/* $NetBSD: ip_nat.c,v 1.27 1999/12/28 07:14:53 darrenr Exp $ */ /* * Copyright (C) 1995-1998 by Darren Reed. * * Redistribution and use in source and binary forms are permitted * provided that this notice is preserved and due credit is given * to the original author and the contributors. * * Added redirect stuff and a LOT of bug fixes. (mcn@EnGarde.com) */ #if !defined(lint) #if defined(__NetBSD__) static const char rcsid[] = "$NetBSD: ip_nat.c,v 1.27 1999/12/28 07:14:53 darrenr Exp $"; #else static const char sccsid[] = "@(#)ip_nat.c 1.11 6/5/96 (C) 1995 Darren Reed"; static const char rcsid[] = "@(#)Id: ip_nat.c,v 2.2.2.10 1999/12/07 12:53:42 darrenr Exp"; #endif #endif #if defined(__FreeBSD__) && defined(KERNEL) && !defined(_KERNEL) #define _KERNEL #endif #include #include #include #include #include #if !defined(_KERNEL) && !defined(KERNEL) # include # include # include #endif #if defined(KERNEL) && (__FreeBSD_version >= 220000) # include # include #else # include #endif #include #include #ifndef linux # include #endif #include #if defined(_KERNEL) && !defined(linux) # include #endif #if !defined(__SVR4) && !defined(__svr4__) # ifndef linux # include # endif #else # include # include # ifdef _KERNEL # include # endif # include # include #endif #if __FreeBSD_version >= 300000 # include #endif #include #if __FreeBSD_version >= 300000 # include # if defined(_KERNEL) && !defined(IPFILTER_LKM) # include "opt_ipfilter.h" # endif #endif #ifdef sun # include #endif #include #include #include #include #ifdef __sgi # ifdef IFF_DRVRLOCK /* IRIX6 */ #include #include # endif #endif #ifdef RFC1825 # include # include extern struct ifnet vpnif; #endif #ifndef linux # include #endif #include #include #include #include "netinet/ip_compat.h" #include #include "netinet/ip_fil.h" #include "netinet/ip_proxy.h" #include "netinet/ip_nat.h" #include "netinet/ip_frag.h" #include "netinet/ip_state.h" #if (__FreeBSD_version >= 300000) # include #endif #ifndef MIN # define MIN(a,b) (((a)<(b))?(a):(b)) #endif #undef SOCKADDR_IN #define SOCKADDR_IN struct sockaddr_in nat_t **nat_table[2] = { NULL, NULL }, *nat_instances = NULL; ipnat_t *nat_list = NULL; u_int ipf_nattable_sz = NAT_TABLE_SZ; u_int ipf_natrules_sz = NAT_SIZE; u_int ipf_rdrrules_sz = RDR_SIZE; u_32_t nat_masks = 0; u_32_t rdr_masks = 0; ipnat_t **nat_rules = NULL; ipnat_t **rdr_rules = NULL; u_long fr_defnatage = DEF_NAT_AGE, fr_defnaticmpage = 6; /* 3 seconds */ natstat_t nat_stats; #if (SOLARIS || defined(__sgi)) && defined(_KERNEL) extern kmutex_t ipf_rw; extern KRWLOCK_T ipf_nat; #endif static int nat_flushtable __P((void)); static int nat_clearlist __P((void)); static void nat_delete __P((struct nat *)); static void nat_delrdr __P((struct ipnat *)); static void nat_delnat __P((struct ipnat *)); int nat_init() { KMALLOCS(nat_table[0], nat_t **, sizeof(nat_t *) * ipf_nattable_sz); if (nat_table[0] != NULL) bzero((char *)nat_table[0], ipf_nattable_sz * sizeof(nat_t *)); else return -1; KMALLOCS(nat_table[1], nat_t **, sizeof(nat_t *) * ipf_nattable_sz); if (nat_table[1] != NULL) bzero((char *)nat_table[1], ipf_nattable_sz * sizeof(nat_t *)); else return -1; KMALLOCS(nat_rules, ipnat_t **, sizeof(ipnat_t *) * ipf_natrules_sz); if (nat_rules != NULL) bzero((char *)nat_rules, ipf_natrules_sz * sizeof(ipnat_t *)); else return -1; KMALLOCS(rdr_rules, ipnat_t **, sizeof(ipnat_t *) * ipf_rdrrules_sz); if (rdr_rules != NULL) bzero((char *)rdr_rules, ipf_rdrrules_sz * sizeof(ipnat_t *)); else return -1; return 0; } void nat_delrdr(n) ipnat_t *n; { ipnat_t **n1; u_32_t iph; u_int hv; iph = n->in_outip & n->in_outmsk; hv = NAT_HASH_FN(iph, ipf_rdrrules_sz); for (n1 = &rdr_rules[hv]; *n1 && (*n1 != n); n1 = &(*n1)->in_rnext) ; if (*n1) *n1 = n->in_rnext; } static void nat_delnat(n) ipnat_t *n; { ipnat_t **n1; u_32_t iph; u_int hv; iph = n->in_inip & n->in_inmsk; hv = NAT_HASH_FN(iph, ipf_natrules_sz); for (n1 = &nat_rules[hv]; *n1 && (*n1 != n); n1 = &(*n1)->in_mnext) ; if (*n1) *n1 = n->in_mnext; } void fix_outcksum(sp, n , len) u_short *sp; u_32_t n; int len; { register u_short sumshort; register u_32_t sum1; if (!n) return; #if SOLARIS2 >= 6 else if (n & NAT_HW_CKSUM) { *sp = n & 0xffff; return; } #endif sum1 = (~ntohs(*sp)) & 0xffff; sum1 += (n); sum1 = (sum1 >> 16) + (sum1 & 0xffff); /* Again */ sum1 = (sum1 >> 16) + (sum1 & 0xffff); sumshort = ~(u_short)sum1; *(sp) = htons(sumshort); } void fix_incksum(sp, n , len) u_short *sp; u_32_t n; int len; { register u_short sumshort; register u_32_t sum1; if (!n) return; #if SOLARIS2 >= 6 else if (n & NAT_HW_CKSUM) { *sp = n & 0xffff; return; } #endif #ifdef sparc sum1 = (~(*sp)) & 0xffff; #else sum1 = (~ntohs(*sp)) & 0xffff; #endif sum1 += ~(n) & 0xffff; sum1 = (sum1 >> 16) + (sum1 & 0xffff); /* Again */ sum1 = (sum1 >> 16) + (sum1 & 0xffff); sumshort = ~(u_short)sum1; *(sp) = htons(sumshort); } /* * How the NAT is organised and works. * * Inside (interface y) NAT Outside (interface x) * -------------------- -+- ------------------------------------- * Packet going | out, processsed by ip_natout() for x * ------------> | ------------> * src=10.1.1.1 | src=192.1.1.1 * | * | in, processed by ip_natin() for x * <------------ | <------------ * dst=10.1.1.1 | dst=192.1.1.1 * -------------------- -+- ------------------------------------- * ip_natout() - changes ip_src and if required, sport * - creates a new mapping, if required. * ip_natin() - changes ip_dst and if required, dport * * In the NAT table, internal source is recorded as "in" and externally * seen as "out". */ /* * Handle ioctls which manipulate the NAT. */ int nat_ioctl(data, cmd, mode) #if defined(__NetBSD__) || defined(__OpenBSD__) u_long cmd; #else int cmd; #endif caddr_t data; int mode; { register ipnat_t *nat, *nt, *n = NULL, **np = NULL; int error = 0, ret, k; ipnat_t natd; u_32_t i, j; #if defined(_KERNEL) && !SOLARIS int s; #endif #if (BSD >= 199306) && defined(_KERNEL) if ((securelevel >= 2) && (mode & FWRITE)) return EPERM; #endif nat = NULL; /* XXX gcc -Wuninitialized */ KMALLOC(nt, ipnat_t *); if ((cmd == SIOCADNAT) || (cmd == SIOCRMNAT)) IRCOPY(data, (char *)&natd, sizeof(natd)); /* * For add/delete, look to see if the NAT entry is already present */ SPL_NET(s); WRITE_ENTER(&ipf_nat); if ((cmd == SIOCADNAT) || (cmd == SIOCRMNAT)) { nat = &natd; nat->in_flags &= IPN_USERFLAGS; if ((nat->in_redir & NAT_MAPBLK) == 0) { nat->in_inip &= nat->in_inmsk; if ((nat->in_flags & IPN_RANGE) == 0) nat->in_outip &= nat->in_outmsk; } for (np = &nat_list; (n = *np); np = &n->in_next) if (!bcmp((char *)&nat->in_flags, (char *)&n->in_flags, IPN_CMPSIZ)) break; } switch (cmd) { #ifdef IPFILTER_LOG case SIOCIPFFB : if (!(mode & FWRITE)) error = EPERM; else *(int *)data = ipflog_clear(IPL_LOGNAT); break; #endif case SIOCADNAT : if (!(mode & FWRITE)) { error = EPERM; break; } if (n) { error = EEXIST; break; } if (nt == NULL) { error = ENOMEM; break; } n = nt; nt = NULL; bcopy((char *)nat, (char *)n, sizeof(*n)); n->in_ifp = (void *)GETUNIT(n->in_ifname); if (!n->in_ifp) n->in_ifp = (void *)-1; if (n->in_plabel[0] != '\0') { n->in_apr = appr_match(n->in_p, n->in_plabel); if (!n->in_apr) { error = ENOENT; break; } } n->in_next = NULL; *np = n; if (n->in_redir & NAT_REDIRECT) { u_int hv; k = countbits(n->in_outmsk); if ((k >= 0) && (k != 32)) rdr_masks |= 1 << k; j = (n->in_outip & n->in_outmsk); hv = NAT_HASH_FN(j, ipf_rdrrules_sz); np = rdr_rules + hv; while (*np != NULL) np = &(*np)->in_rnext; n->in_rnext = NULL; *np = n; } if (n->in_redir & (NAT_MAP|NAT_MAPBLK)) { u_int hv; k = countbits(n->in_inmsk); if ((k >= 0) && (k != 32)) nat_masks |= 1 << k; j = (n->in_inip & n->in_inmsk); hv = NAT_HASH_FN(j, ipf_natrules_sz); np = nat_rules + hv; while (*np != NULL) np = &(*np)->in_mnext; n->in_mnext = NULL; *np = n; } n->in_use = 0; if (n->in_redir & NAT_MAPBLK) n->in_space = USABLE_PORTS * ~ntohl(n->in_outmsk); else if (n->in_flags & IPN_AUTOPORTMAP) n->in_space = USABLE_PORTS * ~ntohl(n->in_inmsk); else if (n->in_flags & IPN_RANGE) n->in_space = ntohl(n->in_outmsk) - ntohl(n->in_outip); else n->in_space = ~ntohl(n->in_outmsk); /* * Calculate the number of valid IP addresses in the output * mapping range. In all cases, the range is inclusive of * the start and ending IP addresses. * If to a CIDR address, lose 2: broadcast + network address * (so subtract 1) * If to a range, add one. * If to a single IP address, set to 1. */ if (n->in_space) { if ((n->in_flags & IPN_RANGE) != 0) n->in_space += 1; else n->in_space -= 1; } else n->in_space = 1; if ((n->in_outmsk != 0xffffffff) && (n->in_outmsk != 0) && ((n->in_flags & IPN_RANGE) == 0)) n->in_nip = ntohl(n->in_outip) + 1; else n->in_nip = ntohl(n->in_outip); if (n->in_redir & NAT_MAP) { n->in_pnext = ntohs(n->in_pmin); /* * Multiply by the number of ports made available. */ if (ntohs(n->in_pmax) >= ntohs(n->in_pmin)) { n->in_space *= (ntohs(n->in_pmax) - ntohs(n->in_pmin) + 1); /* * Because two different sources can map to * different destinations but use the same * local IP#/port #. * If the result is smaller than in_space, then * we may have wrapped around 32bits. */ i = n->in_inmsk; if ((i != 0) && (i != 0xffffffff)) { j = n->in_space * (~ntohl(i) + 1); if (j >= n->in_space) n->in_space = j; else n->in_space = 0xffffffff; } } /* * If no protocol is specified, multiple by 256. */ if ((n->in_flags & IPN_TCPUDP) == 0) { j = n->in_space * 256; if (j >= n->in_space) n->in_space = j; else n->in_space = 0xffffffff; } } /* Otherwise, these fields are preset */ n = NULL; nat_stats.ns_rules++; break; case SIOCRMNAT : if (!(mode & FWRITE)) { error = EPERM; n = NULL; break; } if (!n) { error = ESRCH; break; } if (n->in_redir & NAT_REDIRECT) nat_delrdr(n); if (n->in_redir & (NAT_MAPBLK|NAT_MAP)) nat_delnat(n); if (nat_list == NULL) { nat_masks = 0; rdr_masks = 0; } *np = n->in_next; if (!n->in_use) { if (n->in_apr) appr_free(n->in_apr); KFREE(n); nat_stats.ns_rules--; } else { n->in_flags |= IPN_DELETE; n->in_next = NULL; } n = NULL; break; case SIOCGNATS : MUTEX_DOWNGRADE(&ipf_nat); nat_stats.ns_table[0] = nat_table[0]; nat_stats.ns_table[1] = nat_table[1]; nat_stats.ns_list = nat_list; nat_stats.ns_nattab_sz = ipf_nattable_sz; nat_stats.ns_rultab_sz = ipf_natrules_sz; nat_stats.ns_rdrtab_sz = ipf_rdrrules_sz; nat_stats.ns_instances = nat_instances; nat_stats.ns_apslist = ap_sess_list; IWCOPY((char *)&nat_stats, (char *)data, sizeof(nat_stats)); break; case SIOCGNATL : { natlookup_t nl; MUTEX_DOWNGRADE(&ipf_nat); IRCOPY((char *)data, (char *)&nl, sizeof(nl)); if (nat_lookupredir(&nl)) { IWCOPY((char *)&nl, (char *)data, sizeof(nl)); } else error = ESRCH; break; } case SIOCFLNAT : if (!(mode & FWRITE)) { error = EPERM; break; } ret = nat_flushtable(); MUTEX_DOWNGRADE(&ipf_nat); IWCOPY((caddr_t)&ret, data, sizeof(ret)); break; case SIOCCNATL : if (!(mode & FWRITE)) { error = EPERM; break; } ret = nat_clearlist(); MUTEX_DOWNGRADE(&ipf_nat); IWCOPY((caddr_t)&ret, data, sizeof(ret)); break; case FIONREAD : #ifdef IPFILTER_LOG MUTEX_DOWNGRADE(&ipf_nat); IWCOPY((caddr_t)&iplused[IPL_LOGNAT], (caddr_t)data, sizeof(iplused[IPL_LOGNAT])); #endif break; default : error = EINVAL; break; } RWLOCK_EXIT(&ipf_nat); /* READ/WRITE */ SPL_X(s); if (nt) KFREE(nt); return error; } /* * Delete a nat entry from the various lists and table. */ static void nat_delete(natd) struct nat *natd; { register struct nat **natp, *nat; struct ipnat *ipn; for (natp = natd->nat_hstart[0]; (nat = *natp); natp = &nat->nat_hnext[0]) if (nat == natd) { *natp = nat->nat_hnext[0]; break; } for (natp = natd->nat_hstart[1]; (nat = *natp); natp = &nat->nat_hnext[1]) if (nat == natd) { *natp = nat->nat_hnext[1]; break; } if (natd->nat_fr != NULL) { ATOMIC_DEC(natd->nat_fr->fr_ref); } /* * If there is an active reference from the nat entry to its parent * rule, decrement the rule's reference count and free it too if no * longer being used. */ ipn = natd->nat_ptr; if (ipn != NULL) { ipn->in_space++; ipn->in_use--; if (!ipn->in_use && (ipn->in_flags & IPN_DELETE)) { if (ipn->in_apr) appr_free(ipn->in_apr); KFREE(ipn); nat_stats.ns_rules--; } } /* * If there's a fragment table entry too for this nat entry, then * dereference that as well. */ ipfr_forget((void *)natd); aps_free(natd->nat_aps); nat_stats.ns_inuse--; KFREE(natd); } /* * nat_flushtable - clear the NAT table of all mapping entries. */ static int nat_flushtable() { register nat_t *nat, **natp; register int j = 0; /* * ALL NAT mappings deleted, so lets just make the deletions * quicker. */ if (nat_table[0] != NULL) bzero((char *)nat_table[0], sizeof(nat_table[0]) * ipf_nattable_sz); if (nat_table[1] != NULL) bzero((char *)nat_table[1], sizeof(nat_table[1]) * ipf_nattable_sz); for (natp = &nat_instances; (nat = *natp); ) { *natp = nat->nat_next; nat_delete(nat); j++; } nat_stats.ns_inuse = 0; return j; } /* * nat_clearlist - delete all rules in the active NAT mapping list. */ static int nat_clearlist() { register ipnat_t *n, **np = &nat_list; int i = 0; if (nat_rules != NULL) bzero((char *)nat_rules, sizeof(*nat_rules) * ipf_natrules_sz); if (rdr_rules != NULL) bzero((char *)rdr_rules, sizeof(*rdr_rules) * ipf_rdrrules_sz); while ((n = *np)) { *np = n->in_next; if (!n->in_use) { if (n->in_apr) appr_free(n->in_apr); KFREE(n); nat_stats.ns_rules--; } else { n->in_flags |= IPN_DELETE; n->in_next = NULL; } i++; } nat_masks = 0; rdr_masks = 0; return i; } /* * Create a new NAT table entry. * NOTE: assumes write lock on ipf_nat has been obtained already. */ nat_t *nat_new(np, ip, fin, flags, direction) ipnat_t *np; ip_t *ip; fr_info_t *fin; u_int flags; int direction; { register u_32_t sum1, sum2, sumd, l; u_short port = 0, sport = 0, dport = 0, nport = 0; nat_t *nat, **natp, *natl = NULL; struct in_addr in, inb; tcphdr_t *tcp = NULL; u_short nflags; u_int hv; #if SOLARIS && defined(_KERNEL) && (SOLARIS2 >= 6) qif_t *qf = fin->fin_qif; #endif nflags = flags & np->in_flags; if (flags & IPN_TCPUDP) { tcp = (tcphdr_t *)fin->fin_dp; sport = tcp->th_sport; dport = tcp->th_dport; } /* Give me a new nat */ KMALLOC(nat, nat_t *); if (nat == NULL) return NULL; bzero((char *)nat, sizeof(*nat)); nat->nat_flags = flags; /* * Search the current table for a match. */ if (direction == NAT_OUTBOUND) { /* * Values at which the search for a free resouce starts. */ u_32_t st_ip; u_short st_port; /* * If it's an outbound packet which doesn't match any existing * record, then create a new port */ l = 0; st_ip = np->in_nip; st_port = np->in_pnext; do { port = 0; in.s_addr = np->in_nip; if (l == 0) { natl = nat_maplookup(fin->fin_ifp, flags, ip->ip_src, ip->ip_dst); if (natl != NULL) { in = natl->nat_outip; #ifndef sparc in.s_addr = ntohl(in.s_addr); #endif } } if ((np->in_outmsk == 0xffffffff) && (np->in_pnext == 0)) { if (l > 0) { KFREE(nat); return NULL; } } if (np->in_redir & NAT_MAPBLK) { if ((l >= np->in_ppip) || ((l > 0) && !(flags & IPN_TCPUDP))) { KFREE(nat); return NULL; } /* * map-block - Calculate destination address. */ in.s_addr = ntohl(ip->ip_src.s_addr); in.s_addr &= ntohl(~np->in_inmsk); inb.s_addr = in.s_addr; in.s_addr /= np->in_ippip; in.s_addr &= ntohl(~np->in_outmsk); in.s_addr += ntohl(np->in_outip); /* * Calculate destination port. */ if ((flags & IPN_TCPUDP) && (np->in_ppip != 0)) { port = ntohs(sport) + l; port %= np->in_ppip; port += np->in_ppip * (inb.s_addr % np->in_ippip); port += MAPBLK_MINPORT; port = htons(port); } } else if (!in.s_addr && (np->in_outmsk == 0xffffffff)) { /* * 0/32 - use the interface's IP address. */ if ((l > 0) || fr_ifpaddr(fin->fin_ifp, &in) == -1) { KFREE(nat); return NULL; } in.s_addr = ntohl(in.s_addr); } else if (!in.s_addr && !np->in_outmsk) { /* * 0/0 - use the original source address/port. */ if (l > 0) { KFREE(nat); return NULL; } in.s_addr = ntohl(ip->ip_src.s_addr); } else if ((np->in_outmsk != 0xffffffff) && (np->in_pnext == 0) && ((l > 0) || (natl == NULL))) np->in_nip++; natl = NULL; if ((nflags & IPN_TCPUDP) && ((np->in_redir & NAT_MAPBLK) == 0) && (np->in_flags & IPN_AUTOPORTMAP)) { if ((l > 0) && (l % np->in_ppip == 0)) { if (l > np->in_space) { KFREE(nat); return NULL; } else if ((l > np->in_ppip) && np->in_outmsk != 0xffffffff) np->in_nip++; } if (np->in_ppip != 0) { port = ntohs(sport); port += (l % np->in_ppip); port %= np->in_ppip; port += np->in_ppip * (ntohl(ip->ip_src.s_addr) % np->in_ippip); port += MAPBLK_MINPORT; port = htons(port); } } else if (((np->in_redir & NAT_MAPBLK) == 0) && (nflags & IPN_TCPUDP) && (np->in_pnext != 0)) { port = htons(np->in_pnext++); if (np->in_pnext > ntohs(np->in_pmax)) { np->in_pnext = ntohs(np->in_pmin); if (np->in_outmsk != 0xffffffff) np->in_nip++; } } if (np->in_flags & IPN_RANGE) { if (np->in_nip >= ntohl(np->in_outmsk)) np->in_nip = ntohl(np->in_outip); } else { if ((np->in_outmsk != 0xffffffff) && ((np->in_nip + 1) & ntohl(np->in_outmsk)) > ntohl(np->in_outip)) np->in_nip = ntohl(np->in_outip) + 1; } if (!port && (flags & IPN_TCPUDP)) port = sport; /* * Here we do a lookup of the connection as seen from * the outside. If an IP# pair already exists, try * again. So if you have A->B becomes C->B, you can * also have D->E become C->E but not D->B causing * another C->B. Also take protocol and ports into * account when determining whether a pre-existing * NAT setup will cause an external conflict where * this is appropriate. */ inb.s_addr = htonl(in.s_addr); natl = nat_inlookup(fin->fin_ifp, flags & ~FI_WILD, (u_int)ip->ip_p, ip->ip_dst, inb, (port << 16) | dport); /* * Has the search wrapped around and come back to the * start ? */ if ((natl != NULL) && (np->in_pnext != 0) && (st_port == np->in_pnext) && (np->in_nip != 0) && (st_ip == np->in_nip)) { KFREE(nat); return NULL; } l++; } while (natl != NULL); if (np->in_space > 0) np->in_space--; /* Setup the NAT table */ nat->nat_inip = ip->ip_src; nat->nat_outip.s_addr = htonl(in.s_addr); nat->nat_oip = ip->ip_dst; sum1 = LONG_SUM(ntohl(ip->ip_src.s_addr)) + ntohs(sport); sum2 = LONG_SUM(in.s_addr) + ntohs(port); if (flags & IPN_TCPUDP) { nat->nat_inport = sport; nat->nat_outport = port; /* sport */ nat->nat_oport = dport; } } else { /* * Otherwise, it's an inbound packet. Most likely, we don't * want to rewrite source ports and source addresses. Instead, * we want to rewrite to a fixed internal address and fixed * internal port. */ in.s_addr = ntohl(np->in_inip); if (!(nport = np->in_pnext)) nport = dport; /* * When the redirect-to address is set to 0.0.0.0, just * assume a blank `forwarding' of the packet. We don't * setup any translation for this either. */ if ((in.s_addr == 0) && (nport == dport)) { KFREE(nat); return NULL; } nat->nat_inip.s_addr = htonl(in.s_addr); nat->nat_outip = ip->ip_dst; nat->nat_oip = ip->ip_src; sum1 = LONG_SUM(ntohl(ip->ip_dst.s_addr)) + ntohs(dport); sum2 = LONG_SUM(in.s_addr) + ntohs(nport); if (flags & IPN_TCPUDP) { nat->nat_inport = nport; nat->nat_outport = dport; nat->nat_oport = sport; } } CALC_SUMD(sum1, sum2, sumd); nat->nat_sumd[0] = (sumd & 0xffff) + (sumd >> 16); #if SOLARIS && defined(_KERNEL) && (SOLARIS2 >= 6) if ((flags == IPN_TCP) && dohwcksum && (qf->qf_ill->ill_ick.ick_magic == ICK_M_CTL_MAGIC)) { if (direction == NAT_OUTBOUND) sum1 = LONG_SUM(ntohl(in.s_addr)); else sum1 = LONG_SUM(ntohl(ip->ip_src.s_addr)); sum1 += LONG_SUM(ntohl(ip->ip_dst.s_addr)); sum1 += 30; sum1 = (sum1 & 0xffff) + (sum1 >> 16); nat->nat_sumd[1] = NAT_HW_CKSUM|(sum1 & 0xffff); } else #endif nat->nat_sumd[1] = nat->nat_sumd[0]; if ((flags & IPN_TCPUDP) && ((sport != port) || (dport != nport))) { if (direction == NAT_OUTBOUND) sum1 = LONG_SUM(ntohl(ip->ip_src.s_addr)); else sum1 = LONG_SUM(ntohl(ip->ip_dst.s_addr)); sum2 = LONG_SUM(in.s_addr); CALC_SUMD(sum1, sum2, sumd); nat->nat_ipsumd = (sumd & 0xffff) + (sumd >> 16); } else nat->nat_ipsumd = nat->nat_sumd[0]; in.s_addr = htonl(in.s_addr); nat->nat_next = nat_instances; nat_instances = nat; hv = NAT_HASH_FN(nat->nat_inip.s_addr, ipf_nattable_sz); natp = &nat_table[0][hv]; nat->nat_hstart[0] = natp; nat->nat_hnext[0] = *natp; *natp = nat; hv = NAT_HASH_FN(nat->nat_outip.s_addr, ipf_nattable_sz); natp = &nat_table[1][hv]; nat->nat_hstart[1] = natp; nat->nat_hnext[1] = *natp; *natp = nat; nat->nat_dir = direction; nat->nat_ifp = fin->fin_ifp; nat->nat_ptr = np; nat->nat_p = ip->ip_p; nat->nat_bytes = 0; nat->nat_pkts = 0; nat->nat_age = fr_defnatage; nat->nat_fr = fin->fin_fr; if (nat->nat_fr != NULL) { ATOMIC_INC(nat->nat_fr->fr_ref); } if (direction == NAT_OUTBOUND) { if (flags & IPN_TCPUDP) tcp->th_sport = port; } else { if (flags & IPN_TCPUDP) tcp->th_dport = nport; } nat_stats.ns_added++; nat_stats.ns_inuse++; np->in_use++; return nat; } nat_t *nat_icmpinlookup(ip, fin) ip_t *ip; fr_info_t *fin; { icmphdr_t *icmp; tcphdr_t *tcp = NULL; ip_t *oip; int flags = 0, type; icmp = (icmphdr_t *)fin->fin_dp; /* * Does it at least have the return (basic) IP header ? * Only a basic IP header (no options) should be with an ICMP error * header. */ if ((ip->ip_hl != 5) || (ip->ip_len < ICMPERR_MINPKTLEN)) return NULL; type = icmp->icmp_type; /* * If it's not an error type, then return. */ if ((type != ICMP_UNREACH) && (type != ICMP_SOURCEQUENCH) && (type != ICMP_REDIRECT) && (type != ICMP_TIMXCEED) && (type != ICMP_PARAMPROB)) return NULL; oip = (ip_t *)((char *)fin->fin_dp + 8); if (ip->ip_len < ICMPERR_MAXPKTLEN + ((oip->ip_hl - 5) << 2)) return NULL; if (oip->ip_p == IPPROTO_TCP) flags = IPN_TCP; else if (oip->ip_p == IPPROTO_UDP) flags = IPN_UDP; if (flags & IPN_TCPUDP) { tcp = (tcphdr_t *)((char *)oip + (oip->ip_hl << 2)); return nat_inlookup(fin->fin_ifp, flags, (u_int)oip->ip_p, oip->ip_dst, oip->ip_src, (tcp->th_sport << 16) | tcp->th_dport); } return nat_inlookup(fin->fin_ifp, 0, (u_int)oip->ip_p, oip->ip_dst, oip->ip_src, 0); } /* * This should *ONLY* be used for incoming packets to make sure a NAT'd ICMP * packet gets correctly recognised. */ nat_t *nat_icmpin(ip, fin, nflags) ip_t *ip; fr_info_t *fin; u_int *nflags; { u_32_t sum1, sum2, sumd; struct in_addr in; icmphdr_t *icmp; nat_t *nat; ip_t *oip; int flags = 0; if (!(nat = nat_icmpinlookup(ip, fin))) return NULL; *nflags = IPN_ICMPERR; icmp = (icmphdr_t *)fin->fin_dp; oip = (ip_t *)&icmp->icmp_ip; if (oip->ip_p == IPPROTO_TCP) flags = IPN_TCP; else if (oip->ip_p == IPPROTO_UDP) flags = IPN_UDP; /* * Need to adjust ICMP header to include the real IP#'s and * port #'s. Only apply a checksum change relative to the * IP address change is it will be modified again in ip_natout * for both address and port. Two checksum changes are * necessary for the two header address changes. Be careful * to only modify the checksum once for the port # and twice * for the IP#. */ if (nat->nat_dir == NAT_OUTBOUND) { sum1 = LONG_SUM(ntohl(oip->ip_src.s_addr)); in = nat->nat_inip; oip->ip_src = in; } else { sum1 = LONG_SUM(ntohl(oip->ip_dst.s_addr)); in = nat->nat_outip; oip->ip_dst = in; } sum2 = LONG_SUM(ntohl(in.s_addr)); CALC_SUMD(sum1, sum2, sumd); if (nat->nat_dir == NAT_OUTBOUND) { fix_incksum(&oip->ip_sum, sumd, 0); sumd += (sumd & 0xffff); while (sumd > 0xffff) sumd = (sumd & 0xffff) + (sumd >> 16); fix_outcksum(&icmp->icmp_cksum, sumd, 0); } else { fix_outcksum(&oip->ip_sum, sumd, 0); sumd += (sumd & 0xffff); while (sumd > 0xffff) sumd = (sumd & 0xffff) + (sumd >> 16); fix_incksum(&icmp->icmp_cksum, sumd, 0); } if ((flags & IPN_TCPUDP) != 0) { tcphdr_t *tcp; /* XXX - what if this is bogus hl and we go off the end ? */ tcp = (tcphdr_t *)((((char *)oip) + (oip->ip_hl << 2))); if (nat->nat_dir == NAT_OUTBOUND) { if (tcp->th_sport != nat->nat_inport) { sum1 = ntohs(tcp->th_sport); sum2 = ntohs(nat->nat_inport); CALC_SUMD(sum1, sum2, sumd); tcp->th_sport = nat->nat_inport; fix_outcksum(&icmp->icmp_cksum, sumd, 0); } } else { if (tcp->th_dport != nat->nat_outport) { sum1 = ntohs(tcp->th_dport); sum2 = ntohs(nat->nat_outport); CALC_SUMD(sum1, sum2, sumd); tcp->th_dport = nat->nat_outport; fix_incksum(&icmp->icmp_cksum, sumd, 0); } } } nat->nat_age = fr_defnaticmpage; return nat; } /* * NB: these lookups don't lock access to the list, it assume it has already * been done! */ /* * Lookup a nat entry based on the mapped destination ip address/port and * real source address/port. We use this lookup when receiving a packet, * we're looking for a table entry, based on the destination address. * NOTE: THE PACKET BEING CHECKED (IF FOUND) HAS A MAPPING ALREADY. */ nat_t *nat_inlookup(ifp, flags, p, src, mapdst, ports) void *ifp; register u_int flags, p; struct in_addr src , mapdst; u_32_t ports; { register u_short sport, mapdport; register nat_t *nat; register int nflags; u_int hv; mapdport = ports >> 16; sport = ports & 0xffff; flags &= IPN_TCPUDP; hv = NAT_HASH_FN(mapdst.s_addr, ipf_nattable_sz); nat = nat_table[1][hv]; for (; nat; nat = nat->nat_hnext[1]) { nflags = nat->nat_flags; if ((!ifp || ifp == nat->nat_ifp) && nat->nat_oip.s_addr == src.s_addr && nat->nat_outip.s_addr == mapdst.s_addr && (((p == 0) && (flags == (nat->nat_flags & IPN_TCPUDP))) || (p == nat->nat_p)) && (!flags || (((nat->nat_oport == sport) || (nflags & FI_W_DPORT)) && ((nat->nat_outport == mapdport) || (nflags & FI_W_SPORT))))) return nat; } return NULL; } /* * Lookup a nat entry based on the source 'real' ip address/port and * destination address/port. We use this lookup when sending a packet out, * we're looking for a table entry, based on the source address. * NOTE: THE PACKET BEING CHECKED (IF FOUND) HAS A MAPPING ALREADY. */ nat_t *nat_outlookup(ifp, flags, p, src, dst, ports) void *ifp; register u_int flags, p; struct in_addr src , dst; u_32_t ports; { register u_short sport, dport; register nat_t *nat; register int nflags; u_int hv; sport = ports & 0xffff; dport = ports >> 16; flags &= IPN_TCPUDP; hv = NAT_HASH_FN(src.s_addr, ipf_nattable_sz); nat = nat_table[0][hv]; for (; nat; nat = nat->nat_hnext[0]) { nflags = nat->nat_flags; if ((!ifp || ifp == nat->nat_ifp) && nat->nat_inip.s_addr == src.s_addr && nat->nat_oip.s_addr == dst.s_addr && (((p == 0) && (flags == (nat->nat_flags & IPN_TCPUDP))) || (p == nat->nat_p)) && (!flags || ((nat->nat_inport == sport || nflags & FI_W_SPORT) && (nat->nat_oport == dport || nflags & FI_W_DPORT)))) return nat; } return NULL; } /* * check if an ip address has already been allocated for a given mapping that * is not doing port based translation. */ nat_t *nat_maplookup(ifp, flags, src, dst) void *ifp; register u_int flags; struct in_addr src , dst; { register nat_t *nat; register int oflags; u_int hv; hv = NAT_HASH_FN(src.s_addr, ipf_nattable_sz); nat = nat_table[0][hv]; for (; nat; nat = nat->nat_hnext[0]) { oflags = (flags & IPN_TCPUDP) & nat->nat_ptr->in_flags; if (oflags != 0) continue; if ((!ifp || ifp == nat->nat_ifp) && nat->nat_inip.s_addr == src.s_addr && nat->nat_oip.s_addr == dst.s_addr) return nat; } return NULL; } /* * Lookup the NAT tables to search for a matching redirect */ nat_t *nat_lookupredir(np) register natlookup_t *np; { u_32_t ports; nat_t *nat; ports = (np->nl_outport << 16) | np->nl_inport; /* * If nl_inip is non null, this is a lookup based on the real * ip address. Else, we use the fake. */ if ((nat = nat_outlookup(NULL, np->nl_flags, 0, np->nl_inip, np->nl_outip, ports))) { np->nl_realip = nat->nat_outip; np->nl_realport = nat->nat_outport; } return nat; } /* * Packets going out on the external interface go through this. * Here, the source address requires alteration, if anything. */ int ip_natout(ip, fin) ip_t *ip; fr_info_t *fin; { register ipnat_t *np = NULL; register u_32_t ipa; tcphdr_t *tcp = NULL; u_short nflags = 0, sport = 0, dport = 0, *csump = NULL; struct ifnet *ifp; int natadd = 1; frentry_t *fr; u_int hv, msk; u_32_t iph; nat_t *nat; int i; if (nat_list == NULL) return 0; if ((fr = fin->fin_fr) && !(fr->fr_flags & FR_DUP) && fr->fr_tif.fd_ifp && fr->fr_tif.fd_ifp != (void *)-1) ifp = fr->fr_tif.fd_ifp; else ifp = fin->fin_ifp; if (!(ip->ip_off & IP_OFFMASK) && !(fin->fin_fi.fi_fl & FI_SHORT)) { if (ip->ip_p == IPPROTO_TCP) nflags = IPN_TCP; else if (ip->ip_p == IPPROTO_UDP) nflags = IPN_UDP; if ((nflags & IPN_TCPUDP)) { tcp = (tcphdr_t *)fin->fin_dp; sport = tcp->th_sport; dport = tcp->th_dport; } } ipa = ip->ip_src.s_addr; READ_ENTER(&ipf_nat); if ((ip->ip_off & (IP_OFFMASK|IP_MF)) && (nat = ipfr_nat_knownfrag(ip, fin))) natadd = 0; else if ((nat = nat_outlookup(ifp, nflags, (u_int)ip->ip_p, ip->ip_src, ip->ip_dst, (dport << 16) | sport))) { nflags = nat->nat_flags; if ((nflags & (FI_W_SPORT|FI_W_DPORT)) != 0) { if ((nflags & FI_W_SPORT) && (nat->nat_inport != sport)) nat->nat_inport = sport; else if ((nflags & FI_W_DPORT) && (nat->nat_oport != dport)) nat->nat_oport = dport; if (nat->nat_outport == 0) nat->nat_outport = sport; nat->nat_flags &= ~(FI_W_DPORT|FI_W_SPORT); nflags = nat->nat_flags; } } else { RWLOCK_EXIT(&ipf_nat); WRITE_ENTER(&ipf_nat); /* * If there is no current entry in the nat table for this IP#, * create one for it (if there is a matching rule). */ msk = 0xffffffff; i = 32; maskloop: iph = ipa & htonl(msk); hv = NAT_HASH_FN(iph, ipf_natrules_sz); for (np = nat_rules[hv]; np; np = np->in_mnext) { if ((np->in_ifp == ifp) && np->in_space && (!(np->in_flags & IPN_RF) || (np->in_flags & nflags)) && ((ipa & np->in_inmsk) == np->in_inip) && ((np->in_redir & (NAT_MAP|NAT_MAPBLK)) || (np->in_pnext == sport))) { if (*np->in_plabel && !appr_ok(ip, tcp, np)) continue; /* * If it's a redirection, then we don't want to * create new outgoing port stuff. * Redirections are only for incoming * connections. */ if (!(np->in_redir & (NAT_MAP|NAT_MAPBLK))) continue; if ((nat = nat_new(np, ip, fin, (u_int)nflags, NAT_OUTBOUND))) { np->in_hits++; #ifdef IPFILTER_LOG nat_log(nat, (u_int)np->in_redir); #endif break; } } } if ((np == NULL) && (i > 0)) { do { i--; msk <<= 1; } while ((i >= 0) && ((nat_masks & (1 << i)) == 0)); if (i >= 0) goto maskloop; } MUTEX_DOWNGRADE(&ipf_nat); } if (nat) { np = nat->nat_ptr; if (natadd && fin->fin_fi.fi_fl & FI_FRAG) ipfr_nat_newfrag(ip, fin, 0, nat); ip->ip_src = nat->nat_outip; MUTEX_ENTER(&ipf_rw); nat->nat_age = fr_defnatage; nat->nat_bytes += ip->ip_len; nat->nat_pkts++; MUTEX_EXIT(&ipf_rw); /* * Fix up checksums, not by recalculating them, but * simply computing adjustments. */ #if SOLARIS || defined(__sgi) if (nat->nat_dir == NAT_OUTBOUND) fix_outcksum(&ip->ip_sum, nat->nat_ipsumd, 0); else fix_incksum(&ip->ip_sum, nat->nat_ipsumd, 0); #endif if (!(ip->ip_off & IP_OFFMASK) && !(fin->fin_fi.fi_fl & FI_SHORT)) { if ((nat->nat_outport != 0) && (nflags & IPN_TCPUDP)) { tcp->th_sport = nat->nat_outport; fin->fin_data[0] = ntohs(tcp->th_sport); } if (ip->ip_p == IPPROTO_TCP) { csump = &tcp->th_sum; MUTEX_ENTER(&ipf_rw); fr_tcp_age(&nat->nat_age, nat->nat_tcpstate, ip, fin, 1); if (nat->nat_age < fr_defnaticmpage) nat->nat_age = fr_defnaticmpage; #ifdef LARGE_NAT else if (nat->nat_age > DEF_NAT_AGE) nat->nat_age = DEF_NAT_AGE; #endif /* * Increase this because we may have * "keep state" following this too and * packet storms can occur if this is * removed too quickly. */ if (nat->nat_age == fr_tcpclosed) nat->nat_age = fr_tcplastack; MUTEX_EXIT(&ipf_rw); } else if (ip->ip_p == IPPROTO_UDP) { udphdr_t *udp = (udphdr_t *)tcp; if (udp->uh_sum) csump = &udp->uh_sum; } if (csump) { if (nat->nat_dir == NAT_OUTBOUND) fix_outcksum(csump, nat->nat_sumd[1], ip->ip_len); else fix_incksum(csump, nat->nat_sumd[1], ip->ip_len); } } if ((np->in_apr != NULL) && (np->in_dport == 0 || (tcp != NULL && dport == np->in_dport))) (void) appr_check(ip, fin, nat); ATOMIC_INC(nat_stats.ns_mapped[1]); RWLOCK_EXIT(&ipf_nat); /* READ */ return 1; } RWLOCK_EXIT(&ipf_nat); /* READ/WRITE */ return 0; } /* * Packets coming in from the external interface go through this. * Here, the destination address requires alteration, if anything. */ int ip_natin(ip, fin) ip_t *ip; fr_info_t *fin; { register struct in_addr src; register struct in_addr in; register ipnat_t *np; u_int nflags = 0, natadd = 1, hv, msk; struct ifnet *ifp = fin->fin_ifp; tcphdr_t *tcp = NULL; u_short sport = 0, dport = 0, *csump = NULL; nat_t *nat; u_32_t iph; int i; if (nat_list == NULL) return 0; if (!(ip->ip_off & IP_OFFMASK) && !(fin->fin_fi.fi_fl & FI_SHORT)) { if (ip->ip_p == IPPROTO_TCP) nflags = IPN_TCP; else if (ip->ip_p == IPPROTO_UDP) nflags = IPN_UDP; if ((nflags & IPN_TCPUDP)) { tcp = (tcphdr_t *)fin->fin_dp; dport = tcp->th_dport; sport = tcp->th_sport; } } in = ip->ip_dst; /* make sure the source address is to be redirected */ src = ip->ip_src; READ_ENTER(&ipf_nat); if ((ip->ip_p == IPPROTO_ICMP) && (nat = nat_icmpin(ip, fin, &nflags))) ; else if ((ip->ip_off & IP_OFFMASK) && (nat = ipfr_nat_knownfrag(ip, fin))) natadd = 0; else if ((nat = nat_inlookup(fin->fin_ifp, nflags, (u_int)ip->ip_p, ip->ip_src, in, (dport << 16) | sport))) { nflags = nat->nat_flags; if ((nflags & (FI_W_SPORT|FI_W_DPORT)) != 0) { if ((nat->nat_oport != sport) && (nflags & FI_W_DPORT)) nat->nat_oport = sport; else if ((nat->nat_outport != dport) && (nflags & FI_W_SPORT)) nat->nat_outport = dport; nat->nat_flags &= ~(FI_W_SPORT|FI_W_DPORT); nflags = nat->nat_flags; } } else { RWLOCK_EXIT(&ipf_nat); WRITE_ENTER(&ipf_nat); /* * If there is no current entry in the nat table for this IP#, * create one for it (if there is a matching rule). */ msk = 0xffffffff; i = 32; maskloop: iph = in.s_addr & htonl(msk); hv = NAT_HASH_FN(iph, ipf_rdrrules_sz); for (np = rdr_rules[hv]; np; np = np->in_rnext) if ((np->in_ifp == ifp) && (!np->in_flags || (nflags & np->in_flags)) && ((in.s_addr & np->in_outmsk) == np->in_outip) && ((src.s_addr & np->in_srcmsk) == np->in_srcip) && (np->in_redir & NAT_REDIRECT) && (!np->in_pmin || np->in_pmin == dport) && (!np->in_p || np->in_p == ip->ip_p)) { if ((nat = nat_new(np, ip, fin, nflags, NAT_INBOUND))) { np->in_hits++; #ifdef IPFILTER_LOG nat_log(nat, (u_int)np->in_redir); #endif break; } } if ((np == NULL) && (i > 0)) { do { i--; msk <<= 1; } while ((i >= 0) && ((rdr_masks & (1 << i)) == 0)); if (i >= 0) goto maskloop; } MUTEX_DOWNGRADE(&ipf_nat); } if (nat) { np = nat->nat_ptr; fin->fin_fr = nat->nat_fr; if (natadd && fin->fin_fi.fi_fl & FI_FRAG) ipfr_nat_newfrag(ip, fin, 0, nat); if ((np->in_apr != NULL) && (np->in_dport == 0 || (tcp != NULL && sport == np->in_dport))) (void) appr_check(ip, fin, nat); MUTEX_ENTER(&ipf_rw); if (nflags != IPN_ICMPERR) nat->nat_age = fr_defnatage; nat->nat_bytes += ip->ip_len; nat->nat_pkts++; MUTEX_EXIT(&ipf_rw); ip->ip_dst = nat->nat_inip; fin->fin_fi.fi_dst = nat->nat_inip; /* * Fix up checksums, not by recalculating them, but * simply computing adjustments. */ #if SOLARIS || defined(__sgi) if (nat->nat_dir == NAT_OUTBOUND) fix_incksum(&ip->ip_sum, nat->nat_ipsumd, 0); else fix_outcksum(&ip->ip_sum, nat->nat_ipsumd, 0); #endif if (!(ip->ip_off & IP_OFFMASK) && !(fin->fin_fi.fi_fl & FI_SHORT)) { if ((nat->nat_inport != 0) && (nflags & IPN_TCPUDP)) { tcp->th_dport = nat->nat_inport; fin->fin_data[1] = ntohs(tcp->th_dport); } if (ip->ip_p == IPPROTO_TCP) { csump = &tcp->th_sum; MUTEX_ENTER(&ipf_rw); fr_tcp_age(&nat->nat_age, nat->nat_tcpstate, ip, fin, 0); if (nat->nat_age < fr_defnaticmpage) nat->nat_age = fr_defnaticmpage; #ifdef LARGE_NAT else if (nat->nat_age > DEF_NAT_AGE) nat->nat_age = DEF_NAT_AGE; #endif /* * Increase this because we may have * "keep state" following this too and * packet storms can occur if this is * removed too quickly. */ if (nat->nat_age == fr_tcpclosed) nat->nat_age = fr_tcplastack; MUTEX_EXIT(&ipf_rw); } else if (ip->ip_p == IPPROTO_UDP) { udphdr_t *udp = (udphdr_t *)tcp; if (udp->uh_sum) csump = &udp->uh_sum; } if (csump) { if (nat->nat_dir == NAT_OUTBOUND) fix_incksum(csump, nat->nat_sumd[0], 0); else fix_outcksum(csump, nat->nat_sumd[0], 0); } } ATOMIC_INC(nat_stats.ns_mapped[0]); RWLOCK_EXIT(&ipf_nat); /* READ */ return 1; } RWLOCK_EXIT(&ipf_nat); /* READ/WRITE */ return 0; } /* * Free all memory used by NAT structures allocated at runtime. */ void ip_natunload() { WRITE_ENTER(&ipf_nat); (void) nat_clearlist(); (void) nat_flushtable(); RWLOCK_EXIT(&ipf_nat); if (nat_table[0] != NULL) { KFREES(nat_table[0], sizeof(nat_t *) * ipf_nattable_sz); nat_table[0] = NULL; } if (nat_table[1] != NULL) { KFREES(nat_table[1], sizeof(nat_t *) * ipf_nattable_sz); nat_table[1] = NULL; } if (nat_rules != NULL) { KFREES(nat_rules, sizeof(ipnat_t *) * ipf_natrules_sz); nat_rules = NULL; } if (rdr_rules != NULL) { KFREES(rdr_rules, sizeof(ipnat_t *) * ipf_rdrrules_sz); rdr_rules = NULL; } } /* * Slowly expire held state for NAT entries. Timeouts are set in * expectation of this being called twice per second. */ void ip_natexpire() { register struct nat *nat, **natp; #if defined(_KERNEL) && !SOLARIS int s; #endif SPL_NET(s); WRITE_ENTER(&ipf_nat); for (natp = &nat_instances; (nat = *natp); ) { nat->nat_age--; if (nat->nat_age) { natp = &nat->nat_next; continue; } *natp = nat->nat_next; #ifdef IPFILTER_LOG nat_log(nat, NL_EXPIRE); #endif nat_delete(nat); nat_stats.ns_expire++; } RWLOCK_EXIT(&ipf_nat); SPL_X(s); } /* */ void ip_natsync(ifp) void *ifp; { register ipnat_t *n; register nat_t *nat; register u_32_t sum1, sum2, sumd; struct in_addr in; ipnat_t *np; void *ifp2; #if defined(_KERNEL) && !SOLARIS int s; #endif /* * Change IP addresses for NAT sessions for any protocol except TCP * since it will break the TCP connection anyway. */ SPL_NET(s); WRITE_ENTER(&ipf_nat); for (nat = nat_instances; nat; nat = nat->nat_next) if (((ifp == NULL) || (ifp == nat->nat_ifp)) && !(nat->nat_flags & IPN_TCP) && (np = nat->nat_ptr) && (np->in_outmsk == 0xffffffff) && !np->in_nip) { ifp2 = nat->nat_ifp; /* * Change the map-to address to be the same as the * new one. */ sum1 = nat->nat_outip.s_addr; if (fr_ifpaddr(ifp2, &in) != -1) nat->nat_outip = in; sum2 = nat->nat_outip.s_addr; if (sum1 == sum2) continue; /* * Readjust the checksum adjustment to take into * account the new IP#. */ CALC_SUMD(sum1, sum2, sumd); /* XXX - dont change for TCP when solaris does * hardware checksumming. */ sumd += nat->nat_sumd[0]; nat->nat_sumd[0] = (sumd & 0xffff) + (sumd >> 16); nat->nat_sumd[1] = nat->nat_sumd[0]; } for (n = nat_list; (n != NULL); n = n->in_next) if (n->in_ifp == ifp) { n->in_ifp = (void *)GETUNIT(n->in_ifname); if (!n->in_ifp) n->in_ifp = (void *)-1; } RWLOCK_EXIT(&ipf_nat); SPL_X(s); } #ifdef IPFILTER_LOG void nat_log(nat, type) struct nat *nat; u_int type; { struct ipnat *np; struct natlog natl; void *items[1]; size_t sizes[1]; int rulen, types[1]; natl.nl_inip = nat->nat_inip; natl.nl_outip = nat->nat_outip; natl.nl_origip = nat->nat_oip; natl.nl_bytes = nat->nat_bytes; natl.nl_pkts = nat->nat_pkts; natl.nl_origport = nat->nat_oport; natl.nl_inport = nat->nat_inport; natl.nl_outport = nat->nat_outport; natl.nl_type = type; natl.nl_rule = -1; #ifndef LARGE_NAT if (nat->nat_ptr != NULL) { for (rulen = 0, np = nat_list; np; np = np->in_next, rulen++) if (np == nat->nat_ptr) { natl.nl_rule = rulen; break; } } #endif items[0] = &natl; sizes[0] = sizeof(natl); types[0] = 0; (void) ipllog(IPL_LOGNAT, NULL, items, sizes, types, 1); } #endif