[PATCH v1] media: i2c: ov5640: Implement get_mbus_config

[PATCH v1] media: i2c: ov5640: Implement get_mbus_config

[Marcel Ziswiler]

From: Aishwarya Kothari <aishwarya.kothari@toradex.com>

Implement the introduced get_mbus_config operation to report the
number of used data lanes on the MIPI CSI-2 interface.

Signed-off-by: Aishwarya Kothari <aishwarya.kothari@toradex.com>
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>

---

 drivers/media/i2c/ov5640.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/drivers/media/i2c/ov5640.c b/drivers/media/i2c/ov5640.c
index e0f908af581b..42d43f0d1e1c 100644
--- a/drivers/media/i2c/ov5640.c
+++ b/drivers/media/i2c/ov5640.c
@@ -3733,6 +3733,19 @@ static int ov5640_init_cfg(struct v4l2_subdev *sd,
 	return 0;
 }
 
+static int ov5640_get_mbus_config(struct v4l2_subdev *sd,
+				   unsigned int pad,
+				   struct v4l2_mbus_config *cfg)
+{
+	struct ov5640_dev *sensor = to_ov5640_dev(sd);
+
+	cfg->type = V4L2_MBUS_CSI2_DPHY;
+	cfg->bus.mipi_csi2.num_data_lanes = sensor->ep.bus.mipi_csi2.num_data_lanes;
+	cfg->bus.mipi_csi2.flags = 0;
+
+	return 0;
+}
+
 static const struct v4l2_subdev_core_ops ov5640_core_ops = {
 	.log_status = v4l2_ctrl_subdev_log_status,
 	.subscribe_event = v4l2_ctrl_subdev_subscribe_event,
@@ -3753,6 +3766,7 @@ static const struct v4l2_subdev_pad_ops ov5640_pad_ops = {
 	.get_selection = ov5640_get_selection,
 	.enum_frame_size = ov5640_enum_frame_size,
 	.enum_frame_interval = ov5640_enum_frame_interval,
+	.get_mbus_config = ov5640_get_mbus_config,
 };
 
 static const struct v4l2_subdev_ops ov5640_subdev_ops = {
-- 
2.36.1

Re: [PATCH v1] media: i2c: ov5640: Implement get_mbus_config

[Jacopo Mondi]

Hi Marcel

On Fri, Jan 27, 2023 at 04:12:44PM +0100, Marcel Ziswiler wrote:
> From: Aishwarya Kothari <aishwarya.kothari@toradex.com>
>
> Implement the introduced get_mbus_config operation to report the
> number of used data lanes on the MIPI CSI-2 interface.
>

OV5640 can operate in parallel mode too.

You can check how it currently configured with ov5640_is_csi2() and
populate struct v4l2_mbus_config accordingly.

Thanks
   j

> Signed-off-by: Aishwarya Kothari <aishwarya.kothari@toradex.com>
> Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
>
> ---
>
>  drivers/media/i2c/ov5640.c | 14 ++++++++++++++
>  1 file changed, 14 insertions(+)
>
> diff --git a/drivers/media/i2c/ov5640.c b/drivers/media/i2c/ov5640.c
> index e0f908af581b..42d43f0d1e1c 100644
> --- a/drivers/media/i2c/ov5640.c
> +++ b/drivers/media/i2c/ov5640.c
> @@ -3733,6 +3733,19 @@ static int ov5640_init_cfg(struct v4l2_subdev *sd,
>  	return 0;
>  }
>
> +static int ov5640_get_mbus_config(struct v4l2_subdev *sd,
> +				   unsigned int pad,
> +				   struct v4l2_mbus_config *cfg)
> +{
> +	struct ov5640_dev *sensor = to_ov5640_dev(sd);
> +
> +	cfg->type = V4L2_MBUS_CSI2_DPHY;
> +	cfg->bus.mipi_csi2.num_data_lanes = sensor->ep.bus.mipi_csi2.num_data_lanes;
> +	cfg->bus.mipi_csi2.flags = 0;
> +
> +	return 0;
> +}
> +
>  static const struct v4l2_subdev_core_ops ov5640_core_ops = {
>  	.log_status = v4l2_ctrl_subdev_log_status,
>  	.subscribe_event = v4l2_ctrl_subdev_subscribe_event,
> @@ -3753,6 +3766,7 @@ static const struct v4l2_subdev_pad_ops ov5640_pad_ops = {
>  	.get_selection = ov5640_get_selection,
>  	.enum_frame_size = ov5640_enum_frame_size,
>  	.enum_frame_interval = ov5640_enum_frame_interval,
> +	.get_mbus_config = ov5640_get_mbus_config,
>  };
>
>  static const struct v4l2_subdev_ops ov5640_subdev_ops = {
> --
> 2.36.1
>

Re: [PATCH v1] media: i2c: ov5640: Implement get_mbus_config

[Laurent Pinchart]

On Fri, Jan 27, 2023 at 06:50:03PM +0100, Jacopo Mondi wrote:
> On Fri, Jan 27, 2023 at 04:12:44PM +0100, Marcel Ziswiler wrote:
> > From: Aishwarya Kothari <aishwarya.kothari@toradex.com>
> >
> > Implement the introduced get_mbus_config operation to report the
> > number of used data lanes on the MIPI CSI-2 interface.
> >
> 
> OV5640 can operate in parallel mode too.
> 
> You can check how it currently configured with ov5640_is_csi2() and
> populate struct v4l2_mbus_config accordingly.

I'm also wondering which CSI-2 receiver needs .get_mbus_config() for the
ov5640. The number of lanes is usually specified in DT, on both sides of
the link. It's only when selecting a number of lanes dynamically at
runtime that .get_mbus_config() is needed.

> > Signed-off-by: Aishwarya Kothari <aishwarya.kothari@toradex.com>
> > Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
> >
> > ---
> >
> >  drivers/media/i2c/ov5640.c | 14 ++++++++++++++
> >  1 file changed, 14 insertions(+)
> >
> > diff --git a/drivers/media/i2c/ov5640.c b/drivers/media/i2c/ov5640.c
> > index e0f908af581b..42d43f0d1e1c 100644
> > --- a/drivers/media/i2c/ov5640.c
> > +++ b/drivers/media/i2c/ov5640.c
> > @@ -3733,6 +3733,19 @@ static int ov5640_init_cfg(struct v4l2_subdev *sd,
> >  	return 0;
> >  }
> >
> > +static int ov5640_get_mbus_config(struct v4l2_subdev *sd,
> > +				   unsigned int pad,
> > +				   struct v4l2_mbus_config *cfg)
> > +{
> > +	struct ov5640_dev *sensor = to_ov5640_dev(sd);
> > +
> > +	cfg->type = V4L2_MBUS_CSI2_DPHY;
> > +	cfg->bus.mipi_csi2.num_data_lanes = sensor->ep.bus.mipi_csi2.num_data_lanes;
> > +	cfg->bus.mipi_csi2.flags = 0;
> > +
> > +	return 0;
> > +}
> > +
> >  static const struct v4l2_subdev_core_ops ov5640_core_ops = {
> >  	.log_status = v4l2_ctrl_subdev_log_status,
> >  	.subscribe_event = v4l2_ctrl_subdev_subscribe_event,
> > @@ -3753,6 +3766,7 @@ static const struct v4l2_subdev_pad_ops ov5640_pad_ops = {
> >  	.get_selection = ov5640_get_selection,
> >  	.enum_frame_size = ov5640_enum_frame_size,
> >  	.enum_frame_interval = ov5640_enum_frame_interval,
> > +	.get_mbus_config = ov5640_get_mbus_config,
> >  };
> >
> >  static const struct v4l2_subdev_ops ov5640_subdev_ops = {

-- 
Regards,

Laurent Pinchart

Re: [PATCH v1] media: i2c: ov5640: Implement get_mbus_config

[Marcel Ziswiler]

Hi Jacopo

On Fri, 2023-01-27 at 18:50 +0100, Jacopo Mondi wrote:
> Hi Marcel
> 
> On Fri, Jan 27, 2023 at 04:12:44PM +0100, Marcel Ziswiler wrote:
> > From: Aishwarya Kothari <aishwarya.kothari@toradex.com>
> > 
> > Implement the introduced get_mbus_config operation to report the
> > number of used data lanes on the MIPI CSI-2 interface.
> > 
> 
> OV5640 can operate in parallel mode too.

I admit, we totally neglected this.

> You can check how it currently configured with ov5640_is_csi2() and
> populate struct v4l2_mbus_config accordingly.

Makes sense. Let us incorporate this in a v2.

> Thanks
>    j

Thank you!

Cheers

Marcel

> > Signed-off-by: Aishwarya Kothari <aishwarya.kothari@toradex.com>
> > Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
> > 
> > ---
> > 
> >  drivers/media/i2c/ov5640.c | 14 ++++++++++++++
> >  1 file changed, 14 insertions(+)
> > 
> > diff --git a/drivers/media/i2c/ov5640.c b/drivers/media/i2c/ov5640.c
> > index e0f908af581b..42d43f0d1e1c 100644
> > --- a/drivers/media/i2c/ov5640.c
> > +++ b/drivers/media/i2c/ov5640.c
> > @@ -3733,6 +3733,19 @@ static int ov5640_init_cfg(struct v4l2_subdev *sd,
> >         return 0;
> >  }
> > 
> > +static int ov5640_get_mbus_config(struct v4l2_subdev *sd,
> > +                                  unsigned int pad,
> > +                                  struct v4l2_mbus_config *cfg)
> > +{
> > +       struct ov5640_dev *sensor = to_ov5640_dev(sd);
> > +
> > +       cfg->type = V4L2_MBUS_CSI2_DPHY;
> > +       cfg->bus.mipi_csi2.num_data_lanes = sensor->ep.bus.mipi_csi2.num_data_lanes;
> > +       cfg->bus.mipi_csi2.flags = 0;
> > +
> > +       return 0;
> > +}
> > +
> >  static const struct v4l2_subdev_core_ops ov5640_core_ops = {
> >         .log_status = v4l2_ctrl_subdev_log_status,
> >         .subscribe_event = v4l2_ctrl_subdev_subscribe_event,
> > @@ -3753,6 +3766,7 @@ static const struct v4l2_subdev_pad_ops ov5640_pad_ops = {
> >         .get_selection = ov5640_get_selection,
> >         .enum_frame_size = ov5640_enum_frame_size,
> >         .enum_frame_interval = ov5640_enum_frame_interval,
> > +       .get_mbus_config = ov5640_get_mbus_config,
> >  };
> > 
> >  static const struct v4l2_subdev_ops ov5640_subdev_ops = {
> > --
> > 2.36.1

Re: [PATCH v1] media: i2c: ov5640: Implement get_mbus_config

[Jacopo Mondi]

Hi Laurent

On Fri, Jan 27, 2023 at 08:34:07PM +0200, Laurent Pinchart wrote:
> On Fri, Jan 27, 2023 at 06:50:03PM +0100, Jacopo Mondi wrote:
> > On Fri, Jan 27, 2023 at 04:12:44PM +0100, Marcel Ziswiler wrote:
> > > From: Aishwarya Kothari <aishwarya.kothari@toradex.com>
> > >
> > > Implement the introduced get_mbus_config operation to report the
> > > number of used data lanes on the MIPI CSI-2 interface.
> > >
> >
> > OV5640 can operate in parallel mode too.
> >
> > You can check how it currently configured with ov5640_is_csi2() and
> > populate struct v4l2_mbus_config accordingly.
>
> I'm also wondering which CSI-2 receiver needs .get_mbus_config() for the
> ov5640. The number of lanes is usually specified in DT, on both sides of
> the link. It's only when selecting a number of lanes dynamically at
> runtime that .get_mbus_config() is needed.
>

iirc Aishwarya and Marcel reported issues on i.MX6 so I presume they
need get_mbus_config as a drivers in staging/media/imx/ requires
that:

drivers/staging/media/imx/imx6-mipi-csi2.c
Fetches the remote mbus config to get the number of lanes and make
sure the bus type is CSI-2

drivers/staging/media/imx/imx-media-csi.c
Fetches the remote mbus config to deduce the bus type in use

In both cases I concur the callers can be fixed to parse their
endpoints but looking at commit 7318abface486d6a6389731810f5b60650daedb5
it seems that was not the plan (reason not clear to me)

> > > Signed-off-by: Aishwarya Kothari <aishwarya.kothari@toradex.com>
> > > Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
> > >
> > > ---
> > >
> > >  drivers/media/i2c/ov5640.c | 14 ++++++++++++++
> > >  1 file changed, 14 insertions(+)
> > >
> > > diff --git a/drivers/media/i2c/ov5640.c b/drivers/media/i2c/ov5640.c
> > > index e0f908af581b..42d43f0d1e1c 100644
> > > --- a/drivers/media/i2c/ov5640.c
> > > +++ b/drivers/media/i2c/ov5640.c
> > > @@ -3733,6 +3733,19 @@ static int ov5640_init_cfg(struct v4l2_subdev *sd,
> > >  	return 0;
> > >  }
> > >
> > > +static int ov5640_get_mbus_config(struct v4l2_subdev *sd,
> > > +				   unsigned int pad,
> > > +				   struct v4l2_mbus_config *cfg)
> > > +{
> > > +	struct ov5640_dev *sensor = to_ov5640_dev(sd);
> > > +
> > > +	cfg->type = V4L2_MBUS_CSI2_DPHY;
> > > +	cfg->bus.mipi_csi2.num_data_lanes = sensor->ep.bus.mipi_csi2.num_data_lanes;
> > > +	cfg->bus.mipi_csi2.flags = 0;
> > > +
> > > +	return 0;
> > > +}
> > > +
> > >  static const struct v4l2_subdev_core_ops ov5640_core_ops = {
> > >  	.log_status = v4l2_ctrl_subdev_log_status,
> > >  	.subscribe_event = v4l2_ctrl_subdev_subscribe_event,
> > > @@ -3753,6 +3766,7 @@ static const struct v4l2_subdev_pad_ops ov5640_pad_ops = {
> > >  	.get_selection = ov5640_get_selection,
> > >  	.enum_frame_size = ov5640_enum_frame_size,
> > >  	.enum_frame_interval = ov5640_enum_frame_interval,
> > > +	.get_mbus_config = ov5640_get_mbus_config,
> > >  };
> > >
> > >  static const struct v4l2_subdev_ops ov5640_subdev_ops = {
>
> --
> Regards,
>
> Laurent Pinchart

Re: [PATCH v1] media: i2c: ov5640: Implement get_mbus_config

[Marcel Ziswiler]

Hi Laurent

On Fri, 2023-01-27 at 20:34 +0200, Laurent Pinchart wrote:
> On Fri, Jan 27, 2023 at 06:50:03PM +0100, Jacopo Mondi wrote:
> > On Fri, Jan 27, 2023 at 04:12:44PM +0100, Marcel Ziswiler wrote:
> > > From: Aishwarya Kothari <aishwarya.kothari@toradex.com>
> > > 
> > > Implement the introduced get_mbus_config operation to report the
> > > number of used data lanes on the MIPI CSI-2 interface.
> > > 
> > 
> > OV5640 can operate in parallel mode too.
> > 
> > You can check how it currently configured with ov5640_is_csi2() and
> > populate struct v4l2_mbus_config accordingly.
> 
> I'm also wondering which CSI-2 receiver needs .get_mbus_config() for the
> ov5640. The number of lanes is usually specified in DT, on both sides of
> the link. It's only when selecting a number of lanes dynamically at
> runtime that .get_mbus_config() is needed.

Remember, this is on Apalis iMX6 where we do and always did specify the lanes on both sides of the link ([1],
[2]) or are we missing anything else in that respect?

Fact is, that with this patch (and the two from Hans [3], [4]) all starts working again. So it really seems
required.

[1]
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch/arm/boot/dts/imx6qdl-apalis.dtsi#n710
[2]
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/arch/arm/boot/dts/imx6qdl-apalis.dtsi#n755
[3] https://lore.kernel.org/all/69624c54-7cbd-7362-c468-f8ffea9614be@xs4all.nl/
[4] https://lore.kernel.org/all/c12cfcc5-1d46-7b5f-4a27-f4cd52a1b885@xs4all.nl/

Cheers

Marcel

> > > Signed-off-by: Aishwarya Kothari <aishwarya.kothari@toradex.com>
> > > Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
> > > 
> > > ---
> > > 
> > >  drivers/media/i2c/ov5640.c | 14 ++++++++++++++
> > >  1 file changed, 14 insertions(+)
> > > 
> > > diff --git a/drivers/media/i2c/ov5640.c b/drivers/media/i2c/ov5640.c
> > > index e0f908af581b..42d43f0d1e1c 100644
> > > --- a/drivers/media/i2c/ov5640.c
> > > +++ b/drivers/media/i2c/ov5640.c
> > > @@ -3733,6 +3733,19 @@ static int ov5640_init_cfg(struct v4l2_subdev *sd,
> > >         return 0;
> > >  }
> > > 
> > > +static int ov5640_get_mbus_config(struct v4l2_subdev *sd,
> > > +                                  unsigned int pad,
> > > +                                  struct v4l2_mbus_config *cfg)
> > > +{
> > > +       struct ov5640_dev *sensor = to_ov5640_dev(sd);
> > > +
> > > +       cfg->type = V4L2_MBUS_CSI2_DPHY;
> > > +       cfg->bus.mipi_csi2.num_data_lanes = sensor->ep.bus.mipi_csi2.num_data_lanes;
> > > +       cfg->bus.mipi_csi2.flags = 0;
> > > +
> > > +       return 0;
> > > +}
> > > +
> > >  static const struct v4l2_subdev_core_ops ov5640_core_ops = {
> > >         .log_status = v4l2_ctrl_subdev_log_status,
> > >         .subscribe_event = v4l2_ctrl_subdev_subscribe_event,
> > > @@ -3753,6 +3766,7 @@ static const struct v4l2_subdev_pad_ops ov5640_pad_ops = {
> > >         .get_selection = ov5640_get_selection,
> > >         .enum_frame_size = ov5640_enum_frame_size,
> > >         .enum_frame_interval = ov5640_enum_frame_interval,
> > > +       .get_mbus_config = ov5640_get_mbus_config,
> > >  };
> > > 
> > >  static const struct v4l2_subdev_ops ov5640_subdev_ops = {

Re: [PATCH v1] media: i2c: ov5640: Implement get_mbus_config

[Philipp Zabel]

Hi Jacopo,

On Sat, Jan 28, 2023 at 11:06:11AM +0100, Jacopo Mondi wrote:
> Hi Laurent
> 
> On Fri, Jan 27, 2023 at 08:34:07PM +0200, Laurent Pinchart wrote:
> > On Fri, Jan 27, 2023 at 06:50:03PM +0100, Jacopo Mondi wrote:
> > > On Fri, Jan 27, 2023 at 04:12:44PM +0100, Marcel Ziswiler wrote:
> > > > From: Aishwarya Kothari <aishwarya.kothari@toradex.com>
> > > >
> > > > Implement the introduced get_mbus_config operation to report the
> > > > number of used data lanes on the MIPI CSI-2 interface.
> > > >
> > >
> > > OV5640 can operate in parallel mode too.
> > >
> > > You can check how it currently configured with ov5640_is_csi2() and
> > > populate struct v4l2_mbus_config accordingly.
> >
> > I'm also wondering which CSI-2 receiver needs .get_mbus_config() for the
> > ov5640. The number of lanes is usually specified in DT, on both sides of
> > the link. It's only when selecting a number of lanes dynamically at
> > runtime that .get_mbus_config() is needed.
> >
> 
> iirc Aishwarya and Marcel reported issues on i.MX6 so I presume they
> need get_mbus_config as a drivers in staging/media/imx/ requires
> that:
> 
> drivers/staging/media/imx/imx6-mipi-csi2.c
> Fetches the remote mbus config to get the number of lanes and make
> sure the bus type is CSI-2
> 
> drivers/staging/media/imx/imx-media-csi.c
> Fetches the remote mbus config to deduce the bus type in use
> 
> In both cases I concur the callers can be fixed to parse their
> endpoints but looking at commit 7318abface486d6a6389731810f5b60650daedb5
> it seems that was not the plan (reason not clear to me)

The tc358743 driver dynamically changes the number of active lanes
depending on bandwidth requirements.

regards
Philipp