1 /* 2 * net/tipc/socket.c: TIPC socket API 3 * 4 * Copyright (c) 2001-2007, 2012-2017, Ericsson AB 5 * Copyright (c) 2004-2008, 2010-2013, Wind River Systems 6 * All rights reserved. 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions are met: 10 * 11 * 1. Redistributions of source code must retain the above copyright 12 * notice, this list of conditions and the following disclaimer. 13 * 2. Redistributions in binary form must reproduce the above copyright 14 * notice, this list of conditions and the following disclaimer in the 15 * documentation and/or other materials provided with the distribution. 16 * 3. Neither the names of the copyright holders nor the names of its 17 * contributors may be used to endorse or promote products derived from 18 * this software without specific prior written permission. 19 * 20 * Alternatively, this software may be distributed under the terms of the 21 * GNU General Public License ("GPL") version 2 as published by the Free 22 * Software Foundation. 23 * 24 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" 25 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 26 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 27 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE 28 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR 29 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF 30 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 31 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 32 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) 33 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE 34 * POSSIBILITY OF SUCH DAMAGE. 35 */ 36 37 #include <linux/rhashtable.h> 38 #include <linux/sched/signal.h> 39 40 #include "core.h" 41 #include "name_table.h" 42 #include "node.h" 43 #include "link.h" 44 #include "name_distr.h" 45 #include "socket.h" 46 #include "bcast.h" 47 #include "netlink.h" 48 #include "group.h" 49 #include "trace.h" 50 51 #define CONN_TIMEOUT_DEFAULT 8000 /* default connect timeout = 8s */ 52 #define CONN_PROBING_INTV msecs_to_jiffies(3600000) /* [ms] => 1 h */ 53 #define TIPC_FWD_MSG 1 54 #define TIPC_MAX_PORT 0xffffffff 55 #define TIPC_MIN_PORT 1 56 #define TIPC_ACK_RATE 4 /* ACK at 1/4 of of rcv window size */ 57 58 enum { 59 TIPC_LISTEN = TCP_LISTEN, 60 TIPC_ESTABLISHED = TCP_ESTABLISHED, 61 TIPC_OPEN = TCP_CLOSE, 62 TIPC_DISCONNECTING = TCP_CLOSE_WAIT, 63 TIPC_CONNECTING = TCP_SYN_SENT, 64 }; 65 66 struct sockaddr_pair { 67 struct sockaddr_tipc sock; 68 struct sockaddr_tipc member; 69 }; 70 71 /** 72 * struct tipc_sock - TIPC socket structure 73 * @sk: socket - interacts with 'port' and with user via the socket API 74 * @conn_type: TIPC type used when connection was established 75 * @conn_instance: TIPC instance used when connection was established 76 * @published: non-zero if port has one or more associated names 77 * @max_pkt: maximum packet size "hint" used when building messages sent by port 78 * @maxnagle: maximum size of msg which can be subject to nagle 79 * @portid: unique port identity in TIPC socket hash table 80 * @phdr: preformatted message header used when sending messages 81 * #cong_links: list of congested links 82 * @publications: list of publications for port 83 * @blocking_link: address of the congested link we are currently sleeping on 84 * @pub_count: total # of publications port has made during its lifetime 85 * @conn_timeout: the time we can wait for an unresponded setup request 86 * @dupl_rcvcnt: number of bytes counted twice, in both backlog and rcv queue 87 * @cong_link_cnt: number of congested links 88 * @snt_unacked: # messages sent by socket, and not yet acked by peer 89 * @rcv_unacked: # messages read by user, but not yet acked back to peer 90 * @peer: 'connected' peer for dgram/rdm 91 * @node: hash table node 92 * @mc_method: cookie for use between socket and broadcast layer 93 * @rcu: rcu struct for tipc_sock 94 */ 95 struct tipc_sock { 96 struct sock sk; 97 u32 conn_type; 98 u32 conn_instance; 99 int published; 100 u32 max_pkt; 101 u32 maxnagle; 102 u32 portid; 103 struct tipc_msg phdr; 104 struct list_head cong_links; 105 struct list_head publications; 106 u32 pub_count; 107 atomic_t dupl_rcvcnt; 108 u16 conn_timeout; 109 bool probe_unacked; 110 u16 cong_link_cnt; 111 u16 snt_unacked; 112 u16 snd_win; 113 u16 peer_caps; 114 u16 rcv_unacked; 115 u16 rcv_win; 116 struct sockaddr_tipc peer; 117 struct rhash_head node; 118 struct tipc_mc_method mc_method; 119 struct rcu_head rcu; 120 struct tipc_group *group; 121 u32 oneway; 122 u16 snd_backlog; 123 bool expect_ack; 124 bool nodelay; 125 bool group_is_open; 126 }; 127 128 static int tipc_sk_backlog_rcv(struct sock *sk, struct sk_buff *skb); 129 static void tipc_data_ready(struct sock *sk); 130 static void tipc_write_space(struct sock *sk); 131 static void tipc_sock_destruct(struct sock *sk); 132 static int tipc_release(struct socket *sock); 133 static int tipc_accept(struct socket *sock, struct socket *new_sock, int flags, 134 bool kern); 135 static void tipc_sk_timeout(struct timer_list *t); 136 static int tipc_sk_publish(struct tipc_sock *tsk, uint scope, 137 struct tipc_name_seq const *seq); 138 static int tipc_sk_withdraw(struct tipc_sock *tsk, uint scope, 139 struct tipc_name_seq const *seq); 140 static int tipc_sk_leave(struct tipc_sock *tsk); 141 static struct tipc_sock *tipc_sk_lookup(struct net *net, u32 portid); 142 static int tipc_sk_insert(struct tipc_sock *tsk); 143 static void tipc_sk_remove(struct tipc_sock *tsk); 144 static int __tipc_sendstream(struct socket *sock, struct msghdr *m, size_t dsz); 145 static int __tipc_sendmsg(struct socket *sock, struct msghdr *m, size_t dsz); 146 static void tipc_sk_push_backlog(struct tipc_sock *tsk); 147 148 static const struct proto_ops packet_ops; 149 static const struct proto_ops stream_ops; 150 static const struct proto_ops msg_ops; 151 static struct proto tipc_proto; 152 static const struct rhashtable_params tsk_rht_params; 153 154 static u32 tsk_own_node(struct tipc_sock *tsk) 155 { 156 return msg_prevnode(&tsk->phdr); 157 } 158 159 static u32 tsk_peer_node(struct tipc_sock *tsk) 160 { 161 return msg_destnode(&tsk->phdr); 162 } 163 164 static u32 tsk_peer_port(struct tipc_sock *tsk) 165 { 166 return msg_destport(&tsk->phdr); 167 } 168 169 static bool tsk_unreliable(struct tipc_sock *tsk) 170 { 171 return msg_src_droppable(&tsk->phdr) != 0; 172 } 173 174 static void tsk_set_unreliable(struct tipc_sock *tsk, bool unreliable) 175 { 176 msg_set_src_droppable(&tsk->phdr, unreliable ? 1 : 0); 177 } 178 179 static bool tsk_unreturnable(struct tipc_sock *tsk) 180 { 181 return msg_dest_droppable(&tsk->phdr) != 0; 182 } 183 184 static void tsk_set_unreturnable(struct tipc_sock *tsk, bool unreturnable) 185 { 186 msg_set_dest_droppable(&tsk->phdr, unreturnable ? 1 : 0); 187 } 188 189 static int tsk_importance(struct tipc_sock *tsk) 190 { 191 return msg_importance(&tsk->phdr); 192 } 193 194 static int tsk_set_importance(struct tipc_sock *tsk, int imp) 195 { 196 if (imp > TIPC_CRITICAL_IMPORTANCE) 197 return -EINVAL; 198 msg_set_importance(&tsk->phdr, (u32)imp); 199 return 0; 200 } 201 202 static struct tipc_sock *tipc_sk(const struct sock *sk) 203 { 204 return container_of(sk, struct tipc_sock, sk); 205 } 206 207 static bool tsk_conn_cong(struct tipc_sock *tsk) 208 { 209 return tsk->snt_unacked > tsk->snd_win; 210 } 211 212 static u16 tsk_blocks(int len) 213 { 214 return ((len / FLOWCTL_BLK_SZ) + 1); 215 } 216 217 /* tsk_blocks(): translate a buffer size in bytes to number of 218 * advertisable blocks, taking into account the ratio truesize(len)/len 219 * We can trust that this ratio is always < 4 for len >= FLOWCTL_BLK_SZ 220 */ 221 static u16 tsk_adv_blocks(int len) 222 { 223 return len / FLOWCTL_BLK_SZ / 4; 224 } 225 226 /* tsk_inc(): increment counter for sent or received data 227 * - If block based flow control is not supported by peer we 228 * fall back to message based ditto, incrementing the counter 229 */ 230 static u16 tsk_inc(struct tipc_sock *tsk, int msglen) 231 { 232 if (likely(tsk->peer_caps & TIPC_BLOCK_FLOWCTL)) 233 return ((msglen / FLOWCTL_BLK_SZ) + 1); 234 return 1; 235 } 236 237 /* tsk_set_nagle - enable/disable nagle property by manipulating maxnagle 238 */ 239 static void tsk_set_nagle(struct tipc_sock *tsk) 240 { 241 struct sock *sk = &tsk->sk; 242 243 tsk->maxnagle = 0; 244 if (sk->sk_type != SOCK_STREAM) 245 return; 246 if (tsk->nodelay) 247 return; 248 if (!(tsk->peer_caps & TIPC_NAGLE)) 249 return; 250 /* Limit node local buffer size to avoid receive queue overflow */ 251 if (tsk->max_pkt == MAX_MSG_SIZE) 252 tsk->maxnagle = 1500; 253 else 254 tsk->maxnagle = tsk->max_pkt; 255 } 256 257 /** 258 * tsk_advance_rx_queue - discard first buffer in socket receive queue 259 * 260 * Caller must hold socket lock 261 */ 262 static void tsk_advance_rx_queue(struct sock *sk) 263 { 264 trace_tipc_sk_advance_rx(sk, NULL, TIPC_DUMP_SK_RCVQ, " "); 265 kfree_skb(__skb_dequeue(&sk->sk_receive_queue)); 266 } 267 268 /* tipc_sk_respond() : send response message back to sender 269 */ 270 static void tipc_sk_respond(struct sock *sk, struct sk_buff *skb, int err) 271 { 272 u32 selector; 273 u32 dnode; 274 u32 onode = tipc_own_addr(sock_net(sk)); 275 276 if (!tipc_msg_reverse(onode, &skb, err)) 277 return; 278 279 trace_tipc_sk_rej_msg(sk, skb, TIPC_DUMP_NONE, "@sk_respond!"); 280 dnode = msg_destnode(buf_msg(skb)); 281 selector = msg_origport(buf_msg(skb)); 282 tipc_node_xmit_skb(sock_net(sk), skb, dnode, selector); 283 } 284 285 /** 286 * tsk_rej_rx_queue - reject all buffers in socket receive queue 287 * 288 * Caller must hold socket lock 289 */ 290 static void tsk_rej_rx_queue(struct sock *sk) 291 { 292 struct sk_buff *skb; 293 294 while ((skb = __skb_dequeue(&sk->sk_receive_queue))) 295 tipc_sk_respond(sk, skb, TIPC_ERR_NO_PORT); 296 } 297 298 static bool tipc_sk_connected(struct sock *sk) 299 { 300 return sk->sk_state == TIPC_ESTABLISHED; 301 } 302 303 /* tipc_sk_type_connectionless - check if the socket is datagram socket 304 * @sk: socket 305 * 306 * Returns true if connection less, false otherwise 307 */ 308 static bool tipc_sk_type_connectionless(struct sock *sk) 309 { 310 return sk->sk_type == SOCK_RDM || sk->sk_type == SOCK_DGRAM; 311 } 312 313 /* tsk_peer_msg - verify if message was sent by connected port's peer 314 * 315 * Handles cases where the node's network address has changed from 316 * the default of <0.0.0> to its configured setting. 317 */ 318 static bool tsk_peer_msg(struct tipc_sock *tsk, struct tipc_msg *msg) 319 { 320 struct sock *sk = &tsk->sk; 321 u32 self = tipc_own_addr(sock_net(sk)); 322 u32 peer_port = tsk_peer_port(tsk); 323 u32 orig_node, peer_node; 324 325 if (unlikely(!tipc_sk_connected(sk))) 326 return false; 327 328 if (unlikely(msg_origport(msg) != peer_port)) 329 return false; 330 331 orig_node = msg_orignode(msg); 332 peer_node = tsk_peer_node(tsk); 333 334 if (likely(orig_node == peer_node)) 335 return true; 336 337 if (!orig_node && peer_node == self) 338 return true; 339 340 if (!peer_node && orig_node == self) 341 return true; 342 343 return false; 344 } 345 346 /* tipc_set_sk_state - set the sk_state of the socket 347 * @sk: socket 348 * 349 * Caller must hold socket lock 350 * 351 * Returns 0 on success, errno otherwise 352 */ 353 static int tipc_set_sk_state(struct sock *sk, int state) 354 { 355 int oldsk_state = sk->sk_state; 356 int res = -EINVAL; 357 358 switch (state) { 359 case TIPC_OPEN: 360 res = 0; 361 break; 362 case TIPC_LISTEN: 363 case TIPC_CONNECTING: 364 if (oldsk_state == TIPC_OPEN) 365 res = 0; 366 break; 367 case TIPC_ESTABLISHED: 368 if (oldsk_state == TIPC_CONNECTING || 369 oldsk_state == TIPC_OPEN) 370 res = 0; 371 break; 372 case TIPC_DISCONNECTING: 373 if (oldsk_state == TIPC_CONNECTING || 374 oldsk_state == TIPC_ESTABLISHED) 375 res = 0; 376 break; 377 } 378 379 if (!res) 380 sk->sk_state = state; 381 382 return res; 383 } 384 385 static int tipc_sk_sock_err(struct socket *sock, long *timeout) 386 { 387 struct sock *sk = sock->sk; 388 int err = sock_error(sk); 389 int typ = sock->type; 390 391 if (err) 392 return err; 393 if (typ == SOCK_STREAM || typ == SOCK_SEQPACKET) { 394 if (sk->sk_state == TIPC_DISCONNECTING) 395 return -EPIPE; 396 else if (!tipc_sk_connected(sk)) 397 return -ENOTCONN; 398 } 399 if (!*timeout) 400 return -EAGAIN; 401 if (signal_pending(current)) 402 return sock_intr_errno(*timeout); 403 404 return 0; 405 } 406 407 #define tipc_wait_for_cond(sock_, timeo_, condition_) \ 408 ({ \ 409 DEFINE_WAIT_FUNC(wait_, woken_wake_function); \ 410 struct sock *sk_; \ 411 int rc_; \ 412 \ 413 while ((rc_ = !(condition_))) { \ 414 /* coupled with smp_wmb() in tipc_sk_proto_rcv() */ \ 415 smp_rmb(); \ 416 sk_ = (sock_)->sk; \ 417 rc_ = tipc_sk_sock_err((sock_), timeo_); \ 418 if (rc_) \ 419 break; \ 420 add_wait_queue(sk_sleep(sk_), &wait_); \ 421 release_sock(sk_); \ 422 *(timeo_) = wait_woken(&wait_, TASK_INTERRUPTIBLE, *(timeo_)); \ 423 sched_annotate_sleep(); \ 424 lock_sock(sk_); \ 425 remove_wait_queue(sk_sleep(sk_), &wait_); \ 426 } \ 427 rc_; \ 428 }) 429 430 /** 431 * tipc_sk_create - create a TIPC socket 432 * @net: network namespace (must be default network) 433 * @sock: pre-allocated socket structure 434 * @protocol: protocol indicator (must be 0) 435 * @kern: caused by kernel or by userspace? 436 * 437 * This routine creates additional data structures used by the TIPC socket, 438 * initializes them, and links them together. 439 * 440 * Returns 0 on success, errno otherwise 441 */ 442 static int tipc_sk_create(struct net *net, struct socket *sock, 443 int protocol, int kern) 444 { 445 const struct proto_ops *ops; 446 struct sock *sk; 447 struct tipc_sock *tsk; 448 struct tipc_msg *msg; 449 450 /* Validate arguments */ 451 if (unlikely(protocol != 0)) 452 return -EPROTONOSUPPORT; 453 454 switch (sock->type) { 455 case SOCK_STREAM: 456 ops = &stream_ops; 457 break; 458 case SOCK_SEQPACKET: 459 ops = &packet_ops; 460 break; 461 case SOCK_DGRAM: 462 case SOCK_RDM: 463 ops = &msg_ops; 464 break; 465 default: 466 return -EPROTOTYPE; 467 } 468 469 /* Allocate socket's protocol area */ 470 sk = sk_alloc(net, AF_TIPC, GFP_KERNEL, &tipc_proto, kern); 471 if (sk == NULL) 472 return -ENOMEM; 473 474 tsk = tipc_sk(sk); 475 tsk->max_pkt = MAX_PKT_DEFAULT; 476 tsk->maxnagle = 0; 477 INIT_LIST_HEAD(&tsk->publications); 478 INIT_LIST_HEAD(&tsk->cong_links); 479 msg = &tsk->phdr; 480 481 /* Finish initializing socket data structures */ 482 sock->ops = ops; 483 sock_init_data(sock, sk); 484 tipc_set_sk_state(sk, TIPC_OPEN); 485 if (tipc_sk_insert(tsk)) { 486 pr_warn("Socket create failed; port number exhausted\n"); 487 return -EINVAL; 488 } 489 490 /* Ensure tsk is visible before we read own_addr. */ 491 smp_mb(); 492 493 tipc_msg_init(tipc_own_addr(net), msg, TIPC_LOW_IMPORTANCE, 494 TIPC_NAMED_MSG, NAMED_H_SIZE, 0); 495 496 msg_set_origport(msg, tsk->portid); 497 timer_setup(&sk->sk_timer, tipc_sk_timeout, 0); 498 sk->sk_shutdown = 0; 499 sk->sk_backlog_rcv = tipc_sk_backlog_rcv; 500 sk->sk_rcvbuf = sysctl_tipc_rmem[1]; 501 sk->sk_data_ready = tipc_data_ready; 502 sk->sk_write_space = tipc_write_space; 503 sk->sk_destruct = tipc_sock_destruct; 504 tsk->conn_timeout = CONN_TIMEOUT_DEFAULT; 505 tsk->group_is_open = true; 506 atomic_set(&tsk->dupl_rcvcnt, 0); 507 508 /* Start out with safe limits until we receive an advertised window */ 509 tsk->snd_win = tsk_adv_blocks(RCVBUF_MIN); 510 tsk->rcv_win = tsk->snd_win; 511 512 if (tipc_sk_type_connectionless(sk)) { 513 tsk_set_unreturnable(tsk, true); 514 if (sock->type == SOCK_DGRAM) 515 tsk_set_unreliable(tsk, true); 516 } 517 __skb_queue_head_init(&tsk->mc_method.deferredq); 518 trace_tipc_sk_create(sk, NULL, TIPC_DUMP_NONE, " "); 519 return 0; 520 } 521 522 static void tipc_sk_callback(struct rcu_head *head) 523 { 524 struct tipc_sock *tsk = container_of(head, struct tipc_sock, rcu); 525 526 sock_put(&tsk->sk); 527 } 528 529 /* Caller should hold socket lock for the socket. */ 530 static void __tipc_shutdown(struct socket *sock, int error) 531 { 532 struct sock *sk = sock->sk; 533 struct tipc_sock *tsk = tipc_sk(sk); 534 struct net *net = sock_net(sk); 535 long timeout = msecs_to_jiffies(CONN_TIMEOUT_DEFAULT); 536 u32 dnode = tsk_peer_node(tsk); 537 struct sk_buff *skb; 538 539 /* Avoid that hi-prio shutdown msgs bypass msgs in link wakeup queue */ 540 tipc_wait_for_cond(sock, &timeout, (!tsk->cong_link_cnt && 541 !tsk_conn_cong(tsk))); 542 543 /* Push out delayed messages if in Nagle mode */ 544 tipc_sk_push_backlog(tsk); 545 /* Remove pending SYN */ 546 __skb_queue_purge(&sk->sk_write_queue); 547 548 /* Reject all unreceived messages, except on an active connection 549 * (which disconnects locally & sends a 'FIN+' to peer). 550 */ 551 while ((skb = __skb_dequeue(&sk->sk_receive_queue)) != NULL) { 552 if (TIPC_SKB_CB(skb)->bytes_read) { 553 kfree_skb(skb); 554 continue; 555 } 556 if (!tipc_sk_type_connectionless(sk) && 557 sk->sk_state != TIPC_DISCONNECTING) { 558 tipc_set_sk_state(sk, TIPC_DISCONNECTING); 559 tipc_node_remove_conn(net, dnode, tsk->portid); 560 } 561 tipc_sk_respond(sk, skb, error); 562 } 563 564 if (tipc_sk_type_connectionless(sk)) 565 return; 566 567 if (sk->sk_state != TIPC_DISCONNECTING) { 568 skb = tipc_msg_create(TIPC_CRITICAL_IMPORTANCE, 569 TIPC_CONN_MSG, SHORT_H_SIZE, 0, dnode, 570 tsk_own_node(tsk), tsk_peer_port(tsk), 571 tsk->portid, error); 572 if (skb) 573 tipc_node_xmit_skb(net, skb, dnode, tsk->portid); 574 tipc_node_remove_conn(net, dnode, tsk->portid); 575 tipc_set_sk_state(sk, TIPC_DISCONNECTING); 576 } 577 } 578 579 /** 580 * tipc_release - destroy a TIPC socket 581 * @sock: socket to destroy 582 * 583 * This routine cleans up any messages that are still queued on the socket. 584 * For DGRAM and RDM socket types, all queued messages are rejected. 585 * For SEQPACKET and STREAM socket types, the first message is rejected 586 * and any others are discarded. (If the first message on a STREAM socket 587 * is partially-read, it is discarded and the next one is rejected instead.) 588 * 589 * NOTE: Rejected messages are not necessarily returned to the sender! They 590 * are returned or discarded according to the "destination droppable" setting 591 * specified for the message by the sender. 592 * 593 * Returns 0 on success, errno otherwise 594 */ 595 static int tipc_release(struct socket *sock) 596 { 597 struct sock *sk = sock->sk; 598 struct tipc_sock *tsk; 599 600 /* 601 * Exit if socket isn't fully initialized (occurs when a failed accept() 602 * releases a pre-allocated child socket that was never used) 603 */ 604 if (sk == NULL) 605 return 0; 606 607 tsk = tipc_sk(sk); 608 lock_sock(sk); 609 610 trace_tipc_sk_release(sk, NULL, TIPC_DUMP_ALL, " "); 611 __tipc_shutdown(sock, TIPC_ERR_NO_PORT); 612 sk->sk_shutdown = SHUTDOWN_MASK; 613 tipc_sk_leave(tsk); 614 tipc_sk_withdraw(tsk, 0, NULL); 615 __skb_queue_purge(&tsk->mc_method.deferredq); 616 sk_stop_timer(sk, &sk->sk_timer); 617 tipc_sk_remove(tsk); 618 619 sock_orphan(sk); 620 /* Reject any messages that accumulated in backlog queue */ 621 release_sock(sk); 622 tipc_dest_list_purge(&tsk->cong_links); 623 tsk->cong_link_cnt = 0; 624 call_rcu(&tsk->rcu, tipc_sk_callback); 625 sock->sk = NULL; 626 627 return 0; 628 } 629 630 /** 631 * tipc_bind - associate or disassocate TIPC name(s) with a socket 632 * @sock: socket structure 633 * @uaddr: socket address describing name(s) and desired operation 634 * @uaddr_len: size of socket address data structure 635 * 636 * Name and name sequence binding is indicated using a positive scope value; 637 * a negative scope value unbinds the specified name. Specifying no name 638 * (i.e. a socket address length of 0) unbinds all names from the socket. 639 * 640 * Returns 0 on success, errno otherwise 641 * 642 * NOTE: This routine doesn't need to take the socket lock since it doesn't 643 * access any non-constant socket information. 644 */ 645 static int tipc_bind(struct socket *sock, struct sockaddr *uaddr, 646 int uaddr_len) 647 { 648 struct sock *sk = sock->sk; 649 struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr; 650 struct tipc_sock *tsk = tipc_sk(sk); 651 int res = -EINVAL; 652 653 lock_sock(sk); 654 if (unlikely(!uaddr_len)) { 655 res = tipc_sk_withdraw(tsk, 0, NULL); 656 goto exit; 657 } 658 if (tsk->group) { 659 res = -EACCES; 660 goto exit; 661 } 662 if (uaddr_len < sizeof(struct sockaddr_tipc)) { 663 res = -EINVAL; 664 goto exit; 665 } 666 if (addr->family != AF_TIPC) { 667 res = -EAFNOSUPPORT; 668 goto exit; 669 } 670 671 if (addr->addrtype == TIPC_ADDR_NAME) 672 addr->addr.nameseq.upper = addr->addr.nameseq.lower; 673 else if (addr->addrtype != TIPC_ADDR_NAMESEQ) { 674 res = -EAFNOSUPPORT; 675 goto exit; 676 } 677 678 if ((addr->addr.nameseq.type < TIPC_RESERVED_TYPES) && 679 (addr->addr.nameseq.type != TIPC_TOP_SRV) && 680 (addr->addr.nameseq.type != TIPC_CFG_SRV)) { 681 res = -EACCES; 682 goto exit; 683 } 684 685 res = (addr->scope >= 0) ? 686 tipc_sk_publish(tsk, addr->scope, &addr->addr.nameseq) : 687 tipc_sk_withdraw(tsk, -addr->scope, &addr->addr.nameseq); 688 exit: 689 release_sock(sk); 690 return res; 691 } 692 693 /** 694 * tipc_getname - get port ID of socket or peer socket 695 * @sock: socket structure 696 * @uaddr: area for returned socket address 697 * @uaddr_len: area for returned length of socket address 698 * @peer: 0 = own ID, 1 = current peer ID, 2 = current/former peer ID 699 * 700 * Returns 0 on success, errno otherwise 701 * 702 * NOTE: This routine doesn't need to take the socket lock since it only 703 * accesses socket information that is unchanging (or which changes in 704 * a completely predictable manner). 705 */ 706 static int tipc_getname(struct socket *sock, struct sockaddr *uaddr, 707 int peer) 708 { 709 struct sockaddr_tipc *addr = (struct sockaddr_tipc *)uaddr; 710 struct sock *sk = sock->sk; 711 struct tipc_sock *tsk = tipc_sk(sk); 712 713 memset(addr, 0, sizeof(*addr)); 714 if (peer) { 715 if ((!tipc_sk_connected(sk)) && 716 ((peer != 2) || (sk->sk_state != TIPC_DISCONNECTING))) 717 return -ENOTCONN; 718 addr->addr.id.ref = tsk_peer_port(tsk); 719 addr->addr.id.node = tsk_peer_node(tsk); 720 } else { 721 addr->addr.id.ref = tsk->portid; 722 addr->addr.id.node = tipc_own_addr(sock_net(sk)); 723 } 724 725 addr->addrtype = TIPC_ADDR_ID; 726 addr->family = AF_TIPC; 727 addr->scope = 0; 728 addr->addr.name.domain = 0; 729 730 return sizeof(*addr); 731 } 732 733 /** 734 * tipc_poll - read and possibly block on pollmask 735 * @file: file structure associated with the socket 736 * @sock: socket for which to calculate the poll bits 737 * @wait: ??? 738 * 739 * Returns pollmask value 740 * 741 * COMMENTARY: 742 * It appears that the usual socket locking mechanisms are not useful here 743 * since the pollmask info is potentially out-of-date the moment this routine 744 * exits. TCP and other protocols seem to rely on higher level poll routines 745 * to handle any preventable race conditions, so TIPC will do the same ... 746 * 747 * IMPORTANT: The fact that a read or write operation is indicated does NOT 748 * imply that the operation will succeed, merely that it should be performed 749 * and will not block. 750 */ 751 static __poll_t tipc_poll(struct file *file, struct socket *sock, 752 poll_table *wait) 753 { 754 struct sock *sk = sock->sk; 755 struct tipc_sock *tsk = tipc_sk(sk); 756 __poll_t revents = 0; 757 758 sock_poll_wait(file, sock, wait); 759 trace_tipc_sk_poll(sk, NULL, TIPC_DUMP_ALL, " "); 760 761 if (sk->sk_shutdown & RCV_SHUTDOWN) 762 revents |= EPOLLRDHUP | EPOLLIN | EPOLLRDNORM; 763 if (sk->sk_shutdown == SHUTDOWN_MASK) 764 revents |= EPOLLHUP; 765 766 switch (sk->sk_state) { 767 case TIPC_ESTABLISHED: 768 if (!tsk->cong_link_cnt && !tsk_conn_cong(tsk)) 769 revents |= EPOLLOUT; 770 /* fall through */ 771 case TIPC_LISTEN: 772 case TIPC_CONNECTING: 773 if (!skb_queue_empty_lockless(&sk->sk_receive_queue)) 774 revents |= EPOLLIN | EPOLLRDNORM; 775 break; 776 case TIPC_OPEN: 777 if (tsk->group_is_open && !tsk->cong_link_cnt) 778 revents |= EPOLLOUT; 779 if (!tipc_sk_type_connectionless(sk)) 780 break; 781 if (skb_queue_empty_lockless(&sk->sk_receive_queue)) 782 break; 783 revents |= EPOLLIN | EPOLLRDNORM; 784 break; 785 case TIPC_DISCONNECTING: 786 revents = EPOLLIN | EPOLLRDNORM | EPOLLHUP; 787 break; 788 } 789 return revents; 790 } 791 792 /** 793 * tipc_sendmcast - send multicast message 794 * @sock: socket structure 795 * @seq: destination address 796 * @msg: message to send 797 * @dlen: length of data to send 798 * @timeout: timeout to wait for wakeup 799 * 800 * Called from function tipc_sendmsg(), which has done all sanity checks 801 * Returns the number of bytes sent on success, or errno 802 */ 803 static int tipc_sendmcast(struct socket *sock, struct tipc_name_seq *seq, 804 struct msghdr *msg, size_t dlen, long timeout) 805 { 806 struct sock *sk = sock->sk; 807 struct tipc_sock *tsk = tipc_sk(sk); 808 struct tipc_msg *hdr = &tsk->phdr; 809 struct net *net = sock_net(sk); 810 int mtu = tipc_bcast_get_mtu(net); 811 struct tipc_mc_method *method = &tsk->mc_method; 812 struct sk_buff_head pkts; 813 struct tipc_nlist dsts; 814 int rc; 815 816 if (tsk->group) 817 return -EACCES; 818 819 /* Block or return if any destination link is congested */ 820 rc = tipc_wait_for_cond(sock, &timeout, !tsk->cong_link_cnt); 821 if (unlikely(rc)) 822 return rc; 823 824 /* Lookup destination nodes */ 825 tipc_nlist_init(&dsts, tipc_own_addr(net)); 826 tipc_nametbl_lookup_dst_nodes(net, seq->type, seq->lower, 827 seq->upper, &dsts); 828 if (!dsts.local && !dsts.remote) 829 return -EHOSTUNREACH; 830 831 /* Build message header */ 832 msg_set_type(hdr, TIPC_MCAST_MSG); 833 msg_set_hdr_sz(hdr, MCAST_H_SIZE); 834 msg_set_lookup_scope(hdr, TIPC_CLUSTER_SCOPE); 835 msg_set_destport(hdr, 0); 836 msg_set_destnode(hdr, 0); 837 msg_set_nametype(hdr, seq->type); 838 msg_set_namelower(hdr, seq->lower); 839 msg_set_nameupper(hdr, seq->upper); 840 841 /* Build message as chain of buffers */ 842 __skb_queue_head_init(&pkts); 843 rc = tipc_msg_build(hdr, msg, 0, dlen, mtu, &pkts); 844 845 /* Send message if build was successful */ 846 if (unlikely(rc == dlen)) { 847 trace_tipc_sk_sendmcast(sk, skb_peek(&pkts), 848 TIPC_DUMP_SK_SNDQ, " "); 849 rc = tipc_mcast_xmit(net, &pkts, method, &dsts, 850 &tsk->cong_link_cnt); 851 } 852 853 tipc_nlist_purge(&dsts); 854 855 return rc ? rc : dlen; 856 } 857 858 /** 859 * tipc_send_group_msg - send a message to a member in the group 860 * @net: network namespace 861 * @m: message to send 862 * @mb: group member 863 * @dnode: destination node 864 * @dport: destination port 865 * @dlen: total length of message data 866 */ 867 static int tipc_send_group_msg(struct net *net, struct tipc_sock *tsk, 868 struct msghdr *m, struct tipc_member *mb, 869 u32 dnode, u32 dport, int dlen) 870 { 871 u16 bc_snd_nxt = tipc_group_bc_snd_nxt(tsk->group); 872 struct tipc_mc_method *method = &tsk->mc_method; 873 int blks = tsk_blocks(GROUP_H_SIZE + dlen); 874 struct tipc_msg *hdr = &tsk->phdr; 875 struct sk_buff_head pkts; 876 int mtu, rc; 877 878 /* Complete message header */ 879 msg_set_type(hdr, TIPC_GRP_UCAST_MSG); 880 msg_set_hdr_sz(hdr, GROUP_H_SIZE); 881 msg_set_destport(hdr, dport); 882 msg_set_destnode(hdr, dnode); 883 msg_set_grp_bc_seqno(hdr, bc_snd_nxt); 884 885 /* Build message as chain of buffers */ 886 __skb_queue_head_init(&pkts); 887 mtu = tipc_node_get_mtu(net, dnode, tsk->portid, false); 888 rc = tipc_msg_build(hdr, m, 0, dlen, mtu, &pkts); 889 if (unlikely(rc != dlen)) 890 return rc; 891 892 /* Send message */ 893 rc = tipc_node_xmit(net, &pkts, dnode, tsk->portid); 894 if (unlikely(rc == -ELINKCONG)) { 895 tipc_dest_push(&tsk->cong_links, dnode, 0); 896 tsk->cong_link_cnt++; 897 } 898 899 /* Update send window */ 900 tipc_group_update_member(mb, blks); 901 902 /* A broadcast sent within next EXPIRE period must follow same path */ 903 method->rcast = true; 904 method->mandatory = true; 905 return dlen; 906 } 907 908 /** 909 * tipc_send_group_unicast - send message to a member in the group 910 * @sock: socket structure 911 * @m: message to send 912 * @dlen: total length of message data 913 * @timeout: timeout to wait for wakeup 914 * 915 * Called from function tipc_sendmsg(), which has done all sanity checks 916 * Returns the number of bytes sent on success, or errno 917 */ 918 static int tipc_send_group_unicast(struct socket *sock, struct msghdr *m, 919 int dlen, long timeout) 920 { 921 struct sock *sk = sock->sk; 922 DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name); 923 int blks = tsk_blocks(GROUP_H_SIZE + dlen); 924 struct tipc_sock *tsk = tipc_sk(sk); 925 struct net *net = sock_net(sk); 926 struct tipc_member *mb = NULL; 927 u32 node, port; 928 int rc; 929 930 node = dest->addr.id.node; 931 port = dest->addr.id.ref; 932 if (!port && !node) 933 return -EHOSTUNREACH; 934 935 /* Block or return if destination link or member is congested */ 936 rc = tipc_wait_for_cond(sock, &timeout, 937 !tipc_dest_find(&tsk->cong_links, node, 0) && 938 tsk->group && 939 !tipc_group_cong(tsk->group, node, port, blks, 940 &mb)); 941 if (unlikely(rc)) 942 return rc; 943 944 if (unlikely(!mb)) 945 return -EHOSTUNREACH; 946 947 rc = tipc_send_group_msg(net, tsk, m, mb, node, port, dlen); 948 949 return rc ? rc : dlen; 950 } 951 952 /** 953 * tipc_send_group_anycast - send message to any member with given identity 954 * @sock: socket structure 955 * @m: message to send 956 * @dlen: total length of message data 957 * @timeout: timeout to wait for wakeup 958 * 959 * Called from function tipc_sendmsg(), which has done all sanity checks 960 * Returns the number of bytes sent on success, or errno 961 */ 962 static int tipc_send_group_anycast(struct socket *sock, struct msghdr *m, 963 int dlen, long timeout) 964 { 965 DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name); 966 struct sock *sk = sock->sk; 967 struct tipc_sock *tsk = tipc_sk(sk); 968 struct list_head *cong_links = &tsk->cong_links; 969 int blks = tsk_blocks(GROUP_H_SIZE + dlen); 970 struct tipc_msg *hdr = &tsk->phdr; 971 struct tipc_member *first = NULL; 972 struct tipc_member *mbr = NULL; 973 struct net *net = sock_net(sk); 974 u32 node, port, exclude; 975 struct list_head dsts; 976 u32 type, inst, scope; 977 int lookups = 0; 978 int dstcnt, rc; 979 bool cong; 980 981 INIT_LIST_HEAD(&dsts); 982 983 type = msg_nametype(hdr); 984 inst = dest->addr.name.name.instance; 985 scope = msg_lookup_scope(hdr); 986 987 while (++lookups < 4) { 988 exclude = tipc_group_exclude(tsk->group); 989 990 first = NULL; 991 992 /* Look for a non-congested destination member, if any */ 993 while (1) { 994 if (!tipc_nametbl_lookup(net, type, inst, scope, &dsts, 995 &dstcnt, exclude, false)) 996 return -EHOSTUNREACH; 997 tipc_dest_pop(&dsts, &node, &port); 998 cong = tipc_group_cong(tsk->group, node, port, blks, 999 &mbr); 1000 if (!cong) 1001 break; 1002 if (mbr == first) 1003 break; 1004 if (!first) 1005 first = mbr; 1006 } 1007 1008 /* Start over if destination was not in member list */ 1009 if (unlikely(!mbr)) 1010 continue; 1011 1012 if (likely(!cong && !tipc_dest_find(cong_links, node, 0))) 1013 break; 1014 1015 /* Block or return if destination link or member is congested */ 1016 rc = tipc_wait_for_cond(sock, &timeout, 1017 !tipc_dest_find(cong_links, node, 0) && 1018 tsk->group && 1019 !tipc_group_cong(tsk->group, node, port, 1020 blks, &mbr)); 1021 if (unlikely(rc)) 1022 return rc; 1023 1024 /* Send, unless destination disappeared while waiting */ 1025 if (likely(mbr)) 1026 break; 1027 } 1028 1029 if (unlikely(lookups >= 4)) 1030 return -EHOSTUNREACH; 1031 1032 rc = tipc_send_group_msg(net, tsk, m, mbr, node, port, dlen); 1033 1034 return rc ? rc : dlen; 1035 } 1036 1037 /** 1038 * tipc_send_group_bcast - send message to all members in communication group 1039 * @sk: socket structure 1040 * @m: message to send 1041 * @dlen: total length of message data 1042 * @timeout: timeout to wait for wakeup 1043 * 1044 * Called from function tipc_sendmsg(), which has done all sanity checks 1045 * Returns the number of bytes sent on success, or errno 1046 */ 1047 static int tipc_send_group_bcast(struct socket *sock, struct msghdr *m, 1048 int dlen, long timeout) 1049 { 1050 DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name); 1051 struct sock *sk = sock->sk; 1052 struct net *net = sock_net(sk); 1053 struct tipc_sock *tsk = tipc_sk(sk); 1054 struct tipc_nlist *dsts; 1055 struct tipc_mc_method *method = &tsk->mc_method; 1056 bool ack = method->mandatory && method->rcast; 1057 int blks = tsk_blocks(MCAST_H_SIZE + dlen); 1058 struct tipc_msg *hdr = &tsk->phdr; 1059 int mtu = tipc_bcast_get_mtu(net); 1060 struct sk_buff_head pkts; 1061 int rc = -EHOSTUNREACH; 1062 1063 /* Block or return if any destination link or member is congested */ 1064 rc = tipc_wait_for_cond(sock, &timeout, 1065 !tsk->cong_link_cnt && tsk->group && 1066 !tipc_group_bc_cong(tsk->group, blks)); 1067 if (unlikely(rc)) 1068 return rc; 1069 1070 dsts = tipc_group_dests(tsk->group); 1071 if (!dsts->local && !dsts->remote) 1072 return -EHOSTUNREACH; 1073 1074 /* Complete message header */ 1075 if (dest) { 1076 msg_set_type(hdr, TIPC_GRP_MCAST_MSG); 1077 msg_set_nameinst(hdr, dest->addr.name.name.instance); 1078 } else { 1079 msg_set_type(hdr, TIPC_GRP_BCAST_MSG); 1080 msg_set_nameinst(hdr, 0); 1081 } 1082 msg_set_hdr_sz(hdr, GROUP_H_SIZE); 1083 msg_set_destport(hdr, 0); 1084 msg_set_destnode(hdr, 0); 1085 msg_set_grp_bc_seqno(hdr, tipc_group_bc_snd_nxt(tsk->group)); 1086 1087 /* Avoid getting stuck with repeated forced replicasts */ 1088 msg_set_grp_bc_ack_req(hdr, ack); 1089 1090 /* Build message as chain of buffers */ 1091 __skb_queue_head_init(&pkts); 1092 rc = tipc_msg_build(hdr, m, 0, dlen, mtu, &pkts); 1093 if (unlikely(rc != dlen)) 1094 return rc; 1095 1096 /* Send message */ 1097 rc = tipc_mcast_xmit(net, &pkts, method, dsts, &tsk->cong_link_cnt); 1098 if (unlikely(rc)) 1099 return rc; 1100 1101 /* Update broadcast sequence number and send windows */ 1102 tipc_group_update_bc_members(tsk->group, blks, ack); 1103 1104 /* Broadcast link is now free to choose method for next broadcast */ 1105 method->mandatory = false; 1106 method->expires = jiffies; 1107 1108 return dlen; 1109 } 1110 1111 /** 1112 * tipc_send_group_mcast - send message to all members with given identity 1113 * @sock: socket structure 1114 * @m: message to send 1115 * @dlen: total length of message data 1116 * @timeout: timeout to wait for wakeup 1117 * 1118 * Called from function tipc_sendmsg(), which has done all sanity checks 1119 * Returns the number of bytes sent on success, or errno 1120 */ 1121 static int tipc_send_group_mcast(struct socket *sock, struct msghdr *m, 1122 int dlen, long timeout) 1123 { 1124 struct sock *sk = sock->sk; 1125 DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name); 1126 struct tipc_sock *tsk = tipc_sk(sk); 1127 struct tipc_group *grp = tsk->group; 1128 struct tipc_msg *hdr = &tsk->phdr; 1129 struct net *net = sock_net(sk); 1130 u32 type, inst, scope, exclude; 1131 struct list_head dsts; 1132 u32 dstcnt; 1133 1134 INIT_LIST_HEAD(&dsts); 1135 1136 type = msg_nametype(hdr); 1137 inst = dest->addr.name.name.instance; 1138 scope = msg_lookup_scope(hdr); 1139 exclude = tipc_group_exclude(grp); 1140 1141 if (!tipc_nametbl_lookup(net, type, inst, scope, &dsts, 1142 &dstcnt, exclude, true)) 1143 return -EHOSTUNREACH; 1144 1145 if (dstcnt == 1) { 1146 tipc_dest_pop(&dsts, &dest->addr.id.node, &dest->addr.id.ref); 1147 return tipc_send_group_unicast(sock, m, dlen, timeout); 1148 } 1149 1150 tipc_dest_list_purge(&dsts); 1151 return tipc_send_group_bcast(sock, m, dlen, timeout); 1152 } 1153 1154 /** 1155 * tipc_sk_mcast_rcv - Deliver multicast messages to all destination sockets 1156 * @arrvq: queue with arriving messages, to be cloned after destination lookup 1157 * @inputq: queue with cloned messages, delivered to socket after dest lookup 1158 * 1159 * Multi-threaded: parallel calls with reference to same queues may occur 1160 */ 1161 void tipc_sk_mcast_rcv(struct net *net, struct sk_buff_head *arrvq, 1162 struct sk_buff_head *inputq) 1163 { 1164 u32 self = tipc_own_addr(net); 1165 u32 type, lower, upper, scope; 1166 struct sk_buff *skb, *_skb; 1167 u32 portid, onode; 1168 struct sk_buff_head tmpq; 1169 struct list_head dports; 1170 struct tipc_msg *hdr; 1171 int user, mtyp, hlen; 1172 bool exact; 1173 1174 __skb_queue_head_init(&tmpq); 1175 INIT_LIST_HEAD(&dports); 1176 1177 skb = tipc_skb_peek(arrvq, &inputq->lock); 1178 for (; skb; skb = tipc_skb_peek(arrvq, &inputq->lock)) { 1179 hdr = buf_msg(skb); 1180 user = msg_user(hdr); 1181 mtyp = msg_type(hdr); 1182 hlen = skb_headroom(skb) + msg_hdr_sz(hdr); 1183 onode = msg_orignode(hdr); 1184 type = msg_nametype(hdr); 1185 1186 if (mtyp == TIPC_GRP_UCAST_MSG || user == GROUP_PROTOCOL) { 1187 spin_lock_bh(&inputq->lock); 1188 if (skb_peek(arrvq) == skb) { 1189 __skb_dequeue(arrvq); 1190 __skb_queue_tail(inputq, skb); 1191 } 1192 kfree_skb(skb); 1193 spin_unlock_bh(&inputq->lock); 1194 continue; 1195 } 1196 1197 /* Group messages require exact scope match */ 1198 if (msg_in_group(hdr)) { 1199 lower = 0; 1200 upper = ~0; 1201 scope = msg_lookup_scope(hdr); 1202 exact = true; 1203 } else { 1204 /* TIPC_NODE_SCOPE means "any scope" in this context */ 1205 if (onode == self) 1206 scope = TIPC_NODE_SCOPE; 1207 else 1208 scope = TIPC_CLUSTER_SCOPE; 1209 exact = false; 1210 lower = msg_namelower(hdr); 1211 upper = msg_nameupper(hdr); 1212 } 1213 1214 /* Create destination port list: */ 1215 tipc_nametbl_mc_lookup(net, type, lower, upper, 1216 scope, exact, &dports); 1217 1218 /* Clone message per destination */ 1219 while (tipc_dest_pop(&dports, NULL, &portid)) { 1220 _skb = __pskb_copy(skb, hlen, GFP_ATOMIC); 1221 if (_skb) { 1222 msg_set_destport(buf_msg(_skb), portid); 1223 __skb_queue_tail(&tmpq, _skb); 1224 continue; 1225 } 1226 pr_warn("Failed to clone mcast rcv buffer\n"); 1227 } 1228 /* Append to inputq if not already done by other thread */ 1229 spin_lock_bh(&inputq->lock); 1230 if (skb_peek(arrvq) == skb) { 1231 skb_queue_splice_tail_init(&tmpq, inputq); 1232 kfree_skb(__skb_dequeue(arrvq)); 1233 } 1234 spin_unlock_bh(&inputq->lock); 1235 __skb_queue_purge(&tmpq); 1236 kfree_skb(skb); 1237 } 1238 tipc_sk_rcv(net, inputq); 1239 } 1240 1241 /* tipc_sk_push_backlog(): send accumulated buffers in socket write queue 1242 * when socket is in Nagle mode 1243 */ 1244 static void tipc_sk_push_backlog(struct tipc_sock *tsk) 1245 { 1246 struct sk_buff_head *txq = &tsk->sk.sk_write_queue; 1247 struct net *net = sock_net(&tsk->sk); 1248 u32 dnode = tsk_peer_node(tsk); 1249 struct sk_buff *skb = skb_peek(txq); 1250 int rc; 1251 1252 if (!skb || tsk->cong_link_cnt) 1253 return; 1254 1255 /* Do not send SYN again after congestion */ 1256 if (msg_is_syn(buf_msg(skb))) 1257 return; 1258 1259 tsk->snt_unacked += tsk->snd_backlog; 1260 tsk->snd_backlog = 0; 1261 tsk->expect_ack = true; 1262 rc = tipc_node_xmit(net, txq, dnode, tsk->portid); 1263 if (rc == -ELINKCONG) 1264 tsk->cong_link_cnt = 1; 1265 } 1266 1267 /** 1268 * tipc_sk_conn_proto_rcv - receive a connection mng protocol message 1269 * @tsk: receiving socket 1270 * @skb: pointer to message buffer. 1271 */ 1272 static void tipc_sk_conn_proto_rcv(struct tipc_sock *tsk, struct sk_buff *skb, 1273 struct sk_buff_head *inputq, 1274 struct sk_buff_head *xmitq) 1275 { 1276 struct tipc_msg *hdr = buf_msg(skb); 1277 u32 onode = tsk_own_node(tsk); 1278 struct sock *sk = &tsk->sk; 1279 int mtyp = msg_type(hdr); 1280 bool was_cong; 1281 1282 /* Ignore if connection cannot be validated: */ 1283 if (!tsk_peer_msg(tsk, hdr)) { 1284 trace_tipc_sk_drop_msg(sk, skb, TIPC_DUMP_NONE, "@proto_rcv!"); 1285 goto exit; 1286 } 1287 1288 if (unlikely(msg_errcode(hdr))) { 1289 tipc_set_sk_state(sk, TIPC_DISCONNECTING); 1290 tipc_node_remove_conn(sock_net(sk), tsk_peer_node(tsk), 1291 tsk_peer_port(tsk)); 1292 sk->sk_state_change(sk); 1293 1294 /* State change is ignored if socket already awake, 1295 * - convert msg to abort msg and add to inqueue 1296 */ 1297 msg_set_user(hdr, TIPC_CRITICAL_IMPORTANCE); 1298 msg_set_type(hdr, TIPC_CONN_MSG); 1299 msg_set_size(hdr, BASIC_H_SIZE); 1300 msg_set_hdr_sz(hdr, BASIC_H_SIZE); 1301 __skb_queue_tail(inputq, skb); 1302 return; 1303 } 1304 1305 tsk->probe_unacked = false; 1306 1307 if (mtyp == CONN_PROBE) { 1308 msg_set_type(hdr, CONN_PROBE_REPLY); 1309 if (tipc_msg_reverse(onode, &skb, TIPC_OK)) 1310 __skb_queue_tail(xmitq, skb); 1311 return; 1312 } else if (mtyp == CONN_ACK) { 1313 was_cong = tsk_conn_cong(tsk); 1314 tsk->expect_ack = false; 1315 tipc_sk_push_backlog(tsk); 1316 tsk->snt_unacked -= msg_conn_ack(hdr); 1317 if (tsk->peer_caps & TIPC_BLOCK_FLOWCTL) 1318 tsk->snd_win = msg_adv_win(hdr); 1319 if (was_cong && !tsk_conn_cong(tsk)) 1320 sk->sk_write_space(sk); 1321 } else if (mtyp != CONN_PROBE_REPLY) { 1322 pr_warn("Received unknown CONN_PROTO msg\n"); 1323 } 1324 exit: 1325 kfree_skb(skb); 1326 } 1327 1328 /** 1329 * tipc_sendmsg - send message in connectionless manner 1330 * @sock: socket structure 1331 * @m: message to send 1332 * @dsz: amount of user data to be sent 1333 * 1334 * Message must have an destination specified explicitly. 1335 * Used for SOCK_RDM and SOCK_DGRAM messages, 1336 * and for 'SYN' messages on SOCK_SEQPACKET and SOCK_STREAM connections. 1337 * (Note: 'SYN+' is prohibited on SOCK_STREAM.) 1338 * 1339 * Returns the number of bytes sent on success, or errno otherwise 1340 */ 1341 static int tipc_sendmsg(struct socket *sock, 1342 struct msghdr *m, size_t dsz) 1343 { 1344 struct sock *sk = sock->sk; 1345 int ret; 1346 1347 lock_sock(sk); 1348 ret = __tipc_sendmsg(sock, m, dsz); 1349 release_sock(sk); 1350 1351 return ret; 1352 } 1353 1354 static int __tipc_sendmsg(struct socket *sock, struct msghdr *m, size_t dlen) 1355 { 1356 struct sock *sk = sock->sk; 1357 struct net *net = sock_net(sk); 1358 struct tipc_sock *tsk = tipc_sk(sk); 1359 DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name); 1360 long timeout = sock_sndtimeo(sk, m->msg_flags & MSG_DONTWAIT); 1361 struct list_head *clinks = &tsk->cong_links; 1362 bool syn = !tipc_sk_type_connectionless(sk); 1363 struct tipc_group *grp = tsk->group; 1364 struct tipc_msg *hdr = &tsk->phdr; 1365 struct tipc_name_seq *seq; 1366 struct sk_buff_head pkts; 1367 u32 dport, dnode = 0; 1368 u32 type, inst; 1369 int mtu, rc; 1370 1371 if (unlikely(dlen > TIPC_MAX_USER_MSG_SIZE)) 1372 return -EMSGSIZE; 1373 1374 if (likely(dest)) { 1375 if (unlikely(m->msg_namelen < sizeof(*dest))) 1376 return -EINVAL; 1377 if (unlikely(dest->family != AF_TIPC)) 1378 return -EINVAL; 1379 } 1380 1381 if (grp) { 1382 if (!dest) 1383 return tipc_send_group_bcast(sock, m, dlen, timeout); 1384 if (dest->addrtype == TIPC_ADDR_NAME) 1385 return tipc_send_group_anycast(sock, m, dlen, timeout); 1386 if (dest->addrtype == TIPC_ADDR_ID) 1387 return tipc_send_group_unicast(sock, m, dlen, timeout); 1388 if (dest->addrtype == TIPC_ADDR_MCAST) 1389 return tipc_send_group_mcast(sock, m, dlen, timeout); 1390 return -EINVAL; 1391 } 1392 1393 if (unlikely(!dest)) { 1394 dest = &tsk->peer; 1395 if (!syn && dest->family != AF_TIPC) 1396 return -EDESTADDRREQ; 1397 } 1398 1399 if (unlikely(syn)) { 1400 if (sk->sk_state == TIPC_LISTEN) 1401 return -EPIPE; 1402 if (sk->sk_state != TIPC_OPEN) 1403 return -EISCONN; 1404 if (tsk->published) 1405 return -EOPNOTSUPP; 1406 if (dest->addrtype == TIPC_ADDR_NAME) { 1407 tsk->conn_type = dest->addr.name.name.type; 1408 tsk->conn_instance = dest->addr.name.name.instance; 1409 } 1410 msg_set_syn(hdr, 1); 1411 } 1412 1413 seq = &dest->addr.nameseq; 1414 if (dest->addrtype == TIPC_ADDR_MCAST) 1415 return tipc_sendmcast(sock, seq, m, dlen, timeout); 1416 1417 if (dest->addrtype == TIPC_ADDR_NAME) { 1418 type = dest->addr.name.name.type; 1419 inst = dest->addr.name.name.instance; 1420 dnode = dest->addr.name.domain; 1421 msg_set_type(hdr, TIPC_NAMED_MSG); 1422 msg_set_hdr_sz(hdr, NAMED_H_SIZE); 1423 msg_set_nametype(hdr, type); 1424 msg_set_nameinst(hdr, inst); 1425 msg_set_lookup_scope(hdr, tipc_node2scope(dnode)); 1426 dport = tipc_nametbl_translate(net, type, inst, &dnode); 1427 msg_set_destnode(hdr, dnode); 1428 msg_set_destport(hdr, dport); 1429 if (unlikely(!dport && !dnode)) 1430 return -EHOSTUNREACH; 1431 } else if (dest->addrtype == TIPC_ADDR_ID) { 1432 dnode = dest->addr.id.node; 1433 msg_set_type(hdr, TIPC_DIRECT_MSG); 1434 msg_set_lookup_scope(hdr, 0); 1435 msg_set_destnode(hdr, dnode); 1436 msg_set_destport(hdr, dest->addr.id.ref); 1437 msg_set_hdr_sz(hdr, BASIC_H_SIZE); 1438 } else { 1439 return -EINVAL; 1440 } 1441 1442 /* Block or return if destination link is congested */ 1443 rc = tipc_wait_for_cond(sock, &timeout, 1444 !tipc_dest_find(clinks, dnode, 0)); 1445 if (unlikely(rc)) 1446 return rc; 1447 1448 __skb_queue_head_init(&pkts); 1449 mtu = tipc_node_get_mtu(net, dnode, tsk->portid, false); 1450 rc = tipc_msg_build(hdr, m, 0, dlen, mtu, &pkts); 1451 if (unlikely(rc != dlen)) 1452 return rc; 1453 if (unlikely(syn && !tipc_msg_skb_clone(&pkts, &sk->sk_write_queue))) { 1454 __skb_queue_purge(&pkts); 1455 return -ENOMEM; 1456 } 1457 1458 trace_tipc_sk_sendmsg(sk, skb_peek(&pkts), TIPC_DUMP_SK_SNDQ, " "); 1459 rc = tipc_node_xmit(net, &pkts, dnode, tsk->portid); 1460 if (unlikely(rc == -ELINKCONG)) { 1461 tipc_dest_push(clinks, dnode, 0); 1462 tsk->cong_link_cnt++; 1463 rc = 0; 1464 } 1465 1466 if (unlikely(syn && !rc)) 1467 tipc_set_sk_state(sk, TIPC_CONNECTING); 1468 1469 return rc ? rc : dlen; 1470 } 1471 1472 /** 1473 * tipc_sendstream - send stream-oriented data 1474 * @sock: socket structure 1475 * @m: data to send 1476 * @dsz: total length of data to be transmitted 1477 * 1478 * Used for SOCK_STREAM data. 1479 * 1480 * Returns the number of bytes sent on success (or partial success), 1481 * or errno if no data sent 1482 */ 1483 static int tipc_sendstream(struct socket *sock, struct msghdr *m, size_t dsz) 1484 { 1485 struct sock *sk = sock->sk; 1486 int ret; 1487 1488 lock_sock(sk); 1489 ret = __tipc_sendstream(sock, m, dsz); 1490 release_sock(sk); 1491 1492 return ret; 1493 } 1494 1495 static int __tipc_sendstream(struct socket *sock, struct msghdr *m, size_t dlen) 1496 { 1497 struct sock *sk = sock->sk; 1498 DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name); 1499 long timeout = sock_sndtimeo(sk, m->msg_flags & MSG_DONTWAIT); 1500 struct sk_buff_head *txq = &sk->sk_write_queue; 1501 struct tipc_sock *tsk = tipc_sk(sk); 1502 struct tipc_msg *hdr = &tsk->phdr; 1503 struct net *net = sock_net(sk); 1504 u32 dnode = tsk_peer_node(tsk); 1505 int maxnagle = tsk->maxnagle; 1506 int maxpkt = tsk->max_pkt; 1507 int send, sent = 0; 1508 int blocks, rc = 0; 1509 1510 if (unlikely(dlen > INT_MAX)) 1511 return -EMSGSIZE; 1512 1513 /* Handle implicit connection setup */ 1514 if (unlikely(dest)) { 1515 rc = __tipc_sendmsg(sock, m, dlen); 1516 if (dlen && dlen == rc) { 1517 tsk->peer_caps = tipc_node_get_capabilities(net, dnode); 1518 tsk->snt_unacked = tsk_inc(tsk, dlen + msg_hdr_sz(hdr)); 1519 } 1520 return rc; 1521 } 1522 1523 do { 1524 rc = tipc_wait_for_cond(sock, &timeout, 1525 (!tsk->cong_link_cnt && 1526 !tsk_conn_cong(tsk) && 1527 tipc_sk_connected(sk))); 1528 if (unlikely(rc)) 1529 break; 1530 send = min_t(size_t, dlen - sent, TIPC_MAX_USER_MSG_SIZE); 1531 blocks = tsk->snd_backlog; 1532 if (tsk->oneway++ >= 4 && send <= maxnagle) { 1533 rc = tipc_msg_append(hdr, m, send, maxnagle, txq); 1534 if (unlikely(rc < 0)) 1535 break; 1536 blocks += rc; 1537 if (blocks <= 64 && tsk->expect_ack) { 1538 tsk->snd_backlog = blocks; 1539 sent += send; 1540 break; 1541 } 1542 tsk->expect_ack = true; 1543 } else { 1544 rc = tipc_msg_build(hdr, m, sent, send, maxpkt, txq); 1545 if (unlikely(rc != send)) 1546 break; 1547 blocks += tsk_inc(tsk, send + MIN_H_SIZE); 1548 } 1549 trace_tipc_sk_sendstream(sk, skb_peek(txq), 1550 TIPC_DUMP_SK_SNDQ, " "); 1551 rc = tipc_node_xmit(net, txq, dnode, tsk->portid); 1552 if (unlikely(rc == -ELINKCONG)) { 1553 tsk->cong_link_cnt = 1; 1554 rc = 0; 1555 } 1556 if (likely(!rc)) { 1557 tsk->snt_unacked += blocks; 1558 tsk->snd_backlog = 0; 1559 sent += send; 1560 } 1561 } while (sent < dlen && !rc); 1562 1563 return sent ? sent : rc; 1564 } 1565 1566 /** 1567 * tipc_send_packet - send a connection-oriented message 1568 * @sock: socket structure 1569 * @m: message to send 1570 * @dsz: length of data to be transmitted 1571 * 1572 * Used for SOCK_SEQPACKET messages. 1573 * 1574 * Returns the number of bytes sent on success, or errno otherwise 1575 */ 1576 static int tipc_send_packet(struct socket *sock, struct msghdr *m, size_t dsz) 1577 { 1578 if (dsz > TIPC_MAX_USER_MSG_SIZE) 1579 return -EMSGSIZE; 1580 1581 return tipc_sendstream(sock, m, dsz); 1582 } 1583 1584 /* tipc_sk_finish_conn - complete the setup of a connection 1585 */ 1586 static void tipc_sk_finish_conn(struct tipc_sock *tsk, u32 peer_port, 1587 u32 peer_node) 1588 { 1589 struct sock *sk = &tsk->sk; 1590 struct net *net = sock_net(sk); 1591 struct tipc_msg *msg = &tsk->phdr; 1592 1593 msg_set_syn(msg, 0); 1594 msg_set_destnode(msg, peer_node); 1595 msg_set_destport(msg, peer_port); 1596 msg_set_type(msg, TIPC_CONN_MSG); 1597 msg_set_lookup_scope(msg, 0); 1598 msg_set_hdr_sz(msg, SHORT_H_SIZE); 1599 1600 sk_reset_timer(sk, &sk->sk_timer, jiffies + CONN_PROBING_INTV); 1601 tipc_set_sk_state(sk, TIPC_ESTABLISHED); 1602 tipc_node_add_conn(net, peer_node, tsk->portid, peer_port); 1603 tsk->max_pkt = tipc_node_get_mtu(net, peer_node, tsk->portid, true); 1604 tsk->peer_caps = tipc_node_get_capabilities(net, peer_node); 1605 tsk_set_nagle(tsk); 1606 __skb_queue_purge(&sk->sk_write_queue); 1607 if (tsk->peer_caps & TIPC_BLOCK_FLOWCTL) 1608 return; 1609 1610 /* Fall back to message based flow control */ 1611 tsk->rcv_win = FLOWCTL_MSG_WIN; 1612 tsk->snd_win = FLOWCTL_MSG_WIN; 1613 } 1614 1615 /** 1616 * tipc_sk_set_orig_addr - capture sender's address for received message 1617 * @m: descriptor for message info 1618 * @hdr: received message header 1619 * 1620 * Note: Address is not captured if not requested by receiver. 1621 */ 1622 static void tipc_sk_set_orig_addr(struct msghdr *m, struct sk_buff *skb) 1623 { 1624 DECLARE_SOCKADDR(struct sockaddr_pair *, srcaddr, m->msg_name); 1625 struct tipc_msg *hdr = buf_msg(skb); 1626 1627 if (!srcaddr) 1628 return; 1629 1630 srcaddr->sock.family = AF_TIPC; 1631 srcaddr->sock.addrtype = TIPC_ADDR_ID; 1632 srcaddr->sock.scope = 0; 1633 srcaddr->sock.addr.id.ref = msg_origport(hdr); 1634 srcaddr->sock.addr.id.node = msg_orignode(hdr); 1635 srcaddr->sock.addr.name.domain = 0; 1636 m->msg_namelen = sizeof(struct sockaddr_tipc); 1637 1638 if (!msg_in_group(hdr)) 1639 return; 1640 1641 /* Group message users may also want to know sending member's id */ 1642 srcaddr->member.family = AF_TIPC; 1643 srcaddr->member.addrtype = TIPC_ADDR_NAME; 1644 srcaddr->member.scope = 0; 1645 srcaddr->member.addr.name.name.type = msg_nametype(hdr); 1646 srcaddr->member.addr.name.name.instance = TIPC_SKB_CB(skb)->orig_member; 1647 srcaddr->member.addr.name.domain = 0; 1648 m->msg_namelen = sizeof(*srcaddr); 1649 } 1650 1651 /** 1652 * tipc_sk_anc_data_recv - optionally capture ancillary data for received message 1653 * @m: descriptor for message info 1654 * @skb: received message buffer 1655 * @tsk: TIPC port associated with message 1656 * 1657 * Note: Ancillary data is not captured if not requested by receiver. 1658 * 1659 * Returns 0 if successful, otherwise errno 1660 */ 1661 static int tipc_sk_anc_data_recv(struct msghdr *m, struct sk_buff *skb, 1662 struct tipc_sock *tsk) 1663 { 1664 struct tipc_msg *msg; 1665 u32 anc_data[3]; 1666 u32 err; 1667 u32 dest_type; 1668 int has_name; 1669 int res; 1670 1671 if (likely(m->msg_controllen == 0)) 1672 return 0; 1673 msg = buf_msg(skb); 1674 1675 /* Optionally capture errored message object(s) */ 1676 err = msg ? msg_errcode(msg) : 0; 1677 if (unlikely(err)) { 1678 anc_data[0] = err; 1679 anc_data[1] = msg_data_sz(msg); 1680 res = put_cmsg(m, SOL_TIPC, TIPC_ERRINFO, 8, anc_data); 1681 if (res) 1682 return res; 1683 if (anc_data[1]) { 1684 if (skb_linearize(skb)) 1685 return -ENOMEM; 1686 msg = buf_msg(skb); 1687 res = put_cmsg(m, SOL_TIPC, TIPC_RETDATA, anc_data[1], 1688 msg_data(msg)); 1689 if (res) 1690 return res; 1691 } 1692 } 1693 1694 /* Optionally capture message destination object */ 1695 dest_type = msg ? msg_type(msg) : TIPC_DIRECT_MSG; 1696 switch (dest_type) { 1697 case TIPC_NAMED_MSG: 1698 has_name = 1; 1699 anc_data[0] = msg_nametype(msg); 1700 anc_data[1] = msg_namelower(msg); 1701 anc_data[2] = msg_namelower(msg); 1702 break; 1703 case TIPC_MCAST_MSG: 1704 has_name = 1; 1705 anc_data[0] = msg_nametype(msg); 1706 anc_data[1] = msg_namelower(msg); 1707 anc_data[2] = msg_nameupper(msg); 1708 break; 1709 case TIPC_CONN_MSG: 1710 has_name = (tsk->conn_type != 0); 1711 anc_data[0] = tsk->conn_type; 1712 anc_data[1] = tsk->conn_instance; 1713 anc_data[2] = tsk->conn_instance; 1714 break; 1715 default: 1716 has_name = 0; 1717 } 1718 if (has_name) { 1719 res = put_cmsg(m, SOL_TIPC, TIPC_DESTNAME, 12, anc_data); 1720 if (res) 1721 return res; 1722 } 1723 1724 return 0; 1725 } 1726 1727 static void tipc_sk_send_ack(struct tipc_sock *tsk) 1728 { 1729 struct sock *sk = &tsk->sk; 1730 struct net *net = sock_net(sk); 1731 struct sk_buff *skb = NULL; 1732 struct tipc_msg *msg; 1733 u32 peer_port = tsk_peer_port(tsk); 1734 u32 dnode = tsk_peer_node(tsk); 1735 1736 if (!tipc_sk_connected(sk)) 1737 return; 1738 skb = tipc_msg_create(CONN_MANAGER, CONN_ACK, INT_H_SIZE, 0, 1739 dnode, tsk_own_node(tsk), peer_port, 1740 tsk->portid, TIPC_OK); 1741 if (!skb) 1742 return; 1743 msg = buf_msg(skb); 1744 msg_set_conn_ack(msg, tsk->rcv_unacked); 1745 tsk->rcv_unacked = 0; 1746 1747 /* Adjust to and advertize the correct window limit */ 1748 if (tsk->peer_caps & TIPC_BLOCK_FLOWCTL) { 1749 tsk->rcv_win = tsk_adv_blocks(tsk->sk.sk_rcvbuf); 1750 msg_set_adv_win(msg, tsk->rcv_win); 1751 } 1752 tipc_node_xmit_skb(net, skb, dnode, msg_link_selector(msg)); 1753 } 1754 1755 static int tipc_wait_for_rcvmsg(struct socket *sock, long *timeop) 1756 { 1757 struct sock *sk = sock->sk; 1758 DEFINE_WAIT_FUNC(wait, woken_wake_function); 1759 long timeo = *timeop; 1760 int err = sock_error(sk); 1761 1762 if (err) 1763 return err; 1764 1765 for (;;) { 1766 if (timeo && skb_queue_empty(&sk->sk_receive_queue)) { 1767 if (sk->sk_shutdown & RCV_SHUTDOWN) { 1768 err = -ENOTCONN; 1769 break; 1770 } 1771 add_wait_queue(sk_sleep(sk), &wait); 1772 release_sock(sk); 1773 timeo = wait_woken(&wait, TASK_INTERRUPTIBLE, timeo); 1774 sched_annotate_sleep(); 1775 lock_sock(sk); 1776 remove_wait_queue(sk_sleep(sk), &wait); 1777 } 1778 err = 0; 1779 if (!skb_queue_empty(&sk->sk_receive_queue)) 1780 break; 1781 err = -EAGAIN; 1782 if (!timeo) 1783 break; 1784 err = sock_intr_errno(timeo); 1785 if (signal_pending(current)) 1786 break; 1787 1788 err = sock_error(sk); 1789 if (err) 1790 break; 1791 } 1792 *timeop = timeo; 1793 return err; 1794 } 1795 1796 /** 1797 * tipc_recvmsg - receive packet-oriented message 1798 * @m: descriptor for message info 1799 * @buflen: length of user buffer area 1800 * @flags: receive flags 1801 * 1802 * Used for SOCK_DGRAM, SOCK_RDM, and SOCK_SEQPACKET messages. 1803 * If the complete message doesn't fit in user area, truncate it. 1804 * 1805 * Returns size of returned message data, errno otherwise 1806 */ 1807 static int tipc_recvmsg(struct socket *sock, struct msghdr *m, 1808 size_t buflen, int flags) 1809 { 1810 struct sock *sk = sock->sk; 1811 bool connected = !tipc_sk_type_connectionless(sk); 1812 struct tipc_sock *tsk = tipc_sk(sk); 1813 int rc, err, hlen, dlen, copy; 1814 struct sk_buff_head xmitq; 1815 struct tipc_msg *hdr; 1816 struct sk_buff *skb; 1817 bool grp_evt; 1818 long timeout; 1819 1820 /* Catch invalid receive requests */ 1821 if (unlikely(!buflen)) 1822 return -EINVAL; 1823 1824 lock_sock(sk); 1825 if (unlikely(connected && sk->sk_state == TIPC_OPEN)) { 1826 rc = -ENOTCONN; 1827 goto exit; 1828 } 1829 timeout = sock_rcvtimeo(sk, flags & MSG_DONTWAIT); 1830 1831 /* Step rcv queue to first msg with data or error; wait if necessary */ 1832 do { 1833 rc = tipc_wait_for_rcvmsg(sock, &timeout); 1834 if (unlikely(rc)) 1835 goto exit; 1836 skb = skb_peek(&sk->sk_receive_queue); 1837 hdr = buf_msg(skb); 1838 dlen = msg_data_sz(hdr); 1839 hlen = msg_hdr_sz(hdr); 1840 err = msg_errcode(hdr); 1841 grp_evt = msg_is_grp_evt(hdr); 1842 if (likely(dlen || err)) 1843 break; 1844 tsk_advance_rx_queue(sk); 1845 } while (1); 1846 1847 /* Collect msg meta data, including error code and rejected data */ 1848 tipc_sk_set_orig_addr(m, skb); 1849 rc = tipc_sk_anc_data_recv(m, skb, tsk); 1850 if (unlikely(rc)) 1851 goto exit; 1852 hdr = buf_msg(skb); 1853 1854 /* Capture data if non-error msg, otherwise just set return value */ 1855 if (likely(!err)) { 1856 copy = min_t(int, dlen, buflen); 1857 if (unlikely(copy != dlen)) 1858 m->msg_flags |= MSG_TRUNC; 1859 rc = skb_copy_datagram_msg(skb, hlen, m, copy); 1860 } else { 1861 copy = 0; 1862 rc = 0; 1863 if (err != TIPC_CONN_SHUTDOWN && connected && !m->msg_control) 1864 rc = -ECONNRESET; 1865 } 1866 if (unlikely(rc)) 1867 goto exit; 1868 1869 /* Mark message as group event if applicable */ 1870 if (unlikely(grp_evt)) { 1871 if (msg_grp_evt(hdr) == TIPC_WITHDRAWN) 1872 m->msg_flags |= MSG_EOR; 1873 m->msg_flags |= MSG_OOB; 1874 copy = 0; 1875 } 1876 1877 /* Caption of data or error code/rejected data was successful */ 1878 if (unlikely(flags & MSG_PEEK)) 1879 goto exit; 1880 1881 /* Send group flow control advertisement when applicable */ 1882 if (tsk->group && msg_in_group(hdr) && !grp_evt) { 1883 __skb_queue_head_init(&xmitq); 1884 tipc_group_update_rcv_win(tsk->group, tsk_blocks(hlen + dlen), 1885 msg_orignode(hdr), msg_origport(hdr), 1886 &xmitq); 1887 tipc_node_distr_xmit(sock_net(sk), &xmitq); 1888 } 1889 1890 tsk_advance_rx_queue(sk); 1891 1892 if (likely(!connected)) 1893 goto exit; 1894 1895 /* Send connection flow control advertisement when applicable */ 1896 tsk->rcv_unacked += tsk_inc(tsk, hlen + dlen); 1897 if (tsk->rcv_unacked >= tsk->rcv_win / TIPC_ACK_RATE) 1898 tipc_sk_send_ack(tsk); 1899 exit: 1900 release_sock(sk); 1901 return rc ? rc : copy; 1902 } 1903 1904 /** 1905 * tipc_recvstream - receive stream-oriented data 1906 * @m: descriptor for message info 1907 * @buflen: total size of user buffer area 1908 * @flags: receive flags 1909 * 1910 * Used for SOCK_STREAM messages only. If not enough data is available 1911 * will optionally wait for more; never truncates data. 1912 * 1913 * Returns size of returned message data, errno otherwise 1914 */ 1915 static int tipc_recvstream(struct socket *sock, struct msghdr *m, 1916 size_t buflen, int flags) 1917 { 1918 struct sock *sk = sock->sk; 1919 struct tipc_sock *tsk = tipc_sk(sk); 1920 struct sk_buff *skb; 1921 struct tipc_msg *hdr; 1922 struct tipc_skb_cb *skb_cb; 1923 bool peek = flags & MSG_PEEK; 1924 int offset, required, copy, copied = 0; 1925 int hlen, dlen, err, rc; 1926 bool ack = false; 1927 long timeout; 1928 1929 /* Catch invalid receive attempts */ 1930 if (unlikely(!buflen)) 1931 return -EINVAL; 1932 1933 lock_sock(sk); 1934 1935 if (unlikely(sk->sk_state == TIPC_OPEN)) { 1936 rc = -ENOTCONN; 1937 goto exit; 1938 } 1939 required = sock_rcvlowat(sk, flags & MSG_WAITALL, buflen); 1940 timeout = sock_rcvtimeo(sk, flags & MSG_DONTWAIT); 1941 1942 do { 1943 /* Look at first msg in receive queue; wait if necessary */ 1944 rc = tipc_wait_for_rcvmsg(sock, &timeout); 1945 if (unlikely(rc)) 1946 break; 1947 skb = skb_peek(&sk->sk_receive_queue); 1948 skb_cb = TIPC_SKB_CB(skb); 1949 hdr = buf_msg(skb); 1950 dlen = msg_data_sz(hdr); 1951 hlen = msg_hdr_sz(hdr); 1952 err = msg_errcode(hdr); 1953 1954 /* Discard any empty non-errored (SYN-) message */ 1955 if (unlikely(!dlen && !err)) { 1956 tsk_advance_rx_queue(sk); 1957 continue; 1958 } 1959 1960 /* Collect msg meta data, incl. error code and rejected data */ 1961 if (!copied) { 1962 tipc_sk_set_orig_addr(m, skb); 1963 rc = tipc_sk_anc_data_recv(m, skb, tsk); 1964 if (rc) 1965 break; 1966 hdr = buf_msg(skb); 1967 } 1968 1969 /* Copy data if msg ok, otherwise return error/partial data */ 1970 if (likely(!err)) { 1971 ack = msg_ack_required(hdr); 1972 offset = skb_cb->bytes_read; 1973 copy = min_t(int, dlen - offset, buflen - copied); 1974 rc = skb_copy_datagram_msg(skb, hlen + offset, m, copy); 1975 if (unlikely(rc)) 1976 break; 1977 copied += copy; 1978 offset += copy; 1979 if (unlikely(offset < dlen)) { 1980 if (!peek) 1981 skb_cb->bytes_read = offset; 1982 break; 1983 } 1984 } else { 1985 rc = 0; 1986 if ((err != TIPC_CONN_SHUTDOWN) && !m->msg_control) 1987 rc = -ECONNRESET; 1988 if (copied || rc) 1989 break; 1990 } 1991 1992 if (unlikely(peek)) 1993 break; 1994 1995 tsk_advance_rx_queue(sk); 1996 1997 /* Send connection flow control advertisement when applicable */ 1998 tsk->rcv_unacked += tsk_inc(tsk, hlen + dlen); 1999 if (ack || tsk->rcv_unacked >= tsk->rcv_win / TIPC_ACK_RATE) 2000 tipc_sk_send_ack(tsk); 2001 2002 /* Exit if all requested data or FIN/error received */ 2003 if (copied == buflen || err) 2004 break; 2005 2006 } while (!skb_queue_empty(&sk->sk_receive_queue) || copied < required); 2007 exit: 2008 release_sock(sk); 2009 return copied ? copied : rc; 2010 } 2011 2012 /** 2013 * tipc_write_space - wake up thread if port congestion is released 2014 * @sk: socket 2015 */ 2016 static void tipc_write_space(struct sock *sk) 2017 { 2018 struct socket_wq *wq; 2019 2020 rcu_read_lock(); 2021 wq = rcu_dereference(sk->sk_wq); 2022 if (skwq_has_sleeper(wq)) 2023 wake_up_interruptible_sync_poll(&wq->wait, EPOLLOUT | 2024 EPOLLWRNORM | EPOLLWRBAND); 2025 rcu_read_unlock(); 2026 } 2027 2028 /** 2029 * tipc_data_ready - wake up threads to indicate messages have been received 2030 * @sk: socket 2031 * @len: the length of messages 2032 */ 2033 static void tipc_data_ready(struct sock *sk) 2034 { 2035 struct socket_wq *wq; 2036 2037 rcu_read_lock(); 2038 wq = rcu_dereference(sk->sk_wq); 2039 if (skwq_has_sleeper(wq)) 2040 wake_up_interruptible_sync_poll(&wq->wait, EPOLLIN | 2041 EPOLLRDNORM | EPOLLRDBAND); 2042 rcu_read_unlock(); 2043 } 2044 2045 static void tipc_sock_destruct(struct sock *sk) 2046 { 2047 __skb_queue_purge(&sk->sk_receive_queue); 2048 } 2049 2050 static void tipc_sk_proto_rcv(struct sock *sk, 2051 struct sk_buff_head *inputq, 2052 struct sk_buff_head *xmitq) 2053 { 2054 struct sk_buff *skb = __skb_dequeue(inputq); 2055 struct tipc_sock *tsk = tipc_sk(sk); 2056 struct tipc_msg *hdr = buf_msg(skb); 2057 struct tipc_group *grp = tsk->group; 2058 bool wakeup = false; 2059 2060 switch (msg_user(hdr)) { 2061 case CONN_MANAGER: 2062 tipc_sk_conn_proto_rcv(tsk, skb, inputq, xmitq); 2063 return; 2064 case SOCK_WAKEUP: 2065 tipc_dest_del(&tsk->cong_links, msg_orignode(hdr), 0); 2066 /* coupled with smp_rmb() in tipc_wait_for_cond() */ 2067 smp_wmb(); 2068 tsk->cong_link_cnt--; 2069 wakeup = true; 2070 tipc_sk_push_backlog(tsk); 2071 break; 2072 case GROUP_PROTOCOL: 2073 tipc_group_proto_rcv(grp, &wakeup, hdr, inputq, xmitq); 2074 break; 2075 case TOP_SRV: 2076 tipc_group_member_evt(tsk->group, &wakeup, &sk->sk_rcvbuf, 2077 hdr, inputq, xmitq); 2078 break; 2079 default: 2080 break; 2081 } 2082 2083 if (wakeup) 2084 sk->sk_write_space(sk); 2085 2086 kfree_skb(skb); 2087 } 2088 2089 /** 2090 * tipc_sk_filter_connect - check incoming message for a connection-based socket 2091 * @tsk: TIPC socket 2092 * @skb: pointer to message buffer. 2093 * Returns true if message should be added to receive queue, false otherwise 2094 */ 2095 static bool tipc_sk_filter_connect(struct tipc_sock *tsk, struct sk_buff *skb) 2096 { 2097 struct sock *sk = &tsk->sk; 2098 struct net *net = sock_net(sk); 2099 struct tipc_msg *hdr = buf_msg(skb); 2100 bool con_msg = msg_connected(hdr); 2101 u32 pport = tsk_peer_port(tsk); 2102 u32 pnode = tsk_peer_node(tsk); 2103 u32 oport = msg_origport(hdr); 2104 u32 onode = msg_orignode(hdr); 2105 int err = msg_errcode(hdr); 2106 unsigned long delay; 2107 2108 if (unlikely(msg_mcast(hdr))) 2109 return false; 2110 tsk->oneway = 0; 2111 2112 switch (sk->sk_state) { 2113 case TIPC_CONNECTING: 2114 /* Setup ACK */ 2115 if (likely(con_msg)) { 2116 if (err) 2117 break; 2118 tipc_sk_finish_conn(tsk, oport, onode); 2119 msg_set_importance(&tsk->phdr, msg_importance(hdr)); 2120 /* ACK+ message with data is added to receive queue */ 2121 if (msg_data_sz(hdr)) 2122 return true; 2123 /* Empty ACK-, - wake up sleeping connect() and drop */ 2124 sk->sk_state_change(sk); 2125 msg_set_dest_droppable(hdr, 1); 2126 return false; 2127 } 2128 /* Ignore connectionless message if not from listening socket */ 2129 if (oport != pport || onode != pnode) 2130 return false; 2131 2132 /* Rejected SYN */ 2133 if (err != TIPC_ERR_OVERLOAD) 2134 break; 2135 2136 /* Prepare for new setup attempt if we have a SYN clone */ 2137 if (skb_queue_empty(&sk->sk_write_queue)) 2138 break; 2139 get_random_bytes(&delay, 2); 2140 delay %= (tsk->conn_timeout / 4); 2141 delay = msecs_to_jiffies(delay + 100); 2142 sk_reset_timer(sk, &sk->sk_timer, jiffies + delay); 2143 return false; 2144 case TIPC_OPEN: 2145 case TIPC_DISCONNECTING: 2146 return false; 2147 case TIPC_LISTEN: 2148 /* Accept only SYN message */ 2149 if (!msg_is_syn(hdr) && 2150 tipc_node_get_capabilities(net, onode) & TIPC_SYN_BIT) 2151 return false; 2152 if (!con_msg && !err) 2153 return true; 2154 return false; 2155 case TIPC_ESTABLISHED: 2156 if (!skb_queue_empty(&sk->sk_write_queue)) 2157 tipc_sk_push_backlog(tsk); 2158 /* Accept only connection-based messages sent by peer */ 2159 if (likely(con_msg && !err && pport == oport && pnode == onode)) 2160 return true; 2161 if (!tsk_peer_msg(tsk, hdr)) 2162 return false; 2163 if (!err) 2164 return true; 2165 tipc_set_sk_state(sk, TIPC_DISCONNECTING); 2166 tipc_node_remove_conn(net, pnode, tsk->portid); 2167 sk->sk_state_change(sk); 2168 return true; 2169 default: 2170 pr_err("Unknown sk_state %u\n", sk->sk_state); 2171 } 2172 /* Abort connection setup attempt */ 2173 tipc_set_sk_state(sk, TIPC_DISCONNECTING); 2174 sk->sk_err = ECONNREFUSED; 2175 sk->sk_state_change(sk); 2176 return true; 2177 } 2178 2179 /** 2180 * rcvbuf_limit - get proper overload limit of socket receive queue 2181 * @sk: socket 2182 * @skb: message 2183 * 2184 * For connection oriented messages, irrespective of importance, 2185 * default queue limit is 2 MB. 2186 * 2187 * For connectionless messages, queue limits are based on message 2188 * importance as follows: 2189 * 2190 * TIPC_LOW_IMPORTANCE (2 MB) 2191 * TIPC_MEDIUM_IMPORTANCE (4 MB) 2192 * TIPC_HIGH_IMPORTANCE (8 MB) 2193 * TIPC_CRITICAL_IMPORTANCE (16 MB) 2194 * 2195 * Returns overload limit according to corresponding message importance 2196 */ 2197 static unsigned int rcvbuf_limit(struct sock *sk, struct sk_buff *skb) 2198 { 2199 struct tipc_sock *tsk = tipc_sk(sk); 2200 struct tipc_msg *hdr = buf_msg(skb); 2201 2202 if (unlikely(msg_in_group(hdr))) 2203 return READ_ONCE(sk->sk_rcvbuf); 2204 2205 if (unlikely(!msg_connected(hdr))) 2206 return READ_ONCE(sk->sk_rcvbuf) << msg_importance(hdr); 2207 2208 if (likely(tsk->peer_caps & TIPC_BLOCK_FLOWCTL)) 2209 return READ_ONCE(sk->sk_rcvbuf); 2210 2211 return FLOWCTL_MSG_LIM; 2212 } 2213 2214 /** 2215 * tipc_sk_filter_rcv - validate incoming message 2216 * @sk: socket 2217 * @skb: pointer to message. 2218 * 2219 * Enqueues message on receive queue if acceptable; optionally handles 2220 * disconnect indication for a connected socket. 2221 * 2222 * Called with socket lock already taken 2223 * 2224 */ 2225 static void tipc_sk_filter_rcv(struct sock *sk, struct sk_buff *skb, 2226 struct sk_buff_head *xmitq) 2227 { 2228 bool sk_conn = !tipc_sk_type_connectionless(sk); 2229 struct tipc_sock *tsk = tipc_sk(sk); 2230 struct tipc_group *grp = tsk->group; 2231 struct tipc_msg *hdr = buf_msg(skb); 2232 struct net *net = sock_net(sk); 2233 struct sk_buff_head inputq; 2234 int mtyp = msg_type(hdr); 2235 int limit, err = TIPC_OK; 2236 2237 trace_tipc_sk_filter_rcv(sk, skb, TIPC_DUMP_ALL, " "); 2238 TIPC_SKB_CB(skb)->bytes_read = 0; 2239 __skb_queue_head_init(&inputq); 2240 __skb_queue_tail(&inputq, skb); 2241 2242 if (unlikely(!msg_isdata(hdr))) 2243 tipc_sk_proto_rcv(sk, &inputq, xmitq); 2244 2245 if (unlikely(grp)) 2246 tipc_group_filter_msg(grp, &inputq, xmitq); 2247 2248 if (unlikely(!grp) && mtyp == TIPC_MCAST_MSG) 2249 tipc_mcast_filter_msg(net, &tsk->mc_method.deferredq, &inputq); 2250 2251 /* Validate and add to receive buffer if there is space */ 2252 while ((skb = __skb_dequeue(&inputq))) { 2253 hdr = buf_msg(skb); 2254 limit = rcvbuf_limit(sk, skb); 2255 if ((sk_conn && !tipc_sk_filter_connect(tsk, skb)) || 2256 (!sk_conn && msg_connected(hdr)) || 2257 (!grp && msg_in_group(hdr))) 2258 err = TIPC_ERR_NO_PORT; 2259 else if (sk_rmem_alloc_get(sk) + skb->truesize >= limit) { 2260 trace_tipc_sk_dump(sk, skb, TIPC_DUMP_ALL, 2261 "err_overload2!"); 2262 atomic_inc(&sk->sk_drops); 2263 err = TIPC_ERR_OVERLOAD; 2264 } 2265 2266 if (unlikely(err)) { 2267 if (tipc_msg_reverse(tipc_own_addr(net), &skb, err)) { 2268 trace_tipc_sk_rej_msg(sk, skb, TIPC_DUMP_NONE, 2269 "@filter_rcv!"); 2270 __skb_queue_tail(xmitq, skb); 2271 } 2272 err = TIPC_OK; 2273 continue; 2274 } 2275 __skb_queue_tail(&sk->sk_receive_queue, skb); 2276 skb_set_owner_r(skb, sk); 2277 trace_tipc_sk_overlimit2(sk, skb, TIPC_DUMP_ALL, 2278 "rcvq >90% allocated!"); 2279 sk->sk_data_ready(sk); 2280 } 2281 } 2282 2283 /** 2284 * tipc_sk_backlog_rcv - handle incoming message from backlog queue 2285 * @sk: socket 2286 * @skb: message 2287 * 2288 * Caller must hold socket lock 2289 */ 2290 static int tipc_sk_backlog_rcv(struct sock *sk, struct sk_buff *skb) 2291 { 2292 unsigned int before = sk_rmem_alloc_get(sk); 2293 struct sk_buff_head xmitq; 2294 unsigned int added; 2295 2296 __skb_queue_head_init(&xmitq); 2297 2298 tipc_sk_filter_rcv(sk, skb, &xmitq); 2299 added = sk_rmem_alloc_get(sk) - before; 2300 atomic_add(added, &tipc_sk(sk)->dupl_rcvcnt); 2301 2302 /* Send pending response/rejected messages, if any */ 2303 tipc_node_distr_xmit(sock_net(sk), &xmitq); 2304 return 0; 2305 } 2306 2307 /** 2308 * tipc_sk_enqueue - extract all buffers with destination 'dport' from 2309 * inputq and try adding them to socket or backlog queue 2310 * @inputq: list of incoming buffers with potentially different destinations 2311 * @sk: socket where the buffers should be enqueued 2312 * @dport: port number for the socket 2313 * 2314 * Caller must hold socket lock 2315 */ 2316 static void tipc_sk_enqueue(struct sk_buff_head *inputq, struct sock *sk, 2317 u32 dport, struct sk_buff_head *xmitq) 2318 { 2319 unsigned long time_limit = jiffies + 2; 2320 struct sk_buff *skb; 2321 unsigned int lim; 2322 atomic_t *dcnt; 2323 u32 onode; 2324 2325 while (skb_queue_len(inputq)) { 2326 if (unlikely(time_after_eq(jiffies, time_limit))) 2327 return; 2328 2329 skb = tipc_skb_dequeue(inputq, dport); 2330 if (unlikely(!skb)) 2331 return; 2332 2333 /* Add message directly to receive queue if possible */ 2334 if (!sock_owned_by_user(sk)) { 2335 tipc_sk_filter_rcv(sk, skb, xmitq); 2336 continue; 2337 } 2338 2339 /* Try backlog, compensating for double-counted bytes */ 2340 dcnt = &tipc_sk(sk)->dupl_rcvcnt; 2341 if (!sk->sk_backlog.len) 2342 atomic_set(dcnt, 0); 2343 lim = rcvbuf_limit(sk, skb) + atomic_read(dcnt); 2344 if (likely(!sk_add_backlog(sk, skb, lim))) { 2345 trace_tipc_sk_overlimit1(sk, skb, TIPC_DUMP_ALL, 2346 "bklg & rcvq >90% allocated!"); 2347 continue; 2348 } 2349 2350 trace_tipc_sk_dump(sk, skb, TIPC_DUMP_ALL, "err_overload!"); 2351 /* Overload => reject message back to sender */ 2352 onode = tipc_own_addr(sock_net(sk)); 2353 atomic_inc(&sk->sk_drops); 2354 if (tipc_msg_reverse(onode, &skb, TIPC_ERR_OVERLOAD)) { 2355 trace_tipc_sk_rej_msg(sk, skb, TIPC_DUMP_ALL, 2356 "@sk_enqueue!"); 2357 __skb_queue_tail(xmitq, skb); 2358 } 2359 break; 2360 } 2361 } 2362 2363 /** 2364 * tipc_sk_rcv - handle a chain of incoming buffers 2365 * @inputq: buffer list containing the buffers 2366 * Consumes all buffers in list until inputq is empty 2367 * Note: may be called in multiple threads referring to the same queue 2368 */ 2369 void tipc_sk_rcv(struct net *net, struct sk_buff_head *inputq) 2370 { 2371 struct sk_buff_head xmitq; 2372 u32 dnode, dport = 0; 2373 int err; 2374 struct tipc_sock *tsk; 2375 struct sock *sk; 2376 struct sk_buff *skb; 2377 2378 __skb_queue_head_init(&xmitq); 2379 while (skb_queue_len(inputq)) { 2380 dport = tipc_skb_peek_port(inputq, dport); 2381 tsk = tipc_sk_lookup(net, dport); 2382 2383 if (likely(tsk)) { 2384 sk = &tsk->sk; 2385 if (likely(spin_trylock_bh(&sk->sk_lock.slock))) { 2386 tipc_sk_enqueue(inputq, sk, dport, &xmitq); 2387 spin_unlock_bh(&sk->sk_lock.slock); 2388 } 2389 /* Send pending response/rejected messages, if any */ 2390 tipc_node_distr_xmit(sock_net(sk), &xmitq); 2391 sock_put(sk); 2392 continue; 2393 } 2394 /* No destination socket => dequeue skb if still there */ 2395 skb = tipc_skb_dequeue(inputq, dport); 2396 if (!skb) 2397 return; 2398 2399 /* Try secondary lookup if unresolved named message */ 2400 err = TIPC_ERR_NO_PORT; 2401 if (tipc_msg_lookup_dest(net, skb, &err)) 2402 goto xmit; 2403 2404 /* Prepare for message rejection */ 2405 if (!tipc_msg_reverse(tipc_own_addr(net), &skb, err)) 2406 continue; 2407 2408 trace_tipc_sk_rej_msg(NULL, skb, TIPC_DUMP_NONE, "@sk_rcv!"); 2409 xmit: 2410 dnode = msg_destnode(buf_msg(skb)); 2411 tipc_node_xmit_skb(net, skb, dnode, dport); 2412 } 2413 } 2414 2415 static int tipc_wait_for_connect(struct socket *sock, long *timeo_p) 2416 { 2417 DEFINE_WAIT_FUNC(wait, woken_wake_function); 2418 struct sock *sk = sock->sk; 2419 int done; 2420 2421 do { 2422 int err = sock_error(sk); 2423 if (err) 2424 return err; 2425 if (!*timeo_p) 2426 return -ETIMEDOUT; 2427 if (signal_pending(current)) 2428 return sock_intr_errno(*timeo_p); 2429 2430 add_wait_queue(sk_sleep(sk), &wait); 2431 done = sk_wait_event(sk, timeo_p, 2432 sk->sk_state != TIPC_CONNECTING, &wait); 2433 remove_wait_queue(sk_sleep(sk), &wait); 2434 } while (!done); 2435 return 0; 2436 } 2437 2438 static bool tipc_sockaddr_is_sane(struct sockaddr_tipc *addr) 2439 { 2440 if (addr->family != AF_TIPC) 2441 return false; 2442 if (addr->addrtype == TIPC_SERVICE_RANGE) 2443 return (addr->addr.nameseq.lower <= addr->addr.nameseq.upper); 2444 return (addr->addrtype == TIPC_SERVICE_ADDR || 2445 addr->addrtype == TIPC_SOCKET_ADDR); 2446 } 2447 2448 /** 2449 * tipc_connect - establish a connection to another TIPC port 2450 * @sock: socket structure 2451 * @dest: socket address for destination port 2452 * @destlen: size of socket address data structure 2453 * @flags: file-related flags associated with socket 2454 * 2455 * Returns 0 on success, errno otherwise 2456 */ 2457 static int tipc_connect(struct socket *sock, struct sockaddr *dest, 2458 int destlen, int flags) 2459 { 2460 struct sock *sk = sock->sk; 2461 struct tipc_sock *tsk = tipc_sk(sk); 2462 struct sockaddr_tipc *dst = (struct sockaddr_tipc *)dest; 2463 struct msghdr m = {NULL,}; 2464 long timeout = (flags & O_NONBLOCK) ? 0 : tsk->conn_timeout; 2465 int previous; 2466 int res = 0; 2467 2468 if (destlen != sizeof(struct sockaddr_tipc)) 2469 return -EINVAL; 2470 2471 lock_sock(sk); 2472 2473 if (tsk->group) { 2474 res = -EINVAL; 2475 goto exit; 2476 } 2477 2478 if (dst->family == AF_UNSPEC) { 2479 memset(&tsk->peer, 0, sizeof(struct sockaddr_tipc)); 2480 if (!tipc_sk_type_connectionless(sk)) 2481 res = -EINVAL; 2482 goto exit; 2483 } 2484 if (!tipc_sockaddr_is_sane(dst)) { 2485 res = -EINVAL; 2486 goto exit; 2487 } 2488 /* DGRAM/RDM connect(), just save the destaddr */ 2489 if (tipc_sk_type_connectionless(sk)) { 2490 memcpy(&tsk->peer, dest, destlen); 2491 goto exit; 2492 } else if (dst->addrtype == TIPC_SERVICE_RANGE) { 2493 res = -EINVAL; 2494 goto exit; 2495 } 2496 2497 previous = sk->sk_state; 2498 2499 switch (sk->sk_state) { 2500 case TIPC_OPEN: 2501 /* Send a 'SYN-' to destination */ 2502 m.msg_name = dest; 2503 m.msg_namelen = destlen; 2504 2505 /* If connect is in non-blocking case, set MSG_DONTWAIT to 2506 * indicate send_msg() is never blocked. 2507 */ 2508 if (!timeout) 2509 m.msg_flags = MSG_DONTWAIT; 2510 2511 res = __tipc_sendmsg(sock, &m, 0); 2512 if ((res < 0) && (res != -EWOULDBLOCK)) 2513 goto exit; 2514 2515 /* Just entered TIPC_CONNECTING state; the only 2516 * difference is that return value in non-blocking 2517 * case is EINPROGRESS, rather than EALREADY. 2518 */ 2519 res = -EINPROGRESS; 2520 /* fall through */ 2521 case TIPC_CONNECTING: 2522 if (!timeout) { 2523 if (previous == TIPC_CONNECTING) 2524 res = -EALREADY; 2525 goto exit; 2526 } 2527 timeout = msecs_to_jiffies(timeout); 2528 /* Wait until an 'ACK' or 'RST' arrives, or a timeout occurs */ 2529 res = tipc_wait_for_connect(sock, &timeout); 2530 break; 2531 case TIPC_ESTABLISHED: 2532 res = -EISCONN; 2533 break; 2534 default: 2535 res = -EINVAL; 2536 } 2537 2538 exit: 2539 release_sock(sk); 2540 return res; 2541 } 2542 2543 /** 2544 * tipc_listen - allow socket to listen for incoming connections 2545 * @sock: socket structure 2546 * @len: (unused) 2547 * 2548 * Returns 0 on success, errno otherwise 2549 */ 2550 static int tipc_listen(struct socket *sock, int len) 2551 { 2552 struct sock *sk = sock->sk; 2553 int res; 2554 2555 lock_sock(sk); 2556 res = tipc_set_sk_state(sk, TIPC_LISTEN); 2557 release_sock(sk); 2558 2559 return res; 2560 } 2561 2562 static int tipc_wait_for_accept(struct socket *sock, long timeo) 2563 { 2564 struct sock *sk = sock->sk; 2565 DEFINE_WAIT(wait); 2566 int err; 2567 2568 /* True wake-one mechanism for incoming connections: only 2569 * one process gets woken up, not the 'whole herd'. 2570 * Since we do not 'race & poll' for established sockets 2571 * anymore, the common case will execute the loop only once. 2572 */ 2573 for (;;) { 2574 prepare_to_wait_exclusive(sk_sleep(sk), &wait, 2575 TASK_INTERRUPTIBLE); 2576 if (timeo && skb_queue_empty(&sk->sk_receive_queue)) { 2577 release_sock(sk); 2578 timeo = schedule_timeout(timeo); 2579 lock_sock(sk); 2580 } 2581 err = 0; 2582 if (!skb_queue_empty(&sk->sk_receive_queue)) 2583 break; 2584 err = -EAGAIN; 2585 if (!timeo) 2586 break; 2587 err = sock_intr_errno(timeo); 2588 if (signal_pending(current)) 2589 break; 2590 } 2591 finish_wait(sk_sleep(sk), &wait); 2592 return err; 2593 } 2594 2595 /** 2596 * tipc_accept - wait for connection request 2597 * @sock: listening socket 2598 * @newsock: new socket that is to be connected 2599 * @flags: file-related flags associated with socket 2600 * 2601 * Returns 0 on success, errno otherwise 2602 */ 2603 static int tipc_accept(struct socket *sock, struct socket *new_sock, int flags, 2604 bool kern) 2605 { 2606 struct sock *new_sk, *sk = sock->sk; 2607 struct sk_buff *buf; 2608 struct tipc_sock *new_tsock; 2609 struct tipc_msg *msg; 2610 long timeo; 2611 int res; 2612 2613 lock_sock(sk); 2614 2615 if (sk->sk_state != TIPC_LISTEN) { 2616 res = -EINVAL; 2617 goto exit; 2618 } 2619 timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); 2620 res = tipc_wait_for_accept(sock, timeo); 2621 if (res) 2622 goto exit; 2623 2624 buf = skb_peek(&sk->sk_receive_queue); 2625 2626 res = tipc_sk_create(sock_net(sock->sk), new_sock, 0, kern); 2627 if (res) 2628 goto exit; 2629 security_sk_clone(sock->sk, new_sock->sk); 2630 2631 new_sk = new_sock->sk; 2632 new_tsock = tipc_sk(new_sk); 2633 msg = buf_msg(buf); 2634 2635 /* we lock on new_sk; but lockdep sees the lock on sk */ 2636 lock_sock_nested(new_sk, SINGLE_DEPTH_NESTING); 2637 2638 /* 2639 * Reject any stray messages received by new socket 2640 * before the socket lock was taken (very, very unlikely) 2641 */ 2642 tsk_rej_rx_queue(new_sk); 2643 2644 /* Connect new socket to it's peer */ 2645 tipc_sk_finish_conn(new_tsock, msg_origport(msg), msg_orignode(msg)); 2646 2647 tsk_set_importance(new_tsock, msg_importance(msg)); 2648 if (msg_named(msg)) { 2649 new_tsock->conn_type = msg_nametype(msg); 2650 new_tsock->conn_instance = msg_nameinst(msg); 2651 } 2652 2653 /* 2654 * Respond to 'SYN-' by discarding it & returning 'ACK'-. 2655 * Respond to 'SYN+' by queuing it on new socket. 2656 */ 2657 if (!msg_data_sz(msg)) { 2658 struct msghdr m = {NULL,}; 2659 2660 tsk_advance_rx_queue(sk); 2661 __tipc_sendstream(new_sock, &m, 0); 2662 } else { 2663 __skb_dequeue(&sk->sk_receive_queue); 2664 __skb_queue_head(&new_sk->sk_receive_queue, buf); 2665 skb_set_owner_r(buf, new_sk); 2666 } 2667 release_sock(new_sk); 2668 exit: 2669 release_sock(sk); 2670 return res; 2671 } 2672 2673 /** 2674 * tipc_shutdown - shutdown socket connection 2675 * @sock: socket structure 2676 * @how: direction to close (must be SHUT_RDWR) 2677 * 2678 * Terminates connection (if necessary), then purges socket's receive queue. 2679 * 2680 * Returns 0 on success, errno otherwise 2681 */ 2682 static int tipc_shutdown(struct socket *sock, int how) 2683 { 2684 struct sock *sk = sock->sk; 2685 int res; 2686 2687 if (how != SHUT_RDWR) 2688 return -EINVAL; 2689 2690 lock_sock(sk); 2691 2692 trace_tipc_sk_shutdown(sk, NULL, TIPC_DUMP_ALL, " "); 2693 __tipc_shutdown(sock, TIPC_CONN_SHUTDOWN); 2694 sk->sk_shutdown = SEND_SHUTDOWN; 2695 2696 if (sk->sk_state == TIPC_DISCONNECTING) { 2697 /* Discard any unreceived messages */ 2698 __skb_queue_purge(&sk->sk_receive_queue); 2699 2700 /* Wake up anyone sleeping in poll */ 2701 sk->sk_state_change(sk); 2702 res = 0; 2703 } else { 2704 res = -ENOTCONN; 2705 } 2706 2707 release_sock(sk); 2708 return res; 2709 } 2710 2711 static void tipc_sk_check_probing_state(struct sock *sk, 2712 struct sk_buff_head *list) 2713 { 2714 struct tipc_sock *tsk = tipc_sk(sk); 2715 u32 pnode = tsk_peer_node(tsk); 2716 u32 pport = tsk_peer_port(tsk); 2717 u32 self = tsk_own_node(tsk); 2718 u32 oport = tsk->portid; 2719 struct sk_buff *skb; 2720 2721 if (tsk->probe_unacked) { 2722 tipc_set_sk_state(sk, TIPC_DISCONNECTING); 2723 sk->sk_err = ECONNABORTED; 2724 tipc_node_remove_conn(sock_net(sk), pnode, pport); 2725 sk->sk_state_change(sk); 2726 return; 2727 } 2728 /* Prepare new probe */ 2729 skb = tipc_msg_create(CONN_MANAGER, CONN_PROBE, INT_H_SIZE, 0, 2730 pnode, self, pport, oport, TIPC_OK); 2731 if (skb) 2732 __skb_queue_tail(list, skb); 2733 tsk->probe_unacked = true; 2734 sk_reset_timer(sk, &sk->sk_timer, jiffies + CONN_PROBING_INTV); 2735 } 2736 2737 static void tipc_sk_retry_connect(struct sock *sk, struct sk_buff_head *list) 2738 { 2739 struct tipc_sock *tsk = tipc_sk(sk); 2740 2741 /* Try again later if dest link is congested */ 2742 if (tsk->cong_link_cnt) { 2743 sk_reset_timer(sk, &sk->sk_timer, msecs_to_jiffies(100)); 2744 return; 2745 } 2746 /* Prepare SYN for retransmit */ 2747 tipc_msg_skb_clone(&sk->sk_write_queue, list); 2748 } 2749 2750 static void tipc_sk_timeout(struct timer_list *t) 2751 { 2752 struct sock *sk = from_timer(sk, t, sk_timer); 2753 struct tipc_sock *tsk = tipc_sk(sk); 2754 u32 pnode = tsk_peer_node(tsk); 2755 struct sk_buff_head list; 2756 int rc = 0; 2757 2758 __skb_queue_head_init(&list); 2759 bh_lock_sock(sk); 2760 2761 /* Try again later if socket is busy */ 2762 if (sock_owned_by_user(sk)) { 2763 sk_reset_timer(sk, &sk->sk_timer, jiffies + HZ / 20); 2764 bh_unlock_sock(sk); 2765 sock_put(sk); 2766 return; 2767 } 2768 2769 if (sk->sk_state == TIPC_ESTABLISHED) 2770 tipc_sk_check_probing_state(sk, &list); 2771 else if (sk->sk_state == TIPC_CONNECTING) 2772 tipc_sk_retry_connect(sk, &list); 2773 2774 bh_unlock_sock(sk); 2775 2776 if (!skb_queue_empty(&list)) 2777 rc = tipc_node_xmit(sock_net(sk), &list, pnode, tsk->portid); 2778 2779 /* SYN messages may cause link congestion */ 2780 if (rc == -ELINKCONG) { 2781 tipc_dest_push(&tsk->cong_links, pnode, 0); 2782 tsk->cong_link_cnt = 1; 2783 } 2784 sock_put(sk); 2785 } 2786 2787 static int tipc_sk_publish(struct tipc_sock *tsk, uint scope, 2788 struct tipc_name_seq const *seq) 2789 { 2790 struct sock *sk = &tsk->sk; 2791 struct net *net = sock_net(sk); 2792 struct publication *publ; 2793 u32 key; 2794 2795 if (scope != TIPC_NODE_SCOPE) 2796 scope = TIPC_CLUSTER_SCOPE; 2797 2798 if (tipc_sk_connected(sk)) 2799 return -EINVAL; 2800 key = tsk->portid + tsk->pub_count + 1; 2801 if (key == tsk->portid) 2802 return -EADDRINUSE; 2803 2804 publ = tipc_nametbl_publish(net, seq->type, seq->lower, seq->upper, 2805 scope, tsk->portid, key); 2806 if (unlikely(!publ)) 2807 return -EINVAL; 2808 2809 list_add(&publ->binding_sock, &tsk->publications); 2810 tsk->pub_count++; 2811 tsk->published = 1; 2812 return 0; 2813 } 2814 2815 static int tipc_sk_withdraw(struct tipc_sock *tsk, uint scope, 2816 struct tipc_name_seq const *seq) 2817 { 2818 struct net *net = sock_net(&tsk->sk); 2819 struct publication *publ; 2820 struct publication *safe; 2821 int rc = -EINVAL; 2822 2823 if (scope != TIPC_NODE_SCOPE) 2824 scope = TIPC_CLUSTER_SCOPE; 2825 2826 list_for_each_entry_safe(publ, safe, &tsk->publications, binding_sock) { 2827 if (seq) { 2828 if (publ->scope != scope) 2829 continue; 2830 if (publ->type != seq->type) 2831 continue; 2832 if (publ->lower != seq->lower) 2833 continue; 2834 if (publ->upper != seq->upper) 2835 break; 2836 tipc_nametbl_withdraw(net, publ->type, publ->lower, 2837 publ->upper, publ->key); 2838 rc = 0; 2839 break; 2840 } 2841 tipc_nametbl_withdraw(net, publ->type, publ->lower, 2842 publ->upper, publ->key); 2843 rc = 0; 2844 } 2845 if (list_empty(&tsk->publications)) 2846 tsk->published = 0; 2847 return rc; 2848 } 2849 2850 /* tipc_sk_reinit: set non-zero address in all existing sockets 2851 * when we go from standalone to network mode. 2852 */ 2853 void tipc_sk_reinit(struct net *net) 2854 { 2855 struct tipc_net *tn = net_generic(net, tipc_net_id); 2856 struct rhashtable_iter iter; 2857 struct tipc_sock *tsk; 2858 struct tipc_msg *msg; 2859 2860 rhashtable_walk_enter(&tn->sk_rht, &iter); 2861 2862 do { 2863 rhashtable_walk_start(&iter); 2864 2865 while ((tsk = rhashtable_walk_next(&iter)) && !IS_ERR(tsk)) { 2866 sock_hold(&tsk->sk); 2867 rhashtable_walk_stop(&iter); 2868 lock_sock(&tsk->sk); 2869 msg = &tsk->phdr; 2870 msg_set_prevnode(msg, tipc_own_addr(net)); 2871 msg_set_orignode(msg, tipc_own_addr(net)); 2872 release_sock(&tsk->sk); 2873 rhashtable_walk_start(&iter); 2874 sock_put(&tsk->sk); 2875 } 2876 2877 rhashtable_walk_stop(&iter); 2878 } while (tsk == ERR_PTR(-EAGAIN)); 2879 2880 rhashtable_walk_exit(&iter); 2881 } 2882 2883 static struct tipc_sock *tipc_sk_lookup(struct net *net, u32 portid) 2884 { 2885 struct tipc_net *tn = net_generic(net, tipc_net_id); 2886 struct tipc_sock *tsk; 2887 2888 rcu_read_lock(); 2889 tsk = rhashtable_lookup(&tn->sk_rht, &portid, tsk_rht_params); 2890 if (tsk) 2891 sock_hold(&tsk->sk); 2892 rcu_read_unlock(); 2893 2894 return tsk; 2895 } 2896 2897 static int tipc_sk_insert(struct tipc_sock *tsk) 2898 { 2899 struct sock *sk = &tsk->sk; 2900 struct net *net = sock_net(sk); 2901 struct tipc_net *tn = net_generic(net, tipc_net_id); 2902 u32 remaining = (TIPC_MAX_PORT - TIPC_MIN_PORT) + 1; 2903 u32 portid = prandom_u32() % remaining + TIPC_MIN_PORT; 2904 2905 while (remaining--) { 2906 portid++; 2907 if ((portid < TIPC_MIN_PORT) || (portid > TIPC_MAX_PORT)) 2908 portid = TIPC_MIN_PORT; 2909 tsk->portid = portid; 2910 sock_hold(&tsk->sk); 2911 if (!rhashtable_lookup_insert_fast(&tn->sk_rht, &tsk->node, 2912 tsk_rht_params)) 2913 return 0; 2914 sock_put(&tsk->sk); 2915 } 2916 2917 return -1; 2918 } 2919 2920 static void tipc_sk_remove(struct tipc_sock *tsk) 2921 { 2922 struct sock *sk = &tsk->sk; 2923 struct tipc_net *tn = net_generic(sock_net(sk), tipc_net_id); 2924 2925 if (!rhashtable_remove_fast(&tn->sk_rht, &tsk->node, tsk_rht_params)) { 2926 WARN_ON(refcount_read(&sk->sk_refcnt) == 1); 2927 __sock_put(sk); 2928 } 2929 } 2930 2931 static const struct rhashtable_params tsk_rht_params = { 2932 .nelem_hint = 192, 2933 .head_offset = offsetof(struct tipc_sock, node), 2934 .key_offset = offsetof(struct tipc_sock, portid), 2935 .key_len = sizeof(u32), /* portid */ 2936 .max_size = 1048576, 2937 .min_size = 256, 2938 .automatic_shrinking = true, 2939 }; 2940 2941 int tipc_sk_rht_init(struct net *net) 2942 { 2943 struct tipc_net *tn = net_generic(net, tipc_net_id); 2944 2945 return rhashtable_init(&tn->sk_rht, &tsk_rht_params); 2946 } 2947 2948 void tipc_sk_rht_destroy(struct net *net) 2949 { 2950 struct tipc_net *tn = net_generic(net, tipc_net_id); 2951 2952 /* Wait for socket readers to complete */ 2953 synchronize_net(); 2954 2955 rhashtable_destroy(&tn->sk_rht); 2956 } 2957 2958 static int tipc_sk_join(struct tipc_sock *tsk, struct tipc_group_req *mreq) 2959 { 2960 struct net *net = sock_net(&tsk->sk); 2961 struct tipc_group *grp = tsk->group; 2962 struct tipc_msg *hdr = &tsk->phdr; 2963 struct tipc_name_seq seq; 2964 int rc; 2965 2966 if (mreq->type < TIPC_RESERVED_TYPES) 2967 return -EACCES; 2968 if (mreq->scope > TIPC_NODE_SCOPE) 2969 return -EINVAL; 2970 if (grp) 2971 return -EACCES; 2972 grp = tipc_group_create(net, tsk->portid, mreq, &tsk->group_is_open); 2973 if (!grp) 2974 return -ENOMEM; 2975 tsk->group = grp; 2976 msg_set_lookup_scope(hdr, mreq->scope); 2977 msg_set_nametype(hdr, mreq->type); 2978 msg_set_dest_droppable(hdr, true); 2979 seq.type = mreq->type; 2980 seq.lower = mreq->instance; 2981 seq.upper = seq.lower; 2982 tipc_nametbl_build_group(net, grp, mreq->type, mreq->scope); 2983 rc = tipc_sk_publish(tsk, mreq->scope, &seq); 2984 if (rc) { 2985 tipc_group_delete(net, grp); 2986 tsk->group = NULL; 2987 return rc; 2988 } 2989 /* Eliminate any risk that a broadcast overtakes sent JOINs */ 2990 tsk->mc_method.rcast = true; 2991 tsk->mc_method.mandatory = true; 2992 tipc_group_join(net, grp, &tsk->sk.sk_rcvbuf); 2993 return rc; 2994 } 2995 2996 static int tipc_sk_leave(struct tipc_sock *tsk) 2997 { 2998 struct net *net = sock_net(&tsk->sk); 2999 struct tipc_group *grp = tsk->group; 3000 struct tipc_name_seq seq; 3001 int scope; 3002 3003 if (!grp) 3004 return -EINVAL; 3005 tipc_group_self(grp, &seq, &scope); 3006 tipc_group_delete(net, grp); 3007 tsk->group = NULL; 3008 tipc_sk_withdraw(tsk, scope, &seq); 3009 return 0; 3010 } 3011 3012 /** 3013 * tipc_setsockopt - set socket option 3014 * @sock: socket structure 3015 * @lvl: option level 3016 * @opt: option identifier 3017 * @ov: pointer to new option value 3018 * @ol: length of option value 3019 * 3020 * For stream sockets only, accepts and ignores all IPPROTO_TCP options 3021 * (to ease compatibility). 3022 * 3023 * Returns 0 on success, errno otherwise 3024 */ 3025 static int tipc_setsockopt(struct socket *sock, int lvl, int opt, 3026 char __user *ov, unsigned int ol) 3027 { 3028 struct sock *sk = sock->sk; 3029 struct tipc_sock *tsk = tipc_sk(sk); 3030 struct tipc_group_req mreq; 3031 u32 value = 0; 3032 int res = 0; 3033 3034 if ((lvl == IPPROTO_TCP) && (sock->type == SOCK_STREAM)) 3035 return 0; 3036 if (lvl != SOL_TIPC) 3037 return -ENOPROTOOPT; 3038 3039 switch (opt) { 3040 case TIPC_IMPORTANCE: 3041 case TIPC_SRC_DROPPABLE: 3042 case TIPC_DEST_DROPPABLE: 3043 case TIPC_CONN_TIMEOUT: 3044 case TIPC_NODELAY: 3045 if (ol < sizeof(value)) 3046 return -EINVAL; 3047 if (get_user(value, (u32 __user *)ov)) 3048 return -EFAULT; 3049 break; 3050 case TIPC_GROUP_JOIN: 3051 if (ol < sizeof(mreq)) 3052 return -EINVAL; 3053 if (copy_from_user(&mreq, ov, sizeof(mreq))) 3054 return -EFAULT; 3055 break; 3056 default: 3057 if (ov || ol) 3058 return -EINVAL; 3059 } 3060 3061 lock_sock(sk); 3062 3063 switch (opt) { 3064 case TIPC_IMPORTANCE: 3065 res = tsk_set_importance(tsk, value); 3066 break; 3067 case TIPC_SRC_DROPPABLE: 3068 if (sock->type != SOCK_STREAM) 3069 tsk_set_unreliable(tsk, value); 3070 else 3071 res = -ENOPROTOOPT; 3072 break; 3073 case TIPC_DEST_DROPPABLE: 3074 tsk_set_unreturnable(tsk, value); 3075 break; 3076 case TIPC_CONN_TIMEOUT: 3077 tipc_sk(sk)->conn_timeout = value; 3078 break; 3079 case TIPC_MCAST_BROADCAST: 3080 tsk->mc_method.rcast = false; 3081 tsk->mc_method.mandatory = true; 3082 break; 3083 case TIPC_MCAST_REPLICAST: 3084 tsk->mc_method.rcast = true; 3085 tsk->mc_method.mandatory = true; 3086 break; 3087 case TIPC_GROUP_JOIN: 3088 res = tipc_sk_join(tsk, &mreq); 3089 break; 3090 case TIPC_GROUP_LEAVE: 3091 res = tipc_sk_leave(tsk); 3092 break; 3093 case TIPC_NODELAY: 3094 tsk->nodelay = !!value; 3095 tsk_set_nagle(tsk); 3096 break; 3097 default: 3098 res = -EINVAL; 3099 } 3100 3101 release_sock(sk); 3102 3103 return res; 3104 } 3105 3106 /** 3107 * tipc_getsockopt - get socket option 3108 * @sock: socket structure 3109 * @lvl: option level 3110 * @opt: option identifier 3111 * @ov: receptacle for option value 3112 * @ol: receptacle for length of option value 3113 * 3114 * For stream sockets only, returns 0 length result for all IPPROTO_TCP options 3115 * (to ease compatibility). 3116 * 3117 * Returns 0 on success, errno otherwise 3118 */ 3119 static int tipc_getsockopt(struct socket *sock, int lvl, int opt, 3120 char __user *ov, int __user *ol) 3121 { 3122 struct sock *sk = sock->sk; 3123 struct tipc_sock *tsk = tipc_sk(sk); 3124 struct tipc_name_seq seq; 3125 int len, scope; 3126 u32 value; 3127 int res; 3128 3129 if ((lvl == IPPROTO_TCP) && (sock->type == SOCK_STREAM)) 3130 return put_user(0, ol); 3131 if (lvl != SOL_TIPC) 3132 return -ENOPROTOOPT; 3133 res = get_user(len, ol); 3134 if (res) 3135 return res; 3136 3137 lock_sock(sk); 3138 3139 switch (opt) { 3140 case TIPC_IMPORTANCE: 3141 value = tsk_importance(tsk); 3142 break; 3143 case TIPC_SRC_DROPPABLE: 3144 value = tsk_unreliable(tsk); 3145 break; 3146 case TIPC_DEST_DROPPABLE: 3147 value = tsk_unreturnable(tsk); 3148 break; 3149 case TIPC_CONN_TIMEOUT: 3150 value = tsk->conn_timeout; 3151 /* no need to set "res", since already 0 at this point */ 3152 break; 3153 case TIPC_NODE_RECVQ_DEPTH: 3154 value = 0; /* was tipc_queue_size, now obsolete */ 3155 break; 3156 case TIPC_SOCK_RECVQ_DEPTH: 3157 value = skb_queue_len(&sk->sk_receive_queue); 3158 break; 3159 case TIPC_SOCK_RECVQ_USED: 3160 value = sk_rmem_alloc_get(sk); 3161 break; 3162 case TIPC_GROUP_JOIN: 3163 seq.type = 0; 3164 if (tsk->group) 3165 tipc_group_self(tsk->group, &seq, &scope); 3166 value = seq.type; 3167 break; 3168 default: 3169 res = -EINVAL; 3170 } 3171 3172 release_sock(sk); 3173 3174 if (res) 3175 return res; /* "get" failed */ 3176 3177 if (len < sizeof(value)) 3178 return -EINVAL; 3179 3180 if (copy_to_user(ov, &value, sizeof(value))) 3181 return -EFAULT; 3182 3183 return put_user(sizeof(value), ol); 3184 } 3185 3186 static int tipc_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg) 3187 { 3188 struct net *net = sock_net(sock->sk); 3189 struct tipc_sioc_nodeid_req nr = {0}; 3190 struct tipc_sioc_ln_req lnr; 3191 void __user *argp = (void __user *)arg; 3192 3193 switch (cmd) { 3194 case SIOCGETLINKNAME: 3195 if (copy_from_user(&lnr, argp, sizeof(lnr))) 3196 return -EFAULT; 3197 if (!tipc_node_get_linkname(net, 3198 lnr.bearer_id & 0xffff, lnr.peer, 3199 lnr.linkname, TIPC_MAX_LINK_NAME)) { 3200 if (copy_to_user(argp, &lnr, sizeof(lnr))) 3201 return -EFAULT; 3202 return 0; 3203 } 3204 return -EADDRNOTAVAIL; 3205 case SIOCGETNODEID: 3206 if (copy_from_user(&nr, argp, sizeof(nr))) 3207 return -EFAULT; 3208 if (!tipc_node_get_id(net, nr.peer, nr.node_id)) 3209 return -EADDRNOTAVAIL; 3210 if (copy_to_user(argp, &nr, sizeof(nr))) 3211 return -EFAULT; 3212 return 0; 3213 default: 3214 return -ENOIOCTLCMD; 3215 } 3216 } 3217 3218 static int tipc_socketpair(struct socket *sock1, struct socket *sock2) 3219 { 3220 struct tipc_sock *tsk2 = tipc_sk(sock2->sk); 3221 struct tipc_sock *tsk1 = tipc_sk(sock1->sk); 3222 u32 onode = tipc_own_addr(sock_net(sock1->sk)); 3223 3224 tsk1->peer.family = AF_TIPC; 3225 tsk1->peer.addrtype = TIPC_ADDR_ID; 3226 tsk1->peer.scope = TIPC_NODE_SCOPE; 3227 tsk1->peer.addr.id.ref = tsk2->portid; 3228 tsk1->peer.addr.id.node = onode; 3229 tsk2->peer.family = AF_TIPC; 3230 tsk2->peer.addrtype = TIPC_ADDR_ID; 3231 tsk2->peer.scope = TIPC_NODE_SCOPE; 3232 tsk2->peer.addr.id.ref = tsk1->portid; 3233 tsk2->peer.addr.id.node = onode; 3234 3235 tipc_sk_finish_conn(tsk1, tsk2->portid, onode); 3236 tipc_sk_finish_conn(tsk2, tsk1->portid, onode); 3237 return 0; 3238 } 3239 3240 /* Protocol switches for the various types of TIPC sockets */ 3241 3242 static const struct proto_ops msg_ops = { 3243 .owner = THIS_MODULE, 3244 .family = AF_TIPC, 3245 .release = tipc_release, 3246 .bind = tipc_bind, 3247 .connect = tipc_connect, 3248 .socketpair = tipc_socketpair, 3249 .accept = sock_no_accept, 3250 .getname = tipc_getname, 3251 .poll = tipc_poll, 3252 .ioctl = tipc_ioctl, 3253 .listen = sock_no_listen, 3254 .shutdown = tipc_shutdown, 3255 .setsockopt = tipc_setsockopt, 3256 .getsockopt = tipc_getsockopt, 3257 .sendmsg = tipc_sendmsg, 3258 .recvmsg = tipc_recvmsg, 3259 .mmap = sock_no_mmap, 3260 .sendpage = sock_no_sendpage 3261 }; 3262 3263 static const struct proto_ops packet_ops = { 3264 .owner = THIS_MODULE, 3265 .family = AF_TIPC, 3266 .release = tipc_release, 3267 .bind = tipc_bind, 3268 .connect = tipc_connect, 3269 .socketpair = tipc_socketpair, 3270 .accept = tipc_accept, 3271 .getname = tipc_getname, 3272 .poll = tipc_poll, 3273 .ioctl = tipc_ioctl, 3274 .listen = tipc_listen, 3275 .shutdown = tipc_shutdown, 3276 .setsockopt = tipc_setsockopt, 3277 .getsockopt = tipc_getsockopt, 3278 .sendmsg = tipc_send_packet, 3279 .recvmsg = tipc_recvmsg, 3280 .mmap = sock_no_mmap, 3281 .sendpage = sock_no_sendpage 3282 }; 3283 3284 static const struct proto_ops stream_ops = { 3285 .owner = THIS_MODULE, 3286 .family = AF_TIPC, 3287 .release = tipc_release, 3288 .bind = tipc_bind, 3289 .connect = tipc_connect, 3290 .socketpair = tipc_socketpair, 3291 .accept = tipc_accept, 3292 .getname = tipc_getname, 3293 .poll = tipc_poll, 3294 .ioctl = tipc_ioctl, 3295 .listen = tipc_listen, 3296 .shutdown = tipc_shutdown, 3297 .setsockopt = tipc_setsockopt, 3298 .getsockopt = tipc_getsockopt, 3299 .sendmsg = tipc_sendstream, 3300 .recvmsg = tipc_recvstream, 3301 .mmap = sock_no_mmap, 3302 .sendpage = sock_no_sendpage 3303 }; 3304 3305 static const struct net_proto_family tipc_family_ops = { 3306 .owner = THIS_MODULE, 3307 .family = AF_TIPC, 3308 .create = tipc_sk_create 3309 }; 3310 3311 static struct proto tipc_proto = { 3312 .name = "TIPC", 3313 .owner = THIS_MODULE, 3314 .obj_size = sizeof(struct tipc_sock), 3315 .sysctl_rmem = sysctl_tipc_rmem 3316 }; 3317 3318 /** 3319 * tipc_socket_init - initialize TIPC socket interface 3320 * 3321 * Returns 0 on success, errno otherwise 3322 */ 3323 int tipc_socket_init(void) 3324 { 3325 int res; 3326 3327 res = proto_register(&tipc_proto, 1); 3328 if (res) { 3329 pr_err("Failed to register TIPC protocol type\n"); 3330 goto out; 3331 } 3332 3333 res = sock_register(&tipc_family_ops); 3334 if (res) { 3335 pr_err("Failed to register TIPC socket type\n"); 3336 proto_unregister(&tipc_proto); 3337 goto out; 3338 } 3339 out: 3340 return res; 3341 } 3342 3343 /** 3344 * tipc_socket_stop - stop TIPC socket interface 3345 */ 3346 void tipc_socket_stop(void) 3347 { 3348 sock_unregister(tipc_family_ops.family); 3349 proto_unregister(&tipc_proto); 3350 } 3351 3352 /* Caller should hold socket lock for the passed tipc socket. */ 3353 static int __tipc_nl_add_sk_con(struct sk_buff *skb, struct tipc_sock *tsk) 3354 { 3355 u32 peer_node; 3356 u32 peer_port; 3357 struct nlattr *nest; 3358 3359 peer_node = tsk_peer_node(tsk); 3360 peer_port = tsk_peer_port(tsk); 3361 3362 nest = nla_nest_start_noflag(skb, TIPC_NLA_SOCK_CON); 3363 if (!nest) 3364 return -EMSGSIZE; 3365 3366 if (nla_put_u32(skb, TIPC_NLA_CON_NODE, peer_node)) 3367 goto msg_full; 3368 if (nla_put_u32(skb, TIPC_NLA_CON_SOCK, peer_port)) 3369 goto msg_full; 3370 3371 if (tsk->conn_type != 0) { 3372 if (nla_put_flag(skb, TIPC_NLA_CON_FLAG)) 3373 goto msg_full; 3374 if (nla_put_u32(skb, TIPC_NLA_CON_TYPE, tsk->conn_type)) 3375 goto msg_full; 3376 if (nla_put_u32(skb, TIPC_NLA_CON_INST, tsk->conn_instance)) 3377 goto msg_full; 3378 } 3379 nla_nest_end(skb, nest); 3380 3381 return 0; 3382 3383 msg_full: 3384 nla_nest_cancel(skb, nest); 3385 3386 return -EMSGSIZE; 3387 } 3388 3389 static int __tipc_nl_add_sk_info(struct sk_buff *skb, struct tipc_sock 3390 *tsk) 3391 { 3392 struct net *net = sock_net(skb->sk); 3393 struct sock *sk = &tsk->sk; 3394 3395 if (nla_put_u32(skb, TIPC_NLA_SOCK_REF, tsk->portid) || 3396 nla_put_u32(skb, TIPC_NLA_SOCK_ADDR, tipc_own_addr(net))) 3397 return -EMSGSIZE; 3398 3399 if (tipc_sk_connected(sk)) { 3400 if (__tipc_nl_add_sk_con(skb, tsk)) 3401 return -EMSGSIZE; 3402 } else if (!list_empty(&tsk->publications)) { 3403 if (nla_put_flag(skb, TIPC_NLA_SOCK_HAS_PUBL)) 3404 return -EMSGSIZE; 3405 } 3406 return 0; 3407 } 3408 3409 /* Caller should hold socket lock for the passed tipc socket. */ 3410 static int __tipc_nl_add_sk(struct sk_buff *skb, struct netlink_callback *cb, 3411 struct tipc_sock *tsk) 3412 { 3413 struct nlattr *attrs; 3414 void *hdr; 3415 3416 hdr = genlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, 3417 &tipc_genl_family, NLM_F_MULTI, TIPC_NL_SOCK_GET); 3418 if (!hdr) 3419 goto msg_cancel; 3420 3421 attrs = nla_nest_start_noflag(skb, TIPC_NLA_SOCK); 3422 if (!attrs) 3423 goto genlmsg_cancel; 3424 3425 if (__tipc_nl_add_sk_info(skb, tsk)) 3426 goto attr_msg_cancel; 3427 3428 nla_nest_end(skb, attrs); 3429 genlmsg_end(skb, hdr); 3430 3431 return 0; 3432 3433 attr_msg_cancel: 3434 nla_nest_cancel(skb, attrs); 3435 genlmsg_cancel: 3436 genlmsg_cancel(skb, hdr); 3437 msg_cancel: 3438 return -EMSGSIZE; 3439 } 3440 3441 int tipc_nl_sk_walk(struct sk_buff *skb, struct netlink_callback *cb, 3442 int (*skb_handler)(struct sk_buff *skb, 3443 struct netlink_callback *cb, 3444 struct tipc_sock *tsk)) 3445 { 3446 struct rhashtable_iter *iter = (void *)cb->args[4]; 3447 struct tipc_sock *tsk; 3448 int err; 3449 3450 rhashtable_walk_start(iter); 3451 while ((tsk = rhashtable_walk_next(iter)) != NULL) { 3452 if (IS_ERR(tsk)) { 3453 err = PTR_ERR(tsk); 3454 if (err == -EAGAIN) { 3455 err = 0; 3456 continue; 3457 } 3458 break; 3459 } 3460 3461 sock_hold(&tsk->sk); 3462 rhashtable_walk_stop(iter); 3463 lock_sock(&tsk->sk); 3464 err = skb_handler(skb, cb, tsk); 3465 if (err) { 3466 release_sock(&tsk->sk); 3467 sock_put(&tsk->sk); 3468 goto out; 3469 } 3470 release_sock(&tsk->sk); 3471 rhashtable_walk_start(iter); 3472 sock_put(&tsk->sk); 3473 } 3474 rhashtable_walk_stop(iter); 3475 out: 3476 return skb->len; 3477 } 3478 EXPORT_SYMBOL(tipc_nl_sk_walk); 3479 3480 int tipc_dump_start(struct netlink_callback *cb) 3481 { 3482 return __tipc_dump_start(cb, sock_net(cb->skb->sk)); 3483 } 3484 EXPORT_SYMBOL(tipc_dump_start); 3485 3486 int __tipc_dump_start(struct netlink_callback *cb, struct net *net) 3487 { 3488 /* tipc_nl_name_table_dump() uses cb->args[0...3]. */ 3489 struct rhashtable_iter *iter = (void *)cb->args[4]; 3490 struct tipc_net *tn = tipc_net(net); 3491 3492 if (!iter) { 3493 iter = kmalloc(sizeof(*iter), GFP_KERNEL); 3494 if (!iter) 3495 return -ENOMEM; 3496 3497 cb->args[4] = (long)iter; 3498 } 3499 3500 rhashtable_walk_enter(&tn->sk_rht, iter); 3501 return 0; 3502 } 3503 3504 int tipc_dump_done(struct netlink_callback *cb) 3505 { 3506 struct rhashtable_iter *hti = (void *)cb->args[4]; 3507 3508 rhashtable_walk_exit(hti); 3509 kfree(hti); 3510 return 0; 3511 } 3512 EXPORT_SYMBOL(tipc_dump_done); 3513 3514 int tipc_sk_fill_sock_diag(struct sk_buff *skb, struct netlink_callback *cb, 3515 struct tipc_sock *tsk, u32 sk_filter_state, 3516 u64 (*tipc_diag_gen_cookie)(struct sock *sk)) 3517 { 3518 struct sock *sk = &tsk->sk; 3519 struct nlattr *attrs; 3520 struct nlattr *stat; 3521 3522 /*filter response w.r.t sk_state*/ 3523 if (!(sk_filter_state & (1 << sk->sk_state))) 3524 return 0; 3525 3526 attrs = nla_nest_start_noflag(skb, TIPC_NLA_SOCK); 3527 if (!attrs) 3528 goto msg_cancel; 3529 3530 if (__tipc_nl_add_sk_info(skb, tsk)) 3531 goto attr_msg_cancel; 3532 3533 if (nla_put_u32(skb, TIPC_NLA_SOCK_TYPE, (u32)sk->sk_type) || 3534 nla_put_u32(skb, TIPC_NLA_SOCK_TIPC_STATE, (u32)sk->sk_state) || 3535 nla_put_u32(skb, TIPC_NLA_SOCK_INO, sock_i_ino(sk)) || 3536 nla_put_u32(skb, TIPC_NLA_SOCK_UID, 3537 from_kuid_munged(sk_user_ns(NETLINK_CB(cb->skb).sk), 3538 sock_i_uid(sk))) || 3539 nla_put_u64_64bit(skb, TIPC_NLA_SOCK_COOKIE, 3540 tipc_diag_gen_cookie(sk), 3541 TIPC_NLA_SOCK_PAD)) 3542 goto attr_msg_cancel; 3543 3544 stat = nla_nest_start_noflag(skb, TIPC_NLA_SOCK_STAT); 3545 if (!stat) 3546 goto attr_msg_cancel; 3547 3548 if (nla_put_u32(skb, TIPC_NLA_SOCK_STAT_RCVQ, 3549 skb_queue_len(&sk->sk_receive_queue)) || 3550 nla_put_u32(skb, TIPC_NLA_SOCK_STAT_SENDQ, 3551 skb_queue_len(&sk->sk_write_queue)) || 3552 nla_put_u32(skb, TIPC_NLA_SOCK_STAT_DROP, 3553 atomic_read(&sk->sk_drops))) 3554 goto stat_msg_cancel; 3555 3556 if (tsk->cong_link_cnt && 3557 nla_put_flag(skb, TIPC_NLA_SOCK_STAT_LINK_CONG)) 3558 goto stat_msg_cancel; 3559 3560 if (tsk_conn_cong(tsk) && 3561 nla_put_flag(skb, TIPC_NLA_SOCK_STAT_CONN_CONG)) 3562 goto stat_msg_cancel; 3563 3564 nla_nest_end(skb, stat); 3565 3566 if (tsk->group) 3567 if (tipc_group_fill_sock_diag(tsk->group, skb)) 3568 goto stat_msg_cancel; 3569 3570 nla_nest_end(skb, attrs); 3571 3572 return 0; 3573 3574 stat_msg_cancel: 3575 nla_nest_cancel(skb, stat); 3576 attr_msg_cancel: 3577 nla_nest_cancel(skb, attrs); 3578 msg_cancel: 3579 return -EMSGSIZE; 3580 } 3581 EXPORT_SYMBOL(tipc_sk_fill_sock_diag); 3582 3583 int tipc_nl_sk_dump(struct sk_buff *skb, struct netlink_callback *cb) 3584 { 3585 return tipc_nl_sk_walk(skb, cb, __tipc_nl_add_sk); 3586 } 3587 3588 /* Caller should hold socket lock for the passed tipc socket. */ 3589 static int __tipc_nl_add_sk_publ(struct sk_buff *skb, 3590 struct netlink_callback *cb, 3591 struct publication *publ) 3592 { 3593 void *hdr; 3594 struct nlattr *attrs; 3595 3596 hdr = genlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, 3597 &tipc_genl_family, NLM_F_MULTI, TIPC_NL_PUBL_GET); 3598 if (!hdr) 3599 goto msg_cancel; 3600 3601 attrs = nla_nest_start_noflag(skb, TIPC_NLA_PUBL); 3602 if (!attrs) 3603 goto genlmsg_cancel; 3604 3605 if (nla_put_u32(skb, TIPC_NLA_PUBL_KEY, publ->key)) 3606 goto attr_msg_cancel; 3607 if (nla_put_u32(skb, TIPC_NLA_PUBL_TYPE, publ->type)) 3608 goto attr_msg_cancel; 3609 if (nla_put_u32(skb, TIPC_NLA_PUBL_LOWER, publ->lower)) 3610 goto attr_msg_cancel; 3611 if (nla_put_u32(skb, TIPC_NLA_PUBL_UPPER, publ->upper)) 3612 goto attr_msg_cancel; 3613 3614 nla_nest_end(skb, attrs); 3615 genlmsg_end(skb, hdr); 3616 3617 return 0; 3618 3619 attr_msg_cancel: 3620 nla_nest_cancel(skb, attrs); 3621 genlmsg_cancel: 3622 genlmsg_cancel(skb, hdr); 3623 msg_cancel: 3624 return -EMSGSIZE; 3625 } 3626 3627 /* Caller should hold socket lock for the passed tipc socket. */ 3628 static int __tipc_nl_list_sk_publ(struct sk_buff *skb, 3629 struct netlink_callback *cb, 3630 struct tipc_sock *tsk, u32 *last_publ) 3631 { 3632 int err; 3633 struct publication *p; 3634 3635 if (*last_publ) { 3636 list_for_each_entry(p, &tsk->publications, binding_sock) { 3637 if (p->key == *last_publ) 3638 break; 3639 } 3640 if (p->key != *last_publ) { 3641 /* We never set seq or call nl_dump_check_consistent() 3642 * this means that setting prev_seq here will cause the 3643 * consistence check to fail in the netlink callback 3644 * handler. Resulting in the last NLMSG_DONE message 3645 * having the NLM_F_DUMP_INTR flag set. 3646 */ 3647 cb->prev_seq = 1; 3648 *last_publ = 0; 3649 return -EPIPE; 3650 } 3651 } else { 3652 p = list_first_entry(&tsk->publications, struct publication, 3653 binding_sock); 3654 } 3655 3656 list_for_each_entry_from(p, &tsk->publications, binding_sock) { 3657 err = __tipc_nl_add_sk_publ(skb, cb, p); 3658 if (err) { 3659 *last_publ = p->key; 3660 return err; 3661 } 3662 } 3663 *last_publ = 0; 3664 3665 return 0; 3666 } 3667 3668 int tipc_nl_publ_dump(struct sk_buff *skb, struct netlink_callback *cb) 3669 { 3670 int err; 3671 u32 tsk_portid = cb->args[0]; 3672 u32 last_publ = cb->args[1]; 3673 u32 done = cb->args[2]; 3674 struct net *net = sock_net(skb->sk); 3675 struct tipc_sock *tsk; 3676 3677 if (!tsk_portid) { 3678 struct nlattr **attrs = genl_dumpit_info(cb)->attrs; 3679 struct nlattr *sock[TIPC_NLA_SOCK_MAX + 1]; 3680 3681 if (!attrs[TIPC_NLA_SOCK]) 3682 return -EINVAL; 3683 3684 err = nla_parse_nested_deprecated(sock, TIPC_NLA_SOCK_MAX, 3685 attrs[TIPC_NLA_SOCK], 3686 tipc_nl_sock_policy, NULL); 3687 if (err) 3688 return err; 3689 3690 if (!sock[TIPC_NLA_SOCK_REF]) 3691 return -EINVAL; 3692 3693 tsk_portid = nla_get_u32(sock[TIPC_NLA_SOCK_REF]); 3694 } 3695 3696 if (done) 3697 return 0; 3698 3699 tsk = tipc_sk_lookup(net, tsk_portid); 3700 if (!tsk) 3701 return -EINVAL; 3702 3703 lock_sock(&tsk->sk); 3704 err = __tipc_nl_list_sk_publ(skb, cb, tsk, &last_publ); 3705 if (!err) 3706 done = 1; 3707 release_sock(&tsk->sk); 3708 sock_put(&tsk->sk); 3709 3710 cb->args[0] = tsk_portid; 3711 cb->args[1] = last_publ; 3712 cb->args[2] = done; 3713 3714 return skb->len; 3715 } 3716 3717 /** 3718 * tipc_sk_filtering - check if a socket should be traced 3719 * @sk: the socket to be examined 3720 * @sysctl_tipc_sk_filter[]: the socket tuple for filtering, 3721 * (portid, sock type, name type, name lower, name upper) 3722 * 3723 * Returns true if the socket meets the socket tuple data 3724 * (value 0 = 'any') or when there is no tuple set (all = 0), 3725 * otherwise false 3726 */ 3727 bool tipc_sk_filtering(struct sock *sk) 3728 { 3729 struct tipc_sock *tsk; 3730 struct publication *p; 3731 u32 _port, _sktype, _type, _lower, _upper; 3732 u32 type = 0, lower = 0, upper = 0; 3733 3734 if (!sk) 3735 return true; 3736 3737 tsk = tipc_sk(sk); 3738 3739 _port = sysctl_tipc_sk_filter[0]; 3740 _sktype = sysctl_tipc_sk_filter[1]; 3741 _type = sysctl_tipc_sk_filter[2]; 3742 _lower = sysctl_tipc_sk_filter[3]; 3743 _upper = sysctl_tipc_sk_filter[4]; 3744 3745 if (!_port && !_sktype && !_type && !_lower && !_upper) 3746 return true; 3747 3748 if (_port) 3749 return (_port == tsk->portid); 3750 3751 if (_sktype && _sktype != sk->sk_type) 3752 return false; 3753 3754 if (tsk->published) { 3755 p = list_first_entry_or_null(&tsk->publications, 3756 struct publication, binding_sock); 3757 if (p) { 3758 type = p->type; 3759 lower = p->lower; 3760 upper = p->upper; 3761 } 3762 } 3763 3764 if (!tipc_sk_type_connectionless(sk)) { 3765 type = tsk->conn_type; 3766 lower = tsk->conn_instance; 3767 upper = tsk->conn_instance; 3768 } 3769 3770 if ((_type && _type != type) || (_lower && _lower != lower) || 3771 (_upper && _upper != upper)) 3772 return false; 3773 3774 return true; 3775 } 3776 3777 u32 tipc_sock_get_portid(struct sock *sk) 3778 { 3779 return (sk) ? (tipc_sk(sk))->portid : 0; 3780 } 3781 3782 /** 3783 * tipc_sk_overlimit1 - check if socket rx queue is about to be overloaded, 3784 * both the rcv and backlog queues are considered 3785 * @sk: tipc sk to be checked 3786 * @skb: tipc msg to be checked 3787 * 3788 * Returns true if the socket rx queue allocation is > 90%, otherwise false 3789 */ 3790 3791 bool tipc_sk_overlimit1(struct sock *sk, struct sk_buff *skb) 3792 { 3793 atomic_t *dcnt = &tipc_sk(sk)->dupl_rcvcnt; 3794 unsigned int lim = rcvbuf_limit(sk, skb) + atomic_read(dcnt); 3795 unsigned int qsize = sk->sk_backlog.len + sk_rmem_alloc_get(sk); 3796 3797 return (qsize > lim * 90 / 100); 3798 } 3799 3800 /** 3801 * tipc_sk_overlimit2 - check if socket rx queue is about to be overloaded, 3802 * only the rcv queue is considered 3803 * @sk: tipc sk to be checked 3804 * @skb: tipc msg to be checked 3805 * 3806 * Returns true if the socket rx queue allocation is > 90%, otherwise false 3807 */ 3808 3809 bool tipc_sk_overlimit2(struct sock *sk, struct sk_buff *skb) 3810 { 3811 unsigned int lim = rcvbuf_limit(sk, skb); 3812 unsigned int qsize = sk_rmem_alloc_get(sk); 3813 3814 return (qsize > lim * 90 / 100); 3815 } 3816 3817 /** 3818 * tipc_sk_dump - dump TIPC socket 3819 * @sk: tipc sk to be dumped 3820 * @dqueues: bitmask to decide if any socket queue to be dumped? 3821 * - TIPC_DUMP_NONE: don't dump socket queues 3822 * - TIPC_DUMP_SK_SNDQ: dump socket send queue 3823 * - TIPC_DUMP_SK_RCVQ: dump socket rcv queue 3824 * - TIPC_DUMP_SK_BKLGQ: dump socket backlog queue 3825 * - TIPC_DUMP_ALL: dump all the socket queues above 3826 * @buf: returned buffer of dump data in format 3827 */ 3828 int tipc_sk_dump(struct sock *sk, u16 dqueues, char *buf) 3829 { 3830 int i = 0; 3831 size_t sz = (dqueues) ? SK_LMAX : SK_LMIN; 3832 struct tipc_sock *tsk; 3833 struct publication *p; 3834 bool tsk_connected; 3835 3836 if (!sk) { 3837 i += scnprintf(buf, sz, "sk data: (null)\n"); 3838 return i; 3839 } 3840 3841 tsk = tipc_sk(sk); 3842 tsk_connected = !tipc_sk_type_connectionless(sk); 3843 3844 i += scnprintf(buf, sz, "sk data: %u", sk->sk_type); 3845 i += scnprintf(buf + i, sz - i, " %d", sk->sk_state); 3846 i += scnprintf(buf + i, sz - i, " %x", tsk_own_node(tsk)); 3847 i += scnprintf(buf + i, sz - i, " %u", tsk->portid); 3848 i += scnprintf(buf + i, sz - i, " | %u", tsk_connected); 3849 if (tsk_connected) { 3850 i += scnprintf(buf + i, sz - i, " %x", tsk_peer_node(tsk)); 3851 i += scnprintf(buf + i, sz - i, " %u", tsk_peer_port(tsk)); 3852 i += scnprintf(buf + i, sz - i, " %u", tsk->conn_type); 3853 i += scnprintf(buf + i, sz - i, " %u", tsk->conn_instance); 3854 } 3855 i += scnprintf(buf + i, sz - i, " | %u", tsk->published); 3856 if (tsk->published) { 3857 p = list_first_entry_or_null(&tsk->publications, 3858 struct publication, binding_sock); 3859 i += scnprintf(buf + i, sz - i, " %u", (p) ? p->type : 0); 3860 i += scnprintf(buf + i, sz - i, " %u", (p) ? p->lower : 0); 3861 i += scnprintf(buf + i, sz - i, " %u", (p) ? p->upper : 0); 3862 } 3863 i += scnprintf(buf + i, sz - i, " | %u", tsk->snd_win); 3864 i += scnprintf(buf + i, sz - i, " %u", tsk->rcv_win); 3865 i += scnprintf(buf + i, sz - i, " %u", tsk->max_pkt); 3866 i += scnprintf(buf + i, sz - i, " %x", tsk->peer_caps); 3867 i += scnprintf(buf + i, sz - i, " %u", tsk->cong_link_cnt); 3868 i += scnprintf(buf + i, sz - i, " %u", tsk->snt_unacked); 3869 i += scnprintf(buf + i, sz - i, " %u", tsk->rcv_unacked); 3870 i += scnprintf(buf + i, sz - i, " %u", atomic_read(&tsk->dupl_rcvcnt)); 3871 i += scnprintf(buf + i, sz - i, " %u", sk->sk_shutdown); 3872 i += scnprintf(buf + i, sz - i, " | %d", sk_wmem_alloc_get(sk)); 3873 i += scnprintf(buf + i, sz - i, " %d", sk->sk_sndbuf); 3874 i += scnprintf(buf + i, sz - i, " | %d", sk_rmem_alloc_get(sk)); 3875 i += scnprintf(buf + i, sz - i, " %d", sk->sk_rcvbuf); 3876 i += scnprintf(buf + i, sz - i, " | %d\n", READ_ONCE(sk->sk_backlog.len)); 3877 3878 if (dqueues & TIPC_DUMP_SK_SNDQ) { 3879 i += scnprintf(buf + i, sz - i, "sk_write_queue: "); 3880 i += tipc_list_dump(&sk->sk_write_queue, false, buf + i); 3881 } 3882 3883 if (dqueues & TIPC_DUMP_SK_RCVQ) { 3884 i += scnprintf(buf + i, sz - i, "sk_receive_queue: "); 3885 i += tipc_list_dump(&sk->sk_receive_queue, false, buf + i); 3886 } 3887 3888 if (dqueues & TIPC_DUMP_SK_BKLGQ) { 3889 i += scnprintf(buf + i, sz - i, "sk_backlog:\n head "); 3890 i += tipc_skb_dump(sk->sk_backlog.head, false, buf + i); 3891 if (sk->sk_backlog.tail != sk->sk_backlog.head) { 3892 i += scnprintf(buf + i, sz - i, " tail "); 3893 i += tipc_skb_dump(sk->sk_backlog.tail, false, 3894 buf + i); 3895 } 3896 } 3897 3898 return i; 3899 } 3900