[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 1.165

1.165   ! christos    1: /*     $NetBSD: ip_input.c,v 1.164 2003/02/26 06:31:14 matt Exp $      */
1.89      itojun      2:
                      3: /*
                      4:  * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
                      5:  * All rights reserved.
1.152     itojun      6:  *
1.89      itojun      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. Neither the name of the project nor the names of its contributors
                     16:  *    may be used to endorse or promote products derived from this software
                     17:  *    without specific prior written permission.
1.152     itojun     18:  *
1.89      itojun     19:  * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
                     20:  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
                     21:  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
                     22:  * ARE DISCLAIMED.  IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
                     23:  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
                     24:  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
                     25:  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
                     26:  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
                     27:  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
                     28:  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
                     29:  * SUCH DAMAGE.
                     30:  */
1.76      thorpej    31:
                     32: /*-
                     33:  * Copyright (c) 1998 The NetBSD Foundation, Inc.
                     34:  * All rights reserved.
                     35:  *
                     36:  * This code is derived from software contributed to The NetBSD Foundation
                     37:  * by Public Access Networks Corporation ("Panix").  It was developed under
                     38:  * contract to Panix by Eric Haszlakiewicz and Thor Lancelot Simon.
                     39:  *
                     40:  * Redistribution and use in source and binary forms, with or without
                     41:  * modification, are permitted provided that the following conditions
                     42:  * are met:
                     43:  * 1. Redistributions of source code must retain the above copyright
                     44:  *    notice, this list of conditions and the following disclaimer.
                     45:  * 2. Redistributions in binary form must reproduce the above copyright
                     46:  *    notice, this list of conditions and the following disclaimer in the
                     47:  *    documentation and/or other materials provided with the distribution.
                     48:  * 3. All advertising materials mentioning features or use of this software
                     49:  *    must display the following acknowledgement:
                     50:  *     This product includes software developed by the NetBSD
                     51:  *     Foundation, Inc. and its contributors.
                     52:  * 4. Neither the name of The NetBSD Foundation nor the names of its
                     53:  *    contributors may be used to endorse or promote products derived
                     54:  *    from this software without specific prior written permission.
                     55:  *
                     56:  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
                     57:  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
                     58:  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
                     59:  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
                     60:  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
                     61:  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
                     62:  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
                     63:  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
                     64:  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
                     65:  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
                     66:  * POSSIBILITY OF SUCH DAMAGE.
                     67:  */
1.14      cgd        68:
1.1       cgd        69: /*
1.13      mycroft    70:  * Copyright (c) 1982, 1986, 1988, 1993
                     71:  *     The Regents of the University of California.  All rights reserved.
1.1       cgd        72:  *
                     73:  * Redistribution and use in source and binary forms, with or without
                     74:  * modification, are permitted provided that the following conditions
                     75:  * are met:
                     76:  * 1. Redistributions of source code must retain the above copyright
                     77:  *    notice, this list of conditions and the following disclaimer.
                     78:  * 2. Redistributions in binary form must reproduce the above copyright
                     79:  *    notice, this list of conditions and the following disclaimer in the
                     80:  *    documentation and/or other materials provided with the distribution.
                     81:  * 3. All advertising materials mentioning features or use of this software
                     82:  *    must display the following acknowledgement:
                     83:  *     This product includes software developed by the University of
                     84:  *     California, Berkeley and its contributors.
                     85:  * 4. Neither the name of the University nor the names of its contributors
                     86:  *    may be used to endorse or promote products derived from this software
                     87:  *    without specific prior written permission.
                     88:  *
                     89:  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
                     90:  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
                     91:  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
                     92:  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
                     93:  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
                     94:  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
                     95:  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
                     96:  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
                     97:  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
                     98:  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
                     99:  * SUCH DAMAGE.
                    100:  *
1.14      cgd       101:  *     @(#)ip_input.c  8.2 (Berkeley) 1/4/94
1.1       cgd       102:  */
1.141     lukem     103:
                    104: #include <sys/cdefs.h>
1.165   ! christos  105: __KERNEL_RCSID(0, "$NetBSD: ip_input.c,v 1.164 2003/02/26 06:31:14 matt Exp $");
1.55      scottr    106:
1.62      matt      107: #include "opt_gateway.h"
1.69      mrg       108: #include "opt_pfil_hooks.h"
1.91      thorpej   109: #include "opt_ipsec.h"
1.55      scottr    110: #include "opt_mrouting.h"
1.135     thorpej   111: #include "opt_inet_csum.h"
1.1       cgd       112:
1.5       mycroft   113: #include <sys/param.h>
                    114: #include <sys/systm.h>
                    115: #include <sys/malloc.h>
                    116: #include <sys/mbuf.h>
                    117: #include <sys/domain.h>
                    118: #include <sys/protosw.h>
                    119: #include <sys/socket.h>
1.44      thorpej   120: #include <sys/socketvar.h>
1.5       mycroft   121: #include <sys/errno.h>
                    122: #include <sys/time.h>
                    123: #include <sys/kernel.h>
1.72      thorpej   124: #include <sys/pool.h>
1.28      christos  125: #include <sys/sysctl.h>
1.1       cgd       126:
1.5       mycroft   127: #include <net/if.h>
1.44      thorpej   128: #include <net/if_dl.h>
1.5       mycroft   129: #include <net/route.h>
1.45      mrg       130: #include <net/pfil.h>
1.1       cgd       131:
1.5       mycroft   132: #include <netinet/in.h>
                    133: #include <netinet/in_systm.h>
                    134: #include <netinet/ip.h>
                    135: #include <netinet/in_pcb.h>
                    136: #include <netinet/in_var.h>
                    137: #include <netinet/ip_var.h>
                    138: #include <netinet/ip_icmp.h>
1.89      itojun    139: /* just for gif_ttl */
                    140: #include <netinet/in_gif.h>
                    141: #include "gif.h"
1.144     martin    142: #include <net/if_gre.h>
                    143: #include "gre.h"
1.111     jdolecek  144:
                    145: #ifdef MROUTING
                    146: #include <netinet/ip_mroute.h>
                    147: #endif
1.89      itojun    148:
                    149: #ifdef IPSEC
                    150: #include <netinet6/ipsec.h>
                    151: #include <netkey/key.h>
                    152: #endif
1.44      thorpej   153:
1.1       cgd       154: #ifndef        IPFORWARDING
                    155: #ifdef GATEWAY
                    156: #define        IPFORWARDING    1       /* forward IP packets not for us */
                    157: #else /* GATEWAY */
                    158: #define        IPFORWARDING    0       /* don't forward IP packets not for us */
                    159: #endif /* GATEWAY */
                    160: #endif /* IPFORWARDING */
                    161: #ifndef        IPSENDREDIRECTS
                    162: #define        IPSENDREDIRECTS 1
                    163: #endif
1.26      thorpej   164: #ifndef IPFORWSRCRT
1.47      cjs       165: #define        IPFORWSRCRT     1       /* forward source-routed packets */
                    166: #endif
                    167: #ifndef IPALLOWSRCRT
1.48      mrg       168: #define        IPALLOWSRCRT    1       /* allow source-routed packets */
1.26      thorpej   169: #endif
1.53      kml       170: #ifndef IPMTUDISC
1.153     itojun    171: #define IPMTUDISC      1
1.53      kml       172: #endif
1.60      kml       173: #ifndef IPMTUDISCTIMEOUT
1.61      kml       174: #define IPMTUDISCTIMEOUT (10 * 60)     /* as per RFC 1191 */
1.60      kml       175: #endif
1.53      kml       176:
1.27      thorpej   177: /*
                    178:  * Note: DIRECTED_BROADCAST is handled this way so that previous
                    179:  * configuration using this option will Just Work.
                    180:  */
                    181: #ifndef IPDIRECTEDBCAST
                    182: #ifdef DIRECTED_BROADCAST
                    183: #define IPDIRECTEDBCAST        1
                    184: #else
                    185: #define        IPDIRECTEDBCAST 0
                    186: #endif /* DIRECTED_BROADCAST */
                    187: #endif /* IPDIRECTEDBCAST */
1.1       cgd       188: int    ipforwarding = IPFORWARDING;
                    189: int    ipsendredirects = IPSENDREDIRECTS;
1.13      mycroft   190: int    ip_defttl = IPDEFTTL;
1.26      thorpej   191: int    ip_forwsrcrt = IPFORWSRCRT;
1.27      thorpej   192: int    ip_directedbcast = IPDIRECTEDBCAST;
1.47      cjs       193: int    ip_allowsrcrt = IPALLOWSRCRT;
1.53      kml       194: int    ip_mtudisc = IPMTUDISC;
1.156     itojun    195: int    ip_mtudisc_timeout = IPMTUDISCTIMEOUT;
1.1       cgd       196: #ifdef DIAGNOSTIC
                    197: int    ipprintfs = 0;
                    198: #endif
1.165   ! christos  199: /*
        !           200:  * XXX - Setting ip_checkinterface mostly implements the receive side of
        !           201:  * the Strong ES model described in RFC 1122, but since the routing table
        !           202:  * and transmit implementation do not implement the Strong ES model,
        !           203:  * setting this to 1 results in an odd hybrid.
        !           204:  *
        !           205:  * XXX - ip_checkinterface currently must be disabled if you use ipnat
        !           206:  * to translate the destination address to another local interface.
        !           207:  *
        !           208:  * XXX - ip_checkinterface must be disabled if you add IP aliases
        !           209:  * to the loopback interface instead of the interface where the
        !           210:  * packets for those addresses are received.
        !           211:  */
        !           212: int    ip_checkinterface = 0;
        !           213:
1.1       cgd       214:
1.60      kml       215: struct rttimer_queue *ip_mtudisc_timeout_q = NULL;
                    216:
1.1       cgd       217: extern struct domain inetdomain;
                    218: int    ipqmaxlen = IFQ_MAXLEN;
1.150     matt      219: u_long in_ifaddrhash;                          /* size of hash table - 1 */
                    220: int    in_ifaddrentries;                       /* total number of addrs */
1.22      mycroft   221: struct in_ifaddrhead in_ifaddr;
1.57      tls       222: struct in_ifaddrhashhead *in_ifaddrhashtbl;
1.13      mycroft   223: struct ifqueue ipintrq;
1.63      matt      224: struct ipstat  ipstat;
                    225: u_int16_t      ip_id;
1.75      thorpej   226:
1.121     thorpej   227: #ifdef PFIL_HOOKS
                    228: struct pfil_head inet_pfil_hook;
                    229: #endif
                    230:
1.63      matt      231: struct ipqhead ipq;
1.75      thorpej   232: int    ipq_locked;
1.131     itojun    233: int    ip_nfragpackets = 0;
1.133     itojun    234: int    ip_maxfragpackets = 200;
1.75      thorpej   235:
                    236: static __inline int ipq_lock_try __P((void));
                    237: static __inline void ipq_unlock __P((void));
                    238:
                    239: static __inline int
                    240: ipq_lock_try()
                    241: {
                    242:        int s;
                    243:
1.132     thorpej   244:        /*
1.149     wiz       245:         * Use splvm() -- we're blocking things that would cause
1.132     thorpej   246:         * mbuf allocation.
                    247:         */
                    248:        s = splvm();
1.75      thorpej   249:        if (ipq_locked) {
                    250:                splx(s);
                    251:                return (0);
                    252:        }
                    253:        ipq_locked = 1;
                    254:        splx(s);
                    255:        return (1);
                    256: }
                    257:
                    258: static __inline void
                    259: ipq_unlock()
                    260: {
                    261:        int s;
                    262:
1.132     thorpej   263:        s = splvm();
1.75      thorpej   264:        ipq_locked = 0;
                    265:        splx(s);
                    266: }
                    267:
                    268: #ifdef DIAGNOSTIC
                    269: #define        IPQ_LOCK()                                                      \
                    270: do {                                                                   \
                    271:        if (ipq_lock_try() == 0) {                                      \
                    272:                printf("%s:%d: ipq already locked\n", __FILE__, __LINE__); \
                    273:                panic("ipq_lock");                                      \
                    274:        }                                                               \
1.159     perry     275: } while (/*CONSTCOND*/ 0)
1.75      thorpej   276: #define        IPQ_LOCK_CHECK()                                                \
                    277: do {                                                                   \
                    278:        if (ipq_locked == 0) {                                          \
                    279:                printf("%s:%d: ipq lock not held\n", __FILE__, __LINE__); \
                    280:                panic("ipq lock check");                                \
                    281:        }                                                               \
1.159     perry     282: } while (/*CONSTCOND*/ 0)
1.75      thorpej   283: #else
                    284: #define        IPQ_LOCK()              (void) ipq_lock_try()
                    285: #define        IPQ_LOCK_CHECK()        /* nothing */
                    286: #endif
                    287:
                    288: #define        IPQ_UNLOCK()            ipq_unlock()
1.1       cgd       289:
1.72      thorpej   290: struct pool ipqent_pool;
                    291:
1.135     thorpej   292: #ifdef INET_CSUM_COUNTERS
                    293: #include <sys/device.h>
                    294:
                    295: struct evcnt ip_hwcsum_bad = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
                    296:     NULL, "inet", "hwcsum bad");
                    297: struct evcnt ip_hwcsum_ok = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
                    298:     NULL, "inet", "hwcsum ok");
                    299: struct evcnt ip_swcsum = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
                    300:     NULL, "inet", "swcsum");
                    301:
                    302: #define        INET_CSUM_COUNTER_INCR(ev)      (ev)->ev_count++
                    303:
                    304: #else
                    305:
                    306: #define        INET_CSUM_COUNTER_INCR(ev)      /* nothing */
                    307:
                    308: #endif /* INET_CSUM_COUNTERS */
                    309:
1.1       cgd       310: /*
                    311:  * We need to save the IP options in case a protocol wants to respond
                    312:  * to an incoming packet over the same route if the packet got here
                    313:  * using IP source routing.  This allows connection establishment and
                    314:  * maintenance when the remote end is on a network that is not known
                    315:  * to us.
                    316:  */
                    317: int    ip_nhops = 0;
                    318: static struct ip_srcrt {
                    319:        struct  in_addr dst;                    /* final destination */
                    320:        char    nop;                            /* one NOP to align */
                    321:        char    srcopt[IPOPT_OFFSET + 1];       /* OPTVAL, OLEN and OFFSET */
                    322:        struct  in_addr route[MAX_IPOPTLEN/sizeof(struct in_addr)];
                    323: } ip_srcrt;
                    324:
1.13      mycroft   325: static void save_rte __P((u_char *, struct in_addr));
1.35      mycroft   326:
1.164     matt      327: #ifdef MBUFTRACE
                    328: struct mowner ip_rx_mowner = { "internet", "rx" };
                    329: struct mowner ip_tx_mowner = { "internet", "tx" };
                    330: #endif
                    331:
1.1       cgd       332: /*
                    333:  * IP initialization: fill in IP protocol switch table.
                    334:  * All protocols not implemented in kernel go to raw IP protocol handler.
                    335:  */
1.8       mycroft   336: void
1.1       cgd       337: ip_init()
                    338: {
1.109     augustss  339:        struct protosw *pr;
                    340:        int i;
1.1       cgd       341:
1.72      thorpej   342:        pool_init(&ipqent_pool, sizeof(struct ipqent), 0, 0, 0, "ipqepl",
1.146     thorpej   343:            NULL);
1.72      thorpej   344:
1.1       cgd       345:        pr = pffindproto(PF_INET, IPPROTO_RAW, SOCK_RAW);
                    346:        if (pr == 0)
                    347:                panic("ip_init");
                    348:        for (i = 0; i < IPPROTO_MAX; i++)
                    349:                ip_protox[i] = pr - inetsw;
                    350:        for (pr = inetdomain.dom_protosw;
                    351:            pr < inetdomain.dom_protoswNPROTOSW; pr++)
                    352:                if (pr->pr_domain->dom_family == PF_INET &&
                    353:                    pr->pr_protocol && pr->pr_protocol != IPPROTO_RAW)
                    354:                        ip_protox[pr->pr_protocol] = pr - inetsw;
1.25      cgd       355:        LIST_INIT(&ipq);
1.1       cgd       356:        ip_id = time.tv_sec & 0xffff;
                    357:        ipintrq.ifq_maxlen = ipqmaxlen;
1.22      mycroft   358:        TAILQ_INIT(&in_ifaddr);
1.120     ad        359:        in_ifaddrhashtbl = hashinit(IN_IFADDR_HASH_SIZE, HASH_LIST, M_IFADDR,
                    360:            M_WAITOK, &in_ifaddrhash);
1.160     itojun    361:        ip_mtudisc_timeout_q = rt_timer_queue_create(ip_mtudisc_timeout);
1.73      thorpej   362: #ifdef GATEWAY
                    363:        ipflow_init();
                    364: #endif
1.121     thorpej   365:
                    366: #ifdef PFIL_HOOKS
                    367:        /* Register our Packet Filter hook. */
1.126     thorpej   368:        inet_pfil_hook.ph_type = PFIL_TYPE_AF;
                    369:        inet_pfil_hook.ph_af   = AF_INET;
1.121     thorpej   370:        i = pfil_head_register(&inet_pfil_hook);
                    371:        if (i != 0)
                    372:                printf("ip_init: WARNING: unable to register pfil hook, "
                    373:                    "error %d\n", i);
                    374: #endif /* PFIL_HOOKS */
1.135     thorpej   375:
                    376: #ifdef INET_CSUM_COUNTERS
                    377:        evcnt_attach_static(&ip_hwcsum_bad);
                    378:        evcnt_attach_static(&ip_hwcsum_ok);
                    379:        evcnt_attach_static(&ip_swcsum);
                    380: #endif /* INET_CSUM_COUNTERS */
1.164     matt      381:
                    382: #ifdef MBUFTRACE
                    383:        MOWNER_ATTACH(&ip_tx_mowner);
                    384:        MOWNER_ATTACH(&ip_rx_mowner);
                    385: #endif /* MBUFTRACE */
1.1       cgd       386: }
                    387:
                    388: struct sockaddr_in ipaddr = { sizeof(ipaddr), AF_INET };
                    389: struct route ipforward_rt;
                    390:
                    391: /*
1.89      itojun    392:  * IP software interrupt routine
                    393:  */
                    394: void
                    395: ipintr()
                    396: {
                    397:        int s;
                    398:        struct mbuf *m;
                    399:
                    400:        while (1) {
1.132     thorpej   401:                s = splnet();
1.89      itojun    402:                IF_DEQUEUE(&ipintrq, m);
                    403:                splx(s);
                    404:                if (m == 0)
                    405:                        return;
1.164     matt      406:                MCLAIM(m, &ip_rx_mowner);
1.89      itojun    407:                ip_input(m);
                    408:        }
                    409: }
                    410:
                    411: /*
1.1       cgd       412:  * Ip input routine.  Checksum and byte swap header.  If fragmented
                    413:  * try to reassemble.  Process options.  Pass to next level.
                    414:  */
1.8       mycroft   415: void
1.89      itojun    416: ip_input(struct mbuf *m)
1.1       cgd       417: {
1.109     augustss  418:        struct ip *ip = NULL;
                    419:        struct ipq *fp;
                    420:        struct in_ifaddr *ia;
                    421:        struct ifaddr *ifa;
1.25      cgd       422:        struct ipqent *ipqe;
1.89      itojun    423:        int hlen = 0, mff, len;
1.100     itojun    424:        int downmatch;
1.165   ! christos  425:        int checkif;
1.1       cgd       426:
1.164     matt      427:        MCLAIM(m, &ip_rx_mowner);
1.1       cgd       428: #ifdef DIAGNOSTIC
                    429:        if ((m->m_flags & M_PKTHDR) == 0)
                    430:                panic("ipintr no HDR");
                    431: #endif
1.89      itojun    432: #ifdef IPSEC
                    433:        /*
                    434:         * should the inner packet be considered authentic?
                    435:         * see comment in ah4_input().
                    436:         */
                    437:        if (m) {
                    438:                m->m_flags &= ~M_AUTHIPHDR;
                    439:                m->m_flags &= ~M_AUTHIPDGM;
                    440:        }
                    441: #endif
1.164     matt      442:
1.1       cgd       443:        /*
                    444:         * If no IP addresses have been set yet but the interfaces
                    445:         * are receiving, can't do anything with incoming packets yet.
                    446:         */
1.140     matt      447:        if (TAILQ_FIRST(&in_ifaddr) == 0)
1.1       cgd       448:                goto bad;
                    449:        ipstat.ips_total++;
1.154     thorpej   450:        /*
                    451:         * If the IP header is not aligned, slurp it up into a new
                    452:         * mbuf with space for link headers, in the event we forward
                    453:         * it.  Otherwise, if it is aligned, make sure the entire
                    454:         * base IP header is in the first mbuf of the chain.
                    455:         */
                    456:        if (IP_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) {
                    457:                if ((m = m_copyup(m, sizeof(struct ip),
                    458:                                  (max_linkhdr + 3) & ~3)) == NULL) {
                    459:                        /* XXXJRT new stat, please */
                    460:                        ipstat.ips_toosmall++;
                    461:                        return;
                    462:                }
                    463:        } else if (__predict_false(m->m_len < sizeof (struct ip))) {
                    464:                if ((m = m_pullup(m, sizeof (struct ip))) == NULL) {
                    465:                        ipstat.ips_toosmall++;
                    466:                        return;
                    467:                }
1.1       cgd       468:        }
                    469:        ip = mtod(m, struct ip *);
1.13      mycroft   470:        if (ip->ip_v != IPVERSION) {
                    471:                ipstat.ips_badvers++;
                    472:                goto bad;
                    473:        }
1.1       cgd       474:        hlen = ip->ip_hl << 2;
                    475:        if (hlen < sizeof(struct ip)) { /* minimum header length */
                    476:                ipstat.ips_badhlen++;
                    477:                goto bad;
                    478:        }
                    479:        if (hlen > m->m_len) {
                    480:                if ((m = m_pullup(m, hlen)) == 0) {
                    481:                        ipstat.ips_badhlen++;
1.89      itojun    482:                        return;
1.1       cgd       483:                }
                    484:                ip = mtod(m, struct ip *);
                    485:        }
1.98      thorpej   486:
1.85      hwr       487:        /*
1.99      thorpej   488:         * RFC1122: packets with a multicast source address are
1.98      thorpej   489:         * not allowed.
1.85      hwr       490:         */
                    491:        if (IN_MULTICAST(ip->ip_src.s_addr)) {
1.130     itojun    492:                ipstat.ips_badaddr++;
1.85      hwr       493:                goto bad;
1.129     itojun    494:        }
                    495:
                    496:        /* 127/8 must not appear on wire - RFC1122 */
                    497:        if ((ntohl(ip->ip_dst.s_addr) >> IN_CLASSA_NSHIFT) == IN_LOOPBACKNET ||
                    498:            (ntohl(ip->ip_src.s_addr) >> IN_CLASSA_NSHIFT) == IN_LOOPBACKNET) {
1.130     itojun    499:                if ((m->m_pkthdr.rcvif->if_flags & IFF_LOOPBACK) == 0) {
                    500:                        ipstat.ips_badaddr++;
1.129     itojun    501:                        goto bad;
1.130     itojun    502:                }
1.85      hwr       503:        }
                    504:
1.135     thorpej   505:        switch (m->m_pkthdr.csum_flags &
1.137     thorpej   506:                ((m->m_pkthdr.rcvif->if_csum_flags_rx & M_CSUM_IPv4) |
1.135     thorpej   507:                 M_CSUM_IPv4_BAD)) {
                    508:        case M_CSUM_IPv4|M_CSUM_IPv4_BAD:
                    509:                INET_CSUM_COUNTER_INCR(&ip_hwcsum_bad);
                    510:                goto badcsum;
                    511:
                    512:        case M_CSUM_IPv4:
                    513:                /* Checksum was okay. */
                    514:                INET_CSUM_COUNTER_INCR(&ip_hwcsum_ok);
                    515:                break;
                    516:
                    517:        default:
                    518:                /* Must compute it ourselves. */
                    519:                INET_CSUM_COUNTER_INCR(&ip_swcsum);
                    520:                if (in_cksum(m, hlen) != 0)
                    521:                        goto bad;
                    522:                break;
1.1       cgd       523:        }
                    524:
1.121     thorpej   525:        /* Retrieve the packet length. */
                    526:        len = ntohs(ip->ip_len);
1.81      proff     527:
                    528:        /*
                    529:         * Check for additional length bogosity
                    530:         */
1.84      proff     531:        if (len < hlen) {
1.81      proff     532:                ipstat.ips_badlen++;
                    533:                goto bad;
                    534:        }
1.1       cgd       535:
                    536:        /*
                    537:         * Check that the amount of data in the buffers
                    538:         * is as at least much as the IP header would have us expect.
                    539:         * Trim mbufs if longer than we expect.
                    540:         * Drop packet if shorter than we expect.
                    541:         */
1.35      mycroft   542:        if (m->m_pkthdr.len < len) {
1.1       cgd       543:                ipstat.ips_tooshort++;
                    544:                goto bad;
                    545:        }
1.35      mycroft   546:        if (m->m_pkthdr.len > len) {
1.1       cgd       547:                if (m->m_len == m->m_pkthdr.len) {
1.35      mycroft   548:                        m->m_len = len;
                    549:                        m->m_pkthdr.len = len;
1.1       cgd       550:                } else
1.35      mycroft   551:                        m_adj(m, len - m->m_pkthdr.len);
1.1       cgd       552:        }
                    553:
1.94      itojun    554: #ifdef IPSEC
1.149     wiz       555:        /* ipflow (IP fast forwarding) is not compatible with IPsec. */
1.94      itojun    556:        m->m_flags &= ~M_CANFASTFWD;
                    557: #else
1.64      thorpej   558:        /*
                    559:         * Assume that we can create a fast-forward IP flow entry
                    560:         * based on this packet.
                    561:         */
                    562:        m->m_flags |= M_CANFASTFWD;
1.94      itojun    563: #endif
1.64      thorpej   564:
1.36      mrg       565: #ifdef PFIL_HOOKS
1.33      mrg       566:        /*
1.64      thorpej   567:         * Run through list of hooks for input packets.  If there are any
                    568:         * filters which require that additional packets in the flow are
                    569:         * not fast-forwarded, they must clear the M_CANFASTFWD flag.
                    570:         * Note that filters must _never_ set this flag, as another filter
                    571:         * in the list may have previously cleared it.
1.33      mrg       572:         */
1.127     itojun    573:        /*
                    574:         * let ipfilter look at packet on the wire,
                    575:         * not the decapsulated packet.
                    576:         */
                    577: #ifdef IPSEC
1.136     itojun    578:        if (!ipsec_getnhist(m))
1.127     itojun    579: #else
                    580:        if (1)
                    581: #endif
                    582:        {
                    583:                if (pfil_run_hooks(&inet_pfil_hook, &m, m->m_pkthdr.rcvif,
                    584:                                   PFIL_IN) != 0)
1.121     thorpej   585:                return;
1.127     itojun    586:                if (m == NULL)
                    587:                        return;
                    588:                ip = mtod(m, struct ip *);
1.142     darrenr   589:                hlen = ip->ip_hl << 2;
1.127     itojun    590:        }
1.36      mrg       591: #endif /* PFIL_HOOKS */
1.123     thorpej   592:
                    593: #ifdef ALTQ
                    594:        /* XXX Temporary until ALTQ is changed to use a pfil hook */
                    595:        if (altq_input != NULL && (*altq_input)(m, AF_INET) == 0) {
                    596:                /* packet dropped by traffic conditioner */
                    597:                return;
                    598:        }
                    599: #endif
1.121     thorpej   600:
                    601:        /*
1.1       cgd       602:         * Process options and, if not destined for us,
                    603:         * ship it on.  ip_dooptions returns 1 when an
                    604:         * error was detected (causing an icmp message
                    605:         * to be sent and the original packet to be freed).
                    606:         */
                    607:        ip_nhops = 0;           /* for source routed packets */
                    608:        if (hlen > sizeof (struct ip) && ip_dooptions(m))
1.89      itojun    609:                return;
1.1       cgd       610:
                    611:        /*
1.165   ! christos  612:         * Enable a consistency check between the destination address
        !           613:         * and the arrival interface for a unicast packet (the RFC 1122
        !           614:         * strong ES model) if IP forwarding is disabled and the packet
        !           615:         * is not locally generated.
        !           616:         *
        !           617:         * XXX - Checking also should be disabled if the destination
        !           618:         * address is ipnat'ed to a different interface.
        !           619:         *
        !           620:         * XXX - Checking is incompatible with IP aliases added
        !           621:         * to the loopback interface instead of the interface where
        !           622:         * the packets are received.
        !           623:         *
        !           624:         * XXX - We need to add a per ifaddr flag for this so that
        !           625:         * we get finer grain control.
        !           626:         */
        !           627:        checkif = ip_checkinterface && (ipforwarding == 0) &&
        !           628:            (m->m_pkthdr.rcvif != NULL) &&
        !           629:            ((m->m_pkthdr.rcvif->if_flags & IFF_LOOPBACK) == 0);
        !           630:
        !           631:        /*
1.1       cgd       632:         * Check our list of addresses, to see if the packet is for us.
1.100     itojun    633:         *
                    634:         * Traditional 4.4BSD did not consult IFF_UP at all.
                    635:         * The behavior here is to treat addresses on !IFF_UP interface
                    636:         * as not mine.
1.1       cgd       637:         */
1.100     itojun    638:        downmatch = 0;
1.140     matt      639:        LIST_FOREACH(ia, &IN_IFADDR_HASH(ip->ip_dst.s_addr), ia_hash) {
1.97      itojun    640:                if (in_hosteq(ia->ia_addr.sin_addr, ip->ip_dst)) {
1.165   ! christos  641:                        if (checkif && ia->ia_ifp != m->m_pkthdr.rcvif)
        !           642:                                continue;
1.97      itojun    643:                        if ((ia->ia_ifp->if_flags & IFF_UP) != 0)
                    644:                                break;
1.100     itojun    645:                        else
                    646:                                downmatch++;
1.97      itojun    647:                }
                    648:        }
1.86      thorpej   649:        if (ia != NULL)
                    650:                goto ours;
1.57      tls       651:        if (m->m_pkthdr.rcvif->if_flags & IFF_BROADCAST) {
1.140     matt      652:                TAILQ_FOREACH(ifa, &m->m_pkthdr.rcvif->if_addrlist, ifa_list) {
                    653:                        if (ifa->ifa_addr->sa_family != AF_INET)
                    654:                                continue;
1.57      tls       655:                        ia = ifatoia(ifa);
1.35      mycroft   656:                        if (in_hosteq(ip->ip_dst, ia->ia_broadaddr.sin_addr) ||
                    657:                            in_hosteq(ip->ip_dst, ia->ia_netbroadcast) ||
1.20      mycroft   658:                            /*
                    659:                             * Look for all-0's host part (old broadcast addr),
                    660:                             * either for subnet or net.
                    661:                             */
                    662:                            ip->ip_dst.s_addr == ia->ia_subnet ||
1.18      mycroft   663:                            ip->ip_dst.s_addr == ia->ia_net)
1.1       cgd       664:                                goto ours;
1.57      tls       665:                        /*
                    666:                         * An interface with IP address zero accepts
                    667:                         * all packets that arrive on that interface.
                    668:                         */
                    669:                        if (in_nullhost(ia->ia_addr.sin_addr))
                    670:                                goto ours;
1.1       cgd       671:                }
                    672:        }
1.18      mycroft   673:        if (IN_MULTICAST(ip->ip_dst.s_addr)) {
1.4       hpeyerl   674:                struct in_multi *inm;
                    675: #ifdef MROUTING
                    676:                extern struct socket *ip_mrouter;
1.10      brezak    677:
1.147     matt      678:                if (M_READONLY(m)) {
1.10      brezak    679:                        if ((m = m_pullup(m, hlen)) == 0) {
                    680:                                ipstat.ips_toosmall++;
1.89      itojun    681:                                return;
1.10      brezak    682:                        }
                    683:                        ip = mtod(m, struct ip *);
                    684:                }
1.4       hpeyerl   685:
                    686:                if (ip_mrouter) {
                    687:                        /*
                    688:                         * If we are acting as a multicast router, all
                    689:                         * incoming multicast packets are passed to the
                    690:                         * kernel-level multicast forwarding function.
                    691:                         * The packet is returned (relatively) intact; if
                    692:                         * ip_mforward() returns a non-zero value, the packet
                    693:                         * must be discarded, else it may be accepted below.
                    694:                         *
                    695:                         * (The IP ident field is put in the same byte order
                    696:                         * as expected when ip_mforward() is called from
                    697:                         * ip_output().)
                    698:                         */
1.13      mycroft   699:                        if (ip_mforward(m, m->m_pkthdr.rcvif) != 0) {
                    700:                                ipstat.ips_cantforward++;
1.4       hpeyerl   701:                                m_freem(m);
1.89      itojun    702:                                return;
1.4       hpeyerl   703:                        }
                    704:
                    705:                        /*
                    706:                         * The process-level routing demon needs to receive
                    707:                         * all multicast IGMP packets, whether or not this
                    708:                         * host belongs to their destination groups.
                    709:                         */
                    710:                        if (ip->ip_p == IPPROTO_IGMP)
                    711:                                goto ours;
1.13      mycroft   712:                        ipstat.ips_forward++;
1.4       hpeyerl   713:                }
                    714: #endif
                    715:                /*
                    716:                 * See if we belong to the destination multicast group on the
                    717:                 * arrival interface.
                    718:                 */
                    719:                IN_LOOKUP_MULTI(ip->ip_dst, m->m_pkthdr.rcvif, inm);
                    720:                if (inm == NULL) {
1.13      mycroft   721:                        ipstat.ips_cantforward++;
1.4       hpeyerl   722:                        m_freem(m);
1.89      itojun    723:                        return;
1.4       hpeyerl   724:                }
                    725:                goto ours;
                    726:        }
1.19      mycroft   727:        if (ip->ip_dst.s_addr == INADDR_BROADCAST ||
1.35      mycroft   728:            in_nullhost(ip->ip_dst))
1.1       cgd       729:                goto ours;
                    730:
                    731:        /*
                    732:         * Not for us; forward if possible and desirable.
                    733:         */
                    734:        if (ipforwarding == 0) {
                    735:                ipstat.ips_cantforward++;
                    736:                m_freem(m);
1.100     itojun    737:        } else {
                    738:                /*
                    739:                 * If ip_dst matched any of my address on !IFF_UP interface,
                    740:                 * and there's no IFF_UP interface that matches ip_dst,
                    741:                 * send icmp unreach.  Forwarding it will result in in-kernel
                    742:                 * forwarding loop till TTL goes to 0.
                    743:                 */
                    744:                if (downmatch) {
                    745:                        icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_HOST, 0, 0);
                    746:                        ipstat.ips_cantforward++;
                    747:                        return;
                    748:                }
1.145     itojun    749: #ifdef IPSEC
                    750:                if (ipsec4_in_reject(m, NULL)) {
                    751:                        ipsecstat.in_polvio++;
                    752:                        goto bad;
                    753:                }
                    754: #endif
                    755:
1.1       cgd       756:                ip_forward(m, 0);
1.100     itojun    757:        }
1.89      itojun    758:        return;
1.1       cgd       759:
                    760: ours:
                    761:        /*
                    762:         * If offset or IP_MF are set, must reassemble.
                    763:         * Otherwise, nothing need be done.
                    764:         * (We could look in the reassembly queue to see
                    765:         * if the packet was previously fragmented,
                    766:         * but it's not worth the time; just let them time out.)
                    767:         */
1.155     itojun    768:        if (ip->ip_off & ~htons(IP_DF|IP_RF)) {
                    769:                if (M_READONLY(m)) {
                    770:                        if ((m = m_pullup(m, hlen)) == NULL) {
                    771:                                ipstat.ips_toosmall++;
                    772:                                goto bad;
                    773:                        }
                    774:                        ip = mtod(m, struct ip *);
                    775:                }
                    776:
1.1       cgd       777:                /*
                    778:                 * Look for queue of fragments
                    779:                 * of this datagram.
                    780:                 */
1.75      thorpej   781:                IPQ_LOCK();
1.140     matt      782:                LIST_FOREACH(fp, &ipq, ipq_q)
1.1       cgd       783:                        if (ip->ip_id == fp->ipq_id &&
1.35      mycroft   784:                            in_hosteq(ip->ip_src, fp->ipq_src) &&
                    785:                            in_hosteq(ip->ip_dst, fp->ipq_dst) &&
1.1       cgd       786:                            ip->ip_p == fp->ipq_p)
                    787:                                goto found;
                    788:                fp = 0;
                    789: found:
                    790:
                    791:                /*
                    792:                 * Adjust ip_len to not reflect header,
1.25      cgd       793:                 * set ipqe_mff if more fragments are expected,
1.1       cgd       794:                 * convert offset of this to bytes.
                    795:                 */
1.155     itojun    796:                ip->ip_len = htons(ntohs(ip->ip_len) - hlen);
                    797:                mff = (ip->ip_off & htons(IP_MF)) != 0;
1.25      cgd       798:                if (mff) {
1.16      cgd       799:                        /*
                    800:                         * Make sure that fragments have a data length
                    801:                         * that's a non-zero multiple of 8 bytes.
                    802:                         */
1.155     itojun    803:                        if (ntohs(ip->ip_len) == 0 ||
                    804:                            (ntohs(ip->ip_len) & 0x7) != 0) {
1.16      cgd       805:                                ipstat.ips_badfrags++;
1.75      thorpej   806:                                IPQ_UNLOCK();
1.16      cgd       807:                                goto bad;
                    808:                        }
                    809:                }
1.155     itojun    810:                ip->ip_off = htons((ntohs(ip->ip_off) & IP_OFFMASK) << 3);
1.1       cgd       811:
                    812:                /*
                    813:                 * If datagram marked as having more fragments
                    814:                 * or if this is not the first fragment,
                    815:                 * attempt reassembly; if it succeeds, proceed.
                    816:                 */
1.155     itojun    817:                if (mff || ip->ip_off != htons(0)) {
1.1       cgd       818:                        ipstat.ips_fragments++;
1.72      thorpej   819:                        ipqe = pool_get(&ipqent_pool, PR_NOWAIT);
1.25      cgd       820:                        if (ipqe == NULL) {
                    821:                                ipstat.ips_rcvmemdrop++;
1.75      thorpej   822:                                IPQ_UNLOCK();
1.25      cgd       823:                                goto bad;
                    824:                        }
                    825:                        ipqe->ipqe_mff = mff;
1.50      thorpej   826:                        ipqe->ipqe_m = m;
1.25      cgd       827:                        ipqe->ipqe_ip = ip;
1.50      thorpej   828:                        m = ip_reass(ipqe, fp);
1.75      thorpej   829:                        if (m == 0) {
                    830:                                IPQ_UNLOCK();
1.89      itojun    831:                                return;
1.75      thorpej   832:                        }
1.13      mycroft   833:                        ipstat.ips_reassembled++;
1.50      thorpej   834:                        ip = mtod(m, struct ip *);
1.74      thorpej   835:                        hlen = ip->ip_hl << 2;
1.155     itojun    836:                        ip->ip_len = htons(ntohs(ip->ip_len) + hlen);
1.1       cgd       837:                } else
                    838:                        if (fp)
                    839:                                ip_freef(fp);
1.75      thorpej   840:                IPQ_UNLOCK();
1.79      mycroft   841:        }
1.128     itojun    842:
                    843: #ifdef IPSEC
                    844:        /*
                    845:         * enforce IPsec policy checking if we are seeing last header.
                    846:         * note that we do not visit this with protocols with pcb layer
                    847:         * code - like udp/tcp/raw ip.
                    848:         */
                    849:        if ((inetsw[ip_protox[ip->ip_p]].pr_flags & PR_LASTHDR) != 0 &&
                    850:            ipsec4_in_reject(m, NULL)) {
                    851:                ipsecstat.in_polvio++;
                    852:                goto bad;
                    853:        }
                    854: #endif
1.1       cgd       855:
                    856:        /*
                    857:         * Switch out to protocol's input routine.
                    858:         */
1.82      aidan     859: #if IFA_STATS
1.122     itojun    860:        if (ia && ip)
1.155     itojun    861:                ia->ia_ifa.ifa_data.ifad_inbytes += ntohs(ip->ip_len);
1.82      aidan     862: #endif
1.1       cgd       863:        ipstat.ips_delivered++;
1.89      itojun    864:     {
                    865:        int off = hlen, nh = ip->ip_p;
                    866:
                    867:        (*inetsw[ip_protox[nh]].pr_input)(m, off, nh);
                    868:        return;
                    869:     }
1.1       cgd       870: bad:
                    871:        m_freem(m);
1.135     thorpej   872:        return;
                    873:
                    874: badcsum:
                    875:        ipstat.ips_badsum++;
                    876:        m_freem(m);
1.1       cgd       877: }
                    878:
                    879: /*
                    880:  * Take incoming datagram fragment and try to
                    881:  * reassemble it into whole datagram.  If a chain for
                    882:  * reassembly of this datagram already exists, then it
                    883:  * is given as fp; otherwise have to make a chain.
                    884:  */
1.50      thorpej   885: struct mbuf *
1.25      cgd       886: ip_reass(ipqe, fp)
1.109     augustss  887:        struct ipqent *ipqe;
                    888:        struct ipq *fp;
1.1       cgd       889: {
1.109     augustss  890:        struct mbuf *m = ipqe->ipqe_m;
                    891:        struct ipqent *nq, *p, *q;
1.25      cgd       892:        struct ip *ip;
1.1       cgd       893:        struct mbuf *t;
1.25      cgd       894:        int hlen = ipqe->ipqe_ip->ip_hl << 2;
1.1       cgd       895:        int i, next;
                    896:
1.75      thorpej   897:        IPQ_LOCK_CHECK();
                    898:
1.1       cgd       899:        /*
                    900:         * Presence of header sizes in mbufs
                    901:         * would confuse code below.
                    902:         */
                    903:        m->m_data += hlen;
                    904:        m->m_len -= hlen;
                    905:
                    906:        /*
                    907:         * If first fragment to arrive, create a reassembly queue.
                    908:         */
                    909:        if (fp == 0) {
1.131     itojun    910:                /*
                    911:                 * Enforce upper bound on number of fragmented packets
                    912:                 * for which we attempt reassembly;
                    913:                 * If maxfrag is 0, never accept fragments.
                    914:                 * If maxfrag is -1, accept all fragments without limitation.
                    915:                 */
                    916:                if (ip_maxfragpackets < 0)
                    917:                        ;
                    918:                else if (ip_nfragpackets >= ip_maxfragpackets)
                    919:                        goto dropfrag;
                    920:                ip_nfragpackets++;
1.50      thorpej   921:                MALLOC(fp, struct ipq *, sizeof (struct ipq),
                    922:                    M_FTABLE, M_NOWAIT);
                    923:                if (fp == NULL)
1.1       cgd       924:                        goto dropfrag;
1.25      cgd       925:                LIST_INSERT_HEAD(&ipq, fp, ipq_q);
1.1       cgd       926:                fp->ipq_ttl = IPFRAGTTL;
1.25      cgd       927:                fp->ipq_p = ipqe->ipqe_ip->ip_p;
                    928:                fp->ipq_id = ipqe->ipqe_ip->ip_id;
1.148     matt      929:                TAILQ_INIT(&fp->ipq_fragq);
1.25      cgd       930:                fp->ipq_src = ipqe->ipqe_ip->ip_src;
                    931:                fp->ipq_dst = ipqe->ipqe_ip->ip_dst;
                    932:                p = NULL;
1.1       cgd       933:                goto insert;
                    934:        }
                    935:
                    936:        /*
                    937:         * Find a segment which begins after this one does.
                    938:         */
1.148     matt      939:        for (p = NULL, q = TAILQ_FIRST(&fp->ipq_fragq); q != NULL;
                    940:            p = q, q = TAILQ_NEXT(q, ipqe_q))
1.155     itojun    941:                if (ntohs(q->ipqe_ip->ip_off) > ntohs(ipqe->ipqe_ip->ip_off))
1.1       cgd       942:                        break;
                    943:
                    944:        /*
                    945:         * If there is a preceding segment, it may provide some of
                    946:         * our data already.  If so, drop the data from the incoming
                    947:         * segment.  If it provides all of our data, drop us.
                    948:         */
1.25      cgd       949:        if (p != NULL) {
1.155     itojun    950:                i = ntohs(p->ipqe_ip->ip_off) + ntohs(p->ipqe_ip->ip_len) -
                    951:                    ntohs(ipqe->ipqe_ip->ip_off);
1.1       cgd       952:                if (i > 0) {
1.155     itojun    953:                        if (i >= ntohs(ipqe->ipqe_ip->ip_len))
1.1       cgd       954:                                goto dropfrag;
1.50      thorpej   955:                        m_adj(ipqe->ipqe_m, i);
1.155     itojun    956:                        ipqe->ipqe_ip->ip_off =
                    957:                            htons(ntohs(ipqe->ipqe_ip->ip_off) + i);
                    958:                        ipqe->ipqe_ip->ip_len =
                    959:                            htons(ntohs(ipqe->ipqe_ip->ip_len) - i);
1.1       cgd       960:                }
                    961:        }
                    962:
                    963:        /*
                    964:         * While we overlap succeeding segments trim them or,
                    965:         * if they are completely covered, dequeue them.
                    966:         */
1.155     itojun    967:        for (; q != NULL &&
                    968:            ntohs(ipqe->ipqe_ip->ip_off) + ntohs(ipqe->ipqe_ip->ip_len) >
                    969:            ntohs(q->ipqe_ip->ip_off); q = nq) {
                    970:                i = (ntohs(ipqe->ipqe_ip->ip_off) +
                    971:                    ntohs(ipqe->ipqe_ip->ip_len)) - ntohs(q->ipqe_ip->ip_off);
                    972:                if (i < ntohs(q->ipqe_ip->ip_len)) {
                    973:                        q->ipqe_ip->ip_len =
                    974:                            htons(ntohs(q->ipqe_ip->ip_len) - i);
                    975:                        q->ipqe_ip->ip_off =
                    976:                            htons(ntohs(q->ipqe_ip->ip_off) + i);
1.50      thorpej   977:                        m_adj(q->ipqe_m, i);
1.1       cgd       978:                        break;
                    979:                }
1.148     matt      980:                nq = TAILQ_NEXT(q, ipqe_q);
1.50      thorpej   981:                m_freem(q->ipqe_m);
1.148     matt      982:                TAILQ_REMOVE(&fp->ipq_fragq, q, ipqe_q);
1.72      thorpej   983:                pool_put(&ipqent_pool, q);
1.1       cgd       984:        }
                    985:
                    986: insert:
                    987:        /*
                    988:         * Stick new segment in its place;
                    989:         * check for complete reassembly.
                    990:         */
1.25      cgd       991:        if (p == NULL) {
1.148     matt      992:                TAILQ_INSERT_HEAD(&fp->ipq_fragq, ipqe, ipqe_q);
1.25      cgd       993:        } else {
1.148     matt      994:                TAILQ_INSERT_AFTER(&fp->ipq_fragq, p, ipqe, ipqe_q);
1.25      cgd       995:        }
1.1       cgd       996:        next = 0;
1.148     matt      997:        for (p = NULL, q = TAILQ_FIRST(&fp->ipq_fragq); q != NULL;
                    998:            p = q, q = TAILQ_NEXT(q, ipqe_q)) {
1.155     itojun    999:                if (ntohs(q->ipqe_ip->ip_off) != next)
1.1       cgd      1000:                        return (0);
1.155     itojun   1001:                next += ntohs(q->ipqe_ip->ip_len);
1.1       cgd      1002:        }
1.25      cgd      1003:        if (p->ipqe_mff)
1.1       cgd      1004:                return (0);
                   1005:
                   1006:        /*
1.41      thorpej  1007:         * Reassembly is complete.  Check for a bogus message size and
                   1008:         * concatenate fragments.
1.1       cgd      1009:         */
1.148     matt     1010:        q = TAILQ_FIRST(&fp->ipq_fragq);
1.25      cgd      1011:        ip = q->ipqe_ip;
1.41      thorpej  1012:        if ((next + (ip->ip_hl << 2)) > IP_MAXPACKET) {
                   1013:                ipstat.ips_toolong++;
                   1014:                ip_freef(fp);
                   1015:                return (0);
                   1016:        }
1.50      thorpej  1017:        m = q->ipqe_m;
1.1       cgd      1018:        t = m->m_next;
                   1019:        m->m_next = 0;
                   1020:        m_cat(m, t);
1.148     matt     1021:        nq = TAILQ_NEXT(q, ipqe_q);
1.72      thorpej  1022:        pool_put(&ipqent_pool, q);
1.25      cgd      1023:        for (q = nq; q != NULL; q = nq) {
1.50      thorpej  1024:                t = q->ipqe_m;
1.148     matt     1025:                nq = TAILQ_NEXT(q, ipqe_q);
1.72      thorpej  1026:                pool_put(&ipqent_pool, q);
1.1       cgd      1027:                m_cat(m, t);
                   1028:        }
                   1029:
                   1030:        /*
                   1031:         * Create header for new ip packet by
                   1032:         * modifying header of first packet;
                   1033:         * dequeue and discard fragment reassembly header.
                   1034:         * Make header visible.
                   1035:         */
1.155     itojun   1036:        ip->ip_len = htons(next);
1.25      cgd      1037:        ip->ip_src = fp->ipq_src;
                   1038:        ip->ip_dst = fp->ipq_dst;
                   1039:        LIST_REMOVE(fp, ipq_q);
1.50      thorpej  1040:        FREE(fp, M_FTABLE);
1.131     itojun   1041:        ip_nfragpackets--;
1.1       cgd      1042:        m->m_len += (ip->ip_hl << 2);
                   1043:        m->m_data -= (ip->ip_hl << 2);
                   1044:        /* some debugging cruft by sklower, below, will go away soon */
                   1045:        if (m->m_flags & M_PKTHDR) { /* XXX this should be done elsewhere */
1.109     augustss 1046:                int plen = 0;
1.50      thorpej  1047:                for (t = m; t; t = t->m_next)
                   1048:                        plen += t->m_len;
                   1049:                m->m_pkthdr.len = plen;
1.1       cgd      1050:        }
1.50      thorpej  1051:        return (m);
1.1       cgd      1052:
                   1053: dropfrag:
                   1054:        ipstat.ips_fragdropped++;
                   1055:        m_freem(m);
1.72      thorpej  1056:        pool_put(&ipqent_pool, ipqe);
1.1       cgd      1057:        return (0);
                   1058: }
                   1059:
                   1060: /*
                   1061:  * Free a fragment reassembly header and all
                   1062:  * associated datagrams.
                   1063:  */
1.8       mycroft  1064: void
1.1       cgd      1065: ip_freef(fp)
                   1066:        struct ipq *fp;
                   1067: {
1.109     augustss 1068:        struct ipqent *q, *p;
1.1       cgd      1069:
1.75      thorpej  1070:        IPQ_LOCK_CHECK();
                   1071:
1.148     matt     1072:        for (q = TAILQ_FIRST(&fp->ipq_fragq); q != NULL; q = p) {
                   1073:                p = TAILQ_NEXT(q, ipqe_q);
1.50      thorpej  1074:                m_freem(q->ipqe_m);
1.148     matt     1075:                TAILQ_REMOVE(&fp->ipq_fragq, q, ipqe_q);
1.72      thorpej  1076:                pool_put(&ipqent_pool, q);
1.1       cgd      1077:        }
1.25      cgd      1078:        LIST_REMOVE(fp, ipq_q);
1.50      thorpej  1079:        FREE(fp, M_FTABLE);
1.131     itojun   1080:        ip_nfragpackets--;
1.1       cgd      1081: }
                   1082:
                   1083: /*
                   1084:  * IP timer processing;
                   1085:  * if a timer expires on a reassembly
                   1086:  * queue, discard it.
                   1087:  */
1.8       mycroft  1088: void
1.1       cgd      1089: ip_slowtimo()
                   1090: {
1.109     augustss 1091:        struct ipq *fp, *nfp;
1.24      mycroft  1092:        int s = splsoftnet();
1.1       cgd      1093:
1.75      thorpej  1094:        IPQ_LOCK();
1.140     matt     1095:        for (fp = LIST_FIRST(&ipq); fp != NULL; fp = nfp) {
                   1096:                nfp = LIST_NEXT(fp, ipq_q);
1.25      cgd      1097:                if (--fp->ipq_ttl == 0) {
1.1       cgd      1098:                        ipstat.ips_fragtimeout++;
1.25      cgd      1099:                        ip_freef(fp);
1.1       cgd      1100:                }
                   1101:        }
1.131     itojun   1102:        /*
                   1103:         * If we are over the maximum number of fragments
                   1104:         * (due to the limit being lowered), drain off
                   1105:         * enough to get down to the new limit.
                   1106:         */
                   1107:        if (ip_maxfragpackets < 0)
                   1108:                ;
                   1109:        else {
1.140     matt     1110:                while (ip_nfragpackets > ip_maxfragpackets && LIST_FIRST(&ipq))
                   1111:                        ip_freef(LIST_FIRST(&ipq));
1.131     itojun   1112:        }
1.75      thorpej  1113:        IPQ_UNLOCK();
1.63      matt     1114: #ifdef GATEWAY
                   1115:        ipflow_slowtimo();
                   1116: #endif
1.1       cgd      1117:        splx(s);
                   1118: }
                   1119:
                   1120: /*
                   1121:  * Drain off all datagram fragments.
                   1122:  */
1.8       mycroft  1123: void
1.1       cgd      1124: ip_drain()
                   1125: {
                   1126:
1.75      thorpej  1127:        /*
                   1128:         * We may be called from a device's interrupt context.  If
                   1129:         * the ipq is already busy, just bail out now.
                   1130:         */
                   1131:        if (ipq_lock_try() == 0)
                   1132:                return;
                   1133:
1.140     matt     1134:        while (LIST_FIRST(&ipq) != NULL) {
1.1       cgd      1135:                ipstat.ips_fragdropped++;
1.140     matt     1136:                ip_freef(LIST_FIRST(&ipq));
1.1       cgd      1137:        }
1.75      thorpej  1138:
                   1139:        IPQ_UNLOCK();
1.1       cgd      1140: }
                   1141:
                   1142: /*
                   1143:  * Do option processing on a datagram,
                   1144:  * possibly discarding it if bad options are encountered,
                   1145:  * or forwarding it if source-routed.
                   1146:  * Returns 1 if packet has been forwarded/freed,
                   1147:  * 0 if the packet should be processed further.
                   1148:  */
1.8       mycroft  1149: int
1.1       cgd      1150: ip_dooptions(m)
                   1151:        struct mbuf *m;
                   1152: {
1.109     augustss 1153:        struct ip *ip = mtod(m, struct ip *);
                   1154:        u_char *cp, *cp0;
                   1155:        struct ip_timestamp *ipt;
                   1156:        struct in_ifaddr *ia;
1.1       cgd      1157:        int opt, optlen, cnt, off, code, type = ICMP_PARAMPROB, forward = 0;
1.104     thorpej  1158:        struct in_addr dst;
1.1       cgd      1159:        n_time ntime;
                   1160:
1.13      mycroft  1161:        dst = ip->ip_dst;
1.1       cgd      1162:        cp = (u_char *)(ip + 1);
                   1163:        cnt = (ip->ip_hl << 2) - sizeof (struct ip);
                   1164:        for (; cnt > 0; cnt -= optlen, cp += optlen) {
                   1165:                opt = cp[IPOPT_OPTVAL];
                   1166:                if (opt == IPOPT_EOL)
                   1167:                        break;
                   1168:                if (opt == IPOPT_NOP)
                   1169:                        optlen = 1;
                   1170:                else {
1.113     itojun   1171:                        if (cnt < IPOPT_OLEN + sizeof(*cp)) {
                   1172:                                code = &cp[IPOPT_OLEN] - (u_char *)ip;
                   1173:                                goto bad;
                   1174:                        }
1.1       cgd      1175:                        optlen = cp[IPOPT_OLEN];
1.114     itojun   1176:                        if (optlen < IPOPT_OLEN + sizeof(*cp) || optlen > cnt) {
1.1       cgd      1177:                                code = &cp[IPOPT_OLEN] - (u_char *)ip;
                   1178:                                goto bad;
                   1179:                        }
                   1180:                }
                   1181:                switch (opt) {
                   1182:
                   1183:                default:
                   1184:                        break;
                   1185:
                   1186:                /*
                   1187:                 * Source routing with record.
                   1188:                 * Find interface with current destination address.
                   1189:                 * If none on this machine then drop if strictly routed,
                   1190:                 * or do nothing if loosely routed.
                   1191:                 * Record interface address and bring up next address
                   1192:                 * component.  If strictly routed make sure next
                   1193:                 * address is on directly accessible net.
                   1194:                 */
                   1195:                case IPOPT_LSRR:
                   1196:                case IPOPT_SSRR:
1.47      cjs      1197:                        if (ip_allowsrcrt == 0) {
                   1198:                                type = ICMP_UNREACH;
                   1199:                                code = ICMP_UNREACH_NET_PROHIB;
                   1200:                                goto bad;
                   1201:                        }
1.114     itojun   1202:                        if (optlen < IPOPT_OFFSET + sizeof(*cp)) {
                   1203:                                code = &cp[IPOPT_OLEN] - (u_char *)ip;
                   1204:                                goto bad;
                   1205:                        }
1.1       cgd      1206:                        if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) {
                   1207:                                code = &cp[IPOPT_OFFSET] - (u_char *)ip;
                   1208:                                goto bad;
                   1209:                        }
                   1210:                        ipaddr.sin_addr = ip->ip_dst;
1.19      mycroft  1211:                        ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr)));
1.1       cgd      1212:                        if (ia == 0) {
                   1213:                                if (opt == IPOPT_SSRR) {
                   1214:                                        type = ICMP_UNREACH;
                   1215:                                        code = ICMP_UNREACH_SRCFAIL;
                   1216:                                        goto bad;
                   1217:                                }
                   1218:                                /*
                   1219:                                 * Loose routing, and not at next destination
                   1220:                                 * yet; nothing to do except forward.
                   1221:                                 */
                   1222:                                break;
                   1223:                        }
                   1224:                        off--;                  /* 0 origin */
1.112     sommerfe 1225:                        if ((off + sizeof(struct in_addr)) > optlen) {
1.1       cgd      1226:                                /*
                   1227:                                 * End of source route.  Should be for us.
                   1228:                                 */
                   1229:                                save_rte(cp, ip->ip_src);
                   1230:                                break;
                   1231:                        }
                   1232:                        /*
                   1233:                         * locate outgoing interface
                   1234:                         */
                   1235:                        bcopy((caddr_t)(cp + off), (caddr_t)&ipaddr.sin_addr,
                   1236:                            sizeof(ipaddr.sin_addr));
1.96      thorpej  1237:                        if (opt == IPOPT_SSRR)
                   1238:                                ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr)));
                   1239:                        else
1.1       cgd      1240:                                ia = ip_rtaddr(ipaddr.sin_addr);
                   1241:                        if (ia == 0) {
                   1242:                                type = ICMP_UNREACH;
                   1243:                                code = ICMP_UNREACH_SRCFAIL;
                   1244:                                goto bad;
                   1245:                        }
                   1246:                        ip->ip_dst = ipaddr.sin_addr;
1.20      mycroft  1247:                        bcopy((caddr_t)&ia->ia_addr.sin_addr,
1.1       cgd      1248:                            (caddr_t)(cp + off), sizeof(struct in_addr));
                   1249:                        cp[IPOPT_OFFSET] += sizeof(struct in_addr);
1.13      mycroft  1250:                        /*
                   1251:                         * Let ip_intr's mcast routing check handle mcast pkts
                   1252:                         */
1.18      mycroft  1253:                        forward = !IN_MULTICAST(ip->ip_dst.s_addr);
1.1       cgd      1254:                        break;
                   1255:
                   1256:                case IPOPT_RR:
1.114     itojun   1257:                        if (optlen < IPOPT_OFFSET + sizeof(*cp)) {
                   1258:                                code = &cp[IPOPT_OLEN] - (u_char *)ip;
                   1259:                                goto bad;
                   1260:                        }
1.1       cgd      1261:                        if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) {
                   1262:                                code = &cp[IPOPT_OFFSET] - (u_char *)ip;
                   1263:                                goto bad;
                   1264:                        }
                   1265:                        /*
                   1266:                         * If no space remains, ignore.
                   1267:                         */
                   1268:                        off--;                  /* 0 origin */
1.112     sommerfe 1269:                        if ((off + sizeof(struct in_addr)) > optlen)
1.1       cgd      1270:                                break;
                   1271:                        bcopy((caddr_t)(&ip->ip_dst), (caddr_t)&ipaddr.sin_addr,
                   1272:                            sizeof(ipaddr.sin_addr));
                   1273:                        /*
                   1274:                         * locate outgoing interface; if we're the destination,
                   1275:                         * use the incoming interface (should be same).
                   1276:                         */
1.96      thorpej  1277:                        if ((ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr))))
                   1278:                            == NULL &&
                   1279:                            (ia = ip_rtaddr(ipaddr.sin_addr)) == NULL) {
1.1       cgd      1280:                                type = ICMP_UNREACH;
                   1281:                                code = ICMP_UNREACH_HOST;
                   1282:                                goto bad;
                   1283:                        }
1.20      mycroft  1284:                        bcopy((caddr_t)&ia->ia_addr.sin_addr,
1.1       cgd      1285:                            (caddr_t)(cp + off), sizeof(struct in_addr));
                   1286:                        cp[IPOPT_OFFSET] += sizeof(struct in_addr);
                   1287:                        break;
                   1288:
                   1289:                case IPOPT_TS:
                   1290:                        code = cp - (u_char *)ip;
                   1291:                        ipt = (struct ip_timestamp *)cp;
1.114     itojun   1292:                        if (ipt->ipt_len < 4 || ipt->ipt_len > 40) {
                   1293:                                code = (u_char *)&ipt->ipt_len - (u_char *)ip;
1.1       cgd      1294:                                goto bad;
1.114     itojun   1295:                        }
                   1296:                        if (ipt->ipt_ptr < 5) {
                   1297:                                code = (u_char *)&ipt->ipt_ptr - (u_char *)ip;
                   1298:                                goto bad;
                   1299:                        }
1.15      cgd      1300:                        if (ipt->ipt_ptr > ipt->ipt_len - sizeof (int32_t)) {
1.114     itojun   1301:                                if (++ipt->ipt_oflw == 0) {
                   1302:                                        code = (u_char *)&ipt->ipt_ptr -
                   1303:                                            (u_char *)ip;
1.1       cgd      1304:                                        goto bad;
1.114     itojun   1305:                                }
1.1       cgd      1306:                                break;
                   1307:                        }
1.104     thorpej  1308:                        cp0 = (cp + ipt->ipt_ptr - 1);
1.1       cgd      1309:                        switch (ipt->ipt_flg) {
                   1310:
                   1311:                        case IPOPT_TS_TSONLY:
                   1312:                                break;
                   1313:
                   1314:                        case IPOPT_TS_TSANDADDR:
1.66      thorpej  1315:                                if (ipt->ipt_ptr - 1 + sizeof(n_time) +
1.114     itojun   1316:                                    sizeof(struct in_addr) > ipt->ipt_len) {
                   1317:                                        code = (u_char *)&ipt->ipt_ptr -
                   1318:                                            (u_char *)ip;
1.1       cgd      1319:                                        goto bad;
1.114     itojun   1320:                                }
1.13      mycroft  1321:                                ipaddr.sin_addr = dst;
1.96      thorpej  1322:                                ia = ifatoia(ifaof_ifpforaddr(sintosa(&ipaddr),
                   1323:                                    m->m_pkthdr.rcvif));
1.13      mycroft  1324:                                if (ia == 0)
                   1325:                                        continue;
1.104     thorpej  1326:                                bcopy(&ia->ia_addr.sin_addr,
                   1327:                                    cp0, sizeof(struct in_addr));
1.1       cgd      1328:                                ipt->ipt_ptr += sizeof(struct in_addr);
                   1329:                                break;
                   1330:
                   1331:                        case IPOPT_TS_PRESPEC:
1.66      thorpej  1332:                                if (ipt->ipt_ptr - 1 + sizeof(n_time) +
1.114     itojun   1333:                                    sizeof(struct in_addr) > ipt->ipt_len) {
                   1334:                                        code = (u_char *)&ipt->ipt_ptr -
                   1335:                                            (u_char *)ip;
1.1       cgd      1336:                                        goto bad;
1.114     itojun   1337:                                }
1.104     thorpej  1338:                                bcopy(cp0, &ipaddr.sin_addr,
1.1       cgd      1339:                                    sizeof(struct in_addr));
1.96      thorpej  1340:                                if (ifatoia(ifa_ifwithaddr(sintosa(&ipaddr)))
                   1341:                                    == NULL)
1.1       cgd      1342:                                        continue;
                   1343:                                ipt->ipt_ptr += sizeof(struct in_addr);
                   1344:                                break;
                   1345:
                   1346:                        default:
1.114     itojun   1347:                                /* XXX can't take &ipt->ipt_flg */
                   1348:                                code = (u_char *)&ipt->ipt_ptr -
                   1349:                                    (u_char *)ip + 1;
1.1       cgd      1350:                                goto bad;
                   1351:                        }
                   1352:                        ntime = iptime();
1.107     thorpej  1353:                        cp0 = (u_char *) &ntime; /* XXX grumble, GCC... */
                   1354:                        bcopy(cp0, (caddr_t)cp + ipt->ipt_ptr - 1,
1.1       cgd      1355:                            sizeof(n_time));
                   1356:                        ipt->ipt_ptr += sizeof(n_time);
                   1357:                }
                   1358:        }
                   1359:        if (forward) {
1.26      thorpej  1360:                if (ip_forwsrcrt == 0) {
                   1361:                        type = ICMP_UNREACH;
                   1362:                        code = ICMP_UNREACH_SRCFAIL;
                   1363:                        goto bad;
                   1364:                }
1.1       cgd      1365:                ip_forward(m, 1);
                   1366:                return (1);
1.13      mycroft  1367:        }
                   1368:        return (0);
1.1       cgd      1369: bad:
1.13      mycroft  1370:        icmp_error(m, type, code, 0, 0);
                   1371:        ipstat.ips_badoptions++;
1.1       cgd      1372:        return (1);
                   1373: }
                   1374:
                   1375: /*
                   1376:  * Given address of next destination (final or next hop),
                   1377:  * return internet address info of interface to be used to get there.
                   1378:  */
                   1379: struct in_ifaddr *
                   1380: ip_rtaddr(dst)
                   1381:         struct in_addr dst;
                   1382: {
1.109     augustss 1383:        struct sockaddr_in *sin;
1.1       cgd      1384:
1.19      mycroft  1385:        sin = satosin(&ipforward_rt.ro_dst);
1.1       cgd      1386:
1.35      mycroft  1387:        if (ipforward_rt.ro_rt == 0 || !in_hosteq(dst, sin->sin_addr)) {
1.1       cgd      1388:                if (ipforward_rt.ro_rt) {
                   1389:                        RTFREE(ipforward_rt.ro_rt);
                   1390:                        ipforward_rt.ro_rt = 0;
                   1391:                }
                   1392:                sin->sin_family = AF_INET;
                   1393:                sin->sin_len = sizeof(*sin);
                   1394:                sin->sin_addr = dst;
                   1395:
                   1396:                rtalloc(&ipforward_rt);
                   1397:        }
                   1398:        if (ipforward_rt.ro_rt == 0)
                   1399:                return ((struct in_ifaddr *)0);
1.19      mycroft  1400:        return (ifatoia(ipforward_rt.ro_rt->rt_ifa));
1.1       cgd      1401: }
                   1402:
                   1403: /*
                   1404:  * Save incoming source route for use in replies,
                   1405:  * to be picked up later by ip_srcroute if the receiver is interested.
                   1406:  */
1.13      mycroft  1407: void
1.1       cgd      1408: save_rte(option, dst)
                   1409:        u_char *option;
                   1410:        struct in_addr dst;
                   1411: {
                   1412:        unsigned olen;
                   1413:
                   1414:        olen = option[IPOPT_OLEN];
                   1415: #ifdef DIAGNOSTIC
                   1416:        if (ipprintfs)
1.39      christos 1417:                printf("save_rte: olen %d\n", olen);
1.89      itojun   1418: #endif /* 0 */
1.1       cgd      1419:        if (olen > sizeof(ip_srcrt) - (1 + sizeof(dst)))
                   1420:                return;
                   1421:        bcopy((caddr_t)option, (caddr_t)ip_srcrt.srcopt, olen);
                   1422:        ip_nhops = (olen - IPOPT_OFFSET - 1) / sizeof(struct in_addr);
                   1423:        ip_srcrt.dst = dst;
                   1424: }
                   1425:
                   1426: /*
                   1427:  * Retrieve incoming source route for use in replies,
                   1428:  * in the same form used by setsockopt.
                   1429:  * The first hop is placed before the options, will be removed later.
                   1430:  */
                   1431: struct mbuf *
                   1432: ip_srcroute()
                   1433: {
1.109     augustss 1434:        struct in_addr *p, *q;
                   1435:        struct mbuf *m;
1.1       cgd      1436:
                   1437:        if (ip_nhops == 0)
                   1438:                return ((struct mbuf *)0);
                   1439:        m = m_get(M_DONTWAIT, MT_SOOPTS);
                   1440:        if (m == 0)
                   1441:                return ((struct mbuf *)0);
                   1442:
1.164     matt     1443:        MCLAIM(m, &inetdomain.dom_mowner);
1.13      mycroft  1444: #define OPTSIZ (sizeof(ip_srcrt.nop) + sizeof(ip_srcrt.srcopt))
1.1       cgd      1445:
                   1446:        /* length is (nhops+1)*sizeof(addr) + sizeof(nop + srcrt header) */
                   1447:        m->m_len = ip_nhops * sizeof(struct in_addr) + sizeof(struct in_addr) +
                   1448:            OPTSIZ;
                   1449: #ifdef DIAGNOSTIC
                   1450:        if (ipprintfs)
1.39      christos 1451:                printf("ip_srcroute: nhops %d mlen %d", ip_nhops, m->m_len);
1.1       cgd      1452: #endif
                   1453:
                   1454:        /*
                   1455:         * First save first hop for return route
                   1456:         */
                   1457:        p = &ip_srcrt.route[ip_nhops - 1];
                   1458:        *(mtod(m, struct in_addr *)) = *p--;
                   1459: #ifdef DIAGNOSTIC
                   1460:        if (ipprintfs)
1.39      christos 1461:                printf(" hops %x", ntohl(mtod(m, struct in_addr *)->s_addr));
1.1       cgd      1462: #endif
                   1463:
                   1464:        /*
                   1465:         * Copy option fields and padding (nop) to mbuf.
                   1466:         */
                   1467:        ip_srcrt.nop = IPOPT_NOP;
                   1468:        ip_srcrt.srcopt[IPOPT_OFFSET] = IPOPT_MINOFF;
                   1469:        bcopy((caddr_t)&ip_srcrt.nop,
                   1470:            mtod(m, caddr_t) + sizeof(struct in_addr), OPTSIZ);
                   1471:        q = (struct in_addr *)(mtod(m, caddr_t) +
                   1472:            sizeof(struct in_addr) + OPTSIZ);
                   1473: #undef OPTSIZ
                   1474:        /*
                   1475:         * Record return path as an IP source route,
                   1476:         * reversing the path (pointers are now aligned).
                   1477:         */
                   1478:        while (p >= ip_srcrt.route) {
                   1479: #ifdef DIAGNOSTIC
                   1480:                if (ipprintfs)
1.39      christos 1481:                        printf(" %x", ntohl(q->s_addr));
1.1       cgd      1482: #endif
                   1483:                *q++ = *p--;
                   1484:        }
                   1485:        /*
                   1486:         * Last hop goes to final destination.
                   1487:         */
                   1488:        *q = ip_srcrt.dst;
                   1489: #ifdef DIAGNOSTIC
                   1490:        if (ipprintfs)
1.39      christos 1491:                printf(" %x\n", ntohl(q->s_addr));
1.1       cgd      1492: #endif
                   1493:        return (m);
                   1494: }
                   1495:
                   1496: /*
                   1497:  * Strip out IP options, at higher
                   1498:  * level protocol in the kernel.
                   1499:  * Second argument is buffer to which options
                   1500:  * will be moved, and return value is their length.
                   1501:  * XXX should be deleted; last arg currently ignored.
                   1502:  */
1.8       mycroft  1503: void
1.1       cgd      1504: ip_stripoptions(m, mopt)
1.109     augustss 1505:        struct mbuf *m;
1.1       cgd      1506:        struct mbuf *mopt;
                   1507: {
1.109     augustss 1508:        int i;
1.1       cgd      1509:        struct ip *ip = mtod(m, struct ip *);
1.109     augustss 1510:        caddr_t opts;
1.1       cgd      1511:        int olen;
                   1512:
1.79      mycroft  1513:        olen = (ip->ip_hl << 2) - sizeof (struct ip);
1.1       cgd      1514:        opts = (caddr_t)(ip + 1);
                   1515:        i = m->m_len - (sizeof (struct ip) + olen);
                   1516:        bcopy(opts  + olen, opts, (unsigned)i);
                   1517:        m->m_len -= olen;
                   1518:        if (m->m_flags & M_PKTHDR)
                   1519:                m->m_pkthdr.len -= olen;
1.155     itojun   1520:        ip->ip_len = htons(ntohs(ip->ip_len) - olen);
1.79      mycroft  1521:        ip->ip_hl = sizeof (struct ip) >> 2;
1.1       cgd      1522: }
                   1523:
1.139     matt     1524: const int inetctlerrmap[PRC_NCMDS] = {
1.1       cgd      1525:        0,              0,              0,              0,
                   1526:        0,              EMSGSIZE,       EHOSTDOWN,      EHOSTUNREACH,
                   1527:        EHOSTUNREACH,   EHOSTUNREACH,   ECONNREFUSED,   ECONNREFUSED,
                   1528:        EMSGSIZE,       EHOSTUNREACH,   0,              0,
                   1529:        0,              0,              0,              0,
                   1530:        ENOPROTOOPT
                   1531: };
                   1532:
                   1533: /*
                   1534:  * Forward a packet.  If some error occurs return the sender
                   1535:  * an icmp packet.  Note we can't always generate a meaningful
                   1536:  * icmp message because icmp doesn't have a large enough repertoire
                   1537:  * of codes and types.
                   1538:  *
                   1539:  * If not forwarding, just drop the packet.  This could be confusing
                   1540:  * if ipforwarding was zero but some routing protocol was advancing
                   1541:  * us as a gateway to somewhere.  However, we must let the routing
                   1542:  * protocol deal with that.
                   1543:  *
                   1544:  * The srcrt parameter indicates whether the packet is being forwarded
                   1545:  * via a source route.
                   1546:  */
1.13      mycroft  1547: void
1.1       cgd      1548: ip_forward(m, srcrt)
                   1549:        struct mbuf *m;
                   1550:        int srcrt;
                   1551: {
1.109     augustss 1552:        struct ip *ip = mtod(m, struct ip *);
                   1553:        struct sockaddr_in *sin;
                   1554:        struct rtentry *rt;
1.28      christos 1555:        int error, type = 0, code = 0;
1.1       cgd      1556:        struct mbuf *mcopy;
1.13      mycroft  1557:        n_long dest;
                   1558:        struct ifnet *destifp;
1.89      itojun   1559: #ifdef IPSEC
                   1560:        struct ifnet dummyifp;
                   1561: #endif
1.164     matt     1562:
                   1563:        /*
                   1564:         * We are now in the output path.
                   1565:         */
                   1566:        MCLAIM(m, &ip_tx_mowner);
1.135     thorpej  1567:
                   1568:        /*
                   1569:         * Clear any in-bound checksum flags for this packet.
                   1570:         */
                   1571:        m->m_pkthdr.csum_flags = 0;
1.1       cgd      1572:
1.13      mycroft  1573:        dest = 0;
1.1       cgd      1574: #ifdef DIAGNOSTIC
                   1575:        if (ipprintfs)
1.70      thorpej  1576:                printf("forward: src %2.2x dst %2.2x ttl %x\n",
                   1577:                    ntohl(ip->ip_src.s_addr),
                   1578:                    ntohl(ip->ip_dst.s_addr), ip->ip_ttl);
1.1       cgd      1579: #endif
1.93      sommerfe 1580:        if (m->m_flags & (M_BCAST|M_MCAST) || in_canforward(ip->ip_dst) == 0) {
1.1       cgd      1581:                ipstat.ips_cantforward++;
                   1582:                m_freem(m);
                   1583:                return;
                   1584:        }
                   1585:        if (ip->ip_ttl <= IPTTLDEC) {
1.13      mycroft  1586:                icmp_error(m, ICMP_TIMXCEED, ICMP_TIMXCEED_INTRANS, dest, 0);
1.1       cgd      1587:                return;
                   1588:        }
                   1589:        ip->ip_ttl -= IPTTLDEC;
                   1590:
1.19      mycroft  1591:        sin = satosin(&ipforward_rt.ro_dst);
1.1       cgd      1592:        if ((rt = ipforward_rt.ro_rt) == 0 ||
1.35      mycroft  1593:            !in_hosteq(ip->ip_dst, sin->sin_addr)) {
1.1       cgd      1594:                if (ipforward_rt.ro_rt) {
                   1595:                        RTFREE(ipforward_rt.ro_rt);
                   1596:                        ipforward_rt.ro_rt = 0;
                   1597:                }
                   1598:                sin->sin_family = AF_INET;
1.35      mycroft  1599:                sin->sin_len = sizeof(struct sockaddr_in);
1.1       cgd      1600:                sin->sin_addr = ip->ip_dst;
                   1601:
                   1602:                rtalloc(&ipforward_rt);
                   1603:                if (ipforward_rt.ro_rt == 0) {
1.13      mycroft  1604:                        icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_HOST, dest, 0);
1.1       cgd      1605:                        return;
                   1606:                }
                   1607:                rt = ipforward_rt.ro_rt;
                   1608:        }
                   1609:
                   1610:        /*
1.34      mycroft  1611:         * Save at most 68 bytes of the packet in case
1.1       cgd      1612:         * we need to generate an ICMP message to the src.
1.119     itojun   1613:         * Pullup to avoid sharing mbuf cluster between m and mcopy.
1.1       cgd      1614:         */
1.155     itojun   1615:        mcopy = m_copym(m, 0, imin(ntohs(ip->ip_len), 68), M_DONTWAIT);
1.119     itojun   1616:        if (mcopy)
                   1617:                mcopy = m_pullup(mcopy, ip->ip_hl << 2);
1.1       cgd      1618:
                   1619:        /*
                   1620:         * If forwarding packet using same interface that it came in on,
                   1621:         * perhaps should send a redirect to sender to shortcut a hop.
                   1622:         * Only send redirect if source is sending directly to us,
                   1623:         * and if packet was not source routed (or has any options).
                   1624:         * Also, don't send redirect if forwarding using a default route
                   1625:         * or a route modified by a redirect.
                   1626:         */
                   1627:        if (rt->rt_ifp == m->m_pkthdr.rcvif &&
                   1628:            (rt->rt_flags & (RTF_DYNAMIC|RTF_MODIFIED)) == 0 &&
1.35      mycroft  1629:            !in_nullhost(satosin(rt_key(rt))->sin_addr) &&
1.1       cgd      1630:            ipsendredirects && !srcrt) {
1.19      mycroft  1631:                if (rt->rt_ifa &&
                   1632:                    (ip->ip_src.s_addr & ifatoia(rt->rt_ifa)->ia_subnetmask) ==
                   1633:                    ifatoia(rt->rt_ifa)->ia_subnet) {
1.77      thorpej  1634:                        if (rt->rt_flags & RTF_GATEWAY)
                   1635:                                dest = satosin(rt->rt_gateway)->sin_addr.s_addr;
                   1636:                        else
                   1637:                                dest = ip->ip_dst.s_addr;
                   1638:                        /*
                   1639:                         * Router requirements says to only send host
                   1640:                         * redirects.
                   1641:                         */
                   1642:                        type = ICMP_REDIRECT;
                   1643:                        code = ICMP_REDIRECT_HOST;
1.1       cgd      1644: #ifdef DIAGNOSTIC
1.77      thorpej  1645:                        if (ipprintfs)
                   1646:                                printf("redirect (%d) to %x\n", code,
                   1647:                                    (u_int32_t)dest);
1.1       cgd      1648: #endif
                   1649:                }
                   1650:        }
                   1651:
1.89      itojun   1652: #ifdef IPSEC
1.134     lukem    1653:        /* Don't lookup socket in forwarding case */
1.127     itojun   1654:        (void)ipsec_setsocket(m, NULL);
1.103     itojun   1655: #endif
1.27      thorpej  1656:        error = ip_output(m, (struct mbuf *)0, &ipforward_rt,
                   1657:            (IP_FORWARDING | (ip_directedbcast ? IP_ALLOWBROADCAST : 0)), 0);
1.1       cgd      1658:        if (error)
                   1659:                ipstat.ips_cantforward++;
                   1660:        else {
                   1661:                ipstat.ips_forward++;
                   1662:                if (type)
                   1663:                        ipstat.ips_redirectsent++;
                   1664:                else {
1.63      matt     1665:                        if (mcopy) {
                   1666: #ifdef GATEWAY
1.64      thorpej  1667:                                if (mcopy->m_flags & M_CANFASTFWD)
                   1668:                                        ipflow_create(&ipforward_rt, mcopy);
1.63      matt     1669: #endif
1.1       cgd      1670:                                m_freem(mcopy);
1.63      matt     1671:                        }
1.1       cgd      1672:                        return;
                   1673:                }
                   1674:        }
                   1675:        if (mcopy == NULL)
                   1676:                return;
1.13      mycroft  1677:        destifp = NULL;
                   1678:
1.1       cgd      1679:        switch (error) {
                   1680:
                   1681:        case 0:                         /* forwarded, but need redirect */
                   1682:                /* type, code set above */
                   1683:                break;
                   1684:
                   1685:        case ENETUNREACH:               /* shouldn't happen, checked above */
                   1686:        case EHOSTUNREACH:
                   1687:        case ENETDOWN:
                   1688:        case EHOSTDOWN:
                   1689:        default:
                   1690:                type = ICMP_UNREACH;
                   1691:                code = ICMP_UNREACH_HOST;
                   1692:                break;
                   1693:
                   1694:        case EMSGSIZE:
                   1695:                type = ICMP_UNREACH;
                   1696:                code = ICMP_UNREACH_NEEDFRAG;
1.89      itojun   1697: #ifndef IPSEC
1.13      mycroft  1698:                if (ipforward_rt.ro_rt)
                   1699:                        destifp = ipforward_rt.ro_rt->rt_ifp;
1.89      itojun   1700: #else
                   1701:                /*
                   1702:                 * If the packet is routed over IPsec tunnel, tell the
                   1703:                 * originator the tunnel MTU.
                   1704:                 *      tunnel MTU = if MTU - sizeof(IP) - ESP/AH hdrsiz
                   1705:                 * XXX quickhack!!!
                   1706:                 */
                   1707:                if (ipforward_rt.ro_rt) {
                   1708:                        struct secpolicy *sp;
                   1709:                        int ipsecerror;
1.95      itojun   1710:                        size_t ipsechdr;
1.89      itojun   1711:                        struct route *ro;
                   1712:
                   1713:                        sp = ipsec4_getpolicybyaddr(mcopy,
1.95      itojun   1714:                                                    IPSEC_DIR_OUTBOUND,
                   1715:                                                    IP_FORWARDING,
                   1716:                                                    &ipsecerror);
1.89      itojun   1717:
                   1718:                        if (sp == NULL)
                   1719:                                destifp = ipforward_rt.ro_rt->rt_ifp;
                   1720:                        else {
                   1721:                                /* count IPsec header size */
1.95      itojun   1722:                                ipsechdr = ipsec4_hdrsiz(mcopy,
                   1723:                                                         IPSEC_DIR_OUTBOUND,
                   1724:                                                         NULL);
1.89      itojun   1725:
                   1726:                                /*
                   1727:                                 * find the correct route for outer IPv4
                   1728:                                 * header, compute tunnel MTU.
                   1729:                                 *
                   1730:                                 * XXX BUG ALERT
                   1731:                                 * The "dummyifp" code relies upon the fact
                   1732:                                 * that icmp_error() touches only ifp->if_mtu.
                   1733:                                 */
                   1734:                                /*XXX*/
                   1735:                                destifp = NULL;
                   1736:                                if (sp->req != NULL
1.95      itojun   1737:                                 && sp->req->sav != NULL
                   1738:                                 && sp->req->sav->sah != NULL) {
                   1739:                                        ro = &sp->req->sav->sah->sa_route;
1.89      itojun   1740:                                        if (ro->ro_rt && ro->ro_rt->rt_ifp) {
                   1741:                                                dummyifp.if_mtu =
1.151     itojun   1742:                                                    ro->ro_rt->rt_rmx.rmx_mtu ?
                   1743:                                                    ro->ro_rt->rt_rmx.rmx_mtu :
1.89      itojun   1744:                                                    ro->ro_rt->rt_ifp->if_mtu;
                   1745:                                                dummyifp.if_mtu -= ipsechdr;
                   1746:                                                destifp = &dummyifp;
                   1747:                                        }
                   1748:                                }
                   1749:
                   1750:                                key_freesp(sp);
                   1751:                        }
                   1752:                }
                   1753: #endif /*IPSEC*/
1.1       cgd      1754:                ipstat.ips_cantfrag++;
                   1755:                break;
                   1756:
                   1757:        case ENOBUFS:
1.143     itojun   1758: #if 1
                   1759:                /*
                   1760:                 * a router should not generate ICMP_SOURCEQUENCH as
                   1761:                 * required in RFC1812 Requirements for IP Version 4 Routers.
                   1762:                 * source quench could be a big problem under DoS attacks,
1.149     wiz      1763:                 * or if the underlying interface is rate-limited.
1.143     itojun   1764:                 */
                   1765:                if (mcopy)
                   1766:                        m_freem(mcopy);
                   1767:                return;
                   1768: #else
1.1       cgd      1769:                type = ICMP_SOURCEQUENCH;
                   1770:                code = 0;
                   1771:                break;
1.143     itojun   1772: #endif
1.1       cgd      1773:        }
1.13      mycroft  1774:        icmp_error(mcopy, type, code, dest, destifp);
1.44      thorpej  1775: }
                   1776:
                   1777: void
                   1778: ip_savecontrol(inp, mp, ip, m)
1.109     augustss 1779:        struct inpcb *inp;
                   1780:        struct mbuf **mp;
                   1781:        struct ip *ip;
                   1782:        struct mbuf *m;
1.44      thorpej  1783: {
                   1784:
                   1785:        if (inp->inp_socket->so_options & SO_TIMESTAMP) {
                   1786:                struct timeval tv;
                   1787:
                   1788:                microtime(&tv);
                   1789:                *mp = sbcreatecontrol((caddr_t) &tv, sizeof(tv),
                   1790:                    SCM_TIMESTAMP, SOL_SOCKET);
                   1791:                if (*mp)
                   1792:                        mp = &(*mp)->m_next;
                   1793:        }
                   1794:        if (inp->inp_flags & INP_RECVDSTADDR) {
                   1795:                *mp = sbcreatecontrol((caddr_t) &ip->ip_dst,
                   1796:                    sizeof(struct in_addr), IP_RECVDSTADDR, IPPROTO_IP);
                   1797:                if (*mp)
                   1798:                        mp = &(*mp)->m_next;
                   1799:        }
                   1800: #ifdef notyet
                   1801:        /*
                   1802:         * XXX
                   1803:         * Moving these out of udp_input() made them even more broken
                   1804:         * than they already were.
                   1805:         *      - fenner@parc.xerox.com
                   1806:         */
                   1807:        /* options were tossed already */
                   1808:        if (inp->inp_flags & INP_RECVOPTS) {
                   1809:                *mp = sbcreatecontrol((caddr_t) opts_deleted_above,
                   1810:                    sizeof(struct in_addr), IP_RECVOPTS, IPPROTO_IP);
                   1811:                if (*mp)
                   1812:                        mp = &(*mp)->m_next;
                   1813:        }
                   1814:        /* ip_srcroute doesn't do what we want here, need to fix */
                   1815:        if (inp->inp_flags & INP_RECVRETOPTS) {
                   1816:                *mp = sbcreatecontrol((caddr_t) ip_srcroute(),
                   1817:                    sizeof(struct in_addr), IP_RECVRETOPTS, IPPROTO_IP);
                   1818:                if (*mp)
                   1819:                        mp = &(*mp)->m_next;
                   1820:        }
                   1821: #endif
                   1822:        if (inp->inp_flags & INP_RECVIF) {
                   1823:                struct sockaddr_dl sdl;
                   1824:
                   1825:                sdl.sdl_len = offsetof(struct sockaddr_dl, sdl_data[0]);
                   1826:                sdl.sdl_family = AF_LINK;
                   1827:                sdl.sdl_index = m->m_pkthdr.rcvif ?
                   1828:                    m->m_pkthdr.rcvif->if_index : 0;
                   1829:                sdl.sdl_nlen = sdl.sdl_alen = sdl.sdl_slen = 0;
                   1830:                *mp = sbcreatecontrol((caddr_t) &sdl, sdl.sdl_len,
                   1831:                    IP_RECVIF, IPPROTO_IP);
                   1832:                if (*mp)
                   1833:                        mp = &(*mp)->m_next;
                   1834:        }
1.13      mycroft  1835: }
                   1836:
                   1837: int
                   1838: ip_sysctl(name, namelen, oldp, oldlenp, newp, newlen)
                   1839:        int *name;
                   1840:        u_int namelen;
                   1841:        void *oldp;
                   1842:        size_t *oldlenp;
                   1843:        void *newp;
                   1844:        size_t newlen;
                   1845: {
1.88      sommerfe 1846:        extern int subnetsarelocal, hostzeroisbroadcast;
1.52      thorpej  1847:
1.54      lukem    1848:        int error, old;
                   1849:
1.13      mycroft  1850:        /* All sysctl names at this level are terminal. */
                   1851:        if (namelen != 1)
                   1852:                return (ENOTDIR);
                   1853:
                   1854:        switch (name[0]) {
                   1855:        case IPCTL_FORWARDING:
                   1856:                return (sysctl_int(oldp, oldlenp, newp, newlen, &ipforwarding));
                   1857:        case IPCTL_SENDREDIRECTS:
                   1858:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1859:                        &ipsendredirects));
                   1860:        case IPCTL_DEFTTL:
                   1861:                return (sysctl_int(oldp, oldlenp, newp, newlen, &ip_defttl));
                   1862: #ifdef notyet
                   1863:        case IPCTL_DEFMTU:
                   1864:                return (sysctl_int(oldp, oldlenp, newp, newlen, &ip_mtu));
                   1865: #endif
1.26      thorpej  1866:        case IPCTL_FORWSRCRT:
1.47      cjs      1867:                /* Don't allow this to change in a secure environment.  */
1.26      thorpej  1868:                if (securelevel > 0)
1.46      cjs      1869:                        return (sysctl_rdint(oldp, oldlenp, newp,
                   1870:                            ip_forwsrcrt));
                   1871:                else
                   1872:                        return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1873:                            &ip_forwsrcrt));
1.27      thorpej  1874:        case IPCTL_DIRECTEDBCAST:
                   1875:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1876:                    &ip_directedbcast));
1.47      cjs      1877:        case IPCTL_ALLOWSRCRT:
                   1878:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1879:                    &ip_allowsrcrt));
1.52      thorpej  1880:        case IPCTL_SUBNETSARELOCAL:
                   1881:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1882:                    &subnetsarelocal));
1.53      kml      1883:        case IPCTL_MTUDISC:
1.60      kml      1884:                error = sysctl_int(oldp, oldlenp, newp, newlen,
                   1885:                    &ip_mtudisc);
1.163     itojun   1886:                if (error == 0 && ip_mtudisc == 0)
                   1887:                        rt_timer_queue_remove_all(ip_mtudisc_timeout_q, TRUE);
1.60      kml      1888:                return error;
1.54      lukem    1889:        case IPCTL_ANONPORTMIN:
                   1890:                old = anonportmin;
                   1891:                error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmin);
1.118     itojun   1892:                if (anonportmin >= anonportmax || anonportmin < 0
                   1893:                    || anonportmin > 65535
1.54      lukem    1894: #ifndef IPNOPRIVPORTS
                   1895:                    || anonportmin < IPPORT_RESERVED
                   1896: #endif
                   1897:                    ) {
                   1898:                        anonportmin = old;
                   1899:                        return (EINVAL);
                   1900:                }
                   1901:                return (error);
                   1902:        case IPCTL_ANONPORTMAX:
                   1903:                old = anonportmax;
                   1904:                error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmax);
1.118     itojun   1905:                if (anonportmin >= anonportmax || anonportmax < 0
                   1906:                    || anonportmax > 65535
1.54      lukem    1907: #ifndef IPNOPRIVPORTS
                   1908:                    || anonportmax < IPPORT_RESERVED
                   1909: #endif
                   1910:                    ) {
                   1911:                        anonportmax = old;
                   1912:                        return (EINVAL);
                   1913:                }
1.60      kml      1914:                return (error);
                   1915:        case IPCTL_MTUDISCTIMEOUT:
1.158     itojun   1916:                old = ip_mtudisc_timeout;
1.60      kml      1917:                error = sysctl_int(oldp, oldlenp, newp, newlen,
                   1918:                   &ip_mtudisc_timeout);
1.158     itojun   1919:                if (ip_mtudisc_timeout < 0) {
                   1920:                        ip_mtudisc_timeout = old;
1.156     itojun   1921:                        return (EINVAL);
1.158     itojun   1922:                }
1.163     itojun   1923:                if (error == 0)
1.162     itojun   1924:                        rt_timer_queue_change(ip_mtudisc_timeout_q,
                   1925:                                              ip_mtudisc_timeout);
1.54      lukem    1926:                return (error);
1.65      matt     1927: #ifdef GATEWAY
                   1928:        case IPCTL_MAXFLOWS:
1.67      thorpej  1929:            {
                   1930:                int s;
                   1931:
1.65      matt     1932:                error = sysctl_int(oldp, oldlenp, newp, newlen,
                   1933:                   &ip_maxflows);
1.67      thorpej  1934:                s = splsoftnet();
1.65      matt     1935:                ipflow_reap(0);
1.67      thorpej  1936:                splx(s);
1.65      matt     1937:                return (error);
1.67      thorpej  1938:            }
1.89      itojun   1939: #endif
1.90      itojun   1940:        case IPCTL_HOSTZEROBROADCAST:
                   1941:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1942:                    &hostzeroisbroadcast));
1.89      itojun   1943: #if NGIF > 0
                   1944:        case IPCTL_GIF_TTL:
1.157     itojun   1945:                return (sysctl_int(oldp, oldlenp, newp, newlen,
1.90      itojun   1946:                                  &ip_gif_ttl));
1.144     martin   1947: #endif
                   1948:
                   1949: #if NGRE > 0
                   1950:        case IPCTL_GRE_TTL:
1.157     itojun   1951:                return (sysctl_int(oldp, oldlenp, newp, newlen,
1.144     martin   1952:                                  &ip_gre_ttl));
1.117     tron     1953: #endif
                   1954:
                   1955: #ifndef IPNOPRIVPORTS
                   1956:        case IPCTL_LOWPORTMIN:
                   1957:                old = lowportmin;
                   1958:                error = sysctl_int(oldp, oldlenp, newp, newlen, &lowportmin);
                   1959:                if (lowportmin >= lowportmax
                   1960:                    || lowportmin > IPPORT_RESERVEDMAX
                   1961:                    || lowportmin < IPPORT_RESERVEDMIN
                   1962:                    ) {
                   1963:                        lowportmin = old;
                   1964:                        return (EINVAL);
                   1965:                }
                   1966:                return (error);
                   1967:        case IPCTL_LOWPORTMAX:
                   1968:                old = lowportmax;
                   1969:                error = sysctl_int(oldp, oldlenp, newp, newlen, &lowportmax);
                   1970:                if (lowportmin >= lowportmax
                   1971:                    || lowportmax > IPPORT_RESERVEDMAX
                   1972:                    || lowportmax < IPPORT_RESERVEDMIN
                   1973:                    ) {
                   1974:                        lowportmax = old;
                   1975:                        return (EINVAL);
                   1976:                }
                   1977:                return (error);
1.65      matt     1978: #endif
1.131     itojun   1979:
                   1980:        case IPCTL_MAXFRAGPACKETS:
                   1981:                return (sysctl_int(oldp, oldlenp, newp, newlen,
                   1982:                    &ip_maxfragpackets));
1.88      sommerfe 1983:
1.165   ! christos 1984:        case IPCTL_CHECKINTERFACE:
        !          1985:                return (sysctl_int(oldp, oldlenp, newp, newlen,
        !          1986:                    &ip_checkinterface));
1.13      mycroft  1987:        default:
                   1988:                return (EOPNOTSUPP);
                   1989:        }
                   1990:        /* NOTREACHED */
1.1       cgd      1991: }

CVSweb <webmaster@jp.NetBSD.org>