[PATCH 1/2] serial: sifive: Add suspend and resume operations

[PATCH 1/2] serial: sifive: Add suspend and resume operations

[Nick Hu]

If the Sifive Uart is not used as the wake up source, suspend the uart
before the system enter the suspend state to prevent it woken up by
unexpected uart interrupt. Resume the uart once the system woken up.

Signed-off-by: Nick Hu <nick.hu@sifive.com>
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
---
 drivers/tty/serial/sifive.c | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)

diff --git a/drivers/tty/serial/sifive.c b/drivers/tty/serial/sifive.c
index a19db49327e2..87994cb69007 100644
--- a/drivers/tty/serial/sifive.c
+++ b/drivers/tty/serial/sifive.c
@@ -1022,6 +1022,31 @@ static int sifive_serial_remove(struct platform_device *dev)
 	return 0;
 }
 
+static int sifive_serial_suspend(struct device *dev)
+{
+	int ret = 0;
+	struct sifive_serial_port *ssp = dev_get_drvdata(dev);
+
+	if (ssp && ssp->port.type != PORT_UNKNOWN)
+		ret = uart_suspend_port(&sifive_serial_uart_driver, &ssp->port);
+
+	return ret;
+}
+
+static int sifive_serial_resume(struct device *dev)
+{
+	int ret = 0;
+	struct sifive_serial_port *ssp = dev_get_drvdata(dev);
+
+	if (ssp && ssp->port.type != PORT_UNKNOWN)
+		ret = uart_resume_port(&sifive_serial_uart_driver, &ssp->port);
+
+	return ret;
+}
+
+DEFINE_SIMPLE_DEV_PM_OPS(sifive_uart_pm_ops, sifive_serial_suspend,
+			 sifive_serial_resume);
+
 static const struct of_device_id sifive_serial_of_match[] = {
 	{ .compatible = "sifive,fu540-c000-uart0" },
 	{ .compatible = "sifive,uart0" },
@@ -1034,6 +1059,7 @@ static struct platform_driver sifive_serial_platform_driver = {
 	.remove		= sifive_serial_remove,
 	.driver		= {
 		.name	= SIFIVE_SERIAL_NAME,
+		.pm = pm_sleep_ptr(&sifive_uart_pm_ops),
 		.of_match_table = of_match_ptr(sifive_serial_of_match),
 	},
 };
-- 
2.34.1

Re: [PATCH 1/2] serial: sifive: Add suspend and resume operations

[Conor Dooley]

[-- Attachment #1: Type: text/plain, Size: 405 bytes --]

On Tue, Aug 08, 2023 at 03:26:25PM +0800, Nick Hu wrote:
> If the Sifive Uart is not used as the wake up source, suspend the uart
> before the system enter the suspend state to prevent it woken up by
> unexpected uart interrupt. Resume the uart once the system woken up.
> 
> Signed-off-by: Nick Hu <nick.hu@sifive.com>
> Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>

Where is patch 2/2?

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

Re: [PATCH 1/2] serial: sifive: Add suspend and resume operations

[Ben Dooks]

On 08/08/2023 08:26, Nick Hu wrote:
> If the Sifive Uart is not used as the wake up source, suspend the uart
> before the system enter the suspend state to prevent it woken up by
> unexpected uart interrupt. Resume the uart once the system woken up.
> 
> Signed-off-by: Nick Hu <nick.hu@sifive.com>
> Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>

^ This should be Reviewed-by, as I did review on this earlier.

> ---
>   drivers/tty/serial/sifive.c | 26 ++++++++++++++++++++++++++
>   1 file changed, 26 insertions(+)
> 
> diff --git a/drivers/tty/serial/sifive.c b/drivers/tty/serial/sifive.c
> index a19db49327e2..87994cb69007 100644
> --- a/drivers/tty/serial/sifive.c
> +++ b/drivers/tty/serial/sifive.c
> @@ -1022,6 +1022,31 @@ static int sifive_serial_remove(struct platform_device *dev)
>   	return 0;
>   }
>   
> +static int sifive_serial_suspend(struct device *dev)
> +{
> +	int ret = 0;
> +	struct sifive_serial_port *ssp = dev_get_drvdata(dev);

Minor annyonance is ordering of ssp and ret, I think the showrter one
last is the nicest looking.

> +	if (ssp && ssp->port.type != PORT_UNKNOWN)
> +		ret = uart_suspend_port(&sifive_serial_uart_driver, &ssp->port);

Do we really need a test for ssp being valid if the device is bound.
Not sure if the port.type is also useful?

> +	return ret;
> +}
> +
> +static int sifive_serial_resume(struct device *dev)
> +{
> +	int ret = 0;
> +	struct sifive_serial_port *ssp = dev_get_drvdata(dev);
> +
> +	if (ssp && ssp->port.type != PORT_UNKNOWN)
> +		ret = uart_resume_port(&sifive_serial_uart_driver, &ssp->port);
> +
> +	return ret;
> +}
> +
> +DEFINE_SIMPLE_DEV_PM_OPS(sifive_uart_pm_ops, sifive_serial_suspend,
> +			 sifive_serial_resume);
> +
>   static const struct of_device_id sifive_serial_of_match[] = {
>   	{ .compatible = "sifive,fu540-c000-uart0" },
>   	{ .compatible = "sifive,uart0" },
> @@ -1034,6 +1059,7 @@ static struct platform_driver sifive_serial_platform_driver = {
>   	.remove		= sifive_serial_remove,
>   	.driver		= {
>   		.name	= SIFIVE_SERIAL_NAME,
> +		.pm = pm_sleep_ptr(&sifive_uart_pm_ops),
>   		.of_match_table = of_match_ptr(sifive_serial_of_match),
>   	},
>   };

-- 
Ben Dooks				http://www.codethink.co.uk/
Senior Engineer				Codethink - Providing Genius

https://www.codethink.co.uk/privacy.html

Re: [PATCH 1/2] serial: sifive: Add suspend and resume operations

[Nick Hu]

Hi Conor


On Tue, Aug 8, 2023 at 4:44 PM Conor Dooley <conor@kernel.org> wrote:
>
> On Tue, Aug 08, 2023 at 03:26:25PM +0800, Nick Hu wrote:
> > If the Sifive Uart is not used as the wake up source, suspend the uart
> > before the system enter the suspend state to prevent it woken up by
> > unexpected uart interrupt. Resume the uart once the system woken up.
> >
> > Signed-off-by: Nick Hu <nick.hu@sifive.com>
> > Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
>
> Where is patch 2/2?
This was my bad. There is no patch 2/2. I'll correct it in V2.

Nick

Re: [PATCH 1/2] serial: sifive: Add suspend and resume operations

[Nick Hu]

Hi Ben



On Tue, Aug 8, 2023 at 4:47 PM Ben Dooks <ben.dooks@codethink.co.uk> wrote:
>
> On 08/08/2023 08:26, Nick Hu wrote:
> > If the Sifive Uart is not used as the wake up source, suspend the uart
> > before the system enter the suspend state to prevent it woken up by
> > unexpected uart interrupt. Resume the uart once the system woken up.
> >
> > Signed-off-by: Nick Hu <nick.hu@sifive.com>
> > Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
>
> ^ This should be Reviewed-by, as I did review on this earlier.
>
I'll correct it in V2
> > ---
> >   drivers/tty/serial/sifive.c | 26 ++++++++++++++++++++++++++
> >   1 file changed, 26 insertions(+)
> >
> > diff --git a/drivers/tty/serial/sifive.c b/drivers/tty/serial/sifive.c
> > index a19db49327e2..87994cb69007 100644
> > --- a/drivers/tty/serial/sifive.c
> > +++ b/drivers/tty/serial/sifive.c
> > @@ -1022,6 +1022,31 @@ static int sifive_serial_remove(struct platform_device *dev)
> >       return 0;
> >   }
> >
> > +static int sifive_serial_suspend(struct device *dev)
> > +{
> > +     int ret = 0;
> > +     struct sifive_serial_port *ssp = dev_get_drvdata(dev);
>
> Minor annyonance is ordering of ssp and ret, I think the showrter one
> last is the nicest looking.
>
> > +     if (ssp && ssp->port.type != PORT_UNKNOWN)
> > +             ret = uart_suspend_port(&sifive_serial_uart_driver, &ssp->port);
>
> Do we really need a test for ssp being valid if the device is bound.
> Not sure if the port.type is also useful?
>
You are right. This might be an unnecessary check.
The ssp is bound in the probe function and the PORT_UNKNOWN only be
set when we do the sifive_serial_remove().
And for the ordering of ret, We don't need the ret if we move the check.
We can just return uart_suspend_port(&sifive_serial_uart_driver, &ssp->port);

Will correct it in V2.
> > +     return ret;
> > +}
> > +
> > +static int sifive_serial_resume(struct device *dev)
> > +{
> > +     int ret = 0;
> > +     struct sifive_serial_port *ssp = dev_get_drvdata(dev);
> > +
> > +     if (ssp && ssp->port.type != PORT_UNKNOWN)
> > +             ret = uart_resume_port(&sifive_serial_uart_driver, &ssp->port);
> > +
> > +     return ret;
> > +}
> > +
> > +DEFINE_SIMPLE_DEV_PM_OPS(sifive_uart_pm_ops, sifive_serial_suspend,
> > +                      sifive_serial_resume);
> > +
> >   static const struct of_device_id sifive_serial_of_match[] = {
> >       { .compatible = "sifive,fu540-c000-uart0" },
> >       { .compatible = "sifive,uart0" },
> > @@ -1034,6 +1059,7 @@ static struct platform_driver sifive_serial_platform_driver = {
> >       .remove         = sifive_serial_remove,
> >       .driver         = {
> >               .name   = SIFIVE_SERIAL_NAME,
> > +             .pm = pm_sleep_ptr(&sifive_uart_pm_ops),
> >               .of_match_table = of_match_ptr(sifive_serial_of_match),
> >       },
> >   };
>
> --
> Ben Dooks                               http://www.codethink.co.uk/
> Senior Engineer                         Codethink - Providing Genius
>
> https://www.codethink.co.uk/privacy.html
>