Re: [PATCH v4 3/4] v4l: of: Parse variable length properties --- link-frequencies

[Sylwester Nawrocki <s.nawrocki@samsung.com>]

On 09/04/15 23:25, Sakari Ailus wrote:
> The link-frequencies property is a variable length array of link frequencies
> in an endpoint. The array is needed by an increasing number of drivers, so
> it makes sense to add it to struct v4l2_of_endpoint.
> 
> However, the length of the array is variable and the size of struct
> v4l2_of_endpoint is fixed since it is allocated by the caller. The options
> here are
> 
> 1. to define a fixed maximum limit of link frequencies that has to be the
> global maximum of all boards. This is seen as problematic since the maximum
> could be largish, and everyone hitting the problem would need to submit a
> patch to fix it, or
> 
> 2. parse the property in every driver. This doesn't sound appealing as two
> of the three implementations submitted to linux-media were wrong, and one of
> them was even merged before this was noticed, or
> 
> 3. change the interface so that allocating and releasing memory according to
> the size of the array is possible. This is what the patch does.
> 
> v4l2_of_alloc_parse_endpoint() is just like v4l2_of_parse_endpoint(), but it
> will allocate the memory resources needed to store struct v4l2_of_endpoint
> and the additional arrays pointed to by this struct. A corresponding release
> function v4l2_of_free_endpoint() is provided to release the memory allocated
> by v4l2_of_alloc_parse_endpoint().
> 
> In addition to this, the link-frequencies property is parsed as well, and
> the result is stored to struct v4l2_of_endpoint field link_frequencies.
> 
> Signed-off-by: Sakari Ailus <sakari.ailus@iki.fi>
> Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>

It's a bit sad we need to introduce the ERR_PTR() patter, nevertheless
I can't see a better alternative now. I think in long term we need to
get rid of v4l2_of_parse_endpoint() function.

Acked-by: Sylwester Nawrocki <s.nawrocki@samsung.com>

-- 
Regards,
Sylwester