Add is_disconnected functions to mpsc and mpmc channels

[LevitatingBusinessMan]

Add is_disconnected() functions to the Sender and Receiver of both mpmc an mpsc channels.

std::sync::mpmc::Sender<T>::is_disconnected(&self) -> bool
std::sync::mpmc::Receiver<T>::is_disconnected(&self) -> bool
std::sync::mpsc::Sender<T>::is_disconnected(&self) -> bool
std::sync::mpsc::Receiver<T>::is_disconnected(&self) -> bool

The mpsc methods are locked behind the mpsc_is_disconnected feature gate, which has no tracking issue yet.

ACP: rust-lang/libs-team#748
Tracking issue: #153668

[rustbot]

r? @Mark-Simulacrum

rustbot has assigned @Mark-Simulacrum.
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: @ChrisDenton, libs
• @ChrisDenton, libs expanded to 8 candidates
• Random selection from Mark-Simulacrum, joboet

[LevitatingBusinessMan]

@rustbot label +S-waiting-on-ACP

[a1phyr]

I think that it the doc should mention that if is_disconnected returns false, it does not guaranty anything in the future (eg a send/recv just after might fail anything).

This footgun can cause race conditions, so it should be documented.

View changes since this review

[LevitatingBusinessMan]

This footgun can cause race conditions, so it should be documented.

I added the following text to the documentation:

/// Note that a return value of `false` does not guarantee the channel will/// remain connected. The channel may be disconnected immediately after this method/// returns, so a subsequent [`Sender::send`] may still fail with [`SendError`].

and subsequent versions for Receiver.

[LevitatingBusinessMan]

@rustbot label +S-waiting-on-review -S-waiting-on-ACP

[LevitatingBusinessMan]

Currently I've put the methods for mpmc behind #![feature(mpmc_channel)]. Somebody should tell me if that's a good idea, or that both variants should be gated behind the same feature.

[Mark-Simulacrum]

@bors r+ rollup

I think it seems reasonable either way but this setup seems fine for the feature gate.

[rust-bors]

📌 Commit ce2de64 has been approved by Mark-Simulacrum

It is now in the queue for this repository.