lustre/obdclass/lu_object.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) 2011, 2012, Whamcloud, Inc.
  31  */
  32 /*
  33  * This file is part of Lustre, http://www.lustre.org/
  34  * Lustre is a trademark of Sun Microsystems, Inc.
  35  *
  36  * lustre/obdclass/lu_object.c
  37  *
  38  * Lustre Object.
  39  * These are the only exported functions, they provide some generic
  40  * infrastructure for managing object devices
  41  *
  42  *   Author: Nikita Danilov <nikita.danilov@sun.com>
  43  */
  44
  45 #define DEBUG_SUBSYSTEM S_CLASS
  46
  47 #include <libcfs/libcfs.h>
  48
  49 #ifdef __KERNEL__
  50 # include <linux/module.h>
  51 #endif
  52
  53 /* hash_long() */
  54 #include <libcfs/libcfs_hash.h>
  55 #include <obd_class.h>
  56 #include <obd_support.h>
  57 #include <lustre_disk.h>
  58 #include <lustre_fid.h>
  59 #include <lu_object.h>
  60 #include <libcfs/list.h>
  61 /* lu_time_global_{init,fini}() */
  62 #include <lu_time.h>
  63
  64 static void lu_object_free(const struct lu_env *env, struct lu_object *o);
  65
  66 /**
  67  * Decrease reference counter on object. If last reference is freed, return
  68  * object to the cache, unless lu_object_is_dying(o) holds. In the latter
  69  * case, free object immediately.
  70  */
  71 void lu_object_put(const struct lu_env *env, struct lu_object *o)
  72 {
  73         struct lu_site_bkt_data *bkt;
  74         struct lu_object_header *top;
  75         struct lu_site          *site;
  76         struct lu_object        *orig;
  77         cfs_hash_bd_t            bd;
  78
  79         top  = o->lo_header;
  80         site = o->lo_dev->ld_site;
  81         orig = o;
  82
  83         cfs_hash_bd_get(site->ls_obj_hash, &top->loh_fid, &bd);
  84         bkt = cfs_hash_bd_extra_get(site->ls_obj_hash, &bd);
  85
  86         if (!cfs_hash_bd_dec_and_lock(site->ls_obj_hash, &bd, &top->loh_ref)) {
  87                 if (lu_object_is_dying(top)) {
  88
  89                         /*
  90                          * somebody may be waiting for this, currently only
  91                          * used for cl_object, see cl_object_put_last().
  92                          */
  93                         cfs_waitq_broadcast(&bkt->lsb_marche_funebre);
  94                 }
  95                 return;
  96         }
  97
  98         LASSERT(bkt->lsb_busy > 0);
  99         bkt->lsb_busy--;
 100         /*
 101          * When last reference is released, iterate over object
 102          * layers, and notify them that object is no longer busy.
 103          */
 104         cfs_list_for_each_entry_reverse(o, &top->loh_layers, lo_linkage) {
 105                 if (o->lo_ops->loo_object_release != NULL)
 106                         o->lo_ops->loo_object_release(env, o);
 107         }
 108
 109         if (!lu_object_is_dying(top)) {
 110                 LASSERT(cfs_list_empty(&top->loh_lru));
 111                 cfs_list_add_tail(&top->loh_lru, &bkt->lsb_lru);
 112                 cfs_hash_bd_unlock(site->ls_obj_hash, &bd, 1);
 113                 return;
 114         }
 115
 116         /*
 117          * If object is dying (will not be cached), removed it
 118          * from hash table and LRU.
 119          *
 120          * This is done with hash table and LRU lists locked. As the only
 121          * way to acquire first reference to previously unreferenced
 122          * object is through hash-table lookup (lu_object_find()),
 123          * or LRU scanning (lu_site_purge()), that are done under hash-table
 124          * and LRU lock, no race with concurrent object lookup is possible
 125          * and we can safely destroy object below.
 126          */
 127         cfs_hash_bd_del_locked(site->ls_obj_hash, &bd, &top->loh_hash);
 128         cfs_hash_bd_unlock(site->ls_obj_hash, &bd, 1);
 129         /*
 130          * Object was already removed from hash and lru above, can
 131          * kill it.
 132          */
 133         lu_object_free(env, orig);
 134 }
 135 EXPORT_SYMBOL(lu_object_put);
 136
 137 /**
 138  * Allocate new object.
 139  *
 140  * This follows object creation protocol, described in the comment within
 141  * struct lu_device_operations definition.
 142  */
 143 static struct lu_object *lu_object_alloc(const struct lu_env *env,
 144                                          struct lu_device *dev,
 145                                          const struct lu_fid *f,
 146                                          const struct lu_object_conf *conf)
 147 {
 148         struct lu_object *scan;
 149         struct lu_object *top;
 150         cfs_list_t *layers;
 151         int clean;
 152         int result;
 153         ENTRY;
 154
 155         /*
 156          * Create top-level object slice. This will also create
 157          * lu_object_header.
 158          */
 159         top = dev->ld_ops->ldo_object_alloc(env, NULL, dev);
 160         if (top == NULL)
 161                 RETURN(ERR_PTR(-ENOMEM));
 162         /*
 163          * This is the only place where object fid is assigned. It's constant
 164          * after this point.
 165          */
 166         LASSERT(fid_is_igif(f) || fid_ver(f) == 0);
 167         top->lo_header->loh_fid = *f;
 168         layers = &top->lo_header->loh_layers;
 169         do {
 170                 /*
 171                  * Call ->loo_object_init() repeatedly, until no more new
 172                  * object slices are created.
 173                  */
 174                 clean = 1;
 175                 cfs_list_for_each_entry(scan, layers, lo_linkage) {
 176                         if (scan->lo_flags & LU_OBJECT_ALLOCATED)
 177                                 continue;
 178                         clean = 0;
 179                         scan->lo_header = top->lo_header;
 180                         result = scan->lo_ops->loo_object_init(env, scan, conf);
 181                         if (result != 0) {
 182                                 lu_object_free(env, top);
 183                                 RETURN(ERR_PTR(result));
 184                         }
 185                         scan->lo_flags |= LU_OBJECT_ALLOCATED;
 186                 }
 187         } while (!clean);
 188
 189         cfs_list_for_each_entry_reverse(scan, layers, lo_linkage) {
 190                 if (scan->lo_ops->loo_object_start != NULL) {
 191                         result = scan->lo_ops->loo_object_start(env, scan);
 192                         if (result != 0) {
 193                                 lu_object_free(env, top);
 194                                 RETURN(ERR_PTR(result));
 195                         }
 196                 }
 197         }
 198
 199         lprocfs_counter_incr(dev->ld_site->ls_stats, LU_SS_CREATED);
 200         RETURN(top);
 201 }
 202
 203 /**
 204  * Free an object.
 205  */
 206 static void lu_object_free(const struct lu_env *env, struct lu_object *o)
 207 {
 208         struct lu_site_bkt_data *bkt;
 209         struct lu_site          *site;
 210         struct lu_object        *scan;
 211         cfs_list_t              *layers;
 212         cfs_list_t               splice;
 213
 214         site   = o->lo_dev->ld_site;
 215         layers = &o->lo_header->loh_layers;
 216         bkt    = lu_site_bkt_from_fid(site, &o->lo_header->loh_fid);
 217         /*
 218          * First call ->loo_object_delete() method to release all resources.
 219          */
 220         cfs_list_for_each_entry_reverse(scan, layers, lo_linkage) {
 221                 if (scan->lo_ops->loo_object_delete != NULL)
 222                         scan->lo_ops->loo_object_delete(env, scan);
 223         }
 224
 225         /*
 226          * Then, splice object layers into stand-alone list, and call
 227          * ->loo_object_free() on all layers to free memory. Splice is
 228          * necessary, because lu_object_header is freed together with the
 229          * top-level slice.
 230          */
 231         CFS_INIT_LIST_HEAD(&splice);
 232         cfs_list_splice_init(layers, &splice);
 233         while (!cfs_list_empty(&splice)) {
 234                 /*
 235                  * Free layers in bottom-to-top order, so that object header
 236                  * lives as long as possible and ->loo_object_free() methods
 237                  * can look at its contents.
 238                  */
 239                 o = container_of0(splice.prev, struct lu_object, lo_linkage);
 240                 cfs_list_del_init(&o->lo_linkage);
 241                 LASSERT(o->lo_ops->loo_object_free != NULL);
 242                 o->lo_ops->loo_object_free(env, o);
 243         }
 244
 245         if (cfs_waitq_active(&bkt->lsb_marche_funebre))
 246                 cfs_waitq_broadcast(&bkt->lsb_marche_funebre);
 247 }
 248
 249 /**
 250  * Free \a nr objects from the cold end of the site LRU list.
 251  */
 252 int lu_site_purge(const struct lu_env *env, struct lu_site *s, int nr)
 253 {
 254         struct lu_object_header *h;
 255         struct lu_object_header *temp;
 256         struct lu_site_bkt_data *bkt;
 257         cfs_hash_bd_t            bd;
 258         cfs_hash_bd_t            bd2;
 259         cfs_list_t               dispose;
 260         int                      did_sth;
 261         int                      start;
 262         int                      count;
 263         int                      bnr;
 264         int                      i;
 265
 266         CFS_INIT_LIST_HEAD(&dispose);
 267         /*
 268          * Under LRU list lock, scan LRU list and move unreferenced objects to
 269          * the dispose list, removing them from LRU and hash table.
 270          */
 271         start = s->ls_purge_start;
 272         bnr = (nr == ~0) ? -1 : nr / CFS_HASH_NBKT(s->ls_obj_hash) + 1;
 273  again:
 274         did_sth = 0;
 275         cfs_hash_for_each_bucket(s->ls_obj_hash, &bd, i) {
 276                 if (i < start)
 277                         continue;
 278                 count = bnr;
 279                 cfs_hash_bd_lock(s->ls_obj_hash, &bd, 1);
 280                 bkt = cfs_hash_bd_extra_get(s->ls_obj_hash, &bd);
 281
 282                 cfs_list_for_each_entry_safe(h, temp, &bkt->lsb_lru, loh_lru) {
 283                         LASSERT(cfs_atomic_read(&h->loh_ref) == 0);
 284
 285                         cfs_hash_bd_get(s->ls_obj_hash, &h->loh_fid, &bd2);
 286                         LASSERT(bd.bd_bucket == bd2.bd_bucket);
 287
 288                         cfs_hash_bd_del_locked(s->ls_obj_hash,
 289                                                &bd2, &h->loh_hash);
 290                         cfs_list_move(&h->loh_lru, &dispose);
 291                         if (did_sth == 0)
 292                                 did_sth = 1;
 293
 294                         if (nr != ~0 && --nr == 0)
 295                                 break;
 296
 297                         if (count > 0 && --count == 0)
 298                                 break;
 299
 300                 }
 301                 cfs_hash_bd_unlock(s->ls_obj_hash, &bd, 1);
 302                 cfs_cond_resched();
 303                 /*
 304                  * Free everything on the dispose list. This is safe against
 305                  * races due to the reasons described in lu_object_put().
 306                  */
 307                 while (!cfs_list_empty(&dispose)) {
 308                         h = container_of0(dispose.next,
 309                                           struct lu_object_header, loh_lru);
 310                         cfs_list_del_init(&h->loh_lru);
 311                         lu_object_free(env, lu_object_top(h));
 312                         lprocfs_counter_incr(s->ls_stats, LU_SS_LRU_PURGED);
 313                 }
 314
 315                 if (nr == 0)
 316                         break;
 317         }
 318
 319         if (nr != 0 && did_sth && start != 0) {
 320                 start = 0; /* restart from the first bucket */
 321                 goto again;
 322         }
 323         /* race on s->ls_purge_start, but nobody cares */
 324         s->ls_purge_start = i % CFS_HASH_NBKT(s->ls_obj_hash);
 325
 326         return nr;
 327 }
 328 EXPORT_SYMBOL(lu_site_purge);
 329
 330 /*
 331  * Object printing.
 332  *
 333  * Code below has to jump through certain loops to output object description
 334  * into libcfs_debug_msg-based log. The problem is that lu_object_print()
 335  * composes object description from strings that are parts of _lines_ of
 336  * output (i.e., strings that are not terminated by newline). This doesn't fit
 337  * very well into libcfs_debug_msg() interface that assumes that each message
 338  * supplied to it is a self-contained output line.
 339  *
 340  * To work around this, strings are collected in a temporary buffer
 341  * (implemented as a value of lu_cdebug_key key), until terminating newline
 342  * character is detected.
 343  *
 344  */
 345
 346 enum {
 347         /**
 348          * Maximal line size.
 349          *
 350          * XXX overflow is not handled correctly.
 351          */
 352         LU_CDEBUG_LINE = 512
 353 };
 354
 355 struct lu_cdebug_data {
 356         /**
 357          * Temporary buffer.
 358          */
 359         char lck_area[LU_CDEBUG_LINE];
 360 };
 361
 362 /* context key constructor/destructor: lu_global_key_init, lu_global_key_fini */
 363 LU_KEY_INIT_FINI(lu_global, struct lu_cdebug_data);
 364
 365 /**
 366  * Key, holding temporary buffer. This key is registered very early by
 367  * lu_global_init().
 368  */
 369 struct lu_context_key lu_global_key = {
 370         .lct_tags = LCT_MD_THREAD | LCT_DT_THREAD |
 371                     LCT_MG_THREAD | LCT_CL_THREAD,
 372         .lct_init = lu_global_key_init,
 373         .lct_fini = lu_global_key_fini
 374 };
 375
 376 /**
 377  * Printer function emitting messages through libcfs_debug_msg().
 378  */
 379 int lu_cdebug_printer(const struct lu_env *env,
 380                       void *cookie, const char *format, ...)
 381 {
 382         struct libcfs_debug_msg_data *msgdata = cookie;
 383         struct lu_cdebug_data        *key;
 384         int used;
 385         int complete;
 386         va_list args;
 387
 388         va_start(args, format);
 389
 390         key = lu_context_key_get(&env->le_ctx, &lu_global_key);
 391         LASSERT(key != NULL);
 392
 393         used = strlen(key->lck_area);
 394         complete = format[strlen(format) - 1] == '\n';
 395         /*
 396          * Append new chunk to the buffer.
 397          */
 398         vsnprintf(key->lck_area + used,
 399                   ARRAY_SIZE(key->lck_area) - used, format, args);
 400         if (complete) {
 401                 if (cfs_cdebug_show(msgdata->msg_mask, msgdata->msg_subsys))
 402                         libcfs_debug_msg(msgdata, "%s", key->lck_area);
 403                 key->lck_area[0] = 0;
 404         }
 405         va_end(args);
 406         return 0;
 407 }
 408 EXPORT_SYMBOL(lu_cdebug_printer);
 409
 410 /**
 411  * Print object header.
 412  */
 413 void lu_object_header_print(const struct lu_env *env, void *cookie,
 414                             lu_printer_t printer,
 415                             const struct lu_object_header *hdr)
 416 {
 417         (*printer)(env, cookie, "header@%p[%#lx, %d, "DFID"%s%s%s]",
 418                    hdr, hdr->loh_flags, cfs_atomic_read(&hdr->loh_ref),
 419                    PFID(&hdr->loh_fid),
 420                    cfs_hlist_unhashed(&hdr->loh_hash) ? "" : " hash",
 421                    cfs_list_empty((cfs_list_t *)&hdr->loh_lru) ? \
 422                    "" : " lru",
 423                    hdr->loh_attr & LOHA_EXISTS ? " exist":"");
 424 }
 425 EXPORT_SYMBOL(lu_object_header_print);
 426
 427 /**
 428  * Print human readable representation of the \a o to the \a printer.
 429  */
 430 void lu_object_print(const struct lu_env *env, void *cookie,
 431                      lu_printer_t printer, const struct lu_object *o)
 432 {
 433         static const char ruler[] = "........................................";
 434         struct lu_object_header *top;
 435         int depth;
 436
 437         top = o->lo_header;
 438         lu_object_header_print(env, cookie, printer, top);
 439         (*printer)(env, cookie, "{ \n");
 440         cfs_list_for_each_entry(o, &top->loh_layers, lo_linkage) {
 441                 depth = o->lo_depth + 4;
 442
 443                 /*
 444                  * print `.' \a depth times followed by type name and address
 445                  */
 446                 (*printer)(env, cookie, "%*.*s%s@%p", depth, depth, ruler,
 447                            o->lo_dev->ld_type->ldt_name, o);
 448                 if (o->lo_ops->loo_object_print != NULL)
 449                         o->lo_ops->loo_object_print(env, cookie, printer, o);
 450                 (*printer)(env, cookie, "\n");
 451         }
 452         (*printer)(env, cookie, "} header@%p\n", top);
 453 }
 454 EXPORT_SYMBOL(lu_object_print);
 455
 456 /**
 457  * Check object consistency.
 458  */
 459 int lu_object_invariant(const struct lu_object *o)
 460 {
 461         struct lu_object_header *top;
 462
 463         top = o->lo_header;
 464         cfs_list_for_each_entry(o, &top->loh_layers, lo_linkage) {
 465                 if (o->lo_ops->loo_object_invariant != NULL &&
 466                     !o->lo_ops->loo_object_invariant(o))
 467                         return 0;
 468         }
 469         return 1;
 470 }
 471 EXPORT_SYMBOL(lu_object_invariant);
 472
 473 static struct lu_object *htable_lookup(struct lu_site *s,
 474                                        cfs_hash_bd_t *bd,
 475                                        const struct lu_fid *f,
 476                                        cfs_waitlink_t *waiter,
 477                                        __u64 *version)
 478 {
 479         struct lu_site_bkt_data *bkt;
 480         struct lu_object_header *h;
 481         cfs_hlist_node_t        *hnode;
 482         __u64  ver = cfs_hash_bd_version_get(bd);
 483
 484         if (*version == ver)
 485                 return NULL;
 486
 487         *version = ver;
 488         bkt = cfs_hash_bd_extra_get(s->ls_obj_hash, bd);
 489         /* cfs_hash_bd_lookup_intent is a somehow "internal" function
 490          * of cfs_hash, but we don't want refcount on object right now */
 491         hnode = cfs_hash_bd_lookup_locked(s->ls_obj_hash, bd, (void *)f);
 492         if (hnode == NULL) {
 493                 lprocfs_counter_incr(s->ls_stats, LU_SS_CACHE_MISS);
 494                 return NULL;
 495         }
 496
 497         h = container_of0(hnode, struct lu_object_header, loh_hash);
 498         if (likely(!lu_object_is_dying(h))) {
 499                 lprocfs_counter_incr(s->ls_stats, LU_SS_CACHE_HIT);
 500                 cfs_list_del_init(&h->loh_lru);
 501                 return lu_object_top(h);
 502         }
 503
 504         /*
 505          * Lookup found an object being destroyed this object cannot be
 506          * returned (to assure that references to dying objects are eventually
 507          * drained), and moreover, lookup has to wait until object is freed.
 508          */
 509         cfs_atomic_dec(&h->loh_ref);
 510
 511         cfs_waitlink_init(waiter);
 512         cfs_waitq_add(&bkt->lsb_marche_funebre, waiter);
 513         cfs_set_current_state(CFS_TASK_UNINT);
 514         lprocfs_counter_incr(s->ls_stats, LU_SS_CACHE_DEATH_RACE);
 515         return ERR_PTR(-EAGAIN);
 516 }
 517
 518 /**
 519  * Search cache for an object with the fid \a f. If such object is found,
 520  * return it. Otherwise, create new object, insert it into cache and return
 521  * it. In any case, additional reference is acquired on the returned object.
 522  */
 523 struct lu_object *lu_object_find(const struct lu_env *env,
 524                                  struct lu_device *dev, const struct lu_fid *f,
 525                                  const struct lu_object_conf *conf)
 526 {
 527         return lu_object_find_at(env, dev->ld_site->ls_top_dev, f, conf);
 528 }
 529 EXPORT_SYMBOL(lu_object_find);
 530
 531 static struct lu_object *lu_object_new(const struct lu_env *env,
 532                                        struct lu_device *dev,
 533                                        const struct lu_fid *f,
 534                                        const struct lu_object_conf *conf)
 535 {
 536         struct lu_object        *o;
 537         cfs_hash_t              *hs;
 538         cfs_hash_bd_t            bd;
 539         struct lu_site_bkt_data *bkt;
 540
 541         o = lu_object_alloc(env, dev, f, conf);
 542         if (unlikely(IS_ERR(o)))
 543                 return o;
 544
 545         hs = dev->ld_site->ls_obj_hash;
 546         cfs_hash_bd_get_and_lock(hs, (void *)f, &bd, 1);
 547         bkt = cfs_hash_bd_extra_get(hs, &bd);
 548         cfs_hash_bd_add_locked(hs, &bd, &o->lo_header->loh_hash);
 549         bkt->lsb_busy++;
 550         cfs_hash_bd_unlock(hs, &bd, 1);
 551         return o;
 552 }
 553
 554 /**
 555  * Core logic of lu_object_find*() functions.
 556  */
 557 static struct lu_object *lu_object_find_try(const struct lu_env *env,
 558                                             struct lu_device *dev,
 559                                             const struct lu_fid *f,
 560                                             const struct lu_object_conf *conf,
 561                                             cfs_waitlink_t *waiter)
 562 {
 563         struct lu_object      *o;
 564         struct lu_object      *shadow;
 565         struct lu_site        *s;
 566         cfs_hash_t            *hs;
 567         cfs_hash_bd_t          bd;
 568         __u64                  version = 0;
 569
 570         /*
 571          * This uses standard index maintenance protocol:
 572          *
 573          *     - search index under lock, and return object if found;
 574          *     - otherwise, unlock index, allocate new object;
 575          *     - lock index and search again;
 576          *     - if nothing is found (usual case), insert newly created
 577          *       object into index;
 578          *     - otherwise (race: other thread inserted object), free
 579          *       object just allocated.
 580          *     - unlock index;
 581          *     - return object.
 582          *
 583          * For "LOC_F_NEW" case, we are sure the object is new established.
 584          * It is unnecessary to perform lookup-alloc-lookup-insert, instead,
 585          * just alloc and insert directly.
 586          *
 587          * If dying object is found during index search, add @waiter to the
 588          * site wait-queue and return ERR_PTR(-EAGAIN).
 589          */
 590         if (conf != NULL && conf->loc_flags & LOC_F_NEW)
 591                 return lu_object_new(env, dev, f, conf);
 592
 593         s  = dev->ld_site;
 594         hs = s->ls_obj_hash;
 595         cfs_hash_bd_get_and_lock(hs, (void *)f, &bd, 1);
 596         o = htable_lookup(s, &bd, f, waiter, &version);
 597         cfs_hash_bd_unlock(hs, &bd, 1);
 598         if (o != NULL)
 599                 return o;
 600
 601         /*
 602          * Allocate new object. This may result in rather complicated
 603          * operations, including fld queries, inode loading, etc.
 604          */
 605         o = lu_object_alloc(env, dev, f, conf);
 606         if (unlikely(IS_ERR(o)))
 607                 return o;
 608
 609         LASSERT(lu_fid_eq(lu_object_fid(o), f));
 610
 611         cfs_hash_bd_lock(hs, &bd, 1);
 612
 613         shadow = htable_lookup(s, &bd, f, waiter, &version);
 614         if (likely(shadow == NULL)) {
 615                 struct lu_site_bkt_data *bkt;
 616
 617                 bkt = cfs_hash_bd_extra_get(hs, &bd);
 618                 cfs_hash_bd_add_locked(hs, &bd, &o->lo_header->loh_hash);
 619                 bkt->lsb_busy++;
 620                 cfs_hash_bd_unlock(hs, &bd, 1);
 621                 return o;
 622         }
 623
 624         lprocfs_counter_incr(s->ls_stats, LU_SS_CACHE_RACE);
 625         cfs_hash_bd_unlock(hs, &bd, 1);
 626         lu_object_free(env, o);
 627         return shadow;
 628 }
 629
 630 /**
 631  * Much like lu_object_find(), but top level device of object is specifically
 632  * \a dev rather than top level device of the site. This interface allows
 633  * objects of different "stacking" to be created within the same site.
 634  */
 635 struct lu_object *lu_object_find_at(const struct lu_env *env,
 636                                     struct lu_device *dev,
 637                                     const struct lu_fid *f,
 638                                     const struct lu_object_conf *conf)
 639 {
 640         struct lu_site_bkt_data *bkt;
 641         struct lu_object        *obj;
 642         cfs_waitlink_t           wait;
 643
 644         while (1) {
 645                 obj = lu_object_find_try(env, dev, f, conf, &wait);
 646                 if (obj != ERR_PTR(-EAGAIN))
 647                         return obj;
 648                 /*
 649                  * lu_object_find_try() already added waiter into the
 650                  * wait queue.
 651                  */
 652                 cfs_waitq_wait(&wait, CFS_TASK_UNINT);
 653                 bkt = lu_site_bkt_from_fid(dev->ld_site, (void *)f);
 654                 cfs_waitq_del(&bkt->lsb_marche_funebre, &wait);
 655         }
 656 }
 657 EXPORT_SYMBOL(lu_object_find_at);
 658
 659 /**
 660  * Find object with given fid, and return its slice belonging to given device.
 661  */
 662 struct lu_object *lu_object_find_slice(const struct lu_env *env,
 663                                        struct lu_device *dev,
 664                                        const struct lu_fid *f,
 665                                        const struct lu_object_conf *conf)
 666 {
 667         struct lu_object *top;
 668         struct lu_object *obj;
 669
 670         top = lu_object_find(env, dev, f, conf);
 671         if (!IS_ERR(top)) {
 672                 obj = lu_object_locate(top->lo_header, dev->ld_type);
 673                 if (obj == NULL)
 674                         lu_object_put(env, top);
 675         } else
 676                 obj = top;
 677         return obj;
 678 }
 679 EXPORT_SYMBOL(lu_object_find_slice);
 680
 681 /**
 682  * Global list of all device types.
 683  */
 684 static CFS_LIST_HEAD(lu_device_types);
 685
 686 int lu_device_type_init(struct lu_device_type *ldt)
 687 {
 688         int result;
 689
 690         CFS_INIT_LIST_HEAD(&ldt->ldt_linkage);
 691         result = ldt->ldt_ops->ldto_init(ldt);
 692         if (result == 0)
 693                 cfs_list_add(&ldt->ldt_linkage, &lu_device_types);
 694         return result;
 695 }
 696 EXPORT_SYMBOL(lu_device_type_init);
 697
 698 void lu_device_type_fini(struct lu_device_type *ldt)
 699 {
 700         cfs_list_del_init(&ldt->ldt_linkage);
 701         ldt->ldt_ops->ldto_fini(ldt);
 702 }
 703 EXPORT_SYMBOL(lu_device_type_fini);
 704
 705 void lu_types_stop(void)
 706 {
 707         struct lu_device_type *ldt;
 708
 709         cfs_list_for_each_entry(ldt, &lu_device_types, ldt_linkage) {
 710                 if (ldt->ldt_device_nr == 0)
 711                         ldt->ldt_ops->ldto_stop(ldt);
 712         }
 713 }
 714 EXPORT_SYMBOL(lu_types_stop);
 715
 716 /**
 717  * Global list of all sites on this node
 718  */
 719 static CFS_LIST_HEAD(lu_sites);
 720 static CFS_DEFINE_MUTEX(lu_sites_guard);
 721
 722 /**
 723  * Global environment used by site shrinker.
 724  */
 725 static struct lu_env lu_shrink_env;
 726
 727 struct lu_site_print_arg {
 728         struct lu_env   *lsp_env;
 729         void            *lsp_cookie;
 730         lu_printer_t     lsp_printer;
 731 };
 732
 733 static int
 734 lu_site_obj_print(cfs_hash_t *hs, cfs_hash_bd_t *bd,
 735                   cfs_hlist_node_t *hnode, void *data)
 736 {
 737         struct lu_site_print_arg *arg = (struct lu_site_print_arg *)data;
 738         struct lu_object_header  *h;
 739
 740         h = cfs_hlist_entry(hnode, struct lu_object_header, loh_hash);
 741         if (!cfs_list_empty(&h->loh_layers)) {
 742                 const struct lu_object *o;
 743
 744                 o = lu_object_top(h);
 745                 lu_object_print(arg->lsp_env, arg->lsp_cookie,
 746                                 arg->lsp_printer, o);
 747         } else {
 748                 lu_object_header_print(arg->lsp_env, arg->lsp_cookie,
 749                                        arg->lsp_printer, h);
 750         }
 751         return 0;
 752 }
 753
 754 /**
 755  * Print all objects in \a s.
 756  */
 757 void lu_site_print(const struct lu_env *env, struct lu_site *s, void *cookie,
 758                    lu_printer_t printer)
 759 {
 760         struct lu_site_print_arg arg = {
 761                 .lsp_env     = (struct lu_env *)env,
 762                 .lsp_cookie  = cookie,
 763                 .lsp_printer = printer,
 764         };
 765
 766         cfs_hash_for_each(s->ls_obj_hash, lu_site_obj_print, &arg);
 767 }
 768 EXPORT_SYMBOL(lu_site_print);
 769
 770 enum {
 771         LU_CACHE_PERCENT_MAX     = 50,
 772         LU_CACHE_PERCENT_DEFAULT = 20
 773 };
 774
 775 static unsigned int lu_cache_percent = LU_CACHE_PERCENT_DEFAULT;
 776 CFS_MODULE_PARM(lu_cache_percent, "i", int, 0644,
 777                 "Percentage of memory to be used as lu_object cache");
 778
 779 /**
 780  * Return desired hash table order.
 781  */
 782 static int lu_htable_order(void)
 783 {
 784         unsigned long cache_size;
 785         int bits;
 786
 787         /*
 788          * Calculate hash table size, assuming that we want reasonable
 789          * performance when 20% of total memory is occupied by cache of
 790          * lu_objects.
 791          *
 792          * Size of lu_object is (arbitrary) taken as 1K (together with inode).
 793          */
 794         cache_size = cfs_num_physpages;
 795
 796 #if BITS_PER_LONG == 32
 797         /* limit hashtable size for lowmem systems to low RAM */
 798         if (cache_size > 1 << (30 - CFS_PAGE_SHIFT))
 799                 cache_size = 1 << (30 - CFS_PAGE_SHIFT) * 3 / 4;
 800 #endif
 801
 802         /* clear off unreasonable cache setting. */
 803         if (lu_cache_percent == 0 || lu_cache_percent > LU_CACHE_PERCENT_MAX) {
 804                 CWARN("obdclass: invalid lu_cache_percent: %u, it must be in"
 805                       " the range of (0, %u]. Will use default value: %u.\n",
 806                       lu_cache_percent, LU_CACHE_PERCENT_MAX,
 807                       LU_CACHE_PERCENT_DEFAULT);
 808
 809                 lu_cache_percent = LU_CACHE_PERCENT_DEFAULT;
 810         }
 811         cache_size = cache_size / 100 * lu_cache_percent *
 812                 (CFS_PAGE_SIZE / 1024);
 813
 814         for (bits = 1; (1 << bits) < cache_size; ++bits) {
 815                 ;
 816         }
 817         return bits;
 818 }
 819
 820 static unsigned lu_obj_hop_hash(cfs_hash_t *hs,
 821                                 const void *key, unsigned mask)
 822 {
 823         struct lu_fid  *fid = (struct lu_fid *)key;
 824         __u32           hash;
 825
 826         hash = fid_flatten32(fid);
 827         hash += (hash >> 4) + (hash << 12); /* mixing oid and seq */
 828         hash = cfs_hash_long(hash, hs->hs_bkt_bits);
 829
 830         /* give me another random factor */
 831         hash -= cfs_hash_long((unsigned long)hs, fid_oid(fid) % 11 + 3);
 832
 833         hash <<= hs->hs_cur_bits - hs->hs_bkt_bits;
 834         hash |= (fid_seq(fid) + fid_oid(fid)) & (CFS_HASH_NBKT(hs) - 1);
 835
 836         return hash & mask;
 837 }
 838
 839 static void *lu_obj_hop_object(cfs_hlist_node_t *hnode)
 840 {
 841         return cfs_hlist_entry(hnode, struct lu_object_header, loh_hash);
 842 }
 843
 844 static void *lu_obj_hop_key(cfs_hlist_node_t *hnode)
 845 {
 846         struct lu_object_header *h;
 847
 848         h = cfs_hlist_entry(hnode, struct lu_object_header, loh_hash);
 849         return &h->loh_fid;
 850 }
 851
 852 static int lu_obj_hop_keycmp(const void *key, cfs_hlist_node_t *hnode)
 853 {
 854         struct lu_object_header *h;
 855
 856         h = cfs_hlist_entry(hnode, struct lu_object_header, loh_hash);
 857         return lu_fid_eq(&h->loh_fid, (struct lu_fid *)key);
 858 }
 859
 860 static void lu_obj_hop_get(cfs_hash_t *hs, cfs_hlist_node_t *hnode)
 861 {
 862         struct lu_object_header *h;
 863
 864         h = cfs_hlist_entry(hnode, struct lu_object_header, loh_hash);
 865         if (cfs_atomic_add_return(1, &h->loh_ref) == 1) {
 866                 struct lu_site_bkt_data *bkt;
 867                 cfs_hash_bd_t            bd;
 868
 869                 cfs_hash_bd_get(hs, &h->loh_fid, &bd);
 870                 bkt = cfs_hash_bd_extra_get(hs, &bd);
 871                 bkt->lsb_busy++;
 872         }
 873 }
 874
 875 static void lu_obj_hop_put_locked(cfs_hash_t *hs, cfs_hlist_node_t *hnode)
 876 {
 877         LBUG(); /* we should never called it */
 878 }
 879
 880 cfs_hash_ops_t lu_site_hash_ops = {
 881         .hs_hash        = lu_obj_hop_hash,
 882         .hs_key         = lu_obj_hop_key,
 883         .hs_keycmp      = lu_obj_hop_keycmp,
 884         .hs_object      = lu_obj_hop_object,
 885         .hs_get         = lu_obj_hop_get,
 886         .hs_put_locked  = lu_obj_hop_put_locked,
 887 };
 888
 889 void lu_dev_add_linkage(struct lu_site *s, struct lu_device *d)
 890 {
 891         cfs_spin_lock(&s->ls_ld_lock);
 892         if (cfs_list_empty(&d->ld_linkage))
 893                 cfs_list_add(&d->ld_linkage, &s->ls_ld_linkage);
 894         cfs_spin_unlock(&s->ls_ld_lock);
 895 }
 896 EXPORT_SYMBOL(lu_dev_add_linkage);
 897
 898 void lu_dev_del_linkage(struct lu_site *s, struct lu_device *d)
 899 {
 900         cfs_spin_lock(&s->ls_ld_lock);
 901         cfs_list_del_init(&d->ld_linkage);
 902         cfs_spin_unlock(&s->ls_ld_lock);
 903 }
 904 EXPORT_SYMBOL(lu_dev_del_linkage);
 905
 906 /**
 907  * Initialize site \a s, with \a d as the top level device.
 908  */
 909 #define LU_SITE_BITS_MIN    12
 910 #define LU_SITE_BITS_MAX    24
 911 /**
 912  * total 256 buckets, we don't want too many buckets because:
 913  * - consume too much memory
 914  * - avoid unbalanced LRU list
 915  */
 916 #define LU_SITE_BKT_BITS    8
 917
 918 int lu_site_init(struct lu_site *s, struct lu_device *top)
 919 {
 920         struct lu_site_bkt_data *bkt;
 921         cfs_hash_bd_t bd;
 922         char name[16];
 923         int bits;
 924         int i;
 925         ENTRY;
 926
 927         memset(s, 0, sizeof *s);
 928         bits = lu_htable_order();
 929         snprintf(name, 16, "lu_site_%s", top->ld_type->ldt_name);
 930         for (bits = min(max(LU_SITE_BITS_MIN, bits), LU_SITE_BITS_MAX);
 931              bits >= LU_SITE_BITS_MIN; bits--) {
 932                 s->ls_obj_hash = cfs_hash_create(name, bits, bits,
 933                                                  bits - LU_SITE_BKT_BITS,
 934                                                  sizeof(*bkt), 0, 0,
 935                                                  &lu_site_hash_ops,
 936                                                  CFS_HASH_SPIN_BKTLOCK |
 937                                                  CFS_HASH_NO_ITEMREF |
 938                                                  CFS_HASH_DEPTH |
 939                                                  CFS_HASH_ASSERT_EMPTY);
 940                 if (s->ls_obj_hash != NULL)
 941                         break;
 942         }
 943
 944         if (s->ls_obj_hash == NULL) {
 945                 CERROR("failed to create lu_site hash with bits: %d\n", bits);
 946                 return -ENOMEM;
 947         }
 948
 949         cfs_hash_for_each_bucket(s->ls_obj_hash, &bd, i) {
 950                 bkt = cfs_hash_bd_extra_get(s->ls_obj_hash, &bd);
 951                 CFS_INIT_LIST_HEAD(&bkt->lsb_lru);
 952                 cfs_waitq_init(&bkt->lsb_marche_funebre);
 953         }
 954
 955         s->ls_stats = lprocfs_alloc_stats(LU_SS_LAST_STAT, 0);
 956         if (s->ls_stats == NULL) {
 957                 cfs_hash_putref(s->ls_obj_hash);
 958                 s->ls_obj_hash = NULL;
 959                 return -ENOMEM;
 960         }
 961
 962         lprocfs_counter_init(s->ls_stats, LU_SS_CREATED,
 963                              0, "created", "created");
 964         lprocfs_counter_init(s->ls_stats, LU_SS_CACHE_HIT,
 965                              0, "cache_hit", "cache_hit");
 966         lprocfs_counter_init(s->ls_stats, LU_SS_CACHE_MISS,
 967                              0, "cache_miss", "cache_miss");
 968         lprocfs_counter_init(s->ls_stats, LU_SS_CACHE_RACE,
 969                              0, "cache_race", "cache_race");
 970         lprocfs_counter_init(s->ls_stats, LU_SS_CACHE_DEATH_RACE,
 971                              0, "cache_death_race", "cache_death_race");
 972         lprocfs_counter_init(s->ls_stats, LU_SS_LRU_PURGED,
 973                              0, "lru_purged", "lru_purged");
 974
 975         CFS_INIT_LIST_HEAD(&s->ls_linkage);
 976         s->ls_top_dev = top;
 977         top->ld_site = s;
 978         lu_device_get(top);
 979         lu_ref_add(&top->ld_reference, "site-top", s);
 980
 981         CFS_INIT_LIST_HEAD(&s->ls_ld_linkage);
 982         cfs_spin_lock_init(&s->ls_ld_lock);
 983
 984         lu_dev_add_linkage(s, top);
 985
 986         RETURN(0);
 987 }
 988 EXPORT_SYMBOL(lu_site_init);
 989
 990 /**
 991  * Finalize \a s and release its resources.
 992  */
 993 void lu_site_fini(struct lu_site *s)
 994 {
 995         cfs_mutex_lock(&lu_sites_guard);
 996         cfs_list_del_init(&s->ls_linkage);
 997         cfs_mutex_unlock(&lu_sites_guard);
 998
 999         if (s->ls_obj_hash != NULL) {
1000                 cfs_hash_putref(s->ls_obj_hash);
1001                 s->ls_obj_hash = NULL;
1002         }
1003
1004         if (s->ls_top_dev != NULL) {
1005                 s->ls_top_dev->ld_site = NULL;
1006                 lu_ref_del(&s->ls_top_dev->ld_reference, "site-top", s);
1007                 lu_device_put(s->ls_top_dev);
1008                 s->ls_top_dev = NULL;
1009         }
1010
1011         if (s->ls_stats != NULL)
1012                 lprocfs_free_stats(&s->ls_stats);
1013 }
1014 EXPORT_SYMBOL(lu_site_fini);
1015
1016 /**
1017  * Called when initialization of stack for this site is completed.
1018  */
1019 int lu_site_init_finish(struct lu_site *s)
1020 {
1021         int result;
1022         cfs_mutex_lock(&lu_sites_guard);
1023         result = lu_context_refill(&lu_shrink_env.le_ctx);
1024         if (result == 0)
1025                 cfs_list_add(&s->ls_linkage, &lu_sites);
1026         cfs_mutex_unlock(&lu_sites_guard);
1027         return result;
1028 }
1029 EXPORT_SYMBOL(lu_site_init_finish);
1030
1031 /**
1032  * Acquire additional reference on device \a d
1033  */
1034 void lu_device_get(struct lu_device *d)
1035 {
1036         cfs_atomic_inc(&d->ld_ref);
1037 }
1038 EXPORT_SYMBOL(lu_device_get);
1039
1040 /**
1041  * Release reference on device \a d.
1042  */
1043 void lu_device_put(struct lu_device *d)
1044 {
1045         LASSERT(cfs_atomic_read(&d->ld_ref) > 0);
1046         cfs_atomic_dec(&d->ld_ref);
1047 }
1048 EXPORT_SYMBOL(lu_device_put);
1049
1050 /**
1051  * Initialize device \a d of type \a t.
1052  */
1053 int lu_device_init(struct lu_device *d, struct lu_device_type *t)
1054 {
1055         if (t->ldt_device_nr++ == 0 && t->ldt_ops->ldto_start != NULL)
1056                 t->ldt_ops->ldto_start(t);
1057         memset(d, 0, sizeof *d);
1058         cfs_atomic_set(&d->ld_ref, 0);
1059         d->ld_type = t;
1060         lu_ref_init(&d->ld_reference);
1061         CFS_INIT_LIST_HEAD(&d->ld_linkage);
1062         return 0;
1063 }
1064 EXPORT_SYMBOL(lu_device_init);
1065
1066 /**
1067  * Finalize device \a d.
1068  */
1069 void lu_device_fini(struct lu_device *d)
1070 {
1071         struct lu_device_type *t;
1072
1073         t = d->ld_type;
1074         if (d->ld_obd != NULL) {
1075                 d->ld_obd->obd_lu_dev = NULL;
1076                 d->ld_obd = NULL;
1077         }
1078
1079         lu_ref_fini(&d->ld_reference);
1080         LASSERTF(cfs_atomic_read(&d->ld_ref) == 0,
1081                  "Refcount is %u\n", cfs_atomic_read(&d->ld_ref));
1082         LASSERT(t->ldt_device_nr > 0);
1083         if (--t->ldt_device_nr == 0 && t->ldt_ops->ldto_stop != NULL)
1084                 t->ldt_ops->ldto_stop(t);
1085 }
1086 EXPORT_SYMBOL(lu_device_fini);
1087
1088 /**
1089  * Initialize object \a o that is part of compound object \a h and was created
1090  * by device \a d.
1091  */
1092 int lu_object_init(struct lu_object *o,
1093                    struct lu_object_header *h, struct lu_device *d)
1094 {
1095         memset(o, 0, sizeof *o);
1096         o->lo_header = h;
1097         o->lo_dev    = d;
1098         lu_device_get(d);
1099         o->lo_dev_ref = lu_ref_add(&d->ld_reference, "lu_object", o);
1100         CFS_INIT_LIST_HEAD(&o->lo_linkage);
1101         return 0;
1102 }
1103 EXPORT_SYMBOL(lu_object_init);
1104
1105 /**
1106  * Finalize object and release its resources.
1107  */
1108 void lu_object_fini(struct lu_object *o)
1109 {
1110         struct lu_device *dev = o->lo_dev;
1111
1112         LASSERT(cfs_list_empty(&o->lo_linkage));
1113
1114         if (dev != NULL) {
1115                 lu_ref_del_at(&dev->ld_reference,
1116                               o->lo_dev_ref , "lu_object", o);
1117                 lu_device_put(dev);
1118                 o->lo_dev = NULL;
1119         }
1120 }
1121 EXPORT_SYMBOL(lu_object_fini);
1122
1123 /**
1124  * Add object \a o as first layer of compound object \a h
1125  *
1126  * This is typically called by the ->ldo_object_alloc() method of top-level
1127  * device.
1128  */
1129 void lu_object_add_top(struct lu_object_header *h, struct lu_object *o)
1130 {
1131         cfs_list_move(&o->lo_linkage, &h->loh_layers);
1132 }
1133 EXPORT_SYMBOL(lu_object_add_top);
1134
1135 /**
1136  * Add object \a o as a layer of compound object, going after \a before.
1137  *
1138  * This is typically called by the ->ldo_object_alloc() method of \a
1139  * before->lo_dev.
1140  */
1141 void lu_object_add(struct lu_object *before, struct lu_object *o)
1142 {
1143         cfs_list_move(&o->lo_linkage, &before->lo_linkage);
1144 }
1145 EXPORT_SYMBOL(lu_object_add);
1146
1147 /**
1148  * Initialize compound object.
1149  */
1150 int lu_object_header_init(struct lu_object_header *h)
1151 {
1152         memset(h, 0, sizeof *h);
1153         cfs_atomic_set(&h->loh_ref, 1);
1154         CFS_INIT_HLIST_NODE(&h->loh_hash);
1155         CFS_INIT_LIST_HEAD(&h->loh_lru);
1156         CFS_INIT_LIST_HEAD(&h->loh_layers);
1157         lu_ref_init(&h->loh_reference);
1158         return 0;
1159 }
1160 EXPORT_SYMBOL(lu_object_header_init);
1161
1162 /**
1163  * Finalize compound object.
1164  */
1165 void lu_object_header_fini(struct lu_object_header *h)
1166 {
1167         LASSERT(cfs_list_empty(&h->loh_layers));
1168         LASSERT(cfs_list_empty(&h->loh_lru));
1169         LASSERT(cfs_hlist_unhashed(&h->loh_hash));
1170         lu_ref_fini(&h->loh_reference);
1171 }
1172 EXPORT_SYMBOL(lu_object_header_fini);
1173
1174 /**
1175  * Given a compound object, find its slice, corresponding to the device type
1176  * \a dtype.
1177  */
1178 struct lu_object *lu_object_locate(struct lu_object_header *h,
1179                                    const struct lu_device_type *dtype)
1180 {
1181         struct lu_object *o;
1182
1183         cfs_list_for_each_entry(o, &h->loh_layers, lo_linkage) {
1184                 if (o->lo_dev->ld_type == dtype)
1185                         return o;
1186         }
1187         return NULL;
1188 }
1189 EXPORT_SYMBOL(lu_object_locate);
1190
1191
1192
1193 /**
1194  * Finalize and free devices in the device stack.
1195  *
1196  * Finalize device stack by purging object cache, and calling
1197  * lu_device_type_operations::ldto_device_fini() and
1198  * lu_device_type_operations::ldto_device_free() on all devices in the stack.
1199  */
1200 void lu_stack_fini(const struct lu_env *env, struct lu_device *top)
1201 {
1202         struct lu_site   *site = top->ld_site;
1203         struct lu_device *scan;
1204         struct lu_device *next;
1205
1206         lu_site_purge(env, site, ~0);
1207         for (scan = top; scan != NULL; scan = next) {
1208                 next = scan->ld_type->ldt_ops->ldto_device_fini(env, scan);
1209                 lu_ref_del(&scan->ld_reference, "lu-stack", &lu_site_init);
1210                 lu_device_put(scan);
1211         }
1212
1213         /* purge again. */
1214         lu_site_purge(env, site, ~0);
1215
1216         if (!cfs_hash_is_empty(site->ls_obj_hash)) {
1217                 /*
1218                  * Uh-oh, objects still exist.
1219                  */
1220                 LIBCFS_DEBUG_MSG_DATA_DECL(msgdata, D_ERROR, NULL);
1221
1222                 lu_site_print(env, site, &msgdata, lu_cdebug_printer);
1223         }
1224
1225         for (scan = top; scan != NULL; scan = next) {
1226                 const struct lu_device_type *ldt = scan->ld_type;
1227                 struct obd_type             *type;
1228
1229                 next = ldt->ldt_ops->ldto_device_free(env, scan);
1230                 type = ldt->ldt_obd_type;
1231                 if (type != NULL) {
1232                         type->typ_refcnt--;
1233                         class_put_type(type);
1234                 }
1235         }
1236 }
1237 EXPORT_SYMBOL(lu_stack_fini);
1238
1239 enum {
1240         /**
1241          * Maximal number of tld slots.
1242          */
1243         LU_CONTEXT_KEY_NR = 32
1244 };
1245
1246 static struct lu_context_key *lu_keys[LU_CONTEXT_KEY_NR] = { NULL, };
1247
1248 static DEFINE_SPINLOCK(lu_keys_guard);
1249
1250 /**
1251  * Global counter incremented whenever key is registered, unregistered,
1252  * revived or quiesced. This is used to void unnecessary calls to
1253  * lu_context_refill(). No locking is provided, as initialization and shutdown
1254  * are supposed to be externally serialized.
1255  */
1256 static unsigned key_set_version = 0;
1257
1258 /**
1259  * Register new key.
1260  */
1261 int lu_context_key_register(struct lu_context_key *key)
1262 {
1263         int result;
1264         int i;
1265
1266         LASSERT(key->lct_init != NULL);
1267         LASSERT(key->lct_fini != NULL);
1268         LASSERT(key->lct_tags != 0);
1269         LASSERT(key->lct_owner != NULL);
1270
1271         result = -ENFILE;
1272         cfs_spin_lock(&lu_keys_guard);
1273         for (i = 0; i < ARRAY_SIZE(lu_keys); ++i) {
1274                 if (lu_keys[i] == NULL) {
1275                         key->lct_index = i;
1276                         cfs_atomic_set(&key->lct_used, 1);
1277                         lu_keys[i] = key;
1278                         lu_ref_init(&key->lct_reference);
1279                         result = 0;
1280                         ++key_set_version;
1281                         break;
1282                 }
1283         }
1284         cfs_spin_unlock(&lu_keys_guard);
1285         return result;
1286 }
1287 EXPORT_SYMBOL(lu_context_key_register);
1288
1289 static void key_fini(struct lu_context *ctx, int index)
1290 {
1291         if (ctx->lc_value != NULL && ctx->lc_value[index] != NULL) {
1292                 struct lu_context_key *key;
1293
1294                 key = lu_keys[index];
1295                 LASSERT(key != NULL);
1296                 LASSERT(key->lct_fini != NULL);
1297                 LASSERT(cfs_atomic_read(&key->lct_used) > 1);
1298
1299                 key->lct_fini(ctx, key, ctx->lc_value[index]);
1300                 lu_ref_del(&key->lct_reference, "ctx", ctx);
1301                 cfs_atomic_dec(&key->lct_used);
1302
1303                 LASSERT(key->lct_owner != NULL);
1304                 if ((ctx->lc_tags & LCT_NOREF) == 0) {
1305                         LINVRNT(cfs_module_refcount(key->lct_owner) > 0);
1306                         cfs_module_put(key->lct_owner);
1307                 }
1308                 ctx->lc_value[index] = NULL;
1309         }
1310 }
1311
1312 /**
1313  * Deregister key.
1314  */
1315 void lu_context_key_degister(struct lu_context_key *key)
1316 {
1317         LASSERT(cfs_atomic_read(&key->lct_used) >= 1);
1318         LINVRNT(0 <= key->lct_index && key->lct_index < ARRAY_SIZE(lu_keys));
1319
1320         lu_context_key_quiesce(key);
1321
1322         ++key_set_version;
1323         cfs_spin_lock(&lu_keys_guard);
1324         key_fini(&lu_shrink_env.le_ctx, key->lct_index);
1325         if (lu_keys[key->lct_index]) {
1326                 lu_keys[key->lct_index] = NULL;
1327                 lu_ref_fini(&key->lct_reference);
1328         }
1329         cfs_spin_unlock(&lu_keys_guard);
1330
1331         LASSERTF(cfs_atomic_read(&key->lct_used) == 1,
1332                  "key has instances: %d\n",
1333                  cfs_atomic_read(&key->lct_used));
1334 }
1335 EXPORT_SYMBOL(lu_context_key_degister);
1336
1337 /**
1338  * Register a number of keys. This has to be called after all keys have been
1339  * initialized by a call to LU_CONTEXT_KEY_INIT().
1340  */
1341 int lu_context_key_register_many(struct lu_context_key *k, ...)
1342 {
1343         struct lu_context_key *key = k;
1344         va_list args;
1345         int result;
1346
1347         va_start(args, k);
1348         do {
1349                 result = lu_context_key_register(key);
1350                 if (result)
1351                         break;
1352                 key = va_arg(args, struct lu_context_key *);
1353         } while (key != NULL);
1354         va_end(args);
1355
1356         if (result != 0) {
1357                 va_start(args, k);
1358                 while (k != key) {
1359                         lu_context_key_degister(k);
1360                         k = va_arg(args, struct lu_context_key *);
1361                 }
1362                 va_end(args);
1363         }
1364
1365         return result;
1366 }
1367 EXPORT_SYMBOL(lu_context_key_register_many);
1368
1369 /**
1370  * De-register a number of keys. This is a dual to
1371  * lu_context_key_register_many().
1372  */
1373 void lu_context_key_degister_many(struct lu_context_key *k, ...)
1374 {
1375         va_list args;
1376
1377         va_start(args, k);
1378         do {
1379                 lu_context_key_degister(k);
1380                 k = va_arg(args, struct lu_context_key*);
1381         } while (k != NULL);
1382         va_end(args);
1383 }
1384 EXPORT_SYMBOL(lu_context_key_degister_many);
1385
1386 /**
1387  * Revive a number of keys.
1388  */
1389 void lu_context_key_revive_many(struct lu_context_key *k, ...)
1390 {
1391         va_list args;
1392
1393         va_start(args, k);
1394         do {
1395                 lu_context_key_revive(k);
1396                 k = va_arg(args, struct lu_context_key*);
1397         } while (k != NULL);
1398         va_end(args);
1399 }
1400 EXPORT_SYMBOL(lu_context_key_revive_many);
1401
1402 /**
1403  * Quiescent a number of keys.
1404  */
1405 void lu_context_key_quiesce_many(struct lu_context_key *k, ...)
1406 {
1407         va_list args;
1408
1409         va_start(args, k);
1410         do {
1411                 lu_context_key_quiesce(k);
1412                 k = va_arg(args, struct lu_context_key*);
1413         } while (k != NULL);
1414         va_end(args);
1415 }
1416 EXPORT_SYMBOL(lu_context_key_quiesce_many);
1417
1418 /**
1419  * Return value associated with key \a key in context \a ctx.
1420  */
1421 void *lu_context_key_get(const struct lu_context *ctx,
1422                          const struct lu_context_key *key)
1423 {
1424         LINVRNT(ctx->lc_state == LCS_ENTERED);
1425         LINVRNT(0 <= key->lct_index && key->lct_index < ARRAY_SIZE(lu_keys));
1426         LASSERT(lu_keys[key->lct_index] == key);
1427         return ctx->lc_value[key->lct_index];
1428 }
1429 EXPORT_SYMBOL(lu_context_key_get);
1430
1431 /**
1432  * List of remembered contexts. XXX document me.
1433  */
1434 static CFS_LIST_HEAD(lu_context_remembered);
1435
1436 /**
1437  * Destroy \a key in all remembered contexts. This is used to destroy key
1438  * values in "shared" contexts (like service threads), when a module owning
1439  * the key is about to be unloaded.
1440  */
1441 void lu_context_key_quiesce(struct lu_context_key *key)
1442 {
1443         struct lu_context *ctx;
1444         extern unsigned cl_env_cache_purge(unsigned nr);
1445
1446         if (!(key->lct_tags & LCT_QUIESCENT)) {
1447                 /*
1448                  * XXX layering violation.
1449                  */
1450                 cl_env_cache_purge(~0);
1451                 key->lct_tags |= LCT_QUIESCENT;
1452                 /*
1453                  * XXX memory barrier has to go here.
1454                  */
1455                 cfs_spin_lock(&lu_keys_guard);
1456                 cfs_list_for_each_entry(ctx, &lu_context_remembered,
1457                                         lc_remember)
1458                         key_fini(ctx, key->lct_index);
1459                 cfs_spin_unlock(&lu_keys_guard);
1460                 ++key_set_version;
1461         }
1462 }
1463 EXPORT_SYMBOL(lu_context_key_quiesce);
1464
1465 void lu_context_key_revive(struct lu_context_key *key)
1466 {
1467         key->lct_tags &= ~LCT_QUIESCENT;
1468         ++key_set_version;
1469 }
1470 EXPORT_SYMBOL(lu_context_key_revive);
1471
1472 static void keys_fini(struct lu_context *ctx)
1473 {
1474         int     i;
1475
1476         if (ctx->lc_value == NULL)
1477                 return;
1478
1479         for (i = 0; i < ARRAY_SIZE(lu_keys); ++i)
1480                 key_fini(ctx, i);
1481
1482         OBD_FREE(ctx->lc_value, ARRAY_SIZE(lu_keys) * sizeof ctx->lc_value[0]);
1483         ctx->lc_value = NULL;
1484 }
1485
1486 static int keys_fill(struct lu_context *ctx)
1487 {
1488         int i;
1489
1490         LINVRNT(ctx->lc_value != NULL);
1491         for (i = 0; i < ARRAY_SIZE(lu_keys); ++i) {
1492                 struct lu_context_key *key;
1493
1494                 key = lu_keys[i];
1495                 if (ctx->lc_value[i] == NULL && key != NULL &&
1496                     (key->lct_tags & ctx->lc_tags) &&
1497                     /*
1498                      * Don't create values for a LCT_QUIESCENT key, as this
1499                      * will pin module owning a key.
1500                      */
1501                     !(key->lct_tags & LCT_QUIESCENT)) {
1502                         void *value;
1503
1504                         LINVRNT(key->lct_init != NULL);
1505                         LINVRNT(key->lct_index == i);
1506
1507                         value = key->lct_init(ctx, key);
1508                         if (unlikely(IS_ERR(value)))
1509                                 return PTR_ERR(value);
1510
1511                         LASSERT(key->lct_owner != NULL);
1512                         if (!(ctx->lc_tags & LCT_NOREF))
1513                                 cfs_try_module_get(key->lct_owner);
1514                         lu_ref_add_atomic(&key->lct_reference, "ctx", ctx);
1515                         cfs_atomic_inc(&key->lct_used);
1516                         /*
1517                          * This is the only place in the code, where an
1518                          * element of ctx->lc_value[] array is set to non-NULL
1519                          * value.
1520                          */
1521                         ctx->lc_value[i] = value;
1522                         if (key->lct_exit != NULL)
1523                                 ctx->lc_tags |= LCT_HAS_EXIT;
1524                 }
1525                 ctx->lc_version = key_set_version;
1526         }
1527         return 0;
1528 }
1529
1530 static int keys_init(struct lu_context *ctx)
1531 {
1532         OBD_ALLOC(ctx->lc_value, ARRAY_SIZE(lu_keys) * sizeof ctx->lc_value[0]);
1533         if (likely(ctx->lc_value != NULL))
1534                 return keys_fill(ctx);
1535
1536         return -ENOMEM;
1537 }
1538
1539 /**
1540  * Initialize context data-structure. Create values for all keys.
1541  */
1542 int lu_context_init(struct lu_context *ctx, __u32 tags)
1543 {
1544         int     rc;
1545
1546         memset(ctx, 0, sizeof *ctx);
1547         ctx->lc_state = LCS_INITIALIZED;
1548         ctx->lc_tags = tags;
1549         if (tags & LCT_REMEMBER) {
1550                 cfs_spin_lock(&lu_keys_guard);
1551                 cfs_list_add(&ctx->lc_remember, &lu_context_remembered);
1552                 cfs_spin_unlock(&lu_keys_guard);
1553         } else {
1554                 CFS_INIT_LIST_HEAD(&ctx->lc_remember);
1555         }
1556
1557         rc = keys_init(ctx);
1558         if (rc != 0)
1559                 lu_context_fini(ctx);
1560
1561         return rc;
1562 }
1563 EXPORT_SYMBOL(lu_context_init);
1564
1565 /**
1566  * Finalize context data-structure. Destroy key values.
1567  */
1568 void lu_context_fini(struct lu_context *ctx)
1569 {
1570         LINVRNT(ctx->lc_state == LCS_INITIALIZED || ctx->lc_state == LCS_LEFT);
1571         ctx->lc_state = LCS_FINALIZED;
1572
1573         if ((ctx->lc_tags & LCT_REMEMBER) == 0) {
1574                 LASSERT(cfs_list_empty(&ctx->lc_remember));
1575                 keys_fini(ctx);
1576
1577         } else { /* could race with key degister */
1578                 cfs_spin_lock(&lu_keys_guard);
1579                 keys_fini(ctx);
1580                 cfs_list_del_init(&ctx->lc_remember);
1581                 cfs_spin_unlock(&lu_keys_guard);
1582         }
1583 }
1584 EXPORT_SYMBOL(lu_context_fini);
1585
1586 /**
1587  * Called before entering context.
1588  */
1589 void lu_context_enter(struct lu_context *ctx)
1590 {
1591         LINVRNT(ctx->lc_state == LCS_INITIALIZED || ctx->lc_state == LCS_LEFT);
1592         ctx->lc_state = LCS_ENTERED;
1593 }
1594 EXPORT_SYMBOL(lu_context_enter);
1595
1596 /**
1597  * Called after exiting from \a ctx
1598  */
1599 void lu_context_exit(struct lu_context *ctx)
1600 {
1601         int i;
1602
1603         LINVRNT(ctx->lc_state == LCS_ENTERED);
1604         ctx->lc_state = LCS_LEFT;
1605         if (ctx->lc_tags & LCT_HAS_EXIT && ctx->lc_value != NULL) {
1606                 for (i = 0; i < ARRAY_SIZE(lu_keys); ++i) {
1607                         if (ctx->lc_value[i] != NULL) {
1608                                 struct lu_context_key *key;
1609
1610                                 key = lu_keys[i];
1611                                 LASSERT(key != NULL);
1612                                 if (key->lct_exit != NULL)
1613                                         key->lct_exit(ctx,
1614                                                       key, ctx->lc_value[i]);
1615                         }
1616                 }
1617         }
1618 }
1619 EXPORT_SYMBOL(lu_context_exit);
1620
1621 /**
1622  * Allocate for context all missing keys that were registered after context
1623  * creation. key_set_version is only changed in rare cases when modules
1624  * are loaded and removed.
1625  */
1626 int lu_context_refill(struct lu_context *ctx)
1627 {
1628         return likely(ctx->lc_version == key_set_version) ? 0 : keys_fill(ctx);
1629 }
1630 EXPORT_SYMBOL(lu_context_refill);
1631
1632 /**
1633  * lu_ctx_tags/lu_ses_tags will be updated if there are new types of
1634  * obd being added. Currently, this is only used on client side, specifically
1635  * for echo device client, for other stack (like ptlrpc threads), context are
1636  * predefined when the lu_device type are registered, during the module probe
1637  * phase.
1638  */
1639 __u32 lu_context_tags_default = 0;
1640 __u32 lu_session_tags_default = 0;
1641
1642 void lu_context_tags_update(__u32 tags)
1643 {
1644         cfs_spin_lock(&lu_keys_guard);
1645         lu_context_tags_default |= tags;
1646         key_set_version ++;
1647         cfs_spin_unlock(&lu_keys_guard);
1648 }
1649 EXPORT_SYMBOL(lu_context_tags_update);
1650
1651 void lu_context_tags_clear(__u32 tags)
1652 {
1653         cfs_spin_lock(&lu_keys_guard);
1654         lu_context_tags_default &= ~tags;
1655         key_set_version ++;
1656         cfs_spin_unlock(&lu_keys_guard);
1657 }
1658 EXPORT_SYMBOL(lu_context_tags_clear);
1659
1660 void lu_session_tags_update(__u32 tags)
1661 {
1662         cfs_spin_lock(&lu_keys_guard);
1663         lu_session_tags_default |= tags;
1664         key_set_version ++;
1665         cfs_spin_unlock(&lu_keys_guard);
1666 }
1667 EXPORT_SYMBOL(lu_session_tags_update);
1668
1669 void lu_session_tags_clear(__u32 tags)
1670 {
1671         cfs_spin_lock(&lu_keys_guard);
1672         lu_session_tags_default &= ~tags;
1673         key_set_version ++;
1674         cfs_spin_unlock(&lu_keys_guard);
1675 }
1676 EXPORT_SYMBOL(lu_session_tags_clear);
1677
1678 int lu_env_init(struct lu_env *env, __u32 tags)
1679 {
1680         int result;
1681
1682         env->le_ses = NULL;
1683         result = lu_context_init(&env->le_ctx, tags);
1684         if (likely(result == 0))
1685                 lu_context_enter(&env->le_ctx);
1686         return result;
1687 }
1688 EXPORT_SYMBOL(lu_env_init);
1689
1690 void lu_env_fini(struct lu_env *env)
1691 {
1692         lu_context_exit(&env->le_ctx);
1693         lu_context_fini(&env->le_ctx);
1694         env->le_ses = NULL;
1695 }
1696 EXPORT_SYMBOL(lu_env_fini);
1697
1698 int lu_env_refill(struct lu_env *env)
1699 {
1700         int result;
1701
1702         result = lu_context_refill(&env->le_ctx);
1703         if (result == 0 && env->le_ses != NULL)
1704                 result = lu_context_refill(env->le_ses);
1705         return result;
1706 }
1707 EXPORT_SYMBOL(lu_env_refill);
1708
1709 /**
1710  * Currently, this API will only be used by echo client.
1711  * Because echo client and normal lustre client will share
1712  * same cl_env cache. So echo client needs to refresh
1713  * the env context after it get one from the cache, especially
1714  * when normal client and echo client co-exist in the same client.
1715  */
1716 int lu_env_refill_by_tags(struct lu_env *env, __u32 ctags,
1717                           __u32 stags)
1718 {
1719         int    result;
1720
1721         if ((env->le_ctx.lc_tags & ctags) != ctags) {
1722                 env->le_ctx.lc_version = 0;
1723                 env->le_ctx.lc_tags |= ctags;
1724         }
1725
1726         if (env->le_ses && (env->le_ses->lc_tags & stags) != stags) {
1727                 env->le_ses->lc_version = 0;
1728                 env->le_ses->lc_tags |= stags;
1729         }
1730
1731         result = lu_env_refill(env);
1732
1733         return result;
1734 }
1735 EXPORT_SYMBOL(lu_env_refill_by_tags);
1736
1737 static struct cfs_shrinker *lu_site_shrinker = NULL;
1738
1739 typedef struct lu_site_stats{
1740         unsigned        lss_populated;
1741         unsigned        lss_max_search;
1742         unsigned        lss_total;
1743         unsigned        lss_busy;
1744 } lu_site_stats_t;
1745
1746 static void lu_site_stats_get(cfs_hash_t *hs,
1747                               lu_site_stats_t *stats, int populated)
1748 {
1749         cfs_hash_bd_t bd;
1750         int           i;
1751
1752         cfs_hash_for_each_bucket(hs, &bd, i) {
1753                 struct lu_site_bkt_data *bkt = cfs_hash_bd_extra_get(hs, &bd);
1754                 cfs_hlist_head_t        *hhead;
1755
1756                 cfs_hash_bd_lock(hs, &bd, 1);
1757                 stats->lss_busy  += bkt->lsb_busy;
1758                 stats->lss_total += cfs_hash_bd_count_get(&bd);
1759                 stats->lss_max_search = max((int)stats->lss_max_search,
1760                                             cfs_hash_bd_depmax_get(&bd));
1761                 if (!populated) {
1762                         cfs_hash_bd_unlock(hs, &bd, 1);
1763                         continue;
1764                 }
1765
1766                 cfs_hash_bd_for_each_hlist(hs, &bd, hhead) {
1767                         if (!cfs_hlist_empty(hhead))
1768                                 stats->lss_populated++;
1769                 }
1770                 cfs_hash_bd_unlock(hs, &bd, 1);
1771         }
1772 }
1773
1774 #ifdef __KERNEL__
1775
1776 static int lu_cache_shrink(SHRINKER_ARGS(sc, nr_to_scan, gfp_mask))
1777 {
1778         lu_site_stats_t stats;
1779         struct lu_site *s;
1780         struct lu_site *tmp;
1781         int cached = 0;
1782         int remain = shrink_param(sc, nr_to_scan);
1783         CFS_LIST_HEAD(splice);
1784
1785         if (remain != 0) {
1786                 if (!(shrink_param(sc, gfp_mask) & __GFP_FS))
1787                         return -1;
1788                 CDEBUG(D_INODE, "Shrink %d objects\n", remain);
1789         }
1790
1791         cfs_mutex_lock(&lu_sites_guard);
1792         cfs_list_for_each_entry_safe(s, tmp, &lu_sites, ls_linkage) {
1793                 if (shrink_param(sc, nr_to_scan) != 0) {
1794                         remain = lu_site_purge(&lu_shrink_env, s, remain);
1795                         /*
1796                          * Move just shrunk site to the tail of site list to
1797                          * assure shrinking fairness.
1798                          */
1799                         cfs_list_move_tail(&s->ls_linkage, &splice);
1800                 }
1801
1802                 memset(&stats, 0, sizeof(stats));
1803                 lu_site_stats_get(s->ls_obj_hash, &stats, 0);
1804                 cached += stats.lss_total - stats.lss_busy;
1805                 if (shrink_param(sc, nr_to_scan) && remain <= 0)
1806                         break;
1807         }
1808         cfs_list_splice(&splice, lu_sites.prev);
1809         cfs_mutex_unlock(&lu_sites_guard);
1810
1811         cached = (cached / 100) * sysctl_vfs_cache_pressure;
1812         if (shrink_param(sc, nr_to_scan) == 0)
1813                 CDEBUG(D_INODE, "%d objects cached\n", cached);
1814         return cached;
1815 }
1816
1817 /*
1818  * Debugging stuff.
1819  */
1820
1821 /**
1822  * Environment to be used in debugger, contains all tags.
1823  */
1824 struct lu_env lu_debugging_env;
1825
1826 /**
1827  * Debugging printer function using printk().
1828  */
1829 int lu_printk_printer(const struct lu_env *env,
1830                       void *unused, const char *format, ...)
1831 {
1832         va_list args;
1833
1834         va_start(args, format);
1835         vprintk(format, args);
1836         va_end(args);
1837         return 0;
1838 }
1839
1840 void lu_debugging_setup(void)
1841 {
1842         lu_env_init(&lu_debugging_env, ~0);
1843 }
1844
1845 void lu_context_keys_dump(void)
1846 {
1847         int i;
1848
1849         for (i = 0; i < ARRAY_SIZE(lu_keys); ++i) {
1850                 struct lu_context_key *key;
1851
1852                 key = lu_keys[i];
1853                 if (key != NULL) {
1854                         CERROR("[%d]: %p %x (%p,%p,%p) %d %d \"%s\"@%p\n",
1855                                i, key, key->lct_tags,
1856                                key->lct_init, key->lct_fini, key->lct_exit,
1857                                key->lct_index, cfs_atomic_read(&key->lct_used),
1858                                key->lct_owner ? key->lct_owner->name : "",
1859                                key->lct_owner);
1860                         lu_ref_print(&key->lct_reference);
1861                 }
1862         }
1863 }
1864 EXPORT_SYMBOL(lu_context_keys_dump);
1865 #else  /* !__KERNEL__ */
1866 static int lu_cache_shrink(int nr, unsigned int gfp_mask)
1867 {
1868         return 0;
1869 }
1870 #endif /* __KERNEL__ */
1871
1872 int  cl_global_init(void);
1873 void cl_global_fini(void);
1874 int  lu_ref_global_init(void);
1875 void lu_ref_global_fini(void);
1876
1877 int dt_global_init(void);
1878 void dt_global_fini(void);
1879
1880 int llo_global_init(void);
1881 void llo_global_fini(void);
1882
1883 /**
1884  * Initialization of global lu_* data.
1885  */
1886 int lu_global_init(void)
1887 {
1888         int result;
1889
1890         CDEBUG(D_INFO, "Lustre LU module (%p).\n", &lu_keys);
1891
1892         result = lu_ref_global_init();
1893         if (result != 0)
1894                 return result;
1895
1896         LU_CONTEXT_KEY_INIT(&lu_global_key);
1897         result = lu_context_key_register(&lu_global_key);
1898         if (result != 0)
1899                 return result;
1900         /*
1901          * At this level, we don't know what tags are needed, so allocate them
1902          * conservatively. This should not be too bad, because this
1903          * environment is global.
1904          */
1905         cfs_mutex_lock(&lu_sites_guard);
1906         result = lu_env_init(&lu_shrink_env, LCT_SHRINKER);
1907         cfs_mutex_unlock(&lu_sites_guard);
1908         if (result != 0)
1909                 return result;
1910
1911         /*
1912          * seeks estimation: 3 seeks to read a record from oi, one to read
1913          * inode, one for ea. Unfortunately setting this high value results in
1914          * lu_object/inode cache consuming all the memory.
1915          */
1916         lu_site_shrinker = cfs_set_shrinker(CFS_DEFAULT_SEEKS, lu_cache_shrink);
1917         if (lu_site_shrinker == NULL)
1918                 return -ENOMEM;
1919
1920         result = lu_time_global_init();
1921         if (result)
1922                 GOTO(out, result);
1923
1924 #ifdef __KERNEL__
1925         result = dt_global_init();
1926         if (result)
1927                 GOTO(out, result);
1928
1929         result = llo_global_init();
1930         if (result)
1931                 GOTO(out, result);
1932 #endif
1933         result = cl_global_init();
1934 out:
1935
1936         return result;
1937 }
1938
1939 /**
1940  * Dual to lu_global_init().
1941  */
1942 void lu_global_fini(void)
1943 {
1944         cl_global_fini();
1945 #ifdef __KERNEL__
1946         llo_global_fini();
1947         dt_global_fini();
1948 #endif
1949         lu_time_global_fini();
1950         if (lu_site_shrinker != NULL) {
1951                 cfs_remove_shrinker(lu_site_shrinker);
1952                 lu_site_shrinker = NULL;
1953         }
1954
1955         lu_context_key_degister(&lu_global_key);
1956
1957         /*
1958          * Tear shrinker environment down _after_ de-registering
1959          * lu_global_key, because the latter has a value in the former.
1960          */
1961         cfs_mutex_lock(&lu_sites_guard);
1962         lu_env_fini(&lu_shrink_env);
1963         cfs_mutex_unlock(&lu_sites_guard);
1964
1965         lu_ref_global_fini();
1966 }
1967
1968 struct lu_buf LU_BUF_NULL = {
1969         .lb_buf = NULL,
1970         .lb_len = 0
1971 };
1972 EXPORT_SYMBOL(LU_BUF_NULL);
1973
1974 static __u32 ls_stats_read(struct lprocfs_stats *stats, int idx)
1975 {
1976 #ifdef LPROCFS
1977         struct lprocfs_counter ret;
1978
1979         lprocfs_stats_collect(stats, idx, &ret);
1980         return (__u32)ret.lc_count;
1981 #else
1982         return 0;
1983 #endif
1984 }
1985
1986 /**
1987  * Output site statistical counters into a buffer. Suitable for
1988  * lprocfs_rd_*()-style functions.
1989  */
1990 int lu_site_stats_print(const struct lu_site *s, char *page, int count)
1991 {
1992         lu_site_stats_t stats;
1993
1994         memset(&stats, 0, sizeof(stats));
1995         lu_site_stats_get(s->ls_obj_hash, &stats, 1);
1996
1997         return snprintf(page, count, "%d/%d %d/%d %d %d %d %d %d %d %d\n",
1998                         stats.lss_busy,
1999                         stats.lss_total,
2000                         stats.lss_populated,
2001                         CFS_HASH_NHLIST(s->ls_obj_hash),
2002                         stats.lss_max_search,
2003                         ls_stats_read(s->ls_stats, LU_SS_CREATED),
2004                         ls_stats_read(s->ls_stats, LU_SS_CACHE_HIT),
2005                         ls_stats_read(s->ls_stats, LU_SS_CACHE_MISS),
2006                         ls_stats_read(s->ls_stats, LU_SS_CACHE_RACE),
2007                         ls_stats_read(s->ls_stats, LU_SS_CACHE_DEATH_RACE),
2008                         ls_stats_read(s->ls_stats, LU_SS_LRU_PURGED));
2009 }
2010 EXPORT_SYMBOL(lu_site_stats_print);
2011
2012 const char *lu_time_names[LU_TIME_NR] = {
2013         [LU_TIME_FIND_LOOKUP] = "find_lookup",
2014         [LU_TIME_FIND_ALLOC]  = "find_alloc",
2015         [LU_TIME_FIND_INSERT] = "find_insert"
2016 };
2017 EXPORT_SYMBOL(lu_time_names);
2018
2019 /**
2020  * Helper function to initialize a number of kmem slab caches at once.
2021  */
2022 int lu_kmem_init(struct lu_kmem_descr *caches)
2023 {
2024         int result;
2025         struct lu_kmem_descr *iter = caches;
2026
2027         for (result = 0; iter->ckd_cache != NULL; ++iter) {
2028                 *iter->ckd_cache = cfs_mem_cache_create(iter->ckd_name,
2029                                                         iter->ckd_size,
2030                                                         0, 0);
2031                 if (*iter->ckd_cache == NULL) {
2032                         result = -ENOMEM;
2033                         /* free all previously allocated caches */
2034                         lu_kmem_fini(caches);
2035                         break;
2036                 }
2037         }
2038         return result;
2039 }
2040 EXPORT_SYMBOL(lu_kmem_init);
2041
2042 /**
2043  * Helper function to finalize a number of kmem slab cached at once. Dual to
2044  * lu_kmem_init().
2045  */
2046 void lu_kmem_fini(struct lu_kmem_descr *caches)
2047 {
2048         int rc;
2049
2050         for (; caches->ckd_cache != NULL; ++caches) {
2051                 if (*caches->ckd_cache != NULL) {
2052                         rc = cfs_mem_cache_destroy(*caches->ckd_cache);
2053                         LASSERTF(rc == 0, "couldn't destroy %s slab\n",
2054                                  caches->ckd_name);
2055                         *caches->ckd_cache = NULL;
2056                 }
2057         }
2058 }
2059 EXPORT_SYMBOL(lu_kmem_fini);