1 /*- 2 * SPDX-License-Identifier: BSD-3-Clause 3 * 4 * Copyright (c) 1980, 1986, 1993 5 * The Regents of the University of California. All rights reserved. 6 * 7 * Redistribution and use in source and binary forms, with or without 8 * modification, are permitted provided that the following conditions 9 * are met: 10 * 1. Redistributions of source code must retain the above copyright 11 * notice, this list of conditions and the following disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 3. Neither the name of the University nor the names of its contributors 16 * may be used to endorse or promote products derived from this software 17 * without specific prior written permission. 18 * 19 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 22 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 29 * SUCH DAMAGE. 30 * 31 * @(#)route.h 8.4 (Berkeley) 1/9/95 32 * $FreeBSD$ 33 */ 34 35 #ifndef _NET_ROUTE_H_ 36 #define _NET_ROUTE_H_ 37 38 #include <net/vnet.h> 39 40 /* 41 * Kernel resident routing tables. 42 * 43 * The routing tables are initialized when interface addresses 44 * are set by making entries for all directly connected interfaces. 45 */ 46 47 /* 48 * Struct route consiste of a destination address, 49 * a route entry pointer, link-layer prepend data pointer along 50 * with its length. 51 */ 52 struct route { 53 struct nhop_object *ro_nh; 54 struct llentry *ro_lle; 55 /* 56 * ro_prepend and ro_plen are only used for bpf to pass in a 57 * preformed header. They are not cacheable. 58 */ 59 char *ro_prepend; 60 uint16_t ro_plen; 61 uint16_t ro_flags; 62 uint16_t ro_mtu; /* saved ro_rt mtu */ 63 uint16_t spare; 64 struct sockaddr ro_dst; 65 }; 66 67 #define RT_L2_ME_BIT 2 /* dst L2 addr is our address */ 68 #define RT_MAY_LOOP_BIT 3 /* dst may require loop copy */ 69 #define RT_HAS_HEADER_BIT 4 /* mbuf already have its header prepended */ 70 71 #define RT_L2_ME (1 << RT_L2_ME_BIT) /* 0x0004 */ 72 #define RT_MAY_LOOP (1 << RT_MAY_LOOP_BIT) /* 0x0008 */ 73 #define RT_HAS_HEADER (1 << RT_HAS_HEADER_BIT) /* 0x0010 */ 74 75 #define RT_REJECT 0x0020 /* Destination is reject */ 76 #define RT_BLACKHOLE 0x0040 /* Destination is blackhole */ 77 #define RT_HAS_GW 0x0080 /* Destination has GW */ 78 #define RT_LLE_CACHE 0x0100 /* Cache link layer */ 79 80 struct rt_metrics { 81 u_long rmx_locks; /* Kernel must leave these values alone */ 82 u_long rmx_mtu; /* MTU for this path */ 83 u_long rmx_hopcount; /* max hops expected */ 84 u_long rmx_expire; /* lifetime for route, e.g. redirect */ 85 u_long rmx_recvpipe; /* inbound delay-bandwidth product */ 86 u_long rmx_sendpipe; /* outbound delay-bandwidth product */ 87 u_long rmx_ssthresh; /* outbound gateway buffer limit */ 88 u_long rmx_rtt; /* estimated round trip time */ 89 u_long rmx_rttvar; /* estimated rtt variance */ 90 u_long rmx_pksent; /* packets sent using this route */ 91 u_long rmx_weight; /* route weight */ 92 u_long rmx_nhidx; /* route nexhop index */ 93 u_long rmx_filler[2]; /* will be used for T/TCP later */ 94 }; 95 96 /* 97 * rmx_rtt and rmx_rttvar are stored as microseconds; 98 * RTTTOPRHZ(rtt) converts to a value suitable for use 99 * by a protocol slowtimo counter. 100 */ 101 #define RTM_RTTUNIT 1000000 /* units for rtt, rttvar, as units per sec */ 102 #define RTTTOPRHZ(r) ((r) / (RTM_RTTUNIT / PR_SLOWHZ)) 103 104 /* lle state is exported in rmx_state rt_metrics field */ 105 #define rmx_state rmx_weight 106 107 /* default route weight */ 108 #define RT_DEFAULT_WEIGHT 1 109 #define RT_MAX_WEIGHT 16777215 /* 3 bytes */ 110 111 /* 112 * Keep a generation count of routing table, incremented on route addition, 113 * so we can invalidate caches. This is accessed without a lock, as precision 114 * is not required. 115 */ 116 typedef volatile u_int rt_gen_t; /* tree generation (for adds) */ 117 #define RT_GEN(fibnum, af) rt_tables_get_gen(fibnum, af) 118 119 #define RT_DEFAULT_FIB 0 /* Explicitly mark fib=0 restricted cases */ 120 #define RT_ALL_FIBS -1 /* Announce event for every fib */ 121 #ifdef _KERNEL 122 VNET_DECLARE(uint32_t, _rt_numfibs); /* number of existing route tables */ 123 #define V_rt_numfibs VNET(_rt_numfibs) 124 /* temporary compat arg */ 125 #define rt_numfibs V_rt_numfibs 126 VNET_DECLARE(u_int, rt_add_addr_allfibs); /* Announce interfaces to all fibs */ 127 #define V_rt_add_addr_allfibs VNET(rt_add_addr_allfibs) 128 129 /* Calculate flowid for locally-originated packets */ 130 #define V_fib_hash_outbound VNET(fib_hash_outbound) 131 VNET_DECLARE(u_int, fib_hash_outbound); 132 133 /* Outbound flowid generation rules */ 134 #ifdef RSS 135 136 #define fib4_calc_packet_hash xps_proto_software_hash_v4 137 #define fib6_calc_packet_hash xps_proto_software_hash_v6 138 #define CALC_FLOWID_OUTBOUND_SENDTO true 139 140 #ifdef ROUTE_MPATH 141 #define CALC_FLOWID_OUTBOUND V_fib_hash_outbound 142 #else 143 #define CALC_FLOWID_OUTBOUND false 144 #endif 145 146 #else /* !RSS */ 147 148 #define fib4_calc_packet_hash fib4_calc_software_hash 149 #define fib6_calc_packet_hash fib6_calc_software_hash 150 151 #ifdef ROUTE_MPATH 152 #define CALC_FLOWID_OUTBOUND_SENDTO V_fib_hash_outbound 153 #define CALC_FLOWID_OUTBOUND V_fib_hash_outbound 154 #else 155 #define CALC_FLOWID_OUTBOUND_SENDTO false 156 #define CALC_FLOWID_OUTBOUND false 157 #endif 158 159 #endif /* RSS */ 160 161 162 #endif /* _KERNEL */ 163 164 /* 165 * We distinguish between routes to hosts and routes to networks, 166 * preferring the former if available. For each route we infer 167 * the interface to use from the gateway address supplied when 168 * the route was entered. Routes that forward packets through 169 * gateways are marked so that the output routines know to address the 170 * gateway rather than the ultimate destination. 171 */ 172 #define RTF_UP 0x1 /* route usable */ 173 #define RTF_GATEWAY 0x2 /* destination is a gateway */ 174 #define RTF_HOST 0x4 /* host entry (net otherwise) */ 175 #define RTF_REJECT 0x8 /* host or net unreachable */ 176 #define RTF_DYNAMIC 0x10 /* created dynamically (by redirect) */ 177 #define RTF_MODIFIED 0x20 /* modified dynamically (by redirect) */ 178 #define RTF_DONE 0x40 /* message confirmed */ 179 /* 0x80 unused, was RTF_DELCLONE */ 180 /* 0x100 unused, was RTF_CLONING */ 181 #define RTF_XRESOLVE 0x200 /* external daemon resolves name */ 182 #define RTF_LLINFO 0x400 /* DEPRECATED - exists ONLY for backward 183 compatibility */ 184 #define RTF_LLDATA 0x400 /* used by apps to add/del L2 entries */ 185 #define RTF_STATIC 0x800 /* manually added */ 186 #define RTF_BLACKHOLE 0x1000 /* just discard pkts (during updates) */ 187 #define RTF_PROTO2 0x4000 /* protocol specific routing flag */ 188 #define RTF_PROTO1 0x8000 /* protocol specific routing flag */ 189 /* 0x10000 unused, was RTF_PRCLONING */ 190 /* 0x20000 unused, was RTF_WASCLONED */ 191 #define RTF_PROTO3 0x40000 /* protocol specific routing flag */ 192 #define RTF_FIXEDMTU 0x80000 /* MTU was explicitly specified */ 193 #define RTF_PINNED 0x100000 /* route is immutable */ 194 #define RTF_LOCAL 0x200000 /* route represents a local address */ 195 #define RTF_BROADCAST 0x400000 /* route represents a bcast address */ 196 #define RTF_MULTICAST 0x800000 /* route represents a mcast address */ 197 /* 0x8000000 and up unassigned */ 198 #define RTF_STICKY 0x10000000 /* always route dst->src */ 199 200 /* 0x40000000 unused, was RTF_RNH_LOCKED */ 201 202 #define RTF_GWFLAG_COMPAT 0x80000000 /* a compatibility bit for interacting 203 with existing routing apps */ 204 205 /* Mask of RTF flags that are allowed to be modified by RTM_CHANGE. */ 206 #define RTF_FMASK \ 207 (RTF_PROTO1 | RTF_PROTO2 | RTF_PROTO3 | RTF_BLACKHOLE | \ 208 RTF_REJECT | RTF_STATIC | RTF_STICKY) 209 210 /* 211 * fib_ nexthop API flags. 212 */ 213 214 /* Consumer-visible nexthop info flags */ 215 #define NHF_MULTIPATH 0x0008 /* Nexhop is a nexthop group */ 216 #define NHF_REJECT 0x0010 /* RTF_REJECT */ 217 #define NHF_BLACKHOLE 0x0020 /* RTF_BLACKHOLE */ 218 #define NHF_REDIRECT 0x0040 /* RTF_DYNAMIC|RTF_MODIFIED */ 219 #define NHF_DEFAULT 0x0080 /* Default route */ 220 #define NHF_BROADCAST 0x0100 /* RTF_BROADCAST */ 221 #define NHF_GATEWAY 0x0200 /* RTF_GATEWAY */ 222 #define NHF_HOST 0x0400 /* RTF_HOST */ 223 224 /* Nexthop request flags */ 225 #define NHR_NONE 0x00 /* empty flags field */ 226 #define NHR_REF 0x01 /* reference nexhop */ 227 #define NHR_NODEFAULT 0x02 /* uRPF: do not consider default route */ 228 229 /* Control plane route request flags */ 230 #define NHR_COPY 0x100 /* Copy rte data */ 231 #define NHR_UNLOCKED 0x200 /* Do not lock table */ 232 233 /* 234 * Routing statistics. 235 */ 236 struct rtstat { 237 uint64_t rts_badredirect; /* bogus redirect calls */ 238 uint64_t rts_dynamic; /* routes created by redirects */ 239 uint64_t rts_newgateway; /* routes modified by redirects */ 240 uint64_t rts_unreach; /* lookups which failed */ 241 uint64_t rts_wildcard; /* lookups satisfied by a wildcard */ 242 uint64_t rts_nh_idx_alloc_failure; /* nexthop index alloc failure*/ 243 uint64_t rts_nh_alloc_failure; /* nexthop allocation failure*/ 244 uint64_t rts_add_failure; /* # of route addition failures */ 245 uint64_t rts_add_retry; /* # of route addition retries */ 246 uint64_t rts_del_failure; /* # of route deletion failure */ 247 uint64_t rts_del_retry; /* # of route deletion retries */ 248 }; 249 250 /* 251 * Structures for routing messages. 252 */ 253 struct rt_msghdr { 254 u_short rtm_msglen; /* to skip over non-understood messages */ 255 u_char rtm_version; /* future binary compatibility */ 256 u_char rtm_type; /* message type */ 257 u_short rtm_index; /* index for associated ifp */ 258 u_short _rtm_spare1; 259 int rtm_flags; /* flags, incl. kern & message, e.g. DONE */ 260 int rtm_addrs; /* bitmask identifying sockaddrs in msg */ 261 pid_t rtm_pid; /* identify sender */ 262 int rtm_seq; /* for sender to identify action */ 263 int rtm_errno; /* why failed */ 264 int rtm_fmask; /* bitmask used in RTM_CHANGE message */ 265 u_long rtm_inits; /* which metrics we are initializing */ 266 struct rt_metrics rtm_rmx; /* metrics themselves */ 267 }; 268 269 #define RTM_VERSION 5 /* Up the ante and ignore older versions */ 270 271 /* 272 * Message types. 273 * 274 * The format for each message is annotated below using the following 275 * identifiers: 276 * 277 * (1) struct rt_msghdr 278 * (2) struct ifa_msghdr 279 * (3) struct if_msghdr 280 * (4) struct ifma_msghdr 281 * (5) struct if_announcemsghdr 282 * 283 */ 284 #define RTM_ADD 0x1 /* (1) Add Route */ 285 #define RTM_DELETE 0x2 /* (1) Delete Route */ 286 #define RTM_CHANGE 0x3 /* (1) Change Metrics or flags */ 287 #define RTM_GET 0x4 /* (1) Report Metrics */ 288 #define RTM_LOSING 0x5 /* (1) Kernel Suspects Partitioning */ 289 #define RTM_REDIRECT 0x6 /* (1) Told to use different route */ 290 #define RTM_MISS 0x7 /* (1) Lookup failed on this address */ 291 #define RTM_LOCK 0x8 /* (1) fix specified metrics */ 292 /* 0x9 */ 293 /* 0xa */ 294 #define RTM_RESOLVE 0xb /* (1) req to resolve dst to LL addr */ 295 #define RTM_NEWADDR 0xc /* (2) address being added to iface */ 296 #define RTM_DELADDR 0xd /* (2) address being removed from iface */ 297 #define RTM_IFINFO 0xe /* (3) iface going up/down etc. */ 298 #define RTM_NEWMADDR 0xf /* (4) mcast group membership being added to if */ 299 #define RTM_DELMADDR 0x10 /* (4) mcast group membership being deleted */ 300 #define RTM_IFANNOUNCE 0x11 /* (5) iface arrival/departure */ 301 #define RTM_IEEE80211 0x12 /* (5) IEEE80211 wireless event */ 302 303 /* 304 * Bitmask values for rtm_inits and rmx_locks. 305 */ 306 #define RTV_MTU 0x1 /* init or lock _mtu */ 307 #define RTV_HOPCOUNT 0x2 /* init or lock _hopcount */ 308 #define RTV_EXPIRE 0x4 /* init or lock _expire */ 309 #define RTV_RPIPE 0x8 /* init or lock _recvpipe */ 310 #define RTV_SPIPE 0x10 /* init or lock _sendpipe */ 311 #define RTV_SSTHRESH 0x20 /* init or lock _ssthresh */ 312 #define RTV_RTT 0x40 /* init or lock _rtt */ 313 #define RTV_RTTVAR 0x80 /* init or lock _rttvar */ 314 #define RTV_WEIGHT 0x100 /* init or lock _weight */ 315 316 /* 317 * Bitmask values for rtm_addrs. 318 */ 319 #define RTA_DST 0x1 /* destination sockaddr present */ 320 #define RTA_GATEWAY 0x2 /* gateway sockaddr present */ 321 #define RTA_NETMASK 0x4 /* netmask sockaddr present */ 322 #define RTA_GENMASK 0x8 /* cloning mask sockaddr present */ 323 #define RTA_IFP 0x10 /* interface name sockaddr present */ 324 #define RTA_IFA 0x20 /* interface addr sockaddr present */ 325 #define RTA_AUTHOR 0x40 /* sockaddr for author of redirect */ 326 #define RTA_BRD 0x80 /* for NEWADDR, broadcast or p-p dest addr */ 327 328 /* 329 * Index offsets for sockaddr array for alternate internal encoding. 330 */ 331 #define RTAX_DST 0 /* destination sockaddr present */ 332 #define RTAX_GATEWAY 1 /* gateway sockaddr present */ 333 #define RTAX_NETMASK 2 /* netmask sockaddr present */ 334 #define RTAX_GENMASK 3 /* cloning mask sockaddr present */ 335 #define RTAX_IFP 4 /* interface name sockaddr present */ 336 #define RTAX_IFA 5 /* interface addr sockaddr present */ 337 #define RTAX_AUTHOR 6 /* sockaddr for author of redirect */ 338 #define RTAX_BRD 7 /* for NEWADDR, broadcast or p-p dest addr */ 339 #define RTAX_MAX 8 /* size of array to allocate */ 340 341 struct rtentry; 342 struct nhop_object; 343 typedef int rib_filter_f_t(const struct rtentry *, const struct nhop_object *, 344 void *); 345 346 struct rt_addrinfo { 347 int rti_addrs; /* Route RTF_ flags */ 348 int rti_flags; /* Route RTF_ flags */ 349 struct sockaddr *rti_info[RTAX_MAX]; /* Sockaddr data */ 350 struct ifaddr *rti_ifa; /* value of rt_ifa addr */ 351 struct ifnet *rti_ifp; /* route interface */ 352 rib_filter_f_t *rti_filter; /* filter function */ 353 void *rti_filterdata; /* filter parameters */ 354 u_long rti_mflags; /* metrics RTV_ flags */ 355 u_long rti_spare; /* Will be used for fib */ 356 struct rt_metrics *rti_rmx; /* Pointer to route metrics */ 357 }; 358 359 /* 360 * This macro returns the size of a struct sockaddr when passed 361 * through a routing socket. Basically we round up sa_len to 362 * a multiple of sizeof(long), with a minimum of sizeof(long). 363 * The case sa_len == 0 should only apply to empty structures. 364 */ 365 #define SA_SIZE(sa) \ 366 ( (((struct sockaddr *)(sa))->sa_len == 0) ? \ 367 sizeof(long) : \ 368 1 + ( (((struct sockaddr *)(sa))->sa_len - 1) | (sizeof(long) - 1) ) ) 369 370 #define sa_equal(a, b) ( \ 371 (((const struct sockaddr *)(a))->sa_len == ((const struct sockaddr *)(b))->sa_len) && \ 372 (bcmp((a), (b), ((const struct sockaddr *)(b))->sa_len) == 0)) 373 374 #ifdef _KERNEL 375 376 #define RT_LINK_IS_UP(ifp) (!((ifp)->if_capabilities & IFCAP_LINKSTATE) \ 377 || (ifp)->if_link_state == LINK_STATE_UP) 378 379 #define RO_NHFREE(_ro) do { \ 380 if ((_ro)->ro_nh) { \ 381 NH_FREE((_ro)->ro_nh); \ 382 (_ro)->ro_nh = NULL; \ 383 } \ 384 } while (0) 385 386 #define RO_INVALIDATE_CACHE(ro) do { \ 387 if ((ro)->ro_lle != NULL) { \ 388 LLE_FREE((ro)->ro_lle); \ 389 (ro)->ro_lle = NULL; \ 390 } \ 391 if ((ro)->ro_nh != NULL) { \ 392 NH_FREE((ro)->ro_nh); \ 393 (ro)->ro_nh = NULL; \ 394 } \ 395 } while (0) 396 397 #define RO_GET_FAMILY(ro, dst) ((ro) != NULL && \ 398 (ro)->ro_flags & RT_HAS_GW \ 399 ? (ro)->ro_dst.sa_family : (dst)->sa_family) 400 401 /* 402 * Validate a cached route based on a supplied cookie. If there is an 403 * out-of-date cache, simply free it. Update the generation number 404 * for the new allocation 405 */ 406 #define NH_VALIDATE(ro, cookiep, fibnum) do { \ 407 rt_gen_t cookie = RT_GEN(fibnum, (ro)->ro_dst.sa_family); \ 408 if (*(cookiep) != cookie) { \ 409 RO_INVALIDATE_CACHE(ro); \ 410 *(cookiep) = cookie; \ 411 } \ 412 } while (0) 413 414 struct ifmultiaddr; 415 struct rib_head; 416 417 void rt_ieee80211msg(struct ifnet *, int, void *, size_t); 418 void rt_ifannouncemsg(struct ifnet *, int); 419 void rt_ifmsg(struct ifnet *); 420 void rt_missmsg(int, struct rt_addrinfo *, int, int); 421 void rt_missmsg_fib(int, struct rt_addrinfo *, int, int, int); 422 int rt_addrmsg(int, struct ifaddr *, int); 423 int rt_routemsg(int, struct rtentry *, struct nhop_object *, int); 424 int rt_routemsg_info(int, struct rt_addrinfo *, int); 425 void rt_newmaddrmsg(int, struct ifmultiaddr *); 426 void rt_maskedcopy(struct sockaddr *, struct sockaddr *, struct sockaddr *); 427 struct rib_head *rt_table_init(int, int, u_int); 428 void rt_table_destroy(struct rib_head *); 429 u_int rt_tables_get_gen(uint32_t table, sa_family_t family); 430 431 struct sockaddr *rtsock_fix_netmask(const struct sockaddr *dst, 432 const struct sockaddr *smask, struct sockaddr_storage *dmask); 433 434 void rt_updatemtu(struct ifnet *); 435 436 void rt_flushifroutes(struct ifnet *ifp); 437 438 /* XXX MRT NEW VERSIONS THAT USE FIBs 439 * For now the protocol indepedent versions are the same as the AF_INET ones 440 * but this will change.. 441 */ 442 int rtioctl_fib(u_long, caddr_t, u_int); 443 int rib_lookup_info(uint32_t, const struct sockaddr *, uint32_t, uint32_t, 444 struct rt_addrinfo *); 445 void rib_free_info(struct rt_addrinfo *info); 446 447 /* New API */ 448 void rib_flush_routes_family(int family); 449 struct nhop_object *rib_lookup(uint32_t fibnum, const struct sockaddr *dst, 450 uint32_t flags, uint32_t flowid); 451 const char *rib_print_family(int family); 452 #endif 453 454 #endif 455