-
Notifications
You must be signed in to change notification settings - Fork 90
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
"Pass, then timeout" should not be considered a spurious regression #739
Comments
other possible solutions might be to find some way to introduce retrying or extended second-pass build times that make this work. |
The reason why timeouts regressions are marked as spurious is because the amount of time it takes to build a crate varies wildly between builds, depending on whether the dependencies were cached or not in the target directory. If we didn't mark those as spurious every crater run would contain spurious regressions in the regressed crates list. |
I don't think this is what you meant but since afc6496 spurious regressed crates are included in the Maybe it would help if the crater bot would mention an unusually high amount of spuriously regressed crates in addition to regressed and fixed crates, or spuriously regressed could be a sub-category under regressed? i.e.
|
This sounds like something along the lines of rust-lang/cargo#2644 might help as it would allow to make a two phase build, first deps only and then the crate itself. |
Issue is visible in this crater report:
-Znext-solver=coherence
rust#121848 (comment)Crater was being used to diagnose usage of the new trait solver, which has hang issues in some cases.
This caused rust-lang/rust#130056 to be opened after landing rust-lang/rust#121848 even though we did notice essentially all the nalgebra dependents failed this way:
The text was updated successfully, but these errors were encountered: