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 1.20 retrieving revision 1.82.2.5 diff -u -p -r1.20 -r1.82.2.5 --- src/sys/netinet/ip_input.c 1995/06/04 05:58:26 1.20 +++ src/sys/netinet/ip_input.c 2000/03/02 10:24:18 1.82.2.5 @@ -1,4 +1,41 @@ -/* $NetBSD: ip_input.c,v 1.20 1995/06/04 05:58:26 mycroft Exp $ */ +/* $NetBSD: ip_input.c,v 1.82.2.5 2000/03/02 10:24:18 he Exp $ */ + +/*- + * Copyright (c) 1998 The NetBSD Foundation, Inc. + * All rights reserved. + * + * This code is derived from software contributed to The NetBSD Foundation + * by Public Access Networks Corporation ("Panix"). It was developed under + * contract to Panix by Eric Haszlakiewicz and Thor Lancelot Simon. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. All advertising materials mentioning features or use of this software + * must display the following acknowledgement: + * This product includes software developed by the NetBSD + * Foundation, Inc. and its contributors. + * 4. Neither the name of The NetBSD Foundation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS + * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED + * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS + * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN + * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) + * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE + * POSSIBILITY OF SUCH DAMAGE. + */ /* * Copyright (c) 1982, 1986, 1988, 1993 @@ -35,6 +72,10 @@ * @(#)ip_input.c 8.2 (Berkeley) 1/4/94 */ +#include "opt_gateway.h" +#include "opt_pfil_hooks.h" +#include "opt_mrouting.h" + #include #include #include @@ -42,12 +83,20 @@ #include #include #include +#include #include #include #include +#include +#include + +#include +#include #include +#include #include +#include #include #include @@ -67,19 +116,109 @@ #ifndef IPSENDREDIRECTS #define IPSENDREDIRECTS 1 #endif +#ifndef IPFORWSRCRT +#define IPFORWSRCRT 1 /* forward source-routed packets */ +#endif +#ifndef IPALLOWSRCRT +#define IPALLOWSRCRT 1 /* allow source-routed packets */ +#endif +#ifndef IPMTUDISC +#define IPMTUDISC 0 +#endif +#ifndef IPMTUDISCTIMEOUT +#define IPMTUDISCTIMEOUT (10 * 60) /* as per RFC 1191 */ +#endif + +/* + * Note: DIRECTED_BROADCAST is handled this way so that previous + * configuration using this option will Just Work. + */ +#ifndef IPDIRECTEDBCAST +#ifdef DIRECTED_BROADCAST +#define IPDIRECTEDBCAST 1 +#else +#define IPDIRECTEDBCAST 0 +#endif /* DIRECTED_BROADCAST */ +#endif /* IPDIRECTEDBCAST */ int ipforwarding = IPFORWARDING; int ipsendredirects = IPSENDREDIRECTS; int ip_defttl = IPDEFTTL; +int ip_forwsrcrt = IPFORWSRCRT; +int ip_directedbcast = IPDIRECTEDBCAST; +int ip_allowsrcrt = IPALLOWSRCRT; +int ip_mtudisc = IPMTUDISC; +u_int ip_mtudisc_timeout = IPMTUDISCTIMEOUT; #ifdef DIAGNOSTIC int ipprintfs = 0; #endif +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_ifaddr *in_ifaddr; /* first inet address */ +struct in_ifaddrhead in_ifaddr; +struct in_ifaddrhashhead *in_ifaddrhashtbl; struct ifqueue ipintrq; +struct ipstat ipstat; +u_int16_t ip_id; +int ip_defttl; + +struct ipqhead ipq; +int ipq_locked; + +static __inline int ipq_lock_try __P((void)); +static __inline void ipq_unlock __P((void)); + +static __inline int +ipq_lock_try() +{ + int s; + + s = splimp(); + if (ipq_locked) { + splx(s); + return (0); + } + ipq_locked = 1; + splx(s); + return (1); +} + +static __inline void +ipq_unlock() +{ + int s; + + s = splimp(); + ipq_locked = 0; + splx(s); +} + +#ifdef DIAGNOSTIC +#define IPQ_LOCK() \ +do { \ + if (ipq_lock_try() == 0) { \ + printf("%s:%d: ipq already locked\n", __FILE__, __LINE__); \ + panic("ipq_lock"); \ + } \ +} while (0) +#define IPQ_LOCK_CHECK() \ +do { \ + if (ipq_locked == 0) { \ + printf("%s:%d: ipq lock not held\n", __FILE__, __LINE__); \ + panic("ipq lock check"); \ + } \ +} while (0) +#else +#define IPQ_LOCK() (void) ipq_lock_try() +#define IPQ_LOCK_CHECK() /* nothing */ +#endif + +#define IPQ_UNLOCK() ipq_unlock() + +struct pool ipqent_pool; /* * We need to save the IP options in case a protocol wants to respond @@ -96,12 +235,8 @@ static struct ip_srcrt { struct in_addr route[MAX_IPOPTLEN/sizeof(struct in_addr)]; } ip_srcrt; -#ifdef GATEWAY -extern int if_index; -u_int32_t *ip_ifmatrix; -#endif - static void save_rte __P((u_char *, struct in_addr)); + /* * IP initialization: fill in IP protocol switch table. * All protocols not implemented in kernel go to raw IP protocol handler. @@ -112,6 +247,9 @@ ip_init() register struct protosw *pr; register int i; + pool_init(&ipqent_pool, sizeof(struct ipqent), 0, 0, 0, "ipqepl", + 0, NULL, NULL, M_IPQ); + pr = pffindproto(PF_INET, IPPROTO_RAW, SOCK_RAW); if (pr == 0) panic("ip_init"); @@ -122,13 +260,17 @@ ip_init() if (pr->pr_domain->dom_family == PF_INET && pr->pr_protocol && pr->pr_protocol != IPPROTO_RAW) ip_protox[pr->pr_protocol] = pr - inetsw; - ipq.next = ipq.prev = &ipq; + LIST_INIT(&ipq); 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); + if (ip_mtudisc != 0) + ip_mtudisc_timeout_q = + rt_timer_queue_create(ip_mtudisc_timeout); #ifdef GATEWAY - i = (if_index + 1) * (if_index + 1) * sizeof (u_int32_t); - ip_ifmatrix = (u_int32_t *) malloc(i, M_RTABLE, M_WAITOK); - bzero((char *)ip_ifmatrix, i); + ipflow_init(); #endif } @@ -142,11 +284,18 @@ struct route ipforward_rt; void ipintr() { - register struct ip *ip; + register struct ip *ip = NULL; register struct mbuf *m; register struct ipq *fp; register struct in_ifaddr *ia; - int hlen, s; + register struct ifaddr *ifa; + struct ipqent *ipqe; + int hlen = 0, mff, len, s; +#ifdef PFIL_HOOKS + struct packet_filter_hook *pfh; + struct mbuf *m0; + int rv; +#endif /* PFIL_HOOKS */ next: /* @@ -166,7 +315,7 @@ next: * If no IP addresses have been set yet but the interfaces * are receiving, can't do anything with incoming packets yet. */ - if (in_ifaddr == NULL) + if (in_ifaddr.tqh_first == 0) goto bad; ipstat.ips_total++; if (m->m_len < sizeof (struct ip) && @@ -191,7 +340,17 @@ next: } ip = mtod(m, struct ip *); } - if (ip->ip_sum = in_cksum(m, hlen)) { + + /* + * RFC1122: packets with a multicast source address are + * not allowed. + */ + if (IN_MULTICAST(ip->ip_src.s_addr)) { + /* XXX stat */ + goto bad; + } + + if (in_cksum(m, hlen) != 0) { ipstat.ips_badsum++; goto bad; } @@ -200,12 +359,17 @@ next: * Convert fields to host representation. */ NTOHS(ip->ip_len); - if (ip->ip_len < hlen) { - ipstat.ips_badlen++; + NTOHS(ip->ip_off); + len = ip->ip_len; + + /* + * Check for additional length bogosity + */ + if (len < hlen) + { + ipstat.ips_badlen++; goto bad; } - NTOHS(ip->ip_id); - NTOHS(ip->ip_off); /* * Check that the amount of data in the buffers @@ -213,19 +377,46 @@ next: * Trim mbufs if longer than we expect. * Drop packet if shorter than we expect. */ - if (m->m_pkthdr.len < ip->ip_len) { + if (m->m_pkthdr.len < len) { ipstat.ips_tooshort++; goto bad; } - if (m->m_pkthdr.len > ip->ip_len) { + if (m->m_pkthdr.len > len) { if (m->m_len == m->m_pkthdr.len) { - m->m_len = ip->ip_len; - m->m_pkthdr.len = ip->ip_len; + m->m_len = len; + m->m_pkthdr.len = len; } else - m_adj(m, ip->ip_len - m->m_pkthdr.len); + m_adj(m, len - m->m_pkthdr.len); } /* + * Assume that we can create a fast-forward IP flow entry + * based on this packet. + */ + m->m_flags |= M_CANFASTFWD; + +#ifdef PFIL_HOOKS + /* + * Run through list of hooks for input packets. If there are any + * filters which require that additional packets in the flow are + * not fast-forwarded, they must clear the M_CANFASTFWD flag. + * 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) + goto next; + m = m0; + if (m == NULL) + goto next; + ip = mtod(m, struct ip *); + } +#endif /* PFIL_HOOKS */ + + /* * Process options and, if not destined for us, * ship it on. ip_dooptions returns 1 when an * error was detected (causing an icmp message @@ -238,16 +429,16 @@ next: /* * Check our list of addresses, to see if the packet is for us. */ - for (ia = in_ifaddr; ia; ia = ia->ia_next) { - if (ip->ip_dst.s_addr == ia->ia_addr.sin_addr.s_addr) - goto ours; - if ( -#ifdef DIRECTED_BROADCAST - ia->ia_ifp == m->m_pkthdr.rcvif && -#endif - (ia->ia_ifp->if_flags & IFF_BROADCAST)) { - if (ip->ip_dst.s_addr == ia->ia_broadaddr.sin_addr.s_addr || - ip->ip_dst.s_addr == ia->ia_netbroadcast.s_addr || + INADDR_TO_IA(ip->ip_dst, ia); + 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; + ia = ifatoia(ifa); + if (in_hosteq(ip->ip_dst, ia->ia_broadaddr.sin_addr) || + in_hosteq(ip->ip_dst, ia->ia_netbroadcast) || /* * Look for all-0's host part (old broadcast addr), * either for subnet or net. @@ -255,6 +446,12 @@ next: ip->ip_dst.s_addr == ia->ia_subnet || ip->ip_dst.s_addr == ia->ia_net) goto ours; + /* + * An interface with IP address zero accepts + * all packets that arrive on that interface. + */ + if (in_nullhost(ia->ia_addr.sin_addr)) + goto ours; } } if (IN_MULTICAST(ip->ip_dst.s_addr)) { @@ -283,13 +480,11 @@ next: * as expected when ip_mforward() is called from * ip_output().) */ - ip->ip_id = htons(ip->ip_id); if (ip_mforward(m, m->m_pkthdr.rcvif) != 0) { ipstat.ips_cantforward++; m_freem(m); goto next; } - ip->ip_id = ntohs(ip->ip_id); /* * The process-level routing demon needs to receive @@ -314,7 +509,7 @@ next: goto ours; } if (ip->ip_dst.s_addr == INADDR_BROADCAST || - ip->ip_dst.s_addr == INADDR_ANY) + in_nullhost(ip->ip_dst)) goto ours; /* @@ -335,22 +530,16 @@ ours: * if the packet was previously fragmented, * but it's not worth the time; just let them time out.) */ - if (ip->ip_off &~ IP_DF) { - if (m->m_flags & M_EXT) { /* XXX */ - if ((m = m_pullup(m, sizeof (struct ip))) == 0) { - ipstat.ips_toosmall++; - goto next; - } - ip = mtod(m, struct ip *); - } + if (ip->ip_off & ~(IP_DF|IP_RF)) { /* * Look for queue of fragments * of this datagram. */ - for (fp = ipq.next; fp != &ipq; fp = fp->next) + IPQ_LOCK(); + for (fp = ipq.lh_first; fp != NULL; fp = fp->ipq_q.le_next) if (ip->ip_id == fp->ipq_id && - ip->ip_src.s_addr == fp->ipq_src.s_addr && - ip->ip_dst.s_addr == fp->ipq_dst.s_addr && + in_hosteq(ip->ip_src, fp->ipq_src) && + in_hosteq(ip->ip_dst, fp->ipq_dst) && ip->ip_p == fp->ipq_p) goto found; fp = 0; @@ -358,21 +547,21 @@ found: /* * Adjust ip_len to not reflect header, - * set ip_mff if more fragments are expected, + * set ipqe_mff if more fragments are expected, * convert offset of this to bytes. */ ip->ip_len -= hlen; - ((struct ipasfrag *)ip)->ipf_mff &= ~1; - if (ip->ip_off & IP_MF) { + mff = (ip->ip_off & IP_MF) != 0; + if (mff) { /* * Make sure that fragments have a data length * that's a non-zero multiple of 8 bytes. */ if (ip->ip_len == 0 || (ip->ip_len & 0x7) != 0) { ipstat.ips_badfrags++; + IPQ_UNLOCK(); goto bad; } - ((struct ipasfrag *)ip)->ipf_mff |= 1; } ip->ip_off <<= 3; @@ -381,22 +570,38 @@ found: * or if this is not the first fragment, * attempt reassembly; if it succeeds, proceed. */ - if (((struct ipasfrag *)ip)->ipf_mff & 1 || ip->ip_off) { + if (mff || ip->ip_off) { ipstat.ips_fragments++; - ip = ip_reass((struct ipasfrag *)ip, fp); - if (ip == 0) + ipqe = pool_get(&ipqent_pool, PR_NOWAIT); + if (ipqe == NULL) { + ipstat.ips_rcvmemdrop++; + IPQ_UNLOCK(); + goto bad; + } + ipqe->ipqe_mff = mff; + ipqe->ipqe_m = m; + ipqe->ipqe_ip = ip; + m = ip_reass(ipqe, fp); + if (m == 0) { + IPQ_UNLOCK(); goto next; + } ipstat.ips_reassembled++; - m = dtom(ip); + ip = mtod(m, struct ip *); + hlen = ip->ip_hl << 2; + ip->ip_len += hlen; } else if (fp) ip_freef(fp); - } else - ip->ip_len -= hlen; + IPQ_UNLOCK(); + } /* * Switch out to protocol's input routine. */ +#if IFA_STATS + ia->ia_ifa.ifa_data.ifad_inbytes += ip->ip_len; +#endif ipstat.ips_delivered++; (*inetsw[ip_protox[ip->ip_p]].pr_input)(m, hlen); goto next; @@ -411,17 +616,20 @@ bad: * reassembly of this datagram already exists, then it * is given as fp; otherwise have to make a chain. */ -struct ip * -ip_reass(ip, fp) - register struct ipasfrag *ip; +struct mbuf * +ip_reass(ipqe, fp) + register struct ipqent *ipqe; register struct ipq *fp; { - register struct mbuf *m = dtom(ip); - register struct ipasfrag *q; + register struct mbuf *m = ipqe->ipqe_m; + register struct ipqent *nq, *p, *q; + struct ip *ip; struct mbuf *t; - int hlen = ip->ip_hl << 2; + int hlen = ipqe->ipqe_ip->ip_hl << 2; int i, next; + IPQ_LOCK_CHECK(); + /* * Presence of header sizes in mbufs * would confuse code below. @@ -433,25 +641,27 @@ ip_reass(ip, fp) * If first fragment to arrive, create a reassembly queue. */ if (fp == 0) { - if ((t = m_get(M_DONTWAIT, MT_FTABLE)) == NULL) + MALLOC(fp, struct ipq *, sizeof (struct ipq), + M_FTABLE, M_NOWAIT); + if (fp == NULL) goto dropfrag; - fp = mtod(t, struct ipq *); - insque(fp, &ipq); + LIST_INSERT_HEAD(&ipq, fp, ipq_q); fp->ipq_ttl = IPFRAGTTL; - fp->ipq_p = ip->ip_p; - fp->ipq_id = ip->ip_id; - fp->ipq_next = fp->ipq_prev = (struct ipasfrag *)fp; - fp->ipq_src = ((struct ip *)ip)->ip_src; - fp->ipq_dst = ((struct ip *)ip)->ip_dst; - q = (struct ipasfrag *)fp; + fp->ipq_p = ipqe->ipqe_ip->ip_p; + fp->ipq_id = ipqe->ipqe_ip->ip_id; + LIST_INIT(&fp->ipq_fragq); + fp->ipq_src = ipqe->ipqe_ip->ip_src; + fp->ipq_dst = ipqe->ipqe_ip->ip_dst; + p = NULL; goto insert; } /* * Find a segment which begins after this one does. */ - for (q = fp->ipq_next; q != (struct ipasfrag *)fp; q = q->ipf_next) - if (q->ip_off > ip->ip_off) + for (p = NULL, q = fp->ipq_fragq.lh_first; q != NULL; + p = q, q = q->ipqe_q.le_next) + if (q->ipqe_ip->ip_off > ipqe->ipqe_ip->ip_off) break; /* @@ -459,14 +669,15 @@ ip_reass(ip, fp) * our data already. If so, drop the data from the incoming * segment. If it provides all of our data, drop us. */ - if (q->ipf_prev != (struct ipasfrag *)fp) { - i = q->ipf_prev->ip_off + q->ipf_prev->ip_len - ip->ip_off; + if (p != NULL) { + i = p->ipqe_ip->ip_off + p->ipqe_ip->ip_len - + ipqe->ipqe_ip->ip_off; if (i > 0) { - if (i >= ip->ip_len) + if (i >= ipqe->ipqe_ip->ip_len) goto dropfrag; - m_adj(dtom(ip), i); - ip->ip_off += i; - ip->ip_len -= i; + m_adj(ipqe->ipqe_m, i); + ipqe->ipqe_ip->ip_off += i; + ipqe->ipqe_ip->ip_len -= i; } } @@ -474,17 +685,20 @@ ip_reass(ip, fp) * While we overlap succeeding segments trim them or, * if they are completely covered, dequeue them. */ - while (q != (struct ipasfrag *)fp && ip->ip_off + ip->ip_len > q->ip_off) { - i = (ip->ip_off + ip->ip_len) - q->ip_off; - if (i < q->ip_len) { - q->ip_len -= i; - q->ip_off += i; - m_adj(dtom(q), i); + for (; q != NULL && ipqe->ipqe_ip->ip_off + ipqe->ipqe_ip->ip_len > + q->ipqe_ip->ip_off; q = nq) { + i = (ipqe->ipqe_ip->ip_off + ipqe->ipqe_ip->ip_len) - + q->ipqe_ip->ip_off; + if (i < q->ipqe_ip->ip_len) { + q->ipqe_ip->ip_len -= i; + q->ipqe_ip->ip_off += i; + m_adj(q->ipqe_m, i); break; } - q = q->ipf_next; - m_freem(dtom(q->ipf_prev)); - ip_deq(q->ipf_prev); + nq = q->ipqe_q.le_next; + m_freem(q->ipqe_m); + LIST_REMOVE(q, ipqe_q); + pool_put(&ipqent_pool, q); } insert: @@ -492,28 +706,42 @@ insert: * Stick new segment in its place; * check for complete reassembly. */ - ip_enq(ip, q->ipf_prev); + if (p == NULL) { + LIST_INSERT_HEAD(&fp->ipq_fragq, ipqe, ipqe_q); + } else { + LIST_INSERT_AFTER(p, ipqe, ipqe_q); + } next = 0; - for (q = fp->ipq_next; q != (struct ipasfrag *)fp; q = q->ipf_next) { - if (q->ip_off != next) + for (p = NULL, q = fp->ipq_fragq.lh_first; q != NULL; + p = q, q = q->ipqe_q.le_next) { + if (q->ipqe_ip->ip_off != next) return (0); - next += q->ip_len; + next += q->ipqe_ip->ip_len; } - if (q->ipf_prev->ipf_mff & 1) + if (p->ipqe_mff) return (0); /* - * Reassembly is complete; concatenate fragments. + * Reassembly is complete. Check for a bogus message size and + * concatenate fragments. */ - q = fp->ipq_next; - m = dtom(q); + q = fp->ipq_fragq.lh_first; + ip = q->ipqe_ip; + if ((next + (ip->ip_hl << 2)) > IP_MAXPACKET) { + ipstat.ips_toolong++; + ip_freef(fp); + return (0); + } + m = q->ipqe_m; t = m->m_next; m->m_next = 0; m_cat(m, t); - q = q->ipf_next; - while (q != (struct ipasfrag *)fp) { - t = dtom(q); - q = q->ipf_next; + nq = q->ipqe_q.le_next; + pool_put(&ipqent_pool, q); + for (q = nq; q != NULL; q = nq) { + t = q->ipqe_m; + nq = q->ipqe_q.le_next; + pool_put(&ipqent_pool, q); m_cat(m, t); } @@ -523,28 +751,26 @@ insert: * dequeue and discard fragment reassembly header. * Make header visible. */ - ip = fp->ipq_next; ip->ip_len = next; - ip->ipf_mff &= ~1; - ((struct ip *)ip)->ip_src = fp->ipq_src; - ((struct ip *)ip)->ip_dst = fp->ipq_dst; - remque(fp); - (void) m_free(dtom(fp)); - m = dtom(ip); + ip->ip_src = fp->ipq_src; + ip->ip_dst = fp->ipq_dst; + LIST_REMOVE(fp, ipq_q); + FREE(fp, M_FTABLE); 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; - for (t = m; m; m = m->m_next) - plen += m->m_len; - t->m_pkthdr.len = plen; + for (t = m; t; t = t->m_next) + plen += t->m_len; + m->m_pkthdr.len = plen; } - return ((struct ip *)ip); + return (m); dropfrag: ipstat.ips_fragdropped++; m_freem(m); + pool_put(&ipqent_pool, ipqe); return (0); } @@ -556,42 +782,18 @@ void ip_freef(fp) struct ipq *fp; { - register struct ipasfrag *q, *p; - - for (q = fp->ipq_next; q != (struct ipasfrag *)fp; q = p) { - p = q->ipf_next; - ip_deq(q); - m_freem(dtom(q)); - } - remque(fp); - (void) m_free(dtom(fp)); -} - -/* - * Put an ip fragment on a reassembly chain. - * Like insque, but pointers in middle of structure. - */ -void -ip_enq(p, prev) - register struct ipasfrag *p, *prev; -{ + register struct ipqent *q, *p; - p->ipf_prev = prev; - p->ipf_next = prev->ipf_next; - prev->ipf_next->ipf_prev = p; - prev->ipf_next = p; -} + IPQ_LOCK_CHECK(); -/* - * To ip_enq as remque is to insque. - */ -void -ip_deq(p) - register struct ipasfrag *p; -{ - - p->ipf_prev->ipf_next = p->ipf_next; - p->ipf_next->ipf_prev = p->ipf_prev; + for (q = fp->ipq_fragq.lh_first; q != NULL; q = p) { + p = q->ipqe_q.le_next; + m_freem(q->ipqe_m); + LIST_REMOVE(q, ipqe_q); + pool_put(&ipqent_pool, q); + } + LIST_REMOVE(fp, ipq_q); + FREE(fp, M_FTABLE); } /* @@ -602,22 +804,21 @@ ip_deq(p) void ip_slowtimo() { - register struct ipq *fp; - int s = splnet(); + register struct ipq *fp, *nfp; + int s = splsoftnet(); - fp = ipq.next; - if (fp == 0) { - splx(s); - return; - } - while (fp != &ipq) { - --fp->ipq_ttl; - fp = fp->next; - if (fp->prev->ipq_ttl == 0) { + IPQ_LOCK(); + for (fp = ipq.lh_first; fp != NULL; fp = nfp) { + nfp = fp->ipq_q.le_next; + if (--fp->ipq_ttl == 0) { ipstat.ips_fragtimeout++; - ip_freef(fp->prev); + ip_freef(fp); } } + IPQ_UNLOCK(); +#ifdef GATEWAY + ipflow_slowtimo(); +#endif splx(s); } @@ -628,10 +829,19 @@ void ip_drain() { - while (ipq.next != &ipq) { + /* + * We may be called from a device's interrupt context. If + * the ipq is already busy, just bail out now. + */ + if (ipq_lock_try() == 0) + return; + + while (ipq.lh_first != NULL) { ipstat.ips_fragdropped++; - ip_freef(ipq.next); + ip_freef(ipq.lh_first); } + + IPQ_UNLOCK(); } /* @@ -646,11 +856,11 @@ ip_dooptions(m) struct mbuf *m; { register struct ip *ip = mtod(m, struct ip *); - register u_char *cp; + register u_char *cp, *cp0; register struct ip_timestamp *ipt; register 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; @@ -685,6 +895,11 @@ ip_dooptions(m) */ case IPOPT_LSRR: case IPOPT_SSRR: + if (ip_allowsrcrt == 0) { + type = ICMP_UNREACH; + code = ICMP_UNREACH_NET_PROHIB; + goto bad; + } if ((off = cp[IPOPT_OFFSET]) < IPOPT_MINOFF) { code = &cp[IPOPT_OFFSET] - (u_char *)ip; goto bad; @@ -719,8 +934,7 @@ ip_dooptions(m) if (opt == IPOPT_SSRR) { #define INA struct in_ifaddr * #define SA struct sockaddr * - if ((ia = (INA)ifa_ifwithdstaddr((SA)&ipaddr)) == 0) - ia = (INA)ifa_ifwithnet((SA)&ipaddr); + ia = (INA)ifa_ifwithladdr((SA)&ipaddr); } else ia = ip_rtaddr(ipaddr.sin_addr); if (ia == 0) { @@ -776,14 +990,14 @@ ip_dooptions(m) 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: break; case IPOPT_TS_TSANDADDR: - if (ipt->ipt_ptr + sizeof(n_time) + + if (ipt->ipt_ptr - 1 + sizeof(n_time) + sizeof(struct in_addr) > ipt->ipt_len) goto bad; ipaddr.sin_addr = dst; @@ -791,16 +1005,16 @@ ip_dooptions(m) 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 + sizeof(n_time) + + if (ipt->ipt_ptr - 1 + sizeof(n_time) + sizeof(struct in_addr) > ipt->ipt_len) 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) continue; @@ -811,18 +1025,23 @@ ip_dooptions(m) goto bad; } ntime = iptime(); - bcopy((caddr_t)&ntime, (caddr_t)cp + ipt->ipt_ptr - 1, + cp0 = (u_char *) &ntime; /* XXX GCC BUG */ + bcopy(cp0, (caddr_t)cp + ipt->ipt_ptr - 1, sizeof(n_time)); ipt->ipt_ptr += sizeof(n_time); } } if (forward) { + if (ip_forwsrcrt == 0) { + type = ICMP_UNREACH; + code = ICMP_UNREACH_SRCFAIL; + goto bad; + } ip_forward(m, 1); return (1); } return (0); bad: - ip->ip_len -= ip->ip_hl << 2; /* XXX icmp_error adds in hdr length */ icmp_error(m, type, code, 0, 0); ipstat.ips_badoptions++; return (1); @@ -840,7 +1059,7 @@ ip_rtaddr(dst) sin = satosin(&ipforward_rt.ro_dst); - if (ipforward_rt.ro_rt == 0 || dst.s_addr != sin->sin_addr.s_addr) { + if (ipforward_rt.ro_rt == 0 || !in_hosteq(dst, sin->sin_addr)) { if (ipforward_rt.ro_rt) { RTFREE(ipforward_rt.ro_rt); ipforward_rt.ro_rt = 0; @@ -913,7 +1132,7 @@ ip_srcroute() *(mtod(m, struct in_addr *)) = *p--; #ifdef DIAGNOSTIC if (ipprintfs) - printf(" hops %lx", ntohl(mtod(m, struct in_addr *)->s_addr)); + printf(" hops %x", ntohl(mtod(m, struct in_addr *)->s_addr)); #endif /* @@ -933,7 +1152,7 @@ ip_srcroute() while (p >= ip_srcrt.route) { #ifdef DIAGNOSTIC if (ipprintfs) - printf(" %lx", ntohl(q->s_addr)); + printf(" %x", ntohl(q->s_addr)); #endif *q++ = *p--; } @@ -943,7 +1162,7 @@ ip_srcroute() *q = ip_srcrt.dst; #ifdef DIAGNOSTIC if (ipprintfs) - printf(" %lx\n", ntohl(q->s_addr)); + printf(" %x\n", ntohl(q->s_addr)); #endif return (m); } @@ -965,17 +1184,18 @@ ip_stripoptions(m, mopt) register caddr_t opts; int olen; - olen = (ip->ip_hl<<2) - sizeof (struct ip); + olen = (ip->ip_hl << 2) - sizeof (struct ip); opts = (caddr_t)(ip + 1); i = m->m_len - (sizeof (struct ip) + olen); bcopy(opts + olen, opts, (unsigned)i); m->m_len -= olen; if (m->m_flags & M_PKTHDR) m->m_pkthdr.len -= olen; - ip->ip_hl = sizeof(struct ip) >> 2; + ip->ip_len -= olen; + ip->ip_hl = sizeof (struct ip) >> 2; } -u_char inetctlerrmap[PRC_NCMDS] = { +int inetctlerrmap[PRC_NCMDS] = { 0, 0, 0, 0, 0, EMSGSIZE, EHOSTDOWN, EHOSTUNREACH, EHOSTUNREACH, EHOSTUNREACH, ECONNREFUSED, ECONNREFUSED, @@ -1006,7 +1226,7 @@ ip_forward(m, srcrt) register struct ip *ip = mtod(m, struct ip *); register struct sockaddr_in *sin; register struct rtentry *rt; - int error, type = 0, code; + int error, type = 0, code = 0; struct mbuf *mcopy; n_long dest; struct ifnet *destifp; @@ -1014,15 +1234,15 @@ ip_forward(m, srcrt) dest = 0; #ifdef DIAGNOSTIC if (ipprintfs) - printf("forward: src %x dst %x ttl %x\n", ip->ip_src, - ip->ip_dst, ip->ip_ttl); + printf("forward: src %2.2x dst %2.2x ttl %x\n", + ntohl(ip->ip_src.s_addr), + ntohl(ip->ip_dst.s_addr), ip->ip_ttl); #endif - if (m->m_flags & M_BCAST || in_canforward(ip->ip_dst) == 0) { + if (m->m_flags & (M_BCAST|M_MCAST) || in_canforward(ip->ip_dst) == 0) { ipstat.ips_cantforward++; m_freem(m); return; } - HTONS(ip->ip_id); if (ip->ip_ttl <= IPTTLDEC) { icmp_error(m, ICMP_TIMXCEED, ICMP_TIMXCEED_INTRANS, dest, 0); return; @@ -1031,13 +1251,13 @@ ip_forward(m, srcrt) sin = satosin(&ipforward_rt.ro_dst); if ((rt = ipforward_rt.ro_rt) == 0 || - ip->ip_dst.s_addr != sin->sin_addr.s_addr) { + !in_hosteq(ip->ip_dst, sin->sin_addr)) { if (ipforward_rt.ro_rt) { RTFREE(ipforward_rt.ro_rt); ipforward_rt.ro_rt = 0; } sin->sin_family = AF_INET; - sin->sin_len = sizeof(*sin); + sin->sin_len = sizeof(struct sockaddr_in); sin->sin_addr = ip->ip_dst; rtalloc(&ipforward_rt); @@ -1049,15 +1269,11 @@ ip_forward(m, srcrt) } /* - * Save at most 64 bytes of the packet in case + * Save at most 68 bytes of the packet in case * we need to generate an ICMP message to the src. */ - mcopy = m_copy(m, 0, imin((int)ip->ip_len, 64)); + mcopy = m_copy(m, 0, imin((int)ip->ip_len, 68)); -#ifdef GATEWAY - ip_ifmatrix[rt->rt_ifp->if_index + - if_index * m->m_pkthdr.rcvif->if_index]++; -#endif /* * If forwarding packet using same interface that it came in on, * perhaps should send a redirect to sender to shortcut a hop. @@ -1068,30 +1284,31 @@ ip_forward(m, srcrt) */ if (rt->rt_ifp == m->m_pkthdr.rcvif && (rt->rt_flags & (RTF_DYNAMIC|RTF_MODIFIED)) == 0 && - satosin(rt_key(rt))->sin_addr.s_addr != 0 && + !in_nullhost(satosin(rt_key(rt))->sin_addr) && ipsendredirects && !srcrt) { if (rt->rt_ifa && (ip->ip_src.s_addr & ifatoia(rt->rt_ifa)->ia_subnetmask) == ifatoia(rt->rt_ifa)->ia_subnet) { - if (rt->rt_flags & RTF_GATEWAY) - dest = satosin(rt->rt_gateway)->sin_addr.s_addr; - else - dest = ip->ip_dst.s_addr; - /* Router requirements says to only send host redirects */ - type = ICMP_REDIRECT; - code = ICMP_REDIRECT_HOST; + if (rt->rt_flags & RTF_GATEWAY) + dest = satosin(rt->rt_gateway)->sin_addr.s_addr; + else + dest = ip->ip_dst.s_addr; + /* + * Router requirements says to only send host + * redirects. + */ + type = ICMP_REDIRECT; + code = ICMP_REDIRECT_HOST; #ifdef DIAGNOSTIC - if (ipprintfs) - printf("redirect (%d) to %lx\n", code, (u_int32_t)dest); + if (ipprintfs) + printf("redirect (%d) to %x\n", code, + (u_int32_t)dest); #endif } } - error = ip_output(m, (struct mbuf *)0, &ipforward_rt, IP_FORWARDING -#ifdef DIRECTED_BROADCAST - | IP_ALLOWBROADCAST -#endif - , 0); + error = ip_output(m, (struct mbuf *)0, &ipforward_rt, + (IP_FORWARDING | (ip_directedbcast ? IP_ALLOWBROADCAST : 0)), 0); if (error) ipstat.ips_cantforward++; else { @@ -1099,8 +1316,13 @@ ip_forward(m, srcrt) if (type) ipstat.ips_redirectsent++; else { - if (mcopy) + if (mcopy) { +#ifdef GATEWAY + if (mcopy->m_flags & M_CANFASTFWD) + ipflow_create(&ipforward_rt, mcopy); +#endif m_freem(mcopy); + } return; } } @@ -1139,6 +1361,66 @@ ip_forward(m, srcrt) icmp_error(mcopy, type, code, dest, destifp); } +void +ip_savecontrol(inp, mp, ip, m) + register struct inpcb *inp; + register struct mbuf **mp; + register struct ip *ip; + register struct mbuf *m; +{ + + if (inp->inp_socket->so_options & SO_TIMESTAMP) { + struct timeval tv; + + microtime(&tv); + *mp = sbcreatecontrol((caddr_t) &tv, sizeof(tv), + SCM_TIMESTAMP, SOL_SOCKET); + if (*mp) + mp = &(*mp)->m_next; + } + if (inp->inp_flags & INP_RECVDSTADDR) { + *mp = sbcreatecontrol((caddr_t) &ip->ip_dst, + sizeof(struct in_addr), IP_RECVDSTADDR, IPPROTO_IP); + if (*mp) + mp = &(*mp)->m_next; + } +#ifdef notyet + /* + * XXX + * Moving these out of udp_input() made them even more broken + * than they already were. + * - fenner@parc.xerox.com + */ + /* options were tossed already */ + if (inp->inp_flags & INP_RECVOPTS) { + *mp = sbcreatecontrol((caddr_t) opts_deleted_above, + sizeof(struct in_addr), IP_RECVOPTS, IPPROTO_IP); + if (*mp) + mp = &(*mp)->m_next; + } + /* ip_srcroute doesn't do what we want here, need to fix */ + if (inp->inp_flags & INP_RECVRETOPTS) { + *mp = sbcreatecontrol((caddr_t) ip_srcroute(), + sizeof(struct in_addr), IP_RECVRETOPTS, IPPROTO_IP); + if (*mp) + mp = &(*mp)->m_next; + } +#endif + if (inp->inp_flags & INP_RECVIF) { + struct sockaddr_dl sdl; + + sdl.sdl_len = offsetof(struct sockaddr_dl, sdl_data[0]); + sdl.sdl_family = AF_LINK; + sdl.sdl_index = m->m_pkthdr.rcvif ? + m->m_pkthdr.rcvif->if_index : 0; + sdl.sdl_nlen = sdl.sdl_alen = sdl.sdl_slen = 0; + *mp = sbcreatecontrol((caddr_t) &sdl, sdl.sdl_len, + IP_RECVIF, IPPROTO_IP); + if (*mp) + mp = &(*mp)->m_next; + } +} + int ip_sysctl(name, namelen, oldp, oldlenp, newp, newlen) int *name; @@ -1148,6 +1430,10 @@ ip_sysctl(name, namelen, oldp, oldlenp, void *newp; size_t newlen; { + extern int subnetsarelocal; + + int error, old; + /* All sysctl names at this level are terminal. */ if (namelen != 1) return (ENOTDIR); @@ -1164,6 +1450,78 @@ ip_sysctl(name, namelen, oldp, oldlenp, case IPCTL_DEFMTU: return (sysctl_int(oldp, oldlenp, newp, newlen, &ip_mtu)); #endif + case IPCTL_FORWSRCRT: + /* Don't allow this to change in a secure environment. */ + if (securelevel > 0) + return (sysctl_rdint(oldp, oldlenp, newp, + ip_forwsrcrt)); + else + return (sysctl_int(oldp, oldlenp, newp, newlen, + &ip_forwsrcrt)); + case IPCTL_DIRECTEDBCAST: + return (sysctl_int(oldp, oldlenp, newp, newlen, + &ip_directedbcast)); + case IPCTL_ALLOWSRCRT: + return (sysctl_int(oldp, oldlenp, newp, newlen, + &ip_allowsrcrt)); + case IPCTL_SUBNETSARELOCAL: + return (sysctl_int(oldp, oldlenp, newp, newlen, + &subnetsarelocal)); + case IPCTL_MTUDISC: + error = sysctl_int(oldp, oldlenp, newp, newlen, + &ip_mtudisc); + if (ip_mtudisc != 0 && ip_mtudisc_timeout_q == NULL) { + ip_mtudisc_timeout_q = + rt_timer_queue_create(ip_mtudisc_timeout); + } else if (ip_mtudisc == 0 && ip_mtudisc_timeout_q != NULL) { + rt_timer_queue_destroy(ip_mtudisc_timeout_q, TRUE); + ip_mtudisc_timeout_q = NULL; + } + return error; + case IPCTL_ANONPORTMIN: + old = anonportmin; + error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmin); + if (anonportmin >= anonportmax || anonportmin > 65535 +#ifndef IPNOPRIVPORTS + || anonportmin < IPPORT_RESERVED +#endif + ) { + anonportmin = old; + return (EINVAL); + } + return (error); + case IPCTL_ANONPORTMAX: + old = anonportmax; + error = sysctl_int(oldp, oldlenp, newp, newlen, &anonportmax); + if (anonportmin >= anonportmax || anonportmax > 65535 +#ifndef IPNOPRIVPORTS + || anonportmax < IPPORT_RESERVED +#endif + ) { + anonportmax = old; + return (EINVAL); + } + return (error); + case IPCTL_MTUDISCTIMEOUT: + error = sysctl_int(oldp, oldlenp, newp, newlen, + &ip_mtudisc_timeout); + if (ip_mtudisc_timeout_q != NULL) + rt_timer_queue_change(ip_mtudisc_timeout_q, + ip_mtudisc_timeout); + return (error); +#ifdef GATEWAY + case IPCTL_MAXFLOWS: + { + int s; + + error = sysctl_int(oldp, oldlenp, newp, newlen, + &ip_maxflows); + s = splsoftnet(); + ipflow_reap(0); + splx(s); + return (error); + } +#endif default: return (EOPNOTSUPP); }