-
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
small fix to closure debuginfo #38483
Conversation
r? @arielb1 (rust_highfive has picked a reviewer for you, use r? to override) |
I just hit the wrong button and submitted this without a description, sorry. EDIT(@eddyb): moved the rest of this comment to the PR description. |
r? @eddyb seems ok to me, but I'm probably not best to review |
@bors r+ |
📌 Commit e1d8806 has been approved by |
@bors: p=1 fixes a regression, so I think we'll want to land this soon hopefully |
Rollup of 29 pull requests - Successful merges: #37761, #38006, #38131, #38150, #38158, #38171, #38208, #38215, #38236, #38245, #38289, #38302, #38315, #38346, #38388, #38395, #38398, #38418, #38432, #38451, #38463, #38468, #38470, #38471, #38472, #38478, #38486, #38493, #38498 - Failed merges: #38271, #38483
Marking as beta-accepted. Small (if...dense) patch, regression. cc @rust-lang/compiler |
Definitely we want a test here... |
@nikomatsakis IIRC we had a test but it wasn't running on the bots, @alexcrichton knows more. |
Oh yes #37429 should have been prevented from landing in the first place due to it causing a regression to an existing test that this PR fixes. That test was erroneously not running and has since been fixed to run on all PRs. In that sense I believe we've already got tests checked in for this. |
I'm not sure what dense means but suspect it is bad; if you're referring to the description and the commit it's because I don't fully understand what's going on. @eddyb gets credit for this fix, and they tried to explain but not with much luck, I'm afraid. I forget which test specifically checks this, but there are already tests examining closure-local variables. |
@camlorn I didn't mean dense in a bad way, just meant that there are few lines changed, but the correctness of the changes is non-obvious -- although re-reading they actually look pretty simple. |
This is fallout from #37429. The faking we do to make closures appear to have local variables was broken and made it through CI because the debuginfo tests got turned off.
This probably needs a dedicated test, but I wanted to get this in as quickly as possible because we probably need to backport this to the beta.
r? @alexcrichton