From 41bd9ead35f60823c59367efe4f3d5ade87e756d Mon Sep 17 00:00:00 2001 From: Jonas Paulsson Date: Thu, 20 Feb 2020 10:40:30 -0800 Subject: [PATCH] [SystemZ] Return scalarized costs for vector instructions on older archs. A cost query for a vector instruction should return a cost even without target vector support, and not trigger an assert. VectorCombine does this with an input containing source code vectors. Review: Ulrich Weigand --- .../SystemZ/SystemZTargetTransformInfo.cpp | 261 +++++++++--------- .../CostModel/SystemZ/oldarch-vectors.ll | 13 + 2 files changed, 142 insertions(+), 132 deletions(-) create mode 100644 llvm/test/Analysis/CostModel/SystemZ/oldarch-vectors.ll diff --git a/llvm/lib/Target/SystemZ/SystemZTargetTransformInfo.cpp b/llvm/lib/Target/SystemZ/SystemZTargetTransformInfo.cpp index acec3c533585..df5286bef817 100644 --- a/llvm/lib/Target/SystemZ/SystemZTargetTransformInfo.cpp +++ b/llvm/lib/Target/SystemZ/SystemZTargetTransformInfo.cpp @@ -391,9 +391,57 @@ int SystemZTTIImpl::getArithmeticInstrCost( } } - if (Ty->isVectorTy()) { - assert(ST->hasVector() && - "getArithmeticInstrCost() called with vector type."); + if (!Ty->isVectorTy()) { + // These FP operations are supported with a dedicated instruction for + // float, double and fp128 (base implementation assumes float generally + // costs 2). + if (Opcode == Instruction::FAdd || Opcode == Instruction::FSub || + Opcode == Instruction::FMul || Opcode == Instruction::FDiv) + return 1; + + // There is no native support for FRem. + if (Opcode == Instruction::FRem) + return LIBCALL_COST; + + // Give discount for some combined logical operations if supported. + if (Args.size() == 2 && ST->hasMiscellaneousExtensions3()) { + if (Opcode == Instruction::Xor) { + for (const Value *A : Args) { + if (const Instruction *I = dyn_cast(A)) + if (I->hasOneUse() && + (I->getOpcode() == Instruction::And || + I->getOpcode() == Instruction::Or || + I->getOpcode() == Instruction::Xor)) + return 0; + } + } + else if (Opcode == Instruction::Or || Opcode == Instruction::And) { + for (const Value *A : Args) { + if (const Instruction *I = dyn_cast(A)) + if (I->hasOneUse() && I->getOpcode() == Instruction::Xor) + return 0; + } + } + } + + // Or requires one instruction, although it has custom handling for i64. + if (Opcode == Instruction::Or) + return 1; + + if (Opcode == Instruction::Xor && ScalarBits == 1) { + if (ST->hasLoadStoreOnCond2()) + return 5; // 2 * (li 0; loc 1); xor + return 7; // 2 * ipm sequences ; xor ; shift ; compare + } + + if (DivRemConstPow2) + return (SignedDivRem ? SDivPow2Cost : 1); + if (DivRemConst) + return DivMulSeqCost; + if (SignedDivRem || UnsignedDivRem) + return DivInstrCost; + } + else if (ST->hasVector()) { unsigned VF = Ty->getVectorNumElements(); unsigned NumVectors = getNumVectorRegs(Ty); @@ -454,56 +502,6 @@ int SystemZTTIImpl::getArithmeticInstrCost( return Cost; } } - else { // Scalar: - // These FP operations are supported with a dedicated instruction for - // float, double and fp128 (base implementation assumes float generally - // costs 2). - if (Opcode == Instruction::FAdd || Opcode == Instruction::FSub || - Opcode == Instruction::FMul || Opcode == Instruction::FDiv) - return 1; - - // There is no native support for FRem. - if (Opcode == Instruction::FRem) - return LIBCALL_COST; - - // Give discount for some combined logical operations if supported. - if (Args.size() == 2 && ST->hasMiscellaneousExtensions3()) { - if (Opcode == Instruction::Xor) { - for (const Value *A : Args) { - if (const Instruction *I = dyn_cast(A)) - if (I->hasOneUse() && - (I->getOpcode() == Instruction::And || - I->getOpcode() == Instruction::Or || - I->getOpcode() == Instruction::Xor)) - return 0; - } - } - else if (Opcode == Instruction::Or || Opcode == Instruction::And) { - for (const Value *A : Args) { - if (const Instruction *I = dyn_cast(A)) - if (I->hasOneUse() && I->getOpcode() == Instruction::Xor) - return 0; - } - } - } - - // Or requires one instruction, although it has custom handling for i64. - if (Opcode == Instruction::Or) - return 1; - - if (Opcode == Instruction::Xor && ScalarBits == 1) { - if (ST->hasLoadStoreOnCond2()) - return 5; // 2 * (li 0; loc 1); xor - return 7; // 2 * ipm sequences ; xor ; shift ; compare - } - - if (DivRemConstPow2) - return (SignedDivRem ? SDivPow2Cost : 1); - if (DivRemConst) - return DivMulSeqCost; - if (SignedDivRem || UnsignedDivRem) - return DivInstrCost; - } // Fallback to the default implementation. return BaseT::getArithmeticInstrCost(Opcode, Ty, Op1Info, Op2Info, @@ -513,35 +511,36 @@ int SystemZTTIImpl::getArithmeticInstrCost( int SystemZTTIImpl::getShuffleCost(TTI::ShuffleKind Kind, Type *Tp, int Index, Type *SubTp) { assert (Tp->isVectorTy()); - assert (ST->hasVector() && "getShuffleCost() called."); - unsigned NumVectors = getNumVectorRegs(Tp); + if (ST->hasVector()) { + unsigned NumVectors = getNumVectorRegs(Tp); - // TODO: Since fp32 is expanded, the shuffle cost should always be 0. + // TODO: Since fp32 is expanded, the shuffle cost should always be 0. - // FP128 values are always in scalar registers, so there is no work - // involved with a shuffle, except for broadcast. In that case register - // moves are done with a single instruction per element. - if (Tp->getScalarType()->isFP128Ty()) - return (Kind == TargetTransformInfo::SK_Broadcast ? NumVectors - 1 : 0); + // FP128 values are always in scalar registers, so there is no work + // involved with a shuffle, except for broadcast. In that case register + // moves are done with a single instruction per element. + if (Tp->getScalarType()->isFP128Ty()) + return (Kind == TargetTransformInfo::SK_Broadcast ? NumVectors - 1 : 0); - switch (Kind) { - case TargetTransformInfo::SK_ExtractSubvector: - // ExtractSubvector Index indicates start offset. + switch (Kind) { + case TargetTransformInfo::SK_ExtractSubvector: + // ExtractSubvector Index indicates start offset. - // Extracting a subvector from first index is a noop. - return (Index == 0 ? 0 : NumVectors); + // Extracting a subvector from first index is a noop. + return (Index == 0 ? 0 : NumVectors); - case TargetTransformInfo::SK_Broadcast: - // Loop vectorizer calls here to figure out the extra cost of - // broadcasting a loaded value to all elements of a vector. Since vlrep - // loads and replicates with a single instruction, adjust the returned - // value. - return NumVectors - 1; + case TargetTransformInfo::SK_Broadcast: + // Loop vectorizer calls here to figure out the extra cost of + // broadcasting a loaded value to all elements of a vector. Since vlrep + // loads and replicates with a single instruction, adjust the returned + // value. + return NumVectors - 1; - default: + default: - // SystemZ supports single instruction permutation / replication. - return NumVectors; + // SystemZ supports single instruction permutation / replication. + return NumVectors; + } } return BaseT::getShuffleCost(Kind, Tp, Index, SubTp); @@ -672,8 +671,36 @@ int SystemZTTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst, Type *Src, unsigned DstScalarBits = Dst->getScalarSizeInBits(); unsigned SrcScalarBits = Src->getScalarSizeInBits(); - if (Src->isVectorTy()) { - assert (ST->hasVector() && "getCastInstrCost() called with vector type."); + if (!Src->isVectorTy()) { + assert (!Dst->isVectorTy()); + + if (Opcode == Instruction::SIToFP || Opcode == Instruction::UIToFP) { + if (SrcScalarBits >= 32 || + (I != nullptr && isa(I->getOperand(0)))) + return 1; + return SrcScalarBits > 1 ? 2 /*i8/i16 extend*/ : 5 /*branch seq.*/; + } + + if ((Opcode == Instruction::ZExt || Opcode == Instruction::SExt) && + Src->isIntegerTy(1)) { + if (ST->hasLoadStoreOnCond2()) + return 2; // li 0; loc 1 + + // This should be extension of a compare i1 result, which is done with + // ipm and a varying sequence of instructions. + unsigned Cost = 0; + if (Opcode == Instruction::SExt) + Cost = (DstScalarBits < 64 ? 3 : 4); + if (Opcode == Instruction::ZExt) + Cost = 3; + Type *CmpOpTy = ((I != nullptr) ? getCmpOpsType(I) : nullptr); + if (CmpOpTy != nullptr && CmpOpTy->isFloatingPointTy()) + // If operands of an fp-type was compared, this costs +1. + Cost++; + return Cost; + } + } + else if (ST->hasVector()) { assert (Dst->isVectorTy()); unsigned VF = Src->getVectorNumElements(); unsigned NumDstVectors = getNumVectorRegs(Dst); @@ -759,35 +786,6 @@ int SystemZTTIImpl::getCastInstrCost(unsigned Opcode, Type *Dst, Type *Src, return VF + getScalarizationOverhead(Src, false, true); } } - else { // Scalar - assert (!Dst->isVectorTy()); - - if (Opcode == Instruction::SIToFP || Opcode == Instruction::UIToFP) { - if (SrcScalarBits >= 32 || - (I != nullptr && isa(I->getOperand(0)))) - return 1; - return SrcScalarBits > 1 ? 2 /*i8/i16 extend*/ : 5 /*branch seq.*/; - } - - if ((Opcode == Instruction::ZExt || Opcode == Instruction::SExt) && - Src->isIntegerTy(1)) { - if (ST->hasLoadStoreOnCond2()) - return 2; // li 0; loc 1 - - // This should be extension of a compare i1 result, which is done with - // ipm and a varying sequence of instructions. - unsigned Cost = 0; - if (Opcode == Instruction::SExt) - Cost = (DstScalarBits < 64 ? 3 : 4); - if (Opcode == Instruction::ZExt) - Cost = 3; - Type *CmpOpTy = ((I != nullptr) ? getCmpOpsType(I) : nullptr); - if (CmpOpTy != nullptr && CmpOpTy->isFloatingPointTy()) - // If operands of an fp-type was compared, this costs +1. - Cost++; - return Cost; - } - } return BaseT::getCastInstrCost(Opcode, Dst, Src, I); } @@ -806,8 +804,31 @@ static unsigned getOperandsExtensionCost(const Instruction *I) { int SystemZTTIImpl::getCmpSelInstrCost(unsigned Opcode, Type *ValTy, Type *CondTy, const Instruction *I) { - if (ValTy->isVectorTy()) { - assert (ST->hasVector() && "getCmpSelInstrCost() called with vector type."); + if (!ValTy->isVectorTy()) { + switch (Opcode) { + case Instruction::ICmp: { + // A loaded value compared with 0 with multiple users becomes Load and + // Test. The load is then not foldable, so return 0 cost for the ICmp. + unsigned ScalarBits = ValTy->getScalarSizeInBits(); + if (I != nullptr && ScalarBits >= 32) + if (LoadInst *Ld = dyn_cast(I->getOperand(0))) + if (const ConstantInt *C = dyn_cast(I->getOperand(1))) + if (!Ld->hasOneUse() && Ld->getParent() == I->getParent() && + C->getZExtValue() == 0) + return 0; + + unsigned Cost = 1; + if (ValTy->isIntegerTy() && ValTy->getScalarSizeInBits() <= 16) + Cost += (I != nullptr ? getOperandsExtensionCost(I) : 2); + return Cost; + } + case Instruction::Select: + if (ValTy->isFloatingPointTy()) + return 4; // No load on condition for FP - costs a conditional jump. + return 1; // Load On Condition / Select Register. + } + } + else if (ST->hasVector()) { unsigned VF = ValTy->getVectorNumElements(); // Called with a compare instruction. @@ -856,30 +877,6 @@ int SystemZTTIImpl::getCmpSelInstrCost(unsigned Opcode, Type *ValTy, return getNumVectorRegs(ValTy) /*vsel*/ + PackCost; } } - else { // Scalar - switch (Opcode) { - case Instruction::ICmp: { - // A loaded value compared with 0 with multiple users becomes Load and - // Test. The load is then not foldable, so return 0 cost for the ICmp. - unsigned ScalarBits = ValTy->getScalarSizeInBits(); - if (I != nullptr && ScalarBits >= 32) - if (LoadInst *Ld = dyn_cast(I->getOperand(0))) - if (const ConstantInt *C = dyn_cast(I->getOperand(1))) - if (!Ld->hasOneUse() && Ld->getParent() == I->getParent() && - C->getZExtValue() == 0) - return 0; - - unsigned Cost = 1; - if (ValTy->isIntegerTy() && ValTy->getScalarSizeInBits() <= 16) - Cost += (I != nullptr ? getOperandsExtensionCost(I) : 2); - return Cost; - } - case Instruction::Select: - if (ValTy->isFloatingPointTy()) - return 4; // No load on condition for FP - costs a conditional jump. - return 1; // Load On Condition / Select Register. - } - } return BaseT::getCmpSelInstrCost(Opcode, ValTy, CondTy, nullptr); } diff --git a/llvm/test/Analysis/CostModel/SystemZ/oldarch-vectors.ll b/llvm/test/Analysis/CostModel/SystemZ/oldarch-vectors.ll new file mode 100644 index 000000000000..b56b4b9bdfc8 --- /dev/null +++ b/llvm/test/Analysis/CostModel/SystemZ/oldarch-vectors.ll @@ -0,0 +1,13 @@ +; RUN: opt < %s -cost-model -analyze -mtriple=systemz-unknown -mcpu=z10 +; +; Check that some costs can be returned for vector instructions also without +; vector support. + +define void @fun(<2 x double>* %arg) { +entry: + %add = fadd <2 x double> undef, undef + shufflevector <2 x i32> undef, <2 x i32> undef, <2 x i32> + %conv = fptoui <4 x float> undef to <4 x i32> + %cmp = icmp eq <2 x i64> undef, undef + ret void +}