/*

 * CDDL HEADER START

 *

 * The contents of this file are subject to the terms of the

 * Common Development and Distribution License (the "License").

 * You may not use this file except in compliance with the License.

 *

 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE

 * or http://www.opensolaris.org/os/licensing.

 * See the License for the specific language governing permissions

 * and limitations under the License.

 *

 * When distributing Covered Code, include this CDDL HEADER in each

 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.

 * If applicable, add the following below this CDDL HEADER, with the

 * fields enclosed by brackets "[]" replaced with your own identifying

 * information: Portions Copyright [yyyy] [name of copyright owner]

 *

 * CDDL HEADER END

 */

/*

 * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.

 * Use is subject to license terms.

 */

#pragma ident	"%Z%%M%	%I%	%E% SMI"

/*

 * This file contains all the routines used when modifying on-disk SPA state.

 * This includes opening, importing, destroying, exporting a pool, and syncing a

 * pool.

 */

#include <sys/zfs_context.h>

#include <sys/fm/fs/zfs.h>

#include <sys/spa_impl.h>

#include <sys/zio.h>

#include <sys/zio_checksum.h>

#include <sys/zio_compress.h>

#include <sys/dmu.h>

#include <sys/dmu_tx.h>

#include <sys/zap.h>

#include <sys/zil.h>

#include <sys/vdev_impl.h>

#include <sys/metaslab.h>

#include <sys/uberblock_impl.h>

#include <sys/txg.h>

#include <sys/avl.h>

#include <sys/dmu_traverse.h>

#include <sys/dmu_objset.h>

#include <sys/unique.h>

#include <sys/dsl_pool.h>

#include <sys/dsl_dataset.h>

#include <sys/dsl_dir.h>

#include <sys/dsl_prop.h>

#include <sys/dsl_synctask.h>

#include <sys/fs/zfs.h>

#include <sys/arc.h>

#include <sys/callb.h>

#include <sys/systeminfo.h>

#include <sys/sunddi.h>

#include <sys/spa_boot.h>

#include "zfs_prop.h"

#include "zfs_comutil.h"

int zio_taskq_threads = 8;

static void spa_sync_props(void *arg1, void *arg2, cred_t *cr, dmu_tx_t *tx);

/*

 * ==========================================================================

 * SPA properties routines

 * ==========================================================================

 */

/*

 * Add a (source=src, propname=propval) list to an nvlist.

 */

static void

spa_prop_add_list(nvlist_t *nvl, zpool_prop_t prop, char *strval,

    uint64_t intval, zprop_source_t src)

{

	const char *propname = zpool_prop_to_name(prop);

	nvlist_t *propval;

	VERIFY(nvlist_alloc(&propval, NV_UNIQUE_NAME, KM_SLEEP) == 0);

	VERIFY(nvlist_add_uint64(propval, ZPROP_SOURCE, src) == 0);

	if (strval != NULL)

		VERIFY(nvlist_add_string(propval, ZPROP_VALUE, strval) == 0);

	else

		VERIFY(nvlist_add_uint64(propval, ZPROP_VALUE, intval) == 0);

	VERIFY(nvlist_add_nvlist(nvl, propname, propval) == 0);

	nvlist_free(propval);

}

/*

 * Get property values from the spa configuration.

 */

static void

spa_prop_get_config(spa_t *spa, nvlist_t **nvp)

{

	uint64_t size = spa_get_space(spa);

	uint64_t used = spa_get_alloc(spa);

	uint64_t cap, version;

	zprop_source_t src = ZPROP_SRC_NONE;

	spa_config_dirent_t *dp;

	/*

	 * readonly properties

	 */

	spa_prop_add_list(*nvp, ZPOOL_PROP_NAME, spa->spa_name, 0, src);

	spa_prop_add_list(*nvp, ZPOOL_PROP_SIZE, NULL, size, src);

	spa_prop_add_list(*nvp, ZPOOL_PROP_USED, NULL, used, src);

	spa_prop_add_list(*nvp, ZPOOL_PROP_AVAILABLE, NULL, size - used, src);

	cap = (size == 0) ? 0 : (used * 100 / size);

	spa_prop_add_list(*nvp, ZPOOL_PROP_CAPACITY, NULL, cap, src);

	spa_prop_add_list(*nvp, ZPOOL_PROP_GUID, NULL, spa_guid(spa), src);

	spa_prop_add_list(*nvp, ZPOOL_PROP_HEALTH, NULL,

	    spa->spa_root_vdev->vdev_state, src);

	/*

	 * settable properties that are not stored in the pool property object.

	 */

	version = spa_version(spa);

	if (version == zpool_prop_default_numeric(ZPOOL_PROP_VERSION))

		src = ZPROP_SRC_DEFAULT;

	else

		src = ZPROP_SRC_LOCAL;

	spa_prop_add_list(*nvp, ZPOOL_PROP_VERSION, NULL, version, src);

	if (spa->spa_root != NULL)

		spa_prop_add_list(*nvp, ZPOOL_PROP_ALTROOT, spa->spa_root,

		    0, ZPROP_SRC_LOCAL);

	if ((dp = list_head(&spa->spa_config_list)) != NULL) {

		if (dp->scd_path == NULL) {

			spa_prop_add_list(*nvp, ZPOOL_PROP_CACHEFILE,

			    "none", 0, ZPROP_SRC_LOCAL);

		} else if (strcmp(dp->scd_path, spa_config_path) != 0) {

			spa_prop_add_list(*nvp, ZPOOL_PROP_CACHEFILE,

			    dp->scd_path, 0, ZPROP_SRC_LOCAL);

		}

	}

}

/*

 * Get zpool property values.

 */

int

spa_prop_get(spa_t *spa, nvlist_t **nvp)

{

	zap_cursor_t zc;

	zap_attribute_t za;

	objset_t *mos = spa->spa_meta_objset;

	int err;

	VERIFY(nvlist_alloc(nvp, NV_UNIQUE_NAME, KM_SLEEP) == 0);

	/*

	 * Get properties from the spa config.

	 */

	spa_prop_get_config(spa, nvp);

	mutex_enter(&spa->spa_props_lock);

	/* If no pool property object, no more prop to get. */

	if (spa->spa_pool_props_object == 0) {

		mutex_exit(&spa->spa_props_lock);

		return (0);

	}

	/*

	 * Get properties from the MOS pool property object.

	 */

	for (zap_cursor_init(&zc, mos, spa->spa_pool_props_object);

	    (err = zap_cursor_retrieve(&zc, &za)) == 0;

	    zap_cursor_advance(&zc)) {

		uint64_t intval = 0;

		char *strval = NULL;

		zprop_source_t src = ZPROP_SRC_DEFAULT;

		zpool_prop_t prop;

		if ((prop = zpool_name_to_prop(za.za_name)) == ZPROP_INVAL)

			continue;

		switch (za.za_integer_length) {

		case 8:

			/* integer property */

			if (za.za_first_integer !=

			    zpool_prop_default_numeric(prop))

				src = ZPROP_SRC_LOCAL;

			if (prop == ZPOOL_PROP_BOOTFS) {

				dsl_pool_t *dp;

				dsl_dataset_t *ds = NULL;

				dp = spa_get_dsl(spa);

				rw_enter(&dp->dp_config_rwlock, RW_READER);

					rw_exit(&dp->dp_config_rwlock);

					break;

				}

				strval = kmem_alloc(

				    MAXNAMELEN + strlen(MOS_DIR_NAME) + 1,

				    KM_SLEEP);

				dsl_dataset_name(ds, strval);

				rw_exit(&dp->dp_config_rwlock);

			} else {

				strval = NULL;

				intval = za.za_first_integer;

			}

			spa_prop_add_list(*nvp, prop, strval, intval, src);

			if (strval != NULL)

				kmem_free(strval,

				    MAXNAMELEN + strlen(MOS_DIR_NAME) + 1);

			break;

		case 1:

			/* string property */

			strval = kmem_alloc(za.za_num_integers, KM_SLEEP);

			err = zap_lookup(mos, spa->spa_pool_props_object,

			    za.za_name, 1, za.za_num_integers, strval);

			if (err) {

				kmem_free(strval, za.za_num_integers);

				break;

			}

			spa_prop_add_list(*nvp, prop, strval, 0, src);

			kmem_free(strval, za.za_num_integers);

			break;

		default:

			break;

		}

	}

	zap_cursor_fini(&zc);

	mutex_exit(&spa->spa_props_lock);

out:

	if (err && err != ENOENT) {

		nvlist_free(*nvp);

		*nvp = NULL;

		return (err);

	}

	return (0);

}

/*

 * Validate the given pool properties nvlist and modify the list

 * for the property values to be set.

 */

static int

spa_prop_validate(spa_t *spa, nvlist_t *props)

{

	nvpair_t *elem;

	int error = 0, reset_bootfs = 0;

	uint64_t objnum;

	elem = NULL;

	while ((elem = nvlist_next_nvpair(props, elem)) != NULL) {

		zpool_prop_t prop;

		char *propname, *strval;

		uint64_t intval;

		vdev_t *rvdev;

		char *vdev_type;

		objset_t *os;

		char *slash;

		propname = nvpair_name(elem);

		if ((prop = zpool_name_to_prop(propname)) == ZPROP_INVAL)

			return (EINVAL);

		switch (prop) {

		case ZPOOL_PROP_VERSION:

			error = nvpair_value_uint64(elem, &intval);

			if (!error &&

			    (intval < spa_version(spa) || intval > SPA_VERSION))

				error = EINVAL;

			break;

		case ZPOOL_PROP_DELEGATION:

		case ZPOOL_PROP_AUTOREPLACE:

			error = nvpair_value_uint64(elem, &intval);

			if (!error && intval > 1)

				error = EINVAL;

			break;

		case ZPOOL_PROP_BOOTFS:

			if (spa_version(spa) < SPA_VERSION_BOOTFS) {

				error = ENOTSUP;

				break;

			}

			/*

			 * A bootable filesystem can not be on a RAIDZ pool

			 * nor a striped pool with more than 1 device.

			 */

			rvdev = spa->spa_root_vdev;

			vdev_type =

			    rvdev->vdev_child[0]->vdev_ops->vdev_op_type;

			if (rvdev->vdev_children > 1 ||

			    strcmp(vdev_type, VDEV_TYPE_RAIDZ) == 0 ||

			    strcmp(vdev_type, VDEV_TYPE_MISSING) == 0) {

				error = ENOTSUP;

				break;

			}

			reset_bootfs = 1;

			error = nvpair_value_string(elem, &strval);

			if (!error) {

				if (strval == NULL || strval[0] == '\0') {

					objnum = zpool_prop_default_numeric(

					    ZPOOL_PROP_BOOTFS);

					break;

				}

				if (error = dmu_objset_open(strval, DMU_OST_ZFS,

					break;

				objnum = dmu_objset_id(os);

				dmu_objset_close(os);

			}

			break;

		case ZPOOL_PROP_FAILUREMODE:

			error = nvpair_value_uint64(elem, &intval);

			if (!error && (intval < ZIO_FAILURE_MODE_WAIT ||

			    intval > ZIO_FAILURE_MODE_PANIC))

				error = EINVAL;

			/*

			 * This is a special case which only occurs when

			 * the pool has completely failed. This allows

			 * the user to change the in-core failmode property

			 * without syncing it out to disk (I/Os might

			 * currently be blocked). We do this by returning

			 * EIO to the caller (spa_prop_set) to trick it

			 * into thinking we encountered a property validation

			 * error.

			 */

			if (!error && spa_state(spa) == POOL_STATE_IO_FAILURE) {

				spa->spa_failmode = intval;

				error = EIO;

			}

			break;

		case ZPOOL_PROP_CACHEFILE:

			if ((error = nvpair_value_string(elem, &strval)) != 0)

				break;

			if (strval[0] == '\0')

				break;

			if (strcmp(strval, "none") == 0)

				break;

			if (strval[0] != '/') {

				error = EINVAL;

				break;

			}

			slash = strrchr(strval, '/');

			ASSERT(slash != NULL);

			if (slash[1] == '\0' || strcmp(slash, "/.") == 0 ||

			    strcmp(slash, "/..") == 0)

				error = EINVAL;

			break;

		}

		if (error)

			break;

	}

	if (!error && reset_bootfs) {

		error = nvlist_remove(props,

		    zpool_prop_to_name(ZPOOL_PROP_BOOTFS), DATA_TYPE_STRING);

		if (!error) {

			error = nvlist_add_uint64(props,

			    zpool_prop_to_name(ZPOOL_PROP_BOOTFS), objnum);

		}

	}

	return (error);

}

int

spa_prop_set(spa_t *spa, nvlist_t *nvp)

{

	int error;

	if ((error = spa_prop_validate(spa, nvp)) != 0)

		return (error);

	return (dsl_sync_task_do(spa_get_dsl(spa), NULL, spa_sync_props,

	    spa, nvp, 3));

}

/*

 * If the bootfs property value is dsobj, clear it.

 */

void

spa_prop_clear_bootfs(spa_t *spa, uint64_t dsobj, dmu_tx_t *tx)

{

	if (spa->spa_bootfs == dsobj && spa->spa_pool_props_object != 0) {

		VERIFY(zap_remove(spa->spa_meta_objset,

		    spa->spa_pool_props_object,

		    zpool_prop_to_name(ZPOOL_PROP_BOOTFS), tx) == 0);

		spa->spa_bootfs = 0;

	}

}

/*

 * ==========================================================================

 * SPA state manipulation (open/create/destroy/import/export)

 * ==========================================================================

 */

static int

spa_error_entry_compare(const void *a, const void *b)

{

	spa_error_entry_t *sa = (spa_error_entry_t *)a;

	spa_error_entry_t *sb = (spa_error_entry_t *)b;

	int ret;

	ret = bcmp(&sa->se_bookmark, &sb->se_bookmark,

	    sizeof (zbookmark_t));

	if (ret < 0)

		return (-1);

	else if (ret > 0)

		return (1);

	else

		return (0);

}

/*

 * Utility function which retrieves copies of the current logs and

 * re-initializes them in the process.

 */

void

spa_get_errlists(spa_t *spa, avl_tree_t *last, avl_tree_t *scrub)

{

	ASSERT(MUTEX_HELD(&spa->spa_errlist_lock));

	bcopy(&spa->spa_errlist_last, last, sizeof (avl_tree_t));

	bcopy(&spa->spa_errlist_scrub, scrub, sizeof (avl_tree_t));

	avl_create(&spa->spa_errlist_scrub,

	    spa_error_entry_compare, sizeof (spa_error_entry_t),

	    offsetof(spa_error_entry_t, se_avl));

	avl_create(&spa->spa_errlist_last,

	    spa_error_entry_compare, sizeof (spa_error_entry_t),

	    offsetof(spa_error_entry_t, se_avl));

}

/*

 * Activate an uninitialized pool.

 */

static void

spa_activate(spa_t *spa)

{

	int t;

	ASSERT(spa->spa_state == POOL_STATE_UNINITIALIZED);

	spa->spa_state = POOL_STATE_ACTIVE;

	spa->spa_normal_class = metaslab_class_create();

	spa->spa_log_class = metaslab_class_create();

	for (t = 0; t < ZIO_TYPES; t++) {

		spa->spa_zio_issue_taskq[t] = taskq_create("spa_zio_issue",

		    zio_taskq_threads, maxclsyspri, 50, INT_MAX,

		    TASKQ_PREPOPULATE);

		spa->spa_zio_intr_taskq[t] = taskq_create("spa_zio_intr",

		    zio_taskq_threads, maxclsyspri, 50, INT_MAX,

		    TASKQ_PREPOPULATE);

	}

	list_create(&spa->spa_dirty_list, sizeof (vdev_t),

	    offsetof(vdev_t, vdev_dirty_node));

	list_create(&spa->spa_zio_list, sizeof (zio_t),