linux-kernel - Re: [PATCH] vb2: V4L2_BUF_FLAG

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Mon, 20 Jun 2016 11:53:51 +0200
From:	Hans Verkuil <hverkuil@...all.nl>
To:	Ricardo Ribalda Delgado <ricardo.ribalda@...il.com>,
	Pawel Osciak <pawel@...iak.com>,
	Marek Szyprowski <m.szyprowski@...sung.com>,
	Kyungmin Park <kyungmin.park@...sung.com>,
	Mauro Carvalho Chehab <mchehab@...nel.org>,
	linux-media@...r.kernel.org, linux-kernel@...r.kernel.org,
	hans.verkuil@...co.com, Dimitrios Katsaros <patcherwork@...il.com>
Subject: Re: [PATCH] vb2: V4L2_BUF_FLAG_DONE is set after DQBUF

On 06/20/2016 11:01 AM, Ricardo Ribalda Delgado wrote:
> According to the doc, V4L2_BUF_FLAG_DONE is cleared after DQBUF:
> 
> V4L2_BUF_FLAG_DONE 0x00000004  ... After calling the VIDIOC_QBUF or
> VIDIOC_DQBUF it is always cleared ...
> 
> Unfortunately, it seems that videobuf2 keeps it set after DQBUF. This
> can be tested with vivid and dev_debug:
> 
> [257604.338082] video1: VIDIOC_DQBUF: 71:33:25.00260479 index=3,
> type=vid-cap, flags=0x00002004, field=none, sequence=163,
> memory=userptr, bytesused=460800, offset/userptr=0x344b000,
> length=460800
> 
> This patch changes the order when fill_user_buffer() is called,
> to follow the documentation.
> 
> Reported-by: Dimitrios Katsaros <patcherwork@...il.com>
> Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@...il.com>

Sorry, this won't work. Calling __vb2_dqbuf will overwrite the state
VB2_BUF_STATE_ERROR and so the V4L2_BUF_FLAG_ERROR flag will never be
set. The same is true for the last 'if' in the __fill_v4l2_buffer()
function.

I think it might be better to keep this code and instead change the
vb2_internal_dqbuf function to just clear the DONE flag after calling
vb2_core_dqbuf.

It would be nice to have a v4l2_compliance check for this as well.

Regards,

	Hans

> ---
>  drivers/media/v4l2-core/videobuf2-core.c | 8 ++++----
>  1 file changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/media/v4l2-core/videobuf2-core.c b/drivers/media/v4l2-core/videobuf2-core.c
> index 633fc1ab1d7a..63981f28075e 100644
> --- a/drivers/media/v4l2-core/videobuf2-core.c
> +++ b/drivers/media/v4l2-core/videobuf2-core.c
> @@ -1771,10 +1771,6 @@ int vb2_core_dqbuf(struct vb2_queue *q, unsigned int *pindex, void *pb,
>  	if (pindex)
>  		*pindex = vb->index;
>  
> -	/* Fill buffer information for the userspace */
> -	if (pb)
> -		call_void_bufop(q, fill_user_buffer, vb, pb);
> -
>  	/* Remove from videobuf queue */
>  	list_del(&vb->queued_entry);
>  	q->queued_count--;
> @@ -1784,6 +1780,10 @@ int vb2_core_dqbuf(struct vb2_queue *q, unsigned int *pindex, void *pb,
>  	/* go back to dequeued state */
>  	__vb2_dqbuf(vb);
>  
> +	/* Fill buffer information for the userspace */
> +	if (pb)
> +		call_void_bufop(q, fill_user_buffer, vb, pb);
> +
>  	dprintk(1, "dqbuf of buffer %d, with state %d\n",
>  			vb->index, vb->state);
>  
>