projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: edf100e) | patch
AMDGPU: Add volatile to test loads and stores
authorMatt Arsenault <Matthew.Arsenault@amd.com>
Tue, 12 Apr 2016 13:38:18 +0000 (13:38 +0000)
committerMatt Arsenault <Matthew.Arsenault@amd.com>
Tue, 12 Apr 2016 13:38:18 +0000 (13:38 +0000)
commit44e5483adaf5d7696761fb732486961a51b76393
treec2e8c8a4bab4e0f1b68c3541f4c3cb929747bc0btree | snapshot
parentedf100e4e7479b64260c81f6c6a2a2d308bd97e0commit | diff
AMDGPU: Add volatile to test loads and stores

When the memory vectorizer is enabled, these tests break.
These tests don't really care about the memory instructions,
and it's easier to write check lines with the unmerged loads.

llvm-svn: 266071
19 files changed:
llvm/test/CodeGen/AMDGPU/commute_modifiers.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/dagcombine-reassociate-bug.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/ds_write2.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/ds_write2st64.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/fdiv.f64.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/fma-combine.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/fmax3.f64.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/fmax_legacy.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/fmin_legacy.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/fmul-2-combine-multi-use.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/fmuladd.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/llvm.amdgcn.div.fmas.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/llvm.amdgcn.div.scale.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/mad-combine.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/mad-sub.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/madmk.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/rsq.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/si-instr-info-correct-implicit-operands.ll diff | blob | history
llvm/test/CodeGen/AMDGPU/v_mac.ll diff | blob | history
Domain: System / Toolchain;