sctp: Fix the RTO-doubling on idle-link heartbeats
[safe/jmp/linux-2.6] / net / sctp / transport.c
index a596f53..5c29b14 100644 (file)
@@ -1,23 +1,23 @@
-/* SCTP kernel reference Implementation
+/* SCTP kernel implementation
  * Copyright (c) 1999-2000 Cisco, Inc.
  * Copyright (c) 1999-2001 Motorola, Inc.
  * Copyright (c) 2001-2003 International Business Machines Corp.
  * Copyright (c) 2001 Intel Corp.
  * Copyright (c) 2001 La Monte H.P. Yarroll
  *
- * This file is part of the SCTP kernel reference Implementation
+ * This file is part of the SCTP kernel implementation
  *
  * This module provides the abstraction for an SCTP tranport representing
  * a remote transport address.  For local transport addresses, we just use
  * union sctp_addr.
  *
- * The SCTP reference implementation is free software;
+ * This SCTP implementation is free software;
  * you can redistribute it and/or modify it under the terms of
  * the GNU General Public License as published by
  * the Free Software Foundation; either version 2, or (at your option)
  * any later version.
  *
- * The SCTP reference implementation is distributed in the hope that it
+ * This SCTP implementation is distributed in the hope that it
  * will be useful, but WITHOUT ANY WARRANTY; without even the implied
  *                 ************************
  * warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
@@ -74,11 +74,13 @@ static struct sctp_transport *sctp_transport_init(struct sctp_transport *peer,
         * given destination transport address, set RTO to the protocol
         * parameter 'RTO.Initial'.
         */
+       peer->last_rto = peer->rto = msecs_to_jiffies(sctp_rto_initial);
        peer->rtt = 0;
-       peer->rto = msecs_to_jiffies(sctp_rto_initial);
        peer->rttvar = 0;
        peer->srtt = 0;
        peer->rto_pending = 0;
+       peer->hb_sent = 0;
+       peer->fast_recovery = 0;
 
        peer->last_time_heard = jiffies;
        peer->last_time_used = jiffies;
@@ -99,15 +101,13 @@ static struct sctp_transport *sctp_transport_init(struct sctp_transport *peer,
        INIT_LIST_HEAD(&peer->send_ready);
        INIT_LIST_HEAD(&peer->transports);
 
-       /* Set up the retransmission timer.  */
-       init_timer(&peer->T3_rtx_timer);
-       peer->T3_rtx_timer.function = sctp_generate_t3_rtx_event;
-       peer->T3_rtx_timer.data = (unsigned long)peer;
+       peer->T3_rtx_timer.expires = 0;
+       peer->hb_timer.expires = 0;
 
-       /* Set up the heartbeat timer. */
-       init_timer(&peer->hb_timer);
-       peer->hb_timer.function = sctp_generate_heartbeat_event;
-       peer->hb_timer.data = (unsigned long)peer;
+       setup_timer(&peer->T3_rtx_timer, sctp_generate_t3_rtx_event,
+                       (unsigned long)peer);
+       setup_timer(&peer->hb_timer, sctp_generate_heartbeat_event,
+                       (unsigned long)peer);
 
        /* Initialize the 64-bit random nonce sent with heartbeat. */
        get_random_bytes(&peer->hb_nonce, sizeof(peer->hb_nonce));
@@ -195,7 +195,7 @@ static void sctp_transport_destroy(struct sctp_transport *transport)
 /* Start T3_rtx timer if it is not already running and update the heartbeat
  * timer.  This routine is called every time a DATA chunk is sent.
  */
-void sctp_transport_reset_timers(struct sctp_transport *transport)
+void sctp_transport_reset_timers(struct sctp_transport *transport, int force)
 {
        /* RFC 2960 6.3.2 Retransmission Timer Rules
         *
@@ -205,7 +205,7 @@ void sctp_transport_reset_timers(struct sctp_transport *transport)
         * address.
         */
 
-       if (!timer_pending(&transport->T3_rtx_timer))
+       if (force || !timer_pending(&transport->T3_rtx_timer))
                if (!mod_timer(&transport->T3_rtx_timer,
                               jiffies + transport->rto))
                        sctp_transport_hold(transport);
@@ -241,6 +241,45 @@ void sctp_transport_pmtu(struct sctp_transport *transport)
                transport->pathmtu = SCTP_DEFAULT_MAXSEGMENT;
 }
 
+/* this is a complete rip-off from __sk_dst_check
+ * the cookie is always 0 since this is how it's used in the
+ * pmtu code
+ */
+static struct dst_entry *sctp_transport_dst_check(struct sctp_transport *t)
+{
+       struct dst_entry *dst = t->dst;
+
+       if (dst && dst->obsolete && dst->ops->check(dst, 0) == NULL) {
+               dst_release(t->dst);
+               t->dst = NULL;
+               return NULL;
+       }
+
+       return dst;
+}
+
+void sctp_transport_update_pmtu(struct sctp_transport *t, u32 pmtu)
+{
+       struct dst_entry *dst;
+
+       if (unlikely(pmtu < SCTP_DEFAULT_MINSEGMENT)) {
+               printk(KERN_WARNING "%s: Reported pmtu %d too low, "
+                      "using default minimum of %d\n",
+                      __func__, pmtu,
+                      SCTP_DEFAULT_MINSEGMENT);
+               /* Use default minimum segment size and disable
+                * pmtu discovery on this transport.
+                */
+               t->pathmtu = SCTP_DEFAULT_MINSEGMENT;
+       } else {
+               t->pathmtu = pmtu;
+       }
+
+       dst = sctp_transport_dst_check(t);
+       if (dst)
+               dst->ops->update_pmtu(dst, pmtu);
+}
+
 /* Caches the dst entry and source address for a transport's destination
  * address.
  */
@@ -257,7 +296,7 @@ void sctp_transport_route(struct sctp_transport *transport,
        if (saddr)
                memcpy(&transport->saddr, saddr, sizeof(union sctp_addr));
        else
-               af->get_saddr(asoc, dst, daddr, &transport->saddr);
+               af->get_saddr(opt, asoc, dst, daddr, &transport->saddr);
 
        transport->dst = dst;
        if ((transport->param_flags & SPP_PMTUD_DISABLE) && transport->pathmtu) {
@@ -346,6 +385,7 @@ void sctp_transport_update_rto(struct sctp_transport *tp, __u32 rtt)
                tp->rto = tp->asoc->rto_max;
 
        tp->rtt = rtt;
+       tp->last_rto = tp->rto;
 
        /* Reset rto_pending so that a new RTT measurement is started when a
         * new data chunk is sent.
@@ -353,7 +393,7 @@ void sctp_transport_update_rto(struct sctp_transport *tp, __u32 rtt)
        tp->rto_pending = 0;
 
        SCTP_DEBUG_PRINTK("%s: transport: %p, rtt: %d, srtt: %d "
-                         "rttvar: %d, rto: %ld\n", __FUNCTION__,
+                         "rttvar: %d, rto: %ld\n", __func__,
                          tp, rtt, tp->srtt, tp->rttvar, tp->rto);
 }
 
@@ -368,11 +408,16 @@ void sctp_transport_raise_cwnd(struct sctp_transport *transport,
        cwnd = transport->cwnd;
        flight_size = transport->flight_size;
 
+       /* See if we need to exit Fast Recovery first */
+       if (transport->fast_recovery &&
+           TSN_lte(transport->fast_recovery_exit, sack_ctsn))
+               transport->fast_recovery = 0;
+
        /* The appropriate cwnd increase algorithm is performed if, and only
-        * if the cumulative TSN has advanced and the congestion window is
+        * if the cumulative TSN whould advanced and the congestion window is
         * being fully utilized.
         */
-       if ((transport->asoc->ctsn_ack_point >= sack_ctsn) ||
+       if (TSN_lte(sack_ctsn, transport->asoc->ctsn_ack_point) ||
            (flight_size < cwnd))
                return;
 
@@ -381,17 +426,23 @@ void sctp_transport_raise_cwnd(struct sctp_transport *transport,
        pmtu = transport->asoc->pathmtu;
 
        if (cwnd <= ssthresh) {
-               /* RFC 2960 7.2.1, sctpimpguide-05 2.14.2 When cwnd is less
-                * than or equal to ssthresh an SCTP endpoint MUST use the
-                * slow start algorithm to increase cwnd only if the current
-                * congestion window is being fully utilized and an incoming
-                * SACK advances the Cumulative TSN Ack Point. Only when these
-                * two conditions are met can the cwnd be increased otherwise
-                * the cwnd MUST not be increased. If these conditions are met
-                * then cwnd MUST be increased by at most the lesser of
-                * 1) the total size of the previously outstanding DATA
-                * chunk(s) acknowledged, and 2) the destination's path MTU.
+               /* RFC 4960 7.2.1
+                * o  When cwnd is less than or equal to ssthresh, an SCTP
+                *    endpoint MUST use the slow-start algorithm to increase
+                *    cwnd only if the current congestion window is being fully
+                *    utilized, an incoming SACK advances the Cumulative TSN
+                *    Ack Point, and the data sender is not in Fast Recovery.
+                *    Only when these three conditions are met can the cwnd be
+                *    increased; otherwise, the cwnd MUST not be increased.
+                *    If these conditions are met, then cwnd MUST be increased
+                *    by, at most, the lesser of 1) the total size of the
+                *    previously outstanding DATA chunk(s) acknowledged, and
+                *    2) the destination's path MTU.  This upper bound protects
+                *    against the ACK-Splitting attack outlined in [SAVAGE99].
                 */
+               if (transport->fast_recovery)
+                       return;
+
                if (bytes_acked > pmtu)
                        cwnd += pmtu;
                else
@@ -399,7 +450,7 @@ void sctp_transport_raise_cwnd(struct sctp_transport *transport,
                SCTP_DEBUG_PRINTK("%s: SLOW START: transport: %p, "
                                  "bytes_acked: %d, cwnd: %d, ssthresh: %d, "
                                  "flight_size: %d, pba: %d\n",
-                                 __FUNCTION__,
+                                 __func__,
                                  transport, bytes_acked, cwnd,
                                  ssthresh, flight_size, pba);
        } else {
@@ -425,7 +476,7 @@ void sctp_transport_raise_cwnd(struct sctp_transport *transport,
                SCTP_DEBUG_PRINTK("%s: CONGESTION AVOIDANCE: "
                                  "transport: %p, bytes_acked: %d, cwnd: %d, "
                                  "ssthresh: %d, flight_size: %d, pba: %d\n",
-                                 __FUNCTION__,
+                                 __func__,
                                  transport, bytes_acked, cwnd,
                                  ssthresh, flight_size, pba);
        }
@@ -467,6 +518,13 @@ void sctp_transport_lower_cwnd(struct sctp_transport *transport,
                 *      cwnd = ssthresh
                 *      partial_bytes_acked = 0
                 */
+               if (transport->fast_recovery)
+                       return;
+
+               /* Mark Fast recovery */
+               transport->fast_recovery = 1;
+               transport->fast_recovery_exit = transport->asoc->next_tsn - 1;
+
                transport->ssthresh = max(transport->cwnd/2,
                                          4*transport->asoc->pathmtu);
                transport->cwnd = transport->ssthresh;
@@ -507,11 +565,11 @@ void sctp_transport_lower_cwnd(struct sctp_transport *transport,
                        transport->cwnd = max(transport->cwnd/2,
                                                 4*transport->asoc->pathmtu);
                break;
-       };
+       }
 
        transport->partial_bytes_acked = 0;
        SCTP_DEBUG_PRINTK("%s: transport: %p reason: %d cwnd: "
-                         "%d ssthresh: %d\n", __FUNCTION__,
+                         "%d ssthresh: %d\n", __func__,
                          transport, reason,
                          transport->cwnd, transport->ssthresh);
 }
@@ -526,3 +584,37 @@ unsigned long sctp_transport_timeout(struct sctp_transport *t)
        timeout += jiffies;
        return timeout;
 }
+
+/* Reset transport variables to their initial values */
+void sctp_transport_reset(struct sctp_transport *t)
+{
+       struct sctp_association *asoc = t->asoc;
+
+       /* RFC 2960 (bis), Section 5.2.4
+        * All the congestion control parameters (e.g., cwnd, ssthresh)
+        * related to this peer MUST be reset to their initial values
+        * (see Section 6.2.1)
+        */
+       t->cwnd = min(4*asoc->pathmtu, max_t(__u32, 2*asoc->pathmtu, 4380));
+       t->ssthresh = asoc->peer.i.a_rwnd;
+       t->last_rto = t->rto = asoc->rto_initial;
+       t->rtt = 0;
+       t->srtt = 0;
+       t->rttvar = 0;
+
+       /* Reset these additional varibles so that we have a clean
+        * slate.
+        */
+       t->partial_bytes_acked = 0;
+       t->flight_size = 0;
+       t->error_count = 0;
+       t->rto_pending = 0;
+       t->hb_sent = 0;
+       t->fast_recovery = 0;
+
+       /* Initialize the state information for SFR-CACC */
+       t->cacc.changeover_active = 0;
+       t->cacc.cycling_changeover = 0;
+       t->cacc.next_tsn_at_change = 0;
+       t->cacc.cacc_saw_newack = 0;
+}