On 4/12/19 11:20 AM, Boris Brezillon wrote:
> The fimc-isp-video.c and fimc-lite.c were missing the
> V4L2_CAP_VIDEO_CAPTURE_MPLANE flag when reporting device caps.
>
> Signed-off-by: Boris Brezillon <boris.brezillon@xxxxxxxxxxxxx>
> ---
> Changes in v3:
> - New patch
> ---
> drivers/media/platform/exynos4-is/fimc-isp-video.c | 4 +++-
> drivers/media/platform/exynos4-is/fimc-lite.c | 2 +-
> 2 files changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/media/platform/exynos4-is/fimc-isp-video.c b/drivers/media/platform/exynos4-is/fimc-isp-video.c
> index bb35a2017f21..8cd7b1c6565c 100644
> --- a/drivers/media/platform/exynos4-is/fimc-isp-video.c
> +++ b/drivers/media/platform/exynos4-is/fimc-isp-video.c
> @@ -349,7 +349,9 @@ static int isp_video_querycap(struct file *file, void *priv,
> {
> struct fimc_isp *isp = video_drvdata(file);
>
> - __fimc_vidioc_querycap(&isp->pdev->dev, cap, V4L2_CAP_STREAMING);
> + __fimc_vidioc_querycap(&isp->pdev->dev, cap,
> + V4L2_CAP_STREAMING |
> + V4L2_CAP_VIDEO_CAPTURE_MPLANE);
> return 0;
> }
>
> diff --git a/drivers/media/platform/exynos4-is/fimc-lite.c b/drivers/media/platform/exynos4-is/fimc-lite.c
> index 96f0a8a0dcae..13dfa5e39139 100644
> --- a/drivers/media/platform/exynos4-is/fimc-lite.c
> +++ b/drivers/media/platform/exynos4-is/fimc-lite.c
> @@ -659,7 +659,7 @@ static int fimc_lite_querycap(struct file *file, void *priv,
> snprintf(cap->bus_info, sizeof(cap->bus_info), "platform:%s",
> dev_name(&fimc->pdev->dev));
>
> - cap->device_caps = V4L2_CAP_STREAMING;
> + cap->device_caps = V4L2_CAP_VIDEO_CAPTURE_MPLANE | V4L2_CAP_STREAMING;
> cap->capabilities = cap->device_caps | V4L2_CAP_DEVICE_CAPS;
That's not what should happen. The driver should set device_caps in the struct
video_device, and then it can drop setting cap->device_caps and cap->capabilities
from the querycap implementation (it will be set by the v4l2 core based on the
vdev->device_caps value).
It definitely is a (surprising!) bug that V4L2_CAP_VIDEO_CAPTURE_MPLANE wasn't
set here, but the real thing that you should check MPLANE capable drivers for is
that they set vdev->device_caps.
Regards,
Hans
> return 0;
> }
>
