llvm.org GIT mirror llvm / d94715e lib / Transforms / Scalar / IndVarSimplify.cpp
d94715e

Tree @d94715e (Download .tar.gz)

IndVarSimplify.cpp @d94715eraw · history · blame

   1
   2
   3
   4
   5
   6
   7
   8
   9
  10
  11
  12
  13
  14
  15
  16
  17
  18
  19
  20
  21
  22
  23
  24
  25
  26
  27
  28
  29
  30
  31
  32
  33
  34
  35
  36
  37
  38
  39
  40
  41
  42
  43
  44
  45
  46
  47
  48
  49
  50
  51
  52
  53
  54
  55
  56
  57
  58
  59
  60
  61
  62
  63
  64
  65
  66
  67
  68
  69
  70
  71
  72
  73
  74
  75
  76
  77
  78
  79
  80
  81
  82
  83
  84
  85
  86
  87
  88
  89
  90
  91
  92
  93
  94
  95
  96
  97
  98
  99
 100
 101
 102
 103
 104
 105
 106
 107
 108
 109
 110
 111
 112
 113
 114
 115
 116
 117
 118
 119
 120
 121
 122
 123
 124
 125
 126
 127
 128
 129
 130
 131
 132
 133
 134
 135
 136
 137
 138
 139
 140
 141
 142
 143
 144
 145
 146
 147
 148
 149
 150
 151
 152
 153
 154
 155
 156
 157
 158
 159
 160
 161
 162
 163
 164
 165
 166
 167
 168
 169
 170
 171
 172
 173
 174
 175
 176
 177
 178
 179
 180
 181
 182
 183
 184
 185
 186
 187
 188
 189
 190
 191
 192
 193
 194
 195
 196
 197
 198
 199
 200
 201
 202
 203
 204
 205
 206
 207
 208
 209
 210
 211
 212
 213
 214
 215
 216
 217
 218
 219
 220
 221
 222
 223
 224
 225
 226
 227
 228
 229
 230
 231
 232
 233
 234
 235
 236
 237
 238
 239
 240
 241
 242
 243
 244
 245
 246
 247
 248
 249
 250
 251
 252
 253
 254
 255
 256
 257
 258
 259
 260
 261
 262
 263
 264
 265
 266
 267
 268
 269
 270
 271
 272
 273
 274
 275
 276
 277
 278
 279
 280
 281
 282
 283
 284
 285
 286
 287
 288
 289
 290
 291
 292
 293
 294
 295
 296
 297
 298
 299
 300
 301
 302
 303
 304
 305
 306
 307
 308
 309
 310
 311
 312
 313
 314
 315
 316
 317
 318
 319
 320
 321
 322
 323
 324
 325
 326
 327
 328
 329
 330
 331
 332
 333
 334
 335
 336
 337
 338
 339
 340
 341
 342
 343
 344
 345
 346
 347
 348
 349
 350
 351
 352
 353
 354
 355
 356
 357
 358
 359
 360
 361
 362
 363
 364
 365
 366
 367
 368
 369
 370
 371
 372
 373
 374
 375
 376
 377
 378
 379
 380
 381
 382
 383
 384
 385
 386
 387
 388
 389
 390
 391
 392
 393
 394
 395
 396
 397
 398
 399
 400
 401
 402
 403
 404
 405
 406
 407
 408
 409
 410
 411
 412
 413
 414
 415
 416
 417
 418
 419
 420
 421
 422
 423
 424
 425
 426
 427
 428
 429
 430
 431
 432
 433
 434
 435
 436
 437
 438
 439
 440
 441
 442
 443
 444
 445
 446
 447
 448
 449
 450
 451
 452
 453
 454
 455
 456
 457
 458
 459
 460
 461
 462
 463
 464
 465
 466
 467
 468
 469
 470
 471
 472
 473
 474
 475
 476
 477
 478
 479
 480
 481
 482
 483
 484
 485
 486
 487
 488
 489
 490
 491
 492
 493
 494
 495
 496
 497
 498
 499
 500
 501
 502
 503
 504
 505
 506
 507
 508
 509
 510
 511
 512
 513
 514
 515
 516
 517
 518
 519
 520
 521
 522
 523
 524
 525
 526
 527
 528
 529
 530
 531
 532
 533
 534
 535
 536
 537
 538
 539
 540
 541
 542
 543
 544
 545
 546
 547
 548
 549
 550
 551
 552
 553
 554
 555
 556
 557
 558
 559
 560
 561
 562
 563
 564
 565
 566
 567
 568
 569
 570
 571
 572
 573
 574
 575
 576
 577
 578
 579
 580
 581
 582
 583
 584
 585
 586
 587
 588
 589
 590
 591
 592
 593
 594
 595
 596
 597
 598
 599
 600
 601
 602
 603
 604
 605
 606
 607
 608
 609
 610
 611
 612
 613
 614
 615
 616
 617
 618
 619
 620
 621
 622
 623
 624
 625
 626
 627
 628
 629
 630
 631
 632
 633
 634
 635
 636
 637
 638
 639
 640
 641
 642
 643
 644
 645
 646
 647
 648
 649
 650
 651
 652
 653
 654
 655
 656
 657
 658
 659
 660
 661
 662
 663
 664
 665
 666
 667
 668
 669
 670
 671
 672
 673
 674
 675
 676
 677
 678
 679
 680
 681
 682
 683
 684
 685
 686
 687
 688
 689
 690
 691
 692
 693
 694
 695
 696
 697
 698
 699
 700
 701
 702
 703
 704
 705
 706
 707
 708
 709
 710
 711
 712
 713
 714
 715
 716
 717
 718
 719
 720
 721
 722
 723
 724
 725
 726
 727
 728
 729
 730
 731
 732
 733
 734
 735
 736
 737
 738
 739
 740
 741
 742
 743
 744
 745
 746
 747
 748
 749
 750
 751
 752
 753
 754
 755
 756
 757
 758
 759
 760
 761
 762
 763
 764
 765
 766
 767
 768
 769
 770
 771
 772
 773
 774
 775
 776
 777
 778
 779
 780
 781
 782
 783
 784
 785
 786
 787
 788
 789
 790
 791
 792
 793
 794
 795
 796
 797
 798
 799
 800
 801
 802
 803
 804
 805
 806
 807
 808
 809
 810
 811
 812
 813
 814
 815
 816
 817
 818
 819
 820
 821
 822
 823
 824
 825
 826
 827
 828
 829
 830
 831
 832
 833
 834
 835
 836
 837
 838
 839
 840
 841
 842
 843
 844
 845
 846
 847
 848
 849
 850
 851
 852
 853
 854
 855
 856
 857
 858
 859
 860
 861
 862
 863
 864
 865
 866
 867
 868
 869
 870
 871
 872
 873
 874
 875
 876
 877
 878
 879
 880
 881
 882
 883
 884
 885
 886
 887
 888
 889
 890
 891
 892
 893
 894
 895
 896
 897
 898
 899
 900
 901
 902
 903
 904
 905
 906
 907
 908
 909
 910
 911
 912
 913
 914
 915
 916
 917
 918
 919
 920
 921
 922
 923
 924
 925
 926
 927
 928
 929
 930
 931
 932
 933
 934
 935
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
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
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
//===- IndVarSimplify.cpp - Induction Variable Elimination ----------------===//
//
//                     The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This transformation analyzes and transforms the induction variables (and
// computations derived from them) into simpler forms suitable for subsequent
// analysis and transformation.
//
// If the trip count of a loop is computable, this pass also makes the following
// changes:
//   1. The exit condition for the loop is canonicalized to compare the
//      induction value against the exit value.  This turns loops like:
//        'for (i = 7; i*i < 1000; ++i)' into 'for (i = 0; i != 25; ++i)'
//   2. Any use outside of the loop of an expression derived from the indvar
//      is changed to compute the derived value outside of the loop, eliminating
//      the dependence on the exit value of the induction variable.  If the only
//      purpose of the loop is to compute the exit value of some derived
//      expression, this transformation will make the loop dead.
//
//===----------------------------------------------------------------------===//

#include "llvm/Transforms/Scalar.h"
#include "llvm/ADT/DenseMap.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/LoopPass.h"
#include "llvm/Analysis/ScalarEvolutionExpander.h"
#include "llvm/IR/BasicBlock.h"
#include "llvm/IR/CFG.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/Dominators.h"
#include "llvm/IR/Instructions.h"
#include "llvm/IR/IntrinsicInst.h"
#include "llvm/IR/LLVMContext.h"
#include "llvm/IR/Type.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/raw_ostream.h"
#include "llvm/Target/TargetLibraryInfo.h"
#include "llvm/Transforms/Utils/BasicBlockUtils.h"
#include "llvm/Transforms/Utils/Local.h"
#include "llvm/Transforms/Utils/SimplifyIndVar.h"
using namespace llvm;

#define DEBUG_TYPE "indvars"

STATISTIC(NumWidened     , "Number of indvars widened");
STATISTIC(NumReplaced    , "Number of exit values replaced");
STATISTIC(NumLFTR        , "Number of loop exit tests replaced");
STATISTIC(NumElimExt     , "Number of IV sign/zero extends eliminated");
STATISTIC(NumElimIV      , "Number of congruent IVs eliminated");

// Trip count verification can be enabled by default under NDEBUG if we
// implement a strong expression equivalence checker in SCEV. Until then, we
// use the verify-indvars flag, which may assert in some cases.
static cl::opt<bool> VerifyIndvars(
  "verify-indvars", cl::Hidden,
  cl::desc("Verify the ScalarEvolution result after running indvars"));

static cl::opt<bool> ReduceLiveIVs("liv-reduce", cl::Hidden,
  cl::desc("Reduce live induction variables."));

namespace {
  class IndVarSimplify : public LoopPass {
    LoopInfo        *LI;
    ScalarEvolution *SE;
    DominatorTree   *DT;
    const DataLayout *DL;
    TargetLibraryInfo *TLI;

    SmallVector<WeakVH, 16> DeadInsts;
    bool Changed;
  public:

    static char ID; // Pass identification, replacement for typeid
    IndVarSimplify() : LoopPass(ID), LI(nullptr), SE(nullptr), DT(nullptr),
                       DL(nullptr), Changed(false) {
      initializeIndVarSimplifyPass(*PassRegistry::getPassRegistry());
    }

    bool runOnLoop(Loop *L, LPPassManager &LPM) override;

    void getAnalysisUsage(AnalysisUsage &AU) const override {
      AU.addRequired<DominatorTreeWrapperPass>();
      AU.addRequired<LoopInfo>();
      AU.addRequired<ScalarEvolution>();
      AU.addRequiredID(LoopSimplifyID);
      AU.addRequiredID(LCSSAID);
      AU.addPreserved<ScalarEvolution>();
      AU.addPreservedID(LoopSimplifyID);
      AU.addPreservedID(LCSSAID);
      AU.setPreservesCFG();
    }

  private:
    void releaseMemory() override {
      DeadInsts.clear();
    }

    bool isValidRewrite(Value *FromVal, Value *ToVal);

    void HandleFloatingPointIV(Loop *L, PHINode *PH);
    void RewriteNonIntegerIVs(Loop *L);

    void SimplifyAndExtend(Loop *L, SCEVExpander &Rewriter, LPPassManager &LPM);

    void RewriteLoopExitValues(Loop *L, SCEVExpander &Rewriter);

    Value *LinearFunctionTestReplace(Loop *L, const SCEV *BackedgeTakenCount,
                                     PHINode *IndVar, SCEVExpander &Rewriter);

    void SinkUnusedInvariants(Loop *L);
  };
}

char IndVarSimplify::ID = 0;
INITIALIZE_PASS_BEGIN(IndVarSimplify, "indvars",
                "Induction Variable Simplification", false, false)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(LoopInfo)
INITIALIZE_PASS_DEPENDENCY(ScalarEvolution)
INITIALIZE_PASS_DEPENDENCY(LoopSimplify)
INITIALIZE_PASS_DEPENDENCY(LCSSA)
INITIALIZE_PASS_END(IndVarSimplify, "indvars",
                "Induction Variable Simplification", false, false)

Pass *llvm::createIndVarSimplifyPass() {
  return new IndVarSimplify();
}

/// isValidRewrite - Return true if the SCEV expansion generated by the
/// rewriter can replace the original value. SCEV guarantees that it
/// produces the same value, but the way it is produced may be illegal IR.
/// Ideally, this function will only be called for verification.
bool IndVarSimplify::isValidRewrite(Value *FromVal, Value *ToVal) {
  // If an SCEV expression subsumed multiple pointers, its expansion could
  // reassociate the GEP changing the base pointer. This is illegal because the
  // final address produced by a GEP chain must be inbounds relative to its
  // underlying object. Otherwise basic alias analysis, among other things,
  // could fail in a dangerous way. Ultimately, SCEV will be improved to avoid
  // producing an expression involving multiple pointers. Until then, we must
  // bail out here.
  //
  // Retrieve the pointer operand of the GEP. Don't use GetUnderlyingObject
  // because it understands lcssa phis while SCEV does not.
  Value *FromPtr = FromVal;
  Value *ToPtr = ToVal;
  if (GEPOperator *GEP = dyn_cast<GEPOperator>(FromVal)) {
    FromPtr = GEP->getPointerOperand();
  }
  if (GEPOperator *GEP = dyn_cast<GEPOperator>(ToVal)) {
    ToPtr = GEP->getPointerOperand();
  }
  if (FromPtr != FromVal || ToPtr != ToVal) {
    // Quickly check the common case
    if (FromPtr == ToPtr)
      return true;

    // SCEV may have rewritten an expression that produces the GEP's pointer
    // operand. That's ok as long as the pointer operand has the same base
    // pointer. Unlike GetUnderlyingObject(), getPointerBase() will find the
    // base of a recurrence. This handles the case in which SCEV expansion
    // converts a pointer type recurrence into a nonrecurrent pointer base
    // indexed by an integer recurrence.

    // If the GEP base pointer is a vector of pointers, abort.
    if (!FromPtr->getType()->isPointerTy() || !ToPtr->getType()->isPointerTy())
      return false;

    const SCEV *FromBase = SE->getPointerBase(SE->getSCEV(FromPtr));
    const SCEV *ToBase = SE->getPointerBase(SE->getSCEV(ToPtr));
    if (FromBase == ToBase)
      return true;

    DEBUG(dbgs() << "INDVARS: GEP rewrite bail out "
          << *FromBase << " != " << *ToBase << "\n");

    return false;
  }
  return true;
}

/// Determine the insertion point for this user. By default, insert immediately
/// before the user. SCEVExpander or LICM will hoist loop invariants out of the
/// loop. For PHI nodes, there may be multiple uses, so compute the nearest
/// common dominator for the incoming blocks.
static Instruction *getInsertPointForUses(Instruction *User, Value *Def,
                                          DominatorTree *DT) {
  PHINode *PHI = dyn_cast<PHINode>(User);
  if (!PHI)
    return User;

  Instruction *InsertPt = nullptr;
  for (unsigned i = 0, e = PHI->getNumIncomingValues(); i != e; ++i) {
    if (PHI->getIncomingValue(i) != Def)
      continue;

    BasicBlock *InsertBB = PHI->getIncomingBlock(i);
    if (!InsertPt) {
      InsertPt = InsertBB->getTerminator();
      continue;
    }
    InsertBB = DT->findNearestCommonDominator(InsertPt->getParent(), InsertBB);
    InsertPt = InsertBB->getTerminator();
  }
  assert(InsertPt && "Missing phi operand");
  assert((!isa<Instruction>(Def) ||
          DT->dominates(cast<Instruction>(Def), InsertPt)) &&
         "def does not dominate all uses");
  return InsertPt;
}

//===----------------------------------------------------------------------===//
// RewriteNonIntegerIVs and helpers. Prefer integer IVs.
//===----------------------------------------------------------------------===//

/// ConvertToSInt - Convert APF to an integer, if possible.
static bool ConvertToSInt(const APFloat &APF, int64_t &IntVal) {
  bool isExact = false;
  // See if we can convert this to an int64_t
  uint64_t UIntVal;
  if (APF.convertToInteger(&UIntVal, 64, true, APFloat::rmTowardZero,
                           &isExact) != APFloat::opOK || !isExact)
    return false;
  IntVal = UIntVal;
  return true;
}

/// HandleFloatingPointIV - If the loop has floating induction variable
/// then insert corresponding integer induction variable if possible.
/// For example,
/// for(double i = 0; i < 10000; ++i)
///   bar(i)
/// is converted into
/// for(int i = 0; i < 10000; ++i)
///   bar((double)i);
///
void IndVarSimplify::HandleFloatingPointIV(Loop *L, PHINode *PN) {
  unsigned IncomingEdge = L->contains(PN->getIncomingBlock(0));
  unsigned BackEdge     = IncomingEdge^1;

  // Check incoming value.
  ConstantFP *InitValueVal =
    dyn_cast<ConstantFP>(PN->getIncomingValue(IncomingEdge));

  int64_t InitValue;
  if (!InitValueVal || !ConvertToSInt(InitValueVal->getValueAPF(), InitValue))
    return;

  // Check IV increment. Reject this PN if increment operation is not
  // an add or increment value can not be represented by an integer.
  BinaryOperator *Incr =
    dyn_cast<BinaryOperator>(PN->getIncomingValue(BackEdge));
  if (Incr == nullptr || Incr->getOpcode() != Instruction::FAdd) return;

  // If this is not an add of the PHI with a constantfp, or if the constant fp
  // is not an integer, bail out.
  ConstantFP *IncValueVal = dyn_cast<ConstantFP>(Incr->getOperand(1));
  int64_t IncValue;
  if (IncValueVal == nullptr || Incr->getOperand(0) != PN ||
      !ConvertToSInt(IncValueVal->getValueAPF(), IncValue))
    return;

  // Check Incr uses. One user is PN and the other user is an exit condition
  // used by the conditional terminator.
  Value::user_iterator IncrUse = Incr->user_begin();
  Instruction *U1 = cast<Instruction>(*IncrUse++);
  if (IncrUse == Incr->user_end()) return;
  Instruction *U2 = cast<Instruction>(*IncrUse++);
  if (IncrUse != Incr->user_end()) return;

  // Find exit condition, which is an fcmp.  If it doesn't exist, or if it isn't
  // only used by a branch, we can't transform it.
  FCmpInst *Compare = dyn_cast<FCmpInst>(U1);
  if (!Compare)
    Compare = dyn_cast<FCmpInst>(U2);
  if (!Compare || !Compare->hasOneUse() ||
      !isa<BranchInst>(Compare->user_back()))
    return;

  BranchInst *TheBr = cast<BranchInst>(Compare->user_back());

  // We need to verify that the branch actually controls the iteration count
  // of the loop.  If not, the new IV can overflow and no one will notice.
  // The branch block must be in the loop and one of the successors must be out
  // of the loop.
  assert(TheBr->isConditional() && "Can't use fcmp if not conditional");
  if (!L->contains(TheBr->getParent()) ||
      (L->contains(TheBr->getSuccessor(0)) &&
       L->contains(TheBr->getSuccessor(1))))
    return;


  // If it isn't a comparison with an integer-as-fp (the exit value), we can't
  // transform it.
  ConstantFP *ExitValueVal = dyn_cast<ConstantFP>(Compare->getOperand(1));
  int64_t ExitValue;
  if (ExitValueVal == nullptr ||
      !ConvertToSInt(ExitValueVal->getValueAPF(), ExitValue))
    return;

  // Find new predicate for integer comparison.
  CmpInst::Predicate NewPred = CmpInst::BAD_ICMP_PREDICATE;
  switch (Compare->getPredicate()) {
  default: return;  // Unknown comparison.
  case CmpInst::FCMP_OEQ:
  case CmpInst::FCMP_UEQ: NewPred = CmpInst::ICMP_EQ; break;
  case CmpInst::FCMP_ONE:
  case CmpInst::FCMP_UNE: NewPred = CmpInst::ICMP_NE; break;
  case CmpInst::FCMP_OGT:
  case CmpInst::FCMP_UGT: NewPred = CmpInst::ICMP_SGT; break;
  case CmpInst::FCMP_OGE:
  case CmpInst::FCMP_UGE: NewPred = CmpInst::ICMP_SGE; break;
  case CmpInst::FCMP_OLT:
  case CmpInst::FCMP_ULT: NewPred = CmpInst::ICMP_SLT; break;
  case CmpInst::FCMP_OLE:
  case CmpInst::FCMP_ULE: NewPred = CmpInst::ICMP_SLE; break;
  }

  // We convert the floating point induction variable to a signed i32 value if
  // we can.  This is only safe if the comparison will not overflow in a way
  // that won't be trapped by the integer equivalent operations.  Check for this
  // now.
  // TODO: We could use i64 if it is native and the range requires it.

  // The start/stride/exit values must all fit in signed i32.
  if (!isInt<32>(InitValue) || !isInt<32>(IncValue) || !isInt<32>(ExitValue))
    return;

  // If not actually striding (add x, 0.0), avoid touching the code.
  if (IncValue == 0)
    return;

  // Positive and negative strides have different safety conditions.
  if (IncValue > 0) {
    // If we have a positive stride, we require the init to be less than the
    // exit value.
    if (InitValue >= ExitValue)
      return;

    uint32_t Range = uint32_t(ExitValue-InitValue);
    // Check for infinite loop, either:
    // while (i <= Exit) or until (i > Exit)
    if (NewPred == CmpInst::ICMP_SLE || NewPred == CmpInst::ICMP_SGT) {
      if (++Range == 0) return;  // Range overflows.
    }

    unsigned Leftover = Range % uint32_t(IncValue);

    // If this is an equality comparison, we require that the strided value
    // exactly land on the exit value, otherwise the IV condition will wrap
    // around and do things the fp IV wouldn't.
    if ((NewPred == CmpInst::ICMP_EQ || NewPred == CmpInst::ICMP_NE) &&
        Leftover != 0)
      return;

    // If the stride would wrap around the i32 before exiting, we can't
    // transform the IV.
    if (Leftover != 0 && int32_t(ExitValue+IncValue) < ExitValue)
      return;

  } else {
    // If we have a negative stride, we require the init to be greater than the
    // exit value.
    if (InitValue <= ExitValue)
      return;

    uint32_t Range = uint32_t(InitValue-ExitValue);
    // Check for infinite loop, either:
    // while (i >= Exit) or until (i < Exit)
    if (NewPred == CmpInst::ICMP_SGE || NewPred == CmpInst::ICMP_SLT) {
      if (++Range == 0) return;  // Range overflows.
    }

    unsigned Leftover = Range % uint32_t(-IncValue);

    // If this is an equality comparison, we require that the strided value
    // exactly land on the exit value, otherwise the IV condition will wrap
    // around and do things the fp IV wouldn't.
    if ((NewPred == CmpInst::ICMP_EQ || NewPred == CmpInst::ICMP_NE) &&
        Leftover != 0)
      return;

    // If the stride would wrap around the i32 before exiting, we can't
    // transform the IV.
    if (Leftover != 0 && int32_t(ExitValue+IncValue) > ExitValue)
      return;
  }

  IntegerType *Int32Ty = Type::getInt32Ty(PN->getContext());

  // Insert new integer induction variable.
  PHINode *NewPHI = PHINode::Create(Int32Ty, 2, PN->getName()+".int", PN);
  NewPHI->addIncoming(ConstantInt::get(Int32Ty, InitValue),
                      PN->getIncomingBlock(IncomingEdge));

  Value *NewAdd =
    BinaryOperator::CreateAdd(NewPHI, ConstantInt::get(Int32Ty, IncValue),
                              Incr->getName()+".int", Incr);
  NewPHI->addIncoming(NewAdd, PN->getIncomingBlock(BackEdge));

  ICmpInst *NewCompare = new ICmpInst(TheBr, NewPred, NewAdd,
                                      ConstantInt::get(Int32Ty, ExitValue),
                                      Compare->getName());

  // In the following deletions, PN may become dead and may be deleted.
  // Use a WeakVH to observe whether this happens.
  WeakVH WeakPH = PN;

  // Delete the old floating point exit comparison.  The branch starts using the
  // new comparison.
  NewCompare->takeName(Compare);
  Compare->replaceAllUsesWith(NewCompare);
  RecursivelyDeleteTriviallyDeadInstructions(Compare, TLI);

  // Delete the old floating point increment.
  Incr->replaceAllUsesWith(UndefValue::get(Incr->getType()));
  RecursivelyDeleteTriviallyDeadInstructions(Incr, TLI);

  // If the FP induction variable still has uses, this is because something else
  // in the loop uses its value.  In order to canonicalize the induction
  // variable, we chose to eliminate the IV and rewrite it in terms of an
  // int->fp cast.
  //
  // We give preference to sitofp over uitofp because it is faster on most
  // platforms.
  if (WeakPH) {
    Value *Conv = new SIToFPInst(NewPHI, PN->getType(), "indvar.conv",
                                 PN->getParent()->getFirstInsertionPt());
    PN->replaceAllUsesWith(Conv);
    RecursivelyDeleteTriviallyDeadInstructions(PN, TLI);
  }
  Changed = true;
}

void IndVarSimplify::RewriteNonIntegerIVs(Loop *L) {
  // First step.  Check to see if there are any floating-point recurrences.
  // If there are, change them into integer recurrences, permitting analysis by
  // the SCEV routines.
  //
  BasicBlock *Header = L->getHeader();

  SmallVector<WeakVH, 8> PHIs;
  for (BasicBlock::iterator I = Header->begin();
       PHINode *PN = dyn_cast<PHINode>(I); ++I)
    PHIs.push_back(PN);

  for (unsigned i = 0, e = PHIs.size(); i != e; ++i)
    if (PHINode *PN = dyn_cast_or_null<PHINode>(&*PHIs[i]))
      HandleFloatingPointIV(L, PN);

  // If the loop previously had floating-point IV, ScalarEvolution
  // may not have been able to compute a trip count. Now that we've done some
  // re-writing, the trip count may be computable.
  if (Changed)
    SE->forgetLoop(L);
}

//===----------------------------------------------------------------------===//
// RewriteLoopExitValues - Optimize IV users outside the loop.
// As a side effect, reduces the amount of IV processing within the loop.
//===----------------------------------------------------------------------===//

/// RewriteLoopExitValues - Check to see if this loop has a computable
/// loop-invariant execution count.  If so, this means that we can compute the
/// final value of any expressions that are recurrent in the loop, and
/// substitute the exit values from the loop into any instructions outside of
/// the loop that use the final values of the current expressions.
///
/// This is mostly redundant with the regular IndVarSimplify activities that
/// happen later, except that it's more powerful in some cases, because it's
/// able to brute-force evaluate arbitrary instructions as long as they have
/// constant operands at the beginning of the loop.
void IndVarSimplify::RewriteLoopExitValues(Loop *L, SCEVExpander &Rewriter) {
  // Verify the input to the pass in already in LCSSA form.
  assert(L->isLCSSAForm(*DT));

  SmallVector<BasicBlock*, 8> ExitBlocks;
  L->getUniqueExitBlocks(ExitBlocks);

  // Find all values that are computed inside the loop, but used outside of it.
  // Because of LCSSA, these values will only occur in LCSSA PHI Nodes.  Scan
  // the exit blocks of the loop to find them.
  for (unsigned i = 0, e = ExitBlocks.size(); i != e; ++i) {
    BasicBlock *ExitBB = ExitBlocks[i];

    // If there are no PHI nodes in this exit block, then no values defined
    // inside the loop are used on this path, skip it.
    PHINode *PN = dyn_cast<PHINode>(ExitBB->begin());
    if (!PN) continue;

    unsigned NumPreds = PN->getNumIncomingValues();

    // We would like to be able to RAUW single-incoming value PHI nodes. We
    // have to be certain this is safe even when this is an LCSSA PHI node.
    // While the computed exit value is no longer varying in *this* loop, the
    // exit block may be an exit block for an outer containing loop as well,
    // the exit value may be varying in the outer loop, and thus it may still
    // require an LCSSA PHI node. The safe case is when this is
    // single-predecessor PHI node (LCSSA) and the exit block containing it is
    // part of the enclosing loop, or this is the outer most loop of the nest.
    // In either case the exit value could (at most) be varying in the same
    // loop body as the phi node itself. Thus if it is in turn used outside of
    // an enclosing loop it will only be via a separate LCSSA node.
    bool LCSSASafePhiForRAUW =
        NumPreds == 1 &&
        (!L->getParentLoop() || L->getParentLoop() == LI->getLoopFor(ExitBB));

    // Iterate over all of the PHI nodes.
    BasicBlock::iterator BBI = ExitBB->begin();
    while ((PN = dyn_cast<PHINode>(BBI++))) {
      if (PN->use_empty())
        continue; // dead use, don't replace it

      // SCEV only supports integer expressions for now.
      if (!PN->getType()->isIntegerTy() && !PN->getType()->isPointerTy())
        continue;

      // It's necessary to tell ScalarEvolution about this explicitly so that
      // it can walk the def-use list and forget all SCEVs, as it may not be
      // watching the PHI itself. Once the new exit value is in place, there
      // may not be a def-use connection between the loop and every instruction
      // which got a SCEVAddRecExpr for that loop.
      SE->forgetValue(PN);

      // Iterate over all of the values in all the PHI nodes.
      for (unsigned i = 0; i != NumPreds; ++i) {
        // If the value being merged in is not integer or is not defined
        // in the loop, skip it.
        Value *InVal = PN->getIncomingValue(i);
        if (!isa<Instruction>(InVal))
          continue;

        // If this pred is for a subloop, not L itself, skip it.
        if (LI->getLoopFor(PN->getIncomingBlock(i)) != L)
          continue; // The Block is in a subloop, skip it.

        // Check that InVal is defined in the loop.
        Instruction *Inst = cast<Instruction>(InVal);
        if (!L->contains(Inst))
          continue;

        // Okay, this instruction has a user outside of the current loop
        // and varies predictably *inside* the loop.  Evaluate the value it
        // contains when the loop exits, if possible.
        const SCEV *ExitValue = SE->getSCEVAtScope(Inst, L->getParentLoop());
        if (!SE->isLoopInvariant(ExitValue, L) ||
            !isSafeToExpand(ExitValue, *SE))
          continue;

        // Computing the value outside of the loop brings no benefit if :
        //  - it is definitely used inside the loop in a way which can not be
        //    optimized away.
        //  - no use outside of the loop can take advantage of hoisting the
        //    computation out of the loop
        if (ExitValue->getSCEVType()>=scMulExpr) {
          unsigned NumHardInternalUses = 0;
          unsigned NumSoftExternalUses = 0;
          unsigned NumUses = 0;
          for (auto IB = Inst->user_begin(), IE = Inst->user_end();
               IB != IE && NumUses <= 6; ++IB) {
            Instruction *UseInstr = cast<Instruction>(*IB);
            unsigned Opc = UseInstr->getOpcode();
            NumUses++;
            if (L->contains(UseInstr)) {
              if (Opc == Instruction::Call || Opc == Instruction::Ret)
                NumHardInternalUses++;
            } else {
              if (Opc == Instruction::PHI) {
                // Do not count the Phi as a use. LCSSA may have inserted
                // plenty of trivial ones.
                NumUses--;
                for (auto PB = UseInstr->user_begin(),
                          PE = UseInstr->user_end();
                     PB != PE && NumUses <= 6; ++PB, ++NumUses) {
                  unsigned PhiOpc = cast<Instruction>(*PB)->getOpcode();
                  if (PhiOpc != Instruction::Call && PhiOpc != Instruction::Ret)
                    NumSoftExternalUses++;
                }
                continue;
              }
              if (Opc != Instruction::Call && Opc != Instruction::Ret)
                NumSoftExternalUses++;
            }
          }
          if (NumUses <= 6 && NumHardInternalUses && !NumSoftExternalUses)
            continue;
        }

        Value *ExitVal = Rewriter.expandCodeFor(ExitValue, PN->getType(), Inst);

        DEBUG(dbgs() << "INDVARS: RLEV: AfterLoopVal = " << *ExitVal << '\n'
                     << "  LoopVal = " << *Inst << "\n");

        if (!isValidRewrite(Inst, ExitVal)) {
          DeadInsts.push_back(ExitVal);
          continue;
        }
        Changed = true;
        ++NumReplaced;

        PN->setIncomingValue(i, ExitVal);

        // If this instruction is dead now, delete it. Don't do it now to avoid
        // invalidating iterators.
        if (isInstructionTriviallyDead(Inst, TLI))
          DeadInsts.push_back(Inst);

        // If we determined that this PHI is safe to replace even if an LCSSA
        // PHI, do so.
        if (LCSSASafePhiForRAUW) {
          PN->replaceAllUsesWith(ExitVal);
          PN->eraseFromParent();
        }
      }

      // If we were unable to completely replace the PHI node, clone the PHI
      // and delete the original one. This lets IVUsers and any other maps
      // purge the original user from their records.
      if (!LCSSASafePhiForRAUW) {
        PHINode *NewPN = cast<PHINode>(PN->clone());
        NewPN->takeName(PN);
        NewPN->insertBefore(PN);
        PN->replaceAllUsesWith(NewPN);
        PN->eraseFromParent();
      }
    }
  }

  // The insertion point instruction may have been deleted; clear it out
  // so that the rewriter doesn't trip over it later.
  Rewriter.clearInsertPoint();
}

//===----------------------------------------------------------------------===//
//  IV Widening - Extend the width of an IV to cover its widest uses.
//===----------------------------------------------------------------------===//

namespace {
  // Collect information about induction variables that are used by sign/zero
  // extend operations. This information is recorded by CollectExtend and
  // provides the input to WidenIV.
  struct WideIVInfo {
    PHINode *NarrowIV;
    Type *WidestNativeType; // Widest integer type created [sz]ext
    bool IsSigned;          // Was an sext user seen before a zext?

    WideIVInfo() : NarrowIV(nullptr), WidestNativeType(nullptr),
                   IsSigned(false) {}
  };
}

/// visitCast - Update information about the induction variable that is
/// extended by this sign or zero extend operation. This is used to determine
/// the final width of the IV before actually widening it.
static void visitIVCast(CastInst *Cast, WideIVInfo &WI, ScalarEvolution *SE,
                        const DataLayout *DL) {
  bool IsSigned = Cast->getOpcode() == Instruction::SExt;
  if (!IsSigned && Cast->getOpcode() != Instruction::ZExt)
    return;

  Type *Ty = Cast->getType();
  uint64_t Width = SE->getTypeSizeInBits(Ty);
  if (DL && !DL->isLegalInteger(Width))
    return;

  if (!WI.WidestNativeType) {
    WI.WidestNativeType = SE->getEffectiveSCEVType(Ty);
    WI.IsSigned = IsSigned;
    return;
  }

  // We extend the IV to satisfy the sign of its first user, arbitrarily.
  if (WI.IsSigned != IsSigned)
    return;

  if (Width > SE->getTypeSizeInBits(WI.WidestNativeType))
    WI.WidestNativeType = SE->getEffectiveSCEVType(Ty);
}

namespace {

/// NarrowIVDefUse - Record a link in the Narrow IV def-use chain along with the
/// WideIV that computes the same value as the Narrow IV def.  This avoids
/// caching Use* pointers.
struct NarrowIVDefUse {
  Instruction *NarrowDef;
  Instruction *NarrowUse;
  Instruction *WideDef;

  NarrowIVDefUse(): NarrowDef(nullptr), NarrowUse(nullptr), WideDef(nullptr) {}

  NarrowIVDefUse(Instruction *ND, Instruction *NU, Instruction *WD):
    NarrowDef(ND), NarrowUse(NU), WideDef(WD) {}
};

/// WidenIV - The goal of this transform is to remove sign and zero extends
/// without creating any new induction variables. To do this, it creates a new
/// phi of the wider type and redirects all users, either removing extends or
/// inserting truncs whenever we stop propagating the type.
///
class WidenIV {
  // Parameters
  PHINode *OrigPhi;
  Type *WideType;
  bool IsSigned;

  // Context
  LoopInfo        *LI;
  Loop            *L;
  ScalarEvolution *SE;
  DominatorTree   *DT;

  // Result
  PHINode *WidePhi;
  Instruction *WideInc;
  const SCEV *WideIncExpr;
  SmallVectorImpl<WeakVH> &DeadInsts;

  SmallPtrSet<Instruction*,16> Widened;
  SmallVector<NarrowIVDefUse, 8> NarrowIVUsers;

public:
  WidenIV(const WideIVInfo &WI, LoopInfo *LInfo,
          ScalarEvolution *SEv, DominatorTree *DTree,
          SmallVectorImpl<WeakVH> &DI) :
    OrigPhi(WI.NarrowIV),
    WideType(WI.WidestNativeType),
    IsSigned(WI.IsSigned),
    LI(LInfo),
    L(LI->getLoopFor(OrigPhi->getParent())),
    SE(SEv),
    DT(DTree),
    WidePhi(nullptr),
    WideInc(nullptr),
    WideIncExpr(nullptr),
    DeadInsts(DI) {
    assert(L->getHeader() == OrigPhi->getParent() && "Phi must be an IV");
  }

  PHINode *CreateWideIV(SCEVExpander &Rewriter);

protected:
  Value *getExtend(Value *NarrowOper, Type *WideType, bool IsSigned,
                   Instruction *Use);

  Instruction *CloneIVUser(NarrowIVDefUse DU);

  const SCEVAddRecExpr *GetWideRecurrence(Instruction *NarrowUse);

  const SCEVAddRecExpr* GetExtendedOperandRecurrence(NarrowIVDefUse DU);

  Instruction *WidenIVUse(NarrowIVDefUse DU, SCEVExpander &Rewriter);

  void pushNarrowIVUsers(Instruction *NarrowDef, Instruction *WideDef);
};
} // anonymous namespace

/// isLoopInvariant - Perform a quick domtree based check for loop invariance
/// assuming that V is used within the loop. LoopInfo::isLoopInvariant() seems
/// gratuitous for this purpose.
static bool isLoopInvariant(Value *V, const Loop *L, const DominatorTree *DT) {
  Instruction *Inst = dyn_cast<Instruction>(V);
  if (!Inst)
    return true;

  return DT->properlyDominates(Inst->getParent(), L->getHeader());
}

Value *WidenIV::getExtend(Value *NarrowOper, Type *WideType, bool IsSigned,
                          Instruction *Use) {
  // Set the debug location and conservative insertion point.
  IRBuilder<> Builder(Use);
  // Hoist the insertion point into loop preheaders as far as possible.
  for (const Loop *L = LI->getLoopFor(Use->getParent());
       L && L->getLoopPreheader() && isLoopInvariant(NarrowOper, L, DT);
       L = L->getParentLoop())
    Builder.SetInsertPoint(L->getLoopPreheader()->getTerminator());

  return IsSigned ? Builder.CreateSExt(NarrowOper, WideType) :
                    Builder.CreateZExt(NarrowOper, WideType);
}

/// CloneIVUser - Instantiate a wide operation to replace a narrow
/// operation. This only needs to handle operations that can evaluation to
/// SCEVAddRec. It can safely return 0 for any operation we decide not to clone.
Instruction *WidenIV::CloneIVUser(NarrowIVDefUse DU) {
  unsigned Opcode = DU.NarrowUse->getOpcode();
  switch (Opcode) {
  default:
    return nullptr;
  case Instruction::Add:
  case Instruction::Mul:
  case Instruction::UDiv:
  case Instruction::Sub:
  case Instruction::And:
  case Instruction::Or:
  case Instruction::Xor:
  case Instruction::Shl:
  case Instruction::LShr:
  case Instruction::AShr:
    DEBUG(dbgs() << "Cloning IVUser: " << *DU.NarrowUse << "\n");

    // Replace NarrowDef operands with WideDef. Otherwise, we don't know
    // anything about the narrow operand yet so must insert a [sz]ext. It is
    // probably loop invariant and will be folded or hoisted. If it actually
    // comes from a widened IV, it should be removed during a future call to
    // WidenIVUse.
    Value *LHS = (DU.NarrowUse->getOperand(0) == DU.NarrowDef) ? DU.WideDef :
      getExtend(DU.NarrowUse->getOperand(0), WideType, IsSigned, DU.NarrowUse);
    Value *RHS = (DU.NarrowUse->getOperand(1) == DU.NarrowDef) ? DU.WideDef :
      getExtend(DU.NarrowUse->getOperand(1), WideType, IsSigned, DU.NarrowUse);

    BinaryOperator *NarrowBO = cast<BinaryOperator>(DU.NarrowUse);
    BinaryOperator *WideBO = BinaryOperator::Create(NarrowBO->getOpcode(),
                                                    LHS, RHS,
                                                    NarrowBO->getName());
    IRBuilder<> Builder(DU.NarrowUse);
    Builder.Insert(WideBO);
    if (const OverflowingBinaryOperator *OBO =
        dyn_cast<OverflowingBinaryOperator>(NarrowBO)) {
      if (OBO->hasNoUnsignedWrap()) WideBO->setHasNoUnsignedWrap();
      if (OBO->hasNoSignedWrap()) WideBO->setHasNoSignedWrap();
    }
    return WideBO;
  }
}

/// No-wrap operations can transfer sign extension of their result to their
/// operands. Generate the SCEV value for the widened operation without
/// actually modifying the IR yet. If the expression after extending the
/// operands is an AddRec for this loop, return it.
const SCEVAddRecExpr* WidenIV::GetExtendedOperandRecurrence(NarrowIVDefUse DU) {
  // Handle the common case of add<nsw/nuw>
  if (DU.NarrowUse->getOpcode() != Instruction::Add)
    return nullptr;

  // One operand (NarrowDef) has already been extended to WideDef. Now determine
  // if extending the other will lead to a recurrence.
  unsigned ExtendOperIdx = DU.NarrowUse->getOperand(0) == DU.NarrowDef ? 1 : 0;
  assert(DU.NarrowUse->getOperand(1-ExtendOperIdx) == DU.NarrowDef && "bad DU");

  const SCEV *ExtendOperExpr = nullptr;
  const OverflowingBinaryOperator *OBO =
    cast<OverflowingBinaryOperator>(DU.NarrowUse);
  if (IsSigned && OBO->hasNoSignedWrap())
    ExtendOperExpr = SE->getSignExtendExpr(
      SE->getSCEV(DU.NarrowUse->getOperand(ExtendOperIdx)), WideType);
  else if(!IsSigned && OBO->hasNoUnsignedWrap())
    ExtendOperExpr = SE->getZeroExtendExpr(
      SE->getSCEV(DU.NarrowUse->getOperand(ExtendOperIdx)), WideType);
  else
    return nullptr;

  // When creating this AddExpr, don't apply the current operations NSW or NUW
  // flags. This instruction may be guarded by control flow that the no-wrap
  // behavior depends on. Non-control-equivalent instructions can be mapped to
  // the same SCEV expression, and it would be incorrect to transfer NSW/NUW
  // semantics to those operations.
  const SCEVAddRecExpr *AddRec = dyn_cast<SCEVAddRecExpr>(
    SE->getAddExpr(SE->getSCEV(DU.WideDef), ExtendOperExpr));

  if (!AddRec || AddRec->getLoop() != L)
    return nullptr;
  return AddRec;
}

/// GetWideRecurrence - Is this instruction potentially interesting from
/// IVUsers' perspective after widening it's type? In other words, can the
/// extend be safely hoisted out of the loop with SCEV reducing the value to a
/// recurrence on the same loop. If so, return the sign or zero extended
/// recurrence. Otherwise return NULL.
const SCEVAddRecExpr *WidenIV::GetWideRecurrence(Instruction *NarrowUse) {
  if (!SE->isSCEVable(NarrowUse->getType()))
    return nullptr;

  const SCEV *NarrowExpr = SE->getSCEV(NarrowUse);
  if (SE->getTypeSizeInBits(NarrowExpr->getType())
      >= SE->getTypeSizeInBits(WideType)) {
    // NarrowUse implicitly widens its operand. e.g. a gep with a narrow
    // index. So don't follow this use.
    return nullptr;
  }

  const SCEV *WideExpr = IsSigned ?
    SE->getSignExtendExpr(NarrowExpr, WideType) :
    SE->getZeroExtendExpr(NarrowExpr, WideType);
  const SCEVAddRecExpr *AddRec = dyn_cast<SCEVAddRecExpr>(WideExpr);
  if (!AddRec || AddRec->getLoop() != L)
    return nullptr;
  return AddRec;
}

/// This IV user cannot be widen. Replace this use of the original narrow IV
/// with a truncation of the new wide IV to isolate and eliminate the narrow IV.
static void truncateIVUse(NarrowIVDefUse DU, DominatorTree *DT) {
  DEBUG(dbgs() << "INDVARS: Truncate IV " << *DU.WideDef
        << " for user " << *DU.NarrowUse << "\n");
  IRBuilder<> Builder(getInsertPointForUses(DU.NarrowUse, DU.NarrowDef, DT));
  Value *Trunc = Builder.CreateTrunc(DU.WideDef, DU.NarrowDef->getType());
  DU.NarrowUse->replaceUsesOfWith(DU.NarrowDef, Trunc);
}

/// WidenIVUse - Determine whether an individual user of the narrow IV can be
/// widened. If so, return the wide clone of the user.
Instruction *WidenIV::WidenIVUse(NarrowIVDefUse DU, SCEVExpander &Rewriter) {

  // Stop traversing the def-use chain at inner-loop phis or post-loop phis.
  if (PHINode *UsePhi = dyn_cast<PHINode>(DU.NarrowUse)) {
    if (LI->getLoopFor(UsePhi->getParent()) != L) {
      // For LCSSA phis, sink the truncate outside the loop.
      // After SimplifyCFG most loop exit targets have a single predecessor.
      // Otherwise fall back to a truncate within the loop.
      if (UsePhi->getNumOperands() != 1)
        truncateIVUse(DU, DT);
      else {
        PHINode *WidePhi =
          PHINode::Create(DU.WideDef->getType(), 1, UsePhi->getName() + ".wide",
                          UsePhi);
        WidePhi->addIncoming(DU.WideDef, UsePhi->getIncomingBlock(0));
        IRBuilder<> Builder(WidePhi->getParent()->getFirstInsertionPt());
        Value *Trunc = Builder.CreateTrunc(WidePhi, DU.NarrowDef->getType());
        UsePhi->replaceAllUsesWith(Trunc);
        DeadInsts.push_back(UsePhi);
        DEBUG(dbgs() << "INDVARS: Widen lcssa phi " << *UsePhi
              << " to " << *WidePhi << "\n");
      }
      return nullptr;
    }
  }
  // Our raison d'etre! Eliminate sign and zero extension.
  if (IsSigned ? isa<SExtInst>(DU.NarrowUse) : isa<ZExtInst>(DU.NarrowUse)) {
    Value *NewDef = DU.WideDef;
    if (DU.NarrowUse->getType() != WideType) {
      unsigned CastWidth = SE->getTypeSizeInBits(DU.NarrowUse->getType());
      unsigned IVWidth = SE->getTypeSizeInBits(WideType);
      if (CastWidth < IVWidth) {
        // The cast isn't as wide as the IV, so insert a Trunc.
        IRBuilder<> Builder(DU.NarrowUse);
        NewDef = Builder.CreateTrunc(DU.WideDef, DU.NarrowUse->getType());
      }
      else {
        // A wider extend was hidden behind a narrower one. This may induce
        // another round of IV widening in which the intermediate IV becomes
        // dead. It should be very rare.
        DEBUG(dbgs() << "INDVARS: New IV " << *WidePhi
              << " not wide enough to subsume " << *DU.NarrowUse << "\n");
        DU.NarrowUse->replaceUsesOfWith(DU.NarrowDef, DU.WideDef);
        NewDef = DU.NarrowUse;
      }
    }
    if (NewDef != DU.NarrowUse) {
      DEBUG(dbgs() << "INDVARS: eliminating " << *DU.NarrowUse
            << " replaced by " << *DU.WideDef << "\n");
      ++NumElimExt;
      DU.NarrowUse->replaceAllUsesWith(NewDef);
      DeadInsts.push_back(DU.NarrowUse);
    }
    // Now that the extend is gone, we want to expose it's uses for potential
    // further simplification. We don't need to directly inform SimplifyIVUsers
    // of the new users, because their parent IV will be processed later as a
    // new loop phi. If we preserved IVUsers analysis, we would also want to
    // push the uses of WideDef here.

    // No further widening is needed. The deceased [sz]ext had done it for us.
    return nullptr;
  }

  // Does this user itself evaluate to a recurrence after widening?
  const SCEVAddRecExpr *WideAddRec = GetWideRecurrence(DU.NarrowUse);
  if (!WideAddRec) {
      WideAddRec = GetExtendedOperandRecurrence(DU);
  }
  if (!WideAddRec) {
    // This user does not evaluate to a recurence after widening, so don't
    // follow it. Instead insert a Trunc to kill off the original use,
    // eventually isolating the original narrow IV so it can be removed.
    truncateIVUse(DU, DT);
    return nullptr;
  }
  // Assume block terminators cannot evaluate to a recurrence. We can't to
  // insert a Trunc after a terminator if there happens to be a critical edge.
  assert(DU.NarrowUse != DU.NarrowUse->getParent()->getTerminator() &&
         "SCEV is not expected to evaluate a block terminator");

  // Reuse the IV increment that SCEVExpander created as long as it dominates
  // NarrowUse.
  Instruction *WideUse = nullptr;
  if (WideAddRec == WideIncExpr
      && Rewriter.hoistIVInc(WideInc, DU.NarrowUse))
    WideUse = WideInc;
  else {
    WideUse = CloneIVUser(DU);
    if (!WideUse)
      return nullptr;
  }
  // Evaluation of WideAddRec ensured that the narrow expression could be
  // extended outside the loop without overflow. This suggests that the wide use
  // evaluates to the same expression as the extended narrow use, but doesn't
  // absolutely guarantee it. Hence the following failsafe check. In rare cases
  // where it fails, we simply throw away the newly created wide use.
  if (WideAddRec != SE->getSCEV(WideUse)) {
    DEBUG(dbgs() << "Wide use expression mismatch: " << *WideUse
          << ": " << *SE->getSCEV(WideUse) << " != " << *WideAddRec << "\n");
    DeadInsts.push_back(WideUse);
    return nullptr;
  }

  // Returning WideUse pushes it on the worklist.
  return WideUse;
}

/// pushNarrowIVUsers - Add eligible users of NarrowDef to NarrowIVUsers.
///
void WidenIV::pushNarrowIVUsers(Instruction *NarrowDef, Instruction *WideDef) {
  for (User *U : NarrowDef->users()) {
    Instruction *NarrowUser = cast<Instruction>(U);

    // Handle data flow merges and bizarre phi cycles.
    if (!Widened.insert(NarrowUser))
      continue;

    NarrowIVUsers.push_back(NarrowIVDefUse(NarrowDef, NarrowUser, WideDef));
  }
}

/// CreateWideIV - Process a single induction variable. First use the
/// SCEVExpander to create a wide induction variable that evaluates to the same
/// recurrence as the original narrow IV. Then use a worklist to forward
/// traverse the narrow IV's def-use chain. After WidenIVUse has processed all
/// interesting IV users, the narrow IV will be isolated for removal by
/// DeleteDeadPHIs.
///
/// It would be simpler to delete uses as they are processed, but we must avoid
/// invalidating SCEV expressions.
///
PHINode *WidenIV::CreateWideIV(SCEVExpander &Rewriter) {
  // Is this phi an induction variable?
  const SCEVAddRecExpr *AddRec = dyn_cast<SCEVAddRecExpr>(SE->getSCEV(OrigPhi));
  if (!AddRec)
    return nullptr;

  // Widen the induction variable expression.
  const SCEV *WideIVExpr = IsSigned ?
    SE->getSignExtendExpr(AddRec, WideType) :
    SE->getZeroExtendExpr(AddRec, WideType);

  assert(SE->getEffectiveSCEVType(WideIVExpr->getType()) == WideType &&
         "Expect the new IV expression to preserve its type");

  // Can the IV be extended outside the loop without overflow?
  AddRec = dyn_cast<SCEVAddRecExpr>(WideIVExpr);
  if (!AddRec || AddRec->getLoop() != L)
    return nullptr;

  // An AddRec must have loop-invariant operands. Since this AddRec is
  // materialized by a loop header phi, the expression cannot have any post-loop
  // operands, so they must dominate the loop header.
  assert(SE->properlyDominates(AddRec->getStart(), L->getHeader()) &&
         SE->properlyDominates(AddRec->getStepRecurrence(*SE), L->getHeader())
         && "Loop header phi recurrence inputs do not dominate the loop");

  // The rewriter provides a value for the desired IV expression. This may
  // either find an existing phi or materialize a new one. Either way, we
  // expect a well-formed cyclic phi-with-increments. i.e. any operand not part
  // of the phi-SCC dominates the loop entry.
  Instruction *InsertPt = L->getHeader()->begin();
  WidePhi = cast<PHINode>(Rewriter.expandCodeFor(AddRec, WideType, InsertPt));

  // Remembering the WideIV increment generated by SCEVExpander allows
  // WidenIVUse to reuse it when widening the narrow IV's increment. We don't
  // employ a general reuse mechanism because the call above is the only call to
  // SCEVExpander. Henceforth, we produce 1-to-1 narrow to wide uses.
  if (BasicBlock *LatchBlock = L->getLoopLatch()) {
    WideInc =
      cast<Instruction>(WidePhi->getIncomingValueForBlock(LatchBlock));
    WideIncExpr = SE->getSCEV(WideInc);
  }

  DEBUG(dbgs() << "Wide IV: " << *WidePhi << "\n");
  ++NumWidened;

  // Traverse the def-use chain using a worklist starting at the original IV.
  assert(Widened.empty() && NarrowIVUsers.empty() && "expect initial state" );

  Widened.insert(OrigPhi);
  pushNarrowIVUsers(OrigPhi, WidePhi);

  while (!NarrowIVUsers.empty()) {
    NarrowIVDefUse DU = NarrowIVUsers.pop_back_val();

    // Process a def-use edge. This may replace the use, so don't hold a
    // use_iterator across it.
    Instruction *WideUse = WidenIVUse(DU, Rewriter);

    // Follow all def-use edges from the previous narrow use.
    if (WideUse)
      pushNarrowIVUsers(DU.NarrowUse, WideUse);

    // WidenIVUse may have removed the def-use edge.
    if (DU.NarrowDef->use_empty())
      DeadInsts.push_back(DU.NarrowDef);
  }
  return WidePhi;
}

//===----------------------------------------------------------------------===//
//  Live IV Reduction - Minimize IVs live across the loop.
//===----------------------------------------------------------------------===//


//===----------------------------------------------------------------------===//
//  Simplification of IV users based on SCEV evaluation.
//===----------------------------------------------------------------------===//

namespace {
  class IndVarSimplifyVisitor : public IVVisitor {
    ScalarEvolution *SE;
    const DataLayout *DL;
    PHINode *IVPhi;

  public:
    WideIVInfo WI;

    IndVarSimplifyVisitor(PHINode *IV, ScalarEvolution *SCEV,
                          const DataLayout *DL, const DominatorTree *DTree):
      SE(SCEV), DL(DL), IVPhi(IV) {
      DT = DTree;
      WI.NarrowIV = IVPhi;
      if (ReduceLiveIVs)
        setSplitOverflowIntrinsics();
    }

    // Implement the interface used by simplifyUsersOfIV.
    void visitCast(CastInst *Cast) override { visitIVCast(Cast, WI, SE, DL); }
  };
}

/// SimplifyAndExtend - Iteratively perform simplification on a worklist of IV
/// users. Each successive simplification may push more users which may
/// themselves be candidates for simplification.
///
/// Sign/Zero extend elimination is interleaved with IV simplification.
///
void IndVarSimplify::SimplifyAndExtend(Loop *L,
                                       SCEVExpander &Rewriter,
                                       LPPassManager &LPM) {
  SmallVector<WideIVInfo, 8> WideIVs;

  SmallVector<PHINode*, 8> LoopPhis;
  for (BasicBlock::iterator I = L->getHeader()->begin(); isa<PHINode>(I); ++I) {
    LoopPhis.push_back(cast<PHINode>(I));
  }
  // Each round of simplification iterates through the SimplifyIVUsers worklist
  // for all current phis, then determines whether any IVs can be
  // widened. Widening adds new phis to LoopPhis, inducing another round of
  // simplification on the wide IVs.
  while (!LoopPhis.empty()) {
    // Evaluate as many IV expressions as possible before widening any IVs. This
    // forces SCEV to set no-wrap flags before evaluating sign/zero
    // extension. The first time SCEV attempts to normalize sign/zero extension,
    // the result becomes final. So for the most predictable results, we delay
    // evaluation of sign/zero extend evaluation until needed, and avoid running
    // other SCEV based analysis prior to SimplifyAndExtend.
    do {
      PHINode *CurrIV = LoopPhis.pop_back_val();

      // Information about sign/zero extensions of CurrIV.
      IndVarSimplifyVisitor Visitor(CurrIV, SE, DL, DT);

      Changed |= simplifyUsersOfIV(CurrIV, SE, &LPM, DeadInsts, &Visitor);

      if (Visitor.WI.WidestNativeType) {
        WideIVs.push_back(Visitor.WI);
      }
    } while(!LoopPhis.empty());

    for (; !WideIVs.empty(); WideIVs.pop_back()) {
      WidenIV Widener(WideIVs.back(), LI, SE, DT, DeadInsts);
      if (PHINode *WidePhi = Widener.CreateWideIV(Rewriter)) {
        Changed = true;
        LoopPhis.push_back(WidePhi);
      }
    }
  }
}

//===----------------------------------------------------------------------===//
//  LinearFunctionTestReplace and its kin. Rewrite the loop exit condition.
//===----------------------------------------------------------------------===//

/// Check for expressions that ScalarEvolution generates to compute
/// BackedgeTakenInfo. If these expressions have not been reduced, then
/// expanding them may incur additional cost (albeit in the loop preheader).
static bool isHighCostExpansion(const SCEV *S, BranchInst *BI,
                                SmallPtrSet<const SCEV*, 8> &Processed,
                                ScalarEvolution *SE) {
  if (!Processed.insert(S))
    return false;

  // If the backedge-taken count is a UDiv, it's very likely a UDiv that
  // ScalarEvolution's HowFarToZero or HowManyLessThans produced to compute a
  // precise expression, rather than a UDiv from the user's code. If we can't
  // find a UDiv in the code with some simple searching, assume the former and
  // forego rewriting the loop.
  if (isa<SCEVUDivExpr>(S)) {
    ICmpInst *OrigCond = dyn_cast<ICmpInst>(BI->getCondition());
    if (!OrigCond) return true;
    const SCEV *R = SE->getSCEV(OrigCond->getOperand(1));
    R = SE->getMinusSCEV(R, SE->getConstant(R->getType(), 1));
    if (R != S) {
      const SCEV *L = SE->getSCEV(OrigCond->getOperand(0));
      L = SE->getMinusSCEV(L, SE->getConstant(L->getType(), 1));
      if (L != S)
        return true;
    }
  }

  // Recurse past add expressions, which commonly occur in the
  // BackedgeTakenCount. They may already exist in program code, and if not,
  // they are not too expensive rematerialize.
  if (const SCEVAddExpr *Add = dyn_cast<SCEVAddExpr>(S)) {
    for (SCEVAddExpr::op_iterator I = Add->op_begin(), E = Add->op_end();
         I != E; ++I) {
      if (isHighCostExpansion(*I, BI, Processed, SE))
        return true;
    }
    return false;
  }

  // HowManyLessThans uses a Max expression whenever the loop is not guarded by
  // the exit condition.
  if (isa<SCEVSMaxExpr>(S) || isa<SCEVUMaxExpr>(S))
    return true;

  // If we haven't recognized an expensive SCEV pattern, assume it's an
  // expression produced by program code.
  return false;
}

/// canExpandBackedgeTakenCount - Return true if this loop's backedge taken
/// count expression can be safely and cheaply expanded into an instruction
/// sequence that can be used by LinearFunctionTestReplace.
///
/// TODO: This fails for pointer-type loop counters with greater than one byte
/// strides, consequently preventing LFTR from running. For the purpose of LFTR
/// we could skip this check in the case that the LFTR loop counter (chosen by
/// FindLoopCounter) is also pointer type. Instead, we could directly convert
/// the loop test to an inequality test by checking the target data's alignment
/// of element types (given that the initial pointer value originates from or is
/// used by ABI constrained operation, as opposed to inttoptr/ptrtoint).
/// However, we don't yet have a strong motivation for converting loop tests
/// into inequality tests.
static bool canExpandBackedgeTakenCount(Loop *L, ScalarEvolution *SE) {
  const SCEV *BackedgeTakenCount = SE->getBackedgeTakenCount(L);
  if (isa<SCEVCouldNotCompute>(BackedgeTakenCount) ||
      BackedgeTakenCount->isZero())
    return false;

  if (!L->getExitingBlock())
    return false;

  // Can't rewrite non-branch yet.
  BranchInst *BI = dyn_cast<BranchInst>(L->getExitingBlock()->getTerminator());
  if (!BI)
    return false;

  SmallPtrSet<const SCEV*, 8> Processed;
  if (isHighCostExpansion(BackedgeTakenCount, BI, Processed, SE))
    return false;

  return true;
}

/// getLoopPhiForCounter - Return the loop header phi IFF IncV adds a loop
/// invariant value to the phi.
static PHINode *getLoopPhiForCounter(Value *IncV, Loop *L, DominatorTree *DT) {
  Instruction *IncI = dyn_cast<Instruction>(IncV);
  if (!IncI)
    return nullptr;

  switch (IncI->getOpcode()) {
  case Instruction::Add:
  case Instruction::Sub:
    break;
  case Instruction::GetElementPtr:
    // An IV counter must preserve its type.
    if (IncI->getNumOperands() == 2)
      break;
  default:
    return nullptr;
  }

  PHINode *Phi = dyn_cast<PHINode>(IncI->getOperand(0));
  if (Phi && Phi->getParent() == L->getHeader()) {
    if (isLoopInvariant(IncI->getOperand(1), L, DT))
      return Phi;
    return nullptr;
  }
  if (IncI->getOpcode() == Instruction::GetElementPtr)
    return nullptr;

  // Allow add/sub to be commuted.
  Phi = dyn_cast<PHINode>(IncI->getOperand(1));
  if (Phi && Phi->getParent() == L->getHeader()) {
    if (isLoopInvariant(IncI->getOperand(0), L, DT))
      return Phi;
  }
  return nullptr;
}

/// Return the compare guarding the loop latch, or NULL for unrecognized tests.
static ICmpInst *getLoopTest(Loop *L) {
  assert(L->getExitingBlock() && "expected loop exit");

  BasicBlock *LatchBlock = L->getLoopLatch();
  // Don't bother with LFTR if the loop is not properly simplified.
  if (!LatchBlock)
    return nullptr;

  BranchInst *BI = dyn_cast<BranchInst>(L->getExitingBlock()->getTerminator());
  assert(BI && "expected exit branch");

  return dyn_cast<ICmpInst>(BI->getCondition());
}

/// needsLFTR - LinearFunctionTestReplace policy. Return true unless we can show
/// that the current exit test is already sufficiently canonical.
static bool needsLFTR(Loop *L, DominatorTree *DT) {
  // Do LFTR to simplify the exit condition to an ICMP.
  ICmpInst *Cond = getLoopTest(L);
  if (!Cond)
    return true;

  // Do LFTR to simplify the exit ICMP to EQ/NE
  ICmpInst::Predicate Pred = Cond->getPredicate();
  if (Pred != ICmpInst::ICMP_NE && Pred != ICmpInst::ICMP_EQ)
    return true;

  // Look for a loop invariant RHS
  Value *LHS = Cond->getOperand(0);
  Value *RHS = Cond->getOperand(1);
  if (!isLoopInvariant(RHS, L, DT)) {
    if (!isLoopInvariant(LHS, L, DT))
      return true;
    std::swap(LHS, RHS);
  }
  // Look for a simple IV counter LHS
  PHINode *Phi = dyn_cast<PHINode>(LHS);
  if (!Phi)
    Phi = getLoopPhiForCounter(LHS, L, DT);

  if (!Phi)
    return true;

  // Do LFTR if PHI node is defined in the loop, but is *not* a counter.
  int Idx = Phi->getBasicBlockIndex(L->getLoopLatch());
  if (Idx < 0)
    return true;

  // Do LFTR if the exit condition's IV is *not* a simple counter.
  Value *IncV = Phi->getIncomingValue(Idx);
  return Phi != getLoopPhiForCounter(IncV, L, DT);
}

/// Recursive helper for hasConcreteDef(). Unfortunately, this currently boils
/// down to checking that all operands are constant and listing instructions
/// that may hide undef.
static bool hasConcreteDefImpl(Value *V, SmallPtrSet<Value*, 8> &Visited,
                               unsigned Depth) {
  if (isa<Constant>(V))
    return !isa<UndefValue>(V);

  if (Depth >= 6)
    return false;

  // Conservatively handle non-constant non-instructions. For example, Arguments
  // may be undef.
  Instruction *I = dyn_cast<Instruction>(V);
  if (!I)
    return false;

  // Load and return values may be undef.
  if(I->mayReadFromMemory() || isa<CallInst>(I) || isa<InvokeInst>(I))
    return false;

  // Optimistically handle other instructions.
  for (User::op_iterator OI = I->op_begin(), E = I->op_end(); OI != E; ++OI) {
    if (!Visited.insert(*OI))
      continue;
    if (!hasConcreteDefImpl(*OI, Visited, Depth+1))
      return false;
  }
  return true;
}

/// Return true if the given value is concrete. We must prove that undef can
/// never reach it.
///
/// TODO: If we decide that this is a good approach to checking for undef, we
/// may factor it into a common location.
static bool hasConcreteDef(Value *V) {
  SmallPtrSet<Value*, 8> Visited;
  Visited.insert(V);
  return hasConcreteDefImpl(V, Visited, 0);
}

/// AlmostDeadIV - Return true if this IV has any uses other than the (soon to
/// be rewritten) loop exit test.
static bool AlmostDeadIV(PHINode *Phi, BasicBlock *LatchBlock, Value *Cond) {
  int LatchIdx = Phi->getBasicBlockIndex(LatchBlock);
  Value *IncV = Phi->getIncomingValue(LatchIdx);

  for (User *U : Phi->users())
    if (U != Cond && U != IncV) return false;

  for (User *U : IncV->users())
    if (U != Cond && U != Phi) return false;
  return true;
}

/// FindLoopCounter - Find an affine IV in canonical form.
///
/// BECount may be an i8* pointer type. The pointer difference is already
/// valid count without scaling the address stride, so it remains a pointer
/// expression as far as SCEV is concerned.
///
/// Currently only valid for LFTR. See the comments on hasConcreteDef below.
///
/// FIXME: Accept -1 stride and set IVLimit = IVInit - BECount
///
/// FIXME: Accept non-unit stride as long as SCEV can reduce BECount * Stride.
/// This is difficult in general for SCEV because of potential overflow. But we
/// could at least handle constant BECounts.
static PHINode *
FindLoopCounter(Loop *L, const SCEV *BECount,
                ScalarEvolution *SE, DominatorTree *DT, const DataLayout *DL) {
  uint64_t BCWidth = SE->getTypeSizeInBits(BECount->getType());

  Value *Cond =
    cast<BranchInst>(L->getExitingBlock()->getTerminator())->getCondition();

  // Loop over all of the PHI nodes, looking for a simple counter.
  PHINode *BestPhi = nullptr;
  const SCEV *BestInit = nullptr;
  BasicBlock *LatchBlock = L->getLoopLatch();
  assert(LatchBlock && "needsLFTR should guarantee a loop latch");

  for (BasicBlock::iterator I = L->getHeader()->begin(); isa<PHINode>(I); ++I) {
    PHINode *Phi = cast<PHINode>(I);
    if (!SE->isSCEVable(Phi->getType()))
      continue;

    // Avoid comparing an integer IV against a pointer Limit.
    if (BECount->getType()->isPointerTy() && !Phi->getType()->isPointerTy())
      continue;

    const SCEVAddRecExpr *AR = dyn_cast<SCEVAddRecExpr>(SE->getSCEV(Phi));
    if (!AR || AR->getLoop() != L || !AR->isAffine())
      continue;

    // AR may be a pointer type, while BECount is an integer type.
    // AR may be wider than BECount. With eq/ne tests overflow is immaterial.
    // AR may not be a narrower type, or we may never exit.
    uint64_t PhiWidth = SE->getTypeSizeInBits(AR->getType());
    if (PhiWidth < BCWidth || (DL && !DL->isLegalInteger(PhiWidth)))
      continue;

    const SCEV *Step = dyn_cast<SCEVConstant>(AR->getStepRecurrence(*SE));
    if (!Step || !Step->isOne())
      continue;

    int LatchIdx = Phi->getBasicBlockIndex(LatchBlock);
    Value *IncV = Phi->getIncomingValue(LatchIdx);
    if (getLoopPhiForCounter(IncV, L, DT) != Phi)
      continue;

    // Avoid reusing a potentially undef value to compute other values that may
    // have originally had a concrete definition.
    if (!hasConcreteDef(Phi)) {
      // We explicitly allow unknown phis as long as they are already used by
      // the loop test. In this case we assume that performing LFTR could not
      // increase the number of undef users.
      if (ICmpInst *Cond = getLoopTest(L)) {
        if (Phi != getLoopPhiForCounter(Cond->getOperand(0), L, DT)
            && Phi != getLoopPhiForCounter(Cond->getOperand(1), L, DT)) {
          continue;
        }
      }
    }
    const SCEV *Init = AR->getStart();

    if (BestPhi && !AlmostDeadIV(BestPhi, LatchBlock, Cond)) {
      // Don't force a live loop counter if another IV can be used.
      if (AlmostDeadIV(Phi, LatchBlock, Cond))
        continue;

      // Prefer to count-from-zero. This is a more "canonical" counter form. It
      // also prefers integer to pointer IVs.
      if (BestInit->isZero() != Init->isZero()) {
        if (BestInit->isZero())
          continue;
      }
      // If two IVs both count from zero or both count from nonzero then the
      // narrower is likely a dead phi that has been widened. Use the wider phi
      // to allow the other to be eliminated.
      else if (PhiWidth <= SE->getTypeSizeInBits(BestPhi->getType()))
        continue;
    }
    BestPhi = Phi;
    BestInit = Init;
  }
  return BestPhi;
}

/// genLoopLimit - Help LinearFunctionTestReplace by generating a value that
/// holds the RHS of the new loop test.
static Value *genLoopLimit(PHINode *IndVar, const SCEV *IVCount, Loop *L,
                           SCEVExpander &Rewriter, ScalarEvolution *SE) {
  const SCEVAddRecExpr *AR = dyn_cast<SCEVAddRecExpr>(SE->getSCEV(IndVar));
  assert(AR && AR->getLoop() == L && AR->isAffine() && "bad loop counter");
  const SCEV *IVInit = AR->getStart();

  // IVInit may be a pointer while IVCount is an integer when FindLoopCounter
  // finds a valid pointer IV. Sign extend BECount in order to materialize a
  // GEP. Avoid running SCEVExpander on a new pointer value, instead reusing
  // the existing GEPs whenever possible.
  if (IndVar->getType()->isPointerTy()
      && !IVCount->getType()->isPointerTy()) {

    // IVOffset will be the new GEP offset that is interpreted by GEP as a
    // signed value. IVCount on the other hand represents the loop trip count,
    // which is an unsigned value. FindLoopCounter only allows induction
    // variables that have a positive unit stride of one. This means we don't
    // have to handle the case of negative offsets (yet) and just need to zero
    // extend IVCount.
    Type *OfsTy = SE->getEffectiveSCEVType(IVInit->getType());
    const SCEV *IVOffset = SE->getTruncateOrZeroExtend(IVCount, OfsTy);

    // Expand the code for the iteration count.
    assert(SE->isLoopInvariant(IVOffset, L) &&
           "Computed iteration count is not loop invariant!");
    BranchInst *BI = cast<BranchInst>(L->getExitingBlock()->getTerminator());
    Value *GEPOffset = Rewriter.expandCodeFor(IVOffset, OfsTy, BI);

    Value *GEPBase = IndVar->getIncomingValueForBlock(L->getLoopPreheader());
    assert(AR->getStart() == SE->getSCEV(GEPBase) && "bad loop counter");
    // We could handle pointer IVs other than i8*, but we need to compensate for
    // gep index scaling. See canExpandBackedgeTakenCount comments.
    assert(SE->getSizeOfExpr(IntegerType::getInt64Ty(IndVar->getContext()),
             cast<PointerType>(GEPBase->getType())->getElementType())->isOne()
           && "unit stride pointer IV must be i8*");

    IRBuilder<> Builder(L->getLoopPreheader()->getTerminator());
    return Builder.CreateGEP(GEPBase, GEPOffset, "lftr.limit");
  }
  else {
    // In any other case, convert both IVInit and IVCount to integers before
    // comparing. This may result in SCEV expension of pointers, but in practice
    // SCEV will fold the pointer arithmetic away as such:
    // BECount = (IVEnd - IVInit - 1) => IVLimit = IVInit (postinc).
    //
    // Valid Cases: (1) both integers is most common; (2) both may be pointers
    // for simple memset-style loops.
    //
    // IVInit integer and IVCount pointer would only occur if a canonical IV
    // were generated on top of case #2, which is not expected.

    const SCEV *IVLimit = nullptr;
    // For unit stride, IVCount = Start + BECount with 2's complement overflow.
    // For non-zero Start, compute IVCount here.
    if (AR->getStart()->isZero())
      IVLimit = IVCount;
    else {
      assert(AR->getStepRecurrence(*SE)->isOne() && "only handles unit stride");
      const SCEV *IVInit = AR->getStart();

      // For integer IVs, truncate the IV before computing IVInit + BECount.
      if (SE->getTypeSizeInBits(IVInit->getType())
          > SE->getTypeSizeInBits(IVCount->getType()))
        IVInit = SE->getTruncateExpr(IVInit, IVCount->getType());

      IVLimit = SE->getAddExpr(IVInit, IVCount);
    }
    // Expand the code for the iteration count.
    BranchInst *BI = cast<BranchInst>(L->getExitingBlock()->getTerminator());
    IRBuilder<> Builder(BI);
    assert(SE->isLoopInvariant(IVLimit, L) &&
           "Computed iteration count is not loop invariant!");
    // Ensure that we generate the same type as IndVar, or a smaller integer
    // type. In the presence of null pointer values, we have an integer type
    // SCEV expression (IVInit) for a pointer type IV value (IndVar).
    Type *LimitTy = IVCount->getType()->isPointerTy() ?
      IndVar->getType() : IVCount->getType();
    return Rewriter.expandCodeFor(IVLimit, LimitTy, BI);
  }
}

/// LinearFunctionTestReplace - This method rewrites the exit condition of the
/// loop to be a canonical != comparison against the incremented loop induction
/// variable.  This pass is able to rewrite the exit tests of any loop where the
/// SCEV analysis can determine a loop-invariant trip count of the loop, which
/// is actually a much broader range than just linear tests.
Value *IndVarSimplify::
LinearFunctionTestReplace(Loop *L,
                          const SCEV *BackedgeTakenCount,
                          PHINode *IndVar,
                          SCEVExpander &Rewriter) {
  assert(canExpandBackedgeTakenCount(L, SE) && "precondition");

  // Initialize CmpIndVar and IVCount to their preincremented values.
  Value *CmpIndVar = IndVar;
  const SCEV *IVCount = BackedgeTakenCount;

  // If the exiting block is the same as the backedge block, we prefer to
  // compare against the post-incremented value, otherwise we must compare
  // against the preincremented value.
  if (L->getExitingBlock() == L->getLoopLatch()) {
    // Add one to the "backedge-taken" count to get the trip count.
    // This addition may overflow, which is valid as long as the comparison is
    // truncated to BackedgeTakenCount->getType().
    IVCount = SE->getAddExpr(BackedgeTakenCount,
                             SE->getConstant(BackedgeTakenCount->getType(), 1));
    // The BackedgeTaken expression contains the number of times that the
    // backedge branches to the loop header.  This is one less than the
    // number of times the loop executes, so use the incremented indvar.
    CmpIndVar = IndVar->getIncomingValueForBlock(L->getExitingBlock());
  }

  Value *ExitCnt = genLoopLimit(IndVar, IVCount, L, Rewriter, SE);
  assert(ExitCnt->getType()->isPointerTy() == IndVar->getType()->isPointerTy()
         && "genLoopLimit missed a cast");

  // Insert a new icmp_ne or icmp_eq instruction before the branch.
  BranchInst *BI = cast<BranchInst>(L->getExitingBlock()->getTerminator());
  ICmpInst::Predicate P;
  if (L->contains(BI->getSuccessor(0)))
    P = ICmpInst::ICMP_NE;
  else
    P = ICmpInst::ICMP_EQ;

  DEBUG(dbgs() << "INDVARS: Rewriting loop exit condition to:\n"
               << "      LHS:" << *CmpIndVar << '\n'
               << "       op:\t"
               << (P == ICmpInst::ICMP_NE ? "!=" : "==") << "\n"
               << "      RHS:\t" << *ExitCnt << "\n"
               << "  IVCount:\t" << *IVCount << "\n");

  IRBuilder<> Builder(BI);

  // LFTR can ignore IV overflow and truncate to the width of
  // BECount. This avoids materializing the add(zext(add)) expression.
  unsigned CmpIndVarSize = SE->getTypeSizeInBits(CmpIndVar->getType());
  unsigned ExitCntSize = SE->getTypeSizeInBits(ExitCnt->getType());
  if (CmpIndVarSize > ExitCntSize) {
    const SCEVAddRecExpr *AR = cast<SCEVAddRecExpr>(SE->getSCEV(IndVar));
    const SCEV *ARStart = AR->getStart();
    const SCEV *ARStep = AR->getStepRecurrence(*SE);
    // For constant IVCount, avoid truncation.
    if (isa<SCEVConstant>(ARStart) && isa<SCEVConstant>(IVCount)) {
      const APInt &Start = cast<SCEVConstant>(ARStart)->getValue()->getValue();
      APInt Count = cast<SCEVConstant>(IVCount)->getValue()->getValue();
      // Note that the post-inc value of BackedgeTakenCount may have overflowed
      // above such that IVCount is now zero.
      if (IVCount != BackedgeTakenCount && Count == 0) {
        Count = APInt::getMaxValue(Count.getBitWidth()).zext(CmpIndVarSize);
        ++Count;
      }
      else
        Count = Count.zext(CmpIndVarSize);
      APInt NewLimit;
      if (cast<SCEVConstant>(ARStep)->getValue()->isNegative())
        NewLimit = Start - Count;
      else
        NewLimit = Start + Count;
      ExitCnt = ConstantInt::get(CmpIndVar->getType(), NewLimit);

      DEBUG(dbgs() << "  Widen RHS:\t" << *ExitCnt << "\n");
    } else {
      CmpIndVar = Builder.CreateTrunc(CmpIndVar, ExitCnt->getType(),
                                      "lftr.wideiv");
    }
  }
  Value *Cond = Builder.CreateICmp(P, CmpIndVar, ExitCnt, "exitcond");
  Value *OrigCond = BI->getCondition();
  // It's tempting to use replaceAllUsesWith here to fully replace the old
  // comparison, but that's not immediately safe, since users of the old
  // comparison may not be dominated by the new comparison. Instead, just
  // update the branch to use the new comparison; in the common case this
  // will make old comparison dead.
  BI->setCondition(Cond);
  DeadInsts.push_back(OrigCond);

  ++NumLFTR;
  Changed = true;
  return Cond;
}

//===----------------------------------------------------------------------===//
//  SinkUnusedInvariants. A late subpass to cleanup loop preheaders.
//===----------------------------------------------------------------------===//

/// If there's a single exit block, sink any loop-invariant values that
/// were defined in the preheader but not used inside the loop into the
/// exit block to reduce register pressure in the loop.
void IndVarSimplify::SinkUnusedInvariants(Loop *L) {
  BasicBlock *ExitBlock = L->getExitBlock();
  if (!ExitBlock) return;

  BasicBlock *Preheader = L->getLoopPreheader();
  if (!Preheader) return;

  Instruction *InsertPt = ExitBlock->getFirstInsertionPt();
  BasicBlock::iterator I = Preheader->getTerminator();
  while (I != Preheader->begin()) {
    --I;
    // New instructions were inserted at the end of the preheader.
    if (isa<PHINode>(I))
      break;

    // Don't move instructions which might have side effects, since the side
    // effects need to complete before instructions inside the loop.  Also don't
    // move instructions which might read memory, since the loop may modify
    // memory. Note that it's okay if the instruction might have undefined
    // behavior: LoopSimplify guarantees that the preheader dominates the exit
    // block.
    if (I->mayHaveSideEffects() || I->mayReadFromMemory())
      continue;

    // Skip debug info intrinsics.
    if (isa<DbgInfoIntrinsic>(I))
      continue;

    // Skip landingpad instructions.
    if (isa<LandingPadInst>(I))
      continue;

    // Don't sink alloca: we never want to sink static alloca's out of the
    // entry block, and correctly sinking dynamic alloca's requires
    // checks for stacksave/stackrestore intrinsics.
    // FIXME: Refactor this check somehow?
    if (isa<AllocaInst>(I))
      continue;

    // Determine if there is a use in or before the loop (direct or
    // otherwise).
    bool UsedInLoop = false;
    for (Use &U : I->uses()) {
      Instruction *User = cast<Instruction>(U.getUser());
      BasicBlock *UseBB = User->getParent();
      if (PHINode *P = dyn_cast<PHINode>(User)) {
        unsigned i =
          PHINode::getIncomingValueNumForOperand(U.getOperandNo());
        UseBB = P->getIncomingBlock(i);
      }
      if (UseBB == Preheader || L->contains(UseBB)) {
        UsedInLoop = true;
        break;
      }
    }

    // If there is, the def must remain in the preheader.
    if (UsedInLoop)
      continue;

    // Otherwise, sink it to the exit block.
    Instruction *ToMove = I;
    bool Done = false;

    if (I != Preheader->begin()) {
      // Skip debug info intrinsics.
      do {
        --I;
      } while (isa<DbgInfoIntrinsic>(I) && I != Preheader->begin());

      if (isa<DbgInfoIntrinsic>(I) && I == Preheader->begin())
        Done = true;
    } else {
      Done = true;
    }

    ToMove->moveBefore(InsertPt);
    if (Done) break;
    InsertPt = ToMove;
  }
}

//===----------------------------------------------------------------------===//
//  IndVarSimplify driver. Manage several subpasses of IV simplification.
//===----------------------------------------------------------------------===//

bool IndVarSimplify::runOnLoop(Loop *L, LPPassManager &LPM) {
  if (skipOptnoneFunction(L))
    return false;

  // If LoopSimplify form is not available, stay out of trouble. Some notes:
  //  - LSR currently only supports LoopSimplify-form loops. Indvars'
  //    canonicalization can be a pessimization without LSR to "clean up"
  //    afterwards.
  //  - We depend on having a preheader; in particular,
  //    Loop::getCanonicalInductionVariable only supports loops with preheaders,
  //    and we're in trouble if we can't find the induction variable even when
  //    we've manually inserted one.
  if (!L->isLoopSimplifyForm())
    return false;

  LI = &getAnalysis<LoopInfo>();
  SE = &getAnalysis<ScalarEvolution>();
  DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
  DataLayoutPass *DLP = getAnalysisIfAvailable<DataLayoutPass>();
  DL = DLP ? &DLP->getDataLayout() : nullptr;
  TLI = getAnalysisIfAvailable<TargetLibraryInfo>();

  DeadInsts.clear();
  Changed = false;

  // If there are any floating-point recurrences, attempt to
  // transform them to use integer recurrences.
  RewriteNonIntegerIVs(L);

  const SCEV *BackedgeTakenCount = SE->getBackedgeTakenCount(L);

  // Create a rewriter object which we'll use to transform the code with.
  SCEVExpander Rewriter(*SE, "indvars");
#ifndef NDEBUG
  Rewriter.setDebugType(DEBUG_TYPE);
#endif

  // Eliminate redundant IV users.
  //
  // Simplification works best when run before other consumers of SCEV. We
  // attempt to avoid evaluating SCEVs for sign/zero extend operations until
  // other expressions involving loop IVs have been evaluated. This helps SCEV
  // set no-wrap flags before normalizing sign/zero extension.
  Rewriter.disableCanonicalMode();
  SimplifyAndExtend(L, Rewriter, LPM);

  // Check to see if this loop has a computable loop-invariant execution count.
  // If so, this means that we can compute the final value of any expressions
  // that are recurrent in the loop, and substitute the exit values from the
  // loop into any instructions outside of the loop that use the final values of
  // the current expressions.
  //
  if (!isa<SCEVCouldNotCompute>(BackedgeTakenCount))
    RewriteLoopExitValues(L, Rewriter);

  // Eliminate redundant IV cycles.
  NumElimIV += Rewriter.replaceCongruentIVs(L, DT, DeadInsts);

  // If we have a trip count expression, rewrite the loop's exit condition
  // using it.  We can currently only handle loops with a single exit.
  if (canExpandBackedgeTakenCount(L, SE) && needsLFTR(L, DT)) {
    PHINode *IndVar = FindLoopCounter(L, BackedgeTakenCount, SE, DT, DL);
    if (IndVar) {
      // Check preconditions for proper SCEVExpander operation. SCEV does not
      // express SCEVExpander's dependencies, such as LoopSimplify. Instead any
      // pass that uses the SCEVExpander must do it. This does not work well for
      // loop passes because SCEVExpander makes assumptions about all loops,
      // while LoopPassManager only forces the current loop to be simplified.
      //
      // FIXME: SCEV expansion has no way to bail out, so the caller must
      // explicitly check any assumptions made by SCEV. Brittle.
      const SCEVAddRecExpr *AR = dyn_cast<SCEVAddRecExpr>(BackedgeTakenCount);
      if (!AR || AR->getLoop()->getLoopPreheader())
        (void)LinearFunctionTestReplace(L, BackedgeTakenCount, IndVar,
                                        Rewriter);
    }
  }
  // Clear the rewriter cache, because values that are in the rewriter's cache
  // can be deleted in the loop below, causing the AssertingVH in the cache to
  // trigger.
  Rewriter.clear();

  // Now that we're done iterating through lists, clean up any instructions
  // which are now dead.
  while (!DeadInsts.empty())
    if (Instruction *Inst =
          dyn_cast_or_null<Instruction>(&*DeadInsts.pop_back_val()))
      RecursivelyDeleteTriviallyDeadInstructions(Inst, TLI);

  // The Rewriter may not be used from this point on.

  // Loop-invariant instructions in the preheader that aren't used in the
  // loop may be sunk below the loop to reduce register pressure.
  SinkUnusedInvariants(L);

  // Clean up dead instructions.
  Changed |= DeleteDeadPHIs(L->getHeader(), TLI);
  // Check a post-condition.
  assert(L->isLCSSAForm(*DT) &&
         "Indvars did not leave the loop in lcssa form!");

  // Verify that LFTR, and any other change have not interfered with SCEV's
  // ability to compute trip count.
#ifndef NDEBUG
  if (VerifyIndvars && !isa<SCEVCouldNotCompute>(BackedgeTakenCount)) {
    SE->forgetLoop(L);
    const SCEV *NewBECount = SE->getBackedgeTakenCount(L);
    if (SE->getTypeSizeInBits(BackedgeTakenCount->getType()) <
        SE->getTypeSizeInBits(NewBECount->getType()))
      NewBECount = SE->getTruncateOrNoop(NewBECount,
                                         BackedgeTakenCount->getType());
    else
      BackedgeTakenCount = SE->getTruncateOrNoop(BackedgeTakenCount,
                                                 NewBECount->getType());
    assert(BackedgeTakenCount == NewBECount && "indvars must preserve SCEV");
  }
#endif

  return Changed;
}