raw_ip.c revision 1.54 1 /* $NetBSD: raw_ip.c,v 1.54 2001/01/24 09:04:15 itojun Exp $ */
2
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. All rights reserved.
35 *
36 * Redistribution and use in source and binary forms, with or without
37 * modification, are permitted provided that the following conditions
38 * are met:
39 * 1. Redistributions of source code must retain the above copyright
40 * notice, this list of conditions and the following disclaimer.
41 * 2. Redistributions in binary form must reproduce the above copyright
42 * notice, this list of conditions and the following disclaimer in the
43 * documentation and/or other materials provided with the distribution.
44 * 3. All advertising materials mentioning features or use of this software
45 * must display the following acknowledgement:
46 * This product includes software developed by the University of
47 * California, Berkeley and its contributors.
48 * 4. Neither the name of the University nor the names of its contributors
49 * may be used to endorse or promote products derived from this software
50 * without specific prior written permission.
51 *
52 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
53 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
54 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
55 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
56 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
57 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
58 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
59 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
60 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
61 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
62 * SUCH DAMAGE.
63 *
64 * @(#)raw_ip.c 8.7 (Berkeley) 5/15/95
65 */
66
67 #include "opt_ipsec.h"
68 #include "opt_mrouting.h"
69
70 #include <sys/param.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
80 #include <net/if.h>
81 #include <net/route.h>
82
83 #include <netinet/in.h>
84 #include <netinet/in_systm.h>
85 #include <netinet/ip.h>
86 #include <netinet/ip_var.h>
87 #include <netinet/ip_mroute.h>
88 #include <netinet/ip_icmp.h>
89 #include <netinet/in_pcb.h>
90 #include <netinet/in_var.h>
91
92 #include <machine/stdarg.h>
93
94 #ifdef IPSEC
95 #include <netinet6/ipsec.h>
96 #endif /*IPSEC*/
97
98 struct inpcbtable rawcbtable;
99
100 int rip_bind __P((struct inpcb *, struct mbuf *));
101 int rip_connect __P((struct inpcb *, struct mbuf *));
102 void rip_disconnect __P((struct inpcb *));
103
104 /*
105 * Nominal space allocated to a raw ip socket.
106 */
107 #define RIPSNDQ 8192
108 #define RIPRCVQ 8192
109
110 /*
111 * Raw interface to IP protocol.
112 */
113
114 /*
115 * Initialize raw connection block q.
116 */
117 void
118 rip_init()
119 {
120
121 in_pcbinit(&rawcbtable, 1, 1);
122 }
123
124 static struct sockaddr_in ripsrc = { sizeof(ripsrc), AF_INET };
125
126 /*
127 * Setup generic address and protocol structures
128 * for raw_input routine, then pass them along with
129 * mbuf chain.
130 */
131 void
132 #if __STDC__
133 rip_input(struct mbuf *m, ...)
134 #else
135 rip_input(m, va_alist)
136 struct mbuf *m;
137 va_dcl
138 #endif
139 {
140 int off, proto;
141 struct ip *ip = mtod(m, struct ip *);
142 struct inpcb *inp;
143 struct inpcb *last = 0;
144 struct mbuf *opts = 0;
145 struct sockaddr_in ripsrc;
146 va_list ap;
147
148 va_start(ap, m);
149 off = va_arg(ap, int);
150 proto = va_arg(ap, int);
151 va_end(ap);
152
153 ripsrc.sin_family = AF_INET;
154 ripsrc.sin_len = sizeof(struct sockaddr_in);
155 ripsrc.sin_addr = ip->ip_src;
156 ripsrc.sin_port = 0;
157 bzero((caddr_t)ripsrc.sin_zero, sizeof(ripsrc.sin_zero));
158
159 /*
160 * XXX Compatibility: programs using raw IP expect ip_len
161 * XXX to have the header length subtracted.
162 */
163 ip->ip_len -= ip->ip_hl << 2;
164
165 for (inp = rawcbtable.inpt_queue.cqh_first;
166 inp != (struct inpcb *)&rawcbtable.inpt_queue;
167 inp = inp->inp_queue.cqe_next) {
168 if (inp->inp_ip.ip_p && inp->inp_ip.ip_p != proto)
169 continue;
170 if (!in_nullhost(inp->inp_laddr) &&
171 !in_hosteq(inp->inp_laddr, ip->ip_dst))
172 continue;
173 if (!in_nullhost(inp->inp_faddr) &&
174 !in_hosteq(inp->inp_faddr, ip->ip_src))
175 continue;
176 if (last) {
177 struct mbuf *n;
178 if ((n = m_copy(m, 0, (int)M_COPYALL)) != NULL) {
179 if (last->inp_flags & INP_CONTROLOPTS ||
180 last->inp_socket->so_options & SO_TIMESTAMP)
181 ip_savecontrol(last, &opts, ip, n);
182 if (sbappendaddr(&last->inp_socket->so_rcv,
183 sintosa(&ripsrc), n, opts) == 0) {
184 /* should notify about lost packet */
185 m_freem(n);
186 if (opts)
187 m_freem(opts);
188 } else
189 sorwakeup(last->inp_socket);
190 opts = NULL;
191 }
192 }
193 last = inp;
194 }
195 if (last) {
196 if (last->inp_flags & INP_CONTROLOPTS ||
197 last->inp_socket->so_options & SO_TIMESTAMP)
198 ip_savecontrol(last, &opts, ip, m);
199 if (sbappendaddr(&last->inp_socket->so_rcv,
200 sintosa(&ripsrc), m, opts) == 0) {
201 m_freem(m);
202 if (opts)
203 m_freem(opts);
204 } else
205 sorwakeup(last->inp_socket);
206 } else {
207 if (inetsw[ip_protox[ip->ip_p]].pr_input == rip_input) {
208 icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_PROTOCOL,
209 0, 0);
210 ipstat.ips_noproto++;
211 ipstat.ips_delivered--;
212 } else
213 m_freem(m);
214 }
215 return;
216 }
217
218 /*
219 * Generate IP header and pass packet to ip_output.
220 * Tack on options user may have setup with control call.
221 */
222 int
223 #if __STDC__
224 rip_output(struct mbuf *m, ...)
225 #else
226 rip_output(m, va_alist)
227 struct mbuf *m;
228 va_dcl
229 #endif
230 {
231 struct inpcb *inp;
232 struct ip *ip;
233 struct mbuf *opts;
234 int flags;
235 va_list ap;
236
237 va_start(ap, m);
238 inp = va_arg(ap, struct inpcb *);
239 va_end(ap);
240
241 flags =
242 (inp->inp_socket->so_options & SO_DONTROUTE) | IP_ALLOWBROADCAST
243 | IP_RETURNMTU;
244
245 /*
246 * If the user handed us a complete IP packet, use it.
247 * Otherwise, allocate an mbuf for a header and fill it in.
248 */
249 if ((inp->inp_flags & INP_HDRINCL) == 0) {
250 if ((m->m_pkthdr.len + sizeof(struct ip)) > IP_MAXPACKET) {
251 m_freem(m);
252 return (EMSGSIZE);
253 }
254 M_PREPEND(m, sizeof(struct ip), M_WAIT);
255 ip = mtod(m, struct ip *);
256 ip->ip_tos = 0;
257 ip->ip_off = 0;
258 ip->ip_p = inp->inp_ip.ip_p;
259 ip->ip_len = m->m_pkthdr.len;
260 ip->ip_src = inp->inp_laddr;
261 ip->ip_dst = inp->inp_faddr;
262 ip->ip_ttl = MAXTTL;
263 opts = inp->inp_options;
264 } else {
265 if (m->m_pkthdr.len > IP_MAXPACKET) {
266 m_freem(m);
267 return (EMSGSIZE);
268 }
269 ip = mtod(m, struct ip *);
270 if (m->m_pkthdr.len != ip->ip_len) {
271 m_freem(m);
272 return (EINVAL);
273 }
274 if (ip->ip_id == 0)
275 ip->ip_id = htons(ip_id++);
276 opts = NULL;
277 /* XXX prevent ip_output from overwriting header fields */
278 flags |= IP_RAWOUTPUT;
279 ipstat.ips_rawout++;
280 }
281 #ifdef IPSEC
282 if (ipsec_setsocket(m, inp->inp_socket) != 0) {
283 m_freem(m);
284 return ENOBUFS;
285 }
286 #endif /*IPSEC*/
287 return (ip_output(m, opts, &inp->inp_route, flags, inp->inp_moptions, &inp->inp_errormtu));
288 }
289
290 /*
291 * Raw IP socket option processing.
292 */
293 int
294 rip_ctloutput(op, so, level, optname, m)
295 int op;
296 struct socket *so;
297 int level, optname;
298 struct mbuf **m;
299 {
300 struct inpcb *inp = sotoinpcb(so);
301 int error = 0;
302
303 if (level != IPPROTO_IP) {
304 error = ENOPROTOOPT;
305 if (op == PRCO_SETOPT && *m != 0)
306 (void) m_free(*m);
307 } else switch (op) {
308
309 case PRCO_SETOPT:
310 switch (optname) {
311 case IP_HDRINCL:
312 if (*m == 0 || (*m)->m_len < sizeof (int))
313 error = EINVAL;
314 else {
315 if (*mtod(*m, int *))
316 inp->inp_flags |= INP_HDRINCL;
317 else
318 inp->inp_flags &= ~INP_HDRINCL;
319 }
320 if (*m != 0)
321 (void) m_free(*m);
322 break;
323
324 #ifdef MROUTING
325 case MRT_INIT:
326 case MRT_DONE:
327 case MRT_ADD_VIF:
328 case MRT_DEL_VIF:
329 case MRT_ADD_MFC:
330 case MRT_DEL_MFC:
331 case MRT_ASSERT:
332 error = ip_mrouter_set(so, optname, m);
333 break;
334 #endif
335
336 default:
337 error = ip_ctloutput(op, so, level, optname, m);
338 break;
339 }
340 break;
341
342 case PRCO_GETOPT:
343 switch (optname) {
344 case IP_HDRINCL:
345 *m = m_get(M_WAIT, M_SOOPTS);
346 (*m)->m_len = sizeof (int);
347 *mtod(*m, int *) = inp->inp_flags & INP_HDRINCL ? 1 : 0;
348 break;
349
350 #ifdef MROUTING
351 case MRT_VERSION:
352 case MRT_ASSERT:
353 error = ip_mrouter_get(so, optname, m);
354 break;
355 #endif
356
357 default:
358 error = ip_ctloutput(op, so, level, optname, m);
359 break;
360 }
361 break;
362 }
363 return (error);
364 }
365
366 int
367 rip_bind(inp, nam)
368 struct inpcb *inp;
369 struct mbuf *nam;
370 {
371 struct sockaddr_in *addr = mtod(nam, struct sockaddr_in *);
372
373 if (nam->m_len != sizeof(*addr))
374 return (EINVAL);
375 if (ifnet.tqh_first == 0)
376 return (EADDRNOTAVAIL);
377 if (addr->sin_family != AF_INET &&
378 addr->sin_family != AF_IMPLINK)
379 return (EAFNOSUPPORT);
380 if (!in_nullhost(addr->sin_addr) &&
381 ifa_ifwithaddr(sintosa(addr)) == 0)
382 return (EADDRNOTAVAIL);
383 inp->inp_laddr = addr->sin_addr;
384 return (0);
385 }
386
387 int
388 rip_connect(inp, nam)
389 struct inpcb *inp;
390 struct mbuf *nam;
391 {
392 struct sockaddr_in *addr = mtod(nam, struct sockaddr_in *);
393
394 if (nam->m_len != sizeof(*addr))
395 return (EINVAL);
396 if (ifnet.tqh_first == 0)
397 return (EADDRNOTAVAIL);
398 if (addr->sin_family != AF_INET &&
399 addr->sin_family != AF_IMPLINK)
400 return (EAFNOSUPPORT);
401 inp->inp_faddr = addr->sin_addr;
402 return (0);
403 }
404
405 void
406 rip_disconnect(inp)
407 struct inpcb *inp;
408 {
409
410 inp->inp_faddr = zeroin_addr;
411 }
412
413 u_long rip_sendspace = RIPSNDQ;
414 u_long rip_recvspace = RIPRCVQ;
415
416 /*ARGSUSED*/
417 int
418 rip_usrreq(so, req, m, nam, control, p)
419 struct socket *so;
420 int req;
421 struct mbuf *m, *nam, *control;
422 struct proc *p;
423 {
424 struct inpcb *inp;
425 int s;
426 int error = 0;
427 #ifdef MROUTING
428 extern struct socket *ip_mrouter;
429 #endif
430
431 if (req == PRU_CONTROL)
432 return (in_control(so, (long)m, (caddr_t)nam,
433 (struct ifnet *)control, p));
434
435 if (req == PRU_PURGEIF) {
436 in_purgeif((struct ifnet *)control);
437 in_pcbpurgeif(&rawcbtable, (struct ifnet *)control);
438 return (0);
439 }
440
441 s = splsoftnet();
442 inp = sotoinpcb(so);
443 #ifdef DIAGNOSTIC
444 if (req != PRU_SEND && req != PRU_SENDOOB && control)
445 panic("rip_usrreq: unexpected control mbuf");
446 #endif
447 if (inp == 0 && req != PRU_ATTACH) {
448 error = EINVAL;
449 goto release;
450 }
451
452 switch (req) {
453
454 case PRU_ATTACH:
455 if (inp != 0) {
456 error = EISCONN;
457 break;
458 }
459 if (p == 0 || (error = suser(p->p_ucred, &p->p_acflag))) {
460 error = EACCES;
461 break;
462 }
463 if (so->so_snd.sb_hiwat == 0 || so->so_rcv.sb_hiwat == 0) {
464 error = soreserve(so, rip_sendspace, rip_recvspace);
465 if (error)
466 break;
467 }
468 error = in_pcballoc(so, &rawcbtable);
469 if (error)
470 break;
471 inp = sotoinpcb(so);
472 inp->inp_ip.ip_p = (long)nam;
473 #ifdef IPSEC
474 error = ipsec_init_policy(so, &inp->inp_sp);
475 if (error != 0) {
476 in_pcbdetach(inp);
477 break;
478 }
479 #endif /*IPSEC*/
480 break;
481
482 case PRU_DETACH:
483 #ifdef MROUTING
484 if (so == ip_mrouter)
485 ip_mrouter_done();
486 #endif
487 in_pcbdetach(inp);
488 break;
489
490 case PRU_BIND:
491 error = rip_bind(inp, nam);
492 break;
493
494 case PRU_LISTEN:
495 error = EOPNOTSUPP;
496 break;
497
498 case PRU_CONNECT:
499 error = rip_connect(inp, nam);
500 if (error)
501 break;
502 soisconnected(so);
503 break;
504
505 case PRU_CONNECT2:
506 error = EOPNOTSUPP;
507 break;
508
509 case PRU_DISCONNECT:
510 soisdisconnected(so);
511 rip_disconnect(inp);
512 break;
513
514 /*
515 * Mark the connection as being incapable of further input.
516 */
517 case PRU_SHUTDOWN:
518 socantsendmore(so);
519 break;
520
521 case PRU_RCVD:
522 error = EOPNOTSUPP;
523 break;
524
525 /*
526 * Ship a packet out. The appropriate raw output
527 * routine handles any massaging necessary.
528 */
529 case PRU_SEND:
530 if (control && control->m_len) {
531 m_freem(control);
532 m_freem(m);
533 error = EINVAL;
534 break;
535 }
536 {
537 if (nam) {
538 if ((so->so_state & SS_ISCONNECTED) != 0) {
539 error = EISCONN;
540 goto die;
541 }
542 error = rip_connect(inp, nam);
543 if (error) {
544 die:
545 m_freem(m);
546 break;
547 }
548 } else {
549 if ((so->so_state & SS_ISCONNECTED) == 0) {
550 error = ENOTCONN;
551 goto die;
552 }
553 }
554 error = rip_output(m, inp);
555 if (nam)
556 rip_disconnect(inp);
557 }
558 break;
559
560 case PRU_SENSE:
561 /*
562 * stat: don't bother with a blocksize.
563 */
564 splx(s);
565 return (0);
566
567 case PRU_RCVOOB:
568 error = EOPNOTSUPP;
569 break;
570
571 case PRU_SENDOOB:
572 m_freem(control);
573 m_freem(m);
574 error = EOPNOTSUPP;
575 break;
576
577 case PRU_SOCKADDR:
578 in_setsockaddr(inp, nam);
579 break;
580
581 case PRU_PEERADDR:
582 in_setpeeraddr(inp, nam);
583 break;
584
585 default:
586 panic("rip_usrreq");
587 }
588
589 release:
590 splx(s);
591 return (error);
592 }
593