On Wed, May 18, 2022 at 06:54:08PM +0300, Sakari Ailus wrote:
> The s_stream callback had several issues:
>
> - If pm_runtime_get_sync() fails, the usage_count is not put.
>
> - The sensor wasn't suspended if s_stream(subdev, 1) failed.
>
> Fix this.
Perhaps late, but nevertheless
Reviewed-by: Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx>
> Fixes: ("media: i2c: Add pm_runtime support to ov7251")
> Signed-off-by: Sakari Ailus <sakari.ailus@xxxxxxxxxxxxxxx>
> ---
> drivers/media/i2c/ov7251.c | 5 ++---
> 1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/media/i2c/ov7251.c b/drivers/media/i2c/ov7251.c
> index 0e7be15bc20a7..603a4c7049e69 100644
> --- a/drivers/media/i2c/ov7251.c
> +++ b/drivers/media/i2c/ov7251.c
> @@ -1340,7 +1340,7 @@ static int ov7251_s_stream(struct v4l2_subdev *subdev, int enable)
> if (enable) {
> ret = pm_runtime_get_sync(ov7251->dev);
> if (ret < 0)
> - goto unlock_out;
> + goto err_power_down;
>
> ret = ov7251_pll_configure(ov7251);
> if (ret) {
> @@ -1372,12 +1372,11 @@ static int ov7251_s_stream(struct v4l2_subdev *subdev, int enable)
> pm_runtime_put(ov7251->dev);
> }
>
> -unlock_out:
> mutex_unlock(&ov7251->lock);
> return ret;
>
> err_power_down:
> - pm_runtime_put_noidle(ov7251->dev);
> + pm_runtime_put(ov7251->dev);
> mutex_unlock(&ov7251->lock);
> return ret;
> }
> --
> 2.30.2
>
--
With Best Regards,
Andy Shevchenko
