2019年1月7日(月) 20:30 Sakari Ailus <sakari.ailus@xxxxxxxxxxxxxxx>:
>
> Hi Mita-san,
>
> On Sun, Dec 23, 2018 at 02:12:53AM +0900, Akinobu Mita wrote:
> > The VIDIOC_SUBDEV_G_FMT ioctl for this driver doesn't recognize
> > V4L2_SUBDEV_FORMAT_TRY and always works as if V4L2_SUBDEV_FORMAT_ACTIVE
> > is specified.
> >
> > Cc: Guennadi Liakhovetski <g.liakhovetski@xxxxxx>
> > Cc: Sakari Ailus <sakari.ailus@xxxxxxxxxxxxxxx>
> > Cc: Mauro Carvalho Chehab <mchehab@xxxxxxxxxxxxxxxx>
> > Signed-off-by: Akinobu Mita <akinobu.mita@xxxxxxxxx>
> > ---
> > drivers/media/i2c/mt9m001.c | 27 +++++++++++++++++++++++++++
> > 1 file changed, 27 insertions(+)
> >
> > diff --git a/drivers/media/i2c/mt9m001.c b/drivers/media/i2c/mt9m001.c
> > index a5b94d7..f4afbc9 100644
> > --- a/drivers/media/i2c/mt9m001.c
> > +++ b/drivers/media/i2c/mt9m001.c
> > @@ -331,6 +331,12 @@ static int mt9m001_get_fmt(struct v4l2_subdev *sd,
> > if (format->pad)
> > return -EINVAL;
> >
> > + if (format->which == V4L2_SUBDEV_FORMAT_TRY) {
> > + mf = v4l2_subdev_get_try_format(sd, cfg, 0);
> > + format->format = *mf;
> > + return 0;
> > + }
> > +
> > mf->width = mt9m001->rect.width;
> > mf->height = mt9m001->rect.height;
> > mf->code = mt9m001->fmt->code;
> > @@ -638,6 +644,26 @@ static const struct v4l2_subdev_core_ops mt9m001_subdev_core_ops = {
> > #endif
> > };
> >
> > +static int mt9m001_init_cfg(struct v4l2_subdev *sd,
> > + struct v4l2_subdev_pad_config *cfg)
> > +{
> > + struct i2c_client *client = v4l2_get_subdevdata(sd);
> > + struct mt9m001 *mt9m001 = to_mt9m001(client);
> > + struct v4l2_mbus_framefmt *try_fmt =
> > + v4l2_subdev_get_try_format(sd, cfg, 0);
> > +
> > + try_fmt->width = mt9m001->rect.width;
> > + try_fmt->height = mt9m001->rect.height;
> > + try_fmt->code = mt9m001->fmt->code;
> > + try_fmt->colorspace = mt9m001->fmt->colorspace;
>
> The initial configuration set here should reflect the default, not current
> configuration. This appears to refer to the current one.
You are right. I'll fix this.
