Re: Re: [PATCH net V4 2/2] ax25: Fix NULL pointer dereferences in ax25 timers

[周多明 <duoming@zju.edu.cn>]

Hello,

On Mon, 14 Mar 2022 20:03:00 -0700, Eric Dumazet wrote:
> > There are race conditions that may lead to null pointer dereferences in
> > ax25_heartbeat_expiry(), ax25_t1timer_expiry(), ax25_t2timer_expiry(),
> > ax25_t3timer_expiry() and ax25_idletimer_expiry(), when we use
> > ax25_kill_by_device() to detach the ax25 device.
> >
> > One of the race conditions that cause null pointer dereferences can be
> > shown as below:
> >
> >        (Thread 1)                    |      (Thread 2)
> > ax25_connect()                       |
> >   ax25_std_establish_data_link()     |
> >    ax25_start_t1timer()              |
> >     mod_timer(&ax25->t1timer,..)     |
> >                                      | ax25_kill_by_device()
> >     (wait a time)                    |  ...
> >                                      |  s->ax25_dev = NULL; //(1)
> >     ax25_t1timer_expiry()            |
> >      ax25->ax25_dev->values[..] //(2)|  ...
> >       ...                            |
> >
> > We set null to ax25_cb->ax25_dev in position (1) and dereference
> > the null pointer in position (2).
> >
> > The corresponding fail log is shown below:
> > ===============================================================
> > BUG: kernel NULL pointer dereference, address: 0000000000000050
> > CPU: 1 PID: 0 Comm: swapper/1 Not tainted 5.17.0-rc6-00794-g45690b7d0
> > RIP: 0010:ax25_t1timer_expiry+0x12/0x40
> > ...
> > Call Trace:
> >   call_timer_fn+0x21/0x120
> >   __run_timers.part.0+0x1ca/0x250
> >   run_timer_softirq+0x2c/0x60
> >   __do_softirq+0xef/0x2f3
> >   irq_exit_rcu+0xb6/0x100
> >   sysvec_apic_timer_interrupt+0xa2/0xd0
> > ...
> >
> > This patch uses ax25_disconnect() to delete timers before we set null to
> > ax25_cb->ax25_dev in ax25_kill_by_device().The function ax25_disconnect()
> > will not return until all timers are stopped, because we have changed
> > del_timer() to del_timer_sync(). What`s more, we add condition check in
> > ax25_destroy_socket(), because ax25_stop_heartbeat() will not return,
> > if there is still heartbeat.
> >
> > Signed-off-by: Duoming Zhou <duoming@zju.edu.cn>
> 
> Missing FIxes: tag ?
> 
> 
> > ---
> > Changes in V4:
> >    - Based on [PATCH net V4 1/2] ax25: Fix refcount leaks caused by ax25_cb_del().
> >
> >   net/ax25/af_ax25.c    |  7 ++++---
> >   net/ax25/ax25_timer.c | 10 +++++-----
> >   2 files changed, 9 insertions(+), 8 deletions(-)
> >
> > diff --git a/net/ax25/af_ax25.c b/net/ax25/af_ax25.c
> > index 0886109421a..dc6161a75a1 100644
> > --- a/net/ax25/af_ax25.c
> > +++ b/net/ax25/af_ax25.c
> > @@ -89,20 +89,20 @@ static void ax25_kill_by_device(struct net_device *dev)
> >   			sk = s->sk;
> >   			if (!sk) {
> >   				spin_unlock_bh(&ax25_list_lock);
> > -				s->ax25_dev = NULL;
> >   				ax25_disconnect(s, ENETUNREACH);
> > +				s->ax25_dev = NULL;
> >   				spin_lock_bh(&ax25_list_lock);
> >   				goto again;
> >   			}
> >   			sock_hold(sk);
> >   			spin_unlock_bh(&ax25_list_lock);
> >   			lock_sock(sk);
> > +			ax25_disconnect(s, ENETUNREACH);
> >   			s->ax25_dev = NULL;
> >   			if (sk->sk_wq) {
> >   				dev_put_track(ax25_dev->dev, &ax25_dev->dev_tracker);
> >   				ax25_dev_put(ax25_dev);
> >   			}
> > -			ax25_disconnect(s, ENETUNREACH);
> >   			release_sock(sk);
> >   			spin_lock_bh(&ax25_list_lock);
> >   			sock_put(sk);
> > @@ -307,7 +307,8 @@ void ax25_destroy_socket(ax25_cb *ax25)
> >   
> >   	ax25_cb_del(ax25);
> >   
> > -	ax25_stop_heartbeat(ax25);
> > +	if (!ax25->sk || !sock_flag(ax25->sk, SOCK_DESTROY))
> > +		ax25_stop_heartbeat(ax25);
> >   	ax25_stop_t1timer(ax25);
> >   	ax25_stop_t2timer(ax25);
> >   	ax25_stop_t3timer(ax25);
> > diff --git a/net/ax25/ax25_timer.c b/net/ax25/ax25_timer.c
> > index 85865ebfdfa..99af3d1aeec 100644
> > --- a/net/ax25/ax25_timer.c
> > +++ b/net/ax25/ax25_timer.c
> > @@ -78,27 +78,27 @@ void ax25_start_idletimer(ax25_cb *ax25)
> >   
> >   void ax25_stop_heartbeat(ax25_cb *ax25)
> >   {
> > -	del_timer(&ax25->timer);
> > +	del_timer_sync(&ax25->timer);
> >   }
> >   
> >   void ax25_stop_t1timer(ax25_cb *ax25)
> >   {
> > -	del_timer(&ax25->t1timer);
> > +	del_timer_sync(&ax25->t1timer);
> >   }
> >   
> >   void ax25_stop_t2timer(ax25_cb *ax25)
> >   {
> > -	del_timer(&ax25->t2timer);
> > +	del_timer_sync(&ax25->t2timer);
> >   }
> >   
> >   void ax25_stop_t3timer(ax25_cb *ax25)
> >   {
> > -	del_timer(&ax25->t3timer);
> > +	del_timer_sync(&ax25->t3timer);
> >   }
> >   
> >   void ax25_stop_idletimer(ax25_cb *ax25)
> >   {
> > -	del_timer(&ax25->idletimer);
> > +	del_timer_sync(&ax25->idletimer);
> >   }
> >   
> >   int ax25_t1timer_running(ax25_cb *ax25)
> 
> 
> 
> Are you sure calling del_time_sync() wont deadlock ?
> 
> 
> If the timer handlers need a lock owned by the thread calling 
> del_timer_sync(),
> 
> then this will block forever.

I think there is no deadlock.

Function ax25_kill_by_device() will only hold lock_sock(sk) before using 
ax25_disconnect() to call del_timer_sync().

In timers, we hold bh_lock_sock(sk) or spin_lock_bh(&ax25_list_lock) which
is different from lock_sock(sk).

Best wishes,
Duoming Zhou