Pull ngam-maule-steiner into release branch

[linux-2.6] / net / ipv4 / tcp_ipv4.c

diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 1745dc8d25e66fca6f305b482aaf0cf550f234fd..67c670886c1fda8f3f8beb156e4b23b36888eae1 100644 (file)
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -242,9 +242,14 @@ static int tcp_v4_get_port(struct sock *sk, unsigned short snum)
                tcp_port_rover = rover;
                spin_unlock(&tcp_portalloc_lock);
 
-               /* Exhausted local port range during search? */
+               /* Exhausted local port range during search?  It is not
+                * possible for us to be holding one of the bind hash
+                * locks if this test triggers, because if 'remaining'
+                * drops to zero, we broke out of the do/while loop at
+                * the top level, not from the 'break;' statement.
+                */
                ret = 1;
-               if (remaining <= 0)
+               if (unlikely(remaining <= 0))
                        goto fail;
 
                /* OK, here is the one we will use.  HEAD is
@@ -874,7 +879,7 @@ static struct request_sock *tcp_v4_search_req(struct tcp_sock *tp,
                                              __u16 rport,
                                              __u32 raddr, __u32 laddr)
 {
-       struct tcp_listen_opt *lopt = tp->accept_queue.listen_opt;
+       struct listen_sock *lopt = tp->accept_queue.listen_opt;
        struct request_sock *req, **prev;
 
        for (prev = &lopt->syn_table[tcp_v4_synq_hash(raddr, rport, lopt->hash_rnd)];
@@ -898,7 +903,7 @@ static struct request_sock *tcp_v4_search_req(struct tcp_sock *tp,
 static void tcp_v4_synq_add(struct sock *sk, struct request_sock *req)
 {
        struct tcp_sock *tp = tcp_sk(sk);
-       struct tcp_listen_opt *lopt = tp->accept_queue.listen_opt;
+       struct listen_sock *lopt = tp->accept_queue.listen_opt;
        u32 h = tcp_v4_synq_hash(inet_rsk(req)->rmt_addr, inet_rsk(req)->rmt_port, lopt->hash_rnd);
 
        reqsk_queue_hash_req(&tp->accept_queue, h, req, TCP_TIMEOUT_INIT);
@@ -1366,21 +1371,6 @@ static inline struct ip_options *tcp_v4_save_options(struct sock *sk,
        return dopt;
 }
 
-/*
- * Maximum number of SYN_RECV sockets in queue per LISTEN socket.
- * One SYN_RECV socket costs about 80bytes on a 32bit machine.
- * It would be better to replace it with a global counter for all sockets
- * but then some measure against one socket starving all other sockets
- * would be needed.
- *
- * It was 128 by default. Experiments with real servers show, that
- * it is absolutely not enough even at 100conn/sec. 256 cures most
- * of problems. This value is adjusted to 128 for very small machines
- * (<=32Mb of memory) and to 1024 on normal or better ones (>=256Mb).
- * Further increasing requires to change hash table size.
- */
-int sysctl_max_syn_backlog = 256;
-
 struct request_sock_ops tcp_request_sock_ops = {
        .family         =       PF_INET,
        .obj_size       =       sizeof(struct tcp_request_sock),
@@ -1509,12 +1499,11 @@ int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb)
                         * to destinations, already remembered
                         * to the moment of synflood.
                         */
-                       NETDEBUG(if (net_ratelimit()) \
-                                       printk(KERN_DEBUG "TCP: drop open "
-                                                         "request from %u.%u."
-                                                         "%u.%u/%u\n", \
-                                              NIPQUAD(saddr),
-                                              ntohs(skb->h.th->source)));
+                       LIMIT_NETDEBUG(printk(KERN_DEBUG "TCP: drop open "
+                                             "request from %u.%u."
+                                             "%u.%u/%u\n",
+                                             NIPQUAD(saddr),
+                                             ntohs(skb->h.th->source)));
                        dst_release(dst);
                        goto drop_and_free;
                }
@@ -1642,8 +1631,7 @@ static int tcp_v4_checksum_init(struct sk_buff *skb)
                                  skb->nh.iph->daddr, skb->csum))
                        return 0;
 
-               NETDEBUG(if (net_ratelimit())
-                               printk(KERN_DEBUG "hw tcp v4 csum failed\n"));
+               LIMIT_NETDEBUG(printk(KERN_DEBUG "hw tcp v4 csum failed\n"));
                skb->ip_summed = CHECKSUM_NONE;
        }
        if (skb->len <= 76) {
@@ -2060,9 +2048,10 @@ static int tcp_v4_init_sock(struct sock *sk)
         */
        tp->snd_ssthresh = 0x7fffffff;  /* Infinity */
        tp->snd_cwnd_clamp = ~0;
-       tp->mss_cache_std = tp->mss_cache = 536;
+       tp->mss_cache = 536;
 
        tp->reordering = sysctl_tcp_reordering;
+       tp->ca_ops = &tcp_init_congestion_ops;
 
        sk->sk_state = TCP_CLOSE;
 
@@ -2085,6 +2074,8 @@ int tcp_v4_destroy_sock(struct sock *sk)
 
        tcp_clear_xmit_timers(sk);
 
+       tcp_cleanup_congestion_control(tp);
+
        /* Cleanup up the write buffer. */
        sk_stream_writequeue_purge(sk);
 
@@ -2662,7 +2653,6 @@ EXPORT_SYMBOL(tcp_proc_register);
 EXPORT_SYMBOL(tcp_proc_unregister);
 #endif
 EXPORT_SYMBOL(sysctl_local_port_range);
-EXPORT_SYMBOL(sysctl_max_syn_backlog);
 EXPORT_SYMBOL(sysctl_tcp_low_latency);
 EXPORT_SYMBOL(sysctl_tcp_tw_reuse);