1 /* $FreeBSD: src/sys/netinet/ip_encap.c,v 1.1.2.5 2003/01/23 21:06:45 sam Exp $ */
2 /* $KAME: ip_encap.c,v 1.41 2001/03/15 08:35:08 itojun Exp $ */
5 * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
11 * 1. Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
16 * 3. Neither the name of the project nor the names of its contributors
17 * may be used to endorse or promote products derived from this software
18 * without specific prior written permission.
20 * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
21 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23 * ARE DISCLAIMED. IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
24 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
25 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
26 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
27 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
28 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
29 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
33 * My grandfather said that there's a devil inside tunnelling technology...
35 * We have surprisingly many protocols that want packets with IP protocol
36 * #4 or #41. Here's a list of protocols that want protocol #41:
37 * RFC1933 configured tunnel
38 * RFC1933 automatic tunnel
39 * RFC2401 IPsec tunnel
40 * RFC2473 IPv6 generic packet tunnelling
41 * RFC2529 6over4 tunnel
42 * mobile-ip6 (uses RFC2473)
45 * Here's a list of protocol that want protocol #4:
46 * RFC1853 IPv4-in-IPv4 tunnelling
47 * RFC2003 IPv4 encapsulation within IPv4
48 * RFC2344 reverse tunnelling for mobile-ip4
49 * RFC2401 IPsec tunnel
50 * Well, what can I say. They impose different en/decapsulation mechanism
51 * from each other, so they need separate protocol handler. The only one
52 * we can easily determine by protocol # is IPsec, which always has
53 * AH/ESP/IPComp header right after outer IP header.
55 * So, clearly good old protosw does not work for protocol #4 and #41.
56 * The code will let you match protocol via src/dst address pair.
58 /* XXX is M_NETADDR correct? */
61 #include "opt_inet6.h"
63 #include <sys/param.h>
64 #include <sys/systm.h>
65 #include <sys/socket.h>
66 #include <sys/sockio.h>
68 #include <sys/errno.h>
69 #include <sys/protosw.h>
70 #include <sys/queue.h>
73 #include <net/route.h>
75 #include <netinet/in.h>
76 #include <netinet/in_systm.h>
77 #include <netinet/ip.h>
78 #include <netinet/ip_var.h>
79 #include <netinet/ip_encap.h>
82 #include <netinet/ip6.h>
83 #include <netinet6/ip6_var.h>
84 #include <netinet6/ip6protosw.h>
87 #include <machine/stdarg.h>
89 #include <net/net_osdep.h>
91 #include <sys/kernel.h>
92 #include <sys/malloc.h>
93 #include <sys/thread2.h>
94 MALLOC_DEFINE(M_NETADDR
, "Export Host", "Export host address structure");
96 static void encap_add (struct encaptab
*);
97 static int mask_match (const struct encaptab
*, const struct sockaddr
*,
98 const struct sockaddr
*);
99 static void encap_fillarg (struct mbuf
*, const struct encaptab
*);
101 #ifndef LIST_HEAD_INITIALIZER
102 /* rely upon BSS initialization */
103 LIST_HEAD(, encaptab
) encaptab
;
105 LIST_HEAD(, encaptab
) encaptab
= LIST_HEAD_INITIALIZER(&encaptab
);
108 int (*ipip_input
)(struct mbuf
**, int *, int); /* hook for mrouting */
113 static int initialized
= 0;
120 * we cannot use LIST_INIT() here, since drivers may want to call
121 * encap_attach(), on driver attach. encap_init() will be called
122 * on AF_INET{,6} initialization, which happens after driver
123 * initialization - using LIST_INIT() here can nuke encap_attach()
126 LIST_INIT(&encaptab
);
132 encap4_input(struct mbuf
**mp
, int *offp
, int proto
)
134 struct mbuf
*m
= *mp
;
137 struct sockaddr_in s
, d
;
138 const struct protosw
*psw
;
139 struct encaptab
*ep
, *match
;
142 ip
= mtod(m
, struct ip
*);
146 s
.sin_family
= AF_INET
;
147 s
.sin_len
= sizeof(struct sockaddr_in
);
148 s
.sin_addr
= ip
->ip_src
;
150 d
.sin_family
= AF_INET
;
151 d
.sin_len
= sizeof(struct sockaddr_in
);
152 d
.sin_addr
= ip
->ip_dst
;
156 for (ep
= LIST_FIRST(&encaptab
); ep
; ep
= LIST_NEXT(ep
, chain
)) {
157 if (ep
->af
!= AF_INET
)
159 if (ep
->proto
>= 0 && ep
->proto
!= proto
)
162 prio
= (*ep
->func
)(m
, off
, proto
, ep
->arg
);
165 * it's inbound traffic, we need to match in reverse
168 prio
= mask_match(ep
,
169 (struct sockaddr
*)&d
,
170 (struct sockaddr
*)&s
);
174 * We prioritize the matches by using bit length of the
175 * matches. mask_match() and user-supplied matching function
176 * should return the bit length of the matches (for example,
177 * if both src/dst are matched for IPv4, 64 should be returned).
178 * 0 or negative return value means "it did not match".
180 * The question is, since we have two "mask" portion, we
181 * cannot really define total order between entries.
182 * For example, which of these should be preferred?
183 * mask_match() returns 48 (32 + 16) for both of them.
184 * src=3ffe::/16, dst=3ffe:501::/32
185 * src=3ffe:501::/32, dst=3ffe::/16
187 * We need to loop through all the possible candidates
188 * to get the best match - the search takes O(n) for
189 * n attachments (i.e. interfaces).
193 if (prio
> matchprio
) {
200 /* found a match, "match" has the best one */
202 if (psw
&& psw
->pr_input
) {
203 encap_fillarg(m
, match
);
205 (*psw
->pr_input
)(mp
, offp
, proto
);
209 return(IPPROTO_DONE
);
212 /* for backward compatibility */
213 if (proto
== IPPROTO_IPV4
&& ipip_input
) {
215 ipip_input(mp
, offp
, proto
);
216 return(IPPROTO_DONE
);
219 /* last resort: inject to raw socket */
221 rip_input(mp
, offp
, proto
);
222 return(IPPROTO_DONE
);
228 encap6_input(struct mbuf
**mp
, int *offp
, int proto
)
230 struct mbuf
*m
= *mp
;
232 struct sockaddr_in6 s
, d
;
233 const struct protosw
*psw
;
234 struct encaptab
*ep
, *match
;
237 ip6
= mtod(m
, struct ip6_hdr
*);
240 s
.sin6_family
= AF_INET6
;
241 s
.sin6_len
= sizeof(struct sockaddr_in6
);
242 s
.sin6_addr
= ip6
->ip6_src
;
244 d
.sin6_family
= AF_INET6
;
245 d
.sin6_len
= sizeof(struct sockaddr_in6
);
246 d
.sin6_addr
= ip6
->ip6_dst
;
250 for (ep
= LIST_FIRST(&encaptab
); ep
; ep
= LIST_NEXT(ep
, chain
)) {
251 if (ep
->af
!= AF_INET6
)
253 if (ep
->proto
>= 0 && ep
->proto
!= proto
)
256 prio
= (*ep
->func
)(m
, *offp
, proto
, ep
->arg
);
259 * it's inbound traffic, we need to match in reverse
262 prio
= mask_match(ep
, (struct sockaddr
*)&d
,
263 (struct sockaddr
*)&s
);
266 /* see encap4_input() for issues here */
269 if (prio
> matchprio
) {
278 if (psw
&& psw
->pr_input
) {
279 encap_fillarg(m
, match
);
280 return (*psw
->pr_input
)(mp
, offp
, proto
);
287 /* last resort: inject to raw socket */
288 return rip6_input(mp
, offp
, proto
);
293 encap_add(struct encaptab
*ep
)
296 LIST_INSERT_HEAD(&encaptab
, ep
, chain
);
300 * sp (src ptr) is always my side, and dp (dst ptr) is always remote side.
301 * length of mask (sm and dm) is assumed to be same as sp/dp.
302 * Return value will be necessary as input (cookie) for encap_detach().
304 const struct encaptab
*
305 encap_attach(int af
, int proto
, const struct sockaddr
*sp
,
306 const struct sockaddr
*sm
, const struct sockaddr
*dp
,
307 const struct sockaddr
*dm
, const struct protosw
*psw
, void *arg
)
312 /* sanity check on args */
313 if (sp
->sa_len
> sizeof ep
->src
|| dp
->sa_len
> sizeof ep
->dst
)
315 if (sp
->sa_len
!= dp
->sa_len
)
317 if (af
!= sp
->sa_family
|| af
!= dp
->sa_family
)
320 /* check if anyone have already attached with exactly same config */
321 for (ep
= LIST_FIRST(&encaptab
); ep
; ep
= LIST_NEXT(ep
, chain
)) {
324 if (ep
->proto
!= proto
)
326 if (ep
->src
.ss_len
!= sp
->sa_len
||
327 bcmp(&ep
->src
, sp
, sp
->sa_len
) != 0 ||
328 bcmp(&ep
->srcmask
, sm
, sp
->sa_len
) != 0)
330 if (ep
->dst
.ss_len
!= dp
->sa_len
||
331 bcmp(&ep
->dst
, dp
, dp
->sa_len
) != 0 ||
332 bcmp(&ep
->dstmask
, dm
, dp
->sa_len
) != 0)
338 ep
= kmalloc(sizeof *ep
, M_NETADDR
, M_INTWAIT
| M_ZERO
| M_NULLOK
);
344 bcopy(sp
, &ep
->src
, sp
->sa_len
);
345 bcopy(sm
, &ep
->srcmask
, sp
->sa_len
);
346 bcopy(dp
, &ep
->dst
, dp
->sa_len
);
347 bcopy(dm
, &ep
->dstmask
, dp
->sa_len
);
361 const struct encaptab
*
362 encap_attach_func(int af
, int proto
,
363 int (*func
)(const struct mbuf
*, int, int, void *),
364 const struct protosw
*psw
, void *arg
)
369 /* sanity check on args */
373 ep
= kmalloc(sizeof *ep
, M_NETADDR
, M_INTWAIT
| M_ZERO
| M_NULLOK
);
394 encap_detach(const struct encaptab
*cookie
)
396 const struct encaptab
*ep
= cookie
;
399 for (p
= LIST_FIRST(&encaptab
); p
; p
= LIST_NEXT(p
, chain
)) {
401 LIST_REMOVE(p
, chain
);
402 kfree(p
, M_NETADDR
); /*XXX*/
411 mask_match(const struct encaptab
*ep
, const struct sockaddr
*sp
,
412 const struct sockaddr
*dp
)
414 struct sockaddr_storage s
;
415 struct sockaddr_storage d
;
417 const u_int8_t
*p
, *q
;
421 if (sp
->sa_len
> sizeof s
|| dp
->sa_len
> sizeof d
)
423 if (sp
->sa_family
!= ep
->af
|| dp
->sa_family
!= ep
->af
)
425 if (sp
->sa_len
!= ep
->src
.ss_len
|| dp
->sa_len
!= ep
->dst
.ss_len
)
430 p
= (const u_int8_t
*)sp
;
431 q
= (const u_int8_t
*)&ep
->srcmask
;
433 for (i
= 0 ; i
< sp
->sa_len
; i
++) {
436 matchlen
+= (q
[i
] ? 8 : 0);
439 p
= (const u_int8_t
*)dp
;
440 q
= (const u_int8_t
*)&ep
->dstmask
;
442 for (i
= 0 ; i
< dp
->sa_len
; i
++) {
444 /* XXX rough estimate */
445 matchlen
+= (q
[i
] ? 8 : 0);
448 /* need to overwrite len/family portion as we don't compare them */
449 s
.ss_len
= sp
->sa_len
;
450 s
.ss_family
= sp
->sa_family
;
451 d
.ss_len
= dp
->sa_len
;
452 d
.ss_family
= dp
->sa_family
;
454 if (bcmp(&s
, &ep
->src
, ep
->src
.ss_len
) == 0 &&
455 bcmp(&d
, &ep
->dst
, ep
->dst
.ss_len
) == 0) {
462 encap_fillarg(struct mbuf
*m
, const struct encaptab
*ep
)
466 tag
= m_tag_get(PACKET_TAG_ENCAP
, sizeof(void *), M_NOWAIT
);
468 *(void **)m_tag_data(tag
) = ep
->arg
;
469 m_tag_prepend(m
, tag
);
474 encap_getarg(struct mbuf
*m
)
479 tag
= m_tag_find(m
, PACKET_TAG_ENCAP
, NULL
);
481 p
= *(void **)m_tag_data(tag
);
482 m_tag_delete(m
, tag
);