forked from llvm-mirror/llvm
-
Notifications
You must be signed in to change notification settings - Fork 0
/
MemoryDependenceAnalysis.cpp
1543 lines (1332 loc) · 62.5 KB
/
MemoryDependenceAnalysis.cpp
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
//===- MemoryDependenceAnalysis.cpp - Mem Deps Implementation -------------===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file implements an analysis that determines, for a given memory
// operation, what preceding memory operations it depends on. It builds on
// alias analysis information, and tries to provide a lazy, caching interface to
// a common kind of alias information query.
//
//===----------------------------------------------------------------------===//
#include "llvm/Analysis/MemoryDependenceAnalysis.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/MemoryBuiltins.h"
#include "llvm/Analysis/PHITransAddr.h"
#include "llvm/Analysis/ValueTracking.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/Dominators.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/Instructions.h"
#include "llvm/IR/IntrinsicInst.h"
#include "llvm/IR/LLVMContext.h"
#include "llvm/IR/PredIteratorCache.h"
#include "llvm/Support/Debug.h"
using namespace llvm;
#define DEBUG_TYPE "memdep"
STATISTIC(NumCacheNonLocal, "Number of fully cached non-local responses");
STATISTIC(NumCacheDirtyNonLocal, "Number of dirty cached non-local responses");
STATISTIC(NumUncacheNonLocal, "Number of uncached non-local responses");
STATISTIC(NumCacheNonLocalPtr,
"Number of fully cached non-local ptr responses");
STATISTIC(NumCacheDirtyNonLocalPtr,
"Number of cached, but dirty, non-local ptr responses");
STATISTIC(NumUncacheNonLocalPtr,
"Number of uncached non-local ptr responses");
STATISTIC(NumCacheCompleteNonLocalPtr,
"Number of block queries that were completely cached");
// Limit for the number of instructions to scan in a block.
static const int BlockScanLimit = 100;
char MemoryDependenceAnalysis::ID = 0;
// Register this pass...
INITIALIZE_PASS_BEGIN(MemoryDependenceAnalysis, "memdep",
"Memory Dependence Analysis", false, true)
INITIALIZE_AG_DEPENDENCY(AliasAnalysis)
INITIALIZE_PASS_END(MemoryDependenceAnalysis, "memdep",
"Memory Dependence Analysis", false, true)
MemoryDependenceAnalysis::MemoryDependenceAnalysis()
: FunctionPass(ID), PredCache() {
initializeMemoryDependenceAnalysisPass(*PassRegistry::getPassRegistry());
}
MemoryDependenceAnalysis::~MemoryDependenceAnalysis() {
}
/// Clean up memory in between runs
void MemoryDependenceAnalysis::releaseMemory() {
LocalDeps.clear();
NonLocalDeps.clear();
NonLocalPointerDeps.clear();
ReverseLocalDeps.clear();
ReverseNonLocalDeps.clear();
ReverseNonLocalPtrDeps.clear();
PredCache->clear();
}
/// getAnalysisUsage - Does not modify anything. It uses Alias Analysis.
///
void MemoryDependenceAnalysis::getAnalysisUsage(AnalysisUsage &AU) const {
AU.setPreservesAll();
AU.addRequiredTransitive<AliasAnalysis>();
}
bool MemoryDependenceAnalysis::runOnFunction(Function &) {
AA = &getAnalysis<AliasAnalysis>();
DataLayoutPass *DLP = getAnalysisIfAvailable<DataLayoutPass>();
DL = DLP ? &DLP->getDataLayout() : nullptr;
DominatorTreeWrapperPass *DTWP =
getAnalysisIfAvailable<DominatorTreeWrapperPass>();
DT = DTWP ? &DTWP->getDomTree() : nullptr;
if (!PredCache)
PredCache.reset(new PredIteratorCache());
return false;
}
/// RemoveFromReverseMap - This is a helper function that removes Val from
/// 'Inst's set in ReverseMap. If the set becomes empty, remove Inst's entry.
template <typename KeyTy>
static void RemoveFromReverseMap(DenseMap<Instruction*,
SmallPtrSet<KeyTy, 4> > &ReverseMap,
Instruction *Inst, KeyTy Val) {
typename DenseMap<Instruction*, SmallPtrSet<KeyTy, 4> >::iterator
InstIt = ReverseMap.find(Inst);
assert(InstIt != ReverseMap.end() && "Reverse map out of sync?");
bool Found = InstIt->second.erase(Val);
assert(Found && "Invalid reverse map!"); (void)Found;
if (InstIt->second.empty())
ReverseMap.erase(InstIt);
}
/// GetLocation - If the given instruction references a specific memory
/// location, fill in Loc with the details, otherwise set Loc.Ptr to null.
/// Return a ModRefInfo value describing the general behavior of the
/// instruction.
static
AliasAnalysis::ModRefResult GetLocation(const Instruction *Inst,
AliasAnalysis::Location &Loc,
AliasAnalysis *AA) {
if (const LoadInst *LI = dyn_cast<LoadInst>(Inst)) {
if (LI->isUnordered()) {
Loc = AA->getLocation(LI);
return AliasAnalysis::Ref;
}
if (LI->getOrdering() == Monotonic) {
Loc = AA->getLocation(LI);
return AliasAnalysis::ModRef;
}
Loc = AliasAnalysis::Location();
return AliasAnalysis::ModRef;
}
if (const StoreInst *SI = dyn_cast<StoreInst>(Inst)) {
if (SI->isUnordered()) {
Loc = AA->getLocation(SI);
return AliasAnalysis::Mod;
}
if (SI->getOrdering() == Monotonic) {
Loc = AA->getLocation(SI);
return AliasAnalysis::ModRef;
}
Loc = AliasAnalysis::Location();
return AliasAnalysis::ModRef;
}
if (const VAArgInst *V = dyn_cast<VAArgInst>(Inst)) {
Loc = AA->getLocation(V);
return AliasAnalysis::ModRef;
}
if (const CallInst *CI = isFreeCall(Inst, AA->getTargetLibraryInfo())) {
// calls to free() deallocate the entire structure
Loc = AliasAnalysis::Location(CI->getArgOperand(0));
return AliasAnalysis::Mod;
}
if (const IntrinsicInst *II = dyn_cast<IntrinsicInst>(Inst))
switch (II->getIntrinsicID()) {
case Intrinsic::lifetime_start:
case Intrinsic::lifetime_end:
case Intrinsic::invariant_start:
Loc = AliasAnalysis::Location(II->getArgOperand(1),
cast<ConstantInt>(II->getArgOperand(0))
->getZExtValue(),
II->getMetadata(LLVMContext::MD_tbaa));
// These intrinsics don't really modify the memory, but returning Mod
// will allow them to be handled conservatively.
return AliasAnalysis::Mod;
case Intrinsic::invariant_end:
Loc = AliasAnalysis::Location(II->getArgOperand(2),
cast<ConstantInt>(II->getArgOperand(1))
->getZExtValue(),
II->getMetadata(LLVMContext::MD_tbaa));
// These intrinsics don't really modify the memory, but returning Mod
// will allow them to be handled conservatively.
return AliasAnalysis::Mod;
default:
break;
}
// Otherwise, just do the coarse-grained thing that always works.
if (Inst->mayWriteToMemory())
return AliasAnalysis::ModRef;
if (Inst->mayReadFromMemory())
return AliasAnalysis::Ref;
return AliasAnalysis::NoModRef;
}
/// getCallSiteDependencyFrom - Private helper for finding the local
/// dependencies of a call site.
MemDepResult MemoryDependenceAnalysis::
getCallSiteDependencyFrom(CallSite CS, bool isReadOnlyCall,
BasicBlock::iterator ScanIt, BasicBlock *BB) {
unsigned Limit = BlockScanLimit;
// Walk backwards through the block, looking for dependencies
while (ScanIt != BB->begin()) {
// Limit the amount of scanning we do so we don't end up with quadratic
// running time on extreme testcases.
--Limit;
if (!Limit)
return MemDepResult::getUnknown();
Instruction *Inst = --ScanIt;
// If this inst is a memory op, get the pointer it accessed
AliasAnalysis::Location Loc;
AliasAnalysis::ModRefResult MR = GetLocation(Inst, Loc, AA);
if (Loc.Ptr) {
// A simple instruction.
if (AA->getModRefInfo(CS, Loc) != AliasAnalysis::NoModRef)
return MemDepResult::getClobber(Inst);
continue;
}
if (CallSite InstCS = cast<Value>(Inst)) {
// Debug intrinsics don't cause dependences.
if (isa<DbgInfoIntrinsic>(Inst)) continue;
// If these two calls do not interfere, look past it.
switch (AA->getModRefInfo(CS, InstCS)) {
case AliasAnalysis::NoModRef:
// If the two calls are the same, return InstCS as a Def, so that
// CS can be found redundant and eliminated.
if (isReadOnlyCall && !(MR & AliasAnalysis::Mod) &&
CS.getInstruction()->isIdenticalToWhenDefined(Inst))
return MemDepResult::getDef(Inst);
// Otherwise if the two calls don't interact (e.g. InstCS is readnone)
// keep scanning.
continue;
default:
return MemDepResult::getClobber(Inst);
}
}
// If we could not obtain a pointer for the instruction and the instruction
// touches memory then assume that this is a dependency.
if (MR != AliasAnalysis::NoModRef)
return MemDepResult::getClobber(Inst);
}
// No dependence found. If this is the entry block of the function, it is
// unknown, otherwise it is non-local.
if (BB != &BB->getParent()->getEntryBlock())
return MemDepResult::getNonLocal();
return MemDepResult::getNonFuncLocal();
}
/// isLoadLoadClobberIfExtendedToFullWidth - Return true if LI is a load that
/// would fully overlap MemLoc if done as a wider legal integer load.
///
/// MemLocBase, MemLocOffset are lazily computed here the first time the
/// base/offs of memloc is needed.
static bool
isLoadLoadClobberIfExtendedToFullWidth(const AliasAnalysis::Location &MemLoc,
const Value *&MemLocBase,
int64_t &MemLocOffs,
const LoadInst *LI,
const DataLayout *DL) {
// If we have no target data, we can't do this.
if (!DL) return false;
// If we haven't already computed the base/offset of MemLoc, do so now.
if (!MemLocBase)
MemLocBase = GetPointerBaseWithConstantOffset(MemLoc.Ptr, MemLocOffs, DL);
unsigned Size = MemoryDependenceAnalysis::
getLoadLoadClobberFullWidthSize(MemLocBase, MemLocOffs, MemLoc.Size,
LI, *DL);
return Size != 0;
}
/// getLoadLoadClobberFullWidthSize - This is a little bit of analysis that
/// looks at a memory location for a load (specified by MemLocBase, Offs,
/// and Size) and compares it against a load. If the specified load could
/// be safely widened to a larger integer load that is 1) still efficient,
/// 2) safe for the target, and 3) would provide the specified memory
/// location value, then this function returns the size in bytes of the
/// load width to use. If not, this returns zero.
unsigned MemoryDependenceAnalysis::
getLoadLoadClobberFullWidthSize(const Value *MemLocBase, int64_t MemLocOffs,
unsigned MemLocSize, const LoadInst *LI,
const DataLayout &DL) {
// We can only extend simple integer loads.
if (!isa<IntegerType>(LI->getType()) || !LI->isSimple()) return 0;
// Load widening is hostile to ThreadSanitizer: it may cause false positives
// or make the reports more cryptic (access sizes are wrong).
if (LI->getParent()->getParent()->getAttributes().
hasAttribute(AttributeSet::FunctionIndex, Attribute::SanitizeThread))
return 0;
// Get the base of this load.
int64_t LIOffs = 0;
const Value *LIBase =
GetPointerBaseWithConstantOffset(LI->getPointerOperand(), LIOffs, &DL);
// If the two pointers are not based on the same pointer, we can't tell that
// they are related.
if (LIBase != MemLocBase) return 0;
// Okay, the two values are based on the same pointer, but returned as
// no-alias. This happens when we have things like two byte loads at "P+1"
// and "P+3". Check to see if increasing the size of the "LI" load up to its
// alignment (or the largest native integer type) will allow us to load all
// the bits required by MemLoc.
// If MemLoc is before LI, then no widening of LI will help us out.
if (MemLocOffs < LIOffs) return 0;
// Get the alignment of the load in bytes. We assume that it is safe to load
// any legal integer up to this size without a problem. For example, if we're
// looking at an i8 load on x86-32 that is known 1024 byte aligned, we can
// widen it up to an i32 load. If it is known 2-byte aligned, we can widen it
// to i16.
unsigned LoadAlign = LI->getAlignment();
int64_t MemLocEnd = MemLocOffs+MemLocSize;
// If no amount of rounding up will let MemLoc fit into LI, then bail out.
if (LIOffs+LoadAlign < MemLocEnd) return 0;
// This is the size of the load to try. Start with the next larger power of
// two.
unsigned NewLoadByteSize = LI->getType()->getPrimitiveSizeInBits()/8U;
NewLoadByteSize = NextPowerOf2(NewLoadByteSize);
while (1) {
// If this load size is bigger than our known alignment or would not fit
// into a native integer register, then we fail.
if (NewLoadByteSize > LoadAlign ||
!DL.fitsInLegalInteger(NewLoadByteSize*8))
return 0;
if (LIOffs+NewLoadByteSize > MemLocEnd &&
LI->getParent()->getParent()->getAttributes().
hasAttribute(AttributeSet::FunctionIndex, Attribute::SanitizeAddress))
// We will be reading past the location accessed by the original program.
// While this is safe in a regular build, Address Safety analysis tools
// may start reporting false warnings. So, don't do widening.
return 0;
// If a load of this width would include all of MemLoc, then we succeed.
if (LIOffs+NewLoadByteSize >= MemLocEnd)
return NewLoadByteSize;
NewLoadByteSize <<= 1;
}
}
/// getPointerDependencyFrom - Return the instruction on which a memory
/// location depends. If isLoad is true, this routine ignores may-aliases with
/// read-only operations. If isLoad is false, this routine ignores may-aliases
/// with reads from read-only locations. If possible, pass the query
/// instruction as well; this function may take advantage of the metadata
/// annotated to the query instruction to refine the result.
MemDepResult MemoryDependenceAnalysis::
getPointerDependencyFrom(const AliasAnalysis::Location &MemLoc, bool isLoad,
BasicBlock::iterator ScanIt, BasicBlock *BB,
Instruction *QueryInst) {
const Value *MemLocBase = nullptr;
int64_t MemLocOffset = 0;
unsigned Limit = BlockScanLimit;
bool isInvariantLoad = false;
if (isLoad && QueryInst) {
LoadInst *LI = dyn_cast<LoadInst>(QueryInst);
if (LI && LI->getMetadata(LLVMContext::MD_invariant_load) != nullptr)
isInvariantLoad = true;
}
// Walk backwards through the basic block, looking for dependencies.
while (ScanIt != BB->begin()) {
Instruction *Inst = --ScanIt;
if (IntrinsicInst *II = dyn_cast<IntrinsicInst>(Inst))
// Debug intrinsics don't (and can't) cause dependencies.
if (isa<DbgInfoIntrinsic>(II)) continue;
// Limit the amount of scanning we do so we don't end up with quadratic
// running time on extreme testcases.
--Limit;
if (!Limit)
return MemDepResult::getUnknown();
if (IntrinsicInst *II = dyn_cast<IntrinsicInst>(Inst)) {
// If we reach a lifetime begin or end marker, then the query ends here
// because the value is undefined.
if (II->getIntrinsicID() == Intrinsic::lifetime_start) {
// FIXME: This only considers queries directly on the invariant-tagged
// pointer, not on query pointers that are indexed off of them. It'd
// be nice to handle that at some point (the right approach is to use
// GetPointerBaseWithConstantOffset).
if (AA->isMustAlias(AliasAnalysis::Location(II->getArgOperand(1)),
MemLoc))
return MemDepResult::getDef(II);
continue;
}
}
// Values depend on loads if the pointers are must aliased. This means that
// a load depends on another must aliased load from the same value.
if (LoadInst *LI = dyn_cast<LoadInst>(Inst)) {
// Atomic loads have complications involved.
// FIXME: This is overly conservative.
if (!LI->isUnordered())
return MemDepResult::getClobber(LI);
AliasAnalysis::Location LoadLoc = AA->getLocation(LI);
// If we found a pointer, check if it could be the same as our pointer.
AliasAnalysis::AliasResult R = AA->alias(LoadLoc, MemLoc);
if (isLoad) {
if (R == AliasAnalysis::NoAlias) {
// If this is an over-aligned integer load (for example,
// "load i8* %P, align 4") see if it would obviously overlap with the
// queried location if widened to a larger load (e.g. if the queried
// location is 1 byte at P+1). If so, return it as a load/load
// clobber result, allowing the client to decide to widen the load if
// it wants to.
if (IntegerType *ITy = dyn_cast<IntegerType>(LI->getType()))
if (LI->getAlignment()*8 > ITy->getPrimitiveSizeInBits() &&
isLoadLoadClobberIfExtendedToFullWidth(MemLoc, MemLocBase,
MemLocOffset, LI, DL))
return MemDepResult::getClobber(Inst);
continue;
}
// Must aliased loads are defs of each other.
if (R == AliasAnalysis::MustAlias)
return MemDepResult::getDef(Inst);
#if 0 // FIXME: Temporarily disabled. GVN is cleverly rewriting loads
// in terms of clobbering loads, but since it does this by looking
// at the clobbering load directly, it doesn't know about any
// phi translation that may have happened along the way.
// If we have a partial alias, then return this as a clobber for the
// client to handle.
if (R == AliasAnalysis::PartialAlias)
return MemDepResult::getClobber(Inst);
#endif
// Random may-alias loads don't depend on each other without a
// dependence.
continue;
}
// Stores don't depend on other no-aliased accesses.
if (R == AliasAnalysis::NoAlias)
continue;
// Stores don't alias loads from read-only memory.
if (AA->pointsToConstantMemory(LoadLoc))
continue;
// Stores depend on may/must aliased loads.
return MemDepResult::getDef(Inst);
}
if (StoreInst *SI = dyn_cast<StoreInst>(Inst)) {
// Atomic stores have complications involved.
// FIXME: This is overly conservative.
if (!SI->isUnordered())
return MemDepResult::getClobber(SI);
// If alias analysis can tell that this store is guaranteed to not modify
// the query pointer, ignore it. Use getModRefInfo to handle cases where
// the query pointer points to constant memory etc.
if (AA->getModRefInfo(SI, MemLoc) == AliasAnalysis::NoModRef)
continue;
// Ok, this store might clobber the query pointer. Check to see if it is
// a must alias: in this case, we want to return this as a def.
AliasAnalysis::Location StoreLoc = AA->getLocation(SI);
// If we found a pointer, check if it could be the same as our pointer.
AliasAnalysis::AliasResult R = AA->alias(StoreLoc, MemLoc);
if (R == AliasAnalysis::NoAlias)
continue;
if (R == AliasAnalysis::MustAlias)
return MemDepResult::getDef(Inst);
if (isInvariantLoad)
continue;
return MemDepResult::getClobber(Inst);
}
// If this is an allocation, and if we know that the accessed pointer is to
// the allocation, return Def. This means that there is no dependence and
// the access can be optimized based on that. For example, a load could
// turn into undef.
// Note: Only determine this to be a malloc if Inst is the malloc call, not
// a subsequent bitcast of the malloc call result. There can be stores to
// the malloced memory between the malloc call and its bitcast uses, and we
// need to continue scanning until the malloc call.
const TargetLibraryInfo *TLI = AA->getTargetLibraryInfo();
if (isa<AllocaInst>(Inst) || isNoAliasFn(Inst, TLI)) {
const Value *AccessPtr = GetUnderlyingObject(MemLoc.Ptr, DL);
if (AccessPtr == Inst || AA->isMustAlias(Inst, AccessPtr))
return MemDepResult::getDef(Inst);
// Be conservative if the accessed pointer may alias the allocation.
if (AA->alias(Inst, AccessPtr) != AliasAnalysis::NoAlias)
return MemDepResult::getClobber(Inst);
// If the allocation is not aliased and does not read memory (like
// strdup), it is safe to ignore.
if (isa<AllocaInst>(Inst) ||
isMallocLikeFn(Inst, TLI) || isCallocLikeFn(Inst, TLI))
continue;
}
// See if this instruction (e.g. a call or vaarg) mod/ref's the pointer.
AliasAnalysis::ModRefResult MR = AA->getModRefInfo(Inst, MemLoc);
// If necessary, perform additional analysis.
if (MR == AliasAnalysis::ModRef)
MR = AA->callCapturesBefore(Inst, MemLoc, DT);
switch (MR) {
case AliasAnalysis::NoModRef:
// If the call has no effect on the queried pointer, just ignore it.
continue;
case AliasAnalysis::Mod:
return MemDepResult::getClobber(Inst);
case AliasAnalysis::Ref:
// If the call is known to never store to the pointer, and if this is a
// load query, we can safely ignore it (scan past it).
if (isLoad)
continue;
default:
// Otherwise, there is a potential dependence. Return a clobber.
return MemDepResult::getClobber(Inst);
}
}
// No dependence found. If this is the entry block of the function, it is
// unknown, otherwise it is non-local.
if (BB != &BB->getParent()->getEntryBlock())
return MemDepResult::getNonLocal();
return MemDepResult::getNonFuncLocal();
}
/// getDependency - Return the instruction on which a memory operation
/// depends.
MemDepResult MemoryDependenceAnalysis::getDependency(Instruction *QueryInst) {
Instruction *ScanPos = QueryInst;
// Check for a cached result
MemDepResult &LocalCache = LocalDeps[QueryInst];
// If the cached entry is non-dirty, just return it. Note that this depends
// on MemDepResult's default constructing to 'dirty'.
if (!LocalCache.isDirty())
return LocalCache;
// Otherwise, if we have a dirty entry, we know we can start the scan at that
// instruction, which may save us some work.
if (Instruction *Inst = LocalCache.getInst()) {
ScanPos = Inst;
RemoveFromReverseMap(ReverseLocalDeps, Inst, QueryInst);
}
BasicBlock *QueryParent = QueryInst->getParent();
// Do the scan.
if (BasicBlock::iterator(QueryInst) == QueryParent->begin()) {
// No dependence found. If this is the entry block of the function, it is
// unknown, otherwise it is non-local.
if (QueryParent != &QueryParent->getParent()->getEntryBlock())
LocalCache = MemDepResult::getNonLocal();
else
LocalCache = MemDepResult::getNonFuncLocal();
} else {
AliasAnalysis::Location MemLoc;
AliasAnalysis::ModRefResult MR = GetLocation(QueryInst, MemLoc, AA);
if (MemLoc.Ptr) {
// If we can do a pointer scan, make it happen.
bool isLoad = !(MR & AliasAnalysis::Mod);
if (IntrinsicInst *II = dyn_cast<IntrinsicInst>(QueryInst))
isLoad |= II->getIntrinsicID() == Intrinsic::lifetime_start;
LocalCache = getPointerDependencyFrom(MemLoc, isLoad, ScanPos,
QueryParent, QueryInst);
} else if (isa<CallInst>(QueryInst) || isa<InvokeInst>(QueryInst)) {
CallSite QueryCS(QueryInst);
bool isReadOnly = AA->onlyReadsMemory(QueryCS);
LocalCache = getCallSiteDependencyFrom(QueryCS, isReadOnly, ScanPos,
QueryParent);
} else
// Non-memory instruction.
LocalCache = MemDepResult::getUnknown();
}
// Remember the result!
if (Instruction *I = LocalCache.getInst())
ReverseLocalDeps[I].insert(QueryInst);
return LocalCache;
}
#ifndef NDEBUG
/// AssertSorted - This method is used when -debug is specified to verify that
/// cache arrays are properly kept sorted.
static void AssertSorted(MemoryDependenceAnalysis::NonLocalDepInfo &Cache,
int Count = -1) {
if (Count == -1) Count = Cache.size();
if (Count == 0) return;
for (unsigned i = 1; i != unsigned(Count); ++i)
assert(!(Cache[i] < Cache[i-1]) && "Cache isn't sorted!");
}
#endif
/// getNonLocalCallDependency - Perform a full dependency query for the
/// specified call, returning the set of blocks that the value is
/// potentially live across. The returned set of results will include a
/// "NonLocal" result for all blocks where the value is live across.
///
/// This method assumes the instruction returns a "NonLocal" dependency
/// within its own block.
///
/// This returns a reference to an internal data structure that may be
/// invalidated on the next non-local query or when an instruction is
/// removed. Clients must copy this data if they want it around longer than
/// that.
const MemoryDependenceAnalysis::NonLocalDepInfo &
MemoryDependenceAnalysis::getNonLocalCallDependency(CallSite QueryCS) {
assert(getDependency(QueryCS.getInstruction()).isNonLocal() &&
"getNonLocalCallDependency should only be used on calls with non-local deps!");
PerInstNLInfo &CacheP = NonLocalDeps[QueryCS.getInstruction()];
NonLocalDepInfo &Cache = CacheP.first;
/// DirtyBlocks - This is the set of blocks that need to be recomputed. In
/// the cached case, this can happen due to instructions being deleted etc. In
/// the uncached case, this starts out as the set of predecessors we care
/// about.
SmallVector<BasicBlock*, 32> DirtyBlocks;
if (!Cache.empty()) {
// Okay, we have a cache entry. If we know it is not dirty, just return it
// with no computation.
if (!CacheP.second) {
++NumCacheNonLocal;
return Cache;
}
// If we already have a partially computed set of results, scan them to
// determine what is dirty, seeding our initial DirtyBlocks worklist.
for (NonLocalDepInfo::iterator I = Cache.begin(), E = Cache.end();
I != E; ++I)
if (I->getResult().isDirty())
DirtyBlocks.push_back(I->getBB());
// Sort the cache so that we can do fast binary search lookups below.
std::sort(Cache.begin(), Cache.end());
++NumCacheDirtyNonLocal;
//cerr << "CACHED CASE: " << DirtyBlocks.size() << " dirty: "
// << Cache.size() << " cached: " << *QueryInst;
} else {
// Seed DirtyBlocks with each of the preds of QueryInst's block.
BasicBlock *QueryBB = QueryCS.getInstruction()->getParent();
for (BasicBlock **PI = PredCache->GetPreds(QueryBB); *PI; ++PI)
DirtyBlocks.push_back(*PI);
++NumUncacheNonLocal;
}
// isReadonlyCall - If this is a read-only call, we can be more aggressive.
bool isReadonlyCall = AA->onlyReadsMemory(QueryCS);
SmallPtrSet<BasicBlock*, 64> Visited;
unsigned NumSortedEntries = Cache.size();
DEBUG(AssertSorted(Cache));
// Iterate while we still have blocks to update.
while (!DirtyBlocks.empty()) {
BasicBlock *DirtyBB = DirtyBlocks.back();
DirtyBlocks.pop_back();
// Already processed this block?
if (!Visited.insert(DirtyBB))
continue;
// Do a binary search to see if we already have an entry for this block in
// the cache set. If so, find it.
DEBUG(AssertSorted(Cache, NumSortedEntries));
NonLocalDepInfo::iterator Entry =
std::upper_bound(Cache.begin(), Cache.begin()+NumSortedEntries,
NonLocalDepEntry(DirtyBB));
if (Entry != Cache.begin() && std::prev(Entry)->getBB() == DirtyBB)
--Entry;
NonLocalDepEntry *ExistingResult = nullptr;
if (Entry != Cache.begin()+NumSortedEntries &&
Entry->getBB() == DirtyBB) {
// If we already have an entry, and if it isn't already dirty, the block
// is done.
if (!Entry->getResult().isDirty())
continue;
// Otherwise, remember this slot so we can update the value.
ExistingResult = &*Entry;
}
// If the dirty entry has a pointer, start scanning from it so we don't have
// to rescan the entire block.
BasicBlock::iterator ScanPos = DirtyBB->end();
if (ExistingResult) {
if (Instruction *Inst = ExistingResult->getResult().getInst()) {
ScanPos = Inst;
// We're removing QueryInst's use of Inst.
RemoveFromReverseMap(ReverseNonLocalDeps, Inst,
QueryCS.getInstruction());
}
}
// Find out if this block has a local dependency for QueryInst.
MemDepResult Dep;
if (ScanPos != DirtyBB->begin()) {
Dep = getCallSiteDependencyFrom(QueryCS, isReadonlyCall,ScanPos, DirtyBB);
} else if (DirtyBB != &DirtyBB->getParent()->getEntryBlock()) {
// No dependence found. If this is the entry block of the function, it is
// a clobber, otherwise it is unknown.
Dep = MemDepResult::getNonLocal();
} else {
Dep = MemDepResult::getNonFuncLocal();
}
// If we had a dirty entry for the block, update it. Otherwise, just add
// a new entry.
if (ExistingResult)
ExistingResult->setResult(Dep);
else
Cache.push_back(NonLocalDepEntry(DirtyBB, Dep));
// If the block has a dependency (i.e. it isn't completely transparent to
// the value), remember the association!
if (!Dep.isNonLocal()) {
// Keep the ReverseNonLocalDeps map up to date so we can efficiently
// update this when we remove instructions.
if (Instruction *Inst = Dep.getInst())
ReverseNonLocalDeps[Inst].insert(QueryCS.getInstruction());
} else {
// If the block *is* completely transparent to the load, we need to check
// the predecessors of this block. Add them to our worklist.
for (BasicBlock **PI = PredCache->GetPreds(DirtyBB); *PI; ++PI)
DirtyBlocks.push_back(*PI);
}
}
return Cache;
}
/// getNonLocalPointerDependency - Perform a full dependency query for an
/// access to the specified (non-volatile) memory location, returning the
/// set of instructions that either define or clobber the value.
///
/// This method assumes the pointer has a "NonLocal" dependency within its
/// own block.
///
void MemoryDependenceAnalysis::
getNonLocalPointerDependency(const AliasAnalysis::Location &Loc, bool isLoad,
BasicBlock *FromBB,
SmallVectorImpl<NonLocalDepResult> &Result) {
assert(Loc.Ptr->getType()->isPointerTy() &&
"Can't get pointer deps of a non-pointer!");
Result.clear();
PHITransAddr Address(const_cast<Value *>(Loc.Ptr), DL);
// This is the set of blocks we've inspected, and the pointer we consider in
// each block. Because of critical edges, we currently bail out if querying
// a block with multiple different pointers. This can happen during PHI
// translation.
DenseMap<BasicBlock*, Value*> Visited;
if (!getNonLocalPointerDepFromBB(Address, Loc, isLoad, FromBB,
Result, Visited, true))
return;
Result.clear();
Result.push_back(NonLocalDepResult(FromBB,
MemDepResult::getUnknown(),
const_cast<Value *>(Loc.Ptr)));
}
/// GetNonLocalInfoForBlock - Compute the memdep value for BB with
/// Pointer/PointeeSize using either cached information in Cache or by doing a
/// lookup (which may use dirty cache info if available). If we do a lookup,
/// add the result to the cache.
MemDepResult MemoryDependenceAnalysis::
GetNonLocalInfoForBlock(const AliasAnalysis::Location &Loc,
bool isLoad, BasicBlock *BB,
NonLocalDepInfo *Cache, unsigned NumSortedEntries) {
// Do a binary search to see if we already have an entry for this block in
// the cache set. If so, find it.
NonLocalDepInfo::iterator Entry =
std::upper_bound(Cache->begin(), Cache->begin()+NumSortedEntries,
NonLocalDepEntry(BB));
if (Entry != Cache->begin() && (Entry-1)->getBB() == BB)
--Entry;
NonLocalDepEntry *ExistingResult = nullptr;
if (Entry != Cache->begin()+NumSortedEntries && Entry->getBB() == BB)
ExistingResult = &*Entry;
// If we have a cached entry, and it is non-dirty, use it as the value for
// this dependency.
if (ExistingResult && !ExistingResult->getResult().isDirty()) {
++NumCacheNonLocalPtr;
return ExistingResult->getResult();
}
// Otherwise, we have to scan for the value. If we have a dirty cache
// entry, start scanning from its position, otherwise we scan from the end
// of the block.
BasicBlock::iterator ScanPos = BB->end();
if (ExistingResult && ExistingResult->getResult().getInst()) {
assert(ExistingResult->getResult().getInst()->getParent() == BB &&
"Instruction invalidated?");
++NumCacheDirtyNonLocalPtr;
ScanPos = ExistingResult->getResult().getInst();
// Eliminating the dirty entry from 'Cache', so update the reverse info.
ValueIsLoadPair CacheKey(Loc.Ptr, isLoad);
RemoveFromReverseMap(ReverseNonLocalPtrDeps, ScanPos, CacheKey);
} else {
++NumUncacheNonLocalPtr;
}
// Scan the block for the dependency.
MemDepResult Dep = getPointerDependencyFrom(Loc, isLoad, ScanPos, BB);
// If we had a dirty entry for the block, update it. Otherwise, just add
// a new entry.
if (ExistingResult)
ExistingResult->setResult(Dep);
else
Cache->push_back(NonLocalDepEntry(BB, Dep));
// If the block has a dependency (i.e. it isn't completely transparent to
// the value), remember the reverse association because we just added it
// to Cache!
if (!Dep.isDef() && !Dep.isClobber())
return Dep;
// Keep the ReverseNonLocalPtrDeps map up to date so we can efficiently
// update MemDep when we remove instructions.
Instruction *Inst = Dep.getInst();
assert(Inst && "Didn't depend on anything?");
ValueIsLoadPair CacheKey(Loc.Ptr, isLoad);
ReverseNonLocalPtrDeps[Inst].insert(CacheKey);
return Dep;
}
/// SortNonLocalDepInfoCache - Sort the a NonLocalDepInfo cache, given a certain
/// number of elements in the array that are already properly ordered. This is
/// optimized for the case when only a few entries are added.
static void
SortNonLocalDepInfoCache(MemoryDependenceAnalysis::NonLocalDepInfo &Cache,
unsigned NumSortedEntries) {
switch (Cache.size() - NumSortedEntries) {
case 0:
// done, no new entries.
break;
case 2: {
// Two new entries, insert the last one into place.
NonLocalDepEntry Val = Cache.back();
Cache.pop_back();
MemoryDependenceAnalysis::NonLocalDepInfo::iterator Entry =
std::upper_bound(Cache.begin(), Cache.end()-1, Val);
Cache.insert(Entry, Val);
// FALL THROUGH.
}
case 1:
// One new entry, Just insert the new value at the appropriate position.
if (Cache.size() != 1) {
NonLocalDepEntry Val = Cache.back();
Cache.pop_back();
MemoryDependenceAnalysis::NonLocalDepInfo::iterator Entry =
std::upper_bound(Cache.begin(), Cache.end(), Val);
Cache.insert(Entry, Val);
}
break;
default:
// Added many values, do a full scale sort.
std::sort(Cache.begin(), Cache.end());
break;
}
}
/// getNonLocalPointerDepFromBB - Perform a dependency query based on
/// pointer/pointeesize starting at the end of StartBB. Add any clobber/def
/// results to the results vector and keep track of which blocks are visited in
/// 'Visited'.
///
/// This has special behavior for the first block queries (when SkipFirstBlock
/// is true). In this special case, it ignores the contents of the specified
/// block and starts returning dependence info for its predecessors.
///
/// This function returns false on success, or true to indicate that it could
/// not compute dependence information for some reason. This should be treated
/// as a clobber dependence on the first instruction in the predecessor block.
bool MemoryDependenceAnalysis::
getNonLocalPointerDepFromBB(const PHITransAddr &Pointer,
const AliasAnalysis::Location &Loc,
bool isLoad, BasicBlock *StartBB,
SmallVectorImpl<NonLocalDepResult> &Result,
DenseMap<BasicBlock*, Value*> &Visited,
bool SkipFirstBlock) {
// Look up the cached info for Pointer.
ValueIsLoadPair CacheKey(Pointer.getAddr(), isLoad);
// Set up a temporary NLPI value. If the map doesn't yet have an entry for
// CacheKey, this value will be inserted as the associated value. Otherwise,
// it'll be ignored, and we'll have to check to see if the cached size and
// tbaa tag are consistent with the current query.
NonLocalPointerInfo InitialNLPI;
InitialNLPI.Size = Loc.Size;
InitialNLPI.TBAATag = Loc.TBAATag;
// Get the NLPI for CacheKey, inserting one into the map if it doesn't
// already have one.
std::pair<CachedNonLocalPointerInfo::iterator, bool> Pair =
NonLocalPointerDeps.insert(std::make_pair(CacheKey, InitialNLPI));
NonLocalPointerInfo *CacheInfo = &Pair.first->second;
// If we already have a cache entry for this CacheKey, we may need to do some
// work to reconcile the cache entry and the current query.
if (!Pair.second) {
if (CacheInfo->Size < Loc.Size) {
// The query's Size is greater than the cached one. Throw out the
// cached data and proceed with the query at the greater size.
CacheInfo->Pair = BBSkipFirstBlockPair();
CacheInfo->Size = Loc.Size;
for (NonLocalDepInfo::iterator DI = CacheInfo->NonLocalDeps.begin(),
DE = CacheInfo->NonLocalDeps.end(); DI != DE; ++DI)
if (Instruction *Inst = DI->getResult().getInst())
RemoveFromReverseMap(ReverseNonLocalPtrDeps, Inst, CacheKey);
CacheInfo->NonLocalDeps.clear();
} else if (CacheInfo->Size > Loc.Size) {
// This query's Size is less than the cached one. Conservatively restart
// the query using the greater size.
return getNonLocalPointerDepFromBB(Pointer,
Loc.getWithNewSize(CacheInfo->Size),
isLoad, StartBB, Result, Visited,
SkipFirstBlock);
}
// If the query's TBAATag is inconsistent with the cached one,
// conservatively throw out the cached data and restart the query with
// no tag if needed.
if (CacheInfo->TBAATag != Loc.TBAATag) {
if (CacheInfo->TBAATag) {
CacheInfo->Pair = BBSkipFirstBlockPair();
CacheInfo->TBAATag = nullptr;
for (NonLocalDepInfo::iterator DI = CacheInfo->NonLocalDeps.begin(),
DE = CacheInfo->NonLocalDeps.end(); DI != DE; ++DI)
if (Instruction *Inst = DI->getResult().getInst())
RemoveFromReverseMap(ReverseNonLocalPtrDeps, Inst, CacheKey);
CacheInfo->NonLocalDeps.clear();
}
if (Loc.TBAATag)
return getNonLocalPointerDepFromBB(Pointer, Loc.getWithoutTBAATag(),
isLoad, StartBB, Result, Visited,
SkipFirstBlock);
}
}
NonLocalDepInfo *Cache = &CacheInfo->NonLocalDeps;
// If we have valid cached information for exactly the block we are
// investigating, just return it with no recomputation.
if (CacheInfo->Pair == BBSkipFirstBlockPair(StartBB, SkipFirstBlock)) {
// We have a fully cached result for this query then we can just return the
// cached results and populate the visited set. However, we have to verify
// that we don't already have conflicting results for these blocks. Check
// to ensure that if a block in the results set is in the visited set that
// it was for the same pointer query.
if (!Visited.empty()) {
for (NonLocalDepInfo::iterator I = Cache->begin(), E = Cache->end();
I != E; ++I) {
DenseMap<BasicBlock*, Value*>::iterator VI = Visited.find(I->getBB());
if (VI == Visited.end() || VI->second == Pointer.getAddr())
continue;
// We have a pointer mismatch in a block. Just return clobber, saying
// that something was clobbered in this result. We could also do a
// non-fully cached query, but there is little point in doing this.
return true;
}
}