Update Bazel config for
https://github.com/llvm/llvm-project/commit/6103fdfab4
by deleting the llvm-elfabi target.
Details
- Reviewers
rupprecht - Commits
- rGef648df281e0: [Bazel] Update for 6103fdfab4
Diff Detail
- Repository
- rG LLVM Github Monorepo
Event Timeline
Generally if something's sent for review, it should wait until it's approved before committing - but not everything needs to be sent for review. Changes to the Bazel BUILD files seem to be fine to be committed directly/for post commit review, especially by you :)
Ah I was seeking post-commit review, but wanted to land this trivial fix to unbreak things. What is the right way to say "please review this post-commit" then?
I sometimes include some callout in the commit message "I'm open to other ideas/feedback" or the like (were there any particular questions/aspects of review you wanted to get a second set of eyes on?) - or a reply to the commit email (or on the phab commit entry - though personally I prefer the commit email, since that's the authoritative record & replying to the phab commit entry creates a completely different email thread/subject line, etc)
Ah no I mean I want post-commit as opposed to pre-commit review because this change is trivial and should be uncontroversial :-D I deleted configuration for a target that was deleted.
The LLVM review process seems very weird to me, where review happens in one of several different places that aren't linked. I know I struggle when a commit doesn't have a phab review link attached (I'm not signed up for llvm-commits) and I can't figure out where I should mention some issue. https://llvm.org/docs/CodeReview.html#what-tools-are-used-for-code-review says that post-commit review can happen on phab, so I was attempting to create a place for that to be done in the same way as pre-commit review (I don't really see any reason why we'd want it to be split between tools). I'm of course fine doing whatever the generally-accepted practice in the community is, though threads like https://groups.google.com/g/llvm-dev/c/VTSwu2q5qnc suggest to me that there isn't really a generally-accepted practice.