diff options
Diffstat (limited to 'net/sctp/associola.c')
| -rw-r--r-- | net/sctp/associola.c | 1134 |
1 files changed, 771 insertions, 363 deletions
diff --git a/net/sctp/associola.c b/net/sctp/associola.c index 12b0f582a66..06a9ee6b2d3 100644 --- a/net/sctp/associola.c +++ b/net/sctp/associola.c @@ -1,37 +1,33 @@ -/* SCTP kernel reference Implementation +/* SCTP kernel implementation * (C) Copyright IBM Corp. 2001, 2004 * Copyright (c) 1999-2000 Cisco, Inc. * Copyright (c) 1999-2001 Motorola, Inc. * 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 association. * - * 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. * See the GNU General Public License for more details. * * You should have received a copy of the GNU General Public License - * along with GNU CC; see the file COPYING. If not, write to - * the Free Software Foundation, 59 Temple Place - Suite 330, - * Boston, MA 02111-1307, USA. + * along with GNU CC; see the file COPYING. If not, see + * <http://www.gnu.org/licenses/>. * * Please send any bug reports or fixes you make to the * email address(es): - * lksctp developers <lksctp-developers@lists.sourceforge.net> - * - * Or submit a bug report through the following website: - * http://www.sf.net/projects/lksctp + * lksctp developers <linux-sctp@vger.kernel.org> * * Written or modified by: * La Monte H.P. Yarroll <piggy@acm.org> @@ -43,16 +39,14 @@ * Daisy Chang <daisyc@us.ibm.com> * Ryan Layer <rmlayer@us.ibm.com> * Kevin Gao <kevin.gao@intel.com> - * - * Any bugs reported given to us we will try to fix... any fixes shared will - * be incorporated into the next SCTP release. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/types.h> #include <linux/fcntl.h> #include <linux/poll.h> #include <linux/init.h> -#include <linux/sched.h> #include <linux/slab.h> #include <linux/in.h> @@ -61,8 +55,10 @@ #include <net/sctp/sm.h> /* Forward declarations for internal functions. */ -static void sctp_assoc_bh_rcv(struct sctp_association *asoc); - +static void sctp_select_active_and_retran_path(struct sctp_association *asoc); +static void sctp_assoc_bh_rcv(struct work_struct *work); +static void sctp_assoc_free_asconf_acks(struct sctp_association *asoc); +static void sctp_assoc_free_asconf_queue(struct sctp_association *asoc); /* 1st Level Abstractions. */ @@ -73,21 +69,20 @@ static struct sctp_association *sctp_association_init(struct sctp_association *a sctp_scope_t scope, gfp_t gfp) { + struct net *net = sock_net(sk); struct sctp_sock *sp; int i; + sctp_paramhdr_t *p; + int err; /* 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); - - /* Hold the sock. */ asoc->base.sk = (struct sock *)sk; + + sctp_endpoint_hold(asoc->ep); sock_hold(asoc->base.sk); /* Initialize the common base substructure. */ @@ -95,46 +90,68 @@ static struct sctp_association *sctp_association_init(struct sctp_association *a /* Initialize the object handling fields. */ atomic_set(&asoc->base.refcnt, 1); - asoc->base.dead = 0; - asoc->base.malloced = 0; /* Initialize the bind addr area. */ sctp_bind_addr_init(&asoc->base.bind_addr, ep->base.bind_addr.port); - rwlock_init(&asoc->base.addr_lock); asoc->state = SCTP_STATE_CLOSED; - - /* Set these values from the socket values, a conversion between - * millsecons to seconds/microseconds must also be done. - */ - asoc->cookie_life.tv_sec = sp->assocparams.sasoc_cookie_life / 1000; - asoc->cookie_life.tv_usec = (sp->assocparams.sasoc_cookie_life % 1000) - * 1000; - asoc->pmtu = 0; - asoc->frag_point = 0; + asoc->cookie_life = ms_to_ktime(sp->assocparams.sasoc_cookie_life); + asoc->user_frag = sp->user_frag; /* Set the association max_retrans and RTO values from the * socket values. */ asoc->max_retrans = sp->assocparams.sasoc_asocmaxrxt; + asoc->pf_retrans = net->sctp.pf_retrans; + asoc->rto_initial = msecs_to_jiffies(sp->rtoinfo.srto_initial); asoc->rto_max = msecs_to_jiffies(sp->rtoinfo.srto_max); asoc->rto_min = msecs_to_jiffies(sp->rtoinfo.srto_min); - asoc->overall_error_count = 0; + /* Initialize the association's heartbeat interval based on the + * sock configured value. + */ + asoc->hbinterval = msecs_to_jiffies(sp->hbinterval); - /* Initialize the maximum mumber of new data packets that can be sent + /* Initialize path max retrans value. */ + asoc->pathmaxrxt = sp->pathmaxrxt; + + /* Initialize default path MTU. */ + asoc->pathmtu = sp->pathmtu; + + /* 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->param_flags = sp->param_flags; + + /* Initialize the maximum number of new data packets that can be sent * in a burst. */ - asoc->max_burst = sctp_max_burst; + asoc->max_burst = sp->max_burst; - /* Copy things from the endpoint. */ - for (i = SCTP_EVENT_TIMEOUT_NONE; i < SCTP_NUM_TIMEOUT_TYPES; ++i) { - asoc->timeouts[i] = ep->timeouts[i]; - init_timer(&asoc->timers[i]); - asoc->timers[i].function = sctp_timer_events[i]; - asoc->timers[i].data = (unsigned long) asoc; - } + /* initialize association timers */ + asoc->timeouts[SCTP_EVENT_TIMEOUT_T1_COOKIE] = asoc->rto_initial; + asoc->timeouts[SCTP_EVENT_TIMEOUT_T1_INIT] = asoc->rto_initial; + asoc->timeouts[SCTP_EVENT_TIMEOUT_T2_SHUTDOWN] = asoc->rto_initial; + + /* sctpimpguide Section 2.12.2 + * If the 'T5-shutdown-guard' timer is used, it SHOULD be set to the + * recommended value of 5 times 'RTO.Max'. + */ + asoc->timeouts[SCTP_EVENT_TIMEOUT_T5_SHUTDOWN_GUARD] + = 5 * asoc->rto_max; + + asoc->timeouts[SCTP_EVENT_TIMEOUT_SACK] = asoc->sackdelay; + asoc->timeouts[SCTP_EVENT_TIMEOUT_AUTOCLOSE] = sp->autoclose * HZ; + + /* Initializes the timers */ + for (i = SCTP_EVENT_TIMEOUT_NONE; i < SCTP_NUM_TIMEOUT_TYPES; ++i) + setup_timer(&asoc->timers[i], sctp_timer_events[i], + (unsigned long)asoc); /* Pull default initialization values from the sock options. * Note: This assumes that the values have already been @@ -147,38 +164,27 @@ static struct sctp_association *sctp_association_init(struct sctp_association *a asoc->max_init_timeo = msecs_to_jiffies(sp->initmsg.sinit_max_init_timeo); - /* Allocate storage for the ssnmap after the inbound and outbound - * streams have been negotiated during Init. - */ - asoc->ssnmap = NULL; - /* Set the local window size for receive. * This is also the rcvbuf space per association. * RFC 6 - A SCTP receiver MUST be able to receive a minimum of * 1500 bytes in one SCTP packet. */ - if (sk->sk_rcvbuf < SCTP_DEFAULT_MINWINDOW) + if ((sk->sk_rcvbuf/2) < SCTP_DEFAULT_MINWINDOW) asoc->rwnd = SCTP_DEFAULT_MINWINDOW; else - asoc->rwnd = sk->sk_rcvbuf; + asoc->rwnd = sk->sk_rcvbuf/2; asoc->a_rwnd = asoc->rwnd; - asoc->rwnd_over = 0; - /* Use my own max window until I learn something better. */ asoc->peer.rwnd = SCTP_DEFAULT_MAXWINDOW; - /* Set the sndbuf size for transmit. */ - asoc->sndbuf_used = 0; + /* Initialize the receive memory counter */ + atomic_set(&asoc->rmem_alloc, 0); init_waitqueue_head(&asoc->wait); asoc->c.my_vtag = sctp_generate_tag(ep); - asoc->peer.i.init_tag = 0; /* INIT needs a vtag of 0. */ - asoc->c.peer_vtag = 0; - asoc->c.my_ttag = 0; - asoc->c.peer_ttag = 0; asoc->c.my_port = ep->base.bind_addr.port; asoc->c.initial_tsn = sctp_generate_tsn(ep); @@ -189,7 +195,6 @@ static struct sctp_association *sctp_association_init(struct sctp_association *a asoc->adv_peer_ack_point = asoc->ctsn_ack_point; asoc->highest_sacked = asoc->ctsn_ack_point; asoc->last_cwr_tsn = asoc->ctsn_ack_point; - asoc->unack_data = 0; /* ADDIP Section 4.1 Asconf Chunk Procedures * @@ -204,10 +209,10 @@ static struct sctp_association *sctp_association_init(struct sctp_association *a asoc->addip_serial = asoc->c.initial_tsn; INIT_LIST_HEAD(&asoc->addip_chunk_list); + INIT_LIST_HEAD(&asoc->asconf_ack_list); /* Make an empty list of remote transport addresses. */ INIT_LIST_HEAD(&asoc->peer.transport_addr_list); - asoc->peer.transport_count = 0; /* RFC 2960 5.1 Normal Establishment of an Association * @@ -221,17 +226,19 @@ static struct sctp_association *sctp_association_init(struct sctp_association *a * already received one packet.] */ asoc->peer.sack_needed = 1; + asoc->peer.sack_generation = 1; - /* Assume that the peer recongizes ASCONF until reported otherwise - * via an ERROR chunk. + /* Assume that the peer will tell us if he recognizes ASCONF + * as part of INIT exchange. + * The sctp_addip_noauth option is there for backward compatibility + * and will revert old behavior. */ - asoc->peer.asconf_capable = 1; + if (net->sctp.addip_noauth) + asoc->peer.asconf_capable = 1; /* Create an input queue. */ sctp_inq_init(&asoc->base.inqueue); - sctp_inq_set_th_handler(&asoc->base.inqueue, - (void (*)(void *))sctp_assoc_bh_rcv, - asoc); + sctp_inq_set_th_handler(&asoc->base.inqueue, sctp_assoc_bh_rcv); /* Create an output queue. */ sctp_outq_init(asoc, &asoc->outqueue); @@ -239,33 +246,48 @@ static struct sctp_association *sctp_association_init(struct sctp_association *a 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); - - asoc->need_ecne = 0; - - asoc->assoc_id = 0; - /* Assume that peer would support both address types unless we are * 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; - asoc->default_stream = sp->default_stream; asoc->default_ppid = sp->default_ppid; asoc->default_flags = sp->default_flags; asoc->default_context = sp->default_context; asoc->default_timetolive = sp->default_timetolive; + asoc->default_rcv_context = sp->default_rcv_context; + + /* AUTH related initializations */ + INIT_LIST_HEAD(&asoc->endpoint_shared_keys); + err = sctp_auth_asoc_copy_shkeys(ep, asoc, gfp); + if (err) + goto fail_init; + + asoc->active_key_id = ep->active_key_id; + + /* Save the hmacs and chunks list into this association */ + if (ep->auth_hmacs_list) + memcpy(asoc->c.auth_hmacs, ep->auth_hmacs_list, + ntohs(ep->auth_hmacs_list->param_hdr.length)); + if (ep->auth_chunk_list) + memcpy(asoc->c.auth_chunks, ep->auth_chunk_list, + ntohs(ep->auth_chunk_list->param_hdr.length)); + + /* Get the AUTH random number for this association */ + p = (sctp_paramhdr_t *)asoc->c.auth_random; + p->type = SCTP_PARAM_RANDOM; + p->length = htons(sizeof(sctp_paramhdr_t) + SCTP_AUTH_RANDOM_LENGTH); + get_random_bytes(p+1, SCTP_AUTH_RANDOM_LENGTH); return asoc; fail_init: - sctp_endpoint_put(asoc->ep); sock_put(asoc->base.sk); + sctp_endpoint_put(asoc->ep); return NULL; } @@ -277,16 +299,16 @@ struct sctp_association *sctp_association_new(const struct sctp_endpoint *ep, { struct sctp_association *asoc; - asoc = t_new(struct sctp_association, gfp); + asoc = kzalloc(sizeof(*asoc), gfp); if (!asoc) goto fail; if (!sctp_association_init(asoc, ep, sk, scope, gfp)) goto fail_init; - asoc->base.malloced = 1; SCTP_DBG_OBJCNT_INC(assoc); - SCTP_DEBUG_PRINTK("Created asoc %p\n", asoc); + + pr_debug("Created asoc %p\n", asoc); return asoc; @@ -306,16 +328,23 @@ void sctp_association_free(struct sctp_association *asoc) struct list_head *pos, *temp; int i; - list_del(&asoc->asocs); + /* Only real associations count against the endpoint, so + * don't bother for if this is a temporary association. + */ + if (!list_empty(&asoc->asocs)) { + list_del(&asoc->asocs); - /* Decrement the backlog value for a TCP-style listening socket. */ - if (sctp_style(sk, TCP) && sctp_sstate(sk, LISTENING)) - sk->sk_ack_backlog--; + /* Decrement the backlog value for a TCP-style listening + * socket. + */ + if (sctp_style(sk, TCP) && sctp_sstate(sk, LISTENING)) + sk->sk_ack_backlog--; + } /* Mark as dead, so other users can know this structure is * going away. */ - asoc->base.dead = 1; + asoc->base.dead = true; /* Dispose of any data lying around in the outqueue. */ sctp_outq_free(&asoc->outqueue); @@ -326,6 +355,8 @@ void sctp_association_free(struct sctp_association *asoc) /* 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); @@ -338,32 +369,36 @@ void sctp_association_free(struct sctp_association *asoc) * on our state. */ for (i = SCTP_EVENT_TIMEOUT_NONE; i < SCTP_NUM_TIMEOUT_TYPES; ++i) { - if (timer_pending(&asoc->timers[i]) && - del_timer(&asoc->timers[i])) + if (del_timer(&asoc->timers[i])) sctp_association_put(asoc); } /* Free peer's cached cookie. */ - if (asoc->peer.cookie) { - kfree(asoc->peer.cookie); - } + kfree(asoc->peer.cookie); + kfree(asoc->peer.peer_random); + kfree(asoc->peer.peer_chunks); + kfree(asoc->peer.peer_hmacs); /* Release the transport structures. */ list_for_each_safe(pos, temp, &asoc->peer.transport_addr_list) { transport = list_entry(pos, struct sctp_transport, transports); - list_del(pos); + list_del_rcu(pos); sctp_transport_free(transport); } asoc->peer.transport_count = 0; - /* Free any cached ASCONF_ACK chunk. */ - if (asoc->addip_last_asconf_ack) - sctp_chunk_free(asoc->addip_last_asconf_ack); + sctp_asconf_queue_teardown(asoc); - /* Free any cached ASCONF chunk. */ - if (asoc->addip_last_asconf) - sctp_chunk_free(asoc->addip_last_asconf); + /* Free pending address space being deleted */ + if (asoc->asconf_addr_del_pending != NULL) + kfree(asoc->asconf_addr_del_pending); + + /* AUTH - Free the endpoint shared keys */ + sctp_auth_destroy_keys(&asoc->endpoint_shared_keys); + + /* AUTH - Free the association shared key */ + sctp_auth_key_put(asoc->asoc_shared_key); sctp_association_put(asoc); } @@ -371,7 +406,10 @@ void sctp_association_free(struct sctp_association *asoc) /* Cleanup and free up an association. */ static void sctp_association_destroy(struct sctp_association *asoc) { - SCTP_ASSERT(asoc->base.dead, "Assoc is not dead", return); + if (unlikely(!asoc->base.dead)) { + WARN(1, "Attempt to destroy undead association %p!\n", asoc); + return; + } sctp_endpoint_put(asoc->ep); sock_put(asoc->base.sk); @@ -382,16 +420,25 @@ static void sctp_association_destroy(struct sctp_association *asoc) spin_unlock_bh(&sctp_assocs_id_lock); } - if (asoc->base.malloced) { - kfree(asoc); - SCTP_DBG_OBJCNT_DEC(assoc); - } + WARN_ON(atomic_read(&asoc->rmem_alloc)); + + kfree(asoc); + SCTP_DBG_OBJCNT_DEC(assoc); } /* Change the primary destination address for the peer. */ 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. */ @@ -401,7 +448,8 @@ void sctp_assoc_set_primary(struct sctp_association *asoc, /* If the primary path is changing, assume that the * user wants to use this new path. */ - if (transport->state != SCTP_INACTIVE) + if ((transport->state == SCTP_ACTIVE) || + (transport->state == SCTP_UNKNOWN)) asoc->peer.active_path = transport; /* @@ -414,14 +462,20 @@ void sctp_assoc_set_primary(struct sctp_association *asoc, * 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. @@ -436,11 +490,8 @@ void sctp_assoc_rm_peer(struct sctp_association *asoc, struct list_head *pos; struct sctp_transport *transport; - SCTP_DEBUG_PRINTK_IPADDR("sctp_assoc_rm_peer:association %p addr: ", - " port: %d\n", - asoc, - (&peer->ipaddr), - peer->ipaddr.v4.sin_port); + pr_debug("%s: association:%p addr:%pISpc\n", + __func__, asoc, &peer->ipaddr.sa); /* If we are to remove the current retran_path, update it * to the next peer before removing this peer from the list. @@ -449,7 +500,7 @@ void sctp_assoc_rm_peer(struct sctp_association *asoc, sctp_assoc_update_retran_path(asoc); /* Remove this peer from the list. */ - list_del(&peer->transports); + list_del_rcu(&peer->transports); /* Get the first transport of asoc. */ pos = asoc->peer.transport_addr_list.next; @@ -460,6 +511,8 @@ void sctp_assoc_rm_peer(struct sctp_association *asoc, sctp_assoc_set_primary(asoc, transport); if (asoc->peer.active_path == peer) asoc->peer.active_path = transport; + if (asoc->peer.retran_path == peer) + asoc->peer.retran_path = transport; if (asoc->peer.last_data_from == peer) asoc->peer.last_data_from = transport; @@ -471,6 +524,48 @@ void sctp_assoc_rm_peer(struct sctp_association *asoc, 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 + * retransmitted. + */ + 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); @@ -482,6 +577,7 @@ struct sctp_transport *sctp_assoc_add_peer(struct sctp_association *asoc, const gfp_t gfp, const int peer_state) { + struct net *net = sock_net(asoc->base.sk); struct sctp_transport *peer; struct sctp_sock *sp; unsigned short port; @@ -489,14 +585,10 @@ struct sctp_transport *sctp_assoc_add_peer(struct sctp_association *asoc, sp = sctp_sk(asoc->base.sk); /* AF_INET and AF_INET6 share common port field. */ - port = addr->v4.sin_port; + port = ntohs(addr->v4.sin_port); - SCTP_DEBUG_PRINTK_IPADDR("sctp_assoc_add_peer:association %p addr: ", - " port: %d state:%s\n", - asoc, - addr, - addr->v4.sin_port, - peer_state == SCTP_UNKNOWN?"UNKNOWN":"ACTIVE"); + pr_debug("%s: association:%p addr:%pISpc state:%d\n", __func__, + asoc, &addr->sa, peer_state); /* Set the port if it has not been set yet. */ if (0 == asoc->peer.port) @@ -505,35 +597,70 @@ struct sctp_transport *sctp_assoc_add_peer(struct sctp_association *asoc, /* Check to see if this is a duplicate. */ peer = sctp_assoc_lookup_paddr(asoc, addr); if (peer) { - if (peer_state == SCTP_ACTIVE && - peer->state == SCTP_UNKNOWN) - peer->state = SCTP_ACTIVE; + /* 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) { + peer->state = SCTP_ACTIVE; + } return peer; } - peer = sctp_transport_new(addr, gfp); + peer = sctp_transport_new(net, addr, gfp); if (!peer) return NULL; sctp_transport_set_owner(peer, asoc); + /* Initialize the peer's heartbeat interval based on the + * association configured value. + */ + peer->hbinterval = asoc->hbinterval; + + /* Set the path max_retrans. */ + peer->pathmaxrxt = asoc->pathmaxrxt; + + /* And the partial failure retrans threshold */ + peer->pf_retrans = asoc->pf_retrans; + + /* Initialize the peer's SACK delay timeout based on the + * 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. */ - sctp_transport_pmtu(peer); + 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. * If not and the current association PMTU is higher than the new * peer's PMTU, reset the association PMTU to the new peer's PMTU. */ - if (asoc->pmtu) - asoc->pmtu = min_t(int, peer->pmtu, asoc->pmtu); + if (asoc->pathmtu) + asoc->pathmtu = min_t(int, peer->pathmtu, asoc->pathmtu); else - asoc->pmtu = peer->pmtu; + asoc->pathmtu = peer->pathmtu; - SCTP_DEBUG_PRINTK("sctp_assoc_add_peer:association %p PMTU set to " - "%d\n", asoc, asoc->pmtu); + pr_debug("%s: association:%p PMTU set to %d\n", __func__, asoc, + asoc->pathmtu); - asoc->frag_point = sctp_frag_point(sp, asoc->pmtu); + peer->pmtu_pending = 0; + + asoc->frag_point = sctp_frag_point(asoc, asoc->pathmtu); /* The asoc->peer.port might not be meaningful yet, but * initialize the packet structure anyway. @@ -551,7 +678,7 @@ struct sctp_transport *sctp_assoc_add_peer(struct sctp_association *asoc, * (for example, implementations MAY use the size of the * receiver advertised window). */ - peer->cwnd = min(4*asoc->pmtu, max_t(__u32, 2*asoc->pmtu, 4380)); + peer->cwnd = min(4*asoc->pathmtu, max_t(__u32, 2*asoc->pathmtu, 4380)); /* At this point, we may not have the receiver's advertised window, * so initialize ssthresh to the default value and it will be set @@ -561,26 +688,17 @@ struct sctp_transport *sctp_assoc_add_peer(struct sctp_association *asoc, peer->partial_bytes_acked = 0; peer->flight_size = 0; - - /* By default, enable heartbeat for peer address. */ - peer->hb_allowed = 1; - - /* Initialize the peer's heartbeat interval based on the - * sock configured value. - */ - peer->hb_interval = msecs_to_jiffies(sp->paddrparam.spp_hbinterval); - - /* Set the path max_retrans. */ - peer->max_retrans = sp->paddrparam.spp_pathmaxrxt; + peer->burst_limited = 0; /* Set the transport's RTO.initial value */ peer->rto = asoc->rto_initial; + sctp_max_rto(asoc, peer); /* Set the peer's active state. */ peer->state = peer_state; /* Attach the remote transport to our asoc. */ - list_add_tail(&peer->transports, &asoc->peer.transport_addr_list); + list_add_tail_rcu(&peer->transports, &asoc->peer.transport_addr_list); asoc->peer.transport_count++; /* If we do not yet have a primary path, set one. */ @@ -589,7 +707,8 @@ struct sctp_transport *sctp_assoc_add_peer(struct sctp_association *asoc, 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; } @@ -620,12 +739,11 @@ struct sctp_transport *sctp_assoc_lookup_paddr( const union sctp_addr *address) { struct sctp_transport *t; - struct list_head *pos; /* Cycle through all transports searching for a peer address. */ - list_for_each(pos, &asoc->peer.transport_addr_list) { - t = list_entry(pos, struct sctp_transport, transports); + list_for_each_entry(t, &asoc->peer.transport_addr_list, + transports) { if (sctp_cmp_addr_exact(address, &t->ipaddr)) return t; } @@ -633,6 +751,21 @@ struct sctp_transport *sctp_assoc_lookup_paddr( return NULL; } +/* Remove all transports except a give one */ +void sctp_assoc_del_nonprimary_peers(struct sctp_association *asoc, + struct sctp_transport *primary) +{ + struct sctp_transport *temp; + struct sctp_transport *t; + + list_for_each_entry_safe(t, temp, &asoc->peer.transport_addr_list, + transports) { + /* if the current transport is not the primary one, delete it */ + if (t != primary) + sctp_assoc_rm_peer(asoc, t); + } +} + /* Engage in transport control operations. * Mark the transport up or down and send a notification to the user. * Select and update the new active and retran paths. @@ -642,88 +775,74 @@ void sctp_assoc_control_transport(struct sctp_association *asoc, sctp_transport_cmd_t command, sctp_sn_error_t error) { - struct sctp_transport *t = NULL; - struct sctp_transport *first; - struct sctp_transport *second; struct sctp_ulpevent *event; - struct list_head *pos; + struct sockaddr_storage addr; int spc_state = 0; + bool ulp_notify = true; /* Record the transition on the transport. */ switch (command) { case SCTP_TRANSPORT_UP: + /* If we are moving from UNCONFIRMED state due + * to heartbeat success, report the SCTP_ADDR_CONFIRMED + * state to the user, otherwise report SCTP_ADDR_AVAILABLE. + */ + if (SCTP_UNCONFIRMED == transport->state && + SCTP_HEARTBEAT_SUCCESS == error) + spc_state = SCTP_ADDR_CONFIRMED; + else + spc_state = SCTP_ADDR_AVAILABLE; + /* Don't inform ULP about transition from PF to + * active state and set cwnd to 1 MTU, see SCTP + * Quick failover draft section 5.1, point 5 + */ + if (transport->state == SCTP_PF) { + ulp_notify = false; + transport->cwnd = asoc->pathmtu; + } transport->state = SCTP_ACTIVE; - spc_state = SCTP_ADDR_AVAILABLE; break; case SCTP_TRANSPORT_DOWN: - transport->state = SCTP_INACTIVE; + /* 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; + case SCTP_TRANSPORT_PF: + transport->state = SCTP_PF; + ulp_notify = false; + break; + default: return; - }; - - /* Generate and send a SCTP_PEER_ADDR_CHANGE notification to the - * user. - */ - event = sctp_ulpevent_make_peer_addr_change(asoc, - (struct sockaddr_storage *) &transport->ipaddr, - 0, spc_state, error, GFP_ATOMIC); - if (event) - sctp_ulpq_tail_event(&asoc->ulpq, event); - - /* Select new active and retran paths. */ - - /* Look for the two most recently used active transports. - * - * This code produces the wrong ordering whenever jiffies - * rolls over, but we still get usable transports, so we don't - * worry about it. - */ - first = NULL; second = NULL; - - list_for_each(pos, &asoc->peer.transport_addr_list) { - t = list_entry(pos, struct sctp_transport, transports); - - if (t->state == SCTP_INACTIVE) - continue; - if (!first || t->last_time_heard > first->last_time_heard) { - second = first; - first = t; - } - if (!second || t->last_time_heard > second->last_time_heard) - second = t; } - /* RFC 2960 6.4 Multi-Homed SCTP Endpoints - * - * By default, an endpoint should always transmit to the - * primary path, unless the SCTP user explicitly specifies the - * destination transport address (and possibly source - * transport address) to use. - * - * [If the primary is active but not most recent, bump the most - * recently used transport.] + /* Generate and send a SCTP_PEER_ADDR_CHANGE notification + * to the user. */ - if (asoc->peer.primary_path->state != SCTP_INACTIVE && - first != asoc->peer.primary_path) { - second = first; - first = asoc->peer.primary_path; + if (ulp_notify) { + memset(&addr, 0, sizeof(struct sockaddr_storage)); + memcpy(&addr, &transport->ipaddr, + transport->af_specific->sockaddr_len); + + event = sctp_ulpevent_make_peer_addr_change(asoc, &addr, + 0, spc_state, error, GFP_ATOMIC); + if (event) + sctp_ulpq_tail_event(&asoc->ulpq, event); } - /* If we failed to find a usable transport, just camp on the - * primary, even if it is inactive. - */ - if (!first) { - first = asoc->peer.primary_path; - second = asoc->peer.primary_path; - } - - /* Set the active and retran transports. */ - asoc->peer.active_path = first; - asoc->peer.retran_path = second; + /* Select new active and retran paths. */ + sctp_select_active_and_retran_path(asoc); } /* Hold a reference to an association. */ @@ -779,17 +898,13 @@ int sctp_cmp_addr_exact(const union sctp_addr *ss1, */ struct sctp_chunk *sctp_get_ecne_prepend(struct sctp_association *asoc) { - struct sctp_chunk *chunk; + if (!asoc->need_ecne) + return NULL; /* Send ECNE if needed. * Not being able to allocate a chunk here is not deadly. */ - if (asoc->need_ecne) - chunk = sctp_make_ecne(asoc, asoc->last_ecne_tsn); - else - chunk = NULL; - - return chunk; + return sctp_make_ecne(asoc, asoc->last_ecne_tsn); } /* @@ -800,10 +915,9 @@ struct sctp_transport *sctp_assoc_lookup_tsn(struct sctp_association *asoc, { struct sctp_transport *active; struct sctp_transport *match; - struct list_head *entry, *pos; struct sctp_transport *transport; struct sctp_chunk *chunk; - __u32 key = htonl(tsn); + __be32 key = htonl(tsn); match = NULL; @@ -824,8 +938,8 @@ struct sctp_transport *sctp_assoc_lookup_tsn(struct sctp_association *asoc, active = asoc->peer.active_path; - list_for_each(entry, &active->transmitted) { - chunk = list_entry(entry, struct sctp_chunk, transmitted_list); + list_for_each_entry(chunk, &active->transmitted, + transmitted_list) { if (key == chunk->subh.data_hdr->tsn) { match = active; @@ -834,14 +948,13 @@ struct sctp_transport *sctp_assoc_lookup_tsn(struct sctp_association *asoc, } /* If not found, go search all the other transports. */ - list_for_each(pos, &asoc->peer.transport_addr_list) { - transport = list_entry(pos, struct sctp_transport, transports); + list_for_each_entry(transport, &asoc->peer.transport_addr_list, + transports) { if (transport == active) - break; - list_for_each(entry, &transport->transmitted) { - chunk = list_entry(entry, struct sctp_chunk, - transmitted_list); + continue; + list_for_each_entry(chunk, &transport->transmitted, + transmitted_list) { if (key == chunk->subh.data_hdr->tsn) { match = transport; goto out; @@ -854,15 +967,15 @@ out: /* Is this the association we are looking for? */ struct sctp_transport *sctp_assoc_is_match(struct sctp_association *asoc, + struct net *net, const union sctp_addr *laddr, const union sctp_addr *paddr) { struct sctp_transport *transport; - sctp_read_lock(&asoc->base.addr_lock); - - if ((asoc->base.bind_addr.port == laddr->v4.sin_port) && - (asoc->peer.port == paddr->v4.sin_port)) { + if ((htons(asoc->base.bind_addr.port) == laddr->v4.sin_port) && + (htons(asoc->peer.port) == paddr->v4.sin_port) && + net_eq(sock_net(asoc->base.sk), net)) { transport = sctp_assoc_lookup_paddr(asoc, paddr); if (!transport) goto out; @@ -874,16 +987,18 @@ struct sctp_transport *sctp_assoc_is_match(struct sctp_association *asoc, transport = NULL; out: - sctp_read_unlock(&asoc->base.addr_lock); return transport; } /* Do delayed input processing. This is scheduled by sctp_rcv(). */ -static void sctp_assoc_bh_rcv(struct sctp_association *asoc) +static void sctp_assoc_bh_rcv(struct work_struct *work) { + struct sctp_association *asoc = + container_of(work, struct sctp_association, + base.inqueue.immediate); + struct net *net = sock_net(asoc->base.sk); struct sctp_endpoint *ep; struct sctp_chunk *chunk; - struct sock *sk; struct sctp_inq *inqueue; int state; sctp_subtype_t subtype; @@ -891,7 +1006,6 @@ static void sctp_assoc_bh_rcv(struct sctp_association *asoc) /* The association should be held so we should be safe. */ ep = asoc->ep; - sk = asoc->base.sk; inqueue = &asoc->base.inqueue; sctp_association_hold(asoc); @@ -899,19 +1013,33 @@ static void sctp_assoc_bh_rcv(struct sctp_association *asoc) state = asoc->state; subtype = SCTP_ST_CHUNK(chunk->chunk_hdr->type); + /* SCTP-AUTH, Section 6.3: + * The receiver has a list of chunk types which it expects + * to be received only after an AUTH-chunk. This list has + * been sent to the peer during the association setup. It + * MUST silently discard these chunks if they are not placed + * after an AUTH chunk in the packet. + */ + if (sctp_auth_recv_cid(subtype.chunk, asoc) && !chunk->auth) + continue; + /* Remember where the last DATA chunk came from so we * know where to send the SACK. */ if (sctp_chunk_is_data(chunk)) asoc->peer.last_data_from = chunk->transport; - else - SCTP_INC_STATS(SCTP_MIB_INCTRLCHUNKS); + else { + SCTP_INC_STATS(net, SCTP_MIB_INCTRLCHUNKS); + asoc->stats.ictrlchunks++; + if (chunk->chunk_hdr->type == SCTP_CID_SACK) + asoc->stats.isacks++; + } if (chunk->transport) - chunk->transport->last_time_heard = jiffies; + chunk->transport->last_time_heard = ktime_get(); /* Run through the state machine. */ - error = sctp_do_sm(SCTP_EVENT_T_CHUNK, subtype, + error = sctp_do_sm(net, SCTP_EVENT_T_CHUNK, subtype, state, ep, asoc, chunk, GFP_ATOMIC); /* Check to see if the association is freed in response to @@ -969,15 +1097,21 @@ void sctp_assoc_update(struct sctp_association *asoc, asoc->c = new->c; asoc->peer.rwnd = new->peer.rwnd; asoc->peer.sack_needed = new->peer.sack_needed; + asoc->peer.auth_capable = new->peer.auth_capable; 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); } /* If the case is A (association restart), use @@ -995,14 +1129,25 @@ void sctp_assoc_update(struct sctp_association *asoc, */ sctp_ssnmap_clear(asoc->ssnmap); + /* Flush the ULP reassembly and ordered queue. + * Any data there will now be stale and will + * cause problems. + */ + sctp_ulpq_flush(&asoc->ulpq); + + /* reset the overall association error count so + * that the restarted association doesn't get torn + * down on the next retransmission timer. + */ + asoc->overall_error_count = 0; + } else { /* Add any peer addresses from the new association. */ - list_for_each(pos, &new->peer.transport_addr_list) { - trans = list_entry(pos, struct sctp_transport, - transports); + list_for_each_entry(trans, &new->peer.transport_addr_list, + transports) { if (!sctp_assoc_lookup_paddr(asoc, &trans->ipaddr)) sctp_assoc_add_peer(asoc, &trans->ipaddr, - GFP_ATOMIC, SCTP_ACTIVE); + GFP_ATOMIC, trans->state); } asoc->ctsn_ack_point = asoc->next_tsn - 1; @@ -1012,161 +1157,291 @@ void sctp_assoc_update(struct sctp_association *asoc, asoc->ssnmap = new->ssnmap; new->ssnmap = NULL; } + + if (!asoc->assoc_id) { + /* get a new association id since we don't have one + * yet. + */ + sctp_assoc_set_id(asoc, GFP_ATOMIC); + } } + + /* SCTP-AUTH: Save the peer parameters from the new associations + * and also move the association shared keys over + */ + kfree(asoc->peer.peer_random); + asoc->peer.peer_random = new->peer.peer_random; + new->peer.peer_random = NULL; + + kfree(asoc->peer.peer_chunks); + asoc->peer.peer_chunks = new->peer.peer_chunks; + new->peer.peer_chunks = NULL; + + kfree(asoc->peer.peer_hmacs); + asoc->peer.peer_hmacs = new->peer.peer_hmacs; + new->peer.peer_hmacs = NULL; + + sctp_auth_key_put(asoc->asoc_shared_key); + sctp_auth_asoc_init_active_key(asoc, GFP_ATOMIC); } /* Update the retran path for sending a retransmitted packet. - * Round-robin through the active transports, else round-robin - * through the inactive transports as this is the next best thing - * we can try. + * See also RFC4960, 6.4. Multi-Homed SCTP Endpoints: + * + * When there is outbound data to send and the primary path + * becomes inactive (e.g., due to failures), or where the + * SCTP user explicitly requests to send data to an + * inactive destination transport address, before reporting + * an error to its ULP, the SCTP endpoint should try to send + * the data to an alternate active destination transport + * address if one exists. + * + * When retransmitting data that timed out, if the endpoint + * is multihomed, it should consider each source-destination + * address pair in its retransmission selection policy. + * When retransmitting timed-out data, the endpoint should + * attempt to pick the most divergent source-destination + * pair from the original source-destination pair to which + * the packet was transmitted. + * + * Note: Rules for picking the most divergent source-destination + * pair are an implementation decision and are not specified + * within this document. + * + * Our basic strategy is to round-robin transports in priorities + * according to sctp_state_prio_map[] e.g., if no such + * transport with state SCTP_ACTIVE exists, round-robin through + * SCTP_UNKNOWN, etc. You get the picture. */ -void sctp_assoc_update_retran_path(struct sctp_association *asoc) +static const u8 sctp_trans_state_to_prio_map[] = { + [SCTP_ACTIVE] = 3, /* best case */ + [SCTP_UNKNOWN] = 2, + [SCTP_PF] = 1, + [SCTP_INACTIVE] = 0, /* worst case */ +}; + +static u8 sctp_trans_score(const struct sctp_transport *trans) { - struct sctp_transport *t, *next; - struct list_head *head = &asoc->peer.transport_addr_list; - struct list_head *pos; - - /* Find the next transport in a round-robin fashion. */ - t = asoc->peer.retran_path; - pos = &t->transports; - next = NULL; - - while (1) { - /* Skip the head. */ - if (pos->next == head) - pos = head->next; - else - pos = pos->next; + return sctp_trans_state_to_prio_map[trans->state]; +} + +static struct sctp_transport *sctp_trans_elect_tie(struct sctp_transport *trans1, + struct sctp_transport *trans2) +{ + if (trans1->error_count > trans2->error_count) { + return trans2; + } else if (trans1->error_count == trans2->error_count && + ktime_after(trans2->last_time_heard, + trans1->last_time_heard)) { + return trans2; + } else { + return trans1; + } +} - t = list_entry(pos, struct sctp_transport, transports); +static struct sctp_transport *sctp_trans_elect_best(struct sctp_transport *curr, + struct sctp_transport *best) +{ + u8 score_curr, score_best; - /* Try to find an active transport. */ + if (best == NULL) + return curr; - if (t->state != SCTP_INACTIVE) { - break; - } else { - /* Keep track of the next transport in case - * we don't find any active transport. - */ - if (!next) - next = t; - } + score_curr = sctp_trans_score(curr); + score_best = sctp_trans_score(best); - /* 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; + /* First, try a score-based selection if both transport states + * differ. If we're in a tie, lets try to make a more clever + * decision here based on error counts and last time heard. + */ + if (score_curr > score_best) + return curr; + else if (score_curr == score_best) + return sctp_trans_elect_tie(curr, best); + else + return best; +} + +void sctp_assoc_update_retran_path(struct sctp_association *asoc) +{ + struct sctp_transport *trans = asoc->peer.retran_path; + struct sctp_transport *trans_next = NULL; + + /* We're done as we only have the one and only path. */ + if (asoc->peer.transport_count == 1) + return; + /* If active_path and retran_path are the same and active, + * then this is the only active path. Use it. + */ + if (asoc->peer.active_path == asoc->peer.retran_path && + asoc->peer.active_path->state == SCTP_ACTIVE) + return; + + /* Iterate from retran_path's successor back to retran_path. */ + for (trans = list_next_entry(trans, transports); 1; + trans = list_next_entry(trans, transports)) { + /* Manually skip the head element. */ + if (&trans->transports == &asoc->peer.transport_addr_list) + continue; + if (trans->state == SCTP_UNCONFIRMED) + continue; + trans_next = sctp_trans_elect_best(trans, trans_next); + /* Active is good enough for immediate return. */ + if (trans_next->state == SCTP_ACTIVE) + break; + /* We've reached the end, time to update path. */ + if (trans == asoc->peer.retran_path) break; - } } - asoc->peer.retran_path = t; + asoc->peer.retran_path = trans_next; - SCTP_DEBUG_PRINTK_IPADDR("sctp_assoc_update_retran_path:association" - " %p addr: ", - " port: %d\n", - asoc, - (&t->ipaddr), - t->ipaddr.v4.sin_port); + pr_debug("%s: association:%p updated new path to addr:%pISpc\n", + __func__, asoc, &asoc->peer.retran_path->ipaddr.sa); } -/* Choose the transport for sending a INIT packet. */ -struct sctp_transport *sctp_assoc_choose_init_transport( - struct sctp_association *asoc) +static void sctp_select_active_and_retran_path(struct sctp_association *asoc) { - struct sctp_transport *t; + struct sctp_transport *trans, *trans_pri = NULL, *trans_sec = NULL; + struct sctp_transport *trans_pf = NULL; + + /* Look for the two most recently used active transports. */ + list_for_each_entry(trans, &asoc->peer.transport_addr_list, + transports) { + /* Skip uninteresting transports. */ + if (trans->state == SCTP_INACTIVE || + trans->state == SCTP_UNCONFIRMED) + continue; + /* Keep track of the best PF transport from our + * list in case we don't find an active one. + */ + if (trans->state == SCTP_PF) { + trans_pf = sctp_trans_elect_best(trans, trans_pf); + continue; + } + /* For active transports, pick the most recent ones. */ + if (trans_pri == NULL || + ktime_after(trans->last_time_heard, + trans_pri->last_time_heard)) { + trans_sec = trans_pri; + trans_pri = trans; + } else if (trans_sec == NULL || + ktime_after(trans->last_time_heard, + trans_sec->last_time_heard)) { + trans_sec = trans; + } + } - /* Use the retran path. If the last INIT was sent over the - * retran path, update the retran path and use it. + /* RFC 2960 6.4 Multi-Homed SCTP Endpoints + * + * By default, an endpoint should always transmit to the primary + * path, unless the SCTP user explicitly specifies the + * destination transport address (and possibly source transport + * address) to use. [If the primary is active but not most recent, + * bump the most recently used transport.] */ - 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; + if ((asoc->peer.primary_path->state == SCTP_ACTIVE || + asoc->peer.primary_path->state == SCTP_UNKNOWN) && + asoc->peer.primary_path != trans_pri) { + trans_sec = trans_pri; + trans_pri = asoc->peer.primary_path; } - SCTP_DEBUG_PRINTK_IPADDR("sctp_assoc_update_retran_path:association" - " %p addr: ", - " port: %d\n", - asoc, - (&t->ipaddr), - t->ipaddr.v4.sin_port); + /* We did not find anything useful for a possible retransmission + * path; either primary path that we found is the the same as + * the current one, or we didn't generally find an active one. + */ + if (trans_sec == NULL) + trans_sec = trans_pri; + + /* If we failed to find a usable transport, just camp on the + * primary or retran, even if they are inactive, if possible + * pick a PF iff it's the better choice. + */ + if (trans_pri == NULL) { + trans_pri = sctp_trans_elect_best(asoc->peer.primary_path, + asoc->peer.retran_path); + trans_pri = sctp_trans_elect_best(trans_pri, trans_pf); + trans_sec = asoc->peer.primary_path; + } - return t; + /* Set the active and retran transports. */ + asoc->peer.active_path = trans_pri; + asoc->peer.retran_path = trans_sec; } -/* Choose the transport for sending a SHUTDOWN packet. */ -struct sctp_transport *sctp_assoc_choose_shutdown_transport( - struct sctp_association *asoc) +struct sctp_transport * +sctp_assoc_choose_alter_transport(struct sctp_association *asoc, + struct sctp_transport *last_sent_to) { - /* 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 + /* 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->shutdown_last_sent_to) + if (last_sent_to == NULL) { return asoc->peer.active_path; - else { - if (asoc->shutdown_last_sent_to == asoc->peer.retran_path) + } else { + 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 * transports. This routine is called when a transport's PMTU has changed. */ -void sctp_assoc_sync_pmtu(struct sctp_association *asoc) +void sctp_assoc_sync_pmtu(struct sock *sk, struct sctp_association *asoc) { struct sctp_transport *t; - struct list_head *pos; __u32 pmtu = 0; if (!asoc) return; /* Get the lowest pmtu of all the transports. */ - list_for_each(pos, &asoc->peer.transport_addr_list) { - t = list_entry(pos, struct sctp_transport, transports); - if (!pmtu || (t->pmtu < pmtu)) - pmtu = t->pmtu; + list_for_each_entry(t, &asoc->peer.transport_addr_list, + transports) { + if (t->pmtu_pending && t->dst) { + sctp_transport_update_pmtu(sk, t, dst_mtu(t->dst)); + t->pmtu_pending = 0; + } + if (!pmtu || (t->pathmtu < pmtu)) + pmtu = t->pathmtu; } if (pmtu) { - struct sctp_sock *sp = sctp_sk(asoc->base.sk); - asoc->pmtu = pmtu; - asoc->frag_point = sctp_frag_point(sp, pmtu); + asoc->pathmtu = pmtu; + asoc->frag_point = sctp_frag_point(asoc, pmtu); } - SCTP_DEBUG_PRINTK("%s: asoc:%p, pmtu:%d, frag_point:%d\n", - __FUNCTION__, asoc, asoc->pmtu, asoc->frag_point); + pr_debug("%s: asoc:%p, pmtu:%d, frag_point:%d\n", __func__, asoc, + asoc->pathmtu, asoc->frag_point); } /* Should we send a SACK to update our peer? */ -static inline int sctp_peer_needs_update(struct sctp_association *asoc) +static inline bool sctp_peer_needs_update(struct sctp_association *asoc) { + struct net *net = sock_net(asoc->base.sk); switch (asoc->state) { case SCTP_STATE_ESTABLISHED: case SCTP_STATE_SHUTDOWN_PENDING: 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->pmtu))) - return 1; + ((asoc->rwnd - asoc->a_rwnd) >= max_t(__u32, + (asoc->base.sk->sk_rcvbuf >> net->sctp.rwnd_upd_shift), + asoc->pathmtu))) + return true; break; default: break; } - return 0; + return false; } /* Increase asoc's rwnd by len and send any window update SACK if needed. */ -void sctp_assoc_rwnd_increase(struct sctp_association *asoc, unsigned len) +void sctp_assoc_rwnd_increase(struct sctp_association *asoc, unsigned int len) { struct sctp_chunk *sack; struct timer_list *timer; @@ -1182,9 +1457,20 @@ void sctp_assoc_rwnd_increase(struct sctp_association *asoc, unsigned len) asoc->rwnd += len; } - SCTP_DEBUG_PRINTK("%s: asoc %p rwnd increased by %d to (%u, %u) " - "- %u\n", __FUNCTION__, asoc, len, asoc->rwnd, - asoc->rwnd_over, asoc->a_rwnd); + /* 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; + } + + pr_debug("%s: asoc:%p rwnd increased by %d to (%u, %u) - %u\n", + __func__, asoc, len, asoc->rwnd, asoc->rwnd_over, + asoc->a_rwnd); /* Send a window update SACK if the rwnd has increased by at least the * minimum of the association's PMTU and half of the receive buffer. @@ -1193,9 +1479,11 @@ void sctp_assoc_rwnd_increase(struct sctp_association *asoc, unsigned len) */ if (sctp_peer_needs_update(asoc)) { asoc->a_rwnd = asoc->rwnd; - SCTP_DEBUG_PRINTK("%s: Sending window update SACK- asoc: %p " - "rwnd: %u a_rwnd: %u\n", __FUNCTION__, - asoc, asoc->rwnd, asoc->a_rwnd); + + pr_debug("%s: sending window update SACK- asoc:%p rwnd:%u " + "a_rwnd:%u\n", __func__, asoc, asoc->rwnd, + asoc->a_rwnd); + sack = sctp_make_sack(asoc); if (!sack) return; @@ -1206,47 +1494,70 @@ void sctp_assoc_rwnd_increase(struct sctp_association *asoc, unsigned len) /* Stop the SACK timer. */ timer = &asoc->timers[SCTP_EVENT_TIMEOUT_SACK]; - if (timer_pending(timer) && del_timer(timer)) + if (del_timer(timer)) sctp_association_put(asoc); } } /* Decrease asoc's rwnd by len. */ -void sctp_assoc_rwnd_decrease(struct sctp_association *asoc, unsigned len) +void sctp_assoc_rwnd_decrease(struct sctp_association *asoc, unsigned int len) { - SCTP_ASSERT(asoc->rwnd, "rwnd zero", return); - SCTP_ASSERT(!asoc->rwnd_over, "rwnd_over not zero", return); + int rx_count; + int over = 0; + + if (unlikely(!asoc->rwnd || asoc->rwnd_over)) + pr_debug("%s: association:%p has asoc->rwnd:%u, " + "asoc->rwnd_over:%u!\n", __func__, asoc, + asoc->rwnd, asoc->rwnd_over); + + 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 potential 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", - __FUNCTION__, asoc, len, asoc->rwnd, - asoc->rwnd_over); + + pr_debug("%s: asoc:%p rwnd decreased by %d to (%u, %u, %u)\n", + __func__, asoc, len, asoc->rwnd, 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; if (asoc->peer.ipv6_address) flags |= SCTP_ADDR6_PEERSUPP; - return sctp_bind_addr_copy(&asoc->base.bind_addr, + return sctp_bind_addr_copy(sock_net(asoc->base.sk), + &asoc->base.bind_addr, &asoc->ep->base.bind_addr, scope, gfp, flags); } @@ -1264,22 +1575,119 @@ int sctp_assoc_set_bind_addr_from_cookie(struct sctp_association *asoc, asoc->ep->base.bind_addr.port, gfp); } -/* Lookup laddr in the bind address list of an association. */ -int sctp_assoc_lookup_laddr(struct sctp_association *asoc, +/* Lookup laddr in the bind address list of an association. */ +int sctp_assoc_lookup_laddr(struct sctp_association *asoc, const union sctp_addr *laddr) { - int found; + int found = 0; - sctp_read_lock(&asoc->base.addr_lock); if ((asoc->base.bind_addr.port == ntohs(laddr->v4.sin_port)) && sctp_bind_addr_match(&asoc->base.bind_addr, laddr, - sctp_sk(asoc->base.sk))) { + sctp_sk(asoc->base.sk))) found = 1; - goto out; - } - found = 0; -out: - sctp_read_unlock(&asoc->base.addr_lock); return found; } + +/* Set an association id for a given association */ +int sctp_assoc_set_id(struct sctp_association *asoc, gfp_t gfp) +{ + bool preload = !!(gfp & __GFP_WAIT); + int ret; + + /* If the id is already assigned, keep it. */ + if (asoc->assoc_id) + return 0; + + if (preload) + idr_preload(gfp); + spin_lock_bh(&sctp_assocs_id_lock); + /* 0 is not a valid assoc_id, must be >= 1 */ + ret = idr_alloc_cyclic(&sctp_assocs_id, asoc, 1, 0, GFP_NOWAIT); + spin_unlock_bh(&sctp_assocs_id_lock); + if (preload) + idr_preload_end(); + if (ret < 0) + return ret; + + asoc->assoc_id = (sctp_assoc_t)ret; + return 0; +} + +/* Free the ASCONF queue */ +static void sctp_assoc_free_asconf_queue(struct sctp_association *asoc) +{ + struct sctp_chunk *asconf; + struct sctp_chunk *tmp; + + list_for_each_entry_safe(asconf, tmp, &asoc->addip_chunk_list, list) { + list_del_init(&asconf->list); + sctp_chunk_free(asconf); + } +} + +/* Free asconf_ack cache */ +static void sctp_assoc_free_asconf_acks(struct sctp_association *asoc) +{ + struct sctp_chunk *ack; + struct sctp_chunk *tmp; + + list_for_each_entry_safe(ack, tmp, &asoc->asconf_ack_list, + transmitted_list) { + list_del_init(&ack->transmitted_list); + sctp_chunk_free(ack); + } +} + +/* Clean up the ASCONF_ACK queue */ +void sctp_assoc_clean_asconf_ack_cache(const struct sctp_association *asoc) +{ + struct sctp_chunk *ack; + struct sctp_chunk *tmp; + + /* We can remove all the entries from the queue up to + * the "Peer-Sequence-Number". + */ + list_for_each_entry_safe(ack, tmp, &asoc->asconf_ack_list, + transmitted_list) { + if (ack->subh.addip_hdr->serial == + htonl(asoc->peer.addip_serial)) + break; + + list_del_init(&ack->transmitted_list); + sctp_chunk_free(ack); + } +} + +/* Find the ASCONF_ACK whose serial number matches ASCONF */ +struct sctp_chunk *sctp_assoc_lookup_asconf_ack( + const struct sctp_association *asoc, + __be32 serial) +{ + struct sctp_chunk *ack; + + /* Walk through the list of cached ASCONF-ACKs and find the + * ack chunk whose serial number matches that of the request. + */ + list_for_each_entry(ack, &asoc->asconf_ack_list, transmitted_list) { + if (ack->subh.addip_hdr->serial == serial) { + sctp_chunk_hold(ack); + return ack; + } + } + + return NULL; +} + +void sctp_asconf_queue_teardown(struct sctp_association *asoc) +{ + /* Free any cached ASCONF_ACK chunk. */ + sctp_assoc_free_asconf_acks(asoc); + + /* Free the ASCONF queue. */ + sctp_assoc_free_asconf_queue(asoc); + + /* Free any cached ASCONF chunk. */ + if (asoc->addip_last_asconf) + sctp_chunk_free(asoc->addip_last_asconf); +} |