llvm.org GIT mirror llvm / 28fe7a0 lib / Target / AMDGPU / SIInstrFormats.td
28fe7a0

Tree @28fe7a0 (Download .tar.gz)

SIInstrFormats.td @28fe7a0raw · 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
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
//===-- SIInstrFormats.td - SI Instruction Encodings ----------------------===//
//
//                     The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// SI Instruction format definitions.
//
//===----------------------------------------------------------------------===//

class InstSI <dag outs, dag ins, string asm, list<dag> pattern> :
    AMDGPUInst<outs, ins, asm, pattern>, PredicateControl {

  field bits<1> VM_CNT = 0;
  field bits<1> EXP_CNT = 0;
  field bits<1> LGKM_CNT = 0;

  field bits<1> SALU = 0;
  field bits<1> VALU = 0;

  field bits<1> SOP1 = 0;
  field bits<1> SOP2 = 0;
  field bits<1> SOPC = 0;
  field bits<1> SOPK = 0;
  field bits<1> SOPP = 0;

  field bits<1> VOP1 = 0;
  field bits<1> VOP2 = 0;
  field bits<1> VOP3 = 0;
  field bits<1> VOPC = 0;

  field bits<1> MUBUF = 0;
  field bits<1> MTBUF = 0;
  field bits<1> SMRD = 0;
  field bits<1> DS = 0;
  field bits<1> MIMG = 0;
  field bits<1> FLAT = 0;
  field bits<1> WQM = 0;
  field bits<1> VGPRSpill = 0;

  // This bit tells the assembler to use the 32-bit encoding in case it
  // is unable to infer the encoding from the operands.
  field bits<1> VOPAsmPrefer32Bit = 0;

  // These need to be kept in sync with the enum in SIInstrFlags.
  let TSFlags{0} = VM_CNT;
  let TSFlags{1} = EXP_CNT;
  let TSFlags{2} = LGKM_CNT;

  let TSFlags{3} = SALU;
  let TSFlags{4} = VALU;

  let TSFlags{5} = SOP1;
  let TSFlags{6} = SOP2;
  let TSFlags{7} = SOPC;
  let TSFlags{8} = SOPK;
  let TSFlags{9} = SOPP;

  let TSFlags{10} = VOP1;
  let TSFlags{11} = VOP2;
  let TSFlags{12} = VOP3;
  let TSFlags{13} = VOPC;

  let TSFlags{14} = MUBUF;
  let TSFlags{15} = MTBUF;
  let TSFlags{16} = SMRD;
  let TSFlags{17} = DS;
  let TSFlags{18} = MIMG;
  let TSFlags{19} = FLAT;
  let TSFlags{20} = WQM;
  let TSFlags{21} = VGPRSpill;
  let TSFlags{22} = VOPAsmPrefer32Bit;

  let SchedRW = [Write32Bit];
}

class Enc32 {
  field bits<32> Inst;
  int Size = 4;
}

class Enc64 {
  field bits<64> Inst;
  int Size = 8;
}

class VOPDstOperand <RegisterClass rc> : RegisterOperand <rc, "printVOPDst">;

let Uses = [EXEC] in {

class VOPAnyCommon <dag outs, dag ins, string asm, list<dag> pattern> :
    InstSI <outs, ins, asm, pattern> {

  let mayLoad = 0;
  let mayStore = 0;
  let hasSideEffects = 0;
  let UseNamedOperandTable = 1;
  let VALU = 1;
}

class VOPCCommon <dag ins, string asm, list<dag> pattern> :
    VOPAnyCommon <(outs), ins, asm, pattern> {

  let VOPC = 1;
  let Size = 4;
  let Defs = [VCC];
}

class VOP1Common <dag outs, dag ins, string asm, list<dag> pattern> :
    VOPAnyCommon <outs, ins, asm, pattern> {

  let VOP1 = 1;
  let Size = 4;
}

class VOP2Common <dag outs, dag ins, string asm, list<dag> pattern> :
    VOPAnyCommon <outs, ins, asm, pattern> {

  let VOP2 = 1;
  let Size = 4;
}

class VOP3Common <dag outs, dag ins, string asm, list<dag> pattern> :
    VOPAnyCommon <outs, ins, asm, pattern> {

  // Using complex patterns gives VOP3 patterns a very high complexity rating,
  // but standalone patterns are almost always prefered, so we need to adjust the
  // priority lower.  The goal is to use a high number to reduce complexity to
  // zero (or less than zero).
  let AddedComplexity = -1000;

  let VOP3 = 1;
  let VALU = 1;

  let AsmMatchConverter = "cvtVOP3";
  let isCodeGenOnly = 0;

  int Size = 8;

  // Because SGPRs may be allowed if there are multiple operands, we
  // need a post-isel hook to insert copies in order to avoid
  // violating constant bus requirements.
  let hasPostISelHook = 1;
}

} // End Uses = [EXEC]

//===----------------------------------------------------------------------===//
// Scalar operations
//===----------------------------------------------------------------------===//

class SOP1e <bits<8> op> : Enc32 {
  bits<7> sdst;
  bits<8> ssrc0;

  let Inst{7-0} = ssrc0;
  let Inst{15-8} = op;
  let Inst{22-16} = sdst;
  let Inst{31-23} = 0x17d; //encoding;
}

class SOP2e <bits<7> op> : Enc32 {
  bits<7> sdst;
  bits<8> ssrc0;
  bits<8> ssrc1;

  let Inst{7-0} = ssrc0;
  let Inst{15-8} = ssrc1;
  let Inst{22-16} = sdst;
  let Inst{29-23} = op;
  let Inst{31-30} = 0x2; // encoding
}

class SOPCe <bits<7> op> : Enc32 {
  bits<8> ssrc0;
  bits<8> ssrc1;

  let Inst{7-0} = ssrc0;
  let Inst{15-8} = ssrc1;
  let Inst{22-16} = op;
  let Inst{31-23} = 0x17e;
}

class SOPKe <bits<5> op> : Enc32 {
  bits <7> sdst;
  bits <16> simm16;

  let Inst{15-0} = simm16;
  let Inst{22-16} = sdst;
  let Inst{27-23} = op;
  let Inst{31-28} = 0xb; //encoding
}

class SOPK64e <bits<5> op> : Enc64 {
  bits <7> sdst = 0;
  bits <16> simm16;
  bits <32> imm;

  let Inst{15-0} = simm16;
  let Inst{22-16} = sdst;
  let Inst{27-23} = op;
  let Inst{31-28} = 0xb;

  let Inst{63-32} = imm;
}

class SOPPe <bits<7> op> : Enc32 {
  bits <16> simm16;

  let Inst{15-0} = simm16;
  let Inst{22-16} = op;
  let Inst{31-23} = 0x17f; // encoding
}

class SMRDe <bits<5> op, bits<1> imm> : Enc32 {
  bits<7> sdst;
  bits<7> sbase;
  bits<8> offset;

  let Inst{7-0} = offset;
  let Inst{8} = imm;
  let Inst{14-9} = sbase{6-1};
  let Inst{21-15} = sdst;
  let Inst{26-22} = op;
  let Inst{31-27} = 0x18; //encoding
}

class SMRD_IMMe_ci <bits<5> op> : Enc64 {
  bits<7> sdst;
  bits<7> sbase;
  bits<32> offset;

  let Inst{7-0}   = 0xff;
  let Inst{8}     = 0;
  let Inst{14-9}  = sbase{6-1};
  let Inst{21-15} = sdst;
  let Inst{26-22} = op;
  let Inst{31-27} = 0x18; //encoding
  let Inst{63-32} = offset;
}

let SchedRW = [WriteSALU] in {
class SOP1 <dag outs, dag ins, string asm, list<dag> pattern> :
    InstSI<outs, ins, asm, pattern> {
  let mayLoad = 0;
  let mayStore = 0;
  let hasSideEffects = 0;
  let isCodeGenOnly = 0;
  let SALU = 1;
  let SOP1 = 1;
}

class SOP2 <dag outs, dag ins, string asm, list<dag> pattern> :
    InstSI <outs, ins, asm, pattern> {

  let mayLoad = 0;
  let mayStore = 0;
  let hasSideEffects = 0;
  let isCodeGenOnly = 0;
  let SALU = 1;
  let SOP2 = 1;

  let UseNamedOperandTable = 1;
}

class SOPC <bits<7> op, dag outs, dag ins, string asm, list<dag> pattern> :
  InstSI<outs, ins, asm, pattern>, SOPCe <op> {

  let mayLoad = 0;
  let mayStore = 0;
  let hasSideEffects = 0;
  let SALU = 1;
  let SOPC = 1;
  let isCodeGenOnly = 0;
  let Defs = [SCC];

  let UseNamedOperandTable = 1;
}

class SOPK <dag outs, dag ins, string asm, list<dag> pattern> :
   InstSI <outs, ins , asm, pattern> {

  let mayLoad = 0;
  let mayStore = 0;
  let hasSideEffects = 0;
  let SALU = 1;
  let SOPK = 1;

  let UseNamedOperandTable = 1;
}

class SOPP <bits<7> op, dag ins, string asm, list<dag> pattern = []> :
		InstSI <(outs), ins, asm, pattern >, SOPPe <op> {

  let mayLoad = 0;
  let mayStore = 0;
  let hasSideEffects = 0;
  let SALU = 1;
  let SOPP = 1;

  let UseNamedOperandTable = 1;
}

} // let SchedRW = [WriteSALU]

class SMRD <dag outs, dag ins, string asm, list<dag> pattern> :
    InstSI<outs, ins, asm, pattern> {

  let LGKM_CNT = 1;
  let SMRD = 1;
  let mayStore = 0;
  let mayLoad = 1;
  let hasSideEffects = 0;
  let UseNamedOperandTable = 1;
  let SchedRW = [WriteSMEM];
}

//===----------------------------------------------------------------------===//
// Vector ALU operations
//===----------------------------------------------------------------------===//

class VOP1e <bits<8> op> : Enc32 {
  bits<8> vdst;
  bits<9> src0;

  let Inst{8-0} = src0;
  let Inst{16-9} = op;
  let Inst{24-17} = vdst;
  let Inst{31-25} = 0x3f; //encoding
}

class VOP2e <bits<6> op> : Enc32 {
  bits<8> vdst;
  bits<9> src0;
  bits<8> src1;

  let Inst{8-0} = src0;
  let Inst{16-9} = src1;
  let Inst{24-17} = vdst;
  let Inst{30-25} = op;
  let Inst{31} = 0x0; //encoding
}

class VOP2_MADKe <bits<6> op> : Enc64 {

  bits<8>  vdst;
  bits<9>  src0;
  bits<8>  vsrc1;
  bits<32> src2;

  let Inst{8-0} = src0;
  let Inst{16-9} = vsrc1;
  let Inst{24-17} = vdst;
  let Inst{30-25} = op;
  let Inst{31} = 0x0; // encoding
  let Inst{63-32} = src2;
}

class VOP3e <bits<9> op> : Enc64 {
  bits<8> vdst;
  bits<2> src0_modifiers;
  bits<9> src0;
  bits<2> src1_modifiers;
  bits<9> src1;
  bits<2> src2_modifiers;
  bits<9> src2;
  bits<1> clamp;
  bits<2> omod;

  let Inst{7-0} = vdst;
  let Inst{8} = src0_modifiers{1};
  let Inst{9} = src1_modifiers{1};
  let Inst{10} = src2_modifiers{1};
  let Inst{11} = clamp;
  let Inst{25-17} = op;
  let Inst{31-26} = 0x34; //encoding
  let Inst{40-32} = src0;
  let Inst{49-41} = src1;
  let Inst{58-50} = src2;
  let Inst{60-59} = omod;
  let Inst{61} = src0_modifiers{0};
  let Inst{62} = src1_modifiers{0};
  let Inst{63} = src2_modifiers{0};
}

class VOP3be <bits<9> op> : Enc64 {
  bits<8> vdst;
  bits<2> src0_modifiers;
  bits<9> src0;
  bits<2> src1_modifiers;
  bits<9> src1;
  bits<2> src2_modifiers;
  bits<9> src2;
  bits<7> sdst;
  bits<2> omod;

  let Inst{7-0} = vdst;
  let Inst{14-8} = sdst;
  let Inst{25-17} = op;
  let Inst{31-26} = 0x34; //encoding
  let Inst{40-32} = src0;
  let Inst{49-41} = src1;
  let Inst{58-50} = src2;
  let Inst{60-59} = omod;
  let Inst{61} = src0_modifiers{0};
  let Inst{62} = src1_modifiers{0};
  let Inst{63} = src2_modifiers{0};
}

class VOPCe <bits<8> op> : Enc32 {
  bits<9> src0;
  bits<8> vsrc1;

  let Inst{8-0} = src0;
  let Inst{16-9} = vsrc1;
  let Inst{24-17} = op;
  let Inst{31-25} = 0x3e;
}

class VINTRPe <bits<2> op> : Enc32 {
  bits<8> vdst;
  bits<8> vsrc;
  bits<2> attrchan;
  bits<6> attr;

  let Inst{7-0} = vsrc;
  let Inst{9-8} = attrchan;
  let Inst{15-10} = attr;
  let Inst{17-16} = op;
  let Inst{25-18} = vdst;
  let Inst{31-26} = 0x32; // encoding
}

class DSe <bits<8> op> : Enc64 {
  bits<8> vdst;
  bits<1> gds;
  bits<8> addr;
  bits<8> data0;
  bits<8> data1;
  bits<8> offset0;
  bits<8> offset1;

  let Inst{7-0} = offset0;
  let Inst{15-8} = offset1;
  let Inst{17} = gds;
  let Inst{25-18} = op;
  let Inst{31-26} = 0x36; //encoding
  let Inst{39-32} = addr;
  let Inst{47-40} = data0;
  let Inst{55-48} = data1;
  let Inst{63-56} = vdst;
}

class MUBUFe <bits<7> op> : Enc64 {
  bits<12> offset;
  bits<1> offen;
  bits<1> idxen;
  bits<1> glc;
  bits<1> addr64;
  bits<1> lds;
  bits<8> vaddr;
  bits<8> vdata;
  bits<7> srsrc;
  bits<1> slc;
  bits<1> tfe;
  bits<8> soffset;

  let Inst{11-0} = offset;
  let Inst{12} = offen;
  let Inst{13} = idxen;
  let Inst{14} = glc;
  let Inst{15} = addr64;
  let Inst{16} = lds;
  let Inst{24-18} = op;
  let Inst{31-26} = 0x38; //encoding
  let Inst{39-32} = vaddr;
  let Inst{47-40} = vdata;
  let Inst{52-48} = srsrc{6-2};
  let Inst{54} = slc;
  let Inst{55} = tfe;
  let Inst{63-56} = soffset;
}

class MTBUFe <bits<3> op> : Enc64 {
  bits<8> vdata;
  bits<12> offset;
  bits<1> offen;
  bits<1> idxen;
  bits<1> glc;
  bits<1> addr64;
  bits<4> dfmt;
  bits<3> nfmt;
  bits<8> vaddr;
  bits<7> srsrc;
  bits<1> slc;
  bits<1> tfe;
  bits<8> soffset;

  let Inst{11-0} = offset;
  let Inst{12} = offen;
  let Inst{13} = idxen;
  let Inst{14} = glc;
  let Inst{15} = addr64;
  let Inst{18-16} = op;
  let Inst{22-19} = dfmt;
  let Inst{25-23} = nfmt;
  let Inst{31-26} = 0x3a; //encoding
  let Inst{39-32} = vaddr;
  let Inst{47-40} = vdata;
  let Inst{52-48} = srsrc{6-2};
  let Inst{54} = slc;
  let Inst{55} = tfe;
  let Inst{63-56} = soffset;
}

class MIMGe <bits<7> op> : Enc64 {
  bits<8> vdata;
  bits<4> dmask;
  bits<1> unorm;
  bits<1> glc;
  bits<1> da;
  bits<1> r128;
  bits<1> tfe;
  bits<1> lwe;
  bits<1> slc;
  bits<8> vaddr;
  bits<7> srsrc;
  bits<7> ssamp;

  let Inst{11-8} = dmask;
  let Inst{12} = unorm;
  let Inst{13} = glc;
  let Inst{14} = da;
  let Inst{15} = r128;
  let Inst{16} = tfe;
  let Inst{17} = lwe;
  let Inst{24-18} = op;
  let Inst{25} = slc;
  let Inst{31-26} = 0x3c;
  let Inst{39-32} = vaddr;
  let Inst{47-40} = vdata;
  let Inst{52-48} = srsrc{6-2};
  let Inst{57-53} = ssamp{6-2};
}

class FLATe<bits<7> op> : Enc64 {
  bits<8> addr;
  bits<8> data;
  bits<8> vdst;
  bits<1> slc;
  bits<1> glc;
  bits<1> tfe;

  // 15-0 is reserved.
  let Inst{16} = glc;
  let Inst{17} = slc;
  let Inst{24-18} = op;
  let Inst{31-26} = 0x37; // Encoding.
  let Inst{39-32} = addr;
  let Inst{47-40} = data;
  // 54-48 is reserved.
  let Inst{55} = tfe;
  let Inst{63-56} = vdst;
}

class EXPe : Enc64 {
  bits<4> en;
  bits<6> tgt;
  bits<1> compr;
  bits<1> done;
  bits<1> vm;
  bits<8> vsrc0;
  bits<8> vsrc1;
  bits<8> vsrc2;
  bits<8> vsrc3;

  let Inst{3-0} = en;
  let Inst{9-4} = tgt;
  let Inst{10} = compr;
  let Inst{11} = done;
  let Inst{12} = vm;
  let Inst{31-26} = 0x3e;
  let Inst{39-32} = vsrc0;
  let Inst{47-40} = vsrc1;
  let Inst{55-48} = vsrc2;
  let Inst{63-56} = vsrc3;
}

let Uses = [EXEC] in {

class VOP1 <bits<8> op, dag outs, dag ins, string asm, list<dag> pattern> :
    VOP1Common <outs, ins, asm, pattern>,
    VOP1e<op> {
  let isCodeGenOnly = 0;
}

class VOP2 <bits<6> op, dag outs, dag ins, string asm, list<dag> pattern> :
    VOP2Common <outs, ins, asm, pattern>, VOP2e<op> {
  let isCodeGenOnly = 0;
}

class VOPC <bits<8> op, dag ins, string asm, list<dag> pattern> :
    VOPCCommon <ins, asm, pattern>, VOPCe <op>;

class VINTRPCommon <dag outs, dag ins, string asm, list<dag> pattern> :
    InstSI <outs, ins, asm, pattern> {
  let mayLoad = 1;
  let mayStore = 0;
  let hasSideEffects = 0;
}

} // End Uses = [EXEC]

//===----------------------------------------------------------------------===//
// Vector I/O operations
//===----------------------------------------------------------------------===//

class DS <dag outs, dag ins, string asm, list<dag> pattern> :
    InstSI <outs, ins, asm, pattern> {

  let LGKM_CNT = 1;
  let DS = 1;
  let UseNamedOperandTable = 1;
  let Uses = [M0, EXEC];

  // Most instruction load and store data, so set this as the default.
  let mayLoad = 1;
  let mayStore = 1;

  let hasSideEffects = 0;
  let AsmMatchConverter = "cvtDS";
  let SchedRW = [WriteLDS];
}

class MUBUF <dag outs, dag ins, string asm, list<dag> pattern> :
    InstSI<outs, ins, asm, pattern> {

  let VM_CNT = 1;
  let EXP_CNT = 1;
  let MUBUF = 1;
  let Uses = [EXEC];

  let hasSideEffects = 0;
  let UseNamedOperandTable = 1;
  let AsmMatchConverter = "cvtMubuf";
  let SchedRW = [WriteVMEM];
}

class MTBUF <dag outs, dag ins, string asm, list<dag> pattern> :
    InstSI<outs, ins, asm, pattern> {

  let VM_CNT = 1;
  let EXP_CNT = 1;
  let MTBUF = 1;
  let Uses = [EXEC];

  let hasSideEffects = 0;
  let UseNamedOperandTable = 1;
  let SchedRW = [WriteVMEM];
}

class FLAT <bits<7> op, dag outs, dag ins, string asm, list<dag> pattern> :
    InstSI<outs, ins, asm, pattern>, FLATe <op> {
  let FLAT = 1;
  // Internally, FLAT instruction are executed as both an LDS and a
  // Buffer instruction; so, they increment both VM_CNT and LGKM_CNT
  // and are not considered done until both have been decremented.
  let VM_CNT = 1;
  let LGKM_CNT = 1;

  let Uses = [EXEC, FLAT_SCR]; // M0

  let UseNamedOperandTable = 1;
  let hasSideEffects = 0;
  let AsmMatchConverter = "cvtFlat";
  let SchedRW = [WriteVMEM];
}

class MIMG <bits<7> op, dag outs, dag ins, string asm, list<dag> pattern> :
    InstSI <outs, ins, asm, pattern>, MIMGe <op> {

  let VM_CNT = 1;
  let EXP_CNT = 1;
  let MIMG = 1;
  let Uses = [EXEC];

  let UseNamedOperandTable = 1;
  let hasSideEffects = 0; // XXX ????
}