[Alignment][NFC] Convert StoreInst to MaybeAlign
[llvm-complete.git] / lib / Transforms / InstCombine / InstCombineInternal.h
blob4917a355cadde61f326ccad2bd355797829fc7ce
1 //===- InstCombineInternal.h - InstCombine pass internals -------*- C++ -*-===//
2 //
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 //
7 //===----------------------------------------------------------------------===//
8 //
9 /// \file
10 ///
11 /// This file provides internal interfaces used to implement the InstCombine.
13 //===----------------------------------------------------------------------===//
15 #ifndef LLVM_LIB_TRANSFORMS_INSTCOMBINE_INSTCOMBINEINTERNAL_H
16 #define LLVM_LIB_TRANSFORMS_INSTCOMBINE_INSTCOMBINEINTERNAL_H
18 #include "llvm/ADT/ArrayRef.h"
19 #include "llvm/Analysis/AliasAnalysis.h"
20 #include "llvm/Analysis/InstructionSimplify.h"
21 #include "llvm/Analysis/TargetFolder.h"
22 #include "llvm/Analysis/ValueTracking.h"
23 #include "llvm/IR/Argument.h"
24 #include "llvm/IR/BasicBlock.h"
25 #include "llvm/IR/Constant.h"
26 #include "llvm/IR/Constants.h"
27 #include "llvm/IR/DerivedTypes.h"
28 #include "llvm/IR/IRBuilder.h"
29 #include "llvm/IR/InstVisitor.h"
30 #include "llvm/IR/InstrTypes.h"
31 #include "llvm/IR/Instruction.h"
32 #include "llvm/IR/IntrinsicInst.h"
33 #include "llvm/IR/Intrinsics.h"
34 #include "llvm/IR/PatternMatch.h"
35 #include "llvm/IR/Use.h"
36 #include "llvm/IR/Value.h"
37 #include "llvm/Support/Casting.h"
38 #include "llvm/Support/Compiler.h"
39 #include "llvm/Support/Debug.h"
40 #include "llvm/Support/KnownBits.h"
41 #include "llvm/Support/raw_ostream.h"
42 #include "llvm/Transforms/InstCombine/InstCombineWorklist.h"
43 #include "llvm/Transforms/Utils/Local.h"
44 #include <cassert>
45 #include <cstdint>
47 #define DEBUG_TYPE "instcombine"
49 using namespace llvm::PatternMatch;
51 namespace llvm {
53 class APInt;
54 class AssumptionCache;
55 class BlockFrequencyInfo;
56 class DataLayout;
57 class DominatorTree;
58 class GEPOperator;
59 class GlobalVariable;
60 class LoopInfo;
61 class OptimizationRemarkEmitter;
62 class ProfileSummaryInfo;
63 class TargetLibraryInfo;
64 class User;
66 /// Assign a complexity or rank value to LLVM Values. This is used to reduce
67 /// the amount of pattern matching needed for compares and commutative
68 /// instructions. For example, if we have:
69 /// icmp ugt X, Constant
70 /// or
71 /// xor (add X, Constant), cast Z
72 ///
73 /// We do not have to consider the commuted variants of these patterns because
74 /// canonicalization based on complexity guarantees the above ordering.
75 ///
76 /// This routine maps IR values to various complexity ranks:
77 /// 0 -> undef
78 /// 1 -> Constants
79 /// 2 -> Other non-instructions
80 /// 3 -> Arguments
81 /// 4 -> Cast and (f)neg/not instructions
82 /// 5 -> Other instructions
83 static inline unsigned getComplexity(Value *V) {
84 if (isa<Instruction>(V)) {
85 if (isa<CastInst>(V) || match(V, m_Neg(m_Value())) ||
86 match(V, m_Not(m_Value())) || match(V, m_FNeg(m_Value())))
87 return 4;
88 return 5;
90 if (isa<Argument>(V))
91 return 3;
92 return isa<Constant>(V) ? (isa<UndefValue>(V) ? 0 : 1) : 2;
95 /// Predicate canonicalization reduces the number of patterns that need to be
96 /// matched by other transforms. For example, we may swap the operands of a
97 /// conditional branch or select to create a compare with a canonical (inverted)
98 /// predicate which is then more likely to be matched with other values.
99 static inline bool isCanonicalPredicate(CmpInst::Predicate Pred) {
100 switch (Pred) {
101 case CmpInst::ICMP_NE:
102 case CmpInst::ICMP_ULE:
103 case CmpInst::ICMP_SLE:
104 case CmpInst::ICMP_UGE:
105 case CmpInst::ICMP_SGE:
106 // TODO: There are 16 FCMP predicates. Should others be (not) canonical?
107 case CmpInst::FCMP_ONE:
108 case CmpInst::FCMP_OLE:
109 case CmpInst::FCMP_OGE:
110 return false;
111 default:
112 return true;
116 /// Given an exploded icmp instruction, return true if the comparison only
117 /// checks the sign bit. If it only checks the sign bit, set TrueIfSigned if the
118 /// result of the comparison is true when the input value is signed.
119 inline bool isSignBitCheck(ICmpInst::Predicate Pred, const APInt &RHS,
120 bool &TrueIfSigned) {
121 switch (Pred) {
122 case ICmpInst::ICMP_SLT: // True if LHS s< 0
123 TrueIfSigned = true;
124 return RHS.isNullValue();
125 case ICmpInst::ICMP_SLE: // True if LHS s<= -1
126 TrueIfSigned = true;
127 return RHS.isAllOnesValue();
128 case ICmpInst::ICMP_SGT: // True if LHS s> -1
129 TrueIfSigned = false;
130 return RHS.isAllOnesValue();
131 case ICmpInst::ICMP_SGE: // True if LHS s>= 0
132 TrueIfSigned = false;
133 return RHS.isNullValue();
134 case ICmpInst::ICMP_UGT:
135 // True if LHS u> RHS and RHS == sign-bit-mask - 1
136 TrueIfSigned = true;
137 return RHS.isMaxSignedValue();
138 case ICmpInst::ICMP_UGE:
139 // True if LHS u>= RHS and RHS == sign-bit-mask (2^7, 2^15, 2^31, etc)
140 TrueIfSigned = true;
141 return RHS.isMinSignedValue();
142 case ICmpInst::ICMP_ULT:
143 // True if LHS u< RHS and RHS == sign-bit-mask (2^7, 2^15, 2^31, etc)
144 TrueIfSigned = false;
145 return RHS.isMinSignedValue();
146 case ICmpInst::ICMP_ULE:
147 // True if LHS u<= RHS and RHS == sign-bit-mask - 1
148 TrueIfSigned = false;
149 return RHS.isMaxSignedValue();
150 default:
151 return false;
155 llvm::Optional<std::pair<CmpInst::Predicate, Constant *>>
156 getFlippedStrictnessPredicateAndConstant(CmpInst::Predicate Pred, Constant *C);
158 /// Return the source operand of a potentially bitcasted value while optionally
159 /// checking if it has one use. If there is no bitcast or the one use check is
160 /// not met, return the input value itself.
161 static inline Value *peekThroughBitcast(Value *V, bool OneUseOnly = false) {
162 if (auto *BitCast = dyn_cast<BitCastInst>(V))
163 if (!OneUseOnly || BitCast->hasOneUse())
164 return BitCast->getOperand(0);
166 // V is not a bitcast or V has more than one use and OneUseOnly is true.
167 return V;
170 /// Add one to a Constant
171 static inline Constant *AddOne(Constant *C) {
172 return ConstantExpr::getAdd(C, ConstantInt::get(C->getType(), 1));
175 /// Subtract one from a Constant
176 static inline Constant *SubOne(Constant *C) {
177 return ConstantExpr::getSub(C, ConstantInt::get(C->getType(), 1));
180 /// Return true if the specified value is free to invert (apply ~ to).
181 /// This happens in cases where the ~ can be eliminated. If WillInvertAllUses
182 /// is true, work under the assumption that the caller intends to remove all
183 /// uses of V and only keep uses of ~V.
185 /// See also: canFreelyInvertAllUsersOf()
186 static inline bool isFreeToInvert(Value *V, bool WillInvertAllUses) {
187 // ~(~(X)) -> X.
188 if (match(V, m_Not(m_Value())))
189 return true;
191 // Constants can be considered to be not'ed values.
192 if (match(V, m_AnyIntegralConstant()))
193 return true;
195 // Compares can be inverted if all of their uses are being modified to use the
196 // ~V.
197 if (isa<CmpInst>(V))
198 return WillInvertAllUses;
200 // If `V` is of the form `A + Constant` then `-1 - V` can be folded into `(-1
201 // - Constant) - A` if we are willing to invert all of the uses.
202 if (BinaryOperator *BO = dyn_cast<BinaryOperator>(V))
203 if (BO->getOpcode() == Instruction::Add ||
204 BO->getOpcode() == Instruction::Sub)
205 if (isa<Constant>(BO->getOperand(0)) || isa<Constant>(BO->getOperand(1)))
206 return WillInvertAllUses;
208 // Selects with invertible operands are freely invertible
209 if (match(V, m_Select(m_Value(), m_Not(m_Value()), m_Not(m_Value()))))
210 return WillInvertAllUses;
212 return false;
215 /// Given i1 V, can every user of V be freely adapted if V is changed to !V ?
217 /// See also: isFreeToInvert()
218 static inline bool canFreelyInvertAllUsersOf(Value *V, Value *IgnoredUser) {
219 // Look at every user of V.
220 for (User *U : V->users()) {
221 if (U == IgnoredUser)
222 continue; // Don't consider this user.
224 auto *I = cast<Instruction>(U);
225 switch (I->getOpcode()) {
226 case Instruction::Select:
227 case Instruction::Br:
228 break; // Free to invert by swapping true/false values/destinations.
229 case Instruction::Xor: // Can invert 'xor' if it's a 'not', by ignoring it.
230 if (!match(I, m_Not(m_Value())))
231 return false; // Not a 'not'.
232 break;
233 default:
234 return false; // Don't know, likely not freely invertible.
236 // So far all users were free to invert...
238 return true; // Can freely invert all users!
241 /// Some binary operators require special handling to avoid poison and undefined
242 /// behavior. If a constant vector has undef elements, replace those undefs with
243 /// identity constants if possible because those are always safe to execute.
244 /// If no identity constant exists, replace undef with some other safe constant.
245 static inline Constant *getSafeVectorConstantForBinop(
246 BinaryOperator::BinaryOps Opcode, Constant *In, bool IsRHSConstant) {
247 assert(In->getType()->isVectorTy() && "Not expecting scalars here");
249 Type *EltTy = In->getType()->getVectorElementType();
250 auto *SafeC = ConstantExpr::getBinOpIdentity(Opcode, EltTy, IsRHSConstant);
251 if (!SafeC) {
252 // TODO: Should this be available as a constant utility function? It is
253 // similar to getBinOpAbsorber().
254 if (IsRHSConstant) {
255 switch (Opcode) {
256 case Instruction::SRem: // X % 1 = 0
257 case Instruction::URem: // X %u 1 = 0
258 SafeC = ConstantInt::get(EltTy, 1);
259 break;
260 case Instruction::FRem: // X % 1.0 (doesn't simplify, but it is safe)
261 SafeC = ConstantFP::get(EltTy, 1.0);
262 break;
263 default:
264 llvm_unreachable("Only rem opcodes have no identity constant for RHS");
266 } else {
267 switch (Opcode) {
268 case Instruction::Shl: // 0 << X = 0
269 case Instruction::LShr: // 0 >>u X = 0
270 case Instruction::AShr: // 0 >> X = 0
271 case Instruction::SDiv: // 0 / X = 0
272 case Instruction::UDiv: // 0 /u X = 0
273 case Instruction::SRem: // 0 % X = 0
274 case Instruction::URem: // 0 %u X = 0
275 case Instruction::Sub: // 0 - X (doesn't simplify, but it is safe)
276 case Instruction::FSub: // 0.0 - X (doesn't simplify, but it is safe)
277 case Instruction::FDiv: // 0.0 / X (doesn't simplify, but it is safe)
278 case Instruction::FRem: // 0.0 % X = 0
279 SafeC = Constant::getNullValue(EltTy);
280 break;
281 default:
282 llvm_unreachable("Expected to find identity constant for opcode");
286 assert(SafeC && "Must have safe constant for binop");
287 unsigned NumElts = In->getType()->getVectorNumElements();
288 SmallVector<Constant *, 16> Out(NumElts);
289 for (unsigned i = 0; i != NumElts; ++i) {
290 Constant *C = In->getAggregateElement(i);
291 Out[i] = isa<UndefValue>(C) ? SafeC : C;
293 return ConstantVector::get(Out);
296 /// The core instruction combiner logic.
298 /// This class provides both the logic to recursively visit instructions and
299 /// combine them.
300 class LLVM_LIBRARY_VISIBILITY InstCombiner
301 : public InstVisitor<InstCombiner, Instruction *> {
302 // FIXME: These members shouldn't be public.
303 public:
304 /// A worklist of the instructions that need to be simplified.
305 InstCombineWorklist &Worklist;
307 /// An IRBuilder that automatically inserts new instructions into the
308 /// worklist.
309 using BuilderTy = IRBuilder<TargetFolder, IRBuilderCallbackInserter>;
310 BuilderTy &Builder;
312 private:
313 // Mode in which we are running the combiner.
314 const bool MinimizeSize;
316 /// Enable combines that trigger rarely but are costly in compiletime.
317 const bool ExpensiveCombines;
319 AliasAnalysis *AA;
321 // Required analyses.
322 AssumptionCache &AC;
323 TargetLibraryInfo &TLI;
324 DominatorTree &DT;
325 const DataLayout &DL;
326 const SimplifyQuery SQ;
327 OptimizationRemarkEmitter &ORE;
328 BlockFrequencyInfo *BFI;
329 ProfileSummaryInfo *PSI;
331 // Optional analyses. When non-null, these can both be used to do better
332 // combining and will be updated to reflect any changes.
333 LoopInfo *LI;
335 bool MadeIRChange = false;
337 public:
338 InstCombiner(InstCombineWorklist &Worklist, BuilderTy &Builder,
339 bool MinimizeSize, bool ExpensiveCombines, AliasAnalysis *AA,
340 AssumptionCache &AC, TargetLibraryInfo &TLI, DominatorTree &DT,
341 OptimizationRemarkEmitter &ORE, BlockFrequencyInfo *BFI,
342 ProfileSummaryInfo *PSI, const DataLayout &DL, LoopInfo *LI)
343 : Worklist(Worklist), Builder(Builder), MinimizeSize(MinimizeSize),
344 ExpensiveCombines(ExpensiveCombines), AA(AA), AC(AC), TLI(TLI), DT(DT),
345 DL(DL), SQ(DL, &TLI, &DT, &AC), ORE(ORE), BFI(BFI), PSI(PSI), LI(LI) {}
347 /// Run the combiner over the entire worklist until it is empty.
349 /// \returns true if the IR is changed.
350 bool run();
352 AssumptionCache &getAssumptionCache() const { return AC; }
354 const DataLayout &getDataLayout() const { return DL; }
356 DominatorTree &getDominatorTree() const { return DT; }
358 LoopInfo *getLoopInfo() const { return LI; }
360 TargetLibraryInfo &getTargetLibraryInfo() const { return TLI; }
362 // Visitation implementation - Implement instruction combining for different
363 // instruction types. The semantics are as follows:
364 // Return Value:
365 // null - No change was made
366 // I - Change was made, I is still valid, I may be dead though
367 // otherwise - Change was made, replace I with returned instruction
369 Instruction *visitFNeg(UnaryOperator &I);
370 Instruction *visitAdd(BinaryOperator &I);
371 Instruction *visitFAdd(BinaryOperator &I);
372 Value *OptimizePointerDifference(Value *LHS, Value *RHS, Type *Ty);
373 Instruction *visitSub(BinaryOperator &I);
374 Instruction *visitFSub(BinaryOperator &I);
375 Instruction *visitMul(BinaryOperator &I);
376 Instruction *visitFMul(BinaryOperator &I);
377 Instruction *visitURem(BinaryOperator &I);
378 Instruction *visitSRem(BinaryOperator &I);
379 Instruction *visitFRem(BinaryOperator &I);
380 bool simplifyDivRemOfSelectWithZeroOp(BinaryOperator &I);
381 Instruction *commonRemTransforms(BinaryOperator &I);
382 Instruction *commonIRemTransforms(BinaryOperator &I);
383 Instruction *commonDivTransforms(BinaryOperator &I);
384 Instruction *commonIDivTransforms(BinaryOperator &I);
385 Instruction *visitUDiv(BinaryOperator &I);
386 Instruction *visitSDiv(BinaryOperator &I);
387 Instruction *visitFDiv(BinaryOperator &I);
388 Value *simplifyRangeCheck(ICmpInst *Cmp0, ICmpInst *Cmp1, bool Inverted);
389 Instruction *visitAnd(BinaryOperator &I);
390 Instruction *visitOr(BinaryOperator &I);
391 Instruction *visitXor(BinaryOperator &I);
392 Instruction *visitShl(BinaryOperator &I);
393 Value *reassociateShiftAmtsOfTwoSameDirectionShifts(
394 BinaryOperator *Sh0, const SimplifyQuery &SQ,
395 bool AnalyzeForSignBitExtraction = false);
396 Instruction *canonicalizeCondSignextOfHighBitExtractToSignextHighBitExtract(
397 BinaryOperator &I);
398 Instruction *foldVariableSignZeroExtensionOfVariableHighBitExtract(
399 BinaryOperator &OldAShr);
400 Instruction *visitAShr(BinaryOperator &I);
401 Instruction *visitLShr(BinaryOperator &I);
402 Instruction *commonShiftTransforms(BinaryOperator &I);
403 Instruction *visitFCmpInst(FCmpInst &I);
404 Instruction *visitICmpInst(ICmpInst &I);
405 Instruction *FoldShiftByConstant(Value *Op0, Constant *Op1,
406 BinaryOperator &I);
407 Instruction *commonCastTransforms(CastInst &CI);
408 Instruction *commonPointerCastTransforms(CastInst &CI);
409 Instruction *visitTrunc(TruncInst &CI);
410 Instruction *visitZExt(ZExtInst &CI);
411 Instruction *visitSExt(SExtInst &CI);
412 Instruction *visitFPTrunc(FPTruncInst &CI);
413 Instruction *visitFPExt(CastInst &CI);
414 Instruction *visitFPToUI(FPToUIInst &FI);
415 Instruction *visitFPToSI(FPToSIInst &FI);
416 Instruction *visitUIToFP(CastInst &CI);
417 Instruction *visitSIToFP(CastInst &CI);
418 Instruction *visitPtrToInt(PtrToIntInst &CI);
419 Instruction *visitIntToPtr(IntToPtrInst &CI);
420 Instruction *visitBitCast(BitCastInst &CI);
421 Instruction *visitAddrSpaceCast(AddrSpaceCastInst &CI);
422 Instruction *FoldItoFPtoI(Instruction &FI);
423 Instruction *visitSelectInst(SelectInst &SI);
424 Instruction *visitCallInst(CallInst &CI);
425 Instruction *visitInvokeInst(InvokeInst &II);
426 Instruction *visitCallBrInst(CallBrInst &CBI);
428 Instruction *SliceUpIllegalIntegerPHI(PHINode &PN);
429 Instruction *visitPHINode(PHINode &PN);
430 Instruction *visitGetElementPtrInst(GetElementPtrInst &GEP);
431 Instruction *visitAllocaInst(AllocaInst &AI);
432 Instruction *visitAllocSite(Instruction &FI);
433 Instruction *visitFree(CallInst &FI);
434 Instruction *visitLoadInst(LoadInst &LI);
435 Instruction *visitStoreInst(StoreInst &SI);
436 Instruction *visitAtomicRMWInst(AtomicRMWInst &SI);
437 Instruction *visitBranchInst(BranchInst &BI);
438 Instruction *visitFenceInst(FenceInst &FI);
439 Instruction *visitSwitchInst(SwitchInst &SI);
440 Instruction *visitReturnInst(ReturnInst &RI);
441 Instruction *visitInsertValueInst(InsertValueInst &IV);
442 Instruction *visitInsertElementInst(InsertElementInst &IE);
443 Instruction *visitExtractElementInst(ExtractElementInst &EI);
444 Instruction *visitShuffleVectorInst(ShuffleVectorInst &SVI);
445 Instruction *visitExtractValueInst(ExtractValueInst &EV);
446 Instruction *visitLandingPadInst(LandingPadInst &LI);
447 Instruction *visitVAStartInst(VAStartInst &I);
448 Instruction *visitVACopyInst(VACopyInst &I);
450 /// Specify what to return for unhandled instructions.
451 Instruction *visitInstruction(Instruction &I) { return nullptr; }
453 /// True when DB dominates all uses of DI except UI.
454 /// UI must be in the same block as DI.
455 /// The routine checks that the DI parent and DB are different.
456 bool dominatesAllUses(const Instruction *DI, const Instruction *UI,
457 const BasicBlock *DB) const;
459 /// Try to replace select with select operand SIOpd in SI-ICmp sequence.
460 bool replacedSelectWithOperand(SelectInst *SI, const ICmpInst *Icmp,
461 const unsigned SIOpd);
463 /// Try to replace instruction \p I with value \p V which are pointers
464 /// in different address space.
465 /// \return true if successful.
466 bool replacePointer(Instruction &I, Value *V);
468 private:
469 bool shouldChangeType(unsigned FromBitWidth, unsigned ToBitWidth) const;
470 bool shouldChangeType(Type *From, Type *To) const;
471 Value *dyn_castNegVal(Value *V) const;
472 Type *FindElementAtOffset(PointerType *PtrTy, int64_t Offset,
473 SmallVectorImpl<Value *> &NewIndices);
475 /// Classify whether a cast is worth optimizing.
477 /// This is a helper to decide whether the simplification of
478 /// logic(cast(A), cast(B)) to cast(logic(A, B)) should be performed.
480 /// \param CI The cast we are interested in.
482 /// \return true if this cast actually results in any code being generated and
483 /// if it cannot already be eliminated by some other transformation.
484 bool shouldOptimizeCast(CastInst *CI);
486 /// Try to optimize a sequence of instructions checking if an operation
487 /// on LHS and RHS overflows.
489 /// If this overflow check is done via one of the overflow check intrinsics,
490 /// then CtxI has to be the call instruction calling that intrinsic. If this
491 /// overflow check is done by arithmetic followed by a compare, then CtxI has
492 /// to be the arithmetic instruction.
494 /// If a simplification is possible, stores the simplified result of the
495 /// operation in OperationResult and result of the overflow check in
496 /// OverflowResult, and return true. If no simplification is possible,
497 /// returns false.
498 bool OptimizeOverflowCheck(Instruction::BinaryOps BinaryOp, bool IsSigned,
499 Value *LHS, Value *RHS,
500 Instruction &CtxI, Value *&OperationResult,
501 Constant *&OverflowResult);
503 Instruction *visitCallBase(CallBase &Call);
504 Instruction *tryOptimizeCall(CallInst *CI);
505 bool transformConstExprCastCall(CallBase &Call);
506 Instruction *transformCallThroughTrampoline(CallBase &Call,
507 IntrinsicInst &Tramp);
509 Value *simplifyMaskedLoad(IntrinsicInst &II);
510 Instruction *simplifyMaskedStore(IntrinsicInst &II);
511 Instruction *simplifyMaskedGather(IntrinsicInst &II);
512 Instruction *simplifyMaskedScatter(IntrinsicInst &II);
514 /// Transform (zext icmp) to bitwise / integer operations in order to
515 /// eliminate it.
517 /// \param ICI The icmp of the (zext icmp) pair we are interested in.
518 /// \parem CI The zext of the (zext icmp) pair we are interested in.
519 /// \param DoTransform Pass false to just test whether the given (zext icmp)
520 /// would be transformed. Pass true to actually perform the transformation.
522 /// \return null if the transformation cannot be performed. If the
523 /// transformation can be performed the new instruction that replaces the
524 /// (zext icmp) pair will be returned (if \p DoTransform is false the
525 /// unmodified \p ICI will be returned in this case).
526 Instruction *transformZExtICmp(ICmpInst *ICI, ZExtInst &CI,
527 bool DoTransform = true);
529 Instruction *transformSExtICmp(ICmpInst *ICI, Instruction &CI);
531 bool willNotOverflowSignedAdd(const Value *LHS, const Value *RHS,
532 const Instruction &CxtI) const {
533 return computeOverflowForSignedAdd(LHS, RHS, &CxtI) ==
534 OverflowResult::NeverOverflows;
537 bool willNotOverflowUnsignedAdd(const Value *LHS, const Value *RHS,
538 const Instruction &CxtI) const {
539 return computeOverflowForUnsignedAdd(LHS, RHS, &CxtI) ==
540 OverflowResult::NeverOverflows;
543 bool willNotOverflowAdd(const Value *LHS, const Value *RHS,
544 const Instruction &CxtI, bool IsSigned) const {
545 return IsSigned ? willNotOverflowSignedAdd(LHS, RHS, CxtI)
546 : willNotOverflowUnsignedAdd(LHS, RHS, CxtI);
549 bool willNotOverflowSignedSub(const Value *LHS, const Value *RHS,
550 const Instruction &CxtI) const {
551 return computeOverflowForSignedSub(LHS, RHS, &CxtI) ==
552 OverflowResult::NeverOverflows;
555 bool willNotOverflowUnsignedSub(const Value *LHS, const Value *RHS,
556 const Instruction &CxtI) const {
557 return computeOverflowForUnsignedSub(LHS, RHS, &CxtI) ==
558 OverflowResult::NeverOverflows;
561 bool willNotOverflowSub(const Value *LHS, const Value *RHS,
562 const Instruction &CxtI, bool IsSigned) const {
563 return IsSigned ? willNotOverflowSignedSub(LHS, RHS, CxtI)
564 : willNotOverflowUnsignedSub(LHS, RHS, CxtI);
567 bool willNotOverflowSignedMul(const Value *LHS, const Value *RHS,
568 const Instruction &CxtI) const {
569 return computeOverflowForSignedMul(LHS, RHS, &CxtI) ==
570 OverflowResult::NeverOverflows;
573 bool willNotOverflowUnsignedMul(const Value *LHS, const Value *RHS,
574 const Instruction &CxtI) const {
575 return computeOverflowForUnsignedMul(LHS, RHS, &CxtI) ==
576 OverflowResult::NeverOverflows;
579 bool willNotOverflowMul(const Value *LHS, const Value *RHS,
580 const Instruction &CxtI, bool IsSigned) const {
581 return IsSigned ? willNotOverflowSignedMul(LHS, RHS, CxtI)
582 : willNotOverflowUnsignedMul(LHS, RHS, CxtI);
585 bool willNotOverflow(BinaryOperator::BinaryOps Opcode, const Value *LHS,
586 const Value *RHS, const Instruction &CxtI,
587 bool IsSigned) const {
588 switch (Opcode) {
589 case Instruction::Add: return willNotOverflowAdd(LHS, RHS, CxtI, IsSigned);
590 case Instruction::Sub: return willNotOverflowSub(LHS, RHS, CxtI, IsSigned);
591 case Instruction::Mul: return willNotOverflowMul(LHS, RHS, CxtI, IsSigned);
592 default: llvm_unreachable("Unexpected opcode for overflow query");
596 Value *EmitGEPOffset(User *GEP);
597 Instruction *scalarizePHI(ExtractElementInst &EI, PHINode *PN);
598 Instruction *foldCastedBitwiseLogic(BinaryOperator &I);
599 Instruction *narrowBinOp(TruncInst &Trunc);
600 Instruction *narrowMaskedBinOp(BinaryOperator &And);
601 Instruction *narrowMathIfNoOverflow(BinaryOperator &I);
602 Instruction *narrowRotate(TruncInst &Trunc);
603 Instruction *optimizeBitCastFromPhi(CastInst &CI, PHINode *PN);
605 /// Determine if a pair of casts can be replaced by a single cast.
607 /// \param CI1 The first of a pair of casts.
608 /// \param CI2 The second of a pair of casts.
610 /// \return 0 if the cast pair cannot be eliminated, otherwise returns an
611 /// Instruction::CastOps value for a cast that can replace the pair, casting
612 /// CI1->getSrcTy() to CI2->getDstTy().
614 /// \see CastInst::isEliminableCastPair
615 Instruction::CastOps isEliminableCastPair(const CastInst *CI1,
616 const CastInst *CI2);
618 Value *foldAndOfICmps(ICmpInst *LHS, ICmpInst *RHS, Instruction &CxtI);
619 Value *foldOrOfICmps(ICmpInst *LHS, ICmpInst *RHS, Instruction &CxtI);
620 Value *foldXorOfICmps(ICmpInst *LHS, ICmpInst *RHS, BinaryOperator &I);
622 /// Optimize (fcmp)&(fcmp) or (fcmp)|(fcmp).
623 /// NOTE: Unlike most of instcombine, this returns a Value which should
624 /// already be inserted into the function.
625 Value *foldLogicOfFCmps(FCmpInst *LHS, FCmpInst *RHS, bool IsAnd);
627 Value *foldAndOrOfICmpsOfAndWithPow2(ICmpInst *LHS, ICmpInst *RHS,
628 bool JoinedByAnd, Instruction &CxtI);
629 Value *matchSelectFromAndOr(Value *A, Value *B, Value *C, Value *D);
630 Value *getSelectCondition(Value *A, Value *B);
632 Instruction *foldIntrinsicWithOverflowCommon(IntrinsicInst *II);
634 public:
635 /// Inserts an instruction \p New before instruction \p Old
637 /// Also adds the new instruction to the worklist and returns \p New so that
638 /// it is suitable for use as the return from the visitation patterns.
639 Instruction *InsertNewInstBefore(Instruction *New, Instruction &Old) {
640 assert(New && !New->getParent() &&
641 "New instruction already inserted into a basic block!");
642 BasicBlock *BB = Old.getParent();
643 BB->getInstList().insert(Old.getIterator(), New); // Insert inst
644 Worklist.Add(New);
645 return New;
648 /// Same as InsertNewInstBefore, but also sets the debug loc.
649 Instruction *InsertNewInstWith(Instruction *New, Instruction &Old) {
650 New->setDebugLoc(Old.getDebugLoc());
651 return InsertNewInstBefore(New, Old);
654 /// A combiner-aware RAUW-like routine.
656 /// This method is to be used when an instruction is found to be dead,
657 /// replaceable with another preexisting expression. Here we add all uses of
658 /// I to the worklist, replace all uses of I with the new value, then return
659 /// I, so that the inst combiner will know that I was modified.
660 Instruction *replaceInstUsesWith(Instruction &I, Value *V) {
661 // If there are no uses to replace, then we return nullptr to indicate that
662 // no changes were made to the program.
663 if (I.use_empty()) return nullptr;
665 Worklist.AddUsersToWorkList(I); // Add all modified instrs to worklist.
667 // If we are replacing the instruction with itself, this must be in a
668 // segment of unreachable code, so just clobber the instruction.
669 if (&I == V)
670 V = UndefValue::get(I.getType());
672 LLVM_DEBUG(dbgs() << "IC: Replacing " << I << "\n"
673 << " with " << *V << '\n');
675 I.replaceAllUsesWith(V);
676 return &I;
679 /// Creates a result tuple for an overflow intrinsic \p II with a given
680 /// \p Result and a constant \p Overflow value.
681 Instruction *CreateOverflowTuple(IntrinsicInst *II, Value *Result,
682 Constant *Overflow) {
683 Constant *V[] = {UndefValue::get(Result->getType()), Overflow};
684 StructType *ST = cast<StructType>(II->getType());
685 Constant *Struct = ConstantStruct::get(ST, V);
686 return InsertValueInst::Create(Struct, Result, 0);
689 /// Create and insert the idiom we use to indicate a block is unreachable
690 /// without having to rewrite the CFG from within InstCombine.
691 void CreateNonTerminatorUnreachable(Instruction *InsertAt) {
692 auto &Ctx = InsertAt->getContext();
693 new StoreInst(ConstantInt::getTrue(Ctx),
694 UndefValue::get(Type::getInt1PtrTy(Ctx)),
695 InsertAt);
699 /// Combiner aware instruction erasure.
701 /// When dealing with an instruction that has side effects or produces a void
702 /// value, we can't rely on DCE to delete the instruction. Instead, visit
703 /// methods should return the value returned by this function.
704 Instruction *eraseInstFromFunction(Instruction &I) {
705 LLVM_DEBUG(dbgs() << "IC: ERASE " << I << '\n');
706 assert(I.use_empty() && "Cannot erase instruction that is used!");
707 salvageDebugInfo(I);
709 // Make sure that we reprocess all operands now that we reduced their
710 // use counts.
711 if (I.getNumOperands() < 8) {
712 for (Use &Operand : I.operands())
713 if (auto *Inst = dyn_cast<Instruction>(Operand))
714 Worklist.Add(Inst);
716 Worklist.Remove(&I);
717 I.eraseFromParent();
718 MadeIRChange = true;
719 return nullptr; // Don't do anything with FI
722 void computeKnownBits(const Value *V, KnownBits &Known,
723 unsigned Depth, const Instruction *CxtI) const {
724 llvm::computeKnownBits(V, Known, DL, Depth, &AC, CxtI, &DT);
727 KnownBits computeKnownBits(const Value *V, unsigned Depth,
728 const Instruction *CxtI) const {
729 return llvm::computeKnownBits(V, DL, Depth, &AC, CxtI, &DT);
732 bool isKnownToBeAPowerOfTwo(const Value *V, bool OrZero = false,
733 unsigned Depth = 0,
734 const Instruction *CxtI = nullptr) {
735 return llvm::isKnownToBeAPowerOfTwo(V, DL, OrZero, Depth, &AC, CxtI, &DT);
738 bool MaskedValueIsZero(const Value *V, const APInt &Mask, unsigned Depth = 0,
739 const Instruction *CxtI = nullptr) const {
740 return llvm::MaskedValueIsZero(V, Mask, DL, Depth, &AC, CxtI, &DT);
743 unsigned ComputeNumSignBits(const Value *Op, unsigned Depth = 0,
744 const Instruction *CxtI = nullptr) const {
745 return llvm::ComputeNumSignBits(Op, DL, Depth, &AC, CxtI, &DT);
748 OverflowResult computeOverflowForUnsignedMul(const Value *LHS,
749 const Value *RHS,
750 const Instruction *CxtI) const {
751 return llvm::computeOverflowForUnsignedMul(LHS, RHS, DL, &AC, CxtI, &DT);
754 OverflowResult computeOverflowForSignedMul(const Value *LHS,
755 const Value *RHS,
756 const Instruction *CxtI) const {
757 return llvm::computeOverflowForSignedMul(LHS, RHS, DL, &AC, CxtI, &DT);
760 OverflowResult computeOverflowForUnsignedAdd(const Value *LHS,
761 const Value *RHS,
762 const Instruction *CxtI) const {
763 return llvm::computeOverflowForUnsignedAdd(LHS, RHS, DL, &AC, CxtI, &DT);
766 OverflowResult computeOverflowForSignedAdd(const Value *LHS,
767 const Value *RHS,
768 const Instruction *CxtI) const {
769 return llvm::computeOverflowForSignedAdd(LHS, RHS, DL, &AC, CxtI, &DT);
772 OverflowResult computeOverflowForUnsignedSub(const Value *LHS,
773 const Value *RHS,
774 const Instruction *CxtI) const {
775 return llvm::computeOverflowForUnsignedSub(LHS, RHS, DL, &AC, CxtI, &DT);
778 OverflowResult computeOverflowForSignedSub(const Value *LHS, const Value *RHS,
779 const Instruction *CxtI) const {
780 return llvm::computeOverflowForSignedSub(LHS, RHS, DL, &AC, CxtI, &DT);
783 OverflowResult computeOverflow(
784 Instruction::BinaryOps BinaryOp, bool IsSigned,
785 Value *LHS, Value *RHS, Instruction *CxtI) const;
787 /// Maximum size of array considered when transforming.
788 uint64_t MaxArraySizeForCombine = 0;
790 private:
791 /// Performs a few simplifications for operators which are associative
792 /// or commutative.
793 bool SimplifyAssociativeOrCommutative(BinaryOperator &I);
795 /// Tries to simplify binary operations which some other binary
796 /// operation distributes over.
798 /// It does this by either by factorizing out common terms (eg "(A*B)+(A*C)"
799 /// -> "A*(B+C)") or expanding out if this results in simplifications (eg: "A
800 /// & (B | C) -> (A&B) | (A&C)" if this is a win). Returns the simplified
801 /// value, or null if it didn't simplify.
802 Value *SimplifyUsingDistributiveLaws(BinaryOperator &I);
804 /// Tries to simplify add operations using the definition of remainder.
806 /// The definition of remainder is X % C = X - (X / C ) * C. The add
807 /// expression X % C0 + (( X / C0 ) % C1) * C0 can be simplified to
808 /// X % (C0 * C1)
809 Value *SimplifyAddWithRemainder(BinaryOperator &I);
811 // Binary Op helper for select operations where the expression can be
812 // efficiently reorganized.
813 Value *SimplifySelectsFeedingBinaryOp(BinaryOperator &I, Value *LHS,
814 Value *RHS);
816 /// This tries to simplify binary operations by factorizing out common terms
817 /// (e. g. "(A*B)+(A*C)" -> "A*(B+C)").
818 Value *tryFactorization(BinaryOperator &, Instruction::BinaryOps, Value *,
819 Value *, Value *, Value *);
821 /// Match a select chain which produces one of three values based on whether
822 /// the LHS is less than, equal to, or greater than RHS respectively.
823 /// Return true if we matched a three way compare idiom. The LHS, RHS, Less,
824 /// Equal and Greater values are saved in the matching process and returned to
825 /// the caller.
826 bool matchThreeWayIntCompare(SelectInst *SI, Value *&LHS, Value *&RHS,
827 ConstantInt *&Less, ConstantInt *&Equal,
828 ConstantInt *&Greater);
830 /// Attempts to replace V with a simpler value based on the demanded
831 /// bits.
832 Value *SimplifyDemandedUseBits(Value *V, APInt DemandedMask, KnownBits &Known,
833 unsigned Depth, Instruction *CxtI);
834 bool SimplifyDemandedBits(Instruction *I, unsigned Op,
835 const APInt &DemandedMask, KnownBits &Known,
836 unsigned Depth = 0);
838 /// Helper routine of SimplifyDemandedUseBits. It computes KnownZero/KnownOne
839 /// bits. It also tries to handle simplifications that can be done based on
840 /// DemandedMask, but without modifying the Instruction.
841 Value *SimplifyMultipleUseDemandedBits(Instruction *I,
842 const APInt &DemandedMask,
843 KnownBits &Known,
844 unsigned Depth, Instruction *CxtI);
846 /// Helper routine of SimplifyDemandedUseBits. It tries to simplify demanded
847 /// bit for "r1 = shr x, c1; r2 = shl r1, c2" instruction sequence.
848 Value *simplifyShrShlDemandedBits(
849 Instruction *Shr, const APInt &ShrOp1, Instruction *Shl,
850 const APInt &ShlOp1, const APInt &DemandedMask, KnownBits &Known);
852 /// Tries to simplify operands to an integer instruction based on its
853 /// demanded bits.
854 bool SimplifyDemandedInstructionBits(Instruction &Inst);
856 Value *simplifyAMDGCNMemoryIntrinsicDemanded(IntrinsicInst *II,
857 APInt DemandedElts,
858 int DmaskIdx = -1);
860 Value *SimplifyDemandedVectorElts(Value *V, APInt DemandedElts,
861 APInt &UndefElts, unsigned Depth = 0,
862 bool AllowMultipleUsers = false);
864 /// Canonicalize the position of binops relative to shufflevector.
865 Instruction *foldVectorBinop(BinaryOperator &Inst);
867 /// Given a binary operator, cast instruction, or select which has a PHI node
868 /// as operand #0, see if we can fold the instruction into the PHI (which is
869 /// only possible if all operands to the PHI are constants).
870 Instruction *foldOpIntoPhi(Instruction &I, PHINode *PN);
872 /// Given an instruction with a select as one operand and a constant as the
873 /// other operand, try to fold the binary operator into the select arguments.
874 /// This also works for Cast instructions, which obviously do not have a
875 /// second operand.
876 Instruction *FoldOpIntoSelect(Instruction &Op, SelectInst *SI);
878 /// This is a convenience wrapper function for the above two functions.
879 Instruction *foldBinOpIntoSelectOrPhi(BinaryOperator &I);
881 Instruction *foldAddWithConstant(BinaryOperator &Add);
883 /// Try to rotate an operation below a PHI node, using PHI nodes for
884 /// its operands.
885 Instruction *FoldPHIArgOpIntoPHI(PHINode &PN);
886 Instruction *FoldPHIArgBinOpIntoPHI(PHINode &PN);
887 Instruction *FoldPHIArgGEPIntoPHI(PHINode &PN);
888 Instruction *FoldPHIArgLoadIntoPHI(PHINode &PN);
889 Instruction *FoldPHIArgZextsIntoPHI(PHINode &PN);
891 /// If an integer typed PHI has only one use which is an IntToPtr operation,
892 /// replace the PHI with an existing pointer typed PHI if it exists. Otherwise
893 /// insert a new pointer typed PHI and replace the original one.
894 Instruction *FoldIntegerTypedPHI(PHINode &PN);
896 /// Helper function for FoldPHIArgXIntoPHI() to set debug location for the
897 /// folded operation.
898 void PHIArgMergedDebugLoc(Instruction *Inst, PHINode &PN);
900 Instruction *foldGEPICmp(GEPOperator *GEPLHS, Value *RHS,
901 ICmpInst::Predicate Cond, Instruction &I);
902 Instruction *foldAllocaCmp(ICmpInst &ICI, const AllocaInst *Alloca,
903 const Value *Other);
904 Instruction *foldCmpLoadFromIndexedGlobal(GetElementPtrInst *GEP,
905 GlobalVariable *GV, CmpInst &ICI,
906 ConstantInt *AndCst = nullptr);
907 Instruction *foldFCmpIntToFPConst(FCmpInst &I, Instruction *LHSI,
908 Constant *RHSC);
909 Instruction *foldICmpAddOpConst(Value *X, const APInt &C,
910 ICmpInst::Predicate Pred);
911 Instruction *foldICmpWithCastOp(ICmpInst &ICI);
913 Instruction *foldICmpUsingKnownBits(ICmpInst &Cmp);
914 Instruction *foldICmpWithDominatingICmp(ICmpInst &Cmp);
915 Instruction *foldICmpWithConstant(ICmpInst &Cmp);
916 Instruction *foldICmpInstWithConstant(ICmpInst &Cmp);
917 Instruction *foldICmpInstWithConstantNotInt(ICmpInst &Cmp);
918 Instruction *foldICmpBinOp(ICmpInst &Cmp, const SimplifyQuery &SQ);
919 Instruction *foldICmpEquality(ICmpInst &Cmp);
920 Instruction *foldIRemByPowerOfTwoToBitTest(ICmpInst &I);
921 Instruction *foldSignBitTest(ICmpInst &I);
922 Instruction *foldICmpWithZero(ICmpInst &Cmp);
924 Value *foldUnsignedMultiplicationOverflowCheck(ICmpInst &Cmp);
926 Instruction *foldICmpSelectConstant(ICmpInst &Cmp, SelectInst *Select,
927 ConstantInt *C);
928 Instruction *foldICmpTruncConstant(ICmpInst &Cmp, TruncInst *Trunc,
929 const APInt &C);
930 Instruction *foldICmpAndConstant(ICmpInst &Cmp, BinaryOperator *And,
931 const APInt &C);
932 Instruction *foldICmpXorConstant(ICmpInst &Cmp, BinaryOperator *Xor,
933 const APInt &C);
934 Instruction *foldICmpOrConstant(ICmpInst &Cmp, BinaryOperator *Or,
935 const APInt &C);
936 Instruction *foldICmpMulConstant(ICmpInst &Cmp, BinaryOperator *Mul,
937 const APInt &C);
938 Instruction *foldICmpShlConstant(ICmpInst &Cmp, BinaryOperator *Shl,
939 const APInt &C);
940 Instruction *foldICmpShrConstant(ICmpInst &Cmp, BinaryOperator *Shr,
941 const APInt &C);
942 Instruction *foldICmpSRemConstant(ICmpInst &Cmp, BinaryOperator *UDiv,
943 const APInt &C);
944 Instruction *foldICmpUDivConstant(ICmpInst &Cmp, BinaryOperator *UDiv,
945 const APInt &C);
946 Instruction *foldICmpDivConstant(ICmpInst &Cmp, BinaryOperator *Div,
947 const APInt &C);
948 Instruction *foldICmpSubConstant(ICmpInst &Cmp, BinaryOperator *Sub,
949 const APInt &C);
950 Instruction *foldICmpAddConstant(ICmpInst &Cmp, BinaryOperator *Add,
951 const APInt &C);
952 Instruction *foldICmpAndConstConst(ICmpInst &Cmp, BinaryOperator *And,
953 const APInt &C1);
954 Instruction *foldICmpAndShift(ICmpInst &Cmp, BinaryOperator *And,
955 const APInt &C1, const APInt &C2);
956 Instruction *foldICmpShrConstConst(ICmpInst &I, Value *ShAmt, const APInt &C1,
957 const APInt &C2);
958 Instruction *foldICmpShlConstConst(ICmpInst &I, Value *ShAmt, const APInt &C1,
959 const APInt &C2);
961 Instruction *foldICmpBinOpEqualityWithConstant(ICmpInst &Cmp,
962 BinaryOperator *BO,
963 const APInt &C);
964 Instruction *foldICmpIntrinsicWithConstant(ICmpInst &ICI, IntrinsicInst *II,
965 const APInt &C);
966 Instruction *foldICmpEqIntrinsicWithConstant(ICmpInst &ICI, IntrinsicInst *II,
967 const APInt &C);
969 // Helpers of visitSelectInst().
970 Instruction *foldSelectExtConst(SelectInst &Sel);
971 Instruction *foldSelectOpOp(SelectInst &SI, Instruction *TI, Instruction *FI);
972 Instruction *foldSelectIntoOp(SelectInst &SI, Value *, Value *);
973 Instruction *foldSPFofSPF(Instruction *Inner, SelectPatternFlavor SPF1,
974 Value *A, Value *B, Instruction &Outer,
975 SelectPatternFlavor SPF2, Value *C);
976 Instruction *foldSelectInstWithICmp(SelectInst &SI, ICmpInst *ICI);
978 Instruction *OptAndOp(BinaryOperator *Op, ConstantInt *OpRHS,
979 ConstantInt *AndRHS, BinaryOperator &TheAnd);
981 Value *insertRangeTest(Value *V, const APInt &Lo, const APInt &Hi,
982 bool isSigned, bool Inside);
983 Instruction *PromoteCastOfAllocation(BitCastInst &CI, AllocaInst &AI);
984 bool mergeStoreIntoSuccessor(StoreInst &SI);
986 /// Given an 'or' instruction, check to see if it is part of a bswap idiom.
987 /// If so, return the equivalent bswap intrinsic.
988 Instruction *matchBSwap(BinaryOperator &Or);
990 Instruction *SimplifyAnyMemTransfer(AnyMemTransferInst *MI);
991 Instruction *SimplifyAnyMemSet(AnyMemSetInst *MI);
993 Value *EvaluateInDifferentType(Value *V, Type *Ty, bool isSigned);
995 /// Returns a value X such that Val = X * Scale, or null if none.
997 /// If the multiplication is known not to overflow then NoSignedWrap is set.
998 Value *Descale(Value *Val, APInt Scale, bool &NoSignedWrap);
1001 } // end namespace llvm
1003 #undef DEBUG_TYPE
1005 #endif // LLVM_LIB_TRANSFORMS_INSTCOMBINE_INSTCOMBINEINTERNAL_H