bjoernager/rust
bjoernager/rust
1
Fork 0
Code Issues Pull requests Activity

Rollup merge of #109675 - compiler-errors:object-heck, r=lcnr

Browse source 
Do not consider elaborated projection predicates for objects in new solver

Object types have projection bounds which are elaborated during astconv. There's no need to do it again for projection goals, since that'll give us duplicate projection candidatesd that are distinct up to regions due to the fact that we canonicalize every region to a separate variable. See quick example below the break for a better explanation.

Discussed this with lcnr, and adding a stop-gap until we get something like intersection region constraints (or modify canonicalization to canonicalize identical regions to the same canonical regions) -- after which, this will hopefully not matter and may be removed.

r? `@lcnr`

---

See `tests/ui/traits/new-solver/more-object-bound.rs`:

Consider a goal: `<dyn Iter<'a, ()> as Iterator>::Item = &'a ()`.

After canonicalization: `<dyn Iter<'!0r, (), Item = '!1r ()> as Iterator>::Item == &!'2r ()`
* First object candidate comes from the item bound in the dyn's bounds itself, giving us `<dyn Iter<'!0r, (), Item = '?!r ()> as Iterator>::Item == &!'1r ()`. This gives us one region constraint: `!'1r == !'2r`.
* Second object candidate comes from elaborating the principal trait ref, gives us `<dyn Iter<'!0r, (), Item = '!1r ()> as Iterator>::Item == &!'0r ()`. This gives us one region constraint: `!'0r == !'2r`.
* Oops! Ambiguity!
This commit is contained in:
Matthias Krüger 2023-03-29 21:19:49 +02:00 • committed by GitHub
commit 80e988d7a6
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
4 changed files with 37 additions and 9 deletions
Download patch file Download diff file Expand all files Collapse all files

19
compiler/rustc_trait_selection/src/solve/assembly.rs
View file

@ -6,6 +6,7 @@ use super::trait_goals::structural_traits::*;
use super::{EvalCtxt, SolverMode};
use crate::traits::coherence;
use itertools::Itertools;
use rustc_data_structures::fx::FxIndexSet;
use rustc_hir::def_id::DefId;
use rustc_infer::traits::query::NoSolution;
use rustc_infer::traits::util::elaborate_predicates;
@ -489,9 +490,21 @@ impl<'tcx> EvalCtxt<'_, 'tcx> {
};
let tcx = self.tcx();
for assumption in
elaborate_predicates(tcx, bounds.iter().map(|bound| bound.with_self_ty(tcx, self_ty)))
{
let own_bounds: FxIndexSet<_> =
bounds.iter().map(|bound| bound.with_self_ty(tcx, self_ty)).collect();
for assumption in elaborate_predicates(tcx, own_bounds.iter().copied()) {
// FIXME: Predicates are fully elaborated in the object type's existential bounds
// list. We want to only consider these pre-elaborated projections, and not other
// projection predicates that we reach by elaborating the principal trait ref,
// since that'll cause ambiguity.
//
// We can remove this when we have implemented intersections in responses.
if assumption.to_opt_poly_projection_pred().is_some()
&& !own_bounds.contains(&assumption)
{
continue;
}
match G::consider_object_bound_candidate(self, goal, assumption) {
Ok(result) => {
candidates.push(Candidate { source: CandidateSource::BuiltinImpl, result })