Re: [PATCH v3 00/29] Rework frame descriptors

[Mattijs Korpershoek <mkorpershoek@kernel.org>]

Hi Tomi,

On Thu, Sep 17, 2026 at 11:04, Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> wrote:

> Hi Mattijs, Sakari,
>
> On 26/08/2026 12:58, Mattijs Korpershoek wrote:
>> Hi Sakari,
>> 
>> Thank you for the series.
>> 
>> On Mon, Aug 24, 2026 at 15:14, Sakari Ailus <sakari.ailus@linux.intel.com> wrote:
>> 
>>> Hi folks,
>>>
>>> This smallish set makes frame descriptors dynamically allocated and
>>> implements a single-entry frame descriptor based on the device's format,
>>> using a new helper called v4l2_subdev_get_frame_desc(). All drivers that
>>> do not obtain their frame descriptor from upstream are converted. The
>>> helper also obtains a frame descriptor for the desired type (parallel or
>>> CSI-2) and checks there's at least one entry there. These checks are
>>> removed from drivers that currently perform them. (Some drivers also check
>>> there's exactly a single frame descriptor entry but I think in most cases
>>> this check could be loosened. That could be done after this set.)
>>>
>>> More formats can be added to df-to-mbus conversion as needed. These are
>>> meant to be initial formats that are enough for typical raw sensors (and
>>> one RGB format, too).
>>>
>>> In the long run this information should probably reside in sub-device
>>> state. This set however avoids having all receiver drivers to work with
>>> sub-device drivers (~ 100 of such exist) that have a single stream and so
>>> do not implement get_frame_desc() op.
>> 
>> With this series, I could sucessfully run a capture using a TI AM69-SK
>> with the Arducam FPD V3Link[1] using the following device tree overlays:
>>      ti/k3-am68-sk-v3link-fusion.dtbo ti/k3-v3link-imx219-0-0.dtbo
> I haven't tested, but I'm a bit surprised if it works. The series misses 
> converting ds90ub960.c, which has custom frame desc code, passing 
> uninitialized v4l2_mbus_frame_desc to the callee, and never freeing 
> anything.
>
> I guess the leak doesn't happen with "normal" number of streams, and the 
> uninitialized fields don't happen to hit anything.

I guess I've been lucky then, in my testing.
I used a single stream, and captured 10 frames.

>
>   Tomi