llvm.org GIT mirror llvm / release_39 lib / Target / X86 / X86OptimizeLEAs.cpp
release_39

Tree @release_39 (Download .tar.gz)

X86OptimizeLEAs.cpp @release_39raw · 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
//===-- X86OptimizeLEAs.cpp - optimize usage of LEA instructions ----------===//
//
//                     The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file defines the pass that performs some optimizations with LEA
// instructions in order to improve performance and code size.
// Currently, it does two things:
// 1) If there are two LEA instructions calculating addresses which only differ
//    by displacement inside a basic block, one of them is removed.
// 2) Address calculations in load and store instructions are replaced by
//    existing LEA def registers where possible.
//
//===----------------------------------------------------------------------===//

#include "X86.h"
#include "X86InstrInfo.h"
#include "X86Subtarget.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/CodeGen/LiveVariables.h"
#include "llvm/CodeGen/MachineFunctionPass.h"
#include "llvm/CodeGen/MachineInstrBuilder.h"
#include "llvm/CodeGen/MachineOperand.h"
#include "llvm/CodeGen/MachineRegisterInfo.h"
#include "llvm/CodeGen/Passes.h"
#include "llvm/IR/Function.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/raw_ostream.h"
#include "llvm/Target/TargetInstrInfo.h"

using namespace llvm;

#define DEBUG_TYPE "x86-optimize-LEAs"

static cl::opt<bool>
    DisableX86LEAOpt("disable-x86-lea-opt", cl::Hidden,
                     cl::desc("X86: Disable LEA optimizations."),
                     cl::init(false));

STATISTIC(NumSubstLEAs, "Number of LEA instruction substitutions");
STATISTIC(NumRedundantLEAs, "Number of redundant LEA instructions removed");

class MemOpKey;

/// \brief Returns a hash table key based on memory operands of \p MI. The
/// number of the first memory operand of \p MI is specified through \p N.
static inline MemOpKey getMemOpKey(const MachineInstr &MI, unsigned N);

/// \brief Returns true if two machine operands are identical and they are not
/// physical registers.
static inline bool isIdenticalOp(const MachineOperand &MO1,
                                 const MachineOperand &MO2);

/// \brief Returns true if two address displacement operands are of the same
/// type and use the same symbol/index/address regardless of the offset.
static bool isSimilarDispOp(const MachineOperand &MO1,
                            const MachineOperand &MO2);

/// \brief Returns true if the instruction is LEA.
static inline bool isLEA(const MachineInstr &MI);

/// A key based on instruction's memory operands.
class MemOpKey {
public:
  MemOpKey(const MachineOperand *Base, const MachineOperand *Scale,
           const MachineOperand *Index, const MachineOperand *Segment,
           const MachineOperand *Disp)
      : Disp(Disp) {
    Operands[0] = Base;
    Operands[1] = Scale;
    Operands[2] = Index;
    Operands[3] = Segment;
  }

  bool operator==(const MemOpKey &Other) const {
    // Addresses' bases, scales, indices and segments must be identical.
    for (int i = 0; i < 4; ++i)
      if (!isIdenticalOp(*Operands[i], *Other.Operands[i]))
        return false;

    // Addresses' displacements don't have to be exactly the same. It only
    // matters that they use the same symbol/index/address. Immediates' or
    // offsets' differences will be taken care of during instruction
    // substitution.
    return isSimilarDispOp(*Disp, *Other.Disp);
  }

  // Address' base, scale, index and segment operands.
  const MachineOperand *Operands[4];

  // Address' displacement operand.
  const MachineOperand *Disp;
};

/// Provide DenseMapInfo for MemOpKey.
namespace llvm {
template <> struct DenseMapInfo<MemOpKey> {
  typedef DenseMapInfo<const MachineOperand *> PtrInfo;

  static inline MemOpKey getEmptyKey() {
    return MemOpKey(PtrInfo::getEmptyKey(), PtrInfo::getEmptyKey(),
                    PtrInfo::getEmptyKey(), PtrInfo::getEmptyKey(),
                    PtrInfo::getEmptyKey());
  }

  static inline MemOpKey getTombstoneKey() {
    return MemOpKey(PtrInfo::getTombstoneKey(), PtrInfo::getTombstoneKey(),
                    PtrInfo::getTombstoneKey(), PtrInfo::getTombstoneKey(),
                    PtrInfo::getTombstoneKey());
  }

  static unsigned getHashValue(const MemOpKey &Val) {
    // Checking any field of MemOpKey is enough to determine if the key is
    // empty or tombstone.
    assert(Val.Disp != PtrInfo::getEmptyKey() && "Cannot hash the empty key");
    assert(Val.Disp != PtrInfo::getTombstoneKey() &&
           "Cannot hash the tombstone key");

    hash_code Hash = hash_combine(*Val.Operands[0], *Val.Operands[1],
                                  *Val.Operands[2], *Val.Operands[3]);

    // If the address displacement is an immediate, it should not affect the
    // hash so that memory operands which differ only be immediate displacement
    // would have the same hash. If the address displacement is something else,
    // we should reflect symbol/index/address in the hash.
    switch (Val.Disp->getType()) {
    case MachineOperand::MO_Immediate:
      break;
    case MachineOperand::MO_ConstantPoolIndex:
    case MachineOperand::MO_JumpTableIndex:
      Hash = hash_combine(Hash, Val.Disp->getIndex());
      break;
    case MachineOperand::MO_ExternalSymbol:
      Hash = hash_combine(Hash, Val.Disp->getSymbolName());
      break;
    case MachineOperand::MO_GlobalAddress:
      Hash = hash_combine(Hash, Val.Disp->getGlobal());
      break;
    case MachineOperand::MO_BlockAddress:
      Hash = hash_combine(Hash, Val.Disp->getBlockAddress());
      break;
    case MachineOperand::MO_MCSymbol:
      Hash = hash_combine(Hash, Val.Disp->getMCSymbol());
      break;
    case MachineOperand::MO_MachineBasicBlock:
      Hash = hash_combine(Hash, Val.Disp->getMBB());
      break;
    default:
      llvm_unreachable("Invalid address displacement operand");
    }

    return (unsigned)Hash;
  }

  static bool isEqual(const MemOpKey &LHS, const MemOpKey &RHS) {
    // Checking any field of MemOpKey is enough to determine if the key is
    // empty or tombstone.
    if (RHS.Disp == PtrInfo::getEmptyKey())
      return LHS.Disp == PtrInfo::getEmptyKey();
    if (RHS.Disp == PtrInfo::getTombstoneKey())
      return LHS.Disp == PtrInfo::getTombstoneKey();
    return LHS == RHS;
  }
};
}

static inline MemOpKey getMemOpKey(const MachineInstr &MI, unsigned N) {
  assert((isLEA(MI) || MI.mayLoadOrStore()) &&
         "The instruction must be a LEA, a load or a store");
  return MemOpKey(&MI.getOperand(N + X86::AddrBaseReg),
                  &MI.getOperand(N + X86::AddrScaleAmt),
                  &MI.getOperand(N + X86::AddrIndexReg),
                  &MI.getOperand(N + X86::AddrSegmentReg),
                  &MI.getOperand(N + X86::AddrDisp));
}

static inline bool isIdenticalOp(const MachineOperand &MO1,
                                 const MachineOperand &MO2) {
  return MO1.isIdenticalTo(MO2) &&
         (!MO1.isReg() ||
          !TargetRegisterInfo::isPhysicalRegister(MO1.getReg()));
}

#ifndef NDEBUG
static bool isValidDispOp(const MachineOperand &MO) {
  return MO.isImm() || MO.isCPI() || MO.isJTI() || MO.isSymbol() ||
         MO.isGlobal() || MO.isBlockAddress() || MO.isMCSymbol() || MO.isMBB();
}
#endif

static bool isSimilarDispOp(const MachineOperand &MO1,
                            const MachineOperand &MO2) {
  assert(isValidDispOp(MO1) && isValidDispOp(MO2) &&
         "Address displacement operand is not valid");
  return (MO1.isImm() && MO2.isImm()) ||
         (MO1.isCPI() && MO2.isCPI() && MO1.getIndex() == MO2.getIndex()) ||
         (MO1.isJTI() && MO2.isJTI() && MO1.getIndex() == MO2.getIndex()) ||
         (MO1.isSymbol() && MO2.isSymbol() &&
          MO1.getSymbolName() == MO2.getSymbolName()) ||
         (MO1.isGlobal() && MO2.isGlobal() &&
          MO1.getGlobal() == MO2.getGlobal()) ||
         (MO1.isBlockAddress() && MO2.isBlockAddress() &&
          MO1.getBlockAddress() == MO2.getBlockAddress()) ||
         (MO1.isMCSymbol() && MO2.isMCSymbol() &&
          MO1.getMCSymbol() == MO2.getMCSymbol()) ||
         (MO1.isMBB() && MO2.isMBB() && MO1.getMBB() == MO2.getMBB());
}

static inline bool isLEA(const MachineInstr &MI) {
  unsigned Opcode = MI.getOpcode();
  return Opcode == X86::LEA16r || Opcode == X86::LEA32r ||
         Opcode == X86::LEA64r || Opcode == X86::LEA64_32r;
}

namespace {
class OptimizeLEAPass : public MachineFunctionPass {
public:
  OptimizeLEAPass() : MachineFunctionPass(ID) {}

  const char *getPassName() const override { return "X86 LEA Optimize"; }

  /// \brief Loop over all of the basic blocks, replacing address
  /// calculations in load and store instructions, if it's already
  /// been calculated by LEA. Also, remove redundant LEAs.
  bool runOnMachineFunction(MachineFunction &MF) override;

private:
  typedef DenseMap<MemOpKey, SmallVector<MachineInstr *, 16>> MemOpMap;

  /// \brief Returns a distance between two instructions inside one basic block.
  /// Negative result means, that instructions occur in reverse order.
  int calcInstrDist(const MachineInstr &First, const MachineInstr &Last);

  /// \brief Choose the best \p LEA instruction from the \p List to replace
  /// address calculation in \p MI instruction. Return the address displacement
  /// and the distance between \p MI and the choosen \p BestLEA in
  /// \p AddrDispShift and \p Dist.
  bool chooseBestLEA(const SmallVectorImpl<MachineInstr *> &List,
                     const MachineInstr &MI, MachineInstr *&BestLEA,
                     int64_t &AddrDispShift, int &Dist);

  /// \brief Returns the difference between addresses' displacements of \p MI1
  /// and \p MI2. The numbers of the first memory operands for the instructions
  /// are specified through \p N1 and \p N2.
  int64_t getAddrDispShift(const MachineInstr &MI1, unsigned N1,
                           const MachineInstr &MI2, unsigned N2) const;

  /// \brief Returns true if the \p Last LEA instruction can be replaced by the
  /// \p First. The difference between displacements of the addresses calculated
  /// by these LEAs is returned in \p AddrDispShift. It'll be used for proper
  /// replacement of the \p Last LEA's uses with the \p First's def register.
  bool isReplaceable(const MachineInstr &First, const MachineInstr &Last,
                     int64_t &AddrDispShift) const;

  /// \brief Find all LEA instructions in the basic block. Also, assign position
  /// numbers to all instructions in the basic block to speed up calculation of
  /// distance between them.
  void findLEAs(const MachineBasicBlock &MBB, MemOpMap &LEAs);

  /// \brief Removes redundant address calculations.
  bool removeRedundantAddrCalc(MemOpMap &LEAs);

  /// \brief Removes LEAs which calculate similar addresses.
  bool removeRedundantLEAs(MemOpMap &LEAs);

  DenseMap<const MachineInstr *, unsigned> InstrPos;

  MachineRegisterInfo *MRI;
  const X86InstrInfo *TII;
  const X86RegisterInfo *TRI;

  static char ID;
};
char OptimizeLEAPass::ID = 0;
}

FunctionPass *llvm::createX86OptimizeLEAs() { return new OptimizeLEAPass(); }

int OptimizeLEAPass::calcInstrDist(const MachineInstr &First,
                                   const MachineInstr &Last) {
  // Both instructions must be in the same basic block and they must be
  // presented in InstrPos.
  assert(Last.getParent() == First.getParent() &&
         "Instructions are in different basic blocks");
  assert(InstrPos.find(&First) != InstrPos.end() &&
         InstrPos.find(&Last) != InstrPos.end() &&
         "Instructions' positions are undefined");

  return InstrPos[&Last] - InstrPos[&First];
}

// Find the best LEA instruction in the List to replace address recalculation in
// MI. Such LEA must meet these requirements:
// 1) The address calculated by the LEA differs only by the displacement from
//    the address used in MI.
// 2) The register class of the definition of the LEA is compatible with the
//    register class of the address base register of MI.
// 3) Displacement of the new memory operand should fit in 1 byte if possible.
// 4) The LEA should be as close to MI as possible, and prior to it if
//    possible.
bool OptimizeLEAPass::chooseBestLEA(const SmallVectorImpl<MachineInstr *> &List,
                                    const MachineInstr &MI,
                                    MachineInstr *&BestLEA,
                                    int64_t &AddrDispShift, int &Dist) {
  const MachineFunction *MF = MI.getParent()->getParent();
  const MCInstrDesc &Desc = MI.getDesc();
  int MemOpNo = X86II::getMemoryOperandNo(Desc.TSFlags) +
                X86II::getOperandBias(Desc);

  BestLEA = nullptr;

  // Loop over all LEA instructions.
  for (auto DefMI : List) {
    // Get new address displacement.
    int64_t AddrDispShiftTemp = getAddrDispShift(MI, MemOpNo, *DefMI, 1);

    // Make sure address displacement fits 4 bytes.
    if (!isInt<32>(AddrDispShiftTemp))
      continue;

    // Check that LEA def register can be used as MI address base. Some
    // instructions can use a limited set of registers as address base, for
    // example MOV8mr_NOREX. We could constrain the register class of the LEA
    // def to suit MI, however since this case is very rare and hard to
    // reproduce in a test it's just more reliable to skip the LEA.
    if (TII->getRegClass(Desc, MemOpNo + X86::AddrBaseReg, TRI, *MF) !=
        MRI->getRegClass(DefMI->getOperand(0).getReg()))
      continue;

    // Choose the closest LEA instruction from the list, prior to MI if
    // possible. Note that we took into account resulting address displacement
    // as well. Also note that the list is sorted by the order in which the LEAs
    // occur, so the break condition is pretty simple.
    int DistTemp = calcInstrDist(*DefMI, MI);
    assert(DistTemp != 0 &&
           "The distance between two different instructions cannot be zero");
    if (DistTemp > 0 || BestLEA == nullptr) {
      // Do not update return LEA, if the current one provides a displacement
      // which fits in 1 byte, while the new candidate does not.
      if (BestLEA != nullptr && !isInt<8>(AddrDispShiftTemp) &&
          isInt<8>(AddrDispShift))
        continue;

      BestLEA = DefMI;
      AddrDispShift = AddrDispShiftTemp;
      Dist = DistTemp;
    }

    // FIXME: Maybe we should not always stop at the first LEA after MI.
    if (DistTemp < 0)
      break;
  }

  return BestLEA != nullptr;
}

// Get the difference between the addresses' displacements of the two
// instructions \p MI1 and \p MI2. The numbers of the first memory operands are
// passed through \p N1 and \p N2.
int64_t OptimizeLEAPass::getAddrDispShift(const MachineInstr &MI1, unsigned N1,
                                          const MachineInstr &MI2,
                                          unsigned N2) const {
  const MachineOperand &Op1 = MI1.getOperand(N1 + X86::AddrDisp);
  const MachineOperand &Op2 = MI2.getOperand(N2 + X86::AddrDisp);

  assert(isSimilarDispOp(Op1, Op2) &&
         "Address displacement operands are not compatible");

  // After the assert above we can be sure that both operands are of the same
  // valid type and use the same symbol/index/address, thus displacement shift
  // calculation is rather simple.
  if (Op1.isJTI())
    return 0;
  return Op1.isImm() ? Op1.getImm() - Op2.getImm()
                     : Op1.getOffset() - Op2.getOffset();
}

// Check that the Last LEA can be replaced by the First LEA. To be so,
// these requirements must be met:
// 1) Addresses calculated by LEAs differ only by displacement.
// 2) Def registers of LEAs belong to the same class.
// 3) All uses of the Last LEA def register are replaceable, thus the
//    register is used only as address base.
bool OptimizeLEAPass::isReplaceable(const MachineInstr &First,
                                    const MachineInstr &Last,
                                    int64_t &AddrDispShift) const {
  assert(isLEA(First) && isLEA(Last) &&
         "The function works only with LEA instructions");

  // Get new address displacement.
  AddrDispShift = getAddrDispShift(Last, 1, First, 1);

  // Make sure that LEA def registers belong to the same class. There may be
  // instructions (like MOV8mr_NOREX) which allow a limited set of registers to
  // be used as their operands, so we must be sure that replacing one LEA
  // with another won't lead to putting a wrong register in the instruction.
  if (MRI->getRegClass(First.getOperand(0).getReg()) !=
      MRI->getRegClass(Last.getOperand(0).getReg()))
    return false;

  // Loop over all uses of the Last LEA to check that its def register is
  // used only as address base for memory accesses. If so, it can be
  // replaced, otherwise - no.
  for (auto &MO : MRI->use_operands(Last.getOperand(0).getReg())) {
    MachineInstr &MI = *MO.getParent();

    // Get the number of the first memory operand.
    const MCInstrDesc &Desc = MI.getDesc();
    int MemOpNo = X86II::getMemoryOperandNo(Desc.TSFlags);

    // If the use instruction has no memory operand - the LEA is not
    // replaceable.
    if (MemOpNo < 0)
      return false;

    MemOpNo += X86II::getOperandBias(Desc);

    // If the address base of the use instruction is not the LEA def register -
    // the LEA is not replaceable.
    if (!isIdenticalOp(MI.getOperand(MemOpNo + X86::AddrBaseReg), MO))
      return false;

    // If the LEA def register is used as any other operand of the use
    // instruction - the LEA is not replaceable.
    for (unsigned i = 0; i < MI.getNumOperands(); i++)
      if (i != (unsigned)(MemOpNo + X86::AddrBaseReg) &&
          isIdenticalOp(MI.getOperand(i), MO))
        return false;

    // Check that the new address displacement will fit 4 bytes.
    if (MI.getOperand(MemOpNo + X86::AddrDisp).isImm() &&
        !isInt<32>(MI.getOperand(MemOpNo + X86::AddrDisp).getImm() +
                   AddrDispShift))
      return false;
  }

  return true;
}

void OptimizeLEAPass::findLEAs(const MachineBasicBlock &MBB, MemOpMap &LEAs) {
  unsigned Pos = 0;
  for (auto &MI : MBB) {
    // Assign the position number to the instruction. Note that we are going to
    // move some instructions during the optimization however there will never
    // be a need to move two instructions before any selected instruction. So to
    // avoid multiple positions' updates during moves we just increase position
    // counter by two leaving a free space for instructions which will be moved.
    InstrPos[&MI] = Pos += 2;

    if (isLEA(MI))
      LEAs[getMemOpKey(MI, 1)].push_back(const_cast<MachineInstr *>(&MI));
  }
}

// Try to find load and store instructions which recalculate addresses already
// calculated by some LEA and replace their memory operands with its def
// register.
bool OptimizeLEAPass::removeRedundantAddrCalc(MemOpMap &LEAs) {
  bool Changed = false;

  assert(!LEAs.empty());
  MachineBasicBlock *MBB = (*LEAs.begin()->second.begin())->getParent();

  // Process all instructions in basic block.
  for (auto I = MBB->begin(), E = MBB->end(); I != E;) {
    MachineInstr &MI = *I++;

    // Instruction must be load or store.
    if (!MI.mayLoadOrStore())
      continue;

    // Get the number of the first memory operand.
    const MCInstrDesc &Desc = MI.getDesc();
    int MemOpNo = X86II::getMemoryOperandNo(Desc.TSFlags);

    // If instruction has no memory operand - skip it.
    if (MemOpNo < 0)
      continue;

    MemOpNo += X86II::getOperandBias(Desc);

    // Get the best LEA instruction to replace address calculation.
    MachineInstr *DefMI;
    int64_t AddrDispShift;
    int Dist;
    if (!chooseBestLEA(LEAs[getMemOpKey(MI, MemOpNo)], MI, DefMI, AddrDispShift,
                       Dist))
      continue;

    // If LEA occurs before current instruction, we can freely replace
    // the instruction. If LEA occurs after, we can lift LEA above the
    // instruction and this way to be able to replace it. Since LEA and the
    // instruction have similar memory operands (thus, the same def
    // instructions for these operands), we can always do that, without
    // worries of using registers before their defs.
    if (Dist < 0) {
      DefMI->removeFromParent();
      MBB->insert(MachineBasicBlock::iterator(&MI), DefMI);
      InstrPos[DefMI] = InstrPos[&MI] - 1;

      // Make sure the instructions' position numbers are sane.
      assert(((InstrPos[DefMI] == 1 &&
               MachineBasicBlock::iterator(DefMI) == MBB->begin()) ||
              InstrPos[DefMI] >
                  InstrPos[&*std::prev(MachineBasicBlock::iterator(DefMI))]) &&
             "Instruction positioning is broken");
    }

    // Since we can possibly extend register lifetime, clear kill flags.
    MRI->clearKillFlags(DefMI->getOperand(0).getReg());

    ++NumSubstLEAs;
    DEBUG(dbgs() << "OptimizeLEAs: Candidate to replace: "; MI.dump(););

    // Change instruction operands.
    MI.getOperand(MemOpNo + X86::AddrBaseReg)
        .ChangeToRegister(DefMI->getOperand(0).getReg(), false);
    MI.getOperand(MemOpNo + X86::AddrScaleAmt).ChangeToImmediate(1);
    MI.getOperand(MemOpNo + X86::AddrIndexReg)
        .ChangeToRegister(X86::NoRegister, false);
    MI.getOperand(MemOpNo + X86::AddrDisp).ChangeToImmediate(AddrDispShift);
    MI.getOperand(MemOpNo + X86::AddrSegmentReg)
        .ChangeToRegister(X86::NoRegister, false);

    DEBUG(dbgs() << "OptimizeLEAs: Replaced by: "; MI.dump(););

    Changed = true;
  }

  return Changed;
}

// Try to find similar LEAs in the list and replace one with another.
bool OptimizeLEAPass::removeRedundantLEAs(MemOpMap &LEAs) {
  bool Changed = false;

  // Loop over all entries in the table.
  for (auto &E : LEAs) {
    auto &List = E.second;

    // Loop over all LEA pairs.
    auto I1 = List.begin();
    while (I1 != List.end()) {
      MachineInstr &First = **I1;
      auto I2 = std::next(I1);
      while (I2 != List.end()) {
        MachineInstr &Last = **I2;
        int64_t AddrDispShift;

        // LEAs should be in occurence order in the list, so we can freely
        // replace later LEAs with earlier ones.
        assert(calcInstrDist(First, Last) > 0 &&
               "LEAs must be in occurence order in the list");

        // Check that the Last LEA instruction can be replaced by the First.
        if (!isReplaceable(First, Last, AddrDispShift)) {
          ++I2;
          continue;
        }

        // Loop over all uses of the Last LEA and update their operands. Note
        // that the correctness of this has already been checked in the
        // isReplaceable function.
        for (auto UI = MRI->use_begin(Last.getOperand(0).getReg()),
                  UE = MRI->use_end();
             UI != UE;) {
          MachineOperand &MO = *UI++;
          MachineInstr &MI = *MO.getParent();

          // Get the number of the first memory operand.
          const MCInstrDesc &Desc = MI.getDesc();
          int MemOpNo =
              X86II::getMemoryOperandNo(Desc.TSFlags) +
              X86II::getOperandBias(Desc);

          // Update address base.
          MO.setReg(First.getOperand(0).getReg());

          // Update address disp.
          MachineOperand &Op = MI.getOperand(MemOpNo + X86::AddrDisp);
          if (Op.isImm())
            Op.setImm(Op.getImm() + AddrDispShift);
          else if (!Op.isJTI())
            Op.setOffset(Op.getOffset() + AddrDispShift);
        }

        // Since we can possibly extend register lifetime, clear kill flags.
        MRI->clearKillFlags(First.getOperand(0).getReg());

        ++NumRedundantLEAs;
        DEBUG(dbgs() << "OptimizeLEAs: Remove redundant LEA: "; Last.dump(););

        // By this moment, all of the Last LEA's uses must be replaced. So we
        // can freely remove it.
        assert(MRI->use_empty(Last.getOperand(0).getReg()) &&
               "The LEA's def register must have no uses");
        Last.eraseFromParent();

        // Erase removed LEA from the list.
        I2 = List.erase(I2);

        Changed = true;
      }
      ++I1;
    }
  }

  return Changed;
}

bool OptimizeLEAPass::runOnMachineFunction(MachineFunction &MF) {
  bool Changed = false;

  if (DisableX86LEAOpt || skipFunction(*MF.getFunction()))
    return false;

  MRI = &MF.getRegInfo();
  TII = MF.getSubtarget<X86Subtarget>().getInstrInfo();
  TRI = MF.getSubtarget<X86Subtarget>().getRegisterInfo();

  // Process all basic blocks.
  for (auto &MBB : MF) {
    MemOpMap LEAs;
    InstrPos.clear();

    // Find all LEA instructions in basic block.
    findLEAs(MBB, LEAs);

    // If current basic block has no LEAs, move on to the next one.
    if (LEAs.empty())
      continue;

    // Remove redundant LEA instructions.
    Changed |= removeRedundantLEAs(LEAs);

    // Remove redundant address calculations. Do it only for -Os/-Oz since only
    // a code size gain is expected from this part of the pass.
    if (MF.getFunction()->optForSize())
      Changed |= removeRedundantAddrCalc(LEAs);
  }

  return Changed;
}