Home | History | Annotate | Download | only in inet
      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 #define	TX_FANOUT_SIZE	128
    135 #define	IDLHASHINDEX(X)	\
    136 	((((uintptr_t)(X) >> 2) + ((uintptr_t)(X) >> 9)) & (TX_FANOUT_SIZE - 1))
    137 
    138 /* Data structure to represent addresses */
    139 typedef struct srcid_map {
    140 	struct srcid_map	*sm_next;
    141 	in6_addr_t		sm_addr;	/* Local address */
    142 	uint_t			sm_srcid;	/* source id */
    143 	uint_t			sm_refcnt;	/* > 1 ipif with same addr? */
    144 	zoneid_t		sm_zoneid;	/* zone id */
    145 } srcid_map_t;
    146 
    147 /*
    148  * IP stack instances
    149  */
    150 struct ip_stack {
    151 	netstack_t	*ips_netstack;	/* Common netstack */
    152 
    153 	struct ipparam_s	*ips_param_arr; 	/* ndd variable table */
    154 	struct ipndp_s		*ips_ndp_arr;
    155 
    156 	mib2_ipIfStatsEntry_t	ips_ip_mib;	/* SNMP fixed size info */
    157 	mib2_icmp_t	ips_icmp_mib;
    158 	/*
    159 	 * IPv6 mibs when the interface (ill) is not known.
    160 	 * When the ill is known the per-interface mib in the ill is used.
    161 	 */
    162 	mib2_ipIfStatsEntry_t	ips_ip6_mib;
    163 	mib2_ipv6IfIcmpEntry_t	ips_icmp6_mib;
    164 
    165 	struct igmpstat		ips_igmpstat;
    166 
    167 	kstat_t		*ips_ip_mibkp;	/* kstat exporting ip_mib data */
    168 	kstat_t		*ips_icmp_mibkp; /* kstat exporting icmp_mib data */
    169 	kstat_t		*ips_ip_kstat;
    170 	ip_stat_t	ips_ip_statistics;
    171 	kstat_t		*ips_ip6_kstat;
    172 	ip6_stat_t	ips_ip6_statistics;
    173 
    174 /* ip.c */
    175 	krwlock_t	ips_ip_g_nd_lock;
    176 	kmutex_t	ips_igmp_timer_lock;
    177 	kmutex_t	ips_mld_timer_lock;
    178 	kmutex_t	ips_ip_mi_lock;
    179 	kmutex_t	ips_ip_addr_avail_lock;
    180 	krwlock_t	ips_ill_g_lock;
    181 	krwlock_t	ips_ipsec_capab_ills_lock;
    182 				/* protects the list of IPsec capable ills */
    183 	struct ipsec_capab_ill_s *ips_ipsec_capab_ills_ah;
    184 	struct ipsec_capab_ill_s *ips_ipsec_capab_ills_esp;
    185 
    186 	krwlock_t	ips_ill_g_usesrc_lock;
    187 
    188 	/* Taskq dispatcher for capability operations */
    189 	kmutex_t	ips_capab_taskq_lock;
    190 	kcondvar_t	ips_capab_taskq_cv;
    191 	mblk_t		*ips_capab_taskq_head;
    192 	mblk_t		*ips_capab_taskq_tail;
    193 	kthread_t	*ips_capab_taskq_thread;
    194 	boolean_t	ips_capab_taskq_quit;
    195 
    196 /* ipclassifier.c - keep in ip_stack_t */
    197 	/* ipclassifier hash tables */
    198 	struct connf_s	*ips_rts_clients;
    199 	struct connf_s	*ips_ipcl_conn_fanout;
    200 	struct connf_s	*ips_ipcl_bind_fanout;
    201 	struct connf_s	*ips_ipcl_proto_fanout;
    202 	struct connf_s	*ips_ipcl_proto_fanout_v6;
    203 	struct connf_s	*ips_ipcl_udp_fanout;
    204 	struct connf_s	*ips_ipcl_raw_fanout;
    205 	struct connf_s	*ips_ipcl_iptun_fanout;
    206 	uint_t		ips_ipcl_conn_fanout_size;
    207 	uint_t		ips_ipcl_bind_fanout_size;
    208 	uint_t		ips_ipcl_udp_fanout_size;
    209 	uint_t		ips_ipcl_raw_fanout_size;
    210 	uint_t		ips_ipcl_iptun_fanout_size;
    211 	struct connf_s	*ips_ipcl_globalhash_fanout;
    212 	int		ips_conn_g_index;
    213 
    214 /* ip.c */
    215 	/* Following protected by igmp_timer_lock */
    216 	int 		ips_igmp_time_to_next;	/* Time since last timeout */
    217 	int 		ips_igmp_timer_scheduled_last;
    218 	int		ips_igmp_deferred_next;
    219 	timeout_id_t	ips_igmp_timeout_id;
    220 	boolean_t	ips_igmp_timer_setter_active;
    221 
    222 	/* Following protected by mld_timer_lock */
    223 	int 		ips_mld_time_to_next;	/* Time since last timeout */
    224 	int 		ips_mld_timer_scheduled_last;
    225 	int		ips_mld_deferred_next;
    226 	timeout_id_t	ips_mld_timeout_id;
    227 	boolean_t	ips_mld_timer_setter_active;
    228 
    229 	/* Protected by igmp_slowtimeout_lock */
    230 	timeout_id_t	ips_igmp_slowtimeout_id;
    231 	kmutex_t	ips_igmp_slowtimeout_lock;
    232 
    233 	/* Protected by mld_slowtimeout_lock */
    234 	timeout_id_t	ips_mld_slowtimeout_id;
    235 	kmutex_t	ips_mld_slowtimeout_lock;
    236 
    237 	/* IPv4 forwarding table */
    238 	struct radix_node_head *ips_ip_ftable;
    239 
    240 	/* This is dynamically allocated in ip_ire_init */
    241 	struct irb	 *ips_ip_cache_table;
    242 
    243 #define	IPV6_ABITS		128
    244 #define	IP6_MASK_TABLE_SIZE	(IPV6_ABITS + 1)	/* 129 ptrs */
    245 
    246 	struct irb	*ips_ip_forwarding_table_v6[IP6_MASK_TABLE_SIZE];
    247 	/* This is dynamically allocated in ip_ire_init */
    248 	struct irb	*ips_ip_cache_table_v6;
    249 
    250 	uint32_t	ips_ire_handle;
    251 	/*
    252 	 * ire_ft_init_lock is used while initializing ip_forwarding_table
    253 	 * dynamically in ire_add.
    254 	 */
    255 	kmutex_t	ips_ire_ft_init_lock;
    256 	kmutex_t	ips_ire_handle_lock;	/* Protects ire_handle */
    257 
    258 	uint32_t	ips_ip_cache_table_size;
    259 	uint32_t	ips_ip6_cache_table_size;
    260 	uint32_t	ips_ip6_ftable_hash_size;
    261 
    262 	ire_stats_t 	ips_ire_stats_v4;	/* IPv4 ire statistics */
    263 	ire_stats_t 	ips_ire_stats_v6;	/* IPv6 ire statistics */
    264 
    265 	/* pending binds */
    266 	mblk_t		*ips_ip6_asp_pending_ops;
    267 	mblk_t		*ips_ip6_asp_pending_ops_tail;
    268 
    269 	/* Synchronize updates with table usage */
    270 	mblk_t		*ips_ip6_asp_pending_update; /* pending table updates */
    271 
    272 	boolean_t	ips_ip6_asp_uip;	/* table update in progress */
    273 	kmutex_t	ips_ip6_asp_lock;	/* protect all the above */
    274 	uint32_t	ips_ip6_asp_refcnt;	/* outstanding references */
    275 
    276 	struct ip6_asp	*ips_ip6_asp_table;
    277 	/* The number of policy entries in the table */
    278 	uint_t		ips_ip6_asp_table_count;
    279 
    280 	int		ips_ip_g_forward;
    281 	int		ips_ipv6_forward;
    282 
    283 	time_t		ips_ip_g_frag_timeout;
    284 	clock_t		ips_ip_g_frag_timo_ms;
    285 	time_t		ips_ipv6_frag_timeout;
    286 	clock_t		ips_ipv6_frag_timo_ms;
    287 
    288 	struct conn_s	*ips_ip_g_mrouter;
    289 
    290 	/* Time since last icmp_pkt_err */
    291 	clock_t		ips_icmp_pkt_err_last;
    292 	/* Number of packets sent in burst */
    293 	uint_t		ips_icmp_pkt_err_sent;
    294 
    295 	/* Protected by ip_mi_lock */
    296 	void		*ips_ip_g_head;		/* Instance Data List Head */
    297 
    298 	caddr_t		ips_ip_g_nd;		/* Named Dispatch List Head */
    299 
    300 	/* Multirouting stuff */
    301 	/* Interval (in ms) between consecutive 'bad MTU' warnings */
    302 	hrtime_t	ips_ip_multirt_log_interval;
    303 	/* Time since last warning issued. */
    304 	hrtime_t	ips_multirt_bad_mtu_last_time;
    305 
    306 	struct cgtp_filter_ops *ips_ip_cgtp_filter_ops;	/* CGTP hooks */
    307 	boolean_t	ips_ip_cgtp_filter;	/* Enable/disable CGTP hooks */
    308 
    309 	kmutex_t	ips_ip_trash_timer_lock;
    310 	timeout_id_t	ips_ip_ire_expire_id;	/* IRE expiration timer. */
    311 	struct ipsq_s	*ips_ipsq_g_head;
    312 	uint_t		ips_ill_index;	/* Used to assign interface indicies */
    313 	/* When set search for unused index */
    314 	boolean_t	ips_ill_index_wrap;
    315 
    316 	clock_t		ips_ip_ire_arp_time_elapsed;
    317 			/* Time since IRE cache last flushed */
    318 	clock_t		ips_ip_ire_rd_time_elapsed;
    319 			/* ... redirect IREs last flushed */
    320 	clock_t		ips_ip_ire_pmtu_time_elapsed;
    321 			/* Time since path mtu increase */
    322 
    323 	uint_t		ips_ip_redirect_cnt;
    324 			/* Num of redirect routes in ftable */
    325 	uint_t		ips_ipv6_ire_default_count;
    326 			/* Number of IPv6 IRE_DEFAULT entries */
    327 	uint_t		ips_ipv6_ire_default_index;
    328 			/* Walking IPv6 index used to mod in */
    329 
    330 	uint_t		ips_loopback_packets;
    331 
    332 	/* NDP/NCE structures for IPv4 and IPv6 */
    333 	struct ndp_g_s	*ips_ndp4;
    334 	struct ndp_g_s	*ips_ndp6;
    335 
    336 	/* ip_mroute stuff */
    337 	kmutex_t	ips_ip_g_mrouter_mutex;
    338 
    339 	struct mrtstat	*ips_mrtstat;	/* Stats for netstat */
    340 	int		ips_saved_ip_g_forward;
    341 
    342 	/* numvifs is only a hint about the max interface being used. */
    343 	ushort_t	ips_numvifs;
    344 	kmutex_t	ips_numvifs_mutex;
    345 
    346 	struct vif	*ips_vifs;
    347 	struct mfcb	*ips_mfcs;	/* kernel routing table	*/
    348 	struct tbf	*ips_tbfs;
    349 	/*
    350 	 * One-back cache used to locate a tunnel's vif,
    351 	 * given a datagram's src ip address.
    352 	 */
    353 	ipaddr_t	ips_last_encap_src;
    354 	struct vif	*ips_last_encap_vif;
    355 	kmutex_t	ips_last_encap_lock;	/* Protects the above */
    356 
    357 	/*
    358 	 * reg_vif_num is protected by numvifs_mutex
    359 	 */
    360 	/* Whether or not special PIM assert processing is enabled. */
    361 	ushort_t	ips_reg_vif_num; 	/* Index to Register vif */
    362 	int		ips_pim_assert;
    363 
    364 	union ill_g_head_u *ips_ill_g_heads;   /* ILL List Head */
    365 
    366 	kstat_t		*ips_loopback_ksp;
    367 
    368 	/* Array of conn drain lists */
    369 	struct idl_tx_list_s	*ips_idl_tx_list;
    370 	uint_t		ips_conn_drain_list_cnt; /* Count of conn_drain_list */
    371 
    372 	/*
    373 	 * ID used to assign next free one.
    374 	 * Increases by one. Once it wraps we search for an unused ID.
    375 	 */
    376 	uint_t		ips_ip_src_id;
    377 	boolean_t	ips_srcid_wrapped;
    378 
    379 	struct srcid_map *ips_srcid_head;
    380 	krwlock_t	ips_srcid_lock;
    381 
    382 	uint64_t	ips_ipif_g_seqid;
    383 	union phyint_list_u *ips_phyint_g_list;	/* start of phyint list */
    384 
    385 /* ip_neti.c */
    386 	hook_family_t	ips_ipv4root;
    387 	hook_family_t	ips_ipv6root;
    388 
    389 	net_handle_t		ips_ipv4_net_data;
    390 	net_handle_t		ips_ipv6_net_data;
    391 
    392 	/*
    393 	 * Hooks for firewalling
    394 	 */
    395 	hook_event_t		ips_ip4_physical_in_event;
    396 	hook_event_t		ips_ip4_physical_out_event;
    397 	hook_event_t		ips_ip4_forwarding_event;
    398 	hook_event_t		ips_ip4_loopback_in_event;
    399 	hook_event_t		ips_ip4_loopback_out_event;
    400 	hook_event_t		ips_ip6_physical_in_event;
    401 	hook_event_t		ips_ip6_physical_out_event;
    402 	hook_event_t		ips_ip6_forwarding_event;
    403 	hook_event_t		ips_ip6_loopback_in_event;
    404 	hook_event_t		ips_ip6_loopback_out_event;
    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 	hook_event_token_t	ips_ipv6firewall_physical_in;
    412 	hook_event_token_t	ips_ipv6firewall_physical_out;
    413 	hook_event_token_t	ips_ipv6firewall_forwarding;
    414 	hook_event_token_t	ips_ipv6firewall_loopback_in;
    415 	hook_event_token_t	ips_ipv6firewall_loopback_out;
    416 
    417 	hook_event_t		ips_ip4_nic_events;
    418 	hook_event_t		ips_ip6_nic_events;
    419 	hook_event_token_t	ips_ipv4nicevents;
    420 	hook_event_token_t	ips_ipv6nicevents;
    421 
    422 	net_handle_t		ips_ip4_observe_pr;
    423 	net_handle_t		ips_ip6_observe_pr;
    424 	hook_event_t		ips_ip4_observe;
    425 	hook_event_t		ips_ip6_observe;
    426 	hook_event_token_t	ips_ipv4observing;
    427 	hook_event_token_t	ips_ipv6observing;
    428 
    429 	struct __ldi_ident	*ips_ldi_ident;
    430 
    431 /* ipmp.c */
    432 	krwlock_t		ips_ipmp_lock;
    433 	mod_hash_t		*ips_ipmp_grp_hash;
    434 
    435 /* igmp.c */
    436 	/* multicast restart timers thread logic */
    437 	kmutex_t		ips_mrt_lock;
    438 	uint_t			ips_mrt_flags;
    439 	kcondvar_t		ips_mrt_cv;
    440 	kcondvar_t		ips_mrt_done_cv;
    441 	kthread_t		*ips_mrt_thread;
    442 };
    443 typedef struct ip_stack ip_stack_t;
    444 
    445 /* Finding an ip_stack_t */
    446 #define	CONNQ_TO_IPST(_q)	(Q_TO_CONN(_q)->conn_netstack->netstack_ip)
    447 #define	ILLQ_TO_IPST(_q)	(((ill_t *)(_q)->q_ptr)->ill_ipst)
    448 #define	PHYINT_TO_IPST(phyi)	((phyi)->phyint_ipsq->ipsq_ipst)
    449 
    450 #else /* _KERNEL */
    451 typedef int ip_stack_t;
    452 #endif /* _KERNEL */
    453 
    454 #ifdef	__cplusplus
    455 }
    456 #endif
    457 
    458 #endif	/* _INET_IP_STACK_H */
    459