//===- ScalarEvolution.cpp - Scalar Evolution Analysis ----------*- C++ -*-===// // // The LLVM Compiler Infrastructure // // This file was developed by the LLVM research group and is distributed under // the University of Illinois Open Source License. See LICENSE.TXT for details. // //===----------------------------------------------------------------------===// // // This file contains the implementation of the scalar evolution analysis // engine, which is used primarily to analyze expressions involving induction // variables in loops. // // There are several aspects to this library. First is the representation of // scalar expressions, which are represented as subclasses of the SCEV class. // These classes are used to represent certain types of subexpressions that we // can handle. These classes are reference counted, managed by the SCEVHandle // class. We only create one SCEV of a particular shape, so pointer-comparisons // for equality are legal. // // One important aspect of the SCEV objects is that they are never cyclic, even // if there is a cycle in the dataflow for an expression (ie, a PHI node). If // the PHI node is one of the idioms that we can represent (e.g., a polynomial // recurrence) then we represent it directly as a recurrence node, otherwise we // represent it as a SCEVUnknown node. // // In addition to being able to represent expressions of various types, we also // have folders that are used to build the *canonical* representation for a // particular expression. These folders are capable of using a variety of // rewrite rules to simplify the expressions. // // Once the folders are defined, we can implement the more interesting // higher-level code, such as the code that recognizes PHI nodes of various // types, computes the execution count of a loop, etc. // // TODO: We should use these routines and value representations to implement // dependence analysis! // //===----------------------------------------------------------------------===// // // There are several good references for the techniques used in this analysis. // // Chains of recurrences -- a method to expedite the evaluation // of closed-form functions // Olaf Bachmann, Paul S. Wang, Eugene V. Zima // // On computational properties of chains of recurrences // Eugene V. Zima // // Symbolic Evaluation of Chains of Recurrences for Loop Optimization // Robert A. van Engelen // // Efficient Symbolic Analysis for Optimizing Compilers // Robert A. van Engelen // // Using the chains of recurrences algebra for data dependence testing and // induction variable substitution // MS Thesis, Johnie Birch // //===----------------------------------------------------------------------===// #define DEBUG_TYPE "scalar-evolution" #include "llvm/Analysis/ScalarEvolutionExpressions.h" #include "llvm/Constants.h" #include "llvm/DerivedTypes.h" #include "llvm/GlobalVariable.h" #include "llvm/Instructions.h" #include "llvm/Analysis/ConstantFolding.h" #include "llvm/Analysis/LoopInfo.h" #include "llvm/Assembly/Writer.h" #include "llvm/Transforms/Scalar.h" #include "llvm/Support/CFG.h" #include "llvm/Support/CommandLine.h" #include "llvm/Support/Compiler.h" #include "llvm/Support/ConstantRange.h" #include "llvm/Support/InstIterator.h" #include "llvm/Support/ManagedStatic.h" #include "llvm/Support/MathExtras.h" #include "llvm/Support/Streams.h" #include "llvm/ADT/Statistic.h" #include #include #include using namespace llvm; STATISTIC(NumBruteForceEvaluations, "Number of brute force evaluations needed to " "calculate high-order polynomial exit values"); STATISTIC(NumArrayLenItCounts, "Number of trip counts computed with array length"); STATISTIC(NumTripCountsComputed, "Number of loops with predictable loop counts"); STATISTIC(NumTripCountsNotComputed, "Number of loops without predictable loop counts"); STATISTIC(NumBruteForceTripCountsComputed, "Number of loops with trip counts computed by force"); cl::opt MaxBruteForceIterations("scalar-evolution-max-iterations", cl::ReallyHidden, cl::desc("Maximum number of iterations SCEV will " "symbolically execute a constant derived loop"), cl::init(100)); namespace { RegisterPass R("scalar-evolution", "Scalar Evolution Analysis"); } char ScalarEvolution::ID = 0; //===----------------------------------------------------------------------===// // SCEV class definitions //===----------------------------------------------------------------------===// //===----------------------------------------------------------------------===// // Implementation of the SCEV class. // SCEV::~SCEV() {} void SCEV::dump() const { print(cerr); } /// getValueRange - Return the tightest constant bounds that this value is /// known to have. This method is only valid on integer SCEV objects. ConstantRange SCEV::getValueRange() const { const Type *Ty = getType(); assert(Ty->isInteger() && "Can't get range for a non-integer SCEV!"); // Default to a full range if no better information is available. return ConstantRange(getBitWidth()); } uint32_t SCEV::getBitWidth() const { if (const IntegerType* ITy = dyn_cast(getType())) return ITy->getBitWidth(); return 0; } SCEVCouldNotCompute::SCEVCouldNotCompute() : SCEV(scCouldNotCompute) {} bool SCEVCouldNotCompute::isLoopInvariant(const Loop *L) const { assert(0 && "Attempt to use a SCEVCouldNotCompute object!"); return false; } const Type *SCEVCouldNotCompute::getType() const { assert(0 && "Attempt to use a SCEVCouldNotCompute object!"); return 0; } bool SCEVCouldNotCompute::hasComputableLoopEvolution(const Loop *L) const { assert(0 && "Attempt to use a SCEVCouldNotCompute object!"); return false; } SCEVHandle SCEVCouldNotCompute:: replaceSymbolicValuesWithConcrete(const SCEVHandle &Sym, const SCEVHandle &Conc) const { return this; } void SCEVCouldNotCompute::print(std::ostream &OS) const { OS << "***COULDNOTCOMPUTE***"; } bool SCEVCouldNotCompute::classof(const SCEV *S) { return S->getSCEVType() == scCouldNotCompute; } // SCEVConstants - Only allow the creation of one SCEVConstant for any // particular value. Don't use a SCEVHandle here, or else the object will // never be deleted! static ManagedStatic > SCEVConstants; SCEVConstant::~SCEVConstant() { SCEVConstants->erase(V); } SCEVHandle SCEVConstant::get(ConstantInt *V) { SCEVConstant *&R = (*SCEVConstants)[V]; if (R == 0) R = new SCEVConstant(V); return R; } ConstantRange SCEVConstant::getValueRange() const { return ConstantRange(V->getValue()); } const Type *SCEVConstant::getType() const { return V->getType(); } void SCEVConstant::print(std::ostream &OS) const { WriteAsOperand(OS, V, false); } // SCEVTruncates - Only allow the creation of one SCEVTruncateExpr for any // particular input. Don't use a SCEVHandle here, or else the object will // never be deleted! static ManagedStatic, SCEVTruncateExpr*> > SCEVTruncates; SCEVTruncateExpr::SCEVTruncateExpr(const SCEVHandle &op, const Type *ty) : SCEV(scTruncate), Op(op), Ty(ty) { assert(Op->getType()->isInteger() && Ty->isInteger() && "Cannot truncate non-integer value!"); assert(Op->getType()->getPrimitiveSizeInBits() > Ty->getPrimitiveSizeInBits() && "This is not a truncating conversion!"); } SCEVTruncateExpr::~SCEVTruncateExpr() { SCEVTruncates->erase(std::make_pair(Op, Ty)); } ConstantRange SCEVTruncateExpr::getValueRange() const { return getOperand()->getValueRange().truncate(getBitWidth()); } void SCEVTruncateExpr::print(std::ostream &OS) const { OS << "(truncate " << *Op << " to " << *Ty << ")"; } // SCEVZeroExtends - Only allow the creation of one SCEVZeroExtendExpr for any // particular input. Don't use a SCEVHandle here, or else the object will never // be deleted! static ManagedStatic, SCEVZeroExtendExpr*> > SCEVZeroExtends; SCEVZeroExtendExpr::SCEVZeroExtendExpr(const SCEVHandle &op, const Type *ty) : SCEV(scZeroExtend), Op(op), Ty(ty) { assert(Op->getType()->isInteger() && Ty->isInteger() && "Cannot zero extend non-integer value!"); assert(Op->getType()->getPrimitiveSizeInBits() < Ty->getPrimitiveSizeInBits() && "This is not an extending conversion!"); } SCEVZeroExtendExpr::~SCEVZeroExtendExpr() { SCEVZeroExtends->erase(std::make_pair(Op, Ty)); } ConstantRange SCEVZeroExtendExpr::getValueRange() const { return getOperand()->getValueRange().zeroExtend(getBitWidth()); } void SCEVZeroExtendExpr::print(std::ostream &OS) const { OS << "(zeroextend " << *Op << " to " << *Ty << ")"; } // SCEVCommExprs - Only allow the creation of one SCEVCommutativeExpr for any // particular input. Don't use a SCEVHandle here, or else the object will never // be deleted! static ManagedStatic >, SCEVCommutativeExpr*> > SCEVCommExprs; SCEVCommutativeExpr::~SCEVCommutativeExpr() { SCEVCommExprs->erase(std::make_pair(getSCEVType(), std::vector(Operands.begin(), Operands.end()))); } void SCEVCommutativeExpr::print(std::ostream &OS) const { assert(Operands.size() > 1 && "This plus expr shouldn't exist!"); const char *OpStr = getOperationStr(); OS << "(" << *Operands[0]; for (unsigned i = 1, e = Operands.size(); i != e; ++i) OS << OpStr << *Operands[i]; OS << ")"; } SCEVHandle SCEVCommutativeExpr:: replaceSymbolicValuesWithConcrete(const SCEVHandle &Sym, const SCEVHandle &Conc) const { for (unsigned i = 0, e = getNumOperands(); i != e; ++i) { SCEVHandle H = getOperand(i)->replaceSymbolicValuesWithConcrete(Sym, Conc); if (H != getOperand(i)) { std::vector NewOps; NewOps.reserve(getNumOperands()); for (unsigned j = 0; j != i; ++j) NewOps.push_back(getOperand(j)); NewOps.push_back(H); for (++i; i != e; ++i) NewOps.push_back(getOperand(i)-> replaceSymbolicValuesWithConcrete(Sym, Conc)); if (isa(this)) return SCEVAddExpr::get(NewOps); else if (isa(this)) return SCEVMulExpr::get(NewOps); else assert(0 && "Unknown commutative expr!"); } } return this; } // SCEVSDivs - Only allow the creation of one SCEVSDivExpr for any particular // input. Don't use a SCEVHandle here, or else the object will never be // deleted! static ManagedStatic, SCEVSDivExpr*> > SCEVSDivs; SCEVSDivExpr::~SCEVSDivExpr() { SCEVSDivs->erase(std::make_pair(LHS, RHS)); } void SCEVSDivExpr::print(std::ostream &OS) const { OS << "(" << *LHS << " /s " << *RHS << ")"; } const Type *SCEVSDivExpr::getType() const { return LHS->getType(); } // SCEVAddRecExprs - Only allow the creation of one SCEVAddRecExpr for any // particular input. Don't use a SCEVHandle here, or else the object will never // be deleted! static ManagedStatic >, SCEVAddRecExpr*> > SCEVAddRecExprs; SCEVAddRecExpr::~SCEVAddRecExpr() { SCEVAddRecExprs->erase(std::make_pair(L, std::vector(Operands.begin(), Operands.end()))); } SCEVHandle SCEVAddRecExpr:: replaceSymbolicValuesWithConcrete(const SCEVHandle &Sym, const SCEVHandle &Conc) const { for (unsigned i = 0, e = getNumOperands(); i != e; ++i) { SCEVHandle H = getOperand(i)->replaceSymbolicValuesWithConcrete(Sym, Conc); if (H != getOperand(i)) { std::vector NewOps; NewOps.reserve(getNumOperands()); for (unsigned j = 0; j != i; ++j) NewOps.push_back(getOperand(j)); NewOps.push_back(H); for (++i; i != e; ++i) NewOps.push_back(getOperand(i)-> replaceSymbolicValuesWithConcrete(Sym, Conc)); return get(NewOps, L); } } return this; } bool SCEVAddRecExpr::isLoopInvariant(const Loop *QueryLoop) const { // This recurrence is invariant w.r.t to QueryLoop iff QueryLoop doesn't // contain L and if the start is invariant. return !QueryLoop->contains(L->getHeader()) && getOperand(0)->isLoopInvariant(QueryLoop); } void SCEVAddRecExpr::print(std::ostream &OS) const { OS << "{" << *Operands[0]; for (unsigned i = 1, e = Operands.size(); i != e; ++i) OS << ",+," << *Operands[i]; OS << "}<" << L->getHeader()->getName() + ">"; } // SCEVUnknowns - Only allow the creation of one SCEVUnknown for any particular // value. Don't use a SCEVHandle here, or else the object will never be // deleted! static ManagedStatic > SCEVUnknowns; SCEVUnknown::~SCEVUnknown() { SCEVUnknowns->erase(V); } bool SCEVUnknown::isLoopInvariant(const Loop *L) const { // All non-instruction values are loop invariant. All instructions are loop // invariant if they are not contained in the specified loop. if (Instruction *I = dyn_cast(V)) return !L->contains(I->getParent()); return true; } const Type *SCEVUnknown::getType() const { return V->getType(); } void SCEVUnknown::print(std::ostream &OS) const { WriteAsOperand(OS, V, false); } //===----------------------------------------------------------------------===// // SCEV Utilities //===----------------------------------------------------------------------===// namespace { /// SCEVComplexityCompare - Return true if the complexity of the LHS is less /// than the complexity of the RHS. This comparator is used to canonicalize /// expressions. struct VISIBILITY_HIDDEN SCEVComplexityCompare { bool operator()(SCEV *LHS, SCEV *RHS) { return LHS->getSCEVType() < RHS->getSCEVType(); } }; } /// GroupByComplexity - Given a list of SCEV objects, order them by their /// complexity, and group objects of the same complexity together by value. /// When this routine is finished, we know that any duplicates in the vector are /// consecutive and that complexity is monotonically increasing. /// /// Note that we go take special precautions to ensure that we get determinstic /// results from this routine. In other words, we don't want the results of /// this to depend on where the addresses of various SCEV objects happened to /// land in memory. /// static void GroupByComplexity(std::vector &Ops) { if (Ops.size() < 2) return; // Noop if (Ops.size() == 2) { // This is the common case, which also happens to be trivially simple. // Special case it. if (Ops[0]->getSCEVType() > Ops[1]->getSCEVType()) std::swap(Ops[0], Ops[1]); return; } // Do the rough sort by complexity. std::sort(Ops.begin(), Ops.end(), SCEVComplexityCompare()); // Now that we are sorted by complexity, group elements of the same // complexity. Note that this is, at worst, N^2, but the vector is likely to // be extremely short in practice. Note that we take this approach because we // do not want to depend on the addresses of the objects we are grouping. for (unsigned i = 0, e = Ops.size(); i != e-2; ++i) { SCEV *S = Ops[i]; unsigned Complexity = S->getSCEVType(); // If there are any objects of the same complexity and same value as this // one, group them. for (unsigned j = i+1; j != e && Ops[j]->getSCEVType() == Complexity; ++j) { if (Ops[j] == S) { // Found a duplicate. // Move it to immediately after i'th element. std::swap(Ops[i+1], Ops[j]); ++i; // no need to rescan it. if (i == e-2) return; // Done! } } } } //===----------------------------------------------------------------------===// // Simple SCEV method implementations //===----------------------------------------------------------------------===// /// getIntegerSCEV - Given an integer or FP type, create a constant for the /// specified signed integer value and return a SCEV for the constant. SCEVHandle SCEVUnknown::getIntegerSCEV(int Val, const Type *Ty) { Constant *C; if (Val == 0) C = Constant::getNullValue(Ty); else if (Ty->isFloatingPoint()) C = ConstantFP::get(Ty, Val); else C = ConstantInt::get(Ty, Val); return SCEVUnknown::get(C); } SCEVHandle SCEVUnknown::getIntegerSCEV(const APInt& Val) { return SCEVUnknown::get(ConstantInt::get(Val)); } /// getTruncateOrZeroExtend - Return a SCEV corresponding to a conversion of the /// input value to the specified type. If the type must be extended, it is zero /// extended. static SCEVHandle getTruncateOrZeroExtend(const SCEVHandle &V, const Type *Ty) { const Type *SrcTy = V->getType(); assert(SrcTy->isInteger() && Ty->isInteger() && "Cannot truncate or zero extend with non-integer arguments!"); if (SrcTy->getPrimitiveSizeInBits() == Ty->getPrimitiveSizeInBits()) return V; // No conversion if (SrcTy->getPrimitiveSizeInBits() > Ty->getPrimitiveSizeInBits()) return SCEVTruncateExpr::get(V, Ty); return SCEVZeroExtendExpr::get(V, Ty); } /// getNegativeSCEV - Return a SCEV corresponding to -V = -1*V /// SCEVHandle SCEV::getNegativeSCEV(const SCEVHandle &V) { if (SCEVConstant *VC = dyn_cast(V)) return SCEVUnknown::get(ConstantExpr::getNeg(VC->getValue())); return SCEVMulExpr::get(V, SCEVUnknown::getIntegerSCEV(-1, V->getType())); } /// getMinusSCEV - Return a SCEV corresponding to LHS - RHS. /// SCEVHandle SCEV::getMinusSCEV(const SCEVHandle &LHS, const SCEVHandle &RHS) { // X - Y --> X + -Y return SCEVAddExpr::get(LHS, SCEV::getNegativeSCEV(RHS)); } /// PartialFact - Compute V!/(V-NumSteps)! static SCEVHandle PartialFact(SCEVHandle V, unsigned NumSteps) { // Handle this case efficiently, it is common to have constant iteration // counts while computing loop exit values. if (SCEVConstant *SC = dyn_cast(V)) { const APInt& Val = SC->getValue()->getValue(); APInt Result(Val.getBitWidth(), 1); for (; NumSteps; --NumSteps) Result *= Val-(NumSteps-1); return SCEVUnknown::get(ConstantInt::get(Result)); } const Type *Ty = V->getType(); if (NumSteps == 0) return SCEVUnknown::getIntegerSCEV(1, Ty); SCEVHandle Result = V; for (unsigned i = 1; i != NumSteps; ++i) Result = SCEVMulExpr::get(Result, SCEV::getMinusSCEV(V, SCEVUnknown::getIntegerSCEV(i, Ty))); return Result; } /// evaluateAtIteration - Return the value of this chain of recurrences at /// the specified iteration number. We can evaluate this recurrence by /// multiplying each element in the chain by the binomial coefficient /// corresponding to it. In other words, we can evaluate {A,+,B,+,C,+,D} as: /// /// A*choose(It, 0) + B*choose(It, 1) + C*choose(It, 2) + D*choose(It, 3) /// /// FIXME/VERIFY: I don't trust that this is correct in the face of overflow. /// Is the binomial equation safe using modular arithmetic?? /// SCEVHandle SCEVAddRecExpr::evaluateAtIteration(SCEVHandle It) const { SCEVHandle Result = getStart(); int Divisor = 1; const Type *Ty = It->getType(); for (unsigned i = 1, e = getNumOperands(); i != e; ++i) { SCEVHandle BC = PartialFact(It, i); Divisor *= i; SCEVHandle Val = SCEVSDivExpr::get(SCEVMulExpr::get(BC, getOperand(i)), SCEVUnknown::getIntegerSCEV(Divisor,Ty)); Result = SCEVAddExpr::get(Result, Val); } return Result; } //===----------------------------------------------------------------------===// // SCEV Expression folder implementations //===----------------------------------------------------------------------===// SCEVHandle SCEVTruncateExpr::get(const SCEVHandle &Op, const Type *Ty) { if (SCEVConstant *SC = dyn_cast(Op)) return SCEVUnknown::get( ConstantExpr::getTrunc(SC->getValue(), Ty)); // If the input value is a chrec scev made out of constants, truncate // all of the constants. if (SCEVAddRecExpr *AddRec = dyn_cast(Op)) { std::vector Operands; for (unsigned i = 0, e = AddRec->getNumOperands(); i != e; ++i) // FIXME: This should allow truncation of other expression types! if (isa(AddRec->getOperand(i))) Operands.push_back(get(AddRec->getOperand(i), Ty)); else break; if (Operands.size() == AddRec->getNumOperands()) return SCEVAddRecExpr::get(Operands, AddRec->getLoop()); } SCEVTruncateExpr *&Result = (*SCEVTruncates)[std::make_pair(Op, Ty)]; if (Result == 0) Result = new SCEVTruncateExpr(Op, Ty); return Result; } SCEVHandle SCEVZeroExtendExpr::get(const SCEVHandle &Op, const Type *Ty) { if (SCEVConstant *SC = dyn_cast(Op)) return SCEVUnknown::get( ConstantExpr::getZExt(SC->getValue(), Ty)); // FIXME: If the input value is a chrec scev, and we can prove that the value // did not overflow the old, smaller, value, we can zero extend all of the // operands (often constants). This would allow analysis of something like // this: for (unsigned char X = 0; X < 100; ++X) { int Y = X; } SCEVZeroExtendExpr *&Result = (*SCEVZeroExtends)[std::make_pair(Op, Ty)]; if (Result == 0) Result = new SCEVZeroExtendExpr(Op, Ty); return Result; } // get - Get a canonical add expression, or something simpler if possible. SCEVHandle SCEVAddExpr::get(std::vector &Ops) { assert(!Ops.empty() && "Cannot get empty add!"); if (Ops.size() == 1) return Ops[0]; // Sort by complexity, this groups all similar expression types together. GroupByComplexity(Ops); // If there are any constants, fold them together. unsigned Idx = 0; if (SCEVConstant *LHSC = dyn_cast(Ops[0])) { ++Idx; assert(Idx < Ops.size()); while (SCEVConstant *RHSC = dyn_cast(Ops[Idx])) { // We found two constants, fold them together! Constant *Fold = ConstantInt::get(LHSC->getValue()->getValue() + RHSC->getValue()->getValue()); if (ConstantInt *CI = dyn_cast(Fold)) { Ops[0] = SCEVConstant::get(CI); Ops.erase(Ops.begin()+1); // Erase the folded element if (Ops.size() == 1) return Ops[0]; LHSC = cast(Ops[0]); } else { // If we couldn't fold the expression, move to the next constant. Note // that this is impossible to happen in practice because we always // constant fold constant ints to constant ints. ++Idx; } } // If we are left with a constant zero being added, strip it off. if (cast(Ops[0])->getValue()->isZero()) { Ops.erase(Ops.begin()); --Idx; } } if (Ops.size() == 1) return Ops[0]; // Okay, check to see if the same value occurs in the operand list twice. If // so, merge them together into an multiply expression. Since we sorted the // list, these values are required to be adjacent. const Type *Ty = Ops[0]->getType(); for (unsigned i = 0, e = Ops.size()-1; i != e; ++i) if (Ops[i] == Ops[i+1]) { // X + Y + Y --> X + Y*2 // Found a match, merge the two values into a multiply, and add any // remaining values to the result. SCEVHandle Two = SCEVUnknown::getIntegerSCEV(2, Ty); SCEVHandle Mul = SCEVMulExpr::get(Ops[i], Two); if (Ops.size() == 2) return Mul; Ops.erase(Ops.begin()+i, Ops.begin()+i+2); Ops.push_back(Mul); return SCEVAddExpr::get(Ops); } // Okay, now we know the first non-constant operand. If there are add // operands they would be next. if (Idx < Ops.size()) { bool DeletedAdd = false; while (SCEVAddExpr *Add = dyn_cast(Ops[Idx])) { // If we have an add, expand the add operands onto the end of the operands // list. Ops.insert(Ops.end(), Add->op_begin(), Add->op_end()); Ops.erase(Ops.begin()+Idx); DeletedAdd = true; } // If we deleted at least one add, we added operands to the end of the list, // and they are not necessarily sorted. Recurse to resort and resimplify // any operands we just aquired. if (DeletedAdd) return get(Ops); } // Skip over the add expression until we get to a multiply. while (Idx < Ops.size() && Ops[Idx]->getSCEVType() < scMulExpr) ++Idx; // If we are adding something to a multiply expression, make sure the // something is not already an operand of the multiply. If so, merge it into // the multiply. for (; Idx < Ops.size() && isa(Ops[Idx]); ++Idx) { SCEVMulExpr *Mul = cast(Ops[Idx]); for (unsigned MulOp = 0, e = Mul->getNumOperands(); MulOp != e; ++MulOp) { SCEV *MulOpSCEV = Mul->getOperand(MulOp); for (unsigned AddOp = 0, e = Ops.size(); AddOp != e; ++AddOp) if (MulOpSCEV == Ops[AddOp] && !isa(MulOpSCEV)) { // Fold W + X + (X * Y * Z) --> W + (X * ((Y*Z)+1)) SCEVHandle InnerMul = Mul->getOperand(MulOp == 0); if (Mul->getNumOperands() != 2) { // If the multiply has more than two operands, we must get the // Y*Z term. std::vector MulOps(Mul->op_begin(), Mul->op_end()); MulOps.erase(MulOps.begin()+MulOp); InnerMul = SCEVMulExpr::get(MulOps); } SCEVHandle One = SCEVUnknown::getIntegerSCEV(1, Ty); SCEVHandle AddOne = SCEVAddExpr::get(InnerMul, One); SCEVHandle OuterMul = SCEVMulExpr::get(AddOne, Ops[AddOp]); if (Ops.size() == 2) return OuterMul; if (AddOp < Idx) { Ops.erase(Ops.begin()+AddOp); Ops.erase(Ops.begin()+Idx-1); } else { Ops.erase(Ops.begin()+Idx); Ops.erase(Ops.begin()+AddOp-1); } Ops.push_back(OuterMul); return SCEVAddExpr::get(Ops); } // Check this multiply against other multiplies being added together. for (unsigned OtherMulIdx = Idx+1; OtherMulIdx < Ops.size() && isa(Ops[OtherMulIdx]); ++OtherMulIdx) { SCEVMulExpr *OtherMul = cast(Ops[OtherMulIdx]); // If MulOp occurs in OtherMul, we can fold the two multiplies // together. for (unsigned OMulOp = 0, e = OtherMul->getNumOperands(); OMulOp != e; ++OMulOp) if (OtherMul->getOperand(OMulOp) == MulOpSCEV) { // Fold X + (A*B*C) + (A*D*E) --> X + (A*(B*C+D*E)) SCEVHandle InnerMul1 = Mul->getOperand(MulOp == 0); if (Mul->getNumOperands() != 2) { std::vector MulOps(Mul->op_begin(), Mul->op_end()); MulOps.erase(MulOps.begin()+MulOp); InnerMul1 = SCEVMulExpr::get(MulOps); } SCEVHandle InnerMul2 = OtherMul->getOperand(OMulOp == 0); if (OtherMul->getNumOperands() != 2) { std::vector MulOps(OtherMul->op_begin(), OtherMul->op_end()); MulOps.erase(MulOps.begin()+OMulOp); InnerMul2 = SCEVMulExpr::get(MulOps); } SCEVHandle InnerMulSum = SCEVAddExpr::get(InnerMul1,InnerMul2); SCEVHandle OuterMul = SCEVMulExpr::get(MulOpSCEV, InnerMulSum); if (Ops.size() == 2) return OuterMul; Ops.erase(Ops.begin()+Idx); Ops.erase(Ops.begin()+OtherMulIdx-1); Ops.push_back(OuterMul); return SCEVAddExpr::get(Ops); } } } } // If there are any add recurrences in the operands list, see if any other // added values are loop invariant. If so, we can fold them into the // recurrence. while (Idx < Ops.size() && Ops[Idx]->getSCEVType() < scAddRecExpr) ++Idx; // Scan over all recurrences, trying to fold loop invariants into them. for (; Idx < Ops.size() && isa(Ops[Idx]); ++Idx) { // Scan all of the other operands to this add and add them to the vector if // they are loop invariant w.r.t. the recurrence. std::vector LIOps; SCEVAddRecExpr *AddRec = cast(Ops[Idx]); for (unsigned i = 0, e = Ops.size(); i != e; ++i) if (Ops[i]->isLoopInvariant(AddRec->getLoop())) { LIOps.push_back(Ops[i]); Ops.erase(Ops.begin()+i); --i; --e; } // If we found some loop invariants, fold them into the recurrence. if (!LIOps.empty()) { // NLI + LI + { Start,+,Step} --> NLI + { LI+Start,+,Step } LIOps.push_back(AddRec->getStart()); std::vector AddRecOps(AddRec->op_begin(), AddRec->op_end()); AddRecOps[0] = SCEVAddExpr::get(LIOps); SCEVHandle NewRec = SCEVAddRecExpr::get(AddRecOps, AddRec->getLoop()); // If all of the other operands were loop invariant, we are done. if (Ops.size() == 1) return NewRec; // Otherwise, add the folded AddRec by the non-liv parts. for (unsigned i = 0;; ++i) if (Ops[i] == AddRec) { Ops[i] = NewRec; break; } return SCEVAddExpr::get(Ops); } // Okay, if there weren't any loop invariants to be folded, check to see if // there are multiple AddRec's with the same loop induction variable being // added together. If so, we can fold them. for (unsigned OtherIdx = Idx+1; OtherIdx < Ops.size() && isa(Ops[OtherIdx]);++OtherIdx) if (OtherIdx != Idx) { SCEVAddRecExpr *OtherAddRec = cast(Ops[OtherIdx]); if (AddRec->getLoop() == OtherAddRec->getLoop()) { // Other + {A,+,B} + {C,+,D} --> Other + {A+C,+,B+D} std::vector NewOps(AddRec->op_begin(), AddRec->op_end()); for (unsigned i = 0, e = OtherAddRec->getNumOperands(); i != e; ++i) { if (i >= NewOps.size()) { NewOps.insert(NewOps.end(), OtherAddRec->op_begin()+i, OtherAddRec->op_end()); break; } NewOps[i] = SCEVAddExpr::get(NewOps[i], OtherAddRec->getOperand(i)); } SCEVHandle NewAddRec = SCEVAddRecExpr::get(NewOps, AddRec->getLoop()); if (Ops.size() == 2) return NewAddRec; Ops.erase(Ops.begin()+Idx); Ops.erase(Ops.begin()+OtherIdx-1); Ops.push_back(NewAddRec); return SCEVAddExpr::get(Ops); } } // Otherwise couldn't fold anything into this recurrence. Move onto the // next one. } // Okay, it looks like we really DO need an add expr. Check to see if we // already have one, otherwise create a new one. std::vector SCEVOps(Ops.begin(), Ops.end()); SCEVCommutativeExpr *&Result = (*SCEVCommExprs)[std::make_pair(scAddExpr, SCEVOps)]; if (Result == 0) Result = new SCEVAddExpr(Ops); return Result; } SCEVHandle SCEVMulExpr::get(std::vector &Ops) { assert(!Ops.empty() && "Cannot get empty mul!"); // Sort by complexity, this groups all similar expression types together. GroupByComplexity(Ops); // If there are any constants, fold them together. unsigned Idx = 0; if (SCEVConstant *LHSC = dyn_cast(Ops[0])) { // C1*(C2+V) -> C1*C2 + C1*V if (Ops.size() == 2) if (SCEVAddExpr *Add = dyn_cast(Ops[1])) if (Add->getNumOperands() == 2 && isa(Add->getOperand(0))) return SCEVAddExpr::get(SCEVMulExpr::get(LHSC, Add->getOperand(0)), SCEVMulExpr::get(LHSC, Add->getOperand(1))); ++Idx; while (SCEVConstant *RHSC = dyn_cast(Ops[Idx])) { // We found two constants, fold them together! Constant *Fold = ConstantInt::get(LHSC->getValue()->getValue() * RHSC->getValue()->getValue()); if (ConstantInt *CI = dyn_cast(Fold)) { Ops[0] = SCEVConstant::get(CI); Ops.erase(Ops.begin()+1); // Erase the folded element if (Ops.size() == 1) return Ops[0]; LHSC = cast(Ops[0]); } else { // If we couldn't fold the expression, move to the next constant. Note // that this is impossible to happen in practice because we always // constant fold constant ints to constant ints. ++Idx; } } // If we are left with a constant one being multiplied, strip it off. if (cast(Ops[0])->getValue()->equalsInt(1)) { Ops.erase(Ops.begin()); --Idx; } else if (cast(Ops[0])->getValue()->isZero()) { // If we have a multiply of zero, it will always be zero. return Ops[0]; } } // Skip over the add expression until we get to a multiply. while (Idx < Ops.size() && Ops[Idx]->getSCEVType() < scMulExpr) ++Idx; if (Ops.size() == 1) return Ops[0]; // If there are mul operands inline them all into this expression. if (Idx < Ops.size()) { bool DeletedMul = false; while (SCEVMulExpr *Mul = dyn_cast(Ops[Idx])) { // If we have an mul, expand the mul operands onto the end of the operands // list. Ops.insert(Ops.end(), Mul->op_begin(), Mul->op_end()); Ops.erase(Ops.begin()+Idx); DeletedMul = true; } // If we deleted at least one mul, we added operands to the end of the list, // and they are not necessarily sorted. Recurse to resort and resimplify // any operands we just aquired. if (DeletedMul) return get(Ops); } // If there are any add recurrences in the operands list, see if any other // added values are loop invariant. If so, we can fold them into the // recurrence. while (Idx < Ops.size() && Ops[Idx]->getSCEVType() < scAddRecExpr) ++Idx; // Scan over all recurrences, trying to fold loop invariants into them. for (; Idx < Ops.size() && isa(Ops[Idx]); ++Idx) { // Scan all of the other operands to this mul and add them to the vector if // they are loop invariant w.r.t. the recurrence. std::vector LIOps; SCEVAddRecExpr *AddRec = cast(Ops[Idx]); for (unsigned i = 0, e = Ops.size(); i != e; ++i) if (Ops[i]->isLoopInvariant(AddRec->getLoop())) { LIOps.push_back(Ops[i]); Ops.erase(Ops.begin()+i); --i; --e; } // If we found some loop invariants, fold them into the recurrence. if (!LIOps.empty()) { // NLI * LI * { Start,+,Step} --> NLI * { LI*Start,+,LI*Step } std::vector NewOps; NewOps.reserve(AddRec->getNumOperands()); if (LIOps.size() == 1) { SCEV *Scale = LIOps[0]; for (unsigned i = 0, e = AddRec->getNumOperands(); i != e; ++i) NewOps.push_back(SCEVMulExpr::get(Scale, AddRec->getOperand(i))); } else { for (unsigned i = 0, e = AddRec->getNumOperands(); i != e; ++i) { std::vector MulOps(LIOps); MulOps.push_back(AddRec->getOperand(i)); NewOps.push_back(SCEVMulExpr::get(MulOps)); } } SCEVHandle NewRec = SCEVAddRecExpr::get(NewOps, AddRec->getLoop()); // If all of the other operands were loop invariant, we are done. if (Ops.size() == 1) return NewRec; // Otherwise, multiply the folded AddRec by the non-liv parts. for (unsigned i = 0;; ++i) if (Ops[i] == AddRec) { Ops[i] = NewRec; break; } return SCEVMulExpr::get(Ops); } // Okay, if there weren't any loop invariants to be folded, check to see if // there are multiple AddRec's with the same loop induction variable being // multiplied together. If so, we can fold them. for (unsigned OtherIdx = Idx+1; OtherIdx < Ops.size() && isa(Ops[OtherIdx]);++OtherIdx) if (OtherIdx != Idx) { SCEVAddRecExpr *OtherAddRec = cast(Ops[OtherIdx]); if (AddRec->getLoop() == OtherAddRec->getLoop()) { // F * G --> {A,+,B} * {C,+,D} --> {A*C,+,F*D + G*B + B*D} SCEVAddRecExpr *F = AddRec, *G = OtherAddRec; SCEVHandle NewStart = SCEVMulExpr::get(F->getStart(), G->getStart()); SCEVHandle B = F->getStepRecurrence(); SCEVHandle D = G->getStepRecurrence(); SCEVHandle NewStep = SCEVAddExpr::get(SCEVMulExpr::get(F, D), SCEVMulExpr::get(G, B), SCEVMulExpr::get(B, D)); SCEVHandle NewAddRec = SCEVAddRecExpr::get(NewStart, NewStep, F->getLoop()); if (Ops.size() == 2) return NewAddRec; Ops.erase(Ops.begin()+Idx); Ops.erase(Ops.begin()+OtherIdx-1); Ops.push_back(NewAddRec); return SCEVMulExpr::get(Ops); } } // Otherwise couldn't fold anything into this recurrence. Move onto the // next one. } // Okay, it looks like we really DO need an mul expr. Check to see if we // already have one, otherwise create a new one. std::vector SCEVOps(Ops.begin(), Ops.end()); SCEVCommutativeExpr *&Result = (*SCEVCommExprs)[std::make_pair(scMulExpr, SCEVOps)]; if (Result == 0) Result = new SCEVMulExpr(Ops); return Result; } SCEVHandle SCEVSDivExpr::get(const SCEVHandle &LHS, const SCEVHandle &RHS) { if (SCEVConstant *RHSC = dyn_cast(RHS)) { if (RHSC->getValue()->equalsInt(1)) return LHS; // X sdiv 1 --> x if (RHSC->getValue()->isAllOnesValue()) return SCEV::getNegativeSCEV(LHS); // X sdiv -1 --> -x if (SCEVConstant *LHSC = dyn_cast(LHS)) { Constant *LHSCV = LHSC->getValue(); Constant *RHSCV = RHSC->getValue(); return SCEVUnknown::get(ConstantExpr::getSDiv(LHSCV, RHSCV)); } } // FIXME: implement folding of (X*4)/4 when we know X*4 doesn't overflow. SCEVSDivExpr *&Result = (*SCEVSDivs)[std::make_pair(LHS, RHS)]; if (Result == 0) Result = new SCEVSDivExpr(LHS, RHS); return Result; } /// SCEVAddRecExpr::get - Get a add recurrence expression for the /// specified loop. Simplify the expression as much as possible. SCEVHandle SCEVAddRecExpr::get(const SCEVHandle &Start, const SCEVHandle &Step, const Loop *L) { std::vector Operands; Operands.push_back(Start); if (SCEVAddRecExpr *StepChrec = dyn_cast(Step)) if (StepChrec->getLoop() == L) { Operands.insert(Operands.end(), StepChrec->op_begin(), StepChrec->op_end()); return get(Operands, L); } Operands.push_back(Step); return get(Operands, L); } /// SCEVAddRecExpr::get - Get a add recurrence expression for the /// specified loop. Simplify the expression as much as possible. SCEVHandle SCEVAddRecExpr::get(std::vector &Operands, const Loop *L) { if (Operands.size() == 1) return Operands[0]; if (SCEVConstant *StepC = dyn_cast(Operands.back())) if (StepC->getValue()->isZero()) { Operands.pop_back(); return get(Operands, L); // { X,+,0 } --> X } SCEVAddRecExpr *&Result = (*SCEVAddRecExprs)[std::make_pair(L, std::vector(Operands.begin(), Operands.end()))]; if (Result == 0) Result = new SCEVAddRecExpr(Operands, L); return Result; } SCEVHandle SCEVUnknown::get(Value *V) { if (ConstantInt *CI = dyn_cast(V)) return SCEVConstant::get(CI); SCEVUnknown *&Result = (*SCEVUnknowns)[V]; if (Result == 0) Result = new SCEVUnknown(V); return Result; } //===----------------------------------------------------------------------===// // ScalarEvolutionsImpl Definition and Implementation //===----------------------------------------------------------------------===// // /// ScalarEvolutionsImpl - This class implements the main driver for the scalar /// evolution code. /// namespace { struct VISIBILITY_HIDDEN ScalarEvolutionsImpl { /// F - The function we are analyzing. /// Function &F; /// LI - The loop information for the function we are currently analyzing. /// LoopInfo &LI; /// UnknownValue - This SCEV is used to represent unknown trip counts and /// things. SCEVHandle UnknownValue; /// Scalars - This is a cache of the scalars we have analyzed so far. /// std::map Scalars; /// IterationCounts - Cache the iteration count of the loops for this /// function as they are computed. std::map IterationCounts; /// ConstantEvolutionLoopExitValue - This map contains entries for all of /// the PHI instructions that we attempt to compute constant evolutions for. /// This allows us to avoid potentially expensive recomputation of these /// properties. An instruction maps to null if we are unable to compute its /// exit value. std::map ConstantEvolutionLoopExitValue; public: ScalarEvolutionsImpl(Function &f, LoopInfo &li) : F(f), LI(li), UnknownValue(new SCEVCouldNotCompute()) {} /// getSCEV - Return an existing SCEV if it exists, otherwise analyze the /// expression and create a new one. SCEVHandle getSCEV(Value *V); /// hasSCEV - Return true if the SCEV for this value has already been /// computed. bool hasSCEV(Value *V) const { return Scalars.count(V); } /// setSCEV - Insert the specified SCEV into the map of current SCEVs for /// the specified value. void setSCEV(Value *V, const SCEVHandle &H) { bool isNew = Scalars.insert(std::make_pair(V, H)).second; assert(isNew && "This entry already existed!"); } /// getSCEVAtScope - Compute the value of the specified expression within /// the indicated loop (which may be null to indicate in no loop). If the /// expression cannot be evaluated, return UnknownValue itself. SCEVHandle getSCEVAtScope(SCEV *V, const Loop *L); /// hasLoopInvariantIterationCount - Return true if the specified loop has /// an analyzable loop-invariant iteration count. bool hasLoopInvariantIterationCount(const Loop *L); /// getIterationCount - If the specified loop has a predictable iteration /// count, return it. Note that it is not valid to call this method on a /// loop without a loop-invariant iteration count. SCEVHandle getIterationCount(const Loop *L); /// deleteInstructionFromRecords - This method should be called by the /// client before it removes an instruction from the program, to make sure /// that no dangling references are left around. void deleteInstructionFromRecords(Instruction *I); private: /// createSCEV - We know that there is no SCEV for the specified value. /// Analyze the expression. SCEVHandle createSCEV(Value *V); /// createNodeForPHI - Provide the special handling we need to analyze PHI /// SCEVs. SCEVHandle createNodeForPHI(PHINode *PN); /// ReplaceSymbolicValueWithConcrete - This looks up the computed SCEV value /// for the specified instruction and replaces any references to the /// symbolic value SymName with the specified value. This is used during /// PHI resolution. void ReplaceSymbolicValueWithConcrete(Instruction *I, const SCEVHandle &SymName, const SCEVHandle &NewVal); /// ComputeIterationCount - Compute the number of times the specified loop /// will iterate. SCEVHandle ComputeIterationCount(const Loop *L); /// ComputeLoadConstantCompareIterationCount - Given an exit condition of /// 'setcc load X, cst', try to see if we can compute the trip count. SCEVHandle ComputeLoadConstantCompareIterationCount(LoadInst *LI, Constant *RHS, const Loop *L, ICmpInst::Predicate p); /// ComputeIterationCountExhaustively - If the trip is known to execute a /// constant number of times (the condition evolves only from constants), /// try to evaluate a few iterations of the loop until we get the exit /// condition gets a value of ExitWhen (true or false). If we cannot /// evaluate the trip count of the loop, return UnknownValue. SCEVHandle ComputeIterationCountExhaustively(const Loop *L, Value *Cond, bool ExitWhen); /// HowFarToZero - Return the number of times a backedge comparing the /// specified value to zero will execute. If not computable, return /// UnknownValue. SCEVHandle HowFarToZero(SCEV *V, const Loop *L); /// HowFarToNonZero - Return the number of times a backedge checking the /// specified value for nonzero will execute. If not computable, return /// UnknownValue. SCEVHandle HowFarToNonZero(SCEV *V, const Loop *L); /// HowManyLessThans - Return the number of times a backedge containing the /// specified less-than comparison will execute. If not computable, return /// UnknownValue. SCEVHandle HowManyLessThans(SCEV *LHS, SCEV *RHS, const Loop *L); /// getConstantEvolutionLoopExitValue - If we know that the specified Phi is /// in the header of its containing loop, we know the loop executes a /// constant number of times, and the PHI node is just a recurrence /// involving constants, fold it. Constant *getConstantEvolutionLoopExitValue(PHINode *PN, const APInt& Its, const Loop *L); }; } //===----------------------------------------------------------------------===// // Basic SCEV Analysis and PHI Idiom Recognition Code // /// deleteInstructionFromRecords - This method should be called by the /// client before it removes an instruction from the program, to make sure /// that no dangling references are left around. void ScalarEvolutionsImpl::deleteInstructionFromRecords(Instruction *I) { SmallVector Worklist; if (Scalars.erase(I)) { if (PHINode *PN = dyn_cast(I)) ConstantEvolutionLoopExitValue.erase(PN); Worklist.push_back(I); } while (!Worklist.empty()) { Instruction *II = Worklist.back(); Worklist.pop_back(); for (Instruction::use_iterator UI = II->use_begin(), UE = II->use_end(); UI != UE; ++UI) { Instruction *Inst = dyn_cast(*UI); if (Inst && hasSCEV(Inst) && Scalars.erase(Inst)) { if (PHINode *PN = dyn_cast(II)) ConstantEvolutionLoopExitValue.erase(PN); Worklist.push_back(Inst); } } } } /// getSCEV - Return an existing SCEV if it exists, otherwise analyze the /// expression and create a new one. SCEVHandle ScalarEvolutionsImpl::getSCEV(Value *V) { assert(V->getType() != Type::VoidTy && "Can't analyze void expressions!"); std::map::iterator I = Scalars.find(V); if (I != Scalars.end()) return I->second; SCEVHandle S = createSCEV(V); Scalars.insert(std::make_pair(V, S)); return S; } /// ReplaceSymbolicValueWithConcrete - This looks up the computed SCEV value for /// the specified instruction and replaces any references to the symbolic value /// SymName with the specified value. This is used during PHI resolution. void ScalarEvolutionsImpl:: ReplaceSymbolicValueWithConcrete(Instruction *I, const SCEVHandle &SymName, const SCEVHandle &NewVal) { std::map::iterator SI = Scalars.find(I); if (SI == Scalars.end()) return; SCEVHandle NV = SI->second->replaceSymbolicValuesWithConcrete(SymName, NewVal); if (NV == SI->second) return; // No change. SI->second = NV; // Update the scalars map! // Any instruction values that use this instruction might also need to be // updated! for (Value::use_iterator UI = I->use_begin(), E = I->use_end(); UI != E; ++UI) ReplaceSymbolicValueWithConcrete(cast(*UI), SymName, NewVal); } /// createNodeForPHI - PHI nodes have two cases. Either the PHI node exists in /// a loop header, making it a potential recurrence, or it doesn't. /// SCEVHandle ScalarEvolutionsImpl::createNodeForPHI(PHINode *PN) { if (PN->getNumIncomingValues() == 2) // The loops have been canonicalized. if (const Loop *L = LI.getLoopFor(PN->getParent())) if (L->getHeader() == PN->getParent()) { // If it lives in the loop header, it has two incoming values, one // from outside the loop, and one from inside. unsigned IncomingEdge = L->contains(PN->getIncomingBlock(0)); unsigned BackEdge = IncomingEdge^1; // While we are analyzing this PHI node, handle its value symbolically. SCEVHandle SymbolicName = SCEVUnknown::get(PN); assert(Scalars.find(PN) == Scalars.end() && "PHI node already processed?"); Scalars.insert(std::make_pair(PN, SymbolicName)); // Using this symbolic name for the PHI, analyze the value coming around // the back-edge. SCEVHandle BEValue = getSCEV(PN->getIncomingValue(BackEdge)); // NOTE: If BEValue is loop invariant, we know that the PHI node just // has a special value for the first iteration of the loop. // If the value coming around the backedge is an add with the symbolic // value we just inserted, then we found a simple induction variable! if (SCEVAddExpr *Add = dyn_cast(BEValue)) { // If there is a single occurrence of the symbolic value, replace it // with a recurrence. unsigned FoundIndex = Add->getNumOperands(); for (unsigned i = 0, e = Add->getNumOperands(); i != e; ++i) if (Add->getOperand(i) == SymbolicName) if (FoundIndex == e) { FoundIndex = i; break; } if (FoundIndex != Add->getNumOperands()) { // Create an add with everything but the specified operand. std::vector Ops; for (unsigned i = 0, e = Add->getNumOperands(); i != e; ++i) if (i != FoundIndex) Ops.push_back(Add->getOperand(i)); SCEVHandle Accum = SCEVAddExpr::get(Ops); // This is not a valid addrec if the step amount is varying each // loop iteration, but is not itself an addrec in this loop. if (Accum->isLoopInvariant(L) || (isa(Accum) && cast(Accum)->getLoop() == L)) { SCEVHandle StartVal = getSCEV(PN->getIncomingValue(IncomingEdge)); SCEVHandle PHISCEV = SCEVAddRecExpr::get(StartVal, Accum, L); // Okay, for the entire analysis of this edge we assumed the PHI // to be symbolic. We now need to go back and update all of the // entries for the scalars that use the PHI (except for the PHI // itself) to use the new analyzed value instead of the "symbolic" // value. ReplaceSymbolicValueWithConcrete(PN, SymbolicName, PHISCEV); return PHISCEV; } } } else if (SCEVAddRecExpr *AddRec = dyn_cast(BEValue)) { // Otherwise, this could be a loop like this: // i = 0; for (j = 1; ..; ++j) { .... i = j; } // In this case, j = {1,+,1} and BEValue is j. // Because the other in-value of i (0) fits the evolution of BEValue // i really is an addrec evolution. if (AddRec->getLoop() == L && AddRec->isAffine()) { SCEVHandle StartVal = getSCEV(PN->getIncomingValue(IncomingEdge)); // If StartVal = j.start - j.stride, we can use StartVal as the // initial step of the addrec evolution. if (StartVal == SCEV::getMinusSCEV(AddRec->getOperand(0), AddRec->getOperand(1))) { SCEVHandle PHISCEV = SCEVAddRecExpr::get(StartVal, AddRec->getOperand(1), L); // Okay, for the entire analysis of this edge we assumed the PHI // to be symbolic. We now need to go back and update all of the // entries for the scalars that use the PHI (except for the PHI // itself) to use the new analyzed value instead of the "symbolic" // value. ReplaceSymbolicValueWithConcrete(PN, SymbolicName, PHISCEV); return PHISCEV; } } } return SymbolicName; } // If it's not a loop phi, we can't handle it yet. return SCEVUnknown::get(PN); } /// GetConstantFactor - Determine the largest constant factor that S has. For /// example, turn {4,+,8} -> 4. (S umod result) should always equal zero. static APInt GetConstantFactor(SCEVHandle S) { if (SCEVConstant *C = dyn_cast(S)) { const APInt& V = C->getValue()->getValue(); if (!V.isMinValue()) return V; else // Zero is a multiple of everything. return APInt(C->getBitWidth(), 1).shl(C->getBitWidth()-1); } if (SCEVTruncateExpr *T = dyn_cast(S)) { return GetConstantFactor(T->getOperand()).trunc( cast(T->getType())->getBitWidth()); } if (SCEVZeroExtendExpr *E = dyn_cast(S)) return GetConstantFactor(E->getOperand()).zext( cast(E->getType())->getBitWidth()); if (SCEVAddExpr *A = dyn_cast(S)) { // The result is the min of all operands. APInt Res(GetConstantFactor(A->getOperand(0))); for (unsigned i = 1, e = A->getNumOperands(); i != e && Res.ugt(APInt(Res.getBitWidth(),1)); ++i) { APInt Tmp(GetConstantFactor(A->getOperand(i))); Res = APIntOps::umin(Res, Tmp); } return Res; } if (SCEVMulExpr *M = dyn_cast(S)) { // The result is the product of all the operands. APInt Res(GetConstantFactor(M->getOperand(0))); for (unsigned i = 1, e = M->getNumOperands(); i != e; ++i) { APInt Tmp(GetConstantFactor(M->getOperand(i))); Res *= Tmp; } return Res; } if (SCEVAddRecExpr *A = dyn_cast(S)) { // For now, we just handle linear expressions. if (A->getNumOperands() == 2) { // We want the GCD between the start and the stride value. APInt Start(GetConstantFactor(A->getOperand(0))); if (Start == 1) return Start; APInt Stride(GetConstantFactor(A->getOperand(1))); return APIntOps::GreatestCommonDivisor(Start, Stride); } } // SCEVSDivExpr, SCEVUnknown. return APInt(S->getBitWidth(), 1); } /// createSCEV - We know that there is no SCEV for the specified value. /// Analyze the expression. /// SCEVHandle ScalarEvolutionsImpl::createSCEV(Value *V) { if (Instruction *I = dyn_cast(V)) { switch (I->getOpcode()) { case Instruction::Add: return SCEVAddExpr::get(getSCEV(I->getOperand(0)), getSCEV(I->getOperand(1))); case Instruction::Mul: return SCEVMulExpr::get(getSCEV(I->getOperand(0)), getSCEV(I->getOperand(1))); case Instruction::SDiv: return SCEVSDivExpr::get(getSCEV(I->getOperand(0)), getSCEV(I->getOperand(1))); break; case Instruction::Sub: return SCEV::getMinusSCEV(getSCEV(I->getOperand(0)), getSCEV(I->getOperand(1))); case Instruction::Or: // If the RHS of the Or is a constant, we may have something like: // X*4+1 which got turned into X*4|1. Handle this as an add so loop // optimizations will transparently handle this case. if (ConstantInt *CI = dyn_cast(I->getOperand(1))) { SCEVHandle LHS = getSCEV(I->getOperand(0)); APInt CommonFact(GetConstantFactor(LHS)); assert(!CommonFact.isMinValue() && "Common factor should at least be 1!"); if (CommonFact.ugt(CI->getValue())) { // If the LHS is a multiple that is larger than the RHS, use +. return SCEVAddExpr::get(LHS, getSCEV(I->getOperand(1))); } } break; case Instruction::Xor: // If the RHS of the xor is a signbit, then this is just an add. // Instcombine turns add of signbit into xor as a strength reduction step. if (ConstantInt *CI = dyn_cast(I->getOperand(1))) { if (CI->getValue().isSignBit()) return SCEVAddExpr::get(getSCEV(I->getOperand(0)), getSCEV(I->getOperand(1))); } break; case Instruction::Shl: // Turn shift left of a constant amount into a multiply. if (ConstantInt *SA = dyn_cast(I->getOperand(1))) { uint32_t BitWidth = cast(V->getType())->getBitWidth(); Constant *X = ConstantInt::get( APInt(BitWidth, 1).shl(SA->getLimitedValue(BitWidth))); return SCEVMulExpr::get(getSCEV(I->getOperand(0)), getSCEV(X)); } break; case Instruction::Trunc: return SCEVTruncateExpr::get(getSCEV(I->getOperand(0)), I->getType()); case Instruction::ZExt: return SCEVZeroExtendExpr::get(getSCEV(I->getOperand(0)), I->getType()); case Instruction::BitCast: // BitCasts are no-op casts so we just eliminate the cast. if (I->getType()->isInteger() && I->getOperand(0)->getType()->isInteger()) return getSCEV(I->getOperand(0)); break; case Instruction::PHI: return createNodeForPHI(cast(I)); default: // We cannot analyze this expression. break; } } return SCEVUnknown::get(V); } //===----------------------------------------------------------------------===// // Iteration Count Computation Code // /// getIterationCount - If the specified loop has a predictable iteration /// count, return it. Note that it is not valid to call this method on a /// loop without a loop-invariant iteration count. SCEVHandle ScalarEvolutionsImpl::getIterationCount(const Loop *L) { std::map::iterator I = IterationCounts.find(L); if (I == IterationCounts.end()) { SCEVHandle ItCount = ComputeIterationCount(L); I = IterationCounts.insert(std::make_pair(L, ItCount)).first; if (ItCount != UnknownValue) { assert(ItCount->isLoopInvariant(L) && "Computed trip count isn't loop invariant for loop!"); ++NumTripCountsComputed; } else if (isa(L->getHeader()->begin())) { // Only count loops that have phi nodes as not being computable. ++NumTripCountsNotComputed; } } return I->second; } /// ComputeIterationCount - Compute the number of times the specified loop /// will iterate. SCEVHandle ScalarEvolutionsImpl::ComputeIterationCount(const Loop *L) { // If the loop has a non-one exit block count, we can't analyze it. std::vector ExitBlocks; L->getExitBlocks(ExitBlocks); if (ExitBlocks.size() != 1) return UnknownValue; // Okay, there is one exit block. Try to find the condition that causes the // loop to be exited. BasicBlock *ExitBlock = ExitBlocks[0]; BasicBlock *ExitingBlock = 0; for (pred_iterator PI = pred_begin(ExitBlock), E = pred_end(ExitBlock); PI != E; ++PI) if (L->contains(*PI)) { if (ExitingBlock == 0) ExitingBlock = *PI; else return UnknownValue; // More than one block exiting! } assert(ExitingBlock && "No exits from loop, something is broken!"); // Okay, we've computed the exiting block. See what condition causes us to // exit. // // FIXME: we should be able to handle switch instructions (with a single exit) BranchInst *ExitBr = dyn_cast(ExitingBlock->getTerminator()); if (ExitBr == 0) return UnknownValue; assert(ExitBr->isConditional() && "If unconditional, it can't be in loop!"); // At this point, we know we have a conditional branch that determines whether // the loop is exited. However, we don't know if the branch is executed each // time through the loop. If not, then the execution count of the branch will // not be equal to the trip count of the loop. // // Currently we check for this by checking to see if the Exit branch goes to // the loop header. If so, we know it will always execute the same number of // times as the loop. We also handle the case where the exit block *is* the // loop header. This is common for un-rotated loops. More extensive analysis // could be done to handle more cases here. if (ExitBr->getSuccessor(0) != L->getHeader() && ExitBr->getSuccessor(1) != L->getHeader() && ExitBr->getParent() != L->getHeader()) return UnknownValue; ICmpInst *ExitCond = dyn_cast(ExitBr->getCondition()); // If its not an integer comparison then compute it the hard way. // Note that ICmpInst deals with pointer comparisons too so we must check // the type of the operand. if (ExitCond == 0 || isa(ExitCond->getOperand(0)->getType())) return ComputeIterationCountExhaustively(L, ExitBr->getCondition(), ExitBr->getSuccessor(0) == ExitBlock); // If the condition was exit on true, convert the condition to exit on false ICmpInst::Predicate Cond; if (ExitBr->getSuccessor(1) == ExitBlock) Cond = ExitCond->getPredicate(); else Cond = ExitCond->getInversePredicate(); // Handle common loops like: for (X = "string"; *X; ++X) if (LoadInst *LI = dyn_cast(ExitCond->getOperand(0))) if (Constant *RHS = dyn_cast(ExitCond->getOperand(1))) { SCEVHandle ItCnt = ComputeLoadConstantCompareIterationCount(LI, RHS, L, Cond); if (!isa(ItCnt)) return ItCnt; } SCEVHandle LHS = getSCEV(ExitCond->getOperand(0)); SCEVHandle RHS = getSCEV(ExitCond->getOperand(1)); // Try to evaluate any dependencies out of the loop. SCEVHandle Tmp = getSCEVAtScope(LHS, L); if (!isa(Tmp)) LHS = Tmp; Tmp = getSCEVAtScope(RHS, L); if (!isa(Tmp)) RHS = Tmp; // At this point, we would like to compute how many iterations of the // loop the predicate will return true for these inputs. if (isa(LHS) && !isa(RHS)) { // If there is a constant, force it into the RHS. std::swap(LHS, RHS); Cond = ICmpInst::getSwappedPredicate(Cond); } // FIXME: think about handling pointer comparisons! i.e.: // while (P != P+100) ++P; // If we have a comparison of a chrec against a constant, try to use value // ranges to answer this query. if (SCEVConstant *RHSC = dyn_cast(RHS)) if (SCEVAddRecExpr *AddRec = dyn_cast(LHS)) if (AddRec->getLoop() == L) { // Form the comparison range using the constant of the correct type so // that the ConstantRange class knows to do a signed or unsigned // comparison. ConstantInt *CompVal = RHSC->getValue(); const Type *RealTy = ExitCond->getOperand(0)->getType(); CompVal = dyn_cast( ConstantExpr::getBitCast(CompVal, RealTy)); if (CompVal) { // Form the constant range. ConstantRange CompRange( ICmpInst::makeConstantRange(Cond, CompVal->getValue())); SCEVHandle Ret = AddRec->getNumIterationsInRange(CompRange, false /*Always treat as unsigned range*/); if (!isa(Ret)) return Ret; } } switch (Cond) { case ICmpInst::ICMP_NE: { // while (X != Y) // Convert to: while (X-Y != 0) SCEVHandle TC = HowFarToZero(SCEV::getMinusSCEV(LHS, RHS), L); if (!isa(TC)) return TC; break; } case ICmpInst::ICMP_EQ: { // Convert to: while (X-Y == 0) // while (X == Y) SCEVHandle TC = HowFarToNonZero(SCEV::getMinusSCEV(LHS, RHS), L); if (!isa(TC)) return TC; break; } case ICmpInst::ICMP_SLT: { SCEVHandle TC = HowManyLessThans(LHS, RHS, L); if (!isa(TC)) return TC; break; } case ICmpInst::ICMP_SGT: { SCEVHandle TC = HowManyLessThans(RHS, LHS, L); if (!isa(TC)) return TC; break; } default: #if 0 cerr << "ComputeIterationCount "; if (ExitCond->getOperand(0)->getType()->isUnsigned()) cerr << "[unsigned] "; cerr << *LHS << " " << Instruction::getOpcodeName(Instruction::ICmp) << " " << *RHS << "\n"; #endif break; } return ComputeIterationCountExhaustively(L, ExitCond, ExitBr->getSuccessor(0) == ExitBlock); } static ConstantInt * EvaluateConstantChrecAtConstant(const SCEVAddRecExpr *AddRec, Constant *C) { SCEVHandle InVal = SCEVConstant::get(cast(C)); SCEVHandle Val = AddRec->evaluateAtIteration(InVal); assert(isa(Val) && "Evaluation of SCEV at constant didn't fold correctly?"); return cast(Val)->getValue(); } /// GetAddressedElementFromGlobal - Given a global variable with an initializer /// and a GEP expression (missing the pointer index) indexing into it, return /// the addressed element of the initializer or null if the index expression is /// invalid. static Constant * GetAddressedElementFromGlobal(GlobalVariable *GV, const std::vector &Indices) { Constant *Init = GV->getInitializer(); for (unsigned i = 0, e = Indices.size(); i != e; ++i) { uint64_t Idx = Indices[i]->getZExtValue(); if (ConstantStruct *CS = dyn_cast(Init)) { assert(Idx < CS->getNumOperands() && "Bad struct index!"); Init = cast(CS->getOperand(Idx)); } else if (ConstantArray *CA = dyn_cast(Init)) { if (Idx >= CA->getNumOperands()) return 0; // Bogus program Init = cast(CA->getOperand(Idx)); } else if (isa(Init)) { if (const StructType *STy = dyn_cast(Init->getType())) { assert(Idx < STy->getNumElements() && "Bad struct index!"); Init = Constant::getNullValue(STy->getElementType(Idx)); } else if (const ArrayType *ATy = dyn_cast(Init->getType())) { if (Idx >= ATy->getNumElements()) return 0; // Bogus program Init = Constant::getNullValue(ATy->getElementType()); } else { assert(0 && "Unknown constant aggregate type!"); } return 0; } else { return 0; // Unknown initializer type } } return Init; } /// ComputeLoadConstantCompareIterationCount - Given an exit condition of /// 'setcc load X, cst', try to se if we can compute the trip count. SCEVHandle ScalarEvolutionsImpl:: ComputeLoadConstantCompareIterationCount(LoadInst *LI, Constant *RHS, const Loop *L, ICmpInst::Predicate predicate) { if (LI->isVolatile()) return UnknownValue; // Check to see if the loaded pointer is a getelementptr of a global. GetElementPtrInst *GEP = dyn_cast(LI->getOperand(0)); if (!GEP) return UnknownValue; // Make sure that it is really a constant global we are gepping, with an // initializer, and make sure the first IDX is really 0. GlobalVariable *GV = dyn_cast(GEP->getOperand(0)); if (!GV || !GV->isConstant() || !GV->hasInitializer() || GEP->getNumOperands() < 3 || !isa(GEP->getOperand(1)) || !cast(GEP->getOperand(1))->isNullValue()) return UnknownValue; // Okay, we allow one non-constant index into the GEP instruction. Value *VarIdx = 0; std::vector Indexes; unsigned VarIdxNum = 0; for (unsigned i = 2, e = GEP->getNumOperands(); i != e; ++i) if (ConstantInt *CI = dyn_cast(GEP->getOperand(i))) { Indexes.push_back(CI); } else if (!isa(GEP->getOperand(i))) { if (VarIdx) return UnknownValue; // Multiple non-constant idx's. VarIdx = GEP->getOperand(i); VarIdxNum = i-2; Indexes.push_back(0); } // Okay, we know we have a (load (gep GV, 0, X)) comparison with a constant. // Check to see if X is a loop variant variable value now. SCEVHandle Idx = getSCEV(VarIdx); SCEVHandle Tmp = getSCEVAtScope(Idx, L); if (!isa(Tmp)) Idx = Tmp; // We can only recognize very limited forms of loop index expressions, in // particular, only affine AddRec's like {C1,+,C2}. SCEVAddRecExpr *IdxExpr = dyn_cast(Idx); if (!IdxExpr || !IdxExpr->isAffine() || IdxExpr->isLoopInvariant(L) || !isa(IdxExpr->getOperand(0)) || !isa(IdxExpr->getOperand(1))) return UnknownValue; unsigned MaxSteps = MaxBruteForceIterations; for (unsigned IterationNum = 0; IterationNum != MaxSteps; ++IterationNum) { ConstantInt *ItCst = ConstantInt::get(IdxExpr->getType(), IterationNum); ConstantInt *Val = EvaluateConstantChrecAtConstant(IdxExpr, ItCst); // Form the GEP offset. Indexes[VarIdxNum] = Val; Constant *Result = GetAddressedElementFromGlobal(GV, Indexes); if (Result == 0) break; // Cannot compute! // Evaluate the condition for this iteration. Result = ConstantExpr::getICmp(predicate, Result, RHS); if (!isa(Result)) break; // Couldn't decide for sure if (cast(Result)->getValue().isMinValue()) { #if 0 cerr << "\n***\n*** Computed loop count " << *ItCst << "\n*** From global " << *GV << "*** BB: " << *L->getHeader() << "***\n"; #endif ++NumArrayLenItCounts; return SCEVConstant::get(ItCst); // Found terminating iteration! } } return UnknownValue; } /// CanConstantFold - Return true if we can constant fold an instruction of the /// specified type, assuming that all operands were constants. static bool CanConstantFold(const Instruction *I) { if (isa(I) || isa(I) || isa(I) || isa(I) || isa(I)) return true; if (const CallInst *CI = dyn_cast(I)) if (const Function *F = CI->getCalledFunction()) return canConstantFoldCallTo((Function*)F); // FIXME: elim cast return false; } /// getConstantEvolvingPHI - Given an LLVM value and a loop, return a PHI node /// in the loop that V is derived from. We allow arbitrary operations along the /// way, but the operands of an operation must either be constants or a value /// derived from a constant PHI. If this expression does not fit with these /// constraints, return null. static PHINode *getConstantEvolvingPHI(Value *V, const Loop *L) { // If this is not an instruction, or if this is an instruction outside of the // loop, it can't be derived from a loop PHI. Instruction *I = dyn_cast(V); if (I == 0 || !L->contains(I->getParent())) return 0; if (PHINode *PN = dyn_cast(I)) if (L->getHeader() == I->getParent()) return PN; else // We don't currently keep track of the control flow needed to evaluate // PHIs, so we cannot handle PHIs inside of loops. return 0; // If we won't be able to constant fold this expression even if the operands // are constants, return early. if (!CanConstantFold(I)) return 0; // Otherwise, we can evaluate this instruction if all of its operands are // constant or derived from a PHI node themselves. PHINode *PHI = 0; for (unsigned Op = 0, e = I->getNumOperands(); Op != e; ++Op) if (!(isa(I->getOperand(Op)) || isa(I->getOperand(Op)))) { PHINode *P = getConstantEvolvingPHI(I->getOperand(Op), L); if (P == 0) return 0; // Not evolving from PHI if (PHI == 0) PHI = P; else if (PHI != P) return 0; // Evolving from multiple different PHIs. } // This is a expression evolving from a constant PHI! return PHI; } /// EvaluateExpression - Given an expression that passes the /// getConstantEvolvingPHI predicate, evaluate its value assuming the PHI node /// in the loop has the value PHIVal. If we can't fold this expression for some /// reason, return null. static Constant *EvaluateExpression(Value *V, Constant *PHIVal) { if (isa(V)) return PHIVal; if (GlobalValue *GV = dyn_cast(V)) return GV; if (Constant *C = dyn_cast(V)) return C; Instruction *I = cast(V); std::vector Operands; Operands.resize(I->getNumOperands()); for (unsigned i = 0, e = I->getNumOperands(); i != e; ++i) { Operands[i] = EvaluateExpression(I->getOperand(i), PHIVal); if (Operands[i] == 0) return 0; } return ConstantFoldInstOperands(I, &Operands[0], Operands.size()); } /// getConstantEvolutionLoopExitValue - If we know that the specified Phi is /// in the header of its containing loop, we know the loop executes a /// constant number of times, and the PHI node is just a recurrence /// involving constants, fold it. Constant *ScalarEvolutionsImpl:: getConstantEvolutionLoopExitValue(PHINode *PN, const APInt& Its, const Loop *L){ std::map::iterator I = ConstantEvolutionLoopExitValue.find(PN); if (I != ConstantEvolutionLoopExitValue.end()) return I->second; if (Its.ugt(APInt(Its.getBitWidth(),MaxBruteForceIterations))) return ConstantEvolutionLoopExitValue[PN] = 0; // Not going to evaluate it. Constant *&RetVal = ConstantEvolutionLoopExitValue[PN]; // Since the loop is canonicalized, the PHI node must have two entries. One // entry must be a constant (coming in from outside of the loop), and the // second must be derived from the same PHI. bool SecondIsBackedge = L->contains(PN->getIncomingBlock(1)); Constant *StartCST = dyn_cast(PN->getIncomingValue(!SecondIsBackedge)); if (StartCST == 0) return RetVal = 0; // Must be a constant. Value *BEValue = PN->getIncomingValue(SecondIsBackedge); PHINode *PN2 = getConstantEvolvingPHI(BEValue, L); if (PN2 != PN) return RetVal = 0; // Not derived from same PHI. // Execute the loop symbolically to determine the exit value. if (Its.getActiveBits() >= 32) return RetVal = 0; // More than 2^32-1 iterations?? Not doing it! unsigned NumIterations = Its.getZExtValue(); // must be in range unsigned IterationNum = 0; for (Constant *PHIVal = StartCST; ; ++IterationNum) { if (IterationNum == NumIterations) return RetVal = PHIVal; // Got exit value! // Compute the value of the PHI node for the next iteration. Constant *NextPHI = EvaluateExpression(BEValue, PHIVal); if (NextPHI == PHIVal) return RetVal = NextPHI; // Stopped evolving! if (NextPHI == 0) return 0; // Couldn't evaluate! PHIVal = NextPHI; } } /// ComputeIterationCountExhaustively - If the trip is known to execute a /// constant number of times (the condition evolves only from constants), /// try to evaluate a few iterations of the loop until we get the exit /// condition gets a value of ExitWhen (true or false). If we cannot /// evaluate the trip count of the loop, return UnknownValue. SCEVHandle ScalarEvolutionsImpl:: ComputeIterationCountExhaustively(const Loop *L, Value *Cond, bool ExitWhen) { PHINode *PN = getConstantEvolvingPHI(Cond, L); if (PN == 0) return UnknownValue; // Since the loop is canonicalized, the PHI node must have two entries. One // entry must be a constant (coming in from outside of the loop), and the // second must be derived from the same PHI. bool SecondIsBackedge = L->contains(PN->getIncomingBlock(1)); Constant *StartCST = dyn_cast(PN->getIncomingValue(!SecondIsBackedge)); if (StartCST == 0) return UnknownValue; // Must be a constant. Value *BEValue = PN->getIncomingValue(SecondIsBackedge); PHINode *PN2 = getConstantEvolvingPHI(BEValue, L); if (PN2 != PN) return UnknownValue; // Not derived from same PHI. // Okay, we find a PHI node that defines the trip count of this loop. Execute // the loop symbolically to determine when the condition gets a value of // "ExitWhen". unsigned IterationNum = 0; unsigned MaxIterations = MaxBruteForceIterations; // Limit analysis. for (Constant *PHIVal = StartCST; IterationNum != MaxIterations; ++IterationNum) { ConstantInt *CondVal = dyn_cast_or_null(EvaluateExpression(Cond, PHIVal)); // Couldn't symbolically evaluate. if (!CondVal) return UnknownValue; if (CondVal->getValue() == uint64_t(ExitWhen)) { ConstantEvolutionLoopExitValue[PN] = PHIVal; ++NumBruteForceTripCountsComputed; return SCEVConstant::get(ConstantInt::get(Type::Int32Ty, IterationNum)); } // Compute the value of the PHI node for the next iteration. Constant *NextPHI = EvaluateExpression(BEValue, PHIVal); if (NextPHI == 0 || NextPHI == PHIVal) return UnknownValue; // Couldn't evaluate or not making progress... PHIVal = NextPHI; } // Too many iterations were needed to evaluate. return UnknownValue; } /// getSCEVAtScope - Compute the value of the specified expression within the /// indicated loop (which may be null to indicate in no loop). If the /// expression cannot be evaluated, return UnknownValue. SCEVHandle ScalarEvolutionsImpl::getSCEVAtScope(SCEV *V, const Loop *L) { // FIXME: this should be turned into a virtual method on SCEV! if (isa(V)) return V; // If this instruction is evolves from a constant-evolving PHI, compute the // exit value from the loop without using SCEVs. if (SCEVUnknown *SU = dyn_cast(V)) { if (Instruction *I = dyn_cast(SU->getValue())) { const Loop *LI = this->LI[I->getParent()]; if (LI && LI->getParentLoop() == L) // Looking for loop exit value. if (PHINode *PN = dyn_cast(I)) if (PN->getParent() == LI->getHeader()) { // Okay, there is no closed form solution for the PHI node. Check // to see if the loop that contains it has a known iteration count. // If so, we may be able to force computation of the exit value. SCEVHandle IterationCount = getIterationCount(LI); if (SCEVConstant *ICC = dyn_cast(IterationCount)) { // Okay, we know how many times the containing loop executes. If // this is a constant evolving PHI node, get the final value at // the specified iteration number. Constant *RV = getConstantEvolutionLoopExitValue(PN, ICC->getValue()->getValue(), LI); if (RV) return SCEVUnknown::get(RV); } } // Okay, this is an expression that we cannot symbolically evaluate // into a SCEV. Check to see if it's possible to symbolically evaluate // the arguments into constants, and if so, try to constant propagate the // result. This is particularly useful for computing loop exit values. if (CanConstantFold(I)) { std::vector Operands; Operands.reserve(I->getNumOperands()); for (unsigned i = 0, e = I->getNumOperands(); i != e; ++i) { Value *Op = I->getOperand(i); if (Constant *C = dyn_cast(Op)) { Operands.push_back(C); } else { SCEVHandle OpV = getSCEVAtScope(getSCEV(Op), L); if (SCEVConstant *SC = dyn_cast(OpV)) Operands.push_back(ConstantExpr::getIntegerCast(SC->getValue(), Op->getType(), false)); else if (SCEVUnknown *SU = dyn_cast(OpV)) { if (Constant *C = dyn_cast(SU->getValue())) Operands.push_back(ConstantExpr::getIntegerCast(C, Op->getType(), false)); else return V; } else { return V; } } } Constant *C =ConstantFoldInstOperands(I, &Operands[0], Operands.size()); return SCEVUnknown::get(C); } } // This is some other type of SCEVUnknown, just return it. return V; } if (SCEVCommutativeExpr *Comm = dyn_cast(V)) { // Avoid performing the look-up in the common case where the specified // expression has no loop-variant portions. for (unsigned i = 0, e = Comm->getNumOperands(); i != e; ++i) { SCEVHandle OpAtScope = getSCEVAtScope(Comm->getOperand(i), L); if (OpAtScope != Comm->getOperand(i)) { if (OpAtScope == UnknownValue) return UnknownValue; // Okay, at least one of these operands is loop variant but might be // foldable. Build a new instance of the folded commutative expression. std::vector NewOps(Comm->op_begin(), Comm->op_begin()+i); NewOps.push_back(OpAtScope); for (++i; i != e; ++i) { OpAtScope = getSCEVAtScope(Comm->getOperand(i), L); if (OpAtScope == UnknownValue) return UnknownValue; NewOps.push_back(OpAtScope); } if (isa(Comm)) return SCEVAddExpr::get(NewOps); assert(isa(Comm) && "Only know about add and mul!"); return SCEVMulExpr::get(NewOps); } } // If we got here, all operands are loop invariant. return Comm; } if (SCEVSDivExpr *Div = dyn_cast(V)) { SCEVHandle LHS = getSCEVAtScope(Div->getLHS(), L); if (LHS == UnknownValue) return LHS; SCEVHandle RHS = getSCEVAtScope(Div->getRHS(), L); if (RHS == UnknownValue) return RHS; if (LHS == Div->getLHS() && RHS == Div->getRHS()) return Div; // must be loop invariant return SCEVSDivExpr::get(LHS, RHS); } // If this is a loop recurrence for a loop that does not contain L, then we // are dealing with the final value computed by the loop. if (SCEVAddRecExpr *AddRec = dyn_cast(V)) { if (!L || !AddRec->getLoop()->contains(L->getHeader())) { // To evaluate this recurrence, we need to know how many times the AddRec // loop iterates. Compute this now. SCEVHandle IterationCount = getIterationCount(AddRec->getLoop()); if (IterationCount == UnknownValue) return UnknownValue; IterationCount = getTruncateOrZeroExtend(IterationCount, AddRec->getType()); // If the value is affine, simplify the expression evaluation to just // Start + Step*IterationCount. if (AddRec->isAffine()) return SCEVAddExpr::get(AddRec->getStart(), SCEVMulExpr::get(IterationCount, AddRec->getOperand(1))); // Otherwise, evaluate it the hard way. return AddRec->evaluateAtIteration(IterationCount); } return UnknownValue; } //assert(0 && "Unknown SCEV type!"); return UnknownValue; } /// SolveQuadraticEquation - Find the roots of the quadratic equation for the /// given quadratic chrec {L,+,M,+,N}. This returns either the two roots (which /// might be the same) or two SCEVCouldNotCompute objects. /// static std::pair SolveQuadraticEquation(const SCEVAddRecExpr *AddRec) { assert(AddRec->getNumOperands() == 3 && "This is not a quadratic chrec!"); SCEVConstant *LC = dyn_cast(AddRec->getOperand(0)); SCEVConstant *MC = dyn_cast(AddRec->getOperand(1)); SCEVConstant *NC = dyn_cast(AddRec->getOperand(2)); // We currently can only solve this if the coefficients are constants. if (!LC || !MC || !NC) { SCEV *CNC = new SCEVCouldNotCompute(); return std::make_pair(CNC, CNC); } uint32_t BitWidth = LC->getValue()->getValue().getBitWidth(); const APInt &L = LC->getValue()->getValue(); const APInt &M = MC->getValue()->getValue(); const APInt &N = NC->getValue()->getValue(); APInt Two(BitWidth, 2); APInt Four(BitWidth, 4); { using namespace APIntOps; const APInt& C = L; // Convert from chrec coefficients to polynomial coefficients AX^2+BX+C // The B coefficient is M-N/2 APInt B(M); B -= sdiv(N,Two); // The A coefficient is N/2 APInt A(N.sdiv(Two)); // Compute the B^2-4ac term. APInt SqrtTerm(B); SqrtTerm *= B; SqrtTerm -= Four * (A * C); // Compute sqrt(B^2-4ac). This is guaranteed to be the nearest // integer value or else APInt::sqrt() will assert. APInt SqrtVal(SqrtTerm.sqrt()); // Compute the two solutions for the quadratic formula. // The divisions must be performed as signed divisions. APInt NegB(-B); APInt TwoA( A << 1 ); ConstantInt *Solution1 = ConstantInt::get((NegB + SqrtVal).sdiv(TwoA)); ConstantInt *Solution2 = ConstantInt::get((NegB - SqrtVal).sdiv(TwoA)); return std::make_pair(SCEVUnknown::get(Solution1), SCEVUnknown::get(Solution2)); } // end APIntOps namespace } /// HowFarToZero - Return the number of times a backedge comparing the specified /// value to zero will execute. If not computable, return UnknownValue SCEVHandle ScalarEvolutionsImpl::HowFarToZero(SCEV *V, const Loop *L) { // If the value is a constant if (SCEVConstant *C = dyn_cast(V)) { // If the value is already zero, the branch will execute zero times. if (C->getValue()->isZero()) return C; return UnknownValue; // Otherwise it will loop infinitely. } SCEVAddRecExpr *AddRec = dyn_cast(V); if (!AddRec || AddRec->getLoop() != L) return UnknownValue; if (AddRec->isAffine()) { // If this is an affine expression the execution count of this branch is // equal to: // // (0 - Start/Step) iff Start % Step == 0 // // Get the initial value for the loop. SCEVHandle Start = getSCEVAtScope(AddRec->getStart(), L->getParentLoop()); if (isa(Start)) return UnknownValue; SCEVHandle Step = AddRec->getOperand(1); Step = getSCEVAtScope(Step, L->getParentLoop()); // Figure out if Start % Step == 0. // FIXME: We should add DivExpr and RemExpr operations to our AST. if (SCEVConstant *StepC = dyn_cast(Step)) { if (StepC->getValue()->equalsInt(1)) // N % 1 == 0 return SCEV::getNegativeSCEV(Start); // 0 - Start/1 == -Start if (StepC->getValue()->isAllOnesValue()) // N % -1 == 0 return Start; // 0 - Start/-1 == Start // Check to see if Start is divisible by SC with no remainder. if (SCEVConstant *StartC = dyn_cast(Start)) { ConstantInt *StartCC = StartC->getValue(); Constant *StartNegC = ConstantExpr::getNeg(StartCC); Constant *Rem = ConstantExpr::getSRem(StartNegC, StepC->getValue()); if (Rem->isNullValue()) { Constant *Result =ConstantExpr::getSDiv(StartNegC,StepC->getValue()); return SCEVUnknown::get(Result); } } } } else if (AddRec->isQuadratic() && AddRec->getType()->isInteger()) { // If this is a quadratic (3-term) AddRec {L,+,M,+,N}, find the roots of // the quadratic equation to solve it. std::pair Roots = SolveQuadraticEquation(AddRec); SCEVConstant *R1 = dyn_cast(Roots.first); SCEVConstant *R2 = dyn_cast(Roots.second); if (R1) { #if 0 cerr << "HFTZ: " << *V << " - sol#1: " << *R1 << " sol#2: " << *R2 << "\n"; #endif // Pick the smallest positive root value. if (ConstantInt *CB = dyn_cast(ConstantExpr::getICmp(ICmpInst::ICMP_ULT, R1->getValue(), R2->getValue()))) { if (CB->getZExtValue() == false) std::swap(R1, R2); // R1 is the minimum root now. // We can only use this value if the chrec ends up with an exact zero // value at this index. When solving for "X*X != 5", for example, we // should not accept a root of 2. SCEVHandle Val = AddRec->evaluateAtIteration(R1); if (SCEVConstant *EvalVal = dyn_cast(Val)) if (EvalVal->getValue()->isZero()) return R1; // We found a quadratic root! } } } return UnknownValue; } /// HowFarToNonZero - Return the number of times a backedge checking the /// specified value for nonzero will execute. If not computable, return /// UnknownValue SCEVHandle ScalarEvolutionsImpl::HowFarToNonZero(SCEV *V, const Loop *L) { // Loops that look like: while (X == 0) are very strange indeed. We don't // handle them yet except for the trivial case. This could be expanded in the // future as needed. // If the value is a constant, check to see if it is known to be non-zero // already. If so, the backedge will execute zero times. if (SCEVConstant *C = dyn_cast(V)) { Constant *Zero = Constant::getNullValue(C->getValue()->getType()); Constant *NonZero = ConstantExpr::getICmp(ICmpInst::ICMP_NE, C->getValue(), Zero); if (NonZero == ConstantInt::getTrue()) return getSCEV(Zero); return UnknownValue; // Otherwise it will loop infinitely. } // We could implement others, but I really doubt anyone writes loops like // this, and if they did, they would already be constant folded. return UnknownValue; } /// HowManyLessThans - Return the number of times a backedge containing the /// specified less-than comparison will execute. If not computable, return /// UnknownValue. SCEVHandle ScalarEvolutionsImpl:: HowManyLessThans(SCEV *LHS, SCEV *RHS, const Loop *L) { // Only handle: "ADDREC < LoopInvariant". if (!RHS->isLoopInvariant(L)) return UnknownValue; SCEVAddRecExpr *AddRec = dyn_cast(LHS); if (!AddRec || AddRec->getLoop() != L) return UnknownValue; if (AddRec->isAffine()) { // FORNOW: We only support unit strides. SCEVHandle One = SCEVUnknown::getIntegerSCEV(1, RHS->getType()); if (AddRec->getOperand(1) != One) return UnknownValue; // The number of iterations for "[n,+,1] < m", is m-n. However, we don't // know that m is >= n on input to the loop. If it is, the condition return // true zero times. What we really should return, for full generality, is // SMAX(0, m-n). Since we cannot check this, we will instead check for a // canonical loop form: most do-loops will have a check that dominates the // loop, that only enters the loop if [n-1]= n. // Search for the check. BasicBlock *Preheader = L->getLoopPreheader(); BasicBlock *PreheaderDest = L->getHeader(); if (Preheader == 0) return UnknownValue; BranchInst *LoopEntryPredicate = dyn_cast(Preheader->getTerminator()); if (!LoopEntryPredicate) return UnknownValue; // This might be a critical edge broken out. If the loop preheader ends in // an unconditional branch to the loop, check to see if the preheader has a // single predecessor, and if so, look for its terminator. while (LoopEntryPredicate->isUnconditional()) { PreheaderDest = Preheader; Preheader = Preheader->getSinglePredecessor(); if (!Preheader) return UnknownValue; // Multiple preds. LoopEntryPredicate = dyn_cast(Preheader->getTerminator()); if (!LoopEntryPredicate) return UnknownValue; } // Now that we found a conditional branch that dominates the loop, check to // see if it is the comparison we are looking for. if (ICmpInst *ICI = dyn_cast(LoopEntryPredicate->getCondition())){ Value *PreCondLHS = ICI->getOperand(0); Value *PreCondRHS = ICI->getOperand(1); ICmpInst::Predicate Cond; if (LoopEntryPredicate->getSuccessor(0) == PreheaderDest) Cond = ICI->getPredicate(); else Cond = ICI->getInversePredicate(); switch (Cond) { case ICmpInst::ICMP_UGT: std::swap(PreCondLHS, PreCondRHS); Cond = ICmpInst::ICMP_ULT; break; case ICmpInst::ICMP_SGT: std::swap(PreCondLHS, PreCondRHS); Cond = ICmpInst::ICMP_SLT; break; default: break; } if (Cond == ICmpInst::ICMP_SLT) { if (PreCondLHS->getType()->isInteger()) { if (RHS != getSCEV(PreCondRHS)) return UnknownValue; // Not a comparison against 'm'. if (SCEV::getMinusSCEV(AddRec->getOperand(0), One) != getSCEV(PreCondLHS)) return UnknownValue; // Not a comparison against 'n-1'. } else return UnknownValue; } else if (Cond == ICmpInst::ICMP_ULT) return UnknownValue; // cerr << "Computed Loop Trip Count as: " // << // *SCEV::getMinusSCEV(RHS, AddRec->getOperand(0)) << "\n"; return SCEV::getMinusSCEV(RHS, AddRec->getOperand(0)); } else return UnknownValue; } return UnknownValue; } /// getNumIterationsInRange - Return the number of iterations of this loop that /// produce values in the specified constant range. Another way of looking at /// this is that it returns the first iteration number where the value is not in /// the condition, thus computing the exit count. If the iteration count can't /// be computed, an instance of SCEVCouldNotCompute is returned. SCEVHandle SCEVAddRecExpr::getNumIterationsInRange(ConstantRange Range, bool isSigned) const { if (Range.isFullSet()) // Infinite loop. return new SCEVCouldNotCompute(); // If the start is a non-zero constant, shift the range to simplify things. if (SCEVConstant *SC = dyn_cast(getStart())) if (!SC->getValue()->isZero()) { std::vector Operands(op_begin(), op_end()); Operands[0] = SCEVUnknown::getIntegerSCEV(0, SC->getType()); SCEVHandle Shifted = SCEVAddRecExpr::get(Operands, getLoop()); if (SCEVAddRecExpr *ShiftedAddRec = dyn_cast(Shifted)) return ShiftedAddRec->getNumIterationsInRange( Range.subtract(SC->getValue()->getValue()),isSigned); // This is strange and shouldn't happen. return new SCEVCouldNotCompute(); } // The only time we can solve this is when we have all constant indices. // Otherwise, we cannot determine the overflow conditions. for (unsigned i = 0, e = getNumOperands(); i != e; ++i) if (!isa(getOperand(i))) return new SCEVCouldNotCompute(); // Okay at this point we know that all elements of the chrec are constants and // that the start element is zero. // First check to see if the range contains zero. If not, the first // iteration exits. if (!Range.contains(APInt(getBitWidth(),0))) return SCEVConstant::get(ConstantInt::get(getType(),0)); if (isAffine()) { // If this is an affine expression then we have this situation: // Solve {0,+,A} in Range === Ax in Range // Since we know that zero is in the range, we know that the upper value of // the range must be the first possible exit value. Also note that we // already checked for a full range. const APInt &Upper = Range.getUpper(); APInt A = cast(getOperand(1))->getValue()->getValue(); APInt One(getBitWidth(),1); // The exit value should be (Upper+A-1)/A. APInt ExitVal(Upper); if (A != One) ExitVal = (Upper + A - One).sdiv(A); ConstantInt *ExitValue = ConstantInt::get(ExitVal); // Evaluate at the exit value. If we really did fall out of the valid // range, then we computed our trip count, otherwise wrap around or other // things must have happened. ConstantInt *Val = EvaluateConstantChrecAtConstant(this, ExitValue); if (Range.contains(Val->getValue())) return new SCEVCouldNotCompute(); // Something strange happened // Ensure that the previous value is in the range. This is a sanity check. assert(Range.contains( EvaluateConstantChrecAtConstant(this, ConstantInt::get(ExitVal - One))->getValue()) && "Linear scev computation is off in a bad way!"); return SCEVConstant::get(cast(ExitValue)); } else if (isQuadratic()) { // If this is a quadratic (3-term) AddRec {L,+,M,+,N}, find the roots of the // quadratic equation to solve it. To do this, we must frame our problem in // terms of figuring out when zero is crossed, instead of when // Range.getUpper() is crossed. std::vector NewOps(op_begin(), op_end()); NewOps[0] = SCEV::getNegativeSCEV(SCEVUnknown::get( ConstantInt::get(Range.getUpper()))); SCEVHandle NewAddRec = SCEVAddRecExpr::get(NewOps, getLoop()); // Next, solve the constructed addrec std::pair Roots = SolveQuadraticEquation(cast(NewAddRec)); SCEVConstant *R1 = dyn_cast(Roots.first); SCEVConstant *R2 = dyn_cast(Roots.second); if (R1) { // Pick the smallest positive root value. if (ConstantInt *CB = dyn_cast(ConstantExpr::getICmp(ICmpInst::ICMP_ULT, R1->getValue(), R2->getValue()))) { if (CB->getZExtValue() == false) std::swap(R1, R2); // R1 is the minimum root now. // Make sure the root is not off by one. The returned iteration should // not be in the range, but the previous one should be. When solving // for "X*X < 5", for example, we should not return a root of 2. ConstantInt *R1Val = EvaluateConstantChrecAtConstant(this, R1->getValue()); if (Range.contains(R1Val->getValue())) { // The next iteration must be out of the range... Constant *NextVal = ConstantInt::get(R1->getValue()->getValue()+1); R1Val = EvaluateConstantChrecAtConstant(this, NextVal); if (!Range.contains(R1Val->getValue())) return SCEVUnknown::get(NextVal); return new SCEVCouldNotCompute(); // Something strange happened } // If R1 was not in the range, then it is a good return value. Make // sure that R1-1 WAS in the range though, just in case. Constant *NextVal = ConstantInt::get(R1->getValue()->getValue()-1); R1Val = EvaluateConstantChrecAtConstant(this, NextVal); if (Range.contains(R1Val->getValue())) return R1; return new SCEVCouldNotCompute(); // Something strange happened } } } // Fallback, if this is a general polynomial, figure out the progression // through brute force: evaluate until we find an iteration that fails the // test. This is likely to be slow, but getting an accurate trip count is // incredibly important, we will be able to simplify the exit test a lot, and // we are almost guaranteed to get a trip count in this case. ConstantInt *TestVal = ConstantInt::get(getType(), 0); ConstantInt *EndVal = TestVal; // Stop when we wrap around. do { ++NumBruteForceEvaluations; SCEVHandle Val = evaluateAtIteration(SCEVConstant::get(TestVal)); if (!isa(Val)) // This shouldn't happen. return new SCEVCouldNotCompute(); // Check to see if we found the value! if (!Range.contains(cast(Val)->getValue()->getValue())) return SCEVConstant::get(TestVal); // Increment to test the next index. TestVal = ConstantInt::get(TestVal->getValue()+1); } while (TestVal != EndVal); return new SCEVCouldNotCompute(); } //===----------------------------------------------------------------------===// // ScalarEvolution Class Implementation //===----------------------------------------------------------------------===// bool ScalarEvolution::runOnFunction(Function &F) { Impl = new ScalarEvolutionsImpl(F, getAnalysis()); return false; } void ScalarEvolution::releaseMemory() { delete (ScalarEvolutionsImpl*)Impl; Impl = 0; } void ScalarEvolution::getAnalysisUsage(AnalysisUsage &AU) const { AU.setPreservesAll(); AU.addRequiredTransitive(); } SCEVHandle ScalarEvolution::getSCEV(Value *V) const { return ((ScalarEvolutionsImpl*)Impl)->getSCEV(V); } /// hasSCEV - Return true if the SCEV for this value has already been /// computed. bool ScalarEvolution::hasSCEV(Value *V) const { return ((ScalarEvolutionsImpl*)Impl)->hasSCEV(V); } /// setSCEV - Insert the specified SCEV into the map of current SCEVs for /// the specified value. void ScalarEvolution::setSCEV(Value *V, const SCEVHandle &H) { ((ScalarEvolutionsImpl*)Impl)->setSCEV(V, H); } SCEVHandle ScalarEvolution::getIterationCount(const Loop *L) const { return ((ScalarEvolutionsImpl*)Impl)->getIterationCount(L); } bool ScalarEvolution::hasLoopInvariantIterationCount(const Loop *L) const { return !isa(getIterationCount(L)); } SCEVHandle ScalarEvolution::getSCEVAtScope(Value *V, const Loop *L) const { return ((ScalarEvolutionsImpl*)Impl)->getSCEVAtScope(getSCEV(V), L); } void ScalarEvolution::deleteInstructionFromRecords(Instruction *I) const { return ((ScalarEvolutionsImpl*)Impl)->deleteInstructionFromRecords(I); } static void PrintLoopInfo(std::ostream &OS, const ScalarEvolution *SE, const Loop *L) { // Print all inner loops first for (Loop::iterator I = L->begin(), E = L->end(); I != E; ++I) PrintLoopInfo(OS, SE, *I); cerr << "Loop " << L->getHeader()->getName() << ": "; std::vector ExitBlocks; L->getExitBlocks(ExitBlocks); if (ExitBlocks.size() != 1) cerr << " "; if (SE->hasLoopInvariantIterationCount(L)) { cerr << *SE->getIterationCount(L) << " iterations! "; } else { cerr << "Unpredictable iteration count. "; } cerr << "\n"; } void ScalarEvolution::print(std::ostream &OS, const Module* ) const { Function &F = ((ScalarEvolutionsImpl*)Impl)->F; LoopInfo &LI = ((ScalarEvolutionsImpl*)Impl)->LI; OS << "Classifying expressions for: " << F.getName() << "\n"; for (inst_iterator I = inst_begin(F), E = inst_end(F); I != E; ++I) if (I->getType()->isInteger()) { OS << *I; OS << " --> "; SCEVHandle SV = getSCEV(&*I); SV->print(OS); OS << "\t\t"; if ((*I).getType()->isInteger()) { ConstantRange Bounds = SV->getValueRange(); if (!Bounds.isFullSet()) OS << "Bounds: " << Bounds << " "; } if (const Loop *L = LI.getLoopFor((*I).getParent())) { OS << "Exits: "; SCEVHandle ExitValue = getSCEVAtScope(&*I, L->getParentLoop()); if (isa(ExitValue)) { OS << "<>"; } else { OS << *ExitValue; } } OS << "\n"; } OS << "Determining loop execution counts for: " << F.getName() << "\n"; for (LoopInfo::iterator I = LI.begin(), E = LI.end(); I != E; ++I) PrintLoopInfo(OS, this, *I); }