generated code. For more details, see :ref:`GC Transitions
<gc_transition_args>`.
-Affected Operand Bundles
-^^^^^^^^^^^^^^^^^^^^^^^^
-
-Affected operand bundles are characterized by the ``"affected"`` operand bundle
-tag. These operand bundles indicate that a call, specifically a call to an
-intrinsic like ``llvm.assume``, implies some additional knowledge about the
-values within the bundle. This enables the optimizer to efficiently find these
-relationships. The optimizer will add these automatically.
-
.. _moduleasm:
Module-Level Inline Assembly
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/GetElementPtrTypeIterator.h"
#include "llvm/Support/ErrorHandling.h"
namespace llvm {
+class AssumptionCache;
class DominatorTree;
class LoopInfo;
const DataLayout &DL;
const TargetLibraryInfo &TLI;
+ AssumptionCache &AC;
DominatorTree *DT;
LoopInfo *LI;
public:
BasicAAResult(const DataLayout &DL, const TargetLibraryInfo &TLI,
- DominatorTree *DT = nullptr,
+ AssumptionCache &AC, DominatorTree *DT = nullptr,
LoopInfo *LI = nullptr)
- : AAResultBase(), DL(DL), TLI(TLI), DT(DT), LI(LI) {}
+ : AAResultBase(), DL(DL), TLI(TLI), AC(AC), DT(DT), LI(LI) {}
BasicAAResult(const BasicAAResult &Arg)
- : AAResultBase(Arg), DL(Arg.DL), TLI(Arg.TLI), DT(Arg.DT),
+ : AAResultBase(Arg), DL(Arg.DL), TLI(Arg.TLI), AC(Arg.AC), DT(Arg.DT),
LI(Arg.LI) {}
BasicAAResult(BasicAAResult &&Arg)
- : AAResultBase(std::move(Arg)), DL(Arg.DL), TLI(Arg.TLI),
+ : AAResultBase(std::move(Arg)), DL(Arg.DL), TLI(Arg.TLI), AC(Arg.AC),
DT(Arg.DT), LI(Arg.LI) {}
AliasResult alias(const MemoryLocation &LocA, const MemoryLocation &LocB);
static const Value *
GetLinearExpression(const Value *V, APInt &Scale, APInt &Offset,
unsigned &ZExtBits, unsigned &SExtBits,
- const DataLayout &DL, unsigned Depth,
+ const DataLayout &DL, unsigned Depth, AssumptionCache *AC,
DominatorTree *DT, bool &NSW, bool &NUW);
static bool DecomposeGEPExpression(const Value *V, DecomposedGEP &Decomposed,
- const DataLayout &DL, DominatorTree *DT);
+ const DataLayout &DL, AssumptionCache *AC, DominatorTree *DT);
static bool isGEPBaseAtNegativeOffset(const GEPOperator *GEPOp,
const DecomposedGEP &DecompGEP, const DecomposedGEP &DecompObject,
bool
constantOffsetHeuristic(const SmallVectorImpl<VariableGEPIndex> &VarIndices,
uint64_t V1Size, uint64_t V2Size, int64_t BaseOffset,
- DominatorTree *DT);
+ AssumptionCache *AC, DominatorTree *DT);
bool isValueEqualInPotentialCycles(const Value *V1, const Value *V2);
#include "llvm/IR/CallSite.h"
namespace llvm {
+class AssumptionCache;
class BasicBlock;
class Loop;
class Function;
/// \brief Collect a loop's ephemeral values (those used only by an assume
/// or similar intrinsics in the loop).
- static void collectEphemeralValues(const Loop *L,
+ static void collectEphemeralValues(const Loop *L, AssumptionCache *AC,
SmallPtrSetImpl<const Value *> &EphValues);
/// \brief Collect a functions's ephemeral values (those used only by an
/// assume or similar intrinsics in the function).
- static void collectEphemeralValues(const Function *L,
+ static void collectEphemeralValues(const Function *L, AssumptionCache *AC,
SmallPtrSetImpl<const Value *> &EphValues);
};
class Function;
class Instruction;
class DominatorTree;
+class AssumptionCache;
class DemandedBits {
public:
- DemandedBits(Function &F, DominatorTree &DT) :
- F(F), DT(DT), Analyzed(false) {}
+ DemandedBits(Function &F, AssumptionCache &AC, DominatorTree &DT) :
+ F(F), AC(AC), DT(DT), Analyzed(false) {}
/// Return the bits demanded from instruction I.
APInt getDemandedBits(Instruction *I);
private:
Function &F;
+ AssumptionCache &AC;
DominatorTree &DT;
void performAnalysis();
namespace llvm {
+class AssumptionCache;
class DominatorTree;
class Instruction;
class Value;
class IVUsers {
friend class IVStrideUse;
Loop *L;
+ AssumptionCache *AC;
LoopInfo *LI;
DominatorTree *DT;
ScalarEvolution *SE;
SmallPtrSet<const Value *, 32> EphValues;
public:
- IVUsers(Loop *L, LoopInfo *LI, DominatorTree *DT,
+ IVUsers(Loop *L, AssumptionCache *AC, LoopInfo *LI, DominatorTree *DT,
ScalarEvolution *SE);
IVUsers(IVUsers &&X)
- : L(std::move(X.L)), DT(std::move(X.DT)),
+ : L(std::move(X.L)), AC(std::move(X.AC)), DT(std::move(X.DT)),
SE(std::move(X.SE)), Processed(std::move(X.Processed)),
IVUses(std::move(X.IVUses)), EphValues(std::move(X.EphValues)) {
for (IVStrideUse &U : IVUses)
#define LLVM_ANALYSIS_INLINECOST_H
#include "llvm/Analysis/CallGraphSCCPass.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include <cassert>
#include <climits>
namespace llvm {
+class AssumptionCacheTracker;
class CallSite;
class DataLayout;
class Function;
InlineCost
getInlineCost(CallSite CS, const InlineParams &Params,
TargetTransformInfo &CalleeTTI,
+ std::function<AssumptionCache &(Function &)> &GetAssumptionCache,
ProfileSummaryInfo *PSI);
/// \brief Get an InlineCost with the callee explicitly specified.
InlineCost
getInlineCost(CallSite CS, Function *Callee, const InlineParams &Params,
TargetTransformInfo &CalleeTTI,
+ std::function<AssumptionCache &(Function &)> &GetAssumptionCache,
ProfileSummaryInfo *PSI);
/// \brief Minimal filter to detect invalid constructs for inlining.
namespace llvm {
template<typename T>
class ArrayRef;
+ class AssumptionCache;
class DominatorTree;
class Instruction;
class DataLayout;
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for a Sub, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an FAdd, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an FSub, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an FMul, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for a Mul, fold the result or return null.
Value *SimplifyMulInst(Value *LHS, Value *RHS, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an SDiv, fold the result or return null.
Value *SimplifySDivInst(Value *LHS, Value *RHS, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for a UDiv, fold the result or return null.
Value *SimplifyUDivInst(Value *LHS, Value *RHS, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an FDiv, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an SRem, fold the result or return null.
Value *SimplifySRemInst(Value *LHS, Value *RHS, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for a URem, fold the result or return null.
Value *SimplifyURemInst(Value *LHS, Value *RHS, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an FRem, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for a Shl, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for a LShr, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for a AShr, fold the result or return nulll.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an And, fold the result or return null.
Value *SimplifyAndInst(Value *LHS, Value *RHS, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an Or, fold the result or return null.
Value *SimplifyOrInst(Value *LHS, Value *RHS, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an Xor, fold the result or return null.
Value *SimplifyXorInst(Value *LHS, Value *RHS, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an ICmpInst, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an FCmpInst, fold the result or return null.
FastMathFlags FMF, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for a SelectInst, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for a GetElementPtrInst, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an InsertValueInst, fold the result or return null.
ArrayRef<unsigned> Idxs, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an ExtractValueInst, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an ExtractElementInst, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for a CastInst, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
//=== Helper functions for higher up the class hierarchy.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for a BinaryOperator, fold the result or return null.
const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given operands for an FP BinaryOperator, fold the result or return null.
const FastMathFlags &FMF, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given a function and iterators over arguments, fold the result or return
User::op_iterator ArgEnd, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// Given a function and set of arguments, fold the result or return null.
Value *SimplifyCall(Value *V, ArrayRef<Value *> Args, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr);
/// See if we can compute a simplified version of this instruction. If not,
/// return null.
Value *SimplifyInstruction(Instruction *I, const DataLayout &DL,
const TargetLibraryInfo *TLI = nullptr,
- const DominatorTree *DT = nullptr);
+ const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr);
/// Replace all uses of 'I' with 'SimpleV' and simplify the uses recursively.
///
/// The function returns true if any simplifications were performed.
bool replaceAndRecursivelySimplify(Instruction *I, Value *SimpleV,
const TargetLibraryInfo *TLI = nullptr,
- const DominatorTree *DT = nullptr);
+ const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr);
/// Recursively attempt to simplify an instruction.
///
/// performed.
bool recursivelySimplifyInstruction(Instruction *I,
const TargetLibraryInfo *TLI = nullptr,
- const DominatorTree *DT = nullptr);
+ const DominatorTree *DT = nullptr,
+ AssumptionCache *AC = nullptr);
} // end namespace llvm
#endif
#include "llvm/Pass.h"
namespace llvm {
+ class AssumptionCache;
class Constant;
class ConstantRange;
class DataLayout;
/// This pass computes, caches, and vends lazy value constraint information.
class LazyValueInfo {
friend class LazyValueInfoWrapperPass;
+ AssumptionCache *AC = nullptr;
class TargetLibraryInfo *TLI = nullptr;
DominatorTree *DT = nullptr;
void *PImpl = nullptr;
public:
~LazyValueInfo();
LazyValueInfo() {}
- LazyValueInfo(TargetLibraryInfo *TLI_,
+ LazyValueInfo(AssumptionCache *AC_, TargetLibraryInfo *TLI_,
DominatorTree *DT_)
- : TLI(TLI_), DT(DT_) {}
+ : AC(AC_), TLI(TLI_), DT(DT_) {}
LazyValueInfo(LazyValueInfo &&Arg)
- : TLI(Arg.TLI), DT(Arg.DT), PImpl(Arg.PImpl) {
+ : AC(Arg.AC), TLI(Arg.TLI), DT(Arg.DT), PImpl(Arg.PImpl) {
Arg.PImpl = nullptr;
}
LazyValueInfo &operator=(LazyValueInfo &&Arg) {
releaseMemory();
+ AC = Arg.AC;
TLI = Arg.TLI;
DT = Arg.DT;
PImpl = Arg.PImpl;
class FunctionPass;
class Instruction;
class CallSite;
+class AssumptionCache;
class MemoryDependenceResults;
class PredIteratorCache;
class DominatorTree;
/// Current AA implementation, just a cache.
AliasAnalysis &AA;
+ AssumptionCache &AC;
const TargetLibraryInfo &TLI;
DominatorTree &DT;
PredIteratorCache PredCache;
public:
- MemoryDependenceResults(AliasAnalysis &AA,
+ MemoryDependenceResults(AliasAnalysis &AA, AssumptionCache &AC,
const TargetLibraryInfo &TLI,
DominatorTree &DT)
- : AA(AA), TLI(TLI), DT(DT) {}
+ : AA(AA), AC(AC), TLI(TLI), DT(DT) {}
/// Some methods limit the number of instructions they will examine.
/// The return value of this method is the default limit that will be
#include "llvm/IR/Instruction.h"
namespace llvm {
+ class AssumptionCache;
class DominatorTree;
class DataLayout;
class TargetLibraryInfo;
/// TLI - The target library info if known, otherwise null.
const TargetLibraryInfo *TLI;
+ /// A cache of @llvm.assume calls used by SimplifyInstruction.
+ AssumptionCache *AC;
+
/// InstInputs - The inputs for our symbolic address.
SmallVector<Instruction*, 4> InstInputs;
public:
- PHITransAddr(Value *addr, const DataLayout &DL)
- : Addr(addr), DL(DL), TLI(nullptr) {
+ PHITransAddr(Value *addr, const DataLayout &DL, AssumptionCache *AC)
+ : Addr(addr), DL(DL), TLI(nullptr), AC(AC) {
// If the address is an instruction, the whole thing is considered an input.
if (Instruction *I = dyn_cast<Instruction>(Addr))
InstInputs.push_back(I);
namespace llvm {
class APInt;
+class AssumptionCache;
class Constant;
class ConstantInt;
class DominatorTree;
///
TargetLibraryInfo &TLI;
+ /// The tracker for @llvm.assume intrinsics in this function.
+ AssumptionCache &AC;
+
/// The dominator tree.
///
DominatorTree &DT;
///
ValueExprMapType ValueExprMap;
- /// This is a map of SCEVs to intrinsics (e.g. assumptions) that might affect
- /// (i.e. imply something about) them.
- DenseMap<const SCEV *, SetVector<Value *>> AffectedMap;
-
/// Mark predicate values currently being processed by isImpliedCond.
SmallPtrSet<Value *, 6> PendingLoopPredicates;
ConstantRange getRangeViaFactoring(const SCEV *Start, const SCEV *Stop,
const SCEV *MaxBECount, unsigned BitWidth);
- /// Add to the AffectedMap this SCEV if its operands are in the AffectedMap.
- void addAffectedFromOperands(const SCEV *S);
-
/// We know that there is no SCEV for the specified value. Analyze the
/// expression.
const SCEV *createSCEV(Value *V);
bool isAddRecNeverPoison(const Instruction *I, const Loop *L);
public:
- ScalarEvolution(Function &F, TargetLibraryInfo &TLI,
+ ScalarEvolution(Function &F, TargetLibraryInfo &TLI, AssumptionCache &AC,
DominatorTree &DT, LoopInfo &LI);
~ScalarEvolution();
ScalarEvolution(ScalarEvolution &&Arg);
template <typename T> class ArrayRef;
class APInt;
class AddOperator;
+ class AssumptionCache;
class DataLayout;
class DominatorTree;
class GEPOperator;
/// for all of the elements in the vector.
void computeKnownBits(const Value *V, APInt &KnownZero, APInt &KnownOne,
const DataLayout &DL, unsigned Depth = 0,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
/// Compute known bits from the range metadata.
/// Return true if LHS and RHS have no common bits set.
bool haveNoCommonBitsSet(const Value *LHS, const Value *RHS,
const DataLayout &DL,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
/// wrapper around computeKnownBits.
void ComputeSignBit(const Value *V, bool &KnownZero, bool &KnownOne,
const DataLayout &DL, unsigned Depth = 0,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
/// value is either a power of two or zero.
bool isKnownToBeAPowerOfTwo(const Value *V, const DataLayout &DL,
bool OrZero = false, unsigned Depth = 0,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
/// defined. Supports values with integer or pointer type and vectors of
/// integers.
bool isKnownNonZero(const Value *V, const DataLayout &DL, unsigned Depth = 0,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
/// Returns true if the give value is known to be non-negative.
bool isKnownNonNegative(const Value *V, const DataLayout &DL,
unsigned Depth = 0,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
/// Returns true if the given value is known be positive (i.e. non-negative
/// and non-zero).
bool isKnownPositive(const Value *V, const DataLayout &DL, unsigned Depth = 0,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
/// Returns true if the given value is known be negative (i.e. non-positive
/// and non-zero).
bool isKnownNegative(const Value *V, const DataLayout &DL, unsigned Depth = 0,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
/// Return true if the given values are known to be non-equal when defined.
/// Supports scalar integer types only.
bool isKnownNonEqual(const Value *V1, const Value *V2, const DataLayout &DL,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
/// for all of the elements in the vector.
bool MaskedValueIsZero(const Value *V, const APInt &Mask,
const DataLayout &DL,
- unsigned Depth = 0,
+ unsigned Depth = 0, AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
/// sign bits for the vector element with the mininum number of known sign
/// bits.
unsigned ComputeNumSignBits(const Value *Op, const DataLayout &DL,
- unsigned Depth = 0,
+ unsigned Depth = 0, AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
OverflowResult computeOverflowForUnsignedMul(const Value *LHS,
const Value *RHS,
const DataLayout &DL,
+ AssumptionCache *AC,
const Instruction *CxtI,
const DominatorTree *DT);
OverflowResult computeOverflowForUnsignedAdd(const Value *LHS,
const Value *RHS,
const DataLayout &DL,
+ AssumptionCache *AC,
const Instruction *CxtI,
const DominatorTree *DT);
OverflowResult computeOverflowForSignedAdd(const Value *LHS, const Value *RHS,
const DataLayout &DL,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
/// This version also leverages the sign bit of Add if known.
OverflowResult computeOverflowForSignedAdd(const AddOperator *Add,
const DataLayout &DL,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
const DataLayout &DL,
bool InvertAPred = false,
unsigned Depth = 0,
+ AssumptionCache *AC = nullptr,
const Instruction *CxtI = nullptr,
const DominatorTree *DT = nullptr);
} // end namespace llvm
void initializeAlignmentFromAssumptionsPass(PassRegistry&);
void initializeAlwaysInlinerLegacyPassPass(PassRegistry&);
void initializeArgPromotionPass(PassRegistry&);
+void initializeAssumptionCacheTrackerPass(PassRegistry &);
void initializeAtomicExpandPass(PassRegistry&);
void initializeBBVectorizePass(PassRegistry&);
void initializeBDCELegacyPassPass(PassRegistry &);
#include "llvm/Transforms/Utils/ImportedFunctionsInliningStatistics.h"
namespace llvm {
+class AssumptionCacheTracker;
class CallSite;
class DataLayout;
class InlineCost;
bool InsertLifetime;
protected:
+ AssumptionCacheTracker *ACT;
ProfileSummaryInfo *PSI;
ImportedFunctionsInliningStatistics ImportedFunctionsStats;
};
PreservedAnalyses run(Function &F, FunctionAnalysisManager &AM);
// Glue for old PM.
- bool runImpl(Function &F, ScalarEvolution *SE_, DominatorTree *DT_);
+ bool runImpl(Function &F, AssumptionCache &AC, ScalarEvolution *SE_,
+ DominatorTree *DT_);
// For memory transfers, we need a common alignment for both the source and
// destination. If we have a new alignment for only one operand of a transfer
#include "llvm/ADT/SetVector.h"
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/MemoryDependenceAnalysis.h"
#include "llvm/IR/Dominators.h"
MemoryDependenceResults *MD;
DominatorTree *DT;
const TargetLibraryInfo *TLI;
+ AssumptionCache *AC;
SetVector<BasicBlock *> DeadBlocks;
OptimizationRemarkEmitter *ORE;
typedef SmallVector<gvn::AvailableValueInBlock, 64> AvailValInBlkVect;
typedef SmallVector<BasicBlock *, 64> UnavailBlkVect;
- bool runImpl(Function &F, DominatorTree &RunDT,
+ bool runImpl(Function &F, AssumptionCache &RunAC, DominatorTree &RunDT,
const TargetLibraryInfo &RunTLI, AAResults &RunAA,
MemoryDependenceResults *RunMD, LoopInfo *LI,
OptimizationRemarkEmitter *ORE);
#include "llvm/ADT/STLExtras.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/GlobalsModRef.h"
#include "llvm/Analysis/MemoryDependenceAnalysis.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
MemoryDependenceResults *MD = nullptr;
TargetLibraryInfo *TLI = nullptr;
std::function<AliasAnalysis &()> LookupAliasAnalysis;
+ std::function<AssumptionCache &()> LookupAssumptionCache;
std::function<DominatorTree &()> LookupDomTree;
public:
bool runImpl(Function &F, MemoryDependenceResults *MD_,
TargetLibraryInfo *TLI_,
std::function<AliasAnalysis &()> LookupAliasAnalysis_,
+ std::function<AssumptionCache &()> LookupAssumptionCache_,
std::function<DominatorTree &()> LookupDomTree_);
private:
#include "llvm/ADT/DenseMap.h"
#include "llvm/ADT/SmallVector.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/ScalarEvolution.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/Analysis/TargetTransformInfo.h"
PreservedAnalyses run(Function &F, FunctionAnalysisManager &AM);
// Glue for old PM.
- bool runImpl(Function &F, DominatorTree *DT_,
+ bool runImpl(Function &F, AssumptionCache *AC_, DominatorTree *DT_,
ScalarEvolution *SE_, TargetLibraryInfo *TLI_,
TargetTransformInfo *TTI_);
// to be an index of GEP.
bool requiresSignExtension(Value *Index, GetElementPtrInst *GEP);
+ AssumptionCache *AC;
const DataLayout *DL;
DominatorTree *DT;
ScalarEvolution *SE;
#define LLVM_TRANSFORMS_SCALAR_SROA_H
#include "llvm/ADT/SetVector.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/IR/Dominators.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/PassManager.h"
namespace llvm {
-class AllocaInst;
-class SelectInst;
-class PHINode;
/// A private "module" namespace for types and utilities used by SROA. These
/// are implementation details and should not be used by clients.
class SROA : public PassInfoMixin<SROA> {
LLVMContext *C;
DominatorTree *DT;
+ AssumptionCache *AC;
/// \brief Worklist of alloca instructions to simplify.
///
SetVector<SelectInst *, SmallVector<SelectInst *, 2>> SpeculatableSelects;
public:
- SROA() : C(nullptr), DT(nullptr) {}
+ SROA() : C(nullptr), DT(nullptr), AC(nullptr) {}
/// \brief Run the pass over the function.
PreservedAnalyses run(Function &F, FunctionAnalysisManager &AM);
friend class sroa::SROALegacyPass;
/// Helper used by both the public run method and by the legacy pass.
- PreservedAnalyses runImpl(Function &F, DominatorTree &RunDT);
+ PreservedAnalyses runImpl(Function &F, DominatorTree &RunDT,
+ AssumptionCache &RunAC);
bool presplitLoadsAndStores(AllocaInst &AI, sroa::AllocaSlices &AS);
AllocaInst *rewritePartition(AllocaInst &AI, sroa::AllocaSlices &AS,
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/Twine.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/IR/ValueHandle.h"
#include "llvm/IR/ValueMap.h"
#include "llvm/Transforms/Utils/ValueMapper.h"
class Loop;
class LoopInfo;
class AllocaInst;
+class AssumptionCacheTracker;
class DominatorTree;
/// Return an exact copy of the specified module
/// InlineFunction call, and records the auxiliary results produced by it.
class InlineFunctionInfo {
public:
- explicit InlineFunctionInfo(CallGraph *cg = nullptr)
- : CG(cg) {}
+ explicit InlineFunctionInfo(CallGraph *cg = nullptr,
+ std::function<AssumptionCache &(Function &)>
+ *GetAssumptionCache = nullptr)
+ : CG(cg), GetAssumptionCache(GetAssumptionCache) {}
/// CG - If non-null, InlineFunction will update the callgraph to reflect the
/// changes it makes.
CallGraph *CG;
+ std::function<AssumptionCache &(Function &)> *GetAssumptionCache;
/// StaticAllocas - InlineFunction fills this in with all static allocas that
/// get copied into the caller.
class Value;
class PHINode;
class AllocaInst;
+class AssumptionCache;
class ConstantExpr;
class DataLayout;
class TargetLibraryInfo;
/// parameter, providing the set of loop header that SimplifyCFG should not
/// eliminate.
bool SimplifyCFG(BasicBlock *BB, const TargetTransformInfo &TTI,
- unsigned BonusInstThreshold,
+ unsigned BonusInstThreshold, AssumptionCache *AC = nullptr,
SmallPtrSetImpl<BasicBlock *> *LoopHeaders = nullptr);
/// This function is used to flatten a CFG. For example, it uses parallel-and
unsigned getOrEnforceKnownAlignment(Value *V, unsigned PrefAlign,
const DataLayout &DL,
const Instruction *CxtI = nullptr,
+ AssumptionCache *AC = nullptr,
const DominatorTree *DT = nullptr);
/// Try to infer an alignment for the specified pointer.
static inline unsigned getKnownAlignment(Value *V, const DataLayout &DL,
const Instruction *CxtI = nullptr,
+ AssumptionCache *AC = nullptr,
const DominatorTree *DT = nullptr) {
- return getOrEnforceKnownAlignment(V, 0, DL, CxtI, DT);
+ return getOrEnforceKnownAlignment(V, 0, DL, CxtI, AC, DT);
}
/// Given a getelementptr instruction/constantexpr, emit the code necessary to
#ifndef LLVM_TRANSFORMS_UTILS_LOOPSIMPLIFY_H
#define LLVM_TRANSFORMS_UTILS_LOOPSIMPLIFY_H
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/ScalarEvolution.h"
#include "llvm/IR/Dominators.h"
#include "llvm/IR/PassManager.h"
/// it into a simplified loop nest with preheaders and single backedges. It will
/// update \c AliasAnalysis and \c ScalarEvolution analyses if they're non-null.
bool simplifyLoop(Loop *L, DominatorTree *DT, LoopInfo *LI, ScalarEvolution *SE,
- bool PreserveLCSSA);
+ AssumptionCache *AC, bool PreserveLCSSA);
} // end namespace llvm
namespace llvm {
class AliasSet;
class AliasSetTracker;
+class AssumptionCache;
class BasicBlock;
class DataLayout;
class DominatorTree;
if (WalkingPhi && Location.Ptr) {
PHITransAddr Translator(
const_cast<Value *>(Location.Ptr),
- OriginalAccess->getBlock()->getModule()->getDataLayout());
+ OriginalAccess->getBlock()->getModule()->getDataLayout(), nullptr);
if (!Translator.PHITranslateValue(OriginalAccess->getBlock(),
DefIterator.getPhiArgBlock(), nullptr,
false))
class AllocaInst;
class DominatorTree;
class AliasSetTracker;
+class AssumptionCache;
/// \brief Return true if this alloca is legal for promotion.
///
/// If AST is specified, the specified tracker is updated to reflect changes
/// made to the IR.
void PromoteMemToReg(ArrayRef<AllocaInst *> Allocas, DominatorTree &DT,
- AliasSetTracker *AST = nullptr);
+ AliasSetTracker *AST = nullptr,
+ AssumptionCache *AC = nullptr);
} // End llvm namespace
namespace llvm {
class StringRef;
+class AssumptionCache;
class DominatorTree;
class Loop;
class LoopInfo;
bool AllowRuntime, bool AllowExpensiveTripCount,
bool PreserveCondBr, bool PreserveOnlyFirst,
unsigned TripMultiple, unsigned PeelCount, LoopInfo *LI,
- ScalarEvolution *SE, DominatorTree *DT,
+ ScalarEvolution *SE, DominatorTree *DT, AssumptionCache *AC,
OptimizationRemarkEmitter *ORE, bool PreserveLCSSA);
bool UnrollRuntimeLoopRemainder(Loop *L, unsigned Count,
#include "llvm/ADT/MapVector.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/BasicAliasAnalysis.h"
#include "llvm/Analysis/BlockFrequencyInfo.h"
#include "llvm/Analysis/DemandedBits.h"
TargetLibraryInfo *TLI;
DemandedBits *DB;
AliasAnalysis *AA;
+ AssumptionCache *AC;
std::function<const LoopAccessInfo &(Loop &)> *GetLAA;
OptimizationRemarkEmitter *ORE;
bool runImpl(Function &F, ScalarEvolution &SE_, LoopInfo &LI_,
TargetTransformInfo &TTI_, DominatorTree &DT_,
BlockFrequencyInfo &BFI_, TargetLibraryInfo *TLI_,
- DemandedBits &DB_, AliasAnalysis &AA_,
+ DemandedBits &DB_, AliasAnalysis &AA_, AssumptionCache &AC_,
std::function<const LoopAccessInfo &(Loop &)> &GetLAA_,
OptimizationRemarkEmitter &ORE);
#include "llvm/ADT/MapVector.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/DemandedBits.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/ScalarEvolution.h"
AliasAnalysis *AA = nullptr;
LoopInfo *LI = nullptr;
DominatorTree *DT = nullptr;
+ AssumptionCache *AC = nullptr;
DemandedBits *DB = nullptr;
const DataLayout *DL = nullptr;
// Glue for old PM.
bool runImpl(Function &F, ScalarEvolution *SE_, TargetTransformInfo *TTI_,
TargetLibraryInfo *TLI_, AliasAnalysis *AA_, LoopInfo *LI_,
- DominatorTree *DT_, DemandedBits *DB_);
+ DominatorTree *DT_, AssumptionCache *AC_, DemandedBits *DB_);
private:
/// \brief Collect store and getelementptr instructions and organize them
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/MemoryBuiltins.h"
#include "llvm/Analysis/ValueTracking.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/IR/Constants.h"
#include "llvm/IR/DataLayout.h"
#include "llvm/IR/DerivedTypes.h"
/*static*/ const Value *BasicAAResult::GetLinearExpression(
const Value *V, APInt &Scale, APInt &Offset, unsigned &ZExtBits,
unsigned &SExtBits, const DataLayout &DL, unsigned Depth,
- DominatorTree *DT, bool &NSW, bool &NUW) {
+ AssumptionCache *AC, DominatorTree *DT, bool &NSW, bool &NUW) {
assert(V->getType()->isIntegerTy() && "Not an integer value");
// Limit our recursion depth.
case Instruction::Or:
// X|C == X+C if all the bits in C are unset in X. Otherwise we can't
// analyze it.
- if (!MaskedValueIsZero(BOp->getOperand(0), RHSC->getValue(), DL, 0,
+ if (!MaskedValueIsZero(BOp->getOperand(0), RHSC->getValue(), DL, 0, AC,
BOp, DT)) {
Scale = 1;
Offset = 0;
LLVM_FALLTHROUGH;
case Instruction::Add:
V = GetLinearExpression(BOp->getOperand(0), Scale, Offset, ZExtBits,
- SExtBits, DL, Depth + 1, DT, NSW, NUW);
+ SExtBits, DL, Depth + 1, AC, DT, NSW, NUW);
Offset += RHS;
break;
case Instruction::Sub:
V = GetLinearExpression(BOp->getOperand(0), Scale, Offset, ZExtBits,
- SExtBits, DL, Depth + 1, DT, NSW, NUW);
+ SExtBits, DL, Depth + 1, AC, DT, NSW, NUW);
Offset -= RHS;
break;
case Instruction::Mul:
V = GetLinearExpression(BOp->getOperand(0), Scale, Offset, ZExtBits,
- SExtBits, DL, Depth + 1, DT, NSW, NUW);
+ SExtBits, DL, Depth + 1, AC, DT, NSW, NUW);
Offset *= RHS;
Scale *= RHS;
break;
case Instruction::Shl:
V = GetLinearExpression(BOp->getOperand(0), Scale, Offset, ZExtBits,
- SExtBits, DL, Depth + 1, DT, NSW, NUW);
+ SExtBits, DL, Depth + 1, AC, DT, NSW, NUW);
Offset <<= RHS.getLimitedValue();
Scale <<= RHS.getLimitedValue();
// the semantics of nsw and nuw for left shifts don't match those of
unsigned OldZExtBits = ZExtBits, OldSExtBits = SExtBits;
const Value *Result =
GetLinearExpression(CastOp, Scale, Offset, ZExtBits, SExtBits, DL,
- Depth + 1, DT, NSW, NUW);
+ Depth + 1, AC, DT, NSW, NUW);
// zext(zext(%x)) == zext(%x), and similarly for sext; we'll handle this
// by just incrementing the number of bits we've extended by.
/// depth (MaxLookupSearchDepth). When DataLayout not is around, it just looks
/// through pointer casts.
bool BasicAAResult::DecomposeGEPExpression(const Value *V,
- DecomposedGEP &Decomposed, const DataLayout &DL,
+ DecomposedGEP &Decomposed, const DataLayout &DL, AssumptionCache *AC,
DominatorTree *DT) {
// Limit recursion depth to limit compile time in crazy cases.
unsigned MaxLookup = MaxLookupSearchDepth;
// If it's not a GEP, hand it off to SimplifyInstruction to see if it
// can come up with something. This matches what GetUnderlyingObject does.
if (const Instruction *I = dyn_cast<Instruction>(V))
- // TODO: Get a DominatorTree and use it here (it is now available in
- // this function, but this should be updated when GetUnderlyingObject
- // is updated). TLI should be provided also.
+ // TODO: Get a DominatorTree and AssumptionCache and use them here
+ // (these are both now available in this function, but this should be
+ // updated when GetUnderlyingObject is updated). TLI should be
+ // provided also.
if (const Value *Simplified =
SimplifyInstruction(const_cast<Instruction *>(I), DL)) {
V = Simplified;
APInt IndexScale(Width, 0), IndexOffset(Width, 0);
bool NSW = true, NUW = true;
Index = GetLinearExpression(Index, IndexScale, IndexOffset, ZExtBits,
- SExtBits, DL, 0, DT, NSW, NUW);
+ SExtBits, DL, 0, AC, DT, NSW, NUW);
// The GEP index scale ("Scale") scales C1*V+C2, yielding (C1*V+C2)*Scale.
// This gives us an aggregate computation of (C1*Scale)*V + C2*Scale.
const Value *UnderlyingV2) {
DecomposedGEP DecompGEP1, DecompGEP2;
bool GEP1MaxLookupReached =
- DecomposeGEPExpression(GEP1, DecompGEP1, DL, DT);
+ DecomposeGEPExpression(GEP1, DecompGEP1, DL, &AC, DT);
bool GEP2MaxLookupReached =
- DecomposeGEPExpression(V2, DecompGEP2, DL, DT);
+ DecomposeGEPExpression(V2, DecompGEP2, DL, &AC, DT);
int64_t GEP1BaseOffset = DecompGEP1.StructOffset + DecompGEP1.OtherOffset;
int64_t GEP2BaseOffset = DecompGEP2.StructOffset + DecompGEP2.OtherOffset;
bool SignKnownZero, SignKnownOne;
ComputeSignBit(const_cast<Value *>(V), SignKnownZero, SignKnownOne, DL,
- 0, nullptr, DT);
+ 0, &AC, nullptr, DT);
// Zero-extension widens the variable, and so forces the sign
// bit to zero.
return NoAlias;
if (constantOffsetHeuristic(DecompGEP1.VarIndices, V1Size, V2Size,
- GEP1BaseOffset, DT))
+ GEP1BaseOffset, &AC, DT))
return NoAlias;
}
bool BasicAAResult::constantOffsetHeuristic(
const SmallVectorImpl<VariableGEPIndex> &VarIndices, uint64_t V1Size,
- uint64_t V2Size, int64_t BaseOffset,
+ uint64_t V2Size, int64_t BaseOffset, AssumptionCache *AC,
DominatorTree *DT) {
if (VarIndices.size() != 2 || V1Size == MemoryLocation::UnknownSize ||
V2Size == MemoryLocation::UnknownSize)
bool NSW = true, NUW = true;
unsigned V0ZExtBits = 0, V0SExtBits = 0, V1ZExtBits = 0, V1SExtBits = 0;
const Value *V0 = GetLinearExpression(Var0.V, V0Scale, V0Offset, V0ZExtBits,
- V0SExtBits, DL, 0, DT, NSW, NUW);
+ V0SExtBits, DL, 0, AC, DT, NSW, NUW);
NSW = true;
NUW = true;
const Value *V1 = GetLinearExpression(Var1.V, V1Scale, V1Offset, V1ZExtBits,
- V1SExtBits, DL, 0, DT, NSW, NUW);
+ V1SExtBits, DL, 0, AC, DT, NSW, NUW);
if (V0Scale != V1Scale || V0ZExtBits != V1ZExtBits ||
V0SExtBits != V1SExtBits || !isValueEqualInPotentialCycles(V0, V1))
BasicAAResult BasicAA::run(Function &F, FunctionAnalysisManager &AM) {
return BasicAAResult(F.getParent()->getDataLayout(),
AM.getResult<TargetLibraryAnalysis>(F),
+ AM.getResult<AssumptionAnalysis>(F),
&AM.getResult<DominatorTreeAnalysis>(F),
AM.getCachedResult<LoopAnalysis>(F));
}
INITIALIZE_PASS_BEGIN(BasicAAWrapperPass, "basicaa",
"Basic Alias Analysis (stateless AA impl)", true, true)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
INITIALIZE_PASS_END(BasicAAWrapperPass, "basicaa",
}
bool BasicAAWrapperPass::runOnFunction(Function &F) {
+ auto &ACT = getAnalysis<AssumptionCacheTracker>();
auto &TLIWP = getAnalysis<TargetLibraryInfoWrapperPass>();
auto &DTWP = getAnalysis<DominatorTreeWrapperPass>();
auto *LIWP = getAnalysisIfAvailable<LoopInfoWrapperPass>();
Result.reset(new BasicAAResult(F.getParent()->getDataLayout(), TLIWP.getTLI(),
- &DTWP.getDomTree(),
+ ACT.getAssumptionCache(F), &DTWP.getDomTree(),
LIWP ? &LIWP->getLoopInfo() : nullptr));
return false;
void BasicAAWrapperPass::getAnalysisUsage(AnalysisUsage &AU) const {
AU.setPreservesAll();
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<DominatorTreeWrapperPass>();
AU.addRequired<TargetLibraryInfoWrapperPass>();
}
BasicAAResult llvm::createLegacyPMBasicAAResult(Pass &P, Function &F) {
return BasicAAResult(
F.getParent()->getDataLayout(),
- P.getAnalysis<TargetLibraryInfoWrapperPass>().getTLI());
+ P.getAnalysis<TargetLibraryInfoWrapperPass>().getTLI(),
+ P.getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F));
}
AliasAnalysisSummary.cpp
AliasSetTracker.cpp
Analysis.cpp
+ AssumptionCache.cpp
BasicAliasAnalysis.cpp
BlockFrequencyInfo.cpp
BlockFrequencyInfoImpl.cpp
//
//===----------------------------------------------------------------------===//
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/CodeMetrics.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/TargetTransformInfo.h"
// Find all ephemeral values.
void CodeMetrics::collectEphemeralValues(
- const Loop *L, SmallPtrSetImpl<const Value *> &EphValues) {
+ const Loop *L, AssumptionCache *AC,
+ SmallPtrSetImpl<const Value *> &EphValues) {
SmallPtrSet<const Value *, 32> Visited;
SmallVector<const Value *, 16> Worklist;
- for (auto &B : L->blocks())
- for (auto &I : *B)
- if (auto *II = dyn_cast<IntrinsicInst>(&I))
- if (II->getIntrinsicID() == Intrinsic::assume &&
- EphValues.insert(II).second)
- appendSpeculatableOperands(II, Visited, Worklist);
+ for (auto &AssumeVH : AC->assumptions()) {
+ if (!AssumeVH)
+ continue;
+ Instruction *I = cast<Instruction>(AssumeVH);
+
+ // Filter out call sites outside of the loop so we don't do a function's
+ // worth of work for each of its loops (and, in the common case, ephemeral
+ // values in the loop are likely due to @llvm.assume calls in the loop).
+ if (!L->contains(I->getParent()))
+ continue;
+
+ if (EphValues.insert(I).second)
+ appendSpeculatableOperands(I, Visited, Worklist);
+ }
completeEphemeralValues(Visited, Worklist, EphValues);
}
void CodeMetrics::collectEphemeralValues(
- const Function *F, SmallPtrSetImpl<const Value *> &EphValues) {
+ const Function *F, AssumptionCache *AC,
+ SmallPtrSetImpl<const Value *> &EphValues) {
SmallPtrSet<const Value *, 32> Visited;
SmallVector<const Value *, 16> Worklist;
- for (auto &B : *F)
- for (auto &I : B)
- if (auto *II = dyn_cast<IntrinsicInst>(&I))
- if (II->getIntrinsicID() == Intrinsic::assume &&
- EphValues.insert(II).second)
- appendSpeculatableOperands(II, Visited, Worklist);
+ for (auto &AssumeVH : AC->assumptions()) {
+ if (!AssumeVH)
+ continue;
+ Instruction *I = cast<Instruction>(AssumeVH);
+ assert(I->getParent()->getParent() == F &&
+ "Found assumption for the wrong function!");
+
+ if (EphValues.insert(I).second)
+ appendSpeculatableOperands(I, Visited, Worklist);
+ }
completeEphemeralValues(Visited, Worklist, EphValues);
}
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/StringExtras.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/ValueTracking.h"
#include "llvm/IR/BasicBlock.h"
#include "llvm/IR/CFG.h"
char DemandedBitsWrapperPass::ID = 0;
INITIALIZE_PASS_BEGIN(DemandedBitsWrapperPass, "demanded-bits",
"Demanded bits analysis", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_END(DemandedBitsWrapperPass, "demanded-bits",
"Demanded bits analysis", false, false)
void DemandedBitsWrapperPass::getAnalysisUsage(AnalysisUsage &AU) const {
AU.setPreservesCFG();
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<DominatorTreeWrapperPass>();
AU.setPreservesAll();
}
KnownZero = APInt(BitWidth, 0);
KnownOne = APInt(BitWidth, 0);
computeKnownBits(const_cast<Value *>(V1), KnownZero, KnownOne, DL, 0,
- UserI, &DT);
+ &AC, UserI, &DT);
if (V2) {
KnownZero2 = APInt(BitWidth, 0);
KnownOne2 = APInt(BitWidth, 0);
computeKnownBits(const_cast<Value *>(V2), KnownZero2, KnownOne2, DL,
- 0, UserI, &DT);
+ 0, &AC, UserI, &DT);
}
};
}
bool DemandedBitsWrapperPass::runOnFunction(Function &F) {
+ auto &AC = getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
auto &DT = getAnalysis<DominatorTreeWrapperPass>().getDomTree();
- DB.emplace(F, DT);
+ DB.emplace(F, AC, DT);
return false;
}
DemandedBits DemandedBitsAnalysis::run(Function &F,
FunctionAnalysisManager &AM) {
+ auto &AC = AM.getResult<AssumptionAnalysis>(F);
auto &DT = AM.getResult<DominatorTreeAnalysis>(F);
- return DemandedBits(F, DT);
+ return DemandedBits(F, AC, DT);
}
PreservedAnalyses DemandedBitsPrinterPass::run(Function &F,
#include "llvm/Analysis/IVUsers.h"
#include "llvm/ADT/STLExtras.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/CodeMetrics.h"
#include "llvm/Analysis/LoopPass.h"
#include "llvm/Analysis/LoopPassManager.h"
AM.getResult<FunctionAnalysisManagerLoopProxy>(L).getManager();
Function *F = L.getHeader()->getParent();
- return IVUsers(&L, FAM.getCachedResult<LoopAnalysis>(*F),
+ return IVUsers(&L, FAM.getCachedResult<AssumptionAnalysis>(*F),
+ FAM.getCachedResult<LoopAnalysis>(*F),
FAM.getCachedResult<DominatorTreeAnalysis>(*F),
FAM.getCachedResult<ScalarEvolutionAnalysis>(*F));
}
char IVUsersWrapperPass::ID = 0;
INITIALIZE_PASS_BEGIN(IVUsersWrapperPass, "iv-users",
"Induction Variable Users", false, true)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(LoopInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(ScalarEvolutionWrapperPass)
return IVUses.back();
}
-IVUsers::IVUsers(Loop *L, LoopInfo *LI, DominatorTree *DT, ScalarEvolution *SE)
- : L(L), LI(LI), DT(DT), SE(SE), IVUses() {
+IVUsers::IVUsers(Loop *L, AssumptionCache *AC, LoopInfo *LI, DominatorTree *DT,
+ ScalarEvolution *SE)
+ : L(L), AC(AC), LI(LI), DT(DT), SE(SE), IVUses() {
// Collect ephemeral values so that AddUsersIfInteresting skips them.
EphValues.clear();
- CodeMetrics::collectEphemeralValues(L, EphValues);
+ CodeMetrics::collectEphemeralValues(L, AC, EphValues);
// Find all uses of induction variables in this loop, and categorize
// them by stride. Start by finding all of the PHI nodes in the header for
}
void IVUsersWrapperPass::getAnalysisUsage(AnalysisUsage &AU) const {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<LoopInfoWrapperPass>();
AU.addRequired<DominatorTreeWrapperPass>();
AU.addRequired<ScalarEvolutionWrapperPass>();
}
bool IVUsersWrapperPass::runOnLoop(Loop *L, LPPassManager &LPM) {
+ auto *AC = &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(
+ *L->getHeader()->getParent());
auto *LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo();
auto *DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
auto *SE = &getAnalysis<ScalarEvolutionWrapperPass>().getSE();
- IU.reset(new IVUsers(L, LI, DT, SE));
+ IU.reset(new IVUsers(L, AC, LI, DT, SE));
return false;
}
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/Statistic.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/CodeMetrics.h"
#include "llvm/Analysis/ConstantFolding.h"
#include "llvm/Analysis/InstructionSimplify.h"
/// The TargetTransformInfo available for this compilation.
const TargetTransformInfo &TTI;
+ /// Getter for the cache of @llvm.assume intrinsics.
+ std::function<AssumptionCache &(Function &)> &GetAssumptionCache;
+
/// Profile summary information.
ProfileSummaryInfo *PSI;
public:
CallAnalyzer(const TargetTransformInfo &TTI,
+ std::function<AssumptionCache &(Function &)> &GetAssumptionCache,
ProfileSummaryInfo *PSI, Function &Callee, CallSite CSArg,
const InlineParams &Params)
- : TTI(TTI), PSI(PSI), F(Callee), CandidateCS(CSArg), Params(Params),
- Threshold(Params.DefaultThreshold), Cost(0), IsCallerRecursive(false),
- IsRecursiveCall(false), ExposesReturnsTwice(false),
- HasDynamicAlloca(false), ContainsNoDuplicateCall(false),
- HasReturn(false), HasIndirectBr(false), HasFrameEscape(false),
- AllocatedSize(0), NumInstructions(0), NumVectorInstructions(0),
- FiftyPercentVectorBonus(0), TenPercentVectorBonus(0), VectorBonus(0),
- NumConstantArgs(0), NumConstantOffsetPtrArgs(0), NumAllocaArgs(0),
- NumConstantPtrCmps(0), NumConstantPtrDiffs(0),
- NumInstructionsSimplified(0), SROACostSavings(0),
- SROACostSavingsLost(0) {}
+ : TTI(TTI), GetAssumptionCache(GetAssumptionCache), PSI(PSI), F(Callee),
+ CandidateCS(CSArg), Params(Params), Threshold(Params.DefaultThreshold),
+ Cost(0), IsCallerRecursive(false), IsRecursiveCall(false),
+ ExposesReturnsTwice(false), HasDynamicAlloca(false),
+ ContainsNoDuplicateCall(false), HasReturn(false), HasIndirectBr(false),
+ HasFrameEscape(false), AllocatedSize(0), NumInstructions(0),
+ NumVectorInstructions(0), FiftyPercentVectorBonus(0),
+ TenPercentVectorBonus(0), VectorBonus(0), NumConstantArgs(0),
+ NumConstantOffsetPtrArgs(0), NumAllocaArgs(0), NumConstantPtrCmps(0),
+ NumConstantPtrDiffs(0), NumInstructionsSimplified(0),
+ SROACostSavings(0), SROACostSavingsLost(0) {}
bool analyzeCall(CallSite CS);
// out. Pretend to inline the function, with a custom threshold.
auto IndirectCallParams = Params;
IndirectCallParams.DefaultThreshold = InlineConstants::IndirectCallThreshold;
- CallAnalyzer CA(TTI, PSI, *F, CS, IndirectCallParams);
+ CallAnalyzer CA(TTI, GetAssumptionCache, PSI, *F, CS, IndirectCallParams);
if (CA.analyzeCall(CS)) {
// We were able to inline the indirect call! Subtract the cost from the
// threshold to get the bonus we want to apply, but don't go below zero.
// the ephemeral values multiple times (and they're completely determined by
// the callee, so this is purely duplicate work).
SmallPtrSet<const Value *, 32> EphValues;
- CodeMetrics::collectEphemeralValues(&F, EphValues);
+ CodeMetrics::collectEphemeralValues(&F, &GetAssumptionCache(F), EphValues);
// The worklist of live basic blocks in the callee *after* inlining. We avoid
// adding basic blocks of the callee which can be proven to be dead for this
InlineCost llvm::getInlineCost(
CallSite CS, const InlineParams &Params, TargetTransformInfo &CalleeTTI,
+ std::function<AssumptionCache &(Function &)> &GetAssumptionCache,
ProfileSummaryInfo *PSI) {
- return getInlineCost(CS, CS.getCalledFunction(), Params, CalleeTTI, PSI);
+ return getInlineCost(CS, CS.getCalledFunction(), Params, CalleeTTI,
+ GetAssumptionCache, PSI);
}
InlineCost llvm::getInlineCost(
CallSite CS, Function *Callee, const InlineParams &Params,
- TargetTransformInfo &CalleeTTI, ProfileSummaryInfo *PSI) {
+ TargetTransformInfo &CalleeTTI,
+ std::function<AssumptionCache &(Function &)> &GetAssumptionCache,
+ ProfileSummaryInfo *PSI) {
// Cannot inline indirect calls.
if (!Callee)
DEBUG(llvm::dbgs() << " Analyzing call of " << Callee->getName()
<< "...\n");
- CallAnalyzer CA(CalleeTTI, PSI, *Callee, CS, Params);
+ CallAnalyzer CA(CalleeTTI, GetAssumptionCache, PSI, *Callee, CS, Params);
bool ShouldInline = CA.analyzeCall(CS);
DEBUG(CA.dump());
const DataLayout &DL;
const TargetLibraryInfo *TLI;
const DominatorTree *DT;
+ AssumptionCache *AC;
const Instruction *CxtI;
Query(const DataLayout &DL, const TargetLibraryInfo *tli,
- const DominatorTree *dt, const Instruction *cxti = nullptr)
- : DL(DL), TLI(tli), DT(dt), CxtI(cxti) {}
+ const DominatorTree *dt, AssumptionCache *ac = nullptr,
+ const Instruction *cxti = nullptr)
+ : DL(DL), TLI(tli), DT(dt), AC(ac), CxtI(cxti) {}
};
} // end anonymous namespace
Value *llvm::SimplifyAddInst(Value *Op0, Value *Op1, bool isNSW, bool isNUW,
const DataLayout &DL, const TargetLibraryInfo *TLI,
- const DominatorTree *DT, const Instruction *CxtI) {
- return ::SimplifyAddInst(Op0, Op1, isNSW, isNUW, Query(DL, TLI, DT, CxtI),
+ const DominatorTree *DT, AssumptionCache *AC,
+ const Instruction *CxtI) {
+ return ::SimplifyAddInst(Op0, Op1, isNSW, isNUW, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
unsigned BitWidth = Op1->getType()->getScalarSizeInBits();
APInt KnownZero(BitWidth, 0);
APInt KnownOne(BitWidth, 0);
- computeKnownBits(Op1, KnownZero, KnownOne, Q.DL, 0, Q.CxtI, Q.DT);
+ computeKnownBits(Op1, KnownZero, KnownOne, Q.DL, 0, Q.AC, Q.CxtI, Q.DT);
if (KnownZero == ~APInt::getSignBit(BitWidth)) {
// Op1 is either 0 or the minimum signed value. If the sub is NSW, then
// Op1 must be 0 because negating the minimum signed value is undefined.
Value *llvm::SimplifySubInst(Value *Op0, Value *Op1, bool isNSW, bool isNUW,
const DataLayout &DL, const TargetLibraryInfo *TLI,
- const DominatorTree *DT, const Instruction *CxtI) {
- return ::SimplifySubInst(Op0, Op1, isNSW, isNUW, Query(DL, TLI, DT, CxtI),
+ const DominatorTree *DT, AssumptionCache *AC,
+ const Instruction *CxtI) {
+ return ::SimplifySubInst(Op0, Op1, isNSW, isNUW, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyFAddInst(Value *Op0, Value *Op1, FastMathFlags FMF,
const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyFAddInst(Op0, Op1, FMF, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyFAddInst(Op0, Op1, FMF, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyFSubInst(Value *Op0, Value *Op1, FastMathFlags FMF,
const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyFSubInst(Op0, Op1, FMF, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyFSubInst(Op0, Op1, FMF, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyFMulInst(Value *Op0, Value *Op1, FastMathFlags FMF,
const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyFMulInst(Op0, Op1, FMF, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyFMulInst(Op0, Op1, FMF, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyMulInst(Value *Op0, Value *Op1, const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyMulInst(Op0, Op1, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyMulInst(Op0, Op1, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifySDivInst(Value *Op0, Value *Op1, const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifySDivInst(Op0, Op1, Query(DL, TLI, DT, CxtI),
+ return ::SimplifySDivInst(Op0, Op1, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyUDivInst(Value *Op0, Value *Op1, const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyUDivInst(Op0, Op1, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyUDivInst(Op0, Op1, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyFDivInst(Value *Op0, Value *Op1, FastMathFlags FMF,
const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyFDivInst(Op0, Op1, FMF, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyFDivInst(Op0, Op1, FMF, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifySRemInst(Value *Op0, Value *Op1, const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifySRemInst(Op0, Op1, Query(DL, TLI, DT, CxtI),
+ return ::SimplifySRemInst(Op0, Op1, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyURemInst(Value *Op0, Value *Op1, const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyURemInst(Op0, Op1, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyURemInst(Op0, Op1, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyFRemInst(Value *Op0, Value *Op1, FastMathFlags FMF,
const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyFRemInst(Op0, Op1, FMF, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyFRemInst(Op0, Op1, FMF, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
unsigned BitWidth = Op1->getType()->getScalarSizeInBits();
APInt KnownZero(BitWidth, 0);
APInt KnownOne(BitWidth, 0);
- computeKnownBits(Op1, KnownZero, KnownOne, Q.DL, 0, Q.CxtI, Q.DT);
+ computeKnownBits(Op1, KnownZero, KnownOne, Q.DL, 0, Q.AC, Q.CxtI, Q.DT);
if (KnownOne.getLimitedValue() >= BitWidth)
return UndefValue::get(Op0->getType());
unsigned BitWidth = Op0->getType()->getScalarSizeInBits();
APInt Op0KnownZero(BitWidth, 0);
APInt Op0KnownOne(BitWidth, 0);
- computeKnownBits(Op0, Op0KnownZero, Op0KnownOne, Q.DL, /*Depth=*/0, Q.CxtI,
- Q.DT);
+ computeKnownBits(Op0, Op0KnownZero, Op0KnownOne, Q.DL, /*Depth=*/0, Q.AC,
+ Q.CxtI, Q.DT);
if (Op0KnownOne[0])
return Op0;
}
Value *llvm::SimplifyShlInst(Value *Op0, Value *Op1, bool isNSW, bool isNUW,
const DataLayout &DL, const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyShlInst(Op0, Op1, isNSW, isNUW, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyShlInst(Op0, Op1, isNSW, isNUW, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyLShrInst(Value *Op0, Value *Op1, bool isExact,
const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyLShrInst(Op0, Op1, isExact, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyLShrInst(Op0, Op1, isExact, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
return X;
// Arithmetic shifting an all-sign-bit value is a no-op.
- unsigned NumSignBits = ComputeNumSignBits(Op0, Q.DL, 0, Q.CxtI, Q.DT);
+ unsigned NumSignBits = ComputeNumSignBits(Op0, Q.DL, 0, Q.AC, Q.CxtI, Q.DT);
if (NumSignBits == Op0->getType()->getScalarSizeInBits())
return Op0;
Value *llvm::SimplifyAShrInst(Value *Op0, Value *Op1, bool isExact,
const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyAShrInst(Op0, Op1, isExact, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyAShrInst(Op0, Op1, isExact, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
// A & (-A) = A if A is a power of two or zero.
if (match(Op0, m_Neg(m_Specific(Op1))) ||
match(Op1, m_Neg(m_Specific(Op0)))) {
- if (isKnownToBeAPowerOfTwo(Op0, Q.DL, /*OrZero*/ true, 0, Q.CxtI, Q.DT))
+ if (isKnownToBeAPowerOfTwo(Op0, Q.DL, /*OrZero*/ true, 0, Q.AC, Q.CxtI,
+ Q.DT))
return Op0;
- if (isKnownToBeAPowerOfTwo(Op1, Q.DL, /*OrZero*/ true, 0, Q.CxtI, Q.DT))
+ if (isKnownToBeAPowerOfTwo(Op1, Q.DL, /*OrZero*/ true, 0, Q.AC, Q.CxtI,
+ Q.DT))
return Op1;
}
Value *llvm::SimplifyAndInst(Value *Op0, Value *Op1, const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyAndInst(Op0, Op1, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyAndInst(Op0, Op1, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
match(A, m_Add(m_Value(V1), m_Value(V2)))) {
// Add commutes, try both ways.
if (V1 == B &&
- MaskedValueIsZero(V2, C2->getValue(), Q.DL, 0, Q.CxtI, Q.DT))
+ MaskedValueIsZero(V2, C2->getValue(), Q.DL, 0, Q.AC, Q.CxtI, Q.DT))
return A;
if (V2 == B &&
- MaskedValueIsZero(V1, C2->getValue(), Q.DL, 0, Q.CxtI, Q.DT))
+ MaskedValueIsZero(V1, C2->getValue(), Q.DL, 0, Q.AC, Q.CxtI, Q.DT))
return A;
}
// Or commutes, try both ways.
match(B, m_Add(m_Value(V1), m_Value(V2)))) {
// Add commutes, try both ways.
if (V1 == A &&
- MaskedValueIsZero(V2, C1->getValue(), Q.DL, 0, Q.CxtI, Q.DT))
+ MaskedValueIsZero(V2, C1->getValue(), Q.DL, 0, Q.AC, Q.CxtI, Q.DT))
return B;
if (V2 == A &&
- MaskedValueIsZero(V1, C1->getValue(), Q.DL, 0, Q.CxtI, Q.DT))
+ MaskedValueIsZero(V1, C1->getValue(), Q.DL, 0, Q.AC, Q.CxtI, Q.DT))
return B;
}
}
Value *llvm::SimplifyOrInst(Value *Op0, Value *Op1, const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyOrInst(Op0, Op1, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyOrInst(Op0, Op1, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyXorInst(Value *Op0, Value *Op1, const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyXorInst(Op0, Op1, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyXorInst(Op0, Op1, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
return getTrue(ITy);
case ICmpInst::ICMP_EQ:
case ICmpInst::ICMP_ULE:
- if (isKnownNonZero(LHS, Q.DL, 0, Q.CxtI, Q.DT))
+ if (isKnownNonZero(LHS, Q.DL, 0, Q.AC, Q.CxtI, Q.DT))
return getFalse(ITy);
break;
case ICmpInst::ICMP_NE:
case ICmpInst::ICMP_UGT:
- if (isKnownNonZero(LHS, Q.DL, 0, Q.CxtI, Q.DT))
+ if (isKnownNonZero(LHS, Q.DL, 0, Q.AC, Q.CxtI, Q.DT))
return getTrue(ITy);
break;
case ICmpInst::ICMP_SLT:
- ComputeSignBit(LHS, LHSKnownNonNegative, LHSKnownNegative, Q.DL, 0, Q.CxtI,
- Q.DT);
+ ComputeSignBit(LHS, LHSKnownNonNegative, LHSKnownNegative, Q.DL, 0, Q.AC,
+ Q.CxtI, Q.DT);
if (LHSKnownNegative)
return getTrue(ITy);
if (LHSKnownNonNegative)
return getFalse(ITy);
break;
case ICmpInst::ICMP_SLE:
- ComputeSignBit(LHS, LHSKnownNonNegative, LHSKnownNegative, Q.DL, 0, Q.CxtI,
- Q.DT);
+ ComputeSignBit(LHS, LHSKnownNonNegative, LHSKnownNegative, Q.DL, 0, Q.AC,
+ Q.CxtI, Q.DT);
if (LHSKnownNegative)
return getTrue(ITy);
- if (LHSKnownNonNegative && isKnownNonZero(LHS, Q.DL, 0, Q.CxtI, Q.DT))
+ if (LHSKnownNonNegative && isKnownNonZero(LHS, Q.DL, 0, Q.AC, Q.CxtI, Q.DT))
return getFalse(ITy);
break;
case ICmpInst::ICMP_SGE:
- ComputeSignBit(LHS, LHSKnownNonNegative, LHSKnownNegative, Q.DL, 0, Q.CxtI,
- Q.DT);
+ ComputeSignBit(LHS, LHSKnownNonNegative, LHSKnownNegative, Q.DL, 0, Q.AC,
+ Q.CxtI, Q.DT);
if (LHSKnownNegative)
return getFalse(ITy);
if (LHSKnownNonNegative)
return getTrue(ITy);
break;
case ICmpInst::ICMP_SGT:
- ComputeSignBit(LHS, LHSKnownNonNegative, LHSKnownNegative, Q.DL, 0, Q.CxtI,
- Q.DT);
+ ComputeSignBit(LHS, LHSKnownNonNegative, LHSKnownNegative, Q.DL, 0, Q.AC,
+ Q.CxtI, Q.DT);
if (LHSKnownNegative)
return getFalse(ITy);
- if (LHSKnownNonNegative && isKnownNonZero(LHS, Q.DL, 0, Q.CxtI, Q.DT))
+ if (LHSKnownNonNegative && isKnownNonZero(LHS, Q.DL, 0, Q.AC, Q.CxtI, Q.DT))
return getTrue(ITy);
break;
}
bool RHSKnownNonNegative, RHSKnownNegative;
bool YKnownNonNegative, YKnownNegative;
ComputeSignBit(RHS, RHSKnownNonNegative, RHSKnownNegative, Q.DL, 0,
+ Q.AC, Q.CxtI, Q.DT);
+ ComputeSignBit(Y, YKnownNonNegative, YKnownNegative, Q.DL, 0, Q.AC,
Q.CxtI, Q.DT);
- ComputeSignBit(Y, YKnownNonNegative, YKnownNegative, Q.DL, 0, Q.CxtI,
- Q.DT);
if (RHSKnownNonNegative && YKnownNegative)
return Pred == ICmpInst::ICMP_SLT ? getTrue(ITy) : getFalse(ITy);
if (RHSKnownNegative || YKnownNonNegative)
bool LHSKnownNonNegative, LHSKnownNegative;
bool YKnownNonNegative, YKnownNegative;
ComputeSignBit(LHS, LHSKnownNonNegative, LHSKnownNegative, Q.DL, 0,
+ Q.AC, Q.CxtI, Q.DT);
+ ComputeSignBit(Y, YKnownNonNegative, YKnownNegative, Q.DL, 0, Q.AC,
Q.CxtI, Q.DT);
- ComputeSignBit(Y, YKnownNonNegative, YKnownNegative, Q.DL, 0, Q.CxtI,
- Q.DT);
if (LHSKnownNonNegative && YKnownNegative)
return Pred == ICmpInst::ICMP_SGT ? getTrue(ITy) : getFalse(ITy);
if (LHSKnownNegative || YKnownNonNegative)
break;
case ICmpInst::ICMP_SGT:
case ICmpInst::ICMP_SGE:
- ComputeSignBit(RHS, KnownNonNegative, KnownNegative, Q.DL, 0, Q.CxtI,
- Q.DT);
+ ComputeSignBit(RHS, KnownNonNegative, KnownNegative, Q.DL, 0, Q.AC,
+ Q.CxtI, Q.DT);
if (!KnownNonNegative)
break;
LLVM_FALLTHROUGH;
return getFalse(ITy);
case ICmpInst::ICMP_SLT:
case ICmpInst::ICMP_SLE:
- ComputeSignBit(RHS, KnownNonNegative, KnownNegative, Q.DL, 0, Q.CxtI,
- Q.DT);
+ ComputeSignBit(RHS, KnownNonNegative, KnownNegative, Q.DL, 0, Q.AC,
+ Q.CxtI, Q.DT);
if (!KnownNonNegative)
break;
LLVM_FALLTHROUGH;
break;
case ICmpInst::ICMP_SGT:
case ICmpInst::ICMP_SGE:
- ComputeSignBit(LHS, KnownNonNegative, KnownNegative, Q.DL, 0, Q.CxtI,
- Q.DT);
+ ComputeSignBit(LHS, KnownNonNegative, KnownNegative, Q.DL, 0, Q.AC,
+ Q.CxtI, Q.DT);
if (!KnownNonNegative)
break;
LLVM_FALLTHROUGH;
return getTrue(ITy);
case ICmpInst::ICMP_SLT:
case ICmpInst::ICMP_SLE:
- ComputeSignBit(LHS, KnownNonNegative, KnownNegative, Q.DL, 0, Q.CxtI,
- Q.DT);
+ ComputeSignBit(LHS, KnownNonNegative, KnownNegative, Q.DL, 0, Q.AC,
+ Q.CxtI, Q.DT);
if (!KnownNonNegative)
break;
LLVM_FALLTHROUGH;
// icmp eq|ne X, Y -> false|true if X != Y
if ((Pred == ICmpInst::ICMP_EQ || Pred == ICmpInst::ICMP_NE) &&
- isKnownNonEqual(LHS, RHS, Q.DL, Q.CxtI, Q.DT)) {
+ isKnownNonEqual(LHS, RHS, Q.DL, Q.AC, Q.CxtI, Q.DT)) {
LLVMContext &Ctx = LHS->getType()->getContext();
return Pred == ICmpInst::ICMP_NE ?
ConstantInt::getTrue(Ctx) : ConstantInt::getFalse(Ctx);
unsigned BitWidth = RHSVal->getBitWidth();
APInt LHSKnownZero(BitWidth, 0);
APInt LHSKnownOne(BitWidth, 0);
- computeKnownBits(LHS, LHSKnownZero, LHSKnownOne, Q.DL, /*Depth=*/0,
+ computeKnownBits(LHS, LHSKnownZero, LHSKnownOne, Q.DL, /*Depth=*/0, Q.AC,
Q.CxtI, Q.DT);
if (((LHSKnownZero & *RHSVal) != 0) || ((LHSKnownOne & ~(*RHSVal)) != 0))
return Pred == ICmpInst::ICMP_EQ ? ConstantInt::getFalse(ITy)
Value *llvm::SimplifyICmpInst(unsigned Predicate, Value *LHS, Value *RHS,
const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyICmpInst(Predicate, LHS, RHS, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyICmpInst(Predicate, LHS, RHS, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyFCmpInst(unsigned Predicate, Value *LHS, Value *RHS,
FastMathFlags FMF, const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
return ::SimplifyFCmpInst(Predicate, LHS, RHS, FMF,
- Query(DL, TLI, DT, CxtI), RecursionLimit);
+ Query(DL, TLI, DT, AC, CxtI), RecursionLimit);
}
/// See if V simplifies when its operand Op is replaced with RepOp.
Value *llvm::SimplifySelectInst(Value *Cond, Value *TrueVal, Value *FalseVal,
const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
return ::SimplifySelectInst(Cond, TrueVal, FalseVal,
- Query(DL, TLI, DT, CxtI), RecursionLimit);
+ Query(DL, TLI, DT, AC, CxtI), RecursionLimit);
}
/// Given operands for an GetElementPtrInst, see if we can fold the result.
Value *llvm::SimplifyGEPInst(Type *SrcTy, ArrayRef<Value *> Ops,
const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
return ::SimplifyGEPInst(SrcTy, Ops,
- Query(DL, TLI, DT, CxtI), RecursionLimit);
+ Query(DL, TLI, DT, AC, CxtI), RecursionLimit);
}
/// Given operands for an InsertValueInst, see if we can fold the result.
Value *llvm::SimplifyInsertValueInst(
Value *Agg, Value *Val, ArrayRef<unsigned> Idxs, const DataLayout &DL,
- const TargetLibraryInfo *TLI, const DominatorTree *DT,
+ const TargetLibraryInfo *TLI, const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyInsertValueInst(Agg, Val, Idxs, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyInsertValueInst(Agg, Val, Idxs, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
const DataLayout &DL,
const TargetLibraryInfo *TLI,
const DominatorTree *DT,
+ AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyExtractValueInst(Agg, Idxs, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyExtractValueInst(Agg, Idxs, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyExtractElementInst(
Value *Vec, Value *Idx, const DataLayout &DL, const TargetLibraryInfo *TLI,
- const DominatorTree *DT, const Instruction *CxtI) {
- return ::SimplifyExtractElementInst(Vec, Idx, Query(DL, TLI, DT, CxtI),
+ const DominatorTree *DT, AssumptionCache *AC, const Instruction *CxtI) {
+ return ::SimplifyExtractElementInst(Vec, Idx, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyCastInst(unsigned CastOpc, Value *Op, Type *Ty,
const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyCastInst(CastOpc, Op, Ty, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyCastInst(CastOpc, Op, Ty, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyBinOp(unsigned Opcode, Value *LHS, Value *RHS,
const DataLayout &DL, const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyBinOp(Opcode, LHS, RHS, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyBinOp(Opcode, LHS, RHS, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyFPBinOp(unsigned Opcode, Value *LHS, Value *RHS,
const FastMathFlags &FMF, const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyFPBinOp(Opcode, LHS, RHS, FMF, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyFPBinOp(Opcode, LHS, RHS, FMF, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyCmpInst(unsigned Predicate, Value *LHS, Value *RHS,
const DataLayout &DL, const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
- return ::SimplifyCmpInst(Predicate, LHS, RHS, Query(DL, TLI, DT, CxtI),
+ return ::SimplifyCmpInst(Predicate, LHS, RHS, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyCall(Value *V, User::op_iterator ArgBegin,
User::op_iterator ArgEnd, const DataLayout &DL,
const TargetLibraryInfo *TLI, const DominatorTree *DT,
- const Instruction *CxtI) {
- return ::SimplifyCall(V, ArgBegin, ArgEnd, Query(DL, TLI, DT, CxtI),
+ AssumptionCache *AC, const Instruction *CxtI) {
+ return ::SimplifyCall(V, ArgBegin, ArgEnd, Query(DL, TLI, DT, AC, CxtI),
RecursionLimit);
}
Value *llvm::SimplifyCall(Value *V, ArrayRef<Value *> Args,
const DataLayout &DL, const TargetLibraryInfo *TLI,
- const DominatorTree *DT,
+ const DominatorTree *DT, AssumptionCache *AC,
const Instruction *CxtI) {
return ::SimplifyCall(V, Args.begin(), Args.end(),
- Query(DL, TLI, DT, CxtI), RecursionLimit);
+ Query(DL, TLI, DT, AC, CxtI), RecursionLimit);
}
/// See if we can compute a simplified version of this instruction.
/// If not, this returns null.
Value *llvm::SimplifyInstruction(Instruction *I, const DataLayout &DL,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT) {
+ const DominatorTree *DT, AssumptionCache *AC) {
Value *Result;
switch (I->getOpcode()) {
break;
case Instruction::FAdd:
Result = SimplifyFAddInst(I->getOperand(0), I->getOperand(1),
- I->getFastMathFlags(), DL, TLI, DT, I);
+ I->getFastMathFlags(), DL, TLI, DT, AC, I);
break;
case Instruction::Add:
Result = SimplifyAddInst(I->getOperand(0), I->getOperand(1),
cast<BinaryOperator>(I)->hasNoSignedWrap(),
cast<BinaryOperator>(I)->hasNoUnsignedWrap(), DL,
- TLI, DT, I);
+ TLI, DT, AC, I);
break;
case Instruction::FSub:
Result = SimplifyFSubInst(I->getOperand(0), I->getOperand(1),
- I->getFastMathFlags(), DL, TLI, DT, I);
+ I->getFastMathFlags(), DL, TLI, DT, AC, I);
break;
case Instruction::Sub:
Result = SimplifySubInst(I->getOperand(0), I->getOperand(1),
cast<BinaryOperator>(I)->hasNoSignedWrap(),
cast<BinaryOperator>(I)->hasNoUnsignedWrap(), DL,
- TLI, DT, I);
+ TLI, DT, AC, I);
break;
case Instruction::FMul:
Result = SimplifyFMulInst(I->getOperand(0), I->getOperand(1),
- I->getFastMathFlags(), DL, TLI, DT, I);
+ I->getFastMathFlags(), DL, TLI, DT, AC, I);
break;
case Instruction::Mul:
Result =
- SimplifyMulInst(I->getOperand(0), I->getOperand(1), DL, TLI, DT, I);
+ SimplifyMulInst(I->getOperand(0), I->getOperand(1), DL, TLI, DT, AC, I);
break;
case Instruction::SDiv:
Result = SimplifySDivInst(I->getOperand(0), I->getOperand(1), DL, TLI, DT,
- I);
+ AC, I);
break;
case Instruction::UDiv:
Result = SimplifyUDivInst(I->getOperand(0), I->getOperand(1), DL, TLI, DT,
- I);
+ AC, I);
break;
case Instruction::FDiv:
Result = SimplifyFDivInst(I->getOperand(0), I->getOperand(1),
- I->getFastMathFlags(), DL, TLI, DT, I);
+ I->getFastMathFlags(), DL, TLI, DT, AC, I);
break;
case Instruction::SRem:
Result = SimplifySRemInst(I->getOperand(0), I->getOperand(1), DL, TLI, DT,
- I);
+ AC, I);
break;
case Instruction::URem:
Result = SimplifyURemInst(I->getOperand(0), I->getOperand(1), DL, TLI, DT,
- I);
+ AC, I);
break;
case Instruction::FRem:
Result = SimplifyFRemInst(I->getOperand(0), I->getOperand(1),
- I->getFastMathFlags(), DL, TLI, DT, I);
+ I->getFastMathFlags(), DL, TLI, DT, AC, I);
break;
case Instruction::Shl:
Result = SimplifyShlInst(I->getOperand(0), I->getOperand(1),
cast<BinaryOperator>(I)->hasNoSignedWrap(),
cast<BinaryOperator>(I)->hasNoUnsignedWrap(), DL,
- TLI, DT, I);
+ TLI, DT, AC, I);
break;
case Instruction::LShr:
Result = SimplifyLShrInst(I->getOperand(0), I->getOperand(1),
cast<BinaryOperator>(I)->isExact(), DL, TLI, DT,
- I);
+ AC, I);
break;
case Instruction::AShr:
Result = SimplifyAShrInst(I->getOperand(0), I->getOperand(1),
cast<BinaryOperator>(I)->isExact(), DL, TLI, DT,
- I);
+ AC, I);
break;
case Instruction::And:
Result =
- SimplifyAndInst(I->getOperand(0), I->getOperand(1), DL, TLI, DT, I);
+ SimplifyAndInst(I->getOperand(0), I->getOperand(1), DL, TLI, DT, AC, I);
break;
case Instruction::Or:
Result =
- SimplifyOrInst(I->getOperand(0), I->getOperand(1), DL, TLI, DT, I);
+ SimplifyOrInst(I->getOperand(0), I->getOperand(1), DL, TLI, DT, AC, I);
break;
case Instruction::Xor:
Result =
- SimplifyXorInst(I->getOperand(0), I->getOperand(1), DL, TLI, DT, I);
+ SimplifyXorInst(I->getOperand(0), I->getOperand(1), DL, TLI, DT, AC, I);
break;
case Instruction::ICmp:
Result =
SimplifyICmpInst(cast<ICmpInst>(I)->getPredicate(), I->getOperand(0),
- I->getOperand(1), DL, TLI, DT, I);
+ I->getOperand(1), DL, TLI, DT, AC, I);
break;
case Instruction::FCmp:
Result = SimplifyFCmpInst(cast<FCmpInst>(I)->getPredicate(),
I->getOperand(0), I->getOperand(1),
- I->getFastMathFlags(), DL, TLI, DT, I);
+ I->getFastMathFlags(), DL, TLI, DT, AC, I);
break;
case Instruction::Select:
Result = SimplifySelectInst(I->getOperand(0), I->getOperand(1),
- I->getOperand(2), DL, TLI, DT, I);
+ I->getOperand(2), DL, TLI, DT, AC, I);
break;
case Instruction::GetElementPtr: {
SmallVector<Value*, 8> Ops(I->op_begin(), I->op_end());
Result = SimplifyGEPInst(cast<GetElementPtrInst>(I)->getSourceElementType(),
- Ops, DL, TLI, DT, I);
+ Ops, DL, TLI, DT, AC, I);
break;
}
case Instruction::InsertValue: {
InsertValueInst *IV = cast<InsertValueInst>(I);
Result = SimplifyInsertValueInst(IV->getAggregateOperand(),
IV->getInsertedValueOperand(),
- IV->getIndices(), DL, TLI, DT, I);
+ IV->getIndices(), DL, TLI, DT, AC, I);
break;
}
case Instruction::ExtractValue: {
auto *EVI = cast<ExtractValueInst>(I);
Result = SimplifyExtractValueInst(EVI->getAggregateOperand(),
- EVI->getIndices(), DL, TLI, DT, I);
+ EVI->getIndices(), DL, TLI, DT, AC, I);
break;
}
case Instruction::ExtractElement: {
auto *EEI = cast<ExtractElementInst>(I);
Result = SimplifyExtractElementInst(
- EEI->getVectorOperand(), EEI->getIndexOperand(), DL, TLI, DT, I);
+ EEI->getVectorOperand(), EEI->getIndexOperand(), DL, TLI, DT, AC, I);
break;
}
case Instruction::PHI:
- Result = SimplifyPHINode(cast<PHINode>(I), Query(DL, TLI, DT, I));
+ Result = SimplifyPHINode(cast<PHINode>(I), Query(DL, TLI, DT, AC, I));
break;
case Instruction::Call: {
CallSite CS(cast<CallInst>(I));
Result = SimplifyCall(CS.getCalledValue(), CS.arg_begin(), CS.arg_end(), DL,
- TLI, DT, I);
+ TLI, DT, AC, I);
break;
}
#define HANDLE_CAST_INST(num, opc, clas) case Instruction::opc:
#include "llvm/IR/Instruction.def"
#undef HANDLE_CAST_INST
Result = SimplifyCastInst(I->getOpcode(), I->getOperand(0), I->getType(),
- DL, TLI, DT, I);
+ DL, TLI, DT, AC, I);
break;
}
unsigned BitWidth = I->getType()->getScalarSizeInBits();
APInt KnownZero(BitWidth, 0);
APInt KnownOne(BitWidth, 0);
- computeKnownBits(I, KnownZero, KnownOne, DL, /*Depth*/0, I, DT);
+ computeKnownBits(I, KnownZero, KnownOne, DL, /*Depth*/0, AC, I, DT);
if ((KnownZero | KnownOne).isAllOnesValue())
Result = ConstantInt::get(I->getType(), KnownOne);
}
/// in simplified value does not count toward this.
static bool replaceAndRecursivelySimplifyImpl(Instruction *I, Value *SimpleV,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT) {
+ const DominatorTree *DT,
+ AssumptionCache *AC) {
bool Simplified = false;
SmallSetVector<Instruction *, 8> Worklist;
const DataLayout &DL = I->getModule()->getDataLayout();
I = Worklist[Idx];
// See if this instruction simplifies.
- SimpleV = SimplifyInstruction(I, DL, TLI, DT);
+ SimpleV = SimplifyInstruction(I, DL, TLI, DT, AC);
if (!SimpleV)
continue;
bool llvm::recursivelySimplifyInstruction(Instruction *I,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT) {
- return replaceAndRecursivelySimplifyImpl(I, nullptr, TLI, DT);
+ const DominatorTree *DT,
+ AssumptionCache *AC) {
+ return replaceAndRecursivelySimplifyImpl(I, nullptr, TLI, DT, AC);
}
bool llvm::replaceAndRecursivelySimplify(Instruction *I, Value *SimpleV,
const TargetLibraryInfo *TLI,
- const DominatorTree *DT) {
+ const DominatorTree *DT,
+ AssumptionCache *AC) {
assert(I != SimpleV && "replaceAndRecursivelySimplify(X,X) is not valid!");
assert(SimpleV && "Must provide a simplified value.");
- return replaceAndRecursivelySimplifyImpl(I, SimpleV, TLI, DT);
+ return replaceAndRecursivelySimplifyImpl(I, SimpleV, TLI, DT, AC);
}
#include "llvm/Analysis/LazyValueInfo.h"
#include "llvm/ADT/DenseSet.h"
#include "llvm/ADT/STLExtras.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/ConstantFolding.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/Analysis/ValueTracking.h"
char LazyValueInfoWrapperPass::ID = 0;
INITIALIZE_PASS_BEGIN(LazyValueInfoWrapperPass, "lazy-value-info",
"Lazy Value Information Analysis", false, true)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
INITIALIZE_PASS_END(LazyValueInfoWrapperPass, "lazy-value-info",
"Lazy Value Information Analysis", false, true)
return true;
}
+ AssumptionCache *AC; ///< A pointer to the cache of @llvm.assume calls.
const DataLayout &DL; ///< A mandatory DataLayout
DominatorTree *DT; ///< An optional DT pointer.
/// PredBB to OldSucc has been threaded to be from PredBB to NewSucc.
void threadEdge(BasicBlock *PredBB,BasicBlock *OldSucc,BasicBlock *NewSucc);
- LazyValueInfoImpl(const DataLayout &DL, DominatorTree *DT = nullptr)
- : DL(DL), DT(DT) {}
+ LazyValueInfoImpl(AssumptionCache *AC, const DataLayout &DL,
+ DominatorTree *DT = nullptr)
+ : AC(AC), DL(DL), DT(DT) {}
};
} // end anonymous namespace
if (!BBI)
return;
- for (auto *U : Val->users()) {
- auto *II = dyn_cast<IntrinsicInst>(U);
- if (!II)
+ for (auto &AssumeVH : AC->assumptions()) {
+ if (!AssumeVH)
continue;
- if (II->getIntrinsicID() != Intrinsic::assume)
- continue;
- if (!isValidAssumeForContext(II, BBI, DT))
+ auto *I = cast<CallInst>(AssumeVH);
+ if (!isValidAssumeForContext(I, BBI, DT))
continue;
- BBLV = intersect(BBLV, getValueFromCondition(Val, II->getArgOperand(0)));
+ BBLV = intersect(BBLV, getValueFromCondition(Val, I->getArgOperand(0)));
}
// If guards are not used in the module, don't spend time looking for them
//===----------------------------------------------------------------------===//
/// This lazily constructs the LazyValueInfoImpl.
-static LazyValueInfoImpl &getImpl(void *&PImpl, const DataLayout *DL,
+static LazyValueInfoImpl &getImpl(void *&PImpl, AssumptionCache *AC,
+ const DataLayout *DL,
DominatorTree *DT = nullptr) {
if (!PImpl) {
assert(DL && "getCache() called with a null DataLayout");
- PImpl = new LazyValueInfoImpl(*DL, DT);
+ PImpl = new LazyValueInfoImpl(AC, *DL, DT);
}
return *static_cast<LazyValueInfoImpl*>(PImpl);
}
bool LazyValueInfoWrapperPass::runOnFunction(Function &F) {
+ Info.AC = &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
const DataLayout &DL = F.getParent()->getDataLayout();
DominatorTreeWrapperPass *DTWP =
Info.TLI = &getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
if (Info.PImpl)
- getImpl(Info.PImpl, &DL, Info.DT).clear();
+ getImpl(Info.PImpl, Info.AC, &DL, Info.DT).clear();
// Fully lazy.
return false;
void LazyValueInfoWrapperPass::getAnalysisUsage(AnalysisUsage &AU) const {
AU.setPreservesAll();
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<TargetLibraryInfoWrapperPass>();
}
void LazyValueInfo::releaseMemory() {
// If the cache was allocated, free it.
if (PImpl) {
- delete &getImpl(PImpl, nullptr);
+ delete &getImpl(PImpl, AC, nullptr);
PImpl = nullptr;
}
}
void LazyValueInfoWrapperPass::releaseMemory() { Info.releaseMemory(); }
LazyValueInfo LazyValueAnalysis::run(Function &F, FunctionAnalysisManager &FAM) {
+ auto &AC = FAM.getResult<AssumptionAnalysis>(F);
auto &TLI = FAM.getResult<TargetLibraryAnalysis>(F);
auto *DT = FAM.getCachedResult<DominatorTreeAnalysis>(F);
- return LazyValueInfo(&TLI, DT);
+ return LazyValueInfo(&AC, &TLI, DT);
}
/// Returns true if we can statically tell that this value will never be a
const DataLayout &DL = BB->getModule()->getDataLayout();
LVILatticeVal Result =
- getImpl(PImpl, &DL, DT).getValueInBlock(V, BB, CxtI);
+ getImpl(PImpl, AC, &DL, DT).getValueInBlock(V, BB, CxtI);
if (Result.isConstant())
return Result.getConstant();
unsigned Width = V->getType()->getIntegerBitWidth();
const DataLayout &DL = BB->getModule()->getDataLayout();
LVILatticeVal Result =
- getImpl(PImpl, &DL, DT).getValueInBlock(V, BB, CxtI);
+ getImpl(PImpl, AC, &DL, DT).getValueInBlock(V, BB, CxtI);
if (Result.isUndefined())
return ConstantRange(Width, /*isFullSet=*/false);
if (Result.isConstantRange())
Instruction *CxtI) {
const DataLayout &DL = FromBB->getModule()->getDataLayout();
LVILatticeVal Result =
- getImpl(PImpl, &DL, DT).getValueOnEdge(V, FromBB, ToBB, CxtI);
+ getImpl(PImpl, AC, &DL, DT).getValueOnEdge(V, FromBB, ToBB, CxtI);
if (Result.isConstant())
return Result.getConstant();
Instruction *CxtI) {
const DataLayout &DL = FromBB->getModule()->getDataLayout();
LVILatticeVal Result =
- getImpl(PImpl, &DL, DT).getValueOnEdge(V, FromBB, ToBB, CxtI);
+ getImpl(PImpl, AC, &DL, DT).getValueOnEdge(V, FromBB, ToBB, CxtI);
return getPredicateResult(Pred, C, Result, DL, TLI);
}
return LazyValueInfo::True;
}
const DataLayout &DL = CxtI->getModule()->getDataLayout();
- LVILatticeVal Result = getImpl(PImpl, &DL, DT).getValueAt(V, CxtI);
+ LVILatticeVal Result = getImpl(PImpl, AC, &DL, DT).getValueAt(V, CxtI);
Tristate Ret = getPredicateResult(Pred, C, Result, DL, TLI);
if (Ret != Unknown)
return Ret;
BasicBlock *NewSucc) {
if (PImpl) {
const DataLayout &DL = PredBB->getModule()->getDataLayout();
- getImpl(PImpl, &DL, DT).threadEdge(PredBB, OldSucc, NewSucc);
+ getImpl(PImpl, AC, &DL, DT).threadEdge(PredBB, OldSucc, NewSucc);
}
}
void LazyValueInfo::eraseBlock(BasicBlock *BB) {
if (PImpl) {
const DataLayout &DL = BB->getModule()->getDataLayout();
- getImpl(PImpl, &DL, DT).eraseBlock(BB);
+ getImpl(PImpl, AC, &DL, DT).eraseBlock(BB);
}
}
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/Twine.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/ConstantFolding.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/Loads.h"
Module *Mod;
const DataLayout *DL;
AliasAnalysis *AA;
+ AssumptionCache *AC;
DominatorTree *DT;
TargetLibraryInfo *TLI;
void getAnalysisUsage(AnalysisUsage &AU) const override {
AU.setPreservesAll();
AU.addRequired<AAResultsWrapperPass>();
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<TargetLibraryInfoWrapperPass>();
AU.addRequired<DominatorTreeWrapperPass>();
}
char Lint::ID = 0;
INITIALIZE_PASS_BEGIN(Lint, "lint", "Statically lint-checks LLVM IR",
false, true)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(AAResultsWrapperPass)
Mod = F.getParent();
DL = &F.getParent()->getDataLayout();
AA = &getAnalysis<AAResultsWrapperPass>().getAAResults();
+ AC = &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
TLI = &getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
visit(F);
"Undefined result: Shift count out of range", &I);
}
-static bool isZero(Value *V, const DataLayout &DL, DominatorTree *DT) {
+static bool isZero(Value *V, const DataLayout &DL, DominatorTree *DT,
+ AssumptionCache *AC) {
// Assume undef could be zero.
if (isa<UndefValue>(V))
return true;
if (!VecTy) {
unsigned BitWidth = V->getType()->getIntegerBitWidth();
APInt KnownZero(BitWidth, 0), KnownOne(BitWidth, 0);
- computeKnownBits(V, KnownZero, KnownOne, DL, 0,
+ computeKnownBits(V, KnownZero, KnownOne, DL, 0, AC,
dyn_cast<Instruction>(V), DT);
return KnownZero.isAllOnesValue();
}
}
void Lint::visitSDiv(BinaryOperator &I) {
- Assert(!isZero(I.getOperand(1), I.getModule()->getDataLayout(), DT),
+ Assert(!isZero(I.getOperand(1), I.getModule()->getDataLayout(), DT, AC),
"Undefined behavior: Division by zero", &I);
}
void Lint::visitUDiv(BinaryOperator &I) {
- Assert(!isZero(I.getOperand(1), I.getModule()->getDataLayout(), DT),
+ Assert(!isZero(I.getOperand(1), I.getModule()->getDataLayout(), DT, AC),
"Undefined behavior: Division by zero", &I);
}
void Lint::visitSRem(BinaryOperator &I) {
- Assert(!isZero(I.getOperand(1), I.getModule()->getDataLayout(), DT),
+ Assert(!isZero(I.getOperand(1), I.getModule()->getDataLayout(), DT, AC),
"Undefined behavior: Division by zero", &I);
}
void Lint::visitURem(BinaryOperator &I) {
- Assert(!isZero(I.getOperand(1), I.getModule()->getDataLayout(), DT),
+ Assert(!isZero(I.getOperand(1), I.getModule()->getDataLayout(), DT, AC),
"Undefined behavior: Division by zero", &I);
}
// As a last resort, try SimplifyInstruction or constant folding.
if (Instruction *Inst = dyn_cast<Instruction>(V)) {
- if (Value *W = SimplifyInstruction(Inst, *DL, TLI, DT))
+ if (Value *W = SimplifyInstruction(Inst, *DL, TLI, DT, AC))
return findValueImpl(W, OffsetOk, Visited);
} else if (auto *C = dyn_cast<Constant>(V)) {
if (Value *W = ConstantFoldConstant(C, *DL, TLI))
#include "llvm/ADT/STLExtras.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/MemoryBuiltins.h"
#include "llvm/Analysis/PHITransAddr.h"
#include "llvm/Analysis/OrderedBasicBlock.h"
return;
}
const DataLayout &DL = FromBB->getModule()->getDataLayout();
- PHITransAddr Address(const_cast<Value *>(Loc.Ptr), DL);
+ PHITransAddr Address(const_cast<Value *>(Loc.Ptr), DL, &AC);
// 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
MemoryDependenceResults
MemoryDependenceAnalysis::run(Function &F, FunctionAnalysisManager &AM) {
auto &AA = AM.getResult<AAManager>(F);
+ auto &AC = AM.getResult<AssumptionAnalysis>(F);
auto &TLI = AM.getResult<TargetLibraryAnalysis>(F);
auto &DT = AM.getResult<DominatorTreeAnalysis>(F);
- return MemoryDependenceResults(AA, TLI, DT);
+ return MemoryDependenceResults(AA, AC, TLI, DT);
}
char MemoryDependenceWrapperPass::ID = 0;
INITIALIZE_PASS_BEGIN(MemoryDependenceWrapperPass, "memdep",
"Memory Dependence Analysis", false, true)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(AAResultsWrapperPass)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
void MemoryDependenceWrapperPass::getAnalysisUsage(AnalysisUsage &AU) const {
AU.setPreservesAll();
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<DominatorTreeWrapperPass>();
AU.addRequiredTransitive<AAResultsWrapperPass>();
AU.addRequiredTransitive<TargetLibraryInfoWrapperPass>();
bool MemoryDependenceWrapperPass::runOnFunction(Function &F) {
auto &AA = getAnalysis<AAResultsWrapperPass>().getAAResults();
+ auto &AC = getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
auto &TLI = getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
auto &DT = getAnalysis<DominatorTreeWrapperPass>().getDomTree();
- MemDep.emplace(AA, TLI, DT);
+ MemDep.emplace(AA, AC, TLI, DT);
return false;
}
// Simplify the GEP to handle 'gep x, 0' -> x etc.
if (Value *V = SimplifyGEPInst(GEP->getSourceElementType(),
- GEPOps, DL, TLI, DT)) {
+ GEPOps, DL, TLI, DT, AC)) {
for (unsigned i = 0, e = GEPOps.size(); i != e; ++i)
RemoveInstInputs(GEPOps[i], InstInputs);
}
// See if the add simplifies away.
- if (Value *Res = SimplifyAddInst(LHS, RHS, isNSW, isNUW, DL, TLI, DT)) {
+ if (Value *Res = SimplifyAddInst(LHS, RHS, isNSW, isNUW, DL, TLI, DT, AC)) {
// If we simplified the operands, the LHS is no longer an input, but Res
// is.
RemoveInstInputs(LHS, InstInputs);
SmallVectorImpl<Instruction*> &NewInsts) {
// See if we have a version of this value already available and dominating
// PredBB. If so, there is no need to insert a new instance of it.
- PHITransAddr Tmp(InVal, DL);
+ PHITransAddr Tmp(InVal, DL, AC);
if (!Tmp.PHITranslateValue(CurBB, PredBB, &DT, /*MustDominate=*/true))
return Tmp.getAddr();
#include "llvm/ADT/ScopeExit.h"
#include "llvm/ADT/Sequence.h"
#include "llvm/ADT/SmallPtrSet.h"
-#include "llvm/ADT/SmallSet.h"
#include "llvm/ADT/Statistic.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/ConstantFolding.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/LoopInfo.h"
SCEV *S = new (SCEVAllocator) SCEVTruncateExpr(ID.Intern(SCEVAllocator),
Op, Ty);
UniqueSCEVs.InsertNode(S, IP);
- addAffectedFromOperands(S);
return S;
}
// these to prove lack of overflow. Use this fact to avoid
// doing extra work that may not pay off.
if (!isa<SCEVCouldNotCompute>(MaxBECount) || HasGuards ||
- !AffectedMap.empty()) {
+ !AC.assumptions().empty()) {
// If the backedge is guarded by a comparison with the pre-inc
// value the addrec is safe. Also, if the entry is guarded by
// a comparison with the start value and the backedge is
SCEV *S = new (SCEVAllocator) SCEVZeroExtendExpr(ID.Intern(SCEVAllocator),
Op, Ty);
UniqueSCEVs.InsertNode(S, IP);
- addAffectedFromOperands(S);
return S;
}
// doing extra work that may not pay off.
if (!isa<SCEVCouldNotCompute>(MaxBECount) || HasGuards ||
- !AffectedMap.empty()) {
+ !AC.assumptions().empty()) {
// If the backedge is guarded by a comparison with the pre-inc
// value the addrec is safe. Also, if the entry is guarded by
// a comparison with the start value and the backedge is
SCEV *S = new (SCEVAllocator) SCEVSignExtendExpr(ID.Intern(SCEVAllocator),
Op, Ty);
UniqueSCEVs.InsertNode(S, IP);
- addAffectedFromOperands(S);
return S;
}
S = new (SCEVAllocator) SCEVAddExpr(ID.Intern(SCEVAllocator),
O, Ops.size());
UniqueSCEVs.InsertNode(S, IP);
- addAffectedFromOperands(S);
}
S->setNoWrapFlags(Flags);
return S;
S = new (SCEVAllocator) SCEVMulExpr(ID.Intern(SCEVAllocator),
O, Ops.size());
UniqueSCEVs.InsertNode(S, IP);
- addAffectedFromOperands(S);
}
S->setNoWrapFlags(Flags);
return S;
SCEV *S = new (SCEVAllocator) SCEVUDivExpr(ID.Intern(SCEVAllocator),
LHS, RHS);
UniqueSCEVs.InsertNode(S, IP);
- addAffectedFromOperands(S);
return S;
}
S = new (SCEVAllocator) SCEVAddRecExpr(ID.Intern(SCEVAllocator),
O, Operands.size(), L);
UniqueSCEVs.InsertNode(S, IP);
- addAffectedFromOperands(S);
}
S->setNoWrapFlags(Flags);
return S;
SCEV *S = new (SCEVAllocator) SCEVSMaxExpr(ID.Intern(SCEVAllocator),
O, Ops.size());
UniqueSCEVs.InsertNode(S, IP);
- addAffectedFromOperands(S);
return S;
}
SCEV *S = new (SCEVAllocator) SCEVUMaxExpr(ID.Intern(SCEVAllocator),
O, Ops.size());
UniqueSCEVs.InsertNode(S, IP);
- addAffectedFromOperands(S);
return S;
}
ExprValueMap[Stripped].insert({V, Offset});
}
}
-
- // If this value is an instruction or an argument, and might be affected by
- // an assumption, and its SCEV to the AffectedMap.
- if (isa<Instruction>(V) || isa<Argument>(V)) {
- for (auto *U : V->users()) {
- auto *II = dyn_cast<IntrinsicInst>(U);
- if (!II)
- continue;
- if (II->getIntrinsicID() != Intrinsic::assume)
- continue;
-
- AffectedMap[S].insert(II);
- }
- }
-
return S;
}
-// If one of this SCEV's operands is in the AffectedMap (meaning that it might
-// be affected by an assumption), then this SCEV might be affected by the same
-// assumption.
-void ScalarEvolution::addAffectedFromOperands(const SCEV *S) {
- if (auto *NS = dyn_cast<SCEVNAryExpr>(S))
- for (auto *Op : NS->operands()) {
- auto AMI = AffectedMap.find(Op);
- if (AMI == AffectedMap.end())
- continue;
-
- auto &ISet = AffectedMap[S];
- AMI = AffectedMap.find(Op);
- ISet.insert(AMI->second.begin(), AMI->second.end());
- }
-}
-
const SCEV *ScalarEvolution::getExistingSCEV(Value *V) {
assert(isSCEVable(V->getType()) && "Value is not SCEVable!");
// PHI's incoming blocks are in a different loop, in which case doing so
// risks breaking LCSSA form. Instcombine would normally zap these, but
// it doesn't have DominatorTree information, so it may miss cases.
- if (Value *V = SimplifyInstruction(PN, getDataLayout(), &TLI, &DT))
+ if (Value *V = SimplifyInstruction(PN, getDataLayout(), &TLI, &DT, &AC))
if (LI.replacementPreservesLCSSAForm(PN, V))
return getSCEV(V);
// For a SCEVUnknown, ask ValueTracking.
unsigned BitWidth = getTypeSizeInBits(U->getType());
APInt Zeros(BitWidth, 0), Ones(BitWidth, 0);
- computeKnownBits(U->getValue(), Zeros, Ones, getDataLayout(), 0,
+ computeKnownBits(U->getValue(), Zeros, Ones, getDataLayout(), 0, &AC,
nullptr, &DT);
return Zeros.countTrailingOnes();
}
if (SignHint == ScalarEvolution::HINT_RANGE_UNSIGNED) {
// For a SCEVUnknown, ask ValueTracking.
APInt Zeros(BitWidth, 0), Ones(BitWidth, 0);
- computeKnownBits(U->getValue(), Zeros, Ones, DL, 0, nullptr, &DT);
+ computeKnownBits(U->getValue(), Zeros, Ones, DL, 0, &AC, nullptr, &DT);
if (Ones != ~Zeros + 1)
ConservativeResult =
ConservativeResult.intersectWith(ConstantRange(Ones, ~Zeros + 1));
} else {
assert(SignHint == ScalarEvolution::HINT_RANGE_SIGNED &&
"generalize as needed!");
- unsigned NS = ComputeNumSignBits(U->getValue(), DL, 0, nullptr, &DT);
+ unsigned NS = ComputeNumSignBits(U->getValue(), DL, 0, &AC, nullptr, &DT);
if (NS > 1)
ConservativeResult = ConservativeResult.intersectWith(
ConstantRange(APInt::getSignedMinValue(BitWidth).ashr(NS - 1),
unsigned BitWidth = A.getBitWidth();
APInt KnownZero(BitWidth, 0), KnownOne(BitWidth, 0);
computeKnownBits(BO->LHS, KnownZero, KnownOne, getDataLayout(),
- 0, nullptr, &DT);
+ 0, &AC, nullptr, &DT);
APInt EffectiveMask =
APInt::getLowBitsSet(BitWidth, BitWidth - LZ - TZ).shl(TZ);
// bitwidth(K) iterations.
Value *FirstValue = PN->getIncomingValueForBlock(Predecessor);
bool KnownZero, KnownOne;
- ComputeSignBit(FirstValue, KnownZero, KnownOne, DL, 0,
+ ComputeSignBit(FirstValue, KnownZero, KnownOne, DL, 0, nullptr,
Predecessor->getTerminator(), &DT);
auto *Ty = cast<IntegerType>(RHS->getType());
if (KnownZero)
}
// Check conditions due to any @llvm.assume intrinsics.
- auto CheckAssumptions = [&](const SCEV *S) {
- auto AMI = AffectedMap.find(S);
- if (AMI != AffectedMap.end())
- for (auto *Assume : AMI->second) {
- auto *CI = cast<CallInst>(Assume);
- if (!DT.dominates(CI, Latch->getTerminator()))
- continue;
-
- if (isImpliedCond(Pred, LHS, RHS, CI->getArgOperand(0), false))
- return true;
- }
-
- return false;
- };
+ for (auto &AssumeVH : AC.assumptions()) {
+ if (!AssumeVH)
+ continue;
+ auto *CI = cast<CallInst>(AssumeVH);
+ if (!DT.dominates(CI, Latch->getTerminator()))
+ continue;
- if (CheckAssumptions(LHS) || CheckAssumptions(RHS))
- return true;
+ if (isImpliedCond(Pred, LHS, RHS, CI->getArgOperand(0), false))
+ return true;
+ }
// If the loop is not reachable from the entry block, we risk running into an
// infinite loop as we walk up into the dom tree. These loops do not matter
}
// Check conditions due to any @llvm.assume intrinsics.
- auto CheckAssumptions = [&](const SCEV *S) {
- auto AMI = AffectedMap.find(S);
- if (AMI != AffectedMap.end())
- for (auto *Assume : AMI->second) {
- auto *CI = cast<CallInst>(Assume);
- if (!DT.dominates(CI, L->getHeader()))
- continue;
-
- if (isImpliedCond(Pred, LHS, RHS, CI->getArgOperand(0), false))
- return true;
- }
-
- return false;
- };
+ for (auto &AssumeVH : AC.assumptions()) {
+ if (!AssumeVH)
+ continue;
+ auto *CI = cast<CallInst>(AssumeVH);
+ if (!DT.dominates(CI, L->getHeader()))
+ continue;
- if (CheckAssumptions(LHS) || CheckAssumptions(RHS))
- return true;
+ if (isImpliedCond(Pred, LHS, RHS, CI->getArgOperand(0), false))
+ return true;
+ }
return false;
}
//===----------------------------------------------------------------------===//
ScalarEvolution::ScalarEvolution(Function &F, TargetLibraryInfo &TLI,
- DominatorTree &DT, LoopInfo &LI)
- : F(F), TLI(TLI), DT(DT), LI(LI),
+ AssumptionCache &AC, DominatorTree &DT,
+ LoopInfo &LI)
+ : F(F), TLI(TLI), AC(AC), DT(DT), LI(LI),
CouldNotCompute(new SCEVCouldNotCompute()),
WalkingBEDominatingConds(false), ProvingSplitPredicate(false),
ValuesAtScopes(64), LoopDispositions(64), BlockDispositions(64),
}
ScalarEvolution::ScalarEvolution(ScalarEvolution &&Arg)
- : F(Arg.F), HasGuards(Arg.HasGuards), TLI(Arg.TLI), DT(Arg.DT),
+ : F(Arg.F), HasGuards(Arg.HasGuards), TLI(Arg.TLI), AC(Arg.AC), DT(Arg.DT),
LI(Arg.LI), CouldNotCompute(std::move(Arg.CouldNotCompute)),
ValueExprMap(std::move(Arg.ValueExprMap)),
PendingLoopPredicates(std::move(Arg.PendingLoopPredicates)),
// Gather stringified backedge taken counts for all loops using a fresh
// ScalarEvolution object.
- ScalarEvolution SE2(F, TLI, DT, LI);
+ ScalarEvolution SE2(F, TLI, AC, DT, LI);
for (LoopInfo::reverse_iterator I = LI.rbegin(), E = LI.rend(); I != E; ++I)
getLoopBackedgeTakenCounts(*I, BackedgeDumpsNew, SE2);
ScalarEvolution ScalarEvolutionAnalysis::run(Function &F,
FunctionAnalysisManager &AM) {
return ScalarEvolution(F, AM.getResult<TargetLibraryAnalysis>(F),
+ AM.getResult<AssumptionAnalysis>(F),
AM.getResult<DominatorTreeAnalysis>(F),
AM.getResult<LoopAnalysis>(F));
}
INITIALIZE_PASS_BEGIN(ScalarEvolutionWrapperPass, "scalar-evolution",
"Scalar Evolution Analysis", false, true)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(LoopInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
bool ScalarEvolutionWrapperPass::runOnFunction(Function &F) {
SE.reset(new ScalarEvolution(
F, getAnalysis<TargetLibraryInfoWrapperPass>().getTLI(),
+ getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F),
getAnalysis<DominatorTreeWrapperPass>().getDomTree(),
getAnalysis<LoopInfoWrapperPass>().getLoopInfo()));
return false;
void ScalarEvolutionWrapperPass::getAnalysisUsage(AnalysisUsage &AU) const {
AU.setPreservesAll();
+ AU.addRequiredTransitive<AssumptionCacheTracker>();
AU.addRequiredTransitive<LoopInfoWrapperPass>();
AU.addRequiredTransitive<DominatorTreeWrapperPass>();
AU.addRequiredTransitive<TargetLibraryInfoWrapperPass>();
// so narrow phis can reuse them.
for (PHINode *Phi : Phis) {
auto SimplifyPHINode = [&](PHINode *PN) -> Value * {
- if (Value *V = SimplifyInstruction(PN, DL, &SE.TLI, &SE.DT))
+ if (Value *V = SimplifyInstruction(PN, DL, &SE.TLI, &SE.DT, &SE.AC))
return V;
if (!SE.isSCEVable(PN->getType()))
return nullptr;
#include "llvm/Analysis/ValueTracking.h"
#include "llvm/ADT/Optional.h"
#include "llvm/ADT/SmallPtrSet.h"
-#include "llvm/ADT/SmallSet.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/MemoryBuiltins.h"
#include "llvm/Analysis/Loads.h"
// figuring out if we can use it.
struct Query {
const DataLayout &DL;
+ AssumptionCache *AC;
const Instruction *CxtI;
const DominatorTree *DT;
std::array<const Value *, MaxDepth> Excluded;
unsigned NumExcluded;
- Query(const DataLayout &DL, const Instruction *CxtI, const DominatorTree *DT)
- : DL(DL), CxtI(CxtI), DT(DT), NumExcluded(0) {}
+ Query(const DataLayout &DL, AssumptionCache *AC, const Instruction *CxtI,
+ const DominatorTree *DT)
+ : DL(DL), AC(AC), CxtI(CxtI), DT(DT), NumExcluded(0) {}
Query(const Query &Q, const Value *NewExcl)
- : DL(Q.DL), CxtI(Q.CxtI), DT(Q.DT), NumExcluded(Q.NumExcluded) {
+ : DL(Q.DL), AC(Q.AC), CxtI(Q.CxtI), DT(Q.DT), NumExcluded(Q.NumExcluded) {
Excluded = Q.Excluded;
Excluded[NumExcluded++] = NewExcl;
assert(NumExcluded <= Excluded.size());
void llvm::computeKnownBits(const Value *V, APInt &KnownZero, APInt &KnownOne,
const DataLayout &DL, unsigned Depth,
- const Instruction *CxtI,
+ AssumptionCache *AC, const Instruction *CxtI,
const DominatorTree *DT) {
::computeKnownBits(V, KnownZero, KnownOne, Depth,
- Query(DL, safeCxtI(V, CxtI), DT));
+ Query(DL, AC, safeCxtI(V, CxtI), DT));
}
bool llvm::haveNoCommonBitsSet(const Value *LHS, const Value *RHS,
const DataLayout &DL,
- const Instruction *CxtI,
+ AssumptionCache *AC, const Instruction *CxtI,
const DominatorTree *DT) {
assert(LHS->getType() == RHS->getType() &&
"LHS and RHS should have the same type");
IntegerType *IT = cast<IntegerType>(LHS->getType()->getScalarType());
APInt LHSKnownZero(IT->getBitWidth(), 0), LHSKnownOne(IT->getBitWidth(), 0);
APInt RHSKnownZero(IT->getBitWidth(), 0), RHSKnownOne(IT->getBitWidth(), 0);
- computeKnownBits(LHS, LHSKnownZero, LHSKnownOne, DL, 0, CxtI, DT);
- computeKnownBits(RHS, RHSKnownZero, RHSKnownOne, DL, 0, CxtI, DT);
+ computeKnownBits(LHS, LHSKnownZero, LHSKnownOne, DL, 0, AC, CxtI, DT);
+ computeKnownBits(RHS, RHSKnownZero, RHSKnownOne, DL, 0, AC, CxtI, DT);
return (LHSKnownZero | RHSKnownZero).isAllOnesValue();
}
void llvm::ComputeSignBit(const Value *V, bool &KnownZero, bool &KnownOne,
const DataLayout &DL, unsigned Depth,
- const Instruction *CxtI,
+ AssumptionCache *AC, const Instruction *CxtI,
const DominatorTree *DT) {
::ComputeSignBit(V, KnownZero, KnownOne, Depth,
- Query(DL, safeCxtI(V, CxtI), DT));
+ Query(DL, AC, safeCxtI(V, CxtI), DT));
}
static bool isKnownToBeAPowerOfTwo(const Value *V, bool OrZero, unsigned Depth,
bool llvm::isKnownToBeAPowerOfTwo(const Value *V, const DataLayout &DL,
bool OrZero,
- unsigned Depth, const Instruction *CxtI,
+ unsigned Depth, AssumptionCache *AC,
+ const Instruction *CxtI,
const DominatorTree *DT) {
return ::isKnownToBeAPowerOfTwo(V, OrZero, Depth,
- Query(DL, safeCxtI(V, CxtI), DT));
+ Query(DL, AC, safeCxtI(V, CxtI), DT));
}
static bool isKnownNonZero(const Value *V, unsigned Depth, const Query &Q);
bool llvm::isKnownNonZero(const Value *V, const DataLayout &DL, unsigned Depth,
- const Instruction *CxtI, const DominatorTree *DT) {
- return ::isKnownNonZero(V, Depth, Query(DL, safeCxtI(V, CxtI), DT));
+ AssumptionCache *AC, const Instruction *CxtI,
+ const DominatorTree *DT) {
+ return ::isKnownNonZero(V, Depth, Query(DL, AC, safeCxtI(V, CxtI), DT));
}
bool llvm::isKnownNonNegative(const Value *V, const DataLayout &DL,
- unsigned Depth, const Instruction *CxtI,
+ unsigned Depth,
+ AssumptionCache *AC, const Instruction *CxtI,
const DominatorTree *DT) {
bool NonNegative, Negative;
- ComputeSignBit(V, NonNegative, Negative, DL, Depth, CxtI, DT);
+ ComputeSignBit(V, NonNegative, Negative, DL, Depth, AC, CxtI, DT);
return NonNegative;
}
bool llvm::isKnownPositive(const Value *V, const DataLayout &DL, unsigned Depth,
- const Instruction *CxtI, const DominatorTree *DT) {
+ AssumptionCache *AC, const Instruction *CxtI,
+ const DominatorTree *DT) {
if (auto *CI = dyn_cast<ConstantInt>(V))
return CI->getValue().isStrictlyPositive();
// TODO: We'd doing two recursive queries here. We should factor this such
// that only a single query is needed.
- return isKnownNonNegative(V, DL, Depth, CxtI, DT) &&
- isKnownNonZero(V, DL, Depth, CxtI, DT);
+ return isKnownNonNegative(V, DL, Depth, AC, CxtI, DT) &&
+ isKnownNonZero(V, DL, Depth, AC, CxtI, DT);
}
bool llvm::isKnownNegative(const Value *V, const DataLayout &DL, unsigned Depth,
- const Instruction *CxtI, const DominatorTree *DT) {
+ AssumptionCache *AC, const Instruction *CxtI,
+ const DominatorTree *DT) {
bool NonNegative, Negative;
- ComputeSignBit(V, NonNegative, Negative, DL, Depth, CxtI, DT);
+ ComputeSignBit(V, NonNegative, Negative, DL, Depth, AC, CxtI, DT);
return Negative;
}
static bool isKnownNonEqual(const Value *V1, const Value *V2, const Query &Q);
bool llvm::isKnownNonEqual(const Value *V1, const Value *V2,
- const DataLayout &DL, const Instruction *CxtI,
+ const DataLayout &DL,
+ AssumptionCache *AC, const Instruction *CxtI,
const DominatorTree *DT) {
- return ::isKnownNonEqual(V1, V2, Query(DL, safeCxtI(V1, safeCxtI(V2, CxtI)),
+ return ::isKnownNonEqual(V1, V2, Query(DL, AC,
+ safeCxtI(V1, safeCxtI(V2, CxtI)),
DT));
}
bool llvm::MaskedValueIsZero(const Value *V, const APInt &Mask,
const DataLayout &DL,
- unsigned Depth, const Instruction *CxtI,
- const DominatorTree *DT) {
+ unsigned Depth, AssumptionCache *AC,
+ const Instruction *CxtI, const DominatorTree *DT) {
return ::MaskedValueIsZero(V, Mask, Depth,
- Query(DL, safeCxtI(V, CxtI), DT));
+ Query(DL, AC, safeCxtI(V, CxtI), DT));
}
static unsigned ComputeNumSignBits(const Value *V, unsigned Depth,
const Query &Q);
unsigned llvm::ComputeNumSignBits(const Value *V, const DataLayout &DL,
- unsigned Depth, const Instruction *CxtI,
+ unsigned Depth, AssumptionCache *AC,
+ const Instruction *CxtI,
const DominatorTree *DT) {
- return ::ComputeNumSignBits(V, Depth, Query(DL, safeCxtI(V, CxtI), DT));
+ return ::ComputeNumSignBits(V, Depth, Query(DL, AC, safeCxtI(V, CxtI), DT));
}
static void computeKnownBitsAddSub(bool Add, const Value *Op0, const Value *Op1,
const Query &Q) {
// Use of assumptions is context-sensitive. If we don't have a context, we
// cannot use them!
- if (!Q.CxtI)
+ if (!Q.AC || !Q.CxtI)
return;
unsigned BitWidth = KnownZero.getBitWidth();
- for (auto *U : V->users()) {
- auto *II = dyn_cast<IntrinsicInst>(U);
- if (!II)
- continue;
- if (II->getIntrinsicID() != Intrinsic::assume)
+ for (auto &AssumeVH : Q.AC->assumptions()) {
+ if (!AssumeVH)
continue;
- if (Q.isExcluded(II))
+ CallInst *I = cast<CallInst>(AssumeVH);
+ assert(I->getParent()->getParent() == Q.CxtI->getParent()->getParent() &&
+ "Got assumption for the wrong function!");
+ if (Q.isExcluded(I))
continue;
- Value *Arg = II->getArgOperand(0);
+ // Warning: This loop can end up being somewhat performance sensetive.
+ // We're running this loop for once for each value queried resulting in a
+ // runtime of ~O(#assumes * #values).
- if (Arg == V && isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ assert(I->getCalledFunction()->getIntrinsicID() == Intrinsic::assume &&
+ "must be an assume intrinsic");
+
+ Value *Arg = I->getArgOperand(0);
+
+ if (Arg == V && isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
assert(BitWidth == 1 && "assume operand is not i1?");
KnownZero.clearAllBits();
KnownOne.setAllBits();
return;
}
- // Note that the patterns below need to be kept in sync with the code
- // in InstCombiner::visitCallInst that adds relevant values to each
- // assume's operand bundles.
-
// The remaining tests are all recursive, so bail out if we hit the limit.
if (Depth == MaxDepth)
continue;
ConstantInt *C;
// assume(v = a)
if (match(Arg, m_c_ICmp(Pred, m_V, m_Value(A))) &&
- Pred == ICmpInst::ICMP_EQ && isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ Pred == ICmpInst::ICMP_EQ && isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
KnownZero |= RHSKnownZero;
KnownOne |= RHSKnownOne;
// assume(v & b = a)
} else if (match(Arg,
m_c_ICmp(Pred, m_c_And(m_V, m_Value(B)), m_Value(A))) &&
Pred == ICmpInst::ICMP_EQ &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
APInt MaskKnownZero(BitWidth, 0), MaskKnownOne(BitWidth, 0);
- computeKnownBits(B, MaskKnownZero, MaskKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(B, MaskKnownZero, MaskKnownOne, Depth+1, Query(Q, I));
// For those bits in the mask that are known to be one, we can propagate
// known bits from the RHS to V.
} else if (match(Arg, m_c_ICmp(Pred, m_Not(m_c_And(m_V, m_Value(B))),
m_Value(A))) &&
Pred == ICmpInst::ICMP_EQ &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
APInt MaskKnownZero(BitWidth, 0), MaskKnownOne(BitWidth, 0);
- computeKnownBits(B, MaskKnownZero, MaskKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(B, MaskKnownZero, MaskKnownOne, Depth+1, Query(Q, I));
// For those bits in the mask that are known to be one, we can propagate
// inverted known bits from the RHS to V.
} else if (match(Arg,
m_c_ICmp(Pred, m_c_Or(m_V, m_Value(B)), m_Value(A))) &&
Pred == ICmpInst::ICMP_EQ &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
APInt BKnownZero(BitWidth, 0), BKnownOne(BitWidth, 0);
- computeKnownBits(B, BKnownZero, BKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(B, BKnownZero, BKnownOne, Depth+1, Query(Q, I));
// For those bits in B that are known to be zero, we can propagate known
// bits from the RHS to V.
} else if (match(Arg, m_c_ICmp(Pred, m_Not(m_c_Or(m_V, m_Value(B))),
m_Value(A))) &&
Pred == ICmpInst::ICMP_EQ &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
APInt BKnownZero(BitWidth, 0), BKnownOne(BitWidth, 0);
- computeKnownBits(B, BKnownZero, BKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(B, BKnownZero, BKnownOne, Depth+1, Query(Q, I));
// For those bits in B that are known to be zero, we can propagate
// inverted known bits from the RHS to V.
} else if (match(Arg,
m_c_ICmp(Pred, m_c_Xor(m_V, m_Value(B)), m_Value(A))) &&
Pred == ICmpInst::ICMP_EQ &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
APInt BKnownZero(BitWidth, 0), BKnownOne(BitWidth, 0);
- computeKnownBits(B, BKnownZero, BKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(B, BKnownZero, BKnownOne, Depth+1, Query(Q, I));
// For those bits in B that are known to be zero, we can propagate known
// bits from the RHS to V. For those bits in B that are known to be one,
} else if (match(Arg, m_c_ICmp(Pred, m_Not(m_c_Xor(m_V, m_Value(B))),
m_Value(A))) &&
Pred == ICmpInst::ICMP_EQ &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
APInt BKnownZero(BitWidth, 0), BKnownOne(BitWidth, 0);
- computeKnownBits(B, BKnownZero, BKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(B, BKnownZero, BKnownOne, Depth+1, Query(Q, I));
// For those bits in B that are known to be zero, we can propagate
// inverted known bits from the RHS to V. For those bits in B that are
} else if (match(Arg, m_c_ICmp(Pred, m_Shl(m_V, m_ConstantInt(C)),
m_Value(A))) &&
Pred == ICmpInst::ICMP_EQ &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
// For those bits in RHS that are known, we can propagate them to known
// bits in V shifted to the right by C.
KnownZero |= RHSKnownZero.lshr(C->getZExtValue());
} else if (match(Arg, m_c_ICmp(Pred, m_Not(m_Shl(m_V, m_ConstantInt(C))),
m_Value(A))) &&
Pred == ICmpInst::ICMP_EQ &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
// For those bits in RHS that are known, we can propagate them inverted
// to known bits in V shifted to the right by C.
KnownZero |= RHSKnownOne.lshr(C->getZExtValue());
m_AShr(m_V, m_ConstantInt(C))),
m_Value(A))) &&
Pred == ICmpInst::ICMP_EQ &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
// For those bits in RHS that are known, we can propagate them to known
// bits in V shifted to the right by C.
KnownZero |= RHSKnownZero << C->getZExtValue();
m_AShr(m_V, m_ConstantInt(C)))),
m_Value(A))) &&
Pred == ICmpInst::ICMP_EQ &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
// For those bits in RHS that are known, we can propagate them inverted
// to known bits in V shifted to the right by C.
KnownZero |= RHSKnownOne << C->getZExtValue();
// assume(v >=_s c) where c is non-negative
} else if (match(Arg, m_ICmp(Pred, m_V, m_Value(A))) &&
Pred == ICmpInst::ICMP_SGE &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
if (RHSKnownZero.isNegative()) {
// We know that the sign bit is zero.
// assume(v >_s c) where c is at least -1.
} else if (match(Arg, m_ICmp(Pred, m_V, m_Value(A))) &&
Pred == ICmpInst::ICMP_SGT &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
if (RHSKnownOne.isAllOnesValue() || RHSKnownZero.isNegative()) {
// We know that the sign bit is zero.
// assume(v <=_s c) where c is negative
} else if (match(Arg, m_ICmp(Pred, m_V, m_Value(A))) &&
Pred == ICmpInst::ICMP_SLE &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
if (RHSKnownOne.isNegative()) {
// We know that the sign bit is one.
// assume(v <_s c) where c is non-positive
} else if (match(Arg, m_ICmp(Pred, m_V, m_Value(A))) &&
Pred == ICmpInst::ICMP_SLT &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
if (RHSKnownZero.isAllOnesValue() || RHSKnownOne.isNegative()) {
// We know that the sign bit is one.
// assume(v <=_u c)
} else if (match(Arg, m_ICmp(Pred, m_V, m_Value(A))) &&
Pred == ICmpInst::ICMP_ULE &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
// Whatever high bits in c are zero are known to be zero.
KnownZero |=
// assume(v <_u c)
} else if (match(Arg, m_ICmp(Pred, m_V, m_Value(A))) &&
Pred == ICmpInst::ICMP_ULT &&
- isValidAssumeForContext(II, Q.CxtI, Q.DT)) {
+ isValidAssumeForContext(I, Q.CxtI, Q.DT)) {
APInt RHSKnownZero(BitWidth, 0), RHSKnownOne(BitWidth, 0);
- computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, II));
+ computeKnownBits(A, RHSKnownZero, RHSKnownOne, Depth+1, Query(Q, I));
// Whatever high bits in c are zero are known to be zero (if c is a power
// of 2, then one more).
- if (isKnownToBeAPowerOfTwo(A, false, Depth + 1, Query(Q, II)))
+ if (isKnownToBeAPowerOfTwo(A, false, Depth + 1, Query(Q, I)))
KnownZero |=
APInt::getHighBitsSet(BitWidth, RHSKnownZero.countLeadingOnes()+1);
else
// See if InstructionSimplify knows any relevant tricks.
if (Instruction *I = dyn_cast<Instruction>(V))
- // TODO: Acquire a DominatorTree and use it.
+ // TODO: Acquire a DominatorTree and AssumptionCache and use them.
if (Value *Simplified = SimplifyInstruction(I, DL, nullptr)) {
V = Simplified;
continue;
OverflowResult llvm::computeOverflowForUnsignedMul(const Value *LHS,
const Value *RHS,
const DataLayout &DL,
+ AssumptionCache *AC,
const Instruction *CxtI,
const DominatorTree *DT) {
// Multiplying n * m significant bits yields a result of n + m significant
APInt LHSKnownOne(BitWidth, 0);
APInt RHSKnownZero(BitWidth, 0);
APInt RHSKnownOne(BitWidth, 0);
- computeKnownBits(LHS, LHSKnownZero, LHSKnownOne, DL, /*Depth=*/0, CxtI, DT);
- computeKnownBits(RHS, RHSKnownZero, RHSKnownOne, DL, /*Depth=*/0, CxtI, DT);
+ computeKnownBits(LHS, LHSKnownZero, LHSKnownOne, DL, /*Depth=*/0, AC, CxtI,
+ DT);
+ computeKnownBits(RHS, RHSKnownZero, RHSKnownOne, DL, /*Depth=*/0, AC, CxtI,
+ DT);
// Note that underestimating the number of zero bits gives a more
// conservative answer.
unsigned ZeroBits = LHSKnownZero.countLeadingOnes() +
OverflowResult llvm::computeOverflowForUnsignedAdd(const Value *LHS,
const Value *RHS,
const DataLayout &DL,
+ AssumptionCache *AC,
const Instruction *CxtI,
const DominatorTree *DT) {
bool LHSKnownNonNegative, LHSKnownNegative;
ComputeSignBit(LHS, LHSKnownNonNegative, LHSKnownNegative, DL, /*Depth=*/0,
- CxtI, DT);
+ AC, CxtI, DT);
if (LHSKnownNonNegative || LHSKnownNegative) {
bool RHSKnownNonNegative, RHSKnownNegative;
ComputeSignBit(RHS, RHSKnownNonNegative, RHSKnownNegative, DL, /*Depth=*/0,
- CxtI, DT);
+ AC, CxtI, DT);
if (LHSKnownNegative && RHSKnownNegative) {
// The sign bit is set in both cases: this MUST overflow.
const Value *RHS,
const AddOperator *Add,
const DataLayout &DL,
+ AssumptionCache *AC,
const Instruction *CxtI,
const DominatorTree *DT) {
if (Add && Add->hasNoSignedWrap()) {
bool LHSKnownNonNegative, LHSKnownNegative;
bool RHSKnownNonNegative, RHSKnownNegative;
ComputeSignBit(LHS, LHSKnownNonNegative, LHSKnownNegative, DL, /*Depth=*/0,
- CxtI, DT);
+ AC, CxtI, DT);
ComputeSignBit(RHS, RHSKnownNonNegative, RHSKnownNegative, DL, /*Depth=*/0,
- CxtI, DT);
+ AC, CxtI, DT);
if ((LHSKnownNonNegative && RHSKnownNegative) ||
(LHSKnownNegative && RHSKnownNonNegative)) {
if (LHSOrRHSKnownNonNegative || LHSOrRHSKnownNegative) {
bool AddKnownNonNegative, AddKnownNegative;
ComputeSignBit(Add, AddKnownNonNegative, AddKnownNegative, DL,
- /*Depth=*/0, CxtI, DT);
+ /*Depth=*/0, AC, CxtI, DT);
if ((AddKnownNonNegative && LHSOrRHSKnownNonNegative) ||
(AddKnownNegative && LHSOrRHSKnownNegative)) {
return OverflowResult::NeverOverflows;
OverflowResult llvm::computeOverflowForSignedAdd(const AddOperator *Add,
const DataLayout &DL,
+ AssumptionCache *AC,
const Instruction *CxtI,
const DominatorTree *DT) {
return ::computeOverflowForSignedAdd(Add->getOperand(0), Add->getOperand(1),
- Add, DL, CxtI, DT);
+ Add, DL, AC, CxtI, DT);
}
OverflowResult llvm::computeOverflowForSignedAdd(const Value *LHS,
const Value *RHS,
const DataLayout &DL,
+ AssumptionCache *AC,
const Instruction *CxtI,
const DominatorTree *DT) {
- return ::computeOverflowForSignedAdd(LHS, RHS, nullptr, DL, CxtI, DT);
+ return ::computeOverflowForSignedAdd(LHS, RHS, nullptr, DL, AC, CxtI, DT);
}
bool llvm::isGuaranteedToTransferExecutionToSuccessor(const Instruction *I) {
static bool isTruePredicate(CmpInst::Predicate Pred,
const Value *LHS, const Value *RHS,
const DataLayout &DL, unsigned Depth,
- const Instruction *CxtI, const DominatorTree *DT) {
+ AssumptionCache *AC, const Instruction *CxtI,
+ const DominatorTree *DT) {
assert(!LHS->getType()->isVectorTy() && "TODO: extend to handle vectors!");
if (ICmpInst::isTrueWhenEqual(Pred) && LHS == RHS)
return true;
match(B, m_Or(m_Specific(X), m_APInt(CB)))) {
unsigned BitWidth = CA->getBitWidth();
APInt KnownZero(BitWidth, 0), KnownOne(BitWidth, 0);
- computeKnownBits(X, KnownZero, KnownOne, DL, Depth + 1, CxtI, DT);
+ computeKnownBits(X, KnownZero, KnownOne, DL, Depth + 1, AC, CxtI, DT);
if ((KnownZero & *CA) == *CA && (KnownZero & *CB) == *CB)
return true;
isImpliedCondOperands(CmpInst::Predicate Pred, const Value *ALHS,
const Value *ARHS, const Value *BLHS,
const Value *BRHS, const DataLayout &DL,
- unsigned Depth, const Instruction *CxtI,
- const DominatorTree *DT) {
+ unsigned Depth, AssumptionCache *AC,
+ const Instruction *CxtI, const DominatorTree *DT) {
switch (Pred) {
default:
return None;
case CmpInst::ICMP_SLT:
case CmpInst::ICMP_SLE:
- if (isTruePredicate(CmpInst::ICMP_SLE, BLHS, ALHS, DL, Depth, CxtI,
+ if (isTruePredicate(CmpInst::ICMP_SLE, BLHS, ALHS, DL, Depth, AC, CxtI,
DT) &&
- isTruePredicate(CmpInst::ICMP_SLE, ARHS, BRHS, DL, Depth, CxtI, DT))
+ isTruePredicate(CmpInst::ICMP_SLE, ARHS, BRHS, DL, Depth, AC, CxtI, DT))
return true;
return None;
case CmpInst::ICMP_ULT:
case CmpInst::ICMP_ULE:
- if (isTruePredicate(CmpInst::ICMP_ULE, BLHS, ALHS, DL, Depth, CxtI, DT) &&
- isTruePredicate(CmpInst::ICMP_ULE, ARHS, BRHS, DL, Depth, CxtI, DT))
+ if (isTruePredicate(CmpInst::ICMP_ULE, BLHS, ALHS, DL, Depth, AC, CxtI,
+ DT) &&
+ isTruePredicate(CmpInst::ICMP_ULE, ARHS, BRHS, DL, Depth, AC, CxtI, DT))
return true;
return None;
}
Optional<bool> llvm::isImpliedCondition(const Value *LHS, const Value *RHS,
const DataLayout &DL, bool InvertAPred,
- unsigned Depth, const Instruction *CxtI,
+ unsigned Depth, AssumptionCache *AC,
+ const Instruction *CxtI,
const DominatorTree *DT) {
// A mismatch occurs when we compare a scalar cmp to a vector cmp, for example.
if (LHS->getType() != RHS->getType())
}
if (APred == BPred)
- return isImpliedCondOperands(APred, ALHS, ARHS, BLHS, BRHS, DL, Depth, CxtI,
- DT);
+ return isImpliedCondOperands(APred, ALHS, ARHS, BLHS, BRHS, DL, Depth, AC,
+ CxtI, DT);
return None;
}
#include "llvm/ADT/StringSwitch.h"
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/Analysis/AliasAnalysisEvaluator.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/BasicAliasAnalysis.h"
#include "llvm/Analysis/BlockFrequencyInfo.h"
#include "llvm/Analysis/BlockFrequencyInfoImpl.h"
#define FUNCTION_ANALYSIS(NAME, CREATE_PASS)
#endif
FUNCTION_ANALYSIS("aa", AAManager())
+FUNCTION_ANALYSIS("assumptions", AssumptionAnalysis())
FUNCTION_ANALYSIS("block-freq", BlockFrequencyAnalysis())
FUNCTION_ANALYSIS("branch-prob", BranchProbabilityAnalysis())
FUNCTION_ANALYSIS("domtree", DominatorTreeAnalysis())
FUNCTION_PASS("loop-distribute", LoopDistributePass())
FUNCTION_PASS("loop-vectorize", LoopVectorizePass())
FUNCTION_PASS("print", PrintFunctionPass(dbgs()))
+FUNCTION_PASS("print<assumptions>", AssumptionPrinterPass(dbgs()))
FUNCTION_PASS("print<block-freq>", BlockFrequencyPrinterPass(dbgs()))
FUNCTION_PASS("print<branch-prob>", BranchProbabilityPrinterPass(dbgs()))
FUNCTION_PASS("print<domtree>", DominatorTreePrinterPass(dbgs()))
#include "llvm/Transforms/IPO/AlwaysInliner.h"
#include "llvm/ADT/SetVector.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/CallGraph.h"
#include "llvm/Analysis/InlineCost.h"
#include "llvm/Analysis/ProfileSummaryInfo.h"
char AlwaysInlinerLegacyPass::ID = 0;
INITIALIZE_PASS_BEGIN(AlwaysInlinerLegacyPass, "always-inline",
"Inliner for always_inline functions", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(CallGraphWrapperPass)
INITIALIZE_PASS_DEPENDENCY(ProfileSummaryInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
#include "llvm/ADT/Statistic.h"
#include "llvm/ADT/StringExtras.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/BasicAliasAnalysis.h"
#include "llvm/Analysis/CallGraph.h"
#include "llvm/Analysis/CallGraphSCCPass.h"
///
struct ArgPromotion : public CallGraphSCCPass {
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<TargetLibraryInfoWrapperPass>();
getAAResultsAnalysisUsage(AU);
CallGraphSCCPass::getAnalysisUsage(AU);
char ArgPromotion::ID = 0;
INITIALIZE_PASS_BEGIN(ArgPromotion, "argpromotion",
"Promote 'by reference' arguments to scalars", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(CallGraphWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
INITIALIZE_PASS_END(ArgPromotion, "argpromotion",
#include "llvm/ADT/Statistic.h"
#include "llvm/ADT/StringSwitch.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/BasicAliasAnalysis.h"
#include "llvm/Analysis/CallGraph.h"
#include "llvm/Analysis/CallGraphSCCPass.h"
void getAnalysisUsage(AnalysisUsage &AU) const override {
AU.setPreservesCFG();
+ AU.addRequired<AssumptionCacheTracker>();
getAAResultsAnalysisUsage(AU);
CallGraphSCCPass::getAnalysisUsage(AU);
}
char PostOrderFunctionAttrsLegacyPass::ID = 0;
INITIALIZE_PASS_BEGIN(PostOrderFunctionAttrsLegacyPass, "functionattrs",
"Deduce function attributes", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(CallGraphWrapperPass)
INITIALIZE_PASS_END(PostOrderFunctionAttrsLegacyPass, "functionattrs",
"Deduce function attributes", false, false)
//
//===----------------------------------------------------------------------===//
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/CallGraph.h"
#include "llvm/Analysis/InlineCost.h"
#include "llvm/Analysis/ProfileSummaryInfo.h"
InlineCost getInlineCost(CallSite CS) override {
Function *Callee = CS.getCalledFunction();
TargetTransformInfo &TTI = TTIWP->getTTI(*Callee);
- return llvm::getInlineCost(CS, Params, TTI, PSI);
+ std::function<AssumptionCache &(Function &)> GetAssumptionCache =
+ [&](Function &F) -> AssumptionCache & {
+ return ACT->getAssumptionCache(F);
+ };
+ return llvm::getInlineCost(CS, Params, TTI, GetAssumptionCache, PSI);
}
bool runOnSCC(CallGraphSCC &SCC) override;
char SimpleInliner::ID = 0;
INITIALIZE_PASS_BEGIN(SimpleInliner, "inline", "Function Integration/Inlining",
false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(CallGraphWrapperPass)
INITIALIZE_PASS_DEPENDENCY(ProfileSummaryInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/BasicAliasAnalysis.h"
#include "llvm/Analysis/CallGraph.h"
#include "llvm/Analysis/InlineCost.h"
/// If the derived class implements this method, it should
/// always explicitly call the implementation here.
void Inliner::getAnalysisUsage(AnalysisUsage &AU) const {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<ProfileSummaryInfoWrapperPass>();
AU.addRequired<TargetLibraryInfoWrapperPass>();
getAAResultsAnalysisUsage(AU);
static bool
inlineCallsImpl(CallGraphSCC &SCC, CallGraph &CG,
+ std::function<AssumptionCache &(Function &)> GetAssumptionCache,
ProfileSummaryInfo *PSI, TargetLibraryInfo &TLI,
bool InsertLifetime,
function_ref<InlineCost(CallSite CS)> GetInlineCost,
std::swap(CallSites[i--], CallSites[--FirstCallInSCC]);
InlinedArrayAllocasTy InlinedArrayAllocas;
- InlineFunctionInfo InlineInfo(&CG);
+ InlineFunctionInfo InlineInfo(&CG, &GetAssumptionCache);
// Now that we have all of the call sites, loop over them and inline them if
// it looks profitable to do so.
bool Inliner::inlineCalls(CallGraphSCC &SCC) {
CallGraph &CG = getAnalysis<CallGraphWrapperPass>().getCallGraph();
+ ACT = &getAnalysis<AssumptionCacheTracker>();
PSI = getAnalysis<ProfileSummaryInfoWrapperPass>().getPSI();
auto &TLI = getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
// We compute dedicated AA results for each function in the SCC as needed. We
AAR.emplace(createLegacyPMAAResults(*this, F, *BAR));
return *AAR;
};
- return inlineCallsImpl(SCC, CG, PSI, TLI, InsertLifetime,
+ auto GetAssumptionCache = [&](Function &F) -> AssumptionCache & {
+ return ACT->getAssumptionCache(F);
+ };
+ return inlineCallsImpl(SCC, CG, GetAssumptionCache, PSI, TLI, InsertLifetime,
[this](CallSite CS) { return getInlineCost(CS); },
AARGetter, ImportedFunctionsStats);
}
initializePartialInlinerLegacyPassPass(*PassRegistry::getPassRegistry());
}
+ void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
+ }
bool runOnModule(Module &M) override {
if (skipModule(M))
return false;
- InlineFunctionInfo IFI(nullptr);
+ AssumptionCacheTracker *ACT = &getAnalysis<AssumptionCacheTracker>();
+ std::function<AssumptionCache &(Function &)> GetAssumptionCache =
+ [&ACT](Function &F) -> AssumptionCache & {
+ return ACT->getAssumptionCache(F);
+ };
+ InlineFunctionInfo IFI(nullptr, &GetAssumptionCache);
return PartialInlinerImpl(IFI).run(M);
}
};
}
char PartialInlinerLegacyPass::ID = 0;
-INITIALIZE_PASS(PartialInlinerLegacyPass, "partial-inliner",
- "Partial Inliner", false, false)
+INITIALIZE_PASS_BEGIN(PartialInlinerLegacyPass, "partial-inliner",
+ "Partial Inliner", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
+INITIALIZE_PASS_END(PartialInlinerLegacyPass, "partial-inliner",
+ "Partial Inliner", false, false)
ModulePass *llvm::createPartialInliningPass() {
return new PartialInlinerLegacyPass();
PreservedAnalyses PartialInlinerPass::run(Module &M,
ModuleAnalysisManager &AM) {
- InlineFunctionInfo IFI(nullptr);
+ auto &FAM = AM.getResult<FunctionAnalysisManagerModuleProxy>(M).getManager();
+ std::function<AssumptionCache &(Function &)> GetAssumptionCache =
+ [&FAM](Function &F) -> AssumptionCache & {
+ return FAM.getResult<AssumptionAnalysis>(F);
+ };
+ InlineFunctionInfo IFI(nullptr, &GetAssumptionCache);
if (PartialInlinerImpl(IFI).run(M))
return PreservedAnalyses::none();
return PreservedAnalyses::all();
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/SmallSet.h"
#include "llvm/ADT/StringRef.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/PostDominators.h"
#include "llvm/IR/Constants.h"
class SampleProfileLoader {
public:
SampleProfileLoader(StringRef Name = SampleProfileFile)
- : DT(nullptr), PDT(nullptr), LI(nullptr), Reader(), Samples(nullptr),
- Filename(Name), ProfileIsValid(false), TotalCollectedSamples(0) {}
+ : DT(nullptr), PDT(nullptr), LI(nullptr), ACT(nullptr), Reader(),
+ Samples(nullptr), Filename(Name), ProfileIsValid(false),
+ TotalCollectedSamples(0) {}
bool doInitialization(Module &M);
bool runOnModule(Module &M);
+ void setACT(AssumptionCacheTracker *A) { ACT = A; }
void dump() { Reader->dump(); }
std::unique_ptr<DominatorTreeBase<BasicBlock>> PDT;
std::unique_ptr<LoopInfo> LI;
+ AssumptionCacheTracker *ACT;
+
/// \brief Predecessors for each basic block in the CFG.
BlockEdgeMap Predecessors;
StringRef getPassName() const override { return "Sample profile pass"; }
bool runOnModule(Module &M) override;
+ void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
+ }
+
private:
SampleProfileLoader SampleLoader;
};
bool SampleProfileLoader::inlineHotFunctions(Function &F) {
bool Changed = false;
LLVMContext &Ctx = F.getContext();
+ std::function<AssumptionCache &(Function &)> GetAssumptionCache = [&](
+ Function &F) -> AssumptionCache & { return ACT->getAssumptionCache(F); };
while (true) {
bool LocalChanged = false;
SmallVector<Instruction *, 10> CIS;
}
}
for (auto I : CIS) {
- InlineFunctionInfo IFI(nullptr);
+ InlineFunctionInfo IFI(nullptr, ACT ? &GetAssumptionCache : nullptr);
CallSite CS(I);
Function *CalledFunction = CS.getCalledFunction();
if (!CalledFunction || !CalledFunction->getSubprogram())
}
char SampleProfileLoaderLegacyPass::ID = 0;
-INITIALIZE_PASS(SampleProfileLoaderLegacyPass, "sample-profile",
- "Sample Profile loader", false, false)
+INITIALIZE_PASS_BEGIN(SampleProfileLoaderLegacyPass, "sample-profile",
+ "Sample Profile loader", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
+INITIALIZE_PASS_END(SampleProfileLoaderLegacyPass, "sample-profile",
+ "Sample Profile loader", false, false)
bool SampleProfileLoader::doInitialization(Module &M) {
auto &Ctx = M.getContext();
}
bool SampleProfileLoaderLegacyPass::runOnModule(Module &M) {
+ // FIXME: pass in AssumptionCache correctly for the new pass manager.
+ SampleLoader.setACT(&getAnalysis<AssumptionCacheTracker>());
return SampleLoader.runOnModule(M);
}
return replaceInstUsesWith(I, V);
if (Value *V = SimplifyAddInst(LHS, RHS, I.hasNoSignedWrap(),
- I.hasNoUnsignedWrap(), DL, &TLI, &DT))
+ I.hasNoUnsignedWrap(), DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
// (A*B)+(A*C) -> A*(B+C) etc
return replaceInstUsesWith(I, V);
// A+B --> A|B iff A and B have no bits set in common.
- if (haveNoCommonBitsSet(LHS, RHS, DL, &I, &DT))
+ if (haveNoCommonBitsSet(LHS, RHS, DL, &AC, &I, &DT))
return BinaryOperator::CreateOr(LHS, RHS);
if (Constant *CRHS = dyn_cast<Constant>(RHS)) {
return replaceInstUsesWith(I, V);
if (Value *V =
- SimplifyFAddInst(LHS, RHS, I.getFastMathFlags(), DL, &TLI, &DT))
+ SimplifyFAddInst(LHS, RHS, I.getFastMathFlags(), DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
if (isa<Constant>(RHS)) {
return replaceInstUsesWith(I, V);
if (Value *V = SimplifySubInst(Op0, Op1, I.hasNoSignedWrap(),
- I.hasNoUnsignedWrap(), DL, &TLI, &DT))
+ I.hasNoUnsignedWrap(), DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
// (A*B)-(A*C) -> A*(B-C) etc
return replaceInstUsesWith(I, V);
if (Value *V =
- SimplifyFSubInst(Op0, Op1, I.getFastMathFlags(), DL, &TLI, &DT))
+ SimplifyFSubInst(Op0, Op1, I.getFastMathFlags(), DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
// fsub nsz 0, X ==> fsub nsz -0.0, X
if (Value *V = SimplifyVectorOp(I))
return replaceInstUsesWith(I, V);
- if (Value *V = SimplifyAndInst(Op0, Op1, DL, &TLI, &DT))
+ if (Value *V = SimplifyAndInst(Op0, Op1, DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
// (A|B)&(A|C) -> A|(B&C) etc
Value *Mask = nullptr;
Value *Masked = nullptr;
if (LAnd->getOperand(0) == RAnd->getOperand(0) &&
- isKnownToBeAPowerOfTwo(LAnd->getOperand(1), DL, false, 0, CxtI,
+ isKnownToBeAPowerOfTwo(LAnd->getOperand(1), DL, false, 0, &AC, CxtI,
&DT) &&
- isKnownToBeAPowerOfTwo(RAnd->getOperand(1), DL, false, 0, CxtI,
+ isKnownToBeAPowerOfTwo(RAnd->getOperand(1), DL, false, 0, &AC, CxtI,
&DT)) {
Mask = Builder->CreateOr(LAnd->getOperand(1), RAnd->getOperand(1));
Masked = Builder->CreateAnd(LAnd->getOperand(0), Mask);
} else if (LAnd->getOperand(1) == RAnd->getOperand(1) &&
- isKnownToBeAPowerOfTwo(LAnd->getOperand(0), DL, false, 0, CxtI,
- &DT) &&
- isKnownToBeAPowerOfTwo(RAnd->getOperand(0), DL, false, 0, CxtI,
- &DT)) {
+ isKnownToBeAPowerOfTwo(LAnd->getOperand(0), DL, false, 0, &AC,
+ CxtI, &DT) &&
+ isKnownToBeAPowerOfTwo(RAnd->getOperand(0), DL, false, 0, &AC,
+ CxtI, &DT)) {
Mask = Builder->CreateOr(LAnd->getOperand(0), RAnd->getOperand(0));
Masked = Builder->CreateAnd(LAnd->getOperand(1), Mask);
}
if (Value *V = SimplifyVectorOp(I))
return replaceInstUsesWith(I, V);
- if (Value *V = SimplifyOrInst(Op0, Op1, DL, &TLI, &DT))
+ if (Value *V = SimplifyOrInst(Op0, Op1, DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
// (A&B)|(A&C) -> A&(B|C) etc
if (Value *V = SimplifyVectorOp(I))
return replaceInstUsesWith(I, V);
- if (Value *V = SimplifyXorInst(Op0, Op1, DL, &TLI, &DT))
+ if (Value *V = SimplifyXorInst(Op0, Op1, DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
// (A&B)^(A&C) -> A&(B^C) etc
}
Instruction *InstCombiner::SimplifyMemTransfer(MemIntrinsic *MI) {
- unsigned DstAlign = getKnownAlignment(MI->getArgOperand(0), DL, MI, &DT);
- unsigned SrcAlign = getKnownAlignment(MI->getArgOperand(1), DL, MI, &DT);
+ unsigned DstAlign = getKnownAlignment(MI->getArgOperand(0), DL, MI, &AC, &DT);
+ unsigned SrcAlign = getKnownAlignment(MI->getArgOperand(1), DL, MI, &AC, &DT);
unsigned MinAlign = std::min(DstAlign, SrcAlign);
unsigned CopyAlign = MI->getAlignment();
}
Instruction *InstCombiner::SimplifyMemSet(MemSetInst *MI) {
- unsigned Alignment = getKnownAlignment(MI->getDest(), DL, MI, &DT);
+ unsigned Alignment = getKnownAlignment(MI->getDest(), DL, MI, &AC, &DT);
if (MI->getAlignment() < Alignment) {
MI->setAlignment(ConstantInt::get(MI->getAlignmentType(),
Alignment, false));
Instruction *InstCombiner::visitCallInst(CallInst &CI) {
auto Args = CI.arg_operands();
if (Value *V = SimplifyCall(CI.getCalledValue(), Args.begin(), Args.end(), DL,
- &TLI, &DT))
+ &TLI, &DT, &AC))
return replaceInstUsesWith(CI, V);
if (isFreeCall(&CI, &TLI))
case Intrinsic::ppc_altivec_lvx:
case Intrinsic::ppc_altivec_lvxl:
// Turn PPC lvx -> load if the pointer is known aligned.
- if (getOrEnforceKnownAlignment(II->getArgOperand(0), 16, DL, II,
+ if (getOrEnforceKnownAlignment(II->getArgOperand(0), 16, DL, II, &AC,
&DT) >= 16) {
Value *Ptr = Builder->CreateBitCast(II->getArgOperand(0),
PointerType::getUnqual(II->getType()));
case Intrinsic::ppc_altivec_stvx:
case Intrinsic::ppc_altivec_stvxl:
// Turn stvx -> store if the pointer is known aligned.
- if (getOrEnforceKnownAlignment(II->getArgOperand(1), 16, DL, II,
+ if (getOrEnforceKnownAlignment(II->getArgOperand(1), 16, DL, II, &AC,
&DT) >= 16) {
Type *OpPtrTy =
PointerType::getUnqual(II->getArgOperand(0)->getType());
}
case Intrinsic::ppc_qpx_qvlfs:
// Turn PPC QPX qvlfs -> load if the pointer is known aligned.
- if (getOrEnforceKnownAlignment(II->getArgOperand(0), 16, DL, II,
+ if (getOrEnforceKnownAlignment(II->getArgOperand(0), 16, DL, II, &AC,
&DT) >= 16) {
Type *VTy = VectorType::get(Builder->getFloatTy(),
II->getType()->getVectorNumElements());
break;
case Intrinsic::ppc_qpx_qvlfd:
// Turn PPC QPX qvlfd -> load if the pointer is known aligned.
- if (getOrEnforceKnownAlignment(II->getArgOperand(0), 32, DL, II,
+ if (getOrEnforceKnownAlignment(II->getArgOperand(0), 32, DL, II, &AC,
&DT) >= 32) {
Value *Ptr = Builder->CreateBitCast(II->getArgOperand(0),
PointerType::getUnqual(II->getType()));
break;
case Intrinsic::ppc_qpx_qvstfs:
// Turn PPC QPX qvstfs -> store if the pointer is known aligned.
- if (getOrEnforceKnownAlignment(II->getArgOperand(1), 16, DL, II,
+ if (getOrEnforceKnownAlignment(II->getArgOperand(1), 16, DL, II, &AC,
&DT) >= 16) {
Type *VTy = VectorType::get(Builder->getFloatTy(),
II->getArgOperand(0)->getType()->getVectorNumElements());
break;
case Intrinsic::ppc_qpx_qvstfd:
// Turn PPC QPX qvstfd -> store if the pointer is known aligned.
- if (getOrEnforceKnownAlignment(II->getArgOperand(1), 32, DL, II,
+ if (getOrEnforceKnownAlignment(II->getArgOperand(1), 32, DL, II, &AC,
&DT) >= 32) {
Type *OpPtrTy =
PointerType::getUnqual(II->getArgOperand(0)->getType());
case Intrinsic::arm_neon_vst3lane:
case Intrinsic::arm_neon_vst4lane: {
unsigned MemAlign =
- getKnownAlignment(II->getArgOperand(0), DL, II, &DT);
+ getKnownAlignment(II->getArgOperand(0), DL, II, &AC, &DT);
unsigned AlignArg = II->getNumArgOperands() - 1;
ConstantInt *IntrAlign = dyn_cast<ConstantInt>(II->getArgOperand(AlignArg));
if (IntrAlign && IntrAlign->getZExtValue() < MemAlign) {
if (KnownOne.isAllOnesValue())
return eraseInstFromFunction(*II);
- // For assumptions, add to the associated operand bundle the values to which
- // the assumption might apply.
- // Note: This code must be kept in-sync with the code in
- // computeKnownBitsFromAssume in ValueTracking.
- SmallVector<Value *, 16> Affected;
- auto AddAffected = [&Affected](Value *V) {
- if (isa<Argument>(V)) {
- Affected.push_back(V);
- } else if (auto *I = dyn_cast<Instruction>(V)) {
- Affected.push_back(I);
-
- if (I->getOpcode() == Instruction::BitCast ||
- I->getOpcode() == Instruction::PtrToInt) {
- V = I->getOperand(0);
- if (isa<Instruction>(V) || isa<Argument>(V))
- Affected.push_back(V);
- }
- }
- };
-
- CmpInst::Predicate Pred;
- if (match(IIOperand, m_ICmp(Pred, m_Value(A), m_Value(B)))) {
- AddAffected(A);
- AddAffected(B);
-
- if (Pred == ICmpInst::ICMP_EQ) {
- // For equality comparisons, we handle the case of bit inversion.
- auto AddAffectedFromEq = [&AddAffected](Value *V) {
- Value *A;
- if (match(V, m_Not(m_Value(A)))) {
- AddAffected(A);
- V = A;
- }
-
- Value *B;
- ConstantInt *C;
- if (match(V,
- m_CombineOr(m_And(m_Value(A), m_Value(B)),
- m_CombineOr(m_Or(m_Value(A), m_Value(B)),
- m_Xor(m_Value(A), m_Value(B)))))) {
- AddAffected(A);
- AddAffected(B);
- } else if (match(V,
- m_CombineOr(m_Shl(m_Value(A), m_ConstantInt(C)),
- m_CombineOr(m_LShr(m_Value(A), m_ConstantInt(C)),
- m_AShr(m_Value(A),
- m_ConstantInt(C)))))) {
- AddAffected(A);
- }
- };
-
- AddAffectedFromEq(A);
- AddAffectedFromEq(B);
- }
- }
-
- // If the list of affected values is the same as the existing list then
- // there's nothing more to do here.
- if (!Affected.empty())
- if (auto OB = CI.getOperandBundle("affected"))
- if (Affected.size() == OB.getValue().Inputs.size() &&
- std::equal(Affected.begin(), Affected.end(),
- OB.getValue().Inputs.begin()))
- Affected.clear();
-
- if (!Affected.empty()) {
- Builder->CreateCall(AssumeIntrinsic, IIOperand,
- OperandBundleDef("affected", Affected),
- II->getName());
- return eraseInstFromFunction(*II);
- }
-
break;
}
case Intrinsic::experimental_gc_relocate: {
}
if (Value *V =
- SimplifyICmpInst(I.getPredicate(), Op0, Op1, DL, &TLI, &DT, &I))
+ SimplifyICmpInst(I.getPredicate(), Op0, Op1, DL, &TLI, &DT, &AC, &I))
return replaceInstUsesWith(I, V);
// comparing -val or val with non-zero is the same as just comparing val
// if A is a power of 2.
if (match(Op0, m_And(m_Value(A), m_Not(m_Value(B)))) &&
match(Op1, m_Zero()) &&
- isKnownToBeAPowerOfTwo(A, DL, false, 0, &I, &DT) && I.isEquality())
+ isKnownToBeAPowerOfTwo(A, DL, false, 0, &AC, &I, &DT) && I.isEquality())
return new ICmpInst(I.getInversePredicate(),
Builder->CreateAnd(A, B),
Op1);
Value *Op0 = I.getOperand(0), *Op1 = I.getOperand(1);
if (Value *V = SimplifyFCmpInst(I.getPredicate(), Op0, Op1,
- I.getFastMathFlags(), DL, &TLI, &DT, &I))
+ I.getFastMathFlags(), DL, &TLI, &DT, &AC, &I))
return replaceInstUsesWith(I, V);
// Simplify 'fcmp pred X, X'
#define LLVM_LIB_TRANSFORMS_INSTCOMBINE_INSTCOMBINEINTERNAL_H
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/TargetFolder.h"
#include "llvm/Analysis/ValueTracking.h"
AliasAnalysis *AA;
// Required analyses.
+ AssumptionCache &AC;
TargetLibraryInfo &TLI;
DominatorTree &DT;
const DataLayout &DL;
public:
InstCombiner(InstCombineWorklist &Worklist, BuilderTy *Builder,
bool MinimizeSize, bool ExpensiveCombines, AliasAnalysis *AA,
- TargetLibraryInfo &TLI, DominatorTree &DT, const DataLayout &DL,
- LoopInfo *LI)
+ AssumptionCache &AC, TargetLibraryInfo &TLI,
+ DominatorTree &DT, const DataLayout &DL, LoopInfo *LI)
: Worklist(Worklist), Builder(Builder), MinimizeSize(MinimizeSize),
- ExpensiveCombines(ExpensiveCombines), AA(AA), TLI(TLI), DT(DT), DL(DL),
- LI(LI), MadeIRChange(false) {}
+ ExpensiveCombines(ExpensiveCombines), AA(AA), AC(AC), TLI(TLI), DT(DT),
+ DL(DL), LI(LI), MadeIRChange(false) {}
/// \brief Run the combiner over the entire worklist until it is empty.
///
/// \returns true if the IR is changed.
bool run();
+ AssumptionCache &getAssumptionCache() const { return AC; }
+
const DataLayout &getDataLayout() const { return DL; }
DominatorTree &getDominatorTree() const { return DT; }
void computeKnownBits(Value *V, APInt &KnownZero, APInt &KnownOne,
unsigned Depth, Instruction *CxtI) const {
- return llvm::computeKnownBits(V, KnownZero, KnownOne, DL, Depth, CxtI, &DT);
+ return llvm::computeKnownBits(V, KnownZero, KnownOne, DL, Depth, &AC, CxtI,
+ &DT);
}
bool MaskedValueIsZero(Value *V, const APInt &Mask, unsigned Depth = 0,
Instruction *CxtI = nullptr) const {
- return llvm::MaskedValueIsZero(V, Mask, DL, Depth, CxtI, &DT);
+ return llvm::MaskedValueIsZero(V, Mask, DL, Depth, &AC, CxtI, &DT);
}
unsigned ComputeNumSignBits(Value *Op, unsigned Depth = 0,
Instruction *CxtI = nullptr) const {
- return llvm::ComputeNumSignBits(Op, DL, Depth, CxtI, &DT);
+ return llvm::ComputeNumSignBits(Op, DL, Depth, &AC, CxtI, &DT);
}
void ComputeSignBit(Value *V, bool &KnownZero, bool &KnownOne,
unsigned Depth = 0, Instruction *CxtI = nullptr) const {
- return llvm::ComputeSignBit(V, KnownZero, KnownOne, DL, Depth, CxtI, &DT);
+ return llvm::ComputeSignBit(V, KnownZero, KnownOne, DL, Depth, &AC, CxtI,
+ &DT);
}
OverflowResult computeOverflowForUnsignedMul(Value *LHS, Value *RHS,
const Instruction *CxtI) {
- return llvm::computeOverflowForUnsignedMul(LHS, RHS, DL, CxtI, &DT);
+ return llvm::computeOverflowForUnsignedMul(LHS, RHS, DL, &AC, CxtI, &DT);
}
OverflowResult computeOverflowForUnsignedAdd(Value *LHS, Value *RHS,
const Instruction *CxtI) {
- return llvm::computeOverflowForUnsignedAdd(LHS, RHS, DL, CxtI, &DT);
+ return llvm::computeOverflowForUnsignedAdd(LHS, RHS, DL, &AC, CxtI, &DT);
}
private:
SmallVector<Instruction *, 4> ToDelete;
if (MemTransferInst *Copy = isOnlyCopiedFromConstantGlobal(&AI, ToDelete)) {
unsigned SourceAlign = getOrEnforceKnownAlignment(
- Copy->getSource(), AI.getAlignment(), DL, &AI, &DT);
+ Copy->getSource(), AI.getAlignment(), DL, &AI, &AC, &DT);
if (AI.getAlignment() <= SourceAlign) {
DEBUG(dbgs() << "Found alloca equal to global: " << AI << '\n');
DEBUG(dbgs() << " memcpy = " << *Copy << '\n');
// Attempt to improve the alignment.
unsigned KnownAlign = getOrEnforceKnownAlignment(
- Op, DL.getPrefTypeAlignment(LI.getType()), DL, &LI, &DT);
+ Op, DL.getPrefTypeAlignment(LI.getType()), DL, &LI, &AC, &DT);
unsigned LoadAlign = LI.getAlignment();
unsigned EffectiveLoadAlign =
LoadAlign != 0 ? LoadAlign : DL.getABITypeAlignment(LI.getType());
// Attempt to improve the alignment.
unsigned KnownAlign = getOrEnforceKnownAlignment(
- Ptr, DL.getPrefTypeAlignment(Val->getType()), DL, &SI, &DT);
+ Ptr, DL.getPrefTypeAlignment(Val->getType()), DL, &SI, &AC, &DT);
unsigned StoreAlign = SI.getAlignment();
unsigned EffectiveStoreAlign =
StoreAlign != 0 ? StoreAlign : DL.getABITypeAlignment(Val->getType());
BinaryOperator *I = dyn_cast<BinaryOperator>(V);
if (I && I->isLogicalShift() &&
isKnownToBeAPowerOfTwo(I->getOperand(0), IC.getDataLayout(), false, 0,
- &CxtI, &IC.getDominatorTree())) {
+ &IC.getAssumptionCache(), &CxtI,
+ &IC.getDominatorTree())) {
// We know that this is an exact/nuw shift and that the input is a
// non-zero context as well.
if (Value *V2 = simplifyValueKnownNonZero(I->getOperand(0), IC, CxtI)) {
if (Value *V = SimplifyVectorOp(I))
return replaceInstUsesWith(I, V);
- if (Value *V = SimplifyMulInst(Op0, Op1, DL, &TLI, &DT))
+ if (Value *V = SimplifyMulInst(Op0, Op1, DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
if (Value *V = SimplifyUsingDistributiveLaws(I))
std::swap(Op0, Op1);
if (Value *V =
- SimplifyFMulInst(Op0, Op1, I.getFastMathFlags(), DL, &TLI, &DT))
+ SimplifyFMulInst(Op0, Op1, I.getFastMathFlags(), DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
bool AllowReassociate = I.hasUnsafeAlgebra();
if (Value *V = SimplifyVectorOp(I))
return replaceInstUsesWith(I, V);
- if (Value *V = SimplifyUDivInst(Op0, Op1, DL, &TLI, &DT))
+ if (Value *V = SimplifyUDivInst(Op0, Op1, DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
// Handle the integer div common cases
if (Value *V = SimplifyVectorOp(I))
return replaceInstUsesWith(I, V);
- if (Value *V = SimplifySDivInst(Op0, Op1, DL, &TLI, &DT))
+ if (Value *V = SimplifySDivInst(Op0, Op1, DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
// Handle the integer div common cases
return BO;
}
- if (isKnownToBeAPowerOfTwo(Op1, DL, /*OrZero*/ true, 0, &I, &DT)) {
+ if (isKnownToBeAPowerOfTwo(Op1, DL, /*OrZero*/ true, 0, &AC, &I, &DT)) {
// X sdiv (1 << Y) -> X udiv (1 << Y) ( -> X u>> Y)
// Safe because the only negative value (1 << Y) can take on is
// INT_MIN, and X sdiv INT_MIN == X udiv INT_MIN == 0 if X doesn't have
return replaceInstUsesWith(I, V);
if (Value *V = SimplifyFDivInst(Op0, Op1, I.getFastMathFlags(),
- DL, &TLI, &DT))
+ DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
if (isa<Constant>(Op0))
if (Value *V = SimplifyVectorOp(I))
return replaceInstUsesWith(I, V);
- if (Value *V = SimplifyURemInst(Op0, Op1, DL, &TLI, &DT))
+ if (Value *V = SimplifyURemInst(Op0, Op1, DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
if (Instruction *common = commonIRemTransforms(I))
I.getType());
// X urem Y -> X and Y-1, where Y is a power of 2,
- if (isKnownToBeAPowerOfTwo(Op1, DL, /*OrZero*/ true, 0, &I, &DT)) {
+ if (isKnownToBeAPowerOfTwo(Op1, DL, /*OrZero*/ true, 0, &AC, &I, &DT)) {
Constant *N1 = Constant::getAllOnesValue(I.getType());
Value *Add = Builder->CreateAdd(Op1, N1);
return BinaryOperator::CreateAnd(Op0, Add);
if (Value *V = SimplifyVectorOp(I))
return replaceInstUsesWith(I, V);
- if (Value *V = SimplifySRemInst(Op0, Op1, DL, &TLI, &DT))
+ if (Value *V = SimplifySRemInst(Op0, Op1, DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
// Handle the integer rem common cases
return replaceInstUsesWith(I, V);
if (Value *V = SimplifyFRemInst(Op0, Op1, I.getFastMathFlags(),
- DL, &TLI, &DT))
+ DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
// Handle cases involving: rem X, (select Cond, Y, Z)
// PHINode simplification
//
Instruction *InstCombiner::visitPHINode(PHINode &PN) {
- if (Value *V = SimplifyInstruction(&PN, DL, &TLI, &DT))
+ if (Value *V = SimplifyInstruction(&PN, DL, &TLI, &DT, &AC))
return replaceInstUsesWith(PN, V);
if (Instruction *Result = FoldPHIArgZextsIntoPHI(PN))
for (unsigned i = 0, e = PN.getNumIncomingValues(); i != e; ++i) {
Instruction *CtxI = PN.getIncomingBlock(i)->getTerminator();
Value *VA = PN.getIncomingValue(i);
- if (isKnownNonZero(VA, DL, 0, CtxI, &DT)) {
+ if (isKnownNonZero(VA, DL, 0, &AC, CtxI, &DT)) {
if (!NonZeroConst)
NonZeroConst = GetAnyNonZeroConstInt(PN);
PN.setIncomingValue(i, NonZeroConst);
Type *SelType = SI.getType();
if (Value *V =
- SimplifySelectInst(CondVal, TrueVal, FalseVal, DL, &TLI, &DT))
+ SimplifySelectInst(CondVal, TrueVal, FalseVal, DL, &TLI, &DT, &AC))
return replaceInstUsesWith(SI, V);
if (Instruction *I = canonicalizeSelectToShuffle(SI))
if (Value *V =
SimplifyShlInst(I.getOperand(0), I.getOperand(1), I.hasNoSignedWrap(),
- I.hasNoUnsignedWrap(), DL, &TLI, &DT))
+ I.hasNoUnsignedWrap(), DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
if (Instruction *V = commonShiftTransforms(I))
return replaceInstUsesWith(I, V);
if (Value *V = SimplifyLShrInst(I.getOperand(0), I.getOperand(1), I.isExact(),
- DL, &TLI, &DT))
+ DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
if (Instruction *R = commonShiftTransforms(I))
return replaceInstUsesWith(I, V);
if (Value *V = SimplifyAShrInst(I.getOperand(0), I.getOperand(1), I.isExact(),
- DL, &TLI, &DT))
+ DL, &TLI, &DT, &AC))
return replaceInstUsesWith(I, V);
if (Instruction *R = commonShiftTransforms(I))
Instruction *InstCombiner::visitExtractElementInst(ExtractElementInst &EI) {
if (Value *V = SimplifyExtractElementInst(
- EI.getVectorOperand(), EI.getIndexOperand(), DL, &TLI, &DT))
+ EI.getVectorOperand(), EI.getIndexOperand(), DL, &TLI, &DT, &AC))
return replaceInstUsesWith(EI, V);
// If vector val is constant with all elements the same, replace EI with
#include "llvm/ADT/Statistic.h"
#include "llvm/ADT/StringSwitch.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/BasicAliasAnalysis.h"
#include "llvm/Analysis/CFG.h"
#include "llvm/Analysis/ConstantFolding.h"
if (SI0->getCondition() == SI1->getCondition()) {
Value *SI = nullptr;
if (Value *V = SimplifyBinOp(TopLevelOpcode, SI0->getFalseValue(),
- SI1->getFalseValue(), DL, &TLI, &DT))
+ SI1->getFalseValue(), DL, &TLI, &DT, &AC))
SI = Builder->CreateSelect(SI0->getCondition(),
Builder->CreateBinOp(TopLevelOpcode,
SI0->getTrueValue(),
SI1->getTrueValue()),
V);
if (Value *V = SimplifyBinOp(TopLevelOpcode, SI0->getTrueValue(),
- SI1->getTrueValue(), DL, &TLI, &DT))
+ SI1->getTrueValue(), DL, &TLI, &DT, &AC))
SI = Builder->CreateSelect(
SI0->getCondition(), V,
Builder->CreateBinOp(TopLevelOpcode, SI0->getFalseValue(),
SmallVector<Value*, 8> Ops(GEP.op_begin(), GEP.op_end());
if (Value *V =
- SimplifyGEPInst(GEP.getSourceElementType(), Ops, DL, &TLI, &DT))
+ SimplifyGEPInst(GEP.getSourceElementType(), Ops, DL, &TLI, &DT, &AC))
return replaceInstUsesWith(GEP, V);
Value *PtrOp = GEP.getOperand(0);
return replaceInstUsesWith(EV, Agg);
if (Value *V =
- SimplifyExtractValueInst(Agg, EV.getIndices(), DL, &TLI, &DT))
+ SimplifyExtractValueInst(Agg, EV.getIndices(), DL, &TLI, &DT, &AC))
return replaceInstUsesWith(EV, V);
if (InsertValueInst *IV = dyn_cast<InsertValueInst>(Agg)) {
static bool
combineInstructionsOverFunction(Function &F, InstCombineWorklist &Worklist,
- AliasAnalysis *AA, TargetLibraryInfo &TLI,
- DominatorTree &DT,
+ AliasAnalysis *AA, AssumptionCache &AC,
+ TargetLibraryInfo &TLI, DominatorTree &DT,
bool ExpensiveCombines = true,
LoopInfo *LI = nullptr) {
auto &DL = F.getParent()->getDataLayout();
/// instructions into the worklist when they are created.
IRBuilder<TargetFolder, IRBuilderCallbackInserter> Builder(
F.getContext(), TargetFolder(DL),
- IRBuilderCallbackInserter([&Worklist](Instruction *I) {
+ IRBuilderCallbackInserter([&Worklist, &AC](Instruction *I) {
Worklist.Add(I);
+
+ using namespace llvm::PatternMatch;
+ if (match(I, m_Intrinsic<Intrinsic::assume>()))
+ AC.registerAssumption(cast<CallInst>(I));
}));
// Lower dbg.declare intrinsics otherwise their value may be clobbered
bool Changed = prepareICWorklistFromFunction(F, DL, &TLI, Worklist);
InstCombiner IC(Worklist, &Builder, F.optForMinSize(), ExpensiveCombines,
- AA, TLI, DT, DL, LI);
+ AA, AC, TLI, DT, DL, LI);
Changed |= IC.run();
if (!Changed)
PreservedAnalyses InstCombinePass::run(Function &F,
FunctionAnalysisManager &AM) {
+ auto &AC = AM.getResult<AssumptionAnalysis>(F);
auto &DT = AM.getResult<DominatorTreeAnalysis>(F);
auto &TLI = AM.getResult<TargetLibraryAnalysis>(F);
auto *LI = AM.getCachedResult<LoopAnalysis>(F);
// FIXME: The AliasAnalysis is not yet supported in the new pass manager
- if (!combineInstructionsOverFunction(F, Worklist, nullptr, TLI, DT,
+ if (!combineInstructionsOverFunction(F, Worklist, nullptr, AC, TLI, DT,
ExpensiveCombines, LI))
// No changes, all analyses are preserved.
return PreservedAnalyses::all();
void InstructionCombiningPass::getAnalysisUsage(AnalysisUsage &AU) const {
AU.setPreservesCFG();
AU.addRequired<AAResultsWrapperPass>();
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<TargetLibraryInfoWrapperPass>();
AU.addRequired<DominatorTreeWrapperPass>();
AU.addPreserved<DominatorTreeWrapperPass>();
// Required analyses.
auto AA = &getAnalysis<AAResultsWrapperPass>().getAAResults();
+ auto &AC = getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
auto &TLI = getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
auto &DT = getAnalysis<DominatorTreeWrapperPass>().getDomTree();
auto *LIWP = getAnalysisIfAvailable<LoopInfoWrapperPass>();
auto *LI = LIWP ? &LIWP->getLoopInfo() : nullptr;
- return combineInstructionsOverFunction(F, Worklist, AA, TLI, DT,
+ return combineInstructionsOverFunction(F, Worklist, AA, AC, TLI, DT,
ExpensiveCombines, LI);
}
char InstructionCombiningPass::ID = 0;
INITIALIZE_PASS_BEGIN(InstructionCombiningPass, "instcombine",
"Combine redundant instructions", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(AAResultsWrapperPass)
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/Analysis/GlobalsModRef.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/ScalarEvolutionExpressions.h"
#include "llvm/Analysis/ValueTracking.h"
bool runOnFunction(Function &F) override;
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<ScalarEvolutionWrapperPass>();
AU.addRequired<DominatorTreeWrapperPass>();
static const char aip_name[] = "Alignment from assumptions";
INITIALIZE_PASS_BEGIN(AlignmentFromAssumptions, AA_NAME,
aip_name, false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(ScalarEvolutionWrapperPass)
INITIALIZE_PASS_END(AlignmentFromAssumptions, AA_NAME,
if (skipFunction(F))
return false;
+ auto &AC = getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
ScalarEvolution *SE = &getAnalysis<ScalarEvolutionWrapperPass>().getSE();
DominatorTree *DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
- return Impl.runImpl(F, SE, DT);
+ return Impl.runImpl(F, AC, SE, DT);
}
-bool AlignmentFromAssumptionsPass::runImpl(Function &F, ScalarEvolution *SE_,
+bool AlignmentFromAssumptionsPass::runImpl(Function &F, AssumptionCache &AC,
+ ScalarEvolution *SE_,
DominatorTree *DT_) {
SE = SE_;
DT = DT_;
NewSrcAlignments.clear();
bool Changed = false;
-
- for (auto &B : F)
- for (auto &I : B)
- if (auto *II = dyn_cast<IntrinsicInst>(&I))
- if (II->getIntrinsicID() == Intrinsic::assume)
- Changed |= processAssumption(II);
+ for (auto &AssumeVH : AC.assumptions())
+ if (AssumeVH)
+ Changed |= processAssumption(cast<CallInst>(AssumeVH));
return Changed;
}
PreservedAnalyses
AlignmentFromAssumptionsPass::run(Function &F, FunctionAnalysisManager &AM) {
+ AssumptionCache &AC = AM.getResult<AssumptionAnalysis>(F);
ScalarEvolution &SE = AM.getResult<ScalarEvolutionAnalysis>(F);
DominatorTree &DT = AM.getResult<DominatorTreeAnalysis>(F);
- bool Changed = runImpl(F, &SE, &DT);
+ bool Changed = runImpl(F, AC, &SE, &DT);
// FIXME: We need to invalidate this to avoid PR28400. Is there a better
// solution?
#include "llvm/ADT/Hashing.h"
#include "llvm/ADT/ScopedHashTable.h"
#include "llvm/ADT/Statistic.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/GlobalsModRef.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
const TargetLibraryInfo &TLI;
const TargetTransformInfo &TTI;
DominatorTree &DT;
+ AssumptionCache &AC;
MemorySSA *MSSA;
typedef RecyclingAllocator<
BumpPtrAllocator, ScopedHashTableVal<SimpleValue, Value *>> AllocatorTy;
/// \brief Set up the EarlyCSE runner for a particular function.
EarlyCSE(const TargetLibraryInfo &TLI, const TargetTransformInfo &TTI,
- DominatorTree &DT, MemorySSA *MSSA)
- : TLI(TLI), TTI(TTI), DT(DT), MSSA(MSSA), CurrentGeneration(0) {}
+ DominatorTree &DT, AssumptionCache &AC, MemorySSA *MSSA)
+ : TLI(TLI), TTI(TTI), DT(DT), AC(AC), MSSA(MSSA), CurrentGeneration(0) {}
bool run();
// If the instruction can be simplified (e.g. X+0 = X) then replace it with
// its simpler value.
- if (Value *V = SimplifyInstruction(Inst, DL, &TLI, &DT)) {
+ if (Value *V = SimplifyInstruction(Inst, DL, &TLI, &DT, &AC)) {
DEBUG(dbgs() << "EarlyCSE Simplify: " << *Inst << " to: " << *V << '\n');
bool Killed = false;
if (!Inst->use_empty()) {
auto &TLI = AM.getResult<TargetLibraryAnalysis>(F);
auto &TTI = AM.getResult<TargetIRAnalysis>(F);
auto &DT = AM.getResult<DominatorTreeAnalysis>(F);
+ auto &AC = AM.getResult<AssumptionAnalysis>(F);
auto *MSSA =
UseMemorySSA ? &AM.getResult<MemorySSAAnalysis>(F).getMSSA() : nullptr;
- EarlyCSE CSE(TLI, TTI, DT, MSSA);
+ EarlyCSE CSE(TLI, TTI, DT, AC, MSSA);
if (!CSE.run())
return PreservedAnalyses::all();
auto &TLI = getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
auto &TTI = getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F);
auto &DT = getAnalysis<DominatorTreeWrapperPass>().getDomTree();
+ auto &AC = getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
auto *MSSA =
UseMemorySSA ? &getAnalysis<MemorySSAWrapperPass>().getMSSA() : nullptr;
- EarlyCSE CSE(TLI, TTI, DT, MSSA);
+ EarlyCSE CSE(TLI, TTI, DT, AC, MSSA);
return CSE.run();
}
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<DominatorTreeWrapperPass>();
AU.addRequired<TargetLibraryInfoWrapperPass>();
AU.addRequired<TargetTransformInfoWrapperPass>();
INITIALIZE_PASS_BEGIN(EarlyCSELegacyPass, "early-cse", "Early CSE", false,
false)
INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
INITIALIZE_PASS_END(EarlyCSELegacyPass, "early-cse", "Early CSE", false, false)
INITIALIZE_PASS_BEGIN(EarlyCSEMemSSALegacyPass, "early-cse-memssa",
"Early CSE w/ MemorySSA", false, false)
INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(MemorySSAWrapperPass)
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/CFG.h"
#include "llvm/Analysis/ConstantFolding.h"
#include "llvm/Analysis/GlobalsModRef.h"
// significant! Re-ordering these variables will cause GVN when run alone to
// be less effective! We should fix memdep and basic-aa to not exhibit this
// behavior, but until then don't change the order here.
+ auto &AC = AM.getResult<AssumptionAnalysis>(F);
auto &DT = AM.getResult<DominatorTreeAnalysis>(F);
auto &TLI = AM.getResult<TargetLibraryAnalysis>(F);
auto &AA = AM.getResult<AAManager>(F);
auto &MemDep = AM.getResult<MemoryDependenceAnalysis>(F);
auto *LI = AM.getCachedResult<LoopAnalysis>(F);
auto &ORE = AM.getResult<OptimizationRemarkEmitterAnalysis>(F);
- bool Changed = runImpl(F, DT, TLI, AA, &MemDep, LI, &ORE);
+ bool Changed = runImpl(F, AC, DT, TLI, AA, &MemDep, LI, &ORE);
if (!Changed)
return PreservedAnalyses::all();
PreservedAnalyses PA;
// If all preds have a single successor, then we know it is safe to insert
// the load on the pred (?!?), so we can insert code to materialize the
// pointer if it is not available.
- PHITransAddr Address(LI->getPointerOperand(), DL);
+ PHITransAddr Address(LI->getPointerOperand(), DL, AC);
Value *LoadPtr = nullptr;
LoadPtr = Address.PHITranslateWithInsertion(LoadBB, UnavailablePred,
*DT, NewInsts);
// example if it determines that %y is equal to %x then the instruction
// "%z = and i32 %x, %y" becomes "%z = and i32 %x, %x" which we now simplify.
const DataLayout &DL = I->getModule()->getDataLayout();
- if (Value *V = SimplifyInstruction(I, DL, TLI, DT)) {
+ if (Value *V = SimplifyInstruction(I, DL, TLI, DT, AC)) {
bool Changed = false;
if (!I->use_empty()) {
I->replaceAllUsesWith(V);
}
/// runOnFunction - This is the main transformation entry point for a function.
-bool GVN::runImpl(Function &F, DominatorTree &RunDT,
+bool GVN::runImpl(Function &F, AssumptionCache &RunAC, DominatorTree &RunDT,
const TargetLibraryInfo &RunTLI, AAResults &RunAA,
MemoryDependenceResults *RunMD, LoopInfo *LI,
OptimizationRemarkEmitter *RunORE) {
+ AC = &RunAC;
DT = &RunDT;
VN.setDomTree(DT);
TLI = &RunTLI;
auto *LIWP = getAnalysisIfAvailable<LoopInfoWrapperPass>();
return Impl.runImpl(
- F, getAnalysis<DominatorTreeWrapperPass>().getDomTree(),
+ F, getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F),
+ getAnalysis<DominatorTreeWrapperPass>().getDomTree(),
getAnalysis<TargetLibraryInfoWrapperPass>().getTLI(),
getAnalysis<AAResultsWrapperPass>().getAAResults(),
NoLoads ? nullptr
}
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<DominatorTreeWrapperPass>();
AU.addRequired<TargetLibraryInfoWrapperPass>();
if (!NoLoads)
}
INITIALIZE_PASS_BEGIN(GVNLegacyPass, "gvn", "Global Value Numbering", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(MemoryDependenceWrapperPass)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
auto *L = createClonedLoopStructure(
&OriginalLoop, OriginalLoop.getParentLoop(), PreLoop.Map);
formLCSSARecursively(*L, DT, &LI, &SE);
- simplifyLoop(L, &DT, &LI, &SE, true);
+ simplifyLoop(L, &DT, &LI, &SE, nullptr, true);
// Pre loops are slow paths, we do not need to perform any loop
// optimizations on them.
DisableAllLoopOptsOnLoop(*L);
auto *L = createClonedLoopStructure(
&OriginalLoop, OriginalLoop.getParentLoop(), PostLoop.Map);
formLCSSARecursively(*L, DT, &LI, &SE);
- simplifyLoop(L, &DT, &LI, &SE, true);
+ simplifyLoop(L, &DT, &LI, &SE, nullptr, true);
// Post loops are slow paths, we do not need to perform any loop
// optimizations on them.
DisableAllLoopOptsOnLoop(*L);
}
formLCSSARecursively(OriginalLoop, DT, &LI, &SE);
- simplifyLoop(&OriginalLoop, &DT, &LI, &SE, true);
+ simplifyLoop(&OriginalLoop, &DT, &LI, &SE, nullptr, true);
return true;
}
#define DEBUG_TYPE "loop-data-prefetch"
#include "llvm/ADT/DepthFirstIterator.h"
#include "llvm/ADT/Statistic.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/CodeMetrics.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/LoopInfo.h"
/// Loop prefetch implementation class.
class LoopDataPrefetch {
public:
- LoopDataPrefetch(LoopInfo *LI, ScalarEvolution *SE,
+ LoopDataPrefetch(AssumptionCache *AC, LoopInfo *LI, ScalarEvolution *SE,
const TargetTransformInfo *TTI,
OptimizationRemarkEmitter *ORE)
- : LI(LI), SE(SE), TTI(TTI), ORE(ORE) {}
+ : AC(AC), LI(LI), SE(SE), TTI(TTI), ORE(ORE) {}
bool run();
return TTI->getMaxPrefetchIterationsAhead();
}
+ AssumptionCache *AC;
LoopInfo *LI;
ScalarEvolution *SE;
const TargetTransformInfo *TTI;
}
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addPreserved<DominatorTreeWrapperPass>();
AU.addRequired<LoopInfoWrapperPass>();
AU.addPreserved<LoopInfoWrapperPass>();
char LoopDataPrefetchLegacyPass::ID = 0;
INITIALIZE_PASS_BEGIN(LoopDataPrefetchLegacyPass, "loop-data-prefetch",
"Loop Data Prefetch", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(LoopInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(OptimizationRemarkEmitterWrapperPass)
FunctionAnalysisManager &AM) {
LoopInfo *LI = &AM.getResult<LoopAnalysis>(F);
ScalarEvolution *SE = &AM.getResult<ScalarEvolutionAnalysis>(F);
+ AssumptionCache *AC = &AM.getResult<AssumptionAnalysis>(F);
OptimizationRemarkEmitter *ORE =
&AM.getResult<OptimizationRemarkEmitterAnalysis>(F);
const TargetTransformInfo *TTI = &AM.getResult<TargetIRAnalysis>(F);
- LoopDataPrefetch LDP(LI, SE, TTI, ORE);
+ LoopDataPrefetch LDP(AC, LI, SE, TTI, ORE);
bool Changed = LDP.run();
if (Changed) {
LoopInfo *LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo();
ScalarEvolution *SE = &getAnalysis<ScalarEvolutionWrapperPass>().getSE();
+ AssumptionCache *AC =
+ &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
OptimizationRemarkEmitter *ORE =
&getAnalysis<OptimizationRemarkEmitterWrapperPass>().getORE();
const TargetTransformInfo *TTI =
&getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F);
- LoopDataPrefetch LDP(LI, SE, TTI, ORE);
+ LoopDataPrefetch LDP(AC, LI, SE, TTI, ORE);
return LDP.run();
}
return MadeChange;
SmallPtrSet<const Value *, 32> EphValues;
- CodeMetrics::collectEphemeralValues(L, EphValues);
+ CodeMetrics::collectEphemeralValues(L, AC, EphValues);
// Calculate the number of iterations ahead to prefetch
CodeMetrics Metrics;
#include "llvm/Transforms/Scalar/LoopInstSimplify.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/ADT/Statistic.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/LoopPass.h"
STATISTIC(NumSimplified, "Number of redundant instructions simplified");
static bool SimplifyLoopInst(Loop *L, DominatorTree *DT, LoopInfo *LI,
+ AssumptionCache *AC,
const TargetLibraryInfo *TLI) {
SmallVector<BasicBlock *, 8> ExitBlocks;
L->getUniqueExitBlocks(ExitBlocks);
// Don't bother simplifying unused instructions.
if (!I->use_empty()) {
- Value *V = SimplifyInstruction(I, DL, TLI, DT);
+ Value *V = SimplifyInstruction(I, DL, TLI, DT, AC);
if (V && LI->replacementPreservesLCSSAForm(I, V)) {
// Mark all uses for resimplification next time round the loop.
for (User *U : I->users())
getAnalysisIfAvailable<DominatorTreeWrapperPass>();
DominatorTree *DT = DTWP ? &DTWP->getDomTree() : nullptr;
LoopInfo *LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo();
+ AssumptionCache *AC =
+ &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(
+ *L->getHeader()->getParent());
const TargetLibraryInfo *TLI =
&getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
- return SimplifyLoopInst(L, DT, LI, TLI);
+ return SimplifyLoopInst(L, DT, LI, AC, TLI);
}
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<TargetLibraryInfoWrapperPass>();
AU.setPreservesCFG();
getLoopAnalysisUsage(AU);
// Use getCachedResult because Loop pass cannot trigger a function analysis.
auto *DT = FAM.getCachedResult<DominatorTreeAnalysis>(*F);
auto *LI = FAM.getCachedResult<LoopAnalysis>(*F);
+ auto *AC = FAM.getCachedResult<AssumptionAnalysis>(*F);
const auto *TLI = FAM.getCachedResult<TargetLibraryAnalysis>(*F);
- assert((LI && TLI) && "Analyses for Loop Inst Simplify not available");
+ assert((LI && AC && TLI) && "Analyses for Loop Inst Simplify not available");
- if (!SimplifyLoopInst(&L, DT, LI, TLI))
+ if (!SimplifyLoopInst(&L, DT, LI, AC, TLI))
return PreservedAnalyses::all();
return getLoopPassPreservedAnalyses();
char LoopInstSimplifyLegacyPass::ID = 0;
INITIALIZE_PASS_BEGIN(LoopInstSimplifyLegacyPass, "loop-instsimplify",
"Simplify instructions in loops", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(LoopPass)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
INITIALIZE_PASS_END(LoopInstSimplifyLegacyPass, "loop-instsimplify",
#include "llvm/ADT/SmallVector.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/BlockFrequencyInfo.h"
#include "llvm/Analysis/CodeMetrics.h"
#include "llvm/Analysis/DependenceAnalysis.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/Analysis/BasicAliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/CodeMetrics.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/GlobalsModRef.h"
const unsigned MaxHeaderSize;
LoopInfo *LI;
const TargetTransformInfo *TTI;
+ AssumptionCache *AC;
DominatorTree *DT;
ScalarEvolution *SE;
public:
LoopRotate(unsigned MaxHeaderSize, LoopInfo *LI,
- const TargetTransformInfo *TTI, DominatorTree *DT,
- ScalarEvolution *SE)
- : MaxHeaderSize(MaxHeaderSize), LI(LI), TTI(TTI), DT(DT), SE(SE) {
+ const TargetTransformInfo *TTI, AssumptionCache *AC,
+ DominatorTree *DT, ScalarEvolution *SE)
+ : MaxHeaderSize(MaxHeaderSize), LI(LI), TTI(TTI), AC(AC), DT(DT), SE(SE) {
}
bool processLoop(Loop *L);
// duplicate blocks inside it.
{
SmallPtrSet<const Value *, 32> EphValues;
- CodeMetrics::collectEphemeralValues(L, EphValues);
+ CodeMetrics::collectEphemeralValues(L, AC, EphValues);
CodeMetrics Metrics;
Metrics.analyzeBasicBlock(OrigHeader, *TTI, EphValues);
// With the operands remapped, see if the instruction constant folds or is
// otherwise simplifyable. This commonly occurs because the entry from PHI
// nodes allows icmps and other instructions to fold.
- // FIXME: Provide TLI, and DT to SimplifyInstruction.
+ // FIXME: Provide TLI, DT, AC to SimplifyInstruction.
Value *V = SimplifyInstruction(C, DL);
if (V && LI->replacementPreservesLCSSAForm(C, V)) {
// If so, then delete the temporary instruction and stick the folded value
// Otherwise, stick the new instruction into the new block!
C->setName(Inst->getName());
C->insertBefore(LoopEntryBranch);
+
+ if (auto *II = dyn_cast<IntrinsicInst>(C))
+ if (II->getIntrinsicID() == Intrinsic::assume)
+ AC->registerAssumption(II);
}
}
auto *LI = FAM.getCachedResult<LoopAnalysis>(*F);
const auto *TTI = FAM.getCachedResult<TargetIRAnalysis>(*F);
- assert((LI && TTI) && "Analyses for loop rotation not available");
+ auto *AC = FAM.getCachedResult<AssumptionAnalysis>(*F);
+ assert((LI && TTI && AC) && "Analyses for loop rotation not available");
// Optional analyses.
auto *DT = FAM.getCachedResult<DominatorTreeAnalysis>(*F);
auto *SE = FAM.getCachedResult<ScalarEvolutionAnalysis>(*F);
- LoopRotate LR(DefaultRotationThreshold, LI, TTI, DT, SE);
+ LoopRotate LR(DefaultRotationThreshold, LI, TTI, AC, DT, SE);
bool Changed = LR.processLoop(&L);
if (!Changed)
// LCSSA form makes instruction renaming easier.
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<TargetTransformInfoWrapperPass>();
getLoopAnalysisUsage(AU);
}
auto *LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo();
const auto *TTI = &getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F);
+ auto *AC = &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
auto *DTWP = getAnalysisIfAvailable<DominatorTreeWrapperPass>();
auto *DT = DTWP ? &DTWP->getDomTree() : nullptr;
auto *SEWP = getAnalysisIfAvailable<ScalarEvolutionWrapperPass>();
auto *SE = SEWP ? &SEWP->getSE() : nullptr;
- LoopRotate LR(MaxHeaderSize, LI, TTI, DT, SE);
+ LoopRotate LR(MaxHeaderSize, LI, TTI, AC, DT, SE);
return LR.processLoop(L);
}
};
char LoopRotateLegacyPass::ID = 0;
INITIALIZE_PASS_BEGIN(LoopRotateLegacyPass, "loop-rotate", "Rotate Loops",
false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(LoopPass)
INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)
INITIALIZE_PASS_END(LoopRotateLegacyPass, "loop-rotate", "Rotate Loops", false,
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/Analysis/BasicAliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/DependenceAnalysis.h"
#include "llvm/Analysis/GlobalsModRef.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Transforms/Scalar/LoopUnrollPass.h"
#include "llvm/ADT/SetVector.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/CodeMetrics.h"
#include "llvm/Analysis/GlobalsModRef.h"
#include "llvm/Analysis/InstructionSimplify.h"
static unsigned ApproximateLoopSize(const Loop *L, unsigned &NumCalls,
bool &NotDuplicatable, bool &Convergent,
const TargetTransformInfo &TTI,
- unsigned BEInsns) {
+ AssumptionCache *AC, unsigned BEInsns) {
SmallPtrSet<const Value *, 32> EphValues;
- CodeMetrics::collectEphemeralValues(L, EphValues);
+ CodeMetrics::collectEphemeralValues(L, AC, EphValues);
CodeMetrics Metrics;
for (BasicBlock *BB : L->blocks())
static bool tryToUnrollLoop(Loop *L, DominatorTree &DT, LoopInfo *LI,
ScalarEvolution *SE, const TargetTransformInfo &TTI,
- OptimizationRemarkEmitter &ORE,
+ AssumptionCache &AC, OptimizationRemarkEmitter &ORE,
bool PreserveLCSSA,
Optional<unsigned> ProvidedCount,
Optional<unsigned> ProvidedThreshold,
if (UP.Threshold == 0 && (!UP.Partial || UP.PartialThreshold == 0))
return false;
unsigned LoopSize = ApproximateLoopSize(
- L, NumInlineCandidates, NotDuplicatable, Convergent, TTI, UP.BEInsns);
+ L, NumInlineCandidates, NotDuplicatable, Convergent, TTI, &AC, UP.BEInsns);
DEBUG(dbgs() << " Loop Size = " << LoopSize << "\n");
if (NotDuplicatable) {
DEBUG(dbgs() << " Not unrolling loop which contains non-duplicatable"
// Unroll the loop.
if (!UnrollLoop(L, UP.Count, TripCount, UP.Force, UP.Runtime,
UP.AllowExpensiveTripCount, UseUpperBound, MaxOrZero,
- TripMultiple, UP.PeelCount, LI, SE, &DT, &ORE,
+ TripMultiple, UP.PeelCount, LI, SE, &DT, &AC, &ORE,
PreserveLCSSA))
return false;
ScalarEvolution *SE = &getAnalysis<ScalarEvolutionWrapperPass>().getSE();
const TargetTransformInfo &TTI =
getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F);
+ auto &AC = getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
// For the old PM, we can't use OptimizationRemarkEmitter as an analysis
// pass. Function analyses need to be preserved across loop transformations
// but ORE cannot be preserved (see comment before the pass definition).
OptimizationRemarkEmitter ORE(&F);
bool PreserveLCSSA = mustPreserveAnalysisID(LCSSAID);
- return tryToUnrollLoop(L, DT, LI, SE, TTI, ORE, PreserveLCSSA,
+ return tryToUnrollLoop(L, DT, LI, SE, TTI, AC, ORE, PreserveLCSSA,
ProvidedCount, ProvidedThreshold,
ProvidedAllowPartial, ProvidedRuntime,
ProvidedUpperBound);
/// loop preheaders be inserted into the CFG...
///
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<TargetTransformInfoWrapperPass>();
// FIXME: Loop passes are required to preserve domtree, and for now we just
// recreate dom info if anything gets unrolled.
char LoopUnroll::ID = 0;
INITIALIZE_PASS_BEGIN(LoopUnroll, "loop-unroll", "Unroll loops", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(LoopPass)
INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)
INITIALIZE_PASS_END(LoopUnroll, "loop-unroll", "Unroll loops", false, false)
LoopInfo *LI = FAM.getCachedResult<LoopAnalysis>(*F);
ScalarEvolution *SE = FAM.getCachedResult<ScalarEvolutionAnalysis>(*F);
auto *TTI = FAM.getCachedResult<TargetIRAnalysis>(*F);
+ auto *AC = FAM.getCachedResult<AssumptionAnalysis>(*F);
auto *ORE = FAM.getCachedResult<OptimizationRemarkEmitterAnalysis>(*F);
if (!DT)
report_fatal_error(
if (!TTI)
report_fatal_error(
"LoopUnrollPass: TargetIRAnalysis not cached at a higher level");
+ if (!AC)
+ report_fatal_error(
+ "LoopUnrollPass: AssumptionAnalysis not cached at a higher level");
if (!ORE)
report_fatal_error("LoopUnrollPass: OptimizationRemarkEmitterAnalysis not "
"cached at a higher level");
bool Changed =
- tryToUnrollLoop(&L, *DT, LI, SE, *TTI, *ORE, /*PreserveLCSSA*/ true,
+ tryToUnrollLoop(&L, *DT, LI, SE, *TTI, *AC, *ORE, /*PreserveLCSSA*/ true,
ProvidedCount, ProvidedThreshold, ProvidedAllowPartial,
ProvidedRuntime, ProvidedUpperBound);
#include "llvm/Transforms/Scalar.h"
#include "llvm/ADT/STLExtras.h"
-#include "llvm/ADT/SmallSet.h"
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/GlobalsModRef.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/CodeMetrics.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/LoopInfo.h"
// Analyze loop. Check its size, calculate is it possible to unswitch
// it. Returns true if we can unswitch this loop.
- bool countLoop(const Loop *L, const TargetTransformInfo &TTI);
+ bool countLoop(const Loop *L, const TargetTransformInfo &TTI,
+ AssumptionCache *AC);
// Clean all data related to given loop.
void forgetLoop(const Loop *L);
class LoopUnswitch : public LoopPass {
LoopInfo *LI; // Loop information
LPPassManager *LPM;
+ AssumptionCache *AC;
// Used to check if second loop needs processing after
// RewriteLoopBodyWithConditionConstant rewrites first loop.
/// loop preheaders be inserted into the CFG.
///
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<TargetTransformInfoWrapperPass>();
getLoopAnalysisUsage(AU);
}
// Analyze loop. Check its size, calculate is it possible to unswitch
// it. Returns true if we can unswitch this loop.
-bool LUAnalysisCache::countLoop(const Loop *L, const TargetTransformInfo &TTI) {
+bool LUAnalysisCache::countLoop(const Loop *L, const TargetTransformInfo &TTI,
+ AssumptionCache *AC) {
LoopPropsMapIt PropsIt;
bool Inserted;
// This is a very ad-hoc heuristic.
SmallPtrSet<const Value *, 32> EphValues;
- CodeMetrics::collectEphemeralValues(L, EphValues);
+ CodeMetrics::collectEphemeralValues(L, AC, EphValues);
// FIXME: This is overly conservative because it does not take into
// consideration code simplification opportunities and code that can
char LoopUnswitch::ID = 0;
INITIALIZE_PASS_BEGIN(LoopUnswitch, "loop-unswitch", "Unswitch loops",
false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(LoopPass)
INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)
INITIALIZE_PASS_END(LoopUnswitch, "loop-unswitch", "Unswitch loops",
if (skipLoop(L))
return false;
+ AC = &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(
+ *L->getHeader()->getParent());
LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo();
LPM = &LPM_Ref;
DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
// Analyze loop cost, and stop unswitching if loop content can not be duplicated.
if (!BranchesInfo.countLoop(
currentLoop, getAnalysis<TargetTransformInfoWrapperPass>().getTTI(
- *currentLoop->getHeader()->getParent())))
+ *currentLoop->getHeader()->getParent()),
+ AC))
return false;
// Try trivial unswitch first before loop over other basic blocks in the loop.
}
// Rewrite the code to refer to itself.
- for (unsigned i = 0, e = NewBlocks.size(); i != e; ++i)
- for (Instruction &I : *NewBlocks[i])
+ for (unsigned i = 0, e = NewBlocks.size(); i != e; ++i) {
+ for (Instruction &I : *NewBlocks[i]) {
RemapInstruction(&I, VMap,
RF_NoModuleLevelChanges | RF_IgnoreMissingLocals);
+ if (auto *II = dyn_cast<IntrinsicInst>(&I))
+ if (II->getIntrinsicID() == Intrinsic::assume)
+ AC->registerAssumption(II);
+ }
+ }
// Rewrite the original preheader to select between versions of the loop.
BranchInst *OldBR = cast<BranchInst>(loopPreheader->getTerminator());
// This transformation requires dominator postdominator info
void getAnalysisUsage(AnalysisUsage &AU) const override {
AU.setPreservesCFG();
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<DominatorTreeWrapperPass>();
AU.addRequired<MemoryDependenceWrapperPass>();
AU.addRequired<AAResultsWrapperPass>();
INITIALIZE_PASS_BEGIN(MemCpyOptLegacyPass, "memcpyopt", "MemCpy Optimization",
false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(MemoryDependenceWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
// If it is greater than the memcpy, then we check to see if we can force the
// source of the memcpy to the alignment we need. If we fail, we bail out.
+ AssumptionCache &AC = LookupAssumptionCache();
DominatorTree &DT = LookupDomTree();
if (MDep->getAlignment() < ByValAlign &&
getOrEnforceKnownAlignment(MDep->getSource(), ByValAlign, DL,
- CS.getInstruction(), &DT) < ByValAlign)
+ CS.getInstruction(), &AC, &DT) < ByValAlign)
return false;
// Verify that the copied-from memory doesn't change in between the memcpy and
auto LookupAliasAnalysis = [&]() -> AliasAnalysis & {
return AM.getResult<AAManager>(F);
};
+ auto LookupAssumptionCache = [&]() -> AssumptionCache & {
+ return AM.getResult<AssumptionAnalysis>(F);
+ };
auto LookupDomTree = [&]() -> DominatorTree & {
return AM.getResult<DominatorTreeAnalysis>(F);
};
- bool MadeChange = runImpl(F, &MD, &TLI, LookupAliasAnalysis, LookupDomTree);
+ bool MadeChange = runImpl(F, &MD, &TLI, LookupAliasAnalysis,
+ LookupAssumptionCache, LookupDomTree);
if (!MadeChange)
return PreservedAnalyses::all();
PreservedAnalyses PA;
bool MemCpyOptPass::runImpl(
Function &F, MemoryDependenceResults *MD_, TargetLibraryInfo *TLI_,
std::function<AliasAnalysis &()> LookupAliasAnalysis_,
+ std::function<AssumptionCache &()> LookupAssumptionCache_,
std::function<DominatorTree &()> LookupDomTree_) {
bool MadeChange = false;
MD = MD_;
TLI = TLI_;
LookupAliasAnalysis = std::move(LookupAliasAnalysis_);
+ LookupAssumptionCache = std::move(LookupAssumptionCache_);
LookupDomTree = std::move(LookupDomTree_);
// If we don't have at least memset and memcpy, there is little point of doing
auto LookupAliasAnalysis = [this]() -> AliasAnalysis & {
return getAnalysis<AAResultsWrapperPass>().getAAResults();
};
+ auto LookupAssumptionCache = [this, &F]() -> AssumptionCache & {
+ return getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
+ };
auto LookupDomTree = [this]() -> DominatorTree & {
return getAnalysis<DominatorTreeWrapperPass>().getDomTree();
};
- return Impl.runImpl(F, MD, TLI, LookupAliasAnalysis, LookupDomTree);
+ return Impl.runImpl(F, MD, TLI, LookupAliasAnalysis, LookupAssumptionCache,
+ LookupDomTree);
}
AU.addPreserved<DominatorTreeWrapperPass>();
AU.addPreserved<ScalarEvolutionWrapperPass>();
AU.addPreserved<TargetLibraryInfoWrapperPass>();
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<DominatorTreeWrapperPass>();
AU.addRequired<ScalarEvolutionWrapperPass>();
AU.addRequired<TargetLibraryInfoWrapperPass>();
char NaryReassociateLegacyPass::ID = 0;
INITIALIZE_PASS_BEGIN(NaryReassociateLegacyPass, "nary-reassociate",
"Nary reassociation", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(ScalarEvolutionWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
if (skipFunction(F))
return false;
+ auto *AC = &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
auto *DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
auto *SE = &getAnalysis<ScalarEvolutionWrapperPass>().getSE();
auto *TLI = &getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
auto *TTI = &getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F);
- return Impl.runImpl(F, DT, SE, TLI, TTI);
+ return Impl.runImpl(F, AC, DT, SE, TLI, TTI);
}
PreservedAnalyses NaryReassociatePass::run(Function &F,
FunctionAnalysisManager &AM) {
+ auto *AC = &AM.getResult<AssumptionAnalysis>(F);
auto *DT = &AM.getResult<DominatorTreeAnalysis>(F);
auto *SE = &AM.getResult<ScalarEvolutionAnalysis>(F);
auto *TLI = &AM.getResult<TargetLibraryAnalysis>(F);
auto *TTI = &AM.getResult<TargetIRAnalysis>(F);
- bool Changed = runImpl(F, DT, SE, TLI, TTI);
+ bool Changed = runImpl(F, AC, DT, SE, TLI, TTI);
// FIXME: We need to invalidate this to avoid PR28400. Is there a better
// solution?
return PA;
}
-bool NaryReassociatePass::runImpl(Function &F, DominatorTree *DT_,
- ScalarEvolution *SE_,
+bool NaryReassociatePass::runImpl(Function &F, AssumptionCache *AC_,
+ DominatorTree *DT_, ScalarEvolution *SE_,
TargetLibraryInfo *TLI_,
TargetTransformInfo *TTI_) {
+ AC = AC_;
DT = DT_;
SE = SE_;
TLI = TLI_;
IndexToSplit = SExt->getOperand(0);
} else if (ZExtInst *ZExt = dyn_cast<ZExtInst>(IndexToSplit)) {
// zext can be treated as sext if the source is non-negative.
- if (isKnownNonNegative(ZExt->getOperand(0), *DL, 0, GEP, DT))
+ if (isKnownNonNegative(ZExt->getOperand(0), *DL, 0, AC, GEP, DT))
IndexToSplit = ZExt->getOperand(0);
}
// nsw, we cannot split the add because
// sext(LHS + RHS) != sext(LHS) + sext(RHS).
if (requiresSignExtension(IndexToSplit, GEP) &&
- computeOverflowForSignedAdd(AO, *DL, GEP, DT) !=
+ computeOverflowForSignedAdd(AO, *DL, AC, GEP, DT) !=
OverflowResult::NeverOverflows)
return nullptr;
IndexExprs.push_back(SE->getSCEV(*Index));
// Replace the I-th index with LHS.
IndexExprs[I] = SE->getSCEV(LHS);
- if (isKnownNonNegative(LHS, *DL, 0, GEP, DT) &&
+ if (isKnownNonNegative(LHS, *DL, 0, AC, GEP, DT) &&
DL->getTypeSizeInBits(LHS->getType()) <
DL->getTypeSizeInBits(GEP->getOperand(I)->getType())) {
// Zero-extend LHS if it is non-negative. InstCombine canonicalizes sext to
#include "llvm/ADT/Statistic.h"
#include "llvm/ADT/DenseSet.h"
#include "llvm/ADT/SetVector.h"
-#include "llvm/ADT/SmallSet.h"
#include "llvm/ADT/StringRef.h"
#include "llvm/ADT/MapVector.h"
#include "llvm/IR/BasicBlock.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/Statistic.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/GlobalsModRef.h"
#include "llvm/Analysis/Loads.h"
#include "llvm/Analysis/PtrUseVisitor.h"
NumPromoted += PromotableAllocas.size();
DEBUG(dbgs() << "Promoting allocas with mem2reg...\n");
- PromoteMemToReg(PromotableAllocas, *DT, nullptr);
+ PromoteMemToReg(PromotableAllocas, *DT, nullptr, AC);
PromotableAllocas.clear();
return true;
}
-PreservedAnalyses SROA::runImpl(Function &F, DominatorTree &RunDT) {
+PreservedAnalyses SROA::runImpl(Function &F, DominatorTree &RunDT,
+ AssumptionCache &RunAC) {
DEBUG(dbgs() << "SROA function: " << F.getName() << "\n");
C = &F.getContext();
DT = &RunDT;
+ AC = &RunAC;
BasicBlock &EntryBB = F.getEntryBlock();
for (BasicBlock::iterator I = EntryBB.begin(), E = std::prev(EntryBB.end());
}
PreservedAnalyses SROA::run(Function &F, FunctionAnalysisManager &AM) {
- return runImpl(F, AM.getResult<DominatorTreeAnalysis>(F));
+ return runImpl(F, AM.getResult<DominatorTreeAnalysis>(F),
+ AM.getResult<AssumptionAnalysis>(F));
}
/// A legacy pass for the legacy pass manager that wraps the \c SROA pass.
return false;
auto PA = Impl.runImpl(
- F, getAnalysis<DominatorTreeWrapperPass>().getDomTree());
+ F, getAnalysis<DominatorTreeWrapperPass>().getDomTree(),
+ getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F));
return !PA.areAllPreserved();
}
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<DominatorTreeWrapperPass>();
AU.addPreserved<GlobalsAAWrapperPass>();
AU.setPreservesCFG();
INITIALIZE_PASS_BEGIN(SROALegacyPass, "sroa",
"Scalar Replacement Of Aggregates", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_END(SROALegacyPass, "sroa", "Scalar Replacement Of Aggregates",
false, false)
// Do not trace into "or" unless it is equivalent to "add". If LHS and RHS
// don't have common bits, (LHS | RHS) is equivalent to (LHS + RHS).
if (BO->getOpcode() == Instruction::Or &&
- !haveNoCommonBitsSet(LHS, RHS, DL, BO, DT))
+ !haveNoCommonBitsSet(LHS, RHS, DL, nullptr, BO, DT))
return false;
// In addition, tracing into BO requires that its surrounding s/zext (if
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/Statistic.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/CFG.h"
#include "llvm/Analysis/GlobalsModRef.h"
#include "llvm/Analysis/TargetTransformInfo.h"
/// Call SimplifyCFG on all the blocks in the function,
/// iterating until no more changes are made.
static bool iterativelySimplifyCFG(Function &F, const TargetTransformInfo &TTI,
+ AssumptionCache *AC,
unsigned BonusInstThreshold) {
bool Changed = false;
bool LocalChange = true;
// Loop over all of the basic blocks and remove them if they are unneeded.
for (Function::iterator BBIt = F.begin(); BBIt != F.end(); ) {
- if (SimplifyCFG(&*BBIt++, TTI, BonusInstThreshold, &LoopHeaders)) {
+ if (SimplifyCFG(&*BBIt++, TTI, BonusInstThreshold, AC, &LoopHeaders)) {
LocalChange = true;
++NumSimpl;
}
}
static bool simplifyFunctionCFG(Function &F, const TargetTransformInfo &TTI,
- int BonusInstThreshold) {
+ AssumptionCache *AC, int BonusInstThreshold) {
bool EverChanged = removeUnreachableBlocks(F);
EverChanged |= mergeEmptyReturnBlocks(F);
- EverChanged |= iterativelySimplifyCFG(F, TTI, BonusInstThreshold);
+ EverChanged |= iterativelySimplifyCFG(F, TTI, AC, BonusInstThreshold);
// If neither pass changed anything, we're done.
if (!EverChanged) return false;
return true;
do {
- EverChanged = iterativelySimplifyCFG(F, TTI, BonusInstThreshold);
+ EverChanged = iterativelySimplifyCFG(F, TTI, AC, BonusInstThreshold);
EverChanged |= removeUnreachableBlocks(F);
} while (EverChanged);
PreservedAnalyses SimplifyCFGPass::run(Function &F,
FunctionAnalysisManager &AM) {
auto &TTI = AM.getResult<TargetIRAnalysis>(F);
+ auto &AC = AM.getResult<AssumptionAnalysis>(F);
- if (!simplifyFunctionCFG(F, TTI, BonusInstThreshold))
+ if (!simplifyFunctionCFG(F, TTI, &AC, BonusInstThreshold))
return PreservedAnalyses::all();
PreservedAnalyses PA;
PA.preserve<GlobalsAA>();
if (skipFunction(F) || (PredicateFtor && !PredicateFtor(F)))
return false;
+ AssumptionCache *AC =
+ &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
const TargetTransformInfo &TTI =
getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F);
- return simplifyFunctionCFG(F, TTI, BonusInstThreshold);
+ return simplifyFunctionCFG(F, TTI, AC, BonusInstThreshold);
}
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<TargetTransformInfoWrapperPass>();
AU.addPreserved<GlobalsAAWrapperPass>();
}
INITIALIZE_PASS_BEGIN(CFGSimplifyPass, "simplifycfg", "Simplify the CFG", false,
false)
INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_END(CFGSimplifyPass, "simplifycfg", "Simplify the CFG", false,
false)
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/StringExtras.h"
#include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/CallGraph.h"
#include "llvm/Analysis/CaptureTracking.h"
#include "llvm/Analysis/EHPersonalities.h"
/// If the inlined function has non-byval align arguments, then
/// add @llvm.assume-based alignment assumptions to preserve this information.
static void AddAlignmentAssumptions(CallSite CS, InlineFunctionInfo &IFI) {
- if (!PreserveAlignmentAssumptions)
+ if (!PreserveAlignmentAssumptions || !IFI.GetAssumptionCache)
return;
+ AssumptionCache *AC = IFI.GetAssumptionCache
+ ? &(*IFI.GetAssumptionCache)(*CS.getCaller())
+ : nullptr;
auto &DL = CS.getCaller()->getParent()->getDataLayout();
// To avoid inserting redundant assumptions, we should check for assumptions
// If we can already prove the asserted alignment in the context of the
// caller, then don't bother inserting the assumption.
Value *Arg = CS.getArgument(I->getArgNo());
- if (getKnownAlignment(Arg, DL, CS.getInstruction(), &DT) >= Align)
+ if (getKnownAlignment(Arg, DL, CS.getInstruction(), AC, &DT) >= Align)
continue;
- IRBuilder<>(CS.getInstruction())
- .CreateAlignmentAssumption(DL, Arg, Align);
+ CallInst *NewAssumption = IRBuilder<>(CS.getInstruction())
+ .CreateAlignmentAssumption(DL, Arg, Align);
+ if (AC)
+ AC->registerAssumption(NewAssumption);
}
}
}
if (ByValAlignment <= 1) // 0 = unspecified, 1 = no particular alignment.
return Arg;
+ AssumptionCache *AC =
+ IFI.GetAssumptionCache ? &(*IFI.GetAssumptionCache)(*Caller) : nullptr;
const DataLayout &DL = Caller->getParent()->getDataLayout();
// If the pointer is already known to be sufficiently aligned, or if we can
// round it up to a larger alignment, then we don't need a temporary.
- if (getOrEnforceKnownAlignment(Arg, ByValAlignment, DL, TheCall) >=
+ if (getOrEnforceKnownAlignment(Arg, ByValAlignment, DL, TheCall, AC) >=
ByValAlignment)
return Arg;
// Propagate llvm.mem.parallel_loop_access if necessary.
PropagateParallelLoopAccessMetadata(CS, VMap);
+
+ // Register any cloned assumptions.
+ if (IFI.GetAssumptionCache)
+ for (BasicBlock &NewBlock :
+ make_range(FirstNewBlock->getIterator(), Caller->end()))
+ for (Instruction &I : NewBlock) {
+ if (auto *II = dyn_cast<IntrinsicInst>(&I))
+ if (II->getIntrinsicID() == Intrinsic::assume)
+ (*IFI.GetAssumptionCache)(*Caller).registerAssumption(II);
+ }
}
// If there are any alloca instructions in the block that used to be the entry
// the entries are the same or undef). If so, remove the PHI so it doesn't
// block other optimizations.
if (PHI) {
+ AssumptionCache *AC =
+ IFI.GetAssumptionCache ? &(*IFI.GetAssumptionCache)(*Caller) : nullptr;
auto &DL = Caller->getParent()->getDataLayout();
- if (Value *V = SimplifyInstruction(PHI, DL, nullptr, nullptr)) {
+ if (Value *V = SimplifyInstruction(PHI, DL, nullptr, nullptr, AC)) {
PHI->replaceAllUsesWith(V);
PHI->eraseFromParent();
}
unsigned llvm::getOrEnforceKnownAlignment(Value *V, unsigned PrefAlign,
const DataLayout &DL,
const Instruction *CxtI,
+ AssumptionCache *AC,
const DominatorTree *DT) {
assert(V->getType()->isPointerTy() &&
"getOrEnforceKnownAlignment expects a pointer!");
unsigned BitWidth = DL.getPointerTypeSizeInBits(V->getType());
APInt KnownZero(BitWidth, 0), KnownOne(BitWidth, 0);
- computeKnownBits(V, KnownZero, KnownOne, DL, 0, CxtI, DT);
+ computeKnownBits(V, KnownZero, KnownOne, DL, 0, AC, CxtI, DT);
unsigned TrailZ = KnownZero.countTrailingOnes();
// Avoid trouble with ridiculously large TrailZ values, such as
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/Analysis/BasicAliasAnalysis.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/DependenceAnalysis.h"
#include "llvm/Analysis/GlobalsModRef.h"
#include "llvm/Analysis/InstructionSimplify.h"
/// \brief The first part of loop-nestification is to find a PHI node that tells
/// us how to partition the loops.
-static PHINode *findPHIToPartitionLoops(Loop *L, DominatorTree *DT) {
+static PHINode *findPHIToPartitionLoops(Loop *L, DominatorTree *DT,
+ AssumptionCache *AC) {
const DataLayout &DL = L->getHeader()->getModule()->getDataLayout();
for (BasicBlock::iterator I = L->getHeader()->begin(); isa<PHINode>(I); ) {
PHINode *PN = cast<PHINode>(I);
++I;
- if (Value *V = SimplifyInstruction(PN, DL, nullptr, DT)) {
+ if (Value *V = SimplifyInstruction(PN, DL, nullptr, DT, AC)) {
// This is a degenerate PHI already, don't modify it!
PN->replaceAllUsesWith(V);
PN->eraseFromParent();
///
static Loop *separateNestedLoop(Loop *L, BasicBlock *Preheader,
DominatorTree *DT, LoopInfo *LI,
- ScalarEvolution *SE, bool PreserveLCSSA) {
+ ScalarEvolution *SE, bool PreserveLCSSA,
+ AssumptionCache *AC) {
// Don't try to separate loops without a preheader.
if (!Preheader)
return nullptr;
BasicBlock *Header = L->getHeader();
assert(!Header->isEHPad() && "Can't insert backedge to EH pad");
- PHINode *PN = findPHIToPartitionLoops(L, DT);
+ PHINode *PN = findPHIToPartitionLoops(L, DT, AC);
if (!PN) return nullptr; // No known way to partition.
// Pull out all predecessors that have varying values in the loop. This
/// \brief Simplify one loop and queue further loops for simplification.
static bool simplifyOneLoop(Loop *L, SmallVectorImpl<Loop *> &Worklist,
DominatorTree *DT, LoopInfo *LI,
- ScalarEvolution *SE, bool PreserveLCSSA) {
+ ScalarEvolution *SE, AssumptionCache *AC,
+ bool PreserveLCSSA) {
bool Changed = false;
ReprocessLoop:
// common backedge instead.
if (L->getNumBackEdges() < 8) {
if (Loop *OuterL =
- separateNestedLoop(L, Preheader, DT, LI, SE, PreserveLCSSA)) {
+ separateNestedLoop(L, Preheader, DT, LI, SE, PreserveLCSSA, AC)) {
++NumNested;
// Enqueue the outer loop as it should be processed next in our
// depth-first nest walk.
PHINode *PN;
for (BasicBlock::iterator I = L->getHeader()->begin();
(PN = dyn_cast<PHINode>(I++)); )
- if (Value *V = SimplifyInstruction(PN, DL, nullptr, DT)) {
+ if (Value *V = SimplifyInstruction(PN, DL, nullptr, DT, AC)) {
if (SE) SE->forgetValue(PN);
if (!PreserveLCSSA || LI->replacementPreservesLCSSAForm(PN, V)) {
PN->replaceAllUsesWith(V);
}
bool llvm::simplifyLoop(Loop *L, DominatorTree *DT, LoopInfo *LI,
- ScalarEvolution *SE, bool PreserveLCSSA) {
+ ScalarEvolution *SE, AssumptionCache *AC,
+ bool PreserveLCSSA) {
bool Changed = false;
// Worklist maintains our depth-first queue of loops in this nest to process.
while (!Worklist.empty())
Changed |= simplifyOneLoop(Worklist.pop_back_val(), Worklist, DT, LI, SE,
- PreserveLCSSA);
+ AC, PreserveLCSSA);
return Changed;
}
bool runOnFunction(Function &F) override;
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
+
// We need loop information to identify the loops...
AU.addRequired<DominatorTreeWrapperPass>();
AU.addPreserved<DominatorTreeWrapperPass>();
char LoopSimplify::ID = 0;
INITIALIZE_PASS_BEGIN(LoopSimplify, "loop-simplify",
"Canonicalize natural loops", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(LoopInfoWrapperPass)
INITIALIZE_PASS_END(LoopSimplify, "loop-simplify",
DominatorTree *DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
auto *SEWP = getAnalysisIfAvailable<ScalarEvolutionWrapperPass>();
ScalarEvolution *SE = SEWP ? &SEWP->getSE() : nullptr;
+ AssumptionCache *AC =
+ &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
bool PreserveLCSSA = mustPreserveAnalysisID(LCSSAID);
#ifndef NDEBUG
// Simplify each loop nest in the function.
for (LoopInfo::iterator I = LI->begin(), E = LI->end(); I != E; ++I)
- Changed |= simplifyLoop(*I, DT, LI, SE, PreserveLCSSA);
+ Changed |= simplifyLoop(*I, DT, LI, SE, AC, PreserveLCSSA);
#ifndef NDEBUG
if (PreserveLCSSA) {
LoopInfo *LI = &AM.getResult<LoopAnalysis>(F);
DominatorTree *DT = &AM.getResult<DominatorTreeAnalysis>(F);
ScalarEvolution *SE = AM.getCachedResult<ScalarEvolutionAnalysis>(F);
+ AssumptionCache *AC = &AM.getResult<AssumptionAnalysis>(F);
// FIXME: This pass should verify that the loops on which it's operating
// are in canonical SSA form, and that the pass itself preserves this form.
for (LoopInfo::iterator I = LI->begin(), E = LI->end(); I != E; ++I)
- Changed |= simplifyLoop(*I, DT, LI, SE, true /* PreserveLCSSA */);
+ Changed |= simplifyLoop(*I, DT, LI, SE, AC, true /* PreserveLCSSA */);
// FIXME: We need to invalidate this to avoid PR28400. Is there a better
// solution?
#include "llvm/Transforms/Utils/UnrollLoop.h"
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/Statistic.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/LoopIterator.h"
#include "llvm/Analysis/LoopPass.h"
bool PreserveCondBr, bool PreserveOnlyFirst,
unsigned TripMultiple, unsigned PeelCount, LoopInfo *LI,
ScalarEvolution *SE, DominatorTree *DT,
- OptimizationRemarkEmitter *ORE, bool PreserveLCSSA) {
+ AssumptionCache *AC, OptimizationRemarkEmitter *ORE,
+ bool PreserveLCSSA) {
BasicBlock *Preheader = L->getLoopPreheader();
if (!Preheader) {
}
// Remap all instructions in the most recent iteration
- for (BasicBlock *NewBlock : NewBlocks)
- for (Instruction &I : *NewBlock)
+ for (BasicBlock *NewBlock : NewBlocks) {
+ for (Instruction &I : *NewBlock) {
::remapInstruction(&I, LastValueMap);
+ if (auto *II = dyn_cast<IntrinsicInst>(&I))
+ if (II->getIntrinsicID() == Intrinsic::assume)
+ AC->registerAssumption(II);
+ }
+ }
}
// Loop over the PHI nodes in the original block, setting incoming values.
// loops too).
// TODO: That potentially might be compile-time expensive. We should try
// to fix the loop-simplified form incrementally.
- simplifyLoop(OuterL, DT, LI, SE, PreserveLCSSA);
+ simplifyLoop(OuterL, DT, LI, SE, AC, PreserveLCSSA);
// LCSSA must be performed on the outermost affected loop. The unrolled
// loop's last loop latch is guaranteed to be in the outermost loop after
} else {
// Simplify loops for which we might've broken loop-simplify form.
for (Loop *SubLoop : LoopsToSimplify)
- simplifyLoop(SubLoop, DT, LI, SE, PreserveLCSSA);
+ simplifyLoop(SubLoop, DT, LI, SE, AC, PreserveLCSSA);
}
}
#include "llvm/Transforms/Utils/Mem2Reg.h"
#include "llvm/ADT/Statistic.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/IR/Dominators.h"
#include "llvm/IR/Function.h"
#include "llvm/IR/Instructions.h"
STATISTIC(NumPromoted, "Number of alloca's promoted");
-static bool promoteMemoryToRegister(Function &F, DominatorTree &DT) {
+static bool promoteMemoryToRegister(Function &F, DominatorTree &DT,
+ AssumptionCache &AC) {
std::vector<AllocaInst *> Allocas;
BasicBlock &BB = F.getEntryBlock(); // Get the entry node for the function
bool Changed = false;
if (Allocas.empty())
break;
- PromoteMemToReg(Allocas, DT, nullptr);
+ PromoteMemToReg(Allocas, DT, nullptr, &AC);
NumPromoted += Allocas.size();
Changed = true;
}
PreservedAnalyses PromotePass::run(Function &F, FunctionAnalysisManager &AM) {
auto &DT = AM.getResult<DominatorTreeAnalysis>(F);
- if (!promoteMemoryToRegister(F, DT))
+ auto &AC = AM.getResult<AssumptionAnalysis>(F);
+ if (!promoteMemoryToRegister(F, DT, AC))
return PreservedAnalyses::all();
// FIXME: This should also 'preserve the CFG'.
return false;
DominatorTree &DT = getAnalysis<DominatorTreeWrapperPass>().getDomTree();
- return promoteMemoryToRegister(F, DT);
+ AssumptionCache &AC =
+ getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
+ return promoteMemoryToRegister(F, DT, AC);
}
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<DominatorTreeWrapperPass>();
AU.setPreservesCFG();
}
INITIALIZE_PASS_BEGIN(PromoteLegacyPass, "mem2reg", "Promote Memory to "
"Register",
false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_END(PromoteLegacyPass, "mem2reg", "Promote Memory to Register",
false, false)
/// An AliasSetTracker object to update. If null, don't update it.
AliasSetTracker *AST;
+ /// A cache of @llvm.assume intrinsics used by SimplifyInstruction.
+ AssumptionCache *AC;
+
/// Reverse mapping of Allocas.
DenseMap<AllocaInst *, unsigned> AllocaLookup;
public:
PromoteMem2Reg(ArrayRef<AllocaInst *> Allocas, DominatorTree &DT,
- AliasSetTracker *AST)
+ AliasSetTracker *AST, AssumptionCache *AC)
: Allocas(Allocas.begin(), Allocas.end()), DT(DT),
DIB(*DT.getRoot()->getParent()->getParent(), /*AllowUnresolved*/ false),
- AST(AST) {}
+ AST(AST), AC(AC) {}
void run();
PHINode *PN = I->second;
// If this PHI node merges one value and/or undefs, get the value.
- if (Value *V = SimplifyInstruction(PN, DL, nullptr, &DT)) {
+ if (Value *V = SimplifyInstruction(PN, DL, nullptr, &DT, AC)) {
if (AST && PN->getType()->isPointerTy())
AST->deleteValue(PN);
PN->replaceAllUsesWith(V);
}
void llvm::PromoteMemToReg(ArrayRef<AllocaInst *> Allocas, DominatorTree &DT,
- AliasSetTracker *AST) {
+ AliasSetTracker *AST, AssumptionCache *AC) {
// If there is nothing to do, bail out...
if (Allocas.empty())
return;
- PromoteMem2Reg(Allocas, DT, AST).run();
+ PromoteMem2Reg(Allocas, DT, AST, AC).run();
}
const TargetTransformInfo &TTI;
const DataLayout &DL;
unsigned BonusInstThreshold;
+ AssumptionCache *AC;
SmallPtrSetImpl<BasicBlock *> *LoopHeaders;
Value *isValueEqualityComparison(TerminatorInst *TI);
BasicBlock *GetValueEqualityComparisonCases(
public:
SimplifyCFGOpt(const TargetTransformInfo &TTI, const DataLayout &DL,
- unsigned BonusInstThreshold,
+ unsigned BonusInstThreshold, AssumptionCache *AC,
SmallPtrSetImpl<BasicBlock *> *LoopHeaders)
- : TTI(TTI), DL(DL), BonusInstThreshold(BonusInstThreshold),
+ : TTI(TTI), DL(DL), BonusInstThreshold(BonusInstThreshold), AC(AC),
LoopHeaders(LoopHeaders) {}
bool run(BasicBlock *BB);
/// the PHI, merging the third icmp into the switch.
static bool TryToSimplifyUncondBranchWithICmpInIt(
ICmpInst *ICI, IRBuilder<> &Builder, const DataLayout &DL,
- const TargetTransformInfo &TTI, unsigned BonusInstThreshold) {
+ const TargetTransformInfo &TTI, unsigned BonusInstThreshold,
+ AssumptionCache *AC) {
BasicBlock *BB = ICI->getParent();
// If the block has any PHIs in it or the icmp has multiple uses, it is too
ICI->eraseFromParent();
}
// BB is now empty, so it is likely to simplify away.
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
}
// Ok, the block is reachable from the default dest. If the constant we're
ICI->replaceAllUsesWith(V);
ICI->eraseFromParent();
// BB is now empty, so it is likely to simplify away.
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
}
// The use of the icmp has to be in the 'end' block, by the only PHI node in
/// Compute masked bits for the condition of a switch
/// and use it to remove dead cases.
-static bool EliminateDeadSwitchCases(SwitchInst *SI, const DataLayout &DL) {
+static bool EliminateDeadSwitchCases(SwitchInst *SI, AssumptionCache *AC,
+ const DataLayout &DL) {
Value *Cond = SI->getCondition();
unsigned Bits = Cond->getType()->getIntegerBitWidth();
APInt KnownZero(Bits, 0), KnownOne(Bits, 0);
- computeKnownBits(Cond, KnownZero, KnownOne, DL, 0, SI);
+ computeKnownBits(Cond, KnownZero, KnownOne, DL, 0, AC, SI);
// We can also eliminate cases by determining that their values are outside of
// the limited range of the condition based on how many significant (non-sign)
// bits are in the condition value.
- unsigned ExtraSignBits = ComputeNumSignBits(Cond, DL, 0, SI) - 1;
+ unsigned ExtraSignBits = ComputeNumSignBits(Cond, DL, 0, AC, SI) - 1;
unsigned MaxSignificantBitsInCond = Bits - ExtraSignBits;
// Gather dead cases.
/// phi nodes in a common successor block with only two different
/// constant values, replace the switch with select.
static bool SwitchToSelect(SwitchInst *SI, IRBuilder<> &Builder,
- const DataLayout &DL,
+ AssumptionCache *AC, const DataLayout &DL,
const TargetTransformInfo &TTI) {
Value *const Cond = SI->getCondition();
PHINode *PHI = nullptr;
// see if that predecessor totally determines the outcome of this switch.
if (BasicBlock *OnlyPred = BB->getSinglePredecessor())
if (SimplifyEqualityComparisonWithOnlyPredecessor(SI, OnlyPred, Builder))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
Value *Cond = SI->getCondition();
if (SelectInst *Select = dyn_cast<SelectInst>(Cond))
if (SimplifySwitchOnSelect(SI, Select))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
// If the block only contains the switch, see if we can fold the block
// away into any preds.
++BBI;
if (SI == &*BBI)
if (FoldValueComparisonIntoPredecessors(SI, Builder))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
}
// Try to transform the switch into an icmp and a branch.
if (TurnSwitchRangeIntoICmp(SI, Builder))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
// Remove unreachable cases.
- if (EliminateDeadSwitchCases(SI, DL))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ if (EliminateDeadSwitchCases(SI, AC, DL))
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
- if (SwitchToSelect(SI, Builder, DL, TTI))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ if (SwitchToSelect(SI, Builder, AC, DL, TTI))
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
if (ForwardSwitchConditionToPHI(SI))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
if (SwitchToLookupTable(SI, Builder, DL, TTI))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
if (ReduceSwitchRange(SI, Builder, DL, TTI))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
return false;
}
if (SelectInst *SI = dyn_cast<SelectInst>(IBI->getAddress())) {
if (SimplifyIndirectBrOnSelect(IBI, SI))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
}
return Changed;
}
;
if (I->isTerminator() &&
TryToSimplifyUncondBranchWithICmpInIt(ICI, Builder, DL, TTI,
- BonusInstThreshold))
+ BonusInstThreshold, AC))
return true;
}
// predecessor and use logical operations to update the incoming value
// for PHI nodes in common successor.
if (FoldBranchToCommonDest(BI, BonusInstThreshold))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
return false;
}
// switch.
if (BasicBlock *OnlyPred = BB->getSinglePredecessor())
if (SimplifyEqualityComparisonWithOnlyPredecessor(BI, OnlyPred, Builder))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
// This block must be empty, except for the setcond inst, if it exists.
// Ignore dbg intrinsics.
++I;
if (&*I == BI) {
if (FoldValueComparisonIntoPredecessors(BI, Builder))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
} else if (&*I == cast<Instruction>(BI->getCondition())) {
++I;
// Ignore dbg intrinsics.
while (isa<DbgInfoIntrinsic>(I))
++I;
if (&*I == BI && FoldValueComparisonIntoPredecessors(BI, Builder))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
}
}
: ConstantInt::getFalse(BB->getContext());
BI->setCondition(CI);
RecursivelyDeleteTriviallyDeadInstructions(OldCond);
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
}
}
}
// branches to us and one of our successors, fold the comparison into the
// predecessor and use logical operations to pick the right destination.
if (FoldBranchToCommonDest(BI, BonusInstThreshold))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
// We have a conditional branch to two blocks that are only reachable
// from BI. We know that the condbr dominates the two blocks, so see if
if (BI->getSuccessor(0)->getSinglePredecessor()) {
if (BI->getSuccessor(1)->getSinglePredecessor()) {
if (HoistThenElseCodeToIf(BI, TTI))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
} else {
// If Successor #1 has multiple preds, we may be able to conditionally
// execute Successor #0 if it branches to Successor #1.
if (Succ0TI->getNumSuccessors() == 1 &&
Succ0TI->getSuccessor(0) == BI->getSuccessor(1))
if (SpeculativelyExecuteBB(BI, BI->getSuccessor(0), TTI))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
}
} else if (BI->getSuccessor(1)->getSinglePredecessor()) {
// If Successor #0 has multiple preds, we may be able to conditionally
if (Succ1TI->getNumSuccessors() == 1 &&
Succ1TI->getSuccessor(0) == BI->getSuccessor(0))
if (SpeculativelyExecuteBB(BI, BI->getSuccessor(1), TTI))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
}
// If this is a branch on a phi node in the current block, thread control
if (PHINode *PN = dyn_cast<PHINode>(BI->getCondition()))
if (PN->getParent() == BI->getParent())
if (FoldCondBranchOnPHI(BI, DL))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
// Scan predecessor blocks for conditional branches.
for (pred_iterator PI = pred_begin(BB), E = pred_end(BB); PI != E; ++PI)
if (BranchInst *PBI = dyn_cast<BranchInst>((*PI)->getTerminator()))
if (PBI != BI && PBI->isConditional())
if (SimplifyCondBranchToCondBranch(PBI, BI, DL))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
// Look for diamond patterns.
if (MergeCondStores)
if (BranchInst *PBI = dyn_cast<BranchInst>(PrevBB->getTerminator()))
if (PBI != BI && PBI->isConditional())
if (mergeConditionalStores(PBI, BI))
- return SimplifyCFG(BB, TTI, BonusInstThreshold) | true;
+ return SimplifyCFG(BB, TTI, BonusInstThreshold, AC) | true;
return false;
}
/// of the CFG. It returns true if a modification was made.
///
bool llvm::SimplifyCFG(BasicBlock *BB, const TargetTransformInfo &TTI,
- unsigned BonusInstThreshold,
+ unsigned BonusInstThreshold, AssumptionCache *AC,
SmallPtrSetImpl<BasicBlock *> *LoopHeaders) {
return SimplifyCFGOpt(TTI, BB->getModule()->getDataLayout(),
- BonusInstThreshold, LoopHeaders)
+ BonusInstThreshold, AC, LoopHeaders)
.run(BB);
}
#include "llvm/ADT/DepthFirstIterator.h"
#include "llvm/ADT/SmallPtrSet.h"
#include "llvm/ADT/Statistic.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/InstructionSimplify.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/IR/DataLayout.h"
STATISTIC(NumSimplified, "Number of redundant instructions removed");
static bool runImpl(Function &F, const DominatorTree *DT,
- const TargetLibraryInfo *TLI) {
+ const TargetLibraryInfo *TLI, AssumptionCache *AC) {
const DataLayout &DL = F.getParent()->getDataLayout();
SmallPtrSet<const Instruction *, 8> S1, S2, *ToSimplify = &S1, *Next = &S2;
bool Changed = false;
// Don't waste time simplifying unused instructions.
if (!I->use_empty()) {
- if (Value *V = SimplifyInstruction(I, DL, TLI, DT)) {
+ if (Value *V = SimplifyInstruction(I, DL, TLI, DT, AC)) {
// Mark all uses for resimplification next time round the loop.
for (User *U : I->users())
Next->insert(cast<Instruction>(U));
void getAnalysisUsage(AnalysisUsage &AU) const override {
AU.setPreservesCFG();
AU.addRequired<DominatorTreeWrapperPass>();
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<TargetLibraryInfoWrapperPass>();
}
&getAnalysis<DominatorTreeWrapperPass>().getDomTree();
const TargetLibraryInfo *TLI =
&getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
- return runImpl(F, DT, TLI);
+ AssumptionCache *AC =
+ &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
+ return runImpl(F, DT, TLI, AC);
}
};
}
char InstSimplifier::ID = 0;
INITIALIZE_PASS_BEGIN(InstSimplifier, "instsimplify",
"Remove redundant instructions", false, false)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass)
INITIALIZE_PASS_END(InstSimplifier, "instsimplify",
FunctionAnalysisManager &AM) {
auto &DT = AM.getResult<DominatorTreeAnalysis>(F);
auto &TLI = AM.getResult<TargetLibraryAnalysis>(F);
- bool Changed = runImpl(F, &DT, &TLI);
+ auto &AC = AM.getResult<AssumptionAnalysis>(F);
+ bool Changed = runImpl(F, &DT, &TLI, &AC);
if (!Changed)
return PreservedAnalyses::all();
// FIXME: This should also 'preserve the CFG'.
unsigned BitWidth = Offset->getType()->getIntegerBitWidth();
APInt KnownZero(BitWidth, 0);
APInt KnownOne(BitWidth, 0);
- computeKnownBits(Offset, KnownZero, KnownOne, DL, 0, CI, nullptr);
+ computeKnownBits(Offset, KnownZero, KnownOne, DL, 0, nullptr, CI,
+ nullptr);
KnownZero.flipAllBits();
size_t ArrSize =
cast<ArrayType>(GEP->getSourceElementType())->getNumElements();
if (!Safe) {
APInt KnownZero(BitWidth, 0);
APInt KnownOne(BitWidth, 0);
- computeKnownBits(OpA, KnownZero, KnownOne, DL, 0, OpA, &DT);
+ computeKnownBits(OpA, KnownZero, KnownOne, DL, 0, nullptr, OpA, &DT);
KnownZero &= ~APInt::getHighBitsSet(BitWidth, 1);
if (KnownZero != 0)
Safe = true;
unsigned NewAlign = getOrEnforceKnownAlignment(S0->getPointerOperand(),
StackAdjustedAlignment,
- DL, S0, &DT);
+ DL, S0, nullptr, &DT);
if (NewAlign < StackAdjustedAlignment)
return false;
}
unsigned NewAlign = getOrEnforceKnownAlignment(L0->getPointerOperand(),
StackAdjustedAlignment,
- DL, L0, &DT);
+ DL, L0, nullptr, &DT);
if (NewAlign < StackAdjustedAlignment)
return false;
InnerLoopVectorizer(Loop *OrigLoop, PredicatedScalarEvolution &PSE,
LoopInfo *LI, DominatorTree *DT,
const TargetLibraryInfo *TLI,
- const TargetTransformInfo *TTI,
+ const TargetTransformInfo *TTI, AssumptionCache *AC,
OptimizationRemarkEmitter *ORE, unsigned VecWidth,
unsigned UnrollFactor, LoopVectorizationLegality *LVL,
LoopVectorizationCostModel *CM)
: OrigLoop(OrigLoop), PSE(PSE), LI(LI), DT(DT), TLI(TLI), TTI(TTI),
- ORE(ORE), VF(VecWidth), UF(UnrollFactor),
+ AC(AC), ORE(ORE), VF(VecWidth), UF(UnrollFactor),
Builder(PSE.getSE()->getContext()), Induction(nullptr),
OldInduction(nullptr), VectorLoopValueMap(UnrollFactor, VecWidth),
TripCount(nullptr), VectorTripCount(nullptr), Legal(LVL), Cost(CM),
const TargetLibraryInfo *TLI;
/// Target Transform Info.
const TargetTransformInfo *TTI;
+ /// Assumption Cache.
+ AssumptionCache *AC;
/// Interface to emit optimization remarks.
OptimizationRemarkEmitter *ORE;
InnerLoopUnroller(Loop *OrigLoop, PredicatedScalarEvolution &PSE,
LoopInfo *LI, DominatorTree *DT,
const TargetLibraryInfo *TLI,
- const TargetTransformInfo *TTI,
+ const TargetTransformInfo *TTI, AssumptionCache *AC,
OptimizationRemarkEmitter *ORE, unsigned UnrollFactor,
LoopVectorizationLegality *LVL,
LoopVectorizationCostModel *CM)
- : InnerLoopVectorizer(OrigLoop, PSE, LI, DT, TLI, TTI, ORE, 1,
+ : InnerLoopVectorizer(OrigLoop, PSE, LI, DT, TLI, TTI, AC, ORE, 1,
UnrollFactor, LVL, CM) {}
private:
LoopInfo *LI, LoopVectorizationLegality *Legal,
const TargetTransformInfo &TTI,
const TargetLibraryInfo *TLI, DemandedBits *DB,
+ AssumptionCache *AC,
OptimizationRemarkEmitter *ORE, const Function *F,
const LoopVectorizeHints *Hints)
: TheLoop(L), PSE(PSE), LI(LI), Legal(Legal), TTI(TTI), TLI(TLI), DB(DB),
- ORE(ORE), TheFunction(F), Hints(Hints) {}
+ AC(AC), ORE(ORE), TheFunction(F), Hints(Hints) {}
/// Information about vectorization costs
struct VectorizationFactor {
const TargetLibraryInfo *TLI;
/// Demanded bits analysis.
DemandedBits *DB;
+ /// Assumption cache.
+ AssumptionCache *AC;
/// Interface to emit optimization remarks.
OptimizationRemarkEmitter *ORE;
auto *TLIP = getAnalysisIfAvailable<TargetLibraryInfoWrapperPass>();
auto *TLI = TLIP ? &TLIP->getTLI() : nullptr;
auto *AA = &getAnalysis<AAResultsWrapperPass>().getAAResults();
+ auto *AC = &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
auto *LAA = &getAnalysis<LoopAccessLegacyAnalysis>();
auto *DB = &getAnalysis<DemandedBitsWrapperPass>().getDemandedBits();
auto *ORE = &getAnalysis<OptimizationRemarkEmitterWrapperPass>().getORE();
std::function<const LoopAccessInfo &(Loop &)> GetLAA =
[&](Loop &L) -> const LoopAccessInfo & { return LAA->getInfo(&L); };
- return Impl.runImpl(F, *SE, *LI, *TTI, *DT, *BFI, TLI, *DB, *AA, GetLAA,
- *ORE);
+ return Impl.runImpl(F, *SE, *LI, *TTI, *DT, *BFI, TLI, *DB, *AA, *AC,
+ GetLAA, *ORE);
}
void getAnalysisUsage(AnalysisUsage &AU) const override {
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequiredID(LoopSimplifyID);
AU.addRequiredID(LCSSAID);
AU.addRequired<BlockFrequencyInfoWrapperPass>();
// Add the cloned scalar to the scalar map entry.
Entry[Part][Lane] = Cloned;
+ // If we just cloned a new assumption, add it the assumption cache.
+ if (auto *II = dyn_cast<IntrinsicInst>(Cloned))
+ if (II->getIntrinsicID() == Intrinsic::assume)
+ AC->registerAssumption(II);
+
// End if-block.
if (IfPredicateInstr)
PredicatedInstructions.push_back(std::make_pair(Cloned, Cmp));
INITIALIZE_PASS_DEPENDENCY(BasicAAWrapperPass)
INITIALIZE_PASS_DEPENDENCY(AAResultsWrapperPass)
INITIALIZE_PASS_DEPENDENCY(GlobalsAAWrapperPass)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(BlockFrequencyInfoWrapperPass)
INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
INITIALIZE_PASS_DEPENDENCY(ScalarEvolutionWrapperPass)
void LoopVectorizationCostModel::collectValuesToIgnore() {
// Ignore ephemeral values.
- CodeMetrics::collectEphemeralValues(TheLoop, ValuesToIgnore);
+ CodeMetrics::collectEphemeralValues(TheLoop, AC, ValuesToIgnore);
// Ignore type-promoting instructions we identified during reduction
// detection.
// Add the cloned scalar to the scalar map entry.
Entry[Part][0] = Cloned;
+ // If we just cloned a new assumption, add it the assumption cache.
+ if (auto *II = dyn_cast<IntrinsicInst>(Cloned))
+ if (II->getIntrinsicID() == Intrinsic::assume)
+ AC->registerAssumption(II);
+
// End if-block.
if (IfPredicateInstr)
PredicatedInstructions.push_back(std::make_pair(Cloned, Cmp));
}
// Use the cost model.
- LoopVectorizationCostModel CM(L, PSE, LI, &LVL, *TTI, TLI, DB, ORE, F,
+ LoopVectorizationCostModel CM(L, PSE, LI, &LVL, *TTI, TLI, DB, AC, ORE, F,
&Hints);
CM.collectValuesToIgnore();
assert(IC > 1 && "interleave count should not be 1 or 0");
// If we decided that it is not legal to vectorize the loop, then
// interleave it.
- InnerLoopUnroller Unroller(L, PSE, LI, DT, TLI, TTI, ORE, IC, &LVL, &CM);
+ InnerLoopUnroller Unroller(L, PSE, LI, DT, TLI, TTI, AC, ORE, IC, &LVL,
+ &CM);
Unroller.vectorize();
ORE->emit(OptimizationRemark(LV_NAME, "Interleaved", L->getStartLoc(),
<< NV("InterleaveCount", IC) << ")");
} else {
// If we decided that it is *legal* to vectorize the loop, then do it.
- InnerLoopVectorizer LB(L, PSE, LI, DT, TLI, TTI, ORE, VF.Width, IC, &LVL,
- &CM);
+ InnerLoopVectorizer LB(L, PSE, LI, DT, TLI, TTI, AC, ORE, VF.Width, IC,
+ &LVL, &CM);
LB.vectorize();
++LoopsVectorized;
bool LoopVectorizePass::runImpl(
Function &F, ScalarEvolution &SE_, LoopInfo &LI_, TargetTransformInfo &TTI_,
DominatorTree &DT_, BlockFrequencyInfo &BFI_, TargetLibraryInfo *TLI_,
- DemandedBits &DB_, AliasAnalysis &AA_,
+ DemandedBits &DB_, AliasAnalysis &AA_, AssumptionCache &AC_,
std::function<const LoopAccessInfo &(Loop &)> &GetLAA_,
OptimizationRemarkEmitter &ORE_) {
BFI = &BFI_;
TLI = TLI_;
AA = &AA_;
+ AC = &AC_;
GetLAA = &GetLAA_;
DB = &DB_;
ORE = &ORE_;
auto &BFI = AM.getResult<BlockFrequencyAnalysis>(F);
auto *TLI = AM.getCachedResult<TargetLibraryAnalysis>(F);
auto &AA = AM.getResult<AAManager>(F);
+ auto &AC = AM.getResult<AssumptionAnalysis>(F);
auto &DB = AM.getResult<DemandedBitsAnalysis>(F);
auto &ORE = AM.getResult<OptimizationRemarkEmitterAnalysis>(F);
return LAM.getResult<LoopAccessAnalysis>(L);
};
bool Changed =
- runImpl(F, SE, LI, TTI, DT, BFI, TLI, DB, AA, GetLAA, ORE);
+ runImpl(F, SE, LI, TTI, DT, BFI, TLI, DB, AA, AC, GetLAA, ORE);
if (!Changed)
return PreservedAnalyses::all();
PreservedAnalyses PA;
#include "llvm/ADT/Optional.h"
#include "llvm/ADT/PostOrderIterator.h"
#include "llvm/ADT/SetVector.h"
-#include "llvm/ADT/SmallSet.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/Analysis/CodeMetrics.h"
#include "llvm/Analysis/GlobalsModRef.h"
BoUpSLP(Function *Func, ScalarEvolution *Se, TargetTransformInfo *Tti,
TargetLibraryInfo *TLi, AliasAnalysis *Aa, LoopInfo *Li,
- DominatorTree *Dt, DemandedBits *DB, const DataLayout *DL)
+ DominatorTree *Dt, AssumptionCache *AC, DemandedBits *DB,
+ const DataLayout *DL)
: NumLoadsWantToKeepOrder(0), NumLoadsWantToChangeOrder(0), F(Func),
- SE(Se), TTI(Tti), TLI(TLi), AA(Aa), LI(Li), DT(Dt), DB(DB),
+ SE(Se), TTI(Tti), TLI(TLi), AA(Aa), LI(Li), DT(Dt), AC(AC), DB(DB),
DL(DL), Builder(Se->getContext()) {
- CodeMetrics::collectEphemeralValues(F, EphValues);
+ CodeMetrics::collectEphemeralValues(F, AC, EphValues);
// Use the vector register size specified by the target unless overridden
// by a command-line option.
// TODO: It would be better to limit the vectorization factor based on
AliasAnalysis *AA;
LoopInfo *LI;
DominatorTree *DT;
+ AssumptionCache *AC;
DemandedBits *DB;
const DataLayout *DL;
unsigned MaxVecRegSize; // This is set by TTI or overridden by cl::opt.
// Determine the maximum number of bits required to store the scalar
// values.
for (auto *Scalar : ToDemote) {
- auto NumSignBits = ComputeNumSignBits(Scalar, *DL, 0, 0, DT);
+ auto NumSignBits = ComputeNumSignBits(Scalar, *DL, 0, AC, 0, DT);
auto NumTypeBits = DL->getTypeSizeInBits(Scalar->getType());
MaxBitWidth = std::max<unsigned>(NumTypeBits - NumSignBits, MaxBitWidth);
}
auto *AA = &getAnalysis<AAResultsWrapperPass>().getAAResults();
auto *LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo();
auto *DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
+ auto *AC = &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(F);
auto *DB = &getAnalysis<DemandedBitsWrapperPass>().getDemandedBits();
- return Impl.runImpl(F, SE, TTI, TLI, AA, LI, DT, DB);
+ return Impl.runImpl(F, SE, TTI, TLI, AA, LI, DT, AC, DB);
}
void getAnalysisUsage(AnalysisUsage &AU) const override {
FunctionPass::getAnalysisUsage(AU);
+ AU.addRequired<AssumptionCacheTracker>();
AU.addRequired<ScalarEvolutionWrapperPass>();
AU.addRequired<AAResultsWrapperPass>();
AU.addRequired<TargetTransformInfoWrapperPass>();
auto *AA = &AM.getResult<AAManager>(F);
auto *LI = &AM.getResult<LoopAnalysis>(F);
auto *DT = &AM.getResult<DominatorTreeAnalysis>(F);
+ auto *AC = &AM.getResult<AssumptionAnalysis>(F);
auto *DB = &AM.getResult<DemandedBitsAnalysis>(F);
- bool Changed = runImpl(F, SE, TTI, TLI, AA, LI, DT, DB);
+ bool Changed = runImpl(F, SE, TTI, TLI, AA, LI, DT, AC, DB);
if (!Changed)
return PreservedAnalyses::all();
PreservedAnalyses PA;
TargetTransformInfo *TTI_,
TargetLibraryInfo *TLI_, AliasAnalysis *AA_,
LoopInfo *LI_, DominatorTree *DT_,
- DemandedBits *DB_) {
+ AssumptionCache *AC_, DemandedBits *DB_) {
SE = SE_;
TTI = TTI_;
TLI = TLI_;
AA = AA_;
LI = LI_;
DT = DT_;
+ AC = AC_;
DB = DB_;
DL = &F.getParent()->getDataLayout();
// Use the bottom up slp vectorizer to construct chains that start with
// store instructions.
- BoUpSLP R(&F, SE, TTI, TLI, AA, LI, DT, DB, DL);
+ BoUpSLP R(&F, SE, TTI, TLI, AA, LI, DT, AC, DB, DL);
// A general note: the vectorizer must use BoUpSLP::eraseInstruction() to
// delete instructions.
INITIALIZE_PASS_BEGIN(SLPVectorizer, SV_NAME, lv_name, false, false)
INITIALIZE_PASS_DEPENDENCY(AAResultsWrapperPass)
INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
INITIALIZE_PASS_DEPENDENCY(ScalarEvolutionWrapperPass)
INITIALIZE_PASS_DEPENDENCY(LoopSimplify)
INITIALIZE_PASS_DEPENDENCY(DemandedBitsWrapperPass)
%cmp = icmp slt i32 %iv, 10000
; CHECK: %iv.sext = sext i32 %iv to i64
; CHECK-NEXT: --> {0,+,3}<nuw><nsw><%loop>
- call void @llvm.assume(i1 %cmp) [ "affected"(i32 %iv) ]
+ call void @llvm.assume(i1 %cmp)
%c = load volatile i1, i1* %cond
br i1 %c, label %loop, label %leave
%cmp = icmp ugt i32 %iv.inc, -10000
; CHECK: %iv.zext = zext i32 %iv to i64
; CHECK-NEXT: --> {30000,+,-2}<nw><%loop>
- call void @llvm.assume(i1 %cmp) [ "affected"(i32 %iv.inc) ]
+ call void @llvm.assume(i1 %cmp)
%c = load volatile i1, i1* %cond
br i1 %c, label %loop, label %leave
entry:
%n = and i32 %no, 4294967294
%0 = icmp sgt i32 %n, 0 ; <i1> [#uses=1]
- tail call void @llvm.assume(i1 %0) [ "affected"(i32 %n) ]
+ tail call void @llvm.assume(i1 %0)
br label %bb.nph
bb.nph: ; preds = %entry
define i8 @test2(i8 %a) {
; CHECK-LABEL: @test2
%cmp1 = icmp eq i8 %a, 5
- call void @llvm.assume(i1 %cmp1) [ "affected"(i8 %a) ]
+ call void @llvm.assume(i1 %cmp1)
%cmp2 = icmp eq i8 %a, 3
; CHECK: br i1 false, label %dead, label %exit
br i1 %cmp2, label %dead, label %exit
dead:
%cmp2 = icmp eq i8 %a, 3
; CHECK: call void @llvm.assume(i1 false)
- call void @llvm.assume(i1 %cmp2) [ "affected"(i8 %a) ]
+ call void @llvm.assume(i1 %cmp2)
ret i8 %a
exit:
ret i8 0
define void @_Z3fooR1s(%struct.s* nocapture readonly dereferenceable(8) %x) #0 {
; CHECK-LABEL: @_Z3fooR1s
-; CHECK: call void @llvm.assume(i1 %maskcond) [ "affected"(i64 %maskedptr, i64 %ptrint, double* %{{.*}}) ]
+; CHECK: call void @llvm.assume
; CHECK-NOT: call void @llvm.assume
entry:
; been removed:
; CHECK-LABEL: @foo1
; CHECK-DAG: load i32, i32* %a, align 32
-; CHECK-DAG: call void @llvm.assume(i1 %maskcond) [ "affected"(i64 %maskedptr, i64 %ptrint, i32* %a) ]
+; CHECK-DAG: call void @llvm.assume
; CHECK: ret i32
%ptrint = ptrtoint i32* %a to i64
; Same check as in @foo1, but make sure it works if the assume is first too.
; CHECK-LABEL: @foo2
; CHECK-DAG: load i32, i32* %a, align 32
-; CHECK-DAG: call void @llvm.assume(i1 %maskcond) [ "affected"(i64 %maskedptr, i64 %ptrint, i32* %a) ]
+; CHECK-DAG: call void @llvm.assume
; CHECK: ret i32
%ptrint = ptrtoint i32* %a to i64
; CHECK: ret i32 4
%cmp = icmp eq i32 %a, 4
- tail call void @llvm.assume(i1 %cmp) [ "affected"(i32 %a) ]
+ tail call void @llvm.assume(i1 %cmp)
ret i32 %a
}
%and1 = and i32 %a, 3
; CHECK-LABEL: @bar1
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %and, i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 1
%and = and i32 %a, 7
define i32 @bar2(i32 %a) #0 {
entry:
; CHECK-LABEL: @bar2
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %and, i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 1
%and = and i32 %a, 7
; Don't be fooled by other assumes around.
; CHECK-LABEL: @bar3
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %and, i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 1
tail call void @llvm.assume(i1 %x)
%and1 = and i32 %b, 3
; CHECK-LABEL: @bar4
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %and, i32 %a) ]
-; CHECK: call void @llvm.assume(i1 %cmp2) [ "affected"(i32 %a, i32 %b) ]
+; CHECK: call void @llvm.assume
+; CHECK: call void @llvm.assume
; CHECK: ret i32 1
%and = and i32 %a, 7
ret i32 %conv
; CHECK-LABEL: @icmp1
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 1
}
ret i32 %lnot.ext
; CHECK-LABEL: @icmp2
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 0
}
; CHECK-LABEL: @nonnull2
; CHECK-NOT: !nonnull
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %load) ]
+; CHECK: call void @llvm.assume
}
; Make sure the above canonicalization does not trigger
; CHECK-LABEL: @nonnull3
; CHECK-NOT: !nonnull
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32* %load) ]
+; CHECK: call void @llvm.assume
}
; Make sure the above canonicalization does not trigger
; CHECK-LABEL: @nonnull4
; CHECK-NOT: !nonnull
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32* %load) ]
+; CHECK: call void @llvm.assume
}
define i32 @test1(i32 %a) #0 {
entry:
; CHECK-LABEL: @test1
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %and, i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 5
%and = and i32 %a, 15
define i32 @test2(i32 %a) #0 {
entry:
; CHECK-LABEL: @test2
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %a.not, i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 2
%and = and i32 %a, 15
define i32 @test3(i32 %a) #0 {
entry:
; CHECK-LABEL: @test3
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %v, i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 5
%v = or i32 %a, 4294967280
define i32 @test4(i32 %a) #0 {
entry:
; CHECK-LABEL: @test4
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %a.not, i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 2
%v = or i32 %a, 4294967280
define i32 @test5(i32 %a) #0 {
entry:
; CHECK-LABEL: @test5
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 4
%v = xor i32 %a, 1
define i32 @test6(i32 %a) #0 {
entry:
; CHECK-LABEL: @test6
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %v.mask, i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 5
%v = shl i32 %a, 2
define i32 @test7(i32 %a) #0 {
entry:
; CHECK-LABEL: @test7
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %v.mask, i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 20
%v = lshr i32 %a, 2
define i32 @test8(i32 %a) #0 {
entry:
; CHECK-LABEL: @test8
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %v.mask, i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 20
%v = lshr i32 %a, 2
define i32 @test9(i32 %a) #0 {
entry:
; CHECK-LABEL: @test9
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 0
%cmp = icmp sgt i32 %a, 5
define i32 @test10(i32 %a) #0 {
entry:
; CHECK-LABEL: @test10
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 -2147483648
%cmp = icmp sle i32 %a, -2
define i32 @test11(i32 %a) #0 {
entry:
; CHECK-LABEL: @test11
-; CHECK: call void @llvm.assume(i1 %cmp) [ "affected"(i32 %a) ]
+; CHECK: call void @llvm.assume
; CHECK: ret i32 0
%cmp = icmp ule i32 %a, 256
%b = load i32, i32* @B
%b.and = and i32 %b, 1
%b.cnd = icmp eq i32 %b.and, 1
- call void @llvm.assume(i1 %b.cnd) [ "affected"(i32 %b.and, i32 %b) ]
+ call void @llvm.assume(i1 %b.cnd)
%rhs = add i32 %a, %b
%and = and i32 %a, %rhs
taken:
%res1 = call i8* @escape()
%a = icmp eq i8* %res1, null
- tail call void @llvm.assume(i1 %a) [ "affected"(i8* %res1) ]
+ tail call void @llvm.assume(i1 %a)
br label %done
not_taken:
%res2 = call i8* @escape()
%b = icmp ne i8* %res2, null
- tail call void @llvm.assume(i1 %b) [ "affected"(i8* %res2) ]
+ tail call void @llvm.assume(i1 %b)
br label %done
; An assume that can be used to simplify this comparison dominates each
define i32 @test1(i32 %a, i32 %b) #0 {
entry:
%cmp = icmp sgt i32 %a, 5
- tail call void @llvm.assume(i1 %cmp) [ "affected"(i32 %a) ]
+ tail call void @llvm.assume(i1 %cmp)
%cmp1 = icmp sgt i32 %b, 1234
br i1 %cmp1, label %if.then, label %if.else
define i32 @test2(i32 %a) #0 {
entry:
%cmp = icmp sgt i32 %a, 5
- tail call void @llvm.assume(i1 %cmp) [ "affected"(i32 %a) ]
+ tail call void @llvm.assume(i1 %cmp)
%cmp1 = icmp sgt i32 %a, 3
br i1 %cmp1, label %if.then, label %return
; RUN: opt -S -loop-rotate < %s | FileCheck %s
-; RUN: opt -S -passes='require<targetir>,loop(rotate)' < %s | FileCheck %s
+; RUN: opt -S -passes='require<targetir>,require<assumptions>,loop(rotate)' < %s | FileCheck %s
target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64"
target triple = "x86_64-apple-darwin10.0.0"
; CHECK-LABEL: @reassociate_gep_assume(
; assume(j >= 0)
%cmp = icmp sgt i32 %j, -1
- call void @llvm.assume(i1 %cmp) [ "affected"(i32 %j) ]
+ call void @llvm.assume(i1 %cmp)
%1 = add i32 %i, %j
%cmp2 = icmp sgt i32 %1, -1
- call void @llvm.assume(i1 %cmp2) [ "affected"(i32 %1) ]
+ call void @llvm.assume(i1 %cmp2)
%idxprom.j = zext i32 %j to i64
%2 = getelementptr float, float* %a, i64 %idxprom.j
; CHECK-LABEL: @test5
; CHECK: br i1 [[IGNORE:%.*]], label %true, label %false
%cmp = icmp ult i8 %a, 2
- call void @llvm.assume(i1 %cmp) [ "affected"(i8 %a) ]
+ call void @llvm.assume(i1 %cmp)
switch i8 %a, label %default [i8 1, label %true
i8 0, label %false]
true:
; CHECK: br i1 [[IGNORE:%.*]], label %true, label %false
%and = and i8 %a, 254
%cmp = icmp eq i8 %and, 254
- call void @llvm.assume(i1 %cmp) [ "affected"(i8 %and, i8 %a) ]
+ call void @llvm.assume(i1 %cmp)
switch i8 %a, label %default [i8 255, label %true
i8 254, label %false]
true:
; CHECK: br i1 [[IGNORE:%.*]], label %true, label %false
%and = and i8 %a, 254
%cmp = icmp eq i8 %and, 254
- call void @llvm.assume(i1 %cmp) [ "affected"(i8 %and, i8 %a) ]
+ call void @llvm.assume(i1 %cmp)
switch i8 %a, label %default [i8 255, label %true
i8 254, label %false
i8 0, label %also_dead]
; CHECK: switch i8
%and = and i8 %a, 254
%cmp = icmp eq i8 %and, undef
- call void @llvm.assume(i1 %cmp) [ "affected"(i8 %and, i8 %a) ]
+ call void @llvm.assume(i1 %cmp)
switch i8 %a, label %default [i8 255, label %true
i8 254, label %false]
true:
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/ADT/SetVector.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/BasicAliasAnalysis.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/AsmParser/Parser.h"
Module M;
TargetLibraryInfoImpl TLII;
TargetLibraryInfo TLI;
+ std::unique_ptr<AssumptionCache> AC;
std::unique_ptr<BasicAAResult> BAR;
std::unique_ptr<AAResults> AAR;
AAR.reset(new AAResults(TLI));
// Build the various AA results and register them.
- BAR.reset(new BasicAAResult(M.getDataLayout(), TLI));
+ AC.reset(new AssumptionCache(F));
+ BAR.reset(new BasicAAResult(M.getDataLayout(), TLI, *AC));
AAR->addAAResult(*BAR);
return *AAR;
#include "llvm/Analysis/ScalarEvolutionExpander.h"
#include "llvm/Analysis/ScalarEvolutionExpressions.h"
+#include "llvm/Analysis/AssumptionCache.h"
#include "llvm/Analysis/LoopInfo.h"
#include "llvm/Analysis/TargetLibraryInfo.h"
#include "llvm/ADT/SmallVector.h"
TargetLibraryInfoImpl TLII;
TargetLibraryInfo TLI;
+ std::unique_ptr<AssumptionCache> AC;
std::unique_ptr<DominatorTree> DT;
std::unique_ptr<LoopInfo> LI;
ScalarEvolutionsTest() : M("", Context), TLII(), TLI(TLII) {}
ScalarEvolution buildSE(Function &F) {
+ AC.reset(new AssumptionCache(F));
DT.reset(new DominatorTree(F));
LI.reset(new LoopInfo(*DT));
- return ScalarEvolution(F, TLI, *DT, *LI);
+ return ScalarEvolution(F, TLI, *AC, *DT, *LI);
}
void runWithFunctionAndSE(
InitializeNativeTarget();
InitializeNativeTargetAsmPrinter();
+ // FIXME: It isn't at all clear why this is necesasry, but without it we
+ // fail to initialize the AssumptionCacheTracker.
+ initializeAssumptionCacheTrackerPass(*PassRegistry::getPassRegistry());
+
#ifdef LLVM_ON_WIN32
// On Windows, generate ELF objects by specifying "-elf" in triple
HostTriple += "-elf";
// Things that we need to build after the function is created.
struct TestAnalyses {
DominatorTree DT;
+ AssumptionCache AC;
AAResults AA;
BasicAAResult BAA;
// We need to defer MSSA construction until AA is *entirely* set up, which
MemorySSAWalker *Walker;
TestAnalyses(MemorySSATest &Test)
- : DT(*Test.F), AA(Test.TLI),
- BAA(Test.DL, Test.TLI, &DT) {
+ : DT(*Test.F), AC(*Test.F), AA(Test.TLI),
+ BAA(Test.DL, Test.TLI, AC, &DT) {
AA.addAAResult(BAA);
MSSA = make_unique<MemorySSA>(*Test.F, &AA, &DT);
Walker = MSSA->getWalker();
OSDN Git Service
