Whamcloud - gitweb
LU-73 Make register_shrinker argument more visible.
[fs/lustre-release.git] / lustre / obdclass / lu_object.c
index d5f6f5c..3023736 100644 (file)
@@ -26,7 +26,7 @@
  * GPL HEADER END
  */
 /*
- * Copyright  2008 Sun Microsystems, Inc. All rights reserved
+ * Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved.
  * Use is subject to license terms.
  */
 /*
@@ -73,53 +73,66 @@ static void lu_object_free(const struct lu_env *env, struct lu_object *o);
  */
 void lu_object_put(const struct lu_env *env, struct lu_object *o)
 {
+        struct lu_site_bkt_data *bkt;
         struct lu_object_header *top;
         struct lu_site          *site;
         struct lu_object        *orig;
-        int                      kill_it;
+        cfs_hash_bd_t            bd;
 
-        top = o->lo_header;
+        top  = o->lo_header;
         site = o->lo_dev->ld_site;
         orig = o;
-        kill_it = 0;
-        write_lock(&site->ls_guard);
-        if (atomic_dec_and_test(&top->loh_ref)) {
-                /*
-                 * When last reference is released, iterate over object
-                 * layers, and notify them that object is no longer busy.
-                 */
-                list_for_each_entry_reverse(o, &top->loh_layers, lo_linkage) {
-                        if (o->lo_ops->loo_object_release != NULL)
-                                o->lo_ops->loo_object_release(env, o);
-                }
-                -- site->ls_busy;
+
+        cfs_hash_bd_get(site->ls_obj_hash, &top->loh_fid, &bd);
+        bkt = cfs_hash_bd_extra_get(site->ls_obj_hash, &bd);
+
+        if (!cfs_hash_bd_dec_and_lock(site->ls_obj_hash, &bd, &top->loh_ref)) {
                 if (lu_object_is_dying(top)) {
+
                         /*
-                         * If object is dying (will not be cached), removed it
-                         * from hash table and LRU.
-                         *
-                         * This is done with hash table and LRU lists
-                         * locked. As the only way to acquire first reference
-                         * to previously unreferenced object is through
-                         * hash-table lookup (lu_object_find()), or LRU
-                         * scanning (lu_site_purge()), that are done under
-                         * hash-table and LRU lock, no race with concurrent
-                         * object lookup is possible and we can safely destroy
-                         * object below.
+                         * somebody may be waiting for this, currently only
+                         * used for cl_object, see cl_object_put_last().
                          */
-                        hlist_del_init(&top->loh_hash);
-                        list_del_init(&top->loh_lru);
-                        -- site->ls_total;
-                        kill_it = 1;
+                        cfs_waitq_broadcast(&bkt->lsb_marche_funebre);
                 }
+                return;
         }
-        write_unlock(&site->ls_guard);
-        if (kill_it)
-                /*
-                 * Object was already removed from hash and lru above, can
-                 * kill it.
-                 */
-                lu_object_free(env, orig);
+
+        LASSERT(bkt->lsb_busy > 0);
+        bkt->lsb_busy--;
+        /*
+         * When last reference is released, iterate over object
+         * layers, and notify them that object is no longer busy.
+         */
+        cfs_list_for_each_entry_reverse(o, &top->loh_layers, lo_linkage) {
+                if (o->lo_ops->loo_object_release != NULL)
+                        o->lo_ops->loo_object_release(env, o);
+        }
+
+        if (!lu_object_is_dying(top)) {
+                cfs_hash_bd_unlock(site->ls_obj_hash, &bd, 1);
+                return;
+        }
+
+        /*
+         * If object is dying (will not be cached), removed it
+         * from hash table and LRU.
+         *
+         * This is done with hash table and LRU lists locked. As the only
+         * way to acquire first reference to previously unreferenced
+         * object is through hash-table lookup (lu_object_find()),
+         * or LRU scanning (lu_site_purge()), that are done under hash-table
+         * and LRU lock, no race with concurrent object lookup is possible
+         * and we can safely destroy object below.
+         */
+        cfs_hash_bd_del_locked(site->ls_obj_hash, &bd, &top->loh_hash);
+        cfs_list_del_init(&top->loh_lru);
+        cfs_hash_bd_unlock(site->ls_obj_hash, &bd, 1);
+        /*
+         * Object was already removed from hash and lru above, can
+         * kill it.
+         */
+        lu_object_free(env, orig);
 }
 EXPORT_SYMBOL(lu_object_put);
 
@@ -136,7 +149,7 @@ static struct lu_object *lu_object_alloc(const struct lu_env *env,
 {
         struct lu_object *scan;
         struct lu_object *top;
-        struct list_head *layers;
+        cfs_list_t *layers;
         int clean;
         int result;
         ENTRY;
@@ -153,7 +166,7 @@ static struct lu_object *lu_object_alloc(const struct lu_env *env,
          * after this point.
          */
         LASSERT(fid_is_igif(f) || fid_ver(f) == 0);
-        top->lo_header->loh_fid  = *f;
+        top->lo_header->loh_fid = *f;
         layers = &top->lo_header->loh_layers;
         do {
                 /*
@@ -161,7 +174,7 @@ static struct lu_object *lu_object_alloc(const struct lu_env *env,
                  * object slices are created.
                  */
                 clean = 1;
-                list_for_each_entry(scan, layers, lo_linkage) {
+                cfs_list_for_each_entry(scan, layers, lo_linkage) {
                         if (scan->lo_flags & LU_OBJECT_ALLOCATED)
                                 continue;
                         clean = 0;
@@ -175,7 +188,7 @@ static struct lu_object *lu_object_alloc(const struct lu_env *env,
                 }
         } while (!clean);
 
-        list_for_each_entry_reverse(scan, layers, lo_linkage) {
+        cfs_list_for_each_entry_reverse(scan, layers, lo_linkage) {
                 if (scan->lo_ops->loo_object_start != NULL) {
                         result = scan->lo_ops->loo_object_start(env, scan);
                         if (result != 0) {
@@ -185,7 +198,7 @@ static struct lu_object *lu_object_alloc(const struct lu_env *env,
                 }
         }
 
-        dev->ld_site->ls_stats.s_created ++;
+        lprocfs_counter_incr(dev->ld_site->ls_stats, LU_SS_CREATED);
         RETURN(top);
 }
 
@@ -194,17 +207,19 @@ static struct lu_object *lu_object_alloc(const struct lu_env *env,
  */
 static void lu_object_free(const struct lu_env *env, struct lu_object *o)
 {
-        struct list_head  splice;
-        struct lu_object *scan;
-        struct lu_site   *site;
-        struct list_head *layers;
+        struct lu_site_bkt_data *bkt;
+        struct lu_site          *site;
+        struct lu_object        *scan;
+        cfs_list_t              *layers;
+        cfs_list_t               splice;
 
         site   = o->lo_dev->ld_site;
         layers = &o->lo_header->loh_layers;
+        bkt    = lu_site_bkt_from_fid(site, &o->lo_header->loh_fid);
         /*
          * First call ->loo_object_delete() method to release all resources.
          */
-        list_for_each_entry_reverse(scan, layers, lo_linkage) {
+        cfs_list_for_each_entry_reverse(scan, layers, lo_linkage) {
                 if (scan->lo_ops->loo_object_delete != NULL)
                         scan->lo_ops->loo_object_delete(env, scan);
         }
@@ -216,19 +231,21 @@ static void lu_object_free(const struct lu_env *env, struct lu_object *o)
          * top-level slice.
          */
         CFS_INIT_LIST_HEAD(&splice);
-        list_splice_init(layers, &splice);
-        while (!list_empty(&splice)) {
+        cfs_list_splice_init(layers, &splice);
+        while (!cfs_list_empty(&splice)) {
                 /*
                  * Free layers in bottom-to-top order, so that object header
                  * lives as long as possible and ->loo_object_free() methods
                  * can look at its contents.
                  */
                 o = container_of0(splice.prev, struct lu_object, lo_linkage);
-                list_del_init(&o->lo_linkage);
+                cfs_list_del_init(&o->lo_linkage);
                 LASSERT(o->lo_ops->loo_object_free != NULL);
                 o->lo_ops->loo_object_free(env, o);
         }
-        cfs_waitq_broadcast(&site->ls_marche_funebre);
+
+        if (cfs_waitq_active(&bkt->lsb_marche_funebre))
+                cfs_waitq_broadcast(&bkt->lsb_marche_funebre);
 }
 
 /**
@@ -236,47 +253,91 @@ static void lu_object_free(const struct lu_env *env, struct lu_object *o)
  */
 int lu_site_purge(const struct lu_env *env, struct lu_site *s, int nr)
 {
-        struct list_head         dispose;
         struct lu_object_header *h;
         struct lu_object_header *temp;
+        struct lu_site_bkt_data *bkt;
+        cfs_hash_bd_t            bd;
+        cfs_hash_bd_t            bd2;
+        cfs_list_t               dispose;
+        int                      did_sth;
+        int                      start;
+        int                      count;
+        int                      bnr;
+        int                      i;
 
         CFS_INIT_LIST_HEAD(&dispose);
         /*
          * Under LRU list lock, scan LRU list and move unreferenced objects to
          * the dispose list, removing them from LRU and hash table.
          */
-        write_lock(&s->ls_guard);
-        list_for_each_entry_safe(h, temp, &s->ls_lru, loh_lru) {
+        start = s->ls_purge_start;
+        bnr = (nr == ~0) ? -1 : nr / CFS_HASH_NBKT(s->ls_obj_hash) + 1;
+ again:
+        did_sth = 0;
+        cfs_hash_for_each_bucket(s->ls_obj_hash, &bd, i) {
+                if (i < start)
+                        continue;
+                count = bnr;
+                cfs_hash_bd_lock(s->ls_obj_hash, &bd, 1);
+                bkt = cfs_hash_bd_extra_get(s->ls_obj_hash, &bd);
+
+                cfs_list_for_each_entry_safe(h, temp, &bkt->lsb_lru, loh_lru) {
+                        /*
+                         * Objects are sorted in lru order, and "busy"
+                         * objects (ones with h->loh_ref > 0) naturally tend to
+                         * live near hot end that we scan last. Unfortunately,
+                         * sites usually have small (less then ten) number of
+                         * busy yet rarely accessed objects (some global
+                         * objects, accessed directly through pointers,
+                         * bypassing hash table).
+                         * Currently algorithm scans them over and over again.
+                         * Probably we should move busy objects out of LRU,
+                         * or we can live with that.
+                         */
+                        if (cfs_atomic_read(&h->loh_ref) > 0)
+                                continue;
+
+                        cfs_hash_bd_get(s->ls_obj_hash, &h->loh_fid, &bd2);
+                        LASSERT(bd.bd_bucket == bd2.bd_bucket);
+
+                        cfs_hash_bd_del_locked(s->ls_obj_hash,
+                                               &bd2, &h->loh_hash);
+                        cfs_list_move(&h->loh_lru, &dispose);
+                        if (did_sth == 0)
+                                did_sth = 1;
+
+                        if (nr != ~0 && --nr == 0)
+                                break;
+
+                        if (count > 0 && --count == 0)
+                                break;
+
+                }
+                cfs_hash_bd_unlock(s->ls_obj_hash, &bd, 1);
+                cfs_cond_resched();
                 /*
-                 * Objects are sorted in lru order, and "busy" objects (ones
-                 * with h->loh_ref > 0) naturally tend to live near hot end
-                 * that we scan last. Unfortunately, sites usually have small
-                 * (less then ten) number of busy yet rarely accessed objects
-                 * (some global objects, accessed directly through pointers,
-                 * bypassing hash table). Currently algorithm scans them over
-                 * and over again. Probably we should move busy objects out of
-                 * LRU, or we can live with that.
+                 * Free everything on the dispose list. This is safe against
+                 * races due to the reasons described in lu_object_put().
                  */
-                if (nr-- == 0)
+                while (!cfs_list_empty(&dispose)) {
+                        h = container_of0(dispose.next,
+                                          struct lu_object_header, loh_lru);
+                        cfs_list_del_init(&h->loh_lru);
+                        lu_object_free(env, lu_object_top(h));
+                        lprocfs_counter_incr(s->ls_stats, LU_SS_LRU_PURGED);
+                }
+
+                if (nr == 0)
                         break;
-                if (atomic_read(&h->loh_ref) > 0)
-                        continue;
-                hlist_del_init(&h->loh_hash);
-                list_move(&h->loh_lru, &dispose);
-                s->ls_total --;
         }
-        write_unlock(&s->ls_guard);
-        /*
-         * Free everything on the dispose list. This is safe against races due
-         * to the reasons described in lu_object_put().
-         */
-        while (!list_empty(&dispose)) {
-                h = container_of0(dispose.next,
-                                 struct lu_object_header, loh_lru);
-                list_del_init(&h->loh_lru);
-                lu_object_free(env, lu_object_top(h));
-                s->ls_stats.s_lru_purged ++;
+
+        if (nr != 0 && did_sth && start != 0) {
+                start = 0; /* restart from the first bucket */
+                goto again;
         }
+        /* race on s->ls_purge_start, but nobody cares */
+        s->ls_purge_start = i % CFS_HASH_NBKT(s->ls_obj_hash);
+
         return nr;
 }
 EXPORT_SYMBOL(lu_site_purge);
@@ -351,7 +412,7 @@ int lu_cdebug_printer(const struct lu_env *env,
         vsnprintf(key->lck_area + used,
                   ARRAY_SIZE(key->lck_area) - used, format, args);
         if (complete) {
-                if (cdebug_show(info->lpi_mask, info->lpi_subsys))
+                if (cfs_cdebug_show(info->lpi_mask, info->lpi_subsys))
                         libcfs_debug_msg(NULL, info->lpi_subsys, info->lpi_mask,
                                          (char *)info->lpi_file, info->lpi_fn,
                                          info->lpi_line, "%s", key->lck_area);
@@ -370,10 +431,11 @@ void lu_object_header_print(const struct lu_env *env, void *cookie,
                             const struct lu_object_header *hdr)
 {
         (*printer)(env, cookie, "header@%p[%#lx, %d, "DFID"%s%s%s]",
-                   hdr, hdr->loh_flags, atomic_read(&hdr->loh_ref),
+                   hdr, hdr->loh_flags, cfs_atomic_read(&hdr->loh_ref),
                    PFID(&hdr->loh_fid),
-                   hlist_unhashed(&hdr->loh_hash) ? "" : " hash",
-                   list_empty((struct list_head *)&hdr->loh_lru) ? "" : " lru",
+                   cfs_hlist_unhashed(&hdr->loh_hash) ? "" : " hash",
+                   cfs_list_empty((cfs_list_t *)&hdr->loh_lru) ? \
+                   "" : " lru",
                    hdr->loh_attr & LOHA_EXISTS ? " exist":"");
 }
 EXPORT_SYMBOL(lu_object_header_print);
@@ -391,7 +453,7 @@ void lu_object_print(const struct lu_env *env, void *cookie,
         top = o->lo_header;
         lu_object_header_print(env, cookie, printer, top);
         (*printer)(env, cookie, "{ \n");
-        list_for_each_entry(o, &top->loh_layers, lo_linkage) {
+        cfs_list_for_each_entry(o, &top->loh_layers, lo_linkage) {
                 depth = o->lo_depth + 4;
 
                 /*
@@ -415,7 +477,7 @@ int lu_object_invariant(const struct lu_object *o)
         struct lu_object_header *top;
 
         top = o->lo_header;
-        list_for_each_entry(o, &top->loh_layers, lo_linkage) {
+        cfs_list_for_each_entry(o, &top->loh_layers, lo_linkage) {
                 if (o->lo_ops->loo_object_invariant != NULL &&
                     !o->lo_ops->loo_object_invariant(o))
                         return 0;
@@ -425,52 +487,47 @@ int lu_object_invariant(const struct lu_object *o)
 EXPORT_SYMBOL(lu_object_invariant);
 
 static struct lu_object *htable_lookup(struct lu_site *s,
-                                       const struct hlist_head *bucket,
+                                       cfs_hash_bd_t *bd,
                                        const struct lu_fid *f,
-                                       cfs_waitlink_t *waiter)
+                                       cfs_waitlink_t *waiter,
+                                       __u64 *version)
 {
+        struct lu_site_bkt_data *bkt;
         struct lu_object_header *h;
-        struct hlist_node *scan;
-
-        hlist_for_each_entry(h, scan, bucket, loh_hash) {
-                s->ls_stats.s_cache_check ++;
-                if (likely(lu_fid_eq(&h->loh_fid, f))) {
-                        if (unlikely(lu_object_is_dying(h))) {
-                                /*
-                                 * Lookup found an object being destroyed;
-                                 * this object cannot be returned (to assure
-                                 * that references to dying objects are
-                                 * eventually drained), and moreover, lookup
-                                 * has to wait until object is freed.
-                                 */
-                                cfs_waitlink_init(waiter);
-                                cfs_waitq_add(&s->ls_marche_funebre, waiter);
-                                set_current_state(CFS_TASK_UNINT);
-                                s->ls_stats.s_cache_death_race ++;
-                                return ERR_PTR(-EAGAIN);
-                        }
-                        /* bump reference count... */
-                        if (atomic_add_return(1, &h->loh_ref) == 1)
-                                ++ s->ls_busy;
-                        /* and move to the head of the LRU */
-                        /*
-                         * XXX temporary disable this to measure effects of
-                         * read-write locking.
-                         */
-                        /* list_move_tail(&h->loh_lru, &s->ls_lru); */
-                        s->ls_stats.s_cache_hit ++;
-                        return lu_object_top(h);
-                }
+        cfs_hlist_node_t        *hnode;
+        __u64  ver = cfs_hash_bd_version_get(bd);
+
+        if (*version == ver)
+                return NULL;
+
+        *version = ver;
+        bkt = cfs_hash_bd_extra_get(s->ls_obj_hash, bd);
+        /* cfs_hash_bd_lookup_intent is a somehow "internal" function
+         * of cfs_hash, but we don't want refcount on object right now */
+        hnode = cfs_hash_bd_lookup_locked(s->ls_obj_hash, bd, (void *)f);
+        if (hnode == NULL) {
+                lprocfs_counter_incr(s->ls_stats, LU_SS_CACHE_MISS);
+                return NULL;
         }
-        s->ls_stats.s_cache_miss ++;
-        return NULL;
-}
 
-static __u32 fid_hash(const struct lu_fid *f, int bits)
-{
-        /* all objects with same id and different versions will belong to same
-         * collisions list. */
-        return hash_long(fid_flatten(f), bits);
+        h = container_of0(hnode, struct lu_object_header, loh_hash);
+        if (likely(!lu_object_is_dying(h))) {
+                lprocfs_counter_incr(s->ls_stats, LU_SS_CACHE_HIT);
+                return lu_object_top(h);
+        }
+
+        /*
+         * Lookup found an object being destroyed this object cannot be
+         * returned (to assure that references to dying objects are eventually
+         * drained), and moreover, lookup has to wait until object is freed.
+         */
+        cfs_atomic_dec(&h->loh_ref);
+
+        cfs_waitlink_init(waiter);
+        cfs_waitq_add(&bkt->lsb_marche_funebre, waiter);
+        cfs_set_current_state(CFS_TASK_UNINT);
+        lprocfs_counter_incr(s->ls_stats, LU_SS_CACHE_DEATH_RACE);
+        return ERR_PTR(-EAGAIN);
 }
 
 /**
@@ -495,10 +552,12 @@ static struct lu_object *lu_object_find_try(const struct lu_env *env,
                                             const struct lu_object_conf *conf,
                                             cfs_waitlink_t *waiter)
 {
-        struct lu_site    *s;
-        struct lu_object  *o;
-        struct lu_object  *shadow;
-        struct hlist_head *bucket;
+        struct lu_object      *o;
+        struct lu_object      *shadow;
+        struct lu_site        *s;
+        cfs_hash_t            *hs;
+        cfs_hash_bd_t          bd;
+        __u64                  version = 0;
 
         /*
          * This uses standard index maintenance protocol:
@@ -516,14 +575,11 @@ static struct lu_object *lu_object_find_try(const struct lu_env *env,
          * If dying object is found during index search, add @waiter to the
          * site wait-queue and return ERR_PTR(-EAGAIN).
          */
-
-        s = dev->ld_site;
-        bucket = s->ls_hash + fid_hash(f, s->ls_hash_bits);
-
-        read_lock(&s->ls_guard);
-        o = htable_lookup(s, bucket, f, waiter);
-        read_unlock(&s->ls_guard);
-
+        s  = dev->ld_site;
+        hs = s->ls_obj_hash;
+        cfs_hash_bd_get_and_lock(hs, (void *)f, &bd, 1);
+        o = htable_lookup(s, &bd, f, waiter, &version);
+        cfs_hash_bd_unlock(hs, &bd, 1);
         if (o != NULL)
                 return o;
 
@@ -537,20 +593,23 @@ static struct lu_object *lu_object_find_try(const struct lu_env *env,
 
         LASSERT(lu_fid_eq(lu_object_fid(o), f));
 
-        write_lock(&s->ls_guard);
-        shadow = htable_lookup(s, bucket, f, waiter);
+        cfs_hash_bd_lock(hs, &bd, 1);
+
+        shadow = htable_lookup(s, &bd, f, waiter, &version);
         if (likely(shadow == NULL)) {
-                hlist_add_head(&o->lo_header->loh_hash, bucket);
-                list_add_tail(&o->lo_header->loh_lru, &s->ls_lru);
-                ++ s->ls_busy;
-                ++ s->ls_total;
-                shadow = o;
-                o = NULL;
-        } else
-                s->ls_stats.s_cache_race ++;
-        write_unlock(&s->ls_guard);
-        if (o != NULL)
-                lu_object_free(env, o);
+                struct lu_site_bkt_data *bkt;
+
+                bkt = cfs_hash_bd_extra_get(hs, &bd);
+                cfs_hash_bd_add_locked(hs, &bd, &o->lo_header->loh_hash);
+                cfs_list_add_tail(&o->lo_header->loh_lru, &bkt->lsb_lru);
+                bkt->lsb_busy++;
+                cfs_hash_bd_unlock(hs, &bd, 1);
+                return o;
+        }
+
+        lprocfs_counter_incr(s->ls_stats, LU_SS_CACHE_RACE);
+        cfs_hash_bd_unlock(hs, &bd, 1);
+        lu_object_free(env, o);
         return shadow;
 }
 
@@ -564,22 +623,22 @@ struct lu_object *lu_object_find_at(const struct lu_env *env,
                                     const struct lu_fid *f,
                                     const struct lu_object_conf *conf)
 {
-        struct lu_object *obj;
-        cfs_waitlink_t    wait;
+        struct lu_site_bkt_data *bkt;
+        struct lu_object        *obj;
+        cfs_waitlink_t           wait;
 
         while (1) {
                 obj = lu_object_find_try(env, dev, f, conf, &wait);
-                if (obj == ERR_PTR(-EAGAIN)) {
-                        /*
-                         * lu_object_find_try() already added waiter into the
-                         * wait queue.
-                         */
-                        cfs_waitq_wait(&wait, CFS_TASK_UNINT);
-                        cfs_waitq_del(&dev->ld_site->ls_marche_funebre, &wait);
-                } else
-                        break;
+                if (obj != ERR_PTR(-EAGAIN))
+                        return obj;
+                /*
+                 * lu_object_find_try() already added waiter into the
+                 * wait queue.
+                 */
+                cfs_waitq_wait(&wait, CFS_TASK_UNINT);
+                bkt = lu_site_bkt_from_fid(dev->ld_site, (void *)f);
+                cfs_waitq_del(&bkt->lsb_marche_funebre, &wait);
         }
-        return obj;
 }
 EXPORT_SYMBOL(lu_object_find_at);
 
@@ -617,14 +676,14 @@ int lu_device_type_init(struct lu_device_type *ldt)
         CFS_INIT_LIST_HEAD(&ldt->ldt_linkage);
         result = ldt->ldt_ops->ldto_init(ldt);
         if (result == 0)
-                list_add(&ldt->ldt_linkage, &lu_device_types);
+                cfs_list_add(&ldt->ldt_linkage, &lu_device_types);
         return result;
 }
 EXPORT_SYMBOL(lu_device_type_init);
 
 void lu_device_type_fini(struct lu_device_type *ldt)
 {
-        list_del_init(&ldt->ldt_linkage);
+        cfs_list_del_init(&ldt->ldt_linkage);
         ldt->ldt_ops->ldto_fini(ldt);
 }
 EXPORT_SYMBOL(lu_device_type_fini);
@@ -633,7 +692,7 @@ void lu_types_stop(void)
 {
         struct lu_device_type *ldt;
 
-        list_for_each_entry(ldt, &lu_device_types, ldt_linkage) {
+        cfs_list_for_each_entry(ldt, &lu_device_types, ldt_linkage) {
                 if (ldt->ldt_device_nr == 0)
                         ldt->ldt_ops->ldto_stop(ldt);
         }
@@ -644,38 +703,53 @@ EXPORT_SYMBOL(lu_types_stop);
  * Global list of all sites on this node
  */
 static CFS_LIST_HEAD(lu_sites);
-static DECLARE_MUTEX(lu_sites_guard);
+static CFS_DECLARE_MUTEX(lu_sites_guard);
 
 /**
  * Global environment used by site shrinker.
  */
 static struct lu_env lu_shrink_env;
 
+struct lu_site_print_arg {
+        struct lu_env   *lsp_env;
+        void            *lsp_cookie;
+        lu_printer_t     lsp_printer;
+};
+
+static int
+lu_site_obj_print(cfs_hash_t *hs, cfs_hash_bd_t *bd,
+                  cfs_hlist_node_t *hnode, void *data)
+{
+        struct lu_site_print_arg *arg = (struct lu_site_print_arg *)data;
+        struct lu_object_header  *h;
+
+        h = cfs_hlist_entry(hnode, struct lu_object_header, loh_hash);
+        if (!cfs_list_empty(&h->loh_layers)) {
+                const struct lu_object *o;
+
+                o = lu_object_top(h);
+                lu_object_print(arg->lsp_env, arg->lsp_cookie,
+                                arg->lsp_printer, o);
+        } else {
+                lu_object_header_print(arg->lsp_env, arg->lsp_cookie,
+                                       arg->lsp_printer, h);
+        }
+        return 0;
+}
+
 /**
  * Print all objects in \a s.
  */
 void lu_site_print(const struct lu_env *env, struct lu_site *s, void *cookie,
                    lu_printer_t printer)
 {
-        int i;
-
-        for (i = 0; i < s->ls_hash_size; ++i) {
-                struct lu_object_header *h;
-                struct hlist_node       *scan;
-
-                read_lock(&s->ls_guard);
-                hlist_for_each_entry(h, scan, &s->ls_hash[i], loh_hash) {
+        struct lu_site_print_arg arg = {
+                .lsp_env     = (struct lu_env *)env,
+                .lsp_cookie  = cookie,
+                .lsp_printer = printer,
+        };
 
-                        if (!list_empty(&h->loh_layers)) {
-                                const struct lu_object *obj;
-
-                                obj = lu_object_top(h);
-                                lu_object_print(env, cookie, printer, obj);
-                        } else
-                                lu_object_header_print(env, cookie, printer, h);
-                }
-                read_unlock(&s->ls_guard);
-        }
+        cfs_hash_for_each(s->ls_obj_hash, lu_site_obj_print, &arg);
 }
 EXPORT_SYMBOL(lu_site_print);
 
@@ -698,7 +772,7 @@ static int lu_htable_order(void)
          *
          * Size of lu_object is (arbitrary) taken as 1K (together with inode).
          */
-        cache_size = num_physpages;
+        cache_size = cfs_num_physpages;
 
 #if BITS_PER_LONG == 32
         /* limit hashtable size for lowmem systems to low RAM */
@@ -715,46 +789,139 @@ static int lu_htable_order(void)
         return bits;
 }
 
-static struct lock_class_key lu_site_guard_class;
+static unsigned lu_obj_hop_hash(cfs_hash_t *hs, void *key, unsigned mask)
+{
+        struct lu_fid  *fid = (struct lu_fid *)key;
+        unsigned        hash;
+
+        hash = (fid_seq(fid) + fid_oid(fid)) & (CFS_HASH_NBKT(hs) - 1);
+        hash += fid_hash(fid, hs->hs_bkt_bits) << hs->hs_bkt_bits;
+        return hash & mask;
+}
+
+static void *lu_obj_hop_object(cfs_hlist_node_t *hnode)
+{
+        return cfs_hlist_entry(hnode, struct lu_object_header, loh_hash);
+}
+
+static void *lu_obj_hop_key(cfs_hlist_node_t *hnode)
+{
+        struct lu_object_header *h;
+
+        h = cfs_hlist_entry(hnode, struct lu_object_header, loh_hash);
+        return &h->loh_fid;
+}
+
+static int lu_obj_hop_keycmp(void *key, cfs_hlist_node_t *hnode)
+{
+        struct lu_object_header *h;
+
+        h = cfs_hlist_entry(hnode, struct lu_object_header, loh_hash);
+        return lu_fid_eq(&h->loh_fid, (struct lu_fid *)key);
+}
+
+static void lu_obj_hop_get(cfs_hash_t *hs, cfs_hlist_node_t *hnode)
+{
+        struct lu_object_header *h;
+
+        h = cfs_hlist_entry(hnode, struct lu_object_header, loh_hash);
+        if (cfs_atomic_add_return(1, &h->loh_ref) == 1) {
+                struct lu_site_bkt_data *bkt;
+                cfs_hash_bd_t            bd;
+
+                cfs_hash_bd_get(hs, &h->loh_fid, &bd);
+                bkt = cfs_hash_bd_extra_get(hs, &bd);
+                bkt->lsb_busy++;
+        }
+}
+
+static void lu_obj_hop_put_locked(cfs_hash_t *hs, cfs_hlist_node_t *hnode)
+{
+        LBUG(); /* we should never called it */
+}
+
+cfs_hash_ops_t lu_site_hash_ops = {
+        .hs_hash        = lu_obj_hop_hash,
+        .hs_key         = lu_obj_hop_key,
+        .hs_keycmp      = lu_obj_hop_keycmp,
+        .hs_object      = lu_obj_hop_object,
+        .hs_get         = lu_obj_hop_get,
+        .hs_put_locked  = lu_obj_hop_put_locked,
+};
 
 /**
  * Initialize site \a s, with \a d as the top level device.
  */
+#define LU_SITE_BITS_MIN    12
+#define LU_SITE_BITS_MAX    23
+/**
+ * total 128 buckets, we don't want too many buckets because:
+ * - consume too much memory
+ * - avoid unbalanced LRU list
+ */
+#define LU_SITE_BKT_BITS    7
+
 int lu_site_init(struct lu_site *s, struct lu_device *top)
 {
+        struct lu_site_bkt_data *bkt;
+        cfs_hash_bd_t bd;
         int bits;
-        int size;
         int i;
         ENTRY;
 
         memset(s, 0, sizeof *s);
-        rwlock_init(&s->ls_guard);
-        lockdep_set_class(&s->ls_guard, &lu_site_guard_class);
-        CFS_INIT_LIST_HEAD(&s->ls_lru);
+        bits = lu_htable_order();
+        for (bits = min(max(LU_SITE_BITS_MIN, bits), LU_SITE_BITS_MAX);
+             bits >= LU_SITE_BITS_MIN; bits--) {
+                s->ls_obj_hash = cfs_hash_create("lu_site", bits, bits,
+                                                 bits - LU_SITE_BKT_BITS,
+                                                 sizeof(*bkt), 0, 0,
+                                                 &lu_site_hash_ops,
+                                                 CFS_HASH_SPIN_BKTLOCK |
+                                                 CFS_HASH_NO_ITEMREF |
+                                                 CFS_HASH_DEPTH |
+                                                 CFS_HASH_ASSERT_EMPTY);
+                if (s->ls_obj_hash != NULL)
+                        break;
+        }
+
+        if (s->ls_obj_hash == NULL) {
+                CERROR("failed to create lu_site hash with bits: %d\n", bits);
+                return -ENOMEM;
+        }
+
+        cfs_hash_for_each_bucket(s->ls_obj_hash, &bd, i) {
+                bkt = cfs_hash_bd_extra_get(s->ls_obj_hash, &bd);
+                CFS_INIT_LIST_HEAD(&bkt->lsb_lru);
+                cfs_waitq_init(&bkt->lsb_marche_funebre);
+        }
+
+        s->ls_stats = lprocfs_alloc_stats(LU_SS_LAST_STAT, 0);
+        if (s->ls_stats == NULL) {
+                cfs_hash_putref(s->ls_obj_hash);
+                s->ls_obj_hash = NULL;
+                return -ENOMEM;
+        }
+
+        lprocfs_counter_init(s->ls_stats, LU_SS_CREATED,
+                             0, "created", "created");
+        lprocfs_counter_init(s->ls_stats, LU_SS_CACHE_HIT,
+                             0, "cache_hit", "cache_hit");
+        lprocfs_counter_init(s->ls_stats, LU_SS_CACHE_MISS,
+                             0, "cache_miss", "cache_miss");
+        lprocfs_counter_init(s->ls_stats, LU_SS_CACHE_RACE,
+                             0, "cache_race", "cache_race");
+        lprocfs_counter_init(s->ls_stats, LU_SS_CACHE_DEATH_RACE,
+                             0, "cache_death_race", "cache_death_race");
+        lprocfs_counter_init(s->ls_stats, LU_SS_LRU_PURGED,
+                             0, "lru_purged", "lru_purged");
+
         CFS_INIT_LIST_HEAD(&s->ls_linkage);
-        cfs_waitq_init(&s->ls_marche_funebre);
         s->ls_top_dev = top;
         top->ld_site = s;
         lu_device_get(top);
         lu_ref_add(&top->ld_reference, "site-top", s);
 
-        for (bits = lu_htable_order(), size = 1 << bits;
-             (s->ls_hash =
-              cfs_alloc_large(size * sizeof s->ls_hash[0])) == NULL;
-             --bits, size >>= 1) {
-                /*
-                 * Scale hash table down, until allocation succeeds.
-                 */
-                ;
-        }
-
-        s->ls_hash_size = size;
-        s->ls_hash_bits = bits;
-        s->ls_hash_mask = size - 1;
-
-        for (i = 0; i < size; i++)
-                INIT_HLIST_HEAD(&s->ls_hash[i]);
-
         RETURN(0);
 }
 EXPORT_SYMBOL(lu_site_init);
@@ -764,26 +931,24 @@ EXPORT_SYMBOL(lu_site_init);
  */
 void lu_site_fini(struct lu_site *s)
 {
-        LASSERT(list_empty(&s->ls_lru));
-        LASSERT(s->ls_total == 0);
-
-        down(&lu_sites_guard);
-        list_del_init(&s->ls_linkage);
-        up(&lu_sites_guard);
+        cfs_down(&lu_sites_guard);
+        cfs_list_del_init(&s->ls_linkage);
+        cfs_up(&lu_sites_guard);
 
-        if (s->ls_hash != NULL) {
-                int i;
-                for (i = 0; i < s->ls_hash_size; i++)
-                        LASSERT(hlist_empty(&s->ls_hash[i]));
-                cfs_free_large(s->ls_hash);
-                s->ls_hash = NULL;
+        if (s->ls_obj_hash != NULL) {
+                cfs_hash_putref(s->ls_obj_hash);
+                s->ls_obj_hash = NULL;
         }
+
         if (s->ls_top_dev != NULL) {
                 s->ls_top_dev->ld_site = NULL;
                 lu_ref_del(&s->ls_top_dev->ld_reference, "site-top", s);
                 lu_device_put(s->ls_top_dev);
                 s->ls_top_dev = NULL;
         }
+
+        if (s->ls_stats != NULL)
+                lprocfs_free_stats(&s->ls_stats);
 }
 EXPORT_SYMBOL(lu_site_fini);
 
@@ -793,11 +958,11 @@ EXPORT_SYMBOL(lu_site_fini);
 int lu_site_init_finish(struct lu_site *s)
 {
         int result;
-        down(&lu_sites_guard);
+        cfs_down(&lu_sites_guard);
         result = lu_context_refill(&lu_shrink_env.le_ctx);
         if (result == 0)
-                list_add(&s->ls_linkage, &lu_sites);
-        up(&lu_sites_guard);
+                cfs_list_add(&s->ls_linkage, &lu_sites);
+        cfs_up(&lu_sites_guard);
         return result;
 }
 EXPORT_SYMBOL(lu_site_init_finish);
@@ -807,7 +972,7 @@ EXPORT_SYMBOL(lu_site_init_finish);
  */
 void lu_device_get(struct lu_device *d)
 {
-        atomic_inc(&d->ld_ref);
+        cfs_atomic_inc(&d->ld_ref);
 }
 EXPORT_SYMBOL(lu_device_get);
 
@@ -816,8 +981,8 @@ EXPORT_SYMBOL(lu_device_get);
  */
 void lu_device_put(struct lu_device *d)
 {
-        LASSERT(atomic_read(&d->ld_ref) > 0);
-        atomic_dec(&d->ld_ref);
+        LASSERT(cfs_atomic_read(&d->ld_ref) > 0);
+        cfs_atomic_dec(&d->ld_ref);
 }
 EXPORT_SYMBOL(lu_device_put);
 
@@ -829,7 +994,7 @@ int lu_device_init(struct lu_device *d, struct lu_device_type *t)
         if (t->ldt_device_nr++ == 0 && t->ldt_ops->ldto_start != NULL)
                 t->ldt_ops->ldto_start(t);
         memset(d, 0, sizeof *d);
-        atomic_set(&d->ld_ref, 0);
+        cfs_atomic_set(&d->ld_ref, 0);
         d->ld_type = t;
         lu_ref_init(&d->ld_reference);
         return 0;
@@ -845,15 +1010,13 @@ void lu_device_fini(struct lu_device *d)
 
         t = d->ld_type;
         if (d->ld_obd != NULL) {
-                /* finish lprocfs */
-                lprocfs_obd_cleanup(d->ld_obd);
                 d->ld_obd->obd_lu_dev = NULL;
                 d->ld_obd = NULL;
         }
 
         lu_ref_fini(&d->ld_reference);
-        LASSERTF(atomic_read(&d->ld_ref) == 0,
-                 "Refcount is %u\n", atomic_read(&d->ld_ref));
+        LASSERTF(cfs_atomic_read(&d->ld_ref) == 0,
+                 "Refcount is %u\n", cfs_atomic_read(&d->ld_ref));
         LASSERT(t->ldt_device_nr > 0);
         if (--t->ldt_device_nr == 0 && t->ldt_ops->ldto_stop != NULL)
                 t->ldt_ops->ldto_stop(t);
@@ -884,7 +1047,7 @@ void lu_object_fini(struct lu_object *o)
 {
         struct lu_device *dev = o->lo_dev;
 
-        LASSERT(list_empty(&o->lo_linkage));
+        LASSERT(cfs_list_empty(&o->lo_linkage));
 
         if (dev != NULL) {
                 lu_ref_del_at(&dev->ld_reference,
@@ -903,7 +1066,7 @@ EXPORT_SYMBOL(lu_object_fini);
  */
 void lu_object_add_top(struct lu_object_header *h, struct lu_object *o)
 {
-        list_move(&o->lo_linkage, &h->loh_layers);
+        cfs_list_move(&o->lo_linkage, &h->loh_layers);
 }
 EXPORT_SYMBOL(lu_object_add_top);
 
@@ -915,7 +1078,7 @@ EXPORT_SYMBOL(lu_object_add_top);
  */
 void lu_object_add(struct lu_object *before, struct lu_object *o)
 {
-        list_move(&o->lo_linkage, &before->lo_linkage);
+        cfs_list_move(&o->lo_linkage, &before->lo_linkage);
 }
 EXPORT_SYMBOL(lu_object_add);
 
@@ -925,8 +1088,8 @@ EXPORT_SYMBOL(lu_object_add);
 int lu_object_header_init(struct lu_object_header *h)
 {
         memset(h, 0, sizeof *h);
-        atomic_set(&h->loh_ref, 1);
-        INIT_HLIST_NODE(&h->loh_hash);
+        cfs_atomic_set(&h->loh_ref, 1);
+        CFS_INIT_HLIST_NODE(&h->loh_hash);
         CFS_INIT_LIST_HEAD(&h->loh_lru);
         CFS_INIT_LIST_HEAD(&h->loh_layers);
         lu_ref_init(&h->loh_reference);
@@ -939,9 +1102,9 @@ EXPORT_SYMBOL(lu_object_header_init);
  */
 void lu_object_header_fini(struct lu_object_header *h)
 {
-        LASSERT(list_empty(&h->loh_layers));
-        LASSERT(list_empty(&h->loh_lru));
-        LASSERT(hlist_unhashed(&h->loh_hash));
+        LASSERT(cfs_list_empty(&h->loh_layers));
+        LASSERT(cfs_list_empty(&h->loh_lru));
+        LASSERT(cfs_hlist_unhashed(&h->loh_hash));
         lu_ref_fini(&h->loh_reference);
 }
 EXPORT_SYMBOL(lu_object_header_fini);
@@ -955,7 +1118,7 @@ struct lu_object *lu_object_locate(struct lu_object_header *h,
 {
         struct lu_object *o;
 
-        list_for_each_entry(o, &h->loh_layers, lo_linkage) {
+        cfs_list_for_each_entry(o, &h->loh_layers, lo_linkage) {
                 if (o->lo_dev->ld_type == dtype)
                         return o;
         }
@@ -988,7 +1151,7 @@ void lu_stack_fini(const struct lu_env *env, struct lu_device *top)
         /* purge again. */
         lu_site_purge(env, site, ~0);
 
-        if (!list_empty(&site->ls_lru) || site->ls_total != 0) {
+        if (!cfs_hash_is_empty(site->ls_obj_hash)) {
                 /*
                  * Uh-oh, objects still exist.
                  */
@@ -1020,7 +1183,7 @@ enum {
 
 static struct lu_context_key *lu_keys[LU_CONTEXT_KEY_NR] = { NULL, };
 
-static spinlock_t lu_keys_guard = SPIN_LOCK_UNLOCKED;
+static cfs_spinlock_t lu_keys_guard = CFS_SPIN_LOCK_UNLOCKED;
 
 /**
  * Global counter incremented whenever key is registered, unregistered,
@@ -1044,11 +1207,11 @@ int lu_context_key_register(struct lu_context_key *key)
         LASSERT(key->lct_owner != NULL);
 
         result = -ENFILE;
-        spin_lock(&lu_keys_guard);
+        cfs_spin_lock(&lu_keys_guard);
         for (i = 0; i < ARRAY_SIZE(lu_keys); ++i) {
                 if (lu_keys[i] == NULL) {
                         key->lct_index = i;
-                        atomic_set(&key->lct_used, 1);
+                        cfs_atomic_set(&key->lct_used, 1);
                         lu_keys[i] = key;
                         lu_ref_init(&key->lct_reference);
                         result = 0;
@@ -1056,7 +1219,7 @@ int lu_context_key_register(struct lu_context_key *key)
                         break;
                 }
         }
-        spin_unlock(&lu_keys_guard);
+        cfs_spin_unlock(&lu_keys_guard);
         return result;
 }
 EXPORT_SYMBOL(lu_context_key_register);
@@ -1069,15 +1232,15 @@ static void key_fini(struct lu_context *ctx, int index)
                 key = lu_keys[index];
                 LASSERT(key != NULL);
                 LASSERT(key->lct_fini != NULL);
-                LASSERT(atomic_read(&key->lct_used) > 1);
+                LASSERT(cfs_atomic_read(&key->lct_used) > 1);
 
                 key->lct_fini(ctx, key, ctx->lc_value[index]);
                 lu_ref_del(&key->lct_reference, "ctx", ctx);
-                atomic_dec(&key->lct_used);
+                cfs_atomic_dec(&key->lct_used);
                 LASSERT(key->lct_owner != NULL);
                 if (!(ctx->lc_tags & LCT_NOREF)) {
-                        LASSERT(module_refcount(key->lct_owner) > 0);
-                        module_put(key->lct_owner);
+                        LASSERT(cfs_module_refcount(key->lct_owner) > 0);
+                        cfs_module_put(key->lct_owner);
                 }
                 ctx->lc_value[index] = NULL;
         }
@@ -1088,19 +1251,23 @@ static void key_fini(struct lu_context *ctx, int index)
  */
 void lu_context_key_degister(struct lu_context_key *key)
 {
-        LASSERT(atomic_read(&key->lct_used) >= 1);
+        LASSERT(cfs_atomic_read(&key->lct_used) >= 1);
         LINVRNT(0 <= key->lct_index && key->lct_index < ARRAY_SIZE(lu_keys));
 
         lu_context_key_quiesce(key);
 
         ++key_set_version;
+        cfs_spin_lock(&lu_keys_guard);
         key_fini(&lu_shrink_env.le_ctx, key->lct_index);
+        if (lu_keys[key->lct_index]) {
+                lu_keys[key->lct_index] = NULL;
+                lu_ref_fini(&key->lct_reference);
+        }
+        cfs_spin_unlock(&lu_keys_guard);
 
-        if (atomic_read(&key->lct_used) > 1)
-                CERROR("key has instances.\n");
-        spin_lock(&lu_keys_guard);
-        lu_keys[key->lct_index] = NULL;
-        spin_unlock(&lu_keys_guard);
+        LASSERTF(cfs_atomic_read(&key->lct_used) == 1,
+                 "key has instances: %d\n",
+                 cfs_atomic_read(&key->lct_used));
 }
 EXPORT_SYMBOL(lu_context_key_degister);
 
@@ -1222,10 +1389,11 @@ void lu_context_key_quiesce(struct lu_context_key *key)
                 /*
                  * XXX memory barrier has to go here.
                  */
-                spin_lock(&lu_keys_guard);
-                list_for_each_entry(ctx, &lu_context_remembered, lc_remember)
+                cfs_spin_lock(&lu_keys_guard);
+                cfs_list_for_each_entry(ctx, &lu_context_remembered,
+                                        lc_remember)
                         key_fini(ctx, key->lct_index);
-                spin_unlock(&lu_keys_guard);
+                cfs_spin_unlock(&lu_keys_guard);
                 ++key_set_version;
         }
 }
@@ -1242,6 +1410,7 @@ static void keys_fini(struct lu_context *ctx)
 {
         int i;
 
+        cfs_spin_lock(&lu_keys_guard);
         if (ctx->lc_value != NULL) {
                 for (i = 0; i < ARRAY_SIZE(lu_keys); ++i)
                         key_fini(ctx, i);
@@ -1249,6 +1418,7 @@ static void keys_fini(struct lu_context *ctx)
                          ARRAY_SIZE(lu_keys) * sizeof ctx->lc_value[0]);
                 ctx->lc_value = NULL;
         }
+        cfs_spin_unlock(&lu_keys_guard);
 }
 
 static int keys_fill(struct lu_context *ctx)
@@ -1277,9 +1447,9 @@ static int keys_fill(struct lu_context *ctx)
 
                         LASSERT(key->lct_owner != NULL);
                         if (!(ctx->lc_tags & LCT_NOREF))
-                                try_module_get(key->lct_owner);
+                                cfs_try_module_get(key->lct_owner);
                         lu_ref_add_atomic(&key->lct_reference, "ctx", ctx);
-                        atomic_inc(&key->lct_used);
+                        cfs_atomic_inc(&key->lct_used);
                         /*
                          * This is the only place in the code, where an
                          * element of ctx->lc_value[] array is set to non-NULL
@@ -1318,9 +1488,9 @@ int lu_context_init(struct lu_context *ctx, __u32 tags)
         ctx->lc_state = LCS_INITIALIZED;
         ctx->lc_tags = tags;
         if (tags & LCT_REMEMBER) {
-                spin_lock(&lu_keys_guard);
-                list_add(&ctx->lc_remember, &lu_context_remembered);
-                spin_unlock(&lu_keys_guard);
+                cfs_spin_lock(&lu_keys_guard);
+                cfs_list_add(&ctx->lc_remember, &lu_context_remembered);
+                cfs_spin_unlock(&lu_keys_guard);
         } else
                 CFS_INIT_LIST_HEAD(&ctx->lc_remember);
         return keys_init(ctx);
@@ -1335,9 +1505,9 @@ void lu_context_fini(struct lu_context *ctx)
         LINVRNT(ctx->lc_state == LCS_INITIALIZED || ctx->lc_state == LCS_LEFT);
         ctx->lc_state = LCS_FINALIZED;
         keys_fini(ctx);
-        spin_lock(&lu_keys_guard);
-        list_del_init(&ctx->lc_remember);
-        spin_unlock(&lu_keys_guard);
+        cfs_spin_lock(&lu_keys_guard);
+        cfs_list_del_init(&ctx->lc_remember);
+        cfs_spin_unlock(&lu_keys_guard);
 }
 EXPORT_SYMBOL(lu_context_fini);
 
@@ -1418,38 +1588,83 @@ int lu_env_refill(struct lu_env *env)
 }
 EXPORT_SYMBOL(lu_env_refill);
 
-static struct shrinker *lu_site_shrinker = NULL;
+static struct cfs_shrinker *lu_site_shrinker = NULL;
+
+typedef struct lu_site_stats{
+        unsigned        lss_populated;
+        unsigned        lss_max_search;
+        unsigned        lss_total;
+        unsigned        lss_busy;
+} lu_site_stats_t;
+
+static void lu_site_stats_get(cfs_hash_t *hs,
+                              lu_site_stats_t *stats, int populated)
+{
+        cfs_hash_bd_t bd;
+        int           i;
+
+        cfs_hash_for_each_bucket(hs, &bd, i) {
+                struct lu_site_bkt_data *bkt = cfs_hash_bd_extra_get(hs, &bd);
+                cfs_hlist_head_t        *hhead;
+
+                cfs_hash_bd_lock(hs, &bd, 1);
+                stats->lss_busy  += bkt->lsb_busy;
+                stats->lss_total += cfs_hash_bd_count_get(&bd);
+                stats->lss_max_search = max((int)stats->lss_max_search,
+                                            cfs_hash_bd_depmax_get(&bd));
+                if (!populated) {
+                        cfs_hash_bd_unlock(hs, &bd, 1);
+                        continue;
+                }
+
+                cfs_hash_bd_for_each_hlist(hs, &bd, hhead) {
+                        if (!cfs_hlist_empty(hhead))
+                                stats->lss_populated++;
+                }
+                cfs_hash_bd_unlock(hs, &bd, 1);
+        }
+}
 
 #ifdef __KERNEL__
-static int lu_cache_shrink(int nr, unsigned int gfp_mask)
+static int lu_cache_shrink(SHRINKER_FIRST_ARG int nr_to_scan,
+                           unsigned int gfp_mask)
 {
+        lu_site_stats_t stats;
         struct lu_site *s;
         struct lu_site *tmp;
         int cached = 0;
-        int remain = nr;
+        int remain = nr_to_scan;
         CFS_LIST_HEAD(splice);
 
-        if (nr != 0 && !(gfp_mask & __GFP_FS))
-                return -1;
+        if (nr_to_scan != 0) {
+                if (!(gfp_mask & __GFP_FS))
+                        return -1;
+                CDEBUG(D_INODE, "Shrink %d objects\n", nr_to_scan);
+        }
 
-        down(&lu_sites_guard);
-        list_for_each_entry_safe(s, tmp, &lu_sites, ls_linkage) {
-                if (nr != 0) {
+        cfs_down(&lu_sites_guard);
+        cfs_list_for_each_entry_safe(s, tmp, &lu_sites, ls_linkage) {
+                if (nr_to_scan != 0) {
                         remain = lu_site_purge(&lu_shrink_env, s, remain);
                         /*
                          * Move just shrunk site to the tail of site list to
                          * assure shrinking fairness.
                          */
-                        list_move_tail(&s->ls_linkage, &splice);
+                        cfs_list_move_tail(&s->ls_linkage, &splice);
                 }
-                read_lock(&s->ls_guard);
-                cached += s->ls_total - s->ls_busy;
-                read_unlock(&s->ls_guard);
-                if (remain <= 0)
+
+                memset(&stats, 0, sizeof(stats));
+                lu_site_stats_get(s->ls_obj_hash, &stats, 0);
+                cached += stats.lss_total - stats.lss_busy;
+                if (nr_to_scan && remain <= 0)
                         break;
         }
-        list_splice(&splice, lu_sites.prev);
-        up(&lu_sites_guard);
+        cfs_list_splice(&splice, lu_sites.prev);
+        cfs_up(&lu_sites_guard);
+
+        cached = (cached / 100) * sysctl_vfs_cache_pressure;
+        if (nr_to_scan == 0)
+                CDEBUG(D_INODE, "%d objects cached\n", cached);
         return cached;
 }
 
@@ -1466,7 +1681,7 @@ struct lu_env lu_debugging_env;
  * Debugging printer function using printk().
  */
 int lu_printk_printer(const struct lu_env *env,
-                      void *_, const char *format, ...)
+                      void *unused, const char *format, ...)
 {
         va_list args;
 
@@ -1490,10 +1705,10 @@ void lu_context_keys_dump(void)
 
                 key = lu_keys[i];
                 if (key != NULL) {
-                        CERROR("[%i]: %p %x (%p,%p,%p) %i %i \"%s\"@%p\n",
+                        CERROR("[%d]: %p %x (%p,%p,%p) %d %d \"%s\"@%p\n",
                                i, key, key->lct_tags,
                                key->lct_init, key->lct_fini, key->lct_exit,
-                               key->lct_index, atomic_read(&key->lct_used),
+                               key->lct_index, cfs_atomic_read(&key->lct_used),
                                key->lct_owner ? key->lct_owner->name : "",
                                key->lct_owner);
                         lu_ref_print(&key->lct_reference);
@@ -1528,6 +1743,10 @@ int lu_global_init(void)
 
         CDEBUG(D_CONSOLE, "Lustre LU module (%p).\n", &lu_keys);
 
+        result = lu_ref_global_init();
+        if (result != 0)
+                return result;
+
         LU_CONTEXT_KEY_INIT(&lu_global_key);
         result = lu_context_key_register(&lu_global_key);
         if (result != 0)
@@ -1537,21 +1756,18 @@ int lu_global_init(void)
          * conservatively. This should not be too bad, because this
          * environment is global.
          */
-        down(&lu_sites_guard);
+        cfs_down(&lu_sites_guard);
         result = lu_env_init(&lu_shrink_env, LCT_SHRINKER);
-        up(&lu_sites_guard);
+        cfs_up(&lu_sites_guard);
         if (result != 0)
                 return result;
 
-        result = lu_ref_global_init();
-        if (result != 0)
-                return result;
         /*
          * seeks estimation: 3 seeks to read a record from oi, one to read
          * inode, one for ea. Unfortunately setting this high value results in
          * lu_object/inode cache consuming all the memory.
          */
-        lu_site_shrinker = set_shrinker(DEFAULT_SEEKS, lu_cache_shrink);
+        lu_site_shrinker = cfs_set_shrinker(CFS_DEFAULT_SEEKS, lu_cache_shrink);
         if (lu_site_shrinker == NULL)
                 return -ENOMEM;
 
@@ -1586,7 +1802,7 @@ void lu_global_fini(void)
 #endif
         lu_time_global_fini();
         if (lu_site_shrinker != NULL) {
-                remove_shrinker(lu_site_shrinker);
+                cfs_remove_shrinker(lu_site_shrinker);
                 lu_site_shrinker = NULL;
         }
 
@@ -1596,9 +1812,9 @@ void lu_global_fini(void)
          * Tear shrinker environment down _after_ de-registering
          * lu_global_key, because the latter has a value in the former.
          */
-        down(&lu_sites_guard);
+        cfs_down(&lu_sites_guard);
         lu_env_fini(&lu_shrink_env);
-        up(&lu_sites_guard);
+        cfs_up(&lu_sites_guard);
 
         lu_ref_global_fini();
 }
@@ -1609,119 +1825,43 @@ struct lu_buf LU_BUF_NULL = {
 };
 EXPORT_SYMBOL(LU_BUF_NULL);
 
-/**
- * Output site statistical counters into a buffer. Suitable for
- * lprocfs_rd_*()-style functions.
- */
-int lu_site_stats_print(const struct lu_site *s, char *page, int count)
+static __u32 ls_stats_read(struct lprocfs_stats *stats, int idx)
 {
-        int i;
-        int populated;
+#ifdef LPROCFS
+        struct lprocfs_counter ret;
 
-        /*
-         * How many hash buckets are not-empty? Don't bother with locks: it's
-         * an estimation anyway.
-         */
-        for (i = 0, populated = 0; i < s->ls_hash_size; i++)
-                populated += !hlist_empty(&s->ls_hash[i]);
-
-        return snprintf(page, count, "%d %d %d/%d %d %d %d %d %d %d %d\n",
-                        s->ls_total,
-                        s->ls_busy,
-                        populated,
-                        s->ls_hash_size,
-                        s->ls_stats.s_created,
-                        s->ls_stats.s_cache_hit,
-                        s->ls_stats.s_cache_miss,
-                        s->ls_stats.s_cache_check,
-                        s->ls_stats.s_cache_race,
-                        s->ls_stats.s_cache_death_race,
-                        s->ls_stats.s_lru_purged);
+        lprocfs_stats_collect(stats, idx, &ret);
+        return (__u32)ret.lc_count;
+#else
+        return 0;
+#endif
 }
-EXPORT_SYMBOL(lu_site_stats_print);
 
-#ifdef __KERNEL__
-/*
- * XXX: Functions below logically belong to the fid module, but they are used
- * by dt_store_open(). Put them here until better place is found.
+/**
+ * Output site statistical counters into a buffer. Suitable for
+ * lprocfs_rd_*()-style functions.
  */
-
-void fid_pack(struct lu_fid_pack *pack, const struct lu_fid *fid,
-              struct lu_fid *befider)
+int lu_site_stats_print(const struct lu_site *s, char *page, int count)
 {
-        int recsize;
-        __u64 seq;
-        __u32 oid;
-
-        seq = fid_seq(fid);
-        oid = fid_oid(fid);
-
-        /*
-         * Two cases: compact 6 bytes representation for a common case, and
-         * full 17 byte representation for "unusual" fid.
-         */
-
-        /*
-         * Check that usual case is really usual.
-         */
-        CLASSERT(LUSTRE_SEQ_MAX_WIDTH < 0xffffull);
-
-        /* fid can be packed in six bytes (first byte as length of packed fid,
-         * three bytes of seq and two bytes of oid).
-         * this reduces IO overhead specially for OSD Object Index. */
-
-        if (seq < FID_SEQ_START ||
-            seq > (0xffffffull + FID_SEQ_START) ||
-            oid > 0xffff || fid_ver(fid) != 0) {
-                fid_cpu_to_be(befider, fid);
-                recsize = sizeof *befider;
-        } else {
-                unsigned char *small_befider;
+        lu_site_stats_t stats;
 
-                small_befider = (unsigned char *)befider;
+        memset(&stats, 0, sizeof(stats));
+        lu_site_stats_get(s->ls_obj_hash, &stats, 1);
 
-                small_befider[0] = seq >> 16;
-                small_befider[1] = seq >> 8;
-                small_befider[2] = seq;
-
-                small_befider[3] = oid >> 8;
-                small_befider[4] = oid;
-
-                recsize = 5;
-        }
-        memcpy(pack->fp_area, befider, recsize);
-        pack->fp_len = recsize + 1;
-}
-EXPORT_SYMBOL(fid_pack);
-
-int fid_unpack(const struct lu_fid_pack *pack, struct lu_fid *fid)
-{
-        int result;
-
-        result = 0;
-        switch (pack->fp_len) {
-        case sizeof *fid + 1:
-                memcpy(fid, pack->fp_area, sizeof *fid);
-                fid_be_to_cpu(fid, fid);
-                break;
-        case 6: {
-                const unsigned char *area;
-
-                area = (unsigned char *)pack->fp_area;
-                fid->f_seq = (area[0] << 16) | (area[1] << 8) | area[2];
-                fid->f_seq +=  FID_SEQ_START;
-                fid->f_oid = (area[3] << 8) | area[4];
-                fid->f_ver = 0;
-                break;
-        }
-        default:
-                CERROR("Unexpected packed fid size: %d\n", pack->fp_len);
-                result = -EIO;
-        }
-        return result;
+        return snprintf(page, count, "%d/%d %d/%d %d %d %d %d %d %d %d\n",
+                        stats.lss_busy,
+                        stats.lss_total,
+                        stats.lss_populated,
+                        CFS_HASH_NHLIST(s->ls_obj_hash),
+                        stats.lss_max_search,
+                        ls_stats_read(s->ls_stats, LU_SS_CREATED),
+                        ls_stats_read(s->ls_stats, LU_SS_CACHE_HIT),
+                        ls_stats_read(s->ls_stats, LU_SS_CACHE_MISS),
+                        ls_stats_read(s->ls_stats, LU_SS_CACHE_RACE),
+                        ls_stats_read(s->ls_stats, LU_SS_CACHE_DEATH_RACE),
+                        ls_stats_read(s->ls_stats, LU_SS_LRU_PURGED));
 }
-EXPORT_SYMBOL(fid_unpack);
-#endif  /* #ifdef __KERNEL__ */
+EXPORT_SYMBOL(lu_site_stats_print);
 
 const char *lu_time_names[LU_TIME_NR] = {
         [LU_TIME_FIND_LOOKUP] = "find_lookup",