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 32 #ifndef _NET_ROUTE_H_ 33 #define _NET_ROUTE_H_ 34 35 #ifdef _KERNEL 36 #include <sys/_eventhandler.h> 37 #endif 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_metric; /* route metric */ 94 u_long rmx_filler[1]; 95 }; 96 97 /* 98 * rmx_rtt and rmx_rttvar are stored as microseconds; 99 */ 100 #define RTM_RTTUNIT 1000000 /* units for rtt, rttvar, as units per sec */ 101 102 /* lle state is exported in rmx_state rt_metrics field */ 103 #define rmx_state rmx_weight 104 105 /* default route weight */ 106 #define RT_DEFAULT_WEIGHT 1 107 #define RT_DEFAULT_METRIC 1 108 #define RT_WILDCARD_METRIC 0 109 110 /* 111 * Keep a generation count of routing table, incremented on route addition, 112 * so we can invalidate caches. This is accessed without a lock, as precision 113 * is not required. 114 */ 115 typedef volatile u_int rt_gen_t; /* tree generation (for adds) */ 116 #define RT_GEN(fibnum, af) rt_tables_get_gen(fibnum, af) 117 118 #define RT_DEFAULT_FIB 0 /* Explicitly mark fib=0 restricted cases */ 119 #define RT_ALL_FIBS -1 /* Announce event for every fib */ 120 #ifdef _KERNEL 121 VNET_DECLARE(uint32_t, _rt_numfibs); /* number of existing route tables */ 122 #define V_rt_numfibs VNET(_rt_numfibs) 123 /* temporary compat arg */ 124 #define rt_numfibs V_rt_numfibs 125 VNET_DECLARE(u_int, rt_add_addr_allfibs); /* Announce interfaces to all fibs */ 126 #define V_rt_add_addr_allfibs VNET(rt_add_addr_allfibs) 127 128 /* Calculate flowid for locally-originated packets */ 129 #define V_fib_hash_outbound VNET(fib_hash_outbound) 130 VNET_DECLARE(u_int, fib_hash_outbound); 131 132 typedef void (*rtnumfibs_change_t)(void *, uint32_t); 133 EVENTHANDLER_DECLARE(rtnumfibs_change, rtnumfibs_change_t); 134 135 /* Outbound flowid generation rules */ 136 #ifdef RSS 137 #define fib4_calc_packet_hash xps_proto_software_hash_v4 138 #define fib6_calc_packet_hash xps_proto_software_hash_v6 139 #define CALC_FLOWID_OUTBOUND_SENDTO true 140 #else /* !RSS */ 141 #define fib4_calc_packet_hash fib4_calc_software_hash 142 #define fib6_calc_packet_hash fib6_calc_software_hash 143 #define CALC_FLOWID_OUTBOUND_SENDTO V_fib_hash_outbound 144 #endif /* RSS */ 145 146 147 #endif /* _KERNEL */ 148 149 /* 150 * We distinguish between routes to hosts and routes to networks, 151 * preferring the former if available. For each route we infer 152 * the interface to use from the gateway address supplied when 153 * the route was entered. Routes that forward packets through 154 * gateways are marked so that the output routines know to address the 155 * gateway rather than the ultimate destination. 156 */ 157 #define RTF_UP 0x1 /* route usable */ 158 #define RTF_GATEWAY 0x2 /* destination is a gateway */ 159 #define RTF_HOST 0x4 /* host entry (net otherwise) */ 160 #define RTF_REJECT 0x8 /* host or net unreachable */ 161 #define RTF_DYNAMIC 0x10 /* created dynamically (by redirect) */ 162 #define RTF_MODIFIED 0x20 /* modified dynamically (by redirect) */ 163 #define RTF_DONE 0x40 /* message confirmed */ 164 /* 0x80 unused, was RTF_DELCLONE */ 165 /* 0x100 unused, was RTF_CLONING */ 166 #define RTF_XRESOLVE 0x200 /* external daemon resolves name */ 167 #define RTF_LLINFO 0x400 /* DEPRECATED - exists ONLY for backward 168 compatibility */ 169 #define RTF_LLDATA 0x400 /* used by apps to add/del L2 entries */ 170 #define RTF_STATIC 0x800 /* manually added */ 171 #define RTF_BLACKHOLE 0x1000 /* just discard pkts (during updates) */ 172 #define RTF_PROTO2 0x4000 /* protocol specific routing flag */ 173 #define RTF_PROTO1 0x8000 /* protocol specific routing flag */ 174 /* 0x10000 unused, was RTF_PRCLONING */ 175 /* 0x20000 unused, was RTF_WASCLONED */ 176 #define RTF_PROTO3 0x40000 /* protocol specific routing flag */ 177 #define RTF_FIXEDMTU 0x80000 /* MTU was explicitly specified */ 178 #define RTF_PINNED 0x100000 /* route is immutable */ 179 #define RTF_LOCAL 0x200000 /* route represents a local address */ 180 #define RTF_BROADCAST 0x400000 /* route represents a bcast address */ 181 #define RTF_MULTICAST 0x800000 /* route represents a mcast address */ 182 /* 0x8000000 and up unassigned */ 183 #define RTF_STICKY 0x10000000 /* always route dst->src */ 184 185 /* 0x40000000 unused, was RTF_RNH_LOCKED */ 186 187 #define RTF_GWFLAG_COMPAT 0x80000000 /* a compatibility bit for interacting 188 with existing routing apps */ 189 190 /* Mask of RTF flags that are allowed to be modified by RTM_CHANGE. */ 191 #define RTF_FMASK \ 192 (RTF_PROTO1 | RTF_PROTO2 | RTF_PROTO3 | RTF_BLACKHOLE | \ 193 RTF_REJECT | RTF_STATIC | RTF_STICKY) 194 195 /* 196 * fib_ nexthop API flags. 197 */ 198 199 /* Consumer-visible nexthop info flags */ 200 #define NHF_MULTIPATH 0x0008 /* Nexhop is a nexthop group */ 201 #define NHF_REJECT 0x0010 /* RTF_REJECT */ 202 #define NHF_BLACKHOLE 0x0020 /* RTF_BLACKHOLE */ 203 #define NHF_REDIRECT 0x0040 /* RTF_DYNAMIC|RTF_MODIFIED */ 204 #define NHF_DEFAULT 0x0080 /* Default route */ 205 #define NHF_BROADCAST 0x0100 /* RTF_BROADCAST */ 206 #define NHF_GATEWAY 0x0200 /* RTF_GATEWAY */ 207 #define NHF_HOST 0x0400 /* RTF_HOST */ 208 209 /* Nexthop request flags */ 210 #define NHR_NONE 0x00 /* empty flags field */ 211 #define NHR_REF 0x01 /* reference nexhop */ 212 #define NHR_NODEFAULT 0x02 /* uRPF: do not consider default route */ 213 214 /* Control plane route request flags */ 215 #define NHR_COPY 0x100 /* Copy rte data */ 216 #define NHR_UNLOCKED 0x200 /* Do not lock table */ 217 218 /* 219 * Routing statistics. 220 */ 221 struct rtstat { 222 uint64_t rts_badredirect; /* bogus redirect calls */ 223 uint64_t rts_dynamic; /* routes created by redirects */ 224 uint64_t rts_newgateway; /* routes modified by redirects */ 225 uint64_t rts_unreach; /* lookups which failed */ 226 uint64_t rts_wildcard; /* lookups satisfied by a wildcard */ 227 uint64_t rts_nh_idx_alloc_failure; /* nexthop index alloc failure*/ 228 uint64_t rts_nh_alloc_failure; /* nexthop allocation failure*/ 229 uint64_t rts_add_failure; /* # of route addition failures */ 230 uint64_t rts_add_retry; /* # of route addition retries */ 231 uint64_t rts_del_failure; /* # of route deletion failure */ 232 uint64_t rts_del_retry; /* # of route deletion retries */ 233 uint64_t rts_spare[5]; 234 }; 235 236 /* 237 * Structures for routing messages. 238 */ 239 struct rt_msghdr { 240 u_short rtm_msglen; /* to skip over non-understood messages */ 241 u_char rtm_version; /* future binary compatibility */ 242 u_char rtm_type; /* message type */ 243 u_short rtm_index; /* index for associated ifp */ 244 u_short _rtm_spare1; 245 int rtm_flags; /* flags, incl. kern & message, e.g. DONE */ 246 int rtm_addrs; /* bitmask identifying sockaddrs in msg */ 247 pid_t rtm_pid; /* identify sender */ 248 int rtm_seq; /* for sender to identify action */ 249 int rtm_errno; /* why failed */ 250 int rtm_fmask; /* bitmask used in RTM_CHANGE message */ 251 u_long rtm_inits; /* which metrics we are initializing */ 252 struct rt_metrics rtm_rmx; /* metrics themselves */ 253 }; 254 255 #define RTM_VERSION 5 /* Up the ante and ignore older versions */ 256 257 #ifndef NETLINK_COMPAT 258 /* 259 * Message types. 260 * 261 * The format for each message is annotated below using the following 262 * identifiers: 263 * 264 * (1) struct rt_msghdr 265 * (2) struct ifa_msghdr 266 * (3) struct if_msghdr 267 * (4) struct ifma_msghdr 268 * (5) struct if_announcemsghdr 269 * 270 */ 271 #define RTM_ADD 0x1 /* (1) Add Route */ 272 #define RTM_DELETE 0x2 /* (1) Delete Route */ 273 #define RTM_CHANGE 0x3 /* (1) Change Metrics or flags */ 274 #define RTM_GET 0x4 /* (1) Report Metrics */ 275 #define RTM_LOSING 0x5 /* (1) Kernel Suspects Partitioning */ 276 #define RTM_REDIRECT 0x6 /* (1) Told to use different route */ 277 #define RTM_MISS 0x7 /* (1) Lookup failed on this address */ 278 #define RTM_LOCK 0x8 /* (1) fix specified metrics */ 279 /* 0x9 */ 280 /* 0xa */ 281 #define RTM_RESOLVE 0xb /* (1) req to resolve dst to LL addr */ 282 #define RTM_NEWADDR 0xc /* (2) address being added to iface */ 283 #define RTM_DELADDR 0xd /* (2) address being removed from iface */ 284 #define RTM_IFINFO 0xe /* (3) iface going up/down etc. */ 285 #define RTM_NEWMADDR 0xf /* (4) mcast group membership being added to if */ 286 #define RTM_DELMADDR 0x10 /* (4) mcast group membership being deleted */ 287 #define RTM_IFANNOUNCE 0x11 /* (5) iface arrival/departure */ 288 #define RTM_IEEE80211 0x12 /* (5) IEEE80211 wireless event */ 289 #define RTM_IPFWLOG 0x13 /* (1) IPFW rule match log event */ 290 291 #endif /* NETLINK_COMPAT*/ 292 293 /* 294 * Bitmask values for rtm_inits and rmx_locks. 295 */ 296 #define RTV_MTU 0x1 /* init or lock _mtu */ 297 #define RTV_HOPCOUNT 0x2 /* init or lock _hopcount */ 298 #define RTV_EXPIRE 0x4 /* init or lock _expire */ 299 #define RTV_RPIPE 0x8 /* init or lock _recvpipe */ 300 #define RTV_SPIPE 0x10 /* init or lock _sendpipe */ 301 #define RTV_SSTHRESH 0x20 /* init or lock _ssthresh */ 302 #define RTV_RTT 0x40 /* init or lock _rtt */ 303 #define RTV_RTTVAR 0x80 /* init or lock _rttvar */ 304 #define RTV_WEIGHT 0x100 /* init or lock _weight */ 305 #define RTV_METRIC 0x200 /* init or lock _metric */ 306 307 #ifndef NETLINK_COMPAT 308 309 /* 310 * Bitmask values for rtm_addrs. 311 */ 312 #define RTA_DST 0x1 /* destination sockaddr present */ 313 #define RTA_GATEWAY 0x2 /* gateway sockaddr present */ 314 #define RTA_NETMASK 0x4 /* netmask sockaddr present */ 315 #define RTA_GENMASK 0x8 /* cloning mask sockaddr present */ 316 #define RTA_IFP 0x10 /* interface name sockaddr present */ 317 #define RTA_IFA 0x20 /* interface addr sockaddr present */ 318 #define RTA_AUTHOR 0x40 /* sockaddr for author of redirect */ 319 #define RTA_BRD 0x80 /* for NEWADDR, broadcast or p-p dest addr */ 320 321 #endif /* NETLINK_COMPAT*/ 322 323 /* 324 * Index offsets for sockaddr array for alternate internal encoding. 325 */ 326 #define RTAX_DST 0 /* destination sockaddr present */ 327 #define RTAX_GATEWAY 1 /* gateway sockaddr present */ 328 #define RTAX_NETMASK 2 /* netmask sockaddr present */ 329 #define RTAX_GENMASK 3 /* cloning mask sockaddr present */ 330 #define RTAX_IFP 4 /* interface name sockaddr present */ 331 #define RTAX_IFA 5 /* interface addr sockaddr present */ 332 #define RTAX_AUTHOR 6 /* sockaddr for author of redirect */ 333 #define RTAX_BRD 7 /* for NEWADDR, broadcast or p-p dest addr */ 334 #define RTAX_MAX 8 /* size of array to allocate */ 335 336 struct rtentry; 337 struct nhop_object; 338 typedef int rib_filter_f_t(const struct rtentry *, const struct nhop_object *, 339 void *); 340 341 struct rt_addrinfo { 342 int rti_addrs; /* Route RTF_ flags */ 343 int rti_flags; /* Route RTF_ flags */ 344 struct sockaddr *rti_info[RTAX_MAX]; /* Sockaddr data */ 345 struct ifaddr *rti_ifa; /* value of rt_ifa addr */ 346 struct ifnet *rti_ifp; /* route interface */ 347 rib_filter_f_t *rti_filter; /* filter function */ 348 void *rti_filterdata; /* filter parameters */ 349 u_long rti_mflags; /* metrics RTV_ flags */ 350 u_long rti_spare; /* Will be used for fib */ 351 struct rt_metrics *rti_rmx; /* Pointer to route metrics */ 352 }; 353 354 /* 355 * This macro returns the size of a struct sockaddr when passed 356 * through a routing socket. Basically we round up sa_len to 357 * a multiple of sizeof(long), with a minimum of sizeof(long). 358 * The case sa_len == 0 should only apply to empty structures. 359 */ 360 #define SA_SIZE(sa) \ 361 ( (((struct sockaddr *)(sa))->sa_len == 0) ? \ 362 sizeof(long) : \ 363 1 + ( (((struct sockaddr *)(sa))->sa_len - 1) | (sizeof(long) - 1) ) ) 364 365 #define sa_equal(a, b) ( \ 366 (((const struct sockaddr *)(a))->sa_len == ((const struct sockaddr *)(b))->sa_len) && \ 367 (bcmp((a), (b), ((const struct sockaddr *)(b))->sa_len) == 0)) 368 369 #ifdef _KERNEL 370 371 #define RT_LINK_IS_UP(ifp) (!((ifp)->if_capabilities & IFCAP_LINKSTATE) \ 372 || (ifp)->if_link_state == LINK_STATE_UP) 373 374 #define RO_NHFREE(_ro) do { \ 375 if ((_ro)->ro_nh) { \ 376 NH_FREE((_ro)->ro_nh); \ 377 (_ro)->ro_nh = NULL; \ 378 } \ 379 } while (0) 380 381 #define RO_INVALIDATE_CACHE(ro) do { \ 382 if ((ro)->ro_lle != NULL) { \ 383 LLE_FREE((ro)->ro_lle); \ 384 (ro)->ro_lle = NULL; \ 385 } \ 386 if ((ro)->ro_nh != NULL) { \ 387 NH_FREE((ro)->ro_nh); \ 388 (ro)->ro_nh = NULL; \ 389 } \ 390 } while (0) 391 392 #define RO_GET_FAMILY(ro, dst) ((ro) != NULL && \ 393 (ro)->ro_flags & RT_HAS_GW \ 394 ? (ro)->ro_dst.sa_family : (dst)->sa_family) 395 396 /* 397 * Validate a cached route based on a supplied cookie. If there is an 398 * out-of-date cache, simply free it. Update the generation number 399 * for the new allocation 400 */ 401 #define NH_VALIDATE(ro, cookiep, fibnum) do { \ 402 rt_gen_t cookie = RT_GEN(fibnum, (ro)->ro_dst.sa_family); \ 403 if (*(cookiep) != cookie) { \ 404 RO_INVALIDATE_CACHE(ro); \ 405 *(cookiep) = cookie; \ 406 } \ 407 } while (0) 408 409 struct ifmultiaddr; 410 struct rib_head; 411 412 void rt_ieee80211msg(struct ifnet *, int, void *, size_t); 413 void rt_ifmsg(struct ifnet *, int); 414 void rt_missmsg(int, struct rt_addrinfo *, int, int); 415 void rt_missmsg_fib(int, struct rt_addrinfo *, int, int, int); 416 int rt_addrmsg(int, struct ifaddr *, int); 417 int rt_routemsg(int, struct rtentry *, struct nhop_object *, int); 418 int rt_routemsg_info(int, struct rt_addrinfo *, int); 419 void rt_newmaddrmsg(int, struct ifmultiaddr *); 420 void rt_maskedcopy(const struct sockaddr *, struct sockaddr *, 421 const struct sockaddr *); 422 struct rib_head *rt_table_init(int, int, u_int); 423 void rt_table_destroy(struct rib_head *); 424 u_int rt_tables_get_gen(uint32_t table, sa_family_t family); 425 426 struct sockaddr *rtsock_fix_netmask(const struct sockaddr *dst, 427 const struct sockaddr *smask, struct sockaddr_storage *dmask); 428 429 void rt_updatemtu(struct ifnet *); 430 431 void rt_flushifroutes(struct ifnet *ifp); 432 433 /* XXX MRT NEW VERSIONS THAT USE FIBs 434 * For now the protocol indepedent versions are the same as the AF_INET ones 435 * but this will change.. 436 */ 437 int rtioctl_fib(u_long, caddr_t, u_int); 438 439 /* New API */ 440 void rib_flush_routes_family(int family); 441 struct nhop_object *rib_lookup(uint32_t fibnum, const struct sockaddr *dst, 442 uint32_t flags, uint32_t flowid); 443 const char *rib_print_family(int family); 444 #endif 445 446 #endif 447