2 * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
8 * 1. Redistributions of source code must retain the above copyright
9 * notice, this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright
11 * notice, this list of conditions and the following disclaimer in the
12 * documentation and/or other materials provided with the distribution.
13 * 3. Neither the name of the project nor the names of its contributors
14 * may be used to endorse or promote products derived from this software
15 * without specific prior written permission.
17 * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
31 * Copyright (c) 1982, 1986, 1988, 1993
32 * The Regents of the University of California.
33 * All rights reserved.
35 * Redistribution and use in source and binary forms, with or without
36 * modification, are permitted provided that the following conditions
38 * 1. Redistributions of source code must retain the above copyright
39 * notice, this list of conditions and the following disclaimer.
40 * 2. Redistributions in binary form must reproduce the above copyright
41 * notice, this list of conditions and the following disclaimer in the
42 * documentation and/or other materials provided with the distribution.
43 * 4. Neither the name of the University nor the names of its contributors
44 * may be used to endorse or promote products derived from this software
45 * without specific prior written permission.
47 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
48 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
49 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
50 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
51 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
52 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
53 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
54 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
55 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
56 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
59 * @(#)raw_ip.c 8.2 (Berkeley) 1/4/94
62 #include <sys/cdefs.h>
63 __FBSDID("$FreeBSD$");
65 #include "opt_ipsec.h"
66 #include "opt_inet6.h"
68 #include <sys/param.h>
69 #include <sys/errno.h>
72 #include <sys/malloc.h>
76 #include <sys/protosw.h>
77 #include <sys/signalvar.h>
78 #include <sys/socket.h>
79 #include <sys/socketvar.h>
81 #include <sys/syslog.h>
84 #include <net/if_types.h>
85 #include <net/route.h>
87 #include <netinet/in.h>
88 #include <netinet/in_var.h>
89 #include <netinet/in_systm.h>
90 #include <netinet/icmp6.h>
91 #include <netinet/in_pcb.h>
92 #include <netinet/ip6.h>
93 #include <netinet6/ip6protosw.h>
94 #include <netinet6/ip6_mroute.h>
95 #include <netinet6/in6_pcb.h>
96 #include <netinet6/ip6_var.h>
97 #include <netinet6/nd6.h>
98 #include <netinet6/raw_ip6.h>
99 #include <netinet6/scope6_var.h>
102 #include <netipsec/ipsec.h>
103 #include <netipsec/ipsec6.h>
106 #include <machine/stdarg.h>
108 #define satosin6(sa) ((struct sockaddr_in6 *)(sa))
109 #define ifatoia6(ifa) ((struct in6_ifaddr *)(ifa))
112 * Raw interface to IP6 protocol.
115 extern struct inpcbhead ripcb;
116 extern struct inpcbinfo ripcbinfo;
117 extern u_long rip_sendspace;
118 extern u_long rip_recvspace;
120 struct rip6stat rip6stat;
123 * Hooks for multicast forwarding.
125 struct socket *ip6_mrouter = NULL;
126 int (*ip6_mrouter_set)(struct socket *, struct sockopt *);
127 int (*ip6_mrouter_get)(struct socket *, struct sockopt *);
128 int (*ip6_mrouter_done)(void);
129 int (*ip6_mforward)(struct ip6_hdr *, struct ifnet *, struct mbuf *);
130 int (*mrt6_ioctl)(int, caddr_t);
133 * Setup generic address and protocol structures for raw_input routine, then
134 * pass them along with mbuf chain.
137 rip6_input(struct mbuf **mp, int *offp, int proto)
139 struct mbuf *m = *mp;
140 register struct ip6_hdr *ip6 = mtod(m, struct ip6_hdr *);
141 register struct inpcb *in6p;
142 struct inpcb *last = 0;
143 struct mbuf *opts = NULL;
144 struct sockaddr_in6 fromsa;
146 rip6stat.rip6s_ipackets++;
148 if (faithprefix_p != NULL && (*faithprefix_p)(&ip6->ip6_dst)) {
149 /* XXX Send icmp6 host/port unreach? */
151 return (IPPROTO_DONE);
154 init_sin6(&fromsa, m); /* general init */
156 INP_INFO_RLOCK(&ripcbinfo);
157 LIST_FOREACH(in6p, &ripcb, inp_list) {
158 /* XXX inp locking */
159 if ((in6p->inp_vflag & INP_IPV6) == 0)
161 if (in6p->inp_ip_p &&
162 in6p->inp_ip_p != proto)
164 if (!IN6_IS_ADDR_UNSPECIFIED(&in6p->in6p_laddr) &&
165 !IN6_ARE_ADDR_EQUAL(&in6p->in6p_laddr, &ip6->ip6_dst))
167 if (!IN6_IS_ADDR_UNSPECIFIED(&in6p->in6p_faddr) &&
168 !IN6_ARE_ADDR_EQUAL(&in6p->in6p_faddr, &ip6->ip6_src))
170 if (prison_check_ip6(in6p->inp_cred, &ip6->ip6_dst) != 0)
173 if (in6p->in6p_cksum != -1) {
174 rip6stat.rip6s_isum++;
175 if (in6_cksum(m, proto, *offp,
176 m->m_pkthdr.len - *offp)) {
178 rip6stat.rip6s_badsum++;
183 struct mbuf *n = m_copy(m, 0, (int)M_COPYALL);
187 * Check AH/ESP integrity.
189 if (n && ipsec6_in_reject(n, last)) {
191 ipsec6stat.in_polvio++;
192 /* Do not inject data into pcb. */
196 if (last->inp_flags & INP_CONTROLOPTS ||
197 last->inp_socket->so_options & SO_TIMESTAMP)
198 ip6_savecontrol(last, n, &opts);
199 /* strip intermediate headers */
201 if (sbappendaddr(&last->inp_socket->so_rcv,
202 (struct sockaddr *)&fromsa,
207 rip6stat.rip6s_fullsock++;
209 sorwakeup(last->inp_socket);
216 INP_INFO_RUNLOCK(&ripcbinfo);
219 * Check AH/ESP integrity.
221 if (last && ipsec6_in_reject(m, last)) {
223 ipsec6stat.in_polvio++;
224 ip6stat.ip6s_delivered--;
225 /* Do not inject data into pcb. */
230 if (last->inp_flags & INP_CONTROLOPTS ||
231 last->inp_socket->so_options & SO_TIMESTAMP)
232 ip6_savecontrol(last, m, &opts);
233 /* Strip intermediate headers. */
235 if (sbappendaddr(&last->inp_socket->so_rcv,
236 (struct sockaddr *)&fromsa, m, opts) == 0) {
240 rip6stat.rip6s_fullsock++;
242 sorwakeup(last->inp_socket);
245 rip6stat.rip6s_nosock++;
246 if (m->m_flags & M_MCAST)
247 rip6stat.rip6s_nosockmcast++;
248 if (proto == IPPROTO_NONE)
251 char *prvnxtp = ip6_get_prevhdr(m, *offp); /* XXX */
252 icmp6_error(m, ICMP6_PARAM_PROB,
253 ICMP6_PARAMPROB_NEXTHEADER,
254 prvnxtp - mtod(m, char *));
256 ip6stat.ip6s_delivered--;
258 return (IPPROTO_DONE);
262 rip6_ctlinput(int cmd, struct sockaddr *sa, void *d)
267 struct ip6ctlparam *ip6cp = NULL;
268 const struct sockaddr_in6 *sa6_src = NULL;
270 struct inpcb *(*notify)(struct inpcb *, int) = in6_rtchange;
272 if (sa->sa_family != AF_INET6 ||
273 sa->sa_len != sizeof(struct sockaddr_in6))
276 if ((unsigned)cmd >= PRC_NCMDS)
278 if (PRC_IS_REDIRECT(cmd))
279 notify = in6_rtchange, d = NULL;
280 else if (cmd == PRC_HOSTDEAD)
282 else if (inet6ctlerrmap[cmd] == 0)
286 * If the parameter is from icmp6, decode it.
289 ip6cp = (struct ip6ctlparam *)d;
291 ip6 = ip6cp->ip6c_ip6;
292 off = ip6cp->ip6c_off;
293 cmdarg = ip6cp->ip6c_cmdarg;
294 sa6_src = ip6cp->ip6c_src;
302 (void) in6_pcbnotify(&ripcbinfo, sa, 0,
303 (const struct sockaddr *)sa6_src, 0, cmd, cmdarg, notify);
307 * Generate IPv6 header and pass packet to ip6_output. Tack on options user
308 * may have setup with control call.
312 rip6_output(struct mbuf *m, ...)
314 rip6_output(m, va_alist)
319 struct mbuf *control;
321 struct sockaddr_in6 *dstsock;
322 struct in6_addr *dst;
325 u_int plen = m->m_pkthdr.len;
327 struct ip6_pktopts opt, *optp;
328 struct ifnet *oifp = NULL;
329 int type = 0, code = 0; /* for ICMPv6 output statistics only */
330 int scope_ambiguous = 0;
331 struct in6_addr *in6a;
335 so = va_arg(ap, struct socket *);
336 dstsock = va_arg(ap, struct sockaddr_in6 *);
337 control = va_arg(ap, struct mbuf *);
340 in6p = sotoinpcb(so);
343 dst = &dstsock->sin6_addr;
345 if ((error = ip6_setpktopts(control, &opt,
346 in6p->in6p_outputopts, so->so_cred,
347 so->so_proto->pr_protocol)) != 0) {
352 optp = in6p->in6p_outputopts;
355 * Check and convert scope zone ID into internal form.
357 * XXX: we may still need to determine the zone later.
359 if (!(so->so_state & SS_ISCONNECTED)) {
360 if (dstsock->sin6_scope_id == 0 && !ip6_use_defzone)
362 if ((error = sa6_embedscope(dstsock, ip6_use_defzone)) != 0)
367 * For an ICMPv6 packet, we should know its type and code to update
370 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) {
371 struct icmp6_hdr *icmp6;
372 if (m->m_len < sizeof(struct icmp6_hdr) &&
373 (m = m_pullup(m, sizeof(struct icmp6_hdr))) == NULL) {
377 icmp6 = mtod(m, struct icmp6_hdr *);
378 type = icmp6->icmp6_type;
379 code = icmp6->icmp6_code;
382 M_PREPEND(m, sizeof(*ip6), M_DONTWAIT);
387 ip6 = mtod(m, struct ip6_hdr *);
390 * Source address selection.
392 if ((in6a = in6_selectsrc(dstsock, optp, in6p, NULL, so->so_cred,
393 &oifp, &error)) == NULL) {
395 error = EADDRNOTAVAIL;
398 error = prison_get_ip6(in6p->inp_cred, in6a);
401 ip6->ip6_src = *in6a;
403 if (oifp && scope_ambiguous) {
405 * Application should provide a proper zone ID or the use of
406 * default zone IDs should be enabled. Unfortunately, some
407 * applications do not behave as it should, so we need a
408 * workaround. Even if an appropriate ID is not determined
409 * (when it's required), if we can determine the outgoing
410 * interface. determine the zone ID based on the interface.
412 error = in6_setscope(&dstsock->sin6_addr, oifp, NULL);
416 ip6->ip6_dst = dstsock->sin6_addr;
419 * Fill in the rest of the IPv6 header fields.
421 ip6->ip6_flow = (ip6->ip6_flow & ~IPV6_FLOWINFO_MASK) |
422 (in6p->inp_flow & IPV6_FLOWINFO_MASK);
423 ip6->ip6_vfc = (ip6->ip6_vfc & ~IPV6_VERSION_MASK) |
424 (IPV6_VERSION & IPV6_VERSION_MASK);
427 * ip6_plen will be filled in ip6_output, so not fill it here.
429 ip6->ip6_nxt = in6p->inp_ip_p;
430 ip6->ip6_hlim = in6_selecthlim(in6p, oifp);
432 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6 ||
433 in6p->in6p_cksum != -1) {
438 /* Compute checksum. */
439 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6)
440 off = offsetof(struct icmp6_hdr, icmp6_cksum);
442 off = in6p->in6p_cksum;
443 if (plen < off + 1) {
447 off += sizeof(struct ip6_hdr);
450 while (n && n->m_len <= off) {
456 p = (u_int16_t *)(mtod(n, caddr_t) + off);
458 *p = in6_cksum(m, ip6->ip6_nxt, sizeof(*ip6), plen);
461 error = ip6_output(m, optp, NULL, 0, in6p->in6p_moptions, &oifp, in6p);
462 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) {
464 icmp6_ifoutstat_inc(oifp, type, code);
465 icmp6stat.icp6s_outhist[type]++;
467 rip6stat.rip6s_opackets++;
477 ip6_clearpktopts(&opt, -1);
485 * Raw IPv6 socket option processing.
488 rip6_ctloutput(struct socket *so, struct sockopt *sopt)
492 if (sopt->sopt_level == IPPROTO_ICMPV6)
494 * XXX: is it better to call icmp6_ctloutput() directly
497 return (icmp6_ctloutput(so, sopt));
498 else if (sopt->sopt_level != IPPROTO_IPV6)
503 switch (sopt->sopt_dir) {
505 switch (sopt->sopt_name) {
513 error = ip6_mrouter_get ? ip6_mrouter_get(so, sopt) :
517 error = ip6_raw_ctloutput(so, sopt);
520 error = ip6_ctloutput(so, sopt);
526 switch (sopt->sopt_name) {
534 error = ip6_mrouter_set ? ip6_mrouter_set(so, sopt) :
538 error = ip6_raw_ctloutput(so, sopt);
541 error = ip6_ctloutput(so, sopt);
551 rip6_attach(struct socket *so, int proto, struct thread *td)
554 struct icmp6_filter *filter;
558 KASSERT(inp == NULL, ("rip6_attach: inp != NULL"));
560 error = priv_check(td, PRIV_NETINET_RAW);
563 error = soreserve(so, rip_sendspace, rip_recvspace);
566 MALLOC(filter, struct icmp6_filter *,
567 sizeof(struct icmp6_filter), M_PCB, M_NOWAIT);
570 INP_INFO_WLOCK(&ripcbinfo);
571 error = in_pcballoc(so, &ripcbinfo);
573 INP_INFO_WUNLOCK(&ripcbinfo);
577 inp = (struct inpcb *)so->so_pcb;
578 INP_INFO_WUNLOCK(&ripcbinfo);
579 inp->inp_vflag |= INP_IPV6;
580 inp->inp_ip_p = (long)proto;
581 inp->in6p_hops = -1; /* use kernel default */
582 inp->in6p_cksum = -1;
583 inp->in6p_icmp6filt = filter;
584 ICMP6_FILTER_SETPASSALL(inp->in6p_icmp6filt);
590 rip6_detach(struct socket *so)
595 KASSERT(inp != NULL, ("rip6_detach: inp == NULL"));
597 if (so == ip6_mrouter && ip6_mrouter_done)
600 INP_INFO_WLOCK(&ripcbinfo);
602 FREE(inp->in6p_icmp6filt, M_PCB);
605 INP_INFO_WUNLOCK(&ripcbinfo);
608 /* XXXRW: This can't ever be called. */
610 rip6_abort(struct socket *so)
615 KASSERT(inp != NULL, ("rip6_abort: inp == NULL"));
617 soisdisconnected(so);
621 rip6_close(struct socket *so)
626 KASSERT(inp != NULL, ("rip6_close: inp == NULL"));
628 soisdisconnected(so);
632 rip6_disconnect(struct socket *so)
637 KASSERT(inp != NULL, ("rip6_disconnect: inp == NULL"));
639 if ((so->so_state & SS_ISCONNECTED) == 0)
641 inp->in6p_faddr = in6addr_any;
647 rip6_bind(struct socket *so, struct sockaddr *nam, struct thread *td)
650 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)nam;
651 struct ifaddr *ia = NULL;
655 KASSERT(inp != NULL, ("rip6_bind: inp == NULL"));
657 if (nam->sa_len != sizeof(*addr))
659 if ((error = prison_check_ip6(td->td_ucred, &addr->sin6_addr)) != 0)
661 if (TAILQ_EMPTY(&ifnet) || addr->sin6_family != AF_INET6)
662 return (EADDRNOTAVAIL);
663 if ((error = sa6_embedscope(addr, ip6_use_defzone)) != 0)
666 if (!IN6_IS_ADDR_UNSPECIFIED(&addr->sin6_addr) &&
667 (ia = ifa_ifwithaddr((struct sockaddr *)addr)) == 0)
668 return (EADDRNOTAVAIL);
670 ((struct in6_ifaddr *)ia)->ia6_flags &
671 (IN6_IFF_ANYCAST|IN6_IFF_NOTREADY|
672 IN6_IFF_DETACHED|IN6_IFF_DEPRECATED)) {
673 return (EADDRNOTAVAIL);
675 INP_INFO_WLOCK(&ripcbinfo);
677 inp->in6p_laddr = addr->sin6_addr;
679 INP_INFO_WUNLOCK(&ripcbinfo);
684 rip6_connect(struct socket *so, struct sockaddr *nam, struct thread *td)
687 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)nam;
688 struct in6_addr *in6a = NULL;
689 struct ifnet *ifp = NULL;
690 int error = 0, scope_ambiguous = 0;
693 KASSERT(inp != NULL, ("rip6_connect: inp == NULL"));
695 if (nam->sa_len != sizeof(*addr))
697 if (TAILQ_EMPTY(&ifnet))
698 return (EADDRNOTAVAIL);
699 if (addr->sin6_family != AF_INET6)
700 return (EAFNOSUPPORT);
703 * Application should provide a proper zone ID or the use of default
704 * zone IDs should be enabled. Unfortunately, some applications do
705 * not behave as it should, so we need a workaround. Even if an
706 * appropriate ID is not determined, we'll see if we can determine
707 * the outgoing interface. If we can, determine the zone ID based on
708 * the interface below.
710 if (addr->sin6_scope_id == 0 && !ip6_use_defzone)
712 if ((error = sa6_embedscope(addr, ip6_use_defzone)) != 0)
715 INP_INFO_WLOCK(&ripcbinfo);
717 /* Source address selection. XXX: need pcblookup? */
718 in6a = in6_selectsrc(addr, inp->in6p_outputopts,
719 inp, NULL, so->so_cred,
723 INP_INFO_WUNLOCK(&ripcbinfo);
724 return (error ? error : EADDRNOTAVAIL);
728 if (ifp && scope_ambiguous &&
729 (error = in6_setscope(&addr->sin6_addr, ifp, NULL)) != 0) {
731 INP_INFO_WUNLOCK(&ripcbinfo);
734 inp->in6p_faddr = addr->sin6_addr;
735 inp->in6p_laddr = *in6a;
738 INP_INFO_WUNLOCK(&ripcbinfo);
743 rip6_shutdown(struct socket *so)
748 KASSERT(inp != NULL, ("rip6_shutdown: inp == NULL"));
757 rip6_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *nam,
758 struct mbuf *control, struct thread *td)
761 struct sockaddr_in6 tmp;
762 struct sockaddr_in6 *dst;
766 KASSERT(inp != NULL, ("rip6_send: inp == NULL"));
768 /* Always copy sockaddr to avoid overwrites. */
770 if (so->so_state & SS_ISCONNECTED) {
776 bzero(&tmp, sizeof(tmp));
777 tmp.sin6_family = AF_INET6;
778 tmp.sin6_len = sizeof(struct sockaddr_in6);
780 bcopy(&inp->in6p_faddr, &tmp.sin6_addr,
781 sizeof(struct in6_addr));
789 if (nam->sa_len != sizeof(struct sockaddr_in6)) {
793 tmp = *(struct sockaddr_in6 *)nam;
796 if (dst->sin6_family == AF_UNSPEC) {
798 * XXX: we allow this case for backward
799 * compatibility to buggy applications that
800 * rely on old (and wrong) kernel behavior.
802 log(LOG_INFO, "rip6 SEND: address family is "
803 "unspec. Assume AF_INET6\n");
804 dst->sin6_family = AF_INET6;
805 } else if (dst->sin6_family != AF_INET6) {
807 return(EAFNOSUPPORT);
810 ret = rip6_output(m, so, dst, control);
814 struct pr_usrreqs rip6_usrreqs = {
815 .pru_abort = rip6_abort,
816 .pru_attach = rip6_attach,
817 .pru_bind = rip6_bind,
818 .pru_connect = rip6_connect,
819 .pru_control = in6_control,
820 .pru_detach = rip6_detach,
821 .pru_disconnect = rip6_disconnect,
822 .pru_peeraddr = in6_getpeeraddr,
823 .pru_send = rip6_send,
824 .pru_shutdown = rip6_shutdown,
825 .pru_sockaddr = in6_getsockaddr,
826 .pru_close = rip6_close,