8 files changed, 47 insertions, 87 deletions

diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index 1fbff5fa4241..1aa2dc9e380e 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -1117,6 +1117,7 @@ int inet_sk_rebuild_header(struct sock *sk)
                         },
                 },
                 .proto = sk->sk_protocol,
+                .flags = inet_sk_flowi_flags(sk),
                 .uli_u = {
                         .ports = {
                                 .sport = inet->sport,
diff --git a/net/ipv4/ip_input.c b/net/ipv4/ip_input.c
index 861978a4f1a8..cfb38ac9d698 100644
--- a/net/ipv4/ip_input.c
+++ b/net/ipv4/ip_input.c
@@ -209,9 +209,17 @@ static int ip_local_deliver_finish(struct sk_buff *skb)
 
                 hash = protocol & (MAX_INET_PROTOS - 1);
                 ipprot = rcu_dereference(inet_protos[hash]);
-                if (ipprot != NULL && (net == &init_net || ipprot->netns_ok)) {
+                if (ipprot != NULL) {
                         int ret;
 
+                        if (!net_eq(net, &init_net) && !ipprot->netns_ok) {
+                                if (net_ratelimit())
+                                        printk("%s: proto %d isn't netns-ready\n",
+                                                __func__, protocol);
+                                kfree_skb(skb);
+                                goto out;
+                        }
+
                         if (!ipprot->no_policy) {
                                 if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) {
                                         kfree_skb(skb);
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index b42e082cc170..25924b1eb2ef 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -1945,13 +1945,14 @@ int __init ip_mr_init(void)
                 goto proc_cache_fail;
 #endif
         return 0;
-reg_notif_fail:
-        kmem_cache_destroy(mrt_cachep);
 #ifdef CONFIG_PROC_FS
-proc_vif_fail:
-        unregister_netdevice_notifier(&ip_mr_notifier);
 proc_cache_fail:
         proc_net_remove(&init_net, "ip_mr_vif");
+proc_vif_fail:
+        unregister_netdevice_notifier(&ip_mr_notifier);
 #endif
+reg_notif_fail:
+        del_timer(&ipmr_expire_timer);
+        kmem_cache_destroy(mrt_cachep);
         return err;
 }
diff --git a/net/ipv4/netfilter/nf_nat_rule.c b/net/ipv4/netfilter/nf_nat_rule.c
index bea54a685109..8d489e746b21 100644
--- a/net/ipv4/netfilter/nf_nat_rule.c
+++ b/net/ipv4/netfilter/nf_nat_rule.c
@@ -61,7 +61,7 @@ static struct
 static struct xt_table nat_table = {
         .name           = "nat",
         .valid_hooks    = NAT_VALID_HOOKS,
-        .lock           = __RW_LOCK_UNLOCKED(__nat_table.lock),
+        .lock           = __RW_LOCK_UNLOCKED(nat_table.lock),
         .me             = THIS_MODULE,
         .af             = AF_INET,
 };
diff --git a/net/ipv4/tcp_htcp.c b/net/ipv4/tcp_htcp.c
index af99776146ff..937549b8a921 100644
--- a/net/ipv4/tcp_htcp.c
+++ b/net/ipv4/tcp_htcp.c
@@ -69,9 +69,12 @@ static u32 htcp_cwnd_undo(struct sock *sk)
         const struct tcp_sock *tp = tcp_sk(sk);
         struct htcp *ca = inet_csk_ca(sk);
 
-        ca->last_cong = ca->undo_last_cong;
-        ca->maxRTT = ca->undo_maxRTT;
-        ca->old_maxB = ca->undo_old_maxB;
+        if (ca->undo_last_cong) {
+                ca->last_cong = ca->undo_last_cong;
+                ca->maxRTT = ca->undo_maxRTT;
+                ca->old_maxB = ca->undo_old_maxB;
+                ca->undo_last_cong = 0;
+        }
 
         return max(tp->snd_cwnd, (tp->snd_ssthresh << 7) / ca->beta);
 }
@@ -268,7 +271,10 @@ static void htcp_state(struct sock *sk, u8 new_state)
         case TCP_CA_Open:
                 {
                         struct htcp *ca = inet_csk_ca(sk);
-                        ca->last_cong = jiffies;
+                        if (ca->undo_last_cong) {
+                                ca->last_cong = jiffies;
+                                ca->undo_last_cong = 0;
+                        }
                 }
                 break;
         case TCP_CA_CWR:
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index ba85d8831893..fe3b4bdfd251 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -1028,10 +1028,6 @@ unsigned int tcp_sync_mss(struct sock *sk, u32 pmtu)
 
 /* Compute the current effective MSS, taking SACKs and IP options,
  * and even PMTU discovery events into account.
- *
- * LARGESEND note: !tcp_urg_mode is overkill, only frames up to snd_up
- * cannot be large. However, taking into account rare use of URG, this
- * is not a big flaw.
  */
 unsigned int tcp_current_mss(struct sock *sk, int large_allowed)
 {
@@ -1046,7 +1042,7 @@ unsigned int tcp_current_mss(struct sock *sk, int large_allowed)
 
         mss_now = tp->mss_cache;
 
-        if (large_allowed && sk_can_gso(sk) && !tcp_urg_mode(tp))
+        if (large_allowed && sk_can_gso(sk))
                 doing_tso = 1;
 
         if (dst) {
@@ -1516,6 +1512,10 @@ static int tcp_mtu_probe(struct sock *sk)
  * send_head.  This happens as incoming acks open up the remote
  * window for us.
  *
+ * LARGESEND note: !tcp_urg_mode is overkill, only frames between
+ * snd_up-64k-mss .. snd_up cannot be large. However, taking into
+ * account rare use of URG, this is not a big flaw.
+ *
  * Returns 1, if no segments are in flight and we have queued segments, but
  * cannot send anything now because of SWS or another problem.
  */
@@ -1567,7 +1567,7 @@ static int tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle)
                 }
 
                 limit = mss_now;
-                if (tso_segs > 1)
+                if (tso_segs > 1 && !tcp_urg_mode(tp))
                         limit = tcp_mss_split_point(sk, skb, mss_now,
                                                     cwnd_quota);
 
@@ -1616,6 +1616,7 @@ void __tcp_push_pending_frames(struct sock *sk, unsigned int cur_mss,
  */
 void tcp_push_one(struct sock *sk, unsigned int mss_now)
 {
+        struct tcp_sock *tp = tcp_sk(sk);
         struct sk_buff *skb = tcp_send_head(sk);
         unsigned int tso_segs, cwnd_quota;
 
@@ -1630,7 +1631,7 @@ void tcp_push_one(struct sock *sk, unsigned int mss_now)
                 BUG_ON(!tso_segs);
 
                 limit = mss_now;
-                if (tso_segs > 1)
+                if (tso_segs > 1 && !tcp_urg_mode(tp))
                         limit = tcp_mss_split_point(sk, skb, mss_now,
                                                     cwnd_quota);
 
diff --git a/net/ipv4/tcp_vegas.c b/net/ipv4/tcp_vegas.c
index 14504dada116..a453aac91bd3 100644
--- a/net/ipv4/tcp_vegas.c
+++ b/net/ipv4/tcp_vegas.c
@@ -40,18 +40,14 @@
 
 #include "tcp_vegas.h"
 
-/* Default values of the Vegas variables, in fixed-point representation
- * with V_PARAM_SHIFT bits to the right of the binary point.
- */
-#define V_PARAM_SHIFT 1
-static int alpha = 2<<V_PARAM_SHIFT;
-static int beta  = 4<<V_PARAM_SHIFT;
-static int gamma = 1<<V_PARAM_SHIFT;
+static int alpha = 2;
+static int beta  = 4;
+static int gamma = 1;
 
 module_param(alpha, int, 0644);
-MODULE_PARM_DESC(alpha, "lower bound of packets in network (scale by 2)");
+MODULE_PARM_DESC(alpha, "lower bound of packets in network");
 module_param(beta, int, 0644);
-MODULE_PARM_DESC(beta, "upper bound of packets in network (scale by 2)");
+MODULE_PARM_DESC(beta, "upper bound of packets in network");
 module_param(gamma, int, 0644);
 MODULE_PARM_DESC(gamma, "limit on increase (scale by 2)");
 
@@ -172,49 +168,13 @@ static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 in_flight)
                 return;
         }
 
-        /* The key players are v_beg_snd_una and v_beg_snd_nxt.
-         *
-         * These are so named because they represent the approximate values
-         * of snd_una and snd_nxt at the beginning of the current RTT. More
-         * precisely, they represent the amount of data sent during the RTT.
-         * At the end of the RTT, when we receive an ACK for v_beg_snd_nxt,
-         * we will calculate that (v_beg_snd_nxt - v_beg_snd_una) outstanding
-         * bytes of data have been ACKed during the course of the RTT, giving
-         * an "actual" rate of:
-         *
-         *     (v_beg_snd_nxt - v_beg_snd_una) / (rtt duration)
-         *
-         * Unfortunately, v_beg_snd_una is not exactly equal to snd_una,
-         * because delayed ACKs can cover more than one segment, so they
-         * don't line up nicely with the boundaries of RTTs.
-         *
-         * Another unfortunate fact of life is that delayed ACKs delay the
-         * advance of the left edge of our send window, so that the number
-         * of bytes we send in an RTT is often less than our cwnd will allow.
-         * So we keep track of our cwnd separately, in v_beg_snd_cwnd.
-         */
-
         if (after(ack, vegas->beg_snd_nxt)) {
                 /* Do the Vegas once-per-RTT cwnd adjustment. */
-                u32 old_wnd, old_snd_cwnd;
-
-
-                /* Here old_wnd is essentially the window of data that was
-                 * sent during the previous RTT, and has all
-                 * been acknowledged in the course of the RTT that ended
-                 * with the ACK we just received. Likewise, old_snd_cwnd
-                 * is the cwnd during the previous RTT.
-                 */
-                old_wnd = (vegas->beg_snd_nxt - vegas->beg_snd_una) /
-                        tp->mss_cache;
-                old_snd_cwnd = vegas->beg_snd_cwnd;
 
                 /* Save the extent of the current window so we can use this
                  * at the end of the next RTT.
                  */
-                vegas->beg_snd_una  = vegas->beg_snd_nxt;
                 vegas->beg_snd_nxt  = tp->snd_nxt;
-                vegas->beg_snd_cwnd = tp->snd_cwnd;
 
                 /* We do the Vegas calculations only if we got enough RTT
                  * samples that we can be reasonably sure that we got
@@ -252,22 +212,14 @@ static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 in_flight)
                          *
                          * This is:
                          *     (actual rate in segments) * baseRTT
-                         * We keep it as a fixed point number with
-                         * V_PARAM_SHIFT bits to the right of the binary point.
                          */
-                        target_cwnd = ((u64)old_wnd * vegas->baseRTT);
-                        target_cwnd <<= V_PARAM_SHIFT;
-                        do_div(target_cwnd, rtt);
+                        target_cwnd = tp->snd_cwnd * vegas->baseRTT / rtt;
 
                         /* Calculate the difference between the window we had,
                          * and the window we would like to have. This quantity
                          * is the "Diff" from the Arizona Vegas papers.
-                         *
-                         * Again, this is a fixed point number with
-                         * V_PARAM_SHIFT bits to the right of the binary
-                         * point.
                          */
-                        diff = (old_wnd << V_PARAM_SHIFT) - target_cwnd;
+                        diff = tp->snd_cwnd * (rtt-vegas->baseRTT) / vegas->baseRTT;
 
                         if (diff > gamma && tp->snd_ssthresh > 2 ) {
                                 /* Going too fast. Time to slow down
@@ -282,16 +234,13 @@ static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 in_flight)
                                  * truncation robs us of full link
                                  * utilization.
                                  */
-                                tp->snd_cwnd = min(tp->snd_cwnd,
-                                                   ((u32)target_cwnd >>
-                                                    V_PARAM_SHIFT)+1);
+                                tp->snd_cwnd = min(tp->snd_cwnd, (u32)target_cwnd+1);
 
                         } else if (tp->snd_cwnd <= tp->snd_ssthresh) {
                                 /* Slow start.  */
                                 tcp_slow_start(tp);
                         } else {
                                 /* Congestion avoidance. */
-                                u32 next_snd_cwnd;
 
                                 /* Figure out where we would like cwnd
                                  * to be.
@@ -300,32 +249,25 @@ static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 in_flight)
                                         /* The old window was too fast, so
                                          * we slow down.
                                          */
-                                        next_snd_cwnd = old_snd_cwnd - 1;
+                                        tp->snd_cwnd--;
                                 } else if (diff < alpha) {
                                         /* We don't have enough extra packets
                                          * in the network, so speed up.
                                          */
-                                        next_snd_cwnd = old_snd_cwnd + 1;
+                                        tp->snd_cwnd++;
                                 } else {
                                         /* Sending just as fast as we
                                          * should be.
                                          */
-                                        next_snd_cwnd = old_snd_cwnd;
                                 }
-
-                                /* Adjust cwnd upward or downward, toward the
-                                 * desired value.
-                                 */
-                                if (next_snd_cwnd > tp->snd_cwnd)
-                                        tp->snd_cwnd++;
-                                else if (next_snd_cwnd < tp->snd_cwnd)
-                                        tp->snd_cwnd--;
                         }
 
                         if (tp->snd_cwnd < 2)
                                 tp->snd_cwnd = 2;
                         else if (tp->snd_cwnd > tp->snd_cwnd_clamp)
                                 tp->snd_cwnd = tp->snd_cwnd_clamp;
+
+                        tp->snd_ssthresh = tcp_current_ssthresh(sk);
                 }
 
                 /* Wipe the slate clean for the next RTT. */
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index cf02701ced48..98c1fd09be88 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -633,6 +633,7 @@ int udp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
                                                 .saddr = saddr,
                                                 .tos = tos } },
                                     .proto = sk->sk_protocol,
+                                    .flags = inet_sk_flowi_flags(sk),
                                     .uli_u = { .ports =
                                                { .sport = inet->sport,
                                                  .dport = dport } } };