From 9278dd7b2bd9bab4a7122b074e43e9bdd0d5d20f Mon Sep 17 00:00:00 2001 From: David Sherwood Date: Wed, 29 Mar 2023 12:22:09 +0000 Subject: [PATCH] [LoopVectorize] Fix zext/sext cost calculations when types are shrunk In getInstructionCost if we know a zext/sext is going to be shrunk we should only be changing the destination type, and leave the source type unchanged. For example, we may change a zext from zext <16 x i8> %a to <16 x i32> to zext <16 x i8> %a to <16 x i16> However, we were previously calculating the cost for doing zext <16 x i16> %a to <16 x i16> which is incorrect. Differential Revision: https://reviews.llvm.org/D147152 --- llvm/lib/Transforms/Vectorize/LoopVectorize.cpp | 3 ++- .../Transforms/LoopVectorize/AArch64/type-shrinkage-zext-costs.ll | 6 ++++-- 2 files changed, 6 insertions(+), 3 deletions(-) diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp index fd8596a..e686508 100644 --- a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp +++ b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp @@ -7371,7 +7371,8 @@ LoopVectorizationCostModel::getInstructionCost(Instruction *I, ElementCount VF, VectorTy = largestIntegerVectorType(ToVectorTy(I->getType(), VF), MinVecTy); } else if (Opcode == Instruction::ZExt || Opcode == Instruction::SExt) { - SrcVecTy = largestIntegerVectorType(SrcVecTy, MinVecTy); + // Leave SrcVecTy unchanged - we only shrink the destination element + // type. VectorTy = smallestIntegerVectorType(ToVectorTy(I->getType(), VF), MinVecTy); } diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/type-shrinkage-zext-costs.ll b/llvm/test/Transforms/LoopVectorize/AArch64/type-shrinkage-zext-costs.ll index 091c571..34de689 100644 --- a/llvm/test/Transforms/LoopVectorize/AArch64/type-shrinkage-zext-costs.ll +++ b/llvm/test/Transforms/LoopVectorize/AArch64/type-shrinkage-zext-costs.ll @@ -16,7 +16,8 @@ define void @zext_i8_i16(ptr noalias nocapture readonly %p, ptr noalias nocaptur ; CHECK-COST: LV: Found an estimated cost of 1 for VF vscale x 1 For instruction: %conv = zext i8 %0 to i32 ; CHECK-COST: LV: Found an estimated cost of 1 for VF vscale x 2 For instruction: %conv = zext i8 %0 to i32 ; CHECK-COST: LV: Found an estimated cost of 1 for VF vscale x 4 For instruction: %conv = zext i8 %0 to i32 -; CHECK-COST: LV: Found an estimated cost of 1 for VF vscale x 8 For instruction: %conv = zext i8 %0 to i32 +; CHECK-COST: LV: Found an estimated cost of 0 for VF vscale x 8 For instruction: %conv = zext i8 %0 to i32 + ; CHECK-LABEL: define void @zext_i8_i16 ; CHECK-SAME: (ptr noalias nocapture readonly [[P:%.*]], ptr noalias nocapture [[Q:%.*]], i32 [[LEN:%.*]]) #[[ATTR0:[0-9]+]] { ; CHECK-NEXT: entry: @@ -99,7 +100,8 @@ define void @sext_i8_i16(ptr noalias nocapture readonly %p, ptr noalias nocaptur ; CHECK-COST: LV: Found an estimated cost of 1 for VF vscale x 1 For instruction: %conv = sext i8 %0 to i32 ; CHECK-COST: LV: Found an estimated cost of 1 for VF vscale x 2 For instruction: %conv = sext i8 %0 to i32 ; CHECK-COST: LV: Found an estimated cost of 1 for VF vscale x 4 For instruction: %conv = sext i8 %0 to i32 -; CHECK-COST: LV: Found an estimated cost of 1 for VF vscale x 8 For instruction: %conv = sext i8 %0 to i32 +; CHECK-COST: LV: Found an estimated cost of 0 for VF vscale x 8 For instruction: %conv = sext i8 %0 to i32 + ; CHECK-LABEL: define void @sext_i8_i16 ; CHECK-SAME: (ptr noalias nocapture readonly [[P:%.*]], ptr noalias nocapture [[Q:%.*]], i32 [[LEN:%.*]]) #[[ATTR0]] { ; CHECK-NEXT: entry: -- 2.7.4