[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 1.71

1.71    ! tls         1: /*     $NetBSD: ip_input.c,v 1.70 1998/09/09 04:57:18 thorpej Exp $    */
1.14      cgd         2:
1.1       cgd         3: /*
1.13      mycroft     4:  * Copyright (c) 1982, 1986, 1988, 1993
                      5:  *     The Regents of the University of California.  All rights reserved.
1.1       cgd         6:  *
                      7:  * Redistribution and use in source and binary forms, with or without
                      8:  * modification, are permitted provided that the following conditions
                      9:  * are met:
                     10:  * 1. Redistributions of source code must retain the above copyright
                     11:  *    notice, this list of conditions and the following disclaimer.
                     12:  * 2. Redistributions in binary form must reproduce the above copyright
                     13:  *    notice, this list of conditions and the following disclaimer in the
                     14:  *    documentation and/or other materials provided with the distribution.
                     15:  * 3. All advertising materials mentioning features or use of this software
                     16:  *    must display the following acknowledgement:
                     17:  *     This product includes software developed by the University of
                     18:  *     California, Berkeley and its contributors.
                     19:  * 4. Neither the name of the University nor the names of its contributors
                     20:  *    may be used to endorse or promote products derived from this software
                     21:  *    without specific prior written permission.
                     22:  *
                     23:  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
                     24:  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
                     25:  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
                     26:  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
                     27:  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
                     28:  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
                     29:  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
                     30:  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
                     31:  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
                     32:  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
                     33:  * SUCH DAMAGE.
                     34:  *
1.14      cgd        35:  *     @(#)ip_input.c  8.2 (Berkeley) 1/4/94
1.71    ! tls        36:  */
        !            37:
        !            38: /*-
        !            39:  * Copyright (c) 1998 The NetBSD Foundation, Inc.
        !            40:  * All rights reserved.
        !            41:  *
        !            42:  * This code is derived from software contributed to The NetBSD Foundation
        !            43:  * by Public Access Networks Corporation ("Panix").  It was developed under
        !            44:  * contract to Panix by Eric Haszlakiewicz and Thor Lancelot Simon.
        !            45:  *
        !            46:  * Redistribution and use in source and binary forms, with or without
        !            47:  * modification, are permitted provided that the following conditions
        !            48:  * are met:
        !            49:  * 1. Redistributions of source code must retain the above copyright
        !            50:  *    notice, this list of conditions and the following disclaimer.
        !            51:  * 2. Redistributions in binary form must reproduce the above copyright
        !            52:  *    notice, this list of conditions and the following disclaimer in the
        !            53:  *    documentation and/or other materials provided with the distribution.
        !            54:  * 3. All advertising materials mentioning features or use of this software
        !            55:  *    must display the following acknowledgement:
        !            56:  *      This product includes software developed by the NetBSD
        !            57:  *      Foundation, Inc. and its contributors.
        !            58:  * 4. Neither the name of The NetBSD Foundation nor the names of its
        !            59:  *    contributors may be used to endorse or promote products derived
        !            60:  *    from this software without specific prior written permission.
        !            61:  *
        !            62:  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
        !            63:  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
        !            64:  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
        !            65:  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
        !            66:  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
        !            67:  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
        !            68:  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
        !            69:  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
        !            70:  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
        !            71:  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
        !            72:  * POSSIBILITY OF SUCH DAMAGE.
1.1       cgd        73:  */
1.55      scottr     74:
1.62      matt       75: #include "opt_gateway.h"
1.69      mrg        76: #include "opt_pfil_hooks.h"
1.55      scottr     77: #include "opt_mrouting.h"
1.1       cgd        78:
1.5       mycroft    79: #include <sys/param.h>
                     80: #include <sys/systm.h>
                     81: #include <sys/malloc.h>
                     82: #include <sys/mbuf.h>
                     83: #include <sys/domain.h>
                     84: #include <sys/protosw.h>
                     85: #include <sys/socket.h>
1.44      thorpej    86: #include <sys/socketvar.h>
1.5       mycroft    87: #include <sys/errno.h>
                     88: #include <sys/time.h>
                     89: #include <sys/kernel.h>
1.28      christos   90: #include <sys/proc.h>
                     91:
                     92: #include <vm/vm.h>
                     93: #include <sys/sysctl.h>
1.1       cgd        94:
1.5       mycroft    95: #include <net/if.h>
1.44      thorpej    96: #include <net/if_dl.h>
1.5       mycroft    97: #include <net/route.h>
1.45      mrg        98: #include <net/pfil.h>
1.1       cgd        99:
1.5       mycroft   100: #include <netinet/in.h>
                    101: #include <netinet/in_systm.h>
                    102: #include <netinet/ip.h>
                    103: #include <netinet/in_pcb.h>
                    104: #include <netinet/in_var.h>
                    105: #include <netinet/ip_var.h>
                    106: #include <netinet/ip_icmp.h>
1.44      thorpej   107:
1.1       cgd       108: #ifndef        IPFORWARDING
                    109: #ifdef GATEWAY
                    110: #define        IPFORWARDING    1       /* forward IP packets not for us */
                    111: #else /* GATEWAY */
                    112: #define        IPFORWARDING    0       /* don't forward IP packets not for us */
                    113: #endif /* GATEWAY */
                    114: #endif /* IPFORWARDING */
                    115: #ifndef        IPSENDREDIRECTS
                    116: #define        IPSENDREDIRECTS 1
                    117: #endif
1.26      thorpej   118: #ifndef IPFORWSRCRT
1.47      cjs       119: #define        IPFORWSRCRT     1       /* forward source-routed packets */
                    120: #endif
                    121: #ifndef IPALLOWSRCRT
1.48      mrg       122: #define        IPALLOWSRCRT    1       /* allow source-routed packets */
1.26      thorpej   123: #endif
1.53      kml       124: #ifndef IPMTUDISC
                    125: #define IPMTUDISC      0
                    126: #endif
1.60      kml       127: #ifndef IPMTUDISCTIMEOUT
1.61      kml       128: #define IPMTUDISCTIMEOUT (10 * 60)     /* as per RFC 1191 */
1.60      kml       129: #endif
1.53      kml       130:
1.27      thorpej   131: /*
                    132:  * Note: DIRECTED_BROADCAST is handled this way so that previous
                    133:  * configuration using this option will Just Work.
                    134:  */
                    135: #ifndef IPDIRECTEDBCAST
                    136: #ifdef DIRECTED_BROADCAST
                    137: #define IPDIRECTEDBCAST        1
                    138: #else
                    139: #define        IPDIRECTEDBCAST 0
                    140: #endif /* DIRECTED_BROADCAST */
                    141: #endif /* IPDIRECTEDBCAST */
1.1       cgd       142: int    ipforwarding = IPFORWARDING;
                    143: int    ipsendredirects = IPSENDREDIRECTS;
1.13      mycroft   144: int    ip_defttl = IPDEFTTL;
1.26      thorpej   145: int    ip_forwsrcrt = IPFORWSRCRT;
1.27      thorpej   146: int    ip_directedbcast = IPDIRECTEDBCAST;
1.47      cjs       147: int    ip_allowsrcrt = IPALLOWSRCRT;
1.53      kml       148: int    ip_mtudisc = IPMTUDISC;
1.60      kml       149: u_int  ip_mtudisc_timeout = IPMTUDISCTIMEOUT;
1.1       cgd       150: #ifdef DIAGNOSTIC
                    151: int    ipprintfs = 0;
                    152: #endif
                    153:
1.60      kml       154: struct rttimer_queue *ip_mtudisc_timeout_q = NULL;
                    155:
1.1       cgd       156: extern struct domain inetdomain;
                    157: extern struct protosw inetsw[];
                    158: u_char ip_protox[IPPROTO_MAX];
                    159: int    ipqmaxlen = IFQ_MAXLEN;
1.22      mycroft   160: struct in_ifaddrhead in_ifaddr;
1.57      tls       161: struct in_ifaddrhashhead *in_ifaddrhashtbl;
1.13      mycroft   162: struct ifqueue ipintrq;
1.63      matt      163: struct ipstat  ipstat;
                    164: u_int16_t      ip_id;
                    165: int    ip_defttl;
                    166: struct ipqhead ipq;
1.1       cgd       167:
                    168: /*
                    169:  * We need to save the IP options in case a protocol wants to respond
                    170:  * to an incoming packet over the same route if the packet got here
                    171:  * using IP source routing.  This allows connection establishment and
                    172:  * maintenance when the remote end is on a network that is not known
                    173:  * to us.
                    174:  */
                    175: int    ip_nhops = 0;
                    176: static struct ip_srcrt {
                    177:        struct  in_addr dst;                    /* final destination */
                    178:        char    nop;                            /* one NOP to align */
                    179:        char    srcopt[IPOPT_OFFSET + 1];       /* OPTVAL, OLEN and OFFSET */
                    180:        struct  in_addr route[MAX_IPOPTLEN/sizeof(struct in_addr)];
                    181: } ip_srcrt;
                    182:
1.13      mycroft   183: static void save_rte __P((u_char *, struct in_addr));
1.35      mycroft   184:
1.1       cgd       185: /*
                    186:  * IP initialization: fill in IP protocol switch table.
                    187:  * All protocols not implemented in kernel go to raw IP protocol handler.
                    188:  */
1.8       mycroft   189: void
1.1       cgd       190: ip_init()
                    191: {
                    192:        register struct protosw *pr;
                    193:        register int i;
                    194:
                    195:        pr = pffindproto(PF_INET, IPPROTO_RAW, SOCK_RAW);
                    196:        if (pr == 0)
                    197:                panic("ip_init");
                    198:        for (i = 0; i < IPPROTO_MAX; i++)
                    199:                ip_protox[i] = pr - inetsw;
                    200:        for (pr = inetdomain.dom_protosw;
                    201:            pr < inetdomain.dom_protoswNPROTOSW; pr++)
                    202:                if (pr->pr_domain->dom_family == PF_INET &&
                    203:                    pr->pr_protocol && pr->pr_protocol != IPPROTO_RAW)
                    204:                        ip_protox[pr->pr_protocol] = pr - inetsw;
1.25      cgd       205:        LIST_INIT(&ipq);
1.1       cgd       206:        ip_id = time.tv_sec & 0xffff;
                    207:        ipintrq.ifq_maxlen = ipqmaxlen;
1.22      mycroft   208:        TAILQ_INIT(&in_ifaddr);
1.57      tls       209:        in_ifaddrhashtbl =
                    210:            hashinit(IN_IFADDR_HASH_SIZE, M_IFADDR, M_WAITOK, &in_ifaddrhash);
1.60      kml       211:        if (ip_mtudisc != 0)
                    212:                ip_mtudisc_timeout_q =
                    213:                    rt_timer_queue_create(ip_mtudisc_timeout);
1.1       cgd       214: }
                    215:
                    216: struct sockaddr_in ipaddr = { sizeof(ipaddr), AF_INET };
                    217: struct route ipforward_rt;
                    218:
                    219: /*
                    220:  * Ip input routine.  Checksum and byte swap header.  If fragmented
                    221:  * try to reassemble.  Process options.  Pass to next level.
                    222:  */
1.8       mycroft   223: void
1.1       cgd       224: ipintr()
                    225: {
1.33      mrg       226:        register struct ip *ip = NULL;
1.1       cgd       227:        register struct mbuf *m;
                    228:        register struct ipq *fp;
                    229:        register struct in_ifaddr *ia;
1.57      tls       230:        register struct ifaddr *ifa;
1.25      cgd       231:        struct ipqent *ipqe;
1.35      mycroft   232:        int hlen = 0, mff, len, s;
1.36      mrg       233: #ifdef PFIL_HOOKS
1.33      mrg       234:        struct packet_filter_hook *pfh;
                    235:        struct mbuf *m0;
1.43      mrg       236:        int rv;
1.36      mrg       237: #endif /* PFIL_HOOKS */
1.1       cgd       238:
                    239: next:
                    240:        /*
                    241:         * Get next datagram off input queue and get IP header
                    242:         * in first mbuf.
                    243:         */
                    244:        s = splimp();
                    245:        IF_DEQUEUE(&ipintrq, m);
                    246:        splx(s);
1.13      mycroft   247:        if (m == 0)
1.1       cgd       248:                return;
                    249: #ifdef DIAGNOSTIC
                    250:        if ((m->m_flags & M_PKTHDR) == 0)
                    251:                panic("ipintr no HDR");
                    252: #endif
                    253:        /*
                    254:         * If no IP addresses have been set yet but the interfaces
                    255:         * are receiving, can't do anything with incoming packets yet.
                    256:         */
1.22      mycroft   257:        if (in_ifaddr.tqh_first == 0)
1.1       cgd       258:                goto bad;
                    259:        ipstat.ips_total++;
                    260:        if (m->m_len < sizeof (struct ip) &&
                    261:            (m = m_pullup(m, sizeof (struct ip))) == 0) {
                    262:                ipstat.ips_toosmall++;
                    263:                goto next;
                    264:        }
                    265:        ip = mtod(m, struct ip *);
1.13      mycroft   266:        if (ip->ip_v != IPVERSION) {
                    267:                ipstat.ips_badvers++;
                    268:                goto bad;
                    269:        }
1.1       cgd       270:        hlen = ip->ip_hl << 2;
                    271:        if (hlen < sizeof(struct ip)) { /* minimum header length */
                    272:                ipstat.ips_badhlen++;
                    273:                goto bad;
                    274:        }
                    275:        if (hlen > m->m_len) {
                    276:                if ((m = m_pullup(m, hlen)) == 0) {
                    277:                        ipstat.ips_badhlen++;
                    278:                        goto next;
                    279:                }
                    280:                ip = mtod(m, struct ip *);
                    281:        }
1.28      christos  282:        if ((ip->ip_sum = in_cksum(m, hlen)) != 0) {
1.1       cgd       283:                ipstat.ips_badsum++;
                    284:                goto bad;
                    285:        }
                    286:
                    287:        /*
                    288:         * Convert fields to host representation.
                    289:         */
                    290:        NTOHS(ip->ip_len);
                    291:        NTOHS(ip->ip_id);
                    292:        NTOHS(ip->ip_off);
1.35      mycroft   293:        len = ip->ip_len;
1.1       cgd       294:
                    295:        /*
                    296:         * Check that the amount of data in the buffers
                    297:         * is as at least much as the IP header would have us expect.
                    298:         * Trim mbufs if longer than we expect.
                    299:         * Drop packet if shorter than we expect.
                    300:         */
1.35      mycroft   301:        if (m->m_pkthdr.len < len) {
1.1       cgd       302:                ipstat.ips_tooshort++;
                    303:                goto bad;
                    304:        }
1.35      mycroft   305:        if (m->m_pkthdr.len > len) {
1.1       cgd       306:                if (m->m_len == m->m_pkthdr.len) {
1.35      mycroft   307:                        m->m_len = len;
                    308:                        m->m_pkthdr.len = len;
1.1       cgd       309:                } else
1.35      mycroft   310:                        m_adj(m, len - m->m_pkthdr.len);
1.1       cgd       311:        }
                    312:
1.64      thorpej   313:        /*
                    314:         * Assume that we can create a fast-forward IP flow entry
                    315:         * based on this packet.
                    316:         */
                    317:        m->m_flags |= M_CANFASTFWD;
                    318:
1.36      mrg       319: #ifdef PFIL_HOOKS
1.33      mrg       320:        /*
1.64      thorpej   321:         * Run through list of hooks for input packets.  If there are any
                    322:         * filters which require that additional packets in the flow are
                    323:         * not fast-forwarded, they must clear the M_CANFASTFWD flag.
                    324:         * Note that filters must _never_ set this flag, as another filter
                    325:         * in the list may have previously cleared it.
1.33      mrg       326:         */
                    327:        m0 = m;
1.59      mrg       328:        for (pfh = pfil_hook_get(PFIL_IN); pfh; pfh = pfh->pfil_link.tqe_next)
1.33      mrg       329:                if (pfh->pfil_func) {
1.43      mrg       330:                        rv = pfh->pfil_func(ip, hlen, m->m_pkthdr.rcvif, 0, &m0);
                    331:                        if (rv)
1.40      veego     332:                                goto next;
1.68      sommerfe  333:                        m = m0;
                    334:                        if (m == NULL)
                    335:                                goto next;
                    336:                        ip = mtod(m, struct ip *);
1.33      mrg       337:                }
1.36      mrg       338: #endif /* PFIL_HOOKS */
1.33      mrg       339:
1.1       cgd       340:        /*
                    341:         * Process options and, if not destined for us,
                    342:         * ship it on.  ip_dooptions returns 1 when an
                    343:         * error was detected (causing an icmp message
                    344:         * to be sent and the original packet to be freed).
                    345:         */
                    346:        ip_nhops = 0;           /* for source routed packets */
                    347:        if (hlen > sizeof (struct ip) && ip_dooptions(m))
                    348:                goto next;
                    349:
                    350:        /*
                    351:         * Check our list of addresses, to see if the packet is for us.
                    352:         */
1.57      tls       353:        INADDR_TO_IA(ip->ip_dst, ia);
                    354:        if (ia != NULL) goto ours;
                    355:        if (m->m_pkthdr.rcvif->if_flags & IFF_BROADCAST) {
                    356:                for (ifa = m->m_pkthdr.rcvif->if_addrlist.tqh_first;
                    357:                    ifa != NULL; ifa = ifa->ifa_list.tqe_next) {
                    358:                        if (ifa->ifa_addr->sa_family != AF_INET) continue;
                    359:                        ia = ifatoia(ifa);
1.35      mycroft   360:                        if (in_hosteq(ip->ip_dst, ia->ia_broadaddr.sin_addr) ||
                    361:                            in_hosteq(ip->ip_dst, ia->ia_netbroadcast) ||
1.20      mycroft   362:                            /*
                    363:                             * Look for all-0's host part (old broadcast addr),
                    364:                             * either for subnet or net.
                    365:                             */
                    366:                            ip->ip_dst.s_addr == ia->ia_subnet ||
1.18      mycroft   367:                            ip->ip_dst.s_addr == ia->ia_net)
1.1       cgd       368:                                goto ours;
1.57      tls       369:                        /*
                    370:                         * An interface with IP address zero accepts
                    371:                         * all packets that arrive on that interface.
                    372:                         */
                    373:                        if (in_nullhost(ia->ia_addr.sin_addr))
                    374:                                goto ours;
1.1       cgd       375:                }
                    376:        }
1.18      mycroft   377:        if (IN_MULTICAST(ip->ip_dst.s_addr)) {
1.4       hpeyerl   378:                struct in_multi *inm;
                    379: #ifdef MROUTING
                    380:                extern struct socket *ip_mrouter;
1.10      brezak    381:
                    382:                if (m->m_flags & M_EXT) {
                    383:                        if ((m = m_pullup(m, hlen)) == 0) {
                    384:                                ipstat.ips_toosmall++;
                    385:                                goto next;
                    386:                        }
                    387:                        ip = mtod(m, struct ip *);
                    388:                }
1.4       hpeyerl   389:
                    390:                if (ip_mrouter) {
                    391:                        /*
                    392:                         * If we are acting as a multicast router, all
                    393:                         * incoming multicast packets are passed to the
                    394:                         * kernel-level multicast forwarding function.
                    395:                         * The packet is returned (relatively) intact; if
                    396:                         * ip_mforward() returns a non-zero value, the packet
                    397:                         * must be discarded, else it may be accepted below.
                    398:                         *
                    399:                         * (The IP ident field is put in the same byte order
                    400:                         * as expected when ip_mforward() is called from
                    401:                         * ip_output().)
                    402:                         */
                    403:                        ip->ip_id = htons(ip->ip_id);
1.13      mycroft   404:                        if (ip_mforward(m, m->m_pkthdr.rcvif) != 0) {
                    405:                                ipstat.ips_cantforward++;
1.4       hpeyerl   406:                                m_freem(m);
                    407:                                goto next;
                    408:                        }
                    409:                        ip->ip_id = ntohs(ip->ip_id);
                    410:
                    411:                        /*
                    412:                         * The process-level routing demon needs to receive
                    413:                         * all multicast IGMP packets, whether or not this
                    414:                         * host belongs to their destination groups.
                    415:                         */
                    416:                        if (ip->ip_p == IPPROTO_IGMP)
                    417:                                goto ours;
1.13      mycroft   418:                        ipstat.ips_forward++;
1.4       hpeyerl   419:                }
                    420: #endif
                    421:                /*
                    422:                 * See if we belong to the destination multicast group on the
                    423:                 * arrival interface.
                    424:                 */
                    425:                IN_LOOKUP_MULTI(ip->ip_dst, m->m_pkthdr.rcvif, inm);
                    426:                if (inm == NULL) {
1.13      mycroft   427:                        ipstat.ips_cantforward++;
1.4       hpeyerl   428:                        m_freem(m);
                    429:                        goto next;
                    430:                }
                    431:                goto ours;
                    432:        }
1.19      mycroft   433:        if (ip->ip_dst.s_addr == INADDR_BROADCAST ||
1.35      mycroft   434:            in_nullhost(ip->ip_dst))
1.1       cgd       435:                goto ours;
                    436:
                    437:        /*
                    438:         * Not for us; forward if possible and desirable.
                    439:         */
                    440:        if (ipforwarding == 0) {
                    441:                ipstat.ips_cantforward++;
                    442:                m_freem(m);
                    443:        } else
                    444:                ip_forward(m, 0);
                    445:        goto next;
                    446:
                    447: ours:
                    448:        /*
                    449:         * If offset or IP_MF are set, must reassemble.
                    450:         * Otherwise, nothing need be done.
                    451:         * (We could look in the reassembly queue to see
                    452:         * if the packet was previously fragmented,
                    453:         * but it's not worth the time; just let them time out.)
                    454:         */
1.37      perry     455:        if (ip->ip_off & ~(IP_DF|IP_RF)) {
1.1       cgd       456:                /*
                    457:                 * Look for queue of fragments
                    458:                 * of this datagram.
                    459:                 */
1.25      cgd       460:                for (fp = ipq.lh_first; fp != NULL; fp = fp->ipq_q.le_next)
1.1       cgd       461:                        if (ip->ip_id == fp->ipq_id &&
1.35      mycroft   462:                            in_hosteq(ip->ip_src, fp->ipq_src) &&
                    463:                            in_hosteq(ip->ip_dst, fp->ipq_dst) &&
1.1       cgd       464:                            ip->ip_p == fp->ipq_p)
                    465:                                goto found;
                    466:                fp = 0;
                    467: found:
                    468:
                    469:                /*
                    470:                 * Adjust ip_len to not reflect header,
1.25      cgd       471:                 * set ipqe_mff if more fragments are expected,
1.1       cgd       472:                 * convert offset of this to bytes.
                    473:                 */
                    474:                ip->ip_len -= hlen;
1.25      cgd       475:                mff = (ip->ip_off & IP_MF) != 0;
                    476:                if (mff) {
1.16      cgd       477:                        /*
                    478:                         * Make sure that fragments have a data length
                    479:                         * that's a non-zero multiple of 8 bytes.
                    480:                         */
1.17      cgd       481:                        if (ip->ip_len == 0 || (ip->ip_len & 0x7) != 0) {
1.16      cgd       482:                                ipstat.ips_badfrags++;
                    483:                                goto bad;
                    484:                        }
                    485:                }
1.1       cgd       486:                ip->ip_off <<= 3;
                    487:
                    488:                /*
                    489:                 * If datagram marked as having more fragments
                    490:                 * or if this is not the first fragment,
                    491:                 * attempt reassembly; if it succeeds, proceed.
                    492:                 */
1.25      cgd       493:                if (mff || ip->ip_off) {
1.1       cgd       494:                        ipstat.ips_fragments++;
1.25      cgd       495:                        MALLOC(ipqe, struct ipqent *, sizeof (struct ipqent),
                    496:                            M_IPQ, M_NOWAIT);
                    497:                        if (ipqe == NULL) {
                    498:                                ipstat.ips_rcvmemdrop++;
                    499:                                goto bad;
                    500:                        }
                    501:                        ipqe->ipqe_mff = mff;
1.50      thorpej   502:                        ipqe->ipqe_m = m;
1.25      cgd       503:                        ipqe->ipqe_ip = ip;
1.50      thorpej   504:                        m = ip_reass(ipqe, fp);
                    505:                        if (m == 0)
1.1       cgd       506:                                goto next;
1.13      mycroft   507:                        ipstat.ips_reassembled++;
1.50      thorpej   508:                        ip = mtod(m, struct ip *);
1.1       cgd       509:                } else
                    510:                        if (fp)
                    511:                                ip_freef(fp);
                    512:        } else
                    513:                ip->ip_len -= hlen;
                    514:
                    515:        /*
                    516:         * Switch out to protocol's input routine.
                    517:         */
                    518:        ipstat.ips_delivered++;
                    519:        (*inetsw[ip_protox[ip->ip_p]].pr_input)(m, hlen);
                    520:        goto next;
                    521: bad:
                    522:        m_freem(m);
                    523:        goto next;
                    524: }
                    525:
                    526: /*
                    527:  * Take incoming datagram fragment and try to
                    528:  * reassemble it into whole datagram.  If a chain for
                    529:  * reassembly of this datagram already exists, then it
                    530:  * is given as fp; otherwise have to make a chain.
                    531:  */
1.50      thorpej   532: struct mbuf *
1.25      cgd       533: ip_reass(ipqe, fp)
                    534:        register struct ipqent *ipqe;
1.1       cgd       535:        register struct ipq *fp;
                    536: {
1.50      thorpej   537:        register struct mbuf *m = ipqe->ipqe_m;
1.25      cgd       538:        register struct ipqent *nq, *p, *q;
                    539:        struct ip *ip;
1.1       cgd       540:        struct mbuf *t;
1.25      cgd       541:        int hlen = ipqe->ipqe_ip->ip_hl << 2;
1.1       cgd       542:        int i, next;
                    543:
                    544:        /*
                    545:         * Presence of header sizes in mbufs
                    546:         * would confuse code below.
                    547:         */
                    548:        m->m_data += hlen;
                    549:        m->m_len -= hlen;
                    550:
                    551:        /*
                    552:         * If first fragment to arrive, create a reassembly queue.
                    553:         */
                    554:        if (fp == 0) {
1.50      thorpej   555:                MALLOC(fp, struct ipq *, sizeof (struct ipq),
                    556:                    M_FTABLE, M_NOWAIT);
                    557:                if (fp == NULL)
1.1       cgd       558:                        goto dropfrag;
1.25      cgd       559:                LIST_INSERT_HEAD(&ipq, fp, ipq_q);
1.1       cgd       560:                fp->ipq_ttl = IPFRAGTTL;
1.25      cgd       561:                fp->ipq_p = ipqe->ipqe_ip->ip_p;
                    562:                fp->ipq_id = ipqe->ipqe_ip->ip_id;
                    563:                LIST_INIT(&fp->ipq_fragq);
                    564:                fp->ipq_src = ipqe->ipqe_ip->ip_src;
                    565:                fp->ipq_dst = ipqe->ipqe_ip->ip_dst;
                    566:                p = NULL;
1.1       cgd       567:                goto insert;
                    568:        }
                    569:
                    570:        /*
                    571:         * Find a segment which begins after this one does.
                    572:         */
1.25      cgd       573:        for (p = NULL, q = fp->ipq_fragq.lh_first; q != NULL;
                    574:            p = q, q = q->ipqe_q.le_next)
                    575:                if (q->ipqe_ip->ip_off > ipqe->ipqe_ip->ip_off)
1.1       cgd       576:                        break;
                    577:
                    578:        /*
                    579:         * If there is a preceding segment, it may provide some of
                    580:         * our data already.  If so, drop the data from the incoming
                    581:         * segment.  If it provides all of our data, drop us.
                    582:         */
1.25      cgd       583:        if (p != NULL) {
                    584:                i = p->ipqe_ip->ip_off + p->ipqe_ip->ip_len -
                    585:                    ipqe->ipqe_ip->ip_off;
1.1       cgd       586:                if (i > 0) {
1.25      cgd       587:                        if (i >= ipqe->ipqe_ip->ip_len)
1.1       cgd       588:                                goto dropfrag;
1.50      thorpej   589:                        m_adj(ipqe->ipqe_m, i);
1.25      cgd       590:                        ipqe->ipqe_ip->ip_off += i;
                    591:                        ipqe->ipqe_ip->ip_len -= i;
1.1       cgd       592:                }
                    593:        }
                    594:
                    595:        /*
                    596:         * While we overlap succeeding segments trim them or,
                    597:         * if they are completely covered, dequeue them.
                    598:         */
1.25      cgd       599:        for (; q != NULL && ipqe->ipqe_ip->ip_off + ipqe->ipqe_ip->ip_len >
                    600:            q->ipqe_ip->ip_off; q = nq) {
                    601:                i = (ipqe->ipqe_ip->ip_off + ipqe->ipqe_ip->ip_len) -
                    602:                    q->ipqe_ip->ip_off;
                    603:                if (i < q->ipqe_ip->ip_len) {
                    604:                        q->ipqe_ip->ip_len -= i;
                    605:                        q->ipqe_ip->ip_off += i;
1.50      thorpej   606:                        m_adj(q->ipqe_m, i);
1.1       cgd       607:                        break;
                    608:                }
1.25      cgd       609:                nq = q->ipqe_q.le_next;
1.50      thorpej   610:                m_freem(q->ipqe_m);
1.25      cgd       611:                LIST_REMOVE(q, ipqe_q);
                    612:                FREE(q, M_IPQ);
1.1       cgd       613:        }
                    614:
                    615: insert:
                    616:        /*
                    617:         * Stick new segment in its place;
                    618:         * check for complete reassembly.
                    619:         */
1.25      cgd       620:        if (p == NULL) {
                    621:                LIST_INSERT_HEAD(&fp->ipq_fragq, ipqe, ipqe_q);
                    622:        } else {
                    623:                LIST_INSERT_AFTER(p, ipqe, ipqe_q);
                    624:        }
1.1       cgd       625:        next = 0;
1.25      cgd       626:        for (p = NULL, q = fp->ipq_fragq.lh_first; q != NULL;
                    627:            p = q, q = q->ipqe_q.le_next) {
                    628:                if (q->ipqe_ip->ip_off != next)
1.1       cgd       629:                        return (0);
1.25      cgd       630:                next += q->ipqe_ip->ip_len;
1.1       cgd       631:        }
1.25      cgd       632:        if (p->ipqe_mff)
1.1       cgd       633:                return (0);
                    634:
                    635:        /*
1.41      thorpej   636:         * Reassembly is complete.  Check for a bogus message size and
                    637:         * concatenate fragments.
1.1       cgd       638:         */
1.25      cgd       639:        q = fp->ipq_fragq.lh_first;
                    640:        ip = q->ipqe_ip;
1.41      thorpej   641:        if ((next + (ip->ip_hl << 2)) > IP_MAXPACKET) {
                    642:                ipstat.ips_toolong++;
                    643:                ip_freef(fp);
                    644:                return (0);
                    645:        }
1.50      thorpej   646:        m = q->ipqe_m;
1.1       cgd       647:        t = m->m_next;
                    648:        m->m_next = 0;
                    649:        m_cat(m, t);
1.25      cgd       650:        nq = q->ipqe_q.le_next;
                    651:        FREE(q, M_IPQ);
                    652:        for (q = nq; q != NULL; q = nq) {
1.50      thorpej   653:                t = q->ipqe_m;
1.25      cgd       654:                nq = q->ipqe_q.le_next;
                    655:                FREE(q, M_IPQ);
1.1       cgd       656:                m_cat(m, t);
                    657:        }
                    658:
                    659:        /*
                    660:         * Create header for new ip packet by
                    661:         * modifying header of first packet;
                    662:         * dequeue and discard fragment reassembly header.
                    663:         * Make header visible.
                    664:         */
                    665:        ip->ip_len = next;
1.25      cgd       666:        ip->ip_src = fp->ipq_src;
                    667:        ip->ip_dst = fp->ipq_dst;
                    668:        LIST_REMOVE(fp, ipq_q);
1.50      thorpej   669:        FREE(fp, M_FTABLE);
1.1       cgd       670:        m->m_len += (ip->ip_hl << 2);
                    671:        m->m_data -= (ip->ip_hl << 2);
                    672:        /* some debugging cruft by sklower, below, will go away soon */
                    673:        if (m->m_flags & M_PKTHDR) { /* XXX this should be done elsewhere */
                    674:                register int plen = 0;
1.50      thorpej   675:                for (t = m; t; t = t->m_next)
                    676:                        plen += t->m_len;
                    677:                m->m_pkthdr.len = plen;
1.1       cgd       678:        }
1.50      thorpej   679:        return (m);
1.1       cgd       680:
                    681: dropfrag:
                    682:        ipstat.ips_fragdropped++;
                    683:        m_freem(m);
1.25      cgd       684:        FREE(ipqe, M_IPQ);
1.1       cgd       685:        return (0);
                    686: }
                    687:
                    688: /*
                    689:  * Free a fragment reassembly header and all
                    690:  * associated datagrams.
                    691:  */
1.8       mycroft   692: void
1.1       cgd       693: ip_freef(fp)
                    694:        struct ipq *fp;
                    695: {
1.25      cgd       696:        register struct ipqent *q, *p;
1.1       cgd       697:
1.25      cgd       698:        for (q = fp->ipq_fragq.lh_first; q != NULL; q = p) {
                    699:                p = q->ipqe_q.le_next;
1.50      thorpej   700:                m_freem(q->ipqe_m);
1.25      cgd       701:                LIST_REMOVE(q, ipqe_q);
                    702:                FREE(q, M_IPQ);
1.1       cgd       703:        }
1.25      cgd       704:        LIST_REMOVE(fp, ipq_q);
1.50      thorpej   705:        FREE(fp, M_FTABLE);
1.1       cgd       706: }
                    707:
                    708: /*
                    709:  * IP timer processing;
                    710:  * if a timer expires on a reassembly
                    711:  * queue, discard it.
                    712:  */
1.8       mycroft   713: void
1.1       cgd       714: ip_slowtimo()
                    715: {
1.25      cgd       716:        register struct ipq *fp, *nfp;
1.24      mycroft   717:        int s = splsoftnet();
1.1       cgd       718:
1.25      cgd       719:        for (fp = ipq.lh_first; fp != NULL; fp = nfp) {
                    720:                nfp = fp->ipq_q.le_next;
                    721:                if (--fp->ipq_ttl == 0) {
1.1       cgd       722:                        ipstat.ips_fragtimeout++;
1.25      cgd       723:                        ip_freef(fp);
1.1       cgd       724:                }
                    725:        }
1.63      matt      726: #ifdef GATEWAY
                    727:        ipflow_slowtimo();
                    728: #endif
1.1       cgd       729:        splx(s);
                    730: }
                    731:
                    732: /*
                    733:  * Drain off all datagram fragments.
                    734:  */
1.8       mycroft   735: void
1.1       cgd       736: ip_drain()
                    737: {
                    738:
1.25      cgd       739:        while (ipq.lh_first != NULL) {
1.1       cgd       740:                ipstat.ips_fragdropped++;
1.25      cgd       741:                ip_freef(ipq.lh_first);
1.1       cgd       742:        }
                    743: }
                    744:
                    745: /*
                    746:  * Do option processing on a datagram,
                    747:  * possibly discarding it if bad options are encountered,
                    748:  * or forwarding it if source-routed.
                    749:  * Returns 1 if packet has been forwarded/freed,
                    750:  * 0 if the packet should be processed further.
                    751:  */
1.8       mycroft   752: int
1.1       cgd       753: ip_dooptions(m)
                    754:        struct mbuf *m;
                    755: {
                    756:        register struct ip *ip = mtod(m, struct ip *);
                    757:        register u_char *cp;
                    758:        register struct ip_timestamp *ipt;
                    759:        register struct in_ifaddr *ia;
                    760:        int opt, optlen, cnt, off, code, type = ICMP_PARAMPROB, forward = 0;
1.13      mycroft   761:        struct in_addr *sin, dst;
1.1       cgd       762:        n_time ntime;
                    763:
1.13      mycroft   764:        dst = ip->ip_dst;
1.1       cgd       765:        cp = (u_char *)(ip + 1);
                    766:        cnt = (ip->ip_hl << 2) - sizeof (struct ip);
                    767:        for (; cnt > 0; cnt -= optlen, cp += optlen) {
                    768:                opt = cp[IPOPT_OPTVAL];
                    769:                if (opt == IPOPT_EOL)
                    770:                        break;
                    771:                if (opt == IPOPT_NOP)
                    772:                        optlen = 1;
                    773:                else {
                    774:                        optlen = cp[IPOPT_OLEN];
                    775:                        if (optlen <= 0 || optlen > cnt) {
                    776:                                code = &cp[IPOPT_OLEN] - (u_char *)ip;
                    777:                                goto bad;
                    778:                        }
                    779:                }
                    780:                switch (opt) {
                    781:
                    782:                default:
                    783:                        break;
                    784:
                    785:                /*
                    786:                 * Source routing with record.
                    787:                 * Find interface with current destination address.
                    788:                 * If none on this machine then drop if strictly routed,
                    789:                 * or do nothing if loosely routed.
                    790:                 * Record interface address and bring up next address
                    791:                 * component.  If strictly routed make sure next
                    792:                 * address is on directly accessible net.
                    793:                 */
                    794:                case IPOPT_LSRR:
                    795:                case IPOPT_SSRR:
1.47      cjs       796:                        if (ip_allowsrcrt == 0) {
                    797:                                type = ICMP_UNREACH;
                    798:                                code = ICMP_UNREACH_NET_PROHIB;
                    799:                                goto bad;
                    800:                        }
1.1       cgd       801:                        if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) {
                    802:                                code = &cp[IPOPT_OFFSET] - (u_char *)ip;
                    803:                                goto bad;
                    804:                        }
                    805:                        ipaddr.sin_addr = ip->ip_dst;
1.19      mycroft   806:                        ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr)));
1.1       cgd       807:                        if (ia == 0) {
                    808:                                if (opt == IPOPT_SSRR) {
                    809:                                        type = ICMP_UNREACH;
                    810:                                        code = ICMP_UNREACH_SRCFAIL;
                    811:                                        goto bad;
                    812:                                }
                    813:                                /*
                    814:                                 * Loose routing, and not at next destination
                    815:                                 * yet; nothing to do except forward.
                    816:                                 */
                    817:                                break;
                    818:                        }
                    819:                        off--;                  /* 0 origin */
                    820:                        if (off > optlen - sizeof(struct in_addr)) {
                    821:                                /*
                    822:                                 * End of source route.  Should be for us.
                    823:                                 */
                    824:                                save_rte(cp, ip->ip_src);
                    825:                                break;
                    826:                        }
                    827:                        /*
                    828:                         * locate outgoing interface
                    829:                         */
                    830:                        bcopy((caddr_t)(cp + off), (caddr_t)&ipaddr.sin_addr,
                    831:                            sizeof(ipaddr.sin_addr));
                    832:                        if (opt == IPOPT_SSRR) {
                    833: #define        INA     struct in_ifaddr *
                    834: #define        SA      struct sockaddr *
1.29      mrg       835:                            ia = (INA)ifa_ifwithladdr((SA)&ipaddr);
1.1       cgd       836:                        } else
                    837:                                ia = ip_rtaddr(ipaddr.sin_addr);
                    838:                        if (ia == 0) {
                    839:                                type = ICMP_UNREACH;
                    840:                                code = ICMP_UNREACH_SRCFAIL;
                    841:                                goto bad;
                    842:                        }
                    843:                        ip->ip_dst = ipaddr.sin_addr;
1.20      mycroft   844:                        bcopy((caddr_t)&ia->ia_addr.sin_addr,
1.1       cgd       845:                            (caddr_t)(cp + off), sizeof(struct in_addr));
                    846:                        cp[IPOPT_OFFSET] += sizeof(struct in_addr);
1.13      mycroft   847:                        /*
                    848:                         * Let ip_intr's mcast routing check handle mcast pkts
                    849:                         */
1.18      mycroft   850:                        forward = !IN_MULTICAST(ip->ip_dst.s_addr);
1.1       cgd       851:                        break;
                    852:
                    853:                case IPOPT_RR:
                    854:                        if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) {
                    855:                                code = &cp[IPOPT_OFFSET] - (u_char *)ip;
                    856:                                goto bad;
                    857:                        }
                    858:                        /*
                    859:                         * If no space remains, ignore.
                    860:                         */
                    861:                        off--;                  /* 0 origin */
                    862:                        if (off > optlen - sizeof(struct in_addr))
                    863:                                break;
                    864:                        bcopy((caddr_t)(&ip->ip_dst), (caddr_t)&ipaddr.sin_addr,
                    865:                            sizeof(ipaddr.sin_addr));
                    866:                        /*
                    867:                         * locate outgoing interface; if we're the destination,
                    868:                         * use the incoming interface (should be same).
                    869:                         */
                    870:                        if ((ia = (INA)ifa_ifwithaddr((SA)&ipaddr)) == 0 &&
                    871:                            (ia = ip_rtaddr(ipaddr.sin_addr)) == 0) {
                    872:                                type = ICMP_UNREACH;
                    873:                                code = ICMP_UNREACH_HOST;
                    874:                                goto bad;
                    875:                        }
1.20      mycroft   876:                        bcopy((caddr_t)&ia->ia_addr.sin_addr,
1.1       cgd       877:                            (caddr_t)(cp + off), sizeof(struct in_addr));
                    878:                        cp[IPOPT_OFFSET] += sizeof(struct in_addr);
                    879:                        break;
                    880:
                    881:                case IPOPT_TS:
                    882:                        code = cp - (u_char *)ip;
                    883:                        ipt = (struct ip_timestamp *)cp;
                    884:                        if (ipt->ipt_len < 5)
                    885:                                goto bad;
1.15      cgd       886:                        if (ipt->ipt_ptr > ipt->ipt_len - sizeof (int32_t)) {
1.1       cgd       887:                                if (++ipt->ipt_oflw == 0)
                    888:                                        goto bad;
                    889:                                break;
                    890:                        }
                    891:                        sin = (struct in_addr *)(cp + ipt->ipt_ptr - 1);
                    892:                        switch (ipt->ipt_flg) {
                    893:
                    894:                        case IPOPT_TS_TSONLY:
                    895:                                break;
                    896:
                    897:                        case IPOPT_TS_TSANDADDR:
1.66      thorpej   898:                                if (ipt->ipt_ptr - 1 + sizeof(n_time) +
1.1       cgd       899:                                    sizeof(struct in_addr) > ipt->ipt_len)
                    900:                                        goto bad;
1.13      mycroft   901:                                ipaddr.sin_addr = dst;
                    902:                                ia = (INA)ifaof_ifpforaddr((SA)&ipaddr,
                    903:                                                            m->m_pkthdr.rcvif);
                    904:                                if (ia == 0)
                    905:                                        continue;
1.20      mycroft   906:                                bcopy((caddr_t)&ia->ia_addr.sin_addr,
1.1       cgd       907:                                    (caddr_t)sin, sizeof(struct in_addr));
                    908:                                ipt->ipt_ptr += sizeof(struct in_addr);
                    909:                                break;
                    910:
                    911:                        case IPOPT_TS_PRESPEC:
1.66      thorpej   912:                                if (ipt->ipt_ptr - 1 + sizeof(n_time) +
1.1       cgd       913:                                    sizeof(struct in_addr) > ipt->ipt_len)
                    914:                                        goto bad;
                    915:                                bcopy((caddr_t)sin, (caddr_t)&ipaddr.sin_addr,
                    916:                                    sizeof(struct in_addr));
                    917:                                if (ifa_ifwithaddr((SA)&ipaddr) == 0)
                    918:                                        continue;
                    919:                                ipt->ipt_ptr += sizeof(struct in_addr);
                    920:                                break;
                    921:
                    922:                        default:
                    923:                                goto bad;
                    924:                        }
                    925:                        ntime = iptime();
                    926:                        bcopy((caddr_t)&ntime, (caddr_t)cp + ipt->ipt_ptr - 1,
                    927:                            sizeof(n_time));
                    928:                        ipt->ipt_ptr += sizeof(n_time);
                    929:                }
                    930:        }
                    931:        if (forward) {
1.26      thorpej   932:                if (ip_forwsrcrt == 0) {
                    933:                        type = ICMP_UNREACH;
                    934:                        code = ICMP_UNREACH_SRCFAIL;
                    935:                        goto bad;
                    936:                }
1.1       cgd       937:                ip_forward(m, 1);
                    938:                return (1);
1.13      mycroft   939:        }
                    940:        return (0);
1.1       cgd       941: bad:
1.13      mycroft   942:        ip->ip_len -= ip->ip_hl << 2;   /* XXX icmp_error adds in hdr length */
                    943:        icmp_error(m, type, code, 0, 0);
                    944:        ipstat.ips_badoptions++;
1.1       cgd       945:        return (1);
                    946: }
                    947:
                    948: /*
                    949:  * Given address of next destination (final or next hop),
                    950:  * return internet address info of interface to be used to get there.
                    951:  */
                    952: struct in_ifaddr *
                    953: ip_rtaddr(dst)
                    954:         struct in_addr dst;
                    955: {
                    956:        register struct sockaddr_in *sin;
                    957:
1.19      mycroft   958:        sin = satosin(&ipforward_rt.ro_dst);
1.1       cgd       959:
1.35      mycroft   960:        if (ipforward_rt.ro_rt == 0 || !in_hosteq(dst, sin->sin_addr)) {
1.1       cgd       961:                if (ipforward_rt.ro_rt) {
                    962:                        RTFREE(ipforward_rt.ro_rt);
                    963:                        ipforward_rt.ro_rt = 0;
                    964:                }
                    965:                sin->sin_family = AF_INET;
                    966:                sin->sin_len = sizeof(*sin);
                    967:                sin->sin_addr = dst;
                    968:
                    969:                rtalloc(&ipforward_rt);
                    970:        }
                    971:        if (ipforward_rt.ro_rt == 0)
                    972:                return ((struct in_ifaddr *)0);
1.19      mycroft   973:        return (ifatoia(ipforward_rt.ro_rt->rt_ifa));
1.1       cgd       974: }
                    975:
                    976: /*
                    977:  * Save incoming source route for use in replies,
                    978:  * to be picked up later by ip_srcroute if the receiver is interested.
                    979:  */
1.13      mycroft   980: void
1.1       cgd       981: save_rte(option, dst)
                    982:        u_char *option;
                    983:        struct in_addr dst;
                    984: {
                    985:        unsigned olen;
                    986:
                    987:        olen = option[IPOPT_OLEN];
                    988: #ifdef DIAGNOSTIC
                    989:        if (ipprintfs)
1.39      christos  990:                printf("save_rte: olen %d\n", olen);
1.1       cgd       991: #endif
                    992:        if (olen > sizeof(ip_srcrt) - (1 + sizeof(dst)))
                    993:                return;
                    994:        bcopy((caddr_t)option, (caddr_t)ip_srcrt.srcopt, olen);
                    995:        ip_nhops = (olen - IPOPT_OFFSET - 1) / sizeof(struct in_addr);
                    996:        ip_srcrt.dst = dst;
                    997: }
                    998:
                    999: /*
                   1000:  * Retrieve incoming source route for use in replies,
                   1001:  * in the same form used by setsockopt.
                   1002:  * The first hop is placed before the options, will be removed later.
                   1003:  */
                   1004: struct mbuf *
                   1005: ip_srcroute()
                   1006: {
                   1007:        register struct in_addr *p, *q;
                   1008:        register struct mbuf *m;
                   1009:
                   1010:        if (ip_nhops == 0)
                   1011:                return ((struct mbuf *)0);
                   1012:        m = m_get(M_DONTWAIT, MT_SOOPTS);
                   1013:        if (m == 0)
                   1014:                return ((struct mbuf *)0);
                   1015:
1.13      mycroft  1016: #define OPTSIZ (sizeof(ip_srcrt.nop) + sizeof(ip_srcrt.srcopt))
1.1       cgd      1017:
                   1018:        /* length is (nhops+1)*sizeof(addr) + sizeof(nop + srcrt header) */
                   1019:        m->m_len = ip_nhops * sizeof(struct in_addr) + sizeof(struct in_addr) +
                   1020:            OPTSIZ;
                   1021: #ifdef DIAGNOSTIC
                   1022:        if (ipprintfs)
1.39      christos 1023:                printf("ip_srcroute: nhops %d mlen %d", ip_nhops, m->m_len);
1.1       cgd      1024: #endif
                   1025:
                   1026:        /*
                   1027:         * First save first hop for return route
                   1028:         */
                   1029:        p = &ip_srcrt.route[ip_nhops - 1];
                   1030:        *(mtod(m, struct in_addr *)) = *p--;
                   1031: #ifdef DIAGNOSTIC
                   1032:        if (ipprintfs)
1.39      christos 1033:                printf(" hops %x", ntohl(mtod(m, struct in_addr *)->s_addr));
1.1       cgd      1034: #endif
                   1035:
                   1036:        /*
                   1037:         * Copy option fields and padding (nop) to mbuf.
                   1038:         */
                   1039:        ip_srcrt.nop = IPOPT_NOP;
                   1040:        ip_srcrt.srcopt[IPOPT_OFFSET] = IPOPT_MINOFF;
                   1041:        bcopy((caddr_t)&ip_srcrt.nop,
                   1042:            mtod(m, caddr_t) + sizeof(struct in_addr), OPTSIZ);
                   1043:        q = (struct in_addr *)(mtod(m, caddr_t) +
                   1044:            sizeof(struct in_addr) + OPTSIZ);
                   1045: #undef OPTSIZ
                   1046:        /*
                   1047:         * Record return path as an IP source route,
                   1048:         * reversing the path (pointers are now aligned).
                   1049:         */
                   1050:        while (p >= ip_srcrt.route) {
                   1051: #ifdef DIAGNOSTIC
                   1052:                if (ipprintfs)
1.39      christos 1053:                        printf(" %x", ntohl(q->s_addr));
1.1       cgd      1054: #endif
                   1055:                *q++ = *p--;
                   1056:        }
                   1057:        /*
                   1058:         * Last hop goes to final destination.
                   1059:         */
                   1060:        *q = ip_srcrt.dst;
                   1061: #ifdef DIAGNOSTIC
                   1062:        if (ipprintfs)
1.39      christos 1063:                printf(" %x\n", ntohl(q->s_addr));
1.1       cgd      1064: #endif
                   1065:        return (m);
                   1066: }
                   1067:
                   1068: /*
                   1069:  * Strip out IP options, at higher
                   1070:  * level protocol in the kernel.
                   1071:  * Second argument is buffer to which options
                   1072:  * will be moved, and return value is their length.
                   1073:  * XXX should be deleted; last arg currently ignored.
                   1074:  */
1.8       mycroft  1075: void
1.1       cgd      1076: ip_stripoptions(m, mopt)
                   1077:        register struct mbuf *m;
                   1078:        struct mbuf *mopt;
                   1079: {
                   1080:        register int i;
                   1081:        struct ip *ip = mtod(m, struct ip *);
                   1082:        register caddr_t opts;
                   1083:        int olen;
                   1084:
                   1085:        olen = (ip->ip_hl<<2) - sizeof (struct ip);
                   1086:        opts = (caddr_t)(ip + 1);
                   1087:        i = m->m_len - (sizeof (struct ip) + olen);
                   1088:        bcopy(opts  + olen, opts, (unsigned)i);
                   1089:        m->m_len -= olen;
                   1090:        if (m->m_flags & M_PKTHDR)
                   1091:                m->m_pkthdr.len -= olen;
                   1092:        ip->ip_hl = sizeof(struct ip) >> 2;
                   1093: }
                   1094:
1.23      mycroft  1095: int inetctlerrmap[PRC_NCMDS] = {
1.1       cgd      1096:        0,              0,              0,              0,
                   1097:        0,              EMSGSIZE,       EHOSTDOWN,      EHOSTUNREACH,
                   1098:        EHOSTUNREACH,   EHOSTUNREACH,   ECONNREFUSED,   ECONNREFUSED,
                   1099:        EMSGSIZE,       EHOSTUNREACH,   0,              0,
                   1100:        0,              0,              0,              0,
                   1101:        ENOPROTOOPT
                   1102: };
                   1103:
                   1104: /*
                   1105:  * Forward a packet.  If some error occurs return the sender
                   1106:  * an icmp packet.  Note we can't always generate a meaningful
                   1107:  * icmp message because icmp doesn't have a large enough repertoire
                   1108:  * of codes and types.
                   1109:  *
                   1110:  * If not forwarding, just drop the packet.  This could be confusing
                   1111:  * if ipforwarding was zero but some routing protocol was advancing
                   1112:  * us as a gateway to somewhere.  However, we must let the routing
                   1113:  * protocol deal with that.
                   1114:  *
                   1115:  * The srcrt parameter indicates whether the packet is being forwarded
                   1116:  * via a source route.
                   1117:  */
1.13      mycroft  1118: void
1.1       cgd      1119: ip_forward(m, srcrt)
                   1120:        struct mbuf *m;
                   1121:        int srcrt;
                   1122: {
                   1123:        register struct ip *ip = mtod(m, struct ip *);
                   1124:        register struct sockaddr_in *sin;
                   1125:        register struct rtentry *rt;
1.28      christos 1126:        int error, type = 0, code = 0;
1.1       cgd      1127:        struct mbuf *mcopy;
1.13      mycroft  1128:        n_long dest;
                   1129:        struct ifnet *destifp;
1.1       cgd      1130:
1.13      mycroft  1131:        dest = 0;
1.1       cgd      1132: #ifdef DIAGNOSTIC
                   1133:        if (ipprintfs)
1.70      thorpej  1134:                printf("forward: src %2.2x dst %2.2x ttl %x\n",
                   1135:                    ntohl(ip->ip_src.s_addr),
                   1136:                    ntohl(ip->ip_dst.s_addr), ip->ip_ttl);
1.1       cgd      1137: #endif
                   1138:        if (m->m_flags & M_BCAST || in_canforward(ip->ip_dst) == 0) {
                   1139:                ipstat.ips_cantforward++;
                   1140:                m_freem(m);
                   1141:                return;
                   1142:        }
                   1143:        HTONS(ip->ip_id);
                   1144:        if (ip->ip_ttl <= IPTTLDEC) {
1.13      mycroft  1145:                icmp_error(m, ICMP_TIMXCEED, ICMP_TIMXCEED_INTRANS, dest, 0);
1.1       cgd      1146:                return;
                   1147:        }
                   1148:        ip->ip_ttl -= IPTTLDEC;
                   1149:
1.19      mycroft  1150:        sin = satosin(&ipforward_rt.ro_dst);
1.1       cgd      1151:        if ((rt = ipforward_rt.ro_rt) == 0 ||
1.35      mycroft  1152:            !in_hosteq(ip->ip_dst, sin->sin_addr)) {
1.1       cgd      1153:                if (ipforward_rt.ro_rt) {
                   1154:                        RTFREE(ipforward_rt.ro_rt);
                   1155:                        ipforward_rt.ro_rt = 0;
                   1156:                }
                   1157:                sin->sin_family = AF_INET;
1.35      mycroft  1158:                sin->sin_len = sizeof(struct sockaddr_in);
1.1       cgd      1159:                sin->sin_addr = ip->ip_dst;
                   1160:
                   1161:                rtalloc(&ipforward_rt);
                   1162:                if (ipforward_rt.ro_rt == 0) {
1.13      mycroft  1163:                        icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_HOST, dest, 0);
1.1       cgd      1164:                        return;
                   1165:                }
                   1166:                rt = ipforward_rt.ro_rt;
                   1167:        }
                   1168:
                   1169:        /*
1.34      mycroft  1170:         * Save at most 68 bytes of the packet in case
1.1       cgd      1171:         * we need to generate an ICMP message to the src.
                   1172:         */
1.34      mycroft  1173:        mcopy = m_copy(m, 0, imin((int)ip->ip_len, 68));
1.1       cgd      1174:
                   1175:        /*
                   1176:         * If forwarding packet using same interface that it came in on,
                   1177:         * perhaps should send a redirect to sender to shortcut a hop.
                   1178:         * Only send redirect if source is sending directly to us,
                   1179:         * and if packet was not source routed (or has any options).
                   1180:         * Also, don't send redirect if forwarding using a default route
                   1181:         * or a route modified by a redirect.
                   1182:         */
                   1183:        if (rt->rt_ifp == m->m_pkthdr.rcvif &&
                   1184:            (rt->rt_flags & (RTF_DYNAMIC|RTF_MODIFIED)) == 0 &&
1.35      mycroft  1185:            !in_nullhost(satosin(rt_key(rt))->sin_addr) &&
1.1       cgd      1186:            ipsendredirects && !srcrt) {
1.19      mycroft  1187:                if (rt->rt_ifa &&
                   1188:                    (ip->ip_src.s_addr & ifatoia(rt->rt_ifa)->ia_subnetmask) ==
                   1189:                    ifatoia(rt->rt_ifa)->ia_subnet) {
1.1       cgd      1190:                    if (rt->rt_flags & RTF_GATEWAY)
1.13      mycroft  1191:                        dest = satosin(rt->rt_gateway)->sin_addr.s_addr;
1.1       cgd      1192:                    else
1.13      mycroft  1193:                        dest = ip->ip_dst.s_addr;
                   1194:                    /* Router requirements says to only send host redirects */
1.1       cgd      1195:                    type = ICMP_REDIRECT;
1.13      mycroft  1196:                    code = ICMP_REDIRECT_HOST;
1.1       cgd      1197: #ifdef DIAGNOSTIC
                   1198:                    if (ipprintfs)
1.39      christos 1199:                        printf("redirect (%d) to %x\n", code, (u_int32_t)dest);
1.1       cgd      1200: #endif
                   1201:                }
                   1202:        }
                   1203:
1.27      thorpej  1204:        error = ip_output(m, (struct mbuf *)0, &ipforward_rt,
                   1205:            (IP_FORWARDING | (ip_directedbcast ? IP_ALLOWBROADCAST : 0)), 0);
1.1       cgd      1206:        if (error)
                   1207:                ipstat.ips_cantforward++;
                   1208:        else {
                   1209:                ipstat.ips_forward++;
                   1210:                if (type)
                   1211:                        ipstat.ips_redirectsent++;
                   1212:                else {
1.63      matt     1213:                        if (mcopy) {
                   1214: #ifdef GATEWAY
1.64      thorpej  1215:                                if (mcopy->m_flags & M_CANFASTFWD)
                   1216:                                        ipflow_create(&ipforward_rt, mcopy);
1.63      matt     1217: #endif
1.1       cgd      1218:                                m_freem(mcopy);
1.63      matt     1219:                        }
1.1       cgd      1220:                        return;
                   1221:                }
                   1222:        }
                   1223:        if (mcopy == NULL)
                   1224:                return;
1.13      mycroft  1225:        destifp = NULL;
                   1226:
1.1       cgd      1227:        switch (error) {
                   1228:
                   1229:        case 0:                         /* forwarded, but need redirect */
                   1230:                /* type, code set above */
                   1231:                break;
                   1232:
                   1233:        case ENETUNREACH:               /* shouldn't happen, checked above */
                   1234:        case EHOSTUNREACH:
                   1235:        case ENETDOWN:
                   1236:        case EHOSTDOWN:
                   1237:        default:
                   1238:                type = ICMP_UNREACH;
                   1239:                code = ICMP_UNREACH_HOST;
                   1240:                break;
                   1241:
                   1242:        case EMSGSIZE:
                   1243:                type = ICMP_UNREACH;
                   1244:                code = ICMP_UNREACH_NEEDFRAG;
1.13      mycroft  1245:                if (ipforward_rt.ro_rt)
                   1246:                        destifp = ipforward_rt.ro_rt->rt_ifp;
1.1       cgd      1247:                ipstat.ips_cantfrag++;
                   1248:                break;
                   1249:
                   1250:        case ENOBUFS:
                   1251:                type = ICMP_SOURCEQUENCH;
                   1252:                code = 0;
                   1253:                break;
                   1254:        }
1.13      mycroft  1255:        icmp_error(mcopy, type, code, dest, destifp);
1.44      thorpej  1256: }
                   1257:
                   1258: void
                   1259: ip_savecontrol(inp, mp, ip, m)
                   1260:        register struct inpcb *inp;
                   1261:        register struct mbuf **mp;
                   1262:        register struct ip *ip;
                   1263:        register struct mbuf *m;
                   1264: {
                   1265:
                   1266:        if (inp->inp_socket->so_options & SO_TIMESTAMP) {
                   1267:                struct timeval tv;
                   1268:
                   1269:                microtime(&tv);
                   1270:                *mp = sbcreatecontrol((caddr_t) &tv, sizeof(tv),
                   1271:                    SCM_TIMESTAMP, SOL_SOCKET);
                   1272:                if (*mp)
                   1273:                        mp = &(*mp)->m_next;
                   1274:        }
                   1275:        if (inp->inp_flags & INP_RECVDSTADDR) {
                   1276:                *mp = sbcreatecontrol((caddr_t) &ip->ip_dst,
                   1277:                    sizeof(struct in_addr), IP_RECVDSTADDR, IPPROTO_IP);
                   1278:                if (*mp)
                   1279:                        mp = &(*mp)->m_next;
                   1280:        }
                   1281: #ifdef notyet
                   1282:        /*
                   1283:         * XXX
                   1284:         * Moving these out of udp_input() made them even more broken
                   1285:         * than they already were.
                   1286:         *      - fenner@parc.xerox.com
                   1287:         */
                   1288:        /* options were tossed already */
                   1289:        if (inp->inp_flags & INP_RECVOPTS) {
                   1290:                *mp = sbcreatecontrol((caddr_t) opts_deleted_above,
                   1291:                    sizeof(struct in_addr), IP_RECVOPTS, IPPROTO_IP);
                   1292:                if (*mp)
                   1293:                        mp = &(*mp)->m_next;
                   1294:        }
                   1295:        /* ip_srcroute doesn't do what we want here, need to fix */
                   1296:        if (inp->inp_flags & INP_RECVRETOPTS) {
                   1297:                *mp = sbcreatecontrol((caddr_t) ip_srcroute(),
                   1298:                    sizeof(struct in_addr), IP_RECVRETOPTS, IPPROTO_IP);
                   1299:                if (*mp)
                   1300:                        mp = &(*mp)->m_next;
                   1301:        }
                   1302: #endif
                   1303:        if (inp->inp_flags & INP_RECVIF) {
                   1304:                struct sockaddr_dl sdl;
                   1305:
                   1306:                sdl.sdl_len = offsetof(struct sockaddr_dl, sdl_data[0]);
                   1307:                sdl.sdl_family = AF_LINK;
                   1308:                sdl.sdl_index = m->m_pkthdr.rcvif ?
                   1309:                    m->m_pkthdr.rcvif->if_index : 0;
                   1310:                sdl.sdl_nlen = sdl.sdl_alen = sdl.sdl_slen = 0;
                   1311:                *mp = sbcreatecontrol((caddr_t) &sdl, sdl.sdl_len,
                   1312:                    IP_RECVIF, IPPROTO_IP);
                   1313:                if (*mp)
                   1314:                        mp = &(*mp)->m_next;
                   1315:        }
1.13      mycroft  1316: }
                   1317:
                   1318: int
                   1319: ip_sysctl(name, namelen, oldp, oldlenp, newp, newlen)
                   1320:        int *name;
                   1321:        u_int namelen;
                   1322:        void *oldp;
                   1323:        size_t *oldlenp;
                   1324:        void *newp;
                   1325:        size_t newlen;
                   1326: {
1.52      thorpej  1327:        extern int subnetsarelocal;
                   1328:
1.54      lukem    1329:        int error, old;
                   1330:
1.13      mycroft  1331:        /* All sysctl names at this level are terminal. */
                   1332:        if (namelen != 1)
                   1333:                return (ENOTDIR);
                   1334:
                   1335:        switch (name[0]) {
                   1336:        case IPCTL_FORWARDING:
                   1337:                return (sysctl_int(oldp, oldlenp, newp, newlen, &ipforwarding));
                   1338:        case IPCTL_SENDREDIRECTS:
                   1339:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1340:                        &ipsendredirects));
                   1341:        case IPCTL_DEFTTL:
                   1342:                return (sysctl_int(oldp, oldlenp, newp, newlen, &ip_defttl));
                   1343: #ifdef notyet
                   1344:        case IPCTL_DEFMTU:
                   1345:                return (sysctl_int(oldp, oldlenp, newp, newlen, &ip_mtu));
                   1346: #endif
1.26      thorpej  1347:        case IPCTL_FORWSRCRT:
1.47      cjs      1348:                /* Don't allow this to change in a secure environment.  */
1.26      thorpej  1349:                if (securelevel > 0)
1.46      cjs      1350:                        return (sysctl_rdint(oldp, oldlenp, newp,
                   1351:                            ip_forwsrcrt));
                   1352:                else
                   1353:                        return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1354:                            &ip_forwsrcrt));
1.27      thorpej  1355:        case IPCTL_DIRECTEDBCAST:
                   1356:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1357:                    &ip_directedbcast));
1.47      cjs      1358:        case IPCTL_ALLOWSRCRT:
                   1359:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1360:                    &ip_allowsrcrt));
1.52      thorpej  1361:        case IPCTL_SUBNETSARELOCAL:
                   1362:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1363:                    &subnetsarelocal));
1.53      kml      1364:        case IPCTL_MTUDISC:
1.60      kml      1365:                error = sysctl_int(oldp, oldlenp, newp, newlen,
                   1366:                    &ip_mtudisc);
                   1367:                if (ip_mtudisc != 0 && ip_mtudisc_timeout_q == NULL) {
                   1368:                        ip_mtudisc_timeout_q =
                   1369:                            rt_timer_queue_create(ip_mtudisc_timeout);
                   1370:                } else if (ip_mtudisc == 0 && ip_mtudisc_timeout_q != NULL) {
                   1371:                        rt_timer_queue_destroy(ip_mtudisc_timeout_q, TRUE);
                   1372:                        ip_mtudisc_timeout_q = NULL;
                   1373:                }
                   1374:                return error;
1.54      lukem    1375:        case IPCTL_ANONPORTMIN:
                   1376:                old = anonportmin;
                   1377:                error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmin);
                   1378:                if (anonportmin >= anonportmax || anonportmin > 65535
                   1379: #ifndef IPNOPRIVPORTS
                   1380:                    || anonportmin < IPPORT_RESERVED
                   1381: #endif
                   1382:                    ) {
                   1383:                        anonportmin = old;
                   1384:                        return (EINVAL);
                   1385:                }
                   1386:                return (error);
                   1387:        case IPCTL_ANONPORTMAX:
                   1388:                old = anonportmax;
                   1389:                error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmax);
                   1390:                if (anonportmin >= anonportmax || anonportmax > 65535
                   1391: #ifndef IPNOPRIVPORTS
                   1392:                    || anonportmax < IPPORT_RESERVED
                   1393: #endif
                   1394:                    ) {
                   1395:                        anonportmax = old;
                   1396:                        return (EINVAL);
                   1397:                }
1.60      kml      1398:                return (error);
                   1399:        case IPCTL_MTUDISCTIMEOUT:
                   1400:                error = sysctl_int(oldp, oldlenp, newp, newlen,
                   1401:                   &ip_mtudisc_timeout);
                   1402:                if (ip_mtudisc_timeout_q != NULL)
                   1403:                        rt_timer_queue_change(ip_mtudisc_timeout_q,
                   1404:                                              ip_mtudisc_timeout);
1.54      lukem    1405:                return (error);
1.65      matt     1406: #ifdef GATEWAY
                   1407:        case IPCTL_MAXFLOWS:
1.67      thorpej  1408:            {
                   1409:                int s;
                   1410:
1.65      matt     1411:                error = sysctl_int(oldp, oldlenp, newp, newlen,
                   1412:                   &ip_maxflows);
1.67      thorpej  1413:                s = splsoftnet();
1.65      matt     1414:                ipflow_reap(0);
1.67      thorpej  1415:                splx(s);
1.65      matt     1416:                return (error);
1.67      thorpej  1417:            }
1.65      matt     1418: #endif
1.13      mycroft  1419:        default:
                   1420:                return (EOPNOTSUPP);
                   1421:        }
                   1422:        /* NOTREACHED */
1.1       cgd      1423: }

CVSweb <webmaster@jp.NetBSD.org>