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 2010 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 /* 43 * IP statistics. 44 */ 45 #define IP_STAT(ipst, x) ((ipst)->ips_ip_statistics.x.value.ui64++) 46 #define IP_STAT_UPDATE(ipst, x, n) \ 47 ((ipst)->ips_ip_statistics.x.value.ui64 += (n)) 48 49 typedef struct ip_stat { 50 kstat_named_t ip_udp_fannorm; 51 kstat_named_t ip_udp_fanmb; 52 kstat_named_t ip_recv_pullup; 53 kstat_named_t ip_db_ref; 54 kstat_named_t ip_notaligned; 55 kstat_named_t ip_multimblk; 56 kstat_named_t ip_opt; 57 kstat_named_t ipsec_proto_ahesp; 58 kstat_named_t ip_conn_flputbq; 59 kstat_named_t ip_conn_walk_drain; 60 kstat_named_t ip_out_sw_cksum; 61 kstat_named_t ip_out_sw_cksum_bytes; 62 kstat_named_t ip_in_sw_cksum; 63 kstat_named_t ip_ire_reclaim_calls; 64 kstat_named_t ip_ire_reclaim_deleted; 65 kstat_named_t ip_nce_reclaim_calls; 66 kstat_named_t ip_nce_reclaim_deleted; 67 kstat_named_t ip_dce_reclaim_calls; 68 kstat_named_t ip_dce_reclaim_deleted; 69 kstat_named_t ip_tcp_in_full_hw_cksum_err; 70 kstat_named_t ip_tcp_in_part_hw_cksum_err; 71 kstat_named_t ip_tcp_in_sw_cksum_err; 72 kstat_named_t ip_udp_in_full_hw_cksum_err; 73 kstat_named_t ip_udp_in_part_hw_cksum_err; 74 kstat_named_t ip_udp_in_sw_cksum_err; 75 kstat_named_t conn_in_recvdstaddr; 76 kstat_named_t conn_in_recvopts; 77 kstat_named_t conn_in_recvif; 78 kstat_named_t conn_in_recvslla; 79 kstat_named_t conn_in_recvucred; 80 kstat_named_t conn_in_recvttl; 81 kstat_named_t conn_in_recvhopopts; 82 kstat_named_t conn_in_recvhoplimit; 83 kstat_named_t conn_in_recvdstopts; 84 kstat_named_t conn_in_recvrthdrdstopts; 85 kstat_named_t conn_in_recvrthdr; 86 kstat_named_t conn_in_recvpktinfo; 87 kstat_named_t conn_in_recvtclass; 88 kstat_named_t conn_in_timestamp; 89 } ip_stat_t; 90 91 92 /* 93 * IP6 statistics. 94 */ 95 #define IP6_STAT(ipst, x) ((ipst)->ips_ip6_statistics.x.value.ui64++) 96 #define IP6_STAT_UPDATE(ipst, x, n) \ 97 ((ipst)->ips_ip6_statistics.x.value.ui64 += (n)) 98 99 typedef struct ip6_stat { 100 kstat_named_t ip6_udp_fannorm; 101 kstat_named_t ip6_udp_fanmb; 102 kstat_named_t ip6_recv_pullup; 103 kstat_named_t ip6_db_ref; 104 kstat_named_t ip6_notaligned; 105 kstat_named_t ip6_multimblk; 106 kstat_named_t ipsec_proto_ahesp; 107 kstat_named_t ip6_out_sw_cksum; 108 kstat_named_t ip6_out_sw_cksum_bytes; 109 kstat_named_t ip6_in_sw_cksum; 110 kstat_named_t ip6_tcp_in_full_hw_cksum_err; 111 kstat_named_t ip6_tcp_in_part_hw_cksum_err; 112 kstat_named_t ip6_tcp_in_sw_cksum_err; 113 kstat_named_t ip6_udp_in_full_hw_cksum_err; 114 kstat_named_t ip6_udp_in_part_hw_cksum_err; 115 kstat_named_t ip6_udp_in_sw_cksum_err; 116 } ip6_stat_t; 117 118 typedef struct ire_stats { 119 uint64_t ire_stats_alloced; /* # of ires alloced */ 120 uint64_t ire_stats_freed; /* # of ires freed */ 121 uint64_t ire_stats_inserted; /* # of ires inserted in the bucket */ 122 uint64_t ire_stats_deleted; /* # of ires deleted from the bucket */ 123 } ire_stats_t; 124 125 #define TX_FANOUT_SIZE 128 126 #define IDLHASHINDEX(X) \ 127 ((((uintptr_t)(X) >> 2) + ((uintptr_t)(X) >> 9)) & (TX_FANOUT_SIZE - 1)) 128 129 /* Data structure to represent addresses */ 130 typedef struct srcid_map { 131 struct srcid_map *sm_next; 132 in6_addr_t sm_addr; /* Local address */ 133 uint_t sm_srcid; /* source id */ 134 uint_t sm_refcnt; /* > 1 ipif with same addr? */ 135 zoneid_t sm_zoneid; /* zone id */ 136 } srcid_map_t; 137 138 /* 139 * IP stack instances 140 */ 141 struct ip_stack { 142 netstack_t *ips_netstack; /* Common netstack */ 143 144 uint_t ips_src_generation; /* Both IPv4 and IPv6 */ 145 146 struct mod_prop_info_s *ips_propinfo_tbl; /* ip tunables table */ 147 148 mib2_ipIfStatsEntry_t ips_ip_mib; /* SNMP fixed size info */ 149 mib2_icmp_t ips_icmp_mib; 150 /* 151 * IPv6 mibs when the interface (ill) is not known. 152 * When the ill is known the per-interface mib in the ill is used. 153 */ 154 mib2_ipIfStatsEntry_t ips_ip6_mib; 155 mib2_ipv6IfIcmpEntry_t ips_icmp6_mib; 156 157 struct igmpstat ips_igmpstat; 158 159 kstat_t *ips_ip_mibkp; /* kstat exporting ip_mib data */ 160 kstat_t *ips_icmp_mibkp; /* kstat exporting icmp_mib data */ 161 kstat_t *ips_ip_kstat; 162 ip_stat_t ips_ip_statistics; 163 kstat_t *ips_ip6_kstat; 164 ip6_stat_t ips_ip6_statistics; 165 166 /* ip.c */ 167 kmutex_t ips_igmp_timer_lock; 168 kmutex_t ips_mld_timer_lock; 169 kmutex_t ips_ip_mi_lock; 170 kmutex_t ips_ip_addr_avail_lock; 171 krwlock_t ips_ill_g_lock; 172 173 krwlock_t ips_ill_g_usesrc_lock; 174 175 /* Taskq dispatcher for capability operations */ 176 kmutex_t ips_capab_taskq_lock; 177 kcondvar_t ips_capab_taskq_cv; 178 mblk_t *ips_capab_taskq_head; 179 mblk_t *ips_capab_taskq_tail; 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_v4; 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; /* RAW SCTP sockets */ 192 struct connf_s *ips_ipcl_iptun_fanout; 193 uint_t ips_ipcl_conn_fanout_size; 194 uint_t ips_ipcl_bind_fanout_size; 195 uint_t ips_ipcl_udp_fanout_size; 196 uint_t ips_ipcl_raw_fanout_size; 197 uint_t ips_ipcl_iptun_fanout_size; 198 struct connf_s *ips_ipcl_globalhash_fanout; 199 int ips_conn_g_index; 200 201 /* ip.c */ 202 /* Following protected by igmp_timer_lock */ 203 int ips_igmp_time_to_next; /* Time since last timeout */ 204 int ips_igmp_timer_scheduled_last; 205 int ips_igmp_deferred_next; 206 timeout_id_t ips_igmp_timeout_id; 207 boolean_t ips_igmp_timer_setter_active; 208 209 /* Following protected by mld_timer_lock */ 210 int ips_mld_time_to_next; /* Time since last timeout */ 211 int ips_mld_timer_scheduled_last; 212 int ips_mld_deferred_next; 213 timeout_id_t ips_mld_timeout_id; 214 boolean_t ips_mld_timer_setter_active; 215 216 /* Protected by igmp_slowtimeout_lock */ 217 timeout_id_t ips_igmp_slowtimeout_id; 218 kmutex_t ips_igmp_slowtimeout_lock; 219 220 /* Protected by mld_slowtimeout_lock */ 221 timeout_id_t ips_mld_slowtimeout_id; 222 kmutex_t ips_mld_slowtimeout_lock; 223 224 /* IPv4 forwarding table */ 225 struct radix_node_head *ips_ip_ftable; 226 227 #define IPV6_ABITS 128 228 #define IP6_MASK_TABLE_SIZE (IPV6_ABITS + 1) /* 129 ptrs */ 229 struct irb *ips_ip_forwarding_table_v6[IP6_MASK_TABLE_SIZE]; 230 231 /* 232 * ire_ft_init_lock is used while initializing ip_forwarding_table 233 * dynamically in ire_add. 234 */ 235 kmutex_t ips_ire_ft_init_lock; 236 237 /* 238 * This is the IPv6 counterpart of RADIX_NODE_HEAD_LOCK. It is used 239 * to prevent adds and deletes while we are doing a ftable_lookup 240 * and extracting the ire_generation. 241 */ 242 krwlock_t ips_ip6_ire_head_lock; 243 244 uint32_t ips_ip6_ftable_hash_size; 245 246 ire_stats_t ips_ire_stats_v4; /* IPv4 ire statistics */ 247 ire_stats_t ips_ire_stats_v6; /* IPv6 ire statistics */ 248 249 /* Count how many condemned objects for kmem_cache callbacks */ 250 uint32_t ips_num_ire_condemned; 251 uint32_t ips_num_nce_condemned; 252 uint32_t ips_num_dce_condemned; 253 254 struct ire_s *ips_ire_reject_v4; /* For unreachable dests */ 255 struct ire_s *ips_ire_reject_v6; /* For unreachable dests */ 256 struct ire_s *ips_ire_blackhole_v4; /* For temporary failures */ 257 struct ire_s *ips_ire_blackhole_v6; /* For temporary failures */ 258 259 /* ips_ire_dep_lock protects ire_dep_* relationship between IREs */ 260 krwlock_t ips_ire_dep_lock; 261 262 /* Destination Cache Entries */ 263 struct dce_s *ips_dce_default; 264 uint_t ips_dce_hashsize; 265 struct dcb_s *ips_dce_hash_v4; 266 struct dcb_s *ips_dce_hash_v6; 267 uint_t ips_dce_reclaim_needed; 268 269 /* pending binds */ 270 mblk_t *ips_ip6_asp_pending_ops; 271 mblk_t *ips_ip6_asp_pending_ops_tail; 272 273 /* Synchronize updates with table usage */ 274 mblk_t *ips_ip6_asp_pending_update; /* pending table updates */ 275 276 boolean_t ips_ip6_asp_uip; /* table update in progress */ 277 kmutex_t ips_ip6_asp_lock; /* protect all the above */ 278 uint32_t ips_ip6_asp_refcnt; /* outstanding references */ 279 280 struct ip6_asp *ips_ip6_asp_table; 281 /* The number of policy entries in the table */ 282 uint_t ips_ip6_asp_table_count; 283 284 struct conn_s *ips_ip_g_mrouter; 285 286 /* Time since last icmp_pkt_err */ 287 clock_t ips_icmp_pkt_err_last; 288 /* Number of packets sent in burst */ 289 uint_t ips_icmp_pkt_err_sent; 290 291 /* Protected by ip_mi_lock */ 292 void *ips_ip_g_head; /* IP Instance Data List Head */ 293 void *ips_arp_g_head; /* ARP Instance Data List Head */ 294 295 /* Multirouting stuff */ 296 /* Interval (in ms) between consecutive 'bad MTU' warnings */ 297 hrtime_t ips_ip_multirt_log_interval; 298 /* Time since last warning issued. */ 299 hrtime_t ips_multirt_bad_mtu_last_time; 300 301 /* 302 * CGTP hooks. Enabling and disabling of hooks is controlled by an 303 * IP tunable 'ips_ip_cgtp_filter'. 304 */ 305 struct cgtp_filter_ops *ips_ip_cgtp_filter_ops; 306 307 struct ipsq_s *ips_ipsq_g_head; 308 uint_t ips_ill_index; /* Used to assign interface indicies */ 309 /* When set search for unused index */ 310 boolean_t ips_ill_index_wrap; 311 312 uint_t ips_loopback_packets; 313 314 /* NDP/NCE structures for IPv4 and IPv6 */ 315 struct ndp_g_s *ips_ndp4; 316 struct ndp_g_s *ips_ndp6; 317 318 /* ip_mroute stuff */ 319 kmutex_t ips_ip_g_mrouter_mutex; 320 321 struct mrtstat *ips_mrtstat; /* Stats for netstat */ 322 int ips_saved_ip_forwarding; 323 324 /* numvifs is only a hint about the max interface being used. */ 325 ushort_t ips_numvifs; 326 kmutex_t ips_numvifs_mutex; 327 328 struct vif *ips_vifs; 329 struct mfcb *ips_mfcs; /* kernel routing table */ 330 struct tbf *ips_tbfs; 331 /* 332 * One-back cache used to locate a tunnel's vif, 333 * given a datagram's src ip address. 334 */ 335 ipaddr_t ips_last_encap_src; 336 struct vif *ips_last_encap_vif; 337 kmutex_t ips_last_encap_lock; /* Protects the above */ 338 339 /* 340 * reg_vif_num is protected by numvifs_mutex 341 */ 342 /* Whether or not special PIM assert processing is enabled. */ 343 ushort_t ips_reg_vif_num; /* Index to Register vif */ 344 int ips_pim_assert; 345 346 union ill_g_head_u *ips_ill_g_heads; /* ILL List Head */ 347 348 kstat_t *ips_loopback_ksp; 349 350 /* Array of conn drain lists */ 351 struct idl_tx_list_s *ips_idl_tx_list; 352 uint_t ips_conn_drain_list_cnt; /* Count of conn_drain_list */ 353 354 /* 355 * ID used to assign next free one. 356 * Increases by one. Once it wraps we search for an unused ID. 357 */ 358 uint_t ips_ip_src_id; 359 boolean_t ips_srcid_wrapped; 360 361 struct srcid_map *ips_srcid_head; 362 krwlock_t ips_srcid_lock; 363 364 uint64_t ips_ipif_g_seqid; /* Used only for sctp_addr.c */ 365 union phyint_list_u *ips_phyint_g_list; /* start of phyint list */ 366 367 /* ip_netinfo.c */ 368 hook_family_t ips_ipv4root; 369 hook_family_t ips_ipv6root; 370 hook_family_t ips_arproot; 371 372 net_handle_t ips_ipv4_net_data; 373 net_handle_t ips_ipv6_net_data; 374 net_handle_t ips_arp_net_data; 375 376 /* 377 * Hooks for firewalling 378 */ 379 hook_event_t ips_ip4_physical_in_event; 380 hook_event_t ips_ip4_physical_out_event; 381 hook_event_t ips_ip4_forwarding_event; 382 hook_event_t ips_ip4_loopback_in_event; 383 hook_event_t ips_ip4_loopback_out_event; 384 385 hook_event_t ips_ip6_physical_in_event; 386 hook_event_t ips_ip6_physical_out_event; 387 hook_event_t ips_ip6_forwarding_event; 388 hook_event_t ips_ip6_loopback_in_event; 389 hook_event_t ips_ip6_loopback_out_event; 390 391 hook_event_t ips_arp_physical_in_event; 392 hook_event_t ips_arp_physical_out_event; 393 hook_event_t ips_arp_nic_events; 394 395 hook_event_token_t ips_ipv4firewall_physical_in; 396 hook_event_token_t ips_ipv4firewall_physical_out; 397 hook_event_token_t ips_ipv4firewall_forwarding; 398 hook_event_token_t ips_ipv4firewall_loopback_in; 399 hook_event_token_t ips_ipv4firewall_loopback_out; 400 401 hook_event_token_t ips_ipv6firewall_physical_in; 402 hook_event_token_t ips_ipv6firewall_physical_out; 403 hook_event_token_t ips_ipv6firewall_forwarding; 404 hook_event_token_t ips_ipv6firewall_loopback_in; 405 hook_event_token_t ips_ipv6firewall_loopback_out; 406 407 hook_event_t ips_ip4_nic_events; 408 hook_event_t ips_ip6_nic_events; 409 hook_event_token_t ips_ipv4nicevents; 410 hook_event_token_t ips_ipv6nicevents; 411 412 hook_event_token_t ips_arp_physical_in; 413 hook_event_token_t ips_arp_physical_out; 414 hook_event_token_t ips_arpnicevents; 415 416 net_handle_t ips_ip4_observe_pr; 417 net_handle_t ips_ip6_observe_pr; 418 hook_event_t ips_ip4_observe; 419 hook_event_t ips_ip6_observe; 420 hook_event_token_t ips_ipv4observing; 421 hook_event_token_t ips_ipv6observing; 422 423 struct __ldi_ident *ips_ldi_ident; 424 425 /* ipmp.c */ 426 krwlock_t ips_ipmp_lock; 427 mod_hash_t *ips_ipmp_grp_hash; 428 429 }; 430 typedef struct ip_stack ip_stack_t; 431 432 /* Finding an ip_stack_t */ 433 #define CONNQ_TO_IPST(_q) (Q_TO_CONN(_q)->conn_netstack->netstack_ip) 434 #define ILLQ_TO_IPST(_q) (((ill_t *)(_q)->q_ptr)->ill_ipst) 435 #define PHYINT_TO_IPST(phyi) ((phyi)->phyint_ipsq->ipsq_ipst) 436 437 #else /* _KERNEL */ 438 typedef int ip_stack_t; 439 #endif /* _KERNEL */ 440 441 #ifdef __cplusplus 442 } 443 #endif 444 445 #endif /* _INET_IP_STACK_H */ 446