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 /*
 304  * Raw IP socket option processing.
 305  */
 306 int
 307 rip_ctloutput(so, sopt)
 308         struct socket *so;
 309         struct sockopt *sopt;
 310 {
 311         struct  inpcb *inp = sotoinpcb(so);
 312         int     error, optval;
 313
 314         if (sopt->sopt_level != IPPROTO_IP)
 315                 return (EINVAL);
 316
 317         error = 0;
 318
 319         switch (sopt->sopt_dir) {
 320         case SOPT_GET:
 321                 switch (sopt->sopt_name) {
 322                 case IP_HDRINCL:
 323                         optval = inp->inp_flags & INP_HDRINCL;
 324                         error = sooptcopyout(sopt, &optval, sizeof optval);
 325                         break;
 326
 327         case IP_STRIPHDR:
 328             optval = inp->inp_flags & INP_STRIPHDR;
 329             error = sooptcopyout(sopt, &optval, sizeof optval);
 330             break;
 331
 332                 case IP_FW_ADD:
 333                 case IP_FW_GET:
 334                 case IP_OLD_FW_ADD:
 335                 case IP_OLD_FW_GET:
 336                         if (ip_fw_ctl_ptr == 0)
 337                                 error = ENOPROTOOPT;
 338                         else
 339                                 error = ip_fw_ctl_ptr(sopt);
 340                         break;
 341
 342 #if DUMMYNET
 343                 case IP_DUMMYNET_GET:
 344                         if (ip_dn_ctl_ptr == NULL)
 345                                 error = ENOPROTOOPT ;
 346                         else
 347                                 error = ip_dn_ctl_ptr(sopt);
 348                         break ;
 349 #endif /* DUMMYNET */
 350
 351                 case MRT_INIT:
 352                 case MRT_DONE:
 353                 case MRT_ADD_VIF:
 354                 case MRT_DEL_VIF:
 355                 case MRT_ADD_MFC:
 356                 case MRT_DEL_MFC:
 357                 case MRT_VERSION:
 358                 case MRT_ASSERT:
 359                         error = ip_mrouter_get(so, sopt);
 360                         break;
 361
 362                 default:
 363                         error = ip_ctloutput(so, sopt);
 364                         break;
 365                 }
 366                 break;
 367
 368         case SOPT_SET:
 369                 switch (sopt->sopt_name) {
 370                 case IP_HDRINCL:
 371                         error = sooptcopyin(sopt, &optval, sizeof optval,
 372                                             sizeof optval);
 373                         if (error)
 374                                 break;
 375                         if (optval)
 376                                 inp->inp_flags |= INP_HDRINCL;
 377                         else
 378                                 inp->inp_flags &= ~INP_HDRINCL;
 379                         break;
 380
 381         case IP_STRIPHDR:
 382             error = sooptcopyin(sopt, &optval, sizeof optval,
 383                         sizeof optval);
 384             if (error)
 385                 break;
 386             if (optval)
 387                 inp->inp_flags |= INP_STRIPHDR;
 388             else
 389                 inp->inp_flags &= ~INP_STRIPHDR;
 390             break;
 391
 392
 393                 case IP_FW_ADD:
 394                 case IP_FW_DEL:
 395                 case IP_FW_FLUSH:
 396                 case IP_FW_ZERO:
 397                 case IP_FW_RESETLOG:
 398                 case IP_OLD_FW_ADD:
 399                 case IP_OLD_FW_DEL:
 400                 case IP_OLD_FW_FLUSH:
 401                 case IP_OLD_FW_ZERO:
 402                 case IP_OLD_FW_RESETLOG:
 403                         if (ip_fw_ctl_ptr == 0)
 404                                 error = ENOPROTOOPT;
 405                         else
 406                                 error = ip_fw_ctl_ptr(sopt);
 407                         break;
 408
 409 #if DUMMYNET
 410                 case IP_DUMMYNET_CONFIGURE:
 411                 case IP_DUMMYNET_DEL:
 412                 case IP_DUMMYNET_FLUSH:
 413                         if (ip_dn_ctl_ptr == NULL)
 414                                 error = ENOPROTOOPT ;
 415                         else
 416                                 error = ip_dn_ctl_ptr(sopt);
 417                         break ;
 418 #endif
 419
 420                 case IP_RSVP_ON:
 421                         error = ip_rsvp_init(so);
 422                         break;
 423
 424                 case IP_RSVP_OFF:
 425                         error = ip_rsvp_done();
 426                         break;
 427
 428                         /* XXX - should be combined */
 429                 case IP_RSVP_VIF_ON:
 430                         error = ip_rsvp_vif_init(so, sopt);
 431                         break;
 432
 433                 case IP_RSVP_VIF_OFF:
 434                         error = ip_rsvp_vif_done(so, sopt);
 435                         break;
 436
 437                 case MRT_INIT:
 438                 case MRT_DONE:
 439                 case MRT_ADD_VIF:
 440                 case MRT_DEL_VIF:
 441                 case MRT_ADD_MFC:
 442                 case MRT_DEL_MFC:
 443                 case MRT_VERSION:
 444                 case MRT_ASSERT:
 445                         error = ip_mrouter_set(so, sopt);
 446                         break;
 447
 448                 default:
 449                         error = ip_ctloutput(so, sopt);
 450                         break;
 451                 }
 452                 break;
 453         }
 454
 455         return (error);
 456 }
 457
 458 /*
 459  * This function exists solely to receive the PRC_IFDOWN messages which
 460  * are sent by if_down().  It looks for an ifaddr whose ifa_addr is sa,
 461  * and calls in_ifadown() to remove all routes corresponding to that address.
 462  * It also receives the PRC_IFUP messages from if_up() and reinstalls the
 463  * interface routes.
 464  */
 465 void
 466 rip_ctlinput(cmd, sa, vip)
 467         int cmd;
 468         struct sockaddr *sa;
 469         void *vip;
 470 {
 471         struct in_ifaddr *ia;
 472         struct ifnet *ifp;
 473         int err;
 474         int flags;
 475
 476         switch (cmd) {
 477         case PRC_IFDOWN:
 478                 for (ia = in_ifaddrhead.tqh_first; ia;
 479                      ia = ia->ia_link.tqe_next) {
 480                         if (ia->ia_ifa.ifa_addr == sa
 481                             && (ia->ia_flags & IFA_ROUTE)) {
 482                                 /*
 483                                  * in_ifscrub kills the interface route.
 484                                  */
 485                                 in_ifscrub(ia->ia_ifp, ia);
 486                                 /*
 487                                  * in_ifadown gets rid of all the rest of
 488                                  * the routes.  This is not quite the right
 489                                  * thing to do, but at least if we are running
 490                                  * a routing process they will come back.
 491                                  */
 492                                 in_ifadown(&ia->ia_ifa, 1);
 493                                 break;
 494                         }
 495                 }
 496                 break;
 497
 498         case PRC_IFUP:
 499                 for (ia = in_ifaddrhead.tqh_first; ia;
 500                      ia = ia->ia_link.tqe_next) {
 501                         if (ia->ia_ifa.ifa_addr == sa)
 502                                 break;
 503                 }
 504                 if (ia == 0 || (ia->ia_flags & IFA_ROUTE))
 505                         return;
 506                 flags = RTF_UP;
 507                 ifp = ia->ia_ifa.ifa_ifp;
 508
 509                 if ((ifp->if_flags & IFF_LOOPBACK)
 510                     || (ifp->if_flags & IFF_POINTOPOINT))
 511                         flags |= RTF_HOST;
 512
 513                 err = rtinit(&ia->ia_ifa, RTM_ADD, flags);
 514                 if (err == 0)
 515                         ia->ia_flags |= IFA_ROUTE;
 516                 break;
 517         }
 518 }
 519
 520 u_long  rip_sendspace = RIPSNDQ;
 521 u_long  rip_recvspace = RIPRCVQ;
 522
 523 SYSCTL_INT(_net_inet_raw, OID_AUTO, maxdgram, CTLFLAG_RW,
 524     &rip_sendspace, 0, "Maximum outgoing raw IP datagram size");
 525 SYSCTL_INT(_net_inet_raw, OID_AUTO, recvspace, CTLFLAG_RW,
 526     &rip_recvspace, 0, "Maximum incoming raw IP datagram size");
 527
 528 static int
 529 rip_attach(struct socket *so, int proto, struct proc *p)
 530 {
 531         struct inpcb *inp;
 532         int error, s;
 533
 534         inp = sotoinpcb(so);
 535         if (inp)
 536                 panic("rip_attach");
 537 #if __APPLE__
 538         if ((so->so_state & SS_PRIV) == 0)
 539                 return (EPERM);
 540 #else
 541         if (p && (error = suser(p)) != 0)
 542                 return error;
 543 #endif
 544
 545         error = soreserve(so, rip_sendspace, rip_recvspace);
 546         if (error)
 547                 return error;
 548         s = splnet();
 549         error = in_pcballoc(so, &ripcbinfo, p);
 550         splx(s);
 551         if (error)
 552                 return error;
 553         inp = (struct inpcb *)so->so_pcb;
 554         inp->inp_vflag |= INP_IPV4;
 555         inp->inp_ip_p = proto;
 556         inp->inp_ip_ttl = ip_defttl;
 557         return 0;
 558 }
 559
 560 __private_extern__ int
 561 rip_detach(struct socket *so)
 562 {
 563         struct inpcb *inp;
 564
 565         inp = sotoinpcb(so);
 566         if (inp == 0)
 567                 panic("rip_detach");
 568         if (so == ip_mrouter)
 569                 ip_mrouter_done();
 570         ip_rsvp_force_done(so);
 571         if (so == ip_rsvpd)
 572                 ip_rsvp_done();
 573         in_pcbdetach(inp);
 574         return 0;
 575 }
 576
 577 __private_extern__ int
 578 rip_abort(struct socket *so)
 579 {
 580         soisdisconnected(so);
 581         return rip_detach(so);
 582 }
 583
 584 __private_extern__ int
 585 rip_disconnect(struct socket *so)
 586 {
 587         if ((so->so_state & SS_ISCONNECTED) == 0)
 588                 return ENOTCONN;
 589         return rip_abort(so);
 590 }
 591
 592 __private_extern__ int
 593 rip_bind(struct socket *so, struct sockaddr *nam, struct proc *p)
 594 {
 595         struct inpcb *inp = sotoinpcb(so);
 596         struct sockaddr_in *addr = (struct sockaddr_in *)nam;
 597
 598         if (nam->sa_len != sizeof(*addr))
 599                 return EINVAL;
 600
 601         if (TAILQ_EMPTY(&ifnet) || ((addr->sin_family != AF_INET) &&
 602                                     (addr->sin_family != AF_IMPLINK)) ||
 603             (addr->sin_addr.s_addr &&
 604              ifa_ifwithaddr((struct sockaddr *)addr) == 0))
 605                 return EADDRNOTAVAIL;
 606         inp->inp_laddr = addr->sin_addr;
 607         return 0;
 608 }
 609
 610 __private_extern__ int
 611 rip_connect(struct socket *so, struct sockaddr *nam, struct proc *p)
 612 {
 613         struct inpcb *inp = sotoinpcb(so);
 614         struct sockaddr_in *addr = (struct sockaddr_in *)nam;
 615
 616         if (nam->sa_len != sizeof(*addr))
 617                 return EINVAL;
 618         if (TAILQ_EMPTY(&ifnet))
 619                 return EADDRNOTAVAIL;
 620         if ((addr->sin_family != AF_INET) &&
 621             (addr->sin_family != AF_IMPLINK))
 622                 return EAFNOSUPPORT;
 623         inp->inp_faddr = addr->sin_addr;
 624         soisconnected(so);
 625         return 0;
 626 }
 627
 628 __private_extern__ int
 629 rip_shutdown(struct socket *so)
 630 {
 631         socantsendmore(so);
 632         return 0;
 633 }
 634
 635 __private_extern__ int
 636 rip_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *nam,
 637          struct mbuf *control, struct proc *p)
 638 {
 639         struct inpcb *inp = sotoinpcb(so);
 640         register u_long dst;
 641
 642         if (so->so_state & SS_ISCONNECTED) {
 643                 if (nam) {
 644                         m_freem(m);
 645                         return EISCONN;
 646                 }
 647                 dst = inp->inp_faddr.s_addr;
 648         } else {
 649                 if (nam == NULL) {
 650                         m_freem(m);
 651                         return ENOTCONN;
 652                 }
 653                 dst = ((struct sockaddr_in *)nam)->sin_addr.s_addr;
 654         }
 655         return rip_output(m, so, dst);
 656 }
 657
 658 static int
 659 rip_pcblist SYSCTL_HANDLER_ARGS
 660 {
 661         int error, i, n, s;
 662         struct inpcb *inp, **inp_list;
 663         inp_gen_t gencnt;
 664         struct xinpgen xig;
 665
 666         /*
 667          * The process of preparing the TCB list is too time-consuming and
 668          * resource-intensive to repeat twice on every request.
 669          */
 670         if (req->oldptr == 0) {
 671                 n = ripcbinfo.ipi_count;
 672                 req->oldidx = 2 * (sizeof xig)
 673                         + (n + n/8) * sizeof(struct xinpcb);
 674                 return 0;
 675         }
 676
 677         if (req->newptr != 0)
 678                 return EPERM;
 679
 680         /*
 681          * OK, now we're committed to doing something.
 682          */
 683         s = splnet();
 684         gencnt = ripcbinfo.ipi_gencnt;
 685         n = ripcbinfo.ipi_count;
 686         splx(s);
 687
 688         xig.xig_len = sizeof xig;
 689         xig.xig_count = n;
 690         xig.xig_gen = gencnt;
 691         xig.xig_sogen = so_gencnt;
 692         error = SYSCTL_OUT(req, &xig, sizeof xig);
 693         if (error)
 694                 return error;
 695     /*
 696      * We are done if there is no pcb
 697      */
 698     if (n == 0)
 699         return 0;
 700
 701         inp_list = _MALLOC(n * sizeof *inp_list, M_TEMP, M_WAITOK);
 702         if (inp_list == 0)
 703                 return ENOMEM;
 704
 705         s = splnet();
 706         for (inp = ripcbinfo.listhead->lh_first, i = 0; inp && i < n;
 707              inp = inp->inp_list.le_next) {
 708                 if (inp->inp_gencnt <= gencnt)
 709                         inp_list[i++] = inp;
 710         }
 711         splx(s);
 712         n = i;
 713
 714         error = 0;
 715         for (i = 0; i < n; i++) {
 716                 inp = inp_list[i];
 717                 if (inp->inp_gencnt <= gencnt) {
 718                         struct xinpcb xi;
 719                         xi.xi_len = sizeof xi;
 720                         /* XXX should avoid extra copy */
 721                         bcopy(inp, &xi.xi_inp, sizeof *inp);
 722                         if (inp->inp_socket)
 723                                 sotoxsocket(inp->inp_socket, &xi.xi_socket);
 724                         error = SYSCTL_OUT(req, &xi, sizeof xi);
 725                 }
 726         }
 727         if (!error) {
 728                 /*
 729                  * Give the user an updated idea of our state.
 730                  * If the generation differs from what we told
 731                  * her before, she knows that something happened
 732                  * while we were processing this request, and it
 733                  * might be necessary to retry.
 734                  */
 735                 s = splnet();
 736                 xig.xig_gen = ripcbinfo.ipi_gencnt;
 737                 xig.xig_sogen = so_gencnt;
 738                 xig.xig_count = ripcbinfo.ipi_count;
 739                 splx(s);
 740                 error = SYSCTL_OUT(req, &xig, sizeof xig);
 741         }
 742         FREE(inp_list, M_TEMP);
 743         return error;
 744 }
 745
 746 SYSCTL_PROC(_net_inet_raw, OID_AUTO/*XXX*/, pcblist, CTLFLAG_RD, 0, 0,
 747             rip_pcblist, "S,xinpcb", "List of active raw IP sockets");
 748
 749 struct pr_usrreqs rip_usrreqs = {
 750         rip_abort, pru_accept_notsupp, rip_attach, rip_bind, rip_connect,
 751         pru_connect2_notsupp, in_control, rip_detach, rip_disconnect,
 752         pru_listen_notsupp, in_setpeeraddr, pru_rcvd_notsupp,
 753         pru_rcvoob_notsupp, rip_send, pru_sense_null, rip_shutdown,
 754         in_setsockaddr, sosend, soreceive, sopoll
 755 };