node.c 59.4 KB
Newer Older
P
Per Liden 已提交
1 2
/*
 * net/tipc/node.c: TIPC node management routines
3
 *
4
 * Copyright (c) 2000-2006, 2012-2016, Ericsson AB
5
 * Copyright (c) 2005-2006, 2010-2014, Wind River Systems
P
Per Liden 已提交
6 7
 * All rights reserved.
 *
P
Per Liden 已提交
8
 * Redistribution and use in source and binary forms, with or without
P
Per Liden 已提交
9 10
 * modification, are permitted provided that the following conditions are met:
 *
P
Per Liden 已提交
11 12 13 14 15 16 17 18
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. Neither the names of the copyright holders nor the names of its
 *    contributors may be used to endorse or promote products derived from
 *    this software without specific prior written permission.
P
Per Liden 已提交
19
 *
P
Per Liden 已提交
20 21 22 23 24 25 26 27 28 29 30 31 32 33
 * Alternatively, this software may be distributed under the terms of the
 * GNU General Public License ("GPL") version 2 as published by the Free
 * Software Foundation.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
P
Per Liden 已提交
34 35 36 37
 * POSSIBILITY OF SUCH DAMAGE.
 */

#include "core.h"
38
#include "link.h"
P
Per Liden 已提交
39 40
#include "node.h"
#include "name_distr.h"
41
#include "socket.h"
42
#include "bcast.h"
43
#include "monitor.h"
44
#include "discover.h"
45
#include "netlink.h"
46

47
#define INVALID_NODE_SIG	0x10000
48
#define NODE_CLEANUP_AFTER	300000
49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99

/* Flags used to take different actions according to flag type
 * TIPC_NOTIFY_NODE_DOWN: notify node is down
 * TIPC_NOTIFY_NODE_UP: notify node is up
 * TIPC_DISTRIBUTE_NAME: publish or withdraw link state name type
 */
enum {
	TIPC_NOTIFY_NODE_DOWN		= (1 << 3),
	TIPC_NOTIFY_NODE_UP		= (1 << 4),
	TIPC_NOTIFY_LINK_UP		= (1 << 6),
	TIPC_NOTIFY_LINK_DOWN		= (1 << 7)
};

struct tipc_link_entry {
	struct tipc_link *link;
	spinlock_t lock; /* per link */
	u32 mtu;
	struct sk_buff_head inputq;
	struct tipc_media_addr maddr;
};

struct tipc_bclink_entry {
	struct tipc_link *link;
	struct sk_buff_head inputq1;
	struct sk_buff_head arrvq;
	struct sk_buff_head inputq2;
	struct sk_buff_head namedq;
};

/**
 * struct tipc_node - TIPC node structure
 * @addr: network address of node
 * @ref: reference counter to node object
 * @lock: rwlock governing access to structure
 * @net: the applicable net namespace
 * @hash: links to adjacent nodes in unsorted hash chain
 * @inputq: pointer to input queue containing messages for msg event
 * @namedq: pointer to name table input queue with name table messages
 * @active_links: bearer ids of active links, used as index into links[] array
 * @links: array containing references to all links to node
 * @action_flags: bit mask of different types of node actions
 * @state: connectivity state vs peer node
 * @sync_point: sequence number where synch/failover is finished
 * @list: links to adjacent nodes in sorted list of cluster's nodes
 * @working_links: number of working links to node (both active and standby)
 * @link_cnt: number of links to node
 * @capabilities: bitmap, indicating peer node's functional capabilities
 * @signature: node instance identifier
 * @link_id: local and remote bearer ids of changing link, if any
 * @publ_list: list of publications
 * @rcu: rcu struct for tipc_node
100
 * @delete_at: indicates the time for deleting a down node
101 102 103 104 105 106 107 108 109 110 111 112 113
 */
struct tipc_node {
	u32 addr;
	struct kref kref;
	rwlock_t lock;
	struct net *net;
	struct hlist_node hash;
	int active_links[2];
	struct tipc_link_entry links[MAX_BEARERS];
	struct tipc_bclink_entry bc_entry;
	int action_flags;
	struct list_head list;
	int state;
L
LUU Duc Canh 已提交
114
	bool failover_sent;
115 116 117 118 119 120
	u16 sync_point;
	int link_cnt;
	u16 working_links;
	u16 capabilities;
	u32 signature;
	u32 link_id;
121
	u8 peer_id[16];
122 123 124 125 126
	struct list_head publ_list;
	struct list_head conn_sks;
	unsigned long keepalive_intv;
	struct timer_list timer;
	struct rcu_head rcu;
127
	unsigned long delete_at;
128 129
};

130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153
/* Node FSM states and events:
 */
enum {
	SELF_DOWN_PEER_DOWN    = 0xdd,
	SELF_UP_PEER_UP        = 0xaa,
	SELF_DOWN_PEER_LEAVING = 0xd1,
	SELF_UP_PEER_COMING    = 0xac,
	SELF_COMING_PEER_UP    = 0xca,
	SELF_LEAVING_PEER_DOWN = 0x1d,
	NODE_FAILINGOVER       = 0xf0,
	NODE_SYNCHING          = 0xcc
};

enum {
	SELF_ESTABL_CONTACT_EVT = 0xece,
	SELF_LOST_CONTACT_EVT   = 0x1ce,
	PEER_ESTABL_CONTACT_EVT = 0x9ece,
	PEER_LOST_CONTACT_EVT   = 0x91ce,
	NODE_FAILOVER_BEGIN_EVT = 0xfbe,
	NODE_FAILOVER_END_EVT   = 0xfee,
	NODE_SYNCH_BEGIN_EVT    = 0xcbe,
	NODE_SYNCH_END_EVT      = 0xcee
};

154 155 156 157 158 159
static void __tipc_node_link_down(struct tipc_node *n, int *bearer_id,
				  struct sk_buff_head *xmitq,
				  struct tipc_media_addr **maddr);
static void tipc_node_link_down(struct tipc_node *n, int bearer_id,
				bool delete);
static void node_lost_contact(struct tipc_node *n, struct sk_buff_head *inputq);
160
static void tipc_node_delete(struct tipc_node *node);
161
static void tipc_node_timeout(struct timer_list *t);
162
static void tipc_node_fsm_evt(struct tipc_node *n, int evt);
163
static struct tipc_node *tipc_node_find(struct net *net, u32 addr);
164
static struct tipc_node *tipc_node_find_by_id(struct net *net, u8 *id);
165
static void tipc_node_put(struct tipc_node *node);
166
static bool node_is_up(struct tipc_node *n);
167
static void tipc_node_delete_from_list(struct tipc_node *node);
P
Per Liden 已提交
168

169 170 171 172 173 174 175
struct tipc_sock_conn {
	u32 port;
	u32 peer_port;
	u32 peer_node;
	struct list_head list;
};

176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201
static struct tipc_link *node_active_link(struct tipc_node *n, int sel)
{
	int bearer_id = n->active_links[sel & 1];

	if (unlikely(bearer_id == INVALID_BEARER_ID))
		return NULL;

	return n->links[bearer_id].link;
}

int tipc_node_get_mtu(struct net *net, u32 addr, u32 sel)
{
	struct tipc_node *n;
	int bearer_id;
	unsigned int mtu = MAX_MSG_SIZE;

	n = tipc_node_find(net, addr);
	if (unlikely(!n))
		return mtu;

	bearer_id = n->active_links[sel & 1];
	if (likely(bearer_id != INVALID_BEARER_ID))
		mtu = n->links[bearer_id].mtu;
	tipc_node_put(n);
	return mtu;
}
202

203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223
bool tipc_node_get_id(struct net *net, u32 addr, u8 *id)
{
	u8 *own_id = tipc_own_id(net);
	struct tipc_node *n;

	if (!own_id)
		return true;

	if (addr == tipc_own_addr(net)) {
		memcpy(id, own_id, TIPC_NODEID_LEN);
		return true;
	}
	n = tipc_node_find(net, addr);
	if (!n)
		return false;

	memcpy(id, &n->peer_id, TIPC_NODEID_LEN);
	tipc_node_put(n);
	return true;
}

224 225 226 227 228 229 230 231 232 233 234 235 236
u16 tipc_node_get_capabilities(struct net *net, u32 addr)
{
	struct tipc_node *n;
	u16 caps;

	n = tipc_node_find(net, addr);
	if (unlikely(!n))
		return TIPC_NODE_CAPABILITIES;
	caps = n->capabilities;
	tipc_node_put(n);
	return caps;
}

237 238
static void tipc_node_kref_release(struct kref *kref)
{
239
	struct tipc_node *n = container_of(kref, struct tipc_node, kref);
240

241 242
	kfree(n->bc_entry.link);
	kfree_rcu(n, rcu);
243 244
}

245
static void tipc_node_put(struct tipc_node *node)
246 247 248 249 250 251 252 253 254
{
	kref_put(&node->kref, tipc_node_kref_release);
}

static void tipc_node_get(struct tipc_node *node)
{
	kref_get(&node->kref);
}

255
/*
256 257
 * tipc_node_find - locate specified node object, if it exists
 */
258
static struct tipc_node *tipc_node_find(struct net *net, u32 addr)
259
{
260
	struct tipc_net *tn = tipc_net(net);
261
	struct tipc_node *node;
262
	unsigned int thash = tipc_hashfn(addr);
263

264
	rcu_read_lock();
265 266 267 268 269 270
	hlist_for_each_entry_rcu(node, &tn->node_htable[thash], hash) {
		if (node->addr != addr)
			continue;
		if (!kref_get_unless_zero(&node->kref))
			node = NULL;
		break;
271
	}
272
	rcu_read_unlock();
273
	return node;
274 275
}

276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299
/* tipc_node_find_by_id - locate specified node object by its 128-bit id
 * Note: this function is called only when a discovery request failed
 * to find the node by its 32-bit id, and is not time critical
 */
static struct tipc_node *tipc_node_find_by_id(struct net *net, u8 *id)
{
	struct tipc_net *tn = tipc_net(net);
	struct tipc_node *n;
	bool found = false;

	rcu_read_lock();
	list_for_each_entry_rcu(n, &tn->node_list, list) {
		read_lock_bh(&n->lock);
		if (!memcmp(id, n->peer_id, 16) &&
		    kref_get_unless_zero(&n->kref))
			found = true;
		read_unlock_bh(&n->lock);
		if (found)
			break;
	}
	rcu_read_unlock();
	return found ? n : NULL;
}

300
static void tipc_node_read_lock(struct tipc_node *n)
J
Jon Paul Maloy 已提交
301 302 303 304
{
	read_lock_bh(&n->lock);
}

305
static void tipc_node_read_unlock(struct tipc_node *n)
J
Jon Paul Maloy 已提交
306 307 308 309 310 311 312 313 314
{
	read_unlock_bh(&n->lock);
}

static void tipc_node_write_lock(struct tipc_node *n)
{
	write_lock_bh(&n->lock);
}

315 316 317 318 319
static void tipc_node_write_unlock_fast(struct tipc_node *n)
{
	write_unlock_bh(&n->lock);
}

J
Jon Paul Maloy 已提交
320 321 322 323 324 325
static void tipc_node_write_unlock(struct tipc_node *n)
{
	struct net *net = n->net;
	u32 addr = 0;
	u32 flags = n->action_flags;
	u32 link_id = 0;
326
	u32 bearer_id;
J
Jon Paul Maloy 已提交
327 328 329 330 331 332 333 334 335
	struct list_head *publ_list;

	if (likely(!flags)) {
		write_unlock_bh(&n->lock);
		return;
	}

	addr = n->addr;
	link_id = n->link_id;
336
	bearer_id = link_id & 0xffff;
J
Jon Paul Maloy 已提交
337 338 339 340 341 342 343 344 345 346 347 348 349
	publ_list = &n->publ_list;

	n->action_flags &= ~(TIPC_NOTIFY_NODE_DOWN | TIPC_NOTIFY_NODE_UP |
			     TIPC_NOTIFY_LINK_DOWN | TIPC_NOTIFY_LINK_UP);

	write_unlock_bh(&n->lock);

	if (flags & TIPC_NOTIFY_NODE_DOWN)
		tipc_publ_notify(net, publ_list, addr);

	if (flags & TIPC_NOTIFY_NODE_UP)
		tipc_named_node_up(net, addr);

350 351
	if (flags & TIPC_NOTIFY_LINK_UP) {
		tipc_mon_peer_up(net, addr, bearer_id);
J
Jon Paul Maloy 已提交
352
		tipc_nametbl_publish(net, TIPC_LINK_STATE, addr, addr,
353
				     TIPC_NODE_SCOPE, link_id, link_id);
354 355 356
	}
	if (flags & TIPC_NOTIFY_LINK_DOWN) {
		tipc_mon_peer_down(net, addr, bearer_id);
J
Jon Paul Maloy 已提交
357
		tipc_nametbl_withdraw(net, TIPC_LINK_STATE, addr,
358
				      addr, link_id);
359
	}
J
Jon Paul Maloy 已提交
360 361
}

362 363
static struct tipc_node *tipc_node_create(struct net *net, u32 addr,
					  u8 *peer_id, u16 capabilities)
P
Per Liden 已提交
364
{
365
	struct tipc_net *tn = net_generic(net, tipc_net_id);
366
	struct tipc_node *n, *temp_node;
367 368
	struct tipc_link *l;
	int bearer_id;
J
Jon Paul Maloy 已提交
369
	int i;
P
Per Liden 已提交
370

371
	spin_lock_bh(&tn->node_list_lock);
372
	n = tipc_node_find(net, addr);
373
	if (n) {
374 375
		if (n->capabilities == capabilities)
			goto exit;
376
		/* Same node may come back with new capabilities */
377
		write_lock_bh(&n->lock);
378
		n->capabilities = capabilities;
379 380 381 382 383
		for (bearer_id = 0; bearer_id < MAX_BEARERS; bearer_id++) {
			l = n->links[bearer_id].link;
			if (l)
				tipc_link_update_caps(l, capabilities);
		}
384
		write_unlock_bh(&n->lock);
385
		goto exit;
386
	}
387 388
	n = kzalloc(sizeof(*n), GFP_ATOMIC);
	if (!n) {
389
		pr_warn("Node creation failed, no memory\n");
390
		goto exit;
391
	}
392
	n->addr = addr;
393
	memcpy(&n->peer_id, peer_id, 16);
394 395 396 397 398 399 400 401 402 403 404 405
	n->net = net;
	n->capabilities = capabilities;
	kref_init(&n->kref);
	rwlock_init(&n->lock);
	INIT_HLIST_NODE(&n->hash);
	INIT_LIST_HEAD(&n->list);
	INIT_LIST_HEAD(&n->publ_list);
	INIT_LIST_HEAD(&n->conn_sks);
	skb_queue_head_init(&n->bc_entry.namedq);
	skb_queue_head_init(&n->bc_entry.inputq1);
	__skb_queue_head_init(&n->bc_entry.arrvq);
	skb_queue_head_init(&n->bc_entry.inputq2);
J
Jon Paul Maloy 已提交
406
	for (i = 0; i < MAX_BEARERS; i++)
407 408
		spin_lock_init(&n->links[i].lock);
	n->state = SELF_DOWN_PEER_LEAVING;
409
	n->delete_at = jiffies + msecs_to_jiffies(NODE_CLEANUP_AFTER);
410 411 412
	n->signature = INVALID_NODE_SIG;
	n->active_links[0] = INVALID_BEARER_ID;
	n->active_links[1] = INVALID_BEARER_ID;
413 414
	if (!tipc_link_bc_create(net, tipc_own_addr(net),
				 addr, U16_MAX,
415
				 tipc_link_window(tipc_bc_sndlink(net)),
416 417 418
				 n->capabilities,
				 &n->bc_entry.inputq1,
				 &n->bc_entry.namedq,
419
				 tipc_bc_sndlink(net),
420
				 &n->bc_entry.link)) {
421
		pr_warn("Broadcast rcv link creation failed, no memory\n");
422 423
		kfree(n);
		n = NULL;
424 425
		goto exit;
	}
426
	tipc_node_get(n);
427
	timer_setup(&n->timer, tipc_node_timeout, 0);
428
	n->keepalive_intv = U32_MAX;
429 430 431 432 433 434
	hlist_add_head_rcu(&n->hash, &tn->node_htable[tipc_hashfn(addr)]);
	list_for_each_entry_rcu(temp_node, &tn->node_list, list) {
		if (n->addr < temp_node->addr)
			break;
	}
	list_add_tail_rcu(&n->list, &temp_node->list);
435
exit:
436
	spin_unlock_bh(&tn->node_list_lock);
437
	return n;
P
Per Liden 已提交
438 439
}

440 441
static void tipc_node_calculate_timer(struct tipc_node *n, struct tipc_link *l)
{
442
	unsigned long tol = tipc_link_tolerance(l);
443 444 445
	unsigned long intv = ((tol / 4) > 500) ? 500 : tol / 4;

	/* Link with lowest tolerance determines timer interval */
446 447
	if (intv < n->keepalive_intv)
		n->keepalive_intv = intv;
448

449 450
	/* Ensure link's abort limit corresponds to current tolerance */
	tipc_link_set_abort_limit(l, tol / n->keepalive_intv);
451 452
}

453
static void tipc_node_delete_from_list(struct tipc_node *node)
P
Per Liden 已提交
454
{
455 456
	list_del_rcu(&node->list);
	hlist_del_rcu(&node->hash);
457
	tipc_node_put(node);
458 459 460 461 462
}

static void tipc_node_delete(struct tipc_node *node)
{
	tipc_node_delete_from_list(node);
463 464 465

	del_timer_sync(&node->timer);
	tipc_node_put(node);
P
Per Liden 已提交
466 467
}

468
void tipc_node_stop(struct net *net)
469
{
470
	struct tipc_net *tn = tipc_net(net);
471 472
	struct tipc_node *node, *t_node;

473
	spin_lock_bh(&tn->node_list_lock);
474 475
	list_for_each_entry_safe(node, t_node, &tn->node_list, list)
		tipc_node_delete(node);
476
	spin_unlock_bh(&tn->node_list_lock);
477 478
}

479 480 481 482 483 484 485 486 487 488 489 490
void tipc_node_subscribe(struct net *net, struct list_head *subscr, u32 addr)
{
	struct tipc_node *n;

	if (in_own_node(net, addr))
		return;

	n = tipc_node_find(net, addr);
	if (!n) {
		pr_warn("Node subscribe rejected, unknown node 0x%x\n", addr);
		return;
	}
J
Jon Paul Maloy 已提交
491
	tipc_node_write_lock(n);
492
	list_add_tail(subscr, &n->publ_list);
493
	tipc_node_write_unlock_fast(n);
494 495 496 497 498 499 500 501 502 503 504 505 506 507 508
	tipc_node_put(n);
}

void tipc_node_unsubscribe(struct net *net, struct list_head *subscr, u32 addr)
{
	struct tipc_node *n;

	if (in_own_node(net, addr))
		return;

	n = tipc_node_find(net, addr);
	if (!n) {
		pr_warn("Node unsubscribe rejected, unknown node 0x%x\n", addr);
		return;
	}
J
Jon Paul Maloy 已提交
509
	tipc_node_write_lock(n);
510
	list_del_init(subscr);
511
	tipc_node_write_unlock_fast(n);
512 513 514
	tipc_node_put(n);
}

515
int tipc_node_add_conn(struct net *net, u32 dnode, u32 port, u32 peer_port)
516 517 518
{
	struct tipc_node *node;
	struct tipc_sock_conn *conn;
519
	int err = 0;
520

521
	if (in_own_node(net, dnode))
522 523
		return 0;

524
	node = tipc_node_find(net, dnode);
525 526 527 528 529
	if (!node) {
		pr_warn("Connecting sock to node 0x%x failed\n", dnode);
		return -EHOSTUNREACH;
	}
	conn = kmalloc(sizeof(*conn), GFP_ATOMIC);
530 531 532 533
	if (!conn) {
		err = -EHOSTUNREACH;
		goto exit;
	}
534 535 536 537
	conn->peer_node = dnode;
	conn->port = port;
	conn->peer_port = peer_port;

J
Jon Paul Maloy 已提交
538
	tipc_node_write_lock(node);
539
	list_add_tail(&conn->list, &node->conn_sks);
J
Jon Paul Maloy 已提交
540
	tipc_node_write_unlock(node);
541 542 543
exit:
	tipc_node_put(node);
	return err;
544 545
}

546
void tipc_node_remove_conn(struct net *net, u32 dnode, u32 port)
547 548 549 550
{
	struct tipc_node *node;
	struct tipc_sock_conn *conn, *safe;

551
	if (in_own_node(net, dnode))
552 553
		return;

554
	node = tipc_node_find(net, dnode);
555 556 557
	if (!node)
		return;

J
Jon Paul Maloy 已提交
558
	tipc_node_write_lock(node);
559 560 561 562 563 564
	list_for_each_entry_safe(conn, safe, &node->conn_sks, list) {
		if (port != conn->port)
			continue;
		list_del(&conn->list);
		kfree(conn);
	}
J
Jon Paul Maloy 已提交
565
	tipc_node_write_unlock(node);
566
	tipc_node_put(node);
567 568
}

569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586
static void  tipc_node_clear_links(struct tipc_node *node)
{
	int i;

	for (i = 0; i < MAX_BEARERS; i++) {
		struct tipc_link_entry *le = &node->links[i];

		if (le->link) {
			kfree(le->link);
			le->link = NULL;
			node->link_cnt--;
		}
	}
}

/* tipc_node_cleanup - delete nodes that does not
 * have active links for NODE_CLEANUP_AFTER time
 */
587
static bool tipc_node_cleanup(struct tipc_node *peer)
588 589 590 591
{
	struct tipc_net *tn = tipc_net(peer->net);
	bool deleted = false;

592 593 594 595
	/* If lock held by tipc_node_stop() the node will be deleted anyway */
	if (!spin_trylock_bh(&tn->node_list_lock))
		return false;

596 597 598 599 600 601 602 603 604 605 606 607
	tipc_node_write_lock(peer);

	if (!node_is_up(peer) && time_after(jiffies, peer->delete_at)) {
		tipc_node_clear_links(peer);
		tipc_node_delete_from_list(peer);
		deleted = true;
	}
	tipc_node_write_unlock(peer);
	spin_unlock_bh(&tn->node_list_lock);
	return deleted;
}

608 609
/* tipc_node_timeout - handle expiration of node timer
 */
610
static void tipc_node_timeout(struct timer_list *t)
611
{
612
	struct tipc_node *n = from_timer(n, t, timer);
613
	struct tipc_link_entry *le;
614
	struct sk_buff_head xmitq;
615
	int remains = n->link_cnt;
616 617 618
	int bearer_id;
	int rc = 0;

619 620 621 622 623 624
	if (!node_is_up(n) && tipc_node_cleanup(n)) {
		/*Removing the reference of Timer*/
		tipc_node_put(n);
		return;
	}

625 626
	__skb_queue_head_init(&xmitq);

627 628 629 630 631 632
	/* Initial node interval to value larger (10 seconds), then it will be
	 * recalculated with link lowest tolerance
	 */
	tipc_node_read_lock(n);
	n->keepalive_intv = 10000;
	tipc_node_read_unlock(n);
633
	for (bearer_id = 0; remains && (bearer_id < MAX_BEARERS); bearer_id++) {
J
Jon Paul Maloy 已提交
634
		tipc_node_read_lock(n);
635 636
		le = &n->links[bearer_id];
		if (le->link) {
637
			spin_lock_bh(&le->lock);
638
			/* Link tolerance may change asynchronously: */
639 640
			tipc_node_calculate_timer(n, le->link);
			rc = tipc_link_timeout(le->link, &xmitq);
641 642
			spin_unlock_bh(&le->lock);
			remains--;
643
		}
J
Jon Paul Maloy 已提交
644
		tipc_node_read_unlock(n);
645 646 647
		tipc_bearer_xmit(n->net, bearer_id, &xmitq, &le->maddr);
		if (rc & TIPC_LINK_DOWN_EVT)
			tipc_node_link_down(n, bearer_id, false);
648
	}
649
	mod_timer(&n->timer, jiffies + msecs_to_jiffies(n->keepalive_intv));
650 651
}

P
Per Liden 已提交
652
/**
653 654
 * __tipc_node_link_up - handle addition of link
 * Node lock must be held by caller
P
Per Liden 已提交
655 656
 * Link becomes active (alone or shared) or standby, depending on its priority.
 */
657 658
static void __tipc_node_link_up(struct tipc_node *n, int bearer_id,
				struct sk_buff_head *xmitq)
P
Per Liden 已提交
659
{
660 661
	int *slot0 = &n->active_links[0];
	int *slot1 = &n->active_links[1];
662 663
	struct tipc_link *ol = node_active_link(n, 0);
	struct tipc_link *nl = n->links[bearer_id].link;
664

665
	if (!nl || tipc_link_is_up(nl))
666 667 668 669
		return;

	tipc_link_fsm_evt(nl, LINK_ESTABLISH_EVT);
	if (!tipc_link_is_up(nl))
670 671
		return;

672 673
	n->working_links++;
	n->action_flags |= TIPC_NOTIFY_LINK_UP;
674
	n->link_id = tipc_link_id(nl);
675 676

	/* Leave room for tunnel header when returning 'mtu' to users: */
677
	n->links[bearer_id].mtu = tipc_link_mtu(nl) - INT_H_SIZE;
Y
Ying Xue 已提交
678

679
	tipc_bearer_add_dest(n->net, bearer_id, n->addr);
680
	tipc_bcast_inc_bearer_dst_cnt(n->net, bearer_id);
681

682
	pr_debug("Established link <%s> on network plane %c\n",
683
		 tipc_link_name(nl), tipc_link_plane(nl));
684

685 686 687
	/* Ensure that a STATE message goes first */
	tipc_link_build_state_msg(nl, xmitq);

688 689
	/* First link? => give it both slots */
	if (!ol) {
690 691
		*slot0 = bearer_id;
		*slot1 = bearer_id;
692
		tipc_node_fsm_evt(n, SELF_ESTABL_CONTACT_EVT);
L
LUU Duc Canh 已提交
693
		n->failover_sent = false;
694
		n->action_flags |= TIPC_NOTIFY_NODE_UP;
695
		tipc_link_set_active(nl, true);
696
		tipc_bcast_add_peer(n->net, nl, xmitq);
697
		return;
P
Per Liden 已提交
698
	}
699

700
	/* Second link => redistribute slots */
701 702
	if (tipc_link_prio(nl) > tipc_link_prio(ol)) {
		pr_debug("Old link <%s> becomes standby\n", tipc_link_name(ol));
703
		*slot0 = bearer_id;
704
		*slot1 = bearer_id;
705 706
		tipc_link_set_active(nl, true);
		tipc_link_set_active(ol, false);
707
	} else if (tipc_link_prio(nl) == tipc_link_prio(ol)) {
708
		tipc_link_set_active(nl, true);
709
		*slot1 = bearer_id;
710
	} else {
711
		pr_debug("New link <%s> is standby\n", tipc_link_name(nl));
P
Per Liden 已提交
712 713
	}

714 715
	/* Prepare synchronization with first link */
	tipc_link_tnl_prepare(ol, nl, SYNCH_MSG, xmitq);
P
Per Liden 已提交
716 717 718
}

/**
719 720 721
 * tipc_node_link_up - handle addition of link
 *
 * Link becomes active (alone or shared) or standby, depending on its priority.
P
Per Liden 已提交
722
 */
723 724
static void tipc_node_link_up(struct tipc_node *n, int bearer_id,
			      struct sk_buff_head *xmitq)
P
Per Liden 已提交
725
{
726 727
	struct tipc_media_addr *maddr;

J
Jon Paul Maloy 已提交
728
	tipc_node_write_lock(n);
729
	__tipc_node_link_up(n, bearer_id, xmitq);
730 731
	maddr = &n->links[bearer_id].maddr;
	tipc_bearer_xmit(n->net, bearer_id, xmitq, maddr);
J
Jon Paul Maloy 已提交
732
	tipc_node_write_unlock(n);
733 734 735 736 737 738 739 740 741 742
}

/**
 * __tipc_node_link_down - handle loss of link
 */
static void __tipc_node_link_down(struct tipc_node *n, int *bearer_id,
				  struct sk_buff_head *xmitq,
				  struct tipc_media_addr **maddr)
{
	struct tipc_link_entry *le = &n->links[*bearer_id];
743 744
	int *slot0 = &n->active_links[0];
	int *slot1 = &n->active_links[1];
745
	int i, highest = 0, prio;
746
	struct tipc_link *l, *_l, *tnl;
P
Per Liden 已提交
747

748
	l = n->links[*bearer_id].link;
749
	if (!l || tipc_link_is_reset(l))
750 751
		return;

752 753
	n->working_links--;
	n->action_flags |= TIPC_NOTIFY_LINK_DOWN;
754
	n->link_id = tipc_link_id(l);
755

756
	tipc_bearer_remove_dest(n->net, *bearer_id, n->addr);
757

758
	pr_debug("Lost link <%s> on network plane %c\n",
759
		 tipc_link_name(l), tipc_link_plane(l));
760

761 762 763 764 765 766 767
	/* Select new active link if any available */
	*slot0 = INVALID_BEARER_ID;
	*slot1 = INVALID_BEARER_ID;
	for (i = 0; i < MAX_BEARERS; i++) {
		_l = n->links[i].link;
		if (!_l || !tipc_link_is_up(_l))
			continue;
768 769
		if (_l == l)
			continue;
770 771
		prio = tipc_link_prio(_l);
		if (prio < highest)
772
			continue;
773 774
		if (prio > highest) {
			highest = prio;
775 776 777 778 779 780
			*slot0 = i;
			*slot1 = i;
			continue;
		}
		*slot1 = i;
	}
781

782
	if (!node_is_up(n)) {
783 784 785 786
		if (tipc_link_peer_is_down(l))
			tipc_node_fsm_evt(n, PEER_LOST_CONTACT_EVT);
		tipc_node_fsm_evt(n, SELF_LOST_CONTACT_EVT);
		tipc_link_fsm_evt(l, LINK_RESET_EVT);
787
		tipc_link_reset(l);
788 789
		tipc_link_build_reset_msg(l, xmitq);
		*maddr = &n->links[*bearer_id].maddr;
790
		node_lost_contact(n, &le->inputq);
791
		tipc_bcast_dec_bearer_dst_cnt(n->net, *bearer_id);
792 793
		return;
	}
794
	tipc_bcast_dec_bearer_dst_cnt(n->net, *bearer_id);
795

796
	/* There is still a working link => initiate failover */
797 798
	*bearer_id = n->active_links[0];
	tnl = n->links[*bearer_id].link;
799 800
	tipc_link_fsm_evt(tnl, LINK_SYNCH_END_EVT);
	tipc_node_fsm_evt(n, NODE_SYNCH_END_EVT);
801
	n->sync_point = tipc_link_rcv_nxt(tnl) + (U16_MAX / 2 - 1);
802
	tipc_link_tnl_prepare(l, tnl, FAILOVER_MSG, xmitq);
803
	tipc_link_reset(l);
804
	tipc_link_fsm_evt(l, LINK_RESET_EVT);
805
	tipc_link_fsm_evt(l, LINK_FAILOVER_BEGIN_EVT);
806
	tipc_node_fsm_evt(n, NODE_FAILOVER_BEGIN_EVT);
807
	*maddr = &n->links[*bearer_id].maddr;
808 809 810 811 812
}

static void tipc_node_link_down(struct tipc_node *n, int bearer_id, bool delete)
{
	struct tipc_link_entry *le = &n->links[bearer_id];
813
	struct tipc_link *l = le->link;
814 815
	struct tipc_media_addr *maddr;
	struct sk_buff_head xmitq;
816
	int old_bearer_id = bearer_id;
817

818 819 820
	if (!l)
		return;

821 822
	__skb_queue_head_init(&xmitq);

J
Jon Paul Maloy 已提交
823
	tipc_node_write_lock(n);
824 825 826 827 828 829 830 831 832 833
	if (!tipc_link_is_establishing(l)) {
		__tipc_node_link_down(n, &bearer_id, &xmitq, &maddr);
		if (delete) {
			kfree(l);
			le->link = NULL;
			n->link_cnt--;
		}
	} else {
		/* Defuse pending tipc_node_link_up() */
		tipc_link_fsm_evt(l, LINK_RESET_EVT);
834
	}
J
Jon Paul Maloy 已提交
835
	tipc_node_write_unlock(n);
836 837
	if (delete)
		tipc_mon_remove_peer(n->net, n->addr, old_bearer_id);
838 839
	tipc_bearer_xmit(n->net, bearer_id, &xmitq, maddr);
	tipc_sk_rcv(n->net, &le->inputq);
P
Per Liden 已提交
840 841
}

842
static bool node_is_up(struct tipc_node *n)
P
Per Liden 已提交
843
{
844
	return n->active_links[0] != INVALID_BEARER_ID;
P
Per Liden 已提交
845 846
}

847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862
bool tipc_node_is_up(struct net *net, u32 addr)
{
	struct tipc_node *n;
	bool retval = false;

	if (in_own_node(net, addr))
		return true;

	n = tipc_node_find(net, addr);
	if (!n)
		return false;
	retval = node_is_up(n);
	tipc_node_put(n);
	return retval;
}

863 864 865 866 867 868 869 870 871 872 873 874 875
static u32 tipc_node_suggest_addr(struct net *net, u32 addr)
{
	struct tipc_node *n;

	addr ^= tipc_net(net)->random;
	while ((n = tipc_node_find(net, addr))) {
		tipc_node_put(n);
		addr++;
	}
	return addr;
}

/* tipc_node_try_addr(): Check if addr can be used by peer, suggest other if not
876
 * Returns suggested address if any, otherwise 0
877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898
 */
u32 tipc_node_try_addr(struct net *net, u8 *id, u32 addr)
{
	struct tipc_net *tn = tipc_net(net);
	struct tipc_node *n;

	/* Suggest new address if some other peer is using this one */
	n = tipc_node_find(net, addr);
	if (n) {
		if (!memcmp(n->peer_id, id, NODE_ID_LEN))
			addr = 0;
		tipc_node_put(n);
		if (!addr)
			return 0;
		return tipc_node_suggest_addr(net, addr);
	}

	/* Suggest previously used address if peer is known */
	n = tipc_node_find_by_id(net, id);
	if (n) {
		addr = n->addr;
		tipc_node_put(n);
899
		return addr;
900
	}
901 902

	/* Even this node may be in conflict */
903 904 905
	if (tn->trial_addr == addr)
		return tipc_node_suggest_addr(net, addr);

906
	return 0;
907 908 909 910
}

void tipc_node_check_dest(struct net *net, u32 addr,
			  u8 *peer_id, struct tipc_bearer *b,
911 912 913
			  u16 capabilities, u32 signature,
			  struct tipc_media_addr *maddr,
			  bool *respond, bool *dupl_addr)
914
{
915 916
	struct tipc_node *n;
	struct tipc_link *l;
J
Jon Paul Maloy 已提交
917
	struct tipc_link_entry *le;
918 919 920 921
	bool addr_match = false;
	bool sign_match = false;
	bool link_up = false;
	bool accept_addr = false;
922
	bool reset = true;
923
	char *if_name;
924
	unsigned long intv;
925
	u16 session;
J
Jon Paul Maloy 已提交
926

927 928 929
	*dupl_addr = false;
	*respond = false;

930
	n = tipc_node_create(net, addr, peer_id, capabilities);
931 932
	if (!n)
		return;
933

J
Jon Paul Maloy 已提交
934
	tipc_node_write_lock(n);
935

J
Jon Paul Maloy 已提交
936
	le = &n->links[b->identity];
937 938

	/* Prepare to validate requesting node's signature and media address */
J
Jon Paul Maloy 已提交
939
	l = le->link;
940
	link_up = l && tipc_link_is_up(l);
J
Jon Paul Maloy 已提交
941
	addr_match = l && !memcmp(&le->maddr, maddr, sizeof(*maddr));
942 943 944 945 946 947
	sign_match = (signature == n->signature);

	/* These three flags give us eight permutations: */

	if (sign_match && addr_match && link_up) {
		/* All is fine. Do nothing. */
948
		reset = false;
949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001
	} else if (sign_match && addr_match && !link_up) {
		/* Respond. The link will come up in due time */
		*respond = true;
	} else if (sign_match && !addr_match && link_up) {
		/* Peer has changed i/f address without rebooting.
		 * If so, the link will reset soon, and the next
		 * discovery will be accepted. So we can ignore it.
		 * It may also be an cloned or malicious peer having
		 * chosen the same node address and signature as an
		 * existing one.
		 * Ignore requests until the link goes down, if ever.
		 */
		*dupl_addr = true;
	} else if (sign_match && !addr_match && !link_up) {
		/* Peer link has changed i/f address without rebooting.
		 * It may also be a cloned or malicious peer; we can't
		 * distinguish between the two.
		 * The signature is correct, so we must accept.
		 */
		accept_addr = true;
		*respond = true;
	} else if (!sign_match && addr_match && link_up) {
		/* Peer node rebooted. Two possibilities:
		 *  - Delayed re-discovery; this link endpoint has already
		 *    reset and re-established contact with the peer, before
		 *    receiving a discovery message from that node.
		 *    (The peer happened to receive one from this node first).
		 *  - The peer came back so fast that our side has not
		 *    discovered it yet. Probing from this side will soon
		 *    reset the link, since there can be no working link
		 *    endpoint at the peer end, and the link will re-establish.
		 *  Accept the signature, since it comes from a known peer.
		 */
		n->signature = signature;
	} else if (!sign_match && addr_match && !link_up) {
		/*  The peer node has rebooted.
		 *  Accept signature, since it is a known peer.
		 */
		n->signature = signature;
		*respond = true;
	} else if (!sign_match && !addr_match && link_up) {
		/* Peer rebooted with new address, or a new/duplicate peer.
		 * Ignore until the link goes down, if ever.
		 */
		*dupl_addr = true;
	} else if (!sign_match && !addr_match && !link_up) {
		/* Peer rebooted with new address, or it is a new peer.
		 * Accept signature and address.
		 */
		n->signature = signature;
		accept_addr = true;
		*respond = true;
	}
1002

1003 1004
	if (!accept_addr)
		goto exit;
1005

1006
	/* Now create new link if not already existing */
1007
	if (!l) {
1008
		if (n->link_cnt == 2)
J
Jon Paul Maloy 已提交
1009
			goto exit;
1010

1011
		if_name = strchr(b->name, ':') + 1;
1012
		get_random_bytes(&session, sizeof(u16));
1013
		if (!tipc_link_create(net, if_name, b->identity, b->tolerance,
1014
				      b->net_plane, b->mtu, b->priority,
1015
				      b->window, session,
1016
				      tipc_own_addr(net), addr, peer_id,
1017
				      n->capabilities,
1018 1019 1020
				      tipc_bc_sndlink(n->net), n->bc_entry.link,
				      &le->inputq,
				      &n->bc_entry.namedq, &l)) {
1021 1022 1023
			*respond = false;
			goto exit;
		}
J
Jon Paul Maloy 已提交
1024
		tipc_link_reset(l);
1025
		tipc_link_fsm_evt(l, LINK_RESET_EVT);
1026 1027
		if (n->state == NODE_FAILINGOVER)
			tipc_link_fsm_evt(l, LINK_FAILOVER_BEGIN_EVT);
J
Jon Paul Maloy 已提交
1028 1029
		le->link = l;
		n->link_cnt++;
1030
		tipc_node_calculate_timer(n, l);
1031 1032 1033
		if (n->link_cnt == 1) {
			intv = jiffies + msecs_to_jiffies(n->keepalive_intv);
			if (!mod_timer(&n->timer, intv))
1034
				tipc_node_get(n);
1035
		}
1036
	}
J
Jon Paul Maloy 已提交
1037
	memcpy(&le->maddr, maddr, sizeof(*maddr));
1038
exit:
J
Jon Paul Maloy 已提交
1039
	tipc_node_write_unlock(n);
1040
	if (reset && l && !tipc_link_is_reset(l))
1041
		tipc_node_link_down(n, b->identity, false);
1042
	tipc_node_put(n);
1043 1044
}

1045 1046 1047 1048 1049 1050 1051
void tipc_node_delete_links(struct net *net, int bearer_id)
{
	struct tipc_net *tn = net_generic(net, tipc_net_id);
	struct tipc_node *n;

	rcu_read_lock();
	list_for_each_entry_rcu(n, &tn->node_list, list) {
1052
		tipc_node_link_down(n, bearer_id, true);
1053 1054 1055 1056 1057 1058
	}
	rcu_read_unlock();
}

static void tipc_node_reset_links(struct tipc_node *n)
{
1059
	int i;
1060

J
Jon Maloy 已提交
1061
	pr_warn("Resetting all links to %x\n", n->addr);
1062 1063

	for (i = 0; i < MAX_BEARERS; i++) {
1064
		tipc_node_link_down(n, i, false);
1065 1066 1067
	}
}

1068 1069 1070
/* tipc_node_fsm_evt - node finite state machine
 * Determines when contact is allowed with peer node
 */
1071
static void tipc_node_fsm_evt(struct tipc_node *n, int evt)
1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086
{
	int state = n->state;

	switch (state) {
	case SELF_DOWN_PEER_DOWN:
		switch (evt) {
		case SELF_ESTABL_CONTACT_EVT:
			state = SELF_UP_PEER_COMING;
			break;
		case PEER_ESTABL_CONTACT_EVT:
			state = SELF_COMING_PEER_UP;
			break;
		case SELF_LOST_CONTACT_EVT:
		case PEER_LOST_CONTACT_EVT:
			break;
J
Jon Paul Maloy 已提交
1087 1088 1089 1090
		case NODE_SYNCH_END_EVT:
		case NODE_SYNCH_BEGIN_EVT:
		case NODE_FAILOVER_BEGIN_EVT:
		case NODE_FAILOVER_END_EVT:
1091
		default:
J
Jon Paul Maloy 已提交
1092
			goto illegal_evt;
1093 1094 1095 1096 1097 1098 1099 1100 1101 1102
		}
		break;
	case SELF_UP_PEER_UP:
		switch (evt) {
		case SELF_LOST_CONTACT_EVT:
			state = SELF_DOWN_PEER_LEAVING;
			break;
		case PEER_LOST_CONTACT_EVT:
			state = SELF_LEAVING_PEER_DOWN;
			break;
J
Jon Paul Maloy 已提交
1103 1104 1105 1106 1107 1108
		case NODE_SYNCH_BEGIN_EVT:
			state = NODE_SYNCHING;
			break;
		case NODE_FAILOVER_BEGIN_EVT:
			state = NODE_FAILINGOVER;
			break;
1109 1110
		case SELF_ESTABL_CONTACT_EVT:
		case PEER_ESTABL_CONTACT_EVT:
J
Jon Paul Maloy 已提交
1111 1112
		case NODE_SYNCH_END_EVT:
		case NODE_FAILOVER_END_EVT:
1113 1114
			break;
		default:
J
Jon Paul Maloy 已提交
1115
			goto illegal_evt;
1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126
		}
		break;
	case SELF_DOWN_PEER_LEAVING:
		switch (evt) {
		case PEER_LOST_CONTACT_EVT:
			state = SELF_DOWN_PEER_DOWN;
			break;
		case SELF_ESTABL_CONTACT_EVT:
		case PEER_ESTABL_CONTACT_EVT:
		case SELF_LOST_CONTACT_EVT:
			break;
J
Jon Paul Maloy 已提交
1127 1128 1129 1130
		case NODE_SYNCH_END_EVT:
		case NODE_SYNCH_BEGIN_EVT:
		case NODE_FAILOVER_BEGIN_EVT:
		case NODE_FAILOVER_END_EVT:
1131
		default:
J
Jon Paul Maloy 已提交
1132
			goto illegal_evt;
1133 1134 1135 1136 1137 1138 1139 1140
		}
		break;
	case SELF_UP_PEER_COMING:
		switch (evt) {
		case PEER_ESTABL_CONTACT_EVT:
			state = SELF_UP_PEER_UP;
			break;
		case SELF_LOST_CONTACT_EVT:
J
Jon Paul Maloy 已提交
1141
			state = SELF_DOWN_PEER_DOWN;
1142 1143 1144
			break;
		case SELF_ESTABL_CONTACT_EVT:
		case PEER_LOST_CONTACT_EVT:
J
Jon Paul Maloy 已提交
1145 1146
		case NODE_SYNCH_END_EVT:
		case NODE_FAILOVER_BEGIN_EVT:
1147 1148
			break;
		case NODE_SYNCH_BEGIN_EVT:
J
Jon Paul Maloy 已提交
1149
		case NODE_FAILOVER_END_EVT:
1150
		default:
J
Jon Paul Maloy 已提交
1151
			goto illegal_evt;
1152 1153 1154 1155 1156 1157 1158 1159
		}
		break;
	case SELF_COMING_PEER_UP:
		switch (evt) {
		case SELF_ESTABL_CONTACT_EVT:
			state = SELF_UP_PEER_UP;
			break;
		case PEER_LOST_CONTACT_EVT:
J
Jon Paul Maloy 已提交
1160
			state = SELF_DOWN_PEER_DOWN;
1161 1162 1163 1164
			break;
		case SELF_LOST_CONTACT_EVT:
		case PEER_ESTABL_CONTACT_EVT:
			break;
J
Jon Paul Maloy 已提交
1165 1166 1167 1168
		case NODE_SYNCH_END_EVT:
		case NODE_SYNCH_BEGIN_EVT:
		case NODE_FAILOVER_BEGIN_EVT:
		case NODE_FAILOVER_END_EVT:
1169
		default:
J
Jon Paul Maloy 已提交
1170
			goto illegal_evt;
1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181
		}
		break;
	case SELF_LEAVING_PEER_DOWN:
		switch (evt) {
		case SELF_LOST_CONTACT_EVT:
			state = SELF_DOWN_PEER_DOWN;
			break;
		case SELF_ESTABL_CONTACT_EVT:
		case PEER_ESTABL_CONTACT_EVT:
		case PEER_LOST_CONTACT_EVT:
			break;
J
Jon Paul Maloy 已提交
1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206
		case NODE_SYNCH_END_EVT:
		case NODE_SYNCH_BEGIN_EVT:
		case NODE_FAILOVER_BEGIN_EVT:
		case NODE_FAILOVER_END_EVT:
		default:
			goto illegal_evt;
		}
		break;
	case NODE_FAILINGOVER:
		switch (evt) {
		case SELF_LOST_CONTACT_EVT:
			state = SELF_DOWN_PEER_LEAVING;
			break;
		case PEER_LOST_CONTACT_EVT:
			state = SELF_LEAVING_PEER_DOWN;
			break;
		case NODE_FAILOVER_END_EVT:
			state = SELF_UP_PEER_UP;
			break;
		case NODE_FAILOVER_BEGIN_EVT:
		case SELF_ESTABL_CONTACT_EVT:
		case PEER_ESTABL_CONTACT_EVT:
			break;
		case NODE_SYNCH_BEGIN_EVT:
		case NODE_SYNCH_END_EVT:
1207
		default:
J
Jon Paul Maloy 已提交
1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231
			goto illegal_evt;
		}
		break;
	case NODE_SYNCHING:
		switch (evt) {
		case SELF_LOST_CONTACT_EVT:
			state = SELF_DOWN_PEER_LEAVING;
			break;
		case PEER_LOST_CONTACT_EVT:
			state = SELF_LEAVING_PEER_DOWN;
			break;
		case NODE_SYNCH_END_EVT:
			state = SELF_UP_PEER_UP;
			break;
		case NODE_FAILOVER_BEGIN_EVT:
			state = NODE_FAILINGOVER;
			break;
		case NODE_SYNCH_BEGIN_EVT:
		case SELF_ESTABL_CONTACT_EVT:
		case PEER_ESTABL_CONTACT_EVT:
			break;
		case NODE_FAILOVER_END_EVT:
		default:
			goto illegal_evt;
1232 1233 1234 1235 1236 1237 1238
		}
		break;
	default:
		pr_err("Unknown node fsm state %x\n", state);
		break;
	}
	n->state = state;
J
Jon Paul Maloy 已提交
1239 1240 1241 1242
	return;

illegal_evt:
	pr_err("Illegal node fsm evt %x in state %x\n", evt, state);
1243 1244
}

1245
static void node_lost_contact(struct tipc_node *n,
1246
			      struct sk_buff_head *inputq)
P
Per Liden 已提交
1247
{
1248
	struct tipc_sock_conn *conn, *safe;
1249
	struct tipc_link *l;
1250
	struct list_head *conns = &n->conn_sks;
1251 1252
	struct sk_buff *skb;
	uint i;
P
Per Liden 已提交
1253

J
Jon Maloy 已提交
1254
	pr_debug("Lost contact with %x\n", n->addr);
1255
	n->delete_at = jiffies + msecs_to_jiffies(NODE_CLEANUP_AFTER);
1256

1257
	/* Clean up broadcast state */
1258
	tipc_bcast_remove_peer(n->net, n->bc_entry.link);
P
Per Liden 已提交
1259

1260
	/* Abort any ongoing link failover */
P
Per Liden 已提交
1261
	for (i = 0; i < MAX_BEARERS; i++) {
1262
		l = n->links[i].link;
1263 1264
		if (l)
			tipc_link_fsm_evt(l, LINK_FAILOVER_END_EVT);
P
Per Liden 已提交
1265
	}
1266

1267
	/* Notify publications from this node */
1268
	n->action_flags |= TIPC_NOTIFY_NODE_DOWN;
1269 1270 1271 1272

	/* Notify sockets connected to node */
	list_for_each_entry_safe(conn, safe, conns, list) {
		skb = tipc_msg_create(TIPC_CRITICAL_IMPORTANCE, TIPC_CONN_MSG,
1273
				      SHORT_H_SIZE, 0, tipc_own_addr(n->net),
1274 1275
				      conn->peer_node, conn->port,
				      conn->peer_port, TIPC_ERR_NO_NODE);
1276
		if (likely(skb))
1277
			skb_queue_tail(inputq, skb);
1278 1279 1280
		list_del(&conn->list);
		kfree(conn);
	}
P
Per Liden 已提交
1281 1282
}

E
Erik Hugne 已提交
1283 1284 1285 1286 1287 1288 1289 1290 1291
/**
 * tipc_node_get_linkname - get the name of a link
 *
 * @bearer_id: id of the bearer
 * @node: peer node address
 * @linkname: link name output buffer
 *
 * Returns 0 on success
 */
1292 1293
int tipc_node_get_linkname(struct net *net, u32 bearer_id, u32 addr,
			   char *linkname, size_t len)
E
Erik Hugne 已提交
1294 1295
{
	struct tipc_link *link;
1296
	int err = -EINVAL;
1297
	struct tipc_node *node = tipc_node_find(net, addr);
E
Erik Hugne 已提交
1298

1299 1300 1301 1302 1303 1304
	if (!node)
		return err;

	if (bearer_id >= MAX_BEARERS)
		goto exit;

J
Jon Paul Maloy 已提交
1305
	tipc_node_read_lock(node);
1306
	link = node->links[bearer_id].link;
E
Erik Hugne 已提交
1307
	if (link) {
1308
		strncpy(linkname, tipc_link_name(link), len);
1309
		err = 0;
E
Erik Hugne 已提交
1310
	}
J
Jon Paul Maloy 已提交
1311
	tipc_node_read_unlock(node);
1312
exit:
1313 1314
	tipc_node_put(node);
	return err;
E
Erik Hugne 已提交
1315
}
1316

1317
/* Caller should hold node lock for the passed node */
1318
static int __tipc_nl_add_node(struct tipc_nl_msg *msg, struct tipc_node *node)
1319 1320 1321 1322
{
	void *hdr;
	struct nlattr *attrs;

1323
	hdr = genlmsg_put(msg->skb, msg->portid, msg->seq, &tipc_genl_family,
1324 1325 1326 1327 1328 1329 1330 1331 1332 1333
			  NLM_F_MULTI, TIPC_NL_NODE_GET);
	if (!hdr)
		return -EMSGSIZE;

	attrs = nla_nest_start(msg->skb, TIPC_NLA_NODE);
	if (!attrs)
		goto msg_full;

	if (nla_put_u32(msg->skb, TIPC_NLA_NODE_ADDR, node->addr))
		goto attr_msg_full;
1334
	if (node_is_up(node))
1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350
		if (nla_put_flag(msg->skb, TIPC_NLA_NODE_UP))
			goto attr_msg_full;

	nla_nest_end(msg->skb, attrs);
	genlmsg_end(msg->skb, hdr);

	return 0;

attr_msg_full:
	nla_nest_cancel(msg->skb, attrs);
msg_full:
	genlmsg_cancel(msg->skb, hdr);

	return -EMSGSIZE;
}

1351 1352 1353 1354 1355 1356
/**
 * tipc_node_xmit() is the general link level function for message sending
 * @net: the applicable net namespace
 * @list: chain of buffers containing message
 * @dnode: address of destination node
 * @selector: a number used for deterministic link selection
1357
 * Consumes the buffer chain.
1358
 * Returns 0 if success, otherwise: -ELINKCONG,-EHOSTUNREACH,-EMSGSIZE,-ENOBUF
1359 1360 1361 1362
 */
int tipc_node_xmit(struct net *net, struct sk_buff_head *list,
		   u32 dnode, int selector)
{
J
Jon Paul Maloy 已提交
1363
	struct tipc_link_entry *le = NULL;
1364 1365
	struct tipc_node *n;
	struct sk_buff_head xmitq;
1366 1367 1368 1369 1370 1371 1372
	int bearer_id;
	int rc;

	if (in_own_node(net, dnode)) {
		tipc_sk_rcv(net, list);
		return 0;
	}
1373 1374

	n = tipc_node_find(net, dnode);
1375 1376 1377 1378 1379 1380 1381 1382
	if (unlikely(!n)) {
		skb_queue_purge(list);
		return -EHOSTUNREACH;
	}

	tipc_node_read_lock(n);
	bearer_id = n->active_links[selector & 1];
	if (unlikely(bearer_id == INVALID_BEARER_ID)) {
J
Jon Paul Maloy 已提交
1383
		tipc_node_read_unlock(n);
1384
		tipc_node_put(n);
1385 1386
		skb_queue_purge(list);
		return -EHOSTUNREACH;
1387
	}
J
Jon Paul Maloy 已提交
1388

1389 1390 1391 1392 1393 1394 1395
	__skb_queue_head_init(&xmitq);
	le = &n->links[bearer_id];
	spin_lock_bh(&le->lock);
	rc = tipc_link_xmit(le->link, list, &xmitq);
	spin_unlock_bh(&le->lock);
	tipc_node_read_unlock(n);

1396
	if (unlikely(rc == -ENOBUFS))
1397
		tipc_node_link_down(n, bearer_id, false);
1398 1399
	else
		tipc_bearer_xmit(net, bearer_id, &xmitq, &le->maddr);
1400 1401 1402

	tipc_node_put(n);

1403
	return rc;
1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418
}

/* tipc_node_xmit_skb(): send single buffer to destination
 * Buffers sent via this functon are generally TIPC_SYSTEM_IMPORTANCE
 * messages, which will not be rejected
 * The only exception is datagram messages rerouted after secondary
 * lookup, which are rare and safe to dispose of anyway.
 */
int tipc_node_xmit_skb(struct net *net, struct sk_buff *skb, u32 dnode,
		       u32 selector)
{
	struct sk_buff_head head;

	skb_queue_head_init(&head);
	__skb_queue_tail(&head, skb);
1419
	tipc_node_xmit(net, &head, dnode, selector);
1420 1421 1422
	return 0;
}

1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438
/* tipc_node_distr_xmit(): send single buffer msgs to individual destinations
 * Note: this is only for SYSTEM_IMPORTANCE messages, which cannot be rejected
 */
int tipc_node_distr_xmit(struct net *net, struct sk_buff_head *xmitq)
{
	struct sk_buff *skb;
	u32 selector, dnode;

	while ((skb = __skb_dequeue(xmitq))) {
		selector = msg_origport(buf_msg(skb));
		dnode = msg_destnode(buf_msg(skb));
		tipc_node_xmit_skb(net, skb, dnode, selector);
	}
	return 0;
}

1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449
void tipc_node_broadcast(struct net *net, struct sk_buff *skb)
{
	struct sk_buff *txskb;
	struct tipc_node *n;
	u32 dst;

	rcu_read_lock();
	list_for_each_entry_rcu(n, tipc_nodes(net), list) {
		dst = n->addr;
		if (in_own_node(net, dst))
			continue;
1450
		if (!node_is_up(n))
1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462
			continue;
		txskb = pskb_copy(skb, GFP_ATOMIC);
		if (!txskb)
			break;
		msg_set_destnode(buf_msg(txskb), dst);
		tipc_node_xmit_skb(net, txskb, dst, 0);
	}
	rcu_read_unlock();

	kfree_skb(skb);
}

1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475
static void tipc_node_mcast_rcv(struct tipc_node *n)
{
	struct tipc_bclink_entry *be = &n->bc_entry;

	/* 'arrvq' is under inputq2's lock protection */
	spin_lock_bh(&be->inputq2.lock);
	spin_lock_bh(&be->inputq1.lock);
	skb_queue_splice_tail_init(&be->inputq1, &be->arrvq);
	spin_unlock_bh(&be->inputq1.lock);
	spin_unlock_bh(&be->inputq2.lock);
	tipc_sk_mcast_rcv(n->net, &be->arrvq, &be->inputq2);
}

1476 1477 1478 1479 1480 1481 1482 1483 1484
static void tipc_node_bc_sync_rcv(struct tipc_node *n, struct tipc_msg *hdr,
				  int bearer_id, struct sk_buff_head *xmitq)
{
	struct tipc_link *ucl;
	int rc;

	rc = tipc_bcast_sync_rcv(n->net, n->bc_entry.link, hdr);

	if (rc & TIPC_LINK_DOWN_EVT) {
1485
		tipc_node_reset_links(n);
1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503
		return;
	}

	if (!(rc & TIPC_LINK_SND_STATE))
		return;

	/* If probe message, a STATE response will be sent anyway */
	if (msg_probe(hdr))
		return;

	/* Produce a STATE message carrying broadcast NACK */
	tipc_node_read_lock(n);
	ucl = n->links[bearer_id].link;
	if (ucl)
		tipc_link_build_state_msg(ucl, xmitq);
	tipc_node_read_unlock(n);
}

1504 1505 1506 1507 1508 1509 1510 1511
/**
 * tipc_node_bc_rcv - process TIPC broadcast packet arriving from off-node
 * @net: the applicable net namespace
 * @skb: TIPC packet
 * @bearer_id: id of bearer message arrived on
 *
 * Invoked with no locks held.
 */
1512
static void tipc_node_bc_rcv(struct net *net, struct sk_buff *skb, int bearer_id)
1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539
{
	int rc;
	struct sk_buff_head xmitq;
	struct tipc_bclink_entry *be;
	struct tipc_link_entry *le;
	struct tipc_msg *hdr = buf_msg(skb);
	int usr = msg_user(hdr);
	u32 dnode = msg_destnode(hdr);
	struct tipc_node *n;

	__skb_queue_head_init(&xmitq);

	/* If NACK for other node, let rcv link for that node peek into it */
	if ((usr == BCAST_PROTOCOL) && (dnode != tipc_own_addr(net)))
		n = tipc_node_find(net, dnode);
	else
		n = tipc_node_find(net, msg_prevnode(hdr));
	if (!n) {
		kfree_skb(skb);
		return;
	}
	be = &n->bc_entry;
	le = &n->links[bearer_id];

	rc = tipc_bcast_rcv(net, be->link, skb);

	/* Broadcast ACKs are sent on a unicast link */
1540
	if (rc & TIPC_LINK_SND_STATE) {
J
Jon Paul Maloy 已提交
1541
		tipc_node_read_lock(n);
1542
		tipc_link_build_state_msg(le->link, &xmitq);
J
Jon Paul Maloy 已提交
1543
		tipc_node_read_unlock(n);
1544 1545 1546 1547 1548
	}

	if (!skb_queue_empty(&xmitq))
		tipc_bearer_xmit(net, bearer_id, &xmitq, &le->maddr);

1549 1550
	if (!skb_queue_empty(&be->inputq1))
		tipc_node_mcast_rcv(n);
1551

1552 1553 1554
	/* If reassembly or retransmission failure => reset all links to peer */
	if (rc & TIPC_LINK_DOWN_EVT)
		tipc_node_reset_links(n);
1555

1556 1557 1558
	tipc_node_put(n);
}

1559 1560 1561 1562
/**
 * tipc_node_check_state - check and if necessary update node state
 * @skb: TIPC packet
 * @bearer_id: identity of bearer delivering the packet
1563
 * Returns true if state and msg are ok, otherwise false
1564
 */
1565
static bool tipc_node_check_state(struct tipc_node *n, struct sk_buff *skb,
1566
				  int bearer_id, struct sk_buff_head *xmitq)
1567 1568
{
	struct tipc_msg *hdr = buf_msg(skb);
1569 1570
	int usr = msg_user(hdr);
	int mtyp = msg_type(hdr);
1571
	u16 oseqno = msg_seqno(hdr);
1572 1573
	u16 iseqno = msg_seqno(msg_get_wrapped(hdr));
	u16 exp_pkts = msg_msgcnt(hdr);
1574
	u16 rcv_nxt, syncpt, dlv_nxt, inputq_len;
1575
	int state = n->state;
1576
	struct tipc_link *l, *tnl, *pl = NULL;
1577
	struct tipc_media_addr *maddr;
1578
	int pb_id;
1579

1580 1581 1582
	l = n->links[bearer_id].link;
	if (!l)
		return false;
1583
	rcv_nxt = tipc_link_rcv_nxt(l);
1584 1585


1586 1587
	if (likely((state == SELF_UP_PEER_UP) && (usr != TUNNEL_PROTOCOL)))
		return true;
1588

1589
	/* Find parallel link, if any */
1590 1591 1592
	for (pb_id = 0; pb_id < MAX_BEARERS; pb_id++) {
		if ((pb_id != bearer_id) && n->links[pb_id].link) {
			pl = n->links[pb_id].link;
1593 1594 1595
			break;
		}
	}
1596

1597 1598 1599
	if (!tipc_link_validate_msg(l, hdr))
		return false;

J
Jon Paul Maloy 已提交
1600
	/* Check and update node accesibility if applicable */
1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612
	if (state == SELF_UP_PEER_COMING) {
		if (!tipc_link_is_up(l))
			return true;
		if (!msg_peer_link_is_up(hdr))
			return true;
		tipc_node_fsm_evt(n, PEER_ESTABL_CONTACT_EVT);
	}

	if (state == SELF_DOWN_PEER_LEAVING) {
		if (msg_peer_node_is_up(hdr))
			return false;
		tipc_node_fsm_evt(n, PEER_LOST_CONTACT_EVT);
1613
		return true;
1614 1615
	}

J
Jon Paul Maloy 已提交
1616 1617 1618
	if (state == SELF_LEAVING_PEER_DOWN)
		return false;

1619
	/* Ignore duplicate packets */
1620
	if ((usr != LINK_PROTOCOL) && less(oseqno, rcv_nxt))
1621 1622 1623 1624 1625
		return true;

	/* Initiate or update failover mode if applicable */
	if ((usr == TUNNEL_PROTOCOL) && (mtyp == FAILOVER_MSG)) {
		syncpt = oseqno + exp_pkts - 1;
1626 1627
		if (pl && tipc_link_is_up(pl)) {
			__tipc_node_link_down(n, &pb_id, xmitq, &maddr);
1628 1629
			tipc_skb_queue_splice_tail_init(tipc_link_inputq(pl),
							tipc_link_inputq(l));
1630
		}
L
LUU Duc Canh 已提交
1631 1632 1633 1634 1635 1636 1637 1638
		/* If parallel link was already down, and this happened before
		 * the tunnel link came up, FAILOVER was never sent. Ensure that
		 * FAILOVER is sent to get peer out of NODE_FAILINGOVER state.
		 */
		if (n->state != NODE_FAILINGOVER && !n->failover_sent) {
			tipc_link_create_dummy_tnl_msg(l, xmitq);
			n->failover_sent = true;
		}
1639 1640 1641 1642 1643 1644
		/* If pkts arrive out of order, use lowest calculated syncpt */
		if (less(syncpt, n->sync_point))
			n->sync_point = syncpt;
	}

	/* Open parallel link when tunnel link reaches synch point */
1645
	if ((n->state == NODE_FAILINGOVER) && tipc_link_is_up(l)) {
1646 1647
		if (!more(rcv_nxt, n->sync_point))
			return true;
1648 1649
		tipc_node_fsm_evt(n, NODE_FAILOVER_END_EVT);
		if (pl)
1650
			tipc_link_fsm_evt(pl, LINK_FAILOVER_END_EVT);
1651 1652 1653
		return true;
	}

1654 1655 1656 1657
	/* No synching needed if only one link */
	if (!pl || !tipc_link_is_up(pl))
		return true;

1658 1659
	/* Initiate synch mode if applicable */
	if ((usr == TUNNEL_PROTOCOL) && (mtyp == SYNCH_MSG) && (oseqno == 1)) {
1660
		syncpt = iseqno + exp_pkts - 1;
1661
		if (!tipc_link_is_up(l))
1662
			__tipc_node_link_up(n, bearer_id, xmitq);
1663 1664
		if (n->state == SELF_UP_PEER_UP) {
			n->sync_point = syncpt;
1665
			tipc_link_fsm_evt(l, LINK_SYNCH_BEGIN_EVT);
1666 1667
			tipc_node_fsm_evt(n, NODE_SYNCH_BEGIN_EVT);
		}
1668
	}
1669 1670

	/* Open tunnel link when parallel link reaches synch point */
1671
	if (n->state == NODE_SYNCHING) {
1672 1673 1674 1675 1676 1677
		if (tipc_link_is_synching(l)) {
			tnl = l;
		} else {
			tnl = pl;
			pl = l;
		}
1678 1679
		inputq_len = skb_queue_len(tipc_link_inputq(pl));
		dlv_nxt = tipc_link_rcv_nxt(pl) - inputq_len;
1680
		if (more(dlv_nxt, n->sync_point)) {
1681
			tipc_link_fsm_evt(tnl, LINK_SYNCH_END_EVT);
1682 1683 1684
			tipc_node_fsm_evt(n, NODE_SYNCH_END_EVT);
			return true;
		}
1685 1686
		if (l == pl)
			return true;
1687 1688 1689 1690 1691 1692 1693
		if ((usr == TUNNEL_PROTOCOL) && (mtyp == SYNCH_MSG))
			return true;
		if (usr == LINK_PROTOCOL)
			return true;
		return false;
	}
	return true;
1694 1695
}

1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708
/**
 * tipc_rcv - process TIPC packets/messages arriving from off-node
 * @net: the applicable net namespace
 * @skb: TIPC packet
 * @bearer: pointer to bearer message arrived on
 *
 * Invoked with no locks held. Bearer pointer must point to a valid bearer
 * structure (i.e. cannot be NULL), but bearer can be inactive.
 */
void tipc_rcv(struct net *net, struct sk_buff *skb, struct tipc_bearer *b)
{
	struct sk_buff_head xmitq;
	struct tipc_node *n;
1709
	struct tipc_msg *hdr;
1710
	int bearer_id = b->identity;
1711
	struct tipc_link_entry *le;
1712
	u32 self = tipc_own_addr(net);
1713 1714
	int usr, rc = 0;
	u16 bc_ack;
1715 1716 1717

	__skb_queue_head_init(&xmitq);

1718
	/* Ensure message is well-formed before touching the header */
1719
	if (unlikely(!tipc_msg_validate(&skb)))
1720
		goto discard;
1721 1722 1723
	hdr = buf_msg(skb);
	usr = msg_user(hdr);
	bc_ack = msg_bcast_ack(hdr);
1724

1725
	/* Handle arrival of discovery or broadcast packet */
1726
	if (unlikely(msg_non_seq(hdr))) {
1727 1728
		if (unlikely(usr == LINK_CONFIG))
			return tipc_disc_rcv(net, skb, b);
1729
		else
1730
			return tipc_node_bc_rcv(net, skb, bearer_id);
1731 1732
	}

1733 1734 1735 1736
	/* Discard unicast link messages destined for another node */
	if (unlikely(!msg_short(hdr) && (msg_destnode(hdr) != self)))
		goto discard;

1737 1738 1739 1740
	/* Locate neighboring node that sent packet */
	n = tipc_node_find(net, msg_prevnode(hdr));
	if (unlikely(!n))
		goto discard;
1741
	le = &n->links[bearer_id];
1742

1743 1744
	/* Ensure broadcast reception is in synch with peer's send state */
	if (unlikely(usr == LINK_PROTOCOL))
1745
		tipc_node_bc_sync_rcv(n, hdr, bearer_id, &xmitq);
1746
	else if (unlikely(tipc_link_acked(n->bc_entry.link) != bc_ack))
1747
		tipc_bcast_ack_rcv(net, n->bc_entry.link, hdr);
1748

J
Jon Paul Maloy 已提交
1749 1750 1751
	/* Receive packet directly if conditions permit */
	tipc_node_read_lock(n);
	if (likely((n->state == SELF_UP_PEER_UP) && (usr != TUNNEL_PROTOCOL))) {
J
Jon Paul Maloy 已提交
1752
		spin_lock_bh(&le->lock);
J
Jon Paul Maloy 已提交
1753 1754 1755 1756
		if (le->link) {
			rc = tipc_link_rcv(le->link, skb, &xmitq);
			skb = NULL;
		}
J
Jon Paul Maloy 已提交
1757
		spin_unlock_bh(&le->lock);
1758
	}
J
Jon Paul Maloy 已提交
1759 1760 1761 1762
	tipc_node_read_unlock(n);

	/* Check/update node state before receiving */
	if (unlikely(skb)) {
1763 1764
		if (unlikely(skb_linearize(skb)))
			goto discard;
J
Jon Paul Maloy 已提交
1765 1766 1767 1768 1769 1770 1771 1772 1773
		tipc_node_write_lock(n);
		if (tipc_node_check_state(n, skb, bearer_id, &xmitq)) {
			if (le->link) {
				rc = tipc_link_rcv(le->link, skb, &xmitq);
				skb = NULL;
			}
		}
		tipc_node_write_unlock(n);
	}
1774 1775

	if (unlikely(rc & TIPC_LINK_UP_EVT))
1776 1777
		tipc_node_link_up(n, bearer_id, &xmitq);

1778
	if (unlikely(rc & TIPC_LINK_DOWN_EVT))
1779
		tipc_node_link_down(n, bearer_id, false);
1780

1781 1782
	if (unlikely(!skb_queue_empty(&n->bc_entry.namedq)))
		tipc_named_rcv(net, &n->bc_entry.namedq);
1783

1784 1785 1786
	if (unlikely(!skb_queue_empty(&n->bc_entry.inputq1)))
		tipc_node_mcast_rcv(n);

1787 1788 1789 1790 1791 1792
	if (!skb_queue_empty(&le->inputq))
		tipc_sk_rcv(net, &le->inputq);

	if (!skb_queue_empty(&xmitq))
		tipc_bearer_xmit(net, bearer_id, &xmitq, &le->maddr);

1793 1794 1795 1796 1797
	tipc_node_put(n);
discard:
	kfree_skb(skb);
}

1798 1799
void tipc_node_apply_property(struct net *net, struct tipc_bearer *b,
			      int prop)
1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813
{
	struct tipc_net *tn = tipc_net(net);
	int bearer_id = b->identity;
	struct sk_buff_head xmitq;
	struct tipc_link_entry *e;
	struct tipc_node *n;

	__skb_queue_head_init(&xmitq);

	rcu_read_lock();

	list_for_each_entry_rcu(n, &tn->node_list, list) {
		tipc_node_write_lock(n);
		e = &n->links[bearer_id];
1814 1815 1816 1817 1818 1819 1820
		if (e->link) {
			if (prop == TIPC_NLA_PROP_TOL)
				tipc_link_set_tolerance(e->link, b->tolerance,
							&xmitq);
			else if (prop == TIPC_NLA_PROP_MTU)
				tipc_link_set_mtu(e->link, b->mtu);
		}
1821 1822 1823 1824 1825 1826 1827
		tipc_node_write_unlock(n);
		tipc_bearer_xmit(net, bearer_id, &xmitq, &e->maddr);
	}

	rcu_read_unlock();
}

1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841
int tipc_nl_peer_rm(struct sk_buff *skb, struct genl_info *info)
{
	struct net *net = sock_net(skb->sk);
	struct tipc_net *tn = net_generic(net, tipc_net_id);
	struct nlattr *attrs[TIPC_NLA_NET_MAX + 1];
	struct tipc_node *peer;
	u32 addr;
	int err;

	/* We identify the peer by its net */
	if (!info->attrs[TIPC_NLA_NET])
		return -EINVAL;

	err = nla_parse_nested(attrs, TIPC_NLA_NET_MAX,
1842
			       info->attrs[TIPC_NLA_NET], tipc_nl_net_policy,
1843
			       info->extack);
1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869
	if (err)
		return err;

	if (!attrs[TIPC_NLA_NET_ADDR])
		return -EINVAL;

	addr = nla_get_u32(attrs[TIPC_NLA_NET_ADDR]);

	if (in_own_node(net, addr))
		return -ENOTSUPP;

	spin_lock_bh(&tn->node_list_lock);
	peer = tipc_node_find(net, addr);
	if (!peer) {
		spin_unlock_bh(&tn->node_list_lock);
		return -ENXIO;
	}

	tipc_node_write_lock(peer);
	if (peer->state != SELF_DOWN_PEER_DOWN &&
	    peer->state != SELF_DOWN_PEER_LEAVING) {
		tipc_node_write_unlock(peer);
		err = -EBUSY;
		goto err_out;
	}

1870
	tipc_node_clear_links(peer);
1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881
	tipc_node_write_unlock(peer);
	tipc_node_delete(peer);

	err = 0;
err_out:
	tipc_node_put(peer);
	spin_unlock_bh(&tn->node_list_lock);

	return err;
}

1882 1883 1884
int tipc_nl_node_dump(struct sk_buff *skb, struct netlink_callback *cb)
{
	int err;
1885 1886
	struct net *net = sock_net(skb->sk);
	struct tipc_net *tn = net_generic(net, tipc_net_id);
1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899
	int done = cb->args[0];
	int last_addr = cb->args[1];
	struct tipc_node *node;
	struct tipc_nl_msg msg;

	if (done)
		return 0;

	msg.skb = skb;
	msg.portid = NETLINK_CB(cb->skb).portid;
	msg.seq = cb->nlh->nlmsg_seq;

	rcu_read_lock();
1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914
	if (last_addr) {
		node = tipc_node_find(net, last_addr);
		if (!node) {
			rcu_read_unlock();
			/* We never set seq or call nl_dump_check_consistent()
			 * this means that setting prev_seq here will cause the
			 * consistence check to fail in the netlink callback
			 * handler. Resulting in the NLMSG_DONE message having
			 * the NLM_F_DUMP_INTR flag set if the node state
			 * changed while we released the lock.
			 */
			cb->prev_seq = 1;
			return -EPIPE;
		}
		tipc_node_put(node);
1915 1916
	}

1917
	list_for_each_entry_rcu(node, &tn->node_list, list) {
1918 1919 1920 1921 1922 1923 1924
		if (last_addr) {
			if (node->addr == last_addr)
				last_addr = 0;
			else
				continue;
		}

J
Jon Paul Maloy 已提交
1925
		tipc_node_read_lock(node);
1926 1927 1928
		err = __tipc_nl_add_node(&msg, node);
		if (err) {
			last_addr = node->addr;
J
Jon Paul Maloy 已提交
1929
			tipc_node_read_unlock(node);
1930 1931 1932
			goto out;
		}

J
Jon Paul Maloy 已提交
1933
		tipc_node_read_unlock(node);
1934 1935 1936 1937 1938 1939 1940 1941 1942
	}
	done = 1;
out:
	cb->args[0] = done;
	cb->args[1] = last_addr;
	rcu_read_unlock();

	return skb->len;
}
1943

1944
/* tipc_node_find_by_name - locate owner node of link by link's name
1945 1946 1947 1948 1949 1950
 * @net: the applicable net namespace
 * @name: pointer to link name string
 * @bearer_id: pointer to index in 'node->links' array where the link was found.
 *
 * Returns pointer to node owning the link, or 0 if no matching link is found.
 */
1951 1952 1953
static struct tipc_node *tipc_node_find_by_name(struct net *net,
						const char *link_name,
						unsigned int *bearer_id)
1954 1955
{
	struct tipc_net *tn = net_generic(net, tipc_net_id);
1956 1957
	struct tipc_link *l;
	struct tipc_node *n;
1958 1959 1960 1961 1962
	struct tipc_node *found_node = NULL;
	int i;

	*bearer_id = 0;
	rcu_read_lock();
1963 1964
	list_for_each_entry_rcu(n, &tn->node_list, list) {
		tipc_node_read_lock(n);
1965
		for (i = 0; i < MAX_BEARERS; i++) {
1966 1967
			l = n->links[i].link;
			if (l && !strcmp(tipc_link_name(l), link_name)) {
1968
				*bearer_id = i;
1969
				found_node = n;
1970 1971 1972
				break;
			}
		}
1973
		tipc_node_read_unlock(n);
1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
		if (found_node)
			break;
	}
	rcu_read_unlock();

	return found_node;
}

int tipc_nl_node_set_link(struct sk_buff *skb, struct genl_info *info)
{
	int err;
	int res = 0;
	int bearer_id;
	char *name;
	struct tipc_link *link;
	struct tipc_node *node;
1990
	struct sk_buff_head xmitq;
1991 1992 1993
	struct nlattr *attrs[TIPC_NLA_LINK_MAX + 1];
	struct net *net = sock_net(skb->sk);

1994 1995
	__skb_queue_head_init(&xmitq);

1996 1997 1998 1999 2000
	if (!info->attrs[TIPC_NLA_LINK])
		return -EINVAL;

	err = nla_parse_nested(attrs, TIPC_NLA_LINK_MAX,
			       info->attrs[TIPC_NLA_LINK],
2001
			       tipc_nl_link_policy, info->extack);
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
	if (err)
		return err;

	if (!attrs[TIPC_NLA_LINK_NAME])
		return -EINVAL;

	name = nla_data(attrs[TIPC_NLA_LINK_NAME]);

	if (strcmp(name, tipc_bclink_name) == 0)
		return tipc_nl_bc_link_set(net, attrs);

2013
	node = tipc_node_find_by_name(net, name, &bearer_id);
2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038
	if (!node)
		return -EINVAL;

	tipc_node_read_lock(node);

	link = node->links[bearer_id].link;
	if (!link) {
		res = -EINVAL;
		goto out;
	}

	if (attrs[TIPC_NLA_LINK_PROP]) {
		struct nlattr *props[TIPC_NLA_PROP_MAX + 1];

		err = tipc_nl_parse_link_prop(attrs[TIPC_NLA_LINK_PROP],
					      props);
		if (err) {
			res = err;
			goto out;
		}

		if (props[TIPC_NLA_PROP_TOL]) {
			u32 tol;

			tol = nla_get_u32(props[TIPC_NLA_PROP_TOL]);
2039
			tipc_link_set_tolerance(link, tol, &xmitq);
2040 2041 2042 2043 2044
		}
		if (props[TIPC_NLA_PROP_PRIO]) {
			u32 prio;

			prio = nla_get_u32(props[TIPC_NLA_PROP_PRIO]);
2045
			tipc_link_set_prio(link, prio, &xmitq);
2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056
		}
		if (props[TIPC_NLA_PROP_WIN]) {
			u32 win;

			win = nla_get_u32(props[TIPC_NLA_PROP_WIN]);
			tipc_link_set_queue_limits(link, win);
		}
	}

out:
	tipc_node_read_unlock(node);
2057
	tipc_bearer_xmit(net, bearer_id, &xmitq, &node->links[bearer_id].maddr);
2058 2059 2060 2061 2062 2063
	return res;
}

int tipc_nl_node_get_link(struct sk_buff *skb, struct genl_info *info)
{
	struct net *net = genl_info_net(info);
2064
	struct nlattr *attrs[TIPC_NLA_LINK_MAX + 1];
2065 2066 2067 2068 2069 2070 2071
	struct tipc_nl_msg msg;
	char *name;
	int err;

	msg.portid = info->snd_portid;
	msg.seq = info->snd_seq;

2072
	if (!info->attrs[TIPC_NLA_LINK])
2073
		return -EINVAL;
2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084

	err = nla_parse_nested(attrs, TIPC_NLA_LINK_MAX,
			       info->attrs[TIPC_NLA_LINK],
			       tipc_nl_link_policy, info->extack);
	if (err)
		return err;

	if (!attrs[TIPC_NLA_LINK_NAME])
		return -EINVAL;

	name = nla_data(attrs[TIPC_NLA_LINK_NAME]);
2085 2086 2087 2088 2089 2090 2091

	msg.skb = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL);
	if (!msg.skb)
		return -ENOMEM;

	if (strcmp(name, tipc_bclink_name) == 0) {
		err = tipc_nl_add_bc_link(net, &msg);
2092 2093
		if (err)
			goto err_free;
2094 2095 2096 2097 2098
	} else {
		int bearer_id;
		struct tipc_node *node;
		struct tipc_link *link;

2099
		node = tipc_node_find_by_name(net, name, &bearer_id);
2100 2101 2102 2103
		if (!node) {
			err = -EINVAL;
			goto err_free;
		}
2104 2105 2106 2107 2108

		tipc_node_read_lock(node);
		link = node->links[bearer_id].link;
		if (!link) {
			tipc_node_read_unlock(node);
2109 2110
			err = -EINVAL;
			goto err_free;
2111 2112 2113 2114
		}

		err = __tipc_nl_add_link(net, &msg, link, 0);
		tipc_node_read_unlock(node);
2115 2116
		if (err)
			goto err_free;
2117 2118 2119
	}

	return genlmsg_reply(msg.skb, info);
2120 2121 2122 2123

err_free:
	nlmsg_free(msg.skb);
	return err;
2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141
}

int tipc_nl_node_reset_link_stats(struct sk_buff *skb, struct genl_info *info)
{
	int err;
	char *link_name;
	unsigned int bearer_id;
	struct tipc_link *link;
	struct tipc_node *node;
	struct nlattr *attrs[TIPC_NLA_LINK_MAX + 1];
	struct net *net = sock_net(skb->sk);
	struct tipc_link_entry *le;

	if (!info->attrs[TIPC_NLA_LINK])
		return -EINVAL;

	err = nla_parse_nested(attrs, TIPC_NLA_LINK_MAX,
			       info->attrs[TIPC_NLA_LINK],
2142
			       tipc_nl_link_policy, info->extack);
2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157
	if (err)
		return err;

	if (!attrs[TIPC_NLA_LINK_NAME])
		return -EINVAL;

	link_name = nla_data(attrs[TIPC_NLA_LINK_NAME]);

	if (strcmp(link_name, tipc_bclink_name) == 0) {
		err = tipc_bclink_reset_stats(net);
		if (err)
			return err;
		return 0;
	}

2158
	node = tipc_node_find_by_name(net, link_name, &bearer_id);
2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170
	if (!node)
		return -EINVAL;

	le = &node->links[bearer_id];
	tipc_node_read_lock(node);
	spin_lock_bh(&le->lock);
	link = node->links[bearer_id].link;
	if (!link) {
		spin_unlock_bh(&le->lock);
		tipc_node_read_unlock(node);
		return -EINVAL;
	}
2171
	tipc_link_reset_stats(link);
2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199
	spin_unlock_bh(&le->lock);
	tipc_node_read_unlock(node);
	return 0;
}

/* Caller should hold node lock  */
static int __tipc_nl_add_node_links(struct net *net, struct tipc_nl_msg *msg,
				    struct tipc_node *node, u32 *prev_link)
{
	u32 i;
	int err;

	for (i = *prev_link; i < MAX_BEARERS; i++) {
		*prev_link = i;

		if (!node->links[i].link)
			continue;

		err = __tipc_nl_add_link(net, msg,
					 node->links[i].link, NLM_F_MULTI);
		if (err)
			return err;
	}
	*prev_link = 0;

	return 0;
}

2200
int tipc_nl_node_dump_link(struct sk_buff *skb, struct netlink_callback *cb)
2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269
{
	struct net *net = sock_net(skb->sk);
	struct tipc_net *tn = net_generic(net, tipc_net_id);
	struct tipc_node *node;
	struct tipc_nl_msg msg;
	u32 prev_node = cb->args[0];
	u32 prev_link = cb->args[1];
	int done = cb->args[2];
	int err;

	if (done)
		return 0;

	msg.skb = skb;
	msg.portid = NETLINK_CB(cb->skb).portid;
	msg.seq = cb->nlh->nlmsg_seq;

	rcu_read_lock();
	if (prev_node) {
		node = tipc_node_find(net, prev_node);
		if (!node) {
			/* We never set seq or call nl_dump_check_consistent()
			 * this means that setting prev_seq here will cause the
			 * consistence check to fail in the netlink callback
			 * handler. Resulting in the last NLMSG_DONE message
			 * having the NLM_F_DUMP_INTR flag set.
			 */
			cb->prev_seq = 1;
			goto out;
		}
		tipc_node_put(node);

		list_for_each_entry_continue_rcu(node, &tn->node_list,
						 list) {
			tipc_node_read_lock(node);
			err = __tipc_nl_add_node_links(net, &msg, node,
						       &prev_link);
			tipc_node_read_unlock(node);
			if (err)
				goto out;

			prev_node = node->addr;
		}
	} else {
		err = tipc_nl_add_bc_link(net, &msg);
		if (err)
			goto out;

		list_for_each_entry_rcu(node, &tn->node_list, list) {
			tipc_node_read_lock(node);
			err = __tipc_nl_add_node_links(net, &msg, node,
						       &prev_link);
			tipc_node_read_unlock(node);
			if (err)
				goto out;

			prev_node = node->addr;
		}
	}
	done = 1;
out:
	rcu_read_unlock();

	cb->args[0] = prev_node;
	cb->args[1] = prev_link;
	cb->args[2] = done;

	return skb->len;
}
2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281

int tipc_nl_node_set_monitor(struct sk_buff *skb, struct genl_info *info)
{
	struct nlattr *attrs[TIPC_NLA_MON_MAX + 1];
	struct net *net = sock_net(skb->sk);
	int err;

	if (!info->attrs[TIPC_NLA_MON])
		return -EINVAL;

	err = nla_parse_nested(attrs, TIPC_NLA_MON_MAX,
			       info->attrs[TIPC_NLA_MON],
2282
			       tipc_nl_monitor_policy, info->extack);
2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296
	if (err)
		return err;

	if (attrs[TIPC_NLA_MON_ACTIVATION_THRESHOLD]) {
		u32 val;

		val = nla_get_u32(attrs[TIPC_NLA_MON_ACTIVATION_THRESHOLD]);
		err = tipc_nl_monitor_set_threshold(net, val);
		if (err)
			return err;
	}

	return 0;
}
2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337

static int __tipc_nl_add_monitor_prop(struct net *net, struct tipc_nl_msg *msg)
{
	struct nlattr *attrs;
	void *hdr;
	u32 val;

	hdr = genlmsg_put(msg->skb, msg->portid, msg->seq, &tipc_genl_family,
			  0, TIPC_NL_MON_GET);
	if (!hdr)
		return -EMSGSIZE;

	attrs = nla_nest_start(msg->skb, TIPC_NLA_MON);
	if (!attrs)
		goto msg_full;

	val = tipc_nl_monitor_get_threshold(net);

	if (nla_put_u32(msg->skb, TIPC_NLA_MON_ACTIVATION_THRESHOLD, val))
		goto attr_msg_full;

	nla_nest_end(msg->skb, attrs);
	genlmsg_end(msg->skb, hdr);

	return 0;

attr_msg_full:
	nla_nest_cancel(msg->skb, attrs);
msg_full:
	genlmsg_cancel(msg->skb, hdr);

	return -EMSGSIZE;
}

int tipc_nl_node_get_monitor(struct sk_buff *skb, struct genl_info *info)
{
	struct net *net = sock_net(skb->sk);
	struct tipc_nl_msg msg;
	int err;

	msg.skb = nlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL);
P
Pan Bian 已提交
2338 2339
	if (!msg.skb)
		return -ENOMEM;
2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350
	msg.portid = info->snd_portid;
	msg.seq = info->snd_seq;

	err = __tipc_nl_add_monitor_prop(net, &msg);
	if (err) {
		nlmsg_free(msg.skb);
		return err;
	}

	return genlmsg_reply(msg.skb, info);
}
2351 2352 2353 2354 2355 2356

int tipc_nl_node_dump_monitor(struct sk_buff *skb, struct netlink_callback *cb)
{
	struct net *net = sock_net(skb->sk);
	u32 prev_bearer = cb->args[0];
	struct tipc_nl_msg msg;
2357
	int bearer_id;
2358 2359 2360 2361 2362 2363 2364 2365 2366 2367
	int err;

	if (prev_bearer == MAX_BEARERS)
		return 0;

	msg.skb = skb;
	msg.portid = NETLINK_CB(cb->skb).portid;
	msg.seq = cb->nlh->nlmsg_seq;

	rtnl_lock();
2368
	for (bearer_id = prev_bearer; bearer_id < MAX_BEARERS; bearer_id++) {
2369
		err = __tipc_nl_add_monitor(net, &msg, bearer_id);
2370
		if (err)
2371
			break;
2372 2373
	}
	rtnl_unlock();
2374
	cb->args[0] = bearer_id;
2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401

	return skb->len;
}

int tipc_nl_node_dump_monitor_peer(struct sk_buff *skb,
				   struct netlink_callback *cb)
{
	struct net *net = sock_net(skb->sk);
	u32 prev_node = cb->args[1];
	u32 bearer_id = cb->args[2];
	int done = cb->args[0];
	struct tipc_nl_msg msg;
	int err;

	if (!prev_node) {
		struct nlattr **attrs;
		struct nlattr *mon[TIPC_NLA_MON_MAX + 1];

		err = tipc_nlmsg_parse(cb->nlh, &attrs);
		if (err)
			return err;

		if (!attrs[TIPC_NLA_MON])
			return -EINVAL;

		err = nla_parse_nested(mon, TIPC_NLA_MON_MAX,
				       attrs[TIPC_NLA_MON],
2402
				       tipc_nl_monitor_policy, NULL);
2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433
		if (err)
			return err;

		if (!mon[TIPC_NLA_MON_REF])
			return -EINVAL;

		bearer_id = nla_get_u32(mon[TIPC_NLA_MON_REF]);

		if (bearer_id >= MAX_BEARERS)
			return -EINVAL;
	}

	if (done)
		return 0;

	msg.skb = skb;
	msg.portid = NETLINK_CB(cb->skb).portid;
	msg.seq = cb->nlh->nlmsg_seq;

	rtnl_lock();
	err = tipc_nl_add_monitor_peer(net, &msg, bearer_id, &prev_node);
	if (!err)
		done = 1;

	rtnl_unlock();
	cb->args[0] = done;
	cb->args[1] = prev_node;
	cb->args[2] = bearer_id;

	return skb->len;
}