From: Boqun Feng <boqun.feng@gmail.com>
To: FUJITA Tomonori <fujita.tomonori@gmail.com>
Cc: aliceryhl@google.com, arve@android.com, brauner@kernel.org,
cmllamas@google.com, gregkh@linuxfoundation.org,
ojeda@kernel.org, tkjos@android.com, a.hindborg@kernel.org,
bjorn3_gh@protonmail.com, dakr@kernel.org, gary@garyguo.net,
lossin@kernel.org, tmgross@umich.edu,
rust-for-linux@vger.kernel.org
Subject: Re: [PATCH v1 1/3] rust: revocable: Switch to kernel::sync atomic primitives
Date: Wed, 31 Dec 2025 07:29:17 +0800 [thread overview]
Message-ID: <aVRgTQWhqjKB6xpK@tardis-2.local> (raw)
In-Reply-To: <20251230093718.1852322-2-fujita.tomonori@gmail.com>
On Tue, Dec 30, 2025 at 06:37:16PM +0900, FUJITA Tomonori wrote:
> Convert uses of `AtomicBool` to `Atomic<bool>`.
>
> Signed-off-by: FUJITA Tomonori <fujita.tomonori@gmail.com>
> ---
> rust/kernel/revocable.rs | 24 ++++++++++++++++--------
> 1 file changed, 16 insertions(+), 8 deletions(-)
>
> diff --git a/rust/kernel/revocable.rs b/rust/kernel/revocable.rs
> index 0f4ae673256d..47cd2b45eada 100644
> --- a/rust/kernel/revocable.rs
> +++ b/rust/kernel/revocable.rs
> @@ -7,12 +7,20 @@
>
> use pin_init::Wrapper;
>
> -use crate::{bindings, prelude::*, sync::rcu, types::Opaque};
> +use crate::{
> + bindings,
> + prelude::*,
> + sync::atomic::{
> + ordering::Relaxed,
> + Atomic, //
> + },
> + sync::rcu,
> + types::Opaque, //
> +};
> use core::{
> marker::PhantomData,
> ops::Deref,
> - ptr::drop_in_place,
> - sync::atomic::{AtomicBool, Ordering},
> + ptr::drop_in_place, //
> };
>
> /// An object that can become inaccessible at runtime.
> @@ -65,7 +73,7 @@
> /// ```
> #[pin_data(PinnedDrop)]
> pub struct Revocable<T> {
> - is_available: AtomicBool,
> + is_available: Atomic<bool>,
> #[pin]
> data: Opaque<T>,
I actually think we should use Atomic<i32> instead of Atomic<bool> here,
because in most of the cases `T` is a aligned to at least 4-bytes (e.g.
Devres uses a Revocable<IoMem>) And in that case, using Atomic<bool>
doesn't save extra memory but makes the xchg() operation slower on
architectures that don't support byte-wise RmW instructions (e.g.
riscv).
My rule of thumb for using Atomic<bool> is: when you really care about
the memory usage, and you can guarantee using Atomic<bool> is saving
memory. For a general "flag" usage, Atomic<bool> is not strictly better
than Atomic<i32> in most cases imo.
Regards,
Boqun
> }
> @@ -84,7 +92,7 @@ impl<T> Revocable<T> {
> /// Creates a new revocable instance of the given data.
> pub fn new<E>(data: impl PinInit<T, E>) -> impl PinInit<Self, E> {
> try_pin_init!(Self {
> - is_available: AtomicBool::new(true),
> + is_available: Atomic::new(true),
> data <- Opaque::pin_init(data),
> }? E)
> }
> @@ -98,7 +106,7 @@ pub fn new<E>(data: impl PinInit<T, E>) -> impl PinInit<Self, E> {
> /// because another CPU may be waiting to complete the revocation of this object.
> pub fn try_access(&self) -> Option<RevocableGuard<'_, T>> {
> let guard = rcu::read_lock();
> - if self.is_available.load(Ordering::Relaxed) {
> + if self.is_available.load(Relaxed) {
> // Since `self.is_available` is true, data is initialised and has to remain valid
> // because the RCU read side lock prevents it from being dropped.
> Some(RevocableGuard::new(self.data.get(), guard))
> @@ -116,7 +124,7 @@ pub fn try_access(&self) -> Option<RevocableGuard<'_, T>> {
> /// allowed to sleep because another CPU may be waiting to complete the revocation of this
> /// object.
> pub fn try_access_with_guard<'a>(&'a self, _guard: &'a rcu::Guard) -> Option<&'a T> {
> - if self.is_available.load(Ordering::Relaxed) {
> + if self.is_available.load(Relaxed) {
> // SAFETY: Since `self.is_available` is true, data is initialised and has to remain
> // valid because the RCU read side lock prevents it from being dropped.
> Some(unsafe { &*self.data.get() })
> @@ -157,7 +165,7 @@ pub unsafe fn access(&self) -> &T {
> ///
> /// Callers must ensure that there are no more concurrent users of the revocable object.
> unsafe fn revoke_internal<const SYNC: bool>(&self) -> bool {
> - let revoke = self.is_available.swap(false, Ordering::Relaxed);
> + let revoke = self.is_available.xchg(false, Relaxed);
>
> if revoke {
> if SYNC {
> --
> 2.43.0
>
next prev parent reply other threads:[~2025-12-30 23:29 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-12-30 9:37 [PATCH v1 0/3] rust: Switch to kernel::sync atomic primitives FUJITA Tomonori
2025-12-30 9:37 ` [PATCH v1 1/3] rust: revocable: " FUJITA Tomonori
2025-12-30 23:29 ` Boqun Feng [this message]
2025-12-31 9:35 ` FUJITA Tomonori
2025-12-31 23:41 ` Boqun Feng
2026-01-01 4:34 ` Gary Guo
2025-12-31 10:12 ` Alice Ryhl
2025-12-31 10:33 ` FUJITA Tomonori
2025-12-31 10:54 ` Alice Ryhl
2025-12-30 9:37 ` [PATCH v1 2/3] rust: list: " FUJITA Tomonori
2025-12-31 10:53 ` Alice Ryhl
2025-12-30 9:37 ` [PATCH v1 3/3] rust_binder: " FUJITA Tomonori
2025-12-31 10:52 ` Alice Ryhl
2026-01-05 13:34 ` [PATCH v1 0/3] rust: " Boqun Feng
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=aVRgTQWhqjKB6xpK@tardis-2.local \
--to=boqun.feng@gmail.com \
--cc=a.hindborg@kernel.org \
--cc=aliceryhl@google.com \
--cc=arve@android.com \
--cc=bjorn3_gh@protonmail.com \
--cc=brauner@kernel.org \
--cc=cmllamas@google.com \
--cc=dakr@kernel.org \
--cc=fujita.tomonori@gmail.com \
--cc=gary@garyguo.net \
--cc=gregkh@linuxfoundation.org \
--cc=lossin@kernel.org \
--cc=ojeda@kernel.org \
--cc=rust-for-linux@vger.kernel.org \
--cc=tkjos@android.com \
--cc=tmgross@umich.edu \
/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