[PATCH] D124612: [AArch64][LV] AArch64 does not prefer vectorized addressing

[Tiehu Zhang via Phabricator via llvm-commits]

TiehuZhang created this revision.
TiehuZhang added reviewers: sdesmalen, fhahn, dmgreen.
Herald added subscribers: ctetreau, hiraditya, kristof.beyls.
Herald added a project: All.
TiehuZhang requested review of this revision.
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.

TTI::prefersVectorizedAddressing() try to vectorize the addresses that lead to loads. For aarch64, only gather/scatter (supported by SVE) can deal with vectors of addresses. This patch specializes the hook for AArch64, to return true only when we enable SVE.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D124612

Files:
  llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
  llvm/lib/Target/AArch64/AArch64TargetTransformInfo.h
  llvm/test/Transforms/LoopVectorize/AArch64/gather-do-not-vectorize-addressing.ll
  llvm/test/Transforms/LoopVectorize/AArch64/interleaved-vs-scalar.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D124612.425763.patch
Type: text/x-patch
Size: 9998 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20220428/455f4cd3/attachment.bin>