lustre/obdclass/lu_object.c

   1 /* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
   2  * vim:expandtab:shiftwidth=8:tabstop=8:
   3  *
   4  * Lustre Object.
   5  *
   6  *  Copyright (C) 2006 Cluster File Systems, Inc.
   7  *
   8  *   This file is part of the Lustre file system, http://www.lustre.org
   9  *   Lustre is a trademark of Cluster File Systems, Inc.
  10  *
  11  *   You may have signed or agreed to another license before downloading
  12  *   this software.  If so, you are bound by the terms and conditions
  13  *   of that agreement, and the following does not apply to you.  See the
  14  *   LICENSE file included with this distribution for more information.
  15  *
  16  *   If you did not agree to a different license, then this copy of Lustre
  17  *   is open source software; you can redistribute it and/or modify it
  18  *   under the terms of version 2 of the GNU General Public License as
  19  *   published by the Free Software Foundation.
  20  *
  21  *   In either case, Lustre is distributed in the hope that it will be
  22  *   useful, but WITHOUT ANY WARRANTY; without even the implied warranty
  23  *   of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  24  *   license text for more details.
  25  *
  26  * These are the only exported functions, they provide some generic
  27  * infrastructure for managing object devices
  28  */
  29
  30 #define DEBUG_SUBSYSTEM S_CLASS
  31 #ifndef EXPORT_SYMTAB
  32 # define EXPORT_SYMTAB
  33 #endif
  34
  35 #include <linux/seq_file.h>
  36 #include <linux/module.h>
  37 #include <obd_support.h>
  38 #include <lustre_disk.h>
  39 #include <lu_object.h>
  40 #include <libcfs/list.h>
  41
  42 static void lu_object_free(const struct lu_context *ctx, struct lu_object *o);
  43
  44 /*
  45  * Decrease reference counter on object. If last reference is freed, return
  46  * object to the cache, unless lu_object_is_dying(o) holds. In the latter
  47  * case, free object immediately.
  48  */
  49 void lu_object_put(const struct lu_context *ctxt, struct lu_object *o)
  50 {
  51         struct lu_object_header *top;
  52         struct lu_site          *site;
  53
  54         top = o->lo_header;
  55         site = o->lo_dev->ld_site;
  56         spin_lock(&site->ls_guard);
  57         if (-- top->loh_ref == 0) {
  58                 /*
  59                  * When last reference is released, iterate over object
  60                  * layers, and notify them that object is no longer busy.
  61                  */
  62                 list_for_each_entry(o, &top->loh_layers, lo_linkage) {
  63                         if (o->lo_ops->loo_object_release != NULL)
  64                                 o->lo_ops->loo_object_release(ctxt, o);
  65                 }
  66                 -- site->ls_busy;
  67                 if (lu_object_is_dying(top)) {
  68                         /*
  69                          * If object is dying (will not be cached), removed it
  70                          * from hash table and LRU.
  71                          *
  72                          * This is done with hash table and LRU lists
  73                          * locked. As the only way to acquire first reference
  74                          * to previously unreferenced object is through
  75                          * hash-table lookup (lu_object_find()), or LRU
  76                          * scanning (lu_site_purge()), that are done under
  77                          * hash-table and LRU lock, no race with concurrent
  78                          * object lookup is possible and we can safely destroy
  79                          * object below.
  80                          */
  81                         hlist_del_init(&top->loh_hash);
  82                         list_del_init(&top->loh_lru);
  83                 }
  84         }
  85         spin_unlock(&site->ls_guard);
  86         if (lu_object_is_dying(top))
  87                 /*
  88                  * Object was already removed from hash and lru above, can
  89                  * kill it.
  90                  */
  91                 lu_object_free(ctxt, o);
  92 }
  93 EXPORT_SYMBOL(lu_object_put);
  94
  95 /*
  96  * Allocate new object.
  97  *
  98  * This follows object creation protocol, described in the comment within
  99  * struct lu_device_operations definition.
 100  */
 101 static struct lu_object *lu_object_alloc(const struct lu_context *ctxt,
 102                                          struct lu_site *s,
 103                                          const struct lu_fid *f)
 104 {
 105         struct lu_object *scan;
 106         struct lu_object *top;
 107         int clean;
 108         int result;
 109
 110         /*
 111          * Create top-level object slice. This will also create
 112          * lu_object_header.
 113          */
 114         top = s->ls_top_dev->ld_ops->ldo_object_alloc(ctxt, s->ls_top_dev);
 115         if (IS_ERR(top))
 116                 RETURN(top);
 117         s->ls_total ++;
 118         /*
 119          * This is the only place where object fid is assigned. It's constant
 120          * after this point.
 121          */
 122         top->lo_header->loh_fid = *f;
 123         do {
 124                 /*
 125                  * Call ->loo_object_init() repeatedly, until no more new
 126                  * object slices are created.
 127                  */
 128                 clean = 1;
 129                 list_for_each_entry(scan,
 130                                     &top->lo_header->loh_layers, lo_linkage) {
 131                         if (scan->lo_flags & LU_OBJECT_ALLOCATED)
 132                                 continue;
 133                         clean = 0;
 134                         scan->lo_header = top->lo_header;
 135                         result = scan->lo_ops->loo_object_init(ctxt, scan);
 136                         if (result != 0) {
 137                                 lu_object_free(ctxt, top);
 138                                 RETURN(ERR_PTR(result));
 139                         }
 140                         scan->lo_flags |= LU_OBJECT_ALLOCATED;
 141                 }
 142         } while (!clean);
 143         s->ls_stats.s_created ++;
 144         RETURN(top);
 145 }
 146
 147 /*
 148  * Free object.
 149  */
 150 static void lu_object_free(const struct lu_context *ctx, struct lu_object *o)
 151 {
 152         struct list_head splice;
 153         struct lu_object *scan;
 154
 155         /*
 156          * First call ->loo_object_delete() method to release all resources.
 157          */
 158         list_for_each_entry_reverse(scan,
 159                                     &o->lo_header->loh_layers, lo_linkage) {
 160                 if (scan->lo_ops->loo_object_delete != NULL)
 161                         scan->lo_ops->loo_object_delete(ctx, scan);
 162         }
 163         -- o->lo_dev->ld_site->ls_total;
 164         /*
 165          * Then, splice object layers into stand-alone list, and call
 166          * ->loo_object_free() on all layers to free memory. Splice is
 167          * necessary, because lu_object_header is freed together with the
 168          * top-level slice.
 169          */
 170         INIT_LIST_HEAD(&splice);
 171         list_splice_init(&o->lo_header->loh_layers, &splice);
 172         while (!list_empty(&splice)) {
 173                 o = container_of0(splice.next, struct lu_object, lo_linkage);
 174                 list_del_init(&o->lo_linkage);
 175                 LASSERT(o->lo_ops->loo_object_free != NULL);
 176                 o->lo_ops->loo_object_free(ctx, o);
 177         }
 178 }
 179
 180 /*
 181  * Free @nr objects from the cold end of the site LRU list.
 182  */
 183 void lu_site_purge(const struct lu_context *ctx, struct lu_site *s, int nr)
 184 {
 185         struct list_head         dispose;
 186         struct lu_object_header *h;
 187         struct lu_object_header *temp;
 188
 189         INIT_LIST_HEAD(&dispose);
 190         /*
 191          * Under LRU list lock, scan LRU list and move unreferenced objects to
 192          * the dispose list, removing them from LRU and hash table.
 193          */
 194         spin_lock(&s->ls_guard);
 195         list_for_each_entry_safe(h, temp, &s->ls_lru, loh_lru) {
 196                 if (nr-- == 0)
 197                         break;
 198                 if (h->loh_ref > 0)
 199                         continue;
 200                 hlist_del_init(&h->loh_hash);
 201                 list_move(&h->loh_lru, &dispose);
 202         }
 203         spin_unlock(&s->ls_guard);
 204         /*
 205          * Free everything on the dispose list. This is safe against races due
 206          * to the reasons described in lu_object_put().
 207          */
 208         while (!list_empty(&dispose)) {
 209                 h = container_of0(dispose.next,
 210                                  struct lu_object_header, loh_lru);
 211                 list_del_init(&h->loh_lru);
 212                 lu_object_free(ctx, lu_object_top(h));
 213                 s->ls_stats.s_lru_purged ++;
 214         }
 215 }
 216 EXPORT_SYMBOL(lu_site_purge);
 217
 218 /*
 219  * Print human readable representation of the @o to the @f.
 220  */
 221 int lu_object_print(const struct lu_context *ctx,
 222                     struct seq_file *f, const struct lu_object *o)
 223 {
 224         static char ruler[] = "........................................";
 225         const struct lu_object *scan;
 226         int nob;
 227         int depth;
 228
 229         nob = 0;
 230         scan = o;
 231         list_for_each_entry_continue(scan, &o->lo_linkage, lo_linkage) {
 232                 depth = scan->lo_depth;
 233                 if (depth <= o->lo_depth && scan != o)
 234                         break;
 235                 LASSERT(scan->lo_ops->loo_object_print != NULL);
 236                 /*
 237                  * print `.' @depth times.
 238                  */
 239                 nob += seq_printf(f, "%*.*s", depth, depth, ruler);
 240                 nob += scan->lo_ops->loo_object_print(ctx, f, scan);
 241                 nob += seq_printf(f, "\n");
 242         }
 243         return nob;
 244 }
 245 EXPORT_SYMBOL(lu_object_print);
 246
 247
 248 static struct lu_object *htable_lookup(struct lu_site *s,
 249                                        const struct hlist_head *bucket,
 250                                        const struct lu_fid *f)
 251 {
 252         struct lu_object_header *h;
 253         struct hlist_node *scan;
 254
 255         hlist_for_each_entry(h, scan, bucket, loh_hash) {
 256                 s->ls_stats.s_cache_check ++;
 257                 if (lu_fid_eq(&h->loh_fid, f) && !lu_object_is_dying(h)) {
 258                         /* bump reference count... */
 259                         if (h->loh_ref ++ == 0)
 260                                 ++ s->ls_busy;
 261                         /* and move to the head of the LRU */
 262                         list_move_tail(&h->loh_lru, &s->ls_lru);
 263                         s->ls_stats.s_cache_hit ++;
 264                         return lu_object_top(h);
 265                 }
 266         }
 267         s->ls_stats.s_cache_miss ++;
 268         return NULL;
 269 }
 270
 271 static __u32 fid_hash(const struct lu_fid *f)
 272 {
 273         /* all objects with same id and different versions will belong to same
 274          * collisions list. */
 275         return (fid_seq(f) - 1) * LUSTRE_FID_SEQ_WIDTH + fid_oid(f);
 276 }
 277
 278 /*
 279  * Search cache for an object with the fid @f. If such object is found, return
 280  * it. Otherwise, create new object, insert it into cache and return it. In
 281  * any case, additional reference is acquired on the returned object.
 282  */
 283 struct lu_object *lu_object_find(const struct lu_context *ctxt,
 284                                  struct lu_site *s, const struct lu_fid *f)
 285 {
 286         struct lu_object  *o;
 287         struct lu_object  *shadow;
 288         struct hlist_head *bucket;
 289
 290         /*
 291          * This uses standard index maintenance protocol:
 292          *
 293          *     - search index under lock, and return object if found;
 294          *     - otherwise, unlock index, allocate new object;
 295          *     - lock index and search again;
 296          *     - if nothing is found (usual case), insert newly created
 297          *       object into index;
 298          *     - otherwise (race: other thread inserted object), free
 299          *       object just allocated.
 300          *     - unlock index;
 301          *     - return object.
 302          */
 303
 304         bucket = s->ls_hash + (fid_hash(f) & s->ls_hash_mask);
 305         spin_lock(&s->ls_guard);
 306         o = htable_lookup(s, bucket, f);
 307
 308         spin_unlock(&s->ls_guard);
 309         if (o != NULL)
 310                 return o;
 311         /*
 312          * Allocate new object. This may result in rather complicated
 313          * operations, including fld queries, inode loading, etc.
 314          */
 315         o = lu_object_alloc(ctxt, s, f);
 316         if (IS_ERR(o))
 317                 return o;
 318
 319         LASSERT(lu_fid_eq(lu_object_fid(o), f));
 320
 321         spin_lock(&s->ls_guard);
 322         shadow = htable_lookup(s, bucket, f);
 323         if (shadow == NULL) {
 324                 hlist_add_head(&o->lo_header->loh_hash, bucket);
 325                 list_add_tail(&s->ls_lru, &o->lo_header->loh_lru);
 326                 shadow = o;
 327                 o = NULL;
 328         } else
 329                 s->ls_stats.s_cache_race ++;
 330         spin_unlock(&s->ls_guard);
 331         if (o != NULL)
 332                 lu_object_free(ctxt, o);
 333         return shadow;
 334 }
 335 EXPORT_SYMBOL(lu_object_find);
 336
 337 enum {
 338         LU_SITE_HTABLE_BITS = 8,
 339         LU_SITE_HTABLE_SIZE = (1 << LU_SITE_HTABLE_BITS),
 340         LU_SITE_HTABLE_MASK = LU_SITE_HTABLE_SIZE - 1
 341 };
 342
 343 /*
 344  * Initialize site @s, with @d as the top level device.
 345  */
 346 int lu_site_init(struct lu_site *s, struct lu_device *top)
 347 {
 348         int result;
 349         ENTRY;
 350
 351         memset(s, 0, sizeof *s);
 352         spin_lock_init(&s->ls_guard);
 353         CFS_INIT_LIST_HEAD(&s->ls_lru);
 354         s->ls_top_dev = top;
 355         top->ld_site = s;
 356         lu_device_get(top);
 357         /*
 358          * XXX nikita: fixed size hash-table.
 359          */
 360         s->ls_hash_mask = LU_SITE_HTABLE_MASK;
 361         OBD_ALLOC(s->ls_hash, LU_SITE_HTABLE_SIZE * sizeof s->ls_hash[0]);
 362         if (s->ls_hash != NULL) {
 363                 int i;
 364                 for (i = 0; i < LU_SITE_HTABLE_SIZE; i++)
 365                         INIT_HLIST_HEAD(&s->ls_hash[i]);
 366                 result = 0;
 367         } else {
 368                 result = -ENOMEM;
 369         }
 370
 371         RETURN(result);
 372 }
 373 EXPORT_SYMBOL(lu_site_init);
 374
 375 /*
 376  * Finalize @s and release its resources.
 377  */
 378 void lu_site_fini(struct lu_site *s)
 379 {
 380         LASSERT(list_empty(&s->ls_lru));
 381         LASSERT(s->ls_total == 0);
 382         LASSERT(s->ls_busy == 0);
 383
 384         if (s->ls_hash != NULL) {
 385                 int i;
 386                 for (i = 0; i < LU_SITE_HTABLE_SIZE; i++)
 387                         LASSERT(hlist_empty(&s->ls_hash[i]));
 388                 OBD_FREE(s->ls_hash,
 389                          LU_SITE_HTABLE_SIZE * sizeof s->ls_hash[0]);
 390                 s->ls_hash = NULL;
 391        }
 392        if (s->ls_top_dev != NULL) {
 393                s->ls_top_dev->ld_site = NULL;
 394                lu_device_put(s->ls_top_dev);
 395                s->ls_top_dev = NULL;
 396        }
 397  }
 398 EXPORT_SYMBOL(lu_site_fini);
 399
 400 /*
 401  * Acquire additional reference on device @d
 402  */
 403 void lu_device_get(struct lu_device *d)
 404 {
 405         atomic_inc(&d->ld_ref);
 406 }
 407 EXPORT_SYMBOL(lu_device_get);
 408
 409 /*
 410  * Release reference on device @d.
 411  */
 412 void lu_device_put(struct lu_device *d)
 413 {
 414         atomic_dec(&d->ld_ref);
 415 }
 416 EXPORT_SYMBOL(lu_device_put);
 417
 418 /*
 419  * Initialize device @d of type @t.
 420  */
 421 int lu_device_init(struct lu_device *d, struct lu_device_type *t)
 422 {
 423         memset(d, 0, sizeof *d);
 424         atomic_set(&d->ld_ref, 0);
 425         d->ld_type = t;
 426         return 0;
 427 }
 428 EXPORT_SYMBOL(lu_device_init);
 429
 430 /*
 431  * Finalize device @d.
 432  */
 433 void lu_device_fini(struct lu_device *d)
 434 {
 435         LASSERT(atomic_read(&d->ld_ref) == 0);
 436 }
 437 EXPORT_SYMBOL(lu_device_fini);
 438
 439 /*
 440  * Initialize object @o that is part of compound object @h and was created by
 441  * device @d.
 442  */
 443 int lu_object_init(struct lu_object *o,
 444                    struct lu_object_header *h, struct lu_device *d)
 445 {
 446         memset(o, 0, sizeof *o);
 447         o->lo_header = h;
 448         o->lo_dev    = d;
 449         lu_device_get(d);
 450         CFS_INIT_LIST_HEAD(&o->lo_linkage);
 451         return 0;
 452 }
 453 EXPORT_SYMBOL(lu_object_init);
 454
 455 /*
 456  * Finalize object and release its resources.
 457  */
 458 void lu_object_fini(struct lu_object *o)
 459 {
 460         LASSERT(list_empty(&o->lo_linkage));
 461
 462         if (o->lo_dev != NULL) {
 463                 lu_device_put(o->lo_dev);
 464                 o->lo_dev = NULL;
 465         }
 466 }
 467 EXPORT_SYMBOL(lu_object_fini);
 468
 469 /*
 470  * Add object @o as first layer of compound object @h
 471  *
 472  * This is typically called by the ->ldo_object_alloc() method of top-level
 473  * device.
 474  */
 475 void lu_object_add_top(struct lu_object_header *h, struct lu_object *o)
 476 {
 477         list_move(&o->lo_linkage, &h->loh_layers);
 478 }
 479 EXPORT_SYMBOL(lu_object_add_top);
 480
 481 /*
 482  * Add object @o as a layer of compound object, going after @before.1
 483  *
 484  * This is typically called by the ->ldo_object_alloc() method of
 485  * @before->lo_dev.
 486  */
 487 void lu_object_add(struct lu_object *before, struct lu_object *o)
 488 {
 489         list_move(&o->lo_linkage, &before->lo_linkage);
 490 }
 491 EXPORT_SYMBOL(lu_object_add);
 492
 493 /*
 494  * Initialize compound object.
 495  */
 496 int lu_object_header_init(struct lu_object_header *h)
 497 {
 498         memset(h, 0, sizeof *h);
 499         h->loh_ref = 1;
 500         INIT_HLIST_NODE(&h->loh_hash);
 501         CFS_INIT_LIST_HEAD(&h->loh_lru);
 502         CFS_INIT_LIST_HEAD(&h->loh_layers);
 503         return 0;
 504 }
 505 EXPORT_SYMBOL(lu_object_header_init);
 506
 507 /*
 508  * Finalize compound object.
 509  */
 510 void lu_object_header_fini(struct lu_object_header *h)
 511 {
 512         LASSERT(list_empty(&h->loh_layers));
 513         LASSERT(list_empty(&h->loh_lru));
 514         LASSERT(hlist_unhashed(&h->loh_hash));
 515 }
 516 EXPORT_SYMBOL(lu_object_header_fini);
 517
 518 /*
 519  * Given a compound object, find its slice, corresponding to the device type
 520  * @dtype.
 521  */
 522 struct lu_object *lu_object_locate(struct lu_object_header *h,
 523                                    struct lu_device_type *dtype)
 524 {
 525         struct lu_object *o;
 526
 527         list_for_each_entry(o, &h->loh_layers, lo_linkage) {
 528                 if (o->lo_dev->ld_type == dtype)
 529                         return o;
 530         }
 531         return NULL;
 532 }
 533 EXPORT_SYMBOL(lu_object_locate);
 534
 535 enum {
 536         /*
 537          * Maximal number of tld slots.
 538          */
 539         LU_CONTEXT_KEY_NR = 16
 540 };
 541
 542 static struct lu_context_key *lu_keys[LU_CONTEXT_KEY_NR] = { NULL, };
 543
 544 static spinlock_t lu_keys_guard = SPIN_LOCK_UNLOCKED;
 545
 546 /*
 547  * Register new key.
 548  */
 549 int lu_context_key_register(struct lu_context_key *key)
 550 {
 551         int result;
 552         int i;
 553
 554         result = -ENFILE;
 555         spin_lock(&lu_keys_guard);
 556         for (i = 0; i < ARRAY_SIZE(lu_keys); ++i) {
 557                 if (lu_keys[i] == NULL) {
 558                         key->lct_index = i;
 559                         key->lct_used = 1;
 560                         lu_keys[i] = key;
 561                         result = 0;
 562                         break;
 563                 }
 564         }
 565         spin_unlock(&lu_keys_guard);
 566         return result;
 567 }
 568 EXPORT_SYMBOL(lu_context_key_register);
 569
 570 /*
 571  * Deregister key.
 572  */
 573 void lu_context_key_degister(struct lu_context_key *key)
 574 {
 575         LASSERT(key->lct_used >= 1);
 576         LASSERT(0 <= key->lct_index && key->lct_index < ARRAY_SIZE(lu_keys));
 577
 578         if (key->lct_used > 1)
 579                 CERROR("key has instances.\n");
 580         spin_lock(&lu_keys_guard);
 581         lu_keys[key->lct_index] = NULL;
 582         spin_unlock(&lu_keys_guard);
 583 }
 584 EXPORT_SYMBOL(lu_context_key_degister);
 585
 586 /*
 587  * Return value associated with key @key in context @ctx.
 588  */
 589 void *lu_context_key_get(const struct lu_context *ctx,
 590                          struct lu_context_key *key)
 591 {
 592         LASSERT(0 <= key->lct_index && key->lct_index < ARRAY_SIZE(lu_keys));
 593         return ctx->lc_value[key->lct_index];
 594 }
 595 EXPORT_SYMBOL(lu_context_key_get);
 596
 597 static void keys_fini(struct lu_context *ctx)
 598 {
 599         int i;
 600
 601         if (ctx->lc_value != NULL) {
 602                 for (i = 0; i < ARRAY_SIZE(lu_keys); ++i) {
 603                         if (ctx->lc_value[i] != NULL) {
 604                                 struct lu_context_key *key;
 605
 606                                 key = lu_keys[i];
 607                                 LASSERT(key != NULL);
 608                                 LASSERT(key->lct_fini != NULL);
 609                                 LASSERT(key->lct_used > 1);
 610
 611                                 key->lct_fini(ctx, ctx->lc_value[i]);
 612                                 key->lct_used--;
 613                                 ctx->lc_value[i] = NULL;
 614                         }
 615                 }
 616                 OBD_FREE(ctx->lc_value,
 617                          ARRAY_SIZE(lu_keys) * sizeof ctx->lc_value[0]);
 618                 ctx->lc_value = NULL;
 619         }
 620 }
 621
 622 static int keys_init(struct lu_context *ctx)
 623 {
 624         int i;
 625         int result;
 626
 627         OBD_ALLOC(ctx->lc_value, ARRAY_SIZE(lu_keys) * sizeof ctx->lc_value[0]);
 628         if (ctx->lc_value != NULL) {
 629                 for (i = 0; i < ARRAY_SIZE(lu_keys); ++i) {
 630                         struct lu_context_key *key;
 631
 632                         key = lu_keys[i];
 633                         if (key != NULL) {
 634                                 void *value;
 635
 636                                 LASSERT(key->lct_init != NULL);
 637                                 LASSERT(key->lct_index == i);
 638
 639                                 value = key->lct_init(ctx);
 640                                 if (IS_ERR(value)) {
 641                                         keys_fini(ctx);
 642                                         return PTR_ERR(value);
 643                                 }
 644                                 key->lct_used++;
 645                                 ctx->lc_value[i] = value;
 646                         }
 647                 }
 648                 result = 0;
 649         } else
 650                 result = -ENOMEM;
 651         return result;
 652 }
 653
 654 /*
 655  * Initialize context data-structure. Create values for all keys.
 656  */
 657 int lu_context_init(struct lu_context *ctx)
 658 {
 659         memset(ctx, 0, sizeof *ctx);
 660         keys_init(ctx);
 661         return 0;
 662 }
 663 EXPORT_SYMBOL(lu_context_init);
 664
 665 /*
 666  * Finalize context data-structure. Destroy key values.
 667  */
 668 void lu_context_fini(struct lu_context *ctx)
 669 {
 670         keys_fini(ctx);
 671 }
 672 EXPORT_SYMBOL(lu_context_fini);
 673
 674 /*
 675  * Called before entering context.
 676  */
 677 void lu_context_enter(struct lu_context *ctx)
 678 {
 679 }
 680 EXPORT_SYMBOL(lu_context_enter);
 681
 682 /*
 683  * Called after exiting from @ctx
 684  */
 685 void lu_context_exit(struct lu_context *ctx)
 686 {
 687 }
 688 EXPORT_SYMBOL(lu_context_exit);