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

Fri Apr 26 00:46:57 PDT 2019

lebedev.ri added a comment.

The description on the patch is really unhelpful.
It should explain why the change is needed/should happen, not how great the new intrinsic is, how bad it is not to support it...
Also, as usual, please upload all patches with full context.

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

There wouldn't be any sub op here, because these are constants,
and we don't seem to canonicalize this rotate pattern with constants to fshl in instcombine/aggressiveinstcombine.
That is a separate issue, i'm not sure if it's a bug or not, so please file a bug.

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D61158/new/

https://reviews.llvm.org/D61158