1 /* 2 * CDDL HEADER START 3 * 4 * The contents of this file are subject to the terms of the 5 * Common Development and Distribution License (the "License"). 6 * You may not use this file except in compliance with the License. 7 * 8 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE 9 * or http://www.opensolaris.org/os/licensing. 10 * See the License for the specific language governing permissions 11 * and limitations under the License. 12 * 13 * When distributing Covered Code, include this CDDL HEADER in each 14 * file and include the License file at usr/src/OPENSOLARIS.LICENSE. 15 * If applicable, add the following below this CDDL HEADER, with the 16 * fields enclosed by brackets "[]" replaced with your own identifying 17 * information: Portions Copyright [yyyy] [name of copyright owner] 18 * 19 * CDDL HEADER END 20 */ 21 /* 22 * Copyright 2006 Sun Microsystems, Inc. All rights reserved. 23 * Use is subject to license terms. 24 */ 25 /* Copyright (c) 1990 Mentat Inc. */ 26 27 #ifndef _INET_IP_IRE_H 28 #define _INET_IP_IRE_H 29 30 #pragma ident "%Z%%M% %I% %E% SMI" 31 32 #ifdef __cplusplus 33 extern "C" { 34 #endif 35 36 #define IPV6_LL_PREFIXLEN 10 /* Number of bits in link-local pref */ 37 38 #define IP_FTABLE_HASH_SIZE 32 /* size of each hash table in ptrs */ 39 #define IP_CACHE_TABLE_SIZE 256 40 #define IP_MRTUN_TABLE_SIZE 256 /* Mobile IP reverse tunnel table */ 41 /* size. Only used by mipagent */ 42 #define IP_SRCIF_TABLE_SIZE 256 /* Per interface routing table size */ 43 #define IP_MASK_TABLE_SIZE (IP_ABITS + 1) /* 33 ptrs */ 44 45 #define IP6_FTABLE_HASH_SIZE 32 /* size of each hash table in ptrs */ 46 #define IP6_CACHE_TABLE_SIZE 256 47 #define IP6_MASK_TABLE_SIZE (IPV6_ABITS + 1) /* 129 ptrs */ 48 49 /* 50 * We use the common modulo hash function. In ip_ire_init(), we make 51 * sure that the cache table size is always a power of 2. That's why 52 * we can use & instead of %. Also note that we try hard to make sure 53 * the lower bits of an address capture most info from the whole address. 54 * The reason being that since our hash table is probably a lot smaller 55 * than 2^32 buckets so the lower bits are the most important. 56 */ 57 #define IRE_ADDR_HASH(addr, table_size) \ 58 (((addr) ^ ((addr) >> 8) ^ ((addr) >> 16) ^ ((addr) >> 24)) & \ 59 ((table_size) - 1)) 60 61 /* 62 * Exclusive-or those bytes that are likely to contain the MAC 63 * address. Assumes EUI-64 format for good hashing. 64 */ 65 #define IRE_ADDR_HASH_V6(addr, table_size) \ 66 (((addr).s6_addr32[3] ^ \ 67 (((addr).s6_addr32[3] ^ (addr).s6_addr32[2]) >> 12)) & \ 68 ((table_size) - 1)) 69 /* This assumes that the ftable size is a power of 2. */ 70 #define IRE_ADDR_MASK_HASH_V6(addr, mask, table_size) \ 71 ((((addr).s6_addr8[8] & (mask).s6_addr8[8]) ^ \ 72 ((addr).s6_addr8[9] & (mask).s6_addr8[9]) ^ \ 73 ((addr).s6_addr8[10] & (mask).s6_addr8[10]) ^ \ 74 ((addr).s6_addr8[13] & (mask).s6_addr8[13]) ^ \ 75 ((addr).s6_addr8[14] & (mask).s6_addr8[14]) ^ \ 76 ((addr).s6_addr8[15] & (mask).s6_addr8[15])) & ((table_size) - 1)) 77 78 /* 79 * match parameter definitions for 80 * IRE lookup routines. 81 */ 82 #define MATCH_IRE_DSTONLY 0x0000 /* Match just the address */ 83 #define MATCH_IRE_TYPE 0x0001 /* Match IRE type */ 84 #define MATCH_IRE_SRC 0x0002 /* Match IRE source address */ 85 #define MATCH_IRE_MASK 0x0004 /* Match IRE mask */ 86 #define MATCH_IRE_WQ 0x0008 /* Match IRE Write Q */ 87 #define MATCH_IRE_GW 0x0010 /* Match IRE gateway */ 88 #define MATCH_IRE_IPIF 0x0020 /* Match IRE ipif */ 89 #define MATCH_IRE_RECURSIVE 0x0040 /* Do recursive lookup if necessary */ 90 #define MATCH_IRE_DEFAULT 0x0080 /* Return default route if no route */ 91 /* found. */ 92 #define MATCH_IRE_RJ_BHOLE 0x0100 /* During lookup if we hit an ire */ 93 /* with RTF_REJECT or RTF_BLACKHOLE, */ 94 /* return the ire. No recursive */ 95 /* lookup should be done. */ 96 #define MATCH_IRE_IHANDLE 0x0200 /* Match IRE on ihandle */ 97 #define MATCH_IRE_MARK_HIDDEN 0x0400 /* Match IRE ire_marks with */ 98 /* IRE_MARK_HIDDEN. */ 99 /* 100 * MATCH_IRE_ILL is used whenever we want to specifically match an IRE 101 * whose ire_ipif->ipif_ill or (ill_t *)ire_stq->q_ptr matches a given 102 * ill. When MATCH_IRE_ILL is used to locate an IRE_CACHE, it implies 103 * that the packet will not be load balanced. This is normally used 104 * by in.mpathd to send out failure detection probes. 105 * 106 * MATCH_IRE_ILL_GROUP is used whenever we are not specific about which 107 * interface (ill) the packet should be sent out. This implies that the 108 * packets will be subjected to load balancing and it might go out on 109 * any interface in the group. When there is only interface in the group, 110 * MATCH_IRE_ILL_GROUP becomes MATCH_IRE_ILL. Most of the code uses 111 * MATCH_IRE_ILL_GROUP and MATCH_IRE_ILL is used in very few cases where 112 * we want to disable load balancing. 113 * 114 * MATCH_IRE_PARENT is used whenever we unconditionally want to get the 115 * parent IRE (sire) while recursively searching IREs for an offsubnet 116 * destination. With this flag, even if no IRE_CACHETABLE or IRE_INTERFACE 117 * is found to help resolving IRE_OFFSUBNET in lookup routines, the 118 * IRE_OFFSUBNET sire, if any, is returned to the caller. 119 */ 120 #define MATCH_IRE_ILL_GROUP 0x0800 /* Match IRE on ill or the ill_group. */ 121 #define MATCH_IRE_ILL 0x1000 /* Match IRE on the ill only */ 122 123 #define MATCH_IRE_PARENT 0x2000 /* Match parent ire, if any, */ 124 /* even if ire is not matched. */ 125 #define MATCH_IRE_ZONEONLY 0x4000 /* Match IREs in specified zone, ie */ 126 /* don't match IRE_LOCALs from other */ 127 /* zones or shared IREs */ 128 #define MATCH_IRE_MARK_PRIVATE_ADDR 0x8000 /* Match IRE ire_marks with */ 129 /* IRE_MARK_PRIVATE_ADDR. */ 130 #define MATCH_IRE_SECATTR 0x10000 /* Match gateway security attributes */ 131 132 /* Structure for ire_cache_count() */ 133 typedef struct { 134 int icc_total; /* Total number of IRE_CACHE */ 135 int icc_unused; /* # off/no PMTU unused since last reclaim */ 136 int icc_offlink; /* # offlink without PMTU information */ 137 int icc_pmtu; /* # offlink with PMTU information */ 138 int icc_onlink; /* # onlink */ 139 } ire_cache_count_t; 140 141 /* 142 * Structure for ire_cache_reclaim(). Each field is a fraction i.e. 1 meaning 143 * reclaim all, N meaning reclaim 1/Nth of all entries, 0 meaning reclaim none. 144 */ 145 typedef struct { 146 int icr_unused; /* Fraction for unused since last reclaim */ 147 int icr_offlink; /* Fraction for offlink without PMTU info */ 148 int icr_pmtu; /* Fraction for offlink with PMTU info */ 149 int icr_onlink; /* Fraction for onlink */ 150 } ire_cache_reclaim_t; 151 152 typedef struct { 153 uint64_t ire_stats_alloced; /* # of ires alloced */ 154 uint64_t ire_stats_freed; /* # of ires freed */ 155 uint64_t ire_stats_inserted; /* # of ires inserted in the bucket */ 156 uint64_t ire_stats_deleted; /* # of ires deleted from the bucket */ 157 } ire_stats_t; 158 159 extern ire_stats_t ire_stats_v4; 160 161 /* 162 * We use atomics so that we get an accurate accounting on the ires. 163 * Otherwise we can't determine leaks correctly. 164 */ 165 #define BUMP_IRE_STATS(ire_stats, x) atomic_add_64(&(ire_stats).x, 1) 166 167 extern irb_t *ip_forwarding_table_v6[]; 168 extern irb_t *ip_cache_table_v6; 169 extern irb_t *ip_mrtun_table; 170 extern irb_t *ip_srcif_table; 171 extern kmutex_t ire_ft_init_lock; 172 extern kmutex_t ire_mrtun_lock; 173 extern kmutex_t ire_srcif_table_lock; 174 extern ire_stats_t ire_stats_v6; 175 extern uint_t ire_mrtun_count; 176 extern uint_t ire_srcif_table_count; 177 178 #ifdef _KERNEL 179 struct ts_label_s; 180 181 extern ipaddr_t ip_plen_to_mask(uint_t); 182 extern in6_addr_t *ip_plen_to_mask_v6(uint_t, in6_addr_t *); 183 184 extern int ip_ire_advise(queue_t *, mblk_t *, cred_t *); 185 extern int ip_ire_delete(queue_t *, mblk_t *, cred_t *); 186 extern boolean_t ip_ire_clookup_and_delete(ipaddr_t, ipif_t *); 187 extern void ip_ire_clookup_and_delete_v6(const in6_addr_t *); 188 189 extern int ip_ire_report(queue_t *, mblk_t *, caddr_t, cred_t *); 190 extern int ip_ire_report_mrtun(queue_t *, mblk_t *, caddr_t, cred_t *); 191 extern int ip_ire_report_srcif(queue_t *, mblk_t *, caddr_t, cred_t *); 192 extern int ip_ire_report_v6(queue_t *, mblk_t *, caddr_t, cred_t *); 193 194 extern void ip_ire_req(queue_t *, mblk_t *); 195 196 extern int ip_mask_to_plen(ipaddr_t); 197 extern int ip_mask_to_plen_v6(const in6_addr_t *); 198 199 extern ire_t *ipif_to_ire(const ipif_t *); 200 extern ire_t *ipif_to_ire_v6(const ipif_t *); 201 202 extern int ire_add(ire_t **, queue_t *, mblk_t *, ipsq_func_t); 203 extern int ire_add_mrtun(ire_t **, queue_t *, mblk_t *, ipsq_func_t); 204 extern void ire_add_then_send(queue_t *, ire_t *, mblk_t *); 205 extern int ire_add_v6(ire_t **, queue_t *, mblk_t *, ipsq_func_t); 206 extern int ire_atomic_start(irb_t *irb_ptr, ire_t *ire, queue_t *q, 207 mblk_t *mp, ipsq_func_t func); 208 extern void ire_atomic_end(irb_t *irb_ptr, ire_t *ire); 209 210 extern void ire_cache_count(ire_t *, char *); 211 extern ire_t *ire_cache_lookup(ipaddr_t, zoneid_t, 212 const struct ts_label_s *); 213 extern ire_t *ire_cache_lookup_v6(const in6_addr_t *, zoneid_t, 214 const struct ts_label_s *); 215 extern void ire_cache_reclaim(ire_t *, char *); 216 217 extern void ire_check_bcast_present(ipif_t *, ipaddr_t, int, boolean_t *, 218 boolean_t *); 219 220 extern ire_t *ire_create_mp(uchar_t *, uchar_t *, uchar_t *, uchar_t *, 221 uchar_t *, uint_t, mblk_t *, queue_t *, queue_t *, ushort_t, mblk_t *, 222 ipif_t *, ill_t *, ipaddr_t, uint32_t, uint32_t, uint32_t, const iulp_t *, 223 tsol_gc_t *, tsol_gcgrp_t *); 224 225 extern ire_t *ire_create(uchar_t *, uchar_t *, uchar_t *, uchar_t *, 226 uchar_t *, uint_t *, mblk_t *, queue_t *, queue_t *, ushort_t, mblk_t *, 227 ipif_t *, ill_t *, ipaddr_t, uint32_t, uint32_t, uint32_t, const iulp_t *, 228 tsol_gc_t *, tsol_gcgrp_t *); 229 230 extern ire_t **ire_check_and_create_bcast(ipif_t *, ipaddr_t, 231 ire_t **, int); 232 extern ire_t **ire_create_bcast(ipif_t *, ipaddr_t, ire_t **); 233 extern ire_t *ire_init(ire_t *, uchar_t *, uchar_t *, uchar_t *, 234 uchar_t *, uchar_t *, uint_t *, mblk_t *, queue_t *, queue_t *, ushort_t, 235 mblk_t *, ipif_t *, ill_t *, ipaddr_t, uint32_t, uint32_t, uint32_t, 236 const iulp_t *, tsol_gc_t *, tsol_gcgrp_t *); 237 238 extern boolean_t ire_init_common(ire_t *, uint_t *, mblk_t *, queue_t *, 239 queue_t *, ushort_t, mblk_t *, ipif_t *, ill_t *, uint32_t, 240 uint32_t, uint32_t, uchar_t, const iulp_t *, tsol_gc_t *, tsol_gcgrp_t *); 241 242 extern ire_t *ire_create_v6(const in6_addr_t *, const in6_addr_t *, 243 const in6_addr_t *, const in6_addr_t *, uint_t *, mblk_t *, queue_t *, 244 queue_t *, ushort_t, mblk_t *, ipif_t *, 245 const in6_addr_t *, uint32_t, uint32_t, uint_t, const iulp_t *, 246 tsol_gc_t *, tsol_gcgrp_t *); 247 248 extern ire_t *ire_create_mp_v6(const in6_addr_t *, const in6_addr_t *, 249 const in6_addr_t *, const in6_addr_t *, mblk_t *, queue_t *, 250 queue_t *, ushort_t, mblk_t *, ipif_t *, 251 const in6_addr_t *, uint32_t, uint32_t, uint_t, const iulp_t *, 252 tsol_gc_t *, tsol_gcgrp_t *); 253 254 extern ire_t *ire_init_v6(ire_t *, const in6_addr_t *, const in6_addr_t *, 255 const in6_addr_t *, const in6_addr_t *, uint_t *, mblk_t *, queue_t *, 256 queue_t *, ushort_t, mblk_t *, ipif_t *, 257 const in6_addr_t *, uint32_t, uint32_t, uint_t, const iulp_t *, 258 tsol_gc_t *, tsol_gcgrp_t *); 259 260 extern void ire_clookup_delete_cache_gw(ipaddr_t, zoneid_t); 261 extern void ire_clookup_delete_cache_gw_v6(const in6_addr_t *, zoneid_t); 262 263 extern ire_t *ire_ctable_lookup(ipaddr_t, ipaddr_t, int, const ipif_t *, 264 zoneid_t, const struct ts_label_s *, int); 265 266 extern ire_t *ire_ctable_lookup_v6(const in6_addr_t *, const in6_addr_t *, 267 int, const ipif_t *, zoneid_t, const struct ts_label_s *, int); 268 269 extern void ire_delete(ire_t *); 270 extern void ire_delete_cache_gw(ire_t *, char *); 271 extern void ire_delete_cache_gw_v6(ire_t *, char *); 272 extern void ire_delete_cache_v6(ire_t *, char *); 273 extern void ire_delete_srcif(ire_t *); 274 extern void ire_delete_v6(ire_t *); 275 276 extern void ire_expire(ire_t *, char *); 277 extern void ire_fastpath_flush(ire_t *, void *); 278 extern boolean_t ire_fastpath_update(ire_t *, void *); 279 280 extern void ire_flush_cache_v4(ire_t *, int); 281 extern void ire_flush_cache_v6(ire_t *, int); 282 283 extern ire_t *ire_ftable_lookup(ipaddr_t, ipaddr_t, ipaddr_t, int, 284 const ipif_t *, ire_t **, zoneid_t, uint32_t, 285 const struct ts_label_s *, int); 286 287 extern ire_t *ire_ftable_lookup_v6(const in6_addr_t *, const in6_addr_t *, 288 const in6_addr_t *, int, const ipif_t *, ire_t **, zoneid_t, 289 uint32_t, const struct ts_label_s *, int); 290 291 extern ire_t *ire_ihandle_lookup_onlink(ire_t *); 292 extern ire_t *ire_ihandle_lookup_offlink(ire_t *, ire_t *); 293 extern ire_t *ire_ihandle_lookup_offlink_v6(ire_t *, ire_t *); 294 295 extern ire_t *ire_lookup_local(zoneid_t); 296 extern ire_t *ire_lookup_local_v6(zoneid_t); 297 298 extern ire_t *ire_lookup_multi(ipaddr_t, zoneid_t); 299 extern ire_t *ire_lookup_multi_v6(const in6_addr_t *, zoneid_t); 300 301 extern ire_t *ire_mrtun_lookup(ipaddr_t, ill_t *); 302 303 extern void ire_refrele(ire_t *); 304 extern void ire_refrele_notr(ire_t *); 305 extern ire_t *ire_route_lookup(ipaddr_t, ipaddr_t, ipaddr_t, int, 306 const ipif_t *, ire_t **, zoneid_t, const struct ts_label_s *, int); 307 308 extern ire_t *ire_route_lookup_v6(const in6_addr_t *, const in6_addr_t *, 309 const in6_addr_t *, int, const ipif_t *, ire_t **, zoneid_t, 310 const struct ts_label_s *, int); 311 312 extern ire_t *ire_srcif_table_lookup(ipaddr_t, int, ipif_t *, ill_t *, int); 313 extern ill_t *ire_to_ill(const ire_t *); 314 315 extern void ire_walk(pfv_t, void *); 316 extern void ire_walk_ill(uint_t, uint_t, pfv_t, void *, ill_t *); 317 extern void ire_walk_ill_mrtun(uint_t, uint_t, pfv_t, void *, ill_t *); 318 extern void ire_walk_ill_v4(uint_t, uint_t, pfv_t, void *, ill_t *); 319 extern void ire_walk_ill_v6(uint_t, uint_t, pfv_t, void *, ill_t *); 320 extern void ire_walk_v4(pfv_t, void *, zoneid_t); 321 extern void ire_walk_srcif_table_v4(pfv_t, void *); 322 extern void ire_walk_v6(pfv_t, void *, zoneid_t); 323 324 extern boolean_t ire_multirt_lookup(ire_t **, ire_t **, uint32_t, 325 const struct ts_label_s *); 326 extern boolean_t ire_multirt_need_resolve(ipaddr_t, 327 const struct ts_label_s *); 328 extern boolean_t ire_multirt_lookup_v6(ire_t **, ire_t **, uint32_t, 329 const struct ts_label_s *); 330 extern boolean_t ire_multirt_need_resolve_v6(const in6_addr_t *, 331 const struct ts_label_s *); 332 333 extern ire_t *ipif_lookup_multi_ire(ipif_t *, ipaddr_t); 334 extern ire_t *ipif_lookup_multi_ire_v6(ipif_t *, const in6_addr_t *); 335 336 extern void ire_fastpath_list_dispatch(ill_t *, 337 boolean_t (*)(ire_t *, void *), void *); 338 extern void ire_fastpath_list_delete(ill_t *, ire_t *); 339 340 extern mblk_t *ip_nexthop_route(const struct sockaddr *, char *); 341 extern mblk_t *ip_nexthop(const struct sockaddr *, const char *); 342 343 extern ire_t *ire_get_next_bcast_ire(ire_t *, ire_t *); 344 extern ire_t *ire_get_next_default_ire(ire_t *, ire_t *); 345 346 #endif /* _KERNEL */ 347 348 #ifdef __cplusplus 349 } 350 #endif 351 352 #endif /* _INET_IP_IRE_H */ 353