On 28/08/2024 14:51, Ricardo Ribalda Delgado wrote:
> Hi Hans
>
> On Fri, Aug 9, 2024 at 10:14 AM Hans Verkuil <hverkuil-cisco@xxxxxxxxx> wrote:
>>
>> From: Erling Ljunggren <hljunggr@xxxxxxxxx>
>>
>> Add capability flag to indicate that the device is an EDID-only device.
>>
>> Signed-off-by: Erling Ljunggren <hljunggr@xxxxxxxxx>
>> Signed-off-by: Hans Verkuil <hverkuil-cisco@xxxxxxxxx>
>> ---
>> include/uapi/linux/videodev2.h | 1 +
>> 1 file changed, 1 insertion(+)
>>
>> diff --git a/include/uapi/linux/videodev2.h b/include/uapi/linux/videodev2.h
>> index 4e91362da6da..987c821aed79 100644
>> --- a/include/uapi/linux/videodev2.h
>> +++ b/include/uapi/linux/videodev2.h
>> @@ -502,6 +502,7 @@ struct v4l2_capability {
>> #define V4L2_CAP_META_CAPTURE 0x00800000 /* Is a metadata capture device */
>>
>> #define V4L2_CAP_READWRITE 0x01000000 /* read/write systemcalls */
>> +#define V4L2_CAP_EDID 0x02000000 /* Is an EDID-only device */
>
> Would it make sense to add a check for "EDID-only", I mean, if the
> driver sets this cap, then it should not set V4L2_CAP_STREAMING or others.
>
> The test could be in the core or even in v4l2-compliance.
Good point. It is easy to add this to v4l2-compliance.
I noticed that the v4l2 core doesn't check for invalid caps combinations. It is
left to v4l2-compliance to check that. So I think I'll keep the check there.
Regards,
Hans
>
> Regards!
>
>
>> #define V4L2_CAP_STREAMING 0x04000000 /* streaming I/O ioctls */
>> #define V4L2_CAP_META_OUTPUT 0x08000000 /* Is a metadata output device */
>>
>> --
>> 2.43.0
>>
>>
>
>
> --
> Ricardo Ribalda
>
