raw_ip6.c revision 1.97 1 /* $NetBSD: raw_ip6.c,v 1.97 2008/04/23 06:09:05 thorpej Exp $ */
2 /* $KAME: raw_ip6.c,v 1.82 2001/07/23 18:57:56 jinmei Exp $ */
3
4 /*
5 * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
6 * All rights reserved.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
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.
19 *
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
30 * SUCH DAMAGE.
31 */
32
33 /*
34 * Copyright (c) 1982, 1986, 1988, 1993
35 * The Regents of the University of California. All rights reserved.
36 *
37 * Redistribution and use in source and binary forms, with or without
38 * modification, are permitted provided that the following conditions
39 * are met:
40 * 1. Redistributions of source code must retain the above copyright
41 * notice, this list of conditions and the following disclaimer.
42 * 2. Redistributions in binary form must reproduce the above copyright
43 * notice, this list of conditions and the following disclaimer in the
44 * documentation and/or other materials provided with the distribution.
45 * 3. Neither the name of the University nor the names of its contributors
46 * may be used to endorse or promote products derived from this software
47 * without specific prior written permission.
48 *
49 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
50 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
51 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
52 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
53 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
54 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
55 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
56 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
57 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
58 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
59 * SUCH DAMAGE.
60 *
61 * @(#)raw_ip.c 8.2 (Berkeley) 1/4/94
62 */
63
64 #include <sys/cdefs.h>
65 __KERNEL_RCSID(0, "$NetBSD: raw_ip6.c,v 1.97 2008/04/23 06:09:05 thorpej Exp $");
66
67 #include "opt_ipsec.h"
68
69 #include <sys/param.h>
70 #include <sys/sysctl.h>
71 #include <sys/malloc.h>
72 #include <sys/mbuf.h>
73 #include <sys/socket.h>
74 #include <sys/protosw.h>
75 #include <sys/socketvar.h>
76 #include <sys/errno.h>
77 #include <sys/systm.h>
78 #include <sys/proc.h>
79 #include <sys/kauth.h>
80
81 #include <net/if.h>
82 #include <net/route.h>
83 #include <net/if_types.h>
84 #include <net/net_stats.h>
85
86 #include <netinet/in.h>
87 #include <netinet/in_var.h>
88 #include <netinet/ip6.h>
89 #include <netinet6/ip6_var.h>
90 #include <netinet6/ip6_private.h>
91 #include <netinet6/ip6_mroute.h>
92 #include <netinet/icmp6.h>
93 #include <netinet6/icmp6_private.h>
94 #include <netinet6/in6_pcb.h>
95 #include <netinet6/nd6.h>
96 #include <netinet6/ip6protosw.h>
97 #include <netinet6/scope6_var.h>
98 #include <netinet6/raw_ip6.h>
99
100 #ifdef IPSEC
101 #include <netinet6/ipsec.h>
102 #include <netinet6/ipsec_private.h>
103 #endif /* IPSEC */
104
105 #ifdef FAST_IPSEC
106 #include <netipsec/ipsec.h>
107 #include <netipsec/ipsec_var.h>
108 #include <netipsec/ipsec_private.h>
109 #include <netipsec/ipsec6.h>
110 #endif
111
112 #include "faith.h"
113 #if defined(NFAITH) && 0 < NFAITH
114 #include <net/if_faith.h>
115 #endif
116
117 extern struct inpcbtable rawcbtable;
118 struct inpcbtable raw6cbtable;
119 #define ifatoia6(ifa) ((struct in6_ifaddr *)(ifa))
120
121 /*
122 * Raw interface to IP6 protocol.
123 */
124
125 static percpu_t *rip6stat_percpu;
126
127 #define RIP6_STATINC(x) _NET_STATINC(rip6stat_percpu, x)
128
129 /*
130 * Initialize raw connection block queue.
131 */
132 void
133 rip6_init()
134 {
135
136 in6_pcbinit(&raw6cbtable, 1, 1);
137
138 rip6stat_percpu = percpu_alloc(sizeof(uint64_t) * RIP6_NSTATS);
139 }
140
141 /*
142 * Setup generic address and protocol structures
143 * for raw_input routine, then pass them along with
144 * mbuf chain.
145 */
146 int
147 rip6_input(struct mbuf **mp, int *offp, int proto)
148 {
149 struct mbuf *m = *mp;
150 struct ip6_hdr *ip6 = mtod(m, struct ip6_hdr *);
151 struct inpcb_hdr *inph;
152 struct in6pcb *in6p;
153 struct in6pcb *last = NULL;
154 struct sockaddr_in6 rip6src;
155 struct mbuf *opts = NULL;
156
157 RIP6_STATINC(RIP6_STAT_IPACKETS);
158
159 #if defined(NFAITH) && 0 < NFAITH
160 if (faithprefix(&ip6->ip6_dst)) {
161 /* send icmp6 host unreach? */
162 m_freem(m);
163 return IPPROTO_DONE;
164 }
165 #endif
166
167 /* Be proactive about malicious use of IPv4 mapped address */
168 if (IN6_IS_ADDR_V4MAPPED(&ip6->ip6_src) ||
169 IN6_IS_ADDR_V4MAPPED(&ip6->ip6_dst)) {
170 /* XXX stat */
171 m_freem(m);
172 return IPPROTO_DONE;
173 }
174
175 sockaddr_in6_init(&rip6src, &ip6->ip6_src, 0, 0, 0);
176 if (sa6_recoverscope(&rip6src) != 0) {
177 /* XXX: should be impossible. */
178 m_freem(m);
179 return IPPROTO_DONE;
180 }
181
182 CIRCLEQ_FOREACH(inph, &raw6cbtable.inpt_queue, inph_queue) {
183 in6p = (struct in6pcb *)inph;
184 if (in6p->in6p_af != AF_INET6)
185 continue;
186 if (in6p->in6p_ip6.ip6_nxt &&
187 in6p->in6p_ip6.ip6_nxt != proto)
188 continue;
189 if (!IN6_IS_ADDR_UNSPECIFIED(&in6p->in6p_laddr) &&
190 !IN6_ARE_ADDR_EQUAL(&in6p->in6p_laddr, &ip6->ip6_dst))
191 continue;
192 if (!IN6_IS_ADDR_UNSPECIFIED(&in6p->in6p_faddr) &&
193 !IN6_ARE_ADDR_EQUAL(&in6p->in6p_faddr, &ip6->ip6_src))
194 continue;
195 if (in6p->in6p_cksum != -1) {
196 RIP6_STATINC(RIP6_STAT_ISUM);
197 if (in6_cksum(m, proto, *offp,
198 m->m_pkthdr.len - *offp)) {
199 RIP6_STATINC(RIP6_STAT_BADSUM);
200 continue;
201 }
202 }
203 if (last) {
204 struct mbuf *n;
205
206 #ifdef IPSEC
207 /*
208 * Check AH/ESP integrity.
209 */
210 if (ipsec6_in_reject(m, last)) {
211 IPSEC6_STATINC(IPSEC_STAT_IN_INVAL);
212 /* do not inject data into pcb */
213 } else
214 #endif /* IPSEC */
215 #ifdef FAST_IPSEC
216 /*
217 * Check AH/ESP integrity
218 */
219 if (!ipsec6_in_reject(m,last))
220 #endif /* FAST_IPSEC */
221 if ((n = m_copy(m, 0, (int)M_COPYALL)) != NULL) {
222 if (last->in6p_flags & IN6P_CONTROLOPTS)
223 ip6_savecontrol(last, &opts, ip6, n);
224 /* strip intermediate headers */
225 m_adj(n, *offp);
226 if (sbappendaddr(&last->in6p_socket->so_rcv,
227 (struct sockaddr *)&rip6src, n, opts) == 0) {
228 /* should notify about lost packet */
229 m_freem(n);
230 if (opts)
231 m_freem(opts);
232 RIP6_STATINC(RIP6_STAT_FULLSOCK);
233 } else
234 sorwakeup(last->in6p_socket);
235 opts = NULL;
236 }
237 }
238 last = in6p;
239 }
240 #ifdef IPSEC
241 /*
242 * Check AH/ESP integrity.
243 */
244 if (last && ipsec6_in_reject(m, last)) {
245 m_freem(m);
246 IPSEC6_STATINC(IPSEC_STAT_IN_INVAL);
247 IP6_STATDEC(IP6_STAT_DELIVERED);
248 /* do not inject data into pcb */
249 } else
250 #endif /* IPSEC */
251 #ifdef FAST_IPSEC
252 if (last && ipsec6_in_reject(m, last)) {
253 m_freem(m);
254 /*
255 * XXX ipsec6_in_reject update stat if there is an error
256 * so we just need to update stats by hand in the case of last is
257 * NULL
258 */
259 if (!last)
260 IPSEC6_STATINC(IPSEC_STAT_IN_POLVIO);
261 IP6_STATDEC(IP6_STAT_DELIVERED);
262 /* do not inject data into pcb */
263 } else
264 #endif /* FAST_IPSEC */
265 if (last) {
266 if (last->in6p_flags & IN6P_CONTROLOPTS)
267 ip6_savecontrol(last, &opts, ip6, m);
268 /* strip intermediate headers */
269 m_adj(m, *offp);
270 if (sbappendaddr(&last->in6p_socket->so_rcv,
271 (struct sockaddr *)&rip6src, m, opts) == 0) {
272 m_freem(m);
273 if (opts)
274 m_freem(opts);
275 RIP6_STATINC(RIP6_STAT_FULLSOCK);
276 } else
277 sorwakeup(last->in6p_socket);
278 } else {
279 RIP6_STATINC(RIP6_STAT_NOSOCK);
280 if (m->m_flags & M_MCAST)
281 RIP6_STATINC(RIP6_STAT_NOSOCKMCAST);
282 if (proto == IPPROTO_NONE)
283 m_freem(m);
284 else {
285 u_int8_t *prvnxtp = ip6_get_prevhdr(m, *offp); /* XXX */
286 in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_protounknown);
287 icmp6_error(m, ICMP6_PARAM_PROB,
288 ICMP6_PARAMPROB_NEXTHEADER,
289 prvnxtp - mtod(m, u_int8_t *));
290 }
291 IP6_STATDEC(IP6_STAT_DELIVERED);
292 }
293 return IPPROTO_DONE;
294 }
295
296 void
297 rip6_ctlinput(int cmd, const struct sockaddr *sa, void *d)
298 {
299 struct ip6_hdr *ip6;
300 struct ip6ctlparam *ip6cp = NULL;
301 const struct sockaddr_in6 *sa6_src = NULL;
302 void *cmdarg;
303 void (*notify)(struct in6pcb *, int) = in6_rtchange;
304 int nxt;
305
306 if (sa->sa_family != AF_INET6 ||
307 sa->sa_len != sizeof(struct sockaddr_in6))
308 return;
309
310 if ((unsigned)cmd >= PRC_NCMDS)
311 return;
312 if (PRC_IS_REDIRECT(cmd))
313 notify = in6_rtchange, d = NULL;
314 else if (cmd == PRC_HOSTDEAD)
315 d = NULL;
316 else if (cmd == PRC_MSGSIZE)
317 ; /* special code is present, see below */
318 else if (inet6ctlerrmap[cmd] == 0)
319 return;
320
321 /* if the parameter is from icmp6, decode it. */
322 if (d != NULL) {
323 ip6cp = (struct ip6ctlparam *)d;
324 ip6 = ip6cp->ip6c_ip6;
325 cmdarg = ip6cp->ip6c_cmdarg;
326 sa6_src = ip6cp->ip6c_src;
327 nxt = ip6cp->ip6c_nxt;
328 } else {
329 ip6 = NULL;
330 cmdarg = NULL;
331 sa6_src = &sa6_any;
332 nxt = -1;
333 }
334
335 if (ip6 && cmd == PRC_MSGSIZE) {
336 const struct sockaddr_in6 *sa6 = (const struct sockaddr_in6 *)sa;
337 int valid = 0;
338 struct in6pcb *in6p;
339
340 /*
341 * Check to see if we have a valid raw IPv6 socket
342 * corresponding to the address in the ICMPv6 message
343 * payload, and the protocol (ip6_nxt) meets the socket.
344 * XXX chase extension headers, or pass final nxt value
345 * from icmp6_notify_error()
346 */
347 in6p = NULL;
348 in6p = in6_pcblookup_connect(&raw6cbtable, &sa6->sin6_addr, 0,
349 (const struct in6_addr *)&sa6_src->sin6_addr, 0, 0);
350 #if 0
351 if (!in6p) {
352 /*
353 * As the use of sendto(2) is fairly popular,
354 * we may want to allow non-connected pcb too.
355 * But it could be too weak against attacks...
356 * We should at least check if the local
357 * address (= s) is really ours.
358 */
359 in6p = in6_pcblookup_bind(&raw6cbtable,
360 &sa6->sin6_addr, 0, 0);
361 }
362 #endif
363
364 if (in6p && in6p->in6p_ip6.ip6_nxt &&
365 in6p->in6p_ip6.ip6_nxt == nxt)
366 valid++;
367
368 /*
369 * Depending on the value of "valid" and routing table
370 * size (mtudisc_{hi,lo}wat), we will:
371 * - recalculate the new MTU and create the
372 * corresponding routing entry, or
373 * - ignore the MTU change notification.
374 */
375 icmp6_mtudisc_update((struct ip6ctlparam *)d, valid);
376
377 /*
378 * regardless of if we called icmp6_mtudisc_update(),
379 * we need to call in6_pcbnotify(), to notify path MTU
380 * change to the userland (RFC3542), because some
381 * unconnected sockets may share the same destination
382 * and want to know the path MTU.
383 */
384 }
385
386 (void) in6_pcbnotify(&raw6cbtable, sa, 0,
387 (const struct sockaddr *)sa6_src, 0, cmd, cmdarg, notify);
388 }
389
390 /*
391 * Generate IPv6 header and pass packet to ip6_output.
392 * Tack on options user may have setup with control call.
393 */
394 int
395 rip6_output(struct mbuf *m, struct socket *so, struct sockaddr_in6 *dstsock,
396 struct mbuf *control)
397 {
398 struct in6_addr *dst;
399 struct ip6_hdr *ip6;
400 struct in6pcb *in6p;
401 u_int plen = m->m_pkthdr.len;
402 int error = 0;
403 struct ip6_pktopts opt, *optp = NULL;
404 struct ifnet *oifp = NULL;
405 int type, code; /* for ICMPv6 output statistics only */
406 int priv = 0;
407 int scope_ambiguous = 0;
408 struct in6_addr *in6a;
409
410 in6p = sotoin6pcb(so);
411
412 priv = 0;
413 if (curlwp && !kauth_authorize_generic(curlwp->l_cred,
414 KAUTH_GENERIC_ISSUSER, NULL))
415 priv = 1;
416
417 dst = &dstsock->sin6_addr;
418 if (control) {
419 if ((error = ip6_setpktopts(control, &opt,
420 in6p->in6p_outputopts,
421 priv, so->so_proto->pr_protocol)) != 0) {
422 goto bad;
423 }
424 optp = &opt;
425 } else
426 optp = in6p->in6p_outputopts;
427
428 /*
429 * Check and convert scope zone ID into internal form.
430 * XXX: we may still need to determine the zone later.
431 */
432 if (!(so->so_state & SS_ISCONNECTED)) {
433 if (dstsock->sin6_scope_id == 0 && !ip6_use_defzone)
434 scope_ambiguous = 1;
435 if ((error = sa6_embedscope(dstsock, ip6_use_defzone)) != 0)
436 goto bad;
437 }
438
439 /*
440 * For an ICMPv6 packet, we should know its type and code
441 * to update statistics.
442 */
443 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) {
444 struct icmp6_hdr *icmp6;
445 if (m->m_len < sizeof(struct icmp6_hdr) &&
446 (m = m_pullup(m, sizeof(struct icmp6_hdr))) == NULL) {
447 error = ENOBUFS;
448 goto bad;
449 }
450 icmp6 = mtod(m, struct icmp6_hdr *);
451 type = icmp6->icmp6_type;
452 code = icmp6->icmp6_code;
453 } else {
454 type = 0;
455 code = 0;
456 }
457
458 M_PREPEND(m, sizeof(*ip6), M_DONTWAIT);
459 if (!m) {
460 error = ENOBUFS;
461 goto bad;
462 }
463 ip6 = mtod(m, struct ip6_hdr *);
464
465 /*
466 * Next header might not be ICMP6 but use its pseudo header anyway.
467 */
468 ip6->ip6_dst = *dst;
469
470 /*
471 * Source address selection.
472 */
473 if ((in6a = in6_selectsrc(dstsock, optp, in6p->in6p_moptions,
474 (struct route *)&in6p->in6p_route, &in6p->in6p_laddr, &oifp,
475 &error)) == 0) {
476 if (error == 0)
477 error = EADDRNOTAVAIL;
478 goto bad;
479 }
480 ip6->ip6_src = *in6a;
481
482 if (oifp && scope_ambiguous) {
483 /*
484 * Application should provide a proper zone ID or the use of
485 * default zone IDs should be enabled. Unfortunately, some
486 * applications do not behave as it should, so we need a
487 * workaround. Even if an appropriate ID is not determined
488 * (when it's required), if we can determine the outgoing
489 * interface. determine the zone ID based on the interface.
490 */
491 error = in6_setscope(&dstsock->sin6_addr, oifp, NULL);
492 if (error != 0)
493 goto bad;
494 }
495 ip6->ip6_dst = dstsock->sin6_addr;
496
497 /* fill in the rest of the IPv6 header fields */
498 ip6->ip6_flow = in6p->in6p_flowinfo & IPV6_FLOWINFO_MASK;
499 ip6->ip6_vfc &= ~IPV6_VERSION_MASK;
500 ip6->ip6_vfc |= IPV6_VERSION;
501 /* ip6_plen will be filled in ip6_output, so not fill it here. */
502 ip6->ip6_nxt = in6p->in6p_ip6.ip6_nxt;
503 ip6->ip6_hlim = in6_selecthlim(in6p, oifp);
504
505 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6 ||
506 in6p->in6p_cksum != -1) {
507 int off;
508 u_int16_t sum;
509
510 /* compute checksum */
511 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6)
512 off = offsetof(struct icmp6_hdr, icmp6_cksum);
513 else
514 off = in6p->in6p_cksum;
515 if (plen < off + 1) {
516 error = EINVAL;
517 goto bad;
518 }
519 off += sizeof(struct ip6_hdr);
520
521 sum = 0;
522 m = m_copyback_cow(m, off, sizeof(sum), (void *)&sum,
523 M_DONTWAIT);
524 if (m == NULL) {
525 error = ENOBUFS;
526 goto bad;
527 }
528 sum = in6_cksum(m, ip6->ip6_nxt, sizeof(*ip6), plen);
529 m = m_copyback_cow(m, off, sizeof(sum), (void *)&sum,
530 M_DONTWAIT);
531 if (m == NULL) {
532 error = ENOBUFS;
533 goto bad;
534 }
535 }
536
537 error = ip6_output(m, optp, &in6p->in6p_route, 0,
538 in6p->in6p_moptions, so, &oifp);
539 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) {
540 if (oifp)
541 icmp6_ifoutstat_inc(oifp, type, code);
542 ICMP6_STATINC(ICMP6_STAT_OUTHIST + type);
543 } else
544 RIP6_STATINC(RIP6_STAT_OPACKETS);
545
546 goto freectl;
547
548 bad:
549 if (m)
550 m_freem(m);
551
552 freectl:
553 if (control) {
554 ip6_clearpktopts(&opt, -1);
555 m_freem(control);
556 }
557 return error;
558 }
559
560 /*
561 * Raw IPv6 socket option processing.
562 */
563 int
564 rip6_ctloutput(int op, struct socket *so, int level, int optname,
565 struct mbuf **mp)
566 {
567 int error = 0;
568
569 if (level == SOL_SOCKET && optname == SO_NOHEADER) {
570 /* need to fiddle w/ opt(IPPROTO_IPV6, IPV6_CHECKSUM)? */
571 if (op == PRCO_GETOPT) {
572 *mp = m_intopt(so, 1);
573 return 0;
574 } else if (*mp == NULL || (*mp)->m_len != sizeof(int))
575 error = EINVAL;
576 else if (*mtod(*mp, int *) == 0)
577 error = EINVAL;
578 goto free_m;
579 } else if (level != IPPROTO_IPV6)
580 return ip6_ctloutput(op, so, level, optname, mp);
581
582 switch (optname) {
583 case MRT6_INIT:
584 case MRT6_DONE:
585 case MRT6_ADD_MIF:
586 case MRT6_DEL_MIF:
587 case MRT6_ADD_MFC:
588 case MRT6_DEL_MFC:
589 case MRT6_PIM:
590 if (op == PRCO_SETOPT)
591 error = ip6_mrouter_set(optname, so, *mp);
592 else if (op == PRCO_GETOPT)
593 error = ip6_mrouter_get(optname, so, mp);
594 else
595 error = EINVAL;
596 break;
597 case IPV6_CHECKSUM:
598 return ip6_raw_ctloutput(op, so, level, optname, mp);
599 default:
600 return ip6_ctloutput(op, so, level, optname, mp);
601 }
602 free_m:
603 if (op == PRCO_SETOPT && *mp != NULL)
604 m_free(*mp);
605 return error;
606 }
607
608 extern u_long rip6_sendspace;
609 extern u_long rip6_recvspace;
610
611 int
612 rip6_usrreq(struct socket *so, int req, struct mbuf *m,
613 struct mbuf *nam, struct mbuf *control, struct lwp *l)
614 {
615 struct in6pcb *in6p = sotoin6pcb(so);
616 int s;
617 int error = 0;
618 int priv;
619
620 priv = 0;
621 if (l && !kauth_authorize_generic(l->l_cred,
622 KAUTH_GENERIC_ISSUSER, NULL))
623 priv++;
624
625 if (req == PRU_CONTROL)
626 return in6_control(so, (u_long)m, (void *)nam,
627 (struct ifnet *)control, l);
628
629 if (req == PRU_PURGEIF) {
630 in6_pcbpurgeif0(&raw6cbtable, (struct ifnet *)control);
631 in6_purgeif((struct ifnet *)control);
632 in6_pcbpurgeif(&raw6cbtable, (struct ifnet *)control);
633 return 0;
634 }
635
636 switch (req) {
637 case PRU_ATTACH:
638 if (in6p != NULL)
639 panic("rip6_attach");
640 if (!priv) {
641 error = EACCES;
642 break;
643 }
644 s = splsoftnet();
645 error = soreserve(so, rip6_sendspace, rip6_recvspace);
646 if (error != 0) {
647 splx(s);
648 break;
649 }
650 if ((error = in6_pcballoc(so, &raw6cbtable)) != 0) {
651 splx(s);
652 break;
653 }
654 splx(s);
655 in6p = sotoin6pcb(so);
656 in6p->in6p_ip6.ip6_nxt = (long)nam;
657 in6p->in6p_cksum = -1;
658
659 MALLOC(in6p->in6p_icmp6filt, struct icmp6_filter *,
660 sizeof(struct icmp6_filter), M_PCB, M_NOWAIT);
661 if (in6p->in6p_icmp6filt == NULL) {
662 in6_pcbdetach(in6p);
663 error = ENOMEM;
664 break;
665 }
666 ICMP6_FILTER_SETPASSALL(in6p->in6p_icmp6filt);
667 break;
668
669 case PRU_DISCONNECT:
670 if ((so->so_state & SS_ISCONNECTED) == 0) {
671 error = ENOTCONN;
672 break;
673 }
674 in6p->in6p_faddr = in6addr_any;
675 so->so_state &= ~SS_ISCONNECTED; /* XXX */
676 break;
677
678 case PRU_ABORT:
679 soisdisconnected(so);
680 /* Fallthrough */
681 case PRU_DETACH:
682 if (in6p == NULL)
683 panic("rip6_detach");
684 if (so == ip6_mrouter)
685 ip6_mrouter_done();
686 /* xxx: RSVP */
687 if (in6p->in6p_icmp6filt != NULL) {
688 FREE(in6p->in6p_icmp6filt, M_PCB);
689 in6p->in6p_icmp6filt = NULL;
690 }
691 in6_pcbdetach(in6p);
692 break;
693
694 case PRU_BIND:
695 {
696 struct sockaddr_in6 *addr = mtod(nam, struct sockaddr_in6 *);
697 struct ifaddr *ia = NULL;
698
699 if (nam->m_len != sizeof(*addr)) {
700 error = EINVAL;
701 break;
702 }
703 if (TAILQ_EMPTY(&ifnet) || addr->sin6_family != AF_INET6) {
704 error = EADDRNOTAVAIL;
705 break;
706 }
707 if ((error = sa6_embedscope(addr, ip6_use_defzone)) != 0)
708 break;
709
710 /*
711 * we don't support mapped address here, it would confuse
712 * users so reject it
713 */
714 if (IN6_IS_ADDR_V4MAPPED(&addr->sin6_addr)) {
715 error = EADDRNOTAVAIL;
716 break;
717 }
718 if (!IN6_IS_ADDR_UNSPECIFIED(&addr->sin6_addr) &&
719 (ia = ifa_ifwithaddr((struct sockaddr *)addr)) == 0) {
720 error = EADDRNOTAVAIL;
721 break;
722 }
723 if (ia && ((struct in6_ifaddr *)ia)->ia6_flags &
724 (IN6_IFF_ANYCAST|IN6_IFF_NOTREADY|
725 IN6_IFF_DETACHED|IN6_IFF_DEPRECATED)) {
726 error = EADDRNOTAVAIL;
727 break;
728 }
729 in6p->in6p_laddr = addr->sin6_addr;
730 break;
731 }
732
733 case PRU_CONNECT:
734 {
735 struct sockaddr_in6 *addr = mtod(nam, struct sockaddr_in6 *);
736 struct in6_addr *in6a = NULL;
737 struct ifnet *ifp = NULL;
738 int scope_ambiguous = 0;
739
740 if (nam->m_len != sizeof(*addr)) {
741 error = EINVAL;
742 break;
743 }
744 if (TAILQ_EMPTY(&ifnet)) {
745 error = EADDRNOTAVAIL;
746 break;
747 }
748 if (addr->sin6_family != AF_INET6) {
749 error = EAFNOSUPPORT;
750 break;
751 }
752
753 /*
754 * Application should provide a proper zone ID or the use of
755 * default zone IDs should be enabled. Unfortunately, some
756 * applications do not behave as it should, so we need a
757 * workaround. Even if an appropriate ID is not determined,
758 * we'll see if we can determine the outgoing interface. If we
759 * can, determine the zone ID based on the interface below.
760 */
761 if (addr->sin6_scope_id == 0 && !ip6_use_defzone)
762 scope_ambiguous = 1;
763 if ((error = sa6_embedscope(addr, ip6_use_defzone)) != 0)
764 return error;
765
766 /* Source address selection. XXX: need pcblookup? */
767 in6a = in6_selectsrc(addr, in6p->in6p_outputopts,
768 in6p->in6p_moptions, (struct route *)&in6p->in6p_route,
769 &in6p->in6p_laddr, &ifp, &error);
770 if (in6a == NULL) {
771 if (error == 0)
772 error = EADDRNOTAVAIL;
773 break;
774 }
775 /* XXX: see above */
776 if (ifp && scope_ambiguous &&
777 (error = in6_setscope(&addr->sin6_addr, ifp, NULL)) != 0) {
778 break;
779 }
780 in6p->in6p_laddr = *in6a;
781 in6p->in6p_faddr = addr->sin6_addr;
782 soisconnected(so);
783 break;
784 }
785
786 case PRU_CONNECT2:
787 error = EOPNOTSUPP;
788 break;
789
790 /*
791 * Mark the connection as being incapable of futther input.
792 */
793 case PRU_SHUTDOWN:
794 socantsendmore(so);
795 break;
796 /*
797 * Ship a packet out. The appropriate raw output
798 * routine handles any messaging necessary.
799 */
800 case PRU_SEND:
801 {
802 struct sockaddr_in6 tmp;
803 struct sockaddr_in6 *dst;
804
805 /* always copy sockaddr to avoid overwrites */
806 if (so->so_state & SS_ISCONNECTED) {
807 if (nam) {
808 error = EISCONN;
809 break;
810 }
811 /* XXX */
812 sockaddr_in6_init(&tmp, &in6p->in6p_faddr, 0, 0, 0);
813 dst = &tmp;
814 } else {
815 if (nam == NULL) {
816 error = ENOTCONN;
817 break;
818 }
819 if (nam->m_len != sizeof(tmp)) {
820 error = EINVAL;
821 break;
822 }
823
824 tmp = *mtod(nam, struct sockaddr_in6 *);
825 dst = &tmp;
826
827 if (dst->sin6_family != AF_INET6) {
828 error = EAFNOSUPPORT;
829 break;
830 }
831 }
832 error = rip6_output(m, so, dst, control);
833 m = NULL;
834 break;
835 }
836
837 case PRU_SENSE:
838 /*
839 * stat: don't bother with a blocksize
840 */
841 return 0;
842 /*
843 * Not supported.
844 */
845 case PRU_RCVOOB:
846 case PRU_RCVD:
847 case PRU_LISTEN:
848 case PRU_ACCEPT:
849 case PRU_SENDOOB:
850 error = EOPNOTSUPP;
851 break;
852
853 case PRU_SOCKADDR:
854 in6_setsockaddr(in6p, nam);
855 break;
856
857 case PRU_PEERADDR:
858 in6_setpeeraddr(in6p, nam);
859 break;
860
861 default:
862 panic("rip6_usrreq");
863 }
864 if (m != NULL)
865 m_freem(m);
866 return error;
867 }
868
869 static int
870 sysctl_net_inet6_raw6_stats(SYSCTLFN_ARGS)
871 {
872 netstat_sysctl_context ctx;
873 uint64_t rip6s[RIP6_NSTATS];
874
875 ctx.ctx_stat = rip6stat_percpu;
876 ctx.ctx_counters = rip6s;
877 ctx.ctx_ncounters = RIP6_NSTATS;
878 return (NETSTAT_SYSCTL(&ctx));
879 }
880
881 SYSCTL_SETUP(sysctl_net_inet6_raw6_setup, "sysctl net.inet6.raw6 subtree setup")
882 {
883
884 sysctl_createv(clog, 0, NULL, NULL,
885 CTLFLAG_PERMANENT,
886 CTLTYPE_NODE, "net", NULL,
887 NULL, 0, NULL, 0,
888 CTL_NET, CTL_EOL);
889 sysctl_createv(clog, 0, NULL, NULL,
890 CTLFLAG_PERMANENT,
891 CTLTYPE_NODE, "inet6", NULL,
892 NULL, 0, NULL, 0,
893 CTL_NET, PF_INET6, CTL_EOL);
894 sysctl_createv(clog, 0, NULL, NULL,
895 CTLFLAG_PERMANENT,
896 CTLTYPE_NODE, "raw6",
897 SYSCTL_DESCR("Raw IPv6 settings"),
898 NULL, 0, NULL, 0,
899 CTL_NET, PF_INET6, IPPROTO_RAW, CTL_EOL);
900
901 sysctl_createv(clog, 0, NULL, NULL,
902 CTLFLAG_PERMANENT,
903 CTLTYPE_STRUCT, "pcblist",
904 SYSCTL_DESCR("Raw IPv6 control block list"),
905 sysctl_inpcblist, 0, &raw6cbtable, 0,
906 CTL_NET, PF_INET6, IPPROTO_RAW,
907 CTL_CREATE, CTL_EOL);
908 sysctl_createv(clog, 0, NULL, NULL,
909 CTLFLAG_PERMANENT,
910 CTLTYPE_STRUCT, "stats",
911 SYSCTL_DESCR("Raw IPv6 statistics"),
912 sysctl_net_inet6_raw6_stats, 0, NULL, 0,
913 CTL_NET, PF_INET6, IPPROTO_RAW, RAW6CTL_STATS,
914 CTL_EOL);
915 }
916