[PATCH] D94232: [LoopRotate] Add PrepareForLTO stage, avoid rotating with inline cands (WIP).

[Sanne Wouda via Phabricator via llvm-commits]

sanwou01 added a comment.

So on SPEC 2006 this fixes astar (+9.6%) as well as shakes things up enough to "fix" h264ref (+9.0%, see D93946 <https://reviews.llvm.org/D93946>). Other notable changes are libquantum (+3.2%) and omnetpp (-2.8%). Geomean is +1.5%.

On SPEC 2017 there is still a problem with omnetpp_r (runtime error), which I'll have a look at next, but is otherwise neutral on geomean with only minor wobbles (+1%, -0.5% at most) on individual benchmarks.

With this having a substantial impact on performance, I'm quite keen to help get this in soon, ideally before the LLVM 12 branch, so please let me know if there is anything else I can do to help.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D94232/new/

https://reviews.llvm.org/D94232