1 /*-
2 * SPDX-License-Identifier: BSD-3-Clause
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) 1982, 1986, 1988, 1993
34 * The Regents of the University of California.
35 * 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 __FBSDID("$FreeBSD$");
66
67 #include "opt_ipsec.h"
68 #include "opt_inet6.h"
69 #include "opt_route.h"
70
71 #include <sys/param.h>
72 #include <sys/errno.h>
73 #include <sys/jail.h>
74 #include <sys/kernel.h>
75 #include <sys/lock.h>
76 #include <sys/malloc.h>
77 #include <sys/mbuf.h>
78 #include <sys/priv.h>
79 #include <sys/proc.h>
80 #include <sys/protosw.h>
81 #include <sys/signalvar.h>
82 #include <sys/socket.h>
83 #include <sys/socketvar.h>
84 #include <sys/sx.h>
85 #include <sys/syslog.h>
86
87 #include <net/if.h>
88 #include <net/if_var.h>
89 #include <net/if_types.h>
90 #include <net/route.h>
91 #include <net/vnet.h>
92
93 #include <netinet/in.h>
94 #include <netinet/in_var.h>
95 #include <netinet/in_systm.h>
96 #include <netinet/in_pcb.h>
97
98 #include <netinet/icmp6.h>
99 #include <netinet/ip6.h>
100 #include <netinet/ip_var.h>
101 #include <netinet6/ip6protosw.h>
102 #include <netinet6/ip6_mroute.h>
103 #include <netinet6/in6_pcb.h>
104 #include <netinet6/ip6_var.h>
105 #include <netinet6/nd6.h>
106 #include <netinet6/raw_ip6.h>
107 #include <netinet6/in6_fib.h>
108 #include <netinet6/scope6_var.h>
109 #include <netinet6/send.h>
110
111 #include <netipsec/ipsec_support.h>
112
113 #include <machine/stdarg.h>
114
115 #define satosin6(sa) ((struct sockaddr_in6 *)(sa))
116 #define ifatoia6(ifa) ((struct in6_ifaddr *)(ifa))
117
118 /*
119 * Raw interface to IP6 protocol.
120 */
121
122 VNET_DECLARE(struct inpcbhead, ripcb);
123 VNET_DECLARE(struct inpcbinfo, ripcbinfo);
124 #define V_ripcb VNET(ripcb)
125 #define V_ripcbinfo VNET(ripcbinfo)
126
127 extern u_long rip_sendspace;
128 extern u_long rip_recvspace;
129
130 VNET_PCPUSTAT_DEFINE(struct rip6stat, rip6stat);
131 VNET_PCPUSTAT_SYSINIT(rip6stat);
132
133 #ifdef VIMAGE
134 VNET_PCPUSTAT_SYSUNINIT(rip6stat);
135 #endif /* VIMAGE */
136
137 /*
138 * Hooks for multicast routing. They all default to NULL, so leave them not
139 * initialized and rely on BSS being set to 0.
140 */
141
142 /*
143 * The socket used to communicate with the multicast routing daemon.
144 */
145 VNET_DEFINE(struct socket *, ip6_mrouter);
146
147 /*
148 * The various mrouter functions.
149 */
150 int (*ip6_mrouter_set)(struct socket *, struct sockopt *);
151 int (*ip6_mrouter_get)(struct socket *, struct sockopt *);
152 int (*ip6_mrouter_done)(void);
153 int (*ip6_mforward)(struct ip6_hdr *, struct ifnet *, struct mbuf *);
154 int (*mrt6_ioctl)(u_long, caddr_t);
155
156 /*
157 * Setup generic address and protocol structures for raw_input routine, then
158 * pass them along with mbuf chain.
159 */
160 int
161 rip6_input(struct mbuf **mp, int *offp, int proto)
162 {
163 struct ifnet *ifp;
164 struct mbuf *m = *mp;
165 struct ip6_hdr *ip6 = mtod(m, struct ip6_hdr *);
166 struct inpcb *inp;
167 struct inpcb *last = NULL;
168 struct mbuf *opts = NULL;
169 struct sockaddr_in6 fromsa;
170
171 NET_EPOCH_ASSERT();
172
173 RIP6STAT_INC(rip6s_ipackets);
174
175 init_sin6(&fromsa, m, 0); /* general init */
176
177 ifp = m->m_pkthdr.rcvif;
178
179 CK_LIST_FOREACH(inp, &V_ripcb, inp_list) {
180 /* XXX inp locking */
181 if ((inp->inp_vflag & INP_IPV6) == 0)
182 continue;
183 if (inp->inp_ip_p &&
184 inp->inp_ip_p != proto)
185 continue;
186 if (!IN6_IS_ADDR_UNSPECIFIED(&inp->in6p_laddr) &&
187 !IN6_ARE_ADDR_EQUAL(&inp->in6p_laddr, &ip6->ip6_dst))
188 continue;
189 if (!IN6_IS_ADDR_UNSPECIFIED(&inp->in6p_faddr) &&
190 !IN6_ARE_ADDR_EQUAL(&inp->in6p_faddr, &ip6->ip6_src))
191 continue;
192 if (last != NULL) {
193 struct mbuf *n = m_copym(m, 0, M_COPYALL, M_NOWAIT);
194
195 #if defined(IPSEC) || defined(IPSEC_SUPPORT)
196 /*
197 * Check AH/ESP integrity.
198 */
199 if (IPSEC_ENABLED(ipv6)) {
200 if (n != NULL &&
201 IPSEC_CHECK_POLICY(ipv6, n, last) != 0) {
202 m_freem(n);
203 /* Do not inject data into pcb. */
204 n = NULL;
205 }
206 }
207 #endif /* IPSEC */
208 if (n) {
209 if (last->inp_flags & INP_CONTROLOPTS ||
210 last->inp_socket->so_options & SO_TIMESTAMP)
211 ip6_savecontrol(last, n, &opts);
212 /* strip intermediate headers */
213 m_adj(n, *offp);
214 if (sbappendaddr(&last->inp_socket->so_rcv,
215 (struct sockaddr *)&fromsa,
216 n, opts) == 0) {
217 soroverflow(last->inp_socket);
218 m_freem(n);
219 if (opts)
220 m_freem(opts);
221 RIP6STAT_INC(rip6s_fullsock);
222 } else
223 sorwakeup(last->inp_socket);
224 opts = NULL;
225 }
226 INP_RUNLOCK(last);
227 last = NULL;
228 }
229 INP_RLOCK(inp);
230 if (__predict_false(inp->inp_flags2 & INP_FREED))
231 goto skip_2;
232 if (jailed_without_vnet(inp->inp_cred)) {
233 /*
234 * Allow raw socket in jail to receive multicast;
235 * assume process had PRIV_NETINET_RAW at attach,
236 * and fall through into normal filter path if so.
237 */
238 if (!IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst) &&
239 prison_check_ip6(inp->inp_cred,
240 &ip6->ip6_dst) != 0)
241 goto skip_2;
242 }
243 if (inp->in6p_cksum != -1) {
244 RIP6STAT_INC(rip6s_isum);
245 if (m->m_pkthdr.len - (*offp + inp->in6p_cksum) < 2 ||
246 in6_cksum(m, proto, *offp,
247 m->m_pkthdr.len - *offp)) {
248 RIP6STAT_INC(rip6s_badsum);
249 /*
250 * Drop the received message, don't send an
251 * ICMP6 message. Set proto to IPPROTO_NONE
252 * to achieve that.
253 */
254 proto = IPPROTO_NONE;
255 goto skip_2;
256 }
257 }
258 /*
259 * If this raw socket has multicast state, and we
260 * have received a multicast, check if this socket
261 * should receive it, as multicast filtering is now
262 * the responsibility of the transport layer.
263 */
264 if (inp->in6p_moptions &&
265 IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst)) {
266 /*
267 * If the incoming datagram is for MLD, allow it
268 * through unconditionally to the raw socket.
269 *
270 * Use the M_RTALERT_MLD flag to check for MLD
271 * traffic without having to inspect the mbuf chain
272 * more deeply, as all MLDv1/v2 host messages MUST
273 * contain the Router Alert option.
274 *
275 * In the case of MLDv1, we may not have explicitly
276 * joined the group, and may have set IFF_ALLMULTI
277 * on the interface. im6o_mc_filter() may discard
278 * control traffic we actually need to see.
279 *
280 * Userland multicast routing daemons should continue
281 * filter the control traffic appropriately.
282 */
283 int blocked;
284
285 blocked = MCAST_PASS;
286 if ((m->m_flags & M_RTALERT_MLD) == 0) {
287 struct sockaddr_in6 mcaddr;
288
289 bzero(&mcaddr, sizeof(struct sockaddr_in6));
290 mcaddr.sin6_len = sizeof(struct sockaddr_in6);
291 mcaddr.sin6_family = AF_INET6;
292 mcaddr.sin6_addr = ip6->ip6_dst;
293
294 blocked = im6o_mc_filter(inp->in6p_moptions,
295 ifp,
296 (struct sockaddr *)&mcaddr,
297 (struct sockaddr *)&fromsa);
298 }
299 if (blocked != MCAST_PASS) {
300 IP6STAT_INC(ip6s_notmember);
301 goto skip_2;
302 }
303 }
304 last = inp;
305 continue;
306 skip_2:
307 INP_RUNLOCK(inp);
308 }
309 #if defined(IPSEC) || defined(IPSEC_SUPPORT)
310 /*
311 * Check AH/ESP integrity.
312 */
313 if (IPSEC_ENABLED(ipv6) && last != NULL &&
314 IPSEC_CHECK_POLICY(ipv6, m, last) != 0) {
315 m_freem(m);
316 IP6STAT_DEC(ip6s_delivered);
317 /* Do not inject data into pcb. */
318 INP_RUNLOCK(last);
319 } else
320 #endif /* IPSEC */
321 if (last != NULL) {
322 if (last->inp_flags & INP_CONTROLOPTS ||
323 last->inp_socket->so_options & SO_TIMESTAMP)
324 ip6_savecontrol(last, m, &opts);
325 /* Strip intermediate headers. */
326 m_adj(m, *offp);
327 if (sbappendaddr(&last->inp_socket->so_rcv,
328 (struct sockaddr *)&fromsa, m, opts) == 0) {
329 soroverflow(last->inp_socket);
330 m_freem(m);
331 if (opts)
332 m_freem(opts);
333 RIP6STAT_INC(rip6s_fullsock);
334 } else
335 sorwakeup(last->inp_socket);
336 INP_RUNLOCK(last);
337 } else {
338 RIP6STAT_INC(rip6s_nosock);
339 if (m->m_flags & M_MCAST)
340 RIP6STAT_INC(rip6s_nosockmcast);
341 if (proto == IPPROTO_NONE)
342 m_freem(m);
343 else
344 icmp6_error(m, ICMP6_PARAM_PROB,
345 ICMP6_PARAMPROB_NEXTHEADER,
346 ip6_get_prevhdr(m, *offp));
347 IP6STAT_DEC(ip6s_delivered);
348 }
349 return (IPPROTO_DONE);
350 }
351
352 void
353 rip6_ctlinput(int cmd, struct sockaddr *sa, void *d)
354 {
355 struct ip6ctlparam *ip6cp = NULL;
356 const struct sockaddr_in6 *sa6_src = NULL;
357 void *cmdarg;
358 struct inpcb *(*notify)(struct inpcb *, int) = in6_rtchange;
359
360 if (sa->sa_family != AF_INET6 ||
361 sa->sa_len != sizeof(struct sockaddr_in6))
362 return;
363
364 if ((unsigned)cmd >= PRC_NCMDS)
365 return;
366 if (PRC_IS_REDIRECT(cmd))
367 notify = in6_rtchange, d = NULL;
368 else if (cmd == PRC_HOSTDEAD)
369 d = NULL;
370 else if (inet6ctlerrmap[cmd] == 0)
371 return;
372
373 /*
374 * If the parameter is from icmp6, decode it.
375 */
376 if (d != NULL) {
377 ip6cp = (struct ip6ctlparam *)d;
378 cmdarg = ip6cp->ip6c_cmdarg;
379 sa6_src = ip6cp->ip6c_src;
380 } else {
381 cmdarg = NULL;
382 sa6_src = &sa6_any;
383 }
384
385 (void) in6_pcbnotify(&V_ripcbinfo, sa, 0,
386 (const struct sockaddr *)sa6_src, 0, cmd, cmdarg, notify);
387 }
388
389 /*
390 * Generate IPv6 header and pass packet to ip6_output. Tack on options user
391 * may have setup with control call.
392 */
393 int
394 rip6_output(struct mbuf *m, struct socket *so, ...)
395 {
396 struct epoch_tracker et;
397 struct mbuf *control;
398 struct m_tag *mtag;
399 struct sockaddr_in6 *dstsock;
400 struct ip6_hdr *ip6;
401 struct inpcb *inp;
402 u_int plen = m->m_pkthdr.len;
403 int error = 0;
404 struct ip6_pktopts opt, *optp;
405 struct ifnet *oifp = NULL;
406 int type = 0, code = 0; /* for ICMPv6 output statistics only */
407 int scope_ambiguous = 0;
408 int use_defzone = 0;
409 int hlim = 0;
410 struct in6_addr in6a;
411 va_list ap;
412
413 va_start(ap, so);
414 dstsock = va_arg(ap, struct sockaddr_in6 *);
415 control = va_arg(ap, struct mbuf *);
416 va_end(ap);
417
418 inp = sotoinpcb(so);
419 INP_WLOCK(inp);
420
421 if (control != NULL) {
422 if ((error = ip6_setpktopts(control, &opt,
423 inp->in6p_outputopts, so->so_cred,
424 so->so_proto->pr_protocol)) != 0) {
425 goto bad;
426 }
427 optp = &opt;
428 } else
429 optp = inp->in6p_outputopts;
430
431 /*
432 * Check and convert scope zone ID into internal form.
433 *
434 * XXX: we may still need to determine the zone later.
435 */
436 if (!(so->so_state & SS_ISCONNECTED)) {
437 if (!optp || !optp->ip6po_pktinfo ||
438 !optp->ip6po_pktinfo->ipi6_ifindex)
439 use_defzone = V_ip6_use_defzone;
440 if (dstsock->sin6_scope_id == 0 && !use_defzone)
441 scope_ambiguous = 1;
442 if ((error = sa6_embedscope(dstsock, use_defzone)) != 0)
443 goto bad;
444 }
445
446 /*
447 * For an ICMPv6 packet, we should know its type and code to update
448 * statistics.
449 */
450 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) {
451 struct icmp6_hdr *icmp6;
452 if (m->m_len < sizeof(struct icmp6_hdr) &&
453 (m = m_pullup(m, sizeof(struct icmp6_hdr))) == NULL) {
454 error = ENOBUFS;
455 goto bad;
456 }
457 icmp6 = mtod(m, struct icmp6_hdr *);
458 type = icmp6->icmp6_type;
459 code = icmp6->icmp6_code;
460 }
461
462 M_PREPEND(m, sizeof(*ip6), M_NOWAIT);
463 if (m == NULL) {
464 error = ENOBUFS;
465 goto bad;
466 }
467 ip6 = mtod(m, struct ip6_hdr *);
468
469 #ifdef ROUTE_MPATH
470 if (CALC_FLOWID_OUTBOUND) {
471 uint32_t hash_type, hash_val;
472
473 hash_val = fib6_calc_software_hash(&inp->in6p_laddr,
474 &dstsock->sin6_addr, 0, 0, so->so_proto->pr_protocol,
475 &hash_type);
476 inp->inp_flowid = hash_val;
477 inp->inp_flowtype = hash_type;
478 }
479 #endif
480 /*
481 * Source address selection.
482 */
483 NET_EPOCH_ENTER(et);
484 error = in6_selectsrc_socket(dstsock, optp, inp, so->so_cred,
485 scope_ambiguous, &in6a, &hlim);
486 NET_EPOCH_EXIT(et);
487
488 if (error)
489 goto bad;
490 error = prison_check_ip6(inp->inp_cred, &in6a);
491 if (error != 0)
492 goto bad;
493 ip6->ip6_src = in6a;
494
495 ip6->ip6_dst = dstsock->sin6_addr;
496
497 /*
498 * Fill in the rest of the IPv6 header fields.
499 */
500 ip6->ip6_flow = (ip6->ip6_flow & ~IPV6_FLOWINFO_MASK) |
501 (inp->inp_flow & IPV6_FLOWINFO_MASK);
502 ip6->ip6_vfc = (ip6->ip6_vfc & ~IPV6_VERSION_MASK) |
503 (IPV6_VERSION & IPV6_VERSION_MASK);
504
505 /*
506 * ip6_plen will be filled in ip6_output, so not fill it here.
507 */
508 ip6->ip6_nxt = inp->inp_ip_p;
509 ip6->ip6_hlim = hlim;
510
511 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6 ||
512 inp->in6p_cksum != -1) {
513 struct mbuf *n;
514 int off;
515 u_int16_t *p;
516
517 /* Compute checksum. */
518 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6)
519 off = offsetof(struct icmp6_hdr, icmp6_cksum);
520 else
521 off = inp->in6p_cksum;
522 if (plen < off + 2) {
523 error = EINVAL;
524 goto bad;
525 }
526 off += sizeof(struct ip6_hdr);
527
528 n = m;
529 while (n && n->m_len <= off) {
530 off -= n->m_len;
531 n = n->m_next;
532 }
533 if (!n)
534 goto bad;
535 p = (u_int16_t *)(mtod(n, caddr_t) + off);
536 *p = 0;
537 *p = in6_cksum(m, ip6->ip6_nxt, sizeof(*ip6), plen);
538 }
539
540 /*
541 * Send RA/RS messages to user land for protection, before sending
542 * them to rtadvd/rtsol.
543 */
544 if ((send_sendso_input_hook != NULL) &&
545 so->so_proto->pr_protocol == IPPROTO_ICMPV6) {
546 switch (type) {
547 case ND_ROUTER_ADVERT:
548 case ND_ROUTER_SOLICIT:
549 mtag = m_tag_get(PACKET_TAG_ND_OUTGOING,
550 sizeof(unsigned short), M_NOWAIT);
551 if (mtag == NULL)
552 goto bad;
553 m_tag_prepend(m, mtag);
554 }
555 }
556
557 NET_EPOCH_ENTER(et);
558 error = ip6_output(m, optp, NULL, 0, inp->in6p_moptions, &oifp, inp);
559 NET_EPOCH_EXIT(et);
560 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) {
561 if (oifp)
562 icmp6_ifoutstat_inc(oifp, type, code);
563 ICMP6STAT_INC(icp6s_outhist[type]);
564 } else
565 RIP6STAT_INC(rip6s_opackets);
566
567 goto freectl;
568
569 bad:
570 if (m)
571 m_freem(m);
572
573 freectl:
574 if (control != NULL) {
575 ip6_clearpktopts(&opt, -1);
576 m_freem(control);
577 }
578 INP_WUNLOCK(inp);
579 return (error);
580 }
581
582 /*
583 * Raw IPv6 socket option processing.
584 */
585 int
586 rip6_ctloutput(struct socket *so, struct sockopt *sopt)
587 {
588 struct inpcb *inp;
589 int error;
590
591 if (sopt->sopt_level == IPPROTO_ICMPV6)
592 /*
593 * XXX: is it better to call icmp6_ctloutput() directly
594 * from protosw?
595 */
596 return (icmp6_ctloutput(so, sopt));
597 else if (sopt->sopt_level != IPPROTO_IPV6) {
598 if (sopt->sopt_level == SOL_SOCKET &&
599 sopt->sopt_name == SO_SETFIB) {
600 inp = sotoinpcb(so);
601 INP_WLOCK(inp);
602 inp->inp_inc.inc_fibnum = so->so_fibnum;
603 INP_WUNLOCK(inp);
604 return (0);
605 }
606 return (EINVAL);
607 }
608
609 error = 0;
610
611 switch (sopt->sopt_dir) {
612 case SOPT_GET:
613 switch (sopt->sopt_name) {
614 case MRT6_INIT:
615 case MRT6_DONE:
616 case MRT6_ADD_MIF:
617 case MRT6_DEL_MIF:
618 case MRT6_ADD_MFC:
619 case MRT6_DEL_MFC:
620 case MRT6_PIM:
621 error = ip6_mrouter_get ? ip6_mrouter_get(so, sopt) :
622 EOPNOTSUPP;
623 break;
624 case IPV6_CHECKSUM:
625 error = ip6_raw_ctloutput(so, sopt);
626 break;
627 default:
628 error = ip6_ctloutput(so, sopt);
629 break;
630 }
631 break;
632
633 case SOPT_SET:
634 switch (sopt->sopt_name) {
635 case MRT6_INIT:
636 case MRT6_DONE:
637 case MRT6_ADD_MIF:
638 case MRT6_DEL_MIF:
639 case MRT6_ADD_MFC:
640 case MRT6_DEL_MFC:
641 case MRT6_PIM:
642 error = ip6_mrouter_set ? ip6_mrouter_set(so, sopt) :
643 EOPNOTSUPP;
644 break;
645 case IPV6_CHECKSUM:
646 error = ip6_raw_ctloutput(so, sopt);
647 break;
648 default:
649 error = ip6_ctloutput(so, sopt);
650 break;
651 }
652 break;
653 }
654
655 return (error);
656 }
657
658 static int
659 rip6_attach(struct socket *so, int proto, struct thread *td)
660 {
661 struct inpcb *inp;
662 struct icmp6_filter *filter;
663 int error;
664
665 inp = sotoinpcb(so);
666 KASSERT(inp == NULL, ("rip6_attach: inp != NULL"));
667
668 error = priv_check(td, PRIV_NETINET_RAW);
669 if (error)
670 return (error);
671 error = soreserve(so, rip_sendspace, rip_recvspace);
672 if (error)
673 return (error);
674 filter = malloc(sizeof(struct icmp6_filter), M_PCB, M_NOWAIT);
675 if (filter == NULL)
676 return (ENOMEM);
677 INP_INFO_WLOCK(&V_ripcbinfo);
678 error = in_pcballoc(so, &V_ripcbinfo);
679 if (error) {
680 INP_INFO_WUNLOCK(&V_ripcbinfo);
681 free(filter, M_PCB);
682 return (error);
683 }
684 inp = (struct inpcb *)so->so_pcb;
685 INP_INFO_WUNLOCK(&V_ripcbinfo);
686 inp->inp_vflag |= INP_IPV6;
687 inp->inp_ip_p = (long)proto;
688 inp->in6p_hops = -1; /* use kernel default */
689 inp->in6p_cksum = -1;
690 inp->in6p_icmp6filt = filter;
691 ICMP6_FILTER_SETPASSALL(inp->in6p_icmp6filt);
692 INP_WUNLOCK(inp);
693 return (0);
694 }
695
696 static void
697 rip6_detach(struct socket *so)
698 {
699 struct inpcb *inp;
700
701 inp = sotoinpcb(so);
702 KASSERT(inp != NULL, ("rip6_detach: inp == NULL"));
703
704 if (so == V_ip6_mrouter && ip6_mrouter_done)
705 ip6_mrouter_done();
706 /* xxx: RSVP */
707 INP_INFO_WLOCK(&V_ripcbinfo);
708 INP_WLOCK(inp);
709 free(inp->in6p_icmp6filt, M_PCB);
710 in_pcbdetach(inp);
711 in_pcbfree(inp);
712 INP_INFO_WUNLOCK(&V_ripcbinfo);
713 }
714
715 /* XXXRW: This can't ever be called. */
716 static void
717 rip6_abort(struct socket *so)
718 {
719 struct inpcb *inp;
720
721 inp = sotoinpcb(so);
722 KASSERT(inp != NULL, ("rip6_abort: inp == NULL"));
723
724 soisdisconnected(so);
725 }
726
727 static void
728 rip6_close(struct socket *so)
729 {
730 struct inpcb *inp;
731
732 inp = sotoinpcb(so);
733 KASSERT(inp != NULL, ("rip6_close: inp == NULL"));
734
735 soisdisconnected(so);
736 }
737
738 static int
739 rip6_disconnect(struct socket *so)
740 {
741 struct inpcb *inp;
742
743 inp = sotoinpcb(so);
744 KASSERT(inp != NULL, ("rip6_disconnect: inp == NULL"));
745
746 if ((so->so_state & SS_ISCONNECTED) == 0)
747 return (ENOTCONN);
748 inp->in6p_faddr = in6addr_any;
749 rip6_abort(so);
750 return (0);
751 }
752
753 static int
754 rip6_bind(struct socket *so, struct sockaddr *nam, struct thread *td)
755 {
756 struct epoch_tracker et;
757 struct inpcb *inp;
758 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)nam;
759 struct ifaddr *ifa = NULL;
760 int error = 0;
761
762 inp = sotoinpcb(so);
763 KASSERT(inp != NULL, ("rip6_bind: inp == NULL"));
764
765 if (nam->sa_family != AF_INET6)
766 return (EAFNOSUPPORT);
767 if (nam->sa_len != sizeof(*addr))
768 return (EINVAL);
769 if ((error = prison_check_ip6(td->td_ucred, &addr->sin6_addr)) != 0)
770 return (error);
771 if (CK_STAILQ_EMPTY(&V_ifnet) || addr->sin6_family != AF_INET6)
772 return (EADDRNOTAVAIL);
773 if ((error = sa6_embedscope(addr, V_ip6_use_defzone)) != 0)
774 return (error);
775
776 NET_EPOCH_ENTER(et);
777 if (!IN6_IS_ADDR_UNSPECIFIED(&addr->sin6_addr) &&
778 (ifa = ifa_ifwithaddr((struct sockaddr *)addr)) == NULL) {
779 NET_EPOCH_EXIT(et);
780 return (EADDRNOTAVAIL);
781 }
782 if (ifa != NULL &&
783 ((struct in6_ifaddr *)ifa)->ia6_flags &
784 (IN6_IFF_ANYCAST|IN6_IFF_NOTREADY|
785 IN6_IFF_DETACHED|IN6_IFF_DEPRECATED)) {
786 NET_EPOCH_EXIT(et);
787 return (EADDRNOTAVAIL);
788 }
789 NET_EPOCH_EXIT(et);
790 INP_INFO_WLOCK(&V_ripcbinfo);
791 INP_WLOCK(inp);
792 inp->in6p_laddr = addr->sin6_addr;
793 INP_WUNLOCK(inp);
794 INP_INFO_WUNLOCK(&V_ripcbinfo);
795 return (0);
796 }
797
798 static int
799 rip6_connect(struct socket *so, struct sockaddr *nam, struct thread *td)
800 {
801 struct inpcb *inp;
802 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)nam;
803 struct in6_addr in6a;
804 struct epoch_tracker et;
805 int error = 0, scope_ambiguous = 0;
806
807 inp = sotoinpcb(so);
808 KASSERT(inp != NULL, ("rip6_connect: inp == NULL"));
809
810 if (nam->sa_len != sizeof(*addr))
811 return (EINVAL);
812 if (CK_STAILQ_EMPTY(&V_ifnet))
813 return (EADDRNOTAVAIL);
814 if (addr->sin6_family != AF_INET6)
815 return (EAFNOSUPPORT);
816
817 /*
818 * Application should provide a proper zone ID or the use of default
819 * zone IDs should be enabled. Unfortunately, some applications do
820 * not behave as it should, so we need a workaround. Even if an
821 * appropriate ID is not determined, we'll see if we can determine
822 * the outgoing interface. If we can, determine the zone ID based on
823 * the interface below.
824 */
825 if (addr->sin6_scope_id == 0 && !V_ip6_use_defzone)
826 scope_ambiguous = 1;
827 if ((error = sa6_embedscope(addr, V_ip6_use_defzone)) != 0)
828 return (error);
829
830 INP_INFO_WLOCK(&V_ripcbinfo);
831 INP_WLOCK(inp);
832 /* Source address selection. XXX: need pcblookup? */
833 NET_EPOCH_ENTER(et);
834 error = in6_selectsrc_socket(addr, inp->in6p_outputopts,
835 inp, so->so_cred, scope_ambiguous, &in6a, NULL);
836 NET_EPOCH_EXIT(et);
837 if (error) {
838 INP_WUNLOCK(inp);
839 INP_INFO_WUNLOCK(&V_ripcbinfo);
840 return (error);
841 }
842
843 inp->in6p_faddr = addr->sin6_addr;
844 inp->in6p_laddr = in6a;
845 soisconnected(so);
846 INP_WUNLOCK(inp);
847 INP_INFO_WUNLOCK(&V_ripcbinfo);
848 return (0);
849 }
850
851 static int
852 rip6_shutdown(struct socket *so)
853 {
854 struct inpcb *inp;
855
856 inp = sotoinpcb(so);
857 KASSERT(inp != NULL, ("rip6_shutdown: inp == NULL"));
858
859 INP_WLOCK(inp);
860 socantsendmore(so);
861 INP_WUNLOCK(inp);
862 return (0);
863 }
864
865 static int
866 rip6_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *nam,
867 struct mbuf *control, struct thread *td)
868 {
869 struct inpcb *inp;
870 struct sockaddr_in6 tmp;
871 struct sockaddr_in6 *dst;
872 int error;
873
874 inp = sotoinpcb(so);
875 KASSERT(inp != NULL, ("rip6_send: inp == NULL"));
876
877 /* Always copy sockaddr to avoid overwrites. */
878 /* Unlocked read. */
879 if (so->so_state & SS_ISCONNECTED) {
880 if (nam) {
881 error = EISCONN;
882 goto release;
883 }
884 /* XXX */
885 bzero(&tmp, sizeof(tmp));
886 tmp.sin6_family = AF_INET6;
887 tmp.sin6_len = sizeof(struct sockaddr_in6);
888 INP_RLOCK(inp);
889 bcopy(&inp->in6p_faddr, &tmp.sin6_addr,
890 sizeof(struct in6_addr));
891 INP_RUNLOCK(inp);
892 dst = &tmp;
893 } else {
894 error = 0;
895 if (nam == NULL)
896 error = ENOTCONN;
897 else if (nam->sa_family != AF_INET6)
898 error = EAFNOSUPPORT;
899 else if (nam->sa_len != sizeof(struct sockaddr_in6))
900 error = EINVAL;
901 if (error != 0)
902 goto release;
903 tmp = *(struct sockaddr_in6 *)nam;
904 dst = &tmp;
905
906 if (dst->sin6_family == AF_UNSPEC) {
907 /*
908 * XXX: we allow this case for backward
909 * compatibility to buggy applications that
910 * rely on old (and wrong) kernel behavior.
911 */
912 log(LOG_INFO, "rip6 SEND: address family is "
913 "unspec. Assume AF_INET6\n");
914 dst->sin6_family = AF_INET6;
915 } else if (dst->sin6_family != AF_INET6) {
916 error = EAFNOSUPPORT;
917 goto release;
918 }
919 }
920 return (rip6_output(m, so, dst, control));
921
922 release:
923 if (control != NULL)
924 m_freem(control);
925 m_freem(m);
926 return (error);
927 }
928
929 struct pr_usrreqs rip6_usrreqs = {
930 .pru_abort = rip6_abort,
931 .pru_attach = rip6_attach,
932 .pru_bind = rip6_bind,
933 .pru_connect = rip6_connect,
934 .pru_control = in6_control,
935 .pru_detach = rip6_detach,
936 .pru_disconnect = rip6_disconnect,
937 .pru_peeraddr = in6_getpeeraddr,
938 .pru_send = rip6_send,
939 .pru_shutdown = rip6_shutdown,
940 .pru_sockaddr = in6_getsockaddr,
941 .pru_close = rip6_close,
942 };
Cache object: 6066f43687398b36573b48f3b1651572
|