diff options
| author | Craig Topper <craig.topper@intel.com> | 2018-01-18 07:44:09 +0000 |
|---|---|---|
| committer | Craig Topper <craig.topper@intel.com> | 2018-01-18 07:44:09 +0000 |
| commit | 83b0a98902077b2bc95bde4162e0f210d4b50f61 (patch) | |
| tree | eb9c0decf356cb0b007a2cdff44d6d5bd462cf2c /llvm/test/Transforms/LoopVectorize | |
| parent | 21c8a8fa499c2e5dcafeb82ab2f438e41abe7cb7 (diff) | |
| download | bcm5719-llvm-83b0a98902077b2bc95bde4162e0f210d4b50f61.tar.gz bcm5719-llvm-83b0a98902077b2bc95bde4162e0f210d4b50f61.zip | |
[X86] Use vmovdqu64/vmovdqa64 for unmasked integer vector stores for consistency with loads.
Previously we used 64 for vXi64 stores and 32 for everything else. This change uses 64 for everything just like do for loads.
llvm-svn: 322820
Diffstat (limited to 'llvm/test/Transforms/LoopVectorize')
| -rw-r--r-- | llvm/test/Transforms/LoopVectorize/X86/avx512.ll | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/llvm/test/Transforms/LoopVectorize/X86/avx512.ll b/llvm/test/Transforms/LoopVectorize/X86/avx512.ll index 1eb1cd3f5d7..c7844c11306 100644 --- a/llvm/test/Transforms/LoopVectorize/X86/avx512.ll +++ b/llvm/test/Transforms/LoopVectorize/X86/avx512.ll @@ -7,7 +7,7 @@ target triple = "x86_64-apple-macosx10.9.0" ; loop. ; CHECK-LABEL: f: -; CHECK: vmovdqu32 %zmm{{.}}, +; CHECK: vmovdqu64 %zmm{{.}}, ; CHECK-NOT: %ymm define void @f(i32* %a, i32 %n) { |
