[BACK]Return to ip_input.c CVS log [TXT][DIR] Up to [cvs.NetBSD.org] / src / sys / netinet

Annotation of src/sys/netinet/ip_input.c, Revision 1.68

1.68    ! sommerfe    1: /*     $NetBSD: ip_input.c,v 1.67 1998/06/01 00:39:37 thorpej Exp $    */
1.58      tls         2:
                      3: /*-
                      4:  * Copyright (c) 1998 The NetBSD Foundation, Inc.
                      5:  * All rights reserved.
                      6:  *
                      7:  * This code is derived from software contributed to The NetBSD Foundation
                      8:  * by Public Access Networks Corporation ("Panix").  It was developed under
                      9:  * contract to Panix by Eric Haszlakiewicz and Thor Lancelot Simon.
                     10:  *
                     11:  * Redistribution and use in source and binary forms, with or without
                     12:  * modification, are permitted provided that the following conditions
                     13:  * are met:
                     14:  * 1. Redistributions of source code must retain the above copyright
                     15:  *    notice, this list of conditions and the following disclaimer.
                     16:  * 2. Redistributions in binary form must reproduce the above copyright
                     17:  *    notice, this list of conditions and the following disclaimer in the
                     18:  *    documentation and/or other materials provided with the distribution.
                     19:  * 3. All advertising materials mentioning features or use of this software
                     20:  *    must display the following acknowledgement:
                     21:  *     This product includes software developed by the NetBSD
                     22:  *     Foundation, Inc. and its contributors.
                     23:  * 4. Neither the name of The NetBSD Foundation nor the names of its
                     24:  *    contributors may be used to endorse or promote products derived
                     25:  *    from this software without specific prior written permission.
                     26:  *
                     27:  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
                     28:  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
                     29:  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
                     30:  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
                     31:  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
                     32:  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
                     33:  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
                     34:  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
                     35:  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
                     36:  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
                     37:  * POSSIBILITY OF SUCH DAMAGE.
                     38:  */
1.14      cgd        39:
1.1       cgd        40: /*
1.13      mycroft    41:  * Copyright (c) 1982, 1986, 1988, 1993
                     42:  *     The Regents of the University of California.  All rights reserved.
1.1       cgd        43:  *
                     44:  * Redistribution and use in source and binary forms, with or without
                     45:  * modification, are permitted provided that the following conditions
                     46:  * are met:
                     47:  * 1. Redistributions of source code must retain the above copyright
                     48:  *    notice, this list of conditions and the following disclaimer.
                     49:  * 2. Redistributions in binary form must reproduce the above copyright
                     50:  *    notice, this list of conditions and the following disclaimer in the
                     51:  *    documentation and/or other materials provided with the distribution.
                     52:  * 3. All advertising materials mentioning features or use of this software
                     53:  *    must display the following acknowledgement:
                     54:  *     This product includes software developed by the University of
                     55:  *     California, Berkeley and its contributors.
                     56:  * 4. Neither the name of the University nor the names of its contributors
                     57:  *    may be used to endorse or promote products derived from this software
                     58:  *    without specific prior written permission.
                     59:  *
                     60:  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
                     61:  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
                     62:  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
                     63:  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
                     64:  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
                     65:  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
                     66:  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
                     67:  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
                     68:  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
                     69:  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
                     70:  * SUCH DAMAGE.
                     71:  *
1.14      cgd        72:  *     @(#)ip_input.c  8.2 (Berkeley) 1/4/94
1.1       cgd        73:  */
1.55      scottr     74:
1.62      matt       75: #include "opt_gateway.h"
1.55      scottr     76: #include "opt_mrouting.h"
1.1       cgd        77:
1.5       mycroft    78: #include <sys/param.h>
                     79: #include <sys/systm.h>
                     80: #include <sys/malloc.h>
                     81: #include <sys/mbuf.h>
                     82: #include <sys/domain.h>
                     83: #include <sys/protosw.h>
                     84: #include <sys/socket.h>
1.44      thorpej    85: #include <sys/socketvar.h>
1.5       mycroft    86: #include <sys/errno.h>
                     87: #include <sys/time.h>
                     88: #include <sys/kernel.h>
1.28      christos   89: #include <sys/proc.h>
                     90:
                     91: #include <vm/vm.h>
                     92: #include <sys/sysctl.h>
1.1       cgd        93:
1.5       mycroft    94: #include <net/if.h>
1.44      thorpej    95: #include <net/if_dl.h>
1.5       mycroft    96: #include <net/route.h>
1.45      mrg        97: #include <net/pfil.h>
1.1       cgd        98:
1.5       mycroft    99: #include <netinet/in.h>
                    100: #include <netinet/in_systm.h>
                    101: #include <netinet/ip.h>
                    102: #include <netinet/in_pcb.h>
                    103: #include <netinet/in_var.h>
                    104: #include <netinet/ip_var.h>
                    105: #include <netinet/ip_icmp.h>
1.44      thorpej   106:
1.1       cgd       107: #ifndef        IPFORWARDING
                    108: #ifdef GATEWAY
                    109: #define        IPFORWARDING    1       /* forward IP packets not for us */
                    110: #else /* GATEWAY */
                    111: #define        IPFORWARDING    0       /* don't forward IP packets not for us */
                    112: #endif /* GATEWAY */
                    113: #endif /* IPFORWARDING */
                    114: #ifndef        IPSENDREDIRECTS
                    115: #define        IPSENDREDIRECTS 1
                    116: #endif
1.26      thorpej   117: #ifndef IPFORWSRCRT
1.47      cjs       118: #define        IPFORWSRCRT     1       /* forward source-routed packets */
                    119: #endif
                    120: #ifndef IPALLOWSRCRT
1.48      mrg       121: #define        IPALLOWSRCRT    1       /* allow source-routed packets */
1.26      thorpej   122: #endif
1.53      kml       123: #ifndef IPMTUDISC
                    124: #define IPMTUDISC      0
                    125: #endif
1.60      kml       126: #ifndef IPMTUDISCTIMEOUT
1.61      kml       127: #define IPMTUDISCTIMEOUT (10 * 60)     /* as per RFC 1191 */
1.60      kml       128: #endif
1.53      kml       129:
1.27      thorpej   130: /*
                    131:  * Note: DIRECTED_BROADCAST is handled this way so that previous
                    132:  * configuration using this option will Just Work.
                    133:  */
                    134: #ifndef IPDIRECTEDBCAST
                    135: #ifdef DIRECTED_BROADCAST
                    136: #define IPDIRECTEDBCAST        1
                    137: #else
                    138: #define        IPDIRECTEDBCAST 0
                    139: #endif /* DIRECTED_BROADCAST */
                    140: #endif /* IPDIRECTEDBCAST */
1.1       cgd       141: int    ipforwarding = IPFORWARDING;
                    142: int    ipsendredirects = IPSENDREDIRECTS;
1.13      mycroft   143: int    ip_defttl = IPDEFTTL;
1.26      thorpej   144: int    ip_forwsrcrt = IPFORWSRCRT;
1.27      thorpej   145: int    ip_directedbcast = IPDIRECTEDBCAST;
1.47      cjs       146: int    ip_allowsrcrt = IPALLOWSRCRT;
1.53      kml       147: int    ip_mtudisc = IPMTUDISC;
1.60      kml       148: u_int  ip_mtudisc_timeout = IPMTUDISCTIMEOUT;
1.1       cgd       149: #ifdef DIAGNOSTIC
                    150: int    ipprintfs = 0;
                    151: #endif
                    152:
1.60      kml       153: struct rttimer_queue *ip_mtudisc_timeout_q = NULL;
                    154:
1.1       cgd       155: extern struct domain inetdomain;
                    156: extern struct protosw inetsw[];
                    157: u_char ip_protox[IPPROTO_MAX];
                    158: int    ipqmaxlen = IFQ_MAXLEN;
1.22      mycroft   159: struct in_ifaddrhead in_ifaddr;
1.57      tls       160: struct in_ifaddrhashhead *in_ifaddrhashtbl;
1.13      mycroft   161: struct ifqueue ipintrq;
1.63      matt      162: struct ipstat  ipstat;
                    163: u_int16_t      ip_id;
                    164: int    ip_defttl;
                    165: struct ipqhead ipq;
1.1       cgd       166:
                    167: /*
                    168:  * We need to save the IP options in case a protocol wants to respond
                    169:  * to an incoming packet over the same route if the packet got here
                    170:  * using IP source routing.  This allows connection establishment and
                    171:  * maintenance when the remote end is on a network that is not known
                    172:  * to us.
                    173:  */
                    174: int    ip_nhops = 0;
                    175: static struct ip_srcrt {
                    176:        struct  in_addr dst;                    /* final destination */
                    177:        char    nop;                            /* one NOP to align */
                    178:        char    srcopt[IPOPT_OFFSET + 1];       /* OPTVAL, OLEN and OFFSET */
                    179:        struct  in_addr route[MAX_IPOPTLEN/sizeof(struct in_addr)];
                    180: } ip_srcrt;
                    181:
1.13      mycroft   182: static void save_rte __P((u_char *, struct in_addr));
1.35      mycroft   183:
1.1       cgd       184: /*
                    185:  * IP initialization: fill in IP protocol switch table.
                    186:  * All protocols not implemented in kernel go to raw IP protocol handler.
                    187:  */
1.8       mycroft   188: void
1.1       cgd       189: ip_init()
                    190: {
                    191:        register struct protosw *pr;
                    192:        register int i;
                    193:
                    194:        pr = pffindproto(PF_INET, IPPROTO_RAW, SOCK_RAW);
                    195:        if (pr == 0)
                    196:                panic("ip_init");
                    197:        for (i = 0; i < IPPROTO_MAX; i++)
                    198:                ip_protox[i] = pr - inetsw;
                    199:        for (pr = inetdomain.dom_protosw;
                    200:            pr < inetdomain.dom_protoswNPROTOSW; pr++)
                    201:                if (pr->pr_domain->dom_family == PF_INET &&
                    202:                    pr->pr_protocol && pr->pr_protocol != IPPROTO_RAW)
                    203:                        ip_protox[pr->pr_protocol] = pr - inetsw;
1.25      cgd       204:        LIST_INIT(&ipq);
1.1       cgd       205:        ip_id = time.tv_sec & 0xffff;
                    206:        ipintrq.ifq_maxlen = ipqmaxlen;
1.22      mycroft   207:        TAILQ_INIT(&in_ifaddr);
1.57      tls       208:        in_ifaddrhashtbl =
                    209:            hashinit(IN_IFADDR_HASH_SIZE, M_IFADDR, M_WAITOK, &in_ifaddrhash);
1.60      kml       210:        if (ip_mtudisc != 0)
                    211:                ip_mtudisc_timeout_q =
                    212:                    rt_timer_queue_create(ip_mtudisc_timeout);
1.1       cgd       213: }
                    214:
                    215: struct sockaddr_in ipaddr = { sizeof(ipaddr), AF_INET };
                    216: struct route ipforward_rt;
                    217:
                    218: /*
                    219:  * Ip input routine.  Checksum and byte swap header.  If fragmented
                    220:  * try to reassemble.  Process options.  Pass to next level.
                    221:  */
1.8       mycroft   222: void
1.1       cgd       223: ipintr()
                    224: {
1.33      mrg       225:        register struct ip *ip = NULL;
1.1       cgd       226:        register struct mbuf *m;
                    227:        register struct ipq *fp;
                    228:        register struct in_ifaddr *ia;
1.57      tls       229:        register struct ifaddr *ifa;
1.25      cgd       230:        struct ipqent *ipqe;
1.35      mycroft   231:        int hlen = 0, mff, len, s;
1.36      mrg       232: #ifdef PFIL_HOOKS
1.33      mrg       233:        struct packet_filter_hook *pfh;
                    234:        struct mbuf *m0;
1.43      mrg       235:        int rv;
1.36      mrg       236: #endif /* PFIL_HOOKS */
1.1       cgd       237:
                    238: next:
                    239:        /*
                    240:         * Get next datagram off input queue and get IP header
                    241:         * in first mbuf.
                    242:         */
                    243:        s = splimp();
                    244:        IF_DEQUEUE(&ipintrq, m);
                    245:        splx(s);
1.13      mycroft   246:        if (m == 0)
1.1       cgd       247:                return;
                    248: #ifdef DIAGNOSTIC
                    249:        if ((m->m_flags & M_PKTHDR) == 0)
                    250:                panic("ipintr no HDR");
                    251: #endif
                    252:        /*
                    253:         * If no IP addresses have been set yet but the interfaces
                    254:         * are receiving, can't do anything with incoming packets yet.
                    255:         */
1.22      mycroft   256:        if (in_ifaddr.tqh_first == 0)
1.1       cgd       257:                goto bad;
                    258:        ipstat.ips_total++;
                    259:        if (m->m_len < sizeof (struct ip) &&
                    260:            (m = m_pullup(m, sizeof (struct ip))) == 0) {
                    261:                ipstat.ips_toosmall++;
                    262:                goto next;
                    263:        }
                    264:        ip = mtod(m, struct ip *);
1.13      mycroft   265:        if (ip->ip_v != IPVERSION) {
                    266:                ipstat.ips_badvers++;
                    267:                goto bad;
                    268:        }
1.1       cgd       269:        hlen = ip->ip_hl << 2;
                    270:        if (hlen < sizeof(struct ip)) { /* minimum header length */
                    271:                ipstat.ips_badhlen++;
                    272:                goto bad;
                    273:        }
                    274:        if (hlen > m->m_len) {
                    275:                if ((m = m_pullup(m, hlen)) == 0) {
                    276:                        ipstat.ips_badhlen++;
                    277:                        goto next;
                    278:                }
                    279:                ip = mtod(m, struct ip *);
                    280:        }
1.28      christos  281:        if ((ip->ip_sum = in_cksum(m, hlen)) != 0) {
1.1       cgd       282:                ipstat.ips_badsum++;
                    283:                goto bad;
                    284:        }
                    285:
                    286:        /*
                    287:         * Convert fields to host representation.
                    288:         */
                    289:        NTOHS(ip->ip_len);
                    290:        NTOHS(ip->ip_id);
                    291:        NTOHS(ip->ip_off);
1.35      mycroft   292:        len = ip->ip_len;
1.1       cgd       293:
                    294:        /*
                    295:         * Check that the amount of data in the buffers
                    296:         * is as at least much as the IP header would have us expect.
                    297:         * Trim mbufs if longer than we expect.
                    298:         * Drop packet if shorter than we expect.
                    299:         */
1.35      mycroft   300:        if (m->m_pkthdr.len < len) {
1.1       cgd       301:                ipstat.ips_tooshort++;
                    302:                goto bad;
                    303:        }
1.35      mycroft   304:        if (m->m_pkthdr.len > len) {
1.1       cgd       305:                if (m->m_len == m->m_pkthdr.len) {
1.35      mycroft   306:                        m->m_len = len;
                    307:                        m->m_pkthdr.len = len;
1.1       cgd       308:                } else
1.35      mycroft   309:                        m_adj(m, len - m->m_pkthdr.len);
1.1       cgd       310:        }
                    311:
1.64      thorpej   312:        /*
                    313:         * Assume that we can create a fast-forward IP flow entry
                    314:         * based on this packet.
                    315:         */
                    316:        m->m_flags |= M_CANFASTFWD;
                    317:
1.36      mrg       318: #ifdef PFIL_HOOKS
1.33      mrg       319:        /*
1.64      thorpej   320:         * Run through list of hooks for input packets.  If there are any
                    321:         * filters which require that additional packets in the flow are
                    322:         * not fast-forwarded, they must clear the M_CANFASTFWD flag.
                    323:         * Note that filters must _never_ set this flag, as another filter
                    324:         * in the list may have previously cleared it.
1.33      mrg       325:         */
                    326:        m0 = m;
1.59      mrg       327:        for (pfh = pfil_hook_get(PFIL_IN); pfh; pfh = pfh->pfil_link.tqe_next)
1.33      mrg       328:                if (pfh->pfil_func) {
1.43      mrg       329:                        rv = pfh->pfil_func(ip, hlen, m->m_pkthdr.rcvif, 0, &m0);
                    330:                        if (rv)
1.40      veego     331:                                goto next;
1.68    ! sommerfe  332:                        m = m0;
        !           333:                        if (m == NULL)
        !           334:                                goto next;
        !           335:                        ip = mtod(m, struct ip *);
1.33      mrg       336:                }
1.36      mrg       337: #endif /* PFIL_HOOKS */
1.33      mrg       338:
1.1       cgd       339:        /*
                    340:         * Process options and, if not destined for us,
                    341:         * ship it on.  ip_dooptions returns 1 when an
                    342:         * error was detected (causing an icmp message
                    343:         * to be sent and the original packet to be freed).
                    344:         */
                    345:        ip_nhops = 0;           /* for source routed packets */
                    346:        if (hlen > sizeof (struct ip) && ip_dooptions(m))
                    347:                goto next;
                    348:
                    349:        /*
                    350:         * Check our list of addresses, to see if the packet is for us.
                    351:         */
1.57      tls       352:        INADDR_TO_IA(ip->ip_dst, ia);
                    353:        if (ia != NULL) goto ours;
                    354:        if (m->m_pkthdr.rcvif->if_flags & IFF_BROADCAST) {
                    355:                for (ifa = m->m_pkthdr.rcvif->if_addrlist.tqh_first;
                    356:                    ifa != NULL; ifa = ifa->ifa_list.tqe_next) {
                    357:                        if (ifa->ifa_addr->sa_family != AF_INET) continue;
                    358:                        ia = ifatoia(ifa);
1.35      mycroft   359:                        if (in_hosteq(ip->ip_dst, ia->ia_broadaddr.sin_addr) ||
                    360:                            in_hosteq(ip->ip_dst, ia->ia_netbroadcast) ||
1.20      mycroft   361:                            /*
                    362:                             * Look for all-0's host part (old broadcast addr),
                    363:                             * either for subnet or net.
                    364:                             */
                    365:                            ip->ip_dst.s_addr == ia->ia_subnet ||
1.18      mycroft   366:                            ip->ip_dst.s_addr == ia->ia_net)
1.1       cgd       367:                                goto ours;
1.57      tls       368:                        /*
                    369:                         * An interface with IP address zero accepts
                    370:                         * all packets that arrive on that interface.
                    371:                         */
                    372:                        if (in_nullhost(ia->ia_addr.sin_addr))
                    373:                                goto ours;
1.1       cgd       374:                }
                    375:        }
1.18      mycroft   376:        if (IN_MULTICAST(ip->ip_dst.s_addr)) {
1.4       hpeyerl   377:                struct in_multi *inm;
                    378: #ifdef MROUTING
                    379:                extern struct socket *ip_mrouter;
1.10      brezak    380:
                    381:                if (m->m_flags & M_EXT) {
                    382:                        if ((m = m_pullup(m, hlen)) == 0) {
                    383:                                ipstat.ips_toosmall++;
                    384:                                goto next;
                    385:                        }
                    386:                        ip = mtod(m, struct ip *);
                    387:                }
1.4       hpeyerl   388:
                    389:                if (ip_mrouter) {
                    390:                        /*
                    391:                         * If we are acting as a multicast router, all
                    392:                         * incoming multicast packets are passed to the
                    393:                         * kernel-level multicast forwarding function.
                    394:                         * The packet is returned (relatively) intact; if
                    395:                         * ip_mforward() returns a non-zero value, the packet
                    396:                         * must be discarded, else it may be accepted below.
                    397:                         *
                    398:                         * (The IP ident field is put in the same byte order
                    399:                         * as expected when ip_mforward() is called from
                    400:                         * ip_output().)
                    401:                         */
                    402:                        ip->ip_id = htons(ip->ip_id);
1.13      mycroft   403:                        if (ip_mforward(m, m->m_pkthdr.rcvif) != 0) {
                    404:                                ipstat.ips_cantforward++;
1.4       hpeyerl   405:                                m_freem(m);
                    406:                                goto next;
                    407:                        }
                    408:                        ip->ip_id = ntohs(ip->ip_id);
                    409:
                    410:                        /*
                    411:                         * The process-level routing demon needs to receive
                    412:                         * all multicast IGMP packets, whether or not this
                    413:                         * host belongs to their destination groups.
                    414:                         */
                    415:                        if (ip->ip_p == IPPROTO_IGMP)
                    416:                                goto ours;
1.13      mycroft   417:                        ipstat.ips_forward++;
1.4       hpeyerl   418:                }
                    419: #endif
                    420:                /*
                    421:                 * See if we belong to the destination multicast group on the
                    422:                 * arrival interface.
                    423:                 */
                    424:                IN_LOOKUP_MULTI(ip->ip_dst, m->m_pkthdr.rcvif, inm);
                    425:                if (inm == NULL) {
1.13      mycroft   426:                        ipstat.ips_cantforward++;
1.4       hpeyerl   427:                        m_freem(m);
                    428:                        goto next;
                    429:                }
                    430:                goto ours;
                    431:        }
1.19      mycroft   432:        if (ip->ip_dst.s_addr == INADDR_BROADCAST ||
1.35      mycroft   433:            in_nullhost(ip->ip_dst))
1.1       cgd       434:                goto ours;
                    435:
                    436:        /*
                    437:         * Not for us; forward if possible and desirable.
                    438:         */
                    439:        if (ipforwarding == 0) {
                    440:                ipstat.ips_cantforward++;
                    441:                m_freem(m);
                    442:        } else
                    443:                ip_forward(m, 0);
                    444:        goto next;
                    445:
                    446: ours:
                    447:        /*
                    448:         * If offset or IP_MF are set, must reassemble.
                    449:         * Otherwise, nothing need be done.
                    450:         * (We could look in the reassembly queue to see
                    451:         * if the packet was previously fragmented,
                    452:         * but it's not worth the time; just let them time out.)
                    453:         */
1.37      perry     454:        if (ip->ip_off & ~(IP_DF|IP_RF)) {
1.1       cgd       455:                /*
                    456:                 * Look for queue of fragments
                    457:                 * of this datagram.
                    458:                 */
1.25      cgd       459:                for (fp = ipq.lh_first; fp != NULL; fp = fp->ipq_q.le_next)
1.1       cgd       460:                        if (ip->ip_id == fp->ipq_id &&
1.35      mycroft   461:                            in_hosteq(ip->ip_src, fp->ipq_src) &&
                    462:                            in_hosteq(ip->ip_dst, fp->ipq_dst) &&
1.1       cgd       463:                            ip->ip_p == fp->ipq_p)
                    464:                                goto found;
                    465:                fp = 0;
                    466: found:
                    467:
                    468:                /*
                    469:                 * Adjust ip_len to not reflect header,
1.25      cgd       470:                 * set ipqe_mff if more fragments are expected,
1.1       cgd       471:                 * convert offset of this to bytes.
                    472:                 */
                    473:                ip->ip_len -= hlen;
1.25      cgd       474:                mff = (ip->ip_off & IP_MF) != 0;
                    475:                if (mff) {
1.16      cgd       476:                        /*
                    477:                         * Make sure that fragments have a data length
                    478:                         * that's a non-zero multiple of 8 bytes.
                    479:                         */
1.17      cgd       480:                        if (ip->ip_len == 0 || (ip->ip_len & 0x7) != 0) {
1.16      cgd       481:                                ipstat.ips_badfrags++;
                    482:                                goto bad;
                    483:                        }
                    484:                }
1.1       cgd       485:                ip->ip_off <<= 3;
                    486:
                    487:                /*
                    488:                 * If datagram marked as having more fragments
                    489:                 * or if this is not the first fragment,
                    490:                 * attempt reassembly; if it succeeds, proceed.
                    491:                 */
1.25      cgd       492:                if (mff || ip->ip_off) {
1.1       cgd       493:                        ipstat.ips_fragments++;
1.25      cgd       494:                        MALLOC(ipqe, struct ipqent *, sizeof (struct ipqent),
                    495:                            M_IPQ, M_NOWAIT);
                    496:                        if (ipqe == NULL) {
                    497:                                ipstat.ips_rcvmemdrop++;
                    498:                                goto bad;
                    499:                        }
                    500:                        ipqe->ipqe_mff = mff;
1.50      thorpej   501:                        ipqe->ipqe_m = m;
1.25      cgd       502:                        ipqe->ipqe_ip = ip;
1.50      thorpej   503:                        m = ip_reass(ipqe, fp);
                    504:                        if (m == 0)
1.1       cgd       505:                                goto next;
1.13      mycroft   506:                        ipstat.ips_reassembled++;
1.50      thorpej   507:                        ip = mtod(m, struct ip *);
1.1       cgd       508:                } else
                    509:                        if (fp)
                    510:                                ip_freef(fp);
                    511:        } else
                    512:                ip->ip_len -= hlen;
                    513:
                    514:        /*
                    515:         * Switch out to protocol's input routine.
                    516:         */
                    517:        ipstat.ips_delivered++;
                    518:        (*inetsw[ip_protox[ip->ip_p]].pr_input)(m, hlen);
                    519:        goto next;
                    520: bad:
                    521:        m_freem(m);
                    522:        goto next;
                    523: }
                    524:
                    525: /*
                    526:  * Take incoming datagram fragment and try to
                    527:  * reassemble it into whole datagram.  If a chain for
                    528:  * reassembly of this datagram already exists, then it
                    529:  * is given as fp; otherwise have to make a chain.
                    530:  */
1.50      thorpej   531: struct mbuf *
1.25      cgd       532: ip_reass(ipqe, fp)
                    533:        register struct ipqent *ipqe;
1.1       cgd       534:        register struct ipq *fp;
                    535: {
1.50      thorpej   536:        register struct mbuf *m = ipqe->ipqe_m;
1.25      cgd       537:        register struct ipqent *nq, *p, *q;
                    538:        struct ip *ip;
1.1       cgd       539:        struct mbuf *t;
1.25      cgd       540:        int hlen = ipqe->ipqe_ip->ip_hl << 2;
1.1       cgd       541:        int i, next;
                    542:
                    543:        /*
                    544:         * Presence of header sizes in mbufs
                    545:         * would confuse code below.
                    546:         */
                    547:        m->m_data += hlen;
                    548:        m->m_len -= hlen;
                    549:
                    550:        /*
                    551:         * If first fragment to arrive, create a reassembly queue.
                    552:         */
                    553:        if (fp == 0) {
1.50      thorpej   554:                MALLOC(fp, struct ipq *, sizeof (struct ipq),
                    555:                    M_FTABLE, M_NOWAIT);
                    556:                if (fp == NULL)
1.1       cgd       557:                        goto dropfrag;
1.25      cgd       558:                LIST_INSERT_HEAD(&ipq, fp, ipq_q);
1.1       cgd       559:                fp->ipq_ttl = IPFRAGTTL;
1.25      cgd       560:                fp->ipq_p = ipqe->ipqe_ip->ip_p;
                    561:                fp->ipq_id = ipqe->ipqe_ip->ip_id;
                    562:                LIST_INIT(&fp->ipq_fragq);
                    563:                fp->ipq_src = ipqe->ipqe_ip->ip_src;
                    564:                fp->ipq_dst = ipqe->ipqe_ip->ip_dst;
                    565:                p = NULL;
1.1       cgd       566:                goto insert;
                    567:        }
                    568:
                    569:        /*
                    570:         * Find a segment which begins after this one does.
                    571:         */
1.25      cgd       572:        for (p = NULL, q = fp->ipq_fragq.lh_first; q != NULL;
                    573:            p = q, q = q->ipqe_q.le_next)
                    574:                if (q->ipqe_ip->ip_off > ipqe->ipqe_ip->ip_off)
1.1       cgd       575:                        break;
                    576:
                    577:        /*
                    578:         * If there is a preceding segment, it may provide some of
                    579:         * our data already.  If so, drop the data from the incoming
                    580:         * segment.  If it provides all of our data, drop us.
                    581:         */
1.25      cgd       582:        if (p != NULL) {
                    583:                i = p->ipqe_ip->ip_off + p->ipqe_ip->ip_len -
                    584:                    ipqe->ipqe_ip->ip_off;
1.1       cgd       585:                if (i > 0) {
1.25      cgd       586:                        if (i >= ipqe->ipqe_ip->ip_len)
1.1       cgd       587:                                goto dropfrag;
1.50      thorpej   588:                        m_adj(ipqe->ipqe_m, i);
1.25      cgd       589:                        ipqe->ipqe_ip->ip_off += i;
                    590:                        ipqe->ipqe_ip->ip_len -= i;
1.1       cgd       591:                }
                    592:        }
                    593:
                    594:        /*
                    595:         * While we overlap succeeding segments trim them or,
                    596:         * if they are completely covered, dequeue them.
                    597:         */
1.25      cgd       598:        for (; q != NULL && ipqe->ipqe_ip->ip_off + ipqe->ipqe_ip->ip_len >
                    599:            q->ipqe_ip->ip_off; q = nq) {
                    600:                i = (ipqe->ipqe_ip->ip_off + ipqe->ipqe_ip->ip_len) -
                    601:                    q->ipqe_ip->ip_off;
                    602:                if (i < q->ipqe_ip->ip_len) {
                    603:                        q->ipqe_ip->ip_len -= i;
                    604:                        q->ipqe_ip->ip_off += i;
1.50      thorpej   605:                        m_adj(q->ipqe_m, i);
1.1       cgd       606:                        break;
                    607:                }
1.25      cgd       608:                nq = q->ipqe_q.le_next;
1.50      thorpej   609:                m_freem(q->ipqe_m);
1.25      cgd       610:                LIST_REMOVE(q, ipqe_q);
                    611:                FREE(q, M_IPQ);
1.1       cgd       612:        }
                    613:
                    614: insert:
                    615:        /*
                    616:         * Stick new segment in its place;
                    617:         * check for complete reassembly.
                    618:         */
1.25      cgd       619:        if (p == NULL) {
                    620:                LIST_INSERT_HEAD(&fp->ipq_fragq, ipqe, ipqe_q);
                    621:        } else {
                    622:                LIST_INSERT_AFTER(p, ipqe, ipqe_q);
                    623:        }
1.1       cgd       624:        next = 0;
1.25      cgd       625:        for (p = NULL, q = fp->ipq_fragq.lh_first; q != NULL;
                    626:            p = q, q = q->ipqe_q.le_next) {
                    627:                if (q->ipqe_ip->ip_off != next)
1.1       cgd       628:                        return (0);
1.25      cgd       629:                next += q->ipqe_ip->ip_len;
1.1       cgd       630:        }
1.25      cgd       631:        if (p->ipqe_mff)
1.1       cgd       632:                return (0);
                    633:
                    634:        /*
1.41      thorpej   635:         * Reassembly is complete.  Check for a bogus message size and
                    636:         * concatenate fragments.
1.1       cgd       637:         */
1.25      cgd       638:        q = fp->ipq_fragq.lh_first;
                    639:        ip = q->ipqe_ip;
1.41      thorpej   640:        if ((next + (ip->ip_hl << 2)) > IP_MAXPACKET) {
                    641:                ipstat.ips_toolong++;
                    642:                ip_freef(fp);
                    643:                return (0);
                    644:        }
1.50      thorpej   645:        m = q->ipqe_m;
1.1       cgd       646:        t = m->m_next;
                    647:        m->m_next = 0;
                    648:        m_cat(m, t);
1.25      cgd       649:        nq = q->ipqe_q.le_next;
                    650:        FREE(q, M_IPQ);
                    651:        for (q = nq; q != NULL; q = nq) {
1.50      thorpej   652:                t = q->ipqe_m;
1.25      cgd       653:                nq = q->ipqe_q.le_next;
                    654:                FREE(q, M_IPQ);
1.1       cgd       655:                m_cat(m, t);
                    656:        }
                    657:
                    658:        /*
                    659:         * Create header for new ip packet by
                    660:         * modifying header of first packet;
                    661:         * dequeue and discard fragment reassembly header.
                    662:         * Make header visible.
                    663:         */
                    664:        ip->ip_len = next;
1.25      cgd       665:        ip->ip_src = fp->ipq_src;
                    666:        ip->ip_dst = fp->ipq_dst;
                    667:        LIST_REMOVE(fp, ipq_q);
1.50      thorpej   668:        FREE(fp, M_FTABLE);
1.1       cgd       669:        m->m_len += (ip->ip_hl << 2);
                    670:        m->m_data -= (ip->ip_hl << 2);
                    671:        /* some debugging cruft by sklower, below, will go away soon */
                    672:        if (m->m_flags & M_PKTHDR) { /* XXX this should be done elsewhere */
                    673:                register int plen = 0;
1.50      thorpej   674:                for (t = m; t; t = t->m_next)
                    675:                        plen += t->m_len;
                    676:                m->m_pkthdr.len = plen;
1.1       cgd       677:        }
1.50      thorpej   678:        return (m);
1.1       cgd       679:
                    680: dropfrag:
                    681:        ipstat.ips_fragdropped++;
                    682:        m_freem(m);
1.25      cgd       683:        FREE(ipqe, M_IPQ);
1.1       cgd       684:        return (0);
                    685: }
                    686:
                    687: /*
                    688:  * Free a fragment reassembly header and all
                    689:  * associated datagrams.
                    690:  */
1.8       mycroft   691: void
1.1       cgd       692: ip_freef(fp)
                    693:        struct ipq *fp;
                    694: {
1.25      cgd       695:        register struct ipqent *q, *p;
1.1       cgd       696:
1.25      cgd       697:        for (q = fp->ipq_fragq.lh_first; q != NULL; q = p) {
                    698:                p = q->ipqe_q.le_next;
1.50      thorpej   699:                m_freem(q->ipqe_m);
1.25      cgd       700:                LIST_REMOVE(q, ipqe_q);
                    701:                FREE(q, M_IPQ);
1.1       cgd       702:        }
1.25      cgd       703:        LIST_REMOVE(fp, ipq_q);
1.50      thorpej   704:        FREE(fp, M_FTABLE);
1.1       cgd       705: }
                    706:
                    707: /*
                    708:  * IP timer processing;
                    709:  * if a timer expires on a reassembly
                    710:  * queue, discard it.
                    711:  */
1.8       mycroft   712: void
1.1       cgd       713: ip_slowtimo()
                    714: {
1.25      cgd       715:        register struct ipq *fp, *nfp;
1.24      mycroft   716:        int s = splsoftnet();
1.1       cgd       717:
1.25      cgd       718:        for (fp = ipq.lh_first; fp != NULL; fp = nfp) {
                    719:                nfp = fp->ipq_q.le_next;
                    720:                if (--fp->ipq_ttl == 0) {
1.1       cgd       721:                        ipstat.ips_fragtimeout++;
1.25      cgd       722:                        ip_freef(fp);
1.1       cgd       723:                }
                    724:        }
1.63      matt      725: #ifdef GATEWAY
                    726:        ipflow_slowtimo();
                    727: #endif
1.1       cgd       728:        splx(s);
                    729: }
                    730:
                    731: /*
                    732:  * Drain off all datagram fragments.
                    733:  */
1.8       mycroft   734: void
1.1       cgd       735: ip_drain()
                    736: {
                    737:
1.25      cgd       738:        while (ipq.lh_first != NULL) {
1.1       cgd       739:                ipstat.ips_fragdropped++;
1.25      cgd       740:                ip_freef(ipq.lh_first);
1.1       cgd       741:        }
                    742: }
                    743:
                    744: /*
                    745:  * Do option processing on a datagram,
                    746:  * possibly discarding it if bad options are encountered,
                    747:  * or forwarding it if source-routed.
                    748:  * Returns 1 if packet has been forwarded/freed,
                    749:  * 0 if the packet should be processed further.
                    750:  */
1.8       mycroft   751: int
1.1       cgd       752: ip_dooptions(m)
                    753:        struct mbuf *m;
                    754: {
                    755:        register struct ip *ip = mtod(m, struct ip *);
                    756:        register u_char *cp;
                    757:        register struct ip_timestamp *ipt;
                    758:        register struct in_ifaddr *ia;
                    759:        int opt, optlen, cnt, off, code, type = ICMP_PARAMPROB, forward = 0;
1.13      mycroft   760:        struct in_addr *sin, dst;
1.1       cgd       761:        n_time ntime;
                    762:
1.13      mycroft   763:        dst = ip->ip_dst;
1.1       cgd       764:        cp = (u_char *)(ip + 1);
                    765:        cnt = (ip->ip_hl << 2) - sizeof (struct ip);
                    766:        for (; cnt > 0; cnt -= optlen, cp += optlen) {
                    767:                opt = cp[IPOPT_OPTVAL];
                    768:                if (opt == IPOPT_EOL)
                    769:                        break;
                    770:                if (opt == IPOPT_NOP)
                    771:                        optlen = 1;
                    772:                else {
                    773:                        optlen = cp[IPOPT_OLEN];
                    774:                        if (optlen <= 0 || optlen > cnt) {
                    775:                                code = &cp[IPOPT_OLEN] - (u_char *)ip;
                    776:                                goto bad;
                    777:                        }
                    778:                }
                    779:                switch (opt) {
                    780:
                    781:                default:
                    782:                        break;
                    783:
                    784:                /*
                    785:                 * Source routing with record.
                    786:                 * Find interface with current destination address.
                    787:                 * If none on this machine then drop if strictly routed,
                    788:                 * or do nothing if loosely routed.
                    789:                 * Record interface address and bring up next address
                    790:                 * component.  If strictly routed make sure next
                    791:                 * address is on directly accessible net.
                    792:                 */
                    793:                case IPOPT_LSRR:
                    794:                case IPOPT_SSRR:
1.47      cjs       795:                        if (ip_allowsrcrt == 0) {
                    796:                                type = ICMP_UNREACH;
                    797:                                code = ICMP_UNREACH_NET_PROHIB;
                    798:                                goto bad;
                    799:                        }
1.1       cgd       800:                        if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) {
                    801:                                code = &cp[IPOPT_OFFSET] - (u_char *)ip;
                    802:                                goto bad;
                    803:                        }
                    804:                        ipaddr.sin_addr = ip->ip_dst;
1.19      mycroft   805:                        ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr)));
1.1       cgd       806:                        if (ia == 0) {
                    807:                                if (opt == IPOPT_SSRR) {
                    808:                                        type = ICMP_UNREACH;
                    809:                                        code = ICMP_UNREACH_SRCFAIL;
                    810:                                        goto bad;
                    811:                                }
                    812:                                /*
                    813:                                 * Loose routing, and not at next destination
                    814:                                 * yet; nothing to do except forward.
                    815:                                 */
                    816:                                break;
                    817:                        }
                    818:                        off--;                  /* 0 origin */
                    819:                        if (off > optlen - sizeof(struct in_addr)) {
                    820:                                /*
                    821:                                 * End of source route.  Should be for us.
                    822:                                 */
                    823:                                save_rte(cp, ip->ip_src);
                    824:                                break;
                    825:                        }
                    826:                        /*
                    827:                         * locate outgoing interface
                    828:                         */
                    829:                        bcopy((caddr_t)(cp + off), (caddr_t)&ipaddr.sin_addr,
                    830:                            sizeof(ipaddr.sin_addr));
                    831:                        if (opt == IPOPT_SSRR) {
                    832: #define        INA     struct in_ifaddr *
                    833: #define        SA      struct sockaddr *
1.29      mrg       834:                            ia = (INA)ifa_ifwithladdr((SA)&ipaddr);
1.1       cgd       835:                        } else
                    836:                                ia = ip_rtaddr(ipaddr.sin_addr);
                    837:                        if (ia == 0) {
                    838:                                type = ICMP_UNREACH;
                    839:                                code = ICMP_UNREACH_SRCFAIL;
                    840:                                goto bad;
                    841:                        }
                    842:                        ip->ip_dst = ipaddr.sin_addr;
1.20      mycroft   843:                        bcopy((caddr_t)&ia->ia_addr.sin_addr,
1.1       cgd       844:                            (caddr_t)(cp + off), sizeof(struct in_addr));
                    845:                        cp[IPOPT_OFFSET] += sizeof(struct in_addr);
1.13      mycroft   846:                        /*
                    847:                         * Let ip_intr's mcast routing check handle mcast pkts
                    848:                         */
1.18      mycroft   849:                        forward = !IN_MULTICAST(ip->ip_dst.s_addr);
1.1       cgd       850:                        break;
                    851:
                    852:                case IPOPT_RR:
                    853:                        if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) {
                    854:                                code = &cp[IPOPT_OFFSET] - (u_char *)ip;
                    855:                                goto bad;
                    856:                        }
                    857:                        /*
                    858:                         * If no space remains, ignore.
                    859:                         */
                    860:                        off--;                  /* 0 origin */
                    861:                        if (off > optlen - sizeof(struct in_addr))
                    862:                                break;
                    863:                        bcopy((caddr_t)(&ip->ip_dst), (caddr_t)&ipaddr.sin_addr,
                    864:                            sizeof(ipaddr.sin_addr));
                    865:                        /*
                    866:                         * locate outgoing interface; if we're the destination,
                    867:                         * use the incoming interface (should be same).
                    868:                         */
                    869:                        if ((ia = (INA)ifa_ifwithaddr((SA)&ipaddr)) == 0 &&
                    870:                            (ia = ip_rtaddr(ipaddr.sin_addr)) == 0) {
                    871:                                type = ICMP_UNREACH;
                    872:                                code = ICMP_UNREACH_HOST;
                    873:                                goto bad;
                    874:                        }
1.20      mycroft   875:                        bcopy((caddr_t)&ia->ia_addr.sin_addr,
1.1       cgd       876:                            (caddr_t)(cp + off), sizeof(struct in_addr));
                    877:                        cp[IPOPT_OFFSET] += sizeof(struct in_addr);
                    878:                        break;
                    879:
                    880:                case IPOPT_TS:
                    881:                        code = cp - (u_char *)ip;
                    882:                        ipt = (struct ip_timestamp *)cp;
                    883:                        if (ipt->ipt_len < 5)
                    884:                                goto bad;
1.15      cgd       885:                        if (ipt->ipt_ptr > ipt->ipt_len - sizeof (int32_t)) {
1.1       cgd       886:                                if (++ipt->ipt_oflw == 0)
                    887:                                        goto bad;
                    888:                                break;
                    889:                        }
                    890:                        sin = (struct in_addr *)(cp + ipt->ipt_ptr - 1);
                    891:                        switch (ipt->ipt_flg) {
                    892:
                    893:                        case IPOPT_TS_TSONLY:
                    894:                                break;
                    895:
                    896:                        case IPOPT_TS_TSANDADDR:
1.66      thorpej   897:                                if (ipt->ipt_ptr - 1 + sizeof(n_time) +
1.1       cgd       898:                                    sizeof(struct in_addr) > ipt->ipt_len)
                    899:                                        goto bad;
1.13      mycroft   900:                                ipaddr.sin_addr = dst;
                    901:                                ia = (INA)ifaof_ifpforaddr((SA)&ipaddr,
                    902:                                                            m->m_pkthdr.rcvif);
                    903:                                if (ia == 0)
                    904:                                        continue;
1.20      mycroft   905:                                bcopy((caddr_t)&ia->ia_addr.sin_addr,
1.1       cgd       906:                                    (caddr_t)sin, sizeof(struct in_addr));
                    907:                                ipt->ipt_ptr += sizeof(struct in_addr);
                    908:                                break;
                    909:
                    910:                        case IPOPT_TS_PRESPEC:
1.66      thorpej   911:                                if (ipt->ipt_ptr - 1 + sizeof(n_time) +
1.1       cgd       912:                                    sizeof(struct in_addr) > ipt->ipt_len)
                    913:                                        goto bad;
                    914:                                bcopy((caddr_t)sin, (caddr_t)&ipaddr.sin_addr,
                    915:                                    sizeof(struct in_addr));
                    916:                                if (ifa_ifwithaddr((SA)&ipaddr) == 0)
                    917:                                        continue;
                    918:                                ipt->ipt_ptr += sizeof(struct in_addr);
                    919:                                break;
                    920:
                    921:                        default:
                    922:                                goto bad;
                    923:                        }
                    924:                        ntime = iptime();
                    925:                        bcopy((caddr_t)&ntime, (caddr_t)cp + ipt->ipt_ptr - 1,
                    926:                            sizeof(n_time));
                    927:                        ipt->ipt_ptr += sizeof(n_time);
                    928:                }
                    929:        }
                    930:        if (forward) {
1.26      thorpej   931:                if (ip_forwsrcrt == 0) {
                    932:                        type = ICMP_UNREACH;
                    933:                        code = ICMP_UNREACH_SRCFAIL;
                    934:                        goto bad;
                    935:                }
1.1       cgd       936:                ip_forward(m, 1);
                    937:                return (1);
1.13      mycroft   938:        }
                    939:        return (0);
1.1       cgd       940: bad:
1.13      mycroft   941:        ip->ip_len -= ip->ip_hl << 2;   /* XXX icmp_error adds in hdr length */
                    942:        icmp_error(m, type, code, 0, 0);
                    943:        ipstat.ips_badoptions++;
1.1       cgd       944:        return (1);
                    945: }
                    946:
                    947: /*
                    948:  * Given address of next destination (final or next hop),
                    949:  * return internet address info of interface to be used to get there.
                    950:  */
                    951: struct in_ifaddr *
                    952: ip_rtaddr(dst)
                    953:         struct in_addr dst;
                    954: {
                    955:        register struct sockaddr_in *sin;
                    956:
1.19      mycroft   957:        sin = satosin(&ipforward_rt.ro_dst);
1.1       cgd       958:
1.35      mycroft   959:        if (ipforward_rt.ro_rt == 0 || !in_hosteq(dst, sin->sin_addr)) {
1.1       cgd       960:                if (ipforward_rt.ro_rt) {
                    961:                        RTFREE(ipforward_rt.ro_rt);
                    962:                        ipforward_rt.ro_rt = 0;
                    963:                }
                    964:                sin->sin_family = AF_INET;
                    965:                sin->sin_len = sizeof(*sin);
                    966:                sin->sin_addr = dst;
                    967:
                    968:                rtalloc(&ipforward_rt);
                    969:        }
                    970:        if (ipforward_rt.ro_rt == 0)
                    971:                return ((struct in_ifaddr *)0);
1.19      mycroft   972:        return (ifatoia(ipforward_rt.ro_rt->rt_ifa));
1.1       cgd       973: }
                    974:
                    975: /*
                    976:  * Save incoming source route for use in replies,
                    977:  * to be picked up later by ip_srcroute if the receiver is interested.
                    978:  */
1.13      mycroft   979: void
1.1       cgd       980: save_rte(option, dst)
                    981:        u_char *option;
                    982:        struct in_addr dst;
                    983: {
                    984:        unsigned olen;
                    985:
                    986:        olen = option[IPOPT_OLEN];
                    987: #ifdef DIAGNOSTIC
                    988:        if (ipprintfs)
1.39      christos  989:                printf("save_rte: olen %d\n", olen);
1.1       cgd       990: #endif
                    991:        if (olen > sizeof(ip_srcrt) - (1 + sizeof(dst)))
                    992:                return;
                    993:        bcopy((caddr_t)option, (caddr_t)ip_srcrt.srcopt, olen);
                    994:        ip_nhops = (olen - IPOPT_OFFSET - 1) / sizeof(struct in_addr);
                    995:        ip_srcrt.dst = dst;
                    996: }
                    997:
                    998: /*
                    999:  * Retrieve incoming source route for use in replies,
                   1000:  * in the same form used by setsockopt.
                   1001:  * The first hop is placed before the options, will be removed later.
                   1002:  */
                   1003: struct mbuf *
                   1004: ip_srcroute()
                   1005: {
                   1006:        register struct in_addr *p, *q;
                   1007:        register struct mbuf *m;
                   1008:
                   1009:        if (ip_nhops == 0)
                   1010:                return ((struct mbuf *)0);
                   1011:        m = m_get(M_DONTWAIT, MT_SOOPTS);
                   1012:        if (m == 0)
                   1013:                return ((struct mbuf *)0);
                   1014:
1.13      mycroft  1015: #define OPTSIZ (sizeof(ip_srcrt.nop) + sizeof(ip_srcrt.srcopt))
1.1       cgd      1016:
                   1017:        /* length is (nhops+1)*sizeof(addr) + sizeof(nop + srcrt header) */
                   1018:        m->m_len = ip_nhops * sizeof(struct in_addr) + sizeof(struct in_addr) +
                   1019:            OPTSIZ;
                   1020: #ifdef DIAGNOSTIC
                   1021:        if (ipprintfs)
1.39      christos 1022:                printf("ip_srcroute: nhops %d mlen %d", ip_nhops, m->m_len);
1.1       cgd      1023: #endif
                   1024:
                   1025:        /*
                   1026:         * First save first hop for return route
                   1027:         */
                   1028:        p = &ip_srcrt.route[ip_nhops - 1];
                   1029:        *(mtod(m, struct in_addr *)) = *p--;
                   1030: #ifdef DIAGNOSTIC
                   1031:        if (ipprintfs)
1.39      christos 1032:                printf(" hops %x", ntohl(mtod(m, struct in_addr *)->s_addr));
1.1       cgd      1033: #endif
                   1034:
                   1035:        /*
                   1036:         * Copy option fields and padding (nop) to mbuf.
                   1037:         */
                   1038:        ip_srcrt.nop = IPOPT_NOP;
                   1039:        ip_srcrt.srcopt[IPOPT_OFFSET] = IPOPT_MINOFF;
                   1040:        bcopy((caddr_t)&ip_srcrt.nop,
                   1041:            mtod(m, caddr_t) + sizeof(struct in_addr), OPTSIZ);
                   1042:        q = (struct in_addr *)(mtod(m, caddr_t) +
                   1043:            sizeof(struct in_addr) + OPTSIZ);
                   1044: #undef OPTSIZ
                   1045:        /*
                   1046:         * Record return path as an IP source route,
                   1047:         * reversing the path (pointers are now aligned).
                   1048:         */
                   1049:        while (p >= ip_srcrt.route) {
                   1050: #ifdef DIAGNOSTIC
                   1051:                if (ipprintfs)
1.39      christos 1052:                        printf(" %x", ntohl(q->s_addr));
1.1       cgd      1053: #endif
                   1054:                *q++ = *p--;
                   1055:        }
                   1056:        /*
                   1057:         * Last hop goes to final destination.
                   1058:         */
                   1059:        *q = ip_srcrt.dst;
                   1060: #ifdef DIAGNOSTIC
                   1061:        if (ipprintfs)
1.39      christos 1062:                printf(" %x\n", ntohl(q->s_addr));
1.1       cgd      1063: #endif
                   1064:        return (m);
                   1065: }
                   1066:
                   1067: /*
                   1068:  * Strip out IP options, at higher
                   1069:  * level protocol in the kernel.
                   1070:  * Second argument is buffer to which options
                   1071:  * will be moved, and return value is their length.
                   1072:  * XXX should be deleted; last arg currently ignored.
                   1073:  */
1.8       mycroft  1074: void
1.1       cgd      1075: ip_stripoptions(m, mopt)
                   1076:        register struct mbuf *m;
                   1077:        struct mbuf *mopt;
                   1078: {
                   1079:        register int i;
                   1080:        struct ip *ip = mtod(m, struct ip *);
                   1081:        register caddr_t opts;
                   1082:        int olen;
                   1083:
                   1084:        olen = (ip->ip_hl<<2) - sizeof (struct ip);
                   1085:        opts = (caddr_t)(ip + 1);
                   1086:        i = m->m_len - (sizeof (struct ip) + olen);
                   1087:        bcopy(opts  + olen, opts, (unsigned)i);
                   1088:        m->m_len -= olen;
                   1089:        if (m->m_flags & M_PKTHDR)
                   1090:                m->m_pkthdr.len -= olen;
                   1091:        ip->ip_hl = sizeof(struct ip) >> 2;
                   1092: }
                   1093:
1.23      mycroft  1094: int inetctlerrmap[PRC_NCMDS] = {
1.1       cgd      1095:        0,              0,              0,              0,
                   1096:        0,              EMSGSIZE,       EHOSTDOWN,      EHOSTUNREACH,
                   1097:        EHOSTUNREACH,   EHOSTUNREACH,   ECONNREFUSED,   ECONNREFUSED,
                   1098:        EMSGSIZE,       EHOSTUNREACH,   0,              0,
                   1099:        0,              0,              0,              0,
                   1100:        ENOPROTOOPT
                   1101: };
                   1102:
                   1103: /*
                   1104:  * Forward a packet.  If some error occurs return the sender
                   1105:  * an icmp packet.  Note we can't always generate a meaningful
                   1106:  * icmp message because icmp doesn't have a large enough repertoire
                   1107:  * of codes and types.
                   1108:  *
                   1109:  * If not forwarding, just drop the packet.  This could be confusing
                   1110:  * if ipforwarding was zero but some routing protocol was advancing
                   1111:  * us as a gateway to somewhere.  However, we must let the routing
                   1112:  * protocol deal with that.
                   1113:  *
                   1114:  * The srcrt parameter indicates whether the packet is being forwarded
                   1115:  * via a source route.
                   1116:  */
1.13      mycroft  1117: void
1.1       cgd      1118: ip_forward(m, srcrt)
                   1119:        struct mbuf *m;
                   1120:        int srcrt;
                   1121: {
                   1122:        register struct ip *ip = mtod(m, struct ip *);
                   1123:        register struct sockaddr_in *sin;
                   1124:        register struct rtentry *rt;
1.28      christos 1125:        int error, type = 0, code = 0;
1.1       cgd      1126:        struct mbuf *mcopy;
1.13      mycroft  1127:        n_long dest;
                   1128:        struct ifnet *destifp;
1.1       cgd      1129:
1.13      mycroft  1130:        dest = 0;
1.1       cgd      1131: #ifdef DIAGNOSTIC
                   1132:        if (ipprintfs)
1.39      christos 1133:                printf("forward: src %x dst %x ttl %x\n",
1.35      mycroft  1134:                    ip->ip_src.s_addr, ip->ip_dst.s_addr, ip->ip_ttl);
1.1       cgd      1135: #endif
                   1136:        if (m->m_flags & M_BCAST || in_canforward(ip->ip_dst) == 0) {
                   1137:                ipstat.ips_cantforward++;
                   1138:                m_freem(m);
                   1139:                return;
                   1140:        }
                   1141:        HTONS(ip->ip_id);
                   1142:        if (ip->ip_ttl <= IPTTLDEC) {
1.13      mycroft  1143:                icmp_error(m, ICMP_TIMXCEED, ICMP_TIMXCEED_INTRANS, dest, 0);
1.1       cgd      1144:                return;
                   1145:        }
                   1146:        ip->ip_ttl -= IPTTLDEC;
                   1147:
1.19      mycroft  1148:        sin = satosin(&ipforward_rt.ro_dst);
1.1       cgd      1149:        if ((rt = ipforward_rt.ro_rt) == 0 ||
1.35      mycroft  1150:            !in_hosteq(ip->ip_dst, sin->sin_addr)) {
1.1       cgd      1151:                if (ipforward_rt.ro_rt) {
                   1152:                        RTFREE(ipforward_rt.ro_rt);
                   1153:                        ipforward_rt.ro_rt = 0;
                   1154:                }
                   1155:                sin->sin_family = AF_INET;
1.35      mycroft  1156:                sin->sin_len = sizeof(struct sockaddr_in);
1.1       cgd      1157:                sin->sin_addr = ip->ip_dst;
                   1158:
                   1159:                rtalloc(&ipforward_rt);
                   1160:                if (ipforward_rt.ro_rt == 0) {
1.13      mycroft  1161:                        icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_HOST, dest, 0);
1.1       cgd      1162:                        return;
                   1163:                }
                   1164:                rt = ipforward_rt.ro_rt;
                   1165:        }
                   1166:
                   1167:        /*
1.34      mycroft  1168:         * Save at most 68 bytes of the packet in case
1.1       cgd      1169:         * we need to generate an ICMP message to the src.
                   1170:         */
1.34      mycroft  1171:        mcopy = m_copy(m, 0, imin((int)ip->ip_len, 68));
1.1       cgd      1172:
                   1173:        /*
                   1174:         * If forwarding packet using same interface that it came in on,
                   1175:         * perhaps should send a redirect to sender to shortcut a hop.
                   1176:         * Only send redirect if source is sending directly to us,
                   1177:         * and if packet was not source routed (or has any options).
                   1178:         * Also, don't send redirect if forwarding using a default route
                   1179:         * or a route modified by a redirect.
                   1180:         */
                   1181:        if (rt->rt_ifp == m->m_pkthdr.rcvif &&
                   1182:            (rt->rt_flags & (RTF_DYNAMIC|RTF_MODIFIED)) == 0 &&
1.35      mycroft  1183:            !in_nullhost(satosin(rt_key(rt))->sin_addr) &&
1.1       cgd      1184:            ipsendredirects && !srcrt) {
1.19      mycroft  1185:                if (rt->rt_ifa &&
                   1186:                    (ip->ip_src.s_addr & ifatoia(rt->rt_ifa)->ia_subnetmask) ==
                   1187:                    ifatoia(rt->rt_ifa)->ia_subnet) {
1.1       cgd      1188:                    if (rt->rt_flags & RTF_GATEWAY)
1.13      mycroft  1189:                        dest = satosin(rt->rt_gateway)->sin_addr.s_addr;
1.1       cgd      1190:                    else
1.13      mycroft  1191:                        dest = ip->ip_dst.s_addr;
                   1192:                    /* Router requirements says to only send host redirects */
1.1       cgd      1193:                    type = ICMP_REDIRECT;
1.13      mycroft  1194:                    code = ICMP_REDIRECT_HOST;
1.1       cgd      1195: #ifdef DIAGNOSTIC
                   1196:                    if (ipprintfs)
1.39      christos 1197:                        printf("redirect (%d) to %x\n", code, (u_int32_t)dest);
1.1       cgd      1198: #endif
                   1199:                }
                   1200:        }
                   1201:
1.27      thorpej  1202:        error = ip_output(m, (struct mbuf *)0, &ipforward_rt,
                   1203:            (IP_FORWARDING | (ip_directedbcast ? IP_ALLOWBROADCAST : 0)), 0);
1.1       cgd      1204:        if (error)
                   1205:                ipstat.ips_cantforward++;
                   1206:        else {
                   1207:                ipstat.ips_forward++;
                   1208:                if (type)
                   1209:                        ipstat.ips_redirectsent++;
                   1210:                else {
1.63      matt     1211:                        if (mcopy) {
                   1212: #ifdef GATEWAY
1.64      thorpej  1213:                                if (mcopy->m_flags & M_CANFASTFWD)
                   1214:                                        ipflow_create(&ipforward_rt, mcopy);
1.63      matt     1215: #endif
1.1       cgd      1216:                                m_freem(mcopy);
1.63      matt     1217:                        }
1.1       cgd      1218:                        return;
                   1219:                }
                   1220:        }
                   1221:        if (mcopy == NULL)
                   1222:                return;
1.13      mycroft  1223:        destifp = NULL;
                   1224:
1.1       cgd      1225:        switch (error) {
                   1226:
                   1227:        case 0:                         /* forwarded, but need redirect */
                   1228:                /* type, code set above */
                   1229:                break;
                   1230:
                   1231:        case ENETUNREACH:               /* shouldn't happen, checked above */
                   1232:        case EHOSTUNREACH:
                   1233:        case ENETDOWN:
                   1234:        case EHOSTDOWN:
                   1235:        default:
                   1236:                type = ICMP_UNREACH;
                   1237:                code = ICMP_UNREACH_HOST;
                   1238:                break;
                   1239:
                   1240:        case EMSGSIZE:
                   1241:                type = ICMP_UNREACH;
                   1242:                code = ICMP_UNREACH_NEEDFRAG;
1.13      mycroft  1243:                if (ipforward_rt.ro_rt)
                   1244:                        destifp = ipforward_rt.ro_rt->rt_ifp;
1.1       cgd      1245:                ipstat.ips_cantfrag++;
                   1246:                break;
                   1247:
                   1248:        case ENOBUFS:
                   1249:                type = ICMP_SOURCEQUENCH;
                   1250:                code = 0;
                   1251:                break;
                   1252:        }
1.13      mycroft  1253:        icmp_error(mcopy, type, code, dest, destifp);
1.44      thorpej  1254: }
                   1255:
                   1256: void
                   1257: ip_savecontrol(inp, mp, ip, m)
                   1258:        register struct inpcb *inp;
                   1259:        register struct mbuf **mp;
                   1260:        register struct ip *ip;
                   1261:        register struct mbuf *m;
                   1262: {
                   1263:
                   1264:        if (inp->inp_socket->so_options & SO_TIMESTAMP) {
                   1265:                struct timeval tv;
                   1266:
                   1267:                microtime(&tv);
                   1268:                *mp = sbcreatecontrol((caddr_t) &tv, sizeof(tv),
                   1269:                    SCM_TIMESTAMP, SOL_SOCKET);
                   1270:                if (*mp)
                   1271:                        mp = &(*mp)->m_next;
                   1272:        }
                   1273:        if (inp->inp_flags & INP_RECVDSTADDR) {
                   1274:                *mp = sbcreatecontrol((caddr_t) &ip->ip_dst,
                   1275:                    sizeof(struct in_addr), IP_RECVDSTADDR, IPPROTO_IP);
                   1276:                if (*mp)
                   1277:                        mp = &(*mp)->m_next;
                   1278:        }
                   1279: #ifdef notyet
                   1280:        /*
                   1281:         * XXX
                   1282:         * Moving these out of udp_input() made them even more broken
                   1283:         * than they already were.
                   1284:         *      - fenner@parc.xerox.com
                   1285:         */
                   1286:        /* options were tossed already */
                   1287:        if (inp->inp_flags & INP_RECVOPTS) {
                   1288:                *mp = sbcreatecontrol((caddr_t) opts_deleted_above,
                   1289:                    sizeof(struct in_addr), IP_RECVOPTS, IPPROTO_IP);
                   1290:                if (*mp)
                   1291:                        mp = &(*mp)->m_next;
                   1292:        }
                   1293:        /* ip_srcroute doesn't do what we want here, need to fix */
                   1294:        if (inp->inp_flags & INP_RECVRETOPTS) {
                   1295:                *mp = sbcreatecontrol((caddr_t) ip_srcroute(),
                   1296:                    sizeof(struct in_addr), IP_RECVRETOPTS, IPPROTO_IP);
                   1297:                if (*mp)
                   1298:                        mp = &(*mp)->m_next;
                   1299:        }
                   1300: #endif
                   1301:        if (inp->inp_flags & INP_RECVIF) {
                   1302:                struct sockaddr_dl sdl;
                   1303:
                   1304:                sdl.sdl_len = offsetof(struct sockaddr_dl, sdl_data[0]);
                   1305:                sdl.sdl_family = AF_LINK;
                   1306:                sdl.sdl_index = m->m_pkthdr.rcvif ?
                   1307:                    m->m_pkthdr.rcvif->if_index : 0;
                   1308:                sdl.sdl_nlen = sdl.sdl_alen = sdl.sdl_slen = 0;
                   1309:                *mp = sbcreatecontrol((caddr_t) &sdl, sdl.sdl_len,
                   1310:                    IP_RECVIF, IPPROTO_IP);
                   1311:                if (*mp)
                   1312:                        mp = &(*mp)->m_next;
                   1313:        }
1.13      mycroft  1314: }
                   1315:
                   1316: int
                   1317: ip_sysctl(name, namelen, oldp, oldlenp, newp, newlen)
                   1318:        int *name;
                   1319:        u_int namelen;
                   1320:        void *oldp;
                   1321:        size_t *oldlenp;
                   1322:        void *newp;
                   1323:        size_t newlen;
                   1324: {
1.52      thorpej  1325:        extern int subnetsarelocal;
                   1326:
1.54      lukem    1327:        int error, old;
                   1328:
1.13      mycroft  1329:        /* All sysctl names at this level are terminal. */
                   1330:        if (namelen != 1)
                   1331:                return (ENOTDIR);
                   1332:
                   1333:        switch (name[0]) {
                   1334:        case IPCTL_FORWARDING:
                   1335:                return (sysctl_int(oldp, oldlenp, newp, newlen, &ipforwarding));
                   1336:        case IPCTL_SENDREDIRECTS:
                   1337:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1338:                        &ipsendredirects));
                   1339:        case IPCTL_DEFTTL:
                   1340:                return (sysctl_int(oldp, oldlenp, newp, newlen, &ip_defttl));
                   1341: #ifdef notyet
                   1342:        case IPCTL_DEFMTU:
                   1343:                return (sysctl_int(oldp, oldlenp, newp, newlen, &ip_mtu));
                   1344: #endif
1.26      thorpej  1345:        case IPCTL_FORWSRCRT:
1.47      cjs      1346:                /* Don't allow this to change in a secure environment.  */
1.26      thorpej  1347:                if (securelevel > 0)
1.46      cjs      1348:                        return (sysctl_rdint(oldp, oldlenp, newp,
                   1349:                            ip_forwsrcrt));
                   1350:                else
                   1351:                        return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1352:                            &ip_forwsrcrt));
1.27      thorpej  1353:        case IPCTL_DIRECTEDBCAST:
                   1354:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1355:                    &ip_directedbcast));
1.47      cjs      1356:        case IPCTL_ALLOWSRCRT:
                   1357:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1358:                    &ip_allowsrcrt));
1.52      thorpej  1359:        case IPCTL_SUBNETSARELOCAL:
                   1360:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1361:                    &subnetsarelocal));
1.53      kml      1362:        case IPCTL_MTUDISC:
1.60      kml      1363:                error = sysctl_int(oldp, oldlenp, newp, newlen,
                   1364:                    &ip_mtudisc);
                   1365:                if (ip_mtudisc != 0 && ip_mtudisc_timeout_q == NULL) {
                   1366:                        ip_mtudisc_timeout_q =
                   1367:                            rt_timer_queue_create(ip_mtudisc_timeout);
                   1368:                } else if (ip_mtudisc == 0 && ip_mtudisc_timeout_q != NULL) {
                   1369:                        rt_timer_queue_destroy(ip_mtudisc_timeout_q, TRUE);
                   1370:                        ip_mtudisc_timeout_q = NULL;
                   1371:                }
                   1372:                return error;
1.54      lukem    1373:        case IPCTL_ANONPORTMIN:
                   1374:                old = anonportmin;
                   1375:                error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmin);
                   1376:                if (anonportmin >= anonportmax || anonportmin > 65535
                   1377: #ifndef IPNOPRIVPORTS
                   1378:                    || anonportmin < IPPORT_RESERVED
                   1379: #endif
                   1380:                    ) {
                   1381:                        anonportmin = old;
                   1382:                        return (EINVAL);
                   1383:                }
                   1384:                return (error);
                   1385:        case IPCTL_ANONPORTMAX:
                   1386:                old = anonportmax;
                   1387:                error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmax);
                   1388:                if (anonportmin >= anonportmax || anonportmax > 65535
                   1389: #ifndef IPNOPRIVPORTS
                   1390:                    || anonportmax < IPPORT_RESERVED
                   1391: #endif
                   1392:                    ) {
                   1393:                        anonportmax = old;
                   1394:                        return (EINVAL);
                   1395:                }
1.60      kml      1396:                return (error);
                   1397:        case IPCTL_MTUDISCTIMEOUT:
                   1398:                error = sysctl_int(oldp, oldlenp, newp, newlen,
                   1399:                   &ip_mtudisc_timeout);
                   1400:                if (ip_mtudisc_timeout_q != NULL)
                   1401:                        rt_timer_queue_change(ip_mtudisc_timeout_q,
                   1402:                                              ip_mtudisc_timeout);
1.54      lukem    1403:                return (error);
1.65      matt     1404: #ifdef GATEWAY
                   1405:        case IPCTL_MAXFLOWS:
1.67      thorpej  1406:            {
                   1407:                int s;
                   1408:
1.65      matt     1409:                error = sysctl_int(oldp, oldlenp, newp, newlen,
                   1410:                   &ip_maxflows);
1.67      thorpej  1411:                s = splsoftnet();
1.65      matt     1412:                ipflow_reap(0);
1.67      thorpej  1413:                splx(s);
1.65      matt     1414:                return (error);
1.67      thorpej  1415:            }
1.65      matt     1416: #endif
1.13      mycroft  1417:        default:
                   1418:                return (EOPNOTSUPP);
                   1419:        }
                   1420:        /* NOTREACHED */
1.1       cgd      1421: }

CVSweb <webmaster@jp.NetBSD.org>