| 1 | //===------ NVPTXIRPeephole.cpp - NVPTX IR Peephole --------------===// |
| 2 | // |
| 3 | // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. |
| 4 | // See https://llvm.org/LICENSE.txt for license information. |
| 5 | // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception |
| 6 | // |
| 7 | //===----------------------------------------------------------------------===// |
| 8 | // |
| 9 | // This file implements IR-level peephole optimizations. These transformations |
| 10 | // run late in the NVPTX IR pass pipeline just before the instruction selection. |
| 11 | // |
| 12 | // Currently, it implements the following transformation(s): |
| 13 | // 1. FMA folding (float/double types): |
| 14 | // Transforms FMUL+FADD/FSUB sequences into FMA intrinsics when the |
| 15 | // 'contract' fast-math flag is present. Supported patterns: |
| 16 | // - fadd(fmul(a, b), c) => fma(a, b, c) |
| 17 | // - fadd(c, fmul(a, b)) => fma(a, b, c) |
| 18 | // - fadd(fmul(a, b), fmul(c, d)) => fma(a, b, fmul(c, d)) |
| 19 | // - fsub(fmul(a, b), c) => fma(a, b, fneg(c)) |
| 20 | // - fsub(a, fmul(b, c)) => fma(fneg(b), c, a) |
| 21 | // - fsub(fmul(a, b), fmul(c, d)) => fma(a, b, fneg(fmul(c, d))) |
| 22 | // |
| 23 | //===----------------------------------------------------------------------===// |
| 24 | |
| 25 | #include "NVPTXUtilities.h" |
| 26 | #include "llvm/IR/IRBuilder.h" |
| 27 | #include "llvm/IR/InstIterator.h" |
| 28 | #include "llvm/IR/Instructions.h" |
| 29 | #include "llvm/IR/Intrinsics.h" |
| 30 | |
| 31 | #define DEBUG_TYPE "nvptx-ir-peephole" |
| 32 | |
| 33 | using namespace llvm; |
| 34 | |
| 35 | static bool tryFoldBinaryFMul(BinaryOperator *BI) { |
| 36 | Value *Op0 = BI->getOperand(i_nocapture: 0); |
| 37 | Value *Op1 = BI->getOperand(i_nocapture: 1); |
| 38 | |
| 39 | auto *FMul0 = dyn_cast<BinaryOperator>(Val: Op0); |
| 40 | auto *FMul1 = dyn_cast<BinaryOperator>(Val: Op1); |
| 41 | |
| 42 | BinaryOperator *FMul = nullptr; |
| 43 | Value *OtherOperand = nullptr; |
| 44 | bool IsFirstOperand = false; |
| 45 | |
| 46 | // Either Op0 or Op1 should be a valid FMul |
| 47 | if (FMul0 && FMul0->getOpcode() == Instruction::FMul && FMul0->hasOneUse() && |
| 48 | FMul0->hasAllowContract()) { |
| 49 | FMul = FMul0; |
| 50 | OtherOperand = Op1; |
| 51 | IsFirstOperand = true; |
| 52 | } else if (FMul1 && FMul1->getOpcode() == Instruction::FMul && |
| 53 | FMul1->hasOneUse() && FMul1->hasAllowContract()) { |
| 54 | FMul = FMul1; |
| 55 | OtherOperand = Op0; |
| 56 | IsFirstOperand = false; |
| 57 | } else { |
| 58 | return false; |
| 59 | } |
| 60 | |
| 61 | bool IsFSub = BI->getOpcode() == Instruction::FSub; |
| 62 | LLVM_DEBUG({ |
| 63 | const char *OpName = IsFSub ? "FSub" : "FAdd" ; |
| 64 | dbgs() << "Found " << OpName << " with FMul (single use) as " |
| 65 | << (IsFirstOperand ? "first" : "second" ) << " operand: " << *BI |
| 66 | << "\n" ; |
| 67 | }); |
| 68 | |
| 69 | Value *MulOp0 = FMul->getOperand(i_nocapture: 0); |
| 70 | Value *MulOp1 = FMul->getOperand(i_nocapture: 1); |
| 71 | IRBuilder<> Builder(BI); |
| 72 | Value *FMA = nullptr; |
| 73 | |
| 74 | if (!IsFSub) { |
| 75 | // fadd(fmul(a, b), c) => fma(a, b, c) |
| 76 | // fadd(c, fmul(a, b)) => fma(a, b, c) |
| 77 | FMA = Builder.CreateIntrinsic(ID: Intrinsic::fma, Types: {BI->getType()}, |
| 78 | Args: {MulOp0, MulOp1, OtherOperand}); |
| 79 | } else { |
| 80 | if (IsFirstOperand) { |
| 81 | // fsub(fmul(a, b), c) => fma(a, b, fneg(c)) |
| 82 | Value *NegOtherOp = |
| 83 | Builder.CreateFNegFMF(V: OtherOperand, FMFSource: BI->getFastMathFlags()); |
| 84 | FMA = Builder.CreateIntrinsic(ID: Intrinsic::fma, Types: {BI->getType()}, |
| 85 | Args: {MulOp0, MulOp1, NegOtherOp}); |
| 86 | } else { |
| 87 | // fsub(a, fmul(b, c)) => fma(fneg(b), c, a) |
| 88 | Value *NegMulOp0 = |
| 89 | Builder.CreateFNegFMF(V: MulOp0, FMFSource: FMul->getFastMathFlags()); |
| 90 | FMA = Builder.CreateIntrinsic(ID: Intrinsic::fma, Types: {BI->getType()}, |
| 91 | Args: {NegMulOp0, MulOp1, OtherOperand}); |
| 92 | } |
| 93 | } |
| 94 | |
| 95 | // Combine fast-math flags from the original instructions |
| 96 | auto *FMAInst = cast<Instruction>(Val: FMA); |
| 97 | FastMathFlags BinaryFMF = BI->getFastMathFlags(); |
| 98 | FastMathFlags FMulFMF = FMul->getFastMathFlags(); |
| 99 | FastMathFlags NewFMF = FastMathFlags::intersectRewrite(LHS: BinaryFMF, RHS: FMulFMF) | |
| 100 | FastMathFlags::unionValue(LHS: BinaryFMF, RHS: FMulFMF); |
| 101 | FMAInst->setFastMathFlags(NewFMF); |
| 102 | |
| 103 | LLVM_DEBUG({ |
| 104 | const char *OpName = IsFSub ? "FSub" : "FAdd" ; |
| 105 | dbgs() << "Replacing " << OpName << " with FMA: " << *FMA << "\n" ; |
| 106 | }); |
| 107 | BI->replaceAllUsesWith(V: FMA); |
| 108 | BI->eraseFromParent(); |
| 109 | FMul->eraseFromParent(); |
| 110 | return true; |
| 111 | } |
| 112 | |
| 113 | static bool foldFMA(Function &F) { |
| 114 | bool Changed = false; |
| 115 | |
| 116 | // Iterate and process float/double FAdd/FSub instructions with allow-contract |
| 117 | for (auto &I : llvm::make_early_inc_range(Range: instructions(F))) { |
| 118 | if (auto *BI = dyn_cast<BinaryOperator>(Val: &I)) { |
| 119 | // Only FAdd and FSub are supported. |
| 120 | if (BI->getOpcode() != Instruction::FAdd && |
| 121 | BI->getOpcode() != Instruction::FSub) |
| 122 | continue; |
| 123 | |
| 124 | // At minimum, the instruction should have allow-contract. |
| 125 | if (!BI->hasAllowContract()) |
| 126 | continue; |
| 127 | |
| 128 | // Only float and double are supported. |
| 129 | if (!BI->getType()->isFloatTy() && !BI->getType()->isDoubleTy()) |
| 130 | continue; |
| 131 | |
| 132 | if (tryFoldBinaryFMul(BI)) |
| 133 | Changed = true; |
| 134 | } |
| 135 | } |
| 136 | return Changed; |
| 137 | } |
| 138 | |
| 139 | namespace { |
| 140 | |
| 141 | struct NVPTXIRPeephole : public FunctionPass { |
| 142 | static char ID; |
| 143 | NVPTXIRPeephole() : FunctionPass(ID) {} |
| 144 | bool runOnFunction(Function &F) override; |
| 145 | }; |
| 146 | |
| 147 | } // namespace |
| 148 | |
| 149 | char NVPTXIRPeephole::ID = 0; |
| 150 | INITIALIZE_PASS(NVPTXIRPeephole, "nvptx-ir-peephole" , "NVPTX IR Peephole" , |
| 151 | false, false) |
| 152 | |
| 153 | bool NVPTXIRPeephole::runOnFunction(Function &F) { return foldFMA(F); } |
| 154 | |
| 155 | FunctionPass *llvm::createNVPTXIRPeepholePass() { |
| 156 | return new NVPTXIRPeephole(); |
| 157 | } |
| 158 | |
| 159 | PreservedAnalyses NVPTXIRPeepholePass::run(Function &F, |
| 160 | FunctionAnalysisManager &) { |
| 161 | if (!foldFMA(F)) |
| 162 | return PreservedAnalyses::all(); |
| 163 | |
| 164 | PreservedAnalyses PA; |
| 165 | PA.preserveSet<CFGAnalyses>(); |
| 166 | return PA; |
| 167 | } |
| 168 | |