Re: [PATCH v10 05/38] media: subdev: Add v4l2_subdev_lock_and_return_state()

[Sakari Ailus <sakari.ailus@linux.intel.com>]

Hi Laurent,

On Thu, Dec 16, 2021 at 04:34:22PM +0200, Laurent Pinchart wrote:
> Hi Tomi,
> 
> Thank you for the patch.
> 
> On Tue, Nov 30, 2021 at 04:15:03PM +0200, Tomi Valkeinen wrote:
> > All suitable subdev ops are now passed either the TRY or the ACTIVE
> > state by the v4l2 core. However, other subdev drivers can still call the
> > ops passing NULL as the state, implying the active case.
> > 
> > For all current upstream drivers this doesn't matter, as they do not
> > expect to get a valid state for ACTIVE case. But future drivers which
> > support multiplexed streaming and routing will depend on getting a state
> > for both active and try cases.
> > 
> > For new drivers we can mandate that the pipelines where the drivers are
> > used need to pass the state properly, or preferably, not call such
> > subdev ops at all.
> > 
> > However, if an existing subdev driver is changed to support multiplexed
> > streams, the driver has to consider cases where its ops will be called
> > with NULL state. The problem can easily be solved by using the
> > v4l2_subdev_lock_and_return_state() helper, introduced here.
> > 
> > Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
> > ---
> >  include/media/v4l2-subdev.h | 31 +++++++++++++++++++++++++++++++
> >  1 file changed, 31 insertions(+)
> > 
> > diff --git a/include/media/v4l2-subdev.h b/include/media/v4l2-subdev.h
> > index 1810dde9c7fc..873bbe0686e3 100644
> > --- a/include/media/v4l2-subdev.h
> > +++ b/include/media/v4l2-subdev.h
> > @@ -1317,4 +1317,35 @@ void v4l2_subdev_lock_state(struct v4l2_subdev_state *state);
> >   */
> >  void v4l2_subdev_unlock_state(struct v4l2_subdev_state *state);
> >  
> > +/**
> > + * v4l2_subdev_lock_and_return_state() - Gets locked TRY or ACTIVE subdev state
> > + * @sd: subdevice
> > + * @state: subdevice state as passed to the subdev op
> > + *
> > + * Due to legacy reasons, when subdev drivers call ops in other subdevs they use
> > + * NULL as the state parameter, as subdevs always used to have their active
> > + * state stored privately.
> > + *
> > + * However, newer state-aware subdev drivers, which store their active state in
> > + * a common place, subdev->active_state, expect to always get a proper state as
> > + * a parameter.
> > + *
> > + * These state-aware drivers can use v4l2_subdev_lock_and_return_state() instead
> > + * of v4l2_subdev_lock_state(). v4l2_subdev_lock_and_return_state() solves the
> > + * issue by using subdev->state in case the passed state is NULL.
> > + *
> > + * This is a temporary helper function, and should be removed when we can ensure
> > + * that all drivers pass proper state when calling other subdevs.
> > + */
> > +static inline struct v4l2_subdev_state *
> > +v4l2_subdev_lock_and_return_state(struct v4l2_subdev *sd,
> > +				  struct v4l2_subdev_state *state)
> > +{
> > +	state = state ? state : sd->active_state;
> 
> Can we add a dev_warn() when state is NULL ? This will help speeding up
> the transition.

Wouldn't this produce lots of warnings? I'd rather use dev_warn_once() to
avoid flooding logs.

> 
> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
> 
> > +
> > +	v4l2_subdev_lock_state(state);
> > +
> > +	return state;
> > +}
> > +
> >  #endif
> 

-- 
Sakari Ailus