linux-kernel - Re: [PATCH v5 10/13] media: uvcvideo: Return -EACCES to inactive controls

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 for Android: free password hash cracker in your pocket

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

Date:   Wed, 17 Mar 2021 12:26:40 +0100
From:   Hans Verkuil <hverkuil-cisco@...all.nl>
To:     Ricardo Ribalda <ribalda@...omium.org>,
        Laurent Pinchart <laurent.pinchart@...asonboard.com>,
        Mauro Carvalho Chehab <mchehab@...nel.org>,
        Sergey Senozhatsky <sergey.senozhatsky@...il.com>,
        linux-media@...r.kernel.org, linux-kernel@...r.kernel.org,
        tfiga@...omium.org
Cc:     Hans Verkuil <hverkuil@...all.nl>
Subject: Re: [PATCH v5 10/13] media: uvcvideo: Return -EACCES to inactive
 controls

On 16/03/2021 19:00, Ricardo Ribalda wrote:
> If a control is inactive return -EACCES to let the userspace know that
> the value will not be applied automatically when the control is active
> again.
> 
> Signed-off-by: Ricardo Ribalda <ribalda@...omium.org>
> Suggested-by: Hans Verkuil <hverkuil@...all.nl>
> ---
>  drivers/media/usb/uvc/uvc_ctrl.c | 17 ++++++++++++++++-
>  1 file changed, 16 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/media/usb/uvc/uvc_ctrl.c b/drivers/media/usb/uvc/uvc_ctrl.c
> index ba14733db757..98614e1be829 100644
> --- a/drivers/media/usb/uvc/uvc_ctrl.c
> +++ b/drivers/media/usb/uvc/uvc_ctrl.c
> @@ -1578,6 +1578,18 @@ int uvc_ctrl_begin(struct uvc_video_chain *chain)
>  	return mutex_lock_interruptible(&chain->ctrl_mutex) ? -ERESTARTSYS : 0;
>  }
>  
> +static bool uvc_ctrl_is_inactive(struct uvc_control *ctrl)

This doesn't test if the control is inactive, it tests if it *might* be
inactive. To test if it is really inactive would require checking the value
of the master control.

> +{
> +	struct uvc_control_mapping *map;
> +
> +	list_for_each_entry(map, &ctrl->info.mappings, list) {
> +		if (map->master_id)
> +			return true;
> +	}
> +
> +	return false;
> +}
> +
>  static int uvc_ctrl_commit_entity(struct uvc_device *dev,
>  	struct uvc_entity *entity, int rollback)
>  {
> @@ -1621,8 +1633,11 @@ static int uvc_ctrl_commit_entity(struct uvc_device *dev,
>  
>  		ctrl->dirty = 0;
>  
> -		if (ret < 0)
> +		if (ret < 0) {
> +			if (uvc_ctrl_is_inactive(ctrl))
> +				return -EACCES;

So here you assume that if setting the control failed, and if the control
might be inactive, then it probably was inactive.

I feel a bit uncomfortable by this assumption.

Regards,

	Hans

>  			return ret;
> +		}
>  	}
>  
>  	return 0;
>