[PATCH] D47386: AMDGPU: Use better alignment for kernarg lowering

[Matt Arsenault via Phabricator via llvm-commits]

arsenm created this revision.
arsenm added reviewers: rampitec, kzhuravl, t-tye.
Herald added subscribers: tpr, dstuttard, yaxunl, nhaehnle, wdng.

This was just emitting loads with the ABI alignment
for the raw type. The true alignment is often better,
especially when an illegal vector type was scalarized.
The better alignment allows using a scalar load
more often.


https://reviews.llvm.org/D47386

Files:
  lib/Target/AMDGPU/SIISelLowering.cpp
  lib/Target/AMDGPU/SIISelLowering.h
  test/CodeGen/AMDGPU/fneg-fabs.f16.ll
  test/CodeGen/AMDGPU/half.ll
  test/CodeGen/AMDGPU/kernel-args.ll
  test/CodeGen/AMDGPU/reduce-store-width-alignment.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D47386.148650.patch
Type: text/x-patch
Size: 17899 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20180525/6caaa78c/attachment.bin>