[X86][XOP] XOP's general v16i8 shifts will be used instead of v8i16 shift + mask.

Tweak cost model to match what lowering actually does.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@303013 91177308-0d34-0410-b5e6-96231b3b80d8
3 files changed