Whamcloud - gitweb
LU-11111 lfsck: skip orphan processing
[fs/lustre-release.git] / lustre / osd-ldiskfs / osd_scrub.c
index c94a3b0..6232235 100644 (file)
@@ -6,21 +6,21 @@
  * 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.
-
- * 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 COPYING file that accompanied this code.
-
+ *
+ * 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).
+ *
  * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ * version 2 along with this program; If not, see
+ * http://www.gnu.org/licenses/gpl-2.0.html
  *
  * GPL HEADER END
  */
 /*
- * Copyright (c) 2012, 2016, Intel Corporation.
+ * Copyright (c) 2012, 2017, Intel Corporation.
  */
 /*
  * lustre/osd-ldiskfs/osd_scrub.c
 #define DEBUG_SUBSYSTEM S_LFSCK
 
 #include <linux/kthread.h>
-#include <lustre/lustre_idl.h>
+#include <uapi/linux/lustre/lustre_idl.h>
 #include <lustre_disk.h>
 #include <dt_object.h>
 #include <linux/xattr.h>
+#include <lustre_scrub.h>
+#include <lustre_nodemap.h>
 
 #include "osd_internal.h"
 #include "osd_oi.h"
 #include "osd_scrub.h"
 
-#define HALF_SEC       msecs_to_jiffies(MSEC_PER_SEC >> 1)
-
 #define OSD_OTABLE_MAX_HASH            0x00000000ffffffffULL
 
-#define SCRUB_NEXT_BREAK       1 /* exit current loop and process next group */
-#define SCRUB_NEXT_CONTINUE    2 /* skip current object and process next bit */
-#define SCRUB_NEXT_EXIT        3 /* exit all the loops */
-#define SCRUB_NEXT_WAIT        4 /* wait for free cache slot */
-#define SCRUB_NEXT_CRASH       5 /* simulate system crash during OI scrub */
-#define SCRUB_NEXT_FATAL       6 /* simulate failure during OI scrub */
-#define SCRUB_NEXT_NOSCRUB     7 /* new created object, no scrub on it */
-#define SCRUB_NEXT_NOLMA       8 /* the inode has no FID-in-LMA */
-#define SCRUB_NEXT_OSTOBJ      9 /* for OST-object */
-#define SCRUB_NEXT_OSTOBJ_OLD  10 /* old OST-object, no LMA or no FID-on-OST
-                                   * flags in LMA */
-
-/* misc functions */
-
-static inline struct osd_device *osd_scrub2dev(struct osd_scrub *scrub)
-{
-       return container_of0(scrub, struct osd_device, od_scrub);
-}
-
-static inline struct super_block *osd_scrub2sb(struct osd_scrub *scrub)
-{
-       return osd_sb(osd_scrub2dev(scrub));
-}
-
-static inline int osd_scrub_has_window(struct osd_scrub *scrub,
+static inline int osd_scrub_has_window(struct lustre_scrub *scrub,
                                       struct osd_otable_cache *ooc)
 {
        return scrub->os_pos_current < ooc->ooc_pos_preload + SCRUB_WINDOW_SIZE;
 }
 
-static inline const char *osd_scrub2name(struct osd_scrub *scrub)
-{
-       return osd_dev2name(osd_scrub2dev(scrub));
-}
-
 /**
  * update/insert/delete the specified OI mapping (@fid @id) according to the ops
  *
@@ -104,7 +75,7 @@ static int osd_scrub_refresh_mapping(struct osd_thread_info *info,
        int       rc;
        ENTRY;
 
-       if (dev->od_scrub.os_file.sf_param & SP_DRYRUN && !force)
+       if (dev->od_scrub.os_scrub.os_file.sf_param & SP_DRYRUN && !force)
                RETURN(0);
 
        /* DTO_INDEX_INSERT is enough for other two ops:
@@ -188,199 +159,20 @@ static int osd_scrub_refresh_mapping(struct osd_thread_info *info,
        RETURN(rc);
 }
 
-/* OI_scrub file ops */
-
-static void osd_scrub_file_to_cpu(struct scrub_file *des,
-                                 struct scrub_file *src)
-{
-       memcpy(des->sf_uuid, src->sf_uuid, 16);
-       des->sf_flags   = le64_to_cpu(src->sf_flags);
-       des->sf_magic   = le32_to_cpu(src->sf_magic);
-       des->sf_status  = le16_to_cpu(src->sf_status);
-       des->sf_param   = le16_to_cpu(src->sf_param);
-       des->sf_time_last_complete      =
-                               le64_to_cpu(src->sf_time_last_complete);
-       des->sf_time_latest_start       =
-                               le64_to_cpu(src->sf_time_latest_start);
-       des->sf_time_last_checkpoint    =
-                               le64_to_cpu(src->sf_time_last_checkpoint);
-       des->sf_pos_latest_start        =
-                               le64_to_cpu(src->sf_pos_latest_start);
-       des->sf_pos_last_checkpoint     =
-                               le64_to_cpu(src->sf_pos_last_checkpoint);
-       des->sf_pos_first_inconsistent  =
-                               le64_to_cpu(src->sf_pos_first_inconsistent);
-       des->sf_items_checked           =
-                               le64_to_cpu(src->sf_items_checked);
-       des->sf_items_updated           =
-                               le64_to_cpu(src->sf_items_updated);
-       des->sf_items_failed            =
-                               le64_to_cpu(src->sf_items_failed);
-       des->sf_items_updated_prior     =
-                               le64_to_cpu(src->sf_items_updated_prior);
-       des->sf_run_time        = le32_to_cpu(src->sf_run_time);
-       des->sf_success_count   = le32_to_cpu(src->sf_success_count);
-       des->sf_oi_count        = le16_to_cpu(src->sf_oi_count);
-       des->sf_internal_flags  = le16_to_cpu(src->sf_internal_flags);
-       memcpy(des->sf_oi_bitmap, src->sf_oi_bitmap, SCRUB_OI_BITMAP_SIZE);
-}
-
-static void osd_scrub_file_to_le(struct scrub_file *des,
-                                struct scrub_file *src)
-{
-       memcpy(des->sf_uuid, src->sf_uuid, 16);
-       des->sf_flags   = cpu_to_le64(src->sf_flags);
-       des->sf_magic   = cpu_to_le32(src->sf_magic);
-       des->sf_status  = cpu_to_le16(src->sf_status);
-       des->sf_param   = cpu_to_le16(src->sf_param);
-       des->sf_time_last_complete      =
-                               cpu_to_le64(src->sf_time_last_complete);
-       des->sf_time_latest_start       =
-                               cpu_to_le64(src->sf_time_latest_start);
-       des->sf_time_last_checkpoint    =
-                               cpu_to_le64(src->sf_time_last_checkpoint);
-       des->sf_pos_latest_start        =
-                               cpu_to_le64(src->sf_pos_latest_start);
-       des->sf_pos_last_checkpoint     =
-                               cpu_to_le64(src->sf_pos_last_checkpoint);
-       des->sf_pos_first_inconsistent  =
-                               cpu_to_le64(src->sf_pos_first_inconsistent);
-       des->sf_items_checked           =
-                               cpu_to_le64(src->sf_items_checked);
-       des->sf_items_updated           =
-                               cpu_to_le64(src->sf_items_updated);
-       des->sf_items_failed            =
-                               cpu_to_le64(src->sf_items_failed);
-       des->sf_items_updated_prior     =
-                               cpu_to_le64(src->sf_items_updated_prior);
-       des->sf_run_time        = cpu_to_le32(src->sf_run_time);
-       des->sf_success_count   = cpu_to_le32(src->sf_success_count);
-       des->sf_oi_count        = cpu_to_le16(src->sf_oi_count);
-       des->sf_internal_flags  = cpu_to_le16(src->sf_internal_flags);
-       memcpy(des->sf_oi_bitmap, src->sf_oi_bitmap, SCRUB_OI_BITMAP_SIZE);
-}
-
-static void osd_scrub_file_init(struct osd_scrub *scrub, __u8 *uuid)
-{
-       struct scrub_file *sf = &scrub->os_file;
-
-       memset(sf, 0, sizeof(*sf));
-       memcpy(sf->sf_uuid, uuid, 16);
-       sf->sf_magic = SCRUB_MAGIC_V1;
-       sf->sf_status = SS_INIT;
-}
-
-void osd_scrub_file_reset(struct osd_scrub *scrub, __u8 *uuid, __u64 flags)
-{
-       struct scrub_file *sf = &scrub->os_file;
-
-       CDEBUG(D_LFSCK, "%s: reset OI scrub file, old flags = "
-              "%#llx, add flags = %#llx\n",
-              osd_scrub2name(scrub), sf->sf_flags, flags);
-
-       memcpy(sf->sf_uuid, uuid, 16);
-       sf->sf_status = SS_INIT;
-       sf->sf_flags |= flags;
-       sf->sf_flags &= ~SF_AUTO;
-       sf->sf_run_time = 0;
-       sf->sf_time_latest_start = 0;
-       sf->sf_time_last_checkpoint = 0;
-       sf->sf_pos_latest_start = 0;
-       sf->sf_pos_last_checkpoint = 0;
-       sf->sf_pos_first_inconsistent = 0;
-       sf->sf_items_checked = 0;
-       sf->sf_items_updated = 0;
-       sf->sf_items_failed = 0;
-       if (!scrub->os_in_join)
-               sf->sf_items_updated_prior = 0;
-
-       sf->sf_items_noscrub = 0;
-       sf->sf_items_igif = 0;
-}
-
-static int osd_scrub_file_load(struct osd_scrub *scrub)
-{
-       loff_t  pos  = 0;
-       int     len  = sizeof(scrub->os_file_disk);
-       int     rc;
-
-       rc = osd_ldiskfs_read(scrub->os_inode, &scrub->os_file_disk, len, &pos);
-       if (rc == len) {
-               struct scrub_file *sf = &scrub->os_file;
-
-               osd_scrub_file_to_cpu(sf, &scrub->os_file_disk);
-               if (sf->sf_magic != SCRUB_MAGIC_V1) {
-                       CDEBUG(D_LFSCK, "%s: invalid scrub magic "
-                              "0x%x != 0x%x\n", osd_scrub2name(scrub),
-                              sf->sf_magic, SCRUB_MAGIC_V1);
-                       /* Process it as new scrub file. */
-                       rc = -ENOENT;
-               } else {
-                       rc = 0;
-               }
-       } else if (rc != 0) {
-               CDEBUG(D_LFSCK, "%s: fail to load scrub file, "
-                      "expected = %d: rc = %d\n",
-                      osd_scrub2name(scrub), len, rc);
-               if (rc > 0)
-                       rc = -EFAULT;
-       } else {
-               /* return -ENOENT for empty scrub file case. */
-               rc = -ENOENT;
-       }
-
-       return rc;
-}
-
-int osd_scrub_file_store(struct osd_scrub *scrub)
-{
-       struct osd_device *dev;
-       handle_t          *jh;
-       loff_t             pos     = 0;
-       int                len     = sizeof(scrub->os_file_disk);
-       int                credits;
-       int                rc;
-
-       dev = container_of0(scrub, struct osd_device, od_scrub);
-       credits = osd_dto_credits_noquota[DTO_WRITE_BASE] +
-                 osd_dto_credits_noquota[DTO_WRITE_BLOCK];
-       jh = osd_journal_start_sb(osd_sb(dev), LDISKFS_HT_MISC, credits);
-       if (IS_ERR(jh)) {
-               rc = PTR_ERR(jh);
-               CDEBUG(D_LFSCK, "%s: fail to start trans for scrub store: "
-                      "rc = %d\n", osd_scrub2name(scrub), rc);
-               return rc;
-       }
-
-       osd_scrub_file_to_le(&scrub->os_file_disk, &scrub->os_file);
-       rc = osd_ldiskfs_write_record(scrub->os_inode, &scrub->os_file_disk,
-                                     len, 0, &pos, jh);
-       ldiskfs_journal_stop(jh);
-       if (rc != 0)
-               CDEBUG(D_LFSCK, "%s: fail to store scrub file, "
-                      "expected = %d: rc = %d\n",
-                      osd_scrub2name(scrub), len, rc);
-
-       scrub->os_time_last_checkpoint = cfs_time_current();
-       scrub->os_time_next_checkpoint = scrub->os_time_last_checkpoint +
-                               cfs_time_seconds(SCRUB_CHECKPOINT_INTERVAL);
-       return rc;
-}
-
 static int
 osd_scrub_convert_ff(struct osd_thread_info *info, struct osd_device *dev,
                     struct inode *inode, const struct lu_fid *fid)
 {
-       struct filter_fid_old   *ff      = &info->oti_ff;
-       struct dentry           *dentry  = &info->oti_obj_dentry;
-       struct lu_fid           *tfid    = &info->oti_fid;
-       handle_t                *jh;
-       int                      size    = 0;
-       int                      rc;
-       bool                     reset   = false;
+       struct filter_fid_18_23 *ff = &info->oti_ff_old;
+       struct dentry *dentry = &info->oti_obj_dentry;
+       struct lu_fid *tfid = &info->oti_fid;
+       bool fid_18_23 = false;
+       handle_t *jh;
+       int size = 0;
+       int rc;
        ENTRY;
 
-       if (dev->od_scrub.os_file.sf_param & SP_DRYRUN)
+       if (dev->od_scrub.os_scrub.os_file.sf_param & SP_DRYRUN)
                RETURN(0);
 
        if (fid_is_idif(fid) && dev->od_index_in_idif == 0) {
@@ -416,18 +208,18 @@ osd_scrub_convert_ff(struct osd_thread_info *info, struct osd_device *dev,
        if (rc == sizeof(*ff)) {
                /* 2) delete the old XATTR_NAME_FID */
                ll_vfs_dq_init(inode);
-               rc = inode->i_op->removexattr(dentry, XATTR_NAME_FID);
+               rc = osd_removexattr(dentry, inode, XATTR_NAME_FID);
                if (rc)
                        GOTO(stop, rc);
 
-               reset = true;
-       } else if (rc != -ENODATA && rc != sizeof(struct filter_fid)) {
+               fid_18_23 = true;
+       } else if (rc != -ENODATA && rc < (int)sizeof(struct filter_fid_24_29)) {
                GOTO(stop, rc = -EINVAL);
        }
 
        /* 3) make new LMA and add it */
        rc = osd_ea_fid_set(info, inode, tfid, LMAC_FID_ON_OST, 0);
-       if (reset) {
+       if (fid_18_23) {
                if (rc)
                        /* If failed, we should try to add the old back. */
                        size = sizeof(*ff);
@@ -460,7 +252,7 @@ static int
 osd_scrub_check_update(struct osd_thread_info *info, struct osd_device *dev,
                       struct osd_idmap_cache *oic, int val)
 {
-       struct osd_scrub             *scrub  = &dev->od_scrub;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct scrub_file            *sf     = &scrub->os_file;
        struct lu_fid                *fid    = &oic->oic_fid;
        struct osd_inode_id          *lid    = &oic->oic_lid;
@@ -653,11 +445,9 @@ out:
        RETURN(sf->sf_param & SP_FAILOUT ? rc : 0);
 }
 
-/* OI scrub APIs */
-
-static int osd_scrub_prep(struct osd_device *dev)
+static int osd_scrub_prep(const struct lu_env *env, struct osd_device *dev)
 {
-       struct osd_scrub     *scrub  = &dev->od_scrub;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct ptlrpc_thread *thread = &scrub->os_thread;
        struct scrub_file    *sf     = &scrub->os_file;
        __u32                 flags  = scrub->os_start_flags;
@@ -682,7 +472,7 @@ static int osd_scrub_prep(struct osd_device *dev)
        }
 
        if (flags & SS_RESET)
-               osd_scrub_file_reset(scrub,
+               scrub_file_reset(scrub,
                        LDISKFS_SB(osd_sb(dev))->s_es->s_uuid, 0);
 
        if (flags & SS_AUTO_FULL) {
@@ -719,10 +509,10 @@ static int osd_scrub_prep(struct osd_device *dev)
 
        scrub->os_pos_current = sf->sf_pos_latest_start;
        sf->sf_status = SS_SCANNING;
-       sf->sf_time_latest_start = cfs_time_current_sec();
+       sf->sf_time_latest_start = ktime_get_real_seconds();
        sf->sf_time_last_checkpoint = sf->sf_time_latest_start;
        sf->sf_pos_last_checkpoint = sf->sf_pos_latest_start - 1;
-       rc = osd_scrub_file_store(scrub);
+       rc = scrub_file_store(env, scrub);
        if (rc == 0) {
                spin_lock(&scrub->os_lock);
                thread_set_flags(thread, SVC_RUNNING);
@@ -734,36 +524,15 @@ static int osd_scrub_prep(struct osd_device *dev)
        RETURN(rc);
 }
 
-static int osd_scrub_checkpoint(struct osd_scrub *scrub)
-{
-       struct scrub_file *sf = &scrub->os_file;
-       int                rc;
-
-       if (likely(cfs_time_before(cfs_time_current(),
-                                  scrub->os_time_next_checkpoint) ||
-                  scrub->os_new_checked == 0))
-               return 0;
-
-       down_write(&scrub->os_rwsem);
-       sf->sf_items_checked += scrub->os_new_checked;
-       scrub->os_new_checked = 0;
-       sf->sf_pos_last_checkpoint = scrub->os_pos_current;
-       sf->sf_time_last_checkpoint = cfs_time_current_sec();
-       sf->sf_run_time += cfs_duration_sec(cfs_time_current() + HALF_SEC -
-                                           scrub->os_time_last_checkpoint);
-       rc = osd_scrub_file_store(scrub);
-       up_write(&scrub->os_rwsem);
-
-       return rc;
-}
-
-static int osd_scrub_post(struct osd_scrub *scrub, int result)
+static int osd_scrub_post(const struct lu_env *env, struct osd_device *dev,
+                         int result)
 {
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct scrub_file *sf = &scrub->os_file;
        int rc;
        ENTRY;
 
-       CDEBUG(D_LFSCK, "%s: OI scrub post, result = %d\n",
+       CDEBUG(D_LFSCK, "%s: OI scrub post with result = %d\n",
               osd_scrub2name(scrub), result);
 
        down_write(&scrub->os_rwsem);
@@ -775,11 +544,8 @@ static int osd_scrub_post(struct osd_scrub *scrub, int result)
                scrub->os_new_checked = 0;
                sf->sf_pos_last_checkpoint = scrub->os_pos_current;
        }
-       sf->sf_time_last_checkpoint = cfs_time_current_sec();
+       sf->sf_time_last_checkpoint = ktime_get_real_seconds();
        if (result > 0) {
-               struct osd_device *dev =
-                       container_of0(scrub, struct osd_device, od_scrub);
-
                dev->od_igif_inoi = 1;
                dev->od_check_ff = 0;
                sf->sf_status = SS_COMPLETED;
@@ -798,9 +564,10 @@ static int osd_scrub_post(struct osd_scrub *scrub, int result)
        } else {
                sf->sf_status = SS_FAILED;
        }
-       sf->sf_run_time += cfs_duration_sec(cfs_time_current() + HALF_SEC -
-                                           scrub->os_time_last_checkpoint);
-       rc = osd_scrub_file_store(scrub);
+       sf->sf_run_time += ktime_get_seconds() -
+                          scrub->os_time_last_checkpoint;
+
+       rc = scrub_file_store(env, scrub);
        up_write(&scrub->os_rwsem);
 
        RETURN(rc < 0 ? rc : result);
@@ -820,7 +587,7 @@ typedef int (*osd_iit_exec_policy)(struct osd_thread_info *info,
                                   struct osd_idmap_cache *oic,
                                   bool *noslot, int rc);
 
-static int osd_iit_next(struct osd_iit_param *param, __u32 *pos)
+static int osd_iit_next(struct osd_iit_param *param, __u64 *pos)
 {
        __u32 offset;
 
@@ -907,7 +674,7 @@ static int osd_scrub_get_fid(struct osd_thread_info *info,
                rc = osd_get_idif(info, inode, &info->oti_obj_dentry, fid);
                if (rc == 0) {
                        if (scrub)
-                               /* It is old 2.x (x <= 3) or 1.8 OST-object. */
+                               /* It is 2.3 or older OST-object. */
                                rc = SCRUB_NEXT_OSTOBJ_OLD;
                        return rc;
                }
@@ -918,7 +685,7 @@ static int osd_scrub_get_fid(struct osd_thread_info *info,
                                 * to generate its FID, ignore it directly. */
                                rc = SCRUB_NEXT_CONTINUE;
                        else
-                               /* It is 2.4 OST-object. */
+                               /* It is 2.4 or newer OST-object. */
                                rc = SCRUB_NEXT_OSTOBJ_OLD;
                        return rc;
                }
@@ -927,7 +694,7 @@ static int osd_scrub_get_fid(struct osd_thread_info *info,
                        return rc;
 
                if (!has_lma) {
-                       if (dev->od_scrub.os_convert_igif) {
+                       if (dev->od_scrub.os_scrub.os_convert_igif) {
                                lu_igif_build(fid, inode->i_ino,
                                              inode->i_generation);
                                if (scrub)
@@ -963,9 +730,16 @@ static int osd_iit_iget(struct osd_thread_info *info, struct osd_device *dev,
        /* Not handle the backend root object and agent parent object.
         * They are neither visible to namespace nor have OI mappings. */
        if (unlikely(pos == osd_sb(dev)->s_root->d_inode->i_ino ||
-                    pos == osd_remote_parent_ino(dev)))
+                    is_remote_parent_ino(dev, pos)))
                RETURN(SCRUB_NEXT_CONTINUE);
 
+        /* Skip project quota inode since it is greater than s_first_ino. */
+#ifdef HAVE_PROJECT_QUOTA
+       if (ldiskfs_has_feature_project(sb) &&
+           pos == le32_to_cpu(LDISKFS_SB(sb)->s_es->s_prj_quota_inum))
+               RETURN(SCRUB_NEXT_CONTINUE);
+#endif
+
        osd_id_gen(lid, pos, OSD_OII_NOGEN);
        inode = osd_iget(info, dev, lid);
        if (IS_ERR(inode)) {
@@ -1004,11 +778,11 @@ static int osd_scrub_next(struct osd_thread_info *info, struct osd_device *dev,
                          struct osd_iit_param *param,
                          struct osd_idmap_cache **oic, const bool noslot)
 {
-       struct osd_scrub     *scrub  = &dev->od_scrub;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct ptlrpc_thread *thread = &scrub->os_thread;
-       struct lu_fid        *fid;
-       struct osd_inode_id  *lid;
-       int                   rc;
+       struct lu_fid *fid;
+       struct osd_inode_id *lid;
+       int rc;
 
        if (OBD_FAIL_CHECK(OBD_FAIL_OSD_SCRUB_DELAY) && cfs_fail_val > 0) {
                struct l_wait_info lwi;
@@ -1058,7 +832,7 @@ static int osd_scrub_next(struct osd_thread_info *info, struct osd_device *dev,
        if (rc != 0)
                return rc;
 
-       *oic = &scrub->os_oic;
+       *oic = &dev->od_scrub.os_oic;
        fid = &(*oic)->oic_fid;
        lid = &(*oic)->oic_lid;
        rc = osd_iit_iget(info, dev, fid, lid,
@@ -1071,7 +845,7 @@ static int osd_preload_next(struct osd_thread_info *info,
                            struct osd_idmap_cache **oic, const bool noslot)
 {
        struct osd_otable_cache *ooc = &dev->od_otable_it->ooi_cache;
-       struct osd_scrub *scrub = &dev->od_scrub;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct ptlrpc_thread *thread = &scrub->os_thread;
        int rc;
 
@@ -1091,7 +865,7 @@ static int osd_preload_next(struct osd_thread_info *info,
 }
 
 static inline int
-osd_scrub_wakeup(struct osd_scrub *scrub, struct osd_otable_it *it)
+osd_scrub_wakeup(struct lustre_scrub *scrub, struct osd_otable_it *it)
 {
        spin_lock(&scrub->os_lock);
        if (osd_scrub_has_window(scrub, &it->ooi_cache) ||
@@ -1109,12 +883,12 @@ static int osd_scrub_exec(struct osd_thread_info *info, struct osd_device *dev,
                          struct osd_iit_param *param,
                          struct osd_idmap_cache *oic, bool *noslot, int rc)
 {
-       struct l_wait_info       lwi    = { 0 };
-       struct osd_scrub        *scrub  = &dev->od_scrub;
-       struct scrub_file       *sf     = &scrub->os_file;
-       struct ptlrpc_thread    *thread = &scrub->os_thread;
-       struct osd_otable_it    *it     = dev->od_otable_it;
-       struct osd_otable_cache *ooc    = it ? &it->ooi_cache : NULL;
+       struct l_wait_info lwi = { 0 };
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
+       struct scrub_file *sf = &scrub->os_file;
+       struct ptlrpc_thread *thread = &scrub->os_thread;
+       struct osd_otable_it *it = dev->od_otable_it;
+       struct osd_otable_cache *ooc = it ? &it->ooi_cache : NULL;
 
        switch (rc) {
        case SCRUB_NEXT_NOSCRUB:
@@ -1133,9 +907,9 @@ static int osd_scrub_exec(struct osd_thread_info *info, struct osd_device *dev,
                return rc;
        }
 
-       rc = osd_scrub_checkpoint(scrub);
-       if (rc != 0) {
-               CDEBUG(D_LFSCK, "%s: fail to checkpoint, pos = %u: "
+       rc = scrub_checkpoint(info->oti_env, scrub);
+       if (rc) {
+               CDEBUG(D_LFSCK, "%s: fail to checkpoint, pos = %llu: "
                       "rc = %d\n", osd_scrub2name(scrub),
                       scrub->os_pos_current, rc);
                /* Continue, as long as the scrub itself can go ahead. */
@@ -1191,10 +965,10 @@ static int osd_preload_exec(struct osd_thread_info *info,
 #define SCRUB_IT_ALL   1
 #define SCRUB_IT_CRASH 2
 
-static void osd_scrub_join(struct osd_device *dev, __u32 flags,
-                          bool inconsistent)
+static void osd_scrub_join(const struct lu_env *env, struct osd_device *dev,
+                          __u32 flags, bool inconsistent)
 {
-       struct osd_scrub     *scrub  = &dev->od_scrub;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct ptlrpc_thread *thread = &scrub->os_thread;
        struct scrub_file    *sf     = &scrub->os_file;
        int                   rc;
@@ -1215,8 +989,7 @@ static void osd_scrub_join(struct osd_device *dev, __u32 flags,
                sf->sf_param &= ~SP_DRYRUN;
 
        if (flags & SS_RESET) {
-               osd_scrub_file_reset(scrub,
-                       LDISKFS_SB(osd_sb(dev))->s_es->s_uuid,
+               scrub_file_reset(scrub, LDISKFS_SB(osd_sb(dev))->s_es->s_uuid,
                        inconsistent ? SF_INCONSISTENT : 0);
                sf->sf_status = SS_SCANNING;
        }
@@ -1238,13 +1011,10 @@ static void osd_scrub_join(struct osd_device *dev, __u32 flags,
                sf->sf_pos_latest_start = LDISKFS_FIRST_INO(osd_sb(dev)) + 1;
 
        scrub->os_pos_current = sf->sf_pos_latest_start;
-       sf->sf_time_latest_start = cfs_time_current_sec();
+       sf->sf_time_latest_start = ktime_get_real_seconds();
        sf->sf_time_last_checkpoint = sf->sf_time_latest_start;
        sf->sf_pos_last_checkpoint = sf->sf_pos_latest_start - 1;
-       rc = osd_scrub_file_store(scrub);
-       if (rc != 0)
-               CDEBUG(D_LFSCK, "%s: fail to store scrub file when join "
-                      "the OI scrub: rc = %d\n", osd_scrub2name(scrub), rc);
+       rc = scrub_file_store(env, scrub);
 
        spin_lock(&scrub->os_lock);
        scrub->os_waiting = 0;
@@ -1256,31 +1026,33 @@ static void osd_scrub_join(struct osd_device *dev, __u32 flags,
        wake_up_all(&thread->t_ctl_waitq);
        up_write(&scrub->os_rwsem);
 
+       CDEBUG(D_LFSCK, "%s: joined in the OI scrub with flag %u: rc = %d\n",
+              osd_scrub2name(scrub), flags, rc);
+
        EXIT;
 }
 
 static int osd_inode_iteration(struct osd_thread_info *info,
                               struct osd_device *dev, __u32 max, bool preload)
 {
-       struct osd_scrub     *scrub  = &dev->od_scrub;
+       struct lustre_scrub *scrub  = &dev->od_scrub.os_scrub;
        struct ptlrpc_thread *thread = &scrub->os_thread;
-       struct scrub_file    *sf     = &scrub->os_file;
-       osd_iit_next_policy   next;
-       osd_iit_exec_policy   exec;
-       __u32                *pos;
-       __u32                *count;
+       struct scrub_file *sf = &scrub->os_file;
+       osd_iit_next_policy next;
+       osd_iit_exec_policy exec;
+       __u64 *pos;
+       __u64 *count;
        struct osd_iit_param *param;
-       struct l_wait_info    lwi    = { 0 };
-       __u32                 limit =
-               le32_to_cpu(LDISKFS_SB(osd_sb(dev))->s_es->s_inodes_count);
-       int                   rc;
-       bool                  noslot = true;
+       struct l_wait_info lwi = { 0 };
+       __u32 limit;
+       int rc;
+       bool noslot = true;
        ENTRY;
 
        if (preload)
                goto full;
 
-       param = &scrub->os_iit_param;
+       param = &dev->od_scrub.os_iit_param;
        memset(param, 0, sizeof(*param));
        param->sb = osd_sb(dev);
 
@@ -1305,8 +1077,8 @@ static int osd_inode_iteration(struct osd_thread_info *info,
 
                        if (dev->od_full_scrub_ratio == OFSR_DIRECTLY ||
                            scrub->os_full_scrub) {
-                               osd_scrub_join(dev, SS_AUTO_FULL | SS_RESET,
-                                              true);
+                               osd_scrub_join(info->oti_env, dev,
+                                              SS_AUTO_FULL | SS_RESET, true);
                                goto full;
                        }
 
@@ -1319,7 +1091,7 @@ static int osd_inode_iteration(struct osd_thread_info *info,
                                 * mappings during the partial scan,
                                 * then scan the device completely. */
                                if (used < dev->od_full_scrub_ratio) {
-                                       osd_scrub_join(dev,
+                                       osd_scrub_join(info->oti_env, dev,
                                                SS_AUTO_FULL | SS_RESET, true);
                                        goto full;
                                }
@@ -1392,6 +1164,7 @@ full:
        }
 
        rc = 0;
+       limit = le32_to_cpu(LDISKFS_SB(osd_sb(dev))->s_es->s_inodes_count);
        while (*pos <= limit && *count < max) {
                struct ldiskfs_group_desc *desc;
                bool next_group = false;
@@ -1471,8 +1244,8 @@ out:
 static int osd_otable_it_preload(const struct lu_env *env,
                                 struct osd_otable_it *it)
 {
-       struct osd_device       *dev   = it->ooi_dev;
-       struct osd_scrub        *scrub = &dev->od_scrub;
+       struct osd_device *dev = it->ooi_dev;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct osd_otable_cache *ooc   = &it->ooi_cache;
        int                      rc;
        ENTRY;
@@ -1494,21 +1267,21 @@ static int osd_otable_it_preload(const struct lu_env *env,
 
 static int osd_scrub_main(void *args)
 {
-       struct lu_env         env;
-       struct osd_device    *dev    = (struct osd_device *)args;
-       struct osd_scrub     *scrub  = &dev->od_scrub;
+       struct lu_env env;
+       struct osd_device *dev = (struct osd_device *)args;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct ptlrpc_thread *thread = &scrub->os_thread;
-       int                   rc;
+       int rc;
        ENTRY;
 
-       rc = lu_env_init(&env, LCT_LOCAL);
+       rc = lu_env_init(&env, LCT_LOCAL | LCT_DT_THREAD);
        if (rc != 0) {
                CDEBUG(D_LFSCK, "%s: OI scrub fail to init env: rc = %d\n",
                       osd_scrub2name(scrub), rc);
                GOTO(noenv, rc);
        }
 
-       rc = osd_scrub_prep(dev);
+       rc = osd_scrub_prep(&env, dev);
        if (rc != 0) {
                CDEBUG(D_LFSCK, "%s: OI scrub fail to scrub prep: rc = %d\n",
                       osd_scrub2name(scrub), rc);
@@ -1529,18 +1302,23 @@ static int osd_scrub_main(void *args)
                scrub->os_pos_current = ooc->ooc_pos_preload;
        }
 
-       CDEBUG(D_LFSCK, "%s: OI scrub start, flags = 0x%x, pos = %u\n",
+       CDEBUG(D_LFSCK, "%s: OI scrub start, flags = 0x%x, pos = %llu\n",
               osd_scrub2name(scrub), scrub->os_start_flags,
               scrub->os_pos_current);
 
        rc = osd_inode_iteration(osd_oti_get(&env), dev, ~0U, false);
-       if (unlikely(rc == SCRUB_IT_CRASH))
+       if (unlikely(rc == SCRUB_IT_CRASH)) {
+               spin_lock(&scrub->os_lock);
+               thread_set_flags(&scrub->os_thread, SVC_STOPPING);
+               spin_unlock(&scrub->os_lock);
                GOTO(out, rc = -EINVAL);
+       }
+
        GOTO(post, rc);
 
 post:
-       rc = osd_scrub_post(scrub, rc);
-       CDEBUG(D_LFSCK, "%s: OI scrub: stop, pos = %u: rc = %d\n",
+       rc = osd_scrub_post(&env, dev, rc);
+       CDEBUG(D_LFSCK, "%s: OI scrub: stop, pos = %llu: rc = %d\n",
               osd_scrub2name(scrub), scrub->os_pos_current, rc);
 
 out:
@@ -1548,7 +1326,7 @@ out:
                struct osd_inconsistent_item *oii;
 
                oii = list_entry(scrub->os_inconsistent_items.next,
-                                    struct osd_inconsistent_item, oii_list);
+                                struct osd_inconsistent_item, oii_list);
                list_del_init(&oii->oii_list);
                OBD_FREE_PTR(oii);
        }
@@ -1602,14 +1380,6 @@ static int
 osd_ios_OBJECTS_scan(struct osd_thread_info *info, struct osd_device *dev,
                     struct dentry *dentry, filldir_t filldir);
 
-enum osd_lf_flags {
-       OLF_SCAN_SUBITEMS       = 0x0001,
-       OLF_HIDE_FID            = 0x0002,
-       OLF_SHOW_NAME           = 0x0004,
-       OLF_NO_OI               = 0x0008,
-       OLF_IDX_IN_FID          = 0x0010,
-};
-
 struct osd_lf_map {
        char            *olm_name;
        struct lu_fid    olm_fid;
@@ -1656,8 +1426,8 @@ static const struct osd_lf_map osd_lf_maps[] = {
 
        /* PENDING */
        {
-               .olm_name       = "PENDING",
-               .olm_namelen    = sizeof("PENDING") - 1,
+               .olm_name       = MDT_ORPHAN_DIR,
+               .olm_namelen    = sizeof(MDT_ORPHAN_DIR) - 1,
        },
 
        /* ROOT */
@@ -1793,6 +1563,7 @@ static const struct osd_lf_map osd_lf_maps[] = {
        /* LFSCK */
        {
                .olm_name       = LFSCK_DIR,
+               .olm_flags      = OLF_SCAN_SUBITEMS,
                .olm_namelen    = sizeof(LFSCK_DIR) - 1,
                .olm_scandir    = osd_ios_general_scan,
                .olm_filldir    = osd_ios_varfid_fill,
@@ -1898,6 +1669,29 @@ static const struct osd_lf_map osd_lf_maps[] = {
                .olm_filldir    = osd_ios_lf_fill,
        },
 
+       /* hsm_actions */
+       {
+               .olm_name       = HSM_ACTIONS,
+       },
+
+       /* nodemap */
+       {
+               .olm_name       = LUSTRE_NODEMAP_NAME,
+       },
+
+       /* index_backup */
+       {
+               .olm_name       = INDEX_BACKUP_DIR,
+               .olm_fid        = {
+                       .f_seq  = FID_SEQ_LOCAL_FILE,
+                       .f_oid  = INDEX_BACKUP_OID,
+               },
+               .olm_flags      = OLF_SCAN_SUBITEMS | OLF_NOT_BACKUP,
+               .olm_namelen    = sizeof(INDEX_BACKUP_DIR) - 1,
+               .olm_scandir    = osd_ios_general_scan,
+               .olm_filldir    = osd_ios_varfid_fill,
+       },
+
        {
                .olm_name       = NULL
        }
@@ -1914,6 +1708,7 @@ static const struct osd_lf_map osd_dl_maps[] = {
                },
                .olm_namelen    = sizeof("fid") - 1,
        },
+
        /* .lustre/lost+found */
        {
                .olm_name       = "lost+found",
@@ -1923,6 +1718,7 @@ static const struct osd_lf_map osd_dl_maps[] = {
                },
                .olm_namelen    = sizeof("lost+found") - 1,
        },
+
        {
                .olm_name       = NULL
        }
@@ -1943,6 +1739,7 @@ struct osd_ios_filldir_buf {
        struct osd_thread_info  *oifb_info;
        struct osd_device       *oifb_dev;
        struct dentry           *oifb_dentry;
+       int                      oifb_items;
 };
 
 static inline struct dentry *
@@ -1991,6 +1788,119 @@ osd_ios_new_item(struct osd_device *dev, struct dentry *dentry,
        RETURN(0);
 }
 
+static bool osd_index_need_recreate(const struct lu_env *env,
+                                   struct osd_device *dev, struct inode *inode)
+{
+       struct osd_directory *iam = &osd_oti_get(env)->oti_iam;
+       struct iam_container *bag = &iam->od_container;
+       int rc;
+       ENTRY;
+
+       rc = iam_container_init(bag, &iam->od_descr, inode);
+       if (rc)
+               RETURN(true);
+
+       rc = iam_container_setup(bag);
+       iam_container_fini(bag);
+       if (rc)
+               RETURN(true);
+
+       RETURN(false);
+}
+
+static void osd_ios_index_register(const struct lu_env *env,
+                                  struct osd_device *osd,
+                                  const struct lu_fid *fid,
+                                  struct inode *inode)
+{
+       struct osd_directory *iam = &osd_oti_get(env)->oti_iam;
+       struct iam_container *bag = &iam->od_container;
+       struct super_block *sb = osd_sb(osd);
+       struct iam_descr *descr;
+       __u32 keysize = 0;
+       __u32 recsize = 0;
+       int rc;
+       ENTRY;
+
+       /* Index must be a regular file. */
+       if (!S_ISREG(inode->i_mode))
+               RETURN_EXIT;
+
+       /* Index's size must be block aligned. */
+       if (inode->i_size < sb->s_blocksize ||
+           (inode->i_size & (sb->s_blocksize - 1)) != 0)
+               RETURN_EXIT;
+
+       iam_container_init(bag, &iam->od_descr, inode);
+       rc = iam_container_setup(bag);
+       if (rc)
+               GOTO(fini, rc = 1);
+
+       descr = bag->ic_descr;
+       /* May be regular file with IAM_LFIX_ROOT_MAGIC matched
+        * coincidentally, or corrupted index object, skip it. */
+       if (descr->id_ptr_size != 4)
+               GOTO(fini, rc = 1);
+
+       keysize = descr->id_key_size;
+       recsize = descr->id_rec_size;
+       rc = osd_index_register(osd, fid, keysize, recsize);
+
+       GOTO(fini, rc);
+
+fini:
+       iam_container_fini(bag);
+       if (!rc)
+               CDEBUG(D_LFSCK, "%s: index object "DFID" (%u/%u) registered\n",
+                      osd_name(osd), PFID(fid), keysize, recsize);
+}
+
+static void osd_index_restore(const struct lu_env *env, struct osd_device *dev,
+                             struct lustre_index_restore_unit *liru,
+                             void *buf, int bufsize)
+{
+       struct osd_thread_info *info = osd_oti_get(env);
+       struct osd_inode_id *id = &info->oti_id;
+       struct lu_fid *tgt_fid = &liru->liru_cfid;
+       struct inode *bak_inode = NULL;
+       struct ldiskfs_dir_entry_2 *de = NULL;
+       struct buffer_head *bh = NULL;
+       struct dentry *dentry;
+       char *name = buf;
+       struct lu_fid bak_fid;
+       int rc;
+       ENTRY;
+
+       lustre_fid2lbx(name, tgt_fid, bufsize);
+       dentry = osd_child_dentry_by_inode(env, dev->od_index_backup_inode,
+                                          name, strlen(name));
+       bh = osd_ldiskfs_find_entry(dev->od_index_backup_inode,
+                                   &dentry->d_name, &de, NULL, NULL);
+       if (IS_ERR(bh))
+               GOTO(log, rc = PTR_ERR(bh));
+
+       osd_id_gen(id, le32_to_cpu(de->inode), OSD_OII_NOGEN);
+       brelse(bh);
+       bak_inode = osd_iget_fid(info, dev, id, &bak_fid);
+       if (IS_ERR(bak_inode))
+               GOTO(log, rc = PTR_ERR(bak_inode));
+
+       iput(bak_inode);
+       /* The OI mapping for index may be invalid, since it will be
+        * re-created, not update the OI mapping, just cache it in RAM. */
+       osd_id_gen(id, liru->liru_clid, OSD_OII_NOGEN);
+       osd_add_oi_cache(info, dev, id, tgt_fid);
+       rc = lustre_index_restore(env, &dev->od_dt_dev, &liru->liru_pfid,
+                                 tgt_fid, &bak_fid, liru->liru_name,
+                                 &dev->od_index_backup_list, &dev->od_lock,
+                                 buf, bufsize);
+       GOTO(log, rc);
+
+log:
+       CDEBUG(D_WARNING, "%s: restore index '%s' with "DFID": rc = %d\n",
+              osd_name(dev), liru->liru_name, PFID(tgt_fid), rc);
+}
+
 /**
  * osd_ios_scan_one() - check/fix LMA FID and OI entry for one inode
  *
@@ -2000,12 +1910,14 @@ osd_ios_new_item(struct osd_device *dev, struct dentry *dentry,
  */
 static int
 osd_ios_scan_one(struct osd_thread_info *info, struct osd_device *dev,
-                struct inode *inode, const struct lu_fid *fid, int flags)
+                struct inode *parent, struct inode *inode,
+                const struct lu_fid *fid, const char *name,
+                int namelen, int flags)
 {
        struct lustre_mdt_attrs *lma    = &info->oti_ost_attrs.loa_lma;
        struct osd_inode_id     *id     = &info->oti_id;
        struct osd_inode_id     *id2    = &info->oti_id2;
-       struct osd_scrub        *scrub  = &dev->od_scrub;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct scrub_file       *sf     = &scrub->os_file;
        struct lu_fid            tfid;
        int                      rc;
@@ -2044,6 +1956,28 @@ osd_ios_scan_one(struct osd_thread_info *info, struct osd_device *dev,
                        RETURN(0);
 
                tfid = lma->lma_self_fid;
+               if (lma->lma_compat & LMAC_IDX_BACKUP &&
+                   osd_index_need_recreate(info->oti_env, dev, inode)) {
+                       struct lu_fid *pfid = &info->oti_fid3;
+
+                       if (parent == osd_sb(dev)->s_root->d_inode) {
+                               lu_local_obj_fid(pfid, OSD_FS_ROOT_OID);
+                       } else {
+                               rc = osd_scrub_get_fid(info, dev, parent, pfid,
+                                                      false);
+                               if (rc)
+                                       RETURN(rc);
+                       }
+
+                       rc = lustre_liru_new(&dev->od_index_restore_list, pfid,
+                                       &tfid, inode->i_ino, name, namelen);
+
+                       RETURN(rc);
+               }
+
+               if (!(flags & OLF_NOT_BACKUP))
+                       osd_ios_index_register(info->oti_env, dev, &tfid,
+                                              inode);
        }
 
        rc = osd_oi_lookup(info, dev, &tfid, id2, 0);
@@ -2063,10 +1997,9 @@ osd_ios_scan_one(struct osd_thread_info *info, struct osd_device *dev,
                RETURN(0);
 
        if (!(sf->sf_flags & SF_INCONSISTENT)) {
-               osd_scrub_file_reset(scrub,
-                                    LDISKFS_SB(osd_sb(dev))->s_es->s_uuid,
-                                    SF_INCONSISTENT);
-               rc = osd_scrub_file_store(scrub);
+               scrub_file_reset(scrub, LDISKFS_SB(osd_sb(dev))->s_es->s_uuid,
+                                SF_INCONSISTENT);
+               rc = scrub_file_store(info->oti_env, scrub);
                if (rc != 0)
                        RETURN(rc);
        }
@@ -2081,7 +2014,7 @@ osd_ios_scan_one(struct osd_thread_info *info, struct osd_device *dev,
 
 /**
  * It scans the /lost+found, and for the OST-object (with filter_fid
- * or filter_fid_old), move them back to its proper /O/<seq>/d<x>.
+ * or filter_fid_18_23), move them back to its proper /O/<seq>/d<x>.
  */
 #ifdef HAVE_FILLDIR_USE_CTX
 static int osd_ios_lf_fill(struct dir_context *buf,
@@ -2104,6 +2037,8 @@ static int osd_ios_lf_fill(void *buf,
        int                         rc;
        ENTRY;
 
+       fill_buf->oifb_items++;
+
        /* skip any '.' started names */
        if (name[0] == '.')
                RETURN(0);
@@ -2172,6 +2107,8 @@ static int osd_ios_varfid_fill(void *buf,
        int                         rc;
        ENTRY;
 
+       fill_buf->oifb_items++;
+
        /* skip any '.' started names */
        if (name[0] == '.')
                RETURN(0);
@@ -2180,8 +2117,9 @@ static int osd_ios_varfid_fill(void *buf,
        if (IS_ERR(child))
                RETURN(PTR_ERR(child));
 
-       rc = osd_ios_scan_one(fill_buf->oifb_info, dev, child->d_inode,
-                             NULL, 0);
+       rc = osd_ios_scan_one(fill_buf->oifb_info, dev,
+                             fill_buf->oifb_dentry->d_inode, child->d_inode,
+                             NULL, name, namelen, 0);
        if (rc == 0 && S_ISDIR(child->d_inode->i_mode))
                rc = osd_ios_new_item(dev, child, osd_ios_general_scan,
                                      osd_ios_varfid_fill);
@@ -2206,6 +2144,8 @@ static int osd_ios_dl_fill(void *buf,
        int                         rc       = 0;
        ENTRY;
 
+       fill_buf->oifb_items++;
+
        /* skip any '.' started names */
        if (name[0] == '.')
                RETURN(0);
@@ -2225,8 +2165,9 @@ static int osd_ios_dl_fill(void *buf,
        if (IS_ERR(child))
                RETURN(PTR_ERR(child));
 
-       rc = osd_ios_scan_one(fill_buf->oifb_info, dev, child->d_inode,
-                             &map->olm_fid, map->olm_flags);
+       rc = osd_ios_scan_one(fill_buf->oifb_info, dev,
+                             fill_buf->oifb_dentry->d_inode, child->d_inode,
+                             &map->olm_fid, name, namelen, map->olm_flags);
        dput(child);
 
        RETURN(rc);
@@ -2247,6 +2188,8 @@ static int osd_ios_uld_fill(void *buf,
        int                         rc       = 0;
        ENTRY;
 
+       fill_buf->oifb_items++;
+
        /* skip any non-DFID format name */
        if (name[0] != '[')
                RETURN(0);
@@ -2259,7 +2202,8 @@ static int osd_ios_uld_fill(void *buf,
        sscanf(&name[1], SFID, RFID(&tfid));
        if (fid_is_sane(&tfid))
                rc = osd_ios_scan_one(fill_buf->oifb_info, fill_buf->oifb_dev,
-                                     child->d_inode, &tfid, 0);
+                                     fill_buf->oifb_dentry->d_inode,
+                                     child->d_inode, &tfid, name, namelen, 0);
        else
                rc = -EIO;
        dput(child);
@@ -2283,6 +2227,8 @@ static int osd_ios_root_fill(void *buf,
        int                         rc       = 0;
        ENTRY;
 
+       fill_buf->oifb_items++;
+
        /* skip any '.' started names */
        if (name[0] == '.')
                RETURN(0);
@@ -2303,8 +2249,9 @@ static int osd_ios_root_fill(void *buf,
                RETURN(PTR_ERR(child));
 
        if (!(map->olm_flags & OLF_NO_OI))
-               rc = osd_ios_scan_one(fill_buf->oifb_info, dev, child->d_inode,
-                                     &map->olm_fid, map->olm_flags);
+               rc = osd_ios_scan_one(fill_buf->oifb_info, dev,
+                               fill_buf->oifb_dentry->d_inode, child->d_inode,
+                               &map->olm_fid, name, namelen, map->olm_flags);
        if (rc == 0 && map->olm_flags & OLF_SCAN_SUBITEMS)
                rc = osd_ios_new_item(dev, child, map->olm_scandir,
                                      map->olm_filldir);
@@ -2340,13 +2287,21 @@ osd_ios_general_scan(struct osd_thread_info *info, struct osd_device *dev,
        filp->private_data = NULL;
        set_file_inode(filp, inode);
 
+       do {
+               buf.oifb_items = 0;
 #ifdef HAVE_DIR_CONTEXT
-       buf.ctx.pos = filp->f_pos;
-       rc = fops->iterate(filp, &buf.ctx);
-       filp->f_pos = buf.ctx.pos;
+               buf.ctx.pos = filp->f_pos;
+#ifdef HAVE_ITERATE_SHARED
+               rc = fops->iterate_shared(filp, &buf.ctx);
+#else
+               rc = fops->iterate(filp, &buf.ctx);
+#endif
+               filp->f_pos = buf.ctx.pos;
 #else
-       rc = fops->readdir(filp, &buf, filldir);
+               rc = fops->readdir(filp, &buf, filldir);
 #endif
+       } while (rc >= 0 && buf.oifb_items > 0 &&
+                filp->f_pos != LDISKFS_HTREE_EOF_64BIT);
        fops->release(inode, filp);
 
        RETURN(rc);
@@ -2356,10 +2311,10 @@ static int
 osd_ios_ROOT_scan(struct osd_thread_info *info, struct osd_device *dev,
                  struct dentry *dentry, filldir_t filldir)
 {
-       struct osd_scrub  *scrub  = &dev->od_scrub;
-       struct scrub_file *sf     = &scrub->os_file;
-       struct dentry     *child;
-       int                rc;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
+       struct scrub_file *sf = &scrub->os_file;
+       struct dentry *child;
+       int rc;
        ENTRY;
 
        /* It is existing MDT0 device. We only allow the case of object without
@@ -2383,11 +2338,11 @@ osd_ios_ROOT_scan(struct osd_thread_info *info, struct osd_device *dev,
                if (rc == -ENOENT) {
                        /* It is 1.8 MDT device. */
                        if (!(sf->sf_flags & SF_UPGRADE)) {
-                               osd_scrub_file_reset(scrub,
+                               scrub_file_reset(scrub,
                                        LDISKFS_SB(osd_sb(dev))->s_es->s_uuid,
                                        SF_UPGRADE);
                                sf->sf_internal_flags &= ~SIF_NO_HANDLE_OLD_FID;
-                               rc = osd_scrub_file_store(scrub);
+                               rc = scrub_file_store(info->oti_env, scrub);
                        } else {
                                rc = 0;
                        }
@@ -2416,8 +2371,10 @@ osd_ios_ROOT_scan(struct osd_thread_info *info, struct osd_device *dev,
                 * Usually, it is rare case for the old connected clients
                 * to access the ".lustre" with cached IGIF. So we prefer
                 * to the solution 2). */
-               rc = osd_ios_scan_one(info, dev, child->d_inode,
-                                     &LU_DOT_LUSTRE_FID, 0);
+               rc = osd_ios_scan_one(info, dev, dentry->d_inode,
+                                     child->d_inode, &LU_DOT_LUSTRE_FID,
+                                     dot_lustre_name,
+                                     strlen(dot_lustre_name), 0);
                if (rc == 0)
                        rc = osd_ios_new_item(dev, child, osd_ios_general_scan,
                                              osd_ios_dl_fill);
@@ -2431,22 +2388,24 @@ static int
 osd_ios_OBJECTS_scan(struct osd_thread_info *info, struct osd_device *dev,
                     struct dentry *dentry, filldir_t filldir)
 {
-       struct osd_scrub  *scrub  = &dev->od_scrub;
-       struct scrub_file *sf     = &scrub->os_file;
-       struct dentry     *child;
-       int                rc;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
+       struct scrub_file *sf = &scrub->os_file;
+       struct dentry *child;
+       int rc;
        ENTRY;
 
        if (unlikely(sf->sf_internal_flags & SIF_NO_HANDLE_OLD_FID)) {
                sf->sf_internal_flags &= ~SIF_NO_HANDLE_OLD_FID;
-               rc = osd_scrub_file_store(scrub);
+               rc = scrub_file_store(info->oti_env, scrub);
                if (rc != 0)
                        RETURN(rc);
        }
 
        child = osd_ios_lookup_one_len(ADMIN_USR, dentry, strlen(ADMIN_USR));
        if (!IS_ERR(child)) {
-               rc = osd_ios_scan_one(info, dev, child->d_inode, NULL, 0);
+               rc = osd_ios_scan_one(info, dev, dentry->d_inode,
+                                     child->d_inode, NULL, ADMIN_USR,
+                                     strlen(ADMIN_USR), 0);
                dput(child);
        } else {
                rc = PTR_ERR(child);
@@ -2457,7 +2416,9 @@ osd_ios_OBJECTS_scan(struct osd_thread_info *info, struct osd_device *dev,
 
        child = osd_ios_lookup_one_len(ADMIN_GRP, dentry, strlen(ADMIN_GRP));
        if (!IS_ERR(child)) {
-               rc = osd_ios_scan_one(info, dev, child->d_inode, NULL, 0);
+               rc = osd_ios_scan_one(info, dev, dentry->d_inode,
+                                     child->d_inode, NULL, ADMIN_GRP,
+                                     strlen(ADMIN_GRP), 0);
                dput(child);
        } else {
                rc = PTR_ERR(child);
@@ -2469,30 +2430,26 @@ osd_ios_OBJECTS_scan(struct osd_thread_info *info, struct osd_device *dev,
        RETURN(rc);
 }
 
-static int osd_initial_OI_scrub(struct osd_thread_info *info,
-                               struct osd_device *dev)
+static void osd_initial_OI_scrub(struct osd_thread_info *info,
+                                struct osd_device *dev)
 {
        struct osd_ios_item     *item    = NULL;
        scandir_t                scandir = osd_ios_general_scan;
        filldir_t                filldir = osd_ios_root_fill;
        struct dentry           *dentry  = osd_sb(dev)->s_root;
        const struct osd_lf_map *map     = osd_lf_maps;
-       int                      rc;
        ENTRY;
 
        /* Lookup IGIF in OI by force for initial OI scrub. */
        dev->od_igif_inoi = 1;
 
        while (1) {
-               rc = scandir(info, dev, dentry, filldir);
+               scandir(info, dev, dentry, filldir);
                if (item != NULL) {
                        dput(item->oii_dentry);
                        OBD_FREE_PTR(item);
                }
 
-               if (rc != 0)
-                       break;
-
                if (list_empty(&dev->od_ios_list))
                        break;
 
@@ -2506,17 +2463,6 @@ static int osd_initial_OI_scrub(struct osd_thread_info *info,
                dentry = item->oii_dentry;
        }
 
-       while (!list_empty(&dev->od_ios_list)) {
-               item = list_entry(dev->od_ios_list.next,
-                                 struct osd_ios_item, oii_list);
-               list_del_init(&item->oii_list);
-               dput(item->oii_dentry);
-               OBD_FREE_PTR(item);
-       }
-
-       if (rc != 0)
-               RETURN(rc);
-
        /* There maybe the case that the object has been removed, but its OI
         * mapping is still in the OI file, such as the "CATALOGS" after MDT
         * file-level backup/restore. So here cleanup the stale OI mappings. */
@@ -2540,7 +2486,32 @@ static int osd_initial_OI_scrub(struct osd_thread_info *info,
                map++;
        }
 
-       RETURN(0);
+       if (!list_empty(&dev->od_index_restore_list)) {
+               char *buf;
+
+               OBD_ALLOC_LARGE(buf, INDEX_BACKUP_BUFSIZE);
+               if (!buf)
+                       CERROR("%s: not enough RAM for rebuild index\n",
+                              osd_name(dev));
+
+               while (!list_empty(&dev->od_index_restore_list)) {
+                       struct lustre_index_restore_unit *liru;
+
+                       liru = list_entry(dev->od_index_restore_list.next,
+                                         struct lustre_index_restore_unit,
+                                         liru_link);
+                       list_del(&liru->liru_link);
+                       if (buf)
+                               osd_index_restore(info->oti_env, dev, liru,
+                                                 buf, INDEX_BACKUP_BUFSIZE);
+                       OBD_FREE(liru, liru->liru_len);
+               }
+
+               if (buf)
+                       OBD_FREE_LARGE(buf, INDEX_BACKUP_BUFSIZE);
+       }
+
+       EXIT;
 }
 
 char *osd_lf_fid2name(const struct lu_fid *fid)
@@ -2564,115 +2535,39 @@ char *osd_lf_fid2name(const struct lu_fid *fid)
 
 /* OI scrub start/stop */
 
-static int do_osd_scrub_start(struct osd_device *dev, __u32 flags)
+int osd_scrub_start(const struct lu_env *env, struct osd_device *dev,
+                   __u32 flags)
 {
-       struct osd_scrub     *scrub  = &dev->od_scrub;
-       struct ptlrpc_thread *thread = &scrub->os_thread;
-       struct l_wait_info    lwi    = { 0 };
-       struct task_struct   *task;
-       int                   rc;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
+       int rc;
        ENTRY;
 
        if (dev->od_dt_dev.dd_rdonly)
                RETURN(-EROFS);
 
-       /* os_lock: sync status between stop and scrub thread */
-       spin_lock(&scrub->os_lock);
-
-again:
-       if (thread_is_running(thread)) {
-               spin_unlock(&scrub->os_lock);
-               if (!(scrub->os_file.sf_flags & SF_AUTO ||
-                     scrub->os_partial_scan) ||
-                    (flags & SS_AUTO_PARTIAL))
-                       RETURN(-EALREADY);
-
-               osd_scrub_join(dev, flags, false);
-               spin_lock(&scrub->os_lock);
-               if (!thread_is_running(thread))
-                       goto again;
-
-               spin_unlock(&scrub->os_lock);
-               RETURN(0);
-       }
-
-       if (unlikely(thread_is_stopping(thread))) {
-               spin_unlock(&scrub->os_lock);
-               l_wait_event(thread->t_ctl_waitq,
-                            thread_is_stopped(thread),
-                            &lwi);
-               spin_lock(&scrub->os_lock);
-               goto again;
-       }
-       spin_unlock(&scrub->os_lock);
-
-       if (scrub->os_file.sf_status == SS_COMPLETED) {
-               if (!(flags & SS_SET_FAILOUT))
-                       flags |= SS_CLEAR_FAILOUT;
-
-               if (!(flags & SS_SET_DRYRUN))
-                       flags |= SS_CLEAR_DRYRUN;
-
-               flags |= SS_RESET;
-       }
-
-       scrub->os_start_flags = flags;
-       thread_set_flags(thread, 0);
-       task = kthread_run(osd_scrub_main, dev, "OI_scrub");
-       if (IS_ERR(task)) {
-               rc = PTR_ERR(task);
-               CERROR("%s: cannot start iteration thread: rc = %d\n",
-                      osd_scrub2name(scrub), rc);
-               RETURN(rc);
-       }
-
-       l_wait_event(thread->t_ctl_waitq,
-                    thread_is_running(thread) || thread_is_stopped(thread),
-                    &lwi);
-
-       RETURN(0);
-}
-
-int osd_scrub_start(struct osd_device *dev, __u32 flags)
-{
-       int rc;
-       ENTRY;
-
        /* od_otable_mutex: prevent curcurrent start/stop */
        mutex_lock(&dev->od_otable_mutex);
-       rc = do_osd_scrub_start(dev, flags);
+       rc = scrub_start(osd_scrub_main, scrub, dev, flags);
+       if (rc == -EALREADY) {
+               rc = 0;
+               if ((scrub->os_file.sf_flags & SF_AUTO ||
+                    scrub->os_partial_scan) &&
+                   !(flags & SS_AUTO_PARTIAL))
+                       osd_scrub_join(env, dev, flags, false);
+       }
        mutex_unlock(&dev->od_otable_mutex);
 
-       RETURN(rc == -EALREADY ? 0 : rc);
+       RETURN(rc);
 }
 
-static void do_osd_scrub_stop(struct osd_scrub *scrub)
+void osd_scrub_stop(struct osd_device *dev)
 {
-       struct ptlrpc_thread *thread = &scrub->os_thread;
-       struct l_wait_info    lwi    = { 0 };
-
-       /* os_lock: sync status between stop and scrub thread */
-       spin_lock(&scrub->os_lock);
-       if (!thread_is_init(thread) && !thread_is_stopped(thread)) {
-               thread_set_flags(thread, SVC_STOPPING);
-               spin_unlock(&scrub->os_lock);
-               wake_up_all(&thread->t_ctl_waitq);
-               l_wait_event(thread->t_ctl_waitq,
-                            thread_is_stopped(thread),
-                            &lwi);
-               /* Do not skip the last lock/unlock, which can guarantee that
-                * the caller cannot return until the OI scrub thread exit. */
-               spin_lock(&scrub->os_lock);
-       }
-       spin_unlock(&scrub->os_lock);
-}
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
 
-static void osd_scrub_stop(struct osd_device *dev)
-{
        /* od_otable_mutex: prevent curcurrent start/stop */
        mutex_lock(&dev->od_otable_mutex);
-       dev->od_scrub.os_paused = 1;
-       do_osd_scrub_stop(&dev->od_scrub);
+       scrub->os_paused = 1;
+       scrub_stop(scrub);
        mutex_unlock(&dev->od_otable_mutex);
 }
 
@@ -2682,22 +2577,24 @@ static const char osd_scrub_name[] = "OI_scrub";
 
 int osd_scrub_setup(const struct lu_env *env, struct osd_device *dev)
 {
-       struct osd_thread_info     *info   = osd_oti_get(env);
-       struct osd_scrub           *scrub  = &dev->od_scrub;
-       struct lvfs_run_ctxt       *ctxt   = &scrub->os_ctxt;
-       struct scrub_file          *sf     = &scrub->os_file;
-       struct super_block         *sb     = osd_sb(dev);
-       struct ldiskfs_super_block *es     = LDISKFS_SB(sb)->s_es;
-       struct lvfs_run_ctxt        saved;
-       struct file                *filp;
-       struct inode               *inode;
-       struct lu_fid              *fid    = &info->oti_fid;
-       bool                        dirty  = false;
-       bool                        restored = false;
-       int                         rc     = 0;
+       struct osd_thread_info *info = osd_oti_get(env);
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
+       struct lvfs_run_ctxt *ctxt = &dev->od_scrub.os_ctxt;
+       struct scrub_file *sf = &scrub->os_file;
+       struct super_block *sb = osd_sb(dev);
+       struct ldiskfs_super_block *es = LDISKFS_SB(sb)->s_es;
+       struct lvfs_run_ctxt saved;
+       struct file *filp;
+       struct inode *inode;
+       struct lu_fid *fid = &info->oti_fid;
+       struct osd_inode_id *id = &info->oti_id;
+       struct dt_object *obj;
+       bool dirty = false;
+       bool restored = false;
+       int rc = 0;
        ENTRY;
 
-       memset(scrub, 0, sizeof(*scrub));
+       memset(&dev->od_scrub, 0, sizeof(struct osd_scrub));
        OBD_SET_CTXT_MAGIC(ctxt);
        ctxt->pwdmnt = dev->od_mnt;
        ctxt->pwd = dev->od_mnt->mnt_root;
@@ -2707,32 +2604,42 @@ int osd_scrub_setup(const struct lu_env *env, struct osd_device *dev)
        init_rwsem(&scrub->os_rwsem);
        spin_lock_init(&scrub->os_lock);
        INIT_LIST_HEAD(&scrub->os_inconsistent_items);
+       scrub->os_name = osd_name(dev);
 
        push_ctxt(&saved, ctxt);
-       filp = filp_open(osd_scrub_name, O_RDWR | O_CREAT, 0644);
+       filp = filp_open(osd_scrub_name, O_RDWR |
+                        (dev->od_dt_dev.dd_rdonly ? 0 : O_CREAT), 0644);
        if (IS_ERR(filp)) {
                pop_ctxt(&saved, ctxt);
                RETURN(PTR_ERR(filp));
        }
 
        inode = file_inode(filp);
-       /* 'What the @fid is' is not imporatant, because the object
-        * has no OI mapping, and only is visible inside the OSD.*/
-       lu_igif_build(fid, inode->i_ino, inode->i_generation);
-       rc = osd_ea_fid_set(info, inode, fid, LMAC_NOT_IN_OI, 0);
-       if (rc != 0) {
-               filp_close(filp, NULL);
-               pop_ctxt(&saved, ctxt);
-               RETURN(rc);
+       if (!dev->od_dt_dev.dd_rdonly) {
+               /* 'What the @fid is' is not imporatant, because the object
+                * has no OI mapping, and only is visible inside the OSD.*/
+               lu_igif_build(fid, inode->i_ino, inode->i_generation);
+               rc = osd_ea_fid_set(info, inode, fid, LMAC_NOT_IN_OI, 0);
+               if (rc) {
+                       filp_close(filp, NULL);
+                       pop_ctxt(&saved, ctxt);
+                       RETURN(rc);
+               }
        }
 
-       scrub->os_inode = igrab(inode);
+       osd_id_gen(id, inode->i_ino, inode->i_generation);
+       osd_add_oi_cache(info, dev, id, fid);
        filp_close(filp, NULL);
        pop_ctxt(&saved, ctxt);
 
-       rc = osd_scrub_file_load(scrub);
-       if (rc == -ENOENT) {
-               osd_scrub_file_init(scrub, es->s_uuid);
+       obj = lu2dt(lu_object_find_slice(env, osd2lu_dev(dev), fid, NULL));
+       if (IS_ERR_OR_NULL(obj))
+               RETURN(obj ? PTR_ERR(obj) : -ENOENT);
+
+       scrub->os_obj = obj;
+       rc = scrub_file_load(env, scrub);
+       if (rc == -ENOENT || rc == -EFAULT) {
+               scrub_file_init(scrub, es->s_uuid);
                /* If the "/O" dir does not exist when mount (indicated by
                 * osd_device::od_maybe_new), neither for the "/OI_scrub",
                 * then it is quite probably that the device is a new one,
@@ -2742,14 +2649,14 @@ int osd_scrub_setup(const struct lu_env *env, struct osd_device *dev)
                 * an old device, it can be found and cleared later.
                 *
                 * For the system with "SIF_NO_HANDLE_OLD_FID", we do not
-                * need to check "filter_fid_old" and to convert it to
+                * need to check "filter_fid_18_23" and to convert it to
                 * "filter_fid" for each object, and all the IGIF should
                 * have their FID mapping in OI files already. */
-               if (dev->od_maybe_new)
+               if (dev->od_maybe_new && rc == -ENOENT)
                        sf->sf_internal_flags = SIF_NO_HANDLE_OLD_FID;
                dirty = true;
-       } else if (rc != 0) {
-               GOTO(cleanup_inode, rc);
+       } else if (rc < 0) {
+               GOTO(cleanup_obj, rc);
        } else {
                if (memcmp(sf->sf_uuid, es->s_uuid, 16) != 0) {
                        struct obd_uuid *old_uuid;
@@ -2764,11 +2671,11 @@ int osd_scrub_setup(const struct lu_env *env, struct osd_device *dev)
                        } else {
                                class_uuid_unparse(sf->sf_uuid, old_uuid);
                                class_uuid_unparse(es->s_uuid, new_uuid);
-                               CERROR("%s: UUID has been changed from "
-                                      "%s to %s\n", osd_dev2name(dev),
+                               CDEBUG(D_LFSCK, "%s: UUID has been changed "
+                                      "from %s to %s\n", osd_dev2name(dev),
                                       old_uuid->uuid, new_uuid->uuid);
                        }
-                       osd_scrub_file_reset(scrub, es->s_uuid,SF_INCONSISTENT);
+                       scrub_file_reset(scrub, es->s_uuid, SF_INCONSISTENT);
                        dirty = true;
                        restored = true;
                        if (old_uuid != NULL)
@@ -2779,6 +2686,14 @@ int osd_scrub_setup(const struct lu_env *env, struct osd_device *dev)
                        sf->sf_status = SS_CRASHED;
                        dirty = true;
                }
+
+               if ((sf->sf_oi_count & (sf->sf_oi_count - 1)) != 0) {
+                       LCONSOLE_WARN("%s: invalid oi count %d, set it to %d\n",
+                                     osd_dev2name(dev), sf->sf_oi_count,
+                                     osd_oi_count);
+                       sf->sf_oi_count = osd_oi_count;
+                       dirty = true;
+               }
        }
 
        if (sf->sf_pos_last_checkpoint != 0)
@@ -2787,19 +2702,18 @@ int osd_scrub_setup(const struct lu_env *env, struct osd_device *dev)
                scrub->os_pos_current = LDISKFS_FIRST_INO(sb) + 1;
 
        if (dirty) {
-               rc = osd_scrub_file_store(scrub);
-               if (rc != 0)
-                       GOTO(cleanup_inode, rc);
+               rc = scrub_file_store(env, scrub);
+               if (rc)
+                       GOTO(cleanup_obj, rc);
        }
 
        /* Initialize OI files. */
        rc = osd_oi_init(info, dev, restored);
        if (rc < 0)
-               GOTO(cleanup_inode, rc);
+               GOTO(cleanup_obj, rc);
 
-       rc = osd_initial_OI_scrub(info, dev);
-       if (rc != 0)
-               GOTO(cleanup_oi, rc);
+       if (!dev->od_dt_dev.dd_rdonly)
+               osd_initial_OI_scrub(info, dev);
 
        if (sf->sf_flags & SF_UPGRADE ||
            !(sf->sf_internal_flags & SIF_NO_HANDLE_OLD_FID ||
@@ -2822,7 +2736,8 @@ int osd_scrub_setup(const struct lu_env *env, struct osd_device *dev)
                 * later if found that the system is upgrading. */
                dev->od_igif_inoi = 1;
 
-       if (!dev->od_dt_dev.dd_rdonly && !dev->od_noscrub &&
+       if (!dev->od_dt_dev.dd_rdonly &&
+           dev->od_auto_scrub_interval != AS_NEVER &&
            ((sf->sf_status == SS_PAUSED) ||
             (sf->sf_status == SS_CRASHED &&
              sf->sf_flags & (SF_RECREATED | SF_INCONSISTENT |
@@ -2830,7 +2745,7 @@ int osd_scrub_setup(const struct lu_env *env, struct osd_device *dev)
             (sf->sf_status == SS_INIT &&
              sf->sf_flags & (SF_RECREATED | SF_INCONSISTENT |
                              SF_UPGRADE))))
-               rc = osd_scrub_start(dev, SS_AUTO_FULL);
+               rc = osd_scrub_start(env, dev, SS_AUTO_FULL);
 
        if (rc != 0)
                GOTO(cleanup_oi, rc);
@@ -2845,23 +2760,23 @@ int osd_scrub_setup(const struct lu_env *env, struct osd_device *dev)
        RETURN(0);
 cleanup_oi:
        osd_oi_fini(info, dev);
-cleanup_inode:
-       iput(scrub->os_inode);
-       scrub->os_inode = NULL;
+cleanup_obj:
+       dt_object_put_nocache(env, scrub->os_obj);
+       scrub->os_obj = NULL;
 
        return rc;
 }
 
 void osd_scrub_cleanup(const struct lu_env *env, struct osd_device *dev)
 {
-       struct osd_scrub *scrub = &dev->od_scrub;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
 
        LASSERT(dev->od_otable_it == NULL);
 
-       if (scrub->os_inode != NULL) {
+       if (scrub->os_obj != NULL) {
                osd_scrub_stop(dev);
-               iput(scrub->os_inode);
-               scrub->os_inode = NULL;
+               dt_object_put_nocache(env, scrub->os_obj);
+               scrub->os_obj = NULL;
        }
        if (dev->od_oi_table != NULL)
                osd_oi_fini(osd_oti_get(env), dev);
@@ -2875,7 +2790,7 @@ static struct dt_it *osd_otable_it_init(const struct lu_env *env,
        enum dt_otable_it_flags flags = attr >> DT_OTABLE_IT_FLAGS_SHIFT;
        enum dt_otable_it_valid valid = attr & ~DT_OTABLE_IT_FLAGS_MASK;
        struct osd_device      *dev   = osd_dev(dt->do_lu.lo_dev);
-       struct osd_scrub       *scrub = &dev->od_scrub;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct osd_otable_it   *it;
        __u32                   start = 0;
        int                     rc;
@@ -2913,15 +2828,19 @@ static struct dt_it *osd_otable_it_init(const struct lu_env *env,
                        start |= SS_CLEAR_DRYRUN;
        }
 
-       rc = do_osd_scrub_start(dev, start & ~SS_AUTO_PARTIAL);
-       if (rc < 0 && rc != -EALREADY) {
+       rc = scrub_start(osd_scrub_main, scrub, dev, start & ~SS_AUTO_PARTIAL);
+       if (rc == -EALREADY) {
+               it->ooi_cache.ooc_pos_preload = scrub->os_pos_current;
+       } else  if (rc < 0) {
                dev->od_otable_it = NULL;
                OBD_FREE_PTR(it);
-               GOTO(out, it = ERR_PTR(rc));
+               it = ERR_PTR(rc);
+       } else {
+               /* We have to start from the begining. */
+               it->ooi_cache.ooc_pos_preload =
+                       LDISKFS_FIRST_INO(osd_sb(dev)) + 1;
        }
 
-       it->ooi_cache.ooc_pos_preload = scrub->os_pos_current;
-
        GOTO(out, it);
 
 out:
@@ -2936,7 +2855,7 @@ static void osd_otable_it_fini(const struct lu_env *env, struct dt_it *di)
 
        /* od_otable_mutex: prevent curcurrent init/fini */
        mutex_lock(&dev->od_otable_mutex);
-       do_osd_scrub_stop(&dev->od_scrub);
+       scrub_stop(&dev->od_scrub.os_scrub);
        LASSERT(dev->od_otable_it == it);
 
        dev->od_otable_it = NULL;
@@ -2955,7 +2874,7 @@ static void osd_otable_it_put(const struct lu_env *env, struct dt_it *di)
 }
 
 static inline int
-osd_otable_it_wakeup(struct osd_scrub *scrub, struct osd_otable_it *it)
+osd_otable_it_wakeup(struct lustre_scrub *scrub, struct osd_otable_it *it)
 {
        spin_lock(&scrub->os_lock);
        if (it->ooi_cache.ooc_pos_preload < scrub->os_pos_current ||
@@ -2973,7 +2892,7 @@ static int osd_otable_it_next(const struct lu_env *env, struct dt_it *di)
 {
        struct osd_otable_it    *it     = (struct osd_otable_it *)di;
        struct osd_device       *dev    = it->ooi_dev;
-       struct osd_scrub        *scrub  = &dev->od_scrub;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct osd_otable_cache *ooc    = &it->ooi_cache;
        struct ptlrpc_thread    *thread = &scrub->os_thread;
        struct l_wait_info       lwi    = { 0 };
@@ -3074,7 +2993,7 @@ static int osd_otable_it_load(const struct lu_env *env,
        struct osd_otable_it    *it    = (struct osd_otable_it *)di;
        struct osd_device       *dev   = it->ooi_dev;
        struct osd_otable_cache *ooc   = &it->ooi_cache;
-       struct osd_scrub        *scrub = &dev->od_scrub;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct osd_iit_param    *param = &it->ooi_iit_param;
        int                      rc;
        ENTRY;
@@ -3143,9 +3062,10 @@ int osd_oii_insert(struct osd_device *dev, struct osd_idmap_cache *oic,
                   int insert)
 {
        struct osd_inconsistent_item *oii;
-       struct osd_scrub             *scrub  = &dev->od_scrub;
-       struct ptlrpc_thread         *thread = &scrub->os_thread;
-       int                           wakeup = 0;
+       struct osd_scrub *oscrub = &dev->od_scrub;
+       struct lustre_scrub *lscrub = &oscrub->os_scrub;
+       struct ptlrpc_thread *thread = &lscrub->os_thread;
+       int wakeup = 0;
        ENTRY;
 
        OBD_ALLOC_PTR(oii);
@@ -3156,37 +3076,37 @@ int osd_oii_insert(struct osd_device *dev, struct osd_idmap_cache *oic,
        oii->oii_cache = *oic;
        oii->oii_insert = insert;
 
-       if (scrub->os_partial_scan) {
-               __u64 now = cfs_time_current_sec();
+       if (lscrub->os_partial_scan) {
+               __u64 now = ktime_get_real_seconds();
 
                /* If there haven't been errors in a long time,
                 * decay old count until either the errors are
                 * gone or we reach the current interval. */
-               while (unlikely(scrub->os_bad_oimap_count > 0 &&
-                               scrub->os_bad_oimap_time +
+               while (unlikely(oscrub->os_bad_oimap_count > 0 &&
+                               oscrub->os_bad_oimap_time +
                                SCRUB_BAD_OIMAP_DECAY_INTERVAL < now)) {
-                       scrub->os_bad_oimap_count >>= 1;
-                       scrub->os_bad_oimap_time +=
+                       oscrub->os_bad_oimap_count >>= 1;
+                       oscrub->os_bad_oimap_time +=
                                SCRUB_BAD_OIMAP_DECAY_INTERVAL;
                }
 
-               scrub->os_bad_oimap_time = now;
-               if (++scrub->os_bad_oimap_count >
+               oscrub->os_bad_oimap_time = now;
+               if (++oscrub->os_bad_oimap_count >
                    dev->od_full_scrub_threshold_rate)
-                       scrub->os_full_scrub = 1;
+                       lscrub->os_full_scrub = 1;
        }
 
-       spin_lock(&scrub->os_lock);
+       spin_lock(&lscrub->os_lock);
        if (unlikely(!thread_is_running(thread))) {
-               spin_unlock(&scrub->os_lock);
+               spin_unlock(&lscrub->os_lock);
                OBD_FREE_PTR(oii);
                RETURN(-EAGAIN);
        }
 
-       if (list_empty(&scrub->os_inconsistent_items))
+       if (list_empty(&lscrub->os_inconsistent_items))
                wakeup = 1;
-       list_add_tail(&oii->oii_list, &scrub->os_inconsistent_items);
-       spin_unlock(&scrub->os_lock);
+       list_add_tail(&oii->oii_list, &lscrub->os_inconsistent_items);
+       spin_unlock(&lscrub->os_lock);
 
        if (wakeup != 0)
                wake_up_all(&thread->t_ctl_waitq);
@@ -3197,7 +3117,7 @@ int osd_oii_insert(struct osd_device *dev, struct osd_idmap_cache *oic,
 int osd_oii_lookup(struct osd_device *dev, const struct lu_fid *fid,
                   struct osd_inode_id *id)
 {
-       struct osd_scrub             *scrub = &dev->od_scrub;
+       struct lustre_scrub *scrub = &dev->od_scrub.os_scrub;
        struct osd_inconsistent_item *oii;
        ENTRY;
 
@@ -3214,175 +3134,17 @@ int osd_oii_lookup(struct osd_device *dev, const struct lu_fid *fid,
        RETURN(-ENOENT);
 }
 
-/* OI scrub dump */
-
-static const char *scrub_status_names[] = {
-       "init",
-       "scanning",
-       "completed",
-       "failed",
-       "stopped",
-       "paused",
-       "crashed",
-       NULL
-};
-
-static const char *scrub_flags_names[] = {
-       "recreated",
-       "inconsistent",
-       "auto",
-       "upgrade",
-       NULL
-};
-
-static const char *scrub_param_names[] = {
-       "failout",
-       "dryrun",
-       NULL
-};
-
-static void scrub_bits_dump(struct seq_file *m, int bits, const char *names[],
-                           const char *prefix)
+void osd_scrub_dump(struct seq_file *m, struct osd_device *dev)
 {
-       int flag;
-       int i;
-
-       seq_printf(m, "%s:%c", prefix, bits != 0 ? ' ' : '\n');
-
-       for (i = 0, flag = 1; bits != 0; i++, flag = 1 << i) {
-               if (flag & bits) {
-                       bits &= ~flag;
-                       seq_printf(m, "%s%c", names[i],
-                                  bits != 0 ? ',' : '\n');
-               }
-       }
-}
-
-static void scrub_time_dump(struct seq_file *m, __u64 time, const char *prefix)
-{
-       if (time != 0)
-               seq_printf(m, "%s: %llu seconds\n", prefix,
-                          cfs_time_current_sec() - time);
-       else
-               seq_printf(m, "%s: N/A\n", prefix);
-}
-
-static void scrub_pos_dump(struct seq_file *m, __u64 pos, const char *prefix)
-{
-       if (pos != 0)
-               seq_printf(m, "%s: %llu\n", prefix, pos);
-       else
-               seq_printf(m, "%s: N/A\n", prefix);
-}
+       struct osd_scrub *scrub = &dev->od_scrub;
 
-int osd_scrub_dump(struct seq_file *m, struct osd_device *dev)
-{
-       struct osd_scrub  *scrub   = &dev->od_scrub;
-       struct scrub_file *sf      = &scrub->os_file;
-       __u64              checked;
-       __u64              speed;
-
-       down_read(&scrub->os_rwsem);
-       seq_printf(m, "name: OI_scrub\n"
-                  "magic: 0x%x\n"
-                  "oi_files: %d\n"
-                  "status: %s\n",
-                  sf->sf_magic, (int)sf->sf_oi_count,
-                  scrub_status_names[sf->sf_status]);
-
-       scrub_bits_dump(m, sf->sf_flags, scrub_flags_names, "flags");
-
-       scrub_bits_dump(m, sf->sf_param, scrub_param_names, "param");
-
-       scrub_time_dump(m, sf->sf_time_last_complete,
-                       "time_since_last_completed");
-
-       scrub_time_dump(m, sf->sf_time_latest_start,
-                       "time_since_latest_start");
-
-       scrub_time_dump(m, sf->sf_time_last_checkpoint,
-                       "time_since_last_checkpoint");
-
-       scrub_pos_dump(m, sf->sf_pos_latest_start,
-                       "latest_start_position");
-
-       scrub_pos_dump(m, sf->sf_pos_last_checkpoint,
-                       "last_checkpoint_position");
-
-       scrub_pos_dump(m, sf->sf_pos_first_inconsistent,
-                       "first_failure_position");
-
-       checked = sf->sf_items_checked + scrub->os_new_checked;
-       seq_printf(m, "checked: %llu\n"
-                  "%s: %llu\n"
-                  "failed: %llu\n"
-                  "prior_%s: %llu\n"
-                  "noscrub: %llu\n"
-                  "igif: %llu\n"
-                  "success_count: %u\n",
-                  checked,
-                  sf->sf_param & SP_DRYRUN ? "inconsistent" : "updated",
-                  sf->sf_items_updated, sf->sf_items_failed,
-                  sf->sf_param & SP_DRYRUN ? "inconsistent" : "updated",
-                  sf->sf_items_updated_prior, sf->sf_items_noscrub,
-                  sf->sf_items_igif, sf->sf_success_count);
-
-       speed = checked;
-       if (thread_is_running(&scrub->os_thread)) {
-               cfs_duration_t duration = cfs_time_current() -
-                                         scrub->os_time_last_checkpoint;
-               __u64 new_checked = msecs_to_jiffies(scrub->os_new_checked *
-                                                    MSEC_PER_SEC);
-               __u32 rtime = sf->sf_run_time +
-                             cfs_duration_sec(duration + HALF_SEC);
-
-               if (duration != 0)
-                       do_div(new_checked, duration);
-               if (rtime != 0)
-                       do_div(speed, rtime);
-               seq_printf(m, "run_time: %u seconds\n"
-                          "average_speed: %llu objects/sec\n"
-                          "real-time_speed: %llu objects/sec\n"
-                          "current_position: %u\n"
-                          "lf_scanned: %llu\n"
-                          "lf_%s: %llu\n"
-                          "lf_failed: %llu\n",
-                          rtime, speed, new_checked, scrub->os_pos_current,
-                          scrub->os_lf_scanned,
-                          sf->sf_param & SP_DRYRUN ?
-                               "inconsistent" : "repaired",
-                          scrub->os_lf_repaired,
-                          scrub->os_lf_failed);
-               seq_printf(m, "inodes_per_group: %lu\n"
-                          "current_iit_group: %u\n"
-                          "current_iit_base: %u\n"
-                          "current_iit_offset: %u\n"
-                          "scrub_in_prior: %s\n"
-                          "scrub_full_speed: %s\n"
-                          "partial_scan: %s\n",
-                          LDISKFS_INODES_PER_GROUP(osd_sb(dev)),
-                          scrub->os_iit_param.bg,
-                          scrub->os_iit_param.gbase,
-                          scrub->os_iit_param.offset,
-                          scrub->os_in_prior ? "yes" : "no",
-                          scrub->os_full_speed ? "yes" : "no",
-                          scrub->os_partial_scan ? "yes" : "no");
-       } else {
-               if (sf->sf_run_time != 0)
-                       do_div(speed, sf->sf_run_time);
-               seq_printf(m, "run_time: %u seconds\n"
-                          "average_speed: %llu objects/sec\n"
-                          "real-time_speed: N/A\n"
-                          "current_position: N/A\n"
-                          "lf_scanned: %llu\n"
-                          "lf_%s: %llu\n"
-                          "lf_failed: %llu\n",
-                          sf->sf_run_time, speed, scrub->os_lf_scanned,
-                          sf->sf_param & SP_DRYRUN ?
-                               "inconsistent" : "repaired",
-                          scrub->os_lf_repaired, scrub->os_lf_failed);
-       }
-
-       up_read(&scrub->os_rwsem);
-       return 0;
+       scrub_dump(m, &scrub->os_scrub);
+       seq_printf(m, "lf_scanned: %llu\n"
+                  "lf_%s: %llu\n"
+                  "lf_failed: %llu\n",
+                  scrub->os_lf_scanned,
+                  scrub->os_scrub.os_file.sf_param & SP_DRYRUN ?
+                       "inconsistent" : "repaired",
+                  scrub->os_lf_repaired,
+                  scrub->os_lf_failed);
 }