[PATCH] D29601: AMDGPU: Simplify allowsMisalignedMemoryAccesses

[Matt Arsenault via Phabricator via llvm-commits]

arsenm updated this revision to Diff 87331.
arsenm added a comment.

Missed test update.

This changes the vectorizer to start combining private accesses which will need to be split up later. This isn't exactly ideal, but from an alignment perspective I think is correct. It might be desirable to add a different hook for the vectorizer to check


https://reviews.llvm.org/D29601

Files:
  lib/Target/AMDGPU/SIISelLowering.cpp
  test/CodeGen/AMDGPU/unaligned-load-store.ll
  test/Transforms/LoadStoreVectorizer/AMDGPU/merge-stores-private.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D29601.87331.patch
Type: text/x-patch
Size: 10086 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170207/4a0c1927/attachment.bin>