mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Daniel Almeida <daniel.almeida@collabora.com>
To: "Rafael J. Wysocki" <rafael@kernel.org>,
	"Viresh Kumar" <viresh.kumar@linaro.org>,
	"Danilo Krummrich" <dakr@kernel.org>,
	"Alice Ryhl" <aliceryhl@google.com>,
	"Daniel Almeida" <daniel.almeida@collabora.com>,
	"David Airlie" <airlied@gmail.com>,
	"Simona Vetter" <simona@ffwll.ch>,
	"Michal Wilczynski" <m.wilczynski@samsung.com>,
	"Drew Fustini" <fustini@kernel.org>,
	"Guo Ren" <guoren@kernel.org>, "Fu Wei" <wefu@redhat.com>,
	"Uwe Kleine-König" <ukleinek@kernel.org>,
	"Stephen Boyd" <sboyd@kernel.org>,
	"Brian Masney" <bmasney+clk@redhat.com>,
	"Jerome Brunet" <jbrunet+clk@baylibre.com>,
	"Miguel Ojeda" <ojeda@kernel.org>,
	"Boqun Feng" <boqun@kernel.org>, "Gary Guo" <gary@garyguo.net>,
	"Björn Roy Baron" <bjorn3_gh@protonmail.com>,
	"Benno Lossin" <lossin@kernel.org>,
	"Andreas Hindborg" <a.hindborg@kernel.org>,
	"Trevor Gross" <tmgross@umich.edu>,
	"Tamir Duberstein" <tamird@kernel.org>,
	"Alexandre Courbot" <acourbot@nvidia.com>
Cc: linux-pm@vger.kernel.org, linux-kernel@vger.kernel.org,
	dri-devel@lists.freedesktop.org, linux-riscv@lists.infradead.org,
	linux-pwm@vger.kernel.org, linux-clk@vger.kernel.org,
	rust-for-linux@vger.kernel.org,
	"Boris Brezillon" <boris.brezillon@collabora.com>,
	"Onur Özkan" <work@onurozkan.dev>,
	"Maurice Hieronymus" <mhi@mailbox.org>
Subject: [PATCH v6 6/7] rust: clk: add `ExclusiveClk`
Date: Tue, 13 Oct 2026 14:07:45 -0300	[thread overview]
Message-ID: <20261013-clk-type-state-v6-6-a5e9977b3509@collabora.com> (raw)
In-Reply-To: <20261013-clk-type-state-v6-0-a5e9977b3509@collabora.com>

Some drivers derive their configuration from a clock rate and cannot
tolerate the rate changing underneath them. In C, they claim rate
exclusivity with clk_rate_exclusive_get(), after which the clock
framework refuses rate changes from other consumers until
clk_rate_exclusive_put() is called.

Add ExclusiveClk<T>, which takes a Clk<T> in any state and holds rate
exclusivity until it is dropped. ExclusiveClk::new() calls
clk_rate_exclusive_get(), and dropping the ExclusiveClk or calling
ExclusiveClk::into_inner() calls clk_rate_exclusive_put().
ExclusiveClk dereferences to the Clk<T>, so the rate can still be read
and changed through it. Since a Clk cannot be cloned, the ExclusiveClk
is the only handle to the consumer that holds the exclusivity.

pwm-th1520 is the first user, in the next patch.

Co-developed-by: Maurice Hieronymus <mhi@mailbox.org>
Signed-off-by: Maurice Hieronymus <mhi@mailbox.org>
Signed-off-by: Daniel Almeida <daniel.almeida@collabora.com>
---
 rust/kernel/clk.rs | 100 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 100 insertions(+)

diff --git a/rust/kernel/clk.rs b/rust/kernel/clk.rs
index e2587c613182..8d7ba0fae459 100644
--- a/rust/kernel/clk.rs
+++ b/rust/kernel/clk.rs
@@ -846,6 +846,106 @@ fn drop(&mut self) {
             unsafe { bindings::clk_disable_unprepare(self.clk.as_raw()) };
         }
     }
+
+    /// A [`Clk`] with exclusive control over its rate.
+    ///
+    /// While an [`ExclusiveClk`] exists, no other consumer of the same clock
+    /// may change its rate. The rate can still be changed through the
+    /// [`ExclusiveClk`] itself, which dereferences to the [`Clk`] in its
+    /// current state.
+    ///
+    /// The [`Clk`] is moved into the [`ExclusiveClk`], and a [`Clk`] cannot be
+    /// cloned, so only the [`ExclusiveClk`] refers to the consumer handle that
+    /// holds the exclusivity.
+    ///
+    /// # Invariants
+    ///
+    /// `clk_rate_exclusive_get()` has been called on the wrapped [`Clk`], and
+    /// the matching `clk_rate_exclusive_put()` has not been called yet.
+    ///
+    /// # Examples
+    ///
+    /// ```
+    /// use kernel::clk::{
+    ///     Clk,
+    ///     Enabled,
+    ///     ExclusiveClk,
+    ///     Hertz,
+    /// };
+    /// use kernel::device::{
+    ///     Bound,
+    ///     Device,
+    /// };
+    ///
+    /// fn lock_rate(dev: &Device<Bound>) -> Result<ExclusiveClk<Enabled>> {
+    ///     let clk = Clk::<Enabled>::get(dev, None)?;
+    ///     clk.set_rate(Hertz::from_mhz(24))?;
+    ///
+    ///     ExclusiveClk::new(clk)
+    /// }
+    /// ```
+    pub struct ExclusiveClk<T: ClkState>(Clk<T>);
+
+    impl<T: ClkState> ExclusiveClk<T> {
+        /// Takes exclusive control over the rate of `clk`.
+        ///
+        /// No other consumer of the clock may change its rate until the
+        /// returned [`ExclusiveClk`] is dropped or [`ExclusiveClk::into_inner`]
+        /// is called. This may sleep.
+        ///
+        /// Equivalent to the kernel's [`clk_rate_exclusive_get`] API.
+        ///
+        /// [`clk_rate_exclusive_get`]:
+        /// https://docs.kernel.org/core-api/kernel-api.html#c.clk_rate_exclusive_get
+        #[inline]
+        pub fn new(clk: Clk<T>) -> Result<Self> {
+            // SAFETY: By the type invariants of `Clk`, `clk.as_raw()` is a valid
+            // argument for `clk_rate_exclusive_get()`.
+            to_result(unsafe { bindings::clk_rate_exclusive_get(clk.as_raw()) })?;
+
+            // INVARIANT: `clk_rate_exclusive_get()` succeeded, and the matching
+            // `clk_rate_exclusive_put()` is called when the `ExclusiveClk` is
+            // dropped or consumed by `into_inner()`.
+            Ok(Self(clk))
+        }
+
+        /// Releases the exclusive control over the rate and returns the [`Clk`].
+        ///
+        /// Equivalent to the kernel's [`clk_rate_exclusive_put`] API.
+        ///
+        /// [`clk_rate_exclusive_put`]:
+        /// https://docs.kernel.org/core-api/kernel-api.html#c.clk_rate_exclusive_put
+        #[inline]
+        pub fn into_inner(self) -> Clk<T> {
+            let this = ManuallyDrop::new(self);
+
+            // SAFETY: By the type invariants, `clk_rate_exclusive_get()` was
+            // called on `this.0` and not yet balanced, so this balances it.
+            unsafe { bindings::clk_rate_exclusive_put(this.0.as_raw()) };
+
+            // SAFETY: `this` is wrapped in `ManuallyDrop`, so the `Clk` is moved
+            // out exactly once and the `Drop` of `ExclusiveClk` does not run.
+            unsafe { ptr::read(&this.0) }
+        }
+    }
+
+    impl<T: ClkState> Deref for ExclusiveClk<T> {
+        type Target = Clk<T>;
+
+        #[inline]
+        fn deref(&self) -> &Clk<T> {
+            &self.0
+        }
+    }
+
+    impl<T: ClkState> Drop for ExclusiveClk<T> {
+        #[inline]
+        fn drop(&mut self) {
+            // SAFETY: By the type invariants, `clk_rate_exclusive_get()` was
+            // called on `self.0` and not yet balanced, so this balances it.
+            unsafe { bindings::clk_rate_exclusive_put(self.0.as_raw()) };
+        }
+    }
 }
 
 #[cfg(CONFIG_COMMON_CLK)]

-- 
2.56.0


  parent reply	other threads:[~2026-10-13 17:09 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-13 17:07 [PATCH v6 0/7] Clk improvements Daniel Almeida
2026-10-13 17:07 ` [PATCH v6 1/7] rust: clk: use 'kernel vertical style' for imports Daniel Almeida
2026-10-13 17:07 ` [PATCH v6 2/7] rust: clk: replace `OptionalClk` with `Clk::get_optional()` Daniel Almeida
2026-10-13 17:07 ` [PATCH v6 3/7] rust: clk: use the type-state pattern Daniel Almeida
2026-10-13 17:07 ` [PATCH v6 4/7] rust: clk: add guards for scoped prepare and enable Daniel Almeida
2026-10-13 17:07 ` [PATCH v6 5/7] rust: clk: add devres-managed clks Daniel Almeida
2026-10-13 17:07 ` Daniel Almeida [this message]
2026-10-13 17:14   ` [PATCH v6 6/7] rust: clk: add `ExclusiveClk` Daniel Almeida
2026-10-13 17:07 ` [PATCH v6 7/7] pwm: th1520: Lock clock rate with clk_rate_exclusive_get Daniel Almeida

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=20261013-clk-type-state-v6-6-a5e9977b3509@collabora.com \
    --to=daniel.almeida@collabora.com \
    --cc=a.hindborg@kernel.org \
    --cc=acourbot@nvidia.com \
    --cc=airlied@gmail.com \
    --cc=aliceryhl@google.com \
    --cc=bjorn3_gh@protonmail.com \
    --cc=bmasney+clk@redhat.com \
    --cc=boqun@kernel.org \
    --cc=boris.brezillon@collabora.com \
    --cc=dakr@kernel.org \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=fustini@kernel.org \
    --cc=gary@garyguo.net \
    --cc=guoren@kernel.org \
    --cc=jbrunet+clk@baylibre.com \
    --cc=linux-clk@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pm@vger.kernel.org \
    --cc=linux-pwm@vger.kernel.org \
    --cc=linux-riscv@lists.infradead.org \
    --cc=lossin@kernel.org \
    --cc=m.wilczynski@samsung.com \
    --cc=mhi@mailbox.org \
    --cc=ojeda@kernel.org \
    --cc=rafael@kernel.org \
    --cc=rust-for-linux@vger.kernel.org \
    --cc=sboyd@kernel.org \
    --cc=simona@ffwll.ch \
    --cc=tamird@kernel.org \
    --cc=tmgross@umich.edu \
    --cc=ukleinek@kernel.org \
    --cc=viresh.kumar@linaro.org \
    --cc=wefu@redhat.com \
    --cc=work@onurozkan.dev \
    /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

all inboxes | Powered by JetHome®