Pileus Git - ~andy/linux/blob - drivers/staging/lustre/lnet/selftest/rpc.c

   1 /*
   2  * GPL HEADER START
   3  *
   4  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
   5  *
   6  * This program is free software; you can redistribute it and/or modify
   7  * it under the terms of the GNU General Public License version 2 only,
   8  * as published by the Free Software Foundation.
   9  *
  10  * This program is distributed in the hope that it will be useful, but
  11  * WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13  * General Public License version 2 for more details (a copy is included
  14  * in the LICENSE file that accompanied this code).
  15  *
  16  * You should have received a copy of the GNU General Public License
  17  * version 2 along with this program; If not, see
  18  * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf
  19  *
  20  * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
  21  * CA 95054 USA or visit www.sun.com if you need additional information or
  22  * have any questions.
  23  *
  24  * GPL HEADER END
  25  */
  26 /*
  27  * Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved.
  28  * Use is subject to license terms.
  29  *
  30  * Copyright (c) 2012, Intel Corporation.
  31  */
  32 /*
  33  * This file is part of Lustre, http://www.lustre.org/
  34  * Lustre is a trademark of Sun Microsystems, Inc.
  35  *
  36  * lnet/selftest/rpc.c
  37  *
  38  * Author: Isaac Huang <isaac@clusterfs.com>
  39  *
  40  * 2012-05-13: Liang Zhen <liang@whamcloud.com>
  41  * - percpt data for service to improve smp performance
  42  * - code cleanup
  43  */
  44
  45 #define DEBUG_SUBSYSTEM S_LNET
  46
  47 #include "selftest.h"
  48
  49 typedef enum {
  50         SRPC_STATE_NONE,
  51         SRPC_STATE_NI_INIT,
  52         SRPC_STATE_EQ_INIT,
  53         SRPC_STATE_RUNNING,
  54         SRPC_STATE_STOPPING,
  55 } srpc_state_t;
  56
  57 struct smoketest_rpc {
  58         spinlock_t       rpc_glock;     /* global lock */
  59         srpc_service_t  *rpc_services[SRPC_SERVICE_MAX_ID + 1];
  60         lnet_handle_eq_t rpc_lnet_eq;   /* _the_ LNet event queue */
  61         srpc_state_t     rpc_state;
  62         srpc_counters_t  rpc_counters;
  63         __u64            rpc_matchbits; /* matchbits counter */
  64 } srpc_data;
  65
  66 static inline int
  67 srpc_serv_portal(int svc_id)
  68 {
  69         return svc_id < SRPC_FRAMEWORK_SERVICE_MAX_ID ?
  70                SRPC_FRAMEWORK_REQUEST_PORTAL : SRPC_REQUEST_PORTAL;
  71 }
  72
  73 /* forward ref's */
  74 int srpc_handle_rpc (swi_workitem_t *wi);
  75
  76 void srpc_get_counters (srpc_counters_t *cnt)
  77 {
  78         spin_lock(&srpc_data.rpc_glock);
  79         *cnt = srpc_data.rpc_counters;
  80         spin_unlock(&srpc_data.rpc_glock);
  81 }
  82
  83 void srpc_set_counters (const srpc_counters_t *cnt)
  84 {
  85         spin_lock(&srpc_data.rpc_glock);
  86         srpc_data.rpc_counters = *cnt;
  87         spin_unlock(&srpc_data.rpc_glock);
  88 }
  89
  90 int
  91 srpc_add_bulk_page(srpc_bulk_t *bk, struct page *pg, int i, int nob)
  92 {
  93         nob = min(nob, (int)PAGE_CACHE_SIZE);
  94
  95         LASSERT(nob > 0);
  96         LASSERT(i >= 0 && i < bk->bk_niov);
  97
  98         bk->bk_iovs[i].kiov_offset = 0;
  99         bk->bk_iovs[i].kiov_page   = pg;
 100         bk->bk_iovs[i].kiov_len    = nob;
 101         return nob;
 102 }
 103
 104 void
 105 srpc_free_bulk (srpc_bulk_t *bk)
 106 {
 107         int      i;
 108         struct page *pg;
 109
 110         LASSERT (bk != NULL);
 111
 112         for (i = 0; i < bk->bk_niov; i++) {
 113                 pg = bk->bk_iovs[i].kiov_page;
 114                 if (pg == NULL) break;
 115
 116                 __free_page(pg);
 117         }
 118
 119         LIBCFS_FREE(bk, offsetof(srpc_bulk_t, bk_iovs[bk->bk_niov]));
 120         return;
 121 }
 122
 123 srpc_bulk_t *
 124 srpc_alloc_bulk(int cpt, unsigned bulk_npg, unsigned bulk_len, int sink)
 125 {
 126         srpc_bulk_t  *bk;
 127         int           i;
 128
 129         LASSERT(bulk_npg > 0 && bulk_npg <= LNET_MAX_IOV);
 130
 131         LIBCFS_CPT_ALLOC(bk, lnet_cpt_table(), cpt,
 132                          offsetof(srpc_bulk_t, bk_iovs[bulk_npg]));
 133         if (bk == NULL) {
 134                 CERROR("Can't allocate descriptor for %d pages\n", bulk_npg);
 135                 return NULL;
 136         }
 137
 138         memset(bk, 0, offsetof(srpc_bulk_t, bk_iovs[bulk_npg]));
 139         bk->bk_sink   = sink;
 140         bk->bk_len    = bulk_len;
 141         bk->bk_niov   = bulk_npg;
 142
 143         for (i = 0; i < bulk_npg; i++) {
 144                 struct page *pg;
 145                 int         nob;
 146
 147                 pg = alloc_pages_node(cfs_cpt_spread_node(lnet_cpt_table(), cpt),
 148                                       GFP_IOFS, 0);
 149                 if (pg == NULL) {
 150                         CERROR("Can't allocate page %d of %d\n", i, bulk_npg);
 151                         srpc_free_bulk(bk);
 152                         return NULL;
 153                 }
 154
 155                 nob = srpc_add_bulk_page(bk, pg, i, bulk_len);
 156                 bulk_len -= nob;
 157         }
 158
 159         return bk;
 160 }
 161
 162 static inline __u64
 163 srpc_next_id (void)
 164 {
 165         __u64 id;
 166
 167         spin_lock(&srpc_data.rpc_glock);
 168         id = srpc_data.rpc_matchbits++;
 169         spin_unlock(&srpc_data.rpc_glock);
 170         return id;
 171 }
 172
 173 void
 174 srpc_init_server_rpc(struct srpc_server_rpc *rpc,
 175                      struct srpc_service_cd *scd,
 176                      struct srpc_buffer *buffer)
 177 {
 178         memset(rpc, 0, sizeof(*rpc));
 179         swi_init_workitem(&rpc->srpc_wi, rpc, srpc_handle_rpc,
 180                           srpc_serv_is_framework(scd->scd_svc) ?
 181                           lst_sched_serial : lst_sched_test[scd->scd_cpt]);
 182
 183         rpc->srpc_ev.ev_fired = 1; /* no event expected now */
 184
 185         rpc->srpc_scd      = scd;
 186         rpc->srpc_reqstbuf = buffer;
 187         rpc->srpc_peer     = buffer->buf_peer;
 188         rpc->srpc_self     = buffer->buf_self;
 189         LNetInvalidateHandle(&rpc->srpc_replymdh);
 190 }
 191
 192 static void
 193 srpc_service_fini(struct srpc_service *svc)
 194 {
 195         struct srpc_service_cd  *scd;
 196         struct srpc_server_rpc  *rpc;
 197         struct srpc_buffer      *buf;
 198         struct list_head                *q;
 199         int                     i;
 200
 201         if (svc->sv_cpt_data == NULL)
 202                 return;
 203
 204         cfs_percpt_for_each(scd, i, svc->sv_cpt_data) {
 205                 while (1) {
 206                         if (!list_empty(&scd->scd_buf_posted))
 207                                 q = &scd->scd_buf_posted;
 208                         else if (!list_empty(&scd->scd_buf_blocked))
 209                                 q = &scd->scd_buf_blocked;
 210                         else
 211                                 break;
 212
 213                         while (!list_empty(q)) {
 214                                 buf = list_entry(q->next,
 215                                                      struct srpc_buffer,
 216                                                      buf_list);
 217                                 list_del(&buf->buf_list);
 218                                 LIBCFS_FREE(buf, sizeof(*buf));
 219                         }
 220                 }
 221
 222                 LASSERT(list_empty(&scd->scd_rpc_active));
 223
 224                 while (!list_empty(&scd->scd_rpc_free)) {
 225                         rpc = list_entry(scd->scd_rpc_free.next,
 226                                              struct srpc_server_rpc,
 227                                              srpc_list);
 228                         list_del(&rpc->srpc_list);
 229                         LIBCFS_FREE(rpc, sizeof(*rpc));
 230                 }
 231         }
 232
 233         cfs_percpt_free(svc->sv_cpt_data);
 234         svc->sv_cpt_data = NULL;
 235 }
 236
 237 static int
 238 srpc_service_nrpcs(struct srpc_service *svc)
 239 {
 240         int nrpcs = svc->sv_wi_total / svc->sv_ncpts;
 241
 242         return srpc_serv_is_framework(svc) ?
 243                max(nrpcs, SFW_FRWK_WI_MIN) : max(nrpcs, SFW_TEST_WI_MIN);
 244 }
 245
 246 int srpc_add_buffer(struct swi_workitem *wi);
 247
 248 static int
 249 srpc_service_init(struct srpc_service *svc)
 250 {
 251         struct srpc_service_cd  *scd;
 252         struct srpc_server_rpc  *rpc;
 253         int                     nrpcs;
 254         int                     i;
 255         int                     j;
 256
 257         svc->sv_shuttingdown = 0;
 258
 259         svc->sv_cpt_data = cfs_percpt_alloc(lnet_cpt_table(),
 260                                             sizeof(struct srpc_service_cd));
 261         if (svc->sv_cpt_data == NULL)
 262                 return -ENOMEM;
 263
 264         svc->sv_ncpts = srpc_serv_is_framework(svc) ?
 265                         1 : cfs_cpt_number(lnet_cpt_table());
 266         nrpcs = srpc_service_nrpcs(svc);
 267
 268         cfs_percpt_for_each(scd, i, svc->sv_cpt_data) {
 269                 scd->scd_cpt = i;
 270                 scd->scd_svc = svc;
 271                 spin_lock_init(&scd->scd_lock);
 272                 INIT_LIST_HEAD(&scd->scd_rpc_free);
 273                 INIT_LIST_HEAD(&scd->scd_rpc_active);
 274                 INIT_LIST_HEAD(&scd->scd_buf_posted);
 275                 INIT_LIST_HEAD(&scd->scd_buf_blocked);
 276
 277                 scd->scd_ev.ev_data = scd;
 278                 scd->scd_ev.ev_type = SRPC_REQUEST_RCVD;
 279
 280                 /* NB: don't use lst_sched_serial for adding buffer,
 281                  * see details in srpc_service_add_buffers() */
 282                 swi_init_workitem(&scd->scd_buf_wi, scd,
 283                                   srpc_add_buffer, lst_sched_test[i]);
 284
 285                 if (i != 0 && srpc_serv_is_framework(svc)) {
 286                         /* NB: framework service only needs srpc_service_cd for
 287                          * one partition, but we allocate for all to make
 288                          * it easier to implement, it will waste a little
 289                          * memory but nobody should care about this */
 290                         continue;
 291                 }
 292
 293                 for (j = 0; j < nrpcs; j++) {
 294                         LIBCFS_CPT_ALLOC(rpc, lnet_cpt_table(),
 295                                          i, sizeof(*rpc));
 296                         if (rpc == NULL) {
 297                                 srpc_service_fini(svc);
 298                                 return -ENOMEM;
 299                         }
 300                         list_add(&rpc->srpc_list, &scd->scd_rpc_free);
 301                 }
 302         }
 303
 304         return 0;
 305 }
 306
 307 int
 308 srpc_add_service(struct srpc_service *sv)
 309 {
 310         int id = sv->sv_id;
 311
 312         LASSERT(0 <= id && id <= SRPC_SERVICE_MAX_ID);
 313
 314         if (srpc_service_init(sv) != 0)
 315                 return -ENOMEM;
 316
 317         spin_lock(&srpc_data.rpc_glock);
 318
 319         LASSERT(srpc_data.rpc_state == SRPC_STATE_RUNNING);
 320
 321         if (srpc_data.rpc_services[id] != NULL) {
 322                 spin_unlock(&srpc_data.rpc_glock);
 323                 goto failed;
 324         }
 325
 326         srpc_data.rpc_services[id] = sv;
 327         spin_unlock(&srpc_data.rpc_glock);
 328
 329         CDEBUG(D_NET, "Adding service: id %d, name %s\n", id, sv->sv_name);
 330         return 0;
 331
 332  failed:
 333         srpc_service_fini(sv);
 334         return -EBUSY;
 335 }
 336
 337 int
 338 srpc_remove_service (srpc_service_t *sv)
 339 {
 340         int id = sv->sv_id;
 341
 342         spin_lock(&srpc_data.rpc_glock);
 343
 344         if (srpc_data.rpc_services[id] != sv) {
 345                 spin_unlock(&srpc_data.rpc_glock);
 346                 return -ENOENT;
 347         }
 348
 349         srpc_data.rpc_services[id] = NULL;
 350         spin_unlock(&srpc_data.rpc_glock);
 351         return 0;
 352 }
 353
 354 int
 355 srpc_post_passive_rdma(int portal, int local, __u64 matchbits, void *buf,
 356                        int len, int options, lnet_process_id_t peer,
 357                        lnet_handle_md_t *mdh, srpc_event_t *ev)
 358 {
 359         int              rc;
 360         lnet_md_t        md;
 361         lnet_handle_me_t meh;
 362
 363         rc = LNetMEAttach(portal, peer, matchbits, 0, LNET_UNLINK,
 364                           local ? LNET_INS_LOCAL : LNET_INS_AFTER, &meh);
 365         if (rc != 0) {
 366                 CERROR ("LNetMEAttach failed: %d\n", rc);
 367                 LASSERT (rc == -ENOMEM);
 368                 return -ENOMEM;
 369         }
 370
 371         md.threshold = 1;
 372         md.user_ptr  = ev;
 373         md.start     = buf;
 374         md.length    = len;
 375         md.options   = options;
 376         md.eq_handle = srpc_data.rpc_lnet_eq;
 377
 378         rc = LNetMDAttach(meh, md, LNET_UNLINK, mdh);
 379         if (rc != 0) {
 380                 CERROR ("LNetMDAttach failed: %d\n", rc);
 381                 LASSERT (rc == -ENOMEM);
 382
 383                 rc = LNetMEUnlink(meh);
 384                 LASSERT (rc == 0);
 385                 return -ENOMEM;
 386         }
 387
 388         CDEBUG (D_NET,
 389                 "Posted passive RDMA: peer %s, portal %d, matchbits "LPX64"\n",
 390                 libcfs_id2str(peer), portal, matchbits);
 391         return 0;
 392 }
 393
 394 int
 395 srpc_post_active_rdma(int portal, __u64 matchbits, void *buf, int len,
 396                       int options, lnet_process_id_t peer, lnet_nid_t self,
 397                       lnet_handle_md_t *mdh, srpc_event_t *ev)
 398 {
 399         int       rc;
 400         lnet_md_t md;
 401
 402         md.user_ptr  = ev;
 403         md.start     = buf;
 404         md.length    = len;
 405         md.eq_handle = srpc_data.rpc_lnet_eq;
 406         md.threshold = ((options & LNET_MD_OP_GET) != 0) ? 2 : 1;
 407         md.options   = options & ~(LNET_MD_OP_PUT | LNET_MD_OP_GET);
 408
 409         rc = LNetMDBind(md, LNET_UNLINK, mdh);
 410         if (rc != 0) {
 411                 CERROR ("LNetMDBind failed: %d\n", rc);
 412                 LASSERT (rc == -ENOMEM);
 413                 return -ENOMEM;
 414         }
 415
 416         /* this is kind of an abuse of the LNET_MD_OP_{PUT,GET} options.
 417          * they're only meaningful for MDs attached to an ME (i.e. passive
 418          * buffers... */
 419         if ((options & LNET_MD_OP_PUT) != 0) {
 420                 rc = LNetPut(self, *mdh, LNET_NOACK_REQ, peer,
 421                              portal, matchbits, 0, 0);
 422         } else {
 423                 LASSERT ((options & LNET_MD_OP_GET) != 0);
 424
 425                 rc = LNetGet(self, *mdh, peer, portal, matchbits, 0);
 426         }
 427
 428         if (rc != 0) {
 429                 CERROR ("LNet%s(%s, %d, "LPD64") failed: %d\n",
 430                         ((options & LNET_MD_OP_PUT) != 0) ? "Put" : "Get",
 431                         libcfs_id2str(peer), portal, matchbits, rc);
 432
 433                 /* The forthcoming unlink event will complete this operation
 434                  * with failure, so fall through and return success here.
 435                  */
 436                 rc = LNetMDUnlink(*mdh);
 437                 LASSERT (rc == 0);
 438         } else {
 439                 CDEBUG (D_NET,
 440                         "Posted active RDMA: peer %s, portal %u, matchbits "LPX64"\n",
 441                         libcfs_id2str(peer), portal, matchbits);
 442         }
 443         return 0;
 444 }
 445
 446 int
 447 srpc_post_active_rqtbuf(lnet_process_id_t peer, int service, void *buf,
 448                         int len, lnet_handle_md_t *mdh, srpc_event_t *ev)
 449 {
 450         return srpc_post_active_rdma(srpc_serv_portal(service), service,
 451                                      buf, len, LNET_MD_OP_PUT, peer,
 452                                      LNET_NID_ANY, mdh, ev);
 453 }
 454
 455 int
 456 srpc_post_passive_rqtbuf(int service, int local, void *buf, int len,
 457                          lnet_handle_md_t *mdh, srpc_event_t *ev)
 458 {
 459         lnet_process_id_t any = {0};
 460
 461         any.nid = LNET_NID_ANY;
 462         any.pid = LNET_PID_ANY;
 463
 464         return srpc_post_passive_rdma(srpc_serv_portal(service),
 465                                       local, service, buf, len,
 466                                       LNET_MD_OP_PUT, any, mdh, ev);
 467 }
 468
 469 int
 470 srpc_service_post_buffer(struct srpc_service_cd *scd, struct srpc_buffer *buf)
 471 {
 472         struct srpc_service     *sv = scd->scd_svc;
 473         struct srpc_msg         *msg = &buf->buf_msg;
 474         int                     rc;
 475
 476         LNetInvalidateHandle(&buf->buf_mdh);
 477         list_add(&buf->buf_list, &scd->scd_buf_posted);
 478         scd->scd_buf_nposted++;
 479         spin_unlock(&scd->scd_lock);
 480
 481         rc = srpc_post_passive_rqtbuf(sv->sv_id,
 482                                       !srpc_serv_is_framework(sv),
 483                                       msg, sizeof(*msg), &buf->buf_mdh,
 484                                       &scd->scd_ev);
 485
 486         /* At this point, a RPC (new or delayed) may have arrived in
 487          * msg and its event handler has been called. So we must add
 488          * buf to scd_buf_posted _before_ dropping scd_lock */
 489
 490         spin_lock(&scd->scd_lock);
 491
 492         if (rc == 0) {
 493                 if (!sv->sv_shuttingdown)
 494                         return 0;
 495
 496                 spin_unlock(&scd->scd_lock);
 497                 /* srpc_shutdown_service might have tried to unlink me
 498                  * when my buf_mdh was still invalid */
 499                 LNetMDUnlink(buf->buf_mdh);
 500                 spin_lock(&scd->scd_lock);
 501                 return 0;
 502         }
 503
 504         scd->scd_buf_nposted--;
 505         if (sv->sv_shuttingdown)
 506                 return rc; /* don't allow to change scd_buf_posted */
 507
 508         list_del(&buf->buf_list);
 509         spin_unlock(&scd->scd_lock);
 510
 511         LIBCFS_FREE(buf, sizeof(*buf));
 512
 513         spin_lock(&scd->scd_lock);
 514         return rc;
 515 }
 516
 517 int
 518 srpc_add_buffer(struct swi_workitem *wi)
 519 {
 520         struct srpc_service_cd  *scd = wi->swi_workitem.wi_data;
 521         struct srpc_buffer      *buf;
 522         int                     rc = 0;
 523
 524         /* it's called by workitem scheduler threads, these threads
 525          * should have been set CPT affinity, so buffers will be posted
 526          * on CPT local list of Portal */
 527         spin_lock(&scd->scd_lock);
 528
 529         while (scd->scd_buf_adjust > 0 &&
 530                !scd->scd_svc->sv_shuttingdown) {
 531                 scd->scd_buf_adjust--; /* consume it */
 532                 scd->scd_buf_posting++;
 533
 534                 spin_unlock(&scd->scd_lock);
 535
 536                 LIBCFS_ALLOC(buf, sizeof(*buf));
 537                 if (buf == NULL) {
 538                         CERROR("Failed to add new buf to service: %s\n",
 539                                scd->scd_svc->sv_name);
 540                         spin_lock(&scd->scd_lock);
 541                         rc = -ENOMEM;
 542                         break;
 543                 }
 544
 545                 spin_lock(&scd->scd_lock);
 546                 if (scd->scd_svc->sv_shuttingdown) {
 547                         spin_unlock(&scd->scd_lock);
 548                         LIBCFS_FREE(buf, sizeof(*buf));
 549
 550                         spin_lock(&scd->scd_lock);
 551                         rc = -ESHUTDOWN;
 552                         break;
 553                 }
 554
 555                 rc = srpc_service_post_buffer(scd, buf);
 556                 if (rc != 0)
 557                         break; /* buf has been freed inside */
 558
 559                 LASSERT(scd->scd_buf_posting > 0);
 560                 scd->scd_buf_posting--;
 561                 scd->scd_buf_total++;
 562                 scd->scd_buf_low = MAX(2, scd->scd_buf_total / 4);
 563         }
 564
 565         if (rc != 0) {
 566                 scd->scd_buf_err_stamp = cfs_time_current_sec();
 567                 scd->scd_buf_err = rc;
 568
 569                 LASSERT(scd->scd_buf_posting > 0);
 570                 scd->scd_buf_posting--;
 571         }
 572
 573         spin_unlock(&scd->scd_lock);
 574         return 0;
 575 }
 576
 577 int
 578 srpc_service_add_buffers(struct srpc_service *sv, int nbuffer)
 579 {
 580         struct srpc_service_cd  *scd;
 581         int                     rc = 0;
 582         int                     i;
 583
 584         LASSERTF(nbuffer > 0, "nbuffer must be positive: %d\n", nbuffer);
 585
 586         cfs_percpt_for_each(scd, i, sv->sv_cpt_data) {
 587                 spin_lock(&scd->scd_lock);
 588
 589                 scd->scd_buf_err = 0;
 590                 scd->scd_buf_err_stamp = 0;
 591                 scd->scd_buf_posting = 0;
 592                 scd->scd_buf_adjust = nbuffer;
 593                 /* start to post buffers */
 594                 swi_schedule_workitem(&scd->scd_buf_wi);
 595                 spin_unlock(&scd->scd_lock);
 596
 597                 /* framework service only post buffer for one partition  */
 598                 if (srpc_serv_is_framework(sv))
 599                         break;
 600         }
 601
 602         cfs_percpt_for_each(scd, i, sv->sv_cpt_data) {
 603                 spin_lock(&scd->scd_lock);
 604                 /*
 605                  * NB: srpc_service_add_buffers() can be called inside
 606                  * thread context of lst_sched_serial, and we don't normally
 607                  * allow to sleep inside thread context of WI scheduler
 608                  * because it will block current scheduler thread from doing
 609                  * anything else, even worse, it could deadlock if it's
 610                  * waiting on result from another WI of the same scheduler.
 611                  * However, it's safe at here because scd_buf_wi is scheduled
 612                  * by thread in a different WI scheduler (lst_sched_test),
 613                  * so we don't have any risk of deadlock, though this could
 614                  * block all WIs pending on lst_sched_serial for a moment
 615                  * which is not good but not fatal.
 616                  */
 617                 lst_wait_until(scd->scd_buf_err != 0 ||
 618                                (scd->scd_buf_adjust == 0 &&
 619                                 scd->scd_buf_posting == 0),
 620                                scd->scd_lock, "waiting for adding buffer\n");
 621
 622                 if (scd->scd_buf_err != 0 && rc == 0)
 623                         rc = scd->scd_buf_err;
 624
 625                 spin_unlock(&scd->scd_lock);
 626         }
 627
 628         return rc;
 629 }
 630
 631 void
 632 srpc_service_remove_buffers(struct srpc_service *sv, int nbuffer)
 633 {
 634         struct srpc_service_cd  *scd;
 635         int                     num;
 636         int                     i;
 637
 638         LASSERT(!sv->sv_shuttingdown);
 639
 640         cfs_percpt_for_each(scd, i, sv->sv_cpt_data) {
 641                 spin_lock(&scd->scd_lock);
 642
 643                 num = scd->scd_buf_total + scd->scd_buf_posting;
 644                 scd->scd_buf_adjust -= min(nbuffer, num);
 645
 646                 spin_unlock(&scd->scd_lock);
 647         }
 648 }
 649
 650 /* returns 1 if sv has finished, otherwise 0 */
 651 int
 652 srpc_finish_service(struct srpc_service *sv)
 653 {
 654         struct srpc_service_cd  *scd;
 655         struct srpc_server_rpc  *rpc;
 656         int                     i;
 657
 658         LASSERT(sv->sv_shuttingdown); /* srpc_shutdown_service called */
 659
 660         cfs_percpt_for_each(scd, i, sv->sv_cpt_data) {
 661                 spin_lock(&scd->scd_lock);
 662                 if (!swi_deschedule_workitem(&scd->scd_buf_wi)) {
 663                         spin_unlock(&scd->scd_lock);
 664                         return 0;
 665                 }
 666
 667                 if (scd->scd_buf_nposted > 0) {
 668                         CDEBUG(D_NET, "waiting for %d posted buffers to unlink",
 669                                scd->scd_buf_nposted);
 670                         spin_unlock(&scd->scd_lock);
 671                         return 0;
 672                 }
 673
 674                 if (list_empty(&scd->scd_rpc_active)) {
 675                         spin_unlock(&scd->scd_lock);
 676                         continue;
 677                 }
 678
 679                 rpc = list_entry(scd->scd_rpc_active.next,
 680                                      struct srpc_server_rpc, srpc_list);
 681                 CNETERR("Active RPC %p on shutdown: sv %s, peer %s, "
 682                         "wi %s scheduled %d running %d, "
 683                         "ev fired %d type %d status %d lnet %d\n",
 684                         rpc, sv->sv_name, libcfs_id2str(rpc->srpc_peer),
 685                         swi_state2str(rpc->srpc_wi.swi_state),
 686                         rpc->srpc_wi.swi_workitem.wi_scheduled,
 687                         rpc->srpc_wi.swi_workitem.wi_running,
 688                         rpc->srpc_ev.ev_fired, rpc->srpc_ev.ev_type,
 689                         rpc->srpc_ev.ev_status, rpc->srpc_ev.ev_lnet);
 690                 spin_unlock(&scd->scd_lock);
 691                 return 0;
 692         }
 693
 694         /* no lock needed from now on */
 695         srpc_service_fini(sv);
 696         return 1;
 697 }
 698
 699 /* called with sv->sv_lock held */
 700 void
 701 srpc_service_recycle_buffer(struct srpc_service_cd *scd, srpc_buffer_t *buf)
 702 {
 703         if (!scd->scd_svc->sv_shuttingdown && scd->scd_buf_adjust >= 0) {
 704                 if (srpc_service_post_buffer(scd, buf) != 0) {
 705                         CWARN("Failed to post %s buffer\n",
 706                               scd->scd_svc->sv_name);
 707                 }
 708                 return;
 709         }
 710
 711         /* service is shutting down, or we want to recycle some buffers */
 712         scd->scd_buf_total--;
 713
 714         if (scd->scd_buf_adjust < 0) {
 715                 scd->scd_buf_adjust++;
 716                 if (scd->scd_buf_adjust < 0 &&
 717                     scd->scd_buf_total == 0 && scd->scd_buf_posting == 0) {
 718                         CDEBUG(D_INFO,
 719                                "Try to recyle %d buffers but nothing left\n",
 720                                scd->scd_buf_adjust);
 721                         scd->scd_buf_adjust = 0;
 722                 }
 723         }
 724
 725         spin_unlock(&scd->scd_lock);
 726         LIBCFS_FREE(buf, sizeof(*buf));
 727         spin_lock(&scd->scd_lock);
 728 }
 729
 730 void
 731 srpc_abort_service(struct srpc_service *sv)
 732 {
 733         struct srpc_service_cd  *scd;
 734         struct srpc_server_rpc  *rpc;
 735         int                     i;
 736
 737         CDEBUG(D_NET, "Aborting service: id %d, name %s\n",
 738                sv->sv_id, sv->sv_name);
 739
 740         cfs_percpt_for_each(scd, i, sv->sv_cpt_data) {
 741                 spin_lock(&scd->scd_lock);
 742
 743                 /* schedule in-flight RPCs to notice the abort, NB:
 744                  * racing with incoming RPCs; complete fix should make test
 745                  * RPCs carry session ID in its headers */
 746                 list_for_each_entry(rpc, &scd->scd_rpc_active, srpc_list) {
 747                         rpc->srpc_aborted = 1;
 748                         swi_schedule_workitem(&rpc->srpc_wi);
 749                 }
 750
 751                 spin_unlock(&scd->scd_lock);
 752         }
 753 }
 754
 755 void
 756 srpc_shutdown_service(srpc_service_t *sv)
 757 {
 758         struct srpc_service_cd  *scd;
 759         struct srpc_server_rpc  *rpc;
 760         srpc_buffer_t           *buf;
 761         int                     i;
 762
 763         CDEBUG(D_NET, "Shutting down service: id %d, name %s\n",
 764                sv->sv_id, sv->sv_name);
 765
 766         cfs_percpt_for_each(scd, i, sv->sv_cpt_data)
 767                 spin_lock(&scd->scd_lock);
 768
 769         sv->sv_shuttingdown = 1; /* i.e. no new active RPC */
 770
 771         cfs_percpt_for_each(scd, i, sv->sv_cpt_data)
 772                 spin_unlock(&scd->scd_lock);
 773
 774         cfs_percpt_for_each(scd, i, sv->sv_cpt_data) {
 775                 spin_lock(&scd->scd_lock);
 776
 777                 /* schedule in-flight RPCs to notice the shutdown */
 778                 list_for_each_entry(rpc, &scd->scd_rpc_active, srpc_list)
 779                         swi_schedule_workitem(&rpc->srpc_wi);
 780
 781                 spin_unlock(&scd->scd_lock);
 782
 783                 /* OK to traverse scd_buf_posted without lock, since no one
 784                  * touches scd_buf_posted now */
 785                 list_for_each_entry(buf, &scd->scd_buf_posted, buf_list)
 786                         LNetMDUnlink(buf->buf_mdh);
 787         }
 788 }
 789
 790 int
 791 srpc_send_request (srpc_client_rpc_t *rpc)
 792 {
 793         srpc_event_t *ev = &rpc->crpc_reqstev;
 794         int        rc;
 795
 796         ev->ev_fired = 0;
 797         ev->ev_data  = rpc;
 798         ev->ev_type  = SRPC_REQUEST_SENT;
 799
 800         rc = srpc_post_active_rqtbuf(rpc->crpc_dest, rpc->crpc_service,
 801                                      &rpc->crpc_reqstmsg, sizeof(srpc_msg_t),
 802                                      &rpc->crpc_reqstmdh, ev);
 803         if (rc != 0) {
 804                 LASSERT (rc == -ENOMEM);
 805                 ev->ev_fired = 1;  /* no more event expected */
 806         }
 807         return rc;
 808 }
 809
 810 int
 811 srpc_prepare_reply (srpc_client_rpc_t *rpc)
 812 {
 813         srpc_event_t *ev = &rpc->crpc_replyev;
 814         __u64   *id = &rpc->crpc_reqstmsg.msg_body.reqst.rpyid;
 815         int        rc;
 816
 817         ev->ev_fired = 0;
 818         ev->ev_data  = rpc;
 819         ev->ev_type  = SRPC_REPLY_RCVD;
 820
 821         *id = srpc_next_id();
 822
 823         rc = srpc_post_passive_rdma(SRPC_RDMA_PORTAL, 0, *id,
 824                                     &rpc->crpc_replymsg, sizeof(srpc_msg_t),
 825                                     LNET_MD_OP_PUT, rpc->crpc_dest,
 826                                     &rpc->crpc_replymdh, ev);
 827         if (rc != 0) {
 828                 LASSERT (rc == -ENOMEM);
 829                 ev->ev_fired = 1;  /* no more event expected */
 830         }
 831         return rc;
 832 }
 833
 834 int
 835 srpc_prepare_bulk (srpc_client_rpc_t *rpc)
 836 {
 837         srpc_bulk_t  *bk = &rpc->crpc_bulk;
 838         srpc_event_t *ev = &rpc->crpc_bulkev;
 839         __u64   *id = &rpc->crpc_reqstmsg.msg_body.reqst.bulkid;
 840         int        rc;
 841         int        opt;
 842
 843         LASSERT (bk->bk_niov <= LNET_MAX_IOV);
 844
 845         if (bk->bk_niov == 0) return 0; /* nothing to do */
 846
 847         opt = bk->bk_sink ? LNET_MD_OP_PUT : LNET_MD_OP_GET;
 848         opt |= LNET_MD_KIOV;
 849
 850         ev->ev_fired = 0;
 851         ev->ev_data  = rpc;
 852         ev->ev_type  = SRPC_BULK_REQ_RCVD;
 853
 854         *id = srpc_next_id();
 855
 856         rc = srpc_post_passive_rdma(SRPC_RDMA_PORTAL, 0, *id,
 857                                     &bk->bk_iovs[0], bk->bk_niov, opt,
 858                                     rpc->crpc_dest, &bk->bk_mdh, ev);
 859         if (rc != 0) {
 860                 LASSERT (rc == -ENOMEM);
 861                 ev->ev_fired = 1;  /* no more event expected */
 862         }
 863         return rc;
 864 }
 865
 866 int
 867 srpc_do_bulk (srpc_server_rpc_t *rpc)
 868 {
 869         srpc_event_t  *ev = &rpc->srpc_ev;
 870         srpc_bulk_t   *bk = rpc->srpc_bulk;
 871         __u64     id = rpc->srpc_reqstbuf->buf_msg.msg_body.reqst.bulkid;
 872         int         rc;
 873         int         opt;
 874
 875         LASSERT (bk != NULL);
 876
 877         opt = bk->bk_sink ? LNET_MD_OP_GET : LNET_MD_OP_PUT;
 878         opt |= LNET_MD_KIOV;
 879
 880         ev->ev_fired = 0;
 881         ev->ev_data  = rpc;
 882         ev->ev_type  = bk->bk_sink ? SRPC_BULK_GET_RPLD : SRPC_BULK_PUT_SENT;
 883
 884         rc = srpc_post_active_rdma(SRPC_RDMA_PORTAL, id,
 885                                    &bk->bk_iovs[0], bk->bk_niov, opt,
 886                                    rpc->srpc_peer, rpc->srpc_self,
 887                                    &bk->bk_mdh, ev);
 888         if (rc != 0)
 889                 ev->ev_fired = 1;  /* no more event expected */
 890         return rc;
 891 }
 892
 893 /* only called from srpc_handle_rpc */
 894 void
 895 srpc_server_rpc_done(srpc_server_rpc_t *rpc, int status)
 896 {
 897         struct srpc_service_cd  *scd = rpc->srpc_scd;
 898         struct srpc_service     *sv  = scd->scd_svc;
 899         srpc_buffer_t           *buffer;
 900
 901         LASSERT (status != 0 || rpc->srpc_wi.swi_state == SWI_STATE_DONE);
 902
 903         rpc->srpc_status = status;
 904
 905         CDEBUG_LIMIT (status == 0 ? D_NET : D_NETERROR,
 906                 "Server RPC %p done: service %s, peer %s, status %s:%d\n",
 907                 rpc, sv->sv_name, libcfs_id2str(rpc->srpc_peer),
 908                 swi_state2str(rpc->srpc_wi.swi_state), status);
 909
 910         if (status != 0) {
 911                 spin_lock(&srpc_data.rpc_glock);
 912                 srpc_data.rpc_counters.rpcs_dropped++;
 913                 spin_unlock(&srpc_data.rpc_glock);
 914         }
 915
 916         if (rpc->srpc_done != NULL)
 917                 (*rpc->srpc_done) (rpc);
 918         LASSERT(rpc->srpc_bulk == NULL);
 919
 920         spin_lock(&scd->scd_lock);
 921
 922         if (rpc->srpc_reqstbuf != NULL) {
 923                 /* NB might drop sv_lock in srpc_service_recycle_buffer, but
 924                  * sv won't go away for scd_rpc_active must not be empty */
 925                 srpc_service_recycle_buffer(scd, rpc->srpc_reqstbuf);
 926                 rpc->srpc_reqstbuf = NULL;
 927         }
 928
 929         list_del(&rpc->srpc_list); /* from scd->scd_rpc_active */
 930
 931         /*
 932          * No one can schedule me now since:
 933          * - I'm not on scd_rpc_active.
 934          * - all LNet events have been fired.
 935          * Cancel pending schedules and prevent future schedule attempts:
 936          */
 937         LASSERT(rpc->srpc_ev.ev_fired);
 938         swi_exit_workitem(&rpc->srpc_wi);
 939
 940         if (!sv->sv_shuttingdown && !list_empty(&scd->scd_buf_blocked)) {
 941                 buffer = list_entry(scd->scd_buf_blocked.next,
 942                                         srpc_buffer_t, buf_list);
 943                 list_del(&buffer->buf_list);
 944
 945                 srpc_init_server_rpc(rpc, scd, buffer);
 946                 list_add_tail(&rpc->srpc_list, &scd->scd_rpc_active);
 947                 swi_schedule_workitem(&rpc->srpc_wi);
 948         } else {
 949                 list_add(&rpc->srpc_list, &scd->scd_rpc_free);
 950         }
 951
 952         spin_unlock(&scd->scd_lock);
 953         return;
 954 }
 955
 956 /* handles an incoming RPC */
 957 int
 958 srpc_handle_rpc(swi_workitem_t *wi)
 959 {
 960         struct srpc_server_rpc  *rpc = wi->swi_workitem.wi_data;
 961         struct srpc_service_cd  *scd = rpc->srpc_scd;
 962         struct srpc_service     *sv = scd->scd_svc;
 963         srpc_event_t            *ev = &rpc->srpc_ev;
 964         int                     rc = 0;
 965
 966         LASSERT(wi == &rpc->srpc_wi);
 967
 968         spin_lock(&scd->scd_lock);
 969
 970         if (sv->sv_shuttingdown || rpc->srpc_aborted) {
 971                 spin_unlock(&scd->scd_lock);
 972
 973                 if (rpc->srpc_bulk != NULL)
 974                         LNetMDUnlink(rpc->srpc_bulk->bk_mdh);
 975                 LNetMDUnlink(rpc->srpc_replymdh);
 976
 977                 if (ev->ev_fired) { /* no more event, OK to finish */
 978                         srpc_server_rpc_done(rpc, -ESHUTDOWN);
 979                         return 1;
 980                 }
 981                 return 0;
 982         }
 983
 984         spin_unlock(&scd->scd_lock);
 985
 986         switch (wi->swi_state) {
 987         default:
 988                 LBUG ();
 989         case SWI_STATE_NEWBORN: {
 990                 srpc_msg_t         *msg;
 991                 srpc_generic_reply_t *reply;
 992
 993                 msg = &rpc->srpc_reqstbuf->buf_msg;
 994                 reply = &rpc->srpc_replymsg.msg_body.reply;
 995
 996                 if (msg->msg_magic == 0) {
 997                         /* moaned already in srpc_lnet_ev_handler */
 998                         srpc_server_rpc_done(rpc, EBADMSG);
 999                         return 1;
1000                 }
1001
1002                 srpc_unpack_msg_hdr(msg);
1003                 if (msg->msg_version != SRPC_MSG_VERSION) {
1004                         CWARN("Version mismatch: %u, %u expected, from %s\n",
1005                               msg->msg_version, SRPC_MSG_VERSION,
1006                               libcfs_id2str(rpc->srpc_peer));
1007                         reply->status = EPROTO;
1008                         /* drop through and send reply */
1009                 } else {
1010                         reply->status = 0;
1011                         rc = (*sv->sv_handler)(rpc);
1012                         LASSERT(reply->status == 0 || !rpc->srpc_bulk);
1013                         if (rc != 0) {
1014                                 srpc_server_rpc_done(rpc, rc);
1015                                 return 1;
1016                         }
1017                 }
1018
1019                 wi->swi_state = SWI_STATE_BULK_STARTED;
1020
1021                 if (rpc->srpc_bulk != NULL) {
1022                         rc = srpc_do_bulk(rpc);
1023                         if (rc == 0)
1024                                 return 0; /* wait for bulk */
1025
1026                         LASSERT (ev->ev_fired);
1027                         ev->ev_status = rc;
1028                 }
1029         }
1030         case SWI_STATE_BULK_STARTED:
1031                 LASSERT (rpc->srpc_bulk == NULL || ev->ev_fired);
1032
1033                 if (rpc->srpc_bulk != NULL) {
1034                         rc = ev->ev_status;
1035
1036                         if (sv->sv_bulk_ready != NULL)
1037                                 rc = (*sv->sv_bulk_ready) (rpc, rc);
1038
1039                         if (rc != 0) {
1040                                 srpc_server_rpc_done(rpc, rc);
1041                                 return 1;
1042                         }
1043                 }
1044
1045                 wi->swi_state = SWI_STATE_REPLY_SUBMITTED;
1046                 rc = srpc_send_reply(rpc);
1047                 if (rc == 0)
1048                         return 0; /* wait for reply */
1049                 srpc_server_rpc_done(rpc, rc);
1050                 return 1;
1051
1052         case SWI_STATE_REPLY_SUBMITTED:
1053                 if (!ev->ev_fired) {
1054                         CERROR("RPC %p: bulk %p, service %d\n",
1055                                rpc, rpc->srpc_bulk, sv->sv_id);
1056                         CERROR("Event: status %d, type %d, lnet %d\n",
1057                                ev->ev_status, ev->ev_type, ev->ev_lnet);
1058                         LASSERT (ev->ev_fired);
1059                 }
1060
1061                 wi->swi_state = SWI_STATE_DONE;
1062                 srpc_server_rpc_done(rpc, ev->ev_status);
1063                 return 1;
1064         }
1065
1066         return 0;
1067 }
1068
1069 void
1070 srpc_client_rpc_expired (void *data)
1071 {
1072         srpc_client_rpc_t *rpc = data;
1073
1074         CWARN ("Client RPC expired: service %d, peer %s, timeout %d.\n",
1075                rpc->crpc_service, libcfs_id2str(rpc->crpc_dest),
1076                rpc->crpc_timeout);
1077
1078         spin_lock(&rpc->crpc_lock);
1079
1080         rpc->crpc_timeout = 0;
1081         srpc_abort_rpc(rpc, -ETIMEDOUT);
1082
1083         spin_unlock(&rpc->crpc_lock);
1084
1085         spin_lock(&srpc_data.rpc_glock);
1086         srpc_data.rpc_counters.rpcs_expired++;
1087         spin_unlock(&srpc_data.rpc_glock);
1088 }
1089
1090 inline void
1091 srpc_add_client_rpc_timer (srpc_client_rpc_t *rpc)
1092 {
1093         stt_timer_t *timer = &rpc->crpc_timer;
1094
1095         if (rpc->crpc_timeout == 0) return;
1096
1097         INIT_LIST_HEAD(&timer->stt_list);
1098         timer->stt_data    = rpc;
1099         timer->stt_func    = srpc_client_rpc_expired;
1100         timer->stt_expires = cfs_time_add(rpc->crpc_timeout,
1101                                           cfs_time_current_sec());
1102         stt_add_timer(timer);
1103         return;
1104 }
1105
1106 /*
1107  * Called with rpc->crpc_lock held.
1108  *
1109  * Upon exit the RPC expiry timer is not queued and the handler is not
1110  * running on any CPU. */
1111 void
1112 srpc_del_client_rpc_timer (srpc_client_rpc_t *rpc)
1113 {
1114         /* timer not planted or already exploded */
1115         if (rpc->crpc_timeout == 0)
1116                 return;
1117
1118         /* timer successfully defused */
1119         if (stt_del_timer(&rpc->crpc_timer))
1120                 return;
1121
1122         /* timer detonated, wait for it to explode */
1123         while (rpc->crpc_timeout != 0) {
1124                 spin_unlock(&rpc->crpc_lock);
1125
1126                 schedule();
1127
1128                 spin_lock(&rpc->crpc_lock);
1129         }
1130 }
1131
1132 void
1133 srpc_client_rpc_done (srpc_client_rpc_t *rpc, int status)
1134 {
1135         swi_workitem_t *wi = &rpc->crpc_wi;
1136
1137         LASSERT(status != 0 || wi->swi_state == SWI_STATE_DONE);
1138
1139         spin_lock(&rpc->crpc_lock);
1140
1141         rpc->crpc_closed = 1;
1142         if (rpc->crpc_status == 0)
1143                 rpc->crpc_status = status;
1144
1145         srpc_del_client_rpc_timer(rpc);
1146
1147         CDEBUG_LIMIT ((status == 0) ? D_NET : D_NETERROR,
1148                 "Client RPC done: service %d, peer %s, status %s:%d:%d\n",
1149                 rpc->crpc_service, libcfs_id2str(rpc->crpc_dest),
1150                 swi_state2str(wi->swi_state), rpc->crpc_aborted, status);
1151
1152         /*
1153          * No one can schedule me now since:
1154          * - RPC timer has been defused.
1155          * - all LNet events have been fired.
1156          * - crpc_closed has been set, preventing srpc_abort_rpc from
1157          *   scheduling me.
1158          * Cancel pending schedules and prevent future schedule attempts:
1159          */
1160         LASSERT (!srpc_event_pending(rpc));
1161         swi_exit_workitem(wi);
1162
1163         spin_unlock(&rpc->crpc_lock);
1164
1165         (*rpc->crpc_done)(rpc);
1166         return;
1167 }
1168
1169 /* sends an outgoing RPC */
1170 int
1171 srpc_send_rpc (swi_workitem_t *wi)
1172 {
1173         int             rc = 0;
1174         srpc_client_rpc_t *rpc;
1175         srpc_msg_t      *reply;
1176         int             do_bulk;
1177
1178         LASSERT(wi != NULL);
1179
1180         rpc = wi->swi_workitem.wi_data;
1181
1182         LASSERT (rpc != NULL);
1183         LASSERT (wi == &rpc->crpc_wi);
1184
1185         reply = &rpc->crpc_replymsg;
1186         do_bulk = rpc->crpc_bulk.bk_niov > 0;
1187
1188         spin_lock(&rpc->crpc_lock);
1189
1190         if (rpc->crpc_aborted) {
1191                 spin_unlock(&rpc->crpc_lock);
1192                 goto abort;
1193         }
1194
1195         spin_unlock(&rpc->crpc_lock);
1196
1197         switch (wi->swi_state) {
1198         default:
1199                 LBUG ();
1200         case SWI_STATE_NEWBORN:
1201                 LASSERT (!srpc_event_pending(rpc));
1202
1203                 rc = srpc_prepare_reply(rpc);
1204                 if (rc != 0) {
1205                         srpc_client_rpc_done(rpc, rc);
1206                         return 1;
1207                 }
1208
1209                 rc = srpc_prepare_bulk(rpc);
1210                 if (rc != 0) break;
1211
1212                 wi->swi_state = SWI_STATE_REQUEST_SUBMITTED;
1213                 rc = srpc_send_request(rpc);
1214                 break;
1215
1216         case SWI_STATE_REQUEST_SUBMITTED:
1217                 /* CAVEAT EMPTOR: rqtev, rpyev, and bulkev may come in any
1218                  * order; however, they're processed in a strict order:
1219                  * rqt, rpy, and bulk. */
1220                 if (!rpc->crpc_reqstev.ev_fired) break;
1221
1222                 rc = rpc->crpc_reqstev.ev_status;
1223                 if (rc != 0) break;
1224
1225                 wi->swi_state = SWI_STATE_REQUEST_SENT;
1226                 /* perhaps more events, fall thru */
1227         case SWI_STATE_REQUEST_SENT: {
1228                 srpc_msg_type_t type = srpc_service2reply(rpc->crpc_service);
1229
1230                 if (!rpc->crpc_replyev.ev_fired) break;
1231
1232                 rc = rpc->crpc_replyev.ev_status;
1233                 if (rc != 0) break;
1234
1235                 srpc_unpack_msg_hdr(reply);
1236                 if (reply->msg_type != type ||
1237                     (reply->msg_magic != SRPC_MSG_MAGIC &&
1238                      reply->msg_magic != __swab32(SRPC_MSG_MAGIC))) {
1239                         CWARN ("Bad message from %s: type %u (%d expected),"
1240                                " magic %u (%d expected).\n",
1241                                libcfs_id2str(rpc->crpc_dest),
1242                                reply->msg_type, type,
1243                                reply->msg_magic, SRPC_MSG_MAGIC);
1244                         rc = -EBADMSG;
1245                         break;
1246                 }
1247
1248                 if (do_bulk && reply->msg_body.reply.status != 0) {
1249                         CWARN ("Remote error %d at %s, unlink bulk buffer in "
1250                                "case peer didn't initiate bulk transfer\n",
1251                                reply->msg_body.reply.status,
1252                                libcfs_id2str(rpc->crpc_dest));
1253                         LNetMDUnlink(rpc->crpc_bulk.bk_mdh);
1254                 }
1255
1256                 wi->swi_state = SWI_STATE_REPLY_RECEIVED;
1257         }
1258         case SWI_STATE_REPLY_RECEIVED:
1259                 if (do_bulk && !rpc->crpc_bulkev.ev_fired) break;
1260
1261                 rc = do_bulk ? rpc->crpc_bulkev.ev_status : 0;
1262
1263                 /* Bulk buffer was unlinked due to remote error. Clear error
1264                  * since reply buffer still contains valid data.
1265                  * NB rpc->crpc_done shouldn't look into bulk data in case of
1266                  * remote error. */
1267                 if (do_bulk && rpc->crpc_bulkev.ev_lnet == LNET_EVENT_UNLINK &&
1268                     rpc->crpc_status == 0 && reply->msg_body.reply.status != 0)
1269                         rc = 0;
1270
1271                 wi->swi_state = SWI_STATE_DONE;
1272                 srpc_client_rpc_done(rpc, rc);
1273                 return 1;
1274         }
1275
1276         if (rc != 0) {
1277                 spin_lock(&rpc->crpc_lock);
1278                 srpc_abort_rpc(rpc, rc);
1279                 spin_unlock(&rpc->crpc_lock);
1280         }
1281
1282 abort:
1283         if (rpc->crpc_aborted) {
1284                 LNetMDUnlink(rpc->crpc_reqstmdh);
1285                 LNetMDUnlink(rpc->crpc_replymdh);
1286                 LNetMDUnlink(rpc->crpc_bulk.bk_mdh);
1287
1288                 if (!srpc_event_pending(rpc)) {
1289                         srpc_client_rpc_done(rpc, -EINTR);
1290                         return 1;
1291                 }
1292         }
1293         return 0;
1294 }
1295
1296 srpc_client_rpc_t *
1297 srpc_create_client_rpc (lnet_process_id_t peer, int service,
1298                         int nbulkiov, int bulklen,
1299                         void (*rpc_done)(srpc_client_rpc_t *),
1300                         void (*rpc_fini)(srpc_client_rpc_t *), void *priv)
1301 {
1302         srpc_client_rpc_t *rpc;
1303
1304         LIBCFS_ALLOC(rpc, offsetof(srpc_client_rpc_t,
1305                                    crpc_bulk.bk_iovs[nbulkiov]));
1306         if (rpc == NULL)
1307                 return NULL;
1308
1309         srpc_init_client_rpc(rpc, peer, service, nbulkiov,
1310                              bulklen, rpc_done, rpc_fini, priv);
1311         return rpc;
1312 }
1313
1314 /* called with rpc->crpc_lock held */
1315 void
1316 srpc_abort_rpc (srpc_client_rpc_t *rpc, int why)
1317 {
1318         LASSERT (why != 0);
1319
1320         if (rpc->crpc_aborted || /* already aborted */
1321             rpc->crpc_closed)    /* callback imminent */
1322                 return;
1323
1324         CDEBUG (D_NET,
1325                 "Aborting RPC: service %d, peer %s, state %s, why %d\n",
1326                 rpc->crpc_service, libcfs_id2str(rpc->crpc_dest),
1327                 swi_state2str(rpc->crpc_wi.swi_state), why);
1328
1329         rpc->crpc_aborted = 1;
1330         rpc->crpc_status  = why;
1331         swi_schedule_workitem(&rpc->crpc_wi);
1332         return;
1333 }
1334
1335 /* called with rpc->crpc_lock held */
1336 void
1337 srpc_post_rpc (srpc_client_rpc_t *rpc)
1338 {
1339         LASSERT (!rpc->crpc_aborted);
1340         LASSERT (srpc_data.rpc_state == SRPC_STATE_RUNNING);
1341
1342         CDEBUG (D_NET, "Posting RPC: peer %s, service %d, timeout %d\n",
1343                 libcfs_id2str(rpc->crpc_dest), rpc->crpc_service,
1344                 rpc->crpc_timeout);
1345
1346         srpc_add_client_rpc_timer(rpc);
1347         swi_schedule_workitem(&rpc->crpc_wi);
1348         return;
1349 }
1350
1351
1352 int
1353 srpc_send_reply(struct srpc_server_rpc *rpc)
1354 {
1355         srpc_event_t            *ev = &rpc->srpc_ev;
1356         struct srpc_msg         *msg = &rpc->srpc_replymsg;
1357         struct srpc_buffer      *buffer = rpc->srpc_reqstbuf;
1358         struct srpc_service_cd  *scd = rpc->srpc_scd;
1359         struct srpc_service     *sv = scd->scd_svc;
1360         __u64                   rpyid;
1361         int                     rc;
1362
1363         LASSERT(buffer != NULL);
1364         rpyid = buffer->buf_msg.msg_body.reqst.rpyid;
1365
1366         spin_lock(&scd->scd_lock);
1367
1368         if (!sv->sv_shuttingdown && !srpc_serv_is_framework(sv)) {
1369                 /* Repost buffer before replying since test client
1370                  * might send me another RPC once it gets the reply */
1371                 if (srpc_service_post_buffer(scd, buffer) != 0)
1372                         CWARN("Failed to repost %s buffer\n", sv->sv_name);
1373                 rpc->srpc_reqstbuf = NULL;
1374         }
1375
1376         spin_unlock(&scd->scd_lock);
1377
1378         ev->ev_fired = 0;
1379         ev->ev_data  = rpc;
1380         ev->ev_type  = SRPC_REPLY_SENT;
1381
1382         msg->msg_magic   = SRPC_MSG_MAGIC;
1383         msg->msg_version = SRPC_MSG_VERSION;
1384         msg->msg_type    = srpc_service2reply(sv->sv_id);
1385
1386         rc = srpc_post_active_rdma(SRPC_RDMA_PORTAL, rpyid, msg,
1387                                    sizeof(*msg), LNET_MD_OP_PUT,
1388                                    rpc->srpc_peer, rpc->srpc_self,
1389                                    &rpc->srpc_replymdh, ev);
1390         if (rc != 0)
1391                 ev->ev_fired = 1;  /* no more event expected */
1392         return rc;
1393 }
1394
1395 /* when in kernel always called with LNET_LOCK() held, and in thread context */
1396 void
1397 srpc_lnet_ev_handler(lnet_event_t *ev)
1398 {
1399         struct srpc_service_cd  *scd;
1400         srpc_event_t      *rpcev = ev->md.user_ptr;
1401         srpc_client_rpc_t *crpc;
1402         srpc_server_rpc_t *srpc;
1403         srpc_buffer_t     *buffer;
1404         srpc_service_t    *sv;
1405         srpc_msg_t      *msg;
1406         srpc_msg_type_t    type;
1407
1408         LASSERT (!in_interrupt());
1409
1410         if (ev->status != 0) {
1411                 spin_lock(&srpc_data.rpc_glock);
1412                 srpc_data.rpc_counters.errors++;
1413                 spin_unlock(&srpc_data.rpc_glock);
1414         }
1415
1416         rpcev->ev_lnet = ev->type;
1417
1418         switch (rpcev->ev_type) {
1419         default:
1420                 CERROR("Unknown event: status %d, type %d, lnet %d\n",
1421                        rpcev->ev_status, rpcev->ev_type, rpcev->ev_lnet);
1422                 LBUG ();
1423         case SRPC_REQUEST_SENT:
1424                 if (ev->status == 0 && ev->type != LNET_EVENT_UNLINK) {
1425                         spin_lock(&srpc_data.rpc_glock);
1426                         srpc_data.rpc_counters.rpcs_sent++;
1427                         spin_unlock(&srpc_data.rpc_glock);
1428                 }
1429         case SRPC_REPLY_RCVD:
1430         case SRPC_BULK_REQ_RCVD:
1431                 crpc = rpcev->ev_data;
1432
1433                 if (rpcev != &crpc->crpc_reqstev &&
1434                     rpcev != &crpc->crpc_replyev &&
1435                     rpcev != &crpc->crpc_bulkev) {
1436                         CERROR("rpcev %p, crpc %p, reqstev %p, replyev %p, bulkev %p\n",
1437                                rpcev, crpc, &crpc->crpc_reqstev,
1438                                &crpc->crpc_replyev, &crpc->crpc_bulkev);
1439                         CERROR("Bad event: status %d, type %d, lnet %d\n",
1440                                rpcev->ev_status, rpcev->ev_type, rpcev->ev_lnet);
1441                         LBUG ();
1442                 }
1443
1444                 spin_lock(&crpc->crpc_lock);
1445
1446                 LASSERT(rpcev->ev_fired == 0);
1447                 rpcev->ev_fired  = 1;
1448                 rpcev->ev_status = (ev->type == LNET_EVENT_UNLINK) ?
1449                                                 -EINTR : ev->status;
1450                 swi_schedule_workitem(&crpc->crpc_wi);
1451
1452                 spin_unlock(&crpc->crpc_lock);
1453                 break;
1454
1455         case SRPC_REQUEST_RCVD:
1456                 scd = rpcev->ev_data;
1457                 sv = scd->scd_svc;
1458
1459                 LASSERT(rpcev == &scd->scd_ev);
1460
1461                 spin_lock(&scd->scd_lock);
1462
1463                 LASSERT (ev->unlinked);
1464                 LASSERT (ev->type == LNET_EVENT_PUT ||
1465                          ev->type == LNET_EVENT_UNLINK);
1466                 LASSERT (ev->type != LNET_EVENT_UNLINK ||
1467                          sv->sv_shuttingdown);
1468
1469                 buffer = container_of(ev->md.start, srpc_buffer_t, buf_msg);
1470                 buffer->buf_peer = ev->initiator;
1471                 buffer->buf_self = ev->target.nid;
1472
1473                 LASSERT(scd->scd_buf_nposted > 0);
1474                 scd->scd_buf_nposted--;
1475
1476                 if (sv->sv_shuttingdown) {
1477                         /* Leave buffer on scd->scd_buf_nposted since
1478                          * srpc_finish_service needs to traverse it. */
1479                         spin_unlock(&scd->scd_lock);
1480                         break;
1481                 }
1482
1483                 if (scd->scd_buf_err_stamp != 0 &&
1484                     scd->scd_buf_err_stamp < cfs_time_current_sec()) {
1485                         /* re-enable adding buffer */
1486                         scd->scd_buf_err_stamp = 0;
1487                         scd->scd_buf_err = 0;
1488                 }
1489
1490                 if (scd->scd_buf_err == 0 && /* adding buffer is enabled */
1491                     scd->scd_buf_adjust == 0 &&
1492                     scd->scd_buf_nposted < scd->scd_buf_low) {
1493                         scd->scd_buf_adjust = MAX(scd->scd_buf_total / 2,
1494                                                   SFW_TEST_WI_MIN);
1495                         swi_schedule_workitem(&scd->scd_buf_wi);
1496                 }
1497
1498                 list_del(&buffer->buf_list); /* from scd->scd_buf_posted */
1499                 msg = &buffer->buf_msg;
1500                 type = srpc_service2request(sv->sv_id);
1501
1502                 if (ev->status != 0 || ev->mlength != sizeof(*msg) ||
1503                     (msg->msg_type != type &&
1504                      msg->msg_type != __swab32(type)) ||
1505                     (msg->msg_magic != SRPC_MSG_MAGIC &&
1506                      msg->msg_magic != __swab32(SRPC_MSG_MAGIC))) {
1507                         CERROR ("Dropping RPC (%s) from %s: "
1508                                 "status %d mlength %d type %u magic %u.\n",
1509                                 sv->sv_name, libcfs_id2str(ev->initiator),
1510                                 ev->status, ev->mlength,
1511                                 msg->msg_type, msg->msg_magic);
1512
1513                         /* NB can't call srpc_service_recycle_buffer here since
1514                          * it may call LNetM[DE]Attach. The invalid magic tells
1515                          * srpc_handle_rpc to drop this RPC */
1516                         msg->msg_magic = 0;
1517                 }
1518
1519                 if (!list_empty(&scd->scd_rpc_free)) {
1520                         srpc = list_entry(scd->scd_rpc_free.next,
1521                                               struct srpc_server_rpc,
1522                                               srpc_list);
1523                         list_del(&srpc->srpc_list);
1524
1525                         srpc_init_server_rpc(srpc, scd, buffer);
1526                         list_add_tail(&srpc->srpc_list,
1527                                           &scd->scd_rpc_active);
1528                         swi_schedule_workitem(&srpc->srpc_wi);
1529                 } else {
1530                         list_add_tail(&buffer->buf_list,
1531                                           &scd->scd_buf_blocked);
1532                 }
1533
1534                 spin_unlock(&scd->scd_lock);
1535
1536                 spin_lock(&srpc_data.rpc_glock);
1537                 srpc_data.rpc_counters.rpcs_rcvd++;
1538                 spin_unlock(&srpc_data.rpc_glock);
1539                 break;
1540
1541         case SRPC_BULK_GET_RPLD:
1542                 LASSERT (ev->type == LNET_EVENT_SEND ||
1543                          ev->type == LNET_EVENT_REPLY ||
1544                          ev->type == LNET_EVENT_UNLINK);
1545
1546                 if (!ev->unlinked)
1547                         break; /* wait for final event */
1548
1549         case SRPC_BULK_PUT_SENT:
1550                 if (ev->status == 0 && ev->type != LNET_EVENT_UNLINK) {
1551                         spin_lock(&srpc_data.rpc_glock);
1552
1553                         if (rpcev->ev_type == SRPC_BULK_GET_RPLD)
1554                                 srpc_data.rpc_counters.bulk_get += ev->mlength;
1555                         else
1556                                 srpc_data.rpc_counters.bulk_put += ev->mlength;
1557
1558                         spin_unlock(&srpc_data.rpc_glock);
1559                 }
1560         case SRPC_REPLY_SENT:
1561                 srpc = rpcev->ev_data;
1562                 scd  = srpc->srpc_scd;
1563
1564                 LASSERT(rpcev == &srpc->srpc_ev);
1565
1566                 spin_lock(&scd->scd_lock);
1567
1568                 rpcev->ev_fired  = 1;
1569                 rpcev->ev_status = (ev->type == LNET_EVENT_UNLINK) ?
1570                                    -EINTR : ev->status;
1571                 swi_schedule_workitem(&srpc->srpc_wi);
1572
1573                 spin_unlock(&scd->scd_lock);
1574                 break;
1575         }
1576 }
1577
1578
1579 int
1580 srpc_startup (void)
1581 {
1582         int rc;
1583
1584         memset(&srpc_data, 0, sizeof(struct smoketest_rpc));
1585         spin_lock_init(&srpc_data.rpc_glock);
1586
1587         /* 1 second pause to avoid timestamp reuse */
1588         cfs_pause(cfs_time_seconds(1));
1589         srpc_data.rpc_matchbits = ((__u64) cfs_time_current_sec()) << 48;
1590
1591         srpc_data.rpc_state = SRPC_STATE_NONE;
1592
1593         rc = LNetNIInit(LUSTRE_SRV_LNET_PID);
1594         if (rc < 0) {
1595                 CERROR ("LNetNIInit() has failed: %d\n", rc);
1596                 return rc;
1597         }
1598
1599         srpc_data.rpc_state = SRPC_STATE_NI_INIT;
1600
1601         LNetInvalidateHandle(&srpc_data.rpc_lnet_eq);
1602         rc = LNetEQAlloc(0, srpc_lnet_ev_handler, &srpc_data.rpc_lnet_eq);
1603         if (rc != 0) {
1604                 CERROR("LNetEQAlloc() has failed: %d\n", rc);
1605                 goto bail;
1606         }
1607
1608         rc = LNetSetLazyPortal(SRPC_FRAMEWORK_REQUEST_PORTAL);
1609         LASSERT(rc == 0);
1610         rc = LNetSetLazyPortal(SRPC_REQUEST_PORTAL);
1611         LASSERT(rc == 0);
1612
1613         srpc_data.rpc_state = SRPC_STATE_EQ_INIT;
1614
1615         rc = stt_startup();
1616
1617 bail:
1618         if (rc != 0)
1619                 srpc_shutdown();
1620         else
1621                 srpc_data.rpc_state = SRPC_STATE_RUNNING;
1622
1623         return rc;
1624 }
1625
1626 void
1627 srpc_shutdown (void)
1628 {
1629         int i;
1630         int rc;
1631         int state;
1632
1633         state = srpc_data.rpc_state;
1634         srpc_data.rpc_state = SRPC_STATE_STOPPING;
1635
1636         switch (state) {
1637         default:
1638                 LBUG ();
1639         case SRPC_STATE_RUNNING:
1640                 spin_lock(&srpc_data.rpc_glock);
1641
1642                 for (i = 0; i <= SRPC_SERVICE_MAX_ID; i++) {
1643                         srpc_service_t *sv = srpc_data.rpc_services[i];
1644
1645                         LASSERTF (sv == NULL,
1646                                   "service not empty: id %d, name %s\n",
1647                                   i, sv->sv_name);
1648                 }
1649
1650                 spin_unlock(&srpc_data.rpc_glock);
1651
1652                 stt_shutdown();
1653
1654         case SRPC_STATE_EQ_INIT:
1655                 rc = LNetClearLazyPortal(SRPC_FRAMEWORK_REQUEST_PORTAL);
1656                 rc = LNetClearLazyPortal(SRPC_REQUEST_PORTAL);
1657                 LASSERT (rc == 0);
1658                 rc = LNetEQFree(srpc_data.rpc_lnet_eq);
1659                 LASSERT (rc == 0); /* the EQ should have no user by now */
1660
1661         case SRPC_STATE_NI_INIT:
1662                 LNetNIFini();
1663         }
1664
1665         return;
1666 }