Instantiate predicate binder without recanonicalizing goal in new solver #136997

compiler-errors · 2025-02-14T00:27:23Z

This has the side-effect of making the leak check stronger.

r? lcnr

lcnr · 2025-02-14T13:28:24Z

tests/ui/higher-ranked/leak-check/leak-check-in-selection-6-ambig-unify.rs

 fn main() {
    impls_trait::<(), _, _>()
-    //~^ ERROR type annotations needed
+    //[old]~^ ERROR type annotations needed


this test should be changed to have one step of indirection between the binder and the LeakCheckFailure bound.

lcnr · 2025-02-14T13:30:01Z

hmm 👍 one concern is that this also impacts coherence, so it affects stable.

would need an FCP (and it seems generally useful to explain this to the other types team members, if only to fully reason about this ourselves). I would be fine writing the FCP comment, but would finish some other work first, so it'd take a while

bors · 2025-04-10T01:19:46Z

☔ The latest upstream changes (presumably #139595) made this pull request unmergeable. Please resolve the merge conflicts.

[DO NOT MERGE] bootstrap with `-Znext-solver=globally` A revival of rust-lang#124812. Current status: ~~`./x.py b --stage 2` passes 🎉~~ `try` builds succeed 🎉 🎉 🎉 [first perf run](rust-lang#133502 (comment)) 👻 ### crater This does not detect hangs or memory issues. | date | #crates | #regressions | | ---- | ------- | ------------ | | 2025.04.11 | 100 | 2 | | 2025.04.11 | 1000 | 27 | | 2025.04.17 | 10000 | 456 | | 2025.04.18 | 10000 | 437 | | 2025.04.24 | 10000 | 164 | | 2025.04.26 | 10000 | 108 | | 2025.04.28 | 10000 | 91 | | 2025.05.01 | 10000 | 145 woops | | 2025.05.03 | 624228[^1] | 1585 | | 2025.05.05 | 8964[^2] | 931 | [^1]: a complete crater run [^2]: only testing crates which may have regressed from the above run ### in-flight changes - rust-lang#140561 - rust-lang#140672 - rust-lang#140678 - rust-lang#136997 - rust-lang#139587 - rust-lang#140497 - rust-lang#124852, unsure whether I actually want to land this PR for now - https://github.com/lcnr/rust/tree/opaque-type-method-call - rust-lang#140260 - rust-lang#140375 - rust-lang#140405 - rust-lang#140496 - double recursion limit in the new solver r? `@ghost`

[DO NOT MERGE] bootstrap with `-Znext-solver=globally` A revival of rust-lang#124812. Current status: ~~`./x.py b --stage 2` passes 🎉~~ `try` builds succeed 🎉 🎉 🎉 [first perf run](rust-lang#133502 (comment)) 👻 ### crater This does not detect hangs or memory issues. | date | #crates | #regressions | | ---- | ------- | ------------ | | 2025.04.11 | 100 | 2 | | 2025.04.11 | 1000 | 27 | | 2025.04.17 | 10000 | 456 | | 2025.04.18 | 10000 | 437 | | 2025.04.24 | 10000 | 164 | | 2025.04.26 | 10000 | 108 | | 2025.04.28 | 10000 | 91 | | 2025.05.01 | 10000 | 145 woops | | 2025.05.03 | 624228[^1] | 1585 | | 2025.05.05 | 8964[^2] | 931 | | 2025.05.06 | 4401[^2] | 726 | [^1]: a complete crater run [^2]: only testing crates which may have regressed from the above run ### in-flight changes - rust-lang#140561 - rust-lang#140672 - rust-lang#140678 - rust-lang#136997 - rust-lang#139587 - rust-lang#140497 - rust-lang#124852, unsure whether I actually want to land this PR for now - https://github.com/lcnr/rust/tree/opaque-type-method-call - rust-lang#140260 - rust-lang#140375 - rust-lang#140405 - rust-lang#140496 - double recursion limit in the new solver r? `@ghost`

[DO NOT MERGE] bootstrap with `-Znext-solver=globally` A revival of rust-lang#124812. Current status: ~~`./x.py b --stage 2` passes 🎉~~ `try` builds succeed 🎉 🎉 🎉 [first perf run](rust-lang#133502 (comment)) 👻 ### crater This does not detect hangs or memory issues. | date | #crates | #regressions | | ---- | ------- | ------------ | | 2025.04.11 | 100 | 2 | | 2025.04.11 | 1000 | 27 | | 2025.04.17 | 10000 | 456 | | 2025.04.18 | 10000 | 437 | | 2025.04.24 | 10000 | 164 | | 2025.04.26 | 10000 | 108 | | 2025.04.28 | 10000 | 91 | | 2025.05.01 | 10000 | 145 woops | | 2025.05.03 | 624228[^1] | 1585 | | 2025.05.05 | 8964[^2] | 931 | | 2025.05.06 | 4401[^2] | 726 | [^1]: a complete crater run [^2]: only testing crates which may have regressed from the above run ### in-flight changes - rust-lang#140711 - rust-lang#140713 - rust-lang#140712 - rust-lang#136997 - rust-lang#139587 - rust-lang#140497 - rust-lang#124852, unsure whether I actually want to land this PR for now - https://github.com/lcnr/rust/tree/opaque-type-method-call - rust-lang#140260 - rust-lang#140375 - rust-lang#140405 - look into blanket impls for opaque type infer vars as well - rust-lang#140496 - double recursion limit in the new solver r? `@ghost`

[DO NOT MERGE] bootstrap with `-Znext-solver=globally` A revival of rust-lang#124812. Current status: ~~`./x.py b --stage 2` passes 🎉~~ `try` builds succeed 🎉 🎉 🎉 [first perf run](rust-lang#133502 (comment)) 👻 ### crater This does not detect hangs or memory issues. | date | #crates | #regressions | | ---- | ------- | ------------ | | 2025.04.11 | 100 | 2 | | 2025.04.11 | 1000 | 27 | | 2025.04.17 | 10000 | 456 | | 2025.04.18 | 10000 | 437 | | 2025.04.24 | 10000 | 164 | | 2025.04.26 | 10000 | 108 | | 2025.04.28 | 10000 | 91 | | 2025.05.01 | 10000 | 145 woops | | 2025.05.03 | 624228[^1] | 1585 | | 2025.05.05 | 8964[^2] | 931 | | 2025.05.06 | 4401[^2] | 726 | | 2025.05.07 | 2704[^2] | 668 | [^1]: a complete crater run [^2]: only testing crates which may have regressed from the above run ### in-flight changes - rust-lang#140712 - rust-lang#136997 - rust-lang#139587 - rust-lang#140497 - rust-lang#124852, unsure whether I actually want to land this PR for now - https://github.com/lcnr/rust/tree/opaque-type-method-call - rust-lang#140375 - rust-lang#140405 - look into blanket impls for opaque type infer vars as well - rust-lang#140496 - double recursion limit in the new solver r? `@ghost`

compiler-errors · 2025-06-02T14:42:40Z

I wonder how much this affects perf...

@bors try @rust-timer queue

Instantiate predicate binder without recanonicalizing goal in new solver This has the side-effect of making the leak check stronger. r? lcnr

bors · 2025-06-02T14:43:54Z

⌛ Trying commit a80049f with merge 2f79182...

bors · 2025-06-02T17:08:29Z

☀️ Try build successful - checks-actions
Build commit: 2f79182 (2f79182e4fecd8940876615a520ef94f9ae67a5b)

rust-timer · 2025-06-02T18:47:03Z

Finished benchmarking commit (2f79182): comparison URL.

Overall result: ✅ improvements - no action needed

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. While you can manually mark this PR as fit for rollup, we strongly recommend not doing so since this PR may lead to changes in compiler perf.

@bors rollup=never
@rustbot label: -S-waiting-on-perf -perf-regression

Instruction count

This is the most reliable metric that we have; it was used to determine the overall result at the top of this comment. However, even this metric can sometimes exhibit noise.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.7%	[-0.8%, -0.5%]	4
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (secondary 0.8%)

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	1.7%	[0.4%, 9.6%]	16
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.0%	[-5.6%, -0.6%]	5
All ❌✅ (primary)	-	-	0

Cycles

Results (secondary -0.4%)

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	0.9%	[0.9%, 0.9%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.7%	[-1.0%, -0.4%]	9
All ❌✅ (primary)	-	-	0

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 743.462s -> 741.168s (-0.31%)
Artifact size: 372.27 MiB -> 372.27 MiB (0.00%)

lcnr · 2025-06-03T11:52:38Z

okay, looks like my expectation that recanonicalizing after instantiating the binder being better for caching does not actually hold 😁

so there's pretty much no reason to not land this change :3

lcnr · 2025-09-18T14:30:15Z

closing in favor of #146725 :3

`-Znext-solver` instantiate predicate binder without recanonicalizing goal This strengthens the leak check to match the old trait solver. The new trait solver now also instantiates higher ranked goals in the same scope as candidate selection, so the leak check in each candidate detects placeholder errors involving this higher ranked goal. E.g. let's look at tests/ui/higher-ranked/leak-check/leak-check-in-selection-2.rs ```rust trait Trait<T, U> {} impl<'a> Trait<&'a str, &'a str> for () {} impl<'a> Trait<&'a str, String> for () {} fn impls_trait<T: for<'a> Trait<&'a str, U>, U>() {} fn main() { impls_trait::<(), _>(); } ``` Here proving `(): for<'a> Trait<&'a str, ?u>` via `impl<'a> Trait<&'a str, &'a str> for ()` equates `?u` with `&'!a str` which results in a leak check error as `?u` cannot name `'a`. If this leak check error happens while considering candidates we drop the first impl and infer `?u` to `String`. If not, this remains ambiguous. This behavior is a bit iffy, see the FCP proposal in #119820 for more details on why this current behavior is somewhat undesirable. However, considering placeholders from higher-ranked goals for candidate selection does allow more code to compile and a lot of the code *feels like it should compile*. **This caused us to revert the change of #119820 in #127568.** I originally expected that we can avoid breakage with the new solver differently here, e.g. by considering OR-region constraints. However, doing so is a significant change and I don't have a great idea for how that should work. Matching the old solver behavior for now should not make this cleaner approach any more difficult in the future, so let's just go with what actually allows us to stabilize the new solver for now. This PR changing the new solver to match the behavior of the old one wrt the leak check. As the new solver is already used by default in coherence, this allows more code to compile, see `tests/ui/higher-ranked/leak-check/leak-check-in-selection-7-coherence.rs`: ```rust struct W<T, U>(T, U); trait Trait<T> {} // using this impl results in a higher-ranked region error. impl<'a> Trait<W<&'a str, &'a str>> for () {} impl<'a> Trait<W<&'a str, String>> for () {} trait NotString {} impl NotString for &str {} impl NotString for u32 {} trait Overlap {} impl<T: for<'a> Trait<W<&'a str, U>>, U> Overlap for T {} impl<U: NotString> Overlap for () {} fn main() {} ``` This behavior is quite arbitrary and not something I expect users to rely on in practice, however, it should still go through an FCP imo. r? `@BoxyUwU` originally implemented by `@compiler-errors` in #136997. Closes rust-lang/trait-system-refactor-initiative#120.

`-Znext-solver` instantiate predicate binder without recanonicalizing goal This strengthens the leak check to match the old trait solver. The new trait solver now also instantiates higher ranked goals in the same scope as candidate selection, so the leak check in each candidate detects placeholder errors involving this higher ranked goal. E.g. let's look at tests/ui/higher-ranked/leak-check/leak-check-in-selection-2.rs ```rust trait Trait<T, U> {} impl<'a> Trait<&'a str, &'a str> for () {} impl<'a> Trait<&'a str, String> for () {} fn impls_trait<T: for<'a> Trait<&'a str, U>, U>() {} fn main() { impls_trait::<(), _>(); } ``` Here proving `(): for<'a> Trait<&'a str, ?u>` via `impl<'a> Trait<&'a str, &'a str> for ()` equates `?u` with `&'!a str` which results in a leak check error as `?u` cannot name `'a`. If this leak check error happens while considering candidates we drop the first impl and infer `?u` to `String`. If not, this remains ambiguous. This behavior is a bit iffy, see the FCP proposal in rust-lang/rust#119820 for more details on why this current behavior is somewhat undesirable. However, considering placeholders from higher-ranked goals for candidate selection does allow more code to compile and a lot of the code *feels like it should compile*. **This caused us to revert the change of rust-lang/rust#119820 in rust-lang/rust#127568.** I originally expected that we can avoid breakage with the new solver differently here, e.g. by considering OR-region constraints. However, doing so is a significant change and I don't have a great idea for how that should work. Matching the old solver behavior for now should not make this cleaner approach any more difficult in the future, so let's just go with what actually allows us to stabilize the new solver for now. This PR changing the new solver to match the behavior of the old one wrt the leak check. As the new solver is already used by default in coherence, this allows more code to compile, see `tests/ui/higher-ranked/leak-check/leak-check-in-selection-7-coherence.rs`: ```rust struct W<T, U>(T, U); trait Trait<T> {} // using this impl results in a higher-ranked region error. impl<'a> Trait<W<&'a str, &'a str>> for () {} impl<'a> Trait<W<&'a str, String>> for () {} trait NotString {} impl NotString for &str {} impl NotString for u32 {} trait Overlap {} impl<T: for<'a> Trait<W<&'a str, U>>, U> Overlap for T {} impl<U: NotString> Overlap for () {} fn main() {} ``` This behavior is quite arbitrary and not something I expect users to rely on in practice, however, it should still go through an FCP imo. r? `@BoxyUwU` originally implemented by `@compiler-errors` in rust-lang/rust#136997. Closes rust-lang/trait-system-refactor-initiative#120.

rustbot assigned lcnr Feb 14, 2025

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 Feb 14, 2025

lcnr reviewed Feb 14, 2025

View reviewed changes

Instantiate predicate binder without recanonicalizing goal in new solver

a80049f

lcnr force-pushed the eager-binder branch from 16a95b5 to a80049f Compare May 5, 2025 13:41

lcnr mentioned this pull request May 5, 2025

[DO NOT MERGE] bootstrap with -Znext-solver=globally #133502

Draft

This comment has been minimized.

Sign in to view

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Jun 2, 2025

bors added a commit that referenced this pull request Jun 2, 2025

Auto merge of #136997 - compiler-errors:eager-binder, r=<try>

2f79182

Instantiate predicate binder without recanonicalizing goal in new solver This has the side-effect of making the leak check stronger. r? lcnr

This comment has been minimized.

Sign in to view

rustbot removed the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Jun 2, 2025

lcnr added the needs-fcp This change is insta-stable, or significant enough to need a team FCP to proceed. label Jun 3, 2025

lcnr mentioned this pull request Sep 18, 2025

-Znext-solver instantiate predicate binder without recanonicalizing goal #146725

Merged

compiler-errors closed this Sep 18, 2025

rustbot removed the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Sep 18, 2025

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

Instantiate predicate binder without recanonicalizing goal in new solver #136997

Instantiate predicate binder without recanonicalizing goal in new solver #136997

Uh oh!

compiler-errors commented Feb 14, 2025

Uh oh!

lcnr Feb 14, 2025

Uh oh!

lcnr commented Feb 14, 2025

Uh oh!

bors commented Apr 10, 2025

Uh oh!

compiler-errors commented Jun 2, 2025

Uh oh!

This comment has been minimized.

bors commented Jun 2, 2025

Uh oh!

bors commented Jun 2, 2025

Uh oh!

This comment has been minimized.

rust-timer commented Jun 2, 2025

Uh oh!

lcnr commented Jun 3, 2025

Uh oh!

lcnr commented Sep 18, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Uh oh!

Instantiate predicate binder without recanonicalizing goal in new solver #136997

Instantiate predicate binder without recanonicalizing goal in new solver #136997

Uh oh!

Conversation

compiler-errors commented Feb 14, 2025

Uh oh!

lcnr Feb 14, 2025

Choose a reason for hiding this comment

Uh oh!

lcnr commented Feb 14, 2025

Uh oh!

bors commented Apr 10, 2025

Uh oh!

compiler-errors commented Jun 2, 2025

Uh oh!

This comment has been minimized.

bors commented Jun 2, 2025

Uh oh!

bors commented Jun 2, 2025

Uh oh!

This comment has been minimized.

rust-timer commented Jun 2, 2025

Overall result: ✅ improvements - no action needed

Instruction count

Max RSS (memory usage)

Cycles

Binary size

Uh oh!

lcnr commented Jun 3, 2025

Uh oh!

lcnr commented Sep 18, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants