[PATCH] D14268: AMDGPU: Split x8 and x16 vector loads instead of scalarize

[Matt Arsenault via llvm-commits]

arsenm created this revision.
arsenm added a reviewer: tstellarAMD.
arsenm added a subscriber: llvm-commits.
Herald added a subscriber: arsenm.

The one regression in the builtin tests is in the read2 test which now
(again) has many extra copies, but this should be solved once the pass
is replaced with a DAG combine.

http://reviews.llvm.org/D14268

Files:
  lib/Target/AMDGPU/AMDGPUISelLowering.cpp
  lib/Target/AMDGPU/SIISelLowering.cpp
  test/CodeGen/AMDGPU/cvt_f32_ubyte.ll
  test/CodeGen/AMDGPU/ds_read2_superreg.ll
  test/CodeGen/AMDGPU/global-extload-i32.ll
  test/CodeGen/AMDGPU/half.ll
  test/CodeGen/AMDGPU/load.ll
  test/CodeGen/AMDGPU/merge-stores.ll
  test/CodeGen/AMDGPU/reorder-stores.ll
  test/CodeGen/AMDGPU/salu-to-valu.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D14268.39007.patch
Type: text/x-patch
Size: 23105 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20151103/7e84ffef/attachment.bin>