On 28/12/2022 15:40, Greg Kroah-Hartman wrote:
> From: Hans Verkuil <hverkuil-cisco@xxxxxxxxx>
>
> [ Upstream commit 73278d483378cf850ade923a1107a70297b2602a ]
>
> The patch adding support for dynamically allocated arrays accidentally
> dropped the line setting ctrl->is_new to 1, thus new string values were
> always ignored.
>
> Fixes: fb582cba4492 ("media: v4l2-ctrls: add support for dynamically allocated arrays.")
> Reported-by: Alice Yuan <alice.yuan@xxxxxxx>
> Signed-off-by: Hans Verkuil <hverkuil-cisco@xxxxxxxxx>
> Signed-off-by: Mauro Carvalho Chehab <mchehab@xxxxxxxxxx>
> Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
> ---
> drivers/media/v4l2-core/v4l2-ctrls-api.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/drivers/media/v4l2-core/v4l2-ctrls-api.c b/drivers/media/v4l2-core/v4l2-ctrls-api.c
> index 50d012ba3c02..7781ebd7ee95 100644
> --- a/drivers/media/v4l2-core/v4l2-ctrls-api.c
> +++ b/drivers/media/v4l2-core/v4l2-ctrls-api.c
> @@ -155,6 +155,7 @@ static int user_to_new(struct v4l2_ext_control *c, struct v4l2_ctrl *ctrl)
> * then return an error.
> */
> if (strlen(ctrl->p_new.p_char) == ctrl->maximum && last)
> + ctrl->is_new = 1;
> return -ERANGE;
This is obviously wrong.
Something got messed up in mainline where my original patch:
https://git.linuxtv.org/hverkuil/media_tree.git/commit/?h=for-v6.1a&id=c33692033135acc96dfdd00dca6e8a392c545b0a
got mangled to the version you have here.
I guess I'll have to post a new patch for mainline fixing the mangled fix.
I only now noticed that you posted the backport patch for review on Dec 25th :-(
Oh well, string control support was broken, and is now even more broken.
Regards,
Hans
> }
> return ret;
