[PATCH] serial: core: fix transmit-buffer reset and memleak

[PATCH] serial: core: fix transmit-buffer reset and memleak

[Johan Hovold]

Commit 761ed4a94582 ("tty: serial_core: convert uart_close to use
tty_port_close") converted serial core to use tty_port_close() but
failed to notice that the transmit buffer still needs to be freed on
final close.

Not freeing the transmit buffer means that the buffer is no longer
cleared on next open so that any ioctl() waiting for the buffer to drain
might wait indefinitely (e.g. on termios changes) or that stale data can
end up being transmitted in case tx is restarted.

Furthermore, the buffer of any port that has been opened would leak on
driver unbind.

Note that the port lock is held when clearing the buffer pointer due to
the ldisc race worked around by commit a5ba1d95e46e ("uart: fix race
between uart_put_char() and uart_shutdown()").

Also note that the tty-port shutdown() callback is not called for
console ports so it is not strictly necessary to free the buffer page
after releasing the lock (cf. d72402145ace ("tty/serial: do not free
trasnmit buffer page under port lock")).

Reported-by: Baruch Siach <baruch@tkos.co.il>
Link: https://lore.kernel.org/r/319321886d97c456203d5c6a576a5480d07c3478.1635781688.git.baruch@tkos.co.il
Fixes: 761ed4a94582 ("tty: serial_core: convert uart_close to use tty_port_close")
Cc: stable@vger.kernel.org      # 4.9
Cc: Rob Herring <robh@kernel.org>
Signed-off-by: Johan Hovold <johan@kernel.org>
---
 drivers/tty/serial/serial_core.c | 13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)

diff --git a/drivers/tty/serial/serial_core.c b/drivers/tty/serial/serial_core.c
index 0e2e35ab64c7..58834698739c 100644
--- a/drivers/tty/serial/serial_core.c
+++ b/drivers/tty/serial/serial_core.c
@@ -1542,6 +1542,7 @@ static void uart_tty_port_shutdown(struct tty_port *port)
 {
 	struct uart_state *state = container_of(port, struct uart_state, port);
 	struct uart_port *uport = uart_port_check(state);
+	char *buf;
 
 	/*
 	 * At this point, we stop accepting input.  To do this, we
@@ -1563,8 +1564,18 @@ static void uart_tty_port_shutdown(struct tty_port *port)
 	 */
 	tty_port_set_suspended(port, 0);
 
-	uart_change_pm(state, UART_PM_STATE_OFF);
+	/*
+	 * Free the transmit buffer.
+	 */
+	spin_lock_irq(&uport->lock);
+	buf = state->xmit.buf;
+	state->xmit.buf = NULL;
+	spin_unlock_irq(&uport->lock);
 
+	if (buf)
+		free_page((unsigned long)buf);
+
+	uart_change_pm(state, UART_PM_STATE_OFF);
 }
 
 static void uart_wait_until_sent(struct tty_struct *tty, int timeout)
-- 
2.32.0

Re: [PATCH] serial: core: fix transmit-buffer reset and memleak

[Baruch Siach]

Hi Johan,

On Mon, Nov 08 2021, Johan Hovold wrote:
> Commit 761ed4a94582 ("tty: serial_core: convert uart_close to use
> tty_port_close") converted serial core to use tty_port_close() but
> failed to notice that the transmit buffer still needs to be freed on
> final close.
>
> Not freeing the transmit buffer means that the buffer is no longer
> cleared on next open so that any ioctl() waiting for the buffer to drain
> might wait indefinitely (e.g. on termios changes) or that stale data can
> end up being transmitted in case tx is restarted.
>
> Furthermore, the buffer of any port that has been opened would leak on
> driver unbind.
>
> Note that the port lock is held when clearing the buffer pointer due to
> the ldisc race worked around by commit a5ba1d95e46e ("uart: fix race
> between uart_put_char() and uart_shutdown()").
>
> Also note that the tty-port shutdown() callback is not called for
> console ports so it is not strictly necessary to free the buffer page
> after releasing the lock (cf. d72402145ace ("tty/serial: do not free
> trasnmit buffer page under port lock")).
>
> Reported-by: Baruch Siach <baruch@tkos.co.il>
> Link: https://lore.kernel.org/r/319321886d97c456203d5c6a576a5480d07c3478.1635781688.git.baruch@tkos.co.il
> Fixes: 761ed4a94582 ("tty: serial_core: convert uart_close to use tty_port_close")
> Cc: stable@vger.kernel.org      # 4.9
> Cc: Rob Herring <robh@kernel.org>
> Signed-off-by: Johan Hovold <johan@kernel.org>

Thanks for the analysis and root cause fix. This patch also fixes the
issue for me.

Tested-by: Baruch Siach <baruch@tkos.co.il>

baruch

> ---
>  drivers/tty/serial/serial_core.c | 13 ++++++++++++-
>  1 file changed, 12 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/tty/serial/serial_core.c b/drivers/tty/serial/serial_core.c
> index 0e2e35ab64c7..58834698739c 100644
> --- a/drivers/tty/serial/serial_core.c
> +++ b/drivers/tty/serial/serial_core.c
> @@ -1542,6 +1542,7 @@ static void uart_tty_port_shutdown(struct tty_port *port)
>  {
>  	struct uart_state *state = container_of(port, struct uart_state, port);
>  	struct uart_port *uport = uart_port_check(state);
> +	char *buf;
>  
>  	/*
>  	 * At this point, we stop accepting input.  To do this, we
> @@ -1563,8 +1564,18 @@ static void uart_tty_port_shutdown(struct tty_port *port)
>  	 */
>  	tty_port_set_suspended(port, 0);
>  
> -	uart_change_pm(state, UART_PM_STATE_OFF);
> +	/*
> +	 * Free the transmit buffer.
> +	 */
> +	spin_lock_irq(&uport->lock);
> +	buf = state->xmit.buf;
> +	state->xmit.buf = NULL;
> +	spin_unlock_irq(&uport->lock);
>  
> +	if (buf)
> +		free_page((unsigned long)buf);
> +
> +	uart_change_pm(state, UART_PM_STATE_OFF);
>  }
>  
>  static void uart_wait_until_sent(struct tty_struct *tty, int timeout)


-- 
                                                     ~. .~   Tk Open Systems
=}------------------------------------------------ooO--U--Ooo------------{=
   - baruch@tkos.co.il - tel: +972.52.368.4656, http://www.tkos.co.il -