[PATCH] can: m_can_platform: don't call m_can_class_suspend in runtime suspend

[PATCH] can: m_can_platform: don't call m_can_class_suspend in runtime suspend

[Lucas Stach]

0704c5743694 (can: m_can_platform: remove unnecessary m_can_class_resume()
call) removed the m_can_class_resume() call in the runtime resume path
to get rid of a infinite recursion, so the runtime resume now only handles
the device clocks. Unfortunately it did not remove the complementary
m_can_class_suspend() call in the runtime suspend function, so those paths
are now unbalanced, which causes the pinctrl state to get stuck on the
"sleep" state, which breaks all CAN functionality on SoCs where this state
is defined. Remove the m_can_class_suspend() call to fix this.

Fixes: 0704c5743694 (can: m_can_platform: remove unnecessary
                     m_can_class_resume() call)
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
---
 drivers/net/can/m_can/m_can_platform.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/net/can/m_can/m_can_platform.c b/drivers/net/can/m_can/m_can_platform.c
index 38ea5e600fb8..e6d0cb9ee02f 100644
--- a/drivers/net/can/m_can/m_can_platform.c
+++ b/drivers/net/can/m_can/m_can_platform.c
@@ -144,8 +144,6 @@ static int __maybe_unused m_can_runtime_suspend(struct device *dev)
 	struct net_device *ndev = dev_get_drvdata(dev);
 	struct m_can_classdev *mcan_class = netdev_priv(ndev);
 
-	m_can_class_suspend(dev);
-
 	clk_disable_unprepare(mcan_class->cclk);
 	clk_disable_unprepare(mcan_class->hclk);
 
-- 
2.20.1

[PATCH] can: m_can_platform: don't call m_can_class_suspend in runtime suspend

[Lucas Stach]

0704c5743694 (can: m_can_platform: remove unnecessary m_can_class_resume()
call) removed the m_can_class_resume() call in the runtime resume path
to get rid of a infinite recursion, so the runtime resume now only handles
the device clocks. Unfortunately it did not remove the complementary
m_can_class_suspend() call in the runtime suspend function, so those paths
are now unbalanced, which causes the pinctrl state to get stuck on the
"sleep" state, which breaks all CAN functionality on SoCs where this state
is defined. Remove the m_can_class_suspend() call to fix this.

Fixes: 0704c5743694 (can: m_can_platform: remove unnecessary
                     m_can_class_resume() call)
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
---
 drivers/net/can/m_can/m_can_platform.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/net/can/m_can/m_can_platform.c b/drivers/net/can/m_can/m_can_platform.c
index 38ea5e600fb8..e6d0cb9ee02f 100644
--- a/drivers/net/can/m_can/m_can_platform.c
+++ b/drivers/net/can/m_can/m_can_platform.c
@@ -144,8 +144,6 @@ static int __maybe_unused m_can_runtime_suspend(struct device *dev)
 	struct net_device *ndev = dev_get_drvdata(dev);
 	struct m_can_classdev *mcan_class = netdev_priv(ndev);
 
-	m_can_class_suspend(dev);
-
 	clk_disable_unprepare(mcan_class->cclk);
 	clk_disable_unprepare(mcan_class->hclk);
 
-- 
2.20.1

Re: [PATCH] can: m_can_platform: don't call m_can_class_suspend in runtime suspend

[Dan Murphy]

Lucase

On 8/11/20 3:15 AM, Lucas Stach wrote:
> 0704c5743694 (can: m_can_platform: remove unnecessary m_can_class_resume()
> call) removed the m_can_class_resume() call in the runtime resume path
> to get rid of a infinite recursion, so the runtime resume now only handles
> the device clocks. Unfortunately it did not remove the complementary
> m_can_class_suspend() call in the runtime suspend function, so those paths
> are now unbalanced, which causes the pinctrl state to get stuck on the
> "sleep" state, which breaks all CAN functionality on SoCs where this state
> is defined. Remove the m_can_class_suspend() call to fix this.
>
> Fixes: 0704c5743694 (can: m_can_platform: remove unnecessary
>                       m_can_class_resume() call)

Not sure about this wrap around for the fixes in the commit msg.

Otherwise

Acked-by: Dan Murphy <dmurphy@ti.com>

> Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
> ---
>   drivers/net/can/m_can/m_can_platform.c | 2 --
>   1 file changed, 2 deletions(-)
>
> diff --git a/drivers/net/can/m_can/m_can_platform.c b/drivers/net/can/m_can/m_can_platform.c
> index 38ea5e600fb8..e6d0cb9ee02f 100644
> --- a/drivers/net/can/m_can/m_can_platform.c
> +++ b/drivers/net/can/m_can/m_can_platform.c
> @@ -144,8 +144,6 @@ static int __maybe_unused m_can_runtime_suspend(struct device *dev)
>   	struct net_device *ndev = dev_get_drvdata(dev);
>   	struct m_can_classdev *mcan_class = netdev_priv(ndev);
>   
> -	m_can_class_suspend(dev);
> -
>   	clk_disable_unprepare(mcan_class->cclk);
>   	clk_disable_unprepare(mcan_class->hclk);
>   

pull-request: can 2020-10-06

[Marc Kleine-Budde]

Hello David,

if not too late, I have a pull request of one patch.

The patch is by Lucas Stach and fixes m_can driver by removing an erroneous
call to m_can_class_suspend() in runtime suspend. Which causes the pinctrl
state to get stuck on the "sleep" state, which breaks all CAN functionality on
SoCs where this state is defined.

regards,
Marc

---

The following changes since commit d91dc434f2baa592e9793597421231174d57bbbf:

  Merge tag 'rxrpc-fixes-20201005' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs (2020-10-06 06:18:20 -0700)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can.git tags/linux-can-fixes-for-5.9-20201006

for you to fetch changes up to 81f1f5ae8b3cbd54fdd994c9e9aacdb7b414a802:

  can: m_can_platform: don't call m_can_class_suspend in runtime suspend (2020-10-06 23:29:30 +0200)

----------------------------------------------------------------
linux-can-fixes-for-5.9-20201006

----------------------------------------------------------------
Lucas Stach (1):
      can: m_can_platform: don't call m_can_class_suspend in runtime suspend

 drivers/net/can/m_can/m_can_platform.c | 2 --
 1 file changed, 2 deletions(-)

[PATCH] can: m_can_platform: don't call m_can_class_suspend in runtime suspend

[Marc Kleine-Budde]

From: Lucas Stach <l.stach@pengutronix.de>

    0704c5743694 can: m_can_platform: remove unnecessary m_can_class_resume() call

removed the m_can_class_resume() call in the runtime resume path to get
rid of a infinite recursion, so the runtime resume now only handles the device
clocks.

Unfortunately it did not remove the complementary m_can_class_suspend() call in
the runtime suspend function, so those paths are now unbalanced, which causes
the pinctrl state to get stuck on the "sleep" state, which breaks all CAN
functionality on SoCs where this state is defined. Remove the
m_can_class_suspend() call to fix this.

Fixes: 0704c5743694 can: m_can_platform: remove unnecessary m_can_class_resume() call
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Link: https://lore.kernel.org/r/20200811081545.19921-1-l.stach@pengutronix.de
Acked-by: Dan Murphy <dmurphy@ti.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
---
 drivers/net/can/m_can/m_can_platform.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/net/can/m_can/m_can_platform.c b/drivers/net/can/m_can/m_can_platform.c
index 38ea5e600fb8..e6d0cb9ee02f 100644
--- a/drivers/net/can/m_can/m_can_platform.c
+++ b/drivers/net/can/m_can/m_can_platform.c
@@ -144,8 +144,6 @@ static int __maybe_unused m_can_runtime_suspend(struct device *dev)
 	struct net_device *ndev = dev_get_drvdata(dev);
 	struct m_can_classdev *mcan_class = netdev_priv(ndev);
 
-	m_can_class_suspend(dev);
-
 	clk_disable_unprepare(mcan_class->cclk);
 	clk_disable_unprepare(mcan_class->hclk);
 
-- 
2.28.0