[PATCH v4] bluetooth: Fix WARNING in tty_set_termios()

[PATCH v4] bluetooth: Fix WARNING in tty_set_termios()

[Shuah Khan]

tty_set_termios() has the following WARN_ON which can be triggered with a
syscall to invoke TIOCSETD __NR_ioctl.

WARN_ON(tty->driver->type == TTY_DRIVER_TYPE_PTY &&
                tty->driver->subtype == PTY_TYPE_MASTER);
Reference: https://syzkaller.appspot.com/bug?id=2410d22f1d8e5984217329dd0884b01d99e3e48d

"The problem started with commit 7721383f4199 ("Bluetooth: hci_uart: Support
operational speed during setup") which introduced a new way for how
tty_set_termios() could end up being called for a master pty."

Fix it by by preventing setting the HCI line discipline for PTYs in
hci_uart_tty_open().

Reported-by: syzbot+a950165cbb86bdd023a4@syzkaller.appspotmail.com
Cc: Johan Hovold <johan@kernel.org>
Cc: Marcel Holtmann <marcel@holtmann.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Shuah Khan <shuah@kernel.org>
---
 drivers/bluetooth/hci_ldisc.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/bluetooth/hci_ldisc.c b/drivers/bluetooth/hci_ldisc.c
index fbf7b4df23ab..c8faa4759cb7 100644
--- a/drivers/bluetooth/hci_ldisc.c
+++ b/drivers/bluetooth/hci_ldisc.c
@@ -475,9 +475,9 @@ static int hci_uart_tty_open(struct tty_struct *tty)
 	BT_DBG("tty %p", tty);
 
 	/* Error if the tty has no write op instead of leaving an exploitable
-	 * hole
+	 * hole. In addition check if setting HCI line discipline is allowed.
 	 */
-	if (tty->ops->write == NULL)
+	if (tty->ops->write == NULL || tty->ops->set_termios == NULL)
 		return -EOPNOTSUPP;
 
 	hu = kzalloc(sizeof(struct hci_uart), GFP_KERNEL);
-- 
2.17.1

Re: [PATCH v4] bluetooth: Fix WARNING in tty_set_termios()

[Marcel Holtmann]

Hi Shuah,

> tty_set_termios() has the following WARN_ON which can be triggered with a
> syscall to invoke TIOCSETD __NR_ioctl.
> 
> WARN_ON(tty->driver->type == TTY_DRIVER_TYPE_PTY &&
>                tty->driver->subtype == PTY_TYPE_MASTER);
> Reference: https://syzkaller.appspot.com/bug?id=2410d22f1d8e5984217329dd0884b01d99e3e48d
> 
> "The problem started with commit 7721383f4199 ("Bluetooth: hci_uart: Support
> operational speed during setup") which introduced a new way for how
> tty_set_termios() could end up being called for a master pty."
> 
> Fix it by by preventing setting the HCI line discipline for PTYs in
> hci_uart_tty_open().
> 
> Reported-by: syzbot+a950165cbb86bdd023a4@syzkaller.appspotmail.com
> Cc: Johan Hovold <johan@kernel.org>
> Cc: Marcel Holtmann <marcel@holtmann.org>
> Cc: Al Viro <viro@zeniv.linux.org.uk>
> Signed-off-by: Shuah Khan <shuah@kernel.org>
> ---
> drivers/bluetooth/hci_ldisc.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/bluetooth/hci_ldisc.c b/drivers/bluetooth/hci_ldisc.c
> index fbf7b4df23ab..c8faa4759cb7 100644
> --- a/drivers/bluetooth/hci_ldisc.c
> +++ b/drivers/bluetooth/hci_ldisc.c
> @@ -475,9 +475,9 @@ static int hci_uart_tty_open(struct tty_struct *tty)
> 	BT_DBG("tty %p", tty);
> 
> 	/* Error if the tty has no write op instead of leaving an exploitable
> -	 * hole
> +	 * hole. In addition check if setting HCI line discipline is allowed.
> 	 */

this comment is wrong. The result of setting the HCI line discipline is a call into hci_uart_tty_open. You have to check ops->write and ops->set_termios since both are required. That is it.

> -	if (tty->ops->write == NULL)
> +	if (tty->ops->write == NULL || tty->ops->set_termios == NULL)
> 		return -EOPNOTSUPP;

And while at it, change this to (!tty->ops->write).

Regards

Marcel

Re: [PATCH v4] bluetooth: Fix WARNING in tty_set_termios()

[shuah]

On 2/3/19 10:39 PM, Marcel Holtmann wrote:
> Hi Shuah,
> 
>> tty_set_termios() has the following WARN_ON which can be triggered with a
>> syscall to invoke TIOCSETD __NR_ioctl.
>>
>> WARN_ON(tty->driver->type == TTY_DRIVER_TYPE_PTY &&
>>                 tty->driver->subtype == PTY_TYPE_MASTER);
>> Reference: https://syzkaller.appspot.com/bug?id=2410d22f1d8e5984217329dd0884b01d99e3e48d
>>
>> "The problem started with commit 7721383f4199 ("Bluetooth: hci_uart: Support
>> operational speed during setup") which introduced a new way for how
>> tty_set_termios() could end up being called for a master pty."
>>
>> Fix it by by preventing setting the HCI line discipline for PTYs in
>> hci_uart_tty_open().
>>
>> Reported-by: syzbot+a950165cbb86bdd023a4@syzkaller.appspotmail.com
>> Cc: Johan Hovold <johan@kernel.org>
>> Cc: Marcel Holtmann <marcel@holtmann.org>
>> Cc: Al Viro <viro@zeniv.linux.org.uk>
>> Signed-off-by: Shuah Khan <shuah@kernel.org>
>> ---
>> drivers/bluetooth/hci_ldisc.c | 4 ++--
>> 1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/bluetooth/hci_ldisc.c b/drivers/bluetooth/hci_ldisc.c
>> index fbf7b4df23ab..c8faa4759cb7 100644
>> --- a/drivers/bluetooth/hci_ldisc.c
>> +++ b/drivers/bluetooth/hci_ldisc.c
>> @@ -475,9 +475,9 @@ static int hci_uart_tty_open(struct tty_struct *tty)
>> 	BT_DBG("tty %p", tty);
>>
>> 	/* Error if the tty has no write op instead of leaving an exploitable
>> -	 * hole
>> +	 * hole. In addition check if setting HCI line discipline is allowed.
>> 	 */
> 
> this comment is wrong. The result of setting the HCI line discipline is a call into hci_uart_tty_open. You have to check ops->write and ops->set_termios since both are required. That is it.

I will fix the comment.

> 
>> -	if (tty->ops->write == NULL)
>> +	if (tty->ops->write == NULL || tty->ops->set_termios == NULL)
>> 		return -EOPNOTSUPP;
> 
> And while at it, change this to (!tty->ops->write).
> 

Okay. I will fix both.

thanks,
-- Shuah