icmp.c 20.3 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
/*
 *	Internet Control Message Protocol (ICMPv6)
 *	Linux INET6 implementation
 *
 *	Authors:
 *	Pedro Roque		<roque@di.fc.ul.pt>
 *
 *	$Id: icmp.c,v 1.38 2002/02/08 03:57:19 davem Exp $
 *
 *	Based on net/ipv4/icmp.c
 *
 *	RFC 1885
 *
 *	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.
 */

/*
 *	Changes:
 *
 *	Andi Kleen		:	exception handling
 *	Andi Kleen			add rate limits. never reply to a icmp.
 *					add more length checks and other fixes.
 *	yoshfuji		:	ensure to sent parameter problem for
 *					fragments.
 *	YOSHIFUJI Hideaki @USAGI:	added sysctl for icmp rate limit.
 *	Randy Dunlap and
 *	YOSHIFUJI Hideaki @USAGI:	Per-interface statistics support
 *	Kazunori MIYAZAWA @USAGI:       change output process to use ip6_append_data
 */

#include <linux/module.h>
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/in.h>
#include <linux/kernel.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/skbuff.h>
#include <linux/init.h>
44
#include <linux/netfilter.h>
L
Linus Torvalds 已提交
45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

#ifdef CONFIG_SYSCTL
#include <linux/sysctl.h>
#endif

#include <linux/inet.h>
#include <linux/netdevice.h>
#include <linux/icmpv6.h>

#include <net/ip.h>
#include <net/sock.h>

#include <net/ipv6.h>
#include <net/ip6_checksum.h>
#include <net/protocol.h>
#include <net/raw.h>
#include <net/rawv6.h>
#include <net/transp_v6.h>
#include <net/ip6_route.h>
#include <net/addrconf.h>
#include <net/icmp.h>
66
#include <net/xfrm.h>
L
Linus Torvalds 已提交
67 68 69 70

#include <asm/uaccess.h>
#include <asm/system.h>

71
DEFINE_SNMP_STAT(struct icmpv6_mib, icmpv6_statistics) __read_mostly;
72
EXPORT_SYMBOL(icmpv6_statistics);
73 74
DEFINE_SNMP_STAT(struct icmpv6msg_mib, icmpv6msg_statistics) __read_mostly;
EXPORT_SYMBOL(icmpv6msg_statistics);
L
Linus Torvalds 已提交
75 76 77 78 79 80 81 82 83 84 85

/*
 *	The ICMP socket(s). This is the most convenient way to flow control
 *	our ICMP output as well as maintain a clean interface throughout
 *	all layers. All Socketless IP sends will soon be gone.
 *
 *	On SMP we have one ICMP socket per-cpu.
 */
static DEFINE_PER_CPU(struct socket *, __icmpv6_socket) = NULL;
#define icmpv6_socket	__get_cpu_var(__icmpv6_socket)

86
static int icmpv6_rcv(struct sk_buff *skb);
L
Linus Torvalds 已提交
87 88 89

static struct inet6_protocol icmpv6_protocol = {
	.handler	=	icmpv6_rcv,
90
	.flags		=	INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
L
Linus Torvalds 已提交
91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112
};

static __inline__ int icmpv6_xmit_lock(void)
{
	local_bh_disable();

	if (unlikely(!spin_trylock(&icmpv6_socket->sk->sk_lock.slock))) {
		/* This can happen if the output path (f.e. SIT or
		 * ip6ip6 tunnel) signals dst_link_failure() for an
		 * outgoing ICMP6 packet.
		 */
		local_bh_enable();
		return 1;
	}
	return 0;
}

static __inline__ void icmpv6_xmit_unlock(void)
{
	spin_unlock_bh(&icmpv6_socket->sk->sk_lock.slock);
}

113
/*
L
Linus Torvalds 已提交
114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134
 * Slightly more convenient version of icmpv6_send.
 */
void icmpv6_param_prob(struct sk_buff *skb, int code, int pos)
{
	icmpv6_send(skb, ICMPV6_PARAMPROB, code, pos, skb->dev);
	kfree_skb(skb);
}

/*
 * Figure out, may we reply to this packet with icmp error.
 *
 * We do not reply, if:
 *	- it was icmp error message.
 *	- it is truncated, so that it is known, that protocol is ICMPV6
 *	  (i.e. in the middle of some exthdr)
 *
 *	--ANK (980726)
 */

static int is_ineligible(struct sk_buff *skb)
{
135
	int ptr = (u8 *)(ipv6_hdr(skb) + 1) - skb->data;
L
Linus Torvalds 已提交
136
	int len = skb->len - ptr;
137
	__u8 nexthdr = ipv6_hdr(skb)->nexthdr;
L
Linus Torvalds 已提交
138 139 140 141

	if (len < 0)
		return 1;

142
	ptr = ipv6_skip_exthdr(skb, ptr, &nexthdr);
L
Linus Torvalds 已提交
143 144 145 146 147 148 149 150 151 152 153 154 155 156
	if (ptr < 0)
		return 0;
	if (nexthdr == IPPROTO_ICMPV6) {
		u8 _type, *tp;
		tp = skb_header_pointer(skb,
			ptr+offsetof(struct icmp6hdr, icmp6_type),
			sizeof(_type), &_type);
		if (tp == NULL ||
		    !(*tp & ICMPV6_INFOMSG_MASK))
			return 1;
	}
	return 0;
}

157
static int sysctl_icmpv6_time __read_mostly = 1*HZ;
L
Linus Torvalds 已提交
158

159 160
/*
 * Check the ICMP output rate limit
L
Linus Torvalds 已提交
161 162 163 164 165 166 167 168 169 170 171 172 173 174 175
 */
static inline int icmpv6_xrlim_allow(struct sock *sk, int type,
				     struct flowi *fl)
{
	struct dst_entry *dst;
	int res = 0;

	/* Informational messages are not limited. */
	if (type & ICMPV6_INFOMSG_MASK)
		return 1;

	/* Do not limit pmtu discovery, it would break it. */
	if (type == ICMPV6_PKT_TOOBIG)
		return 1;

176
	/*
L
Linus Torvalds 已提交
177 178 179 180 181 182
	 * Look up the output route.
	 * XXX: perhaps the expire for routing entries cloned by
	 * this lookup should be more aggressive (not longer than timeout).
	 */
	dst = ip6_route_output(sk, fl);
	if (dst->error) {
183 184
		IP6_INC_STATS(ip6_dst_idev(dst),
			      IPSTATS_MIB_OUTNOROUTES);
L
Linus Torvalds 已提交
185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203
	} else if (dst->dev && (dst->dev->flags&IFF_LOOPBACK)) {
		res = 1;
	} else {
		struct rt6_info *rt = (struct rt6_info *)dst;
		int tmo = sysctl_icmpv6_time;

		/* Give more bandwidth to wider prefixes. */
		if (rt->rt6i_dst.plen < 128)
			tmo >>= ((128 - rt->rt6i_dst.plen)>>5);

		res = xrlim_allow(dst, tmo);
	}
	dst_release(dst);
	return res;
}

/*
 *	an inline helper for the "simple" if statement below
 *	checks if parameter problem report is caused by an
204
 *	unrecognized IPv6 option that has the Option Type
L
Linus Torvalds 已提交
205 206 207 208 209 210 211
 *	highest-order two bits set to 10
 */

static __inline__ int opt_unrec(struct sk_buff *skb, __u32 offset)
{
	u8 _optval, *op;

212
	offset += skb_network_offset(skb);
L
Linus Torvalds 已提交
213 214 215 216 217 218 219 220 221 222 223 224 225 226 227
	op = skb_header_pointer(skb, offset, sizeof(_optval), &_optval);
	if (op == NULL)
		return 1;
	return (*op & 0xC0) == 0x80;
}

static int icmpv6_push_pending_frames(struct sock *sk, struct flowi *fl, struct icmp6hdr *thdr, int len)
{
	struct sk_buff *skb;
	struct icmp6hdr *icmp6h;
	int err = 0;

	if ((skb = skb_peek(&sk->sk_write_queue)) == NULL)
		goto out;

228
	icmp6h = icmp6_hdr(skb);
L
Linus Torvalds 已提交
229 230 231 232 233 234 235 236 237 238 239
	memcpy(icmp6h, thdr, sizeof(struct icmp6hdr));
	icmp6h->icmp6_cksum = 0;

	if (skb_queue_len(&sk->sk_write_queue) == 1) {
		skb->csum = csum_partial((char *)icmp6h,
					sizeof(struct icmp6hdr), skb->csum);
		icmp6h->icmp6_cksum = csum_ipv6_magic(&fl->fl6_src,
						      &fl->fl6_dst,
						      len, fl->proto,
						      skb->csum);
	} else {
240
		__wsum tmp_csum = 0;
L
Linus Torvalds 已提交
241 242 243 244 245 246 247

		skb_queue_walk(&sk->sk_write_queue, skb) {
			tmp_csum = csum_add(tmp_csum, skb->csum);
		}

		tmp_csum = csum_partial((char *)icmp6h,
					sizeof(struct icmp6hdr), tmp_csum);
248 249 250 251
		icmp6h->icmp6_cksum = csum_ipv6_magic(&fl->fl6_src,
						      &fl->fl6_dst,
						      len, fl->proto,
						      tmp_csum);
L
Linus Torvalds 已提交
252 253 254 255 256 257 258 259 260
	}
	ip6_push_pending_frames(sk);
out:
	return err;
}

struct icmpv6_msg {
	struct sk_buff	*skb;
	int		offset;
261
	uint8_t		type;
L
Linus Torvalds 已提交
262 263 264 265 266 267
};

static int icmpv6_getfrag(void *from, char *to, int offset, int len, int odd, struct sk_buff *skb)
{
	struct icmpv6_msg *msg = (struct icmpv6_msg *) from;
	struct sk_buff *org_skb = msg->skb;
268
	__wsum csum = 0;
L
Linus Torvalds 已提交
269 270 271 272

	csum = skb_copy_and_csum_bits(org_skb, msg->offset + offset,
				      to, len, csum);
	skb->csum = csum_block_add(skb->csum, csum, odd);
273 274
	if (!(msg->type & ICMPV6_INFOMSG_MASK))
		nf_ct_attach(skb, org_skb);
L
Linus Torvalds 已提交
275 276 277
	return 0;
}

278
#if defined(CONFIG_IPV6_MIP6) || defined(CONFIG_IPV6_MIP6_MODULE)
279 280
static void mip6_addr_swap(struct sk_buff *skb)
{
281
	struct ipv6hdr *iph = ipv6_hdr(skb);
282 283 284 285 286 287 288 289
	struct inet6_skb_parm *opt = IP6CB(skb);
	struct ipv6_destopt_hao *hao;
	struct in6_addr tmp;
	int off;

	if (opt->dsthao) {
		off = ipv6_find_tlv(skb, opt->dsthao, IPV6_TLV_HAO);
		if (likely(off >= 0)) {
290 291
			hao = (struct ipv6_destopt_hao *)
					(skb_network_header(skb) + off);
292 293 294 295 296 297 298 299 300 301
			ipv6_addr_copy(&tmp, &iph->saddr);
			ipv6_addr_copy(&iph->saddr, &hao->addr);
			ipv6_addr_copy(&hao->addr, &tmp);
		}
	}
}
#else
static inline void mip6_addr_swap(struct sk_buff *skb) {}
#endif

L
Linus Torvalds 已提交
302 303 304
/*
 *	Send an ICMP message in response to a packet in error
 */
305
void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info,
L
Linus Torvalds 已提交
306 307 308
		 struct net_device *dev)
{
	struct inet6_dev *idev = NULL;
309
	struct ipv6hdr *hdr = ipv6_hdr(skb);
310 311
	struct sock *sk;
	struct ipv6_pinfo *np;
L
Linus Torvalds 已提交
312 313
	struct in6_addr *saddr = NULL;
	struct dst_entry *dst;
314
	struct dst_entry *dst2;
L
Linus Torvalds 已提交
315 316
	struct icmp6hdr tmp_hdr;
	struct flowi fl;
317
	struct flowi fl2;
L
Linus Torvalds 已提交
318 319 320 321
	struct icmpv6_msg msg;
	int iif = 0;
	int addr_type = 0;
	int len;
322
	int hlimit, tclass;
L
Linus Torvalds 已提交
323 324
	int err = 0;

325 326
	if ((u8 *)hdr < skb->head ||
	    (skb->network_header + sizeof(*hdr)) > skb->tail)
L
Linus Torvalds 已提交
327 328 329
		return;

	/*
330
	 *	Make sure we respect the rules
L
Linus Torvalds 已提交
331 332 333 334 335 336 337 338 339 340 341 342 343 344 345
	 *	i.e. RFC 1885 2.4(e)
	 *	Rule (e.1) is enforced by not using icmpv6_send
	 *	in any code that processes icmp errors.
	 */
	addr_type = ipv6_addr_type(&hdr->daddr);

	if (ipv6_chk_addr(&hdr->daddr, skb->dev, 0))
		saddr = &hdr->daddr;

	/*
	 *	Dest addr check
	 */

	if ((addr_type & IPV6_ADDR_MULTICAST || skb->pkt_type != PACKET_HOST)) {
		if (type != ICMPV6_PKT_TOOBIG &&
346 347
		    !(type == ICMPV6_PARAMPROB &&
		      code == ICMPV6_UNK_OPTION &&
L
Linus Torvalds 已提交
348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363
		      (opt_unrec(skb, info))))
			return;

		saddr = NULL;
	}

	addr_type = ipv6_addr_type(&hdr->saddr);

	/*
	 *	Source addr check
	 */

	if (addr_type & IPV6_ADDR_LINKLOCAL)
		iif = skb->dev->ifindex;

	/*
364 365 366 367
	 *	Must not send error if the source does not uniquely
	 *	identify a single node (RFC2463 Section 2.4).
	 *	We check unspecified / multicast addresses here,
	 *	and anycast addresses will be checked later.
L
Linus Torvalds 已提交
368 369
	 */
	if ((addr_type == IPV6_ADDR_ANY) || (addr_type & IPV6_ADDR_MULTICAST)) {
370
		LIMIT_NETDEBUG(KERN_DEBUG "icmpv6_send: addr_any/mcast source\n");
L
Linus Torvalds 已提交
371 372 373
		return;
	}

374
	/*
L
Linus Torvalds 已提交
375 376 377
	 *	Never answer to a ICMP packet.
	 */
	if (is_ineligible(skb)) {
378
		LIMIT_NETDEBUG(KERN_DEBUG "icmpv6_send: no reply to icmp error\n");
L
Linus Torvalds 已提交
379 380 381
		return;
	}

382 383
	mip6_addr_swap(skb);

L
Linus Torvalds 已提交
384 385 386 387 388 389 390 391
	memset(&fl, 0, sizeof(fl));
	fl.proto = IPPROTO_ICMPV6;
	ipv6_addr_copy(&fl.fl6_dst, &hdr->saddr);
	if (saddr)
		ipv6_addr_copy(&fl.fl6_src, saddr);
	fl.oif = iif;
	fl.fl_icmp_type = type;
	fl.fl_icmp_code = code;
V
Venkat Yekkirala 已提交
392
	security_skb_classify_flow(skb, &fl);
L
Linus Torvalds 已提交
393 394 395 396

	if (icmpv6_xmit_lock())
		return;

397 398 399
	sk = icmpv6_socket->sk;
	np = inet6_sk(sk);

L
Linus Torvalds 已提交
400 401 402 403 404 405 406 407 408 409 410 411 412 413
	if (!icmpv6_xrlim_allow(sk, type, &fl))
		goto out;

	tmp_hdr.icmp6_type = type;
	tmp_hdr.icmp6_code = code;
	tmp_hdr.icmp6_cksum = 0;
	tmp_hdr.icmp6_pointer = htonl(info);

	if (!fl.oif && ipv6_addr_is_multicast(&fl.fl6_dst))
		fl.oif = np->mcast_oif;

	err = ip6_dst_lookup(sk, &dst, &fl);
	if (err)
		goto out;
414 415 416 417 418 419 420 421 422 423

	/*
	 * We won't send icmp if the destination is known
	 * anycast.
	 */
	if (((struct rt6_info *)dst)->rt6i_flags & RTF_ANYCAST) {
		LIMIT_NETDEBUG(KERN_DEBUG "icmpv6_send: acast source\n");
		goto out_dst_release;
	}

424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443
	/* No need to clone since we're just using its address. */
	dst2 = dst;

	err = xfrm_lookup(&dst, &fl, sk, 0);
	switch (err) {
	case 0:
		if (dst != dst2)
			goto route_done;
		break;
	case -EPERM:
		dst = NULL;
		break;
	default:
		goto out;
	}

	if (xfrm_decode_session_reverse(skb, &fl2, AF_INET6))
		goto out;

	if (ip6_dst_lookup(sk, &dst2, &fl))
444
		goto out;
L
Linus Torvalds 已提交
445

446 447 448 449 450 451 452 453 454 455 456 457 458 459
	err = xfrm_lookup(&dst2, &fl, sk, XFRM_LOOKUP_ICMP);
	if (err == -ENOENT) {
		if (!dst)
			goto out;
		goto route_done;
	}

	dst_release(dst);
	dst = dst2;

	if (err)
		goto out;

route_done:
L
Linus Torvalds 已提交
460 461 462 463 464 465 466 467 468
	if (ipv6_addr_is_multicast(&fl.fl6_dst))
		hlimit = np->mcast_hops;
	else
		hlimit = np->hop_limit;
	if (hlimit < 0)
		hlimit = dst_metric(dst, RTAX_HOPLIMIT);
	if (hlimit < 0)
		hlimit = ipv6_get_hoplimit(dst->dev);

469
	tclass = np->tclass;
470 471 472
	if (tclass < 0)
		tclass = 0;

L
Linus Torvalds 已提交
473
	msg.skb = skb;
474
	msg.offset = skb_network_offset(skb);
475
	msg.type = type;
L
Linus Torvalds 已提交
476 477 478 479

	len = skb->len - msg.offset;
	len = min_t(unsigned int, len, IPV6_MIN_MTU - sizeof(struct ipv6hdr) -sizeof(struct icmp6hdr));
	if (len < 0) {
480
		LIMIT_NETDEBUG(KERN_DEBUG "icmp: len problem\n");
L
Linus Torvalds 已提交
481 482 483 484 485 486 487 488
		goto out_dst_release;
	}

	idev = in6_dev_get(skb->dev);

	err = ip6_append_data(sk, icmpv6_getfrag, &msg,
			      len + sizeof(struct icmp6hdr),
			      sizeof(struct icmp6hdr),
489
			      hlimit, tclass, NULL, &fl, (struct rt6_info*)dst,
L
Linus Torvalds 已提交
490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505
			      MSG_DONTWAIT);
	if (err) {
		ip6_flush_pending_frames(sk);
		goto out_put;
	}
	err = icmpv6_push_pending_frames(sk, &fl, &tmp_hdr, len + sizeof(struct icmp6hdr));

out_put:
	if (likely(idev != NULL))
		in6_dev_put(idev);
out_dst_release:
	dst_release(dst);
out:
	icmpv6_xmit_unlock();
}

506 507
EXPORT_SYMBOL(icmpv6_send);

L
Linus Torvalds 已提交
508 509
static void icmpv6_echo_reply(struct sk_buff *skb)
{
510
	struct sock *sk;
L
Linus Torvalds 已提交
511
	struct inet6_dev *idev;
512
	struct ipv6_pinfo *np;
L
Linus Torvalds 已提交
513
	struct in6_addr *saddr = NULL;
514
	struct icmp6hdr *icmph = icmp6_hdr(skb);
L
Linus Torvalds 已提交
515 516 517 518 519 520
	struct icmp6hdr tmp_hdr;
	struct flowi fl;
	struct icmpv6_msg msg;
	struct dst_entry *dst;
	int err = 0;
	int hlimit;
521
	int tclass;
L
Linus Torvalds 已提交
522

523
	saddr = &ipv6_hdr(skb)->daddr;
L
Linus Torvalds 已提交
524 525 526 527 528 529 530 531 532

	if (!ipv6_unicast_destination(skb))
		saddr = NULL;

	memcpy(&tmp_hdr, icmph, sizeof(tmp_hdr));
	tmp_hdr.icmp6_type = ICMPV6_ECHO_REPLY;

	memset(&fl, 0, sizeof(fl));
	fl.proto = IPPROTO_ICMPV6;
533
	ipv6_addr_copy(&fl.fl6_dst, &ipv6_hdr(skb)->saddr);
L
Linus Torvalds 已提交
534 535 536 537
	if (saddr)
		ipv6_addr_copy(&fl.fl6_src, saddr);
	fl.oif = skb->dev->ifindex;
	fl.fl_icmp_type = ICMPV6_ECHO_REPLY;
V
Venkat Yekkirala 已提交
538
	security_skb_classify_flow(skb, &fl);
L
Linus Torvalds 已提交
539 540 541 542

	if (icmpv6_xmit_lock())
		return;

543 544 545
	sk = icmpv6_socket->sk;
	np = inet6_sk(sk);

L
Linus Torvalds 已提交
546 547 548 549 550 551 552
	if (!fl.oif && ipv6_addr_is_multicast(&fl.fl6_dst))
		fl.oif = np->mcast_oif;

	err = ip6_dst_lookup(sk, &dst, &fl);
	if (err)
		goto out;
	if ((err = xfrm_lookup(&dst, &fl, sk, 0)) < 0)
553
		goto out;
L
Linus Torvalds 已提交
554 555 556 557 558 559 560 561 562 563

	if (ipv6_addr_is_multicast(&fl.fl6_dst))
		hlimit = np->mcast_hops;
	else
		hlimit = np->hop_limit;
	if (hlimit < 0)
		hlimit = dst_metric(dst, RTAX_HOPLIMIT);
	if (hlimit < 0)
		hlimit = ipv6_get_hoplimit(dst->dev);

564
	tclass = np->tclass;
565 566 567
	if (tclass < 0)
		tclass = 0;

L
Linus Torvalds 已提交
568 569 570 571
	idev = in6_dev_get(skb->dev);

	msg.skb = skb;
	msg.offset = 0;
572
	msg.type = ICMPV6_ECHO_REPLY;
L
Linus Torvalds 已提交
573 574

	err = ip6_append_data(sk, icmpv6_getfrag, &msg, skb->len + sizeof(struct icmp6hdr),
575
				sizeof(struct icmp6hdr), hlimit, tclass, NULL, &fl,
L
Linus Torvalds 已提交
576 577 578 579 580 581 582 583
				(struct rt6_info*)dst, MSG_DONTWAIT);

	if (err) {
		ip6_flush_pending_frames(sk);
		goto out_put;
	}
	err = icmpv6_push_pending_frames(sk, &fl, &tmp_hdr, skb->len + sizeof(struct icmp6hdr));

584
out_put:
L
Linus Torvalds 已提交
585 586 587
	if (likely(idev != NULL))
		in6_dev_put(idev);
	dst_release(dst);
588
out:
L
Linus Torvalds 已提交
589 590 591
	icmpv6_xmit_unlock();
}

592
static void icmpv6_notify(struct sk_buff *skb, int type, int code, __be32 info)
L
Linus Torvalds 已提交
593 594 595 596 597 598 599 600 601 602 603 604
{
	struct inet6_protocol *ipprot;
	int inner_offset;
	int hash;
	u8 nexthdr;

	if (!pskb_may_pull(skb, sizeof(struct ipv6hdr)))
		return;

	nexthdr = ((struct ipv6hdr *)skb->data)->nexthdr;
	if (ipv6_ext_hdr(nexthdr)) {
		/* now skip over extension headers */
605
		inner_offset = ipv6_skip_exthdr(skb, sizeof(struct ipv6hdr), &nexthdr);
L
Linus Torvalds 已提交
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
		if (inner_offset<0)
			return;
	} else {
		inner_offset = sizeof(struct ipv6hdr);
	}

	/* Checkin header including 8 bytes of inner protocol header. */
	if (!pskb_may_pull(skb, inner_offset+8))
		return;

	/* BUGGG_FUTURE: we should try to parse exthdrs in this packet.
	   Without this we will not able f.e. to make source routed
	   pmtu discovery.
	   Corresponding argument (opt) to notifiers is already added.
	   --ANK (980726)
	 */

	hash = nexthdr & (MAX_INET_PROTOS - 1);

	rcu_read_lock();
	ipprot = rcu_dereference(inet6_protos[hash]);
	if (ipprot && ipprot->err_handler)
		ipprot->err_handler(skb, NULL, type, code, inner_offset, info);
	rcu_read_unlock();

631
	raw6_icmp_error(skb, nexthdr, type, code, inner_offset, info);
L
Linus Torvalds 已提交
632
}
633

L
Linus Torvalds 已提交
634 635 636 637
/*
 *	Handle icmp messages
 */

638
static int icmpv6_rcv(struct sk_buff *skb)
L
Linus Torvalds 已提交
639 640 641 642 643 644 645 646
{
	struct net_device *dev = skb->dev;
	struct inet6_dev *idev = __in6_dev_get(dev);
	struct in6_addr *saddr, *daddr;
	struct ipv6hdr *orig_hdr;
	struct icmp6hdr *hdr;
	int type;

647
	if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb) && skb->sp &&
648 649 650 651 652 653 654 655 656 657 658 659 660 661 662
	    skb->sp->xvec[skb->sp->len - 1]->props.flags & XFRM_STATE_ICMP) {
		int nh;

		if (!pskb_may_pull(skb, sizeof(*hdr) + sizeof(*orig_hdr)))
			goto drop_no_count;

		nh = skb_network_offset(skb);
		skb_set_network_header(skb, sizeof(*hdr));

		if (!xfrm6_policy_check_reverse(NULL, XFRM_POLICY_IN, skb))
			goto drop_no_count;

		skb_set_network_header(skb, nh);
	}

L
Linus Torvalds 已提交
663 664
	ICMP6_INC_STATS_BH(idev, ICMP6_MIB_INMSGS);

665 666
	saddr = &ipv6_hdr(skb)->saddr;
	daddr = &ipv6_hdr(skb)->daddr;
L
Linus Torvalds 已提交
667 668

	/* Perform checksum. */
669
	switch (skb->ip_summed) {
670
	case CHECKSUM_COMPLETE:
671 672 673 674 675
		if (!csum_ipv6_magic(saddr, daddr, skb->len, IPPROTO_ICMPV6,
				     skb->csum))
			break;
		/* fall through */
	case CHECKSUM_NONE:
676 677
		skb->csum = ~csum_unfold(csum_ipv6_magic(saddr, daddr, skb->len,
					     IPPROTO_ICMPV6, 0));
678
		if (__skb_checksum_complete(skb)) {
J
Joe Perches 已提交
679
			LIMIT_NETDEBUG(KERN_DEBUG "ICMPv6 checksum failed [" NIP6_FMT " > " NIP6_FMT "]\n",
680
				       NIP6(*saddr), NIP6(*daddr));
L
Linus Torvalds 已提交
681 682 683 684
			goto discard_it;
		}
	}

685
	__skb_pull(skb, sizeof(*hdr));
L
Linus Torvalds 已提交
686

687
	hdr = icmp6_hdr(skb);
L
Linus Torvalds 已提交
688 689 690

	type = hdr->icmp6_type;

691
	ICMP6MSGIN_INC_STATS_BH(idev, type);
L
Linus Torvalds 已提交
692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709

	switch (type) {
	case ICMPV6_ECHO_REQUEST:
		icmpv6_echo_reply(skb);
		break;

	case ICMPV6_ECHO_REPLY:
		/* we couldn't care less */
		break;

	case ICMPV6_PKT_TOOBIG:
		/* BUGGG_FUTURE: if packet contains rthdr, we cannot update
		   standard destination cache. Seems, only "advanced"
		   destination cache will allow to solve this problem
		   --ANK (980726)
		 */
		if (!pskb_may_pull(skb, sizeof(struct ipv6hdr)))
			goto discard_it;
710
		hdr = icmp6_hdr(skb);
L
Linus Torvalds 已提交
711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751
		orig_hdr = (struct ipv6hdr *) (hdr + 1);
		rt6_pmtu_discovery(&orig_hdr->daddr, &orig_hdr->saddr, dev,
				   ntohl(hdr->icmp6_mtu));

		/*
		 *	Drop through to notify
		 */

	case ICMPV6_DEST_UNREACH:
	case ICMPV6_TIME_EXCEED:
	case ICMPV6_PARAMPROB:
		icmpv6_notify(skb, type, hdr->icmp6_code, hdr->icmp6_mtu);
		break;

	case NDISC_ROUTER_SOLICITATION:
	case NDISC_ROUTER_ADVERTISEMENT:
	case NDISC_NEIGHBOUR_SOLICITATION:
	case NDISC_NEIGHBOUR_ADVERTISEMENT:
	case NDISC_REDIRECT:
		ndisc_rcv(skb);
		break;

	case ICMPV6_MGM_QUERY:
		igmp6_event_query(skb);
		break;

	case ICMPV6_MGM_REPORT:
		igmp6_event_report(skb);
		break;

	case ICMPV6_MGM_REDUCTION:
	case ICMPV6_NI_QUERY:
	case ICMPV6_NI_REPLY:
	case ICMPV6_MLD2_REPORT:
	case ICMPV6_DHAAD_REQUEST:
	case ICMPV6_DHAAD_REPLY:
	case ICMPV6_MOBILE_PREFIX_SOL:
	case ICMPV6_MOBILE_PREFIX_ADV:
		break;

	default:
752
		LIMIT_NETDEBUG(KERN_DEBUG "icmpv6: msg of unknown type\n");
L
Linus Torvalds 已提交
753 754 755 756 757

		/* informational */
		if (type & ICMPV6_INFOMSG_MASK)
			break;

758 759 760
		/*
		 * error of unknown type.
		 * must pass to upper level
L
Linus Torvalds 已提交
761 762 763
		 */

		icmpv6_notify(skb, type, hdr->icmp6_code, hdr->icmp6_mtu);
764 765
	}

L
Linus Torvalds 已提交
766 767 768 769 770
	kfree_skb(skb);
	return 0;

discard_it:
	ICMP6_INC_STATS_BH(idev, ICMP6_MIB_INERRORS);
771
drop_no_count:
L
Linus Torvalds 已提交
772 773 774 775
	kfree_skb(skb);
	return 0;
}

776 777 778 779 780
/*
 * Special lock-class for __icmpv6_socket:
 */
static struct lock_class_key icmpv6_socket_sk_dst_lock_key;

L
Linus Torvalds 已提交
781 782 783 784 785
int __init icmpv6_init(struct net_proto_family *ops)
{
	struct sock *sk;
	int err, i, j;

786
	for_each_possible_cpu(i) {
L
Linus Torvalds 已提交
787 788 789 790 791 792 793 794 795 796 797 798
		err = sock_create_kern(PF_INET6, SOCK_RAW, IPPROTO_ICMPV6,
				       &per_cpu(__icmpv6_socket, i));
		if (err < 0) {
			printk(KERN_ERR
			       "Failed to initialize the ICMP6 control socket "
			       "(err %d).\n",
			       err);
			goto fail;
		}

		sk = per_cpu(__icmpv6_socket, i)->sk;
		sk->sk_allocation = GFP_ATOMIC;
799 800 801 802 803 804 805 806
		/*
		 * Split off their lock-class, because sk->sk_dst_lock
		 * gets used from softirqs, which is safe for
		 * __icmpv6_socket (because those never get directly used
		 * via userspace syscalls), but unsafe for normal sockets.
		 */
		lockdep_set_class(&sk->sk_dst_lock,
				  &icmpv6_socket_sk_dst_lock_key);
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

		/* Enough space for 2 64K ICMP packets, including
		 * sk_buff struct overhead.
		 */
		sk->sk_sndbuf =
			(2 * ((64 * 1024) + sizeof(struct sk_buff)));

		sk->sk_prot->unhash(sk);
	}


	if (inet6_add_protocol(&icmpv6_protocol, IPPROTO_ICMPV6) < 0) {
		printk(KERN_ERR "Failed to register ICMP6 protocol\n");
		err = -EAGAIN;
		goto fail;
	}

	return 0;

 fail:
	for (j = 0; j < i; j++) {
		if (!cpu_possible(j))
			continue;
		sock_release(per_cpu(__icmpv6_socket, j));
	}

	return err;
}

void icmpv6_cleanup(void)
{
	int i;

840
	for_each_possible_cpu(i) {
L
Linus Torvalds 已提交
841 842 843 844 845
		sock_release(per_cpu(__icmpv6_socket, i));
	}
	inet6_del_protocol(&icmpv6_protocol, IPPROTO_ICMPV6);
}

846
static const struct icmp6_err {
L
Linus Torvalds 已提交
847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889
	int err;
	int fatal;
} tab_unreach[] = {
	{	/* NOROUTE */
		.err	= ENETUNREACH,
		.fatal	= 0,
	},
	{	/* ADM_PROHIBITED */
		.err	= EACCES,
		.fatal	= 1,
	},
	{	/* Was NOT_NEIGHBOUR, now reserved */
		.err	= EHOSTUNREACH,
		.fatal	= 0,
	},
	{	/* ADDR_UNREACH	*/
		.err	= EHOSTUNREACH,
		.fatal	= 0,
	},
	{	/* PORT_UNREACH	*/
		.err	= ECONNREFUSED,
		.fatal	= 1,
	},
};

int icmpv6_err_convert(int type, int code, int *err)
{
	int fatal = 0;

	*err = EPROTO;

	switch (type) {
	case ICMPV6_DEST_UNREACH:
		fatal = 1;
		if (code <= ICMPV6_PORT_UNREACH) {
			*err  = tab_unreach[code].err;
			fatal = tab_unreach[code].fatal;
		}
		break;

	case ICMPV6_PKT_TOOBIG:
		*err = EMSGSIZE;
		break;
890

L
Linus Torvalds 已提交
891 892 893 894 895 896 897 898
	case ICMPV6_PARAMPROB:
		*err = EPROTO;
		fatal = 1;
		break;

	case ICMPV6_TIME_EXCEED:
		*err = EHOSTUNREACH;
		break;
899
	}
L
Linus Torvalds 已提交
900 901 902 903

	return fatal;
}

904 905
EXPORT_SYMBOL(icmpv6_err_convert);

L
Linus Torvalds 已提交
906 907 908 909 910 911 912 913 914 915 916 917 918 919
#ifdef CONFIG_SYSCTL
ctl_table ipv6_icmp_table[] = {
	{
		.ctl_name	= NET_IPV6_ICMP_RATELIMIT,
		.procname	= "ratelimit",
		.data		= &sysctl_icmpv6_time,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec
	},
	{ .ctl_name = 0 },
};
#endif