Hi Sakari,
On Friday 16 March 2012 14:36:53 Sakari Ailus wrote:
> On Fri, Mar 16, 2012 at 01:31:39PM +0100, Laurent Pinchart wrote:
> > On Friday 16 March 2012 14:12:11 Sakari Ailus wrote:
> > > On Fri, Mar 16, 2012 at 12:58:30PM +0100, Laurent Pinchart wrote:
> > > > On Thursday 15 March 2012 23:01:39 Sakari Ailus wrote:
> > > > > Provide pixel rate control calculated from external clock and
> > > > > horizontal
> > > > > binning factor.
> > > > >
> > > > > Signed-off-by: Sakari Ailus <sakari.ailus@xxxxxx>
> > > > > ---
> > > > >
> > > > > drivers/media/video/mt9v032.c | 35
> > > > > ++++++++++++++++++++++++++++++++-
> > > > > 1 files changed, 34 insertions(+), 1 deletions(-)
> > > > >
> > > > > diff --git a/drivers/media/video/mt9v032.c
> > > > > b/drivers/media/video/mt9v032.c
> > > > > index 75e253a..e530e8d 100644
> > > > > --- a/drivers/media/video/mt9v032.c
> > > > > +++ b/drivers/media/video/mt9v032.c
> >
> > [snip]
> >
> > > > > +static void mt9v032_configure_pixel_rate(struct v4l2_subdev
> > > > > *subdev,
> > > > > + unsigned int hratio)
> > > > > +{
> > > > > + struct i2c_client *client = v4l2_get_subdevdata(subdev);
> > > > > + struct mt9v032 *mt9v032 = to_mt9v032(subdev);
> > > > > + struct v4l2_ext_controls ctrls;
> > > > > + struct v4l2_ext_control ctrl;
> > > > > +
> > > > > + memset(&ctrls, 0, sizeof(ctrls));
> > > > > + memset(&ctrl, 0, sizeof(ctrl));
> > > > > +
> > > > > + ctrls.count = 1;
> > > > > + ctrls.controls = &ctrl;
> > > > > +
> > > > > + ctrl.id = V4L2_CID_PIXEL_RATE;
> > > > > + ctrl.value64 = EXT_CLK / hratio;
> > > > > +
> > > > > + if (v4l2_s_ext_ctrls(mt9v032->pixel_rate->ctrl_handler, &ctrls)
<
> > > > > 0)
> > > > > + dev_warn(&client->dev, "bug: failed to set pixel rate\n");
> > > >
> > > > What about just calling v4l2_ctrl_s_ctrl() ?
> > >
> > > It's a 64-bit integer control, so it has to be set using
> > > v4l2_s_ext_ctrls().>
> > What about extending v4l2_ctrl_s_ctrl() to support 64-bit integer controls
> > then ?
>
> The second argument to that function is struct v4l2_control:
>
> struct v4l2_control {
> __u32 id;
> __s32 value;
> };
>
> So there's no chance to extend it. This must also be the reason why 64-bit
> controls require using extended controls.
>
> What we could do is to introduce v4l2_s_ext_ctrl without the "s" in the end
> to just set one extended control. I think that would be a reasonable
> approach, as it seems we need the functionality in multiple places.
We're talking about different functions.
int v4l2_ctrl_s_ctrl(struct v4l2_ctrl *ctrl, s32 val);
Extending that one wouldn't break the userspace API. We could use s64 instead
of s32, or add a new function such as v4l2_ctrl_s_ctrl64().
--
Regards,
Laurent Pinchart
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
