From: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
To: Hans Verkuil <hverkuil@xs4all.nl>
Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
Sakari Ailus <sakari.ailus@linux.intel.com>,
linux-media@vger.kernel.org
Subject: Re: vb2_queue type question
Date: Wed, 31 Mar 2021 08:22:07 +0300 [thread overview]
Message-ID: <6eae31f1-3293-c771-19b9-03835c564a1e@ideasonboard.com> (raw)
In-Reply-To: <a964df70-612e-6ae3-a183-f1c34a1c4629@xs4all.nl>
On 30/03/2021 11:56, Hans Verkuil wrote:
> On 3/30/21 10:18 AM, Tomi Valkeinen wrote:
>> Hi Hans,
>>
>> On 26/03/2021 12:18, Hans Verkuil wrote:
>>
>>> The only thing that you do is update the queue type when you set the
>>> video or metadata format. When you start allocating buffers the queue type
>>> of the last set format is used. At that moment any attempt to set the format
>>> to another type will fail since vb2_is_busy(queue) will be true.
>>>
>>> So only the s_fmt ioctl will change the type. The g/try_fmt ioctls just must
>>> keep working as-is.
>>
>> I noticed that v4l2-compliance complains about this. It first tests the
>> format ioctls for both video and metadata buffers, and the last s_fmt is
>> for metadata. Then it tests buffer ioctls, and reqbufs for video buffers
>> fails as the queue is in metadata mode, not video mode.
>>
>> I added a custom .vidioc_reqbufs function to the driver which sets the
>> queue type and then calls vb2_ioctl_reqbufs normally. This makes
>> v4l2-compliance pass.
>>
>> But is that correct change, or should v4l2-compliance be changed?
>>
>> Tomi
>>
>
> Good question.
>
> So currently this is something that is rarely used. The few implementations
> of this rely on the last set format to decide what the queue type will be.
>
> But is this actually something you want? Wouldn't it be better to rely on the
> queue type as passed with VIDIOC_REQBUFS/CREATE_BUFS? That's really the moment
> where you lock in the queue type.
>
> To do this you would have to make your own ioctl op:
>
> int my_ioctl_reqbufs(struct file *file, void *priv,
> struct v4l2_requestbuffers *p)
> {
> struct video_device *vdev = video_devdata(file);
>
> if (p->type != V4L2_BUF_TYPE_VIDEO_CAPTURE &&
> p->type != V4L2_BUF_TYPE_META_CAPTURE)
> return -EINVAL;
> if (p->type != vdev->queue.type && vb2_queue_is_busy(vdev, file))
> return -EBUSY;
> vdev->queue.type = p->type;
> return vb2_ioctl_reqbufs(file, priv, p);
> }
>
> And ditto for create_bufs.
>
> I think this makes more sense than relying on the format.
Thanks! This works fine and I agree, it makes more sense.
Tomi
prev parent reply other threads:[~2021-03-31 5:23 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-03-22 9:18 vb2_queue type question Tomi Valkeinen
2021-03-22 9:49 ` Hans Verkuil
2021-03-22 9:59 ` Laurent Pinchart
2021-03-22 10:08 ` Hans Verkuil
2021-03-22 10:30 ` Tomi Valkeinen
2021-03-26 9:57 ` Tomi Valkeinen
2021-03-26 10:18 ` Hans Verkuil
2021-03-30 8:18 ` Tomi Valkeinen
2021-03-30 8:56 ` Hans Verkuil
2021-03-31 5:22 ` Tomi Valkeinen [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=6eae31f1-3293-c771-19b9-03835c564a1e@ideasonboard.com \
--to=tomi.valkeinen@ideasonboard.com \
--cc=hverkuil@xs4all.nl \
--cc=laurent.pinchart@ideasonboard.com \
--cc=linux-media@vger.kernel.org \
--cc=sakari.ailus@linux.intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).