From: "Onur Özkan" <work@onurozkan.dev>
To: Danilo Krummrich <dakr@kernel.org>
Cc: gregkh@linuxfoundation.org, arve@android.com, tkjos@android.com,
brauner@kernel.org, cmllamas@google.com, aliceryhl@google.com,
boqun@kernel.org, gary@garyguo.net, lyude@redhat.com,
daniel.almeida@collabora.com, juri.lelli@redhat.com,
vincent.guittot@linaro.org, dietmar.eggemann@arm.com,
rostedt@goodmis.org, bsegall@google.com, mgorman@suse.de,
vschneid@redhat.com, kprateek.nayak@amd.com, ojeda@kernel.org,
bjorn3_gh@protonmail.com, lossin@kernel.org,
a.hindborg@kernel.org, tmgross@umich.edu, tamird@kernel.org,
acourbot@nvidia.com, peterz@infradead.org, mingo@redhat.com,
will@kernel.org, longman@redhat.com, viro@zeniv.linux.org.uk,
jack@suse.cz, tj@kernel.org, jiangshanlai@gmail.com,
linux-kernel@vger.kernel.org, rust-for-linux@vger.kernel.org,
linux-fsdevel@vger.kernel.org
Subject: Re: [PATCH 3/5] rust: sync: add WaitQueue infrastructure
Date: Tue, 28 Jul 2026 09:11:46 +0300 [thread overview]
Message-ID: <20260728061150.206946-1-work@onurozkan.dev> (raw)
In-Reply-To: <20260726223613.1242940-4-dakr@kernel.org>
On Mon, 27 Jul 2026 00:36:09 +0200
Danilo Krummrich <dakr@kernel.org> wrote:
> Implement a wait queue wrapping the kernel's struct wait_queue_head,
> with wait_event()-style methods that take a condition closure directly.
>
> The API mirrors the C wait_event() family:
>
> - wait_event()
> - wait_event_interruptible()
> - wait_event_timeout()
> - wait_event_interruptible_timeout()
> - wake_up() / wake_up_all() / wake_up_sync()
>
> Interruptible and timeout variants return Result<(), WaitError>, where
> WaitError maps to ERESTARTSYS (signal) or ETIMEDOUT (timeout).
>
> Signed-off-by: Danilo Krummrich <dakr@kernel.org>
> ---
> rust/kernel/sync.rs | 6 +
> rust/kernel/sync/wait.rs | 388 +++++++++++++++++++++++++++++++++++++++
> 2 files changed, 394 insertions(+)
> create mode 100644 rust/kernel/sync/wait.rs
>
> diff --git a/rust/kernel/sync.rs b/rust/kernel/sync.rs
> index df4f2604ff9b..31f0999e5747 100644
> --- a/rust/kernel/sync.rs
> +++ b/rust/kernel/sync.rs
> @@ -21,6 +21,7 @@
> pub mod rcu;
> mod refcount;
> mod set_once;
> +mod wait;
>
> pub use arc::{Arc, ArcBorrow, UniqueArc};
> pub use completion::Completion;
> @@ -38,6 +39,11 @@
> pub use locked_by::LockedBy;
> pub use refcount::Refcount;
> pub use set_once::SetOnce;
> +pub use wait::{
> + new_waitqueue,
> + WaitError,
> + WaitQueue, //
> +};
>
> /// Represents a lockdep class.
> ///
> diff --git a/rust/kernel/sync/wait.rs b/rust/kernel/sync/wait.rs
> new file mode 100644
> index 000000000000..ba4ee0f8d4d4
> --- /dev/null
> +++ b/rust/kernel/sync/wait.rs
> @@ -0,0 +1,388 @@
> +// SPDX-License-Identifier: GPL-2.0
> +
> +//! Wait queue.
> +//!
> +//! C header: [`include/linux/wait.h`](srctree/include/linux/wait.h)
> +
> +use super::LockClassKey;
> +use crate::{
> + prelude::*,
> + str::CStr,
> + task::{
> + self,
> + TASK_INTERRUPTIBLE,
> + TASK_NORMAL,
> + TASK_UNINTERRUPTIBLE, //
> + },
> + time::Jiffies,
> + types::Opaque,
> +};
> +
> +use core::{
> + pin::Pin,
> + ptr, //
> +};
> +
> +/// Creates a [`WaitQueue`] initialiser with the given name and a newly-created lock class.
> +#[macro_export]
> +macro_rules! new_waitqueue {
> + ($($name:literal)?) => {
> + $crate::sync::WaitQueue::new(
> + $crate::optional_name!($($name)?),
> + $crate::static_lock_class!(),
> + )
> + };
> +}
> +pub use new_waitqueue;
> +
> +/// Exposes the kernel's [`struct wait_queue_head`] as a Rust wait queue.
> +///
> +/// A `WaitQueue` allows a thread to sleep until a caller-supplied condition becomes true,
> +/// re-checking the condition on each wake-up. This matches the C `wait_event()` family of macros.
> +///
> +/// For waiting with a lock guard (the condition variable pattern), use [`CondVar`](super::CondVar)
> +/// instead.
> +///
> +/// Instances of `WaitQueue` need a lock class and to be pinned. The recommended way to create such
> +/// instances is with the [`pin_init!`] and [`new_waitqueue!`] macros.
> +///
> +/// # Examples
> +///
> +/// ```
> +/// use kernel::sync::{
> +/// atomic::{
> +/// Atomic,
> +/// Relaxed,
> +/// },
> +/// new_waitqueue,
> +/// WaitQueue,
> +/// };
> +///
> +/// #[pin_data]
> +/// pub struct Example {
> +/// value: Atomic<i32>,
> +/// #[pin]
> +/// queue: WaitQueue,
> +/// }
> +///
> +/// fn wait_for_value(e: &Example, v: i32) {
> +/// e.queue.wait_event(|| e.value.load(Relaxed) == v);
> +/// }
> +///
> +/// fn set_value(e: &Example, v: i32) {
> +/// e.value.store(v, Relaxed);
> +/// e.queue.wake_up();
> +/// }
> +/// ```
> +///
> +/// [`struct wait_queue_head`]: srctree/include/linux/wait.h
> +#[pin_data]
> +pub struct WaitQueue {
> + #[pin]
> + wait_queue_head: Opaque<bindings::wait_queue_head>,
> +}
> +
> +// SAFETY: `WaitQueue` only uses a `struct wait_queue_head`, which is safe to use on any thread.
> +unsafe impl Send for WaitQueue {}
> +
> +// SAFETY: `WaitQueue` only uses a `struct wait_queue_head`, which is safe to use on multiple
> +// threads concurrently.
> +unsafe impl Sync for WaitQueue {}
> +
> +impl WaitQueue {
> + /// Constructs a new wait queue initialiser.
> + pub fn new(name: &'static CStr, key: Pin<&'static LockClassKey>) -> impl PinInit<Self> {
> + pin_init!(Self {
> + // SAFETY: `slot` is valid while the closure is called and both `name` and `key` have
> + // static lifetimes so they live indefinitely.
> + wait_queue_head <- Opaque::ffi_init(|slot| unsafe {
> + bindings::__init_waitqueue_head(slot, name.as_char_ptr(), key.as_ptr())
> + }),
> + })
> + }
> +
> + /// Returns a raw pointer to the underlying `wait_queue_head`.
> + #[expect(unused)]
> + #[inline]
> + pub(super) fn as_raw(&self) -> *mut bindings::wait_queue_head {
> + self.wait_queue_head.get()
> + }
> +
> + /// Sleeps until the condition returns `true`.
> + ///
> + /// The condition is checked before each sleep and after each wake-up. The wait is
> + /// uninterruptible.
> + #[inline]
> + pub fn wait_event<F: Fn() -> bool>(&self, condition: F) {
> + self.wait_event_timeout_internal(TASK_UNINTERRUPTIBLE, &condition, Jiffies::MAX);
> + }
> +
> + /// Sleeps until the condition returns `true` or a signal is received.
> + ///
> + /// Returns `Ok(())` when the condition is met, or `Err(WaitError::Signal)` if interrupted
> + /// by a signal.
> + #[inline]
> + pub fn wait_event_interruptible<F: Fn() -> bool>(&self, condition: F) -> Result<(), WaitError> {
> + self.wait_event_timeout_internal(TASK_INTERRUPTIBLE, &condition, Jiffies::MAX);
> + if !condition() && current!().signal_pending() {
> + Err(WaitError::Signal)
> + } else {
> + Ok(())
> + }
> + }
> +
> + /// Sleeps until the condition returns `true` or the timeout expires.
> + ///
> + /// Returns `Ok(())` when the condition is met, or `Err(WaitError::Timeout)` if the timeout
> + /// elapsed first.
> + #[inline]
> + pub fn wait_event_timeout<F: Fn() -> bool>(
> + &self,
> + condition: F,
> + jiffies: Jiffies,
> + ) -> Result<(), WaitError> {
> + let remaining = self.wait_event_timeout_internal(TASK_UNINTERRUPTIBLE, &condition, jiffies);
> + if remaining == 0 && !condition() {
> + Err(WaitError::Timeout)
> + } else {
> + Ok(())
> + }
> + }
> +
> + /// Sleeps until the condition returns `true`, a signal is received, or the timeout expires.
> + ///
> + /// Returns `Ok(())` when the condition is met, or `Err(WaitError)` on signal or timeout.
> + #[inline]
> + pub fn wait_event_interruptible_timeout<F: Fn() -> bool>(
> + &self,
> + condition: F,
> + jiffies: Jiffies,
> + ) -> Result<(), WaitError> {
> + let remaining = self.wait_event_timeout_internal(TASK_INTERRUPTIBLE, &condition, jiffies);
> + if condition() {
> + Ok(())
> + } else if current!().signal_pending() {
> + Err(WaitError::Signal)
> + } else if remaining == 0 {
> + Err(WaitError::Timeout)
> + } else {
> + Ok(())
> + }
> + }
> +
> + fn wait_event_timeout_internal(
> + &self,
> + wait_state: c_int,
> + condition: &dyn Fn() -> bool,
> + jiffies: Jiffies,
> + ) -> Jiffies {
Would it make sense to check condition() first and return early?
> + let wait = Opaque::<bindings::wait_queue_entry>::uninit();
> +
> + // SAFETY: `wait` points to valid memory.
> + unsafe { bindings::init_wait(wait.get()) };
> +
> + let mut remaining = jiffies;
> +
> + loop {
> + // SAFETY: Both `wait` and `wait_queue_head` point to valid memory, and `wait` was
> + // initialised by `init_wait()` above.
> + let ret = unsafe {
> + bindings::prepare_to_wait_event(self.wait_queue_head.get(), wait.get(), wait_state)
> + };
> +
> + if condition() {
> + break;
> + }
> +
> + if ret != 0 || remaining == 0 {
> + break;
> + }
> +
> + remaining = task::schedule_timeout(remaining);
> +
> + if condition() {
> + break;
> + }
> + }
> +
> + // SAFETY: Both `wait` and `wait_queue_head` point to valid memory.
> + unsafe { bindings::finish_wait(self.wait_queue_head.get(), wait.get()) };
> +
> + remaining
> + }
> +
> + /// Performs a single exclusive prepare-to-wait / finish-wait cycle, calling `schedule_fn`
> + /// in between.
> + #[expect(unused)]
> + pub(super) fn wait_once_exclusive<F, R>(&self, wait_state: c_int, schedule_fn: F) -> R
> + where
> + F: FnOnce() -> R,
> + {
> + let wait = Opaque::<bindings::wait_queue_entry>::uninit();
> +
> + // SAFETY: `wait` points to valid memory.
> + unsafe { bindings::init_wait(wait.get()) };
> +
> + // SAFETY: Both `wait` and `wait_queue_head` point to valid memory.
> + unsafe {
> + bindings::prepare_to_wait_exclusive(self.wait_queue_head.get(), wait.get(), wait_state)
> + };
> +
> + let ret = schedule_fn();
> +
> + // SAFETY: Both `wait` and `wait_queue_head` point to valid memory.
> + unsafe { bindings::finish_wait(self.wait_queue_head.get(), wait.get()) };
> +
> + ret
> + }
> +
> + /// Wakes up waiters.
> + ///
> + /// Wakes all non-exclusive waiters and one exclusive waiter, if any. Matches C's `wake_up()`.
> + #[inline]
> + pub fn wake_up(&self) {
> + // SAFETY: `wait_queue_head` points to valid memory.
> + unsafe { bindings::__wake_up(self.wait_queue_head.get(), TASK_NORMAL, 1, ptr::null_mut()) };
> + }
> +
> + /// Wakes up all waiters.
> + ///
> + /// Wakes all non-exclusive and all exclusive waiters, if any.
> + /// Matches C's `wake_up_all()`.
> + #[inline]
> + pub fn wake_up_all(&self) {
> + // SAFETY: `wait_queue_head` points to valid memory.
> + unsafe { bindings::__wake_up(self.wait_queue_head.get(), TASK_NORMAL, 0, ptr::null_mut()) };
> + }
> +
> + /// Like [`wake_up()`](Self::wake_up), but hints to the scheduler that the current task is
> + /// about to sleep, so the woken task should be scheduled on the same CPU to avoid unnecessary
> + /// migration. Matches C's `wake_up_sync()`.
> + #[inline]
> + pub fn wake_up_sync(&self) {
> + // SAFETY: `wait_queue_head` points to valid memory.
> + unsafe { bindings::__wake_up_sync(self.wait_queue_head.get(), TASK_NORMAL) };
> + }
> +
> + /// Wakes up all waiters and clears poll registrations.
> + ///
> + /// Used when a wait queue is about to be freed, to ensure epoll items are properly removed.
> + /// Matches C's `wake_up_pollfree()`.
> + #[inline]
> + #[expect(unused)]
> + pub(super) fn wake_up_pollfree(&self) {
> + // SAFETY: `wait_queue_head` points to valid memory.
> + unsafe { bindings::__wake_up_pollfree(self.wait_queue_head.get()) };
> + }
> +}
> +
> +/// Error returned by [`WaitQueue`] wait functions.
> +#[derive(Debug, PartialEq)]
> +pub enum WaitError {
> + /// Interrupted by a signal.
> + Signal,
> + /// The timeout elapsed without the condition being met.
> + Timeout,
> +}
> +
> +impl From<WaitError> for Error {
> + #[inline]
> + fn from(e: WaitError) -> Error {
> + match e {
> + WaitError::Signal => ERESTARTSYS,
> + WaitError::Timeout => ETIMEDOUT,
> + }
> + }
> +}
> +
> +#[macros::kunit_tests(rust_waitqueue)]
> +mod tests {
> + use super::*;
> + use crate::{
> + sync::{
> + atomic::{
> + Atomic,
> + Relaxed, //
> + },
> + Arc,
> + },
> + time::{
> + delay::fsleep,
> + Delta, //
> + },
> + workqueue,
> + };
> +
> + #[pin_data]
> + struct State {
> + value: Atomic<i32>,
> + #[pin]
> + wq: WaitQueue,
> + }
> +
> + impl State {
> + fn new() -> Result<Arc<Self>> {
> + Arc::pin_init(
> + pin_init!(Self {
> + value: Atomic::new(0),
> + wq <- new_waitqueue!(),
> + }),
> + GFP_KERNEL,
> + )
> + }
> + }
> +
> + #[test]
> + fn wait_event_from_work() {
> + let s = State::new().unwrap();
> +
> + let s2 = s.clone();
> + workqueue::system_dfl()
> + .try_spawn(GFP_KERNEL, move || {
> + s2.value.store(1, Relaxed);
> + s2.wq.wake_up();
> + })
> + .unwrap();
> +
> + s.wq.wait_event(|| s.value.load(Relaxed) == 1);
> + assert_eq!(s.value.load(Relaxed), 1);
> + }
> +
> + #[test]
> + fn wait_event_condition_already_true() {
> + let s = State::new().unwrap();
> + s.value.store(1, Relaxed);
> +
> + s.wq.wait_event(|| s.value.load(Relaxed) == 1);
> + assert_eq!(s.value.load(Relaxed), 1);
> + }
> +
> + #[test]
> + fn wait_event_condition_not_yet_met() {
> + let s = State::new().unwrap();
> +
> + let s2 = s.clone();
> + workqueue::system_dfl()
> + .try_spawn(GFP_KERNEL, move || {
> + s2.value.store(1, Relaxed);
> + s2.wq.wake_up_all();
> +
> + fsleep(Delta::from_millis(50));
> +
> + s2.value.store(2, Relaxed);
> + s2.wq.wake_up_all();
> + })
> + .unwrap();
> +
> + s.wq.wait_event(|| s.value.load(Relaxed) == 2);
> + assert_eq!(s.value.load(Relaxed), 2);
> + }
> +
> + #[test]
> + fn wait_event_timeout_expires() {
> + let s = State::new().unwrap();
> +
> + let ret = s.wq.wait_event_timeout(|| s.value.load(Relaxed) == 1, 1);
> + assert_eq!(ret, Err(WaitError::Timeout));
> + }
> +}
> --
> 2.55.0
>
next prev parent reply other threads:[~2026-07-28 6:12 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-07-26 22:36 [PATCH 0/5] rust: sync: add WaitQueue infrastructure Danilo Krummrich
2026-07-26 22:36 ` [PATCH 1/5] rust: task: add safe schedule_timeout() wrapper Danilo Krummrich
2026-07-26 22:36 ` [PATCH 2/5] rust: workqueue: replace deprecated system_wq with system_{percpu,dfl}_wq Danilo Krummrich
2026-07-27 11:53 ` Gary Guo
2026-07-26 22:36 ` [PATCH 3/5] rust: sync: add WaitQueue infrastructure Danilo Krummrich
2026-07-27 12:02 ` Gary Guo
2026-07-27 12:46 ` Danilo Krummrich
2026-07-27 13:21 ` Gary Guo
2026-07-28 6:11 ` Onur Özkan [this message]
2026-07-26 22:36 ` [PATCH 4/5] rust: sync: convert CondVar and PollCondVar to use WaitQueue Danilo Krummrich
2026-07-26 22:36 ` [PATCH 5/5] rust: sync: condvar: use task::schedule_timeout() Danilo Krummrich
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260728061150.206946-1-work@onurozkan.dev \
--to=work@onurozkan.dev \
--cc=a.hindborg@kernel.org \
--cc=acourbot@nvidia.com \
--cc=aliceryhl@google.com \
--cc=arve@android.com \
--cc=bjorn3_gh@protonmail.com \
--cc=boqun@kernel.org \
--cc=brauner@kernel.org \
--cc=bsegall@google.com \
--cc=cmllamas@google.com \
--cc=dakr@kernel.org \
--cc=daniel.almeida@collabora.com \
--cc=dietmar.eggemann@arm.com \
--cc=gary@garyguo.net \
--cc=gregkh@linuxfoundation.org \
--cc=jack@suse.cz \
--cc=jiangshanlai@gmail.com \
--cc=juri.lelli@redhat.com \
--cc=kprateek.nayak@amd.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=longman@redhat.com \
--cc=lossin@kernel.org \
--cc=lyude@redhat.com \
--cc=mgorman@suse.de \
--cc=mingo@redhat.com \
--cc=ojeda@kernel.org \
--cc=peterz@infradead.org \
--cc=rostedt@goodmis.org \
--cc=rust-for-linux@vger.kernel.org \
--cc=tamird@kernel.org \
--cc=tj@kernel.org \
--cc=tkjos@android.com \
--cc=tmgross@umich.edu \
--cc=vincent.guittot@linaro.org \
--cc=viro@zeniv.linux.org.uk \
--cc=vschneid@redhat.com \
--cc=will@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox