public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Andreas Hindborg <nmi@metaspace.dk>
To: Alice Ryhl <aliceryhl@google.com>
Cc: rust-for-linux@vger.kernel.org, "Miguel Ojeda" <ojeda@kernel.org>,
	"Wedson Almeida Filho" <wedsonaf@gmail.com>,
	"Tejun Heo" <tj@kernel.org>,
	"Lai Jiangshan" <jiangshanlai@gmail.com>,
	"Alex Gaynor" <alex.gaynor@gmail.com>,
	"Boqun Feng" <boqun.feng@gmail.com>,
	"Gary Guo" <gary@garyguo.net>,
	"Björn Roy Baron" <bjorn3_gh@protonmail.com>,
	"Benno Lossin" <benno.lossin@proton.me>,
	linux-kernel@vger.kernel.org, patches@lists.linux.dev
Subject: Re: [PATCH v1 1/7] rust: workqueue: add low-level workqueue bindings
Date: Tue, 30 May 2023 10:26:02 +0200	[thread overview]
Message-ID: <87ttvu9q4o.fsf@metaspace.dk> (raw)
In-Reply-To: <20230517203119.3160435-2-aliceryhl@google.com>


Alice Ryhl <aliceryhl@google.com> writes:

> Define basic low-level bindings to a kernel workqueue. The API defined
> here can only be used unsafely. Later commits will provide safe
> wrappers.
>
> Signed-off-by: Alice Ryhl <aliceryhl@google.com>
> ---
>  rust/kernel/lib.rs       |  1 +
>  rust/kernel/workqueue.rs | 99 ++++++++++++++++++++++++++++++++++++++++
>  2 files changed, 100 insertions(+)
>  create mode 100644 rust/kernel/workqueue.rs
>
> diff --git a/rust/kernel/lib.rs b/rust/kernel/lib.rs
> index 676995d4e460..c718524056a6 100644
> --- a/rust/kernel/lib.rs
> +++ b/rust/kernel/lib.rs
> @@ -47,6 +47,7 @@ pub mod str;
>  pub mod sync;
>  pub mod task;
>  pub mod types;
> +pub mod workqueue;
>  
>  #[doc(hidden)]
>  pub use bindings;
> diff --git a/rust/kernel/workqueue.rs b/rust/kernel/workqueue.rs
> new file mode 100644
> index 000000000000..e66b6b50dfae
> --- /dev/null
> +++ b/rust/kernel/workqueue.rs
> @@ -0,0 +1,99 @@
> +// SPDX-License-Identifier: GPL-2.0
> +
> +//! Work queues.
> +//!
> +//! C header: [`include/linux/workqueue.h`](../../../../include/linux/workqueue.h)

I think we need to add workqueue.h to rust/bindings/bindings_helper.h
with this commit. It is probably transitively included as is, but it
would be the right thing to explicitly include it.

BR Andreas

> +
> +use crate::{bindings, types::Opaque};
> +
> +/// A kernel work queue.
> +///
> +/// Wraps the kernel's C `struct workqueue_struct`.
> +///
> +/// It allows work items to be queued to run on thread pools managed by the kernel. Several are
> +/// always available, for example, `system`, `system_highpri`, `system_long`, etc.
> +#[repr(transparent)]
> +pub struct Queue(Opaque<bindings::workqueue_struct>);
> +
> +// SAFETY: Kernel workqueues are usable from any thread.
> +unsafe impl Send for Queue {}
> +unsafe impl Sync for Queue {}
> +
> +impl Queue {
> +    /// Use the provided `struct workqueue_struct` with Rust.
> +    ///
> +    /// # Safety
> +    ///
> +    /// The caller must ensure that the provided raw pointer is not dangling, that it points at a
> +    /// valid workqueue, and that it remains valid until the end of 'a.
> +    pub unsafe fn from_raw<'a>(ptr: *const bindings::workqueue_struct) -> &'a Queue {
> +        // SAFETY: The `Queue` type is `#[repr(transparent)]`, so the pointer cast is valid. The
> +        // caller promises that the pointer is not dangling.
> +        unsafe { &*(ptr as *const Queue) }
> +    }
> +
> +    /// Enqueues a work item.
> +    ///
> +    /// This may fail if the work item is already enqueued in a workqueue.
> +    pub fn enqueue<T: WorkItem + Send + 'static>(&self, w: T) -> T::EnqueueOutput {
> +        let queue_ptr = self.0.get();
> +
> +        // SAFETY: There are two cases.
> +        //
> +        //  1. If `queue_work_on` returns false, then we failed to push the work item to the queue.
> +        //     In this case, we don't touch the work item again.
> +        //
> +        //  2. If `queue_work_on` returns true, then we pushed the work item to the queue. The work
> +        //     queue will call the function pointer in the `work_struct` at some point in the
> +        //     future. We require `T` to be static, so the type has no lifetimes annotated on it.
> +        //     We require `T` to be send, so there are no thread-safety issues to take care of.
> +        //
> +        // In either case we follow the safety requirements of `__enqueue`.
> +        unsafe {
> +            w.__enqueue(move |work_ptr| {
> +                bindings::queue_work_on(bindings::WORK_CPU_UNBOUND as _, queue_ptr, work_ptr)
> +            })
> +        }
> +    }
> +}
> +
> +/// A work item.
> +///
> +/// This is the low-level trait that is designed for being as general as possible.
> +///
> +/// # Safety
> +///
> +/// Implementers must ensure that `__enqueue` behaves as documented.
> +pub unsafe trait WorkItem {
> +    /// The return type of [`Queue::enqueue`].
> +    type EnqueueOutput;
> +
> +    /// Enqueues this work item on a queue using the provided `queue_work_on` method.
> +    ///
> +    /// # Safety
> +    ///
> +    /// Calling this method guarantees that the provided closure will be called with a raw pointer
> +    /// to a `struct work_struct`. The closure should behave in the following way:
> +    ///
> +    ///  1. If the `struct work_struct` cannot be pushed to a workqueue because its already in one,
> +    ///     then the closure should return `false`. It may not access the pointer after returning
> +    ///     `false`.
> +    ///  2. If the `struct work_struct` is successfully added to a workqueue, then the closure
> +    ///     should return `true`. When the workqueue executes the work item, it will do so by
> +    ///     calling the function pointer stored in the `struct work_struct`. The work item ensures
> +    ///     that the raw pointer remains valid until that happens.
> +    ///
> +    /// This method may not have any other failure cases than the closure returning `false`. The
> +    /// output type should reflect this, but it may also be an infallible type if the work item
> +    /// statically ensures that pushing the `struct work_struct` will succeed.
> +    ///
> +    /// If the work item type is annotated with any lifetimes, then the workqueue must call the
> +    /// function pointer before any such lifetime expires. (Or it may forget the work item and
> +    /// never call the function pointer at all.)
> +    ///
> +    /// If the work item type is not [`Send`], then the work item must be executed on the same
> +    /// thread as the call to `__enqueue`.
> +    unsafe fn __enqueue<F>(self, queue_work_on: F) -> Self::EnqueueOutput
> +    where
> +        F: FnOnce(*mut bindings::work_struct) -> bool;
> +}


  parent reply	other threads:[~2023-05-30  8:32 UTC|newest]

Thread overview: 51+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-05-17 20:31 [PATCH v1 0/7] Bindings for the workqueue Alice Ryhl
2023-05-17 20:31 ` [PATCH v1 1/7] rust: workqueue: add low-level workqueue bindings Alice Ryhl
2023-05-18 14:51   ` Martin Rodriguez Reboredo
2023-05-19  9:40     ` Alice Ryhl
2023-05-19 12:04       ` Martin Rodriguez Reboredo
2023-05-23 10:03         ` Alice Ryhl
2023-05-30  8:26   ` Andreas Hindborg [this message]
2023-05-17 20:31 ` [PATCH v1 2/7] rust: add offset_of! macro Alice Ryhl
2023-05-18 14:51   ` Martin Rodriguez Reboredo
2023-05-23 15:48   ` Gary Guo
2023-05-24 12:26     ` Alice Ryhl
2023-05-30  8:40     ` Andreas Hindborg
2023-05-17 20:31 ` [PATCH v1 3/7] rust: sync: add `Arc::{from_raw, into_raw}` Alice Ryhl
2023-05-18 14:51   ` Martin Rodriguez Reboredo
2023-05-23 15:43   ` Gary Guo
2023-05-24 11:19     ` Alice Ryhl
2023-05-24 10:20   ` Andreas Hindborg
2023-05-24 11:11     ` Alice Ryhl
2023-05-25  7:45       ` Andreas Hindborg
2023-05-25 16:32         ` Gary Guo
2023-05-30  7:23           ` Andreas Hindborg
2023-05-17 20:31 ` [PATCH v1 4/7] rust: workqueue: define built-in queues Alice Ryhl
2023-05-18 14:52   ` Martin Rodriguez Reboredo
2023-05-25 11:40   ` Andreas Hindborg
2023-05-31 14:02     ` Alice Ryhl
2023-06-02 10:23       ` Andreas Hindborg (Samsung)
2023-05-17 20:31 ` [PATCH v1 5/7] rust: workqueue: add helper for defining work_struct fields Alice Ryhl
2023-05-18 23:18   ` Martin Rodriguez Reboredo
2023-05-24 14:50   ` Benno Lossin
2023-05-30  8:44   ` Andreas Hindborg
2023-05-31  9:00     ` Alice Ryhl
2023-05-31 10:18       ` Andreas Hindborg
2023-05-17 20:31 ` [PATCH v1 6/7] rust: workqueue: add safe API to workqueue Alice Ryhl
2023-05-19  0:17   ` Martin Rodriguez Reboredo
2023-05-23 11:07     ` Alice Ryhl
2023-05-30  7:19       ` Andreas Hindborg
2023-05-30 13:23         ` Martin Rodriguez Reboredo
2023-05-30 14:13       ` Miguel Ojeda
2023-05-24 14:51   ` Benno Lossin
2023-05-31  9:07     ` Alice Ryhl
2023-05-30  8:51   ` Andreas Hindborg
2023-05-31 14:07     ` Alice Ryhl
2023-05-17 20:31 ` [PATCH v1 7/7] rust: workqueue: add `try_spawn` helper method Alice Ryhl
2023-05-19  0:22   ` Martin Rodriguez Reboredo
2023-05-24 14:52   ` Benno Lossin
2023-05-31 14:03     ` Alice Ryhl
2023-05-17 21:48 ` [PATCH v1 0/7] Bindings for the workqueue Tejun Heo
2023-05-17 22:22   ` Alice Ryhl
2023-05-23 14:08     ` Andreas Hindborg
2023-05-23 14:14 ` Andreas Hindborg
2023-05-24 12:33   ` Alice Ryhl

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=87ttvu9q4o.fsf@metaspace.dk \
    --to=nmi@metaspace.dk \
    --cc=alex.gaynor@gmail.com \
    --cc=aliceryhl@google.com \
    --cc=benno.lossin@proton.me \
    --cc=bjorn3_gh@protonmail.com \
    --cc=boqun.feng@gmail.com \
    --cc=gary@garyguo.net \
    --cc=jiangshanlai@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=ojeda@kernel.org \
    --cc=patches@lists.linux.dev \
    --cc=rust-for-linux@vger.kernel.org \
    --cc=tj@kernel.org \
    --cc=wedsonaf@gmail.com \
    /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