l2tp_core.c 44.8 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0-only
T
Tom Parkin 已提交
2
/* L2TP core.
3 4 5 6 7 8 9 10 11 12 13 14 15 16
 *
 * Copyright (c) 2008,2009,2010 Katalix Systems Ltd
 *
 * This file contains some code of the original L2TPv2 pppol2tp
 * driver, which has the following copyright:
 *
 * Authors:	Martijn van Oosterhout <kleptog@svana.org>
 *		James Chapman (jchapman@katalix.com)
 * Contributors:
 *		Michal Ostrowski <mostrows@speakeasy.net>
 *		Arnaldo Carvalho de Melo <acme@xconectiva.com.br>
 *		David S. Miller (davem@redhat.com)
 */

17 18
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

19 20 21
#include <linux/module.h>
#include <linux/string.h>
#include <linux/list.h>
22
#include <linux/rculist.h>
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
#include <linux/uaccess.h>

#include <linux/kernel.h>
#include <linux/spinlock.h>
#include <linux/kthread.h>
#include <linux/sched.h>
#include <linux/slab.h>
#include <linux/errno.h>
#include <linux/jiffies.h>

#include <linux/netdevice.h>
#include <linux/net.h>
#include <linux/inetdevice.h>
#include <linux/skbuff.h>
#include <linux/init.h>
38
#include <linux/in.h>
39 40
#include <linux/ip.h>
#include <linux/udp.h>
41
#include <linux/l2tp.h>
42 43 44 45 46 47 48 49 50
#include <linux/hash.h>
#include <linux/sort.h>
#include <linux/file.h>
#include <linux/nsproxy.h>
#include <net/net_namespace.h>
#include <net/netns/generic.h>
#include <net/dst.h>
#include <net/ip.h>
#include <net/udp.h>
T
Tom Herbert 已提交
51
#include <net/udp_tunnel.h>
52
#include <net/inet_common.h>
53
#include <net/xfrm.h>
54
#include <net/protocol.h>
55 56 57
#include <net/inet6_connection_sock.h>
#include <net/inet_ecn.h>
#include <net/ip6_route.h>
58
#include <net/ip6_checksum.h>
59 60

#include <asm/byteorder.h>
61
#include <linux/atomic.h>
62 63

#include "l2tp_core.h"
64
#include "trace.h"
65

66 67 68
#define CREATE_TRACE_POINTS
#include "trace.h"

69 70 71 72 73 74 75 76 77 78 79
#define L2TP_DRV_VERSION	"V2.0"

/* L2TP header constants */
#define L2TP_HDRFLAG_T	   0x8000
#define L2TP_HDRFLAG_L	   0x4000
#define L2TP_HDRFLAG_S	   0x0800
#define L2TP_HDRFLAG_O	   0x0200
#define L2TP_HDRFLAG_P	   0x0100

#define L2TP_HDR_VER_MASK  0x000F
#define L2TP_HDR_VER_2	   0x0002
80
#define L2TP_HDR_VER_3	   0x0003
81 82 83 84 85

/* L2TPv3 default L2-specific sublayer */
#define L2TP_SLFLAG_S	   0x40000000
#define L2TP_SL_SEQ_MASK   0x00ffffff

86
#define L2TP_HDR_SIZE_MAX		14
87 88 89 90 91 92 93

/* Default trace flags */
#define L2TP_DEFAULT_DEBUG_FLAGS	0

/* Private data stored for received packets in the skb.
 */
struct l2tp_skb_cb {
94
	u32			ns;
95 96 97 98 99
	u16			has_seq;
	u16			length;
	unsigned long		expires;
};

100
#define L2TP_SKB_CB(skb)	((struct l2tp_skb_cb *)&(skb)->cb[sizeof(struct inet_skb_parm)])
101

102
static struct workqueue_struct *l2tp_wq;
103 104 105 106 107

/* per-net private data for this module */
static unsigned int l2tp_net_id;
struct l2tp_net {
	struct list_head l2tp_tunnel_list;
T
Tom Parkin 已提交
108
	/* Lock for write access to l2tp_tunnel_list */
109
	spinlock_t l2tp_tunnel_list_lock;
110
	struct hlist_head l2tp_session_hlist[L2TP_HASH_SIZE_2];
T
Tom Parkin 已提交
111
	/* Lock for write access to l2tp_session_hlist */
112
	spinlock_t l2tp_session_hlist_lock;
113 114
};

115 116 117 118 119 120 121
#if IS_ENABLED(CONFIG_IPV6)
static bool l2tp_sk_is_v6(struct sock *sk)
{
	return sk->sk_family == PF_INET6 &&
	       !ipv6_addr_v4mapped(&sk->sk_v6_daddr);
}
#endif
122

123
static inline struct l2tp_net *l2tp_pernet(const struct net *net)
124 125 126 127
{
	return net_generic(net, l2tp_net_id);
}

128 129 130 131 132 133 134 135 136 137 138
/* Session hash global list for L2TPv3.
 * The session_id SHOULD be random according to RFC3931, but several
 * L2TP implementations use incrementing session_ids.  So we do a real
 * hash on the session_id, rather than a simple bitmask.
 */
static inline struct hlist_head *
l2tp_session_id_hash_2(struct l2tp_net *pn, u32 session_id)
{
	return &pn->l2tp_session_hlist[hash_32(session_id, L2TP_HASH_BITS_2)];
}

139 140 141 142 143 144 145 146 147 148 149 150
/* Session hash list.
 * The session_id SHOULD be random according to RFC2661, but several
 * L2TP implementations (Cisco and Microsoft) use incrementing
 * session_ids.  So we do a real hash on the session_id, rather than a
 * simple bitmask.
 */
static inline struct hlist_head *
l2tp_session_id_hash(struct l2tp_tunnel *tunnel, u32 session_id)
{
	return &tunnel->session_hlist[hash_32(session_id, L2TP_HASH_BITS)];
}

151
static void l2tp_tunnel_free(struct l2tp_tunnel *tunnel)
152
{
153
	trace_free_tunnel(tunnel);
154 155 156
	sock_put(tunnel->sock);
	/* the tunnel is freed in the socket destructor */
}
157 158 159

static void l2tp_session_free(struct l2tp_session *session)
{
160
	trace_free_session(session);
161 162 163 164
	if (session->tunnel)
		l2tp_tunnel_dec_refcount(session->tunnel);
	kfree(session);
}
165

166 167 168 169 170
struct l2tp_tunnel *l2tp_sk_to_tunnel(struct sock *sk)
{
	struct l2tp_tunnel *tunnel = sk->sk_user_data;

	if (tunnel)
171
		if (WARN_ON(tunnel->magic != L2TP_TUNNEL_MAGIC))
172
			return NULL;
173

174
	return tunnel;
175
}
176
EXPORT_SYMBOL_GPL(l2tp_sk_to_tunnel);
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

void l2tp_tunnel_inc_refcount(struct l2tp_tunnel *tunnel)
{
	refcount_inc(&tunnel->ref_count);
}
EXPORT_SYMBOL_GPL(l2tp_tunnel_inc_refcount);

void l2tp_tunnel_dec_refcount(struct l2tp_tunnel *tunnel)
{
	if (refcount_dec_and_test(&tunnel->ref_count))
		l2tp_tunnel_free(tunnel);
}
EXPORT_SYMBOL_GPL(l2tp_tunnel_dec_refcount);

void l2tp_session_inc_refcount(struct l2tp_session *session)
{
	refcount_inc(&session->ref_count);
}
EXPORT_SYMBOL_GPL(l2tp_session_inc_refcount);

void l2tp_session_dec_refcount(struct l2tp_session *session)
{
	if (refcount_dec_and_test(&session->ref_count))
		l2tp_session_free(session);
}
EXPORT_SYMBOL_GPL(l2tp_session_dec_refcount);
203

204 205 206 207 208 209 210 211
/* Lookup a tunnel. A new reference is held on the returned tunnel. */
struct l2tp_tunnel *l2tp_tunnel_get(const struct net *net, u32 tunnel_id)
{
	const struct l2tp_net *pn = l2tp_pernet(net);
	struct l2tp_tunnel *tunnel;

	rcu_read_lock_bh();
	list_for_each_entry_rcu(tunnel, &pn->l2tp_tunnel_list, list) {
212 213
		if (tunnel->tunnel_id == tunnel_id &&
		    refcount_inc_not_zero(&tunnel->ref_count)) {
214 215 216 217 218 219 220 221 222 223 224
			rcu_read_unlock_bh();

			return tunnel;
		}
	}
	rcu_read_unlock_bh();

	return NULL;
}
EXPORT_SYMBOL_GPL(l2tp_tunnel_get);

225 226 227 228 229 230 231 232
struct l2tp_tunnel *l2tp_tunnel_get_nth(const struct net *net, int nth)
{
	const struct l2tp_net *pn = l2tp_pernet(net);
	struct l2tp_tunnel *tunnel;
	int count = 0;

	rcu_read_lock_bh();
	list_for_each_entry_rcu(tunnel, &pn->l2tp_tunnel_list, list) {
233 234
		if (++count > nth &&
		    refcount_inc_not_zero(&tunnel->ref_count)) {
235 236 237 238 239 240 241 242 243 244
			rcu_read_unlock_bh();
			return tunnel;
		}
	}
	rcu_read_unlock_bh();

	return NULL;
}
EXPORT_SYMBOL_GPL(l2tp_tunnel_get_nth);

245 246
struct l2tp_session *l2tp_tunnel_get_session(struct l2tp_tunnel *tunnel,
					     u32 session_id)
247 248 249 250
{
	struct hlist_head *session_list;
	struct l2tp_session *session;

251
	session_list = l2tp_session_id_hash(tunnel, session_id);
252

253 254 255 256 257
	read_lock_bh(&tunnel->hlist_lock);
	hlist_for_each_entry(session, session_list, hlist)
		if (session->session_id == session_id) {
			l2tp_session_inc_refcount(session);
			read_unlock_bh(&tunnel->hlist_lock);
258

259
			return session;
260
		}
261
	read_unlock_bh(&tunnel->hlist_lock);
262

263 264 265
	return NULL;
}
EXPORT_SYMBOL_GPL(l2tp_tunnel_get_session);
266

267 268 269 270 271 272 273 274 275
struct l2tp_session *l2tp_session_get(const struct net *net, u32 session_id)
{
	struct hlist_head *session_list;
	struct l2tp_session *session;

	session_list = l2tp_session_id_hash_2(l2tp_pernet(net), session_id);

	rcu_read_lock_bh();
	hlist_for_each_entry_rcu(session, session_list, global_hlist)
276 277
		if (session->session_id == session_id) {
			l2tp_session_inc_refcount(session);
278
			rcu_read_unlock_bh();
279 280 281

			return session;
		}
282
	rcu_read_unlock_bh();
283 284 285 286 287

	return NULL;
}
EXPORT_SYMBOL_GPL(l2tp_session_get);

288
struct l2tp_session *l2tp_session_get_nth(struct l2tp_tunnel *tunnel, int nth)
289 290 291 292 293 294 295
{
	int hash;
	struct l2tp_session *session;
	int count = 0;

	read_lock_bh(&tunnel->hlist_lock);
	for (hash = 0; hash < L2TP_HASH_SIZE; hash++) {
296
		hlist_for_each_entry(session, &tunnel->session_hlist[hash], hlist) {
297
			if (++count > nth) {
298
				l2tp_session_inc_refcount(session);
299 300 301 302 303 304 305 306 307 308
				read_unlock_bh(&tunnel->hlist_lock);
				return session;
			}
		}
	}

	read_unlock_bh(&tunnel->hlist_lock);

	return NULL;
}
309
EXPORT_SYMBOL_GPL(l2tp_session_get_nth);
310

311 312 313
/* Lookup a session by interface name.
 * This is very inefficient but is only used by management interfaces.
 */
314
struct l2tp_session *l2tp_session_get_by_ifname(const struct net *net,
315
						const char *ifname)
316 317 318 319 320
{
	struct l2tp_net *pn = l2tp_pernet(net);
	int hash;
	struct l2tp_session *session;

321
	rcu_read_lock_bh();
322
	for (hash = 0; hash < L2TP_HASH_SIZE_2; hash++) {
323
		hlist_for_each_entry_rcu(session, &pn->l2tp_session_hlist[hash], global_hlist) {
324
			if (!strcmp(session->ifname, ifname)) {
325
				l2tp_session_inc_refcount(session);
326
				rcu_read_unlock_bh();
327

328 329 330 331 332
				return session;
			}
		}
	}

333
	rcu_read_unlock_bh();
334 335 336

	return NULL;
}
337
EXPORT_SYMBOL_GPL(l2tp_session_get_by_ifname);
338

339 340
int l2tp_session_register(struct l2tp_session *session,
			  struct l2tp_tunnel *tunnel)
341 342 343 344 345
{
	struct l2tp_session *session_walk;
	struct hlist_head *g_head;
	struct hlist_head *head;
	struct l2tp_net *pn;
346
	int err;
347 348 349 350

	head = l2tp_session_id_hash(tunnel, session->session_id);

	write_lock_bh(&tunnel->hlist_lock);
351 352 353 354 355
	if (!tunnel->acpt_newsess) {
		err = -ENODEV;
		goto err_tlock;
	}

356
	hlist_for_each_entry(session_walk, head, hlist)
357 358 359 360
		if (session_walk->session_id == session->session_id) {
			err = -EEXIST;
			goto err_tlock;
		}
361 362 363

	if (tunnel->version == L2TP_HDR_VER_3) {
		pn = l2tp_pernet(tunnel->l2tp_net);
364
		g_head = l2tp_session_id_hash_2(pn, session->session_id);
365 366

		spin_lock_bh(&pn->l2tp_session_hlist_lock);
367

368 369 370
		/* IP encap expects session IDs to be globally unique, while
		 * UDP encap doesn't.
		 */
371
		hlist_for_each_entry(session_walk, g_head, global_hlist)
372 373 374
			if (session_walk->session_id == session->session_id &&
			    (session_walk->tunnel->encap == L2TP_ENCAPTYPE_IP ||
			     tunnel->encap == L2TP_ENCAPTYPE_IP)) {
375 376 377
				err = -EEXIST;
				goto err_tlock_pnlock;
			}
378

379
		l2tp_tunnel_inc_refcount(tunnel);
380
		hlist_add_head_rcu(&session->global_hlist, g_head);
381

382
		spin_unlock_bh(&pn->l2tp_session_hlist_lock);
383 384
	} else {
		l2tp_tunnel_inc_refcount(tunnel);
385 386 387 388 389
	}

	hlist_add_head(&session->hlist, head);
	write_unlock_bh(&tunnel->hlist_lock);

390 391
	trace_register_session(session);

392 393
	return 0;

394
err_tlock_pnlock:
395
	spin_unlock_bh(&pn->l2tp_session_hlist_lock);
396
err_tlock:
397 398
	write_unlock_bh(&tunnel->hlist_lock);

399
	return err;
400
}
401
EXPORT_SYMBOL_GPL(l2tp_session_register);
402

403 404 405 406 407 408 409 410 411 412 413
/*****************************************************************************
 * Receive data handling
 *****************************************************************************/

/* Queue a skb in order. We come here only if the skb has an L2TP sequence
 * number.
 */
static void l2tp_recv_queue_skb(struct l2tp_session *session, struct sk_buff *skb)
{
	struct sk_buff *skbp;
	struct sk_buff *tmp;
414
	u32 ns = L2TP_SKB_CB(skb)->ns;
415 416 417 418 419

	spin_lock_bh(&session->reorder_q.lock);
	skb_queue_walk_safe(&session->reorder_q, skbp, tmp) {
		if (L2TP_SKB_CB(skbp)->ns > ns) {
			__skb_queue_before(&session->reorder_q, skbp, skb);
420
			atomic_long_inc(&session->stats.rx_oos_packets);
421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442
			goto out;
		}
	}

	__skb_queue_tail(&session->reorder_q, skb);

out:
	spin_unlock_bh(&session->reorder_q.lock);
}

/* Dequeue a single skb.
 */
static void l2tp_recv_dequeue_skb(struct l2tp_session *session, struct sk_buff *skb)
{
	struct l2tp_tunnel *tunnel = session->tunnel;
	int length = L2TP_SKB_CB(skb)->length;

	/* We're about to requeue the skb, so return resources
	 * to its current owner (a socket receive buffer).
	 */
	skb_orphan(skb);

443 444 445 446
	atomic_long_inc(&tunnel->stats.rx_packets);
	atomic_long_add(length, &tunnel->stats.rx_bytes);
	atomic_long_inc(&session->stats.rx_packets);
	atomic_long_add(length, &session->stats.rx_bytes);
447 448 449 450

	if (L2TP_SKB_CB(skb)->has_seq) {
		/* Bump our Nr */
		session->nr++;
451
		session->nr &= session->nr_max;
452
		trace_session_seqnum_update(session);
453 454 455
	}

	/* call private receive handler */
456
	if (session->recv_skb)
457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473
		(*session->recv_skb)(session, skb, L2TP_SKB_CB(skb)->length);
	else
		kfree_skb(skb);
}

/* Dequeue skbs from the session's reorder_q, subject to packet order.
 * Skbs that have been in the queue for too long are simply discarded.
 */
static void l2tp_recv_dequeue(struct l2tp_session *session)
{
	struct sk_buff *skb;
	struct sk_buff *tmp;

	/* If the pkt at the head of the queue has the nr that we
	 * expect to send up next, dequeue it and any other
	 * in-sequence packets behind it.
	 */
474
start:
475 476
	spin_lock_bh(&session->reorder_q.lock);
	skb_queue_walk_safe(&session->reorder_q, skb, tmp) {
477 478 479 480
		struct l2tp_skb_cb *cb = L2TP_SKB_CB(skb);

		/* If the packet has been pending on the queue for too long, discard it */
		if (time_after(jiffies, cb->expires)) {
481 482
			atomic_long_inc(&session->stats.rx_seq_discards);
			atomic_long_inc(&session->stats.rx_errors);
483
			trace_session_pkt_expired(session, cb->ns);
484
			session->reorder_skip = 1;
485 486 487 488 489
			__skb_unlink(skb, &session->reorder_q);
			kfree_skb(skb);
			continue;
		}

490
		if (cb->has_seq) {
491 492
			if (session->reorder_skip) {
				session->reorder_skip = 0;
493 494
				session->nr = cb->ns;
				trace_session_seqnum_reset(session);
495
			}
496
			if (cb->ns != session->nr)
497 498 499 500 501 502 503 504 505
				goto out;
		}
		__skb_unlink(skb, &session->reorder_q);

		/* Process the skb. We release the queue lock while we
		 * do so to let other contexts process the queue.
		 */
		spin_unlock_bh(&session->reorder_q.lock);
		l2tp_recv_dequeue_skb(session, skb);
506
		goto start;
507 508 509 510 511 512
	}

out:
	spin_unlock_bh(&session->reorder_q.lock);
}

513 514 515 516 517 518 519 520 521 522 523 524
static int l2tp_seq_check_rx_window(struct l2tp_session *session, u32 nr)
{
	u32 nws;

	if (nr >= session->nr)
		nws = nr - session->nr;
	else
		nws = (session->nr_max + 1) - (session->nr - nr);

	return nws < session->nr_window_size;
}

525 526 527 528 529
/* If packet has sequence numbers, queue it if acceptable. Returns 0 if
 * acceptable, else non-zero.
 */
static int l2tp_recv_data_seq(struct l2tp_session *session, struct sk_buff *skb)
{
530 531 532
	struct l2tp_skb_cb *cb = L2TP_SKB_CB(skb);

	if (!l2tp_seq_check_rx_window(session, cb->ns)) {
533 534 535
		/* Packet sequence number is outside allowed window.
		 * Discard it.
		 */
536
		trace_session_pkt_outside_rx_window(session, cb->ns);
537 538 539
		goto discard;
	}

540 541 542 543 544
	if (session->reorder_timeout != 0) {
		/* Packet reordering enabled. Add skb to session's
		 * reorder queue, in order of ns.
		 */
		l2tp_recv_queue_skb(session, skb);
545 546 547 548 549 550 551 552
		goto out;
	}

	/* Packet reordering disabled. Discard out-of-sequence packets, while
	 * tracking the number if in-sequence packets after the first OOS packet
	 * is seen. After nr_oos_count_max in-sequence packets, reset the
	 * sequence number to re-enable packet reception.
	 */
553
	if (cb->ns == session->nr) {
554
		skb_queue_tail(&session->reorder_q, skb);
555
	} else {
556
		u32 nr_oos = cb->ns;
557 558 559 560 561 562 563 564 565 566 567 568
		u32 nr_next = (session->nr_oos + 1) & session->nr_max;

		if (nr_oos == nr_next)
			session->nr_oos_count++;
		else
			session->nr_oos_count = 0;

		session->nr_oos = nr_oos;
		if (session->nr_oos_count > session->nr_oos_count_max) {
			session->reorder_skip = 1;
		}
		if (!session->reorder_skip) {
569
			atomic_long_inc(&session->stats.rx_seq_discards);
570
			trace_session_pkt_oos(session, cb->ns);
571 572 573 574 575
			goto discard;
		}
		skb_queue_tail(&session->reorder_q, skb);
	}

576
out:
577 578 579 580 581 582
	return 0;

discard:
	return 1;
}

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
/* Do receive processing of L2TP data frames. We handle both L2TPv2
 * and L2TPv3 data frames here.
 *
 * L2TPv2 Data Message Header
 *
 *  0                   1                   2                   3
 *  0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 * |T|L|x|x|S|x|O|P|x|x|x|x|  Ver  |          Length (opt)         |
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 * |           Tunnel ID           |           Session ID          |
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 * |             Ns (opt)          |             Nr (opt)          |
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 * |      Offset Size (opt)        |    Offset pad... (opt)
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 *
 * Data frames are marked by T=0. All other fields are the same as
 * those in L2TP control frames.
 *
 * L2TPv3 Data Message Header
 *
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 * |                      L2TP Session Header                      |
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 * |                      L2-Specific Sublayer                     |
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 * |                        Tunnel Payload                      ...
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 *
 * L2TPv3 Session Header Over IP
 *
 *  0                   1                   2                   3
 *  0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 * |                           Session ID                          |
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 * |               Cookie (optional, maximum 64 bits)...
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 *                                                                 |
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 *
 * L2TPv3 L2-Specific Sublayer Format
 *
 *  0                   1                   2                   3
 *  0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 * |x|S|x|x|x|x|x|x|              Sequence Number                  |
 * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 *
633 634 635
 * Cookie value and sublayer format are negotiated with the peer when
 * the session is set up. Unlike L2TPv2, we do not need to parse the
 * packet header to determine if optional fields are present.
636 637 638 639 640
 *
 * Caller must already have parsed the frame and determined that it is
 * a data (not control) frame before coming here. Fields up to the
 * session-id have already been parsed and ptr points to the data
 * after the session-id.
641
 */
642 643
void l2tp_recv_common(struct l2tp_session *session, struct sk_buff *skb,
		      unsigned char *ptr, unsigned char *optr, u16 hdrflags,
644
		      int length)
645
{
646
	struct l2tp_tunnel *tunnel = session->tunnel;
647 648
	int offset;

649 650 651
	/* Parse and check optional cookie */
	if (session->peer_cookie_len > 0) {
		if (memcmp(ptr, &session->peer_cookie[0], session->peer_cookie_len)) {
652 653 654
			pr_debug_ratelimited("%s: cookie mismatch (%u/%u). Discarding.\n",
					     tunnel->name, tunnel->tunnel_id,
					     session->session_id);
655
			atomic_long_inc(&session->stats.rx_cookie_discards);
656 657 658 659 660
			goto discard;
		}
		ptr += session->peer_cookie_len;
	}

661 662 663 664 665 666 667 668
	/* Handle the optional sequence numbers. Sequence numbers are
	 * in different places for L2TPv2 and L2TPv3.
	 *
	 * If we are the LAC, enable/disable sequence numbers under
	 * the control of the LNS.  If no sequence numbers present but
	 * we were expecting them, discard frame.
	 */
	L2TP_SKB_CB(skb)->has_seq = 0;
669 670 671
	if (tunnel->version == L2TP_HDR_VER_2) {
		if (hdrflags & L2TP_HDRFLAG_S) {
			/* Store L2TP info in the skb */
672
			L2TP_SKB_CB(skb)->ns = ntohs(*(__be16 *)ptr);
673
			L2TP_SKB_CB(skb)->has_seq = 1;
674 675 676
			ptr += 2;
			/* Skip past nr in the header */
			ptr += 2;
677

678 679
		}
	} else if (session->l2specific_type == L2TP_L2SPECTYPE_DEFAULT) {
680
		u32 l2h = ntohl(*(__be32 *)ptr);
681 682 683

		if (l2h & 0x40000000) {
			/* Store L2TP info in the skb */
684
			L2TP_SKB_CB(skb)->ns = l2h & 0x00ffffff;
685 686
			L2TP_SKB_CB(skb)->has_seq = 1;
		}
687
		ptr += 4;
688 689 690
	}

	if (L2TP_SKB_CB(skb)->has_seq) {
T
Tom Parkin 已提交
691
		/* Received a packet with sequence numbers. If we're the LAC,
692 693 694
		 * check if we sre sending sequence numbers and if not,
		 * configure it so.
		 */
695
		if (!session->lns_mode && !session->send_seq) {
696
			trace_session_seqnum_lns_enable(session);
697
			session->send_seq = 1;
698
			l2tp_session_set_header_len(session, tunnel->version);
699 700 701 702 703 704
		}
	} else {
		/* No sequence numbers.
		 * If user has configured mandatory sequence numbers, discard.
		 */
		if (session->recv_seq) {
705 706
			pr_debug_ratelimited("%s: recv data has no seq numbers when required. Discarding.\n",
					     session->name);
707
			atomic_long_inc(&session->stats.rx_seq_discards);
708 709 710 711 712 713 714 715
			goto discard;
		}

		/* If we're the LAC and we're sending sequence numbers, the
		 * LNS has requested that we no longer send sequence numbers.
		 * If we're the LNS and we're sending sequence numbers, the
		 * LAC is broken. Discard the frame.
		 */
716
		if (!session->lns_mode && session->send_seq) {
717
			trace_session_seqnum_lns_disable(session);
718
			session->send_seq = 0;
719
			l2tp_session_set_header_len(session, tunnel->version);
720
		} else if (session->send_seq) {
721 722
			pr_debug_ratelimited("%s: recv data has no seq numbers when required. Discarding.\n",
					     session->name);
723
			atomic_long_inc(&session->stats.rx_seq_discards);
724 725 726 727
			goto discard;
		}
	}

728 729
	/* Session data offset is defined only for L2TPv2 and is
	 * indicated by an optional 16-bit value in the header.
730 731 732 733 734 735 736
	 */
	if (tunnel->version == L2TP_HDR_VER_2) {
		/* If offset bit set, skip it. */
		if (hdrflags & L2TP_HDRFLAG_O) {
			offset = ntohs(*(__be16 *)ptr);
			ptr += 2 + offset;
		}
737
	}
738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756

	offset = ptr - optr;
	if (!pskb_may_pull(skb, offset))
		goto discard;

	__skb_pull(skb, offset);

	/* Prepare skb for adding to the session's reorder_q.  Hold
	 * packets for max reorder_timeout or 1 second if not
	 * reordering.
	 */
	L2TP_SKB_CB(skb)->length = length;
	L2TP_SKB_CB(skb)->expires = jiffies +
		(session->reorder_timeout ? session->reorder_timeout : HZ);

	/* Add packet to the session's receive queue. Reordering is done here, if
	 * enabled. Saved L2TP protocol info is stored in skb->sb[].
	 */
	if (L2TP_SKB_CB(skb)->has_seq) {
757 758
		if (l2tp_recv_data_seq(session, skb))
			goto discard;
759 760 761 762 763 764 765 766 767 768 769
	} else {
		/* No sequence numbers. Add the skb to the tail of the
		 * reorder queue. This ensures that it will be
		 * delivered after all previous sequenced skbs.
		 */
		skb_queue_tail(&session->reorder_q, skb);
	}

	/* Try to dequeue as many skbs from reorder_q as we can. */
	l2tp_recv_dequeue(session);

770
	return;
771 772

discard:
773
	atomic_long_inc(&session->stats.rx_errors);
774
	kfree_skb(skb);
775
}
776
EXPORT_SYMBOL_GPL(l2tp_recv_common);
777

778 779
/* Drop skbs from the session's reorder_q
 */
780
static void l2tp_session_queue_purge(struct l2tp_session *session)
781 782
{
	struct sk_buff *skb = NULL;
783

784 785 786 787 788 789
	while ((skb = skb_dequeue(&session->reorder_q))) {
		atomic_long_inc(&session->stats.rx_errors);
		kfree_skb(skb);
	}
}

790 791 792 793 794 795
/* Internal UDP receive frame. Do the real work of receiving an L2TP data frame
 * here. The skb is not on a list when we get here.
 * Returns 0 if the packet was a data packet and was successfully passed on.
 * Returns 1 if the packet was not a good data packet and could not be
 * forwarded.  All such packets are passed up to userspace to deal with.
 */
796
static int l2tp_udp_recv_core(struct l2tp_tunnel *tunnel, struct sk_buff *skb)
797 798 799 800 801 802 803 804
{
	struct l2tp_session *session = NULL;
	unsigned char *ptr, *optr;
	u16 hdrflags;
	u32 tunnel_id, session_id;
	u16 version;
	int length;

805
	/* UDP has verifed checksum */
806 807 808 809 810

	/* UDP always verifies the packet length. */
	__skb_pull(skb, sizeof(struct udphdr));

	/* Short packet? */
811
	if (!pskb_may_pull(skb, L2TP_HDR_SIZE_MAX)) {
812 813 814
		pr_debug_ratelimited("%s: recv short packet (len=%d)\n",
				     tunnel->name, skb->len);
		goto invalid;
815 816
	}

817
	/* Point to L2TP header */
818 819
	optr = skb->data;
	ptr = skb->data;
820

821
	/* Get L2TP header flags */
822
	hdrflags = ntohs(*(__be16 *)ptr);
823 824 825 826

	/* Check protocol version */
	version = hdrflags & L2TP_HDR_VER_MASK;
	if (version != tunnel->version) {
827 828 829
		pr_debug_ratelimited("%s: recv protocol version mismatch: got %d expected %d\n",
				     tunnel->name, version, tunnel->version);
		goto invalid;
830 831 832 833 834 835
	}

	/* Get length of L2TP packet */
	length = skb->len;

	/* If type is control packet, it is handled by userspace. */
836
	if (hdrflags & L2TP_HDRFLAG_T)
837
		goto pass;
838 839 840 841 842 843 844 845 846 847

	/* Skip flags */
	ptr += 2;

	if (tunnel->version == L2TP_HDR_VER_2) {
		/* If length is present, skip it */
		if (hdrflags & L2TP_HDRFLAG_L)
			ptr += 2;

		/* Extract tunnel and session ID */
848
		tunnel_id = ntohs(*(__be16 *)ptr);
849
		ptr += 2;
850
		session_id = ntohs(*(__be16 *)ptr);
851 852 853 854
		ptr += 2;
	} else {
		ptr += 2;	/* skip reserved bits */
		tunnel_id = tunnel->tunnel_id;
855
		session_id = ntohl(*(__be32 *)ptr);
856 857 858 859
		ptr += 4;
	}

	/* Find the session context */
860
	session = l2tp_tunnel_get_session(tunnel, session_id);
861
	if (!session || !session->recv_skb) {
862
		if (session)
863 864
			l2tp_session_dec_refcount(session);

865
		/* Not found? Pass to userspace to deal with */
866 867 868
		pr_debug_ratelimited("%s: no session found (%u/%u). Passing up.\n",
				     tunnel->name, tunnel_id, session_id);
		goto pass;
869 870
	}

871 872
	if (tunnel->version == L2TP_HDR_VER_3 &&
	    l2tp_v3_ensure_opt_in_linear(session, skb, &ptr, &optr))
873
		goto invalid;
874

875
	l2tp_recv_common(session, skb, ptr, optr, hdrflags, length);
876
	l2tp_session_dec_refcount(session);
877 878 879

	return 0;

880 881 882 883
invalid:
	atomic_long_inc(&tunnel->stats.rx_invalid);

pass:
884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899
	/* Put UDP header back */
	__skb_push(skb, sizeof(struct udphdr));

	return 1;
}

/* UDP encapsulation receive handler. See net/ipv4/udp.c.
 * Return codes:
 * 0 : success.
 * <0: error
 * >0: skb should be passed up to userspace as UDP.
 */
int l2tp_udp_encap_recv(struct sock *sk, struct sk_buff *skb)
{
	struct l2tp_tunnel *tunnel;

900 901 902 903 904 905
	/* Note that this is called from the encap_rcv hook inside an
	 * RCU-protected region, but without the socket being locked.
	 * Hence we use rcu_dereference_sk_user_data to access the
	 * tunnel data structure rather the usual l2tp_sk_to_tunnel
	 * accessor function.
	 */
906
	tunnel = rcu_dereference_sk_user_data(sk);
907
	if (!tunnel)
908
		goto pass_up;
909 910
	if (WARN_ON(tunnel->magic != L2TP_TUNNEL_MAGIC))
		goto pass_up;
911

912
	if (l2tp_udp_recv_core(tunnel, skb))
913
		goto pass_up;
914 915 916 917 918 919 920 921 922 923 924 925 926 927

	return 0;

pass_up:
	return 1;
}
EXPORT_SYMBOL_GPL(l2tp_udp_encap_recv);

/************************************************************************
 * Transmit handling
 ***********************************************************************/

/* Build an L2TP header for the session into the buffer provided.
 */
928
static int l2tp_build_l2tpv2_header(struct l2tp_session *session, void *buf)
929
{
930
	struct l2tp_tunnel *tunnel = session->tunnel;
931
	__be16 *bufp = buf;
932
	__be16 *optr = buf;
933 934 935 936 937 938 939 940 941 942 943 944 945 946 947
	u16 flags = L2TP_HDR_VER_2;
	u32 tunnel_id = tunnel->peer_tunnel_id;
	u32 session_id = session->peer_session_id;

	if (session->send_seq)
		flags |= L2TP_HDRFLAG_S;

	/* Setup L2TP header. */
	*bufp++ = htons(flags);
	*bufp++ = htons(tunnel_id);
	*bufp++ = htons(session_id);
	if (session->send_seq) {
		*bufp++ = htons(session->ns);
		*bufp++ = 0;
		session->ns++;
948
		session->ns &= 0xffff;
949
		trace_session_seqnum_update(session);
950
	}
951 952

	return bufp - optr;
953 954
}

955
static int l2tp_build_l2tpv3_header(struct l2tp_session *session, void *buf)
956
{
957
	struct l2tp_tunnel *tunnel = session->tunnel;
958 959 960
	char *bufp = buf;
	char *optr = bufp;

961 962 963 964 965
	/* Setup L2TP header. The header differs slightly for UDP and
	 * IP encapsulations. For UDP, there is 4 bytes of flags.
	 */
	if (tunnel->encap == L2TP_ENCAPTYPE_UDP) {
		u16 flags = L2TP_HDR_VER_3;
966
		*((__be16 *)bufp) = htons(flags);
967
		bufp += 2;
968
		*((__be16 *)bufp) = 0;
969 970 971
		bufp += 2;
	}

972
	*((__be32 *)bufp) = htonl(session->peer_session_id);
973 974 975 976 977
	bufp += 4;
	if (session->cookie_len) {
		memcpy(bufp, &session->cookie[0], session->cookie_len);
		bufp += session->cookie_len;
	}
978 979
	if (session->l2specific_type == L2TP_L2SPECTYPE_DEFAULT) {
		u32 l2h = 0;
980

981 982 983 984
		if (session->send_seq) {
			l2h = 0x40000000 | session->ns;
			session->ns++;
			session->ns &= 0xffffff;
985
			trace_session_seqnum_update(session);
986
		}
987 988 989

		*((__be32 *)bufp) = htonl(l2h);
		bufp += 4;
990
	}
991

992
	return bufp - optr;
993 994
}

995 996
/* Queue the packet to IP for output: tunnel socket lock must be held */
static int l2tp_xmit_queue(struct l2tp_tunnel *tunnel, struct sk_buff *skb, struct flowi *fl)
997
{
998
	int err;
999

1000
	skb->ignore_df = 1;
1001
	skb_dst_drop(skb);
1002
#if IS_ENABLED(CONFIG_IPV6)
1003
	if (l2tp_sk_is_v6(tunnel->sock))
1004
		err = inet6_csk_xmit(tunnel->sock, skb, NULL);
1005 1006
	else
#endif
1007
		err = ip_queue_xmit(tunnel->sock, skb, fl);
1008

1009
	return err >= 0 ? NET_XMIT_SUCCESS : NET_XMIT_DROP;
1010 1011
}

1012
static int l2tp_xmit_core(struct l2tp_session *session, struct sk_buff *skb, unsigned int *len)
1013
{
1014
	struct l2tp_tunnel *tunnel = session->tunnel;
1015
	unsigned int data_len = skb->len;
1016
	struct sock *sk = tunnel->sock;
1017
	int headroom, uhlen, udp_len;
1018
	int ret = NET_XMIT_SUCCESS;
1019 1020
	struct inet_sock *inet;
	struct udphdr *uh;
1021 1022 1023 1024 1025

	/* Check that there's enough headroom in the skb to insert IP,
	 * UDP and L2TP headers. If not enough, expand it to
	 * make room. Adjust truesize.
	 */
1026 1027
	uhlen = (tunnel->encap == L2TP_ENCAPTYPE_UDP) ? sizeof(*uh) : 0;
	headroom = NET_SKB_PAD + sizeof(struct iphdr) + uhlen + session->hdr_len;
1028
	if (skb_cow_head(skb, headroom)) {
1029 1030
		kfree_skb(skb);
		return NET_XMIT_DROP;
1031
	}
1032 1033

	/* Setup L2TP header */
1034
	if (tunnel->version == L2TP_HDR_VER_2)
1035
		l2tp_build_l2tpv2_header(session, __skb_push(skb, session->hdr_len));
1036
	else
1037
		l2tp_build_l2tpv3_header(session, __skb_push(skb, session->hdr_len));
1038

1039
	/* Reset skb netfilter state */
1040
	memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
1041
	IPCB(skb)->flags &= ~(IPSKB_XFRM_TUNNEL_SIZE | IPSKB_XFRM_TRANSFORMED | IPSKB_REROUTED);
1042
	nf_reset_ct(skb);
1043

1044 1045
	bh_lock_sock(sk);
	if (sock_owned_by_user(sk)) {
1046 1047
		kfree_skb(skb);
		ret = NET_XMIT_DROP;
1048 1049 1050
		goto out_unlock;
	}

1051 1052 1053 1054 1055 1056 1057 1058 1059
	/* The user-space may change the connection status for the user-space
	 * provided socket at run time: we must check it under the socket lock
	 */
	if (tunnel->fd >= 0 && sk->sk_state != TCP_ESTABLISHED) {
		kfree_skb(skb);
		ret = NET_XMIT_DROP;
		goto out_unlock;
	}

1060 1061 1062 1063 1064
	/* Report transmitted length before we add encap header, which keeps
	 * statistics consistent for both UDP and IP encap tx/rx paths.
	 */
	*len = skb->len;

1065
	inet = inet_sk(sk);
1066 1067 1068 1069 1070 1071 1072 1073
	switch (tunnel->encap) {
	case L2TP_ENCAPTYPE_UDP:
		/* Setup UDP header */
		__skb_push(skb, sizeof(*uh));
		skb_reset_transport_header(skb);
		uh = udp_hdr(skb);
		uh->source = inet->inet_sport;
		uh->dest = inet->inet_dport;
1074
		udp_len = uhlen + session->hdr_len + data_len;
1075 1076 1077
		uh->len = htons(udp_len);

		/* Calculate UDP checksum if configured to do so */
1078
#if IS_ENABLED(CONFIG_IPV6)
1079
		if (l2tp_sk_is_v6(sk))
T
Tom Herbert 已提交
1080 1081 1082
			udp6_set_csum(udp_get_no_check6_tx(sk),
				      skb, &inet6_sk(sk)->saddr,
				      &sk->sk_v6_daddr, udp_len);
1083 1084
		else
#endif
1085 1086
			udp_set_csum(sk->sk_no_check_tx, skb, inet->inet_saddr,
				     inet->inet_daddr, udp_len);
1087 1088 1089 1090
		break;

	case L2TP_ENCAPTYPE_IP:
		break;
1091 1092
	}

1093 1094
	ret = l2tp_xmit_queue(tunnel, skb, &inet->cork.fl);

1095 1096
out_unlock:
	bh_unlock_sock(sk);
1097

1098
	return ret;
1099
}
1100 1101 1102 1103 1104 1105

/* If caller requires the skb to have a ppp header, the header must be
 * inserted in the skb data before calling this function.
 */
int l2tp_xmit_skb(struct l2tp_session *session, struct sk_buff *skb)
{
1106
	unsigned int len = 0;
1107 1108
	int ret;

1109
	ret = l2tp_xmit_core(session, skb, &len);
1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120
	if (ret == NET_XMIT_SUCCESS) {
		atomic_long_inc(&session->tunnel->stats.tx_packets);
		atomic_long_add(len, &session->tunnel->stats.tx_bytes);
		atomic_long_inc(&session->stats.tx_packets);
		atomic_long_add(len, &session->stats.tx_bytes);
	} else {
		atomic_long_inc(&session->tunnel->stats.tx_errors);
		atomic_long_inc(&session->stats.tx_errors);
	}
	return ret;
}
1121 1122 1123 1124 1125 1126 1127 1128 1129 1130
EXPORT_SYMBOL_GPL(l2tp_xmit_skb);

/*****************************************************************************
 * Tinnel and session create/destroy.
 *****************************************************************************/

/* Tunnel socket destruct hook.
 * The tunnel context is deleted only when all session sockets have been
 * closed.
 */
1131
static void l2tp_tunnel_destruct(struct sock *sk)
1132
{
1133
	struct l2tp_tunnel *tunnel = l2tp_sk_to_tunnel(sk);
1134

1135
	if (!tunnel)
1136 1137
		goto end;

1138
	/* Disable udp encapsulation */
1139 1140 1141 1142 1143
	switch (tunnel->encap) {
	case L2TP_ENCAPTYPE_UDP:
		/* No longer an encapsulation socket. See net/ipv4/udp.c */
		(udp_sk(sk))->encap_type = 0;
		(udp_sk(sk))->encap_rcv = NULL;
1144
		(udp_sk(sk))->encap_destroy = NULL;
1145 1146 1147 1148
		break;
	case L2TP_ENCAPTYPE_IP:
		break;
	}
1149 1150 1151 1152 1153

	/* Remove hooks into tunnel socket */
	sk->sk_destruct = tunnel->old_sk_destruct;
	sk->sk_user_data = NULL;

1154 1155 1156
	/* Call the original destructor */
	if (sk->sk_destruct)
		(*sk->sk_destruct)(sk);
1157 1158

	kfree_rcu(tunnel, rcu);
1159 1160 1161 1162
end:
	return;
}

1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186
/* Remove an l2tp session from l2tp_core's hash lists. */
static void l2tp_session_unhash(struct l2tp_session *session)
{
	struct l2tp_tunnel *tunnel = session->tunnel;

	/* Remove the session from core hashes */
	if (tunnel) {
		/* Remove from the per-tunnel hash */
		write_lock_bh(&tunnel->hlist_lock);
		hlist_del_init(&session->hlist);
		write_unlock_bh(&tunnel->hlist_lock);

		/* For L2TPv3 we have a per-net hash: remove from there, too */
		if (tunnel->version != L2TP_HDR_VER_2) {
			struct l2tp_net *pn = l2tp_pernet(tunnel->l2tp_net);

			spin_lock_bh(&pn->l2tp_session_hlist_lock);
			hlist_del_init_rcu(&session->global_hlist);
			spin_unlock_bh(&pn->l2tp_session_hlist_lock);
			synchronize_rcu();
		}
	}
}

1187 1188
/* When the tunnel is closed, all the attached sessions need to go too.
 */
1189
static void l2tp_tunnel_closeall(struct l2tp_tunnel *tunnel)
1190 1191 1192 1193 1194 1195 1196
{
	int hash;
	struct hlist_node *walk;
	struct hlist_node *tmp;
	struct l2tp_session *session;

	write_lock_bh(&tunnel->hlist_lock);
1197
	tunnel->acpt_newsess = false;
1198 1199 1200 1201 1202 1203 1204
	for (hash = 0; hash < L2TP_HASH_SIZE; hash++) {
again:
		hlist_for_each_safe(walk, tmp, &tunnel->session_hlist[hash]) {
			session = hlist_entry(walk, struct l2tp_session, hlist);
			hlist_del_init(&session->hlist);

			write_unlock_bh(&tunnel->hlist_lock);
1205
			l2tp_session_delete(session);
1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218
			write_lock_bh(&tunnel->hlist_lock);

			/* Now restart from the beginning of this hash
			 * chain.  We always remove a session from the
			 * list so we are guaranteed to make forward
			 * progress.
			 */
			goto again;
		}
	}
	write_unlock_bh(&tunnel->hlist_lock);
}

1219 1220 1221
/* Tunnel socket destroy hook for UDP encapsulation */
static void l2tp_udp_encap_destroy(struct sock *sk)
{
1222
	struct l2tp_tunnel *tunnel = l2tp_sk_to_tunnel(sk);
1223 1224 1225

	if (tunnel)
		l2tp_tunnel_delete(tunnel);
1226 1227
}

1228 1229 1230
/* Workqueue tunnel deletion function */
static void l2tp_tunnel_del_work(struct work_struct *work)
{
1231 1232 1233 1234
	struct l2tp_tunnel *tunnel = container_of(work, struct l2tp_tunnel,
						  del_work);
	struct sock *sk = tunnel->sock;
	struct socket *sock = sk->sk_socket;
1235
	struct l2tp_net *pn;
1236 1237 1238

	l2tp_tunnel_closeall(tunnel);

1239
	/* If the tunnel socket was created within the kernel, use
1240
	 * the sk API to release it here.
1241
	 */
1242
	if (tunnel->fd < 0) {
1243
		if (sock) {
1244
			kernel_sock_shutdown(sock, SHUT_RDWR);
1245 1246
			sock_release(sock);
		}
1247
	}
1248

1249 1250 1251 1252 1253 1254
	/* Remove the tunnel struct from the tunnel list */
	pn = l2tp_pernet(tunnel->l2tp_net);
	spin_lock_bh(&pn->l2tp_tunnel_list_lock);
	list_del_rcu(&tunnel->list);
	spin_unlock_bh(&pn->l2tp_tunnel_list_lock);

1255 1256 1257 1258
	/* drop initial ref */
	l2tp_tunnel_dec_refcount(tunnel);

	/* drop workqueue ref */
1259
	l2tp_tunnel_dec_refcount(tunnel);
1260 1261
}

1262 1263 1264
/* Create a socket for the tunnel, if one isn't set up by
 * userspace. This is used for static tunnels where there is no
 * managing L2TP daemon.
1265 1266 1267 1268 1269
 *
 * Since we don't want these sockets to keep a namespace alive by
 * themselves, we drop the socket's namespace refcount after creation.
 * These sockets are freed when the namespace exits using the pernet
 * exit hook.
1270
 */
1271
static int l2tp_tunnel_sock_create(struct net *net,
1272 1273 1274 1275
				   u32 tunnel_id,
				   u32 peer_tunnel_id,
				   struct l2tp_tunnel_cfg *cfg,
				   struct socket **sockp)
1276 1277
{
	int err = -EINVAL;
1278
	struct socket *sock = NULL;
T
Tom Herbert 已提交
1279
	struct udp_port_cfg udp_conf;
1280 1281 1282

	switch (cfg->encap) {
	case L2TP_ENCAPTYPE_UDP:
T
Tom Herbert 已提交
1283 1284
		memset(&udp_conf, 0, sizeof(udp_conf));

1285 1286
#if IS_ENABLED(CONFIG_IPV6)
		if (cfg->local_ip6 && cfg->peer_ip6) {
T
Tom Herbert 已提交
1287 1288 1289 1290 1291 1292
			udp_conf.family = AF_INET6;
			memcpy(&udp_conf.local_ip6, cfg->local_ip6,
			       sizeof(udp_conf.local_ip6));
			memcpy(&udp_conf.peer_ip6, cfg->peer_ip6,
			       sizeof(udp_conf.peer_ip6));
			udp_conf.use_udp6_tx_checksums =
1293
			  !cfg->udp6_zero_tx_checksums;
T
Tom Herbert 已提交
1294
			udp_conf.use_udp6_rx_checksums =
1295
			  !cfg->udp6_zero_rx_checksums;
1296 1297 1298
		} else
#endif
		{
T
Tom Herbert 已提交
1299 1300 1301 1302
			udp_conf.family = AF_INET;
			udp_conf.local_ip = cfg->local_ip;
			udp_conf.peer_ip = cfg->peer_ip;
			udp_conf.use_udp_checksums = cfg->use_udp_checksums;
1303
		}
1304

T
Tom Herbert 已提交
1305 1306 1307 1308 1309 1310
		udp_conf.local_udp_port = htons(cfg->local_udp_port);
		udp_conf.peer_udp_port = htons(cfg->peer_udp_port);

		err = udp_sock_create(net, &udp_conf, &sock);
		if (err < 0)
			goto out;
1311 1312 1313 1314

		break;

	case L2TP_ENCAPTYPE_IP:
1315 1316
#if IS_ENABLED(CONFIG_IPV6)
		if (cfg->local_ip6 && cfg->peer_ip6) {
T
Tom Herbert 已提交
1317 1318
			struct sockaddr_l2tpip6 ip6_addr = {0};

1319
			err = sock_create_kern(net, AF_INET6, SOCK_DGRAM,
1320
					       IPPROTO_L2TP, &sock);
1321 1322
			if (err < 0)
				goto out;
1323

1324 1325 1326 1327
			ip6_addr.l2tp_family = AF_INET6;
			memcpy(&ip6_addr.l2tp_addr, cfg->local_ip6,
			       sizeof(ip6_addr.l2tp_addr));
			ip6_addr.l2tp_conn_id = tunnel_id;
1328
			err = kernel_bind(sock, (struct sockaddr *)&ip6_addr,
1329 1330 1331
					  sizeof(ip6_addr));
			if (err < 0)
				goto out;
1332

1333 1334 1335 1336 1337
			ip6_addr.l2tp_family = AF_INET6;
			memcpy(&ip6_addr.l2tp_addr, cfg->peer_ip6,
			       sizeof(ip6_addr.l2tp_addr));
			ip6_addr.l2tp_conn_id = peer_tunnel_id;
			err = kernel_connect(sock,
1338
					     (struct sockaddr *)&ip6_addr,
1339 1340 1341 1342 1343 1344
					     sizeof(ip6_addr), 0);
			if (err < 0)
				goto out;
		} else
#endif
		{
T
Tom Herbert 已提交
1345 1346
			struct sockaddr_l2tpip ip_addr = {0};

1347
			err = sock_create_kern(net, AF_INET, SOCK_DGRAM,
1348
					       IPPROTO_L2TP, &sock);
1349 1350
			if (err < 0)
				goto out;
1351

1352 1353 1354
			ip_addr.l2tp_family = AF_INET;
			ip_addr.l2tp_addr = cfg->local_ip;
			ip_addr.l2tp_conn_id = tunnel_id;
1355
			err = kernel_bind(sock, (struct sockaddr *)&ip_addr,
1356 1357 1358 1359 1360 1361 1362
					  sizeof(ip_addr));
			if (err < 0)
				goto out;

			ip_addr.l2tp_family = AF_INET;
			ip_addr.l2tp_addr = cfg->peer_ip;
			ip_addr.l2tp_conn_id = peer_tunnel_id;
1363
			err = kernel_connect(sock, (struct sockaddr *)&ip_addr,
1364 1365 1366 1367
					     sizeof(ip_addr), 0);
			if (err < 0)
				goto out;
		}
1368 1369 1370 1371 1372 1373 1374
		break;

	default:
		goto out;
	}

out:
1375
	*sockp = sock;
1376
	if (err < 0 && sock) {
1377
		kernel_sock_shutdown(sock, SHUT_RDWR);
1378
		sock_release(sock);
1379 1380 1381 1382 1383 1384
		*sockp = NULL;
	}

	return err;
}

E
Eric Dumazet 已提交
1385 1386
static struct lock_class_key l2tp_socket_class;

1387
int l2tp_tunnel_create(int fd, int version, u32 tunnel_id, u32 peer_tunnel_id,
1388
		       struct l2tp_tunnel_cfg *cfg, struct l2tp_tunnel **tunnelp)
1389 1390 1391
{
	struct l2tp_tunnel *tunnel = NULL;
	int err;
1392
	enum l2tp_encap_type encap = L2TP_ENCAPTYPE_UDP;
1393

1394
	if (cfg)
1395 1396
		encap = cfg->encap;

1397
	tunnel = kzalloc(sizeof(*tunnel), GFP_KERNEL);
1398
	if (!tunnel) {
1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409
		err = -ENOMEM;
		goto err;
	}

	tunnel->version = version;
	tunnel->tunnel_id = tunnel_id;
	tunnel->peer_tunnel_id = peer_tunnel_id;

	tunnel->magic = L2TP_TUNNEL_MAGIC;
	sprintf(&tunnel->name[0], "tunl %u", tunnel_id);
	rwlock_init(&tunnel->hlist_lock);
1410
	tunnel->acpt_newsess = true;
1411

1412
	tunnel->encap = encap;
1413

1414 1415 1416
	refcount_set(&tunnel->ref_count, 1);
	tunnel->fd = fd;

1417 1418 1419
	/* Init delete workqueue struct */
	INIT_WORK(&tunnel->del_work, l2tp_tunnel_del_work);

1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430
	INIT_LIST_HEAD(&tunnel->list);

	err = 0;
err:
	if (tunnelp)
		*tunnelp = tunnel;

	return err;
}
EXPORT_SYMBOL_GPL(l2tp_tunnel_create);

1431 1432 1433 1434 1435 1436 1437 1438 1439
static int l2tp_validate_socket(const struct sock *sk, const struct net *net,
				enum l2tp_encap_type encap)
{
	if (!net_eq(sock_net(sk), net))
		return -EINVAL;

	if (sk->sk_type != SOCK_DGRAM)
		return -EPROTONOSUPPORT;

1440 1441 1442
	if (sk->sk_family != PF_INET && sk->sk_family != PF_INET6)
		return -EPROTONOSUPPORT;

1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455
	if ((encap == L2TP_ENCAPTYPE_UDP && sk->sk_protocol != IPPROTO_UDP) ||
	    (encap == L2TP_ENCAPTYPE_IP && sk->sk_protocol != IPPROTO_L2TP))
		return -EPROTONOSUPPORT;

	if (sk->sk_user_data)
		return -EBUSY;

	return 0;
}

int l2tp_tunnel_register(struct l2tp_tunnel *tunnel, struct net *net,
			 struct l2tp_tunnel_cfg *cfg)
{
1456
	struct l2tp_tunnel *tunnel_walk;
1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479
	struct l2tp_net *pn;
	struct socket *sock;
	struct sock *sk;
	int ret;

	if (tunnel->fd < 0) {
		ret = l2tp_tunnel_sock_create(net, tunnel->tunnel_id,
					      tunnel->peer_tunnel_id, cfg,
					      &sock);
		if (ret < 0)
			goto err;
	} else {
		sock = sockfd_lookup(tunnel->fd, &ret);
		if (!sock)
			goto err;

		ret = l2tp_validate_socket(sock->sk, net, tunnel->encap);
		if (ret < 0)
			goto err_sock;
	}

	tunnel->l2tp_net = net;
	pn = l2tp_pernet(net);
1480

1481
	spin_lock_bh(&pn->l2tp_tunnel_list_lock);
1482 1483 1484 1485 1486 1487 1488 1489
	list_for_each_entry(tunnel_walk, &pn->l2tp_tunnel_list, list) {
		if (tunnel_walk->tunnel_id == tunnel->tunnel_id) {
			spin_unlock_bh(&pn->l2tp_tunnel_list_lock);

			ret = -EEXIST;
			goto err_sock;
		}
	}
1490 1491 1492
	list_add_rcu(&tunnel->list, &pn->l2tp_tunnel_list);
	spin_unlock_bh(&pn->l2tp_tunnel_list_lock);

1493 1494 1495 1496
	sk = sock->sk;
	sock_hold(sk);
	tunnel->sock = sk;

1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515
	if (tunnel->encap == L2TP_ENCAPTYPE_UDP) {
		struct udp_tunnel_sock_cfg udp_cfg = {
			.sk_user_data = tunnel,
			.encap_type = UDP_ENCAP_L2TPINUDP,
			.encap_rcv = l2tp_udp_encap_recv,
			.encap_destroy = l2tp_udp_encap_destroy,
		};

		setup_udp_tunnel_sock(net, sock, &udp_cfg);
	} else {
		sk->sk_user_data = tunnel;
	}

	tunnel->old_sk_destruct = sk->sk_destruct;
	sk->sk_destruct = &l2tp_tunnel_destruct;
	lockdep_set_class_and_name(&sk->sk_lock.slock, &l2tp_socket_class,
				   "l2tp_sock");
	sk->sk_allocation = GFP_ATOMIC;

1516 1517
	trace_register_tunnel(tunnel);

1518 1519 1520 1521 1522 1523
	if (tunnel->fd >= 0)
		sockfd_put(sock);

	return 0;

err_sock:
1524 1525 1526 1527
	if (tunnel->fd < 0)
		sock_release(sock);
	else
		sockfd_put(sock);
1528 1529 1530 1531 1532
err:
	return ret;
}
EXPORT_SYMBOL_GPL(l2tp_tunnel_register);

1533 1534
/* This function is used by the netlink TUNNEL_DELETE command.
 */
1535
void l2tp_tunnel_delete(struct l2tp_tunnel *tunnel)
1536
{
1537
	if (!test_and_set_bit(0, &tunnel->dead)) {
1538
		trace_delete_tunnel(tunnel);
1539 1540
		l2tp_tunnel_inc_refcount(tunnel);
		queue_work(l2tp_wq, &tunnel->del_work);
1541
	}
1542 1543 1544
}
EXPORT_SYMBOL_GPL(l2tp_tunnel_delete);

1545
void l2tp_session_delete(struct l2tp_session *session)
1546
{
1547
	if (test_and_set_bit(0, &session->dead))
1548
		return;
1549

1550
	trace_delete_session(session);
1551
	l2tp_session_unhash(session);
1552
	l2tp_session_queue_purge(session);
1553
	if (session->session_close)
1554
		(*session->session_close)(session);
1555

1556 1557 1558 1559
	l2tp_session_dec_refcount(session);
}
EXPORT_SYMBOL_GPL(l2tp_session_delete);

1560
/* We come here whenever a session's send_seq, cookie_len or
1561
 * l2specific_type parameters are set.
1562
 */
1563
void l2tp_session_set_header_len(struct l2tp_session *session, int version)
1564 1565 1566 1567 1568 1569
{
	if (version == L2TP_HDR_VER_2) {
		session->hdr_len = 6;
		if (session->send_seq)
			session->hdr_len += 4;
	} else {
1570 1571
		session->hdr_len = 4 + session->cookie_len;
		session->hdr_len += l2tp_get_l2specific_len(session);
1572 1573
		if (session->tunnel->encap == L2TP_ENCAPTYPE_UDP)
			session->hdr_len += 4;
1574 1575
	}
}
1576
EXPORT_SYMBOL_GPL(l2tp_session_set_header_len);
1577

1578 1579
struct l2tp_session *l2tp_session_create(int priv_size, struct l2tp_tunnel *tunnel, u32 session_id,
					 u32 peer_session_id, struct l2tp_session_cfg *cfg)
1580 1581 1582
{
	struct l2tp_session *session;

1583
	session = kzalloc(sizeof(*session) + priv_size, GFP_KERNEL);
1584
	if (session) {
1585 1586 1587 1588 1589
		session->magic = L2TP_SESSION_MAGIC;
		session->tunnel = tunnel;

		session->session_id = session_id;
		session->peer_session_id = peer_session_id;
1590
		session->nr = 0;
1591 1592 1593 1594 1595
		if (tunnel->version == L2TP_HDR_VER_2)
			session->nr_max = 0xffff;
		else
			session->nr_max = 0xffffff;
		session->nr_window_size = session->nr_max / 2;
1596 1597 1598 1599
		session->nr_oos_count_max = 4;

		/* Use NR of first received packet */
		session->reorder_skip = 1;
1600 1601 1602 1603 1604 1605 1606

		sprintf(&session->name[0], "sess %u/%u",
			tunnel->tunnel_id, session->session_id);

		skb_queue_head_init(&session->reorder_q);

		INIT_HLIST_NODE(&session->hlist);
1607
		INIT_HLIST_NODE(&session->global_hlist);
1608 1609

		if (cfg) {
1610
			session->pwtype = cfg->pw_type;
1611 1612 1613
			session->send_seq = cfg->send_seq;
			session->recv_seq = cfg->recv_seq;
			session->lns_mode = cfg->lns_mode;
1614 1615 1616 1617 1618 1619
			session->reorder_timeout = cfg->reorder_timeout;
			session->l2specific_type = cfg->l2specific_type;
			session->cookie_len = cfg->cookie_len;
			memcpy(&session->cookie[0], &cfg->cookie[0], cfg->cookie_len);
			session->peer_cookie_len = cfg->peer_cookie_len;
			memcpy(&session->peer_cookie[0], &cfg->peer_cookie[0], cfg->peer_cookie_len);
1620 1621
		}

1622 1623
		l2tp_session_set_header_len(session, tunnel->version);

1624 1625
		refcount_set(&session->ref_count, 1);

1626
		return session;
1627 1628
	}

1629
	return ERR_PTR(-ENOMEM);
1630 1631 1632 1633 1634 1635 1636 1637 1638
}
EXPORT_SYMBOL_GPL(l2tp_session_create);

/*****************************************************************************
 * Init and cleanup
 *****************************************************************************/

static __net_init int l2tp_init_net(struct net *net)
{
J
Jiri Pirko 已提交
1639
	struct l2tp_net *pn = net_generic(net, l2tp_net_id);
1640
	int hash;
1641 1642

	INIT_LIST_HEAD(&pn->l2tp_tunnel_list);
1643
	spin_lock_init(&pn->l2tp_tunnel_list_lock);
1644

1645 1646 1647
	for (hash = 0; hash < L2TP_HASH_SIZE_2; hash++)
		INIT_HLIST_HEAD(&pn->l2tp_session_hlist[hash]);

1648
	spin_lock_init(&pn->l2tp_session_hlist_lock);
1649

1650 1651 1652
	return 0;
}

1653 1654 1655 1656
static __net_exit void l2tp_exit_net(struct net *net)
{
	struct l2tp_net *pn = l2tp_pernet(net);
	struct l2tp_tunnel *tunnel = NULL;
1657
	int hash;
1658 1659 1660

	rcu_read_lock_bh();
	list_for_each_entry_rcu(tunnel, &pn->l2tp_tunnel_list, list) {
1661
		l2tp_tunnel_delete(tunnel);
1662 1663
	}
	rcu_read_unlock_bh();
1664

1665 1666
	if (l2tp_wq)
		flush_workqueue(l2tp_wq);
1667
	rcu_barrier();
1668 1669 1670

	for (hash = 0; hash < L2TP_HASH_SIZE_2; hash++)
		WARN_ON_ONCE(!hlist_empty(&pn->l2tp_session_hlist[hash]));
1671 1672
}

1673 1674
static struct pernet_operations l2tp_net_ops = {
	.init = l2tp_init_net,
1675
	.exit = l2tp_exit_net,
1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687
	.id   = &l2tp_net_id,
	.size = sizeof(struct l2tp_net),
};

static int __init l2tp_init(void)
{
	int rc = 0;

	rc = register_pernet_device(&l2tp_net_ops);
	if (rc)
		goto out;

1688
	l2tp_wq = alloc_workqueue("l2tp", WQ_UNBOUND, 0);
1689 1690
	if (!l2tp_wq) {
		pr_err("alloc_workqueue failed\n");
1691
		unregister_pernet_device(&l2tp_net_ops);
1692 1693 1694 1695
		rc = -ENOMEM;
		goto out;
	}

1696
	pr_info("L2TP core driver, %s\n", L2TP_DRV_VERSION);
1697 1698 1699 1700 1701 1702 1703 1704

out:
	return rc;
}

static void __exit l2tp_exit(void)
{
	unregister_pernet_device(&l2tp_net_ops);
1705 1706 1707 1708
	if (l2tp_wq) {
		destroy_workqueue(l2tp_wq);
		l2tp_wq = NULL;
	}
1709 1710 1711 1712 1713 1714 1715 1716 1717
}

module_init(l2tp_init);
module_exit(l2tp_exit);

MODULE_AUTHOR("James Chapman <jchapman@katalix.com>");
MODULE_DESCRIPTION("L2TP core");
MODULE_LICENSE("GPL");
MODULE_VERSION(L2TP_DRV_VERSION);
新手
引导
客服 返回
顶部