Hi Laurent,
On 8/30/23 12:49 PM, Laurent Pinchart wrote:
On Wed, Aug 30, 2023 at 10:23:22AM +0530, Umang Jain wrote:
The imx214 driver uses the subdev active state, there's no
Does it ? I don't see a call to v4l2_subdev_init_finalize(), and the
imx214 struct has fmt and crop fields in which it stores the active
format.
Yes, It doesn't. Got confused with one of the pad ops (.init_cfg) which
led me thinking this driver uses active state, whereas I should have
checked v4l2_subdev_init_finalize()
Apologies for the noise.
This patch should be discarded.
need to implement the .get_fmt() operation manually. Use the
v4l2_subdev_get_fmt() helper instead.
Signed-off-by: Umang Jain <umang.jain@xxxxxxxxxxxxxxxx>
---
drivers/media/i2c/imx214.c | 17 +----------------
1 file changed, 1 insertion(+), 16 deletions(-)
diff --git a/drivers/media/i2c/imx214.c b/drivers/media/i2c/imx214.c
index 2f9c8582f940..d8bdc52f552a 100644
--- a/drivers/media/i2c/imx214.c
+++ b/drivers/media/i2c/imx214.c
@@ -548,21 +548,6 @@ __imx214_get_pad_format(struct imx214 *imx214,
}
}
-static int imx214_get_format(struct v4l2_subdev *sd,
- struct v4l2_subdev_state *sd_state,
- struct v4l2_subdev_format *format)
-{
- struct imx214 *imx214 = to_imx214(sd);
-
- mutex_lock(&imx214->mutex);
- format->format = *__imx214_get_pad_format(imx214, sd_state,
- format->pad,
- format->which);
- mutex_unlock(&imx214->mutex);
-
- return 0;
-}
-
static struct v4l2_rect *
__imx214_get_pad_crop(struct imx214 *imx214,
struct v4l2_subdev_state *sd_state,
@@ -842,7 +827,7 @@ static const struct v4l2_subdev_pad_ops imx214_subdev_pad_ops = {
.enum_mbus_code = imx214_enum_mbus_code,
.enum_frame_size = imx214_enum_frame_size,
.enum_frame_interval = imx214_enum_frame_interval,
- .get_fmt = imx214_get_format,
+ .get_fmt = v4l2_subdev_get_fmt,
.set_fmt = imx214_set_format,
.get_selection = imx214_get_selection,
.init_cfg = imx214_entity_init_cfg,
