[PATCH] D18566: [x86] use SSE/AVX ops for non-zero memsets (PR27100)

[H.J Lu via llvm-commits]

hjl.tools added a comment.

For 16-byte memset, we can generate

  	movzbl	4(%esp), %eax
  	imull	$16843009, %eax, %eax
  	movd	%eax, %xmm0
  	movl	dst, %eax
  	pshufd	$0, %xmm0, %xmm0
  	movdqu	%xmm0, (%eax)
  	ret


http://reviews.llvm.org/D18566