link.c 58.0 KB
Newer Older
P
Per Liden 已提交
1 2
/*
 * net/tipc/link.c: TIPC link code
3
 *
4
 * Copyright (c) 1996-2007, 2012-2016, Ericsson AB
5
 * Copyright (c) 2004-2007, 2010-2013, 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 "subscr.h"
P
Per Liden 已提交
39
#include "link.h"
40
#include "bcast.h"
41
#include "socket.h"
P
Per Liden 已提交
42 43
#include "name_distr.h"
#include "discover.h"
44
#include "netlink.h"
45
#include "monitor.h"
P
Per Liden 已提交
46

47 48
#include <linux/pkt_sched.h>

49
struct tipc_stats {
50 51
	u32 sent_pkts;
	u32 recv_pkts;
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
	u32 sent_states;
	u32 recv_states;
	u32 sent_probes;
	u32 recv_probes;
	u32 sent_nacks;
	u32 recv_nacks;
	u32 sent_acks;
	u32 sent_bundled;
	u32 sent_bundles;
	u32 recv_bundled;
	u32 recv_bundles;
	u32 retransmitted;
	u32 sent_fragmented;
	u32 sent_fragments;
	u32 recv_fragmented;
	u32 recv_fragments;
	u32 link_congs;		/* # port sends blocked by congestion */
	u32 deferred_recv;
	u32 duplicates;
	u32 max_queue_sz;	/* send queue size high water mark */
	u32 accu_queue_sz;	/* used for send queue size profiling */
	u32 queue_sz_counts;	/* used for send queue size profiling */
	u32 msg_length_counts;	/* used for message length profiling */
	u32 msg_lengths_total;	/* used for message length profiling */
	u32 msg_length_profile[7]; /* used for msg. length profiling */
};

/**
 * struct tipc_link - TIPC link data structure
 * @addr: network address of link's peer node
 * @name: link name character string
 * @media_addr: media address to use when sending messages over link
 * @timer: link timer
 * @net: pointer to namespace struct
 * @refcnt: reference counter for permanent references (owner node & timer)
 * @peer_session: link session # being used by peer end of link
 * @peer_bearer_id: bearer id used by link's peer endpoint
 * @bearer_id: local bearer id used by link
 * @tolerance: minimum link continuity loss needed to reset link [in ms]
 * @abort_limit: # of unacknowledged continuity probes needed to reset link
 * @state: current state of link FSM
 * @peer_caps: bitmap describing capabilities of peer node
 * @silent_intv_cnt: # of timer intervals without any reception from peer
 * @proto_msg: template for control messages generated by link
 * @pmsg: convenience pointer to "proto_msg" field
 * @priority: current link priority
 * @net_plane: current link network plane ('A' through 'H')
99
 * @mon_state: cookie with information needed by link monitor
100 101 102 103 104 105 106 107 108
 * @backlog_limit: backlog queue congestion thresholds (indexed by importance)
 * @exp_msg_count: # of tunnelled messages expected during link changeover
 * @reset_rcv_checkpt: seq # of last acknowledged message at time of link reset
 * @mtu: current maximum packet size for this link
 * @advertised_mtu: advertised own mtu when link is being established
 * @transmitq: queue for sent, non-acked messages
 * @backlogq: queue for messages waiting to be sent
 * @snt_nxt: next sequence number to use for outbound messages
 * @last_retransmitted: sequence number of most recently retransmitted message
109 110
 * @stale_cnt: counter for number of identical retransmit attempts
 * @stale_limit: time when repeated identical retransmits must force link reset
111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131
 * @ackers: # of peers that needs to ack each packet before it can be released
 * @acked: # last packet acked by a certain peer. Used for broadcast.
 * @rcv_nxt: next sequence number to expect for inbound messages
 * @deferred_queue: deferred queue saved OOS b'cast message received from node
 * @unacked_window: # of inbound messages rx'd without ack'ing back to peer
 * @inputq: buffer queue for messages to be delivered upwards
 * @namedq: buffer queue for name table messages to be delivered upwards
 * @next_out: ptr to first unsent outbound message in queue
 * @wakeupq: linked list of wakeup msgs waiting for link congestion to abate
 * @long_msg_seq_no: next identifier to use for outbound fragmented messages
 * @reasm_buf: head of partially reassembled inbound message fragments
 * @bc_rcvr: marks that this is a broadcast receiver link
 * @stats: collects statistics regarding link activity
 */
struct tipc_link {
	u32 addr;
	char name[TIPC_MAX_LINK_NAME];
	struct net *net;

	/* Management and link supervision data */
	u32 peer_session;
132
	u32 session;
133 134
	u16 snd_nxt_state;
	u16 rcv_nxt_state;
135 136 137 138 139 140 141 142
	u32 peer_bearer_id;
	u32 bearer_id;
	u32 tolerance;
	u32 abort_limit;
	u32 state;
	u16 peer_caps;
	bool active;
	u32 silent_intv_cnt;
143
	char if_name[TIPC_MAX_IF_NAME];
144 145
	u32 priority;
	char net_plane;
146
	struct tipc_mon_state mon_state;
147
	u16 rst_cnt;
148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166

	/* Failover/synch */
	u16 drop_point;
	struct sk_buff *failover_reasm_skb;

	/* Max packet negotiation */
	u16 mtu;
	u16 advertised_mtu;

	/* Sending */
	struct sk_buff_head transmq;
	struct sk_buff_head backlogq;
	struct {
		u16 len;
		u16 limit;
	} backlog[5];
	u16 snd_nxt;
	u16 last_retransm;
	u16 window;
167 168
	u16 stale_cnt;
	unsigned long stale_limit;
169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187

	/* Reception */
	u16 rcv_nxt;
	u32 rcv_unacked;
	struct sk_buff_head deferdq;
	struct sk_buff_head *inputq;
	struct sk_buff_head *namedq;

	/* Congestion handling */
	struct sk_buff_head wakeupq;

	/* Fragmentation/reassembly */
	struct sk_buff *reasm_buf;

	/* Broadcast */
	u16 ackers;
	u16 acked;
	struct tipc_link *bc_rcvlink;
	struct tipc_link *bc_sndlink;
188 189 190 191
	unsigned long prev_retr;
	u16 prev_from;
	u16 prev_to;
	u8 nack_state;
192 193 194 195 196 197
	bool bc_peer_is_up;

	/* Statistics */
	struct tipc_stats stats;
};

198 199 200
/*
 * Error message prefixes
 */
201
static const char *link_co_err = "Link tunneling error, ";
202
static const char *link_rst_msg = "Resetting link ";
203

204 205 206 207 208 209 210 211
/* Send states for broadcast NACKs
 */
enum {
	BC_NACK_SND_CONDITIONAL,
	BC_NACK_SND_UNCONDITIONAL,
	BC_NACK_SND_SUPPRESS,
};

212 213
#define TIPC_BC_RETR_LIMIT 10   /* [ms] */

214 215 216 217
/*
 * Interval between NACKs when packets arrive out of order
 */
#define TIPC_NACK_INTV (TIPC_MIN_LINK_WIN * 2)
218 219 220

/* Wildcard value for link session numbers. When it is known that
 * peer endpoint is down, any session number must be accepted.
221
 */
222
#define ANY_SESSION 0x10000
223

224
/* Link FSM states:
P
Per Liden 已提交
225
 */
226
enum {
227 228 229 230 231 232 233
	LINK_ESTABLISHED     = 0xe,
	LINK_ESTABLISHING    = 0xe  << 4,
	LINK_RESET           = 0x1  << 8,
	LINK_RESETTING       = 0x2  << 12,
	LINK_PEER_RESET      = 0xd  << 16,
	LINK_FAILINGOVER     = 0xf  << 20,
	LINK_SYNCHING        = 0xc  << 24
234 235 236 237
};

/* Link FSM state checking routines
 */
238
static int link_is_up(struct tipc_link *l)
239
{
240
	return l->state & (LINK_ESTABLISHED | LINK_SYNCHING);
241
}
P
Per Liden 已提交
242

243 244
static int tipc_link_proto_rcv(struct tipc_link *l, struct sk_buff *skb,
			       struct sk_buff_head *xmitq);
245
static void tipc_link_build_proto_msg(struct tipc_link *l, int mtyp, bool probe,
246 247
				      bool probe_reply, u16 rcvgap,
				      int tolerance, int priority,
248
				      struct sk_buff_head *xmitq);
249
static void link_print(struct tipc_link *l, const char *str);
250 251
static int tipc_link_build_nack_msg(struct tipc_link *l,
				    struct sk_buff_head *xmitq);
252 253 254
static void tipc_link_build_bc_init_msg(struct tipc_link *l,
					struct sk_buff_head *xmitq);
static bool tipc_link_release_pkts(struct tipc_link *l, u16 to);
255

P
Per Liden 已提交
256
/*
S
Sam Ravnborg 已提交
257
 *  Simple non-static link routines (i.e. referenced outside this file)
P
Per Liden 已提交
258
 */
259
bool tipc_link_is_up(struct tipc_link *l)
P
Per Liden 已提交
260
{
261 262 263
	return link_is_up(l);
}

264 265 266 267 268
bool tipc_link_peer_is_down(struct tipc_link *l)
{
	return l->state == LINK_PEER_RESET;
}

269 270 271 272 273
bool tipc_link_is_reset(struct tipc_link *l)
{
	return l->state & (LINK_RESET | LINK_FAILINGOVER | LINK_ESTABLISHING);
}

274 275 276 277 278
bool tipc_link_is_establishing(struct tipc_link *l)
{
	return l->state == LINK_ESTABLISHING;
}

279 280 281 282 283 284 285 286 287 288 289 290 291
bool tipc_link_is_synching(struct tipc_link *l)
{
	return l->state == LINK_SYNCHING;
}

bool tipc_link_is_failingover(struct tipc_link *l)
{
	return l->state == LINK_FAILINGOVER;
}

bool tipc_link_is_blocked(struct tipc_link *l)
{
	return l->state & (LINK_RESETTING | LINK_PEER_RESET | LINK_FAILINGOVER);
P
Per Liden 已提交
292 293
}

294
static bool link_is_bc_sndlink(struct tipc_link *l)
295 296 297 298
{
	return !l->bc_sndlink;
}

299
static bool link_is_bc_rcvlink(struct tipc_link *l)
300 301 302 303
{
	return ((l->bc_rcvlink == l) && !link_is_bc_sndlink(l));
}

304
int tipc_link_is_active(struct tipc_link *l)
P
Per Liden 已提交
305
{
306 307
	return l->active;
}
308

309 310 311
void tipc_link_set_active(struct tipc_link *l, bool active)
{
	l->active = active;
P
Per Liden 已提交
312 313
}

314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343
u32 tipc_link_id(struct tipc_link *l)
{
	return l->peer_bearer_id << 16 | l->bearer_id;
}

int tipc_link_window(struct tipc_link *l)
{
	return l->window;
}

int tipc_link_prio(struct tipc_link *l)
{
	return l->priority;
}

unsigned long tipc_link_tolerance(struct tipc_link *l)
{
	return l->tolerance;
}

struct sk_buff_head *tipc_link_inputq(struct tipc_link *l)
{
	return l->inputq;
}

char tipc_link_plane(struct tipc_link *l)
{
	return l->net_plane;
}

344 345 346 347 348
void tipc_link_update_caps(struct tipc_link *l, u16 capabilities)
{
	l->peer_caps = capabilities;
}

349 350 351
void tipc_link_add_bc_peer(struct tipc_link *snd_l,
			   struct tipc_link *uc_l,
			   struct sk_buff_head *xmitq)
352
{
353 354 355 356
	struct tipc_link *rcv_l = uc_l->bc_rcvlink;

	snd_l->ackers++;
	rcv_l->acked = snd_l->snd_nxt - 1;
357
	snd_l->state = LINK_ESTABLISHED;
358
	tipc_link_build_bc_init_msg(uc_l, xmitq);
359 360
}

361 362 363
void tipc_link_remove_bc_peer(struct tipc_link *snd_l,
			      struct tipc_link *rcv_l,
			      struct sk_buff_head *xmitq)
364
{
365 366 367
	u16 ack = snd_l->snd_nxt - 1;

	snd_l->ackers--;
368 369
	rcv_l->bc_peer_is_up = true;
	rcv_l->state = LINK_ESTABLISHED;
370 371 372 373 374
	tipc_link_bc_ack_rcv(rcv_l, ack, xmitq);
	tipc_link_reset(rcv_l);
	rcv_l->state = LINK_RESET;
	if (!snd_l->ackers) {
		tipc_link_reset(snd_l);
375
		snd_l->state = LINK_RESET;
376 377
		__skb_queue_purge(xmitq);
	}
378 379 380 381 382 383 384
}

int tipc_link_bc_peers(struct tipc_link *l)
{
	return l->ackers;
}

385 386 387 388 389 390 391 392 393 394 395 396
u16 link_bc_rcv_gap(struct tipc_link *l)
{
	struct sk_buff *skb = skb_peek(&l->deferdq);
	u16 gap = 0;

	if (more(l->snd_nxt, l->rcv_nxt))
		gap = l->snd_nxt - l->rcv_nxt;
	if (skb)
		gap = buf_seqno(skb) - l->rcv_nxt;
	return gap;
}

397 398 399 400 401 402 403 404 405 406
void tipc_link_set_mtu(struct tipc_link *l, int mtu)
{
	l->mtu = mtu;
}

int tipc_link_mtu(struct tipc_link *l)
{
	return l->mtu;
}

407 408 409 410 411 412 413 414 415 416 417 418 419 420 421
u16 tipc_link_rcv_nxt(struct tipc_link *l)
{
	return l->rcv_nxt;
}

u16 tipc_link_acked(struct tipc_link *l)
{
	return l->acked;
}

char *tipc_link_name(struct tipc_link *l)
{
	return l->name;
}

P
Per Liden 已提交
422
/**
423
 * tipc_link_create - create a new link
J
Jon Paul Maloy 已提交
424
 * @n: pointer to associated node
425 426 427 428 429 430 431 432
 * @if_name: associated interface name
 * @bearer_id: id (index) of associated bearer
 * @tolerance: link tolerance to be used by link
 * @net_plane: network plane (A,B,c..) this link belongs to
 * @mtu: mtu to be advertised by link
 * @priority: priority to be used by link
 * @window: send window to be used by link
 * @session: session to be used by link
J
Jon Paul Maloy 已提交
433
 * @ownnode: identity of own node
434
 * @peer: node id of peer node
435
 * @peer_caps: bitmap describing peer node capabilities
436 437
 * @bc_sndlink: the namespace global link used for broadcast sending
 * @bc_rcvlink: the peer specific link used for broadcast reception
J
Jon Paul Maloy 已提交
438 439 440
 * @inputq: queue to put messages ready for delivery
 * @namedq: queue to put binding table update messages ready for delivery
 * @link: return value, pointer to put the created link
441
 *
J
Jon Paul Maloy 已提交
442
 * Returns true if link was created, otherwise false
P
Per Liden 已提交
443
 */
444
bool tipc_link_create(struct net *net, char *if_name, int bearer_id,
445
		      int tolerance, char net_plane, u32 mtu, int priority,
446 447
		      int window, u32 session, u32 self,
		      u32 peer, u8 *peer_id, u16 peer_caps,
448 449 450 451
		      struct tipc_link *bc_sndlink,
		      struct tipc_link *bc_rcvlink,
		      struct sk_buff_head *inputq,
		      struct sk_buff_head *namedq,
J
Jon Paul Maloy 已提交
452
		      struct tipc_link **link)
P
Per Liden 已提交
453
{
454 455
	char peer_str[NODE_ID_STR_LEN] = {0,};
	char self_str[NODE_ID_STR_LEN] = {0,};
J
Jon Paul Maloy 已提交
456
	struct tipc_link *l;
457

J
Jon Paul Maloy 已提交
458 459 460 461
	l = kzalloc(sizeof(*l), GFP_ATOMIC);
	if (!l)
		return false;
	*link = l;
462
	l->session = session;
463

464 465 466 467 468 469 470 471 472 473
	/* Set link name for unicast links only */
	if (peer_id) {
		tipc_nodeid2string(self_str, tipc_own_id(net));
		if (strlen(self_str) > 16)
			sprintf(self_str, "%x", self);
		tipc_nodeid2string(peer_str, peer_id);
		if (strlen(peer_str) > 16)
			sprintf(peer_str, "%x", peer);
	}
	/* Peer i/f name will be completed by reset/activate message */
J
Jon Maloy 已提交
474 475
	snprintf(l->name, sizeof(l->name), "%s:%s-%s:unknown",
		 self_str, if_name, peer_str);
476

477
	strcpy(l->if_name, if_name);
J
Jon Paul Maloy 已提交
478
	l->addr = peer;
479
	l->peer_caps = peer_caps;
480
	l->net = net;
481
	l->peer_session = ANY_SESSION;
482 483 484 485 486 487 488
	l->bearer_id = bearer_id;
	l->tolerance = tolerance;
	l->net_plane = net_plane;
	l->advertised_mtu = mtu;
	l->mtu = mtu;
	l->priority = priority;
	tipc_link_set_queue_limits(l, window);
489
	l->ackers = 1;
490 491
	l->bc_sndlink = bc_sndlink;
	l->bc_rcvlink = bc_rcvlink;
J
Jon Paul Maloy 已提交
492 493 494 495 496 497 498 499 500
	l->inputq = inputq;
	l->namedq = namedq;
	l->state = LINK_RESETTING;
	__skb_queue_head_init(&l->transmq);
	__skb_queue_head_init(&l->backlogq);
	__skb_queue_head_init(&l->deferdq);
	skb_queue_head_init(&l->wakeupq);
	skb_queue_head_init(l->inputq);
	return true;
P
Per Liden 已提交
501 502
}

503 504 505
/**
 * tipc_link_bc_create - create new link to be used for broadcast
 * @n: pointer to associated node
506
 * @mtu: mtu to be used initially if no peers
507 508 509 510 511 512 513
 * @window: send window to be used
 * @inputq: queue to put messages ready for delivery
 * @namedq: queue to put binding table update messages ready for delivery
 * @link: return value, pointer to put the created link
 *
 * Returns true if link was created, otherwise false
 */
514
bool tipc_link_bc_create(struct net *net, u32 ownnode, u32 peer,
515
			 int mtu, int window, u16 peer_caps,
516 517
			 struct sk_buff_head *inputq,
			 struct sk_buff_head *namedq,
518
			 struct tipc_link *bc_sndlink,
519 520 521 522
			 struct tipc_link **link)
{
	struct tipc_link *l;

523
	if (!tipc_link_create(net, "", MAX_BEARERS, 0, 'Z', mtu, 0, window,
524
			      0, ownnode, peer, NULL, peer_caps, bc_sndlink,
525
			      NULL, inputq, namedq, link))
526 527 528 529 530
		return false;

	l = *link;
	strcpy(l->name, tipc_bclink_name);
	tipc_link_reset(l);
531
	l->state = LINK_RESET;
532
	l->ackers = 0;
533
	l->bc_rcvlink = l;
534

535 536 537
	/* Broadcast send link is always up */
	if (link_is_bc_sndlink(l))
		l->state = LINK_ESTABLISHED;
538

539 540 541 542
	/* Disable replicast if even a single peer doesn't support it */
	if (link_is_bc_rcvlink(l) && !(peer_caps & TIPC_BCAST_RCAST))
		tipc_bcast_disable_rcast(net);

543
	return true;
544 545
}

546 547 548 549 550
/**
 * tipc_link_fsm_evt - link finite state machine
 * @l: pointer to link
 * @evt: state machine event to be processed
 */
551
int tipc_link_fsm_evt(struct tipc_link *l, int evt)
552
{
553
	int rc = 0;
554 555

	switch (l->state) {
556
	case LINK_RESETTING:
557
		switch (evt) {
558 559
		case LINK_PEER_RESET_EVT:
			l->state = LINK_PEER_RESET;
560
			break;
561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577
		case LINK_RESET_EVT:
			l->state = LINK_RESET;
			break;
		case LINK_FAILURE_EVT:
		case LINK_FAILOVER_BEGIN_EVT:
		case LINK_ESTABLISH_EVT:
		case LINK_FAILOVER_END_EVT:
		case LINK_SYNCH_BEGIN_EVT:
		case LINK_SYNCH_END_EVT:
		default:
			goto illegal_evt;
		}
		break;
	case LINK_RESET:
		switch (evt) {
		case LINK_PEER_RESET_EVT:
			l->state = LINK_ESTABLISHING;
578
			break;
579 580 581 582 583 584
		case LINK_FAILOVER_BEGIN_EVT:
			l->state = LINK_FAILINGOVER;
		case LINK_FAILURE_EVT:
		case LINK_RESET_EVT:
		case LINK_ESTABLISH_EVT:
		case LINK_FAILOVER_END_EVT:
585
			break;
586 587
		case LINK_SYNCH_BEGIN_EVT:
		case LINK_SYNCH_END_EVT:
588
		default:
589
			goto illegal_evt;
590 591
		}
		break;
592
	case LINK_PEER_RESET:
593
		switch (evt) {
594 595
		case LINK_RESET_EVT:
			l->state = LINK_ESTABLISHING;
596
			break;
597 598 599
		case LINK_PEER_RESET_EVT:
		case LINK_ESTABLISH_EVT:
		case LINK_FAILURE_EVT:
600
			break;
601 602 603 604 605 606 607 608 609 610 611 612
		case LINK_SYNCH_BEGIN_EVT:
		case LINK_SYNCH_END_EVT:
		case LINK_FAILOVER_BEGIN_EVT:
		case LINK_FAILOVER_END_EVT:
		default:
			goto illegal_evt;
		}
		break;
	case LINK_FAILINGOVER:
		switch (evt) {
		case LINK_FAILOVER_END_EVT:
			l->state = LINK_RESET;
613
			break;
614 615 616 617 618 619 620 621
		case LINK_PEER_RESET_EVT:
		case LINK_RESET_EVT:
		case LINK_ESTABLISH_EVT:
		case LINK_FAILURE_EVT:
			break;
		case LINK_FAILOVER_BEGIN_EVT:
		case LINK_SYNCH_BEGIN_EVT:
		case LINK_SYNCH_END_EVT:
622
		default:
623
			goto illegal_evt;
624 625
		}
		break;
626
	case LINK_ESTABLISHING:
627
		switch (evt) {
628 629
		case LINK_ESTABLISH_EVT:
			l->state = LINK_ESTABLISHED;
630
			break;
631 632 633 634
		case LINK_FAILOVER_BEGIN_EVT:
			l->state = LINK_FAILINGOVER;
			break;
		case LINK_RESET_EVT:
635 636
			l->state = LINK_RESET;
			break;
637
		case LINK_FAILURE_EVT:
638
		case LINK_PEER_RESET_EVT:
639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655
		case LINK_SYNCH_BEGIN_EVT:
		case LINK_FAILOVER_END_EVT:
			break;
		case LINK_SYNCH_END_EVT:
		default:
			goto illegal_evt;
		}
		break;
	case LINK_ESTABLISHED:
		switch (evt) {
		case LINK_PEER_RESET_EVT:
			l->state = LINK_PEER_RESET;
			rc |= TIPC_LINK_DOWN_EVT;
			break;
		case LINK_FAILURE_EVT:
			l->state = LINK_RESETTING;
			rc |= TIPC_LINK_DOWN_EVT;
656
			break;
657 658
		case LINK_RESET_EVT:
			l->state = LINK_RESET;
659
			break;
660
		case LINK_ESTABLISH_EVT:
661
		case LINK_SYNCH_END_EVT:
662
			break;
663 664 665 666 667
		case LINK_SYNCH_BEGIN_EVT:
			l->state = LINK_SYNCHING;
			break;
		case LINK_FAILOVER_BEGIN_EVT:
		case LINK_FAILOVER_END_EVT:
668
		default:
669
			goto illegal_evt;
670 671
		}
		break;
672
	case LINK_SYNCHING:
673
		switch (evt) {
674 675 676 677 678 679 680
		case LINK_PEER_RESET_EVT:
			l->state = LINK_PEER_RESET;
			rc |= TIPC_LINK_DOWN_EVT;
			break;
		case LINK_FAILURE_EVT:
			l->state = LINK_RESETTING;
			rc |= TIPC_LINK_DOWN_EVT;
681
			break;
682 683
		case LINK_RESET_EVT:
			l->state = LINK_RESET;
684
			break;
685 686
		case LINK_ESTABLISH_EVT:
		case LINK_SYNCH_BEGIN_EVT:
687
			break;
688 689 690 691 692
		case LINK_SYNCH_END_EVT:
			l->state = LINK_ESTABLISHED;
			break;
		case LINK_FAILOVER_BEGIN_EVT:
		case LINK_FAILOVER_END_EVT:
693
		default:
694
			goto illegal_evt;
695 696 697
		}
		break;
	default:
698
		pr_err("Unknown FSM state %x in %s\n", l->state, l->name);
699
	}
700 701 702 703
	return rc;
illegal_evt:
	pr_err("Illegal FSM event %x in state %x on link %s\n",
	       evt, l->state, l->name);
704 705 706
	return rc;
}

707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751
/* link_profile_stats - update statistical profiling of traffic
 */
static void link_profile_stats(struct tipc_link *l)
{
	struct sk_buff *skb;
	struct tipc_msg *msg;
	int length;

	/* Update counters used in statistical profiling of send traffic */
	l->stats.accu_queue_sz += skb_queue_len(&l->transmq);
	l->stats.queue_sz_counts++;

	skb = skb_peek(&l->transmq);
	if (!skb)
		return;
	msg = buf_msg(skb);
	length = msg_size(msg);

	if (msg_user(msg) == MSG_FRAGMENTER) {
		if (msg_type(msg) != FIRST_FRAGMENT)
			return;
		length = msg_size(msg_get_wrapped(msg));
	}
	l->stats.msg_lengths_total += length;
	l->stats.msg_length_counts++;
	if (length <= 64)
		l->stats.msg_length_profile[0]++;
	else if (length <= 256)
		l->stats.msg_length_profile[1]++;
	else if (length <= 1024)
		l->stats.msg_length_profile[2]++;
	else if (length <= 4096)
		l->stats.msg_length_profile[3]++;
	else if (length <= 16384)
		l->stats.msg_length_profile[4]++;
	else if (length <= 32768)
		l->stats.msg_length_profile[5]++;
	else
		l->stats.msg_length_profile[6]++;
}

/* tipc_link_timeout - perform periodic task as instructed from node timeout
 */
int tipc_link_timeout(struct tipc_link *l, struct sk_buff_head *xmitq)
{
752 753
	int mtyp = 0;
	int rc = 0;
754 755 756
	bool state = false;
	bool probe = false;
	bool setup = false;
757 758
	u16 bc_snt = l->bc_sndlink->snd_nxt - 1;
	u16 bc_acked = l->bc_rcvlink->acked;
759
	struct tipc_mon_state *mstate = &l->mon_state;
760

761 762 763
	switch (l->state) {
	case LINK_ESTABLISHED:
	case LINK_SYNCHING:
764
		mtyp = STATE_MSG;
765 766 767 768
		link_profile_stats(l);
		tipc_mon_get_state(l->net, l->addr, mstate, l->bearer_id);
		if (mstate->reset || (l->silent_intv_cnt > l->abort_limit))
			return tipc_link_fsm_evt(l, LINK_FAILURE_EVT);
769
		state = bc_acked != bc_snt;
770 771 772 773 774 775 776 777
		state |= l->bc_rcvlink->rcv_unacked;
		state |= l->rcv_unacked;
		state |= !skb_queue_empty(&l->transmq);
		state |= !skb_queue_empty(&l->deferdq);
		probe = mstate->probing;
		probe |= l->silent_intv_cnt;
		if (probe || mstate->monitoring)
			l->silent_intv_cnt++;
778 779
		break;
	case LINK_RESET:
780 781
		setup = l->rst_cnt++ <= 4;
		setup |= !(l->rst_cnt % 16);
782
		mtyp = RESET_MSG;
783 784
		break;
	case LINK_ESTABLISHING:
785
		setup = true;
786
		mtyp = ACTIVATE_MSG;
787 788
		break;
	case LINK_PEER_RESET:
789
	case LINK_RESETTING:
790 791 792 793
	case LINK_FAILINGOVER:
		break;
	default:
		break;
794
	}
795

796
	if (state || probe || setup)
797
		tipc_link_build_proto_msg(l, mtyp, probe, 0, 0, 0, 0, xmitq);
798

799 800 801
	return rc;
}

P
Per Liden 已提交
802
/**
803
 * link_schedule_user - schedule a message sender for wakeup after congestion
804 805
 * @l: congested link
 * @hdr: header of message that is being sent
806
 * Create pseudo msg to send back to user when congestion abates
P
Per Liden 已提交
807
 */
808
static int link_schedule_user(struct tipc_link *l, struct tipc_msg *hdr)
P
Per Liden 已提交
809
{
810 811
	u32 dnode = tipc_own_addr(l->net);
	u32 dport = msg_origport(hdr);
812 813 814 815
	struct sk_buff *skb;

	/* Create and schedule wakeup pseudo message */
	skb = tipc_msg_create(SOCK_WAKEUP, 0, INT_H_SIZE, 0,
816
			      dnode, l->addr, dport, 0, 0);
817
	if (!skb)
818
		return -ENOBUFS;
819 820 821 822
	msg_set_dest_droppable(buf_msg(skb), true);
	TIPC_SKB_CB(skb)->chain_imp = msg_importance(hdr);
	skb_queue_tail(&l->wakeupq, skb);
	l->stats.link_congs++;
823
	return -ELINKCONG;
P
Per Liden 已提交
824 825
}

826 827
/**
 * link_prepare_wakeup - prepare users for wakeup after congestion
828 829 830
 * @l: congested link
 * Wake up a number of waiting users, as permitted by available space
 * in the send queue
831
 */
832
void link_prepare_wakeup(struct tipc_link *l)
P
Per Liden 已提交
833
{
834
	struct sk_buff *skb, *tmp;
835
	int imp, i = 0;
836

837 838
	skb_queue_walk_safe(&l->wakeupq, skb, tmp) {
		imp = TIPC_SKB_CB(skb)->chain_imp;
839 840 841 842
		if (l->backlog[imp].len < l->backlog[imp].limit) {
			skb_unlink(skb, &l->wakeupq);
			skb_queue_tail(l->inputq, skb);
		} else if (i++ > 10) {
P
Per Liden 已提交
843
			break;
844
		}
P
Per Liden 已提交
845 846 847
	}
}

848
void tipc_link_reset(struct tipc_link *l)
P
Per Liden 已提交
849
{
850 851
	l->peer_session = ANY_SESSION;
	l->session++;
852 853 854
	l->mtu = l->advertised_mtu;
	__skb_queue_purge(&l->transmq);
	__skb_queue_purge(&l->deferdq);
855
	skb_queue_splice_init(&l->wakeupq, l->inputq);
856 857 858 859 860 861
	__skb_queue_purge(&l->backlogq);
	l->backlog[TIPC_LOW_IMPORTANCE].len = 0;
	l->backlog[TIPC_MEDIUM_IMPORTANCE].len = 0;
	l->backlog[TIPC_HIGH_IMPORTANCE].len = 0;
	l->backlog[TIPC_CRITICAL_IMPORTANCE].len = 0;
	l->backlog[TIPC_SYSTEM_IMPORTANCE].len = 0;
862 863 864 865 866 867 868
	kfree_skb(l->reasm_buf);
	kfree_skb(l->failover_reasm_skb);
	l->reasm_buf = NULL;
	l->failover_reasm_skb = NULL;
	l->rcv_unacked = 0;
	l->snd_nxt = 1;
	l->rcv_nxt = 1;
869 870
	l->snd_nxt_state = 1;
	l->rcv_nxt_state = 1;
871
	l->acked = 0;
872
	l->silent_intv_cnt = 0;
873
	l->rst_cnt = 0;
874
	l->stale_cnt = 0;
875
	l->bc_peer_is_up = false;
876
	memset(&l->mon_state, 0, sizeof(l->mon_state));
877
	tipc_link_reset_stats(l);
P
Per Liden 已提交
878 879
}

880 881 882 883 884 885
/**
 * tipc_link_xmit(): enqueue buffer list according to queue situation
 * @link: link to use
 * @list: chain of buffers containing message
 * @xmitq: returned list of packets to be sent by caller
 *
886
 * Consumes the buffer chain.
887 888 889 890 891 892 893 894
 * Returns 0 if success, or errno: -ELINKCONG, -EMSGSIZE or -ENOBUFS
 * Messages at TIPC_SYSTEM_IMPORTANCE are always accepted
 */
int tipc_link_xmit(struct tipc_link *l, struct sk_buff_head *list,
		   struct sk_buff_head *xmitq)
{
	struct tipc_msg *hdr = buf_msg(skb_peek(list));
	unsigned int maxwin = l->window;
895
	int imp = msg_importance(hdr);
896 897 898
	unsigned int mtu = l->mtu;
	u16 ack = l->rcv_nxt - 1;
	u16 seqno = l->snd_nxt;
899
	u16 bc_ack = l->bc_rcvlink->rcv_nxt - 1;
900 901 902
	struct sk_buff_head *transmq = &l->transmq;
	struct sk_buff_head *backlogq = &l->backlogq;
	struct sk_buff *skb, *_skb, *bskb;
903
	int pkt_cnt = skb_queue_len(list);
904
	int rc = 0;
905

906 907
	if (unlikely(msg_size(hdr) > mtu)) {
		skb_queue_purge(list);
908
		return -EMSGSIZE;
909
	}
910

911 912 913 914 915 916 917 918 919
	/* Allow oversubscription of one data msg per source at congestion */
	if (unlikely(l->backlog[imp].len >= l->backlog[imp].limit)) {
		if (imp == TIPC_SYSTEM_IMPORTANCE) {
			pr_warn("%s<%s>, link overflow", link_rst_msg, l->name);
			return -ENOBUFS;
		}
		rc = link_schedule_user(l, hdr);
	}

920 921 922 923 924
	if (pkt_cnt > 1) {
		l->stats.sent_fragmented++;
		l->stats.sent_fragments += pkt_cnt;
	}

925 926 927 928 929 930
	/* Prepare each packet for sending, and add to relevant queue: */
	while (skb_queue_len(list)) {
		skb = skb_peek(list);
		hdr = buf_msg(skb);
		msg_set_seqno(hdr, seqno);
		msg_set_ack(hdr, ack);
931
		msg_set_bcast_ack(hdr, bc_ack);
932 933 934

		if (likely(skb_queue_len(transmq) < maxwin)) {
			_skb = skb_clone(skb, GFP_ATOMIC);
935 936
			if (!_skb) {
				skb_queue_purge(list);
937
				return -ENOBUFS;
938
			}
939 940 941
			__skb_dequeue(list);
			__skb_queue_tail(transmq, skb);
			__skb_queue_tail(xmitq, _skb);
942
			TIPC_SKB_CB(skb)->ackers = l->ackers;
943
			l->rcv_unacked = 0;
944
			l->stats.sent_pkts++;
945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964
			seqno++;
			continue;
		}
		if (tipc_msg_bundle(skb_peek_tail(backlogq), hdr, mtu)) {
			kfree_skb(__skb_dequeue(list));
			l->stats.sent_bundled++;
			continue;
		}
		if (tipc_msg_make_bundle(&bskb, hdr, mtu, l->addr)) {
			kfree_skb(__skb_dequeue(list));
			__skb_queue_tail(backlogq, bskb);
			l->backlog[msg_importance(buf_msg(bskb))].len++;
			l->stats.sent_bundled++;
			l->stats.sent_bundles++;
			continue;
		}
		l->backlog[imp].len += skb_queue_len(list);
		skb_queue_splice_tail_init(list, backlogq);
	}
	l->snd_nxt = seqno;
965
	return rc;
966 967
}

968 969 970 971 972 973
void tipc_link_advance_backlog(struct tipc_link *l, struct sk_buff_head *xmitq)
{
	struct sk_buff *skb, *_skb;
	struct tipc_msg *hdr;
	u16 seqno = l->snd_nxt;
	u16 ack = l->rcv_nxt - 1;
974
	u16 bc_ack = l->bc_rcvlink->rcv_nxt - 1;
975 976 977 978 979 980 981 982 983 984 985 986 987

	while (skb_queue_len(&l->transmq) < l->window) {
		skb = skb_peek(&l->backlogq);
		if (!skb)
			break;
		_skb = skb_clone(skb, GFP_ATOMIC);
		if (!_skb)
			break;
		__skb_dequeue(&l->backlogq);
		hdr = buf_msg(skb);
		l->backlog[msg_importance(hdr)].len--;
		__skb_queue_tail(&l->transmq, skb);
		__skb_queue_tail(xmitq, _skb);
988
		TIPC_SKB_CB(skb)->ackers = l->ackers;
989
		msg_set_seqno(hdr, seqno);
990 991
		msg_set_ack(hdr, ack);
		msg_set_bcast_ack(hdr, bc_ack);
992
		l->rcv_unacked = 0;
993
		l->stats.sent_pkts++;
994 995 996 997 998
		seqno++;
	}
	l->snd_nxt = seqno;
}

999
static void link_retransmit_failure(struct tipc_link *l, struct sk_buff *skb)
1000
{
1001 1002 1003
	struct tipc_msg *hdr = buf_msg(skb);

	pr_warn("Retransmission failure on link <%s>\n", l->name);
1004
	link_print(l, "State of link ");
1005 1006 1007 1008
	pr_info("Failed msg: usr %u, typ %u, len %u, err %u\n",
		msg_user(hdr), msg_type(hdr), msg_size(hdr), msg_errcode(hdr));
	pr_info("sqno %u, prev: %x, src: %x\n",
		msg_seqno(hdr), msg_prevnode(hdr), msg_orignode(hdr));
1009 1010
}

1011 1012 1013 1014 1015 1016 1017 1018
/* tipc_link_retrans() - retransmit one or more packets
 * @l: the link to transmit on
 * @r: the receiving link ordering the retransmit. Same as l if unicast
 * @from: retransmit from (inclusive) this sequence number
 * @to: retransmit to (inclusive) this sequence number
 * xmitq: queue for accumulating the retransmitted packets
 */
int tipc_link_retrans(struct tipc_link *l, struct tipc_link *r,
1019
		      u16 from, u16 to, struct sk_buff_head *xmitq)
1020 1021
{
	struct sk_buff *_skb, *skb = skb_peek(&l->transmq);
1022
	u16 bc_ack = l->bc_rcvlink->rcv_nxt - 1;
1023 1024
	u16 ack = l->rcv_nxt - 1;
	struct tipc_msg *hdr;
1025 1026 1027 1028 1029

	if (!skb)
		return 0;

	/* Detect repeated retransmit failures on same packet */
1030 1031 1032 1033
	if (r->last_retransm != buf_seqno(skb)) {
		r->last_retransm = buf_seqno(skb);
		r->stale_limit = jiffies + msecs_to_jiffies(l->tolerance);
	} else if (++r->stale_cnt > 99 && time_after(jiffies, r->stale_limit)) {
1034
		link_retransmit_failure(l, skb);
1035 1036
		if (link_is_bc_sndlink(l))
			return TIPC_LINK_DOWN_EVT;
1037
		return tipc_link_fsm_evt(l, LINK_FAILURE_EVT);
1038
	}
1039

1040 1041
	skb_queue_walk(&l->transmq, skb) {
		hdr = buf_msg(skb);
1042 1043 1044 1045
		if (less(msg_seqno(hdr), from))
			continue;
		if (more(msg_seqno(hdr), to))
			break;
1046 1047 1048 1049
		_skb = __pskb_copy(skb, MIN_H_SIZE, GFP_ATOMIC);
		if (!_skb)
			return 0;
		hdr = buf_msg(_skb);
1050 1051
		msg_set_ack(hdr, ack);
		msg_set_bcast_ack(hdr, bc_ack);
1052 1053 1054 1055 1056 1057 1058
		_skb->priority = TC_PRIO_CONTROL;
		__skb_queue_tail(xmitq, _skb);
		l->stats.retransmitted++;
	}
	return 0;
}

1059
/* tipc_data_input - deliver data and name distr msgs to upper layer
1060
 *
1061
 * Consumes buffer if message is of right type
1062 1063
 * Node lock must be held
 */
1064
static bool tipc_data_input(struct tipc_link *l, struct sk_buff *skb,
1065
			    struct sk_buff_head *inputq)
1066
{
1067
	struct sk_buff_head *mc_inputq = l->bc_rcvlink->inputq;
1068 1069 1070
	struct tipc_msg *hdr = buf_msg(skb);

	switch (msg_user(hdr)) {
1071 1072 1073 1074
	case TIPC_LOW_IMPORTANCE:
	case TIPC_MEDIUM_IMPORTANCE:
	case TIPC_HIGH_IMPORTANCE:
	case TIPC_CRITICAL_IMPORTANCE:
1075
		if (unlikely(msg_in_group(hdr) || msg_mcast(hdr))) {
1076
			skb_queue_tail(mc_inputq, skb);
1077 1078
			return true;
		}
1079
		/* else: fall through */
1080
	case CONN_MANAGER:
J
Jon Maloy 已提交
1081
		skb_queue_tail(inputq, skb);
1082
		return true;
1083 1084 1085
	case GROUP_PROTOCOL:
		skb_queue_tail(mc_inputq, skb);
		return true;
1086
	case NAME_DISTRIBUTOR:
1087 1088
		l->bc_rcvlink->state = LINK_ESTABLISHED;
		skb_queue_tail(l->namedq, skb);
1089 1090
		return true;
	case MSG_BUNDLER:
1091
	case TUNNEL_PROTOCOL:
1092
	case MSG_FRAGMENTER:
1093
	case BCAST_PROTOCOL:
1094
		return false;
1095
	default:
1096 1097 1098 1099
		pr_warn("Dropping received illegal msg type\n");
		kfree_skb(skb);
		return false;
	};
1100
}
1101 1102 1103 1104

/* tipc_link_input - process packet that has passed link protocol check
 *
 * Consumes buffer
1105
 */
1106 1107
static int tipc_link_input(struct tipc_link *l, struct sk_buff *skb,
			   struct sk_buff_head *inputq)
1108
{
1109 1110
	struct tipc_msg *hdr = buf_msg(skb);
	struct sk_buff **reasm_skb = &l->reasm_buf;
1111
	struct sk_buff *iskb;
1112
	struct sk_buff_head tmpq;
1113
	int usr = msg_user(hdr);
1114
	int rc = 0;
1115 1116
	int pos = 0;
	int ipos = 0;
1117

1118 1119 1120 1121
	if (unlikely(usr == TUNNEL_PROTOCOL)) {
		if (msg_type(hdr) == SYNCH_MSG) {
			__skb_queue_purge(&l->deferdq);
			goto drop;
1122
		}
1123 1124 1125 1126 1127 1128 1129
		if (!tipc_msg_extract(skb, &iskb, &ipos))
			return rc;
		kfree_skb(skb);
		skb = iskb;
		hdr = buf_msg(skb);
		if (less(msg_seqno(hdr), l->drop_point))
			goto drop;
1130
		if (tipc_data_input(l, skb, inputq))
1131 1132 1133 1134
			return rc;
		usr = msg_user(hdr);
		reasm_skb = &l->failover_reasm_skb;
	}
1135

1136
	if (usr == MSG_BUNDLER) {
1137
		skb_queue_head_init(&tmpq);
1138 1139
		l->stats.recv_bundles++;
		l->stats.recv_bundled += msg_msgcnt(hdr);
1140
		while (tipc_msg_extract(skb, &iskb, &pos))
1141 1142
			tipc_data_input(l, iskb, &tmpq);
		tipc_skb_queue_splice_tail(&tmpq, inputq);
1143
		return 0;
1144 1145 1146 1147
	} else if (usr == MSG_FRAGMENTER) {
		l->stats.recv_fragments++;
		if (tipc_buf_append(reasm_skb, &skb)) {
			l->stats.recv_fragmented++;
1148
			tipc_data_input(l, skb, inputq);
1149 1150
		} else if (!*reasm_skb && !link_is_bc_rcvlink(l)) {
			pr_warn_ratelimited("Unable to build fragment list\n");
1151
			return tipc_link_fsm_evt(l, LINK_FAILURE_EVT);
1152
		}
1153
		return 0;
1154
	} else if (usr == BCAST_PROTOCOL) {
1155
		tipc_bcast_lock(l->net);
1156
		tipc_link_bc_init_rcv(l->bc_rcvlink, hdr);
1157
		tipc_bcast_unlock(l->net);
1158 1159 1160
	}
drop:
	kfree_skb(skb);
1161
	return 0;
1162 1163
}

1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178
static bool tipc_link_release_pkts(struct tipc_link *l, u16 acked)
{
	bool released = false;
	struct sk_buff *skb, *tmp;

	skb_queue_walk_safe(&l->transmq, skb, tmp) {
		if (more(buf_seqno(skb), acked))
			break;
		__skb_unlink(skb, &l->transmq);
		kfree_skb(skb);
		released = true;
	}
	return released;
}

1179
/* tipc_link_build_state_msg: prepare link state message for transmission
1180 1181 1182
 *
 * Note that sending of broadcast ack is coordinated among nodes, to reduce
 * risk of ack storms towards the sender
1183
 */
1184
int tipc_link_build_state_msg(struct tipc_link *l, struct sk_buff_head *xmitq)
1185
{
1186 1187 1188 1189 1190
	if (!l)
		return 0;

	/* Broadcast ACK must be sent via a unicast link => defer to caller */
	if (link_is_bc_rcvlink(l)) {
1191
		if (((l->rcv_nxt ^ tipc_own_addr(l->net)) & 0xf) != 0xf)
1192 1193
			return 0;
		l->rcv_unacked = 0;
1194 1195 1196 1197

		/* Use snd_nxt to store peer's snd_nxt in broadcast rcv link */
		l->snd_nxt = l->rcv_nxt;
		return TIPC_LINK_SND_STATE;
1198 1199 1200
	}

	/* Unicast ACK */
1201 1202
	l->rcv_unacked = 0;
	l->stats.sent_acks++;
1203
	tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, 0, 0, 0, xmitq);
1204
	return 0;
1205 1206
}

1207 1208 1209 1210 1211
/* tipc_link_build_reset_msg: prepare link RESET or ACTIVATE message
 */
void tipc_link_build_reset_msg(struct tipc_link *l, struct sk_buff_head *xmitq)
{
	int mtyp = RESET_MSG;
1212
	struct sk_buff *skb;
1213 1214 1215 1216

	if (l->state == LINK_ESTABLISHING)
		mtyp = ACTIVATE_MSG;

1217
	tipc_link_build_proto_msg(l, mtyp, 0, 0, 0, 0, 0, xmitq);
1218 1219 1220 1221 1222

	/* Inform peer that this endpoint is going down if applicable */
	skb = skb_peek_tail(xmitq);
	if (skb && (l->state == LINK_RESET))
		msg_set_peer_stopping(buf_msg(skb), 1);
1223 1224
}

1225
/* tipc_link_build_nack_msg: prepare link nack message for transmission
1226 1227
 * Note that sending of broadcast NACK is coordinated among nodes, to
 * reduce the risk of NACK storms towards the sender
1228
 */
1229 1230
static int tipc_link_build_nack_msg(struct tipc_link *l,
				    struct sk_buff_head *xmitq)
1231 1232
{
	u32 def_cnt = ++l->stats.deferred_recv;
1233
	int match1, match2;
1234

1235 1236 1237 1238 1239 1240 1241
	if (link_is_bc_rcvlink(l)) {
		match1 = def_cnt & 0xf;
		match2 = tipc_own_addr(l->net) & 0xf;
		if (match1 == match2)
			return TIPC_LINK_SND_STATE;
		return 0;
	}
1242

1243
	if ((skb_queue_len(&l->deferdq) == 1) || !(def_cnt % TIPC_NACK_INTV))
1244
		tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, 0, 0, 0, xmitq);
1245
	return 0;
1246 1247
}

1248
/* tipc_link_rcv - process TIPC packets/messages arriving from off-node
1249
 * @l: the link that should handle the message
1250 1251 1252 1253 1254 1255
 * @skb: TIPC packet
 * @xmitq: queue to place packets to be sent after this call
 */
int tipc_link_rcv(struct tipc_link *l, struct sk_buff *skb,
		  struct sk_buff_head *xmitq)
{
1256
	struct sk_buff_head *defq = &l->deferdq;
1257
	struct tipc_msg *hdr;
1258
	u16 seqno, rcv_nxt, win_lim;
1259 1260
	int rc = 0;

1261
	do {
1262
		hdr = buf_msg(skb);
1263 1264
		seqno = msg_seqno(hdr);
		rcv_nxt = l->rcv_nxt;
1265
		win_lim = rcv_nxt + TIPC_MAX_LINK_WIN;
1266 1267

		/* Verify and update link state */
1268 1269
		if (unlikely(msg_user(hdr) == LINK_PROTOCOL))
			return tipc_link_proto_rcv(l, skb, xmitq);
1270

1271
		if (unlikely(!link_is_up(l))) {
1272 1273 1274
			if (l->state == LINK_ESTABLISHING)
				rc = TIPC_LINK_UP_EVT;
			goto drop;
1275 1276
		}

1277
		/* Don't send probe at next timeout expiration */
1278 1279
		l->silent_intv_cnt = 0;

1280 1281 1282 1283 1284 1285
		/* Drop if outside receive window */
		if (unlikely(less(seqno, rcv_nxt) || more(seqno, win_lim))) {
			l->stats.duplicates++;
			goto drop;
		}

1286 1287
		/* Forward queues and wake up waiting users */
		if (likely(tipc_link_release_pkts(l, msg_ack(hdr)))) {
1288
			l->stale_cnt = 0;
1289 1290 1291 1292 1293
			tipc_link_advance_backlog(l, xmitq);
			if (unlikely(!skb_queue_empty(&l->wakeupq)))
				link_prepare_wakeup(l);
		}

1294 1295
		/* Defer delivery if sequence gap */
		if (unlikely(seqno != rcv_nxt)) {
1296
			__tipc_skb_queue_sorted(defq, seqno, skb);
1297
			rc |= tipc_link_build_nack_msg(l, xmitq);
1298
			break;
1299 1300
		}

1301
		/* Deliver packet */
1302
		l->rcv_nxt++;
1303
		l->stats.recv_pkts++;
1304
		if (!tipc_data_input(l, skb, l->inputq))
1305
			rc |= tipc_link_input(l, skb, l->inputq);
1306
		if (unlikely(++l->rcv_unacked >= TIPC_MIN_LINK_WIN))
1307
			rc |= tipc_link_build_state_msg(l, xmitq);
1308
		if (unlikely(rc & ~TIPC_LINK_SND_STATE))
1309
			break;
1310 1311 1312 1313 1314
	} while ((skb = __skb_dequeue(defq)));

	return rc;
drop:
	kfree_skb(skb);
1315 1316 1317
	return rc;
}

1318
static void tipc_link_build_proto_msg(struct tipc_link *l, int mtyp, bool probe,
1319 1320
				      bool probe_reply, u16 rcvgap,
				      int tolerance, int priority,
1321 1322
				      struct sk_buff_head *xmitq)
{
1323
	struct tipc_link *bcl = l->bc_rcvlink;
1324 1325 1326
	struct sk_buff *skb;
	struct tipc_msg *hdr;
	struct sk_buff_head *dfq = &l->deferdq;
1327
	bool node_up = link_is_up(bcl);
1328 1329 1330
	struct tipc_mon_state *mstate = &l->mon_state;
	int dlen = 0;
	void *data;
1331 1332

	/* Don't send protocol message during reset or link failover */
1333
	if (tipc_link_is_blocked(l))
1334 1335
		return;

1336 1337 1338 1339 1340 1341 1342
	if (!tipc_link_is_up(l) && (mtyp == STATE_MSG))
		return;

	if (!skb_queue_empty(dfq))
		rcvgap = buf_seqno(skb_peek(dfq)) - l->rcv_nxt;

	skb = tipc_msg_create(LINK_PROTOCOL, mtyp, INT_H_SIZE,
1343
			      tipc_max_domain_size, l->addr,
1344 1345 1346 1347 1348
			      tipc_own_addr(l->net), 0, 0, 0);
	if (!skb)
		return;

	hdr = buf_msg(skb);
1349
	data = msg_data(hdr);
1350 1351
	msg_set_session(hdr, l->session);
	msg_set_bearer_id(hdr, l->bearer_id);
1352
	msg_set_net_plane(hdr, l->net_plane);
1353 1354
	msg_set_next_sent(hdr, l->snd_nxt);
	msg_set_ack(hdr, l->rcv_nxt - 1);
1355
	msg_set_bcast_ack(hdr, bcl->rcv_nxt - 1);
1356
	msg_set_bc_ack_invalid(hdr, !node_up);
1357
	msg_set_last_bcast(hdr, l->bc_sndlink->snd_nxt - 1);
1358 1359 1360 1361
	msg_set_link_tolerance(hdr, tolerance);
	msg_set_linkprio(hdr, priority);
	msg_set_redundant_link(hdr, node_up);
	msg_set_seq_gap(hdr, 0);
1362
	msg_set_seqno(hdr, l->snd_nxt + U16_MAX / 2);
1363 1364

	if (mtyp == STATE_MSG) {
1365 1366
		if (l->peer_caps & TIPC_LINK_PROTO_SEQNO)
			msg_set_seqno(hdr, l->snd_nxt_state++);
1367
		msg_set_seq_gap(hdr, rcvgap);
1368
		msg_set_bc_gap(hdr, link_bc_rcv_gap(bcl));
1369
		msg_set_probe(hdr, probe);
1370
		msg_set_is_keepalive(hdr, probe || probe_reply);
1371 1372 1373
		tipc_mon_prep(l->net, data, &dlen, mstate, l->bearer_id);
		msg_set_size(hdr, INT_H_SIZE + dlen);
		skb_trim(skb, INT_H_SIZE + dlen);
1374
		l->stats.sent_states++;
1375
		l->rcv_unacked = 0;
1376 1377 1378
	} else {
		/* RESET_MSG or ACTIVATE_MSG */
		msg_set_max_pkt(hdr, l->advertised_mtu);
1379 1380 1381
		strcpy(data, l->if_name);
		msg_set_size(hdr, INT_H_SIZE + TIPC_MAX_IF_NAME);
		skb_trim(skb, INT_H_SIZE + TIPC_MAX_IF_NAME);
1382
	}
1383 1384 1385 1386
	if (probe)
		l->stats.sent_probes++;
	if (rcvgap)
		l->stats.sent_nacks++;
1387
	skb->priority = TC_PRIO_CONTROL;
1388
	__skb_queue_tail(xmitq, skb);
P
Per Liden 已提交
1389 1390
}

1391
/* tipc_link_tnl_prepare(): prepare and return a list of tunnel packets
1392
 * with contents of the link's transmit and backlog queues.
P
Per Liden 已提交
1393
 */
1394 1395
void tipc_link_tnl_prepare(struct tipc_link *l, struct tipc_link *tnl,
			   int mtyp, struct sk_buff_head *xmitq)
P
Per Liden 已提交
1396
{
1397 1398 1399 1400 1401
	struct sk_buff *skb, *tnlskb;
	struct tipc_msg *hdr, tnlhdr;
	struct sk_buff_head *queue = &l->transmq;
	struct sk_buff_head tmpxq, tnlq;
	u16 pktlen, pktcnt, seqno = l->snd_nxt;
P
Per Liden 已提交
1402

1403
	if (!tnl)
P
Per Liden 已提交
1404 1405
		return;

1406 1407
	skb_queue_head_init(&tnlq);
	skb_queue_head_init(&tmpxq);
1408

1409 1410
	/* At least one packet required for safe algorithm => add dummy */
	skb = tipc_msg_create(TIPC_LOW_IMPORTANCE, TIPC_DIRECT_MSG,
1411
			      BASIC_H_SIZE, 0, l->addr, tipc_own_addr(l->net),
1412 1413 1414
			      0, 0, TIPC_ERR_NO_PORT);
	if (!skb) {
		pr_warn("%sunable to create tunnel packet\n", link_co_err);
P
Per Liden 已提交
1415 1416
		return;
	}
1417 1418 1419 1420 1421
	skb_queue_tail(&tnlq, skb);
	tipc_link_xmit(l, &tnlq, &tmpxq);
	__skb_queue_purge(&tmpxq);

	/* Initialize reusable tunnel packet header */
1422
	tipc_msg_init(tipc_own_addr(l->net), &tnlhdr, TUNNEL_PROTOCOL,
1423 1424 1425 1426 1427 1428
		      mtyp, INT_H_SIZE, l->addr);
	pktcnt = skb_queue_len(&l->transmq) + skb_queue_len(&l->backlogq);
	msg_set_msgcnt(&tnlhdr, pktcnt);
	msg_set_bearer_id(&tnlhdr, l->peer_bearer_id);
tnl:
	/* Wrap each packet into a tunnel packet */
J
Jon Paul Maloy 已提交
1429
	skb_queue_walk(queue, skb) {
1430 1431 1432 1433 1434
		hdr = buf_msg(skb);
		if (queue == &l->backlogq)
			msg_set_seqno(hdr, seqno++);
		pktlen = msg_size(hdr);
		msg_set_size(&tnlhdr, pktlen + INT_H_SIZE);
1435
		tnlskb = tipc_buf_acquire(pktlen + INT_H_SIZE, GFP_ATOMIC);
1436 1437
		if (!tnlskb) {
			pr_warn("%sunable to send packet\n", link_co_err);
P
Per Liden 已提交
1438 1439
			return;
		}
1440 1441 1442
		skb_copy_to_linear_data(tnlskb, &tnlhdr, INT_H_SIZE);
		skb_copy_to_linear_data_offset(tnlskb, INT_H_SIZE, hdr, pktlen);
		__skb_queue_tail(&tnlq, tnlskb);
P
Per Liden 已提交
1443
	}
1444 1445 1446
	if (queue != &l->backlogq) {
		queue = &l->backlogq;
		goto tnl;
1447
	}
1448

1449
	tipc_link_xmit(tnl, &tnlq, xmitq);
P
Per Liden 已提交
1450

1451 1452 1453 1454
	if (mtyp == FAILOVER_MSG) {
		tnl->drop_point = l->rcv_nxt;
		tnl->failover_reasm_skb = l->reasm_buf;
		l->reasm_buf = NULL;
1455
	}
P
Per Liden 已提交
1456 1457
}

1458 1459 1460 1461 1462 1463 1464 1465 1466 1467
/* tipc_link_proto_rcv(): receive link level protocol message :
 * Note that network plane id propagates through the network, and may
 * change at any time. The node with lowest numerical id determines
 * network plane
 */
static int tipc_link_proto_rcv(struct tipc_link *l, struct sk_buff *skb,
			       struct sk_buff_head *xmitq)
{
	struct tipc_msg *hdr = buf_msg(skb);
	u16 rcvgap = 0;
1468 1469
	u16 ack = msg_ack(hdr);
	u16 gap = msg_seq_gap(hdr);
1470 1471 1472
	u16 peers_snd_nxt =  msg_next_sent(hdr);
	u16 peers_tol = msg_link_tolerance(hdr);
	u16 peers_prio = msg_linkprio(hdr);
1473
	u16 rcv_nxt = l->rcv_nxt;
1474
	u16 dlen = msg_data_sz(hdr);
1475
	int mtyp = msg_type(hdr);
1476
	bool reply = msg_probe(hdr);
1477
	void *data;
1478 1479 1480
	char *if_name;
	int rc = 0;

1481
	if (tipc_link_is_blocked(l) || !xmitq)
1482 1483
		goto exit;

1484
	if (tipc_own_addr(l->net) > msg_prevnode(hdr))
1485 1486
		l->net_plane = msg_net_plane(hdr);

1487 1488 1489 1490
	skb_linearize(skb);
	hdr = buf_msg(skb);
	data = msg_data(hdr);

1491
	switch (mtyp) {
1492 1493 1494 1495
	case RESET_MSG:

		/* Ignore duplicate RESET with old session number */
		if ((less_eq(msg_session(hdr), l->peer_session)) &&
1496
		    (l->peer_session != ANY_SESSION))
1497 1498
			break;
		/* fall thru' */
1499

1500 1501 1502 1503 1504 1505 1506 1507
	case ACTIVATE_MSG:

		/* Complete own link name with peer's interface name */
		if_name =  strrchr(l->name, ':') + 1;
		if (sizeof(l->name) - (if_name - l->name) <= TIPC_MAX_IF_NAME)
			break;
		if (msg_data_sz(hdr) < TIPC_MAX_IF_NAME)
			break;
1508
		strncpy(if_name, data, TIPC_MAX_IF_NAME);
1509 1510 1511 1512 1513 1514 1515 1516 1517

		/* Update own tolerance if peer indicates a non-zero value */
		if (in_range(peers_tol, TIPC_MIN_LINK_TOL, TIPC_MAX_LINK_TOL))
			l->tolerance = peers_tol;

		/* Update own priority if peer's priority is higher */
		if (in_range(peers_prio, l->priority + 1, TIPC_MAX_LINK_PRI))
			l->priority = peers_prio;

1518
		/* ACTIVATE_MSG serves as PEER_RESET if link is already down */
1519 1520 1521
		if (msg_peer_stopping(hdr))
			rc = tipc_link_fsm_evt(l, LINK_FAILURE_EVT);
		else if ((mtyp == RESET_MSG) || !link_is_up(l))
1522 1523 1524 1525 1526 1527
			rc = tipc_link_fsm_evt(l, LINK_PEER_RESET_EVT);

		/* ACTIVATE_MSG takes up link if it was already locally reset */
		if ((mtyp == ACTIVATE_MSG) && (l->state == LINK_ESTABLISHING))
			rc = TIPC_LINK_UP_EVT;

1528 1529 1530 1531 1532
		l->peer_session = msg_session(hdr);
		l->peer_bearer_id = msg_bearer_id(hdr);
		if (l->mtu > msg_max_pkt(hdr))
			l->mtu = msg_max_pkt(hdr);
		break;
1533

1534
	case STATE_MSG:
1535

1536 1537 1538 1539 1540
		if (l->peer_caps & TIPC_LINK_PROTO_SEQNO &&
		    less(msg_seqno(hdr), l->rcv_nxt_state))
			break;
		l->rcv_nxt_state = msg_seqno(hdr) + 1;

1541 1542 1543 1544
		/* Update own tolerance if peer indicates a non-zero value */
		if (in_range(peers_tol, TIPC_MIN_LINK_TOL, TIPC_MAX_LINK_TOL))
			l->tolerance = peers_tol;

1545 1546 1547
		/* Update own prio if peer indicates a different value */
		if ((peers_prio != l->priority) &&
		    in_range(peers_prio, 1, TIPC_MAX_LINK_PRI)) {
1548 1549 1550 1551
			l->priority = peers_prio;
			rc = tipc_link_fsm_evt(l, LINK_FAILURE_EVT);
		}

1552 1553 1554 1555
		l->silent_intv_cnt = 0;
		l->stats.recv_states++;
		if (msg_probe(hdr))
			l->stats.recv_probes++;
1556 1557 1558 1559

		if (!link_is_up(l)) {
			if (l->state == LINK_ESTABLISHING)
				rc = TIPC_LINK_UP_EVT;
1560
			break;
1561
		}
1562 1563
		tipc_mon_rcv(l->net, data, dlen, l->addr,
			     &l->mon_state, l->bearer_id);
1564

1565
		/* Send NACK if peer has sent pkts we haven't received yet */
1566
		if (more(peers_snd_nxt, rcv_nxt) && !tipc_link_is_synching(l))
1567
			rcvgap = peers_snd_nxt - l->rcv_nxt;
1568 1569 1570
		if (rcvgap || reply)
			tipc_link_build_proto_msg(l, STATE_MSG, 0, reply,
						  rcvgap, 0, 0, xmitq);
1571
		tipc_link_release_pkts(l, ack);
1572 1573

		/* If NACK, retransmit will now start at right position */
1574
		if (gap) {
1575
			rc = tipc_link_retrans(l, l, ack + 1, ack + gap, xmitq);
1576 1577
			l->stats.recv_nacks++;
		}
1578

1579 1580 1581 1582 1583 1584 1585 1586 1587
		tipc_link_advance_backlog(l, xmitq);
		if (unlikely(!skb_queue_empty(&l->wakeupq)))
			link_prepare_wakeup(l);
	}
exit:
	kfree_skb(skb);
	return rc;
}

1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600
/* tipc_link_build_bc_proto_msg() - create broadcast protocol message
 */
static bool tipc_link_build_bc_proto_msg(struct tipc_link *l, bool bcast,
					 u16 peers_snd_nxt,
					 struct sk_buff_head *xmitq)
{
	struct sk_buff *skb;
	struct tipc_msg *hdr;
	struct sk_buff *dfrd_skb = skb_peek(&l->deferdq);
	u16 ack = l->rcv_nxt - 1;
	u16 gap_to = peers_snd_nxt - 1;

	skb = tipc_msg_create(BCAST_PROTOCOL, STATE_MSG, INT_H_SIZE,
1601
			      0, l->addr, tipc_own_addr(l->net), 0, 0, 0);
1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620
	if (!skb)
		return false;
	hdr = buf_msg(skb);
	msg_set_last_bcast(hdr, l->bc_sndlink->snd_nxt - 1);
	msg_set_bcast_ack(hdr, ack);
	msg_set_bcgap_after(hdr, ack);
	if (dfrd_skb)
		gap_to = buf_seqno(dfrd_skb) - 1;
	msg_set_bcgap_to(hdr, gap_to);
	msg_set_non_seq(hdr, bcast);
	__skb_queue_tail(xmitq, skb);
	return true;
}

/* tipc_link_build_bc_init_msg() - synchronize broadcast link endpoints.
 *
 * Give a newly added peer node the sequence number where it should
 * start receiving and acking broadcast packets.
 */
1621 1622
static void tipc_link_build_bc_init_msg(struct tipc_link *l,
					struct sk_buff_head *xmitq)
1623 1624 1625 1626 1627 1628
{
	struct sk_buff_head list;

	__skb_queue_head_init(&list);
	if (!tipc_link_build_bc_proto_msg(l->bc_rcvlink, false, 0, &list))
		return;
1629
	msg_set_bc_ack_invalid(buf_msg(skb_peek(&list)), true);
1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659
	tipc_link_xmit(l, &list, xmitq);
}

/* tipc_link_bc_init_rcv - receive initial broadcast synch data from peer
 */
void tipc_link_bc_init_rcv(struct tipc_link *l, struct tipc_msg *hdr)
{
	int mtyp = msg_type(hdr);
	u16 peers_snd_nxt = msg_bc_snd_nxt(hdr);

	if (link_is_up(l))
		return;

	if (msg_user(hdr) == BCAST_PROTOCOL) {
		l->rcv_nxt = peers_snd_nxt;
		l->state = LINK_ESTABLISHED;
		return;
	}

	if (l->peer_caps & TIPC_BCAST_SYNCH)
		return;

	if (msg_peer_node_is_up(hdr))
		return;

	/* Compatibility: accept older, less safe initial synch data */
	if ((mtyp == RESET_MSG) || (mtyp == ACTIVATE_MSG))
		l->rcv_nxt = peers_snd_nxt;
}

1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695
/* link_bc_retr eval()- check if the indicated range can be retransmitted now
 * - Adjust permitted range if there is overlap with previous retransmission
 */
static bool link_bc_retr_eval(struct tipc_link *l, u16 *from, u16 *to)
{
	unsigned long elapsed = jiffies_to_msecs(jiffies - l->prev_retr);

	if (less(*to, *from))
		return false;

	/* New retransmission request */
	if ((elapsed > TIPC_BC_RETR_LIMIT) ||
	    less(*to, l->prev_from) || more(*from, l->prev_to)) {
		l->prev_from = *from;
		l->prev_to = *to;
		l->prev_retr = jiffies;
		return true;
	}

	/* Inside range of previous retransmit */
	if (!less(*from, l->prev_from) && !more(*to, l->prev_to))
		return false;

	/* Fully or partially outside previous range => exclude overlap */
	if (less(*from, l->prev_from)) {
		*to = l->prev_from - 1;
		l->prev_from = *from;
	}
	if (more(*to, l->prev_to)) {
		*from = l->prev_to + 1;
		l->prev_to = *to;
	}
	l->prev_retr = jiffies;
	return true;
}

1696 1697
/* tipc_link_bc_sync_rcv - update rcv link according to peer's send state
 */
1698 1699
int tipc_link_bc_sync_rcv(struct tipc_link *l, struct tipc_msg *hdr,
			  struct sk_buff_head *xmitq)
1700
{
1701
	struct tipc_link *snd_l = l->bc_sndlink;
1702
	u16 peers_snd_nxt = msg_bc_snd_nxt(hdr);
1703 1704 1705
	u16 from = msg_bcast_ack(hdr) + 1;
	u16 to = from + msg_bc_gap(hdr) - 1;
	int rc = 0;
1706 1707

	if (!link_is_up(l))
1708
		return rc;
1709 1710

	if (!msg_peer_node_is_up(hdr))
1711
		return rc;
1712

1713 1714 1715 1716 1717
	/* Open when peer ackowledges our bcast init msg (pkt #1) */
	if (msg_ack(hdr))
		l->bc_peer_is_up = true;

	if (!l->bc_peer_is_up)
1718
		return rc;
1719

1720 1721
	l->stats.recv_nacks++;

1722 1723
	/* Ignore if peers_snd_nxt goes beyond receive window */
	if (more(peers_snd_nxt, l->rcv_nxt + l->window))
1724 1725
		return rc;

1726
	if (link_bc_retr_eval(snd_l, &from, &to))
1727
		rc = tipc_link_retrans(snd_l, l, from, to, xmitq);
1728 1729 1730 1731 1732 1733 1734 1735 1736 1737

	l->snd_nxt = peers_snd_nxt;
	if (link_bc_rcv_gap(l))
		rc |= TIPC_LINK_SND_STATE;

	/* Return now if sender supports nack via STATE messages */
	if (l->peer_caps & TIPC_BCAST_STATE_NACK)
		return rc;

	/* Otherwise, be backwards compatible */
1738 1739 1740

	if (!more(peers_snd_nxt, l->rcv_nxt)) {
		l->nack_state = BC_NACK_SND_CONDITIONAL;
1741
		return 0;
1742 1743 1744 1745 1746
	}

	/* Don't NACK if one was recently sent or peeked */
	if (l->nack_state == BC_NACK_SND_SUPPRESS) {
		l->nack_state = BC_NACK_SND_UNCONDITIONAL;
1747
		return 0;
1748 1749 1750 1751 1752 1753
	}

	/* Conditionally delay NACK sending until next synch rcv */
	if (l->nack_state == BC_NACK_SND_CONDITIONAL) {
		l->nack_state = BC_NACK_SND_UNCONDITIONAL;
		if ((peers_snd_nxt - l->rcv_nxt) < TIPC_MIN_LINK_WIN)
1754
			return 0;
1755 1756 1757 1758 1759
	}

	/* Send NACK now but suppress next one */
	tipc_link_build_bc_proto_msg(l, true, peers_snd_nxt, xmitq);
	l->nack_state = BC_NACK_SND_SUPPRESS;
1760
	return 0;
1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796
}

void tipc_link_bc_ack_rcv(struct tipc_link *l, u16 acked,
			  struct sk_buff_head *xmitq)
{
	struct sk_buff *skb, *tmp;
	struct tipc_link *snd_l = l->bc_sndlink;

	if (!link_is_up(l) || !l->bc_peer_is_up)
		return;

	if (!more(acked, l->acked))
		return;

	/* Skip over packets peer has already acked */
	skb_queue_walk(&snd_l->transmq, skb) {
		if (more(buf_seqno(skb), l->acked))
			break;
	}

	/* Update/release the packets peer is acking now */
	skb_queue_walk_from_safe(&snd_l->transmq, skb, tmp) {
		if (more(buf_seqno(skb), acked))
			break;
		if (!--TIPC_SKB_CB(skb)->ackers) {
			__skb_unlink(skb, &snd_l->transmq);
			kfree_skb(skb);
		}
	}
	l->acked = acked;
	tipc_link_advance_backlog(snd_l, xmitq);
	if (unlikely(!skb_queue_empty(&snd_l->wakeupq)))
		link_prepare_wakeup(snd_l);
}

/* tipc_link_bc_nack_rcv(): receive broadcast nack message
1797 1798
 * This function is here for backwards compatibility, since
 * no BCAST_PROTOCOL/STATE messages occur from TIPC v2.5.
1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819
 */
int tipc_link_bc_nack_rcv(struct tipc_link *l, struct sk_buff *skb,
			  struct sk_buff_head *xmitq)
{
	struct tipc_msg *hdr = buf_msg(skb);
	u32 dnode = msg_destnode(hdr);
	int mtyp = msg_type(hdr);
	u16 acked = msg_bcast_ack(hdr);
	u16 from = acked + 1;
	u16 to = msg_bcgap_to(hdr);
	u16 peers_snd_nxt = to + 1;
	int rc = 0;

	kfree_skb(skb);

	if (!tipc_link_is_up(l) || !l->bc_peer_is_up)
		return 0;

	if (mtyp != STATE_MSG)
		return 0;

1820
	if (dnode == tipc_own_addr(l->net)) {
1821
		tipc_link_bc_ack_rcv(l, acked, xmitq);
1822
		rc = tipc_link_retrans(l->bc_sndlink, l, from, to, xmitq);
1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833
		l->stats.recv_nacks++;
		return rc;
	}

	/* Msg for other node => suppress own NACK at next sync if applicable */
	if (more(peers_snd_nxt, l->rcv_nxt) && !less(l->rcv_nxt, from))
		l->nack_state = BC_NACK_SND_SUPPRESS;

	return 0;
}

1834
void tipc_link_set_queue_limits(struct tipc_link *l, u32 win)
P
Per Liden 已提交
1835
{
1836
	int max_bulk = TIPC_MAX_PUBL / (l->mtu / ITEM_SIZE);
1837 1838

	l->window = win;
1839 1840 1841 1842
	l->backlog[TIPC_LOW_IMPORTANCE].limit      = max_t(u16, 50, win);
	l->backlog[TIPC_MEDIUM_IMPORTANCE].limit   = max_t(u16, 100, win * 2);
	l->backlog[TIPC_HIGH_IMPORTANCE].limit     = max_t(u16, 150, win * 3);
	l->backlog[TIPC_CRITICAL_IMPORTANCE].limit = max_t(u16, 200, win * 4);
1843
	l->backlog[TIPC_SYSTEM_IMPORTANCE].limit   = max_bulk;
P
Per Liden 已提交
1844 1845 1846
}

/**
1847
 * link_reset_stats - reset link statistics
1848
 * @l: pointer to link
P
Per Liden 已提交
1849
 */
1850
void tipc_link_reset_stats(struct tipc_link *l)
P
Per Liden 已提交
1851
{
1852
	memset(&l->stats, 0, sizeof(l->stats));
P
Per Liden 已提交
1853 1854
}

1855
static void link_print(struct tipc_link *l, const char *str)
P
Per Liden 已提交
1856
{
1857
	struct sk_buff *hskb = skb_peek(&l->transmq);
1858
	u16 head = hskb ? msg_seqno(buf_msg(hskb)) : l->snd_nxt - 1;
1859
	u16 tail = l->snd_nxt - 1;
1860

1861
	pr_info("%s Link <%s> state %x\n", str, l->name, l->state);
1862 1863 1864
	pr_info("XMTQ: %u [%u-%u], BKLGQ: %u, SNDNX: %u, RCVNX: %u\n",
		skb_queue_len(&l->transmq), head, tail,
		skb_queue_len(&l->backlogq), l->snd_nxt, l->rcv_nxt);
P
Per Liden 已提交
1865
}
1866 1867 1868 1869 1870 1871 1872 1873

/* Parse and validate nested (link) properties valid for media, bearer and link
 */
int tipc_nl_parse_link_prop(struct nlattr *prop, struct nlattr *props[])
{
	int err;

	err = nla_parse_nested(props, TIPC_NLA_PROP_MAX, prop,
1874
			       tipc_nl_prop_policy, NULL);
1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903
	if (err)
		return err;

	if (props[TIPC_NLA_PROP_PRIO]) {
		u32 prio;

		prio = nla_get_u32(props[TIPC_NLA_PROP_PRIO]);
		if (prio > TIPC_MAX_LINK_PRI)
			return -EINVAL;
	}

	if (props[TIPC_NLA_PROP_TOL]) {
		u32 tol;

		tol = nla_get_u32(props[TIPC_NLA_PROP_TOL]);
		if ((tol < TIPC_MIN_LINK_TOL) || (tol > TIPC_MAX_LINK_TOL))
			return -EINVAL;
	}

	if (props[TIPC_NLA_PROP_WIN]) {
		u32 win;

		win = nla_get_u32(props[TIPC_NLA_PROP_WIN]);
		if ((win < TIPC_MIN_LINK_WIN) || (win > TIPC_MAX_LINK_WIN))
			return -EINVAL;
	}

	return 0;
}
1904

1905
static int __tipc_nl_add_stats(struct sk_buff *skb, struct tipc_stats *s)
1906 1907 1908 1909 1910 1911 1912 1913 1914 1915
{
	int i;
	struct nlattr *stats;

	struct nla_map {
		u32 key;
		u32 val;
	};

	struct nla_map map[] = {
1916
		{TIPC_NLA_STATS_RX_INFO, 0},
1917 1918 1919 1920
		{TIPC_NLA_STATS_RX_FRAGMENTS, s->recv_fragments},
		{TIPC_NLA_STATS_RX_FRAGMENTED, s->recv_fragmented},
		{TIPC_NLA_STATS_RX_BUNDLES, s->recv_bundles},
		{TIPC_NLA_STATS_RX_BUNDLED, s->recv_bundled},
1921
		{TIPC_NLA_STATS_TX_INFO, 0},
1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970
		{TIPC_NLA_STATS_TX_FRAGMENTS, s->sent_fragments},
		{TIPC_NLA_STATS_TX_FRAGMENTED, s->sent_fragmented},
		{TIPC_NLA_STATS_TX_BUNDLES, s->sent_bundles},
		{TIPC_NLA_STATS_TX_BUNDLED, s->sent_bundled},
		{TIPC_NLA_STATS_MSG_PROF_TOT, (s->msg_length_counts) ?
			s->msg_length_counts : 1},
		{TIPC_NLA_STATS_MSG_LEN_CNT, s->msg_length_counts},
		{TIPC_NLA_STATS_MSG_LEN_TOT, s->msg_lengths_total},
		{TIPC_NLA_STATS_MSG_LEN_P0, s->msg_length_profile[0]},
		{TIPC_NLA_STATS_MSG_LEN_P1, s->msg_length_profile[1]},
		{TIPC_NLA_STATS_MSG_LEN_P2, s->msg_length_profile[2]},
		{TIPC_NLA_STATS_MSG_LEN_P3, s->msg_length_profile[3]},
		{TIPC_NLA_STATS_MSG_LEN_P4, s->msg_length_profile[4]},
		{TIPC_NLA_STATS_MSG_LEN_P5, s->msg_length_profile[5]},
		{TIPC_NLA_STATS_MSG_LEN_P6, s->msg_length_profile[6]},
		{TIPC_NLA_STATS_RX_STATES, s->recv_states},
		{TIPC_NLA_STATS_RX_PROBES, s->recv_probes},
		{TIPC_NLA_STATS_RX_NACKS, s->recv_nacks},
		{TIPC_NLA_STATS_RX_DEFERRED, s->deferred_recv},
		{TIPC_NLA_STATS_TX_STATES, s->sent_states},
		{TIPC_NLA_STATS_TX_PROBES, s->sent_probes},
		{TIPC_NLA_STATS_TX_NACKS, s->sent_nacks},
		{TIPC_NLA_STATS_TX_ACKS, s->sent_acks},
		{TIPC_NLA_STATS_RETRANSMITTED, s->retransmitted},
		{TIPC_NLA_STATS_DUPLICATES, s->duplicates},
		{TIPC_NLA_STATS_LINK_CONGS, s->link_congs},
		{TIPC_NLA_STATS_MAX_QUEUE, s->max_queue_sz},
		{TIPC_NLA_STATS_AVG_QUEUE, s->queue_sz_counts ?
			(s->accu_queue_sz / s->queue_sz_counts) : 0}
	};

	stats = nla_nest_start(skb, TIPC_NLA_LINK_STATS);
	if (!stats)
		return -EMSGSIZE;

	for (i = 0; i <  ARRAY_SIZE(map); i++)
		if (nla_put_u32(skb, map[i].key, map[i].val))
			goto msg_full;

	nla_nest_end(skb, stats);

	return 0;
msg_full:
	nla_nest_cancel(skb, stats);

	return -EMSGSIZE;
}

/* Caller should hold appropriate locks to protect the link */
1971 1972
int __tipc_nl_add_link(struct net *net, struct tipc_nl_msg *msg,
		       struct tipc_link *link, int nlflags)
1973
{
1974
	u32 self = tipc_own_addr(net);
1975 1976
	struct nlattr *attrs;
	struct nlattr *prop;
1977 1978
	void *hdr;
	int err;
1979

1980
	hdr = genlmsg_put(msg->skb, msg->portid, msg->seq, &tipc_genl_family,
1981
			  nlflags, TIPC_NL_LINK_GET);
1982 1983 1984 1985 1986 1987 1988 1989 1990
	if (!hdr)
		return -EMSGSIZE;

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

	if (nla_put_string(msg->skb, TIPC_NLA_LINK_NAME, link->name))
		goto attr_msg_full;
1991
	if (nla_put_u32(msg->skb, TIPC_NLA_LINK_DEST, tipc_cluster_mask(self)))
1992
		goto attr_msg_full;
1993
	if (nla_put_u32(msg->skb, TIPC_NLA_LINK_MTU, link->mtu))
1994
		goto attr_msg_full;
1995
	if (nla_put_u32(msg->skb, TIPC_NLA_LINK_RX, link->stats.recv_pkts))
1996
		goto attr_msg_full;
1997
	if (nla_put_u32(msg->skb, TIPC_NLA_LINK_TX, link->stats.sent_pkts))
1998 1999 2000 2001 2002
		goto attr_msg_full;

	if (tipc_link_is_up(link))
		if (nla_put_flag(msg->skb, TIPC_NLA_LINK_UP))
			goto attr_msg_full;
2003
	if (link->active)
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
		if (nla_put_flag(msg->skb, TIPC_NLA_LINK_ACTIVE))
			goto attr_msg_full;

	prop = nla_nest_start(msg->skb, TIPC_NLA_LINK_PROP);
	if (!prop)
		goto attr_msg_full;
	if (nla_put_u32(msg->skb, TIPC_NLA_PROP_PRIO, link->priority))
		goto prop_msg_full;
	if (nla_put_u32(msg->skb, TIPC_NLA_PROP_TOL, link->tolerance))
		goto prop_msg_full;
	if (nla_put_u32(msg->skb, TIPC_NLA_PROP_WIN,
2015
			link->window))
2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038
		goto prop_msg_full;
	if (nla_put_u32(msg->skb, TIPC_NLA_PROP_PRIO, link->priority))
		goto prop_msg_full;
	nla_nest_end(msg->skb, prop);

	err = __tipc_nl_add_stats(msg->skb, &link->stats);
	if (err)
		goto attr_msg_full;

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

	return 0;

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

	return -EMSGSIZE;
}
2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051

static int __tipc_nl_add_bc_link_stat(struct sk_buff *skb,
				      struct tipc_stats *stats)
{
	int i;
	struct nlattr *nest;

	struct nla_map {
		__u32 key;
		__u32 val;
	};

	struct nla_map map[] = {
2052
		{TIPC_NLA_STATS_RX_INFO, stats->recv_pkts},
2053 2054 2055 2056
		{TIPC_NLA_STATS_RX_FRAGMENTS, stats->recv_fragments},
		{TIPC_NLA_STATS_RX_FRAGMENTED, stats->recv_fragmented},
		{TIPC_NLA_STATS_RX_BUNDLES, stats->recv_bundles},
		{TIPC_NLA_STATS_RX_BUNDLED, stats->recv_bundled},
2057
		{TIPC_NLA_STATS_TX_INFO, stats->sent_pkts},
2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106
		{TIPC_NLA_STATS_TX_FRAGMENTS, stats->sent_fragments},
		{TIPC_NLA_STATS_TX_FRAGMENTED, stats->sent_fragmented},
		{TIPC_NLA_STATS_TX_BUNDLES, stats->sent_bundles},
		{TIPC_NLA_STATS_TX_BUNDLED, stats->sent_bundled},
		{TIPC_NLA_STATS_RX_NACKS, stats->recv_nacks},
		{TIPC_NLA_STATS_RX_DEFERRED, stats->deferred_recv},
		{TIPC_NLA_STATS_TX_NACKS, stats->sent_nacks},
		{TIPC_NLA_STATS_TX_ACKS, stats->sent_acks},
		{TIPC_NLA_STATS_RETRANSMITTED, stats->retransmitted},
		{TIPC_NLA_STATS_DUPLICATES, stats->duplicates},
		{TIPC_NLA_STATS_LINK_CONGS, stats->link_congs},
		{TIPC_NLA_STATS_MAX_QUEUE, stats->max_queue_sz},
		{TIPC_NLA_STATS_AVG_QUEUE, stats->queue_sz_counts ?
			(stats->accu_queue_sz / stats->queue_sz_counts) : 0}
	};

	nest = nla_nest_start(skb, TIPC_NLA_LINK_STATS);
	if (!nest)
		return -EMSGSIZE;

	for (i = 0; i <  ARRAY_SIZE(map); i++)
		if (nla_put_u32(skb, map[i].key, map[i].val))
			goto msg_full;

	nla_nest_end(skb, nest);

	return 0;
msg_full:
	nla_nest_cancel(skb, nest);

	return -EMSGSIZE;
}

int tipc_nl_add_bc_link(struct net *net, struct tipc_nl_msg *msg)
{
	int err;
	void *hdr;
	struct nlattr *attrs;
	struct nlattr *prop;
	struct tipc_net *tn = net_generic(net, tipc_net_id);
	struct tipc_link *bcl = tn->bcl;

	if (!bcl)
		return 0;

	tipc_bcast_lock(net);

	hdr = genlmsg_put(msg->skb, msg->portid, msg->seq, &tipc_genl_family,
			  NLM_F_MULTI, TIPC_NL_LINK_GET);
I
Insu Yun 已提交
2107 2108
	if (!hdr) {
		tipc_bcast_unlock(net);
2109
		return -EMSGSIZE;
I
Insu Yun 已提交
2110
	}
2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123

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

	/* The broadcast link is always up */
	if (nla_put_flag(msg->skb, TIPC_NLA_LINK_UP))
		goto attr_msg_full;

	if (nla_put_flag(msg->skb, TIPC_NLA_LINK_BROADCAST))
		goto attr_msg_full;
	if (nla_put_string(msg->skb, TIPC_NLA_LINK_NAME, bcl->name))
		goto attr_msg_full;
2124
	if (nla_put_u32(msg->skb, TIPC_NLA_LINK_RX, 0))
2125
		goto attr_msg_full;
2126
	if (nla_put_u32(msg->skb, TIPC_NLA_LINK_TX, 0))
2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156
		goto attr_msg_full;

	prop = nla_nest_start(msg->skb, TIPC_NLA_LINK_PROP);
	if (!prop)
		goto attr_msg_full;
	if (nla_put_u32(msg->skb, TIPC_NLA_PROP_WIN, bcl->window))
		goto prop_msg_full;
	nla_nest_end(msg->skb, prop);

	err = __tipc_nl_add_bc_link_stat(msg->skb, &bcl->stats);
	if (err)
		goto attr_msg_full;

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

	return 0;

prop_msg_full:
	nla_nest_cancel(msg->skb, prop);
attr_msg_full:
	nla_nest_cancel(msg->skb, attrs);
msg_full:
	tipc_bcast_unlock(net);
	genlmsg_cancel(msg->skb, hdr);

	return -EMSGSIZE;
}

2157 2158
void tipc_link_set_tolerance(struct tipc_link *l, u32 tol,
			     struct sk_buff_head *xmitq)
2159 2160
{
	l->tolerance = tol;
2161 2162
	if (link_is_up(l))
		tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, 0, tol, 0, xmitq);
2163 2164
}

2165 2166
void tipc_link_set_prio(struct tipc_link *l, u32 prio,
			struct sk_buff_head *xmitq)
2167 2168
{
	l->priority = prio;
2169
	tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, 0, 0, prio, xmitq);
2170 2171 2172 2173 2174 2175
}

void tipc_link_set_abort_limit(struct tipc_link *l, u32 limit)
{
	l->abort_limit = limit;
}