Hi Sakari,

On Thu, Jun 06, 2019 at 04:02:20PM +0300, Sakari Ailus wrote:
> v4l2_async_notifier_add_fwnode_remote_subdev is a convenience function for
> parsing information on V4L2 fwnode subdevs.
>
> Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
> ---
>  drivers/media/v4l2-core/v4l2-async.c | 23 +++++++++++++++++++++++
>  include/media/v4l2-async.h           | 25 +++++++++++++++++++++++++
>  2 files changed, 48 insertions(+)
>
> diff --git a/drivers/media/v4l2-core/v4l2-async.c b/drivers/media/v4l2-core/v4l2-async.c
> index dc4e83b4f6ba..2ea8afbcbf8f 100644
> --- a/drivers/media/v4l2-core/v4l2-async.c
> +++ b/drivers/media/v4l2-core/v4l2-async.c
> @@ -609,6 +609,29 @@ v4l2_async_notifier_add_fwnode_subdev(struct v4l2_async_notifier *notifier,
>  }
>  EXPORT_SYMBOL_GPL(v4l2_async_notifier_add_fwnode_subdev);
>
> +int
> +v4l2_async_notifier_add_fwnode_remote_subdev(struct v4l2_async_notifier *notif,
> +					     struct fwnode_handle *endpoint,
> +					     struct v4l2_async_subdev *asd)
> +{
> +	struct fwnode_handle *remote_ep;
> +	int ret;
> +
> +	remote_ep = fwnode_graph_get_remote_endpoint(endpoint);
> +	if (!remote_ep)
> +		return -ENOTCONN;
> +
> +	asd->match_type = V4L2_ASYNC_MATCH_FWNODE;
> +	asd->match.fwnode = remote_ep;
> +
> +	ret = v4l2_async_notifier_add_subdev(notif, asd);
> +	if (ret)
> +		fwnode_handle_put(remote_ep);
> +
> +	return ret;
> +}
> +EXPORT_SYMBOL_GPL(v4l2_async_notifier_add_fwnode_remote_subdev);
> +
>  struct v4l2_async_subdev *
>  v4l2_async_notifier_add_i2c_subdev(struct v4l2_async_notifier *notifier,
>  				   int adapter_id, unsigned short address,
> diff --git a/include/media/v4l2-async.h b/include/media/v4l2-async.h
> index b9141ffa188a..55ce3c1672a4 100644
> --- a/include/media/v4l2-async.h
> +++ b/include/media/v4l2-async.h
> @@ -185,6 +185,31 @@ v4l2_async_notifier_add_fwnode_subdev(struct v4l2_async_notifier *notifier,
>  				      unsigned int asd_struct_size);
>
>  /**
> + * v4l2_async_notifier_add_fwnode_remote_subdev - Allocate and add a fwnode
> + *						  remote async subdev to the
> + *						  notifier's master asd_list.
> + *
> + * @notif: pointer to &struct v4l2_async_notifier
> + * @endpoint: local endpoint the remote sub-device to be matched

Not sure I get what you mean here, maybe you're missing a "to" between
the "local endpoint" and "the remote sub-device" ?

> + * @asd: Async sub-device struct allocated by the caller. The &struct
> + *	 v4l2_async_subdev shall be the first member of the driver's async
> + *	 sub-device struct, i.e. both begin at the same memory address.
> + *
> + * Gets the remote endpoint of a given local endpoint, set it up for fwnode
> + * matching and add the async sub-device to the notifier's @asd_list. The

and adds

> + * function also gets a reference of the fwnode which is released later at
> + * notifier cleanup time.
> + *
> + * This is just like @v4l2_async_notifier_add_fwnode_subdev, but with the
> + * exception that the fwnode refers to a local endpoint, not the remote one, and
> + * the function relies on the caller to allocate the async sub-device struct.

This makes v4l2_async_notifier_add_fwnode_subdev behave differently from
v4l2_async_notifier_add_subdev and the here introduced
v4l2_async_notifier_add_remote_subdev in the sense that it's the only
one that allocates the asd for the caller. I'm not sure I see the
advantage of having many functions for doing similar things with a
slightly different interface. What is the reason add_fwnode-subdev
allocates the asd for the caller?

Thanks
  j

> + */
> +int
> +v4l2_async_notifier_add_fwnode_remote_subdev(struct v4l2_async_notifier *notif,
> +					     struct fwnode_handle *endpoint,
> +					     struct v4l2_async_subdev *asd);
> +
> +/**
>   * v4l2_async_notifier_add_i2c_subdev - Allocate and add an i2c async
>   *				subdev to the notifier's master asd_list.
>   *
> --
> 2.11.0
>