 [ Table of Contents | Index ]
Appendix G
G. Index
The page numbers for indexed items have no significance in this Web version of The PowerPC Compiler Writer's Guide; they refer only to the hardcopy version. However, each page number has a functional hyperlink to the indexed item.
Numerics
- 2n in Fortran, 154
- 32-bit implementation, 5
- 32-bit mode, 5
- 64-bit implementation, 5
- 64-bit mode, 5
A
- AA, 9, 209
- ABI, 157, 209
- absolute value, 50, 205
- activation record, 209
- add, 46
- addi, 46
- addic., 46
- addis, 46
- address, 6, 209
- addressing modes, 8
- addze, 53
- AIX
- argument passing rules, 163
- function descriptors, 168
- function return values, 165
- out-of-module function calls, 168
- procedure calling sequence, 163
- procedure interfaces, 157
- procedure prologs and epilogs, 165
- Register Conventions, 158
- run-time stack, 160
- Table Of Contents, 167
- algebraic, 209
- alias, 209
- alignment, 10, 68, 133, 184, 209
- load and store, 133
- Alignment interrupt, 10, 74, 184
- andi., 46, 49
- andis., 46
- API, 209
- argument, 210
- atomic, 165, 210
B
- B, 210
- b, 210
- base, 210
- base address, 210
- basic block, 112, 210
- bclr, 19
- big-endian, 10, 210
- Big-Endian mode, 210
- binary point, 210
- bit manipulation, 65, 208
- Block Address Translation, 210
- blocking, 210
- bounds checking, 144
- branch, 210
- absolute, 9, 24
- branch-and-link, 210
- branch-on-count, 29, 211
- conditional, 24
- conditional to Count Register, 9
- conditional to Link Register, 9
- dynamic prediction, 35, 102
- folding, 211
- instruction performance, 22
- multi-way conditional, 25
- needless activity, 23
- prediction, 18, 35, 102, 172, 211, 214, 230
- registers, 7
- relative, 9
- resolution, 211, 228
- scheduling, 110
- static prediction, 35, 230
- unconditional, 23
- unresolved, 18
- branch history table, 103
- branch target address cache, 102, 211
- Branch-Processing Unit, 7, 211
- BTAC (see branch target address cache)
- bubble, 211
- bypass, 101, 211
C
- CA (see Carry bit)
- cache, 134, 172
- block, 211
- hit, 211
- miss, 106, 211
- touch, 134, 211
- touch instructions, 45, 75
- Carry bit, 46, 212
- CIA, 212
- clean-up code, 131, 212
- clear, 212
- clear least-significant nonzero bit, 141
- cntlzw, 47
- coherence, 212
- coherence block, 212
- committed, 212
- Common Model, 117, 212
- compare instructions, 21
- compiler, 157, 212
- complete, 100, 212
- completion unit, 172
- computing predicates, 38
- condition code, 212
- Condition Register, xix, 6, 7, 21, 160, 212
- Condition Register contention, 23
- Condition Register logical, 40
- conditional increment, 39
- context, 212
- context switch, 103, 212
- context synchronization, 213
- control flow, 17
- conversions, 72
- Count Register, 6, 7, 20, 213
- count trailing zeros, 145
- CPU time, 213
- CR (see Condition Register)
- CR0:7, 213
- CTR (see Count Register)
- cycle, 213
D
- data
- addressing, 9
- alignment, 10
- decode, 99
- decoding a "zero means 2n" field, 153
- delay, 112
- denormal, 214
- denormalized number, 214
- dependence, 17, 214
- antidependence, 17, 209
- control, 17, 213
- data, 17, 213
- def-def, 213
- def-use, 214
- name, 17, 223
- output, 17, 224
- use-def, 231
- direct-store segment, 214
- dispatch, 99
- displacement, 214
- division
- 64-bit unsigned fixed-point, 62
- by a power of 2, 52
- by integer constants, 51
- magic number, 53
- replace with multiplication by reciprocal, 92
- do loop, 29
- double-precision format, 72, 214
- doubleword, 214
- DSP filters, 92
- dynamic branch prediction, 35, 102, 214
- dynamic linking, 167, 214
- dynamic store forwarding, 214
E
- effective address, 214
- endian conversion, 75
- endian orientation, 10, 214
- endian reversal, 49
- exception, 214
- executable module, 157
- execute, 100
- execution synchronization, 215
- execution time, 112, 175, 215
- exponent, 215
- extended mnemonic, xx, 215
- external cache, 215
F
- fall-through path, 23, 215
- FE0, FE1 (see Floating-Point Exception mode bits)
- fetch, 99, 215
- fetch buffer alignment, 134
- FEX (see Floating-Point Enabled Exception Summary bit)
- FI (see Floating-Point Fraction Inexact bit)
- find first string of 1-bits of a given length, 150
- first-class value, 215
- fixed-point, 215
- Fixed-Point Exception Register, 6, 8, 46, 215
- Fixed-Point Unit, 7, 104, 215
- flat memory, 215
- floating-point, 216
- Available bit, 216
- branch elimination, 86-91
- double-precision load and store, 74
- Enabled Exception Summary bit, 94, 216
- environment controls, 11
- Exception Mode bits, 13, 216
- Exception Summary bit, 216
- exceptions, 93
- Floating-Point Status and Control Register, 6, 8, 11, 94, 218
- Fraction Inexact bit, 216
- Fraction Rounded bit, 216
- Inexact Exception bit, 216
- Inexact Exception Enable bit, 12, 216
- Invalid Operation Exception (
 - ) bit, 217 - Invalid Operation Exception ( × 0) bit, 217
- Invalid Operation Exception ( ÷ ) bit, 216
- Invalid Operation Exception (0 ÷ 0) bit, 216
- Invalid Operation Exception (Invalid Compare) bit, 217
- Invalid Operation Exception (Invalid Integer Convert) bit, 217
- Invalid Operation Exception (Invalid Square Root) bit, 217
- Invalid Operation Exception (SNaN) bit, 217
- Invalid Operation Exception (Software Request) bit, 217
- Invalid Operation Exception Enable bit, 12, 216
- Invalid Operation Exception Summary bit, 217
- move instructions, 75
- multiply-add, 79
- Non-IEEE Mode bit, 12, 217
- operand, 218
- Overflow Exception Bit, 217
- Overflow Exception Enable bit, 12, 217
- Register, 6, 8, 68, 72, 159, 217
- Result Flags, 218
- Rounding Control, 13, 74, 218
- scheduling, 105
- single-precision load and store, 74
- Underflow Exception bit, 218
- Underflow Exception Enable bit, 12, 218
- Zero-Divide Exception bit, 218
- Zero-Divide Exception Enable bit, 12, 218
- Floating-Point Unit, 8, 105, 218
- format conversion, 79-86
- floating-point to integer, 80
- integer to floating-point, 83
- rounding to floating-point integer, 85
- forwarding, 99, 101, 218
- FP (see Floating-Point Available bit)
- FPR0:32, 218
- FPRF (see Floating-Point Result Flags)
- FPSCR (see Floating-Point Status and Control Register)
- FPU, 219
- FR (see Floating-Point Fraction Rounded bit)
- fraction, 219
- fres, 77
- frsp, 11, 73, 74
- frsqrte, 77
- FRx, 219
- fsel, 78, 86
- fsqrt, 77
- function, 219
- functional class, 5, 219
- inter-communication, 14
- FX (see Floating-Point Exception Summary bit)
G
- General-Purpose Register, 6, 8, 159, 219
H
- halfword, 219
- hardware overview, 98
- hazard, 100, 219
- control, 102, 213
- data, 100, 213
- read after write, 100, 227
- structural, 103, 230
- write after read, 100, 232
- write after write, 100, 233
- high, 219
- hoist, 219
- home location, 219
I
- IEEE 754, 5, 11, 78-79, 86, 92, 219
- IEEE mode, 219
- if-else, 24
- immediate operand, 219
- implementations, 171
- implicit integer bit, 219
- imprecise, 220
- imprecise interrupt, 220
- incrementing a reversed integer, 152
- index, 220
- indirect, 220
- Inexact Exception, 220
- inline expansion, 220
- instruction, 13
- addressing, 8
- alignment, 10
- arithmetic, 47
- compare, 48
- Condition Register logical, 22
- floating-point arithmetic, 75
- floating-point compare, 76
- floating-point selection, 78
- FPSCR, 76
- load, 9
- logical, 47
- optional, 13, 77
- preferred form, 14
- queue, 171, 220
- reciprocal estimate, 77
- reciprocal square root estimate, 77
- record, 21
- restart, 11, 220
- rotate, 47
- shift, 47
- square root, 77
- store, 9
- instruction usage statistics, 187
- integer
- bit position, 220
- integer log base 10, 154
- interlock, 100, 220
- interprocedural analysis, 158, 220
- iteration, 28
K
- K, 220
L
- latch point, 28
- latency, 175, 220
- LE (see Little-Endian Mode bit)
- leaf procedures, 163
- least-significant, 221
- lifetime analysis, 221
- link bit, 7, 19
- Link Register, 6, 7, 19, 221
- linkage convention, 157, 221
- linkage editor, 157, 221
- little-endian, 221
- Little-Endian Mode bit, 221
- LK, 221
- load, 43, 221
- and reserve, 44
- multiple, 45
- scheduling, 106
- string, 45
- with byte-reversal, 45
- load and store queue, 109, 172, 173, 222
- loader, 157, 221
- load-following-store contention, 107, 221
- loading a constant into a register, 48
- load-store bound, 133, 221
- Load-Store Unit, 106, 130, 222
- load-use delay, 107, 222
- locality, 222
- loop
- optimization, 130, 222
- low, 222
- LR, 222
- LSB, 222
- lsb, 222
M
- Machine State Register, 6, 11, 222
- machine-dependent optimization, 3, 222
- machine-independent optimization, 3, 222
- magic number, 51
- algorithm, 57
- mask, 48, 222
- maximum, 51
- mcrf, 21
- mcrfs, 21, 22
- mcrxr, 21, 22, 46
- memory, 8, 222
- access, 43
- addressing, 8
- coherence (see coherence)
- functions, 68
- models, 8
- memset, 69
- mfcr, 22, 38
- mfctr, 23
- mffs, 73
- mflr, 23
- mfspr, 9
- minimum, 51
- misaligned, 223
- miss penalty, 223
- MMU, 223
- most-significant, 223
- MSB, 223
- msb, 223
- MSR (see Machine State Register)
- mtcrf, 22
- mtctr, 20, 23
- mtlr, 19
- mtspr, 9
- mtxer, 46, 48
- mulld, 47
- mulli, 47
- mullw, 47
- multiple-precision shifts, 66
N
- NaN, 223
- NI (see Floating-Point Non-IEEE-Mode Enable bit)
- NIA, 223
- Non-IEEE mode, 12, 79, 223
- no-op, 14, 224
- normal, 224
- normalize, 224
- normalized number, 224
- Notation, xviii
O
- object module, 157, 224
- OE (see Floating-Point Overflow Exception Enable bit)
- offset, 224
- optimization, 3, 224
- ordinary segment, 224
- out-of-order, 224
- OV (see Overflow bit)
- overflow, 46, 224
- Overflow bit, 46, 224
- OX (see Floating-Point Overflow Exception bit)
P
- page, 225
- path length, 225
- pipeline, 99, 225
- pipeline stages, 99
- pointer, 225
- pointer chasing, 107, 225
- population count, 146
- POWER, 225
- power of 2 crossing, 144
- PowerPC Little-Endian mode, 10
- PR, 225
- precise, 225
- precise interrupt, 93, 225
- precision, 72, 225
- predicate, 38, 225
- prefetch, 225
- preprocessor, 195, 225
- privilege level, 226
- privilege mechanism, 226
- privileged instruction, 226
- problem state, 5, 226
- procedure, 32, 157, 226
- process, 226
- processor starvation, 107, 226
- profile, 36, 226
- program order, 99, 226
- protection, 226
Q
- quadword, 226
- quiet NaN, 227
R
- R0:32, 227
- range test, 26
- RAW (see read after write)
- Rc, 227
- read, 227
- read after write, 100, 227
- record, 21, 46, 227
- re-entrant, 227
- register
- allocation, 101, 227
- dedicated, 158, 213
- exchange, 140
- non-volatile, 158, 223
- renaming, 101, 172
- volatile, 158, 232
- registers, 7
- remainder, 61
- reservation, 44, 227
- reservation granule, 44, 227
- reservation station, 172, 228
- resolution, 18
- retire, 228
- RISC, 1
- rlwinm, 47
- RN (see Floating-Point Rounding Control Field)
- round to a multiple of a given power of 2, 142
- round up or down to next power of 2, 142
- rounding, 72
- run-time environment, 228
- Rx, 228
S
- scheduling, 104, 228
- searching for the specified byte value, 69
- segment, 228
- sequential execution model, 99, 228
- sequential order, 228
- serialization, 104, 174, 228
- set, 228
- SF (see Sixty-Four Bit Mode bit)
- shadow register, 101, 229
- sign extension, 229
- sign function, 139, 205
- signaling NaN, 229
- significand, 229
- single-precision format, 72, 229
- Sixty-Four-Bit mode, 6, 229
- SNaN (see signaling NaN)
- SO (see Summary Overflow bit)
- software pipelining, 229
- spatial locality, 229
- Special-Purpose Register, 160, 229
- speculation, 18, 229
- SPR (see Special-Purpose Register)
- srawi, 53
- stale, 229
- stall, 102, 230
- static branch prediction, 35, 230
- static linking, 170, 230
- status bits, 6, 46
- floating-point, 75
- stfiwx, 77
- sticky bit, 230
- store, 43
- conditional, 44
- multiple, 45
- scheduling, 106
- string, 45
- with byte reversal, 45
- stride, 31, 230
- string, 230
- functions, 68
- strlen, 29
- subf, 46
- subroutine, 230
- Summary Overflow bit, 230
- superscalar, 99
- supervisor state, 230
- switch, 25
- synchronization, 230
- system, 230
- system register, 231
T
- taken, 18, 231
- task, 231
- temporal locality, 231
- thread, 158, 231
- throughput, 231
- tiny, 231
- TLB (see translation-lookaside buffer)
- transfer of sign, 140
- translation-lookaside buffer, 134, 172, 231
- trap, 94, 144, 231
- trap interrupt, 94, 231
- typing, 72
U
- UE (see Floating-Point Underflow Exception Enable bit)
- update, 231
- user mode, 232
- UX (see Floating-Point Underflow Exception bit)
V
- VE (see Floating-Point Invalid Operation Exception Enable bit)
- virtual memory, 232
- VX (see Floating-Point Invalid Operation Exception Summary bit)
- VXCVI (see Floating-Point Invalid Operation Exception (Invalid Integer Convert) bit)
- VXIDI (see Floating-Point Invalid Operation Exception ( ÷ ) bit)
- VXIMZ (see Floating-Point Invalid Operation Exception ( × 0) bit)
- VXISI (see Floating-Point Invalid Operation Exception ( - ) bit)
- VXSNAN (see Floating-Point Invalid Operation Exception (SNaN) bit)
- VXSOFT (see Floating-Point Invalid Operation Exception (Software Request) bit)
- VXSQRT (see Floating-Point Invalid Operation Exception (Invalid Square Root) bit)
- VXVC (see Floating-Point Invalid Operation Exception (Invalid Compare) bit)
- VXZDZ (see Floating-Point Invalid Operation Exception (0 ÷ 0) bit)
W
- WAR (see write after read)
- WAW (see write after write)
- while, 29
- word, 232
- write after read, 100, 232
- write after write, 100, 233
- write back, 100, 233
X
- x = y predicate, 141
- XE (see Floating-Point Inexact Exception Enable bit)
- XER (see Fixed-Point Exception Register)
- XX (see Floating-Point Inexact Exception bit)
Y
- y bit (see static branch prediction bit)
Z
- ZE (see Floating-Point Zero-Divide Exception Enable bit)
- zero extension, 233
- ZX (see Floating-Point Zero-Divide Exception bit)
[ Table of Contents | Index ]
|
