projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 9aa09ef) | patch
[X86] AVX512: Specify compressed displacement for vmovntdqa
authorAdam Nemet <anemet@apple.com>
Wed, 18 Jun 2014 16:51:07 +0000 (16:51 +0000)
committerAdam Nemet <anemet@apple.com>
Wed, 18 Jun 2014 16:51:07 +0000 (16:51 +0000)
commitded81a810c93b94ee4041b0409daf4553e947e9f
tree6cf795edc87444488c1c688da4b9eb60a7b32e97tree | snapshot
parent9aa09ef30f0bc65aeb848f0cdc6b5987fc884d16commit | diff
[X86] AVX512: Specify compressed displacement for vmovntdqa

Use the max 64-bit element size with EVEX_CD8.  This should work since element
size is ignored for a full-vector access (FVM).

llvm-svn: 211175
llvm/lib/Target/X86/X86InstrAVX512.td diff | blob | history
llvm/test/MC/X86/avx512-encodings.s diff | blob | history
Domain: System / Toolchain;