ip_input.c 17.2 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7
/*
 * INET		An implementation of the TCP/IP protocol suite for the LINUX
 *		operating system.  INET is implemented using the  BSD Socket
 *		interface as the means of communication with the user level.
 *
 *		The Internet Protocol (IP) module.
 *
8
 * Authors:	Ross Biro
L
Linus Torvalds 已提交
9 10
 *		Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 *		Donald Becker, <becker@super.org>
11
 *		Alan Cox, <alan@lxorguk.ukuu.org.uk>
L
Linus Torvalds 已提交
12 13 14 15
 *		Richard Underwood
 *		Stefan Becker, <stefanb@yello.ping.de>
 *		Jorge Cwik, <jorge@laser.satlink.net>
 *		Arnt Gulbrandsen, <agulbra@nvg.unit.no>
16
 *
L
Linus Torvalds 已提交
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98
 *
 * Fixes:
 *		Alan Cox	:	Commented a couple of minor bits of surplus code
 *		Alan Cox	:	Undefining IP_FORWARD doesn't include the code
 *					(just stops a compiler warning).
 *		Alan Cox	:	Frames with >=MAX_ROUTE record routes, strict routes or loose routes
 *					are junked rather than corrupting things.
 *		Alan Cox	:	Frames to bad broadcast subnets are dumped
 *					We used to process them non broadcast and
 *					boy could that cause havoc.
 *		Alan Cox	:	ip_forward sets the free flag on the
 *					new frame it queues. Still crap because
 *					it copies the frame but at least it
 *					doesn't eat memory too.
 *		Alan Cox	:	Generic queue code and memory fixes.
 *		Fred Van Kempen :	IP fragment support (borrowed from NET2E)
 *		Gerhard Koerting:	Forward fragmented frames correctly.
 *		Gerhard Koerting: 	Fixes to my fix of the above 8-).
 *		Gerhard Koerting:	IP interface addressing fix.
 *		Linus Torvalds	:	More robustness checks
 *		Alan Cox	:	Even more checks: Still not as robust as it ought to be
 *		Alan Cox	:	Save IP header pointer for later
 *		Alan Cox	:	ip option setting
 *		Alan Cox	:	Use ip_tos/ip_ttl settings
 *		Alan Cox	:	Fragmentation bogosity removed
 *					(Thanks to Mark.Bush@prg.ox.ac.uk)
 *		Dmitry Gorodchanin :	Send of a raw packet crash fix.
 *		Alan Cox	:	Silly ip bug when an overlength
 *					fragment turns up. Now frees the
 *					queue.
 *		Linus Torvalds/ :	Memory leakage on fragmentation
 *		Alan Cox	:	handling.
 *		Gerhard Koerting:	Forwarding uses IP priority hints
 *		Teemu Rantanen	:	Fragment problems.
 *		Alan Cox	:	General cleanup, comments and reformat
 *		Alan Cox	:	SNMP statistics
 *		Alan Cox	:	BSD address rule semantics. Also see
 *					UDP as there is a nasty checksum issue
 *					if you do things the wrong way.
 *		Alan Cox	:	Always defrag, moved IP_FORWARD to the config.in file
 *		Alan Cox	: 	IP options adjust sk->priority.
 *		Pedro Roque	:	Fix mtu/length error in ip_forward.
 *		Alan Cox	:	Avoid ip_chk_addr when possible.
 *	Richard Underwood	:	IP multicasting.
 *		Alan Cox	:	Cleaned up multicast handlers.
 *		Alan Cox	:	RAW sockets demultiplex in the BSD style.
 *		Gunther Mayer	:	Fix the SNMP reporting typo
 *		Alan Cox	:	Always in group 224.0.0.1
 *	Pauline Middelink	:	Fast ip_checksum update when forwarding
 *					Masquerading support.
 *		Alan Cox	:	Multicast loopback error for 224.0.0.1
 *		Alan Cox	:	IP_MULTICAST_LOOP option.
 *		Alan Cox	:	Use notifiers.
 *		Bjorn Ekwall	:	Removed ip_csum (from slhc.c too)
 *		Bjorn Ekwall	:	Moved ip_fast_csum to ip.h (inline!)
 *		Stefan Becker   :       Send out ICMP HOST REDIRECT
 *	Arnt Gulbrandsen	:	ip_build_xmit
 *		Alan Cox	:	Per socket routing cache
 *		Alan Cox	:	Fixed routing cache, added header cache.
 *		Alan Cox	:	Loopback didn't work right in original ip_build_xmit - fixed it.
 *		Alan Cox	:	Only send ICMP_REDIRECT if src/dest are the same net.
 *		Alan Cox	:	Incoming IP option handling.
 *		Alan Cox	:	Set saddr on raw output frames as per BSD.
 *		Alan Cox	:	Stopped broadcast source route explosions.
 *		Alan Cox	:	Can disable source routing
 *		Takeshi Sone    :	Masquerading didn't work.
 *	Dave Bonn,Alan Cox	:	Faster IP forwarding whenever possible.
 *		Alan Cox	:	Memory leaks, tramples, misc debugging.
 *		Alan Cox	:	Fixed multicast (by popular demand 8))
 *		Alan Cox	:	Fixed forwarding (by even more popular demand 8))
 *		Alan Cox	:	Fixed SNMP statistics [I think]
 *	Gerhard Koerting	:	IP fragmentation forwarding fix
 *		Alan Cox	:	Device lock against page fault.
 *		Alan Cox	:	IP_HDRINCL facility.
 *	Werner Almesberger	:	Zero fragment bug
 *		Alan Cox	:	RAW IP frame length bug
 *		Alan Cox	:	Outgoing firewall on build_xmit
 *		A.N.Kuznetsov	:	IP_OPTIONS support throughout the kernel
 *		Alan Cox	:	Multicast routing hooks
 *		Jos Vos		:	Do accounting *before* call_in_firewall
 *	Willy Konynenberg	:	Transparent proxying support
 *
99
 *
L
Linus Torvalds 已提交
100 101 102 103 104
 *
 * To Fix:
 *		IP fragmentation wants rewriting cleanly. The RFC815 algorithm is much more efficient
 *		and could be made very efficient with the addition of some virtual memory hacks to permit
 *		the allocation of a buffer that can then be 'grown' by twiddling page tables.
105
 *		Output fragmentation wants updating along with the buffer management to use a single
L
Linus Torvalds 已提交
106 107 108 109 110 111 112 113 114 115
 *		interleaved copy algorithm so that fragmenting has a one copy overhead. Actual packet
 *		output should probably do its own fragmentation at the UDP/RAW layer. TCP shouldn't cause
 *		fragmentation anyway.
 *
 *		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.
 */

116 117
#define pr_fmt(fmt) "IPv4: " fmt

L
Linus Torvalds 已提交
118 119 120 121 122
#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/string.h>
#include <linux/errno.h>
123
#include <linux/slab.h>
L
Linus Torvalds 已提交
124 125 126 127 128 129

#include <linux/net.h>
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/in.h>
#include <linux/inet.h>
130
#include <linux/inetdevice.h>
L
Linus Torvalds 已提交
131 132 133 134 135 136 137 138 139 140 141 142 143
#include <linux/netdevice.h>
#include <linux/etherdevice.h>

#include <net/snmp.h>
#include <net/ip.h>
#include <net/protocol.h>
#include <net/route.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <net/arp.h>
#include <net/icmp.h>
#include <net/raw.h>
#include <net/checksum.h>
144
#include <net/inet_ecn.h>
L
Linus Torvalds 已提交
145 146 147 148
#include <linux/netfilter_ipv4.h>
#include <net/xfrm.h>
#include <linux/mroute.h>
#include <linux/netlink.h>
T
Thomas Graf 已提交
149
#include <net/dst_metadata.h>
L
Linus Torvalds 已提交
150 151

/*
E
Eric Dumazet 已提交
152
 *	Process Router Attention IP option (RFC 2113)
153
 */
154
bool ip_call_ra_chain(struct sk_buff *skb)
L
Linus Torvalds 已提交
155 156
{
	struct ip_ra_chain *ra;
157
	u8 protocol = ip_hdr(skb)->protocol;
L
Linus Torvalds 已提交
158
	struct sock *last = NULL;
159
	struct net_device *dev = skb->dev;
160
	struct net *net = dev_net(dev);
L
Linus Torvalds 已提交
161

162
	for (ra = rcu_dereference(net->ipv4.ra_chain); ra; ra = rcu_dereference(ra->next)) {
L
Linus Torvalds 已提交
163 164 165 166 167
		struct sock *sk = ra->sk;

		/* If socket is bound to an interface, only report
		 * the packet if it came  from that interface.
		 */
E
Eric Dumazet 已提交
168
		if (sk && inet_sk(sk)->inet_num == protocol &&
L
Linus Torvalds 已提交
169
		    (!sk->sk_bound_dev_if ||
170
		     sk->sk_bound_dev_if == dev->ifindex)) {
171
			if (ip_is_fragment(ip_hdr(skb))) {
172
				if (ip_defrag(net, skb, IP_DEFRAG_CALL_RA_CHAIN))
173
					return true;
L
Linus Torvalds 已提交
174 175 176 177 178 179 180 181 182 183 184 185
			}
			if (last) {
				struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
				if (skb2)
					raw_rcv(last, skb2);
			}
			last = sk;
		}
	}

	if (last) {
		raw_rcv(last, skb);
186
		return true;
L
Linus Torvalds 已提交
187
	}
188
	return false;
L
Linus Torvalds 已提交
189 190
}

191
void ip_protocol_deliver_rcu(struct net *net, struct sk_buff *skb, int protocol)
L
Linus Torvalds 已提交
192
{
193 194
	const struct net_protocol *ipprot;
	int raw, ret;
L
Linus Torvalds 已提交
195

196 197 198 199 200 201 202 203 204
resubmit:
	raw = raw_local_deliver(skb, protocol);

	ipprot = rcu_dereference(inet_protos[protocol]);
	if (ipprot) {
		if (!ipprot->no_policy) {
			if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) {
				kfree_skb(skb);
				return;
L
Linus Torvalds 已提交
205
			}
206 207 208 209 210 211 212 213 214 215 216 217 218 219
			nf_reset(skb);
		}
		ret = ipprot->handler(skb);
		if (ret < 0) {
			protocol = -ret;
			goto resubmit;
		}
		__IP_INC_STATS(net, IPSTATS_MIB_INDELIVERS);
	} else {
		if (!raw) {
			if (xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb)) {
				__IP_INC_STATS(net, IPSTATS_MIB_INUNKNOWNPROTOS);
				icmp_send(skb, ICMP_DEST_UNREACH,
					  ICMP_PROT_UNREACH, 0);
L
Linus Torvalds 已提交
220
			}
221
			kfree_skb(skb);
L
Linus Torvalds 已提交
222
		} else {
223 224
			__IP_INC_STATS(net, IPSTATS_MIB_INDELIVERS);
			consume_skb(skb);
L
Linus Torvalds 已提交
225 226
		}
	}
227 228 229 230 231 232 233 234
}

static int ip_local_deliver_finish(struct net *net, struct sock *sk, struct sk_buff *skb)
{
	__skb_pull(skb, skb_network_header_len(skb));

	rcu_read_lock();
	ip_protocol_deliver_rcu(net, skb, ip_hdr(skb)->protocol);
L
Linus Torvalds 已提交
235 236 237 238 239 240 241
	rcu_read_unlock();

	return 0;
}

/*
 * 	Deliver IP Packets to the higher protocol layers.
242
 */
L
Linus Torvalds 已提交
243 244 245 246 247
int ip_local_deliver(struct sk_buff *skb)
{
	/*
	 *	Reassemble IP fragments.
	 */
248
	struct net *net = dev_net(skb->dev);
L
Linus Torvalds 已提交
249

250
	if (ip_is_fragment(ip_hdr(skb))) {
251
		if (ip_defrag(net, skb, IP_DEFRAG_LOCAL_DELIVER))
L
Linus Torvalds 已提交
252 253 254
			return 0;
	}

255
	return NF_HOOK(NFPROTO_IPV4, NF_INET_LOCAL_IN,
256
		       net, NULL, skb, skb->dev, NULL,
L
Linus Torvalds 已提交
257 258 259
		       ip_local_deliver_finish);
}

260
static inline bool ip_rcv_options(struct sk_buff *skb)
261 262
{
	struct ip_options *opt;
263
	const struct iphdr *iph;
264 265 266 267 268 269 270 271 272 273
	struct net_device *dev = skb->dev;

	/* It looks as overkill, because not all
	   IP options require packet mangling.
	   But it is the easiest for now, especially taking
	   into account that combination of IP options
	   and running sniffer is extremely rare condition.
					      --ANK (980813)
	*/
	if (skb_cow(skb, skb_headroom(skb))) {
E
Eric Dumazet 已提交
274
		__IP_INC_STATS(dev_net(dev), IPSTATS_MIB_INDISCARDS);
275 276 277
		goto drop;
	}

278
	iph = ip_hdr(skb);
279 280
	opt = &(IPCB(skb)->opt);
	opt->optlen = iph->ihl*4 - sizeof(struct iphdr);
281

282
	if (ip_options_compile(dev_net(dev), opt, skb)) {
E
Eric Dumazet 已提交
283
		__IP_INC_STATS(dev_net(dev), IPSTATS_MIB_INHDRERRORS);
284 285 286 287
		goto drop;
	}

	if (unlikely(opt->srr)) {
288 289
		struct in_device *in_dev = __in_dev_get_rcu(dev);

290 291
		if (in_dev) {
			if (!IN_DEV_SOURCE_ROUTE(in_dev)) {
292 293 294 295
				if (IN_DEV_LOG_MARTIANS(in_dev))
					net_info_ratelimited("source route option %pI4 -> %pI4\n",
							     &iph->saddr,
							     &iph->daddr);
296 297 298 299 300 301 302 303
				goto drop;
			}
		}

		if (ip_options_rcv_srr(skb))
			goto drop;
	}

304
	return false;
305
drop:
306
	return true;
307 308
}

E
Edward Cree 已提交
309 310
static int ip_rcv_finish_core(struct net *net, struct sock *sk,
			      struct sk_buff *skb)
L
Linus Torvalds 已提交
311
{
312
	const struct iphdr *iph = ip_hdr(skb);
313
	int (*edemux)(struct sk_buff *skb);
314
	struct net_device *dev = skb->dev;
315 316
	struct rtable *rt;
	int err;
L
Linus Torvalds 已提交
317

318
	if (net->ipv4.sysctl_ip_early_demux &&
319 320 321
	    !skb_dst(skb) &&
	    !skb->sk &&
	    !ip_is_fragment(iph)) {
322 323 324 325
		const struct net_protocol *ipprot;
		int protocol = iph->protocol;

		ipprot = rcu_dereference(inet_protos[protocol]);
326
		if (ipprot && (edemux = READ_ONCE(ipprot->early_demux))) {
327 328 329
			err = edemux(skb);
			if (unlikely(err))
				goto drop_error;
E
Eric Dumazet 已提交
330 331 332
			/* must reload iph, skb->head might have changed */
			iph = ip_hdr(skb);
		}
333 334
	}

L
Linus Torvalds 已提交
335 336 337
	/*
	 *	Initialise the virtual path cache for the packet. It describes
	 *	how the packet travels inside Linux networking.
338
	 */
T
Thomas Graf 已提交
339
	if (!skb_valid_dst(skb)) {
340 341 342 343
		err = ip_route_input_noref(skb, iph->daddr, iph->saddr,
					   iph->tos, dev);
		if (unlikely(err))
			goto drop_error;
L
Linus Torvalds 已提交
344 345
	}

346
#ifdef CONFIG_IP_ROUTE_CLASSID
E
Eric Dumazet 已提交
347
	if (unlikely(skb_dst(skb)->tclassid)) {
E
Eric Dumazet 已提交
348
		struct ip_rt_acct *st = this_cpu_ptr(ip_rt_acct);
E
Eric Dumazet 已提交
349
		u32 idx = skb_dst(skb)->tclassid;
L
Linus Torvalds 已提交
350
		st[idx&0xFF].o_packets++;
351
		st[idx&0xFF].o_bytes += skb->len;
L
Linus Torvalds 已提交
352
		st[(idx>>16)&0xFF].i_packets++;
353
		st[(idx>>16)&0xFF].i_bytes += skb->len;
L
Linus Torvalds 已提交
354 355 356
	}
#endif

357 358
	if (iph->ihl > 5 && ip_rcv_options(skb))
		goto drop;
L
Linus Torvalds 已提交
359

E
Eric Dumazet 已提交
360
	rt = skb_rtable(skb);
361
	if (rt->rt_type == RTN_MULTICAST) {
E
Eric Dumazet 已提交
362
		__IP_UPD_PO_STATS(net, IPSTATS_MIB_INMCAST, skb->len);
363
	} else if (rt->rt_type == RTN_BROADCAST) {
E
Eric Dumazet 已提交
364
		__IP_UPD_PO_STATS(net, IPSTATS_MIB_INBCAST, skb->len);
365 366
	} else if (skb->pkt_type == PACKET_BROADCAST ||
		   skb->pkt_type == PACKET_MULTICAST) {
367
		struct in_device *in_dev = __in_dev_get_rcu(dev);
368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387

		/* RFC 1122 3.3.6:
		 *
		 *   When a host sends a datagram to a link-layer broadcast
		 *   address, the IP destination address MUST be a legal IP
		 *   broadcast or IP multicast address.
		 *
		 *   A host SHOULD silently discard a datagram that is received
		 *   via a link-layer broadcast (see Section 2.4) but does not
		 *   specify an IP multicast or broadcast destination address.
		 *
		 * This doesn't explicitly say L2 *broadcast*, but broadcast is
		 * in a way a form of multicast and the most common use case for
		 * this is 802.11 protecting against cross-station spoofing (the
		 * so-called "hole-196" attack) so do it for both.
		 */
		if (in_dev &&
		    IN_DEV_ORCONF(in_dev, DROP_UNICAST_IN_L2_MULTICAST))
			goto drop;
	}
388

E
Edward Cree 已提交
389
	return NET_RX_SUCCESS;
L
Linus Torvalds 已提交
390 391

drop:
392 393
	kfree_skb(skb);
	return NET_RX_DROP;
394 395 396 397 398

drop_error:
	if (err == -EXDEV)
		__NET_INC_STATS(net, LINUX_MIB_IPRPFILTER);
	goto drop;
L
Linus Torvalds 已提交
399 400
}

E
Edward Cree 已提交
401 402
static int ip_rcv_finish(struct net *net, struct sock *sk, struct sk_buff *skb)
{
403 404 405 406 407 408 409 410
	int ret;

	/* if ingress device is enslaved to an L3 master device pass the
	 * skb to its handler for processing
	 */
	skb = l3mdev_ip_rcv(skb);
	if (!skb)
		return NET_RX_SUCCESS;
E
Edward Cree 已提交
411

412
	ret = ip_rcv_finish_core(net, sk, skb);
E
Edward Cree 已提交
413 414 415 416 417
	if (ret != NET_RX_DROP)
		ret = dst_input(skb);
	return ret;
}

L
Linus Torvalds 已提交
418 419
/*
 * 	Main IP Receive routine.
420
 */
421
static struct sk_buff *ip_rcv_core(struct sk_buff *skb, struct net *net)
L
Linus Torvalds 已提交
422
{
423
	const struct iphdr *iph;
424
	u32 len;
L
Linus Torvalds 已提交
425 426 427 428 429 430 431

	/* When the interface is in promisc. mode, drop all the crap
	 * that it receives, do not try to analyse it.
	 */
	if (skb->pkt_type == PACKET_OTHERHOST)
		goto drop;

432

E
Eric Dumazet 已提交
433
	__IP_UPD_PO_STATS(net, IPSTATS_MIB_IN, skb->len);
L
Linus Torvalds 已提交
434

435 436
	skb = skb_share_check(skb, GFP_ATOMIC);
	if (!skb) {
E
Eric Dumazet 已提交
437
		__IP_INC_STATS(net, IPSTATS_MIB_INDISCARDS);
L
Linus Torvalds 已提交
438 439 440 441 442 443
		goto out;
	}

	if (!pskb_may_pull(skb, sizeof(struct iphdr)))
		goto inhdr_error;

444
	iph = ip_hdr(skb);
L
Linus Torvalds 已提交
445 446

	/*
447
	 *	RFC1122: 3.2.1.2 MUST silently discard any IP frame that fails the checksum.
L
Linus Torvalds 已提交
448 449 450 451 452 453 454 455 456 457
	 *
	 *	Is the datagram acceptable?
	 *
	 *	1.	Length at least the size of an ip header
	 *	2.	Version of 4
	 *	3.	Checksums correctly. [Speed optimisation for later, skip loopback checksums]
	 *	4.	Doesn't have a bogus length
	 */

	if (iph->ihl < 5 || iph->version != 4)
458
		goto inhdr_error;
L
Linus Torvalds 已提交
459

460 461 462
	BUILD_BUG_ON(IPSTATS_MIB_ECT1PKTS != IPSTATS_MIB_NOECTPKTS + INET_ECN_ECT_1);
	BUILD_BUG_ON(IPSTATS_MIB_ECT0PKTS != IPSTATS_MIB_NOECTPKTS + INET_ECN_ECT_0);
	BUILD_BUG_ON(IPSTATS_MIB_CEPKTS != IPSTATS_MIB_NOECTPKTS + INET_ECN_CE);
E
Eric Dumazet 已提交
463 464 465
	__IP_ADD_STATS(net,
		       IPSTATS_MIB_NOECTPKTS + (iph->tos & INET_ECN_MASK),
		       max_t(unsigned short, 1, skb_shinfo(skb)->gso_segs));
466

L
Linus Torvalds 已提交
467 468 469
	if (!pskb_may_pull(skb, iph->ihl*4))
		goto inhdr_error;

470
	iph = ip_hdr(skb);
L
Linus Torvalds 已提交
471

472
	if (unlikely(ip_fast_csum((u8 *)iph, iph->ihl)))
473
		goto csum_error;
L
Linus Torvalds 已提交
474

475
	len = ntohs(iph->tot_len);
476
	if (skb->len < len) {
E
Eric Dumazet 已提交
477
		__IP_INC_STATS(net, IPSTATS_MIB_INTRUNCATEDPKTS);
478 479
		goto drop;
	} else if (len < (iph->ihl*4))
480
		goto inhdr_error;
L
Linus Torvalds 已提交
481

482 483 484 485 486
	/* Our transport medium may have padded the buffer out. Now we know it
	 * is IP we can trim to the true length of the frame.
	 * Note this now means skb->len holds ntohs(iph->tot_len).
	 */
	if (pskb_trim_rcsum(skb, len)) {
E
Eric Dumazet 已提交
487
		__IP_INC_STATS(net, IPSTATS_MIB_INDISCARDS);
488
		goto drop;
L
Linus Torvalds 已提交
489 490
	}

491
	iph = ip_hdr(skb);
492 493
	skb->transport_header = skb->network_header + iph->ihl*4;

494
	/* Remove any debris in the socket control block */
495
	memset(IPCB(skb), 0, sizeof(struct inet_skb_parm));
496
	IPCB(skb)->iif = skb->skb_iif;
497

498 499 500
	/* Must drop socket now because of tproxy. */
	skb_orphan(skb);

501
	return skb;
L
Linus Torvalds 已提交
502

503
csum_error:
E
Eric Dumazet 已提交
504
	__IP_INC_STATS(net, IPSTATS_MIB_CSUMERRORS);
L
Linus Torvalds 已提交
505
inhdr_error:
E
Eric Dumazet 已提交
506
	__IP_INC_STATS(net, IPSTATS_MIB_INHDRERRORS);
L
Linus Torvalds 已提交
507
drop:
508
	kfree_skb(skb);
L
Linus Torvalds 已提交
509
out:
510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528
	return NULL;
}

/*
 * IP receive entry point
 */
int ip_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt,
	   struct net_device *orig_dev)
{
	struct net *net = dev_net(dev);

	skb = ip_rcv_core(skb, net);
	if (skb == NULL)
		return NET_RX_DROP;
	return NF_HOOK(NFPROTO_IPV4, NF_INET_PRE_ROUTING,
		       net, NULL, skb, dev, NULL,
		       ip_rcv_finish);
}

E
Edward Cree 已提交
529
static void ip_sublist_rcv_finish(struct list_head *head)
530 531 532
{
	struct sk_buff *skb, *next;

533
	list_for_each_entry_safe(skb, next, head, list) {
534
		skb_list_del_init(skb);
E
Edward Cree 已提交
535
		dst_input(skb);
536
	}
E
Edward Cree 已提交
537 538 539 540 541 542 543 544 545
}

static void ip_list_rcv_finish(struct net *net, struct sock *sk,
			       struct list_head *head)
{
	struct dst_entry *curr_dst = NULL;
	struct sk_buff *skb, *next;
	struct list_head sublist;

546
	INIT_LIST_HEAD(&sublist);
E
Edward Cree 已提交
547 548 549
	list_for_each_entry_safe(skb, next, head, list) {
		struct dst_entry *dst;

550
		skb_list_del_init(skb);
551 552 553 554 555 556
		/* if ingress device is enslaved to an L3 master device pass the
		 * skb to its handler for processing
		 */
		skb = l3mdev_ip_rcv(skb);
		if (!skb)
			continue;
E
Edward Cree 已提交
557 558 559 560 561 562 563 564 565
		if (ip_rcv_finish_core(net, sk, skb) == NET_RX_DROP)
			continue;

		dst = skb_dst(skb);
		if (curr_dst != dst) {
			/* dispatch old sublist */
			if (!list_empty(&sublist))
				ip_sublist_rcv_finish(&sublist);
			/* start new sublist */
566
			INIT_LIST_HEAD(&sublist);
E
Edward Cree 已提交
567 568
			curr_dst = dst;
		}
569
		list_add_tail(&skb->list, &sublist);
E
Edward Cree 已提交
570 571
	}
	/* dispatch final sublist */
572
	ip_sublist_rcv_finish(&sublist);
E
Edward Cree 已提交
573 574 575 576 577
}

static void ip_sublist_rcv(struct list_head *head, struct net_device *dev,
			   struct net *net)
{
578 579
	NF_HOOK_LIST(NFPROTO_IPV4, NF_INET_PRE_ROUTING, net, NULL,
		     head, dev, NULL, ip_rcv_finish);
E
Edward Cree 已提交
580
	ip_list_rcv_finish(net, NULL, head);
581 582 583 584 585 586 587 588 589 590 591
}

/* Receive a list of IP packets */
void ip_list_rcv(struct list_head *head, struct packet_type *pt,
		 struct net_device *orig_dev)
{
	struct net_device *curr_dev = NULL;
	struct net *curr_net = NULL;
	struct sk_buff *skb, *next;
	struct list_head sublist;

592
	INIT_LIST_HEAD(&sublist);
593 594 595 596
	list_for_each_entry_safe(skb, next, head, list) {
		struct net_device *dev = skb->dev;
		struct net *net = dev_net(dev);

597
		skb_list_del_init(skb);
598 599 600 601 602 603 604
		skb = ip_rcv_core(skb, net);
		if (skb == NULL)
			continue;

		if (curr_dev != dev || curr_net != net) {
			/* dispatch old sublist */
			if (!list_empty(&sublist))
605
				ip_sublist_rcv(&sublist, curr_dev, curr_net);
606
			/* start new sublist */
607
			INIT_LIST_HEAD(&sublist);
608 609 610
			curr_dev = dev;
			curr_net = net;
		}
611
		list_add_tail(&skb->list, &sublist);
612 613
	}
	/* dispatch final sublist */
614
	ip_sublist_rcv(&sublist, curr_dev, curr_net);
L
Linus Torvalds 已提交
615
}