xref: /netbsd/sys/netinet/ip_input.c (revision 61127047)
1 /*	$NetBSD: ip_input.c,v 1.402 2022/09/02 03:50:00 thorpej Exp $	*/
2 
3 /*
4  * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
5  * All rights reserved.
6  *
7  * Redistribution and use in source and binary forms, with or without
8  * modification, are permitted provided that the following conditions
9  * are met:
10  * 1. Redistributions of source code must retain the above copyright
11  *    notice, this list of conditions and the following disclaimer.
12  * 2. Redistributions in binary form must reproduce the above copyright
13  *    notice, this list of conditions and the following disclaimer in the
14  *    documentation and/or other materials provided with the distribution.
15  * 3. Neither the name of the project nor the names of its contributors
16  *    may be used to endorse or promote products derived from this software
17  *    without specific prior written permission.
18  *
19  * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
20  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22  * ARE DISCLAIMED.  IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
23  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
25  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
26  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29  * SUCH DAMAGE.
30  */
31 
32 /*
33  * Copyright (c) 1998 The NetBSD Foundation, Inc.
34  * All rights reserved.
35  *
36  * This code is derived from software contributed to The NetBSD Foundation
37  * by Public Access Networks Corporation ("Panix").  It was developed under
38  * contract to Panix by Eric Haszlakiewicz and Thor Lancelot Simon.
39  *
40  * Redistribution and use in source and binary forms, with or without
41  * modification, are permitted provided that the following conditions
42  * are met:
43  * 1. Redistributions of source code must retain the above copyright
44  *    notice, this list of conditions and the following disclaimer.
45  * 2. Redistributions in binary form must reproduce the above copyright
46  *    notice, this list of conditions and the following disclaimer in the
47  *    documentation and/or other materials provided with the distribution.
48  *
49  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
50  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
51  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
52  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
53  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
54  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
55  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
56  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
57  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
58  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
59  * POSSIBILITY OF SUCH DAMAGE.
60  */
61 
62 /*
63  * Copyright (c) 1982, 1986, 1988, 1993
64  *	The Regents of the University of California.  All rights reserved.
65  *
66  * Redistribution and use in source and binary forms, with or without
67  * modification, are permitted provided that the following conditions
68  * are met:
69  * 1. Redistributions of source code must retain the above copyright
70  *    notice, this list of conditions and the following disclaimer.
71  * 2. Redistributions in binary form must reproduce the above copyright
72  *    notice, this list of conditions and the following disclaimer in the
73  *    documentation and/or other materials provided with the distribution.
74  * 3. Neither the name of the University nor the names of its contributors
75  *    may be used to endorse or promote products derived from this software
76  *    without specific prior written permission.
77  *
78  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
79  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
80  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
81  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
82  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
83  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
84  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
85  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
86  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
87  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
88  * SUCH DAMAGE.
89  *
90  *	@(#)ip_input.c	8.2 (Berkeley) 1/4/94
91  */
92 
93 #include <sys/cdefs.h>
94 __KERNEL_RCSID(0, "$NetBSD: ip_input.c,v 1.402 2022/09/02 03:50:00 thorpej Exp $");
95 
96 #ifdef _KERNEL_OPT
97 #include "opt_inet.h"
98 #include "opt_gateway.h"
99 #include "opt_ipsec.h"
100 #include "opt_mrouting.h"
101 #include "opt_mbuftrace.h"
102 #include "opt_inet_csum.h"
103 #include "opt_net_mpsafe.h"
104 #endif
105 
106 #include "arp.h"
107 
108 #include <sys/param.h>
109 #include <sys/systm.h>
110 #include <sys/cpu.h>
111 #include <sys/mbuf.h>
112 #include <sys/domain.h>
113 #include <sys/protosw.h>
114 #include <sys/socket.h>
115 #include <sys/socketvar.h>
116 #include <sys/errno.h>
117 #include <sys/time.h>
118 #include <sys/kernel.h>
119 #include <sys/pool.h>
120 #include <sys/sysctl.h>
121 #include <sys/kauth.h>
122 
123 #include <net/if.h>
124 #include <net/if_dl.h>
125 #include <net/route.h>
126 #include <net/pktqueue.h>
127 #include <net/pfil.h>
128 
129 #include <netinet/in.h>
130 #include <netinet/in_systm.h>
131 #include <netinet/ip.h>
132 #include <netinet/in_pcb.h>
133 #include <netinet/in_proto.h>
134 #include <netinet/in_var.h>
135 #include <netinet/ip_var.h>
136 #include <netinet/ip_private.h>
137 #include <netinet/ip_icmp.h>
138 /* just for gif_ttl */
139 #include <netinet/in_gif.h>
140 #include "gif.h"
141 #include <net/if_gre.h>
142 #include "gre.h"
143 
144 #ifdef MROUTING
145 #include <netinet/ip_mroute.h>
146 #endif
147 #include <netinet/portalgo.h>
148 
149 #ifdef IPSEC
150 #include <netipsec/ipsec.h>
151 #endif
152 
153 #ifndef	IPFORWARDING
154 #ifdef GATEWAY
155 #define	IPFORWARDING	1	/* forward IP packets not for us */
156 #else
157 #define	IPFORWARDING	0	/* don't forward IP packets not for us */
158 #endif
159 #endif
160 
161 #define IPMTUDISCTIMEOUT (10 * 60)	/* as per RFC 1191 */
162 
163 int ipforwarding = IPFORWARDING;
164 int ipsendredirects = 1;
165 int ip_defttl = IPDEFTTL;
166 int ip_forwsrcrt = 0;
167 int ip_directedbcast = 0;
168 int ip_allowsrcrt = 0;
169 int ip_mtudisc = 1;
170 int ip_mtudisc_timeout = IPMTUDISCTIMEOUT;
171 int ip_do_randomid = 1;
172 
173 /*
174  * XXX - Setting ip_checkinterface mostly implements the receive side of
175  * the Strong ES model described in RFC 1122, but since the routing table
176  * and transmit implementation do not implement the Strong ES model,
177  * setting this to 1 results in an odd hybrid.
178  *
179  * XXX - ip_checkinterface currently must be disabled if you use NAT
180  * to translate the destination address to another local interface.
181  *
182  * XXX - ip_checkinterface must be disabled if you add IP aliases
183  * to the loopback interface instead of the interface where the
184  * packets for those addresses are received.
185  */
186 static int		ip_checkinterface	__read_mostly = 0;
187 
188 struct rttimer_queue *ip_mtudisc_timeout_q = NULL;
189 
190 pktqueue_t *		ip_pktq			__read_mostly;
191 pfil_head_t *		inet_pfil_hook		__read_mostly;
192 percpu_t *		ipstat_percpu		__read_mostly;
193 
194 static percpu_t		*ipforward_rt_percpu	__cacheline_aligned;
195 
196 uint16_t ip_id;
197 
198 #ifdef INET_CSUM_COUNTERS
199 #include <sys/device.h>
200 
201 struct evcnt ip_hwcsum_bad = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
202     NULL, "inet", "hwcsum bad");
203 struct evcnt ip_hwcsum_ok = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
204     NULL, "inet", "hwcsum ok");
205 struct evcnt ip_swcsum = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
206     NULL, "inet", "swcsum");
207 
208 #define	INET_CSUM_COUNTER_INCR(ev)	(ev)->ev_count++
209 
210 EVCNT_ATTACH_STATIC(ip_hwcsum_bad);
211 EVCNT_ATTACH_STATIC(ip_hwcsum_ok);
212 EVCNT_ATTACH_STATIC(ip_swcsum);
213 
214 #else
215 
216 #define	INET_CSUM_COUNTER_INCR(ev)	/* nothing */
217 
218 #endif /* INET_CSUM_COUNTERS */
219 
220 /*
221  * Used to save the IP options in case a protocol wants to respond
222  * to an incoming packet over the same route if the packet got here
223  * using IP source routing.  This allows connection establishment and
224  * maintenance when the remote end is on a network that is not known
225  * to us.
226  */
227 struct ip_srcrt {
228 	int		isr_nhops;		   /* number of hops */
229 	struct in_addr	isr_dst;		   /* final destination */
230 	char		isr_nop;		   /* one NOP to align */
231 	char		isr_hdr[IPOPT_OFFSET + 1]; /* OPTVAL, OLEN & OFFSET */
232 	struct in_addr	isr_routes[MAX_IPOPTLEN/sizeof(struct in_addr)];
233 };
234 
235 static int ip_drainwanted;
236 
237 static void save_rte(struct mbuf *, u_char *, struct in_addr);
238 
239 #ifdef MBUFTRACE
240 struct mowner ip_rx_mowner = MOWNER_INIT("internet", "rx");
241 struct mowner ip_tx_mowner = MOWNER_INIT("internet", "tx");
242 #endif
243 
244 static void		ipintr(void *);
245 static void		ip_input(struct mbuf *, struct ifnet *);
246 static void		ip_forward(struct mbuf *, int, struct ifnet *);
247 static bool		ip_dooptions(struct mbuf *);
248 static struct in_ifaddr *ip_rtaddr(struct in_addr, struct psref *);
249 static void		sysctl_net_inet_ip_setup(struct sysctllog **);
250 
251 static struct in_ifaddr	*ip_match_our_address(struct ifnet *, struct ip *,
252 			    int *);
253 static struct in_ifaddr	*ip_match_our_address_broadcast(struct ifnet *,
254 			    struct ip *);
255 
256 #ifdef NET_MPSAFE
257 #define	SOFTNET_LOCK()		mutex_enter(softnet_lock)
258 #define	SOFTNET_UNLOCK()	mutex_exit(softnet_lock)
259 #else
260 #define	SOFTNET_LOCK()		KASSERT(mutex_owned(softnet_lock))
261 #define	SOFTNET_UNLOCK()	KASSERT(mutex_owned(softnet_lock))
262 #endif
263 
264 /*
265  * IP initialization: fill in IP protocol switch table.
266  * All protocols not implemented in kernel go to raw IP protocol handler.
267  */
268 void
ip_init(void)269 ip_init(void)
270 {
271 	const struct protosw *pr;
272 
273 	ip_pktq = pktq_create(IFQ_MAXLEN, ipintr, NULL);
274 	KASSERT(ip_pktq != NULL);
275 
276 	in_init();
277 	sysctl_net_inet_ip_setup(NULL);
278 
279 	pr = pffindproto(PF_INET, IPPROTO_RAW, SOCK_RAW);
280 	KASSERT(pr != NULL);
281 
282 	for (u_int i = 0; i < IPPROTO_MAX; i++) {
283 		ip_protox[i] = pr - inetsw;
284 	}
285 	for (pr = inetdomain.dom_protosw;
286 	    pr < inetdomain.dom_protoswNPROTOSW; pr++)
287 		if (pr->pr_domain->dom_family == PF_INET &&
288 		    pr->pr_protocol && pr->pr_protocol != IPPROTO_RAW)
289 			ip_protox[pr->pr_protocol] = pr - inetsw;
290 
291 	ip_reass_init();
292 
293 	ip_id = time_uptime & 0xfffff;
294 
295 #ifdef GATEWAY
296 	ipflow_init();
297 #endif
298 
299 	/* Register our Packet Filter hook. */
300 	inet_pfil_hook = pfil_head_create(PFIL_TYPE_AF, (void *)AF_INET);
301 	KASSERT(inet_pfil_hook != NULL);
302 
303 #ifdef MBUFTRACE
304 	MOWNER_ATTACH(&ip_tx_mowner);
305 	MOWNER_ATTACH(&ip_rx_mowner);
306 #endif
307 
308 	ipstat_percpu = percpu_alloc(sizeof(uint64_t) * IP_NSTATS);
309 	ipforward_rt_percpu = rtcache_percpu_alloc();
310 	ip_mtudisc_timeout_q = rt_timer_queue_create(ip_mtudisc_timeout);
311 }
312 
313 static struct in_ifaddr *
ip_match_our_address(struct ifnet * ifp,struct ip * ip,int * downmatch)314 ip_match_our_address(struct ifnet *ifp, struct ip *ip, int *downmatch)
315 {
316 	struct in_ifaddr *ia = NULL;
317 	int checkif;
318 
319 	/*
320 	 * Enable a consistency check between the destination address
321 	 * and the arrival interface for a unicast packet (the RFC 1122
322 	 * strong ES model) if IP forwarding is disabled and the packet
323 	 * is not locally generated.
324 	 *
325 	 * XXX - We need to add a per ifaddr flag for this so that
326 	 * we get finer grain control.
327 	 */
328 	checkif = ip_checkinterface && (ipforwarding == 0) &&
329 	    (ifp->if_flags & IFF_LOOPBACK) == 0;
330 
331 	IN_ADDRHASH_READER_FOREACH(ia, ip->ip_dst.s_addr) {
332 		if (in_hosteq(ia->ia_addr.sin_addr, ip->ip_dst)) {
333 			if (ia->ia4_flags & IN_IFF_NOTREADY)
334 				continue;
335 			if (checkif && ia->ia_ifp != ifp)
336 				continue;
337 			if ((ia->ia_ifp->if_flags & IFF_UP) == 0) {
338 				(*downmatch)++;
339 				continue;
340 			}
341 			if (ia->ia4_flags & IN_IFF_DETACHED &&
342 			    (ifp->if_flags & IFF_LOOPBACK) == 0)
343 				continue;
344 			break;
345 		}
346 	}
347 
348 	return ia;
349 }
350 
351 static struct in_ifaddr *
ip_match_our_address_broadcast(struct ifnet * ifp,struct ip * ip)352 ip_match_our_address_broadcast(struct ifnet *ifp, struct ip *ip)
353 {
354 	struct in_ifaddr *ia = NULL;
355 	struct ifaddr *ifa;
356 
357 	IFADDR_READER_FOREACH(ifa, ifp) {
358 		if (ifa->ifa_addr->sa_family != AF_INET)
359 			continue;
360 		ia = ifatoia(ifa);
361 		if (ia->ia4_flags & IN_IFF_NOTREADY)
362 			continue;
363 		if (ia->ia4_flags & IN_IFF_DETACHED &&
364 		    (ifp->if_flags & IFF_LOOPBACK) == 0)
365 			continue;
366 		if (in_hosteq(ip->ip_dst, ia->ia_broadaddr.sin_addr) ||
367 		    in_hosteq(ip->ip_dst, ia->ia_netbroadcast) ||
368 		    /*
369 		     * Look for all-0's host part (old broadcast addr),
370 		     * either for subnet or net.
371 		     */
372 		    ip->ip_dst.s_addr == ia->ia_subnet ||
373 		    ip->ip_dst.s_addr == ia->ia_net)
374 			goto matched;
375 		/*
376 		 * An interface with IP address zero accepts
377 		 * all packets that arrive on that interface.
378 		 */
379 		if (in_nullhost(ia->ia_addr.sin_addr))
380 			goto matched;
381 	}
382 	ia = NULL;
383 
384 matched:
385 	return ia;
386 }
387 
388 /*
389  * IP software interrupt routine.
390  */
391 static void
ipintr(void * arg __unused)392 ipintr(void *arg __unused)
393 {
394 	struct mbuf *m;
395 
396 	KASSERT(cpu_softintr_p());
397 
398 	SOFTNET_KERNEL_LOCK_UNLESS_NET_MPSAFE();
399 	while ((m = pktq_dequeue(ip_pktq)) != NULL) {
400 		struct ifnet *ifp;
401 		struct psref psref;
402 
403 		ifp = m_get_rcvif_psref(m, &psref);
404 		if (__predict_false(ifp == NULL)) {
405 			IP_STATINC(IP_STAT_IFDROP);
406 			m_freem(m);
407 			continue;
408 		}
409 
410 		ip_input(m, ifp);
411 
412 		m_put_rcvif_psref(ifp, &psref);
413 	}
414 	SOFTNET_KERNEL_UNLOCK_UNLESS_NET_MPSAFE();
415 }
416 
417 /*
418  * IP input routine.  Checksum and byte swap header.  If fragmented
419  * try to reassemble.  Process options.  Pass to next level.
420  */
421 static void
ip_input(struct mbuf * m,struct ifnet * ifp)422 ip_input(struct mbuf *m, struct ifnet *ifp)
423 {
424 	struct ip *ip = NULL;
425 	struct in_ifaddr *ia = NULL;
426 	int hlen = 0, len;
427 	int downmatch;
428 	int srcrt = 0;
429 	int s;
430 
431 	KASSERTMSG(cpu_softintr_p(), "ip_input: not in the software "
432 	    "interrupt handler; synchronization assumptions violated");
433 
434 	MCLAIM(m, &ip_rx_mowner);
435 	KASSERT((m->m_flags & M_PKTHDR) != 0);
436 
437 	/*
438 	 * If no IP addresses have been set yet but the interfaces
439 	 * are receiving, can't do anything with incoming packets yet.
440 	 * Note: we pre-check without locks held.
441 	 */
442 	if (IN_ADDRLIST_READER_EMPTY()) {
443 		IP_STATINC(IP_STAT_IFDROP);
444 		goto out;
445 	}
446 
447 	IP_STATINC(IP_STAT_TOTAL);
448 
449 	/*
450 	 * If the IP header is not aligned, slurp it up into a new
451 	 * mbuf with space for link headers, in the event we forward
452 	 * it.  Otherwise, if it is aligned, make sure the entire
453 	 * base IP header is in the first mbuf of the chain.
454 	 */
455 	if (M_GET_ALIGNED_HDR(&m, struct ip, true) != 0) {
456 		/* XXXJRT new stat, please */
457 		IP_STATINC(IP_STAT_TOOSMALL);
458 		goto out;
459 	}
460 	ip = mtod(m, struct ip *);
461 	if (ip->ip_v != IPVERSION) {
462 		IP_STATINC(IP_STAT_BADVERS);
463 		goto out;
464 	}
465 	hlen = ip->ip_hl << 2;
466 	if (hlen < sizeof(struct ip)) {	/* minimum header length */
467 		IP_STATINC(IP_STAT_BADHLEN);
468 		goto out;
469 	}
470 	if (hlen > m->m_len) {
471 		if ((m = m_pullup(m, hlen)) == NULL) {
472 			IP_STATINC(IP_STAT_BADHLEN);
473 			goto out;
474 		}
475 		ip = mtod(m, struct ip *);
476 	}
477 
478 	/*
479 	 * RFC1122: packets with a multicast source address are
480 	 * not allowed.
481 	 */
482 	if (IN_MULTICAST(ip->ip_src.s_addr)) {
483 		IP_STATINC(IP_STAT_BADADDR);
484 		goto out;
485 	}
486 
487 	/* 127/8 must not appear on wire - RFC1122 */
488 	if ((ntohl(ip->ip_dst.s_addr) >> IN_CLASSA_NSHIFT) == IN_LOOPBACKNET ||
489 	    (ntohl(ip->ip_src.s_addr) >> IN_CLASSA_NSHIFT) == IN_LOOPBACKNET) {
490 		if ((ifp->if_flags & IFF_LOOPBACK) == 0) {
491 			IP_STATINC(IP_STAT_BADADDR);
492 			goto out;
493 		}
494 	}
495 
496 	switch (m->m_pkthdr.csum_flags &
497 		((ifp->if_csum_flags_rx & M_CSUM_IPv4) | M_CSUM_IPv4_BAD)) {
498 	case M_CSUM_IPv4|M_CSUM_IPv4_BAD:
499 		INET_CSUM_COUNTER_INCR(&ip_hwcsum_bad);
500 		IP_STATINC(IP_STAT_BADSUM);
501 		goto out;
502 
503 	case M_CSUM_IPv4:
504 		/* Checksum was okay. */
505 		INET_CSUM_COUNTER_INCR(&ip_hwcsum_ok);
506 		break;
507 
508 	default:
509 		/*
510 		 * Must compute it ourselves.  Maybe skip checksum on
511 		 * loopback interfaces.
512 		 */
513 		if (__predict_true(!(ifp->if_flags & IFF_LOOPBACK) ||
514 		    ip_do_loopback_cksum)) {
515 			INET_CSUM_COUNTER_INCR(&ip_swcsum);
516 			if (in_cksum(m, hlen) != 0) {
517 				IP_STATINC(IP_STAT_BADSUM);
518 				goto out;
519 			}
520 		}
521 		break;
522 	}
523 
524 	/* Retrieve the packet length. */
525 	len = ntohs(ip->ip_len);
526 
527 	/*
528 	 * Check for additional length bogosity
529 	 */
530 	if (len < hlen) {
531 		IP_STATINC(IP_STAT_BADLEN);
532 		goto out;
533 	}
534 
535 	/*
536 	 * Check that the amount of data in the buffers is at least as much
537 	 * as the IP header would have us expect. Trim mbufs if longer than
538 	 * we expect. Drop packet if shorter than we expect.
539 	 */
540 	if (m->m_pkthdr.len < len) {
541 		IP_STATINC(IP_STAT_TOOSHORT);
542 		goto out;
543 	}
544 	if (m->m_pkthdr.len > len) {
545 		if (m->m_len == m->m_pkthdr.len) {
546 			m->m_len = len;
547 			m->m_pkthdr.len = len;
548 		} else
549 			m_adj(m, len - m->m_pkthdr.len);
550 	}
551 
552 	/*
553 	 * Assume that we can create a fast-forward IP flow entry
554 	 * based on this packet.
555 	 */
556 	m->m_flags |= M_CANFASTFWD;
557 
558 	/*
559 	 * Run through list of hooks for input packets.  If there are any
560 	 * filters which require that additional packets in the flow are
561 	 * not fast-forwarded, they must clear the M_CANFASTFWD flag.
562 	 * Note that filters must _never_ set this flag, as another filter
563 	 * in the list may have previously cleared it.
564 	 *
565 	 * Don't call hooks if the packet has already been processed by
566 	 * IPsec (encapsulated, tunnel mode).
567 	 */
568 #if defined(IPSEC)
569 	if (!ipsec_used || !ipsec_skip_pfil(m))
570 #else
571 	if (1)
572 #endif
573 	{
574 		struct in_addr odst = ip->ip_dst;
575 		bool freed;
576 
577 		freed = pfil_run_hooks(inet_pfil_hook, &m, ifp, PFIL_IN) != 0;
578 		if (freed || m == NULL) {
579 			m = NULL;
580 			IP_STATINC(IP_STAT_PFILDROP_IN);
581 			goto out;
582 		}
583 		if (__predict_false(m->m_len < sizeof(struct ip))) {
584 			if ((m = m_pullup(m, sizeof(struct ip))) == NULL) {
585 				IP_STATINC(IP_STAT_TOOSMALL);
586 				goto out;
587 			}
588 		}
589 		ip = mtod(m, struct ip *);
590 		hlen = ip->ip_hl << 2;
591 		if (hlen < sizeof(struct ip)) {	/* minimum header length */
592 			IP_STATINC(IP_STAT_BADHLEN);
593 			goto out;
594 		}
595 		if (hlen > m->m_len) {
596 			if ((m = m_pullup(m, hlen)) == NULL) {
597 				IP_STATINC(IP_STAT_BADHLEN);
598 				goto out;
599 			}
600 			ip = mtod(m, struct ip *);
601 		}
602 
603 		/*
604 		 * XXX The setting of "srcrt" here is to prevent ip_forward()
605 		 * from generating ICMP redirects for packets that have
606 		 * been redirected by a hook back out on to the same LAN that
607 		 * they came from and is not an indication that the packet
608 		 * is being influenced by source routing options.  This
609 		 * allows things like
610 		 * "rdr tlp0 0/0 port 80 -> 1.1.1.200 3128 tcp"
611 		 * where tlp0 is both on the 1.1.1.0/24 network and is the
612 		 * default route for hosts on 1.1.1.0/24.  Of course this
613 		 * also requires a "map tlp0 ..." to complete the story.
614 		 * One might argue whether or not this kind of network config.
615 		 * should be supported in this manner...
616 		 */
617 		srcrt = (odst.s_addr != ip->ip_dst.s_addr);
618 	}
619 
620 #ifdef ALTQ
621 	/* XXX Temporary until ALTQ is changed to use a pfil hook */
622 	if (altq_input) {
623 		SOFTNET_LOCK();
624 		if ((*altq_input)(m, AF_INET) == 0) {
625 			/* Packet dropped by traffic conditioner. */
626 			SOFTNET_UNLOCK();
627 			m = NULL;
628 			goto out;
629 		}
630 		SOFTNET_UNLOCK();
631 	}
632 #endif
633 
634 	/*
635 	 * Process options and, if not destined for us,
636 	 * ship it on.  ip_dooptions returns 1 when an
637 	 * error was detected (causing an icmp message
638 	 * to be sent and the original packet to be freed).
639 	 */
640 	if (hlen > sizeof(struct ip) && ip_dooptions(m)) {
641 		m = NULL;
642 		goto out;
643 	}
644 
645 	/*
646 	 * Check our list of addresses, to see if the packet is for us.
647 	 *
648 	 * Traditional 4.4BSD did not consult IFF_UP at all.
649 	 * The behavior here is to treat addresses on !IFF_UP interface
650 	 * or IN_IFF_NOTREADY addresses as not mine.
651 	 */
652 	downmatch = 0;
653 	s = pserialize_read_enter();
654 	ia = ip_match_our_address(ifp, ip, &downmatch);
655 	if (ia != NULL) {
656 		pserialize_read_exit(s);
657 		goto ours;
658 	}
659 
660 	if (ifp->if_flags & IFF_BROADCAST) {
661 		ia = ip_match_our_address_broadcast(ifp, ip);
662 		if (ia != NULL) {
663 			pserialize_read_exit(s);
664 			goto ours;
665 		}
666 	}
667 	pserialize_read_exit(s);
668 
669 	if (IN_MULTICAST(ip->ip_dst.s_addr)) {
670 #ifdef MROUTING
671 		extern struct socket *ip_mrouter;
672 
673 		if (ip_mrouter) {
674 			/*
675 			 * If we are acting as a multicast router, all
676 			 * incoming multicast packets are passed to the
677 			 * kernel-level multicast forwarding function.
678 			 * The packet is returned (relatively) intact; if
679 			 * ip_mforward() returns a non-zero value, the packet
680 			 * must be discarded, else it may be accepted below.
681 			 *
682 			 * (The IP ident field is put in the same byte order
683 			 * as expected when ip_mforward() is called from
684 			 * ip_output().)
685 			 */
686 			SOFTNET_LOCK();
687 			if (ip_mforward(m, ifp) != 0) {
688 				SOFTNET_UNLOCK();
689 				IP_STATINC(IP_STAT_CANTFORWARD);
690 				goto out;
691 			}
692 			SOFTNET_UNLOCK();
693 
694 			/*
695 			 * The process-level routing demon needs to receive
696 			 * all multicast IGMP packets, whether or not this
697 			 * host belongs to their destination groups.
698 			 */
699 			if (ip->ip_p == IPPROTO_IGMP) {
700 				goto ours;
701 			}
702 			IP_STATINC(IP_STAT_CANTFORWARD);
703 		}
704 #endif
705 		/*
706 		 * See if we belong to the destination multicast group on the
707 		 * arrival interface.
708 		 */
709 		if (!in_multi_group(ip->ip_dst, ifp, 0)) {
710 			IP_STATINC(IP_STAT_CANTFORWARD);
711 			goto out;
712 		}
713 		goto ours;
714 	}
715 	if (ip->ip_dst.s_addr == INADDR_BROADCAST ||
716 	    in_nullhost(ip->ip_dst))
717 		goto ours;
718 
719 	/*
720 	 * Not for us; forward if possible and desirable.
721 	 */
722 	if (ipforwarding == 0) {
723 		IP_STATINC(IP_STAT_CANTFORWARD);
724 		m_freem(m);
725 	} else {
726 		/*
727 		 * If ip_dst matched any of my address on !IFF_UP interface,
728 		 * and there's no IFF_UP interface that matches ip_dst,
729 		 * send icmp unreach.  Forwarding it will result in in-kernel
730 		 * forwarding loop till TTL goes to 0.
731 		 */
732 		if (downmatch) {
733 			icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_HOST, 0, 0);
734 			IP_STATINC(IP_STAT_CANTFORWARD);
735 			return;
736 		}
737 #ifdef IPSEC
738 		/* Check the security policy (SP) for the packet */
739 		if (ipsec_used) {
740 			if (ipsec_ip_input_checkpolicy(m, true) != 0) {
741 				IP_STATINC(IP_STAT_IPSECDROP_IN);
742 				goto out;
743 			}
744 		}
745 #endif
746 		ip_forward(m, srcrt, ifp);
747 	}
748 	return;
749 
750 ours:
751 	/*
752 	 * If offset or IP_MF are set, must reassemble.
753 	 */
754 	if (ip->ip_off & ~htons(IP_DF|IP_RF)) {
755 		/*
756 		 * Pass to IP reassembly mechanism.
757 		 */
758 		if (ip_reass_packet(&m) != 0) {
759 			/* Failed; invalid fragment(s) or packet. */
760 			goto out;
761 		}
762 		if (m == NULL) {
763 			/* More fragments should come; silently return. */
764 			goto out;
765 		}
766 		/*
767 		 * Reassembly is done, we have the final packet.
768 		 * Update cached data in local variable(s).
769 		 */
770 		ip = mtod(m, struct ip *);
771 		hlen = ip->ip_hl << 2;
772 	}
773 
774 	M_VERIFY_PACKET(m);
775 
776 #ifdef IPSEC
777 	/*
778 	 * Enforce IPsec policy checking if we are seeing last header.
779 	 * Note that we do not visit this with protocols with PCB layer
780 	 * code - like UDP/TCP/raw IP.
781 	 */
782 	if (ipsec_used &&
783 	    (inetsw[ip_protox[ip->ip_p]].pr_flags & PR_LASTHDR) != 0) {
784 		if (ipsec_ip_input_checkpolicy(m, false) != 0) {
785 			IP_STATINC(IP_STAT_IPSECDROP_IN);
786 			goto out;
787 		}
788 	}
789 #endif
790 
791 	/*
792 	 * Switch out to protocol's input routine.
793 	 */
794 #if IFA_STATS
795 	if (ia) {
796 		struct in_ifaddr *_ia;
797 		/*
798 		 * Keep a reference from ip_match_our_address with psref
799 		 * is expensive, so explore ia here again.
800 		 */
801 		s = pserialize_read_enter();
802 		_ia = in_get_ia(ip->ip_dst);
803 		_ia->ia_ifa.ifa_data.ifad_inbytes += ntohs(ip->ip_len);
804 		pserialize_read_exit(s);
805 	}
806 #endif
807 	IP_STATINC(IP_STAT_DELIVERED);
808 
809 	const int off = hlen, nh = ip->ip_p;
810 
811 	(*inetsw[ip_protox[nh]].pr_input)(m, off, nh);
812 	return;
813 
814 out:
815 	if (m != NULL)
816 		m_freem(m);
817 }
818 
819 /*
820  * IP timer processing.
821  */
822 void
ip_slowtimo(void)823 ip_slowtimo(void)
824 {
825 
826 	SOFTNET_KERNEL_LOCK_UNLESS_NET_MPSAFE();
827 
828 	ip_reass_slowtimo();
829 
830 	SOFTNET_KERNEL_UNLOCK_UNLESS_NET_MPSAFE();
831 }
832 
833 /*
834  * IP drain processing.
835  */
836 void
ip_drain(void)837 ip_drain(void)
838 {
839 
840 	KERNEL_LOCK(1, NULL);
841 	ip_reass_drain();
842 	KERNEL_UNLOCK_ONE(NULL);
843 }
844 
845 /*
846  * ip_dooptions: perform option processing on a datagram, possibly discarding
847  * it if bad options are encountered, or forwarding it if source-routed.
848  *
849  * => Returns true if packet has been forwarded/freed.
850  * => Returns false if the packet should be processed further.
851  */
852 static bool
ip_dooptions(struct mbuf * m)853 ip_dooptions(struct mbuf *m)
854 {
855 	struct ip *ip = mtod(m, struct ip *);
856 	u_char *cp, *cp0;
857 	struct ip_timestamp *ipt;
858 	struct in_ifaddr *ia;
859 	int opt, optlen, cnt, off, code, type = ICMP_PARAMPROB, forward = 0;
860 	int srr_present, rr_present, ts_present;
861 	struct in_addr dst;
862 	n_time ntime;
863 	struct ifaddr *ifa = NULL;
864 	int s;
865 
866 	srr_present = 0;
867 	rr_present = 0;
868 	ts_present = 0;
869 
870 	dst = ip->ip_dst;
871 	cp = (u_char *)(ip + 1);
872 	cnt = (ip->ip_hl << 2) - sizeof(struct ip);
873 	for (; cnt > 0; cnt -= optlen, cp += optlen) {
874 		opt = cp[IPOPT_OPTVAL];
875 		if (opt == IPOPT_EOL)
876 			break;
877 		if (opt == IPOPT_NOP)
878 			optlen = 1;
879 		else {
880 			if (cnt < IPOPT_OLEN + sizeof(*cp)) {
881 				code = &cp[IPOPT_OLEN] - (u_char *)ip;
882 				goto bad;
883 			}
884 			optlen = cp[IPOPT_OLEN];
885 			if (optlen < IPOPT_OLEN + sizeof(*cp) || optlen > cnt) {
886 				code = &cp[IPOPT_OLEN] - (u_char *)ip;
887 				goto bad;
888 			}
889 		}
890 		switch (opt) {
891 
892 		default:
893 			break;
894 
895 		/*
896 		 * Source routing with record.
897 		 * Find interface with current destination address.
898 		 * If none on this machine then drop if strictly routed,
899 		 * or do nothing if loosely routed.
900 		 * Record interface address and bring up next address
901 		 * component.  If strictly routed make sure next
902 		 * address is on directly accessible net.
903 		 */
904 		case IPOPT_LSRR:
905 		case IPOPT_SSRR: {
906 			struct psref psref;
907 			struct sockaddr_in ipaddr = {
908 			    .sin_len = sizeof(ipaddr),
909 			    .sin_family = AF_INET,
910 			};
911 
912 			if (ip_allowsrcrt == 0) {
913 				type = ICMP_UNREACH;
914 				code = ICMP_UNREACH_NET_PROHIB;
915 				goto bad;
916 			}
917 			if (srr_present++) {
918 				code = &cp[IPOPT_OPTVAL] - (u_char *)ip;
919 				goto bad;
920 			}
921 			if (optlen < IPOPT_OFFSET + sizeof(*cp)) {
922 				code = &cp[IPOPT_OLEN] - (u_char *)ip;
923 				goto bad;
924 			}
925 			if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) {
926 				code = &cp[IPOPT_OFFSET] - (u_char *)ip;
927 				goto bad;
928 			}
929 			ipaddr.sin_addr = ip->ip_dst;
930 
931 			s = pserialize_read_enter();
932 			ifa = ifa_ifwithaddr(sintosa(&ipaddr));
933 			if (ifa == NULL) {
934 				pserialize_read_exit(s);
935 				if (opt == IPOPT_SSRR) {
936 					type = ICMP_UNREACH;
937 					code = ICMP_UNREACH_SRCFAIL;
938 					goto bad;
939 				}
940 				/*
941 				 * Loose routing, and not at next destination
942 				 * yet; nothing to do except forward.
943 				 */
944 				break;
945 			}
946 			pserialize_read_exit(s);
947 
948 			off--;			/* 0 origin */
949 			if ((off + sizeof(struct in_addr)) > optlen) {
950 				/*
951 				 * End of source route.  Should be for us.
952 				 */
953 				save_rte(m, cp, ip->ip_src);
954 				break;
955 			}
956 			/*
957 			 * locate outgoing interface
958 			 */
959 			memcpy((void *)&ipaddr.sin_addr, (void *)(cp + off),
960 			    sizeof(ipaddr.sin_addr));
961 			if (opt == IPOPT_SSRR) {
962 				ifa = ifa_ifwithladdr_psref(sintosa(&ipaddr),
963 				    &psref);
964 				if (ifa != NULL)
965 					ia = ifatoia(ifa);
966 				else
967 					ia = NULL;
968 			} else {
969 				ia = ip_rtaddr(ipaddr.sin_addr, &psref);
970 			}
971 			if (ia == NULL) {
972 				type = ICMP_UNREACH;
973 				code = ICMP_UNREACH_SRCFAIL;
974 				goto bad;
975 			}
976 			ip->ip_dst = ipaddr.sin_addr;
977 			memcpy(cp + off, &ia->ia_addr.sin_addr,
978 			    sizeof(struct in_addr));
979 			ia4_release(ia, &psref);
980 			cp[IPOPT_OFFSET] += sizeof(struct in_addr);
981 			/*
982 			 * Let ip_intr's mcast routing check handle mcast pkts
983 			 */
984 			forward = !IN_MULTICAST(ip->ip_dst.s_addr);
985 			break;
986 		    }
987 
988 		case IPOPT_RR: {
989 			struct psref psref;
990 			struct sockaddr_in ipaddr = {
991 			    .sin_len = sizeof(ipaddr),
992 			    .sin_family = AF_INET,
993 			};
994 
995 			if (rr_present++) {
996 				code = &cp[IPOPT_OPTVAL] - (u_char *)ip;
997 				goto bad;
998 			}
999 			if (optlen < IPOPT_OFFSET + sizeof(*cp)) {
1000 				code = &cp[IPOPT_OLEN] - (u_char *)ip;
1001 				goto bad;
1002 			}
1003 			if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) {
1004 				code = &cp[IPOPT_OFFSET] - (u_char *)ip;
1005 				goto bad;
1006 			}
1007 			/*
1008 			 * If no space remains, ignore.
1009 			 */
1010 			off--;			/* 0 origin */
1011 			if ((off + sizeof(struct in_addr)) > optlen)
1012 				break;
1013 			memcpy((void *)&ipaddr.sin_addr, (void *)&ip->ip_dst,
1014 			    sizeof(ipaddr.sin_addr));
1015 			/*
1016 			 * locate outgoing interface; if we're the destination,
1017 			 * use the incoming interface (should be same).
1018 			 */
1019 			ifa = ifa_ifwithaddr_psref(sintosa(&ipaddr), &psref);
1020 			if (ifa == NULL) {
1021 				ia = ip_rtaddr(ipaddr.sin_addr, &psref);
1022 				if (ia == NULL) {
1023 					type = ICMP_UNREACH;
1024 					code = ICMP_UNREACH_HOST;
1025 					goto bad;
1026 				}
1027 			} else {
1028 				ia = ifatoia(ifa);
1029 			}
1030 			memcpy(cp + off, &ia->ia_addr.sin_addr,
1031 			    sizeof(struct in_addr));
1032 			ia4_release(ia, &psref);
1033 			cp[IPOPT_OFFSET] += sizeof(struct in_addr);
1034 			break;
1035 		    }
1036 
1037 		case IPOPT_TS:
1038 			code = cp - (u_char *)ip;
1039 			ipt = (struct ip_timestamp *)cp;
1040 			if (ts_present++) {
1041 				code = &cp[IPOPT_OPTVAL] - (u_char *)ip;
1042 				goto bad;
1043 			}
1044 			if (ipt->ipt_len < 4 || ipt->ipt_len > 40) {
1045 				code = (u_char *)&ipt->ipt_len - (u_char *)ip;
1046 				goto bad;
1047 			}
1048 			if (ipt->ipt_ptr < 5) {
1049 				code = (u_char *)&ipt->ipt_ptr - (u_char *)ip;
1050 				goto bad;
1051 			}
1052 			if (ipt->ipt_ptr > ipt->ipt_len - sizeof(int32_t)) {
1053 				if (++ipt->ipt_oflw == 0) {
1054 					code = (u_char *)&ipt->ipt_ptr -
1055 					    (u_char *)ip;
1056 					goto bad;
1057 				}
1058 				break;
1059 			}
1060 			cp0 = (cp + ipt->ipt_ptr - 1);
1061 			switch (ipt->ipt_flg) {
1062 
1063 			case IPOPT_TS_TSONLY:
1064 				break;
1065 
1066 			case IPOPT_TS_TSANDADDR: {
1067 				struct ifnet *rcvif;
1068 				int _s, _ss;
1069 				struct sockaddr_in ipaddr = {
1070 				    .sin_len = sizeof(ipaddr),
1071 				    .sin_family = AF_INET,
1072 				};
1073 
1074 				if (ipt->ipt_ptr - 1 + sizeof(n_time) +
1075 				    sizeof(struct in_addr) > ipt->ipt_len) {
1076 					code = (u_char *)&ipt->ipt_ptr -
1077 					    (u_char *)ip;
1078 					goto bad;
1079 				}
1080 				ipaddr.sin_addr = dst;
1081 				_ss = pserialize_read_enter();
1082 				rcvif = m_get_rcvif(m, &_s);
1083 				if (__predict_true(rcvif != NULL)) {
1084 					ifa = ifaof_ifpforaddr(sintosa(&ipaddr),
1085 					    rcvif);
1086 				}
1087 				m_put_rcvif(rcvif, &_s);
1088 				if (ifa == NULL) {
1089 					pserialize_read_exit(_ss);
1090 					break;
1091 				}
1092 				ia = ifatoia(ifa);
1093 				memcpy(cp0, &ia->ia_addr.sin_addr,
1094 				    sizeof(struct in_addr));
1095 				pserialize_read_exit(_ss);
1096 				ipt->ipt_ptr += sizeof(struct in_addr);
1097 				break;
1098 			}
1099 
1100 			case IPOPT_TS_PRESPEC: {
1101 				struct sockaddr_in ipaddr = {
1102 				    .sin_len = sizeof(ipaddr),
1103 				    .sin_family = AF_INET,
1104 				};
1105 
1106 				if (ipt->ipt_ptr - 1 + sizeof(n_time) +
1107 				    sizeof(struct in_addr) > ipt->ipt_len) {
1108 					code = (u_char *)&ipt->ipt_ptr -
1109 					    (u_char *)ip;
1110 					goto bad;
1111 				}
1112 				memcpy(&ipaddr.sin_addr, cp0,
1113 				    sizeof(struct in_addr));
1114 				s = pserialize_read_enter();
1115 				ifa = ifa_ifwithaddr(sintosa(&ipaddr));
1116 				if (ifa == NULL) {
1117 					pserialize_read_exit(s);
1118 					continue;
1119 				}
1120 				pserialize_read_exit(s);
1121 				ipt->ipt_ptr += sizeof(struct in_addr);
1122 				break;
1123 			    }
1124 
1125 			default:
1126 				/* XXX can't take &ipt->ipt_flg */
1127 				code = (u_char *)&ipt->ipt_ptr -
1128 				    (u_char *)ip + 1;
1129 				goto bad;
1130 			}
1131 			ntime = iptime();
1132 			cp0 = (u_char *) &ntime; /* XXX grumble, GCC... */
1133 			memmove((char *)cp + ipt->ipt_ptr - 1, cp0,
1134 			    sizeof(n_time));
1135 			ipt->ipt_ptr += sizeof(n_time);
1136 		}
1137 	}
1138 	if (forward) {
1139 		struct ifnet *rcvif;
1140 		struct psref _psref;
1141 
1142 		if (ip_forwsrcrt == 0) {
1143 			type = ICMP_UNREACH;
1144 			code = ICMP_UNREACH_SRCFAIL;
1145 			goto bad;
1146 		}
1147 
1148 		rcvif = m_get_rcvif_psref(m, &_psref);
1149 		if (__predict_false(rcvif == NULL)) {
1150 			type = ICMP_UNREACH;
1151 			code = ICMP_UNREACH_HOST;
1152 			goto bad;
1153 		}
1154 		ip_forward(m, 1, rcvif);
1155 		m_put_rcvif_psref(rcvif, &_psref);
1156 		return true;
1157 	}
1158 	return false;
1159 bad:
1160 	icmp_error(m, type, code, 0, 0);
1161 	IP_STATINC(IP_STAT_BADOPTIONS);
1162 	return true;
1163 }
1164 
1165 /*
1166  * ip_rtaddr: given address of next destination (final or next hop),
1167  * return internet address info of interface to be used to get there.
1168  */
1169 static struct in_ifaddr *
ip_rtaddr(struct in_addr dst,struct psref * psref)1170 ip_rtaddr(struct in_addr dst, struct psref *psref)
1171 {
1172 	struct rtentry *rt;
1173 	union {
1174 		struct sockaddr		dst;
1175 		struct sockaddr_in	dst4;
1176 	} u;
1177 	struct route *ro;
1178 
1179 	sockaddr_in_init(&u.dst4, &dst, 0);
1180 
1181 	ro = rtcache_percpu_getref(ipforward_rt_percpu);
1182 	rt = rtcache_lookup(ro, &u.dst);
1183 	if (rt == NULL) {
1184 		rtcache_percpu_putref(ipforward_rt_percpu);
1185 		return NULL;
1186 	}
1187 
1188 	ia4_acquire(ifatoia(rt->rt_ifa), psref);
1189 	rtcache_unref(rt, ro);
1190 	rtcache_percpu_putref(ipforward_rt_percpu);
1191 
1192 	return ifatoia(rt->rt_ifa);
1193 }
1194 
1195 /*
1196  * save_rte: save incoming source route for use in replies, to be picked
1197  * up later by ip_srcroute if the receiver is interested.
1198  */
1199 static void
save_rte(struct mbuf * m,u_char * option,struct in_addr dst)1200 save_rte(struct mbuf *m, u_char *option, struct in_addr dst)
1201 {
1202 	struct ip_srcrt *isr;
1203 	struct m_tag *mtag;
1204 	unsigned olen;
1205 
1206 	olen = option[IPOPT_OLEN];
1207 	if (olen > sizeof(isr->isr_hdr) + sizeof(isr->isr_routes))
1208 		return;
1209 
1210 	mtag = m_tag_get(PACKET_TAG_SRCROUTE, sizeof(*isr), M_NOWAIT);
1211 	if (mtag == NULL)
1212 		return;
1213 	isr = (struct ip_srcrt *)(mtag + 1);
1214 
1215 	memcpy(isr->isr_hdr, option, olen);
1216 	isr->isr_nhops = (olen - IPOPT_OFFSET - 1) / sizeof(struct in_addr);
1217 	isr->isr_dst = dst;
1218 	m_tag_prepend(m, mtag);
1219 }
1220 
1221 /*
1222  * Retrieve incoming source route for use in replies,
1223  * in the same form used by setsockopt.
1224  * The first hop is placed before the options, will be removed later.
1225  */
1226 struct mbuf *
ip_srcroute(struct mbuf * m0)1227 ip_srcroute(struct mbuf *m0)
1228 {
1229 	struct in_addr *p, *q;
1230 	struct mbuf *m;
1231 	struct ip_srcrt *isr;
1232 	struct m_tag *mtag;
1233 
1234 	mtag = m_tag_find(m0, PACKET_TAG_SRCROUTE);
1235 	if (mtag == NULL)
1236 		return NULL;
1237 	isr = (struct ip_srcrt *)(mtag + 1);
1238 
1239 	if (isr->isr_nhops == 0)
1240 		return NULL;
1241 
1242 	m = m_get(M_DONTWAIT, MT_SOOPTS);
1243 	if (m == NULL)
1244 		return NULL;
1245 
1246 	MCLAIM(m, &inetdomain.dom_mowner);
1247 #define OPTSIZ	(sizeof(isr->isr_nop) + sizeof(isr->isr_hdr))
1248 
1249 	/* length is (nhops+1)*sizeof(addr) + sizeof(nop + header) */
1250 	m->m_len = (isr->isr_nhops + 1) * sizeof(struct in_addr) + OPTSIZ;
1251 
1252 	/*
1253 	 * First save first hop for return route
1254 	 */
1255 	p = &(isr->isr_routes[isr->isr_nhops - 1]);
1256 	*(mtod(m, struct in_addr *)) = *p--;
1257 
1258 	/*
1259 	 * Copy option fields and padding (nop) to mbuf.
1260 	 */
1261 	isr->isr_nop = IPOPT_NOP;
1262 	isr->isr_hdr[IPOPT_OFFSET] = IPOPT_MINOFF;
1263 	memmove(mtod(m, char *) + sizeof(struct in_addr), &isr->isr_nop,
1264 	    OPTSIZ);
1265 	q = (struct in_addr *)(mtod(m, char *) +
1266 	    sizeof(struct in_addr) + OPTSIZ);
1267 #undef OPTSIZ
1268 	/*
1269 	 * Record return path as an IP source route,
1270 	 * reversing the path (pointers are now aligned).
1271 	 */
1272 	while (p >= isr->isr_routes) {
1273 		*q++ = *p--;
1274 	}
1275 	/*
1276 	 * Last hop goes to final destination.
1277 	 */
1278 	*q = isr->isr_dst;
1279 	m_tag_delete(m0, mtag);
1280 	return m;
1281 }
1282 
1283 const int inetctlerrmap[PRC_NCMDS] = {
1284 	[PRC_MSGSIZE] = EMSGSIZE,
1285 	[PRC_HOSTDEAD] = EHOSTDOWN,
1286 	[PRC_HOSTUNREACH] = EHOSTUNREACH,
1287 	[PRC_UNREACH_NET] = EHOSTUNREACH,
1288 	[PRC_UNREACH_HOST] = EHOSTUNREACH,
1289 	[PRC_UNREACH_PROTOCOL] = ECONNREFUSED,
1290 	[PRC_UNREACH_PORT] = ECONNREFUSED,
1291 	[PRC_UNREACH_SRCFAIL] = EHOSTUNREACH,
1292 	[PRC_PARAMPROB] = ENOPROTOOPT,
1293 };
1294 
1295 void
ip_fasttimo(void)1296 ip_fasttimo(void)
1297 {
1298 	if (ip_drainwanted) {
1299 		ip_drain();
1300 		ip_drainwanted = 0;
1301 	}
1302 }
1303 
1304 void
ip_drainstub(void)1305 ip_drainstub(void)
1306 {
1307 	ip_drainwanted = 1;
1308 }
1309 
1310 /*
1311  * Forward a packet.  If some error occurs return the sender
1312  * an icmp packet.  Note we can't always generate a meaningful
1313  * icmp message because icmp doesn't have a large enough repertoire
1314  * of codes and types.
1315  *
1316  * If not forwarding, just drop the packet.  This could be confusing
1317  * if ipforwarding was zero but some routing protocol was advancing
1318  * us as a gateway to somewhere.  However, we must let the routing
1319  * protocol deal with that.
1320  *
1321  * The srcrt parameter indicates whether the packet is being forwarded
1322  * via a source route.
1323  */
1324 static void
ip_forward(struct mbuf * m,int srcrt,struct ifnet * rcvif)1325 ip_forward(struct mbuf *m, int srcrt, struct ifnet *rcvif)
1326 {
1327 	struct ip *ip = mtod(m, struct ip *);
1328 	struct rtentry *rt;
1329 	int error, type = 0, code = 0, destmtu = 0;
1330 	struct mbuf *mcopy;
1331 	n_long dest;
1332 	union {
1333 		struct sockaddr		dst;
1334 		struct sockaddr_in	dst4;
1335 	} u;
1336 	uint64_t *ips;
1337 	struct route *ro;
1338 
1339 	KASSERTMSG(cpu_softintr_p(), "ip_forward: not in the software "
1340 	    "interrupt handler; synchronization assumptions violated");
1341 
1342 	/*
1343 	 * We are now in the output path.
1344 	 */
1345 	MCLAIM(m, &ip_tx_mowner);
1346 
1347 	/*
1348 	 * Clear any in-bound checksum flags for this packet.
1349 	 */
1350 	m->m_pkthdr.csum_flags = 0;
1351 
1352 	dest = 0;
1353 	if (m->m_flags & (M_BCAST|M_MCAST) || in_canforward(ip->ip_dst) == 0) {
1354 		IP_STATINC(IP_STAT_CANTFORWARD);
1355 		m_freem(m);
1356 		return;
1357 	}
1358 
1359 	if (ip->ip_ttl <= IPTTLDEC) {
1360 		IP_STATINC(IP_STAT_TIMXCEED);
1361 		icmp_error(m, ICMP_TIMXCEED, ICMP_TIMXCEED_INTRANS, dest, 0);
1362 		return;
1363 	}
1364 
1365 	sockaddr_in_init(&u.dst4, &ip->ip_dst, 0);
1366 
1367 	ro = rtcache_percpu_getref(ipforward_rt_percpu);
1368 	rt = rtcache_lookup(ro, &u.dst);
1369 	if (rt == NULL) {
1370 		rtcache_percpu_putref(ipforward_rt_percpu);
1371 		IP_STATINC(IP_STAT_NOROUTE);
1372 		icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_NET, dest, 0);
1373 		return;
1374 	}
1375 
1376 	/*
1377 	 * Save at most 68 bytes of the packet in case
1378 	 * we need to generate an ICMP message to the src.
1379 	 * Pullup to avoid sharing mbuf cluster between m and mcopy.
1380 	 */
1381 	mcopy = m_copym(m, 0, imin(ntohs(ip->ip_len), 68), M_DONTWAIT);
1382 	if (mcopy)
1383 		mcopy = m_pullup(mcopy, ip->ip_hl << 2);
1384 
1385 	ip->ip_ttl -= IPTTLDEC;
1386 
1387 	/*
1388 	 * If forwarding packet using same interface that it came in on,
1389 	 * perhaps should send a redirect to sender to shortcut a hop.
1390 	 * Only send redirect if source is sending directly to us,
1391 	 * and if packet was not source routed (or has any options).
1392 	 * Also, don't send redirect if forwarding using a default route
1393 	 * or a route modified by a redirect.
1394 	 */
1395 	if (rt->rt_ifp == rcvif &&
1396 	    (rt->rt_flags & (RTF_DYNAMIC|RTF_MODIFIED)) == 0 &&
1397 	    !in_nullhost(satocsin(rt_getkey(rt))->sin_addr) &&
1398 	    ipsendredirects && !srcrt) {
1399 		if ((ip->ip_src.s_addr & ifatoia(rt->rt_ifa)->ia_subnetmask) ==
1400 		    ifatoia(rt->rt_ifa)->ia_subnet) {
1401 			if (rt->rt_flags & RTF_GATEWAY)
1402 				dest = satosin(rt->rt_gateway)->sin_addr.s_addr;
1403 			else
1404 				dest = ip->ip_dst.s_addr;
1405 			/*
1406 			 * Router requirements says to only send host
1407 			 * redirects.
1408 			 */
1409 			type = ICMP_REDIRECT;
1410 			code = ICMP_REDIRECT_HOST;
1411 		}
1412 	}
1413 	rtcache_unref(rt, ro);
1414 
1415 	error = ip_output(m, NULL, ro,
1416 	    (IP_FORWARDING | (ip_directedbcast ? IP_ALLOWBROADCAST : 0)),
1417 	    NULL, NULL);
1418 
1419 	if (error) {
1420 		IP_STATINC(IP_STAT_CANTFORWARD);
1421 		goto error;
1422 	}
1423 
1424 	ips = IP_STAT_GETREF();
1425 	ips[IP_STAT_FORWARD]++;
1426 
1427 	if (type) {
1428 		ips[IP_STAT_REDIRECTSENT]++;
1429 		IP_STAT_PUTREF();
1430 		goto redirect;
1431 	}
1432 
1433 	IP_STAT_PUTREF();
1434 	if (mcopy) {
1435 #ifdef GATEWAY
1436 		if (mcopy->m_flags & M_CANFASTFWD)
1437 			ipflow_create(ro, mcopy);
1438 #endif
1439 		m_freem(mcopy);
1440 	}
1441 
1442 	rtcache_percpu_putref(ipforward_rt_percpu);
1443 	return;
1444 
1445 redirect:
1446 error:
1447 	if (mcopy == NULL) {
1448 		rtcache_percpu_putref(ipforward_rt_percpu);
1449 		return;
1450 	}
1451 
1452 	switch (error) {
1453 
1454 	case 0:				/* forwarded, but need redirect */
1455 		/* type, code set above */
1456 		break;
1457 
1458 	case ENETUNREACH:		/* shouldn't happen, checked above */
1459 	case EHOSTUNREACH:
1460 	case ENETDOWN:
1461 	case EHOSTDOWN:
1462 	default:
1463 		type = ICMP_UNREACH;
1464 		code = ICMP_UNREACH_HOST;
1465 		break;
1466 
1467 	case EMSGSIZE:
1468 		type = ICMP_UNREACH;
1469 		code = ICMP_UNREACH_NEEDFRAG;
1470 
1471 		if ((rt = rtcache_validate(ro)) != NULL) {
1472 			destmtu = rt->rt_ifp->if_mtu;
1473 			rtcache_unref(rt, ro);
1474 		}
1475 #ifdef IPSEC
1476 		if (ipsec_used)
1477 			ipsec_mtu(mcopy, &destmtu);
1478 #endif
1479 		IP_STATINC(IP_STAT_CANTFRAG);
1480 		break;
1481 
1482 	case ENOBUFS:
1483 		/*
1484 		 * Do not generate ICMP_SOURCEQUENCH as required in RFC 1812,
1485 		 * Requirements for IP Version 4 Routers.  Source quench can
1486 		 * be a big problem under DoS attacks or if the underlying
1487 		 * interface is rate-limited.
1488 		 */
1489 		if (mcopy)
1490 			m_freem(mcopy);
1491 		rtcache_percpu_putref(ipforward_rt_percpu);
1492 		return;
1493 	}
1494 	icmp_error(mcopy, type, code, dest, destmtu);
1495 	rtcache_percpu_putref(ipforward_rt_percpu);
1496 }
1497 
1498 void
ip_savecontrol(struct inpcb * inp,struct mbuf ** mp,struct ip * ip,struct mbuf * m)1499 ip_savecontrol(struct inpcb *inp, struct mbuf **mp, struct ip *ip,
1500     struct mbuf *m)
1501 {
1502 	struct socket *so = inp->inp_socket;
1503 	int inpflags = inp->inp_flags;
1504 
1505 	if (SOOPT_TIMESTAMP(so->so_options))
1506 		mp = sbsavetimestamp(so->so_options, mp);
1507 
1508 	if (inpflags & INP_RECVDSTADDR) {
1509 		*mp = sbcreatecontrol(&ip->ip_dst,
1510 		    sizeof(struct in_addr), IP_RECVDSTADDR, IPPROTO_IP);
1511 		if (*mp)
1512 			mp = &(*mp)->m_next;
1513 	}
1514 
1515 	if (inpflags & INP_RECVTTL) {
1516 		*mp = sbcreatecontrol(&ip->ip_ttl,
1517 		    sizeof(uint8_t), IP_RECVTTL, IPPROTO_IP);
1518 		if (*mp)
1519 			mp = &(*mp)->m_next;
1520 	}
1521 
1522 	struct psref psref;
1523 	ifnet_t *ifp = m_get_rcvif_psref(m, &psref);
1524 	if (__predict_false(ifp == NULL)) {
1525 #ifdef DIAGNOSTIC
1526 		printf("%s: missing receive interface\n", __func__);
1527 #endif
1528 		return; /* XXX should report error? */
1529 	}
1530 
1531 	if (inpflags & INP_RECVPKTINFO) {
1532 		struct in_pktinfo ipi;
1533 		ipi.ipi_addr = ip->ip_dst;
1534 		ipi.ipi_ifindex = ifp->if_index;
1535 		*mp = sbcreatecontrol(&ipi,
1536 		    sizeof(ipi), IP_PKTINFO, IPPROTO_IP);
1537 		if (*mp)
1538 			mp = &(*mp)->m_next;
1539 	}
1540 	if (inpflags & INP_RECVIF) {
1541 		struct sockaddr_dl sdl;
1542 
1543 		sockaddr_dl_init(&sdl, sizeof(sdl), ifp->if_index, 0, NULL, 0,
1544 		    NULL, 0);
1545 		*mp = sbcreatecontrol(&sdl, sdl.sdl_len, IP_RECVIF, IPPROTO_IP);
1546 		if (*mp)
1547 			mp = &(*mp)->m_next;
1548 	}
1549 	m_put_rcvif_psref(ifp, &psref);
1550 }
1551 
1552 /*
1553  * sysctl helper routine for net.inet.ip.forwsrcrt.
1554  */
1555 static int
sysctl_net_inet_ip_forwsrcrt(SYSCTLFN_ARGS)1556 sysctl_net_inet_ip_forwsrcrt(SYSCTLFN_ARGS)
1557 {
1558 	int error, tmp;
1559 	struct sysctlnode node;
1560 
1561 	node = *rnode;
1562 	tmp = ip_forwsrcrt;
1563 	node.sysctl_data = &tmp;
1564 	error = sysctl_lookup(SYSCTLFN_CALL(&node));
1565 	if (error || newp == NULL)
1566 		return (error);
1567 
1568 	error = kauth_authorize_network(l->l_cred, KAUTH_NETWORK_FORWSRCRT,
1569 	    0, NULL, NULL, NULL);
1570 	if (error)
1571 		return (error);
1572 
1573 	ip_forwsrcrt = tmp;
1574 
1575 	return (0);
1576 }
1577 
1578 /*
1579  * sysctl helper routine for net.inet.ip.mtudisctimeout.  checks the
1580  * range of the new value and tweaks timers if it changes.
1581  */
1582 static int
sysctl_net_inet_ip_pmtudto(SYSCTLFN_ARGS)1583 sysctl_net_inet_ip_pmtudto(SYSCTLFN_ARGS)
1584 {
1585 	int error, tmp;
1586 	struct sysctlnode node;
1587 
1588 	icmp_mtudisc_lock();
1589 
1590 	node = *rnode;
1591 	tmp = ip_mtudisc_timeout;
1592 	node.sysctl_data = &tmp;
1593 	error = sysctl_lookup(SYSCTLFN_CALL(&node));
1594 	if (error || newp == NULL)
1595 		goto out;
1596 	if (tmp < 0) {
1597 		error = EINVAL;
1598 		goto out;
1599 	}
1600 
1601 	ip_mtudisc_timeout = tmp;
1602 	rt_timer_queue_change(ip_mtudisc_timeout_q, ip_mtudisc_timeout);
1603 	error = 0;
1604 out:
1605 	icmp_mtudisc_unlock();
1606 	return error;
1607 }
1608 
1609 static int
sysctl_net_inet_ip_stats(SYSCTLFN_ARGS)1610 sysctl_net_inet_ip_stats(SYSCTLFN_ARGS)
1611 {
1612 
1613 	return (NETSTAT_SYSCTL(ipstat_percpu, IP_NSTATS));
1614 }
1615 
1616 static void
sysctl_net_inet_ip_setup(struct sysctllog ** clog)1617 sysctl_net_inet_ip_setup(struct sysctllog **clog)
1618 {
1619 	const struct sysctlnode *ip_node;
1620 
1621 	sysctl_createv(clog, 0, NULL, NULL,
1622 		       CTLFLAG_PERMANENT,
1623 		       CTLTYPE_NODE, "inet",
1624 		       SYSCTL_DESCR("PF_INET related settings"),
1625 		       NULL, 0, NULL, 0,
1626 		       CTL_NET, PF_INET, CTL_EOL);
1627 	sysctl_createv(clog, 0, NULL, &ip_node,
1628 		       CTLFLAG_PERMANENT,
1629 		       CTLTYPE_NODE, "ip",
1630 		       SYSCTL_DESCR("IPv4 related settings"),
1631 		       NULL, 0, NULL, 0,
1632 		       CTL_NET, PF_INET, IPPROTO_IP, CTL_EOL);
1633 
1634 	sysctl_createv(clog, 0, NULL, NULL,
1635 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1636 		       CTLTYPE_INT, "forwarding",
1637 		       SYSCTL_DESCR("Enable forwarding of INET datagrams"),
1638 		       NULL, 0, &ipforwarding, 0,
1639 		       CTL_NET, PF_INET, IPPROTO_IP,
1640 		       IPCTL_FORWARDING, CTL_EOL);
1641 	sysctl_createv(clog, 0, NULL, NULL,
1642 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1643 		       CTLTYPE_INT, "redirect",
1644 		       SYSCTL_DESCR("Enable sending of ICMP redirect messages"),
1645 		       NULL, 0, &ipsendredirects, 0,
1646 		       CTL_NET, PF_INET, IPPROTO_IP,
1647 		       IPCTL_SENDREDIRECTS, CTL_EOL);
1648 	sysctl_createv(clog, 0, NULL, NULL,
1649 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1650 		       CTLTYPE_INT, "ttl",
1651 		       SYSCTL_DESCR("Default TTL for an INET datagram"),
1652 		       NULL, 0, &ip_defttl, 0,
1653 		       CTL_NET, PF_INET, IPPROTO_IP,
1654 		       IPCTL_DEFTTL, CTL_EOL);
1655 	sysctl_createv(clog, 0, NULL, NULL,
1656 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1657 		       CTLTYPE_INT, "forwsrcrt",
1658 		       SYSCTL_DESCR("Enable forwarding of source-routed "
1659 				    "datagrams"),
1660 		       sysctl_net_inet_ip_forwsrcrt, 0, &ip_forwsrcrt, 0,
1661 		       CTL_NET, PF_INET, IPPROTO_IP,
1662 		       IPCTL_FORWSRCRT, CTL_EOL);
1663 	sysctl_createv(clog, 0, NULL, NULL,
1664 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1665 		       CTLTYPE_INT, "directed-broadcast",
1666 		       SYSCTL_DESCR("Enable forwarding of broadcast datagrams"),
1667 		       NULL, 0, &ip_directedbcast, 0,
1668 		       CTL_NET, PF_INET, IPPROTO_IP,
1669 		       IPCTL_DIRECTEDBCAST, CTL_EOL);
1670 	sysctl_createv(clog, 0, NULL, NULL,
1671 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1672 		       CTLTYPE_INT, "allowsrcrt",
1673 		       SYSCTL_DESCR("Accept source-routed datagrams"),
1674 		       NULL, 0, &ip_allowsrcrt, 0,
1675 		       CTL_NET, PF_INET, IPPROTO_IP,
1676 		       IPCTL_ALLOWSRCRT, CTL_EOL);
1677 
1678 	sysctl_createv(clog, 0, NULL, NULL,
1679 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1680 		       CTLTYPE_INT, "mtudisc",
1681 		       SYSCTL_DESCR("Use RFC1191 Path MTU Discovery"),
1682 		       NULL, 0, &ip_mtudisc, 0,
1683 		       CTL_NET, PF_INET, IPPROTO_IP,
1684 		       IPCTL_MTUDISC, CTL_EOL);
1685 	sysctl_createv(clog, 0, NULL, NULL,
1686 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1687 		       CTLTYPE_INT, "anonportmin",
1688 		       SYSCTL_DESCR("Lowest ephemeral port number to assign"),
1689 		       sysctl_net_inet_ip_ports, 0, &anonportmin, 0,
1690 		       CTL_NET, PF_INET, IPPROTO_IP,
1691 		       IPCTL_ANONPORTMIN, CTL_EOL);
1692 	sysctl_createv(clog, 0, NULL, NULL,
1693 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1694 		       CTLTYPE_INT, "anonportmax",
1695 		       SYSCTL_DESCR("Highest ephemeral port number to assign"),
1696 		       sysctl_net_inet_ip_ports, 0, &anonportmax, 0,
1697 		       CTL_NET, PF_INET, IPPROTO_IP,
1698 		       IPCTL_ANONPORTMAX, CTL_EOL);
1699 	sysctl_createv(clog, 0, NULL, NULL,
1700 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1701 		       CTLTYPE_INT, "mtudisctimeout",
1702 		       SYSCTL_DESCR("Lifetime of a Path MTU Discovered route"),
1703 		       sysctl_net_inet_ip_pmtudto, 0, (void *)&ip_mtudisc_timeout, 0,
1704 		       CTL_NET, PF_INET, IPPROTO_IP,
1705 		       IPCTL_MTUDISCTIMEOUT, CTL_EOL);
1706 #ifndef IPNOPRIVPORTS
1707 	sysctl_createv(clog, 0, NULL, NULL,
1708 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1709 		       CTLTYPE_INT, "lowportmin",
1710 		       SYSCTL_DESCR("Lowest privileged ephemeral port number "
1711 				    "to assign"),
1712 		       sysctl_net_inet_ip_ports, 0, &lowportmin, 0,
1713 		       CTL_NET, PF_INET, IPPROTO_IP,
1714 		       IPCTL_LOWPORTMIN, CTL_EOL);
1715 	sysctl_createv(clog, 0, NULL, NULL,
1716 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1717 		       CTLTYPE_INT, "lowportmax",
1718 		       SYSCTL_DESCR("Highest privileged ephemeral port number "
1719 				    "to assign"),
1720 		       sysctl_net_inet_ip_ports, 0, &lowportmax, 0,
1721 		       CTL_NET, PF_INET, IPPROTO_IP,
1722 		       IPCTL_LOWPORTMAX, CTL_EOL);
1723 #endif /* IPNOPRIVPORTS */
1724 #if NGRE > 0
1725 	sysctl_createv(clog, 0, NULL, NULL,
1726 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1727 		       CTLTYPE_INT, "grettl",
1728 		       SYSCTL_DESCR("Default TTL for a gre tunnel datagram"),
1729 		       NULL, 0, &ip_gre_ttl, 0,
1730 		       CTL_NET, PF_INET, IPPROTO_IP,
1731 		       IPCTL_GRE_TTL, CTL_EOL);
1732 #endif /* NGRE */
1733 	sysctl_createv(clog, 0, NULL, NULL,
1734 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1735 		       CTLTYPE_INT, "checkinterface",
1736 		       SYSCTL_DESCR("Enable receive side of Strong ES model "
1737 				    "from RFC1122"),
1738 		       NULL, 0, &ip_checkinterface, 0,
1739 		       CTL_NET, PF_INET, IPPROTO_IP,
1740 		       IPCTL_CHECKINTERFACE, CTL_EOL);
1741 
1742 	pktq_sysctl_setup(ip_pktq, clog, ip_node, IPCTL_IFQ);
1743 
1744 	sysctl_createv(clog, 0, NULL, NULL,
1745 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1746 		       CTLTYPE_INT, "random_id",
1747 		       SYSCTL_DESCR("Assign random ip_id values"),
1748 		       NULL, 0, &ip_do_randomid, 0,
1749 		       CTL_NET, PF_INET, IPPROTO_IP,
1750 		       IPCTL_RANDOMID, CTL_EOL);
1751 	sysctl_createv(clog, 0, NULL, NULL,
1752 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1753 		       CTLTYPE_INT, "do_loopback_cksum",
1754 		       SYSCTL_DESCR("Perform IP checksum on loopback"),
1755 		       NULL, 0, &ip_do_loopback_cksum, 0,
1756 		       CTL_NET, PF_INET, IPPROTO_IP,
1757 		       IPCTL_LOOPBACKCKSUM, CTL_EOL);
1758 	sysctl_createv(clog, 0, NULL, NULL,
1759 		       CTLFLAG_PERMANENT,
1760 		       CTLTYPE_STRUCT, "stats",
1761 		       SYSCTL_DESCR("IP statistics"),
1762 		       sysctl_net_inet_ip_stats, 0, NULL, 0,
1763 		       CTL_NET, PF_INET, IPPROTO_IP, IPCTL_STATS,
1764 		       CTL_EOL);
1765 #if NARP
1766 	sysctl_createv(clog, 0, NULL, NULL,
1767 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1768 		       CTLTYPE_INT, "dad_count",
1769 		       SYSCTL_DESCR("Number of Duplicate Address Detection "
1770 				    "probes to send"),
1771 		       NULL, 0, &ip_dad_count, 0,
1772 		       CTL_NET, PF_INET, IPPROTO_IP,
1773 		       IPCTL_DAD_COUNT, CTL_EOL);
1774 #endif
1775 
1776 	/* anonportalgo RFC6056 subtree */
1777 	const struct sysctlnode *portalgo_node;
1778 	sysctl_createv(clog, 0, NULL, &portalgo_node,
1779 		       CTLFLAG_PERMANENT,
1780 		       CTLTYPE_NODE, "anonportalgo",
1781 		       SYSCTL_DESCR("Anonymous Port Algorithm Selection (RFC 6056)"),
1782 	    	       NULL, 0, NULL, 0,
1783 		       CTL_NET, PF_INET, IPPROTO_IP, CTL_CREATE, CTL_EOL);
1784 	sysctl_createv(clog, 0, &portalgo_node, NULL,
1785 		       CTLFLAG_PERMANENT,
1786 		       CTLTYPE_STRING, "available",
1787 		       SYSCTL_DESCR("available algorithms"),
1788 		       sysctl_portalgo_available, 0, NULL, PORTALGO_MAXLEN,
1789 		       CTL_CREATE, CTL_EOL);
1790 	sysctl_createv(clog, 0, &portalgo_node, NULL,
1791 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1792 		       CTLTYPE_STRING, "selected",
1793 		       SYSCTL_DESCR("selected algorithm"),
1794 		       sysctl_portalgo_selected4, 0, NULL, PORTALGO_MAXLEN,
1795 		       CTL_CREATE, CTL_EOL);
1796 	sysctl_createv(clog, 0, &portalgo_node, NULL,
1797 		       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
1798 		       CTLTYPE_STRUCT, "reserve",
1799 		       SYSCTL_DESCR("bitmap of reserved ports"),
1800 		       sysctl_portalgo_reserve4, 0, NULL, 0,
1801 		       CTL_CREATE, CTL_EOL);
1802 }
1803 
1804 void
ip_statinc(u_int stat)1805 ip_statinc(u_int stat)
1806 {
1807 
1808 	KASSERT(stat < IP_NSTATS);
1809 	IP_STATINC(stat);
1810 }
1811