[llvm] [AArch64] Add @llvm.experimental.vector.match (PR #101974)
Paul Walker via llvm-commits
llvm-commits at lists.llvm.org
Wed Oct 30 11:16:57 PDT 2024
================
@@ -1771,6 +1771,11 @@ class TargetTransformInfo {
/// This should also apply to lowering for vector funnel shifts (rotates).
bool isVectorShiftByScalarCheap(Type *Ty) const;
+ /// \returns True if the target has hardware support for vector match
+ /// operations between vectors of type `VT` and search vectors of `SearchSize`
+ /// elements, and false otherwise.
+ bool hasVectorMatch(VectorType *VT, unsigned SearchSize) const;
----------------
paulwalker-arm wrote:
Is this really a TTI method?
The generic lowering means it's available to all targets and so the decision to use it would be cost based rather than a specific TTI question like this. I'm not sure of the future intent but based on the current usage I'd suggest moving it into TargetLowering.
https://github.com/llvm/llvm-project/pull/101974
More information about the llvm-commits
mailing list