[PATCH] D15595: AMDGPU: Switch barrier intrinsics to using convergent

[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.

noduplicate prevents unrolling of small loops that happen to have
barriers in them. If a loop has a barrier in it, it is OK to duplicate
it for the unroll.

http://reviews.llvm.org/D15595

Files:
  lib/Target/AMDGPU/AMDGPUIntrinsics.td
  test/CodeGen/AMDGPU/addrspacecast.ll
  test/CodeGen/AMDGPU/array-ptr-calc-i32.ll
  test/CodeGen/AMDGPU/drop-mem-operand-move-smrd.ll
  test/CodeGen/AMDGPU/ds-negative-offset-addressing-mode-loop.ll
  test/CodeGen/AMDGPU/ds-sub-offset.ll
  test/CodeGen/AMDGPU/ds_read2.ll
  test/CodeGen/AMDGPU/ds_read2_superreg.ll
  test/CodeGen/AMDGPU/ds_read2st64.ll
  test/CodeGen/AMDGPU/ds_write2.ll
  test/CodeGen/AMDGPU/ds_write2st64.ll
  test/CodeGen/AMDGPU/flat-address-space.ll
  test/CodeGen/AMDGPU/indirect-private-64.ll
  test/CodeGen/AMDGPU/llvm.AMDGPU.div_fmas.ll
  test/CodeGen/AMDGPU/merge-stores.ll
  test/CodeGen/AMDGPU/schedule-vs-if-nested-loop-failure.ll
  test/CodeGen/AMDGPU/si-triv-disjoint-mem-access.ll
  test/CodeGen/AMDGPU/store-barrier.ll
  test/CodeGen/AMDGPU/wait.ll
  test/Transforms/LoopUnroll/AMDGPU/lit.local.cfg
  test/Transforms/LoopUnroll/AMDGPU/unroll-barrier.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D15595.43079.patch
Type: text/x-patch
Size: 15803 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20151216/bd2b67bb/attachment.bin>