diff mbox series

media: nxp: imx8mq-mipi-csi2: Fix CSI clocks always enabled issue

Message ID 20240929101635.1648234-1-guoniu.zhou@oss.nxp.com (mailing list archive)
State New
Headers show
Series media: nxp: imx8mq-mipi-csi2: Fix CSI clocks always enabled issue | expand

Commit Message

G.N. Zhou (OSS) Sept. 29, 2024, 10:16 a.m. UTC 
From: "Guoniu.zhou" <guoniu.zhou@nxp.com>

Fix CSI clocks always enabled issue after run system suspend/resume.

The clocks for CSI will be enabled in imx8mq_mipi_csi_pm_resume()
when system resume and the state of CSI will be set to ST_POWERED.
It not only cause clock always enabled issue but also system hang
issue on iMX8ULP platform since imx8mq_mipi_csi_pm_suspend() will
access CSI registers if run system suspend/resume again, but the
CSI power domain is not active.

In order to fix this issue, using pm_runtime_force_suspend/resume
instead of calling imx8mq_mipi_csi_pm_suspend/resume directly.

Signed-off-by: Guoniu.zhou <guoniu.zhou@nxp.com>
---
 drivers/media/platform/nxp/imx8mq-mipi-csi2.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

Comments

Laurent Pinchart Sept. 29, 2024, noon UTC | #1 
Hi Guoniu,

Thank you for the patch.

On Sun, Sep 29, 2024 at 06:16:35PM +0800, guoniu.zhou@oss.nxp.com wrote:
> From: "Guoniu.zhou" <guoniu.zhou@nxp.com>
> 
> Fix CSI clocks always enabled issue after run system suspend/resume.
> 
> The clocks for CSI will be enabled in imx8mq_mipi_csi_pm_resume()
> when system resume and the state of CSI will be set to ST_POWERED.
> It not only cause clock always enabled issue but also system hang
> issue on iMX8ULP platform since imx8mq_mipi_csi_pm_suspend() will
> access CSI registers if run system suspend/resume again, but the
> CSI power domain is not active.
> 
> In order to fix this issue, using pm_runtime_force_suspend/resume
> instead of calling imx8mq_mipi_csi_pm_suspend/resume directly.

I think you can simplify suspend/resume handling in the driver by
dropping the system suspend/resume handlers. At system suspend time the
device should already be stopped, because the top-level driver (the CSI
bridge in the i.MX8MQ) should have stopped the pipeline. Similarly, at
system resume time, there should be no need to restart streaming, as the
top-level driver will control that.

I'll send a patch and CC you, could you give it a try ?

> Signed-off-by: Guoniu.zhou <guoniu.zhou@nxp.com>
> ---
>  drivers/media/platform/nxp/imx8mq-mipi-csi2.c | 7 +++++--
>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/media/platform/nxp/imx8mq-mipi-csi2.c b/drivers/media/platform/nxp/imx8mq-mipi-csi2.c
> index d4a6c5532969..15029bb81b35 100644
> --- a/drivers/media/platform/nxp/imx8mq-mipi-csi2.c
> +++ b/drivers/media/platform/nxp/imx8mq-mipi-csi2.c
> @@ -697,8 +697,11 @@ static int imx8mq_mipi_csi_suspend(struct device *dev)
>  {
>  	struct v4l2_subdev *sd = dev_get_drvdata(dev);
>  	struct csi_state *state = mipi_sd_to_csi2_state(sd);
> +	int ret;
>  
> -	imx8mq_mipi_csi_pm_suspend(dev);
> +	ret = pm_runtime_force_suspend(dev);
> +	if (ret < 0)
> +		return ret;
>  
>  	state->state |= ST_SUSPENDED;
>  
> @@ -713,7 +716,7 @@ static int imx8mq_mipi_csi_resume(struct device *dev)
>  	if (!(state->state & ST_SUSPENDED))
>  		return 0;
>  
> -	return imx8mq_mipi_csi_pm_resume(dev);
> +	return pm_runtime_force_resume(dev);
>  }
>  
>  static int imx8mq_mipi_csi_runtime_suspend(struct device *dev)
G.N. Zhou (OSS) Sept. 29, 2024, 12:40 p.m. UTC | #2 
Hi Laurent,

> -----Original Message-----
> From: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
> Sent: Sunday, September 29, 2024 8:01 PM
> To: G.N. Zhou (OSS) <guoniu.zhou@oss.nxp.com>
> Cc: rmfrfs@gmail.com; martink@posteo.de; kernel@puri.sm;
> mchehab@kernel.org; shawnguo@kernel.org; s.hauer@pengutronix.de;
> kernel@pengutronix.de; festevam@gmail.com; imx@lists.linux.dev; linux-
> media@vger.kernel.org; linux-arm-kernel@lists.infradead.org; linux-
> kernel@vger.kernel.org
> Subject: Re: [PATCH] media: nxp: imx8mq-mipi-csi2: Fix CSI clocks always enabled
> issue
> 
> Hi Guoniu,
> 
> Thank you for the patch.
> 
> On Sun, Sep 29, 2024 at 06:16:35PM +0800, guoniu.zhou@oss.nxp.com wrote:
> > From: "Guoniu.zhou" <guoniu.zhou@nxp.com>
> >
> > Fix CSI clocks always enabled issue after run system suspend/resume.
> >
> > The clocks for CSI will be enabled in imx8mq_mipi_csi_pm_resume() when
> > system resume and the state of CSI will be set to ST_POWERED.
> > It not only cause clock always enabled issue but also system hang
> > issue on iMX8ULP platform since imx8mq_mipi_csi_pm_suspend() will
> > access CSI registers if run system suspend/resume again, but the CSI
> > power domain is not active.
> >
> > In order to fix this issue, using pm_runtime_force_suspend/resume
> > instead of calling imx8mq_mipi_csi_pm_suspend/resume directly.
> 
> I think you can simplify suspend/resume handling in the driver by dropping the
> system suspend/resume handlers. At system suspend time the device should
> already be stopped, because the top-level driver (the CSI bridge in the i.MX8MQ)
> should have stopped the pipeline. Similarly, at system resume time, there should
> be no need to restart streaming, as the top-level driver will control that.
> 
> I'll send a patch and CC you, could you give it a try ?

Sure.

> 
> > Signed-off-by: Guoniu.zhou <guoniu.zhou@nxp.com>
> > ---
> >  drivers/media/platform/nxp/imx8mq-mipi-csi2.c | 7 +++++--
> >  1 file changed, 5 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/media/platform/nxp/imx8mq-mipi-csi2.c
> > b/drivers/media/platform/nxp/imx8mq-mipi-csi2.c
> > index d4a6c5532969..15029bb81b35 100644
> > --- a/drivers/media/platform/nxp/imx8mq-mipi-csi2.c
> > +++ b/drivers/media/platform/nxp/imx8mq-mipi-csi2.c
> > @@ -697,8 +697,11 @@ static int imx8mq_mipi_csi_suspend(struct device
> > *dev)  {
> >  	struct v4l2_subdev *sd = dev_get_drvdata(dev);
> >  	struct csi_state *state = mipi_sd_to_csi2_state(sd);
> > +	int ret;
> >
> > -	imx8mq_mipi_csi_pm_suspend(dev);
> > +	ret = pm_runtime_force_suspend(dev);
> > +	if (ret < 0)
> > +		return ret;
> >
> >  	state->state |= ST_SUSPENDED;
> >
> > @@ -713,7 +716,7 @@ static int imx8mq_mipi_csi_resume(struct device *dev)
> >  	if (!(state->state & ST_SUSPENDED))
> >  		return 0;
> >
> > -	return imx8mq_mipi_csi_pm_resume(dev);
> > +	return pm_runtime_force_resume(dev);
> >  }
> >
> >  static int imx8mq_mipi_csi_runtime_suspend(struct device *dev)
> 
> --
> Regards,
> 
> Laurent Pinchart
diff mbox series

Patch

diff --git a/drivers/media/platform/nxp/imx8mq-mipi-csi2.c b/drivers/media/platform/nxp/imx8mq-mipi-csi2.c
index d4a6c5532969..15029bb81b35 100644
--- a/drivers/media/platform/nxp/imx8mq-mipi-csi2.c
+++ b/drivers/media/platform/nxp/imx8mq-mipi-csi2.c
@@ -697,8 +697,11 @@  static int imx8mq_mipi_csi_suspend(struct device *dev)
 {
 	struct v4l2_subdev *sd = dev_get_drvdata(dev);
 	struct csi_state *state = mipi_sd_to_csi2_state(sd);
+	int ret;
 
-	imx8mq_mipi_csi_pm_suspend(dev);
+	ret = pm_runtime_force_suspend(dev);
+	if (ret < 0)
+		return ret;
 
 	state->state |= ST_SUSPENDED;
 
@@ -713,7 +716,7 @@  static int imx8mq_mipi_csi_resume(struct device *dev)
 	if (!(state->state & ST_SUSPENDED))
 		return 0;
 
-	return imx8mq_mipi_csi_pm_resume(dev);
+	return pm_runtime_force_resume(dev);
 }
 
 static int imx8mq_mipi_csi_runtime_suspend(struct device *dev)