| 1 | /* |
| 2 | * Copyright (c) 1982, 1986, 1991 Regents of the University of California. |
| 3 | * 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 | * from: @(#)in_pcb.c 7.14 (Berkeley) 4/20/91 |
| 34 | * $Id: in_pcb.c,v 1.5 1993/12/19 00:52:37 wollman Exp $ |
| 35 | */ |
| 36 | |
| 37 | #include "param.h" |
| 38 | #include "systm.h" |
| 39 | #include "malloc.h" |
| 40 | #include "mbuf.h" |
| 41 | #include "protosw.h" |
| 42 | #include "socket.h" |
| 43 | #include "socketvar.h" |
| 44 | #include "ioctl.h" |
| 45 | |
| 46 | #include "../net/if.h" |
| 47 | #include "../net/route.h" |
| 48 | |
| 49 | #include "in.h" |
| 50 | #include "in_systm.h" |
| 51 | #include "ip.h" |
| 52 | #include "in_pcb.h" |
| 53 | #include "in_var.h" |
| 54 | #ifdef MULTICAST |
| 55 | #include "ip_var.h" |
| 56 | #endif |
| 57 | |
| 58 | int |
| 59 | in_pcballoc(so, head) |
| 60 | struct socket *so; |
| 61 | struct inpcb *head; |
| 62 | { |
| 63 | struct mbuf *m; |
| 64 | register struct inpcb *inp; |
| 65 | |
| 66 | m = m_getclr(M_DONTWAIT, MT_PCB); |
| 67 | if (m == NULL) |
| 68 | return (ENOBUFS); |
| 69 | inp = mtod(m, struct inpcb *); |
| 70 | inp->inp_head = head; |
| 71 | inp->inp_socket = so; |
| 72 | insque(inp, head); |
| 73 | so->so_pcb = (caddr_t)inp; |
| 74 | return (0); |
| 75 | } |
| 76 | |
| 77 | int |
| 78 | in_pcbbind(inp, nam) |
| 79 | register struct inpcb *inp; |
| 80 | struct mbuf *nam; |
| 81 | { |
| 82 | register struct socket *so = inp->inp_socket; |
| 83 | register struct inpcb *head = inp->inp_head; |
| 84 | register struct sockaddr_in *sin; |
| 85 | u_short lport = 0; |
| 86 | |
| 87 | if (in_ifaddr == 0) |
| 88 | return (EADDRNOTAVAIL); |
| 89 | if (inp->inp_lport || inp->inp_laddr.s_addr != INADDR_ANY) |
| 90 | return (EINVAL); |
| 91 | if (nam == 0) |
| 92 | goto noname; |
| 93 | sin = mtod(nam, struct sockaddr_in *); |
| 94 | if (nam->m_len != sizeof (*sin)) |
| 95 | return (EINVAL); |
| 96 | if (sin->sin_addr.s_addr != INADDR_ANY) { |
| 97 | int tport = sin->sin_port; |
| 98 | |
| 99 | sin->sin_port = 0; /* yech... */ |
| 100 | if (ifa_ifwithaddr((struct sockaddr *)sin) == 0) |
| 101 | return (EADDRNOTAVAIL); |
| 102 | sin->sin_port = tport; |
| 103 | } |
| 104 | lport = sin->sin_port; |
| 105 | if (lport) { |
| 106 | u_short aport = ntohs(lport); |
| 107 | int wild = 0; |
| 108 | |
| 109 | /* GROSS */ |
| 110 | if (aport < IPPORT_RESERVED && (so->so_state & SS_PRIV) == 0) |
| 111 | return (EACCES); |
| 112 | /* even GROSSER, but this is the Internet */ |
| 113 | if ((so->so_options & SO_REUSEADDR) == 0 && |
| 114 | ((so->so_proto->pr_flags & PR_CONNREQUIRED) == 0 || |
| 115 | (so->so_options & SO_ACCEPTCONN) == 0)) |
| 116 | wild = INPLOOKUP_WILDCARD; |
| 117 | if (in_pcblookup(head, |
| 118 | zeroin_addr, 0, sin->sin_addr, lport, wild)) |
| 119 | return (EADDRINUSE); |
| 120 | } |
| 121 | inp->inp_laddr = sin->sin_addr; |
| 122 | noname: |
| 123 | if (lport == 0) |
| 124 | do { |
| 125 | if (head->inp_lport++ < IPPORT_RESERVED || |
| 126 | head->inp_lport > IPPORT_USERRESERVED) |
| 127 | head->inp_lport = IPPORT_RESERVED; |
| 128 | lport = htons(head->inp_lport); |
| 129 | } while (in_pcblookup(head, |
| 130 | zeroin_addr, 0, inp->inp_laddr, lport, 0)); |
| 131 | inp->inp_lport = lport; |
| 132 | return (0); |
| 133 | } |
| 134 | |
| 135 | /* |
| 136 | * Connect from a socket to a specified address. |
| 137 | * Both address and port must be specified in argument sin. |
| 138 | * If don't have a local address for this socket yet, |
| 139 | * then pick one. |
| 140 | */ |
| 141 | int |
| 142 | in_pcbconnect(inp, nam) |
| 143 | register struct inpcb *inp; |
| 144 | struct mbuf *nam; |
| 145 | { |
| 146 | struct in_ifaddr *ia; |
| 147 | struct sockaddr_in *ifaddr = 0; |
| 148 | register struct sockaddr_in *sin = mtod(nam, struct sockaddr_in *); |
| 149 | |
| 150 | if (nam->m_len != sizeof (*sin)) |
| 151 | return (EINVAL); |
| 152 | if (sin->sin_family != AF_INET) |
| 153 | return (EAFNOSUPPORT); |
| 154 | #ifdef MULTICAST |
| 155 | if (sin->sin_port == 0 && !IN_MULTICAST(sin->sin_addr.s_addr)) |
| 156 | return (EADDRNOTAVAIL); |
| 157 | #else |
| 158 | if (sin->sin_port == 0) |
| 159 | return (EADDRNOTAVAIL); |
| 160 | #endif |
| 161 | if (in_ifaddr) { |
| 162 | /* |
| 163 | * If the destination address is INADDR_ANY, |
| 164 | * use the primary local address. |
| 165 | * If the supplied address is INADDR_BROADCAST, |
| 166 | * and the primary interface supports broadcast, |
| 167 | * choose the broadcast address for that interface. |
| 168 | */ |
| 169 | #define satosin(sa) ((struct sockaddr_in *)(sa)) |
| 170 | if (sin->sin_addr.s_addr == INADDR_ANY) |
| 171 | sin->sin_addr = IA_SIN(in_ifaddr)->sin_addr; |
| 172 | else if (sin->sin_addr.s_addr == (u_long)INADDR_BROADCAST && |
| 173 | (in_ifaddr->ia_ifp->if_flags & IFF_BROADCAST)) |
| 174 | sin->sin_addr = satosin(&in_ifaddr->ia_broadaddr)->sin_addr; |
| 175 | } |
| 176 | if (inp->inp_laddr.s_addr == INADDR_ANY) { |
| 177 | register struct route *ro; |
| 178 | struct ifnet *ifp; |
| 179 | |
| 180 | ia = (struct in_ifaddr *)0; |
| 181 | /* |
| 182 | * If route is known or can be allocated now, |
| 183 | * our src addr is taken from the i/f, else punt. |
| 184 | */ |
| 185 | ro = &inp->inp_route; |
| 186 | if (ro->ro_rt && |
| 187 | (satosin(&ro->ro_dst)->sin_addr.s_addr != |
| 188 | sin->sin_addr.s_addr || |
| 189 | inp->inp_socket->so_options & SO_DONTROUTE)) { |
| 190 | RTFREE(ro->ro_rt); |
| 191 | ro->ro_rt = (struct rtentry *)0; |
| 192 | } |
| 193 | if ((inp->inp_socket->so_options & SO_DONTROUTE) == 0 && /*XXX*/ |
| 194 | (ro->ro_rt == (struct rtentry *)0 || |
| 195 | ro->ro_rt->rt_ifp == (struct ifnet *)0)) { |
| 196 | /* No route yet, so try to acquire one */ |
| 197 | ro->ro_dst.sa_family = AF_INET; |
| 198 | ro->ro_dst.sa_len = sizeof(struct sockaddr_in); |
| 199 | ((struct sockaddr_in *) &ro->ro_dst)->sin_addr = |
| 200 | sin->sin_addr; |
| 201 | rtalloc(ro); |
| 202 | } |
| 203 | /* |
| 204 | * If we found a route, use the address |
| 205 | * corresponding to the outgoing interface |
| 206 | * unless it is the loopback (in case a route |
| 207 | * to our address on another net goes to loopback). |
| 208 | */ |
| 209 | if (ro->ro_rt && (ifp = ro->ro_rt->rt_ifp) && |
| 210 | (ifp->if_flags & IFF_LOOPBACK) == 0) |
| 211 | for (ia = in_ifaddr; ia; ia = ia->ia_next) |
| 212 | if (ia->ia_ifp == ifp) |
| 213 | break; |
| 214 | if (ia == 0) { |
| 215 | int fport = sin->sin_port; |
| 216 | |
| 217 | sin->sin_port = 0; |
| 218 | ia = (struct in_ifaddr *) |
| 219 | ifa_ifwithdstaddr((struct sockaddr *)sin); |
| 220 | sin->sin_port = fport; |
| 221 | if (ia == 0) |
| 222 | ia = in_iaonnetof(in_netof(sin->sin_addr)); |
| 223 | if (ia == 0) |
| 224 | ia = in_ifaddr; |
| 225 | if (ia == 0) |
| 226 | return (EADDRNOTAVAIL); |
| 227 | } |
| 228 | #ifdef MULTICAST |
| 229 | /* |
| 230 | * If the destination address is multicast and an outgoing |
| 231 | * interface has been set as a multicast option, use the |
| 232 | * address of that interface as our source address. |
| 233 | */ |
| 234 | if (IN_MULTICAST(ntohl(sin->sin_addr.s_addr)) && |
| 235 | inp->inp_moptions != NULL) { |
| 236 | struct ip_moptions *imo; |
| 237 | struct ifnet *ifp; |
| 238 | |
| 239 | imo = inp->inp_moptions; |
| 240 | if (imo->imo_multicast_ifp != NULL) { |
| 241 | ifp = imo->imo_multicast_ifp; |
| 242 | for (ia = in_ifaddr; ia; ia = ia->ia_next) |
| 243 | if (ia->ia_ifp == ifp) |
| 244 | break; |
| 245 | if (ia == 0) |
| 246 | return (EADDRNOTAVAIL); |
| 247 | } |
| 248 | } |
| 249 | #endif |
| 250 | ifaddr = (struct sockaddr_in *)&ia->ia_addr; |
| 251 | } |
| 252 | if (in_pcblookup(inp->inp_head, |
| 253 | sin->sin_addr, |
| 254 | sin->sin_port, |
| 255 | inp->inp_laddr.s_addr ? inp->inp_laddr : ifaddr->sin_addr, |
| 256 | inp->inp_lport, |
| 257 | 0)) |
| 258 | return (EADDRINUSE); |
| 259 | if (inp->inp_laddr.s_addr == INADDR_ANY) { |
| 260 | if (inp->inp_lport == 0) |
| 261 | (void)in_pcbbind(inp, (struct mbuf *)0); |
| 262 | inp->inp_laddr = ifaddr->sin_addr; |
| 263 | } |
| 264 | inp->inp_faddr = sin->sin_addr; |
| 265 | inp->inp_fport = sin->sin_port; |
| 266 | #ifdef MTUDISC |
| 267 | /* |
| 268 | * If the upper layer asked for PMTU discovery services, see |
| 269 | * if we can get an idea of what the MTU should be... |
| 270 | */ |
| 271 | in_pcbmtu(inp); |
| 272 | #endif /* MTUDISC */ |
| 273 | return (0); |
| 274 | } |
| 275 | |
| 276 | void |
| 277 | in_pcbdisconnect(inp) |
| 278 | struct inpcb *inp; |
| 279 | { |
| 280 | |
| 281 | inp->inp_faddr.s_addr = INADDR_ANY; |
| 282 | inp->inp_fport = 0; |
| 283 | #ifdef MTUDISC |
| 284 | inp->inp_flags &= ~INP_MTUDISCOVERED; |
| 285 | #endif |
| 286 | if (inp->inp_socket->so_state & SS_NOFDREF) |
| 287 | in_pcbdetach(inp); |
| 288 | } |
| 289 | |
| 290 | void |
| 291 | in_pcbdetach(inp) |
| 292 | struct inpcb *inp; |
| 293 | { |
| 294 | struct socket *so = inp->inp_socket; |
| 295 | |
| 296 | so->so_pcb = 0; |
| 297 | sofree(so); |
| 298 | if (inp->inp_options) |
| 299 | (void)m_free(inp->inp_options); |
| 300 | if (inp->inp_route.ro_rt) |
| 301 | rtfree(inp->inp_route.ro_rt); |
| 302 | #ifdef MULTICAST |
| 303 | ip_freemoptions(inp->inp_moptions); |
| 304 | #endif |
| 305 | remque(inp); |
| 306 | (void) m_free(dtom(inp)); |
| 307 | } |
| 308 | |
| 309 | void |
| 310 | in_setsockaddr(inp, nam) |
| 311 | register struct inpcb *inp; |
| 312 | struct mbuf *nam; |
| 313 | { |
| 314 | register struct sockaddr_in *sin; |
| 315 | |
| 316 | nam->m_len = sizeof (*sin); |
| 317 | sin = mtod(nam, struct sockaddr_in *); |
| 318 | bzero((caddr_t)sin, sizeof (*sin)); |
| 319 | sin->sin_family = AF_INET; |
| 320 | sin->sin_len = sizeof(*sin); |
| 321 | sin->sin_port = inp->inp_lport; |
| 322 | sin->sin_addr = inp->inp_laddr; |
| 323 | } |
| 324 | |
| 325 | void |
| 326 | in_setpeeraddr(inp, nam) |
| 327 | struct inpcb *inp; |
| 328 | struct mbuf *nam; |
| 329 | { |
| 330 | register struct sockaddr_in *sin; |
| 331 | |
| 332 | nam->m_len = sizeof (*sin); |
| 333 | sin = mtod(nam, struct sockaddr_in *); |
| 334 | bzero((caddr_t)sin, sizeof (*sin)); |
| 335 | sin->sin_family = AF_INET; |
| 336 | sin->sin_len = sizeof(*sin); |
| 337 | sin->sin_port = inp->inp_fport; |
| 338 | sin->sin_addr = inp->inp_faddr; |
| 339 | } |
| 340 | |
| 341 | /* |
| 342 | * Pass some notification to all connections of a protocol |
| 343 | * associated with address dst. The local address and/or port numbers |
| 344 | * may be specified to limit the search. The "usual action" will be |
| 345 | * taken, depending on the ctlinput cmd. The caller must filter any |
| 346 | * cmds that are uninteresting (e.g., no error in the map). |
| 347 | * Call the protocol specific routine (if any) to report |
| 348 | * any errors for each matching socket. |
| 349 | * |
| 350 | * Must be called at splnet. |
| 351 | */ |
| 352 | void |
| 353 | in_pcbnotify(head, dst, fport, laddr, lport, cmd, notify) |
| 354 | struct inpcb *head; |
| 355 | struct sockaddr *dst; |
| 356 | u_short fport, lport; |
| 357 | struct in_addr laddr; |
| 358 | int cmd; |
| 359 | void (*notify)(struct inpcb *, int); |
| 360 | { |
| 361 | register struct inpcb *inp, *oinp; |
| 362 | struct in_addr faddr; |
| 363 | int errno; |
| 364 | |
| 365 | if ((unsigned)cmd > PRC_NCMDS || dst->sa_family != AF_INET) |
| 366 | return; |
| 367 | faddr = ((struct sockaddr_in *)dst)->sin_addr; |
| 368 | if (faddr.s_addr == INADDR_ANY) |
| 369 | return; |
| 370 | |
| 371 | /* |
| 372 | * Redirects go to all references to the destination, |
| 373 | * and use in_rtchange to invalidate the route cache. |
| 374 | * Dead host indications: notify all references to the destination. |
| 375 | * MTU change indications: same thing. |
| 376 | * Otherwise, if we have knowledge of the local port and address, |
| 377 | * deliver only to that socket. |
| 378 | */ |
| 379 | if (PRC_IS_REDIRECT(cmd) || cmd == PRC_HOSTDEAD |
| 380 | || cmd == PRC_MTUCHANGED) { |
| 381 | fport = 0; |
| 382 | lport = 0; |
| 383 | laddr.s_addr = 0; |
| 384 | if (cmd != PRC_HOSTDEAD && cmd != PRC_MTUCHANGED) |
| 385 | notify = in_rtchange; |
| 386 | } |
| 387 | errno = inetctlerrmap[cmd]; |
| 388 | for (inp = head->inp_next; inp != head;) { |
| 389 | if (inp->inp_faddr.s_addr != faddr.s_addr || |
| 390 | inp->inp_socket == 0 || |
| 391 | (lport && inp->inp_lport != lport) || |
| 392 | (laddr.s_addr && inp->inp_laddr.s_addr != laddr.s_addr) || |
| 393 | (fport && inp->inp_fport != fport)) { |
| 394 | inp = inp->inp_next; |
| 395 | continue; |
| 396 | } |
| 397 | oinp = inp; |
| 398 | inp = inp->inp_next; |
| 399 | if (notify) |
| 400 | (*notify)(oinp, errno); |
| 401 | } |
| 402 | } |
| 403 | |
| 404 | /* |
| 405 | * Check for alternatives when higher level complains |
| 406 | * about service problems. For now, invalidate cached |
| 407 | * routing information. If the route was created dynamically |
| 408 | * (by a redirect), time to try a default gateway again. |
| 409 | */ |
| 410 | void |
| 411 | in_losing(inp) |
| 412 | struct inpcb *inp; |
| 413 | { |
| 414 | register struct rtentry *rt; |
| 415 | |
| 416 | if ((rt = inp->inp_route.ro_rt)) { |
| 417 | rt_missmsg(RTM_LOSING, &inp->inp_route.ro_dst, |
| 418 | rt->rt_gateway, (struct sockaddr *)rt_mask(rt), |
| 419 | (struct sockaddr *)0, rt->rt_flags, 0); |
| 420 | if (rt->rt_flags & RTF_DYNAMIC) |
| 421 | (void) rtrequest(RTM_DELETE, rt_key(rt), |
| 422 | rt->rt_gateway, rt_mask(rt), rt->rt_flags, |
| 423 | (struct rtentry **)0); |
| 424 | inp->inp_route.ro_rt = 0; |
| 425 | rtfree(rt); |
| 426 | |
| 427 | #ifdef MTUDISC |
| 428 | /* |
| 429 | * When doing MTU discovery, we want to find out as |
| 430 | * quickly as possible what the MTU of the new route is. |
| 431 | */ |
| 432 | in_pcbmtu(inp); |
| 433 | #endif /* MTUDISC */ |
| 434 | } |
| 435 | } |
| 436 | |
| 437 | /* |
| 438 | * After a routing change, flush old routing |
| 439 | * and allocate a (hopefully) better one. |
| 440 | */ |
| 441 | void |
| 442 | in_rtchange(inp, errno) |
| 443 | register struct inpcb *inp; |
| 444 | int errno; |
| 445 | { |
| 446 | if (inp->inp_route.ro_rt) { |
| 447 | rtfree(inp->inp_route.ro_rt); |
| 448 | inp->inp_route.ro_rt = 0; |
| 449 | #ifdef MTUDISC |
| 450 | /* |
| 451 | * A new route can be allocated the next time |
| 452 | * output is attempted, but make sure to let |
| 453 | * MTU discovery know about it. |
| 454 | */ |
| 455 | in_pcbmtu(inp); |
| 456 | #endif /* MTUDISC */ |
| 457 | } |
| 458 | } |
| 459 | |
| 460 | struct inpcb * |
| 461 | in_pcblookup(head, faddr, fport, laddr, lport, flags) |
| 462 | struct inpcb *head; |
| 463 | struct in_addr faddr, laddr; |
| 464 | u_short fport, lport; |
| 465 | int flags; |
| 466 | { |
| 467 | register struct inpcb *inp, *match = 0; |
| 468 | int matchwild = 3, wildcard; |
| 469 | |
| 470 | for (inp = head->inp_next; inp != head; inp = inp->inp_next) { |
| 471 | if (inp->inp_lport != lport) |
| 472 | continue; |
| 473 | wildcard = 0; |
| 474 | if (inp->inp_laddr.s_addr != INADDR_ANY) { |
| 475 | if (laddr.s_addr == INADDR_ANY) |
| 476 | wildcard++; |
| 477 | else if (inp->inp_laddr.s_addr != laddr.s_addr) |
| 478 | continue; |
| 479 | } else { |
| 480 | if (laddr.s_addr != INADDR_ANY) |
| 481 | wildcard++; |
| 482 | } |
| 483 | if (inp->inp_faddr.s_addr != INADDR_ANY) { |
| 484 | if (faddr.s_addr == INADDR_ANY) |
| 485 | wildcard++; |
| 486 | else if (inp->inp_faddr.s_addr != faddr.s_addr || |
| 487 | inp->inp_fport != fport) |
| 488 | continue; |
| 489 | } else { |
| 490 | if (faddr.s_addr != INADDR_ANY) |
| 491 | wildcard++; |
| 492 | } |
| 493 | if (wildcard && (flags & INPLOOKUP_WILDCARD) == 0) |
| 494 | continue; |
| 495 | if (wildcard < matchwild) { |
| 496 | match = inp; |
| 497 | matchwild = wildcard; |
| 498 | if (matchwild == 0) |
| 499 | break; |
| 500 | } |
| 501 | } |
| 502 | return (match); |
| 503 | } |