xref: /openbmc/linux/net/tipc/socket.c (revision f519cd13)
1 /*
2  * net/tipc/socket.c: TIPC socket API
3  *
4  * Copyright (c) 2001-2007, 2012-2017, Ericsson AB
5  * Copyright (c) 2004-2008, 2010-2013, Wind River Systems
6  * All rights reserved.
7  *
8  * Redistribution and use in source and binary forms, with or without
9  * modification, are permitted provided that the following conditions are met:
10  *
11  * 1. Redistributions of source code must retain the above copyright
12  *    notice, this list of conditions and the following disclaimer.
13  * 2. Redistributions in binary form must reproduce the above copyright
14  *    notice, this list of conditions and the following disclaimer in the
15  *    documentation and/or other materials provided with the distribution.
16  * 3. Neither the names of the copyright holders nor the names of its
17  *    contributors may be used to endorse or promote products derived from
18  *    this software without specific prior written permission.
19  *
20  * Alternatively, this software may be distributed under the terms of the
21  * GNU General Public License ("GPL") version 2 as published by the Free
22  * Software Foundation.
23  *
24  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
25  * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
26  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
27  * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
28  * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
29  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
30  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
31  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
32  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
33  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
34  * POSSIBILITY OF SUCH DAMAGE.
35  */
36 
37 #include <linux/rhashtable.h>
38 #include <linux/sched/signal.h>
39 
40 #include "core.h"
41 #include "name_table.h"
42 #include "node.h"
43 #include "link.h"
44 #include "name_distr.h"
45 #include "socket.h"
46 #include "bcast.h"
47 #include "netlink.h"
48 #include "group.h"
49 #include "trace.h"
50 
51 #define CONN_TIMEOUT_DEFAULT    8000    /* default connect timeout = 8s */
52 #define CONN_PROBING_INTV	msecs_to_jiffies(3600000)  /* [ms] => 1 h */
53 #define TIPC_FWD_MSG		1
54 #define TIPC_MAX_PORT		0xffffffff
55 #define TIPC_MIN_PORT		1
56 #define TIPC_ACK_RATE		4       /* ACK at 1/4 of of rcv window size */
57 
58 enum {
59 	TIPC_LISTEN = TCP_LISTEN,
60 	TIPC_ESTABLISHED = TCP_ESTABLISHED,
61 	TIPC_OPEN = TCP_CLOSE,
62 	TIPC_DISCONNECTING = TCP_CLOSE_WAIT,
63 	TIPC_CONNECTING = TCP_SYN_SENT,
64 };
65 
66 struct sockaddr_pair {
67 	struct sockaddr_tipc sock;
68 	struct sockaddr_tipc member;
69 };
70 
71 /**
72  * struct tipc_sock - TIPC socket structure
73  * @sk: socket - interacts with 'port' and with user via the socket API
74  * @conn_type: TIPC type used when connection was established
75  * @conn_instance: TIPC instance used when connection was established
76  * @published: non-zero if port has one or more associated names
77  * @max_pkt: maximum packet size "hint" used when building messages sent by port
78  * @maxnagle: maximum size of msg which can be subject to nagle
79  * @portid: unique port identity in TIPC socket hash table
80  * @phdr: preformatted message header used when sending messages
81  * #cong_links: list of congested links
82  * @publications: list of publications for port
83  * @blocking_link: address of the congested link we are currently sleeping on
84  * @pub_count: total # of publications port has made during its lifetime
85  * @conn_timeout: the time we can wait for an unresponded setup request
86  * @dupl_rcvcnt: number of bytes counted twice, in both backlog and rcv queue
87  * @cong_link_cnt: number of congested links
88  * @snt_unacked: # messages sent by socket, and not yet acked by peer
89  * @rcv_unacked: # messages read by user, but not yet acked back to peer
90  * @peer: 'connected' peer for dgram/rdm
91  * @node: hash table node
92  * @mc_method: cookie for use between socket and broadcast layer
93  * @rcu: rcu struct for tipc_sock
94  */
95 struct tipc_sock {
96 	struct sock sk;
97 	u32 conn_type;
98 	u32 conn_instance;
99 	int published;
100 	u32 max_pkt;
101 	u32 maxnagle;
102 	u32 portid;
103 	struct tipc_msg phdr;
104 	struct list_head cong_links;
105 	struct list_head publications;
106 	u32 pub_count;
107 	atomic_t dupl_rcvcnt;
108 	u16 conn_timeout;
109 	bool probe_unacked;
110 	u16 cong_link_cnt;
111 	u16 snt_unacked;
112 	u16 snd_win;
113 	u16 peer_caps;
114 	u16 rcv_unacked;
115 	u16 rcv_win;
116 	struct sockaddr_tipc peer;
117 	struct rhash_head node;
118 	struct tipc_mc_method mc_method;
119 	struct rcu_head rcu;
120 	struct tipc_group *group;
121 	u32 oneway;
122 	u16 snd_backlog;
123 	bool expect_ack;
124 	bool nodelay;
125 	bool group_is_open;
126 };
127 
128 static int tipc_sk_backlog_rcv(struct sock *sk, struct sk_buff *skb);
129 static void tipc_data_ready(struct sock *sk);
130 static void tipc_write_space(struct sock *sk);
131 static void tipc_sock_destruct(struct sock *sk);
132 static int tipc_release(struct socket *sock);
133 static int tipc_accept(struct socket *sock, struct socket *new_sock, int flags,
134 		       bool kern);
135 static void tipc_sk_timeout(struct timer_list *t);
136 static int tipc_sk_publish(struct tipc_sock *tsk, uint scope,
137 			   struct tipc_name_seq const *seq);
138 static int tipc_sk_withdraw(struct tipc_sock *tsk, uint scope,
139 			    struct tipc_name_seq const *seq);
140 static int tipc_sk_leave(struct tipc_sock *tsk);
141 static struct tipc_sock *tipc_sk_lookup(struct net *net, u32 portid);
142 static int tipc_sk_insert(struct tipc_sock *tsk);
143 static void tipc_sk_remove(struct tipc_sock *tsk);
144 static int __tipc_sendstream(struct socket *sock, struct msghdr *m, size_t dsz);
145 static int __tipc_sendmsg(struct socket *sock, struct msghdr *m, size_t dsz);
146 static void tipc_sk_push_backlog(struct tipc_sock *tsk);
147 
148 static const struct proto_ops packet_ops;
149 static const struct proto_ops stream_ops;
150 static const struct proto_ops msg_ops;
151 static struct proto tipc_proto;
152 static const struct rhashtable_params tsk_rht_params;
153 
154 static u32 tsk_own_node(struct tipc_sock *tsk)
155 {
156 	return msg_prevnode(&tsk->phdr);
157 }
158 
159 static u32 tsk_peer_node(struct tipc_sock *tsk)
160 {
161 	return msg_destnode(&tsk->phdr);
162 }
163 
164 static u32 tsk_peer_port(struct tipc_sock *tsk)
165 {
166 	return msg_destport(&tsk->phdr);
167 }
168 
169 static  bool tsk_unreliable(struct tipc_sock *tsk)
170 {
171 	return msg_src_droppable(&tsk->phdr) != 0;
172 }
173 
174 static void tsk_set_unreliable(struct tipc_sock *tsk, bool unreliable)
175 {
176 	msg_set_src_droppable(&tsk->phdr, unreliable ? 1 : 0);
177 }
178 
179 static bool tsk_unreturnable(struct tipc_sock *tsk)
180 {
181 	return msg_dest_droppable(&tsk->phdr) != 0;
182 }
183 
184 static void tsk_set_unreturnable(struct tipc_sock *tsk, bool unreturnable)
185 {
186 	msg_set_dest_droppable(&tsk->phdr, unreturnable ? 1 : 0);
187 }
188 
189 static int tsk_importance(struct tipc_sock *tsk)
190 {
191 	return msg_importance(&tsk->phdr);
192 }
193 
194 static int tsk_set_importance(struct tipc_sock *tsk, int imp)
195 {
196 	if (imp > TIPC_CRITICAL_IMPORTANCE)
197 		return -EINVAL;
198 	msg_set_importance(&tsk->phdr, (u32)imp);
199 	return 0;
200 }
201 
202 static struct tipc_sock *tipc_sk(const struct sock *sk)
203 {
204 	return container_of(sk, struct tipc_sock, sk);
205 }
206 
207 static bool tsk_conn_cong(struct tipc_sock *tsk)
208 {
209 	return tsk->snt_unacked > tsk->snd_win;
210 }
211 
212 static u16 tsk_blocks(int len)
213 {
214 	return ((len / FLOWCTL_BLK_SZ) + 1);
215 }
216 
217 /* tsk_blocks(): translate a buffer size in bytes to number of
218  * advertisable blocks, taking into account the ratio truesize(len)/len
219  * We can trust that this ratio is always < 4 for len >= FLOWCTL_BLK_SZ
220  */
221 static u16 tsk_adv_blocks(int len)
222 {
223 	return len / FLOWCTL_BLK_SZ / 4;
224 }
225 
226 /* tsk_inc(): increment counter for sent or received data
227  * - If block based flow control is not supported by peer we
228  *   fall back to message based ditto, incrementing the counter
229  */
230 static u16 tsk_inc(struct tipc_sock *tsk, int msglen)
231 {
232 	if (likely(tsk->peer_caps & TIPC_BLOCK_FLOWCTL))
233 		return ((msglen / FLOWCTL_BLK_SZ) + 1);
234 	return 1;
235 }
236 
237 /* tsk_set_nagle - enable/disable nagle property by manipulating maxnagle
238  */
239 static void tsk_set_nagle(struct tipc_sock *tsk)
240 {
241 	struct sock *sk = &tsk->sk;
242 
243 	tsk->maxnagle = 0;
244 	if (sk->sk_type != SOCK_STREAM)
245 		return;
246 	if (tsk->nodelay)
247 		return;
248 	if (!(tsk->peer_caps & TIPC_NAGLE))
249 		return;
250 	/* Limit node local buffer size to avoid receive queue overflow */
251 	if (tsk->max_pkt == MAX_MSG_SIZE)
252 		tsk->maxnagle = 1500;
253 	else
254 		tsk->maxnagle = tsk->max_pkt;
255 }
256 
257 /**
258  * tsk_advance_rx_queue - discard first buffer in socket receive queue
259  *
260  * Caller must hold socket lock
261  */
262 static void tsk_advance_rx_queue(struct sock *sk)
263 {
264 	trace_tipc_sk_advance_rx(sk, NULL, TIPC_DUMP_SK_RCVQ, " ");
265 	kfree_skb(__skb_dequeue(&sk->sk_receive_queue));
266 }
267 
268 /* tipc_sk_respond() : send response message back to sender
269  */
270 static void tipc_sk_respond(struct sock *sk, struct sk_buff *skb, int err)
271 {
272 	u32 selector;
273 	u32 dnode;
274 	u32 onode = tipc_own_addr(sock_net(sk));
275 
276 	if (!tipc_msg_reverse(onode, &skb, err))
277 		return;
278 
279 	trace_tipc_sk_rej_msg(sk, skb, TIPC_DUMP_NONE, "@sk_respond!");
280 	dnode = msg_destnode(buf_msg(skb));
281 	selector = msg_origport(buf_msg(skb));
282 	tipc_node_xmit_skb(sock_net(sk), skb, dnode, selector);
283 }
284 
285 /**
286  * tsk_rej_rx_queue - reject all buffers in socket receive queue
287  *
288  * Caller must hold socket lock
289  */
290 static void tsk_rej_rx_queue(struct sock *sk, int error)
291 {
292 	struct sk_buff *skb;
293 
294 	while ((skb = __skb_dequeue(&sk->sk_receive_queue)))
295 		tipc_sk_respond(sk, skb, error);
296 }
297 
298 static bool tipc_sk_connected(struct sock *sk)
299 {
300 	return sk->sk_state == TIPC_ESTABLISHED;
301 }
302 
303 /* tipc_sk_type_connectionless - check if the socket is datagram socket
304  * @sk: socket
305  *
306  * Returns true if connection less, false otherwise
307  */
308 static bool tipc_sk_type_connectionless(struct sock *sk)
309 {
310 	return sk->sk_type == SOCK_RDM || sk->sk_type == SOCK_DGRAM;
311 }
312 
313 /* tsk_peer_msg - verify if message was sent by connected port's peer
314  *
315  * Handles cases where the node's network address has changed from
316  * the default of <0.0.0> to its configured setting.
317  */
318 static bool tsk_peer_msg(struct tipc_sock *tsk, struct tipc_msg *msg)
319 {
320 	struct sock *sk = &tsk->sk;
321 	u32 self = tipc_own_addr(sock_net(sk));
322 	u32 peer_port = tsk_peer_port(tsk);
323 	u32 orig_node, peer_node;
324 
325 	if (unlikely(!tipc_sk_connected(sk)))
326 		return false;
327 
328 	if (unlikely(msg_origport(msg) != peer_port))
329 		return false;
330 
331 	orig_node = msg_orignode(msg);
332 	peer_node = tsk_peer_node(tsk);
333 
334 	if (likely(orig_node == peer_node))
335 		return true;
336 
337 	if (!orig_node && peer_node == self)
338 		return true;
339 
340 	if (!peer_node && orig_node == self)
341 		return true;
342 
343 	return false;
344 }
345 
346 /* tipc_set_sk_state - set the sk_state of the socket
347  * @sk: socket
348  *
349  * Caller must hold socket lock
350  *
351  * Returns 0 on success, errno otherwise
352  */
353 static int tipc_set_sk_state(struct sock *sk, int state)
354 {
355 	int oldsk_state = sk->sk_state;
356 	int res = -EINVAL;
357 
358 	switch (state) {
359 	case TIPC_OPEN:
360 		res = 0;
361 		break;
362 	case TIPC_LISTEN:
363 	case TIPC_CONNECTING:
364 		if (oldsk_state == TIPC_OPEN)
365 			res = 0;
366 		break;
367 	case TIPC_ESTABLISHED:
368 		if (oldsk_state == TIPC_CONNECTING ||
369 		    oldsk_state == TIPC_OPEN)
370 			res = 0;
371 		break;
372 	case TIPC_DISCONNECTING:
373 		if (oldsk_state == TIPC_CONNECTING ||
374 		    oldsk_state == TIPC_ESTABLISHED)
375 			res = 0;
376 		break;
377 	}
378 
379 	if (!res)
380 		sk->sk_state = state;
381 
382 	return res;
383 }
384 
385 static int tipc_sk_sock_err(struct socket *sock, long *timeout)
386 {
387 	struct sock *sk = sock->sk;
388 	int err = sock_error(sk);
389 	int typ = sock->type;
390 
391 	if (err)
392 		return err;
393 	if (typ == SOCK_STREAM || typ == SOCK_SEQPACKET) {
394 		if (sk->sk_state == TIPC_DISCONNECTING)
395 			return -EPIPE;
396 		else if (!tipc_sk_connected(sk))
397 			return -ENOTCONN;
398 	}
399 	if (!*timeout)
400 		return -EAGAIN;
401 	if (signal_pending(current))
402 		return sock_intr_errno(*timeout);
403 
404 	return 0;
405 }
406 
407 #define tipc_wait_for_cond(sock_, timeo_, condition_)			       \
408 ({                                                                             \
409 	DEFINE_WAIT_FUNC(wait_, woken_wake_function);                          \
410 	struct sock *sk_;						       \
411 	int rc_;							       \
412 									       \
413 	while ((rc_ = !(condition_))) {					       \
414 		/* coupled with smp_wmb() in tipc_sk_proto_rcv() */            \
415 		smp_rmb();                                                     \
416 		sk_ = (sock_)->sk;					       \
417 		rc_ = tipc_sk_sock_err((sock_), timeo_);		       \
418 		if (rc_)						       \
419 			break;						       \
420 		add_wait_queue(sk_sleep(sk_), &wait_);                         \
421 		release_sock(sk_);					       \
422 		*(timeo_) = wait_woken(&wait_, TASK_INTERRUPTIBLE, *(timeo_)); \
423 		sched_annotate_sleep();				               \
424 		lock_sock(sk_);						       \
425 		remove_wait_queue(sk_sleep(sk_), &wait_);		       \
426 	}								       \
427 	rc_;								       \
428 })
429 
430 /**
431  * tipc_sk_create - create a TIPC socket
432  * @net: network namespace (must be default network)
433  * @sock: pre-allocated socket structure
434  * @protocol: protocol indicator (must be 0)
435  * @kern: caused by kernel or by userspace?
436  *
437  * This routine creates additional data structures used by the TIPC socket,
438  * initializes them, and links them together.
439  *
440  * Returns 0 on success, errno otherwise
441  */
442 static int tipc_sk_create(struct net *net, struct socket *sock,
443 			  int protocol, int kern)
444 {
445 	const struct proto_ops *ops;
446 	struct sock *sk;
447 	struct tipc_sock *tsk;
448 	struct tipc_msg *msg;
449 
450 	/* Validate arguments */
451 	if (unlikely(protocol != 0))
452 		return -EPROTONOSUPPORT;
453 
454 	switch (sock->type) {
455 	case SOCK_STREAM:
456 		ops = &stream_ops;
457 		break;
458 	case SOCK_SEQPACKET:
459 		ops = &packet_ops;
460 		break;
461 	case SOCK_DGRAM:
462 	case SOCK_RDM:
463 		ops = &msg_ops;
464 		break;
465 	default:
466 		return -EPROTOTYPE;
467 	}
468 
469 	/* Allocate socket's protocol area */
470 	sk = sk_alloc(net, AF_TIPC, GFP_KERNEL, &tipc_proto, kern);
471 	if (sk == NULL)
472 		return -ENOMEM;
473 
474 	tsk = tipc_sk(sk);
475 	tsk->max_pkt = MAX_PKT_DEFAULT;
476 	tsk->maxnagle = 0;
477 	INIT_LIST_HEAD(&tsk->publications);
478 	INIT_LIST_HEAD(&tsk->cong_links);
479 	msg = &tsk->phdr;
480 
481 	/* Finish initializing socket data structures */
482 	sock->ops = ops;
483 	sock_init_data(sock, sk);
484 	tipc_set_sk_state(sk, TIPC_OPEN);
485 	if (tipc_sk_insert(tsk)) {
486 		pr_warn("Socket create failed; port number exhausted\n");
487 		return -EINVAL;
488 	}
489 
490 	/* Ensure tsk is visible before we read own_addr. */
491 	smp_mb();
492 
493 	tipc_msg_init(tipc_own_addr(net), msg, TIPC_LOW_IMPORTANCE,
494 		      TIPC_NAMED_MSG, NAMED_H_SIZE, 0);
495 
496 	msg_set_origport(msg, tsk->portid);
497 	timer_setup(&sk->sk_timer, tipc_sk_timeout, 0);
498 	sk->sk_shutdown = 0;
499 	sk->sk_backlog_rcv = tipc_sk_backlog_rcv;
500 	sk->sk_rcvbuf = sysctl_tipc_rmem[1];
501 	sk->sk_data_ready = tipc_data_ready;
502 	sk->sk_write_space = tipc_write_space;
503 	sk->sk_destruct = tipc_sock_destruct;
504 	tsk->conn_timeout = CONN_TIMEOUT_DEFAULT;
505 	tsk->group_is_open = true;
506 	atomic_set(&tsk->dupl_rcvcnt, 0);
507 
508 	/* Start out with safe limits until we receive an advertised window */
509 	tsk->snd_win = tsk_adv_blocks(RCVBUF_MIN);
510 	tsk->rcv_win = tsk->snd_win;
511 
512 	if (tipc_sk_type_connectionless(sk)) {
513 		tsk_set_unreturnable(tsk, true);
514 		if (sock->type == SOCK_DGRAM)
515 			tsk_set_unreliable(tsk, true);
516 	}
517 	__skb_queue_head_init(&tsk->mc_method.deferredq);
518 	trace_tipc_sk_create(sk, NULL, TIPC_DUMP_NONE, " ");
519 	return 0;
520 }
521 
522 static void tipc_sk_callback(struct rcu_head *head)
523 {
524 	struct tipc_sock *tsk = container_of(head, struct tipc_sock, rcu);
525 
526 	sock_put(&tsk->sk);
527 }
528 
529 /* Caller should hold socket lock for the socket. */
530 static void __tipc_shutdown(struct socket *sock, int error)
531 {
532 	struct sock *sk = sock->sk;
533 	struct tipc_sock *tsk = tipc_sk(sk);
534 	struct net *net = sock_net(sk);
535 	long timeout = msecs_to_jiffies(CONN_TIMEOUT_DEFAULT);
536 	u32 dnode = tsk_peer_node(tsk);
537 	struct sk_buff *skb;
538 
539 	/* Avoid that hi-prio shutdown msgs bypass msgs in link wakeup queue */
540 	tipc_wait_for_cond(sock, &timeout, (!tsk->cong_link_cnt &&
541 					    !tsk_conn_cong(tsk)));
542 
543 	/* Push out delayed messages if in Nagle mode */
544 	tipc_sk_push_backlog(tsk);
545 	/* Remove pending SYN */
546 	__skb_queue_purge(&sk->sk_write_queue);
547 
548 	/* Remove partially received buffer if any */
549 	skb = skb_peek(&sk->sk_receive_queue);
550 	if (skb && TIPC_SKB_CB(skb)->bytes_read) {
551 		__skb_unlink(skb, &sk->sk_receive_queue);
552 		kfree_skb(skb);
553 	}
554 
555 	/* Reject all unreceived messages if connectionless */
556 	if (tipc_sk_type_connectionless(sk)) {
557 		tsk_rej_rx_queue(sk, error);
558 		return;
559 	}
560 
561 	switch (sk->sk_state) {
562 	case TIPC_CONNECTING:
563 	case TIPC_ESTABLISHED:
564 		tipc_set_sk_state(sk, TIPC_DISCONNECTING);
565 		tipc_node_remove_conn(net, dnode, tsk->portid);
566 		/* Send a FIN+/- to its peer */
567 		skb = __skb_dequeue(&sk->sk_receive_queue);
568 		if (skb) {
569 			__skb_queue_purge(&sk->sk_receive_queue);
570 			tipc_sk_respond(sk, skb, error);
571 			break;
572 		}
573 		skb = tipc_msg_create(TIPC_CRITICAL_IMPORTANCE,
574 				      TIPC_CONN_MSG, SHORT_H_SIZE, 0, dnode,
575 				      tsk_own_node(tsk), tsk_peer_port(tsk),
576 				      tsk->portid, error);
577 		if (skb)
578 			tipc_node_xmit_skb(net, skb, dnode, tsk->portid);
579 		break;
580 	case TIPC_LISTEN:
581 		/* Reject all SYN messages */
582 		tsk_rej_rx_queue(sk, error);
583 		break;
584 	default:
585 		__skb_queue_purge(&sk->sk_receive_queue);
586 		break;
587 	}
588 }
589 
590 /**
591  * tipc_release - destroy a TIPC socket
592  * @sock: socket to destroy
593  *
594  * This routine cleans up any messages that are still queued on the socket.
595  * For DGRAM and RDM socket types, all queued messages are rejected.
596  * For SEQPACKET and STREAM socket types, the first message is rejected
597  * and any others are discarded.  (If the first message on a STREAM socket
598  * is partially-read, it is discarded and the next one is rejected instead.)
599  *
600  * NOTE: Rejected messages are not necessarily returned to the sender!  They
601  * are returned or discarded according to the "destination droppable" setting
602  * specified for the message by the sender.
603  *
604  * Returns 0 on success, errno otherwise
605  */
606 static int tipc_release(struct socket *sock)
607 {
608 	struct sock *sk = sock->sk;
609 	struct tipc_sock *tsk;
610 
611 	/*
612 	 * Exit if socket isn't fully initialized (occurs when a failed accept()
613 	 * releases a pre-allocated child socket that was never used)
614 	 */
615 	if (sk == NULL)
616 		return 0;
617 
618 	tsk = tipc_sk(sk);
619 	lock_sock(sk);
620 
621 	trace_tipc_sk_release(sk, NULL, TIPC_DUMP_ALL, " ");
622 	__tipc_shutdown(sock, TIPC_ERR_NO_PORT);
623 	sk->sk_shutdown = SHUTDOWN_MASK;
624 	tipc_sk_leave(tsk);
625 	tipc_sk_withdraw(tsk, 0, NULL);
626 	__skb_queue_purge(&tsk->mc_method.deferredq);
627 	sk_stop_timer(sk, &sk->sk_timer);
628 	tipc_sk_remove(tsk);
629 
630 	sock_orphan(sk);
631 	/* Reject any messages that accumulated in backlog queue */
632 	release_sock(sk);
633 	tipc_dest_list_purge(&tsk->cong_links);
634 	tsk->cong_link_cnt = 0;
635 	call_rcu(&tsk->rcu, tipc_sk_callback);
636 	sock->sk = NULL;
637 
638 	return 0;
639 }
640 
641 /**
642  * tipc_bind - associate or disassocate TIPC name(s) with a socket
643  * @sock: socket structure
644  * @uaddr: socket address describing name(s) and desired operation
645  * @uaddr_len: size of socket address data structure
646  *
647  * Name and name sequence binding is indicated using a positive scope value;
648  * a negative scope value unbinds the specified name.  Specifying no name
649  * (i.e. a socket address length of 0) unbinds all names from the socket.
650  *
651  * Returns 0 on success, errno otherwise
652  *
653  * NOTE: This routine doesn't need to take the socket lock since it doesn't
654  *       access any non-constant socket information.
655  */
656 static int tipc_bind(struct socket *sock, struct sockaddr *uaddr,
657 		     int uaddr_len)
658 {
659 	struct sock *sk = sock->sk;
660 	struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr;
661 	struct tipc_sock *tsk = tipc_sk(sk);
662 	int res = -EINVAL;
663 
664 	lock_sock(sk);
665 	if (unlikely(!uaddr_len)) {
666 		res = tipc_sk_withdraw(tsk, 0, NULL);
667 		goto exit;
668 	}
669 	if (tsk->group) {
670 		res = -EACCES;
671 		goto exit;
672 	}
673 	if (uaddr_len < sizeof(struct sockaddr_tipc)) {
674 		res = -EINVAL;
675 		goto exit;
676 	}
677 	if (addr->family != AF_TIPC) {
678 		res = -EAFNOSUPPORT;
679 		goto exit;
680 	}
681 
682 	if (addr->addrtype == TIPC_ADDR_NAME)
683 		addr->addr.nameseq.upper = addr->addr.nameseq.lower;
684 	else if (addr->addrtype != TIPC_ADDR_NAMESEQ) {
685 		res = -EAFNOSUPPORT;
686 		goto exit;
687 	}
688 
689 	if ((addr->addr.nameseq.type < TIPC_RESERVED_TYPES) &&
690 	    (addr->addr.nameseq.type != TIPC_TOP_SRV) &&
691 	    (addr->addr.nameseq.type != TIPC_CFG_SRV)) {
692 		res = -EACCES;
693 		goto exit;
694 	}
695 
696 	res = (addr->scope >= 0) ?
697 		tipc_sk_publish(tsk, addr->scope, &addr->addr.nameseq) :
698 		tipc_sk_withdraw(tsk, -addr->scope, &addr->addr.nameseq);
699 exit:
700 	release_sock(sk);
701 	return res;
702 }
703 
704 /**
705  * tipc_getname - get port ID of socket or peer socket
706  * @sock: socket structure
707  * @uaddr: area for returned socket address
708  * @uaddr_len: area for returned length of socket address
709  * @peer: 0 = own ID, 1 = current peer ID, 2 = current/former peer ID
710  *
711  * Returns 0 on success, errno otherwise
712  *
713  * NOTE: This routine doesn't need to take the socket lock since it only
714  *       accesses socket information that is unchanging (or which changes in
715  *       a completely predictable manner).
716  */
717 static int tipc_getname(struct socket *sock, struct sockaddr *uaddr,
718 			int peer)
719 {
720 	struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr;
721 	struct sock *sk = sock->sk;
722 	struct tipc_sock *tsk = tipc_sk(sk);
723 
724 	memset(addr, 0, sizeof(*addr));
725 	if (peer) {
726 		if ((!tipc_sk_connected(sk)) &&
727 		    ((peer != 2) || (sk->sk_state != TIPC_DISCONNECTING)))
728 			return -ENOTCONN;
729 		addr->addr.id.ref = tsk_peer_port(tsk);
730 		addr->addr.id.node = tsk_peer_node(tsk);
731 	} else {
732 		addr->addr.id.ref = tsk->portid;
733 		addr->addr.id.node = tipc_own_addr(sock_net(sk));
734 	}
735 
736 	addr->addrtype = TIPC_ADDR_ID;
737 	addr->family = AF_TIPC;
738 	addr->scope = 0;
739 	addr->addr.name.domain = 0;
740 
741 	return sizeof(*addr);
742 }
743 
744 /**
745  * tipc_poll - read and possibly block on pollmask
746  * @file: file structure associated with the socket
747  * @sock: socket for which to calculate the poll bits
748  * @wait: ???
749  *
750  * Returns pollmask value
751  *
752  * COMMENTARY:
753  * It appears that the usual socket locking mechanisms are not useful here
754  * since the pollmask info is potentially out-of-date the moment this routine
755  * exits.  TCP and other protocols seem to rely on higher level poll routines
756  * to handle any preventable race conditions, so TIPC will do the same ...
757  *
758  * IMPORTANT: The fact that a read or write operation is indicated does NOT
759  * imply that the operation will succeed, merely that it should be performed
760  * and will not block.
761  */
762 static __poll_t tipc_poll(struct file *file, struct socket *sock,
763 			      poll_table *wait)
764 {
765 	struct sock *sk = sock->sk;
766 	struct tipc_sock *tsk = tipc_sk(sk);
767 	__poll_t revents = 0;
768 
769 	sock_poll_wait(file, sock, wait);
770 	trace_tipc_sk_poll(sk, NULL, TIPC_DUMP_ALL, " ");
771 
772 	if (sk->sk_shutdown & RCV_SHUTDOWN)
773 		revents |= EPOLLRDHUP | EPOLLIN | EPOLLRDNORM;
774 	if (sk->sk_shutdown == SHUTDOWN_MASK)
775 		revents |= EPOLLHUP;
776 
777 	switch (sk->sk_state) {
778 	case TIPC_ESTABLISHED:
779 		if (!tsk->cong_link_cnt && !tsk_conn_cong(tsk))
780 			revents |= EPOLLOUT;
781 		/* fall through */
782 	case TIPC_LISTEN:
783 	case TIPC_CONNECTING:
784 		if (!skb_queue_empty_lockless(&sk->sk_receive_queue))
785 			revents |= EPOLLIN | EPOLLRDNORM;
786 		break;
787 	case TIPC_OPEN:
788 		if (tsk->group_is_open && !tsk->cong_link_cnt)
789 			revents |= EPOLLOUT;
790 		if (!tipc_sk_type_connectionless(sk))
791 			break;
792 		if (skb_queue_empty_lockless(&sk->sk_receive_queue))
793 			break;
794 		revents |= EPOLLIN | EPOLLRDNORM;
795 		break;
796 	case TIPC_DISCONNECTING:
797 		revents = EPOLLIN | EPOLLRDNORM | EPOLLHUP;
798 		break;
799 	}
800 	return revents;
801 }
802 
803 /**
804  * tipc_sendmcast - send multicast message
805  * @sock: socket structure
806  * @seq: destination address
807  * @msg: message to send
808  * @dlen: length of data to send
809  * @timeout: timeout to wait for wakeup
810  *
811  * Called from function tipc_sendmsg(), which has done all sanity checks
812  * Returns the number of bytes sent on success, or errno
813  */
814 static int tipc_sendmcast(struct  socket *sock, struct tipc_name_seq *seq,
815 			  struct msghdr *msg, size_t dlen, long timeout)
816 {
817 	struct sock *sk = sock->sk;
818 	struct tipc_sock *tsk = tipc_sk(sk);
819 	struct tipc_msg *hdr = &tsk->phdr;
820 	struct net *net = sock_net(sk);
821 	int mtu = tipc_bcast_get_mtu(net);
822 	struct tipc_mc_method *method = &tsk->mc_method;
823 	struct sk_buff_head pkts;
824 	struct tipc_nlist dsts;
825 	int rc;
826 
827 	if (tsk->group)
828 		return -EACCES;
829 
830 	/* Block or return if any destination link is congested */
831 	rc = tipc_wait_for_cond(sock, &timeout, !tsk->cong_link_cnt);
832 	if (unlikely(rc))
833 		return rc;
834 
835 	/* Lookup destination nodes */
836 	tipc_nlist_init(&dsts, tipc_own_addr(net));
837 	tipc_nametbl_lookup_dst_nodes(net, seq->type, seq->lower,
838 				      seq->upper, &dsts);
839 	if (!dsts.local && !dsts.remote)
840 		return -EHOSTUNREACH;
841 
842 	/* Build message header */
843 	msg_set_type(hdr, TIPC_MCAST_MSG);
844 	msg_set_hdr_sz(hdr, MCAST_H_SIZE);
845 	msg_set_lookup_scope(hdr, TIPC_CLUSTER_SCOPE);
846 	msg_set_destport(hdr, 0);
847 	msg_set_destnode(hdr, 0);
848 	msg_set_nametype(hdr, seq->type);
849 	msg_set_namelower(hdr, seq->lower);
850 	msg_set_nameupper(hdr, seq->upper);
851 
852 	/* Build message as chain of buffers */
853 	__skb_queue_head_init(&pkts);
854 	rc = tipc_msg_build(hdr, msg, 0, dlen, mtu, &pkts);
855 
856 	/* Send message if build was successful */
857 	if (unlikely(rc == dlen)) {
858 		trace_tipc_sk_sendmcast(sk, skb_peek(&pkts),
859 					TIPC_DUMP_SK_SNDQ, " ");
860 		rc = tipc_mcast_xmit(net, &pkts, method, &dsts,
861 				     &tsk->cong_link_cnt);
862 	}
863 
864 	tipc_nlist_purge(&dsts);
865 
866 	return rc ? rc : dlen;
867 }
868 
869 /**
870  * tipc_send_group_msg - send a message to a member in the group
871  * @net: network namespace
872  * @m: message to send
873  * @mb: group member
874  * @dnode: destination node
875  * @dport: destination port
876  * @dlen: total length of message data
877  */
878 static int tipc_send_group_msg(struct net *net, struct tipc_sock *tsk,
879 			       struct msghdr *m, struct tipc_member *mb,
880 			       u32 dnode, u32 dport, int dlen)
881 {
882 	u16 bc_snd_nxt = tipc_group_bc_snd_nxt(tsk->group);
883 	struct tipc_mc_method *method = &tsk->mc_method;
884 	int blks = tsk_blocks(GROUP_H_SIZE + dlen);
885 	struct tipc_msg *hdr = &tsk->phdr;
886 	struct sk_buff_head pkts;
887 	int mtu, rc;
888 
889 	/* Complete message header */
890 	msg_set_type(hdr, TIPC_GRP_UCAST_MSG);
891 	msg_set_hdr_sz(hdr, GROUP_H_SIZE);
892 	msg_set_destport(hdr, dport);
893 	msg_set_destnode(hdr, dnode);
894 	msg_set_grp_bc_seqno(hdr, bc_snd_nxt);
895 
896 	/* Build message as chain of buffers */
897 	__skb_queue_head_init(&pkts);
898 	mtu = tipc_node_get_mtu(net, dnode, tsk->portid, false);
899 	rc = tipc_msg_build(hdr, m, 0, dlen, mtu, &pkts);
900 	if (unlikely(rc != dlen))
901 		return rc;
902 
903 	/* Send message */
904 	rc = tipc_node_xmit(net, &pkts, dnode, tsk->portid);
905 	if (unlikely(rc == -ELINKCONG)) {
906 		tipc_dest_push(&tsk->cong_links, dnode, 0);
907 		tsk->cong_link_cnt++;
908 	}
909 
910 	/* Update send window */
911 	tipc_group_update_member(mb, blks);
912 
913 	/* A broadcast sent within next EXPIRE period must follow same path */
914 	method->rcast = true;
915 	method->mandatory = true;
916 	return dlen;
917 }
918 
919 /**
920  * tipc_send_group_unicast - send message to a member in the group
921  * @sock: socket structure
922  * @m: message to send
923  * @dlen: total length of message data
924  * @timeout: timeout to wait for wakeup
925  *
926  * Called from function tipc_sendmsg(), which has done all sanity checks
927  * Returns the number of bytes sent on success, or errno
928  */
929 static int tipc_send_group_unicast(struct socket *sock, struct msghdr *m,
930 				   int dlen, long timeout)
931 {
932 	struct sock *sk = sock->sk;
933 	DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name);
934 	int blks = tsk_blocks(GROUP_H_SIZE + dlen);
935 	struct tipc_sock *tsk = tipc_sk(sk);
936 	struct net *net = sock_net(sk);
937 	struct tipc_member *mb = NULL;
938 	u32 node, port;
939 	int rc;
940 
941 	node = dest->addr.id.node;
942 	port = dest->addr.id.ref;
943 	if (!port && !node)
944 		return -EHOSTUNREACH;
945 
946 	/* Block or return if destination link or member is congested */
947 	rc = tipc_wait_for_cond(sock, &timeout,
948 				!tipc_dest_find(&tsk->cong_links, node, 0) &&
949 				tsk->group &&
950 				!tipc_group_cong(tsk->group, node, port, blks,
951 						 &mb));
952 	if (unlikely(rc))
953 		return rc;
954 
955 	if (unlikely(!mb))
956 		return -EHOSTUNREACH;
957 
958 	rc = tipc_send_group_msg(net, tsk, m, mb, node, port, dlen);
959 
960 	return rc ? rc : dlen;
961 }
962 
963 /**
964  * tipc_send_group_anycast - send message to any member with given identity
965  * @sock: socket structure
966  * @m: message to send
967  * @dlen: total length of message data
968  * @timeout: timeout to wait for wakeup
969  *
970  * Called from function tipc_sendmsg(), which has done all sanity checks
971  * Returns the number of bytes sent on success, or errno
972  */
973 static int tipc_send_group_anycast(struct socket *sock, struct msghdr *m,
974 				   int dlen, long timeout)
975 {
976 	DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name);
977 	struct sock *sk = sock->sk;
978 	struct tipc_sock *tsk = tipc_sk(sk);
979 	struct list_head *cong_links = &tsk->cong_links;
980 	int blks = tsk_blocks(GROUP_H_SIZE + dlen);
981 	struct tipc_msg *hdr = &tsk->phdr;
982 	struct tipc_member *first = NULL;
983 	struct tipc_member *mbr = NULL;
984 	struct net *net = sock_net(sk);
985 	u32 node, port, exclude;
986 	struct list_head dsts;
987 	u32 type, inst, scope;
988 	int lookups = 0;
989 	int dstcnt, rc;
990 	bool cong;
991 
992 	INIT_LIST_HEAD(&dsts);
993 
994 	type = msg_nametype(hdr);
995 	inst = dest->addr.name.name.instance;
996 	scope = msg_lookup_scope(hdr);
997 
998 	while (++lookups < 4) {
999 		exclude = tipc_group_exclude(tsk->group);
1000 
1001 		first = NULL;
1002 
1003 		/* Look for a non-congested destination member, if any */
1004 		while (1) {
1005 			if (!tipc_nametbl_lookup(net, type, inst, scope, &dsts,
1006 						 &dstcnt, exclude, false))
1007 				return -EHOSTUNREACH;
1008 			tipc_dest_pop(&dsts, &node, &port);
1009 			cong = tipc_group_cong(tsk->group, node, port, blks,
1010 					       &mbr);
1011 			if (!cong)
1012 				break;
1013 			if (mbr == first)
1014 				break;
1015 			if (!first)
1016 				first = mbr;
1017 		}
1018 
1019 		/* Start over if destination was not in member list */
1020 		if (unlikely(!mbr))
1021 			continue;
1022 
1023 		if (likely(!cong && !tipc_dest_find(cong_links, node, 0)))
1024 			break;
1025 
1026 		/* Block or return if destination link or member is congested */
1027 		rc = tipc_wait_for_cond(sock, &timeout,
1028 					!tipc_dest_find(cong_links, node, 0) &&
1029 					tsk->group &&
1030 					!tipc_group_cong(tsk->group, node, port,
1031 							 blks, &mbr));
1032 		if (unlikely(rc))
1033 			return rc;
1034 
1035 		/* Send, unless destination disappeared while waiting */
1036 		if (likely(mbr))
1037 			break;
1038 	}
1039 
1040 	if (unlikely(lookups >= 4))
1041 		return -EHOSTUNREACH;
1042 
1043 	rc = tipc_send_group_msg(net, tsk, m, mbr, node, port, dlen);
1044 
1045 	return rc ? rc : dlen;
1046 }
1047 
1048 /**
1049  * tipc_send_group_bcast - send message to all members in communication group
1050  * @sk: socket structure
1051  * @m: message to send
1052  * @dlen: total length of message data
1053  * @timeout: timeout to wait for wakeup
1054  *
1055  * Called from function tipc_sendmsg(), which has done all sanity checks
1056  * Returns the number of bytes sent on success, or errno
1057  */
1058 static int tipc_send_group_bcast(struct socket *sock, struct msghdr *m,
1059 				 int dlen, long timeout)
1060 {
1061 	DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name);
1062 	struct sock *sk = sock->sk;
1063 	struct net *net = sock_net(sk);
1064 	struct tipc_sock *tsk = tipc_sk(sk);
1065 	struct tipc_nlist *dsts;
1066 	struct tipc_mc_method *method = &tsk->mc_method;
1067 	bool ack = method->mandatory && method->rcast;
1068 	int blks = tsk_blocks(MCAST_H_SIZE + dlen);
1069 	struct tipc_msg *hdr = &tsk->phdr;
1070 	int mtu = tipc_bcast_get_mtu(net);
1071 	struct sk_buff_head pkts;
1072 	int rc = -EHOSTUNREACH;
1073 
1074 	/* Block or return if any destination link or member is congested */
1075 	rc = tipc_wait_for_cond(sock, &timeout,
1076 				!tsk->cong_link_cnt && tsk->group &&
1077 				!tipc_group_bc_cong(tsk->group, blks));
1078 	if (unlikely(rc))
1079 		return rc;
1080 
1081 	dsts = tipc_group_dests(tsk->group);
1082 	if (!dsts->local && !dsts->remote)
1083 		return -EHOSTUNREACH;
1084 
1085 	/* Complete message header */
1086 	if (dest) {
1087 		msg_set_type(hdr, TIPC_GRP_MCAST_MSG);
1088 		msg_set_nameinst(hdr, dest->addr.name.name.instance);
1089 	} else {
1090 		msg_set_type(hdr, TIPC_GRP_BCAST_MSG);
1091 		msg_set_nameinst(hdr, 0);
1092 	}
1093 	msg_set_hdr_sz(hdr, GROUP_H_SIZE);
1094 	msg_set_destport(hdr, 0);
1095 	msg_set_destnode(hdr, 0);
1096 	msg_set_grp_bc_seqno(hdr, tipc_group_bc_snd_nxt(tsk->group));
1097 
1098 	/* Avoid getting stuck with repeated forced replicasts */
1099 	msg_set_grp_bc_ack_req(hdr, ack);
1100 
1101 	/* Build message as chain of buffers */
1102 	__skb_queue_head_init(&pkts);
1103 	rc = tipc_msg_build(hdr, m, 0, dlen, mtu, &pkts);
1104 	if (unlikely(rc != dlen))
1105 		return rc;
1106 
1107 	/* Send message */
1108 	rc = tipc_mcast_xmit(net, &pkts, method, dsts, &tsk->cong_link_cnt);
1109 	if (unlikely(rc))
1110 		return rc;
1111 
1112 	/* Update broadcast sequence number and send windows */
1113 	tipc_group_update_bc_members(tsk->group, blks, ack);
1114 
1115 	/* Broadcast link is now free to choose method for next broadcast */
1116 	method->mandatory = false;
1117 	method->expires = jiffies;
1118 
1119 	return dlen;
1120 }
1121 
1122 /**
1123  * tipc_send_group_mcast - send message to all members with given identity
1124  * @sock: socket structure
1125  * @m: message to send
1126  * @dlen: total length of message data
1127  * @timeout: timeout to wait for wakeup
1128  *
1129  * Called from function tipc_sendmsg(), which has done all sanity checks
1130  * Returns the number of bytes sent on success, or errno
1131  */
1132 static int tipc_send_group_mcast(struct socket *sock, struct msghdr *m,
1133 				 int dlen, long timeout)
1134 {
1135 	struct sock *sk = sock->sk;
1136 	DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name);
1137 	struct tipc_sock *tsk = tipc_sk(sk);
1138 	struct tipc_group *grp = tsk->group;
1139 	struct tipc_msg *hdr = &tsk->phdr;
1140 	struct net *net = sock_net(sk);
1141 	u32 type, inst, scope, exclude;
1142 	struct list_head dsts;
1143 	u32 dstcnt;
1144 
1145 	INIT_LIST_HEAD(&dsts);
1146 
1147 	type = msg_nametype(hdr);
1148 	inst = dest->addr.name.name.instance;
1149 	scope = msg_lookup_scope(hdr);
1150 	exclude = tipc_group_exclude(grp);
1151 
1152 	if (!tipc_nametbl_lookup(net, type, inst, scope, &dsts,
1153 				 &dstcnt, exclude, true))
1154 		return -EHOSTUNREACH;
1155 
1156 	if (dstcnt == 1) {
1157 		tipc_dest_pop(&dsts, &dest->addr.id.node, &dest->addr.id.ref);
1158 		return tipc_send_group_unicast(sock, m, dlen, timeout);
1159 	}
1160 
1161 	tipc_dest_list_purge(&dsts);
1162 	return tipc_send_group_bcast(sock, m, dlen, timeout);
1163 }
1164 
1165 /**
1166  * tipc_sk_mcast_rcv - Deliver multicast messages to all destination sockets
1167  * @arrvq: queue with arriving messages, to be cloned after destination lookup
1168  * @inputq: queue with cloned messages, delivered to socket after dest lookup
1169  *
1170  * Multi-threaded: parallel calls with reference to same queues may occur
1171  */
1172 void tipc_sk_mcast_rcv(struct net *net, struct sk_buff_head *arrvq,
1173 		       struct sk_buff_head *inputq)
1174 {
1175 	u32 self = tipc_own_addr(net);
1176 	u32 type, lower, upper, scope;
1177 	struct sk_buff *skb, *_skb;
1178 	u32 portid, onode;
1179 	struct sk_buff_head tmpq;
1180 	struct list_head dports;
1181 	struct tipc_msg *hdr;
1182 	int user, mtyp, hlen;
1183 	bool exact;
1184 
1185 	__skb_queue_head_init(&tmpq);
1186 	INIT_LIST_HEAD(&dports);
1187 
1188 	skb = tipc_skb_peek(arrvq, &inputq->lock);
1189 	for (; skb; skb = tipc_skb_peek(arrvq, &inputq->lock)) {
1190 		hdr = buf_msg(skb);
1191 		user = msg_user(hdr);
1192 		mtyp = msg_type(hdr);
1193 		hlen = skb_headroom(skb) + msg_hdr_sz(hdr);
1194 		onode = msg_orignode(hdr);
1195 		type = msg_nametype(hdr);
1196 
1197 		if (mtyp == TIPC_GRP_UCAST_MSG || user == GROUP_PROTOCOL) {
1198 			spin_lock_bh(&inputq->lock);
1199 			if (skb_peek(arrvq) == skb) {
1200 				__skb_dequeue(arrvq);
1201 				__skb_queue_tail(inputq, skb);
1202 			}
1203 			kfree_skb(skb);
1204 			spin_unlock_bh(&inputq->lock);
1205 			continue;
1206 		}
1207 
1208 		/* Group messages require exact scope match */
1209 		if (msg_in_group(hdr)) {
1210 			lower = 0;
1211 			upper = ~0;
1212 			scope = msg_lookup_scope(hdr);
1213 			exact = true;
1214 		} else {
1215 			/* TIPC_NODE_SCOPE means "any scope" in this context */
1216 			if (onode == self)
1217 				scope = TIPC_NODE_SCOPE;
1218 			else
1219 				scope = TIPC_CLUSTER_SCOPE;
1220 			exact = false;
1221 			lower = msg_namelower(hdr);
1222 			upper = msg_nameupper(hdr);
1223 		}
1224 
1225 		/* Create destination port list: */
1226 		tipc_nametbl_mc_lookup(net, type, lower, upper,
1227 				       scope, exact, &dports);
1228 
1229 		/* Clone message per destination */
1230 		while (tipc_dest_pop(&dports, NULL, &portid)) {
1231 			_skb = __pskb_copy(skb, hlen, GFP_ATOMIC);
1232 			if (_skb) {
1233 				msg_set_destport(buf_msg(_skb), portid);
1234 				__skb_queue_tail(&tmpq, _skb);
1235 				continue;
1236 			}
1237 			pr_warn("Failed to clone mcast rcv buffer\n");
1238 		}
1239 		/* Append to inputq if not already done by other thread */
1240 		spin_lock_bh(&inputq->lock);
1241 		if (skb_peek(arrvq) == skb) {
1242 			skb_queue_splice_tail_init(&tmpq, inputq);
1243 			kfree_skb(__skb_dequeue(arrvq));
1244 		}
1245 		spin_unlock_bh(&inputq->lock);
1246 		__skb_queue_purge(&tmpq);
1247 		kfree_skb(skb);
1248 	}
1249 	tipc_sk_rcv(net, inputq);
1250 }
1251 
1252 /* tipc_sk_push_backlog(): send accumulated buffers in socket write queue
1253  *                         when socket is in Nagle mode
1254  */
1255 static void tipc_sk_push_backlog(struct tipc_sock *tsk)
1256 {
1257 	struct sk_buff_head *txq = &tsk->sk.sk_write_queue;
1258 	struct net *net = sock_net(&tsk->sk);
1259 	u32 dnode = tsk_peer_node(tsk);
1260 	struct sk_buff *skb = skb_peek(txq);
1261 	int rc;
1262 
1263 	if (!skb || tsk->cong_link_cnt)
1264 		return;
1265 
1266 	/* Do not send SYN again after congestion */
1267 	if (msg_is_syn(buf_msg(skb)))
1268 		return;
1269 
1270 	tsk->snt_unacked += tsk->snd_backlog;
1271 	tsk->snd_backlog = 0;
1272 	tsk->expect_ack = true;
1273 	rc = tipc_node_xmit(net, txq, dnode, tsk->portid);
1274 	if (rc == -ELINKCONG)
1275 		tsk->cong_link_cnt = 1;
1276 }
1277 
1278 /**
1279  * tipc_sk_conn_proto_rcv - receive a connection mng protocol message
1280  * @tsk: receiving socket
1281  * @skb: pointer to message buffer.
1282  */
1283 static void tipc_sk_conn_proto_rcv(struct tipc_sock *tsk, struct sk_buff *skb,
1284 				   struct sk_buff_head *inputq,
1285 				   struct sk_buff_head *xmitq)
1286 {
1287 	struct tipc_msg *hdr = buf_msg(skb);
1288 	u32 onode = tsk_own_node(tsk);
1289 	struct sock *sk = &tsk->sk;
1290 	int mtyp = msg_type(hdr);
1291 	bool was_cong;
1292 
1293 	/* Ignore if connection cannot be validated: */
1294 	if (!tsk_peer_msg(tsk, hdr)) {
1295 		trace_tipc_sk_drop_msg(sk, skb, TIPC_DUMP_NONE, "@proto_rcv!");
1296 		goto exit;
1297 	}
1298 
1299 	if (unlikely(msg_errcode(hdr))) {
1300 		tipc_set_sk_state(sk, TIPC_DISCONNECTING);
1301 		tipc_node_remove_conn(sock_net(sk), tsk_peer_node(tsk),
1302 				      tsk_peer_port(tsk));
1303 		sk->sk_state_change(sk);
1304 
1305 		/* State change is ignored if socket already awake,
1306 		 * - convert msg to abort msg and add to inqueue
1307 		 */
1308 		msg_set_user(hdr, TIPC_CRITICAL_IMPORTANCE);
1309 		msg_set_type(hdr, TIPC_CONN_MSG);
1310 		msg_set_size(hdr, BASIC_H_SIZE);
1311 		msg_set_hdr_sz(hdr, BASIC_H_SIZE);
1312 		__skb_queue_tail(inputq, skb);
1313 		return;
1314 	}
1315 
1316 	tsk->probe_unacked = false;
1317 
1318 	if (mtyp == CONN_PROBE) {
1319 		msg_set_type(hdr, CONN_PROBE_REPLY);
1320 		if (tipc_msg_reverse(onode, &skb, TIPC_OK))
1321 			__skb_queue_tail(xmitq, skb);
1322 		return;
1323 	} else if (mtyp == CONN_ACK) {
1324 		was_cong = tsk_conn_cong(tsk);
1325 		tsk->expect_ack = false;
1326 		tipc_sk_push_backlog(tsk);
1327 		tsk->snt_unacked -= msg_conn_ack(hdr);
1328 		if (tsk->peer_caps & TIPC_BLOCK_FLOWCTL)
1329 			tsk->snd_win = msg_adv_win(hdr);
1330 		if (was_cong && !tsk_conn_cong(tsk))
1331 			sk->sk_write_space(sk);
1332 	} else if (mtyp != CONN_PROBE_REPLY) {
1333 		pr_warn("Received unknown CONN_PROTO msg\n");
1334 	}
1335 exit:
1336 	kfree_skb(skb);
1337 }
1338 
1339 /**
1340  * tipc_sendmsg - send message in connectionless manner
1341  * @sock: socket structure
1342  * @m: message to send
1343  * @dsz: amount of user data to be sent
1344  *
1345  * Message must have an destination specified explicitly.
1346  * Used for SOCK_RDM and SOCK_DGRAM messages,
1347  * and for 'SYN' messages on SOCK_SEQPACKET and SOCK_STREAM connections.
1348  * (Note: 'SYN+' is prohibited on SOCK_STREAM.)
1349  *
1350  * Returns the number of bytes sent on success, or errno otherwise
1351  */
1352 static int tipc_sendmsg(struct socket *sock,
1353 			struct msghdr *m, size_t dsz)
1354 {
1355 	struct sock *sk = sock->sk;
1356 	int ret;
1357 
1358 	lock_sock(sk);
1359 	ret = __tipc_sendmsg(sock, m, dsz);
1360 	release_sock(sk);
1361 
1362 	return ret;
1363 }
1364 
1365 static int __tipc_sendmsg(struct socket *sock, struct msghdr *m, size_t dlen)
1366 {
1367 	struct sock *sk = sock->sk;
1368 	struct net *net = sock_net(sk);
1369 	struct tipc_sock *tsk = tipc_sk(sk);
1370 	DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name);
1371 	long timeout = sock_sndtimeo(sk, m->msg_flags & MSG_DONTWAIT);
1372 	struct list_head *clinks = &tsk->cong_links;
1373 	bool syn = !tipc_sk_type_connectionless(sk);
1374 	struct tipc_group *grp = tsk->group;
1375 	struct tipc_msg *hdr = &tsk->phdr;
1376 	struct tipc_name_seq *seq;
1377 	struct sk_buff_head pkts;
1378 	u32 dport = 0, dnode = 0;
1379 	u32 type = 0, inst = 0;
1380 	int mtu, rc;
1381 
1382 	if (unlikely(dlen > TIPC_MAX_USER_MSG_SIZE))
1383 		return -EMSGSIZE;
1384 
1385 	if (likely(dest)) {
1386 		if (unlikely(m->msg_namelen < sizeof(*dest)))
1387 			return -EINVAL;
1388 		if (unlikely(dest->family != AF_TIPC))
1389 			return -EINVAL;
1390 	}
1391 
1392 	if (grp) {
1393 		if (!dest)
1394 			return tipc_send_group_bcast(sock, m, dlen, timeout);
1395 		if (dest->addrtype == TIPC_ADDR_NAME)
1396 			return tipc_send_group_anycast(sock, m, dlen, timeout);
1397 		if (dest->addrtype == TIPC_ADDR_ID)
1398 			return tipc_send_group_unicast(sock, m, dlen, timeout);
1399 		if (dest->addrtype == TIPC_ADDR_MCAST)
1400 			return tipc_send_group_mcast(sock, m, dlen, timeout);
1401 		return -EINVAL;
1402 	}
1403 
1404 	if (unlikely(!dest)) {
1405 		dest = &tsk->peer;
1406 		if (!syn && dest->family != AF_TIPC)
1407 			return -EDESTADDRREQ;
1408 	}
1409 
1410 	if (unlikely(syn)) {
1411 		if (sk->sk_state == TIPC_LISTEN)
1412 			return -EPIPE;
1413 		if (sk->sk_state != TIPC_OPEN)
1414 			return -EISCONN;
1415 		if (tsk->published)
1416 			return -EOPNOTSUPP;
1417 		if (dest->addrtype == TIPC_ADDR_NAME) {
1418 			tsk->conn_type = dest->addr.name.name.type;
1419 			tsk->conn_instance = dest->addr.name.name.instance;
1420 		}
1421 		msg_set_syn(hdr, 1);
1422 	}
1423 
1424 	seq = &dest->addr.nameseq;
1425 	if (dest->addrtype == TIPC_ADDR_MCAST)
1426 		return tipc_sendmcast(sock, seq, m, dlen, timeout);
1427 
1428 	if (dest->addrtype == TIPC_ADDR_NAME) {
1429 		type = dest->addr.name.name.type;
1430 		inst = dest->addr.name.name.instance;
1431 		dnode = dest->addr.name.domain;
1432 		dport = tipc_nametbl_translate(net, type, inst, &dnode);
1433 		if (unlikely(!dport && !dnode))
1434 			return -EHOSTUNREACH;
1435 	} else if (dest->addrtype == TIPC_ADDR_ID) {
1436 		dnode = dest->addr.id.node;
1437 	} else {
1438 		return -EINVAL;
1439 	}
1440 
1441 	/* Block or return if destination link is congested */
1442 	rc = tipc_wait_for_cond(sock, &timeout,
1443 				!tipc_dest_find(clinks, dnode, 0));
1444 	if (unlikely(rc))
1445 		return rc;
1446 
1447 	if (dest->addrtype == TIPC_ADDR_NAME) {
1448 		msg_set_type(hdr, TIPC_NAMED_MSG);
1449 		msg_set_hdr_sz(hdr, NAMED_H_SIZE);
1450 		msg_set_nametype(hdr, type);
1451 		msg_set_nameinst(hdr, inst);
1452 		msg_set_lookup_scope(hdr, tipc_node2scope(dnode));
1453 		msg_set_destnode(hdr, dnode);
1454 		msg_set_destport(hdr, dport);
1455 	} else { /* TIPC_ADDR_ID */
1456 		msg_set_type(hdr, TIPC_DIRECT_MSG);
1457 		msg_set_lookup_scope(hdr, 0);
1458 		msg_set_destnode(hdr, dnode);
1459 		msg_set_destport(hdr, dest->addr.id.ref);
1460 		msg_set_hdr_sz(hdr, BASIC_H_SIZE);
1461 	}
1462 
1463 	__skb_queue_head_init(&pkts);
1464 	mtu = tipc_node_get_mtu(net, dnode, tsk->portid, false);
1465 	rc = tipc_msg_build(hdr, m, 0, dlen, mtu, &pkts);
1466 	if (unlikely(rc != dlen))
1467 		return rc;
1468 	if (unlikely(syn && !tipc_msg_skb_clone(&pkts, &sk->sk_write_queue))) {
1469 		__skb_queue_purge(&pkts);
1470 		return -ENOMEM;
1471 	}
1472 
1473 	trace_tipc_sk_sendmsg(sk, skb_peek(&pkts), TIPC_DUMP_SK_SNDQ, " ");
1474 	rc = tipc_node_xmit(net, &pkts, dnode, tsk->portid);
1475 	if (unlikely(rc == -ELINKCONG)) {
1476 		tipc_dest_push(clinks, dnode, 0);
1477 		tsk->cong_link_cnt++;
1478 		rc = 0;
1479 	}
1480 
1481 	if (unlikely(syn && !rc))
1482 		tipc_set_sk_state(sk, TIPC_CONNECTING);
1483 
1484 	return rc ? rc : dlen;
1485 }
1486 
1487 /**
1488  * tipc_sendstream - send stream-oriented data
1489  * @sock: socket structure
1490  * @m: data to send
1491  * @dsz: total length of data to be transmitted
1492  *
1493  * Used for SOCK_STREAM data.
1494  *
1495  * Returns the number of bytes sent on success (or partial success),
1496  * or errno if no data sent
1497  */
1498 static int tipc_sendstream(struct socket *sock, struct msghdr *m, size_t dsz)
1499 {
1500 	struct sock *sk = sock->sk;
1501 	int ret;
1502 
1503 	lock_sock(sk);
1504 	ret = __tipc_sendstream(sock, m, dsz);
1505 	release_sock(sk);
1506 
1507 	return ret;
1508 }
1509 
1510 static int __tipc_sendstream(struct socket *sock, struct msghdr *m, size_t dlen)
1511 {
1512 	struct sock *sk = sock->sk;
1513 	DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name);
1514 	long timeout = sock_sndtimeo(sk, m->msg_flags & MSG_DONTWAIT);
1515 	struct sk_buff_head *txq = &sk->sk_write_queue;
1516 	struct tipc_sock *tsk = tipc_sk(sk);
1517 	struct tipc_msg *hdr = &tsk->phdr;
1518 	struct net *net = sock_net(sk);
1519 	u32 dnode = tsk_peer_node(tsk);
1520 	int maxnagle = tsk->maxnagle;
1521 	int maxpkt = tsk->max_pkt;
1522 	int send, sent = 0;
1523 	int blocks, rc = 0;
1524 
1525 	if (unlikely(dlen > INT_MAX))
1526 		return -EMSGSIZE;
1527 
1528 	/* Handle implicit connection setup */
1529 	if (unlikely(dest)) {
1530 		rc = __tipc_sendmsg(sock, m, dlen);
1531 		if (dlen && dlen == rc) {
1532 			tsk->peer_caps = tipc_node_get_capabilities(net, dnode);
1533 			tsk->snt_unacked = tsk_inc(tsk, dlen + msg_hdr_sz(hdr));
1534 		}
1535 		return rc;
1536 	}
1537 
1538 	do {
1539 		rc = tipc_wait_for_cond(sock, &timeout,
1540 					(!tsk->cong_link_cnt &&
1541 					 !tsk_conn_cong(tsk) &&
1542 					 tipc_sk_connected(sk)));
1543 		if (unlikely(rc))
1544 			break;
1545 		send = min_t(size_t, dlen - sent, TIPC_MAX_USER_MSG_SIZE);
1546 		blocks = tsk->snd_backlog;
1547 		if (tsk->oneway++ >= 4 && send <= maxnagle) {
1548 			rc = tipc_msg_append(hdr, m, send, maxnagle, txq);
1549 			if (unlikely(rc < 0))
1550 				break;
1551 			blocks += rc;
1552 			if (blocks <= 64 && tsk->expect_ack) {
1553 				tsk->snd_backlog = blocks;
1554 				sent += send;
1555 				break;
1556 			}
1557 			tsk->expect_ack = true;
1558 		} else {
1559 			rc = tipc_msg_build(hdr, m, sent, send, maxpkt, txq);
1560 			if (unlikely(rc != send))
1561 				break;
1562 			blocks += tsk_inc(tsk, send + MIN_H_SIZE);
1563 		}
1564 		trace_tipc_sk_sendstream(sk, skb_peek(txq),
1565 					 TIPC_DUMP_SK_SNDQ, " ");
1566 		rc = tipc_node_xmit(net, txq, dnode, tsk->portid);
1567 		if (unlikely(rc == -ELINKCONG)) {
1568 			tsk->cong_link_cnt = 1;
1569 			rc = 0;
1570 		}
1571 		if (likely(!rc)) {
1572 			tsk->snt_unacked += blocks;
1573 			tsk->snd_backlog = 0;
1574 			sent += send;
1575 		}
1576 	} while (sent < dlen && !rc);
1577 
1578 	return sent ? sent : rc;
1579 }
1580 
1581 /**
1582  * tipc_send_packet - send a connection-oriented message
1583  * @sock: socket structure
1584  * @m: message to send
1585  * @dsz: length of data to be transmitted
1586  *
1587  * Used for SOCK_SEQPACKET messages.
1588  *
1589  * Returns the number of bytes sent on success, or errno otherwise
1590  */
1591 static int tipc_send_packet(struct socket *sock, struct msghdr *m, size_t dsz)
1592 {
1593 	if (dsz > TIPC_MAX_USER_MSG_SIZE)
1594 		return -EMSGSIZE;
1595 
1596 	return tipc_sendstream(sock, m, dsz);
1597 }
1598 
1599 /* tipc_sk_finish_conn - complete the setup of a connection
1600  */
1601 static void tipc_sk_finish_conn(struct tipc_sock *tsk, u32 peer_port,
1602 				u32 peer_node)
1603 {
1604 	struct sock *sk = &tsk->sk;
1605 	struct net *net = sock_net(sk);
1606 	struct tipc_msg *msg = &tsk->phdr;
1607 
1608 	msg_set_syn(msg, 0);
1609 	msg_set_destnode(msg, peer_node);
1610 	msg_set_destport(msg, peer_port);
1611 	msg_set_type(msg, TIPC_CONN_MSG);
1612 	msg_set_lookup_scope(msg, 0);
1613 	msg_set_hdr_sz(msg, SHORT_H_SIZE);
1614 
1615 	sk_reset_timer(sk, &sk->sk_timer, jiffies + CONN_PROBING_INTV);
1616 	tipc_set_sk_state(sk, TIPC_ESTABLISHED);
1617 	tipc_node_add_conn(net, peer_node, tsk->portid, peer_port);
1618 	tsk->max_pkt = tipc_node_get_mtu(net, peer_node, tsk->portid, true);
1619 	tsk->peer_caps = tipc_node_get_capabilities(net, peer_node);
1620 	tsk_set_nagle(tsk);
1621 	__skb_queue_purge(&sk->sk_write_queue);
1622 	if (tsk->peer_caps & TIPC_BLOCK_FLOWCTL)
1623 		return;
1624 
1625 	/* Fall back to message based flow control */
1626 	tsk->rcv_win = FLOWCTL_MSG_WIN;
1627 	tsk->snd_win = FLOWCTL_MSG_WIN;
1628 }
1629 
1630 /**
1631  * tipc_sk_set_orig_addr - capture sender's address for received message
1632  * @m: descriptor for message info
1633  * @hdr: received message header
1634  *
1635  * Note: Address is not captured if not requested by receiver.
1636  */
1637 static void tipc_sk_set_orig_addr(struct msghdr *m, struct sk_buff *skb)
1638 {
1639 	DECLARE_SOCKADDR(struct sockaddr_pair *, srcaddr, m->msg_name);
1640 	struct tipc_msg *hdr = buf_msg(skb);
1641 
1642 	if (!srcaddr)
1643 		return;
1644 
1645 	srcaddr->sock.family = AF_TIPC;
1646 	srcaddr->sock.addrtype = TIPC_ADDR_ID;
1647 	srcaddr->sock.scope = 0;
1648 	srcaddr->sock.addr.id.ref = msg_origport(hdr);
1649 	srcaddr->sock.addr.id.node = msg_orignode(hdr);
1650 	srcaddr->sock.addr.name.domain = 0;
1651 	m->msg_namelen = sizeof(struct sockaddr_tipc);
1652 
1653 	if (!msg_in_group(hdr))
1654 		return;
1655 
1656 	/* Group message users may also want to know sending member's id */
1657 	srcaddr->member.family = AF_TIPC;
1658 	srcaddr->member.addrtype = TIPC_ADDR_NAME;
1659 	srcaddr->member.scope = 0;
1660 	srcaddr->member.addr.name.name.type = msg_nametype(hdr);
1661 	srcaddr->member.addr.name.name.instance = TIPC_SKB_CB(skb)->orig_member;
1662 	srcaddr->member.addr.name.domain = 0;
1663 	m->msg_namelen = sizeof(*srcaddr);
1664 }
1665 
1666 /**
1667  * tipc_sk_anc_data_recv - optionally capture ancillary data for received message
1668  * @m: descriptor for message info
1669  * @skb: received message buffer
1670  * @tsk: TIPC port associated with message
1671  *
1672  * Note: Ancillary data is not captured if not requested by receiver.
1673  *
1674  * Returns 0 if successful, otherwise errno
1675  */
1676 static int tipc_sk_anc_data_recv(struct msghdr *m, struct sk_buff *skb,
1677 				 struct tipc_sock *tsk)
1678 {
1679 	struct tipc_msg *msg;
1680 	u32 anc_data[3];
1681 	u32 err;
1682 	u32 dest_type;
1683 	int has_name;
1684 	int res;
1685 
1686 	if (likely(m->msg_controllen == 0))
1687 		return 0;
1688 	msg = buf_msg(skb);
1689 
1690 	/* Optionally capture errored message object(s) */
1691 	err = msg ? msg_errcode(msg) : 0;
1692 	if (unlikely(err)) {
1693 		anc_data[0] = err;
1694 		anc_data[1] = msg_data_sz(msg);
1695 		res = put_cmsg(m, SOL_TIPC, TIPC_ERRINFO, 8, anc_data);
1696 		if (res)
1697 			return res;
1698 		if (anc_data[1]) {
1699 			if (skb_linearize(skb))
1700 				return -ENOMEM;
1701 			msg = buf_msg(skb);
1702 			res = put_cmsg(m, SOL_TIPC, TIPC_RETDATA, anc_data[1],
1703 				       msg_data(msg));
1704 			if (res)
1705 				return res;
1706 		}
1707 	}
1708 
1709 	/* Optionally capture message destination object */
1710 	dest_type = msg ? msg_type(msg) : TIPC_DIRECT_MSG;
1711 	switch (dest_type) {
1712 	case TIPC_NAMED_MSG:
1713 		has_name = 1;
1714 		anc_data[0] = msg_nametype(msg);
1715 		anc_data[1] = msg_namelower(msg);
1716 		anc_data[2] = msg_namelower(msg);
1717 		break;
1718 	case TIPC_MCAST_MSG:
1719 		has_name = 1;
1720 		anc_data[0] = msg_nametype(msg);
1721 		anc_data[1] = msg_namelower(msg);
1722 		anc_data[2] = msg_nameupper(msg);
1723 		break;
1724 	case TIPC_CONN_MSG:
1725 		has_name = (tsk->conn_type != 0);
1726 		anc_data[0] = tsk->conn_type;
1727 		anc_data[1] = tsk->conn_instance;
1728 		anc_data[2] = tsk->conn_instance;
1729 		break;
1730 	default:
1731 		has_name = 0;
1732 	}
1733 	if (has_name) {
1734 		res = put_cmsg(m, SOL_TIPC, TIPC_DESTNAME, 12, anc_data);
1735 		if (res)
1736 			return res;
1737 	}
1738 
1739 	return 0;
1740 }
1741 
1742 static void tipc_sk_send_ack(struct tipc_sock *tsk)
1743 {
1744 	struct sock *sk = &tsk->sk;
1745 	struct net *net = sock_net(sk);
1746 	struct sk_buff *skb = NULL;
1747 	struct tipc_msg *msg;
1748 	u32 peer_port = tsk_peer_port(tsk);
1749 	u32 dnode = tsk_peer_node(tsk);
1750 
1751 	if (!tipc_sk_connected(sk))
1752 		return;
1753 	skb = tipc_msg_create(CONN_MANAGER, CONN_ACK, INT_H_SIZE, 0,
1754 			      dnode, tsk_own_node(tsk), peer_port,
1755 			      tsk->portid, TIPC_OK);
1756 	if (!skb)
1757 		return;
1758 	msg = buf_msg(skb);
1759 	msg_set_conn_ack(msg, tsk->rcv_unacked);
1760 	tsk->rcv_unacked = 0;
1761 
1762 	/* Adjust to and advertize the correct window limit */
1763 	if (tsk->peer_caps & TIPC_BLOCK_FLOWCTL) {
1764 		tsk->rcv_win = tsk_adv_blocks(tsk->sk.sk_rcvbuf);
1765 		msg_set_adv_win(msg, tsk->rcv_win);
1766 	}
1767 	tipc_node_xmit_skb(net, skb, dnode, msg_link_selector(msg));
1768 }
1769 
1770 static int tipc_wait_for_rcvmsg(struct socket *sock, long *timeop)
1771 {
1772 	struct sock *sk = sock->sk;
1773 	DEFINE_WAIT_FUNC(wait, woken_wake_function);
1774 	long timeo = *timeop;
1775 	int err = sock_error(sk);
1776 
1777 	if (err)
1778 		return err;
1779 
1780 	for (;;) {
1781 		if (timeo && skb_queue_empty(&sk->sk_receive_queue)) {
1782 			if (sk->sk_shutdown & RCV_SHUTDOWN) {
1783 				err = -ENOTCONN;
1784 				break;
1785 			}
1786 			add_wait_queue(sk_sleep(sk), &wait);
1787 			release_sock(sk);
1788 			timeo = wait_woken(&wait, TASK_INTERRUPTIBLE, timeo);
1789 			sched_annotate_sleep();
1790 			lock_sock(sk);
1791 			remove_wait_queue(sk_sleep(sk), &wait);
1792 		}
1793 		err = 0;
1794 		if (!skb_queue_empty(&sk->sk_receive_queue))
1795 			break;
1796 		err = -EAGAIN;
1797 		if (!timeo)
1798 			break;
1799 		err = sock_intr_errno(timeo);
1800 		if (signal_pending(current))
1801 			break;
1802 
1803 		err = sock_error(sk);
1804 		if (err)
1805 			break;
1806 	}
1807 	*timeop = timeo;
1808 	return err;
1809 }
1810 
1811 /**
1812  * tipc_recvmsg - receive packet-oriented message
1813  * @m: descriptor for message info
1814  * @buflen: length of user buffer area
1815  * @flags: receive flags
1816  *
1817  * Used for SOCK_DGRAM, SOCK_RDM, and SOCK_SEQPACKET messages.
1818  * If the complete message doesn't fit in user area, truncate it.
1819  *
1820  * Returns size of returned message data, errno otherwise
1821  */
1822 static int tipc_recvmsg(struct socket *sock, struct msghdr *m,
1823 			size_t buflen,	int flags)
1824 {
1825 	struct sock *sk = sock->sk;
1826 	bool connected = !tipc_sk_type_connectionless(sk);
1827 	struct tipc_sock *tsk = tipc_sk(sk);
1828 	int rc, err, hlen, dlen, copy;
1829 	struct sk_buff_head xmitq;
1830 	struct tipc_msg *hdr;
1831 	struct sk_buff *skb;
1832 	bool grp_evt;
1833 	long timeout;
1834 
1835 	/* Catch invalid receive requests */
1836 	if (unlikely(!buflen))
1837 		return -EINVAL;
1838 
1839 	lock_sock(sk);
1840 	if (unlikely(connected && sk->sk_state == TIPC_OPEN)) {
1841 		rc = -ENOTCONN;
1842 		goto exit;
1843 	}
1844 	timeout = sock_rcvtimeo(sk, flags & MSG_DONTWAIT);
1845 
1846 	/* Step rcv queue to first msg with data or error; wait if necessary */
1847 	do {
1848 		rc = tipc_wait_for_rcvmsg(sock, &timeout);
1849 		if (unlikely(rc))
1850 			goto exit;
1851 		skb = skb_peek(&sk->sk_receive_queue);
1852 		hdr = buf_msg(skb);
1853 		dlen = msg_data_sz(hdr);
1854 		hlen = msg_hdr_sz(hdr);
1855 		err = msg_errcode(hdr);
1856 		grp_evt = msg_is_grp_evt(hdr);
1857 		if (likely(dlen || err))
1858 			break;
1859 		tsk_advance_rx_queue(sk);
1860 	} while (1);
1861 
1862 	/* Collect msg meta data, including error code and rejected data */
1863 	tipc_sk_set_orig_addr(m, skb);
1864 	rc = tipc_sk_anc_data_recv(m, skb, tsk);
1865 	if (unlikely(rc))
1866 		goto exit;
1867 	hdr = buf_msg(skb);
1868 
1869 	/* Capture data if non-error msg, otherwise just set return value */
1870 	if (likely(!err)) {
1871 		copy = min_t(int, dlen, buflen);
1872 		if (unlikely(copy != dlen))
1873 			m->msg_flags |= MSG_TRUNC;
1874 		rc = skb_copy_datagram_msg(skb, hlen, m, copy);
1875 	} else {
1876 		copy = 0;
1877 		rc = 0;
1878 		if (err != TIPC_CONN_SHUTDOWN && connected && !m->msg_control)
1879 			rc = -ECONNRESET;
1880 	}
1881 	if (unlikely(rc))
1882 		goto exit;
1883 
1884 	/* Mark message as group event if applicable */
1885 	if (unlikely(grp_evt)) {
1886 		if (msg_grp_evt(hdr) == TIPC_WITHDRAWN)
1887 			m->msg_flags |= MSG_EOR;
1888 		m->msg_flags |= MSG_OOB;
1889 		copy = 0;
1890 	}
1891 
1892 	/* Caption of data or error code/rejected data was successful */
1893 	if (unlikely(flags & MSG_PEEK))
1894 		goto exit;
1895 
1896 	/* Send group flow control advertisement when applicable */
1897 	if (tsk->group && msg_in_group(hdr) && !grp_evt) {
1898 		__skb_queue_head_init(&xmitq);
1899 		tipc_group_update_rcv_win(tsk->group, tsk_blocks(hlen + dlen),
1900 					  msg_orignode(hdr), msg_origport(hdr),
1901 					  &xmitq);
1902 		tipc_node_distr_xmit(sock_net(sk), &xmitq);
1903 	}
1904 
1905 	tsk_advance_rx_queue(sk);
1906 
1907 	if (likely(!connected))
1908 		goto exit;
1909 
1910 	/* Send connection flow control advertisement when applicable */
1911 	tsk->rcv_unacked += tsk_inc(tsk, hlen + dlen);
1912 	if (tsk->rcv_unacked >= tsk->rcv_win / TIPC_ACK_RATE)
1913 		tipc_sk_send_ack(tsk);
1914 exit:
1915 	release_sock(sk);
1916 	return rc ? rc : copy;
1917 }
1918 
1919 /**
1920  * tipc_recvstream - receive stream-oriented data
1921  * @m: descriptor for message info
1922  * @buflen: total size of user buffer area
1923  * @flags: receive flags
1924  *
1925  * Used for SOCK_STREAM messages only.  If not enough data is available
1926  * will optionally wait for more; never truncates data.
1927  *
1928  * Returns size of returned message data, errno otherwise
1929  */
1930 static int tipc_recvstream(struct socket *sock, struct msghdr *m,
1931 			   size_t buflen, int flags)
1932 {
1933 	struct sock *sk = sock->sk;
1934 	struct tipc_sock *tsk = tipc_sk(sk);
1935 	struct sk_buff *skb;
1936 	struct tipc_msg *hdr;
1937 	struct tipc_skb_cb *skb_cb;
1938 	bool peek = flags & MSG_PEEK;
1939 	int offset, required, copy, copied = 0;
1940 	int hlen, dlen, err, rc;
1941 	bool ack = false;
1942 	long timeout;
1943 
1944 	/* Catch invalid receive attempts */
1945 	if (unlikely(!buflen))
1946 		return -EINVAL;
1947 
1948 	lock_sock(sk);
1949 
1950 	if (unlikely(sk->sk_state == TIPC_OPEN)) {
1951 		rc = -ENOTCONN;
1952 		goto exit;
1953 	}
1954 	required = sock_rcvlowat(sk, flags & MSG_WAITALL, buflen);
1955 	timeout = sock_rcvtimeo(sk, flags & MSG_DONTWAIT);
1956 
1957 	do {
1958 		/* Look at first msg in receive queue; wait if necessary */
1959 		rc = tipc_wait_for_rcvmsg(sock, &timeout);
1960 		if (unlikely(rc))
1961 			break;
1962 		skb = skb_peek(&sk->sk_receive_queue);
1963 		skb_cb = TIPC_SKB_CB(skb);
1964 		hdr = buf_msg(skb);
1965 		dlen = msg_data_sz(hdr);
1966 		hlen = msg_hdr_sz(hdr);
1967 		err = msg_errcode(hdr);
1968 
1969 		/* Discard any empty non-errored (SYN-) message */
1970 		if (unlikely(!dlen && !err)) {
1971 			tsk_advance_rx_queue(sk);
1972 			continue;
1973 		}
1974 
1975 		/* Collect msg meta data, incl. error code and rejected data */
1976 		if (!copied) {
1977 			tipc_sk_set_orig_addr(m, skb);
1978 			rc = tipc_sk_anc_data_recv(m, skb, tsk);
1979 			if (rc)
1980 				break;
1981 			hdr = buf_msg(skb);
1982 		}
1983 
1984 		/* Copy data if msg ok, otherwise return error/partial data */
1985 		if (likely(!err)) {
1986 			ack = msg_ack_required(hdr);
1987 			offset = skb_cb->bytes_read;
1988 			copy = min_t(int, dlen - offset, buflen - copied);
1989 			rc = skb_copy_datagram_msg(skb, hlen + offset, m, copy);
1990 			if (unlikely(rc))
1991 				break;
1992 			copied += copy;
1993 			offset += copy;
1994 			if (unlikely(offset < dlen)) {
1995 				if (!peek)
1996 					skb_cb->bytes_read = offset;
1997 				break;
1998 			}
1999 		} else {
2000 			rc = 0;
2001 			if ((err != TIPC_CONN_SHUTDOWN) && !m->msg_control)
2002 				rc = -ECONNRESET;
2003 			if (copied || rc)
2004 				break;
2005 		}
2006 
2007 		if (unlikely(peek))
2008 			break;
2009 
2010 		tsk_advance_rx_queue(sk);
2011 
2012 		/* Send connection flow control advertisement when applicable */
2013 		tsk->rcv_unacked += tsk_inc(tsk, hlen + dlen);
2014 		if (ack || tsk->rcv_unacked >= tsk->rcv_win / TIPC_ACK_RATE)
2015 			tipc_sk_send_ack(tsk);
2016 
2017 		/* Exit if all requested data or FIN/error received */
2018 		if (copied == buflen || err)
2019 			break;
2020 
2021 	} while (!skb_queue_empty(&sk->sk_receive_queue) || copied < required);
2022 exit:
2023 	release_sock(sk);
2024 	return copied ? copied : rc;
2025 }
2026 
2027 /**
2028  * tipc_write_space - wake up thread if port congestion is released
2029  * @sk: socket
2030  */
2031 static void tipc_write_space(struct sock *sk)
2032 {
2033 	struct socket_wq *wq;
2034 
2035 	rcu_read_lock();
2036 	wq = rcu_dereference(sk->sk_wq);
2037 	if (skwq_has_sleeper(wq))
2038 		wake_up_interruptible_sync_poll(&wq->wait, EPOLLOUT |
2039 						EPOLLWRNORM | EPOLLWRBAND);
2040 	rcu_read_unlock();
2041 }
2042 
2043 /**
2044  * tipc_data_ready - wake up threads to indicate messages have been received
2045  * @sk: socket
2046  * @len: the length of messages
2047  */
2048 static void tipc_data_ready(struct sock *sk)
2049 {
2050 	struct socket_wq *wq;
2051 
2052 	rcu_read_lock();
2053 	wq = rcu_dereference(sk->sk_wq);
2054 	if (skwq_has_sleeper(wq))
2055 		wake_up_interruptible_sync_poll(&wq->wait, EPOLLIN |
2056 						EPOLLRDNORM | EPOLLRDBAND);
2057 	rcu_read_unlock();
2058 }
2059 
2060 static void tipc_sock_destruct(struct sock *sk)
2061 {
2062 	__skb_queue_purge(&sk->sk_receive_queue);
2063 }
2064 
2065 static void tipc_sk_proto_rcv(struct sock *sk,
2066 			      struct sk_buff_head *inputq,
2067 			      struct sk_buff_head *xmitq)
2068 {
2069 	struct sk_buff *skb = __skb_dequeue(inputq);
2070 	struct tipc_sock *tsk = tipc_sk(sk);
2071 	struct tipc_msg *hdr = buf_msg(skb);
2072 	struct tipc_group *grp = tsk->group;
2073 	bool wakeup = false;
2074 
2075 	switch (msg_user(hdr)) {
2076 	case CONN_MANAGER:
2077 		tipc_sk_conn_proto_rcv(tsk, skb, inputq, xmitq);
2078 		return;
2079 	case SOCK_WAKEUP:
2080 		tipc_dest_del(&tsk->cong_links, msg_orignode(hdr), 0);
2081 		/* coupled with smp_rmb() in tipc_wait_for_cond() */
2082 		smp_wmb();
2083 		tsk->cong_link_cnt--;
2084 		wakeup = true;
2085 		tipc_sk_push_backlog(tsk);
2086 		break;
2087 	case GROUP_PROTOCOL:
2088 		tipc_group_proto_rcv(grp, &wakeup, hdr, inputq, xmitq);
2089 		break;
2090 	case TOP_SRV:
2091 		tipc_group_member_evt(tsk->group, &wakeup, &sk->sk_rcvbuf,
2092 				      hdr, inputq, xmitq);
2093 		break;
2094 	default:
2095 		break;
2096 	}
2097 
2098 	if (wakeup)
2099 		sk->sk_write_space(sk);
2100 
2101 	kfree_skb(skb);
2102 }
2103 
2104 /**
2105  * tipc_sk_filter_connect - check incoming message for a connection-based socket
2106  * @tsk: TIPC socket
2107  * @skb: pointer to message buffer.
2108  * Returns true if message should be added to receive queue, false otherwise
2109  */
2110 static bool tipc_sk_filter_connect(struct tipc_sock *tsk, struct sk_buff *skb)
2111 {
2112 	struct sock *sk = &tsk->sk;
2113 	struct net *net = sock_net(sk);
2114 	struct tipc_msg *hdr = buf_msg(skb);
2115 	bool con_msg = msg_connected(hdr);
2116 	u32 pport = tsk_peer_port(tsk);
2117 	u32 pnode = tsk_peer_node(tsk);
2118 	u32 oport = msg_origport(hdr);
2119 	u32 onode = msg_orignode(hdr);
2120 	int err = msg_errcode(hdr);
2121 	unsigned long delay;
2122 
2123 	if (unlikely(msg_mcast(hdr)))
2124 		return false;
2125 	tsk->oneway = 0;
2126 
2127 	switch (sk->sk_state) {
2128 	case TIPC_CONNECTING:
2129 		/* Setup ACK */
2130 		if (likely(con_msg)) {
2131 			if (err)
2132 				break;
2133 			tipc_sk_finish_conn(tsk, oport, onode);
2134 			msg_set_importance(&tsk->phdr, msg_importance(hdr));
2135 			/* ACK+ message with data is added to receive queue */
2136 			if (msg_data_sz(hdr))
2137 				return true;
2138 			/* Empty ACK-, - wake up sleeping connect() and drop */
2139 			sk->sk_state_change(sk);
2140 			msg_set_dest_droppable(hdr, 1);
2141 			return false;
2142 		}
2143 		/* Ignore connectionless message if not from listening socket */
2144 		if (oport != pport || onode != pnode)
2145 			return false;
2146 
2147 		/* Rejected SYN */
2148 		if (err != TIPC_ERR_OVERLOAD)
2149 			break;
2150 
2151 		/* Prepare for new setup attempt if we have a SYN clone */
2152 		if (skb_queue_empty(&sk->sk_write_queue))
2153 			break;
2154 		get_random_bytes(&delay, 2);
2155 		delay %= (tsk->conn_timeout / 4);
2156 		delay = msecs_to_jiffies(delay + 100);
2157 		sk_reset_timer(sk, &sk->sk_timer, jiffies + delay);
2158 		return false;
2159 	case TIPC_OPEN:
2160 	case TIPC_DISCONNECTING:
2161 		return false;
2162 	case TIPC_LISTEN:
2163 		/* Accept only SYN message */
2164 		if (!msg_is_syn(hdr) &&
2165 		    tipc_node_get_capabilities(net, onode) & TIPC_SYN_BIT)
2166 			return false;
2167 		if (!con_msg && !err)
2168 			return true;
2169 		return false;
2170 	case TIPC_ESTABLISHED:
2171 		if (!skb_queue_empty(&sk->sk_write_queue))
2172 			tipc_sk_push_backlog(tsk);
2173 		/* Accept only connection-based messages sent by peer */
2174 		if (likely(con_msg && !err && pport == oport && pnode == onode))
2175 			return true;
2176 		if (!tsk_peer_msg(tsk, hdr))
2177 			return false;
2178 		if (!err)
2179 			return true;
2180 		tipc_set_sk_state(sk, TIPC_DISCONNECTING);
2181 		tipc_node_remove_conn(net, pnode, tsk->portid);
2182 		sk->sk_state_change(sk);
2183 		return true;
2184 	default:
2185 		pr_err("Unknown sk_state %u\n", sk->sk_state);
2186 	}
2187 	/* Abort connection setup attempt */
2188 	tipc_set_sk_state(sk, TIPC_DISCONNECTING);
2189 	sk->sk_err = ECONNREFUSED;
2190 	sk->sk_state_change(sk);
2191 	return true;
2192 }
2193 
2194 /**
2195  * rcvbuf_limit - get proper overload limit of socket receive queue
2196  * @sk: socket
2197  * @skb: message
2198  *
2199  * For connection oriented messages, irrespective of importance,
2200  * default queue limit is 2 MB.
2201  *
2202  * For connectionless messages, queue limits are based on message
2203  * importance as follows:
2204  *
2205  * TIPC_LOW_IMPORTANCE       (2 MB)
2206  * TIPC_MEDIUM_IMPORTANCE    (4 MB)
2207  * TIPC_HIGH_IMPORTANCE      (8 MB)
2208  * TIPC_CRITICAL_IMPORTANCE  (16 MB)
2209  *
2210  * Returns overload limit according to corresponding message importance
2211  */
2212 static unsigned int rcvbuf_limit(struct sock *sk, struct sk_buff *skb)
2213 {
2214 	struct tipc_sock *tsk = tipc_sk(sk);
2215 	struct tipc_msg *hdr = buf_msg(skb);
2216 
2217 	if (unlikely(msg_in_group(hdr)))
2218 		return READ_ONCE(sk->sk_rcvbuf);
2219 
2220 	if (unlikely(!msg_connected(hdr)))
2221 		return READ_ONCE(sk->sk_rcvbuf) << msg_importance(hdr);
2222 
2223 	if (likely(tsk->peer_caps & TIPC_BLOCK_FLOWCTL))
2224 		return READ_ONCE(sk->sk_rcvbuf);
2225 
2226 	return FLOWCTL_MSG_LIM;
2227 }
2228 
2229 /**
2230  * tipc_sk_filter_rcv - validate incoming message
2231  * @sk: socket
2232  * @skb: pointer to message.
2233  *
2234  * Enqueues message on receive queue if acceptable; optionally handles
2235  * disconnect indication for a connected socket.
2236  *
2237  * Called with socket lock already taken
2238  *
2239  */
2240 static void tipc_sk_filter_rcv(struct sock *sk, struct sk_buff *skb,
2241 			       struct sk_buff_head *xmitq)
2242 {
2243 	bool sk_conn = !tipc_sk_type_connectionless(sk);
2244 	struct tipc_sock *tsk = tipc_sk(sk);
2245 	struct tipc_group *grp = tsk->group;
2246 	struct tipc_msg *hdr = buf_msg(skb);
2247 	struct net *net = sock_net(sk);
2248 	struct sk_buff_head inputq;
2249 	int mtyp = msg_type(hdr);
2250 	int limit, err = TIPC_OK;
2251 
2252 	trace_tipc_sk_filter_rcv(sk, skb, TIPC_DUMP_ALL, " ");
2253 	TIPC_SKB_CB(skb)->bytes_read = 0;
2254 	__skb_queue_head_init(&inputq);
2255 	__skb_queue_tail(&inputq, skb);
2256 
2257 	if (unlikely(!msg_isdata(hdr)))
2258 		tipc_sk_proto_rcv(sk, &inputq, xmitq);
2259 
2260 	if (unlikely(grp))
2261 		tipc_group_filter_msg(grp, &inputq, xmitq);
2262 
2263 	if (unlikely(!grp) && mtyp == TIPC_MCAST_MSG)
2264 		tipc_mcast_filter_msg(net, &tsk->mc_method.deferredq, &inputq);
2265 
2266 	/* Validate and add to receive buffer if there is space */
2267 	while ((skb = __skb_dequeue(&inputq))) {
2268 		hdr = buf_msg(skb);
2269 		limit = rcvbuf_limit(sk, skb);
2270 		if ((sk_conn && !tipc_sk_filter_connect(tsk, skb)) ||
2271 		    (!sk_conn && msg_connected(hdr)) ||
2272 		    (!grp && msg_in_group(hdr)))
2273 			err = TIPC_ERR_NO_PORT;
2274 		else if (sk_rmem_alloc_get(sk) + skb->truesize >= limit) {
2275 			trace_tipc_sk_dump(sk, skb, TIPC_DUMP_ALL,
2276 					   "err_overload2!");
2277 			atomic_inc(&sk->sk_drops);
2278 			err = TIPC_ERR_OVERLOAD;
2279 		}
2280 
2281 		if (unlikely(err)) {
2282 			if (tipc_msg_reverse(tipc_own_addr(net), &skb, err)) {
2283 				trace_tipc_sk_rej_msg(sk, skb, TIPC_DUMP_NONE,
2284 						      "@filter_rcv!");
2285 				__skb_queue_tail(xmitq, skb);
2286 			}
2287 			err = TIPC_OK;
2288 			continue;
2289 		}
2290 		__skb_queue_tail(&sk->sk_receive_queue, skb);
2291 		skb_set_owner_r(skb, sk);
2292 		trace_tipc_sk_overlimit2(sk, skb, TIPC_DUMP_ALL,
2293 					 "rcvq >90% allocated!");
2294 		sk->sk_data_ready(sk);
2295 	}
2296 }
2297 
2298 /**
2299  * tipc_sk_backlog_rcv - handle incoming message from backlog queue
2300  * @sk: socket
2301  * @skb: message
2302  *
2303  * Caller must hold socket lock
2304  */
2305 static int tipc_sk_backlog_rcv(struct sock *sk, struct sk_buff *skb)
2306 {
2307 	unsigned int before = sk_rmem_alloc_get(sk);
2308 	struct sk_buff_head xmitq;
2309 	unsigned int added;
2310 
2311 	__skb_queue_head_init(&xmitq);
2312 
2313 	tipc_sk_filter_rcv(sk, skb, &xmitq);
2314 	added = sk_rmem_alloc_get(sk) - before;
2315 	atomic_add(added, &tipc_sk(sk)->dupl_rcvcnt);
2316 
2317 	/* Send pending response/rejected messages, if any */
2318 	tipc_node_distr_xmit(sock_net(sk), &xmitq);
2319 	return 0;
2320 }
2321 
2322 /**
2323  * tipc_sk_enqueue - extract all buffers with destination 'dport' from
2324  *                   inputq and try adding them to socket or backlog queue
2325  * @inputq: list of incoming buffers with potentially different destinations
2326  * @sk: socket where the buffers should be enqueued
2327  * @dport: port number for the socket
2328  *
2329  * Caller must hold socket lock
2330  */
2331 static void tipc_sk_enqueue(struct sk_buff_head *inputq, struct sock *sk,
2332 			    u32 dport, struct sk_buff_head *xmitq)
2333 {
2334 	unsigned long time_limit = jiffies + 2;
2335 	struct sk_buff *skb;
2336 	unsigned int lim;
2337 	atomic_t *dcnt;
2338 	u32 onode;
2339 
2340 	while (skb_queue_len(inputq)) {
2341 		if (unlikely(time_after_eq(jiffies, time_limit)))
2342 			return;
2343 
2344 		skb = tipc_skb_dequeue(inputq, dport);
2345 		if (unlikely(!skb))
2346 			return;
2347 
2348 		/* Add message directly to receive queue if possible */
2349 		if (!sock_owned_by_user(sk)) {
2350 			tipc_sk_filter_rcv(sk, skb, xmitq);
2351 			continue;
2352 		}
2353 
2354 		/* Try backlog, compensating for double-counted bytes */
2355 		dcnt = &tipc_sk(sk)->dupl_rcvcnt;
2356 		if (!sk->sk_backlog.len)
2357 			atomic_set(dcnt, 0);
2358 		lim = rcvbuf_limit(sk, skb) + atomic_read(dcnt);
2359 		if (likely(!sk_add_backlog(sk, skb, lim))) {
2360 			trace_tipc_sk_overlimit1(sk, skb, TIPC_DUMP_ALL,
2361 						 "bklg & rcvq >90% allocated!");
2362 			continue;
2363 		}
2364 
2365 		trace_tipc_sk_dump(sk, skb, TIPC_DUMP_ALL, "err_overload!");
2366 		/* Overload => reject message back to sender */
2367 		onode = tipc_own_addr(sock_net(sk));
2368 		atomic_inc(&sk->sk_drops);
2369 		if (tipc_msg_reverse(onode, &skb, TIPC_ERR_OVERLOAD)) {
2370 			trace_tipc_sk_rej_msg(sk, skb, TIPC_DUMP_ALL,
2371 					      "@sk_enqueue!");
2372 			__skb_queue_tail(xmitq, skb);
2373 		}
2374 		break;
2375 	}
2376 }
2377 
2378 /**
2379  * tipc_sk_rcv - handle a chain of incoming buffers
2380  * @inputq: buffer list containing the buffers
2381  * Consumes all buffers in list until inputq is empty
2382  * Note: may be called in multiple threads referring to the same queue
2383  */
2384 void tipc_sk_rcv(struct net *net, struct sk_buff_head *inputq)
2385 {
2386 	struct sk_buff_head xmitq;
2387 	u32 dnode, dport = 0;
2388 	int err;
2389 	struct tipc_sock *tsk;
2390 	struct sock *sk;
2391 	struct sk_buff *skb;
2392 
2393 	__skb_queue_head_init(&xmitq);
2394 	while (skb_queue_len(inputq)) {
2395 		dport = tipc_skb_peek_port(inputq, dport);
2396 		tsk = tipc_sk_lookup(net, dport);
2397 
2398 		if (likely(tsk)) {
2399 			sk = &tsk->sk;
2400 			if (likely(spin_trylock_bh(&sk->sk_lock.slock))) {
2401 				tipc_sk_enqueue(inputq, sk, dport, &xmitq);
2402 				spin_unlock_bh(&sk->sk_lock.slock);
2403 			}
2404 			/* Send pending response/rejected messages, if any */
2405 			tipc_node_distr_xmit(sock_net(sk), &xmitq);
2406 			sock_put(sk);
2407 			continue;
2408 		}
2409 		/* No destination socket => dequeue skb if still there */
2410 		skb = tipc_skb_dequeue(inputq, dport);
2411 		if (!skb)
2412 			return;
2413 
2414 		/* Try secondary lookup if unresolved named message */
2415 		err = TIPC_ERR_NO_PORT;
2416 		if (tipc_msg_lookup_dest(net, skb, &err))
2417 			goto xmit;
2418 
2419 		/* Prepare for message rejection */
2420 		if (!tipc_msg_reverse(tipc_own_addr(net), &skb, err))
2421 			continue;
2422 
2423 		trace_tipc_sk_rej_msg(NULL, skb, TIPC_DUMP_NONE, "@sk_rcv!");
2424 xmit:
2425 		dnode = msg_destnode(buf_msg(skb));
2426 		tipc_node_xmit_skb(net, skb, dnode, dport);
2427 	}
2428 }
2429 
2430 static int tipc_wait_for_connect(struct socket *sock, long *timeo_p)
2431 {
2432 	DEFINE_WAIT_FUNC(wait, woken_wake_function);
2433 	struct sock *sk = sock->sk;
2434 	int done;
2435 
2436 	do {
2437 		int err = sock_error(sk);
2438 		if (err)
2439 			return err;
2440 		if (!*timeo_p)
2441 			return -ETIMEDOUT;
2442 		if (signal_pending(current))
2443 			return sock_intr_errno(*timeo_p);
2444 
2445 		add_wait_queue(sk_sleep(sk), &wait);
2446 		done = sk_wait_event(sk, timeo_p, tipc_sk_connected(sk),
2447 				     &wait);
2448 		remove_wait_queue(sk_sleep(sk), &wait);
2449 	} while (!done);
2450 	return 0;
2451 }
2452 
2453 static bool tipc_sockaddr_is_sane(struct sockaddr_tipc *addr)
2454 {
2455 	if (addr->family != AF_TIPC)
2456 		return false;
2457 	if (addr->addrtype == TIPC_SERVICE_RANGE)
2458 		return (addr->addr.nameseq.lower <= addr->addr.nameseq.upper);
2459 	return (addr->addrtype == TIPC_SERVICE_ADDR ||
2460 		addr->addrtype == TIPC_SOCKET_ADDR);
2461 }
2462 
2463 /**
2464  * tipc_connect - establish a connection to another TIPC port
2465  * @sock: socket structure
2466  * @dest: socket address for destination port
2467  * @destlen: size of socket address data structure
2468  * @flags: file-related flags associated with socket
2469  *
2470  * Returns 0 on success, errno otherwise
2471  */
2472 static int tipc_connect(struct socket *sock, struct sockaddr *dest,
2473 			int destlen, int flags)
2474 {
2475 	struct sock *sk = sock->sk;
2476 	struct tipc_sock *tsk = tipc_sk(sk);
2477 	struct sockaddr_tipc *dst = (struct sockaddr_tipc *)dest;
2478 	struct msghdr m = {NULL,};
2479 	long timeout = (flags & O_NONBLOCK) ? 0 : tsk->conn_timeout;
2480 	int previous;
2481 	int res = 0;
2482 
2483 	if (destlen != sizeof(struct sockaddr_tipc))
2484 		return -EINVAL;
2485 
2486 	lock_sock(sk);
2487 
2488 	if (tsk->group) {
2489 		res = -EINVAL;
2490 		goto exit;
2491 	}
2492 
2493 	if (dst->family == AF_UNSPEC) {
2494 		memset(&tsk->peer, 0, sizeof(struct sockaddr_tipc));
2495 		if (!tipc_sk_type_connectionless(sk))
2496 			res = -EINVAL;
2497 		goto exit;
2498 	}
2499 	if (!tipc_sockaddr_is_sane(dst)) {
2500 		res = -EINVAL;
2501 		goto exit;
2502 	}
2503 	/* DGRAM/RDM connect(), just save the destaddr */
2504 	if (tipc_sk_type_connectionless(sk)) {
2505 		memcpy(&tsk->peer, dest, destlen);
2506 		goto exit;
2507 	} else if (dst->addrtype == TIPC_SERVICE_RANGE) {
2508 		res = -EINVAL;
2509 		goto exit;
2510 	}
2511 
2512 	previous = sk->sk_state;
2513 
2514 	switch (sk->sk_state) {
2515 	case TIPC_OPEN:
2516 		/* Send a 'SYN-' to destination */
2517 		m.msg_name = dest;
2518 		m.msg_namelen = destlen;
2519 
2520 		/* If connect is in non-blocking case, set MSG_DONTWAIT to
2521 		 * indicate send_msg() is never blocked.
2522 		 */
2523 		if (!timeout)
2524 			m.msg_flags = MSG_DONTWAIT;
2525 
2526 		res = __tipc_sendmsg(sock, &m, 0);
2527 		if ((res < 0) && (res != -EWOULDBLOCK))
2528 			goto exit;
2529 
2530 		/* Just entered TIPC_CONNECTING state; the only
2531 		 * difference is that return value in non-blocking
2532 		 * case is EINPROGRESS, rather than EALREADY.
2533 		 */
2534 		res = -EINPROGRESS;
2535 		/* fall through */
2536 	case TIPC_CONNECTING:
2537 		if (!timeout) {
2538 			if (previous == TIPC_CONNECTING)
2539 				res = -EALREADY;
2540 			goto exit;
2541 		}
2542 		timeout = msecs_to_jiffies(timeout);
2543 		/* Wait until an 'ACK' or 'RST' arrives, or a timeout occurs */
2544 		res = tipc_wait_for_connect(sock, &timeout);
2545 		break;
2546 	case TIPC_ESTABLISHED:
2547 		res = -EISCONN;
2548 		break;
2549 	default:
2550 		res = -EINVAL;
2551 	}
2552 
2553 exit:
2554 	release_sock(sk);
2555 	return res;
2556 }
2557 
2558 /**
2559  * tipc_listen - allow socket to listen for incoming connections
2560  * @sock: socket structure
2561  * @len: (unused)
2562  *
2563  * Returns 0 on success, errno otherwise
2564  */
2565 static int tipc_listen(struct socket *sock, int len)
2566 {
2567 	struct sock *sk = sock->sk;
2568 	int res;
2569 
2570 	lock_sock(sk);
2571 	res = tipc_set_sk_state(sk, TIPC_LISTEN);
2572 	release_sock(sk);
2573 
2574 	return res;
2575 }
2576 
2577 static int tipc_wait_for_accept(struct socket *sock, long timeo)
2578 {
2579 	struct sock *sk = sock->sk;
2580 	DEFINE_WAIT(wait);
2581 	int err;
2582 
2583 	/* True wake-one mechanism for incoming connections: only
2584 	 * one process gets woken up, not the 'whole herd'.
2585 	 * Since we do not 'race & poll' for established sockets
2586 	 * anymore, the common case will execute the loop only once.
2587 	*/
2588 	for (;;) {
2589 		prepare_to_wait_exclusive(sk_sleep(sk), &wait,
2590 					  TASK_INTERRUPTIBLE);
2591 		if (timeo && skb_queue_empty(&sk->sk_receive_queue)) {
2592 			release_sock(sk);
2593 			timeo = schedule_timeout(timeo);
2594 			lock_sock(sk);
2595 		}
2596 		err = 0;
2597 		if (!skb_queue_empty(&sk->sk_receive_queue))
2598 			break;
2599 		err = -EAGAIN;
2600 		if (!timeo)
2601 			break;
2602 		err = sock_intr_errno(timeo);
2603 		if (signal_pending(current))
2604 			break;
2605 	}
2606 	finish_wait(sk_sleep(sk), &wait);
2607 	return err;
2608 }
2609 
2610 /**
2611  * tipc_accept - wait for connection request
2612  * @sock: listening socket
2613  * @newsock: new socket that is to be connected
2614  * @flags: file-related flags associated with socket
2615  *
2616  * Returns 0 on success, errno otherwise
2617  */
2618 static int tipc_accept(struct socket *sock, struct socket *new_sock, int flags,
2619 		       bool kern)
2620 {
2621 	struct sock *new_sk, *sk = sock->sk;
2622 	struct sk_buff *buf;
2623 	struct tipc_sock *new_tsock;
2624 	struct tipc_msg *msg;
2625 	long timeo;
2626 	int res;
2627 
2628 	lock_sock(sk);
2629 
2630 	if (sk->sk_state != TIPC_LISTEN) {
2631 		res = -EINVAL;
2632 		goto exit;
2633 	}
2634 	timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK);
2635 	res = tipc_wait_for_accept(sock, timeo);
2636 	if (res)
2637 		goto exit;
2638 
2639 	buf = skb_peek(&sk->sk_receive_queue);
2640 
2641 	res = tipc_sk_create(sock_net(sock->sk), new_sock, 0, kern);
2642 	if (res)
2643 		goto exit;
2644 	security_sk_clone(sock->sk, new_sock->sk);
2645 
2646 	new_sk = new_sock->sk;
2647 	new_tsock = tipc_sk(new_sk);
2648 	msg = buf_msg(buf);
2649 
2650 	/* we lock on new_sk; but lockdep sees the lock on sk */
2651 	lock_sock_nested(new_sk, SINGLE_DEPTH_NESTING);
2652 
2653 	/*
2654 	 * Reject any stray messages received by new socket
2655 	 * before the socket lock was taken (very, very unlikely)
2656 	 */
2657 	tsk_rej_rx_queue(new_sk, TIPC_ERR_NO_PORT);
2658 
2659 	/* Connect new socket to it's peer */
2660 	tipc_sk_finish_conn(new_tsock, msg_origport(msg), msg_orignode(msg));
2661 
2662 	tsk_set_importance(new_tsock, msg_importance(msg));
2663 	if (msg_named(msg)) {
2664 		new_tsock->conn_type = msg_nametype(msg);
2665 		new_tsock->conn_instance = msg_nameinst(msg);
2666 	}
2667 
2668 	/*
2669 	 * Respond to 'SYN-' by discarding it & returning 'ACK'-.
2670 	 * Respond to 'SYN+' by queuing it on new socket.
2671 	 */
2672 	if (!msg_data_sz(msg)) {
2673 		struct msghdr m = {NULL,};
2674 
2675 		tsk_advance_rx_queue(sk);
2676 		__tipc_sendstream(new_sock, &m, 0);
2677 	} else {
2678 		__skb_dequeue(&sk->sk_receive_queue);
2679 		__skb_queue_head(&new_sk->sk_receive_queue, buf);
2680 		skb_set_owner_r(buf, new_sk);
2681 	}
2682 	release_sock(new_sk);
2683 exit:
2684 	release_sock(sk);
2685 	return res;
2686 }
2687 
2688 /**
2689  * tipc_shutdown - shutdown socket connection
2690  * @sock: socket structure
2691  * @how: direction to close (must be SHUT_RDWR)
2692  *
2693  * Terminates connection (if necessary), then purges socket's receive queue.
2694  *
2695  * Returns 0 on success, errno otherwise
2696  */
2697 static int tipc_shutdown(struct socket *sock, int how)
2698 {
2699 	struct sock *sk = sock->sk;
2700 	int res;
2701 
2702 	if (how != SHUT_RDWR)
2703 		return -EINVAL;
2704 
2705 	lock_sock(sk);
2706 
2707 	trace_tipc_sk_shutdown(sk, NULL, TIPC_DUMP_ALL, " ");
2708 	__tipc_shutdown(sock, TIPC_CONN_SHUTDOWN);
2709 	sk->sk_shutdown = SEND_SHUTDOWN;
2710 
2711 	if (sk->sk_state == TIPC_DISCONNECTING) {
2712 		/* Discard any unreceived messages */
2713 		__skb_queue_purge(&sk->sk_receive_queue);
2714 
2715 		/* Wake up anyone sleeping in poll */
2716 		sk->sk_state_change(sk);
2717 		res = 0;
2718 	} else {
2719 		res = -ENOTCONN;
2720 	}
2721 
2722 	release_sock(sk);
2723 	return res;
2724 }
2725 
2726 static void tipc_sk_check_probing_state(struct sock *sk,
2727 					struct sk_buff_head *list)
2728 {
2729 	struct tipc_sock *tsk = tipc_sk(sk);
2730 	u32 pnode = tsk_peer_node(tsk);
2731 	u32 pport = tsk_peer_port(tsk);
2732 	u32 self = tsk_own_node(tsk);
2733 	u32 oport = tsk->portid;
2734 	struct sk_buff *skb;
2735 
2736 	if (tsk->probe_unacked) {
2737 		tipc_set_sk_state(sk, TIPC_DISCONNECTING);
2738 		sk->sk_err = ECONNABORTED;
2739 		tipc_node_remove_conn(sock_net(sk), pnode, pport);
2740 		sk->sk_state_change(sk);
2741 		return;
2742 	}
2743 	/* Prepare new probe */
2744 	skb = tipc_msg_create(CONN_MANAGER, CONN_PROBE, INT_H_SIZE, 0,
2745 			      pnode, self, pport, oport, TIPC_OK);
2746 	if (skb)
2747 		__skb_queue_tail(list, skb);
2748 	tsk->probe_unacked = true;
2749 	sk_reset_timer(sk, &sk->sk_timer, jiffies + CONN_PROBING_INTV);
2750 }
2751 
2752 static void tipc_sk_retry_connect(struct sock *sk, struct sk_buff_head *list)
2753 {
2754 	struct tipc_sock *tsk = tipc_sk(sk);
2755 
2756 	/* Try again later if dest link is congested */
2757 	if (tsk->cong_link_cnt) {
2758 		sk_reset_timer(sk, &sk->sk_timer, msecs_to_jiffies(100));
2759 		return;
2760 	}
2761 	/* Prepare SYN for retransmit */
2762 	tipc_msg_skb_clone(&sk->sk_write_queue, list);
2763 }
2764 
2765 static void tipc_sk_timeout(struct timer_list *t)
2766 {
2767 	struct sock *sk = from_timer(sk, t, sk_timer);
2768 	struct tipc_sock *tsk = tipc_sk(sk);
2769 	u32 pnode = tsk_peer_node(tsk);
2770 	struct sk_buff_head list;
2771 	int rc = 0;
2772 
2773 	__skb_queue_head_init(&list);
2774 	bh_lock_sock(sk);
2775 
2776 	/* Try again later if socket is busy */
2777 	if (sock_owned_by_user(sk)) {
2778 		sk_reset_timer(sk, &sk->sk_timer, jiffies + HZ / 20);
2779 		bh_unlock_sock(sk);
2780 		sock_put(sk);
2781 		return;
2782 	}
2783 
2784 	if (sk->sk_state == TIPC_ESTABLISHED)
2785 		tipc_sk_check_probing_state(sk, &list);
2786 	else if (sk->sk_state == TIPC_CONNECTING)
2787 		tipc_sk_retry_connect(sk, &list);
2788 
2789 	bh_unlock_sock(sk);
2790 
2791 	if (!skb_queue_empty(&list))
2792 		rc = tipc_node_xmit(sock_net(sk), &list, pnode, tsk->portid);
2793 
2794 	/* SYN messages may cause link congestion */
2795 	if (rc == -ELINKCONG) {
2796 		tipc_dest_push(&tsk->cong_links, pnode, 0);
2797 		tsk->cong_link_cnt = 1;
2798 	}
2799 	sock_put(sk);
2800 }
2801 
2802 static int tipc_sk_publish(struct tipc_sock *tsk, uint scope,
2803 			   struct tipc_name_seq const *seq)
2804 {
2805 	struct sock *sk = &tsk->sk;
2806 	struct net *net = sock_net(sk);
2807 	struct publication *publ;
2808 	u32 key;
2809 
2810 	if (scope != TIPC_NODE_SCOPE)
2811 		scope = TIPC_CLUSTER_SCOPE;
2812 
2813 	if (tipc_sk_connected(sk))
2814 		return -EINVAL;
2815 	key = tsk->portid + tsk->pub_count + 1;
2816 	if (key == tsk->portid)
2817 		return -EADDRINUSE;
2818 
2819 	publ = tipc_nametbl_publish(net, seq->type, seq->lower, seq->upper,
2820 				    scope, tsk->portid, key);
2821 	if (unlikely(!publ))
2822 		return -EINVAL;
2823 
2824 	list_add(&publ->binding_sock, &tsk->publications);
2825 	tsk->pub_count++;
2826 	tsk->published = 1;
2827 	return 0;
2828 }
2829 
2830 static int tipc_sk_withdraw(struct tipc_sock *tsk, uint scope,
2831 			    struct tipc_name_seq const *seq)
2832 {
2833 	struct net *net = sock_net(&tsk->sk);
2834 	struct publication *publ;
2835 	struct publication *safe;
2836 	int rc = -EINVAL;
2837 
2838 	if (scope != TIPC_NODE_SCOPE)
2839 		scope = TIPC_CLUSTER_SCOPE;
2840 
2841 	list_for_each_entry_safe(publ, safe, &tsk->publications, binding_sock) {
2842 		if (seq) {
2843 			if (publ->scope != scope)
2844 				continue;
2845 			if (publ->type != seq->type)
2846 				continue;
2847 			if (publ->lower != seq->lower)
2848 				continue;
2849 			if (publ->upper != seq->upper)
2850 				break;
2851 			tipc_nametbl_withdraw(net, publ->type, publ->lower,
2852 					      publ->upper, publ->key);
2853 			rc = 0;
2854 			break;
2855 		}
2856 		tipc_nametbl_withdraw(net, publ->type, publ->lower,
2857 				      publ->upper, publ->key);
2858 		rc = 0;
2859 	}
2860 	if (list_empty(&tsk->publications))
2861 		tsk->published = 0;
2862 	return rc;
2863 }
2864 
2865 /* tipc_sk_reinit: set non-zero address in all existing sockets
2866  *                 when we go from standalone to network mode.
2867  */
2868 void tipc_sk_reinit(struct net *net)
2869 {
2870 	struct tipc_net *tn = net_generic(net, tipc_net_id);
2871 	struct rhashtable_iter iter;
2872 	struct tipc_sock *tsk;
2873 	struct tipc_msg *msg;
2874 
2875 	rhashtable_walk_enter(&tn->sk_rht, &iter);
2876 
2877 	do {
2878 		rhashtable_walk_start(&iter);
2879 
2880 		while ((tsk = rhashtable_walk_next(&iter)) && !IS_ERR(tsk)) {
2881 			sock_hold(&tsk->sk);
2882 			rhashtable_walk_stop(&iter);
2883 			lock_sock(&tsk->sk);
2884 			msg = &tsk->phdr;
2885 			msg_set_prevnode(msg, tipc_own_addr(net));
2886 			msg_set_orignode(msg, tipc_own_addr(net));
2887 			release_sock(&tsk->sk);
2888 			rhashtable_walk_start(&iter);
2889 			sock_put(&tsk->sk);
2890 		}
2891 
2892 		rhashtable_walk_stop(&iter);
2893 	} while (tsk == ERR_PTR(-EAGAIN));
2894 
2895 	rhashtable_walk_exit(&iter);
2896 }
2897 
2898 static struct tipc_sock *tipc_sk_lookup(struct net *net, u32 portid)
2899 {
2900 	struct tipc_net *tn = net_generic(net, tipc_net_id);
2901 	struct tipc_sock *tsk;
2902 
2903 	rcu_read_lock();
2904 	tsk = rhashtable_lookup(&tn->sk_rht, &portid, tsk_rht_params);
2905 	if (tsk)
2906 		sock_hold(&tsk->sk);
2907 	rcu_read_unlock();
2908 
2909 	return tsk;
2910 }
2911 
2912 static int tipc_sk_insert(struct tipc_sock *tsk)
2913 {
2914 	struct sock *sk = &tsk->sk;
2915 	struct net *net = sock_net(sk);
2916 	struct tipc_net *tn = net_generic(net, tipc_net_id);
2917 	u32 remaining = (TIPC_MAX_PORT - TIPC_MIN_PORT) + 1;
2918 	u32 portid = prandom_u32() % remaining + TIPC_MIN_PORT;
2919 
2920 	while (remaining--) {
2921 		portid++;
2922 		if ((portid < TIPC_MIN_PORT) || (portid > TIPC_MAX_PORT))
2923 			portid = TIPC_MIN_PORT;
2924 		tsk->portid = portid;
2925 		sock_hold(&tsk->sk);
2926 		if (!rhashtable_lookup_insert_fast(&tn->sk_rht, &tsk->node,
2927 						   tsk_rht_params))
2928 			return 0;
2929 		sock_put(&tsk->sk);
2930 	}
2931 
2932 	return -1;
2933 }
2934 
2935 static void tipc_sk_remove(struct tipc_sock *tsk)
2936 {
2937 	struct sock *sk = &tsk->sk;
2938 	struct tipc_net *tn = net_generic(sock_net(sk), tipc_net_id);
2939 
2940 	if (!rhashtable_remove_fast(&tn->sk_rht, &tsk->node, tsk_rht_params)) {
2941 		WARN_ON(refcount_read(&sk->sk_refcnt) == 1);
2942 		__sock_put(sk);
2943 	}
2944 }
2945 
2946 static const struct rhashtable_params tsk_rht_params = {
2947 	.nelem_hint = 192,
2948 	.head_offset = offsetof(struct tipc_sock, node),
2949 	.key_offset = offsetof(struct tipc_sock, portid),
2950 	.key_len = sizeof(u32), /* portid */
2951 	.max_size = 1048576,
2952 	.min_size = 256,
2953 	.automatic_shrinking = true,
2954 };
2955 
2956 int tipc_sk_rht_init(struct net *net)
2957 {
2958 	struct tipc_net *tn = net_generic(net, tipc_net_id);
2959 
2960 	return rhashtable_init(&tn->sk_rht, &tsk_rht_params);
2961 }
2962 
2963 void tipc_sk_rht_destroy(struct net *net)
2964 {
2965 	struct tipc_net *tn = net_generic(net, tipc_net_id);
2966 
2967 	/* Wait for socket readers to complete */
2968 	synchronize_net();
2969 
2970 	rhashtable_destroy(&tn->sk_rht);
2971 }
2972 
2973 static int tipc_sk_join(struct tipc_sock *tsk, struct tipc_group_req *mreq)
2974 {
2975 	struct net *net = sock_net(&tsk->sk);
2976 	struct tipc_group *grp = tsk->group;
2977 	struct tipc_msg *hdr = &tsk->phdr;
2978 	struct tipc_name_seq seq;
2979 	int rc;
2980 
2981 	if (mreq->type < TIPC_RESERVED_TYPES)
2982 		return -EACCES;
2983 	if (mreq->scope > TIPC_NODE_SCOPE)
2984 		return -EINVAL;
2985 	if (grp)
2986 		return -EACCES;
2987 	grp = tipc_group_create(net, tsk->portid, mreq, &tsk->group_is_open);
2988 	if (!grp)
2989 		return -ENOMEM;
2990 	tsk->group = grp;
2991 	msg_set_lookup_scope(hdr, mreq->scope);
2992 	msg_set_nametype(hdr, mreq->type);
2993 	msg_set_dest_droppable(hdr, true);
2994 	seq.type = mreq->type;
2995 	seq.lower = mreq->instance;
2996 	seq.upper = seq.lower;
2997 	tipc_nametbl_build_group(net, grp, mreq->type, mreq->scope);
2998 	rc = tipc_sk_publish(tsk, mreq->scope, &seq);
2999 	if (rc) {
3000 		tipc_group_delete(net, grp);
3001 		tsk->group = NULL;
3002 		return rc;
3003 	}
3004 	/* Eliminate any risk that a broadcast overtakes sent JOINs */
3005 	tsk->mc_method.rcast = true;
3006 	tsk->mc_method.mandatory = true;
3007 	tipc_group_join(net, grp, &tsk->sk.sk_rcvbuf);
3008 	return rc;
3009 }
3010 
3011 static int tipc_sk_leave(struct tipc_sock *tsk)
3012 {
3013 	struct net *net = sock_net(&tsk->sk);
3014 	struct tipc_group *grp = tsk->group;
3015 	struct tipc_name_seq seq;
3016 	int scope;
3017 
3018 	if (!grp)
3019 		return -EINVAL;
3020 	tipc_group_self(grp, &seq, &scope);
3021 	tipc_group_delete(net, grp);
3022 	tsk->group = NULL;
3023 	tipc_sk_withdraw(tsk, scope, &seq);
3024 	return 0;
3025 }
3026 
3027 /**
3028  * tipc_setsockopt - set socket option
3029  * @sock: socket structure
3030  * @lvl: option level
3031  * @opt: option identifier
3032  * @ov: pointer to new option value
3033  * @ol: length of option value
3034  *
3035  * For stream sockets only, accepts and ignores all IPPROTO_TCP options
3036  * (to ease compatibility).
3037  *
3038  * Returns 0 on success, errno otherwise
3039  */
3040 static int tipc_setsockopt(struct socket *sock, int lvl, int opt,
3041 			   char __user *ov, unsigned int ol)
3042 {
3043 	struct sock *sk = sock->sk;
3044 	struct tipc_sock *tsk = tipc_sk(sk);
3045 	struct tipc_group_req mreq;
3046 	u32 value = 0;
3047 	int res = 0;
3048 
3049 	if ((lvl == IPPROTO_TCP) && (sock->type == SOCK_STREAM))
3050 		return 0;
3051 	if (lvl != SOL_TIPC)
3052 		return -ENOPROTOOPT;
3053 
3054 	switch (opt) {
3055 	case TIPC_IMPORTANCE:
3056 	case TIPC_SRC_DROPPABLE:
3057 	case TIPC_DEST_DROPPABLE:
3058 	case TIPC_CONN_TIMEOUT:
3059 	case TIPC_NODELAY:
3060 		if (ol < sizeof(value))
3061 			return -EINVAL;
3062 		if (get_user(value, (u32 __user *)ov))
3063 			return -EFAULT;
3064 		break;
3065 	case TIPC_GROUP_JOIN:
3066 		if (ol < sizeof(mreq))
3067 			return -EINVAL;
3068 		if (copy_from_user(&mreq, ov, sizeof(mreq)))
3069 			return -EFAULT;
3070 		break;
3071 	default:
3072 		if (ov || ol)
3073 			return -EINVAL;
3074 	}
3075 
3076 	lock_sock(sk);
3077 
3078 	switch (opt) {
3079 	case TIPC_IMPORTANCE:
3080 		res = tsk_set_importance(tsk, value);
3081 		break;
3082 	case TIPC_SRC_DROPPABLE:
3083 		if (sock->type != SOCK_STREAM)
3084 			tsk_set_unreliable(tsk, value);
3085 		else
3086 			res = -ENOPROTOOPT;
3087 		break;
3088 	case TIPC_DEST_DROPPABLE:
3089 		tsk_set_unreturnable(tsk, value);
3090 		break;
3091 	case TIPC_CONN_TIMEOUT:
3092 		tipc_sk(sk)->conn_timeout = value;
3093 		break;
3094 	case TIPC_MCAST_BROADCAST:
3095 		tsk->mc_method.rcast = false;
3096 		tsk->mc_method.mandatory = true;
3097 		break;
3098 	case TIPC_MCAST_REPLICAST:
3099 		tsk->mc_method.rcast = true;
3100 		tsk->mc_method.mandatory = true;
3101 		break;
3102 	case TIPC_GROUP_JOIN:
3103 		res = tipc_sk_join(tsk, &mreq);
3104 		break;
3105 	case TIPC_GROUP_LEAVE:
3106 		res = tipc_sk_leave(tsk);
3107 		break;
3108 	case TIPC_NODELAY:
3109 		tsk->nodelay = !!value;
3110 		tsk_set_nagle(tsk);
3111 		break;
3112 	default:
3113 		res = -EINVAL;
3114 	}
3115 
3116 	release_sock(sk);
3117 
3118 	return res;
3119 }
3120 
3121 /**
3122  * tipc_getsockopt - get socket option
3123  * @sock: socket structure
3124  * @lvl: option level
3125  * @opt: option identifier
3126  * @ov: receptacle for option value
3127  * @ol: receptacle for length of option value
3128  *
3129  * For stream sockets only, returns 0 length result for all IPPROTO_TCP options
3130  * (to ease compatibility).
3131  *
3132  * Returns 0 on success, errno otherwise
3133  */
3134 static int tipc_getsockopt(struct socket *sock, int lvl, int opt,
3135 			   char __user *ov, int __user *ol)
3136 {
3137 	struct sock *sk = sock->sk;
3138 	struct tipc_sock *tsk = tipc_sk(sk);
3139 	struct tipc_name_seq seq;
3140 	int len, scope;
3141 	u32 value;
3142 	int res;
3143 
3144 	if ((lvl == IPPROTO_TCP) && (sock->type == SOCK_STREAM))
3145 		return put_user(0, ol);
3146 	if (lvl != SOL_TIPC)
3147 		return -ENOPROTOOPT;
3148 	res = get_user(len, ol);
3149 	if (res)
3150 		return res;
3151 
3152 	lock_sock(sk);
3153 
3154 	switch (opt) {
3155 	case TIPC_IMPORTANCE:
3156 		value = tsk_importance(tsk);
3157 		break;
3158 	case TIPC_SRC_DROPPABLE:
3159 		value = tsk_unreliable(tsk);
3160 		break;
3161 	case TIPC_DEST_DROPPABLE:
3162 		value = tsk_unreturnable(tsk);
3163 		break;
3164 	case TIPC_CONN_TIMEOUT:
3165 		value = tsk->conn_timeout;
3166 		/* no need to set "res", since already 0 at this point */
3167 		break;
3168 	case TIPC_NODE_RECVQ_DEPTH:
3169 		value = 0; /* was tipc_queue_size, now obsolete */
3170 		break;
3171 	case TIPC_SOCK_RECVQ_DEPTH:
3172 		value = skb_queue_len(&sk->sk_receive_queue);
3173 		break;
3174 	case TIPC_SOCK_RECVQ_USED:
3175 		value = sk_rmem_alloc_get(sk);
3176 		break;
3177 	case TIPC_GROUP_JOIN:
3178 		seq.type = 0;
3179 		if (tsk->group)
3180 			tipc_group_self(tsk->group, &seq, &scope);
3181 		value = seq.type;
3182 		break;
3183 	default:
3184 		res = -EINVAL;
3185 	}
3186 
3187 	release_sock(sk);
3188 
3189 	if (res)
3190 		return res;	/* "get" failed */
3191 
3192 	if (len < sizeof(value))
3193 		return -EINVAL;
3194 
3195 	if (copy_to_user(ov, &value, sizeof(value)))
3196 		return -EFAULT;
3197 
3198 	return put_user(sizeof(value), ol);
3199 }
3200 
3201 static int tipc_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
3202 {
3203 	struct net *net = sock_net(sock->sk);
3204 	struct tipc_sioc_nodeid_req nr = {0};
3205 	struct tipc_sioc_ln_req lnr;
3206 	void __user *argp = (void __user *)arg;
3207 
3208 	switch (cmd) {
3209 	case SIOCGETLINKNAME:
3210 		if (copy_from_user(&lnr, argp, sizeof(lnr)))
3211 			return -EFAULT;
3212 		if (!tipc_node_get_linkname(net,
3213 					    lnr.bearer_id & 0xffff, lnr.peer,
3214 					    lnr.linkname, TIPC_MAX_LINK_NAME)) {
3215 			if (copy_to_user(argp, &lnr, sizeof(lnr)))
3216 				return -EFAULT;
3217 			return 0;
3218 		}
3219 		return -EADDRNOTAVAIL;
3220 	case SIOCGETNODEID:
3221 		if (copy_from_user(&nr, argp, sizeof(nr)))
3222 			return -EFAULT;
3223 		if (!tipc_node_get_id(net, nr.peer, nr.node_id))
3224 			return -EADDRNOTAVAIL;
3225 		if (copy_to_user(argp, &nr, sizeof(nr)))
3226 			return -EFAULT;
3227 		return 0;
3228 	default:
3229 		return -ENOIOCTLCMD;
3230 	}
3231 }
3232 
3233 static int tipc_socketpair(struct socket *sock1, struct socket *sock2)
3234 {
3235 	struct tipc_sock *tsk2 = tipc_sk(sock2->sk);
3236 	struct tipc_sock *tsk1 = tipc_sk(sock1->sk);
3237 	u32 onode = tipc_own_addr(sock_net(sock1->sk));
3238 
3239 	tsk1->peer.family = AF_TIPC;
3240 	tsk1->peer.addrtype = TIPC_ADDR_ID;
3241 	tsk1->peer.scope = TIPC_NODE_SCOPE;
3242 	tsk1->peer.addr.id.ref = tsk2->portid;
3243 	tsk1->peer.addr.id.node = onode;
3244 	tsk2->peer.family = AF_TIPC;
3245 	tsk2->peer.addrtype = TIPC_ADDR_ID;
3246 	tsk2->peer.scope = TIPC_NODE_SCOPE;
3247 	tsk2->peer.addr.id.ref = tsk1->portid;
3248 	tsk2->peer.addr.id.node = onode;
3249 
3250 	tipc_sk_finish_conn(tsk1, tsk2->portid, onode);
3251 	tipc_sk_finish_conn(tsk2, tsk1->portid, onode);
3252 	return 0;
3253 }
3254 
3255 /* Protocol switches for the various types of TIPC sockets */
3256 
3257 static const struct proto_ops msg_ops = {
3258 	.owner		= THIS_MODULE,
3259 	.family		= AF_TIPC,
3260 	.release	= tipc_release,
3261 	.bind		= tipc_bind,
3262 	.connect	= tipc_connect,
3263 	.socketpair	= tipc_socketpair,
3264 	.accept		= sock_no_accept,
3265 	.getname	= tipc_getname,
3266 	.poll		= tipc_poll,
3267 	.ioctl		= tipc_ioctl,
3268 	.listen		= sock_no_listen,
3269 	.shutdown	= tipc_shutdown,
3270 	.setsockopt	= tipc_setsockopt,
3271 	.getsockopt	= tipc_getsockopt,
3272 	.sendmsg	= tipc_sendmsg,
3273 	.recvmsg	= tipc_recvmsg,
3274 	.mmap		= sock_no_mmap,
3275 	.sendpage	= sock_no_sendpage
3276 };
3277 
3278 static const struct proto_ops packet_ops = {
3279 	.owner		= THIS_MODULE,
3280 	.family		= AF_TIPC,
3281 	.release	= tipc_release,
3282 	.bind		= tipc_bind,
3283 	.connect	= tipc_connect,
3284 	.socketpair	= tipc_socketpair,
3285 	.accept		= tipc_accept,
3286 	.getname	= tipc_getname,
3287 	.poll		= tipc_poll,
3288 	.ioctl		= tipc_ioctl,
3289 	.listen		= tipc_listen,
3290 	.shutdown	= tipc_shutdown,
3291 	.setsockopt	= tipc_setsockopt,
3292 	.getsockopt	= tipc_getsockopt,
3293 	.sendmsg	= tipc_send_packet,
3294 	.recvmsg	= tipc_recvmsg,
3295 	.mmap		= sock_no_mmap,
3296 	.sendpage	= sock_no_sendpage
3297 };
3298 
3299 static const struct proto_ops stream_ops = {
3300 	.owner		= THIS_MODULE,
3301 	.family		= AF_TIPC,
3302 	.release	= tipc_release,
3303 	.bind		= tipc_bind,
3304 	.connect	= tipc_connect,
3305 	.socketpair	= tipc_socketpair,
3306 	.accept		= tipc_accept,
3307 	.getname	= tipc_getname,
3308 	.poll		= tipc_poll,
3309 	.ioctl		= tipc_ioctl,
3310 	.listen		= tipc_listen,
3311 	.shutdown	= tipc_shutdown,
3312 	.setsockopt	= tipc_setsockopt,
3313 	.getsockopt	= tipc_getsockopt,
3314 	.sendmsg	= tipc_sendstream,
3315 	.recvmsg	= tipc_recvstream,
3316 	.mmap		= sock_no_mmap,
3317 	.sendpage	= sock_no_sendpage
3318 };
3319 
3320 static const struct net_proto_family tipc_family_ops = {
3321 	.owner		= THIS_MODULE,
3322 	.family		= AF_TIPC,
3323 	.create		= tipc_sk_create
3324 };
3325 
3326 static struct proto tipc_proto = {
3327 	.name		= "TIPC",
3328 	.owner		= THIS_MODULE,
3329 	.obj_size	= sizeof(struct tipc_sock),
3330 	.sysctl_rmem	= sysctl_tipc_rmem
3331 };
3332 
3333 /**
3334  * tipc_socket_init - initialize TIPC socket interface
3335  *
3336  * Returns 0 on success, errno otherwise
3337  */
3338 int tipc_socket_init(void)
3339 {
3340 	int res;
3341 
3342 	res = proto_register(&tipc_proto, 1);
3343 	if (res) {
3344 		pr_err("Failed to register TIPC protocol type\n");
3345 		goto out;
3346 	}
3347 
3348 	res = sock_register(&tipc_family_ops);
3349 	if (res) {
3350 		pr_err("Failed to register TIPC socket type\n");
3351 		proto_unregister(&tipc_proto);
3352 		goto out;
3353 	}
3354  out:
3355 	return res;
3356 }
3357 
3358 /**
3359  * tipc_socket_stop - stop TIPC socket interface
3360  */
3361 void tipc_socket_stop(void)
3362 {
3363 	sock_unregister(tipc_family_ops.family);
3364 	proto_unregister(&tipc_proto);
3365 }
3366 
3367 /* Caller should hold socket lock for the passed tipc socket. */
3368 static int __tipc_nl_add_sk_con(struct sk_buff *skb, struct tipc_sock *tsk)
3369 {
3370 	u32 peer_node;
3371 	u32 peer_port;
3372 	struct nlattr *nest;
3373 
3374 	peer_node = tsk_peer_node(tsk);
3375 	peer_port = tsk_peer_port(tsk);
3376 
3377 	nest = nla_nest_start_noflag(skb, TIPC_NLA_SOCK_CON);
3378 	if (!nest)
3379 		return -EMSGSIZE;
3380 
3381 	if (nla_put_u32(skb, TIPC_NLA_CON_NODE, peer_node))
3382 		goto msg_full;
3383 	if (nla_put_u32(skb, TIPC_NLA_CON_SOCK, peer_port))
3384 		goto msg_full;
3385 
3386 	if (tsk->conn_type != 0) {
3387 		if (nla_put_flag(skb, TIPC_NLA_CON_FLAG))
3388 			goto msg_full;
3389 		if (nla_put_u32(skb, TIPC_NLA_CON_TYPE, tsk->conn_type))
3390 			goto msg_full;
3391 		if (nla_put_u32(skb, TIPC_NLA_CON_INST, tsk->conn_instance))
3392 			goto msg_full;
3393 	}
3394 	nla_nest_end(skb, nest);
3395 
3396 	return 0;
3397 
3398 msg_full:
3399 	nla_nest_cancel(skb, nest);
3400 
3401 	return -EMSGSIZE;
3402 }
3403 
3404 static int __tipc_nl_add_sk_info(struct sk_buff *skb, struct tipc_sock
3405 			  *tsk)
3406 {
3407 	struct net *net = sock_net(skb->sk);
3408 	struct sock *sk = &tsk->sk;
3409 
3410 	if (nla_put_u32(skb, TIPC_NLA_SOCK_REF, tsk->portid) ||
3411 	    nla_put_u32(skb, TIPC_NLA_SOCK_ADDR, tipc_own_addr(net)))
3412 		return -EMSGSIZE;
3413 
3414 	if (tipc_sk_connected(sk)) {
3415 		if (__tipc_nl_add_sk_con(skb, tsk))
3416 			return -EMSGSIZE;
3417 	} else if (!list_empty(&tsk->publications)) {
3418 		if (nla_put_flag(skb, TIPC_NLA_SOCK_HAS_PUBL))
3419 			return -EMSGSIZE;
3420 	}
3421 	return 0;
3422 }
3423 
3424 /* Caller should hold socket lock for the passed tipc socket. */
3425 static int __tipc_nl_add_sk(struct sk_buff *skb, struct netlink_callback *cb,
3426 			    struct tipc_sock *tsk)
3427 {
3428 	struct nlattr *attrs;
3429 	void *hdr;
3430 
3431 	hdr = genlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq,
3432 			  &tipc_genl_family, NLM_F_MULTI, TIPC_NL_SOCK_GET);
3433 	if (!hdr)
3434 		goto msg_cancel;
3435 
3436 	attrs = nla_nest_start_noflag(skb, TIPC_NLA_SOCK);
3437 	if (!attrs)
3438 		goto genlmsg_cancel;
3439 
3440 	if (__tipc_nl_add_sk_info(skb, tsk))
3441 		goto attr_msg_cancel;
3442 
3443 	nla_nest_end(skb, attrs);
3444 	genlmsg_end(skb, hdr);
3445 
3446 	return 0;
3447 
3448 attr_msg_cancel:
3449 	nla_nest_cancel(skb, attrs);
3450 genlmsg_cancel:
3451 	genlmsg_cancel(skb, hdr);
3452 msg_cancel:
3453 	return -EMSGSIZE;
3454 }
3455 
3456 int tipc_nl_sk_walk(struct sk_buff *skb, struct netlink_callback *cb,
3457 		    int (*skb_handler)(struct sk_buff *skb,
3458 				       struct netlink_callback *cb,
3459 				       struct tipc_sock *tsk))
3460 {
3461 	struct rhashtable_iter *iter = (void *)cb->args[4];
3462 	struct tipc_sock *tsk;
3463 	int err;
3464 
3465 	rhashtable_walk_start(iter);
3466 	while ((tsk = rhashtable_walk_next(iter)) != NULL) {
3467 		if (IS_ERR(tsk)) {
3468 			err = PTR_ERR(tsk);
3469 			if (err == -EAGAIN) {
3470 				err = 0;
3471 				continue;
3472 			}
3473 			break;
3474 		}
3475 
3476 		sock_hold(&tsk->sk);
3477 		rhashtable_walk_stop(iter);
3478 		lock_sock(&tsk->sk);
3479 		err = skb_handler(skb, cb, tsk);
3480 		if (err) {
3481 			release_sock(&tsk->sk);
3482 			sock_put(&tsk->sk);
3483 			goto out;
3484 		}
3485 		release_sock(&tsk->sk);
3486 		rhashtable_walk_start(iter);
3487 		sock_put(&tsk->sk);
3488 	}
3489 	rhashtable_walk_stop(iter);
3490 out:
3491 	return skb->len;
3492 }
3493 EXPORT_SYMBOL(tipc_nl_sk_walk);
3494 
3495 int tipc_dump_start(struct netlink_callback *cb)
3496 {
3497 	return __tipc_dump_start(cb, sock_net(cb->skb->sk));
3498 }
3499 EXPORT_SYMBOL(tipc_dump_start);
3500 
3501 int __tipc_dump_start(struct netlink_callback *cb, struct net *net)
3502 {
3503 	/* tipc_nl_name_table_dump() uses cb->args[0...3]. */
3504 	struct rhashtable_iter *iter = (void *)cb->args[4];
3505 	struct tipc_net *tn = tipc_net(net);
3506 
3507 	if (!iter) {
3508 		iter = kmalloc(sizeof(*iter), GFP_KERNEL);
3509 		if (!iter)
3510 			return -ENOMEM;
3511 
3512 		cb->args[4] = (long)iter;
3513 	}
3514 
3515 	rhashtable_walk_enter(&tn->sk_rht, iter);
3516 	return 0;
3517 }
3518 
3519 int tipc_dump_done(struct netlink_callback *cb)
3520 {
3521 	struct rhashtable_iter *hti = (void *)cb->args[4];
3522 
3523 	rhashtable_walk_exit(hti);
3524 	kfree(hti);
3525 	return 0;
3526 }
3527 EXPORT_SYMBOL(tipc_dump_done);
3528 
3529 int tipc_sk_fill_sock_diag(struct sk_buff *skb, struct netlink_callback *cb,
3530 			   struct tipc_sock *tsk, u32 sk_filter_state,
3531 			   u64 (*tipc_diag_gen_cookie)(struct sock *sk))
3532 {
3533 	struct sock *sk = &tsk->sk;
3534 	struct nlattr *attrs;
3535 	struct nlattr *stat;
3536 
3537 	/*filter response w.r.t sk_state*/
3538 	if (!(sk_filter_state & (1 << sk->sk_state)))
3539 		return 0;
3540 
3541 	attrs = nla_nest_start_noflag(skb, TIPC_NLA_SOCK);
3542 	if (!attrs)
3543 		goto msg_cancel;
3544 
3545 	if (__tipc_nl_add_sk_info(skb, tsk))
3546 		goto attr_msg_cancel;
3547 
3548 	if (nla_put_u32(skb, TIPC_NLA_SOCK_TYPE, (u32)sk->sk_type) ||
3549 	    nla_put_u32(skb, TIPC_NLA_SOCK_TIPC_STATE, (u32)sk->sk_state) ||
3550 	    nla_put_u32(skb, TIPC_NLA_SOCK_INO, sock_i_ino(sk)) ||
3551 	    nla_put_u32(skb, TIPC_NLA_SOCK_UID,
3552 			from_kuid_munged(sk_user_ns(NETLINK_CB(cb->skb).sk),
3553 					 sock_i_uid(sk))) ||
3554 	    nla_put_u64_64bit(skb, TIPC_NLA_SOCK_COOKIE,
3555 			      tipc_diag_gen_cookie(sk),
3556 			      TIPC_NLA_SOCK_PAD))
3557 		goto attr_msg_cancel;
3558 
3559 	stat = nla_nest_start_noflag(skb, TIPC_NLA_SOCK_STAT);
3560 	if (!stat)
3561 		goto attr_msg_cancel;
3562 
3563 	if (nla_put_u32(skb, TIPC_NLA_SOCK_STAT_RCVQ,
3564 			skb_queue_len(&sk->sk_receive_queue)) ||
3565 	    nla_put_u32(skb, TIPC_NLA_SOCK_STAT_SENDQ,
3566 			skb_queue_len(&sk->sk_write_queue)) ||
3567 	    nla_put_u32(skb, TIPC_NLA_SOCK_STAT_DROP,
3568 			atomic_read(&sk->sk_drops)))
3569 		goto stat_msg_cancel;
3570 
3571 	if (tsk->cong_link_cnt &&
3572 	    nla_put_flag(skb, TIPC_NLA_SOCK_STAT_LINK_CONG))
3573 		goto stat_msg_cancel;
3574 
3575 	if (tsk_conn_cong(tsk) &&
3576 	    nla_put_flag(skb, TIPC_NLA_SOCK_STAT_CONN_CONG))
3577 		goto stat_msg_cancel;
3578 
3579 	nla_nest_end(skb, stat);
3580 
3581 	if (tsk->group)
3582 		if (tipc_group_fill_sock_diag(tsk->group, skb))
3583 			goto stat_msg_cancel;
3584 
3585 	nla_nest_end(skb, attrs);
3586 
3587 	return 0;
3588 
3589 stat_msg_cancel:
3590 	nla_nest_cancel(skb, stat);
3591 attr_msg_cancel:
3592 	nla_nest_cancel(skb, attrs);
3593 msg_cancel:
3594 	return -EMSGSIZE;
3595 }
3596 EXPORT_SYMBOL(tipc_sk_fill_sock_diag);
3597 
3598 int tipc_nl_sk_dump(struct sk_buff *skb, struct netlink_callback *cb)
3599 {
3600 	return tipc_nl_sk_walk(skb, cb, __tipc_nl_add_sk);
3601 }
3602 
3603 /* Caller should hold socket lock for the passed tipc socket. */
3604 static int __tipc_nl_add_sk_publ(struct sk_buff *skb,
3605 				 struct netlink_callback *cb,
3606 				 struct publication *publ)
3607 {
3608 	void *hdr;
3609 	struct nlattr *attrs;
3610 
3611 	hdr = genlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq,
3612 			  &tipc_genl_family, NLM_F_MULTI, TIPC_NL_PUBL_GET);
3613 	if (!hdr)
3614 		goto msg_cancel;
3615 
3616 	attrs = nla_nest_start_noflag(skb, TIPC_NLA_PUBL);
3617 	if (!attrs)
3618 		goto genlmsg_cancel;
3619 
3620 	if (nla_put_u32(skb, TIPC_NLA_PUBL_KEY, publ->key))
3621 		goto attr_msg_cancel;
3622 	if (nla_put_u32(skb, TIPC_NLA_PUBL_TYPE, publ->type))
3623 		goto attr_msg_cancel;
3624 	if (nla_put_u32(skb, TIPC_NLA_PUBL_LOWER, publ->lower))
3625 		goto attr_msg_cancel;
3626 	if (nla_put_u32(skb, TIPC_NLA_PUBL_UPPER, publ->upper))
3627 		goto attr_msg_cancel;
3628 
3629 	nla_nest_end(skb, attrs);
3630 	genlmsg_end(skb, hdr);
3631 
3632 	return 0;
3633 
3634 attr_msg_cancel:
3635 	nla_nest_cancel(skb, attrs);
3636 genlmsg_cancel:
3637 	genlmsg_cancel(skb, hdr);
3638 msg_cancel:
3639 	return -EMSGSIZE;
3640 }
3641 
3642 /* Caller should hold socket lock for the passed tipc socket. */
3643 static int __tipc_nl_list_sk_publ(struct sk_buff *skb,
3644 				  struct netlink_callback *cb,
3645 				  struct tipc_sock *tsk, u32 *last_publ)
3646 {
3647 	int err;
3648 	struct publication *p;
3649 
3650 	if (*last_publ) {
3651 		list_for_each_entry(p, &tsk->publications, binding_sock) {
3652 			if (p->key == *last_publ)
3653 				break;
3654 		}
3655 		if (p->key != *last_publ) {
3656 			/* We never set seq or call nl_dump_check_consistent()
3657 			 * this means that setting prev_seq here will cause the
3658 			 * consistence check to fail in the netlink callback
3659 			 * handler. Resulting in the last NLMSG_DONE message
3660 			 * having the NLM_F_DUMP_INTR flag set.
3661 			 */
3662 			cb->prev_seq = 1;
3663 			*last_publ = 0;
3664 			return -EPIPE;
3665 		}
3666 	} else {
3667 		p = list_first_entry(&tsk->publications, struct publication,
3668 				     binding_sock);
3669 	}
3670 
3671 	list_for_each_entry_from(p, &tsk->publications, binding_sock) {
3672 		err = __tipc_nl_add_sk_publ(skb, cb, p);
3673 		if (err) {
3674 			*last_publ = p->key;
3675 			return err;
3676 		}
3677 	}
3678 	*last_publ = 0;
3679 
3680 	return 0;
3681 }
3682 
3683 int tipc_nl_publ_dump(struct sk_buff *skb, struct netlink_callback *cb)
3684 {
3685 	int err;
3686 	u32 tsk_portid = cb->args[0];
3687 	u32 last_publ = cb->args[1];
3688 	u32 done = cb->args[2];
3689 	struct net *net = sock_net(skb->sk);
3690 	struct tipc_sock *tsk;
3691 
3692 	if (!tsk_portid) {
3693 		struct nlattr **attrs = genl_dumpit_info(cb)->attrs;
3694 		struct nlattr *sock[TIPC_NLA_SOCK_MAX + 1];
3695 
3696 		if (!attrs[TIPC_NLA_SOCK])
3697 			return -EINVAL;
3698 
3699 		err = nla_parse_nested_deprecated(sock, TIPC_NLA_SOCK_MAX,
3700 						  attrs[TIPC_NLA_SOCK],
3701 						  tipc_nl_sock_policy, NULL);
3702 		if (err)
3703 			return err;
3704 
3705 		if (!sock[TIPC_NLA_SOCK_REF])
3706 			return -EINVAL;
3707 
3708 		tsk_portid = nla_get_u32(sock[TIPC_NLA_SOCK_REF]);
3709 	}
3710 
3711 	if (done)
3712 		return 0;
3713 
3714 	tsk = tipc_sk_lookup(net, tsk_portid);
3715 	if (!tsk)
3716 		return -EINVAL;
3717 
3718 	lock_sock(&tsk->sk);
3719 	err = __tipc_nl_list_sk_publ(skb, cb, tsk, &last_publ);
3720 	if (!err)
3721 		done = 1;
3722 	release_sock(&tsk->sk);
3723 	sock_put(&tsk->sk);
3724 
3725 	cb->args[0] = tsk_portid;
3726 	cb->args[1] = last_publ;
3727 	cb->args[2] = done;
3728 
3729 	return skb->len;
3730 }
3731 
3732 /**
3733  * tipc_sk_filtering - check if a socket should be traced
3734  * @sk: the socket to be examined
3735  * @sysctl_tipc_sk_filter[]: the socket tuple for filtering,
3736  *  (portid, sock type, name type, name lower, name upper)
3737  *
3738  * Returns true if the socket meets the socket tuple data
3739  * (value 0 = 'any') or when there is no tuple set (all = 0),
3740  * otherwise false
3741  */
3742 bool tipc_sk_filtering(struct sock *sk)
3743 {
3744 	struct tipc_sock *tsk;
3745 	struct publication *p;
3746 	u32 _port, _sktype, _type, _lower, _upper;
3747 	u32 type = 0, lower = 0, upper = 0;
3748 
3749 	if (!sk)
3750 		return true;
3751 
3752 	tsk = tipc_sk(sk);
3753 
3754 	_port = sysctl_tipc_sk_filter[0];
3755 	_sktype = sysctl_tipc_sk_filter[1];
3756 	_type = sysctl_tipc_sk_filter[2];
3757 	_lower = sysctl_tipc_sk_filter[3];
3758 	_upper = sysctl_tipc_sk_filter[4];
3759 
3760 	if (!_port && !_sktype && !_type && !_lower && !_upper)
3761 		return true;
3762 
3763 	if (_port)
3764 		return (_port == tsk->portid);
3765 
3766 	if (_sktype && _sktype != sk->sk_type)
3767 		return false;
3768 
3769 	if (tsk->published) {
3770 		p = list_first_entry_or_null(&tsk->publications,
3771 					     struct publication, binding_sock);
3772 		if (p) {
3773 			type = p->type;
3774 			lower = p->lower;
3775 			upper = p->upper;
3776 		}
3777 	}
3778 
3779 	if (!tipc_sk_type_connectionless(sk)) {
3780 		type = tsk->conn_type;
3781 		lower = tsk->conn_instance;
3782 		upper = tsk->conn_instance;
3783 	}
3784 
3785 	if ((_type && _type != type) || (_lower && _lower != lower) ||
3786 	    (_upper && _upper != upper))
3787 		return false;
3788 
3789 	return true;
3790 }
3791 
3792 u32 tipc_sock_get_portid(struct sock *sk)
3793 {
3794 	return (sk) ? (tipc_sk(sk))->portid : 0;
3795 }
3796 
3797 /**
3798  * tipc_sk_overlimit1 - check if socket rx queue is about to be overloaded,
3799  *			both the rcv and backlog queues are considered
3800  * @sk: tipc sk to be checked
3801  * @skb: tipc msg to be checked
3802  *
3803  * Returns true if the socket rx queue allocation is > 90%, otherwise false
3804  */
3805 
3806 bool tipc_sk_overlimit1(struct sock *sk, struct sk_buff *skb)
3807 {
3808 	atomic_t *dcnt = &tipc_sk(sk)->dupl_rcvcnt;
3809 	unsigned int lim = rcvbuf_limit(sk, skb) + atomic_read(dcnt);
3810 	unsigned int qsize = sk->sk_backlog.len + sk_rmem_alloc_get(sk);
3811 
3812 	return (qsize > lim * 90 / 100);
3813 }
3814 
3815 /**
3816  * tipc_sk_overlimit2 - check if socket rx queue is about to be overloaded,
3817  *			only the rcv queue is considered
3818  * @sk: tipc sk to be checked
3819  * @skb: tipc msg to be checked
3820  *
3821  * Returns true if the socket rx queue allocation is > 90%, otherwise false
3822  */
3823 
3824 bool tipc_sk_overlimit2(struct sock *sk, struct sk_buff *skb)
3825 {
3826 	unsigned int lim = rcvbuf_limit(sk, skb);
3827 	unsigned int qsize = sk_rmem_alloc_get(sk);
3828 
3829 	return (qsize > lim * 90 / 100);
3830 }
3831 
3832 /**
3833  * tipc_sk_dump - dump TIPC socket
3834  * @sk: tipc sk to be dumped
3835  * @dqueues: bitmask to decide if any socket queue to be dumped?
3836  *           - TIPC_DUMP_NONE: don't dump socket queues
3837  *           - TIPC_DUMP_SK_SNDQ: dump socket send queue
3838  *           - TIPC_DUMP_SK_RCVQ: dump socket rcv queue
3839  *           - TIPC_DUMP_SK_BKLGQ: dump socket backlog queue
3840  *           - TIPC_DUMP_ALL: dump all the socket queues above
3841  * @buf: returned buffer of dump data in format
3842  */
3843 int tipc_sk_dump(struct sock *sk, u16 dqueues, char *buf)
3844 {
3845 	int i = 0;
3846 	size_t sz = (dqueues) ? SK_LMAX : SK_LMIN;
3847 	struct tipc_sock *tsk;
3848 	struct publication *p;
3849 	bool tsk_connected;
3850 
3851 	if (!sk) {
3852 		i += scnprintf(buf, sz, "sk data: (null)\n");
3853 		return i;
3854 	}
3855 
3856 	tsk = tipc_sk(sk);
3857 	tsk_connected = !tipc_sk_type_connectionless(sk);
3858 
3859 	i += scnprintf(buf, sz, "sk data: %u", sk->sk_type);
3860 	i += scnprintf(buf + i, sz - i, " %d", sk->sk_state);
3861 	i += scnprintf(buf + i, sz - i, " %x", tsk_own_node(tsk));
3862 	i += scnprintf(buf + i, sz - i, " %u", tsk->portid);
3863 	i += scnprintf(buf + i, sz - i, " | %u", tsk_connected);
3864 	if (tsk_connected) {
3865 		i += scnprintf(buf + i, sz - i, " %x", tsk_peer_node(tsk));
3866 		i += scnprintf(buf + i, sz - i, " %u", tsk_peer_port(tsk));
3867 		i += scnprintf(buf + i, sz - i, " %u", tsk->conn_type);
3868 		i += scnprintf(buf + i, sz - i, " %u", tsk->conn_instance);
3869 	}
3870 	i += scnprintf(buf + i, sz - i, " | %u", tsk->published);
3871 	if (tsk->published) {
3872 		p = list_first_entry_or_null(&tsk->publications,
3873 					     struct publication, binding_sock);
3874 		i += scnprintf(buf + i, sz - i, " %u", (p) ? p->type : 0);
3875 		i += scnprintf(buf + i, sz - i, " %u", (p) ? p->lower : 0);
3876 		i += scnprintf(buf + i, sz - i, " %u", (p) ? p->upper : 0);
3877 	}
3878 	i += scnprintf(buf + i, sz - i, " | %u", tsk->snd_win);
3879 	i += scnprintf(buf + i, sz - i, " %u", tsk->rcv_win);
3880 	i += scnprintf(buf + i, sz - i, " %u", tsk->max_pkt);
3881 	i += scnprintf(buf + i, sz - i, " %x", tsk->peer_caps);
3882 	i += scnprintf(buf + i, sz - i, " %u", tsk->cong_link_cnt);
3883 	i += scnprintf(buf + i, sz - i, " %u", tsk->snt_unacked);
3884 	i += scnprintf(buf + i, sz - i, " %u", tsk->rcv_unacked);
3885 	i += scnprintf(buf + i, sz - i, " %u", atomic_read(&tsk->dupl_rcvcnt));
3886 	i += scnprintf(buf + i, sz - i, " %u", sk->sk_shutdown);
3887 	i += scnprintf(buf + i, sz - i, " | %d", sk_wmem_alloc_get(sk));
3888 	i += scnprintf(buf + i, sz - i, " %d", sk->sk_sndbuf);
3889 	i += scnprintf(buf + i, sz - i, " | %d", sk_rmem_alloc_get(sk));
3890 	i += scnprintf(buf + i, sz - i, " %d", sk->sk_rcvbuf);
3891 	i += scnprintf(buf + i, sz - i, " | %d\n", READ_ONCE(sk->sk_backlog.len));
3892 
3893 	if (dqueues & TIPC_DUMP_SK_SNDQ) {
3894 		i += scnprintf(buf + i, sz - i, "sk_write_queue: ");
3895 		i += tipc_list_dump(&sk->sk_write_queue, false, buf + i);
3896 	}
3897 
3898 	if (dqueues & TIPC_DUMP_SK_RCVQ) {
3899 		i += scnprintf(buf + i, sz - i, "sk_receive_queue: ");
3900 		i += tipc_list_dump(&sk->sk_receive_queue, false, buf + i);
3901 	}
3902 
3903 	if (dqueues & TIPC_DUMP_SK_BKLGQ) {
3904 		i += scnprintf(buf + i, sz - i, "sk_backlog:\n  head ");
3905 		i += tipc_skb_dump(sk->sk_backlog.head, false, buf + i);
3906 		if (sk->sk_backlog.tail != sk->sk_backlog.head) {
3907 			i += scnprintf(buf + i, sz - i, "  tail ");
3908 			i += tipc_skb_dump(sk->sk_backlog.tail, false,
3909 					   buf + i);
3910 		}
3911 	}
3912 
3913 	return i;
3914 }
3915