projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: c5d7c6e) | patch
[InstCombine] fix complexity canonicalization with fake unary vector ops
authorSanjay Patel <spatel@rotateright.com>
Sat, 13 Oct 2018 16:15:37 +0000 (16:15 +0000)
committerSanjay Patel <spatel@rotateright.com>
Sat, 13 Oct 2018 16:15:37 +0000 (16:15 +0000)
commit47579b21e2043a69832a1378e85f1e545abfef08
tree2f0dda4702edd27635e61d71851bceac0bb24032tree | snapshot
parentc5d7c6e5f68476fae3144fb2500bbf6d807555b3commit | diff
[InstCombine] fix complexity canonicalization with fake unary vector ops

This is a preliminary step to avoid regressions when we add
an actual 'fneg' instruction to IR. See D52934 and D53205.

llvm-svn: 344458
llvm/lib/Transforms/InstCombine/InstCombineInternal.h diff | blob | history
llvm/test/Transforms/InstCombine/operand-complexity.ll diff | blob | history
Domain: System / Toolchain;