[PATCH] D54512: [X86] Add -x86-experimental-vector-widening support to reduceVMULWidth and combineMulToPMADDWD

[Craig Topper via Phabricator via llvm-commits]

craig.topper updated this revision to Diff 174249.
craig.topper added a comment.

Add assert messages.  Going to look at the extra shuffle separately. I think we may need to try to reduceVMULWidth before combineMulToPMADDWD on pre-SSE4.1 targets. The mul_4xi8 test case in shrink_vmul.ll shows the same issue even without widening enabled.


https://reviews.llvm.org/D54512

Files:
  lib/Target/X86/X86ISelLowering.cpp
  test/CodeGen/X86/shrink_vmul-widen.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D54512.174249.patch
Type: text/x-patch
Size: 11227 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20181115/758729d2/attachment.bin>