Whamcloud - gitweb
Branch HEAD
[fs/lustre-release.git] / lustre / obdclass / lprocfs_status.c
index 608af88..8787a3e 100644 (file)
@@ -1,26 +1,41 @@
 /* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
  * vim:expandtab:shiftwidth=8:tabstop=8:
  *
- *  Copyright (C) 2002, 2003 Cluster File Systems, Inc.
- *   Author: Hariharan Thantry <thantry@users.sourceforge.net>
+ * GPL HEADER START
  *
- *   This file is part of the Lustre file system, http://www.lustre.org
- *   Lustre is a trademark of Cluster File Systems, Inc.
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
  *
- *   You may have signed or agreed to another license before downloading
- *   this software.  If so, you are bound by the terms and conditions
- *   of that agreement, and the following does not apply to you.  See the
- *   LICENSE file included with this distribution for more information.
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 only,
+ * as published by the Free Software Foundation.
  *
- *   If you did not agree to a different license, then this copy of Lustre
- *   is open source software; you can redistribute it and/or modify it
- *   under the terms of version 2 of the GNU General Public License as
- *   published by the Free Software Foundation.
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License version 2 for more details (a copy is included
+ * in the LICENSE file that accompanied this code).
  *
- *   In either case, Lustre is distributed in the hope that it will be
- *   useful, but WITHOUT ANY WARRANTY; without even the implied warranty
- *   of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- *   license text for more details.
+ * You should have received a copy of the GNU General Public License
+ * version 2 along with this program; If not, see
+ * http://www.sun.com/software/products/lustre/docs/GPLv2.pdf
+ *
+ * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
+ * CA 95054 USA or visit www.sun.com if you need additional information or
+ * have any questions.
+ *
+ * GPL HEADER END
+ */
+/*
+ * Copyright  2008 Sun Microsystems, Inc. All rights reserved
+ * Use is subject to license terms.
+ */
+/*
+ * This file is part of Lustre, http://www.lustre.org/
+ * Lustre is a trademark of Sun Microsystems, Inc.
+ *
+ * lustre/obdclass/lprocfs_status.c
+ *
+ * Author: Hariharan Thantry <thantry@users.sourceforge.net>
  */
 
 #ifndef EXPORT_SYMTAB
@@ -58,14 +73,18 @@ struct proc_dir_entry *lprocfs_srch(struct proc_dir_entry *head,
 
         if (head == NULL)
                 return NULL;
+        LPROCFS_ENTRY();
 
         temp = head->subdir;
         while (temp != NULL) {
-                if (strcmp(temp->name, name) == 0)
+                if (strcmp(temp->name, name) == 0) {
+                        LPROCFS_EXIT();
                         return temp;
+                }
 
                 temp = temp->next;
         }
+        LPROCFS_EXIT();
         return NULL;
 }
 
@@ -92,28 +111,35 @@ static int lprocfs_obd_snprintf(char **page, int end, int *len,
         return n;
 }
 
-int lprocfs_add_simple(struct proc_dir_entry *root, char *name,
-                       read_proc_t *read_proc, write_proc_t *write_proc,
-                       void *data)
+cfs_proc_dir_entry_t *lprocfs_add_simple(struct proc_dir_entry *root,
+                                         char *name,
+                                         read_proc_t *read_proc,
+                                         write_proc_t *write_proc,
+                                         void *data,
+                                         struct file_operations *fops)
 {
-        struct proc_dir_entry *proc;
+        cfs_proc_dir_entry_t *proc;
         mode_t mode = 0;
-        
+
         if (root == NULL || name == NULL)
-                return -EINVAL;
+                return ERR_PTR(-EINVAL);
         if (read_proc)
                 mode = 0444;
         if (write_proc)
                 mode |= 0200;
+        if (fops)
+                mode = 0644;
         proc = create_proc_entry(name, mode, root);
         if (!proc) {
                 CERROR("LprocFS: No memory to create /proc entry %s", name);
-                return -ENOMEM;
+                return ERR_PTR(-ENOMEM);
         }
         proc->read_proc = read_proc;
         proc->write_proc = write_proc;
         proc->data = data;
-        return 0;
+        if (fops)
+                proc->proc_fops = fops;
+        return proc;
 }
 
 struct proc_dir_entry *lprocfs_add_symlink(const char *name,
@@ -148,7 +174,7 @@ static ssize_t lprocfs_fops_read(struct file *f, char __user *buf,
         LPROCFS_ENTRY();
         OBD_FAIL_TIMEOUT(OBD_FAIL_LPROC_REMOVE, 10);
         if (!dp->deleted && dp->read_proc)
-                rc = dp->read_proc(page, &start, *ppos, CFS_PAGE_SIZE, 
+                rc = dp->read_proc(page, &start, *ppos, CFS_PAGE_SIZE,
                         &eof, dp->data);
         LPROCFS_EXIT();
         if (rc <= 0)
@@ -183,7 +209,7 @@ out:
 static ssize_t lprocfs_fops_write(struct file *f, const char __user *buf, size_t size, loff_t *ppos)
 {
         struct proc_dir_entry *dp = PDE(f->f_dentry->d_inode);
-        int rc = 0;
+        int rc = -EIO;
 
         LPROCFS_ENTRY();
         if (!dp->deleted && dp->write_proc)
@@ -228,6 +254,17 @@ struct file_operations lprocfs_evict_client_fops = {
 };
 EXPORT_SYMBOL(lprocfs_evict_client_fops);
 
+/**
+ * Add /proc entries.
+ *
+ * \param root [in]  The parent proc entry on which new entry will be added.
+ * \param list [in]  Array of proc entries to be added.
+ * \param data [in]  The argument to be passed when entries read/write routines
+ *                   are called through /proc file.
+ *
+ * \retval 0   on success
+ *         < 0 on error
+ */
 int lprocfs_add_vars(struct proc_dir_entry *root, struct lprocfs_vars *list,
                      void *data)
 {
@@ -267,10 +304,14 @@ int lprocfs_add_vars(struct proc_dir_entry *root, struct lprocfs_vars *list,
                                             proc_mkdir(cur, cur_root));
                         } else if (proc == NULL) {
                                 mode_t mode = 0;
-                                if (list->read_fptr)
-                                        mode = 0444;
-                                if (list->write_fptr)
-                                        mode |= 0200;
+                                if (list->proc_mode != 0000) {
+                                        mode = list->proc_mode;
+                                } else {
+                                        if (list->read_fptr)
+                                                mode = 0444;
+                                        if (list->write_fptr)
+                                                mode |= 0200;
+                                }
                                 proc = create_proc_entry(cur, mode, cur_root);
                         }
                 }
@@ -309,6 +350,7 @@ void lprocfs_remove(struct proc_dir_entry **rooth)
 
         parent = root->parent;
         LASSERT(parent != NULL);
+        LPROCFS_WRITE_ENTRY(); /* search vs remove race */
 
         while (1) {
                 while (temp->subdir != NULL)
@@ -323,15 +365,14 @@ void lprocfs_remove(struct proc_dir_entry **rooth)
                          "0x%p  %s/%s len %d\n", rm_entry, temp->name,
                          rm_entry->name, (int)strlen(rm_entry->name));
 
-                /* Now, the rm_entry->deleted flags is protected 
+                /* Now, the rm_entry->deleted flags is protected
                  * by _lprocfs_lock. */
-                down_write(&_lprocfs_lock);
                 rm_entry->data = NULL;
                 remove_proc_entry(rm_entry->name, temp);
-                up_write(&_lprocfs_lock);
                 if (temp == parent)
                         break;
         }
+        LPROCFS_WRITE_EXIT();
 }
 
 void lprocfs_remove_proc_entry(const char *name, struct proc_dir_entry *parent)
@@ -414,14 +455,14 @@ int lprocfs_wr_atomic(struct file *file, const char *buffer,
         atomic_t *atm = data;
         int val = 0;
         int rc;
-        
+
         rc = lprocfs_write_helper(buffer, count, &val);
         if (rc < 0)
                 return rc;
 
         if (val <= 0)
                 return -ERANGE;
-                
+
         atomic_set(atm, val);
         return count;
 }
@@ -600,6 +641,137 @@ int lprocfs_rd_conn_uuid(char *page, char **start, off_t off, int count,
         return rc;
 }
 
+#define flag2str(flag) \
+        if (imp->imp_##flag && max - len > 0) \
+                len += snprintf(str + len, max - len, " " #flag);
+        
+/**
+ * Append a space separated list of current set flags to str.
+ */
+static int obd_import_flags2str(struct obd_import *imp, char *str, 
+                                          int max)
+{
+        int len = 0;
+        
+        if (imp->imp_obd->obd_no_recov)
+                len += snprintf(str, max - len, " no_recov");
+
+        flag2str(invalid);
+        flag2str(deactive);
+        flag2str(replayable);
+        flag2str(pingable);
+        flag2str(recon_bk);
+        flag2str(last_recon);
+        return len;
+}
+#undef flags2str
+
+int lprocfs_rd_import(char *page, char **start, off_t off, int count,
+                      int *eof, void *data)
+{
+        struct obd_device *obd = (struct obd_device *)data;
+        struct obd_import *imp;
+        char *imp_state_name = NULL;
+        int rc = 0;
+
+        LASSERT(obd != NULL);
+        LPROCFS_CLIMP_CHECK(obd);
+        imp = obd->u.cli.cl_import;
+        imp_state_name = ptlrpc_import_state_name(imp->imp_state);
+        *eof = 1;
+
+        rc = snprintf(page, count, 
+                      "import: %s\n"
+                      "    target: %s@%s\n"
+                      "    state: %s\n"
+                      "    inflight: %u\n"
+                      "    conn_cnt: %u\n"
+                      "    generation: %u\n"
+                      "    inval_cnt: %u\n"
+                      "    last_replay_transno: "LPU64"\n"
+                      "    peer_committed_transno: "LPU64"\n"
+                      "    last_trasno_checked: "LPU64"\n"
+                      "    flags:",
+                      obd->obd_name,
+                      obd2cli_tgt(obd), imp->imp_connection->c_remote_uuid.uuid,
+                      imp_state_name,
+                      atomic_read(&imp->imp_inflight),
+                      imp->imp_conn_cnt,
+                      imp->imp_generation,
+                      atomic_read(&imp->imp_inval_count),
+                      imp->imp_last_replay_transno,
+                      imp->imp_peer_committed_transno,
+                      imp->imp_last_transno_checked);
+        rc += obd_import_flags2str(imp, page + rc, count - rc);
+        rc += snprintf(page+rc, count - rc, "\n");
+        LPROCFS_CLIMP_EXIT(obd);
+        return rc;
+}
+
+int lprocfs_at_hist_helper(char *page, int count, int rc,
+                           struct adaptive_timeout *at)
+{
+        int i;
+        for (i = 0; i < AT_BINS; i++)
+                rc += snprintf(page + rc, count - rc, "%3u ", at->at_hist[i]);
+        rc += snprintf(page + rc, count - rc, "\n");
+        return rc;
+}
+
+/* See also ptlrpc_lprocfs_rd_timeouts */
+int lprocfs_rd_timeouts(char *page, char **start, off_t off, int count,
+                        int *eof, void *data)
+{
+        struct obd_device *obd = (struct obd_device *)data;
+        struct obd_import *imp;
+        unsigned int cur, worst;
+        time_t now, worstt;
+        struct dhms ts;
+        int i, rc = 0;
+
+        LASSERT(obd != NULL);
+        LPROCFS_CLIMP_CHECK(obd);
+        imp = obd->u.cli.cl_import;
+        *eof = 1;
+
+        now = cfs_time_current_sec();
+
+        /* Some network health info for kicks */
+        s2dhms(&ts, now - imp->imp_last_reply_time);
+        rc += snprintf(page + rc, count - rc,
+                       "%-10s : %ld, "DHMS_FMT" ago\n",
+                       "last reply", imp->imp_last_reply_time, DHMS_VARS(&ts));
+
+
+        cur = at_get(&imp->imp_at.iat_net_latency);
+        worst = imp->imp_at.iat_net_latency.at_worst_ever;
+        worstt = imp->imp_at.iat_net_latency.at_worst_time;
+        s2dhms(&ts, now - worstt);
+        rc += snprintf(page + rc, count - rc,
+                       "%-10s : cur %3u  worst %3u (at %ld, "DHMS_FMT" ago) ",
+                       "network", cur, worst, worstt, DHMS_VARS(&ts));
+        rc = lprocfs_at_hist_helper(page, count, rc,
+                                    &imp->imp_at.iat_net_latency);
+
+        for(i = 0; i < IMP_AT_MAX_PORTALS; i++) {
+                if (imp->imp_at.iat_portal[i] == 0)
+                        break;
+                cur = at_get(&imp->imp_at.iat_service_estimate[i]);
+                worst = imp->imp_at.iat_service_estimate[i].at_worst_ever;
+                worstt = imp->imp_at.iat_service_estimate[i].at_worst_time;
+                s2dhms(&ts, now - worstt);
+                rc += snprintf(page + rc, count - rc,
+                               "portal %-2d  : cur %3u  worst %3u (at %ld, "
+                               DHMS_FMT" ago) ", imp->imp_at.iat_portal[i],
+                               cur, worst, worstt, DHMS_VARS(&ts));
+                rc = lprocfs_at_hist_helper(page, count, rc,
+                                          &imp->imp_at.iat_service_estimate[i]);
+        }
+
+        LPROCFS_CLIMP_EXIT(obd);
+        return rc;
+}
+
 static const char *obd_connect_names[] = {
         "read_only",
         "lov_index",
@@ -632,6 +804,8 @@ static const char *obd_connect_names[] = {
         "change_qunit_size",
         "alt_checksum_algorithm",
         "fid_is_enabled",
+        "version_recovery",
+        "pools",
         NULL
 };
 
@@ -700,7 +874,7 @@ int lprocfs_obd_setup(struct obd_device *obd, struct lprocfs_vars *list)
 
 int lprocfs_obd_cleanup(struct obd_device *obd)
 {
-        if (!obd) 
+        if (!obd)
                 return -EINVAL;
         if (obd->obd_proc_exports_entry) {
                 /* Should be no exports left */
@@ -711,10 +885,8 @@ int lprocfs_obd_cleanup(struct obd_device *obd)
         return 0;
 }
 
-void lprocfs_free_client_stats(void *obj, void *data)
+static void lprocfs_free_client_stats(struct nid_stat *client_stat)
 {
-        struct nid_stat *client_stat = obj;
-
         CDEBUG(D_CONFIG, "stat %p - data %p/%p/%p\n", client_stat,
                client_stat->nid_proc, client_stat->nid_stats,
                client_stat->nid_brw_stats);
@@ -723,7 +895,6 @@ void lprocfs_free_client_stats(void *obj, void *data)
                  client_stat->nid_exp_ref_count);
 
         hlist_del_init(&client_stat->nid_hash);
-        list_del(&client_stat->nid_list);
 
         if (client_stat->nid_proc)
                 lprocfs_remove(&client_stat->nid_proc);
@@ -732,9 +903,12 @@ void lprocfs_free_client_stats(void *obj, void *data)
                 lprocfs_free_stats(&client_stat->nid_stats);
 
         if (client_stat->nid_brw_stats)
-                OBD_FREE(client_stat->nid_brw_stats, sizeof(struct brw_stats));
+                OBD_FREE_PTR(client_stat->nid_brw_stats);
+
+        if (client_stat->nid_ldlm_stats)
+                lprocfs_free_stats(&client_stat->nid_ldlm_stats);
 
-        OBD_FREE(client_stat, sizeof(*client_stat));
+        OBD_FREE_PTR(client_stat);
         return;
 
 }
@@ -745,10 +919,12 @@ void lprocfs_free_per_client_stats(struct obd_device *obd)
         ENTRY;
 
         /* we need extra list - because hash_exit called to early */
+        /* not need locking because all clients is died */
         while(!list_empty(&obd->obd_nid_stats)) {
                 stat = list_entry(obd->obd_nid_stats.next,
                                   struct nid_stat, nid_list);
-                lprocfs_free_client_stats(stat, NULL);
+                list_del_init(&stat->nid_list);
+                lprocfs_free_client_stats(stat);
         }
 
         EXIT;
@@ -891,7 +1067,7 @@ static int lprocfs_stats_seq_show(struct seq_file *p, void *v)
        struct lprocfs_stats *stats = p->private;
        struct lprocfs_counter  *cntr = v;
        struct lprocfs_counter  t, ret = { .lc_min = LC_MIN_INIT };
-       int i, idx, rc;
+       int i, idx, rc = 0;
        unsigned int num_cpu;
 
        if (cntr == &(stats->ls_percpu[0])->lp_cntr[0]) {
@@ -932,6 +1108,9 @@ static int lprocfs_stats_seq_show(struct seq_file *p, void *v)
                ret.lc_sumsquare += t.lc_sumsquare;
        }
 
+       if (ret.lc_count == 0)
+               goto out;
+
        rc = seq_printf(p, "%-25s "LPD64" samples [%s]", cntr->lc_name,
                        ret.lc_count, cntr->lc_units);
        if (rc < 0)
@@ -1069,6 +1248,8 @@ void lprocfs_init_ops_stats(int num_private_stats, struct lprocfs_stats *stats)
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, brw);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, brw_async);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, prep_async_page);
+        LPROCFS_OBD_OP_INIT(num_private_stats, stats, reget_short_lock);
+        LPROCFS_OBD_OP_INIT(num_private_stats, stats, release_short_lock);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, queue_async_io);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, queue_group_io);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, trigger_group_io);
@@ -1088,7 +1269,6 @@ void lprocfs_init_ops_stats(int num_private_stats, struct lprocfs_stats *stats)
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, change_cbdata);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, cancel);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, cancel_unused);
-        LPROCFS_OBD_OP_INIT(num_private_stats, stats, join_lru);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, init_export);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, destroy_export);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, extent_calc);
@@ -1100,9 +1280,18 @@ void lprocfs_init_ops_stats(int num_private_stats, struct lprocfs_stats *stats)
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, import_event);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, notify);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, health_check);
+        LPROCFS_OBD_OP_INIT(num_private_stats, stats, get_uuid);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, quotacheck);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, quotactl);
         LPROCFS_OBD_OP_INIT(num_private_stats, stats, ping);
+        LPROCFS_OBD_OP_INIT(num_private_stats, stats, register_page_removal_cb);
+        LPROCFS_OBD_OP_INIT(num_private_stats,stats,unregister_page_removal_cb);
+        LPROCFS_OBD_OP_INIT(num_private_stats, stats, register_lock_cancel_cb);
+        LPROCFS_OBD_OP_INIT(num_private_stats, stats,unregister_lock_cancel_cb);
+        LPROCFS_OBD_OP_INIT(num_private_stats, stats, pool_new);
+        LPROCFS_OBD_OP_INIT(num_private_stats, stats, pool_rem);
+        LPROCFS_OBD_OP_INIT(num_private_stats, stats, pool_add);
+        LPROCFS_OBD_OP_INIT(num_private_stats, stats, pool_del);
 }
 
 int lprocfs_alloc_obd_stats(struct obd_device *obd, unsigned num_private_stats)
@@ -1145,7 +1334,7 @@ int lprocfs_alloc_obd_stats(struct obd_device *obd, unsigned num_private_stats)
 
 void lprocfs_free_obd_stats(struct obd_device *obd)
 {
-        if (obd->obd_stats) 
+        if (obd->obd_stats)
                 lprocfs_free_stats(&obd->obd_stats);
 }
 
@@ -1167,7 +1356,7 @@ int lprocfs_alloc_md_stats(struct obd_device *obd,
         LASSERT(obd->obd_proc_entry != NULL);
         LASSERT(obd->md_cntr_base == 0);
 
-        num_stats = 1 + MD_COUNTER_OFFSET(get_remote_perm) +
+        num_stats = 1 + MD_COUNTER_OFFSET(revalidate_lock) +
                     num_private_stats;
         stats = lprocfs_alloc_stats(num_stats, 0);
         if (stats == NULL)
@@ -1201,6 +1390,8 @@ int lprocfs_alloc_md_stats(struct obd_device *obd,
         LPROCFS_MD_OP_INIT(num_private_stats, stats, cancel_unused);
         LPROCFS_MD_OP_INIT(num_private_stats, stats, renew_capa);
         LPROCFS_MD_OP_INIT(num_private_stats, stats, get_remote_perm);
+        LPROCFS_MD_OP_INIT(num_private_stats, stats, intent_getattr_async);
+        LPROCFS_MD_OP_INIT(num_private_stats, stats, revalidate_lock);
 
         for (i = num_private_stats; i < num_stats; i++) {
                 if (stats->ls_percpu[0]->lp_cntr[i].lc_name == NULL) {
@@ -1230,6 +1421,28 @@ void lprocfs_free_md_stats(struct obd_device *obd)
         }
 }
 
+void lprocfs_init_ldlm_stats(struct lprocfs_stats *ldlm_stats)
+{
+        lprocfs_counter_init(ldlm_stats,
+                             LDLM_ENQUEUE - LDLM_FIRST_OPC,
+                             0, "ldlm_enqueue", "reqs");
+        lprocfs_counter_init(ldlm_stats,
+                             LDLM_CONVERT - LDLM_FIRST_OPC,
+                             0, "ldlm_convert", "reqs");
+        lprocfs_counter_init(ldlm_stats,
+                             LDLM_CANCEL - LDLM_FIRST_OPC,
+                             0, "ldlm_cancel", "reqs");
+        lprocfs_counter_init(ldlm_stats,
+                             LDLM_BL_CALLBACK - LDLM_FIRST_OPC,
+                             0, "ldlm_bl_callback", "reqs");
+        lprocfs_counter_init(ldlm_stats,
+                             LDLM_CP_CALLBACK - LDLM_FIRST_OPC,
+                             0, "ldlm_cp_callback", "reqs");
+        lprocfs_counter_init(ldlm_stats,
+                             LDLM_GL_CALLBACK - LDLM_FIRST_OPC,
+                             0, "ldlm_gl_callback", "reqs");
+}
+
 int lprocfs_exp_rd_nid(char *page, char **start, off_t off, int count,
                          int *eof,  void *data)
 {
@@ -1246,6 +1459,16 @@ struct exp_uuid_cb_data {
         int                    *len;
 };
 
+static void
+lprocfs_exp_rd_cb_data_init(struct exp_uuid_cb_data *cb_data, char *page,
+                            int count, int *eof, int *len)
+{
+        cb_data->page = page;
+        cb_data->count = count;
+        cb_data->eof = eof;
+        cb_data->len = len;
+}
+
 void lprocfs_exp_print_uuid(void *obj, void *cb_data)
 {
         struct obd_export *exp = (struct obd_export *)obj;
@@ -1267,15 +1490,43 @@ int lprocfs_exp_rd_uuid(char *page, char **start, off_t off, int count,
 
         *eof = 1;
         page[0] = '\0';
-        LASSERT(obd != NULL);
+        lprocfs_exp_rd_cb_data_init(&cb_data, page, count, eof, &len);
+        lustre_hash_for_each_key(obd->obd_nid_hash, &stats->nid,
+                                 lprocfs_exp_print_uuid, &cb_data);
+        return (*cb_data.len);
+}
+
+void lprocfs_exp_print_hash(void *obj, void *cb_data)
+{
+        struct exp_uuid_cb_data *data = cb_data;
+        struct obd_export       *exp = obj;
+        lustre_hash_t           *lh;
+
+        lh = exp->exp_lock_hash;
+        if (lh) {
+                if (!*data->len)
+                        *data->len += lustre_hash_debug_header(data->page,
+                                                               data->count);
+
+                *data->len += lustre_hash_debug_str(lh, data->page + *data->len,
+                                                    data->count);
+        }
+}
+
+int lprocfs_exp_rd_hash(char *page, char **start, off_t off, int count,
+                        int *eof,  void *data)
+{
+        struct nid_stat *stats = (struct nid_stat *)data;
+        struct exp_uuid_cb_data cb_data;
+        struct obd_device *obd = stats->nid_obd;
+        int len = 0;
 
-        cb_data.page = page;
-        cb_data.count = count;
-        cb_data.eof = eof;
-        cb_data.len = &len;
-        lustre_hash_bucket_iterate(obd->obd_nid_hash_body,
-                                   &stats->nid, lprocfs_exp_print_uuid,
-                                   &cb_data);
+        *eof = 1;
+        page[0] = '\0';
+        lprocfs_exp_rd_cb_data_init(&cb_data, page, count, eof, &len);
+
+        lustre_hash_for_each_key(obd->obd_nid_hash, &stats->nid,
+                                 lprocfs_exp_print_hash, &cb_data);
         return (*cb_data.len);
 }
 
@@ -1291,23 +1542,28 @@ EXPORT_SYMBOL(lprocfs_nid_stats_clear_read);
 
 void lprocfs_nid_stats_clear_write_cb(void *obj, void *data)
 {
-        struct nid_stat *client_stat = obj;
+        struct nid_stat *stat = obj;
         int i;
-        if(client_stat->nid_exp_ref_count == 1) {
-                hlist_del_init(&client_stat->nid_hash);
-                lprocfs_free_client_stats(client_stat, data);
-                OBD_FREE(client_stat, sizeof(struct nid_stat));
+        ENTRY;
+        /* object has only hash + iterate_all references.
+         * add/delete blocked by hash bucket lock */
+        CDEBUG(D_INFO,"refcnt %d\n", stat->nid_exp_ref_count);
+        if (stat->nid_exp_ref_count == 2) {
+                hlist_del_init(&stat->nid_hash);
+                stat->nid_exp_ref_count--;
+                spin_lock(&stat->nid_obd->obd_nid_lock);
+                list_move(&stat->nid_list, data);
+                spin_unlock(&stat->nid_obd->obd_nid_lock);
                 EXIT;
                 return;
         }
-
         /* we has reference to object - only clear data*/
-        if (client_stat->nid_stats) {
-                lprocfs_clear_stats(client_stat->nid_stats);
-        }
-        if (client_stat->nid_brw_stats) {
+        if (stat->nid_stats)
+                lprocfs_clear_stats(stat->nid_stats);
+
+        if (stat->nid_brw_stats) {
                 for (i = 0; i < BRW_LAST; i++)
-                        lprocfs_oh_clear(&client_stat->nid_brw_stats->hist[i]);
+                        lprocfs_oh_clear(&stat->nid_brw_stats->hist[i]);
         }
         EXIT;
         return;
@@ -1317,92 +1573,185 @@ int lprocfs_nid_stats_clear_write(struct file *file, const char *buffer,
                                          unsigned long count, void *data)
 {
         struct obd_device *obd = (struct obd_device *)data;
-        lustre_hash_iterate_all(obd->obd_nid_stats_hash_body,
-                                lprocfs_free_client_stats, NULL);
+        struct nid_stat *client_stat;
+        CFS_LIST_HEAD(free_list);
+
+        lustre_hash_for_each(obd->obd_nid_stats_hash,
+                             lprocfs_nid_stats_clear_write_cb, &free_list);
+
+        while (!list_empty(&free_list)) {
+                client_stat = list_entry(free_list.next, struct nid_stat, nid_list);
+                list_del_init(&client_stat->nid_list);
+                lprocfs_free_client_stats(client_stat);
+        }
+
         return count;
 }
 EXPORT_SYMBOL(lprocfs_nid_stats_clear_write);
 
 int lprocfs_exp_setup(struct obd_export *exp, lnet_nid_t *nid, int *newnid)
 {
-        int rc = 0;
-        struct nid_stat *tmp = NULL, *tmp1;
+        struct nid_stat *new_stat, *old_stat;
+        struct nid_stat_uuid *new_ns_uuid;
         struct obd_device *obd = NULL;
+        cfs_proc_dir_entry_t *entry;
+        int rc = 0;
         ENTRY;
 
         *newnid = 0;
 
         if (!exp || !exp->exp_obd || !exp->exp_obd->obd_proc_exports_entry ||
-            !exp->exp_obd->obd_nid_stats_hash_body)
+            !exp->exp_obd->obd_nid_stats_hash)
                 RETURN(-EINVAL);
 
-       /* not test against zero because eric say:
-        * You may only test nid against another nid, or LNET_NID_ANY.  Anything else is
-        * nonsense.*/
+        /* not test against zero because eric say:
+         * You may only test nid against another nid, or LNET_NID_ANY.
+         * Anything else is nonsense.*/
         if (!nid || *nid == LNET_NID_ANY)
                 RETURN(0);
 
         obd = exp->exp_obd;
 
-        CDEBUG(D_CONFIG, "using hash %p\n", obd->obd_nid_stats_hash_body);
+        CDEBUG(D_CONFIG, "using hash %p\n", obd->obd_nid_stats_hash);
 
-        OBD_ALLOC(tmp, sizeof(struct nid_stat));
-        if (tmp == NULL)
+        OBD_ALLOC_PTR(new_stat);
+        if (new_stat == NULL)
                 RETURN(-ENOMEM);
 
-        tmp->nid = *nid;
-        tmp->nid_obd = exp->exp_obd;
-        tmp->nid_exp_ref_count = 1; /* need live in hash after destroy export */
+        OBD_ALLOC_PTR(new_ns_uuid);
+        if (new_ns_uuid == NULL) {
+                OBD_FREE_PTR(new_stat);
+                RETURN(-ENOMEM);
+        }
+        CFS_INIT_LIST_HEAD(&new_ns_uuid->ns_uuid_list);
+        strncpy(new_ns_uuid->ns_uuid.uuid, exp->exp_client_uuid.uuid,
+                sizeof(struct obd_uuid));
+
+        CFS_INIT_LIST_HEAD(&new_stat->nid_uuid_list);
+        new_stat->nid               = *nid;
+        new_stat->nid_obd           = exp->exp_obd;
+        new_stat->nid_exp_ref_count = 1; /* live in hash after destroy export */
+
+       /* protect competitive add to list, not need locking on destroy */
+        spin_lock(&obd->obd_nid_lock);
+        list_add(&new_stat->nid_list, &obd->obd_nid_stats);
+        spin_unlock(&obd->obd_nid_lock);
 
-        tmp1= lustre_hash_findadd_unique(obd->obd_nid_stats_hash_body, nid,
-                                         &tmp->nid_hash);
+        old_stat = lustre_hash_findadd_unique(obd->obd_nid_stats_hash,
+                                              nid, &new_stat->nid_hash);
         CDEBUG(D_INFO, "Found stats %p for nid %s - ref %d\n",
-               tmp1, libcfs_nid2str(*nid), tmp->nid_exp_ref_count);
+               old_stat, libcfs_nid2str(*nid), new_stat->nid_exp_ref_count);
+
+        /* Return -EALREADY here so that we know that the /proc
+         * entry already has been created */
+        if (old_stat != new_stat) {
+                struct nid_stat_uuid *tmp_uuid;
+                int found = 0;
+
+                exp->exp_nid_stats = old_stat;
+                /* We need to decrement the refcount if the uuid was
+                 * already in our list */
+                spin_lock(&obd->obd_nid_lock);
+                list_for_each_entry(tmp_uuid, &old_stat->nid_uuid_list,
+                                    ns_uuid_list) {
+                        if (tmp_uuid && obd_uuid_equals(&tmp_uuid->ns_uuid,
+                                                        &exp->exp_client_uuid)){
+                                found = 1;
+                                --old_stat->nid_exp_ref_count;
+                                break;
+                        }
+                }
+
+                if (!found)
+                        list_add(&new_ns_uuid->ns_uuid_list,
+                                 &old_stat->nid_uuid_list);
+                else
+                        OBD_FREE_PTR(new_ns_uuid);
+
+                spin_unlock(&obd->obd_nid_lock);
 
-        if (tmp1 != tmp) {
-                exp->exp_nid_stats = tmp1;
-                GOTO(destroy_new, rc = 0);
+                GOTO(destroy_new, rc = -EALREADY);
         }
         /* not found - create */
-        tmp->nid_proc = lprocfs_register(libcfs_nid2str(*nid),
-                                         obd->obd_proc_exports_entry, NULL, NULL);
-        if (!tmp->nid_proc) {
-                CERROR("Error making export directory for"
-                       " nid %s\n", libcfs_nid2str(*nid));
-                lustre_hash_delitem(obd->obd_nid_stats_hash_body, nid,
-                                    &tmp->nid_hash);
-                GOTO(destroy_new, rc = -ENOMEM);
+        new_stat->nid_proc = lprocfs_register(libcfs_nid2str(*nid),
+                                              obd->obd_proc_exports_entry,
+                                              NULL, NULL);
+        if (new_stat->nid_proc == NULL) {
+                CERROR("Error making export directory for nid %s\n",
+                       libcfs_nid2str(*nid));
+                GOTO(destroy_new_ns, rc = -ENOMEM);
         }
 
-        rc = lprocfs_add_simple(tmp->nid_proc, "uuid",
-                                lprocfs_exp_rd_uuid, NULL, tmp);
-        if (rc)
-                CWARN("Error adding the uuid file\n");
-
+        /* Add in uuid to our nid_stats list */
         spin_lock(&obd->obd_nid_lock);
-        list_add(&tmp->nid_list, &obd->obd_nid_stats);
+        list_add(&new_ns_uuid->ns_uuid_list, &new_stat->nid_uuid_list);
         spin_unlock(&obd->obd_nid_lock);
 
-        exp->exp_nid_stats = tmp;
+        entry = lprocfs_add_simple(new_stat->nid_proc, "uuid",
+                                   lprocfs_exp_rd_uuid, NULL, new_stat, NULL);
+        if (IS_ERR(entry)) {
+                CWARN("Error adding the NID stats file\n");
+                rc = PTR_ERR(entry);
+                GOTO(destroy_new_ns, rc);
+        }
+
+        entry = lprocfs_add_simple(new_stat->nid_proc, "hash",
+                                   lprocfs_exp_rd_hash, NULL, new_stat, NULL);
+        if (IS_ERR(entry)) {
+                CWARN("Error adding the hash file\n");
+                lprocfs_remove(&new_stat->nid_proc);
+                rc = PTR_ERR(entry);
+                GOTO(destroy_new_ns, rc);
+        }
+
+        exp->exp_nid_stats = new_stat;
         *newnid = 1;
         RETURN(rc);
 
+destroy_new_ns:
+        lustre_hash_del(obd->obd_nid_stats_hash, nid, &new_stat->nid_hash);
+        OBD_FREE_PTR(new_ns_uuid);
+
 destroy_new:
-        OBD_FREE(tmp, sizeof(struct nid_stat));
+        spin_lock(&obd->obd_nid_lock);
+        list_del(&new_stat->nid_list);
+        spin_unlock(&obd->obd_nid_lock);
+        OBD_FREE_PTR(new_stat);
         RETURN(rc);
 }
 
 int lprocfs_exp_cleanup(struct obd_export *exp)
 {
         struct nid_stat *stat = exp->exp_nid_stats;
+        struct nid_stat_uuid *cursor, *tmp;
+        int found = 0;
 
-        if(!stat)
+        if(!stat || !exp->exp_obd)
                 RETURN(0);
 
+        spin_lock(&exp->exp_obd->obd_nid_lock);
+        list_for_each_entry_safe(cursor, tmp,
+                                 &stat->nid_uuid_list,
+                                 ns_uuid_list) {
+                if (cursor && obd_uuid_equals(&cursor->ns_uuid,
+                                              &exp->exp_client_uuid)) {
+                        found = 1;
+                        list_del(&cursor->ns_uuid_list);
+                        OBD_FREE_PTR(cursor);
+                        break;
+                }
+        }
+        spin_unlock(&exp->exp_obd->obd_nid_lock);
+        if (!found)
+                CERROR("obd_export's client uuid %s are not found in its "
+                       "nid_stats list\n", exp->exp_client_uuid.uuid);
+
         stat->nid_exp_ref_count--;
         CDEBUG(D_INFO, "Put stat %p - %d\n", stat, stat->nid_exp_ref_count);
 
         exp->exp_nid_stats = NULL;
+        lprocfs_free_md_stats(exp->exp_obd);
+
         return 0;
 }
 
@@ -1483,7 +1832,7 @@ int lprocfs_read_frac_helper(char *buffer, unsigned long count, long val, int mu
                         2. #echo x.0x > /proc/xxx       output result : x.0x
                         3. #echo x.x0 > /proc/xxx       output result : x.x
                         4. #echo x.xx > /proc/xxx       output result : x.xx
-                        Only reserved 2bits fraction.       
+                        Only reserved 2bits fraction.
                  */
                 for (i = 0; i < (5 - prtn); i++)
                         temp_mult *= 10;
@@ -1562,7 +1911,7 @@ int lprocfs_write_frac_u64_helper(const char *buffer, unsigned long count,
                 units <<= 10;
         }
         /* Specified units override the multiplier */
-        if (units) 
+        if (units)
                 mult = mult < 0 ? -units : units;
 
         frac *= mult;
@@ -1571,7 +1920,7 @@ int lprocfs_write_frac_u64_helper(const char *buffer, unsigned long count,
         return 0;
 }
 
-int lprocfs_seq_create(cfs_proc_dir_entry_t *parent, 
+int lprocfs_seq_create(cfs_proc_dir_entry_t *parent,
                        char *name, mode_t mode,
                        struct file_operations *seq_fops, void *data)
 {
@@ -1593,7 +1942,7 @@ __inline__ int lprocfs_obd_seq_create(struct obd_device *dev, char *name,
                                       struct file_operations *seq_fops,
                                       void *data)
 {
-        return (lprocfs_seq_create(dev->obd_proc_entry, name, 
+        return (lprocfs_seq_create(dev->obd_proc_entry, name,
                                    mode, seq_fops, data));
 }
 EXPORT_SYMBOL(lprocfs_obd_seq_create);
@@ -1639,6 +1988,24 @@ void lprocfs_oh_clear(struct obd_histogram *oh)
 }
 EXPORT_SYMBOL(lprocfs_oh_clear);
 
+int lprocfs_obd_rd_hash(char *page, char **start, off_t off,
+                        int count, int *eof, void *data)
+{
+        struct obd_device *obd = data;
+        int c = 0;
+
+        if (obd == NULL)
+                return 0;
+
+        c += lustre_hash_debug_header(page, count);
+        c += lustre_hash_debug_str(obd->obd_uuid_hash, page + c, count - c);
+        c += lustre_hash_debug_str(obd->obd_nid_hash, page + c, count - c);
+        c += lustre_hash_debug_str(obd->obd_nid_stats_hash, page+c, count-c);
+
+        return c;
+}
+EXPORT_SYMBOL(lprocfs_obd_rd_hash);
+
 int lprocfs_obd_rd_recovery_status(char *page, char **start, off_t off,
                                    int count, int *eof, void *data)
 {
@@ -1665,7 +2032,6 @@ int lprocfs_obd_rd_recovery_status(char *page, char **start, off_t off,
 
         if (lprocfs_obd_snprintf(&page, size, &len, "status: ") <= 0)
                 goto out;
-
         if (obd->obd_max_recoverable_clients == 0) {
                 if (lprocfs_obd_snprintf(&page, size, &len, "INACTIVE\n") <= 0)
                         goto out;
@@ -1677,64 +2043,60 @@ int lprocfs_obd_rd_recovery_status(char *page, char **start, off_t off,
         if (obd->obd_recovering == 0) {
                 if (lprocfs_obd_snprintf(&page, size, &len, "COMPLETE\n") <= 0)
                         goto out;
-
-                if (lprocfs_obd_snprintf(&page, size, &len, "recovery_start: %lu\n",
-                    obd->obd_recovery_start) <= 0)
+                if (lprocfs_obd_snprintf(&page, size, &len,
+                                         "recovery_start: %lu\n",
+                                         obd->obd_recovery_start) <= 0)
                         goto out;
-
-                if (lprocfs_obd_snprintf(&page, size, &len, "recovery_end: %lu\n",
-                    obd->obd_recovery_end) <= 0)
+                if (lprocfs_obd_snprintf(&page, size, &len,
+                                         "recovery_duration: %lu\n",
+                                         obd->obd_recovery_end -
+                                         obd->obd_recovery_start) <= 0)
                         goto out;
-
-                /* Number of clients have have completed recovery */
-                if (lprocfs_obd_snprintf(&page, size, &len, "recovered_clients: %d\n",
-                    obd->obd_max_recoverable_clients - obd->obd_recoverable_clients) <= 0)
+                /* Number of clients that have completed recovery */
+                if (lprocfs_obd_snprintf(&page, size, &len,
+                                         "completed_clients: %d/%d\n",
+                                         obd->obd_max_recoverable_clients -
+                                         obd->obd_recoverable_clients,
+                                         obd->obd_max_recoverable_clients) <= 0)
                         goto out;
-
-                if (lprocfs_obd_snprintf(&page, size, &len, "unrecovered_clients: %d\n",
-                    obd->obd_recoverable_clients) <= 0)
+                if (lprocfs_obd_snprintf(&page, size, &len,
+                                         "replayed_requests: %d\n",
+                                         obd->obd_replayed_requests) <= 0)
                         goto out;
-
-                if (lprocfs_obd_snprintf(&page, size, &len, "last_transno: "LPD64"\n",
-                    obd->obd_next_recovery_transno - 1) <= 0)
+                if (lprocfs_obd_snprintf(&page, size, &len,
+                                         "last_transno: "LPD64"\n",
+                                         obd->obd_next_recovery_transno - 1)<=0)
                         goto out;
-
-                lprocfs_obd_snprintf(&page, size, &len, "replayed_requests: %d\n", obd->obd_replayed_requests);
                 goto fclose;
         }
 
         if (lprocfs_obd_snprintf(&page, size, &len, "RECOVERING\n") <= 0)
                 goto out;
-
         if (lprocfs_obd_snprintf(&page, size, &len, "recovery_start: %lu\n",
-            obd->obd_recovery_start) <= 0)
+                                 obd->obd_recovery_start) <= 0)
                 goto out;
-
-        if (lprocfs_obd_snprintf(&page, size, &len, "time remaining: %lu\n",
-                                 cfs_time_current_sec() >= obd->obd_recovery_end ? 0 :
-                                 obd->obd_recovery_end - cfs_time_current_sec()) <= 0)
+        if (lprocfs_obd_snprintf(&page, size, &len, "time_remaining: %lu\n",
+                           cfs_time_current_sec() >= obd->obd_recovery_end ? 0 :
+                           obd->obd_recovery_end - cfs_time_current_sec()) <= 0)
                 goto out;
-
-        if(lprocfs_obd_snprintf(&page, size, &len, "connected_clients: %d/%d\n",
-                                obd->obd_connected_clients,
-                                obd->obd_max_recoverable_clients) <= 0)
+        if (lprocfs_obd_snprintf(&page, size, &len,"connected_clients: %d/%d\n",
+                                 obd->obd_connected_clients,
+                                 obd->obd_max_recoverable_clients) <= 0)
                 goto out;
-
-        /* Number of clients have have completed recovery */
-        if (lprocfs_obd_snprintf(&page, size, &len, "completed_clients: %d/%d\n",
-                                 obd->obd_max_recoverable_clients - obd->obd_recoverable_clients,
+        /* Number of clients that have completed recovery */
+        if (lprocfs_obd_snprintf(&page, size, &len,"completed_clients: %d/%d\n",
+                                 obd->obd_max_recoverable_clients -
+                                 obd->obd_recoverable_clients,
                                  obd->obd_max_recoverable_clients) <= 0)
                 goto out;
-
-        if (lprocfs_obd_snprintf(&page, size, &len, "replayed_requests: %d/??\n",
+        if (lprocfs_obd_snprintf(&page, size, &len,"replayed_requests: %d/??\n",
                                  obd->obd_replayed_requests) <= 0)
                 goto out;
-
         if (lprocfs_obd_snprintf(&page, size, &len, "queued_requests: %d\n",
                                  obd->obd_requests_queued_for_recovery) <= 0)
                 goto out;
 
-        if (lprocfs_obd_snprintf(&page, size, &len, "next_transno: "LPD64"\n", 
+        if (lprocfs_obd_snprintf(&page, size, &len, "next_transno: "LPD64"\n",
                                  obd->obd_next_recovery_transno) <= 0)
                 goto out;
 
@@ -1751,7 +2113,7 @@ int lprocfs_obd_rd_recovery_maxtime(char *page, char **start, off_t off,
         struct obd_device *obd = (struct obd_device *)data;
         LASSERT(obd != NULL);
 
-        return snprintf(page, count, "%lu\n", 
+        return snprintf(page, count, "%lu\n",
                         obd->obd_recovery_max_time);
 }
 EXPORT_SYMBOL(lprocfs_obd_rd_recovery_maxtime);
@@ -1787,7 +2149,9 @@ EXPORT_SYMBOL(lprocfs_free_stats);
 EXPORT_SYMBOL(lprocfs_clear_stats);
 EXPORT_SYMBOL(lprocfs_register_stats);
 EXPORT_SYMBOL(lprocfs_init_ops_stats);
+EXPORT_SYMBOL(lprocfs_init_ldlm_stats);
 EXPORT_SYMBOL(lprocfs_alloc_obd_stats);
+EXPORT_SYMBOL(lprocfs_alloc_md_stats);
 EXPORT_SYMBOL(lprocfs_free_obd_stats);
 EXPORT_SYMBOL(lprocfs_exp_setup);
 EXPORT_SYMBOL(lprocfs_exp_cleanup);
@@ -1804,7 +2168,9 @@ EXPORT_SYMBOL(lprocfs_rd_server_uuid);
 EXPORT_SYMBOL(lprocfs_rd_conn_uuid);
 EXPORT_SYMBOL(lprocfs_rd_num_exports);
 EXPORT_SYMBOL(lprocfs_rd_numrefs);
-
+EXPORT_SYMBOL(lprocfs_at_hist_helper);
+EXPORT_SYMBOL(lprocfs_rd_import);
+EXPORT_SYMBOL(lprocfs_rd_timeouts);
 EXPORT_SYMBOL(lprocfs_rd_blksize);
 EXPORT_SYMBOL(lprocfs_rd_kbytestotal);
 EXPORT_SYMBOL(lprocfs_rd_kbytesfree);