[PATCH] D143785: [X86] Add Extend shuffle pattern to vNf32 shuffles.

[Noah Goldstein via Phabricator via llvm-commits]

goldstein.w.n added inline comments.


================
Comment at: llvm/test/CodeGen/X86/merge-consecutive-loads-512.ll:310
+; ALL-NEXT:    vpandd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %zmm0, %zmm0
 ; ALL-NEXT:    retq
 ;
----------------
goldstein.w.n wrote:
> RKSimon wrote:
> > Hmm - shuffle combining should already catch this - any idea why it doesn't?
> Yeah, for the float types it is beneficial to do `lowerShuffleAsBlend` before `lowerShuffleAsZeroOrAnyExtend` (so added when added `lowerShuffleAsZeroOrAnyExtend`), `lowerShuffleAsBlend` is what implements the the logic-op shuffles. I should/will update commit message to reflect that.
Moved the blend stuff to a new seperate patch: D143856


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D143785/new/

https://reviews.llvm.org/D143785