2 * net/tipc/link.c: TIPC link code
4 * Copyright (c) 1996-2007, Ericsson AB
5 * Copyright (c) 2004-2007, Wind River Systems
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions are met:
11 * 1. Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
16 * 3. Neither the names of the copyright holders nor the names of its
17 * contributors may be used to endorse or promote products derived from
18 * this software without specific prior written permission.
20 * Alternatively, this software may be distributed under the terms of the
21 * GNU General Public License ("GPL") version 2 as published by the Free
22 * Software Foundation.
24 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
25 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
26 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
27 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
28 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
29 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
30 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
31 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
32 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
33 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
34 * POSSIBILITY OF SUCH DAMAGE.
40 #include "name_distr.h"
46 * Out-of-range value for link session numbers
49 #define INVALID_SESSION 0x10000
55 #define STARTING_EVT 856384768 /* link processing trigger */
56 #define TRAFFIC_MSG_EVT 560815u /* rx'd ??? */
57 #define TIMEOUT_EVT 560817u /* link timer expired */
60 * The following two 'message types' is really just implementation
61 * data conveniently stored in the message header.
62 * They must not be considered part of the protocol
68 * State value stored in 'exp_msg_count'
71 #define START_CHANGEOVER 100000u
74 * struct link_name - deconstructed link name
75 * @addr_local: network address of node at this end
76 * @if_local: name of interface at this end
77 * @addr_peer: network address of node at far end
78 * @if_peer: name of interface at far end
83 char if_local[TIPC_MAX_IF_NAME];
85 char if_peer[TIPC_MAX_IF_NAME];
88 static void link_handle_out_of_seq_msg(struct link *l_ptr,
90 static void link_recv_proto_msg(struct link *l_ptr, struct sk_buff *buf);
91 static int link_recv_changeover_msg(struct link **l_ptr, struct sk_buff **buf);
92 static void link_set_supervision_props(struct link *l_ptr, u32 tolerance);
93 static int link_send_sections_long(struct port *sender,
94 struct iovec const *msg_sect,
95 u32 num_sect, u32 destnode);
96 static void link_check_defragm_bufs(struct link *l_ptr);
97 static void link_state_event(struct link *l_ptr, u32 event);
98 static void link_reset_statistics(struct link *l_ptr);
99 static void link_print(struct link *l_ptr, const char *str);
100 static void link_start(struct link *l_ptr);
101 static int link_send_long_buf(struct link *l_ptr, struct sk_buff *buf);
104 * Simple link routines
107 static unsigned int align(unsigned int i)
109 return (i + 3) & ~3u;
112 static void link_init_max_pkt(struct link *l_ptr)
116 max_pkt = (l_ptr->b_ptr->publ.mtu & ~3);
117 if (max_pkt > MAX_MSG_SIZE)
118 max_pkt = MAX_MSG_SIZE;
120 l_ptr->max_pkt_target = max_pkt;
121 if (l_ptr->max_pkt_target < MAX_PKT_DEFAULT)
122 l_ptr->max_pkt = l_ptr->max_pkt_target;
124 l_ptr->max_pkt = MAX_PKT_DEFAULT;
126 l_ptr->max_pkt_probes = 0;
129 static u32 link_next_sent(struct link *l_ptr)
132 return msg_seqno(buf_msg(l_ptr->next_out));
133 return mod(l_ptr->next_out_no);
136 static u32 link_last_sent(struct link *l_ptr)
138 return mod(link_next_sent(l_ptr) - 1);
142 * Simple non-static link routines (i.e. referenced outside this file)
145 int tipc_link_is_up(struct link *l_ptr)
149 return link_working_working(l_ptr) || link_working_unknown(l_ptr);
152 int tipc_link_is_active(struct link *l_ptr)
154 return (l_ptr->owner->active_links[0] == l_ptr) ||
155 (l_ptr->owner->active_links[1] == l_ptr);
159 * link_name_validate - validate & (optionally) deconstruct link name
160 * @name - ptr to link name string
161 * @name_parts - ptr to area for link name components (or NULL if not needed)
163 * Returns 1 if link name is valid, otherwise 0.
166 static int link_name_validate(const char *name, struct link_name *name_parts)
168 char name_copy[TIPC_MAX_LINK_NAME];
174 u32 z_local, c_local, n_local;
175 u32 z_peer, c_peer, n_peer;
179 /* copy link name & ensure length is OK */
181 name_copy[TIPC_MAX_LINK_NAME - 1] = 0;
182 /* need above in case non-Posix strncpy() doesn't pad with nulls */
183 strncpy(name_copy, name, TIPC_MAX_LINK_NAME);
184 if (name_copy[TIPC_MAX_LINK_NAME - 1] != 0)
187 /* ensure all component parts of link name are present */
189 addr_local = name_copy;
190 if_local = strchr(addr_local, ':');
191 if (if_local == NULL)
194 addr_peer = strchr(if_local, '-');
195 if (addr_peer == NULL)
198 if_local_len = addr_peer - if_local;
199 if_peer = strchr(addr_peer, ':');
203 if_peer_len = strlen(if_peer) + 1;
205 /* validate component parts of link name */
207 if ((sscanf(addr_local, "%u.%u.%u%c",
208 &z_local, &c_local, &n_local, &dummy) != 3) ||
209 (sscanf(addr_peer, "%u.%u.%u%c",
210 &z_peer, &c_peer, &n_peer, &dummy) != 3) ||
211 (z_local > 255) || (c_local > 4095) || (n_local > 4095) ||
212 (z_peer > 255) || (c_peer > 4095) || (n_peer > 4095) ||
213 (if_local_len <= 1) || (if_local_len > TIPC_MAX_IF_NAME) ||
214 (if_peer_len <= 1) || (if_peer_len > TIPC_MAX_IF_NAME) ||
215 (strspn(if_local, tipc_alphabet) != (if_local_len - 1)) ||
216 (strspn(if_peer, tipc_alphabet) != (if_peer_len - 1)))
219 /* return link name components, if necessary */
222 name_parts->addr_local = tipc_addr(z_local, c_local, n_local);
223 strcpy(name_parts->if_local, if_local);
224 name_parts->addr_peer = tipc_addr(z_peer, c_peer, n_peer);
225 strcpy(name_parts->if_peer, if_peer);
231 * link_timeout - handle expiration of link timer
232 * @l_ptr: pointer to link
234 * This routine must not grab "tipc_net_lock" to avoid a potential deadlock conflict
235 * with tipc_link_delete(). (There is no risk that the node will be deleted by
236 * another thread because tipc_link_delete() always cancels the link timer before
237 * tipc_node_delete() is called.)
240 static void link_timeout(struct link *l_ptr)
242 tipc_node_lock(l_ptr->owner);
244 /* update counters used in statistical profiling of send traffic */
246 l_ptr->stats.accu_queue_sz += l_ptr->out_queue_size;
247 l_ptr->stats.queue_sz_counts++;
249 if (l_ptr->out_queue_size > l_ptr->stats.max_queue_sz)
250 l_ptr->stats.max_queue_sz = l_ptr->out_queue_size;
252 if (l_ptr->first_out) {
253 struct tipc_msg *msg = buf_msg(l_ptr->first_out);
254 u32 length = msg_size(msg);
256 if ((msg_user(msg) == MSG_FRAGMENTER) &&
257 (msg_type(msg) == FIRST_FRAGMENT)) {
258 length = msg_size(msg_get_wrapped(msg));
261 l_ptr->stats.msg_lengths_total += length;
262 l_ptr->stats.msg_length_counts++;
264 l_ptr->stats.msg_length_profile[0]++;
265 else if (length <= 256)
266 l_ptr->stats.msg_length_profile[1]++;
267 else if (length <= 1024)
268 l_ptr->stats.msg_length_profile[2]++;
269 else if (length <= 4096)
270 l_ptr->stats.msg_length_profile[3]++;
271 else if (length <= 16384)
272 l_ptr->stats.msg_length_profile[4]++;
273 else if (length <= 32768)
274 l_ptr->stats.msg_length_profile[5]++;
276 l_ptr->stats.msg_length_profile[6]++;
280 /* do all other link processing performed on a periodic basis */
282 link_check_defragm_bufs(l_ptr);
284 link_state_event(l_ptr, TIMEOUT_EVT);
287 tipc_link_push_queue(l_ptr);
289 tipc_node_unlock(l_ptr->owner);
292 static void link_set_timer(struct link *l_ptr, u32 time)
294 k_start_timer(&l_ptr->timer, time);
298 * tipc_link_create - create a new link
299 * @b_ptr: pointer to associated bearer
300 * @peer: network address of node at other end of link
301 * @media_addr: media address to use when sending messages over link
303 * Returns pointer to link.
306 struct link *tipc_link_create(struct bearer *b_ptr, const u32 peer,
307 const struct tipc_media_addr *media_addr)
310 struct tipc_msg *msg;
313 l_ptr = kzalloc(sizeof(*l_ptr), GFP_ATOMIC);
315 warn("Link creation failed, no memory\n");
320 if_name = strchr(b_ptr->publ.name, ':') + 1;
321 sprintf(l_ptr->name, "%u.%u.%u:%s-%u.%u.%u:",
322 tipc_zone(tipc_own_addr), tipc_cluster(tipc_own_addr),
323 tipc_node(tipc_own_addr),
325 tipc_zone(peer), tipc_cluster(peer), tipc_node(peer));
326 /* note: peer i/f is appended to link name by reset/activate */
327 memcpy(&l_ptr->media_addr, media_addr, sizeof(*media_addr));
328 l_ptr->checkpoint = 1;
329 l_ptr->b_ptr = b_ptr;
330 link_set_supervision_props(l_ptr, b_ptr->media->tolerance);
331 l_ptr->state = RESET_UNKNOWN;
333 l_ptr->pmsg = (struct tipc_msg *)&l_ptr->proto_msg;
335 tipc_msg_init(msg, LINK_PROTOCOL, RESET_MSG, INT_H_SIZE, l_ptr->addr);
336 msg_set_size(msg, sizeof(l_ptr->proto_msg));
337 msg_set_session(msg, (tipc_random & 0xffff));
338 msg_set_bearer_id(msg, b_ptr->identity);
339 strcpy((char *)msg_data(msg), if_name);
341 l_ptr->priority = b_ptr->priority;
342 tipc_link_set_queue_limits(l_ptr, b_ptr->media->window);
344 link_init_max_pkt(l_ptr);
346 l_ptr->next_out_no = 1;
347 INIT_LIST_HEAD(&l_ptr->waiting_ports);
349 link_reset_statistics(l_ptr);
351 l_ptr->owner = tipc_node_attach_link(l_ptr);
357 k_init_timer(&l_ptr->timer, (Handler)link_timeout, (unsigned long)l_ptr);
358 list_add_tail(&l_ptr->link_list, &b_ptr->links);
359 tipc_k_signal((Handler)link_start, (unsigned long)l_ptr);
365 * tipc_link_delete - delete a link
366 * @l_ptr: pointer to link
368 * Note: 'tipc_net_lock' is write_locked, bearer is locked.
369 * This routine must not grab the node lock until after link timer cancellation
370 * to avoid a potential deadlock situation.
373 void tipc_link_delete(struct link *l_ptr)
376 err("Attempt to delete non-existent link\n");
380 k_cancel_timer(&l_ptr->timer);
382 tipc_node_lock(l_ptr->owner);
383 tipc_link_reset(l_ptr);
384 tipc_node_detach_link(l_ptr->owner, l_ptr);
385 tipc_link_stop(l_ptr);
386 list_del_init(&l_ptr->link_list);
387 tipc_node_unlock(l_ptr->owner);
388 k_term_timer(&l_ptr->timer);
392 static void link_start(struct link *l_ptr)
394 link_state_event(l_ptr, STARTING_EVT);
398 * link_schedule_port - schedule port for deferred sending
399 * @l_ptr: pointer to link
400 * @origport: reference to sending port
401 * @sz: amount of data to be sent
403 * Schedules port for renewed sending of messages after link congestion
407 static int link_schedule_port(struct link *l_ptr, u32 origport, u32 sz)
411 spin_lock_bh(&tipc_port_list_lock);
412 p_ptr = tipc_port_lock(origport);
416 if (!list_empty(&p_ptr->wait_list))
418 p_ptr->publ.congested = 1;
419 p_ptr->waiting_pkts = 1 + ((sz - 1) / l_ptr->max_pkt);
420 list_add_tail(&p_ptr->wait_list, &l_ptr->waiting_ports);
421 l_ptr->stats.link_congs++;
423 tipc_port_unlock(p_ptr);
425 spin_unlock_bh(&tipc_port_list_lock);
429 void tipc_link_wakeup_ports(struct link *l_ptr, int all)
432 struct port *temp_p_ptr;
433 int win = l_ptr->queue_limit[0] - l_ptr->out_queue_size;
439 if (!spin_trylock_bh(&tipc_port_list_lock))
441 if (link_congested(l_ptr))
443 list_for_each_entry_safe(p_ptr, temp_p_ptr, &l_ptr->waiting_ports,
447 list_del_init(&p_ptr->wait_list);
448 spin_lock_bh(p_ptr->publ.lock);
449 p_ptr->publ.congested = 0;
450 p_ptr->wakeup(&p_ptr->publ);
451 win -= p_ptr->waiting_pkts;
452 spin_unlock_bh(p_ptr->publ.lock);
456 spin_unlock_bh(&tipc_port_list_lock);
460 * link_release_outqueue - purge link's outbound message queue
461 * @l_ptr: pointer to link
464 static void link_release_outqueue(struct link *l_ptr)
466 struct sk_buff *buf = l_ptr->first_out;
467 struct sk_buff *next;
474 l_ptr->first_out = NULL;
475 l_ptr->out_queue_size = 0;
479 * tipc_link_reset_fragments - purge link's inbound message fragments queue
480 * @l_ptr: pointer to link
483 void tipc_link_reset_fragments(struct link *l_ptr)
485 struct sk_buff *buf = l_ptr->defragm_buf;
486 struct sk_buff *next;
493 l_ptr->defragm_buf = NULL;
497 * tipc_link_stop - purge all inbound and outbound messages associated with link
498 * @l_ptr: pointer to link
501 void tipc_link_stop(struct link *l_ptr)
504 struct sk_buff *next;
506 buf = l_ptr->oldest_deferred_in;
513 buf = l_ptr->first_out;
520 tipc_link_reset_fragments(l_ptr);
522 buf_discard(l_ptr->proto_msg_queue);
523 l_ptr->proto_msg_queue = NULL;
526 /* LINK EVENT CODE IS NOT SUPPORTED AT PRESENT */
527 #define link_send_event(fcn, l_ptr, up) do { } while (0)
529 void tipc_link_reset(struct link *l_ptr)
532 u32 prev_state = l_ptr->state;
533 u32 checkpoint = l_ptr->next_in_no;
534 int was_active_link = tipc_link_is_active(l_ptr);
536 msg_set_session(l_ptr->pmsg, ((msg_session(l_ptr->pmsg) + 1) & 0xffff));
538 /* Link is down, accept any session */
539 l_ptr->peer_session = INVALID_SESSION;
541 /* Prepare for max packet size negotiation */
542 link_init_max_pkt(l_ptr);
544 l_ptr->state = RESET_UNKNOWN;
546 if ((prev_state == RESET_UNKNOWN) || (prev_state == RESET_RESET))
549 tipc_node_link_down(l_ptr->owner, l_ptr);
550 tipc_bearer_remove_dest(l_ptr->b_ptr, l_ptr->addr);
552 if (was_active_link && tipc_node_has_active_links(l_ptr->owner) &&
553 l_ptr->owner->permit_changeover) {
554 l_ptr->reset_checkpoint = checkpoint;
555 l_ptr->exp_msg_count = START_CHANGEOVER;
558 /* Clean up all queues: */
560 link_release_outqueue(l_ptr);
561 buf_discard(l_ptr->proto_msg_queue);
562 l_ptr->proto_msg_queue = NULL;
563 buf = l_ptr->oldest_deferred_in;
565 struct sk_buff *next = buf->next;
569 if (!list_empty(&l_ptr->waiting_ports))
570 tipc_link_wakeup_ports(l_ptr, 1);
572 l_ptr->retransm_queue_head = 0;
573 l_ptr->retransm_queue_size = 0;
574 l_ptr->last_out = NULL;
575 l_ptr->first_out = NULL;
576 l_ptr->next_out = NULL;
577 l_ptr->unacked_window = 0;
578 l_ptr->checkpoint = 1;
579 l_ptr->next_out_no = 1;
580 l_ptr->deferred_inqueue_sz = 0;
581 l_ptr->oldest_deferred_in = NULL;
582 l_ptr->newest_deferred_in = NULL;
583 l_ptr->fsm_msg_cnt = 0;
584 l_ptr->stale_count = 0;
585 link_reset_statistics(l_ptr);
587 link_send_event(tipc_cfg_link_event, l_ptr, 0);
588 if (!in_own_cluster(l_ptr->addr))
589 link_send_event(tipc_disc_link_event, l_ptr, 0);
593 static void link_activate(struct link *l_ptr)
595 l_ptr->next_in_no = l_ptr->stats.recv_info = 1;
596 tipc_node_link_up(l_ptr->owner, l_ptr);
597 tipc_bearer_add_dest(l_ptr->b_ptr, l_ptr->addr);
598 link_send_event(tipc_cfg_link_event, l_ptr, 1);
599 if (!in_own_cluster(l_ptr->addr))
600 link_send_event(tipc_disc_link_event, l_ptr, 1);
604 * link_state_event - link finite state machine
605 * @l_ptr: pointer to link
606 * @event: state machine event to process
609 static void link_state_event(struct link *l_ptr, unsigned event)
612 u32 cont_intv = l_ptr->continuity_interval;
614 if (!l_ptr->started && (event != STARTING_EVT))
615 return; /* Not yet. */
617 if (link_blocked(l_ptr)) {
618 if (event == TIMEOUT_EVT)
619 link_set_timer(l_ptr, cont_intv);
620 return; /* Changeover going on */
623 switch (l_ptr->state) {
624 case WORKING_WORKING:
626 case TRAFFIC_MSG_EVT:
630 if (l_ptr->next_in_no != l_ptr->checkpoint) {
631 l_ptr->checkpoint = l_ptr->next_in_no;
632 if (tipc_bclink_acks_missing(l_ptr->owner)) {
633 tipc_link_send_proto_msg(l_ptr, STATE_MSG,
635 l_ptr->fsm_msg_cnt++;
636 } else if (l_ptr->max_pkt < l_ptr->max_pkt_target) {
637 tipc_link_send_proto_msg(l_ptr, STATE_MSG,
639 l_ptr->fsm_msg_cnt++;
641 link_set_timer(l_ptr, cont_intv);
644 l_ptr->state = WORKING_UNKNOWN;
645 l_ptr->fsm_msg_cnt = 0;
646 tipc_link_send_proto_msg(l_ptr, STATE_MSG, 1, 0, 0, 0, 0);
647 l_ptr->fsm_msg_cnt++;
648 link_set_timer(l_ptr, cont_intv / 4);
651 info("Resetting link <%s>, requested by peer\n",
653 tipc_link_reset(l_ptr);
654 l_ptr->state = RESET_RESET;
655 l_ptr->fsm_msg_cnt = 0;
656 tipc_link_send_proto_msg(l_ptr, ACTIVATE_MSG, 0, 0, 0, 0, 0);
657 l_ptr->fsm_msg_cnt++;
658 link_set_timer(l_ptr, cont_intv);
661 err("Unknown link event %u in WW state\n", event);
664 case WORKING_UNKNOWN:
666 case TRAFFIC_MSG_EVT:
668 l_ptr->state = WORKING_WORKING;
669 l_ptr->fsm_msg_cnt = 0;
670 link_set_timer(l_ptr, cont_intv);
673 info("Resetting link <%s>, requested by peer "
674 "while probing\n", l_ptr->name);
675 tipc_link_reset(l_ptr);
676 l_ptr->state = RESET_RESET;
677 l_ptr->fsm_msg_cnt = 0;
678 tipc_link_send_proto_msg(l_ptr, ACTIVATE_MSG, 0, 0, 0, 0, 0);
679 l_ptr->fsm_msg_cnt++;
680 link_set_timer(l_ptr, cont_intv);
683 if (l_ptr->next_in_no != l_ptr->checkpoint) {
684 l_ptr->state = WORKING_WORKING;
685 l_ptr->fsm_msg_cnt = 0;
686 l_ptr->checkpoint = l_ptr->next_in_no;
687 if (tipc_bclink_acks_missing(l_ptr->owner)) {
688 tipc_link_send_proto_msg(l_ptr, STATE_MSG,
690 l_ptr->fsm_msg_cnt++;
692 link_set_timer(l_ptr, cont_intv);
693 } else if (l_ptr->fsm_msg_cnt < l_ptr->abort_limit) {
694 tipc_link_send_proto_msg(l_ptr, STATE_MSG,
696 l_ptr->fsm_msg_cnt++;
697 link_set_timer(l_ptr, cont_intv / 4);
698 } else { /* Link has failed */
699 warn("Resetting link <%s>, peer not responding\n",
701 tipc_link_reset(l_ptr);
702 l_ptr->state = RESET_UNKNOWN;
703 l_ptr->fsm_msg_cnt = 0;
704 tipc_link_send_proto_msg(l_ptr, RESET_MSG,
706 l_ptr->fsm_msg_cnt++;
707 link_set_timer(l_ptr, cont_intv);
711 err("Unknown link event %u in WU state\n", event);
716 case TRAFFIC_MSG_EVT:
719 other = l_ptr->owner->active_links[0];
720 if (other && link_working_unknown(other))
722 l_ptr->state = WORKING_WORKING;
723 l_ptr->fsm_msg_cnt = 0;
724 link_activate(l_ptr);
725 tipc_link_send_proto_msg(l_ptr, STATE_MSG, 1, 0, 0, 0, 0);
726 l_ptr->fsm_msg_cnt++;
727 link_set_timer(l_ptr, cont_intv);
730 l_ptr->state = RESET_RESET;
731 l_ptr->fsm_msg_cnt = 0;
732 tipc_link_send_proto_msg(l_ptr, ACTIVATE_MSG, 1, 0, 0, 0, 0);
733 l_ptr->fsm_msg_cnt++;
734 link_set_timer(l_ptr, cont_intv);
740 tipc_link_send_proto_msg(l_ptr, RESET_MSG, 0, 0, 0, 0, 0);
741 l_ptr->fsm_msg_cnt++;
742 link_set_timer(l_ptr, cont_intv);
745 err("Unknown link event %u in RU state\n", event);
750 case TRAFFIC_MSG_EVT:
752 other = l_ptr->owner->active_links[0];
753 if (other && link_working_unknown(other))
755 l_ptr->state = WORKING_WORKING;
756 l_ptr->fsm_msg_cnt = 0;
757 link_activate(l_ptr);
758 tipc_link_send_proto_msg(l_ptr, STATE_MSG, 1, 0, 0, 0, 0);
759 l_ptr->fsm_msg_cnt++;
760 link_set_timer(l_ptr, cont_intv);
765 tipc_link_send_proto_msg(l_ptr, ACTIVATE_MSG, 0, 0, 0, 0, 0);
766 l_ptr->fsm_msg_cnt++;
767 link_set_timer(l_ptr, cont_intv);
770 err("Unknown link event %u in RR state\n", event);
774 err("Unknown link state %u/%u\n", l_ptr->state, event);
779 * link_bundle_buf(): Append contents of a buffer to
780 * the tail of an existing one.
783 static int link_bundle_buf(struct link *l_ptr,
784 struct sk_buff *bundler,
787 struct tipc_msg *bundler_msg = buf_msg(bundler);
788 struct tipc_msg *msg = buf_msg(buf);
789 u32 size = msg_size(msg);
790 u32 bundle_size = msg_size(bundler_msg);
791 u32 to_pos = align(bundle_size);
792 u32 pad = to_pos - bundle_size;
794 if (msg_user(bundler_msg) != MSG_BUNDLER)
796 if (msg_type(bundler_msg) != OPEN_MSG)
798 if (skb_tailroom(bundler) < (pad + size))
800 if (l_ptr->max_pkt < (to_pos + size))
803 skb_put(bundler, pad + size);
804 skb_copy_to_linear_data_offset(bundler, to_pos, buf->data, size);
805 msg_set_size(bundler_msg, to_pos + size);
806 msg_set_msgcnt(bundler_msg, msg_msgcnt(bundler_msg) + 1);
808 l_ptr->stats.sent_bundled++;
812 static void link_add_to_outqueue(struct link *l_ptr,
814 struct tipc_msg *msg)
816 u32 ack = mod(l_ptr->next_in_no - 1);
817 u32 seqno = mod(l_ptr->next_out_no++);
819 msg_set_word(msg, 2, ((ack << 16) | seqno));
820 msg_set_bcast_ack(msg, l_ptr->owner->bclink.last_in);
822 if (l_ptr->first_out) {
823 l_ptr->last_out->next = buf;
824 l_ptr->last_out = buf;
826 l_ptr->first_out = l_ptr->last_out = buf;
827 l_ptr->out_queue_size++;
831 * tipc_link_send_buf() is the 'full path' for messages, called from
832 * inside TIPC when the 'fast path' in tipc_send_buf
833 * has failed, and from link_send()
836 int tipc_link_send_buf(struct link *l_ptr, struct sk_buff *buf)
838 struct tipc_msg *msg = buf_msg(buf);
839 u32 size = msg_size(msg);
840 u32 dsz = msg_data_sz(msg);
841 u32 queue_size = l_ptr->out_queue_size;
842 u32 imp = tipc_msg_tot_importance(msg);
843 u32 queue_limit = l_ptr->queue_limit[imp];
844 u32 max_packet = l_ptr->max_pkt;
846 msg_set_prevnode(msg, tipc_own_addr); /* If routed message */
848 /* Match msg importance against queue limits: */
850 if (unlikely(queue_size >= queue_limit)) {
851 if (imp <= TIPC_CRITICAL_IMPORTANCE) {
852 return link_schedule_port(l_ptr, msg_origport(msg),
856 if (imp > CONN_MANAGER) {
857 warn("Resetting link <%s>, send queue full", l_ptr->name);
858 tipc_link_reset(l_ptr);
863 /* Fragmentation needed ? */
865 if (size > max_packet)
866 return link_send_long_buf(l_ptr, buf);
868 /* Packet can be queued or sent: */
870 if (queue_size > l_ptr->stats.max_queue_sz)
871 l_ptr->stats.max_queue_sz = queue_size;
873 if (likely(!tipc_bearer_congested(l_ptr->b_ptr, l_ptr) &&
874 !link_congested(l_ptr))) {
875 link_add_to_outqueue(l_ptr, buf, msg);
877 if (likely(tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr))) {
878 l_ptr->unacked_window = 0;
880 tipc_bearer_schedule(l_ptr->b_ptr, l_ptr);
881 l_ptr->stats.bearer_congs++;
882 l_ptr->next_out = buf;
886 /* Congestion: can message be bundled ?: */
888 if ((msg_user(msg) != CHANGEOVER_PROTOCOL) &&
889 (msg_user(msg) != MSG_FRAGMENTER)) {
891 /* Try adding message to an existing bundle */
893 if (l_ptr->next_out &&
894 link_bundle_buf(l_ptr, l_ptr->last_out, buf)) {
895 tipc_bearer_resolve_congestion(l_ptr->b_ptr, l_ptr);
899 /* Try creating a new bundle */
901 if (size <= max_packet * 2 / 3) {
902 struct sk_buff *bundler = tipc_buf_acquire(max_packet);
903 struct tipc_msg bundler_hdr;
906 tipc_msg_init(&bundler_hdr, MSG_BUNDLER, OPEN_MSG,
907 INT_H_SIZE, l_ptr->addr);
908 skb_copy_to_linear_data(bundler, &bundler_hdr,
910 skb_trim(bundler, INT_H_SIZE);
911 link_bundle_buf(l_ptr, bundler, buf);
914 l_ptr->stats.sent_bundles++;
918 if (!l_ptr->next_out)
919 l_ptr->next_out = buf;
920 link_add_to_outqueue(l_ptr, buf, msg);
921 tipc_bearer_resolve_congestion(l_ptr->b_ptr, l_ptr);
926 * tipc_link_send(): same as tipc_link_send_buf(), but the link to use has
927 * not been selected yet, and the the owner node is not locked
928 * Called by TIPC internal users, e.g. the name distributor
931 int tipc_link_send(struct sk_buff *buf, u32 dest, u32 selector)
934 struct tipc_node *n_ptr;
935 int res = -ELINKCONG;
937 read_lock_bh(&tipc_net_lock);
938 n_ptr = tipc_node_find(dest);
940 tipc_node_lock(n_ptr);
941 l_ptr = n_ptr->active_links[selector & 1];
943 res = tipc_link_send_buf(l_ptr, buf);
946 tipc_node_unlock(n_ptr);
950 read_unlock_bh(&tipc_net_lock);
955 * link_send_buf_fast: Entry for data messages where the
956 * destination link is known and the header is complete,
957 * inclusive total message length. Very time critical.
958 * Link is locked. Returns user data length.
961 static int link_send_buf_fast(struct link *l_ptr, struct sk_buff *buf,
964 struct tipc_msg *msg = buf_msg(buf);
965 int res = msg_data_sz(msg);
967 if (likely(!link_congested(l_ptr))) {
968 if (likely(msg_size(msg) <= l_ptr->max_pkt)) {
969 if (likely(list_empty(&l_ptr->b_ptr->cong_links))) {
970 link_add_to_outqueue(l_ptr, buf, msg);
971 if (likely(tipc_bearer_send(l_ptr->b_ptr, buf,
972 &l_ptr->media_addr))) {
973 l_ptr->unacked_window = 0;
976 tipc_bearer_schedule(l_ptr->b_ptr, l_ptr);
977 l_ptr->stats.bearer_congs++;
978 l_ptr->next_out = buf;
982 *used_max_pkt = l_ptr->max_pkt;
984 return tipc_link_send_buf(l_ptr, buf); /* All other cases */
988 * tipc_send_buf_fast: Entry for data messages where the
989 * destination node is known and the header is complete,
990 * inclusive total message length.
991 * Returns user data length.
993 int tipc_send_buf_fast(struct sk_buff *buf, u32 destnode)
996 struct tipc_node *n_ptr;
998 u32 selector = msg_origport(buf_msg(buf)) & 1;
1001 if (destnode == tipc_own_addr)
1002 return tipc_port_recv_msg(buf);
1004 read_lock_bh(&tipc_net_lock);
1005 n_ptr = tipc_node_find(destnode);
1006 if (likely(n_ptr)) {
1007 tipc_node_lock(n_ptr);
1008 l_ptr = n_ptr->active_links[selector];
1009 if (likely(l_ptr)) {
1010 res = link_send_buf_fast(l_ptr, buf, &dummy);
1011 tipc_node_unlock(n_ptr);
1012 read_unlock_bh(&tipc_net_lock);
1015 tipc_node_unlock(n_ptr);
1017 read_unlock_bh(&tipc_net_lock);
1018 res = msg_data_sz(buf_msg(buf));
1019 tipc_reject_msg(buf, TIPC_ERR_NO_NODE);
1025 * tipc_link_send_sections_fast: Entry for messages where the
1026 * destination processor is known and the header is complete,
1027 * except for total message length.
1028 * Returns user data length or errno.
1030 int tipc_link_send_sections_fast(struct port *sender,
1031 struct iovec const *msg_sect,
1035 struct tipc_msg *hdr = &sender->publ.phdr;
1037 struct sk_buff *buf;
1038 struct tipc_node *node;
1040 u32 selector = msg_origport(hdr) & 1;
1044 * Try building message using port's max_pkt hint.
1045 * (Must not hold any locks while building message.)
1048 res = tipc_msg_build(hdr, msg_sect, num_sect, sender->publ.max_pkt,
1049 !sender->user_port, &buf);
1051 read_lock_bh(&tipc_net_lock);
1052 node = tipc_node_find(destaddr);
1054 tipc_node_lock(node);
1055 l_ptr = node->active_links[selector];
1056 if (likely(l_ptr)) {
1058 res = link_send_buf_fast(l_ptr, buf,
1059 &sender->publ.max_pkt);
1060 if (unlikely(res < 0))
1063 tipc_node_unlock(node);
1064 read_unlock_bh(&tipc_net_lock);
1068 /* Exit if build request was invalid */
1070 if (unlikely(res < 0))
1073 /* Exit if link (or bearer) is congested */
1075 if (link_congested(l_ptr) ||
1076 !list_empty(&l_ptr->b_ptr->cong_links)) {
1077 res = link_schedule_port(l_ptr,
1078 sender->publ.ref, res);
1083 * Message size exceeds max_pkt hint; update hint,
1084 * then re-try fast path or fragment the message
1087 sender->publ.max_pkt = l_ptr->max_pkt;
1088 tipc_node_unlock(node);
1089 read_unlock_bh(&tipc_net_lock);
1092 if ((msg_hdr_sz(hdr) + res) <= sender->publ.max_pkt)
1095 return link_send_sections_long(sender, msg_sect,
1096 num_sect, destaddr);
1098 tipc_node_unlock(node);
1100 read_unlock_bh(&tipc_net_lock);
1102 /* Couldn't find a link to the destination node */
1105 return tipc_reject_msg(buf, TIPC_ERR_NO_NODE);
1107 return tipc_port_reject_sections(sender, hdr, msg_sect, num_sect,
1113 * link_send_sections_long(): Entry for long messages where the
1114 * destination node is known and the header is complete,
1115 * inclusive total message length.
1116 * Link and bearer congestion status have been checked to be ok,
1117 * and are ignored if they change.
1119 * Note that fragments do not use the full link MTU so that they won't have
1120 * to undergo refragmentation if link changeover causes them to be sent
1121 * over another link with an additional tunnel header added as prefix.
1122 * (Refragmentation will still occur if the other link has a smaller MTU.)
1124 * Returns user data length or errno.
1126 static int link_send_sections_long(struct port *sender,
1127 struct iovec const *msg_sect,
1132 struct tipc_node *node;
1133 struct tipc_msg *hdr = &sender->publ.phdr;
1134 u32 dsz = msg_data_sz(hdr);
1135 u32 max_pkt, fragm_sz, rest;
1136 struct tipc_msg fragm_hdr;
1137 struct sk_buff *buf, *buf_chain, *prev;
1138 u32 fragm_crs, fragm_rest, hsz, sect_rest;
1139 const unchar *sect_crs;
1145 max_pkt = sender->publ.max_pkt - INT_H_SIZE;
1146 /* leave room for tunnel header in case of link changeover */
1147 fragm_sz = max_pkt - INT_H_SIZE;
1148 /* leave room for fragmentation header in each fragment */
1156 /* Prepare reusable fragment header: */
1158 tipc_msg_init(&fragm_hdr, MSG_FRAGMENTER, FIRST_FRAGMENT,
1159 INT_H_SIZE, msg_destnode(hdr));
1160 msg_set_link_selector(&fragm_hdr, sender->publ.ref);
1161 msg_set_size(&fragm_hdr, max_pkt);
1162 msg_set_fragm_no(&fragm_hdr, 1);
1164 /* Prepare header of first fragment: */
1166 buf_chain = buf = tipc_buf_acquire(max_pkt);
1170 skb_copy_to_linear_data(buf, &fragm_hdr, INT_H_SIZE);
1171 hsz = msg_hdr_sz(hdr);
1172 skb_copy_to_linear_data_offset(buf, INT_H_SIZE, hdr, hsz);
1174 /* Chop up message: */
1176 fragm_crs = INT_H_SIZE + hsz;
1177 fragm_rest = fragm_sz - hsz;
1179 do { /* For all sections */
1183 sect_rest = msg_sect[++curr_sect].iov_len;
1184 sect_crs = (const unchar *)msg_sect[curr_sect].iov_base;
1187 if (sect_rest < fragm_rest)
1192 if (likely(!sender->user_port)) {
1193 if (copy_from_user(buf->data + fragm_crs, sect_crs, sz)) {
1195 for (; buf_chain; buf_chain = buf) {
1196 buf = buf_chain->next;
1197 buf_discard(buf_chain);
1202 skb_copy_to_linear_data_offset(buf, fragm_crs,
1210 if (!fragm_rest && rest) {
1212 /* Initiate new fragment: */
1213 if (rest <= fragm_sz) {
1215 msg_set_type(&fragm_hdr, LAST_FRAGMENT);
1217 msg_set_type(&fragm_hdr, FRAGMENT);
1219 msg_set_size(&fragm_hdr, fragm_sz + INT_H_SIZE);
1220 msg_set_fragm_no(&fragm_hdr, ++fragm_no);
1222 buf = tipc_buf_acquire(fragm_sz + INT_H_SIZE);
1228 skb_copy_to_linear_data(buf, &fragm_hdr, INT_H_SIZE);
1229 fragm_crs = INT_H_SIZE;
1230 fragm_rest = fragm_sz;
1235 * Now we have a buffer chain. Select a link and check
1236 * that packet size is still OK
1238 node = tipc_node_find(destaddr);
1240 tipc_node_lock(node);
1241 l_ptr = node->active_links[sender->publ.ref & 1];
1243 tipc_node_unlock(node);
1246 if (l_ptr->max_pkt < max_pkt) {
1247 sender->publ.max_pkt = l_ptr->max_pkt;
1248 tipc_node_unlock(node);
1249 for (; buf_chain; buf_chain = buf) {
1250 buf = buf_chain->next;
1251 buf_discard(buf_chain);
1257 for (; buf_chain; buf_chain = buf) {
1258 buf = buf_chain->next;
1259 buf_discard(buf_chain);
1261 return tipc_port_reject_sections(sender, hdr, msg_sect, num_sect,
1265 /* Append whole chain to send queue: */
1268 l_ptr->long_msg_seq_no = mod(l_ptr->long_msg_seq_no + 1);
1269 if (!l_ptr->next_out)
1270 l_ptr->next_out = buf_chain;
1271 l_ptr->stats.sent_fragmented++;
1273 struct sk_buff *next = buf->next;
1274 struct tipc_msg *msg = buf_msg(buf);
1276 l_ptr->stats.sent_fragments++;
1277 msg_set_long_msgno(msg, l_ptr->long_msg_seq_no);
1278 link_add_to_outqueue(l_ptr, buf, msg);
1282 /* Send it, if possible: */
1284 tipc_link_push_queue(l_ptr);
1285 tipc_node_unlock(node);
1290 * tipc_link_push_packet: Push one unsent packet to the media
1292 u32 tipc_link_push_packet(struct link *l_ptr)
1294 struct sk_buff *buf = l_ptr->first_out;
1295 u32 r_q_size = l_ptr->retransm_queue_size;
1296 u32 r_q_head = l_ptr->retransm_queue_head;
1298 /* Step to position where retransmission failed, if any, */
1299 /* consider that buffers may have been released in meantime */
1301 if (r_q_size && buf) {
1302 u32 last = lesser(mod(r_q_head + r_q_size),
1303 link_last_sent(l_ptr));
1304 u32 first = msg_seqno(buf_msg(buf));
1306 while (buf && less(first, r_q_head)) {
1307 first = mod(first + 1);
1310 l_ptr->retransm_queue_head = r_q_head = first;
1311 l_ptr->retransm_queue_size = r_q_size = mod(last - first);
1314 /* Continue retransmission now, if there is anything: */
1316 if (r_q_size && buf) {
1317 msg_set_ack(buf_msg(buf), mod(l_ptr->next_in_no - 1));
1318 msg_set_bcast_ack(buf_msg(buf), l_ptr->owner->bclink.last_in);
1319 if (tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr)) {
1320 l_ptr->retransm_queue_head = mod(++r_q_head);
1321 l_ptr->retransm_queue_size = --r_q_size;
1322 l_ptr->stats.retransmitted++;
1325 l_ptr->stats.bearer_congs++;
1330 /* Send deferred protocol message, if any: */
1332 buf = l_ptr->proto_msg_queue;
1334 msg_set_ack(buf_msg(buf), mod(l_ptr->next_in_no - 1));
1335 msg_set_bcast_ack(buf_msg(buf), l_ptr->owner->bclink.last_in);
1336 if (tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr)) {
1337 l_ptr->unacked_window = 0;
1339 l_ptr->proto_msg_queue = NULL;
1342 l_ptr->stats.bearer_congs++;
1347 /* Send one deferred data message, if send window not full: */
1349 buf = l_ptr->next_out;
1351 struct tipc_msg *msg = buf_msg(buf);
1352 u32 next = msg_seqno(msg);
1353 u32 first = msg_seqno(buf_msg(l_ptr->first_out));
1355 if (mod(next - first) < l_ptr->queue_limit[0]) {
1356 msg_set_ack(msg, mod(l_ptr->next_in_no - 1));
1357 msg_set_bcast_ack(msg, l_ptr->owner->bclink.last_in);
1358 if (tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr)) {
1359 if (msg_user(msg) == MSG_BUNDLER)
1360 msg_set_type(msg, CLOSED_MSG);
1361 l_ptr->next_out = buf->next;
1364 l_ptr->stats.bearer_congs++;
1369 return PUSH_FINISHED;
1373 * push_queue(): push out the unsent messages of a link where
1374 * congestion has abated. Node is locked
1376 void tipc_link_push_queue(struct link *l_ptr)
1380 if (tipc_bearer_congested(l_ptr->b_ptr, l_ptr))
1384 res = tipc_link_push_packet(l_ptr);
1387 if (res == PUSH_FAILED)
1388 tipc_bearer_schedule(l_ptr->b_ptr, l_ptr);
1391 static void link_reset_all(unsigned long addr)
1393 struct tipc_node *n_ptr;
1394 char addr_string[16];
1397 read_lock_bh(&tipc_net_lock);
1398 n_ptr = tipc_node_find((u32)addr);
1400 read_unlock_bh(&tipc_net_lock);
1401 return; /* node no longer exists */
1404 tipc_node_lock(n_ptr);
1406 warn("Resetting all links to %s\n",
1407 tipc_addr_string_fill(addr_string, n_ptr->addr));
1409 for (i = 0; i < MAX_BEARERS; i++) {
1410 if (n_ptr->links[i]) {
1411 link_print(n_ptr->links[i], "Resetting link\n");
1412 tipc_link_reset(n_ptr->links[i]);
1416 tipc_node_unlock(n_ptr);
1417 read_unlock_bh(&tipc_net_lock);
1420 static void link_retransmit_failure(struct link *l_ptr, struct sk_buff *buf)
1422 struct tipc_msg *msg = buf_msg(buf);
1424 warn("Retransmission failure on link <%s>\n", l_ptr->name);
1428 /* Handle failure on standard link */
1430 link_print(l_ptr, "Resetting link\n");
1431 tipc_link_reset(l_ptr);
1435 /* Handle failure on broadcast link */
1437 struct tipc_node *n_ptr;
1438 char addr_string[16];
1440 info("Msg seq number: %u, ", msg_seqno(msg));
1441 info("Outstanding acks: %lu\n",
1442 (unsigned long) TIPC_SKB_CB(buf)->handle);
1444 n_ptr = l_ptr->owner->next;
1445 tipc_node_lock(n_ptr);
1447 tipc_addr_string_fill(addr_string, n_ptr->addr);
1448 info("Multicast link info for %s\n", addr_string);
1449 info("Supported: %d, ", n_ptr->bclink.supported);
1450 info("Acked: %u\n", n_ptr->bclink.acked);
1451 info("Last in: %u, ", n_ptr->bclink.last_in);
1452 info("Gap after: %u, ", n_ptr->bclink.gap_after);
1453 info("Gap to: %u\n", n_ptr->bclink.gap_to);
1454 info("Nack sync: %u\n\n", n_ptr->bclink.nack_sync);
1456 tipc_k_signal((Handler)link_reset_all, (unsigned long)n_ptr->addr);
1458 tipc_node_unlock(n_ptr);
1460 l_ptr->stale_count = 0;
1464 void tipc_link_retransmit(struct link *l_ptr, struct sk_buff *buf,
1467 struct tipc_msg *msg;
1474 if (tipc_bearer_congested(l_ptr->b_ptr, l_ptr)) {
1475 if (l_ptr->retransm_queue_size == 0) {
1476 l_ptr->retransm_queue_head = msg_seqno(msg);
1477 l_ptr->retransm_queue_size = retransmits;
1479 err("Unexpected retransmit on link %s (qsize=%d)\n",
1480 l_ptr->name, l_ptr->retransm_queue_size);
1484 /* Detect repeated retransmit failures on uncongested bearer */
1486 if (l_ptr->last_retransmitted == msg_seqno(msg)) {
1487 if (++l_ptr->stale_count > 100) {
1488 link_retransmit_failure(l_ptr, buf);
1492 l_ptr->last_retransmitted = msg_seqno(msg);
1493 l_ptr->stale_count = 1;
1497 while (retransmits && (buf != l_ptr->next_out) && buf) {
1499 msg_set_ack(msg, mod(l_ptr->next_in_no - 1));
1500 msg_set_bcast_ack(msg, l_ptr->owner->bclink.last_in);
1501 if (tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr)) {
1504 l_ptr->stats.retransmitted++;
1506 tipc_bearer_schedule(l_ptr->b_ptr, l_ptr);
1507 l_ptr->stats.bearer_congs++;
1508 l_ptr->retransm_queue_head = msg_seqno(buf_msg(buf));
1509 l_ptr->retransm_queue_size = retransmits;
1514 l_ptr->retransm_queue_head = l_ptr->retransm_queue_size = 0;
1518 * link_insert_deferred_queue - insert deferred messages back into receive chain
1521 static struct sk_buff *link_insert_deferred_queue(struct link *l_ptr,
1522 struct sk_buff *buf)
1526 if (l_ptr->oldest_deferred_in == NULL)
1529 seq_no = msg_seqno(buf_msg(l_ptr->oldest_deferred_in));
1530 if (seq_no == mod(l_ptr->next_in_no)) {
1531 l_ptr->newest_deferred_in->next = buf;
1532 buf = l_ptr->oldest_deferred_in;
1533 l_ptr->oldest_deferred_in = NULL;
1534 l_ptr->deferred_inqueue_sz = 0;
1540 * link_recv_buf_validate - validate basic format of received message
1542 * This routine ensures a TIPC message has an acceptable header, and at least
1543 * as much data as the header indicates it should. The routine also ensures
1544 * that the entire message header is stored in the main fragment of the message
1545 * buffer, to simplify future access to message header fields.
1547 * Note: Having extra info present in the message header or data areas is OK.
1548 * TIPC will ignore the excess, under the assumption that it is optional info
1549 * introduced by a later release of the protocol.
1552 static int link_recv_buf_validate(struct sk_buff *buf)
1554 static u32 min_data_hdr_size[8] = {
1555 SHORT_H_SIZE, MCAST_H_SIZE, LONG_H_SIZE, DIR_MSG_H_SIZE,
1556 MAX_H_SIZE, MAX_H_SIZE, MAX_H_SIZE, MAX_H_SIZE
1559 struct tipc_msg *msg;
1565 if (unlikely(buf->len < MIN_H_SIZE))
1568 msg = skb_header_pointer(buf, 0, sizeof(tipc_hdr), tipc_hdr);
1572 if (unlikely(msg_version(msg) != TIPC_VERSION))
1575 size = msg_size(msg);
1576 hdr_size = msg_hdr_sz(msg);
1577 min_hdr_size = msg_isdata(msg) ?
1578 min_data_hdr_size[msg_type(msg)] : INT_H_SIZE;
1580 if (unlikely((hdr_size < min_hdr_size) ||
1581 (size < hdr_size) ||
1582 (buf->len < size) ||
1583 (size - hdr_size > TIPC_MAX_USER_MSG_SIZE)))
1586 return pskb_may_pull(buf, hdr_size);
1590 * tipc_recv_msg - process TIPC messages arriving from off-node
1591 * @head: pointer to message buffer chain
1592 * @tb_ptr: pointer to bearer message arrived on
1594 * Invoked with no locks held. Bearer pointer must point to a valid bearer
1595 * structure (i.e. cannot be NULL), but bearer can be inactive.
1598 void tipc_recv_msg(struct sk_buff *head, struct tipc_bearer *tb_ptr)
1600 read_lock_bh(&tipc_net_lock);
1602 struct bearer *b_ptr = (struct bearer *)tb_ptr;
1603 struct tipc_node *n_ptr;
1605 struct sk_buff *crs;
1606 struct sk_buff *buf = head;
1607 struct tipc_msg *msg;
1615 /* Ensure bearer is still enabled */
1617 if (unlikely(!b_ptr->active))
1620 /* Ensure message is well-formed */
1622 if (unlikely(!link_recv_buf_validate(buf)))
1625 /* Ensure message data is a single contiguous unit */
1627 if (unlikely(buf_linearize(buf)))
1630 /* Handle arrival of a non-unicast link message */
1634 if (unlikely(msg_non_seq(msg))) {
1635 if (msg_user(msg) == LINK_CONFIG)
1636 tipc_disc_recv_msg(buf, b_ptr);
1638 tipc_bclink_recv_pkt(buf);
1642 if (unlikely(!msg_short(msg) &&
1643 (msg_destnode(msg) != tipc_own_addr)))
1646 /* Discard non-routeable messages destined for another node */
1648 if (unlikely(!msg_isdata(msg) &&
1649 (msg_destnode(msg) != tipc_own_addr))) {
1650 if ((msg_user(msg) != CONN_MANAGER) &&
1651 (msg_user(msg) != MSG_FRAGMENTER))
1655 /* Locate neighboring node that sent message */
1657 n_ptr = tipc_node_find(msg_prevnode(msg));
1658 if (unlikely(!n_ptr))
1660 tipc_node_lock(n_ptr);
1662 /* Don't talk to neighbor during cleanup after last session */
1664 if (n_ptr->cleanup_required) {
1665 tipc_node_unlock(n_ptr);
1669 /* Locate unicast link endpoint that should handle message */
1671 l_ptr = n_ptr->links[b_ptr->identity];
1672 if (unlikely(!l_ptr)) {
1673 tipc_node_unlock(n_ptr);
1677 /* Validate message sequence number info */
1679 seq_no = msg_seqno(msg);
1680 ackd = msg_ack(msg);
1682 /* Release acked messages */
1684 if (less(n_ptr->bclink.acked, msg_bcast_ack(msg))) {
1685 if (tipc_node_is_up(n_ptr) && n_ptr->bclink.supported)
1686 tipc_bclink_acknowledge(n_ptr, msg_bcast_ack(msg));
1689 crs = l_ptr->first_out;
1690 while ((crs != l_ptr->next_out) &&
1691 less_eq(msg_seqno(buf_msg(crs)), ackd)) {
1692 struct sk_buff *next = crs->next;
1699 l_ptr->first_out = crs;
1700 l_ptr->out_queue_size -= released;
1703 /* Try sending any messages link endpoint has pending */
1705 if (unlikely(l_ptr->next_out))
1706 tipc_link_push_queue(l_ptr);
1707 if (unlikely(!list_empty(&l_ptr->waiting_ports)))
1708 tipc_link_wakeup_ports(l_ptr, 0);
1709 if (unlikely(++l_ptr->unacked_window >= TIPC_MIN_LINK_WIN)) {
1710 l_ptr->stats.sent_acks++;
1711 tipc_link_send_proto_msg(l_ptr, STATE_MSG, 0, 0, 0, 0, 0);
1714 /* Now (finally!) process the incoming message */
1717 if (likely(link_working_working(l_ptr))) {
1718 if (likely(seq_no == mod(l_ptr->next_in_no))) {
1719 l_ptr->next_in_no++;
1720 if (unlikely(l_ptr->oldest_deferred_in))
1721 head = link_insert_deferred_queue(l_ptr,
1723 if (likely(msg_is_dest(msg, tipc_own_addr))) {
1725 if (likely(msg_isdata(msg))) {
1726 tipc_node_unlock(n_ptr);
1727 tipc_port_recv_msg(buf);
1730 switch (msg_user(msg)) {
1732 l_ptr->stats.recv_bundles++;
1733 l_ptr->stats.recv_bundled +=
1735 tipc_node_unlock(n_ptr);
1736 tipc_link_recv_bundle(buf);
1738 case ROUTE_DISTRIBUTOR:
1739 tipc_node_unlock(n_ptr);
1742 case NAME_DISTRIBUTOR:
1743 tipc_node_unlock(n_ptr);
1744 tipc_named_recv(buf);
1747 tipc_node_unlock(n_ptr);
1748 tipc_port_recv_proto_msg(buf);
1750 case MSG_FRAGMENTER:
1751 l_ptr->stats.recv_fragments++;
1752 if (tipc_link_recv_fragment(&l_ptr->defragm_buf,
1754 l_ptr->stats.recv_fragmented++;
1758 case CHANGEOVER_PROTOCOL:
1759 type = msg_type(msg);
1760 if (link_recv_changeover_msg(&l_ptr, &buf)) {
1762 seq_no = msg_seqno(msg);
1763 if (type == ORIGINAL_MSG)
1765 goto protocol_check;
1770 tipc_node_unlock(n_ptr);
1771 tipc_net_route_msg(buf);
1774 link_handle_out_of_seq_msg(l_ptr, buf);
1775 head = link_insert_deferred_queue(l_ptr, head);
1776 tipc_node_unlock(n_ptr);
1780 if (msg_user(msg) == LINK_PROTOCOL) {
1781 link_recv_proto_msg(l_ptr, buf);
1782 head = link_insert_deferred_queue(l_ptr, head);
1783 tipc_node_unlock(n_ptr);
1786 link_state_event(l_ptr, TRAFFIC_MSG_EVT);
1788 if (link_working_working(l_ptr)) {
1789 /* Re-insert in front of queue */
1792 tipc_node_unlock(n_ptr);
1795 tipc_node_unlock(n_ptr);
1799 read_unlock_bh(&tipc_net_lock);
1803 * link_defer_buf(): Sort a received out-of-sequence packet
1804 * into the deferred reception queue.
1805 * Returns the increase of the queue length,i.e. 0 or 1
1808 u32 tipc_link_defer_pkt(struct sk_buff **head,
1809 struct sk_buff **tail,
1810 struct sk_buff *buf)
1812 struct sk_buff *prev = NULL;
1813 struct sk_buff *crs = *head;
1814 u32 seq_no = msg_seqno(buf_msg(buf));
1819 if (*head == NULL) {
1820 *head = *tail = buf;
1825 if (less(msg_seqno(buf_msg(*tail)), seq_no)) {
1826 (*tail)->next = buf;
1831 /* Scan through queue and sort it in */
1833 struct tipc_msg *msg = buf_msg(crs);
1835 if (less(seq_no, msg_seqno(msg))) {
1843 if (seq_no == msg_seqno(msg))
1849 /* Message is a duplicate of an existing message */
1856 * link_handle_out_of_seq_msg - handle arrival of out-of-sequence packet
1859 static void link_handle_out_of_seq_msg(struct link *l_ptr,
1860 struct sk_buff *buf)
1862 u32 seq_no = msg_seqno(buf_msg(buf));
1864 if (likely(msg_user(buf_msg(buf)) == LINK_PROTOCOL)) {
1865 link_recv_proto_msg(l_ptr, buf);
1869 /* Record OOS packet arrival (force mismatch on next timeout) */
1871 l_ptr->checkpoint--;
1874 * Discard packet if a duplicate; otherwise add it to deferred queue
1875 * and notify peer of gap as per protocol specification
1878 if (less(seq_no, mod(l_ptr->next_in_no))) {
1879 l_ptr->stats.duplicates++;
1884 if (tipc_link_defer_pkt(&l_ptr->oldest_deferred_in,
1885 &l_ptr->newest_deferred_in, buf)) {
1886 l_ptr->deferred_inqueue_sz++;
1887 l_ptr->stats.deferred_recv++;
1888 if ((l_ptr->deferred_inqueue_sz % 16) == 1)
1889 tipc_link_send_proto_msg(l_ptr, STATE_MSG, 0, 0, 0, 0, 0);
1891 l_ptr->stats.duplicates++;
1895 * Send protocol message to the other endpoint.
1897 void tipc_link_send_proto_msg(struct link *l_ptr, u32 msg_typ, int probe_msg,
1898 u32 gap, u32 tolerance, u32 priority, u32 ack_mtu)
1900 struct sk_buff *buf = NULL;
1901 struct tipc_msg *msg = l_ptr->pmsg;
1902 u32 msg_size = sizeof(l_ptr->proto_msg);
1904 if (link_blocked(l_ptr))
1906 msg_set_type(msg, msg_typ);
1907 msg_set_net_plane(msg, l_ptr->b_ptr->net_plane);
1908 msg_set_bcast_ack(msg, mod(l_ptr->owner->bclink.last_in));
1909 msg_set_last_bcast(msg, tipc_bclink_get_last_sent());
1911 if (msg_typ == STATE_MSG) {
1912 u32 next_sent = mod(l_ptr->next_out_no);
1914 if (!tipc_link_is_up(l_ptr))
1916 if (l_ptr->next_out)
1917 next_sent = msg_seqno(buf_msg(l_ptr->next_out));
1918 msg_set_next_sent(msg, next_sent);
1919 if (l_ptr->oldest_deferred_in) {
1920 u32 rec = msg_seqno(buf_msg(l_ptr->oldest_deferred_in));
1921 gap = mod(rec - mod(l_ptr->next_in_no));
1923 msg_set_seq_gap(msg, gap);
1925 l_ptr->stats.sent_nacks++;
1926 msg_set_link_tolerance(msg, tolerance);
1927 msg_set_linkprio(msg, priority);
1928 msg_set_max_pkt(msg, ack_mtu);
1929 msg_set_ack(msg, mod(l_ptr->next_in_no - 1));
1930 msg_set_probe(msg, probe_msg != 0);
1932 u32 mtu = l_ptr->max_pkt;
1934 if ((mtu < l_ptr->max_pkt_target) &&
1935 link_working_working(l_ptr) &&
1936 l_ptr->fsm_msg_cnt) {
1937 msg_size = (mtu + (l_ptr->max_pkt_target - mtu)/2 + 2) & ~3;
1938 if (l_ptr->max_pkt_probes == 10) {
1939 l_ptr->max_pkt_target = (msg_size - 4);
1940 l_ptr->max_pkt_probes = 0;
1941 msg_size = (mtu + (l_ptr->max_pkt_target - mtu)/2 + 2) & ~3;
1943 l_ptr->max_pkt_probes++;
1946 l_ptr->stats.sent_probes++;
1948 l_ptr->stats.sent_states++;
1949 } else { /* RESET_MSG or ACTIVATE_MSG */
1950 msg_set_ack(msg, mod(l_ptr->reset_checkpoint - 1));
1951 msg_set_seq_gap(msg, 0);
1952 msg_set_next_sent(msg, 1);
1953 msg_set_link_tolerance(msg, l_ptr->tolerance);
1954 msg_set_linkprio(msg, l_ptr->priority);
1955 msg_set_max_pkt(msg, l_ptr->max_pkt_target);
1958 if (tipc_node_has_redundant_links(l_ptr->owner))
1959 msg_set_redundant_link(msg);
1961 msg_clear_redundant_link(msg);
1962 msg_set_linkprio(msg, l_ptr->priority);
1964 /* Ensure sequence number will not fit : */
1966 msg_set_seqno(msg, mod(l_ptr->next_out_no + (0xffff/2)));
1970 if (tipc_bearer_congested(l_ptr->b_ptr, l_ptr)) {
1971 if (!l_ptr->proto_msg_queue) {
1972 l_ptr->proto_msg_queue =
1973 tipc_buf_acquire(sizeof(l_ptr->proto_msg));
1975 buf = l_ptr->proto_msg_queue;
1978 skb_copy_to_linear_data(buf, msg, sizeof(l_ptr->proto_msg));
1981 msg_set_timestamp(msg, jiffies_to_msecs(jiffies));
1983 /* Message can be sent */
1985 buf = tipc_buf_acquire(msg_size);
1989 skb_copy_to_linear_data(buf, msg, sizeof(l_ptr->proto_msg));
1990 msg_set_size(buf_msg(buf), msg_size);
1992 if (tipc_bearer_send(l_ptr->b_ptr, buf, &l_ptr->media_addr)) {
1993 l_ptr->unacked_window = 0;
1998 /* New congestion */
1999 tipc_bearer_schedule(l_ptr->b_ptr, l_ptr);
2000 l_ptr->proto_msg_queue = buf;
2001 l_ptr->stats.bearer_congs++;
2005 * Receive protocol message :
2006 * Note that network plane id propagates through the network, and may
2007 * change at any time. The node with lowest address rules
2010 static void link_recv_proto_msg(struct link *l_ptr, struct sk_buff *buf)
2016 struct tipc_msg *msg = buf_msg(buf);
2018 if (link_blocked(l_ptr))
2021 /* record unnumbered packet arrival (force mismatch on next timeout) */
2023 l_ptr->checkpoint--;
2025 if (l_ptr->b_ptr->net_plane != msg_net_plane(msg))
2026 if (tipc_own_addr > msg_prevnode(msg))
2027 l_ptr->b_ptr->net_plane = msg_net_plane(msg);
2029 l_ptr->owner->permit_changeover = msg_redundant_link(msg);
2031 switch (msg_type(msg)) {
2034 if (!link_working_unknown(l_ptr) &&
2035 (l_ptr->peer_session != INVALID_SESSION)) {
2036 if (msg_session(msg) == l_ptr->peer_session)
2037 break; /* duplicate: ignore */
2041 /* Update link settings according other endpoint's values */
2043 strcpy((strrchr(l_ptr->name, ':') + 1), (char *)msg_data(msg));
2045 msg_tol = msg_link_tolerance(msg);
2046 if (msg_tol > l_ptr->tolerance)
2047 link_set_supervision_props(l_ptr, msg_tol);
2049 if (msg_linkprio(msg) > l_ptr->priority)
2050 l_ptr->priority = msg_linkprio(msg);
2052 max_pkt_info = msg_max_pkt(msg);
2054 if (max_pkt_info < l_ptr->max_pkt_target)
2055 l_ptr->max_pkt_target = max_pkt_info;
2056 if (l_ptr->max_pkt > l_ptr->max_pkt_target)
2057 l_ptr->max_pkt = l_ptr->max_pkt_target;
2059 l_ptr->max_pkt = l_ptr->max_pkt_target;
2061 l_ptr->owner->bclink.supported = (max_pkt_info != 0);
2063 link_state_event(l_ptr, msg_type(msg));
2065 l_ptr->peer_session = msg_session(msg);
2066 l_ptr->peer_bearer_id = msg_bearer_id(msg);
2068 /* Synchronize broadcast sequence numbers */
2069 if (!tipc_node_has_redundant_links(l_ptr->owner))
2070 l_ptr->owner->bclink.last_in = mod(msg_last_bcast(msg));
2074 msg_tol = msg_link_tolerance(msg);
2076 link_set_supervision_props(l_ptr, msg_tol);
2078 if (msg_linkprio(msg) &&
2079 (msg_linkprio(msg) != l_ptr->priority)) {
2080 warn("Resetting link <%s>, priority change %u->%u\n",
2081 l_ptr->name, l_ptr->priority, msg_linkprio(msg));
2082 l_ptr->priority = msg_linkprio(msg);
2083 tipc_link_reset(l_ptr); /* Enforce change to take effect */
2086 link_state_event(l_ptr, TRAFFIC_MSG_EVT);
2087 l_ptr->stats.recv_states++;
2088 if (link_reset_unknown(l_ptr))
2091 if (less_eq(mod(l_ptr->next_in_no), msg_next_sent(msg))) {
2092 rec_gap = mod(msg_next_sent(msg) -
2093 mod(l_ptr->next_in_no));
2096 max_pkt_ack = msg_max_pkt(msg);
2097 if (max_pkt_ack > l_ptr->max_pkt) {
2098 l_ptr->max_pkt = max_pkt_ack;
2099 l_ptr->max_pkt_probes = 0;
2103 if (msg_probe(msg)) {
2104 l_ptr->stats.recv_probes++;
2105 if (msg_size(msg) > sizeof(l_ptr->proto_msg))
2106 max_pkt_ack = msg_size(msg);
2109 /* Protocol message before retransmits, reduce loss risk */
2111 tipc_bclink_check_gap(l_ptr->owner, msg_last_bcast(msg));
2113 if (rec_gap || (msg_probe(msg))) {
2114 tipc_link_send_proto_msg(l_ptr, STATE_MSG,
2115 0, rec_gap, 0, 0, max_pkt_ack);
2117 if (msg_seq_gap(msg)) {
2118 l_ptr->stats.recv_nacks++;
2119 tipc_link_retransmit(l_ptr, l_ptr->first_out,
2130 * tipc_link_tunnel(): Send one message via a link belonging to
2131 * another bearer. Owner node is locked.
2133 static void tipc_link_tunnel(struct link *l_ptr,
2134 struct tipc_msg *tunnel_hdr,
2135 struct tipc_msg *msg,
2138 struct link *tunnel;
2139 struct sk_buff *buf;
2140 u32 length = msg_size(msg);
2142 tunnel = l_ptr->owner->active_links[selector & 1];
2143 if (!tipc_link_is_up(tunnel)) {
2144 warn("Link changeover error, "
2145 "tunnel link no longer available\n");
2148 msg_set_size(tunnel_hdr, length + INT_H_SIZE);
2149 buf = tipc_buf_acquire(length + INT_H_SIZE);
2151 warn("Link changeover error, "
2152 "unable to send tunnel msg\n");
2155 skb_copy_to_linear_data(buf, tunnel_hdr, INT_H_SIZE);
2156 skb_copy_to_linear_data_offset(buf, INT_H_SIZE, msg, length);
2157 tipc_link_send_buf(tunnel, buf);
2163 * changeover(): Send whole message queue via the remaining link
2164 * Owner node is locked.
2167 void tipc_link_changeover(struct link *l_ptr)
2169 u32 msgcount = l_ptr->out_queue_size;
2170 struct sk_buff *crs = l_ptr->first_out;
2171 struct link *tunnel = l_ptr->owner->active_links[0];
2172 struct tipc_msg tunnel_hdr;
2178 if (!l_ptr->owner->permit_changeover) {
2179 warn("Link changeover error, "
2180 "peer did not permit changeover\n");
2184 tipc_msg_init(&tunnel_hdr, CHANGEOVER_PROTOCOL,
2185 ORIGINAL_MSG, INT_H_SIZE, l_ptr->addr);
2186 msg_set_bearer_id(&tunnel_hdr, l_ptr->peer_bearer_id);
2187 msg_set_msgcnt(&tunnel_hdr, msgcount);
2189 if (!l_ptr->first_out) {
2190 struct sk_buff *buf;
2192 buf = tipc_buf_acquire(INT_H_SIZE);
2194 skb_copy_to_linear_data(buf, &tunnel_hdr, INT_H_SIZE);
2195 msg_set_size(&tunnel_hdr, INT_H_SIZE);
2196 tipc_link_send_buf(tunnel, buf);
2198 warn("Link changeover error, "
2199 "unable to send changeover msg\n");
2204 split_bundles = (l_ptr->owner->active_links[0] !=
2205 l_ptr->owner->active_links[1]);
2208 struct tipc_msg *msg = buf_msg(crs);
2210 if ((msg_user(msg) == MSG_BUNDLER) && split_bundles) {
2211 struct tipc_msg *m = msg_get_wrapped(msg);
2212 unchar *pos = (unchar *)m;
2214 msgcount = msg_msgcnt(msg);
2215 while (msgcount--) {
2216 msg_set_seqno(m, msg_seqno(msg));
2217 tipc_link_tunnel(l_ptr, &tunnel_hdr, m,
2218 msg_link_selector(m));
2219 pos += align(msg_size(m));
2220 m = (struct tipc_msg *)pos;
2223 tipc_link_tunnel(l_ptr, &tunnel_hdr, msg,
2224 msg_link_selector(msg));
2230 void tipc_link_send_duplicate(struct link *l_ptr, struct link *tunnel)
2232 struct sk_buff *iter;
2233 struct tipc_msg tunnel_hdr;
2235 tipc_msg_init(&tunnel_hdr, CHANGEOVER_PROTOCOL,
2236 DUPLICATE_MSG, INT_H_SIZE, l_ptr->addr);
2237 msg_set_msgcnt(&tunnel_hdr, l_ptr->out_queue_size);
2238 msg_set_bearer_id(&tunnel_hdr, l_ptr->peer_bearer_id);
2239 iter = l_ptr->first_out;
2241 struct sk_buff *outbuf;
2242 struct tipc_msg *msg = buf_msg(iter);
2243 u32 length = msg_size(msg);
2245 if (msg_user(msg) == MSG_BUNDLER)
2246 msg_set_type(msg, CLOSED_MSG);
2247 msg_set_ack(msg, mod(l_ptr->next_in_no - 1)); /* Update */
2248 msg_set_bcast_ack(msg, l_ptr->owner->bclink.last_in);
2249 msg_set_size(&tunnel_hdr, length + INT_H_SIZE);
2250 outbuf = tipc_buf_acquire(length + INT_H_SIZE);
2251 if (outbuf == NULL) {
2252 warn("Link changeover error, "
2253 "unable to send duplicate msg\n");
2256 skb_copy_to_linear_data(outbuf, &tunnel_hdr, INT_H_SIZE);
2257 skb_copy_to_linear_data_offset(outbuf, INT_H_SIZE, iter->data,
2259 tipc_link_send_buf(tunnel, outbuf);
2260 if (!tipc_link_is_up(l_ptr))
2269 * buf_extract - extracts embedded TIPC message from another message
2270 * @skb: encapsulating message buffer
2271 * @from_pos: offset to extract from
2273 * Returns a new message buffer containing an embedded message. The
2274 * encapsulating message itself is left unchanged.
2277 static struct sk_buff *buf_extract(struct sk_buff *skb, u32 from_pos)
2279 struct tipc_msg *msg = (struct tipc_msg *)(skb->data + from_pos);
2280 u32 size = msg_size(msg);
2283 eb = tipc_buf_acquire(size);
2285 skb_copy_to_linear_data(eb, msg, size);
2290 * link_recv_changeover_msg(): Receive tunneled packet sent
2291 * via other link. Node is locked. Return extracted buffer.
2294 static int link_recv_changeover_msg(struct link **l_ptr,
2295 struct sk_buff **buf)
2297 struct sk_buff *tunnel_buf = *buf;
2298 struct link *dest_link;
2299 struct tipc_msg *msg;
2300 struct tipc_msg *tunnel_msg = buf_msg(tunnel_buf);
2301 u32 msg_typ = msg_type(tunnel_msg);
2302 u32 msg_count = msg_msgcnt(tunnel_msg);
2304 dest_link = (*l_ptr)->owner->links[msg_bearer_id(tunnel_msg)];
2307 if (dest_link == *l_ptr) {
2308 err("Unexpected changeover message on link <%s>\n",
2313 msg = msg_get_wrapped(tunnel_msg);
2315 if (msg_typ == DUPLICATE_MSG) {
2316 if (less(msg_seqno(msg), mod(dest_link->next_in_no)))
2318 *buf = buf_extract(tunnel_buf, INT_H_SIZE);
2320 warn("Link changeover error, duplicate msg dropped\n");
2323 buf_discard(tunnel_buf);
2327 /* First original message ?: */
2329 if (tipc_link_is_up(dest_link)) {
2330 info("Resetting link <%s>, changeover initiated by peer\n",
2332 tipc_link_reset(dest_link);
2333 dest_link->exp_msg_count = msg_count;
2336 } else if (dest_link->exp_msg_count == START_CHANGEOVER) {
2337 dest_link->exp_msg_count = msg_count;
2342 /* Receive original message */
2344 if (dest_link->exp_msg_count == 0) {
2345 warn("Link switchover error, "
2346 "got too many tunnelled messages\n");
2349 dest_link->exp_msg_count--;
2350 if (less(msg_seqno(msg), dest_link->reset_checkpoint)) {
2353 *buf = buf_extract(tunnel_buf, INT_H_SIZE);
2355 buf_discard(tunnel_buf);
2358 warn("Link changeover error, original msg dropped\n");
2363 buf_discard(tunnel_buf);
2368 * Bundler functionality:
2370 void tipc_link_recv_bundle(struct sk_buff *buf)
2372 u32 msgcount = msg_msgcnt(buf_msg(buf));
2373 u32 pos = INT_H_SIZE;
2374 struct sk_buff *obuf;
2376 while (msgcount--) {
2377 obuf = buf_extract(buf, pos);
2379 warn("Link unable to unbundle message(s)\n");
2382 pos += align(msg_size(buf_msg(obuf)));
2383 tipc_net_route_msg(obuf);
2389 * Fragmentation/defragmentation:
2394 * link_send_long_buf: Entry for buffers needing fragmentation.
2395 * The buffer is complete, inclusive total message length.
2396 * Returns user data length.
2398 static int link_send_long_buf(struct link *l_ptr, struct sk_buff *buf)
2400 struct tipc_msg *inmsg = buf_msg(buf);
2401 struct tipc_msg fragm_hdr;
2402 u32 insize = msg_size(inmsg);
2403 u32 dsz = msg_data_sz(inmsg);
2404 unchar *crs = buf->data;
2406 u32 pack_sz = l_ptr->max_pkt;
2407 u32 fragm_sz = pack_sz - INT_H_SIZE;
2411 if (msg_short(inmsg))
2412 destaddr = l_ptr->addr;
2414 destaddr = msg_destnode(inmsg);
2416 if (msg_routed(inmsg))
2417 msg_set_prevnode(inmsg, tipc_own_addr);
2419 /* Prepare reusable fragment header: */
2421 tipc_msg_init(&fragm_hdr, MSG_FRAGMENTER, FIRST_FRAGMENT,
2422 INT_H_SIZE, destaddr);
2423 msg_set_link_selector(&fragm_hdr, msg_link_selector(inmsg));
2424 msg_set_long_msgno(&fragm_hdr, mod(l_ptr->long_msg_seq_no++));
2425 msg_set_fragm_no(&fragm_hdr, fragm_no);
2426 l_ptr->stats.sent_fragmented++;
2428 /* Chop up message: */
2431 struct sk_buff *fragm;
2433 if (rest <= fragm_sz) {
2435 msg_set_type(&fragm_hdr, LAST_FRAGMENT);
2437 fragm = tipc_buf_acquire(fragm_sz + INT_H_SIZE);
2438 if (fragm == NULL) {
2439 warn("Link unable to fragment message\n");
2443 msg_set_size(&fragm_hdr, fragm_sz + INT_H_SIZE);
2444 skb_copy_to_linear_data(fragm, &fragm_hdr, INT_H_SIZE);
2445 skb_copy_to_linear_data_offset(fragm, INT_H_SIZE, crs,
2447 /* Send queued messages first, if any: */
2449 l_ptr->stats.sent_fragments++;
2450 tipc_link_send_buf(l_ptr, fragm);
2451 if (!tipc_link_is_up(l_ptr))
2453 msg_set_fragm_no(&fragm_hdr, ++fragm_no);
2456 msg_set_type(&fragm_hdr, FRAGMENT);
2464 * A pending message being re-assembled must store certain values
2465 * to handle subsequent fragments correctly. The following functions
2466 * help storing these values in unused, available fields in the
2467 * pending message. This makes dynamic memory allocation unecessary.
2470 static void set_long_msg_seqno(struct sk_buff *buf, u32 seqno)
2472 msg_set_seqno(buf_msg(buf), seqno);
2475 static u32 get_fragm_size(struct sk_buff *buf)
2477 return msg_ack(buf_msg(buf));
2480 static void set_fragm_size(struct sk_buff *buf, u32 sz)
2482 msg_set_ack(buf_msg(buf), sz);
2485 static u32 get_expected_frags(struct sk_buff *buf)
2487 return msg_bcast_ack(buf_msg(buf));
2490 static void set_expected_frags(struct sk_buff *buf, u32 exp)
2492 msg_set_bcast_ack(buf_msg(buf), exp);
2495 static u32 get_timer_cnt(struct sk_buff *buf)
2497 return msg_reroute_cnt(buf_msg(buf));
2500 static void incr_timer_cnt(struct sk_buff *buf)
2502 msg_incr_reroute_cnt(buf_msg(buf));
2506 * tipc_link_recv_fragment(): Called with node lock on. Returns
2507 * the reassembled buffer if message is complete.
2509 int tipc_link_recv_fragment(struct sk_buff **pending, struct sk_buff **fb,
2510 struct tipc_msg **m)
2512 struct sk_buff *prev = NULL;
2513 struct sk_buff *fbuf = *fb;
2514 struct tipc_msg *fragm = buf_msg(fbuf);
2515 struct sk_buff *pbuf = *pending;
2516 u32 long_msg_seq_no = msg_long_msgno(fragm);
2520 /* Is there an incomplete message waiting for this fragment? */
2522 while (pbuf && ((msg_seqno(buf_msg(pbuf)) != long_msg_seq_no) ||
2523 (msg_orignode(fragm) != msg_orignode(buf_msg(pbuf))))) {
2528 if (!pbuf && (msg_type(fragm) == FIRST_FRAGMENT)) {
2529 struct tipc_msg *imsg = (struct tipc_msg *)msg_data(fragm);
2530 u32 msg_sz = msg_size(imsg);
2531 u32 fragm_sz = msg_data_sz(fragm);
2532 u32 exp_fragm_cnt = msg_sz/fragm_sz + !!(msg_sz % fragm_sz);
2533 u32 max = TIPC_MAX_USER_MSG_SIZE + LONG_H_SIZE;
2534 if (msg_type(imsg) == TIPC_MCAST_MSG)
2535 max = TIPC_MAX_USER_MSG_SIZE + MCAST_H_SIZE;
2536 if (msg_size(imsg) > max) {
2540 pbuf = tipc_buf_acquire(msg_size(imsg));
2542 pbuf->next = *pending;
2544 skb_copy_to_linear_data(pbuf, imsg,
2545 msg_data_sz(fragm));
2546 /* Prepare buffer for subsequent fragments. */
2548 set_long_msg_seqno(pbuf, long_msg_seq_no);
2549 set_fragm_size(pbuf, fragm_sz);
2550 set_expected_frags(pbuf, exp_fragm_cnt - 1);
2552 warn("Link unable to reassemble fragmented message\n");
2556 } else if (pbuf && (msg_type(fragm) != FIRST_FRAGMENT)) {
2557 u32 dsz = msg_data_sz(fragm);
2558 u32 fsz = get_fragm_size(pbuf);
2559 u32 crs = ((msg_fragm_no(fragm) - 1) * fsz);
2560 u32 exp_frags = get_expected_frags(pbuf) - 1;
2561 skb_copy_to_linear_data_offset(pbuf, crs,
2562 msg_data(fragm), dsz);
2565 /* Is message complete? */
2567 if (exp_frags == 0) {
2569 prev->next = pbuf->next;
2571 *pending = pbuf->next;
2572 msg_reset_reroute_cnt(buf_msg(pbuf));
2577 set_expected_frags(pbuf, exp_frags);
2585 * link_check_defragm_bufs - flush stale incoming message fragments
2586 * @l_ptr: pointer to link
2589 static void link_check_defragm_bufs(struct link *l_ptr)
2591 struct sk_buff *prev = NULL;
2592 struct sk_buff *next = NULL;
2593 struct sk_buff *buf = l_ptr->defragm_buf;
2597 if (!link_working_working(l_ptr))
2600 u32 cnt = get_timer_cnt(buf);
2604 incr_timer_cnt(buf);
2608 prev->next = buf->next;
2610 l_ptr->defragm_buf = buf->next;
2619 static void link_set_supervision_props(struct link *l_ptr, u32 tolerance)
2621 l_ptr->tolerance = tolerance;
2622 l_ptr->continuity_interval =
2623 ((tolerance / 4) > 500) ? 500 : tolerance / 4;
2624 l_ptr->abort_limit = tolerance / (l_ptr->continuity_interval / 4);
2628 void tipc_link_set_queue_limits(struct link *l_ptr, u32 window)
2630 /* Data messages from this node, inclusive FIRST_FRAGM */
2631 l_ptr->queue_limit[TIPC_LOW_IMPORTANCE] = window;
2632 l_ptr->queue_limit[TIPC_MEDIUM_IMPORTANCE] = (window / 3) * 4;
2633 l_ptr->queue_limit[TIPC_HIGH_IMPORTANCE] = (window / 3) * 5;
2634 l_ptr->queue_limit[TIPC_CRITICAL_IMPORTANCE] = (window / 3) * 6;
2635 /* Transiting data messages,inclusive FIRST_FRAGM */
2636 l_ptr->queue_limit[TIPC_LOW_IMPORTANCE + 4] = 300;
2637 l_ptr->queue_limit[TIPC_MEDIUM_IMPORTANCE + 4] = 600;
2638 l_ptr->queue_limit[TIPC_HIGH_IMPORTANCE + 4] = 900;
2639 l_ptr->queue_limit[TIPC_CRITICAL_IMPORTANCE + 4] = 1200;
2640 l_ptr->queue_limit[CONN_MANAGER] = 1200;
2641 l_ptr->queue_limit[CHANGEOVER_PROTOCOL] = 2500;
2642 l_ptr->queue_limit[NAME_DISTRIBUTOR] = 3000;
2643 /* FRAGMENT and LAST_FRAGMENT packets */
2644 l_ptr->queue_limit[MSG_FRAGMENTER] = 4000;
2648 * link_find_link - locate link by name
2649 * @name - ptr to link name string
2650 * @node - ptr to area to be filled with ptr to associated node
2652 * Caller must hold 'tipc_net_lock' to ensure node and bearer are not deleted;
2653 * this also prevents link deletion.
2655 * Returns pointer to link (or 0 if invalid link name).
2658 static struct link *link_find_link(const char *name, struct tipc_node **node)
2660 struct link_name link_name_parts;
2661 struct bearer *b_ptr;
2664 if (!link_name_validate(name, &link_name_parts))
2667 b_ptr = tipc_bearer_find_interface(link_name_parts.if_local);
2671 *node = tipc_node_find(link_name_parts.addr_peer);
2675 l_ptr = (*node)->links[b_ptr->identity];
2676 if (!l_ptr || strcmp(l_ptr->name, name))
2682 struct sk_buff *tipc_link_cmd_config(const void *req_tlv_area, int req_tlv_space,
2685 struct tipc_link_config *args;
2688 struct tipc_node *node;
2691 if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_LINK_CONFIG))
2692 return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
2694 args = (struct tipc_link_config *)TLV_DATA(req_tlv_area);
2695 new_value = ntohl(args->value);
2697 if (!strcmp(args->name, tipc_bclink_name)) {
2698 if ((cmd == TIPC_CMD_SET_LINK_WINDOW) &&
2699 (tipc_bclink_set_queue_limits(new_value) == 0))
2700 return tipc_cfg_reply_none();
2701 return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED
2702 " (cannot change setting on broadcast link)");
2705 read_lock_bh(&tipc_net_lock);
2706 l_ptr = link_find_link(args->name, &node);
2708 read_unlock_bh(&tipc_net_lock);
2709 return tipc_cfg_reply_error_string("link not found");
2712 tipc_node_lock(node);
2715 case TIPC_CMD_SET_LINK_TOL:
2716 if ((new_value >= TIPC_MIN_LINK_TOL) &&
2717 (new_value <= TIPC_MAX_LINK_TOL)) {
2718 link_set_supervision_props(l_ptr, new_value);
2719 tipc_link_send_proto_msg(l_ptr, STATE_MSG,
2720 0, 0, new_value, 0, 0);
2724 case TIPC_CMD_SET_LINK_PRI:
2725 if ((new_value >= TIPC_MIN_LINK_PRI) &&
2726 (new_value <= TIPC_MAX_LINK_PRI)) {
2727 l_ptr->priority = new_value;
2728 tipc_link_send_proto_msg(l_ptr, STATE_MSG,
2729 0, 0, 0, new_value, 0);
2733 case TIPC_CMD_SET_LINK_WINDOW:
2734 if ((new_value >= TIPC_MIN_LINK_WIN) &&
2735 (new_value <= TIPC_MAX_LINK_WIN)) {
2736 tipc_link_set_queue_limits(l_ptr, new_value);
2741 tipc_node_unlock(node);
2743 read_unlock_bh(&tipc_net_lock);
2745 return tipc_cfg_reply_error_string("cannot change link setting");
2747 return tipc_cfg_reply_none();
2751 * link_reset_statistics - reset link statistics
2752 * @l_ptr: pointer to link
2755 static void link_reset_statistics(struct link *l_ptr)
2757 memset(&l_ptr->stats, 0, sizeof(l_ptr->stats));
2758 l_ptr->stats.sent_info = l_ptr->next_out_no;
2759 l_ptr->stats.recv_info = l_ptr->next_in_no;
2762 struct sk_buff *tipc_link_cmd_reset_stats(const void *req_tlv_area, int req_tlv_space)
2766 struct tipc_node *node;
2768 if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_LINK_NAME))
2769 return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
2771 link_name = (char *)TLV_DATA(req_tlv_area);
2772 if (!strcmp(link_name, tipc_bclink_name)) {
2773 if (tipc_bclink_reset_stats())
2774 return tipc_cfg_reply_error_string("link not found");
2775 return tipc_cfg_reply_none();
2778 read_lock_bh(&tipc_net_lock);
2779 l_ptr = link_find_link(link_name, &node);
2781 read_unlock_bh(&tipc_net_lock);
2782 return tipc_cfg_reply_error_string("link not found");
2785 tipc_node_lock(node);
2786 link_reset_statistics(l_ptr);
2787 tipc_node_unlock(node);
2788 read_unlock_bh(&tipc_net_lock);
2789 return tipc_cfg_reply_none();
2793 * percent - convert count to a percentage of total (rounding up or down)
2796 static u32 percent(u32 count, u32 total)
2798 return (count * 100 + (total / 2)) / total;
2802 * tipc_link_stats - print link statistics
2804 * @buf: print buffer area
2805 * @buf_size: size of print buffer area
2807 * Returns length of print buffer data string (or 0 if error)
2810 static int tipc_link_stats(const char *name, char *buf, const u32 buf_size)
2812 struct print_buf pb;
2814 struct tipc_node *node;
2816 u32 profile_total = 0;
2818 if (!strcmp(name, tipc_bclink_name))
2819 return tipc_bclink_stats(buf, buf_size);
2821 tipc_printbuf_init(&pb, buf, buf_size);
2823 read_lock_bh(&tipc_net_lock);
2824 l_ptr = link_find_link(name, &node);
2826 read_unlock_bh(&tipc_net_lock);
2829 tipc_node_lock(node);
2831 if (tipc_link_is_active(l_ptr))
2833 else if (tipc_link_is_up(l_ptr))
2837 tipc_printf(&pb, "Link <%s>\n"
2838 " %s MTU:%u Priority:%u Tolerance:%u ms"
2839 " Window:%u packets\n",
2840 l_ptr->name, status, l_ptr->max_pkt,
2841 l_ptr->priority, l_ptr->tolerance, l_ptr->queue_limit[0]);
2842 tipc_printf(&pb, " RX packets:%u fragments:%u/%u bundles:%u/%u\n",
2843 l_ptr->next_in_no - l_ptr->stats.recv_info,
2844 l_ptr->stats.recv_fragments,
2845 l_ptr->stats.recv_fragmented,
2846 l_ptr->stats.recv_bundles,
2847 l_ptr->stats.recv_bundled);
2848 tipc_printf(&pb, " TX packets:%u fragments:%u/%u bundles:%u/%u\n",
2849 l_ptr->next_out_no - l_ptr->stats.sent_info,
2850 l_ptr->stats.sent_fragments,
2851 l_ptr->stats.sent_fragmented,
2852 l_ptr->stats.sent_bundles,
2853 l_ptr->stats.sent_bundled);
2854 profile_total = l_ptr->stats.msg_length_counts;
2857 tipc_printf(&pb, " TX profile sample:%u packets average:%u octets\n"
2858 " 0-64:%u%% -256:%u%% -1024:%u%% -4096:%u%% "
2859 "-16354:%u%% -32768:%u%% -66000:%u%%\n",
2860 l_ptr->stats.msg_length_counts,
2861 l_ptr->stats.msg_lengths_total / profile_total,
2862 percent(l_ptr->stats.msg_length_profile[0], profile_total),
2863 percent(l_ptr->stats.msg_length_profile[1], profile_total),
2864 percent(l_ptr->stats.msg_length_profile[2], profile_total),
2865 percent(l_ptr->stats.msg_length_profile[3], profile_total),
2866 percent(l_ptr->stats.msg_length_profile[4], profile_total),
2867 percent(l_ptr->stats.msg_length_profile[5], profile_total),
2868 percent(l_ptr->stats.msg_length_profile[6], profile_total));
2869 tipc_printf(&pb, " RX states:%u probes:%u naks:%u defs:%u dups:%u\n",
2870 l_ptr->stats.recv_states,
2871 l_ptr->stats.recv_probes,
2872 l_ptr->stats.recv_nacks,
2873 l_ptr->stats.deferred_recv,
2874 l_ptr->stats.duplicates);
2875 tipc_printf(&pb, " TX states:%u probes:%u naks:%u acks:%u dups:%u\n",
2876 l_ptr->stats.sent_states,
2877 l_ptr->stats.sent_probes,
2878 l_ptr->stats.sent_nacks,
2879 l_ptr->stats.sent_acks,
2880 l_ptr->stats.retransmitted);
2881 tipc_printf(&pb, " Congestion bearer:%u link:%u Send queue max:%u avg:%u\n",
2882 l_ptr->stats.bearer_congs,
2883 l_ptr->stats.link_congs,
2884 l_ptr->stats.max_queue_sz,
2885 l_ptr->stats.queue_sz_counts
2886 ? (l_ptr->stats.accu_queue_sz / l_ptr->stats.queue_sz_counts)
2889 tipc_node_unlock(node);
2890 read_unlock_bh(&tipc_net_lock);
2891 return tipc_printbuf_validate(&pb);
2894 #define MAX_LINK_STATS_INFO 2000
2896 struct sk_buff *tipc_link_cmd_show_stats(const void *req_tlv_area, int req_tlv_space)
2898 struct sk_buff *buf;
2899 struct tlv_desc *rep_tlv;
2902 if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_LINK_NAME))
2903 return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
2905 buf = tipc_cfg_reply_alloc(TLV_SPACE(MAX_LINK_STATS_INFO));
2909 rep_tlv = (struct tlv_desc *)buf->data;
2911 str_len = tipc_link_stats((char *)TLV_DATA(req_tlv_area),
2912 (char *)TLV_DATA(rep_tlv), MAX_LINK_STATS_INFO);
2915 return tipc_cfg_reply_error_string("link not found");
2918 skb_put(buf, TLV_SPACE(str_len));
2919 TLV_SET(rep_tlv, TIPC_TLV_ULTRA_STRING, NULL, str_len);
2925 * tipc_link_get_max_pkt - get maximum packet size to use when sending to destination
2926 * @dest: network address of destination node
2927 * @selector: used to select from set of active links
2929 * If no active link can be found, uses default maximum packet size.
2932 u32 tipc_link_get_max_pkt(u32 dest, u32 selector)
2934 struct tipc_node *n_ptr;
2936 u32 res = MAX_PKT_DEFAULT;
2938 if (dest == tipc_own_addr)
2939 return MAX_MSG_SIZE;
2941 read_lock_bh(&tipc_net_lock);
2942 n_ptr = tipc_node_find(dest);
2944 tipc_node_lock(n_ptr);
2945 l_ptr = n_ptr->active_links[selector & 1];
2947 res = l_ptr->max_pkt;
2948 tipc_node_unlock(n_ptr);
2950 read_unlock_bh(&tipc_net_lock);
2954 static void link_print(struct link *l_ptr, const char *str)
2956 char print_area[256];
2957 struct print_buf pb;
2958 struct print_buf *buf = &pb;
2960 tipc_printbuf_init(buf, print_area, sizeof(print_area));
2962 tipc_printf(buf, str);
2963 tipc_printf(buf, "Link %x<%s>:",
2964 l_ptr->addr, l_ptr->b_ptr->publ.name);
2966 #ifdef CONFIG_TIPC_DEBUG
2967 if (link_reset_reset(l_ptr) || link_reset_unknown(l_ptr))
2970 tipc_printf(buf, ": NXO(%u):", mod(l_ptr->next_out_no));
2971 tipc_printf(buf, "NXI(%u):", mod(l_ptr->next_in_no));
2972 tipc_printf(buf, "SQUE");
2973 if (l_ptr->first_out) {
2974 tipc_printf(buf, "[%u..", msg_seqno(buf_msg(l_ptr->first_out)));
2975 if (l_ptr->next_out)
2976 tipc_printf(buf, "%u..",
2977 msg_seqno(buf_msg(l_ptr->next_out)));
2978 tipc_printf(buf, "%u]", msg_seqno(buf_msg(l_ptr->last_out)));
2979 if ((mod(msg_seqno(buf_msg(l_ptr->last_out)) -
2980 msg_seqno(buf_msg(l_ptr->first_out)))
2981 != (l_ptr->out_queue_size - 1)) ||
2982 (l_ptr->last_out->next != NULL)) {
2983 tipc_printf(buf, "\nSend queue inconsistency\n");
2984 tipc_printf(buf, "first_out= %x ", l_ptr->first_out);
2985 tipc_printf(buf, "next_out= %x ", l_ptr->next_out);
2986 tipc_printf(buf, "last_out= %x ", l_ptr->last_out);
2989 tipc_printf(buf, "[]");
2990 tipc_printf(buf, "SQSIZ(%u)", l_ptr->out_queue_size);
2991 if (l_ptr->oldest_deferred_in) {
2992 u32 o = msg_seqno(buf_msg(l_ptr->oldest_deferred_in));
2993 u32 n = msg_seqno(buf_msg(l_ptr->newest_deferred_in));
2994 tipc_printf(buf, ":RQUE[%u..%u]", o, n);
2995 if (l_ptr->deferred_inqueue_sz != mod((n + 1) - o)) {
2996 tipc_printf(buf, ":RQSIZ(%u)",
2997 l_ptr->deferred_inqueue_sz);
3003 if (link_working_unknown(l_ptr))
3004 tipc_printf(buf, ":WU");
3005 else if (link_reset_reset(l_ptr))
3006 tipc_printf(buf, ":RR");
3007 else if (link_reset_unknown(l_ptr))
3008 tipc_printf(buf, ":RU");
3009 else if (link_working_working(l_ptr))
3010 tipc_printf(buf, ":WW");
3011 tipc_printf(buf, "\n");
3013 tipc_printbuf_validate(buf);
3014 info("%s", print_area);