lnet/klnds/socklnd/socklnd_cb.c

   1 /*
   2  * Copyright (c) 2003, 2010, Oracle and/or its affiliates. All rights reserved.
   3  *
   4  * Copyright (c) 2011, 2012, Intel Corporation.
   5  *
   6  *   Author: Zach Brown <zab@zabbo.net>
   7  *   Author: Peter J. Braam <braam@clusterfs.com>
   8  *   Author: Phil Schwan <phil@clusterfs.com>
   9  *   Author: Eric Barton <eric@bartonsoftware.com>
  10  *
  11  *   This file is part of Portals, http://www.sf.net/projects/sandiaportals/
  12  *
  13  *   Portals is free software; you can redistribute it and/or
  14  *   modify it under the terms of version 2 of the GNU General Public
  15  *   License as published by the Free Software Foundation.
  16  *
  17  *   Portals is distributed in the hope that it will be useful,
  18  *   but WITHOUT ANY WARRANTY; without even the implied warranty of
  19  *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  20  *   GNU General Public License for more details.
  21  *
  22  *   You should have received a copy of the GNU General Public License
  23  *   along with Portals; if not, write to the Free Software
  24  *   Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  25  */
  26
  27 #include "socklnd.h"
  28
  29 ksock_tx_t *
  30 ksocknal_alloc_tx(int type, int size)
  31 {
  32         ksock_tx_t *tx = NULL;
  33
  34         if (type == KSOCK_MSG_NOOP) {
  35                 LASSERT(size == KSOCK_NOOP_TX_SIZE);
  36
  37                 /* searching for a noop tx in free list */
  38                 spin_lock(&ksocknal_data.ksnd_tx_lock);
  39
  40                 if (!cfs_list_empty(&ksocknal_data.ksnd_idle_noop_txs)) {
  41                         tx = cfs_list_entry(ksocknal_data.ksnd_idle_noop_txs. \
  42                                             next, ksock_tx_t, tx_list);
  43                         LASSERT(tx->tx_desc_size == size);
  44                         cfs_list_del(&tx->tx_list);
  45                 }
  46
  47                 spin_unlock(&ksocknal_data.ksnd_tx_lock);
  48         }
  49
  50         if (tx == NULL)
  51                 LIBCFS_ALLOC(tx, size);
  52
  53         if (tx == NULL)
  54                 return NULL;
  55
  56         atomic_set(&tx->tx_refcount, 1);
  57         tx->tx_zc_aborted = 0;
  58         tx->tx_zc_capable = 0;
  59         tx->tx_zc_checked = 0;
  60         tx->tx_desc_size  = size;
  61
  62         atomic_inc(&ksocknal_data.ksnd_nactive_txs);
  63
  64         return tx;
  65 }
  66
  67 ksock_tx_t *
  68 ksocknal_alloc_tx_noop(__u64 cookie, int nonblk)
  69 {
  70         ksock_tx_t *tx;
  71
  72         tx = ksocknal_alloc_tx(KSOCK_MSG_NOOP, KSOCK_NOOP_TX_SIZE);
  73         if (tx == NULL) {
  74                 CERROR("Can't allocate noop tx desc\n");
  75                 return NULL;
  76         }
  77
  78         tx->tx_conn     = NULL;
  79         tx->tx_lnetmsg  = NULL;
  80         tx->tx_kiov     = NULL;
  81         tx->tx_nkiov    = 0;
  82         tx->tx_iov      = tx->tx_frags.virt.iov;
  83         tx->tx_niov     = 1;
  84         tx->tx_nonblk   = nonblk;
  85
  86         socklnd_init_msg(&tx->tx_msg, KSOCK_MSG_NOOP);
  87         tx->tx_msg.ksm_zc_cookies[1] = cookie;
  88
  89         return tx;
  90 }
  91
  92
  93 void
  94 ksocknal_free_tx (ksock_tx_t *tx)
  95 {
  96         atomic_dec(&ksocknal_data.ksnd_nactive_txs);
  97
  98         if (tx->tx_lnetmsg == NULL && tx->tx_desc_size == KSOCK_NOOP_TX_SIZE) {
  99                 /* it's a noop tx */
 100                 spin_lock(&ksocknal_data.ksnd_tx_lock);
 101
 102                 cfs_list_add(&tx->tx_list, &ksocknal_data.ksnd_idle_noop_txs);
 103
 104                 spin_unlock(&ksocknal_data.ksnd_tx_lock);
 105         } else {
 106                 LIBCFS_FREE(tx, tx->tx_desc_size);
 107         }
 108 }
 109
 110 int
 111 ksocknal_send_iov (ksock_conn_t *conn, ksock_tx_t *tx)
 112 {
 113         struct iovec  *iov = tx->tx_iov;
 114         int    nob;
 115         int    rc;
 116
 117         LASSERT (tx->tx_niov > 0);
 118
 119         /* Never touch tx->tx_iov inside ksocknal_lib_send_iov() */
 120         rc = ksocknal_lib_send_iov(conn, tx);
 121
 122         if (rc <= 0)                            /* sent nothing? */
 123                 return (rc);
 124
 125         nob = rc;
 126         LASSERT (nob <= tx->tx_resid);
 127         tx->tx_resid -= nob;
 128
 129         /* "consume" iov */
 130         do {
 131                 LASSERT (tx->tx_niov > 0);
 132
 133                 if (nob < (int) iov->iov_len) {
 134                         iov->iov_base = (void *)((char *)iov->iov_base + nob);
 135                         iov->iov_len -= nob;
 136                         return (rc);
 137                 }
 138
 139                 nob -= iov->iov_len;
 140                 tx->tx_iov = ++iov;
 141                 tx->tx_niov--;
 142         } while (nob != 0);
 143
 144         return (rc);
 145 }
 146
 147 int
 148 ksocknal_send_kiov (ksock_conn_t *conn, ksock_tx_t *tx)
 149 {
 150         lnet_kiov_t    *kiov = tx->tx_kiov;
 151         int     nob;
 152         int     rc;
 153
 154         LASSERT (tx->tx_niov == 0);
 155         LASSERT (tx->tx_nkiov > 0);
 156
 157         /* Never touch tx->tx_kiov inside ksocknal_lib_send_kiov() */
 158         rc = ksocknal_lib_send_kiov(conn, tx);
 159
 160         if (rc <= 0)                            /* sent nothing? */
 161                 return (rc);
 162
 163         nob = rc;
 164         LASSERT (nob <= tx->tx_resid);
 165         tx->tx_resid -= nob;
 166
 167         /* "consume" kiov */
 168         do {
 169                 LASSERT(tx->tx_nkiov > 0);
 170
 171                 if (nob < (int)kiov->kiov_len) {
 172                         kiov->kiov_offset += nob;
 173                         kiov->kiov_len -= nob;
 174                         return rc;
 175                 }
 176
 177                 nob -= (int)kiov->kiov_len;
 178                 tx->tx_kiov = ++kiov;
 179                 tx->tx_nkiov--;
 180         } while (nob != 0);
 181
 182         return (rc);
 183 }
 184
 185 int
 186 ksocknal_transmit (ksock_conn_t *conn, ksock_tx_t *tx)
 187 {
 188         int      rc;
 189         int      bufnob;
 190
 191         if (ksocknal_data.ksnd_stall_tx != 0) {
 192                 cfs_pause(cfs_time_seconds(ksocknal_data.ksnd_stall_tx));
 193         }
 194
 195         LASSERT (tx->tx_resid != 0);
 196
 197         rc = ksocknal_connsock_addref(conn);
 198         if (rc != 0) {
 199                 LASSERT (conn->ksnc_closing);
 200                 return (-ESHUTDOWN);
 201         }
 202
 203         do {
 204                 if (ksocknal_data.ksnd_enomem_tx > 0) {
 205                         /* testing... */
 206                         ksocknal_data.ksnd_enomem_tx--;
 207                         rc = -EAGAIN;
 208                 } else if (tx->tx_niov != 0) {
 209                         rc = ksocknal_send_iov (conn, tx);
 210                 } else {
 211                         rc = ksocknal_send_kiov (conn, tx);
 212                 }
 213
 214                 bufnob = libcfs_sock_wmem_queued(conn->ksnc_sock);
 215                 if (rc > 0)                     /* sent something? */
 216                         conn->ksnc_tx_bufnob += rc; /* account it */
 217
 218                 if (bufnob < conn->ksnc_tx_bufnob) {
 219                         /* allocated send buffer bytes < computed; infer
 220                          * something got ACKed */
 221                         conn->ksnc_tx_deadline =
 222                                 cfs_time_shift(*ksocknal_tunables.ksnd_timeout);
 223                         conn->ksnc_peer->ksnp_last_alive = cfs_time_current();
 224                         conn->ksnc_tx_bufnob = bufnob;
 225                         smp_mb();
 226                 }
 227
 228                 if (rc <= 0) { /* Didn't write anything? */
 229
 230                         if (rc == 0) /* some stacks return 0 instead of -EAGAIN */
 231                                 rc = -EAGAIN;
 232
 233                         /* Check if EAGAIN is due to memory pressure */
 234                         if(rc == -EAGAIN && ksocknal_lib_memory_pressure(conn))
 235                                 rc = -ENOMEM;
 236
 237                         break;
 238                 }
 239
 240                 /* socket's wmem_queued now includes 'rc' bytes */
 241                 atomic_sub (rc, &conn->ksnc_tx_nob);
 242                 rc = 0;
 243
 244         } while (tx->tx_resid != 0);
 245
 246         ksocknal_connsock_decref(conn);
 247         return (rc);
 248 }
 249
 250 int
 251 ksocknal_recv_iov (ksock_conn_t *conn)
 252 {
 253         struct iovec *iov = conn->ksnc_rx_iov;
 254         int     nob;
 255         int     rc;
 256
 257         LASSERT (conn->ksnc_rx_niov > 0);
 258
 259         /* Never touch conn->ksnc_rx_iov or change connection
 260          * status inside ksocknal_lib_recv_iov */
 261         rc = ksocknal_lib_recv_iov(conn);
 262
 263         if (rc <= 0)
 264                 return (rc);
 265
 266         /* received something... */
 267         nob = rc;
 268
 269         conn->ksnc_peer->ksnp_last_alive = cfs_time_current();
 270         conn->ksnc_rx_deadline =
 271                 cfs_time_shift(*ksocknal_tunables.ksnd_timeout);
 272         smp_mb();                       /* order with setting rx_started */
 273         conn->ksnc_rx_started = 1;
 274
 275         conn->ksnc_rx_nob_wanted -= nob;
 276         conn->ksnc_rx_nob_left -= nob;
 277
 278         do {
 279                 LASSERT (conn->ksnc_rx_niov > 0);
 280
 281                 if (nob < (int)iov->iov_len) {
 282                         iov->iov_len -= nob;
 283                         iov->iov_base = (void *)((char *)iov->iov_base + nob);
 284                         return (-EAGAIN);
 285                 }
 286
 287                 nob -= iov->iov_len;
 288                 conn->ksnc_rx_iov = ++iov;
 289                 conn->ksnc_rx_niov--;
 290         } while (nob != 0);
 291
 292         return (rc);
 293 }
 294
 295 int
 296 ksocknal_recv_kiov (ksock_conn_t *conn)
 297 {
 298         lnet_kiov_t   *kiov = conn->ksnc_rx_kiov;
 299         int     nob;
 300         int     rc;
 301         LASSERT (conn->ksnc_rx_nkiov > 0);
 302
 303         /* Never touch conn->ksnc_rx_kiov or change connection
 304          * status inside ksocknal_lib_recv_iov */
 305         rc = ksocknal_lib_recv_kiov(conn);
 306
 307         if (rc <= 0)
 308                 return (rc);
 309
 310         /* received something... */
 311         nob = rc;
 312
 313         conn->ksnc_peer->ksnp_last_alive = cfs_time_current();
 314         conn->ksnc_rx_deadline =
 315                 cfs_time_shift(*ksocknal_tunables.ksnd_timeout);
 316         smp_mb();                       /* order with setting rx_started */
 317         conn->ksnc_rx_started = 1;
 318
 319         conn->ksnc_rx_nob_wanted -= nob;
 320         conn->ksnc_rx_nob_left -= nob;
 321
 322         do {
 323                 LASSERT (conn->ksnc_rx_nkiov > 0);
 324
 325                 if (nob < (int) kiov->kiov_len) {
 326                         kiov->kiov_offset += nob;
 327                         kiov->kiov_len -= nob;
 328                         return -EAGAIN;
 329                 }
 330
 331                 nob -= kiov->kiov_len;
 332                 conn->ksnc_rx_kiov = ++kiov;
 333                 conn->ksnc_rx_nkiov--;
 334         } while (nob != 0);
 335
 336         return 1;
 337 }
 338
 339 int
 340 ksocknal_receive (ksock_conn_t *conn)
 341 {
 342         /* Return 1 on success, 0 on EOF, < 0 on error.
 343          * Caller checks ksnc_rx_nob_wanted to determine
 344          * progress/completion. */
 345         int     rc;
 346         ENTRY;
 347
 348         if (ksocknal_data.ksnd_stall_rx != 0) {
 349                 cfs_pause(cfs_time_seconds (ksocknal_data.ksnd_stall_rx));
 350         }
 351
 352         rc = ksocknal_connsock_addref(conn);
 353         if (rc != 0) {
 354                 LASSERT (conn->ksnc_closing);
 355                 return (-ESHUTDOWN);
 356         }
 357
 358         for (;;) {
 359                 if (conn->ksnc_rx_niov != 0)
 360                         rc = ksocknal_recv_iov (conn);
 361                 else
 362                         rc = ksocknal_recv_kiov (conn);
 363
 364                 if (rc <= 0) {
 365                         /* error/EOF or partial receive */
 366                         if (rc == -EAGAIN) {
 367                                 rc = 1;
 368                         } else if (rc == 0 && conn->ksnc_rx_started) {
 369                                 /* EOF in the middle of a message */
 370                                 rc = -EPROTO;
 371                         }
 372                         break;
 373                 }
 374
 375                 /* Completed a fragment */
 376
 377                 if (conn->ksnc_rx_nob_wanted == 0) {
 378                         rc = 1;
 379                         break;
 380                 }
 381         }
 382
 383         ksocknal_connsock_decref(conn);
 384         RETURN (rc);
 385 }
 386
 387 void
 388 ksocknal_tx_done (lnet_ni_t *ni, ksock_tx_t *tx)
 389 {
 390         lnet_msg_t  *lnetmsg = tx->tx_lnetmsg;
 391         int          rc = (tx->tx_resid == 0 && !tx->tx_zc_aborted) ? 0 : -EIO;
 392         ENTRY;
 393
 394         LASSERT(ni != NULL || tx->tx_conn != NULL);
 395
 396         if (tx->tx_conn != NULL)
 397                 ksocknal_conn_decref(tx->tx_conn);
 398
 399         if (ni == NULL && tx->tx_conn != NULL)
 400                 ni = tx->tx_conn->ksnc_peer->ksnp_ni;
 401
 402         ksocknal_free_tx (tx);
 403         if (lnetmsg != NULL) /* KSOCK_MSG_NOOP go without lnetmsg */
 404                 lnet_finalize (ni, lnetmsg, rc);
 405
 406         EXIT;
 407 }
 408
 409 void
 410 ksocknal_txlist_done (lnet_ni_t *ni, cfs_list_t *txlist, int error)
 411 {
 412         ksock_tx_t *tx;
 413
 414         while (!cfs_list_empty (txlist)) {
 415                 tx = cfs_list_entry (txlist->next, ksock_tx_t, tx_list);
 416
 417                 if (error && tx->tx_lnetmsg != NULL) {
 418                         CNETERR("Deleting packet type %d len %d %s->%s\n",
 419                                 le32_to_cpu (tx->tx_lnetmsg->msg_hdr.type),
 420                                 le32_to_cpu (tx->tx_lnetmsg->msg_hdr.payload_length),
 421                                 libcfs_nid2str(le64_to_cpu(tx->tx_lnetmsg->msg_hdr.src_nid)),
 422                                 libcfs_nid2str(le64_to_cpu(tx->tx_lnetmsg->msg_hdr.dest_nid)));
 423                 } else if (error) {
 424                         CNETERR("Deleting noop packet\n");
 425                 }
 426
 427                 cfs_list_del (&tx->tx_list);
 428
 429                 LASSERT (atomic_read(&tx->tx_refcount) == 1);
 430                 ksocknal_tx_done (ni, tx);
 431         }
 432 }
 433
 434 static void
 435 ksocknal_check_zc_req(ksock_tx_t *tx)
 436 {
 437         ksock_conn_t   *conn = tx->tx_conn;
 438         ksock_peer_t   *peer = conn->ksnc_peer;
 439
 440         /* Set tx_msg.ksm_zc_cookies[0] to a unique non-zero cookie and add tx
 441          * to ksnp_zc_req_list if some fragment of this message should be sent
 442          * zero-copy.  Our peer will send an ACK containing this cookie when
 443          * she has received this message to tell us we can signal completion.
 444          * tx_msg.ksm_zc_cookies[0] remains non-zero while tx is on
 445          * ksnp_zc_req_list. */
 446         LASSERT (tx->tx_msg.ksm_type != KSOCK_MSG_NOOP);
 447         LASSERT (tx->tx_zc_capable);
 448
 449         tx->tx_zc_checked = 1;
 450
 451         if (conn->ksnc_proto == &ksocknal_protocol_v1x ||
 452             !conn->ksnc_zc_capable)
 453                 return;
 454
 455         /* assign cookie and queue tx to pending list, it will be released when
 456          * a matching ack is received. See ksocknal_handle_zcack() */
 457
 458         ksocknal_tx_addref(tx);
 459
 460         spin_lock(&peer->ksnp_lock);
 461
 462         /* ZC_REQ is going to be pinned to the peer */
 463         tx->tx_deadline =
 464                 cfs_time_shift(*ksocknal_tunables.ksnd_timeout);
 465
 466         LASSERT (tx->tx_msg.ksm_zc_cookies[0] == 0);
 467
 468         tx->tx_msg.ksm_zc_cookies[0] = peer->ksnp_zc_next_cookie++;
 469
 470         if (peer->ksnp_zc_next_cookie == 0)
 471                 peer->ksnp_zc_next_cookie = SOCKNAL_KEEPALIVE_PING + 1;
 472
 473         cfs_list_add_tail(&tx->tx_zc_list, &peer->ksnp_zc_req_list);
 474
 475         spin_unlock(&peer->ksnp_lock);
 476 }
 477
 478 static void
 479 ksocknal_uncheck_zc_req(ksock_tx_t *tx)
 480 {
 481         ksock_peer_t   *peer = tx->tx_conn->ksnc_peer;
 482
 483         LASSERT(tx->tx_msg.ksm_type != KSOCK_MSG_NOOP);
 484         LASSERT(tx->tx_zc_capable);
 485
 486         tx->tx_zc_checked = 0;
 487
 488         spin_lock(&peer->ksnp_lock);
 489
 490         if (tx->tx_msg.ksm_zc_cookies[0] == 0) {
 491                 /* Not waiting for an ACK */
 492                 spin_unlock(&peer->ksnp_lock);
 493                 return;
 494         }
 495
 496         tx->tx_msg.ksm_zc_cookies[0] = 0;
 497         cfs_list_del(&tx->tx_zc_list);
 498
 499         spin_unlock(&peer->ksnp_lock);
 500
 501         ksocknal_tx_decref(tx);
 502 }
 503
 504 int
 505 ksocknal_process_transmit (ksock_conn_t *conn, ksock_tx_t *tx)
 506 {
 507         int            rc;
 508
 509         if (tx->tx_zc_capable && !tx->tx_zc_checked)
 510                 ksocknal_check_zc_req(tx);
 511
 512         rc = ksocknal_transmit (conn, tx);
 513
 514         CDEBUG (D_NET, "send(%d) %d\n", tx->tx_resid, rc);
 515
 516         if (tx->tx_resid == 0) {
 517                 /* Sent everything OK */
 518                 LASSERT (rc == 0);
 519
 520                 return (0);
 521         }
 522
 523         if (rc == -EAGAIN)
 524                 return (rc);
 525
 526         if (rc == -ENOMEM) {
 527                 static int counter;
 528
 529                 counter++;   /* exponential backoff warnings */
 530                 if ((counter & (-counter)) == counter)
 531                         CWARN("%u ENOMEM tx %p (%u allocated)\n",
 532                               counter, conn, atomic_read(&libcfs_kmemory));
 533
 534                 /* Queue on ksnd_enomem_conns for retry after a timeout */
 535                 spin_lock_bh(&ksocknal_data.ksnd_reaper_lock);
 536
 537                 /* enomem list takes over scheduler's ref... */
 538                 LASSERT (conn->ksnc_tx_scheduled);
 539                 cfs_list_add_tail(&conn->ksnc_tx_list,
 540                                   &ksocknal_data.ksnd_enomem_conns);
 541                 if (!cfs_time_aftereq(cfs_time_add(cfs_time_current(),
 542                                         SOCKNAL_ENOMEM_RETRY),
 543                                         ksocknal_data.ksnd_reaper_waketime))
 544                         wake_up(&ksocknal_data.ksnd_reaper_waitq);
 545
 546                 spin_unlock_bh(&ksocknal_data.ksnd_reaper_lock);
 547                 return (rc);
 548         }
 549
 550         /* Actual error */
 551         LASSERT (rc < 0);
 552
 553         if (!conn->ksnc_closing) {
 554                 switch (rc) {
 555                 case -ECONNRESET:
 556                         LCONSOLE_WARN("Host %u.%u.%u.%u reset our connection "
 557                                       "while we were sending data; it may have "
 558                                       "rebooted.\n",
 559                                       HIPQUAD(conn->ksnc_ipaddr));
 560                         break;
 561                 default:
 562                         LCONSOLE_WARN("There was an unexpected network error "
 563                                       "while writing to %u.%u.%u.%u: %d.\n",
 564                                       HIPQUAD(conn->ksnc_ipaddr), rc);
 565                         break;
 566                 }
 567                 CDEBUG(D_NET, "[%p] Error %d on write to %s"
 568                        " ip %d.%d.%d.%d:%d\n", conn, rc,
 569                        libcfs_id2str(conn->ksnc_peer->ksnp_id),
 570                        HIPQUAD(conn->ksnc_ipaddr),
 571                        conn->ksnc_port);
 572         }
 573
 574         if (tx->tx_zc_checked)
 575                 ksocknal_uncheck_zc_req(tx);
 576
 577         /* it's not an error if conn is being closed */
 578         ksocknal_close_conn_and_siblings (conn,
 579                                           (conn->ksnc_closing) ? 0 : rc);
 580
 581         return (rc);
 582 }
 583
 584 void
 585 ksocknal_launch_connection_locked (ksock_route_t *route)
 586 {
 587
 588         /* called holding write lock on ksnd_global_lock */
 589
 590         LASSERT (!route->ksnr_scheduled);
 591         LASSERT (!route->ksnr_connecting);
 592         LASSERT ((ksocknal_route_mask() & ~route->ksnr_connected) != 0);
 593
 594         route->ksnr_scheduled = 1;              /* scheduling conn for connd */
 595         ksocknal_route_addref(route);           /* extra ref for connd */
 596
 597         spin_lock_bh(&ksocknal_data.ksnd_connd_lock);
 598
 599         cfs_list_add_tail(&route->ksnr_connd_list,
 600                           &ksocknal_data.ksnd_connd_routes);
 601         wake_up(&ksocknal_data.ksnd_connd_waitq);
 602
 603         spin_unlock_bh(&ksocknal_data.ksnd_connd_lock);
 604 }
 605
 606 void
 607 ksocknal_launch_all_connections_locked (ksock_peer_t *peer)
 608 {
 609         ksock_route_t *route;
 610
 611         /* called holding write lock on ksnd_global_lock */
 612         for (;;) {
 613                 /* launch any/all connections that need it */
 614                 route = ksocknal_find_connectable_route_locked(peer);
 615                 if (route == NULL)
 616                         return;
 617
 618                 ksocknal_launch_connection_locked(route);
 619         }
 620 }
 621
 622 ksock_conn_t *
 623 ksocknal_find_conn_locked(ksock_peer_t *peer, ksock_tx_t *tx, int nonblk)
 624 {
 625         cfs_list_t       *tmp;
 626         ksock_conn_t     *conn;
 627         ksock_conn_t     *typed = NULL;
 628         ksock_conn_t     *fallback = NULL;
 629         int               tnob     = 0;
 630         int               fnob     = 0;
 631
 632         cfs_list_for_each (tmp, &peer->ksnp_conns) {
 633                 ksock_conn_t *c  = cfs_list_entry(tmp, ksock_conn_t, ksnc_list);
 634                 int           nob = atomic_read(&c->ksnc_tx_nob) +
 635                                     libcfs_sock_wmem_queued(c->ksnc_sock);
 636                 int           rc;
 637
 638                 LASSERT (!c->ksnc_closing);
 639                 LASSERT (c->ksnc_proto != NULL &&
 640                          c->ksnc_proto->pro_match_tx != NULL);
 641
 642                 rc = c->ksnc_proto->pro_match_tx(c, tx, nonblk);
 643
 644                 switch (rc) {
 645                 default:
 646                         LBUG();
 647                 case SOCKNAL_MATCH_NO: /* protocol rejected the tx */
 648                         continue;
 649
 650                 case SOCKNAL_MATCH_YES: /* typed connection */
 651                         if (typed == NULL || tnob > nob ||
 652                             (tnob == nob && *ksocknal_tunables.ksnd_round_robin &&
 653                              cfs_time_after(typed->ksnc_tx_last_post, c->ksnc_tx_last_post))) {
 654                                 typed = c;
 655                                 tnob  = nob;
 656                         }
 657                         break;
 658
 659                 case SOCKNAL_MATCH_MAY: /* fallback connection */
 660                         if (fallback == NULL || fnob > nob ||
 661                             (fnob == nob && *ksocknal_tunables.ksnd_round_robin &&
 662                              cfs_time_after(fallback->ksnc_tx_last_post, c->ksnc_tx_last_post))) {
 663                                 fallback = c;
 664                                 fnob     = nob;
 665                         }
 666                         break;
 667                 }
 668         }
 669
 670         /* prefer the typed selection */
 671         conn = (typed != NULL) ? typed : fallback;
 672
 673         if (conn != NULL)
 674                 conn->ksnc_tx_last_post = cfs_time_current();
 675
 676         return conn;
 677 }
 678
 679 void
 680 ksocknal_tx_prep(ksock_conn_t *conn, ksock_tx_t *tx)
 681 {
 682         conn->ksnc_proto->pro_pack(tx);
 683
 684         atomic_add (tx->tx_nob, &conn->ksnc_tx_nob);
 685         ksocknal_conn_addref(conn); /* +1 ref for tx */
 686         tx->tx_conn = conn;
 687 }
 688
 689 void
 690 ksocknal_queue_tx_locked (ksock_tx_t *tx, ksock_conn_t *conn)
 691 {
 692         ksock_sched_t *sched = conn->ksnc_scheduler;
 693         ksock_msg_t   *msg = &tx->tx_msg;
 694         ksock_tx_t    *ztx = NULL;
 695         int            bufnob = 0;
 696
 697         /* called holding global lock (read or irq-write) and caller may
 698          * not have dropped this lock between finding conn and calling me,
 699          * so we don't need the {get,put}connsock dance to deref
 700          * ksnc_sock... */
 701         LASSERT(!conn->ksnc_closing);
 702
 703         CDEBUG (D_NET, "Sending to %s ip %d.%d.%d.%d:%d\n",
 704                 libcfs_id2str(conn->ksnc_peer->ksnp_id),
 705                 HIPQUAD(conn->ksnc_ipaddr),
 706                 conn->ksnc_port);
 707
 708         ksocknal_tx_prep(conn, tx);
 709
 710         /* Ensure the frags we've been given EXACTLY match the number of
 711          * bytes we want to send.  Many TCP/IP stacks disregard any total
 712          * size parameters passed to them and just look at the frags.
 713          *
 714          * We always expect at least 1 mapped fragment containing the
 715          * complete ksocknal message header. */
 716         LASSERT (lnet_iov_nob (tx->tx_niov, tx->tx_iov) +
 717                  lnet_kiov_nob(tx->tx_nkiov, tx->tx_kiov) ==
 718                  (unsigned int)tx->tx_nob);
 719         LASSERT (tx->tx_niov >= 1);
 720         LASSERT (tx->tx_resid == tx->tx_nob);
 721
 722         CDEBUG (D_NET, "Packet %p type %d, nob %d niov %d nkiov %d\n",
 723                 tx, (tx->tx_lnetmsg != NULL) ? tx->tx_lnetmsg->msg_hdr.type:
 724                                                KSOCK_MSG_NOOP,
 725                 tx->tx_nob, tx->tx_niov, tx->tx_nkiov);
 726
 727         /*
 728          * FIXME: SOCK_WMEM_QUEUED and SOCK_ERROR could block in __DARWIN8__
 729          * but they're used inside spinlocks a lot.
 730          */
 731         bufnob = libcfs_sock_wmem_queued(conn->ksnc_sock);
 732         spin_lock_bh(&sched->kss_lock);
 733
 734         if (cfs_list_empty(&conn->ksnc_tx_queue) && bufnob == 0) {
 735                 /* First packet starts the timeout */
 736                 conn->ksnc_tx_deadline =
 737                         cfs_time_shift(*ksocknal_tunables.ksnd_timeout);
 738                 if (conn->ksnc_tx_bufnob > 0) /* something got ACKed */
 739                         conn->ksnc_peer->ksnp_last_alive = cfs_time_current();
 740                 conn->ksnc_tx_bufnob = 0;
 741                 smp_mb(); /* order with adding to tx_queue */
 742         }
 743
 744         if (msg->ksm_type == KSOCK_MSG_NOOP) {
 745                 /* The packet is noop ZC ACK, try to piggyback the ack_cookie
 746                  * on a normal packet so I don't need to send it */
 747                 LASSERT (msg->ksm_zc_cookies[1] != 0);
 748                 LASSERT (conn->ksnc_proto->pro_queue_tx_zcack != NULL);
 749
 750                 if (conn->ksnc_proto->pro_queue_tx_zcack(conn, tx, 0))
 751                         ztx = tx; /* ZC ACK piggybacked on ztx release tx later */
 752
 753         } else {
 754                 /* It's a normal packet - can it piggback a noop zc-ack that
 755                  * has been queued already? */
 756                 LASSERT (msg->ksm_zc_cookies[1] == 0);
 757                 LASSERT (conn->ksnc_proto->pro_queue_tx_msg != NULL);
 758
 759                 ztx = conn->ksnc_proto->pro_queue_tx_msg(conn, tx);
 760                 /* ztx will be released later */
 761         }
 762
 763         if (ztx != NULL) {
 764                 atomic_sub (ztx->tx_nob, &conn->ksnc_tx_nob);
 765                 cfs_list_add_tail(&ztx->tx_list, &sched->kss_zombie_noop_txs);
 766         }
 767
 768         if (conn->ksnc_tx_ready &&      /* able to send */
 769             !conn->ksnc_tx_scheduled) { /* not scheduled to send */
 770                 /* +1 ref for scheduler */
 771                 ksocknal_conn_addref(conn);
 772                 cfs_list_add_tail (&conn->ksnc_tx_list,
 773                                    &sched->kss_tx_conns);
 774                 conn->ksnc_tx_scheduled = 1;
 775                 wake_up(&sched->kss_waitq);
 776         }
 777
 778         spin_unlock_bh(&sched->kss_lock);
 779 }
 780
 781
 782 ksock_route_t *
 783 ksocknal_find_connectable_route_locked (ksock_peer_t *peer)
 784 {
 785         cfs_time_t     now = cfs_time_current();
 786         cfs_list_t    *tmp;
 787         ksock_route_t *route;
 788
 789         cfs_list_for_each (tmp, &peer->ksnp_routes) {
 790                 route = cfs_list_entry (tmp, ksock_route_t, ksnr_list);
 791
 792                 LASSERT (!route->ksnr_connecting || route->ksnr_scheduled);
 793
 794                 if (route->ksnr_scheduled)      /* connections being established */
 795                         continue;
 796
 797                 /* all route types connected ? */
 798                 if ((ksocknal_route_mask() & ~route->ksnr_connected) == 0)
 799                         continue;
 800
 801                 if (!(route->ksnr_retry_interval == 0 || /* first attempt */
 802                       cfs_time_aftereq(now, route->ksnr_timeout))) {
 803                         CDEBUG(D_NET,
 804                                "Too soon to retry route %u.%u.%u.%u "
 805                                "(cnted %d, interval %ld, %ld secs later)\n",
 806                                HIPQUAD(route->ksnr_ipaddr),
 807                                route->ksnr_connected,
 808                                route->ksnr_retry_interval,
 809                                cfs_duration_sec(route->ksnr_timeout - now));
 810                         continue;
 811                 }
 812
 813                 return (route);
 814         }
 815
 816         return (NULL);
 817 }
 818
 819 ksock_route_t *
 820 ksocknal_find_connecting_route_locked (ksock_peer_t *peer)
 821 {
 822         cfs_list_t        *tmp;
 823         ksock_route_t     *route;
 824
 825         cfs_list_for_each (tmp, &peer->ksnp_routes) {
 826                 route = cfs_list_entry (tmp, ksock_route_t, ksnr_list);
 827
 828                 LASSERT (!route->ksnr_connecting || route->ksnr_scheduled);
 829
 830                 if (route->ksnr_scheduled)
 831                         return (route);
 832         }
 833
 834         return (NULL);
 835 }
 836
 837 int
 838 ksocknal_launch_packet (lnet_ni_t *ni, ksock_tx_t *tx, lnet_process_id_t id)
 839 {
 840         ksock_peer_t     *peer;
 841         ksock_conn_t     *conn;
 842         rwlock_t     *g_lock;
 843         int               retry;
 844         int               rc;
 845
 846         LASSERT (tx->tx_conn == NULL);
 847
 848         g_lock = &ksocknal_data.ksnd_global_lock;
 849
 850         for (retry = 0;; retry = 1) {
 851                 read_lock(g_lock);
 852                 peer = ksocknal_find_peer_locked(ni, id);
 853                 if (peer != NULL) {
 854                         if (ksocknal_find_connectable_route_locked(peer) == NULL) {
 855                                 conn = ksocknal_find_conn_locked(peer, tx, tx->tx_nonblk);
 856                                 if (conn != NULL) {
 857                                         /* I've got no routes that need to be
 858                                          * connecting and I do have an actual
 859                                          * connection... */
 860                                         ksocknal_queue_tx_locked (tx, conn);
 861                                         read_unlock(g_lock);
 862                                         return (0);
 863                                 }
 864                         }
 865                 }
 866
 867                 /* I'll need a write lock... */
 868                 read_unlock(g_lock);
 869
 870                 write_lock_bh(g_lock);
 871
 872                 peer = ksocknal_find_peer_locked(ni, id);
 873                 if (peer != NULL)
 874                         break;
 875
 876                 write_unlock_bh(g_lock);
 877
 878                 if ((id.pid & LNET_PID_USERFLAG) != 0) {
 879                         CERROR("Refusing to create a connection to "
 880                                "userspace process %s\n", libcfs_id2str(id));
 881                         return -EHOSTUNREACH;
 882                 }
 883
 884                 if (retry) {
 885                         CERROR("Can't find peer %s\n", libcfs_id2str(id));
 886                         return -EHOSTUNREACH;
 887                 }
 888
 889                 rc = ksocknal_add_peer(ni, id,
 890                                        LNET_NIDADDR(id.nid),
 891                                        lnet_acceptor_port());
 892                 if (rc != 0) {
 893                         CERROR("Can't add peer %s: %d\n",
 894                                libcfs_id2str(id), rc);
 895                         return rc;
 896                 }
 897         }
 898
 899         ksocknal_launch_all_connections_locked(peer);
 900
 901         conn = ksocknal_find_conn_locked(peer, tx, tx->tx_nonblk);
 902         if (conn != NULL) {
 903                 /* Connection exists; queue message on it */
 904                 ksocknal_queue_tx_locked (tx, conn);
 905                 write_unlock_bh(g_lock);
 906                 return (0);
 907         }
 908
 909         if (peer->ksnp_accepting > 0 ||
 910             ksocknal_find_connecting_route_locked (peer) != NULL) {
 911                 /* the message is going to be pinned to the peer */
 912                 tx->tx_deadline =
 913                         cfs_time_shift(*ksocknal_tunables.ksnd_timeout);
 914
 915                 /* Queue the message until a connection is established */
 916                 cfs_list_add_tail (&tx->tx_list, &peer->ksnp_tx_queue);
 917                 write_unlock_bh(g_lock);
 918                 return 0;
 919         }
 920
 921         write_unlock_bh(g_lock);
 922
 923         /* NB Routes may be ignored if connections to them failed recently */
 924         CNETERR("No usable routes to %s\n", libcfs_id2str(id));
 925         return (-EHOSTUNREACH);
 926 }
 927
 928 int
 929 ksocknal_send(lnet_ni_t *ni, void *private, lnet_msg_t *lntmsg)
 930 {
 931         int               mpflag = 1;
 932         int               type = lntmsg->msg_type;
 933         lnet_process_id_t target = lntmsg->msg_target;
 934         unsigned int      payload_niov = lntmsg->msg_niov;
 935         struct iovec     *payload_iov = lntmsg->msg_iov;
 936         lnet_kiov_t      *payload_kiov = lntmsg->msg_kiov;
 937         unsigned int      payload_offset = lntmsg->msg_offset;
 938         unsigned int      payload_nob = lntmsg->msg_len;
 939         ksock_tx_t       *tx;
 940         int               desc_size;
 941         int               rc;
 942
 943         /* NB 'private' is different depending on what we're sending.
 944          * Just ignore it... */
 945
 946         CDEBUG(D_NET, "sending %u bytes in %d frags to %s\n",
 947                payload_nob, payload_niov, libcfs_id2str(target));
 948
 949         LASSERT (payload_nob == 0 || payload_niov > 0);
 950         LASSERT (payload_niov <= LNET_MAX_IOV);
 951         /* payload is either all vaddrs or all pages */
 952         LASSERT (!(payload_kiov != NULL && payload_iov != NULL));
 953         LASSERT (!in_interrupt ());
 954
 955         if (payload_iov != NULL)
 956                 desc_size = offsetof(ksock_tx_t,
 957                                      tx_frags.virt.iov[1 + payload_niov]);
 958         else
 959                 desc_size = offsetof(ksock_tx_t,
 960                                      tx_frags.paged.kiov[payload_niov]);
 961
 962         if (lntmsg->msg_vmflush)
 963                 mpflag = cfs_memory_pressure_get_and_set();
 964         tx = ksocknal_alloc_tx(KSOCK_MSG_LNET, desc_size);
 965         if (tx == NULL) {
 966                 CERROR("Can't allocate tx desc type %d size %d\n",
 967                        type, desc_size);
 968                 if (lntmsg->msg_vmflush)
 969                         cfs_memory_pressure_restore(mpflag);
 970                 return (-ENOMEM);
 971         }
 972
 973         tx->tx_conn = NULL;                     /* set when assigned a conn */
 974         tx->tx_lnetmsg = lntmsg;
 975
 976         if (payload_iov != NULL) {
 977                 tx->tx_kiov = NULL;
 978                 tx->tx_nkiov = 0;
 979                 tx->tx_iov = tx->tx_frags.virt.iov;
 980                 tx->tx_niov = 1 +
 981                               lnet_extract_iov(payload_niov, &tx->tx_iov[1],
 982                                                payload_niov, payload_iov,
 983                                                payload_offset, payload_nob);
 984         } else {
 985                 tx->tx_niov = 1;
 986                 tx->tx_iov = &tx->tx_frags.paged.iov;
 987                 tx->tx_kiov = tx->tx_frags.paged.kiov;
 988                 tx->tx_nkiov = lnet_extract_kiov(payload_niov, tx->tx_kiov,
 989                                                  payload_niov, payload_kiov,
 990                                                  payload_offset, payload_nob);
 991
 992                 if (payload_nob >= *ksocknal_tunables.ksnd_zc_min_payload)
 993                         tx->tx_zc_capable = 1;
 994         }
 995
 996         socklnd_init_msg(&tx->tx_msg, KSOCK_MSG_LNET);
 997
 998         /* The first fragment will be set later in pro_pack */
 999         rc = ksocknal_launch_packet(ni, tx, target);
1000         if (!mpflag)
1001                 cfs_memory_pressure_restore(mpflag);
1002
1003         if (rc == 0)
1004                 return (0);
1005
1006         ksocknal_free_tx(tx);
1007         return (-EIO);
1008 }
1009
1010 int
1011 ksocknal_thread_start(int (*fn)(void *arg), void *arg, char *name)
1012 {
1013         struct task_struct *task = kthread_run(fn, arg, name);
1014
1015         if (IS_ERR(task))
1016                 return PTR_ERR(task);
1017
1018         write_lock_bh(&ksocknal_data.ksnd_global_lock);
1019         ksocknal_data.ksnd_nthreads++;
1020         write_unlock_bh(&ksocknal_data.ksnd_global_lock);
1021         return 0;
1022 }
1023
1024 void
1025 ksocknal_thread_fini (void)
1026 {
1027         write_lock_bh(&ksocknal_data.ksnd_global_lock);
1028         ksocknal_data.ksnd_nthreads--;
1029         write_unlock_bh(&ksocknal_data.ksnd_global_lock);
1030 }
1031
1032 int
1033 ksocknal_new_packet (ksock_conn_t *conn, int nob_to_skip)
1034 {
1035         static char ksocknal_slop_buffer[4096];
1036
1037         int            nob;
1038         unsigned int   niov;
1039         int            skipped;
1040
1041         LASSERT(conn->ksnc_proto != NULL);
1042
1043         if ((*ksocknal_tunables.ksnd_eager_ack & conn->ksnc_type) != 0) {
1044                 /* Remind the socket to ack eagerly... */
1045                 ksocknal_lib_eager_ack(conn);
1046         }
1047
1048         if (nob_to_skip == 0) {         /* right at next packet boundary now */
1049                 conn->ksnc_rx_started = 0;
1050                 smp_mb();                       /* racing with timeout thread */
1051
1052                 switch (conn->ksnc_proto->pro_version) {
1053                 case  KSOCK_PROTO_V2:
1054                 case  KSOCK_PROTO_V3:
1055                         conn->ksnc_rx_state = SOCKNAL_RX_KSM_HEADER;
1056                         conn->ksnc_rx_iov = (struct iovec *)&conn->ksnc_rx_iov_space;
1057                         conn->ksnc_rx_iov[0].iov_base = (char *)&conn->ksnc_msg;
1058
1059                         conn->ksnc_rx_nob_wanted = offsetof(ksock_msg_t, ksm_u);
1060                         conn->ksnc_rx_nob_left = offsetof(ksock_msg_t, ksm_u);
1061                         conn->ksnc_rx_iov[0].iov_len  = offsetof(ksock_msg_t, ksm_u);
1062                         break;
1063
1064                 case KSOCK_PROTO_V1:
1065                         /* Receiving bare lnet_hdr_t */
1066                         conn->ksnc_rx_state = SOCKNAL_RX_LNET_HEADER;
1067                         conn->ksnc_rx_nob_wanted = sizeof(lnet_hdr_t);
1068                         conn->ksnc_rx_nob_left = sizeof(lnet_hdr_t);
1069
1070                         conn->ksnc_rx_iov = (struct iovec *)&conn->ksnc_rx_iov_space;
1071                         conn->ksnc_rx_iov[0].iov_base = (char *)&conn->ksnc_msg.ksm_u.lnetmsg;
1072                         conn->ksnc_rx_iov[0].iov_len  = sizeof (lnet_hdr_t);
1073                         break;
1074
1075                 default:
1076                         LBUG ();
1077                 }
1078                 conn->ksnc_rx_niov = 1;
1079
1080                 conn->ksnc_rx_kiov = NULL;
1081                 conn->ksnc_rx_nkiov = 0;
1082                 conn->ksnc_rx_csum = ~0;
1083                 return (1);
1084         }
1085
1086         /* Set up to skip as much as possible now.  If there's more left
1087          * (ran out of iov entries) we'll get called again */
1088
1089         conn->ksnc_rx_state = SOCKNAL_RX_SLOP;
1090         conn->ksnc_rx_nob_left = nob_to_skip;
1091         conn->ksnc_rx_iov = (struct iovec *)&conn->ksnc_rx_iov_space;
1092         skipped = 0;
1093         niov = 0;
1094
1095         do {
1096                 nob = MIN (nob_to_skip, sizeof (ksocknal_slop_buffer));
1097
1098                 conn->ksnc_rx_iov[niov].iov_base = ksocknal_slop_buffer;
1099                 conn->ksnc_rx_iov[niov].iov_len  = nob;
1100                 niov++;
1101                 skipped += nob;
1102                 nob_to_skip -=nob;
1103
1104         } while (nob_to_skip != 0 &&    /* mustn't overflow conn's rx iov */
1105                  niov < sizeof(conn->ksnc_rx_iov_space) / sizeof (struct iovec));
1106
1107         conn->ksnc_rx_niov = niov;
1108         conn->ksnc_rx_kiov = NULL;
1109         conn->ksnc_rx_nkiov = 0;
1110         conn->ksnc_rx_nob_wanted = skipped;
1111         return (0);
1112 }
1113
1114 int
1115 ksocknal_process_receive (ksock_conn_t *conn)
1116 {
1117         lnet_hdr_t        *lhdr;
1118         lnet_process_id_t *id;
1119         int                rc;
1120
1121         LASSERT (atomic_read(&conn->ksnc_conn_refcount) > 0);
1122
1123         /* NB: sched lock NOT held */
1124         /* SOCKNAL_RX_LNET_HEADER is here for backward compatability */
1125         LASSERT (conn->ksnc_rx_state == SOCKNAL_RX_KSM_HEADER ||
1126                  conn->ksnc_rx_state == SOCKNAL_RX_LNET_PAYLOAD ||
1127                  conn->ksnc_rx_state == SOCKNAL_RX_LNET_HEADER ||
1128                  conn->ksnc_rx_state == SOCKNAL_RX_SLOP);
1129  again:
1130         if (conn->ksnc_rx_nob_wanted != 0) {
1131                 rc = ksocknal_receive(conn);
1132
1133                 if (rc <= 0) {
1134                         LASSERT (rc != -EAGAIN);
1135
1136                         if (rc == 0)
1137                                 CDEBUG (D_NET, "[%p] EOF from %s"
1138                                         " ip %d.%d.%d.%d:%d\n", conn,
1139                                         libcfs_id2str(conn->ksnc_peer->ksnp_id),
1140                                         HIPQUAD(conn->ksnc_ipaddr),
1141                                         conn->ksnc_port);
1142                         else if (!conn->ksnc_closing)
1143                                 CERROR ("[%p] Error %d on read from %s"
1144                                         " ip %d.%d.%d.%d:%d\n",
1145                                         conn, rc,
1146                                         libcfs_id2str(conn->ksnc_peer->ksnp_id),
1147                                         HIPQUAD(conn->ksnc_ipaddr),
1148                                         conn->ksnc_port);
1149
1150                         /* it's not an error if conn is being closed */
1151                         ksocknal_close_conn_and_siblings (conn,
1152                                                           (conn->ksnc_closing) ? 0 : rc);
1153                         return (rc == 0 ? -ESHUTDOWN : rc);
1154                 }
1155
1156                 if (conn->ksnc_rx_nob_wanted != 0) {
1157                         /* short read */
1158                         return (-EAGAIN);
1159                 }
1160         }
1161         switch (conn->ksnc_rx_state) {
1162         case SOCKNAL_RX_KSM_HEADER:
1163                 if (conn->ksnc_flip) {
1164                         __swab32s(&conn->ksnc_msg.ksm_type);
1165                         __swab32s(&conn->ksnc_msg.ksm_csum);
1166                         __swab64s(&conn->ksnc_msg.ksm_zc_cookies[0]);
1167                         __swab64s(&conn->ksnc_msg.ksm_zc_cookies[1]);
1168                 }
1169
1170                 if (conn->ksnc_msg.ksm_type != KSOCK_MSG_NOOP &&
1171                     conn->ksnc_msg.ksm_type != KSOCK_MSG_LNET) {
1172                         CERROR("%s: Unknown message type: %x\n",
1173                                libcfs_id2str(conn->ksnc_peer->ksnp_id),
1174                                conn->ksnc_msg.ksm_type);
1175                         ksocknal_new_packet(conn, 0);
1176                         ksocknal_close_conn_and_siblings(conn, -EPROTO);
1177                         return (-EPROTO);
1178                 }
1179
1180                 if (conn->ksnc_msg.ksm_type == KSOCK_MSG_NOOP &&
1181                     conn->ksnc_msg.ksm_csum != 0 &&     /* has checksum */
1182                     conn->ksnc_msg.ksm_csum != conn->ksnc_rx_csum) {
1183                         /* NOOP Checksum error */
1184                         CERROR("%s: Checksum error, wire:0x%08X data:0x%08X\n",
1185                                libcfs_id2str(conn->ksnc_peer->ksnp_id),
1186                                conn->ksnc_msg.ksm_csum, conn->ksnc_rx_csum);
1187                         ksocknal_new_packet(conn, 0);
1188                         ksocknal_close_conn_and_siblings(conn, -EPROTO);
1189                         return (-EIO);
1190                 }
1191
1192                 if (conn->ksnc_msg.ksm_zc_cookies[1] != 0) {
1193                         __u64 cookie = 0;
1194
1195                         LASSERT (conn->ksnc_proto != &ksocknal_protocol_v1x);
1196
1197                         if (conn->ksnc_msg.ksm_type == KSOCK_MSG_NOOP)
1198                                 cookie = conn->ksnc_msg.ksm_zc_cookies[0];
1199
1200                         rc = conn->ksnc_proto->pro_handle_zcack(conn, cookie,
1201                                                conn->ksnc_msg.ksm_zc_cookies[1]);
1202
1203                         if (rc != 0) {
1204                                 CERROR("%s: Unknown ZC-ACK cookie: "LPU64", "LPU64"\n",
1205                                        libcfs_id2str(conn->ksnc_peer->ksnp_id),
1206                                        cookie, conn->ksnc_msg.ksm_zc_cookies[1]);
1207                                 ksocknal_new_packet(conn, 0);
1208                                 ksocknal_close_conn_and_siblings(conn, -EPROTO);
1209                                 return (rc);
1210                         }
1211                 }
1212
1213                 if (conn->ksnc_msg.ksm_type == KSOCK_MSG_NOOP) {
1214                         ksocknal_new_packet (conn, 0);
1215                         return 0;       /* NOOP is done and just return */
1216                 }
1217
1218                 conn->ksnc_rx_state = SOCKNAL_RX_LNET_HEADER;
1219                 conn->ksnc_rx_nob_wanted = sizeof(ksock_lnet_msg_t);
1220                 conn->ksnc_rx_nob_left = sizeof(ksock_lnet_msg_t);
1221
1222                 conn->ksnc_rx_iov = (struct iovec *)&conn->ksnc_rx_iov_space;
1223                 conn->ksnc_rx_iov[0].iov_base = (char *)&conn->ksnc_msg.ksm_u.lnetmsg;
1224                 conn->ksnc_rx_iov[0].iov_len  = sizeof(ksock_lnet_msg_t);
1225
1226                 conn->ksnc_rx_niov = 1;
1227                 conn->ksnc_rx_kiov = NULL;
1228                 conn->ksnc_rx_nkiov = 0;
1229
1230                 goto again;     /* read lnet header now */
1231
1232         case SOCKNAL_RX_LNET_HEADER:
1233                 /* unpack message header */
1234                 conn->ksnc_proto->pro_unpack(&conn->ksnc_msg);
1235
1236                 if ((conn->ksnc_peer->ksnp_id.pid & LNET_PID_USERFLAG) != 0) {
1237                         /* Userspace peer */
1238                         lhdr = &conn->ksnc_msg.ksm_u.lnetmsg.ksnm_hdr;
1239                         id   = &conn->ksnc_peer->ksnp_id;
1240
1241                         /* Substitute process ID assigned at connection time */
1242                         lhdr->src_pid = cpu_to_le32(id->pid);
1243                         lhdr->src_nid = cpu_to_le64(id->nid);
1244                 }
1245
1246                 conn->ksnc_rx_state = SOCKNAL_RX_PARSE;
1247                 ksocknal_conn_addref(conn);     /* ++ref while parsing */
1248
1249                 rc = lnet_parse(conn->ksnc_peer->ksnp_ni,
1250                                 &conn->ksnc_msg.ksm_u.lnetmsg.ksnm_hdr,
1251                                 conn->ksnc_peer->ksnp_id.nid, conn, 0);
1252                 if (rc < 0) {
1253                         /* I just received garbage: give up on this conn */
1254                         ksocknal_new_packet(conn, 0);
1255                         ksocknal_close_conn_and_siblings (conn, rc);
1256                         ksocknal_conn_decref(conn);
1257                         return (-EPROTO);
1258                 }
1259
1260                 /* I'm racing with ksocknal_recv() */
1261                 LASSERT (conn->ksnc_rx_state == SOCKNAL_RX_PARSE ||
1262                          conn->ksnc_rx_state == SOCKNAL_RX_LNET_PAYLOAD);
1263
1264                 if (conn->ksnc_rx_state != SOCKNAL_RX_LNET_PAYLOAD)
1265                         return 0;
1266
1267                 /* ksocknal_recv() got called */
1268                 goto again;
1269
1270         case SOCKNAL_RX_LNET_PAYLOAD:
1271                 /* payload all received */
1272                 rc = 0;
1273
1274                 if (conn->ksnc_rx_nob_left == 0 &&   /* not truncating */
1275                     conn->ksnc_msg.ksm_csum != 0 &&  /* has checksum */
1276                     conn->ksnc_msg.ksm_csum != conn->ksnc_rx_csum) {
1277                         CERROR("%s: Checksum error, wire:0x%08X data:0x%08X\n",
1278                                libcfs_id2str(conn->ksnc_peer->ksnp_id),
1279                                conn->ksnc_msg.ksm_csum, conn->ksnc_rx_csum);
1280                         rc = -EIO;
1281                 }
1282
1283                 if (rc == 0 && conn->ksnc_msg.ksm_zc_cookies[0] != 0) {
1284                         LASSERT(conn->ksnc_proto != &ksocknal_protocol_v1x);
1285
1286                         lhdr = &conn->ksnc_msg.ksm_u.lnetmsg.ksnm_hdr;
1287                         id   = &conn->ksnc_peer->ksnp_id;
1288
1289                         rc = conn->ksnc_proto->pro_handle_zcreq(conn,
1290                                         conn->ksnc_msg.ksm_zc_cookies[0],
1291                                         *ksocknal_tunables.ksnd_nonblk_zcack ||
1292                                         le64_to_cpu(lhdr->src_nid) != id->nid);
1293                 }
1294
1295                 lnet_finalize(conn->ksnc_peer->ksnp_ni, conn->ksnc_cookie, rc);
1296
1297                 if (rc != 0) {
1298                         ksocknal_new_packet(conn, 0);
1299                         ksocknal_close_conn_and_siblings (conn, rc);
1300                         return (-EPROTO);
1301                 }
1302                 /* Fall through */
1303
1304         case SOCKNAL_RX_SLOP:
1305                 /* starting new packet? */
1306                 if (ksocknal_new_packet (conn, conn->ksnc_rx_nob_left))
1307                         return 0;       /* come back later */
1308                 goto again;             /* try to finish reading slop now */
1309
1310         default:
1311                 break;
1312         }
1313
1314         /* Not Reached */
1315         LBUG ();
1316         return (-EINVAL);                       /* keep gcc happy */
1317 }
1318
1319 int
1320 ksocknal_recv (lnet_ni_t *ni, void *private, lnet_msg_t *msg, int delayed,
1321                unsigned int niov, struct iovec *iov, lnet_kiov_t *kiov,
1322                unsigned int offset, unsigned int mlen, unsigned int rlen)
1323 {
1324         ksock_conn_t  *conn = (ksock_conn_t *)private;
1325         ksock_sched_t *sched = conn->ksnc_scheduler;
1326
1327         LASSERT (mlen <= rlen);
1328         LASSERT (niov <= LNET_MAX_IOV);
1329
1330         conn->ksnc_cookie = msg;
1331         conn->ksnc_rx_nob_wanted = mlen;
1332         conn->ksnc_rx_nob_left   = rlen;
1333
1334         if (mlen == 0 || iov != NULL) {
1335                 conn->ksnc_rx_nkiov = 0;
1336                 conn->ksnc_rx_kiov = NULL;
1337                 conn->ksnc_rx_iov = conn->ksnc_rx_iov_space.iov;
1338                 conn->ksnc_rx_niov =
1339                         lnet_extract_iov(LNET_MAX_IOV, conn->ksnc_rx_iov,
1340                                          niov, iov, offset, mlen);
1341         } else {
1342                 conn->ksnc_rx_niov = 0;
1343                 conn->ksnc_rx_iov  = NULL;
1344                 conn->ksnc_rx_kiov = conn->ksnc_rx_iov_space.kiov;
1345                 conn->ksnc_rx_nkiov =
1346                         lnet_extract_kiov(LNET_MAX_IOV, conn->ksnc_rx_kiov,
1347                                           niov, kiov, offset, mlen);
1348         }
1349
1350         LASSERT (mlen ==
1351                  lnet_iov_nob (conn->ksnc_rx_niov, conn->ksnc_rx_iov) +
1352                  lnet_kiov_nob (conn->ksnc_rx_nkiov, conn->ksnc_rx_kiov));
1353
1354         LASSERT (conn->ksnc_rx_scheduled);
1355
1356         spin_lock_bh(&sched->kss_lock);
1357
1358         switch (conn->ksnc_rx_state) {
1359         case SOCKNAL_RX_PARSE_WAIT:
1360                 cfs_list_add_tail(&conn->ksnc_rx_list, &sched->kss_rx_conns);
1361                 wake_up(&sched->kss_waitq);
1362                 LASSERT(conn->ksnc_rx_ready);
1363                 break;
1364
1365         case SOCKNAL_RX_PARSE:
1366                 /* scheduler hasn't noticed I'm parsing yet */
1367                 break;
1368         }
1369
1370         conn->ksnc_rx_state = SOCKNAL_RX_LNET_PAYLOAD;
1371
1372         spin_unlock_bh(&sched->kss_lock);
1373         ksocknal_conn_decref(conn);
1374         return 0;
1375 }
1376
1377 static inline int
1378 ksocknal_sched_cansleep(ksock_sched_t *sched)
1379 {
1380         int           rc;
1381
1382         spin_lock_bh(&sched->kss_lock);
1383
1384         rc = (!ksocknal_data.ksnd_shuttingdown &&
1385               cfs_list_empty(&sched->kss_rx_conns) &&
1386               cfs_list_empty(&sched->kss_tx_conns));
1387
1388         spin_unlock_bh(&sched->kss_lock);
1389         return rc;
1390 }
1391
1392 int ksocknal_scheduler(void *arg)
1393 {
1394         struct ksock_sched_info *info;
1395         ksock_sched_t           *sched;
1396         ksock_conn_t            *conn;
1397         ksock_tx_t              *tx;
1398         int                     rc;
1399         int                     nloops = 0;
1400         long                    id = (long)arg;
1401
1402         info = ksocknal_data.ksnd_sched_info[KSOCK_THREAD_CPT(id)];
1403         sched = &info->ksi_scheds[KSOCK_THREAD_SID(id)];
1404
1405         cfs_block_allsigs();
1406
1407         rc = cfs_cpt_bind(lnet_cpt_table(), info->ksi_cpt);
1408         if (rc != 0) {
1409                 CERROR("Can't set CPT affinity to %d: %d\n",
1410                        info->ksi_cpt, rc);
1411         }
1412
1413         spin_lock_bh(&sched->kss_lock);
1414
1415         while (!ksocknal_data.ksnd_shuttingdown) {
1416                 int did_something = 0;
1417
1418                 /* Ensure I progress everything semi-fairly */
1419
1420                 if (!cfs_list_empty (&sched->kss_rx_conns)) {
1421                         conn = cfs_list_entry(sched->kss_rx_conns.next,
1422                                               ksock_conn_t, ksnc_rx_list);
1423                         cfs_list_del(&conn->ksnc_rx_list);
1424
1425                         LASSERT(conn->ksnc_rx_scheduled);
1426                         LASSERT(conn->ksnc_rx_ready);
1427
1428                         /* clear rx_ready in case receive isn't complete.
1429                          * Do it BEFORE we call process_recv, since
1430                          * data_ready can set it any time after we release
1431                          * kss_lock. */
1432                         conn->ksnc_rx_ready = 0;
1433                         spin_unlock_bh(&sched->kss_lock);
1434
1435                         rc = ksocknal_process_receive(conn);
1436
1437                         spin_lock_bh(&sched->kss_lock);
1438
1439                         /* I'm the only one that can clear this flag */
1440                         LASSERT(conn->ksnc_rx_scheduled);
1441
1442                         /* Did process_receive get everything it wanted? */
1443                         if (rc == 0)
1444                                 conn->ksnc_rx_ready = 1;
1445
1446                         if (conn->ksnc_rx_state == SOCKNAL_RX_PARSE) {
1447                                 /* Conn blocked waiting for ksocknal_recv()
1448                                  * I change its state (under lock) to signal
1449                                  * it can be rescheduled */
1450                                 conn->ksnc_rx_state = SOCKNAL_RX_PARSE_WAIT;
1451                         } else if (conn->ksnc_rx_ready) {
1452                                 /* reschedule for rx */
1453                                 cfs_list_add_tail (&conn->ksnc_rx_list,
1454                                                    &sched->kss_rx_conns);
1455                         } else {
1456                                 conn->ksnc_rx_scheduled = 0;
1457                                 /* drop my ref */
1458                                 ksocknal_conn_decref(conn);
1459                         }
1460
1461                         did_something = 1;
1462                 }
1463
1464                 if (!cfs_list_empty (&sched->kss_tx_conns)) {
1465                         CFS_LIST_HEAD    (zlist);
1466
1467                         if (!cfs_list_empty(&sched->kss_zombie_noop_txs)) {
1468                                 cfs_list_add(&zlist,
1469                                              &sched->kss_zombie_noop_txs);
1470                                 cfs_list_del_init(&sched->kss_zombie_noop_txs);
1471                         }
1472
1473                         conn = cfs_list_entry(sched->kss_tx_conns.next,
1474                                               ksock_conn_t, ksnc_tx_list);
1475                         cfs_list_del (&conn->ksnc_tx_list);
1476
1477                         LASSERT(conn->ksnc_tx_scheduled);
1478                         LASSERT(conn->ksnc_tx_ready);
1479                         LASSERT(!cfs_list_empty(&conn->ksnc_tx_queue));
1480
1481                         tx = cfs_list_entry(conn->ksnc_tx_queue.next,
1482                                             ksock_tx_t, tx_list);
1483
1484                         if (conn->ksnc_tx_carrier == tx)
1485                                 ksocknal_next_tx_carrier(conn);
1486
1487                         /* dequeue now so empty list => more to send */
1488                         cfs_list_del(&tx->tx_list);
1489
1490                         /* Clear tx_ready in case send isn't complete.  Do
1491                          * it BEFORE we call process_transmit, since
1492                          * write_space can set it any time after we release
1493                          * kss_lock. */
1494                         conn->ksnc_tx_ready = 0;
1495                         spin_unlock_bh(&sched->kss_lock);
1496
1497                         if (!cfs_list_empty(&zlist)) {
1498                                 /* free zombie noop txs, it's fast because
1499                                  * noop txs are just put in freelist */
1500                                 ksocknal_txlist_done(NULL, &zlist, 0);
1501                         }
1502
1503                         rc = ksocknal_process_transmit(conn, tx);
1504
1505                         if (rc == -ENOMEM || rc == -EAGAIN) {
1506                                 /* Incomplete send: replace tx on HEAD of tx_queue */
1507                                 spin_lock_bh(&sched->kss_lock);
1508                                 cfs_list_add(&tx->tx_list,
1509                                              &conn->ksnc_tx_queue);
1510                         } else {
1511                                 /* Complete send; tx -ref */
1512                                 ksocknal_tx_decref(tx);
1513
1514                                 spin_lock_bh(&sched->kss_lock);
1515                                 /* assume space for more */
1516                                 conn->ksnc_tx_ready = 1;
1517                         }
1518
1519                         if (rc == -ENOMEM) {
1520                                 /* Do nothing; after a short timeout, this
1521                                  * conn will be reposted on kss_tx_conns. */
1522                         } else if (conn->ksnc_tx_ready &&
1523                                    !cfs_list_empty (&conn->ksnc_tx_queue)) {
1524                                 /* reschedule for tx */
1525                                 cfs_list_add_tail (&conn->ksnc_tx_list,
1526                                                    &sched->kss_tx_conns);
1527                         } else {
1528                                 conn->ksnc_tx_scheduled = 0;
1529                                 /* drop my ref */
1530                                 ksocknal_conn_decref(conn);
1531                         }
1532
1533                         did_something = 1;
1534                 }
1535                 if (!did_something ||           /* nothing to do */
1536                     ++nloops == SOCKNAL_RESCHED) { /* hogging CPU? */
1537                         spin_unlock_bh(&sched->kss_lock);
1538
1539                         nloops = 0;
1540
1541                         if (!did_something) {   /* wait for something to do */
1542                                 rc = wait_event_interruptible_exclusive(
1543                                         sched->kss_waitq,
1544                                         !ksocknal_sched_cansleep(sched));
1545                                 LASSERT (rc == 0);
1546                         } else {
1547                                 cond_resched();
1548                         }
1549
1550                         spin_lock_bh(&sched->kss_lock);
1551                 }
1552         }
1553
1554         spin_unlock_bh(&sched->kss_lock);
1555         ksocknal_thread_fini();
1556         return 0;
1557 }
1558
1559 /*
1560  * Add connection to kss_rx_conns of scheduler
1561  * and wakeup the scheduler.
1562  */
1563 void ksocknal_read_callback (ksock_conn_t *conn)
1564 {
1565         ksock_sched_t *sched;
1566         ENTRY;
1567
1568         sched = conn->ksnc_scheduler;
1569
1570         spin_lock_bh(&sched->kss_lock);
1571
1572         conn->ksnc_rx_ready = 1;
1573
1574         if (!conn->ksnc_rx_scheduled) {  /* not being progressed */
1575                 cfs_list_add_tail(&conn->ksnc_rx_list,
1576                                   &sched->kss_rx_conns);
1577                 conn->ksnc_rx_scheduled = 1;
1578                 /* extra ref for scheduler */
1579                 ksocknal_conn_addref(conn);
1580
1581                 wake_up (&sched->kss_waitq);
1582         }
1583         spin_unlock_bh(&sched->kss_lock);
1584
1585         EXIT;
1586 }
1587
1588 /*
1589  * Add connection to kss_tx_conns of scheduler
1590  * and wakeup the scheduler.
1591  */
1592 void ksocknal_write_callback(ksock_conn_t *conn)
1593 {
1594         ksock_sched_t *sched;
1595         ENTRY;
1596
1597         sched = conn->ksnc_scheduler;
1598
1599         spin_lock_bh(&sched->kss_lock);
1600
1601         conn->ksnc_tx_ready = 1;
1602
1603         if (!conn->ksnc_tx_scheduled && // not being progressed
1604             !cfs_list_empty(&conn->ksnc_tx_queue)){//packets to send
1605                 cfs_list_add_tail(&conn->ksnc_tx_list, &sched->kss_tx_conns);
1606                 conn->ksnc_tx_scheduled = 1;
1607                 /* extra ref for scheduler */
1608                 ksocknal_conn_addref(conn);
1609
1610                 wake_up(&sched->kss_waitq);
1611         }
1612
1613         spin_unlock_bh(&sched->kss_lock);
1614
1615         EXIT;
1616 }
1617
1618 ksock_proto_t *
1619 ksocknal_parse_proto_version (ksock_hello_msg_t *hello)
1620 {
1621         __u32   version = 0;
1622
1623         if (hello->kshm_magic == LNET_PROTO_MAGIC)
1624                 version = hello->kshm_version;
1625         else if (hello->kshm_magic == __swab32(LNET_PROTO_MAGIC))
1626                 version = __swab32(hello->kshm_version);
1627
1628         if (version != 0) {
1629 #if SOCKNAL_VERSION_DEBUG
1630                 if (*ksocknal_tunables.ksnd_protocol == 1)
1631                         return NULL;
1632
1633                 if (*ksocknal_tunables.ksnd_protocol == 2 &&
1634                     version == KSOCK_PROTO_V3)
1635                         return NULL;
1636 #endif
1637                 if (version == KSOCK_PROTO_V2)
1638                         return &ksocknal_protocol_v2x;
1639
1640                 if (version == KSOCK_PROTO_V3)
1641                         return &ksocknal_protocol_v3x;
1642
1643                 return NULL;
1644         }
1645
1646         if (hello->kshm_magic == le32_to_cpu(LNET_PROTO_TCP_MAGIC)) {
1647                 lnet_magicversion_t *hmv = (lnet_magicversion_t *)hello;
1648
1649                 CLASSERT (sizeof (lnet_magicversion_t) ==
1650                           offsetof (ksock_hello_msg_t, kshm_src_nid));
1651
1652                 if (hmv->version_major == cpu_to_le16 (KSOCK_PROTO_V1_MAJOR) &&
1653                     hmv->version_minor == cpu_to_le16 (KSOCK_PROTO_V1_MINOR))
1654                         return &ksocknal_protocol_v1x;
1655         }
1656
1657         return NULL;
1658 }
1659
1660 int
1661 ksocknal_send_hello (lnet_ni_t *ni, ksock_conn_t *conn,
1662                      lnet_nid_t peer_nid, ksock_hello_msg_t *hello)
1663 {
1664         /* CAVEAT EMPTOR: this byte flips 'ipaddrs' */
1665         ksock_net_t         *net = (ksock_net_t *)ni->ni_data;
1666
1667         LASSERT (hello->kshm_nips <= LNET_MAX_INTERFACES);
1668
1669         /* rely on caller to hold a ref on socket so it wouldn't disappear */
1670         LASSERT (conn->ksnc_proto != NULL);
1671
1672         hello->kshm_src_nid         = ni->ni_nid;
1673         hello->kshm_dst_nid         = peer_nid;
1674         hello->kshm_src_pid         = the_lnet.ln_pid;
1675
1676         hello->kshm_src_incarnation = net->ksnn_incarnation;
1677         hello->kshm_ctype           = conn->ksnc_type;
1678
1679         return conn->ksnc_proto->pro_send_hello(conn, hello);
1680 }
1681
1682 int
1683 ksocknal_invert_type(int type)
1684 {
1685         switch (type)
1686         {
1687         case SOCKLND_CONN_ANY:
1688         case SOCKLND_CONN_CONTROL:
1689                 return (type);
1690         case SOCKLND_CONN_BULK_IN:
1691                 return SOCKLND_CONN_BULK_OUT;
1692         case SOCKLND_CONN_BULK_OUT:
1693                 return SOCKLND_CONN_BULK_IN;
1694         default:
1695                 return (SOCKLND_CONN_NONE);
1696         }
1697 }
1698
1699 int
1700 ksocknal_recv_hello (lnet_ni_t *ni, ksock_conn_t *conn,
1701                      ksock_hello_msg_t *hello, lnet_process_id_t *peerid,
1702                      __u64 *incarnation)
1703 {
1704         /* Return < 0        fatal error
1705          *        0          success
1706          *        EALREADY   lost connection race
1707          *        EPROTO     protocol version mismatch
1708          */
1709         cfs_socket_t        *sock = conn->ksnc_sock;
1710         int                  active = (conn->ksnc_proto != NULL);
1711         int                  timeout;
1712         int                  proto_match;
1713         int                  rc;
1714         ksock_proto_t       *proto;
1715         lnet_process_id_t    recv_id;
1716
1717         /* socket type set on active connections - not set on passive */
1718         LASSERT (!active == !(conn->ksnc_type != SOCKLND_CONN_NONE));
1719
1720         timeout = active ? *ksocknal_tunables.ksnd_timeout :
1721                             lnet_acceptor_timeout();
1722
1723         rc = libcfs_sock_read(sock, &hello->kshm_magic, sizeof (hello->kshm_magic), timeout);
1724         if (rc != 0) {
1725                 CERROR ("Error %d reading HELLO from %u.%u.%u.%u\n",
1726                         rc, HIPQUAD(conn->ksnc_ipaddr));
1727                 LASSERT (rc < 0);
1728                 return rc;
1729         }
1730
1731         if (hello->kshm_magic != LNET_PROTO_MAGIC &&
1732             hello->kshm_magic != __swab32(LNET_PROTO_MAGIC) &&
1733             hello->kshm_magic != le32_to_cpu (LNET_PROTO_TCP_MAGIC)) {
1734                 /* Unexpected magic! */
1735                 CERROR ("Bad magic(1) %#08x (%#08x expected) from "
1736                         "%u.%u.%u.%u\n", __cpu_to_le32 (hello->kshm_magic),
1737                         LNET_PROTO_TCP_MAGIC,
1738                         HIPQUAD(conn->ksnc_ipaddr));
1739                 return -EPROTO;
1740         }
1741
1742         rc = libcfs_sock_read(sock, &hello->kshm_version,
1743                               sizeof(hello->kshm_version), timeout);
1744         if (rc != 0) {
1745                 CERROR ("Error %d reading HELLO from %u.%u.%u.%u\n",
1746                         rc, HIPQUAD(conn->ksnc_ipaddr));
1747                 LASSERT (rc < 0);
1748                 return rc;
1749         }
1750
1751         proto = ksocknal_parse_proto_version(hello);
1752         if (proto == NULL) {
1753                 if (!active) {
1754                         /* unknown protocol from peer, tell peer my protocol */
1755                         conn->ksnc_proto = &ksocknal_protocol_v3x;
1756 #if SOCKNAL_VERSION_DEBUG
1757                         if (*ksocknal_tunables.ksnd_protocol == 2)
1758                                 conn->ksnc_proto = &ksocknal_protocol_v2x;
1759                         else if (*ksocknal_tunables.ksnd_protocol == 1)
1760                                 conn->ksnc_proto = &ksocknal_protocol_v1x;
1761 #endif
1762                         hello->kshm_nips = 0;
1763                         ksocknal_send_hello(ni, conn, ni->ni_nid, hello);
1764                 }
1765
1766                 CERROR ("Unknown protocol version (%d.x expected)"
1767                         " from %u.%u.%u.%u\n",
1768                         conn->ksnc_proto->pro_version,
1769                         HIPQUAD(conn->ksnc_ipaddr));
1770
1771                 return -EPROTO;
1772         }
1773
1774         proto_match = (conn->ksnc_proto == proto);
1775         conn->ksnc_proto = proto;
1776
1777         /* receive the rest of hello message anyway */
1778         rc = conn->ksnc_proto->pro_recv_hello(conn, hello, timeout);
1779         if (rc != 0) {
1780                 CERROR("Error %d reading or checking hello from from %u.%u.%u.%u\n",
1781                        rc, HIPQUAD(conn->ksnc_ipaddr));
1782                 LASSERT (rc < 0);
1783                 return rc;
1784         }
1785
1786         *incarnation = hello->kshm_src_incarnation;
1787
1788         if (hello->kshm_src_nid == LNET_NID_ANY) {
1789                 CERROR("Expecting a HELLO hdr with a NID, but got LNET_NID_ANY"
1790                        "from %u.%u.%u.%u\n", HIPQUAD(conn->ksnc_ipaddr));
1791                 return -EPROTO;
1792         }
1793
1794         if (!active &&
1795             conn->ksnc_port > LNET_ACCEPTOR_MAX_RESERVED_PORT) {
1796                 /* Userspace NAL assigns peer process ID from socket */
1797                 recv_id.pid = conn->ksnc_port | LNET_PID_USERFLAG;
1798                 recv_id.nid = LNET_MKNID(LNET_NIDNET(ni->ni_nid), conn->ksnc_ipaddr);
1799         } else {
1800                 recv_id.nid = hello->kshm_src_nid;
1801                 recv_id.pid = hello->kshm_src_pid;
1802         }
1803
1804         if (!active) {
1805                 *peerid = recv_id;
1806
1807                 /* peer determines type */
1808                 conn->ksnc_type = ksocknal_invert_type(hello->kshm_ctype);
1809                 if (conn->ksnc_type == SOCKLND_CONN_NONE) {
1810                         CERROR ("Unexpected type %d from %s ip %u.%u.%u.%u\n",
1811                                 hello->kshm_ctype, libcfs_id2str(*peerid),
1812                                 HIPQUAD(conn->ksnc_ipaddr));
1813                         return -EPROTO;
1814                 }
1815
1816                 return 0;
1817         }
1818
1819         if (peerid->pid != recv_id.pid ||
1820             peerid->nid != recv_id.nid) {
1821                 LCONSOLE_ERROR_MSG(0x130, "Connected successfully to %s on host"
1822                                    " %u.%u.%u.%u, but they claimed they were "
1823                                    "%s; please check your Lustre "
1824                                    "configuration.\n",
1825                                    libcfs_id2str(*peerid),
1826                                    HIPQUAD(conn->ksnc_ipaddr),
1827                                    libcfs_id2str(recv_id));
1828                 return -EPROTO;
1829         }
1830
1831         if (hello->kshm_ctype == SOCKLND_CONN_NONE) {
1832                 /* Possible protocol mismatch or I lost the connection race */
1833                 return proto_match ? EALREADY : EPROTO;
1834         }
1835
1836         if (ksocknal_invert_type(hello->kshm_ctype) != conn->ksnc_type) {
1837                 CERROR ("Mismatched types: me %d, %s ip %u.%u.%u.%u %d\n",
1838                         conn->ksnc_type, libcfs_id2str(*peerid),
1839                         HIPQUAD(conn->ksnc_ipaddr),
1840                         hello->kshm_ctype);
1841                 return -EPROTO;
1842         }
1843
1844         return 0;
1845 }
1846
1847 int
1848 ksocknal_connect (ksock_route_t *route)
1849 {
1850         CFS_LIST_HEAD    (zombies);
1851         ksock_peer_t     *peer = route->ksnr_peer;
1852         int               type;
1853         int               wanted;
1854         cfs_socket_t     *sock;
1855         cfs_time_t        deadline;
1856         int               retry_later = 0;
1857         int               rc = 0;
1858
1859         deadline = cfs_time_add(cfs_time_current(),
1860                                 cfs_time_seconds(*ksocknal_tunables.ksnd_timeout));
1861
1862         write_lock_bh(&ksocknal_data.ksnd_global_lock);
1863
1864         LASSERT (route->ksnr_scheduled);
1865         LASSERT (!route->ksnr_connecting);
1866
1867         route->ksnr_connecting = 1;
1868
1869         for (;;) {
1870                 wanted = ksocknal_route_mask() & ~route->ksnr_connected;
1871
1872                 /* stop connecting if peer/route got closed under me, or
1873                  * route got connected while queued */
1874                 if (peer->ksnp_closing || route->ksnr_deleted ||
1875                     wanted == 0) {
1876                         retry_later = 0;
1877                         break;
1878                 }
1879
1880                 /* reschedule if peer is connecting to me */
1881                 if (peer->ksnp_accepting > 0) {
1882                         CDEBUG(D_NET,
1883                                "peer %s(%d) already connecting to me, retry later.\n",
1884                                libcfs_nid2str(peer->ksnp_id.nid), peer->ksnp_accepting);
1885                         retry_later = 1;
1886                 }
1887
1888                 if (retry_later) /* needs reschedule */
1889                         break;
1890
1891                 if ((wanted & (1 << SOCKLND_CONN_ANY)) != 0) {
1892                         type = SOCKLND_CONN_ANY;
1893                 } else if ((wanted & (1 << SOCKLND_CONN_CONTROL)) != 0) {
1894                         type = SOCKLND_CONN_CONTROL;
1895                 } else if ((wanted & (1 << SOCKLND_CONN_BULK_IN)) != 0) {
1896                         type = SOCKLND_CONN_BULK_IN;
1897                 } else {
1898                         LASSERT ((wanted & (1 << SOCKLND_CONN_BULK_OUT)) != 0);
1899                         type = SOCKLND_CONN_BULK_OUT;
1900                 }
1901
1902                 write_unlock_bh(&ksocknal_data.ksnd_global_lock);
1903
1904                 if (cfs_time_aftereq(cfs_time_current(), deadline)) {
1905                         rc = -ETIMEDOUT;
1906                         lnet_connect_console_error(rc, peer->ksnp_id.nid,
1907                                                    route->ksnr_ipaddr,
1908                                                    route->ksnr_port);
1909                         goto failed;
1910                 }
1911
1912                 rc = lnet_connect(&sock, peer->ksnp_id.nid,
1913                                   route->ksnr_myipaddr,
1914                                   route->ksnr_ipaddr, route->ksnr_port);
1915                 if (rc != 0)
1916                         goto failed;
1917
1918                 rc = ksocknal_create_conn(peer->ksnp_ni, route, sock, type);
1919                 if (rc < 0) {
1920                         lnet_connect_console_error(rc, peer->ksnp_id.nid,
1921                                                    route->ksnr_ipaddr,
1922                                                    route->ksnr_port);
1923                         goto failed;
1924                 }
1925
1926                 /* A +ve RC means I have to retry because I lost the connection
1927                  * race or I have to renegotiate protocol version */
1928                 retry_later = (rc != 0);
1929                 if (retry_later)
1930                         CDEBUG(D_NET, "peer %s: conn race, retry later.\n",
1931                                libcfs_nid2str(peer->ksnp_id.nid));
1932
1933                 write_lock_bh(&ksocknal_data.ksnd_global_lock);
1934         }
1935
1936         route->ksnr_scheduled = 0;
1937         route->ksnr_connecting = 0;
1938
1939         if (retry_later) {
1940                 /* re-queue for attention; this frees me up to handle
1941                  * the peer's incoming connection request */
1942
1943                 if (rc == EALREADY ||
1944                     (rc == 0 && peer->ksnp_accepting > 0)) {
1945                         /* We want to introduce a delay before next
1946                          * attempt to connect if we lost conn race,
1947                          * but the race is resolved quickly usually,
1948                          * so min_reconnectms should be good heuristic */
1949                         route->ksnr_retry_interval =
1950                                 cfs_time_seconds(*ksocknal_tunables.ksnd_min_reconnectms)/1000;
1951                         route->ksnr_timeout = cfs_time_add(cfs_time_current(),
1952                                                            route->ksnr_retry_interval);
1953                 }
1954
1955                 ksocknal_launch_connection_locked(route);
1956         }
1957
1958         write_unlock_bh(&ksocknal_data.ksnd_global_lock);
1959         return retry_later;
1960
1961  failed:
1962         write_lock_bh(&ksocknal_data.ksnd_global_lock);
1963
1964         route->ksnr_scheduled = 0;
1965         route->ksnr_connecting = 0;
1966
1967         /* This is a retry rather than a new connection */
1968         route->ksnr_retry_interval *= 2;
1969         route->ksnr_retry_interval =
1970                 MAX(route->ksnr_retry_interval,
1971                     cfs_time_seconds(*ksocknal_tunables.ksnd_min_reconnectms)/1000);
1972         route->ksnr_retry_interval =
1973                 MIN(route->ksnr_retry_interval,
1974                     cfs_time_seconds(*ksocknal_tunables.ksnd_max_reconnectms)/1000);
1975
1976         LASSERT (route->ksnr_retry_interval != 0);
1977         route->ksnr_timeout = cfs_time_add(cfs_time_current(),
1978                                            route->ksnr_retry_interval);
1979
1980         if (!cfs_list_empty(&peer->ksnp_tx_queue) &&
1981             peer->ksnp_accepting == 0 &&
1982             ksocknal_find_connecting_route_locked(peer) == NULL) {
1983                 ksock_conn_t *conn;
1984
1985                 /* ksnp_tx_queue is queued on a conn on successful
1986                  * connection for V1.x and V2.x */
1987                 if (!cfs_list_empty (&peer->ksnp_conns)) {
1988                         conn = cfs_list_entry(peer->ksnp_conns.next,
1989                                               ksock_conn_t, ksnc_list);
1990                         LASSERT (conn->ksnc_proto == &ksocknal_protocol_v3x);
1991                 }
1992
1993                 /* take all the blocked packets while I've got the lock and
1994                  * complete below... */
1995                 cfs_list_splice_init(&peer->ksnp_tx_queue, &zombies);
1996         }
1997
1998 #if 0           /* irrelevent with only eager routes */
1999         if (!route->ksnr_deleted) {
2000                 /* make this route least-favourite for re-selection */
2001                 cfs_list_del(&route->ksnr_list);
2002                 cfs_list_add_tail(&route->ksnr_list, &peer->ksnp_routes);
2003         }
2004 #endif
2005         write_unlock_bh(&ksocknal_data.ksnd_global_lock);
2006
2007         ksocknal_peer_failed(peer);
2008         ksocknal_txlist_done(peer->ksnp_ni, &zombies, 1);
2009         return 0;
2010 }
2011
2012 /*
2013  * check whether we need to create more connds.
2014  * It will try to create new thread if it's necessary, @timeout can
2015  * be updated if failed to create, so caller wouldn't keep try while
2016  * running out of resource.
2017  */
2018 static int
2019 ksocknal_connd_check_start(long sec, long *timeout)
2020 {
2021         char name[16];
2022         int rc;
2023         int total = ksocknal_data.ksnd_connd_starting +
2024                     ksocknal_data.ksnd_connd_running;
2025
2026         if (unlikely(ksocknal_data.ksnd_init < SOCKNAL_INIT_ALL)) {
2027                 /* still in initializing */
2028                 return 0;
2029         }
2030
2031         if (total >= *ksocknal_tunables.ksnd_nconnds_max ||
2032             total > ksocknal_data.ksnd_connd_connecting + SOCKNAL_CONND_RESV) {
2033                 /* can't create more connd, or still have enough
2034                  * threads to handle more connecting */
2035                 return 0;
2036         }
2037
2038         if (list_empty(&ksocknal_data.ksnd_connd_routes)) {
2039                 /* no pending connecting request */
2040                 return 0;
2041         }
2042
2043         if (sec - ksocknal_data.ksnd_connd_failed_stamp <= 1) {
2044                 /* may run out of resource, retry later */
2045                 *timeout = cfs_time_seconds(1);
2046                 return 0;
2047         }
2048
2049         if (ksocknal_data.ksnd_connd_starting > 0) {
2050                 /* serialize starting to avoid flood */
2051                 return 0;
2052         }
2053
2054         ksocknal_data.ksnd_connd_starting_stamp = sec;
2055         ksocknal_data.ksnd_connd_starting++;
2056         spin_unlock_bh(&ksocknal_data.ksnd_connd_lock);
2057
2058         /* NB: total is the next id */
2059         snprintf(name, sizeof(name), "socknal_cd%02d", total);
2060         rc = ksocknal_thread_start(ksocknal_connd, NULL, name);
2061
2062         spin_lock_bh(&ksocknal_data.ksnd_connd_lock);
2063         if (rc == 0)
2064                 return 1;
2065
2066         /* we tried ... */
2067         LASSERT(ksocknal_data.ksnd_connd_starting > 0);
2068         ksocknal_data.ksnd_connd_starting--;
2069         ksocknal_data.ksnd_connd_failed_stamp = cfs_time_current_sec();
2070
2071         return 1;
2072 }
2073
2074 /*
2075  * check whether current thread can exit, it will return 1 if there are too
2076  * many threads and no creating in past 120 seconds.
2077  * Also, this function may update @timeout to make caller come back
2078  * again to recheck these conditions.
2079  */
2080 static int
2081 ksocknal_connd_check_stop(long sec, long *timeout)
2082 {
2083         int val;
2084
2085         if (unlikely(ksocknal_data.ksnd_init < SOCKNAL_INIT_ALL)) {
2086                 /* still in initializing */
2087                 return 0;
2088         }
2089
2090         if (ksocknal_data.ksnd_connd_starting > 0) {
2091                 /* in progress of starting new thread */
2092                 return 0;
2093         }
2094
2095         if (ksocknal_data.ksnd_connd_running <=
2096             *ksocknal_tunables.ksnd_nconnds) { /* can't shrink */
2097                 return 0;
2098         }
2099
2100         /* created thread in past 120 seconds? */
2101         val = (int)(ksocknal_data.ksnd_connd_starting_stamp +
2102                     SOCKNAL_CONND_TIMEOUT - sec);
2103
2104         *timeout = (val > 0) ? cfs_time_seconds(val) :
2105                                cfs_time_seconds(SOCKNAL_CONND_TIMEOUT);
2106         if (val > 0)
2107                 return 0;
2108
2109         /* no creating in past 120 seconds */
2110
2111         return ksocknal_data.ksnd_connd_running >
2112                ksocknal_data.ksnd_connd_connecting + SOCKNAL_CONND_RESV;
2113 }
2114
2115 /* Go through connd_routes queue looking for a route that we can process
2116  * right now, @timeout_p can be updated if we need to come back later */
2117 static ksock_route_t *
2118 ksocknal_connd_get_route_locked(signed long *timeout_p)
2119 {
2120         ksock_route_t *route;
2121         cfs_time_t     now;
2122
2123         now = cfs_time_current();
2124
2125         /* connd_routes can contain both pending and ordinary routes */
2126         cfs_list_for_each_entry (route, &ksocknal_data.ksnd_connd_routes,
2127                                  ksnr_connd_list) {
2128
2129                 if (route->ksnr_retry_interval == 0 ||
2130                     cfs_time_aftereq(now, route->ksnr_timeout))
2131                         return route;
2132
2133                 if (*timeout_p == MAX_SCHEDULE_TIMEOUT ||
2134                     (int)*timeout_p > (int)(route->ksnr_timeout - now))
2135                         *timeout_p = (int)(route->ksnr_timeout - now);
2136         }
2137
2138         return NULL;
2139 }
2140
2141 int
2142 ksocknal_connd (void *arg)
2143 {
2144         spinlock_t    *connd_lock = &ksocknal_data.ksnd_connd_lock;
2145         ksock_connreq_t   *cr;
2146         wait_queue_t     wait;
2147         int                nloops = 0;
2148         int                cons_retry = 0;
2149
2150         cfs_block_allsigs ();
2151
2152         init_waitqueue_entry_current(&wait);
2153
2154         spin_lock_bh(connd_lock);
2155
2156         LASSERT(ksocknal_data.ksnd_connd_starting > 0);
2157         ksocknal_data.ksnd_connd_starting--;
2158         ksocknal_data.ksnd_connd_running++;
2159
2160         while (!ksocknal_data.ksnd_shuttingdown) {
2161                 ksock_route_t *route = NULL;
2162                 long sec = cfs_time_current_sec();
2163                 long timeout = MAX_SCHEDULE_TIMEOUT;
2164                 int  dropped_lock = 0;
2165
2166                 if (ksocknal_connd_check_stop(sec, &timeout)) {
2167                         /* wakeup another one to check stop */
2168                         wake_up(&ksocknal_data.ksnd_connd_waitq);
2169                         break;
2170                 }
2171
2172                 if (ksocknal_connd_check_start(sec, &timeout)) {
2173                         /* created new thread */
2174                         dropped_lock = 1;
2175                 }
2176
2177                 if (!cfs_list_empty(&ksocknal_data.ksnd_connd_connreqs)) {
2178                         /* Connection accepted by the listener */
2179                         cr = cfs_list_entry(ksocknal_data.ksnd_connd_connreqs. \
2180                                             next, ksock_connreq_t, ksncr_list);
2181
2182                         cfs_list_del(&cr->ksncr_list);
2183                         spin_unlock_bh(connd_lock);
2184                         dropped_lock = 1;
2185
2186                         ksocknal_create_conn(cr->ksncr_ni, NULL,
2187                                              cr->ksncr_sock, SOCKLND_CONN_NONE);
2188                         lnet_ni_decref(cr->ksncr_ni);
2189                         LIBCFS_FREE(cr, sizeof(*cr));
2190
2191                         spin_lock_bh(connd_lock);
2192                 }
2193
2194                 /* Only handle an outgoing connection request if there
2195                  * is a thread left to handle incoming connections and
2196                  * create new connd */
2197                 if (ksocknal_data.ksnd_connd_connecting + SOCKNAL_CONND_RESV <
2198                     ksocknal_data.ksnd_connd_running) {
2199                         route = ksocknal_connd_get_route_locked(&timeout);
2200                 }
2201                 if (route != NULL) {
2202                         cfs_list_del (&route->ksnr_connd_list);
2203                         ksocknal_data.ksnd_connd_connecting++;
2204                         spin_unlock_bh(connd_lock);
2205                         dropped_lock = 1;
2206
2207                         if (ksocknal_connect(route)) {
2208                                 /* consecutive retry */
2209                                 if (cons_retry++ > SOCKNAL_INSANITY_RECONN) {
2210                                         CWARN("massive consecutive "
2211                                               "re-connecting to %u.%u.%u.%u\n",
2212                                               HIPQUAD(route->ksnr_ipaddr));
2213                                         cons_retry = 0;
2214                                 }
2215                         } else {
2216                                 cons_retry = 0;
2217                         }
2218
2219                         ksocknal_route_decref(route);
2220
2221                         spin_lock_bh(connd_lock);
2222                         ksocknal_data.ksnd_connd_connecting--;
2223                 }
2224
2225                 if (dropped_lock) {
2226                         if (++nloops < SOCKNAL_RESCHED)
2227                                 continue;
2228                         spin_unlock_bh(connd_lock);
2229                         nloops = 0;
2230                         cond_resched();
2231                         spin_lock_bh(connd_lock);
2232                         continue;
2233                 }
2234
2235                 /* Nothing to do for 'timeout'  */
2236                 set_current_state(TASK_INTERRUPTIBLE);
2237                 add_wait_queue_exclusive(&ksocknal_data.ksnd_connd_waitq, &wait);
2238                 spin_unlock_bh(connd_lock);
2239
2240                 nloops = 0;
2241                 waitq_timedwait(&wait, TASK_INTERRUPTIBLE, timeout);
2242
2243                 set_current_state(TASK_RUNNING);
2244                 remove_wait_queue(&ksocknal_data.ksnd_connd_waitq, &wait);
2245                 spin_lock_bh(connd_lock);
2246         }
2247         ksocknal_data.ksnd_connd_running--;
2248         spin_unlock_bh(connd_lock);
2249
2250         ksocknal_thread_fini();
2251         return 0;
2252 }
2253
2254 ksock_conn_t *
2255 ksocknal_find_timed_out_conn (ksock_peer_t *peer)
2256 {
2257         /* We're called with a shared lock on ksnd_global_lock */
2258         ksock_conn_t      *conn;
2259         cfs_list_t        *ctmp;
2260
2261         cfs_list_for_each (ctmp, &peer->ksnp_conns) {
2262                 int     error;
2263                 conn = cfs_list_entry (ctmp, ksock_conn_t, ksnc_list);
2264
2265                 /* Don't need the {get,put}connsock dance to deref ksnc_sock */
2266                 LASSERT (!conn->ksnc_closing);
2267
2268                 /* SOCK_ERROR will reset error code of socket in
2269                  * some platform (like Darwin8.x) */
2270                 error = libcfs_sock_error(conn->ksnc_sock);
2271                 if (error != 0) {
2272                         ksocknal_conn_addref(conn);
2273
2274                         switch (error) {
2275                         case ECONNRESET:
2276                                 CNETERR("A connection with %s "
2277                                         "(%u.%u.%u.%u:%d) was reset; "
2278                                         "it may have rebooted.\n",
2279                                         libcfs_id2str(peer->ksnp_id),
2280                                         HIPQUAD(conn->ksnc_ipaddr),
2281                                         conn->ksnc_port);
2282                                 break;
2283                         case ETIMEDOUT:
2284                                 CNETERR("A connection with %s "
2285                                         "(%u.%u.%u.%u:%d) timed out; the "
2286                                         "network or node may be down.\n",
2287                                         libcfs_id2str(peer->ksnp_id),
2288                                         HIPQUAD(conn->ksnc_ipaddr),
2289                                         conn->ksnc_port);
2290                                 break;
2291                         default:
2292                                 CNETERR("An unexpected network error %d "
2293                                         "occurred with %s "
2294                                         "(%u.%u.%u.%u:%d\n", error,
2295                                         libcfs_id2str(peer->ksnp_id),
2296                                         HIPQUAD(conn->ksnc_ipaddr),
2297                                         conn->ksnc_port);
2298                                 break;
2299                         }
2300
2301                         return (conn);
2302                 }
2303
2304                 if (conn->ksnc_rx_started &&
2305                     cfs_time_aftereq(cfs_time_current(),
2306                                      conn->ksnc_rx_deadline)) {
2307                         /* Timed out incomplete incoming message */
2308                         ksocknal_conn_addref(conn);
2309                         CNETERR("Timeout receiving from %s (%u.%u.%u.%u:%d), "
2310                                 "state %d wanted %d left %d\n",
2311                                 libcfs_id2str(peer->ksnp_id),
2312                                 HIPQUAD(conn->ksnc_ipaddr),
2313                                 conn->ksnc_port,
2314                                 conn->ksnc_rx_state,
2315                                 conn->ksnc_rx_nob_wanted,
2316                                 conn->ksnc_rx_nob_left);
2317                         return (conn);
2318                 }
2319
2320                 if ((!cfs_list_empty(&conn->ksnc_tx_queue) ||
2321                      libcfs_sock_wmem_queued(conn->ksnc_sock) != 0) &&
2322                     cfs_time_aftereq(cfs_time_current(),
2323                                      conn->ksnc_tx_deadline)) {
2324                         /* Timed out messages queued for sending or
2325                          * buffered in the socket's send buffer */
2326                         ksocknal_conn_addref(conn);
2327                         CNETERR("Timeout sending data to %s (%u.%u.%u.%u:%d) "
2328                                 "the network or that node may be down.\n",
2329                                 libcfs_id2str(peer->ksnp_id),
2330                                 HIPQUAD(conn->ksnc_ipaddr),
2331                                 conn->ksnc_port);
2332                         return (conn);
2333                 }
2334         }
2335
2336         return (NULL);
2337 }
2338
2339 static inline void
2340 ksocknal_flush_stale_txs(ksock_peer_t *peer)
2341 {
2342         ksock_tx_t        *tx;
2343         CFS_LIST_HEAD      (stale_txs);
2344
2345         write_lock_bh(&ksocknal_data.ksnd_global_lock);
2346
2347         while (!cfs_list_empty (&peer->ksnp_tx_queue)) {
2348                 tx = cfs_list_entry (peer->ksnp_tx_queue.next,
2349                                      ksock_tx_t, tx_list);
2350
2351                 if (!cfs_time_aftereq(cfs_time_current(),
2352                                       tx->tx_deadline))
2353                         break;
2354
2355                 cfs_list_del (&tx->tx_list);
2356                 cfs_list_add_tail (&tx->tx_list, &stale_txs);
2357         }
2358
2359         write_unlock_bh(&ksocknal_data.ksnd_global_lock);
2360
2361         ksocknal_txlist_done(peer->ksnp_ni, &stale_txs, 1);
2362 }
2363
2364 int
2365 ksocknal_send_keepalive_locked(ksock_peer_t *peer)
2366 {
2367         ksock_sched_t  *sched;
2368         ksock_conn_t   *conn;
2369         ksock_tx_t     *tx;
2370
2371         if (cfs_list_empty(&peer->ksnp_conns)) /* last_alive will be updated by create_conn */
2372                 return 0;
2373
2374         if (peer->ksnp_proto != &ksocknal_protocol_v3x)
2375                 return 0;
2376
2377         if (*ksocknal_tunables.ksnd_keepalive <= 0 ||
2378             cfs_time_before(cfs_time_current(),
2379                             cfs_time_add(peer->ksnp_last_alive,
2380                                          cfs_time_seconds(*ksocknal_tunables.ksnd_keepalive))))
2381                 return 0;
2382
2383         if (cfs_time_before(cfs_time_current(),
2384                             peer->ksnp_send_keepalive))
2385                 return 0;
2386
2387         /* retry 10 secs later, so we wouldn't put pressure
2388          * on this peer if we failed to send keepalive this time */
2389         peer->ksnp_send_keepalive = cfs_time_shift(10);
2390
2391         conn = ksocknal_find_conn_locked(peer, NULL, 1);
2392         if (conn != NULL) {
2393                 sched = conn->ksnc_scheduler;
2394
2395                 spin_lock_bh(&sched->kss_lock);
2396                 if (!cfs_list_empty(&conn->ksnc_tx_queue)) {
2397                         spin_unlock_bh(&sched->kss_lock);
2398                         /* there is an queued ACK, don't need keepalive */
2399                         return 0;
2400                 }
2401
2402                 spin_unlock_bh(&sched->kss_lock);
2403         }
2404
2405         read_unlock(&ksocknal_data.ksnd_global_lock);
2406
2407         /* cookie = 1 is reserved for keepalive PING */
2408         tx = ksocknal_alloc_tx_noop(1, 1);
2409         if (tx == NULL) {
2410                 read_lock(&ksocknal_data.ksnd_global_lock);
2411                 return -ENOMEM;
2412         }
2413
2414         if (ksocknal_launch_packet(peer->ksnp_ni, tx, peer->ksnp_id) == 0) {
2415                 read_lock(&ksocknal_data.ksnd_global_lock);
2416                 return 1;
2417         }
2418
2419         ksocknal_free_tx(tx);
2420         read_lock(&ksocknal_data.ksnd_global_lock);
2421
2422         return -EIO;
2423 }
2424
2425
2426 void
2427 ksocknal_check_peer_timeouts (int idx)
2428 {
2429         cfs_list_t       *peers = &ksocknal_data.ksnd_peers[idx];
2430         ksock_peer_t     *peer;
2431         ksock_conn_t     *conn;
2432         ksock_tx_t       *tx;
2433
2434  again:
2435         /* NB. We expect to have a look at all the peers and not find any
2436          * connections to time out, so we just use a shared lock while we
2437          * take a look... */
2438         read_lock(&ksocknal_data.ksnd_global_lock);
2439
2440         cfs_list_for_each_entry_typed(peer, peers, ksock_peer_t, ksnp_list) {
2441                 cfs_time_t  deadline = 0;
2442                 int         resid = 0;
2443                 int         n     = 0;
2444
2445                 if (ksocknal_send_keepalive_locked(peer) != 0) {
2446                         read_unlock(&ksocknal_data.ksnd_global_lock);
2447                         goto again;
2448                 }
2449
2450                 conn = ksocknal_find_timed_out_conn (peer);
2451
2452                 if (conn != NULL) {
2453                         read_unlock(&ksocknal_data.ksnd_global_lock);
2454
2455                         ksocknal_close_conn_and_siblings (conn, -ETIMEDOUT);
2456
2457                         /* NB we won't find this one again, but we can't
2458                          * just proceed with the next peer, since we dropped
2459                          * ksnd_global_lock and it might be dead already! */
2460                         ksocknal_conn_decref(conn);
2461                         goto again;
2462                 }
2463
2464                 /* we can't process stale txs right here because we're
2465                  * holding only shared lock */
2466                 if (!cfs_list_empty (&peer->ksnp_tx_queue)) {
2467                         ksock_tx_t *tx =
2468                                 cfs_list_entry (peer->ksnp_tx_queue.next,
2469                                                 ksock_tx_t, tx_list);
2470
2471                         if (cfs_time_aftereq(cfs_time_current(),
2472                                              tx->tx_deadline)) {
2473
2474                                 ksocknal_peer_addref(peer);
2475                                 read_unlock(&ksocknal_data.ksnd_global_lock);
2476
2477                                 ksocknal_flush_stale_txs(peer);
2478
2479                                 ksocknal_peer_decref(peer);
2480                                 goto again;
2481                         }
2482                 }
2483
2484                 if (cfs_list_empty(&peer->ksnp_zc_req_list))
2485                         continue;
2486
2487                 spin_lock(&peer->ksnp_lock);
2488                 cfs_list_for_each_entry_typed(tx, &peer->ksnp_zc_req_list,
2489                                               ksock_tx_t, tx_zc_list) {
2490                         if (!cfs_time_aftereq(cfs_time_current(),
2491                                               tx->tx_deadline))
2492                                 break;
2493                         /* ignore the TX if connection is being closed */
2494                         if (tx->tx_conn->ksnc_closing)
2495                                 continue;
2496                         n++;
2497                 }
2498
2499                 if (n == 0) {
2500                         spin_unlock(&peer->ksnp_lock);
2501                         continue;
2502                 }
2503
2504                 tx = cfs_list_entry(peer->ksnp_zc_req_list.next,
2505                                     ksock_tx_t, tx_zc_list);
2506                 deadline = tx->tx_deadline;
2507                 resid    = tx->tx_resid;
2508                 conn     = tx->tx_conn;
2509                 ksocknal_conn_addref(conn);
2510
2511                 spin_unlock(&peer->ksnp_lock);
2512                 read_unlock(&ksocknal_data.ksnd_global_lock);
2513
2514                 CERROR("Total %d stale ZC_REQs for peer %s detected; the "
2515                        "oldest(%p) timed out %ld secs ago, "
2516                        "resid: %d, wmem: %d\n",
2517                        n, libcfs_nid2str(peer->ksnp_id.nid), tx,
2518                        cfs_duration_sec(cfs_time_current() - deadline),
2519                        resid, libcfs_sock_wmem_queued(conn->ksnc_sock));
2520
2521                 ksocknal_close_conn_and_siblings (conn, -ETIMEDOUT);
2522                 ksocknal_conn_decref(conn);
2523                 goto again;
2524         }
2525
2526         read_unlock(&ksocknal_data.ksnd_global_lock);
2527 }
2528
2529 int ksocknal_reaper(void *arg)
2530 {
2531         wait_queue_t     wait;
2532         ksock_conn_t      *conn;
2533         ksock_sched_t     *sched;
2534         cfs_list_t         enomem_conns;
2535         int                nenomem_conns;
2536         cfs_duration_t     timeout;
2537         int                i;
2538         int                peer_index = 0;
2539         cfs_time_t         deadline = cfs_time_current();
2540
2541         cfs_block_allsigs ();
2542
2543         CFS_INIT_LIST_HEAD(&enomem_conns);
2544         init_waitqueue_entry_current(&wait);
2545
2546         spin_lock_bh(&ksocknal_data.ksnd_reaper_lock);
2547
2548         while (!ksocknal_data.ksnd_shuttingdown) {
2549
2550                 if (!cfs_list_empty (&ksocknal_data.ksnd_deathrow_conns)) {
2551                         conn = cfs_list_entry (ksocknal_data. \
2552                                                ksnd_deathrow_conns.next,
2553                                                ksock_conn_t, ksnc_list);
2554                         cfs_list_del (&conn->ksnc_list);
2555
2556                         spin_unlock_bh(&ksocknal_data.ksnd_reaper_lock);
2557
2558                         ksocknal_terminate_conn(conn);
2559                         ksocknal_conn_decref(conn);
2560
2561                         spin_lock_bh(&ksocknal_data.ksnd_reaper_lock);
2562                         continue;
2563                 }
2564
2565                 if (!cfs_list_empty (&ksocknal_data.ksnd_zombie_conns)) {
2566                         conn = cfs_list_entry (ksocknal_data.ksnd_zombie_conns.\
2567                                                next, ksock_conn_t, ksnc_list);
2568                         cfs_list_del (&conn->ksnc_list);
2569
2570                         spin_unlock_bh(&ksocknal_data.ksnd_reaper_lock);
2571
2572                         ksocknal_destroy_conn(conn);
2573
2574                         spin_lock_bh(&ksocknal_data.ksnd_reaper_lock);
2575                         continue;
2576                 }
2577
2578                 if (!cfs_list_empty (&ksocknal_data.ksnd_enomem_conns)) {
2579                         cfs_list_add(&enomem_conns,
2580                                      &ksocknal_data.ksnd_enomem_conns);
2581                         cfs_list_del_init(&ksocknal_data.ksnd_enomem_conns);
2582                 }
2583
2584                 spin_unlock_bh(&ksocknal_data.ksnd_reaper_lock);
2585
2586                 /* reschedule all the connections that stalled with ENOMEM... */
2587                 nenomem_conns = 0;
2588                 while (!cfs_list_empty (&enomem_conns)) {
2589                         conn = cfs_list_entry (enomem_conns.next,
2590                                                ksock_conn_t, ksnc_tx_list);
2591                         cfs_list_del (&conn->ksnc_tx_list);
2592
2593                         sched = conn->ksnc_scheduler;
2594
2595                         spin_lock_bh(&sched->kss_lock);
2596
2597                         LASSERT(conn->ksnc_tx_scheduled);
2598                         conn->ksnc_tx_ready = 1;
2599                         cfs_list_add_tail(&conn->ksnc_tx_list,
2600                                           &sched->kss_tx_conns);
2601                         wake_up(&sched->kss_waitq);
2602
2603                         spin_unlock_bh(&sched->kss_lock);
2604                         nenomem_conns++;
2605                 }
2606
2607                 /* careful with the jiffy wrap... */
2608                 while ((timeout = cfs_time_sub(deadline,
2609                                                cfs_time_current())) <= 0) {
2610                         const int n = 4;
2611                         const int p = 1;
2612                         int       chunk = ksocknal_data.ksnd_peer_hash_size;
2613
2614                         /* Time to check for timeouts on a few more peers: I do
2615                          * checks every 'p' seconds on a proportion of the peer
2616                          * table and I need to check every connection 'n' times
2617                          * within a timeout interval, to ensure I detect a
2618                          * timeout on any connection within (n+1)/n times the
2619                          * timeout interval. */
2620
2621                         if (*ksocknal_tunables.ksnd_timeout > n * p)
2622                                 chunk = (chunk * n * p) /
2623                                         *ksocknal_tunables.ksnd_timeout;
2624                         if (chunk == 0)
2625                                 chunk = 1;
2626
2627                         for (i = 0; i < chunk; i++) {
2628                                 ksocknal_check_peer_timeouts (peer_index);
2629                                 peer_index = (peer_index + 1) %
2630                                              ksocknal_data.ksnd_peer_hash_size;
2631                         }
2632
2633                         deadline = cfs_time_add(deadline, cfs_time_seconds(p));
2634                 }
2635
2636                 if (nenomem_conns != 0) {
2637                         /* Reduce my timeout if I rescheduled ENOMEM conns.
2638                          * This also prevents me getting woken immediately
2639                          * if any go back on my enomem list. */
2640                         timeout = SOCKNAL_ENOMEM_RETRY;
2641                 }
2642                 ksocknal_data.ksnd_reaper_waketime =
2643                         cfs_time_add(cfs_time_current(), timeout);
2644
2645                         set_current_state(TASK_INTERRUPTIBLE);
2646                 add_wait_queue(&ksocknal_data.ksnd_reaper_waitq, &wait);
2647
2648                 if (!ksocknal_data.ksnd_shuttingdown &&
2649                     cfs_list_empty(&ksocknal_data.ksnd_deathrow_conns) &&
2650                     cfs_list_empty(&ksocknal_data.ksnd_zombie_conns))
2651                         waitq_timedwait(&wait, TASK_INTERRUPTIBLE, timeout);
2652
2653                 set_current_state(TASK_RUNNING);
2654                 remove_wait_queue(&ksocknal_data.ksnd_reaper_waitq, &wait);
2655
2656                 spin_lock_bh(&ksocknal_data.ksnd_reaper_lock);
2657         }
2658
2659         spin_unlock_bh(&ksocknal_data.ksnd_reaper_lock);
2660
2661         ksocknal_thread_fini();
2662         return 0;
2663 }