llvm.org GIT mirror llvm / 0392724 lib / Target / AMDGPU / VOP3Instructions.td
0392724

Tree @0392724 (Download .tar.gz)

VOP3Instructions.td @0392724raw · 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
 693
 694
 695
 696
 697
 698
 699
 700
 701
 702
 703
 704
 705
 706
 707
 708
 709
 710
 711
 712
 713
 714
 715
 716
 717
 718
 719
 720
 721
 722
 723
 724
 725
 726
 727
 728
 729
 730
 731
 732
 733
 734
 735
 736
 737
 738
 739
 740
 741
 742
 743
 744
 745
 746
 747
 748
 749
 750
 751
 752
 753
 754
 755
 756
 757
 758
 759
 760
 761
 762
 763
 764
 765
 766
 767
 768
 769
 770
 771
 772
 773
 774
 775
 776
 777
 778
 779
 780
 781
 782
 783
 784
 785
 786
 787
 788
 789
 790
 791
 792
 793
 794
 795
 796
 797
 798
 799
 800
 801
 802
 803
 804
 805
 806
 807
 808
 809
 810
 811
 812
 813
 814
 815
 816
 817
 818
 819
 820
 821
 822
 823
 824
 825
 826
 827
 828
 829
 830
 831
 832
 833
 834
 835
 836
 837
 838
 839
 840
 841
 842
 843
 844
 845
 846
 847
 848
 849
 850
 851
 852
 853
 854
 855
 856
 857
 858
 859
 860
 861
 862
 863
 864
 865
 866
 867
 868
 869
 870
 871
 872
 873
 874
 875
 876
 877
 878
 879
 880
 881
 882
 883
 884
 885
 886
 887
 888
 889
 890
 891
 892
 893
 894
 895
 896
 897
 898
 899
 900
 901
 902
 903
 904
 905
 906
 907
 908
 909
 910
 911
 912
 913
 914
 915
 916
 917
 918
 919
 920
 921
 922
 923
 924
 925
 926
 927
 928
 929
 930
 931
 932
 933
 934
 935
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
//===-- VOP3Instructions.td - Vector Instruction Defintions ---------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//

//===----------------------------------------------------------------------===//
// VOP3 Classes
//===----------------------------------------------------------------------===//

class getVOP3ModPat<VOPProfile P, SDPatternOperator node> {
  dag src0 = !if(P.HasOMod,
    (VOP3Mods0 P.Src0VT:$src0, i32:$src0_modifiers, i1:$clamp, i32:$omod),
    (VOP3Mods0 P.Src0VT:$src0, i32:$src0_modifiers, i1:$clamp));

  list<dag> ret3 = [(set P.DstVT:$vdst,
    (DivergentFragOrOp<node, P>.ret (P.Src0VT src0),
          (P.Src1VT (VOP3Mods P.Src1VT:$src1, i32:$src1_modifiers)),
          (P.Src2VT (VOP3Mods P.Src2VT:$src2, i32:$src2_modifiers))))];

  list<dag> ret2 = [(set P.DstVT:$vdst,
    (DivergentFragOrOp<node, P>.ret (P.Src0VT src0),
          (P.Src1VT (VOP3Mods P.Src1VT:$src1, i32:$src1_modifiers))))];

  list<dag> ret1 = [(set P.DstVT:$vdst,
    (DivergentFragOrOp<node, P>.ret (P.Src0VT src0)))];

  list<dag> ret = !if(!eq(P.NumSrcArgs, 3), ret3,
                  !if(!eq(P.NumSrcArgs, 2), ret2,
                  ret1));
}

class getVOP3PModPat<VOPProfile P, SDPatternOperator node> {
  list<dag> ret3 = [(set P.DstVT:$vdst,
    (DivergentFragOrOp<node, P>.ret (P.Src0VT !if(P.HasClamp, (VOP3PMods0 P.Src0VT:$src0, i32:$src0_modifiers, i1:$clamp),
                                    (VOP3PMods P.Src0VT:$src0, i32:$src0_modifiers))),
          (P.Src1VT (VOP3PMods P.Src1VT:$src1, i32:$src1_modifiers)),
          (P.Src2VT (VOP3PMods P.Src2VT:$src2, i32:$src2_modifiers))))];

  list<dag> ret2 = [(set P.DstVT:$vdst,
    (DivergentFragOrOp<node, P>.ret !if(P.HasClamp, (P.Src0VT (VOP3PMods0 P.Src0VT:$src0, i32:$src0_modifiers, i1:$clamp)),
                          (P.Src0VT (VOP3PMods P.Src0VT:$src0, i32:$src0_modifiers))),
          (P.Src1VT (VOP3PMods P.Src1VT:$src1, i32:$src1_modifiers))))];

  list<dag> ret1 = [(set P.DstVT:$vdst,
    (DivergentFragOrOp<node, P>.ret (P.Src0VT (VOP3PMods0 P.Src0VT:$src0, i32:$src0_modifiers, i1:$clamp))))];

  list<dag> ret = !if(!eq(P.NumSrcArgs, 3), ret3,
                  !if(!eq(P.NumSrcArgs, 2), ret2,
                  ret1));
}

class getVOP3OpSelPat<VOPProfile P, SDPatternOperator node> {
  list<dag> ret3 = [(set P.DstVT:$vdst,
    (DivergentFragOrOp<node, P>.ret (P.Src0VT !if(P.HasClamp, (VOP3OpSel0 P.Src0VT:$src0, i32:$src0_modifiers, i1:$clamp),
                                    (VOP3OpSel P.Src0VT:$src0, i32:$src0_modifiers))),
          (P.Src1VT (VOP3OpSel P.Src1VT:$src1, i32:$src1_modifiers)),
          (P.Src2VT (VOP3OpSel P.Src2VT:$src2, i32:$src2_modifiers))))];

  list<dag> ret2 = [(set P.DstVT:$vdst,
    (DivergentFragOrOp<node, P>.ret !if(P.HasClamp, (P.Src0VT (VOP3OpSel0 P.Src0VT:$src0, i32:$src0_modifiers, i1:$clamp)),
                          (P.Src0VT (VOP3OpSel P.Src0VT:$src0, i32:$src0_modifiers))),
          (P.Src1VT (VOP3OpSel P.Src1VT:$src1, i32:$src1_modifiers))))];

  list<dag> ret1 = [(set P.DstVT:$vdst,
    (DivergentFragOrOp<node, P>.ret (P.Src0VT (VOP3OpSel0 P.Src0VT:$src0, i32:$src0_modifiers, i1:$clamp))))];

  list<dag> ret = !if(!eq(P.NumSrcArgs, 3), ret3,
                  !if(!eq(P.NumSrcArgs, 2), ret2,
                  ret1));
}

class getVOP3OpSelModPat<VOPProfile P, SDPatternOperator node> {
  list<dag> ret3 = [(set P.DstVT:$vdst,
    (DivergentFragOrOp<node, P>.ret (P.Src0VT !if(P.HasClamp, (VOP3OpSelMods0 P.Src0VT:$src0, i32:$src0_modifiers, i1:$clamp),
                                    (VOP3OpSelMods P.Src0VT:$src0, i32:$src0_modifiers))),
          (P.Src1VT (VOP3OpSelMods P.Src1VT:$src1, i32:$src1_modifiers)),
          (P.Src2VT (VOP3OpSelMods P.Src2VT:$src2, i32:$src2_modifiers))))];

  list<dag> ret2 = [(set P.DstVT:$vdst,
    (DivergentFragOrOp<node, P>.ret !if(P.HasClamp, (P.Src0VT (VOP3OpSelMods0 P.Src0VT:$src0, i32:$src0_modifiers, i1:$clamp)),
                          (P.Src0VT (VOP3OpSelMods P.Src0VT:$src0, i32:$src0_modifiers))),
          (P.Src1VT (VOP3OpSelMods P.Src1VT:$src1, i32:$src1_modifiers))))];

  list<dag> ret1 = [(set P.DstVT:$vdst,
    (DivergentFragOrOp<node, P>.ret (P.Src0VT (VOP3OpSelMods0 P.Src0VT:$src0, i32:$src0_modifiers, i1:$clamp))))];

  list<dag> ret = !if(!eq(P.NumSrcArgs, 3), ret3,
                  !if(!eq(P.NumSrcArgs, 2), ret2,
                  ret1));
}

class getVOP3Pat<VOPProfile P, SDPatternOperator node> {
  list<dag> ret3 = [(set P.DstVT:$vdst, (DivergentFragOrOp<node, P>.ret P.Src0VT:$src0, P.Src1VT:$src1, P.Src2VT:$src2))];
  list<dag> ret2 = [(set P.DstVT:$vdst, (DivergentFragOrOp<node, P>.ret P.Src0VT:$src0, P.Src1VT:$src1))];
  list<dag> ret1 = [(set P.DstVT:$vdst, (DivergentFragOrOp<node, P>.ret P.Src0VT:$src0))];
  list<dag> ret = !if(!eq(P.NumSrcArgs, 3), ret3,
                  !if(!eq(P.NumSrcArgs, 2), ret2,
                  ret1));
}

class getVOP3ClampPat<VOPProfile P, SDPatternOperator node> {
  list<dag> ret3 = [(set P.DstVT:$vdst, (node P.Src0VT:$src0, P.Src1VT:$src1, P.Src2VT:$src2, i1:$clamp))];
  list<dag> ret2 = [(set P.DstVT:$vdst, (node P.Src0VT:$src0, P.Src1VT:$src1, i1:$clamp))];
  list<dag> ret1 = [(set P.DstVT:$vdst, (node P.Src0VT:$src0, i1:$clamp))];
  list<dag> ret = !if(!eq(P.NumSrcArgs, 3), ret3,
                  !if(!eq(P.NumSrcArgs, 2), ret2,
                  ret1));
}

class VOP3Inst<string OpName, VOPProfile P, SDPatternOperator node = null_frag, bit VOP3Only = 0> :
  VOP3_Pseudo<OpName, P,
    !if(P.HasOpSel,
        !if(P.HasModifiers,
            getVOP3OpSelModPat<P, node>.ret,
            getVOP3OpSelPat<P, node>.ret),
        !if(P.HasModifiers,
            getVOP3ModPat<P, node>.ret,
            !if(P.HasIntClamp,
                getVOP3ClampPat<P, node>.ret,
                getVOP3Pat<P, node>.ret))),
    VOP3Only, 0, P.HasOpSel> {

  let IntClamp = P.HasIntClamp;
  let AsmMatchConverter =
    !if(P.HasOpSel,
        "cvtVOP3OpSel",
        !if(!or(P.HasModifiers, !or(P.HasOMod, P.HasIntClamp)),
            "cvtVOP3",
            ""));
}

// Special case for v_div_fmas_{f32|f64}, since it seems to be the
// only VOP instruction that implicitly reads VCC.
let Asm64 = " $vdst, $src0_modifiers, $src1_modifiers, $src2_modifiers$clamp$omod" in {
def VOP_F32_F32_F32_F32_VCC : VOPProfile<[f32, f32, f32, f32]> {
  let Outs64 = (outs DstRC.RegClass:$vdst);
}
def VOP_F64_F64_F64_F64_VCC : VOPProfile<[f64, f64, f64, f64]> {
  let Outs64 = (outs DstRC.RegClass:$vdst);
}
}

class VOP3Features<bit Clamp, bit OpSel, bit Packed> {
  bit HasClamp = Clamp;
  bit HasOpSel = OpSel;
  bit IsPacked = Packed;
}

def VOP3_REGULAR : VOP3Features<0, 0, 0>;
def VOP3_CLAMP   : VOP3Features<1, 0, 0>;
def VOP3_OPSEL   : VOP3Features<1, 1, 0>;
def VOP3_PACKED  : VOP3Features<1, 1, 1>;

class VOP3_Profile<VOPProfile P, VOP3Features Features = VOP3_REGULAR> : VOPProfile<P.ArgVT> {

  let HasClamp = !if(Features.HasClamp, 1, P.HasClamp);
  let HasOpSel = !if(Features.HasOpSel, 1, P.HasOpSel);
  let IsPacked = !if(Features.IsPacked, 1, P.IsPacked);

  let HasModifiers = !if(Features.IsPacked, 1, P.HasModifiers);

  // FIXME: Hack to stop printing _e64
  let Outs64 = (outs DstRC.RegClass:$vdst);
  let Asm64 =
    " " # !if(Features.HasOpSel,
              getAsmVOP3OpSel<NumSrcArgs,
                              HasIntClamp,
                              HasSrc0FloatMods,
                              HasSrc1FloatMods,
                              HasSrc2FloatMods>.ret,
              !if(Features.HasClamp,
                  getAsm64<HasDst, NumSrcArgs, HasIntClamp,
                           HasModifiers, HasOMod, DstVT>.ret,
                  P.Asm64));
  let NeedPatGen = P.NeedPatGen;
}

class VOP3b_Profile<ValueType vt> : VOPProfile<[vt, vt, vt, vt]> {
  // v_div_scale_{f32|f64} do not support input modifiers.
  let HasModifiers = 0;
  let HasClamp = 0;
  let HasOMod = 0;
  let Outs64 = (outs DstRC:$vdst, VOPDstS64orS32:$sdst);
  let Asm64 = " $vdst, $sdst, $src0, $src1, $src2";
}

def VOP3b_F32_I1_F32_F32_F32 : VOP3b_Profile<f32> {
  // FIXME: Hack to stop printing _e64
  let DstRC = RegisterOperand<VGPR_32>;
}

def VOP3b_F64_I1_F64_F64_F64 : VOP3b_Profile<f64> {
  // FIXME: Hack to stop printing _e64
  let DstRC = RegisterOperand<VReg_64>;
}

def VOP3b_I64_I1_I32_I32_I64 : VOPProfile<[i64, i32, i32, i64]> {
  let HasClamp = 1;

  // FIXME: Hack to stop printing _e64
  let DstRC = RegisterOperand<VReg_64>;

  let Outs64 = (outs DstRC:$vdst, VOPDstS64orS32:$sdst);
  let Asm64 = " $vdst, $sdst, $src0, $src1, $src2$clamp";
}

//===----------------------------------------------------------------------===//
// VOP3 INTERP
//===----------------------------------------------------------------------===//

class VOP3Interp<string OpName, VOPProfile P, list<dag> pattern = []> :
                 VOP3_Pseudo<OpName, P, pattern> {
  let AsmMatchConverter = "cvtVOP3Interp";
}

def VOP3_INTERP : VOPProfile<[f32, f32, i32, untyped]> {
  let Ins64 = (ins Src0Mod:$src0_modifiers, VRegSrc_32:$src0,
                   Attr:$attr, AttrChan:$attrchan,
                   clampmod:$clamp, omod:$omod);

  let Asm64 = "$vdst, $src0_modifiers, $attr$attrchan$clamp$omod";
}

def VOP3_INTERP_MOV : VOPProfile<[f32, i32, i32, untyped]> {
  let Ins64 = (ins InterpSlot:$src0,
                   Attr:$attr, AttrChan:$attrchan,
                   clampmod:$clamp, omod:$omod);

  let Asm64 = "$vdst, $src0, $attr$attrchan$clamp$omod";

  let HasClamp = 1;
}

class getInterp16Asm <bit HasSrc2, bit HasOMod> {
  string src2 = !if(HasSrc2, ", $src2_modifiers", "");
  string omod = !if(HasOMod, "$omod", "");
  string ret =
    " $vdst, $src0_modifiers, $attr$attrchan"#src2#"$high$clamp"#omod;
}

class getInterp16Ins <bit HasSrc2, bit HasOMod,
                      Operand Src0Mod, Operand Src2Mod> {
  dag ret = !if(HasSrc2,
                !if(HasOMod,
                    (ins Src0Mod:$src0_modifiers, VRegSrc_32:$src0,
                         Attr:$attr, AttrChan:$attrchan,
                         Src2Mod:$src2_modifiers, VRegSrc_32:$src2,
                         highmod:$high, clampmod:$clamp, omod:$omod),
                    (ins Src0Mod:$src0_modifiers, VRegSrc_32:$src0,
                         Attr:$attr, AttrChan:$attrchan,
                         Src2Mod:$src2_modifiers, VRegSrc_32:$src2,
                         highmod:$high, clampmod:$clamp)
                ),
                (ins Src0Mod:$src0_modifiers, VRegSrc_32:$src0,
                     Attr:$attr, AttrChan:$attrchan,
                     highmod:$high, clampmod:$clamp, omod:$omod)
            );
}

class VOP3_INTERP16 <list<ValueType> ArgVT> : VOPProfile<ArgVT> {

  let HasOMod = !if(!eq(DstVT.Value, f16.Value), 0, 1);
  let HasHigh = 1;

  let Outs64 = (outs VGPR_32:$vdst);
  let Ins64 = getInterp16Ins<HasSrc2, HasOMod, Src0Mod, Src2Mod>.ret;
  let Asm64 = getInterp16Asm<HasSrc2, HasOMod>.ret;
}

//===----------------------------------------------------------------------===//
// VOP3 Instructions
//===----------------------------------------------------------------------===//

let isCommutable = 1 in {

def V_MAD_LEGACY_F32 : VOP3Inst <"v_mad_legacy_f32", VOP3_Profile<VOP_F32_F32_F32_F32>>;
def V_MAD_F32 : VOP3Inst <"v_mad_f32", VOP3_Profile<VOP_F32_F32_F32_F32>, fmad>;
def V_MAD_I32_I24 : VOP3Inst <"v_mad_i32_i24", VOP3_Profile<VOP_I32_I32_I32_I32, VOP3_CLAMP>>;
def V_MAD_U32_U24 : VOP3Inst <"v_mad_u32_u24", VOP3_Profile<VOP_I32_I32_I32_I32, VOP3_CLAMP>>;
def V_FMA_F32 : VOP3Inst <"v_fma_f32", VOP3_Profile<VOP_F32_F32_F32_F32>, fma>;
def V_LERP_U8 : VOP3Inst <"v_lerp_u8", VOP3_Profile<VOP_I32_I32_I32_I32>, int_amdgcn_lerp>;

let SchedRW = [WriteDoubleAdd] in {
let FPDPRounding = 1 in {
def V_FMA_F64 : VOP3Inst <"v_fma_f64", VOP3_Profile<VOP_F64_F64_F64_F64>, fma>;
def V_ADD_F64 : VOP3Inst <"v_add_f64", VOP3_Profile<VOP_F64_F64_F64>, fadd, 1>;
def V_MUL_F64 : VOP3Inst <"v_mul_f64", VOP3_Profile<VOP_F64_F64_F64>, fmul, 1>;
} // End FPDPRounding = 1
def V_MIN_F64 : VOP3Inst <"v_min_f64", VOP3_Profile<VOP_F64_F64_F64>, fminnum_like, 1>;
def V_MAX_F64 : VOP3Inst <"v_max_f64", VOP3_Profile<VOP_F64_F64_F64>, fmaxnum_like, 1>;
} // End SchedRW = [WriteDoubleAdd]

let SchedRW = [WriteQuarterRate32] in {
def V_MUL_LO_U32 : VOP3Inst <"v_mul_lo_u32", VOP3_Profile<VOP_I32_I32_I32>, mul>;
def V_MUL_HI_U32 : VOP3Inst <"v_mul_hi_u32", VOP3_Profile<VOP_I32_I32_I32>, mulhu>;
def V_MUL_LO_I32 : VOP3Inst <"v_mul_lo_i32", VOP3_Profile<VOP_I32_I32_I32>>;
def V_MUL_HI_I32 : VOP3Inst <"v_mul_hi_i32", VOP3_Profile<VOP_I32_I32_I32>, mulhs>;
} // End SchedRW = [WriteQuarterRate32]

let Uses = [VCC, EXEC] in {
// v_div_fmas_f32:
//   result = src0 * src1 + src2
//   if (vcc)
//     result *= 2^32
//
def V_DIV_FMAS_F32 : VOP3_Pseudo <"v_div_fmas_f32", VOP_F32_F32_F32_F32_VCC, []> {
  let SchedRW = [WriteFloatFMA];
}
// v_div_fmas_f64:
//   result = src0 * src1 + src2
//   if (vcc)
//     result *= 2^64
//
def V_DIV_FMAS_F64 : VOP3_Pseudo <"v_div_fmas_f64", VOP_F64_F64_F64_F64_VCC, []> {
  let SchedRW = [WriteDouble];
  let FPDPRounding = 1;
}
} // End Uses = [VCC, EXEC]

} // End isCommutable = 1

def V_CUBEID_F32 : VOP3Inst <"v_cubeid_f32", VOP3_Profile<VOP_F32_F32_F32_F32>, int_amdgcn_cubeid>;
def V_CUBESC_F32 : VOP3Inst <"v_cubesc_f32", VOP3_Profile<VOP_F32_F32_F32_F32>, int_amdgcn_cubesc>;
def V_CUBETC_F32 : VOP3Inst <"v_cubetc_f32", VOP3_Profile<VOP_F32_F32_F32_F32>, int_amdgcn_cubetc>;
def V_CUBEMA_F32 : VOP3Inst <"v_cubema_f32", VOP3_Profile<VOP_F32_F32_F32_F32>, int_amdgcn_cubema>;
def V_BFE_U32 : VOP3Inst <"v_bfe_u32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUbfe_u32>;
def V_BFE_I32 : VOP3Inst <"v_bfe_i32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUbfe_i32>;
def V_BFI_B32 : VOP3Inst <"v_bfi_b32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUbfi>;
def V_ALIGNBIT_B32 : VOP3Inst <"v_alignbit_b32", VOP3_Profile<VOP_I32_I32_I32_I32>, int_amdgcn_alignbit>;
def V_ALIGNBYTE_B32 : VOP3Inst <"v_alignbyte_b32", VOP3_Profile<VOP_I32_I32_I32_I32>, int_amdgcn_alignbyte>;
def V_MIN3_F32 : VOP3Inst <"v_min3_f32", VOP3_Profile<VOP_F32_F32_F32_F32>, AMDGPUfmin3>;
def V_MIN3_I32 : VOP3Inst <"v_min3_i32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUsmin3>;
def V_MIN3_U32 : VOP3Inst <"v_min3_u32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUumin3>;
def V_MAX3_F32 : VOP3Inst <"v_max3_f32", VOP3_Profile<VOP_F32_F32_F32_F32>, AMDGPUfmax3>;
def V_MAX3_I32 : VOP3Inst <"v_max3_i32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUsmax3>;
def V_MAX3_U32 : VOP3Inst <"v_max3_u32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUumax3>;
def V_MED3_F32 : VOP3Inst <"v_med3_f32", VOP3_Profile<VOP_F32_F32_F32_F32>, AMDGPUfmed3>;
def V_MED3_I32 : VOP3Inst <"v_med3_i32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUsmed3>;
def V_MED3_U32 : VOP3Inst <"v_med3_u32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUumed3>;
def V_SAD_U8 : VOP3Inst <"v_sad_u8", VOP3_Profile<VOP_I32_I32_I32_I32, VOP3_CLAMP>>;
def V_SAD_HI_U8 : VOP3Inst <"v_sad_hi_u8", VOP3_Profile<VOP_I32_I32_I32_I32, VOP3_CLAMP>>;
def V_SAD_U16 : VOP3Inst <"v_sad_u16", VOP3_Profile<VOP_I32_I32_I32_I32, VOP3_CLAMP>>;
def V_SAD_U32 : VOP3Inst <"v_sad_u32", VOP3_Profile<VOP_I32_I32_I32_I32, VOP3_CLAMP>>;
def V_CVT_PK_U8_F32 : VOP3Inst<"v_cvt_pk_u8_f32", VOP3_Profile<VOP_I32_F32_I32_I32>, int_amdgcn_cvt_pk_u8_f32>;
def V_DIV_FIXUP_F32 : VOP3Inst <"v_div_fixup_f32", VOP3_Profile<VOP_F32_F32_F32_F32>, AMDGPUdiv_fixup>;

let SchedRW = [WriteDoubleAdd], FPDPRounding = 1 in {
def V_DIV_FIXUP_F64 : VOP3Inst <"v_div_fixup_f64", VOP3_Profile<VOP_F64_F64_F64_F64>, AMDGPUdiv_fixup>;
def V_LDEXP_F64 : VOP3Inst <"v_ldexp_f64", VOP3_Profile<VOP_F64_F64_I32>, AMDGPUldexp, 1>;
} // End SchedRW = [WriteDoubleAdd], FPDPRounding = 1

def V_DIV_SCALE_F32 : VOP3_Pseudo <"v_div_scale_f32", VOP3b_F32_I1_F32_F32_F32, [], 1> {
  let SchedRW = [WriteFloatFMA, WriteSALU];
  let AsmMatchConverter = "";
}

// Double precision division pre-scale.
def V_DIV_SCALE_F64 : VOP3_Pseudo <"v_div_scale_f64", VOP3b_F64_I1_F64_F64_F64, [], 1> {
  let SchedRW = [WriteDouble, WriteSALU];
  let AsmMatchConverter = "";
  let FPDPRounding = 1;
}

def V_MSAD_U8 : VOP3Inst <"v_msad_u8", VOP3_Profile<VOP_I32_I32_I32_I32, VOP3_CLAMP>>;

let Constraints = "@earlyclobber $vdst" in {
def V_MQSAD_PK_U16_U8 : VOP3Inst <"v_mqsad_pk_u16_u8", VOP3_Profile<VOP_I64_I64_I32_I64, VOP3_CLAMP>>;
} // End Constraints = "@earlyclobber $vdst"

def V_TRIG_PREOP_F64 : VOP3Inst <"v_trig_preop_f64", VOP3_Profile<VOP_F64_F64_I32>, AMDGPUtrig_preop> {
  let SchedRW = [WriteDouble];
}

let SchedRW = [Write64Bit] in {
let SubtargetPredicate = isGFX6GFX7GFX10, Predicates = [isGFX6GFX7GFX10] in {
def V_LSHL_B64 : VOP3Inst <"v_lshl_b64", VOP3_Profile<VOP_PAT_GEN<VOP_I64_I64_I32>>, shl>;
def V_LSHR_B64 : VOP3Inst <"v_lshr_b64", VOP3_Profile<VOP_PAT_GEN<VOP_I64_I64_I32>>, srl>;
def V_ASHR_I64 : VOP3Inst <"v_ashr_i64", VOP3_Profile<VOP_PAT_GEN<VOP_I64_I64_I32>>, sra>;
def V_MULLIT_F32 : VOP3Inst <"v_mullit_f32", VOP3_Profile<VOP_F32_F32_F32_F32>>;
} // End SubtargetPredicate = isGFX6GFX7GFX10, Predicates = [isGFX6GFX7GFX10]

let SubtargetPredicate = isGFX8Plus in {
def V_LSHLREV_B64 : VOP3Inst <"v_lshlrev_b64", VOP3_Profile<VOP_I64_I32_I64>>;
def V_LSHRREV_B64 : VOP3Inst <"v_lshrrev_b64", VOP3_Profile<VOP_I64_I32_I64>>;
def V_ASHRREV_I64 : VOP3Inst <"v_ashrrev_i64", VOP3_Profile<VOP_I64_I32_I64>>;
} // End SubtargetPredicate = isGFX8Plus
} // End SchedRW = [Write64Bit]

let Predicates = [isGFX8Plus] in {
def : GCNPat <
 (getDivergentFrag<shl>.ret i64:$x, i32:$y),
 (V_LSHLREV_B64 $y, $x)
>;
def : AMDGPUPat <
 (getDivergentFrag<srl>.ret i64:$x, i32:$y),
 (V_LSHRREV_B64 $y, $x)
>;
def : AMDGPUPat <
 (getDivergentFrag<sra>.ret i64:$x, i32:$y),
 (V_ASHRREV_I64 $y, $x)
>;
}


let SchedRW = [Write32Bit] in {
let SubtargetPredicate = isGFX8Plus in {
def V_PERM_B32 : VOP3Inst <"v_perm_b32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUperm>;
} // End SubtargetPredicate = isGFX8Plus
} // End SchedRW = [Write32Bit]

let SubtargetPredicate = isGFX7Plus in {

let Constraints = "@earlyclobber $vdst", SchedRW = [WriteQuarterRate32] in {
def V_QSAD_PK_U16_U8 : VOP3Inst <"v_qsad_pk_u16_u8", VOP3_Profile<VOP_I64_I64_I32_I64, VOP3_CLAMP>>;
def V_MQSAD_U32_U8 : VOP3Inst <"v_mqsad_u32_u8", VOP3_Profile<VOP_V4I32_I64_I32_V4I32, VOP3_CLAMP>>;
} // End Constraints = "@earlyclobber $vdst", SchedRW = [WriteQuarterRate32]

let isCommutable = 1 in {
let SchedRW = [WriteQuarterRate32, WriteSALU] in {
def V_MAD_U64_U32 : VOP3Inst <"v_mad_u64_u32", VOP3b_I64_I1_I32_I32_I64>;
def V_MAD_I64_I32 : VOP3Inst <"v_mad_i64_i32", VOP3b_I64_I1_I32_I32_I64>;
} // End SchedRW = [WriteDouble, WriteSALU]
} // End isCommutable = 1

} // End SubtargetPredicate = isGFX7Plus


def V_DIV_FIXUP_F16 : VOP3Inst <"v_div_fixup_f16", VOP3_Profile<VOP_F16_F16_F16_F16>, AMDGPUdiv_fixup> {
  let Predicates = [Has16BitInsts, isGFX8Only];
  let FPDPRounding = 1;
}
def V_DIV_FIXUP_F16_gfx9 : VOP3Inst <"v_div_fixup_f16_gfx9",
                                      VOP3_Profile<VOP_F16_F16_F16_F16, VOP3_OPSEL>, AMDGPUdiv_fixup> {
  let renamedInGFX9 = 1;
  let Predicates = [Has16BitInsts, isGFX9Plus];
  let FPDPRounding = 1;
}

def V_FMA_F16 : VOP3Inst <"v_fma_f16", VOP3_Profile<VOP_F16_F16_F16_F16>, fma> {
  let Predicates = [Has16BitInsts, isGFX8Only];
  let FPDPRounding = 1;
}
def V_FMA_F16_gfx9 : VOP3Inst <"v_fma_f16_gfx9", VOP3_Profile<VOP_F16_F16_F16_F16, VOP3_OPSEL>, fma> {
  let renamedInGFX9 = 1;
  let Predicates = [Has16BitInsts, isGFX9Plus];
  let FPDPRounding = 1;
}

let SubtargetPredicate = Has16BitInsts, isCommutable = 1 in {

let renamedInGFX9 = 1 in {
def V_MAD_U16 : VOP3Inst <"v_mad_u16", VOP3_Profile<VOP_I16_I16_I16_I16, VOP3_CLAMP>>;
def V_MAD_I16 : VOP3Inst <"v_mad_i16", VOP3_Profile<VOP_I16_I16_I16_I16, VOP3_CLAMP>>;
let FPDPRounding = 1 in {
def V_MAD_F16 : VOP3Inst <"v_mad_f16", VOP3_Profile<VOP_F16_F16_F16_F16>, fmad>;
let Uses = [M0, EXEC] in {
def V_INTERP_P2_F16 : VOP3Interp <"v_interp_p2_f16", VOP3_INTERP16<[f16, f32, i32, f32]>,
       [(set f16:$vdst, (AMDGPUinterp_p2_f16 f32:$src0, (i32 imm:$attrchan),
                                                        (i32 imm:$attr),
                                                        (i32 imm:$src0_modifiers),
                                                        (f32 VRegSrc_32:$src2),
                                                        (i32 imm:$src2_modifiers),
                                                        (i1 imm:$high),
                                                        (i1 imm:$clamp)))]>;
} // End Uses = [M0, EXEC]
} // End FPDPRounding = 1
} // End renamedInGFX9 = 1

let SubtargetPredicate = isGFX9Only in {
def V_MAD_F16_gfx9   : VOP3Inst <"v_mad_f16_gfx9", VOP3_Profile<VOP_F16_F16_F16_F16, VOP3_OPSEL>> {
  let FPDPRounding = 1;
}
} // End SubtargetPredicate = isGFX9Only

let SubtargetPredicate = isGFX9Plus in {
def V_MAD_U16_gfx9   : VOP3Inst <"v_mad_u16_gfx9", VOP3_Profile<VOP_I16_I16_I16_I16, VOP3_OPSEL>>;
def V_MAD_I16_gfx9   : VOP3Inst <"v_mad_i16_gfx9", VOP3_Profile<VOP_I16_I16_I16_I16, VOP3_OPSEL>>;
def V_INTERP_P2_F16_gfx9 : VOP3Interp <"v_interp_p2_f16_gfx9", VOP3_INTERP16<[f16, f32, i32, f32]>>;
} // End SubtargetPredicate = isGFX9Plus

let Uses = [M0, EXEC], FPDPRounding = 1 in {
def V_INTERP_P1LL_F16 : VOP3Interp <"v_interp_p1ll_f16", VOP3_INTERP16<[f32, f32, i32, untyped]>,
       [(set f32:$vdst, (AMDGPUinterp_p1ll_f16 f32:$src0, (i32 imm:$attrchan),
                                                          (i32 imm:$attr),
                                                          (i32 imm:$src0_modifiers),
                                                          (i1 imm:$high),
                                                          (i1 imm:$clamp),
                                                          (i32 imm:$omod)))]>;
def V_INTERP_P1LV_F16 : VOP3Interp <"v_interp_p1lv_f16", VOP3_INTERP16<[f32, f32, i32, f16]>,
       [(set f32:$vdst, (AMDGPUinterp_p1lv_f16 f32:$src0, (i32 imm:$attrchan),
                                                          (i32 imm:$attr),
                                                          (i32 imm:$src0_modifiers),
                                                          (f32 VRegSrc_32:$src2),
                                                          (i32 imm:$src2_modifiers),
                                                          (i1 imm:$high),
                                                          (i1 imm:$clamp),
                                                          (i32 imm:$omod)))]>;
} // End Uses = [M0, EXEC], FPDPRounding = 1

} // End SubtargetPredicate = Has16BitInsts, isCommutable = 1

let SubtargetPredicate = isGFX8GFX9 in {
def V_INTERP_P1_F32_e64  : VOP3Interp <"v_interp_p1_f32", VOP3_INTERP>;
def V_INTERP_P2_F32_e64  : VOP3Interp <"v_interp_p2_f32", VOP3_INTERP>;
def V_INTERP_MOV_F32_e64 : VOP3Interp <"v_interp_mov_f32", VOP3_INTERP_MOV>;
} // End SubtargetPredicate = isGFX8GFX9

let Predicates = [Has16BitInsts, isGFX6GFX7GFX8GFX9] in {

multiclass Ternary_i16_Pats <SDPatternOperator op1, SDPatternOperator op2,
                             Instruction inst, SDPatternOperator op3> {
def : GCNPat <
  (op2 (op1 i16:$src0, i16:$src1), i16:$src2),
  (inst i16:$src0, i16:$src1, i16:$src2, (i1 0))
>;

}

defm: Ternary_i16_Pats<mul, add, V_MAD_U16, zext>;
defm: Ternary_i16_Pats<mul, add, V_MAD_I16, sext>;

} // End Predicates = [Has16BitInsts, isGFX6GFX7GFX8GFX9]

let Predicates = [Has16BitInsts, isGFX10Plus] in {

multiclass Ternary_i16_Pats_gfx9<SDPatternOperator op1, SDPatternOperator op2,
                                 Instruction inst, SDPatternOperator op3> {
def : GCNPat <
  (op2 (op1 i16:$src0, i16:$src1), i16:$src2),
  (inst SRCMODS.NONE, $src0, SRCMODS.NONE, $src1, SRCMODS.NONE, $src2, DSTCLAMP.NONE)
>;

}

defm: Ternary_i16_Pats_gfx9<mul, add, V_MAD_U16_gfx9, zext>;
defm: Ternary_i16_Pats_gfx9<mul, add, V_MAD_I16_gfx9, sext>;

} // End Predicates = [Has16BitInsts, isGFX10Plus]

class ThreeOpFrag<SDPatternOperator op1, SDPatternOperator op2> : PatFrag<
  (ops node:$x, node:$y, node:$z),
  // When the inner operation is used multiple times, selecting 3-op
  // instructions may still be beneficial -- if the other users can be
  // combined similarly. Let's be conservative for now.
  (op2 (HasOneUseBinOp<op1> node:$x, node:$y), node:$z),
  [{
    // Only use VALU ops when the result is divergent.
    if (!N->isDivergent())
      return false;

    // Check constant bus limitations.
    //
    // Note: Use !isDivergent as a conservative proxy for whether the value
    //       is in an SGPR (uniform values can end up in VGPRs as well).
    unsigned ConstantBusUses = 0;
    for (unsigned i = 0; i < 3; ++i) {
      if (!Operands[i]->isDivergent() &&
          !isInlineImmediate(Operands[i].getNode())) {
        ConstantBusUses++;
        // This uses AMDGPU::V_ADD3_U32, but all three operand instructions
        // have the same constant bus limit.
        if (ConstantBusUses > Subtarget->getConstantBusLimit(AMDGPU::V_ADD3_U32))
          return false;
      }
    }

    return true;
  }]
> {
  let PredicateCodeUsesOperands = 1;
}

let SubtargetPredicate = isGFX9Plus in {
def V_PACK_B32_F16 : VOP3Inst <"v_pack_b32_f16", VOP3_Profile<VOP_B32_F16_F16, VOP3_OPSEL>>;
def V_LSHL_ADD_U32 : VOP3Inst <"v_lshl_add_u32", VOP3_Profile<VOP_I32_I32_I32_I32>>;
def V_ADD_LSHL_U32 : VOP3Inst <"v_add_lshl_u32", VOP3_Profile<VOP_I32_I32_I32_I32>>;
def V_ADD3_U32 : VOP3Inst <"v_add3_u32", VOP3_Profile<VOP_I32_I32_I32_I32>>;
def V_LSHL_OR_B32 : VOP3Inst <"v_lshl_or_b32", VOP3_Profile<VOP_I32_I32_I32_I32>>;
def V_AND_OR_B32 : VOP3Inst <"v_and_or_b32", VOP3_Profile<VOP_I32_I32_I32_I32>>;
def V_OR3_B32 : VOP3Inst <"v_or3_b32", VOP3_Profile<VOP_I32_I32_I32_I32>>;

def V_XAD_U32 : VOP3Inst <"v_xad_u32", VOP3_Profile<VOP_I32_I32_I32_I32>>;

def V_MED3_F16 : VOP3Inst <"v_med3_f16", VOP3_Profile<VOP_F16_F16_F16_F16, VOP3_OPSEL>, AMDGPUfmed3>;
def V_MED3_I16 : VOP3Inst <"v_med3_i16", VOP3_Profile<VOP_I16_I16_I16_I16, VOP3_OPSEL>, AMDGPUsmed3>;
def V_MED3_U16 : VOP3Inst <"v_med3_u16", VOP3_Profile<VOP_I16_I16_I16_I16, VOP3_OPSEL>, AMDGPUumed3>;

def V_MIN3_F16 : VOP3Inst <"v_min3_f16", VOP3_Profile<VOP_F16_F16_F16_F16, VOP3_OPSEL>, AMDGPUfmin3>;
def V_MIN3_I16 : VOP3Inst <"v_min3_i16", VOP3_Profile<VOP_I16_I16_I16_I16, VOP3_OPSEL>, AMDGPUsmin3>;
def V_MIN3_U16 : VOP3Inst <"v_min3_u16", VOP3_Profile<VOP_I16_I16_I16_I16, VOP3_OPSEL>, AMDGPUumin3>;

def V_MAX3_F16 : VOP3Inst <"v_max3_f16", VOP3_Profile<VOP_F16_F16_F16_F16, VOP3_OPSEL>, AMDGPUfmax3>;
def V_MAX3_I16 : VOP3Inst <"v_max3_i16", VOP3_Profile<VOP_I16_I16_I16_I16, VOP3_OPSEL>, AMDGPUsmax3>;
def V_MAX3_U16 : VOP3Inst <"v_max3_u16", VOP3_Profile<VOP_I16_I16_I16_I16, VOP3_OPSEL>, AMDGPUumax3>;

def V_ADD_I16 : VOP3Inst <"v_add_i16", VOP3_Profile<VOP_I16_I16_I16, VOP3_OPSEL>>;
def V_SUB_I16 : VOP3Inst <"v_sub_i16", VOP3_Profile<VOP_I16_I16_I16, VOP3_OPSEL>>;

def V_MAD_U32_U16 : VOP3Inst <"v_mad_u32_u16", VOP3_Profile<VOP_I32_I16_I16_I32, VOP3_OPSEL>>;
def V_MAD_I32_I16 : VOP3Inst <"v_mad_i32_i16", VOP3_Profile<VOP_I32_I16_I16_I32, VOP3_OPSEL>>;

def V_CVT_PKNORM_I16_F16 : VOP3Inst <"v_cvt_pknorm_i16_f16", VOP3_Profile<VOP_B32_F16_F16, VOP3_OPSEL>>;
def V_CVT_PKNORM_U16_F16 : VOP3Inst <"v_cvt_pknorm_u16_f16", VOP3_Profile<VOP_B32_F16_F16, VOP3_OPSEL>>;

def V_ADD_I32_gfx9 : VOP3Inst <"v_add_i32_gfx9", VOP3_Profile<VOP_I32_I32_I32>>;
def V_SUB_I32_gfx9 : VOP3Inst <"v_sub_i32_gfx9", VOP3_Profile<VOP_I32_I32_I32>>;


class ThreeOp_i32_Pats <SDPatternOperator op1, SDPatternOperator op2, Instruction inst> : GCNPat <
  // This matches (op2 (op1 i32:$src0, i32:$src1), i32:$src2) with conditions.
  (ThreeOpFrag<op1, op2> i32:$src0, i32:$src1, i32:$src2),
  (inst i32:$src0, i32:$src1, i32:$src2)
>;

def : ThreeOp_i32_Pats<shl, add, V_LSHL_ADD_U32>;
def : ThreeOp_i32_Pats<add, shl, V_ADD_LSHL_U32>;
def : ThreeOp_i32_Pats<add, add, V_ADD3_U32>;
def : ThreeOp_i32_Pats<shl, or, V_LSHL_OR_B32>;
def : ThreeOp_i32_Pats<and, or, V_AND_OR_B32>;
def : ThreeOp_i32_Pats<or, or, V_OR3_B32>;
def : ThreeOp_i32_Pats<xor, add, V_XAD_U32>;

} // End SubtargetPredicate = isGFX9Plus

def VOP3_PERMLANE_Profile : VOP3_Profile<VOPProfile <[i32, i32, i32, i32]>, VOP3_OPSEL> {
  let Src0RC64 = VRegSrc_32;
  let Src1RC64 = SCSrc_b32;
  let Src2RC64 = SCSrc_b32;
  let InsVOP3OpSel = (ins IntOpSelMods:$src0_modifiers, VRegSrc_32:$src0,
                          IntOpSelMods:$src1_modifiers, SCSrc_b32:$src1,
                          IntOpSelMods:$src2_modifiers, SCSrc_b32:$src2,
                          VGPR_32:$vdst_in, op_sel:$op_sel);
  let HasClamp = 0;
  let HasOMod = 0;
}

let SubtargetPredicate = isGFX10Plus in {
  def V_XOR3_B32 : VOP3Inst <"v_xor3_b32", VOP3_Profile<VOP_I32_I32_I32_I32>>;
  def : ThreeOp_i32_Pats<xor, xor, V_XOR3_B32>;

  let Constraints = "$vdst = $vdst_in", DisableEncoding="$vdst_in" in {
    def V_PERMLANE16_B32 : VOP3Inst <"v_permlane16_b32", VOP3_PERMLANE_Profile>;
    def V_PERMLANEX16_B32 : VOP3Inst <"v_permlanex16_b32", VOP3_PERMLANE_Profile>;
  } // End $vdst = $vdst_in, DisableEncoding $vdst_in

  def : GCNPat<
    (int_amdgcn_permlane16 i32:$vdst_in, i32:$src0, i32:$src1, i32:$src2, imm:$fi, imm:$bc),
    (V_PERMLANE16_B32 (as_i1imm $fi), $src0, (as_i1imm $bc), $src1, 0, $src2, $vdst_in)
  >;
  def : GCNPat<
    (int_amdgcn_permlanex16 i32:$vdst_in, i32:$src0, i32:$src1, i32:$src2, imm:$fi, imm:$bc),
    (V_PERMLANEX16_B32 (as_i1imm $fi), $src0, (as_i1imm $bc), $src1, 0, $src2, $vdst_in)
  >;
} // End SubtargetPredicate = isGFX10Plus

//===----------------------------------------------------------------------===//
// Integer Clamp Patterns
//===----------------------------------------------------------------------===//

class getClampPat<VOPProfile P, SDPatternOperator node> {
  dag ret3 = (P.DstVT (node P.Src0VT:$src0, P.Src1VT:$src1, P.Src2VT:$src2));
  dag ret2 = (P.DstVT (node P.Src0VT:$src0, P.Src1VT:$src1));
  dag ret1 = (P.DstVT (node P.Src0VT:$src0));
  dag ret = !if(!eq(P.NumSrcArgs, 3), ret3,
            !if(!eq(P.NumSrcArgs, 2), ret2,
            ret1));
}

class getClampRes<VOPProfile P, Instruction inst> {
  dag ret3 = (inst P.Src0VT:$src0, P.Src1VT:$src1, P.Src2VT:$src2, (i1 0));
  dag ret2 = (inst P.Src0VT:$src0, P.Src1VT:$src1, (i1 0));
  dag ret1 = (inst P.Src0VT:$src0, (i1 0));
  dag ret = !if(!eq(P.NumSrcArgs, 3), ret3,
            !if(!eq(P.NumSrcArgs, 2), ret2,
            ret1));
}

class IntClampPat<VOP3Inst inst, SDPatternOperator node> : GCNPat<
  getClampPat<inst.Pfl, node>.ret,
  getClampRes<inst.Pfl, inst>.ret
>;

def : IntClampPat<V_MAD_I32_I24, AMDGPUmad_i24>;
def : IntClampPat<V_MAD_U32_U24, AMDGPUmad_u24>;

def : IntClampPat<V_SAD_U8, int_amdgcn_sad_u8>;
def : IntClampPat<V_SAD_HI_U8, int_amdgcn_sad_hi_u8>;
def : IntClampPat<V_SAD_U16, int_amdgcn_sad_u16>;

def : IntClampPat<V_MSAD_U8, int_amdgcn_msad_u8>;
def : IntClampPat<V_MQSAD_PK_U16_U8, int_amdgcn_mqsad_pk_u16_u8>;

def : IntClampPat<V_QSAD_PK_U16_U8, int_amdgcn_qsad_pk_u16_u8>;
def : IntClampPat<V_MQSAD_U32_U8, int_amdgcn_mqsad_u32_u8>;


//===----------------------------------------------------------------------===//
// Target-specific instruction encodings.
//===----------------------------------------------------------------------===//

//===----------------------------------------------------------------------===//
// GFX10.
//===----------------------------------------------------------------------===//

let AssemblerPredicate = isGFX10Plus, DecoderNamespace = "GFX10" in {
  multiclass VOP3_Real_gfx10<bits<10> op> {
    def _gfx10 :
      VOP3_Real<!cast<VOP_Pseudo>(NAME), SIEncodingFamily.GFX10>,
      VOP3e_gfx10<op, !cast<VOP_Pseudo>(NAME).Pfl>;
  }
  multiclass VOP3_Real_gfx10_with_name<bits<10> op, string opName,
                                       string asmName> {
    def _gfx10 :
      VOP3_Real<!cast<VOP3_Pseudo>(opName), SIEncodingFamily.GFX10>,
      VOP3e_gfx10<op, !cast<VOP3_Pseudo>(opName).Pfl> {
        VOP3_Pseudo ps = !cast<VOP3_Pseudo>(opName);
        let AsmString = asmName # ps.AsmOperands;
      }
  }
  multiclass VOP3be_Real_gfx10<bits<10> op> {
    def _gfx10 :
      VOP3_Real<!cast<VOP3_Pseudo>(NAME), SIEncodingFamily.GFX10>,
      VOP3be_gfx10<op, !cast<VOP3_Pseudo>(NAME).Pfl>;
  }
  multiclass VOP3Interp_Real_gfx10<bits<10> op> {
    def _gfx10 :
      VOP3_Real<!cast<VOP3_Pseudo>(NAME), SIEncodingFamily.GFX10>,
      VOP3Interp_gfx10<op, !cast<VOP3_Pseudo>(NAME).Pfl>;
  }
  multiclass VOP3OpSel_Real_gfx10<bits<10> op> {
    def _gfx10 :
      VOP3_Real<!cast<VOP3_Pseudo>(NAME), SIEncodingFamily.GFX10>,
      VOP3OpSel_gfx10<op, !cast<VOP3_Pseudo>(NAME).Pfl>;
  }
  multiclass VOP3OpSel_Real_gfx10_with_name<bits<10> op, string opName,
                                            string asmName> {
    def _gfx10 :
      VOP3_Real<!cast<VOP3_Pseudo>(opName), SIEncodingFamily.GFX10>,
      VOP3OpSel_gfx10<op, !cast<VOP3_Pseudo>(opName).Pfl> {
        VOP3_Pseudo ps = !cast<VOP3_Pseudo>(opName);
        let AsmString = asmName # ps.AsmOperands;
      }
  }
} // End AssemblerPredicate = isGFX10Plus, DecoderNamespace = "GFX10"

defm V_READLANE_B32  : VOP3_Real_gfx10<0x360>;

let InOperandList = (ins SSrcOrLds_b32:$src0, SCSrc_b32:$src1, VSrc_b32:$vdst_in) in {
  defm V_WRITELANE_B32 : VOP3_Real_gfx10<0x361>;
} // End InOperandList = (ins SSrcOrLds_b32:$src0, SCSrc_b32:$src1, VSrc_b32:$vdst_in)

defm V_XOR3_B32           : VOP3_Real_gfx10<0x178>;
defm V_LSHLREV_B64        : VOP3_Real_gfx10<0x2ff>;
defm V_LSHRREV_B64        : VOP3_Real_gfx10<0x300>;
defm V_ASHRREV_I64        : VOP3_Real_gfx10<0x301>;
defm V_PERM_B32           : VOP3_Real_gfx10<0x344>;
defm V_XAD_U32            : VOP3_Real_gfx10<0x345>;
defm V_LSHL_ADD_U32       : VOP3_Real_gfx10<0x346>;
defm V_ADD_LSHL_U32       : VOP3_Real_gfx10<0x347>;
defm V_ADD3_U32           : VOP3_Real_gfx10<0x36d>;
defm V_LSHL_OR_B32        : VOP3_Real_gfx10<0x36f>;
defm V_AND_OR_B32         : VOP3_Real_gfx10<0x371>;
defm V_OR3_B32            : VOP3_Real_gfx10<0x372>;

// TODO-GFX10: add MC tests for v_add/sub_nc_i16
defm V_ADD_NC_I16 :
  VOP3OpSel_Real_gfx10_with_name<0x30d, "V_ADD_I16", "v_add_nc_i16">;
defm V_SUB_NC_I16 :
  VOP3OpSel_Real_gfx10_with_name<0x30e, "V_SUB_I16", "v_sub_nc_i16">;
defm V_SUB_NC_I32 :
  VOP3_Real_gfx10_with_name<0x376, "V_SUB_I32_gfx9", "v_sub_nc_i32">;
defm V_ADD_NC_I32 :
  VOP3_Real_gfx10_with_name<0x37f, "V_ADD_I32_gfx9", "v_add_nc_i32">;

defm V_INTERP_P1LL_F16    : VOP3Interp_Real_gfx10<0x342>;
defm V_INTERP_P1LV_F16    : VOP3Interp_Real_gfx10<0x343>;
defm V_INTERP_P2_F16      : VOP3Interp_Real_gfx10<0x35a>;

defm V_PACK_B32_F16       : VOP3OpSel_Real_gfx10<0x311>;
defm V_CVT_PKNORM_I16_F16 : VOP3OpSel_Real_gfx10<0x312>;
defm V_CVT_PKNORM_U16_F16 : VOP3OpSel_Real_gfx10<0x313>;

defm V_MIN3_F16           : VOP3OpSel_Real_gfx10<0x351>;
defm V_MIN3_I16           : VOP3OpSel_Real_gfx10<0x352>;
defm V_MIN3_U16           : VOP3OpSel_Real_gfx10<0x353>;
defm V_MAX3_F16           : VOP3OpSel_Real_gfx10<0x354>;
defm V_MAX3_I16           : VOP3OpSel_Real_gfx10<0x355>;
defm V_MAX3_U16           : VOP3OpSel_Real_gfx10<0x356>;
defm V_MED3_F16           : VOP3OpSel_Real_gfx10<0x357>;
defm V_MED3_I16           : VOP3OpSel_Real_gfx10<0x358>;
defm V_MED3_U16           : VOP3OpSel_Real_gfx10<0x359>;
defm V_MAD_U32_U16        : VOP3OpSel_Real_gfx10<0x373>;
defm V_MAD_I32_I16        : VOP3OpSel_Real_gfx10<0x375>;

defm V_MAD_U16 :
  VOP3OpSel_Real_gfx10_with_name<0x340, "V_MAD_U16_gfx9", "v_mad_u16">;
defm V_FMA_F16 :
  VOP3OpSel_Real_gfx10_with_name<0x34b, "V_FMA_F16_gfx9", "v_fma_f16">;
defm V_MAD_I16 :
  VOP3OpSel_Real_gfx10_with_name<0x35e, "V_MAD_I16_gfx9", "v_mad_i16">;
defm V_DIV_FIXUP_F16 :
  VOP3OpSel_Real_gfx10_with_name<0x35f, "V_DIV_FIXUP_F16_gfx9", "v_div_fixup_f16">;

// FIXME-GFX10-OPSEL: Need to add "selective" opsel support to some of these
// (they do not support SDWA or DPP).
defm V_ADD_NC_U16      : VOP3_Real_gfx10_with_name<0x303, "V_ADD_U16_e64", "v_add_nc_u16">;
defm V_SUB_NC_U16      : VOP3_Real_gfx10_with_name<0x304, "V_SUB_U16_e64", "v_sub_nc_u16">;
defm V_MUL_LO_U16      : VOP3_Real_gfx10_with_name<0x305, "V_MUL_LO_U16_e64", "v_mul_lo_u16">;
defm V_LSHRREV_B16     : VOP3_Real_gfx10_with_name<0x307, "V_LSHRREV_B16_e64", "v_lshrrev_b16">;
defm V_ASHRREV_I16     : VOP3_Real_gfx10_with_name<0x308, "V_ASHRREV_I16_e64", "v_ashrrev_i16">;
defm V_MAX_U16         : VOP3_Real_gfx10_with_name<0x309, "V_MAX_U16_e64", "v_max_u16">;
defm V_MAX_I16         : VOP3_Real_gfx10_with_name<0x30a, "V_MAX_I16_e64", "v_max_i16">;
defm V_MIN_U16         : VOP3_Real_gfx10_with_name<0x30b, "V_MIN_U16_e64", "v_min_u16">;
defm V_MIN_I16         : VOP3_Real_gfx10_with_name<0x30c, "V_MIN_I16_e64", "v_min_i16">;
defm V_LSHLREV_B16     : VOP3_Real_gfx10_with_name<0x314, "V_LSHLREV_B16_e64", "v_lshlrev_b16">;
defm V_PERMLANE16_B32  : VOP3OpSel_Real_gfx10<0x377>;
defm V_PERMLANEX16_B32 : VOP3OpSel_Real_gfx10<0x378>;

//===----------------------------------------------------------------------===//
// GFX7, GFX10.
//===----------------------------------------------------------------------===//

let AssemblerPredicate = isGFX7Only, DecoderNamespace = "GFX7" in {
  multiclass VOP3_Real_gfx7<bits<10> op> {
    def _gfx7 :
      VOP3_Real<!cast<VOP3_Pseudo>(NAME), SIEncodingFamily.SI>,
      VOP3e_gfx6_gfx7<op{8-0}, !cast<VOP3_Pseudo>(NAME).Pfl>;
  }
  multiclass VOP3be_Real_gfx7<bits<10> op> {
    def _gfx7 :
      VOP3_Real<!cast<VOP3_Pseudo>(NAME), SIEncodingFamily.SI>,
      VOP3be_gfx6_gfx7<op{8-0}, !cast<VOP3_Pseudo>(NAME).Pfl>;
  }
} // End AssemblerPredicate = isGFX7Only, DecoderNamespace = "GFX7"

multiclass VOP3_Real_gfx7_gfx10<bits<10> op> :
  VOP3_Real_gfx7<op>, VOP3_Real_gfx10<op>;

multiclass VOP3be_Real_gfx7_gfx10<bits<10> op> :
  VOP3be_Real_gfx7<op>, VOP3be_Real_gfx10<op>;

defm V_QSAD_PK_U16_U8   : VOP3_Real_gfx7_gfx10<0x172>;
defm V_MQSAD_U32_U8     : VOP3_Real_gfx7_gfx10<0x175>;
defm V_MAD_U64_U32      : VOP3be_Real_gfx7_gfx10<0x176>;
defm V_MAD_I64_I32      : VOP3be_Real_gfx7_gfx10<0x177>;

//===----------------------------------------------------------------------===//
// GFX6, GFX7, GFX10.
//===----------------------------------------------------------------------===//

let AssemblerPredicate = isGFX6GFX7, DecoderNamespace = "GFX6GFX7" in {
  multiclass VOP3_Real_gfx6_gfx7<bits<10> op> {
    def _gfx6_gfx7 :
      VOP3_Real<!cast<VOP3_Pseudo>(NAME), SIEncodingFamily.SI>,
      VOP3e_gfx6_gfx7<op{8-0}, !cast<VOP3_Pseudo>(NAME).Pfl>;
  }
  multiclass VOP3be_Real_gfx6_gfx7<bits<10> op> {
    def _gfx6_gfx7 :
      VOP3_Real<!cast<VOP3_Pseudo>(NAME), SIEncodingFamily.SI>,
      VOP3be_gfx6_gfx7<op{8-0}, !cast<VOP3_Pseudo>(NAME).Pfl>;
  }
} // End AssemblerPredicate = isGFX6GFX7, DecoderNamespace = "GFX6GFX7"

multiclass VOP3_Real_gfx6_gfx7_gfx10<bits<10> op> :
  VOP3_Real_gfx6_gfx7<op>, VOP3_Real_gfx10<op>;

multiclass VOP3be_Real_gfx6_gfx7_gfx10<bits<10> op> :
  VOP3be_Real_gfx6_gfx7<op>, VOP3be_Real_gfx10<op>;

defm V_LSHL_B64        : VOP3_Real_gfx6_gfx7<0x161>;
defm V_LSHR_B64        : VOP3_Real_gfx6_gfx7<0x162>;
defm V_ASHR_I64        : VOP3_Real_gfx6_gfx7<0x163>;

defm V_MAD_LEGACY_F32  : VOP3_Real_gfx6_gfx7_gfx10<0x140>;
defm V_MAD_F32         : VOP3_Real_gfx6_gfx7_gfx10<0x141>;
defm V_MAD_I32_I24     : VOP3_Real_gfx6_gfx7_gfx10<0x142>;
defm V_MAD_U32_U24     : VOP3_Real_gfx6_gfx7_gfx10<0x143>;
defm V_CUBEID_F32      : VOP3_Real_gfx6_gfx7_gfx10<0x144>;
defm V_CUBESC_F32      : VOP3_Real_gfx6_gfx7_gfx10<0x145>;
defm V_CUBETC_F32      : VOP3_Real_gfx6_gfx7_gfx10<0x146>;
defm V_CUBEMA_F32      : VOP3_Real_gfx6_gfx7_gfx10<0x147>;
defm V_BFE_U32         : VOP3_Real_gfx6_gfx7_gfx10<0x148>;
defm V_BFE_I32         : VOP3_Real_gfx6_gfx7_gfx10<0x149>;
defm V_BFI_B32         : VOP3_Real_gfx6_gfx7_gfx10<0x14a>;
defm V_FMA_F32         : VOP3_Real_gfx6_gfx7_gfx10<0x14b>;
defm V_FMA_F64         : VOP3_Real_gfx6_gfx7_gfx10<0x14c>;
defm V_LERP_U8         : VOP3_Real_gfx6_gfx7_gfx10<0x14d>;
defm V_ALIGNBIT_B32    : VOP3_Real_gfx6_gfx7_gfx10<0x14e>;
defm V_ALIGNBYTE_B32   : VOP3_Real_gfx6_gfx7_gfx10<0x14f>;
defm V_MULLIT_F32      : VOP3_Real_gfx6_gfx7_gfx10<0x150>;
defm V_MIN3_F32        : VOP3_Real_gfx6_gfx7_gfx10<0x151>;
defm V_MIN3_I32        : VOP3_Real_gfx6_gfx7_gfx10<0x152>;
defm V_MIN3_U32        : VOP3_Real_gfx6_gfx7_gfx10<0x153>;
defm V_MAX3_F32        : VOP3_Real_gfx6_gfx7_gfx10<0x154>;
defm V_MAX3_I32        : VOP3_Real_gfx6_gfx7_gfx10<0x155>;
defm V_MAX3_U32        : VOP3_Real_gfx6_gfx7_gfx10<0x156>;
defm V_MED3_F32        : VOP3_Real_gfx6_gfx7_gfx10<0x157>;
defm V_MED3_I32        : VOP3_Real_gfx6_gfx7_gfx10<0x158>;
defm V_MED3_U32        : VOP3_Real_gfx6_gfx7_gfx10<0x159>;
defm V_SAD_U8          : VOP3_Real_gfx6_gfx7_gfx10<0x15a>;
defm V_SAD_HI_U8       : VOP3_Real_gfx6_gfx7_gfx10<0x15b>;
defm V_SAD_U16         : VOP3_Real_gfx6_gfx7_gfx10<0x15c>;
defm V_SAD_U32         : VOP3_Real_gfx6_gfx7_gfx10<0x15d>;
defm V_CVT_PK_U8_F32   : VOP3_Real_gfx6_gfx7_gfx10<0x15e>;
defm V_DIV_FIXUP_F32   : VOP3_Real_gfx6_gfx7_gfx10<0x15f>;
defm V_DIV_FIXUP_F64   : VOP3_Real_gfx6_gfx7_gfx10<0x160>;
defm V_ADD_F64         : VOP3_Real_gfx6_gfx7_gfx10<0x164>;
defm V_MUL_F64         : VOP3_Real_gfx6_gfx7_gfx10<0x165>;
defm V_MIN_F64         : VOP3_Real_gfx6_gfx7_gfx10<0x166>;
defm V_MAX_F64         : VOP3_Real_gfx6_gfx7_gfx10<0x167>;
defm V_LDEXP_F64       : VOP3_Real_gfx6_gfx7_gfx10<0x168>;
defm V_MUL_LO_U32      : VOP3_Real_gfx6_gfx7_gfx10<0x169>;
defm V_MUL_HI_U32      : VOP3_Real_gfx6_gfx7_gfx10<0x16a>;
defm V_MUL_LO_I32      : VOP3_Real_gfx6_gfx7_gfx10<0x16b>;
defm V_MUL_HI_I32      : VOP3_Real_gfx6_gfx7_gfx10<0x16c>;
defm V_DIV_FMAS_F32    : VOP3_Real_gfx6_gfx7_gfx10<0x16f>;
defm V_DIV_FMAS_F64    : VOP3_Real_gfx6_gfx7_gfx10<0x170>;
defm V_MSAD_U8         : VOP3_Real_gfx6_gfx7_gfx10<0x171>;
defm V_MQSAD_PK_U16_U8 : VOP3_Real_gfx6_gfx7_gfx10<0x173>;
defm V_TRIG_PREOP_F64  : VOP3_Real_gfx6_gfx7_gfx10<0x174>;
defm V_DIV_SCALE_F32   : VOP3be_Real_gfx6_gfx7_gfx10<0x16d>;
defm V_DIV_SCALE_F64   : VOP3be_Real_gfx6_gfx7_gfx10<0x16e>;

//===----------------------------------------------------------------------===//
// GFX8, GFX9 (VI).
//===----------------------------------------------------------------------===//

let AssemblerPredicates = [isGFX8GFX9], DecoderNamespace = "GFX8" in {

multiclass VOP3_Real_vi<bits<10> op> {
  def _vi : VOP3_Real<!cast<VOP_Pseudo>(NAME), SIEncodingFamily.VI>,
            VOP3e_vi <op, !cast<VOP_Pseudo>(NAME).Pfl>;
}

multiclass VOP3be_Real_vi<bits<10> op> {
  def _vi : VOP3_Real<!cast<VOP_Pseudo>(NAME), SIEncodingFamily.VI>,
            VOP3be_vi <op, !cast<VOP_Pseudo>(NAME).Pfl>;
}

multiclass VOP3OpSel_Real_gfx9<bits<10> op> {
  def _vi : VOP3_Real<!cast<VOP_Pseudo>(NAME), SIEncodingFamily.VI>,
            VOP3OpSel_gfx9 <op, !cast<VOP_Pseudo>(NAME).Pfl>;
}

multiclass VOP3Interp_Real_vi<bits<10> op> {
  def _vi : VOP3_Real<!cast<VOP_Pseudo>(NAME), SIEncodingFamily.VI>,
            VOP3Interp_vi <op, !cast<VOP_Pseudo>(NAME).Pfl>;
}

} // End AssemblerPredicates = [isGFX8GFX9], DecoderNamespace = "GFX8"

let AssemblerPredicates = [isGFX8Only], DecoderNamespace = "GFX8" in {

multiclass VOP3_F16_Real_vi<bits<10> op> {
  def _vi : VOP3_Real<!cast<VOP3_Pseudo>(NAME), SIEncodingFamily.VI>,
            VOP3e_vi <op, !cast<VOP3_Pseudo>(NAME).Pfl>;
}

multiclass VOP3Interp_F16_Real_vi<bits<10> op> {
  def _vi : VOP3_Real<!cast<VOP3_Pseudo>(NAME), SIEncodingFamily.VI>,
            VOP3Interp_vi <op, !cast<VOP3_Pseudo>(NAME).Pfl>;
}

} // End AssemblerPredicates = [isGFX8Only], DecoderNamespace = "GFX8"

let AssemblerPredicates = [isGFX9Only], DecoderNamespace = "GFX9" in {

multiclass VOP3_F16_Real_gfx9<bits<10> op, string OpName, string AsmName> {
  def _gfx9 : VOP3_Real<!cast<VOP3_Pseudo>(OpName), SIEncodingFamily.GFX9>,
            VOP3e_vi <op, !cast<VOP3_Pseudo>(OpName).Pfl> {
              VOP3_Pseudo ps = !cast<VOP3_Pseudo>(OpName);
              let AsmString = AsmName # ps.AsmOperands;
            }
}

multiclass VOP3OpSel_F16_Real_gfx9<bits<10> op, string AsmName> {
  def _gfx9 : VOP3_Real<!cast<VOP3_Pseudo>(NAME), SIEncodingFamily.GFX9>,
            VOP3OpSel_gfx9 <op, !cast<VOP3_Pseudo>(NAME).Pfl> {
              VOP3_Pseudo ps = !cast<VOP3_Pseudo>(NAME);
              let AsmString = AsmName # ps.AsmOperands;
            }
}

multiclass VOP3Interp_F16_Real_gfx9<bits<10> op, string OpName, string AsmName> {
  def _gfx9 : VOP3_Real<!cast<VOP3_Pseudo>(OpName), SIEncodingFamily.GFX9>,
            VOP3Interp_vi <op, !cast<VOP3_Pseudo>(OpName).Pfl> {
              VOP3_Pseudo ps = !cast<VOP3_Pseudo>(OpName);
              let AsmString = AsmName # ps.AsmOperands;
            }
}

multiclass VOP3_Real_gfx9<bits<10> op, string AsmName> {
  def _gfx9 : VOP3_Real<!cast<VOP_Pseudo>(NAME), SIEncodingFamily.GFX9>,
              VOP3e_vi <op, !cast<VOP_Pseudo>(NAME).Pfl> {
              VOP_Pseudo ps = !cast<VOP_Pseudo>(NAME);
              let AsmString = AsmName # ps.AsmOperands;
            }
}

} // End AssemblerPredicates = [isGFX9Only], DecoderNamespace = "GFX9"

defm V_MAD_U64_U32      : VOP3be_Real_vi <0x1E8>;
defm V_MAD_I64_I32      : VOP3be_Real_vi <0x1E9>;

defm V_MAD_LEGACY_F32   : VOP3_Real_vi <0x1c0>;
defm V_MAD_F32          : VOP3_Real_vi <0x1c1>;
defm V_MAD_I32_I24      : VOP3_Real_vi <0x1c2>;
defm V_MAD_U32_U24      : VOP3_Real_vi <0x1c3>;
defm V_CUBEID_F32       : VOP3_Real_vi <0x1c4>;
defm V_CUBESC_F32       : VOP3_Real_vi <0x1c5>;
defm V_CUBETC_F32       : VOP3_Real_vi <0x1c6>;
defm V_CUBEMA_F32       : VOP3_Real_vi <0x1c7>;
defm V_BFE_U32          : VOP3_Real_vi <0x1c8>;
defm V_BFE_I32          : VOP3_Real_vi <0x1c9>;
defm V_BFI_B32          : VOP3_Real_vi <0x1ca>;
defm V_FMA_F32          : VOP3_Real_vi <0x1cb>;
defm V_FMA_F64          : VOP3_Real_vi <0x1cc>;
defm V_LERP_U8          : VOP3_Real_vi <0x1cd>;
defm V_ALIGNBIT_B32     : VOP3_Real_vi <0x1ce>;
defm V_ALIGNBYTE_B32    : VOP3_Real_vi <0x1cf>;
defm V_MIN3_F32         : VOP3_Real_vi <0x1d0>;
defm V_MIN3_I32         : VOP3_Real_vi <0x1d1>;
defm V_MIN3_U32         : VOP3_Real_vi <0x1d2>;
defm V_MAX3_F32         : VOP3_Real_vi <0x1d3>;
defm V_MAX3_I32         : VOP3_Real_vi <0x1d4>;
defm V_MAX3_U32         : VOP3_Real_vi <0x1d5>;
defm V_MED3_F32         : VOP3_Real_vi <0x1d6>;
defm V_MED3_I32         : VOP3_Real_vi <0x1d7>;
defm V_MED3_U32         : VOP3_Real_vi <0x1d8>;
defm V_SAD_U8           : VOP3_Real_vi <0x1d9>;
defm V_SAD_HI_U8        : VOP3_Real_vi <0x1da>;
defm V_SAD_U16          : VOP3_Real_vi <0x1db>;
defm V_SAD_U32          : VOP3_Real_vi <0x1dc>;
defm V_CVT_PK_U8_F32    : VOP3_Real_vi <0x1dd>;
defm V_DIV_FIXUP_F32    : VOP3_Real_vi <0x1de>;
defm V_DIV_FIXUP_F64    : VOP3_Real_vi <0x1df>;
defm V_DIV_SCALE_F32    : VOP3be_Real_vi <0x1e0>;
defm V_DIV_SCALE_F64    : VOP3be_Real_vi <0x1e1>;
defm V_DIV_FMAS_F32     : VOP3_Real_vi <0x1e2>;
defm V_DIV_FMAS_F64     : VOP3_Real_vi <0x1e3>;
defm V_MSAD_U8          : VOP3_Real_vi <0x1e4>;
defm V_QSAD_PK_U16_U8   : VOP3_Real_vi <0x1e5>;
defm V_MQSAD_PK_U16_U8  : VOP3_Real_vi <0x1e6>;
defm V_MQSAD_U32_U8     : VOP3_Real_vi <0x1e7>;

defm V_PERM_B32         : VOP3_Real_vi <0x1ed>;

defm V_MAD_F16          : VOP3_F16_Real_vi <0x1ea>;
defm V_MAD_U16          : VOP3_F16_Real_vi <0x1eb>;
defm V_MAD_I16          : VOP3_F16_Real_vi <0x1ec>;
defm V_FMA_F16          : VOP3_F16_Real_vi <0x1ee>;
defm V_DIV_FIXUP_F16    : VOP3_F16_Real_vi <0x1ef>;
defm V_INTERP_P2_F16    : VOP3Interp_F16_Real_vi <0x276>;

let FPDPRounding = 1 in {
defm V_MAD_LEGACY_F16       : VOP3_F16_Real_gfx9 <0x1ea, "V_MAD_F16",       "v_mad_legacy_f16">;
defm V_FMA_LEGACY_F16       : VOP3_F16_Real_gfx9 <0x1ee, "V_FMA_F16",       "v_fma_legacy_f16">;
defm V_DIV_FIXUP_LEGACY_F16 : VOP3_F16_Real_gfx9 <0x1ef, "V_DIV_FIXUP_F16", "v_div_fixup_legacy_f16">;
defm V_INTERP_P2_LEGACY_F16 : VOP3Interp_F16_Real_gfx9 <0x276, "V_INTERP_P2_F16", "v_interp_p2_legacy_f16">;
} // End FPDPRounding = 1

defm V_MAD_LEGACY_U16       : VOP3_F16_Real_gfx9 <0x1eb, "V_MAD_U16",       "v_mad_legacy_u16">;
defm V_MAD_LEGACY_I16       : VOP3_F16_Real_gfx9 <0x1ec, "V_MAD_I16",       "v_mad_legacy_i16">;

defm V_MAD_F16_gfx9         : VOP3OpSel_F16_Real_gfx9 <0x203, "v_mad_f16">;
defm V_MAD_U16_gfx9         : VOP3OpSel_F16_Real_gfx9 <0x204, "v_mad_u16">;
defm V_MAD_I16_gfx9         : VOP3OpSel_F16_Real_gfx9 <0x205, "v_mad_i16">;
defm V_FMA_F16_gfx9         : VOP3OpSel_F16_Real_gfx9 <0x206, "v_fma_f16">;
defm V_DIV_FIXUP_F16_gfx9   : VOP3OpSel_F16_Real_gfx9 <0x207, "v_div_fixup_f16">;
defm V_INTERP_P2_F16_gfx9   : VOP3Interp_F16_Real_gfx9 <0x277, "V_INTERP_P2_F16_gfx9", "v_interp_p2_f16">;

defm V_ADD_I32_gfx9         : VOP3_Real_gfx9 <0x29c, "v_add_i32">;
defm V_SUB_I32_gfx9         : VOP3_Real_gfx9 <0x29d, "v_sub_i32">;

defm V_INTERP_P1_F32_e64  : VOP3Interp_Real_vi <0x270>;
defm V_INTERP_P2_F32_e64  : VOP3Interp_Real_vi <0x271>;
defm V_INTERP_MOV_F32_e64 : VOP3Interp_Real_vi <0x272>;

defm V_INTERP_P1LL_F16  : VOP3Interp_Real_vi <0x274>;
defm V_INTERP_P1LV_F16  : VOP3Interp_Real_vi <0x275>;
defm V_ADD_F64          : VOP3_Real_vi <0x280>;
defm V_MUL_F64          : VOP3_Real_vi <0x281>;
defm V_MIN_F64          : VOP3_Real_vi <0x282>;
defm V_MAX_F64          : VOP3_Real_vi <0x283>;
defm V_LDEXP_F64        : VOP3_Real_vi <0x284>;
defm V_MUL_LO_U32       : VOP3_Real_vi <0x285>;

// removed from VI as identical to V_MUL_LO_U32
let isAsmParserOnly = 1 in {
defm V_MUL_LO_I32       : VOP3_Real_vi <0x285>;
}

defm V_MUL_HI_U32       : VOP3_Real_vi <0x286>;
defm V_MUL_HI_I32       : VOP3_Real_vi <0x287>;

defm V_READLANE_B32     : VOP3_Real_vi <0x289>;
defm V_WRITELANE_B32    : VOP3_Real_vi <0x28a>;

defm V_LSHLREV_B64      : VOP3_Real_vi <0x28f>;
defm V_LSHRREV_B64      : VOP3_Real_vi <0x290>;
defm V_ASHRREV_I64      : VOP3_Real_vi <0x291>;
defm V_TRIG_PREOP_F64   : VOP3_Real_vi <0x292>;

defm V_LSHL_ADD_U32 : VOP3_Real_vi <0x1fd>;
defm V_ADD_LSHL_U32 : VOP3_Real_vi <0x1fe>;
defm V_ADD3_U32 : VOP3_Real_vi <0x1ff>;
defm V_LSHL_OR_B32 : VOP3_Real_vi <0x200>;
defm V_AND_OR_B32 : VOP3_Real_vi <0x201>;
defm V_OR3_B32 : VOP3_Real_vi <0x202>;
defm V_PACK_B32_F16 : VOP3OpSel_Real_gfx9 <0x2a0>;

defm V_XAD_U32 : VOP3_Real_vi <0x1f3>;

defm V_MIN3_F16 : VOP3OpSel_Real_gfx9 <0x1f4>;
defm V_MIN3_I16 : VOP3OpSel_Real_gfx9 <0x1f5>;
defm V_MIN3_U16 : VOP3OpSel_Real_gfx9 <0x1f6>;

defm V_MAX3_F16 : VOP3OpSel_Real_gfx9 <0x1f7>;
defm V_MAX3_I16 : VOP3OpSel_Real_gfx9 <0x1f8>;
defm V_MAX3_U16 : VOP3OpSel_Real_gfx9 <0x1f9>;

defm V_MED3_F16 : VOP3OpSel_Real_gfx9 <0x1fa>;
defm V_MED3_I16 : VOP3OpSel_Real_gfx9 <0x1fb>;
defm V_MED3_U16 : VOP3OpSel_Real_gfx9 <0x1fc>;

defm V_ADD_I16  : VOP3OpSel_Real_gfx9 <0x29e>;
defm V_SUB_I16  : VOP3OpSel_Real_gfx9 <0x29f>;

defm V_MAD_U32_U16 : VOP3OpSel_Real_gfx9 <0x1f1>;
defm V_MAD_I32_I16 : VOP3OpSel_Real_gfx9 <0x1f2>;

defm V_CVT_PKNORM_I16_F16 : VOP3OpSel_Real_gfx9 <0x299>;
defm V_CVT_PKNORM_U16_F16 : VOP3OpSel_Real_gfx9 <0x29a>;