tty: Fix data race between tiocsti() and flush_to_ldisc()
PATCH v2 1/1 tty: Fix data race between tiocsti() and flush_to_ldisc()
The ops->receive_buf() may be accessed concurrently from these two functions. If the driver flushes data to the line discipline receive_buf() method while tiocsti() is waiting for the ops->receive_buf() to finish its work, the data race will happen.
For example:tty_ioctl |tty_ldisc_receive_buf
->tioctsi | ->tty_port_default_receive_buf
| ->tty_ldisc_receive_buf
->hci_uart_tty_receive | ->hci_uart_tty_receive
->h4_recv | ->h4_recvIn this case, the h4 receive buffer will be overwritten by the latecomer, and we will lost the data.
Hence, change tioctsi() function to use the exclusive lock interface from tty_buffer to avoid the data race.
---
V2:
- Remove FIXME comment.
drivers/tty/tty_io.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/tty/tty_io.c +2 -2
@@ -2290,8 +2290,6 @@ static int tty_fasync(int fd, struct file *filp, int on)* Locking: * Called functions take tty_ldiscs_lock * current->signal->tty check is safe without locks -* -* FIXME: may race normal receive processing */ static int tiocsti (struct tty_struct *tty,char __user *p)
@@ -2307,8 +2305,10 @@ static int tiocsti(struct tty_struct *tty, char __user *p) ld =tty_ldisc_ref_wait (tty);if (!ld)return -EIO ; +tty_buffer_lock_exclusive (tty->port);if (ld->ops->receive_buf) ld->ops->receive_buf (tty, &ch, &mbz,1 ); +tty_buffer_unlock_exclusive (tty->port);tty_ldisc_deref (ld);return 0 ; }
--
2.25.1
_______________________________________________
Linux-kernel-mentees mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/linux-kernel-mentees
Quoted 12 lines · Nguyen Dinh Phi
The
ops->receive_buf()may be accessed concurrently from these two functions. If the driver flushes data to the line disciplinereceive_buf()method whiletiocsti()is waiting for theops->receive_buf()to finish its work, the data race will happen.For example:
tty_ioctl|tty_ldisc_receive_buf
->tioctsi | ->tty_port_default_receive_buf| ->tty_ldisc_receive_buf ->hci_uart_tty_receive | ->hci_uart_tty_receive ->h4_recv | ->h4_recv
Quoted 5 lines
In this case, the h4 receive buffer will be overwritten by the latecomer, and we will lost the data.
Hence, change
tioctsi()function to use the exclusive lock interface fromtty_bufferto avoid the data race.
This makes sense. It mimics paste_selection() -- the lock order is preexisting. The normal path processing has the locks inverted, not sure why it doesn't matter :):flush_to_ldisc
mutex_lock(&buf->lock); <--------- A (mutex)
receive_buf()
tty_port_default_receive_buf()
disc = tty_ldisc_ref(tty); <----- B (ldsem for READ)Quoted 23 lines
Signed-off-by Nguyen Dinh Phi <[email protected]>Reported-by [email protected]---
V2:
- Remove FIXME comment.drivers/tty/tty_io.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)drivers/tty/tty_io.c +0
-2@@ -2290,8 +2290,6 @@ static int tty_fasync(int fd, struct file *filp, int on)* Locking: * Called functions take tty_ldiscs_lock * current->signal->tty check is safe without locks -* -* FIXME: may race normal receive processing */ static int tiocsti (struct tty_struct *tty,char __user *p)
drivers/tty/tty_io.c static int tiocsti(struct tty_struct *tty, char __user *p) · line 2305
Quoted 1 line
drivers/tty/tty_io.c +0
-0@@ -2307,8 +2305,10 @@ static int tiocsti(struct tty_struct *tty, char __user *p) ld =tty_ldisc_ref_wait (tty);
Here (and in paste_selection()), it is lock "B (ldsem) for READ".
drivers/tty/tty_io.c static int tiocsti(struct tty_struct *tty, char __user *p) · line 2308
Quoted 3 lines
drivers/tty/tty_io.c +1
-0if (!ld)return -EIO ; +tty_buffer_lock_exclusive (tty->port);
and "A (mutex)".
drivers/tty/tty_io.c static int tiocsti(struct tty_struct *tty, char __user *p) · line 2314
Quoted 6 lines
drivers/tty/tty_io.c +1
-0if (ld->ops->receive_buf) ld->ops->receive_buf (tty, &ch, &mbz,1 ); +tty_buffer_unlock_exclusive (tty->port);tty_ldisc_deref (ld);return 0 ; }
thanks,
Signature · 2 lines
js suse labs