Improve diagnostics for HostEffectPredicate in the new solver
#132345
Conversation
rustbot
added
S-waiting-on-review
compiler-errors
changed the title
HostEffectPredicate in the new solver
|
☔ The latest upstream changes (presumably #132371) made this pull request unmergeable. Please resolve the merge conflicts. |
compiler-errors
force-pushed
the
fx-diag
branch
from
0cbc3ae to
bb76b62
Compare
|
☔ The latest upstream changes (presumably #132479) made this pull request unmergeable. Please resolve the merge conflicts. |
compiler-errors
force-pushed
the
fx-diag
branch
from
bb76b62 to
b921c6a
Compare
|
Some changes occurred to the core trait solver cc @rust-lang/initiative-trait-system-refactor |
|
r? lcnr |
| @@ -16,6 +16,11 @@ error[E0277]: the trait bound `T: ~const Bar` is not satisfied | |||
| LL | type Assoc<T> = C<T> | |||
| | ^^^^ | |||
| | | |||
| note: required for `C<T>` to implement `~const Bar` | |||
There was a problem hiding this comment.
This is the net effect of the change.
| @@ -14,9 +14,9 @@ trait Trait { | |||
|
|
|||
| const fn unqualified<T: Trait>() { | |||
| T::Assoc::func(); | |||
| //~^ ERROR the trait bound `T: ~const Trait` is not satisfied | |||
| //~^ ERROR the trait bound `<T as Trait>::Assoc: ~const Trait` is not satisfied | |||
There was a problem hiding this comment.
We don't treat the GAT where clauses as where clauses for a goal source... should we?
There was a problem hiding this comment.
depends on where and how we prove them. I think for diagnostics, the answer should be yes, for cycle handling maybe no?
| impl_const_condition_bound_count, | ||
| parent_host_pred, | ||
| )); | ||
| impl_const_condition_bound_count += 1; |
There was a problem hiding this comment.
why use a separate counter here? We should only ever track one of the two, shouldn't we?
|
r=me after nit |
Adds derived cause for host effect predicates. Some diagnostics regress, but that's connected to the fact that our predicate visitor doesn't play well with aliases just yet.