net/vmw_vsock/virtio_transport_common.c

   1 // SPDX-License-Identifier: GPL-2.0-only
   2 /*
   3  * common code for virtio vsock
   4  *
   5  * Copyright (C) 2013-2015 Red Hat, Inc.
   6  * Author: Asias He <asias@redhat.com>
   7  *         Stefan Hajnoczi <stefanha@redhat.com>
   8  */
   9 #include <linux/spinlock.h>
  10 #include <linux/module.h>
  11 #include <linux/sched/signal.h>
  12 #include <linux/ctype.h>
  13 #include <linux/list.h>
  14 #include <linux/virtio_vsock.h>
  15 #include <uapi/linux/vsockmon.h>
  16
  17 #include <net/sock.h>
  18 #include <net/af_vsock.h>
  19
  20 #define CREATE_TRACE_POINTS
  21 #include <trace/events/vsock_virtio_transport_common.h>
  22
  23 /* How long to wait for graceful shutdown of a connection */
  24 #define VSOCK_CLOSE_TIMEOUT (8 * HZ)
  25
  26 /* Threshold for detecting small packets to copy */
  27 #define GOOD_COPY_LEN  128
  28
  29 static const struct virtio_transport *
  30 virtio_transport_get_ops(struct vsock_sock *vsk)
  31 {
  32         const struct vsock_transport *t = vsock_core_get_transport(vsk);
  33
  34         if (WARN_ON(!t))
  35                 return NULL;
  36
  37         return container_of(t, struct virtio_transport, transport);
  38 }
  39
  40 /* Returns a new packet on success, otherwise returns NULL.
  41  *
  42  * If NULL is returned, errp is set to a negative errno.
  43  */
  44 static struct sk_buff *
  45 virtio_transport_alloc_skb(struct virtio_vsock_pkt_info *info,
  46                            size_t len,
  47                            u32 src_cid,
  48                            u32 src_port,
  49                            u32 dst_cid,
  50                            u32 dst_port)
  51 {
  52         const size_t skb_len = VIRTIO_VSOCK_SKB_HEADROOM + len;
  53         struct virtio_vsock_hdr *hdr;
  54         struct sk_buff *skb;
  55         void *payload;
  56         int err;
  57
  58         skb = virtio_vsock_alloc_skb(skb_len, GFP_KERNEL);
  59         if (!skb)
  60                 return NULL;
  61
  62         hdr = virtio_vsock_hdr(skb);
  63         hdr->type       = cpu_to_le16(info->type);
  64         hdr->op         = cpu_to_le16(info->op);
  65         hdr->src_cid    = cpu_to_le64(src_cid);
  66         hdr->dst_cid    = cpu_to_le64(dst_cid);
  67         hdr->src_port   = cpu_to_le32(src_port);
  68         hdr->dst_port   = cpu_to_le32(dst_port);
  69         hdr->flags      = cpu_to_le32(info->flags);
  70         hdr->len        = cpu_to_le32(len);
  71
  72         if (info->msg && len > 0) {
  73                 payload = skb_put(skb, len);
  74                 err = memcpy_from_msg(payload, info->msg, len);
  75                 if (err)
  76                         goto out;
  77
  78                 if (msg_data_left(info->msg) == 0 &&
  79                     info->type == VIRTIO_VSOCK_TYPE_SEQPACKET) {
  80                         hdr->flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOM);
  81
  82                         if (info->msg->msg_flags & MSG_EOR)
  83                                 hdr->flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);
  84                 }
  85         }
  86
  87         if (info->reply)
  88                 virtio_vsock_skb_set_reply(skb);
  89
  90         trace_virtio_transport_alloc_pkt(src_cid, src_port,
  91                                          dst_cid, dst_port,
  92                                          len,
  93                                          info->type,
  94                                          info->op,
  95                                          info->flags);
  96
  97         if (info->vsk && !skb_set_owner_sk_safe(skb, sk_vsock(info->vsk))) {
  98                 WARN_ONCE(1, "failed to allocate skb on vsock socket with sk_refcnt == 0\n");
  99                 goto out;
 100         }
 101
 102         return skb;
 103
 104 out:
 105         kfree_skb(skb);
 106         return NULL;
 107 }
 108
 109 /* Packet capture */
 110 static struct sk_buff *virtio_transport_build_skb(void *opaque)
 111 {
 112         struct virtio_vsock_hdr *pkt_hdr;
 113         struct sk_buff *pkt = opaque;
 114         struct af_vsockmon_hdr *hdr;
 115         struct sk_buff *skb;
 116         size_t payload_len;
 117         void *payload_buf;
 118
 119         /* A packet could be split to fit the RX buffer, so we can retrieve
 120          * the payload length from the header and the buffer pointer taking
 121          * care of the offset in the original packet.
 122          */
 123         pkt_hdr = virtio_vsock_hdr(pkt);
 124         payload_len = pkt->len;
 125         payload_buf = pkt->data;
 126
 127         skb = alloc_skb(sizeof(*hdr) + sizeof(*pkt_hdr) + payload_len,
 128                         GFP_ATOMIC);
 129         if (!skb)
 130                 return NULL;
 131
 132         hdr = skb_put(skb, sizeof(*hdr));
 133
 134         /* pkt->hdr is little-endian so no need to byteswap here */
 135         hdr->src_cid = pkt_hdr->src_cid;
 136         hdr->src_port = pkt_hdr->src_port;
 137         hdr->dst_cid = pkt_hdr->dst_cid;
 138         hdr->dst_port = pkt_hdr->dst_port;
 139
 140         hdr->transport = cpu_to_le16(AF_VSOCK_TRANSPORT_VIRTIO);
 141         hdr->len = cpu_to_le16(sizeof(*pkt_hdr));
 142         memset(hdr->reserved, 0, sizeof(hdr->reserved));
 143
 144         switch (le16_to_cpu(pkt_hdr->op)) {
 145         case VIRTIO_VSOCK_OP_REQUEST:
 146         case VIRTIO_VSOCK_OP_RESPONSE:
 147                 hdr->op = cpu_to_le16(AF_VSOCK_OP_CONNECT);
 148                 break;
 149         case VIRTIO_VSOCK_OP_RST:
 150         case VIRTIO_VSOCK_OP_SHUTDOWN:
 151                 hdr->op = cpu_to_le16(AF_VSOCK_OP_DISCONNECT);
 152                 break;
 153         case VIRTIO_VSOCK_OP_RW:
 154                 hdr->op = cpu_to_le16(AF_VSOCK_OP_PAYLOAD);
 155                 break;
 156         case VIRTIO_VSOCK_OP_CREDIT_UPDATE:
 157         case VIRTIO_VSOCK_OP_CREDIT_REQUEST:
 158                 hdr->op = cpu_to_le16(AF_VSOCK_OP_CONTROL);
 159                 break;
 160         default:
 161                 hdr->op = cpu_to_le16(AF_VSOCK_OP_UNKNOWN);
 162                 break;
 163         }
 164
 165         skb_put_data(skb, pkt_hdr, sizeof(*pkt_hdr));
 166
 167         if (payload_len) {
 168                 skb_put_data(skb, payload_buf, payload_len);
 169         }
 170
 171         return skb;
 172 }
 173
 174 void virtio_transport_deliver_tap_pkt(struct sk_buff *skb)
 175 {
 176         if (virtio_vsock_skb_tap_delivered(skb))
 177                 return;
 178
 179         vsock_deliver_tap(virtio_transport_build_skb, skb);
 180         virtio_vsock_skb_set_tap_delivered(skb);
 181 }
 182 EXPORT_SYMBOL_GPL(virtio_transport_deliver_tap_pkt);
 183
 184 static u16 virtio_transport_get_type(struct sock *sk)
 185 {
 186         if (sk->sk_type == SOCK_STREAM)
 187                 return VIRTIO_VSOCK_TYPE_STREAM;
 188         else
 189                 return VIRTIO_VSOCK_TYPE_SEQPACKET;
 190 }
 191
 192 /* This function can only be used on connecting/connected sockets,
 193  * since a socket assigned to a transport is required.
 194  *
 195  * Do not use on listener sockets!
 196  */
 197 static int virtio_transport_send_pkt_info(struct vsock_sock *vsk,
 198                                           struct virtio_vsock_pkt_info *info)
 199 {
 200         u32 src_cid, src_port, dst_cid, dst_port;
 201         const struct virtio_transport *t_ops;
 202         struct virtio_vsock_sock *vvs;
 203         u32 pkt_len = info->pkt_len;
 204         u32 rest_len;
 205         int ret;
 206
 207         info->type = virtio_transport_get_type(sk_vsock(vsk));
 208
 209         t_ops = virtio_transport_get_ops(vsk);
 210         if (unlikely(!t_ops))
 211                 return -EFAULT;
 212
 213         src_cid = t_ops->transport.get_local_cid();
 214         src_port = vsk->local_addr.svm_port;
 215         if (!info->remote_cid) {
 216                 dst_cid = vsk->remote_addr.svm_cid;
 217                 dst_port = vsk->remote_addr.svm_port;
 218         } else {
 219                 dst_cid = info->remote_cid;
 220                 dst_port = info->remote_port;
 221         }
 222
 223         vvs = vsk->trans;
 224
 225         /* virtio_transport_get_credit might return less than pkt_len credit */
 226         pkt_len = virtio_transport_get_credit(vvs, pkt_len);
 227
 228         /* Do not send zero length OP_RW pkt */
 229         if (pkt_len == 0 && info->op == VIRTIO_VSOCK_OP_RW)
 230                 return pkt_len;
 231
 232         rest_len = pkt_len;
 233
 234         do {
 235                 struct sk_buff *skb;
 236                 size_t skb_len;
 237
 238                 skb_len = min_t(u32, VIRTIO_VSOCK_MAX_PKT_BUF_SIZE, rest_len);
 239
 240                 skb = virtio_transport_alloc_skb(info, skb_len,
 241                                                  src_cid, src_port,
 242                                                  dst_cid, dst_port);
 243                 if (!skb) {
 244                         ret = -ENOMEM;
 245                         break;
 246                 }
 247
 248                 virtio_transport_inc_tx_pkt(vvs, skb);
 249
 250                 ret = t_ops->send_pkt(skb);
 251                 if (ret < 0)
 252                         break;
 253
 254                 /* Both virtio and vhost 'send_pkt()' returns 'skb_len',
 255                  * but for reliability use 'ret' instead of 'skb_len'.
 256                  * Also if partial send happens (e.g. 'ret' != 'skb_len')
 257                  * somehow, we break this loop, but account such returned
 258                  * value in 'virtio_transport_put_credit()'.
 259                  */
 260                 rest_len -= ret;
 261
 262                 if (WARN_ONCE(ret != skb_len,
 263                               "'send_pkt()' returns %i, but %zu expected\n",
 264                               ret, skb_len))
 265                         break;
 266         } while (rest_len);
 267
 268         virtio_transport_put_credit(vvs, rest_len);
 269
 270         /* Return number of bytes, if any data has been sent. */
 271         if (rest_len != pkt_len)
 272                 ret = pkt_len - rest_len;
 273
 274         return ret;
 275 }
 276
 277 static bool virtio_transport_inc_rx_pkt(struct virtio_vsock_sock *vvs,
 278                                         u32 len)
 279 {
 280         if (vvs->rx_bytes + len > vvs->buf_alloc)
 281                 return false;
 282
 283         vvs->rx_bytes += len;
 284         return true;
 285 }
 286
 287 static void virtio_transport_dec_rx_pkt(struct virtio_vsock_sock *vvs,
 288                                         u32 len)
 289 {
 290         vvs->rx_bytes -= len;
 291         vvs->fwd_cnt += len;
 292 }
 293
 294 void virtio_transport_inc_tx_pkt(struct virtio_vsock_sock *vvs, struct sk_buff *skb)
 295 {
 296         struct virtio_vsock_hdr *hdr = virtio_vsock_hdr(skb);
 297
 298         spin_lock_bh(&vvs->rx_lock);
 299         vvs->last_fwd_cnt = vvs->fwd_cnt;
 300         hdr->fwd_cnt = cpu_to_le32(vvs->fwd_cnt);
 301         hdr->buf_alloc = cpu_to_le32(vvs->buf_alloc);
 302         spin_unlock_bh(&vvs->rx_lock);
 303 }
 304 EXPORT_SYMBOL_GPL(virtio_transport_inc_tx_pkt);
 305
 306 u32 virtio_transport_get_credit(struct virtio_vsock_sock *vvs, u32 credit)
 307 {
 308         u32 ret;
 309
 310         if (!credit)
 311                 return 0;
 312
 313         spin_lock_bh(&vvs->tx_lock);
 314         ret = vvs->peer_buf_alloc - (vvs->tx_cnt - vvs->peer_fwd_cnt);
 315         if (ret > credit)
 316                 ret = credit;
 317         vvs->tx_cnt += ret;
 318         spin_unlock_bh(&vvs->tx_lock);
 319
 320         return ret;
 321 }
 322 EXPORT_SYMBOL_GPL(virtio_transport_get_credit);
 323
 324 void virtio_transport_put_credit(struct virtio_vsock_sock *vvs, u32 credit)
 325 {
 326         if (!credit)
 327                 return;
 328
 329         spin_lock_bh(&vvs->tx_lock);
 330         vvs->tx_cnt -= credit;
 331         spin_unlock_bh(&vvs->tx_lock);
 332 }
 333 EXPORT_SYMBOL_GPL(virtio_transport_put_credit);
 334
 335 static int virtio_transport_send_credit_update(struct vsock_sock *vsk)
 336 {
 337         struct virtio_vsock_pkt_info info = {
 338                 .op = VIRTIO_VSOCK_OP_CREDIT_UPDATE,
 339                 .vsk = vsk,
 340         };
 341
 342         return virtio_transport_send_pkt_info(vsk, &info);
 343 }
 344
 345 static ssize_t
 346 virtio_transport_stream_do_peek(struct vsock_sock *vsk,
 347                                 struct msghdr *msg,
 348                                 size_t len)
 349 {
 350         struct virtio_vsock_sock *vvs = vsk->trans;
 351         size_t bytes, total = 0, off;
 352         struct sk_buff *skb, *tmp;
 353         int err = -EFAULT;
 354
 355         spin_lock_bh(&vvs->rx_lock);
 356
 357         skb_queue_walk_safe(&vvs->rx_queue, skb,  tmp) {
 358                 off = 0;
 359
 360                 if (total == len)
 361                         break;
 362
 363                 while (total < len && off < skb->len) {
 364                         bytes = len - total;
 365                         if (bytes > skb->len - off)
 366                                 bytes = skb->len - off;
 367
 368                         /* sk_lock is held by caller so no one else can dequeue.
 369                          * Unlock rx_lock since memcpy_to_msg() may sleep.
 370                          */
 371                         spin_unlock_bh(&vvs->rx_lock);
 372
 373                         err = memcpy_to_msg(msg, skb->data + off, bytes);
 374                         if (err)
 375                                 goto out;
 376
 377                         spin_lock_bh(&vvs->rx_lock);
 378
 379                         total += bytes;
 380                         off += bytes;
 381                 }
 382         }
 383
 384         spin_unlock_bh(&vvs->rx_lock);
 385
 386         return total;
 387
 388 out:
 389         if (total)
 390                 err = total;
 391         return err;
 392 }
 393
 394 static ssize_t
 395 virtio_transport_stream_do_dequeue(struct vsock_sock *vsk,
 396                                    struct msghdr *msg,
 397                                    size_t len)
 398 {
 399         struct virtio_vsock_sock *vvs = vsk->trans;
 400         size_t bytes, total = 0;
 401         struct sk_buff *skb;
 402         int err = -EFAULT;
 403         u32 free_space;
 404
 405         spin_lock_bh(&vvs->rx_lock);
 406
 407         if (WARN_ONCE(skb_queue_empty(&vvs->rx_queue) && vvs->rx_bytes,
 408                       "rx_queue is empty, but rx_bytes is non-zero\n")) {
 409                 spin_unlock_bh(&vvs->rx_lock);
 410                 return err;
 411         }
 412
 413         while (total < len && !skb_queue_empty(&vvs->rx_queue)) {
 414                 skb = skb_peek(&vvs->rx_queue);
 415
 416                 bytes = len - total;
 417                 if (bytes > skb->len)
 418                         bytes = skb->len;
 419
 420                 /* sk_lock is held by caller so no one else can dequeue.
 421                  * Unlock rx_lock since memcpy_to_msg() may sleep.
 422                  */
 423                 spin_unlock_bh(&vvs->rx_lock);
 424
 425                 err = memcpy_to_msg(msg, skb->data, bytes);
 426                 if (err)
 427                         goto out;
 428
 429                 spin_lock_bh(&vvs->rx_lock);
 430
 431                 total += bytes;
 432                 skb_pull(skb, bytes);
 433
 434                 if (skb->len == 0) {
 435                         u32 pkt_len = le32_to_cpu(virtio_vsock_hdr(skb)->len);
 436
 437                         virtio_transport_dec_rx_pkt(vvs, pkt_len);
 438                         __skb_unlink(skb, &vvs->rx_queue);
 439                         consume_skb(skb);
 440                 }
 441         }
 442
 443         free_space = vvs->buf_alloc - (vvs->fwd_cnt - vvs->last_fwd_cnt);
 444
 445         spin_unlock_bh(&vvs->rx_lock);
 446
 447         /* To reduce the number of credit update messages,
 448          * don't update credits as long as lots of space is available.
 449          * Note: the limit chosen here is arbitrary. Setting the limit
 450          * too high causes extra messages. Too low causes transmitter
 451          * stalls. As stalls are in theory more expensive than extra
 452          * messages, we set the limit to a high value. TODO: experiment
 453          * with different values.
 454          */
 455         if (free_space < VIRTIO_VSOCK_MAX_PKT_BUF_SIZE)
 456                 virtio_transport_send_credit_update(vsk);
 457
 458         return total;
 459
 460 out:
 461         if (total)
 462                 err = total;
 463         return err;
 464 }
 465
 466 static int virtio_transport_seqpacket_do_dequeue(struct vsock_sock *vsk,
 467                                                  struct msghdr *msg,
 468                                                  int flags)
 469 {
 470         struct virtio_vsock_sock *vvs = vsk->trans;
 471         int dequeued_len = 0;
 472         size_t user_buf_len = msg_data_left(msg);
 473         bool msg_ready = false;
 474         struct sk_buff *skb;
 475
 476         spin_lock_bh(&vvs->rx_lock);
 477
 478         if (vvs->msg_count == 0) {
 479                 spin_unlock_bh(&vvs->rx_lock);
 480                 return 0;
 481         }
 482
 483         while (!msg_ready) {
 484                 struct virtio_vsock_hdr *hdr;
 485                 size_t pkt_len;
 486
 487                 skb = __skb_dequeue(&vvs->rx_queue);
 488                 if (!skb)
 489                         break;
 490                 hdr = virtio_vsock_hdr(skb);
 491                 pkt_len = (size_t)le32_to_cpu(hdr->len);
 492
 493                 if (dequeued_len >= 0) {
 494                         size_t bytes_to_copy;
 495
 496                         bytes_to_copy = min(user_buf_len, pkt_len);
 497
 498                         if (bytes_to_copy) {
 499                                 int err;
 500
 501                                 /* sk_lock is held by caller so no one else can dequeue.
 502                                  * Unlock rx_lock since memcpy_to_msg() may sleep.
 503                                  */
 504                                 spin_unlock_bh(&vvs->rx_lock);
 505
 506                                 err = memcpy_to_msg(msg, skb->data, bytes_to_copy);
 507                                 if (err) {
 508                                         /* Copy of message failed. Rest of
 509                                          * fragments will be freed without copy.
 510                                          */
 511                                         dequeued_len = err;
 512                                 } else {
 513                                         user_buf_len -= bytes_to_copy;
 514                                 }
 515
 516                                 spin_lock_bh(&vvs->rx_lock);
 517                         }
 518
 519                         if (dequeued_len >= 0)
 520                                 dequeued_len += pkt_len;
 521                 }
 522
 523                 if (le32_to_cpu(hdr->flags) & VIRTIO_VSOCK_SEQ_EOM) {
 524                         msg_ready = true;
 525                         vvs->msg_count--;
 526
 527                         if (le32_to_cpu(hdr->flags) & VIRTIO_VSOCK_SEQ_EOR)
 528                                 msg->msg_flags |= MSG_EOR;
 529                 }
 530
 531                 virtio_transport_dec_rx_pkt(vvs, pkt_len);
 532                 kfree_skb(skb);
 533         }
 534
 535         spin_unlock_bh(&vvs->rx_lock);
 536
 537         virtio_transport_send_credit_update(vsk);
 538
 539         return dequeued_len;
 540 }
 541
 542 ssize_t
 543 virtio_transport_stream_dequeue(struct vsock_sock *vsk,
 544                                 struct msghdr *msg,
 545                                 size_t len, int flags)
 546 {
 547         if (flags & MSG_PEEK)
 548                 return virtio_transport_stream_do_peek(vsk, msg, len);
 549         else
 550                 return virtio_transport_stream_do_dequeue(vsk, msg, len);
 551 }
 552 EXPORT_SYMBOL_GPL(virtio_transport_stream_dequeue);
 553
 554 ssize_t
 555 virtio_transport_seqpacket_dequeue(struct vsock_sock *vsk,
 556                                    struct msghdr *msg,
 557                                    int flags)
 558 {
 559         if (flags & MSG_PEEK)
 560                 return -EOPNOTSUPP;
 561
 562         return virtio_transport_seqpacket_do_dequeue(vsk, msg, flags);
 563 }
 564 EXPORT_SYMBOL_GPL(virtio_transport_seqpacket_dequeue);
 565
 566 int
 567 virtio_transport_seqpacket_enqueue(struct vsock_sock *vsk,
 568                                    struct msghdr *msg,
 569                                    size_t len)
 570 {
 571         struct virtio_vsock_sock *vvs = vsk->trans;
 572
 573         spin_lock_bh(&vvs->tx_lock);
 574
 575         if (len > vvs->peer_buf_alloc) {
 576                 spin_unlock_bh(&vvs->tx_lock);
 577                 return -EMSGSIZE;
 578         }
 579
 580         spin_unlock_bh(&vvs->tx_lock);
 581
 582         return virtio_transport_stream_enqueue(vsk, msg, len);
 583 }
 584 EXPORT_SYMBOL_GPL(virtio_transport_seqpacket_enqueue);
 585
 586 int
 587 virtio_transport_dgram_dequeue(struct vsock_sock *vsk,
 588                                struct msghdr *msg,
 589                                size_t len, int flags)
 590 {
 591         return -EOPNOTSUPP;
 592 }
 593 EXPORT_SYMBOL_GPL(virtio_transport_dgram_dequeue);
 594
 595 s64 virtio_transport_stream_has_data(struct vsock_sock *vsk)
 596 {
 597         struct virtio_vsock_sock *vvs = vsk->trans;
 598         s64 bytes;
 599
 600         spin_lock_bh(&vvs->rx_lock);
 601         bytes = vvs->rx_bytes;
 602         spin_unlock_bh(&vvs->rx_lock);
 603
 604         return bytes;
 605 }
 606 EXPORT_SYMBOL_GPL(virtio_transport_stream_has_data);
 607
 608 u32 virtio_transport_seqpacket_has_data(struct vsock_sock *vsk)
 609 {
 610         struct virtio_vsock_sock *vvs = vsk->trans;
 611         u32 msg_count;
 612
 613         spin_lock_bh(&vvs->rx_lock);
 614         msg_count = vvs->msg_count;
 615         spin_unlock_bh(&vvs->rx_lock);
 616
 617         return msg_count;
 618 }
 619 EXPORT_SYMBOL_GPL(virtio_transport_seqpacket_has_data);
 620
 621 static s64 virtio_transport_has_space(struct vsock_sock *vsk)
 622 {
 623         struct virtio_vsock_sock *vvs = vsk->trans;
 624         s64 bytes;
 625
 626         bytes = vvs->peer_buf_alloc - (vvs->tx_cnt - vvs->peer_fwd_cnt);
 627         if (bytes < 0)
 628                 bytes = 0;
 629
 630         return bytes;
 631 }
 632
 633 s64 virtio_transport_stream_has_space(struct vsock_sock *vsk)
 634 {
 635         struct virtio_vsock_sock *vvs = vsk->trans;
 636         s64 bytes;
 637
 638         spin_lock_bh(&vvs->tx_lock);
 639         bytes = virtio_transport_has_space(vsk);
 640         spin_unlock_bh(&vvs->tx_lock);
 641
 642         return bytes;
 643 }
 644 EXPORT_SYMBOL_GPL(virtio_transport_stream_has_space);
 645
 646 int virtio_transport_do_socket_init(struct vsock_sock *vsk,
 647                                     struct vsock_sock *psk)
 648 {
 649         struct virtio_vsock_sock *vvs;
 650
 651         vvs = kzalloc(sizeof(*vvs), GFP_KERNEL);
 652         if (!vvs)
 653                 return -ENOMEM;
 654
 655         vsk->trans = vvs;
 656         vvs->vsk = vsk;
 657         if (psk && psk->trans) {
 658                 struct virtio_vsock_sock *ptrans = psk->trans;
 659
 660                 vvs->peer_buf_alloc = ptrans->peer_buf_alloc;
 661         }
 662
 663         if (vsk->buffer_size > VIRTIO_VSOCK_MAX_BUF_SIZE)
 664                 vsk->buffer_size = VIRTIO_VSOCK_MAX_BUF_SIZE;
 665
 666         vvs->buf_alloc = vsk->buffer_size;
 667
 668         spin_lock_init(&vvs->rx_lock);
 669         spin_lock_init(&vvs->tx_lock);
 670         skb_queue_head_init(&vvs->rx_queue);
 671
 672         return 0;
 673 }
 674 EXPORT_SYMBOL_GPL(virtio_transport_do_socket_init);
 675
 676 /* sk_lock held by the caller */
 677 void virtio_transport_notify_buffer_size(struct vsock_sock *vsk, u64 *val)
 678 {
 679         struct virtio_vsock_sock *vvs = vsk->trans;
 680
 681         if (*val > VIRTIO_VSOCK_MAX_BUF_SIZE)
 682                 *val = VIRTIO_VSOCK_MAX_BUF_SIZE;
 683
 684         vvs->buf_alloc = *val;
 685
 686         virtio_transport_send_credit_update(vsk);
 687 }
 688 EXPORT_SYMBOL_GPL(virtio_transport_notify_buffer_size);
 689
 690 int
 691 virtio_transport_notify_poll_in(struct vsock_sock *vsk,
 692                                 size_t target,
 693                                 bool *data_ready_now)
 694 {
 695         *data_ready_now = vsock_stream_has_data(vsk) >= target;
 696
 697         return 0;
 698 }
 699 EXPORT_SYMBOL_GPL(virtio_transport_notify_poll_in);
 700
 701 int
 702 virtio_transport_notify_poll_out(struct vsock_sock *vsk,
 703                                  size_t target,
 704                                  bool *space_avail_now)
 705 {
 706         s64 free_space;
 707
 708         free_space = vsock_stream_has_space(vsk);
 709         if (free_space > 0)
 710                 *space_avail_now = true;
 711         else if (free_space == 0)
 712                 *space_avail_now = false;
 713
 714         return 0;
 715 }
 716 EXPORT_SYMBOL_GPL(virtio_transport_notify_poll_out);
 717
 718 int virtio_transport_notify_recv_init(struct vsock_sock *vsk,
 719         size_t target, struct vsock_transport_recv_notify_data *data)
 720 {
 721         return 0;
 722 }
 723 EXPORT_SYMBOL_GPL(virtio_transport_notify_recv_init);
 724
 725 int virtio_transport_notify_recv_pre_block(struct vsock_sock *vsk,
 726         size_t target, struct vsock_transport_recv_notify_data *data)
 727 {
 728         return 0;
 729 }
 730 EXPORT_SYMBOL_GPL(virtio_transport_notify_recv_pre_block);
 731
 732 int virtio_transport_notify_recv_pre_dequeue(struct vsock_sock *vsk,
 733         size_t target, struct vsock_transport_recv_notify_data *data)
 734 {
 735         return 0;
 736 }
 737 EXPORT_SYMBOL_GPL(virtio_transport_notify_recv_pre_dequeue);
 738
 739 int virtio_transport_notify_recv_post_dequeue(struct vsock_sock *vsk,
 740         size_t target, ssize_t copied, bool data_read,
 741         struct vsock_transport_recv_notify_data *data)
 742 {
 743         return 0;
 744 }
 745 EXPORT_SYMBOL_GPL(virtio_transport_notify_recv_post_dequeue);
 746
 747 int virtio_transport_notify_send_init(struct vsock_sock *vsk,
 748         struct vsock_transport_send_notify_data *data)
 749 {
 750         return 0;
 751 }
 752 EXPORT_SYMBOL_GPL(virtio_transport_notify_send_init);
 753
 754 int virtio_transport_notify_send_pre_block(struct vsock_sock *vsk,
 755         struct vsock_transport_send_notify_data *data)
 756 {
 757         return 0;
 758 }
 759 EXPORT_SYMBOL_GPL(virtio_transport_notify_send_pre_block);
 760
 761 int virtio_transport_notify_send_pre_enqueue(struct vsock_sock *vsk,
 762         struct vsock_transport_send_notify_data *data)
 763 {
 764         return 0;
 765 }
 766 EXPORT_SYMBOL_GPL(virtio_transport_notify_send_pre_enqueue);
 767
 768 int virtio_transport_notify_send_post_enqueue(struct vsock_sock *vsk,
 769         ssize_t written, struct vsock_transport_send_notify_data *data)
 770 {
 771         return 0;
 772 }
 773 EXPORT_SYMBOL_GPL(virtio_transport_notify_send_post_enqueue);
 774
 775 u64 virtio_transport_stream_rcvhiwat(struct vsock_sock *vsk)
 776 {
 777         return vsk->buffer_size;
 778 }
 779 EXPORT_SYMBOL_GPL(virtio_transport_stream_rcvhiwat);
 780
 781 bool virtio_transport_stream_is_active(struct vsock_sock *vsk)
 782 {
 783         return true;
 784 }
 785 EXPORT_SYMBOL_GPL(virtio_transport_stream_is_active);
 786
 787 bool virtio_transport_stream_allow(u32 cid, u32 port)
 788 {
 789         return true;
 790 }
 791 EXPORT_SYMBOL_GPL(virtio_transport_stream_allow);
 792
 793 int virtio_transport_dgram_bind(struct vsock_sock *vsk,
 794                                 struct sockaddr_vm *addr)
 795 {
 796         return -EOPNOTSUPP;
 797 }
 798 EXPORT_SYMBOL_GPL(virtio_transport_dgram_bind);
 799
 800 bool virtio_transport_dgram_allow(u32 cid, u32 port)
 801 {
 802         return false;
 803 }
 804 EXPORT_SYMBOL_GPL(virtio_transport_dgram_allow);
 805
 806 int virtio_transport_connect(struct vsock_sock *vsk)
 807 {
 808         struct virtio_vsock_pkt_info info = {
 809                 .op = VIRTIO_VSOCK_OP_REQUEST,
 810                 .vsk = vsk,
 811         };
 812
 813         return virtio_transport_send_pkt_info(vsk, &info);
 814 }
 815 EXPORT_SYMBOL_GPL(virtio_transport_connect);
 816
 817 int virtio_transport_shutdown(struct vsock_sock *vsk, int mode)
 818 {
 819         struct virtio_vsock_pkt_info info = {
 820                 .op = VIRTIO_VSOCK_OP_SHUTDOWN,
 821                 .flags = (mode & RCV_SHUTDOWN ?
 822                           VIRTIO_VSOCK_SHUTDOWN_RCV : 0) |
 823                          (mode & SEND_SHUTDOWN ?
 824                           VIRTIO_VSOCK_SHUTDOWN_SEND : 0),
 825                 .vsk = vsk,
 826         };
 827
 828         return virtio_transport_send_pkt_info(vsk, &info);
 829 }
 830 EXPORT_SYMBOL_GPL(virtio_transport_shutdown);
 831
 832 int
 833 virtio_transport_dgram_enqueue(struct vsock_sock *vsk,
 834                                struct sockaddr_vm *remote_addr,
 835                                struct msghdr *msg,
 836                                size_t dgram_len)
 837 {
 838         return -EOPNOTSUPP;
 839 }
 840 EXPORT_SYMBOL_GPL(virtio_transport_dgram_enqueue);
 841
 842 ssize_t
 843 virtio_transport_stream_enqueue(struct vsock_sock *vsk,
 844                                 struct msghdr *msg,
 845                                 size_t len)
 846 {
 847         struct virtio_vsock_pkt_info info = {
 848                 .op = VIRTIO_VSOCK_OP_RW,
 849                 .msg = msg,
 850                 .pkt_len = len,
 851                 .vsk = vsk,
 852         };
 853
 854         return virtio_transport_send_pkt_info(vsk, &info);
 855 }
 856 EXPORT_SYMBOL_GPL(virtio_transport_stream_enqueue);
 857
 858 void virtio_transport_destruct(struct vsock_sock *vsk)
 859 {
 860         struct virtio_vsock_sock *vvs = vsk->trans;
 861
 862         kfree(vvs);
 863 }
 864 EXPORT_SYMBOL_GPL(virtio_transport_destruct);
 865
 866 static int virtio_transport_reset(struct vsock_sock *vsk,
 867                                   struct sk_buff *skb)
 868 {
 869         struct virtio_vsock_pkt_info info = {
 870                 .op = VIRTIO_VSOCK_OP_RST,
 871                 .reply = !!skb,
 872                 .vsk = vsk,
 873         };
 874
 875         /* Send RST only if the original pkt is not a RST pkt */
 876         if (skb && le16_to_cpu(virtio_vsock_hdr(skb)->op) == VIRTIO_VSOCK_OP_RST)
 877                 return 0;
 878
 879         return virtio_transport_send_pkt_info(vsk, &info);
 880 }
 881
 882 /* Normally packets are associated with a socket.  There may be no socket if an
 883  * attempt was made to connect to a socket that does not exist.
 884  */
 885 static int virtio_transport_reset_no_sock(const struct virtio_transport *t,
 886                                           struct sk_buff *skb)
 887 {
 888         struct virtio_vsock_hdr *hdr = virtio_vsock_hdr(skb);
 889         struct virtio_vsock_pkt_info info = {
 890                 .op = VIRTIO_VSOCK_OP_RST,
 891                 .type = le16_to_cpu(hdr->type),
 892                 .reply = true,
 893         };
 894         struct sk_buff *reply;
 895
 896         /* Send RST only if the original pkt is not a RST pkt */
 897         if (le16_to_cpu(hdr->op) == VIRTIO_VSOCK_OP_RST)
 898                 return 0;
 899
 900         if (!t)
 901                 return -ENOTCONN;
 902
 903         reply = virtio_transport_alloc_skb(&info, 0,
 904                                            le64_to_cpu(hdr->dst_cid),
 905                                            le32_to_cpu(hdr->dst_port),
 906                                            le64_to_cpu(hdr->src_cid),
 907                                            le32_to_cpu(hdr->src_port));
 908         if (!reply)
 909                 return -ENOMEM;
 910
 911         return t->send_pkt(reply);
 912 }
 913
 914 /* This function should be called with sk_lock held and SOCK_DONE set */
 915 static void virtio_transport_remove_sock(struct vsock_sock *vsk)
 916 {
 917         struct virtio_vsock_sock *vvs = vsk->trans;
 918
 919         /* We don't need to take rx_lock, as the socket is closing and we are
 920          * removing it.
 921          */
 922         __skb_queue_purge(&vvs->rx_queue);
 923         vsock_remove_sock(vsk);
 924 }
 925
 926 static void virtio_transport_wait_close(struct sock *sk, long timeout)
 927 {
 928         if (timeout) {
 929                 DEFINE_WAIT_FUNC(wait, woken_wake_function);
 930
 931                 add_wait_queue(sk_sleep(sk), &wait);
 932
 933                 do {
 934                         if (sk_wait_event(sk, &timeout,
 935                                           sock_flag(sk, SOCK_DONE), &wait))
 936                                 break;
 937                 } while (!signal_pending(current) && timeout);
 938
 939                 remove_wait_queue(sk_sleep(sk), &wait);
 940         }
 941 }
 942
 943 static void virtio_transport_do_close(struct vsock_sock *vsk,
 944                                       bool cancel_timeout)
 945 {
 946         struct sock *sk = sk_vsock(vsk);
 947
 948         sock_set_flag(sk, SOCK_DONE);
 949         vsk->peer_shutdown = SHUTDOWN_MASK;
 950         if (vsock_stream_has_data(vsk) <= 0)
 951                 sk->sk_state = TCP_CLOSING;
 952         sk->sk_state_change(sk);
 953
 954         if (vsk->close_work_scheduled &&
 955             (!cancel_timeout || cancel_delayed_work(&vsk->close_work))) {
 956                 vsk->close_work_scheduled = false;
 957
 958                 virtio_transport_remove_sock(vsk);
 959
 960                 /* Release refcnt obtained when we scheduled the timeout */
 961                 sock_put(sk);
 962         }
 963 }
 964
 965 static void virtio_transport_close_timeout(struct work_struct *work)
 966 {
 967         struct vsock_sock *vsk =
 968                 container_of(work, struct vsock_sock, close_work.work);
 969         struct sock *sk = sk_vsock(vsk);
 970
 971         sock_hold(sk);
 972         lock_sock(sk);
 973
 974         if (!sock_flag(sk, SOCK_DONE)) {
 975                 (void)virtio_transport_reset(vsk, NULL);
 976
 977                 virtio_transport_do_close(vsk, false);
 978         }
 979
 980         vsk->close_work_scheduled = false;
 981
 982         release_sock(sk);
 983         sock_put(sk);
 984 }
 985
 986 /* User context, vsk->sk is locked */
 987 static bool virtio_transport_close(struct vsock_sock *vsk)
 988 {
 989         struct sock *sk = &vsk->sk;
 990
 991         if (!(sk->sk_state == TCP_ESTABLISHED ||
 992               sk->sk_state == TCP_CLOSING))
 993                 return true;
 994
 995         /* Already received SHUTDOWN from peer, reply with RST */
 996         if ((vsk->peer_shutdown & SHUTDOWN_MASK) == SHUTDOWN_MASK) {
 997                 (void)virtio_transport_reset(vsk, NULL);
 998                 return true;
 999         }
1000
1001         if ((sk->sk_shutdown & SHUTDOWN_MASK) != SHUTDOWN_MASK)
1002                 (void)virtio_transport_shutdown(vsk, SHUTDOWN_MASK);
1003
1004         if (sock_flag(sk, SOCK_LINGER) && !(current->flags & PF_EXITING))
1005                 virtio_transport_wait_close(sk, sk->sk_lingertime);
1006
1007         if (sock_flag(sk, SOCK_DONE)) {
1008                 return true;
1009         }
1010
1011         sock_hold(sk);
1012         INIT_DELAYED_WORK(&vsk->close_work,
1013                           virtio_transport_close_timeout);
1014         vsk->close_work_scheduled = true;
1015         schedule_delayed_work(&vsk->close_work, VSOCK_CLOSE_TIMEOUT);
1016         return false;
1017 }
1018
1019 void virtio_transport_release(struct vsock_sock *vsk)
1020 {
1021         struct sock *sk = &vsk->sk;
1022         bool remove_sock = true;
1023
1024         if (sk->sk_type == SOCK_STREAM || sk->sk_type == SOCK_SEQPACKET)
1025                 remove_sock = virtio_transport_close(vsk);
1026
1027         if (remove_sock) {
1028                 sock_set_flag(sk, SOCK_DONE);
1029                 virtio_transport_remove_sock(vsk);
1030         }
1031 }
1032 EXPORT_SYMBOL_GPL(virtio_transport_release);
1033
1034 static int
1035 virtio_transport_recv_connecting(struct sock *sk,
1036                                  struct sk_buff *skb)
1037 {
1038         struct virtio_vsock_hdr *hdr = virtio_vsock_hdr(skb);
1039         struct vsock_sock *vsk = vsock_sk(sk);
1040         int skerr;
1041         int err;
1042
1043         switch (le16_to_cpu(hdr->op)) {
1044         case VIRTIO_VSOCK_OP_RESPONSE:
1045                 sk->sk_state = TCP_ESTABLISHED;
1046                 sk->sk_socket->state = SS_CONNECTED;
1047                 vsock_insert_connected(vsk);
1048                 sk->sk_state_change(sk);
1049                 break;
1050         case VIRTIO_VSOCK_OP_INVALID:
1051                 break;
1052         case VIRTIO_VSOCK_OP_RST:
1053                 skerr = ECONNRESET;
1054                 err = 0;
1055                 goto destroy;
1056         default:
1057                 skerr = EPROTO;
1058                 err = -EINVAL;
1059                 goto destroy;
1060         }
1061         return 0;
1062
1063 destroy:
1064         virtio_transport_reset(vsk, skb);
1065         sk->sk_state = TCP_CLOSE;
1066         sk->sk_err = skerr;
1067         sk_error_report(sk);
1068         return err;
1069 }
1070
1071 static void
1072 virtio_transport_recv_enqueue(struct vsock_sock *vsk,
1073                               struct sk_buff *skb)
1074 {
1075         struct virtio_vsock_sock *vvs = vsk->trans;
1076         bool can_enqueue, free_pkt = false;
1077         struct virtio_vsock_hdr *hdr;
1078         u32 len;
1079
1080         hdr = virtio_vsock_hdr(skb);
1081         len = le32_to_cpu(hdr->len);
1082
1083         spin_lock_bh(&vvs->rx_lock);
1084
1085         can_enqueue = virtio_transport_inc_rx_pkt(vvs, len);
1086         if (!can_enqueue) {
1087                 free_pkt = true;
1088                 goto out;
1089         }
1090
1091         if (le32_to_cpu(hdr->flags) & VIRTIO_VSOCK_SEQ_EOM)
1092                 vvs->msg_count++;
1093
1094         /* Try to copy small packets into the buffer of last packet queued,
1095          * to avoid wasting memory queueing the entire buffer with a small
1096          * payload.
1097          */
1098         if (len <= GOOD_COPY_LEN && !skb_queue_empty(&vvs->rx_queue)) {
1099                 struct virtio_vsock_hdr *last_hdr;
1100                 struct sk_buff *last_skb;
1101
1102                 last_skb = skb_peek_tail(&vvs->rx_queue);
1103                 last_hdr = virtio_vsock_hdr(last_skb);
1104
1105                 /* If there is space in the last packet queued, we copy the
1106                  * new packet in its buffer. We avoid this if the last packet
1107                  * queued has VIRTIO_VSOCK_SEQ_EOM set, because this is
1108                  * delimiter of SEQPACKET message, so 'pkt' is the first packet
1109                  * of a new message.
1110                  */
1111                 if (skb->len < skb_tailroom(last_skb) &&
1112                     !(le32_to_cpu(last_hdr->flags) & VIRTIO_VSOCK_SEQ_EOM)) {
1113                         memcpy(skb_put(last_skb, skb->len), skb->data, skb->len);
1114                         free_pkt = true;
1115                         last_hdr->flags |= hdr->flags;
1116                         le32_add_cpu(&last_hdr->len, len);
1117                         goto out;
1118                 }
1119         }
1120
1121         __skb_queue_tail(&vvs->rx_queue, skb);
1122
1123 out:
1124         spin_unlock_bh(&vvs->rx_lock);
1125         if (free_pkt)
1126                 kfree_skb(skb);
1127 }
1128
1129 static int
1130 virtio_transport_recv_connected(struct sock *sk,
1131                                 struct sk_buff *skb)
1132 {
1133         struct virtio_vsock_hdr *hdr = virtio_vsock_hdr(skb);
1134         struct vsock_sock *vsk = vsock_sk(sk);
1135         int err = 0;
1136
1137         switch (le16_to_cpu(hdr->op)) {
1138         case VIRTIO_VSOCK_OP_RW:
1139                 virtio_transport_recv_enqueue(vsk, skb);
1140                 vsock_data_ready(sk);
1141                 return err;
1142         case VIRTIO_VSOCK_OP_CREDIT_REQUEST:
1143                 virtio_transport_send_credit_update(vsk);
1144                 break;
1145         case VIRTIO_VSOCK_OP_CREDIT_UPDATE:
1146                 sk->sk_write_space(sk);
1147                 break;
1148         case VIRTIO_VSOCK_OP_SHUTDOWN:
1149                 if (le32_to_cpu(hdr->flags) & VIRTIO_VSOCK_SHUTDOWN_RCV)
1150                         vsk->peer_shutdown |= RCV_SHUTDOWN;
1151                 if (le32_to_cpu(hdr->flags) & VIRTIO_VSOCK_SHUTDOWN_SEND)
1152                         vsk->peer_shutdown |= SEND_SHUTDOWN;
1153                 if (vsk->peer_shutdown == SHUTDOWN_MASK &&
1154                     vsock_stream_has_data(vsk) <= 0 &&
1155                     !sock_flag(sk, SOCK_DONE)) {
1156                         (void)virtio_transport_reset(vsk, NULL);
1157                         virtio_transport_do_close(vsk, true);
1158                 }
1159                 if (le32_to_cpu(virtio_vsock_hdr(skb)->flags))
1160                         sk->sk_state_change(sk);
1161                 break;
1162         case VIRTIO_VSOCK_OP_RST:
1163                 virtio_transport_do_close(vsk, true);
1164                 break;
1165         default:
1166                 err = -EINVAL;
1167                 break;
1168         }
1169
1170         kfree_skb(skb);
1171         return err;
1172 }
1173
1174 static void
1175 virtio_transport_recv_disconnecting(struct sock *sk,
1176                                     struct sk_buff *skb)
1177 {
1178         struct virtio_vsock_hdr *hdr = virtio_vsock_hdr(skb);
1179         struct vsock_sock *vsk = vsock_sk(sk);
1180
1181         if (le16_to_cpu(hdr->op) == VIRTIO_VSOCK_OP_RST)
1182                 virtio_transport_do_close(vsk, true);
1183 }
1184
1185 static int
1186 virtio_transport_send_response(struct vsock_sock *vsk,
1187                                struct sk_buff *skb)
1188 {
1189         struct virtio_vsock_hdr *hdr = virtio_vsock_hdr(skb);
1190         struct virtio_vsock_pkt_info info = {
1191                 .op = VIRTIO_VSOCK_OP_RESPONSE,
1192                 .remote_cid = le64_to_cpu(hdr->src_cid),
1193                 .remote_port = le32_to_cpu(hdr->src_port),
1194                 .reply = true,
1195                 .vsk = vsk,
1196         };
1197
1198         return virtio_transport_send_pkt_info(vsk, &info);
1199 }
1200
1201 static bool virtio_transport_space_update(struct sock *sk,
1202                                           struct sk_buff *skb)
1203 {
1204         struct virtio_vsock_hdr *hdr = virtio_vsock_hdr(skb);
1205         struct vsock_sock *vsk = vsock_sk(sk);
1206         struct virtio_vsock_sock *vvs = vsk->trans;
1207         bool space_available;
1208
1209         /* Listener sockets are not associated with any transport, so we are
1210          * not able to take the state to see if there is space available in the
1211          * remote peer, but since they are only used to receive requests, we
1212          * can assume that there is always space available in the other peer.
1213          */
1214         if (!vvs)
1215                 return true;
1216
1217         /* buf_alloc and fwd_cnt is always included in the hdr */
1218         spin_lock_bh(&vvs->tx_lock);
1219         vvs->peer_buf_alloc = le32_to_cpu(hdr->buf_alloc);
1220         vvs->peer_fwd_cnt = le32_to_cpu(hdr->fwd_cnt);
1221         space_available = virtio_transport_has_space(vsk);
1222         spin_unlock_bh(&vvs->tx_lock);
1223         return space_available;
1224 }
1225
1226 /* Handle server socket */
1227 static int
1228 virtio_transport_recv_listen(struct sock *sk, struct sk_buff *skb,
1229                              struct virtio_transport *t)
1230 {
1231         struct virtio_vsock_hdr *hdr = virtio_vsock_hdr(skb);
1232         struct vsock_sock *vsk = vsock_sk(sk);
1233         struct vsock_sock *vchild;
1234         struct sock *child;
1235         int ret;
1236
1237         if (le16_to_cpu(hdr->op) != VIRTIO_VSOCK_OP_REQUEST) {
1238                 virtio_transport_reset_no_sock(t, skb);
1239                 return -EINVAL;
1240         }
1241
1242         if (sk_acceptq_is_full(sk)) {
1243                 virtio_transport_reset_no_sock(t, skb);
1244                 return -ENOMEM;
1245         }
1246
1247         child = vsock_create_connected(sk);
1248         if (!child) {
1249                 virtio_transport_reset_no_sock(t, skb);
1250                 return -ENOMEM;
1251         }
1252
1253         sk_acceptq_added(sk);
1254
1255         lock_sock_nested(child, SINGLE_DEPTH_NESTING);
1256
1257         child->sk_state = TCP_ESTABLISHED;
1258
1259         vchild = vsock_sk(child);
1260         vsock_addr_init(&vchild->local_addr, le64_to_cpu(hdr->dst_cid),
1261                         le32_to_cpu(hdr->dst_port));
1262         vsock_addr_init(&vchild->remote_addr, le64_to_cpu(hdr->src_cid),
1263                         le32_to_cpu(hdr->src_port));
1264
1265         ret = vsock_assign_transport(vchild, vsk);
1266         /* Transport assigned (looking at remote_addr) must be the same
1267          * where we received the request.
1268          */
1269         if (ret || vchild->transport != &t->transport) {
1270                 release_sock(child);
1271                 virtio_transport_reset_no_sock(t, skb);
1272                 sock_put(child);
1273                 return ret;
1274         }
1275
1276         if (virtio_transport_space_update(child, skb))
1277                 child->sk_write_space(child);
1278
1279         vsock_insert_connected(vchild);
1280         vsock_enqueue_accept(sk, child);
1281         virtio_transport_send_response(vchild, skb);
1282
1283         release_sock(child);
1284
1285         sk->sk_data_ready(sk);
1286         return 0;
1287 }
1288
1289 static bool virtio_transport_valid_type(u16 type)
1290 {
1291         return (type == VIRTIO_VSOCK_TYPE_STREAM) ||
1292                (type == VIRTIO_VSOCK_TYPE_SEQPACKET);
1293 }
1294
1295 /* We are under the virtio-vsock's vsock->rx_lock or vhost-vsock's vq->mutex
1296  * lock.
1297  */
1298 void virtio_transport_recv_pkt(struct virtio_transport *t,
1299                                struct sk_buff *skb)
1300 {
1301         struct virtio_vsock_hdr *hdr = virtio_vsock_hdr(skb);
1302         struct sockaddr_vm src, dst;
1303         struct vsock_sock *vsk;
1304         struct sock *sk;
1305         bool space_available;
1306
1307         vsock_addr_init(&src, le64_to_cpu(hdr->src_cid),
1308                         le32_to_cpu(hdr->src_port));
1309         vsock_addr_init(&dst, le64_to_cpu(hdr->dst_cid),
1310                         le32_to_cpu(hdr->dst_port));
1311
1312         trace_virtio_transport_recv_pkt(src.svm_cid, src.svm_port,
1313                                         dst.svm_cid, dst.svm_port,
1314                                         le32_to_cpu(hdr->len),
1315                                         le16_to_cpu(hdr->type),
1316                                         le16_to_cpu(hdr->op),
1317                                         le32_to_cpu(hdr->flags),
1318                                         le32_to_cpu(hdr->buf_alloc),
1319                                         le32_to_cpu(hdr->fwd_cnt));
1320
1321         if (!virtio_transport_valid_type(le16_to_cpu(hdr->type))) {
1322                 (void)virtio_transport_reset_no_sock(t, skb);
1323                 goto free_pkt;
1324         }
1325
1326         /* The socket must be in connected or bound table
1327          * otherwise send reset back
1328          */
1329         sk = vsock_find_connected_socket(&src, &dst);
1330         if (!sk) {
1331                 sk = vsock_find_bound_socket(&dst);
1332                 if (!sk) {
1333                         (void)virtio_transport_reset_no_sock(t, skb);
1334                         goto free_pkt;
1335                 }
1336         }
1337
1338         if (virtio_transport_get_type(sk) != le16_to_cpu(hdr->type)) {
1339                 (void)virtio_transport_reset_no_sock(t, skb);
1340                 sock_put(sk);
1341                 goto free_pkt;
1342         }
1343
1344         if (!skb_set_owner_sk_safe(skb, sk)) {
1345                 WARN_ONCE(1, "receiving vsock socket has sk_refcnt == 0\n");
1346                 goto free_pkt;
1347         }
1348
1349         vsk = vsock_sk(sk);
1350
1351         lock_sock(sk);
1352
1353         /* Check if sk has been closed before lock_sock */
1354         if (sock_flag(sk, SOCK_DONE)) {
1355                 (void)virtio_transport_reset_no_sock(t, skb);
1356                 release_sock(sk);
1357                 sock_put(sk);
1358                 goto free_pkt;
1359         }
1360
1361         space_available = virtio_transport_space_update(sk, skb);
1362
1363         /* Update CID in case it has changed after a transport reset event */
1364         if (vsk->local_addr.svm_cid != VMADDR_CID_ANY)
1365                 vsk->local_addr.svm_cid = dst.svm_cid;
1366
1367         if (space_available)
1368                 sk->sk_write_space(sk);
1369
1370         switch (sk->sk_state) {
1371         case TCP_LISTEN:
1372                 virtio_transport_recv_listen(sk, skb, t);
1373                 kfree_skb(skb);
1374                 break;
1375         case TCP_SYN_SENT:
1376                 virtio_transport_recv_connecting(sk, skb);
1377                 kfree_skb(skb);
1378                 break;
1379         case TCP_ESTABLISHED:
1380                 virtio_transport_recv_connected(sk, skb);
1381                 break;
1382         case TCP_CLOSING:
1383                 virtio_transport_recv_disconnecting(sk, skb);
1384                 kfree_skb(skb);
1385                 break;
1386         default:
1387                 (void)virtio_transport_reset_no_sock(t, skb);
1388                 kfree_skb(skb);
1389                 break;
1390         }
1391
1392         release_sock(sk);
1393
1394         /* Release refcnt obtained when we fetched this socket out of the
1395          * bound or connected list.
1396          */
1397         sock_put(sk);
1398         return;
1399
1400 free_pkt:
1401         kfree_skb(skb);
1402 }
1403 EXPORT_SYMBOL_GPL(virtio_transport_recv_pkt);
1404
1405 /* Remove skbs found in a queue that have a vsk that matches.
1406  *
1407  * Each skb is freed.
1408  *
1409  * Returns the count of skbs that were reply packets.
1410  */
1411 int virtio_transport_purge_skbs(void *vsk, struct sk_buff_head *queue)
1412 {
1413         struct sk_buff_head freeme;
1414         struct sk_buff *skb, *tmp;
1415         int cnt = 0;
1416
1417         skb_queue_head_init(&freeme);
1418
1419         spin_lock_bh(&queue->lock);
1420         skb_queue_walk_safe(queue, skb, tmp) {
1421                 if (vsock_sk(skb->sk) != vsk)
1422                         continue;
1423
1424                 __skb_unlink(skb, queue);
1425                 __skb_queue_tail(&freeme, skb);
1426
1427                 if (virtio_vsock_skb_reply(skb))
1428                         cnt++;
1429         }
1430         spin_unlock_bh(&queue->lock);
1431
1432         __skb_queue_purge(&freeme);
1433
1434         return cnt;
1435 }
1436 EXPORT_SYMBOL_GPL(virtio_transport_purge_skbs);
1437
1438 int virtio_transport_read_skb(struct vsock_sock *vsk, skb_read_actor_t recv_actor)
1439 {
1440         struct virtio_vsock_sock *vvs = vsk->trans;
1441         struct sock *sk = sk_vsock(vsk);
1442         struct sk_buff *skb;
1443         int off = 0;
1444         int err;
1445
1446         spin_lock_bh(&vvs->rx_lock);
1447         /* Use __skb_recv_datagram() for race-free handling of the receive. It
1448          * works for types other than dgrams.
1449          */
1450         skb = __skb_recv_datagram(sk, &vvs->rx_queue, MSG_DONTWAIT, &off, &err);
1451         spin_unlock_bh(&vvs->rx_lock);
1452
1453         if (!skb)
1454                 return err;
1455
1456         return recv_actor(sk, skb);
1457 }
1458 EXPORT_SYMBOL_GPL(virtio_transport_read_skb);
1459
1460 MODULE_LICENSE("GPL v2");
1461 MODULE_AUTHOR("Asias He");
1462 MODULE_DESCRIPTION("common code for virtio vsock");