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.
60 #include "opt_inet6.h"
62 #include <sys/param.h>
63 #include <sys/systm.h>
64 #include <sys/socket.h>
65 #include <sys/sockio.h>
67 #include <sys/errno.h>
68 #include <sys/protosw.h>
69 #include <sys/queue.h>
72 #include <net/route.h>
74 #include <netinet/in.h>
75 #include <netinet/in_systm.h>
76 #include <netinet/ip.h>
77 #include <netinet/ip_var.h>
78 #include <netinet/ip_encap.h>
81 #include <netinet/ip6.h>
82 #include <netinet6/ip6_var.h>
83 #include <netinet6/ip6protosw.h>
86 #include <machine/stdarg.h>
88 #include <net/net_osdep.h>
90 #include <sys/kernel.h>
91 #include <sys/malloc.h>
92 #include <sys/thread2.h>
93 MALLOC_DEFINE(M_IPENCAP
, "IP Encapsulation", "IP Encapsulation");
95 static void encap_add (struct encaptab
*);
96 static int mask_match (const struct encaptab
*, const struct sockaddr
*,
97 const struct sockaddr
*);
98 static void encap_fillarg (struct mbuf
*, const struct encaptab
*);
100 #ifndef LIST_HEAD_INITIALIZER
101 /* rely upon BSS initialization */
102 LIST_HEAD(, encaptab
) encaptab
;
104 LIST_HEAD(, encaptab
) encaptab
= LIST_HEAD_INITIALIZER(&encaptab
);
107 int (*ipip_input
)(struct mbuf
**, int *, int); /* hook for mrouting */
112 static int initialized
= 0;
119 * we cannot use LIST_INIT() here, since drivers may want to call
120 * encap_attach(), on driver attach. encap_init() will be called
121 * on AF_INET{,6} initialization, which happens after driver
122 * initialization - using LIST_INIT() here can nuke encap_attach()
125 LIST_INIT(&encaptab
);
131 encap4_input(struct mbuf
**mp
, int *offp
, int proto
)
133 struct mbuf
*m
= *mp
;
136 struct sockaddr_in s
, d
;
137 const struct protosw
*psw
;
138 struct encaptab
*ep
, *match
;
144 * Some NICs, noticeably igb(4) and ix(4), use inner IP
145 * datagram to calculate the packet hash, which leads us
148 m
->m_flags
&= ~M_HASH
;
151 lwkt_port_t port
= netisr_hashport(m
->m_pkthdr
.hash
);
153 KASSERT(port
!= &curthread
->td_msgport
,
154 ("mbuf hash recursion"));
155 ip_transport_redispatch(port
, m
, off
);
157 return (IPPROTO_DONE
);
160 ip
= mtod(m
, struct ip
*);
164 s
.sin_family
= AF_INET
;
165 s
.sin_len
= sizeof(struct sockaddr_in
);
166 s
.sin_addr
= ip
->ip_src
;
168 d
.sin_family
= AF_INET
;
169 d
.sin_len
= sizeof(struct sockaddr_in
);
170 d
.sin_addr
= ip
->ip_dst
;
174 for (ep
= LIST_FIRST(&encaptab
); ep
; ep
= LIST_NEXT(ep
, chain
)) {
175 if (ep
->af
!= AF_INET
)
177 if (ep
->proto
>= 0 && ep
->proto
!= proto
)
180 prio
= (*ep
->func
)(m
, off
, proto
, ep
->arg
);
183 * it's inbound traffic, we need to match in reverse
186 prio
= mask_match(ep
,
187 (struct sockaddr
*)&d
,
188 (struct sockaddr
*)&s
);
192 * We prioritize the matches by using bit length of the
193 * matches. mask_match() and user-supplied matching function
194 * should return the bit length of the matches (for example,
195 * if both src/dst are matched for IPv4, 64 should be returned).
196 * 0 or negative return value means "it did not match".
198 * The question is, since we have two "mask" portion, we
199 * cannot really define total order between entries.
200 * For example, which of these should be preferred?
201 * mask_match() returns 48 (32 + 16) for both of them.
202 * src=3ffe::/16, dst=3ffe:501::/32
203 * src=3ffe:501::/32, dst=3ffe::/16
205 * We need to loop through all the possible candidates
206 * to get the best match - the search takes O(n) for
207 * n attachments (i.e. interfaces).
211 if (prio
> matchprio
) {
218 /* found a match, "match" has the best one */
220 if (psw
&& psw
->pr_input
) {
221 encap_fillarg(m
, match
);
223 (*psw
->pr_input
)(mp
, offp
, proto
);
227 return(IPPROTO_DONE
);
230 /* for backward compatibility */
231 if (proto
== IPPROTO_IPV4
&& ipip_input
) {
233 ipip_input(mp
, offp
, proto
);
234 return(IPPROTO_DONE
);
237 /* last resort: inject to raw socket */
239 rip_input(mp
, offp
, proto
);
240 return(IPPROTO_DONE
);
246 encap6_input(struct mbuf
**mp
, int *offp
, int proto
)
248 struct mbuf
*m
= *mp
;
250 struct sockaddr_in6 s
, d
;
251 const struct protosw
*psw
;
252 struct encaptab
*ep
, *match
;
255 ip6
= mtod(m
, struct ip6_hdr
*);
258 s
.sin6_family
= AF_INET6
;
259 s
.sin6_len
= sizeof(struct sockaddr_in6
);
260 s
.sin6_addr
= ip6
->ip6_src
;
262 d
.sin6_family
= AF_INET6
;
263 d
.sin6_len
= sizeof(struct sockaddr_in6
);
264 d
.sin6_addr
= ip6
->ip6_dst
;
268 for (ep
= LIST_FIRST(&encaptab
); ep
; ep
= LIST_NEXT(ep
, chain
)) {
269 if (ep
->af
!= AF_INET6
)
271 if (ep
->proto
>= 0 && ep
->proto
!= proto
)
274 prio
= (*ep
->func
)(m
, *offp
, proto
, ep
->arg
);
277 * it's inbound traffic, we need to match in reverse
280 prio
= mask_match(ep
, (struct sockaddr
*)&d
,
281 (struct sockaddr
*)&s
);
284 /* see encap4_input() for issues here */
287 if (prio
> matchprio
) {
296 if (psw
&& psw
->pr_input
) {
297 encap_fillarg(m
, match
);
298 return (*psw
->pr_input
)(mp
, offp
, proto
);
305 /* last resort: inject to raw socket */
306 return rip6_input(mp
, offp
, proto
);
311 encap_add(struct encaptab
*ep
)
314 LIST_INSERT_HEAD(&encaptab
, ep
, chain
);
318 * sp (src ptr) is always my side, and dp (dst ptr) is always remote side.
319 * length of mask (sm and dm) is assumed to be same as sp/dp.
320 * Return value will be necessary as input (cookie) for encap_detach().
322 const struct encaptab
*
323 encap_attach(int af
, int proto
, const struct sockaddr
*sp
,
324 const struct sockaddr
*sm
, const struct sockaddr
*dp
,
325 const struct sockaddr
*dm
, const struct protosw
*psw
, void *arg
)
330 /* sanity check on args */
331 if (sp
->sa_len
> sizeof ep
->src
|| dp
->sa_len
> sizeof ep
->dst
)
333 if (sp
->sa_len
!= dp
->sa_len
)
335 if (af
!= sp
->sa_family
|| af
!= dp
->sa_family
)
338 /* check if anyone have already attached with exactly same config */
339 for (ep
= LIST_FIRST(&encaptab
); ep
; ep
= LIST_NEXT(ep
, chain
)) {
342 if (ep
->proto
!= proto
)
344 if (ep
->src
.ss_len
!= sp
->sa_len
||
345 bcmp(&ep
->src
, sp
, sp
->sa_len
) != 0 ||
346 bcmp(&ep
->srcmask
, sm
, sp
->sa_len
) != 0)
348 if (ep
->dst
.ss_len
!= dp
->sa_len
||
349 bcmp(&ep
->dst
, dp
, dp
->sa_len
) != 0 ||
350 bcmp(&ep
->dstmask
, dm
, dp
->sa_len
) != 0)
356 ep
= kmalloc(sizeof *ep
, M_IPENCAP
, M_INTWAIT
| M_ZERO
| M_NULLOK
);
362 bcopy(sp
, &ep
->src
, sp
->sa_len
);
363 bcopy(sm
, &ep
->srcmask
, sp
->sa_len
);
364 bcopy(dp
, &ep
->dst
, dp
->sa_len
);
365 bcopy(dm
, &ep
->dstmask
, dp
->sa_len
);
379 const struct encaptab
*
380 encap_attach_func(int af
, int proto
,
381 int (*func
)(const struct mbuf
*, int, int, void *),
382 const struct protosw
*psw
, void *arg
)
387 /* sanity check on args */
391 ep
= kmalloc(sizeof *ep
, M_IPENCAP
, M_INTWAIT
| M_ZERO
| M_NULLOK
);
412 encap_detach(const struct encaptab
*cookie
)
414 const struct encaptab
*ep
= cookie
;
417 for (p
= LIST_FIRST(&encaptab
); p
; p
= LIST_NEXT(p
, chain
)) {
419 LIST_REMOVE(p
, chain
);
420 kfree(p
, M_IPENCAP
); /*XXX*/
429 mask_match(const struct encaptab
*ep
, const struct sockaddr
*sp
,
430 const struct sockaddr
*dp
)
432 struct sockaddr_storage s
;
433 struct sockaddr_storage d
;
435 const u_int8_t
*p
, *q
;
439 if (sp
->sa_len
> sizeof s
|| dp
->sa_len
> sizeof d
)
441 if (sp
->sa_family
!= ep
->af
|| dp
->sa_family
!= ep
->af
)
443 if (sp
->sa_len
!= ep
->src
.ss_len
|| dp
->sa_len
!= ep
->dst
.ss_len
)
448 p
= (const u_int8_t
*)sp
;
449 q
= (const u_int8_t
*)&ep
->srcmask
;
451 for (i
= 0 ; i
< sp
->sa_len
; i
++) {
454 matchlen
+= (q
[i
] ? 8 : 0);
457 p
= (const u_int8_t
*)dp
;
458 q
= (const u_int8_t
*)&ep
->dstmask
;
460 for (i
= 0 ; i
< dp
->sa_len
; i
++) {
462 /* XXX rough estimate */
463 matchlen
+= (q
[i
] ? 8 : 0);
466 /* need to overwrite len/family portion as we don't compare them */
467 s
.ss_len
= sp
->sa_len
;
468 s
.ss_family
= sp
->sa_family
;
469 d
.ss_len
= dp
->sa_len
;
470 d
.ss_family
= dp
->sa_family
;
472 if (bcmp(&s
, &ep
->src
, ep
->src
.ss_len
) == 0 &&
473 bcmp(&d
, &ep
->dst
, ep
->dst
.ss_len
) == 0) {
480 encap_fillarg(struct mbuf
*m
, const struct encaptab
*ep
)
484 tag
= m_tag_get(PACKET_TAG_ENCAP
, sizeof(void *), M_NOWAIT
);
486 *(void **)m_tag_data(tag
) = ep
->arg
;
487 m_tag_prepend(m
, tag
);
492 encap_getarg(struct mbuf
*m
)
497 tag
= m_tag_find(m
, PACKET_TAG_ENCAP
, NULL
);
499 p
= *(void **)m_tag_data(tag
);
500 m_tag_delete(m
, tag
);