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>
71 #include <sys/malloc.h>
75 #include <sys/protosw.h>
76 #include <sys/signalvar.h>
77 #include <sys/socket.h>
78 #include <sys/socketvar.h>
80 #include <sys/syslog.h>
81 #include <sys/vimage.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 INIT_VNET_INET(curvnet);
140 INIT_VNET_INET6(curvnet);
142 INIT_VNET_IPSEC(curvnet);
144 struct mbuf *m = *mp;
145 register struct ip6_hdr *ip6 = mtod(m, struct ip6_hdr *);
146 register struct inpcb *in6p;
147 struct inpcb *last = 0;
148 struct mbuf *opts = NULL;
149 struct sockaddr_in6 fromsa;
151 V_rip6stat.rip6s_ipackets++;
153 if (faithprefix_p != NULL && (*faithprefix_p)(&ip6->ip6_dst)) {
154 /* XXX Send icmp6 host/port unreach? */
156 return (IPPROTO_DONE);
159 init_sin6(&fromsa, m); /* general init */
161 INP_INFO_RLOCK(&V_ripcbinfo);
162 LIST_FOREACH(in6p, &V_ripcb, inp_list) {
163 if ((in6p->in6p_vflag & INP_IPV6) == 0)
165 if (in6p->in6p_ip6_nxt &&
166 in6p->in6p_ip6_nxt != proto)
168 if (!IN6_IS_ADDR_UNSPECIFIED(&in6p->in6p_laddr) &&
169 !IN6_ARE_ADDR_EQUAL(&in6p->in6p_laddr, &ip6->ip6_dst))
171 if (!IN6_IS_ADDR_UNSPECIFIED(&in6p->in6p_faddr) &&
172 !IN6_ARE_ADDR_EQUAL(&in6p->in6p_faddr, &ip6->ip6_src))
175 if (in6p->in6p_cksum != -1) {
176 V_rip6stat.rip6s_isum++;
177 if (in6_cksum(m, proto, *offp,
178 m->m_pkthdr.len - *offp)) {
180 V_rip6stat.rip6s_badsum++;
185 struct mbuf *n = m_copy(m, 0, (int)M_COPYALL);
189 * Check AH/ESP integrity.
191 if (n && ipsec6_in_reject(n, last)) {
193 V_ipsec6stat.in_polvio++;
194 /* Do not inject data into pcb. */
198 if (last->in6p_flags & IN6P_CONTROLOPTS ||
199 last->in6p_socket->so_options & SO_TIMESTAMP)
200 ip6_savecontrol(last, n, &opts);
201 /* strip intermediate headers */
203 if (sbappendaddr(&last->in6p_socket->so_rcv,
204 (struct sockaddr *)&fromsa,
209 V_rip6stat.rip6s_fullsock++;
211 sorwakeup(last->in6p_socket);
218 INP_INFO_RUNLOCK(&V_ripcbinfo);
221 * Check AH/ESP integrity.
223 if (last && ipsec6_in_reject(m, last)) {
225 V_ipsec6stat.in_polvio++;
226 V_ip6stat.ip6s_delivered--;
227 /* Do not inject data into pcb. */
232 if (last->in6p_flags & IN6P_CONTROLOPTS ||
233 last->in6p_socket->so_options & SO_TIMESTAMP)
234 ip6_savecontrol(last, m, &opts);
235 /* Strip intermediate headers. */
237 if (sbappendaddr(&last->in6p_socket->so_rcv,
238 (struct sockaddr *)&fromsa, m, opts) == 0) {
242 V_rip6stat.rip6s_fullsock++;
244 sorwakeup(last->in6p_socket);
247 V_rip6stat.rip6s_nosock++;
248 if (m->m_flags & M_MCAST)
249 V_rip6stat.rip6s_nosockmcast++;
250 if (proto == IPPROTO_NONE)
253 char *prvnxtp = ip6_get_prevhdr(m, *offp); /* XXX */
254 icmp6_error(m, ICMP6_PARAM_PROB,
255 ICMP6_PARAMPROB_NEXTHEADER,
256 prvnxtp - mtod(m, char *));
258 V_ip6stat.ip6s_delivered--;
260 return (IPPROTO_DONE);
264 rip6_ctlinput(int cmd, struct sockaddr *sa, void *d)
266 INIT_VNET_INET(curvnet);
270 struct ip6ctlparam *ip6cp = NULL;
271 const struct sockaddr_in6 *sa6_src = NULL;
273 struct inpcb *(*notify)(struct inpcb *, int) = in6_rtchange;
275 if (sa->sa_family != AF_INET6 ||
276 sa->sa_len != sizeof(struct sockaddr_in6))
279 if ((unsigned)cmd >= PRC_NCMDS)
281 if (PRC_IS_REDIRECT(cmd))
282 notify = in6_rtchange, d = NULL;
283 else if (cmd == PRC_HOSTDEAD)
285 else if (inet6ctlerrmap[cmd] == 0)
289 * If the parameter is from icmp6, decode it.
292 ip6cp = (struct ip6ctlparam *)d;
294 ip6 = ip6cp->ip6c_ip6;
295 off = ip6cp->ip6c_off;
296 cmdarg = ip6cp->ip6c_cmdarg;
297 sa6_src = ip6cp->ip6c_src;
305 (void) in6_pcbnotify(&V_ripcbinfo, sa, 0,
306 (const struct sockaddr *)sa6_src, 0, cmd, cmdarg, notify);
310 * Generate IPv6 header and pass packet to ip6_output. Tack on options user
311 * may have setup with control call.
315 rip6_output(struct mbuf *m, ...)
317 rip6_output(m, va_alist)
322 INIT_VNET_INET6(curvnet);
323 struct mbuf *control;
325 struct sockaddr_in6 *dstsock;
326 struct in6_addr *dst;
329 u_int plen = m->m_pkthdr.len;
331 struct ip6_pktopts opt, *optp;
332 struct ifnet *oifp = NULL;
333 int type = 0, code = 0; /* for ICMPv6 output statistics only */
334 int scope_ambiguous = 0;
335 struct in6_addr *in6a;
339 so = va_arg(ap, struct socket *);
340 dstsock = va_arg(ap, struct sockaddr_in6 *);
341 control = va_arg(ap, struct mbuf *);
344 in6p = sotoin6pcb(so);
347 dst = &dstsock->sin6_addr;
349 if ((error = ip6_setpktopts(control, &opt,
350 in6p->in6p_outputopts, so->so_cred,
351 so->so_proto->pr_protocol)) != 0) {
356 optp = in6p->in6p_outputopts;
359 * Check and convert scope zone ID into internal form.
361 * XXX: we may still need to determine the zone later.
363 if (!(so->so_state & SS_ISCONNECTED)) {
364 if (dstsock->sin6_scope_id == 0 && !V_ip6_use_defzone)
366 if ((error = sa6_embedscope(dstsock, V_ip6_use_defzone)) != 0)
371 * For an ICMPv6 packet, we should know its type and code to update
374 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) {
375 struct icmp6_hdr *icmp6;
376 if (m->m_len < sizeof(struct icmp6_hdr) &&
377 (m = m_pullup(m, sizeof(struct icmp6_hdr))) == NULL) {
381 icmp6 = mtod(m, struct icmp6_hdr *);
382 type = icmp6->icmp6_type;
383 code = icmp6->icmp6_code;
386 M_PREPEND(m, sizeof(*ip6), M_DONTWAIT);
391 ip6 = mtod(m, struct ip6_hdr *);
394 * Source address selection.
396 if ((in6a = in6_selectsrc(dstsock, optp, in6p, NULL, so->so_cred,
397 &oifp, &error)) == NULL) {
399 error = EADDRNOTAVAIL;
402 ip6->ip6_src = *in6a;
404 if (oifp && scope_ambiguous) {
406 * Application should provide a proper zone ID or the use of
407 * default zone IDs should be enabled. Unfortunately, some
408 * applications do not behave as it should, so we need a
409 * workaround. Even if an appropriate ID is not determined
410 * (when it's required), if we can determine the outgoing
411 * interface. determine the zone ID based on the interface.
413 error = in6_setscope(&dstsock->sin6_addr, oifp, NULL);
417 ip6->ip6_dst = dstsock->sin6_addr;
420 * Fill in the rest of the IPv6 header fields.
422 ip6->ip6_flow = (ip6->ip6_flow & ~IPV6_FLOWINFO_MASK) |
423 (in6p->in6p_flowinfo & IPV6_FLOWINFO_MASK);
424 ip6->ip6_vfc = (ip6->ip6_vfc & ~IPV6_VERSION_MASK) |
425 (IPV6_VERSION & IPV6_VERSION_MASK);
428 * ip6_plen will be filled in ip6_output, so not fill it here.
430 ip6->ip6_nxt = in6p->in6p_ip6_nxt;
431 ip6->ip6_hlim = in6_selecthlim(in6p, oifp);
433 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6 ||
434 in6p->in6p_cksum != -1) {
439 /* Compute checksum. */
440 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6)
441 off = offsetof(struct icmp6_hdr, icmp6_cksum);
443 off = in6p->in6p_cksum;
444 if (plen < off + 1) {
448 off += sizeof(struct ip6_hdr);
451 while (n && n->m_len <= off) {
457 p = (u_int16_t *)(mtod(n, caddr_t) + off);
459 *p = in6_cksum(m, ip6->ip6_nxt, sizeof(*ip6), plen);
462 error = ip6_output(m, optp, NULL, 0, in6p->in6p_moptions, &oifp, in6p);
463 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) {
465 icmp6_ifoutstat_inc(oifp, type, code);
466 V_icmp6stat.icp6s_outhist[type]++;
468 V_rip6stat.rip6s_opackets++;
478 ip6_clearpktopts(&opt, -1);
486 * Raw IPv6 socket option processing.
489 rip6_ctloutput(struct socket *so, struct sockopt *sopt)
493 if (sopt->sopt_level == IPPROTO_ICMPV6)
495 * XXX: is it better to call icmp6_ctloutput() directly
498 return (icmp6_ctloutput(so, sopt));
499 else if (sopt->sopt_level != IPPROTO_IPV6)
504 switch (sopt->sopt_dir) {
506 switch (sopt->sopt_name) {
514 error = ip6_mrouter_get ? ip6_mrouter_get(so, sopt) :
518 error = ip6_raw_ctloutput(so, sopt);
521 error = ip6_ctloutput(so, sopt);
527 switch (sopt->sopt_name) {
535 error = ip6_mrouter_set ? ip6_mrouter_set(so, sopt) :
539 error = ip6_raw_ctloutput(so, sopt);
542 error = ip6_ctloutput(so, sopt);
552 rip6_attach(struct socket *so, int proto, struct thread *td)
554 INIT_VNET_INET(so->so_vnet);
556 struct icmp6_filter *filter;
560 KASSERT(inp == NULL, ("rip6_attach: inp != NULL"));
562 error = priv_check(td, PRIV_NETINET_RAW);
565 error = soreserve(so, rip_sendspace, rip_recvspace);
568 MALLOC(filter, struct icmp6_filter *,
569 sizeof(struct icmp6_filter), M_PCB, M_NOWAIT);
572 INP_INFO_WLOCK(&V_ripcbinfo);
573 error = in_pcballoc(so, &V_ripcbinfo);
575 INP_INFO_WUNLOCK(&V_ripcbinfo);
579 inp = (struct inpcb *)so->so_pcb;
580 INP_INFO_WUNLOCK(&V_ripcbinfo);
581 inp->inp_vflag |= INP_IPV6;
582 inp->in6p_ip6_nxt = (long)proto;
583 inp->in6p_hops = -1; /* use kernel default */
584 inp->in6p_cksum = -1;
585 inp->in6p_icmp6filt = filter;
586 ICMP6_FILTER_SETPASSALL(inp->in6p_icmp6filt);
592 rip6_detach(struct socket *so)
594 INIT_VNET_INET(so->so_vnet);
598 KASSERT(inp != NULL, ("rip6_detach: inp == NULL"));
600 if (so == ip6_mrouter && ip6_mrouter_done)
603 INP_INFO_WLOCK(&V_ripcbinfo);
605 FREE(inp->in6p_icmp6filt, M_PCB);
608 INP_INFO_WUNLOCK(&V_ripcbinfo);
611 /* XXXRW: This can't ever be called. */
613 rip6_abort(struct socket *so)
618 KASSERT(inp != NULL, ("rip6_abort: inp == NULL"));
620 soisdisconnected(so);
624 rip6_close(struct socket *so)
629 KASSERT(inp != NULL, ("rip6_close: inp == NULL"));
631 soisdisconnected(so);
635 rip6_disconnect(struct socket *so)
640 KASSERT(inp != NULL, ("rip6_disconnect: inp == NULL"));
642 if ((so->so_state & SS_ISCONNECTED) == 0)
644 inp->in6p_faddr = in6addr_any;
650 rip6_bind(struct socket *so, struct sockaddr *nam, struct thread *td)
652 INIT_VNET_NET(so->so_vnet);
653 INIT_VNET_INET(so->so_vnet);
654 INIT_VNET_INET6(so->so_vnet);
656 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)nam;
657 struct ifaddr *ia = NULL;
661 KASSERT(inp != NULL, ("rip6_bind: inp == NULL"));
663 if (nam->sa_len != sizeof(*addr))
665 if (TAILQ_EMPTY(&V_ifnet) || addr->sin6_family != AF_INET6)
666 return (EADDRNOTAVAIL);
667 if ((error = sa6_embedscope(addr, V_ip6_use_defzone)) != 0)
670 if (!IN6_IS_ADDR_UNSPECIFIED(&addr->sin6_addr) &&
671 (ia = ifa_ifwithaddr((struct sockaddr *)addr)) == 0)
672 return (EADDRNOTAVAIL);
674 ((struct in6_ifaddr *)ia)->ia6_flags &
675 (IN6_IFF_ANYCAST|IN6_IFF_NOTREADY|
676 IN6_IFF_DETACHED|IN6_IFF_DEPRECATED)) {
677 return (EADDRNOTAVAIL);
679 INP_INFO_WLOCK(&V_ripcbinfo);
681 inp->in6p_laddr = addr->sin6_addr;
683 INP_INFO_WUNLOCK(&V_ripcbinfo);
688 rip6_connect(struct socket *so, struct sockaddr *nam, struct thread *td)
690 INIT_VNET_NET(so->so_vnet);
691 INIT_VNET_INET(so->so_vnet);
692 INIT_VNET_INET6(so->so_vnet);
694 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)nam;
695 struct in6_addr *in6a = NULL;
696 struct ifnet *ifp = NULL;
697 int error = 0, scope_ambiguous = 0;
700 KASSERT(inp != NULL, ("rip6_connect: inp == NULL"));
702 if (nam->sa_len != sizeof(*addr))
704 if (TAILQ_EMPTY(&V_ifnet))
705 return (EADDRNOTAVAIL);
706 if (addr->sin6_family != AF_INET6)
707 return (EAFNOSUPPORT);
710 * Application should provide a proper zone ID or the use of default
711 * zone IDs should be enabled. Unfortunately, some applications do
712 * not behave as it should, so we need a workaround. Even if an
713 * appropriate ID is not determined, we'll see if we can determine
714 * the outgoing interface. If we can, determine the zone ID based on
715 * the interface below.
717 if (addr->sin6_scope_id == 0 && !V_ip6_use_defzone)
719 if ((error = sa6_embedscope(addr, V_ip6_use_defzone)) != 0)
722 INP_INFO_WLOCK(&V_ripcbinfo);
724 /* Source address selection. XXX: need pcblookup? */
725 in6a = in6_selectsrc(addr, inp->in6p_outputopts,
726 inp, NULL, so->so_cred,
730 INP_INFO_WUNLOCK(&V_ripcbinfo);
731 return (error ? error : EADDRNOTAVAIL);
735 if (ifp && scope_ambiguous &&
736 (error = in6_setscope(&addr->sin6_addr, ifp, NULL)) != 0) {
738 INP_INFO_WUNLOCK(&V_ripcbinfo);
741 inp->in6p_faddr = addr->sin6_addr;
742 inp->in6p_laddr = *in6a;
745 INP_INFO_WUNLOCK(&V_ripcbinfo);
750 rip6_shutdown(struct socket *so)
755 KASSERT(inp != NULL, ("rip6_shutdown: inp == NULL"));
764 rip6_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *nam,
765 struct mbuf *control, struct thread *td)
767 INIT_VNET_INET(so->so_vnet);
769 struct sockaddr_in6 tmp;
770 struct sockaddr_in6 *dst;
774 KASSERT(inp != NULL, ("rip6_send: inp == NULL"));
776 /* Always copy sockaddr to avoid overwrites. */
778 if (so->so_state & SS_ISCONNECTED) {
784 bzero(&tmp, sizeof(tmp));
785 tmp.sin6_family = AF_INET6;
786 tmp.sin6_len = sizeof(struct sockaddr_in6);
788 bcopy(&inp->in6p_faddr, &tmp.sin6_addr,
789 sizeof(struct in6_addr));
797 if (nam->sa_len != sizeof(struct sockaddr_in6)) {
801 tmp = *(struct sockaddr_in6 *)nam;
804 if (dst->sin6_family == AF_UNSPEC) {
806 * XXX: we allow this case for backward
807 * compatibility to buggy applications that
808 * rely on old (and wrong) kernel behavior.
810 log(LOG_INFO, "rip6 SEND: address family is "
811 "unspec. Assume AF_INET6\n");
812 dst->sin6_family = AF_INET6;
813 } else if (dst->sin6_family != AF_INET6) {
815 return(EAFNOSUPPORT);
818 ret = rip6_output(m, so, dst, control);
822 struct pr_usrreqs rip6_usrreqs = {
823 .pru_abort = rip6_abort,
824 .pru_attach = rip6_attach,
825 .pru_bind = rip6_bind,
826 .pru_connect = rip6_connect,
827 .pru_control = in6_control,
828 .pru_detach = rip6_detach,
829 .pru_disconnect = rip6_disconnect,
830 .pru_peeraddr = in6_getpeeraddr,
831 .pru_send = rip6_send,
832 .pru_shutdown = rip6_shutdown,
833 .pru_sockaddr = in6_getsockaddr,
834 .pru_close = rip6_close,