net/rds/send.c

   1 /*
   2  * Copyright (c) 2006, 2018 Oracle and/or its affiliates. All rights reserved.
   3  *
   4  * This software is available to you under a choice of one of two
   5  * licenses.  You may choose to be licensed under the terms of the GNU
   6  * General Public License (GPL) Version 2, available from the file
   7  * COPYING in the main directory of this source tree, or the
   8  * OpenIB.org BSD license below:
   9  *
  10  *     Redistribution and use in source and binary forms, with or
  11  *     without modification, are permitted provided that the following
  12  *     conditions are met:
  13  *
  14  *      - Redistributions of source code must retain the above
  15  *        copyright notice, this list of conditions and the following
  16  *        disclaimer.
  17  *
  18  *      - Redistributions in binary form must reproduce the above
  19  *        copyright notice, this list of conditions and the following
  20  *        disclaimer in the documentation and/or other materials
  21  *        provided with the distribution.
  22  *
  23  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
  24  * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
  25  * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
  26  * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
  27  * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
  28  * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
  29  * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  30  * SOFTWARE.
  31  *
  32  */
  33 #include <linux/kernel.h>
  34 #include <linux/moduleparam.h>
  35 #include <linux/gfp.h>
  36 #include <net/sock.h>
  37 #include <linux/in.h>
  38 #include <linux/list.h>
  39 #include <linux/ratelimit.h>
  40 #include <linux/export.h>
  41 #include <linux/sizes.h>
  42
  43 #include "rds.h"
  44
  45 /* When transmitting messages in rds_send_xmit, we need to emerge from
  46  * time to time and briefly release the CPU. Otherwise the softlock watchdog
  47  * will kick our shin.
  48  * Also, it seems fairer to not let one busy connection stall all the
  49  * others.
  50  *
  51  * send_batch_count is the number of times we'll loop in send_xmit. Setting
  52  * it to 0 will restore the old behavior (where we looped until we had
  53  * drained the queue).
  54  */
  55 static int send_batch_count = SZ_1K;
  56 module_param(send_batch_count, int, 0444);
  57 MODULE_PARM_DESC(send_batch_count, " batch factor when working the send queue");
  58
  59 static void rds_send_remove_from_sock(struct list_head *messages, int status);
  60
  61 /*
  62  * Reset the send state.  Callers must ensure that this doesn't race with
  63  * rds_send_xmit().
  64  */
  65 void rds_send_path_reset(struct rds_conn_path *cp)
  66 {
  67         struct rds_message *rm, *tmp;
  68         unsigned long flags;
  69
  70         if (cp->cp_xmit_rm) {
  71                 rm = cp->cp_xmit_rm;
  72                 cp->cp_xmit_rm = NULL;
  73                 /* Tell the user the RDMA op is no longer mapped by the
  74                  * transport. This isn't entirely true (it's flushed out
  75                  * independently) but as the connection is down, there's
  76                  * no ongoing RDMA to/from that memory */
  77                 rds_message_unmapped(rm);
  78                 rds_message_put(rm);
  79         }
  80
  81         cp->cp_xmit_sg = 0;
  82         cp->cp_xmit_hdr_off = 0;
  83         cp->cp_xmit_data_off = 0;
  84         cp->cp_xmit_atomic_sent = 0;
  85         cp->cp_xmit_rdma_sent = 0;
  86         cp->cp_xmit_data_sent = 0;
  87
  88         cp->cp_conn->c_map_queued = 0;
  89
  90         cp->cp_unacked_packets = rds_sysctl_max_unacked_packets;
  91         cp->cp_unacked_bytes = rds_sysctl_max_unacked_bytes;
  92
  93         /* Mark messages as retransmissions, and move them to the send q */
  94         spin_lock_irqsave(&cp->cp_lock, flags);
  95         list_for_each_entry_safe(rm, tmp, &cp->cp_retrans, m_conn_item) {
  96                 set_bit(RDS_MSG_ACK_REQUIRED, &rm->m_flags);
  97                 set_bit(RDS_MSG_RETRANSMITTED, &rm->m_flags);
  98         }
  99         list_splice_init(&cp->cp_retrans, &cp->cp_send_queue);
 100         spin_unlock_irqrestore(&cp->cp_lock, flags);
 101 }
 102 EXPORT_SYMBOL_GPL(rds_send_path_reset);
 103
 104 static int acquire_in_xmit(struct rds_conn_path *cp)
 105 {
 106         return test_and_set_bit(RDS_IN_XMIT, &cp->cp_flags) == 0;
 107 }
 108
 109 static void release_in_xmit(struct rds_conn_path *cp)
 110 {
 111         clear_bit(RDS_IN_XMIT, &cp->cp_flags);
 112         smp_mb__after_atomic();
 113         /*
 114          * We don't use wait_on_bit()/wake_up_bit() because our waking is in a
 115          * hot path and finding waiters is very rare.  We don't want to walk
 116          * the system-wide hashed waitqueue buckets in the fast path only to
 117          * almost never find waiters.
 118          */
 119         if (waitqueue_active(&cp->cp_waitq))
 120                 wake_up_all(&cp->cp_waitq);
 121 }
 122
 123 /*
 124  * We're making the conscious trade-off here to only send one message
 125  * down the connection at a time.
 126  *   Pro:
 127  *      - tx queueing is a simple fifo list
 128  *      - reassembly is optional and easily done by transports per conn
 129  *      - no per flow rx lookup at all, straight to the socket
 130  *      - less per-frag memory and wire overhead
 131  *   Con:
 132  *      - queued acks can be delayed behind large messages
 133  *   Depends:
 134  *      - small message latency is higher behind queued large messages
 135  *      - large message latency isn't starved by intervening small sends
 136  */
 137 int rds_send_xmit(struct rds_conn_path *cp)
 138 {
 139         struct rds_connection *conn = cp->cp_conn;
 140         struct rds_message *rm;
 141         unsigned long flags;
 142         unsigned int tmp;
 143         struct scatterlist *sg;
 144         int ret = 0;
 145         LIST_HEAD(to_be_dropped);
 146         int batch_count;
 147         unsigned long send_gen = 0;
 148
 149 restart:
 150         batch_count = 0;
 151
 152         /*
 153          * sendmsg calls here after having queued its message on the send
 154          * queue.  We only have one task feeding the connection at a time.  If
 155          * another thread is already feeding the queue then we back off.  This
 156          * avoids blocking the caller and trading per-connection data between
 157          * caches per message.
 158          */
 159         if (!acquire_in_xmit(cp)) {
 160                 rds_stats_inc(s_send_lock_contention);
 161                 ret = -ENOMEM;
 162                 goto out;
 163         }
 164
 165         if (rds_destroy_pending(cp->cp_conn)) {
 166                 release_in_xmit(cp);
 167                 ret = -ENETUNREACH; /* dont requeue send work */
 168                 goto out;
 169         }
 170
 171         /*
 172          * we record the send generation after doing the xmit acquire.
 173          * if someone else manages to jump in and do some work, we'll use
 174          * this to avoid a goto restart farther down.
 175          *
 176          * The acquire_in_xmit() check above ensures that only one
 177          * caller can increment c_send_gen at any time.
 178          */
 179         send_gen = READ_ONCE(cp->cp_send_gen) + 1;
 180         WRITE_ONCE(cp->cp_send_gen, send_gen);
 181
 182         /*
 183          * rds_conn_shutdown() sets the conn state and then tests RDS_IN_XMIT,
 184          * we do the opposite to avoid races.
 185          */
 186         if (!rds_conn_path_up(cp)) {
 187                 release_in_xmit(cp);
 188                 ret = 0;
 189                 goto out;
 190         }
 191
 192         if (conn->c_trans->xmit_path_prepare)
 193                 conn->c_trans->xmit_path_prepare(cp);
 194
 195         /*
 196          * spin trying to push headers and data down the connection until
 197          * the connection doesn't make forward progress.
 198          */
 199         while (1) {
 200
 201                 rm = cp->cp_xmit_rm;
 202
 203                 /*
 204                  * If between sending messages, we can send a pending congestion
 205                  * map update.
 206                  */
 207                 if (!rm && test_and_clear_bit(0, &conn->c_map_queued)) {
 208                         rm = rds_cong_update_alloc(conn);
 209                         if (IS_ERR(rm)) {
 210                                 ret = PTR_ERR(rm);
 211                                 break;
 212                         }
 213                         rm->data.op_active = 1;
 214                         rm->m_inc.i_conn_path = cp;
 215                         rm->m_inc.i_conn = cp->cp_conn;
 216
 217                         cp->cp_xmit_rm = rm;
 218                 }
 219
 220                 /*
 221                  * If not already working on one, grab the next message.
 222                  *
 223                  * cp_xmit_rm holds a ref while we're sending this message down
 224                  * the connction.  We can use this ref while holding the
 225                  * send_sem.. rds_send_reset() is serialized with it.
 226                  */
 227                 if (!rm) {
 228                         unsigned int len;
 229
 230                         batch_count++;
 231
 232                         /* we want to process as big a batch as we can, but
 233                          * we also want to avoid softlockups.  If we've been
 234                          * through a lot of messages, lets back off and see
 235                          * if anyone else jumps in
 236                          */
 237                         if (batch_count >= send_batch_count)
 238                                 goto over_batch;
 239
 240                         spin_lock_irqsave(&cp->cp_lock, flags);
 241
 242                         if (!list_empty(&cp->cp_send_queue)) {
 243                                 rm = list_entry(cp->cp_send_queue.next,
 244                                                 struct rds_message,
 245                                                 m_conn_item);
 246                                 rds_message_addref(rm);
 247
 248                                 /*
 249                                  * Move the message from the send queue to the retransmit
 250                                  * list right away.
 251                                  */
 252                                 list_move_tail(&rm->m_conn_item,
 253                                                &cp->cp_retrans);
 254                         }
 255
 256                         spin_unlock_irqrestore(&cp->cp_lock, flags);
 257
 258                         if (!rm)
 259                                 break;
 260
 261                         /* Unfortunately, the way Infiniband deals with
 262                          * RDMA to a bad MR key is by moving the entire
 263                          * queue pair to error state. We cold possibly
 264                          * recover from that, but right now we drop the
 265                          * connection.
 266                          * Therefore, we never retransmit messages with RDMA ops.
 267                          */
 268                         if (test_bit(RDS_MSG_FLUSH, &rm->m_flags) ||
 269                             (rm->rdma.op_active &&
 270                             test_bit(RDS_MSG_RETRANSMITTED, &rm->m_flags))) {
 271                                 spin_lock_irqsave(&cp->cp_lock, flags);
 272                                 if (test_and_clear_bit(RDS_MSG_ON_CONN, &rm->m_flags))
 273                                         list_move(&rm->m_conn_item, &to_be_dropped);
 274                                 spin_unlock_irqrestore(&cp->cp_lock, flags);
 275                                 continue;
 276                         }
 277
 278                         /* Require an ACK every once in a while */
 279                         len = ntohl(rm->m_inc.i_hdr.h_len);
 280                         if (cp->cp_unacked_packets == 0 ||
 281                             cp->cp_unacked_bytes < len) {
 282                                 set_bit(RDS_MSG_ACK_REQUIRED, &rm->m_flags);
 283
 284                                 cp->cp_unacked_packets =
 285                                         rds_sysctl_max_unacked_packets;
 286                                 cp->cp_unacked_bytes =
 287                                         rds_sysctl_max_unacked_bytes;
 288                                 rds_stats_inc(s_send_ack_required);
 289                         } else {
 290                                 cp->cp_unacked_bytes -= len;
 291                                 cp->cp_unacked_packets--;
 292                         }
 293
 294                         cp->cp_xmit_rm = rm;
 295                 }
 296
 297                 /* The transport either sends the whole rdma or none of it */
 298                 if (rm->rdma.op_active && !cp->cp_xmit_rdma_sent) {
 299                         rm->m_final_op = &rm->rdma;
 300                         /* The transport owns the mapped memory for now.
 301                          * You can't unmap it while it's on the send queue
 302                          */
 303                         set_bit(RDS_MSG_MAPPED, &rm->m_flags);
 304                         ret = conn->c_trans->xmit_rdma(conn, &rm->rdma);
 305                         if (ret) {
 306                                 clear_bit(RDS_MSG_MAPPED, &rm->m_flags);
 307                                 wake_up_interruptible(&rm->m_flush_wait);
 308                                 break;
 309                         }
 310                         cp->cp_xmit_rdma_sent = 1;
 311
 312                 }
 313
 314                 if (rm->atomic.op_active && !cp->cp_xmit_atomic_sent) {
 315                         rm->m_final_op = &rm->atomic;
 316                         /* The transport owns the mapped memory for now.
 317                          * You can't unmap it while it's on the send queue
 318                          */
 319                         set_bit(RDS_MSG_MAPPED, &rm->m_flags);
 320                         ret = conn->c_trans->xmit_atomic(conn, &rm->atomic);
 321                         if (ret) {
 322                                 clear_bit(RDS_MSG_MAPPED, &rm->m_flags);
 323                                 wake_up_interruptible(&rm->m_flush_wait);
 324                                 break;
 325                         }
 326                         cp->cp_xmit_atomic_sent = 1;
 327
 328                 }
 329
 330                 /*
 331                  * A number of cases require an RDS header to be sent
 332                  * even if there is no data.
 333                  * We permit 0-byte sends; rds-ping depends on this.
 334                  * However, if there are exclusively attached silent ops,
 335                  * we skip the hdr/data send, to enable silent operation.
 336                  */
 337                 if (rm->data.op_nents == 0) {
 338                         int ops_present;
 339                         int all_ops_are_silent = 1;
 340
 341                         ops_present = (rm->atomic.op_active || rm->rdma.op_active);
 342                         if (rm->atomic.op_active && !rm->atomic.op_silent)
 343                                 all_ops_are_silent = 0;
 344                         if (rm->rdma.op_active && !rm->rdma.op_silent)
 345                                 all_ops_are_silent = 0;
 346
 347                         if (ops_present && all_ops_are_silent
 348                             && !rm->m_rdma_cookie)
 349                                 rm->data.op_active = 0;
 350                 }
 351
 352                 if (rm->data.op_active && !cp->cp_xmit_data_sent) {
 353                         rm->m_final_op = &rm->data;
 354
 355                         ret = conn->c_trans->xmit(conn, rm,
 356                                                   cp->cp_xmit_hdr_off,
 357                                                   cp->cp_xmit_sg,
 358                                                   cp->cp_xmit_data_off);
 359                         if (ret <= 0)
 360                                 break;
 361
 362                         if (cp->cp_xmit_hdr_off < sizeof(struct rds_header)) {
 363                                 tmp = min_t(int, ret,
 364                                             sizeof(struct rds_header) -
 365                                             cp->cp_xmit_hdr_off);
 366                                 cp->cp_xmit_hdr_off += tmp;
 367                                 ret -= tmp;
 368                         }
 369
 370                         sg = &rm->data.op_sg[cp->cp_xmit_sg];
 371                         while (ret) {
 372                                 tmp = min_t(int, ret, sg->length -
 373                                                       cp->cp_xmit_data_off);
 374                                 cp->cp_xmit_data_off += tmp;
 375                                 ret -= tmp;
 376                                 if (cp->cp_xmit_data_off == sg->length) {
 377                                         cp->cp_xmit_data_off = 0;
 378                                         sg++;
 379                                         cp->cp_xmit_sg++;
 380                                         BUG_ON(ret != 0 && cp->cp_xmit_sg ==
 381                                                rm->data.op_nents);
 382                                 }
 383                         }
 384
 385                         if (cp->cp_xmit_hdr_off == sizeof(struct rds_header) &&
 386                             (cp->cp_xmit_sg == rm->data.op_nents))
 387                                 cp->cp_xmit_data_sent = 1;
 388                 }
 389
 390                 /*
 391                  * A rm will only take multiple times through this loop
 392                  * if there is a data op. Thus, if the data is sent (or there was
 393                  * none), then we're done with the rm.
 394                  */
 395                 if (!rm->data.op_active || cp->cp_xmit_data_sent) {
 396                         cp->cp_xmit_rm = NULL;
 397                         cp->cp_xmit_sg = 0;
 398                         cp->cp_xmit_hdr_off = 0;
 399                         cp->cp_xmit_data_off = 0;
 400                         cp->cp_xmit_rdma_sent = 0;
 401                         cp->cp_xmit_atomic_sent = 0;
 402                         cp->cp_xmit_data_sent = 0;
 403
 404                         rds_message_put(rm);
 405                 }
 406         }
 407
 408 over_batch:
 409         if (conn->c_trans->xmit_path_complete)
 410                 conn->c_trans->xmit_path_complete(cp);
 411         release_in_xmit(cp);
 412
 413         /* Nuke any messages we decided not to retransmit. */
 414         if (!list_empty(&to_be_dropped)) {
 415                 /* irqs on here, so we can put(), unlike above */
 416                 list_for_each_entry(rm, &to_be_dropped, m_conn_item)
 417                         rds_message_put(rm);
 418                 rds_send_remove_from_sock(&to_be_dropped, RDS_RDMA_DROPPED);
 419         }
 420
 421         /*
 422          * Other senders can queue a message after we last test the send queue
 423          * but before we clear RDS_IN_XMIT.  In that case they'd back off and
 424          * not try and send their newly queued message.  We need to check the
 425          * send queue after having cleared RDS_IN_XMIT so that their message
 426          * doesn't get stuck on the send queue.
 427          *
 428          * If the transport cannot continue (i.e ret != 0), then it must
 429          * call us when more room is available, such as from the tx
 430          * completion handler.
 431          *
 432          * We have an extra generation check here so that if someone manages
 433          * to jump in after our release_in_xmit, we'll see that they have done
 434          * some work and we will skip our goto
 435          */
 436         if (ret == 0) {
 437                 bool raced;
 438
 439                 smp_mb();
 440                 raced = send_gen != READ_ONCE(cp->cp_send_gen);
 441
 442                 if ((test_bit(0, &conn->c_map_queued) ||
 443                     !list_empty(&cp->cp_send_queue)) && !raced) {
 444                         if (batch_count < send_batch_count)
 445                                 goto restart;
 446                         rcu_read_lock();
 447                         if (rds_destroy_pending(cp->cp_conn))
 448                                 ret = -ENETUNREACH;
 449                         else
 450                                 queue_delayed_work(rds_wq, &cp->cp_send_w, 1);
 451                         rcu_read_unlock();
 452                 } else if (raced) {
 453                         rds_stats_inc(s_send_lock_queue_raced);
 454                 }
 455         }
 456 out:
 457         return ret;
 458 }
 459 EXPORT_SYMBOL_GPL(rds_send_xmit);
 460
 461 static void rds_send_sndbuf_remove(struct rds_sock *rs, struct rds_message *rm)
 462 {
 463         u32 len = be32_to_cpu(rm->m_inc.i_hdr.h_len);
 464
 465         assert_spin_locked(&rs->rs_lock);
 466
 467         BUG_ON(rs->rs_snd_bytes < len);
 468         rs->rs_snd_bytes -= len;
 469
 470         if (rs->rs_snd_bytes == 0)
 471                 rds_stats_inc(s_send_queue_empty);
 472 }
 473
 474 static inline int rds_send_is_acked(struct rds_message *rm, u64 ack,
 475                                     is_acked_func is_acked)
 476 {
 477         if (is_acked)
 478                 return is_acked(rm, ack);
 479         return be64_to_cpu(rm->m_inc.i_hdr.h_sequence) <= ack;
 480 }
 481
 482 /*
 483  * This is pretty similar to what happens below in the ACK
 484  * handling code - except that we call here as soon as we get
 485  * the IB send completion on the RDMA op and the accompanying
 486  * message.
 487  */
 488 void rds_rdma_send_complete(struct rds_message *rm, int status)
 489 {
 490         struct rds_sock *rs = NULL;
 491         struct rm_rdma_op *ro;
 492         struct rds_notifier *notifier;
 493         unsigned long flags;
 494         unsigned int notify = 0;
 495
 496         spin_lock_irqsave(&rm->m_rs_lock, flags);
 497
 498         notify =  rm->rdma.op_notify | rm->data.op_notify;
 499         ro = &rm->rdma;
 500         if (test_bit(RDS_MSG_ON_SOCK, &rm->m_flags) &&
 501             ro->op_active && notify && ro->op_notifier) {
 502                 notifier = ro->op_notifier;
 503                 rs = rm->m_rs;
 504                 sock_hold(rds_rs_to_sk(rs));
 505
 506                 notifier->n_status = status;
 507                 spin_lock(&rs->rs_lock);
 508                 list_add_tail(&notifier->n_list, &rs->rs_notify_queue);
 509                 spin_unlock(&rs->rs_lock);
 510
 511                 ro->op_notifier = NULL;
 512         }
 513
 514         spin_unlock_irqrestore(&rm->m_rs_lock, flags);
 515
 516         if (rs) {
 517                 rds_wake_sk_sleep(rs);
 518                 sock_put(rds_rs_to_sk(rs));
 519         }
 520 }
 521 EXPORT_SYMBOL_GPL(rds_rdma_send_complete);
 522
 523 /*
 524  * Just like above, except looks at atomic op
 525  */
 526 void rds_atomic_send_complete(struct rds_message *rm, int status)
 527 {
 528         struct rds_sock *rs = NULL;
 529         struct rm_atomic_op *ao;
 530         struct rds_notifier *notifier;
 531         unsigned long flags;
 532
 533         spin_lock_irqsave(&rm->m_rs_lock, flags);
 534
 535         ao = &rm->atomic;
 536         if (test_bit(RDS_MSG_ON_SOCK, &rm->m_flags)
 537             && ao->op_active && ao->op_notify && ao->op_notifier) {
 538                 notifier = ao->op_notifier;
 539                 rs = rm->m_rs;
 540                 sock_hold(rds_rs_to_sk(rs));
 541
 542                 notifier->n_status = status;
 543                 spin_lock(&rs->rs_lock);
 544                 list_add_tail(&notifier->n_list, &rs->rs_notify_queue);
 545                 spin_unlock(&rs->rs_lock);
 546
 547                 ao->op_notifier = NULL;
 548         }
 549
 550         spin_unlock_irqrestore(&rm->m_rs_lock, flags);
 551
 552         if (rs) {
 553                 rds_wake_sk_sleep(rs);
 554                 sock_put(rds_rs_to_sk(rs));
 555         }
 556 }
 557 EXPORT_SYMBOL_GPL(rds_atomic_send_complete);
 558
 559 /*
 560  * This is the same as rds_rdma_send_complete except we
 561  * don't do any locking - we have all the ingredients (message,
 562  * socket, socket lock) and can just move the notifier.
 563  */
 564 static inline void
 565 __rds_send_complete(struct rds_sock *rs, struct rds_message *rm, int status)
 566 {
 567         struct rm_rdma_op *ro;
 568         struct rm_atomic_op *ao;
 569
 570         ro = &rm->rdma;
 571         if (ro->op_active && ro->op_notify && ro->op_notifier) {
 572                 ro->op_notifier->n_status = status;
 573                 list_add_tail(&ro->op_notifier->n_list, &rs->rs_notify_queue);
 574                 ro->op_notifier = NULL;
 575         }
 576
 577         ao = &rm->atomic;
 578         if (ao->op_active && ao->op_notify && ao->op_notifier) {
 579                 ao->op_notifier->n_status = status;
 580                 list_add_tail(&ao->op_notifier->n_list, &rs->rs_notify_queue);
 581                 ao->op_notifier = NULL;
 582         }
 583
 584         /* No need to wake the app - caller does this */
 585 }
 586
 587 /*
 588  * This removes messages from the socket's list if they're on it.  The list
 589  * argument must be private to the caller, we must be able to modify it
 590  * without locks.  The messages must have a reference held for their
 591  * position on the list.  This function will drop that reference after
 592  * removing the messages from the 'messages' list regardless of if it found
 593  * the messages on the socket list or not.
 594  */
 595 static void rds_send_remove_from_sock(struct list_head *messages, int status)
 596 {
 597         unsigned long flags;
 598         struct rds_sock *rs = NULL;
 599         struct rds_message *rm;
 600
 601         while (!list_empty(messages)) {
 602                 int was_on_sock = 0;
 603
 604                 rm = list_entry(messages->next, struct rds_message,
 605                                 m_conn_item);
 606                 list_del_init(&rm->m_conn_item);
 607
 608                 /*
 609                  * If we see this flag cleared then we're *sure* that someone
 610                  * else beat us to removing it from the sock.  If we race
 611                  * with their flag update we'll get the lock and then really
 612                  * see that the flag has been cleared.
 613                  *
 614                  * The message spinlock makes sure nobody clears rm->m_rs
 615                  * while we're messing with it. It does not prevent the
 616                  * message from being removed from the socket, though.
 617                  */
 618                 spin_lock_irqsave(&rm->m_rs_lock, flags);
 619                 if (!test_bit(RDS_MSG_ON_SOCK, &rm->m_flags))
 620                         goto unlock_and_drop;
 621
 622                 if (rs != rm->m_rs) {
 623                         if (rs) {
 624                                 rds_wake_sk_sleep(rs);
 625                                 sock_put(rds_rs_to_sk(rs));
 626                         }
 627                         rs = rm->m_rs;
 628                         if (rs)
 629                                 sock_hold(rds_rs_to_sk(rs));
 630                 }
 631                 if (!rs)
 632                         goto unlock_and_drop;
 633                 spin_lock(&rs->rs_lock);
 634
 635                 if (test_and_clear_bit(RDS_MSG_ON_SOCK, &rm->m_flags)) {
 636                         struct rm_rdma_op *ro = &rm->rdma;
 637                         struct rds_notifier *notifier;
 638
 639                         list_del_init(&rm->m_sock_item);
 640                         rds_send_sndbuf_remove(rs, rm);
 641
 642                         if (ro->op_active && ro->op_notifier &&
 643                                (ro->op_notify || (ro->op_recverr && status))) {
 644                                 notifier = ro->op_notifier;
 645                                 list_add_tail(&notifier->n_list,
 646                                                 &rs->rs_notify_queue);
 647                                 if (!notifier->n_status)
 648                                         notifier->n_status = status;
 649                                 rm->rdma.op_notifier = NULL;
 650                         }
 651                         was_on_sock = 1;
 652                 }
 653                 spin_unlock(&rs->rs_lock);
 654
 655 unlock_and_drop:
 656                 spin_unlock_irqrestore(&rm->m_rs_lock, flags);
 657                 rds_message_put(rm);
 658                 if (was_on_sock)
 659                         rds_message_put(rm);
 660         }
 661
 662         if (rs) {
 663                 rds_wake_sk_sleep(rs);
 664                 sock_put(rds_rs_to_sk(rs));
 665         }
 666 }
 667
 668 /*
 669  * Transports call here when they've determined that the receiver queued
 670  * messages up to, and including, the given sequence number.  Messages are
 671  * moved to the retrans queue when rds_send_xmit picks them off the send
 672  * queue. This means that in the TCP case, the message may not have been
 673  * assigned the m_ack_seq yet - but that's fine as long as tcp_is_acked
 674  * checks the RDS_MSG_HAS_ACK_SEQ bit.
 675  */
 676 void rds_send_path_drop_acked(struct rds_conn_path *cp, u64 ack,
 677                               is_acked_func is_acked)
 678 {
 679         struct rds_message *rm, *tmp;
 680         unsigned long flags;
 681         LIST_HEAD(list);
 682
 683         spin_lock_irqsave(&cp->cp_lock, flags);
 684
 685         list_for_each_entry_safe(rm, tmp, &cp->cp_retrans, m_conn_item) {
 686                 if (!rds_send_is_acked(rm, ack, is_acked))
 687                         break;
 688
 689                 list_move(&rm->m_conn_item, &list);
 690                 clear_bit(RDS_MSG_ON_CONN, &rm->m_flags);
 691         }
 692
 693         /* order flag updates with spin locks */
 694         if (!list_empty(&list))
 695                 smp_mb__after_atomic();
 696
 697         spin_unlock_irqrestore(&cp->cp_lock, flags);
 698
 699         /* now remove the messages from the sock list as needed */
 700         rds_send_remove_from_sock(&list, RDS_RDMA_SUCCESS);
 701 }
 702 EXPORT_SYMBOL_GPL(rds_send_path_drop_acked);
 703
 704 void rds_send_drop_acked(struct rds_connection *conn, u64 ack,
 705                          is_acked_func is_acked)
 706 {
 707         WARN_ON(conn->c_trans->t_mp_capable);
 708         rds_send_path_drop_acked(&conn->c_path[0], ack, is_acked);
 709 }
 710 EXPORT_SYMBOL_GPL(rds_send_drop_acked);
 711
 712 void rds_send_drop_to(struct rds_sock *rs, struct sockaddr_in *dest)
 713 {
 714         struct rds_message *rm, *tmp;
 715         struct rds_connection *conn;
 716         struct rds_conn_path *cp;
 717         unsigned long flags;
 718         LIST_HEAD(list);
 719
 720         /* get all the messages we're dropping under the rs lock */
 721         spin_lock_irqsave(&rs->rs_lock, flags);
 722
 723         list_for_each_entry_safe(rm, tmp, &rs->rs_send_queue, m_sock_item) {
 724                 if (dest && (dest->sin_addr.s_addr != rm->m_daddr ||
 725                              dest->sin_port != rm->m_inc.i_hdr.h_dport))
 726                         continue;
 727
 728                 list_move(&rm->m_sock_item, &list);
 729                 rds_send_sndbuf_remove(rs, rm);
 730                 clear_bit(RDS_MSG_ON_SOCK, &rm->m_flags);
 731         }
 732
 733         /* order flag updates with the rs lock */
 734         smp_mb__after_atomic();
 735
 736         spin_unlock_irqrestore(&rs->rs_lock, flags);
 737
 738         if (list_empty(&list))
 739                 return;
 740
 741         /* Remove the messages from the conn */
 742         list_for_each_entry(rm, &list, m_sock_item) {
 743
 744                 conn = rm->m_inc.i_conn;
 745                 if (conn->c_trans->t_mp_capable)
 746                         cp = rm->m_inc.i_conn_path;
 747                 else
 748                         cp = &conn->c_path[0];
 749
 750                 spin_lock_irqsave(&cp->cp_lock, flags);
 751                 /*
 752                  * Maybe someone else beat us to removing rm from the conn.
 753                  * If we race with their flag update we'll get the lock and
 754                  * then really see that the flag has been cleared.
 755                  */
 756                 if (!test_and_clear_bit(RDS_MSG_ON_CONN, &rm->m_flags)) {
 757                         spin_unlock_irqrestore(&cp->cp_lock, flags);
 758                         continue;
 759                 }
 760                 list_del_init(&rm->m_conn_item);
 761                 spin_unlock_irqrestore(&cp->cp_lock, flags);
 762
 763                 /*
 764                  * Couldn't grab m_rs_lock in top loop (lock ordering),
 765                  * but we can now.
 766                  */
 767                 spin_lock_irqsave(&rm->m_rs_lock, flags);
 768
 769                 spin_lock(&rs->rs_lock);
 770                 __rds_send_complete(rs, rm, RDS_RDMA_CANCELED);
 771                 spin_unlock(&rs->rs_lock);
 772
 773                 spin_unlock_irqrestore(&rm->m_rs_lock, flags);
 774
 775                 rds_message_put(rm);
 776         }
 777
 778         rds_wake_sk_sleep(rs);
 779
 780         while (!list_empty(&list)) {
 781                 rm = list_entry(list.next, struct rds_message, m_sock_item);
 782                 list_del_init(&rm->m_sock_item);
 783                 rds_message_wait(rm);
 784
 785                 /* just in case the code above skipped this message
 786                  * because RDS_MSG_ON_CONN wasn't set, run it again here
 787                  * taking m_rs_lock is the only thing that keeps us
 788                  * from racing with ack processing.
 789                  */
 790                 spin_lock_irqsave(&rm->m_rs_lock, flags);
 791
 792                 spin_lock(&rs->rs_lock);
 793                 __rds_send_complete(rs, rm, RDS_RDMA_CANCELED);
 794                 spin_unlock(&rs->rs_lock);
 795
 796                 spin_unlock_irqrestore(&rm->m_rs_lock, flags);
 797
 798                 rds_message_put(rm);
 799         }
 800 }
 801
 802 /*
 803  * we only want this to fire once so we use the callers 'queued'.  It's
 804  * possible that another thread can race with us and remove the
 805  * message from the flow with RDS_CANCEL_SENT_TO.
 806  */
 807 static int rds_send_queue_rm(struct rds_sock *rs, struct rds_connection *conn,
 808                              struct rds_conn_path *cp,
 809                              struct rds_message *rm, __be16 sport,
 810                              __be16 dport, int *queued)
 811 {
 812         unsigned long flags;
 813         u32 len;
 814
 815         if (*queued)
 816                 goto out;
 817
 818         len = be32_to_cpu(rm->m_inc.i_hdr.h_len);
 819
 820         /* this is the only place which holds both the socket's rs_lock
 821          * and the connection's c_lock */
 822         spin_lock_irqsave(&rs->rs_lock, flags);
 823
 824         /*
 825          * If there is a little space in sndbuf, we don't queue anything,
 826          * and userspace gets -EAGAIN. But poll() indicates there's send
 827          * room. This can lead to bad behavior (spinning) if snd_bytes isn't
 828          * freed up by incoming acks. So we check the *old* value of
 829          * rs_snd_bytes here to allow the last msg to exceed the buffer,
 830          * and poll() now knows no more data can be sent.
 831          */
 832         if (rs->rs_snd_bytes < rds_sk_sndbuf(rs)) {
 833                 rs->rs_snd_bytes += len;
 834
 835                 /* let recv side know we are close to send space exhaustion.
 836                  * This is probably not the optimal way to do it, as this
 837                  * means we set the flag on *all* messages as soon as our
 838                  * throughput hits a certain threshold.
 839                  */
 840                 if (rs->rs_snd_bytes >= rds_sk_sndbuf(rs) / 2)
 841                         set_bit(RDS_MSG_ACK_REQUIRED, &rm->m_flags);
 842
 843                 list_add_tail(&rm->m_sock_item, &rs->rs_send_queue);
 844                 set_bit(RDS_MSG_ON_SOCK, &rm->m_flags);
 845                 rds_message_addref(rm);
 846                 sock_hold(rds_rs_to_sk(rs));
 847                 rm->m_rs = rs;
 848
 849                 /* The code ordering is a little weird, but we're
 850                    trying to minimize the time we hold c_lock */
 851                 rds_message_populate_header(&rm->m_inc.i_hdr, sport, dport, 0);
 852                 rm->m_inc.i_conn = conn;
 853                 rm->m_inc.i_conn_path = cp;
 854                 rds_message_addref(rm);
 855
 856                 spin_lock(&cp->cp_lock);
 857                 rm->m_inc.i_hdr.h_sequence = cpu_to_be64(cp->cp_next_tx_seq++);
 858                 list_add_tail(&rm->m_conn_item, &cp->cp_send_queue);
 859                 set_bit(RDS_MSG_ON_CONN, &rm->m_flags);
 860                 spin_unlock(&cp->cp_lock);
 861
 862                 rdsdebug("queued msg %p len %d, rs %p bytes %d seq %llu\n",
 863                          rm, len, rs, rs->rs_snd_bytes,
 864                          (unsigned long long)be64_to_cpu(rm->m_inc.i_hdr.h_sequence));
 865
 866                 *queued = 1;
 867         }
 868
 869         spin_unlock_irqrestore(&rs->rs_lock, flags);
 870 out:
 871         return *queued;
 872 }
 873
 874 /*
 875  * rds_message is getting to be quite complicated, and we'd like to allocate
 876  * it all in one go. This figures out how big it needs to be up front.
 877  */
 878 static int rds_rm_size(struct msghdr *msg, int num_sgs)
 879 {
 880         struct cmsghdr *cmsg;
 881         int size = 0;
 882         int cmsg_groups = 0;
 883         int retval;
 884         bool zcopy_cookie = false;
 885
 886         for_each_cmsghdr(cmsg, msg) {
 887                 if (!CMSG_OK(msg, cmsg))
 888                         return -EINVAL;
 889
 890                 if (cmsg->cmsg_level != SOL_RDS)
 891                         continue;
 892
 893                 switch (cmsg->cmsg_type) {
 894                 case RDS_CMSG_RDMA_ARGS:
 895                         cmsg_groups |= 1;
 896                         retval = rds_rdma_extra_size(CMSG_DATA(cmsg));
 897                         if (retval < 0)
 898                                 return retval;
 899                         size += retval;
 900
 901                         break;
 902
 903                 case RDS_CMSG_ZCOPY_COOKIE:
 904                         zcopy_cookie = true;
 905                         /* fall through */
 906
 907                 case RDS_CMSG_RDMA_DEST:
 908                 case RDS_CMSG_RDMA_MAP:
 909                         cmsg_groups |= 2;
 910                         /* these are valid but do no add any size */
 911                         break;
 912
 913                 case RDS_CMSG_ATOMIC_CSWP:
 914                 case RDS_CMSG_ATOMIC_FADD:
 915                 case RDS_CMSG_MASKED_ATOMIC_CSWP:
 916                 case RDS_CMSG_MASKED_ATOMIC_FADD:
 917                         cmsg_groups |= 1;
 918                         size += sizeof(struct scatterlist);
 919                         break;
 920
 921                 default:
 922                         return -EINVAL;
 923                 }
 924
 925         }
 926
 927         if ((msg->msg_flags & MSG_ZEROCOPY) && !zcopy_cookie)
 928                 return -EINVAL;
 929
 930         size += num_sgs * sizeof(struct scatterlist);
 931
 932         /* Ensure (DEST, MAP) are never used with (ARGS, ATOMIC) */
 933         if (cmsg_groups == 3)
 934                 return -EINVAL;
 935
 936         return size;
 937 }
 938
 939 static int rds_cmsg_zcopy(struct rds_sock *rs, struct rds_message *rm,
 940                           struct cmsghdr *cmsg)
 941 {
 942         u32 *cookie;
 943
 944         if (cmsg->cmsg_len < CMSG_LEN(sizeof(*cookie)) ||
 945             !rm->data.op_mmp_znotifier)
 946                 return -EINVAL;
 947         cookie = CMSG_DATA(cmsg);
 948         rm->data.op_mmp_znotifier->z_cookie = *cookie;
 949         return 0;
 950 }
 951
 952 static int rds_cmsg_send(struct rds_sock *rs, struct rds_message *rm,
 953                          struct msghdr *msg, int *allocated_mr)
 954 {
 955         struct cmsghdr *cmsg;
 956         int ret = 0;
 957
 958         for_each_cmsghdr(cmsg, msg) {
 959                 if (!CMSG_OK(msg, cmsg))
 960                         return -EINVAL;
 961
 962                 if (cmsg->cmsg_level != SOL_RDS)
 963                         continue;
 964
 965                 /* As a side effect, RDMA_DEST and RDMA_MAP will set
 966                  * rm->rdma.m_rdma_cookie and rm->rdma.m_rdma_mr.
 967                  */
 968                 switch (cmsg->cmsg_type) {
 969                 case RDS_CMSG_RDMA_ARGS:
 970                         ret = rds_cmsg_rdma_args(rs, rm, cmsg);
 971                         break;
 972
 973                 case RDS_CMSG_RDMA_DEST:
 974                         ret = rds_cmsg_rdma_dest(rs, rm, cmsg);
 975                         break;
 976
 977                 case RDS_CMSG_RDMA_MAP:
 978                         ret = rds_cmsg_rdma_map(rs, rm, cmsg);
 979                         if (!ret)
 980                                 *allocated_mr = 1;
 981                         else if (ret == -ENODEV)
 982                                 /* Accommodate the get_mr() case which can fail
 983                                  * if connection isn't established yet.
 984                                  */
 985                                 ret = -EAGAIN;
 986                         break;
 987                 case RDS_CMSG_ATOMIC_CSWP:
 988                 case RDS_CMSG_ATOMIC_FADD:
 989                 case RDS_CMSG_MASKED_ATOMIC_CSWP:
 990                 case RDS_CMSG_MASKED_ATOMIC_FADD:
 991                         ret = rds_cmsg_atomic(rs, rm, cmsg);
 992                         break;
 993
 994                 case RDS_CMSG_ZCOPY_COOKIE:
 995                         ret = rds_cmsg_zcopy(rs, rm, cmsg);
 996                         break;
 997
 998                 default:
 999                         return -EINVAL;
1000                 }
1001
1002                 if (ret)
1003                         break;
1004         }
1005
1006         return ret;
1007 }
1008
1009 static int rds_send_mprds_hash(struct rds_sock *rs, struct rds_connection *conn)
1010 {
1011         int hash;
1012
1013         if (conn->c_npaths == 0)
1014                 hash = RDS_MPATH_HASH(rs, RDS_MPATH_WORKERS);
1015         else
1016                 hash = RDS_MPATH_HASH(rs, conn->c_npaths);
1017         if (conn->c_npaths == 0 && hash != 0) {
1018                 rds_send_ping(conn, 0);
1019
1020                 /* The underlying connection is not up yet.  Need to wait
1021                  * until it is up to be sure that the non-zero c_path can be
1022                  * used.  But if we are interrupted, we have to use the zero
1023                  * c_path in case the connection ends up being non-MP capable.
1024                  */
1025                 if (conn->c_npaths == 0)
1026                         if (wait_event_interruptible(conn->c_hs_waitq,
1027                                                      conn->c_npaths != 0))
1028                                 hash = 0;
1029                 if (conn->c_npaths == 1)
1030                         hash = 0;
1031         }
1032         return hash;
1033 }
1034
1035 static int rds_rdma_bytes(struct msghdr *msg, size_t *rdma_bytes)
1036 {
1037         struct rds_rdma_args *args;
1038         struct cmsghdr *cmsg;
1039
1040         for_each_cmsghdr(cmsg, msg) {
1041                 if (!CMSG_OK(msg, cmsg))
1042                         return -EINVAL;
1043
1044                 if (cmsg->cmsg_level != SOL_RDS)
1045                         continue;
1046
1047                 if (cmsg->cmsg_type == RDS_CMSG_RDMA_ARGS) {
1048                         if (cmsg->cmsg_len <
1049                             CMSG_LEN(sizeof(struct rds_rdma_args)))
1050                                 return -EINVAL;
1051                         args = CMSG_DATA(cmsg);
1052                         *rdma_bytes += args->remote_vec.bytes;
1053                 }
1054         }
1055         return 0;
1056 }
1057
1058 int rds_sendmsg(struct socket *sock, struct msghdr *msg, size_t payload_len)
1059 {
1060         struct sock *sk = sock->sk;
1061         struct rds_sock *rs = rds_sk_to_rs(sk);
1062         DECLARE_SOCKADDR(struct sockaddr_in *, usin, msg->msg_name);
1063         __be32 daddr;
1064         __be16 dport;
1065         struct rds_message *rm = NULL;
1066         struct rds_connection *conn;
1067         int ret = 0;
1068         int queued = 0, allocated_mr = 0;
1069         int nonblock = msg->msg_flags & MSG_DONTWAIT;
1070         long timeo = sock_sndtimeo(sk, nonblock);
1071         struct rds_conn_path *cpath;
1072         size_t total_payload_len = payload_len, rdma_payload_len = 0;
1073         bool zcopy = ((msg->msg_flags & MSG_ZEROCOPY) &&
1074                       sock_flag(rds_rs_to_sk(rs), SOCK_ZEROCOPY));
1075         int num_sgs = ceil(payload_len, PAGE_SIZE);
1076
1077         /* Mirror Linux UDP mirror of BSD error message compatibility */
1078         /* XXX: Perhaps MSG_MORE someday */
1079         if (msg->msg_flags & ~(MSG_DONTWAIT | MSG_CMSG_COMPAT | MSG_ZEROCOPY)) {
1080                 ret = -EOPNOTSUPP;
1081                 goto out;
1082         }
1083
1084         if (msg->msg_namelen) {
1085                 /* XXX fail non-unicast destination IPs? */
1086                 if (msg->msg_namelen < sizeof(*usin) || usin->sin_family != AF_INET) {
1087                         ret = -EINVAL;
1088                         goto out;
1089                 }
1090                 daddr = usin->sin_addr.s_addr;
1091                 dport = usin->sin_port;
1092         } else {
1093                 /* We only care about consistency with ->connect() */
1094                 lock_sock(sk);
1095                 daddr = rs->rs_conn_addr;
1096                 dport = rs->rs_conn_port;
1097                 release_sock(sk);
1098         }
1099
1100         lock_sock(sk);
1101         if (daddr == 0 || rs->rs_bound_addr == 0) {
1102                 release_sock(sk);
1103                 ret = -ENOTCONN; /* XXX not a great errno */
1104                 goto out;
1105         }
1106         release_sock(sk);
1107
1108         ret = rds_rdma_bytes(msg, &rdma_payload_len);
1109         if (ret)
1110                 goto out;
1111
1112         total_payload_len += rdma_payload_len;
1113         if (max_t(size_t, payload_len, rdma_payload_len) > RDS_MAX_MSG_SIZE) {
1114                 ret = -EMSGSIZE;
1115                 goto out;
1116         }
1117
1118         if (payload_len > rds_sk_sndbuf(rs)) {
1119                 ret = -EMSGSIZE;
1120                 goto out;
1121         }
1122
1123         if (zcopy) {
1124                 if (rs->rs_transport->t_type != RDS_TRANS_TCP) {
1125                         ret = -EOPNOTSUPP;
1126                         goto out;
1127                 }
1128                 num_sgs = iov_iter_npages(&msg->msg_iter, INT_MAX);
1129         }
1130         /* size of rm including all sgs */
1131         ret = rds_rm_size(msg, num_sgs);
1132         if (ret < 0)
1133                 goto out;
1134
1135         rm = rds_message_alloc(ret, GFP_KERNEL);
1136         if (!rm) {
1137                 ret = -ENOMEM;
1138                 goto out;
1139         }
1140
1141         /* Attach data to the rm */
1142         if (payload_len) {
1143                 rm->data.op_sg = rds_message_alloc_sgs(rm, num_sgs);
1144                 if (!rm->data.op_sg) {
1145                         ret = -ENOMEM;
1146                         goto out;
1147                 }
1148                 ret = rds_message_copy_from_user(rm, &msg->msg_iter, zcopy);
1149                 if (ret)
1150                         goto out;
1151         }
1152         rm->data.op_active = 1;
1153
1154         rm->m_daddr = daddr;
1155
1156         /* rds_conn_create has a spinlock that runs with IRQ off.
1157          * Caching the conn in the socket helps a lot. */
1158         if (rs->rs_conn && rs->rs_conn->c_faddr == daddr)
1159                 conn = rs->rs_conn;
1160         else {
1161                 conn = rds_conn_create_outgoing(sock_net(sock->sk),
1162                                                 rs->rs_bound_addr, daddr,
1163                                         rs->rs_transport,
1164                                         sock->sk->sk_allocation);
1165                 if (IS_ERR(conn)) {
1166                         ret = PTR_ERR(conn);
1167                         goto out;
1168                 }
1169                 rs->rs_conn = conn;
1170         }
1171
1172         /* Parse any control messages the user may have included. */
1173         ret = rds_cmsg_send(rs, rm, msg, &allocated_mr);
1174         if (ret) {
1175                 /* Trigger connection so that its ready for the next retry */
1176                 if (ret ==  -EAGAIN)
1177                         rds_conn_connect_if_down(conn);
1178                 goto out;
1179         }
1180
1181         if (rm->rdma.op_active && !conn->c_trans->xmit_rdma) {
1182                 printk_ratelimited(KERN_NOTICE "rdma_op %p conn xmit_rdma %p\n",
1183                                &rm->rdma, conn->c_trans->xmit_rdma);
1184                 ret = -EOPNOTSUPP;
1185                 goto out;
1186         }
1187
1188         if (rm->atomic.op_active && !conn->c_trans->xmit_atomic) {
1189                 printk_ratelimited(KERN_NOTICE "atomic_op %p conn xmit_atomic %p\n",
1190                                &rm->atomic, conn->c_trans->xmit_atomic);
1191                 ret = -EOPNOTSUPP;
1192                 goto out;
1193         }
1194
1195         if (conn->c_trans->t_mp_capable)
1196                 cpath = &conn->c_path[rds_send_mprds_hash(rs, conn)];
1197         else
1198                 cpath = &conn->c_path[0];
1199
1200         if (rds_destroy_pending(conn)) {
1201                 ret = -EAGAIN;
1202                 goto out;
1203         }
1204
1205         rds_conn_path_connect_if_down(cpath);
1206
1207         ret = rds_cong_wait(conn->c_fcong, dport, nonblock, rs);
1208         if (ret) {
1209                 rs->rs_seen_congestion = 1;
1210                 goto out;
1211         }
1212         while (!rds_send_queue_rm(rs, conn, cpath, rm, rs->rs_bound_port,
1213                                   dport, &queued)) {
1214                 rds_stats_inc(s_send_queue_full);
1215
1216                 if (nonblock) {
1217                         ret = -EAGAIN;
1218                         goto out;
1219                 }
1220
1221                 timeo = wait_event_interruptible_timeout(*sk_sleep(sk),
1222                                         rds_send_queue_rm(rs, conn, cpath, rm,
1223                                                           rs->rs_bound_port,
1224                                                           dport,
1225                                                           &queued),
1226                                         timeo);
1227                 rdsdebug("sendmsg woke queued %d timeo %ld\n", queued, timeo);
1228                 if (timeo > 0 || timeo == MAX_SCHEDULE_TIMEOUT)
1229                         continue;
1230
1231                 ret = timeo;
1232                 if (ret == 0)
1233                         ret = -ETIMEDOUT;
1234                 goto out;
1235         }
1236
1237         /*
1238          * By now we've committed to the send.  We reuse rds_send_worker()
1239          * to retry sends in the rds thread if the transport asks us to.
1240          */
1241         rds_stats_inc(s_send_queued);
1242
1243         ret = rds_send_xmit(cpath);
1244         if (ret == -ENOMEM || ret == -EAGAIN) {
1245                 ret = 0;
1246                 rcu_read_lock();
1247                 if (rds_destroy_pending(cpath->cp_conn))
1248                         ret = -ENETUNREACH;
1249                 else
1250                         queue_delayed_work(rds_wq, &cpath->cp_send_w, 1);
1251                 rcu_read_unlock();
1252         }
1253         if (ret)
1254                 goto out;
1255         rds_message_put(rm);
1256         return payload_len;
1257
1258 out:
1259         /* If the user included a RDMA_MAP cmsg, we allocated a MR on the fly.
1260          * If the sendmsg goes through, we keep the MR. If it fails with EAGAIN
1261          * or in any other way, we need to destroy the MR again */
1262         if (allocated_mr)
1263                 rds_rdma_unuse(rs, rds_rdma_cookie_key(rm->m_rdma_cookie), 1);
1264
1265         if (rm)
1266                 rds_message_put(rm);
1267         return ret;
1268 }
1269
1270 /*
1271  * send out a probe. Can be shared by rds_send_ping,
1272  * rds_send_pong, rds_send_hb.
1273  * rds_send_hb should use h_flags
1274  *   RDS_FLAG_HB_PING|RDS_FLAG_ACK_REQUIRED
1275  * or
1276  *   RDS_FLAG_HB_PONG|RDS_FLAG_ACK_REQUIRED
1277  */
1278 static int
1279 rds_send_probe(struct rds_conn_path *cp, __be16 sport,
1280                __be16 dport, u8 h_flags)
1281 {
1282         struct rds_message *rm;
1283         unsigned long flags;
1284         int ret = 0;
1285
1286         rm = rds_message_alloc(0, GFP_ATOMIC);
1287         if (!rm) {
1288                 ret = -ENOMEM;
1289                 goto out;
1290         }
1291
1292         rm->m_daddr = cp->cp_conn->c_faddr;
1293         rm->data.op_active = 1;
1294
1295         rds_conn_path_connect_if_down(cp);
1296
1297         ret = rds_cong_wait(cp->cp_conn->c_fcong, dport, 1, NULL);
1298         if (ret)
1299                 goto out;
1300
1301         spin_lock_irqsave(&cp->cp_lock, flags);
1302         list_add_tail(&rm->m_conn_item, &cp->cp_send_queue);
1303         set_bit(RDS_MSG_ON_CONN, &rm->m_flags);
1304         rds_message_addref(rm);
1305         rm->m_inc.i_conn = cp->cp_conn;
1306         rm->m_inc.i_conn_path = cp;
1307
1308         rds_message_populate_header(&rm->m_inc.i_hdr, sport, dport,
1309                                     cp->cp_next_tx_seq);
1310         rm->m_inc.i_hdr.h_flags |= h_flags;
1311         cp->cp_next_tx_seq++;
1312
1313         if (RDS_HS_PROBE(be16_to_cpu(sport), be16_to_cpu(dport)) &&
1314             cp->cp_conn->c_trans->t_mp_capable) {
1315                 u16 npaths = cpu_to_be16(RDS_MPATH_WORKERS);
1316                 u32 my_gen_num = cpu_to_be32(cp->cp_conn->c_my_gen_num);
1317
1318                 rds_message_add_extension(&rm->m_inc.i_hdr,
1319                                           RDS_EXTHDR_NPATHS, &npaths,
1320                                           sizeof(npaths));
1321                 rds_message_add_extension(&rm->m_inc.i_hdr,
1322                                           RDS_EXTHDR_GEN_NUM,
1323                                           &my_gen_num,
1324                                           sizeof(u32));
1325         }
1326         spin_unlock_irqrestore(&cp->cp_lock, flags);
1327
1328         rds_stats_inc(s_send_queued);
1329         rds_stats_inc(s_send_pong);
1330
1331         /* schedule the send work on rds_wq */
1332         rcu_read_lock();
1333         if (!rds_destroy_pending(cp->cp_conn))
1334                 queue_delayed_work(rds_wq, &cp->cp_send_w, 1);
1335         rcu_read_unlock();
1336
1337         rds_message_put(rm);
1338         return 0;
1339
1340 out:
1341         if (rm)
1342                 rds_message_put(rm);
1343         return ret;
1344 }
1345
1346 int
1347 rds_send_pong(struct rds_conn_path *cp, __be16 dport)
1348 {
1349         return rds_send_probe(cp, 0, dport, 0);
1350 }
1351
1352 void
1353 rds_send_ping(struct rds_connection *conn, int cp_index)
1354 {
1355         unsigned long flags;
1356         struct rds_conn_path *cp = &conn->c_path[cp_index];
1357
1358         spin_lock_irqsave(&cp->cp_lock, flags);
1359         if (conn->c_ping_triggered) {
1360                 spin_unlock_irqrestore(&cp->cp_lock, flags);
1361                 return;
1362         }
1363         conn->c_ping_triggered = 1;
1364         spin_unlock_irqrestore(&cp->cp_lock, flags);
1365         rds_send_probe(cp, cpu_to_be16(RDS_FLAG_PROBE_PORT), 0, 0);
1366 }
1367 EXPORT_SYMBOL_GPL(rds_send_ping);