From: Gary Guo <gary@garyguo.net>
To: "Miguel Ojeda" <ojeda@kernel.org>,
"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>,
"Andreas Hindborg" <a.hindborg@kernel.org>,
"Alice Ryhl" <aliceryhl@google.com>,
"Trevor Gross" <tmgross@umich.edu>,
"Wedson Almeida Filho" <walmeida@microsoft.com>,
"Valentin Obst" <kernel@valentinobst.de>,
"Alex Mantel" <alexmantel93@mailbox.org>
Cc: Will Deacon <will@kernel.org>,
Peter Zijlstra <peterz@infradead.org>,
Mark Rutland <mark.rutland@arm.com>,
Martin Rodriguez Reboredo <yakoyoku@gmail.com>,
rust-for-linux@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: [PATCH 2/3] rust: convert `Arc` to use `Refcount`
Date: Fri, 4 Oct 2024 16:52:23 +0100 [thread overview]
Message-ID: <20241004155247.2210469-3-gary@garyguo.net> (raw)
In-Reply-To: <20241004155247.2210469-1-gary@garyguo.net>
With `Refcount` type created, `Arc` can use `Refcount` instead of
calling into FFI directly.
Cc: Will Deacon <will@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Gary Guo <gary@garyguo.net>
---
rust/kernel/sync/arc.rs | 45 +++++++++++++++++------------------------
1 file changed, 19 insertions(+), 26 deletions(-)
diff --git a/rust/kernel/sync/arc.rs b/rust/kernel/sync/arc.rs
index 3021f30fd822..a1fa58b127c8 100644
--- a/rust/kernel/sync/arc.rs
+++ b/rust/kernel/sync/arc.rs
@@ -8,7 +8,7 @@
//! threads.
//!
//! It is different from the standard library's [`Arc`] in a few ways:
-//! 1. It is backed by the kernel's `refcount_t` type.
+//! 1. It is backed by the kernel's [`Refcount`] type.
//! 2. It does not support weak references, which allows it to be half the size.
//! 3. It saturates the reference count instead of aborting when it goes over a threshold.
//! 4. It does not provide a `get_mut` method, so the ref counted object is pinned.
@@ -18,10 +18,10 @@
use crate::{
alloc::{box_ext::BoxExt, AllocError, Flags},
- bindings,
init::{self, InPlaceInit, Init, PinInit},
+ sync::Refcount,
try_init,
- types::{ForeignOwnable, Opaque},
+ types::ForeignOwnable,
};
use alloc::boxed::Box;
use core::{
@@ -134,7 +134,7 @@ pub struct Arc<T: ?Sized> {
#[pin_data]
#[repr(C)]
struct ArcInner<T: ?Sized> {
- refcount: Opaque<bindings::refcount_t>,
+ refcount: Refcount,
data: T,
}
@@ -146,7 +146,7 @@ impl<T: ?Sized> ArcInner<T> {
/// `ptr` must have been returned by a previous call to [`Arc::into_raw`], and the `Arc` must
/// not yet have been destroyed.
unsafe fn container_of(ptr: *const T) -> NonNull<ArcInner<T>> {
- let refcount_layout = Layout::new::<bindings::refcount_t>();
+ let refcount_layout = Layout::new::<Refcount>();
// SAFETY: The caller guarantees that the pointer is valid.
let val_layout = Layout::for_value(unsafe { &*ptr });
// SAFETY: We're computing the layout of a real struct that existed when compiling this
@@ -199,8 +199,7 @@ impl<T> Arc<T> {
pub fn new(contents: T, flags: Flags) -> Result<Self, AllocError> {
// INVARIANT: The refcount is initialised to a non-zero value.
let value = ArcInner {
- // SAFETY: There are no safety requirements for this FFI call.
- refcount: Opaque::new(unsafe { bindings::REFCOUNT_INIT(1) }),
+ refcount: Refcount::new(1),
data: contents,
};
@@ -308,18 +307,15 @@ pub fn into_unique_or_drop(self) -> Option<Pin<UniqueArc<T>>> {
// We will manually manage the refcount in this method, so we disable the destructor.
let me = ManuallyDrop::new(self);
// SAFETY: We own a refcount, so the pointer is still valid.
- let refcount = unsafe { me.ptr.as_ref() }.refcount.get();
+ let refcount = unsafe { &me.ptr.as_ref().refcount };
// If the refcount reaches a non-zero value, then we have destroyed this `Arc` and will
// return without further touching the `Arc`. If the refcount reaches zero, then there are
// no other arcs, and we can create a `UniqueArc`.
- //
- // SAFETY: We own a refcount, so the pointer is not dangling.
- let is_zero = unsafe { bindings::refcount_dec_and_test(refcount) };
+ let is_zero = refcount.dec_and_test();
if is_zero {
- // SAFETY: We have exclusive access to the arc, so we can perform unsynchronized
- // accesses to the refcount.
- unsafe { core::ptr::write(refcount, bindings::REFCOUNT_INIT(1)) };
+ // We have exclusive access to the arc, so we can modify the refcount at will.
+ refcount.set(1);
// INVARIANT: We own the only refcount to this arc, so we may create a `UniqueArc`. We
// must pin the `UniqueArc` because the values was previously in an `Arc`, and they pin
@@ -376,10 +372,10 @@ fn as_ref(&self) -> &T {
impl<T: ?Sized> Clone for Arc<T> {
fn clone(&self) -> Self {
- // INVARIANT: C `refcount_inc` saturates the refcount, so it cannot overflow to zero.
+ // INVARIANT: `Refcount` saturates the refcount, so it cannot overflow to zero.
// SAFETY: By the type invariant, there is necessarily a reference to the object, so it is
// safe to increment the refcount.
- unsafe { bindings::refcount_inc(self.ptr.as_ref().refcount.get()) };
+ unsafe { self.ptr.as_ref().refcount.inc() };
// SAFETY: We just incremented the refcount. This increment is now owned by the new `Arc`.
unsafe { Self::from_inner(self.ptr) }
@@ -388,16 +384,14 @@ fn clone(&self) -> Self {
impl<T: ?Sized> Drop for Arc<T> {
fn drop(&mut self) {
- // SAFETY: By the type invariant, there is necessarily a reference to the object. We cannot
- // touch `refcount` after it's decremented to a non-zero value because another thread/CPU
- // may concurrently decrement it to zero and free it. It is ok to have a raw pointer to
- // freed/invalid memory as long as it is never dereferenced.
- let refcount = unsafe { self.ptr.as_ref() }.refcount.get();
-
// INVARIANT: If the refcount reaches zero, there are no other instances of `Arc`, and
// this instance is being dropped, so the broken invariant is not observable.
- // SAFETY: Also by the type invariant, we are allowed to decrement the refcount.
- let is_zero = unsafe { bindings::refcount_dec_and_test(refcount) };
+ // SAFETY: By the type invariant, there is necessarily a reference to the object.
+ // NOTE: we cannot touch `refcount` after it's decremented to a non-zero value because
+ // another thread/CPU may concurrently decrement it to zero and free it. However it is okay
+ // to have a transient reference to decrement the refcount, see
+ // https://github.com/rust-lang/rust/issues/55005.
+ let is_zero = unsafe { self.ptr.as_ref().refcount.dec_and_test() };
if is_zero {
// The count reached zero, we must free the memory.
//
@@ -649,8 +643,7 @@ pub fn new_uninit(flags: Flags) -> Result<UniqueArc<MaybeUninit<T>>, AllocError>
// INVARIANT: The refcount is initialised to a non-zero value.
let inner = Box::try_init::<AllocError>(
try_init!(ArcInner {
- // SAFETY: There are no safety requirements for this FFI call.
- refcount: Opaque::new(unsafe { bindings::REFCOUNT_INIT(1) }),
+ refcount: Refcount::new(1),
data <- init::uninit::<T, AllocError>(),
}? AllocError),
flags,
--
2.44.1
next prev parent reply other threads:[~2024-10-04 15:53 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-10-04 15:52 [PATCH 0/3] implement `kernel::sync::Refcount` and convert users Gary Guo
2024-10-04 15:52 ` [PATCH 1/3] rust: implement `kernel::sync::Refcount` Gary Guo
2024-10-04 16:34 ` Dirk Behme
2024-10-04 18:51 ` Andreas Hindborg
2024-10-04 19:08 ` Gary Guo
2024-10-04 19:22 ` Andreas Hindborg
2024-10-05 6:04 ` Dirk Behme
2024-10-05 7:40 ` Greg KH
2024-10-05 13:31 ` Gary Guo
2024-10-05 14:26 ` Gary Guo
2024-10-07 12:30 ` Alice Ryhl
2024-10-04 15:52 ` Gary Guo [this message]
2024-10-05 12:06 ` [PATCH 2/3] rust: convert `Arc` to use `Refcount` Alice Ryhl
2024-10-05 13:12 ` Gary Guo
2024-10-04 15:52 ` [PATCH 3/3] rust: block: convert `block::mq` " Gary Guo
2024-10-04 16:43 ` Dirk Behme
2024-10-04 18:05 ` Andreas Hindborg
2024-10-05 15:08 ` Gary Guo
2024-10-05 18:47 ` Andreas Hindborg
2024-10-04 18:34 ` Andreas Hindborg
2024-10-04 18:43 ` Gary Guo
2024-10-04 19:18 ` Andreas Hindborg
2024-10-05 7:47 ` Greg KH
2024-10-05 9:48 ` Andreas Hindborg
2024-10-05 10:09 ` Greg KH
2024-10-05 10:10 ` Peter Zijlstra
2024-10-05 10:57 ` Andreas Hindborg
2024-10-05 11:05 ` Miguel Ojeda
2024-10-05 11:59 ` Alice Ryhl
2024-10-05 13:23 ` Gary Guo
2024-10-05 14:56 ` Andreas Hindborg
2024-10-10 8:39 ` Andreas Hindborg
2024-10-10 9:06 ` Andreas Hindborg
2024-10-10 9:48 ` Benno Lossin
2024-10-10 11:13 ` Andreas Hindborg
2024-10-05 14:51 ` Andreas Hindborg
2024-10-10 8:41 ` Andreas Hindborg
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=20241004155247.2210469-3-gary@garyguo.net \
--to=gary@garyguo.net \
--cc=a.hindborg@kernel.org \
--cc=alex.gaynor@gmail.com \
--cc=alexmantel93@mailbox.org \
--cc=aliceryhl@google.com \
--cc=benno.lossin@proton.me \
--cc=bjorn3_gh@protonmail.com \
--cc=boqun.feng@gmail.com \
--cc=kernel@valentinobst.de \
--cc=linux-kernel@vger.kernel.org \
--cc=mark.rutland@arm.com \
--cc=ojeda@kernel.org \
--cc=peterz@infradead.org \
--cc=rust-for-linux@vger.kernel.org \
--cc=tmgross@umich.edu \
--cc=walmeida@microsoft.com \
--cc=will@kernel.org \
--cc=yakoyoku@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;
as well as URLs for NNTP newsgroup(s).