Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Sat, 2 Feb 2019 19:30:32 -0800
From:   Myungho Jung <mhjungk@gmail.com>
To:     Marcel Holtmann <marcel@holtmann.org>
Cc:     Johan Hedberg <johan.hedberg@gmail.com>,
        linux-bluetooth@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2] Bluetooth: hci_uart: Switch pty driver to slave side
 in tty_set_termios()
Message-ID: <20190203033032.GA21013@myunghoj-Precision-5530>
References: <20190202072816.GA15630@myunghoj-Precision-5530>
 <3EC8C1C2-27AE-46B7-A345-860A87B94E6B@holtmann.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <3EC8C1C2-27AE-46B7-A345-860A87B94E6B@holtmann.org>
User-Agent: Mutt/1.9.4 (2018-02-28)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Sat, Feb 02, 2019 at 01:15:16PM +0100, Marcel Holtmann wrote:
> Hi Myungho,
> 
> > tty_set_termios() should be called with slave side of pty driver. So, If
> > tty driver is pty master, it needs to be switched to ->link.
> > 
> > Reported-by: syzbot+a950165cbb86bdd023a4@syzkaller.appspotmail.com
> > Signed-off-by: Myungho Jung <mhjungk@gmail.com>
> > ---
> > drivers/bluetooth/hci_ldisc.c | 27 ++++++++++++++++++++++-----
> > 1 file changed, 22 insertions(+), 5 deletions(-)
> > 
> > diff --git a/drivers/bluetooth/hci_ldisc.c b/drivers/bluetooth/hci_ldisc.c
> > index fbf7b4df23ab..0f6e70ea1dc3 100644
> > --- a/drivers/bluetooth/hci_ldisc.c
> > +++ b/drivers/bluetooth/hci_ldisc.c
> > @@ -299,10 +299,21 @@ static int hci_uart_send_frame(struct hci_dev *hdev, struct sk_buff *skb)
> > 	return 0;
> > }
> > 
> > +/* If pty master, return slave side */
> > +static struct tty_struct *hci_uart_get_real_tty(struct tty_struct *tty)
> > +{
> > +	if (tty->driver->type == TTY_DRIVER_TYPE_PTY &&
> > +	    tty->driver->subtype == PTY_TYPE_MASTER)
> > +		tty->link;
> > +
> > +	return tty;
> > +}
> > +
> > /* Flow control or un-flow control the device */
> > void hci_uart_set_flow_control(struct hci_uart *hu, bool enable)
> > {
> > 	struct tty_struct *tty = hu->tty;
> > +	struct tty_struct *real_tty;
> > 	struct ktermios ktermios;
> > 	int status;
> > 	unsigned int set = 0;
> > @@ -314,11 +325,14 @@ void hci_uart_set_flow_control(struct hci_uart *hu, bool enable)
> > 		return;
> > 	}
> > 
> > +	/* termios should be set from slave side if tty driver is pty */
> > +	real_tty = hci_uart_get_real_tty(tty);
> > +
> > 	if (enable) {
> > 		/* Disable hardware flow control */
> > -		ktermios = tty->termios;
> > +		ktermios = real_tty->termios;
> > 		ktermios.c_cflag &= ~CRTSCTS;
> > -		status = tty_set_termios(tty, &ktermios);
> > +		status = tty_set_termios(real_tty, &ktermios);
> > 		BT_DBG("Disabling hardware flow control: %s",
> > 		       status ? "failed" : "success");
> > 
> > @@ -350,9 +364,9 @@ void hci_uart_set_flow_control(struct hci_uart *hu, bool enable)
> > 		BT_DBG("Setting RTS: %s", status ? "failed" : "success");
> > 
> > 		/* Re-enable hardware flow control */
> > -		ktermios = tty->termios;
> > +		ktermios = real_tty->termios;
> > 		ktermios.c_cflag |= CRTSCTS;
> > -		status = tty_set_termios(tty, &ktermios);
> > +		status = tty_set_termios(real_tty, &ktermios);
> > 		BT_DBG("Enabling hardware flow control: %s",
> > 		       status ? "failed" : "success");
> > 	}
> > @@ -367,9 +381,12 @@ void hci_uart_set_speeds(struct hci_uart *hu, unsigned int init_speed,
> > 
> > void hci_uart_set_baudrate(struct hci_uart *hu, unsigned int speed)
> > {
> > -	struct tty_struct *tty = hu->tty;
> > +	struct tty_struct *tty;
> > 	struct ktermios ktermios;
> > 
> > +	/* termios should be set from slave side if tty driver is pty */
> > +	tty = hci_uart_get_real_tty(hu->tty);
> > +
> 
> didn’t we conclude that allowing the HCI line discipline on a PTY master isn't worth the effort? The ptm_unix98_ops doesn’t have the .set_termios and thus, just check that ops->set_termios is present in hci_uart_tty_open and if not fail with EOPNOTSUPP like we do for missing ops->write.
> 
> Regards
> 
> Marcel
> 
Hi Marcel,

I thought there might be some cases that reach here even returning error in
hci_uart_tty_open(). Let me fix it from hci_uart_tty_open.

Thanks,
Myungho