n_tty: add null check for tty->link in packet mode

[PATCH] n_tty: add null check for tty->link in packet mode

Posted by Osama Abdelkader 3 weeks, 2 days ago

Add null check for tty->link before dereferencing in n_tty_read and
n_tty_poll. When the pty master closes, tty->link can be NULL while
the slave is still reading, causing a null pointer dereference.

Signed-off-by: Osama Abdelkader <osama.abdelkader@gmail.com>
---
 drivers/tty/n_tty.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/tty/n_tty.c b/drivers/tty/n_tty.c
index e6a0f5b40d0a..dc04b87364f6 100644
--- a/drivers/tty/n_tty.c
+++ b/drivers/tty/n_tty.c
@@ -2232,7 +2232,7 @@ static ssize_t n_tty_read(struct tty_struct *tty, struct file *file, u8 *kbuf,
 	add_wait_queue(&tty->read_wait, &wait);
 	while (nr) {
 		/* First test for status change. */
-		if (packet && tty->link->ctrl.pktstatus) {
+		if (packet && tty->link && tty->link->ctrl.pktstatus) {
 			u8 cs;
 			if (kb != kbuf)
 				break;
@@ -2444,7 +2444,7 @@ static __poll_t n_tty_poll(struct tty_struct *tty, struct file *file,
 		if (input_available_p(tty, 1))
 			mask |= EPOLLIN | EPOLLRDNORM;
 	}
-	if (tty->ctrl.packet && tty->link->ctrl.pktstatus)
+	if (tty->ctrl.packet && tty->link && tty->link->ctrl.pktstatus)
 		mask |= EPOLLPRI | EPOLLIN | EPOLLRDNORM;
 	if (test_bit(TTY_OTHER_CLOSED, &tty->flags))
 		mask |= EPOLLHUP;
-- 
2.43.0

Re: [PATCH] n_tty: add null check for tty->link in packet mode

Posted by Greg Kroah-Hartman 3 weeks, 1 day ago

On Sat, Mar 14, 2026 at 11:10:44PM +0100, Osama Abdelkader wrote:
> Add null check for tty->link before dereferencing in n_tty_read and
> n_tty_poll. When the pty master closes, tty->link can be NULL while
> the slave is still reading, causing a null pointer dereference.

How can that happen?

> Signed-off-by: Osama Abdelkader <osama.abdelkader@gmail.com>
> ---
>  drivers/tty/n_tty.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/tty/n_tty.c b/drivers/tty/n_tty.c
> index e6a0f5b40d0a..dc04b87364f6 100644
> --- a/drivers/tty/n_tty.c
> +++ b/drivers/tty/n_tty.c
> @@ -2232,7 +2232,7 @@ static ssize_t n_tty_read(struct tty_struct *tty, struct file *file, u8 *kbuf,
>  	add_wait_queue(&tty->read_wait, &wait);
>  	while (nr) {
>  		/* First test for status change. */
> -		if (packet && tty->link->ctrl.pktstatus) {
> +		if (packet && tty->link && tty->link->ctrl.pktstatus) {
>  			u8 cs;
>  			if (kb != kbuf)
>  				break;
> @@ -2444,7 +2444,7 @@ static __poll_t n_tty_poll(struct tty_struct *tty, struct file *file,
>  		if (input_available_p(tty, 1))
>  			mask |= EPOLLIN | EPOLLRDNORM;
>  	}
> -	if (tty->ctrl.packet && tty->link->ctrl.pktstatus)
> +	if (tty->ctrl.packet && tty->link && tty->link->ctrl.pktstatus)

What happens if link changes right after you test it?  Where is the
lock?

And what changed to cause this to show up now?

thanks,

greg k-h

Re: [PATCH] n_tty: add null check for tty->link in packet mode

Posted by Osama Abdelkader 2 weeks ago

On Sun, Mar 15, 2026 at 07:57:53AM +0100, Greg Kroah-Hartman wrote:
> On Sat, Mar 14, 2026 at 11:10:44PM +0100, Osama Abdelkader wrote:
> > Add null check for tty->link before dereferencing in n_tty_read and
> > n_tty_poll. When the pty master closes, tty->link can be NULL while
> > the slave is still reading, causing a null pointer dereference.
> 
> How can that happen?
> 
> > Signed-off-by: Osama Abdelkader <osama.abdelkader@gmail.com>
> > ---
> >  drivers/tty/n_tty.c | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/tty/n_tty.c b/drivers/tty/n_tty.c
> > index e6a0f5b40d0a..dc04b87364f6 100644
> > --- a/drivers/tty/n_tty.c
> > +++ b/drivers/tty/n_tty.c
> > @@ -2232,7 +2232,7 @@ static ssize_t n_tty_read(struct tty_struct *tty, struct file *file, u8 *kbuf,
> >  	add_wait_queue(&tty->read_wait, &wait);
> >  	while (nr) {
> >  		/* First test for status change. */
> > -		if (packet && tty->link->ctrl.pktstatus) {
> > +		if (packet && tty->link && tty->link->ctrl.pktstatus) {
> >  			u8 cs;
> >  			if (kb != kbuf)
> >  				break;
> > @@ -2444,7 +2444,7 @@ static __poll_t n_tty_poll(struct tty_struct *tty, struct file *file,
> >  		if (input_available_p(tty, 1))
> >  			mask |= EPOLLIN | EPOLLRDNORM;
> >  	}
> > -	if (tty->ctrl.packet && tty->link->ctrl.pktstatus)
> > +	if (tty->ctrl.packet && tty->link && tty->link->ctrl.pktstatus)
> 
> What happens if link changes right after you test it?  Where is the
> lock?
> 
> And what changed to cause this to show up now?
> 
> thanks,
> 
> greg k-h

Hi Greg,

I was just thinking about null dereferencing possiblity in tty->link->ctrl.pktstatus. But, you are right
It’s reasonable to drop this patch and reopen it only if I get a solid reproducer or bug report.

BR,
Osama