[1/2] media: i2c: max96714: add HAS_EVENTS subdev flag

Message ID 20240910134428.795273-2-tomm.merciai@gmail.com (mailing list archive)
State Superseded
Delegated to: Sakari Ailus
Headers
Series media: i2c: max96714/max96717 |

Checks

Context Check Description
media-ci/HTML_report success Link
media-ci/report success Link
media-ci/bisect success Link
media-ci/doc success Link
media-ci/build success Link
media-ci/static-upstream success Link
media-ci/abi success Link
media-ci/media-patchstyle success Link
media-ci/checkpatch success Link

Commit Message

Tommaso Merciai Sept. 10, 2024, 1:44 p.m. UTC
  Controls can be exposed to userspace via a v4l-subdevX device, and
userspace has to be able to subscribe to control events so that it is
notified when the control changes value. Add missing HAS_EVENTS flag.

Signed-off-by: Tommaso Merciai <tomm.merciai@gmail.com>
---
 drivers/media/i2c/max96714.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
  

Comments

Laurent Pinchart Sept. 12, 2024, 10:44 a.m. UTC | #1
Hi Tommaso,

On Tue, Sep 10, 2024 at 03:44:27PM +0200, Tommaso Merciai wrote:
> Controls can be exposed to userspace via a v4l-subdevX device, and
> userspace has to be able to subscribe to control events so that it is
> notified when the control changes value. Add missing HAS_EVENTS flag.

How is this supposed to work, given that the driver doesn't implement
.subscribe_event() ?

> Signed-off-by: Tommaso Merciai <tomm.merciai@gmail.com>
> ---
>  drivers/media/i2c/max96714.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/media/i2c/max96714.c b/drivers/media/i2c/max96714.c
> index 159753b13777..94b1bc000e48 100644
> --- a/drivers/media/i2c/max96714.c
> +++ b/drivers/media/i2c/max96714.c
> @@ -602,7 +602,8 @@ static int max96714_create_subdev(struct max96714_priv *priv)
>  		goto err_free_ctrl;
>  	}
>  
> -	priv->sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE | V4L2_SUBDEV_FL_STREAMS;
> +	priv->sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE |
> +			  V4L2_SUBDEV_FL_HAS_EVENTS | V4L2_SUBDEV_FL_STREAMS;
>  	priv->sd.entity.function = MEDIA_ENT_F_VID_IF_BRIDGE;
>  	priv->sd.entity.ops = &max96714_entity_ops;
>
  
Tommaso Merciai Sept. 12, 2024, 11:11 a.m. UTC | #2
Hi Laurent,

On Thu, Sep 12, 2024 at 01:44:09PM +0300, Laurent Pinchart wrote:
> Hi Tommaso,
> 
> On Tue, Sep 10, 2024 at 03:44:27PM +0200, Tommaso Merciai wrote:
> > Controls can be exposed to userspace via a v4l-subdevX device, and
> > userspace has to be able to subscribe to control events so that it is
> > notified when the control changes value. Add missing HAS_EVENTS flag.
> 
> How is this supposed to work, given that the driver doesn't implement
> .subscribe_event() ?

You are completely right, sorry.
I think in both cases I'm missing:

diff --git a/drivers/media/i2c/max96714.c b/drivers/media/i2c/max96714.c
index 94b1bc000e48..2257b6b807ea 100644
--- a/drivers/media/i2c/max96714.c
+++ b/drivers/media/i2c/max96714.c
@@ -17,6 +17,7 @@

 #include <media/v4l2-cci.h>
 #include <media/v4l2-ctrls.h>
+#include <media/v4l2-event.h>
 #include <media/v4l2-fwnode.h>
 #include <media/v4l2-subdev.h>

@@ -488,6 +489,8 @@ static int max96714_log_status(struct v4l2_subdev *sd)

 static const struct v4l2_subdev_core_ops max96714_subdev_core_ops = {
        .log_status = max96714_log_status,
+       .subscribe_event = v4l2_ctrl_subdev_subscribe_event,
+       .unsubscribe_event = v4l2_event_subdev_unsubscribe,
 };

 static const struct v4l2_subdev_video_ops max96714_video_ops = {

Like you suggest. Or I'm wrong?

Thanks & Regards,
Tommaso

> 
> > Signed-off-by: Tommaso Merciai <tomm.merciai@gmail.com>
> > ---
> >  drivers/media/i2c/max96714.c | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/media/i2c/max96714.c b/drivers/media/i2c/max96714.c
> > index 159753b13777..94b1bc000e48 100644
> > --- a/drivers/media/i2c/max96714.c
> > +++ b/drivers/media/i2c/max96714.c
> > @@ -602,7 +602,8 @@ static int max96714_create_subdev(struct max96714_priv *priv)
> >  		goto err_free_ctrl;
> >  	}
> >  
> > -	priv->sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE | V4L2_SUBDEV_FL_STREAMS;
> > +	priv->sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE |
> > +			  V4L2_SUBDEV_FL_HAS_EVENTS | V4L2_SUBDEV_FL_STREAMS;
> >  	priv->sd.entity.function = MEDIA_ENT_F_VID_IF_BRIDGE;
> >  	priv->sd.entity.ops = &max96714_entity_ops;
> >  
> 
> -- 
> Regards,
> 
> Laurent Pinchart
  
Laurent Pinchart Sept. 12, 2024, 12:05 p.m. UTC | #3
On Thu, Sep 12, 2024 at 01:11:13PM +0200, Tommaso Merciai wrote:
> Hi Laurent,
> 
> On Thu, Sep 12, 2024 at 01:44:09PM +0300, Laurent Pinchart wrote:
> > Hi Tommaso,
> > 
> > On Tue, Sep 10, 2024 at 03:44:27PM +0200, Tommaso Merciai wrote:
> > > Controls can be exposed to userspace via a v4l-subdevX device, and
> > > userspace has to be able to subscribe to control events so that it is
> > > notified when the control changes value. Add missing HAS_EVENTS flag.
> > 
> > How is this supposed to work, given that the driver doesn't implement
> > .subscribe_event() ?
> 
> You are completely right, sorry.
> I think in both cases I'm missing:
> 
> diff --git a/drivers/media/i2c/max96714.c b/drivers/media/i2c/max96714.c
> index 94b1bc000e48..2257b6b807ea 100644
> --- a/drivers/media/i2c/max96714.c
> +++ b/drivers/media/i2c/max96714.c
> @@ -17,6 +17,7 @@
> 
>  #include <media/v4l2-cci.h>
>  #include <media/v4l2-ctrls.h>
> +#include <media/v4l2-event.h>
>  #include <media/v4l2-fwnode.h>
>  #include <media/v4l2-subdev.h>
> 
> @@ -488,6 +489,8 @@ static int max96714_log_status(struct v4l2_subdev *sd)
> 
>  static const struct v4l2_subdev_core_ops max96714_subdev_core_ops = {
>         .log_status = max96714_log_status,
> +       .subscribe_event = v4l2_ctrl_subdev_subscribe_event,
> +       .unsubscribe_event = v4l2_event_subdev_unsubscribe,
>  };
> 
>  static const struct v4l2_subdev_video_ops max96714_video_ops = {
> 
> Like you suggest. Or I'm wrong?

That looks better :-)

Out of curiosity, what's your use case for control events ?

> > > Signed-off-by: Tommaso Merciai <tomm.merciai@gmail.com>
> > > ---
> > >  drivers/media/i2c/max96714.c | 3 ++-
> > >  1 file changed, 2 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/drivers/media/i2c/max96714.c b/drivers/media/i2c/max96714.c
> > > index 159753b13777..94b1bc000e48 100644
> > > --- a/drivers/media/i2c/max96714.c
> > > +++ b/drivers/media/i2c/max96714.c
> > > @@ -602,7 +602,8 @@ static int max96714_create_subdev(struct max96714_priv *priv)
> > >  		goto err_free_ctrl;
> > >  	}
> > >  
> > > -	priv->sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE | V4L2_SUBDEV_FL_STREAMS;
> > > +	priv->sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE |
> > > +			  V4L2_SUBDEV_FL_HAS_EVENTS | V4L2_SUBDEV_FL_STREAMS;
> > >  	priv->sd.entity.function = MEDIA_ENT_F_VID_IF_BRIDGE;
> > >  	priv->sd.entity.ops = &max96714_entity_ops;
> > >
  
Sakari Ailus Sept. 12, 2024, 12:14 p.m. UTC | #4
Hi Laurent,

On Thu, Sep 12, 2024 at 03:05:10PM +0300, Laurent Pinchart wrote:
> On Thu, Sep 12, 2024 at 01:11:13PM +0200, Tommaso Merciai wrote:
> > Hi Laurent,
> > 
> > On Thu, Sep 12, 2024 at 01:44:09PM +0300, Laurent Pinchart wrote:
> > > Hi Tommaso,
> > > 
> > > On Tue, Sep 10, 2024 at 03:44:27PM +0200, Tommaso Merciai wrote:
> > > > Controls can be exposed to userspace via a v4l-subdevX device, and
> > > > userspace has to be able to subscribe to control events so that it is
> > > > notified when the control changes value. Add missing HAS_EVENTS flag.
> > > 
> > > How is this supposed to work, given that the driver doesn't implement
> > > .subscribe_event() ?
> > 
> > You are completely right, sorry.
> > I think in both cases I'm missing:
> > 
> > diff --git a/drivers/media/i2c/max96714.c b/drivers/media/i2c/max96714.c
> > index 94b1bc000e48..2257b6b807ea 100644
> > --- a/drivers/media/i2c/max96714.c
> > +++ b/drivers/media/i2c/max96714.c
> > @@ -17,6 +17,7 @@
> > 
> >  #include <media/v4l2-cci.h>
> >  #include <media/v4l2-ctrls.h>
> > +#include <media/v4l2-event.h>
> >  #include <media/v4l2-fwnode.h>
> >  #include <media/v4l2-subdev.h>
> > 
> > @@ -488,6 +489,8 @@ static int max96714_log_status(struct v4l2_subdev *sd)
> > 
> >  static const struct v4l2_subdev_core_ops max96714_subdev_core_ops = {
> >         .log_status = max96714_log_status,
> > +       .subscribe_event = v4l2_ctrl_subdev_subscribe_event,
> > +       .unsubscribe_event = v4l2_event_subdev_unsubscribe,
> >  };
> > 
> >  static const struct v4l2_subdev_video_ops max96714_video_ops = {
> > 
> > Like you suggest. Or I'm wrong?
> 
> That looks better :-)
> 
> Out of curiosity, what's your use case for control events ?

I'm not sure if there's a use case, however control event should be
supported when a driver supports controls.
  
Laurent Pinchart Sept. 12, 2024, 12:18 p.m. UTC | #5
On Thu, Sep 12, 2024 at 12:14:52PM +0000, Sakari Ailus wrote:
> On Thu, Sep 12, 2024 at 03:05:10PM +0300, Laurent Pinchart wrote:
> > On Thu, Sep 12, 2024 at 01:11:13PM +0200, Tommaso Merciai wrote:
> > > On Thu, Sep 12, 2024 at 01:44:09PM +0300, Laurent Pinchart wrote:
> > > > On Tue, Sep 10, 2024 at 03:44:27PM +0200, Tommaso Merciai wrote:
> > > > > Controls can be exposed to userspace via a v4l-subdevX device, and
> > > > > userspace has to be able to subscribe to control events so that it is
> > > > > notified when the control changes value. Add missing HAS_EVENTS flag.
> > > > 
> > > > How is this supposed to work, given that the driver doesn't implement
> > > > .subscribe_event() ?
> > > 
> > > You are completely right, sorry.
> > > I think in both cases I'm missing:
> > > 
> > > diff --git a/drivers/media/i2c/max96714.c b/drivers/media/i2c/max96714.c
> > > index 94b1bc000e48..2257b6b807ea 100644
> > > --- a/drivers/media/i2c/max96714.c
> > > +++ b/drivers/media/i2c/max96714.c
> > > @@ -17,6 +17,7 @@
> > > 
> > >  #include <media/v4l2-cci.h>
> > >  #include <media/v4l2-ctrls.h>
> > > +#include <media/v4l2-event.h>
> > >  #include <media/v4l2-fwnode.h>
> > >  #include <media/v4l2-subdev.h>
> > > 
> > > @@ -488,6 +489,8 @@ static int max96714_log_status(struct v4l2_subdev *sd)
> > > 
> > >  static const struct v4l2_subdev_core_ops max96714_subdev_core_ops = {
> > >         .log_status = max96714_log_status,
> > > +       .subscribe_event = v4l2_ctrl_subdev_subscribe_event,
> > > +       .unsubscribe_event = v4l2_event_subdev_unsubscribe,
> > >  };
> > > 
> > >  static const struct v4l2_subdev_video_ops max96714_video_ops = {
> > > 
> > > Like you suggest. Or I'm wrong?
> > 
> > That looks better :-)
> > 
> > Out of curiosity, what's your use case for control events ?
> 
> I'm not sure if there's a use case, however control event should be
> supported when a driver supports controls.

Not disputing that. That's why I asked "out of curiosity".
  
Tommaso Merciai Sept. 12, 2024, 12:52 p.m. UTC | #6
Hi Laurent, Sakari,
Thanks for your comments.

On Thu, Sep 12, 2024 at 03:18:16PM +0300, Laurent Pinchart wrote:
> On Thu, Sep 12, 2024 at 12:14:52PM +0000, Sakari Ailus wrote:
> > On Thu, Sep 12, 2024 at 03:05:10PM +0300, Laurent Pinchart wrote:
> > > On Thu, Sep 12, 2024 at 01:11:13PM +0200, Tommaso Merciai wrote:
> > > > On Thu, Sep 12, 2024 at 01:44:09PM +0300, Laurent Pinchart wrote:
> > > > > On Tue, Sep 10, 2024 at 03:44:27PM +0200, Tommaso Merciai wrote:
> > > > > > Controls can be exposed to userspace via a v4l-subdevX device, and
> > > > > > userspace has to be able to subscribe to control events so that it is
> > > > > > notified when the control changes value. Add missing HAS_EVENTS flag.
> > > > > 
> > > > > How is this supposed to work, given that the driver doesn't implement
> > > > > .subscribe_event() ?
> > > > 
> > > > You are completely right, sorry.
> > > > I think in both cases I'm missing:
> > > > 
> > > > diff --git a/drivers/media/i2c/max96714.c b/drivers/media/i2c/max96714.c
> > > > index 94b1bc000e48..2257b6b807ea 100644
> > > > --- a/drivers/media/i2c/max96714.c
> > > > +++ b/drivers/media/i2c/max96714.c
> > > > @@ -17,6 +17,7 @@
> > > > 
> > > >  #include <media/v4l2-cci.h>
> > > >  #include <media/v4l2-ctrls.h>
> > > > +#include <media/v4l2-event.h>
> > > >  #include <media/v4l2-fwnode.h>
> > > >  #include <media/v4l2-subdev.h>
> > > > 
> > > > @@ -488,6 +489,8 @@ static int max96714_log_status(struct v4l2_subdev *sd)
> > > > 
> > > >  static const struct v4l2_subdev_core_ops max96714_subdev_core_ops = {
> > > >         .log_status = max96714_log_status,
> > > > +       .subscribe_event = v4l2_ctrl_subdev_subscribe_event,
> > > > +       .unsubscribe_event = v4l2_event_subdev_unsubscribe,
> > > >  };
> > > > 
> > > >  static const struct v4l2_subdev_video_ops max96714_video_ops = {
> > > > 
> > > > Like you suggest. Or I'm wrong?
> > > 
> > > That looks better :-)
> > > 
> > > Out of curiosity, what's your use case for control events ?
> > 
> > I'm not sure if there's a use case, however control event should be
> > supported when a driver supports controls.
> 
> Not disputing that. That's why I asked "out of curiosity".

No use case :'(.
Sakari point me that we are missing events flag, some days ago.

Thanks & Regards,
Tommaso

> 
> -- 
> Regards,
> 
> Laurent Pinchart
  

Patch

diff --git a/drivers/media/i2c/max96714.c b/drivers/media/i2c/max96714.c
index 159753b13777..94b1bc000e48 100644
--- a/drivers/media/i2c/max96714.c
+++ b/drivers/media/i2c/max96714.c
@@ -602,7 +602,8 @@  static int max96714_create_subdev(struct max96714_priv *priv)
 		goto err_free_ctrl;
 	}
 
-	priv->sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE | V4L2_SUBDEV_FL_STREAMS;
+	priv->sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE |
+			  V4L2_SUBDEV_FL_HAS_EVENTS | V4L2_SUBDEV_FL_STREAMS;
 	priv->sd.entity.function = MEDIA_ENT_F_VID_IF_BRIDGE;
 	priv->sd.entity.ops = &max96714_entity_ops;