[PATCH v4 0/3] Implement Send and Sync for clk

[PATCH v4 0/3] Implement Send and Sync for clk

[Alice Ryhl]

The Clk type is thread-safe, so let's mark it as thread-safe in the type
system. This lets us get rid of hacks in drivers.

For Stephen's clk tree.

Signed-off-by: Alice Ryhl <aliceryhl@google.com>
---
Changes in v4:
- Pick up Reviewed-by tags.
- Link to v3: https://lore.kernel.org/r/20251218-clk-send-sync-v3-0-e48b2e2f1eac@google.com

Changes in v3:
- Rebase on v6.19-rc1.
- Pick up tags.
- Add fix for pwm driver as well.
- Link to v2: https://lore.kernel.org/r/20251020-clk-send-sync-v2-0-44ab533ae084@google.com

Changes in v2:
- Rebase on v6.18-rc1.
- Add patch to tyr driver.
- Link to v1: https://lore.kernel.org/r/20250904-clk-send-sync-v1-1-48d023320eb8@google.com

---
Alice Ryhl (3):
      rust: clk: implement Send and Sync
      tyr: remove impl Send/Sync for TyrData
      pwm: th1520: remove impl Send/Sync for Th1520PwmDriverData

 drivers/gpu/drm/tyr/driver.rs | 12 ------------
 drivers/pwm/pwm_th1520.rs     | 15 ---------------
 rust/kernel/clk.rs            |  7 +++++++
 3 files changed, 7 insertions(+), 27 deletions(-)
---
base-commit: 8f0b4cce4481fb22653697cced8d0d04027cb1e8
change-id: 20250904-clk-send-sync-3cfa7f4e1ce2

Best regards,
-- 
Alice Ryhl <aliceryhl@google.com>

[PATCH v4 1/3] rust: clk: implement Send and Sync

[Alice Ryhl]

These traits are required for drivers to embed the Clk type in their own
data structures because driver data structures are usually required to
be Send. Since the Clk type is thread-safe, implement the relevant
traits.

Reviewed-by: Daniel Almeida <daniel.almeida@collabora.com>
Reviewed-by: Danilo Krummrich <dakr@kernel.org>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Boqun Feng <boqun.feng@gmail.com>
Reviewed-by: Gary Guo <gary@garyguo.net>
Signed-off-by: Alice Ryhl <aliceryhl@google.com>
---
 rust/kernel/clk.rs | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/rust/kernel/clk.rs b/rust/kernel/clk.rs
index c1cfaeaa36a22be2e3180b1e9142bb608ab276ea..d192fbd97861212d738d24510eebcd99c9177f2c 100644
--- a/rust/kernel/clk.rs
+++ b/rust/kernel/clk.rs
@@ -129,6 +129,13 @@ mod common_clk {
     #[repr(transparent)]
     pub struct Clk(*mut bindings::clk);
 
+    // SAFETY: It is safe to call `clk_put` on another thread than where `clk_get` was called.
+    unsafe impl Send for Clk {}
+
+    // SAFETY: It is safe to call any combination of the `&self` methods in parallel, as the
+    // methods are synchronized internally.
+    unsafe impl Sync for Clk {}
+
     impl Clk {
         /// Gets [`Clk`] corresponding to a [`Device`] and a connection id.
         ///

-- 
2.52.0.457.g6b5491de43-goog

[PATCH v4 2/3] tyr: remove impl Send/Sync for TyrData

[Alice Ryhl]

Now that clk implements Send and Sync, we no longer need to manually
implement these traits for TyrData. Thus remove the implementations.

The comment also mentions the regulator. However, the regulator had the
traits added in commit 9a200cbdb543 ("rust: regulator: implement Send
and Sync for Regulator<T>"), which is already in mainline.

Reviewed-by: Danilo Krummrich <dakr@kernel.org>
Reviewed-by: Boqun Feng <boqun.feng@gmail.com>
Reviewed-by: Gary Guo <gary@garyguo.net>
Reviewed-by: Daniel Almeida <daniel.almeida@collabora.com>
Signed-off-by: Alice Ryhl <aliceryhl@google.com>
---
 drivers/gpu/drm/tyr/driver.rs | 12 ------------
 1 file changed, 12 deletions(-)

diff --git a/drivers/gpu/drm/tyr/driver.rs b/drivers/gpu/drm/tyr/driver.rs
index 0389c558c0367522471ea78fcf72a6b58c4a3650..09711fb7fe0b1c83b72bffba06f5a76c53244f4d 100644
--- a/drivers/gpu/drm/tyr/driver.rs
+++ b/drivers/gpu/drm/tyr/driver.rs
@@ -53,18 +53,6 @@ pub(crate) struct TyrData {
     pub(crate) gpu_info: GpuInfo,
 }
 
-// Both `Clk` and `Regulator` do not implement `Send` or `Sync`, but they
-// should. There are patches on the mailing list to address this, but they have
-// not landed yet.
-//
-// For now, add this workaround so that this patch compiles with the promise
-// that it will be removed in a future patch.
-//
-// SAFETY: This will be removed in a future patch.
-unsafe impl Send for TyrData {}
-// SAFETY: This will be removed in a future patch.
-unsafe impl Sync for TyrData {}
-
 fn issue_soft_reset(dev: &Device<Bound>, iomem: &Devres<IoMem>) -> Result {
     regs::GPU_CMD.write(dev, iomem, regs::GPU_CMD_SOFT_RESET)?;
 

-- 
2.52.0.457.g6b5491de43-goog

[PATCH v4 3/3] pwm: th1520: remove impl Send/Sync for Th1520PwmDriverData

[Alice Ryhl]

Now that clk implements Send and Sync, we no longer need to manually
implement these traits for Th1520PwmDriverData. Thus remove the
implementations.

Reviewed-by: Gary Guo <gary@garyguo.net>
Reviewed-by: Daniel Almeida <daniel.almeida@collabora.com>
Signed-off-by: Alice Ryhl <aliceryhl@google.com>
---
Cc: linux-riscv@lists.infradead.org
---
 drivers/pwm/pwm_th1520.rs | 15 ---------------
 1 file changed, 15 deletions(-)

diff --git a/drivers/pwm/pwm_th1520.rs b/drivers/pwm/pwm_th1520.rs
index e3b7e77356fc2492077c519073e861beb3e44df9..043dc4dbc6232020195c7b73fad302bbb69652df 100644
--- a/drivers/pwm/pwm_th1520.rs
+++ b/drivers/pwm/pwm_th1520.rs
@@ -97,21 +97,6 @@ struct Th1520PwmDriverData {
     clk: Clk,
 }
 
-// This `unsafe` implementation is a temporary necessity because the underlying `kernel::clk::Clk`
-// type does not yet expose `Send` and `Sync` implementations. This block should be removed
-// as soon as the clock abstraction provides these guarantees directly.
-// TODO: Remove those unsafe impl's when Clk will support them itself.
-
-// SAFETY: The `devres` framework requires the driver's private data to be `Send` and `Sync`.
-// We can guarantee this because the PWM core synchronizes all callbacks, preventing concurrent
-// access to the contained `iomem` and `clk` resources.
-unsafe impl Send for Th1520PwmDriverData {}
-
-// SAFETY: The same reasoning applies as for `Send`. The PWM core's synchronization
-// guarantees that it is safe for multiple threads to have shared access (`&self`)
-// to the driver data during callbacks.
-unsafe impl Sync for Th1520PwmDriverData {}
-
 impl pwm::PwmOps for Th1520PwmDriverData {
     type WfHw = Th1520WfHw;
 

-- 
2.52.0.457.g6b5491de43-goog

Re: [PATCH v4 3/3] pwm: th1520: remove impl Send/Sync for Th1520PwmDriverData

[Michal Wilczynski]

On 1/13/26 16:12, Alice Ryhl wrote:
> Now that clk implements Send and Sync, we no longer need to manually
> implement these traits for Th1520PwmDriverData. Thus remove the
> implementations.
> 
> Reviewed-by: Gary Guo <gary@garyguo.net>
> Reviewed-by: Daniel Almeida <daniel.almeida@collabora.com>
> Signed-off-by: Alice Ryhl <aliceryhl@google.com>
> ---
> Cc: linux-riscv@lists.infradead.org
> ---
>  drivers/pwm/pwm_th1520.rs | 15 ---------------
>  1 file changed, 15 deletions(-)
> 
> diff --git a/drivers/pwm/pwm_th1520.rs b/drivers/pwm/pwm_th1520.rs
> index e3b7e77356fc2492077c519073e861beb3e44df9..043dc4dbc6232020195c7b73fad302bbb69652df 100644
> --- a/drivers/pwm/pwm_th1520.rs
> +++ b/drivers/pwm/pwm_th1520.rs
> @@ -97,21 +97,6 @@ struct Th1520PwmDriverData {
>      clk: Clk,
>  }
>  
> -// This `unsafe` implementation is a temporary necessity because the underlying `kernel::clk::Clk`
> -// type does not yet expose `Send` and `Sync` implementations. This block should be removed
> -// as soon as the clock abstraction provides these guarantees directly.
> -// TODO: Remove those unsafe impl's when Clk will support them itself.
> -
> -// SAFETY: The `devres` framework requires the driver's private data to be `Send` and `Sync`.
> -// We can guarantee this because the PWM core synchronizes all callbacks, preventing concurrent
> -// access to the contained `iomem` and `clk` resources.
> -unsafe impl Send for Th1520PwmDriverData {}
> -
> -// SAFETY: The same reasoning applies as for `Send`. The PWM core's synchronization
> -// guarantees that it is safe for multiple threads to have shared access (`&self`)
> -// to the driver data during callbacks.
> -unsafe impl Sync for Th1520PwmDriverData {}
> -
>  impl pwm::PwmOps for Th1520PwmDriverData {
>      type WfHw = Th1520WfHw;
>  
> 

I thought this was already merged :-).

Reviewed-by: Michal Wilczynski <m.wilczynski@samsung.com>

Re: [PATCH v4 3/3] pwm: th1520: remove impl Send/Sync for Th1520PwmDriverData

[Uwe Kleine-König]

[-- Attachment #1: Type: text/plain, Size: 952 bytes --]

Hello,

On Mon, Jan 19, 2026 at 10:45:56PM +0100, Michal Wilczynski wrote:
> On 1/13/26 16:12, Alice Ryhl wrote:
> > Now that clk implements Send and Sync, we no longer need to manually
> > implement these traits for Th1520PwmDriverData. Thus remove the
> > implementations.
> 
> I thought this was already merged :-).
> 
> Reviewed-by: Michal Wilczynski <m.wilczynski@samsung.com>

If I understand correctly this patch 3 depends on the first patch of
this series so I cannot pick it up via the pwm tree *now*. There is
another patch pending for the th1520 PWM driver, but as of now git seems
to cope well when merging the pwm's tree for-next with this patch.  So
it's fine for me if the series is picked up for 6.20-rc1 via the clock
tree.

Acked-by: Uwe Kleine-König <ukleinek@kernel.org>

(Nitpick: If it was me who picked up patch 3 I would have capitalized
the "remove" in the Subject for consistency.)

Best regards
Uwe

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: [PATCH v4 3/3] pwm: th1520: remove impl Send/Sync for Th1520PwmDriverData

[Uwe Kleine-König]

[-- Attachment #1: Type: text/plain, Size: 1063 bytes --]

Hello,

On Tue, Jan 20, 2026 at 09:48:48AM +0100, Uwe Kleine-König wrote:
> On Mon, Jan 19, 2026 at 10:45:56PM +0100, Michal Wilczynski wrote:
> > On 1/13/26 16:12, Alice Ryhl wrote:
> > > Now that clk implements Send and Sync, we no longer need to manually
> > > implement these traits for Th1520PwmDriverData. Thus remove the
> > > implementations.
> > 
> > I thought this was already merged :-).
> > 
> > Reviewed-by: Michal Wilczynski <m.wilczynski@samsung.com>
> 
> If I understand correctly this patch 3 depends on the first patch of
> this series so I cannot pick it up via the pwm tree *now*. There is
> another patch pending for the th1520 PWM driver, but as of now git seems
> to cope well when merging the pwm's tree for-next with this patch.  So
> it's fine for me if the series is picked up for 6.20-rc1 via the clock
> tree.

That didn't happen (with s/6.20/7.0/), so this patch is still open in my
todo list. Is the first patch still considered for the clk subsystem?
What is the plan for the pwm patch?

Best regards
Uwe

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: [PATCH v4 3/3] pwm: th1520: remove impl Send/Sync for Th1520PwmDriverData

[Alice Ryhl]

On Thu, Mar 05, 2026 at 09:24:33AM +0100, Uwe Kleine-König wrote:
> Hello,
> 
> On Tue, Jan 20, 2026 at 09:48:48AM +0100, Uwe Kleine-König wrote:
> > On Mon, Jan 19, 2026 at 10:45:56PM +0100, Michal Wilczynski wrote:
> > > On 1/13/26 16:12, Alice Ryhl wrote:
> > > > Now that clk implements Send and Sync, we no longer need to manually
> > > > implement these traits for Th1520PwmDriverData. Thus remove the
> > > > implementations.
> > > 
> > > I thought this was already merged :-).
> > > 
> > > Reviewed-by: Michal Wilczynski <m.wilczynski@samsung.com>
> > 
> > If I understand correctly this patch 3 depends on the first patch of
> > this series so I cannot pick it up via the pwm tree *now*. There is
> > another patch pending for the th1520 PWM driver, but as of now git seems
> > to cope well when merging the pwm's tree for-next with this patch.  So
> > it's fine for me if the series is picked up for 6.20-rc1 via the clock
> > tree.
> 
> That didn't happen (with s/6.20/7.0/), so this patch is still open in my
> todo list. Is the first patch still considered for the clk subsystem?
> What is the plan for the pwm patch?

It's been on my todo list to try and contact Stephen to see if he can
take these patches, or if he is okay with them going through another
tree, but I have not gotten around to it. He has not replied on the list
to any version of this series.

Alice