[PATCH] D43792: [PatternMatch, InstSimplify] allow undef elements when matching vector -0.0

[Sanjay Patel via Phabricator via llvm-commits]

spatel created this revision.
spatel added reviewers: craig.topper, RKSimon, efriedma, arsenm.
Herald added subscribers: wdng, mcrosier.

Our undef simplification with FP binops - including constant folding - is incomplete and doesn't match the DAG. 
But I think it's ok to add the FP equivalent of https://reviews.llvm.org/D42818 and use it for the few cases in InstSimplify of -0.0 folds (that's the only current use of m_NegZero() that I see).


https://reviews.llvm.org/D43792

Files:
  include/llvm/IR/PatternMatch.h
  test/Transforms/InstSimplify/floating-point-arithmetic.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D43792.136000.patch
Type: text/x-patch
Size: 4545 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20180226/313495db/attachment.bin>