[PATCH] D61764: [LV] Suppress vectorization in some nontemporal cases

[Warren Ristow via Phabricator via llvm-commits]

wristow updated this revision to Diff 202378.
wristow added a comment.

Updated the patch to allow arbitrary alignment of `float` and `double` nt-stores for SSE4A.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D61764/new/

https://reviews.llvm.org/D61764

Files:
  include/llvm/Analysis/TargetTransformInfo.h
  include/llvm/Analysis/TargetTransformInfoImpl.h
  include/llvm/Transforms/Vectorize/LoopVectorizationLegality.h
  lib/Analysis/TargetTransformInfo.cpp
  lib/Target/X86/X86TargetTransformInfo.cpp
  lib/Target/X86/X86TargetTransformInfo.h
  lib/Transforms/Vectorize/LoopVectorizationLegality.cpp
  lib/Transforms/Vectorize/LoopVectorize.cpp
  test/Transforms/LoopVectorize/X86/nontemporal.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D61764.202378.patch
Type: text/x-patch
Size: 15486 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190531/07be16fe/attachment.bin>