projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 0bacecf) | patch
[X86] AVX512 fallback for floating-point scalar selects
authorZvi Rackover <zvi.rackover@intel.com>
Wed, 26 Oct 2016 14:12:46 +0000 (14:12 +0000)
committerZvi Rackover <zvi.rackover@intel.com>
Wed, 26 Oct 2016 14:12:46 +0000 (14:12 +0000)
commitaa3402b41ef33d49e21243e4fbfcba5bc4440dd3
treed8fd83501862f477d3471043d0298e026cadde38tree | snapshot
parent0bacecfb32ff32ccaa7e2ce950fa6e28f3b53325commit | diff
[X86] AVX512 fallback for floating-point scalar selects

Summary:
In the case where of 'select i1 , f32, f32' or select i1, f64, f64 prefer lowering to masked-moves over branches.

Fixes pr30561

Reviewers: igorb, aymanmus, delena

Differential Revision: https://reviews.llvm.org/D25310

llvm-svn: 285196
llvm/lib/Target/X86/X86ISelLowering.cpp diff | blob | history
llvm/test/CodeGen/X86/avx512-select.ll diff | blob | history
Domain: System / Toolchain;