linux-kernel - Re: [PATCH 1/5] Bluetooth: Add RFCOMM TTY write return error codes

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]

Message-ID: <51F05F49.60609@hurleysoftware.com>
Date:	Wed, 24 Jul 2013 19:12:09 -0400
From:	Peter Hurley <peter@...leysoftware.com>
To:	Dean Jenkins <Dean_Jenkins@...tor.com>
CC:	Andre Naujoks <nautsch2@...il.com>, linux-kernel@...r.kernel.org,
	Jiri Slaby <jslaby@...e.cz>,
	Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: Re: [PATCH 1/5] Bluetooth: Add RFCOMM TTY write return error codes

On 07/02/2013 11:31 AM, Dean Jenkins wrote:
> It appears that rfcomm_tty_write() does not check that the
> passed in TTY device_data is not NULL and also does not check
> that the RFCOMM DLC serial data link pointer is not NULL.
>
> A kernel crash was observed whilst SLIP was bound to /dev/rfcomm0
> but the /dev/rfcomm0 had subsequently disconnected. Unfortunately,
> SLIP attempted to write to the now non-existant RFCOMM TTY device
> which caused a NULL pointer dereference because the device_data
> no longer existed.
>
> Therefore, add NULL pointer checks for the dev and dlc pointers
> and output kernel error debug to show that NULL had been detected.

Dean,

Sorry, I didn't see these until just now.

> Signed-off-by: Dean Jenkins <Dean_Jenkins@...tor.com>
> ---
>   net/bluetooth/rfcomm/tty.c | 14 +++++++++++++-
>   1 file changed, 13 insertions(+), 1 deletion(-)
>
> diff --git a/net/bluetooth/rfcomm/tty.c b/net/bluetooth/rfcomm/tty.c
> index b6e44ad..56d28d1 100644
> --- a/net/bluetooth/rfcomm/tty.c
> +++ b/net/bluetooth/rfcomm/tty.c
> @@ -761,12 +761,24 @@ static void rfcomm_tty_close(struct tty_struct *tty, struct file *filp)
>   static int rfcomm_tty_write(struct tty_struct *tty, const unsigned char *buf, int count)
>   {
>   	struct rfcomm_dev *dev = (struct rfcomm_dev *) tty->driver_data;
> -	struct rfcomm_dlc *dlc = dev->dlc;
> +	struct rfcomm_dlc *dlc;
>   	struct sk_buff *skb;
>   	int err = 0, sent = 0, size;
>
>   	BT_DBG("tty %p count %d", tty, count);
>
> +	if (!dev) {
> +		BT_ERR("RFCOMM TTY device data structure does not exist");
> +		return -ENODEV;
> +	}
> +
> +	dlc = dev->dlc;
> +
> +	if (!dlc) {
> +		BT_ERR("RFCOMM serial data link does not exist");
> +		return -ENOLINK;
> +	}
> +

A number of bugs in rfcomm contributed to the crash you observed.
Several other reporters have also noted crashes stemming from
device disconnect. The rfcomm device must not disappear while a tty
is in-use.

Fixes for these are in-progress in a patch series by Gianluca Anzolin
on linux-bluetooth ML.

Also, for the future, kernel bluetooth patches go to:
   Gustavo Padovan <gustavo@...ovan.org>
with cc's to:
   Johan Hedberg <johan.hedberg@...il.com>
   Marcel Holtmann <marcel@...tmann.org>
   linux-bluetooth@...r.kernel.org

Regards,
Peter Hurley

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/