Message ID | 20200623095945.1402468-3-Jason@zx2c4.com |
---|---|
State | New |
Headers | show |
Series | wireguard fixes for 5.8-rc3 | expand |
Hey Dmitry, The below patch is now in net.git and in Linus' tree. It adds a network namespace deletion notifier, which does some things that impact the logic of the interface and involves a little bit of object lifetime management. You'll see at the very bottom I added a test case to rule out the most pathological behavior. But naturally I'm wondering about the potential of subtle bugs beyond my gasp. It looks like syzkaller doesn't have any coverage of wg_netns_pre_exit -- it's all red. Do you suspect that's because the code is new enough that it just hasn't found it yet? Or should I start looking at teaching syzkaller a bit about interface netns changes? Jason On Tue, Jun 23, 2020 at 4:00 AM Jason A. Donenfeld <Jason@zx2c4.com> wrote: > > Before, we took a reference to the creating netns if the new netns was > different. This caused issues with circular references, with two > wireguard interfaces swapping namespaces. The solution is to rather not > take any extra references at all, but instead simply invalidate the > creating netns pointer when that netns is deleted. > > In order to prevent this from happening again, this commit improves the > rough object leak tracking by allowing it to account for created and > destroyed interfaces, aside from just peers and keys. That then makes it > possible to check for the object leak when having two interfaces take a > reference to each others' namespaces. > > Fixes: e7096c131e51 ("net: WireGuard secure network tunnel") > Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> > --- > drivers/net/wireguard/device.c | 58 ++++++++++------------ > drivers/net/wireguard/device.h | 3 +- > drivers/net/wireguard/netlink.c | 14 ++++-- > drivers/net/wireguard/socket.c | 25 +++++++--- > tools/testing/selftests/wireguard/netns.sh | 13 ++++- > 5 files changed, 67 insertions(+), 46 deletions(-) > > diff --git a/drivers/net/wireguard/device.c b/drivers/net/wireguard/device.c > index 3ac3f8570ca1..a8f151b1b5fa 100644 > --- a/drivers/net/wireguard/device.c > +++ b/drivers/net/wireguard/device.c > @@ -45,17 +45,18 @@ static int wg_open(struct net_device *dev) > if (dev_v6) > dev_v6->cnf.addr_gen_mode = IN6_ADDR_GEN_MODE_NONE; > > + mutex_lock(&wg->device_update_lock); > ret = wg_socket_init(wg, wg->incoming_port); > if (ret < 0) > - return ret; > - mutex_lock(&wg->device_update_lock); > + goto out; > list_for_each_entry(peer, &wg->peer_list, peer_list) { > wg_packet_send_staged_packets(peer); > if (peer->persistent_keepalive_interval) > wg_packet_send_keepalive(peer); > } > +out: > mutex_unlock(&wg->device_update_lock); > - return 0; > + return ret; > } > > #ifdef CONFIG_PM_SLEEP > @@ -225,6 +226,7 @@ static void wg_destruct(struct net_device *dev) > list_del(&wg->device_list); > rtnl_unlock(); > mutex_lock(&wg->device_update_lock); > + rcu_assign_pointer(wg->creating_net, NULL); > wg->incoming_port = 0; > wg_socket_reinit(wg, NULL, NULL); > /* The final references are cleared in the below calls to destroy_workqueue. */ > @@ -240,13 +242,11 @@ static void wg_destruct(struct net_device *dev) > skb_queue_purge(&wg->incoming_handshakes); > free_percpu(dev->tstats); > free_percpu(wg->incoming_handshakes_worker); > - if (wg->have_creating_net_ref) > - put_net(wg->creating_net); > kvfree(wg->index_hashtable); > kvfree(wg->peer_hashtable); > mutex_unlock(&wg->device_update_lock); > > - pr_debug("%s: Interface deleted\n", dev->name); > + pr_debug("%s: Interface destroyed\n", dev->name); > free_netdev(dev); > } > > @@ -292,7 +292,7 @@ static int wg_newlink(struct net *src_net, struct net_device *dev, > struct wg_device *wg = netdev_priv(dev); > int ret = -ENOMEM; > > - wg->creating_net = src_net; > + rcu_assign_pointer(wg->creating_net, src_net); > init_rwsem(&wg->static_identity.lock); > mutex_init(&wg->socket_update_lock); > mutex_init(&wg->device_update_lock); > @@ -393,30 +393,26 @@ static struct rtnl_link_ops link_ops __read_mostly = { > .newlink = wg_newlink, > }; > > -static int wg_netdevice_notification(struct notifier_block *nb, > - unsigned long action, void *data) > +static void wg_netns_pre_exit(struct net *net) > { > - struct net_device *dev = ((struct netdev_notifier_info *)data)->dev; > - struct wg_device *wg = netdev_priv(dev); > - > - ASSERT_RTNL(); > - > - if (action != NETDEV_REGISTER || dev->netdev_ops != &netdev_ops) > - return 0; > + struct wg_device *wg; > > - if (dev_net(dev) == wg->creating_net && wg->have_creating_net_ref) { > - put_net(wg->creating_net); > - wg->have_creating_net_ref = false; > - } else if (dev_net(dev) != wg->creating_net && > - !wg->have_creating_net_ref) { > - wg->have_creating_net_ref = true; > - get_net(wg->creating_net); > + rtnl_lock(); > + list_for_each_entry(wg, &device_list, device_list) { > + if (rcu_access_pointer(wg->creating_net) == net) { > + pr_debug("%s: Creating namespace exiting\n", wg->dev->name); > + netif_carrier_off(wg->dev); > + mutex_lock(&wg->device_update_lock); > + rcu_assign_pointer(wg->creating_net, NULL); > + wg_socket_reinit(wg, NULL, NULL); > + mutex_unlock(&wg->device_update_lock); > + } > } > - return 0; > + rtnl_unlock(); > } > > -static struct notifier_block netdevice_notifier = { > - .notifier_call = wg_netdevice_notification > +static struct pernet_operations pernet_ops = { > + .pre_exit = wg_netns_pre_exit > }; > > int __init wg_device_init(void) > @@ -429,18 +425,18 @@ int __init wg_device_init(void) > return ret; > #endif > > - ret = register_netdevice_notifier(&netdevice_notifier); > + ret = register_pernet_device(&pernet_ops); > if (ret) > goto error_pm; > > ret = rtnl_link_register(&link_ops); > if (ret) > - goto error_netdevice; > + goto error_pernet; > > return 0; > > -error_netdevice: > - unregister_netdevice_notifier(&netdevice_notifier); > +error_pernet: > + unregister_pernet_device(&pernet_ops); > error_pm: > #ifdef CONFIG_PM_SLEEP > unregister_pm_notifier(&pm_notifier); > @@ -451,7 +447,7 @@ int __init wg_device_init(void) > void wg_device_uninit(void) > { > rtnl_link_unregister(&link_ops); > - unregister_netdevice_notifier(&netdevice_notifier); > + unregister_pernet_device(&pernet_ops); > #ifdef CONFIG_PM_SLEEP > unregister_pm_notifier(&pm_notifier); > #endif > diff --git a/drivers/net/wireguard/device.h b/drivers/net/wireguard/device.h > index b15a8be9d816..4d0144e16947 100644 > --- a/drivers/net/wireguard/device.h > +++ b/drivers/net/wireguard/device.h > @@ -40,7 +40,7 @@ struct wg_device { > struct net_device *dev; > struct crypt_queue encrypt_queue, decrypt_queue; > struct sock __rcu *sock4, *sock6; > - struct net *creating_net; > + struct net __rcu *creating_net; > struct noise_static_identity static_identity; > struct workqueue_struct *handshake_receive_wq, *handshake_send_wq; > struct workqueue_struct *packet_crypt_wq; > @@ -56,7 +56,6 @@ struct wg_device { > unsigned int num_peers, device_update_gen; > u32 fwmark; > u16 incoming_port; > - bool have_creating_net_ref; > }; > > int wg_device_init(void); > diff --git a/drivers/net/wireguard/netlink.c b/drivers/net/wireguard/netlink.c > index 802099c8828a..20a4f3c0a0a1 100644 > --- a/drivers/net/wireguard/netlink.c > +++ b/drivers/net/wireguard/netlink.c > @@ -511,11 +511,15 @@ static int wg_set_device(struct sk_buff *skb, struct genl_info *info) > if (flags & ~__WGDEVICE_F_ALL) > goto out; > > - ret = -EPERM; > - if ((info->attrs[WGDEVICE_A_LISTEN_PORT] || > - info->attrs[WGDEVICE_A_FWMARK]) && > - !ns_capable(wg->creating_net->user_ns, CAP_NET_ADMIN)) > - goto out; > + if (info->attrs[WGDEVICE_A_LISTEN_PORT] || info->attrs[WGDEVICE_A_FWMARK]) { > + struct net *net; > + rcu_read_lock(); > + net = rcu_dereference(wg->creating_net); > + ret = !net || !ns_capable(net->user_ns, CAP_NET_ADMIN) ? -EPERM : 0; > + rcu_read_unlock(); > + if (ret) > + goto out; > + } > > ++wg->device_update_gen; > > diff --git a/drivers/net/wireguard/socket.c b/drivers/net/wireguard/socket.c > index f9018027fc13..c33e2c81635f 100644 > --- a/drivers/net/wireguard/socket.c > +++ b/drivers/net/wireguard/socket.c > @@ -347,6 +347,7 @@ static void set_sock_opts(struct socket *sock) > > int wg_socket_init(struct wg_device *wg, u16 port) > { > + struct net *net; > int ret; > struct udp_tunnel_sock_cfg cfg = { > .sk_user_data = wg, > @@ -371,37 +372,47 @@ int wg_socket_init(struct wg_device *wg, u16 port) > }; > #endif > > + rcu_read_lock(); > + net = rcu_dereference(wg->creating_net); > + net = net ? maybe_get_net(net) : NULL; > + rcu_read_unlock(); > + if (unlikely(!net)) > + return -ENONET; > + > #if IS_ENABLED(CONFIG_IPV6) > retry: > #endif > > - ret = udp_sock_create(wg->creating_net, &port4, &new4); > + ret = udp_sock_create(net, &port4, &new4); > if (ret < 0) { > pr_err("%s: Could not create IPv4 socket\n", wg->dev->name); > - return ret; > + goto out; > } > set_sock_opts(new4); > - setup_udp_tunnel_sock(wg->creating_net, new4, &cfg); > + setup_udp_tunnel_sock(net, new4, &cfg); > > #if IS_ENABLED(CONFIG_IPV6) > if (ipv6_mod_enabled()) { > port6.local_udp_port = inet_sk(new4->sk)->inet_sport; > - ret = udp_sock_create(wg->creating_net, &port6, &new6); > + ret = udp_sock_create(net, &port6, &new6); > if (ret < 0) { > udp_tunnel_sock_release(new4); > if (ret == -EADDRINUSE && !port && retries++ < 100) > goto retry; > pr_err("%s: Could not create IPv6 socket\n", > wg->dev->name); > - return ret; > + goto out; > } > set_sock_opts(new6); > - setup_udp_tunnel_sock(wg->creating_net, new6, &cfg); > + setup_udp_tunnel_sock(net, new6, &cfg); > } > #endif > > wg_socket_reinit(wg, new4->sk, new6 ? new6->sk : NULL); > - return 0; > + ret = 0; > +out: > + put_net(net); > + return ret; > } > > void wg_socket_reinit(struct wg_device *wg, struct sock *new4, > diff --git a/tools/testing/selftests/wireguard/netns.sh b/tools/testing/selftests/wireguard/netns.sh > index 17a1f53ceba0..d77f4829f1e0 100755 > --- a/tools/testing/selftests/wireguard/netns.sh > +++ b/tools/testing/selftests/wireguard/netns.sh > @@ -587,9 +587,20 @@ ip0 link set wg0 up > kill $ncat_pid > ip0 link del wg0 > > +# Ensure there aren't circular reference loops > +ip1 link add wg1 type wireguard > +ip2 link add wg2 type wireguard > +ip1 link set wg1 netns $netns2 > +ip2 link set wg2 netns $netns1 > +pp ip netns delete $netns1 > +pp ip netns delete $netns2 > +pp ip netns add $netns1 > +pp ip netns add $netns2 > + > +sleep 2 # Wait for cleanup and grace periods > declare -A objects > while read -t 0.1 -r line 2>/dev/null || [[ $? -ne 142 ]]; do > - [[ $line =~ .*(wg[0-9]+:\ [A-Z][a-z]+\ [0-9]+)\ .*(created|destroyed).* ]] || continue > + [[ $line =~ .*(wg[0-9]+:\ [A-Z][a-z]+\ ?[0-9]*)\ .*(created|destroyed).* ]] || continue > objects["${BASH_REMATCH[1]}"]+="${BASH_REMATCH[2]}" > done < /dev/kmsg > alldeleted=1 > -- > 2.27.0
On Fri, Jun 26, 2020 at 10:59 PM Jason A. Donenfeld <Jason@zx2c4.com> wrote: > > Hey Dmitry, > > The below patch is now in net.git and in Linus' tree. It adds a > network namespace deletion notifier, which does some things that > impact the logic of the interface and involves a little bit of object > lifetime management. You'll see at the very bottom I added a test case > to rule out the most pathological behavior. But naturally I'm > wondering about the potential of subtle bugs beyond my gasp. It looks > like syzkaller doesn't have any coverage of wg_netns_pre_exit -- it's > all red. Do you suspect that's because the code is new enough that it > just hasn't found it yet? Or should I start looking at teaching > syzkaller a bit about interface netns changes? +syzkaller mailing list Hi Jason, Hard to say. syzkaller frequently needs some time (days) to get reasonable coverage of new code. Is wg_netns_pre_exit executed synchronously in the context of a syscall? If not, then it won't be shown as covered. If yes, then what syscall is it? This is related to namespaces, right? syzkaller has some descriptions for namespaces: https://github.com/google/syzkaller/blob/master/sys/linux/namespaces.txt But I don't know if it's enough nor I checked if they actually work. We have this laundry list: https://github.com/google/syzkaller/issues/533 and some interns working on adding more descriptions. If you can identify something that's not covered but can be covered, please add to that list. I think we can even prioritize it then b/c most items on that list don't have anybody actively interested. > On Tue, Jun 23, 2020 at 4:00 AM Jason A. Donenfeld <Jason@zx2c4.com> wrote: > > > > Before, we took a reference to the creating netns if the new netns was > > different. This caused issues with circular references, with two > > wireguard interfaces swapping namespaces. The solution is to rather not > > take any extra references at all, but instead simply invalidate the > > creating netns pointer when that netns is deleted. > > > > In order to prevent this from happening again, this commit improves the > > rough object leak tracking by allowing it to account for created and > > destroyed interfaces, aside from just peers and keys. That then makes it > > possible to check for the object leak when having two interfaces take a > > reference to each others' namespaces. > > > > Fixes: e7096c131e51 ("net: WireGuard secure network tunnel") > > Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> > > --- > > drivers/net/wireguard/device.c | 58 ++++++++++------------ > > drivers/net/wireguard/device.h | 3 +- > > drivers/net/wireguard/netlink.c | 14 ++++-- > > drivers/net/wireguard/socket.c | 25 +++++++--- > > tools/testing/selftests/wireguard/netns.sh | 13 ++++- > > 5 files changed, 67 insertions(+), 46 deletions(-) > > > > diff --git a/drivers/net/wireguard/device.c b/drivers/net/wireguard/device.c > > index 3ac3f8570ca1..a8f151b1b5fa 100644 > > --- a/drivers/net/wireguard/device.c > > +++ b/drivers/net/wireguard/device.c > > @@ -45,17 +45,18 @@ static int wg_open(struct net_device *dev) > > if (dev_v6) > > dev_v6->cnf.addr_gen_mode = IN6_ADDR_GEN_MODE_NONE; > > > > + mutex_lock(&wg->device_update_lock); > > ret = wg_socket_init(wg, wg->incoming_port); > > if (ret < 0) > > - return ret; > > - mutex_lock(&wg->device_update_lock); > > + goto out; > > list_for_each_entry(peer, &wg->peer_list, peer_list) { > > wg_packet_send_staged_packets(peer); > > if (peer->persistent_keepalive_interval) > > wg_packet_send_keepalive(peer); > > } > > +out: > > mutex_unlock(&wg->device_update_lock); > > - return 0; > > + return ret; > > } > > > > #ifdef CONFIG_PM_SLEEP > > @@ -225,6 +226,7 @@ static void wg_destruct(struct net_device *dev) > > list_del(&wg->device_list); > > rtnl_unlock(); > > mutex_lock(&wg->device_update_lock); > > + rcu_assign_pointer(wg->creating_net, NULL); > > wg->incoming_port = 0; > > wg_socket_reinit(wg, NULL, NULL); > > /* The final references are cleared in the below calls to destroy_workqueue. */ > > @@ -240,13 +242,11 @@ static void wg_destruct(struct net_device *dev) > > skb_queue_purge(&wg->incoming_handshakes); > > free_percpu(dev->tstats); > > free_percpu(wg->incoming_handshakes_worker); > > - if (wg->have_creating_net_ref) > > - put_net(wg->creating_net); > > kvfree(wg->index_hashtable); > > kvfree(wg->peer_hashtable); > > mutex_unlock(&wg->device_update_lock); > > > > - pr_debug("%s: Interface deleted\n", dev->name); > > + pr_debug("%s: Interface destroyed\n", dev->name); > > free_netdev(dev); > > } > > > > @@ -292,7 +292,7 @@ static int wg_newlink(struct net *src_net, struct net_device *dev, > > struct wg_device *wg = netdev_priv(dev); > > int ret = -ENOMEM; > > > > - wg->creating_net = src_net; > > + rcu_assign_pointer(wg->creating_net, src_net); > > init_rwsem(&wg->static_identity.lock); > > mutex_init(&wg->socket_update_lock); > > mutex_init(&wg->device_update_lock); > > @@ -393,30 +393,26 @@ static struct rtnl_link_ops link_ops __read_mostly = { > > .newlink = wg_newlink, > > }; > > > > -static int wg_netdevice_notification(struct notifier_block *nb, > > - unsigned long action, void *data) > > +static void wg_netns_pre_exit(struct net *net) > > { > > - struct net_device *dev = ((struct netdev_notifier_info *)data)->dev; > > - struct wg_device *wg = netdev_priv(dev); > > - > > - ASSERT_RTNL(); > > - > > - if (action != NETDEV_REGISTER || dev->netdev_ops != &netdev_ops) > > - return 0; > > + struct wg_device *wg; > > > > - if (dev_net(dev) == wg->creating_net && wg->have_creating_net_ref) { > > - put_net(wg->creating_net); > > - wg->have_creating_net_ref = false; > > - } else if (dev_net(dev) != wg->creating_net && > > - !wg->have_creating_net_ref) { > > - wg->have_creating_net_ref = true; > > - get_net(wg->creating_net); > > + rtnl_lock(); > > + list_for_each_entry(wg, &device_list, device_list) { > > + if (rcu_access_pointer(wg->creating_net) == net) { > > + pr_debug("%s: Creating namespace exiting\n", wg->dev->name); > > + netif_carrier_off(wg->dev); > > + mutex_lock(&wg->device_update_lock); > > + rcu_assign_pointer(wg->creating_net, NULL); > > + wg_socket_reinit(wg, NULL, NULL); > > + mutex_unlock(&wg->device_update_lock); > > + } > > } > > - return 0; > > + rtnl_unlock(); > > } > > > > -static struct notifier_block netdevice_notifier = { > > - .notifier_call = wg_netdevice_notification > > +static struct pernet_operations pernet_ops = { > > + .pre_exit = wg_netns_pre_exit > > }; > > > > int __init wg_device_init(void) > > @@ -429,18 +425,18 @@ int __init wg_device_init(void) > > return ret; > > #endif > > > > - ret = register_netdevice_notifier(&netdevice_notifier); > > + ret = register_pernet_device(&pernet_ops); > > if (ret) > > goto error_pm; > > > > ret = rtnl_link_register(&link_ops); > > if (ret) > > - goto error_netdevice; > > + goto error_pernet; > > > > return 0; > > > > -error_netdevice: > > - unregister_netdevice_notifier(&netdevice_notifier); > > +error_pernet: > > + unregister_pernet_device(&pernet_ops); > > error_pm: > > #ifdef CONFIG_PM_SLEEP > > unregister_pm_notifier(&pm_notifier); > > @@ -451,7 +447,7 @@ int __init wg_device_init(void) > > void wg_device_uninit(void) > > { > > rtnl_link_unregister(&link_ops); > > - unregister_netdevice_notifier(&netdevice_notifier); > > + unregister_pernet_device(&pernet_ops); > > #ifdef CONFIG_PM_SLEEP > > unregister_pm_notifier(&pm_notifier); > > #endif > > diff --git a/drivers/net/wireguard/device.h b/drivers/net/wireguard/device.h > > index b15a8be9d816..4d0144e16947 100644 > > --- a/drivers/net/wireguard/device.h > > +++ b/drivers/net/wireguard/device.h > > @@ -40,7 +40,7 @@ struct wg_device { > > struct net_device *dev; > > struct crypt_queue encrypt_queue, decrypt_queue; > > struct sock __rcu *sock4, *sock6; > > - struct net *creating_net; > > + struct net __rcu *creating_net; > > struct noise_static_identity static_identity; > > struct workqueue_struct *handshake_receive_wq, *handshake_send_wq; > > struct workqueue_struct *packet_crypt_wq; > > @@ -56,7 +56,6 @@ struct wg_device { > > unsigned int num_peers, device_update_gen; > > u32 fwmark; > > u16 incoming_port; > > - bool have_creating_net_ref; > > }; > > > > int wg_device_init(void); > > diff --git a/drivers/net/wireguard/netlink.c b/drivers/net/wireguard/netlink.c > > index 802099c8828a..20a4f3c0a0a1 100644 > > --- a/drivers/net/wireguard/netlink.c > > +++ b/drivers/net/wireguard/netlink.c > > @@ -511,11 +511,15 @@ static int wg_set_device(struct sk_buff *skb, struct genl_info *info) > > if (flags & ~__WGDEVICE_F_ALL) > > goto out; > > > > - ret = -EPERM; > > - if ((info->attrs[WGDEVICE_A_LISTEN_PORT] || > > - info->attrs[WGDEVICE_A_FWMARK]) && > > - !ns_capable(wg->creating_net->user_ns, CAP_NET_ADMIN)) > > - goto out; > > + if (info->attrs[WGDEVICE_A_LISTEN_PORT] || info->attrs[WGDEVICE_A_FWMARK]) { > > + struct net *net; > > + rcu_read_lock(); > > + net = rcu_dereference(wg->creating_net); > > + ret = !net || !ns_capable(net->user_ns, CAP_NET_ADMIN) ? -EPERM : 0; > > + rcu_read_unlock(); > > + if (ret) > > + goto out; > > + } > > > > ++wg->device_update_gen; > > > > diff --git a/drivers/net/wireguard/socket.c b/drivers/net/wireguard/socket.c > > index f9018027fc13..c33e2c81635f 100644 > > --- a/drivers/net/wireguard/socket.c > > +++ b/drivers/net/wireguard/socket.c > > @@ -347,6 +347,7 @@ static void set_sock_opts(struct socket *sock) > > > > int wg_socket_init(struct wg_device *wg, u16 port) > > { > > + struct net *net; > > int ret; > > struct udp_tunnel_sock_cfg cfg = { > > .sk_user_data = wg, > > @@ -371,37 +372,47 @@ int wg_socket_init(struct wg_device *wg, u16 port) > > }; > > #endif > > > > + rcu_read_lock(); > > + net = rcu_dereference(wg->creating_net); > > + net = net ? maybe_get_net(net) : NULL; > > + rcu_read_unlock(); > > + if (unlikely(!net)) > > + return -ENONET; > > + > > #if IS_ENABLED(CONFIG_IPV6) > > retry: > > #endif > > > > - ret = udp_sock_create(wg->creating_net, &port4, &new4); > > + ret = udp_sock_create(net, &port4, &new4); > > if (ret < 0) { > > pr_err("%s: Could not create IPv4 socket\n", wg->dev->name); > > - return ret; > > + goto out; > > } > > set_sock_opts(new4); > > - setup_udp_tunnel_sock(wg->creating_net, new4, &cfg); > > + setup_udp_tunnel_sock(net, new4, &cfg); > > > > #if IS_ENABLED(CONFIG_IPV6) > > if (ipv6_mod_enabled()) { > > port6.local_udp_port = inet_sk(new4->sk)->inet_sport; > > - ret = udp_sock_create(wg->creating_net, &port6, &new6); > > + ret = udp_sock_create(net, &port6, &new6); > > if (ret < 0) { > > udp_tunnel_sock_release(new4); > > if (ret == -EADDRINUSE && !port && retries++ < 100) > > goto retry; > > pr_err("%s: Could not create IPv6 socket\n", > > wg->dev->name); > > - return ret; > > + goto out; > > } > > set_sock_opts(new6); > > - setup_udp_tunnel_sock(wg->creating_net, new6, &cfg); > > + setup_udp_tunnel_sock(net, new6, &cfg); > > } > > #endif > > > > wg_socket_reinit(wg, new4->sk, new6 ? new6->sk : NULL); > > - return 0; > > + ret = 0; > > +out: > > + put_net(net); > > + return ret; > > } > > > > void wg_socket_reinit(struct wg_device *wg, struct sock *new4, > > diff --git a/tools/testing/selftests/wireguard/netns.sh b/tools/testing/selftests/wireguard/netns.sh > > index 17a1f53ceba0..d77f4829f1e0 100755 > > --- a/tools/testing/selftests/wireguard/netns.sh > > +++ b/tools/testing/selftests/wireguard/netns.sh > > @@ -587,9 +587,20 @@ ip0 link set wg0 up > > kill $ncat_pid > > ip0 link del wg0 > > > > +# Ensure there aren't circular reference loops > > +ip1 link add wg1 type wireguard > > +ip2 link add wg2 type wireguard > > +ip1 link set wg1 netns $netns2 > > +ip2 link set wg2 netns $netns1 > > +pp ip netns delete $netns1 > > +pp ip netns delete $netns2 > > +pp ip netns add $netns1 > > +pp ip netns add $netns2 > > + > > +sleep 2 # Wait for cleanup and grace periods > > declare -A objects > > while read -t 0.1 -r line 2>/dev/null || [[ $? -ne 142 ]]; do > > - [[ $line =~ .*(wg[0-9]+:\ [A-Z][a-z]+\ [0-9]+)\ .*(created|destroyed).* ]] || continue > > + [[ $line =~ .*(wg[0-9]+:\ [A-Z][a-z]+\ ?[0-9]*)\ .*(created|destroyed).* ]] || continue > > objects["${BASH_REMATCH[1]}"]+="${BASH_REMATCH[2]}" > > done < /dev/kmsg > > alldeleted=1 > > -- > > 2.27.0
Hi Dmitry, On Sat, Jun 27, 2020 at 2:59 AM Dmitry Vyukov <dvyukov@google.com> wrote: > Hard to say. syzkaller frequently needs some time (days) to get > reasonable coverage of new code. > Is wg_netns_pre_exit executed synchronously in the context of a > syscall? If not, then it won't be shown as covered. If yes, then what > syscall is it? Aw, shucks, you're right. I thought that this would be from the syscall path for deleting a network namespace via netns_put, but sticking a dump_stack() in there, it's clear that namespace cleanup actually happens in a worker: [ 0.407072] wg_netns_pre_exit+0xc/0x98 [ 0.408496] cleanup_net+0x1bd/0x2f0 [ 0.409844] process_one_work+0x17f/0x2d0 [ 0.411327] worker_thread+0x4b/0x3b0 [ 0.412697] ? rescuer_thread+0x360/0x360 [ 0.414169] kthread+0x116/0x140 [ 0.415368] ? __kthread_create_worker+0x110/0x110 [ 0.417125] ret_from_fork+0x1f/0x30 > This is related to namespaces, right? syzkaller has some descriptions > for namespaces: > https://github.com/google/syzkaller/blob/master/sys/linux/namespaces.txt > But I don't know if it's enough nor I checked if they actually work. > We have this laundry list: > https://github.com/google/syzkaller/issues/533 > and some interns working on adding more descriptions. If you can > identify something that's not covered but can be covered, please add > to that list. I think we can even prioritize it then b/c most items on > that list don't have anybody actively interested. Something to add to that list would be: - Create an interface of a given type (wg0->wireguard, for example) - Move it to a namespace - Delete a namespace - Delete an interface Some combination of those things might provoke issues. One would be what the original post of this email thread was about. Another would be that xfrmi OOPS from a few weeks ago, fixed with c95c5f58b35e ("xfrm interface: fix oops when deleting a x-netns interface"), which has a reproducer in the commit message pretty similar to the one here. There's an even older one in xfrmi, fixed with c5d1030f2300 ("xfrm interface: fix list corruption for x-netns"). In short, adding and deleting namespaces and moving interfaces around between them, amidst other syzkaller chaos, might turn up bugs, because in the past humans have found those sorts of bugs doing the same. Jason
On Sat, Jun 27, 2020 at 11:51 AM Jason A. Donenfeld <Jason@zx2c4.com> wrote: > > Hi Dmitry, > > On Sat, Jun 27, 2020 at 2:59 AM Dmitry Vyukov <dvyukov@google.com> wrote: > > Hard to say. syzkaller frequently needs some time (days) to get > > reasonable coverage of new code. > > Is wg_netns_pre_exit executed synchronously in the context of a > > syscall? If not, then it won't be shown as covered. If yes, then what > > syscall is it? > > Aw, shucks, you're right. I thought that this would be from the > syscall path for deleting a network namespace via netns_put, but > sticking a dump_stack() in there, it's clear that namespace cleanup > actually happens in a worker: > > [ 0.407072] wg_netns_pre_exit+0xc/0x98 > [ 0.408496] cleanup_net+0x1bd/0x2f0 > [ 0.409844] process_one_work+0x17f/0x2d0 > [ 0.411327] worker_thread+0x4b/0x3b0 > [ 0.412697] ? rescuer_thread+0x360/0x360 > [ 0.414169] kthread+0x116/0x140 > [ 0.415368] ? __kthread_create_worker+0x110/0x110 > [ 0.417125] ret_from_fork+0x1f/0x30 > > > This is related to namespaces, right? syzkaller has some descriptions > > for namespaces: > > https://github.com/google/syzkaller/blob/master/sys/linux/namespaces.txt > > But I don't know if it's enough nor I checked if they actually work. > > We have this laundry list: > > https://github.com/google/syzkaller/issues/533 > > and some interns working on adding more descriptions. If you can > > identify something that's not covered but can be covered, please add > > to that list. I think we can even prioritize it then b/c most items on > > that list don't have anybody actively interested. > > Something to add to that list would be: > > - Create an interface of a given type (wg0->wireguard, for example) > - Move it to a namespace > - Delete a namespace > - Delete an interface What are the kernel interfaces for these? Some syscalls? /dev files? People working on these will need at least some "entry pointer" to start exploring involved interfaces. > Some combination of those things might provoke issues. One would be > what the original post of this email thread was about. Another would > be that xfrmi OOPS from a few weeks ago, fixed with c95c5f58b35e > ("xfrm interface: fix oops when deleting a x-netns interface"), which > has a reproducer in the commit message pretty similar to the one here. > There's an even older one in xfrmi, fixed with c5d1030f2300 ("xfrm > interface: fix list corruption for x-netns"). > > In short, adding and deleting namespaces and moving interfaces around > between them, amidst other syzkaller chaos, might turn up bugs, > because in the past humans have found those sorts of bugs doing the > same. > > Jason
diff --git a/drivers/net/wireguard/device.c b/drivers/net/wireguard/device.c index 3ac3f8570ca1..a8f151b1b5fa 100644 --- a/drivers/net/wireguard/device.c +++ b/drivers/net/wireguard/device.c @@ -45,17 +45,18 @@ static int wg_open(struct net_device *dev) if (dev_v6) dev_v6->cnf.addr_gen_mode = IN6_ADDR_GEN_MODE_NONE; + mutex_lock(&wg->device_update_lock); ret = wg_socket_init(wg, wg->incoming_port); if (ret < 0) - return ret; - mutex_lock(&wg->device_update_lock); + goto out; list_for_each_entry(peer, &wg->peer_list, peer_list) { wg_packet_send_staged_packets(peer); if (peer->persistent_keepalive_interval) wg_packet_send_keepalive(peer); } +out: mutex_unlock(&wg->device_update_lock); - return 0; + return ret; } #ifdef CONFIG_PM_SLEEP @@ -225,6 +226,7 @@ static void wg_destruct(struct net_device *dev) list_del(&wg->device_list); rtnl_unlock(); mutex_lock(&wg->device_update_lock); + rcu_assign_pointer(wg->creating_net, NULL); wg->incoming_port = 0; wg_socket_reinit(wg, NULL, NULL); /* The final references are cleared in the below calls to destroy_workqueue. */ @@ -240,13 +242,11 @@ static void wg_destruct(struct net_device *dev) skb_queue_purge(&wg->incoming_handshakes); free_percpu(dev->tstats); free_percpu(wg->incoming_handshakes_worker); - if (wg->have_creating_net_ref) - put_net(wg->creating_net); kvfree(wg->index_hashtable); kvfree(wg->peer_hashtable); mutex_unlock(&wg->device_update_lock); - pr_debug("%s: Interface deleted\n", dev->name); + pr_debug("%s: Interface destroyed\n", dev->name); free_netdev(dev); } @@ -292,7 +292,7 @@ static int wg_newlink(struct net *src_net, struct net_device *dev, struct wg_device *wg = netdev_priv(dev); int ret = -ENOMEM; - wg->creating_net = src_net; + rcu_assign_pointer(wg->creating_net, src_net); init_rwsem(&wg->static_identity.lock); mutex_init(&wg->socket_update_lock); mutex_init(&wg->device_update_lock); @@ -393,30 +393,26 @@ static struct rtnl_link_ops link_ops __read_mostly = { .newlink = wg_newlink, }; -static int wg_netdevice_notification(struct notifier_block *nb, - unsigned long action, void *data) +static void wg_netns_pre_exit(struct net *net) { - struct net_device *dev = ((struct netdev_notifier_info *)data)->dev; - struct wg_device *wg = netdev_priv(dev); - - ASSERT_RTNL(); - - if (action != NETDEV_REGISTER || dev->netdev_ops != &netdev_ops) - return 0; + struct wg_device *wg; - if (dev_net(dev) == wg->creating_net && wg->have_creating_net_ref) { - put_net(wg->creating_net); - wg->have_creating_net_ref = false; - } else if (dev_net(dev) != wg->creating_net && - !wg->have_creating_net_ref) { - wg->have_creating_net_ref = true; - get_net(wg->creating_net); + rtnl_lock(); + list_for_each_entry(wg, &device_list, device_list) { + if (rcu_access_pointer(wg->creating_net) == net) { + pr_debug("%s: Creating namespace exiting\n", wg->dev->name); + netif_carrier_off(wg->dev); + mutex_lock(&wg->device_update_lock); + rcu_assign_pointer(wg->creating_net, NULL); + wg_socket_reinit(wg, NULL, NULL); + mutex_unlock(&wg->device_update_lock); + } } - return 0; + rtnl_unlock(); } -static struct notifier_block netdevice_notifier = { - .notifier_call = wg_netdevice_notification +static struct pernet_operations pernet_ops = { + .pre_exit = wg_netns_pre_exit }; int __init wg_device_init(void) @@ -429,18 +425,18 @@ int __init wg_device_init(void) return ret; #endif - ret = register_netdevice_notifier(&netdevice_notifier); + ret = register_pernet_device(&pernet_ops); if (ret) goto error_pm; ret = rtnl_link_register(&link_ops); if (ret) - goto error_netdevice; + goto error_pernet; return 0; -error_netdevice: - unregister_netdevice_notifier(&netdevice_notifier); +error_pernet: + unregister_pernet_device(&pernet_ops); error_pm: #ifdef CONFIG_PM_SLEEP unregister_pm_notifier(&pm_notifier); @@ -451,7 +447,7 @@ int __init wg_device_init(void) void wg_device_uninit(void) { rtnl_link_unregister(&link_ops); - unregister_netdevice_notifier(&netdevice_notifier); + unregister_pernet_device(&pernet_ops); #ifdef CONFIG_PM_SLEEP unregister_pm_notifier(&pm_notifier); #endif diff --git a/drivers/net/wireguard/device.h b/drivers/net/wireguard/device.h index b15a8be9d816..4d0144e16947 100644 --- a/drivers/net/wireguard/device.h +++ b/drivers/net/wireguard/device.h @@ -40,7 +40,7 @@ struct wg_device { struct net_device *dev; struct crypt_queue encrypt_queue, decrypt_queue; struct sock __rcu *sock4, *sock6; - struct net *creating_net; + struct net __rcu *creating_net; struct noise_static_identity static_identity; struct workqueue_struct *handshake_receive_wq, *handshake_send_wq; struct workqueue_struct *packet_crypt_wq; @@ -56,7 +56,6 @@ struct wg_device { unsigned int num_peers, device_update_gen; u32 fwmark; u16 incoming_port; - bool have_creating_net_ref; }; int wg_device_init(void); diff --git a/drivers/net/wireguard/netlink.c b/drivers/net/wireguard/netlink.c index 802099c8828a..20a4f3c0a0a1 100644 --- a/drivers/net/wireguard/netlink.c +++ b/drivers/net/wireguard/netlink.c @@ -511,11 +511,15 @@ static int wg_set_device(struct sk_buff *skb, struct genl_info *info) if (flags & ~__WGDEVICE_F_ALL) goto out; - ret = -EPERM; - if ((info->attrs[WGDEVICE_A_LISTEN_PORT] || - info->attrs[WGDEVICE_A_FWMARK]) && - !ns_capable(wg->creating_net->user_ns, CAP_NET_ADMIN)) - goto out; + if (info->attrs[WGDEVICE_A_LISTEN_PORT] || info->attrs[WGDEVICE_A_FWMARK]) { + struct net *net; + rcu_read_lock(); + net = rcu_dereference(wg->creating_net); + ret = !net || !ns_capable(net->user_ns, CAP_NET_ADMIN) ? -EPERM : 0; + rcu_read_unlock(); + if (ret) + goto out; + } ++wg->device_update_gen; diff --git a/drivers/net/wireguard/socket.c b/drivers/net/wireguard/socket.c index f9018027fc13..c33e2c81635f 100644 --- a/drivers/net/wireguard/socket.c +++ b/drivers/net/wireguard/socket.c @@ -347,6 +347,7 @@ static void set_sock_opts(struct socket *sock) int wg_socket_init(struct wg_device *wg, u16 port) { + struct net *net; int ret; struct udp_tunnel_sock_cfg cfg = { .sk_user_data = wg, @@ -371,37 +372,47 @@ int wg_socket_init(struct wg_device *wg, u16 port) }; #endif + rcu_read_lock(); + net = rcu_dereference(wg->creating_net); + net = net ? maybe_get_net(net) : NULL; + rcu_read_unlock(); + if (unlikely(!net)) + return -ENONET; + #if IS_ENABLED(CONFIG_IPV6) retry: #endif - ret = udp_sock_create(wg->creating_net, &port4, &new4); + ret = udp_sock_create(net, &port4, &new4); if (ret < 0) { pr_err("%s: Could not create IPv4 socket\n", wg->dev->name); - return ret; + goto out; } set_sock_opts(new4); - setup_udp_tunnel_sock(wg->creating_net, new4, &cfg); + setup_udp_tunnel_sock(net, new4, &cfg); #if IS_ENABLED(CONFIG_IPV6) if (ipv6_mod_enabled()) { port6.local_udp_port = inet_sk(new4->sk)->inet_sport; - ret = udp_sock_create(wg->creating_net, &port6, &new6); + ret = udp_sock_create(net, &port6, &new6); if (ret < 0) { udp_tunnel_sock_release(new4); if (ret == -EADDRINUSE && !port && retries++ < 100) goto retry; pr_err("%s: Could not create IPv6 socket\n", wg->dev->name); - return ret; + goto out; } set_sock_opts(new6); - setup_udp_tunnel_sock(wg->creating_net, new6, &cfg); + setup_udp_tunnel_sock(net, new6, &cfg); } #endif wg_socket_reinit(wg, new4->sk, new6 ? new6->sk : NULL); - return 0; + ret = 0; +out: + put_net(net); + return ret; } void wg_socket_reinit(struct wg_device *wg, struct sock *new4, diff --git a/tools/testing/selftests/wireguard/netns.sh b/tools/testing/selftests/wireguard/netns.sh index 17a1f53ceba0..d77f4829f1e0 100755 --- a/tools/testing/selftests/wireguard/netns.sh +++ b/tools/testing/selftests/wireguard/netns.sh @@ -587,9 +587,20 @@ ip0 link set wg0 up kill $ncat_pid ip0 link del wg0 +# Ensure there aren't circular reference loops +ip1 link add wg1 type wireguard +ip2 link add wg2 type wireguard +ip1 link set wg1 netns $netns2 +ip2 link set wg2 netns $netns1 +pp ip netns delete $netns1 +pp ip netns delete $netns2 +pp ip netns add $netns1 +pp ip netns add $netns2 + +sleep 2 # Wait for cleanup and grace periods declare -A objects while read -t 0.1 -r line 2>/dev/null || [[ $? -ne 142 ]]; do - [[ $line =~ .*(wg[0-9]+:\ [A-Z][a-z]+\ [0-9]+)\ .*(created|destroyed).* ]] || continue + [[ $line =~ .*(wg[0-9]+:\ [A-Z][a-z]+\ ?[0-9]*)\ .*(created|destroyed).* ]] || continue objects["${BASH_REMATCH[1]}"]+="${BASH_REMATCH[2]}" done < /dev/kmsg alldeleted=1
Before, we took a reference to the creating netns if the new netns was different. This caused issues with circular references, with two wireguard interfaces swapping namespaces. The solution is to rather not take any extra references at all, but instead simply invalidate the creating netns pointer when that netns is deleted. In order to prevent this from happening again, this commit improves the rough object leak tracking by allowing it to account for created and destroyed interfaces, aside from just peers and keys. That then makes it possible to check for the object leak when having two interfaces take a reference to each others' namespaces. Fixes: e7096c131e51 ("net: WireGuard secure network tunnel") Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> --- drivers/net/wireguard/device.c | 58 ++++++++++------------ drivers/net/wireguard/device.h | 3 +- drivers/net/wireguard/netlink.c | 14 ++++-- drivers/net/wireguard/socket.c | 25 +++++++--- tools/testing/selftests/wireguard/netns.sh | 13 ++++- 5 files changed, 67 insertions(+), 46 deletions(-) -- 2.27.0