-
Notifications
You must be signed in to change notification settings - Fork 12.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Run x test --stage 1
in CI
#99529
Run x test --stage 1
in CI
#99529
Conversation
I think this will only run on PRs and not the main bors merge? that seems not great, it means you can hit semantic conflicts (although I guess those are rare in practice). @pietroalbini how can I tell if this will run the new tests on a full merge or not? |
err pietro is out - will just ask someone random I guess. r? rust-lang/infra-ci |
highfive pls r? rust-lang/infra-ci |
The new builder was only added to the pr: stanza in the configuration, so it won't run on a full build. This is something we should definitely fix. If we're going to run the builder in PR CI, I would probably recommend we drop fulldeps tests and the other bits that are still causing two full compiler builds for it -- IIRC, @jyn514 you were looking into that on Zulip at some point for other reasons, though I forget the details. |
yeah, that was #99141 - it ended up not having any time savings at all, which I think means stage 2 was still being compiled. Haven't had time to investigate - not sure it makes sense to block this change because of that. |
anyway, @Milo123459 I'm happy to approve if you change this to run in the full test suite as well as in PR checks |
@bors r+ rollup=iffy |
I'm not sure it makes sense to run this in PR checks. There's not really a time benefit that I can see, and those builders will add up pretty quickly given how many PRs we have. Hopefully, defects here are pretty rare, so it doesn't seem like a good trade-off to me to spend that CI budget on this builder. Is there some evidence that breaking stage 1 is common enough that we want the early warning at PR time rather than bors time? @bors r- |
ah, fair enough - I still had it in my mind that we were doing this instead of stage2 builds, which isn't the case. Yes, I agree it's better not to run in PR checks. |
Alright, stage1 tests are no longer run in PRs, only in the full test-suite. |
@bors r+ |
☀️ Test successful - checks-actions |
Finished benchmarking commit (482153b): comparison url. Instruction countThis benchmark run did not return any relevant results for this metric. Max RSS (memory usage)Results
CyclesResults
If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf. @rustbot label: -perf-regression Footnotes |
Fixes #99135
r? @jyn514