Fix generic arguments being incorrectly offset in qualified trait casts #8266
Conversation
crates/hir_ty/src/tests/simple.rs
Outdated
| @@ -2564,3 +2564,46 @@ fn f() { | |||
| "#, | |||
| ) | |||
| } | |||
|
|
|||
| #[test] | |||
| fn qualified_as_trait_cast() { | |||
There was a problem hiding this comment.
Does this test belong here or in regressions? or maybe somewhere else?
There was a problem hiding this comment.
I guess it doesn’t matter if you add a cov-mark here.
There was a problem hiding this comment.
Do cov-marks not work across crates? The test(check) is in hir_ty but the fix(hit) is in hir_def and it isnt being hit(but the branch is definitely being executed)
There was a problem hiding this comment.
I'd put it in traits.rs since it's not that simple 😄 (Also this isn't really a cast, just a qualified path.)
There was a problem hiding this comment.
I never know how to call/name this, cause qualified path isn't specific enough in my eyes 😄 (guess I repeated the same mistake I did in the qualify path assist)
There was a problem hiding this comment.
Oups, indeed, cov-marks work only within a single crate. That's because they rely on cfg(test) to emit the code, and that is set only for the leaf crate.
There was a problem hiding this comment.
Ah that makes sense, figured it had to be something like that.
|
bors r+ |
|
Looks like this turned type mismatches into unknown types in
|
|
changelog fix fix generic argument lowering in qualified trait casts (helps with |
We reverse the segments and generic args of the lowered path after building it, this wasn't accounted for when inserting the self parameter in
Type as Traitsegments.Fixes #5886