Hi Hans,
On Thursday 08 May 2014 10:44:12 Hans Verkuil wrote:
> Hi Laurent,
>
> The patch is correct, but I noticed a pre-existing bug that should be
> fixed. See below.
>
> On 05/07/14 17:40, Laurent Pinchart wrote:
> > The digital side of the Black Level Calibration (BLC) function must be
> > disabled when generating a test pattern to avoid artifacts in the image.
> > The driver disables BLC correctly at the hardware level, but the feature
> > gets reenabled by v4l2_ctrl_handler_setup() the next time the device is
> > powered on.
> >
> > Fix this by marking the BLC controls as inactive when generating a test
> > pattern, and ignoring control set requests on inactive controls.
> >
> > Signed-off-by: Laurent Pinchart <laurent.pinchart@xxxxxxxxxxxxxxxx>
> > ---
> >
> > drivers/media/i2c/mt9p031.c | 17 +++++++++++++----
> > 1 file changed, 13 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/media/i2c/mt9p031.c b/drivers/media/i2c/mt9p031.c
> > index 33daace..9102b23 100644
> > --- a/drivers/media/i2c/mt9p031.c
> > +++ b/drivers/media/i2c/mt9p031.c
> > @@ -655,6 +655,9 @@ static int mt9p031_s_ctrl(struct v4l2_ctrl *ctrl)
> > u16 data;
> > int ret;
> >
> > + if (ctrl->flags & V4L2_CTRL_FLAG_INACTIVE)
> > + return 0;
> > +
> > switch (ctrl->id) {
> > case V4L2_CID_EXPOSURE:
> > ret = mt9p031_write(client, MT9P031_SHUTTER_WIDTH_UPPER,
> > @@ -709,8 +712,16 @@ static int mt9p031_s_ctrl(struct v4l2_ctrl *ctrl)
> > MT9P031_READ_MODE_2_ROW_MIR, 0);
> >
> > case V4L2_CID_TEST_PATTERN:
> > + /* The digital side of the Black Level Calibration function must
> > + * be disabled when generating a test pattern to avoid artifacts
> > + * in the image. Activate (deactivate) the BLC-related controls
> > + * when the test pattern is enabled (disabled).
> > + */
> > + v4l2_ctrl_activate(mt9p031->blc_auto, ctrl->val == 0);
> > + v4l2_ctrl_activate(mt9p031->blc_offset, ctrl->val == 0);
> > +
> > if (!ctrl->val) {
> > - /* Restore the black level compensation settings. */
> > + /* Restore the BLC settings. */
> > if (mt9p031->blc_auto->cur.val != 0) {
> > ret = mt9p031_s_ctrl(mt9p031->blc_auto);
>
> This doesn't do what you expect. What you want is to set the blc_auto
> control to the current value, but mt9p031_s_ctrl(mt9p031->blc_auto) will
> set it to the *new* value, which may not be the same. Ditto for doing the
> same for blc_offset.
>
> It's best to just call mt9p031_write directly, rather than going through
> mt9p031_s_ctrl.
Thanks for catching the problem. I'll fix that in a follow-up patch.
> > if (ret < 0)
> >
> > @@ -735,9 +746,7 @@ static int mt9p031_s_ctrl(struct v4l2_ctrl *ctrl)
> > if (ret < 0)
> > return ret;
> >
> > - /* Disable digital black level compensation when using a test
> > - * pattern.
> > - */
> > + /* Disable digital BLC when generating a test pattern. */
> > ret = mt9p031_set_mode2(mt9p031, MT9P031_READ_MODE_2_ROW_BLC,
> > 0);
> > if (ret < 0)
--
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
