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; sockfs is on top */ 90 udpb_timestamp : 1, /* SO_TIMESTAMP "socket" option */ 91 92 udpb_nat_t_endpoint : 1, /* UDP_NAT_T_ENDPOINT option */ 93 udpb_pad_to_bit_31 : 4; 94 } udp_bits_t; 95 96 #define udp_debug udp_bits.udpb_debug 97 #define udp_dontroute udp_bits.udpb_dontroute 98 #define udp_broadcast udp_bits.udpb_broadcast 99 #define udp_useloopback udp_bits.udpb_useloopback 100 101 #define udp_reuseaddr udp_bits.udpb_reuseaddr 102 #define udp_dgram_errind udp_bits.udpb_dgram_errind 103 #define udp_recvdstaddr udp_bits.udpb_recvdstaddr 104 #define udp_recvopts udp_bits.udpb_recvopts 105 106 #define udp_unspec_source udp_bits.udpb_unspec_source 107 #define udp_ip_recvpktinfo udp_bits.udpb_ip_recvpktinfo 108 #define udp_ipv6_recvhoplimit udp_bits.udpb_ipv6_recvhoplimit 109 #define udp_ipv6_recvhopopts udp_bits.udpb_ipv6_recvhopopts 110 111 #define udp_ipv6_recvdstopts udp_bits.udpb_ipv6_recvdstopts 112 #define udp_ipv6_recvrthdr udp_bits.udpb_ipv6_recvrthdr 113 #define udp_ipv6_recvtclass udp_bits.udpb_ipv6_recvtclass 114 #define udp_ipv6_recvpathmtu udp_bits.udpb_ipv6_recvpathmtu 115 116 #define udp_anon_priv_bind udp_bits.udpb_anon_priv_bind 117 #define udp_exclbind udp_bits.udpb_exclbind 118 #define udp_recvif udp_bits.udpb_recvif 119 #define udp_recvslla udp_bits.udpb_recvslla 120 121 #define udp_recvttl udp_bits.udpb_recvttl 122 #define udp_recvucred udp_bits.udpb_recvucred 123 #define udp_old_ipv6_recvdstopts udp_bits.udpb_old_ipv6_recvdstopts 124 #define udp_ipv6_recvrthdrdstopts udp_bits.udpb_ipv6_recvrthdrdstopts 125 126 #define udp_rcvhdr udp_bits.udpb_rcvhdr 127 #define udp_issocket udp_bits.udpb_issocket 128 #define udp_timestamp udp_bits.udpb_timestamp 129 130 #define udp_nat_t_endpoint udp_bits.udpb_nat_t_endpoint 131 132 /* 133 * Bind hash list size and hash function. It has to be a power of 2 for 134 * hashing. 135 */ 136 #define UDP_BIND_FANOUT_SIZE 512 137 #define UDP_BIND_HASH(lport, size) \ 138 ((ntohs((uint16_t)lport)) & (size - 1)) 139 140 /* UDP bind fanout hash structure. */ 141 typedef struct udp_fanout_s { 142 struct udp_s *uf_udp; 143 kmutex_t uf_lock; 144 #if defined(_LP64) || defined(_I32LPx) 145 char uf_pad[48]; 146 #else 147 char uf_pad[56]; 148 #endif 149 } udp_fanout_t; 150 151 /* 152 * dev_q is the write side queue of the entity below IP. 153 * If there is a module below IP, we can't optimize by looking 154 * at q_first of the queue below IP. If the driver is directly 155 * below IP and if the q_first is NULL, we optimize by not doing 156 * the canput check 157 */ 158 #define DEV_Q_FLOW_BLOCKED(dev_q) \ 159 (((dev_q)->q_next != NULL || (dev_q)->q_first != NULL) && \ 160 !canput(dev_q)) 161 162 /* Kstats */ 163 typedef struct udp_stat { /* Class "net" kstats */ 164 kstat_named_t udp_ip_send; 165 kstat_named_t udp_ip_ire_send; 166 kstat_named_t udp_ire_null; 167 kstat_named_t udp_sock_fallback; 168 kstat_named_t udp_out_sw_cksum; 169 kstat_named_t udp_out_sw_cksum_bytes; 170 kstat_named_t udp_out_opt; 171 kstat_named_t udp_out_err_notconn; 172 kstat_named_t udp_out_err_output; 173 kstat_named_t udp_out_err_tudr; 174 kstat_named_t udp_in_pktinfo; 175 kstat_named_t udp_in_recvdstaddr; 176 kstat_named_t udp_in_recvopts; 177 kstat_named_t udp_in_recvif; 178 kstat_named_t udp_in_recvslla; 179 kstat_named_t udp_in_recvucred; 180 kstat_named_t udp_in_recvttl; 181 kstat_named_t udp_in_recvhopopts; 182 kstat_named_t udp_in_recvhoplimit; 183 kstat_named_t udp_in_recvdstopts; 184 kstat_named_t udp_in_recvrtdstopts; 185 kstat_named_t udp_in_recvrthdr; 186 kstat_named_t udp_in_recvpktinfo; 187 kstat_named_t udp_in_recvtclass; 188 kstat_named_t udp_in_timestamp; 189 kstat_named_t udp_ip_rcvpktinfo; 190 kstat_named_t udp_cookie_coll; 191 #ifdef DEBUG 192 kstat_named_t udp_data_conn; 193 kstat_named_t udp_data_notconn; 194 #endif 195 196 } udp_stat_t; 197 198 /* Named Dispatch Parameter Management Structure */ 199 typedef struct udpparam_s { 200 uint32_t udp_param_min; 201 uint32_t udp_param_max; 202 uint32_t udp_param_value; 203 char *udp_param_name; 204 } udpparam_t; 205 206 #define UDP_NUM_EPRIV_PORTS 64 207 208 /* 209 * UDP stack instances 210 */ 211 struct udp_stack { 212 netstack_t *us_netstack; /* Common netstack */ 213 214 uint_t us_bind_fanout_size; 215 udp_fanout_t *us_bind_fanout; 216 217 int us_num_epriv_ports; 218 in_port_t us_epriv_ports[UDP_NUM_EPRIV_PORTS]; 219 220 /* Hint not protected by any lock */ 221 in_port_t us_next_port_to_try; 222 223 IDP us_nd; /* Points to table of UDP ND variables. */ 224 udpparam_t *us_param_arr; /* ndd variable table */ 225 226 kstat_t *us_mibkp; /* kstats exporting mib data */ 227 kstat_t *us_kstat; 228 udp_stat_t us_statistics; 229 230 mib2_udp_t us_udp_mib; /* SNMP fixed size info */ 231 232 /* 233 * The smallest anonymous port in the priviledged port range which UDP 234 * looks for free port. Use in the option UDP_ANONPRIVBIND. 235 */ 236 in_port_t us_min_anonpriv_port; 237 238 ldi_ident_t us_ldi_ident; 239 }; 240 241 typedef struct udp_stack udp_stack_t; 242 243 /* Internal udp control structure, one per open stream */ 244 typedef struct udp_s { 245 krwlock_t udp_rwlock; /* Protects most of udp_t */ 246 t_scalar_t udp_pending_op; /* The current TPI operation */ 247 /* 248 * Following fields up to udp_ipversion protected by conn_lock, 249 * and the fanout lock i.e.uf_lock. Need both locks to change the 250 * field, either lock is sufficient for reading the field. 251 */ 252 uint32_t udp_state; /* TPI state */ 253 in_port_t udp_port; /* Port bound to this stream */ 254 in_port_t udp_dstport; /* Connected port */ 255 in6_addr_t udp_v6src; /* Source address of this stream */ 256 in6_addr_t udp_bound_v6src; /* Explicitly bound address */ 257 in6_addr_t udp_v6dst; /* Connected destination */ 258 /* 259 * IP format that packets transmitted from this struct should use. 260 * Value can be IP4_VERSION or IPV6_VERSION. 261 */ 262 ushort_t udp_ipversion; 263 264 /* Written to only once at the time of opening the endpoint */ 265 sa_family_t udp_family; /* Family from socket() call */ 266 267 /* Following protected by udp_rwlock */ 268 uint32_t udp_flowinfo; /* Connected flow id and tclass */ 269 uint32_t udp_max_hdr_len; /* For write offset in stream head */ 270 uint32_t udp_ip_snd_options_len; /* Len of IPv4 options */ 271 uchar_t *udp_ip_snd_options; /* Ptr to IPv4 options */ 272 uint32_t udp_ip_rcv_options_len; /* Len of IPv4 options recvd */ 273 uchar_t *udp_ip_rcv_options; /* Ptr to IPv4 options recvd */ 274 uchar_t udp_multicast_ttl; /* IP*_MULTICAST_TTL/HOPS */ 275 ipaddr_t udp_multicast_if_addr; /* IP_MULTICAST_IF option */ 276 uint_t udp_multicast_if_index; /* IPV6_MULTICAST_IF option */ 277 int udp_bound_if; /* IP*_BOUND_IF option */ 278 279 /* Written to only once at the time of opening the endpoint */ 280 conn_t *udp_connp; 281 282 /* Following protected by udp_rwlock */ 283 udp_bits_t udp_bits; /* Bit fields defined above */ 284 uint8_t udp_type_of_service; /* IP_TOS option */ 285 uint8_t udp_ttl; /* TTL or hoplimit */ 286 ip6_pkt_t udp_sticky_ipp; /* Sticky options */ 287 uint8_t *udp_sticky_hdrs; /* Prebuilt IPv6 hdrs */ 288 uint_t udp_sticky_hdrs_len; /* Incl. ip6h and any ip6i */ 289 290 /* Following 2 fields protected by the uf_lock */ 291 struct udp_s *udp_bind_hash; /* Bind hash chain */ 292 struct udp_s **udp_ptpbhn; /* Pointer to previous bind hash next. */ 293 294 /* Following protected by udp_rwlock */ 295 kmutex_t udp_recv_lock; /* recv lock */ 296 size_t udp_rcv_disply_hiwat; /* user's view of rcvbuf */ 297 size_t udp_rcv_hiwat; /* receive high watermark */ 298 size_t udp_rcv_lowat; /* receive low watermark */ 299 size_t udp_xmit_hiwat; /* Send buffer high watermark */ 300 size_t udp_xmit_lowat; /* Send buffer low watermark */ 301 uint_t udp_label_len; /* length of security label */ 302 uint_t udp_label_len_v6; /* len of v6 security label */ 303 in6_addr_t udp_v6lastdst; /* most recent destination */ 304 in_port_t udp_lastdstport; /* most recent dest port */ 305 cred_t *udp_last_cred; /* most recent credentials */ 306 cred_t *udp_effective_cred; /* cred with effective label */ 307 308 uint64_t udp_open_time; /* time when this was opened */ 309 pid_t udp_open_pid; /* process id when this was opened */ 310 udp_stack_t *udp_us; /* Stack instance for zone */ 311 int udp_delayed_error; 312 mblk_t *udp_fallback_queue_head; 313 mblk_t *udp_fallback_queue_tail; 314 struct sockaddr_storage udp_delayed_addr; 315 } udp_t; 316 317 /* UDP Protocol header */ 318 /* UDP Protocol header aligned */ 319 typedef struct udpahdr_s { 320 in_port_t uha_src_port; /* Source port */ 321 in_port_t uha_dst_port; /* Destination port */ 322 uint16_t uha_length; /* UDP length */ 323 uint16_t uha_checksum; /* UDP checksum */ 324 } udpha_t; 325 326 #define us_wroff_extra us_param_arr[0].udp_param_value 327 #define us_ipv4_ttl us_param_arr[1].udp_param_value 328 #define us_ipv6_hoplimit us_param_arr[2].udp_param_value 329 #define us_smallest_nonpriv_port us_param_arr[3].udp_param_value 330 #define us_do_checksum us_param_arr[4].udp_param_value 331 #define us_smallest_anon_port us_param_arr[5].udp_param_value 332 #define us_largest_anon_port us_param_arr[6].udp_param_value 333 #define us_xmit_hiwat us_param_arr[7].udp_param_value 334 #define us_xmit_lowat us_param_arr[8].udp_param_value 335 #define us_recv_hiwat us_param_arr[9].udp_param_value 336 #define us_max_buf us_param_arr[10].udp_param_value 337 338 339 #define UDP_STAT(us, x) ((us)->us_statistics.x.value.ui64++) 340 #define UDP_STAT_UPDATE(us, x, n) \ 341 ((us)->us_statistics.x.value.ui64 += (n)) 342 #ifdef DEBUG 343 #define UDP_DBGSTAT(us, x) UDP_STAT(us, x) 344 #else 345 #define UDP_DBGSTAT(us, x) 346 #endif /* DEBUG */ 347 348 extern int udp_opt_default(queue_t *, t_scalar_t, t_scalar_t, uchar_t *); 349 extern int udp_tpi_opt_get(queue_t *, t_scalar_t, t_scalar_t, uchar_t *); 350 extern int udp_tpi_opt_set(queue_t *, uint_t, int, int, uint_t, uchar_t *, 351 uint_t *, uchar_t *, void *, cred_t *, mblk_t *); 352 extern mblk_t *udp_snmp_get(queue_t *, mblk_t *); 353 extern int udp_snmp_set(queue_t *, t_scalar_t, t_scalar_t, uchar_t *, int); 354 extern void udp_close_free(conn_t *); 355 extern void udp_quiesce_conn(conn_t *); 356 extern void udp_ddi_g_init(void); 357 extern void udp_ddi_g_destroy(void); 358 extern void udp_g_q_inactive(udp_stack_t *); 359 extern void udp_output(conn_t *connp, mblk_t *mp, struct sockaddr *addr, 360 socklen_t addrlen); 361 extern void udp_wput(queue_t *, mblk_t *); 362 363 /* 364 * Object to represent database of options to search passed to 365 * {sock,tpi}optcom_req() interface routine to take care of option 366 * management and associated methods. 367 */ 368 extern optdb_obj_t udp_opt_obj; 369 extern uint_t udp_max_optsize; 370 371 extern sock_lower_handle_t udp_create(int, int, int, sock_downcalls_t **, 372 uint_t *, int *, int, cred_t *); 373 extern int udp_fallback(sock_lower_handle_t, queue_t *, boolean_t, 374 so_proto_quiesced_cb_t); 375 376 extern sock_downcalls_t sock_udp_downcalls; 377 378 #endif /* _KERNEL */ 379 380 #ifdef __cplusplus 381 } 382 #endif 383 384 #endif /* _UDP_IMPL_H */ 385