[x86] use PMOVMSK to replace memcmp libcalls for 16-byte equality

[x86] use PMOVMSK to replace memcmp libcalls for 16-byte equality

This is the payoff for D31156 - if a target has efficient comparison instructions for vector-sized equality,
we can replace memcmp calls with inline code that is both smaller and faster.

Differential Revision: https://reviews.llvm.org/D31290

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@298775 91177308-0d34-0410-b5e6-96231b3b80d8