llvm.org GIT mirror llvm / release_35@215010 lib / Target / AArch64 / AArch64LoadStoreOptimizer.cpp
release_35@215010

Tree @release_35@215010 (Download .tar.gz)

AArch64LoadStoreOptimizer.cpp @release_35@215010raw · 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
//=- AArch64LoadStoreOptimizer.cpp - AArch64 load/store opt. pass -*- C++ -*-=//
//
//                     The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file contains a pass that performs load / store related peephole
// optimizations. This pass should be run after register allocation.
//
//===----------------------------------------------------------------------===//

#include "AArch64InstrInfo.h"
#include "MCTargetDesc/AArch64AddressingModes.h"
#include "llvm/ADT/BitVector.h"
#include "llvm/CodeGen/MachineBasicBlock.h"
#include "llvm/CodeGen/MachineFunctionPass.h"
#include "llvm/CodeGen/MachineInstr.h"
#include "llvm/CodeGen/MachineInstrBuilder.h"
#include "llvm/Target/TargetInstrInfo.h"
#include "llvm/Target/TargetMachine.h"
#include "llvm/Target/TargetRegisterInfo.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Debug.h"
#include "llvm/Support/ErrorHandling.h"
#include "llvm/Support/raw_ostream.h"
#include "llvm/ADT/Statistic.h"
using namespace llvm;

#define DEBUG_TYPE "aarch64-ldst-opt"

/// AArch64AllocLoadStoreOpt - Post-register allocation pass to combine
/// load / store instructions to form ldp / stp instructions.

STATISTIC(NumPairCreated, "Number of load/store pair instructions generated");
STATISTIC(NumPostFolded, "Number of post-index updates folded");
STATISTIC(NumPreFolded, "Number of pre-index updates folded");
STATISTIC(NumUnscaledPairCreated,
          "Number of load/store from unscaled generated");

static cl::opt<unsigned> ScanLimit("aarch64-load-store-scan-limit",
                                   cl::init(20), cl::Hidden);

// Place holder while testing unscaled load/store combining
static cl::opt<bool> EnableAArch64UnscaledMemOp(
    "aarch64-unscaled-mem-op", cl::Hidden,
    cl::desc("Allow AArch64 unscaled load/store combining"), cl::init(true));

namespace {
struct AArch64LoadStoreOpt : public MachineFunctionPass {
  static char ID;
  AArch64LoadStoreOpt() : MachineFunctionPass(ID) {}

  const AArch64InstrInfo *TII;
  const TargetRegisterInfo *TRI;

  // Scan the instructions looking for a load/store that can be combined
  // with the current instruction into a load/store pair.
  // Return the matching instruction if one is found, else MBB->end().
  // If a matching instruction is found, MergeForward is set to true if the
  // merge is to remove the first instruction and replace the second with
  // a pair-wise insn, and false if the reverse is true.
  MachineBasicBlock::iterator findMatchingInsn(MachineBasicBlock::iterator I,
                                               bool &MergeForward,
                                               unsigned Limit);
  // Merge the two instructions indicated into a single pair-wise instruction.
  // If MergeForward is true, erase the first instruction and fold its
  // operation into the second. If false, the reverse. Return the instruction
  // following the first instruction (which may change during processing).
  MachineBasicBlock::iterator
  mergePairedInsns(MachineBasicBlock::iterator I,
                   MachineBasicBlock::iterator Paired, bool MergeForward);

  // Scan the instruction list to find a base register update that can
  // be combined with the current instruction (a load or store) using
  // pre or post indexed addressing with writeback. Scan forwards.
  MachineBasicBlock::iterator
  findMatchingUpdateInsnForward(MachineBasicBlock::iterator I, unsigned Limit,
                                int Value);

  // Scan the instruction list to find a base register update that can
  // be combined with the current instruction (a load or store) using
  // pre or post indexed addressing with writeback. Scan backwards.
  MachineBasicBlock::iterator
  findMatchingUpdateInsnBackward(MachineBasicBlock::iterator I, unsigned Limit);

  // Merge a pre-index base register update into a ld/st instruction.
  MachineBasicBlock::iterator
  mergePreIdxUpdateInsn(MachineBasicBlock::iterator I,
                        MachineBasicBlock::iterator Update);

  // Merge a post-index base register update into a ld/st instruction.
  MachineBasicBlock::iterator
  mergePostIdxUpdateInsn(MachineBasicBlock::iterator I,
                         MachineBasicBlock::iterator Update);

  bool optimizeBlock(MachineBasicBlock &MBB);

  bool runOnMachineFunction(MachineFunction &Fn) override;

  const char *getPassName() const override {
    return "AArch64 load / store optimization pass";
  }

private:
  int getMemSize(MachineInstr *MemMI);
};
char AArch64LoadStoreOpt::ID = 0;
}

static bool isUnscaledLdst(unsigned Opc) {
  switch (Opc) {
  default:
    return false;
  case AArch64::STURSi:
    return true;
  case AArch64::STURDi:
    return true;
  case AArch64::STURQi:
    return true;
  case AArch64::STURWi:
    return true;
  case AArch64::STURXi:
    return true;
  case AArch64::LDURSi:
    return true;
  case AArch64::LDURDi:
    return true;
  case AArch64::LDURQi:
    return true;
  case AArch64::LDURWi:
    return true;
  case AArch64::LDURXi:
    return true;
  }
}

// Size in bytes of the data moved by an unscaled load or store
int AArch64LoadStoreOpt::getMemSize(MachineInstr *MemMI) {
  switch (MemMI->getOpcode()) {
  default:
    llvm_unreachable("Opcode has unknown size!");
  case AArch64::STRSui:
  case AArch64::STURSi:
    return 4;
  case AArch64::STRDui:
  case AArch64::STURDi:
    return 8;
  case AArch64::STRQui:
  case AArch64::STURQi:
    return 16;
  case AArch64::STRWui:
  case AArch64::STURWi:
    return 4;
  case AArch64::STRXui:
  case AArch64::STURXi:
    return 8;
  case AArch64::LDRSui:
  case AArch64::LDURSi:
    return 4;
  case AArch64::LDRDui:
  case AArch64::LDURDi:
    return 8;
  case AArch64::LDRQui:
  case AArch64::LDURQi:
    return 16;
  case AArch64::LDRWui:
  case AArch64::LDURWi:
    return 4;
  case AArch64::LDRXui:
  case AArch64::LDURXi:
    return 8;
  }
}

static unsigned getMatchingPairOpcode(unsigned Opc) {
  switch (Opc) {
  default:
    llvm_unreachable("Opcode has no pairwise equivalent!");
  case AArch64::STRSui:
  case AArch64::STURSi:
    return AArch64::STPSi;
  case AArch64::STRDui:
  case AArch64::STURDi:
    return AArch64::STPDi;
  case AArch64::STRQui:
  case AArch64::STURQi:
    return AArch64::STPQi;
  case AArch64::STRWui:
  case AArch64::STURWi:
    return AArch64::STPWi;
  case AArch64::STRXui:
  case AArch64::STURXi:
    return AArch64::STPXi;
  case AArch64::LDRSui:
  case AArch64::LDURSi:
    return AArch64::LDPSi;
  case AArch64::LDRDui:
  case AArch64::LDURDi:
    return AArch64::LDPDi;
  case AArch64::LDRQui:
  case AArch64::LDURQi:
    return AArch64::LDPQi;
  case AArch64::LDRWui:
  case AArch64::LDURWi:
    return AArch64::LDPWi;
  case AArch64::LDRXui:
  case AArch64::LDURXi:
    return AArch64::LDPXi;
  }
}

static unsigned getPreIndexedOpcode(unsigned Opc) {
  switch (Opc) {
  default:
    llvm_unreachable("Opcode has no pre-indexed equivalent!");
  case AArch64::STRSui:
    return AArch64::STRSpre;
  case AArch64::STRDui:
    return AArch64::STRDpre;
  case AArch64::STRQui:
    return AArch64::STRQpre;
  case AArch64::STRWui:
    return AArch64::STRWpre;
  case AArch64::STRXui:
    return AArch64::STRXpre;
  case AArch64::LDRSui:
    return AArch64::LDRSpre;
  case AArch64::LDRDui:
    return AArch64::LDRDpre;
  case AArch64::LDRQui:
    return AArch64::LDRQpre;
  case AArch64::LDRWui:
    return AArch64::LDRWpre;
  case AArch64::LDRXui:
    return AArch64::LDRXpre;
  }
}

static unsigned getPostIndexedOpcode(unsigned Opc) {
  switch (Opc) {
  default:
    llvm_unreachable("Opcode has no post-indexed wise equivalent!");
  case AArch64::STRSui:
    return AArch64::STRSpost;
  case AArch64::STRDui:
    return AArch64::STRDpost;
  case AArch64::STRQui:
    return AArch64::STRQpost;
  case AArch64::STRWui:
    return AArch64::STRWpost;
  case AArch64::STRXui:
    return AArch64::STRXpost;
  case AArch64::LDRSui:
    return AArch64::LDRSpost;
  case AArch64::LDRDui:
    return AArch64::LDRDpost;
  case AArch64::LDRQui:
    return AArch64::LDRQpost;
  case AArch64::LDRWui:
    return AArch64::LDRWpost;
  case AArch64::LDRXui:
    return AArch64::LDRXpost;
  }
}

MachineBasicBlock::iterator
AArch64LoadStoreOpt::mergePairedInsns(MachineBasicBlock::iterator I,
                                      MachineBasicBlock::iterator Paired,
                                      bool MergeForward) {
  MachineBasicBlock::iterator NextI = I;
  ++NextI;
  // If NextI is the second of the two instructions to be merged, we need
  // to skip one further. Either way we merge will invalidate the iterator,
  // and we don't need to scan the new instruction, as it's a pairwise
  // instruction, which we're not considering for further action anyway.
  if (NextI == Paired)
    ++NextI;

  bool IsUnscaled = isUnscaledLdst(I->getOpcode());
  int OffsetStride =
      IsUnscaled && EnableAArch64UnscaledMemOp ? getMemSize(I) : 1;

  unsigned NewOpc = getMatchingPairOpcode(I->getOpcode());
  // Insert our new paired instruction after whichever of the paired
  // instructions MergeForward indicates.
  MachineBasicBlock::iterator InsertionPoint = MergeForward ? Paired : I;
  // Also based on MergeForward is from where we copy the base register operand
  // so we get the flags compatible with the input code.
  MachineOperand &BaseRegOp =
      MergeForward ? Paired->getOperand(1) : I->getOperand(1);

  // Which register is Rt and which is Rt2 depends on the offset order.
  MachineInstr *RtMI, *Rt2MI;
  if (I->getOperand(2).getImm() ==
      Paired->getOperand(2).getImm() + OffsetStride) {
    RtMI = Paired;
    Rt2MI = I;
  } else {
    RtMI = I;
    Rt2MI = Paired;
  }
  // Handle Unscaled
  int OffsetImm = RtMI->getOperand(2).getImm();
  if (IsUnscaled && EnableAArch64UnscaledMemOp)
    OffsetImm /= OffsetStride;

  // Construct the new instruction.
  MachineInstrBuilder MIB = BuildMI(*I->getParent(), InsertionPoint,
                                    I->getDebugLoc(), TII->get(NewOpc))
                                .addOperand(RtMI->getOperand(0))
                                .addOperand(Rt2MI->getOperand(0))
                                .addOperand(BaseRegOp)
                                .addImm(OffsetImm);
  (void)MIB;

  // FIXME: Do we need/want to copy the mem operands from the source
  //        instructions? Probably. What uses them after this?

  DEBUG(dbgs() << "Creating pair load/store. Replacing instructions:\n    ");
  DEBUG(I->print(dbgs()));
  DEBUG(dbgs() << "    ");
  DEBUG(Paired->print(dbgs()));
  DEBUG(dbgs() << "  with instruction:\n    ");
  DEBUG(((MachineInstr *)MIB)->print(dbgs()));
  DEBUG(dbgs() << "\n");

  // Erase the old instructions.
  I->eraseFromParent();
  Paired->eraseFromParent();

  return NextI;
}

/// trackRegDefsUses - Remember what registers the specified instruction uses
/// and modifies.
static void trackRegDefsUses(MachineInstr *MI, BitVector &ModifiedRegs,
                             BitVector &UsedRegs,
                             const TargetRegisterInfo *TRI) {
  for (unsigned i = 0, e = MI->getNumOperands(); i != e; ++i) {
    MachineOperand &MO = MI->getOperand(i);
    if (MO.isRegMask())
      ModifiedRegs.setBitsNotInMask(MO.getRegMask());

    if (!MO.isReg())
      continue;
    unsigned Reg = MO.getReg();
    if (MO.isDef()) {
      for (MCRegAliasIterator AI(Reg, TRI, true); AI.isValid(); ++AI)
        ModifiedRegs.set(*AI);
    } else {
      assert(MO.isUse() && "Reg operand not a def and not a use?!?");
      for (MCRegAliasIterator AI(Reg, TRI, true); AI.isValid(); ++AI)
        UsedRegs.set(*AI);
    }
  }
}

static bool inBoundsForPair(bool IsUnscaled, int Offset, int OffsetStride) {
  if (!IsUnscaled && (Offset > 63 || Offset < -64))
    return false;
  if (IsUnscaled) {
    // Convert the byte-offset used by unscaled into an "element" offset used
    // by the scaled pair load/store instructions.
    int ElemOffset = Offset / OffsetStride;
    if (ElemOffset > 63 || ElemOffset < -64)
      return false;
  }
  return true;
}

// Do alignment, specialized to power of 2 and for signed ints,
// avoiding having to do a C-style cast from uint_64t to int when
// using RoundUpToAlignment from include/llvm/Support/MathExtras.h.
// FIXME: Move this function to include/MathExtras.h?
static int alignTo(int Num, int PowOf2) {
  return (Num + PowOf2 - 1) & ~(PowOf2 - 1);
}

/// findMatchingInsn - Scan the instructions looking for a load/store that can
/// be combined with the current instruction into a load/store pair.
MachineBasicBlock::iterator
AArch64LoadStoreOpt::findMatchingInsn(MachineBasicBlock::iterator I,
                                      bool &MergeForward, unsigned Limit) {
  MachineBasicBlock::iterator E = I->getParent()->end();
  MachineBasicBlock::iterator MBBI = I;
  MachineInstr *FirstMI = I;
  ++MBBI;

  int Opc = FirstMI->getOpcode();
  bool MayLoad = FirstMI->mayLoad();
  bool IsUnscaled = isUnscaledLdst(Opc);
  unsigned Reg = FirstMI->getOperand(0).getReg();
  unsigned BaseReg = FirstMI->getOperand(1).getReg();
  int Offset = FirstMI->getOperand(2).getImm();

  // Early exit if the first instruction modifies the base register.
  // e.g., ldr x0, [x0]
  // Early exit if the offset if not possible to match. (6 bits of positive
  // range, plus allow an extra one in case we find a later insn that matches
  // with Offset-1
  if (FirstMI->modifiesRegister(BaseReg, TRI))
    return E;
  int OffsetStride =
      IsUnscaled && EnableAArch64UnscaledMemOp ? getMemSize(FirstMI) : 1;
  if (!inBoundsForPair(IsUnscaled, Offset, OffsetStride))
    return E;

  // Track which registers have been modified and used between the first insn
  // (inclusive) and the second insn.
  BitVector ModifiedRegs, UsedRegs;
  ModifiedRegs.resize(TRI->getNumRegs());
  UsedRegs.resize(TRI->getNumRegs());
  for (unsigned Count = 0; MBBI != E && Count < Limit; ++MBBI) {
    MachineInstr *MI = MBBI;
    // Skip DBG_VALUE instructions. Otherwise debug info can affect the
    // optimization by changing how far we scan.
    if (MI->isDebugValue())
      continue;

    // Now that we know this is a real instruction, count it.
    ++Count;

    if (Opc == MI->getOpcode() && MI->getOperand(2).isImm()) {
      // If we've found another instruction with the same opcode, check to see
      // if the base and offset are compatible with our starting instruction.
      // These instructions all have scaled immediate operands, so we just
      // check for +1/-1. Make sure to check the new instruction offset is
      // actually an immediate and not a symbolic reference destined for
      // a relocation.
      //
      // Pairwise instructions have a 7-bit signed offset field. Single insns
      // have a 12-bit unsigned offset field. To be a valid combine, the
      // final offset must be in range.
      unsigned MIBaseReg = MI->getOperand(1).getReg();
      int MIOffset = MI->getOperand(2).getImm();
      if (BaseReg == MIBaseReg && ((Offset == MIOffset + OffsetStride) ||
                                   (Offset + OffsetStride == MIOffset))) {
        int MinOffset = Offset < MIOffset ? Offset : MIOffset;
        // If this is a volatile load/store that otherwise matched, stop looking
        // as something is going on that we don't have enough information to
        // safely transform. Similarly, stop if we see a hint to avoid pairs.
        if (MI->hasOrderedMemoryRef() || TII->isLdStPairSuppressed(MI))
          return E;
        // If the resultant immediate offset of merging these instructions
        // is out of range for a pairwise instruction, bail and keep looking.
        bool MIIsUnscaled = isUnscaledLdst(MI->getOpcode());
        if (!inBoundsForPair(MIIsUnscaled, MinOffset, OffsetStride)) {
          trackRegDefsUses(MI, ModifiedRegs, UsedRegs, TRI);
          continue;
        }
        // If the alignment requirements of the paired (scaled) instruction
        // can't express the offset of the unscaled input, bail and keep
        // looking.
        if (IsUnscaled && EnableAArch64UnscaledMemOp &&
            (alignTo(MinOffset, OffsetStride) != MinOffset)) {
          trackRegDefsUses(MI, ModifiedRegs, UsedRegs, TRI);
          continue;
        }
        // If the destination register of the loads is the same register, bail
        // and keep looking. A load-pair instruction with both destination
        // registers the same is UNPREDICTABLE and will result in an exception.
        if (MayLoad && Reg == MI->getOperand(0).getReg()) {
          trackRegDefsUses(MI, ModifiedRegs, UsedRegs, TRI);
          continue;
        }

        // If the Rt of the second instruction was not modified or used between
        // the two instructions, we can combine the second into the first.
        if (!ModifiedRegs[MI->getOperand(0).getReg()] &&
            !UsedRegs[MI->getOperand(0).getReg()]) {
          MergeForward = false;
          return MBBI;
        }

        // Likewise, if the Rt of the first instruction is not modified or used
        // between the two instructions, we can combine the first into the
        // second.
        if (!ModifiedRegs[FirstMI->getOperand(0).getReg()] &&
            !UsedRegs[FirstMI->getOperand(0).getReg()]) {
          MergeForward = true;
          return MBBI;
        }
        // Unable to combine these instructions due to interference in between.
        // Keep looking.
      }
    }

    // If the instruction wasn't a matching load or store, but does (or can)
    // modify memory, stop searching, as we don't have alias analysis or
    // anything like that to tell us whether the access is tromping on the
    // locations we care about. The big one we want to catch is calls.
    //
    // FIXME: Theoretically, we can do better than that for SP and FP based
    // references since we can effectively know where those are touching. It's
    // unclear if it's worth the extra code, though. Most paired instructions
    // will be sequential, perhaps with a few intervening non-memory related
    // instructions.
    if (MI->mayStore() || MI->isCall())
      return E;
    // Likewise, if we're matching a store instruction, we don't want to
    // move across a load, as it may be reading the same location.
    if (FirstMI->mayStore() && MI->mayLoad())
      return E;

    // Update modified / uses register lists.
    trackRegDefsUses(MI, ModifiedRegs, UsedRegs, TRI);

    // Otherwise, if the base register is modified, we have no match, so
    // return early.
    if (ModifiedRegs[BaseReg])
      return E;
  }
  return E;
}

MachineBasicBlock::iterator
AArch64LoadStoreOpt::mergePreIdxUpdateInsn(MachineBasicBlock::iterator I,
                                           MachineBasicBlock::iterator Update) {
  assert((Update->getOpcode() == AArch64::ADDXri ||
          Update->getOpcode() == AArch64::SUBXri) &&
         "Unexpected base register update instruction to merge!");
  MachineBasicBlock::iterator NextI = I;
  // Return the instruction following the merged instruction, which is
  // the instruction following our unmerged load. Unless that's the add/sub
  // instruction we're merging, in which case it's the one after that.
  if (++NextI == Update)
    ++NextI;

  int Value = Update->getOperand(2).getImm();
  assert(AArch64_AM::getShiftValue(Update->getOperand(3).getImm()) == 0 &&
         "Can't merge 1 << 12 offset into pre-indexed load / store");
  if (Update->getOpcode() == AArch64::SUBXri)
    Value = -Value;

  unsigned NewOpc = getPreIndexedOpcode(I->getOpcode());
  MachineInstrBuilder MIB =
      BuildMI(*I->getParent(), I, I->getDebugLoc(), TII->get(NewOpc))
          .addOperand(Update->getOperand(0))
          .addOperand(I->getOperand(0))
          .addOperand(I->getOperand(1))
          .addImm(Value);
  (void)MIB;

  DEBUG(dbgs() << "Creating pre-indexed load/store.");
  DEBUG(dbgs() << "    Replacing instructions:\n    ");
  DEBUG(I->print(dbgs()));
  DEBUG(dbgs() << "    ");
  DEBUG(Update->print(dbgs()));
  DEBUG(dbgs() << "  with instruction:\n    ");
  DEBUG(((MachineInstr *)MIB)->print(dbgs()));
  DEBUG(dbgs() << "\n");

  // Erase the old instructions for the block.
  I->eraseFromParent();
  Update->eraseFromParent();

  return NextI;
}

MachineBasicBlock::iterator AArch64LoadStoreOpt::mergePostIdxUpdateInsn(
    MachineBasicBlock::iterator I, MachineBasicBlock::iterator Update) {
  assert((Update->getOpcode() == AArch64::ADDXri ||
          Update->getOpcode() == AArch64::SUBXri) &&
         "Unexpected base register update instruction to merge!");
  MachineBasicBlock::iterator NextI = I;
  // Return the instruction following the merged instruction, which is
  // the instruction following our unmerged load. Unless that's the add/sub
  // instruction we're merging, in which case it's the one after that.
  if (++NextI == Update)
    ++NextI;

  int Value = Update->getOperand(2).getImm();
  assert(AArch64_AM::getShiftValue(Update->getOperand(3).getImm()) == 0 &&
         "Can't merge 1 << 12 offset into post-indexed load / store");
  if (Update->getOpcode() == AArch64::SUBXri)
    Value = -Value;

  unsigned NewOpc = getPostIndexedOpcode(I->getOpcode());
  MachineInstrBuilder MIB =
      BuildMI(*I->getParent(), I, I->getDebugLoc(), TII->get(NewOpc))
          .addOperand(Update->getOperand(0))
          .addOperand(I->getOperand(0))
          .addOperand(I->getOperand(1))
          .addImm(Value);
  (void)MIB;

  DEBUG(dbgs() << "Creating post-indexed load/store.");
  DEBUG(dbgs() << "    Replacing instructions:\n    ");
  DEBUG(I->print(dbgs()));
  DEBUG(dbgs() << "    ");
  DEBUG(Update->print(dbgs()));
  DEBUG(dbgs() << "  with instruction:\n    ");
  DEBUG(((MachineInstr *)MIB)->print(dbgs()));
  DEBUG(dbgs() << "\n");

  // Erase the old instructions for the block.
  I->eraseFromParent();
  Update->eraseFromParent();

  return NextI;
}

static bool isMatchingUpdateInsn(MachineInstr *MI, unsigned BaseReg,
                                 int Offset) {
  switch (MI->getOpcode()) {
  default:
    break;
  case AArch64::SUBXri:
    // Negate the offset for a SUB instruction.
    Offset *= -1;
  // FALLTHROUGH
  case AArch64::ADDXri:
    // Make sure it's a vanilla immediate operand, not a relocation or
    // anything else we can't handle.
    if (!MI->getOperand(2).isImm())
      break;
    // Watch out for 1 << 12 shifted value.
    if (AArch64_AM::getShiftValue(MI->getOperand(3).getImm()))
      break;
    // If the instruction has the base register as source and dest and the
    // immediate will fit in a signed 9-bit integer, then we have a match.
    if (MI->getOperand(0).getReg() == BaseReg &&
        MI->getOperand(1).getReg() == BaseReg &&
        MI->getOperand(2).getImm() <= 255 &&
        MI->getOperand(2).getImm() >= -256) {
      // If we have a non-zero Offset, we check that it matches the amount
      // we're adding to the register.
      if (!Offset || Offset == MI->getOperand(2).getImm())
        return true;
    }
    break;
  }
  return false;
}

MachineBasicBlock::iterator AArch64LoadStoreOpt::findMatchingUpdateInsnForward(
    MachineBasicBlock::iterator I, unsigned Limit, int Value) {
  MachineBasicBlock::iterator E = I->getParent()->end();
  MachineInstr *MemMI = I;
  MachineBasicBlock::iterator MBBI = I;
  const MachineFunction &MF = *MemMI->getParent()->getParent();

  unsigned DestReg = MemMI->getOperand(0).getReg();
  unsigned BaseReg = MemMI->getOperand(1).getReg();
  int Offset = MemMI->getOperand(2).getImm() *
               TII->getRegClass(MemMI->getDesc(), 0, TRI, MF)->getSize();

  // If the base register overlaps the destination register, we can't
  // merge the update.
  if (DestReg == BaseReg || TRI->isSubRegister(BaseReg, DestReg))
    return E;

  // Scan forward looking for post-index opportunities.
  // Updating instructions can't be formed if the memory insn already
  // has an offset other than the value we're looking for.
  if (Offset != Value)
    return E;

  // Track which registers have been modified and used between the first insn
  // (inclusive) and the second insn.
  BitVector ModifiedRegs, UsedRegs;
  ModifiedRegs.resize(TRI->getNumRegs());
  UsedRegs.resize(TRI->getNumRegs());
  ++MBBI;
  for (unsigned Count = 0; MBBI != E; ++MBBI) {
    MachineInstr *MI = MBBI;
    // Skip DBG_VALUE instructions. Otherwise debug info can affect the
    // optimization by changing how far we scan.
    if (MI->isDebugValue())
      continue;

    // Now that we know this is a real instruction, count it.
    ++Count;

    // If we found a match, return it.
    if (isMatchingUpdateInsn(MI, BaseReg, Value))
      return MBBI;

    // Update the status of what the instruction clobbered and used.
    trackRegDefsUses(MI, ModifiedRegs, UsedRegs, TRI);

    // Otherwise, if the base register is used or modified, we have no match, so
    // return early.
    if (ModifiedRegs[BaseReg] || UsedRegs[BaseReg])
      return E;
  }
  return E;
}

MachineBasicBlock::iterator AArch64LoadStoreOpt::findMatchingUpdateInsnBackward(
    MachineBasicBlock::iterator I, unsigned Limit) {
  MachineBasicBlock::iterator B = I->getParent()->begin();
  MachineBasicBlock::iterator E = I->getParent()->end();
  MachineInstr *MemMI = I;
  MachineBasicBlock::iterator MBBI = I;
  const MachineFunction &MF = *MemMI->getParent()->getParent();

  unsigned DestReg = MemMI->getOperand(0).getReg();
  unsigned BaseReg = MemMI->getOperand(1).getReg();
  int Offset = MemMI->getOperand(2).getImm();
  unsigned RegSize = TII->getRegClass(MemMI->getDesc(), 0, TRI, MF)->getSize();

  // If the load/store is the first instruction in the block, there's obviously
  // not any matching update. Ditto if the memory offset isn't zero.
  if (MBBI == B || Offset != 0)
    return E;
  // If the base register overlaps the destination register, we can't
  // merge the update.
  if (DestReg == BaseReg || TRI->isSubRegister(BaseReg, DestReg))
    return E;

  // Track which registers have been modified and used between the first insn
  // (inclusive) and the second insn.
  BitVector ModifiedRegs, UsedRegs;
  ModifiedRegs.resize(TRI->getNumRegs());
  UsedRegs.resize(TRI->getNumRegs());
  --MBBI;
  for (unsigned Count = 0; MBBI != B; --MBBI) {
    MachineInstr *MI = MBBI;
    // Skip DBG_VALUE instructions. Otherwise debug info can affect the
    // optimization by changing how far we scan.
    if (MI->isDebugValue())
      continue;

    // Now that we know this is a real instruction, count it.
    ++Count;

    // If we found a match, return it.
    if (isMatchingUpdateInsn(MI, BaseReg, RegSize))
      return MBBI;

    // Update the status of what the instruction clobbered and used.
    trackRegDefsUses(MI, ModifiedRegs, UsedRegs, TRI);

    // Otherwise, if the base register is used or modified, we have no match, so
    // return early.
    if (ModifiedRegs[BaseReg] || UsedRegs[BaseReg])
      return E;
  }
  return E;
}

bool AArch64LoadStoreOpt::optimizeBlock(MachineBasicBlock &MBB) {
  bool Modified = false;
  // Two tranformations to do here:
  // 1) Find loads and stores that can be merged into a single load or store
  //    pair instruction.
  //      e.g.,
  //        ldr x0, [x2]
  //        ldr x1, [x2, #8]
  //        ; becomes
  //        ldp x0, x1, [x2]
  // 2) Find base register updates that can be merged into the load or store
  //    as a base-reg writeback.
  //      e.g.,
  //        ldr x0, [x2]
  //        add x2, x2, #4
  //        ; becomes
  //        ldr x0, [x2], #4

  for (MachineBasicBlock::iterator MBBI = MBB.begin(), E = MBB.end();
       MBBI != E;) {
    MachineInstr *MI = MBBI;
    switch (MI->getOpcode()) {
    default:
      // Just move on to the next instruction.
      ++MBBI;
      break;
    case AArch64::STRSui:
    case AArch64::STRDui:
    case AArch64::STRQui:
    case AArch64::STRXui:
    case AArch64::STRWui:
    case AArch64::LDRSui:
    case AArch64::LDRDui:
    case AArch64::LDRQui:
    case AArch64::LDRXui:
    case AArch64::LDRWui:
    // do the unscaled versions as well
    case AArch64::STURSi:
    case AArch64::STURDi:
    case AArch64::STURQi:
    case AArch64::STURWi:
    case AArch64::STURXi:
    case AArch64::LDURSi:
    case AArch64::LDURDi:
    case AArch64::LDURQi:
    case AArch64::LDURWi:
    case AArch64::LDURXi: {
      // If this is a volatile load/store, don't mess with it.
      if (MI->hasOrderedMemoryRef()) {
        ++MBBI;
        break;
      }
      // Make sure this is a reg+imm (as opposed to an address reloc).
      if (!MI->getOperand(2).isImm()) {
        ++MBBI;
        break;
      }
      // Check if this load/store has a hint to avoid pair formation.
      // MachineMemOperands hints are set by the AArch64StorePairSuppress pass.
      if (TII->isLdStPairSuppressed(MI)) {
        ++MBBI;
        break;
      }
      // Look ahead up to ScanLimit instructions for a pairable instruction.
      bool MergeForward = false;
      MachineBasicBlock::iterator Paired =
          findMatchingInsn(MBBI, MergeForward, ScanLimit);
      if (Paired != E) {
        // Merge the loads into a pair. Keeping the iterator straight is a
        // pain, so we let the merge routine tell us what the next instruction
        // is after it's done mucking about.
        MBBI = mergePairedInsns(MBBI, Paired, MergeForward);

        Modified = true;
        ++NumPairCreated;
        if (isUnscaledLdst(MI->getOpcode()))
          ++NumUnscaledPairCreated;
        break;
      }
      ++MBBI;
      break;
    }
      // FIXME: Do the other instructions.
    }
  }

  for (MachineBasicBlock::iterator MBBI = MBB.begin(), E = MBB.end();
       MBBI != E;) {
    MachineInstr *MI = MBBI;
    // Do update merging. It's simpler to keep this separate from the above
    // switch, though not strictly necessary.
    int Opc = MI->getOpcode();
    switch (Opc) {
    default:
      // Just move on to the next instruction.
      ++MBBI;
      break;
    case AArch64::STRSui:
    case AArch64::STRDui:
    case AArch64::STRQui:
    case AArch64::STRXui:
    case AArch64::STRWui:
    case AArch64::LDRSui:
    case AArch64::LDRDui:
    case AArch64::LDRQui:
    case AArch64::LDRXui:
    case AArch64::LDRWui:
    // do the unscaled versions as well
    case AArch64::STURSi:
    case AArch64::STURDi:
    case AArch64::STURQi:
    case AArch64::STURWi:
    case AArch64::STURXi:
    case AArch64::LDURSi:
    case AArch64::LDURDi:
    case AArch64::LDURQi:
    case AArch64::LDURWi:
    case AArch64::LDURXi: {
      // Make sure this is a reg+imm (as opposed to an address reloc).
      if (!MI->getOperand(2).isImm()) {
        ++MBBI;
        break;
      }
      // Look ahead up to ScanLimit instructions for a mergable instruction.
      MachineBasicBlock::iterator Update =
          findMatchingUpdateInsnForward(MBBI, ScanLimit, 0);
      if (Update != E) {
        // Merge the update into the ld/st.
        MBBI = mergePostIdxUpdateInsn(MBBI, Update);
        Modified = true;
        ++NumPostFolded;
        break;
      }
      // Don't know how to handle pre/post-index versions, so move to the next
      // instruction.
      if (isUnscaledLdst(Opc)) {
        ++MBBI;
        break;
      }

      // Look back to try to find a pre-index instruction. For example,
      // add x0, x0, #8
      // ldr x1, [x0]
      //   merged into:
      // ldr x1, [x0, #8]!
      Update = findMatchingUpdateInsnBackward(MBBI, ScanLimit);
      if (Update != E) {
        // Merge the update into the ld/st.
        MBBI = mergePreIdxUpdateInsn(MBBI, Update);
        Modified = true;
        ++NumPreFolded;
        break;
      }

      // Look forward to try to find a post-index instruction. For example,
      // ldr x1, [x0, #64]
      // add x0, x0, #64
      //   merged into:
      // ldr x1, [x0, #64]!

      // The immediate in the load/store is scaled by the size of the register
      // being loaded. The immediate in the add we're looking for,
      // however, is not, so adjust here.
      int Value = MI->getOperand(2).getImm() *
                  TII->getRegClass(MI->getDesc(), 0, TRI, *(MBB.getParent()))
                      ->getSize();
      Update = findMatchingUpdateInsnForward(MBBI, ScanLimit, Value);
      if (Update != E) {
        // Merge the update into the ld/st.
        MBBI = mergePreIdxUpdateInsn(MBBI, Update);
        Modified = true;
        ++NumPreFolded;
        break;
      }

      // Nothing found. Just move to the next instruction.
      ++MBBI;
      break;
    }
      // FIXME: Do the other instructions.
    }
  }

  return Modified;
}

bool AArch64LoadStoreOpt::runOnMachineFunction(MachineFunction &Fn) {
  const TargetMachine &TM = Fn.getTarget();
  TII = static_cast<const AArch64InstrInfo *>(TM.getInstrInfo());
  TRI = TM.getRegisterInfo();

  bool Modified = false;
  for (auto &MBB : Fn)
    Modified |= optimizeBlock(MBB);

  return Modified;
}

// FIXME: Do we need/want a pre-alloc pass like ARM has to try to keep
// loads and stores near one another?

/// createARMLoadStoreOptimizationPass - returns an instance of the load / store
/// optimization pass.
FunctionPass *llvm::createAArch64LoadStoreOptimizationPass() {
  return new AArch64LoadStoreOpt();
}