linux-kernel - Re: [PATCH] usb: iowarrior: fix access to freed data structure

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] [day] [month] [year] [list]

Message-ID: <20191009073607.GB27819@localhost>
Date:   Wed, 9 Oct 2019 09:36:07 +0200
From:   Johan Hovold <johan@...nel.org>
To:     Valentin Vidic <vvidic@...entin-vidic.from.hr>
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Oliver Neukum <oneukum@...e.com>, linux-usb@...r.kernel.org,
        linux-kernel@...r.kernel.org,
        syzbot+0761012cebf7bdb38137@...kaller.appspotmail.com
Subject: Re: [PATCH] usb: iowarrior: fix access to freed data structure

On Wed, Oct 09, 2019 at 12:23:07AM +0200, Valentin Vidic wrote:
> struct iowarrior gets freed prematurely in iowarrior_release while
> it is still being referenced from usb_interface, so let only
> iowarrior_disconnect call iowarrior_delete.

The proposed fix is broken since release() may happen long after
disconnect(), in which case you'd now leak memory.

> Fixes: KMSAN: uninit-value in iowarrior_disconnect

Fixes tags are supposed to refer to the commit introducing the issue.

> Reported-by: syzbot+0761012cebf7bdb38137@...kaller.appspotmail.com
> Signed-off-by: Valentin Vidic <vvidic@...entin-vidic.from.hr>
> ---
>  drivers/usb/misc/iowarrior.c | 35 +++++++++++++++--------------------
>  1 file changed, 15 insertions(+), 20 deletions(-)
> 
> diff --git a/drivers/usb/misc/iowarrior.c b/drivers/usb/misc/iowarrior.c
> index f5bed9f29e56..0492ea76c4bf 100644
> --- a/drivers/usb/misc/iowarrior.c
> +++ b/drivers/usb/misc/iowarrior.c
> @@ -638,7 +638,6 @@ static int iowarrior_open(struct inode *inode, struct file *file)
>  static int iowarrior_release(struct inode *inode, struct file *file)
>  {
>  	struct iowarrior *dev;
> -	int retval = 0;
>  
>  	dev = file->private_data;
>  	if (!dev)
> @@ -650,27 +649,23 @@ static int iowarrior_release(struct inode *inode, struct file *file)
>  	mutex_lock(&dev->mutex);
>  
>  	if (dev->opened <= 0) {
> -		retval = -ENODEV;	/* close called more than once */
>  		mutex_unlock(&dev->mutex);
> -	} else {
> -		dev->opened = 0;	/* we're closing now */
> -		retval = 0;
> -		if (dev->present) {
> -			/*
> -			   The device is still connected so we only shutdown
> -			   pending read-/write-ops.
> -			 */
> -			usb_kill_urb(dev->int_in_urb);
> -			wake_up_interruptible(&dev->read_wait);
> -			wake_up_interruptible(&dev->write_wait);
> -			mutex_unlock(&dev->mutex);
> -		} else {
> -			/* The device was unplugged, cleanup resources */
> -			mutex_unlock(&dev->mutex);
> -			iowarrior_delete(dev);

So you cannot just remove this bit.

> -		}
> +		return -ENODEV;	/* close called more than once */
>  	}
> -	return retval;
> +
> +	dev->opened = 0;	/* we're closing now */
> +	if (dev->present) {
> +		/*
> +		 * The device is still connected so we only shutdown
> +		 * pending read/write ops.
> +		 */
> +		usb_kill_urb(dev->int_in_urb);
> +		wake_up_interruptible(&dev->read_wait);
> +		wake_up_interruptible(&dev->write_wait);
> +	}
> +
> +	mutex_unlock(&dev->mutex);
> +	return 0;
>  }
>  
>  static __poll_t iowarrior_poll(struct file *file, poll_table * wait)

Johan