net/vmw_vsock/af_vsock.c

   1 /*
   2  * VMware vSockets Driver
   3  *
   4  * Copyright (C) 2007-2013 VMware, Inc. All rights reserved.
   5  *
   6  * This program is free software; you can redistribute it and/or modify it
   7  * under the terms of the GNU General Public License as published by the Free
   8  * Software Foundation version 2 and no later version.
   9  *
  10  * This program is distributed in the hope that it will be useful, but WITHOUT
  11  * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
  12  * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  13  * more details.
  14  */
  15
  16 /* Implementation notes:
  17  *
  18  * - There are two kinds of sockets: those created by user action (such as
  19  * calling socket(2)) and those created by incoming connection request packets.
  20  *
  21  * - There are two "global" tables, one for bound sockets (sockets that have
  22  * specified an address that they are responsible for) and one for connected
  23  * sockets (sockets that have established a connection with another socket).
  24  * These tables are "global" in that all sockets on the system are placed
  25  * within them. - Note, though, that the bound table contains an extra entry
  26  * for a list of unbound sockets and SOCK_DGRAM sockets will always remain in
  27  * that list. The bound table is used solely for lookup of sockets when packets
  28  * are received and that's not necessary for SOCK_DGRAM sockets since we create
  29  * a datagram handle for each and need not perform a lookup.  Keeping SOCK_DGRAM
  30  * sockets out of the bound hash buckets will reduce the chance of collisions
  31  * when looking for SOCK_STREAM sockets and prevents us from having to check the
  32  * socket type in the hash table lookups.
  33  *
  34  * - Sockets created by user action will either be "client" sockets that
  35  * initiate a connection or "server" sockets that listen for connections; we do
  36  * not support simultaneous connects (two "client" sockets connecting).
  37  *
  38  * - "Server" sockets are referred to as listener sockets throughout this
  39  * implementation because they are in the VSOCK_SS_LISTEN state.  When a
  40  * connection request is received (the second kind of socket mentioned above),
  41  * we create a new socket and refer to it as a pending socket.  These pending
  42  * sockets are placed on the pending connection list of the listener socket.
  43  * When future packets are received for the address the listener socket is
  44  * bound to, we check if the source of the packet is from one that has an
  45  * existing pending connection.  If it does, we process the packet for the
  46  * pending socket.  When that socket reaches the connected state, it is removed
  47  * from the listener socket's pending list and enqueued in the listener
  48  * socket's accept queue.  Callers of accept(2) will accept connected sockets
  49  * from the listener socket's accept queue.  If the socket cannot be accepted
  50  * for some reason then it is marked rejected.  Once the connection is
  51  * accepted, it is owned by the user process and the responsibility for cleanup
  52  * falls with that user process.
  53  *
  54  * - It is possible that these pending sockets will never reach the connected
  55  * state; in fact, we may never receive another packet after the connection
  56  * request.  Because of this, we must schedule a cleanup function to run in the
  57  * future, after some amount of time passes where a connection should have been
  58  * established.  This function ensures that the socket is off all lists so it
  59  * cannot be retrieved, then drops all references to the socket so it is cleaned
  60  * up (sock_put() -> sk_free() -> our sk_destruct implementation).  Note this
  61  * function will also cleanup rejected sockets, those that reach the connected
  62  * state but leave it before they have been accepted.
  63  *
  64  * - Sockets created by user action will be cleaned up when the user process
  65  * calls close(2), causing our release implementation to be called. Our release
  66  * implementation will perform some cleanup then drop the last reference so our
  67  * sk_destruct implementation is invoked.  Our sk_destruct implementation will
  68  * perform additional cleanup that's common for both types of sockets.
  69  *
  70  * - A socket's reference count is what ensures that the structure won't be
  71  * freed.  Each entry in a list (such as the "global" bound and connected tables
  72  * and the listener socket's pending list and connected queue) ensures a
  73  * reference.  When we defer work until process context and pass a socket as our
  74  * argument, we must ensure the reference count is increased to ensure the
  75  * socket isn't freed before the function is run; the deferred function will
  76  * then drop the reference.
  77  */
  78
  79 #include <linux/types.h>
  80 #include <linux/bitops.h>
  81 #include <linux/cred.h>
  82 #include <linux/init.h>
  83 #include <linux/io.h>
  84 #include <linux/kernel.h>
  85 #include <linux/kmod.h>
  86 #include <linux/list.h>
  87 #include <linux/miscdevice.h>
  88 #include <linux/module.h>
  89 #include <linux/mutex.h>
  90 #include <linux/net.h>
  91 #include <linux/poll.h>
  92 #include <linux/random.h>
  93 #include <linux/skbuff.h>
  94 #include <linux/smp.h>
  95 #include <linux/socket.h>
  96 #include <linux/stddef.h>
  97 #include <linux/unistd.h>
  98 #include <linux/wait.h>
  99 #include <linux/workqueue.h>
 100 #include <net/sock.h>
 101 #include <net/af_vsock.h>
 102
 103 static int __vsock_bind(struct sock *sk, struct sockaddr_vm *addr);
 104 static void vsock_sk_destruct(struct sock *sk);
 105 static int vsock_queue_rcv_skb(struct sock *sk, struct sk_buff *skb);
 106
 107 /* Protocol family. */
 108 static struct proto vsock_proto = {
 109         .name = "AF_VSOCK",
 110         .owner = THIS_MODULE,
 111         .obj_size = sizeof(struct vsock_sock),
 112 };
 113
 114 /* The default peer timeout indicates how long we will wait for a peer response
 115  * to a control message.
 116  */
 117 #define VSOCK_DEFAULT_CONNECT_TIMEOUT (2 * HZ)
 118
 119 static const struct vsock_transport *transport;
 120 static DEFINE_MUTEX(vsock_register_mutex);
 121
 122 /**** EXPORTS ****/
 123
 124 /* Get the ID of the local context.  This is transport dependent. */
 125
 126 int vm_sockets_get_local_cid(void)
 127 {
 128         return transport->get_local_cid();
 129 }
 130 EXPORT_SYMBOL_GPL(vm_sockets_get_local_cid);
 131
 132 /**** UTILS ****/
 133
 134 /* Each bound VSocket is stored in the bind hash table and each connected
 135  * VSocket is stored in the connected hash table.
 136  *
 137  * Unbound sockets are all put on the same list attached to the end of the hash
 138  * table (vsock_unbound_sockets).  Bound sockets are added to the hash table in
 139  * the bucket that their local address hashes to (vsock_bound_sockets(addr)
 140  * represents the list that addr hashes to).
 141  *
 142  * Specifically, we initialize the vsock_bind_table array to a size of
 143  * VSOCK_HASH_SIZE + 1 so that vsock_bind_table[0] through
 144  * vsock_bind_table[VSOCK_HASH_SIZE - 1] are for bound sockets and
 145  * vsock_bind_table[VSOCK_HASH_SIZE] is for unbound sockets.  The hash function
 146  * mods with VSOCK_HASH_SIZE to ensure this.
 147  */
 148 #define VSOCK_HASH_SIZE         251
 149 #define MAX_PORT_RETRIES        24
 150
 151 #define VSOCK_HASH(addr)        ((addr)->svm_port % VSOCK_HASH_SIZE)
 152 #define vsock_bound_sockets(addr) (&vsock_bind_table[VSOCK_HASH(addr)])
 153 #define vsock_unbound_sockets     (&vsock_bind_table[VSOCK_HASH_SIZE])
 154
 155 /* XXX This can probably be implemented in a better way. */
 156 #define VSOCK_CONN_HASH(src, dst)                               \
 157         (((src)->svm_cid ^ (dst)->svm_port) % VSOCK_HASH_SIZE)
 158 #define vsock_connected_sockets(src, dst)               \
 159         (&vsock_connected_table[VSOCK_CONN_HASH(src, dst)])
 160 #define vsock_connected_sockets_vsk(vsk)                                \
 161         vsock_connected_sockets(&(vsk)->remote_addr, &(vsk)->local_addr)
 162
 163 static struct list_head vsock_bind_table[VSOCK_HASH_SIZE + 1];
 164 static struct list_head vsock_connected_table[VSOCK_HASH_SIZE];
 165 static DEFINE_SPINLOCK(vsock_table_lock);
 166
 167 /* Autobind this socket to the local address if necessary. */
 168 static int vsock_auto_bind(struct vsock_sock *vsk)
 169 {
 170         struct sock *sk = sk_vsock(vsk);
 171         struct sockaddr_vm local_addr;
 172
 173         if (vsock_addr_bound(&vsk->local_addr))
 174                 return 0;
 175         vsock_addr_init(&local_addr, VMADDR_CID_ANY, VMADDR_PORT_ANY);
 176         return __vsock_bind(sk, &local_addr);
 177 }
 178
 179 static void vsock_init_tables(void)
 180 {
 181         int i;
 182
 183         for (i = 0; i < ARRAY_SIZE(vsock_bind_table); i++)
 184                 INIT_LIST_HEAD(&vsock_bind_table[i]);
 185
 186         for (i = 0; i < ARRAY_SIZE(vsock_connected_table); i++)
 187                 INIT_LIST_HEAD(&vsock_connected_table[i]);
 188 }
 189
 190 static void __vsock_insert_bound(struct list_head *list,
 191                                  struct vsock_sock *vsk)
 192 {
 193         sock_hold(&vsk->sk);
 194         list_add(&vsk->bound_table, list);
 195 }
 196
 197 static void __vsock_insert_connected(struct list_head *list,
 198                                      struct vsock_sock *vsk)
 199 {
 200         sock_hold(&vsk->sk);
 201         list_add(&vsk->connected_table, list);
 202 }
 203
 204 static void __vsock_remove_bound(struct vsock_sock *vsk)
 205 {
 206         list_del_init(&vsk->bound_table);
 207         sock_put(&vsk->sk);
 208 }
 209
 210 static void __vsock_remove_connected(struct vsock_sock *vsk)
 211 {
 212         list_del_init(&vsk->connected_table);
 213         sock_put(&vsk->sk);
 214 }
 215
 216 static struct sock *__vsock_find_bound_socket(struct sockaddr_vm *addr)
 217 {
 218         struct vsock_sock *vsk;
 219
 220         list_for_each_entry(vsk, vsock_bound_sockets(addr), bound_table)
 221                 if (addr->svm_port == vsk->local_addr.svm_port)
 222                         return sk_vsock(vsk);
 223
 224         return NULL;
 225 }
 226
 227 static struct sock *__vsock_find_connected_socket(struct sockaddr_vm *src,
 228                                                   struct sockaddr_vm *dst)
 229 {
 230         struct vsock_sock *vsk;
 231
 232         list_for_each_entry(vsk, vsock_connected_sockets(src, dst),
 233                             connected_table) {
 234                 if (vsock_addr_equals_addr(src, &vsk->remote_addr) &&
 235                     dst->svm_port == vsk->local_addr.svm_port) {
 236                         return sk_vsock(vsk);
 237                 }
 238         }
 239
 240         return NULL;
 241 }
 242
 243 static bool __vsock_in_bound_table(struct vsock_sock *vsk)
 244 {
 245         return !list_empty(&vsk->bound_table);
 246 }
 247
 248 static bool __vsock_in_connected_table(struct vsock_sock *vsk)
 249 {
 250         return !list_empty(&vsk->connected_table);
 251 }
 252
 253 static void vsock_insert_unbound(struct vsock_sock *vsk)
 254 {
 255         spin_lock_bh(&vsock_table_lock);
 256         __vsock_insert_bound(vsock_unbound_sockets, vsk);
 257         spin_unlock_bh(&vsock_table_lock);
 258 }
 259
 260 void vsock_insert_connected(struct vsock_sock *vsk)
 261 {
 262         struct list_head *list = vsock_connected_sockets(
 263                 &vsk->remote_addr, &vsk->local_addr);
 264
 265         spin_lock_bh(&vsock_table_lock);
 266         __vsock_insert_connected(list, vsk);
 267         spin_unlock_bh(&vsock_table_lock);
 268 }
 269 EXPORT_SYMBOL_GPL(vsock_insert_connected);
 270
 271 void vsock_remove_bound(struct vsock_sock *vsk)
 272 {
 273         spin_lock_bh(&vsock_table_lock);
 274         __vsock_remove_bound(vsk);
 275         spin_unlock_bh(&vsock_table_lock);
 276 }
 277 EXPORT_SYMBOL_GPL(vsock_remove_bound);
 278
 279 void vsock_remove_connected(struct vsock_sock *vsk)
 280 {
 281         spin_lock_bh(&vsock_table_lock);
 282         __vsock_remove_connected(vsk);
 283         spin_unlock_bh(&vsock_table_lock);
 284 }
 285 EXPORT_SYMBOL_GPL(vsock_remove_connected);
 286
 287 struct sock *vsock_find_bound_socket(struct sockaddr_vm *addr)
 288 {
 289         struct sock *sk;
 290
 291         spin_lock_bh(&vsock_table_lock);
 292         sk = __vsock_find_bound_socket(addr);
 293         if (sk)
 294                 sock_hold(sk);
 295
 296         spin_unlock_bh(&vsock_table_lock);
 297
 298         return sk;
 299 }
 300 EXPORT_SYMBOL_GPL(vsock_find_bound_socket);
 301
 302 struct sock *vsock_find_connected_socket(struct sockaddr_vm *src,
 303                                          struct sockaddr_vm *dst)
 304 {
 305         struct sock *sk;
 306
 307         spin_lock_bh(&vsock_table_lock);
 308         sk = __vsock_find_connected_socket(src, dst);
 309         if (sk)
 310                 sock_hold(sk);
 311
 312         spin_unlock_bh(&vsock_table_lock);
 313
 314         return sk;
 315 }
 316 EXPORT_SYMBOL_GPL(vsock_find_connected_socket);
 317
 318 static bool vsock_in_bound_table(struct vsock_sock *vsk)
 319 {
 320         bool ret;
 321
 322         spin_lock_bh(&vsock_table_lock);
 323         ret = __vsock_in_bound_table(vsk);
 324         spin_unlock_bh(&vsock_table_lock);
 325
 326         return ret;
 327 }
 328
 329 static bool vsock_in_connected_table(struct vsock_sock *vsk)
 330 {
 331         bool ret;
 332
 333         spin_lock_bh(&vsock_table_lock);
 334         ret = __vsock_in_connected_table(vsk);
 335         spin_unlock_bh(&vsock_table_lock);
 336
 337         return ret;
 338 }
 339
 340 void vsock_for_each_connected_socket(void (*fn)(struct sock *sk))
 341 {
 342         int i;
 343
 344         spin_lock_bh(&vsock_table_lock);
 345
 346         for (i = 0; i < ARRAY_SIZE(vsock_connected_table); i++) {
 347                 struct vsock_sock *vsk;
 348                 list_for_each_entry(vsk, &vsock_connected_table[i],
 349                                     connected_table)
 350                         fn(sk_vsock(vsk));
 351         }
 352
 353         spin_unlock_bh(&vsock_table_lock);
 354 }
 355 EXPORT_SYMBOL_GPL(vsock_for_each_connected_socket);
 356
 357 void vsock_add_pending(struct sock *listener, struct sock *pending)
 358 {
 359         struct vsock_sock *vlistener;
 360         struct vsock_sock *vpending;
 361
 362         vlistener = vsock_sk(listener);
 363         vpending = vsock_sk(pending);
 364
 365         sock_hold(pending);
 366         sock_hold(listener);
 367         list_add_tail(&vpending->pending_links, &vlistener->pending_links);
 368 }
 369 EXPORT_SYMBOL_GPL(vsock_add_pending);
 370
 371 void vsock_remove_pending(struct sock *listener, struct sock *pending)
 372 {
 373         struct vsock_sock *vpending = vsock_sk(pending);
 374
 375         list_del_init(&vpending->pending_links);
 376         sock_put(listener);
 377         sock_put(pending);
 378 }
 379 EXPORT_SYMBOL_GPL(vsock_remove_pending);
 380
 381 void vsock_enqueue_accept(struct sock *listener, struct sock *connected)
 382 {
 383         struct vsock_sock *vlistener;
 384         struct vsock_sock *vconnected;
 385
 386         vlistener = vsock_sk(listener);
 387         vconnected = vsock_sk(connected);
 388
 389         sock_hold(connected);
 390         sock_hold(listener);
 391         list_add_tail(&vconnected->accept_queue, &vlistener->accept_queue);
 392 }
 393 EXPORT_SYMBOL_GPL(vsock_enqueue_accept);
 394
 395 static struct sock *vsock_dequeue_accept(struct sock *listener)
 396 {
 397         struct vsock_sock *vlistener;
 398         struct vsock_sock *vconnected;
 399
 400         vlistener = vsock_sk(listener);
 401
 402         if (list_empty(&vlistener->accept_queue))
 403                 return NULL;
 404
 405         vconnected = list_entry(vlistener->accept_queue.next,
 406                                 struct vsock_sock, accept_queue);
 407
 408         list_del_init(&vconnected->accept_queue);
 409         sock_put(listener);
 410         /* The caller will need a reference on the connected socket so we let
 411          * it call sock_put().
 412          */
 413
 414         return sk_vsock(vconnected);
 415 }
 416
 417 static bool vsock_is_accept_queue_empty(struct sock *sk)
 418 {
 419         struct vsock_sock *vsk = vsock_sk(sk);
 420         return list_empty(&vsk->accept_queue);
 421 }
 422
 423 static bool vsock_is_pending(struct sock *sk)
 424 {
 425         struct vsock_sock *vsk = vsock_sk(sk);
 426         return !list_empty(&vsk->pending_links);
 427 }
 428
 429 static int vsock_send_shutdown(struct sock *sk, int mode)
 430 {
 431         return transport->shutdown(vsock_sk(sk), mode);
 432 }
 433
 434 static void vsock_pending_work(struct work_struct *work)
 435 {
 436         struct sock *sk;
 437         struct sock *listener;
 438         struct vsock_sock *vsk;
 439         bool cleanup;
 440
 441         vsk = container_of(work, struct vsock_sock, pending_work.work);
 442         sk = sk_vsock(vsk);
 443         listener = vsk->listener;
 444         cleanup = true;
 445
 446         lock_sock(listener);
 447         lock_sock(sk);
 448
 449         if (vsock_is_pending(sk)) {
 450                 vsock_remove_pending(listener, sk);
 451         } else if (!vsk->rejected) {
 452                 /* We are not on the pending list and accept() did not reject
 453                  * us, so we must have been accepted by our user process.  We
 454                  * just need to drop our references to the sockets and be on
 455                  * our way.
 456                  */
 457                 cleanup = false;
 458                 goto out;
 459         }
 460
 461         listener->sk_ack_backlog--;
 462
 463         /* We need to remove ourself from the global connected sockets list so
 464          * incoming packets can't find this socket, and to reduce the reference
 465          * count.
 466          */
 467         if (vsock_in_connected_table(vsk))
 468                 vsock_remove_connected(vsk);
 469
 470         sk->sk_state = SS_FREE;
 471
 472 out:
 473         release_sock(sk);
 474         release_sock(listener);
 475         if (cleanup)
 476                 sock_put(sk);
 477
 478         sock_put(sk);
 479         sock_put(listener);
 480 }
 481
 482 /**** SOCKET OPERATIONS ****/
 483
 484 static int __vsock_bind_stream(struct vsock_sock *vsk,
 485                                struct sockaddr_vm *addr)
 486 {
 487         static u32 port = 0;
 488         struct sockaddr_vm new_addr;
 489
 490         if (!port)
 491                 port = LAST_RESERVED_PORT + 1 +
 492                         prandom_u32_max(U32_MAX - LAST_RESERVED_PORT);
 493
 494         vsock_addr_init(&new_addr, addr->svm_cid, addr->svm_port);
 495
 496         if (addr->svm_port == VMADDR_PORT_ANY) {
 497                 bool found = false;
 498                 unsigned int i;
 499
 500                 for (i = 0; i < MAX_PORT_RETRIES; i++) {
 501                         if (port <= LAST_RESERVED_PORT)
 502                                 port = LAST_RESERVED_PORT + 1;
 503
 504                         new_addr.svm_port = port++;
 505
 506                         if (!__vsock_find_bound_socket(&new_addr)) {
 507                                 found = true;
 508                                 break;
 509                         }
 510                 }
 511
 512                 if (!found)
 513                         return -EADDRNOTAVAIL;
 514         } else {
 515                 /* If port is in reserved range, ensure caller
 516                  * has necessary privileges.
 517                  */
 518                 if (addr->svm_port <= LAST_RESERVED_PORT &&
 519                     !capable(CAP_NET_BIND_SERVICE)) {
 520                         return -EACCES;
 521                 }
 522
 523                 if (__vsock_find_bound_socket(&new_addr))
 524                         return -EADDRINUSE;
 525         }
 526
 527         vsock_addr_init(&vsk->local_addr, new_addr.svm_cid, new_addr.svm_port);
 528
 529         /* Remove stream sockets from the unbound list and add them to the hash
 530          * table for easy lookup by its address.  The unbound list is simply an
 531          * extra entry at the end of the hash table, a trick used by AF_UNIX.
 532          */
 533         __vsock_remove_bound(vsk);
 534         __vsock_insert_bound(vsock_bound_sockets(&vsk->local_addr), vsk);
 535
 536         return 0;
 537 }
 538
 539 static int __vsock_bind_dgram(struct vsock_sock *vsk,
 540                               struct sockaddr_vm *addr)
 541 {
 542         return transport->dgram_bind(vsk, addr);
 543 }
 544
 545 static int __vsock_bind(struct sock *sk, struct sockaddr_vm *addr)
 546 {
 547         struct vsock_sock *vsk = vsock_sk(sk);
 548         u32 cid;
 549         int retval;
 550
 551         /* First ensure this socket isn't already bound. */
 552         if (vsock_addr_bound(&vsk->local_addr))
 553                 return -EINVAL;
 554
 555         /* Now bind to the provided address or select appropriate values if
 556          * none are provided (VMADDR_CID_ANY and VMADDR_PORT_ANY).  Note that
 557          * like AF_INET prevents binding to a non-local IP address (in most
 558          * cases), we only allow binding to the local CID.
 559          */
 560         cid = transport->get_local_cid();
 561         if (addr->svm_cid != cid && addr->svm_cid != VMADDR_CID_ANY)
 562                 return -EADDRNOTAVAIL;
 563
 564         switch (sk->sk_socket->type) {
 565         case SOCK_STREAM:
 566                 spin_lock_bh(&vsock_table_lock);
 567                 retval = __vsock_bind_stream(vsk, addr);
 568                 spin_unlock_bh(&vsock_table_lock);
 569                 break;
 570
 571         case SOCK_DGRAM:
 572                 retval = __vsock_bind_dgram(vsk, addr);
 573                 break;
 574
 575         default:
 576                 retval = -EINVAL;
 577                 break;
 578         }
 579
 580         return retval;
 581 }
 582
 583 static void vsock_connect_timeout(struct work_struct *work);
 584
 585 struct sock *__vsock_create(struct net *net,
 586                             struct socket *sock,
 587                             struct sock *parent,
 588                             gfp_t priority,
 589                             unsigned short type,
 590                             int kern)
 591 {
 592         struct sock *sk;
 593         struct vsock_sock *psk;
 594         struct vsock_sock *vsk;
 595
 596         sk = sk_alloc(net, AF_VSOCK, priority, &vsock_proto, kern);
 597         if (!sk)
 598                 return NULL;
 599
 600         sock_init_data(sock, sk);
 601
 602         /* sk->sk_type is normally set in sock_init_data, but only if sock is
 603          * non-NULL. We make sure that our sockets always have a type by
 604          * setting it here if needed.
 605          */
 606         if (!sock)
 607                 sk->sk_type = type;
 608
 609         vsk = vsock_sk(sk);
 610         vsock_addr_init(&vsk->local_addr, VMADDR_CID_ANY, VMADDR_PORT_ANY);
 611         vsock_addr_init(&vsk->remote_addr, VMADDR_CID_ANY, VMADDR_PORT_ANY);
 612
 613         sk->sk_destruct = vsock_sk_destruct;
 614         sk->sk_backlog_rcv = vsock_queue_rcv_skb;
 615         sk->sk_state = 0;
 616         sock_reset_flag(sk, SOCK_DONE);
 617
 618         INIT_LIST_HEAD(&vsk->bound_table);
 619         INIT_LIST_HEAD(&vsk->connected_table);
 620         vsk->listener = NULL;
 621         INIT_LIST_HEAD(&vsk->pending_links);
 622         INIT_LIST_HEAD(&vsk->accept_queue);
 623         vsk->rejected = false;
 624         vsk->sent_request = false;
 625         vsk->ignore_connecting_rst = false;
 626         vsk->peer_shutdown = 0;
 627         INIT_DELAYED_WORK(&vsk->connect_work, vsock_connect_timeout);
 628         INIT_DELAYED_WORK(&vsk->pending_work, vsock_pending_work);
 629
 630         psk = parent ? vsock_sk(parent) : NULL;
 631         if (parent) {
 632                 vsk->trusted = psk->trusted;
 633                 vsk->owner = get_cred(psk->owner);
 634                 vsk->connect_timeout = psk->connect_timeout;
 635                 security_sk_clone(parent, sk);
 636         } else {
 637                 vsk->trusted = ns_capable_noaudit(&init_user_ns, CAP_NET_ADMIN);
 638                 vsk->owner = get_current_cred();
 639                 vsk->connect_timeout = VSOCK_DEFAULT_CONNECT_TIMEOUT;
 640         }
 641
 642         if (transport->init(vsk, psk) < 0) {
 643                 sk_free(sk);
 644                 return NULL;
 645         }
 646
 647         if (sock)
 648                 vsock_insert_unbound(vsk);
 649
 650         return sk;
 651 }
 652 EXPORT_SYMBOL_GPL(__vsock_create);
 653
 654 static void __vsock_release(struct sock *sk)
 655 {
 656         if (sk) {
 657                 struct sk_buff *skb;
 658                 struct sock *pending;
 659                 struct vsock_sock *vsk;
 660
 661                 vsk = vsock_sk(sk);
 662                 pending = NULL; /* Compiler warning. */
 663
 664                 if (vsock_in_bound_table(vsk))
 665                         vsock_remove_bound(vsk);
 666
 667                 if (vsock_in_connected_table(vsk))
 668                         vsock_remove_connected(vsk);
 669
 670                 transport->release(vsk);
 671
 672                 lock_sock(sk);
 673                 sock_orphan(sk);
 674                 sk->sk_shutdown = SHUTDOWN_MASK;
 675
 676                 while ((skb = skb_dequeue(&sk->sk_receive_queue)))
 677                         kfree_skb(skb);
 678
 679                 /* Clean up any sockets that never were accepted. */
 680                 while ((pending = vsock_dequeue_accept(sk)) != NULL) {
 681                         __vsock_release(pending);
 682                         sock_put(pending);
 683                 }
 684
 685                 release_sock(sk);
 686                 sock_put(sk);
 687         }
 688 }
 689
 690 static void vsock_sk_destruct(struct sock *sk)
 691 {
 692         struct vsock_sock *vsk = vsock_sk(sk);
 693
 694         transport->destruct(vsk);
 695
 696         /* When clearing these addresses, there's no need to set the family and
 697          * possibly register the address family with the kernel.
 698          */
 699         vsock_addr_init(&vsk->local_addr, VMADDR_CID_ANY, VMADDR_PORT_ANY);
 700         vsock_addr_init(&vsk->remote_addr, VMADDR_CID_ANY, VMADDR_PORT_ANY);
 701
 702         put_cred(vsk->owner);
 703 }
 704
 705 static int vsock_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
 706 {
 707         int err;
 708
 709         err = sock_queue_rcv_skb(sk, skb);
 710         if (err)
 711                 kfree_skb(skb);
 712
 713         return err;
 714 }
 715
 716 s64 vsock_stream_has_data(struct vsock_sock *vsk)
 717 {
 718         return transport->stream_has_data(vsk);
 719 }
 720 EXPORT_SYMBOL_GPL(vsock_stream_has_data);
 721
 722 s64 vsock_stream_has_space(struct vsock_sock *vsk)
 723 {
 724         return transport->stream_has_space(vsk);
 725 }
 726 EXPORT_SYMBOL_GPL(vsock_stream_has_space);
 727
 728 static int vsock_release(struct socket *sock)
 729 {
 730         __vsock_release(sock->sk);
 731         sock->sk = NULL;
 732         sock->state = SS_FREE;
 733
 734         return 0;
 735 }
 736
 737 static int
 738 vsock_bind(struct socket *sock, struct sockaddr *addr, int addr_len)
 739 {
 740         int err;
 741         struct sock *sk;
 742         struct sockaddr_vm *vm_addr;
 743
 744         sk = sock->sk;
 745
 746         if (vsock_addr_cast(addr, addr_len, &vm_addr) != 0)
 747                 return -EINVAL;
 748
 749         lock_sock(sk);
 750         err = __vsock_bind(sk, vm_addr);
 751         release_sock(sk);
 752
 753         return err;
 754 }
 755
 756 static int vsock_getname(struct socket *sock,
 757                          struct sockaddr *addr, int *addr_len, int peer)
 758 {
 759         int err;
 760         struct sock *sk;
 761         struct vsock_sock *vsk;
 762         struct sockaddr_vm *vm_addr;
 763
 764         sk = sock->sk;
 765         vsk = vsock_sk(sk);
 766         err = 0;
 767
 768         lock_sock(sk);
 769
 770         if (peer) {
 771                 if (sock->state != SS_CONNECTED) {
 772                         err = -ENOTCONN;
 773                         goto out;
 774                 }
 775                 vm_addr = &vsk->remote_addr;
 776         } else {
 777                 vm_addr = &vsk->local_addr;
 778         }
 779
 780         if (!vm_addr) {
 781                 err = -EINVAL;
 782                 goto out;
 783         }
 784
 785         /* sys_getsockname() and sys_getpeername() pass us a
 786          * MAX_SOCK_ADDR-sized buffer and don't set addr_len.  Unfortunately
 787          * that macro is defined in socket.c instead of .h, so we hardcode its
 788          * value here.
 789          */
 790         BUILD_BUG_ON(sizeof(*vm_addr) > 128);
 791         memcpy(addr, vm_addr, sizeof(*vm_addr));
 792         *addr_len = sizeof(*vm_addr);
 793
 794 out:
 795         release_sock(sk);
 796         return err;
 797 }
 798
 799 static int vsock_shutdown(struct socket *sock, int mode)
 800 {
 801         int err;
 802         struct sock *sk;
 803
 804         /* User level uses SHUT_RD (0) and SHUT_WR (1), but the kernel uses
 805          * RCV_SHUTDOWN (1) and SEND_SHUTDOWN (2), so we must increment mode
 806          * here like the other address families do.  Note also that the
 807          * increment makes SHUT_RDWR (2) into RCV_SHUTDOWN | SEND_SHUTDOWN (3),
 808          * which is what we want.
 809          */
 810         mode++;
 811
 812         if ((mode & ~SHUTDOWN_MASK) || !mode)
 813                 return -EINVAL;
 814
 815         /* If this is a STREAM socket and it is not connected then bail out
 816          * immediately.  If it is a DGRAM socket then we must first kick the
 817          * socket so that it wakes up from any sleeping calls, for example
 818          * recv(), and then afterwards return the error.
 819          */
 820
 821         sk = sock->sk;
 822
 823         lock_sock(sk);
 824         if (sock->state == SS_UNCONNECTED) {
 825                 err = -ENOTCONN;
 826                 if (sk->sk_type == SOCK_STREAM)
 827                         goto out;
 828         } else {
 829                 sock->state = SS_DISCONNECTING;
 830                 err = 0;
 831         }
 832
 833         /* Receive and send shutdowns are treated alike. */
 834         mode = mode & (RCV_SHUTDOWN | SEND_SHUTDOWN);
 835         if (mode) {
 836                 sk->sk_shutdown |= mode;
 837                 sk->sk_state_change(sk);
 838
 839                 if (sk->sk_type == SOCK_STREAM) {
 840                         sock_reset_flag(sk, SOCK_DONE);
 841                         vsock_send_shutdown(sk, mode);
 842                 }
 843         }
 844
 845 out:
 846         release_sock(sk);
 847         return err;
 848 }
 849
 850 static unsigned int vsock_poll(struct file *file, struct socket *sock,
 851                                poll_table *wait)
 852 {
 853         struct sock *sk;
 854         unsigned int mask;
 855         struct vsock_sock *vsk;
 856
 857         sk = sock->sk;
 858         vsk = vsock_sk(sk);
 859
 860         poll_wait(file, sk_sleep(sk), wait);
 861         mask = 0;
 862
 863         if (sk->sk_err)
 864                 /* Signify that there has been an error on this socket. */
 865                 mask |= POLLERR;
 866
 867         /* INET sockets treat local write shutdown and peer write shutdown as a
 868          * case of POLLHUP set.
 869          */
 870         if ((sk->sk_shutdown == SHUTDOWN_MASK) ||
 871             ((sk->sk_shutdown & SEND_SHUTDOWN) &&
 872              (vsk->peer_shutdown & SEND_SHUTDOWN))) {
 873                 mask |= POLLHUP;
 874         }
 875
 876         if (sk->sk_shutdown & RCV_SHUTDOWN ||
 877             vsk->peer_shutdown & SEND_SHUTDOWN) {
 878                 mask |= POLLRDHUP;
 879         }
 880
 881         if (sock->type == SOCK_DGRAM) {
 882                 /* For datagram sockets we can read if there is something in
 883                  * the queue and write as long as the socket isn't shutdown for
 884                  * sending.
 885                  */
 886                 if (!skb_queue_empty(&sk->sk_receive_queue) ||
 887                     (sk->sk_shutdown & RCV_SHUTDOWN)) {
 888                         mask |= POLLIN | POLLRDNORM;
 889                 }
 890
 891                 if (!(sk->sk_shutdown & SEND_SHUTDOWN))
 892                         mask |= POLLOUT | POLLWRNORM | POLLWRBAND;
 893
 894         } else if (sock->type == SOCK_STREAM) {
 895                 lock_sock(sk);
 896
 897                 /* Listening sockets that have connections in their accept
 898                  * queue can be read.
 899                  */
 900                 if (sk->sk_state == VSOCK_SS_LISTEN
 901                     && !vsock_is_accept_queue_empty(sk))
 902                         mask |= POLLIN | POLLRDNORM;
 903
 904                 /* If there is something in the queue then we can read. */
 905                 if (transport->stream_is_active(vsk) &&
 906                     !(sk->sk_shutdown & RCV_SHUTDOWN)) {
 907                         bool data_ready_now = false;
 908                         int ret = transport->notify_poll_in(
 909                                         vsk, 1, &data_ready_now);
 910                         if (ret < 0) {
 911                                 mask |= POLLERR;
 912                         } else {
 913                                 if (data_ready_now)
 914                                         mask |= POLLIN | POLLRDNORM;
 915
 916                         }
 917                 }
 918
 919                 /* Sockets whose connections have been closed, reset, or
 920                  * terminated should also be considered read, and we check the
 921                  * shutdown flag for that.
 922                  */
 923                 if (sk->sk_shutdown & RCV_SHUTDOWN ||
 924                     vsk->peer_shutdown & SEND_SHUTDOWN) {
 925                         mask |= POLLIN | POLLRDNORM;
 926                 }
 927
 928                 /* Connected sockets that can produce data can be written. */
 929                 if (sk->sk_state == SS_CONNECTED) {
 930                         if (!(sk->sk_shutdown & SEND_SHUTDOWN)) {
 931                                 bool space_avail_now = false;
 932                                 int ret = transport->notify_poll_out(
 933                                                 vsk, 1, &space_avail_now);
 934                                 if (ret < 0) {
 935                                         mask |= POLLERR;
 936                                 } else {
 937                                         if (space_avail_now)
 938                                                 /* Remove POLLWRBAND since INET
 939                                                  * sockets are not setting it.
 940                                                  */
 941                                                 mask |= POLLOUT | POLLWRNORM;
 942
 943                                 }
 944                         }
 945                 }
 946
 947                 /* Simulate INET socket poll behaviors, which sets
 948                  * POLLOUT|POLLWRNORM when peer is closed and nothing to read,
 949                  * but local send is not shutdown.
 950                  */
 951                 if (sk->sk_state == SS_UNCONNECTED) {
 952                         if (!(sk->sk_shutdown & SEND_SHUTDOWN))
 953                                 mask |= POLLOUT | POLLWRNORM;
 954
 955                 }
 956
 957                 release_sock(sk);
 958         }
 959
 960         return mask;
 961 }
 962
 963 static int vsock_dgram_sendmsg(struct socket *sock, struct msghdr *msg,
 964                                size_t len)
 965 {
 966         int err;
 967         struct sock *sk;
 968         struct vsock_sock *vsk;
 969         struct sockaddr_vm *remote_addr;
 970
 971         if (msg->msg_flags & MSG_OOB)
 972                 return -EOPNOTSUPP;
 973
 974         /* For now, MSG_DONTWAIT is always assumed... */
 975         err = 0;
 976         sk = sock->sk;
 977         vsk = vsock_sk(sk);
 978
 979         lock_sock(sk);
 980
 981         err = vsock_auto_bind(vsk);
 982         if (err)
 983                 goto out;
 984
 985
 986         /* If the provided message contains an address, use that.  Otherwise
 987          * fall back on the socket's remote handle (if it has been connected).
 988          */
 989         if (msg->msg_name &&
 990             vsock_addr_cast(msg->msg_name, msg->msg_namelen,
 991                             &remote_addr) == 0) {
 992                 /* Ensure this address is of the right type and is a valid
 993                  * destination.
 994                  */
 995
 996                 if (remote_addr->svm_cid == VMADDR_CID_ANY)
 997                         remote_addr->svm_cid = transport->get_local_cid();
 998
 999                 if (!vsock_addr_bound(remote_addr)) {
1000                         err = -EINVAL;
1001                         goto out;
1002                 }
1003         } else if (sock->state == SS_CONNECTED) {
1004                 remote_addr = &vsk->remote_addr;
1005
1006                 if (remote_addr->svm_cid == VMADDR_CID_ANY)
1007                         remote_addr->svm_cid = transport->get_local_cid();
1008
1009                 /* XXX Should connect() or this function ensure remote_addr is
1010                  * bound?
1011                  */
1012                 if (!vsock_addr_bound(&vsk->remote_addr)) {
1013                         err = -EINVAL;
1014                         goto out;
1015                 }
1016         } else {
1017                 err = -EINVAL;
1018                 goto out;
1019         }
1020
1021         if (!transport->dgram_allow(remote_addr->svm_cid,
1022                                     remote_addr->svm_port)) {
1023                 err = -EINVAL;
1024                 goto out;
1025         }
1026
1027         err = transport->dgram_enqueue(vsk, remote_addr, msg, len);
1028
1029 out:
1030         release_sock(sk);
1031         return err;
1032 }
1033
1034 static int vsock_dgram_connect(struct socket *sock,
1035                                struct sockaddr *addr, int addr_len, int flags)
1036 {
1037         int err;
1038         struct sock *sk;
1039         struct vsock_sock *vsk;
1040         struct sockaddr_vm *remote_addr;
1041
1042         sk = sock->sk;
1043         vsk = vsock_sk(sk);
1044
1045         err = vsock_addr_cast(addr, addr_len, &remote_addr);
1046         if (err == -EAFNOSUPPORT && remote_addr->svm_family == AF_UNSPEC) {
1047                 lock_sock(sk);
1048                 vsock_addr_init(&vsk->remote_addr, VMADDR_CID_ANY,
1049                                 VMADDR_PORT_ANY);
1050                 sock->state = SS_UNCONNECTED;
1051                 release_sock(sk);
1052                 return 0;
1053         } else if (err != 0)
1054                 return -EINVAL;
1055
1056         lock_sock(sk);
1057
1058         err = vsock_auto_bind(vsk);
1059         if (err)
1060                 goto out;
1061
1062         if (!transport->dgram_allow(remote_addr->svm_cid,
1063                                     remote_addr->svm_port)) {
1064                 err = -EINVAL;
1065                 goto out;
1066         }
1067
1068         memcpy(&vsk->remote_addr, remote_addr, sizeof(vsk->remote_addr));
1069         sock->state = SS_CONNECTED;
1070
1071 out:
1072         release_sock(sk);
1073         return err;
1074 }
1075
1076 static int vsock_dgram_recvmsg(struct socket *sock, struct msghdr *msg,
1077                                size_t len, int flags)
1078 {
1079         return transport->dgram_dequeue(vsock_sk(sock->sk), msg, len, flags);
1080 }
1081
1082 static const struct proto_ops vsock_dgram_ops = {
1083         .family = PF_VSOCK,
1084         .owner = THIS_MODULE,
1085         .release = vsock_release,
1086         .bind = vsock_bind,
1087         .connect = vsock_dgram_connect,
1088         .socketpair = sock_no_socketpair,
1089         .accept = sock_no_accept,
1090         .getname = vsock_getname,
1091         .poll = vsock_poll,
1092         .ioctl = sock_no_ioctl,
1093         .listen = sock_no_listen,
1094         .shutdown = vsock_shutdown,
1095         .setsockopt = sock_no_setsockopt,
1096         .getsockopt = sock_no_getsockopt,
1097         .sendmsg = vsock_dgram_sendmsg,
1098         .recvmsg = vsock_dgram_recvmsg,
1099         .mmap = sock_no_mmap,
1100         .sendpage = sock_no_sendpage,
1101 };
1102
1103 static void vsock_connect_timeout(struct work_struct *work)
1104 {
1105         struct sock *sk;
1106         struct vsock_sock *vsk;
1107
1108         vsk = container_of(work, struct vsock_sock, connect_work.work);
1109         sk = sk_vsock(vsk);
1110
1111         lock_sock(sk);
1112         if (sk->sk_state == SS_CONNECTING &&
1113             (sk->sk_shutdown != SHUTDOWN_MASK)) {
1114                 sk->sk_state = SS_UNCONNECTED;
1115                 sk->sk_err = ETIMEDOUT;
1116                 sk->sk_error_report(sk);
1117         }
1118         release_sock(sk);
1119
1120         sock_put(sk);
1121 }
1122
1123 static int vsock_stream_connect(struct socket *sock, struct sockaddr *addr,
1124                                 int addr_len, int flags)
1125 {
1126         int err;
1127         struct sock *sk;
1128         struct vsock_sock *vsk;
1129         struct sockaddr_vm *remote_addr;
1130         long timeout;
1131         DEFINE_WAIT(wait);
1132
1133         err = 0;
1134         sk = sock->sk;
1135         vsk = vsock_sk(sk);
1136
1137         lock_sock(sk);
1138
1139         /* XXX AF_UNSPEC should make us disconnect like AF_INET. */
1140         switch (sock->state) {
1141         case SS_CONNECTED:
1142                 err = -EISCONN;
1143                 goto out;
1144         case SS_DISCONNECTING:
1145                 err = -EINVAL;
1146                 goto out;
1147         case SS_CONNECTING:
1148                 /* This continues on so we can move sock into the SS_CONNECTED
1149                  * state once the connection has completed (at which point err
1150                  * will be set to zero also).  Otherwise, we will either wait
1151                  * for the connection or return -EALREADY should this be a
1152                  * non-blocking call.
1153                  */
1154                 err = -EALREADY;
1155                 break;
1156         default:
1157                 if ((sk->sk_state == VSOCK_SS_LISTEN) ||
1158                     vsock_addr_cast(addr, addr_len, &remote_addr) != 0) {
1159                         err = -EINVAL;
1160                         goto out;
1161                 }
1162
1163                 /* The hypervisor and well-known contexts do not have socket
1164                  * endpoints.
1165                  */
1166                 if (!transport->stream_allow(remote_addr->svm_cid,
1167                                              remote_addr->svm_port)) {
1168                         err = -ENETUNREACH;
1169                         goto out;
1170                 }
1171
1172                 /* Set the remote address that we are connecting to. */
1173                 memcpy(&vsk->remote_addr, remote_addr,
1174                        sizeof(vsk->remote_addr));
1175
1176                 err = vsock_auto_bind(vsk);
1177                 if (err)
1178                         goto out;
1179
1180                 sk->sk_state = SS_CONNECTING;
1181
1182                 err = transport->connect(vsk);
1183                 if (err < 0)
1184                         goto out;
1185
1186                 /* Mark sock as connecting and set the error code to in
1187                  * progress in case this is a non-blocking connect.
1188                  */
1189                 sock->state = SS_CONNECTING;
1190                 err = -EINPROGRESS;
1191         }
1192
1193         /* The receive path will handle all communication until we are able to
1194          * enter the connected state.  Here we wait for the connection to be
1195          * completed or a notification of an error.
1196          */
1197         timeout = vsk->connect_timeout;
1198         prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
1199
1200         while (sk->sk_state != SS_CONNECTED && sk->sk_err == 0) {
1201                 if (flags & O_NONBLOCK) {
1202                         /* If we're not going to block, we schedule a timeout
1203                          * function to generate a timeout on the connection
1204                          * attempt, in case the peer doesn't respond in a
1205                          * timely manner. We hold on to the socket until the
1206                          * timeout fires.
1207                          */
1208                         sock_hold(sk);
1209                         schedule_delayed_work(&vsk->connect_work, timeout);
1210
1211                         /* Skip ahead to preserve error code set above. */
1212                         goto out_wait;
1213                 }
1214
1215                 release_sock(sk);
1216                 timeout = schedule_timeout(timeout);
1217                 lock_sock(sk);
1218
1219                 if (signal_pending(current)) {
1220                         err = sock_intr_errno(timeout);
1221                         sk->sk_state = SS_UNCONNECTED;
1222                         sock->state = SS_UNCONNECTED;
1223                         goto out_wait;
1224                 } else if (timeout == 0) {
1225                         err = -ETIMEDOUT;
1226                         sk->sk_state = SS_UNCONNECTED;
1227                         sock->state = SS_UNCONNECTED;
1228                         goto out_wait;
1229                 }
1230
1231                 prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
1232         }
1233
1234         if (sk->sk_err) {
1235                 err = -sk->sk_err;
1236                 sk->sk_state = SS_UNCONNECTED;
1237                 sock->state = SS_UNCONNECTED;
1238         } else {
1239                 err = 0;
1240         }
1241
1242 out_wait:
1243         finish_wait(sk_sleep(sk), &wait);
1244 out:
1245         release_sock(sk);
1246         return err;
1247 }
1248
1249 static int vsock_accept(struct socket *sock, struct socket *newsock, int flags)
1250 {
1251         struct sock *listener;
1252         int err;
1253         struct sock *connected;
1254         struct vsock_sock *vconnected;
1255         long timeout;
1256         DEFINE_WAIT(wait);
1257
1258         err = 0;
1259         listener = sock->sk;
1260
1261         lock_sock(listener);
1262
1263         if (sock->type != SOCK_STREAM) {
1264                 err = -EOPNOTSUPP;
1265                 goto out;
1266         }
1267
1268         if (listener->sk_state != VSOCK_SS_LISTEN) {
1269                 err = -EINVAL;
1270                 goto out;
1271         }
1272
1273         /* Wait for children sockets to appear; these are the new sockets
1274          * created upon connection establishment.
1275          */
1276         timeout = sock_rcvtimeo(listener, flags & O_NONBLOCK);
1277         prepare_to_wait(sk_sleep(listener), &wait, TASK_INTERRUPTIBLE);
1278
1279         while ((connected = vsock_dequeue_accept(listener)) == NULL &&
1280                listener->sk_err == 0) {
1281                 release_sock(listener);
1282                 timeout = schedule_timeout(timeout);
1283                 finish_wait(sk_sleep(listener), &wait);
1284                 lock_sock(listener);
1285
1286                 if (signal_pending(current)) {
1287                         err = sock_intr_errno(timeout);
1288                         goto out;
1289                 } else if (timeout == 0) {
1290                         err = -EAGAIN;
1291                         goto out;
1292                 }
1293
1294                 prepare_to_wait(sk_sleep(listener), &wait, TASK_INTERRUPTIBLE);
1295         }
1296         finish_wait(sk_sleep(listener), &wait);
1297
1298         if (listener->sk_err)
1299                 err = -listener->sk_err;
1300
1301         if (connected) {
1302                 listener->sk_ack_backlog--;
1303
1304                 lock_sock(connected);
1305                 vconnected = vsock_sk(connected);
1306
1307                 /* If the listener socket has received an error, then we should
1308                  * reject this socket and return.  Note that we simply mark the
1309                  * socket rejected, drop our reference, and let the cleanup
1310                  * function handle the cleanup; the fact that we found it in
1311                  * the listener's accept queue guarantees that the cleanup
1312                  * function hasn't run yet.
1313                  */
1314                 if (err) {
1315                         vconnected->rejected = true;
1316                 } else {
1317                         newsock->state = SS_CONNECTED;
1318                         sock_graft(connected, newsock);
1319                 }
1320
1321                 release_sock(connected);
1322                 sock_put(connected);
1323         }
1324
1325 out:
1326         release_sock(listener);
1327         return err;
1328 }
1329
1330 static int vsock_listen(struct socket *sock, int backlog)
1331 {
1332         int err;
1333         struct sock *sk;
1334         struct vsock_sock *vsk;
1335
1336         sk = sock->sk;
1337
1338         lock_sock(sk);
1339
1340         if (sock->type != SOCK_STREAM) {
1341                 err = -EOPNOTSUPP;
1342                 goto out;
1343         }
1344
1345         if (sock->state != SS_UNCONNECTED) {
1346                 err = -EINVAL;
1347                 goto out;
1348         }
1349
1350         vsk = vsock_sk(sk);
1351
1352         if (!vsock_addr_bound(&vsk->local_addr)) {
1353                 err = -EINVAL;
1354                 goto out;
1355         }
1356
1357         sk->sk_max_ack_backlog = backlog;
1358         sk->sk_state = VSOCK_SS_LISTEN;
1359
1360         err = 0;
1361
1362 out:
1363         release_sock(sk);
1364         return err;
1365 }
1366
1367 static int vsock_stream_setsockopt(struct socket *sock,
1368                                    int level,
1369                                    int optname,
1370                                    char __user *optval,
1371                                    unsigned int optlen)
1372 {
1373         int err;
1374         struct sock *sk;
1375         struct vsock_sock *vsk;
1376         u64 val;
1377
1378         if (level != AF_VSOCK)
1379                 return -ENOPROTOOPT;
1380
1381 #define COPY_IN(_v)                                       \
1382         do {                                              \
1383                 if (optlen < sizeof(_v)) {                \
1384                         err = -EINVAL;                    \
1385                         goto exit;                        \
1386                 }                                         \
1387                 if (copy_from_user(&_v, optval, sizeof(_v)) != 0) {     \
1388                         err = -EFAULT;                                  \
1389                         goto exit;                                      \
1390                 }                                                       \
1391         } while (0)
1392
1393         err = 0;
1394         sk = sock->sk;
1395         vsk = vsock_sk(sk);
1396
1397         lock_sock(sk);
1398
1399         switch (optname) {
1400         case SO_VM_SOCKETS_BUFFER_SIZE:
1401                 COPY_IN(val);
1402                 transport->set_buffer_size(vsk, val);
1403                 break;
1404
1405         case SO_VM_SOCKETS_BUFFER_MAX_SIZE:
1406                 COPY_IN(val);
1407                 transport->set_max_buffer_size(vsk, val);
1408                 break;
1409
1410         case SO_VM_SOCKETS_BUFFER_MIN_SIZE:
1411                 COPY_IN(val);
1412                 transport->set_min_buffer_size(vsk, val);
1413                 break;
1414
1415         case SO_VM_SOCKETS_CONNECT_TIMEOUT: {
1416                 struct timeval tv;
1417                 COPY_IN(tv);
1418                 if (tv.tv_sec >= 0 && tv.tv_usec < USEC_PER_SEC &&
1419                     tv.tv_sec < (MAX_SCHEDULE_TIMEOUT / HZ - 1)) {
1420                         vsk->connect_timeout = tv.tv_sec * HZ +
1421                             DIV_ROUND_UP(tv.tv_usec, (1000000 / HZ));
1422                         if (vsk->connect_timeout == 0)
1423                                 vsk->connect_timeout =
1424                                     VSOCK_DEFAULT_CONNECT_TIMEOUT;
1425
1426                 } else {
1427                         err = -ERANGE;
1428                 }
1429                 break;
1430         }
1431
1432         default:
1433                 err = -ENOPROTOOPT;
1434                 break;
1435         }
1436
1437 #undef COPY_IN
1438
1439 exit:
1440         release_sock(sk);
1441         return err;
1442 }
1443
1444 static int vsock_stream_getsockopt(struct socket *sock,
1445                                    int level, int optname,
1446                                    char __user *optval,
1447                                    int __user *optlen)
1448 {
1449         int err;
1450         int len;
1451         struct sock *sk;
1452         struct vsock_sock *vsk;
1453         u64 val;
1454
1455         if (level != AF_VSOCK)
1456                 return -ENOPROTOOPT;
1457
1458         err = get_user(len, optlen);
1459         if (err != 0)
1460                 return err;
1461
1462 #define COPY_OUT(_v)                            \
1463         do {                                    \
1464                 if (len < sizeof(_v))           \
1465                         return -EINVAL;         \
1466                                                 \
1467                 len = sizeof(_v);               \
1468                 if (copy_to_user(optval, &_v, len) != 0)        \
1469                         return -EFAULT;                         \
1470                                                                 \
1471         } while (0)
1472
1473         err = 0;
1474         sk = sock->sk;
1475         vsk = vsock_sk(sk);
1476
1477         switch (optname) {
1478         case SO_VM_SOCKETS_BUFFER_SIZE:
1479                 val = transport->get_buffer_size(vsk);
1480                 COPY_OUT(val);
1481                 break;
1482
1483         case SO_VM_SOCKETS_BUFFER_MAX_SIZE:
1484                 val = transport->get_max_buffer_size(vsk);
1485                 COPY_OUT(val);
1486                 break;
1487
1488         case SO_VM_SOCKETS_BUFFER_MIN_SIZE:
1489                 val = transport->get_min_buffer_size(vsk);
1490                 COPY_OUT(val);
1491                 break;
1492
1493         case SO_VM_SOCKETS_CONNECT_TIMEOUT: {
1494                 struct timeval tv;
1495                 tv.tv_sec = vsk->connect_timeout / HZ;
1496                 tv.tv_usec =
1497                     (vsk->connect_timeout -
1498                      tv.tv_sec * HZ) * (1000000 / HZ);
1499                 COPY_OUT(tv);
1500                 break;
1501         }
1502         default:
1503                 return -ENOPROTOOPT;
1504         }
1505
1506         err = put_user(len, optlen);
1507         if (err != 0)
1508                 return -EFAULT;
1509
1510 #undef COPY_OUT
1511
1512         return 0;
1513 }
1514
1515 static int vsock_stream_sendmsg(struct socket *sock, struct msghdr *msg,
1516                                 size_t len)
1517 {
1518         struct sock *sk;
1519         struct vsock_sock *vsk;
1520         ssize_t total_written;
1521         long timeout;
1522         int err;
1523         struct vsock_transport_send_notify_data send_data;
1524         DEFINE_WAIT_FUNC(wait, woken_wake_function);
1525
1526         sk = sock->sk;
1527         vsk = vsock_sk(sk);
1528         total_written = 0;
1529         err = 0;
1530
1531         if (msg->msg_flags & MSG_OOB)
1532                 return -EOPNOTSUPP;
1533
1534         lock_sock(sk);
1535
1536         /* Callers should not provide a destination with stream sockets. */
1537         if (msg->msg_namelen) {
1538                 err = sk->sk_state == SS_CONNECTED ? -EISCONN : -EOPNOTSUPP;
1539                 goto out;
1540         }
1541
1542         /* Send data only if both sides are not shutdown in the direction. */
1543         if (sk->sk_shutdown & SEND_SHUTDOWN ||
1544             vsk->peer_shutdown & RCV_SHUTDOWN) {
1545                 err = -EPIPE;
1546                 goto out;
1547         }
1548
1549         if (sk->sk_state != SS_CONNECTED ||
1550             !vsock_addr_bound(&vsk->local_addr)) {
1551                 err = -ENOTCONN;
1552                 goto out;
1553         }
1554
1555         if (!vsock_addr_bound(&vsk->remote_addr)) {
1556                 err = -EDESTADDRREQ;
1557                 goto out;
1558         }
1559
1560         /* Wait for room in the produce queue to enqueue our user's data. */
1561         timeout = sock_sndtimeo(sk, msg->msg_flags & MSG_DONTWAIT);
1562
1563         err = transport->notify_send_init(vsk, &send_data);
1564         if (err < 0)
1565                 goto out;
1566
1567         while (total_written < len) {
1568                 ssize_t written;
1569
1570                 add_wait_queue(sk_sleep(sk), &wait);
1571                 while (vsock_stream_has_space(vsk) == 0 &&
1572                        sk->sk_err == 0 &&
1573                        !(sk->sk_shutdown & SEND_SHUTDOWN) &&
1574                        !(vsk->peer_shutdown & RCV_SHUTDOWN)) {
1575
1576                         /* Don't wait for non-blocking sockets. */
1577                         if (timeout == 0) {
1578                                 err = -EAGAIN;
1579                                 remove_wait_queue(sk_sleep(sk), &wait);
1580                                 goto out_err;
1581                         }
1582
1583                         err = transport->notify_send_pre_block(vsk, &send_data);
1584                         if (err < 0) {
1585                                 remove_wait_queue(sk_sleep(sk), &wait);
1586                                 goto out_err;
1587                         }
1588
1589                         release_sock(sk);
1590                         timeout = wait_woken(&wait, TASK_INTERRUPTIBLE, timeout);
1591                         lock_sock(sk);
1592                         if (signal_pending(current)) {
1593                                 err = sock_intr_errno(timeout);
1594                                 remove_wait_queue(sk_sleep(sk), &wait);
1595                                 goto out_err;
1596                         } else if (timeout == 0) {
1597                                 err = -EAGAIN;
1598                                 remove_wait_queue(sk_sleep(sk), &wait);
1599                                 goto out_err;
1600                         }
1601                 }
1602                 remove_wait_queue(sk_sleep(sk), &wait);
1603
1604                 /* These checks occur both as part of and after the loop
1605                  * conditional since we need to check before and after
1606                  * sleeping.
1607                  */
1608                 if (sk->sk_err) {
1609                         err = -sk->sk_err;
1610                         goto out_err;
1611                 } else if ((sk->sk_shutdown & SEND_SHUTDOWN) ||
1612                            (vsk->peer_shutdown & RCV_SHUTDOWN)) {
1613                         err = -EPIPE;
1614                         goto out_err;
1615                 }
1616
1617                 err = transport->notify_send_pre_enqueue(vsk, &send_data);
1618                 if (err < 0)
1619                         goto out_err;
1620
1621                 /* Note that enqueue will only write as many bytes as are free
1622                  * in the produce queue, so we don't need to ensure len is
1623                  * smaller than the queue size.  It is the caller's
1624                  * responsibility to check how many bytes we were able to send.
1625                  */
1626
1627                 written = transport->stream_enqueue(
1628                                 vsk, msg,
1629                                 len - total_written);
1630                 if (written < 0) {
1631                         err = -ENOMEM;
1632                         goto out_err;
1633                 }
1634
1635                 total_written += written;
1636
1637                 err = transport->notify_send_post_enqueue(
1638                                 vsk, written, &send_data);
1639                 if (err < 0)
1640                         goto out_err;
1641
1642         }
1643
1644 out_err:
1645         if (total_written > 0)
1646                 err = total_written;
1647 out:
1648         release_sock(sk);
1649         return err;
1650 }
1651
1652
1653 static int
1654 vsock_stream_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
1655                      int flags)
1656 {
1657         struct sock *sk;
1658         struct vsock_sock *vsk;
1659         int err;
1660         size_t target;
1661         ssize_t copied;
1662         long timeout;
1663         struct vsock_transport_recv_notify_data recv_data;
1664
1665         DEFINE_WAIT(wait);
1666
1667         sk = sock->sk;
1668         vsk = vsock_sk(sk);
1669         err = 0;
1670
1671         lock_sock(sk);
1672
1673         if (sk->sk_state != SS_CONNECTED) {
1674                 /* Recvmsg is supposed to return 0 if a peer performs an
1675                  * orderly shutdown. Differentiate between that case and when a
1676                  * peer has not connected or a local shutdown occured with the
1677                  * SOCK_DONE flag.
1678                  */
1679                 if (sock_flag(sk, SOCK_DONE))
1680                         err = 0;
1681                 else
1682                         err = -ENOTCONN;
1683
1684                 goto out;
1685         }
1686
1687         if (flags & MSG_OOB) {
1688                 err = -EOPNOTSUPP;
1689                 goto out;
1690         }
1691
1692         /* We don't check peer_shutdown flag here since peer may actually shut
1693          * down, but there can be data in the queue that a local socket can
1694          * receive.
1695          */
1696         if (sk->sk_shutdown & RCV_SHUTDOWN) {
1697                 err = 0;
1698                 goto out;
1699         }
1700
1701         /* It is valid on Linux to pass in a zero-length receive buffer.  This
1702          * is not an error.  We may as well bail out now.
1703          */
1704         if (!len) {
1705                 err = 0;
1706                 goto out;
1707         }
1708
1709         /* We must not copy less than target bytes into the user's buffer
1710          * before returning successfully, so we wait for the consume queue to
1711          * have that much data to consume before dequeueing.  Note that this
1712          * makes it impossible to handle cases where target is greater than the
1713          * queue size.
1714          */
1715         target = sock_rcvlowat(sk, flags & MSG_WAITALL, len);
1716         if (target >= transport->stream_rcvhiwat(vsk)) {
1717                 err = -ENOMEM;
1718                 goto out;
1719         }
1720         timeout = sock_rcvtimeo(sk, flags & MSG_DONTWAIT);
1721         copied = 0;
1722
1723         err = transport->notify_recv_init(vsk, target, &recv_data);
1724         if (err < 0)
1725                 goto out;
1726
1727
1728         while (1) {
1729                 s64 ready;
1730
1731                 prepare_to_wait(sk_sleep(sk), &wait, TASK_INTERRUPTIBLE);
1732                 ready = vsock_stream_has_data(vsk);
1733
1734                 if (ready == 0) {
1735                         if (sk->sk_err != 0 ||
1736                             (sk->sk_shutdown & RCV_SHUTDOWN) ||
1737                             (vsk->peer_shutdown & SEND_SHUTDOWN)) {
1738                                 finish_wait(sk_sleep(sk), &wait);
1739                                 break;
1740                         }
1741                         /* Don't wait for non-blocking sockets. */
1742                         if (timeout == 0) {
1743                                 err = -EAGAIN;
1744                                 finish_wait(sk_sleep(sk), &wait);
1745                                 break;
1746                         }
1747
1748                         err = transport->notify_recv_pre_block(
1749                                         vsk, target, &recv_data);
1750                         if (err < 0) {
1751                                 finish_wait(sk_sleep(sk), &wait);
1752                                 break;
1753                         }
1754                         release_sock(sk);
1755                         timeout = schedule_timeout(timeout);
1756                         lock_sock(sk);
1757
1758                         if (signal_pending(current)) {
1759                                 err = sock_intr_errno(timeout);
1760                                 finish_wait(sk_sleep(sk), &wait);
1761                                 break;
1762                         } else if (timeout == 0) {
1763                                 err = -EAGAIN;
1764                                 finish_wait(sk_sleep(sk), &wait);
1765                                 break;
1766                         }
1767                 } else {
1768                         ssize_t read;
1769
1770                         finish_wait(sk_sleep(sk), &wait);
1771
1772                         if (ready < 0) {
1773                                 /* Invalid queue pair content. XXX This should
1774                                 * be changed to a connection reset in a later
1775                                 * change.
1776                                 */
1777
1778                                 err = -ENOMEM;
1779                                 goto out;
1780                         }
1781
1782                         err = transport->notify_recv_pre_dequeue(
1783                                         vsk, target, &recv_data);
1784                         if (err < 0)
1785                                 break;
1786
1787                         read = transport->stream_dequeue(
1788                                         vsk, msg,
1789                                         len - copied, flags);
1790                         if (read < 0) {
1791                                 err = -ENOMEM;
1792                                 break;
1793                         }
1794
1795                         copied += read;
1796
1797                         err = transport->notify_recv_post_dequeue(
1798                                         vsk, target, read,
1799                                         !(flags & MSG_PEEK), &recv_data);
1800                         if (err < 0)
1801                                 goto out;
1802
1803                         if (read >= target || flags & MSG_PEEK)
1804                                 break;
1805
1806                         target -= read;
1807                 }
1808         }
1809
1810         if (sk->sk_err)
1811                 err = -sk->sk_err;
1812         else if (sk->sk_shutdown & RCV_SHUTDOWN)
1813                 err = 0;
1814
1815         if (copied > 0)
1816                 err = copied;
1817
1818 out:
1819         release_sock(sk);
1820         return err;
1821 }
1822
1823 static const struct proto_ops vsock_stream_ops = {
1824         .family = PF_VSOCK,
1825         .owner = THIS_MODULE,
1826         .release = vsock_release,
1827         .bind = vsock_bind,
1828         .connect = vsock_stream_connect,
1829         .socketpair = sock_no_socketpair,
1830         .accept = vsock_accept,
1831         .getname = vsock_getname,
1832         .poll = vsock_poll,
1833         .ioctl = sock_no_ioctl,
1834         .listen = vsock_listen,
1835         .shutdown = vsock_shutdown,
1836         .setsockopt = vsock_stream_setsockopt,
1837         .getsockopt = vsock_stream_getsockopt,
1838         .sendmsg = vsock_stream_sendmsg,
1839         .recvmsg = vsock_stream_recvmsg,
1840         .mmap = sock_no_mmap,
1841         .sendpage = sock_no_sendpage,
1842 };
1843
1844 static int vsock_create(struct net *net, struct socket *sock,
1845                         int protocol, int kern)
1846 {
1847         if (!sock)
1848                 return -EINVAL;
1849
1850         if (protocol && protocol != PF_VSOCK)
1851                 return -EPROTONOSUPPORT;
1852
1853         switch (sock->type) {
1854         case SOCK_DGRAM:
1855                 sock->ops = &vsock_dgram_ops;
1856                 break;
1857         case SOCK_STREAM:
1858                 sock->ops = &vsock_stream_ops;
1859                 break;
1860         default:
1861                 return -ESOCKTNOSUPPORT;
1862         }
1863
1864         sock->state = SS_UNCONNECTED;
1865
1866         return __vsock_create(net, sock, NULL, GFP_KERNEL, 0, kern) ? 0 : -ENOMEM;
1867 }
1868
1869 static const struct net_proto_family vsock_family_ops = {
1870         .family = AF_VSOCK,
1871         .create = vsock_create,
1872         .owner = THIS_MODULE,
1873 };
1874
1875 static long vsock_dev_do_ioctl(struct file *filp,
1876                                unsigned int cmd, void __user *ptr)
1877 {
1878         u32 __user *p = ptr;
1879         int retval = 0;
1880
1881         switch (cmd) {
1882         case IOCTL_VM_SOCKETS_GET_LOCAL_CID:
1883                 if (put_user(transport->get_local_cid(), p) != 0)
1884                         retval = -EFAULT;
1885                 break;
1886
1887         default:
1888                 pr_err("Unknown ioctl %d\n", cmd);
1889                 retval = -EINVAL;
1890         }
1891
1892         return retval;
1893 }
1894
1895 static long vsock_dev_ioctl(struct file *filp,
1896                             unsigned int cmd, unsigned long arg)
1897 {
1898         return vsock_dev_do_ioctl(filp, cmd, (void __user *)arg);
1899 }
1900
1901 #ifdef CONFIG_COMPAT
1902 static long vsock_dev_compat_ioctl(struct file *filp,
1903                                    unsigned int cmd, unsigned long arg)
1904 {
1905         return vsock_dev_do_ioctl(filp, cmd, compat_ptr(arg));
1906 }
1907 #endif
1908
1909 static const struct file_operations vsock_device_ops = {
1910         .owner          = THIS_MODULE,
1911         .unlocked_ioctl = vsock_dev_ioctl,
1912 #ifdef CONFIG_COMPAT
1913         .compat_ioctl   = vsock_dev_compat_ioctl,
1914 #endif
1915         .open           = nonseekable_open,
1916 };
1917
1918 static struct miscdevice vsock_device = {
1919         .name           = "vsock",
1920         .fops           = &vsock_device_ops,
1921 };
1922
1923 int __vsock_core_init(const struct vsock_transport *t, struct module *owner)
1924 {
1925         int err = mutex_lock_interruptible(&vsock_register_mutex);
1926
1927         if (err)
1928                 return err;
1929
1930         if (transport) {
1931                 err = -EBUSY;
1932                 goto err_busy;
1933         }
1934
1935         /* Transport must be the owner of the protocol so that it can't
1936          * unload while there are open sockets.
1937          */
1938         vsock_proto.owner = owner;
1939         transport = t;
1940
1941         vsock_init_tables();
1942
1943         vsock_device.minor = MISC_DYNAMIC_MINOR;
1944         err = misc_register(&vsock_device);
1945         if (err) {
1946                 pr_err("Failed to register misc device\n");
1947                 goto err_reset_transport;
1948         }
1949
1950         err = proto_register(&vsock_proto, 1);  /* we want our slab */
1951         if (err) {
1952                 pr_err("Cannot register vsock protocol\n");
1953                 goto err_deregister_misc;
1954         }
1955
1956         err = sock_register(&vsock_family_ops);
1957         if (err) {
1958                 pr_err("could not register af_vsock (%d) address family: %d\n",
1959                        AF_VSOCK, err);
1960                 goto err_unregister_proto;
1961         }
1962
1963         mutex_unlock(&vsock_register_mutex);
1964         return 0;
1965
1966 err_unregister_proto:
1967         proto_unregister(&vsock_proto);
1968 err_deregister_misc:
1969         misc_deregister(&vsock_device);
1970 err_reset_transport:
1971         transport = NULL;
1972 err_busy:
1973         mutex_unlock(&vsock_register_mutex);
1974         return err;
1975 }
1976 EXPORT_SYMBOL_GPL(__vsock_core_init);
1977
1978 void vsock_core_exit(void)
1979 {
1980         mutex_lock(&vsock_register_mutex);
1981
1982         misc_deregister(&vsock_device);
1983         sock_unregister(AF_VSOCK);
1984         proto_unregister(&vsock_proto);
1985
1986         /* We do not want the assignment below re-ordered. */
1987         mb();
1988         transport = NULL;
1989
1990         mutex_unlock(&vsock_register_mutex);
1991 }
1992 EXPORT_SYMBOL_GPL(vsock_core_exit);
1993
1994 MODULE_AUTHOR("VMware, Inc.");
1995 MODULE_DESCRIPTION("VMware Virtual Socket Family");
1996 MODULE_VERSION("1.0.1.0-k");
1997 MODULE_LICENSE("GPL v2");