ip_gre.c 31.5 KB
Newer Older
L
Linus Torvalds 已提交
1
/*
2
 *	Linux NET3:	GRE over IP protocol decoder.
L
Linus Torvalds 已提交
3 4 5 6 7 8 9 10 11 12
 *
 *	Authors: Alexey Kuznetsov (kuznet@ms2.inr.ac.ru)
 *
 *	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.
 *
 */

13
#include <linux/capability.h>
L
Linus Torvalds 已提交
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <asm/uaccess.h>
#include <linux/skbuff.h>
#include <linux/netdevice.h>
#include <linux/in.h>
#include <linux/tcp.h>
#include <linux/udp.h>
#include <linux/if_arp.h>
#include <linux/mroute.h>
#include <linux/init.h>
#include <linux/in6.h>
#include <linux/inetdevice.h>
#include <linux/igmp.h>
#include <linux/netfilter_ipv4.h>
30
#include <linux/if_ether.h>
L
Linus Torvalds 已提交
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

#include <net/sock.h>
#include <net/ip.h>
#include <net/icmp.h>
#include <net/protocol.h>
#include <net/ipip.h>
#include <net/arp.h>
#include <net/checksum.h>
#include <net/dsfield.h>
#include <net/inet_ecn.h>
#include <net/xfrm.h>

#ifdef CONFIG_IPV6
#include <net/ipv6.h>
#include <net/ip6_fib.h>
#include <net/ip6_route.h>
#endif

/*
   Problems & solutions
   --------------------

   1. The most important issue is detecting local dead loops.
   They would cause complete host lockup in transmit, which
   would be "resolved" by stack overflow or, if queueing is enabled,
   with infinite looping in net_bh.

   We cannot track such dead loops during route installation,
   it is infeasible task. The most general solutions would be
   to keep skb->encapsulation counter (sort of local ttl),
   and silently drop packet when it expires. It is the best
   solution, but it supposes maintaing new variable in ALL
   skb, even if no tunneling is used.

65
   Current solution: t->recursion lock breaks dead loops. It looks
L
Linus Torvalds 已提交
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
   like dev->tbusy flag, but I preferred new variable, because
   the semantics is different. One day, when hard_start_xmit
   will be multithreaded we will have to use skb->encapsulation.



   2. Networking dead loops would not kill routers, but would really
   kill network. IP hop limit plays role of "t->recursion" in this case,
   if we copy it from packet being encapsulated to upper header.
   It is very good solution, but it introduces two problems:

   - Routing protocols, using packets with ttl=1 (OSPF, RIP2),
     do not work over tunnels.
   - traceroute does not work. I planned to relay ICMP from tunnel,
     so that this problem would be solved and traceroute output
     would even more informative. This idea appeared to be wrong:
     only Linux complies to rfc1812 now (yes, guys, Linux is the only
     true router now :-)), all routers (at least, in neighbourhood of mine)
     return only 8 bytes of payload. It is the end.

   Hence, if we want that OSPF worked or traceroute said something reasonable,
   we should search for another solution.

   One of them is to parse packet trying to detect inner encapsulation
   made by our node. It is difficult or even impossible, especially,
   taking into account fragmentation. TO be short, tt is not solution at all.

   Current solution: The solution was UNEXPECTEDLY SIMPLE.
   We force DF flag on tunnels with preconfigured hop limit,
   that is ALL. :-) Well, it does not remove the problem completely,
   but exponential growth of network traffic is changed to linear
   (branches, that exceed pmtu are pruned) and tunnel mtu
   fastly degrades to value <68, where looping stops.
   Yes, it is not good if there exists a router in the loop,
   which does not force DF, even when encapsulating packets have DF set.
   But it is not our problem! Nobody could accuse us, we made
   all that we could make. Even if it is your gated who injected
   fatal route to network, even if it were you who configured
   fatal static route: you are innocent. :-)



   3. Really, ipv4/ipip.c, ipv4/ip_gre.c and ipv6/sit.c contain
   practically identical code. It would be good to glue them
   together, but it is not very evident, how to make them modular.
   sit is integral part of IPv6, ipip and gre are naturally modular.
   We could extract common parts (hash table, ioctl etc)
   to a separate module (ip_tunnel.c).

   Alexey Kuznetsov.
 */

static int ipgre_tunnel_init(struct net_device *dev);
static void ipgre_tunnel_setup(struct net_device *dev);

/* Fallback tunnel: no source, no destination, no key, no options */

static int ipgre_fb_tunnel_init(struct net_device *dev);

static struct net_device *ipgre_fb_tunnel_dev;

/* Tunnel hash table */

/*
   4 hash tables:

   3: (remote,local)
   2: (remote,*)
   1: (*,local)
   0: (*,*)

   We require exact key match i.e. if a key is present in packet
   it will match only tunnel with the same key; if it is not present,
   it will match only keyless tunnel.

   All keysless packets, if not matched configured keyless tunnels
   will match fallback tunnel.
 */

#define HASH_SIZE  16
A
Al Viro 已提交
146
#define HASH(addr) (((__force u32)addr^((__force u32)addr>>4))&0xF)
L
Linus Torvalds 已提交
147 148 149 150 151 152 153 154 155 156 157 158

static struct ip_tunnel *tunnels[4][HASH_SIZE];

#define tunnels_r_l	(tunnels[3])
#define tunnels_r	(tunnels[2])
#define tunnels_l	(tunnels[1])
#define tunnels_wc	(tunnels[0])

static DEFINE_RWLOCK(ipgre_lock);

/* Given src, dst and key, find appropriate for input tunnel. */

A
Al Viro 已提交
159
static struct ip_tunnel * ipgre_tunnel_lookup(__be32 remote, __be32 local, __be32 key)
L
Linus Torvalds 已提交
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
{
	unsigned h0 = HASH(remote);
	unsigned h1 = HASH(key);
	struct ip_tunnel *t;

	for (t = tunnels_r_l[h0^h1]; t; t = t->next) {
		if (local == t->parms.iph.saddr && remote == t->parms.iph.daddr) {
			if (t->parms.i_key == key && (t->dev->flags&IFF_UP))
				return t;
		}
	}
	for (t = tunnels_r[h0^h1]; t; t = t->next) {
		if (remote == t->parms.iph.daddr) {
			if (t->parms.i_key == key && (t->dev->flags&IFF_UP))
				return t;
		}
	}
	for (t = tunnels_l[h1]; t; t = t->next) {
		if (local == t->parms.iph.saddr ||
		     (local == t->parms.iph.daddr && MULTICAST(local))) {
			if (t->parms.i_key == key && (t->dev->flags&IFF_UP))
				return t;
		}
	}
	for (t = tunnels_wc[h1]; t; t = t->next) {
		if (t->parms.i_key == key && (t->dev->flags&IFF_UP))
			return t;
	}

	if (ipgre_fb_tunnel_dev->flags&IFF_UP)
190
		return netdev_priv(ipgre_fb_tunnel_dev);
L
Linus Torvalds 已提交
191 192 193
	return NULL;
}

194
static struct ip_tunnel **__ipgre_bucket(struct ip_tunnel_parm *parms)
L
Linus Torvalds 已提交
195
{
196 197 198
	__be32 remote = parms->iph.daddr;
	__be32 local = parms->iph.saddr;
	__be32 key = parms->i_key;
L
Linus Torvalds 已提交
199 200 201 202 203 204 205 206 207 208 209 210 211
	unsigned h = HASH(key);
	int prio = 0;

	if (local)
		prio |= 1;
	if (remote && !MULTICAST(remote)) {
		prio |= 2;
		h ^= HASH(remote);
	}

	return &tunnels[prio][h];
}

212 213 214 215 216
static inline struct ip_tunnel **ipgre_bucket(struct ip_tunnel *t)
{
	return __ipgre_bucket(&t->parms);
}

L
Linus Torvalds 已提交
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
static void ipgre_tunnel_link(struct ip_tunnel *t)
{
	struct ip_tunnel **tp = ipgre_bucket(t);

	t->next = *tp;
	write_lock_bh(&ipgre_lock);
	*tp = t;
	write_unlock_bh(&ipgre_lock);
}

static void ipgre_tunnel_unlink(struct ip_tunnel *t)
{
	struct ip_tunnel **tp;

	for (tp = ipgre_bucket(t); *tp; tp = &(*tp)->next) {
		if (t == *tp) {
			write_lock_bh(&ipgre_lock);
			*tp = t->next;
			write_unlock_bh(&ipgre_lock);
			break;
		}
	}
}

static struct ip_tunnel * ipgre_tunnel_locate(struct ip_tunnel_parm *parms, int create)
{
A
Al Viro 已提交
243 244 245
	__be32 remote = parms->iph.daddr;
	__be32 local = parms->iph.saddr;
	__be32 key = parms->i_key;
L
Linus Torvalds 已提交
246 247 248 249
	struct ip_tunnel *t, **tp, *nt;
	struct net_device *dev;
	char name[IFNAMSIZ];

250
	for (tp = __ipgre_bucket(parms); (t = *tp) != NULL; tp = &t->next) {
L
Linus Torvalds 已提交
251 252 253 254 255 256 257 258 259 260 261 262 263 264
		if (local == t->parms.iph.saddr && remote == t->parms.iph.daddr) {
			if (key == t->parms.i_key)
				return t;
		}
	}
	if (!create)
		return NULL;

	if (parms->name[0])
		strlcpy(name, parms->name, IFNAMSIZ);
	else {
		int i;
		for (i=1; i<100; i++) {
			sprintf(name, "gre%d", i);
265
			if (__dev_get_by_name(&init_net, name) == NULL)
L
Linus Torvalds 已提交
266 267 268 269 270 271 272 273 274 275 276
				break;
		}
		if (i==100)
			goto failed;
	}

	dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
	if (!dev)
	  return NULL;

	dev->init = ipgre_tunnel_init;
277
	nt = netdev_priv(dev);
L
Linus Torvalds 已提交
278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294
	nt->parms = *parms;

	if (register_netdevice(dev) < 0) {
		free_netdev(dev);
		goto failed;
	}

	dev_hold(dev);
	ipgre_tunnel_link(nt);
	return nt;

failed:
	return NULL;
}

static void ipgre_tunnel_uninit(struct net_device *dev)
{
295
	ipgre_tunnel_unlink(netdev_priv(dev));
L
Linus Torvalds 已提交
296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317
	dev_put(dev);
}


static void ipgre_err(struct sk_buff *skb, u32 info)
{
#ifndef I_WISH_WORLD_WERE_PERFECT

/* It is not :-( All the routers (except for Linux) return only
   8 bytes of packet payload. It means, that precise relaying of
   ICMP in the real Internet is absolutely infeasible.

   Moreover, Cisco "wise men" put GRE key to the third word
   in GRE header. It makes impossible maintaining even soft state for keyed
   GRE tunnels with enabled checksum. Tell them "thank you".

   Well, I wonder, rfc1812 was written by Cisco employee,
   what the hell these idiots break standrads established
   by themself???
 */

	struct iphdr *iph = (struct iphdr*)skb->data;
A
Al Viro 已提交
318
	__be16	     *p = (__be16*)(skb->data+(iph->ihl<<2));
L
Linus Torvalds 已提交
319
	int grehlen = (iph->ihl<<2) + 4;
320 321
	const int type = icmp_hdr(skb)->type;
	const int code = icmp_hdr(skb)->code;
L
Linus Torvalds 已提交
322
	struct ip_tunnel *t;
A
Al Viro 已提交
323
	__be16 flags;
L
Linus Torvalds 已提交
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

	flags = p[0];
	if (flags&(GRE_CSUM|GRE_KEY|GRE_SEQ|GRE_ROUTING|GRE_VERSION)) {
		if (flags&(GRE_VERSION|GRE_ROUTING))
			return;
		if (flags&GRE_KEY) {
			grehlen += 4;
			if (flags&GRE_CSUM)
				grehlen += 4;
		}
	}

	/* If only 8 bytes returned, keyed message will be dropped here */
	if (skb_headlen(skb) < grehlen)
		return;

	switch (type) {
	default:
	case ICMP_PARAMETERPROB:
		return;

	case ICMP_DEST_UNREACH:
		switch (code) {
		case ICMP_SR_FAILED:
		case ICMP_PORT_UNREACH:
			/* Impossible event. */
			return;
		case ICMP_FRAG_NEEDED:
			/* Soft state for pmtu is maintained by IP core. */
			return;
		default:
			/* All others are translated to HOST_UNREACH.
			   rfc2003 contains "deep thoughts" about NET_UNREACH,
			   I believe they are just ether pollution. --ANK
			 */
			break;
		}
		break;
	case ICMP_TIME_EXCEEDED:
		if (code != ICMP_EXC_TTL)
			return;
		break;
	}

	read_lock(&ipgre_lock);
A
Al Viro 已提交
369
	t = ipgre_tunnel_lookup(iph->daddr, iph->saddr, (flags&GRE_KEY) ? *(((__be32*)p) + (grehlen>>2) - 1) : 0);
L
Linus Torvalds 已提交
370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386
	if (t == NULL || t->parms.iph.daddr == 0 || MULTICAST(t->parms.iph.daddr))
		goto out;

	if (t->parms.iph.ttl == 0 && type == ICMP_TIME_EXCEEDED)
		goto out;

	if (jiffies - t->err_time < IPTUNNEL_ERR_TIMEO)
		t->err_count++;
	else
		t->err_count = 1;
	t->err_time = jiffies;
out:
	read_unlock(&ipgre_lock);
	return;
#else
	struct iphdr *iph = (struct iphdr*)dp;
	struct iphdr *eiph;
A
Al Viro 已提交
387
	__be16	     *p = (__be16*)(dp+(iph->ihl<<2));
388 389
	const int type = icmp_hdr(skb)->type;
	const int code = icmp_hdr(skb)->code;
L
Linus Torvalds 已提交
390 391
	int rel_type = 0;
	int rel_code = 0;
392 393
	__be32 rel_info = 0;
	__u32 n = 0;
A
Al Viro 已提交
394
	__be16 flags;
L
Linus Torvalds 已提交
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
	int grehlen = (iph->ihl<<2) + 4;
	struct sk_buff *skb2;
	struct flowi fl;
	struct rtable *rt;

	if (p[1] != htons(ETH_P_IP))
		return;

	flags = p[0];
	if (flags&(GRE_CSUM|GRE_KEY|GRE_SEQ|GRE_ROUTING|GRE_VERSION)) {
		if (flags&(GRE_VERSION|GRE_ROUTING))
			return;
		if (flags&GRE_CSUM)
			grehlen += 4;
		if (flags&GRE_KEY)
			grehlen += 4;
		if (flags&GRE_SEQ)
			grehlen += 4;
	}
	if (len < grehlen + sizeof(struct iphdr))
		return;
	eiph = (struct iphdr*)(dp + grehlen);

	switch (type) {
	default:
		return;
	case ICMP_PARAMETERPROB:
422
		n = ntohl(icmp_hdr(skb)->un.gateway) >> 24;
423
		if (n < (iph->ihl<<2))
L
Linus Torvalds 已提交
424 425 426 427 428 429
			return;

		/* So... This guy found something strange INSIDE encapsulated
		   packet. Well, he is fool, but what can we do ?
		 */
		rel_type = ICMP_PARAMETERPROB;
430 431
		n -= grehlen;
		rel_info = htonl(n << 24);
L
Linus Torvalds 已提交
432 433 434 435 436 437 438 439 440 441
		break;

	case ICMP_DEST_UNREACH:
		switch (code) {
		case ICMP_SR_FAILED:
		case ICMP_PORT_UNREACH:
			/* Impossible event. */
			return;
		case ICMP_FRAG_NEEDED:
			/* And it is the only really necessary thing :-) */
442
			n = ntohs(icmp_hdr(skb)->un.frag.mtu);
443
			if (n < grehlen+68)
L
Linus Torvalds 已提交
444
				return;
445
			n -= grehlen;
L
Linus Torvalds 已提交
446
			/* BSD 4.2 MORE DOES NOT EXIST IN NATURE. */
447
			if (n > ntohs(eiph->tot_len))
L
Linus Torvalds 已提交
448
				return;
449
			rel_info = htonl(n);
L
Linus Torvalds 已提交
450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473
			break;
		default:
			/* All others are translated to HOST_UNREACH.
			   rfc2003 contains "deep thoughts" about NET_UNREACH,
			   I believe, it is just ether pollution. --ANK
			 */
			rel_type = ICMP_DEST_UNREACH;
			rel_code = ICMP_HOST_UNREACH;
			break;
		}
		break;
	case ICMP_TIME_EXCEEDED:
		if (code != ICMP_EXC_TTL)
			return;
		break;
	}

	/* Prepare fake skb to feed it to icmp_send */
	skb2 = skb_clone(skb, GFP_ATOMIC);
	if (skb2 == NULL)
		return;
	dst_release(skb2->dst);
	skb2->dst = NULL;
	skb_pull(skb2, skb->data - (u8*)eiph);
474
	skb_reset_network_header(skb2);
L
Linus Torvalds 已提交
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

	/* Try to guess incoming interface */
	memset(&fl, 0, sizeof(fl));
	fl.fl4_dst = eiph->saddr;
	fl.fl4_tos = RT_TOS(eiph->tos);
	fl.proto = IPPROTO_GRE;
	if (ip_route_output_key(&rt, &fl)) {
		kfree_skb(skb2);
		return;
	}
	skb2->dev = rt->u.dst.dev;

	/* route "incoming" packet */
	if (rt->rt_flags&RTCF_LOCAL) {
		ip_rt_put(rt);
		rt = NULL;
		fl.fl4_dst = eiph->daddr;
		fl.fl4_src = eiph->saddr;
		fl.fl4_tos = eiph->tos;
		if (ip_route_output_key(&rt, &fl) ||
		    rt->u.dst.dev->type != ARPHRD_IPGRE) {
			ip_rt_put(rt);
			kfree_skb(skb2);
			return;
		}
	} else {
		ip_rt_put(rt);
		if (ip_route_input(skb2, eiph->daddr, eiph->saddr, eiph->tos, skb2->dev) ||
		    skb2->dst->dev->type != ARPHRD_IPGRE) {
			kfree_skb(skb2);
			return;
		}
	}

	/* change mtu on this route */
	if (type == ICMP_DEST_UNREACH && code == ICMP_FRAG_NEEDED) {
511
		if (n > dst_mtu(skb2->dst)) {
L
Linus Torvalds 已提交
512 513 514
			kfree_skb(skb2);
			return;
		}
515
		skb2->dst->ops->update_pmtu(skb2->dst, n);
L
Linus Torvalds 已提交
516
	} else if (type == ICMP_TIME_EXCEEDED) {
517
		struct ip_tunnel *t = netdev_priv(skb2->dev);
L
Linus Torvalds 已提交
518 519 520 521 522 523 524 525 526 527 528 529 530 531 532
		if (t->parms.iph.ttl) {
			rel_type = ICMP_DEST_UNREACH;
			rel_code = ICMP_HOST_UNREACH;
		}
	}

	icmp_send(skb2, rel_type, rel_code, rel_info);
	kfree_skb(skb2);
#endif
}

static inline void ipgre_ecn_decapsulate(struct iphdr *iph, struct sk_buff *skb)
{
	if (INET_ECN_is_ce(iph->tos)) {
		if (skb->protocol == htons(ETH_P_IP)) {
533
			IP_ECN_set_ce(ip_hdr(skb));
L
Linus Torvalds 已提交
534
		} else if (skb->protocol == htons(ETH_P_IPV6)) {
535
			IP6_ECN_set_ce(ipv6_hdr(skb));
L
Linus Torvalds 已提交
536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554
		}
	}
}

static inline u8
ipgre_ecn_encapsulate(u8 tos, struct iphdr *old_iph, struct sk_buff *skb)
{
	u8 inner = 0;
	if (skb->protocol == htons(ETH_P_IP))
		inner = old_iph->tos;
	else if (skb->protocol == htons(ETH_P_IPV6))
		inner = ipv6_get_dsfield((struct ipv6hdr *)old_iph);
	return INET_ECN_encapsulate(tos, inner);
}

static int ipgre_rcv(struct sk_buff *skb)
{
	struct iphdr *iph;
	u8     *h;
A
Al Viro 已提交
555
	__be16    flags;
556
	__sum16   csum = 0;
A
Al Viro 已提交
557
	__be32 key = 0;
L
Linus Torvalds 已提交
558 559 560 561 562 563 564
	u32    seqno = 0;
	struct ip_tunnel *tunnel;
	int    offset = 4;

	if (!pskb_may_pull(skb, 16))
		goto drop_nolock;

565
	iph = ip_hdr(skb);
L
Linus Torvalds 已提交
566
	h = skb->data;
A
Al Viro 已提交
567
	flags = *(__be16*)h;
L
Linus Torvalds 已提交
568 569 570 571 572 573 574 575 576

	if (flags&(GRE_CSUM|GRE_KEY|GRE_ROUTING|GRE_SEQ|GRE_VERSION)) {
		/* - Version must be 0.
		   - We do not support routing headers.
		 */
		if (flags&(GRE_VERSION|GRE_ROUTING))
			goto drop_nolock;

		if (flags&GRE_CSUM) {
577
			switch (skb->ip_summed) {
578
			case CHECKSUM_COMPLETE:
579
				csum = csum_fold(skb->csum);
580 581 582 583 584 585
				if (!csum)
					break;
				/* fall through */
			case CHECKSUM_NONE:
				skb->csum = 0;
				csum = __skb_checksum_complete(skb);
586
				skb->ip_summed = CHECKSUM_COMPLETE;
L
Linus Torvalds 已提交
587 588 589 590
			}
			offset += 4;
		}
		if (flags&GRE_KEY) {
A
Al Viro 已提交
591
			key = *(__be32*)(h + offset);
L
Linus Torvalds 已提交
592 593 594
			offset += 4;
		}
		if (flags&GRE_SEQ) {
A
Al Viro 已提交
595
			seqno = ntohl(*(__be32*)(h + offset));
L
Linus Torvalds 已提交
596 597 598 599 600 601 602 603
			offset += 4;
		}
	}

	read_lock(&ipgre_lock);
	if ((tunnel = ipgre_tunnel_lookup(iph->saddr, iph->daddr, key)) != NULL) {
		secpath_reset(skb);

A
Al Viro 已提交
604
		skb->protocol = *(__be16*)(h + 2);
L
Linus Torvalds 已提交
605 606 607 608 609
		/* WCCP version 1 and 2 protocol decoding.
		 * - Change protocol to IP
		 * - When dealing with WCCPv2, Skip extra 4 bytes in GRE header
		 */
		if (flags == 0 &&
610 611
		    skb->protocol == htons(ETH_P_WCCP)) {
			skb->protocol = htons(ETH_P_IP);
612
			if ((*(h + offset) & 0xF0) != 0x40)
L
Linus Torvalds 已提交
613 614 615
				offset += 4;
		}

616
		skb_reset_mac_header(skb);
617 618
		__pskb_pull(skb, offset);
		skb_reset_network_header(skb);
619
		skb_postpull_rcsum(skb, skb_transport_header(skb), offset);
L
Linus Torvalds 已提交
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
		skb->pkt_type = PACKET_HOST;
#ifdef CONFIG_NET_IPGRE_BROADCAST
		if (MULTICAST(iph->daddr)) {
			/* Looped back packet, drop it! */
			if (((struct rtable*)skb->dst)->fl.iif == 0)
				goto drop;
			tunnel->stat.multicast++;
			skb->pkt_type = PACKET_BROADCAST;
		}
#endif

		if (((flags&GRE_CSUM) && csum) ||
		    (!(flags&GRE_CSUM) && tunnel->parms.i_flags&GRE_CSUM)) {
			tunnel->stat.rx_crc_errors++;
			tunnel->stat.rx_errors++;
			goto drop;
		}
		if (tunnel->parms.i_flags&GRE_SEQ) {
			if (!(flags&GRE_SEQ) ||
			    (tunnel->i_seqno && (s32)(seqno - tunnel->i_seqno) < 0)) {
				tunnel->stat.rx_fifo_errors++;
				tunnel->stat.rx_errors++;
				goto drop;
			}
			tunnel->i_seqno = seqno + 1;
		}
		tunnel->stat.rx_packets++;
		tunnel->stat.rx_bytes += skb->len;
		skb->dev = tunnel->dev;
		dst_release(skb->dst);
		skb->dst = NULL;
		nf_reset(skb);
		ipgre_ecn_decapsulate(iph, skb);
		netif_rx(skb);
		read_unlock(&ipgre_lock);
		return(0);
	}
657
	icmp_send(skb, ICMP_DEST_UNREACH, ICMP_PORT_UNREACH, 0);
L
Linus Torvalds 已提交
658 659 660 661 662 663 664 665 666 667

drop:
	read_unlock(&ipgre_lock);
drop_nolock:
	kfree_skb(skb);
	return(0);
}

static int ipgre_tunnel_xmit(struct sk_buff *skb, struct net_device *dev)
{
668
	struct ip_tunnel *tunnel = netdev_priv(dev);
L
Linus Torvalds 已提交
669
	struct net_device_stats *stats = &tunnel->stat;
670
	struct iphdr  *old_iph = ip_hdr(skb);
L
Linus Torvalds 已提交
671 672
	struct iphdr  *tiph;
	u8     tos;
A
Al Viro 已提交
673
	__be16 df;
L
Linus Torvalds 已提交
674 675 676 677 678
	struct rtable *rt;     			/* Route to the other host */
	struct net_device *tdev;			/* Device to other host */
	struct iphdr  *iph;			/* Our new IP header */
	int    max_headroom;			/* The extra header space needed */
	int    gre_hlen;
A
Al Viro 已提交
679
	__be32 dst;
L
Linus Torvalds 已提交
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
	int    mtu;

	if (tunnel->recursion++) {
		tunnel->stat.collisions++;
		goto tx_error;
	}

	if (dev->hard_header) {
		gre_hlen = 0;
		tiph = (struct iphdr*)skb->data;
	} else {
		gre_hlen = tunnel->hlen;
		tiph = &tunnel->parms.iph;
	}

	if ((dst = tiph->daddr) == 0) {
		/* NBMA tunnel */

		if (skb->dst == NULL) {
			tunnel->stat.tx_fifo_errors++;
			goto tx_error;
		}

		if (skb->protocol == htons(ETH_P_IP)) {
			rt = (struct rtable*)skb->dst;
			if ((dst = rt->rt_gateway) == 0)
				goto tx_error_icmp;
		}
#ifdef CONFIG_IPV6
		else if (skb->protocol == htons(ETH_P_IPV6)) {
			struct in6_addr *addr6;
			int addr_type;
			struct neighbour *neigh = skb->dst->neighbour;

			if (neigh == NULL)
				goto tx_error;

			addr6 = (struct in6_addr*)&neigh->primary_key;
			addr_type = ipv6_addr_type(addr6);

			if (addr_type == IPV6_ADDR_ANY) {
721
				addr6 = &ipv6_hdr(skb)->daddr;
L
Linus Torvalds 已提交
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
				addr_type = ipv6_addr_type(addr6);
			}

			if ((addr_type & IPV6_ADDR_COMPATv4) == 0)
				goto tx_error_icmp;

			dst = addr6->s6_addr32[3];
		}
#endif
		else
			goto tx_error;
	}

	tos = tiph->tos;
	if (tos&1) {
		if (skb->protocol == htons(ETH_P_IP))
			tos = old_iph->tos;
		tos &= ~1;
	}

	{
		struct flowi fl = { .oif = tunnel->parms.link,
				    .nl_u = { .ip4_u =
					      { .daddr = dst,
						.saddr = tiph->saddr,
						.tos = RT_TOS(tos) } },
				    .proto = IPPROTO_GRE };
		if (ip_route_output_key(&rt, &fl)) {
			tunnel->stat.tx_carrier_errors++;
			goto tx_error;
		}
	}
	tdev = rt->u.dst.dev;

	if (tdev == dev) {
		ip_rt_put(rt);
		tunnel->stat.collisions++;
		goto tx_error;
	}

	df = tiph->frag_off;
	if (df)
		mtu = dst_mtu(&rt->u.dst) - tunnel->hlen;
	else
		mtu = skb->dst ? dst_mtu(skb->dst) : dev->mtu;

	if (skb->dst)
		skb->dst->ops->update_pmtu(skb->dst, mtu);

	if (skb->protocol == htons(ETH_P_IP)) {
		df |= (old_iph->frag_off&htons(IP_DF));

		if ((old_iph->frag_off&htons(IP_DF)) &&
		    mtu < ntohs(old_iph->tot_len)) {
			icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, htonl(mtu));
			ip_rt_put(rt);
			goto tx_error;
		}
	}
#ifdef CONFIG_IPV6
	else if (skb->protocol == htons(ETH_P_IPV6)) {
		struct rt6_info *rt6 = (struct rt6_info*)skb->dst;

		if (rt6 && mtu < dst_mtu(skb->dst) && mtu >= IPV6_MIN_MTU) {
			if ((tunnel->parms.iph.daddr && !MULTICAST(tunnel->parms.iph.daddr)) ||
			    rt6->rt6i_dst.plen == 128) {
				rt6->rt6i_flags |= RTF_MODIFIED;
				skb->dst->metrics[RTAX_MTU-1] = mtu;
			}
		}

		if (mtu >= IPV6_MIN_MTU && mtu < skb->len - tunnel->hlen + gre_hlen) {
			icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, dev);
			ip_rt_put(rt);
			goto tx_error;
		}
	}
#endif

	if (tunnel->err_count > 0) {
		if (jiffies - tunnel->err_time < IPTUNNEL_ERR_TIMEO) {
			tunnel->err_count--;

			dst_link_failure(skb);
		} else
			tunnel->err_count = 0;
	}

	max_headroom = LL_RESERVED_SPACE(tdev) + gre_hlen;

812 813
	if (skb_headroom(skb) < max_headroom || skb_shared(skb)||
	    (skb_cloned(skb) && !skb_clone_writable(skb, 0))) {
L
Linus Torvalds 已提交
814 815 816
		struct sk_buff *new_skb = skb_realloc_headroom(skb, max_headroom);
		if (!new_skb) {
			ip_rt_put(rt);
817
			stats->tx_dropped++;
L
Linus Torvalds 已提交
818 819 820 821 822 823 824 825
			dev_kfree_skb(skb);
			tunnel->recursion--;
			return 0;
		}
		if (skb->sk)
			skb_set_owner_w(new_skb, skb->sk);
		dev_kfree_skb(skb);
		skb = new_skb;
826
		old_iph = ip_hdr(skb);
L
Linus Torvalds 已提交
827 828
	}

829
	skb->transport_header = skb->network_header;
830 831
	skb_push(skb, gre_hlen);
	skb_reset_network_header(skb);
L
Linus Torvalds 已提交
832
	memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
833 834
	IPCB(skb)->flags &= ~(IPSKB_XFRM_TUNNEL_SIZE | IPSKB_XFRM_TRANSFORMED |
			      IPSKB_REROUTED);
L
Linus Torvalds 已提交
835 836 837 838 839 840 841
	dst_release(skb->dst);
	skb->dst = &rt->u.dst;

	/*
	 *	Push down and install the IPIP header.
	 */

842
	iph 			=	ip_hdr(skb);
L
Linus Torvalds 已提交
843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861
	iph->version		=	4;
	iph->ihl		=	sizeof(struct iphdr) >> 2;
	iph->frag_off		=	df;
	iph->protocol		=	IPPROTO_GRE;
	iph->tos		=	ipgre_ecn_encapsulate(tos, old_iph, skb);
	iph->daddr		=	rt->rt_dst;
	iph->saddr		=	rt->rt_src;

	if ((iph->ttl = tiph->ttl) == 0) {
		if (skb->protocol == htons(ETH_P_IP))
			iph->ttl = old_iph->ttl;
#ifdef CONFIG_IPV6
		else if (skb->protocol == htons(ETH_P_IPV6))
			iph->ttl = ((struct ipv6hdr*)old_iph)->hop_limit;
#endif
		else
			iph->ttl = dst_metric(&rt->u.dst, RTAX_HOPLIMIT);
	}

A
Al Viro 已提交
862 863
	((__be16*)(iph+1))[0] = tunnel->parms.o_flags;
	((__be16*)(iph+1))[1] = skb->protocol;
L
Linus Torvalds 已提交
864 865

	if (tunnel->parms.o_flags&(GRE_KEY|GRE_CSUM|GRE_SEQ)) {
A
Al Viro 已提交
866
		__be32 *ptr = (__be32*)(((u8*)iph) + tunnel->hlen - 4);
L
Linus Torvalds 已提交
867 868 869 870 871 872 873 874 875 876 877 878

		if (tunnel->parms.o_flags&GRE_SEQ) {
			++tunnel->o_seqno;
			*ptr = htonl(tunnel->o_seqno);
			ptr--;
		}
		if (tunnel->parms.o_flags&GRE_KEY) {
			*ptr = tunnel->parms.o_key;
			ptr--;
		}
		if (tunnel->parms.o_flags&GRE_CSUM) {
			*ptr = 0;
879
			*(__sum16*)ptr = ip_compute_csum((void*)(iph+1), skb->len - sizeof(struct iphdr));
L
Linus Torvalds 已提交
880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916
		}
	}

	nf_reset(skb);

	IPTUNNEL_XMIT();
	tunnel->recursion--;
	return 0;

tx_error_icmp:
	dst_link_failure(skb);

tx_error:
	stats->tx_errors++;
	dev_kfree_skb(skb);
	tunnel->recursion--;
	return 0;
}

static int
ipgre_tunnel_ioctl (struct net_device *dev, struct ifreq *ifr, int cmd)
{
	int err = 0;
	struct ip_tunnel_parm p;
	struct ip_tunnel *t;

	switch (cmd) {
	case SIOCGETTUNNEL:
		t = NULL;
		if (dev == ipgre_fb_tunnel_dev) {
			if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p))) {
				err = -EFAULT;
				break;
			}
			t = ipgre_tunnel_locate(&p, 0);
		}
		if (t == NULL)
917
			t = netdev_priv(dev);
L
Linus Torvalds 已提交
918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956
		memcpy(&p, &t->parms, sizeof(p));
		if (copy_to_user(ifr->ifr_ifru.ifru_data, &p, sizeof(p)))
			err = -EFAULT;
		break;

	case SIOCADDTUNNEL:
	case SIOCCHGTUNNEL:
		err = -EPERM;
		if (!capable(CAP_NET_ADMIN))
			goto done;

		err = -EFAULT;
		if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p)))
			goto done;

		err = -EINVAL;
		if (p.iph.version != 4 || p.iph.protocol != IPPROTO_GRE ||
		    p.iph.ihl != 5 || (p.iph.frag_off&htons(~IP_DF)) ||
		    ((p.i_flags|p.o_flags)&(GRE_VERSION|GRE_ROUTING)))
			goto done;
		if (p.iph.ttl)
			p.iph.frag_off |= htons(IP_DF);

		if (!(p.i_flags&GRE_KEY))
			p.i_key = 0;
		if (!(p.o_flags&GRE_KEY))
			p.o_key = 0;

		t = ipgre_tunnel_locate(&p, cmd == SIOCADDTUNNEL);

		if (dev != ipgre_fb_tunnel_dev && cmd == SIOCCHGTUNNEL) {
			if (t != NULL) {
				if (t->dev != dev) {
					err = -EEXIST;
					break;
				}
			} else {
				unsigned nflags=0;

957
				t = netdev_priv(dev);
L
Linus Torvalds 已提交
958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005

				if (MULTICAST(p.iph.daddr))
					nflags = IFF_BROADCAST;
				else if (p.iph.daddr)
					nflags = IFF_POINTOPOINT;

				if ((dev->flags^nflags)&(IFF_POINTOPOINT|IFF_BROADCAST)) {
					err = -EINVAL;
					break;
				}
				ipgre_tunnel_unlink(t);
				t->parms.iph.saddr = p.iph.saddr;
				t->parms.iph.daddr = p.iph.daddr;
				t->parms.i_key = p.i_key;
				t->parms.o_key = p.o_key;
				memcpy(dev->dev_addr, &p.iph.saddr, 4);
				memcpy(dev->broadcast, &p.iph.daddr, 4);
				ipgre_tunnel_link(t);
				netdev_state_change(dev);
			}
		}

		if (t) {
			err = 0;
			if (cmd == SIOCCHGTUNNEL) {
				t->parms.iph.ttl = p.iph.ttl;
				t->parms.iph.tos = p.iph.tos;
				t->parms.iph.frag_off = p.iph.frag_off;
			}
			if (copy_to_user(ifr->ifr_ifru.ifru_data, &t->parms, sizeof(p)))
				err = -EFAULT;
		} else
			err = (cmd == SIOCADDTUNNEL ? -ENOBUFS : -ENOENT);
		break;

	case SIOCDELTUNNEL:
		err = -EPERM;
		if (!capable(CAP_NET_ADMIN))
			goto done;

		if (dev == ipgre_fb_tunnel_dev) {
			err = -EFAULT;
			if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p)))
				goto done;
			err = -ENOENT;
			if ((t = ipgre_tunnel_locate(&p, 0)) == NULL)
				goto done;
			err = -EPERM;
1006
			if (t == netdev_priv(ipgre_fb_tunnel_dev))
L
Linus Torvalds 已提交
1007 1008 1009
				goto done;
			dev = t->dev;
		}
1010 1011
		unregister_netdevice(dev);
		err = 0;
L
Linus Torvalds 已提交
1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023
		break;

	default:
		err = -EINVAL;
	}

done:
	return err;
}

static struct net_device_stats *ipgre_tunnel_get_stats(struct net_device *dev)
{
1024
	return &(((struct ip_tunnel*)netdev_priv(dev))->stat);
L
Linus Torvalds 已提交
1025 1026 1027 1028
}

static int ipgre_tunnel_change_mtu(struct net_device *dev, int new_mtu)
{
1029
	struct ip_tunnel *tunnel = netdev_priv(dev);
L
Linus Torvalds 已提交
1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045
	if (new_mtu < 68 || new_mtu > 0xFFF8 - tunnel->hlen)
		return -EINVAL;
	dev->mtu = new_mtu;
	return 0;
}

#ifdef CONFIG_NET_IPGRE_BROADCAST
/* Nice toy. Unfortunately, useless in real life :-)
   It allows to construct virtual multiprotocol broadcast "LAN"
   over the Internet, provided multicast routing is tuned.


   I have no idea was this bicycle invented before me,
   so that I had to set ARPHRD_IPGRE to a random value.
   I have an impression, that Cisco could make something similar,
   but this feature is apparently missing in IOS<=11.2(8).
1046

L
Linus Torvalds 已提交
1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068
   I set up 10.66.66/24 and fec0:6666:6666::0/96 as virtual networks
   with broadcast 224.66.66.66. If you have access to mbone, play with me :-)

   ping -t 255 224.66.66.66

   If nobody answers, mbone does not work.

   ip tunnel add Universe mode gre remote 224.66.66.66 local <Your_real_addr> ttl 255
   ip addr add 10.66.66.<somewhat>/24 dev Universe
   ifconfig Universe up
   ifconfig Universe add fe80::<Your_real_addr>/10
   ifconfig Universe add fec0:6666:6666::<Your_real_addr>/96
   ftp 10.66.66.66
   ...
   ftp fec0:6666:6666::193.233.7.65
   ...

 */

static int ipgre_header(struct sk_buff *skb, struct net_device *dev, unsigned short type,
			void *daddr, void *saddr, unsigned len)
{
1069
	struct ip_tunnel *t = netdev_priv(dev);
L
Linus Torvalds 已提交
1070
	struct iphdr *iph = (struct iphdr *)skb_push(skb, t->hlen);
A
Al Viro 已提交
1071
	__be16 *p = (__be16*)(iph+1);
L
Linus Torvalds 已提交
1072 1073 1074 1075 1076 1077

	memcpy(iph, &t->parms.iph, sizeof(struct iphdr));
	p[0]		= t->parms.o_flags;
	p[1]		= htons(type);

	/*
1078
	 *	Set the source hardware address.
L
Linus Torvalds 已提交
1079
	 */
1080

L
Linus Torvalds 已提交
1081 1082 1083 1084 1085 1086 1087 1088 1089
	if (saddr)
		memcpy(&iph->saddr, saddr, 4);

	if (daddr) {
		memcpy(&iph->daddr, daddr, 4);
		return t->hlen;
	}
	if (iph->daddr && !MULTICAST(iph->daddr))
		return t->hlen;
1090

L
Linus Torvalds 已提交
1091 1092 1093 1094 1095
	return -t->hlen;
}

static int ipgre_open(struct net_device *dev)
{
1096
	struct ip_tunnel *t = netdev_priv(dev);
L
Linus Torvalds 已提交
1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109

	if (MULTICAST(t->parms.iph.daddr)) {
		struct flowi fl = { .oif = t->parms.link,
				    .nl_u = { .ip4_u =
					      { .daddr = t->parms.iph.daddr,
						.saddr = t->parms.iph.saddr,
						.tos = RT_TOS(t->parms.iph.tos) } },
				    .proto = IPPROTO_GRE };
		struct rtable *rt;
		if (ip_route_output_key(&rt, &fl))
			return -EADDRNOTAVAIL;
		dev = rt->u.dst.dev;
		ip_rt_put(rt);
1110
		if (__in_dev_get_rtnl(dev) == NULL)
L
Linus Torvalds 已提交
1111 1112
			return -EADDRNOTAVAIL;
		t->mlink = dev->ifindex;
1113
		ip_mc_inc_group(__in_dev_get_rtnl(dev), t->parms.iph.daddr);
L
Linus Torvalds 已提交
1114 1115 1116 1117 1118 1119
	}
	return 0;
}

static int ipgre_close(struct net_device *dev)
{
1120
	struct ip_tunnel *t = netdev_priv(dev);
L
Linus Torvalds 已提交
1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143
	if (MULTICAST(t->parms.iph.daddr) && t->mlink) {
		struct in_device *in_dev = inetdev_by_index(t->mlink);
		if (in_dev) {
			ip_mc_dec_group(in_dev, t->parms.iph.daddr);
			in_dev_put(in_dev);
		}
	}
	return 0;
}

#endif

static void ipgre_tunnel_setup(struct net_device *dev)
{
	dev->uninit		= ipgre_tunnel_uninit;
	dev->destructor 	= free_netdev;
	dev->hard_start_xmit	= ipgre_tunnel_xmit;
	dev->get_stats		= ipgre_tunnel_get_stats;
	dev->do_ioctl		= ipgre_tunnel_ioctl;
	dev->change_mtu		= ipgre_tunnel_change_mtu;

	dev->type		= ARPHRD_IPGRE;
	dev->hard_header_len 	= LL_MAX_HEADER + sizeof(struct iphdr) + 4;
1144
	dev->mtu		= ETH_DATA_LEN - sizeof(struct iphdr) - 4;
L
Linus Torvalds 已提交
1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155
	dev->flags		= IFF_NOARP;
	dev->iflink		= 0;
	dev->addr_len		= 4;
}

static int ipgre_tunnel_init(struct net_device *dev)
{
	struct net_device *tdev = NULL;
	struct ip_tunnel *tunnel;
	struct iphdr *iph;
	int hlen = LL_MAX_HEADER;
1156
	int mtu = ETH_DATA_LEN;
L
Linus Torvalds 已提交
1157 1158
	int addend = sizeof(struct iphdr) + 4;

1159
	tunnel = netdev_priv(dev);
L
Linus Torvalds 已提交
1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197
	iph = &tunnel->parms.iph;

	tunnel->dev = dev;
	strcpy(tunnel->parms.name, dev->name);

	memcpy(dev->dev_addr, &tunnel->parms.iph.saddr, 4);
	memcpy(dev->broadcast, &tunnel->parms.iph.daddr, 4);

	/* Guess output device to choose reasonable mtu and hard_header_len */

	if (iph->daddr) {
		struct flowi fl = { .oif = tunnel->parms.link,
				    .nl_u = { .ip4_u =
					      { .daddr = iph->daddr,
						.saddr = iph->saddr,
						.tos = RT_TOS(iph->tos) } },
				    .proto = IPPROTO_GRE };
		struct rtable *rt;
		if (!ip_route_output_key(&rt, &fl)) {
			tdev = rt->u.dst.dev;
			ip_rt_put(rt);
		}

		dev->flags |= IFF_POINTOPOINT;

#ifdef CONFIG_NET_IPGRE_BROADCAST
		if (MULTICAST(iph->daddr)) {
			if (!iph->saddr)
				return -EINVAL;
			dev->flags = IFF_BROADCAST;
			dev->hard_header = ipgre_header;
			dev->open = ipgre_open;
			dev->stop = ipgre_close;
		}
#endif
	}

	if (!tdev && tunnel->parms.link)
1198
		tdev = __dev_get_by_index(&init_net, tunnel->parms.link);
L
Linus Torvalds 已提交
1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220

	if (tdev) {
		hlen = tdev->hard_header_len;
		mtu = tdev->mtu;
	}
	dev->iflink = tunnel->parms.link;

	/* Precalculate GRE options length */
	if (tunnel->parms.o_flags&(GRE_CSUM|GRE_KEY|GRE_SEQ)) {
		if (tunnel->parms.o_flags&GRE_CSUM)
			addend += 4;
		if (tunnel->parms.o_flags&GRE_KEY)
			addend += 4;
		if (tunnel->parms.o_flags&GRE_SEQ)
			addend += 4;
	}
	dev->hard_header_len = hlen + addend;
	dev->mtu = mtu - addend;
	tunnel->hlen = addend;
	return 0;
}

A
Adrian Bunk 已提交
1221
static int __init ipgre_fb_tunnel_init(struct net_device *dev)
L
Linus Torvalds 已提交
1222
{
1223
	struct ip_tunnel *tunnel = netdev_priv(dev);
L
Linus Torvalds 已提交
1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280
	struct iphdr *iph = &tunnel->parms.iph;

	tunnel->dev = dev;
	strcpy(tunnel->parms.name, dev->name);

	iph->version		= 4;
	iph->protocol		= IPPROTO_GRE;
	iph->ihl		= 5;
	tunnel->hlen		= sizeof(struct iphdr) + 4;

	dev_hold(dev);
	tunnels_wc[0]		= tunnel;
	return 0;
}


static struct net_protocol ipgre_protocol = {
	.handler	=	ipgre_rcv,
	.err_handler	=	ipgre_err,
};


/*
 *	And now the modules code and kernel interface.
 */

static int __init ipgre_init(void)
{
	int err;

	printk(KERN_INFO "GRE over IPv4 tunneling driver\n");

	if (inet_add_protocol(&ipgre_protocol, IPPROTO_GRE) < 0) {
		printk(KERN_INFO "ipgre init: can't add protocol\n");
		return -EAGAIN;
	}

	ipgre_fb_tunnel_dev = alloc_netdev(sizeof(struct ip_tunnel), "gre0",
					   ipgre_tunnel_setup);
	if (!ipgre_fb_tunnel_dev) {
		err = -ENOMEM;
		goto err1;
	}

	ipgre_fb_tunnel_dev->init = ipgre_fb_tunnel_init;

	if ((err = register_netdev(ipgre_fb_tunnel_dev)))
		goto err2;
out:
	return err;
err2:
	free_netdev(ipgre_fb_tunnel_dev);
err1:
	inet_del_protocol(&ipgre_protocol, IPPROTO_GRE);
	goto out;
}

1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295
static void __exit ipgre_destroy_tunnels(void)
{
	int prio;

	for (prio = 0; prio < 4; prio++) {
		int h;
		for (h = 0; h < HASH_SIZE; h++) {
			struct ip_tunnel *t;
			while ((t = tunnels[prio][h]) != NULL)
				unregister_netdevice(t->dev);
		}
	}
}

static void __exit ipgre_fini(void)
L
Linus Torvalds 已提交
1296 1297 1298 1299
{
	if (inet_del_protocol(&ipgre_protocol, IPPROTO_GRE) < 0)
		printk(KERN_INFO "ipgre close: can't remove protocol\n");

1300 1301 1302
	rtnl_lock();
	ipgre_destroy_tunnels();
	rtnl_unlock();
L
Linus Torvalds 已提交
1303 1304 1305 1306 1307
}

module_init(ipgre_init);
module_exit(ipgre_fini);
MODULE_LICENSE("GPL");