Remove redundant match arm for RPITITs #113340
Conversation
rustbot
added
S-waiting-on-review
| @@ -122,13 +122,6 @@ fn variance_of_opaque(tcx: TyCtxt<'_>, item_def_id: LocalDefId) -> &[ty::Varianc | |||
| { | |||
| self.visit_opaque(*def_id, substs) | |||
| } | |||
| // FIXME(-Zlower-impl-trait-in-trait-to-assoc-ty) check whether this is necessary | |||
There was a problem hiding this comment.
Are you certain that this is correct to remove, or did you just remove this because no UI tests failed?
There was a problem hiding this comment.
Isn't it just equivalent to the previous arm?. Like in this new scheme, def_kind is going to be OpaqueTy so this branch is never gonna be exercised.
There was a problem hiding this comment.
Are you sure it's not being called on any of the GAT def ids?
There was a problem hiding this comment.
Ok, going to double check tomorrow. I guess you may be implying that for impls is executed because for traits shouldn't but I wouldn't be sure what would be the effect of calling this on impl's GAT. Going to check again tomorrow.
There was a problem hiding this comment.
In any case, given that all test pass here would be good to come up with a test case that fails with this branch removed and works if we add this back.
There was a problem hiding this comment.
For traits, walking the fn_sig will give you back projections, so it may happen. Yeah, this at least needs a bit more checking.
|
☔ The latest upstream changes (presumably #113377) made this pull request unmergeable. Please resolve the merge conflicts. |
spastorino
changed the title
|
I've just rebased this PR and didn't yet properly go over the things we've discussed. |
|
☔ The latest upstream changes (presumably #113308) made this pull request unmergeable. Please resolve the merge conflicts. |
This goes on top of #112988.
r? @compiler-errors