[PATCH] D153547: [CostModel] Use min/max intrinsics for vecreduce.min/max costs

[Nikita Popov via Phabricator via llvm-commits]

nikic added inline comments.


================
Comment at: llvm/include/llvm/Analysis/TargetTransformInfo.h:1397
   InstructionCost getMinMaxReductionCost(
-      VectorType *Ty, VectorType *CondTy, bool IsUnsigned,
-      FastMathFlags FMF = FastMathFlags(),
+      unsigned Opcode, VectorType *Ty, FastMathFlags FMF = FastMathFlags(),
       TTI::TargetCostKind CostKind = TTI::TCK_RecipThroughput) const;
----------------
Opcode -> IntrinsicID? We don't usually use the term opcode for intrinsics. Could also use `Intrinsic::ID` type here (unless it's not available).


================
Comment at: llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp:13826
       }
       Intrinsic::ID Id = getMinMaxReductionIntrinsicOp(RdxKind);
       ScalarCost = EvaluateScalarCost([&]() {
----------------
Use this value instead of your switch?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D153547/new/

https://reviews.llvm.org/D153547