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         kstat_named_t   ip6_frag_mdt_pkt_out;
 117         kstat_named_t   ip6_frag_mdt_discarded;
 118         kstat_named_t   ip6_frag_mdt_allocfail;
 119         kstat_named_t   ip6_frag_mdt_addpdescfail;
 120         kstat_named_t   ip6_frag_mdt_allocd;
 121 } ip6_stat_t;
 122 
 123 typedef struct ire_stats {
 124         uint64_t ire_stats_alloced;     /* # of ires alloced */
 125         uint64_t ire_stats_freed;       /* # of ires freed */
 126         uint64_t ire_stats_inserted;    /* # of ires inserted in the bucket */
 127         uint64_t ire_stats_deleted;     /* # of ires deleted from the bucket */
 128 } ire_stats_t;
 129 
 130 #define TX_FANOUT_SIZE  128
 131 #define IDLHASHINDEX(X) \
 132         ((((uintptr_t)(X) >> 2) + ((uintptr_t)(X) >> 9)) & (TX_FANOUT_SIZE - 1))
 133 
 134 /* Data structure to represent addresses */
 135 typedef struct srcid_map {
 136         struct srcid_map        *sm_next;
 137         in6_addr_t              sm_addr;        /* Local address */
 138         uint_t                  sm_srcid;       /* source id */
 139         uint_t                  sm_refcnt;      /* > 1 ipif with same addr? */
 140         zoneid_t                sm_zoneid;      /* zone id */
 141 } srcid_map_t;
 142 
 143 /*
 144  * IP stack instances
 145  */
 146 struct ip_stack {
 147         netstack_t      *ips_netstack;  /* Common netstack */
 148 
 149         uint_t                  ips_src_generation;     /* Both IPv4 and IPv6 */
 150 
 151         struct mod_prop_info_s  *ips_propinfo_tbl;      /* ip tunables table */
 152 
 153         mib2_ipIfStatsEntry_t   ips_ip_mib;     /* SNMP fixed size info */
 154         mib2_icmp_t     ips_icmp_mib;
 155         /*
 156          * IPv6 mibs when the interface (ill) is not known.
 157          * When the ill is known the per-interface mib in the ill is used.
 158          */
 159         mib2_ipIfStatsEntry_t   ips_ip6_mib;
 160         mib2_ipv6IfIcmpEntry_t  ips_icmp6_mib;
 161 
 162         struct igmpstat         ips_igmpstat;
 163 
 164         kstat_t         *ips_ip_mibkp;  /* kstat exporting ip_mib data */
 165         kstat_t         *ips_icmp_mibkp; /* kstat exporting icmp_mib data */
 166         kstat_t         *ips_ip_kstat;
 167         ip_stat_t       ips_ip_statistics;
 168         kstat_t         *ips_ip6_kstat;
 169         ip6_stat_t      ips_ip6_statistics;
 170 
 171 /* ip.c */
 172         kmutex_t        ips_igmp_timer_lock;
 173         kmutex_t        ips_mld_timer_lock;
 174         kmutex_t        ips_ip_mi_lock;
 175         kmutex_t        ips_ip_addr_avail_lock;
 176         krwlock_t       ips_ill_g_lock;
 177 
 178         krwlock_t       ips_ill_g_usesrc_lock;
 179 
 180         /* Taskq dispatcher for capability operations */
 181         kmutex_t        ips_capab_taskq_lock;
 182         kcondvar_t      ips_capab_taskq_cv;
 183         mblk_t          *ips_capab_taskq_head;
 184         mblk_t          *ips_capab_taskq_tail;
 185         kthread_t       *ips_capab_taskq_thread;
 186         boolean_t       ips_capab_taskq_quit;
 187 
 188 /* ipclassifier.c - keep in ip_stack_t */
 189         /* ipclassifier hash tables */
 190         struct connf_s  *ips_rts_clients;
 191         struct connf_s  *ips_ipcl_conn_fanout;
 192         struct connf_s  *ips_ipcl_bind_fanout;
 193         struct connf_s  *ips_ipcl_proto_fanout_v4;
 194         struct connf_s  *ips_ipcl_proto_fanout_v6;
 195         struct connf_s  *ips_ipcl_udp_fanout;
 196         struct connf_s  *ips_ipcl_raw_fanout;           /* RAW SCTP sockets */
 197         struct connf_s  *ips_ipcl_iptun_fanout;
 198         struct connf_s  *ips_ipcl_dccp_conn_fanout;
 199         struct connf_s  *ips_ipcl_dccp_bind_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         uint_t          ips_ipcl_dccp_conn_fanout_size;
 206         uint_t          ips_ipcl_dccp_bind_fanout_size;
 207         struct connf_s  *ips_ipcl_globalhash_fanout;
 208         int             ips_conn_g_index;
 209 
 210 /* ip.c */
 211         /* Following protected by igmp_timer_lock */
 212         int             ips_igmp_time_to_next;  /* Time since last timeout */
 213         int             ips_igmp_timer_scheduled_last;
 214         int             ips_igmp_deferred_next;
 215         timeout_id_t    ips_igmp_timeout_id;
 216         boolean_t       ips_igmp_timer_setter_active;
 217 
 218         /* Following protected by mld_timer_lock */
 219         int             ips_mld_time_to_next;   /* Time since last timeout */
 220         int             ips_mld_timer_scheduled_last;
 221         int             ips_mld_deferred_next;
 222         timeout_id_t    ips_mld_timeout_id;
 223         boolean_t       ips_mld_timer_setter_active;
 224 
 225         /* Protected by igmp_slowtimeout_lock */
 226         timeout_id_t    ips_igmp_slowtimeout_id;
 227         kmutex_t        ips_igmp_slowtimeout_lock;
 228 
 229         /* Protected by mld_slowtimeout_lock */
 230         timeout_id_t    ips_mld_slowtimeout_id;
 231         kmutex_t        ips_mld_slowtimeout_lock;
 232 
 233         /* IPv4 forwarding table */
 234         struct radix_node_head *ips_ip_ftable;
 235 
 236 #define IPV6_ABITS              128
 237 #define IP6_MASK_TABLE_SIZE     (IPV6_ABITS + 1)        /* 129 ptrs */
 238         struct irb      *ips_ip_forwarding_table_v6[IP6_MASK_TABLE_SIZE];
 239 
 240         /*
 241          * ire_ft_init_lock is used while initializing ip_forwarding_table
 242          * dynamically in ire_add.
 243          */
 244         kmutex_t        ips_ire_ft_init_lock;
 245 
 246         /*
 247          * This is the IPv6 counterpart of RADIX_NODE_HEAD_LOCK. It is used
 248          * to prevent adds and deletes while we are doing a ftable_lookup
 249          * and extracting the ire_generation.
 250          */
 251         krwlock_t       ips_ip6_ire_head_lock;
 252 
 253         uint32_t        ips_ip6_ftable_hash_size;
 254 
 255         ire_stats_t     ips_ire_stats_v4;       /* IPv4 ire statistics */
 256         ire_stats_t     ips_ire_stats_v6;       /* IPv6 ire statistics */
 257 
 258         /* Count how many condemned objects for kmem_cache callbacks */
 259         uint32_t        ips_num_ire_condemned;
 260         uint32_t        ips_num_nce_condemned;
 261         uint32_t        ips_num_dce_condemned;
 262 
 263         struct ire_s    *ips_ire_reject_v4;     /* For unreachable dests */
 264         struct ire_s    *ips_ire_reject_v6;     /* For unreachable dests */
 265         struct ire_s    *ips_ire_blackhole_v4;  /* For temporary failures */
 266         struct ire_s    *ips_ire_blackhole_v6;  /* For temporary failures */
 267 
 268         /* ips_ire_dep_lock protects ire_dep_* relationship between IREs */
 269         krwlock_t       ips_ire_dep_lock;
 270 
 271         /* Destination Cache Entries */
 272         struct dce_s    *ips_dce_default;
 273         uint_t          ips_dce_hashsize;
 274         struct dcb_s    *ips_dce_hash_v4;
 275         struct dcb_s    *ips_dce_hash_v6;
 276 
 277         /* pending binds */
 278         mblk_t          *ips_ip6_asp_pending_ops;
 279         mblk_t          *ips_ip6_asp_pending_ops_tail;
 280 
 281         /* Synchronize updates with table usage */
 282         mblk_t          *ips_ip6_asp_pending_update; /* pending table updates */
 283 
 284         boolean_t       ips_ip6_asp_uip;        /* table update in progress */
 285         kmutex_t        ips_ip6_asp_lock;       /* protect all the above */
 286         uint32_t        ips_ip6_asp_refcnt;     /* outstanding references */
 287 
 288         struct ip6_asp  *ips_ip6_asp_table;
 289         /* The number of policy entries in the table */
 290         uint_t          ips_ip6_asp_table_count;
 291 
 292         struct conn_s   *ips_ip_g_mrouter;
 293 
 294         /* Time since last icmp_pkt_err */
 295         clock_t         ips_icmp_pkt_err_last;
 296         /* Number of packets sent in burst */
 297         uint_t          ips_icmp_pkt_err_sent;
 298 
 299         /* Protected by ip_mi_lock */
 300         void            *ips_ip_g_head; /* IP Instance Data List Head */
 301         void            *ips_arp_g_head; /* ARP Instance Data List Head */
 302 
 303         /* Multirouting stuff */
 304         /* Interval (in ms) between consecutive 'bad MTU' warnings */
 305         hrtime_t        ips_ip_multirt_log_interval;
 306         /* Time since last warning issued. */
 307         hrtime_t        ips_multirt_bad_mtu_last_time;
 308 
 309         /*
 310          * CGTP hooks. Enabling and disabling of hooks is controlled by an
 311          * IP tunable 'ips_ip_cgtp_filter'.
 312          */
 313         struct cgtp_filter_ops *ips_ip_cgtp_filter_ops;
 314 
 315         struct ipsq_s   *ips_ipsq_g_head;
 316         uint_t          ips_ill_index;  /* Used to assign interface indicies */
 317         /* When set search for unused index */
 318         boolean_t       ips_ill_index_wrap;
 319 
 320         uint_t          ips_loopback_packets;
 321 
 322         /* NDP/NCE structures for IPv4 and IPv6 */
 323         struct ndp_g_s  *ips_ndp4;
 324         struct ndp_g_s  *ips_ndp6;
 325 
 326         /* ip_mroute stuff */
 327         kmutex_t        ips_ip_g_mrouter_mutex;
 328 
 329         struct mrtstat  *ips_mrtstat;   /* Stats for netstat */
 330         int             ips_saved_ip_forwarding;
 331 
 332         /* numvifs is only a hint about the max interface being used. */
 333         ushort_t        ips_numvifs;
 334         kmutex_t        ips_numvifs_mutex;
 335 
 336         struct vif      *ips_vifs;
 337         struct mfcb     *ips_mfcs;      /* kernel routing table */
 338         struct tbf      *ips_tbfs;
 339         /*
 340          * One-back cache used to locate a tunnel's vif,
 341          * given a datagram's src ip address.
 342          */
 343         ipaddr_t        ips_last_encap_src;
 344         struct vif      *ips_last_encap_vif;
 345         kmutex_t        ips_last_encap_lock;    /* Protects the above */
 346 
 347         /*
 348          * reg_vif_num is protected by numvifs_mutex
 349          */
 350         /* Whether or not special PIM assert processing is enabled. */
 351         ushort_t        ips_reg_vif_num;        /* Index to Register vif */
 352         int             ips_pim_assert;
 353 
 354         union ill_g_head_u *ips_ill_g_heads;   /* ILL List Head */
 355 
 356         kstat_t         *ips_loopback_ksp;
 357 
 358         /* Array of conn drain lists */
 359         struct idl_tx_list_s    *ips_idl_tx_list;
 360         uint_t          ips_conn_drain_list_cnt; /* Count of conn_drain_list */
 361 
 362         /*
 363          * ID used to assign next free one.
 364          * Increases by one. Once it wraps we search for an unused ID.
 365          */
 366         uint_t          ips_ip_src_id;
 367         boolean_t       ips_srcid_wrapped;
 368 
 369         struct srcid_map *ips_srcid_head;
 370         krwlock_t       ips_srcid_lock;
 371 
 372         uint64_t        ips_ipif_g_seqid;       /* Used only for sctp_addr.c */
 373         union phyint_list_u *ips_phyint_g_list; /* start of phyint list */
 374 
 375 /* ip_netinfo.c */
 376         hook_family_t   ips_ipv4root;
 377         hook_family_t   ips_ipv6root;
 378         hook_family_t   ips_arproot;
 379 
 380         net_handle_t            ips_ipv4_net_data;
 381         net_handle_t            ips_ipv6_net_data;
 382         net_handle_t            ips_arp_net_data;
 383 
 384         /*
 385          * Hooks for firewalling
 386          */
 387         hook_event_t            ips_ip4_physical_in_event;
 388         hook_event_t            ips_ip4_physical_out_event;
 389         hook_event_t            ips_ip4_forwarding_event;
 390         hook_event_t            ips_ip4_loopback_in_event;
 391         hook_event_t            ips_ip4_loopback_out_event;
 392 
 393         hook_event_t            ips_ip6_physical_in_event;
 394         hook_event_t            ips_ip6_physical_out_event;
 395         hook_event_t            ips_ip6_forwarding_event;
 396         hook_event_t            ips_ip6_loopback_in_event;
 397         hook_event_t            ips_ip6_loopback_out_event;
 398 
 399         hook_event_t            ips_arp_physical_in_event;
 400         hook_event_t            ips_arp_physical_out_event;
 401         hook_event_t            ips_arp_nic_events;
 402 
 403         hook_event_token_t      ips_ipv4firewall_physical_in;
 404         hook_event_token_t      ips_ipv4firewall_physical_out;
 405         hook_event_token_t      ips_ipv4firewall_forwarding;
 406         hook_event_token_t      ips_ipv4firewall_loopback_in;
 407         hook_event_token_t      ips_ipv4firewall_loopback_out;
 408 
 409         hook_event_token_t      ips_ipv6firewall_physical_in;
 410         hook_event_token_t      ips_ipv6firewall_physical_out;
 411         hook_event_token_t      ips_ipv6firewall_forwarding;
 412         hook_event_token_t      ips_ipv6firewall_loopback_in;
 413         hook_event_token_t      ips_ipv6firewall_loopback_out;
 414 
 415         hook_event_t            ips_ip4_nic_events;
 416         hook_event_t            ips_ip6_nic_events;
 417         hook_event_token_t      ips_ipv4nicevents;
 418         hook_event_token_t      ips_ipv6nicevents;
 419 
 420         hook_event_token_t      ips_arp_physical_in;
 421         hook_event_token_t      ips_arp_physical_out;
 422         hook_event_token_t      ips_arpnicevents;
 423 
 424         net_handle_t            ips_ip4_observe_pr;
 425         net_handle_t            ips_ip6_observe_pr;
 426         hook_event_t            ips_ip4_observe;
 427         hook_event_t            ips_ip6_observe;
 428         hook_event_token_t      ips_ipv4observing;
 429         hook_event_token_t      ips_ipv6observing;
 430 
 431         struct __ldi_ident      *ips_ldi_ident;
 432 
 433 /* ipmp.c */
 434         krwlock_t               ips_ipmp_lock;
 435         mod_hash_t              *ips_ipmp_grp_hash;
 436 
 437 };
 438 typedef struct ip_stack ip_stack_t;
 439 
 440 /* Finding an ip_stack_t */
 441 #define CONNQ_TO_IPST(_q)       (Q_TO_CONN(_q)->conn_netstack->netstack_ip)
 442 #define ILLQ_TO_IPST(_q)        (((ill_t *)(_q)->q_ptr)->ill_ipst)
 443 #define PHYINT_TO_IPST(phyi)    ((phyi)->phyint_ipsq->ipsq_ipst)
 444 
 445 #else /* _KERNEL */
 446 typedef int ip_stack_t;
 447 #endif /* _KERNEL */
 448 
 449 #ifdef  __cplusplus
 450 }
 451 #endif
 452 
 453 #endif  /* _INET_IP_STACK_H */