projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: a27f54b) | patch
[AVX-512] Don't add an additional rounding mode operand to the avx512 vcvtps2ph intri...
authorCraig Topper <craig.topper@gmail.com>
Wed, 21 Sep 2016 03:58:44 +0000 (03:58 +0000)
committerCraig Topper <craig.topper@gmail.com>
Wed, 21 Sep 2016 03:58:44 +0000 (03:58 +0000)
commitd868870f17148361b30b1adce9b1243b9bd4e021
tree7729f1075fdf626e79f91ce9211be90462a2dabatree | snapshot
parenta27f54b4d9ae684be3c040058aafdbcbf5c6d92ccommit | diff
[AVX-512] Don't add an additional rounding mode operand to the avx512 vcvtps2ph intrinsic lowering.

There was no way to control its value so it was always FROUND_CURRENT making it unnecessary. The true rounding mode is encoded in the immediate operand of the instruction.

This also removes the pattern from the rb form of the instructions since there is no way to specify the FROUND_NO_EXC rounding mode it required.

llvm-svn: 282052
llvm/lib/Target/X86/X86InstrAVX512.td diff | blob | history
llvm/lib/Target/X86/X86InstrFragmentsSIMD.td diff | blob | history
llvm/lib/Target/X86/X86IntrinsicsInfo.h diff | blob | history
Domain: System / Toolchain;