Re: [PATCH 2/2] media: entity: Add media_entity_pad_from_dt_regs() function

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hej,

Thanks for your feedback.

On 2017-04-28 13:43:39 +0300, Sakari Ailus wrote:
> Hejssan!!!
> 
> On Fri, Apr 28, 2017 at 12:33:23AM +0200, Niklas Söderlund wrote:
> > This is a wrapper around the media entity pad_from_dt_regs operation.
> > 
> > Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@xxxxxxxxxxxx>
> > ---
> >  drivers/media/media-entity.c | 21 +++++++++++++++++++++
> >  include/media/media-entity.h | 22 ++++++++++++++++++++++
> >  2 files changed, 43 insertions(+)
> > 
> > diff --git a/drivers/media/media-entity.c b/drivers/media/media-entity.c
> > index 5640ca29da8c9bbc..6ef76186d552724e 100644
> > --- a/drivers/media/media-entity.c
> > +++ b/drivers/media/media-entity.c
> > @@ -386,6 +386,27 @@ struct media_entity *media_graph_walk_next(struct media_graph *graph)
> >  }
> >  EXPORT_SYMBOL_GPL(media_graph_walk_next);
> >  
> > +int media_entity_pad_from_dt_regs(struct media_entity *entity,
> > +				  int port_reg, int reg, unsigned int *pad)
> > +{
> > +	int ret;
> > +
> > +	if (!entity->ops || !entity->ops->pad_from_dt_regs) {
> > +		*pad = port_reg;
> 
> I don't think we should bind the port number in firmware to a pad in V4L2
> sub-device interface.
> 
> How about looking for a source pad in the entity instead? That's what some
> drivers do.

Sure that sounds like a nice approach, will do this for next version.

Would it make sens to extend this operation with a 'direction' parameter 
which could take either MEDIA_PAD_FL_SOURCE or MEDIA_PAD_FL_SINK and if 
!entity->ops->pad_from_dt_regs find the first pad that matches this 
'direction' and if it do exist add a check to make sure the pad that is 
return matches that 'direction'?

> 
> > +		return 0;
> > +	}
> > +
> > +	ret = entity->ops->pad_from_dt_regs(port_reg, reg, pad);
> > +	if (ret)
> > +		return ret;
> > +
> > +	if (*pad >= entity->num_pads)
> > +		return -EINVAL;
> > +
> > +	return 0;
> > +}
> > +EXPORT_SYMBOL_GPL(media_entity_pad_from_dt_regs);
> > +
> >  /* -----------------------------------------------------------------------------
> >   * Pipeline management
> >   */
> > diff --git a/include/media/media-entity.h b/include/media/media-entity.h
> > index 47efaf4d825e671b..c60a3713d0a21baf 100644
> > --- a/include/media/media-entity.h
> > +++ b/include/media/media-entity.h
> > @@ -820,6 +820,28 @@ struct media_pad *media_entity_remote_pad(struct media_pad *pad);
> >  struct media_entity *media_entity_get(struct media_entity *entity);
> >  
> >  /**
> > + * media_entity_pad_from_dt_regs - Get pad number from DT regs
> > + *
> > + * @entity: The entity
> > + * @port_reg: DT port
> > + * @reg: DT reg
> > + * @pad: Pointer to pad which will be filled in
> > + *
> > + * This function can be used to resolve the media pad number from
> > + * DT port and reg numbers. This is useful for devices which
> > + * uses more complex mappings of media pads then that the
> > + * DT port number is equivalent to the media pad number.
> > + *
> > + * If the entity do not implement the pad_from_dt_regs() operation
> > + * this function assumes DT port is equivalent to media pad number
> > + * and sets @pad to @port_reg.
> > + *
> > + * Return: 0 on success else -EINVAL.
> 
> -EINVAL suggests the user provided bad parameters, but this isn't the case
> here. How about e.g. -ENXIO?


I reasoned that if a port_reg and reg supplied did result in a  pad 
match the user would have given pad parameters. But sure there might be 
cases where that assumtion might not be true. So I see no problem of 
changing this to -ENXIO in next version.

> 
> > + */
> > +int media_entity_pad_from_dt_regs(struct media_entity *entity,
> > +				  int port_reg, int reg, unsigned int *pad);
> > +
> > +/**
> >   * media_graph_walk_init - Allocate resources used by graph walk.
> >   *
> >   * @graph: Media graph structure that will be used to walk the graph
> 
> -- 
> Hälsningar,
> 
> Sakari Ailus
> e-mail: sakari.ailus@xxxxxx	XMPP: sailus@xxxxxxxxxxxxxx

-- 
Regards,
Niklas Söderlund



[Index of Archives]     [Linux Samsung SOC]     [Linux Wireless]     [Linux Kernel]     [ATH6KL]     [Linux Bluetooth]     [Linux Netdev]     [Kernel Newbies]     [IDE]     [Security]     [Git]     [Netfilter]     [Bugtraq]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux ATA RAID]     [Samba]     [Device Mapper]

  Powered by Linux