always rerun if we normalize local opaques by adwinwhite · Pull Request #161795 · rust-lang/rust

adwinwhite · 2026-08-26T02:15:24Z

The problem is that we rerun goals too late such that we compute auto trait for wrongly-marked-as-rigid opaques.
We call type_of on those opaques which would trigger query cycle if we're already in typeck/borrowck of the same opaques.

It's fixed by eagerly rerun if we ever normalize a local opaque.
To be more consersative, we could also rerun if we normalize remote opaque. But that causes regression for wg-grammar. Since we wouldn't be in typeck/borrowck if we're in post analysis mode, we don't have the query cycle problem. We might get away with not rerunning there.

OpaqueInStorage removal doesn't affect behavior and can be done in a followup.

r? @lcnr

adwinwhite · 2026-08-26T02:15:44Z

@bors try @rust-timer queue

always rerun if we normalize local opaques

rust-bors · 2026-08-26T03:14:31Z

☀️ Try build successful (CI)
Build commit: 82f99bb (82f99bb2d2238f8f809e1242ecfb758690bd0267)
Base parent: 0f33d09 (0f33d0912709e847199218ebab88e2311872f364)

rust-timer · 2026-08-26T03:57:10Z

Finished benchmarking commit (82f99bb): comparison URL.

Overall result: ❌✅ regressions and improvements - please read:

Benchmarking means the PR may be perf-sensitive. It's automatically marked not fit for rolling up. Overriding is possible but disadvised: it risks changing compiler perf.

Next, please: If you can, justify the regressions found in this try perf run in writing along with @rustbot label: +perf-regression-triaged. If not, fix the regressions and do another perf run. Neutral or positive results will clear the label automatically.

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

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	6.8%	[4.8%, 10.5%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.3%	[-0.4%, -0.2%]	7
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

This perf run didn't have relevant results for this metric.

Cycles

Results (secondary 5.6%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	5.6%	[3.8%, 8.4%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Binary size

Results (primary -0.0%, secondary -0.0%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.0%	[-0.0%, -0.0%]	4
Improvements ✅ (secondary)	-0.0%	[-0.0%, -0.0%]	4
All ❌✅ (primary)	-0.0%	[-0.0%, -0.0%]	4

Bootstrap: 475.39s -> 474.699s (-0.15%)
Artifact size: 402.69 MiB -> 402.62 MiB (-0.02%)

lcnr · 2026-08-27T07:46:50Z

+                    self.opaque_accesses.rerun_always(RerunReason::NormalizeOpaqueType)?;
                } else {
                    self.opaque_accesses
                        .rerun_if_in_post_analysis(RerunReason::NormalizeOpaqueTypeRemoteCrate)?;


that's also wrong, is it not. We also need to just rerun_always. We need to make sure we never mark an alias as rigid if it's not. We should add this invariant as a comment to ErasedNonCoherence

View changes since the review

Doing that basically undoes our perf improvement for wg-grammar. For remote opaques, we won't have query cycle from type_of. Uncertain about other potential problems caused by improperly marked rigid aliases.

okay, yeah 🤔 that's interesting. So instead:

always reveal auto traits for rigid aliases, and don't allow "incorrectly rigid local opaques"?

please comment this on ErasedNonCoherence and also IsRigid I think

adwinwhite · 2026-08-28T03:08:13Z

@bors try @rust-timer queue

always rerun if we normalize local opaques

rust-bors · 2026-08-28T04:07:07Z

☀️ Try build successful (CI)
Build commit: ca934c3 (ca934c31f1843211718c06bfac5663240c9b03dc)
Base parent: e457a7b (e457a7b0d326d67b4322ef0d11bd715cfaeda48f)

rust-timer · 2026-08-28T04:50:14Z

Finished benchmarking commit (ca934c3): comparison URL.

Overall result: ❌✅ regressions and improvements - please read:

Benchmarking means the PR may be perf-sensitive. It's automatically marked not fit for rolling up. Overriding is possible but disadvised: it risks changing compiler perf.

Next, please: If you can, justify the regressions found in this try perf run in writing along with @rustbot label: +perf-regression-triaged. If not, fix the regressions and do another perf run. Neutral or positive results will clear the label automatically.

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

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	151.2%	[4.8%, 310.3%]	6
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.1%	[-0.1%, -0.1%]	2
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (primary 2.9%, secondary 16.9%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	2.9%	[2.9%, 2.9%]	1
Regressions ❌ (secondary)	16.9%	[14.9%, 18.5%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.9%	[2.9%, 2.9%]	1

Cycles

Results (primary -2.2%, secondary 135.8%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	159.0%	[4.3%, 330.4%]	6
Improvements ✅ (primary)	-2.2%	[-2.4%, -2.0%]	2
Improvements ✅ (secondary)	-3.6%	[-3.6%, -3.6%]	1
All ❌✅ (primary)	-2.2%	[-2.4%, -2.0%]	2

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 476.343s -> 475.472s (-0.18%)
Artifact size: 402.74 MiB -> 402.85 MiB (0.03%)

lcnr

r=me after nits

View changes since this review

rustbot · 2026-09-04T02:09:28Z

This PR was rebased onto a different main commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.

adwinwhite · 2026-09-04T02:18:14Z

The actual perf run is the first one and the regression concentrates in a stress test which is very sensitive to changes on opaques.
We discussed it on zulip and think it's acceptable to fix a pass->fail bug.

@rustbot label: +perf-regression-triaged

@bors r=lcnr rollup=never

rust-bors · 2026-09-04T02:18:17Z

📌 Commit 8720f91 has been approved by lcnr

It is now in the queue for this repository.

🌲 The tree is currently closed for pull requests below priority 10. This pull request will be tested once the tree is reopened.

Reason for tree closure: broken external dependency

@lcnr

always rerun if we normalize local opaques Fixes #135062 The problem is that we rerun goals too late such that we compute auto trait for wrongly-marked-as-rigid opaques. We call `type_of` on those opaques which would trigger query cycle if we're already in typeck/borrowck of the same opaques. It's fixed by eagerly rerun if we ever normalize a local opaque. To be more consersative, we could also rerun if we normalize remote opaque. But that causes regression for `wg-grammar`. Since we wouldn't be in typeck/borrowck if we're in post analysis mode, we don't have the query cycle problem. We **might** get away with not rerunning there. `OpaqueInStorage` removal doesn't affect behavior and can be done in a followup. r? @lcnr

rust-log-analyzer · 2026-09-04T07:44:35Z

The job x86_64-msvc-1 failed! Check out the build log: (web) (plain enhanced) (plain)

Click to see the possible cause of the failure (guessed by this bot)

rust-bors · 2026-09-04T07:45:03Z

💔 Test for 9db98a2 failed: CI. Failed job:

auto - x86_64-msvc-1 (web logs, enhanced plaintext logs)

ShoyuVanilla · 2026-09-04T09:00:17Z

@bors retry

rust-bors · 2026-09-04T14:06:54Z

☀️ Test successful - CI
Approved by: lcnr
Duration: 3h 9m 58s
Pushing 0ed41eb to main...

github-actions · 2026-09-04T14:10:18Z

What is this?

This is an experimental post-merge analysis report that shows differences in test outcomes between the merged PR and its parent PR.

Comparing b924f94 (parent) -> 0ed41eb (this PR)

Test differences

Show 42 test diffs

Stage 1

[ui] tests/ui/traits/next-solver/normalize/always_rerun_normalizing_opaques.rs: [missing] -> pass (J0)
[ui (polonius)] tests/ui/traits/next-solver/normalize/always_rerun_normalizing_opaques.rs: [missing] -> pass (J1)

Stage 2

[ui] tests/ui/traits/next-solver/normalize/always_rerun_normalizing_opaques.rs: [missing] -> pass (J2)

Additionally, 39 doctest diffs were found. These are ignored, as they are noisy.

Job group index

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard 0ed41eb4142dda2df61eb1145a312c1a9d62eb56 --output-dir test-dashboard

And then open test-dashboard/index.html in your browser to see an overview of all executed tests.

Job duration changes

x86_64-gnu: 1h 49m -> 2h 42m (+47.6%)
x86_64-gnu-parallel-frontend: 1h 24m -> 2h 1m (+44.6%)
dist-android: 25m -> 31m 59s (+27.9%)
dist-riscv64-linux-gnu: 1h 36m -> 1h 10m (-26.8%)
dist-s390x-linux: 1h 14m -> 1h 33m (+25.4%)
x86_64-gnu-distcheck: 2h 26m -> 1h 49m (-25.4%)
x86_64-msvc-ext1: 2h 19m -> 1h 44m (-25.2%)
dist-sparcv9-solaris: 1h 20m -> 1h 40m (+25.0%)
x86_64-gnu-tools: 55m 37s -> 1h 9m (+24.5%)
pr-check-2: 42m 9s -> 32m 4s (-23.9%)

How to interpret the job duration changes?

Job durations can vary a lot, based on the actual runner instance
that executed the job, system noise, invalidated caches, etc. The table above is provided
mostly for t-infra members, for simpler debugging of potential CI slow-downs.

rust-timer · 2026-09-04T14:50:36Z

Finished benchmarking commit (0ed41eb): comparison URL.

Overall result: ❌✅ regressions and improvements - please read:

Our benchmarks found a performance regression caused by this PR.
This might be an actual regression, but it can also be just noise.

Next Steps:

If the regression was expected or you think it can be justified,
please write a comment with sufficient written justification, and add
@rustbot label: +perf-regression-triaged to it, to mark the regression as triaged.
If you think that you know of a way to resolve the regression, try to create
a new PR with a fix for the regression.
If you do not understand the regression or you think that it is just noise,
you can ask the @rust-lang/wg-compiler-performance working group for help (members of this group
were already notified of this PR).

@rustbot label: +perf-regression
cc @rust-lang/wg-compiler-performance

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	6.8%	[4.7%, 10.7%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.1%	[-0.1%, -0.1%]	1
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (primary 6.0%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	6.0%	[3.1%, 9.0%]	2
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	6.0%	[3.1%, 9.0%]	2

Cycles

Results (primary -2.9%, secondary 5.4%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	5.4%	[3.9%, 7.8%]	3
Improvements ✅ (primary)	-2.9%	[-2.9%, -2.9%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-2.9%	[-2.9%, -2.9%]	1

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 480.684s -> 476.718s (-0.83%)
Artifact size: 403.39 MiB -> 403.29 MiB (-0.02%)

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 Aug 26, 2026

rust-bors Bot pushed a commit that referenced this pull request Aug 26, 2026

Auto merge of #161795 - adwinwhite:hidden-is-opaque, r=<try>

82f99bb

always rerun if we normalize local opaques

This comment has been minimized.

Sign in to view

rustbot added perf-regression Performance regression. and removed S-waiting-on-perf Status: Waiting on a perf run to be completed. labels Aug 26, 2026

lcnr reviewed Aug 27, 2026

View reviewed changes

adwinwhite force-pushed the hidden-is-opaque branch from cdb50b8 to 009cc84 Compare August 28, 2026 03:07

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 Aug 28, 2026

rust-bors Bot pushed a commit that referenced this pull request Aug 28, 2026

Auto merge of #161795 - adwinwhite:hidden-is-opaque, r=<try>

ca934c3

always rerun if we normalize local opaques

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 Aug 28, 2026

adwinwhite force-pushed the hidden-is-opaque branch from 009cc84 to 64fea5e Compare August 31, 2026 03:18

rustbot assigned lcnr Aug 31, 2026

adwinwhite marked this pull request as ready for review August 31, 2026 03:36

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Aug 31, 2026

adwinwhite force-pushed the hidden-is-opaque branch from 64fea5e to 0a26490 Compare September 3, 2026 07:50

lcnr reviewed Sep 3, 2026

View reviewed changes

Comment thread compiler/rustc_type_ir/src/infer_ctxt.rs

Comment thread compiler/rustc_type_ir/src/ty_kind.rs Outdated

always rerun if we normalize local opaques

8720f91

adwinwhite force-pushed the hidden-is-opaque branch from 0a26490 to 8720f91 Compare September 4, 2026 02:09

rustbot added the perf-regression-triaged The performance regression has been triaged. label Sep 4, 2026

rust-bors Bot added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Sep 4, 2026

This comment has been minimized.

Sign in to view

rust-bors Bot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Sep 4, 2026

rust-bors Bot added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Sep 4, 2026

This comment has been minimized.

Sign in to view

rust-bors Bot added merged-by-bors This PR was explicitly merged by bors. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Sep 4, 2026

rust-bors Bot merged commit 0ed41eb into rust-lang:main Sep 4, 2026
14 checks passed

rustbot added this to the 1.100.0 milestone Sep 4, 2026

Uh oh!

Uh oh!

Conversation

adwinwhite commented Aug 26, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

adwinwhite commented Aug 26, 2026

Uh oh!

This comment has been minimized.

This comment has been minimized.

rust-bors Bot commented Aug 26, 2026

Uh oh!

This comment has been minimized.

rust-timer commented Aug 26, 2026

Overall result: ❌✅ regressions and improvements - please read:

Uh oh!

lcnr Aug 27, 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!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

adwinwhite commented Aug 28, 2026

Uh oh!

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

rust-bors Bot commented Aug 28, 2026

Uh oh!

This comment has been minimized.

rust-timer commented Aug 28, 2026

Overall result: ❌✅ regressions and improvements - please read:

Uh oh!

lcnr left a comment • 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!

Uh oh!

rustbot commented Sep 4, 2026

Uh oh!

adwinwhite commented Sep 4, 2026

Uh oh!

rust-bors Bot commented Sep 4, 2026

Uh oh!

This comment has been minimized.

rust-log-analyzer commented Sep 4, 2026

Uh oh!

rust-bors Bot commented Sep 4, 2026

Uh oh!

ShoyuVanilla commented Sep 4, 2026

Uh oh!

This comment has been minimized.

rust-bors Bot commented Sep 4, 2026

Uh oh!

Uh oh!

github-actions Bot commented Sep 4, 2026

Test differences

Stage 1

Stage 2

Job duration changes

Uh oh!

rust-timer commented Sep 4, 2026

Overall result: ❌✅ regressions and improvements - please read:

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

adwinwhite commented Aug 26, 2026 •

edited by rustbot

Loading

lcnr Aug 27, 2026 •

edited by rustbot

Loading

lcnr left a comment •

edited by rustbot

Loading