bsd/netinet/raw_ip.c

   1 /*
   2  * Copyright (c) 2000 Apple Computer, Inc. All rights reserved.
   3  *
   4  * @APPLE_LICENSE_HEADER_START@
   5  *
   6  * Copyright (c) 1999-2003 Apple Computer, Inc.  All Rights Reserved.
   7  *
   8  * This file contains Original Code and/or Modifications of Original Code
   9  * as defined in and that are subject to the Apple Public Source License
  10  * Version 2.0 (the 'License'). You may not use this file except in
  11  * compliance with the License. Please obtain a copy of the License at
  12  * http://www.opensource.apple.com/apsl/ and read it before using this
  13  * file.
  14  *
  15  * The Original Code and all software distributed under the License are
  16  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  17  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  18  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  19  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  20  * Please see the License for the specific language governing rights and
  21  * limitations under the License.
  22  *
  23  * @APPLE_LICENSE_HEADER_END@
  24  */
  25 /*
  26  * Copyright (c) 1982, 1986, 1988, 1993
  27  *      The Regents of the University of California.  All rights reserved.
  28  *
  29  * Redistribution and use in source and binary forms, with or without
  30  * modification, are permitted provided that the following conditions
  31  * are met:
  32  * 1. Redistributions of source code must retain the above copyright
  33  *    notice, this list of conditions and the following disclaimer.
  34  * 2. Redistributions in binary form must reproduce the above copyright
  35  *    notice, this list of conditions and the following disclaimer in the
  36  *    documentation and/or other materials provided with the distribution.
  37  * 3. All advertising materials mentioning features or use of this software
  38  *    must display the following acknowledgement:
  39  *      This product includes software developed by the University of
  40  *      California, Berkeley and its contributors.
  41  * 4. Neither the name of the University nor the names of its contributors
  42  *    may be used to endorse or promote products derived from this software
  43  *    without specific prior written permission.
  44  *
  45  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  46  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  47  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  48  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  49  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  50  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  51  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  52  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  53  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  54  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  55  * SUCH DAMAGE.
  56  *
  57  *      @(#)raw_ip.c    8.7 (Berkeley) 5/15/95
  58  */
  59
  60 #include <sys/param.h>
  61 #include <sys/systm.h>
  62 #include <sys/kernel.h>
  63 #include <sys/malloc.h>
  64 #include <sys/mbuf.h>
  65 #include <sys/proc.h>
  66 #include <sys/protosw.h>
  67 #include <sys/socket.h>
  68 #include <sys/socketvar.h>
  69 #include <sys/sysctl.h>
  70
  71 #if __FreeBSD__
  72 #include <vm/vm_zone.h>
  73 #endif
  74
  75 #include <net/if.h>
  76 #include <net/route.h>
  77
  78 #define _IP_VHL
  79 #include <netinet/in.h>
  80 #include <netinet/in_systm.h>
  81 #include <netinet/ip.h>
  82 #include <netinet/in_pcb.h>
  83 #include <netinet/in_var.h>
  84 #include <netinet/ip_var.h>
  85 #include <netinet/ip_mroute.h>
  86
  87 #include <netinet/ip_fw.h>
  88
  89 #if IPSEC
  90 #include <netinet6/ipsec.h>
  91 #endif /*IPSEC*/
  92
  93 #if DUMMYNET
  94 #include <netinet/ip_dummynet.h>
  95 #endif
  96
  97 #if IPSEC
  98 extern int ipsec_bypass;
  99 #endif
 100
 101 struct  inpcbhead ripcb;
 102 struct  inpcbinfo ripcbinfo;
 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         LIST_INIT(&ripcb);
 121         ripcbinfo.listhead = &ripcb;
 122         /*
 123          * XXX We don't use the hash list for raw IP, but it's easier
 124          * to allocate a one entry hash list than it is to check all
 125          * over the place for hashbase == NULL.
 126          */
 127         ripcbinfo.hashbase = hashinit(1, M_PCB, &ripcbinfo.hashmask);
 128         ripcbinfo.porthashbase = hashinit(1, M_PCB, &ripcbinfo.porthashmask);
 129
 130         ripcbinfo.ipi_zone = (void *) zinit(sizeof(struct inpcb),
 131                                             (4096 * sizeof(struct inpcb)),
 132                                             4096, "ripzone");
 133
 134 }
 135
 136 static struct   sockaddr_in ripsrc = { sizeof(ripsrc), AF_INET };
 137 /*
 138  * Setup generic address and protocol structures
 139  * for raw_input routine, then pass them along with
 140  * mbuf chain.
 141  */
 142 void
 143 rip_input(m, iphlen)
 144         struct mbuf *m;
 145         int iphlen;
 146 {
 147         register struct ip *ip = mtod(m, struct ip *);
 148         register struct inpcb *inp;
 149         struct inpcb *last = 0;
 150         struct mbuf *opts = 0;
 151
 152         ripsrc.sin_addr = ip->ip_src;
 153         LIST_FOREACH(inp, &ripcb, inp_list) {
 154 #if INET6
 155                 if ((inp->inp_vflag & INP_IPV4) == 0)
 156                         continue;
 157 #endif
 158                 if (inp->inp_ip_p && (inp->inp_ip_p != ip->ip_p))
 159                         continue;
 160                 if (inp->inp_laddr.s_addr &&
 161                   inp->inp_laddr.s_addr != ip->ip_dst.s_addr)
 162                         continue;
 163                 if (inp->inp_faddr.s_addr &&
 164                   inp->inp_faddr.s_addr != ip->ip_src.s_addr)
 165                         continue;
 166                 if (last) {
 167                         struct mbuf *n = m_copy(m, 0, (int)M_COPYALL);
 168
 169 #if IPSEC
 170                         /* check AH/ESP integrity. */
 171                         if (ipsec_bypass == 0 && n && ipsec4_in_reject_so(n, last->inp_socket)) {
 172                                 m_freem(n);
 173                                 ipsecstat.in_polvio++;
 174                                 /* do not inject data to pcb */
 175                         } else
 176 #endif /*IPSEC*/
 177                         if (n) {
 178                                 if (last->inp_flags & INP_CONTROLOPTS ||
 179                                     last->inp_socket->so_options & SO_TIMESTAMP)
 180                                     ip_savecontrol(last, &opts, ip, n);
 181                                 if (last->inp_flags & INP_STRIPHDR) {
 182                                         n->m_len -= iphlen;
 183                                         n->m_pkthdr.len -= iphlen;
 184                                         n->m_data += iphlen;
 185                                 }
 186                                 if (sbappendaddr(&last->inp_socket->so_rcv,
 187                                     (struct sockaddr *)&ripsrc, n,
 188                                     opts) == 0) {
 189                                         /* should notify about lost packet */
 190                                     kprintf("rip_input can't append to socket\n");
 191                                         m_freem(n);
 192                                         if (opts)
 193                                             m_freem(opts);
 194                                 } else
 195                                         sorwakeup(last->inp_socket);
 196                                 opts = 0;
 197                         }
 198                 }
 199                 last = inp;
 200         }
 201 #if IPSEC
 202         /* check AH/ESP integrity. */
 203         if (ipsec_bypass == 0 && last && ipsec4_in_reject_so(m, last->inp_socket)) {
 204                 m_freem(m);
 205                 ipsecstat.in_polvio++;
 206                 ipstat.ips_delivered--;
 207                 /* do not inject data to pcb */
 208         } else
 209 #endif /*IPSEC*/
 210         if (last) {
 211                 if (last->inp_flags & INP_CONTROLOPTS ||
 212                     last->inp_socket->so_options & SO_TIMESTAMP)
 213                         ip_savecontrol(last, &opts, ip, m);
 214         if (last->inp_flags & INP_STRIPHDR) {
 215             m->m_len -= iphlen;
 216             m->m_pkthdr.len -= iphlen;
 217             m->m_data += iphlen;
 218         }
 219                 if (sbappendaddr(&last->inp_socket->so_rcv,
 220                     (struct sockaddr *)&ripsrc, m, opts) == 0) {
 221                     kprintf("rip_input(2) can't append to socket\n");
 222                         m_freem(m);
 223                         if (opts)
 224                             m_freem(opts);
 225                 } else
 226                         sorwakeup(last->inp_socket);
 227         } else {
 228                 m_freem(m);
 229                 ipstat.ips_noproto++;
 230                 ipstat.ips_delivered--;
 231         }
 232 }
 233
 234 /*
 235  * Generate IP header and pass packet to ip_output.
 236  * Tack on options user may have setup with control call.
 237  */
 238 int
 239 rip_output(m, so, dst)
 240         register struct mbuf *m;
 241         struct socket *so;
 242         u_long dst;
 243 {
 244         register struct ip *ip;
 245         register struct inpcb *inp = sotoinpcb(so);
 246         int flags = (so->so_options & SO_DONTROUTE) | IP_ALLOWBROADCAST;
 247
 248         /*
 249          * If the user handed us a complete IP packet, use it.
 250          * Otherwise, allocate an mbuf for a header and fill it in.
 251          */
 252         if ((inp->inp_flags & INP_HDRINCL) == 0) {
 253                 if (m->m_pkthdr.len + sizeof(struct ip) > IP_MAXPACKET) {
 254                         m_freem(m);
 255                         return(EMSGSIZE);
 256                 }
 257                 M_PREPEND(m, sizeof(struct ip), M_WAIT);
 258                 ip = mtod(m, struct ip *);
 259                 ip->ip_tos = inp->inp_ip_tos;
 260                 ip->ip_off = 0;
 261                 ip->ip_p = inp->inp_ip_p;
 262                 ip->ip_len = m->m_pkthdr.len;
 263                 ip->ip_src = inp->inp_laddr;
 264                 ip->ip_dst.s_addr = dst;
 265                 ip->ip_ttl = inp->inp_ip_ttl;
 266         } else {
 267                 if (m->m_pkthdr.len > IP_MAXPACKET) {
 268                         m_freem(m);
 269                         return(EMSGSIZE);
 270                 }
 271                 ip = mtod(m, struct ip *);
 272                 /* don't allow both user specified and setsockopt options,
 273                    and don't allow packet length sizes that will crash */
 274                 if (((IP_VHL_HL(ip->ip_vhl) != (sizeof (*ip) >> 2))
 275                      && inp->inp_options)
 276                     || (ip->ip_len > m->m_pkthdr.len)
 277                     || (ip->ip_len < (IP_VHL_HL(ip->ip_vhl) << 2))) {
 278                         m_freem(m);
 279                         return EINVAL;
 280                 }
 281                 if (ip->ip_id == 0)
 282 #if RANDOM_IP_ID
 283                         ip->ip_id = ip_randomid();
 284 #else
 285                         ip->ip_id = htons(ip_id++);
 286 #endif
 287                 /* XXX prevent ip_output from overwriting header fields */
 288                 flags |= IP_RAWOUTPUT;
 289                 ipstat.ips_rawout++;
 290         }
 291
 292 #if IPSEC
 293         if (ipsec_bypass == 0 && ipsec_setsocket(m, so) != 0) {
 294                 m_freem(m);
 295                 return ENOBUFS;
 296         }
 297 #endif /*IPSEC*/
 298
 299         return (ip_output(m, inp->inp_options, &inp->inp_route, flags,
 300                           inp->inp_moptions));
 301 }
 302
 303 int
 304 load_ipfw()
 305 {
 306         kern_return_t   err;
 307
 308         /* Load the kext by the identifier */
 309         err = kmod_load_extension("com.apple.nke.IPFirewall");
 310         if (err) return err;
 311
 312         if (ip_fw_ctl_ptr == NULL) {
 313                 /* Wait for the kext to finish loading */
 314                 err = tsleep(&ip_fw_ctl_ptr, PWAIT | PCATCH, "load_ipfw_kext", 5 * 60 /* 5 seconds */);
 315         }
 316
 317         return err == 0 && ip_fw_ctl_ptr == NULL ? -1 : err;
 318 }
 319
 320 /*
 321  * Raw IP socket option processing.
 322  */
 323 int
 324 rip_ctloutput(so, sopt)
 325         struct socket *so;
 326         struct sockopt *sopt;
 327 {
 328         struct  inpcb *inp = sotoinpcb(so);
 329         int     error, optval;
 330
 331         if (sopt->sopt_level != IPPROTO_IP)
 332                 return (EINVAL);
 333
 334         error = 0;
 335
 336         switch (sopt->sopt_dir) {
 337         case SOPT_GET:
 338                 switch (sopt->sopt_name) {
 339                 case IP_HDRINCL:
 340                         optval = inp->inp_flags & INP_HDRINCL;
 341                         error = sooptcopyout(sopt, &optval, sizeof optval);
 342                         break;
 343
 344         case IP_STRIPHDR:
 345             optval = inp->inp_flags & INP_STRIPHDR;
 346             error = sooptcopyout(sopt, &optval, sizeof optval);
 347             break;
 348
 349                 case IP_FW_ADD:
 350                 case IP_FW_GET:
 351                 case IP_OLD_FW_ADD:
 352                 case IP_OLD_FW_GET:
 353                         if (ip_fw_ctl_ptr == 0)
 354                                 error = load_ipfw();
 355                         if (ip_fw_ctl_ptr && error == 0)
 356                                 error = ip_fw_ctl_ptr(sopt);
 357                         else
 358                                 error = ENOPROTOOPT;
 359                         break;
 360
 361 #if DUMMYNET
 362                 case IP_DUMMYNET_GET:
 363                         if (ip_dn_ctl_ptr == NULL)
 364                                 error = ENOPROTOOPT ;
 365                         else
 366                                 error = ip_dn_ctl_ptr(sopt);
 367                         break ;
 368 #endif /* DUMMYNET */
 369
 370                 case MRT_INIT:
 371                 case MRT_DONE:
 372                 case MRT_ADD_VIF:
 373                 case MRT_DEL_VIF:
 374                 case MRT_ADD_MFC:
 375                 case MRT_DEL_MFC:
 376                 case MRT_VERSION:
 377                 case MRT_ASSERT:
 378                         error = ip_mrouter_get(so, sopt);
 379                         break;
 380
 381                 default:
 382                         error = ip_ctloutput(so, sopt);
 383                         break;
 384                 }
 385                 break;
 386
 387         case SOPT_SET:
 388                 switch (sopt->sopt_name) {
 389                 case IP_HDRINCL:
 390                         error = sooptcopyin(sopt, &optval, sizeof optval,
 391                                             sizeof optval);
 392                         if (error)
 393                                 break;
 394                         if (optval)
 395                                 inp->inp_flags |= INP_HDRINCL;
 396                         else
 397                                 inp->inp_flags &= ~INP_HDRINCL;
 398                         break;
 399
 400         case IP_STRIPHDR:
 401             error = sooptcopyin(sopt, &optval, sizeof optval,
 402                         sizeof optval);
 403             if (error)
 404                 break;
 405             if (optval)
 406                 inp->inp_flags |= INP_STRIPHDR;
 407             else
 408                 inp->inp_flags &= ~INP_STRIPHDR;
 409             break;
 410
 411
 412                 case IP_FW_ADD:
 413                 case IP_FW_DEL:
 414                 case IP_FW_FLUSH:
 415                 case IP_FW_ZERO:
 416                 case IP_FW_RESETLOG:
 417                 case IP_OLD_FW_ADD:
 418                 case IP_OLD_FW_DEL:
 419                 case IP_OLD_FW_FLUSH:
 420                 case IP_OLD_FW_ZERO:
 421                 case IP_OLD_FW_RESETLOG:
 422                         if (ip_fw_ctl_ptr == 0)
 423                                 error = load_ipfw();
 424                         if (ip_fw_ctl_ptr && error == 0)
 425                                 error = ip_fw_ctl_ptr(sopt);
 426                         else
 427                                 error = ENOPROTOOPT;
 428                         break;
 429
 430 #if DUMMYNET
 431                 case IP_DUMMYNET_CONFIGURE:
 432                 case IP_DUMMYNET_DEL:
 433                 case IP_DUMMYNET_FLUSH:
 434                         if (ip_dn_ctl_ptr == NULL)
 435                                 error = ENOPROTOOPT ;
 436                         else
 437                                 error = ip_dn_ctl_ptr(sopt);
 438                         break ;
 439 #endif
 440
 441                 case IP_RSVP_ON:
 442                         error = ip_rsvp_init(so);
 443                         break;
 444
 445                 case IP_RSVP_OFF:
 446                         error = ip_rsvp_done();
 447                         break;
 448
 449                         /* XXX - should be combined */
 450                 case IP_RSVP_VIF_ON:
 451                         error = ip_rsvp_vif_init(so, sopt);
 452                         break;
 453
 454                 case IP_RSVP_VIF_OFF:
 455                         error = ip_rsvp_vif_done(so, sopt);
 456                         break;
 457
 458                 case MRT_INIT:
 459                 case MRT_DONE:
 460                 case MRT_ADD_VIF:
 461                 case MRT_DEL_VIF:
 462                 case MRT_ADD_MFC:
 463                 case MRT_DEL_MFC:
 464                 case MRT_VERSION:
 465                 case MRT_ASSERT:
 466                         error = ip_mrouter_set(so, sopt);
 467                         break;
 468
 469                 default:
 470                         error = ip_ctloutput(so, sopt);
 471                         break;
 472                 }
 473                 break;
 474         }
 475
 476         return (error);
 477 }
 478
 479 /*
 480  * This function exists solely to receive the PRC_IFDOWN messages which
 481  * are sent by if_down().  It looks for an ifaddr whose ifa_addr is sa,
 482  * and calls in_ifadown() to remove all routes corresponding to that address.
 483  * It also receives the PRC_IFUP messages from if_up() and reinstalls the
 484  * interface routes.
 485  */
 486 void
 487 rip_ctlinput(cmd, sa, vip)
 488         int cmd;
 489         struct sockaddr *sa;
 490         void *vip;
 491 {
 492         struct in_ifaddr *ia;
 493         struct ifnet *ifp;
 494         int err;
 495         int flags;
 496
 497         switch (cmd) {
 498         case PRC_IFDOWN:
 499                 for (ia = in_ifaddrhead.tqh_first; ia;
 500                      ia = ia->ia_link.tqe_next) {
 501                         if (ia->ia_ifa.ifa_addr == sa
 502                             && (ia->ia_flags & IFA_ROUTE)) {
 503                                 /*
 504                                  * in_ifscrub kills the interface route.
 505                                  */
 506                                 in_ifscrub(ia->ia_ifp, ia);
 507                                 /*
 508                                  * in_ifadown gets rid of all the rest of
 509                                  * the routes.  This is not quite the right
 510                                  * thing to do, but at least if we are running
 511                                  * a routing process they will come back.
 512                                  */
 513                                 in_ifadown(&ia->ia_ifa, 1);
 514                                 break;
 515                         }
 516                 }
 517                 break;
 518
 519         case PRC_IFUP:
 520                 for (ia = in_ifaddrhead.tqh_first; ia;
 521                      ia = ia->ia_link.tqe_next) {
 522                         if (ia->ia_ifa.ifa_addr == sa)
 523                                 break;
 524                 }
 525                 if (ia == 0 || (ia->ia_flags & IFA_ROUTE))
 526                         return;
 527                 flags = RTF_UP;
 528                 ifp = ia->ia_ifa.ifa_ifp;
 529
 530                 if ((ifp->if_flags & IFF_LOOPBACK)
 531                     || (ifp->if_flags & IFF_POINTOPOINT))
 532                         flags |= RTF_HOST;
 533
 534                 err = rtinit(&ia->ia_ifa, RTM_ADD, flags);
 535                 if (err == 0)
 536                         ia->ia_flags |= IFA_ROUTE;
 537                 break;
 538         }
 539 }
 540
 541 u_long  rip_sendspace = RIPSNDQ;
 542 u_long  rip_recvspace = RIPRCVQ;
 543
 544 SYSCTL_INT(_net_inet_raw, OID_AUTO, maxdgram, CTLFLAG_RW,
 545     &rip_sendspace, 0, "Maximum outgoing raw IP datagram size");
 546 SYSCTL_INT(_net_inet_raw, OID_AUTO, recvspace, CTLFLAG_RW,
 547     &rip_recvspace, 0, "Maximum incoming raw IP datagram size");
 548
 549 static int
 550 rip_attach(struct socket *so, int proto, struct proc *p)
 551 {
 552         struct inpcb *inp;
 553         int error, s;
 554
 555         inp = sotoinpcb(so);
 556         if (inp)
 557                 panic("rip_attach");
 558 #if __APPLE__
 559         if ((so->so_state & SS_PRIV) == 0)
 560                 return (EPERM);
 561 #else
 562         if (p && (error = suser(p)) != 0)
 563                 return error;
 564 #endif
 565
 566         error = soreserve(so, rip_sendspace, rip_recvspace);
 567         if (error)
 568                 return error;
 569         s = splnet();
 570         error = in_pcballoc(so, &ripcbinfo, p);
 571         splx(s);
 572         if (error)
 573                 return error;
 574         inp = (struct inpcb *)so->so_pcb;
 575         inp->inp_vflag |= INP_IPV4;
 576         inp->inp_ip_p = proto;
 577         inp->inp_ip_ttl = ip_defttl;
 578         return 0;
 579 }
 580
 581 __private_extern__ int
 582 rip_detach(struct socket *so)
 583 {
 584         struct inpcb *inp;
 585
 586         inp = sotoinpcb(so);
 587         if (inp == 0)
 588                 panic("rip_detach");
 589         if (so == ip_mrouter)
 590                 ip_mrouter_done();
 591         ip_rsvp_force_done(so);
 592         if (so == ip_rsvpd)
 593                 ip_rsvp_done();
 594         in_pcbdetach(inp);
 595         return 0;
 596 }
 597
 598 __private_extern__ int
 599 rip_abort(struct socket *so)
 600 {
 601         soisdisconnected(so);
 602         return rip_detach(so);
 603 }
 604
 605 __private_extern__ int
 606 rip_disconnect(struct socket *so)
 607 {
 608         if ((so->so_state & SS_ISCONNECTED) == 0)
 609                 return ENOTCONN;
 610         return rip_abort(so);
 611 }
 612
 613 __private_extern__ int
 614 rip_bind(struct socket *so, struct sockaddr *nam, struct proc *p)
 615 {
 616         struct inpcb *inp = sotoinpcb(so);
 617         struct sockaddr_in *addr = (struct sockaddr_in *)nam;
 618
 619         if (nam->sa_len != sizeof(*addr))
 620                 return EINVAL;
 621
 622         if (TAILQ_EMPTY(&ifnet) || ((addr->sin_family != AF_INET) &&
 623                                     (addr->sin_family != AF_IMPLINK)) ||
 624             (addr->sin_addr.s_addr &&
 625              ifa_ifwithaddr((struct sockaddr *)addr) == 0))
 626                 return EADDRNOTAVAIL;
 627         inp->inp_laddr = addr->sin_addr;
 628         return 0;
 629 }
 630
 631 __private_extern__ int
 632 rip_connect(struct socket *so, struct sockaddr *nam, struct proc *p)
 633 {
 634         struct inpcb *inp = sotoinpcb(so);
 635         struct sockaddr_in *addr = (struct sockaddr_in *)nam;
 636
 637         if (nam->sa_len != sizeof(*addr))
 638                 return EINVAL;
 639         if (TAILQ_EMPTY(&ifnet))
 640                 return EADDRNOTAVAIL;
 641         if ((addr->sin_family != AF_INET) &&
 642             (addr->sin_family != AF_IMPLINK))
 643                 return EAFNOSUPPORT;
 644         inp->inp_faddr = addr->sin_addr;
 645         soisconnected(so);
 646         return 0;
 647 }
 648
 649 __private_extern__ int
 650 rip_shutdown(struct socket *so)
 651 {
 652         socantsendmore(so);
 653         return 0;
 654 }
 655
 656 __private_extern__ int
 657 rip_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *nam,
 658          struct mbuf *control, struct proc *p)
 659 {
 660         struct inpcb *inp = sotoinpcb(so);
 661         register u_long dst;
 662
 663         if (so->so_state & SS_ISCONNECTED) {
 664                 if (nam) {
 665                         m_freem(m);
 666                         return EISCONN;
 667                 }
 668                 dst = inp->inp_faddr.s_addr;
 669         } else {
 670                 if (nam == NULL) {
 671                         m_freem(m);
 672                         return ENOTCONN;
 673                 }
 674                 dst = ((struct sockaddr_in *)nam)->sin_addr.s_addr;
 675         }
 676         return rip_output(m, so, dst);
 677 }
 678
 679 static int
 680 rip_pcblist SYSCTL_HANDLER_ARGS
 681 {
 682         int error, i, n, s;
 683         struct inpcb *inp, **inp_list;
 684         inp_gen_t gencnt;
 685         struct xinpgen xig;
 686
 687         /*
 688          * The process of preparing the TCB list is too time-consuming and
 689          * resource-intensive to repeat twice on every request.
 690          */
 691         if (req->oldptr == 0) {
 692                 n = ripcbinfo.ipi_count;
 693                 req->oldidx = 2 * (sizeof xig)
 694                         + (n + n/8) * sizeof(struct xinpcb);
 695                 return 0;
 696         }
 697
 698         if (req->newptr != 0)
 699                 return EPERM;
 700
 701         /*
 702          * OK, now we're committed to doing something.
 703          */
 704         s = splnet();
 705         gencnt = ripcbinfo.ipi_gencnt;
 706         n = ripcbinfo.ipi_count;
 707         splx(s);
 708
 709         xig.xig_len = sizeof xig;
 710         xig.xig_count = n;
 711         xig.xig_gen = gencnt;
 712         xig.xig_sogen = so_gencnt;
 713         error = SYSCTL_OUT(req, &xig, sizeof xig);
 714         if (error)
 715                 return error;
 716     /*
 717      * We are done if there is no pcb
 718      */
 719     if (n == 0)
 720         return 0;
 721
 722         inp_list = _MALLOC(n * sizeof *inp_list, M_TEMP, M_WAITOK);
 723         if (inp_list == 0)
 724                 return ENOMEM;
 725
 726         s = splnet();
 727         for (inp = ripcbinfo.listhead->lh_first, i = 0; inp && i < n;
 728              inp = inp->inp_list.le_next) {
 729                 if (inp->inp_gencnt <= gencnt)
 730                         inp_list[i++] = inp;
 731         }
 732         splx(s);
 733         n = i;
 734
 735         error = 0;
 736         for (i = 0; i < n; i++) {
 737                 inp = inp_list[i];
 738                 if (inp->inp_gencnt <= gencnt) {
 739                         struct xinpcb xi;
 740                         xi.xi_len = sizeof xi;
 741                         /* XXX should avoid extra copy */
 742                         bcopy(inp, &xi.xi_inp, sizeof *inp);
 743                         if (inp->inp_socket)
 744                                 sotoxsocket(inp->inp_socket, &xi.xi_socket);
 745                         error = SYSCTL_OUT(req, &xi, sizeof xi);
 746                 }
 747         }
 748         if (!error) {
 749                 /*
 750                  * Give the user an updated idea of our state.
 751                  * If the generation differs from what we told
 752                  * her before, she knows that something happened
 753                  * while we were processing this request, and it
 754                  * might be necessary to retry.
 755                  */
 756                 s = splnet();
 757                 xig.xig_gen = ripcbinfo.ipi_gencnt;
 758                 xig.xig_sogen = so_gencnt;
 759                 xig.xig_count = ripcbinfo.ipi_count;
 760                 splx(s);
 761                 error = SYSCTL_OUT(req, &xig, sizeof xig);
 762         }
 763         FREE(inp_list, M_TEMP);
 764         return error;
 765 }
 766
 767 SYSCTL_PROC(_net_inet_raw, OID_AUTO/*XXX*/, pcblist, CTLFLAG_RD, 0, 0,
 768             rip_pcblist, "S,xinpcb", "List of active raw IP sockets");
 769
 770 struct pr_usrreqs rip_usrreqs = {
 771         rip_abort, pru_accept_notsupp, rip_attach, rip_bind, rip_connect,
 772         pru_connect2_notsupp, in_control, rip_detach, rip_disconnect,
 773         pru_listen_notsupp, in_setpeeraddr, pru_rcvd_notsupp,
 774         pru_rcvoob_notsupp, rip_send, pru_sense_null, rip_shutdown,
 775         in_setsockaddr, sosend, soreceive, sopoll
 776 };