@@ -164,6 +164,7 @@ struct ip_tunnel_net {
struct rtnl_link_ops *rtnl_link_ops;
struct hlist_head tunnels[IP_TNL_HASH_SIZE];
struct ip_tunnel __rcu *collect_md_tun;
+ struct ip_tunnel __rcu *fb_tun;
int type;
};
@@ -162,8 +162,9 @@ struct ip_tunnel *ip_tunnel_lookup(struct ip_tunnel_net *itn,
if (t && t->dev->flags & IFF_UP)
return t;
- if (itn->fb_tunnel_dev && itn->fb_tunnel_dev->flags & IFF_UP)
- return netdev_priv(itn->fb_tunnel_dev);
+ t = rcu_dereference(itn->fb_tun);
+ if (t && t->dev->flags & IFF_UP)
+ return t;
return NULL;
}
@@ -1059,6 +1060,7 @@ int ip_tunnel_init_net(struct net *net, unsigned int ip_tnl_net_id,
it_init_net = net_generic(&init_net, ip_tnl_net_id);
itn->type = it_init_net->type;
itn->fb_tunnel_dev = NULL;
+ RCU_INIT_POINTER(itn->fb_tun, NULL);
return 0;
}
@@ -1074,8 +1076,9 @@ int ip_tunnel_init_net(struct net *net, unsigned int ip_tnl_net_id,
if (!IS_ERR(itn->fb_tunnel_dev)) {
itn->fb_tunnel_dev->features |= NETIF_F_NETNS_LOCAL;
itn->fb_tunnel_dev->mtu = ip_tunnel_bind_dev(itn->fb_tunnel_dev);
- ip_tunnel_add(itn, netdev_priv(itn->fb_tunnel_dev));
itn->type = itn->fb_tunnel_dev->type;
+ rcu_assign_pointer(itn->fb_tun,
+ netdev_priv(itn->fb_tunnel_dev));
}
rtnl_unlock();
@@ -1262,6 +1265,8 @@ void ip_tunnel_uninit(struct net_device *dev)
/* fb_tunnel_dev will be unregisted in net-exit call. */
if (itn->fb_tunnel_dev != dev)
ip_tunnel_del(itn, netdev_priv(dev));
+ else
+ rcu_assign_pointer(itn->fb_tun, NULL);
dst_cache_reset(&tunnel->dst_cache);
}
In the datapath, the ip_tunnel_lookup() is used and it internally uses fallback tunnel device pointer, which is fb_tunnel_dev. This pointer is protected by RTNL. It's not enough to be used in the datapath. So, this pointer would be used after an interface is deleted. It eventually results in the use-after-free problem. In order to avoid the problem, the new tunnel pointer variable is added, which indicates a fallback tunnel device's tunnel pointer. This is protected by both RTNL and RCU. So, it's safe to be used in the datapath. Test commands: ip netns add A ip netns add B ip link add eth0 type veth peer name eth1 ip link set eth0 netns A ip link set eth1 netns B ip netns exec A ip link set lo up ip netns exec A ip link set eth0 up ip netns exec A ip link add gre1 type gre local 10.0.0.1 \ remote 10.0.0.2 ip netns exec A ip link set gre1 up ip netns exec A ip a a 10.0.100.1/24 dev gre1 ip netns exec A ip a a 10.0.0.1/24 dev eth0 ip netns exec B ip link set lo up ip netns exec B ip link set eth1 up ip netns exec B ip link add gre1 type gre local 10.0.0.2 \ remote 10.0.0.1 ip netns exec B ip link set gre1 up ip netns exec B ip a a 10.0.100.2/24 dev gre1 ip netns exec B ip a a 10.0.0.2/24 dev eth1 ip netns exec A hping3 10.0.100.2 -2 --flood -d 60000 & ip netns del B Splat looks like: [ 133.319668][ C3] BUG: KASAN: use-after-free in ip_tunnel_lookup+0x9d6/0xde0 [ 133.343852][ C3] Read of size 4 at addr ffff8880b1701c84 by task hping3/1222 [ 133.344724][ C3] [ 133.345002][ C3] CPU: 3 PID: 1222 Comm: hping3 Not tainted 5.7.0+ #591 [ 133.345814][ C3] Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006 [ 133.373336][ C3] Call Trace: [ 133.374792][ C3] <IRQ> [ 133.375205][ C3] dump_stack+0x96/0xdb [ 133.375789][ C3] print_address_description.constprop.6+0x2cc/0x450 [ 133.376720][ C3] ? ip_tunnel_lookup+0x9d6/0xde0 [ 133.377431][ C3] ? ip_tunnel_lookup+0x9d6/0xde0 [ 133.378130][ C3] ? ip_tunnel_lookup+0x9d6/0xde0 [ 133.378851][ C3] kasan_report+0x154/0x190 [ 133.379494][ C3] ? ip_tunnel_lookup+0x9d6/0xde0 [ 133.380200][ C3] ip_tunnel_lookup+0x9d6/0xde0 [ 133.380894][ C3] __ipgre_rcv+0x1ab/0xaa0 [ip_gre] [ 133.381630][ C3] ? rcu_read_lock_sched_held+0xc0/0xc0 [ 133.382429][ C3] gre_rcv+0x304/0x1910 [ip_gre] [ ... ] Fixes: c54419321455 ("GRE: Refactor GRE tunneling code.") Signed-off-by: Taehee Yoo <ap420073@gmail.com> --- include/net/ip_tunnels.h | 1 + net/ipv4/ip_tunnel.c | 11 ++++++++--- 2 files changed, 9 insertions(+), 3 deletions(-)