[PATCH] D38506: [X86][SSE] Improve shuffling combining with horizontal operations

[Peter Cordes via Phabricator via llvm-commits]

pcordes accepted this revision.
pcordes added a comment.

ASM output changes are all obvious improvements.



================
Comment at: test/CodeGen/X86/vector-compare-results.ll:3532
+; SSE42-NEXT:    pcmpgtd {{[0-9]+}}(%rsp), %xmm7
+; SSE42-NEXT:    pextrb $12, %xmm7, %eax
 ; SSE42-NEXT:    andb $1, %al
----------------
The extra instructions before the dumb stuff are gone again now.  Yay?

Hopefully this is a sign that it's resistant to doing extra work in real extract situations, too.


Repository:
  rL LLVM

https://reviews.llvm.org/D38506