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 /* 23 * Copyright 2009 Sun Microsystems, Inc. All rights reserved. 24 * Use is subject to license terms. 25 */ 26 27 #ifndef _INET_IP_STACK_H 28 #define _INET_IP_STACK_H 29 30 #ifdef __cplusplus 31 extern "C" { 32 #endif 33 34 #include <sys/netstack.h> 35 #include <netinet/igmp_var.h> 36 #include <sys/modhash.h> 37 38 #ifdef _KERNEL 39 #include <sys/list.h> 40 41 /* 42 * IP statistics. 43 */ 44 #define IP_STAT(ipst, x) ((ipst)->ips_ip_statistics.x.value.ui64++) 45 #define IP_STAT_UPDATE(ipst, x, n) \ 46 ((ipst)->ips_ip_statistics.x.value.ui64 += (n)) 47 48 typedef struct ip_stat { 49 kstat_named_t ipsec_fanout_proto; 50 kstat_named_t ip_udp_fannorm; 51 kstat_named_t ip_udp_fanmb; 52 kstat_named_t ip_udp_fanothers; 53 kstat_named_t ip_udp_fast_path; 54 kstat_named_t ip_udp_slow_path; 55 kstat_named_t ip_udp_input_err; 56 kstat_named_t ip_tcppullup; 57 kstat_named_t ip_tcpoptions; 58 kstat_named_t ip_multipkttcp; 59 kstat_named_t ip_tcp_fast_path; 60 kstat_named_t ip_tcp_slow_path; 61 kstat_named_t ip_tcp_input_error; 62 kstat_named_t ip_db_ref; 63 kstat_named_t ip_notaligned1; 64 kstat_named_t ip_notaligned2; 65 kstat_named_t ip_multimblk3; 66 kstat_named_t ip_multimblk4; 67 kstat_named_t ip_ipoptions; 68 kstat_named_t ip_classify_fail; 69 kstat_named_t ip_opt; 70 kstat_named_t ip_udp_rput_local; 71 kstat_named_t ipsec_proto_ahesp; 72 kstat_named_t ip_conn_flputbq; 73 kstat_named_t ip_conn_walk_drain; 74 kstat_named_t ip_out_sw_cksum; 75 kstat_named_t ip_in_sw_cksum; 76 kstat_named_t ip_trash_ire_reclaim_calls; 77 kstat_named_t ip_trash_ire_reclaim_success; 78 kstat_named_t ip_ire_arp_timer_expired; 79 kstat_named_t ip_ire_redirect_timer_expired; 80 kstat_named_t ip_ire_pmtu_timer_expired; 81 kstat_named_t ip_input_multi_squeue; 82 kstat_named_t ip_tcp_in_full_hw_cksum_err; 83 kstat_named_t ip_tcp_in_part_hw_cksum_err; 84 kstat_named_t ip_tcp_in_sw_cksum_err; 85 kstat_named_t ip_tcp_out_sw_cksum_bytes; 86 kstat_named_t ip_udp_in_full_hw_cksum_err; 87 kstat_named_t ip_udp_in_part_hw_cksum_err; 88 kstat_named_t ip_udp_in_sw_cksum_err; 89 kstat_named_t ip_udp_out_sw_cksum_bytes; 90 kstat_named_t ip_frag_mdt_pkt_out; 91 kstat_named_t ip_frag_mdt_discarded; 92 kstat_named_t ip_frag_mdt_allocfail; 93 kstat_named_t ip_frag_mdt_addpdescfail; 94 kstat_named_t ip_frag_mdt_allocd; 95 } ip_stat_t; 96 97 98 /* 99 * IP6 statistics. 100 */ 101 #define IP6_STAT(ipst, x) ((ipst)->ips_ip6_statistics.x.value.ui64++) 102 #define IP6_STAT_UPDATE(ipst, x, n) \ 103 ((ipst)->ips_ip6_statistics.x.value.ui64 += (n)) 104 105 typedef struct ip6_stat { 106 kstat_named_t ip6_udp_fast_path; 107 kstat_named_t ip6_udp_slow_path; 108 kstat_named_t ip6_udp_fannorm; 109 kstat_named_t ip6_udp_fanmb; 110 kstat_named_t ip6_out_sw_cksum; 111 kstat_named_t ip6_in_sw_cksum; 112 kstat_named_t ip6_tcp_in_full_hw_cksum_err; 113 kstat_named_t ip6_tcp_in_part_hw_cksum_err; 114 kstat_named_t ip6_tcp_in_sw_cksum_err; 115 kstat_named_t ip6_tcp_out_sw_cksum_bytes; 116 kstat_named_t ip6_udp_in_full_hw_cksum_err; 117 kstat_named_t ip6_udp_in_part_hw_cksum_err; 118 kstat_named_t ip6_udp_in_sw_cksum_err; 119 kstat_named_t ip6_udp_out_sw_cksum_bytes; 120 kstat_named_t ip6_frag_mdt_pkt_out; 121 kstat_named_t ip6_frag_mdt_discarded; 122 kstat_named_t ip6_frag_mdt_allocfail; 123 kstat_named_t ip6_frag_mdt_addpdescfail; 124 kstat_named_t ip6_frag_mdt_allocd; 125 } ip6_stat_t; 126 127 typedef struct ire_stats { 128 uint64_t ire_stats_alloced; /* # of ires alloced */ 129 uint64_t ire_stats_freed; /* # of ires freed */ 130 uint64_t ire_stats_inserted; /* # of ires inserted in the bucket */ 131 uint64_t ire_stats_deleted; /* # of ires deleted from the bucket */ 132 } ire_stats_t; 133 134 135 /* 136 * IP stack instances 137 */ 138 struct ip_stack { 139 netstack_t *ips_netstack; /* Common netstack */ 140 141 struct ipparam_s *ips_param_arr; /* ndd variable table */ 142 struct ipndp_s *ips_ndp_arr; 143 144 mib2_ipIfStatsEntry_t ips_ip_mib; /* SNMP fixed size info */ 145 mib2_icmp_t ips_icmp_mib; 146 /* 147 * IPv6 mibs when the interface (ill) is not known. 148 * When the ill is known the per-interface mib in the ill is used. 149 */ 150 mib2_ipIfStatsEntry_t ips_ip6_mib; 151 mib2_ipv6IfIcmpEntry_t ips_icmp6_mib; 152 153 struct igmpstat ips_igmpstat; 154 155 kstat_t *ips_ip_mibkp; /* kstat exporting ip_mib data */ 156 kstat_t *ips_icmp_mibkp; /* kstat exporting icmp_mib data */ 157 kstat_t *ips_ip_kstat; 158 ip_stat_t ips_ip_statistics; 159 kstat_t *ips_ip6_kstat; 160 ip6_stat_t ips_ip6_statistics; 161 162 /* ip.c */ 163 krwlock_t ips_ip_g_nd_lock; 164 kmutex_t ips_igmp_timer_lock; 165 kmutex_t ips_mld_timer_lock; 166 kmutex_t ips_ip_mi_lock; 167 kmutex_t ips_ip_addr_avail_lock; 168 krwlock_t ips_ill_g_lock; 169 krwlock_t ips_ipsec_capab_ills_lock; 170 /* protects the list of IPsec capable ills */ 171 struct ipsec_capab_ill_s *ips_ipsec_capab_ills_ah; 172 struct ipsec_capab_ill_s *ips_ipsec_capab_ills_esp; 173 174 krwlock_t ips_ill_g_usesrc_lock; 175 176 /* Taskq dispatcher for capability operations */ 177 kmutex_t ips_capab_taskq_lock; 178 kcondvar_t ips_capab_taskq_cv; 179 list_t ips_capab_taskq_list; 180 kthread_t *ips_capab_taskq_thread; 181 boolean_t ips_capab_taskq_quit; 182 183 /* ipclassifier.c - keep in ip_stack_t */ 184 /* ipclassifier hash tables */ 185 struct connf_s *ips_rts_clients; 186 struct connf_s *ips_ipcl_conn_fanout; 187 struct connf_s *ips_ipcl_bind_fanout; 188 struct connf_s *ips_ipcl_proto_fanout; 189 struct connf_s *ips_ipcl_proto_fanout_v6; 190 struct connf_s *ips_ipcl_udp_fanout; 191 struct connf_s *ips_ipcl_raw_fanout; 192 uint_t ips_ipcl_conn_fanout_size; 193 uint_t ips_ipcl_bind_fanout_size; 194 uint_t ips_ipcl_udp_fanout_size; 195 uint_t ips_ipcl_raw_fanout_size; 196 struct connf_s *ips_ipcl_globalhash_fanout; 197 int ips_conn_g_index; 198 199 /* ip.c */ 200 /* Following protected by igmp_timer_lock */ 201 int ips_igmp_time_to_next; /* Time since last timeout */ 202 int ips_igmp_timer_scheduled_last; 203 int ips_igmp_deferred_next; 204 timeout_id_t ips_igmp_timeout_id; 205 boolean_t ips_igmp_timer_setter_active; 206 207 /* Following protected by mld_timer_lock */ 208 int ips_mld_time_to_next; /* Time since last timeout */ 209 int ips_mld_timer_scheduled_last; 210 int ips_mld_deferred_next; 211 timeout_id_t ips_mld_timeout_id; 212 boolean_t ips_mld_timer_setter_active; 213 214 /* Protected by igmp_slowtimeout_lock */ 215 timeout_id_t ips_igmp_slowtimeout_id; 216 kmutex_t ips_igmp_slowtimeout_lock; 217 218 /* Protected by mld_slowtimeout_lock */ 219 timeout_id_t ips_mld_slowtimeout_id; 220 kmutex_t ips_mld_slowtimeout_lock; 221 222 /* IPv4 forwarding table */ 223 struct radix_node_head *ips_ip_ftable; 224 225 /* This is dynamically allocated in ip_ire_init */ 226 struct irb *ips_ip_cache_table; 227 228 #define IPV6_ABITS 128 229 #define IP6_MASK_TABLE_SIZE (IPV6_ABITS + 1) /* 129 ptrs */ 230 231 struct irb *ips_ip_forwarding_table_v6[IP6_MASK_TABLE_SIZE]; 232 /* This is dynamically allocated in ip_ire_init */ 233 struct irb *ips_ip_cache_table_v6; 234 235 uint32_t ips_ire_handle; 236 /* 237 * ire_ft_init_lock is used while initializing ip_forwarding_table 238 * dynamically in ire_add. 239 */ 240 kmutex_t ips_ire_ft_init_lock; 241 kmutex_t ips_ire_handle_lock; /* Protects ire_handle */ 242 243 uint32_t ips_ip_cache_table_size; 244 uint32_t ips_ip6_cache_table_size; 245 uint32_t ips_ip6_ftable_hash_size; 246 247 ire_stats_t ips_ire_stats_v4; /* IPv4 ire statistics */ 248 ire_stats_t ips_ire_stats_v6; /* IPv6 ire statistics */ 249 250 /* pending binds */ 251 mblk_t *ips_ip6_asp_pending_ops; 252 mblk_t *ips_ip6_asp_pending_ops_tail; 253 254 /* Synchronize updates with table usage */ 255 mblk_t *ips_ip6_asp_pending_update; /* pending table updates */ 256 257 boolean_t ips_ip6_asp_uip; /* table update in progress */ 258 kmutex_t ips_ip6_asp_lock; /* protect all the above */ 259 uint32_t ips_ip6_asp_refcnt; /* outstanding references */ 260 261 struct ip6_asp *ips_ip6_asp_table; 262 /* The number of policy entries in the table */ 263 uint_t ips_ip6_asp_table_count; 264 265 int ips_ip_g_forward; 266 int ips_ipv6_forward; 267 268 time_t ips_ip_g_frag_timeout; 269 clock_t ips_ip_g_frag_timo_ms; 270 271 struct conn_s *ips_ip_g_mrouter; 272 273 /* Time since last icmp_pkt_err */ 274 clock_t ips_icmp_pkt_err_last; 275 /* Number of packets sent in burst */ 276 uint_t ips_icmp_pkt_err_sent; 277 278 /* Protected by ip_mi_lock */ 279 void *ips_ip_g_head; /* Instance Data List Head */ 280 281 caddr_t ips_ip_g_nd; /* Named Dispatch List Head */ 282 283 /* Multirouting stuff */ 284 /* Interval (in ms) between consecutive 'bad MTU' warnings */ 285 hrtime_t ips_ip_multirt_log_interval; 286 /* Time since last warning issued. */ 287 hrtime_t ips_multirt_bad_mtu_last_time; 288 289 struct cgtp_filter_ops *ips_ip_cgtp_filter_ops; /* CGTP hooks */ 290 boolean_t ips_ip_cgtp_filter; /* Enable/disable CGTP hooks */ 291 292 kmutex_t ips_ip_trash_timer_lock; 293 timeout_id_t ips_ip_ire_expire_id; /* IRE expiration timer. */ 294 struct ipsq_s *ips_ipsq_g_head; 295 uint_t ips_ill_index; /* Used to assign interface indicies */ 296 /* When set search for unused index */ 297 boolean_t ips_ill_index_wrap; 298 299 clock_t ips_ip_ire_arp_time_elapsed; 300 /* Time since IRE cache last flushed */ 301 clock_t ips_ip_ire_rd_time_elapsed; 302 /* ... redirect IREs last flushed */ 303 clock_t ips_ip_ire_pmtu_time_elapsed; 304 /* Time since path mtu increase */ 305 306 uint_t ips_ip_redirect_cnt; 307 /* Num of redirect routes in ftable */ 308 uint_t ips_ipv6_ire_default_count; 309 /* Number of IPv6 IRE_DEFAULT entries */ 310 uint_t ips_ipv6_ire_default_index; 311 /* Walking IPv6 index used to mod in */ 312 313 uint_t ips_loopback_packets; 314 315 /* NDP/NCE structures for IPv4 and IPv6 */ 316 struct ndp_g_s *ips_ndp4; 317 struct ndp_g_s *ips_ndp6; 318 319 /* ip_mroute stuff */ 320 kmutex_t ips_ip_g_mrouter_mutex; 321 322 struct mrtstat *ips_mrtstat; /* Stats for netstat */ 323 int ips_saved_ip_g_forward; 324 325 /* numvifs is only a hint about the max interface being used. */ 326 ushort_t ips_numvifs; 327 kmutex_t ips_numvifs_mutex; 328 329 struct vif *ips_vifs; 330 struct mfcb *ips_mfcs; /* kernel routing table */ 331 struct tbf *ips_tbfs; 332 /* 333 * One-back cache used to locate a tunnel's vif, 334 * given a datagram's src ip address. 335 */ 336 ipaddr_t ips_last_encap_src; 337 struct vif *ips_last_encap_vif; 338 kmutex_t ips_last_encap_lock; /* Protects the above */ 339 340 /* 341 * reg_vif_num is protected by numvifs_mutex 342 */ 343 /* Whether or not special PIM assert processing is enabled. */ 344 ushort_t ips_reg_vif_num; /* Index to Register vif */ 345 int ips_pim_assert; 346 347 union ill_g_head_u *ips_ill_g_heads; /* ILL List Head */ 348 349 kstat_t *ips_loopback_ksp; 350 351 struct idl_s *ips_conn_drain_list; /* Array of conn drain lists */ 352 uint_t ips_conn_drain_list_cnt; /* Count of conn_drain_list */ 353 int ips_conn_drain_list_index; /* Next drain_list */ 354 355 /* 356 * ID used to assign next free one. 357 * Increases by one. Once it wraps we search for an unused ID. 358 */ 359 uint_t ips_ip_src_id; 360 boolean_t ips_srcid_wrapped; 361 362 struct srcid_map *ips_srcid_head; 363 krwlock_t ips_srcid_lock; 364 365 uint64_t ips_ipif_g_seqid; 366 union phyint_list_u *ips_phyint_g_list; /* start of phyint list */ 367 368 /* ip_neti.c */ 369 hook_family_t ips_ipv4root; 370 hook_family_t ips_ipv6root; 371 372 /* 373 * Hooks for firewalling 374 */ 375 hook_event_t ips_ip4_physical_in_event; 376 hook_event_t ips_ip4_physical_out_event; 377 hook_event_t ips_ip4_forwarding_event; 378 hook_event_t ips_ip4_loopback_in_event; 379 hook_event_t ips_ip4_loopback_out_event; 380 hook_event_t ips_ip4_nic_events; 381 hook_event_t ips_ip6_physical_in_event; 382 hook_event_t ips_ip6_physical_out_event; 383 hook_event_t ips_ip6_forwarding_event; 384 hook_event_t ips_ip6_loopback_in_event; 385 hook_event_t ips_ip6_loopback_out_event; 386 hook_event_t ips_ip6_nic_events; 387 388 hook_event_token_t ips_ipv4firewall_physical_in; 389 hook_event_token_t ips_ipv4firewall_physical_out; 390 hook_event_token_t ips_ipv4firewall_forwarding; 391 hook_event_token_t ips_ipv4firewall_loopback_in; 392 hook_event_token_t ips_ipv4firewall_loopback_out; 393 hook_event_token_t ips_ipv4nicevents; 394 hook_event_token_t ips_ipv6firewall_physical_in; 395 hook_event_token_t ips_ipv6firewall_physical_out; 396 hook_event_token_t ips_ipv6firewall_forwarding; 397 hook_event_token_t ips_ipv6firewall_loopback_in; 398 hook_event_token_t ips_ipv6firewall_loopback_out; 399 hook_event_token_t ips_ipv6nicevents; 400 401 net_handle_t ips_ipv4_net_data; 402 net_handle_t ips_ipv6_net_data; 403 404 boolean_t ips_ipobs_enabled; 405 list_t ips_ipobs_cb_list; 406 kmutex_t ips_ipobs_cb_lock; 407 uint_t ips_ipobs_cb_nwalkers; 408 kcondvar_t ips_ipobs_cb_cv; 409 410 struct __ldi_ident *ips_ldi_ident; 411 412 /* ipmp.c */ 413 krwlock_t ips_ipmp_lock; 414 mod_hash_t *ips_ipmp_grp_hash; 415 416 /* igmp.c */ 417 /* multicast restart timers thread logic */ 418 kmutex_t ips_mrt_lock; 419 uint_t ips_mrt_flags; 420 kcondvar_t ips_mrt_cv; 421 kcondvar_t ips_mrt_done_cv; 422 kthread_t *ips_mrt_thread; 423 }; 424 typedef struct ip_stack ip_stack_t; 425 426 /* Finding an ip_stack_t */ 427 #define CONNQ_TO_IPST(_q) (Q_TO_CONN(_q)->conn_netstack->netstack_ip) 428 #define ILLQ_TO_IPST(_q) (((ill_t *)(_q)->q_ptr)->ill_ipst) 429 #define PHYINT_TO_IPST(phyi) ((phyi)->phyint_ipsq->ipsq_ipst) 430 431 #else /* _KERNEL */ 432 typedef int ip_stack_t; 433 #endif /* _KERNEL */ 434 435 #ifdef __cplusplus 436 } 437 #endif 438 439 #endif /* _INET_IP_STACK_H */ 440