/* $NetBSD: in_pcb.c,v 1.202 2022/11/04 09:05:41 ozaki-r Exp $ */ /* * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project. * All rights reserved. * * 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. Neither the name of the project 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 PROJECT 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 PROJECT 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) 1998, 2011 The NetBSD Foundation, Inc. * All rights reserved. * * This code is derived from software contributed to The NetBSD Foundation * by Coyote Point Systems, Inc. * 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. * * 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, 1991, 1993, 1995 * The Regents of the University of California. All rights reserved. * * 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. Neither the name of the University 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 REGENTS 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 REGENTS 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. * * @(#)in_pcb.c 8.4 (Berkeley) 5/24/95 */ #include __KERNEL_RCSID(0, "$NetBSD: in_pcb.c,v 1.202 2022/11/04 09:05:41 ozaki-r Exp $"); #ifdef _KERNEL_OPT #include "opt_inet.h" #include "opt_ipsec.h" #endif #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #ifdef INET6 #include #include #include #endif #ifdef IPSEC #include #include #endif /* IPSEC */ #include struct in_addr zeroin_addr; #define INPCBHASH_PORT(table, lport) \ &(table)->inpt_porthashtbl[ntohs(lport) & (table)->inpt_porthash] #define INPCBHASH_BIND(table, laddr, lport) \ &(table)->inpt_bindhashtbl[ \ ((ntohl((laddr).s_addr) + ntohs(lport))) & (table)->inpt_bindhash] #define INPCBHASH_CONNECT(table, faddr, fport, laddr, lport) \ &(table)->inpt_connecthashtbl[ \ ((ntohl((faddr).s_addr) + ntohs(fport)) + \ (ntohl((laddr).s_addr) + ntohs(lport))) & (table)->inpt_connecthash] int anonportmin = IPPORT_ANONMIN; int anonportmax = IPPORT_ANONMAX; int lowportmin = IPPORT_RESERVEDMIN; int lowportmax = IPPORT_RESERVEDMAX; static pool_cache_t in4pcb_pool_cache; #ifdef INET6 static pool_cache_t in6pcb_pool_cache; #endif static int inpcb_poolinit(void) { in4pcb_pool_cache = pool_cache_init(sizeof(struct in4pcb), coherency_unit, 0, 0, "in4pcbpl", NULL, IPL_NET, NULL, NULL, NULL); #ifdef INET6 in6pcb_pool_cache = pool_cache_init(sizeof(struct in6pcb), coherency_unit, 0, 0, "in6pcbpl", NULL, IPL_NET, NULL, NULL, NULL); #endif return 0; } void inpcb_init(struct inpcbtable *table, int bindhashsize, int connecthashsize) { static ONCE_DECL(control); TAILQ_INIT(&table->inpt_queue); table->inpt_porthashtbl = hashinit(bindhashsize, HASH_LIST, true, &table->inpt_porthash); table->inpt_bindhashtbl = hashinit(bindhashsize, HASH_LIST, true, &table->inpt_bindhash); table->inpt_connecthashtbl = hashinit(connecthashsize, HASH_LIST, true, &table->inpt_connecthash); table->inpt_lastlow = IPPORT_RESERVEDMAX; table->inpt_lastport = (in_port_t)anonportmax; RUN_ONCE(&control, inpcb_poolinit); } /* * inpcb_create: construct a new PCB and associated with a given socket. * Sets the PCB state to INP_ATTACHED and makes PCB globally visible. */ int inpcb_create(struct socket *so, void *v) { struct inpcbtable *table = v; struct inpcb *inp; int s; #ifdef INET6 KASSERT(soaf(so) == AF_INET || soaf(so) == AF_INET6); if (soaf(so) == AF_INET) inp = pool_cache_get(in4pcb_pool_cache, PR_NOWAIT); else inp = pool_cache_get(in6pcb_pool_cache, PR_NOWAIT); #else KASSERT(soaf(so) == AF_INET); inp = pool_cache_get(in4pcb_pool_cache, PR_NOWAIT); #endif if (inp == NULL) return ENOBUFS; if (soaf(so) == AF_INET) memset(inp, 0, sizeof(struct in4pcb)); #ifdef INET6 else memset(inp, 0, sizeof(struct in6pcb)); #endif inp->inp_af = soaf(so); inp->inp_table = table; inp->inp_socket = so; inp->inp_portalgo = PORTALGO_DEFAULT; inp->inp_bindportonsend = false; if (inp->inp_af == AF_INET) { in4p_errormtu(inp) = -1; in4p_prefsrcip(inp).s_addr = INADDR_ANY; } #ifdef INET6 else { in6p_hops6(inp) = -1; /* use kernel default */ if (ip6_v6only) inp->inp_flags |= IN6P_IPV6_V6ONLY; } #endif #if defined(IPSEC) if (ipsec_enabled) { int error = ipsec_init_pcbpolicy(so, &inp->inp_sp); if (error != 0) { #ifdef INET6 if (inp->inp_af == AF_INET) pool_cache_put(in4pcb_pool_cache, inp); else pool_cache_put(in6pcb_pool_cache, inp); #else KASSERT(inp->inp_af == AF_INET); pool_cache_put(in4pcb_pool_cache, inp); #endif return error; } inp->inp_sp->sp_inp = inp; } #endif so->so_pcb = inp; s = splsoftnet(); TAILQ_INSERT_HEAD(&table->inpt_queue, inp, inp_queue); LIST_INSERT_HEAD(INPCBHASH_PORT(table, inp->inp_lport), inp, inp_lhash); inpcb_set_state(inp, INP_ATTACHED); splx(s); return 0; } static int inpcb_set_port(struct sockaddr_in *sin, struct inpcb *inp, kauth_cred_t cred) { struct inpcbtable *table = inp->inp_table; struct socket *so = inp->inp_socket; in_port_t *lastport; in_port_t lport = 0; enum kauth_network_req req; int error; if (inp->inp_flags & INP_LOWPORT) { #ifndef IPNOPRIVPORTS req = KAUTH_REQ_NETWORK_BIND_PRIVPORT; #else req = KAUTH_REQ_NETWORK_BIND_PORT; #endif lastport = &table->inpt_lastlow; } else { req = KAUTH_REQ_NETWORK_BIND_PORT; lastport = &table->inpt_lastport; } /* XXX-kauth: KAUTH_REQ_NETWORK_BIND_AUTOASSIGN_{,PRIV}PORT */ error = kauth_authorize_network(cred, KAUTH_NETWORK_BIND, req, so, sin, NULL); if (error) return EACCES; /* * Use RFC6056 randomized port selection */ error = portalgo_randport(&lport, inp, cred); if (error) return error; inp->inp_flags |= INP_ANONPORT; *lastport = lport; lport = htons(lport); inp->inp_lport = lport; inpcb_set_state(inp, INP_BOUND); return 0; } int inpcb_bindableaddr(const struct inpcb *inp, struct sockaddr_in *sin, kauth_cred_t cred) { int error = EADDRNOTAVAIL; struct ifaddr *ifa = NULL; int s; if (sin->sin_family != AF_INET) return EAFNOSUPPORT; s = pserialize_read_enter(); if (IN_MULTICAST(sin->sin_addr.s_addr)) { /* Always succeed; port reuse handled in inpcb_bind_port(). */ } else if (!in_nullhost(sin->sin_addr)) { struct in_ifaddr *ia; ia = in_get_ia(sin->sin_addr); /* check for broadcast addresses */ if (ia == NULL) { ifa = ifa_ifwithaddr(sintosa(sin)); if (ifa != NULL) ia = ifatoia(ifa); else if ((inp->inp_flags & INP_BINDANY) != 0) { error = 0; goto error; } } if (ia == NULL) goto error; if (ia->ia4_flags & IN_IFF_DUPLICATED) goto error; } error = 0; error: pserialize_read_exit(s); return error; } static int inpcb_bind_addr(struct inpcb *inp, struct sockaddr_in *sin, kauth_cred_t cred) { int error; error = inpcb_bindableaddr(inp, sin, cred); if (error == 0) in4p_laddr(inp) = sin->sin_addr; return error; } static int inpcb_bind_port(struct inpcb *inp, struct sockaddr_in *sin, kauth_cred_t cred) { struct inpcbtable *table = inp->inp_table; struct socket *so = inp->inp_socket; int reuseport = (so->so_options & SO_REUSEPORT); int wild = 0, error; if (IN_MULTICAST(sin->sin_addr.s_addr)) { /* * Treat SO_REUSEADDR as SO_REUSEPORT for multicast; * allow complete duplication of binding if * SO_REUSEPORT is set, or if SO_REUSEADDR is set * and a multicast address is bound on both * new and duplicated sockets. */ if (so->so_options & (SO_REUSEADDR | SO_REUSEPORT)) reuseport = SO_REUSEADDR|SO_REUSEPORT; } if (sin->sin_port == 0) { error = inpcb_set_port(sin, inp, cred); if (error) return error; } else { struct inpcb *t; vestigial_inpcb_t vestige; #ifdef INET6 struct inpcb *t6; struct in6_addr mapped; #endif enum kauth_network_req req; if ((so->so_options & (SO_REUSEADDR|SO_REUSEPORT)) == 0) wild = 1; #ifndef IPNOPRIVPORTS if (ntohs(sin->sin_port) < IPPORT_RESERVED) req = KAUTH_REQ_NETWORK_BIND_PRIVPORT; else #endif /* !IPNOPRIVPORTS */ req = KAUTH_REQ_NETWORK_BIND_PORT; error = kauth_authorize_network(cred, KAUTH_NETWORK_BIND, req, so, sin, NULL); if (error) return EACCES; #ifdef INET6 in6_in_2_v4mapin6(&sin->sin_addr, &mapped); t6 = in6pcb_lookup_local(table, &mapped, sin->sin_port, wild, &vestige); if (t6 && (reuseport & t6->inp_socket->so_options) == 0) return EADDRINUSE; if (!t6 && vestige.valid) { if (!!reuseport != !!vestige.reuse_port) { return EADDRINUSE; } } #endif /* XXX-kauth */ if (so->so_uidinfo->ui_uid && !IN_MULTICAST(sin->sin_addr.s_addr)) { t = inpcb_lookup_local(table, sin->sin_addr, sin->sin_port, 1, &vestige); /* * XXX: investigate ramifications of loosening this * restriction so that as long as both ports have * SO_REUSEPORT allow the bind */ if (t && (!in_nullhost(sin->sin_addr) || !in_nullhost(in4p_laddr(t)) || (t->inp_socket->so_options & SO_REUSEPORT) == 0) && (so->so_uidinfo->ui_uid != t->inp_socket->so_uidinfo->ui_uid)) { return EADDRINUSE; } if (!t && vestige.valid) { if ((!in_nullhost(sin->sin_addr) || !in_nullhost(vestige.laddr.v4) || !vestige.reuse_port) && so->so_uidinfo->ui_uid != vestige.uid) { return EADDRINUSE; } } } t = inpcb_lookup_local(table, sin->sin_addr, sin->sin_port, wild, &vestige); if (t && (reuseport & t->inp_socket->so_options) == 0) return EADDRINUSE; if (!t && vestige.valid && !(reuseport && vestige.reuse_port)) return EADDRINUSE; inp->inp_lport = sin->sin_port; inpcb_set_state(inp, INP_BOUND); } LIST_REMOVE(inp, inp_lhash); LIST_INSERT_HEAD(INPCBHASH_PORT(table, inp->inp_lport), inp, inp_lhash); return 0; } /* * inpcb_bind: assign a local IP address and port number to the PCB. * * If the address is not a wildcard, verify that it corresponds to a * local interface. If a port is specified and it is privileged, then * check the permission. Check whether the address or port is in use, * and if so, whether we can re-use them. */ int inpcb_bind(void *v, struct sockaddr_in *sin, struct lwp *l) { struct inpcb *inp = v; struct sockaddr_in lsin; int error; if (inp->inp_af != AF_INET) return EINVAL; if (inp->inp_lport || !in_nullhost(in4p_laddr(inp))) return EINVAL; if (NULL != sin) { if (sin->sin_len != sizeof(*sin)) return EINVAL; } else { lsin = *((const struct sockaddr_in *) inp->inp_socket->so_proto->pr_domain->dom_sa_any); sin = &lsin; } /* Bind address. */ error = inpcb_bind_addr(inp, sin, l->l_cred); if (error) return error; /* Bind port. */ error = inpcb_bind_port(inp, sin, l->l_cred); if (error) { in4p_laddr(inp).s_addr = INADDR_ANY; return error; } return 0; } /* * inpcb_connect: connect from a socket to a specified address, i.e., * assign a foreign IP address and port number to the PCB. * * Both address and port must be specified in the name argument. * If there is no local address for this socket yet, then pick one. */ int inpcb_connect(void *v, struct sockaddr_in *sin, struct lwp *l) { struct inpcb *inp = v; vestigial_inpcb_t vestige; int error; struct in_addr laddr; if (inp->inp_af != AF_INET) return EINVAL; if (sin->sin_len != sizeof (*sin)) return EINVAL; if (sin->sin_family != AF_INET) return EAFNOSUPPORT; if (sin->sin_port == 0) return EADDRNOTAVAIL; if (IN_MULTICAST(sin->sin_addr.s_addr) && inp->inp_socket->so_type == SOCK_STREAM) return EADDRNOTAVAIL; if (!IN_ADDRLIST_READER_EMPTY()) { /* * If the destination address is INADDR_ANY, * use any local address (likely loopback). * If the supplied address is INADDR_BROADCAST, * use the broadcast address of an interface * which supports broadcast. (loopback does not) */ if (in_nullhost(sin->sin_addr)) { /* XXX racy */ sin->sin_addr = IN_ADDRLIST_READER_FIRST()->ia_addr.sin_addr; } else if (sin->sin_addr.s_addr == INADDR_BROADCAST) { struct in_ifaddr *ia; int s = pserialize_read_enter(); IN_ADDRLIST_READER_FOREACH(ia) { if (ia->ia_ifp->if_flags & IFF_BROADCAST) { sin->sin_addr = ia->ia_broadaddr.sin_addr; break; } } pserialize_read_exit(s); } } /* * If we haven't bound which network number to use as ours, * we will use the number of the outgoing interface. * This depends on having done a routing lookup, which * we will probably have to do anyway, so we might * as well do it now. On the other hand if we are * sending to multiple destinations we may have already * done the lookup, so see if we can use the route * from before. In any case, we only * chose a port number once, even if sending to multiple * destinations. */ if (in_nullhost(in4p_laddr(inp))) { int xerror; struct in_ifaddr *ia, *_ia; int s; struct psref psref; int bound; bound = curlwp_bind(); ia = in_selectsrc(sin, &inp->inp_route, inp->inp_socket->so_options, inp->inp_moptions, &xerror, &psref); if (ia == NULL) { curlwp_bindx(bound); if (xerror == 0) xerror = EADDRNOTAVAIL; return xerror; } s = pserialize_read_enter(); _ia = in_get_ia(IA_SIN(ia)->sin_addr); if (_ia == NULL && (inp->inp_flags & INP_BINDANY) == 0) { pserialize_read_exit(s); ia4_release(ia, &psref); curlwp_bindx(bound); return EADDRNOTAVAIL; } pserialize_read_exit(s); laddr = IA_SIN(ia)->sin_addr; ia4_release(ia, &psref); curlwp_bindx(bound); } else laddr = in4p_laddr(inp); if (inpcb_lookup(inp->inp_table, sin->sin_addr, sin->sin_port, laddr, inp->inp_lport, &vestige) != NULL || vestige.valid) { return EADDRINUSE; } if (in_nullhost(in4p_laddr(inp))) { if (inp->inp_lport == 0) { error = inpcb_bind(inp, NULL, l); /* * This used to ignore the return value * completely, but we need to check for * ephemeral port shortage. * And attempts to request low ports if not root. */ if (error != 0) return error; } in4p_laddr(inp) = laddr; } in4p_faddr(inp) = sin->sin_addr; inp->inp_fport = sin->sin_port; /* Late bind, if needed */ if (inp->inp_bindportonsend) { struct sockaddr_in lsin = *((const struct sockaddr_in *) inp->inp_socket->so_proto->pr_domain->dom_sa_any); lsin.sin_addr = in4p_laddr(inp); lsin.sin_port = 0; if ((error = inpcb_bind_port(inp, &lsin, l->l_cred)) != 0) return error; } inpcb_set_state(inp, INP_CONNECTED); #if defined(IPSEC) if (ipsec_enabled && inp->inp_socket->so_type == SOCK_STREAM) ipsec_pcbconn(inp->inp_sp); #endif return 0; } /* * inpcb_disconnect: remove any foreign IP/port association. * * Note: destroys the PCB if socket was closed. */ void inpcb_disconnect(void *v) { struct inpcb *inp = v; if (inp->inp_af != AF_INET) return; in4p_faddr(inp) = zeroin_addr; inp->inp_fport = 0; inpcb_set_state(inp, INP_BOUND); #if defined(IPSEC) if (ipsec_enabled) ipsec_pcbdisconn(inp->inp_sp); #endif if (inp->inp_socket->so_state & SS_NOFDREF) inpcb_destroy(inp); } /* * inpcb_destroy: destroy PCB as well as the associated socket. */ void inpcb_destroy(void *v) { struct inpcb *inp = v; struct socket *so = inp->inp_socket; int s; KASSERT(inp->inp_af == AF_INET || inp->inp_af == AF_INET6); #if defined(IPSEC) if (ipsec_enabled) ipsec_delete_pcbpolicy(inp); #endif so->so_pcb = NULL; s = splsoftnet(); inpcb_set_state(inp, INP_ATTACHED); LIST_REMOVE(inp, inp_lhash); TAILQ_REMOVE(&inp->inp_table->inpt_queue, inp, inp_queue); splx(s); if (inp->inp_options) { m_free(inp->inp_options); } rtcache_free(&inp->inp_route); ip_freemoptions(inp->inp_moptions); #ifdef INET6 if (inp->inp_af == AF_INET6) { if (in6p_outputopts(inp) != NULL) { ip6_clearpktopts(in6p_outputopts(inp), -1); free(in6p_outputopts(inp), M_IP6OPT); } ip6_freemoptions(in6p_moptions(inp)); } #endif sofree(so); /* drops the socket's lock */ #ifdef INET6 if (inp->inp_af == AF_INET) pool_cache_put(in4pcb_pool_cache, inp); else pool_cache_put(in6pcb_pool_cache, inp); #else KASSERT(inp->inp_af == AF_INET); pool_cache_put(in4pcb_pool_cache, inp); #endif mutex_enter(softnet_lock); /* reacquire the softnet_lock */ } /* * inpcb_fetch_sockaddr: fetch the local IP address and port number. */ void inpcb_fetch_sockaddr(struct inpcb *inp, struct sockaddr_in *sin) { if (inp->inp_af != AF_INET) return; sockaddr_in_init(sin, &in4p_laddr(inp), inp->inp_lport); } /* * inpcb_fetch_peeraddr: fetch the foreign IP address and port number. */ void inpcb_fetch_peeraddr(struct inpcb *inp, struct sockaddr_in *sin) { if (inp->inp_af != AF_INET) return; sockaddr_in_init(sin, &in4p_faddr(inp), inp->inp_fport); } /* * inpcb_notify: pass some notification to all connections of a protocol * associated with destination address. The local address and/or port * numbers may be specified to limit the search. The "usual action" will * be taken, depending on the command. * * The caller must filter any commands that are not interesting (e.g., * no error in the map). Call the protocol specific routine (if any) to * report any errors for each matching socket. * * Must be called at splsoftnet. */ int inpcb_notify(struct inpcbtable *table, struct in_addr faddr, u_int fport_arg, struct in_addr laddr, u_int lport_arg, int errno, void (*notify)(struct inpcb *, int)) { struct inpcbhead *head; struct inpcb *inp; in_port_t fport = fport_arg, lport = lport_arg; int nmatch; if (in_nullhost(faddr) || notify == NULL) return 0; nmatch = 0; head = INPCBHASH_CONNECT(table, faddr, fport, laddr, lport); LIST_FOREACH(inp, head, inp_hash) { if (inp->inp_af != AF_INET) continue; if (in_hosteq(in4p_faddr(inp), faddr) && inp->inp_fport == fport && inp->inp_lport == lport && in_hosteq(in4p_laddr(inp), laddr)) { (*notify)(inp, errno); nmatch++; } } return nmatch; } void inpcb_notifyall(struct inpcbtable *table, struct in_addr faddr, int errno, void (*notify)(struct inpcb *, int)) { struct inpcb *inp; if (in_nullhost(faddr) || notify == NULL) return; TAILQ_FOREACH(inp, &table->inpt_queue, inp_queue) { if (inp->inp_af != AF_INET) continue; if (in_hosteq(in4p_faddr(inp), faddr)) (*notify)(inp, errno); } } void in_purgeifmcast(struct ip_moptions *imo, struct ifnet *ifp) { int i, gap; /* The owner of imo should be protected by solock */ KASSERT(ifp != NULL); if (imo == NULL) return; /* * Unselect the outgoing interface if it is being * detached. */ if (imo->imo_multicast_if_index == ifp->if_index) imo->imo_multicast_if_index = 0; /* * Drop multicast group membership if we joined * through the interface being detached. */ for (i = 0, gap = 0; i < imo->imo_num_memberships; i++) { if (imo->imo_membership[i]->inm_ifp == ifp) { in_delmulti(imo->imo_membership[i]); gap++; } else if (gap != 0) imo->imo_membership[i - gap] = imo->imo_membership[i]; } imo->imo_num_memberships -= gap; } void inpcb_purgeif0(struct inpcbtable *table, struct ifnet *ifp) { struct inpcb *inp; TAILQ_FOREACH(inp, &table->inpt_queue, inp_queue) { bool need_unlock = false; if (inp->inp_af != AF_INET) continue; /* The caller holds either one of inps' lock */ if (!inp_locked(inp)) { inp_lock(inp); need_unlock = true; } in_purgeifmcast(inp->inp_moptions, ifp); if (need_unlock) inp_unlock(inp); } } void inpcb_purgeif(struct inpcbtable *table, struct ifnet *ifp) { struct rtentry *rt; struct inpcb *inp; TAILQ_FOREACH(inp, &table->inpt_queue, inp_queue) { if (inp->inp_af != AF_INET) continue; if ((rt = rtcache_validate(&inp->inp_route)) != NULL && rt->rt_ifp == ifp) { rtcache_unref(rt, &inp->inp_route); inpcb_rtchange(inp, 0); } else rtcache_unref(rt, &inp->inp_route); } } /* * inpcb_losing: check for alternatives when higher level complains about * service problems. For now, invalidate cached routing information. * If the route was created dynamically (by a redirect), time to try a * default gateway again. */ void inpcb_losing(struct inpcb *inp) { struct rtentry *rt; struct rt_addrinfo info; if (inp->inp_af != AF_INET) return; if ((rt = rtcache_validate(&inp->inp_route)) == NULL) return; memset(&info, 0, sizeof(info)); info.rti_info[RTAX_DST] = rtcache_getdst(&inp->inp_route); info.rti_info[RTAX_GATEWAY] = rt->rt_gateway; info.rti_info[RTAX_NETMASK] = rt_mask(rt); rt_missmsg(RTM_LOSING, &info, rt->rt_flags, 0); if (rt->rt_flags & RTF_DYNAMIC) { int error; struct rtentry *nrt; error = rtrequest(RTM_DELETE, rt_getkey(rt), rt->rt_gateway, rt_mask(rt), rt->rt_flags, &nrt); rtcache_unref(rt, &inp->inp_route); if (error == 0) { rt_newmsg_dynamic(RTM_DELETE, nrt); rt_free(nrt); } } else rtcache_unref(rt, &inp->inp_route); /* * A new route can be allocated * the next time output is attempted. */ rtcache_free(&inp->inp_route); } /* * inpcb_rtchange: after a routing change, flush old routing. * A new route can be allocated the next time output is attempted. */ void inpcb_rtchange(struct inpcb *inp, int errno) { if (inp->inp_af != AF_INET) return; rtcache_free(&inp->inp_route); /* XXX SHOULD NOTIFY HIGHER-LEVEL PROTOCOLS */ } /* * inpcb_lookup_local: find a PCB by looking at the local port and matching * the local address or resolving the wildcards. Primarily used to detect * when the local address is already in use. */ struct inpcb * inpcb_lookup_local(struct inpcbtable *table, struct in_addr laddr, u_int lport_arg, int lookup_wildcard, vestigial_inpcb_t *vp) { struct inpcbhead *head; struct inpcb *inp; struct inpcb *match = NULL; int matchwild = 3; int wildcard; in_port_t lport = lport_arg; if (vp) vp->valid = 0; head = INPCBHASH_PORT(table, lport); LIST_FOREACH(inp, head, inp_lhash) { if (inp->inp_af != AF_INET) continue; if (inp->inp_lport != lport) continue; /* * check if inp's faddr and laddr match with ours. * our faddr is considered null. * count the number of wildcard matches. (0 - 2) * * null null match * A null wildcard match * null B wildcard match * A B non match * A A match */ wildcard = 0; if (!in_nullhost(in4p_faddr(inp))) wildcard++; if (in_nullhost(in4p_laddr(inp))) { if (!in_nullhost(laddr)) wildcard++; } else { if (in_nullhost(laddr)) wildcard++; else { if (!in_hosteq(in4p_laddr(inp), laddr)) continue; } } if (wildcard && !lookup_wildcard) continue; /* * prefer an address with less wildcards. */ if (wildcard < matchwild) { match = inp; matchwild = wildcard; if (matchwild == 0) break; } } if (match && matchwild == 0) return match; if (vp && table->vestige) { void *state = (*table->vestige->init_ports4)(laddr, lport_arg, lookup_wildcard); vestigial_inpcb_t better; bool has_better = false; while (table->vestige && (*table->vestige->next_port4)(state, vp)) { if (vp->lport != lport) continue; wildcard = 0; if (!in_nullhost(vp->faddr.v4)) wildcard++; if (in_nullhost(vp->laddr.v4)) { if (!in_nullhost(laddr)) wildcard++; } else { if (in_nullhost(laddr)) wildcard++; else { if (!in_hosteq(vp->laddr.v4, laddr)) continue; } } if (wildcard && !lookup_wildcard) continue; if (wildcard < matchwild) { better = *vp; has_better = true; matchwild = wildcard; if (matchwild == 0) break; } } if (has_better) { *vp = better; return 0; } } return match; } #ifdef DIAGNOSTIC int inpcb_notifymiss = 0; #endif /* * inpcb_lookup: perform a full 4-tuple PCB lookup. */ struct inpcb * inpcb_lookup(struct inpcbtable *table, struct in_addr faddr, u_int fport_arg, struct in_addr laddr, u_int lport_arg, vestigial_inpcb_t *vp) { struct inpcbhead *head; struct inpcb *inp; in_port_t fport = fport_arg, lport = lport_arg; if (vp) vp->valid = 0; head = INPCBHASH_CONNECT(table, faddr, fport, laddr, lport); LIST_FOREACH(inp, head, inp_hash) { if (inp->inp_af != AF_INET) continue; if (in_hosteq(in4p_faddr(inp), faddr) && inp->inp_fport == fport && inp->inp_lport == lport && in_hosteq(in4p_laddr(inp), laddr)) goto out; } if (vp && table->vestige) { if ((*table->vestige->lookup4)(faddr, fport_arg, laddr, lport_arg, vp)) return 0; } #ifdef DIAGNOSTIC if (inpcb_notifymiss) { printf("inpcb_lookup: faddr=%08x fport=%d laddr=%08x lport=%d\n", ntohl(faddr.s_addr), ntohs(fport), ntohl(laddr.s_addr), ntohs(lport)); } #endif return 0; out: /* Move this PCB to the head of hash chain. */ if (inp != LIST_FIRST(head)) { LIST_REMOVE(inp, inp_hash); LIST_INSERT_HEAD(head, inp, inp_hash); } return inp; } /* * inpcb_lookup_bound: find a PCB by looking at the local address and port. * Primarily used to find the listening (i.e., already bound) socket. */ struct inpcb * inpcb_lookup_bound(struct inpcbtable *table, struct in_addr laddr, u_int lport_arg) { struct inpcbhead *head; struct inpcb *inp; in_port_t lport = lport_arg; head = INPCBHASH_BIND(table, laddr, lport); LIST_FOREACH(inp, head, inp_hash) { if (inp->inp_af != AF_INET) continue; if (inp->inp_lport == lport && in_hosteq(in4p_laddr(inp), laddr)) goto out; } head = INPCBHASH_BIND(table, zeroin_addr, lport); LIST_FOREACH(inp, head, inp_hash) { if (inp->inp_af != AF_INET) continue; if (inp->inp_lport == lport && in_hosteq(in4p_laddr(inp), zeroin_addr)) goto out; } #ifdef DIAGNOSTIC if (inpcb_notifymiss) { printf("inpcb_lookup_bound: laddr=%08x lport=%d\n", ntohl(laddr.s_addr), ntohs(lport)); } #endif return 0; out: /* Move this PCB to the head of hash chain. */ if (inp != LIST_FIRST(head)) { LIST_REMOVE(inp, inp_hash); LIST_INSERT_HEAD(head, inp, inp_hash); } return inp; } void inpcb_set_state(struct inpcb *inp, int state) { #ifdef INET6 if (inp->inp_af == AF_INET6) { in6pcb_set_state(inp, state); return; } #else if (inp->inp_af != AF_INET) return; #endif if (inp->inp_state > INP_ATTACHED) LIST_REMOVE(inp, inp_hash); switch (state) { case INP_BOUND: LIST_INSERT_HEAD(INPCBHASH_BIND(inp->inp_table, in4p_laddr(inp), inp->inp_lport), inp, inp_hash); break; case INP_CONNECTED: LIST_INSERT_HEAD(INPCBHASH_CONNECT(inp->inp_table, in4p_faddr(inp), inp->inp_fport, in4p_laddr(inp), inp->inp_lport), inp, inp_hash); break; } inp->inp_state = state; } struct rtentry * inpcb_rtentry(struct inpcb *inp) { struct route *ro; union { struct sockaddr dst; struct sockaddr_in dst4; } u; #ifdef INET6 if (inp->inp_af == AF_INET6) return in6pcb_rtentry(inp); #endif if (inp->inp_af != AF_INET) return NULL; ro = &inp->inp_route; sockaddr_in_init(&u.dst4, &in4p_faddr(inp), 0); return rtcache_lookup(ro, &u.dst); } void inpcb_rtentry_unref(struct rtentry *rt, struct inpcb *inp) { rtcache_unref(rt, &inp->inp_route); }