AVX-512: Shorten implementation of lowerV16X32VectorShuffle()
authorElena Demikhovsky <elena.demikhovsky@intel.com>
Tue, 2 Jun 2015 13:43:18 +0000 (13:43 +0000)
committerElena Demikhovsky <elena.demikhovsky@intel.com>
Tue, 2 Jun 2015 13:43:18 +0000 (13:43 +0000)
commitccbc17f896309be5bfe203a17c6c35bd420c335d
tree9909d7e55bfdfe6be8273b9009a67817b7353060
parent330e5f16d10736964e1cbec3fdf597b90c0ba6f7
AVX-512: Shorten implementation of lowerV16X32VectorShuffle()
using lowerVectorShuffleWithSHUFPS() and other shuffle-helpers routines.
Added matching of VALIGN instruction.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@238830 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/X86/X86ISelLowering.cpp
test/CodeGen/X86/avx512-shuffle.ll