1 /* $KAME: ip_encap.c,v 1.41 2001/03/15 08:35:08 itojun Exp $ */
2
3 /*-
4 * SPDX-License-Identifier: BSD-3-Clause
5 *
6 * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
7 * Copyright (c) 2018 Andrey V. Elsukov <ae@FreeBSD.org>
8 * All rights reserved.
9 *
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
12 * are met:
13 * 1. Redistributions of source code must retain the above copyright
14 * notice, this list of conditions and the following disclaimer.
15 * 2. Redistributions in binary form must reproduce the above copyright
16 * notice, this list of conditions and the following disclaimer in the
17 * documentation and/or other materials provided with the distribution.
18 * 3. Neither the name of the project nor the names of its contributors
19 * may be used to endorse or promote products derived from this software
20 * without specific prior written permission.
21 *
22 * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
25 * ARE DISCLAIMED. IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32 * SUCH DAMAGE.
33 */
34 /*
35 * My grandfather said that there's a devil inside tunnelling technology...
36 *
37 * We have surprisingly many protocols that want packets with IP protocol
38 * #4 or #41. Here's a list of protocols that want protocol #41:
39 * RFC1933 configured tunnel
40 * RFC1933 automatic tunnel
41 * RFC2401 IPsec tunnel
42 * RFC2473 IPv6 generic packet tunnelling
43 * RFC2529 6over4 tunnel
44 * mobile-ip6 (uses RFC2473)
45 * RFC3056 6to4 tunnel
46 * isatap tunnel
47 * Here's a list of protocol that want protocol #4:
48 * RFC1853 IPv4-in-IPv4 tunnelling
49 * RFC2003 IPv4 encapsulation within IPv4
50 * RFC2344 reverse tunnelling for mobile-ip4
51 * RFC2401 IPsec tunnel
52 * Well, what can I say. They impose different en/decapsulation mechanism
53 * from each other, so they need separate protocol handler. The only one
54 * we can easily determine by protocol # is IPsec, which always has
55 * AH/ESP/IPComp header right after outer IP header.
56 *
57 * So, clearly good old protosw does not work for protocol #4 and #41.
58 * The code will let you match protocol via src/dst address pair.
59 */
60
61 #include "opt_inet.h"
62 #include "opt_inet6.h"
63
64 #include <sys/param.h>
65 #include <sys/systm.h>
66 #include <sys/eventhandler.h>
67 #include <sys/kernel.h>
68 #include <sys/lock.h>
69 #include <sys/malloc.h>
70 #include <sys/mutex.h>
71 #include <sys/mbuf.h>
72 #include <sys/errno.h>
73 #include <sys/socket.h>
74
75 #include <net/if.h>
76 #include <net/if_var.h>
77
78 #include <netinet/in.h>
79 #include <netinet/ip_var.h>
80 #include <netinet/ip_encap.h>
81
82 #ifdef INET6
83 #include <netinet6/ip6_var.h>
84 #endif
85
86 static MALLOC_DEFINE(M_NETADDR, "encap_export_host",
87 "Export host address structure");
88
89 struct encaptab {
90 CK_LIST_ENTRY(encaptab) chain;
91 int proto;
92 int min_length;
93 int exact_match;
94 void *arg;
95
96 encap_lookup_t lookup;
97 encap_check_t check;
98 encap_input_t input;
99 };
100
101 struct srcaddrtab {
102 CK_LIST_ENTRY(srcaddrtab) chain;
103
104 encap_srcaddr_t srcaddr;
105 void *arg;
106 };
107
108 CK_LIST_HEAD(encaptab_head, encaptab);
109 CK_LIST_HEAD(srcaddrtab_head, srcaddrtab);
110 #ifdef INET
111 static struct encaptab_head ipv4_encaptab = CK_LIST_HEAD_INITIALIZER();
112 static struct srcaddrtab_head ipv4_srcaddrtab = CK_LIST_HEAD_INITIALIZER();
113 #endif
114 #ifdef INET6
115 static struct encaptab_head ipv6_encaptab = CK_LIST_HEAD_INITIALIZER();
116 static struct srcaddrtab_head ipv6_srcaddrtab = CK_LIST_HEAD_INITIALIZER();
117 #endif
118
119 static struct mtx encapmtx, srcaddrmtx;
120 MTX_SYSINIT(encapmtx, &encapmtx, "encapmtx", MTX_DEF);
121 MTX_SYSINIT(srcaddrmtx, &srcaddrmtx, "srcaddrmtx", MTX_DEF);
122 #define ENCAP_WLOCK() mtx_lock(&encapmtx)
123 #define ENCAP_WUNLOCK() mtx_unlock(&encapmtx)
124 #define ENCAP_RLOCK_TRACKER struct epoch_tracker encap_et
125 #define ENCAP_RLOCK() NET_EPOCH_ENTER(encap_et)
126 #define ENCAP_RUNLOCK() NET_EPOCH_EXIT(encap_et)
127 #define ENCAP_WAIT() NET_EPOCH_WAIT()
128
129 #define SRCADDR_WLOCK() mtx_lock(&srcaddrmtx)
130 #define SRCADDR_WUNLOCK() mtx_unlock(&srcaddrmtx)
131 #define SRCADDR_RLOCK_TRACKER struct epoch_tracker srcaddr_et
132 #define SRCADDR_RLOCK() \
133 epoch_enter_preempt(net_epoch_preempt, &srcaddr_et)
134 #define SRCADDR_RUNLOCK() \
135 epoch_exit_preempt(net_epoch_preempt, &srcaddr_et)
136 #define SRCADDR_WAIT() epoch_wait_preempt(net_epoch_preempt)
137
138 /*
139 * ifaddr_event_ext handler.
140 *
141 * Tunnelling interfaces may request the kernel to notify when
142 * some interface addresses appears or disappears. Usually tunnelling
143 * interface must use an address configured on the local machine as
144 * ingress address to be able receive datagramms and do not send
145 * spoofed packets.
146 */
147 static void
srcaddr_change_event(void * arg __unused,struct ifnet * ifp,struct ifaddr * ifa,int event)148 srcaddr_change_event(void *arg __unused, struct ifnet *ifp,
149 struct ifaddr *ifa, int event)
150 {
151 SRCADDR_RLOCK_TRACKER;
152 struct srcaddrtab_head *head;
153 struct srcaddrtab *p;
154
155 /* Support for old ifaddr_event. */
156 EVENTHANDLER_INVOKE(ifaddr_event, ifp);
157
158 switch (ifa->ifa_addr->sa_family) {
159 #ifdef INET
160 case AF_INET:
161 head = &ipv4_srcaddrtab;
162 break;
163 #endif
164 #ifdef INET6
165 case AF_INET6:
166 head = &ipv6_srcaddrtab;
167 break;
168 #endif
169 default:
170 /* ignore event */
171 return;
172 }
173
174 SRCADDR_RLOCK();
175 CK_LIST_FOREACH(p, head, chain) {
176 (*p->srcaddr)(p->arg, ifa->ifa_addr, event);
177 }
178 SRCADDR_RUNLOCK();
179 }
180 EVENTHANDLER_DEFINE(ifaddr_event_ext, srcaddr_change_event, NULL, 0);
181
182 static struct srcaddrtab *
encap_register_srcaddr(struct srcaddrtab_head * head,encap_srcaddr_t func,void * arg,int mflags)183 encap_register_srcaddr(struct srcaddrtab_head *head, encap_srcaddr_t func,
184 void *arg, int mflags)
185 {
186 struct srcaddrtab *p, *tmp;
187
188 if (func == NULL)
189 return (NULL);
190 p = malloc(sizeof(*p), M_NETADDR, mflags);
191 if (p == NULL)
192 return (NULL);
193 p->srcaddr = func;
194 p->arg = arg;
195
196 SRCADDR_WLOCK();
197 CK_LIST_FOREACH(tmp, head, chain) {
198 if (func == tmp->srcaddr && arg == tmp->arg)
199 break;
200 }
201 if (tmp == NULL)
202 CK_LIST_INSERT_HEAD(head, p, chain);
203 SRCADDR_WUNLOCK();
204
205 if (tmp != NULL) {
206 free(p, M_NETADDR);
207 p = tmp;
208 }
209 return (p);
210 }
211
212 static int
encap_unregister_srcaddr(struct srcaddrtab_head * head,const struct srcaddrtab * cookie)213 encap_unregister_srcaddr(struct srcaddrtab_head *head,
214 const struct srcaddrtab *cookie)
215 {
216 struct srcaddrtab *p;
217
218 SRCADDR_WLOCK();
219 CK_LIST_FOREACH(p, head, chain) {
220 if (p == cookie) {
221 CK_LIST_REMOVE(p, chain);
222 SRCADDR_WUNLOCK();
223 SRCADDR_WAIT();
224 free(p, M_NETADDR);
225 return (0);
226 }
227 }
228 SRCADDR_WUNLOCK();
229 return (EINVAL);
230 }
231
232 static struct encaptab *
encap_attach(struct encaptab_head * head,const struct encap_config * cfg,void * arg,int mflags)233 encap_attach(struct encaptab_head *head, const struct encap_config *cfg,
234 void *arg, int mflags)
235 {
236 struct encaptab *ep, *tmp;
237
238 if (cfg == NULL || cfg->input == NULL ||
239 (cfg->check == NULL && cfg->lookup == NULL) ||
240 (cfg->lookup != NULL && cfg->exact_match != ENCAP_DRV_LOOKUP) ||
241 (cfg->exact_match == ENCAP_DRV_LOOKUP && cfg->lookup == NULL))
242 return (NULL);
243
244 ep = malloc(sizeof(*ep), M_NETADDR, mflags);
245 if (ep == NULL)
246 return (NULL);
247
248 ep->proto = cfg->proto;
249 ep->min_length = cfg->min_length;
250 ep->exact_match = cfg->exact_match;
251 ep->arg = arg;
252 ep->lookup = cfg->exact_match == ENCAP_DRV_LOOKUP ? cfg->lookup: NULL;
253 ep->check = cfg->exact_match != ENCAP_DRV_LOOKUP ? cfg->check: NULL;
254 ep->input = cfg->input;
255
256 ENCAP_WLOCK();
257 CK_LIST_FOREACH(tmp, head, chain) {
258 if (tmp->exact_match <= ep->exact_match)
259 break;
260 }
261 if (tmp == NULL)
262 CK_LIST_INSERT_HEAD(head, ep, chain);
263 else
264 CK_LIST_INSERT_BEFORE(tmp, ep, chain);
265 ENCAP_WUNLOCK();
266 return (ep);
267 }
268
269 static int
encap_detach(struct encaptab_head * head,const struct encaptab * cookie)270 encap_detach(struct encaptab_head *head, const struct encaptab *cookie)
271 {
272 struct encaptab *ep;
273
274 ENCAP_WLOCK();
275 CK_LIST_FOREACH(ep, head, chain) {
276 if (ep == cookie) {
277 CK_LIST_REMOVE(ep, chain);
278 ENCAP_WUNLOCK();
279 ENCAP_WAIT();
280 free(ep, M_NETADDR);
281 return (0);
282 }
283 }
284 ENCAP_WUNLOCK();
285 return (EINVAL);
286 }
287
288 static int
encap_input(struct encaptab_head * head,struct mbuf * m,int off,int proto)289 encap_input(struct encaptab_head *head, struct mbuf *m, int off, int proto)
290 {
291 ENCAP_RLOCK_TRACKER;
292 struct encaptab *ep, *match;
293 void *arg;
294 int matchprio, ret;
295
296 match = NULL;
297 matchprio = 0;
298
299 ENCAP_RLOCK();
300 CK_LIST_FOREACH(ep, head, chain) {
301 if (ep->proto >= 0 && ep->proto != proto)
302 continue;
303 if (ep->min_length > m->m_pkthdr.len)
304 continue;
305 if (ep->exact_match == ENCAP_DRV_LOOKUP)
306 ret = (*ep->lookup)(m, off, proto, &arg);
307 else
308 ret = (*ep->check)(m, off, proto, ep->arg);
309 if (ret <= 0)
310 continue;
311 if (ret > matchprio) {
312 match = ep;
313 if (ep->exact_match != ENCAP_DRV_LOOKUP)
314 arg = ep->arg;
315 /*
316 * No need to continue the search, we got the
317 * exact match.
318 */
319 if (ret >= ep->exact_match)
320 break;
321 matchprio = ret;
322 }
323 }
324
325 if (match != NULL) {
326 /* found a match, "match" has the best one */
327 ret = (*match->input)(m, off, proto, arg);
328 ENCAP_RUNLOCK();
329 MPASS(ret == IPPROTO_DONE);
330 return (IPPROTO_DONE);
331 }
332 ENCAP_RUNLOCK();
333 return (0);
334 }
335
336 #ifdef INET
337 const struct srcaddrtab *
ip_encap_register_srcaddr(encap_srcaddr_t func,void * arg,int mflags)338 ip_encap_register_srcaddr(encap_srcaddr_t func, void *arg, int mflags)
339 {
340
341 return (encap_register_srcaddr(&ipv4_srcaddrtab, func, arg, mflags));
342 }
343
344 int
ip_encap_unregister_srcaddr(const struct srcaddrtab * cookie)345 ip_encap_unregister_srcaddr(const struct srcaddrtab *cookie)
346 {
347
348 return (encap_unregister_srcaddr(&ipv4_srcaddrtab, cookie));
349 }
350
351 const struct encaptab *
ip_encap_attach(const struct encap_config * cfg,void * arg,int mflags)352 ip_encap_attach(const struct encap_config *cfg, void *arg, int mflags)
353 {
354
355 return (encap_attach(&ipv4_encaptab, cfg, arg, mflags));
356 }
357
358 int
ip_encap_detach(const struct encaptab * cookie)359 ip_encap_detach(const struct encaptab *cookie)
360 {
361
362 return (encap_detach(&ipv4_encaptab, cookie));
363 }
364
365 int
encap4_input(struct mbuf ** mp,int * offp,int proto)366 encap4_input(struct mbuf **mp, int *offp, int proto)
367 {
368
369 if (encap_input(&ipv4_encaptab, *mp, *offp, proto) != IPPROTO_DONE)
370 return (rip_input(mp, offp, proto));
371 return (IPPROTO_DONE);
372 }
373 #endif /* INET */
374
375 #ifdef INET6
376 const struct srcaddrtab *
ip6_encap_register_srcaddr(encap_srcaddr_t func,void * arg,int mflags)377 ip6_encap_register_srcaddr(encap_srcaddr_t func, void *arg, int mflags)
378 {
379
380 return (encap_register_srcaddr(&ipv6_srcaddrtab, func, arg, mflags));
381 }
382
383 int
ip6_encap_unregister_srcaddr(const struct srcaddrtab * cookie)384 ip6_encap_unregister_srcaddr(const struct srcaddrtab *cookie)
385 {
386
387 return (encap_unregister_srcaddr(&ipv6_srcaddrtab, cookie));
388 }
389
390 const struct encaptab *
ip6_encap_attach(const struct encap_config * cfg,void * arg,int mflags)391 ip6_encap_attach(const struct encap_config *cfg, void *arg, int mflags)
392 {
393
394 return (encap_attach(&ipv6_encaptab, cfg, arg, mflags));
395 }
396
397 int
ip6_encap_detach(const struct encaptab * cookie)398 ip6_encap_detach(const struct encaptab *cookie)
399 {
400
401 return (encap_detach(&ipv6_encaptab, cookie));
402 }
403
404 int
encap6_input(struct mbuf ** mp,int * offp,int proto)405 encap6_input(struct mbuf **mp, int *offp, int proto)
406 {
407
408 if (encap_input(&ipv6_encaptab, *mp, *offp, proto) != IPPROTO_DONE)
409 return (rip6_input(mp, offp, proto));
410 return (IPPROTO_DONE);
411 }
412 #endif /* INET6 */
413