[PATCH] D40059: AMDGPU: Select DS insts without m0 initialization

[Matt Arsenault via Phabricator via llvm-commits]

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

GFX9 stopped using m0 for most DS instructions. Select
a different instruction without the use. I think this will
be less error prone than trying to manually maintain m0
uses as needed.


https://reviews.llvm.org/D40059

Files:
  lib/Target/AMDGPU/AMDGPU.td
  lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp
  lib/Target/AMDGPU/AMDGPUSubtarget.h
  lib/Target/AMDGPU/DSInstructions.td
  lib/Target/AMDGPU/SIInstrInfo.td
  lib/Target/AMDGPU/SILoadStoreOptimizer.cpp
  test/CodeGen/AMDGPU/atomic_cmp_swap_local.ll
  test/CodeGen/AMDGPU/atomic_load_add.ll
  test/CodeGen/AMDGPU/atomic_load_sub.ll
  test/CodeGen/AMDGPU/ds_read2.ll
  test/CodeGen/AMDGPU/ds_read2st64.ll
  test/CodeGen/AMDGPU/ds_write2.ll
  test/CodeGen/AMDGPU/ds_write2st64.ll
  test/CodeGen/AMDGPU/hsa-metadata-kernel-debug-props.ll
  test/CodeGen/AMDGPU/indirect-addressing-si.ll
  test/CodeGen/AMDGPU/llvm.amdgcn.atomic.dec.ll
  test/CodeGen/AMDGPU/llvm.amdgcn.atomic.inc.ll
  test/CodeGen/AMDGPU/load-hi16.ll
  test/CodeGen/AMDGPU/load-local-f32.ll
  test/CodeGen/AMDGPU/load-local-f64.ll
  test/CodeGen/AMDGPU/load-local-i1.ll
  test/CodeGen/AMDGPU/load-local-i16.ll
  test/CodeGen/AMDGPU/load-local-i32.ll
  test/CodeGen/AMDGPU/load-local-i64.ll
  test/CodeGen/AMDGPU/load-local-i8.ll
  test/CodeGen/AMDGPU/local-64.ll
  test/CodeGen/AMDGPU/local-atomics.ll
  test/CodeGen/AMDGPU/local-atomics64.ll
  test/CodeGen/AMDGPU/packed-op-sel.ll
  test/CodeGen/AMDGPU/store-local.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D40059.122954.patch
Type: text/x-patch
Size: 196373 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20171115/74fe31d5/attachment-0001.bin>