summaryrefslogtreecommitdiff
path: root/net/rxrpc/transport.c
blob: 465efc86fccf1f253cf93a4e5705017e6e1dd83d (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
/* transport.c: Rx Transport routines
 *
 * Copyright (C) 2002 Red Hat, Inc. All Rights Reserved.
 * Written by David Howells (dhowells@redhat.com)
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version
 * 2 of the License, or (at your option) any later version.
 */

#include <linux/sched.h>
#include <linux/slab.h>
#include <linux/module.h>
#include <rxrpc/transport.h>
#include <rxrpc/peer.h>
#include <rxrpc/connection.h>
#include <rxrpc/call.h>
#include <rxrpc/message.h>
#include <rxrpc/krxiod.h>
#include <rxrpc/krxsecd.h>
#include <linux/udp.h>
#include <linux/in.h>
#include <linux/in6.h>
#include <linux/icmp.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <net/ip.h>
#if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE)
#include <linux/ipv6.h>	/* this should _really_ be in errqueue.h.. */
#endif
#include <linux/errqueue.h>
#include <asm/uaccess.h>
#include <asm/checksum.h>
#include "internal.h"

struct errormsg {
	struct cmsghdr			cmsg;		/* control message header */
	struct sock_extended_err	ee;		/* extended error information */
	struct sockaddr_in		icmp_src;	/* ICMP packet source address */
};

static DEFINE_SPINLOCK(rxrpc_transports_lock);
static struct list_head rxrpc_transports = LIST_HEAD_INIT(rxrpc_transports);

__RXACCT_DECL(atomic_t rxrpc_transport_count);
LIST_HEAD(rxrpc_proc_transports);
DECLARE_RWSEM(rxrpc_proc_transports_sem);

static void rxrpc_data_ready(struct sock *sk, int count);
static void rxrpc_error_report(struct sock *sk);
static int rxrpc_trans_receive_new_call(struct rxrpc_transport *trans,
					struct list_head *msgq);
static void rxrpc_trans_receive_error_report(struct rxrpc_transport *trans);

/*****************************************************************************/
/*
 * create a new transport endpoint using the specified UDP port
 */
int rxrpc_create_transport(unsigned short port,
			   struct rxrpc_transport **_trans)
{
	struct rxrpc_transport *trans;
	struct sockaddr_in sin;
	mm_segment_t oldfs;
	struct sock *sock;
	int ret, opt;

	_enter("%hu", port);

	trans = kzalloc(sizeof(struct rxrpc_transport), GFP_KERNEL);
	if (!trans)
		return -ENOMEM;

	atomic_set(&trans->usage, 1);
	INIT_LIST_HEAD(&trans->services);
	INIT_LIST_HEAD(&trans->link);
	INIT_LIST_HEAD(&trans->krxiodq_link);
	spin_lock_init(&trans->lock);
	INIT_LIST_HEAD(&trans->peer_active);
	INIT_LIST_HEAD(&trans->peer_graveyard);
	spin_lock_init(&trans->peer_gylock);
	init_waitqueue_head(&trans->peer_gy_waitq);
	rwlock_init(&trans->peer_lock);
	atomic_set(&trans->peer_count, 0);
	trans->port = port;

	/* create a UDP socket to be my actual transport endpoint */
	ret = sock_create_kern(PF_INET, SOCK_DGRAM, IPPROTO_UDP, &trans->socket);
	if (ret < 0)
		goto error;

	/* use the specified port */
	if (port) {
		memset(&sin, 0, sizeof(sin));
		sin.sin_family = AF_INET;
		sin.sin_port = htons(port);
		ret = trans->socket->ops->bind(trans->socket,
					       (struct sockaddr *) &sin,
					       sizeof(sin));
		if (ret < 0)
			goto error;
	}

	opt = 1;
	oldfs = get_fs();
	set_fs(KERNEL_DS);
	ret = trans->socket->ops->setsockopt(trans->socket, SOL_IP, IP_RECVERR,
					     (char *) &opt, sizeof(opt));
	set_fs(oldfs);

	spin_lock(&rxrpc_transports_lock);
	list_add(&trans->link, &rxrpc_transports);
	spin_unlock(&rxrpc_transports_lock);

	/* set the socket up */
	sock = trans->socket->sk;
	sock->sk_user_data	= trans;
	sock->sk_data_ready	= rxrpc_data_ready;
	sock->sk_error_report	= rxrpc_error_report;

	down_write(&rxrpc_proc_transports_sem);
	list_add_tail(&trans->proc_link, &rxrpc_proc_transports);
	up_write(&rxrpc_proc_transports_sem);

	__RXACCT(atomic_inc(&rxrpc_transport_count));

	*_trans = trans;
	_leave(" = 0 (%p)", trans);
	return 0;

 error:
	/* finish cleaning up the transport (not really needed here, but...) */
	if (trans->socket)
		trans->socket->ops->shutdown(trans->socket, 2);

	/* close the socket */
	if (trans->socket) {
		trans->socket->sk->sk_user_data = NULL;
		sock_release(trans->socket);
		trans->socket = NULL;
	}

	kfree(trans);


	_leave(" = %d", ret);
	return ret;
} /* end rxrpc_create_transport() */

/*****************************************************************************/
/*
 * destroy a transport endpoint
 */
void rxrpc_put_transport(struct rxrpc_transport *trans)
{
	_enter("%p{u=%d p=%hu}",
	       trans, atomic_read(&trans->usage), trans->port);

	BUG_ON(atomic_read(&trans->usage) <= 0);

	/* to prevent a race, the decrement and the dequeue must be
	 * effectively atomic */
	spin_lock(&rxrpc_transports_lock);
	if (likely(!atomic_dec_and_test(&trans->usage))) {
		spin_unlock(&rxrpc_transports_lock);
		_leave("");
		return;
	}

	list_del(&trans->link);
	spin_unlock(&rxrpc_transports_lock);

	/* finish cleaning up the transport */
	if (trans->socket)
		trans->socket->ops->shutdown(trans->socket, 2);

	rxrpc_krxsecd_clear_transport(trans);
	rxrpc_krxiod_dequeue_transport(trans);

	/* discard all peer information */
	rxrpc_peer_clearall(trans);

	down_write(&rxrpc_proc_transports_sem);
	list_del(&trans->proc_link);
	up_write(&rxrpc_proc_transports_sem);
	__RXACCT(atomic_dec(&rxrpc_transport_count));

	/* close the socket */
	if (trans->socket) {
		trans->socket->sk->sk_user_data = NULL;
		sock_release(trans->socket);
		trans->socket = NULL;
	}

	kfree(trans);

	_leave("");
} /* end rxrpc_put_transport() */

/*****************************************************************************/
/*
 * add a service to a transport to be listened upon
 */
int rxrpc_add_service(struct rxrpc_transport *trans,
		      struct rxrpc_service *newsrv)
{
	struct rxrpc_service *srv;
	struct list_head *_p;
	int ret = -EEXIST;

	_enter("%p{%hu},%p{%hu}",
	       trans, trans->port, newsrv, newsrv->service_id);

	/* verify that the service ID is not already present */
	spin_lock(&trans->lock);

	list_for_each(_p, &trans->services) {
		srv = list_entry(_p, struct rxrpc_service, link);
		if (srv->service_id == newsrv->service_id)
			goto out;
	}

	/* okay - add the transport to the list */
	list_add_tail(&newsrv->link, &trans->services);
	rxrpc_get_transport(trans);
	ret = 0;

 out:
	spin_unlock(&trans->lock);

	_leave("= %d", ret);
	return ret;
} /* end rxrpc_add_service() */

/*****************************************************************************/
/*
 * remove a service from a transport
 */
void rxrpc_del_service(struct rxrpc_transport *trans, struct rxrpc_service *srv)
{
	_enter("%p{%hu},%p{%hu}", trans, trans->port, srv, srv->service_id);

	spin_lock(&trans->lock);
	list_del(&srv->link);
	spin_unlock(&trans->lock);

	rxrpc_put_transport(trans);

	_leave("");
} /* end rxrpc_del_service() */

/*****************************************************************************/
/*
 * INET callback when data has been received on the socket.
 */
static void rxrpc_data_ready(struct sock *sk, int count)
{
	struct rxrpc_transport *trans;

	_enter("%p{t=%p},%d", sk, sk->sk_user_data, count);

	/* queue the transport for attention by krxiod */
	trans = (struct rxrpc_transport *) sk->sk_user_data;
	if (trans)
		rxrpc_krxiod_queue_transport(trans);

	/* wake up anyone waiting on the socket */
	if (sk->sk_sleep && waitqueue_active(sk->sk_sleep))
		wake_up_interruptible(sk->sk_sleep);

	_leave("");
} /* end rxrpc_data_ready() */

/*****************************************************************************/
/*
 * INET callback when an ICMP error packet is received
 * - sk->err is error (EHOSTUNREACH, EPROTO or EMSGSIZE)
 */
static void rxrpc_error_report(struct sock *sk)
{
	struct rxrpc_transport *trans;

	_enter("%p{t=%p}", sk, sk->sk_user_data);

	/* queue the transport for attention by krxiod */
	trans = (struct rxrpc_transport *) sk->sk_user_data;
	if (trans) {
		trans->error_rcvd = 1;
		rxrpc_krxiod_queue_transport(trans);
	}

	/* wake up anyone waiting on the socket */
	if (sk->sk_sleep && waitqueue_active(sk->sk_sleep))
		wake_up_interruptible(sk->sk_sleep);

	_leave("");
} /* end rxrpc_error_report() */

/*****************************************************************************/
/*
 * split a message up, allocating message records and filling them in
 * from the contents of a socket buffer
 */
static int rxrpc_incoming_msg(struct rxrpc_transport *trans,
			      struct sk_buff *pkt,
			      struct list_head *msgq)
{
	struct rxrpc_message *msg;
	int ret;

	_enter("");

	msg = kzalloc(sizeof(struct rxrpc_message), GFP_KERNEL);
	if (!msg) {
		_leave(" = -ENOMEM");
		return -ENOMEM;
	}

	atomic_set(&msg->usage, 1);
	list_add_tail(&msg->link,msgq);

	/* dig out the Rx routing parameters */
	if (skb_copy_bits(pkt, sizeof(struct udphdr),
			  &msg->hdr, sizeof(msg->hdr)) < 0) {
		ret = -EBADMSG;
		goto error;
	}

	msg->trans = trans;
	msg->state = RXRPC_MSG_RECEIVED;
	skb_get_timestamp(pkt, &msg->stamp);
	if (msg->stamp.tv_sec == 0) {
		do_gettimeofday(&msg->stamp); 
		if (pkt->sk) 
			sock_enable_timestamp(pkt->sk);
	} 
	msg->seq = ntohl(msg->hdr.seq);

	/* attach the packet */
	skb_get(pkt);
	msg->pkt = pkt;

	msg->offset = sizeof(struct udphdr) + sizeof(struct rxrpc_header);
	msg->dsize = msg->pkt->len - msg->offset;

	_net("Rx Received packet from %s (%08x;%08x,%1x,%d,%s,%02x,%d,%d)",
	     msg->hdr.flags & RXRPC_CLIENT_INITIATED ? "client" : "server",
	     ntohl(msg->hdr.epoch),
	     (ntohl(msg->hdr.cid) & RXRPC_CIDMASK) >> RXRPC_CIDSHIFT,
	     ntohl(msg->hdr.cid) & RXRPC_CHANNELMASK,
	     ntohl(msg->hdr.callNumber),
	     rxrpc_pkts[msg->hdr.type],
	     msg->hdr.flags,
	     ntohs(msg->hdr.serviceId),
	     msg->hdr.securityIndex);

	__RXACCT(atomic_inc(&rxrpc_message_count));

	/* split off jumbo packets */
	while (msg->hdr.type == RXRPC_PACKET_TYPE_DATA &&
	       msg->hdr.flags & RXRPC_JUMBO_PACKET
	       ) {
		struct rxrpc_jumbo_header jumbo;
		struct rxrpc_message *jumbomsg = msg;

		_debug("split jumbo packet");

		/* quick sanity check */
		ret = -EBADMSG;
		if (msg->dsize <
		    RXRPC_JUMBO_DATALEN + sizeof(struct rxrpc_jumbo_header))
			goto error;
		if (msg->hdr.flags & RXRPC_LAST_PACKET)
			goto error;

		/* dig out the secondary header */
		if (skb_copy_bits(pkt, msg->offset + RXRPC_JUMBO_DATALEN,
				  &jumbo, sizeof(jumbo)) < 0)
			goto error;

		/* allocate a new message record */
		ret = -ENOMEM;
		msg = kmalloc(sizeof(struct rxrpc_message), GFP_KERNEL);
		if (!msg)
			goto error;

		memcpy(msg, jumbomsg, sizeof(*msg));
		list_add_tail(&msg->link, msgq);

		/* adjust the jumbo packet */
		jumbomsg->dsize = RXRPC_JUMBO_DATALEN;

		/* attach the packet here too */
		skb_get(pkt);

		/* adjust the parameters */
		msg->seq++;
		msg->hdr.seq = htonl(msg->seq);
		msg->hdr.serial = htonl(ntohl(msg->hdr.serial) + 1);
		msg->offset += RXRPC_JUMBO_DATALEN +
			sizeof(struct rxrpc_jumbo_header);
		msg->dsize -= RXRPC_JUMBO_DATALEN +
			sizeof(struct rxrpc_jumbo_header);
		msg->hdr.flags = jumbo.flags;
		msg->hdr._rsvd = jumbo._rsvd;

		_net("Rx Split jumbo packet from %s"
		     " (%08x;%08x,%1x,%d,%s,%02x,%d,%d)",
		     msg->hdr.flags & RXRPC_CLIENT_INITIATED ? "client" : "server",
		     ntohl(msg->hdr.epoch),
		     (ntohl(msg->hdr.cid) & RXRPC_CIDMASK) >> RXRPC_CIDSHIFT,
		     ntohl(msg->hdr.cid) & RXRPC_CHANNELMASK,
		     ntohl(msg->hdr.callNumber),
		     rxrpc_pkts[msg->hdr.type],
		     msg->hdr.flags,
		     ntohs(msg->hdr.serviceId),
		     msg->hdr.securityIndex);

		__RXACCT(atomic_inc(&rxrpc_message_count));
	}

	_leave(" = 0 #%d", atomic_read(&rxrpc_message_count));
	return 0;

 error:
	while (!list_empty(msgq)) {
		msg = list_entry(msgq->next, struct rxrpc_message, link);
		list_del_init(&msg->link);

		rxrpc_put_message(msg);
	}

	_leave(" = %d", ret);
	return ret;
} /* end rxrpc_incoming_msg() */

/*****************************************************************************/
/*
 * accept a new call
 * - called from krxiod in process context
 */
void rxrpc_trans_receive_packet(struct rxrpc_transport *trans)
{
	struct rxrpc_message *msg;
	struct rxrpc_peer *peer;
	struct sk_buff *pkt;
	int ret;
	__be32 addr;
	__be16 port;

	LIST_HEAD(msgq);

	_enter("%p{%d}", trans, trans->port);

	for (;;) {
		/* deal with outstanting errors first */
		if (trans->error_rcvd)
			rxrpc_trans_receive_error_report(trans);

		/* attempt to receive a packet */
		pkt = skb_recv_datagram(trans->socket->sk, 0, 1, &ret);
		if (!pkt) {
			if (ret == -EAGAIN) {
				_leave(" EAGAIN");
				return;
			}

			/* an icmp error may have occurred */
			rxrpc_krxiod_queue_transport(trans);
			_leave(" error %d\n", ret);
			return;
		}

		/* we'll probably need to checksum it (didn't call
		 * sock_recvmsg) */
		if (skb_checksum_complete(pkt)) {
			kfree_skb(pkt);
			rxrpc_krxiod_queue_transport(trans);
			_leave(" CSUM failed");
			return;
		}

		addr = pkt->nh.iph->saddr;
		port = pkt->h.uh->source;

		_net("Rx Received UDP packet from %08x:%04hu",
		     ntohl(addr), ntohs(port));

		/* unmarshall the Rx parameters and split jumbo packets */
		ret = rxrpc_incoming_msg(trans, pkt, &msgq);
		if (ret < 0) {
			kfree_skb(pkt);
			rxrpc_krxiod_queue_transport(trans);
			_leave(" bad packet");
			return;
		}

		BUG_ON(list_empty(&msgq));

		msg = list_entry(msgq.next, struct rxrpc_message, link);

		/* locate the record for the peer from which it
		 * originated */
		ret = rxrpc_peer_lookup(trans, addr, &peer);
		if (ret < 0) {
			kdebug("Rx No connections from that peer");
			rxrpc_trans_immediate_abort(trans, msg, -EINVAL);
			goto finished_msg;
		}

		/* try and find a matching connection */
		ret = rxrpc_connection_lookup(peer, msg, &msg->conn);
		if (ret < 0) {
			kdebug("Rx Unknown Connection");
			rxrpc_trans_immediate_abort(trans, msg, -EINVAL);
			rxrpc_put_peer(peer);
			goto finished_msg;
		}
		rxrpc_put_peer(peer);

		/* deal with the first packet of a new call */
		if (msg->hdr.flags & RXRPC_CLIENT_INITIATED &&
		    msg->hdr.type == RXRPC_PACKET_TYPE_DATA &&
		    ntohl(msg->hdr.seq) == 1
		    ) {
			_debug("Rx New server call");
			rxrpc_trans_receive_new_call(trans, &msgq);
			goto finished_msg;
		}

		/* deal with subsequent packet(s) of call */
		_debug("Rx Call packet");
		while (!list_empty(&msgq)) {
			msg = list_entry(msgq.next, struct rxrpc_message, link);
			list_del_init(&msg->link);

			ret = rxrpc_conn_receive_call_packet(msg->conn, NULL, msg);
			if (ret < 0) {
				rxrpc_trans_immediate_abort(trans, msg, ret);
				rxrpc_put_message(msg);
				goto finished_msg;
			}

			rxrpc_put_message(msg);
		}

		goto finished_msg;

		/* dispose of the packets */
	finished_msg:
		while (!list_empty(&msgq)) {
			msg = list_entry(msgq.next, struct rxrpc_message, link);
			list_del_init(&msg->link);

			rxrpc_put_message(msg);
		}
		kfree_skb(pkt);
	}

	_leave("");

} /* end rxrpc_trans_receive_packet() */

/*****************************************************************************/
/*
 * accept a new call from a client trying to connect to one of my services
 * - called in process context
 */
static int rxrpc_trans_receive_new_call(struct rxrpc_transport *trans,
					struct list_head *msgq)
{
	struct rxrpc_message *msg;

	_enter("");

	/* only bother with the first packet */
	msg = list_entry(msgq->next, struct rxrpc_message, link);
	list_del_init(&msg->link);
	rxrpc_krxsecd_queue_incoming_call(msg);
	rxrpc_put_message(msg);

	_leave(" = 0");

	return 0;
} /* end rxrpc_trans_receive_new_call() */

/*****************************************************************************/
/*
 * perform an immediate abort without connection or call structures
 */
int rxrpc_trans_immediate_abort(struct rxrpc_transport *trans,
				struct rxrpc_message *msg,
				int error)
{
	struct rxrpc_header ahdr;
	struct sockaddr_in sin;
	struct msghdr msghdr;
	struct kvec iov[2];
	__be32 _error;
	int len, ret;

	_enter("%p,%p,%d", trans, msg, error);

	/* don't abort an abort packet */
	if (msg->hdr.type == RXRPC_PACKET_TYPE_ABORT) {
		_leave(" = 0");
		return 0;
	}

	_error = htonl(-error);

	/* set up the message to be transmitted */
	memcpy(&ahdr, &msg->hdr, sizeof(ahdr));
	ahdr.epoch	= msg->hdr.epoch;
	ahdr.serial	= htonl(1);
	ahdr.seq	= 0;
	ahdr.type	= RXRPC_PACKET_TYPE_ABORT;
	ahdr.flags	= RXRPC_LAST_PACKET;
	ahdr.flags	|= ~msg->hdr.flags & RXRPC_CLIENT_INITIATED;

	iov[0].iov_len	= sizeof(ahdr);
	iov[0].iov_base	= &ahdr;
	iov[1].iov_len	= sizeof(_error);
	iov[1].iov_base	= &_error;

	len = sizeof(ahdr) + sizeof(_error);

	memset(&sin,0,sizeof(sin));
	sin.sin_family		= AF_INET;
	sin.sin_port		= msg->pkt->h.uh->source;
	sin.sin_addr.s_addr	= msg->pkt->nh.iph->saddr;

	msghdr.msg_name		= &sin;
	msghdr.msg_namelen	= sizeof(sin);
	msghdr.msg_control	= NULL;
	msghdr.msg_controllen	= 0;
	msghdr.msg_flags	= MSG_DONTWAIT;

	_net("Sending message type %d of %d bytes to %08x:%d",
	     ahdr.type,
	     len,
	     ntohl(sin.sin_addr.s_addr),
	     ntohs(sin.sin_port));

	/* send the message */
	ret = kernel_sendmsg(trans->socket, &msghdr, iov, 2, len);

	_leave(" = %d", ret);
	return ret;
} /* end rxrpc_trans_immediate_abort() */

/*****************************************************************************/
/*
 * receive an ICMP error report and percolate it to all connections
 * heading to the affected host or port
 */
static void rxrpc_trans_receive_error_report(struct rxrpc_transport *trans)
{
	struct rxrpc_connection *conn;
	struct sockaddr_in sin;
	struct rxrpc_peer *peer;
	struct list_head connq, *_p;
	struct errormsg emsg;
	struct msghdr msg;
	__be16 port;
	int local, err;

	_enter("%p", trans);

	for (;;) {
		trans->error_rcvd = 0;

		/* try and receive an error message */
		msg.msg_name	= &sin;
		msg.msg_namelen	= sizeof(sin);
		msg.msg_control	= &emsg;
		msg.msg_controllen = sizeof(emsg);
		msg.msg_flags	= 0;

		err = kernel_recvmsg(trans->socket, &msg, NULL, 0, 0,
				   MSG_ERRQUEUE | MSG_DONTWAIT | MSG_TRUNC);

		if (err == -EAGAIN) {
			_leave("");
			return;
		}

		if (err < 0) {
			printk("%s: unable to recv an error report: %d\n",
			       __FUNCTION__, err);
			_leave("");
			return;
		}

		msg.msg_controllen = (char *) msg.msg_control - (char *) &emsg;

		if (msg.msg_controllen < sizeof(emsg.cmsg) ||
		    msg.msg_namelen < sizeof(sin)) {
			printk("%s: short control message"
			       " (nlen=%u clen=%Zu fl=%x)\n",
			       __FUNCTION__,
			       msg.msg_namelen,
			       msg.msg_controllen,
			       msg.msg_flags);
			continue;
		}

		_net("Rx Received control message"
		     " { len=%Zu level=%u type=%u }",
		     emsg.cmsg.cmsg_len,
		     emsg.cmsg.cmsg_level,
		     emsg.cmsg.cmsg_type);

		if (sin.sin_family != AF_INET) {
			printk("Rx Ignoring error report with non-INET address"
			       " (fam=%u)",
			       sin.sin_family);
			continue;
		}

		_net("Rx Received message pertaining to host addr=%x port=%hu",
		     ntohl(sin.sin_addr.s_addr), ntohs(sin.sin_port));

		if (emsg.cmsg.cmsg_level != SOL_IP ||
		    emsg.cmsg.cmsg_type != IP_RECVERR) {
			printk("Rx Ignoring unknown error report"
			       " { level=%u type=%u }",
			       emsg.cmsg.cmsg_level,
			       emsg.cmsg.cmsg_type);
			continue;
		}

		if (msg.msg_controllen < sizeof(emsg.cmsg) + sizeof(emsg.ee)) {
			printk("%s: short error message (%Zu)\n",
			       __FUNCTION__, msg.msg_controllen);
			_leave("");
			return;
		}

		port = sin.sin_port;

		switch (emsg.ee.ee_origin) {
		case SO_EE_ORIGIN_ICMP:
			local = 0;
			switch (emsg.ee.ee_type) {
			case ICMP_DEST_UNREACH:
				switch (emsg.ee.ee_code) {
				case ICMP_NET_UNREACH:
					_net("Rx Received ICMP Network Unreachable");
					port = 0;
					err = -ENETUNREACH;
					break;
				case ICMP_HOST_UNREACH:
					_net("Rx Received ICMP Host Unreachable");
					port = 0;
					err = -EHOSTUNREACH;
					break;
				case ICMP_PORT_UNREACH:
					_net("Rx Received ICMP Port Unreachable");
					err = -ECONNREFUSED;
					break;
				case ICMP_NET_UNKNOWN:
					_net("Rx Received ICMP Unknown Network");
					port = 0;
					err = -ENETUNREACH;
					break;
				case ICMP_HOST_UNKNOWN:
					_net("Rx Received ICMP Unknown Host");
					port = 0;
					err = -EHOSTUNREACH;
					break;
				default:
					_net("Rx Received ICMP DestUnreach { code=%u }",
					     emsg.ee.ee_code);
					err = emsg.ee.ee_errno;
					break;
				}
				break;

			case ICMP_TIME_EXCEEDED:
				_net("Rx Received ICMP TTL Exceeded");
				err = emsg.ee.ee_errno;
				break;

			default:
				_proto("Rx Received ICMP error { type=%u code=%u }",
				       emsg.ee.ee_type, emsg.ee.ee_code);
				err = emsg.ee.ee_errno;
				break;
			}
			break;

		case SO_EE_ORIGIN_LOCAL:
			_proto("Rx Received local error { error=%d }",
			       emsg.ee.ee_errno);
			local = 1;
			err = emsg.ee.ee_errno;
			break;

		case SO_EE_ORIGIN_NONE:
		case SO_EE_ORIGIN_ICMP6:
		default:
			_proto("Rx Received error report { orig=%u }",
			       emsg.ee.ee_origin);
			local = 0;
			err = emsg.ee.ee_errno;
			break;
		}

		/* find all the connections between this transport and the
		 * affected destination */
		INIT_LIST_HEAD(&connq);

		if (rxrpc_peer_lookup(trans, sin.sin_addr.s_addr,
				      &peer) == 0) {
			read_lock(&peer->conn_lock);
			list_for_each(_p, &peer->conn_active) {
				conn = list_entry(_p, struct rxrpc_connection,
						  link);
				if (port && conn->addr.sin_port != port)
					continue;
				if (!list_empty(&conn->err_link))
					continue;

				rxrpc_get_connection(conn);
				list_add_tail(&conn->err_link, &connq);
			}
			read_unlock(&peer->conn_lock);

			/* service all those connections */
			while (!list_empty(&connq)) {
				conn = list_entry(connq.next,
						  struct rxrpc_connection,
						  err_link);
				list_del(&conn->err_link);

				rxrpc_conn_handle_error(conn, local, err);

				rxrpc_put_connection(conn);
			}

			rxrpc_put_peer(peer);
		}
	}

	_leave("");
	return;
} /* end rxrpc_trans_receive_error_report() */