net/rxrpc/recvmsg.c

   1 // SPDX-License-Identifier: GPL-2.0-or-later
   2 /* RxRPC recvmsg() implementation
   3  *
   4  * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
   5  * Written by David Howells (dhowells@redhat.com)
   6  */
   7
   8 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
   9
  10 #include <linux/net.h>
  11 #include <linux/skbuff.h>
  12 #include <linux/export.h>
  13 #include <linux/sched/signal.h>
  14
  15 #include <net/sock.h>
  16 #include <net/af_rxrpc.h>
  17 #include "ar-internal.h"
  18
  19 /*
  20  * Post a call for attention by the socket or kernel service.  Further
  21  * notifications are suppressed by putting recvmsg_link on a dummy queue.
  22  */
  23 void rxrpc_notify_socket(struct rxrpc_call *call)
  24 {
  25         struct rxrpc_sock *rx;
  26         struct sock *sk;
  27
  28         _enter("%d", call->debug_id);
  29
  30         if (!list_empty(&call->recvmsg_link))
  31                 return;
  32
  33         rcu_read_lock();
  34
  35         rx = rcu_dereference(call->socket);
  36         sk = &rx->sk;
  37         if (rx && sk->sk_state < RXRPC_CLOSE) {
  38                 if (call->notify_rx) {
  39                         spin_lock_bh(&call->notify_lock);
  40                         call->notify_rx(sk, call, call->user_call_ID);
  41                         spin_unlock_bh(&call->notify_lock);
  42                 } else {
  43                         write_lock_bh(&rx->recvmsg_lock);
  44                         if (list_empty(&call->recvmsg_link)) {
  45                                 rxrpc_get_call(call, rxrpc_call_got);
  46                                 list_add_tail(&call->recvmsg_link, &rx->recvmsg_q);
  47                         }
  48                         write_unlock_bh(&rx->recvmsg_lock);
  49
  50                         if (!sock_flag(sk, SOCK_DEAD)) {
  51                                 _debug("call %ps", sk->sk_data_ready);
  52                                 sk->sk_data_ready(sk);
  53                         }
  54                 }
  55         }
  56
  57         rcu_read_unlock();
  58         _leave("");
  59 }
  60
  61 /*
  62  * Transition a call to the complete state.
  63  */
  64 bool __rxrpc_set_call_completion(struct rxrpc_call *call,
  65                                  enum rxrpc_call_completion compl,
  66                                  u32 abort_code,
  67                                  int error)
  68 {
  69         if (call->state < RXRPC_CALL_COMPLETE) {
  70                 call->abort_code = abort_code;
  71                 call->error = error;
  72                 call->completion = compl;
  73                 call->state = RXRPC_CALL_COMPLETE;
  74                 trace_rxrpc_call_complete(call);
  75                 wake_up(&call->waitq);
  76                 rxrpc_notify_socket(call);
  77                 return true;
  78         }
  79         return false;
  80 }
  81
  82 bool rxrpc_set_call_completion(struct rxrpc_call *call,
  83                                enum rxrpc_call_completion compl,
  84                                u32 abort_code,
  85                                int error)
  86 {
  87         bool ret = false;
  88
  89         if (call->state < RXRPC_CALL_COMPLETE) {
  90                 write_lock_bh(&call->state_lock);
  91                 ret = __rxrpc_set_call_completion(call, compl, abort_code, error);
  92                 write_unlock_bh(&call->state_lock);
  93         }
  94         return ret;
  95 }
  96
  97 /*
  98  * Record that a call successfully completed.
  99  */
 100 bool __rxrpc_call_completed(struct rxrpc_call *call)
 101 {
 102         return __rxrpc_set_call_completion(call, RXRPC_CALL_SUCCEEDED, 0, 0);
 103 }
 104
 105 bool rxrpc_call_completed(struct rxrpc_call *call)
 106 {
 107         bool ret = false;
 108
 109         if (call->state < RXRPC_CALL_COMPLETE) {
 110                 write_lock_bh(&call->state_lock);
 111                 ret = __rxrpc_call_completed(call);
 112                 write_unlock_bh(&call->state_lock);
 113         }
 114         return ret;
 115 }
 116
 117 /*
 118  * Record that a call is locally aborted.
 119  */
 120 bool __rxrpc_abort_call(const char *why, struct rxrpc_call *call,
 121                         rxrpc_seq_t seq, u32 abort_code, int error)
 122 {
 123         trace_rxrpc_abort(call->debug_id, why, call->cid, call->call_id, seq,
 124                           abort_code, error);
 125         return __rxrpc_set_call_completion(call, RXRPC_CALL_LOCALLY_ABORTED,
 126                                            abort_code, error);
 127 }
 128
 129 bool rxrpc_abort_call(const char *why, struct rxrpc_call *call,
 130                       rxrpc_seq_t seq, u32 abort_code, int error)
 131 {
 132         bool ret;
 133
 134         write_lock_bh(&call->state_lock);
 135         ret = __rxrpc_abort_call(why, call, seq, abort_code, error);
 136         write_unlock_bh(&call->state_lock);
 137         return ret;
 138 }
 139
 140 /*
 141  * Pass a call terminating message to userspace.
 142  */
 143 static int rxrpc_recvmsg_term(struct rxrpc_call *call, struct msghdr *msg)
 144 {
 145         u32 tmp = 0;
 146         int ret;
 147
 148         switch (call->completion) {
 149         case RXRPC_CALL_SUCCEEDED:
 150                 ret = 0;
 151                 if (rxrpc_is_service_call(call))
 152                         ret = put_cmsg(msg, SOL_RXRPC, RXRPC_ACK, 0, &tmp);
 153                 break;
 154         case RXRPC_CALL_REMOTELY_ABORTED:
 155                 tmp = call->abort_code;
 156                 ret = put_cmsg(msg, SOL_RXRPC, RXRPC_ABORT, 4, &tmp);
 157                 break;
 158         case RXRPC_CALL_LOCALLY_ABORTED:
 159                 tmp = call->abort_code;
 160                 ret = put_cmsg(msg, SOL_RXRPC, RXRPC_ABORT, 4, &tmp);
 161                 break;
 162         case RXRPC_CALL_NETWORK_ERROR:
 163                 tmp = -call->error;
 164                 ret = put_cmsg(msg, SOL_RXRPC, RXRPC_NET_ERROR, 4, &tmp);
 165                 break;
 166         case RXRPC_CALL_LOCAL_ERROR:
 167                 tmp = -call->error;
 168                 ret = put_cmsg(msg, SOL_RXRPC, RXRPC_LOCAL_ERROR, 4, &tmp);
 169                 break;
 170         default:
 171                 pr_err("Invalid terminal call state %u\n", call->state);
 172                 BUG();
 173                 break;
 174         }
 175
 176         trace_rxrpc_recvmsg(call, rxrpc_recvmsg_terminal, call->rx_hard_ack,
 177                             call->rx_pkt_offset, call->rx_pkt_len, ret);
 178         return ret;
 179 }
 180
 181 /*
 182  * End the packet reception phase.
 183  */
 184 static void rxrpc_end_rx_phase(struct rxrpc_call *call, rxrpc_serial_t serial)
 185 {
 186         _enter("%d,%s", call->debug_id, rxrpc_call_states[call->state]);
 187
 188         trace_rxrpc_receive(call, rxrpc_receive_end, 0, call->rx_top);
 189         ASSERTCMP(call->rx_hard_ack, ==, call->rx_top);
 190
 191         if (call->state == RXRPC_CALL_CLIENT_RECV_REPLY)
 192                 rxrpc_propose_delay_ACK(call, serial, rxrpc_propose_ack_terminal_ack);
 193
 194         write_lock_bh(&call->state_lock);
 195
 196         switch (call->state) {
 197         case RXRPC_CALL_CLIENT_RECV_REPLY:
 198                 __rxrpc_call_completed(call);
 199                 write_unlock_bh(&call->state_lock);
 200                 break;
 201
 202         case RXRPC_CALL_SERVER_RECV_REQUEST:
 203                 call->state = RXRPC_CALL_SERVER_ACK_REQUEST;
 204                 call->expect_req_by = jiffies + MAX_JIFFY_OFFSET;
 205                 write_unlock_bh(&call->state_lock);
 206                 rxrpc_propose_delay_ACK(call, serial,
 207                                         rxrpc_propose_ack_processing_op);
 208                 break;
 209         default:
 210                 write_unlock_bh(&call->state_lock);
 211                 break;
 212         }
 213 }
 214
 215 /*
 216  * Discard a packet we've used up and advance the Rx window by one.
 217  */
 218 static void rxrpc_rotate_rx_window(struct rxrpc_call *call)
 219 {
 220         struct rxrpc_skb_priv *sp;
 221         struct sk_buff *skb;
 222         rxrpc_serial_t serial;
 223         rxrpc_seq_t hard_ack, top;
 224         bool last = false;
 225         u8 subpacket;
 226         int ix;
 227
 228         _enter("%d", call->debug_id);
 229
 230         hard_ack = call->rx_hard_ack;
 231         top = smp_load_acquire(&call->rx_top);
 232         ASSERT(before(hard_ack, top));
 233
 234         hard_ack++;
 235         ix = hard_ack & RXRPC_RXTX_BUFF_MASK;
 236         skb = call->rxtx_buffer[ix];
 237         rxrpc_see_skb(skb, rxrpc_skb_rotated);
 238         sp = rxrpc_skb(skb);
 239
 240         subpacket = call->rxtx_annotations[ix] & RXRPC_RX_ANNO_SUBPACKET;
 241         serial = sp->hdr.serial + subpacket;
 242
 243         if (subpacket == sp->nr_subpackets - 1 &&
 244             sp->rx_flags & RXRPC_SKB_INCL_LAST)
 245                 last = true;
 246
 247         call->rxtx_buffer[ix] = NULL;
 248         call->rxtx_annotations[ix] = 0;
 249         /* Barrier against rxrpc_input_data(). */
 250         smp_store_release(&call->rx_hard_ack, hard_ack);
 251
 252         rxrpc_free_skb(skb, rxrpc_skb_freed);
 253
 254         trace_rxrpc_receive(call, rxrpc_receive_rotate, serial, hard_ack);
 255         if (last) {
 256                 rxrpc_end_rx_phase(call, serial);
 257         } else {
 258                 /* Check to see if there's an ACK that needs sending. */
 259                 if (atomic_inc_return(&call->ackr_nr_consumed) > 2 &&
 260                     !test_and_set_bit(RXRPC_CALL_IDLE_ACK_PENDING, &call->flags)) {
 261                         rxrpc_send_ACK(call, RXRPC_ACK_IDLE, serial,
 262                                        rxrpc_propose_ack_rotate_rx);
 263                         rxrpc_transmit_ack_packets(call->peer->local);
 264                 }
 265         }
 266 }
 267
 268 /*
 269  * Decrypt and verify a (sub)packet.  The packet's length may be changed due to
 270  * padding, but if this is the case, the packet length will be resident in the
 271  * socket buffer.  Note that we can't modify the master skb info as the skb may
 272  * be the home to multiple subpackets.
 273  */
 274 static int rxrpc_verify_packet(struct rxrpc_call *call, struct sk_buff *skb,
 275                                u8 annotation,
 276                                unsigned int offset, unsigned int len)
 277 {
 278         struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
 279         rxrpc_seq_t seq = sp->hdr.seq;
 280         u16 cksum = sp->hdr.cksum;
 281         u8 subpacket = annotation & RXRPC_RX_ANNO_SUBPACKET;
 282
 283         _enter("");
 284
 285         /* For all but the head jumbo subpacket, the security checksum is in a
 286          * jumbo header immediately prior to the data.
 287          */
 288         if (subpacket > 0) {
 289                 __be16 tmp;
 290                 if (skb_copy_bits(skb, offset - 2, &tmp, 2) < 0)
 291                         BUG();
 292                 cksum = ntohs(tmp);
 293                 seq += subpacket;
 294         }
 295
 296         return call->security->verify_packet(call, skb, offset, len,
 297                                              seq, cksum);
 298 }
 299
 300 /*
 301  * Locate the data within a packet.  This is complicated by:
 302  *
 303  * (1) An skb may contain a jumbo packet - so we have to find the appropriate
 304  *     subpacket.
 305  *
 306  * (2) The (sub)packets may be encrypted and, if so, the encrypted portion
 307  *     contains an extra header which includes the true length of the data,
 308  *     excluding any encrypted padding.
 309  */
 310 static int rxrpc_locate_data(struct rxrpc_call *call, struct sk_buff *skb,
 311                              u8 *_annotation,
 312                              unsigned int *_offset, unsigned int *_len,
 313                              bool *_last)
 314 {
 315         struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
 316         unsigned int offset = sizeof(struct rxrpc_wire_header);
 317         unsigned int len;
 318         bool last = false;
 319         int ret;
 320         u8 annotation = *_annotation;
 321         u8 subpacket = annotation & RXRPC_RX_ANNO_SUBPACKET;
 322
 323         /* Locate the subpacket */
 324         offset += subpacket * RXRPC_JUMBO_SUBPKTLEN;
 325         len = skb->len - offset;
 326         if (subpacket < sp->nr_subpackets - 1)
 327                 len = RXRPC_JUMBO_DATALEN;
 328         else if (sp->rx_flags & RXRPC_SKB_INCL_LAST)
 329                 last = true;
 330
 331         if (!(annotation & RXRPC_RX_ANNO_VERIFIED)) {
 332                 ret = rxrpc_verify_packet(call, skb, annotation, offset, len);
 333                 if (ret < 0)
 334                         return ret;
 335                 *_annotation |= RXRPC_RX_ANNO_VERIFIED;
 336         }
 337
 338         *_offset = offset;
 339         *_len = len;
 340         *_last = last;
 341         call->security->locate_data(call, skb, _offset, _len);
 342         return 0;
 343 }
 344
 345 /*
 346  * Deliver messages to a call.  This keeps processing packets until the buffer
 347  * is filled and we find either more DATA (returns 0) or the end of the DATA
 348  * (returns 1).  If more packets are required, it returns -EAGAIN.
 349  */
 350 static int rxrpc_recvmsg_data(struct socket *sock, struct rxrpc_call *call,
 351                               struct msghdr *msg, struct iov_iter *iter,
 352                               size_t len, int flags, size_t *_offset)
 353 {
 354         struct rxrpc_skb_priv *sp;
 355         struct sk_buff *skb;
 356         rxrpc_serial_t serial;
 357         rxrpc_seq_t hard_ack, top, seq;
 358         size_t remain;
 359         bool rx_pkt_last;
 360         unsigned int rx_pkt_offset, rx_pkt_len;
 361         int ix, copy, ret = -EAGAIN, ret2;
 362
 363         rx_pkt_offset = call->rx_pkt_offset;
 364         rx_pkt_len = call->rx_pkt_len;
 365         rx_pkt_last = call->rx_pkt_last;
 366
 367         if (call->state >= RXRPC_CALL_SERVER_ACK_REQUEST) {
 368                 seq = call->rx_hard_ack;
 369                 ret = 1;
 370                 goto done;
 371         }
 372
 373         /* Barriers against rxrpc_input_data(). */
 374         hard_ack = call->rx_hard_ack;
 375         seq = hard_ack + 1;
 376
 377         while (top = smp_load_acquire(&call->rx_top),
 378                before_eq(seq, top)
 379                ) {
 380                 ix = seq & RXRPC_RXTX_BUFF_MASK;
 381                 skb = call->rxtx_buffer[ix];
 382                 if (!skb) {
 383                         trace_rxrpc_recvmsg(call, rxrpc_recvmsg_hole, seq,
 384                                             rx_pkt_offset, rx_pkt_len, 0);
 385                         rxrpc_transmit_ack_packets(call->peer->local);
 386                         break;
 387                 }
 388                 smp_rmb();
 389                 rxrpc_see_skb(skb, rxrpc_skb_seen);
 390                 sp = rxrpc_skb(skb);
 391
 392                 if (!(flags & MSG_PEEK)) {
 393                         serial = sp->hdr.serial;
 394                         serial += call->rxtx_annotations[ix] & RXRPC_RX_ANNO_SUBPACKET;
 395                         trace_rxrpc_receive(call, rxrpc_receive_front,
 396                                             serial, seq);
 397                 }
 398
 399                 if (msg)
 400                         sock_recv_timestamp(msg, sock->sk, skb);
 401
 402                 if (rx_pkt_offset == 0) {
 403                         ret2 = rxrpc_locate_data(call, skb,
 404                                                  &call->rxtx_annotations[ix],
 405                                                  &rx_pkt_offset, &rx_pkt_len,
 406                                                  &rx_pkt_last);
 407                         trace_rxrpc_recvmsg(call, rxrpc_recvmsg_next, seq,
 408                                             rx_pkt_offset, rx_pkt_len, ret2);
 409                         if (ret2 < 0) {
 410                                 ret = ret2;
 411                                 goto out;
 412                         }
 413                 } else {
 414                         trace_rxrpc_recvmsg(call, rxrpc_recvmsg_cont, seq,
 415                                             rx_pkt_offset, rx_pkt_len, 0);
 416                 }
 417
 418                 /* We have to handle short, empty and used-up DATA packets. */
 419                 remain = len - *_offset;
 420                 copy = rx_pkt_len;
 421                 if (copy > remain)
 422                         copy = remain;
 423                 if (copy > 0) {
 424                         ret2 = skb_copy_datagram_iter(skb, rx_pkt_offset, iter,
 425                                                       copy);
 426                         if (ret2 < 0) {
 427                                 ret = ret2;
 428                                 goto out;
 429                         }
 430
 431                         /* handle piecemeal consumption of data packets */
 432                         rx_pkt_offset += copy;
 433                         rx_pkt_len -= copy;
 434                         *_offset += copy;
 435                 }
 436
 437                 if (rx_pkt_len > 0) {
 438                         trace_rxrpc_recvmsg(call, rxrpc_recvmsg_full, seq,
 439                                             rx_pkt_offset, rx_pkt_len, 0);
 440                         ASSERTCMP(*_offset, ==, len);
 441                         ret = 0;
 442                         break;
 443                 }
 444
 445                 /* The whole packet has been transferred. */
 446                 if (!(flags & MSG_PEEK))
 447                         rxrpc_rotate_rx_window(call);
 448                 rx_pkt_offset = 0;
 449                 rx_pkt_len = 0;
 450
 451                 if (rx_pkt_last) {
 452                         ASSERTCMP(seq, ==, READ_ONCE(call->rx_top));
 453                         ret = 1;
 454                         goto out;
 455                 }
 456
 457                 seq++;
 458         }
 459
 460 out:
 461         if (!(flags & MSG_PEEK)) {
 462                 call->rx_pkt_offset = rx_pkt_offset;
 463                 call->rx_pkt_len = rx_pkt_len;
 464                 call->rx_pkt_last = rx_pkt_last;
 465         }
 466 done:
 467         trace_rxrpc_recvmsg(call, rxrpc_recvmsg_data_return, seq,
 468                             rx_pkt_offset, rx_pkt_len, ret);
 469         if (ret == -EAGAIN)
 470                 set_bit(RXRPC_CALL_RX_UNDERRUN, &call->flags);
 471         return ret;
 472 }
 473
 474 /*
 475  * Receive a message from an RxRPC socket
 476  * - we need to be careful about two or more threads calling recvmsg
 477  *   simultaneously
 478  */
 479 int rxrpc_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
 480                   int flags)
 481 {
 482         struct rxrpc_call *call;
 483         struct rxrpc_sock *rx = rxrpc_sk(sock->sk);
 484         struct list_head *l;
 485         size_t copied = 0;
 486         long timeo;
 487         int ret;
 488
 489         DEFINE_WAIT(wait);
 490
 491         trace_rxrpc_recvmsg(NULL, rxrpc_recvmsg_enter, 0, 0, 0, 0);
 492
 493         if (flags & (MSG_OOB | MSG_TRUNC))
 494                 return -EOPNOTSUPP;
 495
 496         timeo = sock_rcvtimeo(&rx->sk, flags & MSG_DONTWAIT);
 497
 498 try_again:
 499         lock_sock(&rx->sk);
 500
 501         /* Return immediately if a client socket has no outstanding calls */
 502         if (RB_EMPTY_ROOT(&rx->calls) &&
 503             list_empty(&rx->recvmsg_q) &&
 504             rx->sk.sk_state != RXRPC_SERVER_LISTENING) {
 505                 release_sock(&rx->sk);
 506                 return -EAGAIN;
 507         }
 508
 509         if (list_empty(&rx->recvmsg_q)) {
 510                 ret = -EWOULDBLOCK;
 511                 if (timeo == 0) {
 512                         call = NULL;
 513                         goto error_no_call;
 514                 }
 515
 516                 release_sock(&rx->sk);
 517
 518                 /* Wait for something to happen */
 519                 prepare_to_wait_exclusive(sk_sleep(&rx->sk), &wait,
 520                                           TASK_INTERRUPTIBLE);
 521                 ret = sock_error(&rx->sk);
 522                 if (ret)
 523                         goto wait_error;
 524
 525                 if (list_empty(&rx->recvmsg_q)) {
 526                         if (signal_pending(current))
 527                                 goto wait_interrupted;
 528                         trace_rxrpc_recvmsg(NULL, rxrpc_recvmsg_wait,
 529                                             0, 0, 0, 0);
 530                         timeo = schedule_timeout(timeo);
 531                 }
 532                 finish_wait(sk_sleep(&rx->sk), &wait);
 533                 goto try_again;
 534         }
 535
 536         /* Find the next call and dequeue it if we're not just peeking.  If we
 537          * do dequeue it, that comes with a ref that we will need to release.
 538          */
 539         write_lock_bh(&rx->recvmsg_lock);
 540         l = rx->recvmsg_q.next;
 541         call = list_entry(l, struct rxrpc_call, recvmsg_link);
 542         if (!(flags & MSG_PEEK))
 543                 list_del_init(&call->recvmsg_link);
 544         else
 545                 rxrpc_get_call(call, rxrpc_call_got);
 546         write_unlock_bh(&rx->recvmsg_lock);
 547
 548         trace_rxrpc_recvmsg(call, rxrpc_recvmsg_dequeue, 0, 0, 0, 0);
 549
 550         /* We're going to drop the socket lock, so we need to lock the call
 551          * against interference by sendmsg.
 552          */
 553         if (!mutex_trylock(&call->user_mutex)) {
 554                 ret = -EWOULDBLOCK;
 555                 if (flags & MSG_DONTWAIT)
 556                         goto error_requeue_call;
 557                 ret = -ERESTARTSYS;
 558                 if (mutex_lock_interruptible(&call->user_mutex) < 0)
 559                         goto error_requeue_call;
 560         }
 561
 562         release_sock(&rx->sk);
 563
 564         if (test_bit(RXRPC_CALL_RELEASED, &call->flags))
 565                 BUG();
 566
 567         if (test_bit(RXRPC_CALL_HAS_USERID, &call->flags)) {
 568                 if (flags & MSG_CMSG_COMPAT) {
 569                         unsigned int id32 = call->user_call_ID;
 570
 571                         ret = put_cmsg(msg, SOL_RXRPC, RXRPC_USER_CALL_ID,
 572                                        sizeof(unsigned int), &id32);
 573                 } else {
 574                         unsigned long idl = call->user_call_ID;
 575
 576                         ret = put_cmsg(msg, SOL_RXRPC, RXRPC_USER_CALL_ID,
 577                                        sizeof(unsigned long), &idl);
 578                 }
 579                 if (ret < 0)
 580                         goto error_unlock_call;
 581         }
 582
 583         if (msg->msg_name && call->peer) {
 584                 struct sockaddr_rxrpc *srx = msg->msg_name;
 585                 size_t len = sizeof(call->peer->srx);
 586
 587                 memcpy(msg->msg_name, &call->peer->srx, len);
 588                 srx->srx_service = call->service_id;
 589                 msg->msg_namelen = len;
 590         }
 591
 592         switch (READ_ONCE(call->state)) {
 593         case RXRPC_CALL_CLIENT_RECV_REPLY:
 594         case RXRPC_CALL_SERVER_RECV_REQUEST:
 595         case RXRPC_CALL_SERVER_ACK_REQUEST:
 596                 ret = rxrpc_recvmsg_data(sock, call, msg, &msg->msg_iter, len,
 597                                          flags, &copied);
 598                 if (ret == -EAGAIN)
 599                         ret = 0;
 600
 601                 rxrpc_transmit_ack_packets(call->peer->local);
 602                 if (after(call->rx_top, call->rx_hard_ack) &&
 603                     call->rxtx_buffer[(call->rx_hard_ack + 1) & RXRPC_RXTX_BUFF_MASK])
 604                         rxrpc_notify_socket(call);
 605                 break;
 606         default:
 607                 ret = 0;
 608                 break;
 609         }
 610
 611         if (ret < 0)
 612                 goto error_unlock_call;
 613
 614         if (call->state == RXRPC_CALL_COMPLETE) {
 615                 ret = rxrpc_recvmsg_term(call, msg);
 616                 if (ret < 0)
 617                         goto error_unlock_call;
 618                 if (!(flags & MSG_PEEK))
 619                         rxrpc_release_call(rx, call);
 620                 msg->msg_flags |= MSG_EOR;
 621                 ret = 1;
 622         }
 623
 624         if (ret == 0)
 625                 msg->msg_flags |= MSG_MORE;
 626         else
 627                 msg->msg_flags &= ~MSG_MORE;
 628         ret = copied;
 629
 630 error_unlock_call:
 631         mutex_unlock(&call->user_mutex);
 632         rxrpc_put_call(call, rxrpc_call_put);
 633         trace_rxrpc_recvmsg(call, rxrpc_recvmsg_return, 0, 0, 0, ret);
 634         return ret;
 635
 636 error_requeue_call:
 637         if (!(flags & MSG_PEEK)) {
 638                 write_lock_bh(&rx->recvmsg_lock);
 639                 list_add(&call->recvmsg_link, &rx->recvmsg_q);
 640                 write_unlock_bh(&rx->recvmsg_lock);
 641                 trace_rxrpc_recvmsg(call, rxrpc_recvmsg_requeue, 0, 0, 0, 0);
 642         } else {
 643                 rxrpc_put_call(call, rxrpc_call_put);
 644         }
 645 error_no_call:
 646         release_sock(&rx->sk);
 647 error_trace:
 648         trace_rxrpc_recvmsg(call, rxrpc_recvmsg_return, 0, 0, 0, ret);
 649         return ret;
 650
 651 wait_interrupted:
 652         ret = sock_intr_errno(timeo);
 653 wait_error:
 654         finish_wait(sk_sleep(&rx->sk), &wait);
 655         call = NULL;
 656         goto error_trace;
 657 }
 658
 659 /**
 660  * rxrpc_kernel_recv_data - Allow a kernel service to receive data/info
 661  * @sock: The socket that the call exists on
 662  * @call: The call to send data through
 663  * @iter: The buffer to receive into
 664  * @_len: The amount of data we want to receive (decreased on return)
 665  * @want_more: True if more data is expected to be read
 666  * @_abort: Where the abort code is stored if -ECONNABORTED is returned
 667  * @_service: Where to store the actual service ID (may be upgraded)
 668  *
 669  * Allow a kernel service to receive data and pick up information about the
 670  * state of a call.  Returns 0 if got what was asked for and there's more
 671  * available, 1 if we got what was asked for and we're at the end of the data
 672  * and -EAGAIN if we need more data.
 673  *
 674  * Note that we may return -EAGAIN to drain empty packets at the end of the
 675  * data, even if we've already copied over the requested data.
 676  *
 677  * *_abort should also be initialised to 0.
 678  */
 679 int rxrpc_kernel_recv_data(struct socket *sock, struct rxrpc_call *call,
 680                            struct iov_iter *iter, size_t *_len,
 681                            bool want_more, u32 *_abort, u16 *_service)
 682 {
 683         size_t offset = 0;
 684         int ret;
 685
 686         _enter("{%d,%s},%zu,%d",
 687                call->debug_id, rxrpc_call_states[call->state],
 688                *_len, want_more);
 689
 690         ASSERTCMP(call->state, !=, RXRPC_CALL_SERVER_SECURING);
 691
 692         mutex_lock(&call->user_mutex);
 693
 694         switch (READ_ONCE(call->state)) {
 695         case RXRPC_CALL_CLIENT_RECV_REPLY:
 696         case RXRPC_CALL_SERVER_RECV_REQUEST:
 697         case RXRPC_CALL_SERVER_ACK_REQUEST:
 698                 ret = rxrpc_recvmsg_data(sock, call, NULL, iter,
 699                                          *_len, 0, &offset);
 700                 *_len -= offset;
 701                 if (ret < 0)
 702                         goto out;
 703
 704                 /* We can only reach here with a partially full buffer if we
 705                  * have reached the end of the data.  We must otherwise have a
 706                  * full buffer or have been given -EAGAIN.
 707                  */
 708                 if (ret == 1) {
 709                         if (iov_iter_count(iter) > 0)
 710                                 goto short_data;
 711                         if (!want_more)
 712                                 goto read_phase_complete;
 713                         ret = 0;
 714                         goto out;
 715                 }
 716
 717                 if (!want_more)
 718                         goto excess_data;
 719                 goto out;
 720
 721         case RXRPC_CALL_COMPLETE:
 722                 goto call_complete;
 723
 724         default:
 725                 ret = -EINPROGRESS;
 726                 goto out;
 727         }
 728
 729 read_phase_complete:
 730         ret = 1;
 731 out:
 732         rxrpc_transmit_ack_packets(call->peer->local);
 733         if (_service)
 734                 *_service = call->service_id;
 735         mutex_unlock(&call->user_mutex);
 736         _leave(" = %d [%zu,%d]", ret, iov_iter_count(iter), *_abort);
 737         return ret;
 738
 739 short_data:
 740         trace_rxrpc_rx_eproto(call, 0, tracepoint_string("short_data"));
 741         ret = -EBADMSG;
 742         goto out;
 743 excess_data:
 744         trace_rxrpc_rx_eproto(call, 0, tracepoint_string("excess_data"));
 745         ret = -EMSGSIZE;
 746         goto out;
 747 call_complete:
 748         *_abort = call->abort_code;
 749         ret = call->error;
 750         if (call->completion == RXRPC_CALL_SUCCEEDED) {
 751                 ret = 1;
 752                 if (iov_iter_count(iter) > 0)
 753                         ret = -ECONNRESET;
 754         }
 755         goto out;
 756 }
 757 EXPORT_SYMBOL(rxrpc_kernel_recv_data);