Implementation: `#[feature(nonpoison_once)]` by connortsui20 · Pull Request #144653 · rust-lang/rust

connortsui20 · 2025-07-29T20:30:24Z

Tracking Issue: #134645

This PR continues the effort made in #144022 by adding the implementation of nonpoison::once.

Many of the changes here are similar to the changes made to implement nonpoison::mutex.

Note that even though there were FIXMEs for changing OnceLock and LazyLock to use the nonpoison variant, that is not correct as both rely on poisoning behavior to operate correctly.

Related PRs

nonpoison_rwlock implementation: Implementation: #[feature(nonpoison_rwlock)] #144648
nonpoison_condvar implementation: Implementation: #[feature(nonpoison_condvar)] #144651

rustbot · 2025-07-29T20:30:28Z

r? @ChrisDenton

rustbot has assigned @ChrisDenton.
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

…ark-Simulacrum Implementation: `#[feature(nonpoison_rwlock)]` Tracking Issue: rust-lang#134645 This PR continues the effort made in rust-lang#144022 by adding the implementation of `nonpoison::rwlock`. Many of the changes here are similar to the changes made to implement `nonpoison::mutex`. The only real difference is that this PR includes a reorganizing of the existing `poison::rwlock` file that hopefully makes both variants more readable. ### Related PRs - `nonpoison_condvar` implementation: rust-lang#144651 - `nonpoison_once` implementation: rust-lang#144653

ChrisDenton · 2025-08-23T07:52:50Z

Sorry for the delay, I've not been keep up with reviews atm. I'll assign another reviewer.

r? libs

Rollup merge of #144648 - connortsui20:nonpoison_rwlock, r=Mark-Simulacrum Implementation: `#[feature(nonpoison_rwlock)]` Tracking Issue: #134645 This PR continues the effort made in #144022 by adding the implementation of `nonpoison::rwlock`. Many of the changes here are similar to the changes made to implement `nonpoison::mutex`. The only real difference is that this PR includes a reorganizing of the existing `poison::rwlock` file that hopefully makes both variants more readable. ### Related PRs - `nonpoison_condvar` implementation: #144651 - `nonpoison_once` implementation: #144653

bors · 2025-08-23T09:16:43Z

☔ The latest upstream changes (presumably #145773) made this pull request unmergeable. Please resolve the merge conflicts.

Adds the equivalent `nonpoison` types to the `poison::once` module. These types and implementations are gated under the `nonpoison_once` feature gate. Signed-off-by: Connor Tsui <[email protected]>

Signed-off-by: Connor Tsui <[email protected]>

Both implementations require poisoning support. Signed-off-by: Connor Tsui <[email protected]>

rustbot · 2025-08-23T13:30:26Z

This PR was rebased onto a different master 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.

connortsui20 · 2025-08-23T13:53:43Z

library/std/src/sync/nonpoison/once.rs

For reviewers: it might be good to compare this directly with the poison::once module.

(for copy paste)

delta library/std/src/sync/poison/once.rs library/std/src/sync/nonpoison/once.rs

diff library/std/src/sync/poison/once.rs library/std/src/sync/nonpoison/once.rs

…acrum Implementation: `#[feature(nonpoison_rwlock)]` Tracking Issue: rust-lang/rust#134645 This PR continues the effort made in rust-lang/rust#144022 by adding the implementation of `nonpoison::rwlock`. Many of the changes here are similar to the changes made to implement `nonpoison::mutex`. The only real difference is that this PR includes a reorganizing of the existing `poison::rwlock` file that hopefully makes both variants more readable. ### Related PRs - `nonpoison_condvar` implementation: rust-lang/rust#144651 - `nonpoison_once` implementation: rust-lang/rust#144653

…ark-Simulacrum Implementation: `#[feature(nonpoison_rwlock)]` Tracking Issue: rust-lang#134645 This PR continues the effort made in rust-lang#144022 by adding the implementation of `nonpoison::rwlock`. Many of the changes here are similar to the changes made to implement `nonpoison::mutex`. The only real difference is that this PR includes a reorganizing of the existing `poison::rwlock` file that hopefully makes both variants more readable. ### Related PRs - `nonpoison_condvar` implementation: rust-lang#144651 - `nonpoison_once` implementation: rust-lang#144653

connortsui20 · 2025-08-30T10:18:49Z

library/std/src/sync/nonpoison/once.rs

+    #[inline]
+    #[unstable(feature = "nonpoison_once", issue = "134645")]
+    #[track_caller]
+    pub fn call_once<F>(&self, f: F)


This has the exact same behavior as the poison::Once::call_once_force method. Could that be confusing (since it is named differently but is identical)?

connortsui20 · 2025-08-30T10:19:15Z

library/std/src/sync/nonpoison/once.rs

+    /// [`call_once()`]: Once::call_once
+    /// [`poison::Once::wait_force`]: crate::sync::poison::Once::wait_force
+    #[unstable(feature = "nonpoison_once", issue = "134645")]
+    pub fn wait(&self) {


Similarly, this has the same behavior as wait_force.

connortsui20 · 2025-08-30T10:20:33Z

library/std/tests/sync/once.rs

Note that there are no "new" tests here, just refactored tests for the macro

…joboet Implementation: `#[feature(nonpoison_condvar)]` Tracking Issue: rust-lang#134645 This PR continues the effort made in rust-lang#144022 by adding the implementation of `nonpoison::condvar`. Many of the changes here are similar to the changes made to implement `nonpoison::mutex`. There are two other changes here. The first is that the `Barrier` implementation is migrated to use the `nonpoison::Condvar` instead of the `poison` variant. The second (which might be subject to some discussion) is that `WaitTimeoutResult` is moved up to `mod.rs`, as both `condvar` variants need that type (and I do not know if there is a better place to put it now). ### Related PRs - `nonpoison_rwlock` implementation: rust-lang#144648 - `nonpoison_once` implementation: rust-lang#144653

Rollup merge of #144651 - connortsui20:nonpoison_condvar, r=joboet Implementation: `#[feature(nonpoison_condvar)]` Tracking Issue: #134645 This PR continues the effort made in #144022 by adding the implementation of `nonpoison::condvar`. Many of the changes here are similar to the changes made to implement `nonpoison::mutex`. There are two other changes here. The first is that the `Barrier` implementation is migrated to use the `nonpoison::Condvar` instead of the `poison` variant. The second (which might be subject to some discussion) is that `WaitTimeoutResult` is moved up to `mod.rs`, as both `condvar` variants need that type (and I do not know if there is a better place to put it now). ### Related PRs - `nonpoison_rwlock` implementation: #144648 - `nonpoison_once` implementation: #144653

joboet · 2025-08-31T12:14:19Z

I'm not sure whether it's actually worth it to add a non-poisoning variant here. Contrary to Mutex and RwLock, the non-poisoning variants (call_once_force and wait_force) are already available for Once and poisoning is pretty much irrelevant from a performance perspective.

connortsui20 · 2025-08-31T12:29:10Z

I think I'm inclined to agree (see this comment: #134645 (comment))

Should Once be taken out of the poison module then? Either that or I would argue we should move OnceLock and LazyLock into poison since they also have poisoning semantics.

Also I'm going to close this PR and move this to the tracking issue itself.

connortsui20 · 2025-08-31T12:51:29Z

cc @joboet @tgross35 #134645 (comment)

…joboet Implementation: `#[feature(nonpoison_condvar)]` Tracking Issue: rust-lang#134645 This PR continues the effort made in rust-lang#144022 by adding the implementation of `nonpoison::condvar`. Many of the changes here are similar to the changes made to implement `nonpoison::mutex`. There are two other changes here. The first is that the `Barrier` implementation is migrated to use the `nonpoison::Condvar` instead of the `poison` variant. The second (which might be subject to some discussion) is that `WaitTimeoutResult` is moved up to `mod.rs`, as both `condvar` variants need that type (and I do not know if there is a better place to put it now). ### Related PRs - `nonpoison_rwlock` implementation: rust-lang#144648 - `nonpoison_once` implementation: rust-lang#144653

rustbot assigned ChrisDenton Jul 29, 2025

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-libs Relevant to the library team, which will review and decide on the PR/issue. labels Jul 29, 2025

This was referenced Jul 29, 2025

Implementation: #[feature(nonpoison_rwlock)] #144648

Merged

Implementation: #[feature(nonpoison_condvar)] #144651

Merged

This comment has been minimized.

Sign in to view

connortsui20 force-pushed the nonpoison_once branch from 43b1e0c to b201b42 Compare July 30, 2025 08:30

rustbot assigned Mark-Simulacrum and unassigned ChrisDenton Aug 23, 2025

connortsui20 added 3 commits August 23, 2025 09:23

add nonpoison::once implementation

ca534ee

Adds the equivalent `nonpoison` types to the `poison::once` module. These types and implementations are gated under the `nonpoison_once` feature gate. Signed-off-by: Connor Tsui <[email protected]>

add nonpoison and poison once tests

3a74360

Signed-off-by: Connor Tsui <[email protected]>

remove FIXMEs in once_lock and lazy_lock

2287b60

Both implementations require poisoning support. Signed-off-by: Connor Tsui <[email protected]>

connortsui20 force-pushed the nonpoison_once branch from b201b42 to 2287b60 Compare August 23, 2025 13:30

connortsui20 commented Aug 23, 2025

View reviewed changes

connortsui20 commented Aug 30, 2025

View reviewed changes

library/std/tests/sync/once.rs

Copy link

Contributor Author

connortsui20 Aug 30, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Note that there are no "new" tests here, just refactored tests for the macro

connortsui20 closed this Aug 31, 2025

rustbot removed the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Aug 31, 2025

connortsui20 deleted the nonpoison_once branch August 31, 2025 12:32

connortsui20 mentioned this pull request Aug 31, 2025

Tracking Issue for sync_nonpoison and nonpoison_{condvar,mutex,rwlock} #134645

Open

8 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Implementation: `#[feature(nonpoison_once)]`#144653

Implementation: `#[feature(nonpoison_once)]`#144653
connortsui20 wants to merge 3 commits intorust-lang:masterfrom
connortsui20:nonpoison_once

connortsui20 commented Jul 29, 2025 •

edited

Loading

Uh oh!

rustbot commented Jul 29, 2025

Uh oh!

This comment has been minimized.

ChrisDenton commented Aug 23, 2025

Uh oh!

bors commented Aug 23, 2025

Uh oh!

rustbot commented Aug 23, 2025

Uh oh!

connortsui20 Aug 23, 2025 •

edited

Loading

Uh oh!

connortsui20 Aug 30, 2025

Uh oh!

connortsui20 Aug 30, 2025

Uh oh!

connortsui20 Aug 30, 2025

Uh oh!

joboet commented Aug 31, 2025

Uh oh!

connortsui20 commented Aug 31, 2025

Uh oh!

connortsui20 commented Aug 31, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

Uh oh!

Conversation

connortsui20 commented Jul 29, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Related PRs

Uh oh!

rustbot commented Jul 29, 2025

Uh oh!

This comment has been minimized.

ChrisDenton commented Aug 23, 2025

Uh oh!

bors commented Aug 23, 2025

Uh oh!

rustbot commented Aug 23, 2025

Uh oh!

connortsui20 Aug 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

connortsui20 Aug 30, 2025

Choose a reason for hiding this comment

Uh oh!

connortsui20 Aug 30, 2025

Choose a reason for hiding this comment

Uh oh!

connortsui20 Aug 30, 2025

Choose a reason for hiding this comment

Uh oh!

joboet commented Aug 31, 2025

Uh oh!

connortsui20 commented Aug 31, 2025

Uh oh!

connortsui20 commented Aug 31, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

connortsui20 commented Jul 29, 2025 •

edited

Loading

connortsui20 Aug 23, 2025 •

edited

Loading