ipip.c 21.9 KB
Newer Older
L
Linus Torvalds 已提交
1
/*
2
 *	Linux NET3:	IP/IP protocol decoder.
L
Linus Torvalds 已提交
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
 *
 *	Version: $Id: ipip.c,v 1.50 2001/10/02 02:22:36 davem Exp $
 *
 *	Authors:
 *		Sam Lantinga (slouken@cs.ucdavis.edu)  02/01/95
 *
 *	Fixes:
 *		Alan Cox	:	Merged and made usable non modular (its so tiny its silly as
 *					a module taking up 2 pages).
 *		Alan Cox	: 	Fixed bug with 1.3.18 and IPIP not working (now needs to set skb->h.iph)
 *					to keep ip_forward happy.
 *		Alan Cox	:	More fixes for 1.3.21, and firewall fix. Maybe this will work soon 8).
 *		Kai Schulte	:	Fixed #defines for IP_FIREWALL->FIREWALL
 *              David Woodhouse :       Perform some basic ICMP handling.
 *                                      IPIP Routing without decapsulation.
 *              Carlos Picoto   :       GRE over IP support
 *		Alexey Kuznetsov:	Reworked. Really, now it is truncated version of ipv4/ip_gre.c.
 *					I do not want to merge them together.
 *
 *	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.
 *
 */

/* tunnel.c: an IP tunnel driver

	The purpose of this driver is to provide an IP tunnel through
	which you can tunnel network traffic transparently across subnets.

	This was written by looking at Nick Holloway's dummy driver
	Thanks for the great code!

		-Sam Lantinga	(slouken@cs.ucdavis.edu)  02/01/95
38

L
Linus Torvalds 已提交
39 40 41 42 43 44
	Minor tweaks:
		Cleaned up the code a little and added some pre-1.3.0 tweaks.
		dev->hard_header/hard_header_len changed to use no headers.
		Comments/bracketing tweaked.
		Made the tunnels use dev->name not tunnel: when error reporting.
		Added tx_dropped stat
45

L
Linus Torvalds 已提交
46 47 48 49 50 51 52 53 54
		-Alan Cox	(Alan.Cox@linux.org) 21 March 95

	Reworked:
		Changed to tunnel to destination gateway in addition to the
			tunnel's pointopoint address
		Almost completely rewritten
		Note:  There is currently no firewall or ICMP handling done.

		-Sam Lantinga	(slouken@cs.ucdavis.edu) 02/13/96
55

L
Linus Torvalds 已提交
56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77
*/

/* Things I wish I had known when writing the tunnel driver:

	When the tunnel_xmit() function is called, the skb contains the
	packet to be sent (plus a great deal of extra info), and dev
	contains the tunnel device that _we_ are.

	When we are passed a packet, we are expected to fill in the
	source address with our source IP address.

	What is the proper way to allocate, copy and free a buffer?
	After you allocate it, it is a "0 length" chunk of memory
	starting at zero.  If you want to add headers to the buffer
	later, you'll have to call "skb_reserve(skb, amount)" with
	the amount of memory you want reserved.  Then, you call
	"skb_put(skb, amount)" with the amount of space you want in
	the buffer.  skb_put() returns a pointer to the top (#0) of
	that buffer.  skb->len is set to the amount of space you have
	"allocated" with skb_put().  You can then write up to skb->len
	bytes to that buffer.  If you need more, you can call skb_put()
	again with the additional amount of space you need.  You can
78
	find out how much more space you can allocate by calling
L
Linus Torvalds 已提交
79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94
	"skb_tailroom(skb)".
	Now, to add header space, call "skb_push(skb, header_len)".
	This creates space at the beginning of the buffer and returns
	a pointer to this new space.  If later you need to strip a
	header from a buffer, call "skb_pull(skb, header_len)".
	skb_headroom() will return how much space is left at the top
	of the buffer (before the main data).  Remember, this headroom
	space must be reserved before the skb_put() function is called.
	*/

/*
   This version of net/ipv4/ipip.c is cloned of net/ipv4/ip_gre.c

   For comments look at net/ipv4/ip_gre.c --ANK
 */

95

96
#include <linux/capability.h>
L
Linus Torvalds 已提交
97 98 99 100 101 102 103 104 105 106 107 108 109
#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/netfilter_ipv4.h>
110
#include <linux/if_ether.h>
L
Linus Torvalds 已提交
111 112 113 114 115 116 117 118 119

#include <net/sock.h>
#include <net/ip.h>
#include <net/icmp.h>
#include <net/ipip.h>
#include <net/inet_ecn.h>
#include <net/xfrm.h>

#define HASH_SIZE  16
A
Al Viro 已提交
120
#define HASH(addr) (((__force u32)addr^((__force u32)addr>>4))&0xF)
L
Linus Torvalds 已提交
121 122 123 124 125 126 127 128 129 130 131 132 133 134 135

static int ipip_fb_tunnel_init(struct net_device *dev);
static int ipip_tunnel_init(struct net_device *dev);
static void ipip_tunnel_setup(struct net_device *dev);

static struct net_device *ipip_fb_tunnel_dev;

static struct ip_tunnel *tunnels_r_l[HASH_SIZE];
static struct ip_tunnel *tunnels_r[HASH_SIZE];
static struct ip_tunnel *tunnels_l[HASH_SIZE];
static struct ip_tunnel *tunnels_wc[1];
static struct ip_tunnel **tunnels[4] = { tunnels_wc, tunnels_l, tunnels_r, tunnels_r_l };

static DEFINE_RWLOCK(ipip_lock);

A
Al Viro 已提交
136
static struct ip_tunnel * ipip_tunnel_lookup(__be32 remote, __be32 local)
L
Linus Torvalds 已提交
137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161
{
	unsigned h0 = HASH(remote);
	unsigned h1 = HASH(local);
	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 && (t->dev->flags&IFF_UP))
			return t;
	}
	for (t = tunnels_r[h0]; t; t = t->next) {
		if (remote == t->parms.iph.daddr && (t->dev->flags&IFF_UP))
			return t;
	}
	for (t = tunnels_l[h1]; t; t = t->next) {
		if (local == t->parms.iph.saddr && (t->dev->flags&IFF_UP))
			return t;
	}
	if ((t = tunnels_wc[0]) != NULL && (t->dev->flags&IFF_UP))
		return t;
	return NULL;
}

static struct ip_tunnel **ipip_bucket(struct ip_tunnel *t)
{
A
Al Viro 已提交
162 163
	__be32 remote = t->parms.iph.daddr;
	__be32 local = t->parms.iph.saddr;
L
Linus Torvalds 已提交
164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204
	unsigned h = 0;
	int prio = 0;

	if (remote) {
		prio |= 2;
		h ^= HASH(remote);
	}
	if (local) {
		prio |= 1;
		h ^= HASH(local);
	}
	return &tunnels[prio][h];
}


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

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

static void ipip_tunnel_link(struct ip_tunnel *t)
{
	struct ip_tunnel **tp = ipip_bucket(t);

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

static struct ip_tunnel * ipip_tunnel_locate(struct ip_tunnel_parm *parms, int create)
{
A
Al Viro 已提交
205 206
	__be32 remote = parms->iph.daddr;
	__be32 local = parms->iph.saddr;
L
Linus Torvalds 已提交
207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244
	struct ip_tunnel *t, **tp, *nt;
	struct net_device *dev;
	unsigned h = 0;
	int prio = 0;
	char name[IFNAMSIZ];

	if (remote) {
		prio |= 2;
		h ^= HASH(remote);
	}
	if (local) {
		prio |= 1;
		h ^= HASH(local);
	}
	for (tp = &tunnels[prio][h]; (t = *tp) != NULL; tp = &t->next) {
		if (local == t->parms.iph.saddr && remote == t->parms.iph.daddr)
			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, "tunl%d", i);
			if (__dev_get_by_name(name) == NULL)
				break;
		}
		if (i==100)
			goto failed;
	}

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

245
	nt = netdev_priv(dev);
L
Linus Torvalds 已提交
246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269
	SET_MODULE_OWNER(dev);
	dev->init = ipip_tunnel_init;
	nt->parms = *parms;

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

	dev_hold(dev);
	ipip_tunnel_link(nt);
	return nt;

failed:
	return NULL;
}

static void ipip_tunnel_uninit(struct net_device *dev)
{
	if (dev == ipip_fb_tunnel_dev) {
		write_lock_bh(&ipip_lock);
		tunnels_wc[0] = NULL;
		write_unlock_bh(&ipip_lock);
	} else
270
		ipip_tunnel_unlink(netdev_priv(dev));
L
Linus Torvalds 已提交
271 272 273
	dev_put(dev);
}

H
Herbert Xu 已提交
274
static int ipip_err(struct sk_buff *skb, u32 info)
L
Linus Torvalds 已提交
275 276 277 278 279 280 281 282 283 284 285
{
#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.
 */
	struct iphdr *iph = (struct iphdr*)skb->data;
	int type = skb->h.icmph->type;
	int code = skb->h.icmph->code;
	struct ip_tunnel *t;
H
Herbert Xu 已提交
286
	int err;
L
Linus Torvalds 已提交
287 288 289 290

	switch (type) {
	default:
	case ICMP_PARAMETERPROB:
H
Herbert Xu 已提交
291
		return 0;
L
Linus Torvalds 已提交
292 293 294 295 296 297

	case ICMP_DEST_UNREACH:
		switch (code) {
		case ICMP_SR_FAILED:
		case ICMP_PORT_UNREACH:
			/* Impossible event. */
H
Herbert Xu 已提交
298
			return 0;
L
Linus Torvalds 已提交
299 300
		case ICMP_FRAG_NEEDED:
			/* Soft state for pmtu is maintained by IP core. */
H
Herbert Xu 已提交
301
			return 0;
L
Linus Torvalds 已提交
302 303 304 305 306 307 308 309 310 311
		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)
H
Herbert Xu 已提交
312
			return 0;
L
Linus Torvalds 已提交
313 314 315
		break;
	}

H
Herbert Xu 已提交
316 317
	err = -ENOENT;

L
Linus Torvalds 已提交
318 319 320 321
	read_lock(&ipip_lock);
	t = ipip_tunnel_lookup(iph->daddr, iph->saddr);
	if (t == NULL || t->parms.iph.daddr == 0)
		goto out;
H
Herbert Xu 已提交
322 323

	err = 0;
L
Linus Torvalds 已提交
324 325 326 327 328 329 330 331 332 333
	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(&ipip_lock);
H
Herbert Xu 已提交
334
	return err;
L
Linus Torvalds 已提交
335 336 337 338 339 340 341 342
#else
	struct iphdr *iph = (struct iphdr*)dp;
	int hlen = iph->ihl<<2;
	struct iphdr *eiph;
	int type = skb->h.icmph->type;
	int code = skb->h.icmph->code;
	int rel_type = 0;
	int rel_code = 0;
343 344
	__be32 rel_info = 0;
	__u32 n = 0;
L
Linus Torvalds 已提交
345 346 347 348 349
	struct sk_buff *skb2;
	struct flowi fl;
	struct rtable *rt;

	if (len < hlen + sizeof(struct iphdr))
H
Herbert Xu 已提交
350
		return 0;
L
Linus Torvalds 已提交
351 352 353 354
	eiph = (struct iphdr*)(dp + hlen);

	switch (type) {
	default:
H
Herbert Xu 已提交
355
		return 0;
L
Linus Torvalds 已提交
356
	case ICMP_PARAMETERPROB:
357 358
		n = ntohl(skb->h.icmph->un.gateway) >> 24;
		if (n < hlen)
H
Herbert Xu 已提交
359
			return 0;
L
Linus Torvalds 已提交
360 361 362 363 364

		/* So... This guy found something strange INSIDE encapsulated
		   packet. Well, he is fool, but what can we do ?
		 */
		rel_type = ICMP_PARAMETERPROB;
365
		rel_info = htonl((n - hlen) << 24);
L
Linus Torvalds 已提交
366 367 368 369 370 371 372
		break;

	case ICMP_DEST_UNREACH:
		switch (code) {
		case ICMP_SR_FAILED:
		case ICMP_PORT_UNREACH:
			/* Impossible event. */
H
Herbert Xu 已提交
373
			return 0;
L
Linus Torvalds 已提交
374 375
		case ICMP_FRAG_NEEDED:
			/* And it is the only really necessary thing :-) */
376 377
			n = ntohs(skb->h.icmph->un.frag.mtu);
			if (n < hlen+68)
H
Herbert Xu 已提交
378
				return 0;
379
			n -= hlen;
L
Linus Torvalds 已提交
380
			/* BSD 4.2 MORE DOES NOT EXIST IN NATURE. */
381
			if (n > ntohs(eiph->tot_len))
H
Herbert Xu 已提交
382
				return 0;
383
			rel_info = htonl(n);
L
Linus Torvalds 已提交
384 385 386 387 388 389 390 391 392 393 394 395 396
			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)
H
Herbert Xu 已提交
397
			return 0;
L
Linus Torvalds 已提交
398 399 400 401 402 403
		break;
	}

	/* Prepare fake skb to feed it to icmp_send */
	skb2 = skb_clone(skb, GFP_ATOMIC);
	if (skb2 == NULL)
H
Herbert Xu 已提交
404
		return 0;
L
Linus Torvalds 已提交
405 406 407
	dst_release(skb2->dst);
	skb2->dst = NULL;
	skb_pull(skb2, skb->data - (u8*)eiph);
408
	skb_reset_network_header(skb2);
L
Linus Torvalds 已提交
409 410 411 412 413 414 415 416

	/* Try to guess incoming interface */
	memset(&fl, 0, sizeof(fl));
	fl.fl4_daddr = eiph->saddr;
	fl.fl4_tos = RT_TOS(eiph->tos);
	fl.proto = IPPROTO_IPIP;
	if (ip_route_output_key(&rt, &key)) {
		kfree_skb(skb2);
H
Herbert Xu 已提交
417
		return 0;
L
Linus Torvalds 已提交
418 419 420 421 422 423 424 425 426 427 428 429 430 431
	}
	skb2->dev = rt->u.dst.dev;

	/* route "incoming" packet */
	if (rt->rt_flags&RTCF_LOCAL) {
		ip_rt_put(rt);
		rt = NULL;
		fl.fl4_daddr = 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_TUNNEL) {
			ip_rt_put(rt);
			kfree_skb(skb2);
H
Herbert Xu 已提交
432
			return 0;
L
Linus Torvalds 已提交
433 434 435 436 437 438
		}
	} else {
		ip_rt_put(rt);
		if (ip_route_input(skb2, eiph->daddr, eiph->saddr, eiph->tos, skb2->dev) ||
		    skb2->dst->dev->type != ARPHRD_TUNNEL) {
			kfree_skb(skb2);
H
Herbert Xu 已提交
439
			return 0;
L
Linus Torvalds 已提交
440 441 442 443 444
		}
	}

	/* change mtu on this route */
	if (type == ICMP_DEST_UNREACH && code == ICMP_FRAG_NEEDED) {
445
		if (n > dst_mtu(skb2->dst)) {
L
Linus Torvalds 已提交
446
			kfree_skb(skb2);
H
Herbert Xu 已提交
447
			return 0;
L
Linus Torvalds 已提交
448
		}
449
		skb2->dst->ops->update_pmtu(skb2->dst, n);
L
Linus Torvalds 已提交
450
	} else if (type == ICMP_TIME_EXCEEDED) {
451
		struct ip_tunnel *t = netdev_priv(skb2->dev);
L
Linus Torvalds 已提交
452 453 454 455 456 457 458 459
		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);
H
Herbert Xu 已提交
460
	return 0;
L
Linus Torvalds 已提交
461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489
#endif
}

static inline void ipip_ecn_decapsulate(struct iphdr *outer_iph, struct sk_buff *skb)
{
	struct iphdr *inner_iph = skb->nh.iph;

	if (INET_ECN_is_ce(outer_iph->tos))
		IP_ECN_set_ce(inner_iph);
}

static int ipip_rcv(struct sk_buff *skb)
{
	struct iphdr *iph;
	struct ip_tunnel *tunnel;

	iph = skb->nh.iph;

	read_lock(&ipip_lock);
	if ((tunnel = ipip_tunnel_lookup(iph->saddr, iph->daddr)) != NULL) {
		if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) {
			read_unlock(&ipip_lock);
			kfree_skb(skb);
			return 0;
		}

		secpath_reset(skb);

		skb->mac.raw = skb->nh.raw;
490
		skb_reset_network_header(skb);
L
Linus Torvalds 已提交
491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516
		skb->protocol = htons(ETH_P_IP);
		skb->pkt_type = PACKET_HOST;

		tunnel->stat.rx_packets++;
		tunnel->stat.rx_bytes += skb->len;
		skb->dev = tunnel->dev;
		dst_release(skb->dst);
		skb->dst = NULL;
		nf_reset(skb);
		ipip_ecn_decapsulate(iph, skb);
		netif_rx(skb);
		read_unlock(&ipip_lock);
		return 0;
	}
	read_unlock(&ipip_lock);

	return -1;
}

/*
 *	This function assumes it is being called from dev_queue_xmit()
 *	and that skb is filled properly by that function.
 */

static int ipip_tunnel_xmit(struct sk_buff *skb, struct net_device *dev)
{
517
	struct ip_tunnel *tunnel = netdev_priv(dev);
L
Linus Torvalds 已提交
518 519 520
	struct net_device_stats *stats = &tunnel->stat;
	struct iphdr  *tiph = &tunnel->parms.iph;
	u8     tos = tunnel->parms.iph.tos;
A
Al Viro 已提交
521
	__be16 df = tiph->frag_off;
L
Linus Torvalds 已提交
522 523 524 525 526
	struct rtable *rt;     			/* Route to the other host */
	struct net_device *tdev;			/* Device to other host */
	struct iphdr  *old_iph = skb->nh.iph;
	struct iphdr  *iph;			/* Our new IP header */
	int    max_headroom;			/* The extra header space needed */
A
Al Viro 已提交
527
	__be32 dst = tiph->daddr;
L
Linus Torvalds 已提交
528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608
	int    mtu;

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

	if (skb->protocol != htons(ETH_P_IP))
		goto tx_error;

	if (tos&1)
		tos = old_iph->tos;

	if (!dst) {
		/* NBMA tunnel */
		if ((rt = (struct rtable*)skb->dst) == NULL) {
			tunnel->stat.tx_fifo_errors++;
			goto tx_error;
		}
		if ((dst = rt->rt_gateway) == 0)
			goto tx_error_icmp;
	}

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

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

	if (tiph->frag_off)
		mtu = dst_mtu(&rt->u.dst) - sizeof(struct iphdr);
	else
		mtu = skb->dst ? dst_mtu(skb->dst) : dev->mtu;

	if (mtu < 68) {
		tunnel->stat.collisions++;
		ip_rt_put(rt);
		goto tx_error;
	}
	if (skb->dst)
		skb->dst->ops->update_pmtu(skb->dst, mtu);

	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;
	}

	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;
	}

	/*
	 * Okay, now see if we can stuff it in the buffer as-is.
	 */
	max_headroom = (LL_RESERVED_SPACE(tdev)+sizeof(struct iphdr));

	if (skb_headroom(skb) < max_headroom || skb_cloned(skb) || skb_shared(skb)) {
		struct sk_buff *new_skb = skb_realloc_headroom(skb, max_headroom);
		if (!new_skb) {
			ip_rt_put(rt);
609
			stats->tx_dropped++;
L
Linus Torvalds 已提交
610 611 612 613 614 615 616 617 618 619 620 621
			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;
		old_iph = skb->nh.iph;
	}

	skb->h.raw = skb->nh.raw;
622 623
	skb_push(skb, sizeof(struct iphdr));
	skb_reset_network_header(skb);
L
Linus Torvalds 已提交
624
	memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
625 626
	IPCB(skb)->flags &= ~(IPSKB_XFRM_TUNNEL_SIZE | IPSKB_XFRM_TRANSFORMED |
			      IPSKB_REROUTED);
L
Linus Torvalds 已提交
627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678
	dst_release(skb->dst);
	skb->dst = &rt->u.dst;

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

	iph 			=	skb->nh.iph;
	iph->version		=	4;
	iph->ihl		=	sizeof(struct iphdr)>>2;
	iph->frag_off		=	df;
	iph->protocol		=	IPPROTO_IPIP;
	iph->tos		=	INET_ECN_encapsulate(tos, old_iph->tos);
	iph->daddr		=	rt->rt_dst;
	iph->saddr		=	rt->rt_src;

	if ((iph->ttl = tiph->ttl) == 0)
		iph->ttl	=	old_iph->ttl;

	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
ipip_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 == ipip_fb_tunnel_dev) {
			if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p))) {
				err = -EFAULT;
				break;
			}
			t = ipip_tunnel_locate(&p, 0);
		}
		if (t == NULL)
679
			t = netdev_priv(dev);
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
		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_IPIP ||
		    p.iph.ihl != 5 || (p.iph.frag_off&htons(~IP_DF)))
			goto done;
		if (p.iph.ttl)
			p.iph.frag_off |= htons(IP_DF);

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

		if (dev != ipip_fb_tunnel_dev && cmd == SIOCCHGTUNNEL) {
			if (t != NULL) {
				if (t->dev != dev) {
					err = -EEXIST;
					break;
				}
			} else {
				if (((dev->flags&IFF_POINTOPOINT) && !p.iph.daddr) ||
				    (!(dev->flags&IFF_POINTOPOINT) && p.iph.daddr)) {
					err = -EINVAL;
					break;
				}
716
				t = netdev_priv(dev);
L
Linus Torvalds 已提交
717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756
				ipip_tunnel_unlink(t);
				t->parms.iph.saddr = p.iph.saddr;
				t->parms.iph.daddr = p.iph.daddr;
				memcpy(dev->dev_addr, &p.iph.saddr, 4);
				memcpy(dev->broadcast, &p.iph.daddr, 4);
				ipip_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 == ipip_fb_tunnel_dev) {
			err = -EFAULT;
			if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p)))
				goto done;
			err = -ENOENT;
			if ((t = ipip_tunnel_locate(&p, 0)) == NULL)
				goto done;
			err = -EPERM;
			if (t->dev == ipip_fb_tunnel_dev)
				goto done;
			dev = t->dev;
		}
757 758
		unregister_netdevice(dev);
		err = 0;
L
Linus Torvalds 已提交
759 760 761 762 763 764 765 766 767 768 769 770
		break;

	default:
		err = -EINVAL;
	}

done:
	return err;
}

static struct net_device_stats *ipip_tunnel_get_stats(struct net_device *dev)
{
771
	return &(((struct ip_tunnel*)netdev_priv(dev))->stat);
L
Linus Torvalds 已提交
772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793
}

static int ipip_tunnel_change_mtu(struct net_device *dev, int new_mtu)
{
	if (new_mtu < 68 || new_mtu > 0xFFF8 - sizeof(struct iphdr))
		return -EINVAL;
	dev->mtu = new_mtu;
	return 0;
}

static void ipip_tunnel_setup(struct net_device *dev)
{
	SET_MODULE_OWNER(dev);
	dev->uninit		= ipip_tunnel_uninit;
	dev->hard_start_xmit	= ipip_tunnel_xmit;
	dev->get_stats		= ipip_tunnel_get_stats;
	dev->do_ioctl		= ipip_tunnel_ioctl;
	dev->change_mtu		= ipip_tunnel_change_mtu;
	dev->destructor		= free_netdev;

	dev->type		= ARPHRD_TUNNEL;
	dev->hard_header_len 	= LL_MAX_HEADER + sizeof(struct iphdr);
794
	dev->mtu		= ETH_DATA_LEN - sizeof(struct iphdr);
L
Linus Torvalds 已提交
795 796 797 798 799 800 801 802 803 804 805
	dev->flags		= IFF_NOARP;
	dev->iflink		= 0;
	dev->addr_len		= 4;
}

static int ipip_tunnel_init(struct net_device *dev)
{
	struct net_device *tdev = NULL;
	struct ip_tunnel *tunnel;
	struct iphdr *iph;

806
	tunnel = netdev_priv(dev);
L
Linus Torvalds 已提交
807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843
	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);

	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_IPIP };
		struct rtable *rt;
		if (!ip_route_output_key(&rt, &fl)) {
			tdev = rt->u.dst.dev;
			ip_rt_put(rt);
		}
		dev->flags |= IFF_POINTOPOINT;
	}

	if (!tdev && tunnel->parms.link)
		tdev = __dev_get_by_index(tunnel->parms.link);

	if (tdev) {
		dev->hard_header_len = tdev->hard_header_len + sizeof(struct iphdr);
		dev->mtu = tdev->mtu - sizeof(struct iphdr);
	}
	dev->iflink = tunnel->parms.link;

	return 0;
}

static int __init ipip_fb_tunnel_init(struct net_device *dev)
{
844
	struct ip_tunnel *tunnel = netdev_priv(dev);
L
Linus Torvalds 已提交
845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861
	struct iphdr *iph = &tunnel->parms.iph;

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

	iph->version		= 4;
	iph->protocol		= IPPROTO_IPIP;
	iph->ihl		= 5;

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

static struct xfrm_tunnel ipip_handler = {
	.handler	=	ipip_rcv,
	.err_handler	=	ipip_err,
H
Herbert Xu 已提交
862
	.priority	=	1,
L
Linus Torvalds 已提交
863 864 865 866 867 868 869 870 871 872 873
};

static char banner[] __initdata =
	KERN_INFO "IPv4 over IPv4 tunneling driver\n";

static int __init ipip_init(void)
{
	int err;

	printk(banner);

874
	if (xfrm4_tunnel_register(&ipip_handler, AF_INET)) {
L
Linus Torvalds 已提交
875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895
		printk(KERN_INFO "ipip init: can't register tunnel\n");
		return -EAGAIN;
	}

	ipip_fb_tunnel_dev = alloc_netdev(sizeof(struct ip_tunnel),
					   "tunl0",
					   ipip_tunnel_setup);
	if (!ipip_fb_tunnel_dev) {
		err = -ENOMEM;
		goto err1;
	}

	ipip_fb_tunnel_dev->init = ipip_fb_tunnel_init;

	if ((err = register_netdev(ipip_fb_tunnel_dev)))
		goto err2;
 out:
	return err;
 err2:
	free_netdev(ipip_fb_tunnel_dev);
 err1:
896
	xfrm4_tunnel_deregister(&ipip_handler, AF_INET);
L
Linus Torvalds 已提交
897 898 899
	goto out;
}

900 901 902 903 904 905 906 907 908 909 910 911 912 913
static void __exit ipip_destroy_tunnels(void)
{
	int prio;

	for (prio = 1; 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);
		}
	}
}

L
Linus Torvalds 已提交
914 915
static void __exit ipip_fini(void)
{
916
	if (xfrm4_tunnel_deregister(&ipip_handler, AF_INET))
L
Linus Torvalds 已提交
917 918
		printk(KERN_INFO "ipip close: can't deregister tunnel\n");

919 920 921 922
	rtnl_lock();
	ipip_destroy_tunnels();
	unregister_netdevice(ipip_fb_tunnel_dev);
	rtnl_unlock();
L
Linus Torvalds 已提交
923 924 925 926 927
}

module_init(ipip_init);
module_exit(ipip_fini);
MODULE_LICENSE("GPL");