diff --git a/lib/Analysis/ScalarEvolution.cpp b/lib/Analysis/ScalarEvolution.cpp index 06dbde58c10..ff8fbfdd76e 100644 --- a/lib/Analysis/ScalarEvolution.cpp +++ b/lib/Analysis/ScalarEvolution.cpp @@ -1,4 +1,4 @@ -//===- ScalarEvolution.cpp - Scalar Evolution Analysis ----------*- C++ -*-===// +//===- ScalarEvolution.cpp - Scalar Evolution Analysis --------------------===// // // The LLVM Compiler Infrastructure // @@ -2061,71 +2061,66 @@ const SCEV *ScalarEvolution::getMulExpr(SmallVectorImpl &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. + + // {A1,+,A2,+,...,+,An} * {B1,+,B2,+,...,+,Bn} + // = {x=1 in [ sum y=x..2x [ sum z=max(y-x, y-n)..min(x,n) [ + // choose(x, 2x)*choose(2x-y, x-z)*A_{y-z}*B_z + // ]]],+,...up to x=2n}. + // Note that the arguments to choose() are always integers with values + // known at compile time, never SCEV objects. + // + // The implementation avoids pointless extra computations when the two + // addrec's are of different length (mathematically, it's equivalent to + // an infinite stream of zeros on the right). + bool OpsModified = false; for (unsigned OtherIdx = Idx+1; - OtherIdx < Ops.size() && isa(Ops[OtherIdx]); + OtherIdx != Ops.size() && isa(Ops[OtherIdx]); ++OtherIdx) { - if (AddRecLoop != cast(Ops[OtherIdx])->getLoop()) + const SCEVAddRecExpr *OtherAddRec = + dyn_cast(Ops[OtherIdx]); + if (!OtherAddRec || OtherAddRec->getLoop() != AddRecLoop) continue; - // {A1,+,A2,+,...,+,An} * {B1,+,B2,+,...,+,Bn} - // = {x=1 in [ sum y=x..2x [ sum z=max(y-x, y-n)..min(x,n) [ - // choose(x, 2x)*choose(2x-y, x-z)*A_{y-z}*B_z - // ]]],+,...up to x=2n}. - // Note that the arguments to choose() are always integers with values - // known at compile time, never SCEV objects. - // - // The implementation avoids pointless extra computations when the two - // addrec's are of different length (mathematically, it's equivalent to - // an infinite stream of zeros on the right). - bool OpsModified = false; - for (; OtherIdx != Ops.size() && isa(Ops[OtherIdx]); - ++OtherIdx) { - const SCEVAddRecExpr *OtherAddRec = - dyn_cast(Ops[OtherIdx]); - if (!OtherAddRec || OtherAddRec->getLoop() != AddRecLoop) - continue; - - bool Overflow = false; - Type *Ty = AddRec->getType(); - bool LargerThan64Bits = getTypeSizeInBits(Ty) > 64; - SmallVector AddRecOps; - for (int x = 0, xe = AddRec->getNumOperands() + - OtherAddRec->getNumOperands() - 1; x != xe && !Overflow; ++x) { - const SCEV *Term = getConstant(Ty, 0); - for (int y = x, ye = 2*x+1; y != ye && !Overflow; ++y) { - uint64_t Coeff1 = Choose(x, 2*x - y, Overflow); - for (int z = std::max(y-x, y-(int)AddRec->getNumOperands()+1), - ze = std::min(x+1, (int)OtherAddRec->getNumOperands()); - z < ze && !Overflow; ++z) { - uint64_t Coeff2 = Choose(2*x - y, x-z, Overflow); - uint64_t Coeff; - if (LargerThan64Bits) - Coeff = umul_ov(Coeff1, Coeff2, Overflow); - else - Coeff = Coeff1*Coeff2; - const SCEV *CoeffTerm = getConstant(Ty, Coeff); - const SCEV *Term1 = AddRec->getOperand(y-z); - const SCEV *Term2 = OtherAddRec->getOperand(z); - Term = getAddExpr(Term, getMulExpr(CoeffTerm, Term1,Term2)); - } + bool Overflow = false; + Type *Ty = AddRec->getType(); + bool LargerThan64Bits = getTypeSizeInBits(Ty) > 64; + SmallVector AddRecOps; + for (int x = 0, xe = AddRec->getNumOperands() + + OtherAddRec->getNumOperands() - 1; x != xe && !Overflow; ++x) { + const SCEV *Term = getConstant(Ty, 0); + for (int y = x, ye = 2*x+1; y != ye && !Overflow; ++y) { + uint64_t Coeff1 = Choose(x, 2*x - y, Overflow); + for (int z = std::max(y-x, y-(int)AddRec->getNumOperands()+1), + ze = std::min(x+1, (int)OtherAddRec->getNumOperands()); + z < ze && !Overflow; ++z) { + uint64_t Coeff2 = Choose(2*x - y, x-z, Overflow); + uint64_t Coeff; + if (LargerThan64Bits) + Coeff = umul_ov(Coeff1, Coeff2, Overflow); + else + Coeff = Coeff1*Coeff2; + const SCEV *CoeffTerm = getConstant(Ty, Coeff); + const SCEV *Term1 = AddRec->getOperand(y-z); + const SCEV *Term2 = OtherAddRec->getOperand(z); + Term = getAddExpr(Term, getMulExpr(CoeffTerm, Term1,Term2)); } - AddRecOps.push_back(Term); - } - if (!Overflow) { - const SCEV *NewAddRec = getAddRecExpr(AddRecOps, AddRec->getLoop(), - SCEV::FlagAnyWrap); - if (Ops.size() == 2) return NewAddRec; - Ops[Idx] = NewAddRec; - Ops.erase(Ops.begin() + OtherIdx); --OtherIdx; - OpsModified = true; - AddRec = dyn_cast(NewAddRec); - if (!AddRec) - break; } + AddRecOps.push_back(Term); + } + if (!Overflow) { + const SCEV *NewAddRec = getAddRecExpr(AddRecOps, AddRec->getLoop(), + SCEV::FlagAnyWrap); + if (Ops.size() == 2) return NewAddRec; + Ops[Idx] = NewAddRec; + Ops.erase(Ops.begin() + OtherIdx); --OtherIdx; + OpsModified = true; + AddRec = dyn_cast(NewAddRec); + if (!AddRec) + break; } - if (OpsModified) - return getMulExpr(Ops); } + if (OpsModified) + return getMulExpr(Ops); // Otherwise couldn't fold anything into this recurrence. Move onto the // next one.