Page MenuHomePhabricator

[bugpoint] construct a plausible TargetMachine for simplifyCFG
Needs ReviewPublic

Authored by vtjnash on Feb 25 2022, 11:15 AM.

Details

Reviewers
bjope
aeubanks
Summary

Added in https://reviews.llvm.org/D22845, I do not really know why
simplifyCFG here needs the TTI, so I assumed it would be better if it
had the precise information for the TM it is trying to bugpoint, and not
only the minimal DataLayout Concept.

(We can sprinkle a bit of const into here once https://reviews.llvm.org/D120518 lands)

Diff Detail

Event Timeline

vtjnash created this revision.Feb 25 2022, 11:15 AM
vtjnash requested review of this revision.Feb 25 2022, 11:15 AM
Herald added a project: Restricted Project. · View Herald TranscriptFeb 25 2022, 11:15 AM
vtjnash edited the summary of this revision. (Show Details)Feb 25 2022, 11:19 AM
vtjnash added reviewers: bjope, aeubanks.
vtjnash edited the summary of this revision. (Show Details)
vtjnash added subscribers: wsmoses, vchuravy.
Herald added a project: Restricted Project. · View Herald TranscriptMar 8 2022, 1:57 PM
bjope added a comment.Mar 10 2022, 1:55 AM

Sorry, I don't know much about bugpoint to understand if this is feasible or not. I doubt that the CFG simplification done by bugpoint really need to take target specific costs etc into account? We just want to simplify the flow, but not optimize it for a given target, right?
I don't find the motivation that strong and there are no test cases etc included that show any gains.

But maybe you could try to find some reviewers that know more about llvm-bugpoint. Unfortunately it seems like there haven't been many functional commits to the tool lately (looking at git log), so I do not really have any immediate candidates to add as reviewers.