Bubble up opaque <eq> opaque operations instead of picking an order

[oli-obk]

In case we are in Bubble mode (meaning every opaque type that is defined in the current crate is treated as if it were in its defining scope), we don't try to register an opaque type as the hidden type of another opaque type, but instead bubble up an obligation to equate them at the query caller site. Usually that means we have a DefiningAnchor::Bind and thus can reliably figure out whether an opaque type is in its defining scope. Where we can't, we'll error out, so the default is sound.

With this change we start using AliasTyEq predicates in the old solver, too.

fixes #108498

But also regresses tests/ui/impl-trait/anon_scope_creep.rs. Our use of Bubble for check_opaque_type_well_formed is going to keep biting us.

r? @lcnr @compiler-errors

[rustbot]

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @lcnr (or someone else) soon.

Please see the contribution instructions for more information. Namely, in order to ensure the minimum review times lag, PR authors and assigned reviewers should ensure that the review label (S-waiting-on-review and S-waiting-on-author) stays updated, invoking these commands when appropriate:

• @rustbot author: the review is finished, PR author should check the comments and take action accordingly
• @rustbot review: the author is ready for a review, this PR will be queued again in the reviewer's queue

[compiler-errors]

@bors try

[bors]

⌛ Trying commit d64fc80 with merge cde356bd79f103d9897ada6234862d1d3e511791...

[bors]

☀️ Try build successful - checks-actions
Build commit: cde356bd79f103d9897ada6234862d1d3e511791 (cde356bd79f103d9897ada6234862d1d3e511791)

[oli-obk]

@craterbot check

[craterbot]

👌 Experiment pr-114586 created and queued.
🤖 Automatically detected try build cde356bd79f103d9897ada6234862d1d3e511791
🔍 You can check out the queue and this experiment's details.

ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

[compiler-errors]

Given that this works in the new solver already:

Suggested change

	// check-pass
	// revisions: current next
	//[next] compile-flags: -Ztrait-solver=next
	// check-pass

[craterbot]

🚧 Experiment pr-114586 is now running

ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

[craterbot]

🎉 Experiment pr-114586 is completed!
📊 37 regressed and 3 fixed (343590 total)
📰 Open the full report.

⚠️ If you notice any spurious failure please add them to the blacklist!
ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

[oli-obk]

All of the regressions are spurious (disk failures or panics in the metadata encoder, probably also due to disk failures).

[lcnr]

r=me unless @compiler-errors also wants to take another look

[oli-obk]

@bors r=lcnr,compiler-errors

[bors]

📌 Commit 930affa has been approved by lcnr,compiler-errors

It is now in the queue for this repository.

[bors]

⌛ Testing commit 930affa with merge e2b3676...

[bors]

☀️ Test successful - checks-actions
Approved by: lcnr,compiler-errors
Pushing e2b3676 to master...

[rust-timer]

Finished benchmarking commit (e2b3676): comparison URL.

Overall result: ✅ improvements - no action needed

@rustbot label: -perf-regression

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.4%	[-0.4%, -0.4%]	1
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results

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)	4.8%	[4.2%, 5.4%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Cycles

Results

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)	1.3%	[1.3%, 1.3%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Binary size

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

Bootstrap: 630.454s -> 633.716s (0.52%)
Artifact size: 317.84 MiB -> 317.88 MiB (0.02%)

[aliemjay]

This does not bubble up the AliasRelate obligation to the query caller. Instead, the AliasRelate obligation does not make progress, forcing the query to be ambiguous and unfortunately we swallow ambiguities silently in MIR typeck instead of ICEing as expected.

Here is an example that regressed under the PR:

#![feature(type_alias_impl_trait)]
type Opaque = impl Sized; //~ ERROR unconstrained opaque type
fn get_rpit() -> impl Sized {}
fn query(_: impl FnOnce() -> Opaque) {}
fn test(_: Opaque) {
    query(get_rpit); // although it is constrained here
}