X-Git-Url: http://plrg.eecs.uci.edu/git/?a=blobdiff_plain;f=lib%2FTransforms%2FScalar%2FEarlyCSE.cpp;h=6fa194e57092f3576a9b7662467da9b976c7f810;hb=23c0f3b5a2197452c6a25a74fe2fc624763547a1;hp=258a961df4238b7112474fe6e74189d4ea92067e;hpb=86a1c32e67b23c5e9e42dff9eb86e99ba15bb42f;p=oota-llvm.git diff --git a/lib/Transforms/Scalar/EarlyCSE.cpp b/lib/Transforms/Scalar/EarlyCSE.cpp index 258a961df42..6fa194e5709 100644 --- a/lib/Transforms/Scalar/EarlyCSE.cpp +++ b/lib/Transforms/Scalar/EarlyCSE.cpp @@ -12,21 +12,31 @@ // //===----------------------------------------------------------------------===// -#define DEBUG_TYPE "early-cse" -#include "llvm/Transforms/Scalar.h" -#include "llvm/Instructions.h" -#include "llvm/Pass.h" -#include "llvm/Analysis/Dominators.h" +#include "llvm/Transforms/Scalar/EarlyCSE.h" +#include "llvm/ADT/Hashing.h" +#include "llvm/ADT/ScopedHashTable.h" +#include "llvm/ADT/Statistic.h" +#include "llvm/Analysis/GlobalsModRef.h" +#include "llvm/Analysis/AssumptionCache.h" #include "llvm/Analysis/InstructionSimplify.h" -#include "llvm/Target/TargetData.h" -#include "llvm/Target/TargetLibraryInfo.h" -#include "llvm/Transforms/Utils/Local.h" +#include "llvm/Analysis/TargetLibraryInfo.h" +#include "llvm/Analysis/TargetTransformInfo.h" +#include "llvm/IR/DataLayout.h" +#include "llvm/IR/Dominators.h" +#include "llvm/IR/Instructions.h" +#include "llvm/IR/IntrinsicInst.h" +#include "llvm/IR/PatternMatch.h" +#include "llvm/Pass.h" #include "llvm/Support/Debug.h" #include "llvm/Support/RecyclingAllocator.h" -#include "llvm/ADT/ScopedHashTable.h" -#include "llvm/ADT/Statistic.h" +#include "llvm/Support/raw_ostream.h" +#include "llvm/Transforms/Scalar.h" +#include "llvm/Transforms/Utils/Local.h" #include using namespace llvm; +using namespace llvm::PatternMatch; + +#define DEBUG_TYPE "early-cse" STATISTIC(NumSimplify, "Number of instructions simplified or DCE'd"); STATISTIC(NumCSE, "Number of instructions CSE'd"); @@ -34,54 +44,44 @@ STATISTIC(NumCSELoad, "Number of load instructions CSE'd"); STATISTIC(NumCSECall, "Number of call instructions CSE'd"); STATISTIC(NumDSE, "Number of trivial dead stores removed"); -static unsigned getHash(const void *V) { - return DenseMapInfo::getHashValue(V); -} - //===----------------------------------------------------------------------===// // SimpleValue //===----------------------------------------------------------------------===// namespace { - /// SimpleValue - Instances of this struct represent available values in the - /// scoped hash table. - struct SimpleValue { - Instruction *Inst; +/// \brief Struct representing the available values in the scoped hash table. +struct SimpleValue { + Instruction *Inst; - SimpleValue(Instruction *I) : Inst(I) { - assert((isSentinel() || canHandle(I)) && "Inst can't be handled!"); - } + SimpleValue(Instruction *I) : Inst(I) { + assert((isSentinel() || canHandle(I)) && "Inst can't be handled!"); + } - bool isSentinel() const { - return Inst == DenseMapInfo::getEmptyKey() || - Inst == DenseMapInfo::getTombstoneKey(); - } + bool isSentinel() const { + return Inst == DenseMapInfo::getEmptyKey() || + Inst == DenseMapInfo::getTombstoneKey(); + } - static bool canHandle(Instruction *Inst) { - // This can only handle non-void readnone functions. - if (CallInst *CI = dyn_cast(Inst)) - return CI->doesNotAccessMemory() && !CI->getType()->isVoidTy(); - return isa(Inst) || isa(Inst) || - isa(Inst) || isa(Inst) || - isa(Inst) || isa(Inst) || - isa(Inst) || isa(Inst) || - isa(Inst) || isa(Inst); - } - }; + static bool canHandle(Instruction *Inst) { + // This can only handle non-void readnone functions. + if (CallInst *CI = dyn_cast(Inst)) + return CI->doesNotAccessMemory() && !CI->getType()->isVoidTy(); + return isa(Inst) || isa(Inst) || + isa(Inst) || isa(Inst) || + isa(Inst) || isa(Inst) || + isa(Inst) || isa(Inst) || + isa(Inst) || isa(Inst); + } +}; } namespace llvm { -// SimpleValue is POD. -template<> struct isPodLike { - static const bool value = true; -}; - -template<> struct DenseMapInfo { +template <> struct DenseMapInfo { static inline SimpleValue getEmptyKey() { - return DenseMapInfo::getEmptyKey(); + return DenseMapInfo::getEmptyKey(); } static inline SimpleValue getTombstoneKey() { - return DenseMapInfo::getTombstoneKey(); + return DenseMapInfo::getTombstoneKey(); } static unsigned getHashValue(SimpleValue Val); static bool isEqual(SimpleValue LHS, SimpleValue RHS); @@ -90,35 +90,58 @@ template<> struct DenseMapInfo { unsigned DenseMapInfo::getHashValue(SimpleValue Val) { Instruction *Inst = Val.Inst; - // Hash in all of the operands as pointers. - unsigned Res = 0; - for (unsigned i = 0, e = Inst->getNumOperands(); i != e; ++i) - Res ^= getHash(Inst->getOperand(i)) << (i & 0xF); + if (BinaryOperator *BinOp = dyn_cast(Inst)) { + Value *LHS = BinOp->getOperand(0); + Value *RHS = BinOp->getOperand(1); + if (BinOp->isCommutative() && BinOp->getOperand(0) > BinOp->getOperand(1)) + std::swap(LHS, RHS); + + if (isa(BinOp)) { + // Hash the overflow behavior + unsigned Overflow = + BinOp->hasNoSignedWrap() * OverflowingBinaryOperator::NoSignedWrap | + BinOp->hasNoUnsignedWrap() * + OverflowingBinaryOperator::NoUnsignedWrap; + return hash_combine(BinOp->getOpcode(), Overflow, LHS, RHS); + } - if (CastInst *CI = dyn_cast(Inst)) - Res ^= getHash(CI->getType()); - else if (CmpInst *CI = dyn_cast(Inst)) - Res ^= CI->getPredicate(); - else if (const ExtractValueInst *EVI = dyn_cast(Inst)) { - for (ExtractValueInst::idx_iterator I = EVI->idx_begin(), - E = EVI->idx_end(); I != E; ++I) - Res ^= *I; - } else if (const InsertValueInst *IVI = dyn_cast(Inst)) { - for (InsertValueInst::idx_iterator I = IVI->idx_begin(), - E = IVI->idx_end(); I != E; ++I) - Res ^= *I; - } else { - // nothing extra to hash in. - assert((isa(Inst) || - isa(Inst) || isa(Inst) || - isa(Inst) || isa(Inst) || - isa(Inst) || isa(Inst)) && - "Invalid/unknown instruction"); + return hash_combine(BinOp->getOpcode(), LHS, RHS); + } + + if (CmpInst *CI = dyn_cast(Inst)) { + Value *LHS = CI->getOperand(0); + Value *RHS = CI->getOperand(1); + CmpInst::Predicate Pred = CI->getPredicate(); + if (Inst->getOperand(0) > Inst->getOperand(1)) { + std::swap(LHS, RHS); + Pred = CI->getSwappedPredicate(); + } + return hash_combine(Inst->getOpcode(), Pred, LHS, RHS); } + if (CastInst *CI = dyn_cast(Inst)) + return hash_combine(CI->getOpcode(), CI->getType(), CI->getOperand(0)); + + if (const ExtractValueInst *EVI = dyn_cast(Inst)) + return hash_combine(EVI->getOpcode(), EVI->getOperand(0), + hash_combine_range(EVI->idx_begin(), EVI->idx_end())); + + if (const InsertValueInst *IVI = dyn_cast(Inst)) + return hash_combine(IVI->getOpcode(), IVI->getOperand(0), + IVI->getOperand(1), + hash_combine_range(IVI->idx_begin(), IVI->idx_end())); + + assert((isa(Inst) || isa(Inst) || + isa(Inst) || isa(Inst) || + isa(Inst) || isa(Inst) || + isa(Inst)) && + "Invalid/unknown instruction"); + // Mix in the opcode. - return (Res << 1) ^ Inst->getOpcode(); + return hash_combine( + Inst->getOpcode(), + hash_combine_range(Inst->value_op_begin(), Inst->value_op_end())); } bool DenseMapInfo::isEqual(SimpleValue LHS, SimpleValue RHS) { @@ -127,8 +150,44 @@ bool DenseMapInfo::isEqual(SimpleValue LHS, SimpleValue RHS) { if (LHS.isSentinel() || RHS.isSentinel()) return LHSI == RHSI; - if (LHSI->getOpcode() != RHSI->getOpcode()) return false; - return LHSI->isIdenticalTo(RHSI); + if (LHSI->getOpcode() != RHSI->getOpcode()) + return false; + if (LHSI->isIdenticalTo(RHSI)) + return true; + + // If we're not strictly identical, we still might be a commutable instruction + if (BinaryOperator *LHSBinOp = dyn_cast(LHSI)) { + if (!LHSBinOp->isCommutative()) + return false; + + assert(isa(RHSI) && + "same opcode, but different instruction type?"); + BinaryOperator *RHSBinOp = cast(RHSI); + + // Check overflow attributes + if (isa(LHSBinOp)) { + assert(isa(RHSBinOp) && + "same opcode, but different operator type?"); + if (LHSBinOp->hasNoUnsignedWrap() != RHSBinOp->hasNoUnsignedWrap() || + LHSBinOp->hasNoSignedWrap() != RHSBinOp->hasNoSignedWrap()) + return false; + } + + // Commuted equality + return LHSBinOp->getOperand(0) == RHSBinOp->getOperand(1) && + LHSBinOp->getOperand(1) == RHSBinOp->getOperand(0); + } + if (CmpInst *LHSCmp = dyn_cast(LHSI)) { + assert(isa(RHSI) && + "same opcode, but different instruction type?"); + CmpInst *RHSCmp = cast(RHSI); + // Commuted equality + return LHSCmp->getOperand(0) == RHSCmp->getOperand(1) && + LHSCmp->getOperand(1) == RHSCmp->getOperand(0) && + LHSCmp->getSwappedPredicate() == RHSCmp->getPredicate(); + } + + return false; } //===----------------------------------------------------------------------===// @@ -136,62 +195,52 @@ bool DenseMapInfo::isEqual(SimpleValue LHS, SimpleValue RHS) { //===----------------------------------------------------------------------===// namespace { - /// CallValue - Instances of this struct represent available call values in - /// the scoped hash table. - struct CallValue { - Instruction *Inst; +/// \brief Struct representing the available call values in the scoped hash +/// table. +struct CallValue { + Instruction *Inst; - CallValue(Instruction *I) : Inst(I) { - assert((isSentinel() || canHandle(I)) && "Inst can't be handled!"); - } + CallValue(Instruction *I) : Inst(I) { + assert((isSentinel() || canHandle(I)) && "Inst can't be handled!"); + } - bool isSentinel() const { - return Inst == DenseMapInfo::getEmptyKey() || - Inst == DenseMapInfo::getTombstoneKey(); - } + bool isSentinel() const { + return Inst == DenseMapInfo::getEmptyKey() || + Inst == DenseMapInfo::getTombstoneKey(); + } - static bool canHandle(Instruction *Inst) { - // Don't value number anything that returns void. - if (Inst->getType()->isVoidTy()) - return false; + static bool canHandle(Instruction *Inst) { + // Don't value number anything that returns void. + if (Inst->getType()->isVoidTy()) + return false; - CallInst *CI = dyn_cast(Inst); - if (CI == 0 || !CI->onlyReadsMemory()) - return false; - return true; - } - }; + CallInst *CI = dyn_cast(Inst); + if (!CI || !CI->onlyReadsMemory()) + return false; + return true; + } +}; } namespace llvm { - // CallValue is POD. - template<> struct isPodLike { - static const bool value = true; - }; - - template<> struct DenseMapInfo { - static inline CallValue getEmptyKey() { - return DenseMapInfo::getEmptyKey(); - } - static inline CallValue getTombstoneKey() { - return DenseMapInfo::getTombstoneKey(); - } - static unsigned getHashValue(CallValue Val); - static bool isEqual(CallValue LHS, CallValue RHS); - }; +template <> struct DenseMapInfo { + static inline CallValue getEmptyKey() { + return DenseMapInfo::getEmptyKey(); + } + static inline CallValue getTombstoneKey() { + return DenseMapInfo::getTombstoneKey(); + } + static unsigned getHashValue(CallValue Val); + static bool isEqual(CallValue LHS, CallValue RHS); +}; } + unsigned DenseMapInfo::getHashValue(CallValue Val) { Instruction *Inst = Val.Inst; - // Hash in all of the operands as pointers. - unsigned Res = 0; - for (unsigned i = 0, e = Inst->getNumOperands(); i != e; ++i) { - assert(!Inst->getOperand(i)->getType()->isMetadataTy() && - "Cannot value number calls with metadata operands"); - Res ^= getHash(Inst->getOperand(i)) << (i & 0xF); - } - - // Mix in the opcode. - return (Res << 1) ^ Inst->getOpcode(); + // Hash all of the operands as pointers and mix in the opcode. + return hash_combine( + Inst->getOpcode(), + hash_combine_range(Inst->value_op_begin(), Inst->value_op_end())); } bool DenseMapInfo::isEqual(CallValue LHS, CallValue RHS) { @@ -201,102 +250,119 @@ bool DenseMapInfo::isEqual(CallValue LHS, CallValue RHS) { return LHSI->isIdenticalTo(RHSI); } - //===----------------------------------------------------------------------===// -// EarlyCSE pass. +// EarlyCSE implementation //===----------------------------------------------------------------------===// namespace { - -/// EarlyCSE - This pass does a simple depth-first walk over the dominator -/// tree, eliminating trivially redundant instructions and using instsimplify -/// to canonicalize things as it goes. It is intended to be fast and catch -/// obvious cases so that instcombine and other passes are more effective. It -/// is expected that a later pass of GVN will catch the interesting/hard -/// cases. -class EarlyCSE : public FunctionPass { +/// \brief A simple and fast domtree-based CSE pass. +/// +/// This pass does a simple depth-first walk over the dominator tree, +/// eliminating trivially redundant instructions and using instsimplify to +/// canonicalize things as it goes. It is intended to be fast and catch obvious +/// cases so that instcombine and other passes are more effective. It is +/// expected that a later pass of GVN will catch the interesting/hard cases. +class EarlyCSE { public: - const TargetData *TD; - const TargetLibraryInfo *TLI; - DominatorTree *DT; - typedef RecyclingAllocator > AllocatorTy; - typedef ScopedHashTable, + const TargetLibraryInfo &TLI; + const TargetTransformInfo &TTI; + DominatorTree &DT; + AssumptionCache &AC; + typedef RecyclingAllocator< + BumpPtrAllocator, ScopedHashTableVal> AllocatorTy; + typedef ScopedHashTable, AllocatorTy> ScopedHTType; - /// AvailableValues - This scoped hash table contains the current values of - /// all of our simple scalar expressions. As we walk down the domtree, we - /// look to see if instructions are in this: if so, we replace them with what - /// we find, otherwise we insert them so that dominated values can succeed in - /// their lookup. - ScopedHTType *AvailableValues; - - /// AvailableLoads - This scoped hash table contains the current values - /// of loads. This allows us to get efficient access to dominating loads when + /// \brief A scoped hash table of the current values of all of our simple + /// scalar expressions. + /// + /// As we walk down the domtree, we look to see if instructions are in this: + /// if so, we replace them with what we find, otherwise we insert them so + /// that dominated values can succeed in their lookup. + ScopedHTType AvailableValues; + + /// A scoped hash table of the current values of previously encounted memory + /// locations. + /// + /// This allows us to get efficient access to dominating loads or stores when /// we have a fully redundant load. In addition to the most recent load, we /// keep track of a generation count of the read, which is compared against - /// the current generation count. The current generation count is - /// incremented after every possibly writing memory operation, which ensures - /// that we only CSE loads with other loads that have no intervening store. + /// the current generation count. The current generation count is incremented + /// after every possibly writing memory operation, which ensures that we only + /// CSE loads with other loads that have no intervening store. Ordering + /// events (such as fences or atomic instructions) increment the generation + /// count as well; essentially, we model these as writes to all possible + /// locations. Note that atomic and/or volatile loads and stores can be + /// present the table; it is the responsibility of the consumer to inspect + /// the atomicity/volatility if needed. + struct LoadValue { + Value *Data; + unsigned Generation; + int MatchingId; + bool IsAtomic; + LoadValue() + : Data(nullptr), Generation(0), MatchingId(-1), IsAtomic(false) {} + LoadValue(Value *Data, unsigned Generation, unsigned MatchingId, + bool IsAtomic) + : Data(Data), Generation(Generation), MatchingId(MatchingId), + IsAtomic(IsAtomic) {} + }; typedef RecyclingAllocator > > LoadMapAllocator; - typedef ScopedHashTable, - DenseMapInfo, LoadMapAllocator> LoadHTType; - LoadHTType *AvailableLoads; - - /// AvailableCalls - This scoped hash table contains the current values - /// of read-only call values. It uses the same generation count as loads. - typedef ScopedHashTable > CallHTType; - CallHTType *AvailableCalls; - - /// CurrentGeneration - This is the current generation of the memory value. + ScopedHashTableVal> + LoadMapAllocator; + typedef ScopedHashTable, + LoadMapAllocator> LoadHTType; + LoadHTType AvailableLoads; + + /// \brief A scoped hash table of the current values of read-only call + /// values. + /// + /// It uses the same generation count as loads. + typedef ScopedHashTable> CallHTType; + CallHTType AvailableCalls; + + /// \brief This is the current generation of the memory value. unsigned CurrentGeneration; - static char ID; - explicit EarlyCSE() : FunctionPass(ID) { - initializeEarlyCSEPass(*PassRegistry::getPassRegistry()); - } + /// \brief Set up the EarlyCSE runner for a particular function. + EarlyCSE(const TargetLibraryInfo &TLI, const TargetTransformInfo &TTI, + DominatorTree &DT, AssumptionCache &AC) + : TLI(TLI), TTI(TTI), DT(DT), AC(AC), CurrentGeneration(0) {} - bool runOnFunction(Function &F); + bool run(); private: - - // NodeScope - almost a POD, but needs to call the constructors for the - // scoped hash tables so that a new scope gets pushed on. These are RAII so - // that the scope gets popped when the NodeScope is destroyed. + // Almost a POD, but needs to call the constructors for the scoped hash + // tables so that a new scope gets pushed on. These are RAII so that the + // scope gets popped when the NodeScope is destroyed. class NodeScope { - public: - NodeScope(ScopedHTType *availableValues, - LoadHTType *availableLoads, - CallHTType *availableCalls) : - Scope(*availableValues), - LoadScope(*availableLoads), - CallScope(*availableCalls) {} - - private: - NodeScope(const NodeScope&) LLVM_DELETED_FUNCTION; - void operator=(const NodeScope&) LLVM_DELETED_FUNCTION; + public: + NodeScope(ScopedHTType &AvailableValues, LoadHTType &AvailableLoads, + CallHTType &AvailableCalls) + : Scope(AvailableValues), LoadScope(AvailableLoads), + CallScope(AvailableCalls) {} + + private: + NodeScope(const NodeScope &) = delete; + void operator=(const NodeScope &) = delete; ScopedHTType::ScopeTy Scope; LoadHTType::ScopeTy LoadScope; CallHTType::ScopeTy CallScope; }; - // StackNode - contains all the needed information to create a stack for - // doing a depth first tranversal of the tree. This includes scopes for - // values, loads, and calls as well as the generation. There is a child - // iterator so that the children do not need to be store spearately. + // Contains all the needed information to create a stack for doing a depth + // first tranversal of the tree. This includes scopes for values, loads, and + // calls as well as the generation. There is a child iterator so that the + // children do not need to be store spearately. class StackNode { - public: - StackNode(ScopedHTType *availableValues, - LoadHTType *availableLoads, - CallHTType *availableCalls, - unsigned cg, DomTreeNode *n, - DomTreeNode::iterator child, DomTreeNode::iterator end) : - CurrentGeneration(cg), ChildGeneration(cg), Node(n), - ChildIter(child), EndIter(end), - Scopes(availableValues, availableLoads, availableCalls), - Processed(false) {} + public: + StackNode(ScopedHTType &AvailableValues, LoadHTType &AvailableLoads, + CallHTType &AvailableCalls, unsigned cg, DomTreeNode *n, + DomTreeNode::iterator child, DomTreeNode::iterator end) + : CurrentGeneration(cg), ChildGeneration(cg), Node(n), ChildIter(child), + EndIter(end), Scopes(AvailableValues, AvailableLoads, AvailableCalls), + Processed(false) {} // Accessors. unsigned currentGeneration() { return CurrentGeneration; } @@ -313,9 +379,9 @@ private: bool isProcessed() { return Processed; } void process() { Processed = true; } - private: - StackNode(const StackNode&) LLVM_DELETED_FUNCTION; - void operator=(const StackNode&) LLVM_DELETED_FUNCTION; + private: + StackNode(const StackNode &) = delete; + void operator=(const StackNode &) = delete; // Members. unsigned CurrentGeneration; @@ -327,29 +393,121 @@ private: bool Processed; }; + /// \brief Wrapper class to handle memory instructions, including loads, + /// stores and intrinsic loads and stores defined by the target. + class ParseMemoryInst { + public: + ParseMemoryInst(Instruction *Inst, const TargetTransformInfo &TTI) + : IsTargetMemInst(false), Inst(Inst) { + if (IntrinsicInst *II = dyn_cast(Inst)) + if (TTI.getTgtMemIntrinsic(II, Info) && Info.NumMemRefs == 1) + IsTargetMemInst = true; + } + bool isLoad() const { + if (IsTargetMemInst) return Info.ReadMem; + return isa(Inst); + } + bool isStore() const { + if (IsTargetMemInst) return Info.WriteMem; + return isa(Inst); + } + bool isSimple() const { + if (IsTargetMemInst) return Info.IsSimple; + if (LoadInst *LI = dyn_cast(Inst)) { + return LI->isSimple(); + } else if (StoreInst *SI = dyn_cast(Inst)) { + return SI->isSimple(); + } + return Inst->isAtomic(); + } + bool isAtomic() const { + if (IsTargetMemInst) { + assert(Info.IsSimple && "need to refine IsSimple in TTI"); + return false; + } + return Inst->isAtomic(); + } + bool isUnordered() const { + if (IsTargetMemInst) { + assert(Info.IsSimple && "need to refine IsSimple in TTI"); + return true; + } + if (LoadInst *LI = dyn_cast(Inst)) { + return LI->isUnordered(); + } else if (StoreInst *SI = dyn_cast(Inst)) { + return SI->isUnordered(); + } + // Conservative answer + return !Inst->isAtomic(); + } + + bool isVolatile() const { + if (IsTargetMemInst) { + assert(Info.IsSimple && "need to refine IsSimple in TTI"); + return false; + } + if (LoadInst *LI = dyn_cast(Inst)) { + return LI->isVolatile(); + } else if (StoreInst *SI = dyn_cast(Inst)) { + return SI->isVolatile(); + } + // Conservative answer + return true; + } + + + bool isMatchingMemLoc(const ParseMemoryInst &Inst) const { + return (getPointerOperand() == Inst.getPointerOperand() && + getMatchingId() == Inst.getMatchingId()); + } + bool isValid() const { return getPointerOperand() != nullptr; } + + // For regular (non-intrinsic) loads/stores, this is set to -1. For + // intrinsic loads/stores, the id is retrieved from the corresponding + // field in the MemIntrinsicInfo structure. That field contains + // non-negative values only. + int getMatchingId() const { + if (IsTargetMemInst) return Info.MatchingId; + return -1; + } + Value *getPointerOperand() const { + if (IsTargetMemInst) return Info.PtrVal; + if (LoadInst *LI = dyn_cast(Inst)) { + return LI->getPointerOperand(); + } else if (StoreInst *SI = dyn_cast(Inst)) { + return SI->getPointerOperand(); + } + return nullptr; + } + bool mayReadFromMemory() const { + if (IsTargetMemInst) return Info.ReadMem; + return Inst->mayReadFromMemory(); + } + bool mayWriteToMemory() const { + if (IsTargetMemInst) return Info.WriteMem; + return Inst->mayWriteToMemory(); + } + + private: + bool IsTargetMemInst; + MemIntrinsicInfo Info; + Instruction *Inst; + }; + bool processNode(DomTreeNode *Node); - // This transformation requires dominator postdominator info - virtual void getAnalysisUsage(AnalysisUsage &AU) const { - AU.addRequired(); - AU.addRequired(); - AU.setPreservesCFG(); + Value *getOrCreateResult(Value *Inst, Type *ExpectedType) const { + if (LoadInst *LI = dyn_cast(Inst)) + return LI; + else if (StoreInst *SI = dyn_cast(Inst)) + return SI->getValueOperand(); + assert(isa(Inst) && "Instruction not supported"); + return TTI.getOrCreateResultFromMemIntrinsic(cast(Inst), + ExpectedType); } }; } -char EarlyCSE::ID = 0; - -// createEarlyCSEPass - The public interface to this file. -FunctionPass *llvm::createEarlyCSEPass() { - return new EarlyCSE(); -} - -INITIALIZE_PASS_BEGIN(EarlyCSE, "early-cse", "Early CSE", false, false) -INITIALIZE_PASS_DEPENDENCY(DominatorTree) -INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfo) -INITIALIZE_PASS_END(EarlyCSE, "early-cse", "Early CSE", false, false) - bool EarlyCSE::processNode(DomTreeNode *Node) { BasicBlock *BB = Node->getBlock(); @@ -359,24 +517,49 @@ bool EarlyCSE::processNode(DomTreeNode *Node) { // have invalidated the live-out memory values of our parent value. For now, // just be conservative and invalidate memory if this block has multiple // predecessors. - if (BB->getSinglePredecessor() == 0) + if (!BB->getSinglePredecessor()) ++CurrentGeneration; + // If this node has a single predecessor which ends in a conditional branch, + // we can infer the value of the branch condition given that we took this + // path. We need the single predeccesor to ensure there's not another path + // which reaches this block where the condition might hold a different + // value. Since we're adding this to the scoped hash table (like any other + // def), it will have been popped if we encounter a future merge block. + if (BasicBlock *Pred = BB->getSinglePredecessor()) + if (auto *BI = dyn_cast(Pred->getTerminator())) + if (BI->isConditional()) + if (auto *CondInst = dyn_cast(BI->getCondition())) + if (SimpleValue::canHandle(CondInst)) { + assert(BI->getSuccessor(0) == BB || BI->getSuccessor(1) == BB); + auto *ConditionalConstant = (BI->getSuccessor(0) == BB) ? + ConstantInt::getTrue(BB->getContext()) : + ConstantInt::getFalse(BB->getContext()); + AvailableValues.insert(CondInst, ConditionalConstant); + DEBUG(dbgs() << "EarlyCSE CVP: Add conditional value for '" + << CondInst->getName() << "' as " << *ConditionalConstant + << " in " << BB->getName() << "\n"); + // Replace all dominated uses with the known value + replaceDominatedUsesWith(CondInst, ConditionalConstant, DT, + BasicBlockEdge(Pred, BB)); + } + /// LastStore - Keep track of the last non-volatile store that we saw... for /// as long as there in no instruction that reads memory. If we see a store /// to the same location, we delete the dead store. This zaps trivial dead /// stores which can occur in bitfield code among other things. - StoreInst *LastStore = 0; + Instruction *LastStore = nullptr; bool Changed = false; + const DataLayout &DL = BB->getModule()->getDataLayout(); // See if any instructions in the block can be eliminated. If so, do it. If // not, add them to AvailableValues. - for (BasicBlock::iterator I = BB->begin(), E = BB->end(); I != E; ) { - Instruction *Inst = I++; + for (BasicBlock::iterator I = BB->begin(), E = BB->end(); I != E;) { + Instruction *Inst = &*I++; // Dead instructions should just be removed. - if (isInstructionTriviallyDead(Inst, TLI)) { + if (isInstructionTriviallyDead(Inst, &TLI)) { DEBUG(dbgs() << "EarlyCSE DCE: " << *Inst << '\n'); Inst->eraseFromParent(); Changed = true; @@ -384,9 +567,18 @@ bool EarlyCSE::processNode(DomTreeNode *Node) { continue; } + // Skip assume intrinsics, they don't really have side effects (although + // they're marked as such to ensure preservation of control dependencies), + // and this pass will not disturb any of the assumption's control + // dependencies. + if (match(Inst, m_Intrinsic())) { + DEBUG(dbgs() << "EarlyCSE skipping assumption: " << *Inst << '\n'); + continue; + } + // If the instruction can be simplified (e.g. X+0 = X) then replace it with // its simpler value. - if (Value *V = SimplifyInstruction(Inst, TD, TLI, DT)) { + if (Value *V = SimplifyInstruction(Inst, DL, &TLI, &DT, &AC)) { DEBUG(dbgs() << "EarlyCSE Simplify: " << *Inst << " to: " << *V << '\n'); Inst->replaceAllUsesWith(V); Inst->eraseFromParent(); @@ -398,7 +590,7 @@ bool EarlyCSE::processNode(DomTreeNode *Node) { // If this is a simple instruction that we can value number, process it. if (SimpleValue::canHandle(Inst)) { // See if the instruction has an available value. If so, use it. - if (Value *V = AvailableValues->lookup(Inst)) { + if (Value *V = AvailableValues.lookup(Inst)) { DEBUG(dbgs() << "EarlyCSE CSE: " << *Inst << " to: " << *V << '\n'); Inst->replaceAllUsesWith(V); Inst->eraseFromParent(); @@ -408,52 +600,70 @@ bool EarlyCSE::processNode(DomTreeNode *Node) { } // Otherwise, just remember that this value is available. - AvailableValues->insert(Inst, Inst); + AvailableValues.insert(Inst, Inst); continue; } + ParseMemoryInst MemInst(Inst, TTI); // If this is a non-volatile load, process it. - if (LoadInst *LI = dyn_cast(Inst)) { - // Ignore volatile loads. - if (!LI->isSimple()) { - LastStore = 0; - continue; + if (MemInst.isValid() && MemInst.isLoad()) { + // (conservatively) we can't peak past the ordering implied by this + // operation, but we can add this load to our set of available values + if (MemInst.isVolatile() || !MemInst.isUnordered()) { + LastStore = nullptr; + ++CurrentGeneration; } // If we have an available version of this load, and if it is the right // generation, replace this instruction. - std::pair InVal = - AvailableLoads->lookup(Inst->getOperand(0)); - if (InVal.first != 0 && InVal.second == CurrentGeneration) { - DEBUG(dbgs() << "EarlyCSE CSE LOAD: " << *Inst << " to: " - << *InVal.first << '\n'); - if (!Inst->use_empty()) Inst->replaceAllUsesWith(InVal.first); - Inst->eraseFromParent(); - Changed = true; - ++NumCSELoad; - continue; + LoadValue InVal = AvailableLoads.lookup(MemInst.getPointerOperand()); + if (InVal.Data != nullptr && InVal.Generation == CurrentGeneration && + InVal.MatchingId == MemInst.getMatchingId() && + // We don't yet handle removing loads with ordering of any kind. + !MemInst.isVolatile() && MemInst.isUnordered() && + // We can't replace an atomic load with one which isn't also atomic. + InVal.IsAtomic >= MemInst.isAtomic()) { + Value *Op = getOrCreateResult(InVal.Data, Inst->getType()); + if (Op != nullptr) { + DEBUG(dbgs() << "EarlyCSE CSE LOAD: " << *Inst + << " to: " << *InVal.Data << '\n'); + if (!Inst->use_empty()) + Inst->replaceAllUsesWith(Op); + Inst->eraseFromParent(); + Changed = true; + ++NumCSELoad; + continue; + } } // Otherwise, remember that we have this instruction. - AvailableLoads->insert(Inst->getOperand(0), - std::pair(Inst, CurrentGeneration)); - LastStore = 0; + AvailableLoads.insert( + MemInst.getPointerOperand(), + LoadValue(Inst, CurrentGeneration, MemInst.getMatchingId(), + MemInst.isAtomic())); + LastStore = nullptr; continue; } // If this instruction may read from memory, forget LastStore. - if (Inst->mayReadFromMemory()) - LastStore = 0; + // Load/store intrinsics will indicate both a read and a write to + // memory. The target may override this (e.g. so that a store intrinsic + // does not read from memory, and thus will be treated the same as a + // regular store for commoning purposes). + if (Inst->mayReadFromMemory() && + !(MemInst.isValid() && !MemInst.mayReadFromMemory())) + LastStore = nullptr; // If this is a read-only call, process it. if (CallValue::canHandle(Inst)) { // If we have an available version of this call, and if it is the right // generation, replace this instruction. - std::pair InVal = AvailableCalls->lookup(Inst); - if (InVal.first != 0 && InVal.second == CurrentGeneration) { - DEBUG(dbgs() << "EarlyCSE CSE CALL: " << *Inst << " to: " - << *InVal.first << '\n'); - if (!Inst->use_empty()) Inst->replaceAllUsesWith(InVal.first); + std::pair InVal = AvailableCalls.lookup(Inst); + if (InVal.first != nullptr && InVal.second == CurrentGeneration) { + DEBUG(dbgs() << "EarlyCSE CSE CALL: " << *Inst + << " to: " << *InVal.first << '\n'); + if (!Inst->use_empty()) + Inst->replaceAllUsesWith(InVal.first); Inst->eraseFromParent(); Changed = true; ++NumCSECall; @@ -461,29 +671,45 @@ bool EarlyCSE::processNode(DomTreeNode *Node) { } // Otherwise, remember that we have this instruction. - AvailableCalls->insert(Inst, - std::pair(Inst, CurrentGeneration)); + AvailableCalls.insert( + Inst, std::pair(Inst, CurrentGeneration)); continue; } + // A release fence requires that all stores complete before it, but does + // not prevent the reordering of following loads 'before' the fence. As a + // result, we don't need to consider it as writing to memory and don't need + // to advance the generation. We do need to prevent DSE across the fence, + // but that's handled above. + if (FenceInst *FI = dyn_cast(Inst)) + if (FI->getOrdering() == Release) { + assert(Inst->mayReadFromMemory() && "relied on to prevent DSE above"); + continue; + } + // Okay, this isn't something we can CSE at all. Check to see if it is // something that could modify memory. If so, our available memory values // cannot be used so bump the generation count. if (Inst->mayWriteToMemory()) { ++CurrentGeneration; - if (StoreInst *SI = dyn_cast(Inst)) { + if (MemInst.isValid() && MemInst.isStore()) { // We do a trivial form of DSE if there are two stores to the same - // location with no intervening loads. Delete the earlier store. - if (LastStore && - LastStore->getPointerOperand() == SI->getPointerOperand()) { - DEBUG(dbgs() << "EarlyCSE DEAD STORE: " << *LastStore << " due to: " - << *Inst << '\n'); - LastStore->eraseFromParent(); - Changed = true; - ++NumDSE; - LastStore = 0; - continue; + // location with no intervening loads. Delete the earlier store. Note + // that we can delete an earlier simple store even if the following one + // is ordered/volatile/atomic store. + if (LastStore) { + ParseMemoryInst LastStoreMemInst(LastStore, TTI); + assert(LastStoreMemInst.isSimple() && "Violated invariant"); + if (LastStoreMemInst.isMatchingMemLoc(MemInst)) { + DEBUG(dbgs() << "EarlyCSE DEAD STORE: " << *LastStore + << " due to: " << *Inst << '\n'); + LastStore->eraseFromParent(); + Changed = true; + ++NumDSE; + LastStore = nullptr; + } + // fallthrough - we can exploit information about this store } // Okay, we just invalidated anything we knew about loaded values. Try @@ -491,12 +717,19 @@ bool EarlyCSE::processNode(DomTreeNode *Node) { // version of the pointer. It is safe to forward from volatile stores // to non-volatile loads, so we don't have to check for volatility of // the store. - AvailableLoads->insert(SI->getPointerOperand(), - std::pair(SI->getValueOperand(), CurrentGeneration)); - - // Remember that this was the last store we saw for DSE. - if (SI->isSimple()) - LastStore = SI; + AvailableLoads.insert( + MemInst.getPointerOperand(), + LoadValue(Inst, CurrentGeneration, MemInst.getMatchingId(), + MemInst.isAtomic())); + + // Remember that this was the last normal store we saw for DSE. + // Note that we can't delete an earlier atomic or volatile store in + // favor of a later one which isn't. We could in principle remove an + // earlier unordered store if the later one is also unordered. + if (MemInst.isSimple()) + LastStore = Inst; + else + LastStore = nullptr; } } } @@ -504,31 +737,20 @@ bool EarlyCSE::processNode(DomTreeNode *Node) { return Changed; } - -bool EarlyCSE::runOnFunction(Function &F) { +bool EarlyCSE::run() { + // Note, deque is being used here because there is significant performance + // gains over vector when the container becomes very large due to the + // specific access patterns. For more information see the mailing list + // discussion on this: + // http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20120116/135228.html std::deque nodesToProcess; - TD = getAnalysisIfAvailable(); - TLI = &getAnalysis(); - DT = &getAnalysis(); - - // Tables that the pass uses when walking the domtree. - ScopedHTType AVTable; - AvailableValues = &AVTable; - LoadHTType LoadTable; - AvailableLoads = &LoadTable; - CallHTType CallTable; - AvailableCalls = &CallTable; - - CurrentGeneration = 0; bool Changed = false; // Process the root node. - nodesToProcess.push_front( - new StackNode(AvailableValues, AvailableLoads, AvailableCalls, - CurrentGeneration, DT->getRootNode(), - DT->getRootNode()->begin(), - DT->getRootNode()->end())); + nodesToProcess.push_back(new StackNode( + AvailableValues, AvailableLoads, AvailableCalls, CurrentGeneration, + DT.getRootNode(), DT.getRootNode()->begin(), DT.getRootNode()->end())); // Save the current generation. unsigned LiveOutGeneration = CurrentGeneration; @@ -537,7 +759,7 @@ bool EarlyCSE::runOnFunction(Function &F) { while (!nodesToProcess.empty()) { // Grab the first item off the stack. Set the current generation, remove // the node from the stack, and process it. - StackNode *NodeToProcess = nodesToProcess.front(); + StackNode *NodeToProcess = nodesToProcess.back(); // Initialize class members. CurrentGeneration = NodeToProcess->currentGeneration(); @@ -551,17 +773,15 @@ bool EarlyCSE::runOnFunction(Function &F) { } else if (NodeToProcess->childIter() != NodeToProcess->end()) { // Push the next child onto the stack. DomTreeNode *child = NodeToProcess->nextChild(); - nodesToProcess.push_front( - new StackNode(AvailableValues, - AvailableLoads, - AvailableCalls, - NodeToProcess->childGeneration(), child, - child->begin(), child->end())); + nodesToProcess.push_back( + new StackNode(AvailableValues, AvailableLoads, AvailableCalls, + NodeToProcess->childGeneration(), child, child->begin(), + child->end())); } else { // It has been processed, and there are no more children to process, // so delete it and pop it off the stack. delete NodeToProcess; - nodesToProcess.pop_front(); + nodesToProcess.pop_back(); } } // while (!nodes...) @@ -570,3 +790,75 @@ bool EarlyCSE::runOnFunction(Function &F) { return Changed; } + +PreservedAnalyses EarlyCSEPass::run(Function &F, + AnalysisManager *AM) { + auto &TLI = AM->getResult(F); + auto &TTI = AM->getResult(F); + auto &DT = AM->getResult(F); + auto &AC = AM->getResult(F); + + EarlyCSE CSE(TLI, TTI, DT, AC); + + if (!CSE.run()) + return PreservedAnalyses::all(); + + // CSE preserves the dominator tree because it doesn't mutate the CFG. + // FIXME: Bundle this with other CFG-preservation. + PreservedAnalyses PA; + PA.preserve(); + return PA; +} + +namespace { +/// \brief A simple and fast domtree-based CSE pass. +/// +/// This pass does a simple depth-first walk over the dominator tree, +/// eliminating trivially redundant instructions and using instsimplify to +/// canonicalize things as it goes. It is intended to be fast and catch obvious +/// cases so that instcombine and other passes are more effective. It is +/// expected that a later pass of GVN will catch the interesting/hard cases. +class EarlyCSELegacyPass : public FunctionPass { +public: + static char ID; + + EarlyCSELegacyPass() : FunctionPass(ID) { + initializeEarlyCSELegacyPassPass(*PassRegistry::getPassRegistry()); + } + + bool runOnFunction(Function &F) override { + if (skipOptnoneFunction(F)) + return false; + + auto &TLI = getAnalysis().getTLI(); + auto &TTI = getAnalysis().getTTI(F); + auto &DT = getAnalysis().getDomTree(); + auto &AC = getAnalysis().getAssumptionCache(F); + + EarlyCSE CSE(TLI, TTI, DT, AC); + + return CSE.run(); + } + + void getAnalysisUsage(AnalysisUsage &AU) const override { + AU.addRequired(); + AU.addRequired(); + AU.addRequired(); + AU.addRequired(); + AU.addPreserved(); + AU.setPreservesCFG(); + } +}; +} + +char EarlyCSELegacyPass::ID = 0; + +FunctionPass *llvm::createEarlyCSEPass() { return new EarlyCSELegacyPass(); } + +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)