projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: a3b7527) | patch
[AVX-512] Add scalar vfmsub/vfnmsub mask3 intrinsics
authorCraig Topper <craig.topper@gmail.com>
Wed, 26 Oct 2016 04:59:58 +0000 (04:59 +0000)
committerCraig Topper <craig.topper@gmail.com>
Wed, 26 Oct 2016 04:59:58 +0000 (04:59 +0000)
commit812d3d30aede80b5f9e5b1fc0484700ecaf1afef
tree029ecae74275eb078a716136ec5ad8ca1fa66b1etree | snapshot
parenta3b7527ccb4400be6978b6ed6283afc202ecd753commit | diff
[AVX-512] Add scalar vfmsub/vfnmsub mask3 intrinsics

Summary: Clang's intrinsic header currently tries to negate the third operand of a vfmadd mask3 in order to create vfmsub, but this fails isel. This patch adds scalar vfmsub and vfnmsub mask3 that we can use instead to avoid the negate. This is consistent with the packed instructions.

Reviewers: igorb, delena

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D25933

llvm-svn: 285173
llvm/include/llvm/IR/IntrinsicsX86.td diff | blob | history
llvm/lib/Target/X86/X86IntrinsicsInfo.h diff | blob | history
llvm/test/CodeGen/X86/avx512-intrinsics.ll diff | blob | history
Domain: System / Toolchain;