[PATCH] D61158: [SimplifyCFG] use fshr instead of shl/lshr/or

[Shawn Landden via Phabricator via llvm-commits]

shawnl created this revision.
shawnl added a reviewer: jmolloy.
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.
shawnl updated this revision to Diff 196749.

We already try (but fail due to lack of sub op)
to convert this to fshr in AggressiveInstCombine.cpp:92.

If a target cannot lower fshr/fshl then it needs to be fixed.
Performance on targets that have such an instruction should
not be penalized.

If you really don't want this I can remove it, but fshr is NOT experimental.


https://reviews.llvm.org/D61158

Files:
  lib/Transforms/Utils/SimplifyCFG.cpp
  test/Transforms/SimplifyCFG/rangereduce.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D61158.196749.patch
Type: text/x-patch
Size: 5121 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190425/11849224/attachment-0001.bin>