Please note that diffs are not public domain; they are subject to the copyright notices on the relevant files. =================================================================== RCS file: /ftp/cvs/cvsroot/src/sys/netinet/ip_input.c,v rcsdiff: /ftp/cvs/cvsroot/src/sys/netinet/ip_input.c,v: warning: Unknown phrases like `commitid ...;' are present. retrieving revision 1.92.2.1 retrieving revision 1.137.2.1 diff -u -p -r1.92.2.1 -r1.137.2.1 --- src/sys/netinet/ip_input.c 1999/12/27 18:36:16 1.92.2.1 +++ src/sys/netinet/ip_input.c 2001/11/12 21:19:24 1.137.2.1 @@ -1,4 +1,4 @@ -/* $NetBSD: ip_input.c,v 1.92.2.1 1999/12/27 18:36:16 wrstuden Exp $ */ +/* $NetBSD: ip_input.c,v 1.137.2.1 2001/11/12 21:19:24 thorpej Exp $ */ /* * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project. @@ -105,6 +105,7 @@ #include "opt_pfil_hooks.h" #include "opt_ipsec.h" #include "opt_mrouting.h" +#include "opt_inet_csum.h" #include #include @@ -117,10 +118,7 @@ #include #include #include -#include #include - -#include #include #include @@ -139,10 +137,13 @@ #include #include "gif.h" +#ifdef MROUTING +#include +#endif + #ifdef IPSEC #include #include -#include #endif #ifndef IPFORWARDING @@ -194,18 +195,21 @@ int ipprintfs = 0; struct rttimer_queue *ip_mtudisc_timeout_q = NULL; extern struct domain inetdomain; -extern struct protosw inetsw[]; -u_char ip_protox[IPPROTO_MAX]; int ipqmaxlen = IFQ_MAXLEN; struct in_ifaddrhead in_ifaddr; struct in_ifaddrhashhead *in_ifaddrhashtbl; struct ifqueue ipintrq; struct ipstat ipstat; u_int16_t ip_id; -int ip_defttl; + +#ifdef PFIL_HOOKS +struct pfil_head inet_pfil_hook; +#endif struct ipqhead ipq; int ipq_locked; +int ip_nfragpackets = 0; +int ip_maxfragpackets = 200; static __inline int ipq_lock_try __P((void)); static __inline void ipq_unlock __P((void)); @@ -215,7 +219,11 @@ ipq_lock_try() { int s; - s = splimp(); + /* + * Use splvm() -- we're bloking things that would cause + * mbuf allocation. + */ + s = splvm(); if (ipq_locked) { splx(s); return (0); @@ -230,7 +238,7 @@ ipq_unlock() { int s; - s = splimp(); + s = splvm(); ipq_locked = 0; splx(s); } @@ -259,6 +267,24 @@ do { \ struct pool ipqent_pool; +#ifdef INET_CSUM_COUNTERS +#include + +struct evcnt ip_hwcsum_bad = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, + NULL, "inet", "hwcsum bad"); +struct evcnt ip_hwcsum_ok = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, + NULL, "inet", "hwcsum ok"); +struct evcnt ip_swcsum = EVCNT_INITIALIZER(EVCNT_TYPE_MISC, + NULL, "inet", "swcsum"); + +#define INET_CSUM_COUNTER_INCR(ev) (ev)->ev_count++ + +#else + +#define INET_CSUM_COUNTER_INCR(ev) /* nothing */ + +#endif /* INET_CSUM_COUNTERS */ + /* * We need to save the IP options in case a protocol wants to respond * to an incoming packet over the same route if the packet got here @@ -283,8 +309,8 @@ static void save_rte __P((u_char *, stru void ip_init() { - register struct protosw *pr; - register int i; + struct protosw *pr; + int i; pool_init(&ipqent_pool, sizeof(struct ipqent), 0, 0, 0, "ipqepl", 0, NULL, NULL, M_IPQ); @@ -303,14 +329,30 @@ ip_init() ip_id = time.tv_sec & 0xffff; ipintrq.ifq_maxlen = ipqmaxlen; TAILQ_INIT(&in_ifaddr); - in_ifaddrhashtbl = - hashinit(IN_IFADDR_HASH_SIZE, M_IFADDR, M_WAITOK, &in_ifaddrhash); + in_ifaddrhashtbl = hashinit(IN_IFADDR_HASH_SIZE, HASH_LIST, M_IFADDR, + M_WAITOK, &in_ifaddrhash); if (ip_mtudisc != 0) ip_mtudisc_timeout_q = rt_timer_queue_create(ip_mtudisc_timeout); #ifdef GATEWAY ipflow_init(); #endif + +#ifdef PFIL_HOOKS + /* Register our Packet Filter hook. */ + inet_pfil_hook.ph_type = PFIL_TYPE_AF; + inet_pfil_hook.ph_af = AF_INET; + i = pfil_head_register(&inet_pfil_hook); + if (i != 0) + printf("ip_init: WARNING: unable to register pfil hook, " + "error %d\n", i); +#endif /* PFIL_HOOKS */ + +#ifdef INET_CSUM_COUNTERS + evcnt_attach_static(&ip_hwcsum_bad); + evcnt_attach_static(&ip_hwcsum_ok); + evcnt_attach_static(&ip_swcsum); +#endif /* INET_CSUM_COUNTERS */ } struct sockaddr_in ipaddr = { sizeof(ipaddr), AF_INET }; @@ -326,7 +368,7 @@ ipintr() struct mbuf *m; while (1) { - s = splimp(); + s = splnet(); IF_DEQUEUE(&ipintrq, m); splx(s); if (m == 0) @@ -342,17 +384,13 @@ ipintr() void ip_input(struct mbuf *m) { - register struct ip *ip = NULL; - register struct ipq *fp; - register struct in_ifaddr *ia; - register struct ifaddr *ifa; + struct ip *ip = NULL; + struct ipq *fp; + struct in_ifaddr *ia; + struct ifaddr *ifa; struct ipqent *ipqe; int hlen = 0, mff, len; -#ifdef PFIL_HOOKS - struct packet_filter_hook *pfh; - struct mbuf *m0; - int rv; -#endif /* PFIL_HOOKS */ + int downmatch; #ifdef DIAGNOSTIC if ((m->m_flags & M_PKTHDR) == 0) @@ -372,7 +410,7 @@ ip_input(struct mbuf *m) * If no IP addresses have been set yet but the interfaces * are receiving, can't do anything with incoming packets yet. */ - if (in_ifaddr.tqh_first == 0) + if (TAILQ_FIRST(&in_ifaddr) == 0) goto bad; ipstat.ips_total++; if (m->m_len < sizeof (struct ip) && @@ -397,25 +435,47 @@ ip_input(struct mbuf *m) } ip = mtod(m, struct ip *); } + /* - * we drop packets that have a multicast address as source - * as wanted by rfc 1112 + * RFC1122: packets with a multicast source address are + * not allowed. */ if (IN_MULTICAST(ip->ip_src.s_addr)) { + ipstat.ips_badaddr++; goto bad; } - if (in_cksum(m, hlen) != 0) { - ipstat.ips_badsum++; - goto bad; + /* 127/8 must not appear on wire - RFC1122 */ + if ((ntohl(ip->ip_dst.s_addr) >> IN_CLASSA_NSHIFT) == IN_LOOPBACKNET || + (ntohl(ip->ip_src.s_addr) >> IN_CLASSA_NSHIFT) == IN_LOOPBACKNET) { + if ((m->m_pkthdr.rcvif->if_flags & IFF_LOOPBACK) == 0) { + ipstat.ips_badaddr++; + goto bad; + } } - /* - * Convert fields to host representation. - */ - NTOHS(ip->ip_len); - NTOHS(ip->ip_off); - len = ip->ip_len; + switch (m->m_pkthdr.csum_flags & + ((m->m_pkthdr.rcvif->if_csum_flags_rx & M_CSUM_IPv4) | + M_CSUM_IPv4_BAD)) { + case M_CSUM_IPv4|M_CSUM_IPv4_BAD: + INET_CSUM_COUNTER_INCR(&ip_hwcsum_bad); + goto badcsum; + + case M_CSUM_IPv4: + /* Checksum was okay. */ + INET_CSUM_COUNTER_INCR(&ip_hwcsum_ok); + break; + + default: + /* Must compute it ourselves. */ + INET_CSUM_COUNTER_INCR(&ip_swcsum); + if (in_cksum(m, hlen) != 0) + goto bad; + break; + } + + /* Retrieve the packet length. */ + len = ntohs(ip->ip_len); /* * Check for additional length bogosity @@ -462,19 +522,39 @@ ip_input(struct mbuf *m) * Note that filters must _never_ set this flag, as another filter * in the list may have previously cleared it. */ - m0 = m; - for (pfh = pfil_hook_get(PFIL_IN); pfh; pfh = pfh->pfil_link.tqe_next) - if (pfh->pfil_func) { - rv = pfh->pfil_func(ip, hlen, m->m_pkthdr.rcvif, 0, &m0); - if (rv) - return; - m = m0; - if (m == NULL) - return; - ip = mtod(m, struct ip *); - } + /* + * let ipfilter look at packet on the wire, + * not the decapsulated packet. + */ +#ifdef IPSEC + if (!ipsec_getnhist(m)) +#else + if (1) +#endif + { + if (pfil_run_hooks(&inet_pfil_hook, &m, m->m_pkthdr.rcvif, + PFIL_IN) != 0) + return; + if (m == NULL) + return; + ip = mtod(m, struct ip *); + } #endif /* PFIL_HOOKS */ +#ifdef ALTQ + /* XXX Temporary until ALTQ is changed to use a pfil hook */ + if (altq_input != NULL && (*altq_input)(m, AF_INET) == 0) { + /* packet dropped by traffic conditioner */ + return; + } +#endif + + /* + * Convert fields to host representation. + */ + NTOHS(ip->ip_len); + NTOHS(ip->ip_off); + /* * Process options and, if not destined for us, * ship it on. ip_dooptions returns 1 when an @@ -487,14 +567,26 @@ ip_input(struct mbuf *m) /* * Check our list of addresses, to see if the packet is for us. - */ - INADDR_TO_IA(ip->ip_dst, ia); + * + * Traditional 4.4BSD did not consult IFF_UP at all. + * The behavior here is to treat addresses on !IFF_UP interface + * as not mine. + */ + downmatch = 0; + LIST_FOREACH(ia, &IN_IFADDR_HASH(ip->ip_dst.s_addr), ia_hash) { + if (in_hosteq(ia->ia_addr.sin_addr, ip->ip_dst)) { + if ((ia->ia_ifp->if_flags & IFF_UP) != 0) + break; + else + downmatch++; + } + } if (ia != NULL) goto ours; if (m->m_pkthdr.rcvif->if_flags & IFF_BROADCAST) { - for (ifa = m->m_pkthdr.rcvif->if_addrlist.tqh_first; - ifa != NULL; ifa = ifa->ifa_list.tqe_next) { - if (ifa->ifa_addr->sa_family != AF_INET) continue; + TAILQ_FOREACH(ifa, &m->m_pkthdr.rcvif->if_addrlist, ifa_list) { + if (ifa->ifa_addr->sa_family != AF_INET) + continue; ia = ifatoia(ifa); if (in_hosteq(ip->ip_dst, ia->ia_broadaddr.sin_addr) || in_hosteq(ip->ip_dst, ia->ia_netbroadcast) || @@ -577,8 +669,20 @@ ip_input(struct mbuf *m) if (ipforwarding == 0) { ipstat.ips_cantforward++; m_freem(m); - } else + } else { + /* + * If ip_dst matched any of my address on !IFF_UP interface, + * and there's no IFF_UP interface that matches ip_dst, + * send icmp unreach. Forwarding it will result in in-kernel + * forwarding loop till TTL goes to 0. + */ + if (downmatch) { + icmp_error(m, ICMP_UNREACH, ICMP_UNREACH_HOST, 0, 0); + ipstat.ips_cantforward++; + return; + } ip_forward(m, 0); + } return; ours: @@ -595,7 +699,7 @@ ours: * of this datagram. */ IPQ_LOCK(); - for (fp = ipq.lh_first; fp != NULL; fp = fp->ipq_q.le_next) + LIST_FOREACH(fp, &ipq, ipq_q) if (ip->ip_id == fp->ipq_id && in_hosteq(ip->ip_src, fp->ipq_src) && in_hosteq(ip->ip_dst, fp->ipq_dst) && @@ -655,11 +759,25 @@ found: IPQ_UNLOCK(); } +#ifdef IPSEC + /* + * enforce IPsec policy checking if we are seeing last header. + * note that we do not visit this with protocols with pcb layer + * code - like udp/tcp/raw ip. + */ + if ((inetsw[ip_protox[ip->ip_p]].pr_flags & PR_LASTHDR) != 0 && + ipsec4_in_reject(m, NULL)) { + ipsecstat.in_polvio++; + goto bad; + } +#endif + /* * Switch out to protocol's input routine. */ #if IFA_STATS - ia->ia_ifa.ifa_data.ifad_inbytes += ip->ip_len; + if (ia && ip) + ia->ia_ifa.ifa_data.ifad_inbytes += ip->ip_len; #endif ipstat.ips_delivered++; { @@ -670,6 +788,11 @@ found: } bad: m_freem(m); + return; + +badcsum: + ipstat.ips_badsum++; + m_freem(m); } /* @@ -680,11 +803,11 @@ bad: */ struct mbuf * ip_reass(ipqe, fp) - register struct ipqent *ipqe; - register struct ipq *fp; + struct ipqent *ipqe; + struct ipq *fp; { - register struct mbuf *m = ipqe->ipqe_m; - register struct ipqent *nq, *p, *q; + struct mbuf *m = ipqe->ipqe_m; + struct ipqent *nq, *p, *q; struct ip *ip; struct mbuf *t; int hlen = ipqe->ipqe_ip->ip_hl << 2; @@ -703,6 +826,17 @@ ip_reass(ipqe, fp) * If first fragment to arrive, create a reassembly queue. */ if (fp == 0) { + /* + * Enforce upper bound on number of fragmented packets + * for which we attempt reassembly; + * If maxfrag is 0, never accept fragments. + * If maxfrag is -1, accept all fragments without limitation. + */ + if (ip_maxfragpackets < 0) + ; + else if (ip_nfragpackets >= ip_maxfragpackets) + goto dropfrag; + ip_nfragpackets++; MALLOC(fp, struct ipq *, sizeof (struct ipq), M_FTABLE, M_NOWAIT); if (fp == NULL) @@ -721,8 +855,8 @@ ip_reass(ipqe, fp) /* * Find a segment which begins after this one does. */ - for (p = NULL, q = fp->ipq_fragq.lh_first; q != NULL; - p = q, q = q->ipqe_q.le_next) + for (p = NULL, q = LIST_FIRST(&fp->ipq_fragq); q != NULL; + p = q, q = LIST_NEXT(q, ipqe_q)) if (q->ipqe_ip->ip_off > ipqe->ipqe_ip->ip_off) break; @@ -757,7 +891,7 @@ ip_reass(ipqe, fp) m_adj(q->ipqe_m, i); break; } - nq = q->ipqe_q.le_next; + nq = LIST_NEXT(q, ipqe_q); m_freem(q->ipqe_m); LIST_REMOVE(q, ipqe_q); pool_put(&ipqent_pool, q); @@ -774,8 +908,8 @@ insert: LIST_INSERT_AFTER(p, ipqe, ipqe_q); } next = 0; - for (p = NULL, q = fp->ipq_fragq.lh_first; q != NULL; - p = q, q = q->ipqe_q.le_next) { + for (p = NULL, q = LIST_FIRST(&fp->ipq_fragq); q != NULL; + p = q, q = LIST_NEXT(q, ipqe_q)) { if (q->ipqe_ip->ip_off != next) return (0); next += q->ipqe_ip->ip_len; @@ -787,7 +921,7 @@ insert: * Reassembly is complete. Check for a bogus message size and * concatenate fragments. */ - q = fp->ipq_fragq.lh_first; + q = LIST_FIRST(&fp->ipq_fragq); ip = q->ipqe_ip; if ((next + (ip->ip_hl << 2)) > IP_MAXPACKET) { ipstat.ips_toolong++; @@ -798,11 +932,11 @@ insert: t = m->m_next; m->m_next = 0; m_cat(m, t); - nq = q->ipqe_q.le_next; + nq = LIST_NEXT(q, ipqe_q); pool_put(&ipqent_pool, q); for (q = nq; q != NULL; q = nq) { t = q->ipqe_m; - nq = q->ipqe_q.le_next; + nq = LIST_NEXT(q, ipqe_q); pool_put(&ipqent_pool, q); m_cat(m, t); } @@ -814,17 +948,16 @@ insert: * Make header visible. */ ip->ip_len = next; - ip->ip_ttl = 0; /* xxx */ - ip->ip_sum = 0; ip->ip_src = fp->ipq_src; ip->ip_dst = fp->ipq_dst; LIST_REMOVE(fp, ipq_q); FREE(fp, M_FTABLE); + ip_nfragpackets--; m->m_len += (ip->ip_hl << 2); m->m_data -= (ip->ip_hl << 2); /* some debugging cruft by sklower, below, will go away soon */ if (m->m_flags & M_PKTHDR) { /* XXX this should be done elsewhere */ - register int plen = 0; + int plen = 0; for (t = m; t; t = t->m_next) plen += t->m_len; m->m_pkthdr.len = plen; @@ -846,18 +979,19 @@ void ip_freef(fp) struct ipq *fp; { - register struct ipqent *q, *p; + struct ipqent *q, *p; IPQ_LOCK_CHECK(); - for (q = fp->ipq_fragq.lh_first; q != NULL; q = p) { - p = q->ipqe_q.le_next; + for (q = LIST_FIRST(&fp->ipq_fragq); q != NULL; q = p) { + p = LIST_NEXT(q, ipqe_q); m_freem(q->ipqe_m); LIST_REMOVE(q, ipqe_q); pool_put(&ipqent_pool, q); } LIST_REMOVE(fp, ipq_q); FREE(fp, M_FTABLE); + ip_nfragpackets--; } /* @@ -868,17 +1002,28 @@ ip_freef(fp) void ip_slowtimo() { - register struct ipq *fp, *nfp; + struct ipq *fp, *nfp; int s = splsoftnet(); IPQ_LOCK(); - for (fp = ipq.lh_first; fp != NULL; fp = nfp) { - nfp = fp->ipq_q.le_next; + for (fp = LIST_FIRST(&ipq); fp != NULL; fp = nfp) { + nfp = LIST_NEXT(fp, ipq_q); if (--fp->ipq_ttl == 0) { ipstat.ips_fragtimeout++; ip_freef(fp); } } + /* + * If we are over the maximum number of fragments + * (due to the limit being lowered), drain off + * enough to get down to the new limit. + */ + if (ip_maxfragpackets < 0) + ; + else { + while (ip_nfragpackets > ip_maxfragpackets && LIST_FIRST(&ipq)) + ip_freef(LIST_FIRST(&ipq)); + } IPQ_UNLOCK(); #ifdef GATEWAY ipflow_slowtimo(); @@ -900,9 +1045,9 @@ ip_drain() if (ipq_lock_try() == 0) return; - while (ipq.lh_first != NULL) { + while (LIST_FIRST(&ipq) != NULL) { ipstat.ips_fragdropped++; - ip_freef(ipq.lh_first); + ip_freef(LIST_FIRST(&ipq)); } IPQ_UNLOCK(); @@ -919,12 +1064,12 @@ int ip_dooptions(m) struct mbuf *m; { - register struct ip *ip = mtod(m, struct ip *); - register u_char *cp; - register struct ip_timestamp *ipt; - register struct in_ifaddr *ia; + struct ip *ip = mtod(m, struct ip *); + u_char *cp, *cp0; + struct ip_timestamp *ipt; + struct in_ifaddr *ia; int opt, optlen, cnt, off, code, type = ICMP_PARAMPROB, forward = 0; - struct in_addr *sin, dst; + struct in_addr dst; n_time ntime; dst = ip->ip_dst; @@ -937,8 +1082,12 @@ ip_dooptions(m) if (opt == IPOPT_NOP) optlen = 1; else { + if (cnt < IPOPT_OLEN + sizeof(*cp)) { + code = &cp[IPOPT_OLEN] - (u_char *)ip; + goto bad; + } optlen = cp[IPOPT_OLEN]; - if (optlen <= 0 || optlen > cnt) { + if (optlen < IPOPT_OLEN + sizeof(*cp) || optlen > cnt) { code = &cp[IPOPT_OLEN] - (u_char *)ip; goto bad; } @@ -964,6 +1113,10 @@ ip_dooptions(m) code = ICMP_UNREACH_NET_PROHIB; goto bad; } + if (optlen < IPOPT_OFFSET + sizeof(*cp)) { + code = &cp[IPOPT_OLEN] - (u_char *)ip; + goto bad; + } if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) { code = &cp[IPOPT_OFFSET] - (u_char *)ip; goto bad; @@ -983,7 +1136,7 @@ ip_dooptions(m) break; } off--; /* 0 origin */ - if (off > optlen - sizeof(struct in_addr)) { + if ((off + sizeof(struct in_addr)) > optlen) { /* * End of source route. Should be for us. */ @@ -995,11 +1148,9 @@ ip_dooptions(m) */ bcopy((caddr_t)(cp + off), (caddr_t)&ipaddr.sin_addr, sizeof(ipaddr.sin_addr)); - if (opt == IPOPT_SSRR) { -#define INA struct in_ifaddr * -#define SA struct sockaddr * - ia = (INA)ifa_ifwithladdr((SA)&ipaddr); - } else + if (opt == IPOPT_SSRR) + ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr))); + else ia = ip_rtaddr(ipaddr.sin_addr); if (ia == 0) { type = ICMP_UNREACH; @@ -1017,6 +1168,10 @@ ip_dooptions(m) break; case IPOPT_RR: + if (optlen < IPOPT_OFFSET + sizeof(*cp)) { + code = &cp[IPOPT_OLEN] - (u_char *)ip; + goto bad; + } if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) { code = &cp[IPOPT_OFFSET] - (u_char *)ip; goto bad; @@ -1025,7 +1180,7 @@ ip_dooptions(m) * If no space remains, ignore. */ off--; /* 0 origin */ - if (off > optlen - sizeof(struct in_addr)) + if ((off + sizeof(struct in_addr)) > optlen) break; bcopy((caddr_t)(&ip->ip_dst), (caddr_t)&ipaddr.sin_addr, sizeof(ipaddr.sin_addr)); @@ -1033,8 +1188,9 @@ ip_dooptions(m) * locate outgoing interface; if we're the destination, * use the incoming interface (should be same). */ - if ((ia = (INA)ifa_ifwithaddr((SA)&ipaddr)) == 0 && - (ia = ip_rtaddr(ipaddr.sin_addr)) == 0) { + if ((ia = ifatoia(ifa_ifwithaddr(sintosa(&ipaddr)))) + == NULL && + (ia = ip_rtaddr(ipaddr.sin_addr)) == NULL) { type = ICMP_UNREACH; code = ICMP_UNREACH_HOST; goto bad; @@ -1047,14 +1203,23 @@ ip_dooptions(m) case IPOPT_TS: code = cp - (u_char *)ip; ipt = (struct ip_timestamp *)cp; - if (ipt->ipt_len < 5) + if (ipt->ipt_len < 4 || ipt->ipt_len > 40) { + code = (u_char *)&ipt->ipt_len - (u_char *)ip; goto bad; + } + if (ipt->ipt_ptr < 5) { + code = (u_char *)&ipt->ipt_ptr - (u_char *)ip; + goto bad; + } if (ipt->ipt_ptr > ipt->ipt_len - sizeof (int32_t)) { - if (++ipt->ipt_oflw == 0) + if (++ipt->ipt_oflw == 0) { + code = (u_char *)&ipt->ipt_ptr - + (u_char *)ip; goto bad; + } break; } - sin = (struct in_addr *)(cp + ipt->ipt_ptr - 1); + cp0 = (cp + ipt->ipt_ptr - 1); switch (ipt->ipt_flg) { case IPOPT_TS_TSONLY: @@ -1062,34 +1227,45 @@ ip_dooptions(m) case IPOPT_TS_TSANDADDR: if (ipt->ipt_ptr - 1 + sizeof(n_time) + - sizeof(struct in_addr) > ipt->ipt_len) + sizeof(struct in_addr) > ipt->ipt_len) { + code = (u_char *)&ipt->ipt_ptr - + (u_char *)ip; goto bad; + } ipaddr.sin_addr = dst; - ia = (INA)ifaof_ifpforaddr((SA)&ipaddr, - m->m_pkthdr.rcvif); + ia = ifatoia(ifaof_ifpforaddr(sintosa(&ipaddr), + m->m_pkthdr.rcvif)); if (ia == 0) continue; - bcopy((caddr_t)&ia->ia_addr.sin_addr, - (caddr_t)sin, sizeof(struct in_addr)); + bcopy(&ia->ia_addr.sin_addr, + cp0, sizeof(struct in_addr)); ipt->ipt_ptr += sizeof(struct in_addr); break; case IPOPT_TS_PRESPEC: if (ipt->ipt_ptr - 1 + sizeof(n_time) + - sizeof(struct in_addr) > ipt->ipt_len) + sizeof(struct in_addr) > ipt->ipt_len) { + code = (u_char *)&ipt->ipt_ptr - + (u_char *)ip; goto bad; - bcopy((caddr_t)sin, (caddr_t)&ipaddr.sin_addr, + } + bcopy(cp0, &ipaddr.sin_addr, sizeof(struct in_addr)); - if (ifa_ifwithaddr((SA)&ipaddr) == 0) + if (ifatoia(ifa_ifwithaddr(sintosa(&ipaddr))) + == NULL) continue; ipt->ipt_ptr += sizeof(struct in_addr); break; default: + /* XXX can't take &ipt->ipt_flg */ + code = (u_char *)&ipt->ipt_ptr - + (u_char *)ip + 1; goto bad; } ntime = iptime(); - bcopy((caddr_t)&ntime, (caddr_t)cp + ipt->ipt_ptr - 1, + cp0 = (u_char *) &ntime; /* XXX grumble, GCC... */ + bcopy(cp0, (caddr_t)cp + ipt->ipt_ptr - 1, sizeof(n_time)); ipt->ipt_ptr += sizeof(n_time); } @@ -1118,7 +1294,7 @@ struct in_ifaddr * ip_rtaddr(dst) struct in_addr dst; { - register struct sockaddr_in *sin; + struct sockaddr_in *sin; sin = satosin(&ipforward_rt.ro_dst); @@ -1169,8 +1345,8 @@ save_rte(option, dst) struct mbuf * ip_srcroute() { - register struct in_addr *p, *q; - register struct mbuf *m; + struct in_addr *p, *q; + struct mbuf *m; if (ip_nhops == 0) return ((struct mbuf *)0); @@ -1239,12 +1415,12 @@ ip_srcroute() */ void ip_stripoptions(m, mopt) - register struct mbuf *m; + struct mbuf *m; struct mbuf *mopt; { - register int i; + int i; struct ip *ip = mtod(m, struct ip *); - register caddr_t opts; + caddr_t opts; int olen; olen = (ip->ip_hl << 2) - sizeof (struct ip); @@ -1258,7 +1434,7 @@ ip_stripoptions(m, mopt) ip->ip_hl = sizeof (struct ip) >> 2; } -int inetctlerrmap[PRC_NCMDS] = { +const int inetctlerrmap[PRC_NCMDS] = { 0, 0, 0, 0, 0, EMSGSIZE, EHOSTDOWN, EHOSTUNREACH, EHOSTUNREACH, EHOSTUNREACH, ECONNREFUSED, ECONNREFUSED, @@ -1286,9 +1462,9 @@ ip_forward(m, srcrt) struct mbuf *m; int srcrt; { - register struct ip *ip = mtod(m, struct ip *); - register struct sockaddr_in *sin; - register struct rtentry *rt; + struct ip *ip = mtod(m, struct ip *); + struct sockaddr_in *sin; + struct rtentry *rt; int error, type = 0, code = 0; struct mbuf *mcopy; n_long dest; @@ -1297,6 +1473,11 @@ ip_forward(m, srcrt) struct ifnet dummyifp; #endif + /* + * Clear any in-bound checksum flags for this packet. + */ + m->m_pkthdr.csum_flags = 0; + dest = 0; #ifdef DIAGNOSTIC if (ipprintfs) @@ -1337,8 +1518,11 @@ ip_forward(m, srcrt) /* * Save at most 68 bytes of the packet in case * we need to generate an ICMP message to the src. + * Pullup to avoid sharing mbuf cluster between m and mcopy. */ - mcopy = m_copy(m, 0, imin((int)ip->ip_len, 68)); + mcopy = m_copym(m, 0, imin((int)ip->ip_len, 68), M_DONTWAIT); + if (mcopy) + mcopy = m_pullup(mcopy, ip->ip_hl << 2); /* * If forwarding packet using same interface that it came in on, @@ -1374,8 +1558,9 @@ ip_forward(m, srcrt) } #ifdef IPSEC - m->m_pkthdr.rcvif = NULL; -#endif /*IPSEC*/ + /* Don't lookup socket in forwarding case */ + (void)ipsec_setsocket(m, NULL); +#endif error = ip_output(m, (struct mbuf *)0, &ipforward_rt, (IP_FORWARDING | (ip_directedbcast ? IP_ALLOWBROADCAST : 0)), 0); if (error) @@ -1430,18 +1615,21 @@ ip_forward(m, srcrt) if (ipforward_rt.ro_rt) { struct secpolicy *sp; int ipsecerror; - int ipsechdr; + size_t ipsechdr; struct route *ro; sp = ipsec4_getpolicybyaddr(mcopy, - IP_FORWARDING, - &ipsecerror); + IPSEC_DIR_OUTBOUND, + IP_FORWARDING, + &ipsecerror); if (sp == NULL) destifp = ipforward_rt.ro_rt->rt_ifp; else { /* count IPsec header size */ - ipsechdr = ipsec4_hdrsiz(mcopy, NULL); + ipsechdr = ipsec4_hdrsiz(mcopy, + IPSEC_DIR_OUTBOUND, + NULL); /* * find the correct route for outer IPv4 @@ -1454,8 +1642,9 @@ ip_forward(m, srcrt) /*XXX*/ destifp = NULL; if (sp->req != NULL - && sp->req->sa != NULL) { - ro = &sp->req->sa->saidx->sa_route; + && sp->req->sav != NULL + && sp->req->sav->sah != NULL) { + ro = &sp->req->sav->sah->sa_route; if (ro->ro_rt && ro->ro_rt->rt_ifp) { dummyifp.if_mtu = ro->ro_rt->rt_ifp->if_mtu; @@ -1481,10 +1670,10 @@ ip_forward(m, srcrt) void ip_savecontrol(inp, mp, ip, m) - register struct inpcb *inp; - register struct mbuf **mp; - register struct ip *ip; - register struct mbuf *m; + struct inpcb *inp; + struct mbuf **mp; + struct ip *ip; + struct mbuf *m; { if (inp->inp_socket->so_options & SO_TIMESTAMP) { @@ -1599,7 +1788,8 @@ ip_sysctl(name, namelen, oldp, oldlenp, case IPCTL_ANONPORTMIN: old = anonportmin; error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmin); - if (anonportmin >= anonportmax || anonportmin > 65535 + if (anonportmin >= anonportmax || anonportmin < 0 + || anonportmin > 65535 #ifndef IPNOPRIVPORTS || anonportmin < IPPORT_RESERVED #endif @@ -1611,7 +1801,8 @@ ip_sysctl(name, namelen, oldp, oldlenp, case IPCTL_ANONPORTMAX: old = anonportmax; error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmax); - if (anonportmin >= anonportmax || anonportmax > 65535 + if (anonportmin >= anonportmax || anonportmax < 0 + || anonportmax > 65535 #ifndef IPNOPRIVPORTS || anonportmax < IPPORT_RESERVED #endif @@ -1649,6 +1840,35 @@ ip_sysctl(name, namelen, oldp, oldlenp, &ip_gif_ttl)); #endif +#ifndef IPNOPRIVPORTS + case IPCTL_LOWPORTMIN: + old = lowportmin; + error = sysctl_int(oldp, oldlenp, newp, newlen, &lowportmin); + if (lowportmin >= lowportmax + || lowportmin > IPPORT_RESERVEDMAX + || lowportmin < IPPORT_RESERVEDMIN + ) { + lowportmin = old; + return (EINVAL); + } + return (error); + case IPCTL_LOWPORTMAX: + old = lowportmax; + error = sysctl_int(oldp, oldlenp, newp, newlen, &lowportmax); + if (lowportmin >= lowportmax + || lowportmax > IPPORT_RESERVEDMAX + || lowportmax < IPPORT_RESERVEDMIN + ) { + lowportmax = old; + return (EINVAL); + } + return (error); +#endif + + case IPCTL_MAXFRAGPACKETS: + return (sysctl_int(oldp, oldlenp, newp, newlen, + &ip_maxfragpackets)); + default: return (EOPNOTSUPP); }