[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 1.250.6.2

1.250.6.2! dyoung      1: /*     $NetBSD: ip_input.c,v 1.250 2007/07/19 20:48:55 dyoung Exp $    */
        !             2:
        !             3: /*
        !             4:  * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
        !             5:  * All rights reserved.
        !             6:  *
        !             7:  * Redistribution and use in source and binary forms, with or without
        !             8:  * modification, are permitted provided that the following conditions
        !             9:  * are met:
        !            10:  * 1. Redistributions of source code must retain the above copyright
        !            11:  *    notice, this list of conditions and the following disclaimer.
        !            12:  * 2. Redistributions in binary form must reproduce the above copyright
        !            13:  *    notice, this list of conditions and the following disclaimer in the
        !            14:  *    documentation and/or other materials provided with the distribution.
        !            15:  * 3. 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.
        !            18:  *
        !            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:  */
        !            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:  */
        !            68:
        !            69: /*
        !            70:  * Copyright (c) 1982, 1986, 1988, 1993
        !            71:  *     The Regents of the University of California.  All rights reserved.
        !            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. Neither the name of the University nor the names of its contributors
        !            82:  *    may be used to endorse or promote products derived from this software
        !            83:  *    without specific prior written permission.
        !            84:  *
        !            85:  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
        !            86:  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
        !            87:  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
        !            88:  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
        !            89:  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
        !            90:  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
        !            91:  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
        !            92:  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
        !            93:  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
        !            94:  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
        !            95:  * SUCH DAMAGE.
        !            96:  *
        !            97:  *     @(#)ip_input.c  8.2 (Berkeley) 1/4/94
        !            98:  */
        !            99:
        !           100: #include <sys/cdefs.h>
        !           101: __KERNEL_RCSID(0, "$NetBSD: ip_input.c,v 1.250 2007/07/19 20:48:55 dyoung Exp $");
        !           102:
        !           103: #include "opt_inet.h"
        !           104: #include "opt_gateway.h"
        !           105: #include "opt_pfil_hooks.h"
        !           106: #include "opt_ipsec.h"
        !           107: #include "opt_mrouting.h"
        !           108: #include "opt_mbuftrace.h"
        !           109: #include "opt_inet_csum.h"
        !           110:
        !           111: #include <sys/param.h>
        !           112: #include <sys/systm.h>
        !           113: #include <sys/malloc.h>
        !           114: #include <sys/mbuf.h>
        !           115: #include <sys/domain.h>
        !           116: #include <sys/protosw.h>
        !           117: #include <sys/socket.h>
        !           118: #include <sys/socketvar.h>
        !           119: #include <sys/errno.h>
        !           120: #include <sys/time.h>
        !           121: #include <sys/kernel.h>
        !           122: #include <sys/pool.h>
        !           123: #include <sys/sysctl.h>
        !           124: #include <sys/kauth.h>
        !           125:
        !           126: #include <net/if.h>
        !           127: #include <net/if_dl.h>
        !           128: #include <net/route.h>
        !           129: #include <net/pfil.h>
        !           130:
        !           131: #include <netinet/in.h>
        !           132: #include <netinet/in_systm.h>
        !           133: #include <netinet/ip.h>
        !           134: #include <netinet/in_pcb.h>
        !           135: #include <netinet/in_proto.h>
        !           136: #include <netinet/in_var.h>
        !           137: #include <netinet/ip_var.h>
        !           138: #include <netinet/ip_icmp.h>
        !           139: /* just for gif_ttl */
        !           140: #include <netinet/in_gif.h>
        !           141: #include "gif.h"
        !           142: #include <net/if_gre.h>
        !           143: #include "gre.h"
        !           144:
        !           145: #ifdef MROUTING
        !           146: #include <netinet/ip_mroute.h>
        !           147: #endif
        !           148:
        !           149: #ifdef IPSEC
        !           150: #include <netinet6/ipsec.h>
        !           151: #include <netkey/key.h>
        !           152: #endif
        !           153: #ifdef FAST_IPSEC
        !           154: #include <netipsec/ipsec.h>
        !           155: #include <netipsec/key.h>
        !           156: #endif /* FAST_IPSEC*/
        !           157:
        !           158: #ifndef        IPFORWARDING
        !           159: #ifdef GATEWAY
        !           160: #define        IPFORWARDING    1       /* forward IP packets not for us */
        !           161: #else /* GATEWAY */
        !           162: #define        IPFORWARDING    0       /* don't forward IP packets not for us */
        !           163: #endif /* GATEWAY */
        !           164: #endif /* IPFORWARDING */
        !           165: #ifndef        IPSENDREDIRECTS
        !           166: #define        IPSENDREDIRECTS 1
        !           167: #endif
        !           168: #ifndef IPFORWSRCRT
        !           169: #define        IPFORWSRCRT     1       /* forward source-routed packets */
        !           170: #endif
        !           171: #ifndef IPALLOWSRCRT
        !           172: #define        IPALLOWSRCRT    1       /* allow source-routed packets */
        !           173: #endif
        !           174: #ifndef IPMTUDISC
        !           175: #define IPMTUDISC      1
        !           176: #endif
        !           177: #ifndef IPMTUDISCTIMEOUT
        !           178: #define IPMTUDISCTIMEOUT (10 * 60)     /* as per RFC 1191 */
        !           179: #endif
        !           180:
        !           181: /*
        !           182:  * Note: DIRECTED_BROADCAST is handled this way so that previous
        !           183:  * configuration using this option will Just Work.
        !           184:  */
        !           185: #ifndef IPDIRECTEDBCAST
        !           186: #ifdef DIRECTED_BROADCAST
        !           187: #define IPDIRECTEDBCAST        1
        !           188: #else
        !           189: #define        IPDIRECTEDBCAST 0
        !           190: #endif /* DIRECTED_BROADCAST */
        !           191: #endif /* IPDIRECTEDBCAST */
        !           192: int    ipforwarding = IPFORWARDING;
        !           193: int    ipsendredirects = IPSENDREDIRECTS;
        !           194: int    ip_defttl = IPDEFTTL;
        !           195: int    ip_forwsrcrt = IPFORWSRCRT;
        !           196: int    ip_directedbcast = IPDIRECTEDBCAST;
        !           197: int    ip_allowsrcrt = IPALLOWSRCRT;
        !           198: int    ip_mtudisc = IPMTUDISC;
        !           199: int    ip_mtudisc_timeout = IPMTUDISCTIMEOUT;
        !           200: #ifdef DIAGNOSTIC
        !           201: int    ipprintfs = 0;
        !           202: #endif
        !           203:
        !           204: int    ip_do_randomid = 0;
        !           205:
        !           206: /*
        !           207:  * XXX - Setting ip_checkinterface mostly implements the receive side of
        !           208:  * the Strong ES model described in RFC 1122, but since the routing table
        !           209:  * and transmit implementation do not implement the Strong ES model,
        !           210:  * setting this to 1 results in an odd hybrid.
        !           211:  *
        !           212:  * XXX - ip_checkinterface currently must be disabled if you use ipnat
        !           213:  * to translate the destination address to another local interface.
        !           214:  *
        !           215:  * XXX - ip_checkinterface must be disabled if you add IP aliases
        !           216:  * to the loopback interface instead of the interface where the
        !           217:  * packets for those addresses are received.
        !           218:  */
        !           219: int    ip_checkinterface = 0;
        !           220:
        !           221:
        !           222: struct rttimer_queue *ip_mtudisc_timeout_q = NULL;
        !           223:
        !           224: int    ipqmaxlen = IFQ_MAXLEN;
        !           225: u_long in_ifaddrhash;                          /* size of hash table - 1 */
        !           226: int    in_ifaddrentries;                       /* total number of addrs */
        !           227: struct in_ifaddrhead in_ifaddrhead;
        !           228: struct in_ifaddrhashhead *in_ifaddrhashtbl;
        !           229: u_long in_multihash;                           /* size of hash table - 1 */
        !           230: int    in_multientries;                        /* total number of addrs */
        !           231: struct in_multihashhead *in_multihashtbl;
        !           232: struct ifqueue ipintrq;
        !           233: struct ipstat  ipstat;
        !           234: uint16_t ip_id;
        !           235:
        !           236: #ifdef PFIL_HOOKS
        !           237: struct pfil_head inet_pfil_hook;
        !           238: #endif
        !           239:
        !           240: /*
        !           241:  * Cached copy of nmbclusters. If nbclusters is different,
        !           242:  * recalculate IP parameters derived from nmbclusters.
        !           243:  */
        !           244: static int     ip_nmbclusters;                 /* copy of nmbclusters */
        !           245: static void    ip_nmbclusters_changed(void);   /* recalc limits */
        !           246:
        !           247: #define CHECK_NMBCLUSTER_PARAMS()                              \
        !           248: do {                                                           \
        !           249:        if (__predict_false(ip_nmbclusters != nmbclusters))     \
        !           250:                ip_nmbclusters_changed();                       \
        !           251: } while (/*CONSTCOND*/0)
        !           252:
        !           253: /* IP datagram reassembly queues (hashed) */
        !           254: #define IPREASS_NHASH_LOG2      6
        !           255: #define IPREASS_NHASH           (1 << IPREASS_NHASH_LOG2)
        !           256: #define IPREASS_HMASK           (IPREASS_NHASH - 1)
        !           257: #define IPREASS_HASH(x,y) \
        !           258:        (((((x) & 0xF) | ((((x) >> 8) & 0xF) << 4)) ^ (y)) & IPREASS_HMASK)
        !           259: struct ipqhead ipq[IPREASS_NHASH];
        !           260: int    ipq_locked;
        !           261: static int     ip_nfragpackets;        /* packets in reass queue */
        !           262: static int     ip_nfrags;              /* total fragments in reass queues */
        !           263:
        !           264: int    ip_maxfragpackets = 200;        /* limit on packets. XXX sysctl */
        !           265: int    ip_maxfrags;                    /* limit on fragments. XXX sysctl */
        !           266:
        !           267:
        !           268: /*
        !           269:  * Additive-Increase/Multiplicative-Decrease (AIMD) strategy for
        !           270:  * IP reassembly queue buffer managment.
        !           271:  *
        !           272:  * We keep a count of total IP fragments (NB: not fragmented packets!)
        !           273:  * awaiting reassembly (ip_nfrags) and a limit (ip_maxfrags) on fragments.
        !           274:  * If ip_nfrags exceeds ip_maxfrags the limit, we drop half the
        !           275:  * total fragments in  reassembly queues.This AIMD policy avoids
        !           276:  * repeatedly deleting single packets under heavy fragmentation load
        !           277:  * (e.g., from lossy NFS peers).
        !           278:  */
        !           279: static u_int   ip_reass_ttl_decr(u_int ticks);
        !           280: static void    ip_reass_drophalf(void);
        !           281:
        !           282:
        !           283: static inline int ipq_lock_try(void);
        !           284: static inline void ipq_unlock(void);
        !           285:
        !           286: static inline int
        !           287: ipq_lock_try(void)
        !           288: {
        !           289:        int s;
        !           290:
        !           291:        /*
        !           292:         * Use splvm() -- we're blocking things that would cause
        !           293:         * mbuf allocation.
        !           294:         */
        !           295:        s = splvm();
        !           296:        if (ipq_locked) {
        !           297:                splx(s);
        !           298:                return (0);
        !           299:        }
        !           300:        ipq_locked = 1;
        !           301:        splx(s);
        !           302:        return (1);
        !           303: }
        !           304:
        !           305: static inline void
        !           306: ipq_unlock(void)
        !           307: {
        !           308:        int s;
        !           309:
        !           310:        s = splvm();
        !           311:        ipq_locked = 0;
        !           312:        splx(s);
        !           313: }
        !           314:
        !           315: #ifdef DIAGNOSTIC
        !           316: #define        IPQ_LOCK()                                                      \
        !           317: do {                                                                   \
        !           318:        if (ipq_lock_try() == 0) {                                      \
        !           319:                printf("%s:%d: ipq already locked\n", __FILE__, __LINE__); \
        !           320:                panic("ipq_lock");                                      \
        !           321:        }                                                               \
        !           322: } while (/*CONSTCOND*/ 0)
        !           323: #define        IPQ_LOCK_CHECK()                                                \
        !           324: do {                                                                   \
        !           325:        if (ipq_locked == 0) {                                          \
        !           326:                printf("%s:%d: ipq lock not held\n", __FILE__, __LINE__); \
        !           327:                panic("ipq lock check");                                \
        !           328:        }                                                               \
        !           329: } while (/*CONSTCOND*/ 0)
        !           330: #else
        !           331: #define        IPQ_LOCK()              (void) ipq_lock_try()
        !           332: #define        IPQ_LOCK_CHECK()        /* nothing */
        !           333: #endif
        !           334:
        !           335: #define        IPQ_UNLOCK()            ipq_unlock()
        !           336:
        !           337: POOL_INIT(inmulti_pool, sizeof(struct in_multi), 0, 0, 0, "inmltpl", NULL,
        !           338:     IPL_SOFTNET);
        !           339: POOL_INIT(ipqent_pool, sizeof(struct ipqent), 0, 0, 0, "ipqepl", NULL,
        !           340:     IPL_VM);
        !           341:
        !           342: #ifdef INET_CSUM_COUNTERS
        !           343: #include <sys/device.h>
        !           344:
        !           345: struct evcnt ip_hwcsum_bad = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
        !           346:     NULL, "inet", "hwcsum bad");
        !           347: struct evcnt ip_hwcsum_ok = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
        !           348:     NULL, "inet", "hwcsum ok");
        !           349: struct evcnt ip_swcsum = EVCNT_INITIALIZER(EVCNT_TYPE_MISC,
        !           350:     NULL, "inet", "swcsum");
        !           351:
        !           352: #define        INET_CSUM_COUNTER_INCR(ev)      (ev)->ev_count++
        !           353:
        !           354: EVCNT_ATTACH_STATIC(ip_hwcsum_bad);
        !           355: EVCNT_ATTACH_STATIC(ip_hwcsum_ok);
        !           356: EVCNT_ATTACH_STATIC(ip_swcsum);
        !           357:
        !           358: #else
        !           359:
        !           360: #define        INET_CSUM_COUNTER_INCR(ev)      /* nothing */
        !           361:
        !           362: #endif /* INET_CSUM_COUNTERS */
        !           363:
        !           364: /*
        !           365:  * We need to save the IP options in case a protocol wants to respond
        !           366:  * to an incoming packet over the same route if the packet got here
        !           367:  * using IP source routing.  This allows connection establishment and
        !           368:  * maintenance when the remote end is on a network that is not known
        !           369:  * to us.
        !           370:  */
        !           371: int    ip_nhops = 0;
        !           372: static struct ip_srcrt {
        !           373:        struct  in_addr dst;                    /* final destination */
        !           374:        char    nop;                            /* one NOP to align */
        !           375:        char    srcopt[IPOPT_OFFSET + 1];       /* OPTVAL, OLEN and OFFSET */
        !           376:        struct  in_addr route[MAX_IPOPTLEN/sizeof(struct in_addr)];
        !           377: } ip_srcrt;
        !           378:
        !           379: static void save_rte(u_char *, struct in_addr);
        !           380:
        !           381: #ifdef MBUFTRACE
        !           382: struct mowner ip_rx_mowner = MOWNER_INIT("internet", "rx");
        !           383: struct mowner ip_tx_mowner = MOWNER_INIT("internet", "tx");
        !           384: #endif
        !           385:
        !           386: /*
        !           387:  * Compute IP limits derived from the value of nmbclusters.
        !           388:  */
        !           389: static void
        !           390: ip_nmbclusters_changed(void)
        !           391: {
        !           392:        ip_maxfrags = nmbclusters / 4;
        !           393:        ip_nmbclusters =  nmbclusters;
        !           394: }
        !           395:
        !           396: /*
        !           397:  * IP initialization: fill in IP protocol switch table.
        !           398:  * All protocols not implemented in kernel go to raw IP protocol handler.
        !           399:  */
        !           400: void
        !           401: ip_init(void)
        !           402: {
        !           403:        const struct protosw *pr;
        !           404:        int i;
        !           405:
        !           406:        pr = pffindproto(PF_INET, IPPROTO_RAW, SOCK_RAW);
        !           407:        if (pr == 0)
        !           408:                panic("ip_init");
        !           409:        for (i = 0; i < IPPROTO_MAX; i++)
        !           410:                ip_protox[i] = pr - inetsw;
        !           411:        for (pr = inetdomain.dom_protosw;
        !           412:            pr < inetdomain.dom_protoswNPROTOSW; pr++)
        !           413:                if (pr->pr_domain->dom_family == PF_INET &&
        !           414:                    pr->pr_protocol && pr->pr_protocol != IPPROTO_RAW)
        !           415:                        ip_protox[pr->pr_protocol] = pr - inetsw;
        !           416:
        !           417:        for (i = 0; i < IPREASS_NHASH; i++)
        !           418:                LIST_INIT(&ipq[i]);
        !           419:
        !           420:        ip_id = time_second & 0xfffff;
        !           421:
        !           422:        ipintrq.ifq_maxlen = ipqmaxlen;
        !           423:        ip_nmbclusters_changed();
        !           424:
        !           425:        TAILQ_INIT(&in_ifaddrhead);
        !           426:        in_ifaddrhashtbl = hashinit(IN_IFADDR_HASH_SIZE, HASH_LIST, M_IFADDR,
        !           427:            M_WAITOK, &in_ifaddrhash);
        !           428:        in_multihashtbl = hashinit(IN_IFADDR_HASH_SIZE, HASH_LIST, M_IPMADDR,
        !           429:            M_WAITOK, &in_multihash);
        !           430:        ip_mtudisc_timeout_q = rt_timer_queue_create(ip_mtudisc_timeout);
        !           431: #ifdef GATEWAY
        !           432:        ipflow_init(ip_hashsize);
        !           433: #endif
        !           434:
        !           435: #ifdef PFIL_HOOKS
        !           436:        /* Register our Packet Filter hook. */
        !           437:        inet_pfil_hook.ph_type = PFIL_TYPE_AF;
        !           438:        inet_pfil_hook.ph_af   = AF_INET;
        !           439:        i = pfil_head_register(&inet_pfil_hook);
        !           440:        if (i != 0)
        !           441:                printf("ip_init: WARNING: unable to register pfil hook, "
        !           442:                    "error %d\n", i);
        !           443: #endif /* PFIL_HOOKS */
        !           444:
        !           445: #ifdef MBUFTRACE
        !           446:        MOWNER_ATTACH(&ip_tx_mowner);
        !           447:        MOWNER_ATTACH(&ip_rx_mowner);
        !           448: #endif /* MBUFTRACE */
        !           449: }
        !           450:
        !           451: struct sockaddr_in ipaddr = {
        !           452:        .sin_len = sizeof(ipaddr),
        !           453:        .sin_family = AF_INET,
        !           454: };
        !           455: struct route ipforward_rt;
        !           456:
        !           457: /*
        !           458:  * IP software interrupt routine
        !           459:  */
        !           460: void
        !           461: ipintr(void)
        !           462: {
        !           463:        int s;
        !           464:        struct mbuf *m;
        !           465:
        !           466:        while (!IF_IS_EMPTY(&ipintrq)) {
        !           467:                s = splnet();
        !           468:                IF_DEQUEUE(&ipintrq, m);
        !           469:                splx(s);
        !           470:                if (m == 0)
        !           471:                        return;
        !           472:                MCLAIM(m, &ip_rx_mowner);
        !           473:                ip_input(m);
        !           474:        }
        !           475: }
        !           476:
        !           477: /*
        !           478:  * Ip input routine.  Checksum and byte swap header.  If fragmented
        !           479:  * try to reassemble.  Process options.  Pass to next level.
        !           480:  */
        !           481: void
        !           482: ip_input(struct mbuf *m)
        !           483: {
        !           484:        struct ip *ip = NULL;
        !           485:        struct ipq *fp;
        !           486:        struct in_ifaddr *ia;
        !           487:        struct ifaddr *ifa;
        !           488:        struct ipqent *ipqe;
        !           489:        int hlen = 0, mff, len;
        !           490:        int downmatch;
        !           491:        int checkif;
        !           492:        int srcrt = 0;
        !           493:        int s;
        !           494:        u_int hash;
        !           495: #ifdef FAST_IPSEC
        !           496:        struct m_tag *mtag;
        !           497:        struct tdb_ident *tdbi;
        !           498:        struct secpolicy *sp;
        !           499:        int error;
        !           500: #endif /* FAST_IPSEC */
        !           501:
        !           502:        MCLAIM(m, &ip_rx_mowner);
        !           503: #ifdef DIAGNOSTIC
        !           504:        if ((m->m_flags & M_PKTHDR) == 0)
        !           505:                panic("ipintr no HDR");
        !           506: #endif
        !           507:
        !           508:        /*
        !           509:         * If no IP addresses have been set yet but the interfaces
        !           510:         * are receiving, can't do anything with incoming packets yet.
        !           511:         */
        !           512:        if (TAILQ_FIRST(&in_ifaddrhead) == 0)
        !           513:                goto bad;
        !           514:        ipstat.ips_total++;
        !           515:        /*
        !           516:         * If the IP header is not aligned, slurp it up into a new
        !           517:         * mbuf with space for link headers, in the event we forward
        !           518:         * it.  Otherwise, if it is aligned, make sure the entire
        !           519:         * base IP header is in the first mbuf of the chain.
        !           520:         */
        !           521:        if (IP_HDR_ALIGNED_P(mtod(m, void *)) == 0) {
        !           522:                if ((m = m_copyup(m, sizeof(struct ip),
        !           523:                                  (max_linkhdr + 3) & ~3)) == NULL) {
        !           524:                        /* XXXJRT new stat, please */
        !           525:                        ipstat.ips_toosmall++;
        !           526:                        return;
        !           527:                }
        !           528:        } else if (__predict_false(m->m_len < sizeof (struct ip))) {
        !           529:                if ((m = m_pullup(m, sizeof (struct ip))) == NULL) {
        !           530:                        ipstat.ips_toosmall++;
        !           531:                        return;
        !           532:                }
        !           533:        }
        !           534:        ip = mtod(m, struct ip *);
        !           535:        if (ip->ip_v != IPVERSION) {
        !           536:                ipstat.ips_badvers++;
        !           537:                goto bad;
        !           538:        }
        !           539:        hlen = ip->ip_hl << 2;
        !           540:        if (hlen < sizeof(struct ip)) { /* minimum header length */
        !           541:                ipstat.ips_badhlen++;
        !           542:                goto bad;
        !           543:        }
        !           544:        if (hlen > m->m_len) {
        !           545:                if ((m = m_pullup(m, hlen)) == 0) {
        !           546:                        ipstat.ips_badhlen++;
        !           547:                        return;
        !           548:                }
        !           549:                ip = mtod(m, struct ip *);
        !           550:        }
        !           551:
        !           552:        /*
        !           553:         * RFC1122: packets with a multicast source address are
        !           554:         * not allowed.
        !           555:         */
        !           556:        if (IN_MULTICAST(ip->ip_src.s_addr)) {
        !           557:                ipstat.ips_badaddr++;
        !           558:                goto bad;
        !           559:        }
        !           560:
        !           561:        /* 127/8 must not appear on wire - RFC1122 */
        !           562:        if ((ntohl(ip->ip_dst.s_addr) >> IN_CLASSA_NSHIFT) == IN_LOOPBACKNET ||
        !           563:            (ntohl(ip->ip_src.s_addr) >> IN_CLASSA_NSHIFT) == IN_LOOPBACKNET) {
        !           564:                if ((m->m_pkthdr.rcvif->if_flags & IFF_LOOPBACK) == 0) {
        !           565:                        ipstat.ips_badaddr++;
        !           566:                        goto bad;
        !           567:                }
        !           568:        }
        !           569:
        !           570:        switch (m->m_pkthdr.csum_flags &
        !           571:                ((m->m_pkthdr.rcvif->if_csum_flags_rx & M_CSUM_IPv4) |
        !           572:                 M_CSUM_IPv4_BAD)) {
        !           573:        case M_CSUM_IPv4|M_CSUM_IPv4_BAD:
        !           574:                INET_CSUM_COUNTER_INCR(&ip_hwcsum_bad);
        !           575:                goto badcsum;
        !           576:
        !           577:        case M_CSUM_IPv4:
        !           578:                /* Checksum was okay. */
        !           579:                INET_CSUM_COUNTER_INCR(&ip_hwcsum_ok);
        !           580:                break;
        !           581:
        !           582:        default:
        !           583:                /*
        !           584:                 * Must compute it ourselves.  Maybe skip checksum on
        !           585:                 * loopback interfaces.
        !           586:                 */
        !           587:                if (__predict_true(!(m->m_pkthdr.rcvif->if_flags &
        !           588:                                     IFF_LOOPBACK) || ip_do_loopback_cksum)) {
        !           589:                        INET_CSUM_COUNTER_INCR(&ip_swcsum);
        !           590:                        if (in_cksum(m, hlen) != 0)
        !           591:                                goto badcsum;
        !           592:                }
        !           593:                break;
        !           594:        }
        !           595:
        !           596:        /* Retrieve the packet length. */
        !           597:        len = ntohs(ip->ip_len);
        !           598:
        !           599:        /*
        !           600:         * Check for additional length bogosity
        !           601:         */
        !           602:        if (len < hlen) {
        !           603:                ipstat.ips_badlen++;
        !           604:                goto bad;
        !           605:        }
        !           606:
        !           607:        /*
        !           608:         * Check that the amount of data in the buffers
        !           609:         * is as at least much as the IP header would have us expect.
        !           610:         * Trim mbufs if longer than we expect.
        !           611:         * Drop packet if shorter than we expect.
        !           612:         */
        !           613:        if (m->m_pkthdr.len < len) {
        !           614:                ipstat.ips_tooshort++;
        !           615:                goto bad;
        !           616:        }
        !           617:        if (m->m_pkthdr.len > len) {
        !           618:                if (m->m_len == m->m_pkthdr.len) {
        !           619:                        m->m_len = len;
        !           620:                        m->m_pkthdr.len = len;
        !           621:                } else
        !           622:                        m_adj(m, len - m->m_pkthdr.len);
        !           623:        }
        !           624:
        !           625: #if defined(IPSEC)
        !           626:        /* ipflow (IP fast forwarding) is not compatible with IPsec. */
        !           627:        m->m_flags &= ~M_CANFASTFWD;
        !           628: #else
        !           629:        /*
        !           630:         * Assume that we can create a fast-forward IP flow entry
        !           631:         * based on this packet.
        !           632:         */
        !           633:        m->m_flags |= M_CANFASTFWD;
        !           634: #endif
        !           635:
        !           636: #ifdef PFIL_HOOKS
        !           637:        /*
        !           638:         * Run through list of hooks for input packets.  If there are any
        !           639:         * filters which require that additional packets in the flow are
        !           640:         * not fast-forwarded, they must clear the M_CANFASTFWD flag.
        !           641:         * Note that filters must _never_ set this flag, as another filter
        !           642:         * in the list may have previously cleared it.
        !           643:         */
        !           644:        /*
        !           645:         * let ipfilter look at packet on the wire,
        !           646:         * not the decapsulated packet.
        !           647:         */
        !           648: #ifdef IPSEC
        !           649:        if (!ipsec_getnhist(m))
        !           650: #elif defined(FAST_IPSEC)
        !           651:        if (!ipsec_indone(m))
        !           652: #else
        !           653:        if (1)
        !           654: #endif
        !           655:        {
        !           656:                struct in_addr odst;
        !           657:
        !           658:                odst = ip->ip_dst;
        !           659:                if (pfil_run_hooks(&inet_pfil_hook, &m, m->m_pkthdr.rcvif,
        !           660:                    PFIL_IN) != 0)
        !           661:                        return;
        !           662:                if (m == NULL)
        !           663:                        return;
        !           664:                ip = mtod(m, struct ip *);
        !           665:                hlen = ip->ip_hl << 2;
        !           666:                /*
        !           667:                 * XXX The setting of "srcrt" here is to prevent ip_forward()
        !           668:                 * from generating ICMP redirects for packets that have
        !           669:                 * been redirected by a hook back out on to the same LAN that
        !           670:                 * they came from and is not an indication that the packet
        !           671:                 * is being inffluenced by source routing options.  This
        !           672:                 * allows things like
        !           673:                 * "rdr tlp0 0/0 port 80 -> 1.1.1.200 3128 tcp"
        !           674:                 * where tlp0 is both on the 1.1.1.0/24 network and is the
        !           675:                 * default route for hosts on 1.1.1.0/24.  Of course this
        !           676:                 * also requires a "map tlp0 ..." to complete the story.
        !           677:                 * One might argue whether or not this kind of network config.
        !           678:                 * should be supported in this manner...
        !           679:                 */
        !           680:                srcrt = (odst.s_addr != ip->ip_dst.s_addr);
        !           681:        }
        !           682: #endif /* PFIL_HOOKS */
        !           683:
        !           684: #ifdef ALTQ
        !           685:        /* XXX Temporary until ALTQ is changed to use a pfil hook */
        !           686:        if (altq_input != NULL && (*altq_input)(m, AF_INET) == 0) {
        !           687:                /* packet dropped by traffic conditioner */
        !           688:                return;
        !           689:        }
        !           690: #endif
        !           691:
        !           692:        /*
        !           693:         * Process options and, if not destined for us,
        !           694:         * ship it on.  ip_dooptions returns 1 when an
        !           695:         * error was detected (causing an icmp message
        !           696:         * to be sent and the original packet to be freed).
        !           697:         */
        !           698:        ip_nhops = 0;           /* for source routed packets */
        !           699:        if (hlen > sizeof (struct ip) && ip_dooptions(m))
        !           700:                return;
        !           701:
        !           702:        /*
        !           703:         * Enable a consistency check between the destination address
        !           704:         * and the arrival interface for a unicast packet (the RFC 1122
        !           705:         * strong ES model) if IP forwarding is disabled and the packet
        !           706:         * is not locally generated.
        !           707:         *
        !           708:         * XXX - Checking also should be disabled if the destination
        !           709:         * address is ipnat'ed to a different interface.
        !           710:         *
        !           711:         * XXX - Checking is incompatible with IP aliases added
        !           712:         * to the loopback interface instead of the interface where
        !           713:         * the packets are received.
        !           714:         *
        !           715:         * XXX - We need to add a per ifaddr flag for this so that
        !           716:         * we get finer grain control.
        !           717:         */
        !           718:        checkif = ip_checkinterface && (ipforwarding == 0) &&
        !           719:            (m->m_pkthdr.rcvif != NULL) &&
        !           720:            ((m->m_pkthdr.rcvif->if_flags & IFF_LOOPBACK) == 0);
        !           721:
        !           722:        /*
        !           723:         * Check our list of addresses, to see if the packet is for us.
        !           724:         *
        !           725:         * Traditional 4.4BSD did not consult IFF_UP at all.
        !           726:         * The behavior here is to treat addresses on !IFF_UP interface
        !           727:         * as not mine.
        !           728:         */
        !           729:        downmatch = 0;
        !           730:        LIST_FOREACH(ia, &IN_IFADDR_HASH(ip->ip_dst.s_addr), ia_hash) {
        !           731:                if (in_hosteq(ia->ia_addr.sin_addr, ip->ip_dst)) {
        !           732:                        if (checkif && ia->ia_ifp != m->m_pkthdr.rcvif)
        !           733:                                continue;
        !           734:                        if ((ia->ia_ifp->if_flags & IFF_UP) != 0)
        !           735:                                break;
        !           736:                        else
        !           737:                                downmatch++;
        !           738:                }
        !           739:        }
        !           740:        if (ia != NULL)
        !           741:                goto ours;
        !           742:        if (m->m_pkthdr.rcvif && m->m_pkthdr.rcvif->if_flags & IFF_BROADCAST) {
        !           743:                IFADDR_FOREACH(ifa, m->m_pkthdr.rcvif) {
        !           744:                        if (ifa->ifa_addr->sa_family != AF_INET)
        !           745:                                continue;
        !           746:                        ia = ifatoia(ifa);
        !           747:                        if (in_hosteq(ip->ip_dst, ia->ia_broadaddr.sin_addr) ||
        !           748:                            in_hosteq(ip->ip_dst, ia->ia_netbroadcast) ||
        !           749:                            /*
        !           750:                             * Look for all-0's host part (old broadcast addr),
        !           751:                             * either for subnet or net.
        !           752:                             */
        !           753:                            ip->ip_dst.s_addr == ia->ia_subnet ||
        !           754:                            ip->ip_dst.s_addr == ia->ia_net)
        !           755:                                goto ours;
        !           756:                        /*
        !           757:                         * An interface with IP address zero accepts
        !           758:                         * all packets that arrive on that interface.
        !           759:                         */
        !           760:                        if (in_nullhost(ia->ia_addr.sin_addr))
        !           761:                                goto ours;
        !           762:                }
        !           763:        }
        !           764:        if (IN_MULTICAST(ip->ip_dst.s_addr)) {
        !           765:                struct in_multi *inm;
        !           766: #ifdef MROUTING
        !           767:                extern struct socket *ip_mrouter;
        !           768:
        !           769:                if (ip_mrouter) {
        !           770:                        /*
        !           771:                         * If we are acting as a multicast router, all
        !           772:                         * incoming multicast packets are passed to the
        !           773:                         * kernel-level multicast forwarding function.
        !           774:                         * The packet is returned (relatively) intact; if
        !           775:                         * ip_mforward() returns a non-zero value, the packet
        !           776:                         * must be discarded, else it may be accepted below.
        !           777:                         *
        !           778:                         * (The IP ident field is put in the same byte order
        !           779:                         * as expected when ip_mforward() is called from
        !           780:                         * ip_output().)
        !           781:                         */
        !           782:                        if (ip_mforward(m, m->m_pkthdr.rcvif) != 0) {
        !           783:                                ipstat.ips_cantforward++;
        !           784:                                m_freem(m);
        !           785:                                return;
        !           786:                        }
        !           787:
        !           788:                        /*
        !           789:                         * The process-level routing demon needs to receive
        !           790:                         * all multicast IGMP packets, whether or not this
        !           791:                         * host belongs to their destination groups.
        !           792:                         */
        !           793:                        if (ip->ip_p == IPPROTO_IGMP)
        !           794:                                goto ours;
        !           795:                        ipstat.ips_forward++;
        !           796:                }
        !           797: #endif
        !           798:                /*
        !           799:                 * See if we belong to the destination multicast group on the
        !           800:                 * arrival interface.
        !           801:                 */
        !           802:                IN_LOOKUP_MULTI(ip->ip_dst, m->m_pkthdr.rcvif, inm);
        !           803:                if (inm == NULL) {
        !           804:                        ipstat.ips_cantforward++;
        !           805:                        m_freem(m);
        !           806:                        return;
        !           807:                }
        !           808:                goto ours;
        !           809:        }
        !           810:        if (ip->ip_dst.s_addr == INADDR_BROADCAST ||
        !           811:            in_nullhost(ip->ip_dst))
        !           812:                goto ours;
        !           813:
        !           814:        /*
        !           815:         * Not for us; forward if possible and desirable.
        !           816:         */
        !           817:        if (ipforwarding == 0) {
        !           818:                ipstat.ips_cantforward++;
        !           819:                m_freem(m);
        !           820:        } else {
        !           821:                /*
        !           822:                 * If ip_dst matched any of my address on !IFF_UP interface,
        !           823:                 * and there's no IFF_UP interface that matches ip_dst,
        !           824:                 * send icmp unreach.  Forwarding it will result in in-kernel
        !           825:                 * forwarding loop till TTL goes to 0.
        !           826:                 */
        !           827:                if (downmatch) {
        !           828:                        icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_HOST, 0, 0);
        !           829:                        ipstat.ips_cantforward++;
        !           830:                        return;
        !           831:                }
        !           832: #ifdef IPSEC
        !           833:                if (ipsec4_in_reject(m, NULL)) {
        !           834:                        ipsecstat.in_polvio++;
        !           835:                        goto bad;
        !           836:                }
        !           837: #endif
        !           838: #ifdef FAST_IPSEC
        !           839:                mtag = m_tag_find(m, PACKET_TAG_IPSEC_IN_DONE, NULL);
        !           840:                s = splsoftnet();
        !           841:                if (mtag != NULL) {
        !           842:                        tdbi = (struct tdb_ident *)(mtag + 1);
        !           843:                        sp = ipsec_getpolicy(tdbi, IPSEC_DIR_INBOUND);
        !           844:                } else {
        !           845:                        sp = ipsec_getpolicybyaddr(m, IPSEC_DIR_INBOUND,
        !           846:                                                   IP_FORWARDING, &error);
        !           847:                }
        !           848:                if (sp == NULL) {       /* NB: can happen if error */
        !           849:                        splx(s);
        !           850:                        /*XXX error stat???*/
        !           851:                        DPRINTF(("ip_input: no SP for forwarding\n"));  /*XXX*/
        !           852:                        goto bad;
        !           853:                }
        !           854:
        !           855:                /*
        !           856:                 * Check security policy against packet attributes.
        !           857:                 */
        !           858:                error = ipsec_in_reject(sp, m);
        !           859:                KEY_FREESP(&sp);
        !           860:                splx(s);
        !           861:                if (error) {
        !           862:                        ipstat.ips_cantforward++;
        !           863:                        goto bad;
        !           864:                }
        !           865:
        !           866:                /*
        !           867:                 * Peek at the outbound SP for this packet to determine if
        !           868:                 * it's a Fast Forward candidate.
        !           869:                 */
        !           870:                mtag = m_tag_find(m, PACKET_TAG_IPSEC_PENDING_TDB, NULL);
        !           871:                if (mtag != NULL)
        !           872:                        m->m_flags &= ~M_CANFASTFWD;
        !           873:                else {
        !           874:                        s = splsoftnet();
        !           875:                        sp = ipsec4_checkpolicy(m, IPSEC_DIR_OUTBOUND,
        !           876:                            (IP_FORWARDING |
        !           877:                             (ip_directedbcast ? IP_ALLOWBROADCAST : 0)),
        !           878:                            &error, NULL);
        !           879:                        if (sp != NULL) {
        !           880:                                m->m_flags &= ~M_CANFASTFWD;
        !           881:                                KEY_FREESP(&sp);
        !           882:                        }
        !           883:                        splx(s);
        !           884:                }
        !           885: #endif /* FAST_IPSEC */
        !           886:
        !           887:                ip_forward(m, srcrt);
        !           888:        }
        !           889:        return;
        !           890:
        !           891: ours:
        !           892:        /*
        !           893:         * If offset or IP_MF are set, must reassemble.
        !           894:         * Otherwise, nothing need be done.
        !           895:         * (We could look in the reassembly queue to see
        !           896:         * if the packet was previously fragmented,
        !           897:         * but it's not worth the time; just let them time out.)
        !           898:         */
        !           899:        if (ip->ip_off & ~htons(IP_DF|IP_RF)) {
        !           900:
        !           901:                /*
        !           902:                 * Look for queue of fragments
        !           903:                 * of this datagram.
        !           904:                 */
        !           905:                IPQ_LOCK();
        !           906:                hash = IPREASS_HASH(ip->ip_src.s_addr, ip->ip_id);
        !           907:                LIST_FOREACH(fp, &ipq[hash], ipq_q) {
        !           908:                        if (ip->ip_id == fp->ipq_id &&
        !           909:                            in_hosteq(ip->ip_src, fp->ipq_src) &&
        !           910:                            in_hosteq(ip->ip_dst, fp->ipq_dst) &&
        !           911:                            ip->ip_p == fp->ipq_p)
        !           912:                                goto found;
        !           913:
        !           914:                }
        !           915:                fp = 0;
        !           916: found:
        !           917:
        !           918:                /*
        !           919:                 * Adjust ip_len to not reflect header,
        !           920:                 * set ipqe_mff if more fragments are expected,
        !           921:                 * convert offset of this to bytes.
        !           922:                 */
        !           923:                ip->ip_len = htons(ntohs(ip->ip_len) - hlen);
        !           924:                mff = (ip->ip_off & htons(IP_MF)) != 0;
        !           925:                if (mff) {
        !           926:                        /*
        !           927:                         * Make sure that fragments have a data length
        !           928:                         * that's a non-zero multiple of 8 bytes.
        !           929:                         */
        !           930:                        if (ntohs(ip->ip_len) == 0 ||
        !           931:                            (ntohs(ip->ip_len) & 0x7) != 0) {
        !           932:                                ipstat.ips_badfrags++;
        !           933:                                IPQ_UNLOCK();
        !           934:                                goto bad;
        !           935:                        }
        !           936:                }
        !           937:                ip->ip_off = htons((ntohs(ip->ip_off) & IP_OFFMASK) << 3);
        !           938:
        !           939:                /*
        !           940:                 * If datagram marked as having more fragments
        !           941:                 * or if this is not the first fragment,
        !           942:                 * attempt reassembly; if it succeeds, proceed.
        !           943:                 */
        !           944:                if (mff || ip->ip_off != htons(0)) {
        !           945:                        ipstat.ips_fragments++;
        !           946:                        s = splvm();
        !           947:                        ipqe = pool_get(&ipqent_pool, PR_NOWAIT);
        !           948:                        splx(s);
        !           949:                        if (ipqe == NULL) {
        !           950:                                ipstat.ips_rcvmemdrop++;
        !           951:                                IPQ_UNLOCK();
        !           952:                                goto bad;
        !           953:                        }
        !           954:                        ipqe->ipqe_mff = mff;
        !           955:                        ipqe->ipqe_m = m;
        !           956:                        ipqe->ipqe_ip = ip;
        !           957:                        m = ip_reass(ipqe, fp, &ipq[hash]);
        !           958:                        if (m == 0) {
        !           959:                                IPQ_UNLOCK();
        !           960:                                return;
        !           961:                        }
        !           962:                        ipstat.ips_reassembled++;
        !           963:                        ip = mtod(m, struct ip *);
        !           964:                        hlen = ip->ip_hl << 2;
        !           965:                        ip->ip_len = htons(ntohs(ip->ip_len) + hlen);
        !           966:                } else
        !           967:                        if (fp)
        !           968:                                ip_freef(fp);
        !           969:                IPQ_UNLOCK();
        !           970:        }
        !           971:
        !           972: #if defined(IPSEC)
        !           973:        /*
        !           974:         * enforce IPsec policy checking if we are seeing last header.
        !           975:         * note that we do not visit this with protocols with pcb layer
        !           976:         * code - like udp/tcp/raw ip.
        !           977:         */
        !           978:        if ((inetsw[ip_protox[ip->ip_p]].pr_flags & PR_LASTHDR) != 0 &&
        !           979:            ipsec4_in_reject(m, NULL)) {
        !           980:                ipsecstat.in_polvio++;
        !           981:                goto bad;
        !           982:        }
        !           983: #endif
        !           984: #ifdef FAST_IPSEC
        !           985:        /*
        !           986:         * enforce IPsec policy checking if we are seeing last header.
        !           987:         * note that we do not visit this with protocols with pcb layer
        !           988:         * code - like udp/tcp/raw ip.
        !           989:         */
        !           990:        if ((inetsw[ip_protox[ip->ip_p]].pr_flags & PR_LASTHDR) != 0) {
        !           991:                /*
        !           992:                 * Check if the packet has already had IPsec processing
        !           993:                 * done.  If so, then just pass it along.  This tag gets
        !           994:                 * set during AH, ESP, etc. input handling, before the
        !           995:                 * packet is returned to the ip input queue for delivery.
        !           996:                 */
        !           997:                mtag = m_tag_find(m, PACKET_TAG_IPSEC_IN_DONE, NULL);
        !           998:                s = splsoftnet();
        !           999:                if (mtag != NULL) {
        !          1000:                        tdbi = (struct tdb_ident *)(mtag + 1);
        !          1001:                        sp = ipsec_getpolicy(tdbi, IPSEC_DIR_INBOUND);
        !          1002:                } else {
        !          1003:                        sp = ipsec_getpolicybyaddr(m, IPSEC_DIR_INBOUND,
        !          1004:                                                   IP_FORWARDING, &error);
        !          1005:                }
        !          1006:                if (sp != NULL) {
        !          1007:                        /*
        !          1008:                         * Check security policy against packet attributes.
        !          1009:                         */
        !          1010:                        error = ipsec_in_reject(sp, m);
        !          1011:                        KEY_FREESP(&sp);
        !          1012:                } else {
        !          1013:                        /* XXX error stat??? */
        !          1014:                        error = EINVAL;
        !          1015: DPRINTF(("ip_input: no SP, packet discarded\n"));/*XXX*/
        !          1016:                        goto bad;
        !          1017:                }
        !          1018:                splx(s);
        !          1019:                if (error)
        !          1020:                        goto bad;
        !          1021:        }
        !          1022: #endif /* FAST_IPSEC */
        !          1023:
        !          1024:        /*
        !          1025:         * Switch out to protocol's input routine.
        !          1026:         */
        !          1027: #if IFA_STATS
        !          1028:        if (ia && ip)
        !          1029:                ia->ia_ifa.ifa_data.ifad_inbytes += ntohs(ip->ip_len);
        !          1030: #endif
        !          1031:        ipstat.ips_delivered++;
        !          1032:     {
        !          1033:        int off = hlen, nh = ip->ip_p;
        !          1034:
        !          1035:        (*inetsw[ip_protox[nh]].pr_input)(m, off, nh);
        !          1036:        return;
        !          1037:     }
        !          1038: bad:
        !          1039:        m_freem(m);
        !          1040:        return;
        !          1041:
        !          1042: badcsum:
        !          1043:        ipstat.ips_badsum++;
        !          1044:        m_freem(m);
        !          1045: }
        !          1046:
        !          1047: /*
        !          1048:  * Take incoming datagram fragment and try to
        !          1049:  * reassemble it into whole datagram.  If a chain for
        !          1050:  * reassembly of this datagram already exists, then it
        !          1051:  * is given as fp; otherwise have to make a chain.
        !          1052:  */
        !          1053: struct mbuf *
        !          1054: ip_reass(struct ipqent *ipqe, struct ipq *fp, struct ipqhead *ipqhead)
        !          1055: {
        !          1056:        struct mbuf *m = ipqe->ipqe_m;
        !          1057:        struct ipqent *nq, *p, *q;
        !          1058:        struct ip *ip;
        !          1059:        struct mbuf *t;
        !          1060:        int hlen = ipqe->ipqe_ip->ip_hl << 2;
        !          1061:        int i, next, s;
        !          1062:
        !          1063:        IPQ_LOCK_CHECK();
        !          1064:
        !          1065:        /*
        !          1066:         * Presence of header sizes in mbufs
        !          1067:         * would confuse code below.
        !          1068:         */
        !          1069:        m->m_data += hlen;
        !          1070:        m->m_len -= hlen;
        !          1071:
        !          1072: #ifdef notyet
        !          1073:        /* make sure fragment limit is up-to-date */
        !          1074:        CHECK_NMBCLUSTER_PARAMS();
        !          1075:
        !          1076:        /* If we have too many fragments, drop the older half. */
        !          1077:        if (ip_nfrags >= ip_maxfrags)
        !          1078:                ip_reass_drophalf(void);
        !          1079: #endif
        !          1080:
        !          1081:        /*
        !          1082:         * We are about to add a fragment; increment frag count.
        !          1083:         */
        !          1084:        ip_nfrags++;
        !          1085:
        !          1086:        /*
        !          1087:         * If first fragment to arrive, create a reassembly queue.
        !          1088:         */
        !          1089:        if (fp == 0) {
        !          1090:                /*
        !          1091:                 * Enforce upper bound on number of fragmented packets
        !          1092:                 * for which we attempt reassembly;
        !          1093:                 * If maxfrag is 0, never accept fragments.
        !          1094:                 * If maxfrag is -1, accept all fragments without limitation.
        !          1095:                 */
        !          1096:                if (ip_maxfragpackets < 0)
        !          1097:                        ;
        !          1098:                else if (ip_nfragpackets >= ip_maxfragpackets)
        !          1099:                        goto dropfrag;
        !          1100:                ip_nfragpackets++;
        !          1101:                MALLOC(fp, struct ipq *, sizeof (struct ipq),
        !          1102:                    M_FTABLE, M_NOWAIT);
        !          1103:                if (fp == NULL)
        !          1104:                        goto dropfrag;
        !          1105:                LIST_INSERT_HEAD(ipqhead, fp, ipq_q);
        !          1106:                fp->ipq_nfrags = 1;
        !          1107:                fp->ipq_ttl = IPFRAGTTL;
        !          1108:                fp->ipq_p = ipqe->ipqe_ip->ip_p;
        !          1109:                fp->ipq_id = ipqe->ipqe_ip->ip_id;
        !          1110:                TAILQ_INIT(&fp->ipq_fragq);
        !          1111:                fp->ipq_src = ipqe->ipqe_ip->ip_src;
        !          1112:                fp->ipq_dst = ipqe->ipqe_ip->ip_dst;
        !          1113:                p = NULL;
        !          1114:                goto insert;
        !          1115:        } else {
        !          1116:                fp->ipq_nfrags++;
        !          1117:        }
        !          1118:
        !          1119:        /*
        !          1120:         * Find a segment which begins after this one does.
        !          1121:         */
        !          1122:        for (p = NULL, q = TAILQ_FIRST(&fp->ipq_fragq); q != NULL;
        !          1123:            p = q, q = TAILQ_NEXT(q, ipqe_q))
        !          1124:                if (ntohs(q->ipqe_ip->ip_off) > ntohs(ipqe->ipqe_ip->ip_off))
        !          1125:                        break;
        !          1126:
        !          1127:        /*
        !          1128:         * If there is a preceding segment, it may provide some of
        !          1129:         * our data already.  If so, drop the data from the incoming
        !          1130:         * segment.  If it provides all of our data, drop us.
        !          1131:         */
        !          1132:        if (p != NULL) {
        !          1133:                i = ntohs(p->ipqe_ip->ip_off) + ntohs(p->ipqe_ip->ip_len) -
        !          1134:                    ntohs(ipqe->ipqe_ip->ip_off);
        !          1135:                if (i > 0) {
        !          1136:                        if (i >= ntohs(ipqe->ipqe_ip->ip_len))
        !          1137:                                goto dropfrag;
        !          1138:                        m_adj(ipqe->ipqe_m, i);
        !          1139:                        ipqe->ipqe_ip->ip_off =
        !          1140:                            htons(ntohs(ipqe->ipqe_ip->ip_off) + i);
        !          1141:                        ipqe->ipqe_ip->ip_len =
        !          1142:                            htons(ntohs(ipqe->ipqe_ip->ip_len) - i);
        !          1143:                }
        !          1144:        }
        !          1145:
        !          1146:        /*
        !          1147:         * While we overlap succeeding segments trim them or,
        !          1148:         * if they are completely covered, dequeue them.
        !          1149:         */
        !          1150:        for (; q != NULL &&
        !          1151:            ntohs(ipqe->ipqe_ip->ip_off) + ntohs(ipqe->ipqe_ip->ip_len) >
        !          1152:            ntohs(q->ipqe_ip->ip_off); q = nq) {
        !          1153:                i = (ntohs(ipqe->ipqe_ip->ip_off) +
        !          1154:                    ntohs(ipqe->ipqe_ip->ip_len)) - ntohs(q->ipqe_ip->ip_off);
        !          1155:                if (i < ntohs(q->ipqe_ip->ip_len)) {
        !          1156:                        q->ipqe_ip->ip_len =
        !          1157:                            htons(ntohs(q->ipqe_ip->ip_len) - i);
        !          1158:                        q->ipqe_ip->ip_off =
        !          1159:                            htons(ntohs(q->ipqe_ip->ip_off) + i);
        !          1160:                        m_adj(q->ipqe_m, i);
        !          1161:                        break;
        !          1162:                }
        !          1163:                nq = TAILQ_NEXT(q, ipqe_q);
        !          1164:                m_freem(q->ipqe_m);
        !          1165:                TAILQ_REMOVE(&fp->ipq_fragq, q, ipqe_q);
        !          1166:                s = splvm();
        !          1167:                pool_put(&ipqent_pool, q);
        !          1168:                splx(s);
        !          1169:                fp->ipq_nfrags--;
        !          1170:                ip_nfrags--;
        !          1171:        }
        !          1172:
        !          1173: insert:
        !          1174:        /*
        !          1175:         * Stick new segment in its place;
        !          1176:         * check for complete reassembly.
        !          1177:         */
        !          1178:        if (p == NULL) {
        !          1179:                TAILQ_INSERT_HEAD(&fp->ipq_fragq, ipqe, ipqe_q);
        !          1180:        } else {
        !          1181:                TAILQ_INSERT_AFTER(&fp->ipq_fragq, p, ipqe, ipqe_q);
        !          1182:        }
        !          1183:        next = 0;
        !          1184:        for (p = NULL, q = TAILQ_FIRST(&fp->ipq_fragq); q != NULL;
        !          1185:            p = q, q = TAILQ_NEXT(q, ipqe_q)) {
        !          1186:                if (ntohs(q->ipqe_ip->ip_off) != next)
        !          1187:                        return (0);
        !          1188:                next += ntohs(q->ipqe_ip->ip_len);
        !          1189:        }
        !          1190:        if (p->ipqe_mff)
        !          1191:                return (0);
        !          1192:
        !          1193:        /*
        !          1194:         * Reassembly is complete.  Check for a bogus message size and
        !          1195:         * concatenate fragments.
        !          1196:         */
        !          1197:        q = TAILQ_FIRST(&fp->ipq_fragq);
        !          1198:        ip = q->ipqe_ip;
        !          1199:        if ((next + (ip->ip_hl << 2)) > IP_MAXPACKET) {
        !          1200:                ipstat.ips_toolong++;
        !          1201:                ip_freef(fp);
        !          1202:                return (0);
        !          1203:        }
        !          1204:        m = q->ipqe_m;
        !          1205:        t = m->m_next;
        !          1206:        m->m_next = 0;
        !          1207:        m_cat(m, t);
        !          1208:        nq = TAILQ_NEXT(q, ipqe_q);
        !          1209:        s = splvm();
        !          1210:        pool_put(&ipqent_pool, q);
        !          1211:        splx(s);
        !          1212:        for (q = nq; q != NULL; q = nq) {
        !          1213:                t = q->ipqe_m;
        !          1214:                nq = TAILQ_NEXT(q, ipqe_q);
        !          1215:                s = splvm();
        !          1216:                pool_put(&ipqent_pool, q);
        !          1217:                splx(s);
        !          1218:                m_cat(m, t);
        !          1219:        }
        !          1220:        ip_nfrags -= fp->ipq_nfrags;
        !          1221:
        !          1222:        /*
        !          1223:         * Create header for new ip packet by
        !          1224:         * modifying header of first packet;
        !          1225:         * dequeue and discard fragment reassembly header.
        !          1226:         * Make header visible.
        !          1227:         */
        !          1228:        ip->ip_len = htons(next);
        !          1229:        ip->ip_src = fp->ipq_src;
        !          1230:        ip->ip_dst = fp->ipq_dst;
        !          1231:        LIST_REMOVE(fp, ipq_q);
        !          1232:        FREE(fp, M_FTABLE);
        !          1233:        ip_nfragpackets--;
        !          1234:        m->m_len += (ip->ip_hl << 2);
        !          1235:        m->m_data -= (ip->ip_hl << 2);
        !          1236:        /* some debugging cruft by sklower, below, will go away soon */
        !          1237:        if (m->m_flags & M_PKTHDR) { /* XXX this should be done elsewhere */
        !          1238:                int plen = 0;
        !          1239:                for (t = m; t; t = t->m_next)
        !          1240:                        plen += t->m_len;
        !          1241:                m->m_pkthdr.len = plen;
        !          1242:                m->m_pkthdr.csum_flags = 0;
        !          1243:        }
        !          1244:        return (m);
        !          1245:
        !          1246: dropfrag:
        !          1247:        if (fp != 0)
        !          1248:                fp->ipq_nfrags--;
        !          1249:        ip_nfrags--;
        !          1250:        ipstat.ips_fragdropped++;
        !          1251:        m_freem(m);
        !          1252:        s = splvm();
        !          1253:        pool_put(&ipqent_pool, ipqe);
        !          1254:        splx(s);
        !          1255:        return (0);
        !          1256: }
        !          1257:
        !          1258: /*
        !          1259:  * Free a fragment reassembly header and all
        !          1260:  * associated datagrams.
        !          1261:  */
        !          1262: void
        !          1263: ip_freef(struct ipq *fp)
        !          1264: {
        !          1265:        struct ipqent *q, *p;
        !          1266:        u_int nfrags = 0;
        !          1267:        int s;
        !          1268:
        !          1269:        IPQ_LOCK_CHECK();
        !          1270:
        !          1271:        for (q = TAILQ_FIRST(&fp->ipq_fragq); q != NULL; q = p) {
        !          1272:                p = TAILQ_NEXT(q, ipqe_q);
        !          1273:                m_freem(q->ipqe_m);
        !          1274:                nfrags++;
        !          1275:                TAILQ_REMOVE(&fp->ipq_fragq, q, ipqe_q);
        !          1276:                s = splvm();
        !          1277:                pool_put(&ipqent_pool, q);
        !          1278:                splx(s);
        !          1279:        }
        !          1280:
        !          1281:        if (nfrags != fp->ipq_nfrags)
        !          1282:            printf("ip_freef: nfrags %d != %d\n", fp->ipq_nfrags, nfrags);
        !          1283:        ip_nfrags -= nfrags;
        !          1284:        LIST_REMOVE(fp, ipq_q);
        !          1285:        FREE(fp, M_FTABLE);
        !          1286:        ip_nfragpackets--;
        !          1287: }
        !          1288:
        !          1289: /*
        !          1290:  * IP reassembly TTL machinery for  multiplicative drop.
        !          1291:  */
        !          1292: static u_int   fragttl_histo[(IPFRAGTTL+1)];
        !          1293:
        !          1294:
        !          1295: /*
        !          1296:  * Decrement TTL of all reasembly queue entries by `ticks'.
        !          1297:  * Count number of distinct fragments (as opposed to partial, fragmented
        !          1298:  * datagrams) in the reassembly queue.  While we  traverse the entire
        !          1299:  * reassembly queue, compute and return the median TTL over all fragments.
        !          1300:  */
        !          1301: static u_int
        !          1302: ip_reass_ttl_decr(u_int ticks)
        !          1303: {
        !          1304:        u_int nfrags, median, dropfraction, keepfraction;
        !          1305:        struct ipq *fp, *nfp;
        !          1306:        int i;
        !          1307:
        !          1308:        nfrags = 0;
        !          1309:        memset(fragttl_histo, 0, sizeof fragttl_histo);
        !          1310:
        !          1311:        for (i = 0; i < IPREASS_NHASH; i++) {
        !          1312:                for (fp = LIST_FIRST(&ipq[i]); fp != NULL; fp = nfp) {
        !          1313:                        fp->ipq_ttl = ((fp->ipq_ttl  <= ticks) ?
        !          1314:                                       0 : fp->ipq_ttl - ticks);
        !          1315:                        nfp = LIST_NEXT(fp, ipq_q);
        !          1316:                        if (fp->ipq_ttl == 0) {
        !          1317:                                ipstat.ips_fragtimeout++;
        !          1318:                                ip_freef(fp);
        !          1319:                        } else {
        !          1320:                                nfrags += fp->ipq_nfrags;
        !          1321:                                fragttl_histo[fp->ipq_ttl] += fp->ipq_nfrags;
        !          1322:                        }
        !          1323:                }
        !          1324:        }
        !          1325:
        !          1326:        KASSERT(ip_nfrags == nfrags);
        !          1327:
        !          1328:        /* Find median (or other drop fraction) in histogram. */
        !          1329:        dropfraction = (ip_nfrags / 2);
        !          1330:        keepfraction = ip_nfrags - dropfraction;
        !          1331:        for (i = IPFRAGTTL, median = 0; i >= 0; i--) {
        !          1332:                median +=  fragttl_histo[i];
        !          1333:                if (median >= keepfraction)
        !          1334:                        break;
        !          1335:        }
        !          1336:
        !          1337:        /* Return TTL of median (or other fraction). */
        !          1338:        return (u_int)i;
        !          1339: }
        !          1340:
        !          1341: void
        !          1342: ip_reass_drophalf(void)
        !          1343: {
        !          1344:
        !          1345:        u_int median_ticks;
        !          1346:        /*
        !          1347:         * Compute median TTL of all fragments, and count frags
        !          1348:         * with that TTL or lower (roughly half of all fragments).
        !          1349:         */
        !          1350:        median_ticks = ip_reass_ttl_decr(0);
        !          1351:
        !          1352:        /* Drop half. */
        !          1353:        median_ticks = ip_reass_ttl_decr(median_ticks);
        !          1354:
        !          1355: }
        !          1356:
        !          1357: /*
        !          1358:  * IP timer processing;
        !          1359:  * if a timer expires on a reassembly
        !          1360:  * queue, discard it.
        !          1361:  */
        !          1362: void
        !          1363: ip_slowtimo(void)
        !          1364: {
        !          1365:        static u_int dropscanidx = 0;
        !          1366:        u_int i;
        !          1367:        u_int median_ttl;
        !          1368:        int s = splsoftnet();
        !          1369:
        !          1370:        IPQ_LOCK();
        !          1371:
        !          1372:        /* Age TTL of all fragments by 1 tick .*/
        !          1373:        median_ttl = ip_reass_ttl_decr(1);
        !          1374:
        !          1375:        /* make sure fragment limit is up-to-date */
        !          1376:        CHECK_NMBCLUSTER_PARAMS();
        !          1377:
        !          1378:        /* If we have too many fragments, drop the older half. */
        !          1379:        if (ip_nfrags > ip_maxfrags)
        !          1380:                ip_reass_ttl_decr(median_ttl);
        !          1381:
        !          1382:        /*
        !          1383:         * If we are over the maximum number of fragmented packets
        !          1384:         * (due to the limit being lowered), drain off
        !          1385:         * enough to get down to the new limit. Start draining
        !          1386:         * from the reassembly hashqueue most recently drained.
        !          1387:         */
        !          1388:        if (ip_maxfragpackets < 0)
        !          1389:                ;
        !          1390:        else {
        !          1391:                int wrapped = 0;
        !          1392:
        !          1393:                i = dropscanidx;
        !          1394:                while (ip_nfragpackets > ip_maxfragpackets && wrapped == 0) {
        !          1395:                        while (LIST_FIRST(&ipq[i]) != NULL)
        !          1396:                                ip_freef(LIST_FIRST(&ipq[i]));
        !          1397:                        if (++i >= IPREASS_NHASH) {
        !          1398:                                i = 0;
        !          1399:                        }
        !          1400:                        /*
        !          1401:                         * Dont scan forever even if fragment counters are
        !          1402:                         * wrong: stop after scanning entire reassembly queue.
        !          1403:                         */
        !          1404:                        if (i == dropscanidx)
        !          1405:                            wrapped = 1;
        !          1406:                }
        !          1407:                dropscanidx = i;
        !          1408:        }
        !          1409:        IPQ_UNLOCK();
        !          1410:        splx(s);
        !          1411: }
        !          1412:
        !          1413: /*
        !          1414:  * Drain off all datagram fragments.
        !          1415:  */
        !          1416: void
        !          1417: ip_drain(void)
        !          1418: {
        !          1419:
        !          1420:        /*
        !          1421:         * We may be called from a device's interrupt context.  If
        !          1422:         * the ipq is already busy, just bail out now.
        !          1423:         */
        !          1424:        if (ipq_lock_try() == 0)
        !          1425:                return;
        !          1426:
        !          1427:        /*
        !          1428:         * Drop half the total fragments now. If more mbufs are needed,
        !          1429:         *  we will be called again soon.
        !          1430:         */
        !          1431:        ip_reass_drophalf();
        !          1432:
        !          1433:        IPQ_UNLOCK();
        !          1434: }
        !          1435:
        !          1436: /*
        !          1437:  * Do option processing on a datagram,
        !          1438:  * possibly discarding it if bad options are encountered,
        !          1439:  * or forwarding it if source-routed.
        !          1440:  * Returns 1 if packet has been forwarded/freed,
        !          1441:  * 0 if the packet should be processed further.
        !          1442:  */
        !          1443: int
        !          1444: ip_dooptions(struct mbuf *m)
        !          1445: {
        !          1446:        struct ip *ip = mtod(m, struct ip *);
        !          1447:        u_char *cp, *cp0;
        !          1448:        struct ip_timestamp *ipt;
        !          1449:        struct in_ifaddr *ia;
        !          1450:        int opt, optlen, cnt, off, code, type = ICMP_PARAMPROB, forward = 0;
        !          1451:        struct in_addr dst;
        !          1452:        n_time ntime;
        !          1453:
        !          1454:        dst = ip->ip_dst;
        !          1455:        cp = (u_char *)(ip + 1);
        !          1456:        cnt = (ip->ip_hl << 2) - sizeof (struct ip);
        !          1457:        for (; cnt > 0; cnt -= optlen, cp += optlen) {
        !          1458:                opt = cp[IPOPT_OPTVAL];
        !          1459:                if (opt == IPOPT_EOL)
        !          1460:                        break;
        !          1461:                if (opt == IPOPT_NOP)
        !          1462:                        optlen = 1;
        !          1463:                else {
        !          1464:                        if (cnt < IPOPT_OLEN + sizeof(*cp)) {
        !          1465:                                code = &cp[IPOPT_OLEN] - (u_char *)ip;
        !          1466:                                goto bad;
        !          1467:                        }
        !          1468:                        optlen = cp[IPOPT_OLEN];
        !          1469:                        if (optlen < IPOPT_OLEN + sizeof(*cp) || optlen > cnt) {
        !          1470:                                code = &cp[IPOPT_OLEN] - (u_char *)ip;
        !          1471:                                goto bad;
        !          1472:                        }
        !          1473:                }
        !          1474:                switch (opt) {
        !          1475:
        !          1476:                default:
        !          1477:                        break;
        !          1478:
        !          1479:                /*
        !          1480:                 * Source routing with record.
        !          1481:                 * Find interface with current destination address.
        !          1482:                 * If none on this machine then drop if strictly routed,
        !          1483:                 * or do nothing if loosely routed.
        !          1484:                 * Record interface address and bring up next address
        !          1485:                 * component.  If strictly routed make sure next
        !          1486:                 * address is on directly accessible net.
        !          1487:                 */
        !          1488:                case IPOPT_LSRR:
        !          1489:                case IPOPT_SSRR:
        !          1490:                        if (ip_allowsrcrt == 0) {
        !          1491:                                type = ICMP_UNREACH;
        !          1492:                                code = ICMP_UNREACH_NET_PROHIB;
        !          1493:                                goto bad;
        !          1494:                        }
        !          1495:                        if (optlen < IPOPT_OFFSET + sizeof(*cp)) {
        !          1496:                                code = &cp[IPOPT_OLEN] - (u_char *)ip;
        !          1497:                                goto bad;
        !          1498:                        }
        !          1499:                        if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) {
        !          1500:                                code = &cp[IPOPT_OFFSET] - (u_char *)ip;
        !          1501:                                goto bad;
        !          1502:                        }
        !          1503:                        ipaddr.sin_addr = ip->ip_dst;
        !          1504:                        ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr)));
        !          1505:                        if (ia == 0) {
        !          1506:                                if (opt == IPOPT_SSRR) {
        !          1507:                                        type = ICMP_UNREACH;
        !          1508:                                        code = ICMP_UNREACH_SRCFAIL;
        !          1509:                                        goto bad;
        !          1510:                                }
        !          1511:                                /*
        !          1512:                                 * Loose routing, and not at next destination
        !          1513:                                 * yet; nothing to do except forward.
        !          1514:                                 */
        !          1515:                                break;
        !          1516:                        }
        !          1517:                        off--;                  /* 0 origin */
        !          1518:                        if ((off + sizeof(struct in_addr)) > optlen) {
        !          1519:                                /*
        !          1520:                                 * End of source route.  Should be for us.
        !          1521:                                 */
        !          1522:                                save_rte(cp, ip->ip_src);
        !          1523:                                break;
        !          1524:                        }
        !          1525:                        /*
        !          1526:                         * locate outgoing interface
        !          1527:                         */
        !          1528:                        bcopy((void *)(cp + off), (void *)&ipaddr.sin_addr,
        !          1529:                            sizeof(ipaddr.sin_addr));
        !          1530:                        if (opt == IPOPT_SSRR)
        !          1531:                                ia = ifatoia(ifa_ifwithladdr(sintosa(&ipaddr)));
        !          1532:                        else
        !          1533:                                ia = ip_rtaddr(ipaddr.sin_addr);
        !          1534:                        if (ia == 0) {
        !          1535:                                type = ICMP_UNREACH;
        !          1536:                                code = ICMP_UNREACH_SRCFAIL;
        !          1537:                                goto bad;
        !          1538:                        }
        !          1539:                        ip->ip_dst = ipaddr.sin_addr;
        !          1540:                        bcopy((void *)&ia->ia_addr.sin_addr,
        !          1541:                            (void *)(cp + off), sizeof(struct in_addr));
        !          1542:                        cp[IPOPT_OFFSET] += sizeof(struct in_addr);
        !          1543:                        /*
        !          1544:                         * Let ip_intr's mcast routing check handle mcast pkts
        !          1545:                         */
        !          1546:                        forward = !IN_MULTICAST(ip->ip_dst.s_addr);
        !          1547:                        break;
        !          1548:
        !          1549:                case IPOPT_RR:
        !          1550:                        if (optlen < IPOPT_OFFSET + sizeof(*cp)) {
        !          1551:                                code = &cp[IPOPT_OLEN] - (u_char *)ip;
        !          1552:                                goto bad;
        !          1553:                        }
        !          1554:                        if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) {
        !          1555:                                code = &cp[IPOPT_OFFSET] - (u_char *)ip;
        !          1556:                                goto bad;
        !          1557:                        }
        !          1558:                        /*
        !          1559:                         * If no space remains, ignore.
        !          1560:                         */
        !          1561:                        off--;                  /* 0 origin */
        !          1562:                        if ((off + sizeof(struct in_addr)) > optlen)
        !          1563:                                break;
        !          1564:                        bcopy((void *)(&ip->ip_dst), (void *)&ipaddr.sin_addr,
        !          1565:                            sizeof(ipaddr.sin_addr));
        !          1566:                        /*
        !          1567:                         * locate outgoing interface; if we're the destination,
        !          1568:                         * use the incoming interface (should be same).
        !          1569:                         */
        !          1570:                        if ((ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr))))
        !          1571:                            == NULL &&
        !          1572:                            (ia = ip_rtaddr(ipaddr.sin_addr)) == NULL) {
        !          1573:                                type = ICMP_UNREACH;
        !          1574:                                code = ICMP_UNREACH_HOST;
        !          1575:                                goto bad;
        !          1576:                        }
        !          1577:                        bcopy((void *)&ia->ia_addr.sin_addr,
        !          1578:                            (void *)(cp + off), sizeof(struct in_addr));
        !          1579:                        cp[IPOPT_OFFSET] += sizeof(struct in_addr);
        !          1580:                        break;
        !          1581:
        !          1582:                case IPOPT_TS:
        !          1583:                        code = cp - (u_char *)ip;
        !          1584:                        ipt = (struct ip_timestamp *)cp;
        !          1585:                        if (ipt->ipt_len < 4 || ipt->ipt_len > 40) {
        !          1586:                                code = (u_char *)&ipt->ipt_len - (u_char *)ip;
        !          1587:                                goto bad;
        !          1588:                        }
        !          1589:                        if (ipt->ipt_ptr < 5) {
        !          1590:                                code = (u_char *)&ipt->ipt_ptr - (u_char *)ip;
        !          1591:                                goto bad;
        !          1592:                        }
        !          1593:                        if (ipt->ipt_ptr > ipt->ipt_len - sizeof (int32_t)) {
        !          1594:                                if (++ipt->ipt_oflw == 0) {
        !          1595:                                        code = (u_char *)&ipt->ipt_ptr -
        !          1596:                                            (u_char *)ip;
        !          1597:                                        goto bad;
        !          1598:                                }
        !          1599:                                break;
        !          1600:                        }
        !          1601:                        cp0 = (cp + ipt->ipt_ptr - 1);
        !          1602:                        switch (ipt->ipt_flg) {
        !          1603:
        !          1604:                        case IPOPT_TS_TSONLY:
        !          1605:                                break;
        !          1606:
        !          1607:                        case IPOPT_TS_TSANDADDR:
        !          1608:                                if (ipt->ipt_ptr - 1 + sizeof(n_time) +
        !          1609:                                    sizeof(struct in_addr) > ipt->ipt_len) {
        !          1610:                                        code = (u_char *)&ipt->ipt_ptr -
        !          1611:                                            (u_char *)ip;
        !          1612:                                        goto bad;
        !          1613:                                }
        !          1614:                                ipaddr.sin_addr = dst;
        !          1615:                                ia = ifatoia(ifaof_ifpforaddr(sintosa(&ipaddr),
        !          1616:                                    m->m_pkthdr.rcvif));
        !          1617:                                if (ia == 0)
        !          1618:                                        continue;
        !          1619:                                bcopy(&ia->ia_addr.sin_addr,
        !          1620:                                    cp0, sizeof(struct in_addr));
        !          1621:                                ipt->ipt_ptr += sizeof(struct in_addr);
        !          1622:                                break;
        !          1623:
        !          1624:                        case IPOPT_TS_PRESPEC:
        !          1625:                                if (ipt->ipt_ptr - 1 + sizeof(n_time) +
        !          1626:                                    sizeof(struct in_addr) > ipt->ipt_len) {
        !          1627:                                        code = (u_char *)&ipt->ipt_ptr -
        !          1628:                                            (u_char *)ip;
        !          1629:                                        goto bad;
        !          1630:                                }
        !          1631:                                bcopy(cp0, &ipaddr.sin_addr,
        !          1632:                                    sizeof(struct in_addr));
        !          1633:                                if (ifatoia(ifa_ifwithaddr(sintosa(&ipaddr)))
        !          1634:                                    == NULL)
        !          1635:                                        continue;
        !          1636:                                ipt->ipt_ptr += sizeof(struct in_addr);
        !          1637:                                break;
        !          1638:
        !          1639:                        default:
        !          1640:                                /* XXX can't take &ipt->ipt_flg */
        !          1641:                                code = (u_char *)&ipt->ipt_ptr -
        !          1642:                                    (u_char *)ip + 1;
        !          1643:                                goto bad;
        !          1644:                        }
        !          1645:                        ntime = iptime();
        !          1646:                        cp0 = (u_char *) &ntime; /* XXX grumble, GCC... */
        !          1647:                        memmove((char *)cp + ipt->ipt_ptr - 1, cp0,
        !          1648:                            sizeof(n_time));
        !          1649:                        ipt->ipt_ptr += sizeof(n_time);
        !          1650:                }
        !          1651:        }
        !          1652:        if (forward) {
        !          1653:                if (ip_forwsrcrt == 0) {
        !          1654:                        type = ICMP_UNREACH;
        !          1655:                        code = ICMP_UNREACH_SRCFAIL;
        !          1656:                        goto bad;
        !          1657:                }
        !          1658:                ip_forward(m, 1);
        !          1659:                return (1);
        !          1660:        }
        !          1661:        return (0);
        !          1662: bad:
        !          1663:        icmp_error(m, type, code, 0, 0);
        !          1664:        ipstat.ips_badoptions++;
        !          1665:        return (1);
        !          1666: }
        !          1667:
        !          1668: /*
        !          1669:  * Given address of next destination (final or next hop),
        !          1670:  * return internet address info of interface to be used to get there.
        !          1671:  */
        !          1672: struct in_ifaddr *
        !          1673: ip_rtaddr(struct in_addr dst)
        !          1674: {
        !          1675:        struct rtentry *rt;
        !          1676:        union {
        !          1677:                struct sockaddr         dst;
        !          1678:                struct sockaddr_in      dst4;
        !          1679:        } u;
        !          1680:
        !          1681:        sockaddr_in_init(&u.dst4, &dst, 0);
        !          1682:
        !          1683:        if ((rt = rtcache_lookup(&ipforward_rt, &u.dst)) == NULL)
        !          1684:                return NULL;
        !          1685:
        !          1686:        return ifatoia(rt->rt_ifa);
        !          1687: }
        !          1688:
        !          1689: /*
        !          1690:  * Save incoming source route for use in replies,
        !          1691:  * to be picked up later by ip_srcroute if the receiver is interested.
        !          1692:  */
        !          1693: void
        !          1694: save_rte(u_char *option, struct in_addr dst)
        !          1695: {
        !          1696:        unsigned olen;
        !          1697:
        !          1698:        olen = option[IPOPT_OLEN];
        !          1699: #ifdef DIAGNOSTIC
        !          1700:        if (ipprintfs)
        !          1701:                printf("save_rte: olen %d\n", olen);
        !          1702: #endif /* 0 */
        !          1703:        if (olen > sizeof(ip_srcrt) - (1 + sizeof(dst)))
        !          1704:                return;
        !          1705:        bcopy((void *)option, (void *)ip_srcrt.srcopt, olen);
        !          1706:        ip_nhops = (olen - IPOPT_OFFSET - 1) / sizeof(struct in_addr);
        !          1707:        ip_srcrt.dst = dst;
        !          1708: }
        !          1709:
        !          1710: /*
        !          1711:  * Retrieve incoming source route for use in replies,
        !          1712:  * in the same form used by setsockopt.
        !          1713:  * The first hop is placed before the options, will be removed later.
        !          1714:  */
        !          1715: struct mbuf *
        !          1716: ip_srcroute(void)
        !          1717: {
        !          1718:        struct in_addr *p, *q;
        !          1719:        struct mbuf *m;
        !          1720:
        !          1721:        if (ip_nhops == 0)
        !          1722:                return NULL;
        !          1723:        m = m_get(M_DONTWAIT, MT_SOOPTS);
        !          1724:        if (m == 0)
        !          1725:                return NULL;
        !          1726:
        !          1727:        MCLAIM(m, &inetdomain.dom_mowner);
        !          1728: #define OPTSIZ (sizeof(ip_srcrt.nop) + sizeof(ip_srcrt.srcopt))
        !          1729:
        !          1730:        /* length is (nhops+1)*sizeof(addr) + sizeof(nop + srcrt header) */
        !          1731:        m->m_len = ip_nhops * sizeof(struct in_addr) + sizeof(struct in_addr) +
        !          1732:            OPTSIZ;
        !          1733: #ifdef DIAGNOSTIC
        !          1734:        if (ipprintfs)
        !          1735:                printf("ip_srcroute: nhops %d mlen %d", ip_nhops, m->m_len);
        !          1736: #endif
        !          1737:
        !          1738:        /*
        !          1739:         * First save first hop for return route
        !          1740:         */
        !          1741:        p = &ip_srcrt.route[ip_nhops - 1];
        !          1742:        *(mtod(m, struct in_addr *)) = *p--;
        !          1743: #ifdef DIAGNOSTIC
        !          1744:        if (ipprintfs)
        !          1745:                printf(" hops %x", ntohl(mtod(m, struct in_addr *)->s_addr));
        !          1746: #endif
        !          1747:
        !          1748:        /*
        !          1749:         * Copy option fields and padding (nop) to mbuf.
        !          1750:         */
        !          1751:        ip_srcrt.nop = IPOPT_NOP;
        !          1752:        ip_srcrt.srcopt[IPOPT_OFFSET] = IPOPT_MINOFF;
        !          1753:        memmove(mtod(m, char *) + sizeof(struct in_addr), &ip_srcrt.nop,
        !          1754:            OPTSIZ);
        !          1755:        q = (struct in_addr *)(mtod(m, char *) +
        !          1756:            sizeof(struct in_addr) + OPTSIZ);
        !          1757: #undef OPTSIZ
        !          1758:        /*
        !          1759:         * Record return path as an IP source route,
        !          1760:         * reversing the path (pointers are now aligned).
        !          1761:         */
        !          1762:        while (p >= ip_srcrt.route) {
        !          1763: #ifdef DIAGNOSTIC
        !          1764:                if (ipprintfs)
        !          1765:                        printf(" %x", ntohl(q->s_addr));
        !          1766: #endif
        !          1767:                *q++ = *p--;
        !          1768:        }
        !          1769:        /*
        !          1770:         * Last hop goes to final destination.
        !          1771:         */
        !          1772:        *q = ip_srcrt.dst;
        !          1773: #ifdef DIAGNOSTIC
        !          1774:        if (ipprintfs)
        !          1775:                printf(" %x\n", ntohl(q->s_addr));
        !          1776: #endif
        !          1777:        return (m);
        !          1778: }
        !          1779:
        !          1780: /*
        !          1781:  * Strip out IP options, at higher
        !          1782:  * level protocol in the kernel.
        !          1783:  * Second argument is buffer to which options
        !          1784:  * will be moved, and return value is their length.
        !          1785:  * XXX should be deleted; last arg currently ignored.
        !          1786:  */
        !          1787: void
        !          1788: ip_stripoptions(struct mbuf *m, struct mbuf *mopt)
        !          1789: {
        !          1790:        int i;
        !          1791:        struct ip *ip = mtod(m, struct ip *);
        !          1792:        void *opts;
        !          1793:        int olen;
        !          1794:
        !          1795:        olen = (ip->ip_hl << 2) - sizeof (struct ip);
        !          1796:        opts = (void *)(ip + 1);
        !          1797:        i = m->m_len - (sizeof (struct ip) + olen);
        !          1798:        memmove(opts, (char *)opts + olen, (unsigned)i);
        !          1799:        m->m_len -= olen;
        !          1800:        if (m->m_flags & M_PKTHDR)
        !          1801:                m->m_pkthdr.len -= olen;
        !          1802:        ip->ip_len = htons(ntohs(ip->ip_len) - olen);
        !          1803:        ip->ip_hl = sizeof (struct ip) >> 2;
        !          1804: }
        !          1805:
        !          1806: const int inetctlerrmap[PRC_NCMDS] = {
        !          1807:        0,              0,              0,              0,
        !          1808:        0,              EMSGSIZE,       EHOSTDOWN,      EHOSTUNREACH,
        !          1809:        EHOSTUNREACH,   EHOSTUNREACH,   ECONNREFUSED,   ECONNREFUSED,
        !          1810:        EMSGSIZE,       EHOSTUNREACH,   0,              0,
        !          1811:        0,              0,              0,              0,
        !          1812:        ENOPROTOOPT
        !          1813: };
        !          1814:
        !          1815: /*
        !          1816:  * Forward a packet.  If some error occurs return the sender
        !          1817:  * an icmp packet.  Note we can't always generate a meaningful
        !          1818:  * icmp message because icmp doesn't have a large enough repertoire
        !          1819:  * of codes and types.
        !          1820:  *
        !          1821:  * If not forwarding, just drop the packet.  This could be confusing
        !          1822:  * if ipforwarding was zero but some routing protocol was advancing
        !          1823:  * us as a gateway to somewhere.  However, we must let the routing
        !          1824:  * protocol deal with that.
        !          1825:  *
        !          1826:  * The srcrt parameter indicates whether the packet is being forwarded
        !          1827:  * via a source route.
        !          1828:  */
        !          1829: void
        !          1830: ip_forward(struct mbuf *m, int srcrt)
        !          1831: {
        !          1832:        struct ip *ip = mtod(m, struct ip *);
        !          1833:        struct rtentry *rt;
        !          1834:        int error, type = 0, code = 0, destmtu = 0;
        !          1835:        struct mbuf *mcopy;
        !          1836:        n_long dest;
        !          1837:        union {
        !          1838:                struct sockaddr         dst;
        !          1839:                struct sockaddr_in      dst4;
        !          1840:        } u;
        !          1841:
        !          1842:        /*
        !          1843:         * We are now in the output path.
        !          1844:         */
        !          1845:        MCLAIM(m, &ip_tx_mowner);
        !          1846:
        !          1847:        /*
        !          1848:         * Clear any in-bound checksum flags for this packet.
        !          1849:         */
        !          1850:        m->m_pkthdr.csum_flags = 0;
        !          1851:
        !          1852:        dest = 0;
        !          1853: #ifdef DIAGNOSTIC
        !          1854:        if (ipprintfs) {
        !          1855:                printf("forward: src %s ", inet_ntoa(ip->ip_src));
        !          1856:                printf("dst %s ttl %x\n", inet_ntoa(ip->ip_dst), ip->ip_ttl);
        !          1857:        }
        !          1858: #endif
        !          1859:        if (m->m_flags & (M_BCAST|M_MCAST) || in_canforward(ip->ip_dst) == 0) {
        !          1860:                ipstat.ips_cantforward++;
        !          1861:                m_freem(m);
        !          1862:                return;
        !          1863:        }
        !          1864:        if (ip->ip_ttl <= IPTTLDEC) {
        !          1865:                icmp_error(m, ICMP_TIMXCEED, ICMP_TIMXCEED_INTRANS, dest, 0);
        !          1866:                return;
        !          1867:        }
        !          1868:
        !          1869:        sockaddr_in_init(&u.dst4, &ip->ip_dst, 0);
        !          1870:        if ((rt = rtcache_lookup(&ipforward_rt, &u.dst)) == NULL) {
        !          1871:                icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_NET, dest, 0);
        !          1872:                return;
        !          1873:        }
        !          1874:
        !          1875:        /*
        !          1876:         * Save at most 68 bytes of the packet in case
        !          1877:         * we need to generate an ICMP message to the src.
        !          1878:         * Pullup to avoid sharing mbuf cluster between m and mcopy.
        !          1879:         */
        !          1880:        mcopy = m_copym(m, 0, imin(ntohs(ip->ip_len), 68), M_DONTWAIT);
        !          1881:        if (mcopy)
        !          1882:                mcopy = m_pullup(mcopy, ip->ip_hl << 2);
        !          1883:
        !          1884:        ip->ip_ttl -= IPTTLDEC;
        !          1885:
        !          1886:        /*
        !          1887:         * If forwarding packet using same interface that it came in on,
        !          1888:         * perhaps should send a redirect to sender to shortcut a hop.
        !          1889:         * Only send redirect if source is sending directly to us,
        !          1890:         * and if packet was not source routed (or has any options).
        !          1891:         * Also, don't send redirect if forwarding using a default route
        !          1892:         * or a route modified by a redirect.
        !          1893:         */
        !          1894:        if (rt->rt_ifp == m->m_pkthdr.rcvif &&
        !          1895:            (rt->rt_flags & (RTF_DYNAMIC|RTF_MODIFIED)) == 0 &&
        !          1896:            !in_nullhost(satocsin(rt_getkey(rt))->sin_addr) &&
        !          1897:            ipsendredirects && !srcrt) {
        !          1898:                if (rt->rt_ifa &&
        !          1899:                    (ip->ip_src.s_addr & ifatoia(rt->rt_ifa)->ia_subnetmask) ==
        !          1900:                    ifatoia(rt->rt_ifa)->ia_subnet) {
        !          1901:                        if (rt->rt_flags & RTF_GATEWAY)
        !          1902:                                dest = satosin(rt->rt_gateway)->sin_addr.s_addr;
        !          1903:                        else
        !          1904:                                dest = ip->ip_dst.s_addr;
        !          1905:                        /*
        !          1906:                         * Router requirements says to only send host
        !          1907:                         * redirects.
        !          1908:                         */
        !          1909:                        type = ICMP_REDIRECT;
        !          1910:                        code = ICMP_REDIRECT_HOST;
        !          1911: #ifdef DIAGNOSTIC
        !          1912:                        if (ipprintfs)
        !          1913:                                printf("redirect (%d) to %x\n", code,
        !          1914:                                    (u_int32_t)dest);
        !          1915: #endif
        !          1916:                }
        !          1917:        }
        !          1918:
        !          1919:        error = ip_output(m, NULL, &ipforward_rt,
        !          1920:            (IP_FORWARDING | (ip_directedbcast ? IP_ALLOWBROADCAST : 0)),
        !          1921:            (struct ip_moptions *)NULL, (struct socket *)NULL);
        !          1922:
        !          1923:        if (error)
        !          1924:                ipstat.ips_cantforward++;
        !          1925:        else {
        !          1926:                ipstat.ips_forward++;
        !          1927:                if (type)
        !          1928:                        ipstat.ips_redirectsent++;
        !          1929:                else {
        !          1930:                        if (mcopy) {
        !          1931: #ifdef GATEWAY
        !          1932:                                if (mcopy->m_flags & M_CANFASTFWD)
        !          1933:                                        ipflow_create(&ipforward_rt, mcopy);
        !          1934: #endif
        !          1935:                                m_freem(mcopy);
        !          1936:                        }
        !          1937:                        return;
        !          1938:                }
        !          1939:        }
        !          1940:        if (mcopy == NULL)
        !          1941:                return;
        !          1942:
        !          1943:        switch (error) {
        !          1944:
        !          1945:        case 0:                         /* forwarded, but need redirect */
        !          1946:                /* type, code set above */
        !          1947:                break;
        !          1948:
        !          1949:        case ENETUNREACH:               /* shouldn't happen, checked above */
        !          1950:        case EHOSTUNREACH:
        !          1951:        case ENETDOWN:
        !          1952:        case EHOSTDOWN:
        !          1953:        default:
        !          1954:                type = ICMP_UNREACH;
        !          1955:                code = ICMP_UNREACH_HOST;
        !          1956:                break;
        !          1957:
        !          1958:        case EMSGSIZE:
        !          1959:                type = ICMP_UNREACH;
        !          1960:                code = ICMP_UNREACH_NEEDFRAG;
        !          1961: #if !defined(IPSEC) && !defined(FAST_IPSEC)
        !          1962:                if (ipforward_rt.ro_rt != NULL)
        !          1963:                        destmtu = ipforward_rt.ro_rt->rt_ifp->if_mtu;
        !          1964: #else
        !          1965:                /*
        !          1966:                 * If the packet is routed over IPsec tunnel, tell the
        !          1967:                 * originator the tunnel MTU.
        !          1968:                 *      tunnel MTU = if MTU - sizeof(IP) - ESP/AH hdrsiz
        !          1969:                 * XXX quickhack!!!
        !          1970:                 */
        !          1971:                if (ipforward_rt.ro_rt != NULL) {
        !          1972:                        struct secpolicy *sp;
        !          1973:                        int ipsecerror;
        !          1974:                        size_t ipsechdr;
        !          1975:                        struct route *ro;
        !          1976:
        !          1977:                        sp = ipsec4_getpolicybyaddr(mcopy,
        !          1978:                            IPSEC_DIR_OUTBOUND, IP_FORWARDING,
        !          1979:                            &ipsecerror);
        !          1980:
        !          1981:                        if (sp == NULL)
        !          1982:                                destmtu = ipforward_rt.ro_rt->rt_ifp->if_mtu;
        !          1983:                        else {
        !          1984:                                /* count IPsec header size */
        !          1985:                                ipsechdr = ipsec4_hdrsiz(mcopy,
        !          1986:                                    IPSEC_DIR_OUTBOUND, NULL);
        !          1987:
        !          1988:                                /*
        !          1989:                                 * find the correct route for outer IPv4
        !          1990:                                 * header, compute tunnel MTU.
        !          1991:                                 */
        !          1992:
        !          1993:                                if (sp->req != NULL
        !          1994:                                 && sp->req->sav != NULL
        !          1995:                                 && sp->req->sav->sah != NULL) {
        !          1996:                                        ro = &sp->req->sav->sah->sa_route;
        !          1997:                                        if (ro->ro_rt && ro->ro_rt->rt_ifp) {
        !          1998:                                                destmtu =
        !          1999:                                                    ro->ro_rt->rt_rmx.rmx_mtu ?
        !          2000:                                                    ro->ro_rt->rt_rmx.rmx_mtu :
        !          2001:                                                    ro->ro_rt->rt_ifp->if_mtu;
        !          2002:                                                destmtu -= ipsechdr;
        !          2003:                                        }
        !          2004:                                }
        !          2005:
        !          2006: #ifdef IPSEC
        !          2007:                                key_freesp(sp);
        !          2008: #else
        !          2009:                                KEY_FREESP(&sp);
        !          2010: #endif
        !          2011:                        }
        !          2012:                }
        !          2013: #endif /*IPSEC*/
        !          2014:                ipstat.ips_cantfrag++;
        !          2015:                break;
        !          2016:
        !          2017:        case ENOBUFS:
        !          2018: #if 1
        !          2019:                /*
        !          2020:                 * a router should not generate ICMP_SOURCEQUENCH as
        !          2021:                 * required in RFC1812 Requirements for IP Version 4 Routers.
        !          2022:                 * source quench could be a big problem under DoS attacks,
        !          2023:                 * or if the underlying interface is rate-limited.
        !          2024:                 */
        !          2025:                if (mcopy)
        !          2026:                        m_freem(mcopy);
        !          2027:                return;
        !          2028: #else
        !          2029:                type = ICMP_SOURCEQUENCH;
        !          2030:                code = 0;
        !          2031:                break;
        !          2032: #endif
        !          2033:        }
        !          2034:        icmp_error(mcopy, type, code, dest, destmtu);
        !          2035: }
        !          2036:
        !          2037: void
        !          2038: ip_savecontrol(struct inpcb *inp, struct mbuf **mp, struct ip *ip,
        !          2039:     struct mbuf *m)
        !          2040: {
        !          2041:
        !          2042:        if (inp->inp_socket->so_options & SO_TIMESTAMP) {
        !          2043:                struct timeval tv;
        !          2044:
        !          2045:                microtime(&tv);
        !          2046:                *mp = sbcreatecontrol((void *) &tv, sizeof(tv),
        !          2047:                    SCM_TIMESTAMP, SOL_SOCKET);
        !          2048:                if (*mp)
        !          2049:                        mp = &(*mp)->m_next;
        !          2050:        }
        !          2051:        if (inp->inp_flags & INP_RECVDSTADDR) {
        !          2052:                *mp = sbcreatecontrol((void *) &ip->ip_dst,
        !          2053:                    sizeof(struct in_addr), IP_RECVDSTADDR, IPPROTO_IP);
        !          2054:                if (*mp)
        !          2055:                        mp = &(*mp)->m_next;
        !          2056:        }
        !          2057: #ifdef notyet
        !          2058:        /*
        !          2059:         * XXX
        !          2060:         * Moving these out of udp_input() made them even more broken
        !          2061:         * than they already were.
        !          2062:         *      - fenner@parc.xerox.com
        !          2063:         */
        !          2064:        /* options were tossed already */
        !          2065:        if (inp->inp_flags & INP_RECVOPTS) {
        !          2066:                *mp = sbcreatecontrol((void *) opts_deleted_above,
        !          2067:                    sizeof(struct in_addr), IP_RECVOPTS, IPPROTO_IP);
        !          2068:                if (*mp)
        !          2069:                        mp = &(*mp)->m_next;
        !          2070:        }
        !          2071:        /* ip_srcroute doesn't do what we want here, need to fix */
        !          2072:        if (inp->inp_flags & INP_RECVRETOPTS) {
        !          2073:                *mp = sbcreatecontrol((void *) ip_srcroute(),
        !          2074:                    sizeof(struct in_addr), IP_RECVRETOPTS, IPPROTO_IP);
        !          2075:                if (*mp)
        !          2076:                        mp = &(*mp)->m_next;
        !          2077:        }
        !          2078: #endif
        !          2079:        if (inp->inp_flags & INP_RECVIF) {
        !          2080:                struct sockaddr_dl sdl;
        !          2081:
        !          2082:                sdl.sdl_len = offsetof(struct sockaddr_dl, sdl_data[0]);
        !          2083:                sdl.sdl_family = AF_LINK;
        !          2084:                sdl.sdl_index = m->m_pkthdr.rcvif ?
        !          2085:                    m->m_pkthdr.rcvif->if_index : 0;
        !          2086:                sdl.sdl_nlen = sdl.sdl_alen = sdl.sdl_slen = 0;
        !          2087:                *mp = sbcreatecontrol((void *) &sdl, sdl.sdl_len,
        !          2088:                    IP_RECVIF, IPPROTO_IP);
        !          2089:                if (*mp)
        !          2090:                        mp = &(*mp)->m_next;
        !          2091:        }
        !          2092: }
        !          2093:
        !          2094: /*
        !          2095:  * sysctl helper routine for net.inet.ip.forwsrcrt.
        !          2096:  */
        !          2097: static int
        !          2098: sysctl_net_inet_ip_forwsrcrt(SYSCTLFN_ARGS)
        !          2099: {
        !          2100:        int error, tmp;
        !          2101:        struct sysctlnode node;
        !          2102:
        !          2103:        node = *rnode;
        !          2104:        tmp = ip_forwsrcrt;
        !          2105:        node.sysctl_data = &tmp;
        !          2106:        error = sysctl_lookup(SYSCTLFN_CALL(&node));
        !          2107:        if (error || newp == NULL)
        !          2108:                return (error);
        !          2109:
        !          2110:        if (kauth_authorize_network(l->l_cred, KAUTH_NETWORK_FORWSRCRT,
        !          2111:            0, NULL, NULL, NULL))
        !          2112:                return (EPERM);
        !          2113:
        !          2114:        ip_forwsrcrt = tmp;
        !          2115:
        !          2116:        return (0);
        !          2117: }
        !          2118:
        !          2119: /*
        !          2120:  * sysctl helper routine for net.inet.ip.mtudisctimeout.  checks the
        !          2121:  * range of the new value and tweaks timers if it changes.
        !          2122:  */
        !          2123: static int
        !          2124: sysctl_net_inet_ip_pmtudto(SYSCTLFN_ARGS)
        !          2125: {
        !          2126:        int error, tmp;
        !          2127:        struct sysctlnode node;
        !          2128:
        !          2129:        node = *rnode;
        !          2130:        tmp = ip_mtudisc_timeout;
        !          2131:        node.sysctl_data = &tmp;
        !          2132:        error = sysctl_lookup(SYSCTLFN_CALL(&node));
        !          2133:        if (error || newp == NULL)
        !          2134:                return (error);
        !          2135:        if (tmp < 0)
        !          2136:                return (EINVAL);
        !          2137:
        !          2138:        ip_mtudisc_timeout = tmp;
        !          2139:        rt_timer_queue_change(ip_mtudisc_timeout_q, ip_mtudisc_timeout);
        !          2140:
        !          2141:        return (0);
        !          2142: }
        !          2143:
        !          2144: #ifdef GATEWAY
        !          2145: /*
        !          2146:  * sysctl helper routine for net.inet.ip.maxflows.
        !          2147:  */
        !          2148: static int
        !          2149: sysctl_net_inet_ip_maxflows(SYSCTLFN_ARGS)
        !          2150: {
        !          2151:        int s;
        !          2152:
        !          2153:        s = sysctl_lookup(SYSCTLFN_CALL(rnode));
        !          2154:        if (s || newp == NULL)
        !          2155:                return (s);
        !          2156:
        !          2157:        s = splsoftnet();
        !          2158:        ipflow_reap(0);
        !          2159:        splx(s);
        !          2160:
        !          2161:        return (0);
        !          2162: }
        !          2163:
        !          2164: static int
        !          2165: sysctl_net_inet_ip_hashsize(SYSCTLFN_ARGS)
        !          2166: {
        !          2167:        int error, tmp;
        !          2168:        struct sysctlnode node;
        !          2169:
        !          2170:        node = *rnode;
        !          2171:        tmp = ip_hashsize;
        !          2172:        node.sysctl_data = &tmp;
        !          2173:        error = sysctl_lookup(SYSCTLFN_CALL(&node));
        !          2174:        if (error || newp == NULL)
        !          2175:                return (error);
        !          2176:
        !          2177:        if ((tmp & (tmp - 1)) == 0 && tmp != 0) {
        !          2178:                /*
        !          2179:                 * Can only fail due to malloc()
        !          2180:                 */
        !          2181:                if (ipflow_invalidate_all(tmp))
        !          2182:                        return ENOMEM;
        !          2183:        } else {
        !          2184:                /*
        !          2185:                 * EINVAL if not a power of 2
        !          2186:                 */
        !          2187:                return EINVAL;
        !          2188:        }
        !          2189:
        !          2190:        return (0);
        !          2191: }
        !          2192: #endif /* GATEWAY */
        !          2193:
        !          2194:
        !          2195: SYSCTL_SETUP(sysctl_net_inet_ip_setup, "sysctl net.inet.ip subtree setup")
        !          2196: {
        !          2197:        extern int subnetsarelocal, hostzeroisbroadcast;
        !          2198:
        !          2199:        sysctl_createv(clog, 0, NULL, NULL,
        !          2200:                       CTLFLAG_PERMANENT,
        !          2201:                       CTLTYPE_NODE, "net", NULL,
        !          2202:                       NULL, 0, NULL, 0,
        !          2203:                       CTL_NET, CTL_EOL);
        !          2204:        sysctl_createv(clog, 0, NULL, NULL,
        !          2205:                       CTLFLAG_PERMANENT,
        !          2206:                       CTLTYPE_NODE, "inet",
        !          2207:                       SYSCTL_DESCR("PF_INET related settings"),
        !          2208:                       NULL, 0, NULL, 0,
        !          2209:                       CTL_NET, PF_INET, CTL_EOL);
        !          2210:        sysctl_createv(clog, 0, NULL, NULL,
        !          2211:                       CTLFLAG_PERMANENT,
        !          2212:                       CTLTYPE_NODE, "ip",
        !          2213:                       SYSCTL_DESCR("IPv4 related settings"),
        !          2214:                       NULL, 0, NULL, 0,
        !          2215:                       CTL_NET, PF_INET, IPPROTO_IP, CTL_EOL);
        !          2216:
        !          2217:        sysctl_createv(clog, 0, NULL, NULL,
        !          2218:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2219:                       CTLTYPE_INT, "forwarding",
        !          2220:                       SYSCTL_DESCR("Enable forwarding of INET datagrams"),
        !          2221:                       NULL, 0, &ipforwarding, 0,
        !          2222:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2223:                       IPCTL_FORWARDING, CTL_EOL);
        !          2224:        sysctl_createv(clog, 0, NULL, NULL,
        !          2225:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2226:                       CTLTYPE_INT, "redirect",
        !          2227:                       SYSCTL_DESCR("Enable sending of ICMP redirect messages"),
        !          2228:                       NULL, 0, &ipsendredirects, 0,
        !          2229:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2230:                       IPCTL_SENDREDIRECTS, CTL_EOL);
        !          2231:        sysctl_createv(clog, 0, NULL, NULL,
        !          2232:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2233:                       CTLTYPE_INT, "ttl",
        !          2234:                       SYSCTL_DESCR("Default TTL for an INET datagram"),
        !          2235:                       NULL, 0, &ip_defttl, 0,
        !          2236:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2237:                       IPCTL_DEFTTL, CTL_EOL);
        !          2238: #ifdef IPCTL_DEFMTU
        !          2239:        sysctl_createv(clog, 0, NULL, NULL,
        !          2240:                       CTLFLAG_PERMANENT /* |CTLFLAG_READWRITE? */,
        !          2241:                       CTLTYPE_INT, "mtu",
        !          2242:                       SYSCTL_DESCR("Default MTA for an INET route"),
        !          2243:                       NULL, 0, &ip_mtu, 0,
        !          2244:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2245:                       IPCTL_DEFMTU, CTL_EOL);
        !          2246: #endif /* IPCTL_DEFMTU */
        !          2247:        sysctl_createv(clog, 0, NULL, NULL,
        !          2248:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2249:                       CTLTYPE_INT, "forwsrcrt",
        !          2250:                       SYSCTL_DESCR("Enable forwarding of source-routed "
        !          2251:                                    "datagrams"),
        !          2252:                       sysctl_net_inet_ip_forwsrcrt, 0, &ip_forwsrcrt, 0,
        !          2253:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2254:                       IPCTL_FORWSRCRT, CTL_EOL);
        !          2255:        sysctl_createv(clog, 0, NULL, NULL,
        !          2256:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2257:                       CTLTYPE_INT, "directed-broadcast",
        !          2258:                       SYSCTL_DESCR("Enable forwarding of broadcast datagrams"),
        !          2259:                       NULL, 0, &ip_directedbcast, 0,
        !          2260:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2261:                       IPCTL_DIRECTEDBCAST, CTL_EOL);
        !          2262:        sysctl_createv(clog, 0, NULL, NULL,
        !          2263:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2264:                       CTLTYPE_INT, "allowsrcrt",
        !          2265:                       SYSCTL_DESCR("Accept source-routed datagrams"),
        !          2266:                       NULL, 0, &ip_allowsrcrt, 0,
        !          2267:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2268:                       IPCTL_ALLOWSRCRT, CTL_EOL);
        !          2269:        sysctl_createv(clog, 0, NULL, NULL,
        !          2270:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2271:                       CTLTYPE_INT, "subnetsarelocal",
        !          2272:                       SYSCTL_DESCR("Whether logical subnets are considered "
        !          2273:                                    "local"),
        !          2274:                       NULL, 0, &subnetsarelocal, 0,
        !          2275:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2276:                       IPCTL_SUBNETSARELOCAL, CTL_EOL);
        !          2277:        sysctl_createv(clog, 0, NULL, NULL,
        !          2278:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2279:                       CTLTYPE_INT, "mtudisc",
        !          2280:                       SYSCTL_DESCR("Use RFC1191 Path MTU Discovery"),
        !          2281:                       NULL, 0, &ip_mtudisc, 0,
        !          2282:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2283:                       IPCTL_MTUDISC, CTL_EOL);
        !          2284:        sysctl_createv(clog, 0, NULL, NULL,
        !          2285:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2286:                       CTLTYPE_INT, "anonportmin",
        !          2287:                       SYSCTL_DESCR("Lowest ephemeral port number to assign"),
        !          2288:                       sysctl_net_inet_ip_ports, 0, &anonportmin, 0,
        !          2289:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2290:                       IPCTL_ANONPORTMIN, CTL_EOL);
        !          2291:        sysctl_createv(clog, 0, NULL, NULL,
        !          2292:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2293:                       CTLTYPE_INT, "anonportmax",
        !          2294:                       SYSCTL_DESCR("Highest ephemeral port number to assign"),
        !          2295:                       sysctl_net_inet_ip_ports, 0, &anonportmax, 0,
        !          2296:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2297:                       IPCTL_ANONPORTMAX, CTL_EOL);
        !          2298:        sysctl_createv(clog, 0, NULL, NULL,
        !          2299:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2300:                       CTLTYPE_INT, "mtudisctimeout",
        !          2301:                       SYSCTL_DESCR("Lifetime of a Path MTU Discovered route"),
        !          2302:                       sysctl_net_inet_ip_pmtudto, 0, &ip_mtudisc_timeout, 0,
        !          2303:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2304:                       IPCTL_MTUDISCTIMEOUT, CTL_EOL);
        !          2305: #ifdef GATEWAY
        !          2306:        sysctl_createv(clog, 0, NULL, NULL,
        !          2307:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2308:                       CTLTYPE_INT, "maxflows",
        !          2309:                       SYSCTL_DESCR("Number of flows for fast forwarding"),
        !          2310:                       sysctl_net_inet_ip_maxflows, 0, &ip_maxflows, 0,
        !          2311:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2312:                       IPCTL_MAXFLOWS, CTL_EOL);
        !          2313:        sysctl_createv(clog, 0, NULL, NULL,
        !          2314:                        CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2315:                        CTLTYPE_INT, "hashsize",
        !          2316:                        SYSCTL_DESCR("Size of hash table for fast forwarding (IPv4)"),
        !          2317:                        sysctl_net_inet_ip_hashsize, 0, &ip_hashsize, 0,
        !          2318:                        CTL_NET, PF_INET, IPPROTO_IP,
        !          2319:                        CTL_CREATE, CTL_EOL);
        !          2320: #endif /* GATEWAY */
        !          2321:        sysctl_createv(clog, 0, NULL, NULL,
        !          2322:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2323:                       CTLTYPE_INT, "hostzerobroadcast",
        !          2324:                       SYSCTL_DESCR("All zeroes address is broadcast address"),
        !          2325:                       NULL, 0, &hostzeroisbroadcast, 0,
        !          2326:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2327:                       IPCTL_HOSTZEROBROADCAST, CTL_EOL);
        !          2328: #if NGIF > 0
        !          2329:        sysctl_createv(clog, 0, NULL, NULL,
        !          2330:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2331:                       CTLTYPE_INT, "gifttl",
        !          2332:                       SYSCTL_DESCR("Default TTL for a gif tunnel datagram"),
        !          2333:                       NULL, 0, &ip_gif_ttl, 0,
        !          2334:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2335:                       IPCTL_GIF_TTL, CTL_EOL);
        !          2336: #endif /* NGIF */
        !          2337: #ifndef IPNOPRIVPORTS
        !          2338:        sysctl_createv(clog, 0, NULL, NULL,
        !          2339:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2340:                       CTLTYPE_INT, "lowportmin",
        !          2341:                       SYSCTL_DESCR("Lowest privileged ephemeral port number "
        !          2342:                                    "to assign"),
        !          2343:                       sysctl_net_inet_ip_ports, 0, &lowportmin, 0,
        !          2344:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2345:                       IPCTL_LOWPORTMIN, CTL_EOL);
        !          2346:        sysctl_createv(clog, 0, NULL, NULL,
        !          2347:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2348:                       CTLTYPE_INT, "lowportmax",
        !          2349:                       SYSCTL_DESCR("Highest privileged ephemeral port number "
        !          2350:                                    "to assign"),
        !          2351:                       sysctl_net_inet_ip_ports, 0, &lowportmax, 0,
        !          2352:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2353:                       IPCTL_LOWPORTMAX, CTL_EOL);
        !          2354: #endif /* IPNOPRIVPORTS */
        !          2355:        sysctl_createv(clog, 0, NULL, NULL,
        !          2356:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2357:                       CTLTYPE_INT, "maxfragpackets",
        !          2358:                       SYSCTL_DESCR("Maximum number of fragments to retain for "
        !          2359:                                    "possible reassembly"),
        !          2360:                       NULL, 0, &ip_maxfragpackets, 0,
        !          2361:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2362:                       IPCTL_MAXFRAGPACKETS, CTL_EOL);
        !          2363: #if NGRE > 0
        !          2364:        sysctl_createv(clog, 0, NULL, NULL,
        !          2365:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2366:                       CTLTYPE_INT, "grettl",
        !          2367:                       SYSCTL_DESCR("Default TTL for a gre tunnel datagram"),
        !          2368:                       NULL, 0, &ip_gre_ttl, 0,
        !          2369:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2370:                       IPCTL_GRE_TTL, CTL_EOL);
        !          2371: #endif /* NGRE */
        !          2372:        sysctl_createv(clog, 0, NULL, NULL,
        !          2373:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2374:                       CTLTYPE_INT, "checkinterface",
        !          2375:                       SYSCTL_DESCR("Enable receive side of Strong ES model "
        !          2376:                                    "from RFC1122"),
        !          2377:                       NULL, 0, &ip_checkinterface, 0,
        !          2378:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2379:                       IPCTL_CHECKINTERFACE, CTL_EOL);
        !          2380:        sysctl_createv(clog, 0, NULL, NULL,
        !          2381:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2382:                       CTLTYPE_INT, "random_id",
        !          2383:                       SYSCTL_DESCR("Assign random ip_id values"),
        !          2384:                       NULL, 0, &ip_do_randomid, 0,
        !          2385:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2386:                       IPCTL_RANDOMID, CTL_EOL);
        !          2387:        sysctl_createv(clog, 0, NULL, NULL,
        !          2388:                       CTLFLAG_PERMANENT|CTLFLAG_READWRITE,
        !          2389:                       CTLTYPE_INT, "do_loopback_cksum",
        !          2390:                       SYSCTL_DESCR("Perform IP checksum on loopback"),
        !          2391:                       NULL, 0, &ip_do_loopback_cksum, 0,
        !          2392:                       CTL_NET, PF_INET, IPPROTO_IP,
        !          2393:                       IPCTL_LOOPBACKCKSUM, CTL_EOL);
        !          2394:        sysctl_createv(clog, 0, NULL, NULL,
        !          2395:                       CTLFLAG_PERMANENT,
        !          2396:                       CTLTYPE_STRUCT, "stats",
        !          2397:                       SYSCTL_DESCR("IP statistics"),
        !          2398:                       NULL, 0, &ipstat, sizeof(ipstat),
        !          2399:                       CTL_NET, PF_INET, IPPROTO_IP, IPCTL_STATS,
        !          2400:                       CTL_EOL);
        !          2401: }

CVSweb <webmaster@jp.NetBSD.org>