ipv6: invert join/leave anycast rtnl/socket locking order

Commit baf606d9c9b1 ("ipv4,ipv6: grab rtnl before locking the socket")
missed to update two setsockopt options, IPV6_JOIN_ANYCAST and
IPV6_LEAVE_ANYCAST, causing a lock inverstion regarding to the updated ones.

As ipv6_sock_ac_join and ipv6_sock_ac_leave are only called from
do_ipv6_setsockopt, we are good to just move the rtnl lock upper.

Fixes: baf606d9c9b1 ("ipv4,ipv6: grab rtnl before locking the socket")
Reported-by: Ying Huang <ying.huang@intel.com>
Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/net/ipv6/anycast.c b/net/ipv6/anycast.c
index baf2742..9e6b0ee 100644 (file)
--- a/net/ipv6/anycast.c
+++ b/net/ipv6/anycast.c
@@ -60,6 +60,8 @@ int ipv6_sock_ac_join(struct sock *sk, int ifindex, const struct in6_addr *addr)
        int     ishost = !net->ipv6.devconf_all->forwarding;
        int     err = 0;
 
+       ASSERT_RTNL();
+
        if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
                return -EPERM;
        if (ipv6_addr_is_multicast(addr))
@@ -73,7 +75,6 @@ int ipv6_sock_ac_join(struct sock *sk, int ifindex, const struct in6_addr *addr)
        pac->acl_next = NULL;
        pac->acl_addr = *addr;
 
-       rtnl_lock();
        if (ifindex == 0) {
                struct rt6_info *rt;
 
@@ -130,7 +131,6 @@ int ipv6_sock_ac_join(struct sock *sk, int ifindex, const struct in6_addr *addr)
        }
 
 error:
-       rtnl_unlock();
        if (pac)
                sock_kfree_s(sk, pac, sizeof(*pac));
        return err;
@@ -146,7 +146,8 @@ int ipv6_sock_ac_drop(struct sock *sk, int ifindex, const struct in6_addr *addr)
        struct ipv6_ac_socklist *pac, *prev_pac;
        struct net *net = sock_net(sk);
 
-       rtnl_lock();
+       ASSERT_RTNL();
+
        prev_pac = NULL;
        for (pac = np->ipv6_ac_list; pac; pac = pac->acl_next) {
                if ((ifindex == 0 || pac->acl_ifindex == ifindex) &&
@@ -154,10 +155,8 @@ int ipv6_sock_ac_drop(struct sock *sk, int ifindex, const struct in6_addr *addr)
                        break;
                prev_pac = pac;
        }
-       if (!pac) {
-               rtnl_unlock();
+       if (!pac)
                return -ENOENT;
-       }
        if (prev_pac)
                prev_pac->acl_next = pac->acl_next;
        else
@@ -166,7 +165,6 @@ int ipv6_sock_ac_drop(struct sock *sk, int ifindex, const struct in6_addr *addr)
        dev = __dev_get_by_index(net, pac->acl_ifindex);
        if (dev)
                ipv6_dev_ac_dec(dev, &pac->acl_addr);
-       rtnl_unlock();
 
        sock_kfree_s(sk, pac, sizeof(*pac));
        return 0;

diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c
index cc58837..9b2cb14 100644 (file)
--- a/net/ipv6/ipv6_sockglue.c
+++ b/net/ipv6/ipv6_sockglue.c
@@ -122,6 +122,8 @@ static bool setsockopt_needs_rtnl(int optname)
        switch (optname) {
        case IPV6_ADD_MEMBERSHIP:
        case IPV6_DROP_MEMBERSHIP:
+       case IPV6_JOIN_ANYCAST:
+       case IPV6_LEAVE_ANYCAST:
        case MCAST_JOIN_GROUP:
        case MCAST_LEAVE_GROUP:
        case MCAST_JOIN_SOURCE_GROUP: