static void sctp_assoc_bh_rcv(struct work_struct *work);
static void sctp_assoc_free_asconf_acks(struct sctp_association *asoc);
+/* Keep track of the new idr low so that we don't re-use association id
+ * numbers too fast. It is protected by they idr spin lock is in the
+ * range of 1 - INT_MAX.
+ */
+static u32 idr_low = 1;
+
/* 1st Level Abstractions. */
/* Retrieve the SCTP per socket area. */
sp = sctp_sk((struct sock *)sk);
- /* Init all variables to a known value. */
- memset(asoc, 0, sizeof(struct sctp_association));
-
/* Discarding const is appropriate here. */
asoc->ep = (struct sctp_endpoint *)ep;
sctp_endpoint_hold(asoc->ep);
asoc->cookie_life.tv_usec = (sp->assocparams.sasoc_cookie_life % 1000)
* 1000;
asoc->frag_point = 0;
+ asoc->user_frag = sp->user_frag;
/* Set the association max_retrans and RTO values from the
* socket values.
/* Set association default SACK delay */
asoc->sackdelay = msecs_to_jiffies(sp->sackdelay);
+ asoc->sackfreq = sp->sackfreq;
/* Set the association default flags controlling
* Heartbeat, SACK delay, and Path MTU Discovery.
asoc->timeouts[SCTP_EVENT_TIMEOUT_HEARTBEAT] = 0;
asoc->timeouts[SCTP_EVENT_TIMEOUT_SACK] = asoc->sackdelay;
asoc->timeouts[SCTP_EVENT_TIMEOUT_AUTOCLOSE] =
- sp->autoclose * HZ;
+ (unsigned long)sp->autoclose * HZ;
/* Initilizes the timers */
for (i = SCTP_EVENT_TIMEOUT_NONE; i < SCTP_NUM_TIMEOUT_TYPES; ++i)
asoc->a_rwnd = asoc->rwnd;
asoc->rwnd_over = 0;
+ asoc->rwnd_press = 0;
/* Use my own max window until I learn something better. */
asoc->peer.rwnd = SCTP_DEFAULT_MAXWINDOW;
* already received one packet.]
*/
asoc->peer.sack_needed = 1;
+ asoc->peer.sack_cnt = 0;
/* Assume that the peer will tell us if he recognizes ASCONF
* as part of INIT exchange.
if (!sctp_ulpq_init(&asoc->ulpq, asoc))
goto fail_init;
- /* Set up the tsn tracking. */
- sctp_tsnmap_init(&asoc->peer.tsn_map, SCTP_TSN_MAP_SIZE, 0);
+ memset(&asoc->peer.tsn_map, 0, sizeof(struct sctp_tsnmap));
asoc->need_ecne = 0;
* told otherwise.
*/
asoc->peer.ipv4_address = 1;
- asoc->peer.ipv6_address = 1;
+ if (asoc->base.sk->sk_family == PF_INET6)
+ asoc->peer.ipv6_address = 1;
INIT_LIST_HEAD(&asoc->asocs);
asoc->autoclose = sp->autoclose;
/* Dispose of any pending chunks on the inqueue. */
sctp_inq_free(&asoc->base.inqueue);
+ sctp_tsnmap_free(&asoc->peer.tsn_map);
+
/* Free ssnmap storage. */
sctp_ssnmap_free(asoc->ssnmap);
spin_unlock_bh(&sctp_assocs_id_lock);
}
- BUG_TRAP(!atomic_read(&asoc->rmem_alloc));
+ WARN_ON(atomic_read(&asoc->rmem_alloc));
if (asoc->base.malloced) {
kfree(asoc);
void sctp_assoc_set_primary(struct sctp_association *asoc,
struct sctp_transport *transport)
{
+ int changeover = 0;
+
+ /* it's a changeover only if we already have a primary path
+ * that we are changing
+ */
+ if (asoc->peer.primary_path != NULL &&
+ asoc->peer.primary_path != transport)
+ changeover = 1 ;
+
asoc->peer.primary_path = transport;
/* Set a default msg_name for events. */
* to this destination address earlier. The sender MUST set
* CYCLING_CHANGEOVER to indicate that this switch is a
* double switch to the same destination address.
+ *
+ * Really, only bother is we have data queued or outstanding on
+ * the association.
*/
+ if (!asoc->outqueue.outstanding_bytes && !asoc->outqueue.out_qlen)
+ return;
+
if (transport->cacc.changeover_active)
- transport->cacc.cycling_changeover = 1;
+ transport->cacc.cycling_changeover = changeover;
/* 2) The sender MUST set CHANGEOVER_ACTIVE to indicate that
* a changeover has occurred.
*/
- transport->cacc.changeover_active = 1;
+ transport->cacc.changeover_active = changeover;
/* 3) The sender MUST store the next TSN to be sent in
* next_tsn_at_change.
if (asoc->init_last_sent_to == peer)
asoc->init_last_sent_to = NULL;
+ /* If we remove the transport an SHUTDOWN was last sent to, set it
+ * to NULL. Combined with the update of the retran path above, this
+ * will cause the next SHUTDOWN to be sent to the next available
+ * transport, maintaining the cycle.
+ */
+ if (asoc->shutdown_last_sent_to == peer)
+ asoc->shutdown_last_sent_to = NULL;
+
+ /* If we remove the transport an ASCONF was last sent to, set it to
+ * NULL.
+ */
+ if (asoc->addip_last_asconf &&
+ asoc->addip_last_asconf->transport == peer)
+ asoc->addip_last_asconf->transport = NULL;
+
+ /* If we have something on the transmitted list, we have to
+ * save it off. The best place is the active path.
+ */
+ if (!list_empty(&peer->transmitted)) {
+ struct sctp_transport *active = asoc->peer.active_path;
+ struct sctp_chunk *ch;
+
+ /* Reset the transport of each chunk on this list */
+ list_for_each_entry(ch, &peer->transmitted,
+ transmitted_list) {
+ ch->transport = NULL;
+ ch->rtt_in_progress = 0;
+ }
+
+ list_splice_tail_init(&peer->transmitted,
+ &active->transmitted);
+
+ /* Start a T3 timer here in case it wasn't running so
+ * that these migrated packets have a chance to get
+ * retrnasmitted.
+ */
+ if (!timer_pending(&active->T3_rtx_timer))
+ if (!mod_timer(&active->T3_rtx_timer,
+ jiffies + active->rto))
+ sctp_transport_hold(active);
+ }
+
asoc->peer.transport_count--;
sctp_transport_free(peer);
/* Check to see if this is a duplicate. */
peer = sctp_assoc_lookup_paddr(asoc, addr);
if (peer) {
+ /* An UNKNOWN state is only set on transports added by
+ * user in sctp_connectx() call. Such transports should be
+ * considered CONFIRMED per RFC 4960, Section 5.4.
+ */
if (peer->state == SCTP_UNKNOWN) {
- if (peer_state == SCTP_ACTIVE)
- peer->state = SCTP_ACTIVE;
- if (peer_state == SCTP_UNCONFIRMED)
- peer->state = SCTP_UNCONFIRMED;
+ peer->state = SCTP_ACTIVE;
}
return peer;
}
* association configured value.
*/
peer->sackdelay = asoc->sackdelay;
+ peer->sackfreq = asoc->sackfreq;
/* Enable/disable heartbeat, SACK delay, and path MTU discovery
* based on association setting.
*/
peer->param_flags = asoc->param_flags;
+ sctp_transport_route(peer, NULL, sp);
+
/* Initialize the pmtu of the transport. */
- if (peer->param_flags & SPP_PMTUD_ENABLE)
- sctp_transport_pmtu(peer);
- else if (asoc->pathmtu)
- peer->pathmtu = asoc->pathmtu;
- else
- peer->pathmtu = SCTP_DEFAULT_MAXSEGMENT;
+ if (peer->param_flags & SPP_PMTUD_DISABLE) {
+ if (asoc->pathmtu)
+ peer->pathmtu = asoc->pathmtu;
+ else
+ peer->pathmtu = SCTP_DEFAULT_MAXSEGMENT;
+ }
/* If this is the first transport addr on this association,
* initialize the association PMTU to the peer's PMTU.
SCTP_DEBUG_PRINTK("sctp_assoc_add_peer:association %p PMTU set to "
"%d\n", asoc, asoc->pathmtu);
+ peer->pmtu_pending = 0;
- asoc->frag_point = sctp_frag_point(sp, asoc->pathmtu);
+ asoc->frag_point = sctp_frag_point(asoc, asoc->pathmtu);
/* The asoc->peer.port might not be meaningful yet, but
* initialize the packet structure anyway.
peer->partial_bytes_acked = 0;
peer->flight_size = 0;
+ peer->burst_limited = 0;
/* Set the transport's RTO.initial value */
peer->rto = asoc->rto_initial;
asoc->peer.retran_path = peer;
}
- if (asoc->peer.active_path == asoc->peer.retran_path) {
+ if (asoc->peer.active_path == asoc->peer.retran_path &&
+ peer->state != SCTP_UNCONFIRMED) {
asoc->peer.retran_path = peer;
}
if (t != primary)
sctp_assoc_rm_peer(asoc, t);
}
-
- return;
}
/* Engage in transport control operations.
break;
case SCTP_TRANSPORT_DOWN:
- /* if the transort was never confirmed, do not transition it
- * to inactive state.
+ /* If the transport was never confirmed, do not transition it
+ * to inactive state. Also, release the cached route since
+ * there may be a better route next time.
*/
if (transport->state != SCTP_UNCONFIRMED)
transport->state = SCTP_INACTIVE;
+ else {
+ dst_release(transport->dst);
+ transport->dst = NULL;
+ }
spc_state = SCTP_ADDR_UNREACHABLE;
break;
asoc->peer.rwnd = new->peer.rwnd;
asoc->peer.sack_needed = new->peer.sack_needed;
asoc->peer.i = new->peer.i;
- sctp_tsnmap_init(&asoc->peer.tsn_map, SCTP_TSN_MAP_SIZE,
- asoc->peer.i.initial_tsn);
+ sctp_tsnmap_init(&asoc->peer.tsn_map, SCTP_TSN_MAP_INITIAL,
+ asoc->peer.i.initial_tsn, GFP_ATOMIC);
/* Remove any peer addresses not present in the new association. */
list_for_each_safe(pos, temp, &asoc->peer.transport_addr_list) {
trans = list_entry(pos, struct sctp_transport, transports);
- if (!sctp_assoc_lookup_paddr(new, &trans->ipaddr))
- sctp_assoc_del_peer(asoc, &trans->ipaddr);
+ if (!sctp_assoc_lookup_paddr(new, &trans->ipaddr)) {
+ sctp_assoc_rm_peer(asoc, trans);
+ continue;
+ }
if (asoc->state >= SCTP_STATE_ESTABLISHED)
sctp_transport_reset(trans);
struct list_head *head = &asoc->peer.transport_addr_list;
struct list_head *pos;
+ if (asoc->peer.transport_count == 1)
+ return;
+
/* Find the next transport in a round-robin fashion. */
t = asoc->peer.retran_path;
pos = &t->transports;
t = list_entry(pos, struct sctp_transport, transports);
+ /* We have exhausted the list, but didn't find any
+ * other active transports. If so, use the next
+ * transport.
+ */
+ if (t == asoc->peer.retran_path) {
+ t = next;
+ break;
+ }
+
/* Try to find an active transport. */
if ((t->state == SCTP_ACTIVE) ||
/* Keep track of the next transport in case
* we don't find any active transport.
*/
- if (!next)
+ if (t->state != SCTP_UNCONFIRMED && !next)
next = t;
}
-
- /* We have exhausted the list, but didn't find any
- * other active transports. If so, use the next
- * transport.
- */
- if (t == asoc->peer.retran_path) {
- t = next;
- break;
- }
}
- asoc->peer.retran_path = t;
+ if (t)
+ asoc->peer.retran_path = t;
SCTP_DEBUG_PRINTK_IPADDR("sctp_assoc_update_retran_path:association"
" %p addr: ",
ntohs(t->ipaddr.v4.sin_port));
}
-/* Choose the transport for sending a INIT packet. */
-struct sctp_transport *sctp_assoc_choose_init_transport(
- struct sctp_association *asoc)
+/* Choose the transport for sending retransmit packet. */
+struct sctp_transport *sctp_assoc_choose_alter_transport(
+ struct sctp_association *asoc, struct sctp_transport *last_sent_to)
{
- struct sctp_transport *t;
-
- /* Use the retran path. If the last INIT was sent over the
+ /* If this is the first time packet is sent, use the active path,
+ * else use the retran path. If the last packet was sent over the
* retran path, update the retran path and use it.
*/
- if (!asoc->init_last_sent_to) {
- t = asoc->peer.active_path;
- } else {
- if (asoc->init_last_sent_to == asoc->peer.retran_path)
- sctp_assoc_update_retran_path(asoc);
- t = asoc->peer.retran_path;
- }
-
- SCTP_DEBUG_PRINTK_IPADDR("sctp_assoc_update_retran_path:association"
- " %p addr: ",
- " port: %d\n",
- asoc,
- (&t->ipaddr),
- ntohs(t->ipaddr.v4.sin_port));
-
- return t;
-}
-
-/* Choose the transport for sending a SHUTDOWN packet. */
-struct sctp_transport *sctp_assoc_choose_shutdown_transport(
- struct sctp_association *asoc)
-{
- /* If this is the first time SHUTDOWN is sent, use the active path,
- * else use the retran path. If the last SHUTDOWN was sent over the
- * retran path, update the retran path and use it.
- */
- if (!asoc->shutdown_last_sent_to)
+ if (!last_sent_to)
return asoc->peer.active_path;
else {
- if (asoc->shutdown_last_sent_to == asoc->peer.retran_path)
+ if (last_sent_to == asoc->peer.retran_path)
sctp_assoc_update_retran_path(asoc);
return asoc->peer.retran_path;
}
-
}
/* Update the association's pmtu and frag_point by going through all the
}
if (pmtu) {
- struct sctp_sock *sp = sctp_sk(asoc->base.sk);
asoc->pathmtu = pmtu;
- asoc->frag_point = sctp_frag_point(sp, pmtu);
+ asoc->frag_point = sctp_frag_point(asoc, pmtu);
}
SCTP_DEBUG_PRINTK("%s: asoc:%p, pmtu:%d, frag_point:%d\n",
case SCTP_STATE_SHUTDOWN_RECEIVED:
case SCTP_STATE_SHUTDOWN_SENT:
if ((asoc->rwnd > asoc->a_rwnd) &&
- ((asoc->rwnd - asoc->a_rwnd) >=
- min_t(__u32, (asoc->base.sk->sk_rcvbuf >> 1), asoc->pathmtu)))
+ ((asoc->rwnd - asoc->a_rwnd) >= max_t(__u32,
+ (asoc->base.sk->sk_rcvbuf >> sctp_rwnd_upd_shift),
+ asoc->pathmtu)))
return 1;
break;
default:
asoc->rwnd += len;
}
+ /* If we had window pressure, start recovering it
+ * once our rwnd had reached the accumulated pressure
+ * threshold. The idea is to recover slowly, but up
+ * to the initial advertised window.
+ */
+ if (asoc->rwnd_press && asoc->rwnd >= asoc->rwnd_press) {
+ int change = min(asoc->pathmtu, asoc->rwnd_press);
+ asoc->rwnd += change;
+ asoc->rwnd_press -= change;
+ }
+
SCTP_DEBUG_PRINTK("%s: asoc %p rwnd increased by %d to (%u, %u) "
"- %u\n", __func__, asoc, len, asoc->rwnd,
asoc->rwnd_over, asoc->a_rwnd);
/* Decrease asoc's rwnd by len. */
void sctp_assoc_rwnd_decrease(struct sctp_association *asoc, unsigned len)
{
+ int rx_count;
+ int over = 0;
+
SCTP_ASSERT(asoc->rwnd, "rwnd zero", return);
SCTP_ASSERT(!asoc->rwnd_over, "rwnd_over not zero", return);
+
+ if (asoc->ep->rcvbuf_policy)
+ rx_count = atomic_read(&asoc->rmem_alloc);
+ else
+ rx_count = atomic_read(&asoc->base.sk->sk_rmem_alloc);
+
+ /* If we've reached or overflowed our receive buffer, announce
+ * a 0 rwnd if rwnd would still be positive. Store the
+ * the pottential pressure overflow so that the window can be restored
+ * back to original value.
+ */
+ if (rx_count >= asoc->base.sk->sk_rcvbuf)
+ over = 1;
+
if (asoc->rwnd >= len) {
asoc->rwnd -= len;
+ if (over) {
+ asoc->rwnd_press += asoc->rwnd;
+ asoc->rwnd = 0;
+ }
} else {
asoc->rwnd_over = len - asoc->rwnd;
asoc->rwnd = 0;
}
- SCTP_DEBUG_PRINTK("%s: asoc %p rwnd decreased by %d to (%u, %u)\n",
+ SCTP_DEBUG_PRINTK("%s: asoc %p rwnd decreased by %d to (%u, %u, %u)\n",
__func__, asoc, len, asoc->rwnd,
- asoc->rwnd_over);
+ asoc->rwnd_over, asoc->rwnd_press);
}
/* Build the bind address list for the association based on info from the
* local endpoint and the remote peer.
*/
int sctp_assoc_set_bind_addr_from_ep(struct sctp_association *asoc,
- gfp_t gfp)
+ sctp_scope_t scope, gfp_t gfp)
{
- sctp_scope_t scope;
int flags;
/* Use scoping rules to determine the subset of addresses from
* the endpoint.
*/
- scope = sctp_scope(&asoc->peer.active_path->ipaddr);
flags = (PF_INET6 == asoc->base.sk->sk_family) ? SCTP_ADDR6_ALLOWED : 0;
if (asoc->peer.ipv4_address)
flags |= SCTP_ADDR4_PEERSUPP;
{
int assoc_id;
int error = 0;
+
+ /* If the id is already assigned, keep it. */
+ if (asoc->assoc_id)
+ return error;
retry:
if (unlikely(!idr_pre_get(&sctp_assocs_id, gfp)))
return -ENOMEM;
spin_lock_bh(&sctp_assocs_id_lock);
error = idr_get_new_above(&sctp_assocs_id, (void *)asoc,
- 1, &assoc_id);
+ idr_low, &assoc_id);
+ if (!error) {
+ idr_low = assoc_id + 1;
+ if (idr_low == INT_MAX)
+ idr_low = 1;
+ }
spin_unlock_bh(&sctp_assocs_id_lock);
if (error == -EAGAIN)
goto retry;