aboutsummaryrefslogtreecommitdiffstats
path: root/contrib/libs/llvm12/lib/Transforms/InstCombine/InstCombinePHI.cpp
blob: b211b0813611a20424e7647404b9ca314a77f25a (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
//===- InstCombinePHI.cpp -------------------------------------------------===//
//
// 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 implements the visitPHINode function.
//
//===----------------------------------------------------------------------===//

#include "InstCombineInternal.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/ValueTracking.h"
#include "llvm/IR/PatternMatch.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Transforms/InstCombine/InstCombiner.h"
#include "llvm/Transforms/Utils/Local.h"

using namespace llvm;
using namespace llvm::PatternMatch;

#define DEBUG_TYPE "instcombine"

static cl::opt<unsigned>
MaxNumPhis("instcombine-max-num-phis", cl::init(512),
           cl::desc("Maximum number phis to handle in intptr/ptrint folding"));

STATISTIC(NumPHIsOfInsertValues,
          "Number of phi-of-insertvalue turned into insertvalue-of-phis");
STATISTIC(NumPHIsOfExtractValues,
          "Number of phi-of-extractvalue turned into extractvalue-of-phi");
STATISTIC(NumPHICSEs, "Number of PHI's that got CSE'd");

/// The PHI arguments will be folded into a single operation with a PHI node
/// as input. The debug location of the single operation will be the merged
/// locations of the original PHI node arguments.
void InstCombinerImpl::PHIArgMergedDebugLoc(Instruction *Inst, PHINode &PN) {
  auto *FirstInst = cast<Instruction>(PN.getIncomingValue(0));
  Inst->setDebugLoc(FirstInst->getDebugLoc());
  // We do not expect a CallInst here, otherwise, N-way merging of DebugLoc
  // will be inefficient.
  assert(!isa<CallInst>(Inst));

  for (unsigned i = 1; i != PN.getNumIncomingValues(); ++i) {
    auto *I = cast<Instruction>(PN.getIncomingValue(i));
    Inst->applyMergedLocation(Inst->getDebugLoc(), I->getDebugLoc());
  }
}

// Replace Integer typed PHI PN if the PHI's value is used as a pointer value.
// If there is an existing pointer typed PHI that produces the same value as PN,
// replace PN and the IntToPtr operation with it. Otherwise, synthesize a new
// PHI node:
//
// Case-1:
// bb1:
//     int_init = PtrToInt(ptr_init)
//     br label %bb2
// bb2:
//    int_val = PHI([int_init, %bb1], [int_val_inc, %bb2]
//    ptr_val = PHI([ptr_init, %bb1], [ptr_val_inc, %bb2]
//    ptr_val2 = IntToPtr(int_val)
//    ...
//    use(ptr_val2)
//    ptr_val_inc = ...
//    inc_val_inc = PtrToInt(ptr_val_inc)
//
// ==>
// bb1:
//     br label %bb2
// bb2:
//    ptr_val = PHI([ptr_init, %bb1], [ptr_val_inc, %bb2]
//    ...
//    use(ptr_val)
//    ptr_val_inc = ...
//
// Case-2:
// bb1:
//    int_ptr = BitCast(ptr_ptr)
//    int_init = Load(int_ptr)
//    br label %bb2
// bb2:
//    int_val = PHI([int_init, %bb1], [int_val_inc, %bb2]
//    ptr_val2 = IntToPtr(int_val)
//    ...
//    use(ptr_val2)
//    ptr_val_inc = ...
//    inc_val_inc = PtrToInt(ptr_val_inc)
// ==>
// bb1:
//    ptr_init = Load(ptr_ptr)
//    br label %bb2
// bb2:
//    ptr_val = PHI([ptr_init, %bb1], [ptr_val_inc, %bb2]
//    ...
//    use(ptr_val)
//    ptr_val_inc = ...
//    ...
//
Instruction *InstCombinerImpl::foldIntegerTypedPHI(PHINode &PN) {
  if (!PN.getType()->isIntegerTy())
    return nullptr;
  if (!PN.hasOneUse())
    return nullptr;

  auto *IntToPtr = dyn_cast<IntToPtrInst>(PN.user_back());
  if (!IntToPtr)
    return nullptr;

  // Check if the pointer is actually used as pointer:
  auto HasPointerUse = [](Instruction *IIP) {
    for (User *U : IIP->users()) {
      Value *Ptr = nullptr;
      if (LoadInst *LoadI = dyn_cast<LoadInst>(U)) {
        Ptr = LoadI->getPointerOperand();
      } else if (StoreInst *SI = dyn_cast<StoreInst>(U)) {
        Ptr = SI->getPointerOperand();
      } else if (GetElementPtrInst *GI = dyn_cast<GetElementPtrInst>(U)) {
        Ptr = GI->getPointerOperand();
      }

      if (Ptr && Ptr == IIP)
        return true;
    }
    return false;
  };

  if (!HasPointerUse(IntToPtr))
    return nullptr;

  if (DL.getPointerSizeInBits(IntToPtr->getAddressSpace()) !=
      DL.getTypeSizeInBits(IntToPtr->getOperand(0)->getType()))
    return nullptr;

  SmallVector<Value *, 4> AvailablePtrVals;
  for (unsigned i = 0; i != PN.getNumIncomingValues(); ++i) {
    Value *Arg = PN.getIncomingValue(i);

    // First look backward:
    if (auto *PI = dyn_cast<PtrToIntInst>(Arg)) {
      AvailablePtrVals.emplace_back(PI->getOperand(0));
      continue;
    }

    // Next look forward:
    Value *ArgIntToPtr = nullptr;
    for (User *U : Arg->users()) {
      if (isa<IntToPtrInst>(U) && U->getType() == IntToPtr->getType() &&
          (DT.dominates(cast<Instruction>(U), PN.getIncomingBlock(i)) ||
           cast<Instruction>(U)->getParent() == PN.getIncomingBlock(i))) {
        ArgIntToPtr = U;
        break;
      }
    }

    if (ArgIntToPtr) {
      AvailablePtrVals.emplace_back(ArgIntToPtr);
      continue;
    }

    // If Arg is defined by a PHI, allow it. This will also create
    // more opportunities iteratively.
    if (isa<PHINode>(Arg)) {
      AvailablePtrVals.emplace_back(Arg);
      continue;
    }

    // For a single use integer load:
    auto *LoadI = dyn_cast<LoadInst>(Arg);
    if (!LoadI)
      return nullptr;

    if (!LoadI->hasOneUse())
      return nullptr;

    // Push the integer typed Load instruction into the available
    // value set, and fix it up later when the pointer typed PHI
    // is synthesized.
    AvailablePtrVals.emplace_back(LoadI);
  }

  // Now search for a matching PHI
  auto *BB = PN.getParent();
  assert(AvailablePtrVals.size() == PN.getNumIncomingValues() &&
         "Not enough available ptr typed incoming values");
  PHINode *MatchingPtrPHI = nullptr;
  unsigned NumPhis = 0;
  for (auto II = BB->begin(); II != BB->end(); II++, NumPhis++) {
    // FIXME: consider handling this in AggressiveInstCombine
    PHINode *PtrPHI = dyn_cast<PHINode>(II);
    if (!PtrPHI)
      break;
    if (NumPhis > MaxNumPhis)
      return nullptr;
    if (PtrPHI == &PN || PtrPHI->getType() != IntToPtr->getType())
      continue;
    MatchingPtrPHI = PtrPHI;
    for (unsigned i = 0; i != PtrPHI->getNumIncomingValues(); ++i) {
      if (AvailablePtrVals[i] !=
          PtrPHI->getIncomingValueForBlock(PN.getIncomingBlock(i))) {
        MatchingPtrPHI = nullptr;
        break;
      }
    }

    if (MatchingPtrPHI)
      break;
  }

  if (MatchingPtrPHI) {
    assert(MatchingPtrPHI->getType() == IntToPtr->getType() &&
           "Phi's Type does not match with IntToPtr");
    // The PtrToCast + IntToPtr will be simplified later
    return CastInst::CreateBitOrPointerCast(MatchingPtrPHI,
                                            IntToPtr->getOperand(0)->getType());
  }

  // If it requires a conversion for every PHI operand, do not do it.
  if (all_of(AvailablePtrVals, [&](Value *V) {
        return (V->getType() != IntToPtr->getType()) || isa<IntToPtrInst>(V);
      }))
    return nullptr;

  // If any of the operand that requires casting is a terminator
  // instruction, do not do it. Similarly, do not do the transform if the value
  // is PHI in a block with no insertion point, for example, a catchswitch
  // block, since we will not be able to insert a cast after the PHI.
  if (any_of(AvailablePtrVals, [&](Value *V) {
        if (V->getType() == IntToPtr->getType())
          return false;
        auto *Inst = dyn_cast<Instruction>(V);
        if (!Inst)
          return false;
        if (Inst->isTerminator())
          return true;
        auto *BB = Inst->getParent();
        if (isa<PHINode>(Inst) && BB->getFirstInsertionPt() == BB->end())
          return true;
        return false;
      }))
    return nullptr;

  PHINode *NewPtrPHI = PHINode::Create(
      IntToPtr->getType(), PN.getNumIncomingValues(), PN.getName() + ".ptr");

  InsertNewInstBefore(NewPtrPHI, PN);
  SmallDenseMap<Value *, Instruction *> Casts;
  for (unsigned i = 0; i != PN.getNumIncomingValues(); ++i) {
    auto *IncomingBB = PN.getIncomingBlock(i);
    auto *IncomingVal = AvailablePtrVals[i];

    if (IncomingVal->getType() == IntToPtr->getType()) {
      NewPtrPHI->addIncoming(IncomingVal, IncomingBB);
      continue;
    }

#ifndef NDEBUG
    LoadInst *LoadI = dyn_cast<LoadInst>(IncomingVal);
    assert((isa<PHINode>(IncomingVal) ||
            IncomingVal->getType()->isPointerTy() ||
            (LoadI && LoadI->hasOneUse())) &&
           "Can not replace LoadInst with multiple uses");
#endif
    // Need to insert a BitCast.
    // For an integer Load instruction with a single use, the load + IntToPtr
    // cast will be simplified into a pointer load:
    // %v = load i64, i64* %a.ip, align 8
    // %v.cast = inttoptr i64 %v to float **
    // ==>
    // %v.ptrp = bitcast i64 * %a.ip to float **
    // %v.cast = load float *, float ** %v.ptrp, align 8
    Instruction *&CI = Casts[IncomingVal];
    if (!CI) {
      CI = CastInst::CreateBitOrPointerCast(IncomingVal, IntToPtr->getType(),
                                            IncomingVal->getName() + ".ptr");
      if (auto *IncomingI = dyn_cast<Instruction>(IncomingVal)) {
        BasicBlock::iterator InsertPos(IncomingI);
        InsertPos++;
        BasicBlock *BB = IncomingI->getParent();
        if (isa<PHINode>(IncomingI))
          InsertPos = BB->getFirstInsertionPt();
        assert(InsertPos != BB->end() && "should have checked above");
        InsertNewInstBefore(CI, *InsertPos);
      } else {
        auto *InsertBB = &IncomingBB->getParent()->getEntryBlock();
        InsertNewInstBefore(CI, *InsertBB->getFirstInsertionPt());
      }
    }
    NewPtrPHI->addIncoming(CI, IncomingBB);
  }

  // The PtrToCast + IntToPtr will be simplified later
  return CastInst::CreateBitOrPointerCast(NewPtrPHI,
                                          IntToPtr->getOperand(0)->getType());
}

/// If we have something like phi [insertvalue(a,b,0), insertvalue(c,d,0)],
/// turn this into a phi[a,c] and phi[b,d] and a single insertvalue.
Instruction *
InstCombinerImpl::foldPHIArgInsertValueInstructionIntoPHI(PHINode &PN) {
  auto *FirstIVI = cast<InsertValueInst>(PN.getIncomingValue(0));

  // Scan to see if all operands are `insertvalue`'s with the same indicies,
  // and all have a single use.
  for (unsigned i = 1; i != PN.getNumIncomingValues(); ++i) {
    auto *I = dyn_cast<InsertValueInst>(PN.getIncomingValue(i));
    if (!I || !I->hasOneUser() || I->getIndices() != FirstIVI->getIndices())
      return nullptr;
  }

  // For each operand of an `insertvalue`
  std::array<PHINode *, 2> NewOperands;
  for (int OpIdx : {0, 1}) {
    auto *&NewOperand = NewOperands[OpIdx];
    // Create a new PHI node to receive the values the operand has in each
    // incoming basic block.
    NewOperand = PHINode::Create(
        FirstIVI->getOperand(OpIdx)->getType(), PN.getNumIncomingValues(),
        FirstIVI->getOperand(OpIdx)->getName() + ".pn");
    // And populate each operand's PHI with said values.
    for (auto Incoming : zip(PN.blocks(), PN.incoming_values()))
      NewOperand->addIncoming(
          cast<InsertValueInst>(std::get<1>(Incoming))->getOperand(OpIdx),
          std::get<0>(Incoming));
    InsertNewInstBefore(NewOperand, PN);
  }

  // And finally, create `insertvalue` over the newly-formed PHI nodes.
  auto *NewIVI = InsertValueInst::Create(NewOperands[0], NewOperands[1],
                                         FirstIVI->getIndices(), PN.getName());

  PHIArgMergedDebugLoc(NewIVI, PN);
  ++NumPHIsOfInsertValues;
  return NewIVI;
}

/// If we have something like phi [extractvalue(a,0), extractvalue(b,0)],
/// turn this into a phi[a,b] and a single extractvalue.
Instruction *
InstCombinerImpl::foldPHIArgExtractValueInstructionIntoPHI(PHINode &PN) {
  auto *FirstEVI = cast<ExtractValueInst>(PN.getIncomingValue(0));

  // Scan to see if all operands are `extractvalue`'s with the same indicies,
  // and all have a single use.
  for (unsigned i = 1; i != PN.getNumIncomingValues(); ++i) {
    auto *I = dyn_cast<ExtractValueInst>(PN.getIncomingValue(i));
    if (!I || !I->hasOneUser() || I->getIndices() != FirstEVI->getIndices() ||
        I->getAggregateOperand()->getType() !=
            FirstEVI->getAggregateOperand()->getType())
      return nullptr;
  }

  // Create a new PHI node to receive the values the aggregate operand has
  // in each incoming basic block.
  auto *NewAggregateOperand = PHINode::Create(
      FirstEVI->getAggregateOperand()->getType(), PN.getNumIncomingValues(),
      FirstEVI->getAggregateOperand()->getName() + ".pn");
  // And populate the PHI with said values.
  for (auto Incoming : zip(PN.blocks(), PN.incoming_values()))
    NewAggregateOperand->addIncoming(
        cast<ExtractValueInst>(std::get<1>(Incoming))->getAggregateOperand(),
        std::get<0>(Incoming));
  InsertNewInstBefore(NewAggregateOperand, PN);

  // And finally, create `extractvalue` over the newly-formed PHI nodes.
  auto *NewEVI = ExtractValueInst::Create(NewAggregateOperand,
                                          FirstEVI->getIndices(), PN.getName());

  PHIArgMergedDebugLoc(NewEVI, PN);
  ++NumPHIsOfExtractValues;
  return NewEVI;
}

/// If we have something like phi [add (a,b), add(a,c)] and if a/b/c and the
/// adds all have a single user, turn this into a phi and a single binop.
Instruction *InstCombinerImpl::foldPHIArgBinOpIntoPHI(PHINode &PN) {
  Instruction *FirstInst = cast<Instruction>(PN.getIncomingValue(0));
  assert(isa<BinaryOperator>(FirstInst) || isa<CmpInst>(FirstInst));
  unsigned Opc = FirstInst->getOpcode();
  Value *LHSVal = FirstInst->getOperand(0);
  Value *RHSVal = FirstInst->getOperand(1);

  Type *LHSType = LHSVal->getType();
  Type *RHSType = RHSVal->getType();

  // Scan to see if all operands are the same opcode, and all have one user.
  for (unsigned i = 1; i != PN.getNumIncomingValues(); ++i) {
    Instruction *I = dyn_cast<Instruction>(PN.getIncomingValue(i));
    if (!I || I->getOpcode() != Opc || !I->hasOneUser() ||
        // Verify type of the LHS matches so we don't fold cmp's of different
        // types.
        I->getOperand(0)->getType() != LHSType ||
        I->getOperand(1)->getType() != RHSType)
      return nullptr;

    // If they are CmpInst instructions, check their predicates
    if (CmpInst *CI = dyn_cast<CmpInst>(I))
      if (CI->getPredicate() != cast<CmpInst>(FirstInst)->getPredicate())
        return nullptr;

    // Keep track of which operand needs a phi node.
    if (I->getOperand(0) != LHSVal) LHSVal = nullptr;
    if (I->getOperand(1) != RHSVal) RHSVal = nullptr;
  }

  // If both LHS and RHS would need a PHI, don't do this transformation,
  // because it would increase the number of PHIs entering the block,
  // which leads to higher register pressure. This is especially
  // bad when the PHIs are in the header of a loop.
  if (!LHSVal && !RHSVal)
    return nullptr;

  // Otherwise, this is safe to transform!

  Value *InLHS = FirstInst->getOperand(0);
  Value *InRHS = FirstInst->getOperand(1);
  PHINode *NewLHS = nullptr, *NewRHS = nullptr;
  if (!LHSVal) {
    NewLHS = PHINode::Create(LHSType, PN.getNumIncomingValues(),
                             FirstInst->getOperand(0)->getName() + ".pn");
    NewLHS->addIncoming(InLHS, PN.getIncomingBlock(0));
    InsertNewInstBefore(NewLHS, PN);
    LHSVal = NewLHS;
  }

  if (!RHSVal) {
    NewRHS = PHINode::Create(RHSType, PN.getNumIncomingValues(),
                             FirstInst->getOperand(1)->getName() + ".pn");
    NewRHS->addIncoming(InRHS, PN.getIncomingBlock(0));
    InsertNewInstBefore(NewRHS, PN);
    RHSVal = NewRHS;
  }

  // Add all operands to the new PHIs.
  if (NewLHS || NewRHS) {
    for (unsigned i = 1, e = PN.getNumIncomingValues(); i != e; ++i) {
      Instruction *InInst = cast<Instruction>(PN.getIncomingValue(i));
      if (NewLHS) {
        Value *NewInLHS = InInst->getOperand(0);
        NewLHS->addIncoming(NewInLHS, PN.getIncomingBlock(i));
      }
      if (NewRHS) {
        Value *NewInRHS = InInst->getOperand(1);
        NewRHS->addIncoming(NewInRHS, PN.getIncomingBlock(i));
      }
    }
  }

  if (CmpInst *CIOp = dyn_cast<CmpInst>(FirstInst)) {
    CmpInst *NewCI = CmpInst::Create(CIOp->getOpcode(), CIOp->getPredicate(),
                                     LHSVal, RHSVal);
    PHIArgMergedDebugLoc(NewCI, PN);
    return NewCI;
  }

  BinaryOperator *BinOp = cast<BinaryOperator>(FirstInst);
  BinaryOperator *NewBinOp =
    BinaryOperator::Create(BinOp->getOpcode(), LHSVal, RHSVal);

  NewBinOp->copyIRFlags(PN.getIncomingValue(0));

  for (unsigned i = 1, e = PN.getNumIncomingValues(); i != e; ++i)
    NewBinOp->andIRFlags(PN.getIncomingValue(i));

  PHIArgMergedDebugLoc(NewBinOp, PN);
  return NewBinOp;
}

Instruction *InstCombinerImpl::foldPHIArgGEPIntoPHI(PHINode &PN) {
  GetElementPtrInst *FirstInst =cast<GetElementPtrInst>(PN.getIncomingValue(0));

  SmallVector<Value*, 16> FixedOperands(FirstInst->op_begin(),
                                        FirstInst->op_end());
  // This is true if all GEP bases are allocas and if all indices into them are
  // constants.
  bool AllBasePointersAreAllocas = true;

  // We don't want to replace this phi if the replacement would require
  // more than one phi, which leads to higher register pressure. This is
  // especially bad when the PHIs are in the header of a loop.
  bool NeededPhi = false;

  bool AllInBounds = true;

  // Scan to see if all operands are the same opcode, and all have one user.
  for (unsigned i = 1; i != PN.getNumIncomingValues(); ++i) {
    GetElementPtrInst *GEP =
        dyn_cast<GetElementPtrInst>(PN.getIncomingValue(i));
    if (!GEP || !GEP->hasOneUser() || GEP->getType() != FirstInst->getType() ||
        GEP->getNumOperands() != FirstInst->getNumOperands())
      return nullptr;

    AllInBounds &= GEP->isInBounds();

    // Keep track of whether or not all GEPs are of alloca pointers.
    if (AllBasePointersAreAllocas &&
        (!isa<AllocaInst>(GEP->getOperand(0)) ||
         !GEP->hasAllConstantIndices()))
      AllBasePointersAreAllocas = false;

    // Compare the operand lists.
    for (unsigned op = 0, e = FirstInst->getNumOperands(); op != e; ++op) {
      if (FirstInst->getOperand(op) == GEP->getOperand(op))
        continue;

      // Don't merge two GEPs when two operands differ (introducing phi nodes)
      // if one of the PHIs has a constant for the index.  The index may be
      // substantially cheaper to compute for the constants, so making it a
      // variable index could pessimize the path.  This also handles the case
      // for struct indices, which must always be constant.
      if (isa<ConstantInt>(FirstInst->getOperand(op)) ||
          isa<ConstantInt>(GEP->getOperand(op)))
        return nullptr;

      if (FirstInst->getOperand(op)->getType() !=GEP->getOperand(op)->getType())
        return nullptr;

      // If we already needed a PHI for an earlier operand, and another operand
      // also requires a PHI, we'd be introducing more PHIs than we're
      // eliminating, which increases register pressure on entry to the PHI's
      // block.
      if (NeededPhi)
        return nullptr;

      FixedOperands[op] = nullptr;  // Needs a PHI.
      NeededPhi = true;
    }
  }

  // If all of the base pointers of the PHI'd GEPs are from allocas, don't
  // bother doing this transformation.  At best, this will just save a bit of
  // offset calculation, but all the predecessors will have to materialize the
  // stack address into a register anyway.  We'd actually rather *clone* the
  // load up into the predecessors so that we have a load of a gep of an alloca,
  // which can usually all be folded into the load.
  if (AllBasePointersAreAllocas)
    return nullptr;

  // Otherwise, this is safe to transform.  Insert PHI nodes for each operand
  // that is variable.
  SmallVector<PHINode*, 16> OperandPhis(FixedOperands.size());

  bool HasAnyPHIs = false;
  for (unsigned i = 0, e = FixedOperands.size(); i != e; ++i) {
    if (FixedOperands[i]) continue;  // operand doesn't need a phi.
    Value *FirstOp = FirstInst->getOperand(i);
    PHINode *NewPN = PHINode::Create(FirstOp->getType(), e,
                                     FirstOp->getName()+".pn");
    InsertNewInstBefore(NewPN, PN);

    NewPN->addIncoming(FirstOp, PN.getIncomingBlock(0));
    OperandPhis[i] = NewPN;
    FixedOperands[i] = NewPN;
    HasAnyPHIs = true;
  }


  // Add all operands to the new PHIs.
  if (HasAnyPHIs) {
    for (unsigned i = 1, e = PN.getNumIncomingValues(); i != e; ++i) {
      GetElementPtrInst *InGEP =cast<GetElementPtrInst>(PN.getIncomingValue(i));
      BasicBlock *InBB = PN.getIncomingBlock(i);

      for (unsigned op = 0, e = OperandPhis.size(); op != e; ++op)
        if (PHINode *OpPhi = OperandPhis[op])
          OpPhi->addIncoming(InGEP->getOperand(op), InBB);
    }
  }

  Value *Base = FixedOperands[0];
  GetElementPtrInst *NewGEP =
      GetElementPtrInst::Create(FirstInst->getSourceElementType(), Base,
                                makeArrayRef(FixedOperands).slice(1));
  if (AllInBounds) NewGEP->setIsInBounds();
  PHIArgMergedDebugLoc(NewGEP, PN);
  return NewGEP;
}

/// Return true if we know that it is safe to sink the load out of the block
/// that defines it. This means that it must be obvious the value of the load is
/// not changed from the point of the load to the end of the block it is in.
///
/// Finally, it is safe, but not profitable, to sink a load targeting a
/// non-address-taken alloca.  Doing so will cause us to not promote the alloca
/// to a register.
static bool isSafeAndProfitableToSinkLoad(LoadInst *L) {
  BasicBlock::iterator BBI = L->getIterator(), E = L->getParent()->end();

  for (++BBI; BBI != E; ++BBI)
    if (BBI->mayWriteToMemory()) {
      // Calls that only access inaccessible memory do not block sinking the
      // load.
      if (auto *CB = dyn_cast<CallBase>(BBI))
        if (CB->onlyAccessesInaccessibleMemory())
          continue;
      return false;
    }

  // Check for non-address taken alloca.  If not address-taken already, it isn't
  // profitable to do this xform.
  if (AllocaInst *AI = dyn_cast<AllocaInst>(L->getOperand(0))) {
    bool isAddressTaken = false;
    for (User *U : AI->users()) {
      if (isa<LoadInst>(U)) continue;
      if (StoreInst *SI = dyn_cast<StoreInst>(U)) {
        // If storing TO the alloca, then the address isn't taken.
        if (SI->getOperand(1) == AI) continue;
      }
      isAddressTaken = true;
      break;
    }

    if (!isAddressTaken && AI->isStaticAlloca())
      return false;
  }

  // If this load is a load from a GEP with a constant offset from an alloca,
  // then we don't want to sink it.  In its present form, it will be
  // load [constant stack offset].  Sinking it will cause us to have to
  // materialize the stack addresses in each predecessor in a register only to
  // do a shared load from register in the successor.
  if (GetElementPtrInst *GEP = dyn_cast<GetElementPtrInst>(L->getOperand(0)))
    if (AllocaInst *AI = dyn_cast<AllocaInst>(GEP->getOperand(0)))
      if (AI->isStaticAlloca() && GEP->hasAllConstantIndices())
        return false;

  return true;
}

Instruction *InstCombinerImpl::foldPHIArgLoadIntoPHI(PHINode &PN) {
  LoadInst *FirstLI = cast<LoadInst>(PN.getIncomingValue(0));

  // FIXME: This is overconservative; this transform is allowed in some cases
  // for atomic operations.
  if (FirstLI->isAtomic())
    return nullptr;

  // When processing loads, we need to propagate two bits of information to the
  // sunk load: whether it is volatile, and what its alignment is.  We currently
  // don't sink loads when some have their alignment specified and some don't.
  // visitLoadInst will propagate an alignment onto the load when TD is around,
  // and if TD isn't around, we can't handle the mixed case.
  bool isVolatile = FirstLI->isVolatile();
  Align LoadAlignment = FirstLI->getAlign();
  unsigned LoadAddrSpace = FirstLI->getPointerAddressSpace();

  // We can't sink the load if the loaded value could be modified between the
  // load and the PHI.
  if (FirstLI->getParent() != PN.getIncomingBlock(0) ||
      !isSafeAndProfitableToSinkLoad(FirstLI))
    return nullptr;

  // If the PHI is of volatile loads and the load block has multiple
  // successors, sinking it would remove a load of the volatile value from
  // the path through the other successor.
  if (isVolatile &&
      FirstLI->getParent()->getTerminator()->getNumSuccessors() != 1)
    return nullptr;

  // Check to see if all arguments are the same operation.
  for (unsigned i = 1, e = PN.getNumIncomingValues(); i != e; ++i) {
    LoadInst *LI = dyn_cast<LoadInst>(PN.getIncomingValue(i));
    if (!LI || !LI->hasOneUser())
      return nullptr;

    // We can't sink the load if the loaded value could be modified between
    // the load and the PHI.
    if (LI->isVolatile() != isVolatile ||
        LI->getParent() != PN.getIncomingBlock(i) ||
        LI->getPointerAddressSpace() != LoadAddrSpace ||
        !isSafeAndProfitableToSinkLoad(LI))
      return nullptr;

    LoadAlignment = std::min(LoadAlignment, Align(LI->getAlign()));

    // If the PHI is of volatile loads and the load block has multiple
    // successors, sinking it would remove a load of the volatile value from
    // the path through the other successor.
    if (isVolatile &&
        LI->getParent()->getTerminator()->getNumSuccessors() != 1)
      return nullptr;
  }

  // Okay, they are all the same operation.  Create a new PHI node of the
  // correct type, and PHI together all of the LHS's of the instructions.
  PHINode *NewPN = PHINode::Create(FirstLI->getOperand(0)->getType(),
                                   PN.getNumIncomingValues(),
                                   PN.getName()+".in");

  Value *InVal = FirstLI->getOperand(0);
  NewPN->addIncoming(InVal, PN.getIncomingBlock(0));
  LoadInst *NewLI =
      new LoadInst(FirstLI->getType(), NewPN, "", isVolatile, LoadAlignment);

  unsigned KnownIDs[] = {
    LLVMContext::MD_tbaa,
    LLVMContext::MD_range,
    LLVMContext::MD_invariant_load,
    LLVMContext::MD_alias_scope,
    LLVMContext::MD_noalias,
    LLVMContext::MD_nonnull,
    LLVMContext::MD_align,
    LLVMContext::MD_dereferenceable,
    LLVMContext::MD_dereferenceable_or_null,
    LLVMContext::MD_access_group,
  };

  for (unsigned ID : KnownIDs)
    NewLI->setMetadata(ID, FirstLI->getMetadata(ID));

  // Add all operands to the new PHI and combine TBAA metadata.
  for (unsigned i = 1, e = PN.getNumIncomingValues(); i != e; ++i) {
    LoadInst *LI = cast<LoadInst>(PN.getIncomingValue(i));
    combineMetadata(NewLI, LI, KnownIDs, true);
    Value *NewInVal = LI->getOperand(0);
    if (NewInVal != InVal)
      InVal = nullptr;
    NewPN->addIncoming(NewInVal, PN.getIncomingBlock(i));
  }

  if (InVal) {
    // The new PHI unions all of the same values together.  This is really
    // common, so we handle it intelligently here for compile-time speed.
    NewLI->setOperand(0, InVal);
    delete NewPN;
  } else {
    InsertNewInstBefore(NewPN, PN);
  }

  // If this was a volatile load that we are merging, make sure to loop through
  // and mark all the input loads as non-volatile.  If we don't do this, we will
  // insert a new volatile load and the old ones will not be deletable.
  if (isVolatile)
    for (Value *IncValue : PN.incoming_values())
      cast<LoadInst>(IncValue)->setVolatile(false);

  PHIArgMergedDebugLoc(NewLI, PN);
  return NewLI;
}

/// TODO: This function could handle other cast types, but then it might
/// require special-casing a cast from the 'i1' type. See the comment in
/// FoldPHIArgOpIntoPHI() about pessimizing illegal integer types.
Instruction *InstCombinerImpl::foldPHIArgZextsIntoPHI(PHINode &Phi) {
  // We cannot create a new instruction after the PHI if the terminator is an
  // EHPad because there is no valid insertion point.
  if (Instruction *TI = Phi.getParent()->getTerminator())
    if (TI->isEHPad())
      return nullptr;

  // Early exit for the common case of a phi with two operands. These are
  // handled elsewhere. See the comment below where we check the count of zexts
  // and constants for more details.
  unsigned NumIncomingValues = Phi.getNumIncomingValues();
  if (NumIncomingValues < 3)
    return nullptr;

  // Find the narrower type specified by the first zext.
  Type *NarrowType = nullptr;
  for (Value *V : Phi.incoming_values()) {
    if (auto *Zext = dyn_cast<ZExtInst>(V)) {
      NarrowType = Zext->getSrcTy();
      break;
    }
  }
  if (!NarrowType)
    return nullptr;

  // Walk the phi operands checking that we only have zexts or constants that
  // we can shrink for free. Store the new operands for the new phi.
  SmallVector<Value *, 4> NewIncoming;
  unsigned NumZexts = 0;
  unsigned NumConsts = 0;
  for (Value *V : Phi.incoming_values()) {
    if (auto *Zext = dyn_cast<ZExtInst>(V)) {
      // All zexts must be identical and have one user.
      if (Zext->getSrcTy() != NarrowType || !Zext->hasOneUser())
        return nullptr;
      NewIncoming.push_back(Zext->getOperand(0));
      NumZexts++;
    } else if (auto *C = dyn_cast<Constant>(V)) {
      // Make sure that constants can fit in the new type.
      Constant *Trunc = ConstantExpr::getTrunc(C, NarrowType);
      if (ConstantExpr::getZExt(Trunc, C->getType()) != C)
        return nullptr;
      NewIncoming.push_back(Trunc);
      NumConsts++;
    } else {
      // If it's not a cast or a constant, bail out.
      return nullptr;
    }
  }

  // The more common cases of a phi with no constant operands or just one
  // variable operand are handled by FoldPHIArgOpIntoPHI() and foldOpIntoPhi()
  // respectively. foldOpIntoPhi() wants to do the opposite transform that is
  // performed here. It tries to replicate a cast in the phi operand's basic
  // block to expose other folding opportunities. Thus, InstCombine will
  // infinite loop without this check.
  if (NumConsts == 0 || NumZexts < 2)
    return nullptr;

  // All incoming values are zexts or constants that are safe to truncate.
  // Create a new phi node of the narrow type, phi together all of the new
  // operands, and zext the result back to the original type.
  PHINode *NewPhi = PHINode::Create(NarrowType, NumIncomingValues,
                                    Phi.getName() + ".shrunk");
  for (unsigned i = 0; i != NumIncomingValues; ++i)
    NewPhi->addIncoming(NewIncoming[i], Phi.getIncomingBlock(i));

  InsertNewInstBefore(NewPhi, Phi);
  return CastInst::CreateZExtOrBitCast(NewPhi, Phi.getType());
}

/// If all operands to a PHI node are the same "unary" operator and they all are
/// only used by the PHI, PHI together their inputs, and do the operation once,
/// to the result of the PHI.
Instruction *InstCombinerImpl::foldPHIArgOpIntoPHI(PHINode &PN) {
  // We cannot create a new instruction after the PHI if the terminator is an
  // EHPad because there is no valid insertion point.
  if (Instruction *TI = PN.getParent()->getTerminator())
    if (TI->isEHPad())
      return nullptr;

  Instruction *FirstInst = cast<Instruction>(PN.getIncomingValue(0));

  if (isa<GetElementPtrInst>(FirstInst))
    return foldPHIArgGEPIntoPHI(PN);
  if (isa<LoadInst>(FirstInst))
    return foldPHIArgLoadIntoPHI(PN);
  if (isa<InsertValueInst>(FirstInst))
    return foldPHIArgInsertValueInstructionIntoPHI(PN);
  if (isa<ExtractValueInst>(FirstInst))
    return foldPHIArgExtractValueInstructionIntoPHI(PN);

  // Scan the instruction, looking for input operations that can be folded away.
  // If all input operands to the phi are the same instruction (e.g. a cast from
  // the same type or "+42") we can pull the operation through the PHI, reducing
  // code size and simplifying code.
  Constant *ConstantOp = nullptr;
  Type *CastSrcTy = nullptr;

  if (isa<CastInst>(FirstInst)) {
    CastSrcTy = FirstInst->getOperand(0)->getType();

    // Be careful about transforming integer PHIs.  We don't want to pessimize
    // the code by turning an i32 into an i1293.
    if (PN.getType()->isIntegerTy() && CastSrcTy->isIntegerTy()) {
      if (!shouldChangeType(PN.getType(), CastSrcTy))
        return nullptr;
    }
  } else if (isa<BinaryOperator>(FirstInst) || isa<CmpInst>(FirstInst)) {
    // Can fold binop, compare or shift here if the RHS is a constant,
    // otherwise call FoldPHIArgBinOpIntoPHI.
    ConstantOp = dyn_cast<Constant>(FirstInst->getOperand(1));
    if (!ConstantOp)
      return foldPHIArgBinOpIntoPHI(PN);
  } else {
    return nullptr;  // Cannot fold this operation.
  }

  // Check to see if all arguments are the same operation.
  for (unsigned i = 1, e = PN.getNumIncomingValues(); i != e; ++i) {
    Instruction *I = dyn_cast<Instruction>(PN.getIncomingValue(i));
    if (!I || !I->hasOneUser() || !I->isSameOperationAs(FirstInst))
      return nullptr;
    if (CastSrcTy) {
      if (I->getOperand(0)->getType() != CastSrcTy)
        return nullptr;  // Cast operation must match.
    } else if (I->getOperand(1) != ConstantOp) {
      return nullptr;
    }
  }

  // Okay, they are all the same operation.  Create a new PHI node of the
  // correct type, and PHI together all of the LHS's of the instructions.
  PHINode *NewPN = PHINode::Create(FirstInst->getOperand(0)->getType(),
                                   PN.getNumIncomingValues(),
                                   PN.getName()+".in");

  Value *InVal = FirstInst->getOperand(0);
  NewPN->addIncoming(InVal, PN.getIncomingBlock(0));

  // Add all operands to the new PHI.
  for (unsigned i = 1, e = PN.getNumIncomingValues(); i != e; ++i) {
    Value *NewInVal = cast<Instruction>(PN.getIncomingValue(i))->getOperand(0);
    if (NewInVal != InVal)
      InVal = nullptr;
    NewPN->addIncoming(NewInVal, PN.getIncomingBlock(i));
  }

  Value *PhiVal;
  if (InVal) {
    // The new PHI unions all of the same values together.  This is really
    // common, so we handle it intelligently here for compile-time speed.
    PhiVal = InVal;
    delete NewPN;
  } else {
    InsertNewInstBefore(NewPN, PN);
    PhiVal = NewPN;
  }

  // Insert and return the new operation.
  if (CastInst *FirstCI = dyn_cast<CastInst>(FirstInst)) {
    CastInst *NewCI = CastInst::Create(FirstCI->getOpcode(), PhiVal,
                                       PN.getType());
    PHIArgMergedDebugLoc(NewCI, PN);
    return NewCI;
  }

  if (BinaryOperator *BinOp = dyn_cast<BinaryOperator>(FirstInst)) {
    BinOp = BinaryOperator::Create(BinOp->getOpcode(), PhiVal, ConstantOp);
    BinOp->copyIRFlags(PN.getIncomingValue(0));

    for (unsigned i = 1, e = PN.getNumIncomingValues(); i != e; ++i)
      BinOp->andIRFlags(PN.getIncomingValue(i));

    PHIArgMergedDebugLoc(BinOp, PN);
    return BinOp;
  }

  CmpInst *CIOp = cast<CmpInst>(FirstInst);
  CmpInst *NewCI = CmpInst::Create(CIOp->getOpcode(), CIOp->getPredicate(),
                                   PhiVal, ConstantOp);
  PHIArgMergedDebugLoc(NewCI, PN);
  return NewCI;
}

/// Return true if this PHI node is only used by a PHI node cycle that is dead.
static bool DeadPHICycle(PHINode *PN,
                         SmallPtrSetImpl<PHINode*> &PotentiallyDeadPHIs) {
  if (PN->use_empty()) return true;
  if (!PN->hasOneUse()) return false;

  // Remember this node, and if we find the cycle, return.
  if (!PotentiallyDeadPHIs.insert(PN).second)
    return true;

  // Don't scan crazily complex things.
  if (PotentiallyDeadPHIs.size() == 16)
    return false;

  if (PHINode *PU = dyn_cast<PHINode>(PN->user_back()))
    return DeadPHICycle(PU, PotentiallyDeadPHIs);

  return false;
}

/// Return true if this phi node is always equal to NonPhiInVal.
/// This happens with mutually cyclic phi nodes like:
///   z = some value; x = phi (y, z); y = phi (x, z)
static bool PHIsEqualValue(PHINode *PN, Value *NonPhiInVal,
                           SmallPtrSetImpl<PHINode*> &ValueEqualPHIs) {
  // See if we already saw this PHI node.
  if (!ValueEqualPHIs.insert(PN).second)
    return true;

  // Don't scan crazily complex things.
  if (ValueEqualPHIs.size() == 16)
    return false;

  // Scan the operands to see if they are either phi nodes or are equal to
  // the value.
  for (Value *Op : PN->incoming_values()) {
    if (PHINode *OpPN = dyn_cast<PHINode>(Op)) {
      if (!PHIsEqualValue(OpPN, NonPhiInVal, ValueEqualPHIs))
        return false;
    } else if (Op != NonPhiInVal)
      return false;
  }

  return true;
}

/// Return an existing non-zero constant if this phi node has one, otherwise
/// return constant 1.
static ConstantInt *GetAnyNonZeroConstInt(PHINode &PN) {
  assert(isa<IntegerType>(PN.getType()) && "Expect only integer type phi");
  for (Value *V : PN.operands())
    if (auto *ConstVA = dyn_cast<ConstantInt>(V))
      if (!ConstVA->isZero())
        return ConstVA;
  return ConstantInt::get(cast<IntegerType>(PN.getType()), 1);
}

namespace {
struct PHIUsageRecord {
  unsigned PHIId;     // The ID # of the PHI (something determinstic to sort on)
  unsigned Shift;     // The amount shifted.
  Instruction *Inst;  // The trunc instruction.

  PHIUsageRecord(unsigned pn, unsigned Sh, Instruction *User)
    : PHIId(pn), Shift(Sh), Inst(User) {}

  bool operator<(const PHIUsageRecord &RHS) const {
    if (PHIId < RHS.PHIId) return true;
    if (PHIId > RHS.PHIId) return false;
    if (Shift < RHS.Shift) return true;
    if (Shift > RHS.Shift) return false;
    return Inst->getType()->getPrimitiveSizeInBits() <
           RHS.Inst->getType()->getPrimitiveSizeInBits();
  }
};

struct LoweredPHIRecord {
  PHINode *PN;        // The PHI that was lowered.
  unsigned Shift;     // The amount shifted.
  unsigned Width;     // The width extracted.

  LoweredPHIRecord(PHINode *pn, unsigned Sh, Type *Ty)
    : PN(pn), Shift(Sh), Width(Ty->getPrimitiveSizeInBits()) {}

  // Ctor form used by DenseMap.
  LoweredPHIRecord(PHINode *pn, unsigned Sh)
    : PN(pn), Shift(Sh), Width(0) {}
};
} // namespace

namespace llvm {
  template<>
  struct DenseMapInfo<LoweredPHIRecord> {
    static inline LoweredPHIRecord getEmptyKey() {
      return LoweredPHIRecord(nullptr, 0);
    }
    static inline LoweredPHIRecord getTombstoneKey() {
      return LoweredPHIRecord(nullptr, 1);
    }
    static unsigned getHashValue(const LoweredPHIRecord &Val) {
      return DenseMapInfo<PHINode*>::getHashValue(Val.PN) ^ (Val.Shift>>3) ^
             (Val.Width>>3);
    }
    static bool isEqual(const LoweredPHIRecord &LHS,
                        const LoweredPHIRecord &RHS) {
      return LHS.PN == RHS.PN && LHS.Shift == RHS.Shift &&
             LHS.Width == RHS.Width;
    }
  };
} // namespace llvm


/// This is an integer PHI and we know that it has an illegal type: see if it is
/// only used by trunc or trunc(lshr) operations. If so, we split the PHI into
/// the various pieces being extracted. This sort of thing is introduced when
/// SROA promotes an aggregate to large integer values.
///
/// TODO: The user of the trunc may be an bitcast to float/double/vector or an
/// inttoptr.  We should produce new PHIs in the right type.
///
Instruction *InstCombinerImpl::SliceUpIllegalIntegerPHI(PHINode &FirstPhi) {
  // PHIUsers - Keep track of all of the truncated values extracted from a set
  // of PHIs, along with their offset.  These are the things we want to rewrite.
  SmallVector<PHIUsageRecord, 16> PHIUsers;

  // PHIs are often mutually cyclic, so we keep track of a whole set of PHI
  // nodes which are extracted from. PHIsToSlice is a set we use to avoid
  // revisiting PHIs, PHIsInspected is a ordered list of PHIs that we need to
  // check the uses of (to ensure they are all extracts).
  SmallVector<PHINode*, 8> PHIsToSlice;
  SmallPtrSet<PHINode*, 8> PHIsInspected;

  PHIsToSlice.push_back(&FirstPhi);
  PHIsInspected.insert(&FirstPhi);

  for (unsigned PHIId = 0; PHIId != PHIsToSlice.size(); ++PHIId) {
    PHINode *PN = PHIsToSlice[PHIId];

    // Scan the input list of the PHI.  If any input is an invoke, and if the
    // input is defined in the predecessor, then we won't be split the critical
    // edge which is required to insert a truncate.  Because of this, we have to
    // bail out.
    for (unsigned i = 0, e = PN->getNumIncomingValues(); i != e; ++i) {
      InvokeInst *II = dyn_cast<InvokeInst>(PN->getIncomingValue(i));
      if (!II) continue;
      if (II->getParent() != PN->getIncomingBlock(i))
        continue;

      // If we have a phi, and if it's directly in the predecessor, then we have
      // a critical edge where we need to put the truncate.  Since we can't
      // split the edge in instcombine, we have to bail out.
      return nullptr;
    }

    for (User *U : PN->users()) {
      Instruction *UserI = cast<Instruction>(U);

      // If the user is a PHI, inspect its uses recursively.
      if (PHINode *UserPN = dyn_cast<PHINode>(UserI)) {
        if (PHIsInspected.insert(UserPN).second)
          PHIsToSlice.push_back(UserPN);
        continue;
      }

      // Truncates are always ok.
      if (isa<TruncInst>(UserI)) {
        PHIUsers.push_back(PHIUsageRecord(PHIId, 0, UserI));
        continue;
      }

      // Otherwise it must be a lshr which can only be used by one trunc.
      if (UserI->getOpcode() != Instruction::LShr ||
          !UserI->hasOneUse() || !isa<TruncInst>(UserI->user_back()) ||
          !isa<ConstantInt>(UserI->getOperand(1)))
        return nullptr;

      // Bail on out of range shifts.
      unsigned SizeInBits = UserI->getType()->getScalarSizeInBits();
      if (cast<ConstantInt>(UserI->getOperand(1))->getValue().uge(SizeInBits))
        return nullptr;

      unsigned Shift = cast<ConstantInt>(UserI->getOperand(1))->getZExtValue();
      PHIUsers.push_back(PHIUsageRecord(PHIId, Shift, UserI->user_back()));
    }
  }

  // If we have no users, they must be all self uses, just nuke the PHI.
  if (PHIUsers.empty())
    return replaceInstUsesWith(FirstPhi, UndefValue::get(FirstPhi.getType()));

  // If this phi node is transformable, create new PHIs for all the pieces
  // extracted out of it.  First, sort the users by their offset and size.
  array_pod_sort(PHIUsers.begin(), PHIUsers.end());

  LLVM_DEBUG(dbgs() << "SLICING UP PHI: " << FirstPhi << '\n';
             for (unsigned i = 1, e = PHIsToSlice.size(); i != e; ++i) dbgs()
             << "AND USER PHI #" << i << ": " << *PHIsToSlice[i] << '\n';);

  // PredValues - This is a temporary used when rewriting PHI nodes.  It is
  // hoisted out here to avoid construction/destruction thrashing.
  DenseMap<BasicBlock*, Value*> PredValues;

  // ExtractedVals - Each new PHI we introduce is saved here so we don't
  // introduce redundant PHIs.
  DenseMap<LoweredPHIRecord, PHINode*> ExtractedVals;

  for (unsigned UserI = 0, UserE = PHIUsers.size(); UserI != UserE; ++UserI) {
    unsigned PHIId = PHIUsers[UserI].PHIId;
    PHINode *PN = PHIsToSlice[PHIId];
    unsigned Offset = PHIUsers[UserI].Shift;
    Type *Ty = PHIUsers[UserI].Inst->getType();

    PHINode *EltPHI;

    // If we've already lowered a user like this, reuse the previously lowered
    // value.
    if ((EltPHI = ExtractedVals[LoweredPHIRecord(PN, Offset, Ty)]) == nullptr) {

      // Otherwise, Create the new PHI node for this user.
      EltPHI = PHINode::Create(Ty, PN->getNumIncomingValues(),
                               PN->getName()+".off"+Twine(Offset), PN);
      assert(EltPHI->getType() != PN->getType() &&
             "Truncate didn't shrink phi?");

      for (unsigned i = 0, e = PN->getNumIncomingValues(); i != e; ++i) {
        BasicBlock *Pred = PN->getIncomingBlock(i);
        Value *&PredVal = PredValues[Pred];

        // If we already have a value for this predecessor, reuse it.
        if (PredVal) {
          EltPHI->addIncoming(PredVal, Pred);
          continue;
        }

        // Handle the PHI self-reuse case.
        Value *InVal = PN->getIncomingValue(i);
        if (InVal == PN) {
          PredVal = EltPHI;
          EltPHI->addIncoming(PredVal, Pred);
          continue;
        }

        if (PHINode *InPHI = dyn_cast<PHINode>(PN)) {
          // If the incoming value was a PHI, and if it was one of the PHIs we
          // already rewrote it, just use the lowered value.
          if (Value *Res = ExtractedVals[LoweredPHIRecord(InPHI, Offset, Ty)]) {
            PredVal = Res;
            EltPHI->addIncoming(PredVal, Pred);
            continue;
          }
        }

        // Otherwise, do an extract in the predecessor.
        Builder.SetInsertPoint(Pred->getTerminator());
        Value *Res = InVal;
        if (Offset)
          Res = Builder.CreateLShr(Res, ConstantInt::get(InVal->getType(),
                                                          Offset), "extract");
        Res = Builder.CreateTrunc(Res, Ty, "extract.t");
        PredVal = Res;
        EltPHI->addIncoming(Res, Pred);

        // If the incoming value was a PHI, and if it was one of the PHIs we are
        // rewriting, we will ultimately delete the code we inserted.  This
        // means we need to revisit that PHI to make sure we extract out the
        // needed piece.
        if (PHINode *OldInVal = dyn_cast<PHINode>(PN->getIncomingValue(i)))
          if (PHIsInspected.count(OldInVal)) {
            unsigned RefPHIId =
                find(PHIsToSlice, OldInVal) - PHIsToSlice.begin();
            PHIUsers.push_back(PHIUsageRecord(RefPHIId, Offset,
                                              cast<Instruction>(Res)));
            ++UserE;
          }
      }
      PredValues.clear();

      LLVM_DEBUG(dbgs() << "  Made element PHI for offset " << Offset << ": "
                        << *EltPHI << '\n');
      ExtractedVals[LoweredPHIRecord(PN, Offset, Ty)] = EltPHI;
    }

    // Replace the use of this piece with the PHI node.
    replaceInstUsesWith(*PHIUsers[UserI].Inst, EltPHI);
  }

  // Replace all the remaining uses of the PHI nodes (self uses and the lshrs)
  // with undefs.
  Value *Undef = UndefValue::get(FirstPhi.getType());
  for (unsigned i = 1, e = PHIsToSlice.size(); i != e; ++i)
    replaceInstUsesWith(*PHIsToSlice[i], Undef);
  return replaceInstUsesWith(FirstPhi, Undef);
}

static Value *SimplifyUsingControlFlow(InstCombiner &Self, PHINode &PN,
                                       const DominatorTree &DT) {
  // Simplify the following patterns:
  //       if (cond)
  //       /       \
  //      ...      ...
  //       \       /
  //    phi [true] [false]
  if (!PN.getType()->isIntegerTy(1))
    return nullptr;

  if (PN.getNumOperands() != 2)
    return nullptr;

  // Make sure all inputs are constants.
  if (!all_of(PN.operands(), [](Value *V) { return isa<ConstantInt>(V); }))
    return nullptr;

  BasicBlock *BB = PN.getParent();
  // Do not bother with unreachable instructions.
  if (!DT.isReachableFromEntry(BB))
    return nullptr;

  // Same inputs.
  if (PN.getOperand(0) == PN.getOperand(1))
    return PN.getOperand(0);

  BasicBlock *TruePred = nullptr, *FalsePred = nullptr;
  for (auto *Pred : predecessors(BB)) {
    auto *Input = cast<ConstantInt>(PN.getIncomingValueForBlock(Pred));
    if (Input->isAllOnesValue())
      TruePred = Pred;
    else
      FalsePred = Pred;
  }
  assert(TruePred && FalsePred && "Must be!");

  // Check which edge of the dominator dominates the true input. If it is the
  // false edge, we should invert the condition.
  auto *IDom = DT.getNode(BB)->getIDom()->getBlock();
  auto *BI = dyn_cast<BranchInst>(IDom->getTerminator());
  if (!BI || BI->isUnconditional())
    return nullptr;

  // Check that edges outgoing from the idom's terminators dominate respective
  // inputs of the Phi.
  BasicBlockEdge TrueOutEdge(IDom, BI->getSuccessor(0));
  BasicBlockEdge FalseOutEdge(IDom, BI->getSuccessor(1));

  BasicBlockEdge TrueIncEdge(TruePred, BB);
  BasicBlockEdge FalseIncEdge(FalsePred, BB);

  auto *Cond = BI->getCondition();
  if (DT.dominates(TrueOutEdge, TrueIncEdge) &&
      DT.dominates(FalseOutEdge, FalseIncEdge))
    // This Phi is actually equivalent to branching condition of IDom.
    return Cond;
  else if (DT.dominates(TrueOutEdge, FalseIncEdge) &&
           DT.dominates(FalseOutEdge, TrueIncEdge)) {
    // This Phi is actually opposite to branching condition of IDom. We invert
    // the condition that will potentially open up some opportunities for
    // sinking.
    auto InsertPt = BB->getFirstInsertionPt();
    if (InsertPt != BB->end()) {
      Self.Builder.SetInsertPoint(&*InsertPt);
      return Self.Builder.CreateNot(Cond);
    }
  }

  return nullptr;
}

// PHINode simplification
//
Instruction *InstCombinerImpl::visitPHINode(PHINode &PN) {
  if (Value *V = SimplifyInstruction(&PN, SQ.getWithInstruction(&PN)))
    return replaceInstUsesWith(PN, V);

  if (Instruction *Result = foldPHIArgZextsIntoPHI(PN))
    return Result;

  // If all PHI operands are the same operation, pull them through the PHI,
  // reducing code size.
  if (isa<Instruction>(PN.getIncomingValue(0)) &&
      isa<Instruction>(PN.getIncomingValue(1)) &&
      cast<Instruction>(PN.getIncomingValue(0))->getOpcode() ==
          cast<Instruction>(PN.getIncomingValue(1))->getOpcode() &&
      PN.getIncomingValue(0)->hasOneUser())
    if (Instruction *Result = foldPHIArgOpIntoPHI(PN))
      return Result;

  // If this is a trivial cycle in the PHI node graph, remove it.  Basically, if
  // this PHI only has a single use (a PHI), and if that PHI only has one use (a
  // PHI)... break the cycle.
  if (PN.hasOneUse()) {
    if (Instruction *Result = foldIntegerTypedPHI(PN))
      return Result;

    Instruction *PHIUser = cast<Instruction>(PN.user_back());
    if (PHINode *PU = dyn_cast<PHINode>(PHIUser)) {
      SmallPtrSet<PHINode*, 16> PotentiallyDeadPHIs;
      PotentiallyDeadPHIs.insert(&PN);
      if (DeadPHICycle(PU, PotentiallyDeadPHIs))
        return replaceInstUsesWith(PN, UndefValue::get(PN.getType()));
    }

    // If this phi has a single use, and if that use just computes a value for
    // the next iteration of a loop, delete the phi.  This occurs with unused
    // induction variables, e.g. "for (int j = 0; ; ++j);".  Detecting this
    // common case here is good because the only other things that catch this
    // are induction variable analysis (sometimes) and ADCE, which is only run
    // late.
    if (PHIUser->hasOneUse() &&
        (isa<BinaryOperator>(PHIUser) || isa<GetElementPtrInst>(PHIUser)) &&
        PHIUser->user_back() == &PN) {
      return replaceInstUsesWith(PN, UndefValue::get(PN.getType()));
    }
    // When a PHI is used only to be compared with zero, it is safe to replace
    // an incoming value proved as known nonzero with any non-zero constant.
    // For example, in the code below, the incoming value %v can be replaced
    // with any non-zero constant based on the fact that the PHI is only used to
    // be compared with zero and %v is a known non-zero value:
    // %v = select %cond, 1, 2
    // %p = phi [%v, BB] ...
    //      icmp eq, %p, 0
    auto *CmpInst = dyn_cast<ICmpInst>(PHIUser);
    // FIXME: To be simple, handle only integer type for now.
    if (CmpInst && isa<IntegerType>(PN.getType()) && CmpInst->isEquality() &&
        match(CmpInst->getOperand(1), m_Zero())) {
      ConstantInt *NonZeroConst = nullptr;
      bool MadeChange = false;
      for (unsigned i = 0, e = PN.getNumIncomingValues(); i != e; ++i) {
        Instruction *CtxI = PN.getIncomingBlock(i)->getTerminator();
        Value *VA = PN.getIncomingValue(i);
        if (isKnownNonZero(VA, DL, 0, &AC, CtxI, &DT)) {
          if (!NonZeroConst)
            NonZeroConst = GetAnyNonZeroConstInt(PN);

          if (NonZeroConst != VA) {
            replaceOperand(PN, i, NonZeroConst);
            MadeChange = true;
          }
        }
      }
      if (MadeChange)
        return &PN;
    }
  }

  // We sometimes end up with phi cycles that non-obviously end up being the
  // same value, for example:
  //   z = some value; x = phi (y, z); y = phi (x, z)
  // where the phi nodes don't necessarily need to be in the same block.  Do a
  // quick check to see if the PHI node only contains a single non-phi value, if
  // so, scan to see if the phi cycle is actually equal to that value.
  {
    unsigned InValNo = 0, NumIncomingVals = PN.getNumIncomingValues();
    // Scan for the first non-phi operand.
    while (InValNo != NumIncomingVals &&
           isa<PHINode>(PN.getIncomingValue(InValNo)))
      ++InValNo;

    if (InValNo != NumIncomingVals) {
      Value *NonPhiInVal = PN.getIncomingValue(InValNo);

      // Scan the rest of the operands to see if there are any conflicts, if so
      // there is no need to recursively scan other phis.
      for (++InValNo; InValNo != NumIncomingVals; ++InValNo) {
        Value *OpVal = PN.getIncomingValue(InValNo);
        if (OpVal != NonPhiInVal && !isa<PHINode>(OpVal))
          break;
      }

      // If we scanned over all operands, then we have one unique value plus
      // phi values.  Scan PHI nodes to see if they all merge in each other or
      // the value.
      if (InValNo == NumIncomingVals) {
        SmallPtrSet<PHINode*, 16> ValueEqualPHIs;
        if (PHIsEqualValue(&PN, NonPhiInVal, ValueEqualPHIs))
          return replaceInstUsesWith(PN, NonPhiInVal);
      }
    }
  }

  // If there are multiple PHIs, sort their operands so that they all list
  // the blocks in the same order. This will help identical PHIs be eliminated
  // by other passes. Other passes shouldn't depend on this for correctness
  // however.
  PHINode *FirstPN = cast<PHINode>(PN.getParent()->begin());
  if (&PN != FirstPN)
    for (unsigned i = 0, e = FirstPN->getNumIncomingValues(); i != e; ++i) {
      BasicBlock *BBA = PN.getIncomingBlock(i);
      BasicBlock *BBB = FirstPN->getIncomingBlock(i);
      if (BBA != BBB) {
        Value *VA = PN.getIncomingValue(i);
        unsigned j = PN.getBasicBlockIndex(BBB);
        Value *VB = PN.getIncomingValue(j);
        PN.setIncomingBlock(i, BBB);
        PN.setIncomingValue(i, VB);
        PN.setIncomingBlock(j, BBA);
        PN.setIncomingValue(j, VA);
        // NOTE: Instcombine normally would want us to "return &PN" if we
        // modified any of the operands of an instruction.  However, since we
        // aren't adding or removing uses (just rearranging them) we don't do
        // this in this case.
      }
    }

  // Is there an identical PHI node in this basic block?
  for (PHINode &IdenticalPN : PN.getParent()->phis()) {
    // Ignore the PHI node itself.
    if (&IdenticalPN == &PN)
      continue;
    // Note that even though we've just canonicalized this PHI, due to the
    // worklist visitation order, there are no guarantess that *every* PHI
    // has been canonicalized, so we can't just compare operands ranges.
    if (!PN.isIdenticalToWhenDefined(&IdenticalPN))
      continue;
    // Just use that PHI instead then.
    ++NumPHICSEs;
    return replaceInstUsesWith(PN, &IdenticalPN);
  }

  // If this is an integer PHI and we know that it has an illegal type, see if
  // it is only used by trunc or trunc(lshr) operations.  If so, we split the
  // PHI into the various pieces being extracted.  This sort of thing is
  // introduced when SROA promotes an aggregate to a single large integer type.
  if (PN.getType()->isIntegerTy() &&
      !DL.isLegalInteger(PN.getType()->getPrimitiveSizeInBits()))
    if (Instruction *Res = SliceUpIllegalIntegerPHI(PN))
      return Res;

  // Ultimately, try to replace this Phi with a dominating condition.
  if (auto *V = SimplifyUsingControlFlow(*this, PN, DT))
    return replaceInstUsesWith(PN, V);

  return nullptr;
}