[PATCH] D159036: [AMDGPU] Accept arbitrary sized sources in CalculateByteProvider

[Jeffrey Byrnes via Phabricator via llvm-commits]

jrbyrnes updated this revision to Diff 554812.
jrbyrnes added a comment.

Account for max scalar size of i256. Factor out common code.

Minor regressions in load-hi16.ll are caused by problematic heuristic hasNon16BitAccesses which will be addressed separately.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D159036/new/

https://reviews.llvm.org/D159036

Files:
  llvm/lib/Target/AMDGPU/SIISelLowering.cpp
  llvm/test/CodeGen/AMDGPU/idot4u.ll
  llvm/test/CodeGen/AMDGPU/insert_vector_elt.v2i16.ll
  llvm/test/CodeGen/AMDGPU/load-hi16.ll
  llvm/test/CodeGen/AMDGPU/permute_i8.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D159036.554812.patch
Type: text/x-patch
Size: 79608 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230830/8563b890/attachment.bin>