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 /*
28 * Copyright 2019 Joyent, Inc.
29 */
30
31 #ifndef _INET_IP_STACK_H
32 #define _INET_IP_STACK_H
33
34 #ifdef __cplusplus
35 extern "C" {
36 #endif
37
38 #include <sys/netstack.h>
39 #include <netinet/igmp_var.h>
40 #include <sys/modhash.h>
41
42 #ifdef _KERNEL
43 #include <sys/list.h>
44
45
46 /*
47 * IP statistics.
48 */
49 #define IP_STAT(ipst, x) ((ipst)->ips_ip_statistics.x.value.ui64++)
50 #define IP_STAT_UPDATE(ipst, x, n) \
51 ((ipst)->ips_ip_statistics.x.value.ui64 += (n))
52
53 typedef struct ip_stat {
54 kstat_named_t ip_udp_fannorm;
55 kstat_named_t ip_udp_fanmb;
56 kstat_named_t ip_recv_pullup;
57 kstat_named_t ip_db_ref;
58 kstat_named_t ip_notaligned;
59 kstat_named_t ip_multimblk;
60 kstat_named_t ip_opt;
61 kstat_named_t ipsec_proto_ahesp;
62 kstat_named_t ip_conn_flputbq;
63 kstat_named_t ip_conn_walk_drain;
64 kstat_named_t ip_out_sw_cksum;
65 kstat_named_t ip_out_sw_cksum_bytes;
66 kstat_named_t ip_in_sw_cksum;
67 kstat_named_t ip_ire_reclaim_calls;
68 kstat_named_t ip_ire_reclaim_deleted;
69 kstat_named_t ip_nce_reclaim_calls;
70 kstat_named_t ip_nce_reclaim_deleted;
71 kstat_named_t ip_nce_mcast_reclaim_calls;
72 kstat_named_t ip_nce_mcast_reclaim_deleted;
73 kstat_named_t ip_nce_mcast_reclaim_tqfail;
74 kstat_named_t ip_dce_reclaim_calls;
75 kstat_named_t ip_dce_reclaim_deleted;
76 kstat_named_t ip_tcp_in_full_hw_cksum_err;
77 kstat_named_t ip_tcp_in_part_hw_cksum_err;
78 kstat_named_t ip_tcp_in_sw_cksum_err;
79 kstat_named_t ip_udp_in_full_hw_cksum_err;
80 kstat_named_t ip_udp_in_part_hw_cksum_err;
81 kstat_named_t ip_udp_in_sw_cksum_err;
82 kstat_named_t conn_in_recvdstaddr;
83 kstat_named_t conn_in_recvopts;
84 kstat_named_t conn_in_recvif;
85 kstat_named_t conn_in_recvslla;
86 kstat_named_t conn_in_recvucred;
87 kstat_named_t conn_in_recvttl;
88 kstat_named_t conn_in_recvhopopts;
89 kstat_named_t conn_in_recvhoplimit;
90 kstat_named_t conn_in_recvdstopts;
91 kstat_named_t conn_in_recvrthdrdstopts;
92 kstat_named_t conn_in_recvrthdr;
93 kstat_named_t conn_in_recvpktinfo;
94 kstat_named_t conn_in_recvtclass;
95 kstat_named_t conn_in_timestamp;
96 } ip_stat_t;
97
98
99 /*
100 * IP6 statistics.
101 */
102 #define IP6_STAT(ipst, x) ((ipst)->ips_ip6_statistics.x.value.ui64++)
103 #define IP6_STAT_UPDATE(ipst, x, n) \
104 ((ipst)->ips_ip6_statistics.x.value.ui64 += (n))
105
106 typedef struct ip6_stat {
107 kstat_named_t ip6_udp_fannorm;
108 kstat_named_t ip6_udp_fanmb;
109 kstat_named_t ip6_recv_pullup;
110 kstat_named_t ip6_db_ref;
111 kstat_named_t ip6_notaligned;
112 kstat_named_t ip6_multimblk;
113 kstat_named_t ipsec_proto_ahesp;
114 kstat_named_t ip6_out_sw_cksum;
115 kstat_named_t ip6_out_sw_cksum_bytes;
116 kstat_named_t ip6_in_sw_cksum;
117 kstat_named_t ip6_tcp_in_full_hw_cksum_err;
118 kstat_named_t ip6_tcp_in_part_hw_cksum_err;
119 kstat_named_t ip6_tcp_in_sw_cksum_err;
120 kstat_named_t ip6_udp_in_full_hw_cksum_err;
121 kstat_named_t ip6_udp_in_part_hw_cksum_err;
122 kstat_named_t ip6_udp_in_sw_cksum_err;
123 } ip6_stat_t;
124
125 typedef struct ire_stats {
126 uint64_t ire_stats_alloced; /* # of ires alloced */
127 uint64_t ire_stats_freed; /* # of ires freed */
128 uint64_t ire_stats_inserted; /* # of ires inserted in the bucket */
129 uint64_t ire_stats_deleted; /* # of ires deleted from the bucket */
130 } ire_stats_t;
131
132 #define TX_FANOUT_SIZE 128
133 #define IDLHASHINDEX(X) \
134 ((((uintptr_t)(X) >> 2) + ((uintptr_t)(X) >> 9)) & (TX_FANOUT_SIZE - 1))
135
136 /* Data structure to represent addresses */
137 typedef struct srcid_map {
138 struct srcid_map *sm_next;
139 in6_addr_t sm_addr; /* Local address */
140 uint_t sm_srcid; /* source id */
141 uint_t sm_refcnt; /* > 1 ipif with same addr? */
142 zoneid_t sm_zoneid; /* zone id */
143 } srcid_map_t;
144
145 /*
146 * IP stack instances
147 */
148 struct ip_stack {
149 netstack_t *ips_netstack; /* Common netstack */
150
151 uint_t ips_src_generation; /* Both IPv4 and IPv6 */
152
153 struct mod_prop_info_s *ips_propinfo_tbl; /* ip tunables table */
154
155 mib2_ipIfStatsEntry_t ips_ip_mib; /* SNMP fixed size info */
156 mib2_icmp_t ips_icmp_mib;
157 /*
158 * IPv6 mibs when the interface (ill) is not known.
159 * When the ill is known the per-interface mib in the ill is used.
160 */
161 mib2_ipIfStatsEntry_t ips_ip6_mib;
162 mib2_ipv6IfIcmpEntry_t ips_icmp6_mib;
163
164 struct igmpstat ips_igmpstat;
165
166 kstat_t *ips_ip_mibkp; /* kstat exporting ip_mib data */
167 kstat_t *ips_icmp_mibkp; /* kstat exporting icmp_mib data */
168 kstat_t *ips_ip_kstat;
169 ip_stat_t ips_ip_statistics;
170 kstat_t *ips_ip6_kstat;
171 ip6_stat_t ips_ip6_statistics;
172
173 /* ip.c */
174 kmutex_t ips_igmp_timer_lock;
175 kmutex_t ips_mld_timer_lock;
176 kmutex_t ips_ip_mi_lock;
177 kmutex_t ips_ip_addr_avail_lock;
178 krwlock_t ips_ill_g_lock;
179
180 krwlock_t ips_ill_g_usesrc_lock;
181
182 /* Taskq dispatcher for capability operations */
183 kmutex_t ips_capab_taskq_lock;
184 kcondvar_t ips_capab_taskq_cv;
185 mblk_t *ips_capab_taskq_head;
186 mblk_t *ips_capab_taskq_tail;
187 kthread_t *ips_capab_taskq_thread;
188 boolean_t ips_capab_taskq_quit;
189
190 /* ipclassifier.c - keep in ip_stack_t */
191 /* ipclassifier hash tables */
192 struct connf_s *ips_rts_clients;
193 struct connf_s *ips_ipcl_conn_fanout;
194 struct connf_s *ips_ipcl_bind_fanout;
195 struct connf_s *ips_ipcl_proto_fanout_v4;
196 struct connf_s *ips_ipcl_proto_fanout_v6;
197 struct connf_s *ips_ipcl_udp_fanout;
198 struct connf_s *ips_ipcl_raw_fanout; /* RAW SCTP sockets */
199 struct connf_s *ips_ipcl_iptun_fanout;
200 uint_t ips_ipcl_conn_fanout_size;
201 uint_t ips_ipcl_bind_fanout_size;
202 uint_t ips_ipcl_udp_fanout_size;
203 uint_t ips_ipcl_raw_fanout_size;
204 uint_t ips_ipcl_iptun_fanout_size;
205 struct connf_s *ips_ipcl_globalhash_fanout;
206 int ips_conn_g_index;
207
208 /* ip.c */
209 /* Following protected by igmp_timer_lock */
210 int ips_igmp_time_to_next; /* Time since last timeout */
211 int ips_igmp_timer_scheduled_last;
212 int ips_igmp_deferred_next;
213 timeout_id_t ips_igmp_timeout_id;
214 boolean_t ips_igmp_timer_setter_active;
215 boolean_t ips_igmp_timer_quiesce;
216
217 /* Following protected by mld_timer_lock */
218 int ips_mld_time_to_next; /* Time since last timeout */
219 int ips_mld_timer_scheduled_last;
220 int ips_mld_deferred_next;
221 timeout_id_t ips_mld_timeout_id;
222 boolean_t ips_mld_timer_setter_active;
223 boolean_t ips_mld_timer_quiesce;
224
225 /* Protected by igmp_slowtimeout_lock */
226 timeout_id_t ips_igmp_slowtimeout_id;
227 kmutex_t ips_igmp_slowtimeout_lock;
228 boolean_t ips_igmp_slowtimeout_quiesce;
229
230 /* Protected by mld_slowtimeout_lock */
231 timeout_id_t ips_mld_slowtimeout_id;
232 kmutex_t ips_mld_slowtimeout_lock;
233 boolean_t ips_mld_slowtimeout_quiesce;
234
235 /* IPv4 forwarding table */
236 struct radix_node_head *ips_ip_ftable;
237
238 #define IPV6_ABITS 128
239 #define IP6_MASK_TABLE_SIZE (IPV6_ABITS + 1) /* 129 ptrs */
240 struct irb *ips_ip_forwarding_table_v6[IP6_MASK_TABLE_SIZE];
241
242 /*
243 * ire_ft_init_lock is used while initializing ip_forwarding_table
244 * dynamically in ire_add.
245 */
246 kmutex_t ips_ire_ft_init_lock;
247
248 /*
249 * This is the IPv6 counterpart of RADIX_NODE_HEAD_LOCK. It is used
250 * to prevent adds and deletes while we are doing a ftable_lookup
251 * and extracting the ire_generation.
252 */
253 krwlock_t ips_ip6_ire_head_lock;
254
255 uint32_t ips_ip6_ftable_hash_size;
256
257 ire_stats_t ips_ire_stats_v4; /* IPv4 ire statistics */
258 ire_stats_t ips_ire_stats_v6; /* IPv6 ire statistics */
259
260 /* Count how many condemned objects for kmem_cache callbacks */
261 uint32_t ips_num_ire_condemned;
262 uint32_t ips_num_nce_condemned;
263 uint32_t ips_num_dce_condemned;
264
265 struct ire_s *ips_ire_reject_v4; /* For unreachable dests */
266 struct ire_s *ips_ire_reject_v6; /* For unreachable dests */
267 struct ire_s *ips_ire_blackhole_v4; /* For temporary failures */
268 struct ire_s *ips_ire_blackhole_v6; /* For temporary failures */
269
270 /* ips_ire_dep_lock protects ire_dep_* relationship between IREs */
271 krwlock_t ips_ire_dep_lock;
272
273 /* Destination Cache Entries */
274 struct dce_s *ips_dce_default;
275 uint_t ips_dce_hashsize;
276 struct dcb_s *ips_dce_hash_v4;
277 struct dcb_s *ips_dce_hash_v6;
278 uint_t ips_dce_reclaim_needed;
279
280 /* pending binds */
281 mblk_t *ips_ip6_asp_pending_ops;
282 mblk_t *ips_ip6_asp_pending_ops_tail;
283
284 /* Synchronize updates with table usage */
285 mblk_t *ips_ip6_asp_pending_update; /* pending table updates */
286
287 boolean_t ips_ip6_asp_uip; /* table update in progress */
288 kmutex_t ips_ip6_asp_lock; /* protect all the above */
289 uint32_t ips_ip6_asp_refcnt; /* outstanding references */
290
291 struct ip6_asp *ips_ip6_asp_table;
292 /* The number of policy entries in the table */
293 uint_t ips_ip6_asp_table_count;
294
295 struct conn_s *ips_ip_g_mrouter;
296
297 /* Time since last icmp_pkt_err */
298 clock_t ips_icmp_pkt_err_last;
299 /* Number of packets sent in burst */
300 uint_t ips_icmp_pkt_err_sent;
301
302 /* Protected by ip_mi_lock */
303 void *ips_ip_g_head; /* IP Instance Data List Head */
304 void *ips_arp_g_head; /* ARP Instance Data List Head */
305
306 /* Multirouting stuff */
307 /* Interval (in ms) between consecutive 'bad MTU' warnings */
308 hrtime_t ips_ip_multirt_log_interval;
309 /* Time since last warning issued. */
310 hrtime_t ips_multirt_bad_mtu_last_time;
311
312 /*
313 * CGTP hooks. Enabling and disabling of hooks is controlled by an
314 * IP tunable 'ips_ip_cgtp_filter'.
315 */
316 struct cgtp_filter_ops *ips_ip_cgtp_filter_ops;
317
318 struct ipsq_s *ips_ipsq_g_head;
319 uint_t ips_ill_index; /* Used to assign interface indicies */
320 /* When set search for unused index */
321 boolean_t ips_ill_index_wrap;
322
323 uint_t ips_loopback_packets;
324
325 /* NDP/NCE structures for IPv4 and IPv6 */
326 struct ndp_g_s *ips_ndp4;
327 struct ndp_g_s *ips_ndp6;
328
329 /* ip_mroute stuff */
330 kmutex_t ips_ip_g_mrouter_mutex;
331
332 struct mrtstat *ips_mrtstat; /* Stats for netstat */
333 int ips_saved_ip_forwarding;
334
335 /* numvifs is only a hint about the max interface being used. */
336 ushort_t ips_numvifs;
337 kmutex_t ips_numvifs_mutex;
338
339 struct vif *ips_vifs;
340 struct mfcb *ips_mfcs; /* kernel routing table */
341 struct tbf *ips_tbfs;
342 /*
343 * One-back cache used to locate a tunnel's vif,
344 * given a datagram's src ip address.
345 */
346 ipaddr_t ips_last_encap_src;
347 struct vif *ips_last_encap_vif;
348 kmutex_t ips_last_encap_lock; /* Protects the above */
349
350 /*
351 * reg_vif_num is protected by numvifs_mutex
352 */
353 /* Whether or not special PIM assert processing is enabled. */
354 ushort_t ips_reg_vif_num; /* Index to Register vif */
355 int ips_pim_assert;
356
357 union ill_g_head_u *ips_ill_g_heads; /* ILL List Head */
358
359 kstat_t *ips_loopback_ksp;
360
361 /* Array of conn drain lists */
362 struct idl_tx_list_s *ips_idl_tx_list;
363 uint_t ips_conn_drain_list_cnt; /* Count of conn_drain_list */
364
365 /*
366 * ID used to assign next free one.
367 * Increases by one. Once it wraps we search for an unused ID.
368 */
369 uint_t ips_ip_src_id;
370 boolean_t ips_srcid_wrapped;
371
372 struct srcid_map *ips_srcid_head;
373 krwlock_t ips_srcid_lock;
374
375 uint64_t ips_ipif_g_seqid; /* Used only for sctp_addr.c */
376 union phyint_list_u *ips_phyint_g_list; /* start of phyint list */
377
378 /* ip_netinfo.c */
379 hook_family_t ips_ipv4root;
380 hook_family_t ips_ipv6root;
381 hook_family_t ips_arproot;
382
383 net_handle_t ips_ipv4_net_data;
384 net_handle_t ips_ipv6_net_data;
385 net_handle_t ips_arp_net_data;
386
387 /*
388 * Hooks for firewalling
389 */
390 hook_event_t ips_ip4_physical_in_event;
391 hook_event_t ips_ip4_physical_out_event;
392 hook_event_t ips_ip4_forwarding_event;
393 hook_event_t ips_ip4_loopback_in_event;
394 hook_event_t ips_ip4_loopback_out_event;
395
396 hook_event_t ips_ip6_physical_in_event;
397 hook_event_t ips_ip6_physical_out_event;
398 hook_event_t ips_ip6_forwarding_event;
399 hook_event_t ips_ip6_loopback_in_event;
400 hook_event_t ips_ip6_loopback_out_event;
401
402 hook_event_t ips_arp_physical_in_event;
403 hook_event_t ips_arp_physical_out_event;
404 hook_event_t ips_arp_nic_events;
405
406 hook_event_token_t ips_ipv4firewall_physical_in;
407 hook_event_token_t ips_ipv4firewall_physical_out;
408 hook_event_token_t ips_ipv4firewall_forwarding;
409 hook_event_token_t ips_ipv4firewall_loopback_in;
410 hook_event_token_t ips_ipv4firewall_loopback_out;
411
412 hook_event_token_t ips_ipv6firewall_physical_in;
413 hook_event_token_t ips_ipv6firewall_physical_out;
414 hook_event_token_t ips_ipv6firewall_forwarding;
415 hook_event_token_t ips_ipv6firewall_loopback_in;
416 hook_event_token_t ips_ipv6firewall_loopback_out;
417
418 hook_event_t ips_ip4_nic_events;
419 hook_event_t ips_ip6_nic_events;
420 hook_event_token_t ips_ipv4nicevents;
421 hook_event_token_t ips_ipv6nicevents;
422
423 hook_event_token_t ips_arp_physical_in;
424 hook_event_token_t ips_arp_physical_out;
425 hook_event_token_t ips_arpnicevents;
426
427 net_handle_t ips_ip4_observe_pr;
428 net_handle_t ips_ip6_observe_pr;
429 hook_event_t ips_ip4_observe;
430 hook_event_t ips_ip6_observe;
431 hook_event_token_t ips_ipv4observing;
432 hook_event_token_t ips_ipv6observing;
433
434 struct __ldi_ident *ips_ldi_ident;
435
436 /* ipmp.c */
437 krwlock_t ips_ipmp_lock;
438 mod_hash_t *ips_ipmp_grp_hash;
439
440 };
441 typedef struct ip_stack ip_stack_t;
442
443 /* Finding an ip_stack_t */
444 #define CONNQ_TO_IPST(_q) (Q_TO_CONN(_q)->conn_netstack->netstack_ip)
445 #define ILLQ_TO_IPST(_q) (((ill_t *)(_q)->q_ptr)->ill_ipst)
446 #define PHYINT_TO_IPST(phyi) ((phyi)->phyint_ipsq->ipsq_ipst)
447
448 #else /* _KERNEL */
449 typedef int ip_stack_t;
450 #endif /* _KERNEL */
451
452 #ifdef __cplusplus
453 }
454 #endif
455
456 #endif /* _INET_IP_STACK_H */