Log the name of the dependency that changed #13883
Conversation
|
Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @epage (or someone else) some time within the next two weeks. Please see the contribution instructions for more information. Namely, in order to ensure the minimum review times lag, PR authors and assigned reviewers should ensure that the review label (
|
rustbot
added
A-rebuild-detection
| @@ -622,7 +622,7 @@ fn dylib() { | |||
| [FRESH] registry-shared v0.0.1 | |||
| [COMPILING] registry v0.0.1 | |||
| [RUNNING] `rustc --crate-name registry [..] | |||
| [DIRTY] bar v0.0.0 ([..]): dependency info changed | |||
| [DIRTY] bar v0.0.0 ([..]): dependency info changed (registry_shared) | |||
There was a problem hiding this comment.
This message looks strange to me. Should it not output registry rather than registry_shared here?
There was a problem hiding this comment.
Part of the problem is we only report the first dependency, rather than all.
I suspect what is happening here is we are reusing registry_shared with a version from a couple builds back. This is a different hash than the immediately prior build, so it requires rebuilding bar but this isn't quite clear without more context.
I wonder if this change will cause more confusion than good and we should defer to #2904 which would add a cargo report subcommand for reporting why the last command rebuilt and we can include a lot more context.
There was a problem hiding this comment.
See #13859 (comment), yet it might need some design discussions before proceed.
|
Going to close this as it needs more designs before making a PR. Thank you for being interested in this. |
|
Fair enough, but the issue should probably not be |
Fixes #13859 by adding the name of the dependency to the "dependency info changed" log message.
There is a surprising test output after this. See code review.