X-Git-Url: http://demsky.eecs.uci.edu/git/?a=blobdiff_plain;f=lib%2FTransforms%2FScalar%2FReassociate.cpp;h=0da37469505b1140d795e49358f425a85e0d4a0c;hb=0b8c9a80f20772c3793201ab5b251d3520b9cea3;hp=a2d6f08297344f39871d08e08d3ad26dcd8dc01d;hpb=f3f55a9bc1ce62fad7faecff7bd83565d569dee8;p=oota-llvm.git diff --git a/lib/Transforms/Scalar/Reassociate.cpp b/lib/Transforms/Scalar/Reassociate.cpp index a2d6f082973..0da37469505 100644 --- a/lib/Transforms/Scalar/Reassociate.cpp +++ b/lib/Transforms/Scalar/Reassociate.cpp @@ -8,7 +8,7 @@ //===----------------------------------------------------------------------===// // // This pass reassociates commutative expressions in an order that is designed -// to promote better constant propagation, GCSE, LICM, PRE... +// to promote better constant propagation, GCSE, LICM, PRE, etc. // // For example: 4 + (x + 5) -> x + (4 + 5) // @@ -22,25 +22,27 @@ #define DEBUG_TYPE "reassociate" #include "llvm/Transforms/Scalar.h" -#include "llvm/Constants.h" -#include "llvm/DerivedTypes.h" -#include "llvm/Function.h" -#include "llvm/Instructions.h" -#include "llvm/IntrinsicInst.h" -#include "llvm/Pass.h" +#include "llvm/ADT/DenseMap.h" +#include "llvm/ADT/PostOrderIterator.h" +#include "llvm/ADT/STLExtras.h" +#include "llvm/ADT/SetVector.h" +#include "llvm/ADT/Statistic.h" #include "llvm/Assembly/Writer.h" +#include "llvm/IR/Constants.h" +#include "llvm/IR/DerivedTypes.h" +#include "llvm/IR/Function.h" +#include "llvm/IR/IRBuilder.h" +#include "llvm/IR/Instructions.h" +#include "llvm/IR/IntrinsicInst.h" +#include "llvm/Pass.h" #include "llvm/Support/CFG.h" #include "llvm/Support/Debug.h" #include "llvm/Support/ValueHandle.h" #include "llvm/Support/raw_ostream.h" -#include "llvm/ADT/PostOrderIterator.h" -#include "llvm/ADT/Statistic.h" -#include "llvm/ADT/DenseMap.h" +#include "llvm/Transforms/Utils/Local.h" #include -#include using namespace llvm; -STATISTIC(NumLinear , "Number of insts linearized"); STATISTIC(NumChanged, "Number of insts reassociated"); STATISTIC(NumAnnihil, "Number of expr tree annihilated"); STATISTIC(NumFactor , "Number of multiplies factored"); @@ -59,26 +61,68 @@ namespace { #ifndef NDEBUG /// PrintOps - Print out the expression identified in the Ops list. /// -static void PrintOps(Instruction *I, const std::vector &Ops) { +static void PrintOps(Instruction *I, const SmallVectorImpl &Ops) { Module *M = I->getParent()->getParent()->getParent(); - errs() << Instruction::getOpcodeName(I->getOpcode()) << " " + dbgs() << Instruction::getOpcodeName(I->getOpcode()) << " " << *Ops[0].Op->getType() << '\t'; for (unsigned i = 0, e = Ops.size(); i != e; ++i) { - errs() << "[ "; - WriteAsOperand(errs(), Ops[i].Op, false, M); - errs() << ", #" << Ops[i].Rank << "] "; + dbgs() << "[ "; + WriteAsOperand(dbgs(), Ops[i].Op, false, M); + dbgs() << ", #" << Ops[i].Rank << "] "; } } #endif - + +namespace { + /// \brief Utility class representing a base and exponent pair which form one + /// factor of some product. + struct Factor { + Value *Base; + unsigned Power; + + Factor(Value *Base, unsigned Power) : Base(Base), Power(Power) {} + + /// \brief Sort factors by their Base. + struct BaseSorter { + bool operator()(const Factor &LHS, const Factor &RHS) { + return LHS.Base < RHS.Base; + } + }; + + /// \brief Compare factors for equal bases. + struct BaseEqual { + bool operator()(const Factor &LHS, const Factor &RHS) { + return LHS.Base == RHS.Base; + } + }; + + /// \brief Sort factors in descending order by their power. + struct PowerDescendingSorter { + bool operator()(const Factor &LHS, const Factor &RHS) { + return LHS.Power > RHS.Power; + } + }; + + /// \brief Compare factors for equal powers. + struct PowerEqual { + bool operator()(const Factor &LHS, const Factor &RHS) { + return LHS.Power == RHS.Power; + } + }; + }; +} + namespace { class Reassociate : public FunctionPass { - std::map RankMap; - std::map, unsigned> ValueRankMap; + DenseMap RankMap; + DenseMap, unsigned> ValueRankMap; + SetVector > RedoInsts; bool MadeChange; public: static char ID; // Pass identification, replacement for typeid - Reassociate() : FunctionPass(&ID) {} + Reassociate() : FunctionPass(ID) { + initializeReassociatePass(*PassRegistry::getPassRegistry()); + } bool runOnFunction(Function &F); @@ -89,44 +133,46 @@ namespace { void BuildRankMap(Function &F); unsigned getRank(Value *V); void ReassociateExpression(BinaryOperator *I); - void RewriteExprTree(BinaryOperator *I, std::vector &Ops, - unsigned Idx = 0); - Value *OptimizeExpression(BinaryOperator *I, std::vector &Ops); - Value *OptimizeAdd(std::vector &Ops); - void LinearizeExprTree(BinaryOperator *I, std::vector &Ops); - void LinearizeExpr(BinaryOperator *I); + void RewriteExprTree(BinaryOperator *I, SmallVectorImpl &Ops); + Value *OptimizeExpression(BinaryOperator *I, + SmallVectorImpl &Ops); + Value *OptimizeAdd(Instruction *I, SmallVectorImpl &Ops); + bool collectMultiplyFactors(SmallVectorImpl &Ops, + SmallVectorImpl &Factors); + Value *buildMinimalMultiplyDAG(IRBuilder<> &Builder, + SmallVectorImpl &Factors); + Value *OptimizeMul(BinaryOperator *I, SmallVectorImpl &Ops); Value *RemoveFactorFromExpression(Value *V, Value *Factor); - void ReassociateBB(BasicBlock *BB); - - void RemoveDeadBinaryOp(Value *V); + void EraseInst(Instruction *I); + void OptimizeInst(Instruction *I); }; } char Reassociate::ID = 0; -static RegisterPass X("reassociate", "Reassociate expressions"); +INITIALIZE_PASS(Reassociate, "reassociate", + "Reassociate expressions", false, false) // Public interface to the Reassociate pass FunctionPass *llvm::createReassociatePass() { return new Reassociate(); } -void Reassociate::RemoveDeadBinaryOp(Value *V) { - Instruction *Op = dyn_cast(V); - if (!Op || !isa(Op) || !isa(Op) || !Op->use_empty()) - return; - - Value *LHS = Op->getOperand(0), *RHS = Op->getOperand(1); - RemoveDeadBinaryOp(LHS); - RemoveDeadBinaryOp(RHS); +/// isReassociableOp - Return true if V is an instruction of the specified +/// opcode and if it only has one use. +static BinaryOperator *isReassociableOp(Value *V, unsigned Opcode) { + if (V->hasOneUse() && isa(V) && + cast(V)->getOpcode() == Opcode) + return cast(V); + return 0; } - static bool isUnmovableInstruction(Instruction *I) { if (I->getOpcode() == Instruction::PHI || + I->getOpcode() == Instruction::LandingPad || I->getOpcode() == Instruction::Alloca || I->getOpcode() == Instruction::Load || I->getOpcode() == Instruction::Invoke || (I->getOpcode() == Instruction::Call && !isa(I)) || - I->getOpcode() == Instruction::UDiv || + I->getOpcode() == Instruction::UDiv || I->getOpcode() == Instruction::SDiv || I->getOpcode() == Instruction::FDiv || I->getOpcode() == Instruction::URem || @@ -159,13 +205,14 @@ void Reassociate::BuildRankMap(Function &F) { } unsigned Reassociate::getRank(Value *V) { - if (isa(V)) return ValueRankMap[V]; // Function argument... - Instruction *I = dyn_cast(V); - if (I == 0) return 0; // Otherwise it's a global or constant, rank 0. + if (I == 0) { + if (isa(V)) return ValueRankMap[V]; // Function argument. + return 0; // Otherwise it's a global or constant, rank 0. + } - unsigned &CachedRank = ValueRankMap[I]; - if (CachedRank) return CachedRank; // Rank already known? + if (unsigned Rank = ValueRankMap[I]) + return Rank; // Rank already known? // If this is an expression, return the 1+MAX(rank(LHS), rank(RHS)) so that // we can reassociate expressions for code motion! Since we do not recurse @@ -178,200 +225,539 @@ unsigned Reassociate::getRank(Value *V) { // If this is a not or neg instruction, do not count it for rank. This // assures us that X and ~X will have the same rank. - if (!I->getType()->isInteger() || + if (!I->getType()->isIntegerTy() || (!BinaryOperator::isNot(I) && !BinaryOperator::isNeg(I))) ++Rank; - //DEBUG(errs() << "Calculated Rank[" << V->getName() << "] = " + //DEBUG(dbgs() << "Calculated Rank[" << V->getName() << "] = " // << Rank << "\n"); - return CachedRank = Rank; -} - -/// isReassociableOp - Return true if V is an instruction of the specified -/// opcode and if it only has one use. -static BinaryOperator *isReassociableOp(Value *V, unsigned Opcode) { - if ((V->hasOneUse() || V->use_empty()) && isa(V) && - cast(V)->getOpcode() == Opcode) - return cast(V); - return 0; + return ValueRankMap[I] = Rank; } /// LowerNegateToMultiply - Replace 0-X with X*-1. /// -static Instruction *LowerNegateToMultiply(Instruction *Neg, - std::map, unsigned> &ValueRankMap) { +static BinaryOperator *LowerNegateToMultiply(Instruction *Neg) { Constant *Cst = Constant::getAllOnesValue(Neg->getType()); - Instruction *Res = BinaryOperator::CreateMul(Neg->getOperand(1), Cst, "",Neg); - ValueRankMap.erase(Neg); + BinaryOperator *Res = + BinaryOperator::CreateMul(Neg->getOperand(1), Cst, "",Neg); + Neg->setOperand(1, Constant::getNullValue(Neg->getType())); // Drop use of op. Res->takeName(Neg); Neg->replaceAllUsesWith(Res); - Neg->eraseFromParent(); + Res->setDebugLoc(Neg->getDebugLoc()); return Res; } -// Given an expression of the form '(A+B)+(D+C)', turn it into '(((A+B)+C)+D)'. -// Note that if D is also part of the expression tree that we recurse to -// linearize it as well. Besides that case, this does not recurse into A,B, or -// C. -void Reassociate::LinearizeExpr(BinaryOperator *I) { - BinaryOperator *LHS = cast(I->getOperand(0)); - BinaryOperator *RHS = cast(I->getOperand(1)); - assert(isReassociableOp(LHS, I->getOpcode()) && - isReassociableOp(RHS, I->getOpcode()) && - "Not an expression that needs linearization?"); - - DEBUG(errs() << "Linear" << *LHS << '\n' << *RHS << '\n' << *I << '\n'); - - // Move the RHS instruction to live immediately before I, avoiding breaking - // dominator properties. - RHS->moveBefore(I); - - // Move operands around to do the linearization. - I->setOperand(1, RHS->getOperand(0)); - RHS->setOperand(0, LHS); - I->setOperand(0, RHS); +/// CarmichaelShift - Returns k such that lambda(2^Bitwidth) = 2^k, where lambda +/// is the Carmichael function. This means that x^(2^k) === 1 mod 2^Bitwidth for +/// every odd x, i.e. x^(2^k) = 1 for every odd x in Bitwidth-bit arithmetic. +/// Note that 0 <= k < Bitwidth, and if Bitwidth > 3 then x^(2^k) = 0 for every +/// even x in Bitwidth-bit arithmetic. +static unsigned CarmichaelShift(unsigned Bitwidth) { + if (Bitwidth < 3) + return Bitwidth - 1; + return Bitwidth - 2; +} - ++NumLinear; - MadeChange = true; - DEBUG(errs() << "Linearized: " << *I << '\n'); +/// IncorporateWeight - Add the extra weight 'RHS' to the existing weight 'LHS', +/// reducing the combined weight using any special properties of the operation. +/// The existing weight LHS represents the computation X op X op ... op X where +/// X occurs LHS times. The combined weight represents X op X op ... op X with +/// X occurring LHS + RHS times. If op is "Xor" for example then the combined +/// operation is equivalent to X if LHS + RHS is odd, or 0 if LHS + RHS is even; +/// the routine returns 1 in LHS in the first case, and 0 in LHS in the second. +static void IncorporateWeight(APInt &LHS, const APInt &RHS, unsigned Opcode) { + // If we were working with infinite precision arithmetic then the combined + // weight would be LHS + RHS. But we are using finite precision arithmetic, + // and the APInt sum LHS + RHS may not be correct if it wraps (it is correct + // for nilpotent operations and addition, but not for idempotent operations + // and multiplication), so it is important to correctly reduce the combined + // weight back into range if wrapping would be wrong. + + // If RHS is zero then the weight didn't change. + if (RHS.isMinValue()) + return; + // If LHS is zero then the combined weight is RHS. + if (LHS.isMinValue()) { + LHS = RHS; + return; + } + // From this point on we know that neither LHS nor RHS is zero. + + if (Instruction::isIdempotent(Opcode)) { + // Idempotent means X op X === X, so any non-zero weight is equivalent to a + // weight of 1. Keeping weights at zero or one also means that wrapping is + // not a problem. + assert(LHS == 1 && RHS == 1 && "Weights not reduced!"); + return; // Return a weight of 1. + } + if (Instruction::isNilpotent(Opcode)) { + // Nilpotent means X op X === 0, so reduce weights modulo 2. + assert(LHS == 1 && RHS == 1 && "Weights not reduced!"); + LHS = 0; // 1 + 1 === 0 modulo 2. + return; + } + if (Opcode == Instruction::Add) { + // TODO: Reduce the weight by exploiting nsw/nuw? + LHS += RHS; + return; + } - // If D is part of this expression tree, tail recurse. - if (isReassociableOp(I->getOperand(1), I->getOpcode())) - LinearizeExpr(I); + assert(Opcode == Instruction::Mul && "Unknown associative operation!"); + unsigned Bitwidth = LHS.getBitWidth(); + // If CM is the Carmichael number then a weight W satisfying W >= CM+Bitwidth + // can be replaced with W-CM. That's because x^W=x^(W-CM) for every Bitwidth + // bit number x, since either x is odd in which case x^CM = 1, or x is even in + // which case both x^W and x^(W - CM) are zero. By subtracting off multiples + // of CM like this weights can always be reduced to the range [0, CM+Bitwidth) + // which by a happy accident means that they can always be represented using + // Bitwidth bits. + // TODO: Reduce the weight by exploiting nsw/nuw? (Could do much better than + // the Carmichael number). + if (Bitwidth > 3) { + /// CM - The value of Carmichael's lambda function. + APInt CM = APInt::getOneBitSet(Bitwidth, CarmichaelShift(Bitwidth)); + // Any weight W >= Threshold can be replaced with W - CM. + APInt Threshold = CM + Bitwidth; + assert(LHS.ult(Threshold) && RHS.ult(Threshold) && "Weights not reduced!"); + // For Bitwidth 4 or more the following sum does not overflow. + LHS += RHS; + while (LHS.uge(Threshold)) + LHS -= CM; + } else { + // To avoid problems with overflow do everything the same as above but using + // a larger type. + unsigned CM = 1U << CarmichaelShift(Bitwidth); + unsigned Threshold = CM + Bitwidth; + assert(LHS.getZExtValue() < Threshold && RHS.getZExtValue() < Threshold && + "Weights not reduced!"); + unsigned Total = LHS.getZExtValue() + RHS.getZExtValue(); + while (Total >= Threshold) + Total -= CM; + LHS = Total; + } } - -/// LinearizeExprTree - Given an associative binary expression tree, traverse -/// all of the uses putting it into canonical form. This forces a left-linear -/// form of the the expression (((a+b)+c)+d), and collects information about the -/// rank of the non-tree operands. +typedef std::pair RepeatedValue; + +/// LinearizeExprTree - Given an associative binary expression, return the leaf +/// nodes in Ops along with their weights (how many times the leaf occurs). The +/// original expression is the same as +/// (Ops[0].first op Ops[0].first op ... Ops[0].first) <- Ops[0].second times +/// op +/// (Ops[1].first op Ops[1].first op ... Ops[1].first) <- Ops[1].second times +/// op +/// ... +/// op +/// (Ops[N].first op Ops[N].first op ... Ops[N].first) <- Ops[N].second times +/// +/// Note that the values Ops[0].first, ..., Ops[N].first are all distinct. +/// +/// This routine may modify the function, in which case it returns 'true'. The +/// changes it makes may well be destructive, changing the value computed by 'I' +/// to something completely different. Thus if the routine returns 'true' then +/// you MUST either replace I with a new expression computed from the Ops array, +/// or use RewriteExprTree to put the values back in. +/// +/// A leaf node is either not a binary operation of the same kind as the root +/// node 'I' (i.e. is not a binary operator at all, or is, but with a different +/// opcode), or is the same kind of binary operator but has a use which either +/// does not belong to the expression, or does belong to the expression but is +/// a leaf node. Every leaf node has at least one use that is a non-leaf node +/// of the expression, while for non-leaf nodes (except for the root 'I') every +/// use is a non-leaf node of the expression. +/// +/// For example: +/// expression graph node names +/// +/// + | I +/// / \ | +/// + + | A, B +/// / \ / \ | +/// * + * | C, D, E +/// / \ / \ / \ | +/// + * | F, G +/// +/// The leaf nodes are C, E, F and G. The Ops array will contain (maybe not in +/// that order) (C, 1), (E, 1), (F, 2), (G, 2). +/// +/// The expression is maximal: if some instruction is a binary operator of the +/// same kind as 'I', and all of its uses are non-leaf nodes of the expression, +/// then the instruction also belongs to the expression, is not a leaf node of +/// it, and its operands also belong to the expression (but may be leaf nodes). /// -/// NOTE: These intentionally destroys the expression tree operands (turning -/// them into undef values) to reduce #uses of the values. This means that the -/// caller MUST use something like RewriteExprTree to put the values back in. +/// NOTE: This routine will set operands of non-leaf non-root nodes to undef in +/// order to ensure that every non-root node in the expression has *exactly one* +/// use by a non-leaf node of the expression. This destruction means that the +/// caller MUST either replace 'I' with a new expression or use something like +/// RewriteExprTree to put the values back in if the routine indicates that it +/// made a change by returning 'true'. /// -void Reassociate::LinearizeExprTree(BinaryOperator *I, - std::vector &Ops) { - Value *LHS = I->getOperand(0), *RHS = I->getOperand(1); +/// In the above example either the right operand of A or the left operand of B +/// will be replaced by undef. If it is B's operand then this gives: +/// +/// + | I +/// / \ | +/// + + | A, B - operand of B replaced with undef +/// / \ \ | +/// * + * | C, D, E +/// / \ / \ / \ | +/// + * | F, G +/// +/// Note that such undef operands can only be reached by passing through 'I'. +/// For example, if you visit operands recursively starting from a leaf node +/// then you will never see such an undef operand unless you get back to 'I', +/// which requires passing through a phi node. +/// +/// Note that this routine may also mutate binary operators of the wrong type +/// that have all uses inside the expression (i.e. only used by non-leaf nodes +/// of the expression) if it can turn them into binary operators of the right +/// type and thus make the expression bigger. + +static bool LinearizeExprTree(BinaryOperator *I, + SmallVectorImpl &Ops) { + DEBUG(dbgs() << "LINEARIZE: " << *I << '\n'); + unsigned Bitwidth = I->getType()->getScalarType()->getPrimitiveSizeInBits(); unsigned Opcode = I->getOpcode(); + assert(Instruction::isAssociative(Opcode) && + Instruction::isCommutative(Opcode) && + "Expected an associative and commutative operation!"); + + // Visit all operands of the expression, keeping track of their weight (the + // number of paths from the expression root to the operand, or if you like + // the number of times that operand occurs in the linearized expression). + // For example, if I = X + A, where X = A + B, then I, X and B have weight 1 + // while A has weight two. + + // Worklist of non-leaf nodes (their operands are in the expression too) along + // with their weights, representing a certain number of paths to the operator. + // If an operator occurs in the worklist multiple times then we found multiple + // ways to get to it. + SmallVector, 8> Worklist; // (Op, Weight) + Worklist.push_back(std::make_pair(I, APInt(Bitwidth, 1))); + bool MadeChange = false; + + // Leaves of the expression are values that either aren't the right kind of + // operation (eg: a constant, or a multiply in an add tree), or are, but have + // some uses that are not inside the expression. For example, in I = X + X, + // X = A + B, the value X has two uses (by I) that are in the expression. If + // X has any other uses, for example in a return instruction, then we consider + // X to be a leaf, and won't analyze it further. When we first visit a value, + // if it has more than one use then at first we conservatively consider it to + // be a leaf. Later, as the expression is explored, we may discover some more + // uses of the value from inside the expression. If all uses turn out to be + // from within the expression (and the value is a binary operator of the right + // kind) then the value is no longer considered to be a leaf, and its operands + // are explored. + + // Leaves - Keeps track of the set of putative leaves as well as the number of + // paths to each leaf seen so far. + typedef DenseMap LeafMap; + LeafMap Leaves; // Leaf -> Total weight so far. + SmallVector LeafOrder; // Ensure deterministic leaf output order. - // First step, linearize the expression if it is in ((A+B)+(C+D)) form. - BinaryOperator *LHSBO = isReassociableOp(LHS, Opcode); - BinaryOperator *RHSBO = isReassociableOp(RHS, Opcode); +#ifndef NDEBUG + SmallPtrSet Visited; // For sanity checking the iteration scheme. +#endif + while (!Worklist.empty()) { + std::pair P = Worklist.pop_back_val(); + I = P.first; // We examine the operands of this binary operator. + + for (unsigned OpIdx = 0; OpIdx < 2; ++OpIdx) { // Visit operands. + Value *Op = I->getOperand(OpIdx); + APInt Weight = P.second; // Number of paths to this operand. + DEBUG(dbgs() << "OPERAND: " << *Op << " (" << Weight << ")\n"); + assert(!Op->use_empty() && "No uses, so how did we get to it?!"); + + // If this is a binary operation of the right kind with only one use then + // add its operands to the expression. + if (BinaryOperator *BO = isReassociableOp(Op, Opcode)) { + assert(Visited.insert(Op) && "Not first visit!"); + DEBUG(dbgs() << "DIRECT ADD: " << *Op << " (" << Weight << ")\n"); + Worklist.push_back(std::make_pair(BO, Weight)); + continue; + } - // If this is a multiply expression tree and it contains internal negations, - // transform them into multiplies by -1 so they can be reassociated. - if (I->getOpcode() == Instruction::Mul) { - if (!LHSBO && LHS->hasOneUse() && BinaryOperator::isNeg(LHS)) { - LHS = LowerNegateToMultiply(cast(LHS), ValueRankMap); - LHSBO = isReassociableOp(LHS, Opcode); - } - if (!RHSBO && RHS->hasOneUse() && BinaryOperator::isNeg(RHS)) { - RHS = LowerNegateToMultiply(cast(RHS), ValueRankMap); - RHSBO = isReassociableOp(RHS, Opcode); - } - } + // Appears to be a leaf. Is the operand already in the set of leaves? + LeafMap::iterator It = Leaves.find(Op); + if (It == Leaves.end()) { + // Not in the leaf map. Must be the first time we saw this operand. + assert(Visited.insert(Op) && "Not first visit!"); + if (!Op->hasOneUse()) { + // This value has uses not accounted for by the expression, so it is + // not safe to modify. Mark it as being a leaf. + DEBUG(dbgs() << "ADD USES LEAF: " << *Op << " (" << Weight << ")\n"); + LeafOrder.push_back(Op); + Leaves[Op] = Weight; + continue; + } + // No uses outside the expression, try morphing it. + } else if (It != Leaves.end()) { + // Already in the leaf map. + assert(Visited.count(Op) && "In leaf map but not visited!"); + + // Update the number of paths to the leaf. + IncorporateWeight(It->second, Weight, Opcode); + +#if 0 // TODO: Re-enable once PR13021 is fixed. + // The leaf already has one use from inside the expression. As we want + // exactly one such use, drop this new use of the leaf. + assert(!Op->hasOneUse() && "Only one use, but we got here twice!"); + I->setOperand(OpIdx, UndefValue::get(I->getType())); + MadeChange = true; - if (!LHSBO) { - if (!RHSBO) { - // Neither the LHS or RHS as part of the tree, thus this is a leaf. As - // such, just remember these operands and their rank. - Ops.push_back(ValueEntry(getRank(LHS), LHS)); - Ops.push_back(ValueEntry(getRank(RHS), RHS)); - - // Clear the leaves out. - I->setOperand(0, UndefValue::get(I->getType())); - I->setOperand(1, UndefValue::get(I->getType())); - return; + // If the leaf is a binary operation of the right kind and we now see + // that its multiple original uses were in fact all by nodes belonging + // to the expression, then no longer consider it to be a leaf and add + // its operands to the expression. + if (BinaryOperator *BO = isReassociableOp(Op, Opcode)) { + DEBUG(dbgs() << "UNLEAF: " << *Op << " (" << It->second << ")\n"); + Worklist.push_back(std::make_pair(BO, It->second)); + Leaves.erase(It); + continue; + } +#endif + + // If we still have uses that are not accounted for by the expression + // then it is not safe to modify the value. + if (!Op->hasOneUse()) + continue; + + // No uses outside the expression, try morphing it. + Weight = It->second; + Leaves.erase(It); // Since the value may be morphed below. + } + + // At this point we have a value which, first of all, is not a binary + // expression of the right kind, and secondly, is only used inside the + // expression. This means that it can safely be modified. See if we + // can usefully morph it into an expression of the right kind. + assert((!isa(Op) || + cast(Op)->getOpcode() != Opcode) && + "Should have been handled above!"); + assert(Op->hasOneUse() && "Has uses outside the expression tree!"); + + // If this is a multiply expression, turn any internal negations into + // multiplies by -1 so they can be reassociated. + BinaryOperator *BO = dyn_cast(Op); + if (Opcode == Instruction::Mul && BO && BinaryOperator::isNeg(BO)) { + DEBUG(dbgs() << "MORPH LEAF: " << *Op << " (" << Weight << ") TO "); + BO = LowerNegateToMultiply(BO); + DEBUG(dbgs() << *BO << 'n'); + Worklist.push_back(std::make_pair(BO, Weight)); + MadeChange = true; + continue; + } + + // Failed to morph into an expression of the right type. This really is + // a leaf. + DEBUG(dbgs() << "ADD LEAF: " << *Op << " (" << Weight << ")\n"); + assert(!isReassociableOp(Op, Opcode) && "Value was morphed?"); + LeafOrder.push_back(Op); + Leaves[Op] = Weight; } - - // Turn X+(Y+Z) -> (Y+Z)+X - std::swap(LHSBO, RHSBO); - std::swap(LHS, RHS); - bool Success = !I->swapOperands(); - assert(Success && "swapOperands failed"); - Success = false; - MadeChange = true; - } else if (RHSBO) { - // Turn (A+B)+(C+D) -> (((A+B)+C)+D). This guarantees the the RHS is not - // part of the expression tree. - LinearizeExpr(I); - LHS = LHSBO = cast(I->getOperand(0)); - RHS = I->getOperand(1); - RHSBO = 0; } - // Okay, now we know that the LHS is a nested expression and that the RHS is - // not. Perform reassociation. - assert(!isReassociableOp(RHS, Opcode) && "LinearizeExpr failed!"); - - // Move LHS right before I to make sure that the tree expression dominates all - // values. - LHSBO->moveBefore(I); + // The leaves, repeated according to their weights, represent the linearized + // form of the expression. + for (unsigned i = 0, e = LeafOrder.size(); i != e; ++i) { + Value *V = LeafOrder[i]; + LeafMap::iterator It = Leaves.find(V); + if (It == Leaves.end()) + // Node initially thought to be a leaf wasn't. + continue; + assert(!isReassociableOp(V, Opcode) && "Shouldn't be a leaf!"); + APInt Weight = It->second; + if (Weight.isMinValue()) + // Leaf already output or weight reduction eliminated it. + continue; + // Ensure the leaf is only output once. + It->second = 0; + Ops.push_back(std::make_pair(V, Weight)); + } - // Linearize the expression tree on the LHS. - LinearizeExprTree(LHSBO, Ops); + // For nilpotent operations or addition there may be no operands, for example + // because the expression was "X xor X" or consisted of 2^Bitwidth additions: + // in both cases the weight reduces to 0 causing the value to be skipped. + if (Ops.empty()) { + Constant *Identity = ConstantExpr::getBinOpIdentity(Opcode, I->getType()); + assert(Identity && "Associative operation without identity!"); + Ops.push_back(std::make_pair(Identity, APInt(Bitwidth, 1))); + } - // Remember the RHS operand and its rank. - Ops.push_back(ValueEntry(getRank(RHS), RHS)); - - // Clear the RHS leaf out. - I->setOperand(1, UndefValue::get(I->getType())); + return MadeChange; } // RewriteExprTree - Now that the operands for this expression tree are -// linearized and optimized, emit them in-order. This function is written to be -// tail recursive. +// linearized and optimized, emit them in-order. void Reassociate::RewriteExprTree(BinaryOperator *I, - std::vector &Ops, - unsigned i) { - if (i+2 == Ops.size()) { - if (I->getOperand(0) != Ops[i].Op || - I->getOperand(1) != Ops[i+1].Op) { - Value *OldLHS = I->getOperand(0); - DEBUG(errs() << "RA: " << *I << '\n'); - I->setOperand(0, Ops[i].Op); - I->setOperand(1, Ops[i+1].Op); - DEBUG(errs() << "TO: " << *I << '\n'); + SmallVectorImpl &Ops) { + assert(Ops.size() > 1 && "Single values should be used directly!"); + + // Since our optimizations should never increase the number of operations, the + // new expression can usually be written reusing the existing binary operators + // from the original expression tree, without creating any new instructions, + // though the rewritten expression may have a completely different topology. + // We take care to not change anything if the new expression will be the same + // as the original. If more than trivial changes (like commuting operands) + // were made then we are obliged to clear out any optional subclass data like + // nsw flags. + + /// NodesToRewrite - Nodes from the original expression available for writing + /// the new expression into. + SmallVector NodesToRewrite; + unsigned Opcode = I->getOpcode(); + BinaryOperator *Op = I; + + /// NotRewritable - The operands being written will be the leaves of the new + /// expression and must not be used as inner nodes (via NodesToRewrite) by + /// mistake. Inner nodes are always reassociable, and usually leaves are not + /// (if they were they would have been incorporated into the expression and so + /// would not be leaves), so most of the time there is no danger of this. But + /// in rare cases a leaf may become reassociable if an optimization kills uses + /// of it, or it may momentarily become reassociable during rewriting (below) + /// due it being removed as an operand of one of its uses. Ensure that misuse + /// of leaf nodes as inner nodes cannot occur by remembering all of the future + /// leaves and refusing to reuse any of them as inner nodes. + SmallPtrSet NotRewritable; + for (unsigned i = 0, e = Ops.size(); i != e; ++i) + NotRewritable.insert(Ops[i].Op); + + // ExpressionChanged - Non-null if the rewritten expression differs from the + // original in some non-trivial way, requiring the clearing of optional flags. + // Flags are cleared from the operator in ExpressionChanged up to I inclusive. + BinaryOperator *ExpressionChanged = 0; + for (unsigned i = 0; ; ++i) { + // The last operation (which comes earliest in the IR) is special as both + // operands will come from Ops, rather than just one with the other being + // a subexpression. + if (i+2 == Ops.size()) { + Value *NewLHS = Ops[i].Op; + Value *NewRHS = Ops[i+1].Op; + Value *OldLHS = Op->getOperand(0); + Value *OldRHS = Op->getOperand(1); + + if (NewLHS == OldLHS && NewRHS == OldRHS) + // Nothing changed, leave it alone. + break; + + if (NewLHS == OldRHS && NewRHS == OldLHS) { + // The order of the operands was reversed. Swap them. + DEBUG(dbgs() << "RA: " << *Op << '\n'); + Op->swapOperands(); + DEBUG(dbgs() << "TO: " << *Op << '\n'); + MadeChange = true; + ++NumChanged; + break; + } + + // The new operation differs non-trivially from the original. Overwrite + // the old operands with the new ones. + DEBUG(dbgs() << "RA: " << *Op << '\n'); + if (NewLHS != OldLHS) { + BinaryOperator *BO = isReassociableOp(OldLHS, Opcode); + if (BO && !NotRewritable.count(BO)) + NodesToRewrite.push_back(BO); + Op->setOperand(0, NewLHS); + } + if (NewRHS != OldRHS) { + BinaryOperator *BO = isReassociableOp(OldRHS, Opcode); + if (BO && !NotRewritable.count(BO)) + NodesToRewrite.push_back(BO); + Op->setOperand(1, NewRHS); + } + DEBUG(dbgs() << "TO: " << *Op << '\n'); + + ExpressionChanged = Op; MadeChange = true; ++NumChanged; - - // If we reassociated a tree to fewer operands (e.g. (1+a+2) -> (a+3) - // delete the extra, now dead, nodes. - RemoveDeadBinaryOp(OldLHS); + + break; + } + + // Not the last operation. The left-hand side will be a sub-expression + // while the right-hand side will be the current element of Ops. + Value *NewRHS = Ops[i].Op; + if (NewRHS != Op->getOperand(1)) { + DEBUG(dbgs() << "RA: " << *Op << '\n'); + if (NewRHS == Op->getOperand(0)) { + // The new right-hand side was already present as the left operand. If + // we are lucky then swapping the operands will sort out both of them. + Op->swapOperands(); + } else { + // Overwrite with the new right-hand side. + BinaryOperator *BO = isReassociableOp(Op->getOperand(1), Opcode); + if (BO && !NotRewritable.count(BO)) + NodesToRewrite.push_back(BO); + Op->setOperand(1, NewRHS); + ExpressionChanged = Op; + } + DEBUG(dbgs() << "TO: " << *Op << '\n'); + MadeChange = true; + ++NumChanged; + } + + // Now deal with the left-hand side. If this is already an operation node + // from the original expression then just rewrite the rest of the expression + // into it. + BinaryOperator *BO = isReassociableOp(Op->getOperand(0), Opcode); + if (BO && !NotRewritable.count(BO)) { + Op = BO; + continue; } - return; - } - assert(i+2 < Ops.size() && "Ops index out of range!"); - if (I->getOperand(1) != Ops[i].Op) { - DEBUG(errs() << "RA: " << *I << '\n'); - I->setOperand(1, Ops[i].Op); - DEBUG(errs() << "TO: " << *I << '\n'); + // Otherwise, grab a spare node from the original expression and use that as + // the left-hand side. If there are no nodes left then the optimizers made + // an expression with more nodes than the original! This usually means that + // they did something stupid but it might mean that the problem was just too + // hard (finding the mimimal number of multiplications needed to realize a + // multiplication expression is NP-complete). Whatever the reason, smart or + // stupid, create a new node if there are none left. + BinaryOperator *NewOp; + if (NodesToRewrite.empty()) { + Constant *Undef = UndefValue::get(I->getType()); + NewOp = BinaryOperator::Create(Instruction::BinaryOps(Opcode), + Undef, Undef, "", I); + } else { + NewOp = NodesToRewrite.pop_back_val(); + } + + DEBUG(dbgs() << "RA: " << *Op << '\n'); + Op->setOperand(0, NewOp); + DEBUG(dbgs() << "TO: " << *Op << '\n'); + ExpressionChanged = Op; MadeChange = true; ++NumChanged; + Op = NewOp; } - - BinaryOperator *LHS = cast(I->getOperand(0)); - assert(LHS->getOpcode() == I->getOpcode() && - "Improper expression tree!"); - - // Compactify the tree instructions together with each other to guarantee - // that the expression tree is dominated by all of Ops. - LHS->moveBefore(I); - RewriteExprTree(LHS, Ops, i+1); -} - + // If the expression changed non-trivially then clear out all subclass data + // starting from the operator specified in ExpressionChanged, and compactify + // the operators to just before the expression root to guarantee that the + // expression tree is dominated by all of Ops. + if (ExpressionChanged) + do { + ExpressionChanged->clearSubclassOptionalData(); + if (ExpressionChanged == I) + break; + ExpressionChanged->moveBefore(I); + ExpressionChanged = cast(*ExpressionChanged->use_begin()); + } while (1); + + // Throw away any left over nodes from the original expression. + for (unsigned i = 0, e = NodesToRewrite.size(); i != e; ++i) + RedoInsts.insert(NodesToRewrite[i]); +} -// NegateValue - Insert instructions before the instruction pointed to by BI, -// that computes the negative version of the value specified. The negative -// version of the value is returned, and BI is left pointing at the instruction -// that should be processed next by the reassociation pass. -// +/// NegateValue - Insert instructions before the instruction pointed to by BI, +/// that computes the negative version of the value specified. The negative +/// version of the value is returned, and BI is left pointing at the instruction +/// that should be processed next by the reassociation pass. static Value *NegateValue(Value *V, Instruction *BI) { + if (Constant *C = dyn_cast(V)) + return ConstantExpr::getNeg(C); + // We are trying to expose opportunity for reassociation. One of the things // that we want to do to achieve this is to push a negation as deep into an // expression chain as possible, to expose the add instructions. In practice, @@ -379,27 +765,57 @@ static Value *NegateValue(Value *V, Instruction *BI) { // X = -(A+12+C+D) into X = -A + -12 + -C + -D = -12 + -A + -C + -D // so that later, a: Y = 12+X could get reassociated with the -12 to eliminate // the constants. We assume that instcombine will clean up the mess later if - // we introduce tons of unnecessary negation instructions... + // we introduce tons of unnecessary negation instructions. // - if (Instruction *I = dyn_cast(V)) - if (I->getOpcode() == Instruction::Add && I->hasOneUse()) { - // Push the negates through the add. - I->setOperand(0, NegateValue(I->getOperand(0), BI)); - I->setOperand(1, NegateValue(I->getOperand(1), BI)); - - // We must move the add instruction here, because the neg instructions do - // not dominate the old add instruction in general. By moving it, we are - // assured that the neg instructions we just inserted dominate the - // instruction we are about to insert after them. - // - I->moveBefore(BI); - I->setName(I->getName()+".neg"); - return I; + if (BinaryOperator *I = isReassociableOp(V, Instruction::Add)) { + // Push the negates through the add. + I->setOperand(0, NegateValue(I->getOperand(0), BI)); + I->setOperand(1, NegateValue(I->getOperand(1), BI)); + + // We must move the add instruction here, because the neg instructions do + // not dominate the old add instruction in general. By moving it, we are + // assured that the neg instructions we just inserted dominate the + // instruction we are about to insert after them. + // + I->moveBefore(BI); + I->setName(I->getName()+".neg"); + return I; + } + + // Okay, we need to materialize a negated version of V with an instruction. + // Scan the use lists of V to see if we have one already. + for (Value::use_iterator UI = V->use_begin(), E = V->use_end(); UI != E;++UI){ + User *U = *UI; + if (!BinaryOperator::isNeg(U)) continue; + + // We found one! Now we have to make sure that the definition dominates + // this use. We do this by moving it to the entry block (if it is a + // non-instruction value) or right after the definition. These negates will + // be zapped by reassociate later, so we don't need much finesse here. + BinaryOperator *TheNeg = cast(U); + + // Verify that the negate is in this function, V might be a constant expr. + if (TheNeg->getParent()->getParent() != BI->getParent()->getParent()) + continue; + + BasicBlock::iterator InsertPt; + if (Instruction *InstInput = dyn_cast(V)) { + if (InvokeInst *II = dyn_cast(InstInput)) { + InsertPt = II->getNormalDest()->begin(); + } else { + InsertPt = InstInput; + ++InsertPt; + } + while (isa(InsertPt)) ++InsertPt; + } else { + InsertPt = TheNeg->getParent()->getParent()->getEntryBlock().begin(); } + TheNeg->moveBefore(InsertPt); + return TheNeg; + } // Insert a 'neg' instruction that subtracts the value from zero to get the // negation. - // return BinaryOperator::CreateNeg(V, V->getName() + ".neg", BI); } @@ -409,7 +825,7 @@ static bool ShouldBreakUpSubtract(Instruction *Sub) { // If this is a negation, we can't split it up! if (BinaryOperator::isNeg(Sub)) return false; - + // Don't bother to break this up unless either the LHS is an associable add or // subtract or if this is only used by one. if (isReassociableOp(Sub->getOperand(0), Instruction::Add) || @@ -418,74 +834,67 @@ static bool ShouldBreakUpSubtract(Instruction *Sub) { if (isReassociableOp(Sub->getOperand(1), Instruction::Add) || isReassociableOp(Sub->getOperand(1), Instruction::Sub)) return true; - if (Sub->hasOneUse() && + if (Sub->hasOneUse() && (isReassociableOp(Sub->use_back(), Instruction::Add) || isReassociableOp(Sub->use_back(), Instruction::Sub))) return true; - + return false; } /// BreakUpSubtract - If we have (X-Y), and if either X is an add, or if this is /// only used by an add, transform this into (X+(0-Y)) to promote better /// reassociation. -static Instruction *BreakUpSubtract(Instruction *Sub, - std::map, unsigned> &ValueRankMap) { - // Convert a subtract into an add and a neg instruction... so that sub - // instructions can be commuted with other add instructions... +static BinaryOperator *BreakUpSubtract(Instruction *Sub) { + // Convert a subtract into an add and a neg instruction. This allows sub + // instructions to be commuted with other add instructions. // - // Calculate the negative value of Operand 1 of the sub instruction... - // and set it as the RHS of the add instruction we just made... + // Calculate the negative value of Operand 1 of the sub instruction, + // and set it as the RHS of the add instruction we just made. // Value *NegVal = NegateValue(Sub->getOperand(1), Sub); - Instruction *New = + BinaryOperator *New = BinaryOperator::CreateAdd(Sub->getOperand(0), NegVal, "", Sub); + Sub->setOperand(0, Constant::getNullValue(Sub->getType())); // Drop use of op. + Sub->setOperand(1, Constant::getNullValue(Sub->getType())); // Drop use of op. New->takeName(Sub); // Everyone now refers to the add instruction. - ValueRankMap.erase(Sub); Sub->replaceAllUsesWith(New); - Sub->eraseFromParent(); + New->setDebugLoc(Sub->getDebugLoc()); - DEBUG(errs() << "Negated: " << *New << '\n'); + DEBUG(dbgs() << "Negated: " << *New << '\n'); return New; } /// ConvertShiftToMul - If this is a shift of a reassociable multiply or is used /// by one, change this into a multiply by a constant to assist with further /// reassociation. -static Instruction *ConvertShiftToMul(Instruction *Shl, - std::map, unsigned> &ValueRankMap) { - // If an operand of this shift is a reassociable multiply, or if the shift - // is used by a reassociable multiply or add, turn into a multiply. - if (isReassociableOp(Shl->getOperand(0), Instruction::Mul) || - (Shl->hasOneUse() && - (isReassociableOp(Shl->use_back(), Instruction::Mul) || - isReassociableOp(Shl->use_back(), Instruction::Add)))) { - Constant *MulCst = ConstantInt::get(Shl->getType(), 1); - MulCst = ConstantExpr::getShl(MulCst, cast(Shl->getOperand(1))); - - Instruction *Mul = - BinaryOperator::CreateMul(Shl->getOperand(0), MulCst, "", Shl); - ValueRankMap.erase(Shl); - Mul->takeName(Shl); - Shl->replaceAllUsesWith(Mul); - Shl->eraseFromParent(); - return Mul; - } - return 0; +static BinaryOperator *ConvertShiftToMul(Instruction *Shl) { + Constant *MulCst = ConstantInt::get(Shl->getType(), 1); + MulCst = ConstantExpr::getShl(MulCst, cast(Shl->getOperand(1))); + + BinaryOperator *Mul = + BinaryOperator::CreateMul(Shl->getOperand(0), MulCst, "", Shl); + Shl->setOperand(0, UndefValue::get(Shl->getType())); // Drop use of op. + Mul->takeName(Shl); + Shl->replaceAllUsesWith(Mul); + Mul->setDebugLoc(Shl->getDebugLoc()); + return Mul; } -// Scan backwards and forwards among values with the same rank as element i to -// see if X exists. If X does not exist, return i. -static unsigned FindInOperandList(std::vector &Ops, unsigned i, +/// FindInOperandList - Scan backwards and forwards among values with the same +/// rank as element i to see if X exists. If X does not exist, return i. This +/// is useful when scanning for 'x' when we see '-x' because they both get the +/// same rank. +static unsigned FindInOperandList(SmallVectorImpl &Ops, unsigned i, Value *X) { unsigned XRank = Ops[i].Rank; unsigned e = Ops.size(); for (unsigned j = i+1; j != e && Ops[j].Rank == XRank; ++j) if (Ops[j].Op == X) return j; - // Scan backwards + // Scan backwards. for (unsigned j = i-1; j != ~0U && Ops[j].Rank == XRank; --j) if (Ops[j].Op == X) return j; @@ -494,66 +903,100 @@ static unsigned FindInOperandList(std::vector &Ops, unsigned i, /// EmitAddTreeOfValues - Emit a tree of add instructions, summing Ops together /// and returning the result. Insert the tree before I. -static Value *EmitAddTreeOfValues(Instruction *I, SmallVectorImpl &Ops){ +static Value *EmitAddTreeOfValues(Instruction *I, + SmallVectorImpl &Ops){ if (Ops.size() == 1) return Ops.back(); - + Value *V1 = Ops.back(); Ops.pop_back(); Value *V2 = EmitAddTreeOfValues(I, Ops); return BinaryOperator::CreateAdd(V2, V1, "tmp", I); } -/// RemoveFactorFromExpression - If V is an expression tree that is a +/// RemoveFactorFromExpression - If V is an expression tree that is a /// multiplication sequence, and if this sequence contains a multiply by Factor, /// remove Factor from the tree and return the new tree. Value *Reassociate::RemoveFactorFromExpression(Value *V, Value *Factor) { BinaryOperator *BO = isReassociableOp(V, Instruction::Mul); if (!BO) return 0; - - std::vector Factors; - LinearizeExprTree(BO, Factors); + + SmallVector Tree; + MadeChange |= LinearizeExprTree(BO, Tree); + SmallVector Factors; + Factors.reserve(Tree.size()); + for (unsigned i = 0, e = Tree.size(); i != e; ++i) { + RepeatedValue E = Tree[i]; + Factors.append(E.second.getZExtValue(), + ValueEntry(getRank(E.first), E.first)); + } bool FoundFactor = false; - for (unsigned i = 0, e = Factors.size(); i != e; ++i) + bool NeedsNegate = false; + for (unsigned i = 0, e = Factors.size(); i != e; ++i) { if (Factors[i].Op == Factor) { FoundFactor = true; Factors.erase(Factors.begin()+i); break; } + + // If this is a negative version of this factor, remove it. + if (ConstantInt *FC1 = dyn_cast(Factor)) + if (ConstantInt *FC2 = dyn_cast(Factors[i].Op)) + if (FC1->getValue() == -FC2->getValue()) { + FoundFactor = NeedsNegate = true; + Factors.erase(Factors.begin()+i); + break; + } + } + if (!FoundFactor) { // Make sure to restore the operands to the expression tree. RewriteExprTree(BO, Factors); return 0; } - - if (Factors.size() == 1) return Factors[0].Op; - - RewriteExprTree(BO, Factors); - return BO; + + BasicBlock::iterator InsertPt = BO; ++InsertPt; + + // If this was just a single multiply, remove the multiply and return the only + // remaining operand. + if (Factors.size() == 1) { + RedoInsts.insert(BO); + V = Factors[0].Op; + } else { + RewriteExprTree(BO, Factors); + V = BO; + } + + if (NeedsNegate) + V = BinaryOperator::CreateNeg(V, "neg", InsertPt); + + return V; } /// FindSingleUseMultiplyFactors - If V is a single-use multiply, recursively /// add its operands as factors, otherwise add V to the list of factors. +/// +/// Ops is the top-level list of add operands we're trying to factor. static void FindSingleUseMultiplyFactors(Value *V, - SmallVectorImpl &Factors) { - BinaryOperator *BO; - if ((!V->hasOneUse() && !V->use_empty()) || - !(BO = dyn_cast(V)) || - BO->getOpcode() != Instruction::Mul) { + SmallVectorImpl &Factors, + const SmallVectorImpl &Ops) { + BinaryOperator *BO = isReassociableOp(V, Instruction::Mul); + if (!BO) { Factors.push_back(V); return; } - + // Otherwise, add the LHS and RHS to the list of factors. - FindSingleUseMultiplyFactors(BO->getOperand(1), Factors); - FindSingleUseMultiplyFactors(BO->getOperand(0), Factors); + FindSingleUseMultiplyFactors(BO->getOperand(1), Factors, Ops); + FindSingleUseMultiplyFactors(BO->getOperand(0), Factors, Ops); } /// OptimizeAndOrXor - Optimize a series of operands to an 'and', 'or', or 'xor' /// instruction. This optimizes based on identities. If it can be reduced to /// a single Value, it is returned, otherwise the Ops list is mutated as /// necessary. -static Value *OptimizeAndOrXor(unsigned Opcode, std::vector &Ops) { +static Value *OptimizeAndOrXor(unsigned Opcode, + SmallVectorImpl &Ops) { // Scan the operand lists looking for X and ~X pairs, along with X,X pairs. // If we find any, we can simplify the expression. X&~X == 0, X|~X == -1. for (unsigned i = 0, e = Ops.size(); i != e; ++i) { @@ -563,38 +1006,35 @@ static Value *OptimizeAndOrXor(unsigned Opcode, std::vector &Ops) { Value *X = BinaryOperator::getNotArgument(Ops[i].Op); unsigned FoundX = FindInOperandList(Ops, i, X); if (FoundX != i) { - if (Opcode == Instruction::And) { // ...&X&~X = 0 - ++NumAnnihil; + if (Opcode == Instruction::And) // ...&X&~X = 0 return Constant::getNullValue(X->getType()); - } - - if (Opcode == Instruction::Or) { // ...|X|~X = -1 - ++NumAnnihil; + + if (Opcode == Instruction::Or) // ...|X|~X = -1 return Constant::getAllOnesValue(X->getType()); - } } } - + // Next, check for duplicate pairs of values, which we assume are next to // each other, due to our sorting criteria. assert(i < Ops.size()); if (i+1 != Ops.size() && Ops[i+1].Op == Ops[i].Op) { if (Opcode == Instruction::And || Opcode == Instruction::Or) { - // Drop duplicate values. + // Drop duplicate values for And and Or. Ops.erase(Ops.begin()+i); --i; --e; ++NumAnnihil; - } else { - assert(Opcode == Instruction::Xor); - if (e == 2) { - ++NumAnnihil; - return Constant::getNullValue(Ops[0].Op->getType()); - } - // ... X^X -> ... - Ops.erase(Ops.begin()+i, Ops.begin()+i+2); - i -= 1; e -= 2; - ++NumAnnihil; + continue; } + + // Drop pairs of values for Xor. + assert(Opcode == Instruction::Xor); + if (e == 2) + return Constant::getNullValue(Ops[0].Op->getType()); + + // Y ^ X^X -> Y + Ops.erase(Ops.begin()+i, Ops.begin()+i+2); + i -= 1; e -= 2; + ++NumAnnihil; } } return 0; @@ -603,26 +1043,66 @@ static Value *OptimizeAndOrXor(unsigned Opcode, std::vector &Ops) { /// OptimizeAdd - Optimize a series of operands to an 'add' instruction. This /// optimizes based on identities. If it can be reduced to a single Value, it /// is returned, otherwise the Ops list is mutated as necessary. -Value *Reassociate::OptimizeAdd(std::vector &Ops) { +Value *Reassociate::OptimizeAdd(Instruction *I, + SmallVectorImpl &Ops) { // Scan the operand lists looking for X and -X pairs. If we find any, we - // can simplify the expression. X+-X == 0. + // can simplify the expression. X+-X == 0. While we're at it, scan for any + // duplicates. We want to canonicalize Y+Y+Y+Z -> 3*Y+Z. + // + // TODO: We could handle "X + ~X" -> "-1" if we wanted, since "-X = ~X+1". + // for (unsigned i = 0, e = Ops.size(); i != e; ++i) { - assert(i < Ops.size()); + Value *TheOp = Ops[i].Op; + // Check to see if we've seen this operand before. If so, we factor all + // instances of the operand together. Due to our sorting criteria, we know + // that these need to be next to each other in the vector. + if (i+1 != Ops.size() && Ops[i+1].Op == TheOp) { + // Rescan the list, remove all instances of this operand from the expr. + unsigned NumFound = 0; + do { + Ops.erase(Ops.begin()+i); + ++NumFound; + } while (i != Ops.size() && Ops[i].Op == TheOp); + + DEBUG(errs() << "\nFACTORING [" << NumFound << "]: " << *TheOp << '\n'); + ++NumFactor; + + // Insert a new multiply. + Value *Mul = ConstantInt::get(cast(I->getType()), NumFound); + Mul = BinaryOperator::CreateMul(TheOp, Mul, "factor", I); + + // Now that we have inserted a multiply, optimize it. This allows us to + // handle cases that require multiple factoring steps, such as this: + // (X*2) + (X*2) + (X*2) -> (X*2)*3 -> X*6 + RedoInsts.insert(cast(Mul)); + + // If every add operand was a duplicate, return the multiply. + if (Ops.empty()) + return Mul; + + // Otherwise, we had some input that didn't have the dupe, such as + // "A + A + B" -> "A*2 + B". Add the new multiply to the list of + // things being added by this operation. + Ops.insert(Ops.begin(), ValueEntry(getRank(Mul), Mul)); + + --i; + e = Ops.size(); + continue; + } + // Check for X and -X in the operand list. - if (!BinaryOperator::isNeg(Ops[i].Op)) + if (!BinaryOperator::isNeg(TheOp)) continue; - - Value *X = BinaryOperator::getNegArgument(Ops[i].Op); + + Value *X = BinaryOperator::getNegArgument(TheOp); unsigned FoundX = FindInOperandList(Ops, i, X); if (FoundX == i) continue; - + // Remove X and -X from the operand list. - if (Ops.size() == 2) { - ++NumAnnihil; + if (Ops.size() == 2) return Constant::getNullValue(X->getType()); - } - + Ops.erase(Ops.begin()+i); if (i < FoundX) --FoundX; @@ -633,233 +1113,479 @@ Value *Reassociate::OptimizeAdd(std::vector &Ops) { --i; // Revisit element. e -= 2; // Removed two elements. } + + // Scan the operand list, checking to see if there are any common factors + // between operands. Consider something like A*A+A*B*C+D. We would like to + // reassociate this to A*(A+B*C)+D, which reduces the number of multiplies. + // To efficiently find this, we count the number of times a factor occurs + // for any ADD operands that are MULs. + DenseMap FactorOccurrences; + + // Keep track of each multiply we see, to avoid triggering on (X*4)+(X*4) + // where they are actually the same multiply. + unsigned MaxOcc = 0; + Value *MaxOccVal = 0; + for (unsigned i = 0, e = Ops.size(); i != e; ++i) { + BinaryOperator *BOp = isReassociableOp(Ops[i].Op, Instruction::Mul); + if (!BOp) + continue; + + // Compute all of the factors of this added value. + SmallVector Factors; + FindSingleUseMultiplyFactors(BOp, Factors, Ops); + assert(Factors.size() > 1 && "Bad linearize!"); + + // Add one to FactorOccurrences for each unique factor in this op. + SmallPtrSet Duplicates; + for (unsigned i = 0, e = Factors.size(); i != e; ++i) { + Value *Factor = Factors[i]; + if (!Duplicates.insert(Factor)) continue; + + unsigned Occ = ++FactorOccurrences[Factor]; + if (Occ > MaxOcc) { MaxOcc = Occ; MaxOccVal = Factor; } + + // If Factor is a negative constant, add the negated value as a factor + // because we can percolate the negate out. Watch for minint, which + // cannot be positivified. + if (ConstantInt *CI = dyn_cast(Factor)) + if (CI->isNegative() && !CI->isMinValue(true)) { + Factor = ConstantInt::get(CI->getContext(), -CI->getValue()); + assert(!Duplicates.count(Factor) && + "Shouldn't have two constant factors, missed a canonicalize"); + + unsigned Occ = ++FactorOccurrences[Factor]; + if (Occ > MaxOcc) { MaxOcc = Occ; MaxOccVal = Factor; } + } + } + } + + // If any factor occurred more than one time, we can pull it out. + if (MaxOcc > 1) { + DEBUG(errs() << "\nFACTORING [" << MaxOcc << "]: " << *MaxOccVal << '\n'); + ++NumFactor; + + // Create a new instruction that uses the MaxOccVal twice. If we don't do + // this, we could otherwise run into situations where removing a factor + // from an expression will drop a use of maxocc, and this can cause + // RemoveFactorFromExpression on successive values to behave differently. + Instruction *DummyInst = BinaryOperator::CreateAdd(MaxOccVal, MaxOccVal); + SmallVector NewMulOps; + for (unsigned i = 0; i != Ops.size(); ++i) { + // Only try to remove factors from expressions we're allowed to. + BinaryOperator *BOp = isReassociableOp(Ops[i].Op, Instruction::Mul); + if (!BOp) + continue; + + if (Value *V = RemoveFactorFromExpression(Ops[i].Op, MaxOccVal)) { + // The factorized operand may occur several times. Convert them all in + // one fell swoop. + for (unsigned j = Ops.size(); j != i;) { + --j; + if (Ops[j].Op == Ops[i].Op) { + NewMulOps.push_back(V); + Ops.erase(Ops.begin()+j); + } + } + --i; + } + } + + // No need for extra uses anymore. + delete DummyInst; + + unsigned NumAddedValues = NewMulOps.size(); + Value *V = EmitAddTreeOfValues(I, NewMulOps); + + // Now that we have inserted the add tree, optimize it. This allows us to + // handle cases that require multiple factoring steps, such as this: + // A*A*B + A*A*C --> A*(A*B+A*C) --> A*(A*(B+C)) + assert(NumAddedValues > 1 && "Each occurrence should contribute a value"); + (void)NumAddedValues; + if (Instruction *VI = dyn_cast(V)) + RedoInsts.insert(VI); + + // Create the multiply. + Instruction *V2 = BinaryOperator::CreateMul(V, MaxOccVal, "tmp", I); + + // Rerun associate on the multiply in case the inner expression turned into + // a multiply. We want to make sure that we keep things in canonical form. + RedoInsts.insert(V2); + + // If every add operand included the factor (e.g. "A*B + A*C"), then the + // entire result expression is just the multiply "A*(B+C)". + if (Ops.empty()) + return V2; + + // Otherwise, we had some input that didn't have the factor, such as + // "A*B + A*C + D" -> "A*(B+C) + D". Add the new multiply to the list of + // things being added by this operation. + Ops.insert(Ops.begin(), ValueEntry(getRank(V2), V2)); + } + + return 0; +} + +namespace { + /// \brief Predicate tests whether a ValueEntry's op is in a map. + struct IsValueInMap { + const DenseMap ⤅ + + IsValueInMap(const DenseMap &Map) : Map(Map) {} + + bool operator()(const ValueEntry &Entry) { + return Map.find(Entry.Op) != Map.end(); + } + }; +} + +/// \brief Build up a vector of value/power pairs factoring a product. +/// +/// Given a series of multiplication operands, build a vector of factors and +/// the powers each is raised to when forming the final product. Sort them in +/// the order of descending power. +/// +/// (x*x) -> [(x, 2)] +/// ((x*x)*x) -> [(x, 3)] +/// ((((x*y)*x)*y)*x) -> [(x, 3), (y, 2)] +/// +/// \returns Whether any factors have a power greater than one. +bool Reassociate::collectMultiplyFactors(SmallVectorImpl &Ops, + SmallVectorImpl &Factors) { + // FIXME: Have Ops be (ValueEntry, Multiplicity) pairs, simplifying this. + // Compute the sum of powers of simplifiable factors. + unsigned FactorPowerSum = 0; + for (unsigned Idx = 1, Size = Ops.size(); Idx < Size; ++Idx) { + Value *Op = Ops[Idx-1].Op; + + // Count the number of occurrences of this value. + unsigned Count = 1; + for (; Idx < Size && Ops[Idx].Op == Op; ++Idx) + ++Count; + // Track for simplification all factors which occur 2 or more times. + if (Count > 1) + FactorPowerSum += Count; + } + + // We can only simplify factors if the sum of the powers of our simplifiable + // factors is 4 or higher. When that is the case, we will *always* have + // a simplification. This is an important invariant to prevent cyclicly + // trying to simplify already minimal formations. + if (FactorPowerSum < 4) + return false; + + // Now gather the simplifiable factors, removing them from Ops. + FactorPowerSum = 0; + for (unsigned Idx = 1; Idx < Ops.size(); ++Idx) { + Value *Op = Ops[Idx-1].Op; + + // Count the number of occurrences of this value. + unsigned Count = 1; + for (; Idx < Ops.size() && Ops[Idx].Op == Op; ++Idx) + ++Count; + if (Count == 1) + continue; + // Move an even number of occurrences to Factors. + Count &= ~1U; + Idx -= Count; + FactorPowerSum += Count; + Factors.push_back(Factor(Op, Count)); + Ops.erase(Ops.begin()+Idx, Ops.begin()+Idx+Count); + } + + // None of the adjustments above should have reduced the sum of factor powers + // below our mininum of '4'. + assert(FactorPowerSum >= 4); + + std::sort(Factors.begin(), Factors.end(), Factor::PowerDescendingSorter()); + return true; +} + +/// \brief Build a tree of multiplies, computing the product of Ops. +static Value *buildMultiplyTree(IRBuilder<> &Builder, + SmallVectorImpl &Ops) { + if (Ops.size() == 1) + return Ops.back(); + + Value *LHS = Ops.pop_back_val(); + do { + LHS = Builder.CreateMul(LHS, Ops.pop_back_val()); + } while (!Ops.empty()); + + return LHS; +} + +/// \brief Build a minimal multiplication DAG for (a^x)*(b^y)*(c^z)*... +/// +/// Given a vector of values raised to various powers, where no two values are +/// equal and the powers are sorted in decreasing order, compute the minimal +/// DAG of multiplies to compute the final product, and return that product +/// value. +Value *Reassociate::buildMinimalMultiplyDAG(IRBuilder<> &Builder, + SmallVectorImpl &Factors) { + assert(Factors[0].Power); + SmallVector OuterProduct; + for (unsigned LastIdx = 0, Idx = 1, Size = Factors.size(); + Idx < Size && Factors[Idx].Power > 0; ++Idx) { + if (Factors[Idx].Power != Factors[LastIdx].Power) { + LastIdx = Idx; + continue; + } + + // We want to multiply across all the factors with the same power so that + // we can raise them to that power as a single entity. Build a mini tree + // for that. + SmallVector InnerProduct; + InnerProduct.push_back(Factors[LastIdx].Base); + do { + InnerProduct.push_back(Factors[Idx].Base); + ++Idx; + } while (Idx < Size && Factors[Idx].Power == Factors[LastIdx].Power); + + // Reset the base value of the first factor to the new expression tree. + // We'll remove all the factors with the same power in a second pass. + Value *M = Factors[LastIdx].Base = buildMultiplyTree(Builder, InnerProduct); + if (Instruction *MI = dyn_cast(M)) + RedoInsts.insert(MI); + + LastIdx = Idx; + } + // Unique factors with equal powers -- we've folded them into the first one's + // base. + Factors.erase(std::unique(Factors.begin(), Factors.end(), + Factor::PowerEqual()), + Factors.end()); + + // Iteratively collect the base of each factor with an add power into the + // outer product, and halve each power in preparation for squaring the + // expression. + for (unsigned Idx = 0, Size = Factors.size(); Idx != Size; ++Idx) { + if (Factors[Idx].Power & 1) + OuterProduct.push_back(Factors[Idx].Base); + Factors[Idx].Power >>= 1; + } + if (Factors[0].Power) { + Value *SquareRoot = buildMinimalMultiplyDAG(Builder, Factors); + OuterProduct.push_back(SquareRoot); + OuterProduct.push_back(SquareRoot); + } + if (OuterProduct.size() == 1) + return OuterProduct.front(); + + Value *V = buildMultiplyTree(Builder, OuterProduct); + return V; +} + +Value *Reassociate::OptimizeMul(BinaryOperator *I, + SmallVectorImpl &Ops) { + // We can only optimize the multiplies when there is a chain of more than + // three, such that a balanced tree might require fewer total multiplies. + if (Ops.size() < 4) + return 0; + + // Try to turn linear trees of multiplies without other uses of the + // intermediate stages into minimal multiply DAGs with perfect sub-expression + // re-use. + SmallVector Factors; + if (!collectMultiplyFactors(Ops, Factors)) + return 0; // All distinct factors, so nothing left for us to do. + + IRBuilder<> Builder(I); + Value *V = buildMinimalMultiplyDAG(Builder, Factors); + if (Ops.empty()) + return V; + + ValueEntry NewEntry = ValueEntry(getRank(V), V); + Ops.insert(std::lower_bound(Ops.begin(), Ops.end(), NewEntry), NewEntry); return 0; } Value *Reassociate::OptimizeExpression(BinaryOperator *I, - std::vector &Ops) { + SmallVectorImpl &Ops) { // Now that we have the linearized expression tree, try to optimize it. // Start by folding any constants that we found. - bool IterateOptimization = false; - if (Ops.size() == 1) return Ops[0].Op; - + Constant *Cst = 0; unsigned Opcode = I->getOpcode(); - - if (Constant *V1 = dyn_cast(Ops[Ops.size()-2].Op)) - if (Constant *V2 = dyn_cast(Ops.back().Op)) { - Ops.pop_back(); - Ops.back().Op = ConstantExpr::get(Opcode, V1, V2); - return OptimizeExpression(I, Ops); - } + while (!Ops.empty() && isa(Ops.back().Op)) { + Constant *C = cast(Ops.pop_back_val().Op); + Cst = Cst ? ConstantExpr::get(Opcode, C, Cst) : C; + } + // If there was nothing but constants then we are done. + if (Ops.empty()) + return Cst; + + // Put the combined constant back at the end of the operand list, except if + // there is no point. For example, an add of 0 gets dropped here, while a + // multiplication by zero turns the whole expression into zero. + if (Cst && Cst != ConstantExpr::getBinOpIdentity(Opcode, I->getType())) { + if (Cst == ConstantExpr::getBinOpAbsorber(Opcode, I->getType())) + return Cst; + Ops.push_back(ValueEntry(0, Cst)); + } - // Check for destructive annihilation due to a constant being used. - if (ConstantInt *CstVal = dyn_cast(Ops.back().Op)) - switch (Opcode) { - default: break; - case Instruction::And: - if (CstVal->isZero()) { // ... & 0 -> 0 - ++NumAnnihil; - return CstVal; - } - if (CstVal->isAllOnesValue()) // ... & -1 -> ... - Ops.pop_back(); - break; - case Instruction::Mul: - if (CstVal->isZero()) { // ... * 0 -> 0 - ++NumAnnihil; - return CstVal; - } - - if (cast(CstVal)->isOne()) - Ops.pop_back(); // ... * 1 -> ... - break; - case Instruction::Or: - if (CstVal->isAllOnesValue()) { // ... | -1 -> -1 - ++NumAnnihil; - return CstVal; - } - // FALLTHROUGH! - case Instruction::Add: - case Instruction::Xor: - if (CstVal->isZero()) // ... [|^+] 0 -> ... - Ops.pop_back(); - break; - } if (Ops.size() == 1) return Ops[0].Op; // Handle destructive annihilation due to identities between elements in the // argument list here. + unsigned NumOps = Ops.size(); switch (Opcode) { default: break; case Instruction::And: case Instruction::Or: - case Instruction::Xor: { - unsigned NumOps = Ops.size(); + case Instruction::Xor: if (Value *Result = OptimizeAndOrXor(Opcode, Ops)) return Result; - IterateOptimization |= Ops.size() != NumOps; break; - } - case Instruction::Add: { - unsigned NumOps = Ops.size(); - if (Value *Result = OptimizeAdd(Ops)) + case Instruction::Add: + if (Value *Result = OptimizeAdd(I, Ops)) return Result; - IterateOptimization |= Ops.size() != NumOps; - } - - // Scan the operand list, checking to see if there are any common factors - // between operands. Consider something like A*A+A*B*C+D. We would like to - // reassociate this to A*(A+B*C)+D, which reduces the number of multiplies. - // To efficiently find this, we count the number of times a factor occurs - // for any ADD operands that are MULs. - DenseMap FactorOccurrences; - unsigned MaxOcc = 0; - Value *MaxOccVal = 0; - for (unsigned i = 0, e = Ops.size(); i != e; ++i) { - BinaryOperator *BOp = dyn_cast(Ops[i].Op); - if (BOp == 0 || BOp->getOpcode() != Instruction::Mul || !BOp->use_empty()) - continue; - - // Compute all of the factors of this added value. - SmallVector Factors; - FindSingleUseMultiplyFactors(BOp, Factors); - assert(Factors.size() > 1 && "Bad linearize!"); - - // Add one to FactorOccurrences for each unique factor in this op. - if (Factors.size() == 2) { - unsigned Occ = ++FactorOccurrences[Factors[0]]; - if (Occ > MaxOcc) { MaxOcc = Occ; MaxOccVal = Factors[0]; } - if (Factors[0] != Factors[1]) { // Don't double count A*A. - Occ = ++FactorOccurrences[Factors[1]]; - if (Occ > MaxOcc) { MaxOcc = Occ; MaxOccVal = Factors[1]; } - } - } else { - SmallPtrSet Duplicates; - for (unsigned i = 0, e = Factors.size(); i != e; ++i) { - if (!Duplicates.insert(Factors[i])) continue; - - unsigned Occ = ++FactorOccurrences[Factors[i]]; - if (Occ > MaxOcc) { MaxOcc = Occ; MaxOccVal = Factors[i]; } - } - } - } + break; - // If any factor occurred more than one time, we can pull it out. - if (MaxOcc > 1) { - DEBUG(errs() << "\nFACTORING [" << MaxOcc << "]: " << *MaxOccVal << "\n"); - - // Create a new instruction that uses the MaxOccVal twice. If we don't do - // this, we could otherwise run into situations where removing a factor - // from an expression will drop a use of maxocc, and this can cause - // RemoveFactorFromExpression on successive values to behave differently. - Instruction *DummyInst = BinaryOperator::CreateAdd(MaxOccVal, MaxOccVal); - SmallVector NewMulOps; - for (unsigned i = 0, e = Ops.size(); i != e; ++i) { - if (Value *V = RemoveFactorFromExpression(Ops[i].Op, MaxOccVal)) { - NewMulOps.push_back(V); - Ops.erase(Ops.begin()+i); - --i; --e; - } - } - - // No need for extra uses anymore. - delete DummyInst; - - unsigned NumAddedValues = NewMulOps.size(); - Value *V = EmitAddTreeOfValues(I, NewMulOps); - Value *V2 = BinaryOperator::CreateMul(V, MaxOccVal, "tmp", I); - - // Now that we have inserted V and its sole use, optimize it. This allows - // us to handle cases that require multiple factoring steps, such as this: - // A*A*B + A*A*C --> A*(A*B+A*C) --> A*(A*(B+C)) - if (NumAddedValues > 1) - ReassociateExpression(cast(V)); - - ++NumFactor; - - if (Ops.empty()) - return V2; - - // Add the new value to the list of things being added. - Ops.insert(Ops.begin(), ValueEntry(getRank(V2), V2)); - - // Rewrite the tree so that there is now a use of V. - RewriteExprTree(I, Ops); - return OptimizeExpression(I, Ops); - } + case Instruction::Mul: + if (Value *Result = OptimizeMul(I, Ops)) + return Result; break; - //case Instruction::Mul: } - if (IterateOptimization) + if (Ops.size() != NumOps) return OptimizeExpression(I, Ops); return 0; } +/// EraseInst - Zap the given instruction, adding interesting operands to the +/// work list. +void Reassociate::EraseInst(Instruction *I) { + assert(isInstructionTriviallyDead(I) && "Trivially dead instructions only!"); + SmallVector Ops(I->op_begin(), I->op_end()); + // Erase the dead instruction. + ValueRankMap.erase(I); + RedoInsts.remove(I); + I->eraseFromParent(); + // Optimize its operands. + SmallPtrSet Visited; // Detect self-referential nodes. + for (unsigned i = 0, e = Ops.size(); i != e; ++i) + if (Instruction *Op = dyn_cast(Ops[i])) { + // If this is a node in an expression tree, climb to the expression root + // and add that since that's where optimization actually happens. + unsigned Opcode = Op->getOpcode(); + while (Op->hasOneUse() && Op->use_back()->getOpcode() == Opcode && + Visited.insert(Op)) + Op = Op->use_back(); + RedoInsts.insert(Op); + } +} -/// ReassociateBB - Inspect all of the instructions in this basic block, -/// reassociating them as we go. -void Reassociate::ReassociateBB(BasicBlock *BB) { - for (BasicBlock::iterator BBI = BB->begin(); BBI != BB->end(); ) { - Instruction *BI = BBI++; - if (BI->getOpcode() == Instruction::Shl && - isa(BI->getOperand(1))) - if (Instruction *NI = ConvertShiftToMul(BI, ValueRankMap)) { - MadeChange = true; - BI = NI; - } +/// OptimizeInst - Inspect and optimize the given instruction. Note that erasing +/// instructions is not allowed. +void Reassociate::OptimizeInst(Instruction *I) { + // Only consider operations that we understand. + if (!isa(I)) + return; + + if (I->getOpcode() == Instruction::Shl && + isa(I->getOperand(1))) + // If an operand of this shift is a reassociable multiply, or if the shift + // is used by a reassociable multiply or add, turn into a multiply. + if (isReassociableOp(I->getOperand(0), Instruction::Mul) || + (I->hasOneUse() && + (isReassociableOp(I->use_back(), Instruction::Mul) || + isReassociableOp(I->use_back(), Instruction::Add)))) { + Instruction *NI = ConvertShiftToMul(I); + RedoInsts.insert(I); + MadeChange = true; + I = NI; + } - // Reject cases where it is pointless to do this. - if (!isa(BI) || BI->getType()->isFloatingPoint() || - isa(BI->getType())) - continue; // Floating point ops are not associative. + // Floating point binary operators are not associative, but we can still + // commute (some) of them, to canonicalize the order of their operands. + // This can potentially expose more CSE opportunities, and makes writing + // other transformations simpler. + if ((I->getType()->isFloatingPointTy() || I->getType()->isVectorTy())) { + // FAdd and FMul can be commuted. + if (I->getOpcode() != Instruction::FMul && + I->getOpcode() != Instruction::FAdd) + return; + + Value *LHS = I->getOperand(0); + Value *RHS = I->getOperand(1); + unsigned LHSRank = getRank(LHS); + unsigned RHSRank = getRank(RHS); + + // Sort the operands by rank. + if (RHSRank < LHSRank) { + I->setOperand(0, RHS); + I->setOperand(1, LHS); + } + + return; + } - // If this is a subtract instruction which is not already in negate form, - // see if we can convert it to X+-Y. - if (BI->getOpcode() == Instruction::Sub) { - if (ShouldBreakUpSubtract(BI)) { - BI = BreakUpSubtract(BI, ValueRankMap); + // Do not reassociate boolean (i1) expressions. We want to preserve the + // original order of evaluation for short-circuited comparisons that + // SimplifyCFG has folded to AND/OR expressions. If the expression + // is not further optimized, it is likely to be transformed back to a + // short-circuited form for code gen, and the source order may have been + // optimized for the most likely conditions. + if (I->getType()->isIntegerTy(1)) + return; + + // If this is a subtract instruction which is not already in negate form, + // see if we can convert it to X+-Y. + if (I->getOpcode() == Instruction::Sub) { + if (ShouldBreakUpSubtract(I)) { + Instruction *NI = BreakUpSubtract(I); + RedoInsts.insert(I); + MadeChange = true; + I = NI; + } else if (BinaryOperator::isNeg(I)) { + // Otherwise, this is a negation. See if the operand is a multiply tree + // and if this is not an inner node of a multiply tree. + if (isReassociableOp(I->getOperand(1), Instruction::Mul) && + (!I->hasOneUse() || + !isReassociableOp(I->use_back(), Instruction::Mul))) { + Instruction *NI = LowerNegateToMultiply(I); + RedoInsts.insert(I); MadeChange = true; - } else if (BinaryOperator::isNeg(BI)) { - // Otherwise, this is a negation. See if the operand is a multiply tree - // and if this is not an inner node of a multiply tree. - if (isReassociableOp(BI->getOperand(1), Instruction::Mul) && - (!BI->hasOneUse() || - !isReassociableOp(BI->use_back(), Instruction::Mul))) { - BI = LowerNegateToMultiply(BI, ValueRankMap); - MadeChange = true; - } + I = NI; } } + } - // If this instruction is a commutative binary operator, process it. - if (!BI->isAssociative()) continue; - BinaryOperator *I = cast(BI); + // If this instruction is an associative binary operator, process it. + if (!I->isAssociative()) return; + BinaryOperator *BO = cast(I); - // If this is an interior node of a reassociable tree, ignore it until we - // get to the root of the tree, to avoid N^2 analysis. - if (I->hasOneUse() && isReassociableOp(I->use_back(), I->getOpcode())) - continue; + // If this is an interior node of a reassociable tree, ignore it until we + // get to the root of the tree, to avoid N^2 analysis. + unsigned Opcode = BO->getOpcode(); + if (BO->hasOneUse() && BO->use_back()->getOpcode() == Opcode) + return; - // If this is an add tree that is used by a sub instruction, ignore it - // until we process the subtract. - if (I->hasOneUse() && I->getOpcode() == Instruction::Add && - cast(I->use_back())->getOpcode() == Instruction::Sub) - continue; + // If this is an add tree that is used by a sub instruction, ignore it + // until we process the subtract. + if (BO->hasOneUse() && BO->getOpcode() == Instruction::Add && + cast(BO->use_back())->getOpcode() == Instruction::Sub) + return; - ReassociateExpression(I); - } + ReassociateExpression(BO); } void Reassociate::ReassociateExpression(BinaryOperator *I) { - - // First, walk the expression tree, linearizing the tree, collecting - std::vector Ops; - LinearizeExprTree(I, Ops); - - DEBUG(errs() << "RAIn:\t"; PrintOps(I, Ops); errs() << "\n"); - + + // First, walk the expression tree, linearizing the tree, collecting the + // operand information. + SmallVector Tree; + MadeChange |= LinearizeExprTree(I, Tree); + SmallVector Ops; + Ops.reserve(Tree.size()); + for (unsigned i = 0, e = Tree.size(); i != e; ++i) { + RepeatedValue E = Tree[i]; + Ops.append(E.second.getZExtValue(), + ValueEntry(getRank(E.first), E.first)); + } + + DEBUG(dbgs() << "RAIn:\t"; PrintOps(I, Ops); dbgs() << '\n'); + // Now that we have linearized the tree to a list and have gathered all of // the operands and their ranks, sort the operands by their rank. Use a // stable_sort so that values with equal ranks will have their relative @@ -867,18 +1593,24 @@ void Reassociate::ReassociateExpression(BinaryOperator *I) { // this sorts so that the highest ranking values end up at the beginning of // the vector. std::stable_sort(Ops.begin(), Ops.end()); - + // OptimizeExpression - Now that we have the expression tree in a convenient // sorted form, optimize it globally if possible. if (Value *V = OptimizeExpression(I, Ops)) { + if (V == I) + // Self-referential expression in unreachable code. + return; // This expression tree simplified to something that isn't a tree, // eliminate it. - DEBUG(errs() << "Reassoc to scalar: " << *V << "\n"); + DEBUG(dbgs() << "Reassoc to scalar: " << *V << '\n'); I->replaceAllUsesWith(V); - RemoveDeadBinaryOp(I); + if (Instruction *VI = dyn_cast(V)) + VI->setDebugLoc(I->getDebugLoc()); + RedoInsts.insert(I); + ++NumAnnihil; return; } - + // We want to sink immediates as deeply as possible except in the case where // this is a multiply tree used only by an add, and the immediate is a -1. // In this case we reassociate to put the negation on the outside so that we @@ -887,36 +1619,60 @@ void Reassociate::ReassociateExpression(BinaryOperator *I) { cast(I->use_back())->getOpcode() == Instruction::Add && isa(Ops.back().Op) && cast(Ops.back().Op)->isAllOnesValue()) { - Ops.insert(Ops.begin(), Ops.back()); - Ops.pop_back(); + ValueEntry Tmp = Ops.pop_back_val(); + Ops.insert(Ops.begin(), Tmp); } - - DEBUG(errs() << "RAOut:\t"; PrintOps(I, Ops); errs() << "\n"); - + + DEBUG(dbgs() << "RAOut:\t"; PrintOps(I, Ops); dbgs() << '\n'); + if (Ops.size() == 1) { + if (Ops[0].Op == I) + // Self-referential expression in unreachable code. + return; + // This expression tree simplified to something that isn't a tree, // eliminate it. I->replaceAllUsesWith(Ops[0].Op); - RemoveDeadBinaryOp(I); - } else { - // Now that we ordered and optimized the expressions, splat them back into - // the expression tree, removing any unneeded nodes. - RewriteExprTree(I, Ops); + if (Instruction *OI = dyn_cast(Ops[0].Op)) + OI->setDebugLoc(I->getDebugLoc()); + RedoInsts.insert(I); + return; } -} + // Now that we ordered and optimized the expressions, splat them back into + // the expression tree, removing any unneeded nodes. + RewriteExprTree(I, Ops); +} bool Reassociate::runOnFunction(Function &F) { - // Recalculate the rank map for F + // Calculate the rank map for F BuildRankMap(F); MadeChange = false; - for (Function::iterator FI = F.begin(), FE = F.end(); FI != FE; ++FI) - ReassociateBB(FI); + for (Function::iterator BI = F.begin(), BE = F.end(); BI != BE; ++BI) { + // Optimize every instruction in the basic block. + for (BasicBlock::iterator II = BI->begin(), IE = BI->end(); II != IE; ) + if (isInstructionTriviallyDead(II)) { + EraseInst(II++); + } else { + OptimizeInst(II); + assert(II->getParent() == BI && "Moved to a different block!"); + ++II; + } - // We are done with the rank map... + // If this produced extra instructions to optimize, handle them now. + while (!RedoInsts.empty()) { + Instruction *I = RedoInsts.pop_back_val(); + if (isInstructionTriviallyDead(I)) + EraseInst(I); + else + OptimizeInst(I); + } + } + + // We are done with the rank map. RankMap.clear(); ValueRankMap.clear(); + return MadeChange; } -