fix: Check the fallback map before queueing child in `visible_parent_map` breadth-first search - #160464

zalanlevai · 2026-08-03T20:11:04Z

The fix in #159881 extended the visible_parent_map breadth-first search (BFS) to include items nested within "fallback" items, such as #[doc(hidden)] modules and re-exports. However, the fix resulted in fallback entries not being checked before enqueueing new items into the BFS, resulting in fallback items being explored in the BFS multiple times, once for each fallback parent, up until a non-fallback visible parent is found instead, if any.

This PR adds a check before new fallback items are recorded and the BFS queue is extended that ensures that the item has not already been searched through a fallback parent item.

Fixes the libc performance regression described in #160439. (Fixes #160439.)

Unfortunately, I could not think of a way to write a test case that covers the regression that this PR fixes, as it would require a massive crate with a very large number of "fallback" parents, and the test would be performance-based.

rustbot · 2026-08-03T20:11:09Z

r? @nnethercote

rustbot has assigned @nnethercote.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

Owners of files modified in this PR: compiler
compiler expanded to 75 candidates
Random selection from 19 candidates

zalanlevai · 2026-08-03T20:34:27Z

@rustbot label +perf-regression

zalanlevai · 2026-08-03T20:39:14Z

Actually, after looking at PRs labeled perf-regression, I think I used the label incorrectly as this PR fixes a performance regression, rather than introducing one.

@rustbot label -perf-regression

tgross35 · 2026-08-03T21:04:05Z

That makes me think, maybe we could have a perf job that runs rustdoc on libc or any other larger crate? That would have caught this.

Cc @Kobzol

zalanlevai · 2026-08-03T21:31:35Z

Huh, those CI failures on x86_64-gnu-gcc seem strange to me. All seven test failures are caused by the following unexpected warning on stderr:

libgccjit.so: warning: : ‘retain’ attribute ignored [-Wattributes]

Is my reading correct that the CI failure seems unrelated to this change and could be caused by something else?

tgross35 · 2026-08-03T21:35:13Z

Looks like that job may be having some issues #t-infra > Tree ops @ 💬

Kobzol · 2026-08-04T07:51:17Z

That makes me think, maybe we could have a perf job that runs rustdoc on libc or any other larger crate? That would have caught this.

Cc @Kobzol

We do actually run rustdoc on libc on every merge :) https://perf.rust-lang.org/compare.html?debug=false&check=false&opt=false&nonRelevant=true

tgross35 · 2026-08-04T09:28:40Z

Oh I was wondering why it didn't show up in https://perf.rust-lang.org/compare.html?start=b04d3c8c22822b7f10dab5a563b4e2afff1861c1&end=28c66af27578a28bc8cad1c5957be4b1e2e7fb8b&stat=instructions%3Au&nonRelevant=true&opt=false&debug=false&check=false&doc-json=false. But testing again locally, it looks like libc is actually fine but ctest in the same repo was hitting the crash.

@zalanlevai if you have a build locally, cargo doc -p ctest --no-deps in https://github.com/rust-lang/libc is how I'm able to reproduce the issue

zalanlevai · 2026-08-04T09:45:53Z

Thank you @tgross35 for the reproducer! I can confirm that running cargo doc -p ctest --no-deps on rust-lang/libc locally runs perfectly fine with this fix applied (and that I can reproduce the hang locally if I remove my changes).

zalanlevai · 2026-08-04T09:52:51Z

Should I force CI checks to re-run on this PR, now that the x86_64-gnu-gcc failures have been resolved?

Kobzol · 2026-08-04T09:59:38Z

Yeah, you can just repush the latest commit.

rustbot · 2026-08-04T10:04:30Z

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.

nnethercote · 2026-08-05T10:05:22Z

I'm going to pass this on to the reviewer of #159881:

r? @fee1-dead

oli-obk

r=me with the sorting avoided

View changes since this review

rustbot · 2026-08-05T14:51:14Z

Reminder, once the PR becomes ready for a review, use @rustbot ready.

…map` BFS

zalanlevai · 2026-08-05T16:42:26Z

I have made the change we discussed. I also found a way to test the behaviour through path printing in diagnostics. Just like with regular entries, we expect to store the topmost fallback entry in the BFS to produce the shortest visible paths. This condition was failed for fallback items before this change. The added regression test checks that we do not override fallback entries with later ones by ensuring that we use the shortest available fallback path.

@rustbot ready

oli-obk

@bors r+

View changes since this review

rust-bors · 2026-08-06T05:07:08Z

📌 Commit 640032d has been approved by oli-obk

It is now in the queue for this repository.

…fallback-entry-check, r=oli-obk fix: Check the fallback map before queueing child in `visible_parent_map` breadth-first search The fix in rust-lang#159881 extended the `visible_parent_map` breadth-first search (BFS) to include items nested within "fallback" items, such as `#[doc(hidden)]` modules and re-exports. However, the fix resulted in fallback entries not being checked before enqueueing new items into the BFS, resulting in fallback items being explored in the BFS multiple times, once for each fallback parent, up until a non-fallback visible parent is found instead, if any. This PR adds a check before new fallback items are recorded and the BFS queue is extended that ensures that the item has not already been searched through a fallback parent item. Fixes the `libc` performance regression described in rust-lang#160439. (Fixes rust-lang#160439.) Unfortunately, I could not think of a way to write a test case that covers the regression that this PR fixes, as it would require a massive crate with a very large number of "fallback" parents, and the test would be performance-based.

…uwer Rollup of 17 pull requests Successful merges: - #159530 (Cap socket send length to c_int::MAX on Apple targets) - #159506 (Allow associated const equality constraints with GCA) - #160006 (Account for desugaring in method call move errors) - #160415 (Split `aarch64-apple{,-macos-26}` => `aarch64-apple{,-macos-26}-{1,2}` jobs) - #160464 (fix: Check the fallback map before queueing child in `visible_parent_map` breadth-first search) - #160555 (Split `apply_primary_terminator_effect`) - #160592 (Suggest if-let chain continuation on unclosed delimiter) - #160600 (Avoid the std DLL copy alongside rustc) - #160156 (check_consts: exhaustively match on CastKind) - #160211 (Rename `#[unroll]` => `#[rustc_unroll]` to mitigate nameres ambiguity) - #160304 (Add tests for fixed new solver issues) - #160398 (rustc_abi: Add `LayoutData::is_variant_uninhabited` method) - #160546 (Update error message in documentation comments) - #160568 (Use `VisitorResult` helper macros) - #160571 (Add regression test for array type recovery in generic arguments) - #160588 (add a test showing polonius alpha is not a subset of datalog polonius) - #160617 (Add a suggestion to MissingUnsafeOnExtern diagnostic)

Rollup merge of #160464 - zalanlevai:fix-visible-parent-map-fallback-entry-check, r=oli-obk fix: Check the fallback map before queueing child in `visible_parent_map` breadth-first search The fix in #159881 extended the `visible_parent_map` breadth-first search (BFS) to include items nested within "fallback" items, such as `#[doc(hidden)]` modules and re-exports. However, the fix resulted in fallback entries not being checked before enqueueing new items into the BFS, resulting in fallback items being explored in the BFS multiple times, once for each fallback parent, up until a non-fallback visible parent is found instead, if any. This PR adds a check before new fallback items are recorded and the BFS queue is extended that ensures that the item has not already been searched through a fallback parent item. Fixes the `libc` performance regression described in #160439. (Fixes #160439.) Unfortunately, I could not think of a way to write a test case that covers the regression that this PR fixes, as it would require a massive crate with a very large number of "fallback" parents, and the test would be performance-based.

JonathanBrouwer · 2026-08-06T20:10:16Z

@rust-timer build e577f4b

rust-timer · 2026-08-07T09:02:16Z

Finished benchmarking commit (e577f4b): 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)	0.4%	[0.2%, 0.5%]	7
Improvements ✅ (primary)	-0.4%	[-0.4%, -0.4%]	1
Improvements ✅ (secondary)	-0.4%	[-0.5%, -0.2%]	7
All ❌✅ (primary)	-0.4%	[-0.4%, -0.4%]	1

Max RSS (memory usage)

Results (primary -6.6%)

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)	-	-	0
Improvements ✅ (primary)	-6.6%	[-6.6%, -6.6%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-6.6%	[-6.6%, -6.6%]	1

Cycles

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

Binary size

Results (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
Improvements ✅ (secondary)	-0.0%	[-0.0%, -0.0%]	1
All ❌✅ (primary)	-	-	0

Bootstrap: 458.35s -> 458.321s (-0.01%)
Artifact size: 398.64 MiB -> 398.62 MiB (-0.01%)

rustbot assigned nnethercote Aug 3, 2026

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. labels Aug 3, 2026

zalanlevai mentioned this pull request Aug 3, 2026

cargo doc freezes on libc with nightly-2026-08-03 #160439

Closed

zalanlevai changed the title ~~fix: Check the fallback map before queueing child in visible_parent_map breadth-first-search~~ fix: Check the fallback map before queueing child in visible_parent_map breadth-first search Aug 3, 2026

rustbot added the perf-regression Performance regression. label Aug 3, 2026

rustbot removed the perf-regression Performance regression. label Aug 3, 2026

This comment has been minimized.

Sign in to view

lqd mentioned this pull request Aug 4, 2026

dedup fallback #160487

Closed

zalanlevai force-pushed the fix-visible-parent-map-fallback-entry-check branch from 464edf2 to 45cd627 Compare August 4, 2026 10:04

rustbot assigned fee1-dead and unassigned nnethercote Aug 5, 2026

oli-obk requested changes Aug 5, 2026

View reviewed changes

Comment thread compiler/rustc_metadata/src/rmeta/decoder/cstore_impl.rs Outdated

rustbot added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Aug 5, 2026

oli-obk self-assigned this Aug 5, 2026

oli-obk unassigned fee1-dead Aug 5, 2026

fix: Check the fallback map before queueing child in `visible_parent_…

640032d

…map` BFS

zalanlevai force-pushed the fix-visible-parent-map-fallback-entry-check branch from 45cd627 to 640032d Compare August 5, 2026 16:42

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 5, 2026

oli-obk approved these changes Aug 6, 2026

View reviewed changes

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 Aug 6, 2026

JonathanBrouwer mentioned this pull request Aug 6, 2026

Rollup of 13 pull requests #160609

Closed

JonathanBrouwer mentioned this pull request Aug 6, 2026

Rollup of 13 pull requests #160610

Closed

JonathanBrouwer mentioned this pull request Aug 6, 2026

Rollup of 17 pull requests #160615

Closed

This was referenced Aug 6, 2026

Rollup of 16 pull requests #160623

Closed

Rollup of 17 pull requests #160627

Merged

rust-bors Bot merged commit c5549dd into rust-lang:main Aug 6, 2026
13 checks passed

rustbot added this to the 1.99.0 milestone Aug 6, 2026

This comment has been minimized.

Sign in to view

rustbot added the perf-regression Performance regression. label Aug 7, 2026

Uh oh!

Uh oh!

Conversation

zalanlevai commented Aug 3, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rustbot commented Aug 3, 2026

Uh oh!

zalanlevai commented Aug 3, 2026

Uh oh!

zalanlevai commented Aug 3, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

tgross35 commented Aug 3, 2026

Uh oh!

zalanlevai commented Aug 3, 2026

Uh oh!

tgross35 commented Aug 3, 2026

Uh oh!

This comment has been minimized.

Kobzol commented Aug 4, 2026

Uh oh!

tgross35 commented Aug 4, 2026

Uh oh!

zalanlevai commented Aug 4, 2026

Uh oh!

zalanlevai commented Aug 4, 2026

Uh oh!

Kobzol commented Aug 4, 2026

Uh oh!

rustbot commented Aug 4, 2026

Uh oh!

nnethercote commented Aug 5, 2026

Uh oh!

oli-obk 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!

rustbot commented Aug 5, 2026

Uh oh!

zalanlevai commented Aug 5, 2026

Uh oh!

oli-obk 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!

rust-bors Bot commented Aug 6, 2026

Uh oh!

Uh oh!

JonathanBrouwer commented Aug 6, 2026

Uh oh!

This comment has been minimized.

rust-timer commented Aug 7, 2026

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

Instruction count

Max RSS (memory usage)

Cycles

Binary size

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

10 participants

zalanlevai commented Aug 3, 2026 •

edited by rustbot

Loading

zalanlevai commented Aug 3, 2026 •

edited

Loading

oli-obk left a comment •

edited by rustbot

Loading

oli-obk left a comment •

edited by rustbot

Loading