/* $NetBSD: if_bridge.c,v 1.46 2006/11/23 04:07:07 rpaulo Exp $ */ /* * Copyright 2001 Wasabi Systems, Inc. * All rights reserved. * * Written by Jason R. Thorpe for Wasabi Systems, Inc. * * 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 for the NetBSD Project by * Wasabi Systems, Inc. * 4. The name of Wasabi Systems, Inc. may not be used to endorse * or promote products derived from this software without specific prior * written permission. * * THIS SOFTWARE IS PROVIDED BY WASABI SYSTEMS, INC. ``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 WASABI SYSTEMS, INC * 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) 1999, 2000 Jason L. Wright (jason@thought.net) * 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. All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed by Jason L. Wright * 4. The name of the author may not be used to endorse or promote products * derived from this software without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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. * * OpenBSD: if_bridge.c,v 1.60 2001/06/15 03:38:33 itojun Exp */ /* * Network interface bridge support. * * TODO: * * - Currently only supports Ethernet-like interfaces (Ethernet, * 802.11, VLANs on Ethernet, etc.) Figure out a nice way * to bridge other types of interfaces (FDDI-FDDI, and maybe * consider heterogenous bridges). */ #include __KERNEL_RCSID(0, "$NetBSD: if_bridge.c,v 1.46 2006/11/23 04:07:07 rpaulo Exp $"); #include "opt_bridge_ipf.h" #include "opt_inet.h" #include "opt_pfil_hooks.h" #include "bpfilter.h" #include #include #include #include #include #include #include #include #include #include #if NBPFILTER > 0 #include #endif #include #include #include #include #include #include #if defined(BRIDGE_IPF) && defined(PFIL_HOOKS) /* Used for bridge_ip[6]_checkbasic */ #include #include #include #include #include #include #include #endif /* BRIDGE_IPF && PFIL_HOOKS */ /* * Size of the route hash table. Must be a power of two. */ #ifndef BRIDGE_RTHASH_SIZE #define BRIDGE_RTHASH_SIZE 1024 #endif #define BRIDGE_RTHASH_MASK (BRIDGE_RTHASH_SIZE - 1) #include "carp.h" #if NCARP > 0 #include #include #include #endif /* * Maximum number of addresses to cache. */ #ifndef BRIDGE_RTABLE_MAX #define BRIDGE_RTABLE_MAX 100 #endif /* * Spanning tree defaults. */ #define BSTP_DEFAULT_MAX_AGE (20 * 256) #define BSTP_DEFAULT_HELLO_TIME (2 * 256) #define BSTP_DEFAULT_FORWARD_DELAY (15 * 256) #define BSTP_DEFAULT_HOLD_TIME (1 * 256) #define BSTP_DEFAULT_BRIDGE_PRIORITY 0x8000 #define BSTP_DEFAULT_PORT_PRIORITY 0x80 #define BSTP_DEFAULT_PATH_COST 55 /* * Timeout (in seconds) for entries learned dynamically. */ #ifndef BRIDGE_RTABLE_TIMEOUT #define BRIDGE_RTABLE_TIMEOUT (20 * 60) /* same as ARP */ #endif /* * Number of seconds between walks of the route list. */ #ifndef BRIDGE_RTABLE_PRUNE_PERIOD #define BRIDGE_RTABLE_PRUNE_PERIOD (5 * 60) #endif int bridge_rtable_prune_period = BRIDGE_RTABLE_PRUNE_PERIOD; static struct pool bridge_rtnode_pool; void bridgeattach(int); static int bridge_clone_create(struct if_clone *, int); static int bridge_clone_destroy(struct ifnet *); static int bridge_ioctl(struct ifnet *, u_long, caddr_t); static int bridge_init(struct ifnet *); static void bridge_stop(struct ifnet *, int); static void bridge_start(struct ifnet *); static void bridge_forward(struct bridge_softc *, struct mbuf *m); static void bridge_timer(void *); static void bridge_broadcast(struct bridge_softc *, struct ifnet *, struct mbuf *); static int bridge_rtupdate(struct bridge_softc *, const uint8_t *, struct ifnet *, int, uint8_t); static struct ifnet *bridge_rtlookup(struct bridge_softc *, const uint8_t *); static void bridge_rttrim(struct bridge_softc *); static void bridge_rtage(struct bridge_softc *); static void bridge_rtflush(struct bridge_softc *, int); static int bridge_rtdaddr(struct bridge_softc *, const uint8_t *); static void bridge_rtdelete(struct bridge_softc *, struct ifnet *ifp); static int bridge_rtable_init(struct bridge_softc *); static void bridge_rtable_fini(struct bridge_softc *); static struct bridge_rtnode *bridge_rtnode_lookup(struct bridge_softc *, const uint8_t *); static int bridge_rtnode_insert(struct bridge_softc *, struct bridge_rtnode *); static void bridge_rtnode_destroy(struct bridge_softc *, struct bridge_rtnode *); static struct bridge_iflist *bridge_lookup_member(struct bridge_softc *, const char *name); static struct bridge_iflist *bridge_lookup_member_if(struct bridge_softc *, struct ifnet *ifp); static void bridge_delete_member(struct bridge_softc *, struct bridge_iflist *); static int bridge_ioctl_add(struct bridge_softc *, void *); static int bridge_ioctl_del(struct bridge_softc *, void *); static int bridge_ioctl_gifflags(struct bridge_softc *, void *); static int bridge_ioctl_sifflags(struct bridge_softc *, void *); static int bridge_ioctl_scache(struct bridge_softc *, void *); static int bridge_ioctl_gcache(struct bridge_softc *, void *); static int bridge_ioctl_gifs(struct bridge_softc *, void *); static int bridge_ioctl_rts(struct bridge_softc *, void *); static int bridge_ioctl_saddr(struct bridge_softc *, void *); static int bridge_ioctl_sto(struct bridge_softc *, void *); static int bridge_ioctl_gto(struct bridge_softc *, void *); static int bridge_ioctl_daddr(struct bridge_softc *, void *); static int bridge_ioctl_flush(struct bridge_softc *, void *); static int bridge_ioctl_gpri(struct bridge_softc *, void *); static int bridge_ioctl_spri(struct bridge_softc *, void *); static int bridge_ioctl_ght(struct bridge_softc *, void *); static int bridge_ioctl_sht(struct bridge_softc *, void *); static int bridge_ioctl_gfd(struct bridge_softc *, void *); static int bridge_ioctl_sfd(struct bridge_softc *, void *); static int bridge_ioctl_gma(struct bridge_softc *, void *); static int bridge_ioctl_sma(struct bridge_softc *, void *); static int bridge_ioctl_sifprio(struct bridge_softc *, void *); static int bridge_ioctl_sifcost(struct bridge_softc *, void *); #if defined(BRIDGE_IPF) && defined(PFIL_HOOKS) static int bridge_ioctl_gfilt(struct bridge_softc *, void *); static int bridge_ioctl_sfilt(struct bridge_softc *, void *); static int bridge_ipf(void *, struct mbuf **, struct ifnet *, int); static int bridge_ip_checkbasic(struct mbuf **mp); # ifdef INET6 static int bridge_ip6_checkbasic(struct mbuf **mp); # endif /* INET6 */ #endif /* BRIDGE_IPF && PFIL_HOOKS */ struct bridge_control { int (*bc_func)(struct bridge_softc *, void *); int bc_argsize; int bc_flags; }; #define BC_F_COPYIN 0x01 /* copy arguments in */ #define BC_F_COPYOUT 0x02 /* copy arguments out */ #define BC_F_SUSER 0x04 /* do super-user check */ static const struct bridge_control bridge_control_table[] = { { bridge_ioctl_add, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_del, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gifflags, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_COPYOUT }, { bridge_ioctl_sifflags, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_scache, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gcache, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_gifs, sizeof(struct ifbifconf), BC_F_COPYIN|BC_F_COPYOUT }, { bridge_ioctl_rts, sizeof(struct ifbaconf), BC_F_COPYIN|BC_F_COPYOUT }, { bridge_ioctl_saddr, sizeof(struct ifbareq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_sto, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gto, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_daddr, sizeof(struct ifbareq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_flush, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gpri, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_spri, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_ght, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_sht, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gfd, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_sfd, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_gma, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_sma, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_sifprio, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, { bridge_ioctl_sifcost, sizeof(struct ifbreq), BC_F_COPYIN|BC_F_SUSER }, #if defined(BRIDGE_IPF) && defined(PFIL_HOOKS) { bridge_ioctl_gfilt, sizeof(struct ifbrparam), BC_F_COPYOUT }, { bridge_ioctl_sfilt, sizeof(struct ifbrparam), BC_F_COPYIN|BC_F_SUSER }, #endif /* BRIDGE_IPF && PFIL_HOOKS */ }; static const int bridge_control_table_size = sizeof(bridge_control_table) / sizeof(bridge_control_table[0]); static LIST_HEAD(, bridge_softc) bridge_list; static struct if_clone bridge_cloner = IF_CLONE_INITIALIZER("bridge", bridge_clone_create, bridge_clone_destroy); /* * bridgeattach: * * Pseudo-device attach routine. */ void bridgeattach(int n) { pool_init(&bridge_rtnode_pool, sizeof(struct bridge_rtnode), 0, 0, 0, "brtpl", NULL); LIST_INIT(&bridge_list); if_clone_attach(&bridge_cloner); } /* * bridge_clone_create: * * Create a new bridge instance. */ static int bridge_clone_create(struct if_clone *ifc, int unit) { struct bridge_softc *sc; struct ifnet *ifp; int s; sc = malloc(sizeof(*sc), M_DEVBUF, M_WAITOK); memset(sc, 0, sizeof(*sc)); ifp = &sc->sc_if; sc->sc_brtmax = BRIDGE_RTABLE_MAX; sc->sc_brttimeout = BRIDGE_RTABLE_TIMEOUT; sc->sc_bridge_max_age = BSTP_DEFAULT_MAX_AGE; sc->sc_bridge_hello_time = BSTP_DEFAULT_HELLO_TIME; sc->sc_bridge_forward_delay = BSTP_DEFAULT_FORWARD_DELAY; sc->sc_bridge_priority = BSTP_DEFAULT_BRIDGE_PRIORITY; sc->sc_hold_time = BSTP_DEFAULT_HOLD_TIME; sc->sc_filter_flags = 0; /* Initialize our routing table. */ bridge_rtable_init(sc); callout_init(&sc->sc_brcallout); callout_init(&sc->sc_bstpcallout); LIST_INIT(&sc->sc_iflist); snprintf(ifp->if_xname, sizeof(ifp->if_xname), "%s%d", ifc->ifc_name, unit); ifp->if_softc = sc; ifp->if_mtu = ETHERMTU; ifp->if_ioctl = bridge_ioctl; ifp->if_output = bridge_output; ifp->if_start = bridge_start; ifp->if_stop = bridge_stop; ifp->if_init = bridge_init; ifp->if_type = IFT_BRIDGE; ifp->if_addrlen = 0; ifp->if_dlt = DLT_EN10MB; ifp->if_hdrlen = ETHER_HDR_LEN; if_attach(ifp); if_alloc_sadl(ifp); s = splnet(); LIST_INSERT_HEAD(&bridge_list, sc, sc_list); splx(s); return (0); } /* * bridge_clone_destroy: * * Destroy a bridge instance. */ static int bridge_clone_destroy(struct ifnet *ifp) { struct bridge_softc *sc = ifp->if_softc; struct bridge_iflist *bif; int s; s = splnet(); bridge_stop(ifp, 1); while ((bif = LIST_FIRST(&sc->sc_iflist)) != NULL) bridge_delete_member(sc, bif); LIST_REMOVE(sc, sc_list); splx(s); if_detach(ifp); /* Tear down the routing table. */ bridge_rtable_fini(sc); free(sc, M_DEVBUF); return (0); } /* * bridge_ioctl: * * Handle a control request from the operator. */ static int bridge_ioctl(struct ifnet *ifp, u_long cmd, caddr_t data) { struct bridge_softc *sc = ifp->if_softc; struct lwp *l = curlwp; /* XXX */ union { struct ifbreq ifbreq; struct ifbifconf ifbifconf; struct ifbareq ifbareq; struct ifbaconf ifbaconf; struct ifbrparam ifbrparam; } args; struct ifdrv *ifd = (struct ifdrv *) data; const struct bridge_control *bc; int s, error = 0; s = splnet(); switch (cmd) { case SIOCGDRVSPEC: case SIOCSDRVSPEC: if (ifd->ifd_cmd >= bridge_control_table_size) { error = EINVAL; break; } bc = &bridge_control_table[ifd->ifd_cmd]; if (cmd == SIOCGDRVSPEC && (bc->bc_flags & BC_F_COPYOUT) == 0) { error = EINVAL; break; } else if (cmd == SIOCSDRVSPEC && (bc->bc_flags & BC_F_COPYOUT) != 0) { error = EINVAL; break; } if (bc->bc_flags & BC_F_SUSER) { error = kauth_authorize_generic(l->l_cred, KAUTH_GENERIC_ISSUSER, &l->l_acflag); if (error) break; } if (ifd->ifd_len != bc->bc_argsize || ifd->ifd_len > sizeof(args)) { error = EINVAL; break; } memset(&args, 0, sizeof(args)); if (bc->bc_flags & BC_F_COPYIN) { error = copyin(ifd->ifd_data, &args, ifd->ifd_len); if (error) break; } error = (*bc->bc_func)(sc, &args); if (error) break; if (bc->bc_flags & BC_F_COPYOUT) error = copyout(&args, ifd->ifd_data, ifd->ifd_len); break; case SIOCSIFFLAGS: if ((ifp->if_flags & (IFF_UP|IFF_RUNNING)) == IFF_RUNNING) { /* * If interface is marked down and it is running, * then stop and disable it. */ (*ifp->if_stop)(ifp, 1); } else if ((ifp->if_flags & (IFF_UP|IFF_RUNNING)) == IFF_UP) { /* * If interface is marked up and it is stopped, then * start it. */ error = (*ifp->if_init)(ifp); } break; default: error = ENOTTY; break; } splx(s); return (error); } /* * bridge_lookup_member: * * Lookup a bridge member interface. Must be called at splnet(). */ static struct bridge_iflist * bridge_lookup_member(struct bridge_softc *sc, const char *name) { struct bridge_iflist *bif; struct ifnet *ifp; LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { ifp = bif->bif_ifp; if (strcmp(ifp->if_xname, name) == 0) return (bif); } return (NULL); } /* * bridge_lookup_member_if: * * Lookup a bridge member interface by ifnet*. Must be called at splnet(). */ static struct bridge_iflist * bridge_lookup_member_if(struct bridge_softc *sc, struct ifnet *member_ifp) { struct bridge_iflist *bif; LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { if (bif->bif_ifp == member_ifp) return (bif); } return (NULL); } /* * bridge_delete_member: * * Delete the specified member interface. */ static void bridge_delete_member(struct bridge_softc *sc, struct bridge_iflist *bif) { struct ifnet *ifs = bif->bif_ifp; switch (ifs->if_type) { case IFT_ETHER: /* * Take the interface out of promiscuous mode. */ (void) ifpromisc(ifs, 0); break; default: #ifdef DIAGNOSTIC panic("bridge_delete_member: impossible"); #endif break; } ifs->if_bridge = NULL; LIST_REMOVE(bif, bif_next); bridge_rtdelete(sc, ifs); free(bif, M_DEVBUF); if (sc->sc_if.if_flags & IFF_RUNNING) bstp_initialization(sc); } static int bridge_ioctl_add(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif = NULL; struct ifnet *ifs; int error = 0; ifs = ifunit(req->ifbr_ifsname); if (ifs == NULL) return (ENOENT); if (sc->sc_if.if_mtu != ifs->if_mtu) return (EINVAL); if (ifs->if_bridge == sc) return (EEXIST); if (ifs->if_bridge != NULL) return (EBUSY); bif = malloc(sizeof(*bif), M_DEVBUF, M_NOWAIT); if (bif == NULL) return (ENOMEM); switch (ifs->if_type) { case IFT_ETHER: /* * Place the interface into promiscuous mode. */ error = ifpromisc(ifs, 1); if (error) goto out; break; default: error = EINVAL; goto out; } bif->bif_ifp = ifs; bif->bif_flags = IFBIF_LEARNING | IFBIF_DISCOVER; bif->bif_priority = BSTP_DEFAULT_PORT_PRIORITY; bif->bif_path_cost = BSTP_DEFAULT_PATH_COST; ifs->if_bridge = sc; LIST_INSERT_HEAD(&sc->sc_iflist, bif, bif_next); if (sc->sc_if.if_flags & IFF_RUNNING) bstp_initialization(sc); else bstp_stop(sc); out: if (error) { if (bif != NULL) free(bif, M_DEVBUF); } return (error); } static int bridge_ioctl_del(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif; bif = bridge_lookup_member(sc, req->ifbr_ifsname); if (bif == NULL) return (ENOENT); bridge_delete_member(sc, bif); return (0); } static int bridge_ioctl_gifflags(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif; bif = bridge_lookup_member(sc, req->ifbr_ifsname); if (bif == NULL) return (ENOENT); req->ifbr_ifsflags = bif->bif_flags; req->ifbr_state = bif->bif_state; req->ifbr_priority = bif->bif_priority; req->ifbr_path_cost = bif->bif_path_cost; req->ifbr_portno = bif->bif_ifp->if_index & 0xff; return (0); } static int bridge_ioctl_sifflags(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif; bif = bridge_lookup_member(sc, req->ifbr_ifsname); if (bif == NULL) return (ENOENT); if (req->ifbr_ifsflags & IFBIF_STP) { switch (bif->bif_ifp->if_type) { case IFT_ETHER: /* These can do spanning tree. */ break; default: /* Nothing else can. */ return (EINVAL); } } bif->bif_flags = req->ifbr_ifsflags; if (sc->sc_if.if_flags & IFF_RUNNING) bstp_initialization(sc); return (0); } static int bridge_ioctl_scache(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; sc->sc_brtmax = param->ifbrp_csize; bridge_rttrim(sc); return (0); } static int bridge_ioctl_gcache(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; param->ifbrp_csize = sc->sc_brtmax; return (0); } static int bridge_ioctl_gifs(struct bridge_softc *sc, void *arg) { struct ifbifconf *bifc = arg; struct bridge_iflist *bif; struct ifbreq breq; int count, len, error = 0; count = 0; LIST_FOREACH(bif, &sc->sc_iflist, bif_next) count++; if (bifc->ifbic_len == 0) { bifc->ifbic_len = sizeof(breq) * count; return (0); } count = 0; len = bifc->ifbic_len; memset(&breq, 0, sizeof breq); LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { if (len < sizeof(breq)) break; strlcpy(breq.ifbr_ifsname, bif->bif_ifp->if_xname, sizeof(breq.ifbr_ifsname)); breq.ifbr_ifsflags = bif->bif_flags; breq.ifbr_state = bif->bif_state; breq.ifbr_priority = bif->bif_priority; breq.ifbr_path_cost = bif->bif_path_cost; breq.ifbr_portno = bif->bif_ifp->if_index & 0xff; error = copyout(&breq, bifc->ifbic_req + count, sizeof(breq)); if (error) break; count++; len -= sizeof(breq); } bifc->ifbic_len = sizeof(breq) * count; return (error); } static int bridge_ioctl_rts(struct bridge_softc *sc, void *arg) { struct ifbaconf *bac = arg; struct bridge_rtnode *brt; struct ifbareq bareq; int count = 0, error = 0, len; if (bac->ifbac_len == 0) return (0); len = bac->ifbac_len; LIST_FOREACH(brt, &sc->sc_rtlist, brt_list) { if (len < sizeof(bareq)) goto out; memset(&bareq, 0, sizeof(bareq)); strlcpy(bareq.ifba_ifsname, brt->brt_ifp->if_xname, sizeof(bareq.ifba_ifsname)); memcpy(bareq.ifba_dst, brt->brt_addr, sizeof(brt->brt_addr)); if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) { bareq.ifba_expire = brt->brt_expire - time_uptime; } else bareq.ifba_expire = 0; bareq.ifba_flags = brt->brt_flags; error = copyout(&bareq, bac->ifbac_req + count, sizeof(bareq)); if (error) goto out; count++; len -= sizeof(bareq); } out: bac->ifbac_len = sizeof(bareq) * count; return (error); } static int bridge_ioctl_saddr(struct bridge_softc *sc, void *arg) { struct ifbareq *req = arg; struct bridge_iflist *bif; int error; bif = bridge_lookup_member(sc, req->ifba_ifsname); if (bif == NULL) return (ENOENT); error = bridge_rtupdate(sc, req->ifba_dst, bif->bif_ifp, 1, req->ifba_flags); return (error); } static int bridge_ioctl_sto(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; sc->sc_brttimeout = param->ifbrp_ctime; return (0); } static int bridge_ioctl_gto(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; param->ifbrp_ctime = sc->sc_brttimeout; return (0); } static int bridge_ioctl_daddr(struct bridge_softc *sc, void *arg) { struct ifbareq *req = arg; return (bridge_rtdaddr(sc, req->ifba_dst)); } static int bridge_ioctl_flush(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; bridge_rtflush(sc, req->ifbr_ifsflags); return (0); } static int bridge_ioctl_gpri(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; param->ifbrp_prio = sc->sc_bridge_priority; return (0); } static int bridge_ioctl_spri(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; sc->sc_bridge_priority = param->ifbrp_prio; if (sc->sc_if.if_flags & IFF_RUNNING) bstp_initialization(sc); return (0); } static int bridge_ioctl_ght(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; param->ifbrp_hellotime = sc->sc_bridge_hello_time >> 8; return (0); } static int bridge_ioctl_sht(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; if (param->ifbrp_hellotime == 0) return (EINVAL); sc->sc_bridge_hello_time = param->ifbrp_hellotime << 8; if (sc->sc_if.if_flags & IFF_RUNNING) bstp_initialization(sc); return (0); } static int bridge_ioctl_gfd(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; param->ifbrp_fwddelay = sc->sc_bridge_forward_delay >> 8; return (0); } static int bridge_ioctl_sfd(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; if (param->ifbrp_fwddelay == 0) return (EINVAL); sc->sc_bridge_forward_delay = param->ifbrp_fwddelay << 8; if (sc->sc_if.if_flags & IFF_RUNNING) bstp_initialization(sc); return (0); } static int bridge_ioctl_gma(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; param->ifbrp_maxage = sc->sc_bridge_max_age >> 8; return (0); } static int bridge_ioctl_sma(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; if (param->ifbrp_maxage == 0) return (EINVAL); sc->sc_bridge_max_age = param->ifbrp_maxage << 8; if (sc->sc_if.if_flags & IFF_RUNNING) bstp_initialization(sc); return (0); } static int bridge_ioctl_sifprio(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif; bif = bridge_lookup_member(sc, req->ifbr_ifsname); if (bif == NULL) return (ENOENT); bif->bif_priority = req->ifbr_priority; if (sc->sc_if.if_flags & IFF_RUNNING) bstp_initialization(sc); return (0); } #if defined(BRIDGE_IPF) && defined(PFIL_HOOKS) static int bridge_ioctl_gfilt(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; param->ifbrp_filter = sc->sc_filter_flags; return (0); } static int bridge_ioctl_sfilt(struct bridge_softc *sc, void *arg) { struct ifbrparam *param = arg; uint32_t nflags, oflags; if (param->ifbrp_filter & ~IFBF_FILT_MASK) return (EINVAL); nflags = param->ifbrp_filter; oflags = sc->sc_filter_flags; if ((nflags & IFBF_FILT_USEIPF) && !(oflags & IFBF_FILT_USEIPF)) { pfil_add_hook((void *)bridge_ipf, NULL, PFIL_IN|PFIL_OUT, &sc->sc_if.if_pfil); } if (!(nflags & IFBF_FILT_USEIPF) && (oflags & IFBF_FILT_USEIPF)) { pfil_remove_hook((void *)bridge_ipf, NULL, PFIL_IN|PFIL_OUT, &sc->sc_if.if_pfil); } sc->sc_filter_flags = nflags; return (0); } #endif /* BRIDGE_IPF && PFIL_HOOKS */ static int bridge_ioctl_sifcost(struct bridge_softc *sc, void *arg) { struct ifbreq *req = arg; struct bridge_iflist *bif; bif = bridge_lookup_member(sc, req->ifbr_ifsname); if (bif == NULL) return (ENOENT); bif->bif_path_cost = req->ifbr_path_cost; if (sc->sc_if.if_flags & IFF_RUNNING) bstp_initialization(sc); return (0); } /* * bridge_ifdetach: * * Detach an interface from a bridge. Called when a member * interface is detaching. */ void bridge_ifdetach(struct ifnet *ifp) { struct bridge_softc *sc = ifp->if_bridge; struct ifbreq breq; memset(&breq, 0, sizeof(breq)); snprintf(breq.ifbr_ifsname, sizeof(breq.ifbr_ifsname), ifp->if_xname); (void) bridge_ioctl_del(sc, &breq); } /* * bridge_init: * * Initialize a bridge interface. */ static int bridge_init(struct ifnet *ifp) { struct bridge_softc *sc = ifp->if_softc; if (ifp->if_flags & IFF_RUNNING) return (0); callout_reset(&sc->sc_brcallout, bridge_rtable_prune_period * hz, bridge_timer, sc); ifp->if_flags |= IFF_RUNNING; bstp_initialization(sc); return (0); } /* * bridge_stop: * * Stop the bridge interface. */ static void bridge_stop(struct ifnet *ifp, int disable) { struct bridge_softc *sc = ifp->if_softc; if ((ifp->if_flags & IFF_RUNNING) == 0) return; callout_stop(&sc->sc_brcallout); bstp_stop(sc); IF_PURGE(&ifp->if_snd); bridge_rtflush(sc, IFBF_FLUSHDYN); ifp->if_flags &= ~IFF_RUNNING; } /* * bridge_enqueue: * * Enqueue a packet on a bridge member interface. * * NOTE: must be called at splnet(). */ void bridge_enqueue(struct bridge_softc *sc, struct ifnet *dst_ifp, struct mbuf *m, int runfilt) { ALTQ_DECL(struct altq_pktattr pktattr;) int len, error; short mflags; /* * Clear any in-bound checksum flags for this packet. */ m->m_pkthdr.csum_flags = 0; #ifdef PFIL_HOOKS if (runfilt) { if (pfil_run_hooks(&sc->sc_if.if_pfil, &m, dst_ifp, PFIL_OUT) != 0) { if (m != NULL) m_freem(m); return; } if (m == NULL) return; } #endif /* PFIL_HOOKS */ #ifdef ALTQ /* * If ALTQ is enabled on the member interface, do * classification; the queueing discipline might * not require classification, but might require * the address family/header pointer in the pktattr. */ if (ALTQ_IS_ENABLED(&dst_ifp->if_snd)) { /* XXX IFT_ETHER */ altq_etherclassify(&dst_ifp->if_snd, m, &pktattr); } #endif /* ALTQ */ len = m->m_pkthdr.len; m->m_flags |= M_PROTO1; mflags = m->m_flags; IFQ_ENQUEUE(&dst_ifp->if_snd, m, &pktattr, error); if (error) { /* mbuf is already freed */ sc->sc_if.if_oerrors++; return; } sc->sc_if.if_opackets++; sc->sc_if.if_obytes += len; dst_ifp->if_obytes += len; if (mflags & M_MCAST) { sc->sc_if.if_omcasts++; dst_ifp->if_omcasts++; } if ((dst_ifp->if_flags & IFF_OACTIVE) == 0) (*dst_ifp->if_start)(dst_ifp); } /* * bridge_output: * * Send output from a bridge member interface. This * performs the bridging function for locally originated * packets. * * The mbuf has the Ethernet header already attached. We must * enqueue or free the mbuf before returning. */ int bridge_output(struct ifnet *ifp, struct mbuf *m, struct sockaddr *sa, struct rtentry *rt) { struct ether_header *eh; struct ifnet *dst_if; struct bridge_softc *sc; int s; if (m->m_len < ETHER_HDR_LEN) { m = m_pullup(m, ETHER_HDR_LEN); if (m == NULL) return (0); } eh = mtod(m, struct ether_header *); sc = ifp->if_bridge; s = splnet(); /* * If bridge is down, but the original output interface is up, * go ahead and send out that interface. Otherwise, the packet * is dropped below. */ if ((sc->sc_if.if_flags & IFF_RUNNING) == 0) { dst_if = ifp; goto sendunicast; } /* * If the packet is a multicast, or we don't know a better way to * get there, send to all interfaces. */ if (ETHER_IS_MULTICAST(eh->ether_dhost)) dst_if = NULL; else dst_if = bridge_rtlookup(sc, eh->ether_dhost); if (dst_if == NULL) { struct bridge_iflist *bif; struct mbuf *mc; int used = 0; LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { dst_if = bif->bif_ifp; if ((dst_if->if_flags & IFF_RUNNING) == 0) continue; /* * If this is not the original output interface, * and the interface is participating in spanning * tree, make sure the port is in a state that * allows forwarding. */ if (dst_if != ifp && (bif->bif_flags & IFBIF_STP) != 0) { switch (bif->bif_state) { case BSTP_IFSTATE_BLOCKING: case BSTP_IFSTATE_LISTENING: case BSTP_IFSTATE_DISABLED: continue; } } if (LIST_NEXT(bif, bif_next) == NULL) { used = 1; mc = m; } else { mc = m_copym(m, 0, M_COPYALL, M_NOWAIT); if (mc == NULL) { sc->sc_if.if_oerrors++; continue; } } bridge_enqueue(sc, dst_if, mc, 0); } if (used == 0) m_freem(m); splx(s); return (0); } sendunicast: /* * XXX Spanning tree consideration here? */ if ((dst_if->if_flags & IFF_RUNNING) == 0) { m_freem(m); splx(s); return (0); } bridge_enqueue(sc, dst_if, m, 0); splx(s); return (0); } /* * bridge_start: * * Start output on a bridge. * * NOTE: This routine should never be called in this implementation. */ static void bridge_start(struct ifnet *ifp) { printf("%s: bridge_start() called\n", ifp->if_xname); } /* * bridge_forward: * * The forwarding function of the bridge. */ static void bridge_forward(struct bridge_softc *sc, struct mbuf *m) { struct bridge_iflist *bif; struct ifnet *src_if, *dst_if; struct ether_header *eh; src_if = m->m_pkthdr.rcvif; sc->sc_if.if_ipackets++; sc->sc_if.if_ibytes += m->m_pkthdr.len; /* * Look up the bridge_iflist. */ bif = bridge_lookup_member_if(sc, src_if); if (bif == NULL) { /* Interface is not a bridge member (anymore?) */ m_freem(m); return; } if (bif->bif_flags & IFBIF_STP) { switch (bif->bif_state) { case BSTP_IFSTATE_BLOCKING: case BSTP_IFSTATE_LISTENING: case BSTP_IFSTATE_DISABLED: m_freem(m); return; } } eh = mtod(m, struct ether_header *); /* * If the interface is learning, and the source * address is valid and not multicast, record * the address. */ if ((bif->bif_flags & IFBIF_LEARNING) != 0 && ETHER_IS_MULTICAST(eh->ether_shost) == 0 && (eh->ether_shost[0] == 0 && eh->ether_shost[1] == 0 && eh->ether_shost[2] == 0 && eh->ether_shost[3] == 0 && eh->ether_shost[4] == 0 && eh->ether_shost[5] == 0) == 0) { (void) bridge_rtupdate(sc, eh->ether_shost, src_if, 0, IFBAF_DYNAMIC); } if ((bif->bif_flags & IFBIF_STP) != 0 && bif->bif_state == BSTP_IFSTATE_LEARNING) { m_freem(m); return; } /* * At this point, the port either doesn't participate * in spanning tree or it is in the forwarding state. */ /* * If the packet is unicast, destined for someone on * "this" side of the bridge, drop it. */ if ((m->m_flags & (M_BCAST|M_MCAST)) == 0) { dst_if = bridge_rtlookup(sc, eh->ether_dhost); if (src_if == dst_if) { m_freem(m); return; } } else { /* ...forward it to all interfaces. */ sc->sc_if.if_imcasts++; dst_if = NULL; } #ifdef PFIL_HOOKS if (pfil_run_hooks(&sc->sc_if.if_pfil, &m, m->m_pkthdr.rcvif, PFIL_IN) != 0) { if (m != NULL) m_freem(m); return; } if (m == NULL) return; #endif /* PFIL_HOOKS */ if (dst_if == NULL) { bridge_broadcast(sc, src_if, m); return; } /* * At this point, we're dealing with a unicast frame * going to a different interface. */ if ((dst_if->if_flags & IFF_RUNNING) == 0) { m_freem(m); return; } bif = bridge_lookup_member_if(sc, dst_if); if (bif == NULL) { /* Not a member of the bridge (anymore?) */ m_freem(m); return; } if (bif->bif_flags & IFBIF_STP) { switch (bif->bif_state) { case BSTP_IFSTATE_DISABLED: case BSTP_IFSTATE_BLOCKING: m_freem(m); return; } } bridge_enqueue(sc, dst_if, m, 1); } /* * bridge_input: * * Receive input from a member interface. Queue the packet for * bridging if it is not for us. */ struct mbuf * bridge_input(struct ifnet *ifp, struct mbuf *m) { struct bridge_softc *sc = ifp->if_bridge; struct bridge_iflist *bif; struct ether_header *eh; struct mbuf *mc; if ((sc->sc_if.if_flags & IFF_RUNNING) == 0) return (m); bif = bridge_lookup_member_if(sc, ifp); if (bif == NULL) return (m); eh = mtod(m, struct ether_header *); if (m->m_flags & (M_BCAST|M_MCAST)) { /* Tap off 802.1D packets; they do not get forwarded. */ if (memcmp(eh->ether_dhost, bstp_etheraddr, ETHER_ADDR_LEN) == 0) { m = bstp_input(ifp, m); if (m == NULL) return (NULL); } if (bif->bif_flags & IFBIF_STP) { switch (bif->bif_state) { case BSTP_IFSTATE_BLOCKING: case BSTP_IFSTATE_LISTENING: case BSTP_IFSTATE_DISABLED: return (m); } } /* * Make a deep copy of the packet and enqueue the copy * for bridge processing; return the original packet for * local processing. */ mc = m_dup(m, 0, M_COPYALL, M_NOWAIT); if (mc == NULL) return (m); /* Perform the bridge forwarding function with the copy. */ bridge_forward(sc, mc); /* Return the original packet for local processing. */ return (m); } if (bif->bif_flags & IFBIF_STP) { switch (bif->bif_state) { case BSTP_IFSTATE_BLOCKING: case BSTP_IFSTATE_LISTENING: case BSTP_IFSTATE_DISABLED: return (m); } } /* * Unicast. Make sure it's not for us. */ LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { /* It is destined for us. */ if (memcmp(LLADDR(bif->bif_ifp->if_sadl), eh->ether_dhost, ETHER_ADDR_LEN) == 0 #if NCARP > 0 || (bif->bif_ifp->if_carp && carp_ourether(bif->bif_ifp->if_carp, eh, IFT_ETHER, 0) != NULL) #endif /* NCARP > 0 */ ) { if (bif->bif_flags & IFBIF_LEARNING) (void) bridge_rtupdate(sc, eh->ether_shost, ifp, 0, IFBAF_DYNAMIC); m->m_pkthdr.rcvif = bif->bif_ifp; return (m); } /* We just received a packet that we sent out. */ if (memcmp(LLADDR(bif->bif_ifp->if_sadl), eh->ether_shost, ETHER_ADDR_LEN) == 0 #if NCARP > 0 || (bif->bif_ifp->if_carp && carp_ourether(bif->bif_ifp->if_carp, eh, IFT_ETHER, 1) != NULL) #endif /* NCARP > 0 */ ) { m_freem(m); return (NULL); } } /* Perform the bridge forwarding function. */ bridge_forward(sc, m); return (NULL); } /* * bridge_broadcast: * * Send a frame to all interfaces that are members of * the bridge, except for the one on which the packet * arrived. */ static void bridge_broadcast(struct bridge_softc *sc, struct ifnet *src_if, struct mbuf *m) { struct bridge_iflist *bif; struct mbuf *mc; struct ifnet *dst_if; int used = 0; LIST_FOREACH(bif, &sc->sc_iflist, bif_next) { dst_if = bif->bif_ifp; if (dst_if == src_if) continue; if (bif->bif_flags & IFBIF_STP) { switch (bif->bif_state) { case BSTP_IFSTATE_BLOCKING: case BSTP_IFSTATE_DISABLED: continue; } } if ((bif->bif_flags & IFBIF_DISCOVER) == 0 && (m->m_flags & (M_BCAST|M_MCAST)) == 0) continue; if ((dst_if->if_flags & IFF_RUNNING) == 0) continue; if (LIST_NEXT(bif, bif_next) == NULL) { mc = m; used = 1; } else { mc = m_copym(m, 0, M_COPYALL, M_DONTWAIT); if (mc == NULL) { sc->sc_if.if_oerrors++; continue; } } bridge_enqueue(sc, dst_if, mc, 1); } if (used == 0) m_freem(m); } /* * bridge_rtupdate: * * Add a bridge routing entry. */ static int bridge_rtupdate(struct bridge_softc *sc, const uint8_t *dst, struct ifnet *dst_if, int setflags, uint8_t flags) { struct bridge_rtnode *brt; int error, s; /* * A route for this destination might already exist. If so, * update it, otherwise create a new one. */ if ((brt = bridge_rtnode_lookup(sc, dst)) == NULL) { if (sc->sc_brtcnt >= sc->sc_brtmax) return (ENOSPC); /* * Allocate a new bridge forwarding node, and * initialize the expiration time and Ethernet * address. */ s = splnet(); brt = pool_get(&bridge_rtnode_pool, PR_NOWAIT); splx(s); if (brt == NULL) return (ENOMEM); memset(brt, 0, sizeof(*brt)); brt->brt_expire = time_uptime + sc->sc_brttimeout; brt->brt_flags = IFBAF_DYNAMIC; memcpy(brt->brt_addr, dst, ETHER_ADDR_LEN); if ((error = bridge_rtnode_insert(sc, brt)) != 0) { s = splnet(); pool_put(&bridge_rtnode_pool, brt); splx(s); return (error); } } brt->brt_ifp = dst_if; if (setflags) { brt->brt_flags = flags; if (flags & IFBAF_STATIC) brt->brt_expire = 0; else brt->brt_expire = time_uptime + sc->sc_brttimeout; } return (0); } /* * bridge_rtlookup: * * Lookup the destination interface for an address. */ static struct ifnet * bridge_rtlookup(struct bridge_softc *sc, const uint8_t *addr) { struct bridge_rtnode *brt; if ((brt = bridge_rtnode_lookup(sc, addr)) == NULL) return (NULL); return (brt->brt_ifp); } /* * bridge_rttrim: * * Trim the routine table so that we have a number * of routing entries less than or equal to the * maximum number. */ static void bridge_rttrim(struct bridge_softc *sc) { struct bridge_rtnode *brt, *nbrt; /* Make sure we actually need to do this. */ if (sc->sc_brtcnt <= sc->sc_brtmax) return; /* Force an aging cycle; this might trim enough addresses. */ bridge_rtage(sc); if (sc->sc_brtcnt <= sc->sc_brtmax) return; for (brt = LIST_FIRST(&sc->sc_rtlist); brt != NULL; brt = nbrt) { nbrt = LIST_NEXT(brt, brt_list); if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) { bridge_rtnode_destroy(sc, brt); if (sc->sc_brtcnt <= sc->sc_brtmax) return; } } } /* * bridge_timer: * * Aging timer for the bridge. */ static void bridge_timer(void *arg) { struct bridge_softc *sc = arg; int s; s = splnet(); bridge_rtage(sc); splx(s); if (sc->sc_if.if_flags & IFF_RUNNING) callout_reset(&sc->sc_brcallout, bridge_rtable_prune_period * hz, bridge_timer, sc); } /* * bridge_rtage: * * Perform an aging cycle. */ static void bridge_rtage(struct bridge_softc *sc) { struct bridge_rtnode *brt, *nbrt; for (brt = LIST_FIRST(&sc->sc_rtlist); brt != NULL; brt = nbrt) { nbrt = LIST_NEXT(brt, brt_list); if ((brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) { if (time_uptime >= brt->brt_expire) bridge_rtnode_destroy(sc, brt); } } } /* * bridge_rtflush: * * Remove all dynamic addresses from the bridge. */ static void bridge_rtflush(struct bridge_softc *sc, int full) { struct bridge_rtnode *brt, *nbrt; for (brt = LIST_FIRST(&sc->sc_rtlist); brt != NULL; brt = nbrt) { nbrt = LIST_NEXT(brt, brt_list); if (full || (brt->brt_flags & IFBAF_TYPEMASK) == IFBAF_DYNAMIC) bridge_rtnode_destroy(sc, brt); } } /* * bridge_rtdaddr: * * Remove an address from the table. */ static int bridge_rtdaddr(struct bridge_softc *sc, const uint8_t *addr) { struct bridge_rtnode *brt; if ((brt = bridge_rtnode_lookup(sc, addr)) == NULL) return (ENOENT); bridge_rtnode_destroy(sc, brt); return (0); } /* * bridge_rtdelete: * * Delete routes to a speicifc member interface. */ static void bridge_rtdelete(struct bridge_softc *sc, struct ifnet *ifp) { struct bridge_rtnode *brt, *nbrt; for (brt = LIST_FIRST(&sc->sc_rtlist); brt != NULL; brt = nbrt) { nbrt = LIST_NEXT(brt, brt_list); if (brt->brt_ifp == ifp) bridge_rtnode_destroy(sc, brt); } } /* * bridge_rtable_init: * * Initialize the route table for this bridge. */ static int bridge_rtable_init(struct bridge_softc *sc) { int i; sc->sc_rthash = malloc(sizeof(*sc->sc_rthash) * BRIDGE_RTHASH_SIZE, M_DEVBUF, M_NOWAIT); if (sc->sc_rthash == NULL) return (ENOMEM); for (i = 0; i < BRIDGE_RTHASH_SIZE; i++) LIST_INIT(&sc->sc_rthash[i]); sc->sc_rthash_key = arc4random(); LIST_INIT(&sc->sc_rtlist); return (0); } /* * bridge_rtable_fini: * * Deconstruct the route table for this bridge. */ static void bridge_rtable_fini(struct bridge_softc *sc) { free(sc->sc_rthash, M_DEVBUF); } /* * The following hash function is adapted from "Hash Functions" by Bob Jenkins * ("Algorithm Alley", Dr. Dobbs Journal, September 1997). */ #define mix(a, b, c) \ do { \ a -= b; a -= c; a ^= (c >> 13); \ b -= c; b -= a; b ^= (a << 8); \ c -= a; c -= b; c ^= (b >> 13); \ a -= b; a -= c; a ^= (c >> 12); \ b -= c; b -= a; b ^= (a << 16); \ c -= a; c -= b; c ^= (b >> 5); \ a -= b; a -= c; a ^= (c >> 3); \ b -= c; b -= a; b ^= (a << 10); \ c -= a; c -= b; c ^= (b >> 15); \ } while (/*CONSTCOND*/0) static inline uint32_t bridge_rthash(struct bridge_softc *sc, const uint8_t *addr) { uint32_t a = 0x9e3779b9, b = 0x9e3779b9, c = sc->sc_rthash_key; b += addr[5] << 8; b += addr[4]; a += addr[3] << 24; a += addr[2] << 16; a += addr[1] << 8; a += addr[0]; mix(a, b, c); return (c & BRIDGE_RTHASH_MASK); } #undef mix /* * bridge_rtnode_lookup: * * Look up a bridge route node for the specified destination. */ static struct bridge_rtnode * bridge_rtnode_lookup(struct bridge_softc *sc, const uint8_t *addr) { struct bridge_rtnode *brt; uint32_t hash; int dir; hash = bridge_rthash(sc, addr); LIST_FOREACH(brt, &sc->sc_rthash[hash], brt_hash) { dir = memcmp(addr, brt->brt_addr, ETHER_ADDR_LEN); if (dir == 0) return (brt); if (dir > 0) return (NULL); } return (NULL); } /* * bridge_rtnode_insert: * * Insert the specified bridge node into the route table. We * assume the entry is not already in the table. */ static int bridge_rtnode_insert(struct bridge_softc *sc, struct bridge_rtnode *brt) { struct bridge_rtnode *lbrt; uint32_t hash; int dir; hash = bridge_rthash(sc, brt->brt_addr); lbrt = LIST_FIRST(&sc->sc_rthash[hash]); if (lbrt == NULL) { LIST_INSERT_HEAD(&sc->sc_rthash[hash], brt, brt_hash); goto out; } do { dir = memcmp(brt->brt_addr, lbrt->brt_addr, ETHER_ADDR_LEN); if (dir == 0) return (EEXIST); if (dir > 0) { LIST_INSERT_BEFORE(lbrt, brt, brt_hash); goto out; } if (LIST_NEXT(lbrt, brt_hash) == NULL) { LIST_INSERT_AFTER(lbrt, brt, brt_hash); goto out; } lbrt = LIST_NEXT(lbrt, brt_hash); } while (lbrt != NULL); #ifdef DIAGNOSTIC panic("bridge_rtnode_insert: impossible"); #endif out: LIST_INSERT_HEAD(&sc->sc_rtlist, brt, brt_list); sc->sc_brtcnt++; return (0); } /* * bridge_rtnode_destroy: * * Destroy a bridge rtnode. */ static void bridge_rtnode_destroy(struct bridge_softc *sc, struct bridge_rtnode *brt) { int s = splnet(); LIST_REMOVE(brt, brt_hash); LIST_REMOVE(brt, brt_list); sc->sc_brtcnt--; pool_put(&bridge_rtnode_pool, brt); splx(s); } #if defined(BRIDGE_IPF) && defined(PFIL_HOOKS) extern struct pfil_head inet_pfil_hook; /* XXX */ extern struct pfil_head inet6_pfil_hook; /* XXX */ /* * Send bridge packets through IPF if they are one of the types IPF can deal * with, or if they are ARP or REVARP. (IPF will pass ARP and REVARP without * question.) */ static int bridge_ipf(void *arg, struct mbuf **mp, struct ifnet *ifp, int dir) { int snap, error; struct ether_header *eh1, eh2; struct llc llc1; u_int16_t ether_type; snap = 0; error = -1; /* Default error if not error == 0 */ eh1 = mtod(*mp, struct ether_header *); ether_type = ntohs(eh1->ether_type); /* * Check for SNAP/LLC. */ if (ether_type < ETHERMTU) { struct llc *llc2 = (struct llc *)(eh1 + 1); if ((*mp)->m_len >= ETHER_HDR_LEN + 8 && llc2->llc_dsap == LLC_SNAP_LSAP && llc2->llc_ssap == LLC_SNAP_LSAP && llc2->llc_control == LLC_UI) { ether_type = htons(llc2->llc_un.type_snap.ether_type); snap = 1; } } /* * If we're trying to filter bridge traffic, don't look at anything * other than IP and ARP traffic. If the filter doesn't understand * IPv6, don't allow IPv6 through the bridge either. This is lame * since if we really wanted, say, an AppleTalk filter, we are hosed, * but of course we don't have an AppleTalk filter to begin with. * (Note that since IPF doesn't understand ARP it will pass *ALL* * ARP traffic.) */ switch (ether_type) { case ETHERTYPE_ARP: case ETHERTYPE_REVARP: return 0; /* Automatically pass */ case ETHERTYPE_IP: # ifdef INET6 case ETHERTYPE_IPV6: # endif /* INET6 */ break; default: goto bad; } /* Strip off the Ethernet header and keep a copy. */ m_copydata(*mp, 0, ETHER_HDR_LEN, (caddr_t) &eh2); m_adj(*mp, ETHER_HDR_LEN); /* Strip off snap header, if present */ if (snap) { m_copydata(*mp, 0, sizeof(struct llc), (caddr_t) &llc1); m_adj(*mp, sizeof(struct llc)); } /* * Check basic packet sanity and run IPF through pfil. */ switch (ether_type) { case ETHERTYPE_IP : error = (dir == PFIL_IN) ? bridge_ip_checkbasic(mp) : 0; if (error == 0) error = pfil_run_hooks(&inet_pfil_hook, mp, ifp, dir); break; # ifdef INET6 case ETHERTYPE_IPV6 : error = (dir == PFIL_IN) ? bridge_ip6_checkbasic(mp) : 0; if (error == 0) error = pfil_run_hooks(&inet6_pfil_hook, mp, ifp, dir); break; # endif default : error = 0; break; } if (*mp == NULL) return error; if (error != 0) goto bad; error = -1; /* * Finally, put everything back the way it was and return */ if (snap) { M_PREPEND(*mp, sizeof(struct llc), M_DONTWAIT); if (*mp == NULL) return error; bcopy(&llc1, mtod(*mp, caddr_t), sizeof(struct llc)); } M_PREPEND(*mp, ETHER_HDR_LEN, M_DONTWAIT); if (*mp == NULL) return error; bcopy(&eh2, mtod(*mp, caddr_t), ETHER_HDR_LEN); return 0; bad: m_freem(*mp); *mp = NULL; return error; } /* * Perform basic checks on header size since * IPF assumes ip_input has already processed * it for it. Cut-and-pasted from ip_input.c. * Given how simple the IPv6 version is, * does the IPv4 version really need to be * this complicated? * * XXX Should we update ipstat here, or not? * XXX Right now we update ipstat but not * XXX csum_counter. */ static int bridge_ip_checkbasic(struct mbuf **mp) { struct mbuf *m = *mp; struct ip *ip; int len, hlen; if (*mp == NULL) return -1; if (IP_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) { if ((m = m_copyup(m, sizeof(struct ip), (max_linkhdr + 3) & ~3)) == NULL) { /* XXXJRT new stat, please */ ipstat.ips_toosmall++; goto bad; } } else if (__predict_false(m->m_len < sizeof (struct ip))) { if ((m = m_pullup(m, sizeof (struct ip))) == NULL) { ipstat.ips_toosmall++; goto bad; } } ip = mtod(m, struct ip *); if (ip == NULL) goto bad; if (ip->ip_v != IPVERSION) { ipstat.ips_badvers++; goto bad; } hlen = ip->ip_hl << 2; if (hlen < sizeof(struct ip)) { /* minimum header length */ ipstat.ips_badhlen++; goto bad; } if (hlen > m->m_len) { if ((m = m_pullup(m, hlen)) == 0) { ipstat.ips_badhlen++; goto bad; } ip = mtod(m, struct ip *); if (ip == NULL) goto bad; } switch (m->m_pkthdr.csum_flags & ((m->m_pkthdr.rcvif->if_csum_flags_rx & M_CSUM_IPv4) | M_CSUM_IPv4_BAD)) { case M_CSUM_IPv4|M_CSUM_IPv4_BAD: /* INET_CSUM_COUNTER_INCR(&ip_hwcsum_bad); */ goto bad; case M_CSUM_IPv4: /* Checksum was okay. */ /* INET_CSUM_COUNTER_INCR(&ip_hwcsum_ok); */ break; default: /* Must compute it ourselves. */ /* INET_CSUM_COUNTER_INCR(&ip_swcsum); */ if (in_cksum(m, hlen) != 0) goto bad; break; } /* Retrieve the packet length. */ len = ntohs(ip->ip_len); /* * Check for additional length bogosity */ if (len < hlen) { ipstat.ips_badlen++; goto bad; } /* * Check that the amount of data in the buffers * is as at least much as the IP header would have us expect. * Drop packet if shorter than we expect. */ if (m->m_pkthdr.len < len) { ipstat.ips_tooshort++; goto bad; } /* Checks out, proceed */ *mp = m; return 0; bad: *mp = m; return -1; } # ifdef INET6 /* * Same as above, but for IPv6. * Cut-and-pasted from ip6_input.c. * XXX Should we update ip6stat, or not? */ static int bridge_ip6_checkbasic(struct mbuf **mp) { struct mbuf *m = *mp; struct ip6_hdr *ip6; /* * If the IPv6 header is not aligned, slurp it up into a new * mbuf with space for link headers, in the event we forward * it. Otherwise, if it is aligned, make sure the entire base * IPv6 header is in the first mbuf of the chain. */ if (IP6_HDR_ALIGNED_P(mtod(m, caddr_t)) == 0) { struct ifnet *inifp = m->m_pkthdr.rcvif; if ((m = m_copyup(m, sizeof(struct ip6_hdr), (max_linkhdr + 3) & ~3)) == NULL) { /* XXXJRT new stat, please */ ip6stat.ip6s_toosmall++; in6_ifstat_inc(inifp, ifs6_in_hdrerr); goto bad; } } else if (__predict_false(m->m_len < sizeof(struct ip6_hdr))) { struct ifnet *inifp = m->m_pkthdr.rcvif; if ((m = m_pullup(m, sizeof(struct ip6_hdr))) == NULL) { ip6stat.ip6s_toosmall++; in6_ifstat_inc(inifp, ifs6_in_hdrerr); goto bad; } } ip6 = mtod(m, struct ip6_hdr *); if ((ip6->ip6_vfc & IPV6_VERSION_MASK) != IPV6_VERSION) { ip6stat.ip6s_badvers++; in6_ifstat_inc(m->m_pkthdr.rcvif, ifs6_in_hdrerr); goto bad; } /* Checks out, proceed */ *mp = m; return 0; bad: *mp = m; return -1; } # endif /* INET6 */ #endif /* BRIDGE_IPF && PFIL_HOOKS */