2 * net/tipc/node.c: TIPC node management routines
4 * Copyright (c) 2000-2006, 2012-2015, Ericsson AB
5 * Copyright (c) 2005-2006, 2010-2014, 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"
45 /* Node FSM states and events:
48 SELF_DOWN_PEER_DOWN
= 0xdd,
49 SELF_UP_PEER_UP
= 0xaa,
50 SELF_DOWN_PEER_LEAVING
= 0xd1,
51 SELF_UP_PEER_COMING
= 0xac,
52 SELF_COMING_PEER_UP
= 0xca,
53 SELF_LEAVING_PEER_DOWN
= 0x1d,
54 NODE_FAILINGOVER
= 0xf0,
59 SELF_ESTABL_CONTACT_EVT
= 0xece,
60 SELF_LOST_CONTACT_EVT
= 0x1ce,
61 PEER_ESTABL_CONTACT_EVT
= 0x9ece,
62 PEER_LOST_CONTACT_EVT
= 0x91ce,
63 NODE_FAILOVER_BEGIN_EVT
= 0xfbe,
64 NODE_FAILOVER_END_EVT
= 0xfee,
65 NODE_SYNCH_BEGIN_EVT
= 0xcbe,
66 NODE_SYNCH_END_EVT
= 0xcee
69 static void __tipc_node_link_down(struct tipc_node
*n
, int *bearer_id
,
70 struct sk_buff_head
*xmitq
,
71 struct tipc_media_addr
**maddr
);
72 static void tipc_node_link_down(struct tipc_node
*n
, int bearer_id
,
74 static void node_lost_contact(struct tipc_node
*n
, struct sk_buff_head
*inputq
);
75 static void tipc_node_delete(struct tipc_node
*node
);
76 static void tipc_node_timeout(unsigned long data
);
77 static void tipc_node_fsm_evt(struct tipc_node
*n
, int evt
);
79 struct tipc_sock_conn
{
83 struct list_head list
;
86 static const struct nla_policy tipc_nl_node_policy
[TIPC_NLA_NODE_MAX
+ 1] = {
87 [TIPC_NLA_NODE_UNSPEC
] = { .type
= NLA_UNSPEC
},
88 [TIPC_NLA_NODE_ADDR
] = { .type
= NLA_U32
},
89 [TIPC_NLA_NODE_UP
] = { .type
= NLA_FLAG
}
93 * A trivial power-of-two bitmask technique is used for speed, since this
94 * operation is done for every incoming TIPC packet. The number of hash table
95 * entries has been chosen so that no hash chain exceeds 8 nodes and will
96 * usually be much smaller (typically only a single node).
98 static unsigned int tipc_hashfn(u32 addr
)
100 return addr
& (NODE_HTABLE_SIZE
- 1);
103 static void tipc_node_kref_release(struct kref
*kref
)
105 struct tipc_node
*node
= container_of(kref
, struct tipc_node
, kref
);
107 tipc_node_delete(node
);
110 void tipc_node_put(struct tipc_node
*node
)
112 kref_put(&node
->kref
, tipc_node_kref_release
);
115 static void tipc_node_get(struct tipc_node
*node
)
117 kref_get(&node
->kref
);
121 * tipc_node_find - locate specified node object, if it exists
123 struct tipc_node
*tipc_node_find(struct net
*net
, u32 addr
)
125 struct tipc_net
*tn
= net_generic(net
, tipc_net_id
);
126 struct tipc_node
*node
;
128 if (unlikely(!in_own_cluster_exact(net
, addr
)))
132 hlist_for_each_entry_rcu(node
, &tn
->node_htable
[tipc_hashfn(addr
)],
134 if (node
->addr
== addr
) {
144 struct tipc_node
*tipc_node_create(struct net
*net
, u32 addr
, u16 capabilities
)
146 struct tipc_net
*tn
= net_generic(net
, tipc_net_id
);
147 struct tipc_node
*n_ptr
, *temp_node
;
149 spin_lock_bh(&tn
->node_list_lock
);
150 n_ptr
= tipc_node_find(net
, addr
);
153 n_ptr
= kzalloc(sizeof(*n_ptr
), GFP_ATOMIC
);
155 pr_warn("Node creation failed, no memory\n");
160 n_ptr
->capabilities
= capabilities
;
161 kref_init(&n_ptr
->kref
);
162 spin_lock_init(&n_ptr
->lock
);
163 INIT_HLIST_NODE(&n_ptr
->hash
);
164 INIT_LIST_HEAD(&n_ptr
->list
);
165 INIT_LIST_HEAD(&n_ptr
->publ_list
);
166 INIT_LIST_HEAD(&n_ptr
->conn_sks
);
167 skb_queue_head_init(&n_ptr
->bc_entry
.namedq
);
168 skb_queue_head_init(&n_ptr
->bc_entry
.inputq1
);
169 __skb_queue_head_init(&n_ptr
->bc_entry
.arrvq
);
170 skb_queue_head_init(&n_ptr
->bc_entry
.inputq2
);
171 hlist_add_head_rcu(&n_ptr
->hash
, &tn
->node_htable
[tipc_hashfn(addr
)]);
172 list_for_each_entry_rcu(temp_node
, &tn
->node_list
, list
) {
173 if (n_ptr
->addr
< temp_node
->addr
)
176 list_add_tail_rcu(&n_ptr
->list
, &temp_node
->list
);
177 n_ptr
->state
= SELF_DOWN_PEER_LEAVING
;
178 n_ptr
->signature
= INVALID_NODE_SIG
;
179 n_ptr
->active_links
[0] = INVALID_BEARER_ID
;
180 n_ptr
->active_links
[1] = INVALID_BEARER_ID
;
181 if (!tipc_link_bc_create(n_ptr
, tipc_own_addr(net
), n_ptr
->addr
,
182 U16_MAX
, tipc_bc_sndlink(net
)->window
,
184 &n_ptr
->bc_entry
.inputq1
,
185 &n_ptr
->bc_entry
.namedq
,
186 tipc_bc_sndlink(net
),
187 &n_ptr
->bc_entry
.link
)) {
188 pr_warn("Broadcast rcv link creation failed, no memory\n");
193 tipc_node_get(n_ptr
);
194 setup_timer(&n_ptr
->timer
, tipc_node_timeout
, (unsigned long)n_ptr
);
195 n_ptr
->keepalive_intv
= U32_MAX
;
197 spin_unlock_bh(&tn
->node_list_lock
);
201 static void tipc_node_calculate_timer(struct tipc_node
*n
, struct tipc_link
*l
)
203 unsigned long tol
= l
->tolerance
;
204 unsigned long intv
= ((tol
/ 4) > 500) ? 500 : tol
/ 4;
205 unsigned long keepalive_intv
= msecs_to_jiffies(intv
);
207 /* Link with lowest tolerance determines timer interval */
208 if (keepalive_intv
< n
->keepalive_intv
)
209 n
->keepalive_intv
= keepalive_intv
;
211 /* Ensure link's abort limit corresponds to current interval */
212 l
->abort_limit
= l
->tolerance
/ jiffies_to_msecs(n
->keepalive_intv
);
215 static void tipc_node_delete(struct tipc_node
*node
)
217 list_del_rcu(&node
->list
);
218 hlist_del_rcu(&node
->hash
);
219 kfree(node
->bc_entry
.link
);
220 kfree_rcu(node
, rcu
);
223 void tipc_node_stop(struct net
*net
)
225 struct tipc_net
*tn
= net_generic(net
, tipc_net_id
);
226 struct tipc_node
*node
, *t_node
;
228 spin_lock_bh(&tn
->node_list_lock
);
229 list_for_each_entry_safe(node
, t_node
, &tn
->node_list
, list
) {
230 if (del_timer(&node
->timer
))
234 spin_unlock_bh(&tn
->node_list_lock
);
237 int tipc_node_add_conn(struct net
*net
, u32 dnode
, u32 port
, u32 peer_port
)
239 struct tipc_node
*node
;
240 struct tipc_sock_conn
*conn
;
243 if (in_own_node(net
, dnode
))
246 node
= tipc_node_find(net
, dnode
);
248 pr_warn("Connecting sock to node 0x%x failed\n", dnode
);
249 return -EHOSTUNREACH
;
251 conn
= kmalloc(sizeof(*conn
), GFP_ATOMIC
);
256 conn
->peer_node
= dnode
;
258 conn
->peer_port
= peer_port
;
260 tipc_node_lock(node
);
261 list_add_tail(&conn
->list
, &node
->conn_sks
);
262 tipc_node_unlock(node
);
268 void tipc_node_remove_conn(struct net
*net
, u32 dnode
, u32 port
)
270 struct tipc_node
*node
;
271 struct tipc_sock_conn
*conn
, *safe
;
273 if (in_own_node(net
, dnode
))
276 node
= tipc_node_find(net
, dnode
);
280 tipc_node_lock(node
);
281 list_for_each_entry_safe(conn
, safe
, &node
->conn_sks
, list
) {
282 if (port
!= conn
->port
)
284 list_del(&conn
->list
);
287 tipc_node_unlock(node
);
291 /* tipc_node_timeout - handle expiration of node timer
293 static void tipc_node_timeout(unsigned long data
)
295 struct tipc_node
*n
= (struct tipc_node
*)data
;
296 struct tipc_link_entry
*le
;
297 struct sk_buff_head xmitq
;
301 __skb_queue_head_init(&xmitq
);
303 for (bearer_id
= 0; bearer_id
< MAX_BEARERS
; bearer_id
++) {
305 le
= &n
->links
[bearer_id
];
307 /* Link tolerance may change asynchronously: */
308 tipc_node_calculate_timer(n
, le
->link
);
309 rc
= tipc_link_timeout(le
->link
, &xmitq
);
312 tipc_bearer_xmit(n
->net
, bearer_id
, &xmitq
, &le
->maddr
);
313 if (rc
& TIPC_LINK_DOWN_EVT
)
314 tipc_node_link_down(n
, bearer_id
, false);
316 if (!mod_timer(&n
->timer
, jiffies
+ n
->keepalive_intv
))
322 * __tipc_node_link_up - handle addition of link
323 * Node lock must be held by caller
324 * Link becomes active (alone or shared) or standby, depending on its priority.
326 static void __tipc_node_link_up(struct tipc_node
*n
, int bearer_id
,
327 struct sk_buff_head
*xmitq
)
329 int *slot0
= &n
->active_links
[0];
330 int *slot1
= &n
->active_links
[1];
331 struct tipc_link
*ol
= node_active_link(n
, 0);
332 struct tipc_link
*nl
= n
->links
[bearer_id
].link
;
337 tipc_link_fsm_evt(nl
, LINK_ESTABLISH_EVT
);
338 if (!tipc_link_is_up(nl
))
342 n
->action_flags
|= TIPC_NOTIFY_LINK_UP
;
343 n
->link_id
= nl
->peer_bearer_id
<< 16 | bearer_id
;
345 /* Leave room for tunnel header when returning 'mtu' to users: */
346 n
->links
[bearer_id
].mtu
= nl
->mtu
- INT_H_SIZE
;
348 tipc_bearer_add_dest(n
->net
, bearer_id
, n
->addr
);
350 pr_debug("Established link <%s> on network plane %c\n",
351 nl
->name
, nl
->net_plane
);
353 /* First link? => give it both slots */
357 tipc_node_fsm_evt(n
, SELF_ESTABL_CONTACT_EVT
);
358 n
->action_flags
|= TIPC_NOTIFY_NODE_UP
;
359 tipc_bcast_add_peer(n
->net
, n
->addr
, nl
, xmitq
);
363 /* Second link => redistribute slots */
364 if (nl
->priority
> ol
->priority
) {
365 pr_debug("Old link <%s> becomes standby\n", ol
->name
);
368 } else if (nl
->priority
== ol
->priority
) {
371 pr_debug("New link <%s> is standby\n", nl
->name
);
374 /* Prepare synchronization with first link */
375 tipc_link_tnl_prepare(ol
, nl
, SYNCH_MSG
, xmitq
);
379 * tipc_node_link_up - handle addition of link
381 * Link becomes active (alone or shared) or standby, depending on its priority.
383 static void tipc_node_link_up(struct tipc_node
*n
, int bearer_id
,
384 struct sk_buff_head
*xmitq
)
387 __tipc_node_link_up(n
, bearer_id
, xmitq
);
392 * __tipc_node_link_down - handle loss of link
394 static void __tipc_node_link_down(struct tipc_node
*n
, int *bearer_id
,
395 struct sk_buff_head
*xmitq
,
396 struct tipc_media_addr
**maddr
)
398 struct tipc_link_entry
*le
= &n
->links
[*bearer_id
];
399 int *slot0
= &n
->active_links
[0];
400 int *slot1
= &n
->active_links
[1];
402 struct tipc_link
*l
, *_l
, *tnl
;
404 l
= n
->links
[*bearer_id
].link
;
405 if (!l
|| tipc_link_is_reset(l
))
409 n
->action_flags
|= TIPC_NOTIFY_LINK_DOWN
;
410 n
->link_id
= l
->peer_bearer_id
<< 16 | *bearer_id
;
412 tipc_bearer_remove_dest(n
->net
, *bearer_id
, n
->addr
);
414 pr_debug("Lost link <%s> on network plane %c\n",
415 l
->name
, l
->net_plane
);
417 /* Select new active link if any available */
418 *slot0
= INVALID_BEARER_ID
;
419 *slot1
= INVALID_BEARER_ID
;
420 for (i
= 0; i
< MAX_BEARERS
; i
++) {
421 _l
= n
->links
[i
].link
;
422 if (!_l
|| !tipc_link_is_up(_l
))
426 if (_l
->priority
< highest
)
428 if (_l
->priority
> highest
) {
429 highest
= _l
->priority
;
437 if (!tipc_node_is_up(n
)) {
438 if (tipc_link_peer_is_down(l
))
439 tipc_node_fsm_evt(n
, PEER_LOST_CONTACT_EVT
);
440 tipc_node_fsm_evt(n
, SELF_LOST_CONTACT_EVT
);
441 tipc_link_fsm_evt(l
, LINK_RESET_EVT
);
443 tipc_link_build_reset_msg(l
, xmitq
);
444 *maddr
= &n
->links
[*bearer_id
].maddr
;
445 node_lost_contact(n
, &le
->inputq
);
449 /* There is still a working link => initiate failover */
450 tnl
= node_active_link(n
, 0);
451 tipc_link_fsm_evt(tnl
, LINK_SYNCH_END_EVT
);
452 tipc_node_fsm_evt(n
, NODE_SYNCH_END_EVT
);
453 n
->sync_point
= tnl
->rcv_nxt
+ (U16_MAX
/ 2 - 1);
454 tipc_link_tnl_prepare(l
, tnl
, FAILOVER_MSG
, xmitq
);
456 tipc_link_fsm_evt(l
, LINK_RESET_EVT
);
457 tipc_link_fsm_evt(l
, LINK_FAILOVER_BEGIN_EVT
);
458 tipc_node_fsm_evt(n
, NODE_FAILOVER_BEGIN_EVT
);
459 *maddr
= &n
->links
[tnl
->bearer_id
].maddr
;
460 *bearer_id
= tnl
->bearer_id
;
463 static void tipc_node_link_down(struct tipc_node
*n
, int bearer_id
, bool delete)
465 struct tipc_link_entry
*le
= &n
->links
[bearer_id
];
466 struct tipc_link
*l
= le
->link
;
467 struct tipc_media_addr
*maddr
;
468 struct sk_buff_head xmitq
;
473 __skb_queue_head_init(&xmitq
);
476 if (!tipc_link_is_establishing(l
)) {
477 __tipc_node_link_down(n
, &bearer_id
, &xmitq
, &maddr
);
484 /* Defuse pending tipc_node_link_up() */
485 tipc_link_fsm_evt(l
, LINK_RESET_EVT
);
488 tipc_bearer_xmit(n
->net
, bearer_id
, &xmitq
, maddr
);
489 tipc_sk_rcv(n
->net
, &le
->inputq
);
492 bool tipc_node_is_up(struct tipc_node
*n
)
494 return n
->active_links
[0] != INVALID_BEARER_ID
;
497 void tipc_node_check_dest(struct net
*net
, u32 onode
,
498 struct tipc_bearer
*b
,
499 u16 capabilities
, u32 signature
,
500 struct tipc_media_addr
*maddr
,
501 bool *respond
, bool *dupl_addr
)
505 struct tipc_link_entry
*le
;
506 bool addr_match
= false;
507 bool sign_match
= false;
508 bool link_up
= false;
509 bool accept_addr
= false;
516 n
= tipc_node_create(net
, onode
, capabilities
);
522 le
= &n
->links
[b
->identity
];
524 /* Prepare to validate requesting node's signature and media address */
526 link_up
= l
&& tipc_link_is_up(l
);
527 addr_match
= l
&& !memcmp(&le
->maddr
, maddr
, sizeof(*maddr
));
528 sign_match
= (signature
== n
->signature
);
530 /* These three flags give us eight permutations: */
532 if (sign_match
&& addr_match
&& link_up
) {
533 /* All is fine. Do nothing. */
535 } else if (sign_match
&& addr_match
&& !link_up
) {
536 /* Respond. The link will come up in due time */
538 } else if (sign_match
&& !addr_match
&& link_up
) {
539 /* Peer has changed i/f address without rebooting.
540 * If so, the link will reset soon, and the next
541 * discovery will be accepted. So we can ignore it.
542 * It may also be an cloned or malicious peer having
543 * chosen the same node address and signature as an
545 * Ignore requests until the link goes down, if ever.
548 } else if (sign_match
&& !addr_match
&& !link_up
) {
549 /* Peer link has changed i/f address without rebooting.
550 * It may also be a cloned or malicious peer; we can't
551 * distinguish between the two.
552 * The signature is correct, so we must accept.
556 } else if (!sign_match
&& addr_match
&& link_up
) {
557 /* Peer node rebooted. Two possibilities:
558 * - Delayed re-discovery; this link endpoint has already
559 * reset and re-established contact with the peer, before
560 * receiving a discovery message from that node.
561 * (The peer happened to receive one from this node first).
562 * - The peer came back so fast that our side has not
563 * discovered it yet. Probing from this side will soon
564 * reset the link, since there can be no working link
565 * endpoint at the peer end, and the link will re-establish.
566 * Accept the signature, since it comes from a known peer.
568 n
->signature
= signature
;
569 } else if (!sign_match
&& addr_match
&& !link_up
) {
570 /* The peer node has rebooted.
571 * Accept signature, since it is a known peer.
573 n
->signature
= signature
;
575 } else if (!sign_match
&& !addr_match
&& link_up
) {
576 /* Peer rebooted with new address, or a new/duplicate peer.
577 * Ignore until the link goes down, if ever.
580 } else if (!sign_match
&& !addr_match
&& !link_up
) {
581 /* Peer rebooted with new address, or it is a new peer.
582 * Accept signature and address.
584 n
->signature
= signature
;
592 /* Now create new link if not already existing */
594 if (n
->link_cnt
== 2) {
595 pr_warn("Cannot establish 3rd link to %x\n", n
->addr
);
598 if_name
= strchr(b
->name
, ':') + 1;
599 if (!tipc_link_create(n
, if_name
, b
->identity
, b
->tolerance
,
600 b
->net_plane
, b
->mtu
, b
->priority
,
601 b
->window
, mod(tipc_net(net
)->random
),
602 tipc_own_addr(net
), onode
,
603 n
->capabilities
, &le
->maddr
,
604 tipc_bc_sndlink(n
->net
), n
->bc_entry
.link
,
606 &n
->bc_entry
.namedq
, &l
)) {
611 tipc_link_fsm_evt(l
, LINK_RESET_EVT
);
612 if (n
->state
== NODE_FAILINGOVER
)
613 tipc_link_fsm_evt(l
, LINK_FAILOVER_BEGIN_EVT
);
616 tipc_node_calculate_timer(n
, l
);
617 if (n
->link_cnt
== 1)
618 if (!mod_timer(&n
->timer
, jiffies
+ n
->keepalive_intv
))
621 memcpy(&le
->maddr
, maddr
, sizeof(*maddr
));
624 if (reset
&& !tipc_link_is_reset(l
))
625 tipc_node_link_down(n
, b
->identity
, false);
629 void tipc_node_delete_links(struct net
*net
, int bearer_id
)
631 struct tipc_net
*tn
= net_generic(net
, tipc_net_id
);
635 list_for_each_entry_rcu(n
, &tn
->node_list
, list
) {
636 tipc_node_link_down(n
, bearer_id
, true);
641 static void tipc_node_reset_links(struct tipc_node
*n
)
643 char addr_string
[16];
646 pr_warn("Resetting all links to %s\n",
647 tipc_addr_string_fill(addr_string
, n
->addr
));
649 for (i
= 0; i
< MAX_BEARERS
; i
++) {
650 tipc_node_link_down(n
, i
, false);
654 /* tipc_node_fsm_evt - node finite state machine
655 * Determines when contact is allowed with peer node
657 static void tipc_node_fsm_evt(struct tipc_node
*n
, int evt
)
659 int state
= n
->state
;
662 case SELF_DOWN_PEER_DOWN
:
664 case SELF_ESTABL_CONTACT_EVT
:
665 state
= SELF_UP_PEER_COMING
;
667 case PEER_ESTABL_CONTACT_EVT
:
668 state
= SELF_COMING_PEER_UP
;
670 case SELF_LOST_CONTACT_EVT
:
671 case PEER_LOST_CONTACT_EVT
:
673 case NODE_SYNCH_END_EVT
:
674 case NODE_SYNCH_BEGIN_EVT
:
675 case NODE_FAILOVER_BEGIN_EVT
:
676 case NODE_FAILOVER_END_EVT
:
681 case SELF_UP_PEER_UP
:
683 case SELF_LOST_CONTACT_EVT
:
684 state
= SELF_DOWN_PEER_LEAVING
;
686 case PEER_LOST_CONTACT_EVT
:
687 state
= SELF_LEAVING_PEER_DOWN
;
689 case NODE_SYNCH_BEGIN_EVT
:
690 state
= NODE_SYNCHING
;
692 case NODE_FAILOVER_BEGIN_EVT
:
693 state
= NODE_FAILINGOVER
;
695 case SELF_ESTABL_CONTACT_EVT
:
696 case PEER_ESTABL_CONTACT_EVT
:
697 case NODE_SYNCH_END_EVT
:
698 case NODE_FAILOVER_END_EVT
:
704 case SELF_DOWN_PEER_LEAVING
:
706 case PEER_LOST_CONTACT_EVT
:
707 state
= SELF_DOWN_PEER_DOWN
;
709 case SELF_ESTABL_CONTACT_EVT
:
710 case PEER_ESTABL_CONTACT_EVT
:
711 case SELF_LOST_CONTACT_EVT
:
713 case NODE_SYNCH_END_EVT
:
714 case NODE_SYNCH_BEGIN_EVT
:
715 case NODE_FAILOVER_BEGIN_EVT
:
716 case NODE_FAILOVER_END_EVT
:
721 case SELF_UP_PEER_COMING
:
723 case PEER_ESTABL_CONTACT_EVT
:
724 state
= SELF_UP_PEER_UP
;
726 case SELF_LOST_CONTACT_EVT
:
727 state
= SELF_DOWN_PEER_LEAVING
;
729 case SELF_ESTABL_CONTACT_EVT
:
730 case PEER_LOST_CONTACT_EVT
:
731 case NODE_SYNCH_END_EVT
:
732 case NODE_FAILOVER_BEGIN_EVT
:
734 case NODE_SYNCH_BEGIN_EVT
:
735 case NODE_FAILOVER_END_EVT
:
740 case SELF_COMING_PEER_UP
:
742 case SELF_ESTABL_CONTACT_EVT
:
743 state
= SELF_UP_PEER_UP
;
745 case PEER_LOST_CONTACT_EVT
:
746 state
= SELF_LEAVING_PEER_DOWN
;
748 case SELF_LOST_CONTACT_EVT
:
749 case PEER_ESTABL_CONTACT_EVT
:
751 case NODE_SYNCH_END_EVT
:
752 case NODE_SYNCH_BEGIN_EVT
:
753 case NODE_FAILOVER_BEGIN_EVT
:
754 case NODE_FAILOVER_END_EVT
:
759 case SELF_LEAVING_PEER_DOWN
:
761 case SELF_LOST_CONTACT_EVT
:
762 state
= SELF_DOWN_PEER_DOWN
;
764 case SELF_ESTABL_CONTACT_EVT
:
765 case PEER_ESTABL_CONTACT_EVT
:
766 case PEER_LOST_CONTACT_EVT
:
768 case NODE_SYNCH_END_EVT
:
769 case NODE_SYNCH_BEGIN_EVT
:
770 case NODE_FAILOVER_BEGIN_EVT
:
771 case NODE_FAILOVER_END_EVT
:
776 case NODE_FAILINGOVER
:
778 case SELF_LOST_CONTACT_EVT
:
779 state
= SELF_DOWN_PEER_LEAVING
;
781 case PEER_LOST_CONTACT_EVT
:
782 state
= SELF_LEAVING_PEER_DOWN
;
784 case NODE_FAILOVER_END_EVT
:
785 state
= SELF_UP_PEER_UP
;
787 case NODE_FAILOVER_BEGIN_EVT
:
788 case SELF_ESTABL_CONTACT_EVT
:
789 case PEER_ESTABL_CONTACT_EVT
:
791 case NODE_SYNCH_BEGIN_EVT
:
792 case NODE_SYNCH_END_EVT
:
799 case SELF_LOST_CONTACT_EVT
:
800 state
= SELF_DOWN_PEER_LEAVING
;
802 case PEER_LOST_CONTACT_EVT
:
803 state
= SELF_LEAVING_PEER_DOWN
;
805 case NODE_SYNCH_END_EVT
:
806 state
= SELF_UP_PEER_UP
;
808 case NODE_FAILOVER_BEGIN_EVT
:
809 state
= NODE_FAILINGOVER
;
811 case NODE_SYNCH_BEGIN_EVT
:
812 case SELF_ESTABL_CONTACT_EVT
:
813 case PEER_ESTABL_CONTACT_EVT
:
815 case NODE_FAILOVER_END_EVT
:
821 pr_err("Unknown node fsm state %x\n", state
);
828 pr_err("Illegal node fsm evt %x in state %x\n", evt
, state
);
831 bool tipc_node_filter_pkt(struct tipc_node
*n
, struct tipc_msg
*hdr
)
833 int state
= n
->state
;
835 if (likely(state
== SELF_UP_PEER_UP
))
838 if (state
== SELF_LEAVING_PEER_DOWN
)
841 if (state
== SELF_DOWN_PEER_LEAVING
) {
842 if (msg_peer_node_is_up(hdr
))
849 static void node_lost_contact(struct tipc_node
*n
,
850 struct sk_buff_head
*inputq
)
852 char addr_string
[16];
853 struct tipc_sock_conn
*conn
, *safe
;
855 struct list_head
*conns
= &n
->conn_sks
;
859 pr_debug("Lost contact with %s\n",
860 tipc_addr_string_fill(addr_string
, n
->addr
));
862 /* Clean up broadcast state */
863 tipc_bcast_remove_peer(n
->net
, n
->addr
, n
->bc_entry
.link
);
865 /* Abort any ongoing link failover */
866 for (i
= 0; i
< MAX_BEARERS
; i
++) {
867 l
= n
->links
[i
].link
;
869 tipc_link_fsm_evt(l
, LINK_FAILOVER_END_EVT
);
872 /* Notify publications from this node */
873 n
->action_flags
|= TIPC_NOTIFY_NODE_DOWN
;
875 /* Notify sockets connected to node */
876 list_for_each_entry_safe(conn
, safe
, conns
, list
) {
877 skb
= tipc_msg_create(TIPC_CRITICAL_IMPORTANCE
, TIPC_CONN_MSG
,
878 SHORT_H_SIZE
, 0, tipc_own_addr(n
->net
),
879 conn
->peer_node
, conn
->port
,
880 conn
->peer_port
, TIPC_ERR_NO_NODE
);
882 skb_queue_tail(inputq
, skb
);
883 list_del(&conn
->list
);
889 * tipc_node_get_linkname - get the name of a link
891 * @bearer_id: id of the bearer
892 * @node: peer node address
893 * @linkname: link name output buffer
895 * Returns 0 on success
897 int tipc_node_get_linkname(struct net
*net
, u32 bearer_id
, u32 addr
,
898 char *linkname
, size_t len
)
900 struct tipc_link
*link
;
902 struct tipc_node
*node
= tipc_node_find(net
, addr
);
907 if (bearer_id
>= MAX_BEARERS
)
910 tipc_node_lock(node
);
911 link
= node
->links
[bearer_id
].link
;
913 strncpy(linkname
, link
->name
, len
);
917 tipc_node_unlock(node
);
922 void tipc_node_unlock(struct tipc_node
*node
)
924 struct net
*net
= node
->net
;
926 u32 flags
= node
->action_flags
;
928 struct list_head
*publ_list
;
930 if (likely(!flags
)) {
931 spin_unlock_bh(&node
->lock
);
936 link_id
= node
->link_id
;
937 publ_list
= &node
->publ_list
;
939 node
->action_flags
&= ~(TIPC_NOTIFY_NODE_DOWN
| TIPC_NOTIFY_NODE_UP
|
940 TIPC_NOTIFY_LINK_DOWN
| TIPC_NOTIFY_LINK_UP
|
941 TIPC_WAKEUP_BCAST_USERS
| TIPC_BCAST_MSG_EVT
|
944 spin_unlock_bh(&node
->lock
);
946 if (flags
& TIPC_NOTIFY_NODE_DOWN
)
947 tipc_publ_notify(net
, publ_list
, addr
);
949 if (flags
& TIPC_WAKEUP_BCAST_USERS
)
950 tipc_bclink_wakeup_users(net
);
952 if (flags
& TIPC_NOTIFY_NODE_UP
)
953 tipc_named_node_up(net
, addr
);
955 if (flags
& TIPC_NOTIFY_LINK_UP
)
956 tipc_nametbl_publish(net
, TIPC_LINK_STATE
, addr
, addr
,
957 TIPC_NODE_SCOPE
, link_id
, addr
);
959 if (flags
& TIPC_NOTIFY_LINK_DOWN
)
960 tipc_nametbl_withdraw(net
, TIPC_LINK_STATE
, addr
,
963 if (flags
& TIPC_BCAST_MSG_EVT
)
964 tipc_bclink_input(net
);
966 if (flags
& TIPC_BCAST_RESET
)
967 tipc_node_reset_links(node
);
970 /* Caller should hold node lock for the passed node */
971 static int __tipc_nl_add_node(struct tipc_nl_msg
*msg
, struct tipc_node
*node
)
974 struct nlattr
*attrs
;
976 hdr
= genlmsg_put(msg
->skb
, msg
->portid
, msg
->seq
, &tipc_genl_family
,
977 NLM_F_MULTI
, TIPC_NL_NODE_GET
);
981 attrs
= nla_nest_start(msg
->skb
, TIPC_NLA_NODE
);
985 if (nla_put_u32(msg
->skb
, TIPC_NLA_NODE_ADDR
, node
->addr
))
987 if (tipc_node_is_up(node
))
988 if (nla_put_flag(msg
->skb
, TIPC_NLA_NODE_UP
))
991 nla_nest_end(msg
->skb
, attrs
);
992 genlmsg_end(msg
->skb
, hdr
);
997 nla_nest_cancel(msg
->skb
, attrs
);
999 genlmsg_cancel(msg
->skb
, hdr
);
1004 static struct tipc_link
*tipc_node_select_link(struct tipc_node
*n
, int sel
,
1006 struct tipc_media_addr
**maddr
)
1008 int id
= n
->active_links
[sel
& 1];
1010 if (unlikely(id
< 0))
1014 *maddr
= &n
->links
[id
].maddr
;
1015 return n
->links
[id
].link
;
1019 * tipc_node_xmit() is the general link level function for message sending
1020 * @net: the applicable net namespace
1021 * @list: chain of buffers containing message
1022 * @dnode: address of destination node
1023 * @selector: a number used for deterministic link selection
1024 * Consumes the buffer chain, except when returning -ELINKCONG
1025 * Returns 0 if success, otherwise errno: -ELINKCONG,-EHOSTUNREACH,-EMSGSIZE
1027 int tipc_node_xmit(struct net
*net
, struct sk_buff_head
*list
,
1028 u32 dnode
, int selector
)
1030 struct tipc_link
*l
= NULL
;
1031 struct tipc_node
*n
;
1032 struct sk_buff_head xmitq
;
1033 struct tipc_media_addr
*maddr
;
1035 int rc
= -EHOSTUNREACH
;
1037 __skb_queue_head_init(&xmitq
);
1038 n
= tipc_node_find(net
, dnode
);
1041 l
= tipc_node_select_link(n
, selector
, &bearer_id
, &maddr
);
1043 rc
= tipc_link_xmit(l
, list
, &xmitq
);
1044 tipc_node_unlock(n
);
1045 if (unlikely(rc
== -ENOBUFS
))
1046 tipc_node_link_down(n
, bearer_id
, false);
1050 tipc_bearer_xmit(net
, bearer_id
, &xmitq
, maddr
);
1053 if (likely(in_own_node(net
, dnode
))) {
1054 tipc_sk_rcv(net
, list
);
1060 /* tipc_node_xmit_skb(): send single buffer to destination
1061 * Buffers sent via this functon are generally TIPC_SYSTEM_IMPORTANCE
1062 * messages, which will not be rejected
1063 * The only exception is datagram messages rerouted after secondary
1064 * lookup, which are rare and safe to dispose of anyway.
1065 * TODO: Return real return value, and let callers use
1066 * tipc_wait_for_sendpkt() where applicable
1068 int tipc_node_xmit_skb(struct net
*net
, struct sk_buff
*skb
, u32 dnode
,
1071 struct sk_buff_head head
;
1074 skb_queue_head_init(&head
);
1075 __skb_queue_tail(&head
, skb
);
1076 rc
= tipc_node_xmit(net
, &head
, dnode
, selector
);
1077 if (rc
== -ELINKCONG
)
1083 * tipc_node_bc_rcv - process TIPC broadcast packet arriving from off-node
1084 * @net: the applicable net namespace
1086 * @bearer_id: id of bearer message arrived on
1088 * Invoked with no locks held.
1090 void tipc_node_bc_rcv(struct net
*net
, struct sk_buff
*skb
, int bearer_id
)
1093 struct sk_buff_head xmitq
;
1094 struct tipc_bclink_entry
*be
;
1095 struct tipc_link_entry
*le
;
1096 struct tipc_msg
*hdr
= buf_msg(skb
);
1097 int usr
= msg_user(hdr
);
1098 u32 dnode
= msg_destnode(hdr
);
1099 struct tipc_node
*n
;
1101 __skb_queue_head_init(&xmitq
);
1103 /* If NACK for other node, let rcv link for that node peek into it */
1104 if ((usr
== BCAST_PROTOCOL
) && (dnode
!= tipc_own_addr(net
)))
1105 n
= tipc_node_find(net
, dnode
);
1107 n
= tipc_node_find(net
, msg_prevnode(hdr
));
1113 le
= &n
->links
[bearer_id
];
1115 rc
= tipc_bcast_rcv(net
, be
->link
, skb
);
1117 /* Broadcast link reset may happen at reassembly failure */
1118 if (rc
& TIPC_LINK_DOWN_EVT
)
1119 tipc_node_reset_links(n
);
1121 /* Broadcast ACKs are sent on a unicast link */
1122 if (rc
& TIPC_LINK_SND_BC_ACK
) {
1124 tipc_link_build_ack_msg(le
->link
, &xmitq
);
1125 tipc_node_unlock(n
);
1128 if (!skb_queue_empty(&xmitq
))
1129 tipc_bearer_xmit(net
, bearer_id
, &xmitq
, &le
->maddr
);
1131 /* Deliver. 'arrvq' is under inputq2's lock protection */
1132 if (!skb_queue_empty(&be
->inputq1
)) {
1133 spin_lock_bh(&be
->inputq2
.lock
);
1134 spin_lock_bh(&be
->inputq1
.lock
);
1135 skb_queue_splice_tail_init(&be
->inputq1
, &be
->arrvq
);
1136 spin_unlock_bh(&be
->inputq1
.lock
);
1137 spin_unlock_bh(&be
->inputq2
.lock
);
1138 tipc_sk_mcast_rcv(net
, &be
->arrvq
, &be
->inputq2
);
1144 * tipc_node_check_state - check and if necessary update node state
1146 * @bearer_id: identity of bearer delivering the packet
1147 * Returns true if state is ok, otherwise consumes buffer and returns false
1149 static bool tipc_node_check_state(struct tipc_node
*n
, struct sk_buff
*skb
,
1150 int bearer_id
, struct sk_buff_head
*xmitq
)
1152 struct tipc_msg
*hdr
= buf_msg(skb
);
1153 int usr
= msg_user(hdr
);
1154 int mtyp
= msg_type(hdr
);
1155 u16 oseqno
= msg_seqno(hdr
);
1156 u16 iseqno
= msg_seqno(msg_get_wrapped(hdr
));
1157 u16 exp_pkts
= msg_msgcnt(hdr
);
1158 u16 rcv_nxt
, syncpt
, dlv_nxt
;
1159 int state
= n
->state
;
1160 struct tipc_link
*l
, *tnl
, *pl
= NULL
;
1161 struct tipc_media_addr
*maddr
;
1164 l
= n
->links
[bearer_id
].link
;
1167 rcv_nxt
= l
->rcv_nxt
;
1170 if (likely((state
== SELF_UP_PEER_UP
) && (usr
!= TUNNEL_PROTOCOL
)))
1173 /* Find parallel link, if any */
1174 for (i
= 0; i
< MAX_BEARERS
; i
++) {
1175 if ((i
!= bearer_id
) && n
->links
[i
].link
) {
1176 pl
= n
->links
[i
].link
;
1181 /* Update node accesibility if applicable */
1182 if (state
== SELF_UP_PEER_COMING
) {
1183 if (!tipc_link_is_up(l
))
1185 if (!msg_peer_link_is_up(hdr
))
1187 tipc_node_fsm_evt(n
, PEER_ESTABL_CONTACT_EVT
);
1190 if (state
== SELF_DOWN_PEER_LEAVING
) {
1191 if (msg_peer_node_is_up(hdr
))
1193 tipc_node_fsm_evt(n
, PEER_LOST_CONTACT_EVT
);
1196 /* Ignore duplicate packets */
1197 if ((usr
!= LINK_PROTOCOL
) && less(oseqno
, rcv_nxt
))
1200 /* Initiate or update failover mode if applicable */
1201 if ((usr
== TUNNEL_PROTOCOL
) && (mtyp
== FAILOVER_MSG
)) {
1202 syncpt
= oseqno
+ exp_pkts
- 1;
1203 if (pl
&& tipc_link_is_up(pl
)) {
1204 pb_id
= pl
->bearer_id
;
1205 __tipc_node_link_down(n
, &pb_id
, xmitq
, &maddr
);
1206 tipc_skb_queue_splice_tail_init(pl
->inputq
, l
->inputq
);
1208 /* If pkts arrive out of order, use lowest calculated syncpt */
1209 if (less(syncpt
, n
->sync_point
))
1210 n
->sync_point
= syncpt
;
1213 /* Open parallel link when tunnel link reaches synch point */
1214 if ((n
->state
== NODE_FAILINGOVER
) && tipc_link_is_up(l
)) {
1215 if (!more(rcv_nxt
, n
->sync_point
))
1217 tipc_node_fsm_evt(n
, NODE_FAILOVER_END_EVT
);
1219 tipc_link_fsm_evt(pl
, LINK_FAILOVER_END_EVT
);
1223 /* No synching needed if only one link */
1224 if (!pl
|| !tipc_link_is_up(pl
))
1227 /* Initiate synch mode if applicable */
1228 if ((usr
== TUNNEL_PROTOCOL
) && (mtyp
== SYNCH_MSG
) && (oseqno
== 1)) {
1229 syncpt
= iseqno
+ exp_pkts
- 1;
1230 if (!tipc_link_is_up(l
)) {
1231 tipc_link_fsm_evt(l
, LINK_ESTABLISH_EVT
);
1232 __tipc_node_link_up(n
, bearer_id
, xmitq
);
1234 if (n
->state
== SELF_UP_PEER_UP
) {
1235 n
->sync_point
= syncpt
;
1236 tipc_link_fsm_evt(l
, LINK_SYNCH_BEGIN_EVT
);
1237 tipc_node_fsm_evt(n
, NODE_SYNCH_BEGIN_EVT
);
1239 if (less(syncpt
, n
->sync_point
))
1240 n
->sync_point
= syncpt
;
1243 /* Open tunnel link when parallel link reaches synch point */
1244 if ((n
->state
== NODE_SYNCHING
) && tipc_link_is_synching(l
)) {
1245 if (tipc_link_is_synching(l
)) {
1251 dlv_nxt
= pl
->rcv_nxt
- mod(skb_queue_len(pl
->inputq
));
1252 if (more(dlv_nxt
, n
->sync_point
)) {
1253 tipc_link_fsm_evt(tnl
, LINK_SYNCH_END_EVT
);
1254 tipc_node_fsm_evt(n
, NODE_SYNCH_END_EVT
);
1259 if ((usr
== TUNNEL_PROTOCOL
) && (mtyp
== SYNCH_MSG
))
1261 if (usr
== LINK_PROTOCOL
)
1269 * tipc_rcv - process TIPC packets/messages arriving from off-node
1270 * @net: the applicable net namespace
1272 * @bearer: pointer to bearer message arrived on
1274 * Invoked with no locks held. Bearer pointer must point to a valid bearer
1275 * structure (i.e. cannot be NULL), but bearer can be inactive.
1277 void tipc_rcv(struct net
*net
, struct sk_buff
*skb
, struct tipc_bearer
*b
)
1279 struct sk_buff_head xmitq
;
1280 struct tipc_node
*n
;
1281 struct tipc_msg
*hdr
= buf_msg(skb
);
1282 int usr
= msg_user(hdr
);
1283 int bearer_id
= b
->identity
;
1284 struct tipc_link_entry
*le
;
1285 u16 bc_ack
= msg_bcast_ack(hdr
);
1288 __skb_queue_head_init(&xmitq
);
1290 /* Ensure message is well-formed */
1291 if (unlikely(!tipc_msg_validate(skb
)))
1294 /* Handle arrival of discovery or broadcast packet */
1295 if (unlikely(msg_non_seq(hdr
))) {
1296 if (unlikely(usr
== LINK_CONFIG
))
1297 return tipc_disc_rcv(net
, skb
, b
);
1299 return tipc_node_bc_rcv(net
, skb
, bearer_id
);
1302 /* Locate neighboring node that sent packet */
1303 n
= tipc_node_find(net
, msg_prevnode(hdr
));
1306 le
= &n
->links
[bearer_id
];
1308 /* Ensure broadcast reception is in synch with peer's send state */
1309 if (unlikely(usr
== LINK_PROTOCOL
))
1310 tipc_bcast_sync_rcv(net
, n
->bc_entry
.link
, hdr
);
1311 else if (unlikely(n
->bc_entry
.link
->acked
!= bc_ack
))
1312 tipc_bcast_ack_rcv(net
, n
->bc_entry
.link
, bc_ack
);
1316 /* Is reception permitted at the moment ? */
1317 if (!tipc_node_filter_pkt(n
, hdr
))
1320 /* Check and if necessary update node state */
1321 if (likely(tipc_node_check_state(n
, skb
, bearer_id
, &xmitq
))) {
1322 rc
= tipc_link_rcv(le
->link
, skb
, &xmitq
);
1326 tipc_node_unlock(n
);
1328 if (unlikely(rc
& TIPC_LINK_UP_EVT
))
1329 tipc_node_link_up(n
, bearer_id
, &xmitq
);
1331 if (unlikely(rc
& TIPC_LINK_DOWN_EVT
))
1332 tipc_node_link_down(n
, bearer_id
, false);
1334 if (unlikely(!skb_queue_empty(&n
->bc_entry
.namedq
)))
1335 tipc_named_rcv(net
, &n
->bc_entry
.namedq
);
1337 if (!skb_queue_empty(&le
->inputq
))
1338 tipc_sk_rcv(net
, &le
->inputq
);
1340 if (!skb_queue_empty(&xmitq
))
1341 tipc_bearer_xmit(net
, bearer_id
, &xmitq
, &le
->maddr
);
1348 int tipc_nl_node_dump(struct sk_buff
*skb
, struct netlink_callback
*cb
)
1351 struct net
*net
= sock_net(skb
->sk
);
1352 struct tipc_net
*tn
= net_generic(net
, tipc_net_id
);
1353 int done
= cb
->args
[0];
1354 int last_addr
= cb
->args
[1];
1355 struct tipc_node
*node
;
1356 struct tipc_nl_msg msg
;
1362 msg
.portid
= NETLINK_CB(cb
->skb
).portid
;
1363 msg
.seq
= cb
->nlh
->nlmsg_seq
;
1367 node
= tipc_node_find(net
, last_addr
);
1370 /* We never set seq or call nl_dump_check_consistent()
1371 * this means that setting prev_seq here will cause the
1372 * consistence check to fail in the netlink callback
1373 * handler. Resulting in the NLMSG_DONE message having
1374 * the NLM_F_DUMP_INTR flag set if the node state
1375 * changed while we released the lock.
1380 tipc_node_put(node
);
1383 list_for_each_entry_rcu(node
, &tn
->node_list
, list
) {
1385 if (node
->addr
== last_addr
)
1391 tipc_node_lock(node
);
1392 err
= __tipc_nl_add_node(&msg
, node
);
1394 last_addr
= node
->addr
;
1395 tipc_node_unlock(node
);
1399 tipc_node_unlock(node
);
1404 cb
->args
[1] = last_addr
;