summaryrefslogtreecommitdiffstats
path: root/src/proto_quic.c
blob: d4a5a13d702975e4a7ad980c56eb0c54148438f3 (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
/*
 * AF_INET/AF_INET6 QUIC protocol layer.
 *
 * Copyright 2020 Frederic Lecaille <flecaille@haproxy.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 <ctype.h>
#include <errno.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>

#include <sys/param.h>
#include <sys/socket.h>
#include <sys/types.h>

#include <netinet/udp.h>
#include <netinet/in.h>

#include <import/ebtree-t.h>

#include <haproxy/api.h>
#include <haproxy/arg.h>
#include <haproxy/cbuf.h>
#include <haproxy/connection.h>
#include <haproxy/errors.h>
#include <haproxy/fd.h>
#include <haproxy/global.h>
#include <haproxy/list.h>
#include <haproxy/listener.h>
#include <haproxy/log.h>
#include <haproxy/namespace.h>
#include <haproxy/port_range.h>
#include <haproxy/protocol.h>
#include <haproxy/proto_quic.h>
#include <haproxy/proto_udp.h>
#include <haproxy/proxy-t.h>
#include <haproxy/quic_conn.h>
#include <haproxy/quic_sock.h>
#include <haproxy/sock.h>
#include <haproxy/sock_inet.h>
#include <haproxy/task.h>
#include <haproxy/tools.h>

/* per-thread quic datagram handlers */
struct quic_dghdlr *quic_dghdlrs;

/* Size of the internal buffer of QUIC RX buffer at the fd level */
#define QUIC_RX_BUFSZ  (1UL << 18)

DECLARE_STATIC_POOL(pool_head_quic_rxbuf, "quic_rxbuf", QUIC_RX_BUFSZ);

static void quic_add_listener(struct protocol *proto, struct listener *listener);
static int quic_bind_listener(struct listener *listener, char *errmsg, int errlen);
static int quic_connect_server(struct connection *conn, int flags);
static void quic_enable_listener(struct listener *listener);
static void quic_disable_listener(struct listener *listener);

/* Note: must not be declared <const> as its list will be overwritten */
struct protocol proto_quic4 = {
	.name           = "quic4",

	/* connection layer */
	.xprt_type      = PROTO_TYPE_STREAM,
	.listen         = quic_bind_listener,
	.enable         = quic_enable_listener,
	.disable        = quic_disable_listener,
	.add            = quic_add_listener,
	.unbind         = default_unbind_listener,
	.suspend        = default_suspend_listener,
	.resume         = default_resume_listener,
	.accept_conn    = quic_sock_accept_conn,
	.get_src        = quic_sock_get_src,
	.get_dst        = quic_sock_get_dst,
	.connect        = quic_connect_server,

	/* binding layer */
	.rx_suspend     = udp_suspend_receiver,
	.rx_resume      = udp_resume_receiver,

	/* address family */
	.fam            = &proto_fam_inet4,

	/* socket layer */
	.proto_type     = PROTO_TYPE_DGRAM,
	.sock_type      = SOCK_DGRAM,
	.sock_prot      = IPPROTO_UDP,
	.rx_enable      = sock_enable,
	.rx_disable     = sock_disable,
	.rx_unbind      = sock_unbind,
	.rx_listening   = quic_sock_accepting_conn,
	.default_iocb   = quic_sock_fd_iocb,
	.receivers      = LIST_HEAD_INIT(proto_quic4.receivers),
	.nb_receivers   = 0,
};

INITCALL1(STG_REGISTER, protocol_register, &proto_quic4);

/* Note: must not be declared <const> as its list will be overwritten */
struct protocol proto_quic6 = {
	.name            = "quic6",

	/* connection layer */
	.xprt_type      = PROTO_TYPE_STREAM,
	.listen         = quic_bind_listener,
	.enable         = quic_enable_listener,
	.disable        = quic_disable_listener,
	.add            = quic_add_listener,
	.unbind         = default_unbind_listener,
	.suspend        = default_suspend_listener,
	.resume         = default_resume_listener,
	.accept_conn    = quic_sock_accept_conn,
	.get_src        = quic_sock_get_src,
	.get_dst        = quic_sock_get_dst,
	.connect        = quic_connect_server,

	/* binding layer */
	.rx_suspend     = udp_suspend_receiver,
	.rx_resume      = udp_resume_receiver,

	/* address family */
	.fam            = &proto_fam_inet6,

	/* socket layer */
	.proto_type     = PROTO_TYPE_DGRAM,
	.sock_type      = SOCK_DGRAM,
	.sock_prot      = IPPROTO_UDP,
	.rx_enable      = sock_enable,
	.rx_disable     = sock_disable,
	.rx_unbind      = sock_unbind,
	.rx_listening   = quic_sock_accepting_conn,
	.default_iocb   = quic_sock_fd_iocb,
	.receivers      = LIST_HEAD_INIT(proto_quic6.receivers),
	.nb_receivers   = 0,
};

INITCALL1(STG_REGISTER, protocol_register, &proto_quic6);

/* Binds ipv4/ipv6 address <local> to socket <fd>, unless <flags> is set, in which
 * case we try to bind <remote>. <flags> is a 2-bit field consisting of :
 *  - 0 : ignore remote address (may even be a NULL pointer)
 *  - 1 : use provided address
 *  - 2 : use provided port
 *  - 3 : use both
 *
 * The function supports multiple foreign binding methods :
 *   - linux_tproxy: we directly bind to the foreign address
 * The second one can be used as a fallback for the first one.
 * This function returns 0 when everything's OK, 1 if it could not bind, to the
 * local address, 2 if it could not bind to the foreign address.
 */
int quic_bind_socket(int fd, int flags, struct sockaddr_storage *local, struct sockaddr_storage *remote)
{
	struct sockaddr_storage bind_addr;
	int foreign_ok = 0;
	int ret;
	static THREAD_LOCAL int ip_transp_working = 1;
	static THREAD_LOCAL int ip6_transp_working = 1;

	switch (local->ss_family) {
	case AF_INET:
		if (flags && ip_transp_working) {
			/* This deserves some explanation. Some platforms will support
			 * multiple combinations of certain methods, so we try the
			 * supported ones until one succeeds.
			 */
			if (sock_inet4_make_foreign(fd))
				foreign_ok = 1;
			else
				ip_transp_working = 0;
		}
		break;
	case AF_INET6:
		if (flags && ip6_transp_working) {
			if (sock_inet6_make_foreign(fd))
				foreign_ok = 1;
			else
				ip6_transp_working = 0;
		}
		break;
	}

	if (flags) {
		memset(&bind_addr, 0, sizeof(bind_addr));
		bind_addr.ss_family = remote->ss_family;
		switch (remote->ss_family) {
		case AF_INET:
			if (flags & 1)
				((struct sockaddr_in *)&bind_addr)->sin_addr = ((struct sockaddr_in *)remote)->sin_addr;
			if (flags & 2)
				((struct sockaddr_in *)&bind_addr)->sin_port = ((struct sockaddr_in *)remote)->sin_port;
			break;
		case AF_INET6:
			if (flags & 1)
				((struct sockaddr_in6 *)&bind_addr)->sin6_addr = ((struct sockaddr_in6 *)remote)->sin6_addr;
			if (flags & 2)
				((struct sockaddr_in6 *)&bind_addr)->sin6_port = ((struct sockaddr_in6 *)remote)->sin6_port;
			break;
		default:
			/* we don't want to try to bind to an unknown address family */
			foreign_ok = 0;
		}
	}

	setsockopt(fd, SOL_SOCKET, SO_REUSEADDR, &one, sizeof(one));
	if (foreign_ok) {
		if (is_inet_addr(&bind_addr)) {
			ret = bind(fd, (struct sockaddr *)&bind_addr, get_addr_len(&bind_addr));
			if (ret < 0)
				return 2;
		}
	}
	else {
		if (is_inet_addr(local)) {
			ret = bind(fd, (struct sockaddr *)local, get_addr_len(local));
			if (ret < 0)
				return 1;
		}
	}

	if (!flags)
		return 0;

	if (!foreign_ok)
		/* we could not bind to a foreign address */
		return 2;

	return 0;
}

/*
 * This function initiates a QUIC connection establishment to the target assigned
 * to connection <conn> using (si->{target,dst}). A source address may be
 * pointed to by conn->src in case of transparent proxying. Normal source
 * bind addresses are still determined locally (due to the possible need of a
 * source port). conn->target may point either to a valid server or to a backend,
 * depending on conn->target. Only OBJ_TYPE_PROXY and OBJ_TYPE_SERVER are
 * supported. The <data> parameter is a boolean indicating whether there are data
 * waiting for being sent or not, in order to adjust data write polling and on
 * some platforms, the ability to avoid an empty initial ACK. The <flags> argument
 * is not used.
 *
 * Note that a pending send_proxy message accounts for data.
 *
 * It can return one of :
 *  - SF_ERR_NONE if everything's OK
 *  - SF_ERR_SRVTO if there are no more servers
 *  - SF_ERR_SRVCL if the connection was refused by the server
 *  - SF_ERR_PRXCOND if the connection has been limited by the proxy (maxconn)
 *  - SF_ERR_RESOURCE if a system resource is lacking (eg: fd limits, ports, ...)
 *  - SF_ERR_INTERNAL for any other purely internal errors
 * Additionally, in the case of SF_ERR_RESOURCE, an emergency log will be emitted.
 *
 * The connection's fd is inserted only when SF_ERR_NONE is returned, otherwise
 * it's invalid and the caller has nothing to do.
 */

int quic_connect_server(struct connection *conn, int flags)
{
	int fd;
	struct server *srv;
	struct proxy *be;
	struct conn_src *src;
	struct sockaddr_storage *addr;

	BUG_ON(!conn->dst);

	conn->flags |= CO_FL_WAIT_L4_CONN; /* connection in progress */

	switch (obj_type(conn->target)) {
	case OBJ_TYPE_PROXY:
		be = __objt_proxy(conn->target);
		srv = NULL;
		break;
	case OBJ_TYPE_SERVER:
		srv = __objt_server(conn->target);
		be = srv->proxy;
		break;
	default:
		conn->flags |= CO_FL_ERROR;
		return SF_ERR_INTERNAL;
	}

	fd = conn->handle.fd = sock_create_server_socket(conn);

	if (fd == -1) {
		qfprintf(stderr, "Cannot get a server socket.\n");

		if (errno == ENFILE) {
			conn->err_code = CO_ER_SYS_FDLIM;
			send_log(be, LOG_EMERG,
				 "Proxy %s reached system FD limit (maxsock=%d). Please check system tunables.\n",
				 be->id, global.maxsock);
		}
		else if (errno == EMFILE) {
			conn->err_code = CO_ER_PROC_FDLIM;
			send_log(be, LOG_EMERG,
				 "Proxy %s reached process FD limit (maxsock=%d). Please check 'ulimit-n' and restart.\n",
				 be->id, global.maxsock);
		}
		else if (errno == ENOBUFS || errno == ENOMEM) {
			conn->err_code = CO_ER_SYS_MEMLIM;
			send_log(be, LOG_EMERG,
				 "Proxy %s reached system memory limit (maxsock=%d). Please check system tunables.\n",
				 be->id, global.maxsock);
		}
		else if (errno == EAFNOSUPPORT || errno == EPROTONOSUPPORT) {
			conn->err_code = CO_ER_NOPROTO;
		}
		else
			conn->err_code = CO_ER_SOCK_ERR;

		/* this is a resource error */
		conn->flags |= CO_FL_ERROR;
		return SF_ERR_RESOURCE;
	}

	if (fd >= global.maxsock) {
		/* do not log anything there, it's a normal condition when this option
		 * is used to serialize connections to a server !
		 */
		ha_alert("socket(): not enough free sockets. Raise -n argument. Giving up.\n");
		close(fd);
		conn->err_code = CO_ER_CONF_FDLIM;
		conn->flags |= CO_FL_ERROR;
		return SF_ERR_PRXCOND; /* it is a configuration limit */
	}

	if (fd_set_nonblock(fd) == -1) {
		qfprintf(stderr,"Cannot set client socket to non blocking mode.\n");
		close(fd);
		conn->err_code = CO_ER_SOCK_ERR;
		conn->flags |= CO_FL_ERROR;
		return SF_ERR_INTERNAL;
	}

	if (master == 1 && fd_set_cloexec(fd) == -1) {
		ha_alert("Cannot set CLOEXEC on client socket.\n");
		close(fd);
		conn->err_code = CO_ER_SOCK_ERR;
		conn->flags |= CO_FL_ERROR;
		return SF_ERR_INTERNAL;
	}

	/* allow specific binding :
	 * - server-specific at first
	 * - proxy-specific next
	 */
	if (srv && srv->conn_src.opts & CO_SRC_BIND)
		src = &srv->conn_src;
	else if (be->conn_src.opts & CO_SRC_BIND)
		src = &be->conn_src;
	else
		src = NULL;

	if (src) {
		int ret, flags = 0;

		if (conn->src && is_inet_addr(conn->src)) {
			switch (src->opts & CO_SRC_TPROXY_MASK) {
			case CO_SRC_TPROXY_CLI:
				conn_set_private(conn);
				/* fall through */
			case CO_SRC_TPROXY_ADDR:
				flags = 3;
				break;
			case CO_SRC_TPROXY_CIP:
			case CO_SRC_TPROXY_DYN:
				conn_set_private(conn);
				flags = 1;
				break;
			}
		}

#ifdef SO_BINDTODEVICE
		/* Note: this might fail if not CAP_NET_RAW */
		if (src->iface_name)
			setsockopt(fd, SOL_SOCKET, SO_BINDTODEVICE, src->iface_name, src->iface_len + 1);
#endif

		if (src->sport_range) {
			int attempts = 10; /* should be more than enough to find a spare port */
			struct sockaddr_storage sa;

			ret = 1;
			memcpy(&sa, &src->source_addr, sizeof(sa));

			do {
				/* note: in case of retry, we may have to release a previously
				 * allocated port, hence this loop's construct.
				 */
				port_range_release_port(fdinfo[fd].port_range, fdinfo[fd].local_port);
				fdinfo[fd].port_range = NULL;

				if (!attempts)
					break;
				attempts--;

				fdinfo[fd].local_port = port_range_alloc_port(src->sport_range);
				if (!fdinfo[fd].local_port) {
					conn->err_code = CO_ER_PORT_RANGE;
					break;
				}

				fdinfo[fd].port_range = src->sport_range;
				set_host_port(&sa, fdinfo[fd].local_port);

				ret = quic_bind_socket(fd, flags, &sa, conn->src);
				if (ret != 0)
					conn->err_code = CO_ER_CANT_BIND;
			} while (ret != 0); /* binding NOK */
		}
		else {
#ifdef IP_BIND_ADDRESS_NO_PORT
			static THREAD_LOCAL int bind_address_no_port = 1;
			setsockopt(fd, IPPROTO_IP, IP_BIND_ADDRESS_NO_PORT, (const void *) &bind_address_no_port, sizeof(int));
#endif
			ret = quic_bind_socket(fd, flags, &src->source_addr, conn->src);
			if (ret != 0)
				conn->err_code = CO_ER_CANT_BIND;
		}

		if (unlikely(ret != 0)) {
			port_range_release_port(fdinfo[fd].port_range, fdinfo[fd].local_port);
			fdinfo[fd].port_range = NULL;
			close(fd);

			if (ret == 1) {
				ha_alert("Cannot bind to source address before connect() for backend %s. Aborting.\n",
					 be->id);
				send_log(be, LOG_EMERG,
					 "Cannot bind to source address before connect() for backend %s.\n",
					 be->id);
			} else {
				ha_alert("Cannot bind to tproxy source address before connect() for backend %s. Aborting.\n",
					 be->id);
				send_log(be, LOG_EMERG,
					 "Cannot bind to tproxy source address before connect() for backend %s.\n",
					 be->id);
			}
			conn->flags |= CO_FL_ERROR;
			return SF_ERR_RESOURCE;
		}
	}

	if (global.tune.server_sndbuf)
                setsockopt(fd, SOL_SOCKET, SO_SNDBUF, &global.tune.server_sndbuf, sizeof(global.tune.server_sndbuf));

	if (global.tune.server_rcvbuf)
                setsockopt(fd, SOL_SOCKET, SO_RCVBUF, &global.tune.server_rcvbuf, sizeof(global.tune.server_rcvbuf));

	addr = (conn->flags & CO_FL_SOCKS4) ? &srv->socks4_addr : conn->dst;
	if (connect(fd, (const struct sockaddr *)addr, get_addr_len(addr)) == -1) {
		if (errno == EINPROGRESS || errno == EALREADY) {
			/* common case, let's wait for connect status */
			conn->flags |= CO_FL_WAIT_L4_CONN;
		}
		else if (errno == EISCONN) {
			/* should normally not happen but if so, indicates that it's OK */
			conn->flags &= ~CO_FL_WAIT_L4_CONN;
		}
		else if (errno == EAGAIN || errno == EWOULDBLOCK || errno == EADDRINUSE || errno == EADDRNOTAVAIL) {
			char *msg;
			if (errno == EAGAIN || errno == EWOULDBLOCK || errno == EADDRNOTAVAIL) {
				msg = "no free ports";
				conn->err_code = CO_ER_FREE_PORTS;
			}
			else {
				msg = "local address already in use";
				conn->err_code = CO_ER_ADDR_INUSE;
			}

			qfprintf(stderr,"Connect() failed for backend %s: %s.\n", be->id, msg);
			port_range_release_port(fdinfo[fd].port_range, fdinfo[fd].local_port);
			fdinfo[fd].port_range = NULL;
			close(fd);
			send_log(be, LOG_ERR, "Connect() failed for backend %s: %s.\n", be->id, msg);
			conn->flags |= CO_FL_ERROR;
			return SF_ERR_RESOURCE;
		} else if (errno == ETIMEDOUT) {
			//qfprintf(stderr,"Connect(): ETIMEDOUT");
			port_range_release_port(fdinfo[fd].port_range, fdinfo[fd].local_port);
			fdinfo[fd].port_range = NULL;
			close(fd);
			conn->err_code = CO_ER_SOCK_ERR;
			conn->flags |= CO_FL_ERROR;
			return SF_ERR_SRVTO;
		} else {
			// (errno == ECONNREFUSED || errno == ENETUNREACH || errno == EACCES || errno == EPERM)
			//qfprintf(stderr,"Connect(): %d", errno);
			port_range_release_port(fdinfo[fd].port_range, fdinfo[fd].local_port);
			fdinfo[fd].port_range = NULL;
			close(fd);
			conn->err_code = CO_ER_SOCK_ERR;
			conn->flags |= CO_FL_ERROR;
			return SF_ERR_SRVCL;
		}
	}
	else {
		/* connect() == 0, this is great! */
		conn->flags &= ~CO_FL_WAIT_L4_CONN;
	}

	conn_ctrl_init(conn);       /* registers the FD */
	HA_ATOMIC_OR(&fdtab[fd].state, FD_LINGER_RISK);  /* close hard if needed */

	if (conn->flags & CO_FL_WAIT_L4_CONN) {
		fd_want_send(fd);
		fd_cant_send(fd);
		fd_cant_recv(fd);
	}

	return SF_ERR_NONE;  /* connection is OK */
}

/* Add listener <listener> to protocol <proto>. Technically speaking we just
 * initialize a few entries which should be doable during quic_bind_listener().
 * The end of the initialization goes on with the default function.
 */
static void quic_add_listener(struct protocol *proto, struct listener *listener)
{
	listener->flags |= LI_F_QUIC_LISTENER;
	listener->rx.flags |= RX_F_LOCAL_ACCEPT;

	default_add_listener(proto, listener);
}

/* Allocate the RX buffers for <l> listener.
 * Return 1 if succeeded, 0 if not.
 */
static int quic_alloc_rxbufs_listener(struct listener *l)
{
	int i;
	struct quic_receiver_buf *tmp;

	MT_LIST_INIT(&l->rx.rxbuf_list);
	for (i = 0; i < global.nbthread; i++) {
		struct quic_receiver_buf *rxbuf;
		char *buf;

		rxbuf = calloc(1, sizeof(*rxbuf));
		if (!rxbuf)
			goto err;

		buf = pool_alloc(pool_head_quic_rxbuf);
		if (!buf) {
			free(rxbuf);
			goto err;
		}

		rxbuf->buf = b_make(buf, QUIC_RX_BUFSZ, 0, 0);
		LIST_INIT(&rxbuf->dgram_list);
		MT_LIST_APPEND(&l->rx.rxbuf_list, &rxbuf->rxbuf_el);
	}

	return 1;

 err:
	while ((tmp = MT_LIST_POP(&l->rx.rxbuf_list, typeof(tmp), rxbuf_el))) {
		pool_free(pool_head_quic_rxbuf, tmp->buf.area);
		free(tmp);
	}
	return 0;
}

/* This function tries to bind a QUIC4/6 listener. It may return a warning or
 * an error message in <errmsg> if the message is at most <errlen> bytes long
 * (including '\0'). Note that <errmsg> may be NULL if <errlen> is also zero.
 * The return value is composed from ERR_ABORT, ERR_WARN,
 * ERR_ALERT, ERR_RETRYABLE and ERR_FATAL. ERR_NONE indicates that everything
 * was alright and that no message was returned. ERR_RETRYABLE means that an
 * error occurred but that it may vanish after a retry (eg: port in use), and
 * ERR_FATAL indicates a non-fixable error. ERR_WARN and ERR_ALERT do not alter
 * the meaning of the error, but just indicate that a message is present which
 * should be displayed with the respective level. Last, ERR_ABORT indicates
 * that it's pointless to try to start other listeners. No error message is
 * returned if errlen is NULL.
 */
static int quic_bind_listener(struct listener *listener, char *errmsg, int errlen)
{
	const struct sockaddr_storage addr = listener->rx.addr;
	int fd, err = ERR_NONE;
	char *msg = NULL;

	/* ensure we never return garbage */
	if (errlen)
		*errmsg = 0;

	if (listener->state != LI_ASSIGNED)
		return ERR_NONE; /* already bound */

	if (!(listener->rx.flags & RX_F_BOUND)) {
		msg = "receiving socket not bound";
		goto udp_return;
	}

	/* Set IP_PKTINFO to retrieve destination address on recv. */
	fd = listener->rx.fd;
	switch (addr.ss_family) {
	case AF_INET:
#if defined(IP_PKTINFO)
		setsockopt(fd, IPPROTO_IP, IP_PKTINFO, &one, sizeof(one));
#elif defined(IP_RECVDSTADDR)
		setsockopt(fd, IPPROTO_IP, IP_RECVDSTADDR, &one, sizeof(one));
#endif /* IP_PKTINFO || IP_RECVDSTADDR */
		break;
	case AF_INET6:
#ifdef IPV6_RECVPKTINFO
		setsockopt(fd, IPPROTO_IPV6, IPV6_RECVPKTINFO, &one, sizeof(one));
#endif
		break;
	default:
		break;
	}

	if (!quic_alloc_rxbufs_listener(listener)) {
		msg = "could not initialize tx/rx rings";
		err |= ERR_WARN;
		goto udp_return;
	}

	listener_set_state(listener, LI_LISTEN);

 udp_return:
	if (msg && errlen) {
		char pn[INET6_ADDRSTRLEN];

		addr_to_str(&listener->rx.addr, pn, sizeof(pn));
		snprintf(errmsg, errlen, "%s for [%s:%d]", msg, pn, get_host_port(&listener->rx.addr));
	}
	return err;
}

/* Enable receipt of incoming connections for listener <l>. The receiver must
 * still be valid. Does nothing in early boot (needs fd_updt).
 */
static void quic_enable_listener(struct listener *l)
{
	/* FIXME: The following statements are incorrect. This
	 * is the responsibility of the QUIC xprt to stop accepting new
	 * connections.
	 */
	if (fd_updt)
		fd_want_recv(l->rx.fd);
}

/* Disable receipt of incoming connections for listener <l>. The receiver must
 * still be valid. Does nothing in early boot (needs fd_updt).
 */
static void quic_disable_listener(struct listener *l)
{
	/* FIXME: The following statements are incorrect. This
	 * is the responsibility of the QUIC xprt to start accepting new
	 * connections again.
	 */
	if (fd_updt)
		fd_stop_recv(l->rx.fd);
}

static int quic_alloc_dghdlrs(void)
{
	int i;

	quic_dghdlrs = calloc(global.nbthread, sizeof(*quic_dghdlrs));
	if (!quic_dghdlrs) {
		ha_alert("Failed to allocate the quic datagram handlers.\n");
		return 0;
	}

	for (i = 0; i < global.nbthread; i++) {
		struct quic_dghdlr *dghdlr = &quic_dghdlrs[i];

		dghdlr->task = tasklet_new();
		if (!dghdlr->task) {
			ha_alert("Failed to allocate the quic datagram handler on thread %d.\n", i);
			return 0;
		}

		tasklet_set_tid(dghdlr->task, i);
		dghdlr->task->context = dghdlr;
		dghdlr->task->process = quic_lstnr_dghdlr;

		dghdlr->odcids = EB_ROOT_UNIQUE;
		dghdlr->cids = EB_ROOT_UNIQUE;

		MT_LIST_INIT(&dghdlr->dgrams);
	}

	return 1;
}
REGISTER_POST_CHECK(quic_alloc_dghdlrs);

static int quic_deallocate_dghdlrs(void)
{
	int i;

	if (quic_dghdlrs) {
		for (i = 0; i < global.nbthread; ++i)
			tasklet_free(quic_dghdlrs[i].task);
		free(quic_dghdlrs);
	}

	return 1;
}
REGISTER_POST_DEINIT(quic_deallocate_dghdlrs);

/*
 * Local variables:
 *  c-indent-level: 8
 *  c-basic-offset: 8
 * End:
 */