use ProjectionPredicate instead of AliasRelate by khyperia · Pull Request #158606 · rust-lang/rust

khyperia · 2026-06-30T09:03:04Z

View all comments

part of #155345

this PR is split into three commits:

refactoring super_combine_{tys,consts}, should not change behavior at all, just reorganizing code
convert to use ProjectionPredicate instead of AliasRelate in super_combine_tys and generalization (this is the important change for this PR)
deleting the now-dead AliasRelate predicate

r? @lcnr

rustbot · 2026-06-30T09:03:08Z

Some changes occurred to the core trait solver

cc @rust-lang/initiative-trait-system-refactor

changes to inspect_obligations.rs

cc @lcnr

changes to the core type system

cc @lcnr

This PR changes rustc_public

cc @oli-obk, @celinval, @ouz-a, @makai410

khyperia · 2026-06-30T09:08:40Z

 fn test<T: Foo1<Assoc1 = <T as Foo2>::Assoc2> + Foo2<Assoc2 = <T as Foo1>::Assoc1>>() {
    needs_bar::<T::Assoc1>();
    //~^ ERROR: the trait bound `<T as Foo2>::Assoc2: Bar` is not satisfied
+    //~| ERROR: the size for values of type `<T as Foo2>::Assoc2` cannot be known at compilation time


I'm unsure about this one, to be honest. It makes sense why it can't prove Sized, it's divergent, but I don't know why printing a failure to prove sized was suppressed before, why recursive-self-normalization.rs does still not print this (which is direct recursion rather than a 2-cycle), or if the fact this is printing an error now is indicative of an underlying issue.

View changes since the review

This error might be related #155388? I personally wouldn't worry too much about emitting a new error here. Our error reporting logic is quite brittle with changes like this

lcnr · 2026-06-30T10:52:48Z

-                    }
-                    coercion.obligations.extend(ocx.into_pending_obligations());
-                    continue;
-                }


fun, this was dead code before 😁

View changes since the review

haha yeah, saw this when I was deleting it and I was like "uuuh wait, why... AliasRelate in the old solver?? huh??" :P

lcnr · 2026-06-30T10:53:44Z

-            | ty::PredicateKind::NormalizesTo(_) => VisibleForLeakCheck::No,
-            ty::PredicateKind::AliasRelate(_, _, _) => VisibleForLeakCheck::Yes,
-        };
+        let vis = VisibleForLeakCheck::No;


can inline that further 😁

View changes since the review

lcnr · 2026-07-01T08:13:26Z

+    /// However, in case the just-normalized `term` contains further aliases, we
+    /// emit nested `ProjectionPredicate` goals to normalize it further, which
+    /// are eagerly evaluated and may fail.


so i think the reason we get errors here is that term has not been just-normalized.

Because any alias we try to normalize when adding goals we should also have tried to normalize earlier. I would expect that by adding a normalize call to places where we currently don't normalize explicitly before calling this fn we'd stop panicking here.

However, I don't necessary want to expect that stuff is already normalized in this function, so not ICEing if the equate fails is still fine🤔

View changes since the review

right yeah, like, it's been "partially" normalized? e.g. <SomeStruct as std::ptr::Pointee>::Metadata --> <TailOfThatStruct as std::ptr::Pointee>::Metadata is a partial step in the progress of the entire expression being normalized (it's been "just-normalized" in that sense), but, TailOfThatStruct here has not been further normalized

we normalized the original alias, but fetching an assoc type from an impl/or a struct field gives you a separate unnormalized type

ok, pushed a rephrasing! I thiiink I've addressed all feedback now.

lcnr · 2026-07-02T08:59:47Z

small feedback wrt to tests, I'd prefer to go

//@ flags n stuff

// comments

source

in UI tests because that makes it easier for me to notice flags where putting them in the middle always takes me a second to figure out

after that r=me

@bors delegate+

rust-bors · 2026-07-02T08:59:50Z

✌️ @khyperia, you can now approve this pull request!

If @lcnr told you to "r=me" after making some further change, then please make that change and post @bors r=lcnr.

View changes since this delegation.

khyperia · 2026-07-02T12:59:12Z

@bors r=lcnr

rust-bors · 2026-07-02T12:59:16Z

📌 Commit c77583e has been approved by lcnr

It is now in the queue for this repository.

🌲 The tree is currently closed for pull requests below priority 1000. This pull request will be tested once the tree is reopened.

@lcnr

use ProjectionPredicate instead of AliasRelate part of rust-lang#155345 this PR is split into three commits: - refactoring `super_combine_{tys,consts}`, should not change behavior at all, just reorganizing code - convert to use ProjectionPredicate instead of AliasRelate in super_combine_tys and generalization (this is the important change for this PR) - deleting the now-dead AliasRelate predicate r? @lcnr

…uwer Rollup of 22 pull requests Successful merges: - #155429 (Support `u128`/`i128` c-variadic arguments) - #158100 (Emit retags in codegen to support BorrowSanitizer (part 4)) - #158494 (Improve E0277 diagnostics for conditionally implemented traits) - #158606 (use ProjectionPredicate instead of AliasRelate) - #158627 (Simplify option-iterator flattening in the compiler) - #158658 (Update LLVM submodule) - #158665 (Revert "Remove redundant dyn-compatibility check.") - #158021 (Remove old MinGW workaround) - #158473 (Add `riscv32imfc-unknown-none-elf` bare-metal target) - #158549 (process::exec: using appropriate exit code on vxworks.) - #158585 (Improve diagnostic for too many super keywords) - #158637 (hir_ty_lowering: avoid self type lookup for inherent aliases) - #158651 (ptr doc: reduce use of unsafe block to where needed) - #158669 (Remove `src/tools/test-float-parse/Cargo.lock`) - #158674 (library: Polish transmute's `split_at_stdlib` example) - #158677 (Add extra splat tests) - #158680 (Avoid ICE for `NonZero<char>` in improper_ctypes) - #158681 (Remove unnecessary `Hash` derives from MIR types) - #158682 (Avoid delayed bug for disabled on_type_error arguments) - #158684 (Add missing generic test coverage for ```#[splat]```) - #158687 (Streamline `MacEager`) - #158688 (Cleanup attribute docs and add links to other mentioned attributes)

khyperia added 2 commits June 30, 2026 09:16

refactor super_combine matches (no behavior change)

248a48b

use ProjectionPredicate instead of AliasRelate

1fdd4de

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. WG-trait-system-refactor The Rustc Trait System Refactor Initiative (-Znext-solver) labels Jun 30, 2026

rustbot assigned lcnr Jun 30, 2026

khyperia commented Jun 30, 2026

View reviewed changes

This comment has been minimized.

Sign in to view

lcnr reviewed Jun 30, 2026

View reviewed changes

Comment thread compiler/rustc_next_trait_solver/src/solve/normalizes_to.rs

This comment has been minimized.

Sign in to view

yeet AliasRelate

e9df4e0

khyperia force-pushed the yeet-aliasrelate branch from d0b358b to 47e4f26 Compare June 30, 2026 14:52

BoxyUwU self-assigned this Jun 30, 2026

lcnr reviewed Jul 1, 2026

View reviewed changes

khyperia force-pushed the yeet-aliasrelate branch from 47e4f26 to 551a033 Compare July 1, 2026 09:09

This comment has been minimized.

Sign in to view

khyperia force-pushed the yeet-aliasrelate branch 2 times, most recently from 203f3b2 to 9b818d1 Compare July 2, 2026 05:10

khyperia added 3 commits July 2, 2026 12:30

the .eq() in instantiate_normalizes_to_term may fail

9ccafdc

VisibleForLeakCheck::No in evaluate_root_goal_for_proof_tree

1f5a905

instantiate_normalizes_to_as_rigid

c77583e

khyperia force-pushed the yeet-aliasrelate branch from 9b818d1 to c77583e Compare July 2, 2026 10:32

rust-bors Bot added the S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. label Jul 2, 2026

rust-bors Bot removed the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Jul 2, 2026

JonathanBrouwer mentioned this pull request Jul 2, 2026

Rollup of 22 pull requests #158700

Open

Uh oh!

Uh oh!

Conversation

khyperia commented Jun 30, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rustbot commented Jun 30, 2026

Uh oh!

khyperia Jun 30, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lcnr Jun 30, 2026

Choose a reason for hiding this comment

Uh oh!

This comment has been minimized.

lcnr Jun 30, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

khyperia Jun 30, 2026

Choose a reason for hiding this comment

Uh oh!

lcnr Jun 30, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

This comment has been minimized.

lcnr Jul 1, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

khyperia Jul 1, 2026

Choose a reason for hiding this comment

Uh oh!

lcnr Jul 1, 2026

Choose a reason for hiding this comment

Uh oh!

khyperia Jul 1, 2026

Choose a reason for hiding this comment

Uh oh!

This comment has been minimized.

lcnr commented Jul 2, 2026

Uh oh!

rust-bors Bot commented Jul 2, 2026

Uh oh!

khyperia commented Jul 2, 2026

Uh oh!

rust-bors Bot commented Jul 2, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

khyperia commented Jun 30, 2026 •

edited by rustbot

Loading

khyperia Jun 30, 2026 •

edited by rustbot

Loading

lcnr Jun 30, 2026 •

edited by rustbot

Loading

lcnr Jun 30, 2026 •

edited by rustbot

Loading

lcnr Jul 1, 2026 •

edited by rustbot

Loading