raw_ip.c revision 34881
1/* 2 * Copyright (c) 1982, 1986, 1988, 1993 3 * The Regents of the University of California. All rights reserved. 4 * 5 * Redistribution and use in source and binary forms, with or without 6 * modification, are permitted provided that the following conditions 7 * are met: 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. 20 * 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 31 * SUCH DAMAGE. 32 * 33 * @(#)raw_ip.c 8.7 (Berkeley) 5/15/95 34 * $Id: raw_ip.c,v 1.51 1998/01/27 09:15:07 davidg Exp $ 35 */ 36 37#include <sys/param.h> 38#include <sys/systm.h> 39#include <sys/kernel.h> 40#include <sys/malloc.h> 41#include <sys/mbuf.h> 42#include <sys/proc.h> 43#include <sys/protosw.h> 44#include <sys/socket.h> 45#include <sys/socketvar.h> 46#include <sys/sysctl.h> 47#include <vm/vm_zone.h> 48 49#include <net/if.h> 50#include <net/route.h> 51 52#define _IP_VHL 53#include <netinet/in.h> 54#include <netinet/in_systm.h> 55#include <netinet/ip.h> 56#include <netinet/in_pcb.h> 57#include <netinet/in_var.h> 58#include <netinet/ip_var.h> 59#include <netinet/ip_mroute.h> 60 61#include <netinet/ip_fw.h> 62 63#if !defined(COMPAT_IPFW) || COMPAT_IPFW == 1 64#undef COMPAT_IPFW 65#define COMPAT_IPFW 1 66#else 67#undef COMPAT_IPFW 68#endif 69 70static struct inpcbhead ripcb; 71static struct inpcbinfo ripcbinfo; 72 73/* 74 * Nominal space allocated to a raw ip socket. 75 */ 76#define RIPSNDQ 8192 77#define RIPRCVQ 8192 78 79/* 80 * Raw interface to IP protocol. 81 */ 82 83/* 84 * Initialize raw connection block q. 85 */ 86void 87rip_init() 88{ 89 LIST_INIT(&ripcb); 90 ripcbinfo.listhead = &ripcb; 91 /* 92 * XXX We don't use the hash list for raw IP, but it's easier 93 * to allocate a one entry hash list than it is to check all 94 * over the place for hashbase == NULL. 95 */ 96 ripcbinfo.hashbase = hashinit(1, M_PCB, &ripcbinfo.hashmask); 97 ripcbinfo.porthashbase = hashinit(1, M_PCB, &ripcbinfo.porthashmask); 98 ripcbinfo.ipi_zone = zinit("ripcb", sizeof(struct inpcb), 99 nmbclusters/4, ZONE_INTERRUPT, 0); 100} 101 102static struct sockaddr_in ripsrc = { sizeof(ripsrc), AF_INET }; 103/* 104 * Setup generic address and protocol structures 105 * for raw_input routine, then pass them along with 106 * mbuf chain. 107 */ 108void 109rip_input(m, iphlen) 110 struct mbuf *m; 111 int iphlen; 112{ 113 register struct ip *ip = mtod(m, struct ip *); 114 register struct inpcb *inp; 115 struct inpcb *last = 0; 116 struct mbuf *opts = 0; 117 118 ripsrc.sin_addr = ip->ip_src; 119 for (inp = ripcb.lh_first; inp != NULL; inp = inp->inp_list.le_next) { 120 if (inp->inp_ip_p && inp->inp_ip_p != ip->ip_p) 121 continue; 122 if (inp->inp_laddr.s_addr && 123 inp->inp_laddr.s_addr != ip->ip_dst.s_addr) 124 continue; 125 if (inp->inp_faddr.s_addr && 126 inp->inp_faddr.s_addr != ip->ip_src.s_addr) 127 continue; 128 if (last) { 129 struct mbuf *n = m_copy(m, 0, (int)M_COPYALL); 130 if (n) { 131 if (last->inp_flags & INP_CONTROLOPTS || 132 last->inp_socket->so_options & SO_TIMESTAMP) 133 ip_savecontrol(last, &opts, ip, n); 134 if (sbappendaddr(&last->inp_socket->so_rcv, 135 (struct sockaddr *)&ripsrc, n, 136 opts) == 0) { 137 /* should notify about lost packet */ 138 m_freem(n); 139 if (opts) 140 m_freem(opts); 141 } else 142 sorwakeup(last->inp_socket); 143 opts = 0; 144 } 145 } 146 last = inp; 147 } 148 if (last) { 149 if (last->inp_flags & INP_CONTROLOPTS || 150 last->inp_socket->so_options & SO_TIMESTAMP) 151 ip_savecontrol(last, &opts, ip, m); 152 if (sbappendaddr(&last->inp_socket->so_rcv, 153 (struct sockaddr *)&ripsrc, m, opts) == 0) { 154 m_freem(m); 155 if (opts) 156 m_freem(opts); 157 } else 158 sorwakeup(last->inp_socket); 159 } else { 160 m_freem(m); 161 ipstat.ips_noproto++; 162 ipstat.ips_delivered--; 163 } 164} 165 166/* 167 * Generate IP header and pass packet to ip_output. 168 * Tack on options user may have setup with control call. 169 */ 170int 171rip_output(m, so, dst) 172 register struct mbuf *m; 173 struct socket *so; 174 u_long dst; 175{ 176 register struct ip *ip; 177 register struct inpcb *inp = sotoinpcb(so); 178 int flags = (so->so_options & SO_DONTROUTE) | IP_ALLOWBROADCAST; 179 180 /* 181 * If the user handed us a complete IP packet, use it. 182 * Otherwise, allocate an mbuf for a header and fill it in. 183 */ 184 if ((inp->inp_flags & INP_HDRINCL) == 0) { 185 if (m->m_pkthdr.len + sizeof(struct ip) > IP_MAXPACKET) { 186 m_freem(m); 187 return(EMSGSIZE); 188 } 189 M_PREPEND(m, sizeof(struct ip), M_WAIT); 190 ip = mtod(m, struct ip *); 191 ip->ip_tos = 0; 192 ip->ip_off = 0; 193 ip->ip_p = inp->inp_ip_p; 194 ip->ip_len = m->m_pkthdr.len; 195 ip->ip_src = inp->inp_laddr; 196 ip->ip_dst.s_addr = dst; 197 ip->ip_ttl = MAXTTL; 198 } else { 199 if (m->m_pkthdr.len > IP_MAXPACKET) { 200 m_freem(m); 201 return(EMSGSIZE); 202 } 203 ip = mtod(m, struct ip *); 204 /* don't allow both user specified and setsockopt options, 205 and don't allow packet length sizes that will crash */ 206 if (((IP_VHL_HL(ip->ip_vhl) != (sizeof (*ip) >> 2)) 207 && inp->inp_options) 208 || (ip->ip_len > m->m_pkthdr.len) 209 || (ip->ip_len < (IP_VHL_HL(ip->ip_vhl) << 2))) { 210 m_freem(m); 211 return EINVAL; 212 } 213 if (ip->ip_id == 0) 214 ip->ip_id = htons(ip_id++); 215 /* XXX prevent ip_output from overwriting header fields */ 216 flags |= IP_RAWOUTPUT; 217 ipstat.ips_rawout++; 218 } 219 return (ip_output(m, inp->inp_options, &inp->inp_route, flags, 220 inp->inp_moptions)); 221} 222 223/* 224 * Raw IP socket option processing. 225 */ 226int 227rip_ctloutput(op, so, level, optname, m, p) 228 int op; 229 struct socket *so; 230 int level, optname; 231 struct mbuf **m; 232 struct proc *p; 233{ 234 register struct inpcb *inp = sotoinpcb(so); 235 register int error; 236 237 if (level != IPPROTO_IP) { 238 if (op == PRCO_SETOPT && *m) 239 (void)m_free(*m); 240 return (EINVAL); 241 } 242 243 switch (optname) { 244 245 case IP_HDRINCL: 246 error = 0; 247 if (op == PRCO_SETOPT) { 248 if (m == 0 || *m == 0 || (*m)->m_len < sizeof (int)) 249 error = EINVAL; 250 else if (*mtod(*m, int *)) 251 inp->inp_flags |= INP_HDRINCL; 252 else 253 inp->inp_flags &= ~INP_HDRINCL; 254 if (*m) 255 (void)m_free(*m); 256 } else { 257 *m = m_get(M_WAIT, MT_SOOPTS); 258 (*m)->m_len = sizeof (int); 259 *mtod(*m, int *) = inp->inp_flags & INP_HDRINCL; 260 } 261 return (error); 262 263#ifdef COMPAT_IPFW 264 case IP_FW_GET: 265 if (ip_fw_ctl_ptr == NULL || op == PRCO_SETOPT) { 266 if (*m) (void)m_free(*m); 267 return(EINVAL); 268 } 269 return (*ip_fw_ctl_ptr)(optname, m); 270 271 case IP_FW_ADD: 272 case IP_FW_DEL: 273 case IP_FW_FLUSH: 274 case IP_FW_ZERO: 275 if (ip_fw_ctl_ptr == NULL || op != PRCO_SETOPT) { 276 if (*m) (void)m_free(*m); 277 return(EINVAL); 278 } 279 return (*ip_fw_ctl_ptr)(optname, m); 280 281 case IP_NAT: 282 if (ip_nat_ctl_ptr == NULL) { 283 if (*m) (void)m_free(*m); 284 return(EINVAL); 285 } 286 return (*ip_nat_ctl_ptr)(op, m); 287 288#endif 289 case IP_RSVP_ON: 290 return ip_rsvp_init(so); 291 break; 292 293 case IP_RSVP_OFF: 294 return ip_rsvp_done(); 295 break; 296 297 case IP_RSVP_VIF_ON: 298 return ip_rsvp_vif_init(so, *m); 299 300 case IP_RSVP_VIF_OFF: 301 return ip_rsvp_vif_done(so, *m); 302 303 case MRT_INIT: 304 case MRT_DONE: 305 case MRT_ADD_VIF: 306 case MRT_DEL_VIF: 307 case MRT_ADD_MFC: 308 case MRT_DEL_MFC: 309 case MRT_VERSION: 310 case MRT_ASSERT: 311 if (op == PRCO_SETOPT) { 312 error = ip_mrouter_set(optname, so, *m); 313 if (*m) 314 (void)m_free(*m); 315 } else if (op == PRCO_GETOPT) { 316 error = ip_mrouter_get(optname, so, m); 317 } else 318 error = EINVAL; 319 return (error); 320 } 321 return (ip_ctloutput(op, so, level, optname, m, p)); 322} 323 324/* 325 * This function exists solely to receive the PRC_IFDOWN messages which 326 * are sent by if_down(). It looks for an ifaddr whose ifa_addr is sa, 327 * and calls in_ifadown() to remove all routes corresponding to that address. 328 * It also receives the PRC_IFUP messages from if_up() and reinstalls the 329 * interface routes. 330 */ 331void 332rip_ctlinput(cmd, sa, vip) 333 int cmd; 334 struct sockaddr *sa; 335 void *vip; 336{ 337 struct in_ifaddr *ia; 338 struct ifnet *ifp; 339 int err; 340 int flags; 341 342 switch(cmd) { 343 case PRC_IFDOWN: 344 for (ia = in_ifaddrhead.tqh_first; ia; 345 ia = ia->ia_link.tqe_next) { 346 if (ia->ia_ifa.ifa_addr == sa 347 && (ia->ia_flags & IFA_ROUTE)) { 348 /* 349 * in_ifscrub kills the interface route. 350 */ 351 in_ifscrub(ia->ia_ifp, ia); 352 /* 353 * in_ifadown gets rid of all the rest of 354 * the routes. This is not quite the right 355 * thing to do, but at least if we are running 356 * a routing process they will come back. 357 */ 358 in_ifadown(&ia->ia_ifa); 359 break; 360 } 361 } 362 break; 363 364 case PRC_IFUP: 365 for (ia = in_ifaddrhead.tqh_first; ia; 366 ia = ia->ia_link.tqe_next) { 367 if (ia->ia_ifa.ifa_addr == sa) 368 break; 369 } 370 if (ia == 0 || (ia->ia_flags & IFA_ROUTE)) 371 return; 372 flags = RTF_UP; 373 ifp = ia->ia_ifa.ifa_ifp; 374 375 if ((ifp->if_flags & IFF_LOOPBACK) 376 || (ifp->if_flags & IFF_POINTOPOINT)) 377 flags |= RTF_HOST; 378 379 err = rtinit(&ia->ia_ifa, RTM_ADD, flags); 380 if (err == 0) 381 ia->ia_flags |= IFA_ROUTE; 382 break; 383 } 384} 385 386static u_long rip_sendspace = RIPSNDQ; 387static u_long rip_recvspace = RIPRCVQ; 388 389SYSCTL_INT(_net_inet_raw, OID_AUTO, maxdgram, CTLFLAG_RW, &rip_sendspace, 390 0, ""); 391SYSCTL_INT(_net_inet_raw, OID_AUTO, recvspace, CTLFLAG_RW, &rip_recvspace, 392 0, ""); 393 394static int 395rip_attach(struct socket *so, int proto, struct proc *p) 396{ 397 struct inpcb *inp; 398 int error, s; 399 400 inp = sotoinpcb(so); 401 if (inp) 402 panic("rip_attach"); 403 if (p && (error = suser(p->p_ucred, &p->p_acflag)) != 0) 404 return error; 405 406 s = splnet(); 407 error = in_pcballoc(so, &ripcbinfo, p); 408 splx(s); 409 if (error) 410 return error; 411 error = soreserve(so, rip_sendspace, rip_recvspace); 412 if (error) 413 return error; 414 inp = (struct inpcb *)so->so_pcb; 415 inp->inp_ip_p = proto; 416 return 0; 417} 418 419static int 420rip_detach(struct socket *so) 421{ 422 struct inpcb *inp; 423 424 inp = sotoinpcb(so); 425 if (inp == 0) 426 panic("rip_detach"); 427 if (so == ip_mrouter) 428 ip_mrouter_done(); 429 ip_rsvp_force_done(so); 430 if (so == ip_rsvpd) 431 ip_rsvp_done(); 432 in_pcbdetach(inp); 433 return 0; 434} 435 436static int 437rip_abort(struct socket *so) 438{ 439 soisdisconnected(so); 440 return rip_detach(so); 441} 442 443static int 444rip_disconnect(struct socket *so) 445{ 446 if ((so->so_state & SS_ISCONNECTED) == 0) 447 return ENOTCONN; 448 return rip_abort(so); 449} 450 451static int 452rip_bind(struct socket *so, struct sockaddr *nam, struct proc *p) 453{ 454 struct inpcb *inp = sotoinpcb(so); 455 struct sockaddr_in *addr = (struct sockaddr_in *)nam; 456 457 if (nam->sa_len != sizeof(*addr)) 458 return EINVAL; 459 460 if (TAILQ_EMPTY(&ifnet) || ((addr->sin_family != AF_INET) && 461 (addr->sin_family != AF_IMPLINK)) || 462 (addr->sin_addr.s_addr && 463 ifa_ifwithaddr((struct sockaddr *)addr) == 0)) 464 return EADDRNOTAVAIL; 465 inp->inp_laddr = addr->sin_addr; 466 return 0; 467} 468 469static int 470rip_connect(struct socket *so, struct sockaddr *nam, struct proc *p) 471{ 472 struct inpcb *inp = sotoinpcb(so); 473 struct sockaddr_in *addr = (struct sockaddr_in *)nam; 474 475 if (nam->sa_len != sizeof(*addr)) 476 return EINVAL; 477 if (TAILQ_EMPTY(&ifnet)) 478 return EADDRNOTAVAIL; 479 if ((addr->sin_family != AF_INET) && 480 (addr->sin_family != AF_IMPLINK)) 481 return EAFNOSUPPORT; 482 inp->inp_faddr = addr->sin_addr; 483 soisconnected(so); 484 return 0; 485} 486 487static int 488rip_shutdown(struct socket *so) 489{ 490 socantsendmore(so); 491 return 0; 492} 493 494static int 495rip_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *nam, 496 struct mbuf *control, struct proc *p) 497{ 498 struct inpcb *inp = sotoinpcb(so); 499 register u_long dst; 500 501 if (so->so_state & SS_ISCONNECTED) { 502 if (nam) { 503 m_freem(m); 504 return EISCONN; 505 } 506 dst = inp->inp_faddr.s_addr; 507 } else { 508 if (nam == NULL) { 509 m_freem(m); 510 return ENOTCONN; 511 } 512 dst = ((struct sockaddr_in *)nam)->sin_addr.s_addr; 513 } 514 return rip_output(m, so, dst); 515} 516 517struct pr_usrreqs rip_usrreqs = { 518 rip_abort, pru_accept_notsupp, rip_attach, rip_bind, rip_connect, 519 pru_connect2_notsupp, in_control, rip_detach, rip_disconnect, 520 pru_listen_notsupp, in_setpeeraddr, pru_rcvd_notsupp, 521 pru_rcvoob_notsupp, rip_send, pru_sense_null, rip_shutdown, 522 in_setsockaddr, sosend, soreceive, sopoll 523}; 524