Mordante added a comment. In D129048#3669568 <https://reviews.llvm.org/D129048#3669568>, @aaron.ballman wrote:
> In D129048#3669069 <https://reviews.llvm.org/D129048#3669069>, @Mordante > wrote: > >> In D129048#3668905 <https://reviews.llvm.org/D129048#3668905>, >> @aaron.ballman wrote: >> >>> In D129048#3668846 <https://reviews.llvm.org/D129048#3668846>, @ldionne >>> wrote: >>> >>>> In D129048#3668594 <https://reviews.llvm.org/D129048#3668594>, @philnik >>>> wrote: >>>> >>>>> Also, please wait for #libc <https://reviews.llvm.org/tag/libc/> approval >>>>> next time. >>>> >>>> This, x1000. >>>> >>>> We go through the trouble of having excellent pre-commit testing and >>>> automatic review groups assigned to reviews, please don't bypass that. >>> >>> We certainly will keep this in mind for the future, thanks for pointing it >>> out! However, the precommit CI behavior confused multiple Clang >>> contributors (I also thought the bot was misconfigured because Clang tests >>> are never run against old Clang versions), the output did not clearly >>> identify what was wrong or how to address it, and the difficulties with >>> testing this locally are all things the libc++ maintainers should also keep >>> in mind. >> >> What can we do to make it easier for Clang contributors to understand the >> libc++ CI? >> Maybe we can discuss it on Discord. > > That's a great question to be asking, so thank you! I'm not on Discord (still > on IRC though), but I'm happy to chat about it via whatever means works for > us. > > At a high-level, I think it boils down to familiarity. If we can get the > libc++ CI to run as part of precommit CI (assuming precommit CI can be made > to run with reliable results, which is a bit of a question mark), then I > think that will help get all clang contributors more familiar with the libc++ > testing behavior over time, because we'll be far more used to seeing it when > we cause a failure. It would have also helped to catch the cause for the > initial revert where everyone thought the patch was ready to land. Another > thing that would help would be to get the libc++ test bots into the LLVM lab > (https://lab.llvm.org/buildbot/#/builders) and ensure that they're sending > failure emails to everyone on the blame list including people who land a > patch on behalf of others. It looks like we have one builder for libc++, but > it doesn't appear to be working recently: > https://lab.llvm.org/buildbot/#/builders/156. As @ldionne said we mainly use a pre-commit CI instead of a post-commit CI. I agree that it's probably a lack of familiarity, so let's try to remedy that. That will make everybody happier. I'll reach out to you on IRC either tomorrow or next week. Then we can discuss how we can get the clang contributors more familiar with libc++'s way of testing. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D129048/new/ https://reviews.llvm.org/D129048 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
