[PATCH] can: flexcan: disable transceiver during system PM

[PATCH] can: flexcan: disable transceiver during system PM

[haibo.chen]

From: Haibo Chen <haibo.chen@nxp.com>

During system PM, if no wakeup requirement, disable transceiver to
save power.

Signed-off-by: Haibo Chen <haibo.chen@nxp.com>
---
 drivers/net/can/flexcan/flexcan-core.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/net/can/flexcan/flexcan-core.c b/drivers/net/can/flexcan/flexcan-core.c
index b347a1c93536..74f3137295fb 100644
--- a/drivers/net/can/flexcan/flexcan-core.c
+++ b/drivers/net/can/flexcan/flexcan-core.c
@@ -2292,6 +2292,7 @@ static int __maybe_unused flexcan_suspend(struct device *device)
 				return err;
 
 			flexcan_chip_interrupts_disable(dev);
+			flexcan_transceiver_disable(priv);
 
 			err = pinctrl_pm_select_sleep_state(device);
 			if (err)
@@ -2325,6 +2326,10 @@ static int __maybe_unused flexcan_resume(struct device *device)
 			if (err)
 				return err;
 
+			err = flexcan_transceiver_enable(priv);
+			if (err)
+				return err;
+
 			err = flexcan_chip_start(dev);
 			if (err)
 				return err;
-- 
2.34.1

Re: [PATCH] can: flexcan: disable transceiver during system PM

[Vincent Mailhol]

On 04/03/2025 at 12:37, haibo.chen@nxp.com wrote:
> From: Haibo Chen <haibo.chen@nxp.com>
> 
> During system PM, if no wakeup requirement, disable transceiver to
> save power.
> 
> Signed-off-by: Haibo Chen <haibo.chen@nxp.com>> ---
>  drivers/net/can/flexcan/flexcan-core.c | 5 +++++
>  1 file changed, 5 insertions(+)
> 
> diff --git a/drivers/net/can/flexcan/flexcan-core.c b/drivers/net/can/flexcan/flexcan-core.c
> index b347a1c93536..74f3137295fb 100644
> --- a/drivers/net/can/flexcan/flexcan-core.c
> +++ b/drivers/net/can/flexcan/flexcan-core.c
> @@ -2292,6 +2292,7 @@ static int __maybe_unused flexcan_suspend(struct device *device)
>  				return err;
>  
>  			flexcan_chip_interrupts_disable(dev);
> +			flexcan_transceiver_disable(priv);

flexcan_transceiver_disable() may return an error. Can you check its
return value?

>  			err = pinctrl_pm_select_sleep_state(device);
>  			if (err)
> @@ -2325,6 +2326,10 @@ static int __maybe_unused flexcan_resume(struct device *device)
>  			if (err)
>  				return err;
>  
> +			err = flexcan_transceiver_enable(priv);
> +			if (err)
> +				return err;
> +
>  			err = flexcan_chip_start(dev);
>  			if (err)
>  				return err;

If flexcan_chip_start() fails, shouldn't you have a clean-up branch that
will flexcan_transceiver_disable()? Or do you consider that it is
acceptable to keep the transceiver enabled if an error occurs?

Speaking of which, if flexcan_resume() fails, priv->can.state would
still be set to CAN_STATE_ERROR_ACTIVE.


Yours sincerely,
Vincent Mailhol

RE: [PATCH] can: flexcan: disable transceiver during system PM

[Bough Chen]

> -----Original Message-----
> From: Vincent Mailhol <mailhol.vincent@wanadoo.fr>
> Sent: 2025年3月4日 12:06
> To: Bough Chen <haibo.chen@nxp.com>; mkl@pengutronix.de
> Cc: Ciprian Marian Costea (OSS) <ciprianmarian.costea@oss.nxp.com>;
> u.kleine-koenig@baylibre.com; linux-can@vger.kernel.org;
> linux-kernel@vger.kernel.org; imx@lists.linux.dev
> Subject: Re: [PATCH] can: flexcan: disable transceiver during system PM
> 
> On 04/03/2025 at 12:37, haibo.chen@nxp.com wrote:
> > From: Haibo Chen <haibo.chen@nxp.com>
> >
> > During system PM, if no wakeup requirement, disable transceiver to
> > save power.
> >
> > Signed-off-by: Haibo Chen <haibo.chen@nxp.com>> ---
> > drivers/net/can/flexcan/flexcan-core.c | 5 +++++
> >  1 file changed, 5 insertions(+)
> >
> > diff --git a/drivers/net/can/flexcan/flexcan-core.c
> > b/drivers/net/can/flexcan/flexcan-core.c
> > index b347a1c93536..74f3137295fb 100644
> > --- a/drivers/net/can/flexcan/flexcan-core.c
> > +++ b/drivers/net/can/flexcan/flexcan-core.c
> > @@ -2292,6 +2292,7 @@ static int __maybe_unused flexcan_suspend(struct
> device *device)
> >  				return err;
> >
> >  			flexcan_chip_interrupts_disable(dev);
> > +			flexcan_transceiver_disable(priv);
> 
> flexcan_transceiver_disable() may return an error. Can you check its return
> value?

Yes, sure.

> 
> >  			err = pinctrl_pm_select_sleep_state(device);
> >  			if (err)
> > @@ -2325,6 +2326,10 @@ static int __maybe_unused flexcan_resume(struct
> device *device)
> >  			if (err)
> >  				return err;
> >
> > +			err = flexcan_transceiver_enable(priv);
> > +			if (err)
> > +				return err;
> > +
> >  			err = flexcan_chip_start(dev);
> >  			if (err)
> >  				return err;
> 
> If flexcan_chip_start() fails, shouldn't you have a clean-up branch that will
> flexcan_transceiver_disable()? Or do you consider that it is acceptable to keep
> the transceiver enabled if an error occurs?

Good suggestion, I will add a clean-up branch.

> 
> Speaking of which, if flexcan_resume() fails, priv->can.state would still be set to
> CAN_STATE_ERROR_ACTIVE.

Thanks for your quick response, I will take all these suggestions into consideration and prepare v2 patch.

Regards
Haibo Chen
> 
> 
> Yours sincerely,
> Vincent Mailhol