aboutsummaryrefslogtreecommitdiffstats
path: root/contrib/libs/llvm16/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp
blob: 41bc65620ff62d2014ee9feaf292eab02ad5b021 (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
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
//===- InstCombineLoadStoreAlloca.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 visit functions for load, store and alloca.
//
//===----------------------------------------------------------------------===//

#include "InstCombineInternal.h"
#include "llvm/ADT/MapVector.h"
#include "llvm/ADT/SmallString.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/Analysis/Loads.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/DebugInfoMetadata.h"
#include "llvm/IR/IntrinsicInst.h"
#include "llvm/IR/LLVMContext.h"
#include "llvm/IR/PatternMatch.h"
#include "llvm/Transforms/InstCombine/InstCombiner.h"
#include "llvm/Transforms/Utils/Local.h"
using namespace llvm;
using namespace PatternMatch;

#define DEBUG_TYPE "instcombine"

STATISTIC(NumDeadStore, "Number of dead stores eliminated");
STATISTIC(NumGlobalCopies, "Number of allocas copied from constant global");

static cl::opt<unsigned> MaxCopiedFromConstantUsers(
    "instcombine-max-copied-from-constant-users", cl::init(128),
    cl::desc("Maximum users to visit in copy from constant transform"),
    cl::Hidden);

/// isOnlyCopiedFromConstantMemory - Recursively walk the uses of a (derived)
/// pointer to an alloca.  Ignore any reads of the pointer, return false if we
/// see any stores or other unknown uses.  If we see pointer arithmetic, keep
/// track of whether it moves the pointer (with IsOffset) but otherwise traverse
/// the uses.  If we see a memcpy/memmove that targets an unoffseted pointer to
/// the alloca, and if the source pointer is a pointer to a constant memory
/// location, we can optimize this.
static bool
isOnlyCopiedFromConstantMemory(AAResults *AA, AllocaInst *V,
                               MemTransferInst *&TheCopy,
                               SmallVectorImpl<Instruction *> &ToDelete) {
  // We track lifetime intrinsics as we encounter them.  If we decide to go
  // ahead and replace the value with the memory location, this lets the caller
  // quickly eliminate the markers.

  using ValueAndIsOffset = PointerIntPair<Value *, 1, bool>;
  SmallVector<ValueAndIsOffset, 32> Worklist;
  SmallPtrSet<ValueAndIsOffset, 32> Visited;
  Worklist.emplace_back(V, false);
  while (!Worklist.empty()) {
    ValueAndIsOffset Elem = Worklist.pop_back_val();
    if (!Visited.insert(Elem).second)
      continue;
    if (Visited.size() > MaxCopiedFromConstantUsers)
      return false;

    const auto [Value, IsOffset] = Elem;
    for (auto &U : Value->uses()) {
      auto *I = cast<Instruction>(U.getUser());

      if (auto *LI = dyn_cast<LoadInst>(I)) {
        // Ignore non-volatile loads, they are always ok.
        if (!LI->isSimple()) return false;
        continue;
      }

      if (isa<PHINode, SelectInst>(I)) {
        // We set IsOffset=true, to forbid the memcpy from occurring after the
        // phi: If one of the phi operands is not based on the alloca, we
        // would incorrectly omit a write.
        Worklist.emplace_back(I, true);
        continue;
      }
      if (isa<BitCastInst, AddrSpaceCastInst>(I)) {
        // If uses of the bitcast are ok, we are ok.
        Worklist.emplace_back(I, IsOffset);
        continue;
      }
      if (auto *GEP = dyn_cast<GetElementPtrInst>(I)) {
        // If the GEP has all zero indices, it doesn't offset the pointer. If it
        // doesn't, it does.
        Worklist.emplace_back(I, IsOffset || !GEP->hasAllZeroIndices());
        continue;
      }

      if (auto *Call = dyn_cast<CallBase>(I)) {
        // If this is the function being called then we treat it like a load and
        // ignore it.
        if (Call->isCallee(&U))
          continue;

        unsigned DataOpNo = Call->getDataOperandNo(&U);
        bool IsArgOperand = Call->isArgOperand(&U);

        // Inalloca arguments are clobbered by the call.
        if (IsArgOperand && Call->isInAllocaArgument(DataOpNo))
          return false;

        // If this call site doesn't modify the memory, then we know it is just
        // a load (but one that potentially returns the value itself), so we can
        // ignore it if we know that the value isn't captured.
        bool NoCapture = Call->doesNotCapture(DataOpNo);
        if ((Call->onlyReadsMemory() && (Call->use_empty() || NoCapture)) ||
            (Call->onlyReadsMemory(DataOpNo) && NoCapture))
          continue;

        // If this is being passed as a byval argument, the caller is making a
        // copy, so it is only a read of the alloca.
        if (IsArgOperand && Call->isByValArgument(DataOpNo))
          continue;
      }

      // Lifetime intrinsics can be handled by the caller.
      if (I->isLifetimeStartOrEnd()) {
        assert(I->use_empty() && "Lifetime markers have no result to use!");
        ToDelete.push_back(I);
        continue;
      }

      // If this is isn't our memcpy/memmove, reject it as something we can't
      // handle.
      MemTransferInst *MI = dyn_cast<MemTransferInst>(I);
      if (!MI)
        return false;

      // If the transfer is volatile, reject it.
      if (MI->isVolatile())
        return false;

      // If the transfer is using the alloca as a source of the transfer, then
      // ignore it since it is a load (unless the transfer is volatile).
      if (U.getOperandNo() == 1)
        continue;

      // If we already have seen a copy, reject the second one.
      if (TheCopy) return false;

      // If the pointer has been offset from the start of the alloca, we can't
      // safely handle this.
      if (IsOffset) return false;

      // If the memintrinsic isn't using the alloca as the dest, reject it.
      if (U.getOperandNo() != 0) return false;

      // If the source of the memcpy/move is not constant, reject it.
      if (isModSet(AA->getModRefInfoMask(MI->getSource())))
        return false;

      // Otherwise, the transform is safe.  Remember the copy instruction.
      TheCopy = MI;
    }
  }
  return true;
}

/// isOnlyCopiedFromConstantMemory - Return true if the specified alloca is only
/// modified by a copy from a constant memory location. If we can prove this, we
/// can replace any uses of the alloca with uses of the memory location
/// directly.
static MemTransferInst *
isOnlyCopiedFromConstantMemory(AAResults *AA,
                               AllocaInst *AI,
                               SmallVectorImpl<Instruction *> &ToDelete) {
  MemTransferInst *TheCopy = nullptr;
  if (isOnlyCopiedFromConstantMemory(AA, AI, TheCopy, ToDelete))
    return TheCopy;
  return nullptr;
}

/// Returns true if V is dereferenceable for size of alloca.
static bool isDereferenceableForAllocaSize(const Value *V, const AllocaInst *AI,
                                           const DataLayout &DL) {
  if (AI->isArrayAllocation())
    return false;
  uint64_t AllocaSize = DL.getTypeStoreSize(AI->getAllocatedType());
  if (!AllocaSize)
    return false;
  return isDereferenceableAndAlignedPointer(V, AI->getAlign(),
                                            APInt(64, AllocaSize), DL);
}

static Instruction *simplifyAllocaArraySize(InstCombinerImpl &IC,
                                            AllocaInst &AI, DominatorTree &DT) {
  // Check for array size of 1 (scalar allocation).
  if (!AI.isArrayAllocation()) {
    // i32 1 is the canonical array size for scalar allocations.
    if (AI.getArraySize()->getType()->isIntegerTy(32))
      return nullptr;

    // Canonicalize it.
    return IC.replaceOperand(AI, 0, IC.Builder.getInt32(1));
  }

  // Convert: alloca Ty, C - where C is a constant != 1 into: alloca [C x Ty], 1
  if (const ConstantInt *C = dyn_cast<ConstantInt>(AI.getArraySize())) {
    if (C->getValue().getActiveBits() <= 64) {
      Type *NewTy = ArrayType::get(AI.getAllocatedType(), C->getZExtValue());
      AllocaInst *New = IC.Builder.CreateAlloca(NewTy, AI.getAddressSpace(),
                                                nullptr, AI.getName());
      New->setAlignment(AI.getAlign());

      replaceAllDbgUsesWith(AI, *New, *New, DT);

      // Scan to the end of the allocation instructions, to skip over a block of
      // allocas if possible...also skip interleaved debug info
      //
      BasicBlock::iterator It(New);
      while (isa<AllocaInst>(*It) || isa<DbgInfoIntrinsic>(*It))
        ++It;

      // Now that I is pointing to the first non-allocation-inst in the block,
      // insert our getelementptr instruction...
      //
      Type *IdxTy = IC.getDataLayout().getIntPtrType(AI.getType());
      Value *NullIdx = Constant::getNullValue(IdxTy);
      Value *Idx[2] = {NullIdx, NullIdx};
      Instruction *GEP = GetElementPtrInst::CreateInBounds(
          NewTy, New, Idx, New->getName() + ".sub");
      IC.InsertNewInstBefore(GEP, *It);

      // Now make everything use the getelementptr instead of the original
      // allocation.
      return IC.replaceInstUsesWith(AI, GEP);
    }
  }

  if (isa<UndefValue>(AI.getArraySize()))
    return IC.replaceInstUsesWith(AI, Constant::getNullValue(AI.getType()));

  // Ensure that the alloca array size argument has type intptr_t, so that
  // any casting is exposed early.
  Type *IntPtrTy = IC.getDataLayout().getIntPtrType(AI.getType());
  if (AI.getArraySize()->getType() != IntPtrTy) {
    Value *V = IC.Builder.CreateIntCast(AI.getArraySize(), IntPtrTy, false);
    return IC.replaceOperand(AI, 0, V);
  }

  return nullptr;
}

namespace {
// If I and V are pointers in different address space, it is not allowed to
// use replaceAllUsesWith since I and V have different types. A
// non-target-specific transformation should not use addrspacecast on V since
// the two address space may be disjoint depending on target.
//
// This class chases down uses of the old pointer until reaching the load
// instructions, then replaces the old pointer in the load instructions with
// the new pointer. If during the chasing it sees bitcast or GEP, it will
// create new bitcast or GEP with the new pointer and use them in the load
// instruction.
class PointerReplacer {
public:
  PointerReplacer(InstCombinerImpl &IC, Instruction &Root)
    : IC(IC), Root(Root) {}

  bool collectUsers();
  void replacePointer(Value *V);

private:
  bool collectUsersRecursive(Instruction &I);
  void replace(Instruction *I);
  Value *getReplacement(Value *I);
  bool isAvailable(Instruction *I) const {
    return I == &Root || Worklist.contains(I);
  }

  SmallPtrSet<Instruction *, 32> ValuesToRevisit;
  SmallSetVector<Instruction *, 4> Worklist;
  MapVector<Value *, Value *> WorkMap;
  InstCombinerImpl &IC;
  Instruction &Root;
};
} // end anonymous namespace

bool PointerReplacer::collectUsers() {
  if (!collectUsersRecursive(Root))
    return false;

  // Ensure that all outstanding (indirect) users of I
  // are inserted into the Worklist. Return false
  // otherwise.
  for (auto *Inst : ValuesToRevisit)
    if (!Worklist.contains(Inst))
      return false;
  return true;
}

bool PointerReplacer::collectUsersRecursive(Instruction &I) {
  for (auto *U : I.users()) {
    auto *Inst = cast<Instruction>(&*U);
    if (auto *Load = dyn_cast<LoadInst>(Inst)) {
      if (Load->isVolatile())
        return false;
      Worklist.insert(Load);
    } else if (auto *PHI = dyn_cast<PHINode>(Inst)) {
      // All incoming values must be instructions for replacability
      if (any_of(PHI->incoming_values(),
                 [](Value *V) { return !isa<Instruction>(V); }))
        return false;

      // If at least one incoming value of the PHI is not in Worklist,
      // store the PHI for revisiting and skip this iteration of the
      // loop.
      if (any_of(PHI->incoming_values(), [this](Value *V) {
            return !isAvailable(cast<Instruction>(V));
          })) {
        ValuesToRevisit.insert(Inst);
        continue;
      }

      Worklist.insert(PHI);
      if (!collectUsersRecursive(*PHI))
        return false;
    } else if (auto *SI = dyn_cast<SelectInst>(Inst)) {
      if (!isa<Instruction>(SI->getTrueValue()) ||
          !isa<Instruction>(SI->getFalseValue()))
        return false;

      if (!isAvailable(cast<Instruction>(SI->getTrueValue())) ||
          !isAvailable(cast<Instruction>(SI->getFalseValue()))) {
        ValuesToRevisit.insert(Inst);
        continue;
      }
      Worklist.insert(SI);
      if (!collectUsersRecursive(*SI))
        return false;
    } else if (isa<GetElementPtrInst, BitCastInst>(Inst)) {
      Worklist.insert(Inst);
      if (!collectUsersRecursive(*Inst))
        return false;
    } else if (auto *MI = dyn_cast<MemTransferInst>(Inst)) {
      if (MI->isVolatile())
        return false;
      Worklist.insert(Inst);
    } else if (Inst->isLifetimeStartOrEnd()) {
      continue;
    } else {
      LLVM_DEBUG(dbgs() << "Cannot handle pointer user: " << *U << '\n');
      return false;
    }
  }

  return true;
}

Value *PointerReplacer::getReplacement(Value *V) { return WorkMap.lookup(V); }

void PointerReplacer::replace(Instruction *I) {
  if (getReplacement(I))
    return;

  if (auto *LT = dyn_cast<LoadInst>(I)) {
    auto *V = getReplacement(LT->getPointerOperand());
    assert(V && "Operand not replaced");
    auto *NewI = new LoadInst(LT->getType(), V, "", LT->isVolatile(),
                              LT->getAlign(), LT->getOrdering(),
                              LT->getSyncScopeID());
    NewI->takeName(LT);
    copyMetadataForLoad(*NewI, *LT);

    IC.InsertNewInstWith(NewI, *LT);
    IC.replaceInstUsesWith(*LT, NewI);
    WorkMap[LT] = NewI;
  } else if (auto *PHI = dyn_cast<PHINode>(I)) {
    Type *NewTy = getReplacement(PHI->getIncomingValue(0))->getType();
    auto *NewPHI = PHINode::Create(NewTy, PHI->getNumIncomingValues(),
                                   PHI->getName(), PHI);
    for (unsigned int I = 0; I < PHI->getNumIncomingValues(); ++I)
      NewPHI->addIncoming(getReplacement(PHI->getIncomingValue(I)),
                          PHI->getIncomingBlock(I));
    WorkMap[PHI] = NewPHI;
  } else if (auto *GEP = dyn_cast<GetElementPtrInst>(I)) {
    auto *V = getReplacement(GEP->getPointerOperand());
    assert(V && "Operand not replaced");
    SmallVector<Value *, 8> Indices;
    Indices.append(GEP->idx_begin(), GEP->idx_end());
    auto *NewI =
        GetElementPtrInst::Create(GEP->getSourceElementType(), V, Indices);
    IC.InsertNewInstWith(NewI, *GEP);
    NewI->takeName(GEP);
    WorkMap[GEP] = NewI;
  } else if (auto *BC = dyn_cast<BitCastInst>(I)) {
    auto *V = getReplacement(BC->getOperand(0));
    assert(V && "Operand not replaced");
    auto *NewT = PointerType::getWithSamePointeeType(
        cast<PointerType>(BC->getType()),
        V->getType()->getPointerAddressSpace());
    auto *NewI = new BitCastInst(V, NewT);
    IC.InsertNewInstWith(NewI, *BC);
    NewI->takeName(BC);
    WorkMap[BC] = NewI;
  } else if (auto *SI = dyn_cast<SelectInst>(I)) {
    auto *NewSI = SelectInst::Create(
        SI->getCondition(), getReplacement(SI->getTrueValue()),
        getReplacement(SI->getFalseValue()), SI->getName(), nullptr, SI);
    IC.InsertNewInstWith(NewSI, *SI);
    NewSI->takeName(SI);
    WorkMap[SI] = NewSI;
  } else if (auto *MemCpy = dyn_cast<MemTransferInst>(I)) {
    auto *SrcV = getReplacement(MemCpy->getRawSource());
    // The pointer may appear in the destination of a copy, but we don't want to
    // replace it.
    if (!SrcV) {
      assert(getReplacement(MemCpy->getRawDest()) &&
             "destination not in replace list");
      return;
    }

    IC.Builder.SetInsertPoint(MemCpy);
    auto *NewI = IC.Builder.CreateMemTransferInst(
        MemCpy->getIntrinsicID(), MemCpy->getRawDest(), MemCpy->getDestAlign(),
        SrcV, MemCpy->getSourceAlign(), MemCpy->getLength(),
        MemCpy->isVolatile());
    AAMDNodes AAMD = MemCpy->getAAMetadata();
    if (AAMD)
      NewI->setAAMetadata(AAMD);

    IC.eraseInstFromFunction(*MemCpy);
    WorkMap[MemCpy] = NewI;
  } else {
    llvm_unreachable("should never reach here");
  }
}

void PointerReplacer::replacePointer(Value *V) {
#ifndef NDEBUG
  auto *PT = cast<PointerType>(Root.getType());
  auto *NT = cast<PointerType>(V->getType());
  assert(PT != NT && PT->hasSameElementTypeAs(NT) && "Invalid usage");
#endif
  WorkMap[&Root] = V;

  for (Instruction *Workitem : Worklist)
    replace(Workitem);
}

Instruction *InstCombinerImpl::visitAllocaInst(AllocaInst &AI) {
  if (auto *I = simplifyAllocaArraySize(*this, AI, DT))
    return I;

  if (AI.getAllocatedType()->isSized()) {
    // Move all alloca's of zero byte objects to the entry block and merge them
    // together.  Note that we only do this for alloca's, because malloc should
    // allocate and return a unique pointer, even for a zero byte allocation.
    if (DL.getTypeAllocSize(AI.getAllocatedType()).getKnownMinValue() == 0) {
      // For a zero sized alloca there is no point in doing an array allocation.
      // This is helpful if the array size is a complicated expression not used
      // elsewhere.
      if (AI.isArrayAllocation())
        return replaceOperand(AI, 0,
            ConstantInt::get(AI.getArraySize()->getType(), 1));

      // Get the first instruction in the entry block.
      BasicBlock &EntryBlock = AI.getParent()->getParent()->getEntryBlock();
      Instruction *FirstInst = EntryBlock.getFirstNonPHIOrDbg();
      if (FirstInst != &AI) {
        // If the entry block doesn't start with a zero-size alloca then move
        // this one to the start of the entry block.  There is no problem with
        // dominance as the array size was forced to a constant earlier already.
        AllocaInst *EntryAI = dyn_cast<AllocaInst>(FirstInst);
        if (!EntryAI || !EntryAI->getAllocatedType()->isSized() ||
            DL.getTypeAllocSize(EntryAI->getAllocatedType())
                    .getKnownMinValue() != 0) {
          AI.moveBefore(FirstInst);
          return &AI;
        }

        // Replace this zero-sized alloca with the one at the start of the entry
        // block after ensuring that the address will be aligned enough for both
        // types.
        const Align MaxAlign = std::max(EntryAI->getAlign(), AI.getAlign());
        EntryAI->setAlignment(MaxAlign);
        if (AI.getType() != EntryAI->getType())
          return new BitCastInst(EntryAI, AI.getType());
        return replaceInstUsesWith(AI, EntryAI);
      }
    }
  }

  // Check to see if this allocation is only modified by a memcpy/memmove from
  // a memory location whose alignment is equal to or exceeds that of the
  // allocation. If this is the case, we can change all users to use the
  // constant memory location instead.  This is commonly produced by the CFE by
  // constructs like "void foo() { int A[] = {1,2,3,4,5,6,7,8,9...}; }" if 'A'
  // is only subsequently read.
  SmallVector<Instruction *, 4> ToDelete;
  if (MemTransferInst *Copy = isOnlyCopiedFromConstantMemory(AA, &AI, ToDelete)) {
    Value *TheSrc = Copy->getSource();
    Align AllocaAlign = AI.getAlign();
    Align SourceAlign = getOrEnforceKnownAlignment(
      TheSrc, AllocaAlign, DL, &AI, &AC, &DT);
    if (AllocaAlign <= SourceAlign &&
        isDereferenceableForAllocaSize(TheSrc, &AI, DL) &&
        !isa<Instruction>(TheSrc)) {
      // FIXME: Can we sink instructions without violating dominance when TheSrc
      // is an instruction instead of a constant or argument?
      LLVM_DEBUG(dbgs() << "Found alloca equal to global: " << AI << '\n');
      LLVM_DEBUG(dbgs() << "  memcpy = " << *Copy << '\n');
      unsigned SrcAddrSpace = TheSrc->getType()->getPointerAddressSpace();
      auto *DestTy = PointerType::get(AI.getAllocatedType(), SrcAddrSpace);
      if (AI.getAddressSpace() == SrcAddrSpace) {
        for (Instruction *Delete : ToDelete)
          eraseInstFromFunction(*Delete);

        Value *Cast = Builder.CreateBitCast(TheSrc, DestTy);
        Instruction *NewI = replaceInstUsesWith(AI, Cast);
        eraseInstFromFunction(*Copy);
        ++NumGlobalCopies;
        return NewI;
      }

      PointerReplacer PtrReplacer(*this, AI);
      if (PtrReplacer.collectUsers()) {
        for (Instruction *Delete : ToDelete)
          eraseInstFromFunction(*Delete);

        Value *Cast = Builder.CreateBitCast(TheSrc, DestTy);
        PtrReplacer.replacePointer(Cast);
        ++NumGlobalCopies;
      }
    }
  }

  // At last, use the generic allocation site handler to aggressively remove
  // unused allocas.
  return visitAllocSite(AI);
}

// Are we allowed to form a atomic load or store of this type?
static bool isSupportedAtomicType(Type *Ty) {
  return Ty->isIntOrPtrTy() || Ty->isFloatingPointTy();
}

/// Helper to combine a load to a new type.
///
/// This just does the work of combining a load to a new type. It handles
/// metadata, etc., and returns the new instruction. The \c NewTy should be the
/// loaded *value* type. This will convert it to a pointer, cast the operand to
/// that pointer type, load it, etc.
///
/// Note that this will create all of the instructions with whatever insert
/// point the \c InstCombinerImpl currently is using.
LoadInst *InstCombinerImpl::combineLoadToNewType(LoadInst &LI, Type *NewTy,
                                                 const Twine &Suffix) {
  assert((!LI.isAtomic() || isSupportedAtomicType(NewTy)) &&
         "can't fold an atomic load to requested type");

  Value *Ptr = LI.getPointerOperand();
  unsigned AS = LI.getPointerAddressSpace();
  Type *NewPtrTy = NewTy->getPointerTo(AS);
  Value *NewPtr = nullptr;
  if (!(match(Ptr, m_BitCast(m_Value(NewPtr))) &&
        NewPtr->getType() == NewPtrTy))
    NewPtr = Builder.CreateBitCast(Ptr, NewPtrTy);

  LoadInst *NewLoad = Builder.CreateAlignedLoad(
      NewTy, NewPtr, LI.getAlign(), LI.isVolatile(), LI.getName() + Suffix);
  NewLoad->setAtomic(LI.getOrdering(), LI.getSyncScopeID());
  copyMetadataForLoad(*NewLoad, LI);
  return NewLoad;
}

/// Combine a store to a new type.
///
/// Returns the newly created store instruction.
static StoreInst *combineStoreToNewValue(InstCombinerImpl &IC, StoreInst &SI,
                                         Value *V) {
  assert((!SI.isAtomic() || isSupportedAtomicType(V->getType())) &&
         "can't fold an atomic store of requested type");

  Value *Ptr = SI.getPointerOperand();
  unsigned AS = SI.getPointerAddressSpace();
  SmallVector<std::pair<unsigned, MDNode *>, 8> MD;
  SI.getAllMetadata(MD);

  StoreInst *NewStore = IC.Builder.CreateAlignedStore(
      V, IC.Builder.CreateBitCast(Ptr, V->getType()->getPointerTo(AS)),
      SI.getAlign(), SI.isVolatile());
  NewStore->setAtomic(SI.getOrdering(), SI.getSyncScopeID());
  for (const auto &MDPair : MD) {
    unsigned ID = MDPair.first;
    MDNode *N = MDPair.second;
    // Note, essentially every kind of metadata should be preserved here! This
    // routine is supposed to clone a store instruction changing *only its
    // type*. The only metadata it makes sense to drop is metadata which is
    // invalidated when the pointer type changes. This should essentially
    // never be the case in LLVM, but we explicitly switch over only known
    // metadata to be conservatively correct. If you are adding metadata to
    // LLVM which pertains to stores, you almost certainly want to add it
    // here.
    switch (ID) {
    case LLVMContext::MD_dbg:
    case LLVMContext::MD_DIAssignID:
    case LLVMContext::MD_tbaa:
    case LLVMContext::MD_prof:
    case LLVMContext::MD_fpmath:
    case LLVMContext::MD_tbaa_struct:
    case LLVMContext::MD_alias_scope:
    case LLVMContext::MD_noalias:
    case LLVMContext::MD_nontemporal:
    case LLVMContext::MD_mem_parallel_loop_access:
    case LLVMContext::MD_access_group:
      // All of these directly apply.
      NewStore->setMetadata(ID, N);
      break;
    case LLVMContext::MD_invariant_load:
    case LLVMContext::MD_nonnull:
    case LLVMContext::MD_noundef:
    case LLVMContext::MD_range:
    case LLVMContext::MD_align:
    case LLVMContext::MD_dereferenceable:
    case LLVMContext::MD_dereferenceable_or_null:
      // These don't apply for stores.
      break;
    }
  }

  return NewStore;
}

/// Returns true if instruction represent minmax pattern like:
///   select ((cmp load V1, load V2), V1, V2).
static bool isMinMaxWithLoads(Value *V, Type *&LoadTy) {
  assert(V->getType()->isPointerTy() && "Expected pointer type.");
  // Ignore possible ty* to ixx* bitcast.
  V = InstCombiner::peekThroughBitcast(V);
  // Check that select is select ((cmp load V1, load V2), V1, V2) - minmax
  // pattern.
  CmpInst::Predicate Pred;
  Instruction *L1;
  Instruction *L2;
  Value *LHS;
  Value *RHS;
  if (!match(V, m_Select(m_Cmp(Pred, m_Instruction(L1), m_Instruction(L2)),
                         m_Value(LHS), m_Value(RHS))))
    return false;
  LoadTy = L1->getType();
  return (match(L1, m_Load(m_Specific(LHS))) &&
          match(L2, m_Load(m_Specific(RHS)))) ||
         (match(L1, m_Load(m_Specific(RHS))) &&
          match(L2, m_Load(m_Specific(LHS))));
}

/// Combine loads to match the type of their uses' value after looking
/// through intervening bitcasts.
///
/// The core idea here is that if the result of a load is used in an operation,
/// we should load the type most conducive to that operation. For example, when
/// loading an integer and converting that immediately to a pointer, we should
/// instead directly load a pointer.
///
/// However, this routine must never change the width of a load or the number of
/// loads as that would introduce a semantic change. This combine is expected to
/// be a semantic no-op which just allows loads to more closely model the types
/// of their consuming operations.
///
/// Currently, we also refuse to change the precise type used for an atomic load
/// or a volatile load. This is debatable, and might be reasonable to change
/// later. However, it is risky in case some backend or other part of LLVM is
/// relying on the exact type loaded to select appropriate atomic operations.
static Instruction *combineLoadToOperationType(InstCombinerImpl &IC,
                                               LoadInst &Load) {
  // FIXME: We could probably with some care handle both volatile and ordered
  // atomic loads here but it isn't clear that this is important.
  if (!Load.isUnordered())
    return nullptr;

  if (Load.use_empty())
    return nullptr;

  // swifterror values can't be bitcasted.
  if (Load.getPointerOperand()->isSwiftError())
    return nullptr;

  // Fold away bit casts of the loaded value by loading the desired type.
  // Note that we should not do this for pointer<->integer casts,
  // because that would result in type punning.
  if (Load.hasOneUse()) {
    // Don't transform when the type is x86_amx, it makes the pass that lower
    // x86_amx type happy.
    Type *LoadTy = Load.getType();
    if (auto *BC = dyn_cast<BitCastInst>(Load.user_back())) {
      assert(!LoadTy->isX86_AMXTy() && "Load from x86_amx* should not happen!");
      if (BC->getType()->isX86_AMXTy())
        return nullptr;
    }

    if (auto *CastUser = dyn_cast<CastInst>(Load.user_back())) {
      Type *DestTy = CastUser->getDestTy();
      if (CastUser->isNoopCast(IC.getDataLayout()) &&
          LoadTy->isPtrOrPtrVectorTy() == DestTy->isPtrOrPtrVectorTy() &&
          (!Load.isAtomic() || isSupportedAtomicType(DestTy))) {
        LoadInst *NewLoad = IC.combineLoadToNewType(Load, DestTy);
        CastUser->replaceAllUsesWith(NewLoad);
        IC.eraseInstFromFunction(*CastUser);
        return &Load;
      }
    }
  }

  // FIXME: We should also canonicalize loads of vectors when their elements are
  // cast to other types.
  return nullptr;
}

static Instruction *unpackLoadToAggregate(InstCombinerImpl &IC, LoadInst &LI) {
  // FIXME: We could probably with some care handle both volatile and atomic
  // stores here but it isn't clear that this is important.
  if (!LI.isSimple())
    return nullptr;

  Type *T = LI.getType();
  if (!T->isAggregateType())
    return nullptr;

  StringRef Name = LI.getName();

  if (auto *ST = dyn_cast<StructType>(T)) {
    // If the struct only have one element, we unpack.
    auto NumElements = ST->getNumElements();
    if (NumElements == 1) {
      LoadInst *NewLoad = IC.combineLoadToNewType(LI, ST->getTypeAtIndex(0U),
                                                  ".unpack");
      NewLoad->setAAMetadata(LI.getAAMetadata());
      return IC.replaceInstUsesWith(LI, IC.Builder.CreateInsertValue(
        PoisonValue::get(T), NewLoad, 0, Name));
    }

    // We don't want to break loads with padding here as we'd loose
    // the knowledge that padding exists for the rest of the pipeline.
    const DataLayout &DL = IC.getDataLayout();
    auto *SL = DL.getStructLayout(ST);
    if (SL->hasPadding())
      return nullptr;

    const auto Align = LI.getAlign();
    auto *Addr = LI.getPointerOperand();
    auto *IdxType = Type::getInt32Ty(T->getContext());
    auto *Zero = ConstantInt::get(IdxType, 0);

    Value *V = PoisonValue::get(T);
    for (unsigned i = 0; i < NumElements; i++) {
      Value *Indices[2] = {
        Zero,
        ConstantInt::get(IdxType, i),
      };
      auto *Ptr = IC.Builder.CreateInBoundsGEP(ST, Addr, ArrayRef(Indices),
                                               Name + ".elt");
      auto *L = IC.Builder.CreateAlignedLoad(
          ST->getElementType(i), Ptr,
          commonAlignment(Align, SL->getElementOffset(i)), Name + ".unpack");
      // Propagate AA metadata. It'll still be valid on the narrowed load.
      L->setAAMetadata(LI.getAAMetadata());
      V = IC.Builder.CreateInsertValue(V, L, i);
    }

    V->setName(Name);
    return IC.replaceInstUsesWith(LI, V);
  }

  if (auto *AT = dyn_cast<ArrayType>(T)) {
    auto *ET = AT->getElementType();
    auto NumElements = AT->getNumElements();
    if (NumElements == 1) {
      LoadInst *NewLoad = IC.combineLoadToNewType(LI, ET, ".unpack");
      NewLoad->setAAMetadata(LI.getAAMetadata());
      return IC.replaceInstUsesWith(LI, IC.Builder.CreateInsertValue(
        PoisonValue::get(T), NewLoad, 0, Name));
    }

    // Bail out if the array is too large. Ideally we would like to optimize
    // arrays of arbitrary size but this has a terrible impact on compile time.
    // The threshold here is chosen arbitrarily, maybe needs a little bit of
    // tuning.
    if (NumElements > IC.MaxArraySizeForCombine)
      return nullptr;

    const DataLayout &DL = IC.getDataLayout();
    auto EltSize = DL.getTypeAllocSize(ET);
    const auto Align = LI.getAlign();

    auto *Addr = LI.getPointerOperand();
    auto *IdxType = Type::getInt64Ty(T->getContext());
    auto *Zero = ConstantInt::get(IdxType, 0);

    Value *V = PoisonValue::get(T);
    uint64_t Offset = 0;
    for (uint64_t i = 0; i < NumElements; i++) {
      Value *Indices[2] = {
        Zero,
        ConstantInt::get(IdxType, i),
      };
      auto *Ptr = IC.Builder.CreateInBoundsGEP(AT, Addr, ArrayRef(Indices),
                                               Name + ".elt");
      auto *L = IC.Builder.CreateAlignedLoad(AT->getElementType(), Ptr,
                                             commonAlignment(Align, Offset),
                                             Name + ".unpack");
      L->setAAMetadata(LI.getAAMetadata());
      V = IC.Builder.CreateInsertValue(V, L, i);
      Offset += EltSize;
    }

    V->setName(Name);
    return IC.replaceInstUsesWith(LI, V);
  }

  return nullptr;
}

// If we can determine that all possible objects pointed to by the provided
// pointer value are, not only dereferenceable, but also definitively less than
// or equal to the provided maximum size, then return true. Otherwise, return
// false (constant global values and allocas fall into this category).
//
// FIXME: This should probably live in ValueTracking (or similar).
static bool isObjectSizeLessThanOrEq(Value *V, uint64_t MaxSize,
                                     const DataLayout &DL) {
  SmallPtrSet<Value *, 4> Visited;
  SmallVector<Value *, 4> Worklist(1, V);

  do {
    Value *P = Worklist.pop_back_val();
    P = P->stripPointerCasts();

    if (!Visited.insert(P).second)
      continue;

    if (SelectInst *SI = dyn_cast<SelectInst>(P)) {
      Worklist.push_back(SI->getTrueValue());
      Worklist.push_back(SI->getFalseValue());
      continue;
    }

    if (PHINode *PN = dyn_cast<PHINode>(P)) {
      append_range(Worklist, PN->incoming_values());
      continue;
    }

    if (GlobalAlias *GA = dyn_cast<GlobalAlias>(P)) {
      if (GA->isInterposable())
        return false;
      Worklist.push_back(GA->getAliasee());
      continue;
    }

    // If we know how big this object is, and it is less than MaxSize, continue
    // searching. Otherwise, return false.
    if (AllocaInst *AI = dyn_cast<AllocaInst>(P)) {
      if (!AI->getAllocatedType()->isSized())
        return false;

      ConstantInt *CS = dyn_cast<ConstantInt>(AI->getArraySize());
      if (!CS)
        return false;

      TypeSize TS = DL.getTypeAllocSize(AI->getAllocatedType());
      if (TS.isScalable())
        return false;
      // Make sure that, even if the multiplication below would wrap as an
      // uint64_t, we still do the right thing.
      if ((CS->getValue().zext(128) * APInt(128, TS.getFixedValue()))
              .ugt(MaxSize))
        return false;
      continue;
    }

    if (GlobalVariable *GV = dyn_cast<GlobalVariable>(P)) {
      if (!GV->hasDefinitiveInitializer() || !GV->isConstant())
        return false;

      uint64_t InitSize = DL.getTypeAllocSize(GV->getValueType());
      if (InitSize > MaxSize)
        return false;
      continue;
    }

    return false;
  } while (!Worklist.empty());

  return true;
}

// If we're indexing into an object of a known size, and the outer index is
// not a constant, but having any value but zero would lead to undefined
// behavior, replace it with zero.
//
// For example, if we have:
// @f.a = private unnamed_addr constant [1 x i32] [i32 12], align 4
// ...
// %arrayidx = getelementptr inbounds [1 x i32]* @f.a, i64 0, i64 %x
// ... = load i32* %arrayidx, align 4
// Then we know that we can replace %x in the GEP with i64 0.
//
// FIXME: We could fold any GEP index to zero that would cause UB if it were
// not zero. Currently, we only handle the first such index. Also, we could
// also search through non-zero constant indices if we kept track of the
// offsets those indices implied.
static bool canReplaceGEPIdxWithZero(InstCombinerImpl &IC,
                                     GetElementPtrInst *GEPI, Instruction *MemI,
                                     unsigned &Idx) {
  if (GEPI->getNumOperands() < 2)
    return false;

  // Find the first non-zero index of a GEP. If all indices are zero, return
  // one past the last index.
  auto FirstNZIdx = [](const GetElementPtrInst *GEPI) {
    unsigned I = 1;
    for (unsigned IE = GEPI->getNumOperands(); I != IE; ++I) {
      Value *V = GEPI->getOperand(I);
      if (const ConstantInt *CI = dyn_cast<ConstantInt>(V))
        if (CI->isZero())
          continue;

      break;
    }

    return I;
  };

  // Skip through initial 'zero' indices, and find the corresponding pointer
  // type. See if the next index is not a constant.
  Idx = FirstNZIdx(GEPI);
  if (Idx == GEPI->getNumOperands())
    return false;
  if (isa<Constant>(GEPI->getOperand(Idx)))
    return false;

  SmallVector<Value *, 4> Ops(GEPI->idx_begin(), GEPI->idx_begin() + Idx);
  Type *SourceElementType = GEPI->getSourceElementType();
  // Size information about scalable vectors is not available, so we cannot
  // deduce whether indexing at n is undefined behaviour or not. Bail out.
  if (isa<ScalableVectorType>(SourceElementType))
    return false;

  Type *AllocTy = GetElementPtrInst::getIndexedType(SourceElementType, Ops);
  if (!AllocTy || !AllocTy->isSized())
    return false;
  const DataLayout &DL = IC.getDataLayout();
  uint64_t TyAllocSize = DL.getTypeAllocSize(AllocTy).getFixedValue();

  // If there are more indices after the one we might replace with a zero, make
  // sure they're all non-negative. If any of them are negative, the overall
  // address being computed might be before the base address determined by the
  // first non-zero index.
  auto IsAllNonNegative = [&]() {
    for (unsigned i = Idx+1, e = GEPI->getNumOperands(); i != e; ++i) {
      KnownBits Known = IC.computeKnownBits(GEPI->getOperand(i), 0, MemI);
      if (Known.isNonNegative())
        continue;
      return false;
    }

    return true;
  };

  // FIXME: If the GEP is not inbounds, and there are extra indices after the
  // one we'll replace, those could cause the address computation to wrap
  // (rendering the IsAllNonNegative() check below insufficient). We can do
  // better, ignoring zero indices (and other indices we can prove small
  // enough not to wrap).
  if (Idx+1 != GEPI->getNumOperands() && !GEPI->isInBounds())
    return false;

  // Note that isObjectSizeLessThanOrEq will return true only if the pointer is
  // also known to be dereferenceable.
  return isObjectSizeLessThanOrEq(GEPI->getOperand(0), TyAllocSize, DL) &&
         IsAllNonNegative();
}

// If we're indexing into an object with a variable index for the memory
// access, but the object has only one element, we can assume that the index
// will always be zero. If we replace the GEP, return it.
template <typename T>
static Instruction *replaceGEPIdxWithZero(InstCombinerImpl &IC, Value *Ptr,
                                          T &MemI) {
  if (GetElementPtrInst *GEPI = dyn_cast<GetElementPtrInst>(Ptr)) {
    unsigned Idx;
    if (canReplaceGEPIdxWithZero(IC, GEPI, &MemI, Idx)) {
      Instruction *NewGEPI = GEPI->clone();
      NewGEPI->setOperand(Idx,
        ConstantInt::get(GEPI->getOperand(Idx)->getType(), 0));
      NewGEPI->insertBefore(GEPI);
      MemI.setOperand(MemI.getPointerOperandIndex(), NewGEPI);
      return NewGEPI;
    }
  }

  return nullptr;
}

static bool canSimplifyNullStoreOrGEP(StoreInst &SI) {
  if (NullPointerIsDefined(SI.getFunction(), SI.getPointerAddressSpace()))
    return false;

  auto *Ptr = SI.getPointerOperand();
  if (GetElementPtrInst *GEPI = dyn_cast<GetElementPtrInst>(Ptr))
    Ptr = GEPI->getOperand(0);
  return (isa<ConstantPointerNull>(Ptr) &&
          !NullPointerIsDefined(SI.getFunction(), SI.getPointerAddressSpace()));
}

static bool canSimplifyNullLoadOrGEP(LoadInst &LI, Value *Op) {
  if (GetElementPtrInst *GEPI = dyn_cast<GetElementPtrInst>(Op)) {
    const Value *GEPI0 = GEPI->getOperand(0);
    if (isa<ConstantPointerNull>(GEPI0) &&
        !NullPointerIsDefined(LI.getFunction(), GEPI->getPointerAddressSpace()))
      return true;
  }
  if (isa<UndefValue>(Op) ||
      (isa<ConstantPointerNull>(Op) &&
       !NullPointerIsDefined(LI.getFunction(), LI.getPointerAddressSpace())))
    return true;
  return false;
}

Instruction *InstCombinerImpl::visitLoadInst(LoadInst &LI) {
  Value *Op = LI.getOperand(0);

  // Try to canonicalize the loaded type.
  if (Instruction *Res = combineLoadToOperationType(*this, LI))
    return Res;

  // Attempt to improve the alignment.
  Align KnownAlign = getOrEnforceKnownAlignment(
      Op, DL.getPrefTypeAlign(LI.getType()), DL, &LI, &AC, &DT);
  if (KnownAlign > LI.getAlign())
    LI.setAlignment(KnownAlign);

  // Replace GEP indices if possible.
  if (Instruction *NewGEPI = replaceGEPIdxWithZero(*this, Op, LI)) {
      Worklist.push(NewGEPI);
      return &LI;
  }

  if (Instruction *Res = unpackLoadToAggregate(*this, LI))
    return Res;

  // Do really simple store-to-load forwarding and load CSE, to catch cases
  // where there are several consecutive memory accesses to the same location,
  // separated by a few arithmetic operations.
  bool IsLoadCSE = false;
  if (Value *AvailableVal = FindAvailableLoadedValue(&LI, *AA, &IsLoadCSE)) {
    if (IsLoadCSE)
      combineMetadataForCSE(cast<LoadInst>(AvailableVal), &LI, false);

    return replaceInstUsesWith(
        LI, Builder.CreateBitOrPointerCast(AvailableVal, LI.getType(),
                                           LI.getName() + ".cast"));
  }

  // None of the following transforms are legal for volatile/ordered atomic
  // loads.  Most of them do apply for unordered atomics.
  if (!LI.isUnordered()) return nullptr;

  // load(gep null, ...) -> unreachable
  // load null/undef -> unreachable
  // TODO: Consider a target hook for valid address spaces for this xforms.
  if (canSimplifyNullLoadOrGEP(LI, Op)) {
    // Insert a new store to null instruction before the load to indicate
    // that this code is not reachable.  We do this instead of inserting
    // an unreachable instruction directly because we cannot modify the
    // CFG.
    StoreInst *SI = new StoreInst(PoisonValue::get(LI.getType()),
                                  Constant::getNullValue(Op->getType()), &LI);
    SI->setDebugLoc(LI.getDebugLoc());
    return replaceInstUsesWith(LI, PoisonValue::get(LI.getType()));
  }

  if (Op->hasOneUse()) {
    // Change select and PHI nodes to select values instead of addresses: this
    // helps alias analysis out a lot, allows many others simplifications, and
    // exposes redundancy in the code.
    //
    // Note that we cannot do the transformation unless we know that the
    // introduced loads cannot trap!  Something like this is valid as long as
    // the condition is always false: load (select bool %C, int* null, int* %G),
    // but it would not be valid if we transformed it to load from null
    // unconditionally.
    //
    if (SelectInst *SI = dyn_cast<SelectInst>(Op)) {
      // load (select (Cond, &V1, &V2))  --> select(Cond, load &V1, load &V2).
      Align Alignment = LI.getAlign();
      if (isSafeToLoadUnconditionally(SI->getOperand(1), LI.getType(),
                                      Alignment, DL, SI) &&
          isSafeToLoadUnconditionally(SI->getOperand(2), LI.getType(),
                                      Alignment, DL, SI)) {
        LoadInst *V1 =
            Builder.CreateLoad(LI.getType(), SI->getOperand(1),
                               SI->getOperand(1)->getName() + ".val");
        LoadInst *V2 =
            Builder.CreateLoad(LI.getType(), SI->getOperand(2),
                               SI->getOperand(2)->getName() + ".val");
        assert(LI.isUnordered() && "implied by above");
        V1->setAlignment(Alignment);
        V1->setAtomic(LI.getOrdering(), LI.getSyncScopeID());
        V2->setAlignment(Alignment);
        V2->setAtomic(LI.getOrdering(), LI.getSyncScopeID());
        return SelectInst::Create(SI->getCondition(), V1, V2);
      }

      // load (select (cond, null, P)) -> load P
      if (isa<ConstantPointerNull>(SI->getOperand(1)) &&
          !NullPointerIsDefined(SI->getFunction(),
                                LI.getPointerAddressSpace()))
        return replaceOperand(LI, 0, SI->getOperand(2));

      // load (select (cond, P, null)) -> load P
      if (isa<ConstantPointerNull>(SI->getOperand(2)) &&
          !NullPointerIsDefined(SI->getFunction(),
                                LI.getPointerAddressSpace()))
        return replaceOperand(LI, 0, SI->getOperand(1));
    }
  }
  return nullptr;
}

/// Look for extractelement/insertvalue sequence that acts like a bitcast.
///
/// \returns underlying value that was "cast", or nullptr otherwise.
///
/// For example, if we have:
///
///     %E0 = extractelement <2 x double> %U, i32 0
///     %V0 = insertvalue [2 x double] undef, double %E0, 0
///     %E1 = extractelement <2 x double> %U, i32 1
///     %V1 = insertvalue [2 x double] %V0, double %E1, 1
///
/// and the layout of a <2 x double> is isomorphic to a [2 x double],
/// then %V1 can be safely approximated by a conceptual "bitcast" of %U.
/// Note that %U may contain non-undef values where %V1 has undef.
static Value *likeBitCastFromVector(InstCombinerImpl &IC, Value *V) {
  Value *U = nullptr;
  while (auto *IV = dyn_cast<InsertValueInst>(V)) {
    auto *E = dyn_cast<ExtractElementInst>(IV->getInsertedValueOperand());
    if (!E)
      return nullptr;
    auto *W = E->getVectorOperand();
    if (!U)
      U = W;
    else if (U != W)
      return nullptr;
    auto *CI = dyn_cast<ConstantInt>(E->getIndexOperand());
    if (!CI || IV->getNumIndices() != 1 || CI->getZExtValue() != *IV->idx_begin())
      return nullptr;
    V = IV->getAggregateOperand();
  }
  if (!match(V, m_Undef()) || !U)
    return nullptr;

  auto *UT = cast<VectorType>(U->getType());
  auto *VT = V->getType();
  // Check that types UT and VT are bitwise isomorphic.
  const auto &DL = IC.getDataLayout();
  if (DL.getTypeStoreSizeInBits(UT) != DL.getTypeStoreSizeInBits(VT)) {
    return nullptr;
  }
  if (auto *AT = dyn_cast<ArrayType>(VT)) {
    if (AT->getNumElements() != cast<FixedVectorType>(UT)->getNumElements())
      return nullptr;
  } else {
    auto *ST = cast<StructType>(VT);
    if (ST->getNumElements() != cast<FixedVectorType>(UT)->getNumElements())
      return nullptr;
    for (const auto *EltT : ST->elements()) {
      if (EltT != UT->getElementType())
        return nullptr;
    }
  }
  return U;
}

/// Combine stores to match the type of value being stored.
///
/// The core idea here is that the memory does not have any intrinsic type and
/// where we can we should match the type of a store to the type of value being
/// stored.
///
/// However, this routine must never change the width of a store or the number of
/// stores as that would introduce a semantic change. This combine is expected to
/// be a semantic no-op which just allows stores to more closely model the types
/// of their incoming values.
///
/// Currently, we also refuse to change the precise type used for an atomic or
/// volatile store. This is debatable, and might be reasonable to change later.
/// However, it is risky in case some backend or other part of LLVM is relying
/// on the exact type stored to select appropriate atomic operations.
///
/// \returns true if the store was successfully combined away. This indicates
/// the caller must erase the store instruction. We have to let the caller erase
/// the store instruction as otherwise there is no way to signal whether it was
/// combined or not: IC.EraseInstFromFunction returns a null pointer.
static bool combineStoreToValueType(InstCombinerImpl &IC, StoreInst &SI) {
  // FIXME: We could probably with some care handle both volatile and ordered
  // atomic stores here but it isn't clear that this is important.
  if (!SI.isUnordered())
    return false;

  // swifterror values can't be bitcasted.
  if (SI.getPointerOperand()->isSwiftError())
    return false;

  Value *V = SI.getValueOperand();

  // Fold away bit casts of the stored value by storing the original type.
  if (auto *BC = dyn_cast<BitCastInst>(V)) {
    assert(!BC->getType()->isX86_AMXTy() &&
           "store to x86_amx* should not happen!");
    V = BC->getOperand(0);
    // Don't transform when the type is x86_amx, it makes the pass that lower
    // x86_amx type happy.
    if (V->getType()->isX86_AMXTy())
      return false;
    if (!SI.isAtomic() || isSupportedAtomicType(V->getType())) {
      combineStoreToNewValue(IC, SI, V);
      return true;
    }
  }

  if (Value *U = likeBitCastFromVector(IC, V))
    if (!SI.isAtomic() || isSupportedAtomicType(U->getType())) {
      combineStoreToNewValue(IC, SI, U);
      return true;
    }

  // FIXME: We should also canonicalize stores of vectors when their elements
  // are cast to other types.
  return false;
}

static bool unpackStoreToAggregate(InstCombinerImpl &IC, StoreInst &SI) {
  // FIXME: We could probably with some care handle both volatile and atomic
  // stores here but it isn't clear that this is important.
  if (!SI.isSimple())
    return false;

  Value *V = SI.getValueOperand();
  Type *T = V->getType();

  if (!T->isAggregateType())
    return false;

  if (auto *ST = dyn_cast<StructType>(T)) {
    // If the struct only have one element, we unpack.
    unsigned Count = ST->getNumElements();
    if (Count == 1) {
      V = IC.Builder.CreateExtractValue(V, 0);
      combineStoreToNewValue(IC, SI, V);
      return true;
    }

    // We don't want to break loads with padding here as we'd loose
    // the knowledge that padding exists for the rest of the pipeline.
    const DataLayout &DL = IC.getDataLayout();
    auto *SL = DL.getStructLayout(ST);
    if (SL->hasPadding())
      return false;

    const auto Align = SI.getAlign();

    SmallString<16> EltName = V->getName();
    EltName += ".elt";
    auto *Addr = SI.getPointerOperand();
    SmallString<16> AddrName = Addr->getName();
    AddrName += ".repack";

    auto *IdxType = Type::getInt32Ty(ST->getContext());
    auto *Zero = ConstantInt::get(IdxType, 0);
    for (unsigned i = 0; i < Count; i++) {
      Value *Indices[2] = {
        Zero,
        ConstantInt::get(IdxType, i),
      };
      auto *Ptr =
          IC.Builder.CreateInBoundsGEP(ST, Addr, ArrayRef(Indices), AddrName);
      auto *Val = IC.Builder.CreateExtractValue(V, i, EltName);
      auto EltAlign = commonAlignment(Align, SL->getElementOffset(i));
      llvm::Instruction *NS = IC.Builder.CreateAlignedStore(Val, Ptr, EltAlign);
      NS->setAAMetadata(SI.getAAMetadata());
    }

    return true;
  }

  if (auto *AT = dyn_cast<ArrayType>(T)) {
    // If the array only have one element, we unpack.
    auto NumElements = AT->getNumElements();
    if (NumElements == 1) {
      V = IC.Builder.CreateExtractValue(V, 0);
      combineStoreToNewValue(IC, SI, V);
      return true;
    }

    // Bail out if the array is too large. Ideally we would like to optimize
    // arrays of arbitrary size but this has a terrible impact on compile time.
    // The threshold here is chosen arbitrarily, maybe needs a little bit of
    // tuning.
    if (NumElements > IC.MaxArraySizeForCombine)
      return false;

    const DataLayout &DL = IC.getDataLayout();
    auto EltSize = DL.getTypeAllocSize(AT->getElementType());
    const auto Align = SI.getAlign();

    SmallString<16> EltName = V->getName();
    EltName += ".elt";
    auto *Addr = SI.getPointerOperand();
    SmallString<16> AddrName = Addr->getName();
    AddrName += ".repack";

    auto *IdxType = Type::getInt64Ty(T->getContext());
    auto *Zero = ConstantInt::get(IdxType, 0);

    uint64_t Offset = 0;
    for (uint64_t i = 0; i < NumElements; i++) {
      Value *Indices[2] = {
        Zero,
        ConstantInt::get(IdxType, i),
      };
      auto *Ptr =
          IC.Builder.CreateInBoundsGEP(AT, Addr, ArrayRef(Indices), AddrName);
      auto *Val = IC.Builder.CreateExtractValue(V, i, EltName);
      auto EltAlign = commonAlignment(Align, Offset);
      Instruction *NS = IC.Builder.CreateAlignedStore(Val, Ptr, EltAlign);
      NS->setAAMetadata(SI.getAAMetadata());
      Offset += EltSize;
    }

    return true;
  }

  return false;
}

/// equivalentAddressValues - Test if A and B will obviously have the same
/// value. This includes recognizing that %t0 and %t1 will have the same
/// value in code like this:
///   %t0 = getelementptr \@a, 0, 3
///   store i32 0, i32* %t0
///   %t1 = getelementptr \@a, 0, 3
///   %t2 = load i32* %t1
///
static bool equivalentAddressValues(Value *A, Value *B) {
  // Test if the values are trivially equivalent.
  if (A == B) return true;

  // Test if the values come form identical arithmetic instructions.
  // This uses isIdenticalToWhenDefined instead of isIdenticalTo because
  // its only used to compare two uses within the same basic block, which
  // means that they'll always either have the same value or one of them
  // will have an undefined value.
  if (isa<BinaryOperator>(A) ||
      isa<CastInst>(A) ||
      isa<PHINode>(A) ||
      isa<GetElementPtrInst>(A))
    if (Instruction *BI = dyn_cast<Instruction>(B))
      if (cast<Instruction>(A)->isIdenticalToWhenDefined(BI))
        return true;

  // Otherwise they may not be equivalent.
  return false;
}

/// Converts store (bitcast (load (bitcast (select ...)))) to
/// store (load (select ...)), where select is minmax:
/// select ((cmp load V1, load V2), V1, V2).
static bool removeBitcastsFromLoadStoreOnMinMax(InstCombinerImpl &IC,
                                                StoreInst &SI) {
  // bitcast?
  if (!match(SI.getPointerOperand(), m_BitCast(m_Value())))
    return false;
  // load? integer?
  Value *LoadAddr;
  if (!match(SI.getValueOperand(), m_Load(m_BitCast(m_Value(LoadAddr)))))
    return false;
  auto *LI = cast<LoadInst>(SI.getValueOperand());
  if (!LI->getType()->isIntegerTy())
    return false;
  Type *CmpLoadTy;
  if (!isMinMaxWithLoads(LoadAddr, CmpLoadTy))
    return false;

  // Make sure the type would actually change.
  // This condition can be hit with chains of bitcasts.
  if (LI->getType() == CmpLoadTy)
    return false;

  // Make sure we're not changing the size of the load/store.
  const auto &DL = IC.getDataLayout();
  if (DL.getTypeStoreSizeInBits(LI->getType()) !=
      DL.getTypeStoreSizeInBits(CmpLoadTy))
    return false;

  if (!all_of(LI->users(), [LI, LoadAddr](User *U) {
        auto *SI = dyn_cast<StoreInst>(U);
        return SI && SI->getPointerOperand() != LI &&
               InstCombiner::peekThroughBitcast(SI->getPointerOperand()) !=
                   LoadAddr &&
               !SI->getPointerOperand()->isSwiftError();
      }))
    return false;

  IC.Builder.SetInsertPoint(LI);
  LoadInst *NewLI = IC.combineLoadToNewType(*LI, CmpLoadTy);
  // Replace all the stores with stores of the newly loaded value.
  for (auto *UI : LI->users()) {
    auto *USI = cast<StoreInst>(UI);
    IC.Builder.SetInsertPoint(USI);
    combineStoreToNewValue(IC, *USI, NewLI);
  }
  IC.replaceInstUsesWith(*LI, PoisonValue::get(LI->getType()));
  IC.eraseInstFromFunction(*LI);
  return true;
}

Instruction *InstCombinerImpl::visitStoreInst(StoreInst &SI) {
  Value *Val = SI.getOperand(0);
  Value *Ptr = SI.getOperand(1);

  // Try to canonicalize the stored type.
  if (combineStoreToValueType(*this, SI))
    return eraseInstFromFunction(SI);

  // Attempt to improve the alignment.
  const Align KnownAlign = getOrEnforceKnownAlignment(
      Ptr, DL.getPrefTypeAlign(Val->getType()), DL, &SI, &AC, &DT);
  if (KnownAlign > SI.getAlign())
    SI.setAlignment(KnownAlign);

  // Try to canonicalize the stored type.
  if (unpackStoreToAggregate(*this, SI))
    return eraseInstFromFunction(SI);

  if (removeBitcastsFromLoadStoreOnMinMax(*this, SI))
    return eraseInstFromFunction(SI);

  // Replace GEP indices if possible.
  if (Instruction *NewGEPI = replaceGEPIdxWithZero(*this, Ptr, SI)) {
      Worklist.push(NewGEPI);
      return &SI;
  }

  // Don't hack volatile/ordered stores.
  // FIXME: Some bits are legal for ordered atomic stores; needs refactoring.
  if (!SI.isUnordered()) return nullptr;

  // If the RHS is an alloca with a single use, zapify the store, making the
  // alloca dead.
  if (Ptr->hasOneUse()) {
    if (isa<AllocaInst>(Ptr))
      return eraseInstFromFunction(SI);
    if (GetElementPtrInst *GEP = dyn_cast<GetElementPtrInst>(Ptr)) {
      if (isa<AllocaInst>(GEP->getOperand(0))) {
        if (GEP->getOperand(0)->hasOneUse())
          return eraseInstFromFunction(SI);
      }
    }
  }

  // If we have a store to a location which is known constant, we can conclude
  // that the store must be storing the constant value (else the memory
  // wouldn't be constant), and this must be a noop.
  if (!isModSet(AA->getModRefInfoMask(Ptr)))
    return eraseInstFromFunction(SI);

  // Do really simple DSE, to catch cases where there are several consecutive
  // stores to the same location, separated by a few arithmetic operations. This
  // situation often occurs with bitfield accesses.
  BasicBlock::iterator BBI(SI);
  for (unsigned ScanInsts = 6; BBI != SI.getParent()->begin() && ScanInsts;
       --ScanInsts) {
    --BBI;
    // Don't count debug info directives, lest they affect codegen,
    // and we skip pointer-to-pointer bitcasts, which are NOPs.
    if (BBI->isDebugOrPseudoInst() ||
        (isa<BitCastInst>(BBI) && BBI->getType()->isPointerTy())) {
      ScanInsts++;
      continue;
    }

    if (StoreInst *PrevSI = dyn_cast<StoreInst>(BBI)) {
      // Prev store isn't volatile, and stores to the same location?
      if (PrevSI->isUnordered() &&
          equivalentAddressValues(PrevSI->getOperand(1), SI.getOperand(1)) &&
          PrevSI->getValueOperand()->getType() ==
              SI.getValueOperand()->getType()) {
        ++NumDeadStore;
        // Manually add back the original store to the worklist now, so it will
        // be processed after the operands of the removed store, as this may
        // expose additional DSE opportunities.
        Worklist.push(&SI);
        eraseInstFromFunction(*PrevSI);
        return nullptr;
      }
      break;
    }

    // If this is a load, we have to stop.  However, if the loaded value is from
    // the pointer we're loading and is producing the pointer we're storing,
    // then *this* store is dead (X = load P; store X -> P).
    if (LoadInst *LI = dyn_cast<LoadInst>(BBI)) {
      if (LI == Val && equivalentAddressValues(LI->getOperand(0), Ptr)) {
        assert(SI.isUnordered() && "can't eliminate ordering operation");
        return eraseInstFromFunction(SI);
      }

      // Otherwise, this is a load from some other location.  Stores before it
      // may not be dead.
      break;
    }

    // Don't skip over loads, throws or things that can modify memory.
    if (BBI->mayWriteToMemory() || BBI->mayReadFromMemory() || BBI->mayThrow())
      break;
  }

  // store X, null    -> turns into 'unreachable' in SimplifyCFG
  // store X, GEP(null, Y) -> turns into 'unreachable' in SimplifyCFG
  if (canSimplifyNullStoreOrGEP(SI)) {
    if (!isa<PoisonValue>(Val))
      return replaceOperand(SI, 0, PoisonValue::get(Val->getType()));
    return nullptr;  // Do not modify these!
  }

  // store undef, Ptr -> noop
  // FIXME: This is technically incorrect because it might overwrite a poison
  // value. Change to PoisonValue once #52930 is resolved.
  if (isa<UndefValue>(Val))
    return eraseInstFromFunction(SI);

  return nullptr;
}

/// Try to transform:
///   if () { *P = v1; } else { *P = v2 }
/// or:
///   *P = v1; if () { *P = v2; }
/// into a phi node with a store in the successor.
bool InstCombinerImpl::mergeStoreIntoSuccessor(StoreInst &SI) {
  if (!SI.isUnordered())
    return false; // This code has not been audited for volatile/ordered case.

  // Check if the successor block has exactly 2 incoming edges.
  BasicBlock *StoreBB = SI.getParent();
  BasicBlock *DestBB = StoreBB->getTerminator()->getSuccessor(0);
  if (!DestBB->hasNPredecessors(2))
    return false;

  // Capture the other block (the block that doesn't contain our store).
  pred_iterator PredIter = pred_begin(DestBB);
  if (*PredIter == StoreBB)
    ++PredIter;
  BasicBlock *OtherBB = *PredIter;

  // Bail out if all of the relevant blocks aren't distinct. This can happen,
  // for example, if SI is in an infinite loop.
  if (StoreBB == DestBB || OtherBB == DestBB)
    return false;

  // Verify that the other block ends in a branch and is not otherwise empty.
  BasicBlock::iterator BBI(OtherBB->getTerminator());
  BranchInst *OtherBr = dyn_cast<BranchInst>(BBI);
  if (!OtherBr || BBI == OtherBB->begin())
    return false;

  // If the other block ends in an unconditional branch, check for the 'if then
  // else' case. There is an instruction before the branch.
  StoreInst *OtherStore = nullptr;
  if (OtherBr->isUnconditional()) {
    --BBI;
    // Skip over debugging info and pseudo probes.
    while (BBI->isDebugOrPseudoInst() ||
           (isa<BitCastInst>(BBI) && BBI->getType()->isPointerTy())) {
      if (BBI==OtherBB->begin())
        return false;
      --BBI;
    }
    // If this isn't a store, isn't a store to the same location, or is not the
    // right kind of store, bail out.
    OtherStore = dyn_cast<StoreInst>(BBI);
    if (!OtherStore || OtherStore->getOperand(1) != SI.getOperand(1) ||
        !SI.isSameOperationAs(OtherStore))
      return false;
  } else {
    // Otherwise, the other block ended with a conditional branch. If one of the
    // destinations is StoreBB, then we have the if/then case.
    if (OtherBr->getSuccessor(0) != StoreBB &&
        OtherBr->getSuccessor(1) != StoreBB)
      return false;

    // Okay, we know that OtherBr now goes to Dest and StoreBB, so this is an
    // if/then triangle. See if there is a store to the same ptr as SI that
    // lives in OtherBB.
    for (;; --BBI) {
      // Check to see if we find the matching store.
      if ((OtherStore = dyn_cast<StoreInst>(BBI))) {
        if (OtherStore->getOperand(1) != SI.getOperand(1) ||
            !SI.isSameOperationAs(OtherStore))
          return false;
        break;
      }
      // If we find something that may be using or overwriting the stored
      // value, or if we run out of instructions, we can't do the transform.
      if (BBI->mayReadFromMemory() || BBI->mayThrow() ||
          BBI->mayWriteToMemory() || BBI == OtherBB->begin())
        return false;
    }

    // In order to eliminate the store in OtherBr, we have to make sure nothing
    // reads or overwrites the stored value in StoreBB.
    for (BasicBlock::iterator I = StoreBB->begin(); &*I != &SI; ++I) {
      // FIXME: This should really be AA driven.
      if (I->mayReadFromMemory() || I->mayThrow() || I->mayWriteToMemory())
        return false;
    }
  }

  // Insert a PHI node now if we need it.
  Value *MergedVal = OtherStore->getOperand(0);
  // The debug locations of the original instructions might differ. Merge them.
  DebugLoc MergedLoc = DILocation::getMergedLocation(SI.getDebugLoc(),
                                                     OtherStore->getDebugLoc());
  if (MergedVal != SI.getOperand(0)) {
    PHINode *PN = PHINode::Create(MergedVal->getType(), 2, "storemerge");
    PN->addIncoming(SI.getOperand(0), SI.getParent());
    PN->addIncoming(OtherStore->getOperand(0), OtherBB);
    MergedVal = InsertNewInstBefore(PN, DestBB->front());
    PN->setDebugLoc(MergedLoc);
  }

  // Advance to a place where it is safe to insert the new store and insert it.
  BBI = DestBB->getFirstInsertionPt();
  StoreInst *NewSI =
      new StoreInst(MergedVal, SI.getOperand(1), SI.isVolatile(), SI.getAlign(),
                    SI.getOrdering(), SI.getSyncScopeID());
  InsertNewInstBefore(NewSI, *BBI);
  NewSI->setDebugLoc(MergedLoc);
  NewSI->mergeDIAssignID({&SI, OtherStore});

  // If the two stores had AA tags, merge them.
  AAMDNodes AATags = SI.getAAMetadata();
  if (AATags)
    NewSI->setAAMetadata(AATags.merge(OtherStore->getAAMetadata()));

  // Nuke the old stores.
  eraseInstFromFunction(SI);
  eraseInstFromFunction(*OtherStore);
  return true;
}