Merge remote-tracking branch 'ipsec/master'

[karo-tx-linux.git] / net / ipv4 / inet_hashtables.c

diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c
index 89120196a94934e3bb1f201eef9ad7f936ad828b..ccc5980797fcdb9ed3a1003db47e4e8cb7180279 100644 (file)
--- a/net/ipv4/inet_hashtables.c
+++ b/net/ipv4/inet_hashtables.c
@@ -126,7 +126,7 @@ void inet_put_port(struct sock *sk)
 }
 EXPORT_SYMBOL(inet_put_port);
 
-int __inet_inherit_port(struct sock *sk, struct sock *child)
+int __inet_inherit_port(const struct sock *sk, struct sock *child)
 {
        struct inet_hashinfo *table = sk->sk_prot->h.hashinfo;
        unsigned short port = inet_sk(child)->inet_num;
@@ -137,6 +137,10 @@ int __inet_inherit_port(struct sock *sk, struct sock *child)
 
        spin_lock(&head->lock);
        tb = inet_csk(sk)->icsk_bind_hash;
+       if (unlikely(!tb)) {
+               spin_unlock(&head->lock);
+               return -ENOENT;
+       }
        if (tb->port != port) {
                /* NOTE: using tproxy and redirecting skbs to a proxy
                 * on a different listener port breaks the assumption
@@ -185,6 +189,8 @@ static inline int compute_score(struct sock *sk, struct net *net,
                                return -1;
                        score += 4;
                }
+               if (sk->sk_incoming_cpu == raw_smp_processor_id())
+                       score++;
        }
        return score;
 }
@@ -398,14 +404,18 @@ static u32 inet_sk_port_offset(const struct sock *sk)
                                          inet->inet_dport);
 }
 
-void __inet_hash_nolisten(struct sock *sk, struct sock *osk)
+/* insert a socket into ehash, and eventually remove another one
+ * (The another one can be a SYN_RECV or TIMEWAIT
+ */
+bool inet_ehash_insert(struct sock *sk, struct sock *osk)
 {
        struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo;
        struct hlist_nulls_head *list;
        struct inet_ehash_bucket *head;
        spinlock_t *lock;
+       bool ret = true;
 
-       WARN_ON(!sk_unhashed(sk));
+       WARN_ON_ONCE(!sk_unhashed(sk));
 
        sk->sk_hash = sk_ehashfn(sk);
        head = inet_ehash_bucket(hashinfo, sk->sk_hash);
@@ -413,24 +423,41 @@ void __inet_hash_nolisten(struct sock *sk, struct sock *osk)
        lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
 
        spin_lock(lock);
-       __sk_nulls_add_node_rcu(sk, list);
        if (osk) {
-               WARN_ON(sk->sk_hash != osk->sk_hash);
-               sk_nulls_del_node_init_rcu(osk);
+               WARN_ON_ONCE(sk->sk_hash != osk->sk_hash);
+               ret = sk_nulls_del_node_init_rcu(osk);
        }
+       if (ret)
+               __sk_nulls_add_node_rcu(sk, list);
        spin_unlock(lock);
-       sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
+       return ret;
+}
+
+bool inet_ehash_nolisten(struct sock *sk, struct sock *osk)
+{
+       bool ok = inet_ehash_insert(sk, osk);
+
+       if (ok) {
+               sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
+       } else {
+               percpu_counter_inc(sk->sk_prot->orphan_count);
+               sk->sk_state = TCP_CLOSE;
+               sock_set_flag(sk, SOCK_DEAD);
+               inet_csk_destroy_sock(sk);
+       }
+       return ok;
 }
-EXPORT_SYMBOL_GPL(__inet_hash_nolisten);
+EXPORT_SYMBOL_GPL(inet_ehash_nolisten);
 
 void __inet_hash(struct sock *sk, struct sock *osk)
 {
        struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo;
        struct inet_listen_hashbucket *ilb;
 
-       if (sk->sk_state != TCP_LISTEN)
-               return __inet_hash_nolisten(sk, osk);
-
+       if (sk->sk_state != TCP_LISTEN) {
+               inet_ehash_nolisten(sk, osk);
+               return;
+       }
        WARN_ON(!sk_unhashed(sk));
        ilb = &hashinfo->listening_hash[inet_sk_listen_hashfn(sk)];
 
@@ -551,7 +578,7 @@ ok:
                inet_bind_hash(sk, tb, port);
                if (sk_unhashed(sk)) {
                        inet_sk(sk)->inet_sport = htons(port);
-                       __inet_hash_nolisten(sk, (struct sock *)tw);
+                       inet_ehash_nolisten(sk, (struct sock *)tw);
                }
                if (tw)
                        inet_twsk_bind_unhash(tw, hinfo);
@@ -568,7 +595,7 @@ ok:
        tb  = inet_csk(sk)->icsk_bind_hash;
        spin_lock_bh(&head->lock);
        if (sk_head(&tb->owners) == sk && !sk->sk_bind_node.next) {
-               __inet_hash_nolisten(sk, NULL);
+               inet_ehash_nolisten(sk, NULL);
                spin_unlock_bh(&head->lock);
                return 0;
        } else {