Re: [PATCH] media: v4l2: Fix the _MPLANE format check in v4l_enum_fmt()

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 6/6/19 8:51 AM, Boris Brezillon wrote:
> On Thu, 6 Jun 2019 12:53:57 +0900
> Tomasz Figa <tfiga@xxxxxxxxxxxx> wrote:
> 
>> On Thu, Jun 6, 2019 at 1:46 AM Boris Brezillon
>> <boris.brezillon@xxxxxxxxxxxxx> wrote:
>>>
>>> CAP_M2M_MPLANE means the device supports _MPLANE formats for both
>>> capture and output. Adjust the check to avoid EINVAL errors on
>>> such devices.
>>>
>>> Fixes: 366c719d6479 ("media: v4l2: Get rid of ->vidioc_enum_fmt_vid_{cap,out}_mplane")
>>> Reported-by: Maxime Jourdan <mjourdan@xxxxxxxxxxxx>
>>> Signed-off-by: Boris Brezillon <boris.brezillon@xxxxxxxxxxxxx>
>>> ---
>>>  drivers/media/v4l2-core/v4l2-ioctl.c | 9 +++++++--
>>>  1 file changed, 7 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/media/v4l2-core/v4l2-ioctl.c b/drivers/media/v4l2-core/v4l2-ioctl.c
>>> index b4c73e8f23c5..ace9b9761bed 100644
>>> --- a/drivers/media/v4l2-core/v4l2-ioctl.c
>>> +++ b/drivers/media/v4l2-core/v4l2-ioctl.c
>>> @@ -1385,6 +1385,7 @@ static int v4l_enum_fmt(const struct v4l2_ioctl_ops *ops,
>>>         struct video_device *vdev = video_devdata(file);
>>>         struct v4l2_fmtdesc *p = arg;
>>>         int ret = check_fmt(file, p->type);
>>> +       u32 cap_mask;
>>>
>>>         if (ret)
>>>                 return ret;
>>> @@ -1393,7 +1394,9 @@ static int v4l_enum_fmt(const struct v4l2_ioctl_ops *ops,
>>>         switch (p->type) {
>>>         case V4L2_BUF_TYPE_VIDEO_CAPTURE:
>>>         case V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE:
>>> -               if (!!(vdev->device_caps & V4L2_CAP_VIDEO_CAPTURE_MPLANE) !=
>>> +               cap_mask = V4L2_CAP_VIDEO_CAPTURE_MPLANE |
>>> +                          V4L2_CAP_VIDEO_M2M_MPLANE;
>>> +               if (!!(vdev->device_caps & cap_mask) !=  
>>
>> Shouldn't devices that report V4L2_CAP_VIDEO_M2M_MPLANE have also
>> V4L2_CAP_VIDEO_CAPTURE_MPLANE and V4L2_CAP_VIDEO_OUTPUT_MPLANE
>> reported anyway?
> 
> That's the other option, force drivers that set
> V4L2_CAP_VIDEO_M2M_MPLANE to also set
> V4L2_CAP_VIDEO_{CAPTURE,OUTPUT}_MPLANE (or we can let the core do it).
> 

No, we decided at some point not to do that. The M2M cap makes it explicit
that this is a memory to memory device, and it avoids applications from
trying to use it as a webcam since most apps just check for the VIDEO_CAPTURE
capability.

Regards,

	Hans



[Index of Archives]     [Linux Input]     [Video for Linux]     [Gstreamer Embedded]     [Mplayer Users]     [Linux USB Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]

  Powered by Linux