/*
 * GPL HEADER START
 *
 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
 *
 * 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 LICENSE file that accompanied this code).
 *
 * You should have received a copy of the GNU General Public License
 * version 2 along with this program; If not, see
 * http://www.gnu.org/licenses/gpl-2.0.html
 *
 * GPL HEADER END
 */
/*
 * Copyright (c) 2007, 2010, Oracle and/or its affiliates. All rights reserved.
 * Use is subject to license terms.
 *
 * Copyright (c) 2011, 2017, Intel Corporation.
 */
/*
 * This file is part of Lustre, http://www.lustre.org/
 * Lustre is a trademark of Sun Microsystems, Inc.
 *
 * lustre/mdd/mdd_device.c
 *
 * Lustre Metadata Server (mdd) routines
 *
 * Author: Wang Di <wangdi@clusterfs.com>
 */

#define DEBUG_SUBSYSTEM S_MDS

#include <linux/module.h>
#include <linux/kthread.h>
#include <obd_class.h>
#include <uapi/linux/lustre/lustre_ioctl.h>
#include <lustre_mds.h>
#include <obd_support.h>
#include <lu_object.h>
#include <uapi/linux/lustre/lustre_param.h>
#include <lustre_fid.h>
#include <lustre_nodemap.h>
#include <lustre_barrier.h>

#include "mdd_internal.h"

static const struct md_device_operations mdd_ops;
static struct lu_device_type mdd_device_type;

static const char mdd_root_dir_name[] = "ROOT";
static const char mdd_obf_dir_name[] = "fid";
static const char mdd_lpf_dir_name[] = "lost+found";

/* Slab for MDD object allocation */
struct kmem_cache *mdd_object_kmem;

static struct lu_kmem_descr mdd_caches[] = {
	{
		.ckd_cache = &mdd_object_kmem,
		.ckd_name  = "mdd_obj",
		.ckd_size  = sizeof(struct mdd_object)
	},
	{
		.ckd_cache = NULL
	}
};

static int mdd_connect_to_next(const struct lu_env *env, struct mdd_device *m,
			       const char *nextdev)
{
	struct obd_connect_data *data = NULL;
	struct lu_device	*lud = mdd2lu_dev(m);
	struct obd_device       *obd;
	int			 rc;
	ENTRY;

	LASSERT(m->mdd_child_exp == NULL);

	OBD_ALLOC(data, sizeof(*data));
	if (data == NULL)
		GOTO(out, rc = -ENOMEM);

	obd = class_name2obd(nextdev);
	if (obd == NULL) {
		CERROR("can't locate next device: %s\n", nextdev);
		GOTO(out, rc = -ENOTCONN);
	}

	data->ocd_connect_flags = OBD_CONNECT_VERSION;
	data->ocd_version = LUSTRE_VERSION_CODE;

	rc = obd_connect(NULL, &m->mdd_child_exp, obd, &obd->obd_uuid, data, NULL);
	if (rc) {
		CERROR("cannot connect to next dev %s (%d)\n", nextdev, rc);
		GOTO(out, rc);
	}

	lud->ld_site = m->mdd_child_exp->exp_obd->obd_lu_dev->ld_site;
	LASSERT(lud->ld_site);
	m->mdd_child = lu2dt_dev(m->mdd_child_exp->exp_obd->obd_lu_dev);
	m->mdd_bottom = lu2dt_dev(lud->ld_site->ls_bottom_dev);
	lu_dev_add_linkage(lud->ld_site, lud);

out:
	if (data)
		OBD_FREE(data, sizeof(*data));
	RETURN(rc);
}

static int mdd_init0(const struct lu_env *env, struct mdd_device *mdd,
		struct lu_device_type *t, struct lustre_cfg *lcfg)
{
	int rc = -EINVAL;
	const char *dev;
	ENTRY;

	/* LU-8040 Set defaults here, before values configs */
	mdd->mdd_cl.mc_flags = 0; /* off by default */
	mdd->mdd_cl.mc_mask = CHANGELOG_DEFMASK;
	mdd->mdd_cl.mc_deniednext = 60; /* 60 secs by default */

	dev = lustre_cfg_string(lcfg, 0);
	if (dev == NULL)
		RETURN(rc);

	mdd->mdd_md_dev.md_lu_dev.ld_obd = class_name2obd(dev);
	if (mdd->mdd_md_dev.md_lu_dev.ld_obd == NULL)
		RETURN(rc);
	mdd->mdd_md_dev.md_lu_dev.ld_ops = &mdd_lu_ops;
	mdd->mdd_md_dev.md_ops = &mdd_ops;

	rc = mdd_connect_to_next(env, mdd, lustre_cfg_string(lcfg, 3));
	if (rc != 0)
		RETURN(rc);

	mdd->mdd_atime_diff = MAX_ATIME_DIFF;
	/* sync permission changes */
	mdd->mdd_sync_permission = 1;
	/* enable changelog garbage collection */
	mdd->mdd_changelog_gc = 1;
	/* with a significant amount of idle time */
	mdd->mdd_changelog_max_idle_time = CHLOG_MAX_IDLE_TIME;
	/* or a significant amount of late indexes */
	mdd->mdd_changelog_max_idle_indexes = CHLOG_MAX_IDLE_INDEXES;
	/* with a reasonable interval between each check */
	mdd->mdd_changelog_min_gc_interval = CHLOG_MIN_GC_INTERVAL;
	/* with a very few number of free catalog entries */
	mdd->mdd_changelog_min_free_cat_entries = CHLOG_MIN_FREE_CAT_ENTRIES;
	/* special default striping for files created with O_APPEND */
	mdd->mdd_append_stripe_count = 1;
	mdd->mdd_append_pool[0] = '\0';

	dt_conf_get(env, mdd->mdd_child, &mdd->mdd_dt_conf);

	/* we are using service name but not mdd obd name
	 * for compatibility reasons.
	 * It is passed from MDT in lustre_cfg[2] buffer */
	rc = mdd_procfs_init(mdd, lustre_cfg_string(lcfg, 2));
	if (rc < 0)
		obd_disconnect(mdd->mdd_child_exp);

        RETURN(rc);
}

static struct lu_device *mdd_device_fini(const struct lu_env *env,
                                         struct lu_device *d)
{
	struct mdd_device *mdd = lu2mdd_dev(d);

	if (d->ld_site)
		lu_dev_del_linkage(d->ld_site, d);

	mdd_procfs_fini(mdd);
	return NULL;
}

static int changelog_init_cb(const struct lu_env *env, struct llog_handle *llh,
			     struct llog_rec_hdr *hdr, void *data)
{
	struct mdd_device *mdd = (struct mdd_device *)data;
	struct llog_changelog_rec *rec = (struct llog_changelog_rec *)hdr;

	LASSERT(llh->lgh_hdr->llh_flags & LLOG_F_IS_PLAIN);
	LASSERT(rec->cr_hdr.lrh_type == CHANGELOG_REC);

	CDEBUG(D_INFO,
	       "seeing record at index %d/%d/%llu t=%x %.*s in log"
	       DFID"\n", hdr->lrh_index, rec->cr_hdr.lrh_index,
	       rec->cr.cr_index, rec->cr.cr_type, rec->cr.cr_namelen,
	       changelog_rec_name(&rec->cr), PFID(&llh->lgh_id.lgl_oi.oi_fid));

	mdd->mdd_cl.mc_index = rec->cr.cr_index;
	return LLOG_PROC_BREAK;
}

static int changelog_user_init_cb(const struct lu_env *env,
				  struct llog_handle *llh,
				  struct llog_rec_hdr *hdr, void *data)
{
	struct mdd_device *mdd = (struct mdd_device *)data;
	struct llog_changelog_user_rec *rec =
		(struct llog_changelog_user_rec *)hdr;

	LASSERT(llh->lgh_hdr->llh_flags & LLOG_F_IS_PLAIN);
	LASSERT(rec->cur_hdr.lrh_type == CHANGELOG_USER_REC);

	CDEBUG(D_INFO, "seeing user at index %d/%d id=%d endrec=%llu"
	       " in log "DFID"\n", hdr->lrh_index, rec->cur_hdr.lrh_index,
	       rec->cur_id, rec->cur_endrec, PFID(&llh->lgh_id.lgl_oi.oi_fid));

	spin_lock(&mdd->mdd_cl.mc_user_lock);
	mdd->mdd_cl.mc_lastuser = rec->cur_id;
	mdd->mdd_cl.mc_users++;
	if (rec->cur_endrec > mdd->mdd_cl.mc_index)
		mdd->mdd_cl.mc_index = rec->cur_endrec;
	spin_unlock(&mdd->mdd_cl.mc_user_lock);

	return LLOG_PROC_BREAK;
}

struct changelog_orphan_data {
	__u64 index;
	struct mdd_device *mdd;
};

/* find oldest changelog record index */
static int changelog_detect_orphan_cb(const struct lu_env *env,
				      struct llog_handle *llh,
				      struct llog_rec_hdr *hdr, void *data)
{
	struct mdd_device *mdd = ((struct changelog_orphan_data *)data)->mdd;
	struct llog_changelog_rec *rec = container_of(hdr,
						      struct llog_changelog_rec,
						      cr_hdr);

	LASSERT(llh->lgh_hdr->llh_flags & LLOG_F_IS_PLAIN);

	if (rec->cr_hdr.lrh_type != CHANGELOG_REC) {
		CWARN("%s: invalid record at index %d in log "DFID"\n",
		      mdd2obd_dev(mdd)->obd_name, hdr->lrh_index,
		      PFID(&llh->lgh_id.lgl_oi.oi_fid));
		/* try to find some next valid record and thus allow to recover
		 * from a corrupted LLOG, instead to assert and force a crash
		 */
		return 0;
	}

	CDEBUG(D_INFO, "%s: seeing record at index %d/%d/%llu t=%x %.*s in log "
	       DFID"\n", mdd2obd_dev(mdd)->obd_name, hdr->lrh_index,
	       rec->cr_hdr.lrh_index, rec->cr.cr_index, rec->cr.cr_type,
	       rec->cr.cr_namelen, changelog_rec_name(&rec->cr),
	       PFID(&llh->lgh_id.lgl_oi.oi_fid));

	((struct changelog_orphan_data *)data)->index = rec->cr.cr_index;
	return LLOG_PROC_BREAK;
}

/* find oldest changelog user index */
static int changelog_user_detect_orphan_cb(const struct lu_env *env,
					   struct llog_handle *llh,
					   struct llog_rec_hdr *hdr, void *data)
{
	struct mdd_device *mdd = ((struct changelog_orphan_data *)data)->mdd;
	struct llog_changelog_user_rec *rec = container_of(hdr,
						struct llog_changelog_user_rec,
						cur_hdr);

	LASSERT(llh->lgh_hdr->llh_flags & LLOG_F_IS_PLAIN);

	if (rec->cur_hdr.lrh_type != CHANGELOG_USER_REC) {
		CWARN("%s: invalid user at index %d in log "DFID"\n",
		      mdd2obd_dev(mdd)->obd_name, hdr->lrh_index,
		      PFID(&llh->lgh_id.lgl_oi.oi_fid));
		/* try to find some next valid record and thus allow to recover
		 * from a corrupted LLOG, instead to assert and force a crash
		 */
		return 0;
	}

	CDEBUG(D_INFO, "%s: seeing user at index %d/%d id=%d endrec=%llu in "
	       "log "DFID"\n", mdd2obd_dev(mdd)->obd_name, hdr->lrh_index,
	       rec->cur_hdr.lrh_index, rec->cur_id, rec->cur_endrec,
	       PFID(&llh->lgh_id.lgl_oi.oi_fid));

	if (((struct changelog_orphan_data *)data)->index == 0 ||
	    rec->cur_endrec < ((struct changelog_orphan_data *)data)->index)
		((struct changelog_orphan_data *)data)->index = rec->cur_endrec;

	return 0;
}

struct changelog_cancel_cookie {
	long long endrec;
	struct mdd_device *mdd;
};

static int llog_changelog_cancel_cb(const struct lu_env *env,
				    struct llog_handle *llh,
				    struct llog_rec_hdr *hdr, void *data)
{
	struct llog_changelog_rec *rec = (struct llog_changelog_rec *)hdr;
	struct llog_cookie	 cookie;
	struct changelog_cancel_cookie *cl_cookie =
		(struct changelog_cancel_cookie *)data;
	int			 rc;

	ENTRY;

	/* This is always a (sub)log, not the catalog */
	LASSERT(llh->lgh_hdr->llh_flags & LLOG_F_IS_PLAIN);

	/* if current context is GC-thread allow it to stop upon umount
	 * remaining records cleanup will occur upon next mount
	 *
	 * also during testing, wait for GC-thread to be released
	 *
	 * XXX this requires the GC-thread to not fork a sub-thread via
	 * llog[_cat]_process_or_fork() and we may think to also implement
	 * this shutdown mechanism for manually started user unregister which
	 * can also take a long time if huge backlog of records
	 */
	if (unlikely(cl_cookie->mdd->mdd_cl.mc_gc_task == current)) {
		/* wait to be released */
		while (CFS_FAIL_CHECK_QUIET(OBD_FAIL_FORCE_GC_THREAD))
			schedule();

		if (kthread_should_stop())
			RETURN(LLOG_PROC_BREAK);
	}

	if (rec->cr.cr_index > cl_cookie->endrec)
		/* records are in order, so we're done */
		RETURN(LLOG_PROC_BREAK);

	cookie.lgc_lgl = llh->lgh_id;
	cookie.lgc_index = hdr->lrh_index;

	/* cancel them one at a time.  I suppose we could store up the cookies
	 * and cancel them all at once; probably more efficient, but this is
	 * done as a user call, so who cares... */
	rc = llog_cat_cancel_records(env, llh->u.phd.phd_cat_handle, 1,
				     &cookie);
	RETURN(rc < 0 ? rc : 0);
}

static int llog_changelog_cancel(const struct lu_env *env,
				 struct llog_ctxt *ctxt,
				 struct changelog_cancel_cookie *cookie)
{
	struct llog_handle	*cathandle = ctxt->loc_handle;
	int			 rc;

	ENTRY;

	/* This should only be called with the catalog handle */
	LASSERT(cathandle->lgh_hdr->llh_flags & LLOG_F_IS_CAT);

	rc = llog_cat_process(env, cathandle, llog_changelog_cancel_cb,
			      cookie, 0, 0);
	if (rc >= 0)
		/* 0 or 1 means we're done */
		rc = 0;
	else
		CERROR("%s: cancel idx %u of catalog "DFID": rc = %d\n",
		       ctxt->loc_obd->obd_name, cathandle->lgh_last_idx,
		       PFID(&cathandle->lgh_id.lgl_oi.oi_fid), rc);

	RETURN(rc);
}

static struct llog_operations changelog_orig_logops;

static int
mdd_changelog_write_header(const struct lu_env *env, struct mdd_device *mdd,
			   int markerflags);

static int
mdd_changelog_on(const struct lu_env *env, struct mdd_device *mdd)
{
	int rc = 0;

	if ((mdd->mdd_cl.mc_flags & CLM_ON) != 0)
		return rc;

	LCONSOLE_INFO("%s: changelog on\n", mdd2obd_dev(mdd)->obd_name);
	if (mdd->mdd_cl.mc_flags & CLM_ERR) {
		CERROR("Changelogs cannot be enabled due to error "
		       "condition (see %s log).\n",
		       mdd2obd_dev(mdd)->obd_name);
		rc = -ESRCH;
	} else {
		spin_lock(&mdd->mdd_cl.mc_lock);
		mdd->mdd_cl.mc_flags |= CLM_ON;
		spin_unlock(&mdd->mdd_cl.mc_lock);
		rc = mdd_changelog_write_header(env, mdd, CLM_START);
	}
	return rc;
}

static int
mdd_changelog_off(const struct lu_env *env, struct mdd_device *mdd)
{
	int rc = 0;

	if ((mdd->mdd_cl.mc_flags & CLM_ON) != CLM_ON)
		return rc;

	LCONSOLE_INFO("%s: changelog off\n", mdd2obd_dev(mdd)->obd_name);
	rc = mdd_changelog_write_header(env, mdd, CLM_FINI);
	spin_lock(&mdd->mdd_cl.mc_lock);
	mdd->mdd_cl.mc_flags &= ~CLM_ON;
	spin_unlock(&mdd->mdd_cl.mc_lock);

	return rc;
}

static int mdd_changelog_llog_init(const struct lu_env *env,
				   struct mdd_device *mdd)
{
	struct obd_device	*obd = mdd2obd_dev(mdd);
	struct llog_ctxt	*ctxt = NULL, *uctxt = NULL;
	struct changelog_orphan_data changelog_orphan = { .index = 0,
							  .mdd = mdd },
				     user_orphan = { .index = 0,
						     .mdd = mdd };
	int			 rc;

	ENTRY;

	/* LU-2844 mdd setup failure should not cause umount oops */
	if (OBD_FAIL_CHECK(OBD_FAIL_MDS_CHANGELOG_INIT))
		RETURN(-EIO);

	OBD_SET_CTXT_MAGIC(&obd->obd_lvfs_ctxt);
	obd->obd_lvfs_ctxt.dt = mdd->mdd_bottom;
	rc = llog_setup(env, obd, &obd->obd_olg, LLOG_CHANGELOG_ORIG_CTXT,
			obd, &changelog_orig_logops);
	if (rc) {
		CERROR("%s: changelog llog setup failed: rc = %d\n",
		       obd->obd_name, rc);
		RETURN(rc);
	}

	ctxt = llog_get_context(obd, LLOG_CHANGELOG_ORIG_CTXT);
	LASSERT(ctxt);

	rc = llog_open_create(env, ctxt, &ctxt->loc_handle, NULL,
			      CHANGELOG_CATALOG);
	if (rc)
		GOTO(out_cleanup, rc);

	rc = llog_init_handle(env, ctxt->loc_handle, LLOG_F_IS_CAT, NULL);
	if (rc)
		GOTO(out_close, rc);

	rc = llog_cat_reverse_process(env, ctxt->loc_handle,
				      changelog_init_cb, mdd);

	if (rc < 0) {
		CERROR("%s: changelog init failed: rc = %d\n", obd->obd_name,
		       rc);
		GOTO(out_close, rc);
	}

	CDEBUG(D_IOCTL, "changelog starting index=%llu\n",
	       mdd->mdd_cl.mc_index);

	/* setup user changelog */
	rc = llog_setup(env, obd, &obd->obd_olg, LLOG_CHANGELOG_USER_ORIG_CTXT,
			obd, &changelog_orig_logops);
	if (rc) {
		CERROR("%s: changelog users llog setup failed: rc = %d\n",
		       obd->obd_name, rc);
		GOTO(out_close, rc);
	}

	uctxt = llog_get_context(obd, LLOG_CHANGELOG_USER_ORIG_CTXT);
	LASSERT(ctxt);

	rc = llog_open_create(env, uctxt, &uctxt->loc_handle, NULL,
			      CHANGELOG_USERS);
	if (rc)
		GOTO(out_ucleanup, rc);

	rc = llog_init_handle(env, uctxt->loc_handle, LLOG_F_IS_CAT, NULL);
	if (rc)
		GOTO(out_uclose, rc);

	rc = llog_cat_reverse_process(env, uctxt->loc_handle,
				      changelog_user_init_cb, mdd);
	if (rc < 0) {
		CERROR("%s: changelog user init failed: rc = %d\n",
		       obd->obd_name, rc);
		GOTO(out_uclose, rc);
	}

	/* If we have registered users, assume we want changelogs on */
	if (mdd->mdd_cl.mc_lastuser > 0) {
		rc = mdd_changelog_on(env, mdd);
		if (rc < 0)
			GOTO(out_uclose, rc);
	}

	/* find and clear any orphan changelog records (1st record index <
	 * smallest of all users current index), likely to come from an
	 * interrupted manual or GC-thread purge, as its user record had
	 * been deleted first
	 * XXX we may wait for a still registered user clear operation to
	 * do the job, but it may then take a long time to reach the user's
	 * real targetted records if a huge purge backlog is still to be
	 * processed as a long time idle user record could have been deleted
	 * XXX we may need to run end of purge as a separate thread
	 */
	rc = llog_cat_process(env, ctxt->loc_handle, changelog_detect_orphan_cb,
			      &changelog_orphan, 0, 0);
	if (rc < 0) {
		CERROR("%s: changelog detect orphan failed: rc = %d\n",
		       obd->obd_name, rc);
		GOTO(out_uclose, rc);
	}
	rc = llog_cat_process(env, uctxt->loc_handle,
			      changelog_user_detect_orphan_cb,
			      &user_orphan, 0, 0);
	if (rc < 0) {
		CERROR("%s: changelog user detect orphan failed: rc = %d\n",
		       obd->obd_name, rc);
		GOTO(out_uclose, rc);
	}
	if (unlikely(changelog_orphan.index < user_orphan.index)) {
		struct changelog_cancel_cookie cl_cookie = {
			.endrec = user_orphan.index,
			.mdd = mdd,
		};

		CWARN("%s : orphan changelog records found, starting from "
		      "index %llu to index %llu, being cleared now\n",
		      obd->obd_name, changelog_orphan.index, user_orphan.index);

		/* XXX we may need to run end of purge as a separate thread */
		rc = llog_changelog_cancel(env, ctxt, &cl_cookie);
		if (rc < 0) {
			CERROR("%s: purge of changelog orphan records failed: "
			       "rc = %d\n", obd->obd_name, rc);
			GOTO(out_uclose, rc);
		}
	}

	llog_ctxt_put(ctxt);
	llog_ctxt_put(uctxt);
	RETURN(0);
out_uclose:
	llog_cat_close(env, uctxt->loc_handle);
out_ucleanup:
	llog_cleanup(env, uctxt);
out_close:
	llog_cat_close(env, ctxt->loc_handle);
out_cleanup:
	llog_cleanup(env, ctxt);
	return rc;
}

static int mdd_changelog_init(const struct lu_env *env, struct mdd_device *mdd)
{
	struct obd_device	*obd = mdd2obd_dev(mdd);
	int			 rc;

	mdd->mdd_cl.mc_index = 0;
	spin_lock_init(&mdd->mdd_cl.mc_lock);
	mdd->mdd_cl.mc_starttime = ktime_get();
	spin_lock_init(&mdd->mdd_cl.mc_user_lock);
	mdd->mdd_cl.mc_lastuser = 0;

	/* ensure a GC check will, and a thread run may, occur upon start */
	mdd->mdd_cl.mc_gc_time = 0;
	mdd->mdd_cl.mc_gc_task = MDD_CHLG_GC_NONE;

	rc = mdd_changelog_llog_init(env, mdd);
	if (rc) {
		CERROR("%s: changelog setup during init failed: rc = %d\n",
		       obd->obd_name, rc);
		mdd->mdd_cl.mc_flags |= CLM_ERR;
	}

	return rc;
}

static void mdd_changelog_fini(const struct lu_env *env,
			       struct mdd_device *mdd)
{
	struct obd_device	*obd = mdd2obd_dev(mdd);
	struct llog_ctxt	*ctxt;

	mdd->mdd_cl.mc_flags = 0;

again:
	/* stop GC-thread if running */
	spin_lock(&mdd->mdd_cl.mc_lock);
	if (likely(mdd->mdd_cl.mc_gc_task == MDD_CHLG_GC_NONE)) {
		/* avoid any attempt to run a GC-thread */
		mdd->mdd_cl.mc_gc_task = current;
		spin_unlock(&mdd->mdd_cl.mc_lock);
	} else {
		struct task_struct *gc_task;

		if (unlikely(mdd->mdd_cl.mc_gc_task == MDD_CHLG_GC_NEED ||
			     mdd->mdd_cl.mc_gc_task == MDD_CHLG_GC_START)) {
			/* need to wait for birthing GC-thread to be started
			 * and to have set mc_gc_task to itself
			 */
			spin_unlock(&mdd->mdd_cl.mc_lock);
			schedule_timeout(usecs_to_jiffies(10));
			/* go back to fully check if GC-thread has started or
			 * even already exited or if a new one is starting...
			 */
			goto again;
		}
		/* take a reference on task_struct to avoid it to be freed
		 * upon exit
		 */
		gc_task = mdd->mdd_cl.mc_gc_task;
		get_task_struct(gc_task);
		spin_unlock(&mdd->mdd_cl.mc_lock);
		kthread_stop(gc_task);
		put_task_struct(gc_task);
	}

	ctxt = llog_get_context(obd, LLOG_CHANGELOG_ORIG_CTXT);
	if (ctxt) {
		llog_cat_close(env, ctxt->loc_handle);
		llog_cleanup(env, ctxt);
	}
	ctxt = llog_get_context(obd, LLOG_CHANGELOG_USER_ORIG_CTXT);
	if (ctxt) {
		llog_cat_close(env, ctxt->loc_handle);
		llog_cleanup(env, ctxt);
	}
}

/** Remove entries with indicies up to and including \a endrec from the
 *  changelog
 * \param mdd
 * \param endrec
 * \retval 0 ok
 */
static int
mdd_changelog_llog_cancel(const struct lu_env *env, struct mdd_device *mdd,
			  long long endrec)
{
        struct obd_device *obd = mdd2obd_dev(mdd);
        struct llog_ctxt *ctxt;
        long long unsigned cur;
	struct changelog_cancel_cookie cookie;
        int rc;

        ctxt = llog_get_context(obd, LLOG_CHANGELOG_ORIG_CTXT);
        if (ctxt == NULL)
                return -ENXIO;

	spin_lock(&mdd->mdd_cl.mc_lock);
	cur = (long long)mdd->mdd_cl.mc_index;
	spin_unlock(&mdd->mdd_cl.mc_lock);
        if (endrec > cur)
                endrec = cur;

        /* purge to "0" is shorthand for everything */
        if (endrec == 0)
                endrec = cur;

        /* If purging all records, write a header entry so we don't have an
           empty catalog and we're sure to have a valid starting index next
           time.  In case of crash, we just restart with old log so we're
           allright. */
        if (endrec == cur) {
                /* XXX: transaction is started by llog itself */
                rc = mdd_changelog_write_header(env, mdd, CLM_PURGE);
                if (rc)
                      goto out;
        }

        /* Some records were purged, so reset repeat-access time (so we
           record new mtime update records, so users can see a file has been
           changed since the last purge) */
	mdd->mdd_cl.mc_starttime = ktime_get();

	cookie.endrec = endrec;
	cookie.mdd = mdd;
	rc = llog_changelog_cancel(env, ctxt, &cookie);
out:
        llog_ctxt_put(ctxt);
        return rc;
}

/** Add a CL_MARK record to the changelog
 * \param mdd
 * \param markerflags - CLM_*
 * \retval 0 ok
 */
int mdd_changelog_write_header(const struct lu_env *env,
			       struct mdd_device *mdd, int markerflags)
{
	struct obd_device		*obd = mdd2obd_dev(mdd);
	struct llog_changelog_rec	*rec;
	struct lu_buf			*buf;
	struct llog_ctxt		*ctxt;
	int				 reclen;
	int				 len = strlen(obd->obd_name);
	int				 rc;

	ENTRY;

	if (mdd->mdd_cl.mc_mask & (1 << CL_MARK)) {
		mdd->mdd_cl.mc_starttime = ktime_get();
		RETURN(0);
	}

	reclen = llog_data_len(sizeof(*rec) + len);
	buf = lu_buf_check_and_alloc(&mdd_env_info(env)->mti_big_buf, reclen);
	if (buf->lb_buf == NULL)
		RETURN(-ENOMEM);
	rec = buf->lb_buf;

        rec->cr.cr_flags = CLF_VERSION;
        rec->cr.cr_type = CL_MARK;
        rec->cr.cr_namelen = len;
	memcpy(changelog_rec_name(&rec->cr), obd->obd_name, rec->cr.cr_namelen);
        /* Status and action flags */
	rec->cr.cr_markerflags = mdd->mdd_cl.mc_flags | markerflags;
	rec->cr_hdr.lrh_len = llog_data_len(changelog_rec_size(&rec->cr) +
					    rec->cr.cr_namelen);
	rec->cr_hdr.lrh_type = CHANGELOG_REC;
	rec->cr.cr_time = cl_time();

	ctxt = llog_get_context(obd, LLOG_CHANGELOG_ORIG_CTXT);
	LASSERT(ctxt);

	rc = llog_cat_add(env, ctxt->loc_handle, &rec->cr_hdr, NULL);
	if (rc > 0)
		rc = 0;
	llog_ctxt_put(ctxt);

	/* assume on or off event; reset repeat-access time */
	mdd->mdd_cl.mc_starttime = ktime_get();
	RETURN(rc);
}

/**
 * Lookup method for "fid" object. Only filenames with correct SEQ:OID format
 * are valid. We also check if object with passed fid exists or not.
 */
static int obf_lookup(const struct lu_env *env, struct md_object *p,
                      const struct lu_name *lname, struct lu_fid *f,
                      struct md_op_spec *spec)
{
        char *name = (char *)lname->ln_name;
        struct mdd_device *mdd = mdo2mdd(p);
        struct mdd_object *child;
        int rc = 0;

        while (*name == '[')
                name++;

        sscanf(name, SFID, RFID(f));
        if (!fid_is_sane(f)) {
		CWARN("%s: Trying to lookup invalid FID [%s] in %s/%s, FID "
		      "format should be "DFID"\n", mdd2obd_dev(mdd)->obd_name,
		      lname->ln_name, dot_lustre_name, mdd_obf_dir_name,
		      (__u64)FID_SEQ_NORMAL, 1, 0);
                GOTO(out, rc = -EINVAL);
        }

	if (!fid_is_norm(f) && !fid_is_igif(f) && !fid_is_root(f) &&
	    !fid_seq_is_dot(f->f_seq)) {
		CWARN("%s: Trying to lookup invalid FID "DFID" in %s/%s, "
		      "sequence should be >= %#llx or within [%#llx,"
		      "%#llx].\n", mdd2obd_dev(mdd)->obd_name, PFID(f),
		      dot_lustre_name, mdd_obf_dir_name, (__u64)FID_SEQ_NORMAL,
		      (__u64)FID_SEQ_IGIF, (__u64)FID_SEQ_IGIF_MAX);
		GOTO(out, rc = -EINVAL);
	}

        /* Check if object with this fid exists */
        child = mdd_object_find(env, mdd, f);
        if (IS_ERR(child))
                GOTO(out, rc = PTR_ERR(child));

        if (mdd_object_exists(child) == 0)
                rc = -ENOENT;

        mdd_object_put(env, child);

out:
        return rc;
}

static int mdd_dummy_create(const struct lu_env *env,
			    struct md_object *pobj,
			    const struct lu_name *lname,
			    struct md_object *child,
			    struct md_op_spec *spec,
			    struct md_attr* ma)
{
	return -EPERM;
}

static int mdd_dummy_rename(const struct lu_env *env,
			    struct md_object *src_pobj,
			    struct md_object *tgt_pobj,
			    const struct lu_fid *lf,
			    const struct lu_name *lsname,
			    struct md_object *tobj,
			    const struct lu_name *ltname,
			    struct md_attr *ma)
{
	return -EPERM;
}

static int mdd_dummy_link(const struct lu_env *env,
			  struct md_object *tgt_obj,
			  struct md_object *src_obj,
			  const struct lu_name *lname,
			  struct md_attr *ma)
{
	return -EPERM;
}

static int mdd_dummy_unlink(const struct lu_env *env,
			    struct md_object *pobj,
			    struct md_object *cobj,
			    const struct lu_name *lname,
			    struct md_attr *ma,
			    int no_name)
{
	return -EPERM;
}

int mdd_create(const struct lu_env *env, struct md_object *pobj,
		      const struct lu_name *lname, struct md_object *child,
		      struct md_op_spec *spec, struct md_attr *ma);
static int mdd_obf_create(const struct lu_env *env, struct md_object *pobj,
		      const struct lu_name *lname, struct md_object *child,
		      struct md_op_spec *spec, struct md_attr *ma)
{
	if (spec->sp_cr_flags & MDS_OPEN_VOLATILE)
		return mdd_create(env, pobj, lname, child, spec, ma);
	RETURN(-EPERM);
}

static struct md_dir_operations mdd_obf_dir_ops = {
	.mdo_lookup = obf_lookup,
	.mdo_create = mdd_obf_create,
	.mdo_rename = mdd_dummy_rename,
	.mdo_link   = mdd_dummy_link,
	.mdo_unlink = mdd_dummy_unlink
};

static struct md_dir_operations mdd_lpf_dir_ops = {
	.mdo_lookup = mdd_lookup,
	.mdo_create = mdd_dummy_create,
	.mdo_rename = mdd_dummy_rename,
	.mdo_link   = mdd_dummy_link,
	.mdo_unlink = mdd_dummy_unlink
};

static struct md_object *mdo_locate(const struct lu_env *env,
				    struct md_device *md,
				    const struct lu_fid *fid)
{
	struct lu_object *obj;
	struct md_object *mdo;

	obj = lu_object_find(env, &md->md_lu_dev, fid, NULL);
	if (!IS_ERR(obj)) {
		obj = lu_object_locate(obj->lo_header, md->md_lu_dev.ld_type);
		LASSERT(obj != NULL);
		mdo = lu2md(obj);
	} else {
		mdo = ERR_PTR(PTR_ERR(obj));
	}
	return mdo;
}

static int mdd_lpf_setup(const struct lu_env *env, struct mdd_device *m)
{
	struct md_object	*mdo;
	struct mdd_object	*mdd_lpf;
	struct lu_fid		 fid	= LU_LPF_FID;
	int			 rc;
	ENTRY;

	rc = mdd_local_file_create(env, m, mdd_object_fid(m->mdd_dot_lustre),
				   mdd_lpf_dir_name, S_IFDIR | S_IRUSR | S_IXUSR,
				   &fid);
	if (rc != 0)
		RETURN(rc);

	mdo = mdo_locate(env, &m->mdd_md_dev, &fid);
	if (IS_ERR(mdo))
		RETURN(PTR_ERR(mdo));

	LASSERT(lu_object_exists(&mdo->mo_lu));

	mdd_lpf = md2mdd_obj(mdo);
	mdd_lpf->mod_obj.mo_dir_ops = &mdd_lpf_dir_ops;
	m->mdd_dot_lustre_objs.mdd_lpf = mdd_lpf;

	RETURN(0);
}

/**
 * Create special in-memory "fid" object for open-by-fid.
 */
static int mdd_obf_setup(const struct lu_env *env, struct mdd_device *m)
{
	struct md_object	*mdo;
	struct mdd_object	*mdd_obf;
	struct lu_fid		 fid = LU_OBF_FID;
	int			 rc;

	rc = mdd_local_file_create(env, m, mdd_object_fid(m->mdd_dot_lustre),
				   mdd_obf_dir_name, S_IFDIR | S_IXUSR, &fid);
	if (rc < 0)
		RETURN(rc);

	mdo = mdo_locate(env, &m->mdd_md_dev, &fid);
	if (IS_ERR(mdo))
		RETURN(PTR_ERR(mdo));

	LASSERT(lu_object_exists(&mdo->mo_lu));

	mdd_obf = md2mdd_obj(mdo);
	mdd_obf->mod_obj.mo_dir_ops = &mdd_obf_dir_ops;
	m->mdd_dot_lustre_objs.mdd_obf = mdd_obf;

	return 0;
}

static void mdd_dot_lustre_cleanup(const struct lu_env *env,
				   struct mdd_device *m)
{
	if (m->mdd_dot_lustre_objs.mdd_lpf != NULL) {
		mdd_object_put(env, m->mdd_dot_lustre_objs.mdd_lpf);
		m->mdd_dot_lustre_objs.mdd_lpf = NULL;
	}
	if (m->mdd_dot_lustre_objs.mdd_obf != NULL) {
		mdd_object_put(env, m->mdd_dot_lustre_objs.mdd_obf);
		m->mdd_dot_lustre_objs.mdd_obf = NULL;
	}
	if (m->mdd_dot_lustre != NULL) {
		mdd_object_put(env, m->mdd_dot_lustre);
		m->mdd_dot_lustre = NULL;
	}
}

/** Setup ".lustre" directory object */
static int mdd_dot_lustre_setup(const struct lu_env *env, struct mdd_device *m)
{
	struct md_object	*mdo;
	struct lu_fid		 fid;
	int			 rc;

	ENTRY;
	/* Create ".lustre" directory in ROOT. */
	fid = LU_DOT_LUSTRE_FID;
	rc = mdd_local_file_create(env, m, &m->mdd_root_fid,
				   dot_lustre_name,
				   S_IFDIR | S_IRUGO | S_IWUSR | S_IXUGO,
				   &fid);
	if (rc < 0)
		RETURN(rc);
	mdo = mdo_locate(env, &m->mdd_md_dev, &fid);
	if (IS_ERR(mdo))
		RETURN(PTR_ERR(mdo));
	LASSERT(lu_object_exists(&mdo->mo_lu));

	m->mdd_dot_lustre = md2mdd_obj(mdo);

	rc = mdd_obf_setup(env, m);
	if (rc) {
		CERROR("%s: error initializing \"fid\" object: rc = %d.\n",
		       mdd2obd_dev(m)->obd_name, rc);
		GOTO(out, rc);
	}

	rc = mdd_lpf_setup(env, m);
	if (rc != 0) {
		CERROR("%s: error initializing \"lost+found\": rc = %d.\n",
		       mdd2obd_dev(m)->obd_name, rc);
		GOTO(out, rc);
	}

	RETURN(0);

out:
	mdd_dot_lustre_cleanup(env, m);

	return rc;
}

/**
 * set llog methods and create LLOG_AGENT_ORIG_CTXT llog
 * object in obd_device
 */
static int mdd_hsm_actions_llog_init(const struct lu_env *env,
				     struct mdd_device *m)
{
	struct obd_device	*obd = mdd2obd_dev(m);
	struct llog_ctxt	*ctxt = NULL;
	int			 rc;
	ENTRY;

	OBD_SET_CTXT_MAGIC(&obd->obd_lvfs_ctxt);
	obd->obd_lvfs_ctxt.dt = m->mdd_bottom;

	rc = llog_setup(env, obd, &obd->obd_olg, LLOG_AGENT_ORIG_CTXT,
			obd, &llog_common_cat_ops);
	if (rc) {
		CERROR("%s: hsm actions llog setup failed: rc = %d\n",
			obd->obd_name, rc);
		RETURN(rc);
	}

	ctxt = llog_get_context(obd, LLOG_AGENT_ORIG_CTXT);
	LASSERT(ctxt);

	rc = llog_open_create(env, ctxt, &ctxt->loc_handle, NULL,
			      HSM_ACTIONS);
	if (rc) {
		CERROR("%s: hsm actions llog open_create failed: rc = %d\n",
			obd->obd_name, rc);
		GOTO(out_cleanup, rc);
	}

	rc = llog_init_handle(env, ctxt->loc_handle, LLOG_F_IS_CAT, NULL);
	if (rc)
		GOTO(out_close, rc);

	llog_ctxt_put(ctxt);
	RETURN(0);

out_close:
	llog_cat_close(env, ctxt->loc_handle);
	ctxt->loc_handle = NULL;
out_cleanup:
	llog_cleanup(env, ctxt);

	return rc;
}

/**
 * cleanup the context created by llog_setup_named()
 */
static int mdd_hsm_actions_llog_fini(const struct lu_env *env,
				     struct mdd_device *m)
{
	struct obd_device	*obd = mdd2obd_dev(m);
	struct llog_ctxt	*lctxt;
	ENTRY;

	lctxt = llog_get_context(obd, LLOG_AGENT_ORIG_CTXT);
	if (lctxt) {
		llog_cat_close(env, lctxt->loc_handle);
		lctxt->loc_handle = NULL;
		llog_cleanup(env, lctxt);
	}

	RETURN(0);
}

static void mdd_device_shutdown(const struct lu_env *env, struct mdd_device *m,
				struct lustre_cfg *cfg)
{
	barrier_deregister(m->mdd_bottom);
	lfsck_degister(env, m->mdd_bottom);
	mdd_hsm_actions_llog_fini(env, m);
	mdd_changelog_fini(env, m);
	mdd_orphan_index_fini(env, m);
	mdd_dot_lustre_cleanup(env, m);
	if (mdd2obd_dev(m)->u.obt.obt_nodemap_config_file) {
		nm_config_file_deregister_tgt(env,
				mdd2obd_dev(m)->u.obt.obt_nodemap_config_file);
		mdd2obd_dev(m)->u.obt.obt_nodemap_config_file = NULL;
	}
	if (m->mdd_los != NULL) {
		local_oid_storage_fini(env, m->mdd_los);
		m->mdd_los = NULL;
	}
	lu_site_purge(env, mdd2lu_dev(m)->ld_site, ~0);

	if (m->mdd_child_exp)
		obd_disconnect(m->mdd_child_exp);
}

static int mdd_process_config(const struct lu_env *env,
                              struct lu_device *d, struct lustre_cfg *cfg)
{
        struct mdd_device *m    = lu2mdd_dev(d);
        struct dt_device  *dt   = m->mdd_child;
        struct lu_device  *next = &dt->dd_lu_dev;
        int rc;
        ENTRY;

        switch (cfg->lcfg_command) {
	case LCFG_PARAM: {
		ssize_t count;

		count = class_modify_config(cfg, PARAM_MDD, &m->mdd_kobj);
		rc = count > 0 ? 0 : count;
		if (rc)
			/* we don't understand; pass it on */
			rc = next->ld_ops->ldo_process_config(env, next, cfg);
		break;
	}
        case LCFG_SETUP:
                rc = next->ld_ops->ldo_process_config(env, next, cfg);
                if (rc)
                        GOTO(out, rc);
		dt_conf_get(env, dt, &m->mdd_dt_conf);
                break;
	case LCFG_PRE_CLEANUP:
		rc = next->ld_ops->ldo_process_config(env, next, cfg);
		mdd_generic_thread_stop(&m->mdd_orphan_cleanup_thread);
		break;
	case LCFG_CLEANUP:
		rc = next->ld_ops->ldo_process_config(env, next, cfg);
		lu_dev_del_linkage(d->ld_site, d);
		mdd_device_shutdown(env, m, cfg);
		break;
        default:
                rc = next->ld_ops->ldo_process_config(env, next, cfg);
                break;
        }
out:
        RETURN(rc);
}

static int mdd_recovery_complete(const struct lu_env *env,
				 struct lu_device *d)
{
	struct mdd_device *mdd = lu2mdd_dev(d);
	struct lu_device *next;
	int rc;
	ENTRY;

	LASSERT(mdd != NULL);
	next = &mdd->mdd_child->dd_lu_dev;

	if (!mdd->mdd_bottom->dd_rdonly)
		mdd_orphan_cleanup(env, mdd);
	rc = next->ld_ops->ldo_recovery_complete(env, next);

	RETURN(rc);
}

int mdd_local_file_create(const struct lu_env *env, struct mdd_device *mdd,
			  const struct lu_fid *pfid, const char *name,
			  __u32 mode, struct lu_fid *fid)
{
	struct dt_object *parent, *dto;
	int rc = 0;

	ENTRY;

	LASSERT(!fid_is_zero(pfid));
	parent = dt_locate(env, mdd->mdd_bottom, pfid);
	if (unlikely(IS_ERR(parent)))
		RETURN(PTR_ERR(parent));

	/* create local file/dir, if @fid is passed then try to use it */
	if (fid_is_zero(fid))
		dto = local_file_find_or_create(env, mdd->mdd_los, parent,
						name, mode);
	else
		dto = local_file_find_or_create_with_fid(env, mdd->mdd_bottom,
							 fid, parent, name,
							 mode);
	if (IS_ERR(dto))
		GOTO(out_put, rc = PTR_ERR(dto));
	*fid = *lu_object_fid(&dto->do_lu);
	/* since stack is not fully set up the local_storage uses own stack
	 * and we should drop its object from cache */
	dt_object_put_nocache(env, dto);
	EXIT;
out_put:
	dt_object_put(env, parent);
	return rc;
}

static int mdd_lfsck_out_notify(const struct lu_env *env, void *data,
				enum lfsck_events event)
{
	return 0;
}

static int mdd_prepare(const struct lu_env *env,
                       struct lu_device *pdev,
                       struct lu_device *cdev)
{
	struct mdd_device *mdd = lu2mdd_dev(cdev);
	struct lu_device *next = &mdd->mdd_child->dd_lu_dev;
	struct nm_config_file *nodemap_config;
	struct obd_device_target *obt = &mdd2obd_dev(mdd)->u.obt;
	struct lu_fid fid;
	int rc;

	ENTRY;

	rc = next->ld_ops->ldo_prepare(env, cdev, next);
	if (rc)
		RETURN(rc);

	/* Setup local dirs */
	fid.f_seq = FID_SEQ_LOCAL_NAME;
	fid.f_oid = 1;
	fid.f_ver = 0;
	rc = local_oid_storage_init(env, mdd->mdd_bottom, &fid,
				    &mdd->mdd_los);
	if (rc)
		RETURN(rc);

	rc = dt_root_get(env, mdd->mdd_child, &mdd->mdd_local_root_fid);
	if (rc < 0)
		GOTO(out_los, rc);

	lu_root_fid(&fid);
	if (mdd_seq_site(mdd)->ss_node_id == 0) {
		rc = mdd_local_file_create(env, mdd, &mdd->mdd_local_root_fid,
					   mdd_root_dir_name, S_IFDIR |
					   S_IRUGO | S_IWUSR | S_IXUGO, &fid);
		if (rc != 0) {
			CERROR("%s: create root fid failed: rc = %d\n",
			       mdd2obd_dev(mdd)->obd_name, rc);
			GOTO(out_los, rc);
		}

		mdd->mdd_root_fid = fid;
		rc = mdd_dot_lustre_setup(env, mdd);
		if (rc != 0) {
			CERROR("%s: initializing .lustre failed: rc = %d\n",
			       mdd2obd_dev(mdd)->obd_name, rc);
			GOTO(out_los, rc);
		}
	} else {
		/* Normal client usually send root access to MDT0 directly,
		 * the root FID on non-MDT0 will only be used by echo client. */
		mdd->mdd_root_fid = fid;
	}

	rc = mdd_orphan_index_init(env, mdd);
	if (rc < 0)
		GOTO(out_dot, rc);

	rc = mdd_changelog_init(env, mdd);
	if (rc != 0) {
		CERROR("%s: failed to initialize changelog: rc = %d\n",
		       mdd2obd_dev(mdd)->obd_name, rc);
		GOTO(out_orph, rc);
	}

	rc = mdd_hsm_actions_llog_init(env, mdd);
	if (rc != 0)
		GOTO(out_changelog, rc);

	nodemap_config = nm_config_file_register_tgt(env, mdd->mdd_bottom,
						     mdd->mdd_los);
	if (IS_ERR(nodemap_config)) {
		rc = PTR_ERR(nodemap_config);
		if (rc != -EROFS)
			GOTO(out_hsm, rc);
	} else {
		obt->obt_nodemap_config_file = nodemap_config;
	}

	rc = lfsck_register(env, mdd->mdd_bottom, mdd->mdd_child,
			    mdd2obd_dev(mdd), mdd_lfsck_out_notify,
			    mdd, true);
	if (rc != 0) {
		CERROR("%s: failed to initialize lfsck: rc = %d\n",
		       mdd2obd_dev(mdd)->obd_name, rc);
		GOTO(out_nodemap, rc);
	}

	rc = barrier_register(mdd->mdd_bottom, mdd->mdd_child);
	if (rc) {
		CERROR("%s: failed to register to barrier: rc = %d\n",
		       mdd2obd_dev(mdd)->obd_name, rc);
		GOTO(out_lfsck, rc);
	}

	RETURN(0);

out_lfsck:
	lfsck_degister(env, mdd->mdd_bottom);
out_nodemap:
	nm_config_file_deregister_tgt(env, obt->obt_nodemap_config_file);
	obt->obt_nodemap_config_file = NULL;
out_hsm:
	mdd_hsm_actions_llog_fini(env, mdd);
out_changelog:
	mdd_changelog_fini(env, mdd);
out_orph:
	mdd_orphan_index_fini(env, mdd);
out_dot:
	if (mdd_seq_site(mdd)->ss_node_id == 0)
		mdd_dot_lustre_cleanup(env, mdd);
out_los:
	local_oid_storage_fini(env, mdd->mdd_los);
	mdd->mdd_los = NULL;

	return rc;
}

const struct lu_device_operations mdd_lu_ops = {
        .ldo_object_alloc      = mdd_object_alloc,
        .ldo_process_config    = mdd_process_config,
        .ldo_recovery_complete = mdd_recovery_complete,
        .ldo_prepare           = mdd_prepare,
};

static int mdd_root_get(const struct lu_env *env,
			struct md_device *m, struct lu_fid *f)
{
	struct mdd_device *mdd = lu2mdd_dev(&m->md_lu_dev);

	ENTRY;
	*f = mdd->mdd_root_fid;
	RETURN(0);
}

/*
 * No permission check is needed.
 */
static int mdd_statfs(const struct lu_env *env, struct md_device *m,
		      struct obd_statfs *sfs)
{
	struct mdd_device *mdd = lu2mdd_dev(&m->md_lu_dev);
	int rc;

	ENTRY;

	rc = mdd_child_ops(mdd)->dt_statfs(env, mdd->mdd_child, sfs, NULL);

	sfs->os_namelen = min_t(__u32, sfs->os_namelen, NAME_MAX);

	RETURN(rc);
}

static const struct dt_device_param *mdd_dtconf_get(const struct lu_env *env,
						    struct md_device *m)
{
	struct mdd_device *mdd = lu2mdd_dev(&m->md_lu_dev);

	return &mdd->mdd_dt_conf;
}

static int mdd_llog_ctxt_get(const struct lu_env *env, struct md_device *m,
                             int idx, void **h)
{
        struct mdd_device *mdd = lu2mdd_dev(&m->md_lu_dev);

        *h = llog_group_get_ctxt(&mdd2obd_dev(mdd)->obd_olg, idx);
        return (*h == NULL ? -ENOENT : 0);
}

static struct lu_device *mdd_device_free(const struct lu_env *env,
					 struct lu_device *lu)
{
	struct mdd_device *m = lu2mdd_dev(lu);
	ENTRY;

	LASSERT(atomic_read(&lu->ld_ref) == 0);
	md_device_fini(&m->mdd_md_dev);
	OBD_FREE_PTR(m);
	RETURN(NULL);
}

static struct lu_device *mdd_device_alloc(const struct lu_env *env,
                                          struct lu_device_type *t,
                                          struct lustre_cfg *lcfg)
{
        struct lu_device  *l;
        struct mdd_device *m;

        OBD_ALLOC_PTR(m);
        if (m == NULL) {
                l = ERR_PTR(-ENOMEM);
        } else {
		int rc;

                l = mdd2lu_dev(m);
		md_device_init(&m->mdd_md_dev, t);
		rc = mdd_init0(env, m, t, lcfg);
		if (rc != 0) {
			mdd_device_free(env, l);
			l = ERR_PTR(rc);
		}
        }

        return l;
}

/*
 * we use exports to track all mdd users
 */
static int mdd_obd_connect(const struct lu_env *env, struct obd_export **exp,
			   struct obd_device *obd, struct obd_uuid *cluuid,
			   struct obd_connect_data *data, void *localdata)
{
	struct mdd_device    *mdd = lu2mdd_dev(obd->obd_lu_dev);
	struct lustre_handle  conn;
	int                   rc;
	ENTRY;

	CDEBUG(D_CONFIG, "connect #%d\n", mdd->mdd_connects);

	rc = class_connect(&conn, obd, cluuid);
	if (rc)
		RETURN(rc);

	*exp = class_conn2export(&conn);

	/* Why should there ever be more than 1 connect? */
	LASSERT(mdd->mdd_connects == 0);
	mdd->mdd_connects++;

	RETURN(0);
}

/*
 * once last export (we don't count self-export) disappeared
 * mdd can be released
 */
static int mdd_obd_disconnect(struct obd_export *exp)
{
	struct obd_device *obd = exp->exp_obd;
	struct mdd_device *mdd = lu2mdd_dev(obd->obd_lu_dev);
	int                rc, release = 0;
	ENTRY;

	mdd->mdd_connects--;
	if (mdd->mdd_connects == 0)
		release = 1;

	rc = class_disconnect(exp);

	if (rc == 0 && release)
		class_manual_cleanup(obd);
	RETURN(rc);
}

static int mdd_obd_get_info(const struct lu_env *env, struct obd_export *exp,
			    __u32 keylen, void *key, __u32 *vallen, void *val)
{
	int rc = -EINVAL;

	if (KEY_IS(KEY_OSP_CONNECTED)) {
		struct obd_device	*obd = exp->exp_obd;
		struct mdd_device	*mdd;

		if (!obd->obd_set_up || obd->obd_stopping)
			RETURN(-EAGAIN);

		mdd = lu2mdd_dev(obd->obd_lu_dev);
		LASSERT(mdd);
		rc = obd_get_info(env, mdd->mdd_child_exp, keylen, key, vallen,
				  val);
		RETURN(rc);
	}

	RETURN(rc);
}

static int mdd_obd_set_info_async(const struct lu_env *env,
				  struct obd_export *exp,
				  __u32 keylen, void *key,
				  __u32 vallen, void *val,
				  struct ptlrpc_request_set *set)
{
	struct obd_device	*obd = exp->exp_obd;
	struct mdd_device	*mdd;
	int			 rc;

	if (!obd->obd_set_up || obd->obd_stopping)
		RETURN(-EAGAIN);

	mdd = lu2mdd_dev(obd->obd_lu_dev);
	LASSERT(mdd);
	rc = obd_set_info_async(env, mdd->mdd_child_exp, keylen, key,
				vallen, val, set);
	RETURN(rc);
}

static const struct obd_ops mdd_obd_device_ops = {
	.o_owner	= THIS_MODULE,
	.o_connect	= mdd_obd_connect,
	.o_disconnect	= mdd_obd_disconnect,
	.o_get_info     = mdd_obd_get_info,
	.o_set_info_async = mdd_obd_set_info_async,
};

static int mdd_changelog_user_register(const struct lu_env *env,
				       struct mdd_device *mdd, int *id)
{
        struct llog_ctxt *ctxt;
        struct llog_changelog_user_rec *rec;
        int rc;
        ENTRY;

        ctxt = llog_get_context(mdd2obd_dev(mdd),
				LLOG_CHANGELOG_USER_ORIG_CTXT);
        if (ctxt == NULL)
                RETURN(-ENXIO);

        OBD_ALLOC_PTR(rec);
        if (rec == NULL) {
                llog_ctxt_put(ctxt);
                RETURN(-ENOMEM);
        }

	CFS_RACE(CFS_FAIL_CHLOG_USER_REG_UNREG_RACE);

        rec->cur_hdr.lrh_len = sizeof(*rec);
        rec->cur_hdr.lrh_type = CHANGELOG_USER_REC;
	spin_lock(&mdd->mdd_cl.mc_user_lock);
	if (mdd->mdd_cl.mc_lastuser == (unsigned int)(-1)) {
		spin_unlock(&mdd->mdd_cl.mc_user_lock);
		CERROR("Maximum number of changelog users exceeded!\n");
		GOTO(out, rc = -EOVERFLOW);
	}
	*id = rec->cur_id = ++mdd->mdd_cl.mc_lastuser;
	mdd->mdd_cl.mc_users++;
	rec->cur_endrec = mdd->mdd_cl.mc_index;

	rec->cur_time = (__u32)ktime_get_real_seconds();
	if (OBD_FAIL_CHECK(OBD_FAIL_TIME_IN_CHLOG_USER))
		rec->cur_time = 0;

	spin_unlock(&mdd->mdd_cl.mc_user_lock);

	rc = llog_cat_add(env, ctxt->loc_handle, &rec->cur_hdr, NULL);
	if (rc) {
		CWARN("%s: Failed to register changelog user %d: rc=%d\n",
		      mdd2obd_dev(mdd)->obd_name, *id, rc);
		spin_lock(&mdd->mdd_cl.mc_user_lock);
		mdd->mdd_cl.mc_users--;
		spin_unlock(&mdd->mdd_cl.mc_user_lock);
		GOTO(out, rc);
	}

        CDEBUG(D_IOCTL, "Registered changelog user %d\n", *id);

	/* Assume we want it on since somebody registered */
	rc = mdd_changelog_on(env, mdd);
	if (rc)
		GOTO(out, rc);

out:
        OBD_FREE_PTR(rec);
        llog_ctxt_put(ctxt);
        RETURN(rc);
}

struct mdd_changelog_user_purge {
	struct mdd_device *mcup_mdd;
	__u32 mcup_id;
	__u32 mcup_usercount;
	__u64 mcup_minrec;
	bool mcup_found;
};

/**
 * changelog_user_purge callback
 *
 * Is called once per user.
 *
 * Check to see the user requested is available from rec.
 * Truncate the changelog.
 * Keep track of the total number of users (calls).
 */
static int mdd_changelog_user_purge_cb(const struct lu_env *env,
				       struct llog_handle *llh,
				       struct llog_rec_hdr *hdr, void *data)
{
	struct llog_changelog_user_rec	*rec;
	struct mdd_changelog_user_purge	*mcup = data;
	struct llog_cookie               cookie;
	int				 rc;

	ENTRY;

	if ((llh->lgh_hdr->llh_flags & LLOG_F_IS_PLAIN) == 0)
		RETURN(-ENXIO);

	rec = container_of(hdr, struct llog_changelog_user_rec, cur_hdr);

	mcup->mcup_usercount++;

	if (rec->cur_id != mcup->mcup_id) {
		/* truncate to the lowest endrec that is not this user */
		mcup->mcup_minrec = min(mcup->mcup_minrec,
					rec->cur_endrec);
		RETURN(0);
	}

	/* Unregister this user */
	cookie.lgc_lgl = llh->lgh_id;
	cookie.lgc_index = hdr->lrh_index;

	rc = llog_cat_cancel_records(env, llh->u.phd.phd_cat_handle,
				     1, &cookie);
	if (rc == 0) {
		mcup->mcup_found = true;
		mcup->mcup_usercount--;
		spin_lock(&mcup->mcup_mdd->mdd_cl.mc_user_lock);
		mcup->mcup_mdd->mdd_cl.mc_users--;
		spin_unlock(&mcup->mcup_mdd->mdd_cl.mc_user_lock);
	}

	RETURN(rc);
}

int mdd_changelog_user_purge(const struct lu_env *env,
			     struct mdd_device *mdd, __u32 id)
{
	struct mdd_changelog_user_purge mcup = {
		.mcup_mdd = mdd,
		.mcup_id = id,
		.mcup_found = false,
		.mcup_usercount = 0,
		.mcup_minrec = ULLONG_MAX,
	};
	struct llog_ctxt *ctxt;
	int rc;

	ENTRY;

	CDEBUG(D_IOCTL, "%s: Purge request: id=%u\n",
	       mdd2obd_dev(mdd)->obd_name, id);

	ctxt = llog_get_context(mdd2obd_dev(mdd),
				LLOG_CHANGELOG_USER_ORIG_CTXT);
	if (ctxt == NULL ||
	    (ctxt->loc_handle->lgh_hdr->llh_flags & LLOG_F_IS_CAT) == 0)
		GOTO(out, rc = -ENXIO);

	rc = llog_cat_process(env, ctxt->loc_handle,
			      mdd_changelog_user_purge_cb, &mcup,
			      0, 0);

	OBD_FAIL_TIMEOUT(OBD_FAIL_LLOG_PURGE_DELAY, cfs_fail_val);

	if ((rc == 0) && (mcup.mcup_usercount == 0)) {
		spin_lock(&mdd->mdd_cl.mc_user_lock);
		if (mdd->mdd_cl.mc_users == 0) {
			/* No more users; turn changelogs off */
			CDEBUG(D_IOCTL, "turning off changelogs\n");
			rc = mdd_changelog_off(env, mdd);
		}
		spin_unlock(&mdd->mdd_cl.mc_user_lock);
	}

	if ((rc == 0) && mcup.mcup_found) {
		CDEBUG(D_IOCTL, "%s: Purging changelog entries for user %d "
		       "record=%llu\n",
		       mdd2obd_dev(mdd)->obd_name, id, mcup.mcup_minrec);
		/* Cancelling record 0 destroys the entire changelog, make sure
		   we don't do that unless we mean it. */
		if (mcup.mcup_minrec != 0 || mcup.mcup_usercount == 0) {
			rc = mdd_changelog_llog_cancel(env, mdd,
						       mcup.mcup_minrec);
		}
	} else {
		CWARN("%s: No changelog for user %u; rc=%d\n",
		      mdd2obd_dev(mdd)->obd_name, id, rc);
		GOTO(out, rc = -ENOENT);
	}

	CFS_RACE(CFS_FAIL_CHLOG_USER_REG_UNREG_RACE);

	EXIT;
out:
	if (ctxt != NULL)
		llog_ctxt_put(ctxt);

	return rc;
}

struct mdd_changelog_user_clear {
	__u64 mcuc_endrec;
	__u64 mcuc_minrec;
	__u32 mcuc_id;
	bool mcuc_flush;
};

/**
 * changelog_clear callback
 *
 * Is called once per user.
 *
 * Check to see the user requested is available from rec.
 * Check the oldest (smallest) record for boundary conditions.
 * Truncate the changelog.
 */
static int mdd_changelog_clear_cb(const struct lu_env *env,
				  struct llog_handle *llh,
				  struct llog_rec_hdr *hdr,
				  void *data)
{
	struct llog_changelog_user_rec *rec;
	struct mdd_changelog_user_clear *mcuc = data;
	int rc;

	ENTRY;

	if ((llh->lgh_hdr->llh_flags & LLOG_F_IS_PLAIN) == 0)
		RETURN(-ENXIO);

	rec = container_of(hdr, struct llog_changelog_user_rec, cur_hdr);

	/* Does the changelog id match the requested id? */
	if (rec->cur_id != mcuc->mcuc_id) {
		mcuc->mcuc_minrec = min(mcuc->mcuc_minrec,
					rec->cur_endrec);
		RETURN(0);
	}

	/* cur_endrec is the oldest purgeable record, make sure we're newer */
	if (rec->cur_endrec > mcuc->mcuc_endrec) {
		CDEBUG(D_IOCTL, "Request %llu out of range: %llu\n",
		       mcuc->mcuc_endrec, rec->cur_endrec);
		RETURN(-EINVAL);
	}

	/* Flag that we've met all the range and user checks.
	 * We now know the record to flush.
	 */
	rec->cur_endrec = mcuc->mcuc_endrec;

	rec->cur_time = (__u32)ktime_get_real_seconds();
	if (OBD_FAIL_CHECK(OBD_FAIL_TIME_IN_CHLOG_USER))
		rec->cur_time = 0;

	mcuc->mcuc_flush = true;

	CDEBUG(D_IOCTL, "Rewriting changelog user %u endrec to %llu\n",
	       mcuc->mcuc_id, rec->cur_endrec);

	/* Update the endrec */
	rc = llog_write(env, llh, hdr, hdr->lrh_index);

	RETURN(rc);
}

/**
 * Clear a changelog up to entry specified by endrec for user id.
 */
static int mdd_changelog_clear(const struct lu_env *env,
			       struct mdd_device *mdd, __u32 id,
			       __u64 endrec)
{
	struct mdd_changelog_user_clear mcuc = {
		.mcuc_id = id,
		.mcuc_minrec = endrec,
		.mcuc_flush = false,
	};
	struct llog_ctxt *ctxt;
	__u64 start_rec;
	int rc;

	ENTRY;

	CDEBUG(D_IOCTL, "%s: Purge request: id=%u, endrec=%llu\n",
	       mdd2obd_dev(mdd)->obd_name, id, endrec);
	/* start_rec is the newest (largest value) entry in the changelogs*/
	spin_lock(&mdd->mdd_cl.mc_lock);
	start_rec = mdd->mdd_cl.mc_index;
	spin_unlock(&mdd->mdd_cl.mc_lock);

	if (start_rec < endrec) {
		CDEBUG(D_IOCTL, "%s: Could not clear changelog, requested "\
		       "address out of range\n", mdd2obd_dev(mdd)->obd_name);
		RETURN(-EINVAL);
	}

	if (endrec == 0) {
		mcuc.mcuc_endrec = start_rec;
		mcuc.mcuc_minrec = ULLONG_MAX;
	} else {
		mcuc.mcuc_endrec = endrec;
	}

	ctxt = llog_get_context(mdd2obd_dev(mdd),
				LLOG_CHANGELOG_USER_ORIG_CTXT);
	if (ctxt == NULL ||
	    (ctxt->loc_handle->lgh_hdr->llh_flags & LLOG_F_IS_CAT) == 0)
		GOTO(out, rc = -ENXIO);

	rc = llog_cat_process(env, ctxt->loc_handle,
			      mdd_changelog_clear_cb, (void *)&mcuc,
			      0, 0);

	if (rc < 0) {
		CWARN("%s: Failure to clear the changelog for user %d: %d\n",
		      mdd2obd_dev(mdd)->obd_name, id, rc);
	} else if (mcuc.mcuc_flush) {
		/* Cancelling record 0 destroys the entire changelog, make sure
		   we don't do that unless we mean it. */
		if (mcuc.mcuc_minrec != 0) {
			CDEBUG(D_IOCTL, "%s: Purging changelog entries up "\
			       "to %llu\n", mdd2obd_dev(mdd)->obd_name,
			       mcuc.mcuc_minrec);

			rc = mdd_changelog_llog_cancel(env, mdd,
						      mcuc.mcuc_minrec);
		}
	} else {
		CWARN("%s: No entry for user %d\n",
		      mdd2obd_dev(mdd)->obd_name, id);
		rc = -ENOENT;
	}

	EXIT;
out:
	if (ctxt != NULL)
		llog_ctxt_put(ctxt);

	return rc;
}

/** mdd_iocontrol
 * May be called remotely from mdt_iocontrol_handle or locally from
 * mdt_iocontrol. Data may be freeform - remote handling doesn't enforce
 * an obd_ioctl_data format (but local ioctl handler does).
 * \param cmd - ioc
 * \param len - data len
 * \param karg - ioctl data, in kernel space
 */
static int mdd_iocontrol(const struct lu_env *env, struct md_device *m,
                         unsigned int cmd, int len, void *karg)
{
	struct mdd_device *mdd = lu2mdd_dev(&m->md_lu_dev);
	struct obd_ioctl_data *data = karg;
	int rc;
	ENTRY;

	/* Doesn't use obd_ioctl_data */
	switch (cmd) {
	case OBD_IOC_CHANGELOG_CLEAR: {
		struct changelog_setinfo *cs = karg;

		if (unlikely(!barrier_entry(mdd->mdd_bottom)))
			RETURN(-EINPROGRESS);

		rc = mdd_changelog_clear(env, mdd, cs->cs_id,
					 cs->cs_recno);
		barrier_exit(mdd->mdd_bottom);
		RETURN(rc);
	}
	case OBD_IOC_START_LFSCK: {
		rc = lfsck_start(env, mdd->mdd_bottom,
				 (struct lfsck_start_param *)karg);
		RETURN(rc);
	}
	case OBD_IOC_STOP_LFSCK: {
		rc = lfsck_stop(env, mdd->mdd_bottom,
				(struct lfsck_stop *)karg);
		RETURN(rc);
	}
	case OBD_IOC_QUERY_LFSCK: {
		rc = lfsck_query(env, mdd->mdd_bottom, NULL, NULL,
				 (struct lfsck_query *)karg);
		RETURN(rc);
	}
	}

	/* Below ioctls use obd_ioctl_data */
	if (len != sizeof(*data)) {
		CERROR("Bad ioctl size %d\n", len);
		RETURN(-EINVAL);
	}
	if (data->ioc_version != OBD_IOCTL_VERSION) {
		CERROR("Bad magic %x != %x\n", data->ioc_version,
		       OBD_IOCTL_VERSION);
		RETURN(-EINVAL);
	}

	switch (cmd) {
	case OBD_IOC_CHANGELOG_REG:
		if (unlikely(!barrier_entry(mdd->mdd_bottom)))
			RETURN(-EINPROGRESS);

		rc = mdd_changelog_user_register(env, mdd, &data->ioc_u32_1);
		barrier_exit(mdd->mdd_bottom);
		break;
	case OBD_IOC_CHANGELOG_DEREG:
		if (unlikely(!barrier_entry(mdd->mdd_bottom)))
			RETURN(-EINPROGRESS);

		/* explicitly clear changelog first, to protect from crash in
		 * the middle of purge that would lead to unregistered consumer
		 * but pending changelog entries
		 */
		rc = mdd_changelog_clear(env, mdd, data->ioc_u32_1, 0);
		if (!rc)
			rc = mdd_changelog_user_purge(env,
						      mdd, data->ioc_u32_1);

		barrier_exit(mdd->mdd_bottom);
		break;
	default:
		rc = -ENOTTY;
	}

	RETURN(rc);
}

/* type constructor/destructor: mdd_type_init, mdd_type_fini */
LU_TYPE_INIT_FINI(mdd, &mdd_thread_key);

static const struct md_device_operations mdd_ops = {
	.mdo_statfs         = mdd_statfs,
	.mdo_root_get	    = mdd_root_get,
	.mdo_llog_ctxt_get  = mdd_llog_ctxt_get,
	.mdo_iocontrol      = mdd_iocontrol,
	.mdo_dtconf_get     = mdd_dtconf_get,
};

static struct lu_device_type_operations mdd_device_type_ops = {
        .ldto_init = mdd_type_init,
        .ldto_fini = mdd_type_fini,

        .ldto_start = mdd_type_start,
        .ldto_stop  = mdd_type_stop,

        .ldto_device_alloc = mdd_device_alloc,
        .ldto_device_free  = mdd_device_free,

        .ldto_device_fini    = mdd_device_fini
};

static struct lu_device_type mdd_device_type = {
        .ldt_tags     = LU_DEVICE_MD,
        .ldt_name     = LUSTRE_MDD_NAME,
        .ldt_ops      = &mdd_device_type_ops,
        .ldt_ctx_tags = LCT_MD_THREAD
};

/* context key constructor: mdd_key_init */
LU_KEY_INIT(mdd, struct mdd_thread_info);

static void mdd_key_fini(const struct lu_context *ctx,
			 struct lu_context_key *key, void *data)
{
	struct mdd_thread_info *info = data;

	lu_buf_free(&info->mti_big_buf);
	lu_buf_free(&info->mti_link_buf);
	lu_buf_free(&info->mti_xattr_buf);

	OBD_FREE_PTR(info);
}

/* context key: mdd_thread_key */
LU_CONTEXT_KEY_DEFINE(mdd, LCT_MD_THREAD);

int mdd_generic_thread_start(struct mdd_generic_thread *thread,
			     int (*func)(void *), void *data, char *name)
{
	struct task_struct	*task;

	LASSERT(thread->mgt_init == false);
	init_completion(&thread->mgt_started);
	init_completion(&thread->mgt_finished);
	thread->mgt_data = data;
	thread->mgt_abort = false;
	thread->mgt_init = true;

	task = kthread_run(func, thread, name);
	if (IS_ERR(task)) {
		complete(&thread->mgt_finished);
		return PTR_ERR(task);
	}
	wait_for_completion(&thread->mgt_started);
	return 0;
}

void mdd_generic_thread_stop(struct mdd_generic_thread *thread)
{
	if (thread->mgt_init == true) {
		thread->mgt_abort = true;
		wait_for_completion(&thread->mgt_finished);
	}
}

static int __init mdd_init(void)
{
	int rc;

	rc = lu_kmem_init(mdd_caches);
	if (rc)
		return rc;

	changelog_orig_logops = llog_common_cat_ops;
	changelog_orig_logops.lop_write_rec = mdd_changelog_write_rec;

	rc = class_register_type(&mdd_obd_device_ops, NULL, false, NULL,
				 LUSTRE_MDD_NAME, &mdd_device_type);
	if (rc)
		lu_kmem_fini(mdd_caches);
	return rc;
}

static void __exit mdd_exit(void)
{
	class_unregister_type(LUSTRE_MDD_NAME);
	lu_kmem_fini(mdd_caches);
}

MODULE_AUTHOR("OpenSFS, Inc. <http://www.lustre.org/>");
MODULE_DESCRIPTION("Lustre Meta-data Device Driver ("LUSTRE_MDD_NAME")");
MODULE_VERSION(LUSTRE_VERSION_STRING);
MODULE_LICENSE("GPL");

module_init(mdd_init);
module_exit(mdd_exit);