bsd/netinet/raw_ip.c

   1 /*
   2  * Copyright (c) 2000 Apple Computer, Inc. All rights reserved.
   3  *
   4  * @APPLE_LICENSE_OSREFERENCE_HEADER_START@
   5  *
   6  * This file contains Original Code and/or Modifications of Original Code
   7  * as defined in and that are subject to the Apple Public Source License
   8  * Version 2.0 (the 'License'). You may not use this file except in
   9  * compliance with the License.  The rights granted to you under the
  10  * License may not be used to create, or enable the creation or
  11  * redistribution of, unlawful or unlicensed copies of an Apple operating
  12  * system, or to circumvent, violate, or enable the circumvention or
  13  * violation of, any terms of an Apple operating system software license
  14  * agreement.
  15  *
  16  * Please obtain a copy of the License at
  17  * http://www.opensource.apple.com/apsl/ and read it before using this
  18  * file.
  19  *
  20  * The Original Code and all software distributed under the License are
  21  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  22  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  23  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  24  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  25  * Please see the License for the specific language governing rights and
  26  * limitations under the License.
  27  *
  28  * @APPLE_LICENSE_OSREFERENCE_HEADER_END@
  29  */
  30 /*
  31  * Copyright (c) 1982, 1986, 1988, 1993
  32  *      The Regents of the University of California.  All rights reserved.
  33  *
  34  * Redistribution and use in source and binary forms, with or without
  35  * modification, are permitted provided that the following conditions
  36  * are met:
  37  * 1. Redistributions of source code must retain the above copyright
  38  *    notice, this list of conditions and the following disclaimer.
  39  * 2. Redistributions in binary form must reproduce the above copyright
  40  *    notice, this list of conditions and the following disclaimer in the
  41  *    documentation and/or other materials provided with the distribution.
  42  * 3. All advertising materials mentioning features or use of this software
  43  *    must display the following acknowledgement:
  44  *      This product includes software developed by the University of
  45  *      California, Berkeley and its contributors.
  46  * 4. Neither the name of the University nor the names of its contributors
  47  *    may be used to endorse or promote products derived from this software
  48  *    without specific prior written permission.
  49  *
  50  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  51  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  52  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  53  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  54  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  55  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  56  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  57  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  58  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  59  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  60  * SUCH DAMAGE.
  61  *
  62  *      @(#)raw_ip.c    8.7 (Berkeley) 5/15/95
  63  */
  64
  65 #include <sys/param.h>
  66 #include <sys/systm.h>
  67 #include <sys/kernel.h>
  68 #include <sys/malloc.h>
  69 #include <sys/mbuf.h>
  70 #include <sys/proc.h>
  71 #include <sys/domain.h>
  72 #include <sys/protosw.h>
  73 #include <sys/socket.h>
  74 #include <sys/socketvar.h>
  75 #include <sys/sysctl.h>
  76
  77 #if __FreeBSD__
  78 #include <vm/vm_zone.h>
  79 #endif
  80
  81 #include <net/if.h>
  82 #include <net/route.h>
  83
  84 #define _IP_VHL
  85 #include <netinet/in.h>
  86 #include <netinet/in_systm.h>
  87 #include <netinet/ip.h>
  88 #include <netinet/in_pcb.h>
  89 #include <netinet/in_var.h>
  90 #include <netinet/ip_var.h>
  91 #include <netinet/ip_mroute.h>
  92
  93 #include <netinet/ip_fw.h>
  94
  95 #if IPSEC
  96 #include <netinet6/ipsec.h>
  97 #endif /*IPSEC*/
  98
  99 #if DUMMYNET
 100 #include <netinet/ip_dummynet.h>
 101 #endif
 102
 103 #if IPSEC
 104 extern int ipsec_bypass;
 105 extern lck_mtx_t *sadb_mutex;
 106 #endif
 107
 108 extern u_long  route_generation;
 109 struct  inpcbhead ripcb;
 110 struct  inpcbinfo ripcbinfo;
 111
 112 /* control hooks for ipfw and dummynet */
 113 ip_fw_ctl_t *ip_fw_ctl_ptr;
 114 #if DUMMYNET
 115 ip_dn_ctl_t *ip_dn_ctl_ptr;
 116 #endif /* DUMMYNET */
 117
 118 /*
 119  * Nominal space allocated to a raw ip socket.
 120  */
 121 #define RIPSNDQ         8192
 122 #define RIPRCVQ         8192
 123
 124 /*
 125  * Raw interface to IP protocol.
 126  */
 127
 128 /*
 129  * Initialize raw connection block q.
 130  */
 131 void
 132 rip_init()
 133 {
 134         struct inpcbinfo *pcbinfo;
 135
 136         LIST_INIT(&ripcb);
 137         ripcbinfo.listhead = &ripcb;
 138         /*
 139          * XXX We don't use the hash list for raw IP, but it's easier
 140          * to allocate a one entry hash list than it is to check all
 141          * over the place for hashbase == NULL.
 142          */
 143         ripcbinfo.hashbase = hashinit(1, M_PCB, &ripcbinfo.hashmask);
 144         ripcbinfo.porthashbase = hashinit(1, M_PCB, &ripcbinfo.porthashmask);
 145
 146         ripcbinfo.ipi_zone = (void *) zinit(sizeof(struct inpcb),
 147                                             (4096 * sizeof(struct inpcb)),
 148                                             4096, "ripzone");
 149
 150         pcbinfo = &ripcbinfo;
 151         /*
 152          * allocate lock group attribute and group for udp pcb mutexes
 153          */
 154         pcbinfo->mtx_grp_attr = lck_grp_attr_alloc_init();
 155
 156         pcbinfo->mtx_grp = lck_grp_alloc_init("ripcb", pcbinfo->mtx_grp_attr);
 157
 158         /*
 159          * allocate the lock attribute for udp pcb mutexes
 160          */
 161         pcbinfo->mtx_attr = lck_attr_alloc_init();
 162
 163         if ((pcbinfo->mtx = lck_rw_alloc_init(pcbinfo->mtx_grp, pcbinfo->mtx_attr)) == NULL)
 164                 return; /* pretty much dead if this fails... */
 165
 166 }
 167
 168 static struct   sockaddr_in ripsrc = { sizeof(ripsrc), AF_INET };
 169 /*
 170  * Setup generic address and protocol structures
 171  * for raw_input routine, then pass them along with
 172  * mbuf chain.
 173  */
 174 void
 175 rip_input(m, iphlen)
 176         struct mbuf *m;
 177         int iphlen;
 178 {
 179         register struct ip *ip = mtod(m, struct ip *);
 180         register struct inpcb *inp;
 181         struct inpcb *last = 0;
 182         struct mbuf *opts = 0;
 183         int skipit;
 184
 185         ripsrc.sin_addr = ip->ip_src;
 186         lck_rw_lock_shared(ripcbinfo.mtx);
 187         LIST_FOREACH(inp, &ripcb, inp_list) {
 188 #if INET6
 189                 if ((inp->inp_vflag & INP_IPV4) == 0)
 190                         continue;
 191 #endif
 192                 if (inp->inp_ip_p && (inp->inp_ip_p != ip->ip_p))
 193                         continue;
 194                 if (inp->inp_laddr.s_addr &&
 195                   inp->inp_laddr.s_addr != ip->ip_dst.s_addr)
 196                         continue;
 197                 if (inp->inp_faddr.s_addr &&
 198                   inp->inp_faddr.s_addr != ip->ip_src.s_addr)
 199                         continue;
 200                 if (last) {
 201                         struct mbuf *n = m_copy(m, 0, (int)M_COPYALL);
 202
 203 #if IPSEC
 204                         /* check AH/ESP integrity. */
 205                         skipit = 0;
 206                         if (ipsec_bypass == 0 && n) {
 207                                 lck_mtx_lock(sadb_mutex);
 208                                 if (ipsec4_in_reject_so(n, last->inp_socket)) {
 209                                         m_freem(n);
 210                                         ipsecstat.in_polvio++;
 211                                         /* do not inject data to pcb */
 212                                         skipit = 1;
 213                                 }
 214                                 lck_mtx_unlock(sadb_mutex);
 215                         }
 216 #endif /*IPSEC*/
 217                         if (n && skipit == 0) {
 218                                 int error = 0;
 219                                 if (last->inp_flags & INP_CONTROLOPTS ||
 220                                     last->inp_socket->so_options & SO_TIMESTAMP)
 221                                     ip_savecontrol(last, &opts, ip, n);
 222                                 if (last->inp_flags & INP_STRIPHDR) {
 223                                         n->m_len -= iphlen;
 224                                         n->m_pkthdr.len -= iphlen;
 225                                         n->m_data += iphlen;
 226                                 }
 227 // ###LOCK need to lock that socket?
 228                                 if (sbappendaddr(&last->inp_socket->so_rcv,
 229                                     (struct sockaddr *)&ripsrc, n,
 230                                     opts, &error) != 0) {
 231                                         sorwakeup(last->inp_socket);
 232                                 }
 233                                 else {
 234                                         if (error) {
 235                                                 /* should notify about lost packet */
 236                                                 kprintf("rip_input can't append to socket\n");
 237                                         }
 238                                 }
 239                                 opts = 0;
 240                         }
 241                 }
 242                 last = inp;
 243         }
 244         lck_rw_done(ripcbinfo.mtx);
 245 #if IPSEC
 246         /* check AH/ESP integrity. */
 247         skipit = 0;
 248         if (ipsec_bypass == 0 && last) {
 249                 lck_mtx_lock(sadb_mutex);
 250                 if (ipsec4_in_reject_so(m, last->inp_socket)) {
 251                         m_freem(m);
 252                         ipsecstat.in_polvio++;
 253                         ipstat.ips_delivered--;
 254                         /* do not inject data to pcb */
 255                         skipit = 1;
 256                 }
 257                 lck_mtx_unlock(sadb_mutex);
 258         }
 259 #endif /*IPSEC*/
 260         if (skipit == 0) {
 261                 if (last) {
 262                         if (last->inp_flags & INP_CONTROLOPTS ||
 263                                 last->inp_socket->so_options & SO_TIMESTAMP)
 264                                 ip_savecontrol(last, &opts, ip, m);
 265                         if (last->inp_flags & INP_STRIPHDR) {
 266                                 m->m_len -= iphlen;
 267                                 m->m_pkthdr.len -= iphlen;
 268                                 m->m_data += iphlen;
 269                         }
 270                         if (sbappendaddr(&last->inp_socket->so_rcv,
 271                                 (struct sockaddr *)&ripsrc, m, opts, NULL) != 0) {
 272                                 sorwakeup(last->inp_socket);
 273                         } else {
 274                                 kprintf("rip_input(2) can't append to socket\n");
 275                         }
 276                 } else {
 277                         m_freem(m);
 278                         ipstat.ips_noproto++;
 279                         ipstat.ips_delivered--;
 280                 }
 281         }
 282 }
 283
 284 /*
 285  * Generate IP header and pass packet to ip_output.
 286  * Tack on options user may have setup with control call.
 287  */
 288 int
 289 rip_output(m, so, dst)
 290         register struct mbuf *m;
 291         struct socket *so;
 292         u_long dst;
 293 {
 294         register struct ip *ip;
 295         register struct inpcb *inp = sotoinpcb(so);
 296         int flags = (so->so_options & SO_DONTROUTE) | IP_ALLOWBROADCAST;
 297
 298         /*
 299          * If the user handed us a complete IP packet, use it.
 300          * Otherwise, allocate an mbuf for a header and fill it in.
 301          */
 302         if ((inp->inp_flags & INP_HDRINCL) == 0) {
 303                 if (m->m_pkthdr.len + sizeof(struct ip) > IP_MAXPACKET) {
 304                         m_freem(m);
 305                         return(EMSGSIZE);
 306                 }
 307                 M_PREPEND(m, sizeof(struct ip), M_WAIT);
 308                 ip = mtod(m, struct ip *);
 309                 ip->ip_tos = inp->inp_ip_tos;
 310                 ip->ip_off = 0;
 311                 ip->ip_p = inp->inp_ip_p;
 312                 ip->ip_len = m->m_pkthdr.len;
 313                 ip->ip_src = inp->inp_laddr;
 314                 ip->ip_dst.s_addr = dst;
 315                 ip->ip_ttl = inp->inp_ip_ttl;
 316         } else {
 317                 if (m->m_pkthdr.len > IP_MAXPACKET) {
 318                         m_freem(m);
 319                         return(EMSGSIZE);
 320                 }
 321                 ip = mtod(m, struct ip *);
 322                 /* don't allow both user specified and setsockopt options,
 323                    and don't allow packet length sizes that will crash */
 324                 if (((IP_VHL_HL(ip->ip_vhl) != (sizeof (*ip) >> 2))
 325                      && inp->inp_options)
 326                     || (ip->ip_len > m->m_pkthdr.len)
 327                     || (ip->ip_len < (IP_VHL_HL(ip->ip_vhl) << 2))) {
 328                         m_freem(m);
 329                         return EINVAL;
 330                 }
 331                 if (ip->ip_id == 0)
 332 #if RANDOM_IP_ID
 333                         ip->ip_id = ip_randomid();
 334 #else
 335                         ip->ip_id = htons(ip_id++);
 336 #endif
 337                 /* XXX prevent ip_output from overwriting header fields */
 338                 flags |= IP_RAWOUTPUT;
 339                 ipstat.ips_rawout++;
 340         }
 341
 342 #if IPSEC
 343         if (ipsec_bypass == 0 && ipsec_setsocket(m, so) != 0) {
 344                 m_freem(m);
 345                 return ENOBUFS;
 346         }
 347 #endif /*IPSEC*/
 348
 349         if (inp->inp_route.ro_rt && inp->inp_route.ro_rt->generation_id != route_generation) {
 350                 rtfree(inp->inp_route.ro_rt);
 351                 inp->inp_route.ro_rt = (struct rtentry *)0;
 352         }
 353
 354         return (ip_output_list(m, 0, inp->inp_options, &inp->inp_route, flags,
 355                           inp->inp_moptions));
 356 }
 357
 358 extern int
 359 load_ipfw()
 360 {
 361         kern_return_t   err;
 362
 363         ipfw_init();
 364
 365 #if DUMMYNET
 366         if (!DUMMYNET_LOADED)
 367                 ip_dn_init();
 368 #endif /* DUMMYNET */
 369         err = 0;
 370
 371         return err == 0 && ip_fw_ctl_ptr == NULL ? -1 : err;
 372 }
 373
 374 /*
 375  * Raw IP socket option processing.
 376  */
 377 int
 378 rip_ctloutput(so, sopt)
 379         struct socket *so;
 380         struct sockopt *sopt;
 381 {
 382         struct  inpcb *inp = sotoinpcb(so);
 383         int     error, optval;
 384
 385         if (sopt->sopt_level != IPPROTO_IP)
 386                 return (EINVAL);
 387
 388         error = 0;
 389
 390         switch (sopt->sopt_dir) {
 391         case SOPT_GET:
 392                 switch (sopt->sopt_name) {
 393                 case IP_HDRINCL:
 394                         optval = inp->inp_flags & INP_HDRINCL;
 395                         error = sooptcopyout(sopt, &optval, sizeof optval);
 396                         break;
 397
 398         case IP_STRIPHDR:
 399             optval = inp->inp_flags & INP_STRIPHDR;
 400             error = sooptcopyout(sopt, &optval, sizeof optval);
 401             break;
 402
 403                 case IP_FW_ADD:
 404                 case IP_FW_GET:
 405                 case IP_OLD_FW_ADD:
 406                 case IP_OLD_FW_GET:
 407                         if (ip_fw_ctl_ptr == 0)
 408                                 error = load_ipfw();
 409                         if (ip_fw_ctl_ptr && error == 0)
 410                                 error = ip_fw_ctl_ptr(sopt);
 411                         else
 412                                 error = ENOPROTOOPT;
 413                         break;
 414
 415 #if DUMMYNET
 416                 case IP_DUMMYNET_GET:
 417                         if (DUMMYNET_LOADED)
 418                                 error = ip_dn_ctl_ptr(sopt);
 419                         else
 420                                 error = ENOPROTOOPT;
 421                         break ;
 422 #endif /* DUMMYNET */
 423
 424                 case MRT_INIT:
 425                 case MRT_DONE:
 426                 case MRT_ADD_VIF:
 427                 case MRT_DEL_VIF:
 428                 case MRT_ADD_MFC:
 429                 case MRT_DEL_MFC:
 430                 case MRT_VERSION:
 431                 case MRT_ASSERT:
 432                         error = ip_mrouter_get(so, sopt);
 433                         break;
 434
 435                 default:
 436                         error = ip_ctloutput(so, sopt);
 437                         break;
 438                 }
 439                 break;
 440
 441         case SOPT_SET:
 442                 switch (sopt->sopt_name) {
 443                 case IP_HDRINCL:
 444                         error = sooptcopyin(sopt, &optval, sizeof optval,
 445                                             sizeof optval);
 446                         if (error)
 447                                 break;
 448                         if (optval)
 449                                 inp->inp_flags |= INP_HDRINCL;
 450                         else
 451                                 inp->inp_flags &= ~INP_HDRINCL;
 452                         break;
 453
 454         case IP_STRIPHDR:
 455             error = sooptcopyin(sopt, &optval, sizeof optval,
 456                         sizeof optval);
 457             if (error)
 458                 break;
 459             if (optval)
 460                 inp->inp_flags |= INP_STRIPHDR;
 461             else
 462                 inp->inp_flags &= ~INP_STRIPHDR;
 463             break;
 464
 465
 466                 case IP_FW_ADD:
 467                 case IP_FW_DEL:
 468                 case IP_FW_FLUSH:
 469                 case IP_FW_ZERO:
 470                 case IP_FW_RESETLOG:
 471                 case IP_OLD_FW_ADD:
 472                 case IP_OLD_FW_DEL:
 473                 case IP_OLD_FW_FLUSH:
 474                 case IP_OLD_FW_ZERO:
 475                 case IP_OLD_FW_RESETLOG:
 476                         if (ip_fw_ctl_ptr == 0)
 477                                 error = load_ipfw();
 478                         if (ip_fw_ctl_ptr && error == 0)
 479                                 error = ip_fw_ctl_ptr(sopt);
 480                         else
 481                                 error = ENOPROTOOPT;
 482                         break;
 483
 484 #if DUMMYNET
 485                 case IP_DUMMYNET_CONFIGURE:
 486                 case IP_DUMMYNET_DEL:
 487                 case IP_DUMMYNET_FLUSH:
 488                         if (DUMMYNET_LOADED)
 489                                 error = ip_dn_ctl_ptr(sopt);
 490                         else
 491                                 error = ENOPROTOOPT ;
 492                         break ;
 493 #endif
 494
 495                 case IP_RSVP_ON:
 496                         error = ip_rsvp_init(so);
 497                         break;
 498
 499                 case IP_RSVP_OFF:
 500                         error = ip_rsvp_done();
 501                         break;
 502
 503                         /* XXX - should be combined */
 504                 case IP_RSVP_VIF_ON:
 505                         error = ip_rsvp_vif_init(so, sopt);
 506                         break;
 507
 508                 case IP_RSVP_VIF_OFF:
 509                         error = ip_rsvp_vif_done(so, sopt);
 510                         break;
 511
 512                 case MRT_INIT:
 513                 case MRT_DONE:
 514                 case MRT_ADD_VIF:
 515                 case MRT_DEL_VIF:
 516                 case MRT_ADD_MFC:
 517                 case MRT_DEL_MFC:
 518                 case MRT_VERSION:
 519                 case MRT_ASSERT:
 520                         error = ip_mrouter_set(so, sopt);
 521                         break;
 522
 523                 default:
 524                         error = ip_ctloutput(so, sopt);
 525                         break;
 526                 }
 527                 break;
 528         }
 529
 530         return (error);
 531 }
 532
 533 /*
 534  * This function exists solely to receive the PRC_IFDOWN messages which
 535  * are sent by if_down().  It looks for an ifaddr whose ifa_addr is sa,
 536  * and calls in_ifadown() to remove all routes corresponding to that address.
 537  * It also receives the PRC_IFUP messages from if_up() and reinstalls the
 538  * interface routes.
 539  */
 540 void
 541 rip_ctlinput(cmd, sa, vip)
 542         int cmd;
 543         struct sockaddr *sa;
 544         void *vip;
 545 {
 546         struct in_ifaddr *ia;
 547         struct ifnet *ifp;
 548         int err;
 549         int flags;
 550
 551         switch (cmd) {
 552         case PRC_IFDOWN:
 553                 lck_mtx_lock(rt_mtx);
 554                 for (ia = in_ifaddrhead.tqh_first; ia;
 555                      ia = ia->ia_link.tqe_next) {
 556                         if (ia->ia_ifa.ifa_addr == sa
 557                             && (ia->ia_flags & IFA_ROUTE)) {
 558                                 /*
 559                                  * in_ifscrub kills the interface route.
 560                                  */
 561                                 in_ifscrub(ia->ia_ifp, ia, 1);
 562                                 /*
 563                                  * in_ifadown gets rid of all the rest of
 564                                  * the routes.  This is not quite the right
 565                                  * thing to do, but at least if we are running
 566                                  * a routing process they will come back.
 567                                  */
 568                                 in_ifadown(&ia->ia_ifa, 1);
 569                                 break;
 570                         }
 571                 }
 572                 lck_mtx_unlock(rt_mtx);
 573                 break;
 574
 575         case PRC_IFUP:
 576                 lck_mtx_lock(rt_mtx);
 577                 for (ia = in_ifaddrhead.tqh_first; ia;
 578                      ia = ia->ia_link.tqe_next) {
 579                         if (ia->ia_ifa.ifa_addr == sa)
 580                                 break;
 581                 }
 582                 if (ia == 0 || (ia->ia_flags & IFA_ROUTE)) {
 583                         lck_mtx_unlock(rt_mtx);
 584                         return;
 585                 }
 586                 flags = RTF_UP;
 587                 ifp = ia->ia_ifa.ifa_ifp;
 588
 589                 if ((ifp->if_flags & IFF_LOOPBACK)
 590                     || (ifp->if_flags & IFF_POINTOPOINT))
 591                         flags |= RTF_HOST;
 592
 593                 err = rtinit_locked(&ia->ia_ifa, RTM_ADD, flags);
 594                 lck_mtx_unlock(rt_mtx);
 595                 if (err == 0)
 596                         ia->ia_flags |= IFA_ROUTE;
 597                 break;
 598         }
 599 }
 600
 601 u_long  rip_sendspace = RIPSNDQ;
 602 u_long  rip_recvspace = RIPRCVQ;
 603
 604 SYSCTL_INT(_net_inet_raw, OID_AUTO, maxdgram, CTLFLAG_RW,
 605     &rip_sendspace, 0, "Maximum outgoing raw IP datagram size");
 606 SYSCTL_INT(_net_inet_raw, OID_AUTO, recvspace, CTLFLAG_RW,
 607     &rip_recvspace, 0, "Maximum incoming raw IP datagram size");
 608
 609 static int
 610 rip_attach(struct socket *so, int proto, struct proc *p)
 611 {
 612         struct inpcb *inp;
 613         int error, s;
 614
 615         inp = sotoinpcb(so);
 616         if (inp)
 617                 panic("rip_attach");
 618 #if __APPLE__
 619         if ((so->so_state & SS_PRIV) == 0)
 620                 return (EPERM);
 621 #else
 622         if (p && (error = suser(p)) != 0)
 623                 return error;
 624 #endif
 625
 626         error = soreserve(so, rip_sendspace, rip_recvspace);
 627         if (error)
 628                 return error;
 629         s = splnet();
 630         error = in_pcballoc(so, &ripcbinfo, p);
 631         splx(s);
 632         if (error)
 633                 return error;
 634         inp = (struct inpcb *)so->so_pcb;
 635         inp->inp_vflag |= INP_IPV4;
 636         inp->inp_ip_p = proto;
 637         inp->inp_ip_ttl = ip_defttl;
 638         return 0;
 639 }
 640
 641 __private_extern__ int
 642 rip_detach(struct socket *so)
 643 {
 644         struct inpcb *inp;
 645
 646         inp = sotoinpcb(so);
 647         if (inp == 0)
 648                 panic("rip_detach");
 649         if (so == ip_mrouter)
 650                 ip_mrouter_done();
 651         ip_rsvp_force_done(so);
 652         if (so == ip_rsvpd)
 653                 ip_rsvp_done();
 654         in_pcbdetach(inp);
 655         return 0;
 656 }
 657
 658 __private_extern__ int
 659 rip_abort(struct socket *so)
 660 {
 661         soisdisconnected(so);
 662         return rip_detach(so);
 663 }
 664
 665 __private_extern__ int
 666 rip_disconnect(struct socket *so)
 667 {
 668         if ((so->so_state & SS_ISCONNECTED) == 0)
 669                 return ENOTCONN;
 670         return rip_abort(so);
 671 }
 672
 673 __private_extern__ int
 674 rip_bind(struct socket *so, struct sockaddr *nam, struct proc *p)
 675 {
 676         struct inpcb *inp = sotoinpcb(so);
 677         struct sockaddr_in *addr = (struct sockaddr_in *)nam;
 678         struct ifaddr *ifa = NULL;
 679
 680         if (nam->sa_len != sizeof(*addr))
 681                 return EINVAL;
 682
 683         if (TAILQ_EMPTY(&ifnet_head) || ((addr->sin_family != AF_INET) &&
 684                                     (addr->sin_family != AF_IMPLINK)) ||
 685             (addr->sin_addr.s_addr &&
 686              (ifa = ifa_ifwithaddr((struct sockaddr *)addr)) == 0)) {
 687                 return EADDRNOTAVAIL;
 688         }
 689         else if (ifa) {
 690                 ifafree(ifa);
 691                 ifa = NULL;
 692         }
 693         inp->inp_laddr = addr->sin_addr;
 694         return 0;
 695 }
 696
 697 __private_extern__ int
 698 rip_connect(struct socket *so, struct sockaddr *nam, struct proc *p)
 699 {
 700         struct inpcb *inp = sotoinpcb(so);
 701         struct sockaddr_in *addr = (struct sockaddr_in *)nam;
 702
 703         if (nam->sa_len != sizeof(*addr))
 704                 return EINVAL;
 705         if (TAILQ_EMPTY(&ifnet_head))
 706                 return EADDRNOTAVAIL;
 707         if ((addr->sin_family != AF_INET) &&
 708             (addr->sin_family != AF_IMPLINK))
 709                 return EAFNOSUPPORT;
 710         inp->inp_faddr = addr->sin_addr;
 711         soisconnected(so);
 712         return 0;
 713 }
 714
 715 __private_extern__ int
 716 rip_shutdown(struct socket *so)
 717 {
 718         socantsendmore(so);
 719         return 0;
 720 }
 721
 722 __private_extern__ int
 723 rip_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *nam,
 724          struct mbuf *control, struct proc *p)
 725 {
 726         struct inpcb *inp = sotoinpcb(so);
 727         register u_long dst;
 728
 729         if (so->so_state & SS_ISCONNECTED) {
 730                 if (nam) {
 731                         m_freem(m);
 732                         return EISCONN;
 733                 }
 734                 dst = inp->inp_faddr.s_addr;
 735         } else {
 736                 if (nam == NULL) {
 737                         m_freem(m);
 738                         return ENOTCONN;
 739                 }
 740                 dst = ((struct sockaddr_in *)nam)->sin_addr.s_addr;
 741         }
 742         return rip_output(m, so, dst);
 743 }
 744
 745 /* note: rip_unlock is called from different protos  instead of the generic socket_unlock,
 746  * it will handle the socket dealloc on last reference
 747  * */
 748 int
 749 rip_unlock(struct socket *so, int refcount, int debug)
 750 {
 751         int lr_saved;
 752         struct inpcb *inp = sotoinpcb(so);
 753
 754         if (debug == 0)
 755                 lr_saved = (unsigned int) __builtin_return_address(0);
 756         else lr_saved = debug;
 757
 758         if (refcount) {
 759                 if (so->so_usecount <= 0)
 760                         panic("rip_unlock: bad refoucnt so=%x val=%x\n", so, so->so_usecount);
 761                 so->so_usecount--;
 762                 if (so->so_usecount == 0 && (inp->inp_wantcnt == WNT_STOPUSING)) {
 763                         /* cleanup after last reference */
 764                         lck_mtx_unlock(so->so_proto->pr_domain->dom_mtx);
 765                         lck_rw_lock_exclusive(ripcbinfo.mtx);
 766                         in_pcbdispose(inp);
 767                         lck_rw_done(ripcbinfo.mtx);
 768                         return(0);
 769                 }
 770         }
 771         so->unlock_lr[so->next_unlock_lr] = (u_int *)lr_saved;
 772         so->next_unlock_lr = (so->next_unlock_lr+1) % SO_LCKDBG_MAX;
 773         lck_mtx_unlock(so->so_proto->pr_domain->dom_mtx);
 774         return(0);
 775 }
 776
 777 static int
 778 rip_pcblist SYSCTL_HANDLER_ARGS
 779 {
 780         int error, i, n, s;
 781         struct inpcb *inp, **inp_list;
 782         inp_gen_t gencnt;
 783         struct xinpgen xig;
 784
 785         /*
 786          * The process of preparing the TCB list is too time-consuming and
 787          * resource-intensive to repeat twice on every request.
 788          */
 789         lck_rw_lock_exclusive(ripcbinfo.mtx);
 790         if (req->oldptr == USER_ADDR_NULL) {
 791                 n = ripcbinfo.ipi_count;
 792                 req->oldidx = 2 * (sizeof xig)
 793                         + (n + n/8) * sizeof(struct xinpcb);
 794                 lck_rw_done(ripcbinfo.mtx);
 795                 return 0;
 796         }
 797
 798         if (req->newptr != USER_ADDR_NULL) {
 799                 lck_rw_done(ripcbinfo.mtx);
 800                 return EPERM;
 801         }
 802
 803         /*
 804          * OK, now we're committed to doing something.
 805          */
 806         gencnt = ripcbinfo.ipi_gencnt;
 807         n = ripcbinfo.ipi_count;
 808
 809         bzero(&xig, sizeof(xig));
 810         xig.xig_len = sizeof xig;
 811         xig.xig_count = n;
 812         xig.xig_gen = gencnt;
 813         xig.xig_sogen = so_gencnt;
 814         error = SYSCTL_OUT(req, &xig, sizeof xig);
 815         if (error) {
 816                 lck_rw_done(ripcbinfo.mtx);
 817                 return error;
 818         }
 819     /*
 820      * We are done if there is no pcb
 821      */
 822     if (n == 0) {
 823         lck_rw_done(ripcbinfo.mtx);
 824         return 0;
 825     }
 826
 827         inp_list = _MALLOC(n * sizeof *inp_list, M_TEMP, M_WAITOK);
 828         if (inp_list == 0) {
 829                 lck_rw_done(ripcbinfo.mtx);
 830                 return ENOMEM;
 831         }
 832
 833         for (inp = ripcbinfo.listhead->lh_first, i = 0; inp && i < n;
 834              inp = inp->inp_list.le_next) {
 835                 if (inp->inp_gencnt <= gencnt && inp->inp_state != INPCB_STATE_DEAD)
 836                         inp_list[i++] = inp;
 837         }
 838         n = i;
 839
 840         error = 0;
 841         for (i = 0; i < n; i++) {
 842                 inp = inp_list[i];
 843                 if (inp->inp_gencnt <= gencnt && inp->inp_state != INPCB_STATE_DEAD) {
 844                         struct xinpcb xi;
 845
 846                         bzero(&xi, sizeof(xi));
 847                         xi.xi_len = sizeof xi;
 848                         /* XXX should avoid extra copy */
 849                         inpcb_to_compat(inp, &xi.xi_inp);
 850                         if (inp->inp_socket)
 851                                 sotoxsocket(inp->inp_socket, &xi.xi_socket);
 852                         error = SYSCTL_OUT(req, &xi, sizeof xi);
 853                 }
 854         }
 855         if (!error) {
 856                 /*
 857                  * Give the user an updated idea of our state.
 858                  * If the generation differs from what we told
 859                  * her before, she knows that something happened
 860                  * while we were processing this request, and it
 861                  * might be necessary to retry.
 862                  */
 863                 bzero(&xig, sizeof(xig));
 864                 xig.xig_len = sizeof xig;
 865                 xig.xig_gen = ripcbinfo.ipi_gencnt;
 866                 xig.xig_sogen = so_gencnt;
 867                 xig.xig_count = ripcbinfo.ipi_count;
 868                 error = SYSCTL_OUT(req, &xig, sizeof xig);
 869         }
 870         FREE(inp_list, M_TEMP);
 871         lck_rw_done(ripcbinfo.mtx);
 872         return error;
 873 }
 874
 875 SYSCTL_PROC(_net_inet_raw, OID_AUTO/*XXX*/, pcblist, CTLFLAG_RD, 0, 0,
 876             rip_pcblist, "S,xinpcb", "List of active raw IP sockets");
 877
 878 struct pr_usrreqs rip_usrreqs = {
 879         rip_abort, pru_accept_notsupp, rip_attach, rip_bind, rip_connect,
 880         pru_connect2_notsupp, in_control, rip_detach, rip_disconnect,
 881         pru_listen_notsupp, in_setpeeraddr, pru_rcvd_notsupp,
 882         pru_rcvoob_notsupp, rip_send, pru_sense_null, rip_shutdown,
 883         in_setsockaddr, sosend, soreceive, pru_sopoll_notsupp
 884 };