aboutsummaryrefslogtreecommitdiffstats
path: root/contrib/libs/llvm12/lib/Target/X86/X86ScheduleZnver2.td
blob: 48da0d6329b19906f514b67779c87d6c9f4b2ff2 (plain) (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
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
//=- X86ScheduleZnver2.td - X86 Znver2 Scheduling -------------*- tablegen -*-=//
//
// 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
//
//===----------------------------------------------------------------------===//
//
// This file defines the machine model for Znver2 to support instruction
// scheduling and other instruction cost heuristics.
//
//===----------------------------------------------------------------------===//

def Znver2Model : SchedMachineModel {
  // Zen can decode 4 instructions per cycle.
  let IssueWidth = 4;
  // Based on the reorder buffer we define MicroOpBufferSize
  let MicroOpBufferSize = 224;
  let LoadLatency = 4;
  let MispredictPenalty = 17;
  let HighLatency = 25;
  let PostRAScheduler = 1;

  // FIXME: This variable is required for incomplete model.
  // We haven't catered all instructions.
  // So, we reset the value of this variable so as to
  // say that the model is incomplete.
  let CompleteModel = 0;
}

let SchedModel = Znver2Model in {

// Zen can issue micro-ops to 10 different units in one cycle.
// These are
//  * Four integer ALU units (ZALU0, ZALU1, ZALU2, ZALU3)
//  * Three AGU units (ZAGU0, ZAGU1, ZAGU2)
//  * Four FPU units (ZFPU0, ZFPU1, ZFPU2, ZFPU3)
// AGUs feed load store queues @two loads and 1 store per cycle.

// Four ALU units are defined below
def Zn2ALU0 : ProcResource<1>;
def Zn2ALU1 : ProcResource<1>;
def Zn2ALU2 : ProcResource<1>;
def Zn2ALU3 : ProcResource<1>;

// Three AGU units are defined below
def Zn2AGU0 : ProcResource<1>;
def Zn2AGU1 : ProcResource<1>;
def Zn2AGU2 : ProcResource<1>;

// Four FPU units are defined below
def Zn2FPU0 : ProcResource<1>;
def Zn2FPU1 : ProcResource<1>;
def Zn2FPU2 : ProcResource<1>;
def Zn2FPU3 : ProcResource<1>;

// FPU grouping
def Zn2FPU013  : ProcResGroup<[Zn2FPU0, Zn2FPU1, Zn2FPU3]>;
def Zn2FPU01   : ProcResGroup<[Zn2FPU0, Zn2FPU1]>;
def Zn2FPU12   : ProcResGroup<[Zn2FPU1, Zn2FPU2]>;
def Zn2FPU13   : ProcResGroup<[Zn2FPU1, Zn2FPU3]>;
def Zn2FPU23   : ProcResGroup<[Zn2FPU2, Zn2FPU3]>;
def Zn2FPU02   : ProcResGroup<[Zn2FPU0, Zn2FPU2]>;
def Zn2FPU03   : ProcResGroup<[Zn2FPU0, Zn2FPU3]>;

// Below are the grouping of the units.
// Micro-ops to be issued to multiple units are tackled this way.

// ALU grouping
// Zn2ALU03 - 0,3 grouping
def Zn2ALU03: ProcResGroup<[Zn2ALU0, Zn2ALU3]>;

// 64 Entry (16x4 entries) Int Scheduler
def Zn2ALU : ProcResGroup<[Zn2ALU0, Zn2ALU1, Zn2ALU2, Zn2ALU3]> {
  let BufferSize=64;
}

// 28 Entry (14x2) AGU group. AGUs can't be used for all ALU operations
// but are relevant for some instructions
def Zn2AGU : ProcResGroup<[Zn2AGU0, Zn2AGU1, Zn2AGU2]> {
  let BufferSize=28;
}

// Integer Multiplication issued on ALU1.
def Zn2Multiplier : ProcResource<1>;

// Integer division issued on ALU2.
def Zn2Divider : ProcResource<1>;

// 4 Cycles load-to use Latency is captured
def : ReadAdvance<ReadAfterLd, 4>;

// 7 Cycles vector load-to use Latency is captured
def : ReadAdvance<ReadAfterVecLd, 7>;
def : ReadAdvance<ReadAfterVecXLd, 7>;
def : ReadAdvance<ReadAfterVecYLd, 7>;

def : ReadAdvance<ReadInt2Fpu, 0>;

// The Integer PRF for Zen is 168 entries, and it holds the architectural and
// speculative version of the 64-bit integer registers.
// Reference: "Software Optimization Guide for AMD Family 17h Processors"
def Zn2IntegerPRF : RegisterFile<168, [GR64, CCR]>;

// 36 Entry (9x4 entries) floating-point Scheduler
def Zn2FPU     : ProcResGroup<[Zn2FPU0, Zn2FPU1, Zn2FPU2, Zn2FPU3]> {
  let BufferSize=36;
}

// The Zen FP Retire Queue renames SIMD and FP uOps onto a pool of 160 128-bit
// registers. Operations on 256-bit data types are cracked into two COPs.
// Reference: "Software Optimization Guide for AMD Family 17h Processors"
def Zn2FpuPRF: RegisterFile<160, [VR64, VR128, VR256], [1, 1, 2]>;

// The unit can track up to 192 macro ops in-flight.
// The retire unit handles in-order commit of up to 8 macro ops per cycle.
// Reference: "Software Optimization Guide for AMD Family 17h Processors"
// To be noted, the retire unit is shared between integer and FP ops.
// In SMT mode it is 96 entry per thread. But, we do not use the conservative
// value here because there is currently no way to fully mode the SMT mode,
// so there is no point in trying.
def Zn2RCU : RetireControlUnit<192, 8>;

// (a folded load is an instruction that loads and does some operation)
// Ex: ADDPD xmm,[mem]-> This instruction has two micro-ops
// Instructions with folded loads are usually micro-fused, so they only appear
// as two micro-ops.
//      a. load and
//      b. addpd
// This multiclass is for folded loads for integer units.
multiclass Zn2WriteResPair<X86FoldableSchedWrite SchedRW,
                          list<ProcResourceKind> ExePorts,
                          int Lat, list<int> Res = [], int UOps = 1,
                          int LoadLat = 4, int LoadUOps = 1> {
  // Register variant takes 1-cycle on Execution Port.
  def : WriteRes<SchedRW, ExePorts> {
    let Latency = Lat;
    let ResourceCycles = Res;
    let NumMicroOps = UOps;
  }

  // Memory variant also uses a cycle on Zn2AGU
  // adds LoadLat cycles to the latency (default = 4).
  def : WriteRes<SchedRW.Folded, !listconcat([Zn2AGU], ExePorts)> {
    let Latency = !add(Lat, LoadLat);
    let ResourceCycles = !if(!empty(Res), [], !listconcat([1], Res));
    let NumMicroOps = !add(UOps, LoadUOps);
  }
}

// This multiclass is for folded loads for floating point units.
multiclass Zn2WriteResFpuPair<X86FoldableSchedWrite SchedRW,
                          list<ProcResourceKind> ExePorts,
                          int Lat, list<int> Res = [], int UOps = 1,
                          int LoadLat = 7, int LoadUOps = 0> {
  // Register variant takes 1-cycle on Execution Port.
  def : WriteRes<SchedRW, ExePorts> {
    let Latency = Lat;
    let ResourceCycles = Res;
    let NumMicroOps = UOps;
  }

  // Memory variant also uses a cycle on Zn2AGU
  // adds LoadLat cycles to the latency (default = 7).
  def : WriteRes<SchedRW.Folded, !listconcat([Zn2AGU], ExePorts)> {
    let Latency = !add(Lat, LoadLat);
    let ResourceCycles = !if(!empty(Res), [], !listconcat([1], Res));
    let NumMicroOps = !add(UOps, LoadUOps);
  }
}

// WriteRMW is set for instructions with Memory write
// operation in codegen
def : WriteRes<WriteRMW, [Zn2AGU]>;

def : WriteRes<WriteStore,   [Zn2AGU]>;
def : WriteRes<WriteStoreNT, [Zn2AGU]>;
def : WriteRes<WriteMove,    [Zn2ALU]>;
def : WriteRes<WriteLoad,    [Zn2AGU]> { let Latency = 8; }

def : WriteRes<WriteZero,  []>;
def : WriteRes<WriteLEA, [Zn2ALU]>;
defm : Zn2WriteResPair<WriteALU,   [Zn2ALU], 1>;
defm : Zn2WriteResPair<WriteADC,   [Zn2ALU], 1>;

defm : Zn2WriteResPair<WriteIMul8,     [Zn2ALU1, Zn2Multiplier], 4>;

defm : X86WriteRes<WriteBSWAP32, [Zn2ALU], 1, [4], 1>;
defm : X86WriteRes<WriteBSWAP64, [Zn2ALU], 1, [4], 1>;
defm : X86WriteRes<WriteCMPXCHG, [Zn2ALU], 3, [1], 1>;
defm : X86WriteRes<WriteCMPXCHGRMW,[Zn2ALU,Zn2AGU], 8, [1,1], 5>;
defm : X86WriteRes<WriteXCHG, [Zn2ALU], 1, [2], 2>;

defm : Zn2WriteResPair<WriteShift, [Zn2ALU], 1>;
defm : Zn2WriteResPair<WriteShiftCL,  [Zn2ALU], 1>;
defm : Zn2WriteResPair<WriteRotate,   [Zn2ALU], 1>;
defm : Zn2WriteResPair<WriteRotateCL, [Zn2ALU], 1>;

defm : X86WriteRes<WriteSHDrri, [Zn2ALU], 1, [1], 1>;
defm : X86WriteResUnsupported<WriteSHDrrcl>;
defm : X86WriteResUnsupported<WriteSHDmri>;
defm : X86WriteResUnsupported<WriteSHDmrcl>;

defm : Zn2WriteResPair<WriteJump,  [Zn2ALU], 1>;
defm : Zn2WriteResFpuPair<WriteCRC32, [Zn2FPU0], 3>;

defm : Zn2WriteResPair<WriteCMOV,   [Zn2ALU], 1>;
def  : WriteRes<WriteSETCC,  [Zn2ALU]>;
def  : WriteRes<WriteSETCCStore,  [Zn2ALU, Zn2AGU]>;
defm : X86WriteRes<WriteLAHFSAHF, [Zn2ALU], 2, [1], 2>;

defm : X86WriteRes<WriteBitTest,         [Zn2ALU], 1, [1], 1>;
defm : X86WriteRes<WriteBitTestImmLd,    [Zn2ALU,Zn2AGU], 5, [1,1], 2>;
defm : X86WriteRes<WriteBitTestRegLd,    [Zn2ALU,Zn2AGU], 5, [1,1], 2>;
defm : X86WriteRes<WriteBitTestSet,      [Zn2ALU], 2, [1], 2>;

// Bit counts.
defm : Zn2WriteResPair<WriteBSF, [Zn2ALU], 3>;
defm : Zn2WriteResPair<WriteBSR, [Zn2ALU], 4>;
defm : Zn2WriteResPair<WriteLZCNT,          [Zn2ALU], 1>;
defm : Zn2WriteResPair<WriteTZCNT,          [Zn2ALU], 2>;
defm : Zn2WriteResPair<WritePOPCNT,         [Zn2ALU], 1>;

// Treat misc copies as a move.
def : InstRW<[WriteMove], (instrs COPY)>;

// BMI1 BEXTR, BMI2 BZHI
defm : Zn2WriteResPair<WriteBEXTR, [Zn2ALU], 1>;
defm : Zn2WriteResPair<WriteBZHI, [Zn2ALU], 1>;

// IDIV
defm : Zn2WriteResPair<WriteDiv8,   [Zn2ALU2, Zn2Divider], 15, [1,15], 1>;
defm : Zn2WriteResPair<WriteDiv16,  [Zn2ALU2, Zn2Divider], 17, [1,17], 2>;
defm : Zn2WriteResPair<WriteDiv32,  [Zn2ALU2, Zn2Divider], 25, [1,25], 2>;
defm : Zn2WriteResPair<WriteDiv64,  [Zn2ALU2, Zn2Divider], 41, [1,41], 2>;
defm : Zn2WriteResPair<WriteIDiv8,  [Zn2ALU2, Zn2Divider], 15, [1,15], 1>;
defm : Zn2WriteResPair<WriteIDiv16, [Zn2ALU2, Zn2Divider], 17, [1,17], 2>;
defm : Zn2WriteResPair<WriteIDiv32, [Zn2ALU2, Zn2Divider], 25, [1,25], 2>;
defm : Zn2WriteResPair<WriteIDiv64, [Zn2ALU2, Zn2Divider], 41, [1,41], 2>;

// IMULH
def  : WriteRes<WriteIMulH, [Zn2ALU1, Zn2Multiplier]>{
  let Latency = 4;
}

// Floating point operations
defm : X86WriteRes<WriteFLoad,         [Zn2AGU], 8, [1], 1>;
defm : X86WriteRes<WriteFLoadX,        [Zn2AGU], 8, [1], 1>;
defm : X86WriteRes<WriteFLoadY,        [Zn2AGU], 8, [1], 1>;
defm : X86WriteRes<WriteFMaskedLoad,   [Zn2AGU,Zn2FPU01], 8, [1,1], 1>;
defm : X86WriteRes<WriteFMaskedLoadY,  [Zn2AGU,Zn2FPU01], 8, [1,1], 2>;
defm : X86WriteRes<WriteFMaskedStore32,  [Zn2AGU,Zn2FPU01], 4, [1,1], 1>;
defm : X86WriteRes<WriteFMaskedStore32Y, [Zn2AGU,Zn2FPU01], 5, [1,2], 2>;
defm : X86WriteRes<WriteFMaskedStore64,  [Zn2AGU,Zn2FPU01], 4, [1,1], 1>;
defm : X86WriteRes<WriteFMaskedStore64Y, [Zn2AGU,Zn2FPU01], 5, [1,2], 2>;

defm : X86WriteRes<WriteFStore,        [Zn2AGU], 1, [1], 1>;
defm : X86WriteRes<WriteFStoreX,       [Zn2AGU], 1, [1], 1>;
defm : X86WriteRes<WriteFStoreY,       [Zn2AGU], 1, [1], 1>;
defm : X86WriteRes<WriteFStoreNT,      [Zn2AGU,Zn2FPU2], 8, [1,1], 1>;
defm : X86WriteRes<WriteFStoreNTX,     [Zn2AGU], 1, [1], 1>;
defm : X86WriteRes<WriteFStoreNTY,     [Zn2AGU], 1, [1], 1>;
defm : X86WriteRes<WriteFMove,         [Zn2FPU], 1, [1], 1>;
defm : X86WriteRes<WriteFMoveX,        [Zn2FPU], 1, [1], 1>;
defm : X86WriteRes<WriteFMoveY,        [Zn2FPU], 1, [1], 1>;

defm : Zn2WriteResFpuPair<WriteFAdd,      [Zn2FPU0],  3>;
defm : Zn2WriteResFpuPair<WriteFAddX,     [Zn2FPU0],  3>;
defm : Zn2WriteResFpuPair<WriteFAddY,     [Zn2FPU0],  3>;
defm : X86WriteResPairUnsupported<WriteFAddZ>;
defm : Zn2WriteResFpuPair<WriteFAdd64,    [Zn2FPU0],  3>;
defm : Zn2WriteResFpuPair<WriteFAdd64X,   [Zn2FPU0],  3>;
defm : Zn2WriteResFpuPair<WriteFAdd64Y,   [Zn2FPU0],  3>;
defm : X86WriteResPairUnsupported<WriteFAdd64Z>;
defm : Zn2WriteResFpuPair<WriteFCmp,      [Zn2FPU0],  1>;
defm : Zn2WriteResFpuPair<WriteFCmpX,     [Zn2FPU0],  1>;
defm : Zn2WriteResFpuPair<WriteFCmpY,     [Zn2FPU0],  1>;
defm : X86WriteResPairUnsupported<WriteFCmpZ>;
defm : Zn2WriteResFpuPair<WriteFCmp64,    [Zn2FPU0],  1>;
defm : Zn2WriteResFpuPair<WriteFCmp64X,   [Zn2FPU0],  1>;
defm : Zn2WriteResFpuPair<WriteFCmp64Y,   [Zn2FPU0],  1>;
defm : X86WriteResPairUnsupported<WriteFCmp64Z>;
defm : Zn2WriteResFpuPair<WriteFCom,      [Zn2FPU0],  3>;
defm : Zn2WriteResFpuPair<WriteFComX,     [Zn2FPU0],  3>;
defm : Zn2WriteResFpuPair<WriteFBlend,    [Zn2FPU01], 1>;
defm : Zn2WriteResFpuPair<WriteFBlendY,   [Zn2FPU01], 1>;
defm : X86WriteResPairUnsupported<WriteFBlendZ>;
defm : Zn2WriteResFpuPair<WriteFVarBlend, [Zn2FPU01], 1>;
defm : Zn2WriteResFpuPair<WriteFVarBlendY,[Zn2FPU01], 1>;
defm : X86WriteResPairUnsupported<WriteFVarBlendZ>;
defm : Zn2WriteResFpuPair<WriteVarBlend,  [Zn2FPU0],  1>;
defm : Zn2WriteResFpuPair<WriteVarBlendY, [Zn2FPU0],  1>;
defm : X86WriteResPairUnsupported<WriteVarBlendZ>;
defm : Zn2WriteResFpuPair<WriteCvtSS2I,   [Zn2FPU3],  5>;
defm : Zn2WriteResFpuPair<WriteCvtPS2I,   [Zn2FPU3],  5>;
defm : Zn2WriteResFpuPair<WriteCvtPS2IY,  [Zn2FPU3],  5>;
defm : X86WriteResPairUnsupported<WriteCvtPS2IZ>;
defm : Zn2WriteResFpuPair<WriteCvtSD2I,   [Zn2FPU3],  5>;
defm : Zn2WriteResFpuPair<WriteCvtPD2I,   [Zn2FPU3],  5>;
defm : Zn2WriteResFpuPair<WriteCvtPD2IY,  [Zn2FPU3],  5>;
defm : X86WriteResPairUnsupported<WriteCvtPD2IZ>;
defm : Zn2WriteResFpuPair<WriteCvtI2SS,   [Zn2FPU3],  5>;
defm : Zn2WriteResFpuPair<WriteCvtI2PS,   [Zn2FPU3],  5>;
defm : Zn2WriteResFpuPair<WriteCvtI2PSY,  [Zn2FPU3],  5>;
defm : X86WriteResPairUnsupported<WriteCvtI2PSZ>;
defm : Zn2WriteResFpuPair<WriteCvtI2SD,   [Zn2FPU3],  5>;
defm : Zn2WriteResFpuPair<WriteCvtI2PD,   [Zn2FPU3],  5>;
defm : Zn2WriteResFpuPair<WriteCvtI2PDY,  [Zn2FPU3],  5>;
defm : X86WriteResPairUnsupported<WriteCvtI2PDZ>;
defm : Zn2WriteResFpuPair<WriteFDiv,      [Zn2FPU3], 15>;
defm : Zn2WriteResFpuPair<WriteFDivX,     [Zn2FPU3], 15>;
defm : X86WriteResPairUnsupported<WriteFDivZ>;
defm : Zn2WriteResFpuPair<WriteFDiv64,    [Zn2FPU3], 15>;
defm : Zn2WriteResFpuPair<WriteFDiv64X,   [Zn2FPU3], 15>;
defm : X86WriteResPairUnsupported<WriteFDiv64Z>;
defm : Zn2WriteResFpuPair<WriteFSign,     [Zn2FPU3],  2>;
defm : Zn2WriteResFpuPair<WriteFRnd,      [Zn2FPU3],  3, [1], 1, 7, 0>;
defm : Zn2WriteResFpuPair<WriteFRndY,     [Zn2FPU3],  3, [1], 1, 7, 0>;
defm : X86WriteResPairUnsupported<WriteFRndZ>;
defm : Zn2WriteResFpuPair<WriteFLogic,    [Zn2FPU],   1>;
defm : Zn2WriteResFpuPair<WriteFLogicY,   [Zn2FPU],   1>;
defm : X86WriteResPairUnsupported<WriteFLogicZ>;
defm : Zn2WriteResFpuPair<WriteFTest,     [Zn2FPU],   1>;
defm : Zn2WriteResFpuPair<WriteFTestY,    [Zn2FPU],   1>;
defm : X86WriteResPairUnsupported<WriteFTestZ>;
defm : Zn2WriteResFpuPair<WriteFShuffle,  [Zn2FPU12], 1>;
defm : Zn2WriteResFpuPair<WriteFShuffleY, [Zn2FPU12], 1>;
defm : X86WriteResPairUnsupported<WriteFShuffleZ>;
defm : Zn2WriteResFpuPair<WriteFVarShuffle, [Zn2FPU12], 3>;
defm : Zn2WriteResFpuPair<WriteFVarShuffleY,[Zn2FPU12], 3>;
defm : X86WriteResPairUnsupported<WriteFVarShuffleZ>;
defm : Zn2WriteResFpuPair<WriteFMul,      [Zn2FPU01], 3, [1], 1, 7, 1>;
defm : Zn2WriteResFpuPair<WriteFMulX,     [Zn2FPU01], 3, [1], 1, 7, 1>;
defm : Zn2WriteResFpuPair<WriteFMulY,     [Zn2FPU01], 3, [1], 1, 7, 1>;
defm : X86WriteResPairUnsupported<WriteFMulZ>;
defm : Zn2WriteResFpuPair<WriteFMul64,    [Zn2FPU01], 3, [1], 1, 7, 1>;
defm : Zn2WriteResFpuPair<WriteFMul64X,   [Zn2FPU01], 3, [1], 1, 7, 1>;
defm : Zn2WriteResFpuPair<WriteFMul64Y,   [Zn2FPU01], 3, [1], 1, 7, 1>;
defm : X86WriteResPairUnsupported<WriteFMul64Z>;
defm : Zn2WriteResFpuPair<WriteFMA,       [Zn2FPU03], 5>;
defm : Zn2WriteResFpuPair<WriteFMAX,      [Zn2FPU03], 5>;
defm : Zn2WriteResFpuPair<WriteFMAY,      [Zn2FPU03], 5>;
defm : X86WriteResPairUnsupported<WriteFMAZ>;
defm : Zn2WriteResFpuPair<WriteFRcp,      [Zn2FPU01], 5>;
defm : Zn2WriteResFpuPair<WriteFRcpX,     [Zn2FPU01], 5>;
defm : Zn2WriteResFpuPair<WriteFRcpY,     [Zn2FPU01], 5, [1], 1, 7, 2>;
defm : X86WriteResPairUnsupported<WriteFRcpZ>;
defm : Zn2WriteResFpuPair<WriteFRsqrtX,   [Zn2FPU01], 5, [1], 1, 7, 1>;
defm : X86WriteResPairUnsupported<WriteFRsqrtZ>;
defm : Zn2WriteResFpuPair<WriteFSqrt,     [Zn2FPU3], 20, [20]>;
defm : Zn2WriteResFpuPair<WriteFSqrtX,    [Zn2FPU3], 20, [20]>;
defm : Zn2WriteResFpuPair<WriteFSqrtY,    [Zn2FPU3], 28, [28], 1, 7, 1>;
defm : X86WriteResPairUnsupported<WriteFSqrtZ>;
defm : Zn2WriteResFpuPair<WriteFSqrt64,   [Zn2FPU3], 20, [20]>;
defm : Zn2WriteResFpuPair<WriteFSqrt64X,  [Zn2FPU3], 20, [20]>;
defm : Zn2WriteResFpuPair<WriteFSqrt64Y,  [Zn2FPU3], 20, [20], 1, 7, 1>;
defm : X86WriteResPairUnsupported<WriteFSqrt64Z>;
defm : Zn2WriteResFpuPair<WriteFSqrt80,   [Zn2FPU3], 20, [20]>;

// Vector integer operations which uses FPU units
defm : X86WriteRes<WriteVecLoad,         [Zn2AGU], 8, [1], 1>;
defm : X86WriteRes<WriteVecLoadX,        [Zn2AGU], 8, [1], 1>;
defm : X86WriteRes<WriteVecLoadY,        [Zn2AGU], 8, [1], 1>;
defm : X86WriteRes<WriteVecLoadNT,       [Zn2AGU], 8, [1], 1>;
defm : X86WriteRes<WriteVecLoadNTY,      [Zn2AGU], 8, [1], 1>;
defm : X86WriteRes<WriteVecMaskedLoad,   [Zn2AGU,Zn2FPU01], 8, [1,2], 2>;
defm : X86WriteRes<WriteVecMaskedLoadY,  [Zn2AGU,Zn2FPU01], 8, [1,2], 2>;
defm : X86WriteRes<WriteVecStore,        [Zn2AGU], 1, [1], 1>;
defm : X86WriteRes<WriteVecStoreX,       [Zn2AGU], 1, [1], 1>;
defm : X86WriteRes<WriteVecStoreY,       [Zn2AGU], 1, [1], 1>;
defm : X86WriteRes<WriteVecStoreNT,      [Zn2AGU], 1, [1], 1>;
defm : X86WriteRes<WriteVecStoreNTY,     [Zn2AGU], 1, [1], 1>;
defm : X86WriteRes<WriteVecMaskedStore32,  [Zn2AGU,Zn2FPU01], 4, [1,1], 1>;
defm : X86WriteRes<WriteVecMaskedStore32Y, [Zn2AGU,Zn2FPU01], 5, [1,2], 2>;
defm : X86WriteRes<WriteVecMaskedStore64,  [Zn2AGU,Zn2FPU01], 4, [1,1], 1>;
defm : X86WriteRes<WriteVecMaskedStore64Y, [Zn2AGU,Zn2FPU01], 5, [1,2], 2>;
defm : X86WriteRes<WriteVecMove,         [Zn2FPU], 1, [1], 1>;
defm : X86WriteRes<WriteVecMoveX,        [Zn2FPU], 1, [1], 1>;
defm : X86WriteRes<WriteVecMoveY,        [Zn2FPU], 2, [1], 2>;
defm : X86WriteRes<WriteVecMoveToGpr,    [Zn2FPU2], 2, [1], 1>;
defm : X86WriteRes<WriteVecMoveFromGpr,  [Zn2FPU2], 3, [1], 1>;
defm : X86WriteRes<WriteEMMS,            [Zn2FPU], 2, [1], 1>;

defm : Zn2WriteResFpuPair<WriteVecShift,   [Zn2FPU],   1>;
defm : Zn2WriteResFpuPair<WriteVecShiftX,  [Zn2FPU2],  1>;
defm : Zn2WriteResFpuPair<WriteVecShiftY,  [Zn2FPU2],  1>;
defm : X86WriteResPairUnsupported<WriteVecShiftZ>;
defm : Zn2WriteResFpuPair<WriteVecShiftImm,  [Zn2FPU], 1>;
defm : Zn2WriteResFpuPair<WriteVecShiftImmX, [Zn2FPU], 1>;
defm : Zn2WriteResFpuPair<WriteVecShiftImmY, [Zn2FPU], 1>;
defm : X86WriteResPairUnsupported<WriteVecShiftImmZ>;
defm : Zn2WriteResFpuPair<WriteVecLogic,   [Zn2FPU],   1>;
defm : Zn2WriteResFpuPair<WriteVecLogicX,  [Zn2FPU],   1>;
defm : Zn2WriteResFpuPair<WriteVecLogicY,  [Zn2FPU],   1>;
defm : X86WriteResPairUnsupported<WriteVecLogicZ>;
defm : Zn2WriteResFpuPair<WriteVecTest,    [Zn2FPU12], 1, [2], 1, 7, 1>;
defm : Zn2WriteResFpuPair<WriteVecTestY,   [Zn2FPU12], 1, [2], 1, 7, 1>;
defm : X86WriteResPairUnsupported<WriteVecTestZ>;
defm : Zn2WriteResFpuPair<WriteVecALU,     [Zn2FPU],   1>;
defm : Zn2WriteResFpuPair<WriteVecALUX,    [Zn2FPU],   1>;
defm : Zn2WriteResFpuPair<WriteVecALUY,    [Zn2FPU],   1>;
defm : X86WriteResPairUnsupported<WriteVecALUZ>;
defm : Zn2WriteResFpuPair<WriteVecIMul,    [Zn2FPU0],  4>;
defm : Zn2WriteResFpuPair<WriteVecIMulX,   [Zn2FPU0],  4>;
defm : Zn2WriteResFpuPair<WriteVecIMulY,   [Zn2FPU0],  4>;
defm : X86WriteResPairUnsupported<WriteVecIMulZ>;
defm : Zn2WriteResFpuPair<WritePMULLD,     [Zn2FPU0],  4, [1], 1, 7, 1>;
defm : Zn2WriteResFpuPair<WritePMULLDY,    [Zn2FPU0],  4, [1], 1, 7, 1>;
defm : X86WriteResPairUnsupported<WritePMULLDZ>;
defm : Zn2WriteResFpuPair<WriteShuffle,    [Zn2FPU],   1>;
defm : Zn2WriteResFpuPair<WriteShuffleX,   [Zn2FPU],   1>;
defm : Zn2WriteResFpuPair<WriteShuffleY,   [Zn2FPU],   1>;
defm : X86WriteResPairUnsupported<WriteShuffleZ>;
defm : Zn2WriteResFpuPair<WriteVarShuffle, [Zn2FPU],   1>;
defm : Zn2WriteResFpuPair<WriteVarShuffleX,[Zn2FPU],   1>;
defm : Zn2WriteResFpuPair<WriteVarShuffleY,[Zn2FPU],   1>;
defm : X86WriteResPairUnsupported<WriteVarShuffleZ>;
defm : Zn2WriteResFpuPair<WriteBlend,      [Zn2FPU01], 1>;
defm : Zn2WriteResFpuPair<WriteBlendY,     [Zn2FPU01], 1>;
defm : X86WriteResPairUnsupported<WriteBlendZ>;
defm : Zn2WriteResFpuPair<WriteShuffle256, [Zn2FPU],   2>;
defm : Zn2WriteResFpuPair<WriteVarShuffle256, [Zn2FPU],   2>;
defm : Zn2WriteResFpuPair<WritePSADBW,     [Zn2FPU0],  3>;
defm : Zn2WriteResFpuPair<WritePSADBWX,    [Zn2FPU0],  3>;
defm : Zn2WriteResFpuPair<WritePSADBWY,    [Zn2FPU0],  3>;
defm : X86WriteResPairUnsupported<WritePSADBWZ>;
defm : Zn2WriteResFpuPair<WritePHMINPOS,   [Zn2FPU0],  4>;

// Vector Shift Operations
defm : Zn2WriteResFpuPair<WriteVarVecShift,  [Zn2FPU12], 3>;
defm : Zn2WriteResFpuPair<WriteVarVecShiftY, [Zn2FPU12], 3>;
defm : X86WriteResPairUnsupported<WriteVarVecShiftZ>;

// Vector insert/extract operations.
defm : Zn2WriteResFpuPair<WriteVecInsert,   [Zn2FPU],   1>;

def : WriteRes<WriteVecExtract, [Zn2FPU12, Zn2FPU2]> {
  let Latency = 2;
  let ResourceCycles = [1, 2];
}
def : WriteRes<WriteVecExtractSt, [Zn2AGU, Zn2FPU12, Zn2FPU2]> {
  let Latency = 5;
  let NumMicroOps = 2;
  let ResourceCycles = [1, 2, 3];
}

// MOVMSK Instructions.
def : WriteRes<WriteFMOVMSK, [Zn2FPU2]>;
def : WriteRes<WriteMMXMOVMSK, [Zn2FPU2]>;
def : WriteRes<WriteVecMOVMSK, [Zn2FPU2]>;

def : WriteRes<WriteVecMOVMSKY, [Zn2FPU2]> {
  let NumMicroOps = 2;
  let Latency = 2;
  let ResourceCycles = [2];
}

// AES Instructions.
defm : Zn2WriteResFpuPair<WriteAESDecEnc, [Zn2FPU01], 4>;
defm : Zn2WriteResFpuPair<WriteAESIMC,    [Zn2FPU01], 4>;
defm : Zn2WriteResFpuPair<WriteAESKeyGen, [Zn2FPU01], 4>;

def : WriteRes<WriteFence,  [Zn2AGU]>;
def : WriteRes<WriteNop, []>;

// Following instructions with latency=100 are microcoded.
// We set long latency so as to block the entire pipeline.
defm : Zn2WriteResFpuPair<WriteFShuffle256, [Zn2FPU], 100>;
defm : Zn2WriteResFpuPair<WriteFVarShuffle256, [Zn2FPU], 100>;

// Microcoded Instructions
def Zn2WriteMicrocoded : SchedWriteRes<[]> {
  let Latency = 100;
}
defm : Zn2WriteResPair<WriteDPPS, [], 15>;
defm : Zn2WriteResPair<WriteFHAdd, [], 7>;
defm : Zn2WriteResPair<WriteFHAddY, [], 7>;
defm : Zn2WriteResPair<WritePHAdd, [], 3>;
defm : Zn2WriteResPair<WritePHAddX, [], 3>;
defm : Zn2WriteResPair<WritePHAddY, [], 3>;

def : SchedAlias<WriteMicrocoded, Zn2WriteMicrocoded>;
def : SchedAlias<WriteFCMOV, Zn2WriteMicrocoded>;
def : SchedAlias<WriteSystem, Zn2WriteMicrocoded>;
def : SchedAlias<WriteMPSAD, Zn2WriteMicrocoded>;
def : SchedAlias<WriteMPSADY, Zn2WriteMicrocoded>;
def : SchedAlias<WriteMPSADLd, Zn2WriteMicrocoded>;
def : SchedAlias<WriteMPSADYLd, Zn2WriteMicrocoded>;
def : SchedAlias<WriteCLMul, Zn2WriteMicrocoded>;
def : SchedAlias<WriteCLMulLd, Zn2WriteMicrocoded>;
def : SchedAlias<WritePCmpIStrM, Zn2WriteMicrocoded>;
def : SchedAlias<WritePCmpIStrMLd, Zn2WriteMicrocoded>;
def : SchedAlias<WritePCmpEStrI, Zn2WriteMicrocoded>;
def : SchedAlias<WritePCmpEStrILd, Zn2WriteMicrocoded>;
def : SchedAlias<WritePCmpEStrM, Zn2WriteMicrocoded>;
def : SchedAlias<WritePCmpEStrMLd, Zn2WriteMicrocoded>;
def : SchedAlias<WritePCmpIStrI, Zn2WriteMicrocoded>;
def : SchedAlias<WritePCmpIStrILd, Zn2WriteMicrocoded>;
def : SchedAlias<WriteLDMXCSR, Zn2WriteMicrocoded>;
def : SchedAlias<WriteSTMXCSR, Zn2WriteMicrocoded>;

//=== Regex based InstRW ===//
// Notation:
// - r: register.
// - m = memory.
// - i = immediate
// - mm: 64 bit mmx register.
// - x = 128 bit xmm register.
// - (x)mm = mmx or xmm register.
// - y = 256 bit ymm register.
// - v = any vector register.

//=== Integer Instructions ===//
//-- Move instructions --//
// MOV.
// r16,m.
def : InstRW<[WriteALULd, ReadAfterLd], (instregex "MOV16rm")>;

// MOVSX, MOVZX.
// r,m.
def : InstRW<[WriteLoad], (instregex "MOV(S|Z)X32rm(8|16)")>;

// XCHG.
// r,r.
def Zn2WriteXCHG : SchedWriteRes<[Zn2ALU]> {
  let NumMicroOps = 2;
}

def : InstRW<[Zn2WriteXCHG], (instregex "^XCHG(8|16|32|64)rr", "^XCHG(16|32|64)ar")>;

// r,m.
def Zn2WriteXCHGrm : SchedWriteRes<[Zn2AGU, Zn2ALU]> {
  let Latency = 5;
  let NumMicroOps = 2;
}
def : InstRW<[Zn2WriteXCHGrm, ReadAfterLd], (instregex "^XCHG(8|16|32|64)rm")>;

def : InstRW<[WriteMicrocoded], (instrs XLAT)>;

// POP16.
// r.
def Zn2WritePop16r : SchedWriteRes<[Zn2AGU]>{
  let Latency = 5;
  let NumMicroOps = 2;
}
def : InstRW<[Zn2WritePop16r], (instregex "POP16rmm")>;
def : InstRW<[WriteMicrocoded], (instregex "POPF(16|32)")>;
def : InstRW<[WriteMicrocoded], (instregex "POPA(16|32)")>;


// PUSH.
// r. Has default values.
// m.
def Zn2WritePUSH : SchedWriteRes<[Zn2AGU]>{
  let Latency = 4;
}
def : InstRW<[Zn2WritePUSH], (instregex "PUSH(16|32)rmm")>;

//PUSHF
def : InstRW<[WriteMicrocoded], (instregex "PUSHF(16|32)")>;

// PUSHA.
def Zn2WritePushA : SchedWriteRes<[Zn2AGU]> {
  let Latency = 8;
}
def : InstRW<[Zn2WritePushA], (instregex "PUSHA(16|32)")>;

//LAHF
def : InstRW<[WriteMicrocoded], (instrs LAHF)>;

// MOVBE.
// r,m.
def Zn2WriteMOVBE : SchedWriteRes<[Zn2AGU, Zn2ALU]> {
  let Latency = 5;
}
def : InstRW<[Zn2WriteMOVBE, ReadAfterLd], (instregex "MOVBE(16|32|64)rm")>;

// m16,r16.
def : InstRW<[Zn2WriteMOVBE], (instregex "MOVBE(16|32|64)mr")>;

//-- Arithmetic instructions --//

// ADD SUB.
// m,r/i.
def : InstRW<[WriteALULd], (instregex "(ADD|SUB)(8|16|32|64)m(r|i)",
                          "(ADD|SUB)(8|16|32|64)mi8",
                          "(ADD|SUB)64mi32")>;

// ADC SBB.
// m,r/i.
def : InstRW<[WriteALULd],
             (instregex "(ADC|SBB)(8|16|32|64)m(r|i)",
              "(ADC|SBB)(16|32|64)mi8",
              "(ADC|SBB)64mi32")>;

// INC DEC NOT NEG.
// m.
def : InstRW<[WriteALULd],
             (instregex "(INC|DEC|NOT|NEG)(8|16|32|64)m")>;

// MUL IMUL.
// r16.
def Zn2WriteMul16 : SchedWriteRes<[Zn2ALU1, Zn2Multiplier]> {
  let Latency = 3;
}
def Zn2WriteMul16Imm : SchedWriteRes<[Zn2ALU1, Zn2Multiplier]> {
  let Latency = 4;
}
def : SchedAlias<WriteIMul16, Zn2WriteMul16>;
def : SchedAlias<WriteIMul16Imm, Zn2WriteMul16Imm>;
def : SchedAlias<WriteIMul16Reg, Zn2WriteMul16>;

// m16.
def Zn2WriteMul16Ld : SchedWriteRes<[Zn2AGU, Zn2ALU1, Zn2Multiplier]> {
  let Latency = 7;
}
def : SchedAlias<WriteIMul16Ld, Zn2WriteMul16Ld>;
def : SchedAlias<WriteIMul16ImmLd, Zn2WriteMul16Ld>;
def : SchedAlias<WriteIMul16RegLd, Zn2WriteMul16Ld>;

// r32.
def Zn2WriteMul32 : SchedWriteRes<[Zn2ALU1, Zn2Multiplier]> {
  let Latency = 3;
}
def : SchedAlias<WriteIMul32, Zn2WriteMul32>;
def : SchedAlias<WriteIMul32Imm, Zn2WriteMul32>;
def : SchedAlias<WriteIMul32Reg, Zn2WriteMul32>;

// m32.
def Zn2WriteMul32Ld : SchedWriteRes<[Zn2AGU, Zn2ALU1, Zn2Multiplier]> {
  let Latency = 7;
}
def : SchedAlias<WriteIMul32Ld, Zn2WriteMul32Ld>;
def : SchedAlias<WriteIMul32ImmLd, Zn2WriteMul32Ld>;
def : SchedAlias<WriteIMul32RegLd, Zn2WriteMul32Ld>;

// r64.
def Zn2WriteMul64 : SchedWriteRes<[Zn2ALU1, Zn2Multiplier]> {
  let Latency = 4;
  let NumMicroOps = 2;
}
def : SchedAlias<WriteIMul64, Zn2WriteMul64>;
def : SchedAlias<WriteIMul64Imm, Zn2WriteMul64>;
def : SchedAlias<WriteIMul64Reg, Zn2WriteMul64>;

// m64.
def Zn2WriteMul64Ld : SchedWriteRes<[Zn2AGU, Zn2ALU1, Zn2Multiplier]> {
  let Latency = 8;
  let NumMicroOps = 2;
}
def : SchedAlias<WriteIMul64Ld, Zn2WriteMul64Ld>;
def : SchedAlias<WriteIMul64ImmLd, Zn2WriteMul64Ld>;
def : SchedAlias<WriteIMul64RegLd, Zn2WriteMul64Ld>;

// MULX.
// r32,r32,r32.
def Zn2WriteMulX32 : SchedWriteRes<[Zn2ALU1, Zn2Multiplier]> {
  let Latency = 3;
  let ResourceCycles = [1, 2];
}
def : InstRW<[Zn2WriteMulX32], (instrs MULX32rr)>;

// r32,r32,m32.
def Zn2WriteMulX32Ld : SchedWriteRes<[Zn2AGU, Zn2ALU1, Zn2Multiplier]> {
  let Latency = 7;
  let ResourceCycles = [1, 2, 2];
}
def : InstRW<[Zn2WriteMulX32Ld, ReadAfterLd], (instrs MULX32rm)>;

// r64,r64,r64.
def Zn2WriteMulX64 : SchedWriteRes<[Zn2ALU1]> {
  let Latency = 3;
}
def : InstRW<[Zn2WriteMulX64], (instrs MULX64rr)>;

// r64,r64,m64.
def Zn2WriteMulX64Ld : SchedWriteRes<[Zn2AGU, Zn2ALU1, Zn2Multiplier]> {
  let Latency = 7;
}
def : InstRW<[Zn2WriteMulX64Ld, ReadAfterLd], (instrs MULX64rm)>;

//-- Control transfer instructions --//

// J(E|R)CXZ.
def Zn2WriteJCXZ : SchedWriteRes<[Zn2ALU03]>;
def : InstRW<[Zn2WriteJCXZ], (instrs JCXZ, JECXZ, JRCXZ)>;

// INTO
def : InstRW<[WriteMicrocoded], (instrs INTO)>;

// LOOP.
def Zn2WriteLOOP : SchedWriteRes<[Zn2ALU03]>;
def : InstRW<[Zn2WriteLOOP], (instrs LOOP)>;

// LOOP(N)E, LOOP(N)Z
def Zn2WriteLOOPE : SchedWriteRes<[Zn2ALU03]>;
def : InstRW<[Zn2WriteLOOPE], (instrs LOOPE, LOOPNE)>;

// CALL.
// r.
def Zn2WriteCALLr : SchedWriteRes<[Zn2AGU, Zn2ALU03]>;
def : InstRW<[Zn2WriteCALLr], (instregex "CALL(16|32)r")>;

def : InstRW<[WriteMicrocoded], (instregex "CALL(16|32)m")>;

// RET.
def Zn2WriteRET : SchedWriteRes<[Zn2ALU03]> {
  let NumMicroOps = 2;
}
def : InstRW<[Zn2WriteRET], (instregex "RET(L|Q|W)", "LRET(L|Q|W)",
                            "IRET(16|32|64)")>;

//-- Logic instructions --//

// AND OR XOR.
// m,r/i.
def : InstRW<[WriteALULd],
             (instregex "(AND|OR|XOR)(8|16|32|64)m(r|i)",
              "(AND|OR|XOR)(8|16|32|64)mi8", "(AND|OR|XOR)64mi32")>;

// Define ALU latency variants
def Zn2WriteALULat2 : SchedWriteRes<[Zn2ALU]> {
  let Latency = 2;
}
def Zn2WriteALULat2Ld : SchedWriteRes<[Zn2AGU, Zn2ALU]> {
  let Latency = 6;
}

// BT.
// m,i.
def : InstRW<[WriteShiftLd], (instregex "BT(16|32|64)mi8")>;

// BTR BTS BTC.
// r,r,i.
def Zn2WriteBTRSC : SchedWriteRes<[Zn2ALU]> {
  let Latency = 2;
  let NumMicroOps = 2;
}
def : InstRW<[Zn2WriteBTRSC], (instregex "BT(R|S|C)(16|32|64)r(r|i8)")>;

// m,r,i.
def Zn2WriteBTRSCm : SchedWriteRes<[Zn2AGU, Zn2ALU]> {
  let Latency = 6;
  let NumMicroOps = 2;
}
// m,r,i.
def : SchedAlias<WriteBitTestSetImmRMW, Zn2WriteBTRSCm>;
def : SchedAlias<WriteBitTestSetRegRMW, Zn2WriteBTRSCm>;

// BLSI BLSMSK BLSR.
// r,r.
def : SchedAlias<WriteBLS, Zn2WriteALULat2>;
// r,m.
def : SchedAlias<WriteBLSLd, Zn2WriteALULat2Ld>;

// CLD STD.
def : InstRW<[WriteALU], (instrs STD, CLD)>;

// PDEP PEXT.
// r,r,r.
def : InstRW<[WriteMicrocoded], (instregex "PDEP(32|64)rr", "PEXT(32|64)rr")>;
// r,r,m.
def : InstRW<[WriteMicrocoded], (instregex "PDEP(32|64)rm", "PEXT(32|64)rm")>;

// RCR RCL.
// m,i.
def : InstRW<[WriteMicrocoded], (instregex "RC(R|L)(8|16|32|64)m(1|i|CL)")>;

// SHR SHL SAR.
// m,i.
def : InstRW<[WriteShiftLd], (instregex "S(A|H)(R|L)(8|16|32|64)m(i|1)")>;

// SHRD SHLD.
// m,r
def : InstRW<[WriteShiftLd], (instregex "SH(R|L)D(16|32|64)mri8")>;

// r,r,cl.
def : InstRW<[WriteMicrocoded], (instregex "SH(R|L)D(16|32|64)rrCL")>;

// m,r,cl.
def : InstRW<[WriteMicrocoded], (instregex "SH(R|L)D(16|32|64)mrCL")>;

//-- Misc instructions --//
// CMPXCHG8B.
def Zn2WriteCMPXCHG8B : SchedWriteRes<[Zn2AGU, Zn2ALU]> {
  let NumMicroOps = 18;
}
def : InstRW<[Zn2WriteCMPXCHG8B], (instrs CMPXCHG8B)>;

def : InstRW<[WriteMicrocoded], (instrs CMPXCHG16B)>;

// LEAVE
def Zn2WriteLEAVE : SchedWriteRes<[Zn2ALU, Zn2AGU]> {
  let Latency = 8;
  let NumMicroOps = 2;
}
def : InstRW<[Zn2WriteLEAVE], (instregex "LEAVE")>;

// PAUSE.
def : InstRW<[WriteMicrocoded], (instrs PAUSE)>;

// RDTSC.
def : InstRW<[WriteMicrocoded], (instregex "RDTSC")>;

// RDPMC.
def : InstRW<[WriteMicrocoded], (instrs RDPMC)>;

// RDRAND.
def : InstRW<[WriteMicrocoded], (instregex "RDRAND(16|32|64)r")>;

// XGETBV.
def : InstRW<[WriteMicrocoded], (instregex "XGETBV")>;

//-- String instructions --//
// CMPS.
def : InstRW<[WriteMicrocoded], (instregex "CMPS(B|L|Q|W)")>;

// LODSB/W.
def : InstRW<[WriteMicrocoded], (instregex "LODS(B|W)")>;

// LODSD/Q.
def : InstRW<[WriteMicrocoded], (instregex "LODS(L|Q)")>;

// MOVS.
def : InstRW<[WriteMicrocoded], (instregex "MOVS(B|L|Q|W)")>;

// SCAS.
def : InstRW<[WriteMicrocoded], (instregex "SCAS(B|W|L|Q)")>;

// STOS
def : InstRW<[WriteMicrocoded], (instregex "STOS(B|L|Q|W)")>;

// XADD.
def Zn2XADD : SchedWriteRes<[Zn2ALU]>;
def : InstRW<[Zn2XADD], (instregex "XADD(8|16|32|64)rr")>;
def : InstRW<[WriteMicrocoded], (instregex "XADD(8|16|32|64)rm")>;

//=== Floating Point x87 Instructions ===//
//-- Move instructions --//

def Zn2WriteFLDr : SchedWriteRes<[Zn2FPU13]> ;

def Zn2WriteSTr: SchedWriteRes<[Zn2FPU23]> {
  let Latency = 5;
  let NumMicroOps = 2;
}

// LD_F.
// r.
def : InstRW<[Zn2WriteFLDr], (instregex "LD_Frr")>;

// m.
def Zn2WriteLD_F80m : SchedWriteRes<[Zn2AGU, Zn2FPU13]> {
  let NumMicroOps = 2;
}
def : InstRW<[Zn2WriteLD_F80m], (instregex "LD_F80m")>;

// FBLD.
def : InstRW<[WriteMicrocoded], (instregex "FBLDm")>;

// FST(P).
// r.
def : InstRW<[Zn2WriteSTr], (instregex "ST_(F|FP)rr")>;

// m80.
def Zn2WriteST_FP80m : SchedWriteRes<[Zn2AGU, Zn2FPU23]> {
  let Latency = 5;
}
def : InstRW<[Zn2WriteST_FP80m], (instregex "ST_FP80m")>;

// FBSTP.
// m80.
def : InstRW<[WriteMicrocoded], (instregex "FBSTPm")>;

def Zn2WriteFXCH : SchedWriteRes<[Zn2FPU]>;

// FXCHG.
def : InstRW<[Zn2WriteFXCH], (instrs XCH_F)>;

// FILD.
def Zn2WriteFILD : SchedWriteRes<[Zn2AGU, Zn2FPU3]> {
  let Latency = 11;
  let NumMicroOps = 2;
}
def : InstRW<[Zn2WriteFILD], (instregex "ILD_F(16|32|64)m")>;

// FIST(P) FISTTP.
def Zn2WriteFIST : SchedWriteRes<[Zn2AGU, Zn2FPU23]> {
  let Latency = 12;
}
def : InstRW<[Zn2WriteFIST], (instregex "IS(T|TT)_(F|FP)(16|32|64)m")>;

def Zn2WriteFPU13 : SchedWriteRes<[Zn2AGU, Zn2FPU13]> {
  let Latency = 8;
}

def Zn2WriteFPU3 : SchedWriteRes<[Zn2AGU, Zn2FPU3]> {
  let Latency = 11;
}

// FLDZ.
def : SchedAlias<WriteFLD0, Zn2WriteFPU13>;

// FLD1.
def : SchedAlias<WriteFLD1, Zn2WriteFPU3>;

// FLDPI FLDL2E etc.
def : SchedAlias<WriteFLDC, Zn2WriteFPU3>;

// FNSTSW.
// AX.
def : InstRW<[WriteMicrocoded], (instrs FNSTSW16r)>;

// m16.
def : InstRW<[WriteMicrocoded], (instrs FNSTSWm)>;

// FLDCW.
def : InstRW<[WriteMicrocoded], (instrs FLDCW16m)>;

// FNSTCW.
def : InstRW<[WriteMicrocoded], (instrs FNSTCW16m)>;

// FINCSTP FDECSTP.
def : InstRW<[Zn2WriteFPU3], (instrs FINCSTP, FDECSTP)>;

// FFREE.
def : InstRW<[Zn2WriteFPU3], (instregex "FFREE")>;

// FNSAVE.
def : InstRW<[WriteMicrocoded], (instregex "FSAVEm")>;

// FRSTOR.
def : InstRW<[WriteMicrocoded], (instregex "FRSTORm")>;

//-- Arithmetic instructions --//

def Zn2WriteFPU3Lat1 : SchedWriteRes<[Zn2FPU3]> ;

def Zn2WriteFPU0Lat1 : SchedWriteRes<[Zn2FPU0]> ;

def Zn2WriteFPU0Lat1Ld : SchedWriteRes<[Zn2AGU, Zn2FPU0]> {
  let Latency = 8;
}

// FCHS.
def : InstRW<[Zn2WriteFPU3Lat1], (instregex "CHS_F")>;

// FCOM(P) FUCOM(P).
// r.
def : InstRW<[Zn2WriteFPU0Lat1], (instregex "COM(P?)_FST0r", "UCOM_F(P?)r")>;
// m.
def : InstRW<[Zn2WriteFPU0Lat1Ld], (instregex "FCOM(P?)(32|64)m")>;

// FCOMPP FUCOMPP.
// r.
def : InstRW<[Zn2WriteFPU0Lat1], (instrs FCOMPP, UCOM_FPPr)>;

def Zn2WriteFPU02 : SchedWriteRes<[Zn2AGU, Zn2FPU02]>
{
  let Latency = 9;
}

// FCOMI(P) FUCOMI(P).
// m.
def : InstRW<[Zn2WriteFPU02], (instrs COM_FIPr, COM_FIr, UCOM_FIPr, UCOM_FIr)>;

def Zn2WriteFPU03 : SchedWriteRes<[Zn2AGU, Zn2FPU03]>
{
  let Latency = 12;
  let NumMicroOps = 2;
  let ResourceCycles = [1,3];
}

// FICOM(P).
def : InstRW<[Zn2WriteFPU03], (instregex "FICOM(P?)(16|32)m")>;

// FTST.
def : InstRW<[Zn2WriteFPU0Lat1], (instregex "TST_F")>;

// FXAM.
def : InstRW<[Zn2WriteFPU3Lat1], (instrs FXAM)>;

// FPREM.
def : InstRW<[WriteMicrocoded], (instrs FPREM)>;

// FPREM1.
def : InstRW<[WriteMicrocoded], (instrs FPREM1)>;

// FRNDINT.
def : InstRW<[WriteMicrocoded], (instrs FRNDINT)>;

// FSCALE.
def : InstRW<[WriteMicrocoded], (instrs FSCALE)>;

// FXTRACT.
def : InstRW<[WriteMicrocoded], (instrs FXTRACT)>;

// FNOP.
def : InstRW<[Zn2WriteFPU0Lat1], (instrs FNOP)>;

// WAIT.
def : InstRW<[Zn2WriteFPU0Lat1], (instrs WAIT)>;

// FNCLEX.
def : InstRW<[WriteMicrocoded], (instrs FNCLEX)>;

// FNINIT.
def : InstRW<[WriteMicrocoded], (instrs FNINIT)>;

//=== Integer MMX and XMM Instructions ===//

// PACKSSWB/DW.
// mm <- mm.
def Zn2WriteFPU12 : SchedWriteRes<[Zn2FPU12]> ;
def Zn2WriteFPU12Y : SchedWriteRes<[Zn2FPU12]> {
  let Latency = 4;
  let NumMicroOps = 2;
}
def Zn2WriteFPU12m : SchedWriteRes<[Zn2AGU, Zn2FPU12]> ;
def Zn2WriteFPU12Ym : SchedWriteRes<[Zn2AGU, Zn2FPU12]> {
  let Latency = 8;
  let NumMicroOps = 2;
}

def : InstRW<[Zn2WriteFPU12], (instrs MMX_PACKSSDWirr,
                                     MMX_PACKSSWBirr,
                                     MMX_PACKUSWBirr)>;
def : InstRW<[Zn2WriteFPU12m], (instrs MMX_PACKSSDWirm,
                                      MMX_PACKSSWBirm,
                                      MMX_PACKUSWBirm)>;

// VPMOVSX/ZX BW BD BQ WD WQ DQ.
// y <- x.
def : InstRW<[Zn2WriteFPU12Y], (instregex "VPMOV(SX|ZX)(BW|BD|BQ|WD|WQ|DQ)Yrr")>;
def : InstRW<[Zn2WriteFPU12Ym], (instregex "VPMOV(SX|ZX)(BW|BD|BQ|WD|WQ|DQ)Yrm")>;

def Zn2WriteFPU013 : SchedWriteRes<[Zn2FPU013]> ;
def Zn2WriteFPU013Y : SchedWriteRes<[Zn2FPU013]> ;
def Zn2WriteFPU013m : SchedWriteRes<[Zn2AGU, Zn2FPU013]> {
  let Latency = 8;
  let NumMicroOps = 2;
}
def Zn2WriteFPU013Ld : SchedWriteRes<[Zn2AGU, Zn2FPU013]> {
  let Latency = 8;
  let NumMicroOps = 2;
}
def Zn2WriteFPU013LdY : SchedWriteRes<[Zn2AGU, Zn2FPU013]> {
  let Latency = 8;
  let NumMicroOps = 2;
}

// PBLENDW.
// x,x,i / v,v,v,i
def : InstRW<[Zn2WriteFPU013], (instregex "(V?)PBLENDWrri")>;
// ymm
def : InstRW<[Zn2WriteFPU013Y], (instrs VPBLENDWYrri)>;

// x,m,i / v,v,m,i
def : InstRW<[Zn2WriteFPU013Ld], (instregex "(V?)PBLENDWrmi")>;
// y,m,i
def : InstRW<[Zn2WriteFPU013LdY], (instrs VPBLENDWYrmi)>;

def Zn2WriteFPU01 : SchedWriteRes<[Zn2FPU01]> ;
def Zn2WriteFPU01Y : SchedWriteRes<[Zn2FPU01]> {
  let NumMicroOps = 2;
}

// VPBLENDD.
// v,v,v,i.
def : InstRW<[Zn2WriteFPU01], (instrs VPBLENDDrri)>;
// ymm
def : InstRW<[Zn2WriteFPU01Y], (instrs VPBLENDDYrri)>;

// v,v,m,i
def Zn2WriteFPU01Op2 : SchedWriteRes<[Zn2AGU, Zn2FPU01]> {
  let NumMicroOps = 2;
  let Latency = 8;
  let ResourceCycles = [1, 2];
}
def Zn2WriteFPU01Op2Y : SchedWriteRes<[Zn2AGU, Zn2FPU01]> {
  let NumMicroOps = 2;
  let Latency = 9;
  let ResourceCycles = [1, 3];
}
def : InstRW<[Zn2WriteFPU01Op2], (instrs VPBLENDDrmi)>;
def : InstRW<[Zn2WriteFPU01Op2Y], (instrs VPBLENDDYrmi)>;

// MASKMOVQ.
def : InstRW<[WriteMicrocoded], (instregex "MMX_MASKMOVQ(64)?")>;

// MASKMOVDQU.
def : InstRW<[WriteMicrocoded], (instregex "(V?)MASKMOVDQU(64)?")>;

// VPMASKMOVD.
// ymm
def : InstRW<[WriteMicrocoded],
                               (instregex "VPMASKMOVD(Y?)rm")>;
// m, v,v.
def : InstRW<[WriteMicrocoded], (instregex "VPMASKMOV(D|Q)(Y?)mr")>;

// VPBROADCAST B/W.
// x, m8/16.
def Zn2WriteVPBROADCAST128Ld : SchedWriteRes<[Zn2AGU, Zn2FPU12]> {
  let Latency = 8;
  let NumMicroOps = 2;
  let ResourceCycles = [1, 2];
}
def : InstRW<[Zn2WriteVPBROADCAST128Ld],
                                     (instregex "VPBROADCAST(B|W)rm")>;

// y, m8/16
def Zn2WriteVPBROADCAST256Ld : SchedWriteRes<[Zn2AGU, Zn2FPU1]> {
  let Latency = 8;
  let NumMicroOps = 2;
  let ResourceCycles = [1, 2];
}
def : InstRW<[Zn2WriteVPBROADCAST256Ld],
                                     (instregex "VPBROADCAST(B|W)Yrm")>;

// VPGATHER.
def : InstRW<[WriteMicrocoded], (instregex "VPGATHER(Q|D)(Q|D)(Y?)rm")>;

//-- Arithmetic instructions --//

// PCMPGTQ.
def Zn2WritePCMPGTQr : SchedWriteRes<[Zn2FPU03]>;
def : InstRW<[Zn2WritePCMPGTQr], (instregex "(V?)PCMPGTQ(Y?)rr")>;

// x <- x,m.
def Zn2WritePCMPGTQm : SchedWriteRes<[Zn2AGU, Zn2FPU03]> {
  let Latency = 8;
}
// ymm.
def Zn2WritePCMPGTQYm : SchedWriteRes<[Zn2AGU, Zn2FPU03]> {
  let Latency = 8;
}
def : InstRW<[Zn2WritePCMPGTQm], (instregex "(V?)PCMPGTQrm")>;
def : InstRW<[Zn2WritePCMPGTQYm], (instrs VPCMPGTQYrm)>;

//-- Logic instructions --//

// PSLL,PSRL,PSRA W/D/Q.
// x,x / v,v,x.
def Zn2WritePShift  : SchedWriteRes<[Zn2FPU2]> {
  let Latency = 3;
}
def Zn2WritePShiftY : SchedWriteRes<[Zn2FPU2]> {
  let Latency = 3;
}

// PSLL,PSRL DQ.
def : InstRW<[Zn2WritePShift], (instregex "(V?)PS(R|L)LDQri")>;
def : InstRW<[Zn2WritePShiftY], (instregex "(V?)PS(R|L)LDQYri")>;

//=== Floating Point XMM and YMM Instructions ===//
//-- Move instructions --//

// VPERM2F128.
def : InstRW<[WriteMicrocoded], (instrs VPERM2F128rr)>;
def : InstRW<[WriteMicrocoded], (instrs VPERM2F128rm)>;

def Zn2WriteBROADCAST : SchedWriteRes<[Zn2AGU, Zn2FPU13]> {
  let NumMicroOps = 2;
  let Latency = 8;
}
// VBROADCASTF128.
def : InstRW<[Zn2WriteBROADCAST], (instrs VBROADCASTF128)>;

// EXTRACTPS.
// r32,x,i.
def Zn2WriteEXTRACTPSr : SchedWriteRes<[Zn2FPU12, Zn2FPU2]> {
  let Latency = 2;
  let ResourceCycles = [1, 2];
}
def : InstRW<[Zn2WriteEXTRACTPSr], (instregex "(V?)EXTRACTPSrr")>;

def Zn2WriteEXTRACTPSm : SchedWriteRes<[Zn2AGU,Zn2FPU12, Zn2FPU2]> {
  let Latency = 5;
  let NumMicroOps = 2;
  let ResourceCycles = [5, 1, 2];
}
// m32,x,i.
def : InstRW<[Zn2WriteEXTRACTPSm], (instregex "(V?)EXTRACTPSmr")>;

// VEXTRACTF128.
// x,y,i.
def : InstRW<[Zn2WriteFPU013], (instrs VEXTRACTF128rr)>;

// m128,y,i.
def : InstRW<[Zn2WriteFPU013m], (instrs VEXTRACTF128mr)>;

def Zn2WriteVINSERT128r: SchedWriteRes<[Zn2FPU013]> {
  let Latency = 2;
//  let ResourceCycles = [2];
}
def Zn2WriteVINSERT128Ld: SchedWriteRes<[Zn2AGU,Zn2FPU013]> {
  let Latency = 9;
  let NumMicroOps = 2;
}
// VINSERTF128.
// y,y,x,i.
def : InstRW<[Zn2WriteVINSERT128r], (instrs VINSERTF128rr)>;
def : InstRW<[Zn2WriteVINSERT128Ld], (instrs VINSERTF128rm)>;

// VGATHER.
def : InstRW<[WriteMicrocoded], (instregex "VGATHER(Q|D)(PD|PS)(Y?)rm")>;

//-- Conversion instructions --//
def Zn2WriteCVTPD2PSr: SchedWriteRes<[Zn2FPU3]> {
  let Latency = 3;
}
def Zn2WriteCVTPD2PSYr: SchedWriteRes<[Zn2FPU3]> {
  let Latency = 3;
}

// CVTPD2PS.
// x,x.
def : SchedAlias<WriteCvtPD2PS,  Zn2WriteCVTPD2PSr>;
// y,y.
def : SchedAlias<WriteCvtPD2PSY, Zn2WriteCVTPD2PSYr>;
// z,z.
defm : X86WriteResUnsupported<WriteCvtPD2PSZ>;

def Zn2WriteCVTPD2PSLd: SchedWriteRes<[Zn2AGU,Zn2FPU03]> {
  let Latency = 10;
  let NumMicroOps = 2;
}
// x,m128.
def : SchedAlias<WriteCvtPD2PSLd, Zn2WriteCVTPD2PSLd>;

// x,m256.
def Zn2WriteCVTPD2PSYLd : SchedWriteRes<[Zn2AGU, Zn2FPU3]> {
  let Latency = 10;
}
def : SchedAlias<WriteCvtPD2PSYLd, Zn2WriteCVTPD2PSYLd>;
// z,m512
defm : X86WriteResUnsupported<WriteCvtPD2PSZLd>;

// CVTSD2SS.
// x,x.
// Same as WriteCVTPD2PSr
def : SchedAlias<WriteCvtSD2SS, Zn2WriteCVTPD2PSr>;

// x,m64.
def : SchedAlias<WriteCvtSD2SSLd, Zn2WriteCVTPD2PSLd>;

// CVTPS2PD.
// x,x.
def Zn2WriteCVTPS2PDr : SchedWriteRes<[Zn2FPU3]> {
  let Latency = 3;
}
def : SchedAlias<WriteCvtPS2PD, Zn2WriteCVTPS2PDr>;

// x,m64.
// y,m128.
def Zn2WriteCVTPS2PDLd : SchedWriteRes<[Zn2AGU, Zn2FPU3]> {
  let Latency = 10;
  let NumMicroOps = 2;
}
def : SchedAlias<WriteCvtPS2PDLd, Zn2WriteCVTPS2PDLd>;
def : SchedAlias<WriteCvtPS2PDYLd, Zn2WriteCVTPS2PDLd>;
defm : X86WriteResUnsupported<WriteCvtPS2PDZLd>;

// y,x.
def Zn2WriteVCVTPS2PDY : SchedWriteRes<[Zn2FPU3]> {
  let Latency = 3;
}
def : SchedAlias<WriteCvtPS2PDY, Zn2WriteVCVTPS2PDY>;
defm : X86WriteResUnsupported<WriteCvtPS2PDZ>;

// CVTSS2SD.
// x,x.
def Zn2WriteCVTSS2SDr : SchedWriteRes<[Zn2FPU3]> {
  let Latency = 3;
}
def : SchedAlias<WriteCvtSS2SD, Zn2WriteCVTSS2SDr>;

// x,m32.
def Zn2WriteCVTSS2SDLd : SchedWriteRes<[Zn2AGU, Zn2FPU3]> {
  let Latency = 10;
  let NumMicroOps = 2;
  let ResourceCycles = [1, 2];
}
def : SchedAlias<WriteCvtSS2SDLd, Zn2WriteCVTSS2SDLd>;

def Zn2WriteCVTDQ2PDr: SchedWriteRes<[Zn2FPU12,Zn2FPU3]> {
  let Latency = 3;
}
// CVTDQ2PD.
// x,x.
def : InstRW<[Zn2WriteCVTDQ2PDr], (instregex "(V)?CVTDQ2P(D|S)rr")>;

// Same as xmm
// y,x.
def : InstRW<[Zn2WriteCVTDQ2PDr], (instrs VCVTDQ2PDYrr)>;
def : InstRW<[Zn2WriteCVTDQ2PDr], (instrs VCVTDQ2PSYrr)>;

def Zn2WriteCVTPD2DQr: SchedWriteRes<[Zn2FPU12, Zn2FPU3]> {
  let Latency = 3;
}
// CVT(T)P(D|S)2DQ.
// x,x.
def : InstRW<[Zn2WriteCVTPD2DQr], (instregex "(V?)CVT(T?)P(D|S)2DQrr")>;

def Zn2WriteCVTPD2DQLd: SchedWriteRes<[Zn2AGU,Zn2FPU12,Zn2FPU3]> {
  let Latency = 10;
  let NumMicroOps = 2;
}
// x,m128.
def : InstRW<[Zn2WriteCVTPD2DQLd], (instregex "(V?)CVT(T?)PD2DQrm")>;
// same as xmm handling
// x,y.
def : InstRW<[Zn2WriteCVTPD2DQr], (instregex "VCVT(T?)PD2DQYrr")>;
// x,m256.
def : InstRW<[Zn2WriteCVTPD2DQLd], (instregex "VCVT(T?)PD2DQYrm")>;

def Zn2WriteCVTPS2PIr: SchedWriteRes<[Zn2FPU3]> {
  let Latency = 4;
}
// CVT(T)PS2PI.
// mm,x.
def : InstRW<[Zn2WriteCVTPS2PIr], (instregex "MMX_CVT(T?)PS2PIirr")>;

// CVTPI2PD.
// x,mm.
def : InstRW<[Zn2WriteCVTPS2PDr], (instrs MMX_CVTPI2PDirr)>;

// CVT(T)PD2PI.
// mm,x.
def : InstRW<[Zn2WriteCVTPS2PIr], (instregex "MMX_CVT(T?)PD2PIirr")>;

def Zn2WriteCVSTSI2SSr: SchedWriteRes<[Zn2FPU3]> {
  let Latency = 3;
}

// same as CVTPD2DQr
// CVT(T)SS2SI.
// r32,x.
def : InstRW<[Zn2WriteCVTPD2DQr], (instregex "(V?)CVT(T?)SS2SI(64)?rr")>;
// same as CVTPD2DQm
// r32,m32.
def : InstRW<[Zn2WriteCVTPD2DQLd], (instregex "(V?)CVT(T?)SS2SI(64)?rm")>;

def Zn2WriteCVSTSI2SDr: SchedWriteRes<[Zn2FPU013, Zn2FPU3]> {
  let Latency = 3;
}
// CVTSI2SD.
// x,r32/64.
def : InstRW<[Zn2WriteCVSTSI2SDr], (instregex "(V?)CVTSI(64)?2SDrr")>;


def Zn2WriteCVSTSI2SIr: SchedWriteRes<[Zn2FPU3, Zn2FPU2]> {
  let Latency = 4;
}
def Zn2WriteCVSTSI2SILd: SchedWriteRes<[Zn2AGU, Zn2FPU3, Zn2FPU2]> {
  let Latency = 11;
}
// CVTSD2SI.
// r32/64
def : InstRW<[Zn2WriteCVSTSI2SIr], (instregex "(V?)CVT(T?)SD2SI(64)?rr")>;
// r32,m32.
def : InstRW<[Zn2WriteCVSTSI2SILd], (instregex "(V?)CVT(T?)SD2SI(64)?rm")>;

// VCVTPS2PH.
// x,v,i.
def : SchedAlias<WriteCvtPS2PH,    Zn2WriteMicrocoded>;
def : SchedAlias<WriteCvtPS2PHY,   Zn2WriteMicrocoded>;
defm : X86WriteResUnsupported<WriteCvtPS2PHZ>;
// m,v,i.
def : SchedAlias<WriteCvtPS2PHSt,  Zn2WriteMicrocoded>;
def : SchedAlias<WriteCvtPS2PHYSt, Zn2WriteMicrocoded>;
defm : X86WriteResUnsupported<WriteCvtPS2PHZSt>;

// VCVTPH2PS.
// v,x.
def : SchedAlias<WriteCvtPH2PS,    Zn2WriteMicrocoded>;
def : SchedAlias<WriteCvtPH2PSY,   Zn2WriteMicrocoded>;
defm : X86WriteResUnsupported<WriteCvtPH2PSZ>;
// v,m.
def : SchedAlias<WriteCvtPH2PSLd,  Zn2WriteMicrocoded>;
def : SchedAlias<WriteCvtPH2PSYLd, Zn2WriteMicrocoded>;
defm : X86WriteResUnsupported<WriteCvtPH2PSZLd>;

//-- SSE4A instructions --//
// EXTRQ
def Zn2WriteEXTRQ: SchedWriteRes<[Zn2FPU12, Zn2FPU2]> {
  let Latency = 3;
}
def : InstRW<[Zn2WriteEXTRQ], (instregex "EXTRQ")>;

// INSERTQ
def Zn2WriteINSERTQ: SchedWriteRes<[Zn2FPU03,Zn2FPU1]> {
  let Latency = 4;
}
def : InstRW<[Zn2WriteINSERTQ], (instregex "INSERTQ")>;

//-- SHA instructions --//
// SHA256MSG2
def : InstRW<[WriteMicrocoded], (instregex "SHA256MSG2(Y?)r(r|m)")>;

// SHA1MSG1, SHA256MSG1
// x,x.
def Zn2WriteSHA1MSG1r : SchedWriteRes<[Zn2FPU12]> {
  let Latency = 2;
}
def : InstRW<[Zn2WriteSHA1MSG1r], (instregex "SHA(1|256)MSG1rr")>;
// x,m.
def Zn2WriteSHA1MSG1Ld : SchedWriteRes<[Zn2AGU, Zn2FPU12]> {
  let Latency = 9;
}
def : InstRW<[Zn2WriteSHA1MSG1Ld], (instregex "SHA(1|256)MSG1rm")>;

// SHA1MSG2
// x,x.
def Zn2WriteSHA1MSG2r : SchedWriteRes<[Zn2FPU12]> ;
def : InstRW<[Zn2WriteSHA1MSG2r], (instregex "SHA1MSG2rr")>;
// x,m.
def Zn2WriteSHA1MSG2Ld : SchedWriteRes<[Zn2AGU, Zn2FPU12]> {
  let Latency = 8;
}
def : InstRW<[Zn2WriteSHA1MSG2Ld], (instregex "SHA1MSG2rm")>;

// SHA1NEXTE
// x,x.
def Zn2WriteSHA1NEXTEr : SchedWriteRes<[Zn2FPU1]> ;
def : InstRW<[Zn2WriteSHA1NEXTEr], (instregex "SHA1NEXTErr")>;
// x,m.
def Zn2WriteSHA1NEXTELd : SchedWriteRes<[Zn2AGU, Zn2FPU1]> {
  let Latency = 8;
}
def : InstRW<[Zn2WriteSHA1NEXTELd], (instregex "SHA1NEXTErm")>;

// SHA1RNDS4
// x,x.
def Zn2WriteSHA1RNDS4r : SchedWriteRes<[Zn2FPU1]> {
  let Latency = 6;
}
def : InstRW<[Zn2WriteSHA1RNDS4r], (instregex "SHA1RNDS4rr")>;
// x,m.
def Zn2WriteSHA1RNDS4Ld : SchedWriteRes<[Zn2AGU, Zn2FPU1]> {
  let Latency = 13;
}
def : InstRW<[Zn2WriteSHA1RNDS4Ld], (instregex "SHA1RNDS4rm")>;

// SHA256RNDS2
// x,x.
def Zn2WriteSHA256RNDS2r : SchedWriteRes<[Zn2FPU1]> {
  let Latency = 4;
}
def : InstRW<[Zn2WriteSHA256RNDS2r], (instregex "SHA256RNDS2rr")>;
// x,m.
def Zn2WriteSHA256RNDS2Ld : SchedWriteRes<[Zn2AGU, Zn2FPU1]> {
  let Latency = 11;
}
def : InstRW<[Zn2WriteSHA256RNDS2Ld], (instregex "SHA256RNDS2rm")>;

//-- Arithmetic instructions --//

// VDIVPS.
// TODO - convert to Zn2WriteResFpuPair
// y,y,y.
def Zn2WriteVDIVPSYr : SchedWriteRes<[Zn2FPU3]> {
  let Latency = 10;
  let ResourceCycles = [10];
}
def : SchedAlias<WriteFDivY,   Zn2WriteVDIVPSYr>;

// y,y,m256.
def Zn2WriteVDIVPSYLd : SchedWriteRes<[Zn2AGU, Zn2FPU3]> {
  let Latency = 17;
  let NumMicroOps = 2;
  let ResourceCycles = [1, 17];
}
def : SchedAlias<WriteFDivYLd,  Zn2WriteVDIVPSYLd>;

// VDIVPD.
// TODO - convert to Zn2WriteResFpuPair
// y,y,y.
def Zn2WriteVDIVPDY : SchedWriteRes<[Zn2FPU3]> {
  let Latency = 13;
  let ResourceCycles = [13];
}
def : SchedAlias<WriteFDiv64Y, Zn2WriteVDIVPDY>;

// y,y,m256.
def Zn2WriteVDIVPDYLd : SchedWriteRes<[Zn2AGU, Zn2FPU3]> {
  let Latency = 20;
  let NumMicroOps = 2;
  let ResourceCycles = [1,20];
}
def : SchedAlias<WriteFDiv64YLd, Zn2WriteVDIVPDYLd>;

// DPPS.
// x,x,i / v,v,v,i.
def : SchedAlias<WriteDPPSY,  Zn2WriteMicrocoded>;

// x,m,i / v,v,m,i.
def : SchedAlias<WriteDPPSYLd,Zn2WriteMicrocoded>;

// DPPD.
// x,x,i.
def : SchedAlias<WriteDPPD,   Zn2WriteMicrocoded>;

// x,m,i.
def : SchedAlias<WriteDPPDLd, Zn2WriteMicrocoded>;

// RSQRTSS
// TODO - convert to Zn2WriteResFpuPair
// x,x.
def Zn2WriteRSQRTSSr : SchedWriteRes<[Zn2FPU02]> {
  let Latency = 5;
}
def : SchedAlias<WriteFRsqrt, Zn2WriteRSQRTSSr>;

// x,m128.
def Zn2WriteRSQRTSSLd: SchedWriteRes<[Zn2AGU, Zn2FPU02]> {
  let Latency = 12;
  let NumMicroOps = 2;
  let ResourceCycles = [1,2];
}
def : SchedAlias<WriteFRsqrtLd, Zn2WriteRSQRTSSLd>;

// RSQRTPS
// TODO - convert to Zn2WriteResFpuPair
// y,y.
def Zn2WriteRSQRTPSYr : SchedWriteRes<[Zn2FPU01]> {
  let Latency = 5;
  let NumMicroOps = 2;
  let ResourceCycles = [2];
}
def : SchedAlias<WriteFRsqrtY, Zn2WriteRSQRTPSYr>;

// y,m256.
def Zn2WriteRSQRTPSYLd : SchedWriteRes<[Zn2AGU, Zn2FPU01]> {
  let Latency = 12;
  let NumMicroOps = 2;
}
def : SchedAlias<WriteFRsqrtYLd, Zn2WriteRSQRTPSYLd>;

//-- Other instructions --//

// VZEROUPPER.
def : InstRW<[WriteALU], (instrs VZEROUPPER)>;

// VZEROALL.
def : InstRW<[WriteMicrocoded], (instrs VZEROALL)>;

} // SchedModel