//===- Reassociate.cpp - Reassociate binary expressions -------------------===// // // The LLVM Compiler Infrastructure // // This file is distributed under the University of Illinois Open Source // License. See LICENSE.TXT for details. // //===----------------------------------------------------------------------===// // // This pass reassociates commutative expressions in an order that is designed // to promote better constant propagation, GCSE, LICM, PRE, etc. // // For example: 4 + (x + 5) -> x + (4 + 5) // // In the implementation of this algorithm, constants are assigned rank = 0, // function arguments are rank = 1, and other values are assigned ranks // corresponding to the reverse post order traversal of current function // (starting at 2), which effectively gives values in deep loops higher rank // than values not in loops. // //===----------------------------------------------------------------------===// #define DEBUG_TYPE "reassociate" #include "llvm/Transforms/Scalar.h" #include "llvm/Transforms/Utils/Local.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/Assembly/Writer.h" #include "llvm/Support/CFG.h" #include "llvm/Support/IRBuilder.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/STLExtras.h" #include "llvm/ADT/Statistic.h" #include "llvm/ADT/DenseMap.h" #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"); namespace { struct ValueEntry { unsigned Rank; Value *Op; ValueEntry(unsigned R, Value *O) : Rank(R), Op(O) {} }; inline bool operator<(const ValueEntry &LHS, const ValueEntry &RHS) { return LHS.Rank > RHS.Rank; // Sort so that highest rank goes to start. } } #ifndef NDEBUG /// PrintOps - Print out the expression identified in the Ops list. /// static void PrintOps(Instruction *I, const SmallVectorImpl &Ops) { Module *M = I->getParent()->getParent()->getParent(); dbgs() << Instruction::getOpcodeName(I->getOpcode()) << " " << *Ops[0].Op->getType() << '\t'; for (unsigned i = 0, e = Ops.size(); i != e; ++i) { 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 { DenseMap RankMap; DenseMap, unsigned> ValueRankMap; SmallVector RedoInsts; SmallVector DeadInsts; bool MadeChange; public: static char ID; // Pass identification, replacement for typeid Reassociate() : FunctionPass(ID) { initializeReassociatePass(*PassRegistry::getPassRegistry()); } bool runOnFunction(Function &F); virtual void getAnalysisUsage(AnalysisUsage &AU) const { AU.setPreservesCFG(); } private: void BuildRankMap(Function &F); unsigned getRank(Value *V); Value *ReassociateExpression(BinaryOperator *I); void RewriteExprTree(BinaryOperator *I, SmallVectorImpl &Ops, unsigned Idx = 0); 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); void LinearizeExprTree(BinaryOperator *I, SmallVectorImpl &Ops); void LinearizeExpr(BinaryOperator *I); Value *RemoveFactorFromExpression(Value *V, Value *Factor); void ReassociateInst(BasicBlock::iterator &BBI); void RemoveDeadBinaryOp(Value *V); }; } char Reassociate::ID = 0; 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)) return; Value *LHS = Op->getOperand(0), *RHS = Op->getOperand(1); ValueRankMap.erase(Op); DeadInsts.push_back(Op); RemoveDeadBinaryOp(LHS); RemoveDeadBinaryOp(RHS); } 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::SDiv || I->getOpcode() == Instruction::FDiv || I->getOpcode() == Instruction::URem || I->getOpcode() == Instruction::SRem || I->getOpcode() == Instruction::FRem) return true; return false; } void Reassociate::BuildRankMap(Function &F) { unsigned i = 2; // Assign distinct ranks to function arguments for (Function::arg_iterator I = F.arg_begin(), E = F.arg_end(); I != E; ++I) ValueRankMap[&*I] = ++i; ReversePostOrderTraversal RPOT(&F); for (ReversePostOrderTraversal::rpo_iterator I = RPOT.begin(), E = RPOT.end(); I != E; ++I) { BasicBlock *BB = *I; unsigned BBRank = RankMap[BB] = ++i << 16; // Walk the basic block, adding precomputed ranks for any instructions that // we cannot move. This ensures that the ranks for these instructions are // all different in the block. for (BasicBlock::iterator I = BB->begin(), E = BB->end(); I != E; ++I) if (isUnmovableInstruction(I)) ValueRankMap[&*I] = ++BBRank; } } unsigned Reassociate::getRank(Value *V) { Instruction *I = dyn_cast(V); if (I == 0) { if (isa(V)) return ValueRankMap[V]; // Function argument. return 0; // Otherwise it's a global or constant, rank 0. } 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 // for PHI nodes, we cannot have infinite recursion here, because there // cannot be loops in the value graph that do not go through PHI nodes. unsigned Rank = 0, MaxRank = RankMap[I->getParent()]; for (unsigned i = 0, e = I->getNumOperands(); i != e && Rank != MaxRank; ++i) Rank = std::max(Rank, getRank(I->getOperand(i))); // 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()->isIntegerTy() || (!BinaryOperator::isNot(I) && !BinaryOperator::isNeg(I))) ++Rank; //DEBUG(dbgs() << "Calculated Rank[" << V->getName() << "] = " // << Rank << "\n"); return ValueRankMap[I] = 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; } /// LowerNegateToMultiply - Replace 0-X with X*-1. /// static Instruction *LowerNegateToMultiply(Instruction *Neg, DenseMap, unsigned> &ValueRankMap) { Constant *Cst = Constant::getAllOnesValue(Neg->getType()); Instruction *Res = BinaryOperator::CreateMul(Neg->getOperand(1), Cst, "",Neg); ValueRankMap.erase(Neg); Res->takeName(Neg); Neg->replaceAllUsesWith(Res); Res->setDebugLoc(Neg->getDebugLoc()); Neg->eraseFromParent(); 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 = isReassociableOp(I->getOperand(0), I->getOpcode()); BinaryOperator *RHS = isReassociableOp(I->getOperand(1), I->getOpcode()); assert(LHS && RHS && "Not an expression that needs linearization?"); DEBUG({ dbgs() << "Linear:\n"; dbgs() << '\t' << *LHS << "\t\n" << *RHS << "\t\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); // Conservatively clear all the optional flags, which may not hold // after the reassociation. I->clearSubclassOptionalData(); LHS->clearSubclassOptionalData(); RHS->clearSubclassOptionalData(); ++NumLinear; MadeChange = true; DEBUG(dbgs() << "Linearized: " << *I << '\n'); // If D is part of this expression tree, tail recurse. if (isReassociableOp(I->getOperand(1), I->getOpcode())) LinearizeExpr(I); } /// 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 expression (((a+b)+c)+d), and collects information about the /// rank of the non-tree operands. /// /// 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. /// void Reassociate::LinearizeExprTree(BinaryOperator *I, SmallVectorImpl &Ops) { Value *LHS = I->getOperand(0), *RHS = I->getOperand(1); unsigned Opcode = I->getOpcode(); // First step, linearize the expression if it is in ((A+B)+(C+D)) form. BinaryOperator *LHSBO = isReassociableOp(LHS, Opcode); BinaryOperator *RHSBO = isReassociableOp(RHS, Opcode); // 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); } } 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; } // Turn X+(Y+Z) -> (Y+Z)+X std::swap(LHSBO, RHSBO); std::swap(LHS, RHS); bool Success = !I->swapOperands(); assert(Success && "swapOperands failed"); (void)Success; MadeChange = true; } else if (RHSBO) { // Turn (A+B)+(C+D) -> (((A+B)+C)+D). This guarantees 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); // Linearize the expression tree on the LHS. LinearizeExprTree(LHSBO, Ops); // 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())); } // 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. void Reassociate::RewriteExprTree(BinaryOperator *I, SmallVectorImpl &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(dbgs() << "RA: " << *I << '\n'); I->setOperand(0, Ops[i].Op); I->setOperand(1, Ops[i+1].Op); // Clear all the optional flags, which may not hold after the // reassociation if the expression involved more than just this operation. if (Ops.size() != 2) I->clearSubclassOptionalData(); DEBUG(dbgs() << "TO: " << *I << '\n'); 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); } return; } assert(i+2 < Ops.size() && "Ops index out of range!"); if (I->getOperand(1) != Ops[i].Op) { DEBUG(dbgs() << "RA: " << *I << '\n'); I->setOperand(1, Ops[i].Op); // Conservatively clear all the optional flags, which may not hold // after the reassociation. I->clearSubclassOptionalData(); DEBUG(dbgs() << "TO: " << *I << '\n'); MadeChange = true; ++NumChanged; } 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); } /// 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, // this means that we turn this: // 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. // 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; } // 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); } /// ShouldBreakUpSubtract - Return true if we should break up this subtract of /// X-Y into (X + -Y). 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) || isReassociableOp(Sub->getOperand(0), Instruction::Sub)) return true; if (isReassociableOp(Sub->getOperand(1), Instruction::Add) || isReassociableOp(Sub->getOperand(1), Instruction::Sub)) return true; 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, DenseMap, unsigned> &ValueRankMap) { // 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. // Value *NegVal = NegateValue(Sub->getOperand(1), Sub); Instruction *New = BinaryOperator::CreateAdd(Sub->getOperand(0), NegVal, "", Sub); New->takeName(Sub); // Everyone now refers to the add instruction. ValueRankMap.erase(Sub); Sub->replaceAllUsesWith(New); New->setDebugLoc(Sub->getDebugLoc()); Sub->eraseFromParent(); 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, DenseMap, 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); Mul->setDebugLoc(Shl->getDebugLoc()); Shl->eraseFromParent(); return Mul; } return 0; } /// 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. for (unsigned j = i-1; j != ~0U && Ops[j].Rank == XRank; --j) if (Ops[j].Op == X) return j; return 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){ 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 /// 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; SmallVector Factors; LinearizeExprTree(BO, Factors); bool FoundFactor = false; 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; } 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) { ValueRankMap.erase(BO); DeadInsts.push_back(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, const SmallVectorImpl &Ops, bool IsRoot) { BinaryOperator *BO; if (!(V->hasOneUse() || V->use_empty()) || // More than one use. !(BO = dyn_cast(V)) || BO->getOpcode() != Instruction::Mul) { Factors.push_back(V); return; } // If this value has a single use because it is another input to the add // tree we're reassociating and we dropped its use, it actually has two // uses and we can't factor it. if (!IsRoot) { for (unsigned i = 0, e = Ops.size(); i != e; ++i) if (Ops[i].Op == V) { Factors.push_back(V); return; } } // Otherwise, add the LHS and RHS to the list of factors. FindSingleUseMultiplyFactors(BO->getOperand(1), Factors, Ops, false); FindSingleUseMultiplyFactors(BO->getOperand(0), Factors, Ops, false); } /// 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, 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) { // First, check for X and ~X in the operand list. assert(i < Ops.size()); if (BinaryOperator::isNot(Ops[i].Op)) { // Cannot occur for ^. Value *X = BinaryOperator::getNotArgument(Ops[i].Op); unsigned FoundX = FindInOperandList(Ops, i, X); if (FoundX != i) { if (Opcode == Instruction::And) // ...&X&~X = 0 return Constant::getNullValue(X->getType()); 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 for And and Or. Ops.erase(Ops.begin()+i); --i; --e; ++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; } /// 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(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. 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) { 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.push_back(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(TheOp)) continue; 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) return Constant::getNullValue(X->getType()); Ops.erase(Ops.begin()+i); if (i < FoundX) --FoundX; else --i; // Need to back up an extra one. Ops.erase(Ops.begin()+FoundX); ++NumAnnihil; --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 = 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, Ops, true); 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 = dyn_cast(Ops[i].Op); if (BOp == 0 || BOp->getOpcode() != Instruction::Mul || !BOp->use_empty()) 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; RedoInsts.push_back(V); // Create the multiply. Value *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.push_back(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) { unsigned FactorPowerSum = 0; DenseMap FactorCounts; for (unsigned LastIdx = 0, Idx = 0, Size = Ops.size(); Idx < Size; ++Idx) { // Note that 'use_empty' uses means the only use is in the linearized tree // represented by Ops -- we remove the values from the actual operations to // reduce their use count. if (!Ops[Idx].Op->use_empty()) { if (LastIdx == Idx) ++LastIdx; continue; } if (LastIdx == Idx || Ops[LastIdx].Op != Ops[Idx].Op) { LastIdx = Idx; continue; } // Track for simplification all factors which occur 2 or more times. DenseMap::iterator CountIt; bool Inserted; llvm::tie(CountIt, Inserted) = FactorCounts.insert(std::make_pair(Ops[Idx].Op, 2)); if (Inserted) { FactorPowerSum += 2; Factors.push_back(Factor(Ops[Idx].Op, 2)); } else { ++CountIt->second; ++FactorPowerSum; } } // 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; // Remove all the operands which are in the map. Ops.erase(std::remove_if(Ops.begin(), Ops.end(), IsValueInMap(FactorCounts)), Ops.end()); // Record the adjusted power for the simplification factors. We add back into // the Ops list any values with an odd power, and make the power even. This // allows the outer-most multiplication tree to remain in tact during // simplification. unsigned OldOpsSize = Ops.size(); for (unsigned Idx = 0, Size = Factors.size(); Idx != Size; ++Idx) { Factors[Idx].Power = FactorCounts[Factors[Idx].Base]; if (Factors[Idx].Power & 1) { Ops.push_back(ValueEntry(getRank(Factors[Idx].Base), Factors[Idx].Base)); --Factors[Idx].Power; --FactorPowerSum; } } // None of the adjustments above should have reduced the sum of factor powers // below our mininum of '4'. assert(FactorPowerSum >= 4); // Patch up the sort of the ops vector by sorting the factors we added back // onto the back, and merging the two sequences. if (OldOpsSize != Ops.size()) { SmallVectorImpl::iterator MiddleIt = Ops.begin() + OldOpsSize; std::sort(MiddleIt, Ops.end()); std::inplace_merge(Ops.begin(), MiddleIt, Ops.end()); } 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. Factors[LastIdx].Base = buildMultiplyTree(Builder, InnerProduct); RedoInsts.push_back(Factors[LastIdx].Base); 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); RedoInsts.push_back(V); 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, 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; 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); } // 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()) // X & 0 -> 0 return CstVal; if (CstVal->isAllOnesValue()) // X & -1 -> X Ops.pop_back(); break; case Instruction::Mul: if (CstVal->isZero()) { // X * 0 -> 0 ++NumAnnihil; return CstVal; } if (cast(CstVal)->isOne()) Ops.pop_back(); // X * 1 -> X break; case Instruction::Or: if (CstVal->isAllOnesValue()) // X | -1 -> -1 return CstVal; // FALLTHROUGH! case Instruction::Add: case Instruction::Xor: if (CstVal->isZero()) // X [|^+] 0 -> X 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: if (Value *Result = OptimizeAndOrXor(Opcode, Ops)) return Result; break; case Instruction::Add: if (Value *Result = OptimizeAdd(I, Ops)) return Result; break; case Instruction::Mul: if (Value *Result = OptimizeMul(I, Ops)) return Result; break; } if (IterateOptimization || Ops.size() != NumOps) return OptimizeExpression(I, Ops); return 0; } /// ReassociateInst - Inspect and reassociate the instruction at the /// given position, post-incrementing the position. void Reassociate::ReassociateInst(BasicBlock::iterator &BBI) { Instruction *BI = BBI++; if (BI->getOpcode() == Instruction::Shl && isa(BI->getOperand(1))) if (Instruction *NI = ConvertShiftToMul(BI, ValueRankMap)) { MadeChange = true; BI = NI; } // 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 (isa(BI) && (BI->getType()->isFloatingPointTy() || BI->getType()->isVectorTy())) { // FAdd and FMul can be commuted. if (BI->getOpcode() != Instruction::FMul && BI->getOpcode() != Instruction::FAdd) return; Value *LHS = BI->getOperand(0); Value *RHS = BI->getOperand(1); unsigned LHSRank = getRank(LHS); unsigned RHSRank = getRank(RHS); // Sort the operands by rank. if (RHSRank < LHSRank) { BI->setOperand(0, RHS); BI->setOperand(1, LHS); } return; } // Do not reassociate operations that we do not understand. if (!isa(BI)) return; // 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 (BI->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 (BI->getOpcode() == Instruction::Sub) { if (ShouldBreakUpSubtract(BI)) { BI = BreakUpSubtract(BI, ValueRankMap); // Reset the BBI iterator in case BreakUpSubtract changed the // instruction it points to. BBI = BI; ++BBI; 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; } } } // If this instruction is a commutative binary operator, process it. if (!BI->isAssociative()) return; BinaryOperator *I = cast(BI); // 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())) 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) return; ReassociateExpression(I); } Value *Reassociate::ReassociateExpression(BinaryOperator *I) { // First, walk the expression tree, linearizing the tree, collecting the // operand information. SmallVector Ops; LinearizeExprTree(I, Ops); 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 // positions maintained (and so the compiler is deterministic). Note that // 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)) { // This expression tree simplified to something that isn't a tree, // eliminate it. DEBUG(dbgs() << "Reassoc to scalar: " << *V << '\n'); I->replaceAllUsesWith(V); if (Instruction *VI = dyn_cast(V)) VI->setDebugLoc(I->getDebugLoc()); RemoveDeadBinaryOp(I); ++NumAnnihil; return V; } // 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 // can fold the negation into the add: (-X)*Y + Z -> Z-X*Y if (I->getOpcode() == Instruction::Mul && I->hasOneUse() && cast(I->use_back())->getOpcode() == Instruction::Add && isa(Ops.back().Op) && cast(Ops.back().Op)->isAllOnesValue()) { ValueEntry Tmp = Ops.pop_back_val(); Ops.insert(Ops.begin(), Tmp); } DEBUG(dbgs() << "RAOut:\t"; PrintOps(I, Ops); dbgs() << '\n'); if (Ops.size() == 1) { // This expression tree simplified to something that isn't a tree, // eliminate it. I->replaceAllUsesWith(Ops[0].Op); if (Instruction *OI = dyn_cast(Ops[0].Op)) OI->setDebugLoc(I->getDebugLoc()); RemoveDeadBinaryOp(I); return Ops[0].Op; } // Now that we ordered and optimized the expressions, splat them back into // the expression tree, removing any unneeded nodes. RewriteExprTree(I, Ops); return I; } bool Reassociate::runOnFunction(Function &F) { // Recalculate the rank map for F BuildRankMap(F); MadeChange = false; for (Function::iterator FI = F.begin(), FE = F.end(); FI != FE; ++FI) for (BasicBlock::iterator BBI = FI->begin(); BBI != FI->end(); ) ReassociateInst(BBI); // Now that we're done, revisit any instructions which are likely to // have secondary reassociation opportunities. while (!RedoInsts.empty()) if (Value *V = RedoInsts.pop_back_val()) { BasicBlock::iterator BBI = cast(V); ReassociateInst(BBI); } // Now that we're done, delete any instructions which are no longer used. while (!DeadInsts.empty()) if (Value *V = DeadInsts.pop_back_val()) RecursivelyDeleteTriviallyDeadInstructions(V); // We are done with the rank map. RankMap.clear(); ValueRankMap.clear(); return MadeChange; }