[PATCH] D152258: [LV] Add mask support for vectorizing interleaved groups
Graham Hunter via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Thu Jun 15 02:02:16 PDT 2023
huntergr added a comment.
I think it would be worthwhile to have a test covering conditional execution inside the original source loop as well as the active lane mask for tail folding. See `llvm/test/Transforms/LoopVectorize/interleaved-accesses-masked-group.ll` for examples.
================
Comment at: llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp:2824-2825
+ // VF.
+ if (!VecTy->isScalableTy() && (UseMaskForCond || UseMaskForGaps))
+ return InstructionCost::getInvalid();
+
----------------
We could use SVE masking for fixed length as well, but that can be done at a later date if you want.
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D152258/new/
https://reviews.llvm.org/D152258
More information about the llvm-commits
mailing list