Re: [PATCH v8 1/9] media: v4l: Support passing media pad argument to v4l2_get_link_freq()

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

Moi,

On Fri, Dec 20, 2024 at 03:09:05PM +0200, Tomi Valkeinen wrote:
> Hi,
> 
> On 17/12/2024 23:54, Sakari Ailus wrote:
> > v4l2_get_link_freq() accepts a V4L2 control handler for now, but it needs
> > to take struct media_pad argument in order to obtain the link frequency
> > using get_mbus_config() pad op. Prepare for this by allowing struct
> > media_pad as well.
> > 
> > Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
> > ---
> >   drivers/media/v4l2-core/v4l2-common.c | 21 ++++++++++++++++++---
> >   include/media/v4l2-common.h           | 19 ++++++++++++++++---
> >   2 files changed, 34 insertions(+), 6 deletions(-)
> > 
> > diff --git a/drivers/media/v4l2-core/v4l2-common.c b/drivers/media/v4l2-core/v4l2-common.c
> > index 0a2f4f0d0a07..9fe74c7e064f 100644
> > --- a/drivers/media/v4l2-core/v4l2-common.c
> > +++ b/drivers/media/v4l2-core/v4l2-common.c
> > @@ -466,8 +466,8 @@ int v4l2_fill_pixfmt(struct v4l2_pix_format *pixfmt, u32 pixelformat,
> >   }
> >   EXPORT_SYMBOL_GPL(v4l2_fill_pixfmt);
> > -s64 v4l2_get_link_freq(struct v4l2_ctrl_handler *handler, unsigned int mul,
> > -		       unsigned int div)
> > +s64 __v4l2_get_link_freq_ctrl(struct v4l2_ctrl_handler *handler,
> > +			      unsigned int mul, unsigned int div)
> >   {
> >   	struct v4l2_ctrl *ctrl;
> >   	s64 freq;
> > @@ -502,7 +502,22 @@ s64 v4l2_get_link_freq(struct v4l2_ctrl_handler *handler, unsigned int mul,
> >   	return freq > 0 ? freq : -EINVAL;
> >   }
> > -EXPORT_SYMBOL_GPL(v4l2_get_link_freq);
> > +EXPORT_SYMBOL_GPL(__v4l2_get_link_freq_ctrl);
> > +
> > +#ifdef CONFIG_MEDIA_CONTROLLER
> > +s64 __v4l2_get_link_freq_pad(struct media_pad *pad, unsigned int mul,
> > +			     unsigned int div)
> > +{
> > +	struct v4l2_subdev *sd;
> > +
> > +	sd = media_entity_to_v4l2_subdev(pad->entity);
> > +	if (!sd)
> > +		return -ENODEV;
> > +
> > +	return __v4l2_get_link_freq_ctrl(sd->ctrl_handler, mul, div);
> > +}
> > +EXPORT_SYMBOL_GPL(__v4l2_get_link_freq_pad);
> > +#endif /* CONFIG_MEDIA_CONTROLLER */
> >   /*
> >    * Simplify a fraction using a simple continued fraction decomposition. The
> > diff --git a/include/media/v4l2-common.h b/include/media/v4l2-common.h
> > index 63ad36f04f72..fda903bb3674 100644
> > --- a/include/media/v4l2-common.h
> > +++ b/include/media/v4l2-common.h
> > @@ -525,7 +525,8 @@ int v4l2_fill_pixfmt_mp(struct v4l2_pix_format_mplane *pixfmt, u32 pixelformat,
> >   /**
> >    * v4l2_get_link_freq - Get link rate from transmitter
> >    *
> > - * @handler: The transmitter's control handler
> > + * @pad: The transmitter's media pad (or control handler for non-MC users or
> > + *	 compatibility reasons, don't use in new code)
> >    * @mul: The multiplier between pixel rate and link frequency. Bits per pixel on
> >    *	 D-PHY, samples per clock on parallel. 0 otherwise.
> >    * @div: The divisor between pixel rate and link frequency. Number of data lanes
> > @@ -541,8 +542,20 @@ int v4l2_fill_pixfmt_mp(struct v4l2_pix_format_mplane *pixfmt, u32 pixelformat,
> >    * * %-ENOENT: Link frequency or pixel rate control not found
> >    * * %-EINVAL: Invalid link frequency value
> >    */
> > -s64 v4l2_get_link_freq(struct v4l2_ctrl_handler *handler, unsigned int mul,
> > -		       unsigned int div);
> > +#ifdef CONFIG_MEDIA_CONTROLLER
> > +#define v4l2_get_link_freq(pad, mul, div)				\
> > +	_Generic(pad,							\
> > +		 struct media_pad *: __v4l2_get_link_freq_pad,		\
> > +		 struct v4l2_ctrl_handler *: __v4l2_get_link_freq_ctrl)	\
> > +	(pad, mul, div)
> 
> I can't decide if I love or hate _Generic =).

I really like it. It enables certain things that didn't use to be possible
in C, and really helps in transitions like this. We also have
container_of_const() etc. for added constness checking.

> 
> > +s64 __v4l2_get_link_freq_pad(struct media_pad *pad, unsigned int mul,
> > +			     unsigned int div);
> > +#else
> > +#define v4l2_get_link_freq(handler, mul, div)		\
> > +	__v4l2_get_link_freq_ctrl(handler, mul, div)
> > +#endif
> > +s64 __v4l2_get_link_freq_ctrl(struct v4l2_ctrl_handler *handler,
> > +			      unsigned int mul, unsigned int div);
> >   void v4l2_simplify_fraction(u32 *numerator, u32 *denominator,
> >   		unsigned int n_terms, unsigned int threshold);
> 
> This makes sense in the transition period. But I wonder, should we later
> have:
> 
> v4l2_get_link_freq(struct v4l2_subdev *sd, unsigned int pad, ...) like all
> the subdev ops? Perhaps not... Subdev ops take subdev as the first argument
> as, well, they're subdev ops.
> 
> It did look to me that in many drivers we already have the source subdev and
> pad stored somewhere, and getting the media_pad to call v4l2_get_link_freq()
> is an extra step.

I could add a couple of macros to cover that need. ;-) But to be fair,
drivers should instead store the media pad of the sensor or even better,
not store it at all but use media_entity_remote_pad_unique() to obtain it
at runtime.

-- 
Terveisin,

Sakari Ailus