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 * Copyright 2009 Sun Microsystems, Inc. All rights reserved. 23 * Use is subject to license terms. 24 */ 25 26 #ifndef _UDP_IMPL_H 27 #define _UDP_IMPL_H 28 29 /* 30 * UDP implementation private declarations. These interfaces are 31 * used to build the IP module and are not meant to be accessed 32 * by any modules except IP itself. They are undocumented and are 33 * subject to change without notice. 34 */ 35 36 #ifdef __cplusplus 37 extern "C" { 38 #endif 39 40 #ifdef _KERNEL 41 42 #include <sys/int_types.h> 43 #include <sys/netstack.h> 44 45 #include <netinet/in.h> 46 #include <netinet/ip6.h> 47 48 #include <inet/common.h> 49 #include <inet/ip.h> 50 #include <inet/optcom.h> 51 52 #define UDP_MOD_ID 5607 53 54 typedef struct udp_bits_s { 55 56 uint32_t 57 58 udpb_debug : 1, /* SO_DEBUG "socket" option. */ 59 udpb_dontroute : 1, /* SO_DONTROUTE "socket" option. */ 60 udpb_broadcast : 1, /* SO_BROADCAST "socket" option. */ 61 udpb_useloopback : 1, /* SO_USELOOPBACK "socket" option */ 62 63 udpb_reuseaddr : 1, /* SO_REUSEADDR "socket" option. */ 64 udpb_dgram_errind : 1, /* SO_DGRAM_ERRIND option */ 65 udpb_recvdstaddr : 1, /* IP_RECVDSTADDR option */ 66 udpb_recvopts : 1, /* IP_RECVOPTS option */ 67 68 udpb_unspec_source : 1, /* IP*_UNSPEC_SRC option */ 69 udpb_ip_recvpktinfo : 1, /* IPV6_RECVPKTINFO option */ 70 udpb_ipv6_recvhoplimit : 1, /* IPV6_RECVHOPLIMIT option */ 71 udpb_ipv6_recvhopopts : 1, /* IPV6_RECVHOPOPTS option */ 72 73 udpb_ipv6_recvdstopts : 1, /* IPV6_RECVDSTOPTS option */ 74 udpb_ipv6_recvrthdr : 1, /* IPV6_RECVRTHDR option */ 75 udpb_ipv6_recvtclass : 1, /* IPV6_RECVTCLASS */ 76 udpb_ipv6_recvpathmtu : 1, /* IPV6_RECVPATHMTU */ 77 78 udpb_anon_priv_bind : 1, 79 udpb_exclbind : 1, /* ``exclusive'' binding */ 80 udpb_recvif : 1, /* IP_RECVIF option */ 81 udpb_recvslla : 1, /* IP_RECVSLLA option */ 82 83 udpb_recvttl : 1, /* IP_RECVTTL option */ 84 udpb_recvucred : 1, /* IP_RECVUCRED option */ 85 udpb_old_ipv6_recvdstopts : 1, /* old form of IPV6_DSTOPTS */ 86 udpb_ipv6_recvrthdrdstopts : 1, /* IPV6_RECVRTHDRDSTOPTS */ 87 88 udpb_rcvhdr : 1, /* UDP_RCVHDR option */ 89 udpb_issocket : 1, /* socket mode */ 90 udpb_direct_sockfs : 1, /* direct calls to/from sockfs */ 91 udpb_timestamp : 1, /* SO_TIMESTAMP "socket" option */ 92 93 udpb_nat_t_endpoint : 1, /* UDP_NAT_T_ENDPOINT option */ 94 udpb_pad_to_bit_31 : 3; 95 } udp_bits_t; 96 97 #define udp_debug udp_bits.udpb_debug 98 #define udp_dontroute udp_bits.udpb_dontroute 99 #define udp_broadcast udp_bits.udpb_broadcast 100 #define udp_useloopback udp_bits.udpb_useloopback 101 102 #define udp_reuseaddr udp_bits.udpb_reuseaddr 103 #define udp_dgram_errind udp_bits.udpb_dgram_errind 104 #define udp_recvdstaddr udp_bits.udpb_recvdstaddr 105 #define udp_recvopts udp_bits.udpb_recvopts 106 107 #define udp_unspec_source udp_bits.udpb_unspec_source 108 #define udp_ip_recvpktinfo udp_bits.udpb_ip_recvpktinfo 109 #define udp_ipv6_recvhoplimit udp_bits.udpb_ipv6_recvhoplimit 110 #define udp_ipv6_recvhopopts udp_bits.udpb_ipv6_recvhopopts 111 112 #define udp_ipv6_recvdstopts udp_bits.udpb_ipv6_recvdstopts 113 #define udp_ipv6_recvrthdr udp_bits.udpb_ipv6_recvrthdr 114 #define udp_ipv6_recvtclass udp_bits.udpb_ipv6_recvtclass 115 #define udp_ipv6_recvpathmtu udp_bits.udpb_ipv6_recvpathmtu 116 117 #define udp_anon_priv_bind udp_bits.udpb_anon_priv_bind 118 #define udp_exclbind udp_bits.udpb_exclbind 119 #define udp_recvif udp_bits.udpb_recvif 120 #define udp_recvslla udp_bits.udpb_recvslla 121 122 #define udp_recvttl udp_bits.udpb_recvttl 123 #define udp_recvucred udp_bits.udpb_recvucred 124 #define udp_old_ipv6_recvdstopts udp_bits.udpb_old_ipv6_recvdstopts 125 #define udp_ipv6_recvrthdrdstopts udp_bits.udpb_ipv6_recvrthdrdstopts 126 127 #define udp_rcvhdr udp_bits.udpb_rcvhdr 128 #define udp_issocket udp_bits.udpb_issocket 129 #define udp_direct_sockfs udp_bits.udpb_direct_sockfs 130 #define udp_timestamp udp_bits.udpb_timestamp 131 132 #define udp_nat_t_endpoint udp_bits.udpb_nat_t_endpoint 133 134 /* 135 * Bind hash list size and hash function. It has to be a power of 2 for 136 * hashing. 137 */ 138 #define UDP_BIND_FANOUT_SIZE 512 139 #define UDP_BIND_HASH(lport, size) \ 140 ((ntohs((uint16_t)lport)) & (size - 1)) 141 142 /* UDP bind fanout hash structure. */ 143 typedef struct udp_fanout_s { 144 struct udp_s *uf_udp; 145 kmutex_t uf_lock; 146 #if defined(_LP64) || defined(_I32LPx) 147 char uf_pad[48]; 148 #else 149 char uf_pad[56]; 150 #endif 151 } udp_fanout_t; 152 153 /* 154 * dev_q is the write side queue of the entity below IP. 155 * If there is a module below IP, we can't optimize by looking 156 * at q_first of the queue below IP. If the driver is directly 157 * below IP and if the q_first is NULL, we optimize by not doing 158 * the canput check 159 */ 160 #define DEV_Q_FLOW_BLOCKED(dev_q) \ 161 (((dev_q)->q_next != NULL || (dev_q)->q_first != NULL) && \ 162 !canput(dev_q)) 163 164 /* Kstats */ 165 typedef struct udp_stat { /* Class "net" kstats */ 166 kstat_named_t udp_ip_send; 167 kstat_named_t udp_ip_ire_send; 168 kstat_named_t udp_ire_null; 169 kstat_named_t udp_drain; 170 kstat_named_t udp_sock_fallback; 171 kstat_named_t udp_rrw_busy; 172 kstat_named_t udp_rrw_msgcnt; 173 kstat_named_t udp_out_sw_cksum; 174 kstat_named_t udp_out_sw_cksum_bytes; 175 kstat_named_t udp_out_opt; 176 kstat_named_t udp_out_err_notconn; 177 kstat_named_t udp_out_err_output; 178 kstat_named_t udp_out_err_tudr; 179 kstat_named_t udp_in_pktinfo; 180 kstat_named_t udp_in_recvdstaddr; 181 kstat_named_t udp_in_recvopts; 182 kstat_named_t udp_in_recvif; 183 kstat_named_t udp_in_recvslla; 184 kstat_named_t udp_in_recvucred; 185 kstat_named_t udp_in_recvttl; 186 kstat_named_t udp_in_recvhopopts; 187 kstat_named_t udp_in_recvhoplimit; 188 kstat_named_t udp_in_recvdstopts; 189 kstat_named_t udp_in_recvrtdstopts; 190 kstat_named_t udp_in_recvrthdr; 191 kstat_named_t udp_in_recvpktinfo; 192 kstat_named_t udp_in_recvtclass; 193 kstat_named_t udp_in_timestamp; 194 kstat_named_t udp_ip_rcvpktinfo; 195 kstat_named_t udp_cookie_coll; 196 #ifdef DEBUG 197 kstat_named_t udp_data_conn; 198 kstat_named_t udp_data_notconn; 199 #endif 200 201 } udp_stat_t; 202 203 /* Named Dispatch Parameter Management Structure */ 204 typedef struct udpparam_s { 205 uint32_t udp_param_min; 206 uint32_t udp_param_max; 207 uint32_t udp_param_value; 208 char *udp_param_name; 209 } udpparam_t; 210 211 #define UDP_NUM_EPRIV_PORTS 64 212 213 /* 214 * UDP stack instances 215 */ 216 struct udp_stack { 217 netstack_t *us_netstack; /* Common netstack */ 218 219 uint_t us_bind_fanout_size; 220 udp_fanout_t *us_bind_fanout; 221 222 int us_num_epriv_ports; 223 in_port_t us_epriv_ports[UDP_NUM_EPRIV_PORTS]; 224 225 /* Hint not protected by any lock */ 226 in_port_t us_next_port_to_try; 227 228 IDP us_nd; /* Points to table of UDP ND variables. */ 229 udpparam_t *us_param_arr; /* ndd variable table */ 230 231 kstat_t *us_mibkp; /* kstats exporting mib data */ 232 kstat_t *us_kstat; 233 udp_stat_t us_statistics; 234 235 mib2_udp_t us_udp_mib; /* SNMP fixed size info */ 236 237 /* 238 * This controls the rate some ndd info report functions can be used 239 * by non-priviledged users. It stores the last time such info is 240 * requested. When those report functions are called again, this 241 * is checked with the current time and compare with the ndd param 242 * udp_ndd_get_info_interval. 243 */ 244 clock_t us_last_ndd_get_info_time; 245 246 /* 247 * The smallest anonymous port in the priviledged port range which UDP 248 * looks for free port. Use in the option UDP_ANONPRIVBIND. 249 */ 250 in_port_t us_min_anonpriv_port; 251 252 ldi_ident_t us_ldi_ident; 253 }; 254 255 typedef struct udp_stack udp_stack_t; 256 257 /* Internal udp control structure, one per open stream */ 258 typedef struct udp_s { 259 krwlock_t udp_rwlock; /* Protects most of udp_t */ 260 t_scalar_t udp_pending_op; /* The current TPI operation */ 261 /* 262 * Following fields up to udp_ipversion protected by conn_lock, 263 * and the fanout lock i.e.uf_lock. Need both locks to change the 264 * field, either lock is sufficient for reading the field. 265 */ 266 uint32_t udp_state; /* TPI state */ 267 in_port_t udp_port; /* Port bound to this stream */ 268 in_port_t udp_dstport; /* Connected port */ 269 in6_addr_t udp_v6src; /* Source address of this stream */ 270 in6_addr_t udp_bound_v6src; /* Explicitly bound address */ 271 in6_addr_t udp_v6dst; /* Connected destination */ 272 /* 273 * IP format that packets transmitted from this struct should use. 274 * Value can be IP4_VERSION or IPV6_VERSION. 275 */ 276 ushort_t udp_ipversion; 277 278 /* Written to only once at the time of opening the endpoint */ 279 sa_family_t udp_family; /* Family from socket() call */ 280 281 /* Following protected by udp_rwlock */ 282 uint32_t udp_flowinfo; /* Connected flow id and tclass */ 283 uint32_t udp_max_hdr_len; /* For write offset in stream head */ 284 uint32_t udp_ip_snd_options_len; /* Len of IPv4 options */ 285 uchar_t *udp_ip_snd_options; /* Ptr to IPv4 options */ 286 uint32_t udp_ip_rcv_options_len; /* Len of IPv4 options recvd */ 287 uchar_t *udp_ip_rcv_options; /* Ptr to IPv4 options recvd */ 288 uchar_t udp_multicast_ttl; /* IP*_MULTICAST_TTL/HOPS */ 289 ipaddr_t udp_multicast_if_addr; /* IP_MULTICAST_IF option */ 290 uint_t udp_multicast_if_index; /* IPV6_MULTICAST_IF option */ 291 int udp_bound_if; /* IP*_BOUND_IF option */ 292 293 /* Written to only once at the time of opening the endpoint */ 294 conn_t *udp_connp; 295 296 /* Following protected by udp_rwlock */ 297 udp_bits_t udp_bits; /* Bit fields defined above */ 298 uint8_t udp_type_of_service; /* IP_TOS option */ 299 uint8_t udp_ttl; /* TTL or hoplimit */ 300 ip6_pkt_t udp_sticky_ipp; /* Sticky options */ 301 uint8_t *udp_sticky_hdrs; /* Prebuilt IPv6 hdrs */ 302 uint_t udp_sticky_hdrs_len; /* Incl. ip6h and any ip6i */ 303 304 /* Following 2 fields protected by the uf_lock */ 305 struct udp_s *udp_bind_hash; /* Bind hash chain */ 306 struct udp_s **udp_ptpbhn; /* Pointer to previous bind hash next. */ 307 308 kmutex_t udp_drain_lock; /* lock for udp_rcv_list */ 309 /* Protected by udp_drain_lock */ 310 boolean_t udp_drain_qfull; /* drain queue is full */ 311 312 /* Following protected by udp_rwlock */ 313 mblk_t *udp_rcv_list_head; /* b_next chain of mblks */ 314 mblk_t *udp_rcv_list_tail; /* last mblk in chain */ 315 kmutex_t udp_recv_lock; /* recv lock */ 316 uint_t udp_rcv_cnt; /* total data in rcv_list */ 317 uint_t udp_rcv_msgcnt; /* total msgs in rcv_list */ 318 size_t udp_rcv_disply_hiwat; /* user's view of rcvbuf */ 319 size_t udp_rcv_hiwat; /* receive high watermark */ 320 size_t udp_rcv_lowat; /* receive low watermark */ 321 size_t udp_xmit_hiwat; /* Send buffer high watermark */ 322 size_t udp_xmit_lowat; /* Send buffer low watermark */ 323 uint_t udp_label_len; /* length of security label */ 324 uint_t udp_label_len_v6; /* len of v6 security label */ 325 in6_addr_t udp_v6lastdst; /* most recent destination */ 326 in_port_t udp_lastdstport; /* most recent dest port */ 327 328 uint64_t udp_open_time; /* time when this was opened */ 329 pid_t udp_open_pid; /* process id when this was opened */ 330 udp_stack_t *udp_us; /* Stack instance for zone */ 331 int udp_delayed_error; 332 mblk_t *udp_fallback_queue_head; 333 mblk_t *udp_fallback_queue_tail; 334 struct sockaddr_storage udp_delayed_addr; 335 } udp_t; 336 337 /* UDP Protocol header */ 338 /* UDP Protocol header aligned */ 339 typedef struct udpahdr_s { 340 in_port_t uha_src_port; /* Source port */ 341 in_port_t uha_dst_port; /* Destination port */ 342 uint16_t uha_length; /* UDP length */ 343 uint16_t uha_checksum; /* UDP checksum */ 344 } udpha_t; 345 346 #define us_wroff_extra us_param_arr[0].udp_param_value 347 #define us_ipv4_ttl us_param_arr[1].udp_param_value 348 #define us_ipv6_hoplimit us_param_arr[2].udp_param_value 349 #define us_smallest_nonpriv_port us_param_arr[3].udp_param_value 350 #define us_do_checksum us_param_arr[4].udp_param_value 351 #define us_smallest_anon_port us_param_arr[5].udp_param_value 352 #define us_largest_anon_port us_param_arr[6].udp_param_value 353 #define us_xmit_hiwat us_param_arr[7].udp_param_value 354 #define us_xmit_lowat us_param_arr[8].udp_param_value 355 #define us_recv_hiwat us_param_arr[9].udp_param_value 356 #define us_max_buf us_param_arr[10].udp_param_value 357 #define us_ndd_get_info_interval us_param_arr[11].udp_param_value 358 359 360 #define UDP_STAT(us, x) ((us)->us_statistics.x.value.ui64++) 361 #define UDP_STAT_UPDATE(us, x, n) \ 362 ((us)->us_statistics.x.value.ui64 += (n)) 363 #ifdef DEBUG 364 #define UDP_DBGSTAT(us, x) UDP_STAT(us, x) 365 #else 366 #define UDP_DBGSTAT(us, x) 367 #endif /* DEBUG */ 368 369 extern int udp_opt_default(queue_t *, t_scalar_t, t_scalar_t, uchar_t *); 370 extern int udp_tpi_opt_get(queue_t *, t_scalar_t, t_scalar_t, uchar_t *); 371 extern int udp_tpi_opt_set(queue_t *, uint_t, int, int, uint_t, uchar_t *, 372 uint_t *, uchar_t *, void *, cred_t *, mblk_t *); 373 extern mblk_t *udp_snmp_get(queue_t *, mblk_t *); 374 extern int udp_snmp_set(queue_t *, t_scalar_t, t_scalar_t, uchar_t *, int); 375 extern void udp_close_free(conn_t *); 376 extern void udp_quiesce_conn(conn_t *); 377 extern void udp_ddi_g_init(void); 378 extern void udp_ddi_g_destroy(void); 379 extern void udp_g_q_inactive(udp_stack_t *); 380 extern void udp_output(conn_t *connp, mblk_t *mp, struct sockaddr *addr, 381 socklen_t addrlen); 382 extern void udp_wput(queue_t *, mblk_t *); 383 384 /* 385 * Object to represent database of options to search passed to 386 * {sock,tpi}optcom_req() interface routine to take care of option 387 * management and associated methods. 388 */ 389 extern optdb_obj_t udp_opt_obj; 390 extern uint_t udp_max_optsize; 391 392 extern sock_lower_handle_t udp_create(int, int, int, sock_downcalls_t **, 393 uint_t *, int *, int, cred_t *); 394 extern void udp_fallback(sock_lower_handle_t, queue_t *, boolean_t, 395 so_proto_quiesced_cb_t); 396 397 extern sock_downcalls_t sock_udp_downcalls; 398 399 #endif /* _KERNEL */ 400 401 #ifdef __cplusplus 402 } 403 #endif 404 405 #endif /* _UDP_IMPL_H */ 406