link.c 68.9 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"
T
Tuong Lien 已提交
46
#include "trace.h"
P
Per Liden 已提交
47

48 49
#include <linux/pkt_sched.h>

50
struct tipc_stats {
51 52
	u32 sent_pkts;
	u32 recv_pkts;
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
	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')
100
 * @mon_state: cookie with information needed by link monitor
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
109
 * @prev_from: sequence number of most previous retransmission request
110 111
 * @stale_cnt: counter for number of identical retransmit attempts
 * @stale_limit: time when repeated identical retransmits must force link reset
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 */
132 133
	u16 peer_session;
	u16 session;
134 135
	u16 snd_nxt_state;
	u16 rcv_nxt_state;
136 137 138 139 140 141
	u32 peer_bearer_id;
	u32 bearer_id;
	u32 tolerance;
	u32 abort_limit;
	u32 state;
	u16 peer_caps;
142
	bool in_session;
143 144
	bool active;
	u32 silent_intv_cnt;
145
	char if_name[TIPC_MAX_IF_NAME];
146 147
	u32 priority;
	char net_plane;
148
	struct tipc_mon_state mon_state;
149
	u16 rst_cnt;
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;
167
	u16 prev_from;
168
	u16 window;
169 170
	u16 stale_cnt;
	unsigned long stale_limit;
171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189

	/* 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;
190
	u8 nack_state;
191 192 193 194 195 196
	bool bc_peer_is_up;

	/* Statistics */
	struct tipc_stats stats;
};

197 198 199
/*
 * Error message prefixes
 */
200
static const char *link_co_err = "Link tunneling error, ";
201
static const char *link_rst_msg = "Resetting link ";
202

203 204 205 206 207 208 209 210
/* Send states for broadcast NACKs
 */
enum {
	BC_NACK_SND_CONDITIONAL,
	BC_NACK_SND_UNCONDITIONAL,
	BC_NACK_SND_SUPPRESS,
};

211
#define TIPC_BC_RETR_LIM msecs_to_jiffies(10)   /* [ms] */
212
#define TIPC_UC_RETR_TIME (jiffies + msecs_to_jiffies(1))
213

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

219
/* Link FSM states:
P
Per Liden 已提交
220
 */
221
enum {
222 223 224 225 226 227 228
	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
229 230 231 232
};

/* Link FSM state checking routines
 */
233
static int link_is_up(struct tipc_link *l)
234
{
235
	return l->state & (LINK_ESTABLISHED | LINK_SYNCHING);
236
}
P
Per Liden 已提交
237

238 239
static int tipc_link_proto_rcv(struct tipc_link *l, struct sk_buff *skb,
			       struct sk_buff_head *xmitq);
240
static void tipc_link_build_proto_msg(struct tipc_link *l, int mtyp, bool probe,
241 242
				      bool probe_reply, u16 rcvgap,
				      int tolerance, int priority,
243
				      struct sk_buff_head *xmitq);
244
static void link_print(struct tipc_link *l, const char *str);
245 246
static int tipc_link_build_nack_msg(struct tipc_link *l,
				    struct sk_buff_head *xmitq);
247 248 249
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);
250 251 252 253
static u16 tipc_build_gap_ack_blks(struct tipc_link *l, void *data);
static void tipc_link_advance_transmq(struct tipc_link *l, u16 acked, u16 gap,
				      struct tipc_gap_ack_blks *ga,
				      struct sk_buff_head *xmitq);
254

P
Per Liden 已提交
255
/*
S
Sam Ravnborg 已提交
256
 *  Simple non-static link routines (i.e. referenced outside this file)
P
Per Liden 已提交
257
 */
258
bool tipc_link_is_up(struct tipc_link *l)
P
Per Liden 已提交
259
{
260 261 262
	return link_is_up(l);
}

263 264 265 266 267
bool tipc_link_peer_is_down(struct tipc_link *l)
{
	return l->state == LINK_PEER_RESET;
}

268 269 270 271 272
bool tipc_link_is_reset(struct tipc_link *l)
{
	return l->state & (LINK_RESET | LINK_FAILINGOVER | LINK_ESTABLISHING);
}

273 274 275 276 277
bool tipc_link_is_establishing(struct tipc_link *l)
{
	return l->state == LINK_ESTABLISHING;
}

278 279 280 281 282 283 284 285 286 287 288 289 290
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
291 292
}

293
static bool link_is_bc_sndlink(struct tipc_link *l)
294 295 296 297
{
	return !l->bc_sndlink;
}

298
static bool link_is_bc_rcvlink(struct tipc_link *l)
299 300 301 302
{
	return ((l->bc_rcvlink == l) && !link_is_bc_sndlink(l));
}

303 304 305
void tipc_link_set_active(struct tipc_link *l, bool active)
{
	l->active = active;
P
Per Liden 已提交
306 307
}

308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337
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;
}

338 339 340 341 342
void tipc_link_update_caps(struct tipc_link *l, u16 capabilities)
{
	l->peer_caps = capabilities;
}

343 344 345
void tipc_link_add_bc_peer(struct tipc_link *snd_l,
			   struct tipc_link *uc_l,
			   struct sk_buff_head *xmitq)
346
{
347 348 349 350
	struct tipc_link *rcv_l = uc_l->bc_rcvlink;

	snd_l->ackers++;
	rcv_l->acked = snd_l->snd_nxt - 1;
351
	snd_l->state = LINK_ESTABLISHED;
352
	tipc_link_build_bc_init_msg(uc_l, xmitq);
353 354
}

355 356 357
void tipc_link_remove_bc_peer(struct tipc_link *snd_l,
			      struct tipc_link *rcv_l,
			      struct sk_buff_head *xmitq)
358
{
359 360 361
	u16 ack = snd_l->snd_nxt - 1;

	snd_l->ackers--;
362 363
	rcv_l->bc_peer_is_up = true;
	rcv_l->state = LINK_ESTABLISHED;
364
	tipc_link_bc_ack_rcv(rcv_l, ack, xmitq);
T
Tuong Lien 已提交
365
	trace_tipc_link_reset(rcv_l, TIPC_DUMP_ALL, "bclink removed!");
366 367 368
	tipc_link_reset(rcv_l);
	rcv_l->state = LINK_RESET;
	if (!snd_l->ackers) {
T
Tuong Lien 已提交
369
		trace_tipc_link_reset(snd_l, TIPC_DUMP_ALL, "zero ackers!");
370
		tipc_link_reset(snd_l);
371
		snd_l->state = LINK_RESET;
372 373
		__skb_queue_purge(xmitq);
	}
374 375 376 377 378 379 380
}

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

Y
YueHaibing 已提交
381
static u16 link_bc_rcv_gap(struct tipc_link *l)
382 383 384 385 386 387 388 389 390 391 392
{
	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;
}

393 394 395 396 397 398 399 400 401 402
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;
}

403 404 405 406 407 408 409 410 411 412 413 414 415 416 417
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;
}

L
LUU Duc Canh 已提交
418 419 420 421 422
u32 tipc_link_state(struct tipc_link *l)
{
	return l->state;
}

P
Per Liden 已提交
423
/**
424
 * tipc_link_create - create a new link
J
Jon Paul Maloy 已提交
425
 * @n: pointer to associated node
426 427 428 429 430 431 432 433
 * @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
434
 * @ownnode: identity of own node
435
 * @peer: node id of peer node
436
 * @peer_caps: bitmap describing peer node capabilities
437 438
 * @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
439 440 441
 * @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
442
 *
J
Jon Paul Maloy 已提交
443
 * Returns true if link was created, otherwise false
P
Per Liden 已提交
444
 */
445
bool tipc_link_create(struct net *net, char *if_name, int bearer_id,
446
		      int tolerance, char net_plane, u32 mtu, int priority,
447 448
		      int window, u32 session, u32 self,
		      u32 peer, u8 *peer_id, u16 peer_caps,
449 450 451 452
		      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
453
		      struct tipc_link **link)
P
Per Liden 已提交
454
{
455 456
	char peer_str[NODE_ID_STR_LEN] = {0,};
	char self_str[NODE_ID_STR_LEN] = {0,};
J
Jon Paul Maloy 已提交
457
	struct tipc_link *l;
458

J
Jon Paul Maloy 已提交
459 460 461 462
	l = kzalloc(sizeof(*l), GFP_ATOMIC);
	if (!l)
		return false;
	*link = l;
463
	l->session = session;
464

465 466 467 468 469 470 471 472 473 474
	/* 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
475 476
	snprintf(l->name, sizeof(l->name), "%s:%s-%s:unknown",
		 self_str, if_name, peer_str);
477

478
	strcpy(l->if_name, if_name);
J
Jon Paul Maloy 已提交
479
	l->addr = peer;
480
	l->peer_caps = peer_caps;
481
	l->net = net;
482
	l->in_session = false;
483 484
	l->bearer_id = bearer_id;
	l->tolerance = tolerance;
485 486
	if (bc_rcvlink)
		bc_rcvlink->tolerance = tolerance;
487 488 489 490 491
	l->net_plane = net_plane;
	l->advertised_mtu = mtu;
	l->mtu = mtu;
	l->priority = priority;
	tipc_link_set_queue_limits(l, window);
492
	l->ackers = 1;
493 494
	l->bc_sndlink = bc_sndlink;
	l->bc_rcvlink = bc_rcvlink;
J
Jon Paul Maloy 已提交
495 496 497 498 499 500 501 502 503
	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
504 505
}

506 507 508
/**
 * tipc_link_bc_create - create new link to be used for broadcast
 * @n: pointer to associated node
509
 * @mtu: mtu to be used initially if no peers
510 511 512 513 514 515 516
 * @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
 */
517
bool tipc_link_bc_create(struct net *net, u32 ownnode, u32 peer,
518
			 int mtu, int window, u16 peer_caps,
519 520
			 struct sk_buff_head *inputq,
			 struct sk_buff_head *namedq,
521
			 struct tipc_link *bc_sndlink,
522 523 524 525
			 struct tipc_link **link)
{
	struct tipc_link *l;

526
	if (!tipc_link_create(net, "", MAX_BEARERS, 0, 'Z', mtu, 0, window,
527
			      0, ownnode, peer, NULL, peer_caps, bc_sndlink,
528
			      NULL, inputq, namedq, link))
529 530 531 532
		return false;

	l = *link;
	strcpy(l->name, tipc_bclink_name);
T
Tuong Lien 已提交
533
	trace_tipc_link_reset(l, TIPC_DUMP_ALL, "bclink created!");
534
	tipc_link_reset(l);
535
	l->state = LINK_RESET;
536
	l->ackers = 0;
537
	l->bc_rcvlink = l;
538

539 540 541
	/* Broadcast send link is always up */
	if (link_is_bc_sndlink(l))
		l->state = LINK_ESTABLISHED;
542

543 544 545 546
	/* 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);

547
	return true;
548 549
}

550 551 552 553 554
/**
 * tipc_link_fsm_evt - link finite state machine
 * @l: pointer to link
 * @evt: state machine event to be processed
 */
555
int tipc_link_fsm_evt(struct tipc_link *l, int evt)
556
{
557
	int rc = 0;
T
Tuong Lien 已提交
558
	int old_state = l->state;
559 560

	switch (l->state) {
561
	case LINK_RESETTING:
562
		switch (evt) {
563 564
		case LINK_PEER_RESET_EVT:
			l->state = LINK_PEER_RESET;
565
			break;
566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582
		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;
583
			break;
584 585 586 587 588 589
		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:
590
			break;
591 592
		case LINK_SYNCH_BEGIN_EVT:
		case LINK_SYNCH_END_EVT:
593
		default:
594
			goto illegal_evt;
595 596
		}
		break;
597
	case LINK_PEER_RESET:
598
		switch (evt) {
599 600
		case LINK_RESET_EVT:
			l->state = LINK_ESTABLISHING;
601
			break;
602 603 604
		case LINK_PEER_RESET_EVT:
		case LINK_ESTABLISH_EVT:
		case LINK_FAILURE_EVT:
605
			break;
606 607 608 609 610 611 612 613 614 615 616 617
		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;
618
			break;
619 620 621 622 623 624 625 626
		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:
627
		default:
628
			goto illegal_evt;
629 630
		}
		break;
631
	case LINK_ESTABLISHING:
632
		switch (evt) {
633 634
		case LINK_ESTABLISH_EVT:
			l->state = LINK_ESTABLISHED;
635
			break;
636 637 638 639
		case LINK_FAILOVER_BEGIN_EVT:
			l->state = LINK_FAILINGOVER;
			break;
		case LINK_RESET_EVT:
640 641
			l->state = LINK_RESET;
			break;
642
		case LINK_FAILURE_EVT:
643
		case LINK_PEER_RESET_EVT:
644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660
		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;
661
			break;
662 663
		case LINK_RESET_EVT:
			l->state = LINK_RESET;
664
			break;
665
		case LINK_ESTABLISH_EVT:
666
		case LINK_SYNCH_END_EVT:
667
			break;
668 669 670 671 672
		case LINK_SYNCH_BEGIN_EVT:
			l->state = LINK_SYNCHING;
			break;
		case LINK_FAILOVER_BEGIN_EVT:
		case LINK_FAILOVER_END_EVT:
673
		default:
674
			goto illegal_evt;
675 676
		}
		break;
677
	case LINK_SYNCHING:
678
		switch (evt) {
679 680 681 682 683 684 685
		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;
686
			break;
687 688
		case LINK_RESET_EVT:
			l->state = LINK_RESET;
689
			break;
690 691
		case LINK_ESTABLISH_EVT:
		case LINK_SYNCH_BEGIN_EVT:
692
			break;
693 694 695 696 697
		case LINK_SYNCH_END_EVT:
			l->state = LINK_ESTABLISHED;
			break;
		case LINK_FAILOVER_BEGIN_EVT:
		case LINK_FAILOVER_END_EVT:
698
		default:
699
			goto illegal_evt;
700 701 702
		}
		break;
	default:
703
		pr_err("Unknown FSM state %x in %s\n", l->state, l->name);
704
	}
T
Tuong Lien 已提交
705
	trace_tipc_link_fsm(l->name, old_state, l->state, evt);
706 707 708 709
	return rc;
illegal_evt:
	pr_err("Illegal FSM event %x in state %x on link %s\n",
	       evt, l->state, l->name);
T
Tuong Lien 已提交
710
	trace_tipc_link_fsm(l->name, old_state, l->state, evt);
711 712 713
	return rc;
}

714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754
/* 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]++;
}

T
Tuong Lien 已提交
755 756 757 758 759 760 761 762 763 764 765 766
/**
 * tipc_link_too_silent - check if link is "too silent"
 * @l: tipc link to be checked
 *
 * Returns true if the link 'silent_intv_cnt' is about to reach the
 * 'abort_limit' value, otherwise false
 */
bool tipc_link_too_silent(struct tipc_link *l)
{
	return (l->silent_intv_cnt + 2 > l->abort_limit);
}

767 768 769 770
/* tipc_link_timeout - perform periodic task as instructed from node timeout
 */
int tipc_link_timeout(struct tipc_link *l, struct sk_buff_head *xmitq)
{
771 772
	int mtyp = 0;
	int rc = 0;
773 774 775
	bool state = false;
	bool probe = false;
	bool setup = false;
776 777
	u16 bc_snt = l->bc_sndlink->snd_nxt - 1;
	u16 bc_acked = l->bc_rcvlink->acked;
778
	struct tipc_mon_state *mstate = &l->mon_state;
779

T
Tuong Lien 已提交
780 781
	trace_tipc_link_timeout(l, TIPC_DUMP_NONE, " ");
	trace_tipc_link_too_silent(l, TIPC_DUMP_ALL, " ");
782 783 784
	switch (l->state) {
	case LINK_ESTABLISHED:
	case LINK_SYNCHING:
785
		mtyp = STATE_MSG;
786 787 788 789
		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);
790
		state = bc_acked != bc_snt;
791 792 793 794 795 796 797 798
		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++;
799 800
		break;
	case LINK_RESET:
801 802
		setup = l->rst_cnt++ <= 4;
		setup |= !(l->rst_cnt % 16);
803
		mtyp = RESET_MSG;
804 805
		break;
	case LINK_ESTABLISHING:
806
		setup = true;
807
		mtyp = ACTIVATE_MSG;
808 809
		break;
	case LINK_PEER_RESET:
810
	case LINK_RESETTING:
811 812 813 814
	case LINK_FAILINGOVER:
		break;
	default:
		break;
815
	}
816

817
	if (state || probe || setup)
818
		tipc_link_build_proto_msg(l, mtyp, probe, 0, 0, 0, 0, xmitq);
819

820 821 822
	return rc;
}

P
Per Liden 已提交
823
/**
824
 * link_schedule_user - schedule a message sender for wakeup after congestion
825 826
 * @l: congested link
 * @hdr: header of message that is being sent
827
 * Create pseudo msg to send back to user when congestion abates
P
Per Liden 已提交
828
 */
829
static int link_schedule_user(struct tipc_link *l, struct tipc_msg *hdr)
P
Per Liden 已提交
830
{
831 832
	u32 dnode = tipc_own_addr(l->net);
	u32 dport = msg_origport(hdr);
833 834 835 836
	struct sk_buff *skb;

	/* Create and schedule wakeup pseudo message */
	skb = tipc_msg_create(SOCK_WAKEUP, 0, INT_H_SIZE, 0,
837
			      dnode, l->addr, dport, 0, 0);
838
	if (!skb)
839
		return -ENOBUFS;
840 841 842 843
	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++;
T
Tuong Lien 已提交
844
	trace_tipc_link_conges(l, TIPC_DUMP_ALL, "wakeup scheduled!");
845
	return -ELINKCONG;
P
Per Liden 已提交
846 847
}

848 849
/**
 * link_prepare_wakeup - prepare users for wakeup after congestion
850 851 852
 * @l: congested link
 * Wake up a number of waiting users, as permitted by available space
 * in the send queue
853
 */
Y
YueHaibing 已提交
854
static void link_prepare_wakeup(struct tipc_link *l)
P
Per Liden 已提交
855
{
856
	struct sk_buff *skb, *tmp;
857
	int imp, i = 0;
858

859 860
	skb_queue_walk_safe(&l->wakeupq, skb, tmp) {
		imp = TIPC_SKB_CB(skb)->chain_imp;
861 862 863 864
		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
865
			break;
866
		}
P
Per Liden 已提交
867 868 869
	}
}

870
void tipc_link_reset(struct tipc_link *l)
P
Per Liden 已提交
871
{
872 873 874 875
	struct sk_buff_head list;

	__skb_queue_head_init(&list);

876
	l->in_session = false;
877
	l->session++;
878
	l->mtu = l->advertised_mtu;
879

880
	spin_lock_bh(&l->wakeupq.lock);
881 882 883
	skb_queue_splice_init(&l->wakeupq, &list);
	spin_unlock_bh(&l->wakeupq.lock);

884
	spin_lock_bh(&l->inputq->lock);
885
	skb_queue_splice_init(&list, l->inputq);
886 887
	spin_unlock_bh(&l->inputq->lock);

888 889
	__skb_queue_purge(&l->transmq);
	__skb_queue_purge(&l->deferdq);
890 891 892 893 894 895
	__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;
896 897 898 899 900 901 902
	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;
903 904
	l->snd_nxt_state = 1;
	l->rcv_nxt_state = 1;
905
	l->acked = 0;
906
	l->silent_intv_cnt = 0;
907
	l->rst_cnt = 0;
908
	l->stale_cnt = 0;
909
	l->bc_peer_is_up = false;
910
	memset(&l->mon_state, 0, sizeof(l->mon_state));
911
	tipc_link_reset_stats(l);
P
Per Liden 已提交
912 913
}

914 915 916 917 918 919
/**
 * 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
 *
920
 * Consumes the buffer chain.
921 922 923 924 925 926 927 928
 * 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;
929
	int imp = msg_importance(hdr);
930 931 932
	unsigned int mtu = l->mtu;
	u16 ack = l->rcv_nxt - 1;
	u16 seqno = l->snd_nxt;
933
	u16 bc_ack = l->bc_rcvlink->rcv_nxt - 1;
934 935 936
	struct sk_buff_head *transmq = &l->transmq;
	struct sk_buff_head *backlogq = &l->backlogq;
	struct sk_buff *skb, *_skb, *bskb;
937
	int pkt_cnt = skb_queue_len(list);
938
	int rc = 0;
939

940 941
	if (unlikely(msg_size(hdr) > mtu)) {
		skb_queue_purge(list);
942
		return -EMSGSIZE;
943
	}
944

945 946 947 948 949 950 951 952 953
	/* 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);
	}

954 955 956 957 958
	if (pkt_cnt > 1) {
		l->stats.sent_fragmented++;
		l->stats.sent_fragments += pkt_cnt;
	}

959 960 961 962 963 964
	/* 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);
965
		msg_set_bcast_ack(hdr, bc_ack);
966 967 968

		if (likely(skb_queue_len(transmq) < maxwin)) {
			_skb = skb_clone(skb, GFP_ATOMIC);
969 970
			if (!_skb) {
				skb_queue_purge(list);
971
				return -ENOBUFS;
972
			}
973 974
			__skb_dequeue(list);
			__skb_queue_tail(transmq, skb);
975 976 977 978
			/* next retransmit attempt */
			if (link_is_bc_sndlink(l))
				TIPC_SKB_CB(skb)->nxt_retr =
					jiffies + TIPC_BC_RETR_LIM;
979
			__skb_queue_tail(xmitq, _skb);
980
			TIPC_SKB_CB(skb)->ackers = l->ackers;
981
			l->rcv_unacked = 0;
982
			l->stats.sent_pkts++;
983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002
			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;
1003
	return rc;
1004 1005
}

Y
YueHaibing 已提交
1006 1007
static void tipc_link_advance_backlog(struct tipc_link *l,
				      struct sk_buff_head *xmitq)
1008 1009 1010 1011 1012
{
	struct sk_buff *skb, *_skb;
	struct tipc_msg *hdr;
	u16 seqno = l->snd_nxt;
	u16 ack = l->rcv_nxt - 1;
1013
	u16 bc_ack = l->bc_rcvlink->rcv_nxt - 1;
1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025

	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);
1026 1027 1028 1029
		/* next retransmit attempt */
		if (link_is_bc_sndlink(l))
			TIPC_SKB_CB(skb)->nxt_retr = jiffies + TIPC_BC_RETR_LIM;

1030
		__skb_queue_tail(xmitq, _skb);
1031
		TIPC_SKB_CB(skb)->ackers = l->ackers;
1032
		msg_set_seqno(hdr, seqno);
1033 1034
		msg_set_ack(hdr, ack);
		msg_set_bcast_ack(hdr, bc_ack);
1035
		l->rcv_unacked = 0;
1036
		l->stats.sent_pkts++;
1037 1038 1039 1040 1041
		seqno++;
	}
	l->snd_nxt = seqno;
}

1042
static void link_retransmit_failure(struct tipc_link *l, struct sk_buff *skb)
1043
{
1044 1045 1046
	struct tipc_msg *hdr = buf_msg(skb);

	pr_warn("Retransmission failure on link <%s>\n", l->name);
1047
	link_print(l, "State of link ");
1048 1049 1050 1051
	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));
1052 1053
}

1054 1055 1056 1057 1058 1059 1060
/* 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
 */
Y
YueHaibing 已提交
1061 1062
static int tipc_link_retrans(struct tipc_link *l, struct tipc_link *r,
			     u16 from, u16 to, struct sk_buff_head *xmitq)
1063 1064
{
	struct sk_buff *_skb, *skb = skb_peek(&l->transmq);
1065
	u16 bc_ack = l->bc_rcvlink->rcv_nxt - 1;
1066 1067
	u16 ack = l->rcv_nxt - 1;
	struct tipc_msg *hdr;
1068 1069 1070

	if (!skb)
		return 0;
1071 1072
	if (less(to, from))
		return 0;
1073

T
Tuong Lien 已提交
1074
	trace_tipc_link_retrans(r, from, to, &l->transmq);
1075
	/* Detect repeated retransmit failures on same packet */
1076 1077
	if (r->prev_from != from) {
		r->prev_from = from;
1078
		r->stale_limit = jiffies + msecs_to_jiffies(r->tolerance);
1079
		r->stale_cnt = 0;
1080
	} else if (++r->stale_cnt > 99 && time_after(jiffies, r->stale_limit)) {
1081
		link_retransmit_failure(l, skb);
T
Tuong Lien 已提交
1082 1083 1084
		trace_tipc_list_dump(&l->transmq, true, "retrans failure!");
		trace_tipc_link_dump(l, TIPC_DUMP_NONE, "retrans failure!");
		trace_tipc_link_dump(r, TIPC_DUMP_NONE, "retrans failure!");
1085 1086
		if (link_is_bc_sndlink(l))
			return TIPC_LINK_DOWN_EVT;
1087
		return tipc_link_fsm_evt(l, LINK_FAILURE_EVT);
1088
	}
1089

1090 1091
	skb_queue_walk(&l->transmq, skb) {
		hdr = buf_msg(skb);
1092 1093 1094 1095
		if (less(msg_seqno(hdr), from))
			continue;
		if (more(msg_seqno(hdr), to))
			break;
1096 1097 1098 1099 1100
		if (link_is_bc_sndlink(l)) {
			if (time_before(jiffies, TIPC_SKB_CB(skb)->nxt_retr))
				continue;
			TIPC_SKB_CB(skb)->nxt_retr = jiffies + TIPC_BC_RETR_LIM;
		}
1101 1102 1103 1104
		_skb = __pskb_copy(skb, MIN_H_SIZE, GFP_ATOMIC);
		if (!_skb)
			return 0;
		hdr = buf_msg(_skb);
1105 1106
		msg_set_ack(hdr, ack);
		msg_set_bcast_ack(hdr, bc_ack);
1107 1108 1109 1110 1111 1112 1113
		_skb->priority = TC_PRIO_CONTROL;
		__skb_queue_tail(xmitq, _skb);
		l->stats.retransmitted++;
	}
	return 0;
}

1114
/* tipc_data_input - deliver data and name distr msgs to upper layer
1115
 *
1116
 * Consumes buffer if message is of right type
1117 1118
 * Node lock must be held
 */
1119
static bool tipc_data_input(struct tipc_link *l, struct sk_buff *skb,
1120
			    struct sk_buff_head *inputq)
1121
{
1122
	struct sk_buff_head *mc_inputq = l->bc_rcvlink->inputq;
1123 1124 1125
	struct tipc_msg *hdr = buf_msg(skb);

	switch (msg_user(hdr)) {
1126 1127 1128 1129
	case TIPC_LOW_IMPORTANCE:
	case TIPC_MEDIUM_IMPORTANCE:
	case TIPC_HIGH_IMPORTANCE:
	case TIPC_CRITICAL_IMPORTANCE:
1130
		if (unlikely(msg_in_group(hdr) || msg_mcast(hdr))) {
1131
			skb_queue_tail(mc_inputq, skb);
1132 1133
			return true;
		}
1134
		/* fall through */
1135
	case CONN_MANAGER:
J
Jon Maloy 已提交
1136
		skb_queue_tail(inputq, skb);
1137
		return true;
1138 1139 1140
	case GROUP_PROTOCOL:
		skb_queue_tail(mc_inputq, skb);
		return true;
1141
	case NAME_DISTRIBUTOR:
1142 1143
		l->bc_rcvlink->state = LINK_ESTABLISHED;
		skb_queue_tail(l->namedq, skb);
1144 1145
		return true;
	case MSG_BUNDLER:
1146
	case TUNNEL_PROTOCOL:
1147
	case MSG_FRAGMENTER:
1148
	case BCAST_PROTOCOL:
1149
		return false;
1150
	default:
1151 1152
		pr_warn("Dropping received illegal msg type\n");
		kfree_skb(skb);
1153
		return true;
1154
	};
1155
}
1156 1157 1158 1159

/* tipc_link_input - process packet that has passed link protocol check
 *
 * Consumes buffer
1160
 */
1161 1162
static int tipc_link_input(struct tipc_link *l, struct sk_buff *skb,
			   struct sk_buff_head *inputq)
1163
{
1164 1165
	struct tipc_msg *hdr = buf_msg(skb);
	struct sk_buff **reasm_skb = &l->reasm_buf;
1166
	struct sk_buff *iskb;
1167
	struct sk_buff_head tmpq;
1168
	int usr = msg_user(hdr);
1169
	int rc = 0;
1170 1171
	int pos = 0;
	int ipos = 0;
1172

1173 1174 1175 1176
	if (unlikely(usr == TUNNEL_PROTOCOL)) {
		if (msg_type(hdr) == SYNCH_MSG) {
			__skb_queue_purge(&l->deferdq);
			goto drop;
1177
		}
1178 1179 1180 1181 1182 1183 1184
		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;
1185
		if (tipc_data_input(l, skb, inputq))
1186 1187 1188 1189
			return rc;
		usr = msg_user(hdr);
		reasm_skb = &l->failover_reasm_skb;
	}
1190

1191
	if (usr == MSG_BUNDLER) {
1192
		skb_queue_head_init(&tmpq);
1193 1194
		l->stats.recv_bundles++;
		l->stats.recv_bundled += msg_msgcnt(hdr);
1195
		while (tipc_msg_extract(skb, &iskb, &pos))
1196 1197
			tipc_data_input(l, iskb, &tmpq);
		tipc_skb_queue_splice_tail(&tmpq, inputq);
1198
		return 0;
1199 1200 1201 1202
	} else if (usr == MSG_FRAGMENTER) {
		l->stats.recv_fragments++;
		if (tipc_buf_append(reasm_skb, &skb)) {
			l->stats.recv_fragmented++;
1203
			tipc_data_input(l, skb, inputq);
1204 1205
		} else if (!*reasm_skb && !link_is_bc_rcvlink(l)) {
			pr_warn_ratelimited("Unable to build fragment list\n");
1206
			return tipc_link_fsm_evt(l, LINK_FAILURE_EVT);
1207
		}
1208
		return 0;
1209
	} else if (usr == BCAST_PROTOCOL) {
1210
		tipc_bcast_lock(l->net);
1211
		tipc_link_bc_init_rcv(l->bc_rcvlink, hdr);
1212
		tipc_bcast_unlock(l->net);
1213 1214 1215
	}
drop:
	kfree_skb(skb);
1216
	return 0;
1217 1218
}

1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233
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;
}

1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308
/* tipc_build_gap_ack_blks - build Gap ACK blocks
 * @l: tipc link that data have come with gaps in sequence if any
 * @data: data buffer to store the Gap ACK blocks after built
 *
 * returns the actual allocated memory size
 */
static u16 tipc_build_gap_ack_blks(struct tipc_link *l, void *data)
{
	struct sk_buff *skb = skb_peek(&l->deferdq);
	struct tipc_gap_ack_blks *ga = data;
	u16 len, expect, seqno = 0;
	u8 n = 0;

	if (!skb)
		goto exit;

	expect = buf_seqno(skb);
	skb_queue_walk(&l->deferdq, skb) {
		seqno = buf_seqno(skb);
		if (unlikely(more(seqno, expect))) {
			ga->gacks[n].ack = htons(expect - 1);
			ga->gacks[n].gap = htons(seqno - expect);
			if (++n >= MAX_GAP_ACK_BLKS) {
				pr_info_ratelimited("Too few Gap ACK blocks!\n");
				goto exit;
			}
		} else if (unlikely(less(seqno, expect))) {
			pr_warn("Unexpected skb in deferdq!\n");
			continue;
		}
		expect = seqno + 1;
	}

	/* last block */
	ga->gacks[n].ack = htons(seqno);
	ga->gacks[n].gap = 0;
	n++;

exit:
	len = tipc_gap_ack_blks_sz(n);
	ga->len = htons(len);
	ga->gack_cnt = n;
	return len;
}

/* tipc_link_advance_transmq - advance TIPC link transmq queue by releasing
 *			       acked packets, also doing retransmissions if
 *			       gaps found
 * @l: tipc link with transmq queue to be advanced
 * @acked: seqno of last packet acked by peer without any gaps before
 * @gap: # of gap packets
 * @ga: buffer pointer to Gap ACK blocks from peer
 * @xmitq: queue for accumulating the retransmitted packets if any
 */
static void tipc_link_advance_transmq(struct tipc_link *l, u16 acked, u16 gap,
				      struct tipc_gap_ack_blks *ga,
				      struct sk_buff_head *xmitq)
{
	struct sk_buff *skb, *_skb, *tmp;
	struct tipc_msg *hdr;
	u16 bc_ack = l->bc_rcvlink->rcv_nxt - 1;
	u16 ack = l->rcv_nxt - 1;
	u16 seqno;
	u16 n = 0;

	skb_queue_walk_safe(&l->transmq, skb, tmp) {
		seqno = buf_seqno(skb);

next_gap_ack:
		if (less_eq(seqno, acked)) {
			/* release skb */
			__skb_unlink(skb, &l->transmq);
			kfree_skb(skb);
		} else if (less_eq(seqno, acked + gap)) {
			/* retransmit skb */
1309 1310 1311 1312
			if (time_before(jiffies, TIPC_SKB_CB(skb)->nxt_retr))
				continue;
			TIPC_SKB_CB(skb)->nxt_retr = TIPC_UC_RETR_TIME;

1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333
			_skb = __pskb_copy(skb, MIN_H_SIZE, GFP_ATOMIC);
			if (!_skb)
				continue;
			hdr = buf_msg(_skb);
			msg_set_ack(hdr, ack);
			msg_set_bcast_ack(hdr, bc_ack);
			_skb->priority = TC_PRIO_CONTROL;
			__skb_queue_tail(xmitq, _skb);
			l->stats.retransmitted++;
		} else {
			/* retry with Gap ACK blocks if any */
			if (!ga || n >= ga->gack_cnt)
				break;
			acked = ntohs(ga->gacks[n].ack);
			gap = ntohs(ga->gacks[n].gap);
			n++;
			goto next_gap_ack;
		}
	}
}

1334
/* tipc_link_build_state_msg: prepare link state message for transmission
1335 1336 1337
 *
 * Note that sending of broadcast ack is coordinated among nodes, to reduce
 * risk of ack storms towards the sender
1338
 */
1339
int tipc_link_build_state_msg(struct tipc_link *l, struct sk_buff_head *xmitq)
1340
{
1341 1342 1343 1344 1345
	if (!l)
		return 0;

	/* Broadcast ACK must be sent via a unicast link => defer to caller */
	if (link_is_bc_rcvlink(l)) {
1346
		if (((l->rcv_nxt ^ tipc_own_addr(l->net)) & 0xf) != 0xf)
1347 1348
			return 0;
		l->rcv_unacked = 0;
1349 1350 1351 1352

		/* Use snd_nxt to store peer's snd_nxt in broadcast rcv link */
		l->snd_nxt = l->rcv_nxt;
		return TIPC_LINK_SND_STATE;
1353 1354 1355
	}

	/* Unicast ACK */
1356 1357
	l->rcv_unacked = 0;
	l->stats.sent_acks++;
1358
	tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, 0, 0, 0, xmitq);
1359
	return 0;
1360 1361
}

1362 1363 1364 1365 1366
/* 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;
1367
	struct sk_buff *skb;
1368 1369 1370 1371

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

1372
	tipc_link_build_proto_msg(l, mtyp, 0, 0, 0, 0, 0, xmitq);
1373 1374 1375 1376 1377

	/* 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);
1378 1379
}

1380
/* tipc_link_build_nack_msg: prepare link nack message for transmission
1381 1382
 * Note that sending of broadcast NACK is coordinated among nodes, to
 * reduce the risk of NACK storms towards the sender
1383
 */
1384 1385
static int tipc_link_build_nack_msg(struct tipc_link *l,
				    struct sk_buff_head *xmitq)
1386 1387
{
	u32 def_cnt = ++l->stats.deferred_recv;
1388
	u32 defq_len = skb_queue_len(&l->deferdq);
1389
	int match1, match2;
1390

1391 1392 1393 1394 1395 1396 1397
	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;
	}
1398

1399
	if (defq_len >= 3 && !((defq_len - 3) % 16))
1400
		tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, 0, 0, 0, xmitq);
1401
	return 0;
1402 1403
}

1404
/* tipc_link_rcv - process TIPC packets/messages arriving from off-node
1405
 * @l: the link that should handle the message
1406 1407 1408 1409 1410 1411
 * @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)
{
1412
	struct sk_buff_head *defq = &l->deferdq;
1413
	struct tipc_msg *hdr = buf_msg(skb);
1414
	u16 seqno, rcv_nxt, win_lim;
1415 1416
	int rc = 0;

1417 1418 1419 1420 1421 1422 1423
	/* Verify and update link state */
	if (unlikely(msg_user(hdr) == LINK_PROTOCOL))
		return tipc_link_proto_rcv(l, skb, xmitq);

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

1424
	do {
1425
		hdr = buf_msg(skb);
1426 1427
		seqno = msg_seqno(hdr);
		rcv_nxt = l->rcv_nxt;
1428
		win_lim = rcv_nxt + TIPC_MAX_LINK_WIN;
1429

1430
		if (unlikely(!link_is_up(l))) {
1431 1432 1433
			if (l->state == LINK_ESTABLISHING)
				rc = TIPC_LINK_UP_EVT;
			goto drop;
1434 1435
		}

1436 1437 1438 1439 1440 1441
		/* Drop if outside receive window */
		if (unlikely(less(seqno, rcv_nxt) || more(seqno, win_lim))) {
			l->stats.duplicates++;
			goto drop;
		}

1442 1443
		/* Forward queues and wake up waiting users */
		if (likely(tipc_link_release_pkts(l, msg_ack(hdr)))) {
1444
			l->stale_cnt = 0;
1445 1446 1447 1448 1449
			tipc_link_advance_backlog(l, xmitq);
			if (unlikely(!skb_queue_empty(&l->wakeupq)))
				link_prepare_wakeup(l);
		}

1450 1451
		/* Defer delivery if sequence gap */
		if (unlikely(seqno != rcv_nxt)) {
1452
			__tipc_skb_queue_sorted(defq, seqno, skb);
1453
			rc |= tipc_link_build_nack_msg(l, xmitq);
1454
			break;
1455 1456
		}

1457
		/* Deliver packet */
1458
		l->rcv_nxt++;
1459
		l->stats.recv_pkts++;
1460
		if (!tipc_data_input(l, skb, l->inputq))
1461
			rc |= tipc_link_input(l, skb, l->inputq);
1462
		if (unlikely(++l->rcv_unacked >= TIPC_MIN_LINK_WIN))
1463
			rc |= tipc_link_build_state_msg(l, xmitq);
1464
		if (unlikely(rc & ~TIPC_LINK_SND_STATE))
1465
			break;
1466
	} while ((skb = __tipc_skb_dequeue(defq, l->rcv_nxt)));
1467 1468 1469 1470

	return rc;
drop:
	kfree_skb(skb);
1471 1472 1473
	return rc;
}

1474
static void tipc_link_build_proto_msg(struct tipc_link *l, int mtyp, bool probe,
1475 1476
				      bool probe_reply, u16 rcvgap,
				      int tolerance, int priority,
1477 1478
				      struct sk_buff_head *xmitq)
{
1479
	struct tipc_link *bcl = l->bc_rcvlink;
1480 1481 1482
	struct sk_buff *skb;
	struct tipc_msg *hdr;
	struct sk_buff_head *dfq = &l->deferdq;
1483
	bool node_up = link_is_up(bcl);
1484 1485 1486
	struct tipc_mon_state *mstate = &l->mon_state;
	int dlen = 0;
	void *data;
1487
	u16 glen = 0;
1488 1489

	/* Don't send protocol message during reset or link failover */
1490
	if (tipc_link_is_blocked(l))
1491 1492
		return;

1493 1494 1495 1496 1497 1498 1499
	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,
1500 1501
			      tipc_max_domain_size + MAX_GAP_ACK_BLKS_SZ,
			      l->addr, tipc_own_addr(l->net), 0, 0, 0);
1502 1503 1504 1505
	if (!skb)
		return;

	hdr = buf_msg(skb);
1506
	data = msg_data(hdr);
1507 1508
	msg_set_session(hdr, l->session);
	msg_set_bearer_id(hdr, l->bearer_id);
1509
	msg_set_net_plane(hdr, l->net_plane);
1510 1511
	msg_set_next_sent(hdr, l->snd_nxt);
	msg_set_ack(hdr, l->rcv_nxt - 1);
1512
	msg_set_bcast_ack(hdr, bcl->rcv_nxt - 1);
1513
	msg_set_bc_ack_invalid(hdr, !node_up);
1514
	msg_set_last_bcast(hdr, l->bc_sndlink->snd_nxt - 1);
1515 1516 1517 1518
	msg_set_link_tolerance(hdr, tolerance);
	msg_set_linkprio(hdr, priority);
	msg_set_redundant_link(hdr, node_up);
	msg_set_seq_gap(hdr, 0);
1519
	msg_set_seqno(hdr, l->snd_nxt + U16_MAX / 2);
1520 1521

	if (mtyp == STATE_MSG) {
1522 1523
		if (l->peer_caps & TIPC_LINK_PROTO_SEQNO)
			msg_set_seqno(hdr, l->snd_nxt_state++);
1524
		msg_set_seq_gap(hdr, rcvgap);
1525
		msg_set_bc_gap(hdr, link_bc_rcv_gap(bcl));
1526
		msg_set_probe(hdr, probe);
1527
		msg_set_is_keepalive(hdr, probe || probe_reply);
1528 1529 1530 1531 1532
		if (l->peer_caps & TIPC_GAP_ACK_BLOCK)
			glen = tipc_build_gap_ack_blks(l, data);
		tipc_mon_prep(l->net, data + glen, &dlen, mstate, l->bearer_id);
		msg_set_size(hdr, INT_H_SIZE + glen + dlen);
		skb_trim(skb, INT_H_SIZE + glen + dlen);
1533
		l->stats.sent_states++;
1534
		l->rcv_unacked = 0;
1535 1536
	} else {
		/* RESET_MSG or ACTIVATE_MSG */
1537 1538 1539 1540
		if (mtyp == ACTIVATE_MSG) {
			msg_set_dest_session_valid(hdr, 1);
			msg_set_dest_session(hdr, l->peer_session);
		}
1541
		msg_set_max_pkt(hdr, l->advertised_mtu);
1542 1543 1544
		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);
1545
	}
1546 1547 1548 1549
	if (probe)
		l->stats.sent_probes++;
	if (rcvgap)
		l->stats.sent_nacks++;
1550
	skb->priority = TC_PRIO_CONTROL;
1551
	__skb_queue_tail(xmitq, skb);
T
Tuong Lien 已提交
1552
	trace_tipc_proto_build(skb, false, l->name);
P
Per Liden 已提交
1553 1554
}

L
LUU Duc Canh 已提交
1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584
void tipc_link_create_dummy_tnl_msg(struct tipc_link *l,
				    struct sk_buff_head *xmitq)
{
	u32 onode = tipc_own_addr(l->net);
	struct tipc_msg *hdr, *ihdr;
	struct sk_buff_head tnlq;
	struct sk_buff *skb;
	u32 dnode = l->addr;

	skb_queue_head_init(&tnlq);
	skb = tipc_msg_create(TUNNEL_PROTOCOL, FAILOVER_MSG,
			      INT_H_SIZE, BASIC_H_SIZE,
			      dnode, onode, 0, 0, 0);
	if (!skb) {
		pr_warn("%sunable to create tunnel packet\n", link_co_err);
		return;
	}

	hdr = buf_msg(skb);
	msg_set_msgcnt(hdr, 1);
	msg_set_bearer_id(hdr, l->peer_bearer_id);

	ihdr = (struct tipc_msg *)msg_data(hdr);
	tipc_msg_init(onode, ihdr, TIPC_LOW_IMPORTANCE, TIPC_DIRECT_MSG,
		      BASIC_H_SIZE, dnode);
	msg_set_errcode(ihdr, TIPC_ERR_NO_PORT);
	__skb_queue_tail(&tnlq, skb);
	tipc_link_xmit(l, &tnlq, xmitq);
}

1585
/* tipc_link_tnl_prepare(): prepare and return a list of tunnel packets
1586
 * with contents of the link's transmit and backlog queues.
P
Per Liden 已提交
1587
 */
1588 1589
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
1590
{
1591 1592 1593 1594 1595
	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
1596

1597
	if (!tnl)
P
Per Liden 已提交
1598 1599
		return;

1600 1601
	skb_queue_head_init(&tnlq);
	skb_queue_head_init(&tmpxq);
1602

1603 1604
	/* At least one packet required for safe algorithm => add dummy */
	skb = tipc_msg_create(TIPC_LOW_IMPORTANCE, TIPC_DIRECT_MSG,
1605
			      BASIC_H_SIZE, 0, l->addr, tipc_own_addr(l->net),
1606 1607 1608
			      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
1609 1610
		return;
	}
1611 1612 1613 1614 1615
	skb_queue_tail(&tnlq, skb);
	tipc_link_xmit(l, &tnlq, &tmpxq);
	__skb_queue_purge(&tmpxq);

	/* Initialize reusable tunnel packet header */
1616
	tipc_msg_init(tipc_own_addr(l->net), &tnlhdr, TUNNEL_PROTOCOL,
1617 1618 1619 1620 1621 1622
		      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
1623
	skb_queue_walk(queue, skb) {
1624 1625 1626 1627 1628
		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);
1629
		tnlskb = tipc_buf_acquire(pktlen + INT_H_SIZE, GFP_ATOMIC);
1630 1631
		if (!tnlskb) {
			pr_warn("%sunable to send packet\n", link_co_err);
P
Per Liden 已提交
1632 1633
			return;
		}
1634 1635 1636
		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
1637
	}
1638 1639 1640
	if (queue != &l->backlogq) {
		queue = &l->backlogq;
		goto tnl;
1641
	}
1642

1643
	tipc_link_xmit(tnl, &tnlq, xmitq);
P
Per Liden 已提交
1644

1645 1646 1647 1648
	if (mtyp == FAILOVER_MSG) {
		tnl->drop_point = l->rcv_nxt;
		tnl->failover_reasm_skb = l->reasm_buf;
		l->reasm_buf = NULL;
1649
	}
P
Per Liden 已提交
1650 1651
}

1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680
/* tipc_link_validate_msg(): validate message against current link state
 * Returns true if message should be accepted, otherwise false
 */
bool tipc_link_validate_msg(struct tipc_link *l, struct tipc_msg *hdr)
{
	u16 curr_session = l->peer_session;
	u16 session = msg_session(hdr);
	int mtyp = msg_type(hdr);

	if (msg_user(hdr) != LINK_PROTOCOL)
		return true;

	switch (mtyp) {
	case RESET_MSG:
		if (!l->in_session)
			return true;
		/* Accept only RESET with new session number */
		return more(session, curr_session);
	case ACTIVATE_MSG:
		if (!l->in_session)
			return true;
		/* Accept only ACTIVATE with new or current session number */
		return !less(session, curr_session);
	case STATE_MSG:
		/* Accept only STATE with current session number */
		if (!l->in_session)
			return false;
		if (session != curr_session)
			return false;
1681 1682 1683
		/* Extra sanity check */
		if (!link_is_up(l) && msg_ack(hdr))
			return false;
1684 1685 1686 1687 1688 1689 1690 1691 1692
		if (!(l->peer_caps & TIPC_LINK_PROTO_SEQNO))
			return true;
		/* Accept only STATE with new sequence number */
		return !less(msg_seqno(hdr), l->rcv_nxt_state);
	default:
		return false;
	}
}

1693 1694 1695 1696 1697 1698 1699 1700 1701
/* 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);
1702
	struct tipc_gap_ack_blks *ga = NULL;
1703
	u16 rcvgap = 0;
1704 1705
	u16 ack = msg_ack(hdr);
	u16 gap = msg_seq_gap(hdr);
1706 1707 1708
	u16 peers_snd_nxt =  msg_next_sent(hdr);
	u16 peers_tol = msg_link_tolerance(hdr);
	u16 peers_prio = msg_linkprio(hdr);
1709
	u16 rcv_nxt = l->rcv_nxt;
1710
	u16 dlen = msg_data_sz(hdr);
1711
	int mtyp = msg_type(hdr);
1712
	bool reply = msg_probe(hdr);
1713
	u16 glen = 0;
1714
	void *data;
1715 1716 1717
	char *if_name;
	int rc = 0;

T
Tuong Lien 已提交
1718
	trace_tipc_proto_rcv(skb, false, l->name);
1719
	if (tipc_link_is_blocked(l) || !xmitq)
1720 1721
		goto exit;

1722
	if (tipc_own_addr(l->net) > msg_prevnode(hdr))
1723 1724
		l->net_plane = msg_net_plane(hdr);

1725 1726 1727 1728
	skb_linearize(skb);
	hdr = buf_msg(skb);
	data = msg_data(hdr);

T
Tuong Lien 已提交
1729 1730 1731
	if (!tipc_link_validate_msg(l, hdr)) {
		trace_tipc_skb_dump(skb, false, "PROTO invalid (1)!");
		trace_tipc_link_dump(l, TIPC_DUMP_NONE, "PROTO invalid (1)!");
1732
		goto exit;
T
Tuong Lien 已提交
1733
	}
1734

1735
	switch (mtyp) {
1736 1737 1738 1739 1740 1741 1742 1743
	case RESET_MSG:
	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;
1744
		strncpy(if_name, data, TIPC_MAX_IF_NAME);
1745 1746

		/* Update own tolerance if peer indicates a non-zero value */
1747
		if (in_range(peers_tol, TIPC_MIN_LINK_TOL, TIPC_MAX_LINK_TOL)) {
1748
			l->tolerance = peers_tol;
1749 1750
			l->bc_rcvlink->tolerance = peers_tol;
		}
1751 1752 1753 1754
		/* 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;

J
Jon Maloy 已提交
1755 1756
		/* If peer is going down we want full re-establish cycle */
		if (msg_peer_stopping(hdr)) {
1757
			rc = tipc_link_fsm_evt(l, LINK_FAILURE_EVT);
J
Jon Maloy 已提交
1758 1759
			break;
		}
1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770

		/* If this endpoint was re-created while peer was ESTABLISHING
		 * it doesn't know current session number. Force re-synch.
		 */
		if (mtyp == ACTIVATE_MSG && msg_dest_session_valid(hdr) &&
		    l->session != msg_dest_session(hdr)) {
			if (less(l->session, msg_dest_session(hdr)))
				l->session = msg_dest_session(hdr) + 1;
			break;
		}

J
Jon Maloy 已提交
1771 1772
		/* ACTIVATE_MSG serves as PEER_RESET if link is already down */
		if (mtyp == RESET_MSG || !link_is_up(l))
1773 1774 1775
			rc = tipc_link_fsm_evt(l, LINK_PEER_RESET_EVT);

		/* ACTIVATE_MSG takes up link if it was already locally reset */
J
Jon Maloy 已提交
1776
		if (mtyp == ACTIVATE_MSG && l->state == LINK_ESTABLISHING)
1777 1778
			rc = TIPC_LINK_UP_EVT;

1779
		l->peer_session = msg_session(hdr);
1780
		l->in_session = true;
1781 1782 1783 1784
		l->peer_bearer_id = msg_bearer_id(hdr);
		if (l->mtu > msg_max_pkt(hdr))
			l->mtu = msg_max_pkt(hdr);
		break;
1785

1786
	case STATE_MSG:
1787 1788
		l->rcv_nxt_state = msg_seqno(hdr) + 1;

1789
		/* Update own tolerance if peer indicates a non-zero value */
1790
		if (in_range(peers_tol, TIPC_MIN_LINK_TOL, TIPC_MAX_LINK_TOL)) {
1791
			l->tolerance = peers_tol;
1792 1793
			l->bc_rcvlink->tolerance = peers_tol;
		}
1794 1795 1796
		/* Update own prio if peer indicates a different value */
		if ((peers_prio != l->priority) &&
		    in_range(peers_prio, 1, TIPC_MAX_LINK_PRI)) {
1797 1798 1799 1800
			l->priority = peers_prio;
			rc = tipc_link_fsm_evt(l, LINK_FAILURE_EVT);
		}

1801 1802 1803 1804
		l->silent_intv_cnt = 0;
		l->stats.recv_states++;
		if (msg_probe(hdr))
			l->stats.recv_probes++;
1805 1806 1807 1808

		if (!link_is_up(l)) {
			if (l->state == LINK_ESTABLISHING)
				rc = TIPC_LINK_UP_EVT;
1809
			break;
1810
		}
1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821

		/* Receive Gap ACK blocks from peer if any */
		if (l->peer_caps & TIPC_GAP_ACK_BLOCK) {
			ga = (struct tipc_gap_ack_blks *)data;
			glen = ntohs(ga->len);
			/* sanity check: if failed, ignore Gap ACK blocks */
			if (glen != tipc_gap_ack_blks_sz(ga->gack_cnt))
				ga = NULL;
		}

		tipc_mon_rcv(l->net, data + glen, dlen - glen, l->addr,
1822
			     &l->mon_state, l->bearer_id);
1823

1824
		/* Send NACK if peer has sent pkts we haven't received yet */
1825
		if (more(peers_snd_nxt, rcv_nxt) && !tipc_link_is_synching(l))
1826
			rcvgap = peers_snd_nxt - l->rcv_nxt;
1827 1828 1829
		if (rcvgap || reply)
			tipc_link_build_proto_msg(l, STATE_MSG, 0, reply,
						  rcvgap, 0, 0, xmitq);
1830 1831

		tipc_link_advance_transmq(l, ack, gap, ga, xmitq);
1832 1833

		/* If NACK, retransmit will now start at right position */
1834
		if (gap)
1835
			l->stats.recv_nacks++;
1836

1837 1838 1839 1840 1841 1842 1843 1844 1845
		tipc_link_advance_backlog(l, xmitq);
		if (unlikely(!skb_queue_empty(&l->wakeupq)))
			link_prepare_wakeup(l);
	}
exit:
	kfree_skb(skb);
	return rc;
}

1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858
/* 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,
1859
			      0, l->addr, tipc_own_addr(l->net), 0, 0, 0);
1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878
	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.
 */
1879 1880
static void tipc_link_build_bc_init_msg(struct tipc_link *l,
					struct sk_buff_head *xmitq)
1881 1882 1883 1884 1885 1886
{
	struct sk_buff_head list;

	__skb_queue_head_init(&list);
	if (!tipc_link_build_bc_proto_msg(l->bc_rcvlink, false, 0, &list))
		return;
1887
	msg_set_bc_ack_invalid(buf_msg(skb_peek(&list)), true);
1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919
	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;
}

/* tipc_link_bc_sync_rcv - update rcv link according to peer's send state
 */
1920 1921
int tipc_link_bc_sync_rcv(struct tipc_link *l, struct tipc_msg *hdr,
			  struct sk_buff_head *xmitq)
1922
{
1923
	struct tipc_link *snd_l = l->bc_sndlink;
1924
	u16 peers_snd_nxt = msg_bc_snd_nxt(hdr);
1925 1926 1927
	u16 from = msg_bcast_ack(hdr) + 1;
	u16 to = from + msg_bc_gap(hdr) - 1;
	int rc = 0;
1928 1929

	if (!link_is_up(l))
1930
		return rc;
1931 1932

	if (!msg_peer_node_is_up(hdr))
1933
		return rc;
1934

1935 1936 1937 1938 1939
	/* 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)
1940
		return rc;
1941

1942 1943
	l->stats.recv_nacks++;

1944 1945
	/* Ignore if peers_snd_nxt goes beyond receive window */
	if (more(peers_snd_nxt, l->rcv_nxt + l->window))
1946 1947
		return rc;

1948
	rc = tipc_link_retrans(snd_l, l, from, to, xmitq);
1949 1950 1951 1952 1953 1954 1955 1956 1957 1958

	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 */
1959 1960 1961

	if (!more(peers_snd_nxt, l->rcv_nxt)) {
		l->nack_state = BC_NACK_SND_CONDITIONAL;
1962
		return 0;
1963 1964 1965 1966 1967
	}

	/* 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;
1968
		return 0;
1969 1970 1971 1972 1973 1974
	}

	/* 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)
1975
			return 0;
1976 1977 1978 1979 1980
	}

	/* 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;
1981
	return 0;
1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995
}

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;

T
Tuong Lien 已提交
1996
	trace_tipc_link_bc_ack(l, l->acked, acked, &snd_l->transmq);
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
	/* 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
2019 2020
 * This function is here for backwards compatibility, since
 * no BCAST_PROTOCOL/STATE messages occur from TIPC v2.5.
2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041
 */
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;

2042
	if (dnode == tipc_own_addr(l->net)) {
2043
		tipc_link_bc_ack_rcv(l, acked, xmitq);
2044
		rc = tipc_link_retrans(l->bc_sndlink, l, from, to, xmitq);
2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055
		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;
}

2056
void tipc_link_set_queue_limits(struct tipc_link *l, u32 win)
P
Per Liden 已提交
2057
{
2058
	int max_bulk = TIPC_MAX_PUBL / (l->mtu / ITEM_SIZE);
2059 2060

	l->window = win;
2061 2062 2063 2064
	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);
2065
	l->backlog[TIPC_SYSTEM_IMPORTANCE].limit   = max_bulk;
P
Per Liden 已提交
2066 2067 2068
}

/**
2069
 * link_reset_stats - reset link statistics
2070
 * @l: pointer to link
P
Per Liden 已提交
2071
 */
2072
void tipc_link_reset_stats(struct tipc_link *l)
P
Per Liden 已提交
2073
{
2074
	memset(&l->stats, 0, sizeof(l->stats));
P
Per Liden 已提交
2075 2076
}

2077
static void link_print(struct tipc_link *l, const char *str)
P
Per Liden 已提交
2078
{
2079
	struct sk_buff *hskb = skb_peek(&l->transmq);
2080
	u16 head = hskb ? msg_seqno(buf_msg(hskb)) : l->snd_nxt - 1;
2081
	u16 tail = l->snd_nxt - 1;
2082

2083
	pr_info("%s Link <%s> state %x\n", str, l->name, l->state);
2084 2085 2086
	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
2087
}
2088 2089 2090 2091 2092 2093 2094 2095

/* 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,
2096
			       tipc_nl_prop_policy, NULL);
2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125
	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;
}
2126

2127
static int __tipc_nl_add_stats(struct sk_buff *skb, struct tipc_stats *s)
2128 2129 2130 2131 2132 2133 2134 2135 2136 2137
{
	int i;
	struct nlattr *stats;

	struct nla_map {
		u32 key;
		u32 val;
	};

	struct nla_map map[] = {
2138
		{TIPC_NLA_STATS_RX_INFO, 0},
2139 2140 2141 2142
		{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},
2143
		{TIPC_NLA_STATS_TX_INFO, 0},
2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192
		{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 */
2193 2194
int __tipc_nl_add_link(struct net *net, struct tipc_nl_msg *msg,
		       struct tipc_link *link, int nlflags)
2195
{
2196
	u32 self = tipc_own_addr(net);
2197 2198
	struct nlattr *attrs;
	struct nlattr *prop;
2199 2200
	void *hdr;
	int err;
2201

2202
	hdr = genlmsg_put(msg->skb, msg->portid, msg->seq, &tipc_genl_family,
2203
			  nlflags, TIPC_NL_LINK_GET);
2204 2205 2206 2207 2208 2209 2210 2211 2212
	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;
2213
	if (nla_put_u32(msg->skb, TIPC_NLA_LINK_DEST, tipc_cluster_mask(self)))
2214
		goto attr_msg_full;
2215
	if (nla_put_u32(msg->skb, TIPC_NLA_LINK_MTU, link->mtu))
2216
		goto attr_msg_full;
2217
	if (nla_put_u32(msg->skb, TIPC_NLA_LINK_RX, link->stats.recv_pkts))
2218
		goto attr_msg_full;
2219
	if (nla_put_u32(msg->skb, TIPC_NLA_LINK_TX, link->stats.sent_pkts))
2220 2221 2222 2223 2224
		goto attr_msg_full;

	if (tipc_link_is_up(link))
		if (nla_put_flag(msg->skb, TIPC_NLA_LINK_UP))
			goto attr_msg_full;
2225
	if (link->active)
2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236
		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,
2237
			link->window))
2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260
		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;
}
2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273

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[] = {
2274
		{TIPC_NLA_STATS_RX_INFO, stats->recv_pkts},
2275 2276 2277 2278
		{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},
2279
		{TIPC_NLA_STATS_TX_INFO, stats->sent_pkts},
2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319
		{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);
2320 2321
	u32 bc_mode = tipc_bcast_get_broadcast_mode(net);
	u32 bc_ratio = tipc_bcast_get_broadcast_ratio(net);
2322 2323 2324 2325 2326 2327 2328 2329 2330
	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
2331 2332
	if (!hdr) {
		tipc_bcast_unlock(net);
2333
		return -EMSGSIZE;
I
Insu Yun 已提交
2334
	}
2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347

	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;
2348
	if (nla_put_u32(msg->skb, TIPC_NLA_LINK_RX, 0))
2349
		goto attr_msg_full;
2350
	if (nla_put_u32(msg->skb, TIPC_NLA_LINK_TX, 0))
2351 2352 2353 2354 2355 2356 2357
		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;
2358 2359 2360 2361 2362 2363
	if (nla_put_u32(msg->skb, TIPC_NLA_PROP_BROADCAST, bc_mode))
		goto prop_msg_full;
	if (bc_mode & BCLINK_MODE_SEL)
		if (nla_put_u32(msg->skb, TIPC_NLA_PROP_BROADCAST_RATIO,
				bc_ratio))
			goto prop_msg_full;
2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386
	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;
}

2387 2388
void tipc_link_set_tolerance(struct tipc_link *l, u32 tol,
			     struct sk_buff_head *xmitq)
2389 2390
{
	l->tolerance = tol;
2391 2392
	if (l->bc_rcvlink)
		l->bc_rcvlink->tolerance = tol;
2393 2394
	if (link_is_up(l))
		tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, 0, tol, 0, xmitq);
2395 2396
}

2397 2398
void tipc_link_set_prio(struct tipc_link *l, u32 prio,
			struct sk_buff_head *xmitq)
2399 2400
{
	l->priority = prio;
2401
	tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, 0, 0, prio, xmitq);
2402 2403 2404 2405 2406 2407
}

void tipc_link_set_abort_limit(struct tipc_link *l, u32 limit)
{
	l->abort_limit = limit;
}
T
Tuong Lien 已提交
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 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526

char *tipc_link_name_ext(struct tipc_link *l, char *buf)
{
	if (!l)
		scnprintf(buf, TIPC_MAX_LINK_NAME, "null");
	else if (link_is_bc_sndlink(l))
		scnprintf(buf, TIPC_MAX_LINK_NAME, "broadcast-sender");
	else if (link_is_bc_rcvlink(l))
		scnprintf(buf, TIPC_MAX_LINK_NAME,
			  "broadcast-receiver, peer %x", l->addr);
	else
		memcpy(buf, l->name, TIPC_MAX_LINK_NAME);

	return buf;
}

/**
 * tipc_link_dump - dump TIPC link data
 * @l: tipc link to be dumped
 * @dqueues: bitmask to decide if any link queue to be dumped?
 *           - TIPC_DUMP_NONE: don't dump link queues
 *           - TIPC_DUMP_TRANSMQ: dump link transmq queue
 *           - TIPC_DUMP_BACKLOGQ: dump link backlog queue
 *           - TIPC_DUMP_DEFERDQ: dump link deferd queue
 *           - TIPC_DUMP_INPUTQ: dump link input queue
 *           - TIPC_DUMP_WAKEUP: dump link wakeup queue
 *           - TIPC_DUMP_ALL: dump all the link queues above
 * @buf: returned buffer of dump data in format
 */
int tipc_link_dump(struct tipc_link *l, u16 dqueues, char *buf)
{
	int i = 0;
	size_t sz = (dqueues) ? LINK_LMAX : LINK_LMIN;
	struct sk_buff_head *list;
	struct sk_buff *hskb, *tskb;
	u32 len;

	if (!l) {
		i += scnprintf(buf, sz, "link data: (null)\n");
		return i;
	}

	i += scnprintf(buf, sz, "link data: %x", l->addr);
	i += scnprintf(buf + i, sz - i, " %x", l->state);
	i += scnprintf(buf + i, sz - i, " %u", l->in_session);
	i += scnprintf(buf + i, sz - i, " %u", l->session);
	i += scnprintf(buf + i, sz - i, " %u", l->peer_session);
	i += scnprintf(buf + i, sz - i, " %u", l->snd_nxt);
	i += scnprintf(buf + i, sz - i, " %u", l->rcv_nxt);
	i += scnprintf(buf + i, sz - i, " %u", l->snd_nxt_state);
	i += scnprintf(buf + i, sz - i, " %u", l->rcv_nxt_state);
	i += scnprintf(buf + i, sz - i, " %x", l->peer_caps);
	i += scnprintf(buf + i, sz - i, " %u", l->silent_intv_cnt);
	i += scnprintf(buf + i, sz - i, " %u", l->rst_cnt);
	i += scnprintf(buf + i, sz - i, " %u", l->prev_from);
	i += scnprintf(buf + i, sz - i, " %u", l->stale_cnt);
	i += scnprintf(buf + i, sz - i, " %u", l->acked);

	list = &l->transmq;
	len = skb_queue_len(list);
	hskb = skb_peek(list);
	tskb = skb_peek_tail(list);
	i += scnprintf(buf + i, sz - i, " | %u %u %u", len,
		       (hskb) ? msg_seqno(buf_msg(hskb)) : 0,
		       (tskb) ? msg_seqno(buf_msg(tskb)) : 0);

	list = &l->deferdq;
	len = skb_queue_len(list);
	hskb = skb_peek(list);
	tskb = skb_peek_tail(list);
	i += scnprintf(buf + i, sz - i, " | %u %u %u", len,
		       (hskb) ? msg_seqno(buf_msg(hskb)) : 0,
		       (tskb) ? msg_seqno(buf_msg(tskb)) : 0);

	list = &l->backlogq;
	len = skb_queue_len(list);
	hskb = skb_peek(list);
	tskb = skb_peek_tail(list);
	i += scnprintf(buf + i, sz - i, " | %u %u %u", len,
		       (hskb) ? msg_seqno(buf_msg(hskb)) : 0,
		       (tskb) ? msg_seqno(buf_msg(tskb)) : 0);

	list = l->inputq;
	len = skb_queue_len(list);
	hskb = skb_peek(list);
	tskb = skb_peek_tail(list);
	i += scnprintf(buf + i, sz - i, " | %u %u %u\n", len,
		       (hskb) ? msg_seqno(buf_msg(hskb)) : 0,
		       (tskb) ? msg_seqno(buf_msg(tskb)) : 0);

	if (dqueues & TIPC_DUMP_TRANSMQ) {
		i += scnprintf(buf + i, sz - i, "transmq: ");
		i += tipc_list_dump(&l->transmq, false, buf + i);
	}
	if (dqueues & TIPC_DUMP_BACKLOGQ) {
		i += scnprintf(buf + i, sz - i,
			       "backlogq: <%u %u %u %u %u>, ",
			       l->backlog[TIPC_LOW_IMPORTANCE].len,
			       l->backlog[TIPC_MEDIUM_IMPORTANCE].len,
			       l->backlog[TIPC_HIGH_IMPORTANCE].len,
			       l->backlog[TIPC_CRITICAL_IMPORTANCE].len,
			       l->backlog[TIPC_SYSTEM_IMPORTANCE].len);
		i += tipc_list_dump(&l->backlogq, false, buf + i);
	}
	if (dqueues & TIPC_DUMP_DEFERDQ) {
		i += scnprintf(buf + i, sz - i, "deferdq: ");
		i += tipc_list_dump(&l->deferdq, false, buf + i);
	}
	if (dqueues & TIPC_DUMP_INPUTQ) {
		i += scnprintf(buf + i, sz - i, "inputq: ");
		i += tipc_list_dump(l->inputq, false, buf + i);
	}
	if (dqueues & TIPC_DUMP_WAKEUP) {
		i += scnprintf(buf + i, sz - i, "wakeup: ");
		i += tipc_list_dump(&l->wakeupq, false, buf + i);
	}

	return i;
}