[X86] Enable commuting of VUNPCKHPD to VMOVLHPS to enable load folding by using VMOVL...
authorCraig Topper <craig.topper@intel.com>
Wed, 18 Jul 2018 07:31:32 +0000 (07:31 +0000)
committerCraig Topper <craig.topper@intel.com>
Wed, 18 Jul 2018 07:31:32 +0000 (07:31 +0000)
commit92ea7a7b4877f931770a2bf2c997db0d4bf8bf86
tree5a139120a83011a1760130f014f94cfaf7f16b03
parenta2bbfa21ce952fd02bfb584e214b1397dc81f53e
[X86] Enable commuting of VUNPCKHPD to VMOVLHPS to enable load folding by using VMOVLPS with a modified address.

This required an annoying amount of tablegen multiclass changes to make only VUNPCKHPDZ128rr commutable.

llvm-svn: 337357
llvm/lib/Target/X86/X86InstrAVX512.td
llvm/lib/Target/X86/X86InstrInfo.cpp
llvm/lib/Target/X86/X86InstrSSE.td
llvm/test/CodeGen/X86/fma.ll
llvm/test/CodeGen/X86/vector-shuffle-128-v2.ll