projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 7f6ba17) | patch
Allow AVX vrsqrtps generation.
authorSanjay Patel <spatel@rotateright.com>
Fri, 24 Oct 2014 17:59:18 +0000 (17:59 +0000)
committerSanjay Patel <spatel@rotateright.com>
Fri, 24 Oct 2014 17:59:18 +0000 (17:59 +0000)
commit51fa1bcef3346f3d127ae405dc0de6011c19be24
tree4618f570c68d4b6f7f26c5f2ffbffb15c540a5f2tree | snapshot
parent7f6ba1758b0f00b19da9dac0a439e18bb4631107commit | diff
Allow AVX vrsqrtps generation.

This is a follow-on to r220570 that allows a 256-bit (v8f32)
version of vrsqrtps to be generated.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220579 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/X86/X86ISelLowering.cpp diff | blob | history
test/CodeGen/X86/sqrt-fastmath.ll diff | blob | history
C/C++ LLVM-based compilers that forbids OOTA behaviors