
Enabled exception type Program interrupt, and
then only when ESR[PIE] is 0.
This is an implementation-dependent field of the
ESR and is not part of the PowerPC Book-E Archi-
tecture.
Program Interrupt—Compare
0 Instruction which caused the exception is not a
floating-point compare type instruction
1 Instruction which caused the exception is a
floating-point compare type instruction.
28
This field is only defined for a Floating-Point
Enabled exception type Program interrupt, and
then only when ESR[PIE] is 0.
This is an implementation-dependent field of the
ESR and is not part of the PowerPC Book-E Archi-
tecture.
Program Interrupt—Condition Register Field
If ESR[PCRE]=1, this field indicates which CR field
was to be updated by the floating-point instruction
which caused the exception.
29:31
This field is only defined for a Floating-Point
Enabled exception type Program interrupt, and
then only when ESR[PIE] is 0.
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
IAC1–IAC4
Instruction Address Compare Registers
PPC440x5 CPU Core User’s Manual
Preliminary
IAC1–IAC4
SPR 0x138–0x13B Supervisor R/W
0
29 30 31
Figure 10-24. Instruction Address Compare Registers (IAC1–IAC4)
0:29
Instruction Address Compare (IAC) word address
Reserved
30:31
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
ICDBDR
Instruction Cache Debug Data Register
PPC440x5 CPU Core User’s Manual
Preliminary
ICDBDR
SPR 0x3D3 Supervisor Read-Only
0
31
Figure 10-25. Instruction Cache Debug Data Register (ICDBDR)
0:31
Instruction machine code from instruction cache
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
ICDBTRH
Instruction Cache Debug Tag Register High
PPC440x5 CPU Core User’s Manual
Preliminary
ICDBTRH
SPR 0x39F Supervisor Read-Only
TPAR
TEA
0
23 24 25 26 27 28
31
DAPAR
V
Figure 10-26. Instruction Cache Debug Tag Register High (ICDBTRH)
Bits 0:23 of the 32-bit effective address associated
with the cache line read by icread.
0:23
Tag Effective Address
Cache Line Valid
0 Cache line is not valid.
1 Cache line is valid.
The valid indicator for the cache line read by
icread.
24
V
The parity bits for the address tag for the cache
line read by icread, if CCR0[CRPE] is set.
25:26
TPAR
DAPAR
Tag Parity
The parity bit for the instruction word at the 32-bit
effective address specified in the icread instruc-
tion, if CCR0[CRPE] is set.
27
Instruction Data parity
Reserved
28:31
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
ICDBTRL
Instruction Cache Debug Tag Register Low
PPC440x5 CPU Core User’s Manual
Preliminary
ICDBTRL
SPR 0x39E Supervisor Read-Only
TID
TS
0
21 22 23 24
TD
31
Figure 10-27. Instruction Cache Debug Tag Register Low (ICDBTRL)
0:21
Reserved
The address space portion of the virtual address
associated with the cache line read by icread.
22
TS
Translation Space
Translation ID (TID) Disable
0 TID enable
1 TID disable
TID Disable field for the memory page associated
with the cache line read by icread.
23
TD
TID
TID field portion of the virtual address associated
with the cache line read by icread.
24:31
Translation ID
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
INV0–INV3
Instruction Cache Normal Victim 0–3
PPC440x5 CPU Core User’s Manual
Preliminary
INV0–INV3
SPR 0x370–0x373 Supervisor R/W
VNDXA
VNDXC
0
7
8
15 16
23 24
31
VNDXB
VNDXD
Figure 10-28. Instruction Cache Normal Victim Registers (INV0–INV3)
For all victim index fields, the number of bits used
to select the cache way for replacement depends
on the implemented cache size. See Table 4-3, on
page -98
Victim Index A (for cache lines with EA[25:26] =
0b00)
0:7
VNDXA
VNDXB
for more information.
Victim Index B (for cache lines with EA[25:26] =
0b01)
8:15
Victim Index C (for cache lines with EA[25:26] =
0b10)
16:23
24:31
VNDXC
Victim Index D (for cache lines with EA[25:26] =
0b11)
VNDXD
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
ITV0–ITV3
Instruction Cache Transient Victim 0–3
PPC440x5 CPU Core User’s Manual
Preliminary
ITV0–ITV3
SPR 0x374–0x377 Supervisor R/W
VNDXA
VNDXC
0
7
8
15 16
23 24
31
VNDXB
VNDXD
Figure 10-29. Instruction Cache Transient Victim Registers (ITV0–ITV3)
For all victim index fields, the number of bits used
to select the cache way for replacement depends
on the implemented cache size. See Table 4-3, on
page -98
Victim Index A (for cache lines with EA[25:26] =
0b00)
0:7
VNDXA
VNDXB
for more information.
Victim Index B (for cache lines with EA[25:26] =
0b01)
8:15
Victim Index C (for cache lines with EA[25:26] =
0b10)
16:23
24:31
VNDXC
Victim Index D (for cache lines with EA[25:26] =
0b11)
VNDXD
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
IVLIM
Instruction Cache Victim Limit
PPC440x5 CPU Core User’s Manual
Preliminary
IVLIM
SPR 0x399 Supervisor R/W
TFLOOR
NFLOOR
0
1
2
9
10
12 13
20 21
23 24
31
TCEILING
Figure 10-30. Instruction Cache Victim Limit (IVLIM)
0:1
Reserved
The number of bits in the TFLOOR field varies,
depending on the implemented cache size. See
Table 4-3, on page -98 for more information.
2:9
TFLOOR
Transient Floor
Reserved
10:12
13:20
21:23
24:31
The number of bits in the TCEILING field varies,
depending on the implemented cache size. See
Table 4-3, on page -98 for more information.
TCEILING
Transient Ceiling
Reserved
The number of bits in the NFLOOR field varies,
depending on the implemented cache size. See
Table 4-3, on page -98 for more information.
NFLOOR
Normal Floor
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
IVOR0–IVOR15
Interrupt Vector Offset Registers
Preliminary
PPC440x5 CPU Core User’s Manual
IVOR0–IVOR15
SPR 0x190–0x19F Supervisor R/W
IVO
0
15 16
27 28
31
Figure 10-31. Interrupt Vector Offset Registers (IVOR0–IVOR15)
0:15
Reserved
16:27
28:31
IVO
Interrupt Vector Offset
Reserved
Table 10-3. Interrupt Types Associated with each IVOR
IVOR
IVOR0
Interrupt Type
Critical Input
IVOR1
IVOR2
IVOR3
IVOR4
IVOR5
IVOR6
IVOR7
IVOR8
IVOR9
IVOR10
IVOR11
IVOR12
IVOR13
IVOR14
IVOR15
Machine Check
Data Storage
Instruction Storage
External Input
Alignment
Program
Floating Point Unavailable
System Call
Auxiliary Processor Unavailable
Decrementer
Fixed Interval Timer
Watchdog Timer
Data TLB Error
Instruction TLB Error
Debug
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
IVPR
Interrupt Vector Prefix Register
PPC440x5 CPU Core User’s Manual
Preliminary
IVPR
SPR 0x03F Supervisor R/W
IVP
0
15 16
31
Figure 10-32. Interrupt Vector Prefix Register (IVPR)
0:15
IVP
Interrupt Vector Prefix
Reserved
16:31
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
MCSR
Machine Check Status Register
PPC440x5 CPU Core User’s Manual
Preliminary
MCSR
SPR 0x23C Supervisor Read/Clear
DRB
2
TLBP
IMPE
8
MCS
0
DCSP
6
1
3
4
5
7
9
31
IB
ICP
DCFP
DWB
Figure 10-34. Machine Check Status Register (MCSR)
Set when a machine check exception occurs
that is handled in the asynchronous fashion.
One of MCSR bits 1:7 will be set simulta-
neously to indicate the exception type. When
MSR[ME] and this bit are both set, Machine
Check interrupt is taken.
Machine Check Summary
0 No async machine check exception pending
1 Async machine check exception pending
0
MCS
Instruction PLB Error
0 Exception not caused by Instruction Read PLB
interrupt request (IRQ)
1 Exception caused by Instruction Read PLB interrupt
request (IRQ)
1
2
3
IB
Data Read PLB Error
0 Exception not caused by Data Read PLB interrupt
request (IRQ)
1 Exception caused by Data Read PLB interrupt
request (IRQ)
DRB
DWB
Data Write PLB Error
0 Exception not caused by Data Write PLB interrupt
request (IRQ)
1 Exception caused by Data Write PLB interrupt
request (IRQ)
Translation Lookaside Buffer Parity Error
0 Exception not caused by TLB parity error
1 Exception caused by TLB parity error
4
5
TLBP
ICP
Instruction Cache Parity Error
0 Exception not caused by I-cache parity error
1 Exception caused by I-cache parity error
Set if and only If the DCU parity error was dis-
covered during a DCU Search operation.
Data Cache Search Parity Error
0 Exception not caused by DCU Search parity error
1 Exception caused by DCU Search parity error
6
7
DCSP
See Data Cache Parity Operations on
page 129.
Set if and only If the DCU parity error was dis-
covered during a DCU Flush operation.
Data Cache Flush Parity Error
0 Exception not caused by DCU Flush parity error
1 Exception caused by DCU Flush parity error
DCFP
IMPE
See Data Cache Parity Operations on
page 129.
Imprecise Machine Check Exception
0 No imprecise machine check exception occurred.
1 Imprecise machine check exception occurred.
Set if a machine check exception occurs that
sets MCSR[MCS] (or would if it were not
already set) and MSR[ME] = 0.
8
9:31
Reserved
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
MCSRR0
Machine Check Save/Restore Register 0
PPC440x5 CPU Core User’s Manual
Preliminary
MCSRR0
SPR 0x23A Supervisor R/W
See Machine Check Save/Restore Register 0 (MCSRR0) on page 169.
0
29 30 31
Figure 10-35. Machine Check Save/Restore Register 0 (MCSRR0)
0:29
Return address for machine check interrupts
Reserved
30:31
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
MCSRR1
Machine Check Save/Restore Register 1
PPC440x5 CPU Core User’s Manual
Preliminary
MCSRR1
SPR 0x23B Supervisor R/W
IS
EE
DE
FP
WE
FE0
0
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
31
FE1
CE
ME
DWE
DS
PR
Figure 0-2. Machine Check Save/Restore Register 1 (MCSRR1)
0:31
Copy of the MSR at the time of a machine check interrupt.
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
MMUCR
Memory Management Control Register
PPC440x5 CPU Core User’s Manual
Preliminary
MMUCR
SPR 0x3B2 Supervisor R/W
STID
SWOA
U2SWOAE
IULXE
10 11 12 13 14 15 16
DULXE
0
6
7
8
9
23 24
31
STS
U1TE
Figure 10-36. Memory Management Unit Control Register (MMUCR)
0:6
Reserved
Store Without Allocate
0 Cacheable store misses allocate a line in the
data cache.
7
SWOA
If MMUCR[U2SWOAE] = 1, this field is ignored.
1 Cacheable store misses do not allocate a line
in the data cache.
8
Reserved
U1 Transient Enable
0 Disable U1 storage attribute as transient
storage attribute.
9
U1TE
1 Enable U1 storage attribute as transient
storage attribute.
U2 Store without Allocate Enable
0 Disable U2 storage attribute control of store
without allocate.
1 Enable U2 storage attribute control of store
without allocate.
If MMUCR[U2SWOAE] = 1, the U2 storage
attribute overrides MMUCR[SWOA].
10
U2SWOAE
11
12
Reserved
Data Cache Unlock Exception Enable
0 Data cache unlock exception is disabled.
1 Data cache unlock exception is enabled.
dcbf in user mode causes a Cache Locking
exception type Data Storage interrupt when
MMUCR[DULXE] is 1.
DULXE
IULXE
Instruction Cache Unlock Exception Enable
0 Instruction cache unlock exception is disabled.
1 Instruction cache unlock exception is enabled.
icbi in user mode causes a Cache Locking
exception type Data Storage interrupt when
MMUCR[IULXE] is 1.
13
14
Reserved
Specifies the value of the translation space (TS)
field for tlbsx[.]
15
STS
Search Translation Space
Reserved
16:23
Specifies the value of the TID field for the
tlbsx[.]; also used to transfer a TLB entry’s TID
value for tlbre and tlbwe .
24:31
STID
Search Translation ID
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
MSR
Machine State Register
PPC440x5 CPU Core User’s Manual
Preliminary
MSR
Supervisor R/W
EE
FP
WE
FE0
DE
IS
0
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
31
FE1
CE
ME
DWE
DS
PR
Figure 10-37. Machine State Register (MSR)
0:12
Reserved
Wait State Enable
0 The processor is not in the wait state.
1 The processor is in the wait state.
If MSR[WE] = 1, the processor remains in the wait
state until an interrupt is taken, a reset occurs, or
an external debug tool clears WE.
13
WE
CE
Critical Interrupt Enable
0 Critical Input and Watchdog Timer interrupts are
disabled.
1 Critical Input and Watchdog Timer interrupts are
enabled.
14
15
16
Reserved
External Interrupt Enable
0 External Input, Decrementer, and Fixed Interval
Timer interrupts are disabled.
1 External Input, Decrementer, and Fixed Interval
Timer interrupts are enabled.
EE
PR
Problem State
0 Supervisor state (privileged instructions can be
executed)
1 Problem state (privileged instructions can not be
executed)
17
Floating Point Available
0 The processor cannot execute floating-point
instructions
1 The processor can execute floating-point
instructions
18
19
20
FP
Machine Check Enable
0 Machine Check interrupts are disabled
1 Machine Check interrupts are enabled.
ME
FE0
Floating-point exception mode 0
0 If MSR[FE1] = 0, ignore exceptions mode; if
MSR[FE1] = 1, imprecise nonrecoverable mode
1 If MSR[FE1] = 0, imprecise recoverable mode; if
MSR[FE1] = 1, precise mode
Debug Wait Enable
0 Disable debug wait mode.
1 Enable debug wait mode.
21
22
DWE
DE
Debug interrupt Enable
0 Debug interrupts are disabled.
1 Debug interrupts are enabled.
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
MSR (cont.)
Machine State Register
Preliminary
PPC440x5 CPU Core User’s Manual
Floating-point exception mode 1
0 If MSR[FE0] = 0, ignore exceptions mode; if
MSR[FE0] = 1, imprecise recoverable mode
1 If MSR[FE0] = 0, imprecise non-recoverable
mode; if MSR[FE0] = 1, precise mode
23
FE1
24:25
Reserved
Instruction Address Space
0 All instruction storage accesses are directed to
address space 0 (TS = 0 in the relevant TLB
entry).
26
IS
1 All instruction storage accesses are directed to
address space 1 (TS = 1 in the relevant TLB
entry).
Data Address Space
0 All data storage accesses are directed to
address space 0 (TS = 0 in the relevant TLB
entry).
27
DS
1 All data storage accesses are directed to
address space 1 (TS = 1 in the relevant TLB
entry).
28:31
Reserved
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
PIR
Processor Identification Register
Preliminary
PPC440x5 CPU Core User’s Manual
PIR
SPR 0x11E Supervisor Read-Only
PIN
0
27 28
31
Figure 10-39. Processor Identification Register (PIR)
0:27
Reserved
28:31
PIN
Processor Identification Number (PIN)
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
PVR
Processor Version Register
PPC440x5 CPU Core User’s Manual
Preliminary
PVR
SPR 0x11F Supervisor Read-Only
OWN
0
11 12
31
PVN
Figure 10-40. Processor Version Register (PVR)
0:11
OWN
Owner Identifier
Identifies the owner of a core.
Implementation-specific value identifying the spe-
cific version and use of a processor core within a
chip.
12:31
PVN
Processor Version Number
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
RSTCFG
Reset Configuration
Preliminary
PPC440x5 CPU Core User’s Manual
RSTCFG
SPR 39B Supervisor Read-Only
U0
15 16 17 18 19 20
U1 U3
U2
E
0
23 24 25
27 28
31
ERPN
Figure 10-41. Reset Configuration
0:15
Reserved
U0 Storage Attribute
0 U0 storage attribute is disabled
1 U0 storage attribute is enabled
16
U0
U1
U2
U3
See Table 5-1 on page 135.
See Table 5-1 on page 135.
See Table 5-1 on page 135.
See Table 5-1 on page 135.
U1 Storage Attribute
0 U1 storage attribute is disabled
1 U1 storage attribute is enabled
17
18
U2 Storage Attribute
0 U2 storage attribute is disabled
1 U2 storage attribute is enabled
U3 Storage Attribute
0 U3 storage attribute is disabled
19
1 U3 storage attribute is enabled
20:23
24
Reserved
E Storage Attribute
0 Accesses to the page are big endian.
E
1 Accesses to the page are little endian.
25:27
Reserved
This TLB field is prepended to the translated
address to form a 36-bit real address. See Table
5.4 Address Translation on page 140 and Table
5-3 Page Size and Real Address Formation on
page 142.
28:31
ERPN
Extended Real Page Number
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
SPRG0–SPRG7
Special Purpose Register General
PPC440x5 CPU Core User’s Manual
Preliminary
SPRG0–SPRG7
SPR 0x104–0x107 (User/Supervisor Read-Only); SPR 0x110–0x113 (Supervisor R/W);
SPR 0x114–0x117 (Supervisor Write-Only)
0
31
Figure 10-42. Special Purpose Registers General (SPRG0–SPRG7)
0:31
General data
Software value; no hardware usage.
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
SRR0
Save/Restore Register 0
Preliminary
PPC440x5 CPU Core User’s Manual
SRR0
SPR 0x01A Supervisor R/W
0
29 30 31
Figure 10-43. Save/Restore Register 0 (SRR0)
0:29
Return address for non-critical interrupts
Reserved
30:31
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
SRR1
Save/Restore Register 1
PPC440x5 CPU Core User’s Manual
Preliminary
SRR1
SPR 0x01B Supervisor R/W
IS
EE
DE
FP
WE
FE0
0
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
31
FE1
CE
ME
DWE
DS
PR
Figure 10-44. Save/Restore Register 1 (SRR1)
Copy of the MSR at the time of a non-critical inter-
rupt.
0:31
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
TBL
Time Base Lower
Preliminary
PPC440x5 CPU Core User’s Manual
TBL
SPR 0x10C (User/Supervisor Read-Only); SPR 0x11C (Supervisor Write-Only)
0
31
Figure 10-45. Time Base Lower (TBL)
0:31
Time Base Lower
Low-order 32 bits of time base.
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
TBU
Time Base Upper
PPC440x5 CPU Core User’s Manual
Preliminary
TBU
SPR 0x10D (User/Supervisor Read-Only); SPR 0x11D (Supervisor Write-Only)
0
31
Figure 10-46. Time Base Upper (TBU)
0:31
Time Base Upper
High-order 32 bits of time base.
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
TCR
Timer Control Register
Preliminary
PPC440x5 CPU Core User’s Manual
TCR
SPR 0x154 Supervisor R/W
WP
WIE
FP FIE
0
1
2
3
4
5
6
7
8
9
10
31
WRC
DIE
ARE
Figure 10-47. Timer Control Register (TCR)
Watchdog Timer Period
21
00 2 time base clocks
25
01 2 time base clocks
0:1
WP
29
10 2 time base clocks
33
11 2 time base clocks
TCR[WRC] resets to 0b00.
Watchdog Timer Reset Control
00 No Watchdog Timer reset will occur.
Type of reset to cause upon Watchdog Timer excep-
tion with TSR[ENW,WIS]=0b11.
01 Core reset
10 Chip reset
11 System reset
2:3
WRC
WIE
This field can be set by software, but cannot be
cleared by software, except by a software-induced
reset.
Watchdog Timer Interrupt Enable
0 Disable Watchdog Timer interrupt.
1 Enable Watchdog Timer interrupt.
4
5
Decrementer Interrupt Enable
0 Disable Decrementer interrupt.
1 Enable Decrementer interrupt.
DIE
Fixed Interval Timer (FIT) Period
13
00 2 time base clocks
17
01 2 time base clocks
6:7
8
FP
21
10 2 time base clocks
25
11 2 time base clocks
FIT Interrupt Enable
0 Disable Fixed Interval Timer interrupt.
FIE
1 Enable Fixed Interval Timer interrupt.
Auto-Reload Enable
0 Disable auto reload.
1 Enable auto reload.
9
ARE
TCR[ARE] resets to 0b0.
10:31
Reserved
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
TSR
Timer Status Register
PPC440x5 CPU Core User’s Manual
Preliminary
TSR
SPR 0x150 Supervisor Read/Clear
ENW WRS
FIS
0
1
2
3
4
5
6
31
WIS
DIS
Figure 10-48. Timer Status Register (TSR)
Enable Next Watchdog Timer Exception
0 Action on next Watchdog Timer exception is to set
TSR[ENW] = 1.
1 Action on next Watchdog Timer exception is governed
by TSR[WIS].
0
ENW
WIS
Watchdog Timer Interrupt Status
0 Watchdog Timer exception has not occurred.
1
1 Watchdog Timer exception has occurred.
Watchdog Timer Reset Status
00 No Watchdog Timer reset has occurred.
01 Core reset was forced by Watchdog Timer.
10 Chip reset was forced by Watchdog Timer.
11 System reset was forced by Watchdog Timer.
2:3
4
WRS
Decrementer Interrupt Status
0 Decrementer exception has not occurred.
1 Decrementer exception has occurred.
DIS
FIS
Fixed Interval Timer (FIT) Interrupt Status
0 Fixed Interval Timer exception has not occurred.
5
1 Fixed Interval Timer exception has occurred.
6:31
Reserved
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
USPRG0
User Special Purpose Register General 0
PPC440x5 CPU Core User’s Manual
Preliminary
USPRG0
SPR 0x100 (User R/W)
0
31
Figure 10-49. User Special Purpose Register General (USPRG0)
0:31
General data
Software value; no hardware usage.
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
XER
Integer Exception Register
PPC440x5 CPU Core User’s Manual
Preliminary
XER
SPR 0x001 User R/W
TBC
CA
2
SO
0
1
3
24 25
31
OV
Figure 10-50. Integer Exception Register (XER)
Summary Overflow
0 No overflow has occurred.
1 Overflow has occurred.
Can be set by mtspr or by integer or auxiliary
processor instructions with the [o] option; can be
reset by mtspr or by mcrxr.
0
SO
Can be set by mtspr or by integer or allocated
instructions with the [o] option; can be reset by
mtspr, by mcrxr, or by integer or allocated
instructions with the [o] option.
Overflow
0 No overflow has occurred.
1 Overflow has occurred.
1
OV
Can be set by mtspr or by certain integer arith-
metic and shift instructions; can be reset by
mtspr, by mcrxr, or by certain integer arithmetic
and shift instructions.
Carry
0 Carry has not occurred.
1 Carry has occurred.
2
CA
3:24
Reserved
Used as a byte count by lswx and stswx; written
by dlmzb[.] and by mtspr.
25:31
TBC
Transfer Byte Count
regsumm440core.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Appendix A. Instruction Summary
This appendix describes the various instruction formats, and lists all of the PPC440x5 instructions summa-
rized alphabetically and by opcode.
Appendix A.1 on page 519 illustrates the PPC440x5 instruction forms (allowed arrangements of fields within
instructions).
Appendix A.2 on page 524 lists all PPC440x5 mnemonics, including extended mnemonics. A short functional
description is included for each mnemonic.
Appendix A.3 on page 557 identifies those opcodes which are allocated by PowerPC Book-E for implementa-
tion-dependent usage, including auxiliary processors.
Appendix A.4 on page 557 identifies those opcodes which are identified by PowerPC Book-E as “preserved”
for compatibility with previous versions of the architecture.
Appendix A.5 on page 558 indentifies those opcodes which are “reserved” for use by future versions of the
architecture.
Appendix A.6 on page 559, lists all instructions implemented within the PPC440x5 core, sorted by primary
and secondary opcodes. Extended mnemonics are not included in the opcode list, but allocated, preserved,
and reserved-nop opcodes are included.
A.1 Instruction Formats
Instructions are four bytes long. Instruction addresses are always word-aligned.
Instruction bits 0 through 5 always contain the primary opcode. Many instructions have an extended opcode
in another field. Remaining instruction bits contain additional fields. All instruction fields belong to one of the
following categories:
• Defined
These instructions contain values, such as opcodes, that cannot be altered. The instruction format dia-
grams specify the values of defined fields.
• Variable
These fields contain operands, such as GPR selectors and immediate values, that can vary from execu-
tion to execution. The instruction format diagrams specify the operands in the variable fields.
• Reserved
Bits in reserved fields should be set to 0. In the instruction format diagrams, /, //, or /// indicate reserved
fields.
If any bit in a defined field does not contain the expected value, the instruction is illegal and an illegal instruc-
tion exception occurs. If any bit in a reserved field does not contain 0, the instruction form is invalid; its result
is architecturally undefined. The PPC440x5 core executes all invalid instruction forms without causing an
illegal instruction exception.
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
A.1.1 Instruction Fields
PPC440x5 instructions contain various combinations of the following fields, as indicated in the instruction
format diagrams that follow the field definitions. Numbers, enclosed in parentheses, that follow the field
names indicate bit positions; bit fields are indicated by starting and stopping bit positions separated by
colons.
AA (30)
Absolute address bit.
0 The immediate field represents an address relative to the current instruction address
(CIA). The effective address (EA) of the branch is either the sum of the LI field sign-
extended to 32 bits and the branch instruction address, or the sum of the BD field
sign-extended to 32 bits and the branch instruction address.
1 The immediate field represents an absolute address. The EA of the branch is either
the LI field or the BD field, sign-extended to 32 bits.
BA (11:15)
BB (16:20)
BD (16:29)
Specifies a bit in the CR used as a source of a CR-logical instruction.
Specifies a bit in the CR used as a source of a CR-logical instruction.
An immediate field specifying a 14-bit signed twos complement branch displacement.
This field is concatenated on the right with 0b00 and sign-extended to 32 bits.
BF (6:8)
Specifies a field in the CR used as a target in a compare or mcrf instruction.
Specifies a field in the CR used as a source in a mcrf instruction.
BFA (11:13)
BI (11:15)
Specifies a bit in the CR used as a source for the condition of a conditional branch
instruction.
BO (6:10)
BT (6:10)
D (16:31)
Specifies a bit in the CR used as a target as the result of a CR-Logical instruction.
Specifies a 16-bit signed two’s-complement integer displacement for load/store
instructions.
DCRF (11:20) Specifies a device control register (DCR). This field represents the DCR Number
(DCRN) with the upper and lower five bits reversed (that is,
DCRF = DCRN[5:9] || DCRN[0:4]).
FXM (12:19) Field mask used to identify CR fields to be updated by the mtcrf instruction.
IM (16:31)
LI (6:29)
An immediate field used to specify a 16-bit value (either signed integer or unsigned).
An immediate field specifying a 24-bit signed twos complement branch displacement;
this field is concatenated on the right with b'00' and sign-extended to 32 bits.
LK (31)
Link bit.
0 Do not update the link register (LR).
1 Update the LR with the address of the next instruction.
MB (21:25)
ME (26:30)
Mask begin.
Used in rotate-and-mask instructions to specify the beginning bit of a mask.
Mask end.
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Used in rotate-and-mask instructions to specify the ending bit of a mask.
NB (16:20)
OPCD (0:5)
Specifies the number of bytes to move in an immediate string load or store.
Primary opcode. Primary opcodes, in decimal, appear in the instruction format diagrams
presented with individual instructions. The OPCD field name does not appear in
instruction descriptions.
OE (21)
Enables setting the OV and SO fields in the fixed-point exception register (XER) for
extended arithmetic.
RA (11:15)
RB (16:20)
Rc (31)
A GPR used as a source or target.
A GPR used as a source.
Record bit.
0 Do not set the CR.
1 Set the CR to reflect the result of an operation.
set.
RS (6:10)
RT (6:10)
SH (16:20)
A GPR used as a source.
A GPR used as a target.
Specifies a shift amount.
SPRF (11:20) Specifies a special purpose register (SPR). This field represents the SPR Number
(SPRN) with the upper and lower five bits reversed (that is,
SPRF = SPRN[5:9] || SPRN[0:4]).
TO (6:10)
Specifies the conditions on which to trap, as described under tw and twi instructions.
Specifies the portion of a TLB entry to be read/written by tlbre/tlbwe.
WS (16:20)
XO (21:30)
Extended opcode for instructions without an OE field. Extended opcodes, in decimal,
appear in the instruction format diagrams presented with individual instructions. The XO
field name does not appear in instruction descriptions.
XO (22:30)
Extended opcode for instructions with an OE field. Extended opcodes, in decimal,
appear in the instruction format diagrams presented with individual instructions. The XO
field name does not appear in instruction descriptions.
A.1.2 Instruction Format Diagrams
tions of instruction fields. Table A-5 on page 559 indicates which “form” is utilized by each PPC440x5
opcode. Fields indicated by slashes (/, //, or ///) are reserved. The figures are adapted from the PowerPC
User Instruction Set Architecture.
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
A.1.2.1 I-Form
OPCD
LI
0
6
31
Figure A-1. I Instruction Format
A.1.2.2 B-Form
OPCD
BO
BI
BD
AA LK
0
6
11
16
30 31
Figure A-2. B Instruction Format
A.1.2.3 SC-Form
OPCD
///
///
///
1
/
0
6
11
16
30 31
Figure A-3. SC Instruction Format
A.1.2.4 D-Form
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
RT
RS
RS
RS
RA
RA
RA
RA
RA
RA
RA
D
SI
D
UI
SI
UI
SI
BF
BF
/
/
L
L
TO
0
6
11
16
31
Figure A-4. D Instruction Format
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
A.1.2.5 X-Form
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
OPCD
0
RT
RT
RT
RT
RT
RT
RS
RS
RS
RS
RS
RS
RS
RS
RS
RA
RA
RA
RA
///
RB
RB
NB
WS
RB
///
XO
Rc
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
XO
/
/
/
/
///
/
RA
RA
RA
RA
RA
RA
RA
///
RB
RB
RB
NB
WS
SH
///
Rc
1
/
/
/
Rc
Rc
RB
///
/
///
/
BF
BF
BF
BF
BF
/
L
RA
RB
///
/
//
//
//
//
BFA
//
Rc
///
///
///
/
U
Rc
///
///
/
TO
BT
///
RA
///
RB
///
/
Rc
/
RA
///
RB
///
///
/
///
///
E
//
/
6
11
16
21
31
Figure A-5. X Instruction Format
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
A.1.2.6 XL-Form
OPCD
OPCD
OPCD
OPCD
0
BT
BC
BA
BI
BB
///
XO
XO
XO
XO
/
LK
/
BF
//
BFA
//
///
///
///
///
/
6
11
16
21
31
Figure A-6. XL Instruction Format
A.1.2.7 XFX-Form
OPCD
OPCD
OPCD
OPCD
OPCD
RT
RT
RT
RS
RS
SPRF
DCRF
FXM
XO
XO
XO
XO
XO
/
/
/
/
/
SPRF
DCRF
16
/
/
0
6
11
21
31
Figure A-7. XFX Instruction Format
A.1.2.8 XO-Form
OPCD
OPCD
OPCD
RT
RT
RT
RA
RA
RA
RB
RB
///
OE
OE
/
XO
Rc
Rc
Rc
31
XO
XO
0
6
11
16
21 22
Figure A-8. XO Instruction Format
A.1.2.9 M-Form
ME
ME
OPCD
OPCD
RS
RS
RA
RA
RB
SH
MB
MB
Rc
Rc
31
0
6
11
16
21
26
Figure A-9. M Instruction Format
A.2 Alphabetical Summary of Implemented Instructions
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1 summarizes the PPC440x5 instruction set, including required extended mnemonics. All
mnemonics are listed alphabetically, without regard to whether the mnemonic is realized in hardware or soft-
ware. When an instruction supports multiple hardware mnemonics (for example, b, ba, bl, bla are all forms of
b), the instruction is alphabetized under the root form. The hardware instructions are described in detail in
Chapter 9, “Instruction Set,” which is also alphabetized under the root form. Section 9 also describes the
instruction operands and notation.
Programming Note: Bit 4 of the BO instruction field provides a hint about the most likely
a detailed description of branch prediction.) Assemblers should set
BO = 0 unless a specific reason exists otherwise. In the BO field values
4
4
assembler must enable the programmer to specify branch prediction. To
do this, the assembler supports suffixes for the conditional branch
mnemonics:
+ Predict branch to be taken.
– Predict branch not to be taken.
For example, bc also could be coded as bc+ or bc–, and bne also could
be coded bne+ or bne–. These alternate codings set BO = 1 only if the
4
requested prediction differs from the standard prediction. See Branch
Table A-1. PPC440x5 Instruction Syntax Summary
Other Registers
Changed
Mnemonic
Operands
Function
Page
add
CR[CR0]
add.
Add (RA) to (RB).
Place result in RT.
RT, RA, RB
XER[SO, OV]
addo
CR[CR0]
addo.
XER[SO, OV]
addc
CR[CR0]
addc.
addco
Add (RA) to (RB).
RT, RA, RB
RT, RA, RB
Place result in RT.
XER[SO, OV]
Place carry-out in XER[CA].
CR[CR0]
addco.
XER[SO, OV]
adde
CR[CR0]
adde.
addeo
Add XER[CA], (RA), (RB).
Place result in RT.
XER[SO, OV]
Place carry-out in XER[CA].
CR[CR0]
addeo.
addi
XER[SO, OV]
Add EXTS(IM) to (RA|0).
Place result in RT.
RT, RA, IM
RT, RA, IM
Add EXTS(IM) to (RA|0).
Place result in RT.
addic
Place carry-out in XER[CA].
Add EXTS(IM) to (RA|0).
Place result in RT.
RT, RA, IM
CR[CR0]
addic.
Place carry-out in XER[CA].
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
16
0) to (RA|0).
Add (IM ||
RT, RA, IM
addis
Place result in RT.
addme
addme.
addmeo
CR[CR0]
Add XER[CA], (RA), (-1).
Place result in RT.
RT, RA
RT, RA
XER[SO, OV]
Place carry-out in XER[CA].
CR[CR0]
addmeo.
XER[SO, OV]
addze
CR[CR0]
addze.
addzeo
Add XER[CA] to (RA).
Place result in RT.
XER[SO, OV]
Place carry-out in XER[CA].
CR[CR0]
addzeo.
XER[SO, OV]
and
AND (RS) with (RB).
Place result in RA.
RA, RS, RB
RA, RS, RB
CR[CR0]
and.
andc
andc.
AND (RS) with ¬(RB).
Place result in RA.
CR[CR0]
CR[CR0]
16
AND (RS) with ( 0 || IM).
RA, RS, IM
RA, RS, IM
andi.
Place result in RA.
16
AND (RS) with (IM ||
0).
CR[CR0]
andis.
Place result in RA.
Branch unconditional relative.
b
LI ← (target – CIA)
6:29
2
NIA ← CIA + EXTS(LI || 0)
Branch unconditional absolute.
ba
LI ← target
6:29
2
NIA ← EXTS(LI || 0)
target
Branch unconditional relative.
(LR) ← CIA + 4
(LR) ← CIA + 4
bl
LI ← (target – CIA)
6:29
2
NIA ← CIA + EXTS(LI || 0)
Branch unconditional absolute.
bla
bc
LI ← target
6:29
2
NIA ← EXTS(LI || 0)
Branch conditional relative.
CTR if BO = 0
2
BD ← (target – CIA)
16:29
2
NIA ← CIA + EXTS(BD || 0)
Branch conditional absolute.
CTR if BO = 0
2
bca
bcl
bcla
BD ← target
16:29
2
NIA ← EXTS(BD || 0)
BO, BI, target
Branch conditional relative.
CTR if BO = 0
2
(LR) ← CIA + 4
BD ← (target – CIA)
16:29
2
NIA ← CIA + EXTS(BD || 0)
Branch conditional absolute.
CTR if BO = 0
2
(LR) ← CIA + 4
BD ← target
16:29
2
NIA ← EXTS(BD || 0)
CTR if BO = 0
2
bcctr
Branch conditional to address in CTR.
BO, BI
Using (CTR) at exit from instruction,
2
CTR if BO = 0
2
bcctrl
NIA ← CTR
|| 0
(LR) ← CIA + 4
0:29
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Mnemonic
Operands
Function
Page
Changed
CTR if BO = 0
2
bclr
Branch conditional to address in LR.
BO, BI
Using (LR) at entry to instruction,
2
CTR if BO = 0
2
bclrl
bctr
bctrl
NIA ← LR
|| 0
(LR) ← CIA + 4
(LR) ← CIA + 4
0:29
Branch unconditionally to address in CTR.
Extended mnemonic for
bcctr 20,0
Extended mnemonic for
bcctrl 20,0
Decrement CTR.
Branch if CTR ≠ 0
Extended mnemonic for
bdnz
bc 16,0,target
Extended mnemonic for
bca 16,0,target
bdnza
bdnzl
target
Extended mnemonic for
bcl 16,0,target
(LR) ← CIA + 4
(LR) ← CIA + 4
Extended mnemonic for
bcla 16,0,target
bdnzla
Decrement CTR.
Branch if CTR ≠ 0 to address in LR.
Extended mnemonic for
bclr 16,0
bdnzlr
bdnzlrl
bdnzf
Extended mnemonic for
bclrl 16,0
(LR) ← CIA + 4
Decrement CTR.
Branch if CTR ≠ 0 AND CR
= 0
cr_bit
Extended mnemonic for
bc 0,cr_bit,target
Extended mnemonic for
bca 0,cr_bit,target
bdnzfa
bdnzfl
bdnzfla
cr_bit, target
Extended mnemonic for
bcl 0,cr_bit,target
(LR) ← CIA + 4
(LR) ← CIA + 4
Extended mnemonic for
bcla 0,cr_bit,target
Decrement CTR.
Branch if CTR ≠ 0 AND CR
= 0 to address in LR.
cr_bit
bdnzflr
bdnzflrl
Extended mnemonic for
cr_bit
bclr 0,cr_bit
Extended mnemonic for
bclrl 0,cr_bit
(LR) ← CIA + 4
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Decrement CTR.
Branch if CTR ≠ 0 AND CR
= 1.
cr_bit
bdnzt
Extended mnemonic for
bc 8,cr_bit,target
Extended mnemonic for
bca 8,cr_bit,target
bdnzta
bdnztl
bdnztla
cr_bit, target
Extended mnemonic for
bcl 8,cr_bit,target
(LR) ← CIA + 4
(LR) ← CIA + 4
Extended mnemonic for
bcla 8,cr_bit,target
Decrement CTR.
Branch if CTR ≠ 0 AND CR
= 1 to address in LR.
cr_bit
bdnztlr
bdnztlrl
bdz
Extended mnemonic for
cr_bit
bclr 8,cr_bit
Extended mnemonic for
bclrl 8,cr_bit
(LR) ← CIA + 4
Decrement CTR.
Branch if CTR = 0
Extended mnemonic for
bc 18,0,target
Extended mnemonic for
bca 18,0,target
bdza
bdzl
target
Extended mnemonic for
bcl 18,0,target
(LR) ← CIA + 4
(LR) ← CIA + 4
Extended mnemonic for
bcla 18,0,target
bdzla
Decrement CTR.
Branch if CTR = 0 to address in LR.
Extended mnemonic for
bdzlr
bdzlrl
bdzf
bclr 18,0
Extended mnemonic for
bclrl 18,0
(LR) ← CIA + 4
Decrement CTR.
Branch if CTR = 0 AND CR
= 0
cr_bit
Extended mnemonic for
bc 2,cr_bit,target
Extended mnemonic for
bca 2,cr_bit,target
bdzfa
bdzfl
cr_bit, target
Extended mnemonic for
bcl 2,cr_bit,target
(LR) ← CIA + 4
(LR) ← CIA + 4
Extended mnemonic for
bcla 2,cr_bit,target
bdzfla
Decrement CTR.
Branch if CTR = 0 AND CR
Extended mnemonic for
bclr 2,cr_bit
= 0 to address in LR.
cr_bit
bdzflr
cr_bit
Extended mnemonic for
bclrl 2,cr_bit
(LR) ← CIA + 4
bdzflrl
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Mnemonic
Operands
Function
Page
Changed
Decrement CTR.
Branch if CTR = 0 AND CR
Extended mnemonic for
= 1.
cr_bit
bdzt
bc 10,cr_bit,target
Extended mnemonic for
bca 10,cr_bit,target
bdzta
bdztl
cr_bit, target
Extended mnemonic for
bcl 10,cr_bit,target
(LR) ← CIA + 4
(LR) ← CIA + 4
Extended mnemonic for
bcla 10,cr_bit,target
bdztla
Decrement CTR.
Branch if CTR = 0 AND CR
Extended mnemonic for
bclr 10,cr_bit
= 1to address in LR.
cr_bit
bdztlr
bdztlrl
beq
cr_bit
Extended mnemonic for
bclrl 10,cr_bit
(LR) ← CIA + 4
Branch if equal.
UseCR[CR0] if cr_field is omitted.
Extended mnemonic for
bc 12,4∗cr_field+2,target
Extended mnemonic for
bca 12,4∗cr_field+2,target
Extended mnemonic for
bcl 12,4∗cr_field+2,target
Extended mnemonic for
bcla 12,4∗cr_field+2,target
beqa
beql
[cr_field], target
(LR) ← CIA + 4
(LR) ← CIA + 4
beqla
Branch if equal to address in CTR.
UseCR[CR0] if cr_field is omitted.
Extended mnemonic for
bcctr 12,4∗cr_field+2
Extended mnemonic for
beqctr
beqctrl
beqlr
[cr_field]
[cr_field]
(LR) ← CIA + 4
bcctrl 12,4∗cr_field+2
Branch if equal to address in LR.
UseCR[CR0] if cr_field is omitted.
Extended mnemonic for
bclr 12,4∗cr_field+2
Extended mnemonic for
bclrl 12,4∗cr_field+2
(LR) ← CIA + 4
beqlrl
bf
Branch if CR
= 0.
Extendedcrm_bnitemonic for
bc 4,cr_bit,target
Extended mnemonic for
bca 4,cr_bit,target
bfa
bfl
cr_bit, target
Extended mnemonic for
bcl 4,cr_bit,target
(LR) ← CIA + 4
(LR) ← CIA + 4
Extended mnemonic for
bcla 4,cr_bit,target
bfla
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Branch if CR
= 0 to address in CTR.
Extendedcrm_bnitemonic for
bfctr
bcctr 4,cr_bit
cr_bit
Extended mnemonic for
bcctrl 4,cr_bit
(LR) ← CIA + 4
(LR) ← CIA + 4
bfctrl
bflr
Branch if CR
= 0 to address in LR.
Extendedcrm_bnitemonic for
bclr 4,cr_bit
cr_bit
Extended mnemonic for
bclrl 4,cr_bit
bflrl
Branch if greater than or equal.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bge
bc 4,4∗cr_field+0,target
Extended mnemonic for
bca 4,4∗cr_field+0,target
Extended mnemonic for
bcl 4,4∗cr_field+0,target
Extended mnemonic for
bcla 4,4∗cr_field+0,target
bgea
bgel
[cr_field], target
(LR) ← CIA + 4
(LR) ← CIA + 4
bgela
Branch if greater than or equal to address in CTR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bcctr 4,4∗cr_field+0
Extended mnemonic for
bgectr
bgectrl
bgelr
bgelrl
bgt
[cr_field]
[cr_field]
(LR) ← CIA + 4
bcctrl 4,4∗cr_field+0
Branch if greater than or equal to address in LR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bclr 4,4∗cr_field+0
Extended mnemonic for
bclrl 4,4∗cr_field+0
Branch if greater than.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
(LR) ← CIA + 4
bc 12,4∗cr_field+1,target
Extended mnemonic for
bca 12,4∗cr_field+1,target
Extended mnemonic for
bcl 12,4∗cr_field+1,target
Extended mnemonic for
bcla 12,4∗cr_field+1,target
bgta
bgtl
[cr_field], target
(LR) ← CIA + 4
(LR) ← CIA + 4
bgtla
Branch if greater than to address in CTR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bcctr 12,4∗cr_field+1
Extended mnemonic for
bgtctr
bgtctrl
[cr_field]
(LR) ← CIA + 4
bcctrl 12,4∗cr_field+1
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Mnemonic
Operands
Function
Page
Changed
Branch if greater than to address in LR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bgtlr
[cr_field]
bclr 12,4∗cr_field+1
Extended mnemonic for
bclrl 12,4∗cr_field+1
Branch if less than or equal.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
(LR) ← CIA + 4
bgtlrl
ble
bc 4,4∗cr_field+1,target
Extended mnemonic for
bca 4,4∗cr_field+1,target
Extended mnemonic for
bcl 4,4∗cr_field+1,target
Extended mnemonic for
bcla 4,4∗cr_field+1,target
blea
blel
[cr_field], target
(LR) ← CIA + 4
(LR) ← CIA + 4
blela
Branch if less than or equal to address in CTR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bcctr 4,4∗cr_field+1
Extended mnemonic for
blectr
blectrl
blelr
[cr_field]
[cr_field]
(LR) ← CIA + 4
bcctrl 4,4∗cr_field+1
Branch if less than or equal to address in LR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bclr 4,4∗cr_field+1
Extended mnemonic for
bclrl 4,4∗cr_field+1
Branch unconditionally to address in LR.
Extended mnemonic for
bclr 20,0
(LR) ← CIA + 4
(LR) ← CIA + 4
blelrl
blr
Extended mnemonic for
bclrl 20,0
blrl
Branch if less than.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
blt
bc 12,4∗cr_field+0,target
Extended mnemonic for
bca 12,4∗cr_field+0,target
Extended mnemonic for
bcl 12,4∗cr_field+0,target
Extended mnemonic for
bcla 12,4∗cr_field+0,target
blta
bltl
[cr_field], target
(LR) ← CIA + 4
(LR) ← CIA + 4
bltla
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Branch if less than to address in CTR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bcctr 12,4∗cr_field+0
Extended mnemonic for
bltctr
[cr_field]
(LR) ← CIA + 4
bltctrl
bltlr
bcctrl 12,4∗cr_field+0
Branch if less than to address in LR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
[cr_field]
bclr 12,4∗cr_field+0
Extended mnemonic for
bclrl 12,4∗cr_field+0
Branch if not equal.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
(LR) ← CIA + 4
bltlrl
bne
bc 4,4∗cr_field+2,target
Extended mnemonic for
bca 4,4∗cr_field+2,target
Extended mnemonic for
bcl 4,4∗cr_field+2,target
Extended mnemonic for
bcla 4,4∗cr_field+2,target
bnea
bnel
[cr_field], target
(LR) ← CIA + 4
(LR) ← CIA + 4
bnela
Branch if not equal to address in CTR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bcctr 4,4∗cr_field+2
Extended mnemonic for
bnectr
bnectrl
bnelr
bnelrl
bng
[cr_field]
[cr_field]
(LR) ← CIA + 4
bcctrl 4,4∗cr_field+2
Branch if not equal to address in LR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bclr 4,4∗cr_field+2
Extended mnemonic for
bclrl 4,4∗cr_field+2
Branch if not greater than.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
(LR) ← CIA + 4
bc 4,4∗cr_field+1,target
Extended mnemonic for
bca 4,4∗cr_field+1,target
Extended mnemonic for
bcl 4,4∗cr_field+1,target
Extended mnemonic for
bcla 4,4∗cr_field+1,target
bnga
bngl
[cr_field], target
(LR) ← CIA + 4
(LR) ← CIA + 4
bngla
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Mnemonic
Operands
Function
Page
Changed
Branch if not greater than to address in CTR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bngctr
[cr_field]
bcctr 4,4∗cr_field+1
Extended mnemonic for
bcctrl 4,4∗cr_field+1
Branch if not greater than to address in LR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bclr 4,4∗cr_field+1
Extended mnemonic for
bclrl 4,4∗cr_field+1
Branch if not less than.
(LR) ← CIA + 4
bngctrl
bnglr
bnglrl
bnl
[cr_field]
(LR) ← CIA + 4
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bc 4,4∗cr_field+0,target
Extended mnemonic for
bca 4,4∗cr_field+0,target
Extended mnemonic for
bcl 4,4∗cr_field+0,target
Extended mnemonic for
bcla 4,4∗cr_field+0,target
bnla
bnll
[cr_field], target
(LR) ← CIA + 4
(LR) ← CIA + 4
bnlla
Branch if not less than to address in CTR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bcctr 4,4∗cr_field+0
Extended mnemonic for
bnlctr
bnlctrl
bnllr
[cr_field]
[cr_field]
(LR) ← CIA + 4
bcctrl 4,4∗cr_field+0
Branch if not less than to address in LR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bclr 4,4∗cr_field+0
Extended mnemonic for
bclrl 4,4∗cr_field+0
Branch if not summary overflow.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
(LR) ← CIA + 4
bnllrl
bns
bc 4,4∗cr_field+3,target
Extended mnemonic for
bca 4,4∗cr_field+3,target
Extended mnemonic for
bcl 4,4∗cr_field+3,target
Extended mnemonic for
bcla 4,4∗cr_field+3,target
bnsa
bnsl
[cr_field], target
(LR) ← CIA + 4
(LR) ← CIA + 4
bnsla
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Branch if not summary overflow to address in CTR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bnsctr
[cr_field]
bcctr 4,4∗cr_field+3
Extended mnemonic for
bcctrl 4,4∗cr_field+3
Branch if not summary overflow to address in LR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bclr 4,4∗cr_field+3
Extended mnemonic for
bclrl 4,4∗cr_field+3
Branch if not unordered.
(LR) ← CIA + 4
bnsctrl
bnslr
bnslrl
bnu
[cr_field]
(LR) ← CIA + 4
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bc 4,4∗cr_field+3,target
Extended mnemonic for
bca 4,4∗cr_field+3,target
Extended mnemonic for
bcl 4,4∗cr_field+3,target
bnua
bnul
[cr_field], target
(LR) ← CIA + 4
(LR) ← CIA + 4
Extended mnemonic for
bcla 4,4∗cr_field+3,target
bnula
Branch if not unordered to address in CTR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bcctr 4,4∗cr_field+3
Extended mnemonic for
bnuctr
bnuctrl
bnulr
bnulrl
bso
[cr_field]
[cr_field]
(LR) ← CIA + 4
bcctrl 4,4∗cr_field+3
Branch if not unordered to address in LR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bclr 4,4∗cr_field+3
Extended mnemonic for
(LR) ← CIA + 4
bclrl 4,4∗cr_field+3
Branch if summary overflow.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bc 12,4∗cr_field+3,target
Extended mnemonic for
bca 12,4∗cr_field+3,target
Extended mnemonic for
bcl 12,4∗cr_field+3,target
Extended mnemonic for
bcla 12,4∗cr_field+3,target
bsoa
bsol
[cr_field], target
(LR) ← CIA + 4
(LR) ← CIA + 4
bsola
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Mnemonic
Operands
Function
Page
Changed
Branch if summary overflow to address in CTR.
UseCR[CR0] if cr_field is omitted.
Extended mnemonic for
bsoctr
[cr_field]
bcctr 12,4∗cr_field+3
Extended mnemonic for
bcctrl 12,4∗cr_field+3
Branch if summary overflow to address in LR.
UseCR[CR0] if cr_field is omitted.
Extended mnemonic for
(LR) ← CIA + 4
bsoctrl
bsolr
[cr_field]
bclr 12,4∗cr_field+3
Extended mnemonic for
bclrl 12,4∗cr_field+3
(LR) ← CIA + 4
bsolrl
bt
Branch if CR
= 1.
Extendedcrm_bnitemonic for
bc 12,cr_bit,target
Extended mnemonic for
bca 12,cr_bit,target
bta
btl
cr_bit, target
Extended mnemonic for
bcl 12,cr_bit,target
(LR) ← CIA + 4
(LR) ← CIA + 4
Extended mnemonic for
bcla 12,cr_bit,target
btla
Branch if CR
= 1 to address in CTR.
Extendedcrm_bnitemonic for
btctr
bcctr 12,cr_bit
cr_bit
cr_bit
Extended mnemonic for
bcctrl 12,cr_bit
(LR) ← CIA + 4
btctrl
Branch if CR
= 1,
cr_bit
to address in LR.
Extended mnemonic for
bclr 12,cr_bit
btlr
btlrl
bun
Extended mnemonic for
bclrl 12,cr_bit
(LR) ← CIA + 4
Branch if unordered.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bc 12,4∗cr_field+3,target
Extended mnemonic for
bca 12,4∗cr_field+3,target
Extended mnemonic for
bcl 12,4∗cr_field+3,target
Extended mnemonic for
bcla 12,4∗cr_field+3,target
buna
bunl
[cr_field], target
(LR) ← CIA + 4
(LR) ← CIA + 4
bunla
Branch if unordered to address in CTR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bcctr 12,4∗cr_field+3
Extended mnemonic for
bunctr
bunctrl
[cr_field]
(LR) ← CIA + 4
bcctrl 12,4∗cr_field+3
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Branch if unordered,
to address in LR.
Use CR[CR0] if cr_field is omitted.
Extended mnemonic for
bunlr
[cr_field]
bclr 12,4∗cr_field+3
Extended mnemonic for
bclrl 12,4∗cr_field+3
Clear left immediate. (n < 32)
(LR) ← CIA + 4
bunlrl
clrlwi
clrlwi.
n
(RA)
←
0
0:n-1
Extended mnemonic for
RA, RS, n
RA, RS, b, n
RA, RS, n
rlwinm RA,RS,0,n,31
Extended mnemonic for
rlwinm. RA,RS,0,n,31
CR[CR0]
Clear left and shift left immediate.
(n ≤ b < 32)
(RA)
(RA)
(RA)
← (RS)
b–n:31–n
32–n:31
0:b–n–1
b:31
n
←
0
clrlslwi
b–n
←
0
Extended mnemonic for
rlwinm RA,RS,n,b−n,31−n
Extended mnemonic for
rlwinm. RA,RS,n,b−n,31−n
Clear right immediate. (n < 32)
CR[CR0]
CR[CR0]
clrlslwi.
clrrwi
n
(RA)
←
0
32–n:31
Extended mnemonic for
rlwinm RA,RS,0,0,31−n
Extended mnemonic for
rlwinm. RA,RS,0,0,31−n
Compare (RA) to (RB), signed.
Results in CR[CRn], where n = BF.
Compare (RA) to EXTS(IM), signed.
clrrwi.
cmp
BF, 0, RA, RB
BF, 0, RA, IM
BF, 0, RA, RB
BF, 0, RA, IM
cmpi
cmpl
cmpli
Results in CR[CRn], where n = BF.
Compare (RA) to (RB), unsigned.
Results in CR[CRn], where n = BF.
16
Compare (RA) to ( 0 || IM), unsigned.
Results in CR[CRn], where n = BF.
Compare Logical Word.
UseCR[CR0] if BF is omitted.
Extended mnemonic for
cmpl BF,0,RA,RB
[BF,] RA, RB
[BF,] RA, IM
[BF,] RA, RB
cmplw
cmplwi
cmpw
Compare Logical Word Immediate.
UseCR[CR0] if BF is omitted.
Extended mnemonic for
cmpli BF,0,RA,IM
Compare Word.
UseCR[CR0] if BF is omitted.
Extended mnemonic for
cmp BF,0,RA,RB
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Compare Word Immediate.
UseCR[CR0] if BF is omitted.
Extended mnemonic for
[BF,] RA, IM
cmpwi
cmpi BF,0,RA,IM
cntlzw
cntlzw.
Count leading zeros in RS.
Place result in RA.
RA, RS
CR[CR0]
AND bit (CR ) with (CR ).
BA
BB
BT, BA, BB
BT, BA, BB
crand
Place result in CR
.
BT
AND bit (CR ) with ¬(CR ).
BA
BB
crandc
Place result in CR
.
BT
Condition register clear.
Extended mnemonic for
crxor bx,bx,bx
bx
crclr
Equivalence of bit CR
BA
with CR
⊕ CR
.
BB
)
BT, BA, BB
bx, by
creqv
crmove
CR
← ¬(CR
BT
BA
BB
Condition register move.
Extended mnemonic for
cror bx,by,by
NAND bit (CR ) with (CR ).
BA
BB
BT, BA, BB
BT, BA, BB
crnand
crnor
Place result in CR
.
BT
NOR bit (CR ) with (CR ).
BA
BB
Place result in CR
.
BT
Condition register not.
Extended mnemonic for
bx, by
crnot
crnor bx,by,by
OR bit (CR ) with (CR ).
BA
BB
BT, BA, BB
BT, BA, BB
cror
Place result in CR
.
BT
OR bit (CR ) with ¬(CR ).
BA
BB
crorc
Place result in CR
.
BT
Condition register set.
Extended mnemonic for
bx
crset
creqv bx,bx,bx
XOR bit (CR ) with (CR ).
BA
BB
BT, BA, BB
RA, RB
crxor
dcba
dcbf
Place result in CR
.
BT
Treated as a no-op.
Flush (store, then invalidate) the data cache block which con-
tains the effective address (RA|0) + (RB).
RA, RB
Invalidate the data cache block which contains the effective
address (RA|0) + (RB).
RA, RB
RA, RB
RA, RB
RA,RB
dcbi
Store the data cache block which contains the effective address
(RA|0) + (RB).
dcbst
dcbt
Load the data cache block which contains the effective address
(RA|0) + (RB).
Load the data cache block which contains the effective address
(RA|0) + (RB).
dcbtst
Zero the data cache block which contains the effective address
(RA|0) + (RB).
RA, RB
RA, RB
dcbz
Invalidate the data cache array.
dccci
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Read tag and data information from the data cache line selected
using effective address bits 17:26. The effective address is cal-
culated by (RA|0) + (RB).
RT, RA, RB
dcread
Place the data word selected by effective address bits 27:29 in
GPR RT; place the tag information in DCDBTRH and DCDB-
TRL.
divw
CR[CR0]
divw.
divwo
Divide (RA) by (RB), signed.
Place result in RT.
RT, RA, RB
RT, RA, RB
XER[SO, OV]
CR[CR0]
divwo.
XER[SO, OV]
divwu
CR[CR0]
divwu.
divwuo
Divide (RA) by (RB), unsigned.
Place result in RT.
XER[SO, OV]
CR[CR0]
divwuo.
XER[SO, OV]
dlmzb
d ← (RS) || (RB)
i, x, y ← 0
XER[TBC], RA
do while (x < 8) ∧ (y = 0)
x ← x + 1
if d
= 0 then
i:i + 7
y ← 1
else
i ← i + 8
(RA) ← x
RA, RS, RB
XER[TBC] ← x
XER[TBC], RA,
CR[CR0]
dlmzb.
if Rc = 1 then
CR[CR0] ←XER[SO]
3
if y = 1 then
if x < 5 then
CR[CR0]
else
CR[CR0]
← 0b010
← 0b100
← 0b001
0:2
0:2
else
CR[CR0]
0:2
eqv
Equivalence of (RS) with (RB).
(RA) ← ¬((RS) ⊕ (RB))
RA, RS, RB
RA, RS, n, b
CR[CR0]
CR[CR0]
eqv.
E(RxAtr)act and←left (juRsSti)fy immediate. (n > 0)
–
0:n 1
b:b+n–1
–
32
n
(RA)
←
0
extlwi
extlwi.
extrwi
extrwi.
n:31
Extended mnemonic for
rlwinm RA,RS,b,0,n−1
Extended mnemonic for
rlwinm. RA,RS,b,0,n−1
E(RxAtr)act and rig←ht ju(sRtiSfy) immediate. (n > 0)
32–n:31
b:b+n–1
–
32
n
(RA)
←
0
0:31–n
Extended mnemonic for
RA, RS, n, b
rlwinm RA,RS,b+n,32−n,31
Extended mnemonic for
rlwinm. RA,RS,b+n,32−n,31
CR[CR0]
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
extsb
Extend the sign of byte (RS)
.
24:31
RA, RS
Place the result in RA.
CR[CR0]
extsb.
extsh
extsh.
Extend the sign of halfword (RS)
Place the result in RA.
.
16:31
RA, RS
RA, RB
CR[CR0]
Invalidate the instruction cache block which contains the effec-
tive address (RA|0) + (RB).
icbi
Load the instruction cache block which contains the effective
address (RA|0) + (RB).
RA, RB
RA, RB
icbt
Invalidate the instruction cache array.
iccci
Read tag and data information from the instruction cache line
selected using effective address bits 17:26. The effective
address is calculated by (RA|0) + (RB).
RA, RB
icread
Place the instruction selected by effective address bits 27:29 in
ICDBDR; place the tag information in ICDBTRH and ICDBTRL.
I(nRsAe)rt from left←imm(RedSia) te. (n > 0)
b:b+n–1
0:n–1
inslwi
inslwi.
insrwi
insrwi.
Extended mnemonic for
RA, RS, n, b
rlwimi RA,RS,32−b,b,b+n−1
Extended mnemonic for
rlwimi. RA,RS,32−b,b,b+n−1
Insert from right immediate. (n > 0)
CR[CR0]
CR[CR0]
(RA)
← (RS)
32–n:31
b:b+n–1
Extended mnemonic for
RA, RS, n, b
rlwimi RA,RS,32−b−n,b,b+n−1
Extended mnemonic for
rlwimi. RA,RS,32−b−n,b,b+n−1
RT, RA, RB,
CRb
isel
RT ← (RA|0) if CRb = 1, else RT ← (RB)
Synchronize execution context by flushing the prefetch queue.
isync
Load address. (RA ≠ 0)
D is an offset from a base address that is assumed to be (RA).
RT, D(RA)
(RT) ← (RA) + EXTS(D)
la
Extended mnemonic for
addi RT,RA,D
Load byte from EA = (RA|0) + EXTS(D) and pad left with zeroes,
RT, D(RA)
RT, D(RA)
lbz
24
(RT) ←
0 || MS(EA,1).
Load byte from EA = (RA|0) + EXTS(D) and pad left with zeroes,
24
(RT) ←
0 || MS(EA,1).
lbzu
Update the base address,
(RA) ← EA.
Load byte from EA = (RA|0) + (RB) and pad left with zeroes,
24
(RT) ←
0 || MS(EA,1).
RT, RA, RB
lbzux
Update the base address,
(RA) ← EA.
Load byte from EA = (RA|0) + (RB) and pad left with zeroes,
RT, RA, RB
RT, D(RA)
lbzx
lha
24
(RT) ←
0 || MS(EA,1).
Load halfword from EA = (RA|0) + EXTS(D) and sign extend,
(RT) ← EXTS(MS(EA,2)).
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Load halfword from EA = (RA|0) + EXTS(D) and sign extend,
(RT) ← EXTS(MS(EA,2)).
RT, D(RA)
lhau
Update the base address,
(RA) ← EA.
Load halfword from EA = (RA|0) + (RB) and sign extend,
(RT) ← EXTS(MS(EA,2)).
RT, RA, RB
lhaux
Update the base address,
(RA) ← EA.
Load halfword from EA = (RA|0) + (RB) and sign extend,
RT, RA, RB
RT, RA, RB
lhax
(RT) ← EXTS(MS(EA,2)).
Load halfword from EA = (RA|0) + (RB), then reverse byte order
and pad left with zeroes,
lhbrx
16
(RT) ← 0 || MS(EA+1,1) || MS(EA,1).
Load halfword from EA = (RA|0) + EXTS(D) and pad left with
zeroes,
RT, D(RA)
RT, D(RA)
lhz
16
(RT) ←
0 || MS(EA,2).
Load halfword from EA = (RA|0) + EXTS(D) and pad left with
zeroes,
16
(RT) ←
0 || MS(EA,2).
lhzu
Update the base address,
(RA) ← EA.
Load halfword from EA = (RA|0) + (RB) and pad left with zeroes,
16
(RT) ←
0 || MS(EA,2).
RT, RA, RB
RT, RA, RB
RT, IM
lhzux
lhzx
li
Update the base address,
(RA) ← EA.
Load halfword from EA = (RA|0) + (RB) and pad left with zeroes,
16
(RT) ←
0 || MS(EA,2).
Load immediate.
(RT) ← EXTS(IM)
Extended mnemonic for
addi RT,0,value
Load immediate shifted.
16
(RT) ← (IM ||
0)
RT, IM
lis
Extended mnemonic for
addis RT,0,value
Load multiple words starting from EA = (RA|0) + EXTS(D).
Place into consecutive registers RT through GPR(31).
RA is not altered unless RA = GPR(31).
RT, D(RA)
lmw
Load consecutive bytes from EA=(RA|0).
Number of bytes n=32 if NB=0, else n=NB.
Stack bytes into words in CEIL(n/4)
RT, RA, NB
consecutive registers starting with RT, to
RFINAL ← ((RT + CEIL(n/4) – 1) % 32).
lswi
GPR(0) is consecutive to GPR(31).
RA is not altered unless RA = R
.
FINAL
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Load consecutive bytes from EA=(RA|0)+(RB).
Number of bytes n=XER[TBC].
Stack bytes into words in CEIL(n/4)
consecutive registers starting with RT, to
RFINAL ← ((RT + CEIL(n/4) – 1) % 32).
RT, RA, RB
lswx
GPR(0) is consecutive to GPR(31).
RA is not altered unless RA = R
RB is not altered unless RB = R
.
.
FINAL
FINAL
If n=0, content of RT is undefined.
Load word from EA = (RA|0) + (RB) and place in RT,
(RT) ← MS(EA,4).
RT, RA, RB
lwarx
Set the Reservation bit.
Load word from EA = (RA|0) + (RB) then reverse byte order,
RT, RA, RB
RT, D(RA)
lwbrx
lwz
(RT) ← MS(EA+3,1) || MS(EA+2,1) ||
MS(EA+1,1) || MS(EA,1).
Load word from EA = (RA|0) + EXTS(D) and place in RT,
(RT) ← MS(EA,4).
Load word from EA = (RA|0) + EXTS(D) and place in RT,
(RT) ← MS(EA,4).
RT, D(RA)
lwzu
Update the base address,
(RA) ← EA.
Load word from EA = (RA|0) + (RB) and place in RT,
(RT) ← MS(EA,4).
RT, RA, RB
RT, RA, RB
lwzux
lwzx
Update the base address,
(RA) ← EA.
Load word from EA = (RA|0) + (RB) and place in RT,
(RT) ← MS(EA,4).
macchw
macchw.
macchwo
CR[CR0]
prod0:31 ← (RA)16:31 × (RB)
0:15
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
temp0:32 ← prod
+ (RT)
XER[SO, OV]
0:31
(RT) ← temp
1:32
CR[CR0]
macchwo.
XER[SO, OV]
macchwu
macchwu.
macchwuo
CR[CR0]
prod0:31 ← (RA)16:31 × (RB)
0:15
0:15
temp0:32 ← prod
+ (RT)
XER[SO, OV]
0:31
(RT) ← temp
1:32
CR[CR0]
macchwuo.
XER[SO, OV]
macchws
macchws.
macchwso
prod0:31 ← (RA)16:31 × (RB)
temp0:32 ← prod + (RT)
CR[CR0]
0:31
if ((prod = RT ) ∧ (RT ≠ temp )) then
XER[SO, OV]
0
0
0
1
31
(RT) ← (RT ∨ (¬RT ))
0
0
CR[CR0]
macchwso.
else (RT) ← temp
1:32
XER[SO, OV]
macchwsu
macchwsu.
macchwsuo
CR[CR0]
prod0:31 ← (RA)16:31 × (RB)
0:15
temp0:32 ← prod
(RT) ← (temp
+ (RT)
XER[SO, OV]
0:31
31
∨
temp )
1:32
0
CR[CR0]
macchwsuo.
XER[SO, OV]
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
machhw
machhw.
machhwo
CR[CR0]
prod0:31 ← (RA)16:31 × (RB)
0:15
RT, RA, RB
temp0:32 ← prod
+ (RT)
XER[SO, OV]
0:31
(RT) ← temp
1:32
CR[CR0]
machhwo.
XER[SO, OV]
machhwu
machhwu.
machhwuo
CR[CR0]
prod0:31 ← (RA)16:31 × (RB)
0:15
0:15
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
temp0:32 ← prod
+ (RT)
0:31
XER[SO, OV]
(RT) ← temp
1:32
CR[CR0]
machhwuo.
XER[SO, OV]
machhws
machhws.
machhwso
prod0:31 ← (RA)16:31 × (RB)
temp0:32 ← prod + (RT)
CR[CR0]
0:31
if ((prod = RT ) ∧ (RT ≠ temp )) then
XER[SO, OV]
0
0
0
1
31
(RT) ← (RT ∨ (¬RT ))
0
0
CR[CR0]
machhwso.
else (RT) ← temp
1:32
XER[SO, OV]
machhwsu
machhwsu.
machhwsuo
CR[CR0]
prod0:31 ← (RA)16:31 × (RB)
0:15
temp0:32 ← prod
(RT) ← (temp
+ (RT)
XER[SO, OV]
0:31
31
∨
temp )
1:32
0
CR[CR0]
machhwsuo.
XER[SO, OV]
maclhw
maclhw.
maclhwo
CR[CR0]
prod0:31 ← (RA)16:31 × (RB)
0:15
temp0:32 ← prod
+ (RT)
XER[SO, OV]
0:31
(RT) ← temp
1:32
CR[CR0]
maclhwo.
XER[SO, OV]
maclhwu
maclhwu.
maclhwuo
CR[CR0]
prod0:31 ← (RA)16:31 × (RB)
0:15
0:15
temp0:32 ← prod
+ (RT)
XER[SO, OV]
0:31
(RT) ← temp
1:32
CR[CR0]
maclhwuo.
XER[SO, OV]
maclhws
maclhws.
maclhwso
prod0:31 ← (RA)16:31 × (RB)
temp0:32 ← prod + (RT)
CR[CR0]
0:31
if ((prod = RT ) ∧ (RT ≠ temp )) then
XER[SO, OV]
0
0
0
1
31
(RT) ← (RT ∨ (¬RT ))
0
0
CR[CR0]
maclhwso.
else (RT) ← temp
1:32
XER[SO, OV]
maclhwsu
maclhwsu.
maclhwsuo
CR[CR0]
prod0:31 ← (RA)16:31 × (RB)
0:15
temp0:32 ← prod
(RT) ← (temp
+ (RT)
XER[SO, OV]
0:31
31
∨
temp )
1:32
0
CR[CR0]
maclhwsuo.
mbar
XER[SO, OV]
Storage synchronization. All loads and stores that precede the
mbar instruction complete before any loads and stores that fol-
low the instruction access main storage.
Move CR field, (CR[CRn]) ← (CR[CRm])
BF, BFA
BF
mcrf
where m ← BFA and n ← BF
Move XER[0:3] into field CRn, where n←BF.
CR[CRn] ← (XER[SO, OV, CA])
mcrxr
3
(XER[SO, OV, CA]) ←
0
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Move from CR to RT,
RT
mfcr
(RT) ← (CR).
Move from DCR to RT,
RT, DCRN
RT
mfdcr
(RT) ← (DCR(DCRN)).
Move from MSR to RT,
mfmsr
(RT) ← (MSR).
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
mfccr0
mfccr1
mfctr
mfdac1
mfdac2
mfdbcr0
mfdbcr1
mfdbsr
mfdccr
mfdcwr
mfdvc1
mfdvc2
mfesr
mfiac1
mfiac2
mfiac3
mfiac4
mficcr
mficdbdr
mfivpr
mflr
Move from special purpose register (SPR) SPRN.
Extended mnemonic for
mfspr RT,SPRN
mfmcsr
mfmcsrr0
mfmcsrr1
mfpid
RT
mfpit
mfpvr
mfsgr
mfsprg0
mfsprg1
mfsprg2
mfsprg3
mfsprg4
mfsprg5
mfsprg6
mfsprg7
mfsrr0
mfsrr1
mfsrr2
mfsrr3
mfsu0r
mftbl
mftbu
mftcr
mftsr
mfxer
Move from SPR to RT,
RT, SPRN
mfspr
(RT) ← (SPR(SPRN)).
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Move register.
(RT) ← (RS)
mr
Extended mnemonic for
RT, RS
or RT,RS,RS
Extended mnemonic for
or. RT,RS,RS
CR[CR0]
mr.
Synchronization. All instructions that precede msync complete
before any instructions that follow msync begin.
When msync completes, all storage accesses initiated prior to
msync will have completed.
msync
Move to Condition Register.
Extended mnemonic for
RS
mtcr
mtcrf 0xFF,RS
Move some or all of the contents of RS into CR as specified by
FXM field,
4
4
4
4
FXM, RS
mtcrf
mask ← (FXM ) || (FXM ) || ... ||
0
6
1
7
(FXM ) || (FXM ).
(CR)←((RS) ∧ mask) ∨ (CR) ∧ ¬mask).
Move to DCR from RS,
(DCR(DCRN)) ← (RS).
Move to MSR from RS,
DCRN, RS
RS
mtdcr
mtmsr
(MSR) ← (RS).
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
mtccr0
mtccr1
mtctr
mtdac1
mtdac2
mtdbcr0
mtdbcr1
mtdbsr
mtdccr
mtdcwr
mtdvc1
mtdvc2
mtesr
mtiac1
mtiac2
mtiac3
mtiac4
mticcr
mticdbdr
mtivpr
mtlr
Move to SPR SPRN.
Extended mnemonic for
mtspr SPRN,RS
mtmcsr
mtmcsrr0
mtmcsrr1
mtpid
RS
mtpit
mtpvr
mtsgr
mtsprg0
mtsprg1
mtsprg2
mtsprg3
mtsprg4
mtsprg5
mtsprg6
mtsprg7
mtsrr0
mtsrr1
mtsrr2
mtsrr3
mtsu0r
mttbl
mttbu
mttcr
mttsr
mtxer
mtzpr
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Move to SPR from RS,
SPRN, RS
mtspr
(SPR(SPRN)) ← (RS).
mulchw
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
(RT) :31 ← (RA)16:31 × (RB)
(signed)
0:15
0
mulchw.
mulchwu
mulchwu.
mulhhw
CR[CR0]
(RT) :31 ← (RA)16:31 × (RB)
(unsigned)
0:15
0
CR[CR0]
CR[CR0]
CR[CR0]
(RT)
← (RA)0:15 × (RB
(signed)
:
0:15
0 31
mulhhw.
mulhhwu
mulhhwu.
mulhw
(RT) :31 ← (RA)0:15 × (RB)
(unsigned)
0:15
0
Multiply (RA) and (RB), signed.
Place high-order result in RT.
RT, RA, RB
RT, RA, RB
prod
(RT) ← prod
← (RA) × (RB) (signed).
CR[CR0]
CR[CR0]
mulhw.
0:63
0:31.
Multiply (RA) and (RB), unsigned.
mulhwu
mulhwu.
Place high-order result in RT.
prod
← (RA) × (RB) (unsigned).
0:63
(RT) ← prod
0:31.
mullhw
RT, RA, RB
RT, RA, RB
(RT)
← (RA)16:31 × (RB
(signed)
:
16:31
16:31
0 31
mullhw.
mullhwu
mullhwu.
CR[CR0]
CR[CR0]
(RT)16:31 ← (RA)0:15 × (RB)
(unsigned)
Multiply (RA) and IM, signed.
Pprlaocde low-←orde(rRrAes)u×lt inIMR(Ts.igned)
RT, RA, IM
mulli
0:47
(RT) ← prod
16:47
mullw
mullw.
mullwo
Multiply (RA) and (RB), signed.
Place low-order result in RT.
CR[CR0]
RT, RA, RB
RA, RS, RB
RT, RA
XER[SO, OV]
prod
← (RA) × (RB) (signed).
0:63
(RT) ← prod
CR[CR0]
32:63.
mullwo.
XER[SO, OV]
nand
nand.
neg
NAND (RS) with (RB).
Place result in RA.
CR[CR0]
CR[CR0]
neg.
nego
Negative (two’s complement) of RA.
XER[SO, OV]
(RT) ← ¬(RA) + 1
CR[CR0]
nego.
XER[SO, OV]
nmacchw
nmacchw.
nmacchwo
CR[CR0]
prod0:31 ← (RA)16:31 × (RB)
0:15
RT, RA, RB
temp0:32 ← –prod
+ (RT)
XER[SO, OV]
0:31
(RT) ← temp
1:32
CR[CR0]
nmacchwo.
XER[SO, OV]
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Page
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
prod0:31 ← (RA)16:31 × (RB)
nmacchws
0:15
nmacchws.
nmacchwso
CR[CR0]
temp0:32 ← –prod
+ (RT)
0:31
RT, RA, RB
if ((prod = RT ) ∧ (RT ≠ temp )) then
XER[SO, OV]
0
0
0
1
31
(RT) ← (RT ∨ (¬RT ))
0
0
CR[CR0]
nmacchwso.
else (RT) ← temp
1:32
XER[SO, OV]
nmachhw
nmachhw.
nmachhwo
CR[CR0]
prod0:31 ← (RA)0:15 × (RB)
0:15
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
temp0:32 ← –prod
+ (RT)
XER[SO, OV]
0:31
(RT) ← temp
1:32
CR[CR0]
nmachhwo.
XER[SO, OV]
nmachhws
nmachhws.
nmachhwso
prod0:31 ← (RA)0:15 × (RB)
temp0:32 ← –prod + (RT)
0:15
CR[CR0]
0:31
if ((prod = RT ) ∧ (RT ≠ temp )) then
XER[SO, OV]
0
0
0
1
31
(RT) ← (RT ∨ (¬RT ))
0
0
CR[CR0]
nmachhwso.
else (RT) ← temp
1:32
XER[SO, OV]
nmaclhw
nmaclhw.
nmaclhwo
CR[CR0]
prod0:31 ← (RA16:31 × (RB)
16:31
temp0:32 ← –prod
+ (RT)
XER[SO, OV]
0:31
(RT) ← temp
1:32
CR[CR0]
nmaclhwo.
XER[SO, OV]
nmaclhws
nmaclhws.
nmaclhwso
prod0:31 ← (RA)16:31 × (RB)
16:31
CR[CR0]
temp0:32 ← –prod
+ (RT)
0:31
if ((prod = RT ) ∧ (RT ≠ temp )) then
XER[SO, OV]
0
0
0
1
31
(RT) ← (RT ∨ (¬RT ))
0
0
CR[CR0]
nmaclhwso.
else (RT) ← temp
1:32
XER[SO, OV]
Preferred no-op, triggers optimizations based on no-ops.
Extended mnemonic for
ori 0,0,0
nop
nor
NOR (RS) with (RB).
Place result in RA.
RA, RS, RB
RA, RS
CR[CR0]
nor.
Complement register.
(RA) ← ¬(RS)
not
Extended mnemonic for
nor RA,RS,RS
Extended mnemonic for
nor. RA,RS,RS
CR[CR0]
not.
or
OR (RS) with (RB).
Place result in RA.
RA, RS, RB
RA, RS, RB
CR[CR0]
CR[CR0]
or.
orc
orc.
OR (RS) with ¬(RB).
Place result in RA.
16
OR (RS) with ( 0 || IM).
RA, RS, IM
RA, RS, IM
ori
Place result in RA.
16
OR (RS) with (IM ||
0).
oris
Place result in RA.
Return from critical interrupt
rfci
(PC) ← (CSRR0).
(MSR) ← (CSRR1).
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Return from interrupt.
rfi
(PC) ← (SRR0).
(MSR) ← (SRR1).
Return from machine check interrupt
rfmci
(PC) ← (MCSRR0).
(MSR) ← (MCSRR1).
Rotate left word immediate, then insert according to mask.
rlwimi
rlwimi.
rlwinm
rlwinm.
rlwnm
rlwnm.
RA, RS, SH,
MB, ME
r ← ROTL((RS), SH)
m ← MASK(MB, ME)
CR[CR0]
(RA) ← (r ∧ m) ∨ ((RA) ∧ ¬m)
Rotate left word immediate, then AND with mask.
RA, RS, SH,
MB, ME
r ← ROTL((RS), SH)
m ← MASK(MB, ME)
(RA) ← (r ∧ m)
CR[CR0]
CR[CR0]
Rotate left word, then AND with mask.
RA, RS, RB,
MB, ME
r ← ROTL((RS), (RB)
m ← MASK(MB, ME)
(RA) ← (r ∧ m)
)
27:31
Rotate left.
(RA) ← ROTL((RS), (RB)
)
27:31
rotlw
Extended mnemonic for
RA, RS, RB
RA, RS, n
RA, RS, n
rlwnm RA,RS,RB,0,31
Extended mnemonic for
rlwnm. RA,RS,RB,0,31
CR[CR0]
CR[CR0]
CR[CR0]
rotlw.
rotlwi
rotlwi.
rotrwi
rotrwi.
Rotate left immediate.
(RA) ← ROTL((RS), n)
Extended mnemonic for
rlwinm RA,RS,n,0,31
Extended mnemonic for
rlwinm. RA,RS,n,0,31
Rotate right immediate.
(RA) ← ROTL((RS), 32−n)
Extended mnemonic for
rlwinm RA,RS,32−n,0,31
Extended mnemonic for
rlwinm. RA,RS,32−n,0,31
System call exception is generated.
(SRR1) ← (MSR)
sc
(SRR0) ← (PC)
PC ← EVPR
|| 0x0C00
0:15
(MSR[WE, PR, EE, PE, DR, IR]) ← 0
Shift left (RS) by (RB)
.
slw
27:31
n ← (RB)
27:31.
r ← ROTL((RS), n).
RA, RS, RB
if (RB)
= 0 then m ← MASK(0, 31 – n)
32
CR[CR0]
slw.
26
else m ←
(RA) ← r ∧ 0m.
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
S(RhAift) left imm←edia(RteS. )(n < 32)
32-n:31
Extended mnemonic for
0:31-n
n:31
n
(RA)
←
0
slwi
RA, RS, n
rlwinm RA,RS,n,0,31−n
Extended mnemonic for
rlwinm. RA,RS,n,0,31−n
CR[CR0]
slwi.
sraw
Shift right algebraic (RS) by (RB)
.
27:31
n ← (RB)
27:31.
r ← ROTL((RS), 32 – n).
if (RB)
= 0 then m ← MASK(n, 31)
32
26
RA, RS, RB
else m ←
0
CR[CR0]
sraw.
s ← (RS)
0
32
(RA) ← (r ∧ m) ∨ (
XER[CA] ← s ∧ ((r ∧s¬∧m)¬≠m0))..
Shift right algebraic (RS) by SH.
srawi
n ← SH.
r ← ROTL((RS), 32 – n).
m ← MASK(n, 31).
RA, RS, SH
RA, RS, RB
CR[CR0]
CR[CR0]
srawi.
s ← (RS)
0
32
(RA) ← (r ∧ m) ∨ (
XER[CA] ← s ∧ ((r ∧s¬∧m)¬≠0m))..
Shift right (RS) by (RB)
.
srw
27:31
n ← (RB)
27:31.
r ← ROTL((RS), 32 – n).
if (RB)
26
= 0 then m ← MASK(n, 31)
srw.
32
else m ←
0
(RA) ← r ∧ m.
Shift right immediate. (n < 32)
(RA)
(RA)
← (RS)
n:31
0:n-1
0:31-n
n
←
0
srwi
Extended mnemonic for
RA, RS, n
rlwinm RA,RS,32−n,n,31
Extended mnemonic for
rlwinm. RA,RS,32−n,n,31
CR[CR0]
srwi.
stb
Store byte (RS)
in memory at
24:31
EA = (RA|0) + EXTS(D).
RS, D(RA)
RS, D(RA)
Store byte (RS) in memory at
24:31
EA = (RA|0) + EXTS(D).
Update the base address,
(RA) ← EA.
stbu
Store byte (RS)
in memory at
24:31
EA = (RA|0) + (RB).
Update the base address,
(RA) ← EA.
24:31
EA = (RA|0) + (RB).
RS, RA, RB
stbux
Store byte (RS)
in memory at
RS, RA, RB
RS, D(RA)
stbx
sth
Store halfword (RS)
in memory at
16:31
EA = (RA|0) + EXTS(D).
Store halfword (RS)
16:31
byte-reversed in memory at EA =
(RA|0) + (RB).
RS, RA, RB
sthbrx
MS(EA, 2) ← (RS)
|| (RS)
24:31
16:23
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Store halfword (RS)
16:31
in memory at
EA = (RA|0) + EXTS(D).
Update the base address,
(RA) ← EA.
RS, D(RA)
sthu
Store halfword (RS)
in memory at
in memory at
16:31
EA = (RA|0) + (RB).
Update the base address,
(RA) ← EA.
RS, RA, RB
sthux
Store halfword (RS)
16:31
RS, RA, RB
RS, D(RA)
sthx
EA = (RA|0) + (RB).
Store consecutive words from RS through GPR(31) in memory
starting at
stmw
EA = (RA|0) + EXTS(D).
Store consecutive bytes in memory starting at EA=(RA|0).
Number of bytes n=32 if NB=0, else n=NB.
RS, RA, NB
RS, RA, RB
stswi
Bytes are unstacked from CEIL(n/4)
consecutive registers starting with RS.
GPR(0) is consecutive to GPR(31).
Store consecutive bytes in memory starting at EA=(RA|0)+(RB).
Number of bytes n=XER[TBC].
stswx
Bytes are unstacked from CEIL(n/4)
consecutive registers starting with RS.
GPR(0) is consecutive to GPR(31).
Store word (RS) in memory at
EA = (RA|0) + EXTS(D).
RS, D(RA)
stw
Store word (RS) byte-reversed in memory at EA = (RA|0) +
(RB).
RS, RA, RB
stwbrx
MS(EA, 4) ← (RS)
|| (RS)
||
24:31
|| (RS)
0:7
16:23
(RS)
8:15
Store word (RS) in memory at EA = (RA|0) + (RB)
only if reservation bit is set.
if RESERVE = 1 then
MS(EA, 4) ← (RS)
RESERVE ← 0
RS, RA, RB
RS, D(RA)
stwcx.
stwu
2
(CR[CR0]) ← 0 || 1 || XER
so
else
2
(CR[CR0]) ← 0 || 0 || XER
so.
Store word (RS) in memory at
EA = (RA|0) + EXTS(D).
Update the base address,
(RA) ← EA.
Store word (RS) in memory at
EA = (RA|0) + (RB).
RS, RA, RB
RS, RA, RB
stwux
stwx
Update the base address,
(RA) ← EA.
Store word (RS) in memory at
EA = (RA|0) + (RB).
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Subtract (RB) from (RA).
(RT) ← ¬(RB) + (RA) + 1.
sub
Extended mnemonic for
subf RT,RB,RA
Extended mnemonic for
subf. RT,RB,RA
CR[CR0]
sub.
RT, RA, RB
Extended mnemonic for
subfo RT,RB,RA
XER[SO, OV]
subo
subo.
Extended mnemonic for
subfo. RT,RB,RA
CR[CR0]
XER[SO, OV]
Subtract (RB) from (RA).
(RT) ← ¬(RB) + (RA) + 1.
Place carry-out in XER[CA].
Extended mnemonic for
subc
subfc RT,RB,RA
Extended mnemonic for
subfc. RT,RB,RA
RT, RA, RB
CR[CR0]
subc.
Extended mnemonic for
subfco RT,RB,RA
XER[SO, OV]
subco
subco.
Extended mnemonic for
subfco. RT,RB,RA
CR[CR0]
XER[SO, OV]
subf
CR[CR0]
subf.
subfo
Subtract (RA) from (RB).
RT, RA, RB
RT, RA, RB
XER[SO, OV]
(RT) ← ¬(RA) + (RB) + 1.
CR[CR0]
subfo.
XER[SO, OV]
subfc
CR[CR0]
subfc.
subfco
Subtract (RA) from (RB).
(RT) ← ¬(RA) + (RB) + 1.
Place carry-out in XER[CA].
XER[SO, OV]
CR[CR0]
subfco.
XER[SO, OV]
subfe
CR[CR0]
subfe.
subfeo
Subtract (RA) from (RB) with carry-in.
(RT) ← ¬(RA) + (RB) + XER[CA].
Place carry-out in XER[CA].
RT, RA, RB
RT, RA, IM
RT, RA, RB
XER[SO, OV]
CR[CR0]
subfeo.
XER[SO, OV]
Subtract (RA) from EXTS(IM).
(RT) ← ¬(RA) + EXTS(IM) + 1.
Place carry-out in XER[CA].
subfic
subfme
subfme.
subfmeo
CR[CR0]
Subtract (RA) from (–1) with carry-in.
(RT) ← ¬(RA) + (–1) + XER[CA].
Place carry-out in XER[CA].
XER[SO, OV]
CR[CR0]
subfmeo.
XER[SO, OV]
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
subfze
CR[CR0]
subfze.
subfzeo
Subtract (RA) from zero with carry-in.
(RT) ← ¬(RA) + XER[CA].
Place carry-out in XER[CA].
RT, RA, RB
XER[SO, OV]
CR[CR0]
subfzeo.
XER[SO, OV]
Subtract EXTS(IM) from (RA|0).
Place result in RT.
Extended mnemonic for
addi RT,RA,−IM
Subtract EXTS(IM) from (RA).
Place result in RT.
Place carry-out in XER[CA].
Extended mnemonic for
RT, RA, IM
RT, RA, IM
subi
subic
addic RT,RA,−IM
Subtract EXTS(IM) from (RA).
Place result in RT.
RT, RA, IM
RT, RA, IM
Place carry-out in XER[CA].
Extended mnemonic for
CR[CR0]
subic.
subis
addic. RT,RA,−IM
16
Subtract (IM ||
0) from (RA|0).
Place result in RT.
Extended mnemonic for
addis RT,RA,−IM
tlbentry ← TLB[(RA)
]
26:31
if WS = 0
(RT)0:27 ← tlbentry[EPN,V,TS,SIZE]
4
(RT)28:31 ← 0
MMUCR[STID] ← tlbentry[TID]
else if WS = 1
(RT)0:21 ← tlbentry[RPN]
6
RT, RA,WS
(RT)22:27 ← 0
tlbre
(RT)28:31 ← tlbentry[ERPN]
else if WS = 2
16
(RT)
←
0
0:15
(RT)16:24 ← tlbentry[U0,U1,U2,U3,W,I,M,G,E]
(RT)25 ← 0
(RT)26:31 ← tlbentry[UX,UW,UR,SX,SW,SR]
else (RT), MMUCR[STID] ← undefined
Search the TLB for a valid entry that translates the EA.
tlbsx
EA = (RA|0) + (RB)
if Rc = 1
CR[CR0] ← 0
0
CR[CR0] ← 0
1
CR[CR0] ← XER[SO}
3
if Valid TLB entry matching EA and MMUCR[STID,STS] is in
the TLB then
RT,RA,RB
CR[CR0]
tlbsx.
(RT) ← Index of matching TLB Entry
if Rc = 1
CR[CR0] ← 1
2
else
(RT) ← Undefined
if Rc = 1
CR[CR0] ← 0
2
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
tlbsync does not complete until all previous TLB-update instruc-
tions executed by this processor have been received and com-
pleted by all other processors.
tlbsync
For the PPC440x5 core, tlbsync is a no-op.
tlbentry ← TLB[(RA)
]
26:31
if WS = 0
tlbentry[EPN,V,TS,SIZE] ← (RS)
0:27
tlbentry[TID] ← MMUCR[STID]
else if WS = 1
RS, RA,WS
tlbwe
tlbentry[RPN] ← (RS)
0:21
tlbentry[ERPN] ← (RS)
28:31
else if WS = 2
tlbentry[U0,U1,U2,U3,W,I,M,G,E] ← (RS)
16:24
26:31
tlbentry[UX,UW,UR,SX,SW,SR] ← (RS)
else tlbentry ← undefined
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Trap unconditionally.
Extended mnemonic for
trap
tw 31,0,0
Trap if (RA) equal to (RB).
Extended mnemonic for
tw 4,RA,RB
tweq
twge
twgt
twle
Trap if (RA) greater than or equal to (RB).
Extended mnemonic for
tw 12,RA,RB
Trap if (RA) greater than (RB).
Extended mnemonic for
tw 8,RA,RB
Trap if (RA) less than or equal to (RB).
Extended mnemonic for
tw 20,RA,RB
Trap if (RA) logically greater than or equal to (RB).
Extended mnemonic for
tw 5,RA,RB
twlge
twlgt
twlle
twllt
Trap if (RA) logically greater than (RB).
Extended mnemonic for
tw 1,RA,RB
Trap if (RA) logically less than or equal to (RB).
Extended mnemonic for
tw 6,RA,RB
RA, RB
Trap if (RA) logically less than (RB).
Extended mnemonic for
tw 2,RA,RB
Trap if (RA) logically not greater than (RB).
Extended mnemonic for
tw 6,RA,RB
twlng
twlnl
twlt
Trap if (RA) logically not less than (RB).
Extended mnemonic for
tw 5,RA,RB
Trap if (RA) less than (RB).
Extended mnemonic for
tw 16,RA,RB
Trap if (RA) not equal to (RB).
Extended mnemonic for
tw 24,RA,RB
twne
twng
Trap if (RA) not greater than (RB).
Extended mnemonic for
tw 20,RA,RB
Trap if (RA) not less than (RB).
Extended mnemonic for
tw 12,RA,RB
twnl
tw
Trap exception is generated if, comparing (RA) with (RB), any
condition specified by TO is true.
TO, RA, RB
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
Function
Page
Trap if (RA) equal to EXTS(IM).
Extended mnemonic for
tweqi
wi 4,RA,IM
Trap if (RA) greater than or equal to EXTS(IM).
Extended mnemonic for
twi 12,RA,IM
twgei
twgti
twlei
Trap if (RA) greater than EXTS(IM).
Extended mnemonic for
twi 8,RA,IM
Trap if (RA) less than or equal to EXTS(IM).
Extended mnemonic for
twi 20,RA,IM
Trap if (RA) logically greater than or equal to EXTS(IM).
Extended mnemonic for
wi 5,RA,IM
twlgei
twlgti
twllei
twllti
twlngi
twlnli
twlti
Trap if (RA) logically greater than EXTS(IM).
Extended mnemonic for
twi 1,RA,IM
Trap if (RA) logically less than or equal to EXTS(IM).
Extended mnemonic for
twi 6,RA,IM
RA, IM
Trap if (RA) logically less than EXTS(IM).
Extended mnemonic for
twi 2,RA,IM
Trap if (RA) logically not greater than EXTS(IM).
Extended mnemonic for
twi 6,RA,IM
Trap if (RA) logically not less than EXTS(IM).
Extended mnemonic for
twi 5,RA,IM
Trap if (RA) less than EXTS(IM).
Extended mnemonic for
twi 16,RA,IM
Trap if (RA) not equal to EXTS(IM).
Extended mnemonic for
twi 24,RA,IM
twnei
twngi
Trap if (RA) not greater than EXTS(IM).
Extended mnemonic for
twi 20,RA,IM
Trap if (RA) not less than EXTS(IM).
Extended mnemonic for
twi 12,RA,IM
twnli
twi
Trap exception is generated if, comparing (RA) with EXTS(IM),
any condition specified by TO is true.
TO, RA, IM
RS
E
Write value of RS to MSR[EE].
16
wrtee
wrteei
xor
Write value of E to MSR[EE].
XOR (RS) with (RB).
Place result in RA.
RA, RS, RB
CR[CR0]
xor.
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-1. PPC440x5 Instruction Syntax Summary (continued)
Other Registers
Changed
Mnemonic
Operands
RA, RS, IM
RA, RS, IM
Function
Page
16
XOR (RS) with ( 0 || IM).
xori
xoris
Place result in RA.
16
XOR (RS) with (IM ||
0).
Place result in RA.
A.3 Allocated Instruction Opcodes
Allocated instructions are provided for purposes that are outside the scope of PowerPC Book-E architecture,
and are for implementation-dependent and application-specific use, including use within auxiliary processors.
the table, the character “u” designates a secondary opcode bit which can be set to any value. In some cases,
the decimal value of a secondary opcode is shown in parentheses after the binary value.
Table A-2. Allocated Opcodes
Primary
Opcode
Extended
Opcodes
PPC440x5
Usage
All instruction encodings (bits 6:31) except 0x00000000
0
None
(the instruction encoding of 0x00000000 is and always will be
reserved-illegal)
4
All instruction encodings (bits 6:31)
19
Secondary opcodes (bits 21:30) = 0buuuuu0u11u
None
Secondary opcodes (bits 21:30) = 0buuuuu0011u
Secondary opcodes (bits 21:30) = 0buuuuu0u110
Secondary opcode (bits 21:30) = 0b0101010110 (342)
Secondary opcode (bits 21:30) = 0b0101110110 (374)
Secondary opcode (bits 21:30) = 0b1100110110 (822)
31
59
63
Secondary opcodes (bits 21:30) = 0buuuuu0u10u
None
None
Secondary opcodes (bits 21:30) = 0buuuuu0u10u (except second-
ary opcode decimal 12, which is the fsrp defined instruction)
All of the allocated opcodes listed in the table above are available for use by auxiliary processors attached to
the PPC440x5, except for those which have already been implemented within the PPC440x5 core for certain
implementation-specific purposes. As indicated in the table above, this is the case for certain secondary
opcodes within primary opcodes 4 and 31. These opcodes are identified in Table A-5 on page 559, along with
all of the defined, preserved, and reserved-nop class opcodes which are implemented within the PPC440x5
core.
A.4 Preserved Instruction Opcodes
The preserved instruction class is provided to support backward compatibility with the PowerPC Architecture,
and/or earlier versions of the PowerPC Book-E architecture. This instruction class includes opcodes which
were defined for these previous architectures, but which are no longer defined for PowerPC Book-E.
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
opcode is shown in parentheses after the binary value.
Table A-3. Preserved Opcodes
Primary
Opcode
Extended
Opcode
Preserved
Mnemonic
PPC440x5
Usage
31
31
31
31
31
31
31
31
31
0b0011010010 (210)
0b0011110010 (242)
0b0101110010 (370)
0b0100110010 (306)
0b0101110011 (371)
0b1001010011 (595)
0b1010010011 (659)
0b0100110110 (310)
0b0110110110 (438)
mtsr
mtsrin
tlbia
tlbie
mftb
Yes
mfsr
mfsrin
eciwx
ecowx
As indicated in the table above, the only preserved opcode which is implemented within the PPC440x5 core
is the mftb instruction. See Preserved Instruction Class on page 55 for more information on PPC440x5
support for this instruction. All other preserved instructions are treated as reserved by PPC440x5 and will
cause Illegal Instruction exception type Program interrupts if their execution is attempted.
The preserved opcode for mftb is included in Table A-5 on page 559, along with all of the defined, allocated,
and reserved-nop class opcodes which are implemented within the PPC440x5 core.
A.5 Reserved Instruction Opcodes
This class of instructions consists of all instruction primary opcodes (and associated extended opcodes, if
applicable) which do not belong to either the defined, allocated, or preserved instruction classes.
Reserved instructions are available for future versions of PowerPC Book-E architecture. That is, future
versions of PowerPC Book-E may define any of these instructions to perform new functions or make them
available for implementation-dependent use as allocated instructions. There are two types of reserved
instructions: reserved-illegal and reserved-nop.
designates a secondary opcode bit which can be set to any value. All other reserved opcodes are in the
reserved-illegal class.
Table A-4. Reserved-nop Opcodes
Primary
Opcode
Extended
Opcode
31
0b10uuu10010
As shown in the table, there are a total of eight (8) secondary opcodes in the reserved-nop class. The
PPC440x5 implements all of the reserved-nop instruction opcodes as true no-ops. These opcodes are
included in Table A-5 on page 559, along with all of the defined, allocated, and preserved class opcodes
which are implemented within the PPC440x5 core.
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
A.6 Implemented Instructions Sorted by Opcode
Table A-5 on page 559 lists all of the instructions which have been implemented within the PPC440x5 core,
sorted by primary and secondary opcode. These include defined, allocated, preserved, and reserved-nop
class instructions (see Instruction Classes on page 53 for a more detailed description of each of these
instruction classes). Opcodes which are not implemented in the PPC440x5 core are not shown in the table,
and consist of the following:
• Defined instructions
These include the floating-point operations (which may be implemented in an auxiliary processor and
executed via the AP interface), as well as the 64-bit operations and the tlbiva and mfapidi instructions,
all of which are handled as reserved-illegal instructions by the PPC440x5.
• Allocated instructions
These include all of the allocated opcodes identified in Table A-2 on page 557 which are not already
implemented within the PPC440x5 core. If not implemented within an attached auxiliary processor, these
instructions will be handled as reserved-illegal by the PPC440x5.
• Preserved instructions
reserved-illegal by the PPC440x5.
• Reserved instructions
reserved-nop opcodes identified in Table A-4 on page 558. These instructions by definition are all in the
reserved-illegal class and will be handled as such by the PPC440x5.
All PowerPC Book-E instructions are four bytes long and word aligned. All instructions have a primary opcode
field (shown as field OPCD in Figure A-1 through Figure A-9, beginning on page 522) in bits 0:5. Some
The “Form” indicated in the table refers to the arrangement of valid field combinations within the four-byte
Form X has a 10-bit secondary opcode field, while form XO uses only the low-order 9-bits of that field.
Form XO uses the high-order secondary opcode bit (the tenth bit) as a variable; therefore, every form XO
instruction really consumes two secondary opcodes from the 10-bit secondary-opcode space. The implicitly
consumed secondary opcode is listed in parentheses for form XO instructions in the table below.
Table A-5. PPC440x5 Instructions by Opcode
Primary
Opcode
Secondary
Opcode
Form
Mnemonic
Operands
TO, RA, IM
Page
3
D
X
twi
mulhhwu
4
4
8
RT, RA, RB
mulhhwu.
machhwu
machhwu.
machhwuo
machhwuo.
12 (524)
XO
RT, RA, RB
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-5. PPC440x5 Instructions by Opcode (continued)
Primary
Opcode
Secondary
Opcode
Form
Mnemonic
Operands
Page
mulhhw
4
40
X
RT, RA, RB
mulhhw.
machhw
machhw.
4
4
4
4
44 (556)
46 (558)
76 (588)
108 (620)
XO
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
machhwo
machhwo.
nmachhw
nmachhw.
nmachhwo
nmachhwo.
machhwsu
machhwsu.
machhwsuo
machhwsuo.
machhws
machhws.
machhwso
machhwso.
nmachhws
nmachhws.
nmachhwso
nmachhwso.
mulchwu
XO
XO
XO
4
4
4
4
4
110 (622)
136
XO
X
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
mulchwu.
macchwu
macchwu.
macchwuo
machhwuo.
mulchw
140 (652)
168
XO
X
mulchw.
macchw
macchw.
172 (684)
XO
macchwo
macchwo.
nmacchw
nmacchw.
nmacchwo
nmacchwo.
macchwsu
macchwsu.
macchwsuo
macchwsuo.
4
4
174 (686)
204 (716)
XO
XO
RT, RA, RB
RT, RA, RB
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-5. PPC440x5 Instructions by Opcode (continued)
Primary
Opcode
Secondary
Opcode
Form
Mnemonic
Operands
Page
macchws
macchws.
macchwso
macchwso.
nmacchws
nmacchws.
nmacchwso
nmacchwso.
mullhwu
4
236 (748)
XO
RT, RA, RB
4
4
4
4
4
238 (750)
392
XO
X
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
mullhwu.
maclhwu
maclhwu.
maclhwuo
maclhwuo.
mullhw
396 (908)
424
XO
X
mullhw.
maclhw
maclhw.
428 (940)
XO
maclhwo
maclhwo.
nmaclhw
nmaclhw.
nmaclhwo
nmaclhwo.
maclhwsu
maclhwsu.
maclhwsuo
maclhwsuo.
maclhws
4
4
4
4
430 (942)
460 (972)
492 (1004)
494 (1006)
XO
XO
XO
XO
RT, RA, RB
RT, RA, RB
RT, RA, RB
RT, RA, RB
maclhws.
maclhwso
maclhwso.
nmaclhws
nmaclhws.
nmaclhwso
nmaclhwso.
mulli
7
D
D
D
D
D
D
D
D
RT, RA, IM
RT, RA, IM
BF, 0, RA, IM
BF, 0, RA, IM
RT, RA, IM
RT, RA, IM
RT, RA, IM
RT, RA, IM
8
subfic
cmpli
cmpi
10
11
12
13
14
15
addic
addic.
addi
addis
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-5. PPC440x5 Instructions by Opcode (continued)
Primary
Opcode
Secondary
Opcode
Form
Mnemonic
Operands
Page
bc
bca
16
B
BO, BI, target
bcl
bcla
sc
17
18
SC
I
b
ba
target
bl
bla
19
19
0
XL
XL
BF, BFA
BO, BI
mcrf
bclr
16
bclrl
crnor
rfmci
19
19
19
19
19
19
19
19
19
19
19
19
33
XL
XL
XL
XL
XL
XL
XL
XL
XL
XL
XL
XL
BT, BA, BB
38
50
rfi
51
rfci
129
150
193
225
257
289
417
449
BT, BA, BB
crandc
isync
crxor
crnand
crand
creqv
crorc
cror
BT, BA, BB
BT, BA, BB
BT, BA, BB
BT, BA, BB
BT, BA, BB
BT, BA, BB
bcctr
bcctrl
rlwimi
rlwimi.
rlwinm
rlwinm.
rlwnm
rlwnm.
ori
19
20
21
23
528
XL
M
BO, BI
RA, RS, SH, MB, ME
RA, RS, SH, MB, ME
RA, RS, RB, MB, ME
M
M
24
25
26
27
28
29
31
31
D
D
D
D
D
D
X
X
RA, RS, IM
RA, RS, IM
RA, RS, IM
RA, RS, IM
RA, RS, IM
RA, RS, IM
BF, 0, RA, RB
TO, RA, RB
oris
xori
xoris
andi.
andis.
cmp
tw
0
4
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-5. PPC440x5 Instructions by Opcode (continued)
Primary
Opcode
Secondary
Opcode
Form
Mnemonic
Operands
Page
subfc
subfc.
subfco
subfco.
addc
31
8 (520)
XO
RT, RA, RB
addc.
addco
addco.
mulhwu
mulhwu.
isel
31
31
10 (522)
11 (523)
XO
XO
RT, RA, RB
RT, RA, RB
31
31
31
31
31
15
19
20
22
23
XO
X
RT, RA, RB
RT
mfcr
X
RT, RA, RB
RA, RB
lwarx
icbt
X
X
RT, RA, RB
lwzx
slw
31
31
24
26
X
X
RA, RS, RB
RA, RS
slw.
cntlzw
cntlzw.
and
31
31
28
32
X
X
RA, RS, RB
and.
BF, 0, RA, RB
cmpl
subf
subf.
subfo
subfo.
dcbst
lwzux
andc
31
40 (552)
XO
RT, RA, RB
31
31
54
55
X
X
RA, RB
RT, RA, RB
31
31
31
60
X
RA, RS, RB
RT, RA, RB
RA, RS, RB
andc.
mulhw
mulhw.
dlmzb
dlmzb.
mfmsr
dcbf
75 (587)
78
XO
X
31
31
31
83
86
87
X
X
X
RT
RA, RB
RT, RA, RB
lbzx
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-5. PPC440x5 Instructions by Opcode (continued)
Primary
Opcode
Secondary
Opcode
Form
Mnemonic
Operands
Page
neg
neg.
31
104 (616)
XO
RT, RA
nego
nego.
31
31
31
119
124
131
X
X
X
RT, RA, RB
RA, RS, RB
RS
lbzux
nor
nor.
wrtee
subfe
subfe.
subfeo
subfeo.
adde
31
31
136 (648)
138 (650)
XO
XO
RT, RA, RB
RT, RA, RB
adde.
addeo
addeo.
mtcrf
31
31
31
31
31
31
144
146
150
151
163
183
XFX
X
FXM, RS
RS
mtmsr
stwcx.
stwx
X
RS, RA, RB
RS, RA, RB
E
X
X
wrteei
stwux
subfze
subfze.
subfzeo
subfzeo.
addze
addze.
addzeo
addzeo.
stbx
X
RS, RA, RB
31
200 (712)
XO
RT, RA, RB
31
31
31
202 (714)
215
XO
X
RT, RA
RS, RA, RB
RT, RA, RB
subfme
subfme.
subfmeo
subfmeo.
addme
addme.
addmeo
addmeo.
232 (744)
XO
31
234 (746)
XO
RT, RA
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-5. PPC440x5 Instructions by Opcode (continued)
Primary
Opcode
Secondary
Opcode
Form
Mnemonic
Operands
Page
mullw
mullw.
mullwo
mullwo.
dcbtst
stbux
icbt
31
235 (747)
XO
RT, RA, RB
31
31
31
246
247
262
X
X
X
RA,RB
RS, RA, RB
RA, RB
add
add.
31
266 (778)
XO
RT, RA, RB
addo
addo.
dcbt
31
31
278
279
X
X
RA, RB
RT, RA, RB
lhzx
eqv
31
31
31
284
311
316
X
X
X
RA, RS, RB
RT, RA, RB
RA, RS, RB
eqv.
lhzux
xor
xor.
31
31
31
31
31
31
323
339
343
371
375
407
XFX
XFX
X
RT, DCRN
RT, SPRN
RT, RA, RB
RT, SPRN
RT, RA, RB
RS, RA, RB
mfdcr
mfspr
lhax
XFX
X
mftb
lhaux
sthx
X
orc
31
31
31
412
439
444
X
X
X
RA, RS, RB
RS, RA, RB
RA, RS, RB
orc.
sthux
or
or.
31
31
451
454
XFX
X
DCRN, RS
RA, RB
mtdcr
dccci
divwu
divwu.
divwuo
divwuo.
mtspr
dcbi
31
459 (971)
XO
RT, RA, RB
31
31
467
470
XFX
X
SPRN, RS
RA, RB
nand
nand.
dcread
31
31
476
486
X
X
RA, RS, RB
RT, RA, RB
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
Table A-5. PPC440x5 Instructions by Opcode (continued)
Primary
Opcode
Secondary
Opcode
Form
Mnemonic
Operands
Page
divw
divw.
31
491 (1003)
XO
X
RT, RA, RB
divwo
divwo.
mcrxr
31
31
31
31
512
530
533
534
BF
Reserved-nop
lswx
X
X
RT, RA, RB
RT, RA, RB
lwbrx
srw
31
536
X
RA, RS, RB
srw.
31
31
31
31
562
566
594
597
Reserved-nop
tlbsync
Reserved-nop
lswi
X
X
X
RT, RA, NB
31
31
31
31
598
626
658
661
sync
Reserved-nop
Reserved-nop
stswx
X
X
RS, RA, RB
RS, RA, RB
31
31
31
31
31
31
662
690
722
725
754
758
stwbrx
Reserved-nop
Reserved-nop
stswi
X
RS, RA, NB
Reserved-nop
dcba
X
X
RA, RB
31
790
RT, RA, RB
lhbrx
sraw
31
792
X
RA, RS, RB
RA, RS, SH
sraw.
srawi
srawi.
eieio
tlbsx
tlbsx.
sthbrx
extsh
extsh.
tlbre
31
31
31
31
31
31
31
824
854
914
918
922
946
954
X
X
X
X
X
X
X
RT,RA,RB
RS, RA, RB
RA, RS
RT, RA,WS
RA, RS
extsb
extsb.
iccci
31
31
31
966
978
982
X
X
X
RA, RB
RS, RA,WS
RA, RB
tlbwe
icbi
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Table A-5. PPC440x5 Instructions by Opcode (continued)
Primary
Opcode
Secondary
Opcode
Form
Mnemonic
Operands
Page
31
998
1014
X
X
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
RA, RB
RA, RB
icread
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
dcbz
lwz
RT, D(RA)
RT, D(RA)
RT, D(RA)
RT, D(RA)
RS, D(RA)
RS, D(RA)
RS, D(RA)
RS, D(RA)
RT, D(RA)
RT, D(RA)
RT, D(RA)
RT, D(RA)
RS, D(RA)
RS, D(RA)
RT, D(RA)
RS, D(RA)
lwzu
lbz
lbzu
stw
stwu
stb
stbu
lhz
lhzu
lha
lhau
sth
sthu
lmw
stmw
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
instalfa.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Appendix B. PPC440x5 Core Compiler Optimizations
This appendix describes some potential optimizations for compilers.
1. Place target addresses (subroutine entry points) on cache line boundaries (32-bytes)
2. Up to five instructions between a load and a use of the load result. Assuming a data cache hit, the worst
case scenario for the PPC440x5 core is five instructions between a load-use, in order to avoid any bub-
bles. The five instructions are:
• One dispatch, together with the load
• Two the cycle after
• Two the cycle after that
In the next cycle, the use of the load result can dispatch. Therefore, the compiler should try to schedule
as many as five instructions between the load and use of the load result. However, if some of the instruc-
tion pairs between the load-use have pipeline dependencies (such that they cannot dispatch together),
there is no benefit in including the extra instructions between the load-use, and other scheduling optimi-
zations could be made.
In the worst case of instruction pairings, the maximum performance can be achieved with only two
instructions between the load and use of the load result. This is the case when the load instruction pairs
with the instruction before it (instead of after it), and then the next two instructions require the same pipe,
so only one can dispatch during the cycle after the load, and then third instruction after the load needs the
same pipe as the second, so they cannot dispatch together either. In such a case, the third instruction
tion pairings can dispatch together.
3. Pair instructions for dual dispatch. The rules for instruction dispatch in the PPC440x5 core are as follows:
loads and stores can only use the L-Pipe. Branches, CR-updates, XER-updates (“o” forms of arithmetic
instructions), multiply, divide, system instructions (such as rfi and sc), and any SPR accesses (mtspr,
mfspr) can only use the I-Pipe. All other instructions (primarily non-CR-updating and non-XER-updating
arithmetic and logic instructions) can use either the J-Pipe or the I-Pipe. Instructions should be paired so
that they can dispatch as pairs. For example, pair loads and stores with any other instructions. Pair CR-
updates with non-CR-updating instructions and so on.
4. Do not bother to try to schedule instructions between CR-updates and branches that are conditional on
those CR-updates (with some exceptions).
The exceptions are for CR-updates caused by multiply, divide, multiply-accumulate, mtcrf, tlbsx., and
stwcx. instructions. If a branch depends on the CR result of one of these instructions, one or more
instructions should be scheduled (if possible) between the CR update and the branch. Of course, it is also
issue in pairs, and a CR-update and a branch both issue to the I-Pipe, so they cannot issue together.
(The compiler should try to set things up so a CR-update and a following branch (regardless of any CR-
dependency by the branch) can issue in pairs.) This can mean the CR-update can get paired with the
instruction before it, and the branch with the instruction after it, such that there is dual issue in both
cycles. However, if this pairing is not possible, an instruction should be inserted (if possible, of course; do
not create no-ops for no reason) between the CR-update and the branch to allow the dual issue.
The point of this item is to explain that there is no need to separate the CR-update and the branch simply
for the sake of the CR-dependency. That is, there is no extra cycle penalty associated with the CR-
update/branch CR-dependency, beyond the “standard” penalty of the inability to dual issue, unless the
CR-update is one of the types mentioned above.
optimize.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
If the CR-update is MAC or a 16 × 32 multiply, 1 to 3 instructions should be scheduled between the CR-
update and the branch (0 or 1 instruction, depending on whether the CR-update pairs with the instruction
before or after, or 1 to 2 instructions to issue between the issue of the CR-update and the issue of the
branch, depending on whether there is a single-issue or dual-issue opportunity for the instruction(s)
which are scheduled between the CR-update and the branch).
Similarly, if the CR-update is 32 × 32 multiply, divide, tlbsx., or stwcx., schedule 3 to 5 instructions
between the CR-update and the branch (two issue cycles of 2 to 4 instructions between, plus the 0 to 1
issuing with the CR-update).
Finally, if the CR-update is mtcrf, schedule 5 to 7 instructions between (3 cycles of issue between them).
5. Avoid the use of string/multiple instructions (with some exceptions).
The exceptions have to do with cache effects (more cache misses due to more instructions if you use
separate loads/stores instead of a string/multiple), and the specialized behavior of a string, where the
bytes are inserted into the more-significant portion of the GPR, in preparation for a “string compare” oper-
ation to determine which string is “greater” than another. If the string/multiple is for a relatively small num-
ber of registers (or the expansion into discrete loads/stores is known to not have an overall detrimental
cache impact), and if a string is being used only for a copy operation and the size is known, performance
can be improved by using discrete loads/stores. Essentially, due to hazard determination within the pro-
cessor, string/multiples impose a couple of cycles of extra, “false” penalty on both the front-end and the
back-end. On the other hand, if this penalty is amortized over a large number of registers (say 16 or so),
the impact of the extra stalls is probably negligible.
6. Insert 10 or so instructions within a bdnz loop (loop unrolling).
7. Put 4 to 8 instructions between mtlr/mtctr and blr/bctr
8. Put 1 to 3 instructions between 16 × 32 multiply and the use of the result.
9. Put 2 to 5 instructions between 32 × 32 multiply and the use of the result.
10. Use the “without allocate” attribute appropriately on block copy operations, such as calls to the library
memcpy function, or implicit structure copies.
11. Block move operations. If moving a block of memory using a series of load/store operations, perform the
load/store operations in the following order: L1-L2-L3-S1-S2-S3, and repeat. Having the second and third
loads between the first load and the first store fills the two-cycle load-use penalty.
optimize.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
big endian
branching control
ppc440x5IX.fm.
September 12, 2002
Page 572 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
coherence
coherency
compare
control
conventions
structure mapping
instructions
integer
C
cache line
See also cache block
cache management instructions
summary
D
DAC
debug events
applied to instructions that result in multiple storage
registers
code
ppc440x5IX.fm.
September 12, 2002
Page 573 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
data cache
debug
debug events
data read PLB interface requests
data write PLB interface requests
debug modes
dcba
debugging
registers
dcbt
dcbtst
DCC (data cache controller)
dcread
DCRs
ppc440x5IX.fm.
September 12, 2002
Page 574 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
extended mnemonics
DVC
debug events
applied to instructions that result in multiple storage
registers
E
effective address
exception
exception priorities for
integer load, store, and cache management instruc-
ppc440x5IX.fm.
September 12, 2002
Page 575 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
ppc440x5IX.fm.
September 12, 2002
Page 576 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
F
features
ppc440x5IX.fm.
September 12, 2002
Page 577 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
instruction
floating-point load and store instructions, exception priori-
G
GPRs
H, I, J, K
IAC
debug events
registers
icbt
ICC (instruction cache controller)
ppc440x5IX.fm.
September 12, 2002
Page 578 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
ppc440x5IX.fm.
September 12, 2002
Page 579 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
instruction set
summary
Instructions
classes
ppc440x5IX.fm.
September 12, 2002
Page 580 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
instructions
integer instructions
classes
DAC debug events applied to
instructions that result in multiple storage accesses,
integer load, store, and cache management instructions,
interfaces
interrupt
DVC debug events applied to
instructions that result in multiple storage accesses,
instruction
floating-point load and store, exception priorities for,
instruction cache management instruction summary,
integer compare
type
integer load, store, and cache management, exception
ppc440x5IX.fm.
September 12, 2002
Page 581 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
L
interrupt and exception handling registers
interrupt classes
little endian
interrupts
types
M
ppc440x5IX.fm.
September 12, 2002
Page 582 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
MMU
PowerPC Book-E MMU Architecture, nonsupported
O
TLB management instructions
operands
operations
ordering
P
N
ppc440x5IX.fm.
September 12, 2002
Page 583 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
processor control instructions
R
register
requirements
software
reset
return from interrupt instructions, exception priorities for,
ppc440x5IX.fm.
September 12, 2002
Page 584 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
storage attributes
storage control instructions
store miss
S
structure mapping
software
SPRs
ppc440x5IX.fm.
September 12, 2002
Page 585 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
TLB management instructions
synchronization
trap instructions
T
time base
timers
TLB
entry fields
ppc440x5IX.fm.
September 12, 2002
Page 586 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
U, V, W
units
X
ppc440x5IX.fm.
September 12, 2002
Page 587 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
PPC440x5 CPU Core
Preliminary
ppc440x5IX.fm.
September 12, 2002
Page 588 of 583
Download from Www.Somanuals.com. All Manuals Search And Download.
User’s Manual
Preliminary
PPC440x5 CPU Core
Revision Log
Revision Date
7/25/2002
Contents of Modification
Reformatted to division standard template, no content revisions.
Content and format revisions summarized by chapter:
Ch. 2: CCR1 updated
Ch. 4: CCR0 and CCR1 updated. Sections on data cache parity insertion and simulating
parity errors revised.
Ch. 6: Revised MCSR, MCSRR0, MCSRR1, and descriptions of Machine Check Interrupt
handling.
9/12/2002
Ch. 9:Added extended mnemonics to mfspr, mtspr for MCSR, MCSRR0, MCSRR1
Ch10:Fixed format problems in MCSR, MCSRR0, MCSRR1, MSR, SRR1
Appendix A: Added extended mnemonics to mfspr, mtspr for MCSR, MCSRR0, MCSRR1
revlog.fm.
September 12, 2002
Download from Www.Somanuals.com. All Manuals Search And Download.
Download from Www.Somanuals.com. All Manuals Search And Download.
|