2 * Copyright (c) 1982, 1986, 1988, 1993
3 * The Regents of the University of California. All rights reserved.
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. All advertising materials mentioning features or use of this software
14 * must display the following acknowledgement:
15 * This product includes software developed by the University of
16 * California, Berkeley and its contributors.
17 * 4. Neither the name of the University nor the names of its contributors
18 * may be used to endorse or promote products derived from this software
19 * without specific prior written permission.
21 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
33 * @(#)raw_ip.c 8.7 (Berkeley) 5/15/95
37 #include "opt_inet6.h"
38 #include "opt_ipsec.h"
39 #include "opt_random_ip_id.h"
41 #include <sys/param.h>
42 #include <sys/systm.h>
43 #include <sys/kernel.h>
44 #include <sys/malloc.h>
46 #include <sys/protosw.h>
47 #include <sys/socket.h>
48 #include <sys/socketvar.h>
49 #include <sys/sysctl.h>
51 #include <vm/vm_zone.h>
54 #include <net/route.h>
57 #include <netinet/in.h>
58 #include <netinet/in_systm.h>
59 #include <netinet/ip.h>
60 #include <netinet/in_pcb.h>
61 #include <netinet/in_var.h>
62 #include <netinet/ip_var.h>
63 #include <netinet/ip_mroute.h>
65 #include <netinet/ip_fw.h>
68 #include <netinet6/ipsec.h>
73 #include <netinet/ip_dummynet.h>
76 struct inpcbhead ripcb;
77 struct inpcbinfo ripcbinfo;
80 * Nominal space allocated to a raw ip socket.
86 * Raw interface to IP protocol.
90 * Initialize raw connection block q.
96 ripcbinfo.listhead = &ripcb;
98 * XXX We don't use the hash list for raw IP, but it's easier
99 * to allocate a one entry hash list than it is to check all
100 * over the place for hashbase == NULL.
102 ripcbinfo.hashbase = hashinit(1, M_PCB, &ripcbinfo.hashmask);
103 ripcbinfo.porthashbase = hashinit(1, M_PCB, &ripcbinfo.porthashmask);
104 ripcbinfo.ipi_zone = zinit("ripcb", sizeof(struct inpcb),
105 maxsockets, ZONE_INTERRUPT, 0);
108 static struct sockaddr_in ripsrc = { sizeof(ripsrc), AF_INET };
110 * Setup generic address and protocol structures
111 * for raw_input routine, then pass them along with
115 rip_input(m, off, proto)
119 register struct ip *ip = mtod(m, struct ip *);
120 register struct inpcb *inp;
121 struct inpcb *last = 0;
122 struct mbuf *opts = 0;
124 ripsrc.sin_addr = ip->ip_src;
125 LIST_FOREACH(inp, &ripcb, inp_list) {
127 if ((inp->inp_vflag & INP_IPV4) == 0)
130 if (inp->inp_ip_p && inp->inp_ip_p != proto)
132 if (inp->inp_laddr.s_addr &&
133 inp->inp_laddr.s_addr != ip->ip_dst.s_addr)
135 if (inp->inp_faddr.s_addr &&
136 inp->inp_faddr.s_addr != ip->ip_src.s_addr)
139 struct mbuf *n = m_copy(m, 0, (int)M_COPYALL);
142 /* check AH/ESP integrity. */
143 if (n && ipsec4_in_reject_so(n, last->inp_socket)) {
145 ipsecstat.in_polvio++;
146 /* do not inject data to pcb */
150 if (last->inp_flags & INP_CONTROLOPTS ||
151 last->inp_socket->so_options & SO_TIMESTAMP)
152 ip_savecontrol(last, &opts, ip, n);
153 if (sbappendaddr(&last->inp_socket->so_rcv,
154 (struct sockaddr *)&ripsrc, n,
156 /* should notify about lost packet */
161 sorwakeup(last->inp_socket);
168 /* check AH/ESP integrity. */
169 if (last && ipsec4_in_reject_so(m, last->inp_socket)) {
171 ipsecstat.in_polvio++;
172 ipstat.ips_delivered--;
173 /* do not inject data to pcb */
177 if (last->inp_flags & INP_CONTROLOPTS ||
178 last->inp_socket->so_options & SO_TIMESTAMP)
179 ip_savecontrol(last, &opts, ip, m);
180 if (sbappendaddr(&last->inp_socket->so_rcv,
181 (struct sockaddr *)&ripsrc, m, opts) == 0) {
186 sorwakeup(last->inp_socket);
189 ipstat.ips_noproto++;
190 ipstat.ips_delivered--;
195 * Generate IP header and pass packet to ip_output.
196 * Tack on options user may have setup with control call.
199 rip_output(m, so, dst)
204 register struct ip *ip;
205 register struct inpcb *inp = sotoinpcb(so);
206 int flags = (so->so_options & SO_DONTROUTE) | IP_ALLOWBROADCAST;
209 * If the user handed us a complete IP packet, use it.
210 * Otherwise, allocate an mbuf for a header and fill it in.
212 if ((inp->inp_flags & INP_HDRINCL) == 0) {
213 if (m->m_pkthdr.len + sizeof(struct ip) > IP_MAXPACKET) {
217 M_PREPEND(m, sizeof(struct ip), M_TRYWAIT);
218 ip = mtod(m, struct ip *);
219 ip->ip_tos = inp->inp_ip_tos;
221 ip->ip_p = inp->inp_ip_p;
222 ip->ip_len = m->m_pkthdr.len;
223 ip->ip_src = inp->inp_laddr;
224 ip->ip_dst.s_addr = dst;
225 ip->ip_ttl = inp->inp_ip_ttl;
227 if (m->m_pkthdr.len > IP_MAXPACKET) {
231 ip = mtod(m, struct ip *);
232 /* don't allow both user specified and setsockopt options,
233 and don't allow packet length sizes that will crash */
234 if (((IP_VHL_HL(ip->ip_vhl) != (sizeof (*ip) >> 2))
236 || (ip->ip_len > m->m_pkthdr.len)
237 || (ip->ip_len < (IP_VHL_HL(ip->ip_vhl) << 2))) {
243 ip->ip_id = ip_randomid();
245 ip->ip_id = htons(ip_id++);
247 /* XXX prevent ip_output from overwriting header fields */
248 flags |= IP_RAWOUTPUT;
253 if (ipsec_setsocket(m, so) != 0) {
259 return (ip_output(m, inp->inp_options, &inp->inp_route, flags,
264 * Raw IP socket option processing.
267 rip_ctloutput(so, sopt)
269 struct sockopt *sopt;
271 struct inpcb *inp = sotoinpcb(so);
274 if (sopt->sopt_level != IPPROTO_IP)
279 switch (sopt->sopt_dir) {
281 switch (sopt->sopt_name) {
283 optval = inp->inp_flags & INP_HDRINCL;
284 error = sooptcopyout(sopt, &optval, sizeof optval);
289 if (ip_fw_ctl_ptr == 0)
292 error = ip_fw_ctl_ptr(sopt);
296 case IP_DUMMYNET_GET:
297 if (ip_dn_ctl_ptr == NULL)
298 error = ENOPROTOOPT ;
300 error = ip_dn_ctl_ptr(sopt);
302 #endif /* DUMMYNET */
312 error = ip_mrouter_get(so, sopt);
316 error = ip_ctloutput(so, sopt);
322 switch (sopt->sopt_name) {
324 error = sooptcopyin(sopt, &optval, sizeof optval,
329 inp->inp_flags |= INP_HDRINCL;
331 inp->inp_flags &= ~INP_HDRINCL;
339 if (ip_fw_ctl_ptr == 0)
342 error = ip_fw_ctl_ptr(sopt);
346 case IP_DUMMYNET_CONFIGURE:
347 case IP_DUMMYNET_DEL:
348 case IP_DUMMYNET_FLUSH:
349 if (ip_dn_ctl_ptr == NULL)
350 error = ENOPROTOOPT ;
352 error = ip_dn_ctl_ptr(sopt);
357 error = ip_rsvp_init(so);
361 error = ip_rsvp_done();
364 /* XXX - should be combined */
366 error = ip_rsvp_vif_init(so, sopt);
369 case IP_RSVP_VIF_OFF:
370 error = ip_rsvp_vif_done(so, sopt);
381 error = ip_mrouter_set(so, sopt);
385 error = ip_ctloutput(so, sopt);
395 * This function exists solely to receive the PRC_IFDOWN messages which
396 * are sent by if_down(). It looks for an ifaddr whose ifa_addr is sa,
397 * and calls in_ifadown() to remove all routes corresponding to that address.
398 * It also receives the PRC_IFUP messages from if_up() and reinstalls the
402 rip_ctlinput(cmd, sa, vip)
407 struct in_ifaddr *ia;
414 TAILQ_FOREACH(ia, &in_ifaddrhead, ia_link) {
415 if (ia->ia_ifa.ifa_addr == sa
416 && (ia->ia_flags & IFA_ROUTE)) {
418 * in_ifscrub kills the interface route.
420 in_ifscrub(ia->ia_ifp, ia);
422 * in_ifadown gets rid of all the rest of
423 * the routes. This is not quite the right
424 * thing to do, but at least if we are running
425 * a routing process they will come back.
427 in_ifadown(&ia->ia_ifa, 0);
434 TAILQ_FOREACH(ia, &in_ifaddrhead, ia_link) {
435 if (ia->ia_ifa.ifa_addr == sa)
438 if (ia == 0 || (ia->ia_flags & IFA_ROUTE))
441 ifp = ia->ia_ifa.ifa_ifp;
443 if ((ifp->if_flags & IFF_LOOPBACK)
444 || (ifp->if_flags & IFF_POINTOPOINT))
447 err = rtinit(&ia->ia_ifa, RTM_ADD, flags);
449 ia->ia_flags |= IFA_ROUTE;
454 u_long rip_sendspace = RIPSNDQ;
455 u_long rip_recvspace = RIPRCVQ;
457 SYSCTL_INT(_net_inet_raw, OID_AUTO, maxdgram, CTLFLAG_RW,
458 &rip_sendspace, 0, "Maximum outgoing raw IP datagram size");
459 SYSCTL_INT(_net_inet_raw, OID_AUTO, recvspace, CTLFLAG_RW,
460 &rip_recvspace, 0, "Maximum incoming raw IP datagram size");
463 rip_attach(struct socket *so, int proto, struct proc *p)
471 if (p && (error = suser(p)) != 0)
474 error = soreserve(so, rip_sendspace, rip_recvspace);
478 error = in_pcballoc(so, &ripcbinfo, p);
482 inp = (struct inpcb *)so->so_pcb;
483 inp->inp_vflag |= INP_IPV4;
484 inp->inp_ip_p = proto;
485 inp->inp_ip_ttl = ip_defttl;
487 error = ipsec_init_policy(so, &inp->inp_sp);
497 rip_detach(struct socket *so)
504 if (so == ip_mrouter)
506 ip_rsvp_force_done(so);
514 rip_abort(struct socket *so)
516 soisdisconnected(so);
517 return rip_detach(so);
521 rip_disconnect(struct socket *so)
523 if ((so->so_state & SS_ISCONNECTED) == 0)
525 return rip_abort(so);
529 rip_bind(struct socket *so, struct sockaddr *nam, struct proc *p)
531 struct inpcb *inp = sotoinpcb(so);
532 struct sockaddr_in *addr = (struct sockaddr_in *)nam;
534 if (nam->sa_len != sizeof(*addr))
537 if (TAILQ_EMPTY(&ifnet) || ((addr->sin_family != AF_INET) &&
538 (addr->sin_family != AF_IMPLINK)) ||
539 (addr->sin_addr.s_addr &&
540 ifa_ifwithaddr((struct sockaddr *)addr) == 0))
541 return EADDRNOTAVAIL;
542 inp->inp_laddr = addr->sin_addr;
547 rip_connect(struct socket *so, struct sockaddr *nam, struct proc *p)
549 struct inpcb *inp = sotoinpcb(so);
550 struct sockaddr_in *addr = (struct sockaddr_in *)nam;
552 if (nam->sa_len != sizeof(*addr))
554 if (TAILQ_EMPTY(&ifnet))
555 return EADDRNOTAVAIL;
556 if ((addr->sin_family != AF_INET) &&
557 (addr->sin_family != AF_IMPLINK))
559 inp->inp_faddr = addr->sin_addr;
565 rip_shutdown(struct socket *so)
572 rip_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *nam,
573 struct mbuf *control, struct proc *p)
575 struct inpcb *inp = sotoinpcb(so);
578 if (so->so_state & SS_ISCONNECTED) {
583 dst = inp->inp_faddr.s_addr;
589 dst = ((struct sockaddr_in *)nam)->sin_addr.s_addr;
591 return rip_output(m, so, dst);
595 rip_pcblist(SYSCTL_HANDLER_ARGS)
598 struct inpcb *inp, **inp_list;
603 * The process of preparing the TCB list is too time-consuming and
604 * resource-intensive to repeat twice on every request.
606 if (req->oldptr == 0) {
607 n = ripcbinfo.ipi_count;
608 req->oldidx = 2 * (sizeof xig)
609 + (n + n/8) * sizeof(struct xinpcb);
613 if (req->newptr != 0)
617 * OK, now we're committed to doing something.
620 gencnt = ripcbinfo.ipi_gencnt;
621 n = ripcbinfo.ipi_count;
624 xig.xig_len = sizeof xig;
626 xig.xig_gen = gencnt;
627 xig.xig_sogen = so_gencnt;
628 error = SYSCTL_OUT(req, &xig, sizeof xig);
632 inp_list = malloc(n * sizeof *inp_list, M_TEMP, M_WAITOK);
637 for (inp = LIST_FIRST(ripcbinfo.listhead), i = 0; inp && i < n;
638 inp = LIST_NEXT(inp, inp_list)) {
639 if (inp->inp_gencnt <= gencnt)
646 for (i = 0; i < n; i++) {
648 if (inp->inp_gencnt <= gencnt) {
650 xi.xi_len = sizeof xi;
651 /* XXX should avoid extra copy */
652 bcopy(inp, &xi.xi_inp, sizeof *inp);
654 sotoxsocket(inp->inp_socket, &xi.xi_socket);
655 error = SYSCTL_OUT(req, &xi, sizeof xi);
660 * Give the user an updated idea of our state.
661 * If the generation differs from what we told
662 * her before, she knows that something happened
663 * while we were processing this request, and it
664 * might be necessary to retry.
667 xig.xig_gen = ripcbinfo.ipi_gencnt;
668 xig.xig_sogen = so_gencnt;
669 xig.xig_count = ripcbinfo.ipi_count;
671 error = SYSCTL_OUT(req, &xig, sizeof xig);
673 free(inp_list, M_TEMP);
677 SYSCTL_PROC(_net_inet_raw, OID_AUTO/*XXX*/, pcblist, CTLFLAG_RD, 0, 0,
678 rip_pcblist, "S,xinpcb", "List of active raw IP sockets");
680 struct pr_usrreqs rip_usrreqs = {
681 rip_abort, pru_accept_notsupp, rip_attach, rip_bind, rip_connect,
682 pru_connect2_notsupp, in_control, rip_detach, rip_disconnect,
683 pru_listen_notsupp, in_setpeeraddr, pru_rcvd_notsupp,
684 pru_rcvoob_notsupp, rip_send, pru_sense_null, rip_shutdown,
685 in_setsockaddr, sosend, soreceive, sopoll