| Message ID | 42a31507d9b1d4874bb9a8570bd9f17225aaa553.1552027854.git.lucien.xin@gmail.com |
|---|---|
| State | Accepted |
| Delegated to: | David Miller |
| Headers | show |
| Series | [net] route: set the deleted fnhe fnhe_daddr to 0 in ip_del_fnhe to fix a race | expand |
On 3/7/19 11:50 PM, Xin Long wrote: > The race occurs in __mkroute_output() when 2 threads lookup a dst: > > CPU A CPU B > find_exception() > find_exception() [fnhe expires] > ip_del_fnhe() [fnhe is deleted] > rt_bind_exception() > > In rt_bind_exception() it will bind a deleted fnhe with the new dst, and > this dst will get no chance to be freed. It causes a dev defcnt leak and > consecutive dmesg warnings: > > unregister_netdevice: waiting for ethX to become free. Usage count = 1 > > Especially thinks Jon to identify the issue. > > This patch fixes it by setting fnhe_daddr to 0 in ip_del_fnhe() to stop > binding the deleted fnhe with a new dst when checking fnhe's fnhe_daddr > and daddr in rt_bind_exception(). > > It works as both ip_del_fnhe() and rt_bind_exception() are protected by > fnhe_lock and the fhne is freed by kfree_rcu(). > > Fixes: deed49df7390 ("route: check and remove route cache when we get route") > Signed-off-by: Jon Maxwell <jmaxwell37@gmail.com> > Signed-off-by: Xin Long <lucien.xin@gmail.com> > --- > net/ipv4/route.c | 4 ++++ > 1 file changed, 4 insertions(+) > Reviewed-by: David Ahern <dsahern@gmail.com>
From: Xin Long <lucien.xin@gmail.com> Date: Fri, 8 Mar 2019 14:50:54 +0800 > The race occurs in __mkroute_output() when 2 threads lookup a dst: > > CPU A CPU B > find_exception() > find_exception() [fnhe expires] > ip_del_fnhe() [fnhe is deleted] > rt_bind_exception() > > In rt_bind_exception() it will bind a deleted fnhe with the new dst, and > this dst will get no chance to be freed. It causes a dev defcnt leak and > consecutive dmesg warnings: > > unregister_netdevice: waiting for ethX to become free. Usage count = 1 > > Especially thinks Jon to identify the issue. I did "s/thinks/thanks/" here. > This patch fixes it by setting fnhe_daddr to 0 in ip_del_fnhe() to stop > binding the deleted fnhe with a new dst when checking fnhe's fnhe_daddr > and daddr in rt_bind_exception(). > > It works as both ip_del_fnhe() and rt_bind_exception() are protected by > fnhe_lock and the fhne is freed by kfree_rcu(). > > Fixes: deed49df7390 ("route: check and remove route cache when we get route") > Signed-off-by: Jon Maxwell <jmaxwell37@gmail.com> > Signed-off-by: Xin Long <lucien.xin@gmail.com> Applied and queues up for -stable, thank you.
diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 5163b64..b0eb4fa 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -1303,6 +1303,10 @@ static void ip_del_fnhe(struct fib_nh *nh, __be32 daddr) if (fnhe->fnhe_daddr == daddr) { rcu_assign_pointer(*fnhe_p, rcu_dereference_protected( fnhe->fnhe_next, lockdep_is_held(&fnhe_lock))); + /* set fnhe_daddr to 0 to ensure it won't bind with + * new dsts in rt_bind_exception(). + */ + fnhe->fnhe_daddr = 0; fnhe_flush_routes(fnhe); kfree_rcu(fnhe, rcu); break;
The race occurs in __mkroute_output() when 2 threads lookup a dst: CPU A CPU B find_exception() find_exception() [fnhe expires] ip_del_fnhe() [fnhe is deleted] rt_bind_exception() In rt_bind_exception() it will bind a deleted fnhe with the new dst, and this dst will get no chance to be freed. It causes a dev defcnt leak and consecutive dmesg warnings: unregister_netdevice: waiting for ethX to become free. Usage count = 1 Especially thinks Jon to identify the issue. This patch fixes it by setting fnhe_daddr to 0 in ip_del_fnhe() to stop binding the deleted fnhe with a new dst when checking fnhe's fnhe_daddr and daddr in rt_bind_exception(). It works as both ip_del_fnhe() and rt_bind_exception() are protected by fnhe_lock and the fhne is freed by kfree_rcu(). Fixes: deed49df7390 ("route: check and remove route cache when we get route") Signed-off-by: Jon Maxwell <jmaxwell37@gmail.com> Signed-off-by: Xin Long <lucien.xin@gmail.com> --- net/ipv4/route.c | 4 ++++ 1 file changed, 4 insertions(+)