llvm.org GIT mirror llvm / 945c85d lib / Target / AMDGPU / AMDGPU.td
945c85d

Tree @945c85d (Download .tar.gz)

AMDGPU.td @945c85draw · history · blame

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
//===-- AMDGPU.td - AMDGPU Tablegen files --------*- tablegen -*-===//
//
//                     The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===------------------------------------------------------------===//

include "llvm/Target/Target.td"

//===------------------------------------------------------------===//
// Subtarget Features (device properties)
//===------------------------------------------------------------===//

def FeatureFP64 : SubtargetFeature<"fp64",
  "FP64",
  "true",
  "Enable double precision operations"
>;

def FeatureFastFMAF32 : SubtargetFeature<"fast-fmaf",
  "FastFMAF32",
  "true",
  "Assuming f32 fma is at least as fast as mul + add"
>;

def HalfRate64Ops : SubtargetFeature<"half-rate-64-ops",
  "HalfRate64Ops",
  "true",
  "Most fp64 instructions are half rate instead of quarter"
>;

def FeatureR600ALUInst : SubtargetFeature<"R600ALUInst",
  "R600ALUInst",
  "false",
  "Older version of ALU instructions encoding"
>;

def FeatureVertexCache : SubtargetFeature<"HasVertexCache",
  "HasVertexCache",
  "true",
  "Specify use of dedicated vertex cache"
>;

def FeatureCaymanISA : SubtargetFeature<"caymanISA",
  "CaymanISA",
  "true",
  "Use Cayman ISA"
>;

def FeatureCFALUBug : SubtargetFeature<"cfalubug",
  "CFALUBug",
  "true",
  "GPU has CF_ALU bug"
>;

def FeatureFlatAddressSpace : SubtargetFeature<"flat-address-space",
  "FlatAddressSpace",
  "true",
  "Support flat address space"
>;

def FeatureUnalignedBufferAccess : SubtargetFeature<"unaligned-buffer-access",
  "UnalignedBufferAccess",
  "true",
  "Support unaligned global loads and stores"
>;

def FeatureUnalignedScratchAccess : SubtargetFeature<"unaligned-scratch-access",
  "UnalignedScratchAccess",
  "true",
  "Support unaligned scratch loads and stores"
>;

// XNACK is disabled if SH_MEM_CONFIG.ADDRESS_MODE = GPUVM on chips that support
// XNACK. The current default kernel driver setting is:
// - graphics ring: XNACK disabled
// - compute ring: XNACK enabled
//
// If XNACK is enabled, the VMEM latency can be worse.
// If XNACK is disabled, the 2 SGPRs can be used for general purposes.
def FeatureXNACK : SubtargetFeature<"xnack",
  "EnableXNACK",
  "true",
  "Enable XNACK support"
>;

def FeatureSGPRInitBug : SubtargetFeature<"sgpr-init-bug",
  "SGPRInitBug",
  "true",
  "VI SGPR initilization bug requiring a fixed SGPR allocation size"
>;

class SubtargetFeatureFetchLimit <string Value> :
                          SubtargetFeature <"fetch"#Value,
  "TexVTXClauseSize",
  Value,
  "Limit the maximum number of fetches in a clause to "#Value
>;

def FeatureFetchLimit8 : SubtargetFeatureFetchLimit <"8">;
def FeatureFetchLimit16 : SubtargetFeatureFetchLimit <"16">;

class SubtargetFeatureWavefrontSize <int Value> : SubtargetFeature<
  "wavefrontsize"#Value,
  "WavefrontSize",
  !cast<string>(Value),
  "The number of threads per wavefront"
>;

def FeatureWavefrontSize16 : SubtargetFeatureWavefrontSize<16>;
def FeatureWavefrontSize32 : SubtargetFeatureWavefrontSize<32>;
def FeatureWavefrontSize64 : SubtargetFeatureWavefrontSize<64>;

class SubtargetFeatureLDSBankCount <int Value> : SubtargetFeature <
  "ldsbankcount"#Value,
  "LDSBankCount",
  !cast<string>(Value),
  "The number of LDS banks per compute unit."
>;

def FeatureLDSBankCount16 : SubtargetFeatureLDSBankCount<16>;
def FeatureLDSBankCount32 : SubtargetFeatureLDSBankCount<32>;

class SubtargetFeatureLocalMemorySize <int Value> : SubtargetFeature<
  "localmemorysize"#Value,
  "LocalMemorySize",
  !cast<string>(Value),
  "The size of local memory in bytes"
>;

def FeatureGCN : SubtargetFeature<"gcn",
  "IsGCN",
  "true",
  "GCN or newer GPU"
>;

def FeatureGCN1Encoding : SubtargetFeature<"gcn1-encoding",
  "GCN1Encoding",
  "true",
  "Encoding format for SI and CI"
>;

def FeatureGCN3Encoding : SubtargetFeature<"gcn3-encoding",
  "GCN3Encoding",
  "true",
  "Encoding format for VI"
>;

def FeatureCIInsts : SubtargetFeature<"ci-insts",
  "CIInsts",
  "true",
  "Additional intstructions for CI+"
>;

def FeatureSMemRealTime : SubtargetFeature<"s-memrealtime",
  "HasSMemRealTime",
  "true",
  "Has s_memrealtime instruction"
>;

def FeatureInv2PiInlineImm : SubtargetFeature<"inv-2pi-inline-imm",
  "HasInv2PiInlineImm",
  "true",
  "Has 1 / (2 * pi) as inline immediate"
>;

def Feature16BitInsts : SubtargetFeature<"16-bit-insts",
  "Has16BitInsts",
  "true",
  "Has i16/f16 instructions"
>;

def FeatureMovrel : SubtargetFeature<"movrel",
  "HasMovrel",
  "true",
  "Has v_movrel*_b32 instructions"
>;

def FeatureVGPRIndexMode : SubtargetFeature<"vgpr-index-mode",
  "HasVGPRIndexMode",
  "true",
  "Has VGPR mode register indexing"
>;

def FeatureScalarStores : SubtargetFeature<"scalar-stores",
  "HasScalarStores",
  "true",
  "Has store scalar memory instructions"
>;

def FeatureSDWA : SubtargetFeature<"sdwa",
  "HasSDWA",
  "true",
  "Support SDWA (Sub-DWORD Addressing) extension"
>;

def FeatureDPP : SubtargetFeature<"dpp",
  "HasDPP",
  "true",
  "Support DPP (Data Parallel Primitives) extension"
>;

//===------------------------------------------------------------===//
// Subtarget Features (options and debugging)
//===------------------------------------------------------------===//

// Some instructions do not support denormals despite this flag. Using
// fp32 denormals also causes instructions to run at the double
// precision rate for the device.
def FeatureFP32Denormals : SubtargetFeature<"fp32-denormals",
  "FP32Denormals",
  "true",
  "Enable single precision denormal handling"
>;

// Denormal handling for fp64 and fp16 is controlled by the same
// config register when fp16 supported.
// TODO: Do we need a separate f16 setting when not legal?
def FeatureFP64FP16Denormals : SubtargetFeature<"fp64-fp16-denormals",
  "FP64FP16Denormals",
  "true",
  "Enable double and half precision denormal handling",
  [FeatureFP64]
>;

def FeatureFP64Denormals : SubtargetFeature<"fp64-denormals",
  "FP64FP16Denormals",
  "true",
  "Enable double and half precision denormal handling",
  [FeatureFP64, FeatureFP64FP16Denormals]
>;

def FeatureFP16Denormals : SubtargetFeature<"fp16-denormals",
  "FP64FP16Denormals",
  "true",
  "Enable half precision denormal handling",
  [FeatureFP64FP16Denormals]
>;

def FeatureFPExceptions : SubtargetFeature<"fp-exceptions",
  "FPExceptions",
  "true",
  "Enable floating point exceptions"
>;

class FeatureMaxPrivateElementSize<int size> : SubtargetFeature<
  "max-private-element-size-"#size,
  "MaxPrivateElementSize",
  !cast<string>(size),
  "Maximum private access size may be "#size
>;

def FeatureMaxPrivateElementSize4 : FeatureMaxPrivateElementSize<4>;
def FeatureMaxPrivateElementSize8 : FeatureMaxPrivateElementSize<8>;
def FeatureMaxPrivateElementSize16 : FeatureMaxPrivateElementSize<16>;

def FeatureVGPRSpilling : SubtargetFeature<"vgpr-spilling",
  "EnableVGPRSpilling",
  "true",
  "Enable spilling of VGPRs to scratch memory"
>;

def FeatureDumpCode : SubtargetFeature <"DumpCode",
  "DumpCode",
  "true",
  "Dump MachineInstrs in the CodeEmitter"
>;

def FeatureDumpCodeLower : SubtargetFeature <"dumpcode",
  "DumpCode",
  "true",
  "Dump MachineInstrs in the CodeEmitter"
>;

def FeaturePromoteAlloca : SubtargetFeature <"promote-alloca",
  "EnablePromoteAlloca",
  "true",
  "Enable promote alloca pass"
>;

// XXX - This should probably be removed once enabled by default
def FeatureEnableLoadStoreOpt : SubtargetFeature <"load-store-opt",
  "EnableLoadStoreOpt",
  "true",
  "Enable SI load/store optimizer pass"
>;

// Performance debugging feature. Allow using DS instruction immediate
// offsets even if the base pointer can't be proven to be base. On SI,
// base pointer values that won't give the same result as a 16-bit add
// are not safe to fold, but this will override the conservative test
// for the base pointer.
def FeatureEnableUnsafeDSOffsetFolding : SubtargetFeature <
  "unsafe-ds-offset-folding",
  "EnableUnsafeDSOffsetFolding",
  "true",
  "Force using DS instruction immediate offsets on SI"
>;

def FeatureEnableSIScheduler : SubtargetFeature<"si-scheduler",
  "EnableSIScheduler",
  "true",
  "Enable SI Machine Scheduler"
>;

// Unless +-flat-for-global is specified, turn on FlatForGlobal for
// all OS-es on VI and newer hardware to avoid assertion failures due
// to missing ADDR64 variants of MUBUF instructions.
// FIXME: moveToVALU should be able to handle converting addr64 MUBUF
// instructions.

def FeatureFlatForGlobal : SubtargetFeature<"flat-for-global",
  "FlatForGlobal",
  "true",
  "Force to generate flat instruction for global"
>;

// Dummy feature used to disable assembler instructions.
def FeatureDisable : SubtargetFeature<"",
  "FeatureDisable","true",
  "Dummy feature to disable assembler instructions"
>;

class SubtargetFeatureGeneration <string Value,
                                  list<SubtargetFeature> Implies> :
        SubtargetFeature <Value, "Gen", "AMDGPUSubtarget::"#Value,
                          Value#" GPU generation", Implies>;

def FeatureLocalMemorySize0 : SubtargetFeatureLocalMemorySize<0>;
def FeatureLocalMemorySize32768 : SubtargetFeatureLocalMemorySize<32768>;
def FeatureLocalMemorySize65536 : SubtargetFeatureLocalMemorySize<65536>;

def FeatureR600 : SubtargetFeatureGeneration<"R600",
  [FeatureR600ALUInst, FeatureFetchLimit8, FeatureLocalMemorySize0]
>;

def FeatureR700 : SubtargetFeatureGeneration<"R700",
  [FeatureFetchLimit16, FeatureLocalMemorySize0]
>;

def FeatureEvergreen : SubtargetFeatureGeneration<"EVERGREEN",
  [FeatureFetchLimit16, FeatureLocalMemorySize32768]
>;

def FeatureNorthernIslands : SubtargetFeatureGeneration<"NORTHERN_ISLANDS",
  [FeatureFetchLimit16, FeatureWavefrontSize64,
   FeatureLocalMemorySize32768]
>;

def FeatureSouthernIslands : SubtargetFeatureGeneration<"SOUTHERN_ISLANDS",
  [FeatureFP64, FeatureLocalMemorySize32768,
  FeatureWavefrontSize64, FeatureGCN, FeatureGCN1Encoding,
  FeatureLDSBankCount32, FeatureMovrel]
>;

def FeatureSeaIslands : SubtargetFeatureGeneration<"SEA_ISLANDS",
  [FeatureFP64, FeatureLocalMemorySize65536,
  FeatureWavefrontSize64, FeatureGCN, FeatureFlatAddressSpace,
  FeatureGCN1Encoding, FeatureCIInsts, FeatureMovrel]
>;

def FeatureVolcanicIslands : SubtargetFeatureGeneration<"VOLCANIC_ISLANDS",
  [FeatureFP64, FeatureLocalMemorySize65536,
   FeatureWavefrontSize64, FeatureFlatAddressSpace, FeatureGCN,
   FeatureGCN3Encoding, FeatureCIInsts, Feature16BitInsts,
   FeatureSMemRealTime, FeatureVGPRIndexMode, FeatureMovrel,
   FeatureScalarStores, FeatureInv2PiInlineImm, FeatureSDWA,
   FeatureDPP
  ]
>;

class SubtargetFeatureISAVersion <int Major, int Minor, int Stepping,
                                  list<SubtargetFeature> Implies>
                                 : SubtargetFeature <
  "isaver"#Major#"."#Minor#"."#Stepping,
  "IsaVersion",
  "ISAVersion"#Major#"_"#Minor#"_"#Stepping,
  "Instruction set version number",
  Implies
>;

def FeatureISAVersion7_0_0 : SubtargetFeatureISAVersion <7,0,0,
  [FeatureSeaIslands,
   FeatureLDSBankCount32]>;

def FeatureISAVersion7_0_1 : SubtargetFeatureISAVersion <7,0,1,
  [FeatureSeaIslands,
   HalfRate64Ops,
   FeatureLDSBankCount32,
   FeatureFastFMAF32]>;

def FeatureISAVersion7_0_2 : SubtargetFeatureISAVersion <7,0,2,
  [FeatureSeaIslands,
   FeatureLDSBankCount16]>;

def FeatureISAVersion8_0_0 : SubtargetFeatureISAVersion <8,0,0,
  [FeatureVolcanicIslands,
   FeatureLDSBankCount32,
   FeatureSGPRInitBug]>;

def FeatureISAVersion8_0_1 : SubtargetFeatureISAVersion <8,0,1,
  [FeatureVolcanicIslands,
   FeatureLDSBankCount32,
   FeatureXNACK]>;

def FeatureISAVersion8_0_2 : SubtargetFeatureISAVersion <8,0,2,
  [FeatureVolcanicIslands,
   FeatureLDSBankCount32,
   FeatureSGPRInitBug]>;

def FeatureISAVersion8_0_3 : SubtargetFeatureISAVersion <8,0,3,
  [FeatureVolcanicIslands,
   FeatureLDSBankCount32]>;

def FeatureISAVersion8_0_4 : SubtargetFeatureISAVersion <8,0,4,
  [FeatureVolcanicIslands,
   FeatureLDSBankCount32]>;

def FeatureISAVersion8_1_0 : SubtargetFeatureISAVersion <8,1,0,
  [FeatureVolcanicIslands,
   FeatureLDSBankCount16,
   FeatureXNACK]>;

//===----------------------------------------------------------------------===//
// Debugger related subtarget features.
//===----------------------------------------------------------------------===//

def FeatureDebuggerInsertNops : SubtargetFeature<
  "amdgpu-debugger-insert-nops",
  "DebuggerInsertNops",
  "true",
  "Insert one nop instruction for each high level source statement"
>;

def FeatureDebuggerReserveRegs : SubtargetFeature<
  "amdgpu-debugger-reserve-regs",
  "DebuggerReserveRegs",
  "true",
  "Reserve registers for debugger usage"
>;

def FeatureDebuggerEmitPrologue : SubtargetFeature<
  "amdgpu-debugger-emit-prologue",
  "DebuggerEmitPrologue",
  "true",
  "Emit debugger prologue"
>;

//===----------------------------------------------------------------------===//

def AMDGPUInstrInfo : InstrInfo {
  let guessInstructionProperties = 1;
  let noNamedPositionallyEncodedOperands = 1;
}

def AMDGPUAsmParser : AsmParser {
  // Some of the R600 registers have the same name, so this crashes.
  // For example T0_XYZW and T0_XY both have the asm name T0.
  let ShouldEmitMatchRegisterName = 0;
}

def AMDGPUAsmWriter : AsmWriter {
  int PassSubtarget = 1;
}

def AMDGPUAsmVariants {
  string Default = "Default";
  int Default_ID = 0;
  string VOP3 = "VOP3";
  int VOP3_ID = 1;
  string SDWA = "SDWA";
  int SDWA_ID = 2;
  string DPP = "DPP";
  int DPP_ID = 3;
  string Disable = "Disable";
  int Disable_ID = 4;
}

def DefaultAMDGPUAsmParserVariant : AsmParserVariant {
  let Variant = AMDGPUAsmVariants.Default_ID;
  let Name = AMDGPUAsmVariants.Default;
}

def VOP3AsmParserVariant : AsmParserVariant {
  let Variant = AMDGPUAsmVariants.VOP3_ID;
  let Name = AMDGPUAsmVariants.VOP3;
}

def SDWAAsmParserVariant : AsmParserVariant {
  let Variant = AMDGPUAsmVariants.SDWA_ID;
  let Name = AMDGPUAsmVariants.SDWA;
}

def DPPAsmParserVariant : AsmParserVariant {
  let Variant = AMDGPUAsmVariants.DPP_ID;
  let Name = AMDGPUAsmVariants.DPP;
}

def AMDGPU : Target {
  // Pull in Instruction Info:
  let InstructionSet = AMDGPUInstrInfo;
  let AssemblyParsers = [AMDGPUAsmParser];
  let AssemblyParserVariants = [DefaultAMDGPUAsmParserVariant,
                                VOP3AsmParserVariant,
                                SDWAAsmParserVariant,
                                DPPAsmParserVariant];
  let AssemblyWriters = [AMDGPUAsmWriter];
}

// Dummy Instruction itineraries for pseudo instructions
def ALU_NULL : FuncUnit;
def NullALU : InstrItinClass;

//===----------------------------------------------------------------------===//
// Predicate helper class
//===----------------------------------------------------------------------===//

def TruePredicate : Predicate<"true">;

def isSICI : Predicate<
  "Subtarget->getGeneration() == AMDGPUSubtarget::SOUTHERN_ISLANDS ||"
  "Subtarget->getGeneration() == AMDGPUSubtarget::SEA_ISLANDS"
>, AssemblerPredicate<"FeatureGCN1Encoding">;

def isVI : Predicate <
  "Subtarget->getGeneration() >= AMDGPUSubtarget::VOLCANIC_ISLANDS">,
  AssemblerPredicate<"FeatureGCN3Encoding">;

def isCIVI : Predicate <
  "Subtarget->getGeneration() == AMDGPUSubtarget::SEA_ISLANDS || "
  "Subtarget->getGeneration() == AMDGPUSubtarget::VOLCANIC_ISLANDS"
>, AssemblerPredicate<"FeatureCIInsts">;

def HasFlatAddressSpace : Predicate<"Subtarget->hasFlatAddressSpace()">;

def Has16BitInsts : Predicate<"Subtarget->has16BitInsts()">;

def HasSDWA : Predicate<"Subtarget->hasSDWA()">,
  AssemblerPredicate<"FeatureSDWA">;

def HasDPP : Predicate<"Subtarget->hasDPP()">,
  AssemblerPredicate<"FeatureDPP">;

class PredicateControl {
  Predicate SubtargetPredicate;
  Predicate SIAssemblerPredicate = isSICI;
  Predicate VIAssemblerPredicate = isVI;
  list<Predicate> AssemblerPredicates = [];
  Predicate AssemblerPredicate = TruePredicate;
  list<Predicate> OtherPredicates = [];
  list<Predicate> Predicates = !listconcat([SubtargetPredicate, AssemblerPredicate],
                                            AssemblerPredicates,
                                            OtherPredicates);
}

// Include AMDGPU TD files
include "R600Schedule.td"
include "SISchedule.td"
include "Processors.td"
include "AMDGPUInstrInfo.td"
include "AMDGPUIntrinsics.td"
include "AMDGPURegisterInfo.td"
include "AMDGPURegisterBanks.td"
include "AMDGPUInstructions.td"
include "AMDGPUCallingConv.td"