2008-11-20 23:01:55 +03:00
|
|
|
/*
|
|
|
|
* 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
|
|
|
|
*/
|
|
|
|
/*
|
2010-05-29 00:45:14 +04:00
|
|
|
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
|
2013-03-08 22:41:28 +04:00
|
|
|
* Copyright (c) 2013 by Delphix. All rights reserved.
|
2013-05-23 21:07:25 +04:00
|
|
|
* Copyright (c) 2013 Martin Matuska. All rights reserved.
|
2008-11-20 23:01:55 +03:00
|
|
|
*/
|
|
|
|
|
|
|
|
#include <sys/dmu.h>
|
|
|
|
#include <sys/dmu_objset.h>
|
|
|
|
#include <sys/dmu_tx.h>
|
|
|
|
#include <sys/dsl_dataset.h>
|
|
|
|
#include <sys/dsl_dir.h>
|
|
|
|
#include <sys/dsl_prop.h>
|
|
|
|
#include <sys/dsl_synctask.h>
|
|
|
|
#include <sys/dsl_deleg.h>
|
|
|
|
#include <sys/spa.h>
|
2010-05-29 00:45:14 +04:00
|
|
|
#include <sys/metaslab.h>
|
2008-11-20 23:01:55 +03:00
|
|
|
#include <sys/zap.h>
|
|
|
|
#include <sys/zio.h>
|
|
|
|
#include <sys/arc.h>
|
|
|
|
#include <sys/sunddi.h>
|
2013-12-07 02:20:22 +04:00
|
|
|
#include <sys/zvol.h>
|
2008-11-20 23:01:55 +03:00
|
|
|
#include "zfs_namecheck.h"
|
|
|
|
|
|
|
|
static uint64_t dsl_dir_space_towrite(dsl_dir_t *dd);
|
|
|
|
|
|
|
|
/* ARGSUSED */
|
|
|
|
static void
|
|
|
|
dsl_dir_evict(dmu_buf_t *db, void *arg)
|
|
|
|
{
|
|
|
|
dsl_dir_t *dd = arg;
|
|
|
|
int t;
|
2013-11-01 23:26:11 +04:00
|
|
|
ASSERTV(dsl_pool_t *dp = dd->dd_pool);
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
for (t = 0; t < TXG_SIZE; t++) {
|
|
|
|
ASSERT(!txg_list_member(&dp->dp_dirty_dirs, dd, t));
|
|
|
|
ASSERT(dd->dd_tempreserved[t] == 0);
|
|
|
|
ASSERT(dd->dd_space_towrite[t] == 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (dd->dd_parent)
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_rele(dd->dd_parent, dd);
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
spa_close(dd->dd_pool->dp_spa, dd);
|
|
|
|
|
|
|
|
/*
|
2010-05-29 00:45:14 +04:00
|
|
|
* The props callback list should have been cleaned up by
|
|
|
|
* objset_evict().
|
2008-11-20 23:01:55 +03:00
|
|
|
*/
|
|
|
|
list_destroy(&dd->dd_prop_cbs);
|
|
|
|
mutex_destroy(&dd->dd_lock);
|
|
|
|
kmem_free(dd, sizeof (dsl_dir_t));
|
|
|
|
}
|
|
|
|
|
|
|
|
int
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_hold_obj(dsl_pool_t *dp, uint64_t ddobj,
|
2008-11-20 23:01:55 +03:00
|
|
|
const char *tail, void *tag, dsl_dir_t **ddp)
|
|
|
|
{
|
|
|
|
dmu_buf_t *dbuf;
|
|
|
|
dsl_dir_t *dd;
|
|
|
|
int err;
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
ASSERT(dsl_pool_config_held(dp));
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
err = dmu_bonus_hold(dp->dp_meta_objset, ddobj, tag, &dbuf);
|
2013-09-04 16:00:57 +04:00
|
|
|
if (err != 0)
|
2008-11-20 23:01:55 +03:00
|
|
|
return (err);
|
|
|
|
dd = dmu_buf_get_user(dbuf);
|
|
|
|
#ifdef ZFS_DEBUG
|
|
|
|
{
|
|
|
|
dmu_object_info_t doi;
|
|
|
|
dmu_object_info_from_db(dbuf, &doi);
|
|
|
|
ASSERT3U(doi.doi_type, ==, DMU_OT_DSL_DIR);
|
2008-12-03 23:09:06 +03:00
|
|
|
ASSERT3U(doi.doi_bonus_size, >=, sizeof (dsl_dir_phys_t));
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
#endif
|
|
|
|
if (dd == NULL) {
|
|
|
|
dsl_dir_t *winner;
|
|
|
|
|
2012-05-07 21:49:51 +04:00
|
|
|
dd = kmem_zalloc(sizeof (dsl_dir_t), KM_PUSHPAGE);
|
2008-11-20 23:01:55 +03:00
|
|
|
dd->dd_object = ddobj;
|
|
|
|
dd->dd_dbuf = dbuf;
|
|
|
|
dd->dd_pool = dp;
|
|
|
|
dd->dd_phys = dbuf->db_data;
|
|
|
|
mutex_init(&dd->dd_lock, NULL, MUTEX_DEFAULT, NULL);
|
|
|
|
|
|
|
|
list_create(&dd->dd_prop_cbs, sizeof (dsl_prop_cb_record_t),
|
|
|
|
offsetof(dsl_prop_cb_record_t, cbr_node));
|
|
|
|
|
2010-05-29 00:45:14 +04:00
|
|
|
dsl_dir_snap_cmtime_update(dd);
|
|
|
|
|
2008-11-20 23:01:55 +03:00
|
|
|
if (dd->dd_phys->dd_parent_obj) {
|
2013-09-04 16:00:57 +04:00
|
|
|
err = dsl_dir_hold_obj(dp, dd->dd_phys->dd_parent_obj,
|
2008-11-20 23:01:55 +03:00
|
|
|
NULL, dd, &dd->dd_parent);
|
2013-09-04 16:00:57 +04:00
|
|
|
if (err != 0)
|
2008-12-03 23:09:06 +03:00
|
|
|
goto errout;
|
2008-11-20 23:01:55 +03:00
|
|
|
if (tail) {
|
|
|
|
#ifdef ZFS_DEBUG
|
|
|
|
uint64_t foundobj;
|
|
|
|
|
|
|
|
err = zap_lookup(dp->dp_meta_objset,
|
|
|
|
dd->dd_parent->dd_phys->dd_child_dir_zapobj,
|
|
|
|
tail, sizeof (foundobj), 1, &foundobj);
|
|
|
|
ASSERT(err || foundobj == ddobj);
|
|
|
|
#endif
|
|
|
|
(void) strcpy(dd->dd_myname, tail);
|
|
|
|
} else {
|
|
|
|
err = zap_value_search(dp->dp_meta_objset,
|
|
|
|
dd->dd_parent->dd_phys->dd_child_dir_zapobj,
|
|
|
|
ddobj, 0, dd->dd_myname);
|
|
|
|
}
|
2013-09-04 16:00:57 +04:00
|
|
|
if (err != 0)
|
2008-12-03 23:09:06 +03:00
|
|
|
goto errout;
|
2008-11-20 23:01:55 +03:00
|
|
|
} else {
|
|
|
|
(void) strcpy(dd->dd_myname, spa_name(dp->dp_spa));
|
|
|
|
}
|
|
|
|
|
2010-05-29 00:45:14 +04:00
|
|
|
if (dsl_dir_is_clone(dd)) {
|
|
|
|
dmu_buf_t *origin_bonus;
|
|
|
|
dsl_dataset_phys_t *origin_phys;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We can't open the origin dataset, because
|
|
|
|
* that would require opening this dsl_dir.
|
|
|
|
* Just look at its phys directly instead.
|
|
|
|
*/
|
|
|
|
err = dmu_bonus_hold(dp->dp_meta_objset,
|
|
|
|
dd->dd_phys->dd_origin_obj, FTAG, &origin_bonus);
|
2013-09-04 16:00:57 +04:00
|
|
|
if (err != 0)
|
2010-05-29 00:45:14 +04:00
|
|
|
goto errout;
|
|
|
|
origin_phys = origin_bonus->db_data;
|
|
|
|
dd->dd_origin_txg =
|
|
|
|
origin_phys->ds_creation_txg;
|
|
|
|
dmu_buf_rele(origin_bonus, FTAG);
|
|
|
|
}
|
|
|
|
|
2008-11-20 23:01:55 +03:00
|
|
|
winner = dmu_buf_set_user_ie(dbuf, dd, &dd->dd_phys,
|
|
|
|
dsl_dir_evict);
|
|
|
|
if (winner) {
|
|
|
|
if (dd->dd_parent)
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_rele(dd->dd_parent, dd);
|
2008-11-20 23:01:55 +03:00
|
|
|
mutex_destroy(&dd->dd_lock);
|
|
|
|
kmem_free(dd, sizeof (dsl_dir_t));
|
|
|
|
dd = winner;
|
|
|
|
} else {
|
|
|
|
spa_open_ref(dp->dp_spa, dd);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The dsl_dir_t has both open-to-close and instantiate-to-evict
|
|
|
|
* holds on the spa. We need the open-to-close holds because
|
|
|
|
* otherwise the spa_refcnt wouldn't change when we open a
|
|
|
|
* dir which the spa also has open, so we could incorrectly
|
|
|
|
* think it was OK to unload/export/destroy the pool. We need
|
|
|
|
* the instantiate-to-evict hold because the dsl_dir_t has a
|
|
|
|
* pointer to the dd_pool, which has a pointer to the spa_t.
|
|
|
|
*/
|
|
|
|
spa_open_ref(dp->dp_spa, tag);
|
|
|
|
ASSERT3P(dd->dd_pool, ==, dp);
|
|
|
|
ASSERT3U(dd->dd_object, ==, ddobj);
|
|
|
|
ASSERT3P(dd->dd_dbuf, ==, dbuf);
|
|
|
|
*ddp = dd;
|
|
|
|
return (0);
|
2008-12-03 23:09:06 +03:00
|
|
|
|
|
|
|
errout:
|
|
|
|
if (dd->dd_parent)
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_rele(dd->dd_parent, dd);
|
2008-12-03 23:09:06 +03:00
|
|
|
mutex_destroy(&dd->dd_lock);
|
|
|
|
kmem_free(dd, sizeof (dsl_dir_t));
|
|
|
|
dmu_buf_rele(dbuf, tag);
|
|
|
|
return (err);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_rele(dsl_dir_t *dd, void *tag)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
|
|
|
dprintf_dd(dd, "%s\n", "");
|
|
|
|
spa_close(dd->dd_pool->dp_spa, tag);
|
|
|
|
dmu_buf_rele(dd->dd_dbuf, tag);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* buf must be long enough (MAXNAMELEN + strlen(MOS_DIR_NAME) + 1 should do) */
|
|
|
|
void
|
|
|
|
dsl_dir_name(dsl_dir_t *dd, char *buf)
|
|
|
|
{
|
|
|
|
if (dd->dd_parent) {
|
|
|
|
dsl_dir_name(dd->dd_parent, buf);
|
|
|
|
(void) strcat(buf, "/");
|
|
|
|
} else {
|
|
|
|
buf[0] = '\0';
|
|
|
|
}
|
|
|
|
if (!MUTEX_HELD(&dd->dd_lock)) {
|
|
|
|
/*
|
|
|
|
* recursive mutex so that we can use
|
|
|
|
* dprintf_dd() with dd_lock held
|
|
|
|
*/
|
|
|
|
mutex_enter(&dd->dd_lock);
|
|
|
|
(void) strcat(buf, dd->dd_myname);
|
|
|
|
mutex_exit(&dd->dd_lock);
|
|
|
|
} else {
|
|
|
|
(void) strcat(buf, dd->dd_myname);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2012-12-15 04:13:40 +04:00
|
|
|
/* Calculate name length, avoiding all the strcat calls of dsl_dir_name */
|
2008-11-20 23:01:55 +03:00
|
|
|
int
|
|
|
|
dsl_dir_namelen(dsl_dir_t *dd)
|
|
|
|
{
|
|
|
|
int result = 0;
|
|
|
|
|
|
|
|
if (dd->dd_parent) {
|
|
|
|
/* parent's name + 1 for the "/" */
|
|
|
|
result = dsl_dir_namelen(dd->dd_parent) + 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!MUTEX_HELD(&dd->dd_lock)) {
|
|
|
|
/* see dsl_dir_name */
|
|
|
|
mutex_enter(&dd->dd_lock);
|
|
|
|
result += strlen(dd->dd_myname);
|
|
|
|
mutex_exit(&dd->dd_lock);
|
|
|
|
} else {
|
|
|
|
result += strlen(dd->dd_myname);
|
|
|
|
}
|
|
|
|
|
|
|
|
return (result);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
getcomponent(const char *path, char *component, const char **nextp)
|
|
|
|
{
|
|
|
|
char *p;
|
2013-09-04 16:00:57 +04:00
|
|
|
|
2009-07-03 02:44:48 +04:00
|
|
|
if ((path == NULL) || (path[0] == '\0'))
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(ENOENT));
|
2008-11-20 23:01:55 +03:00
|
|
|
/* This would be a good place to reserve some namespace... */
|
|
|
|
p = strpbrk(path, "/@");
|
|
|
|
if (p && (p[1] == '/' || p[1] == '@')) {
|
|
|
|
/* two separators in a row */
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(EINVAL));
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
if (p == NULL || p == path) {
|
|
|
|
/*
|
|
|
|
* if the first thing is an @ or /, it had better be an
|
|
|
|
* @ and it had better not have any more ats or slashes,
|
|
|
|
* and it had better have something after the @.
|
|
|
|
*/
|
|
|
|
if (p != NULL &&
|
|
|
|
(p[0] != '@' || strpbrk(path+1, "/@") || p[1] == '\0'))
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(EINVAL));
|
2008-11-20 23:01:55 +03:00
|
|
|
if (strlen(path) >= MAXNAMELEN)
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(ENAMETOOLONG));
|
2008-11-20 23:01:55 +03:00
|
|
|
(void) strcpy(component, path);
|
|
|
|
p = NULL;
|
|
|
|
} else if (p[0] == '/') {
|
2013-09-04 16:00:57 +04:00
|
|
|
if (p - path >= MAXNAMELEN)
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(ENAMETOOLONG));
|
2008-11-20 23:01:55 +03:00
|
|
|
(void) strncpy(component, path, p - path);
|
2013-09-04 16:00:57 +04:00
|
|
|
component[p - path] = '\0';
|
2008-11-20 23:01:55 +03:00
|
|
|
p++;
|
|
|
|
} else if (p[0] == '@') {
|
|
|
|
/*
|
|
|
|
* if the next separator is an @, there better not be
|
|
|
|
* any more slashes.
|
|
|
|
*/
|
|
|
|
if (strchr(path, '/'))
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(EINVAL));
|
2013-09-04 16:00:57 +04:00
|
|
|
if (p - path >= MAXNAMELEN)
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(ENAMETOOLONG));
|
2008-11-20 23:01:55 +03:00
|
|
|
(void) strncpy(component, path, p - path);
|
2013-09-04 16:00:57 +04:00
|
|
|
component[p - path] = '\0';
|
2008-11-20 23:01:55 +03:00
|
|
|
} else {
|
2013-09-04 16:00:57 +04:00
|
|
|
panic("invalid p=%p", (void *)p);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
*nextp = p;
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2013-09-04 16:00:57 +04:00
|
|
|
* Return the dsl_dir_t, and possibly the last component which couldn't
|
|
|
|
* be found in *tail. The name must be in the specified dsl_pool_t. This
|
|
|
|
* thread must hold the dp_config_rwlock for the pool. Returns NULL if the
|
|
|
|
* path is bogus, or if tail==NULL and we couldn't parse the whole name.
|
|
|
|
* (*tail)[0] == '@' means that the last component is a snapshot.
|
2008-11-20 23:01:55 +03:00
|
|
|
*/
|
|
|
|
int
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_hold(dsl_pool_t *dp, const char *name, void *tag,
|
2008-11-20 23:01:55 +03:00
|
|
|
dsl_dir_t **ddp, const char **tailp)
|
|
|
|
{
|
2010-08-26 21:53:31 +04:00
|
|
|
char *buf;
|
2013-09-04 16:00:57 +04:00
|
|
|
const char *spaname, *next, *nextnext = NULL;
|
2008-11-20 23:01:55 +03:00
|
|
|
int err;
|
|
|
|
dsl_dir_t *dd;
|
|
|
|
uint64_t ddobj;
|
|
|
|
|
2013-02-06 22:15:13 +04:00
|
|
|
buf = kmem_alloc(MAXNAMELEN, KM_PUSHPAGE);
|
2008-11-20 23:01:55 +03:00
|
|
|
err = getcomponent(name, buf, &next);
|
2013-09-04 16:00:57 +04:00
|
|
|
if (err != 0)
|
2010-08-26 21:53:31 +04:00
|
|
|
goto error;
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
/* Make sure the name is in the specified pool. */
|
|
|
|
spaname = spa_name(dp->dp_spa);
|
|
|
|
if (strcmp(buf, spaname) != 0) {
|
2013-03-08 22:41:28 +04:00
|
|
|
err = SET_ERROR(EINVAL);
|
2013-09-04 16:00:57 +04:00
|
|
|
goto error;
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
ASSERT(dsl_pool_config_held(dp));
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
err = dsl_dir_hold_obj(dp, dp->dp_root_dir_obj, NULL, tag, &dd);
|
|
|
|
if (err != 0) {
|
2010-08-26 21:53:31 +04:00
|
|
|
goto error;
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
while (next != NULL) {
|
|
|
|
dsl_dir_t *child_ds;
|
|
|
|
err = getcomponent(next, buf, &nextnext);
|
2013-09-04 16:00:57 +04:00
|
|
|
if (err != 0)
|
2008-11-20 23:01:55 +03:00
|
|
|
break;
|
|
|
|
ASSERT(next[0] != '\0');
|
|
|
|
if (next[0] == '@')
|
|
|
|
break;
|
|
|
|
dprintf("looking up %s in obj%lld\n",
|
|
|
|
buf, dd->dd_phys->dd_child_dir_zapobj);
|
|
|
|
|
|
|
|
err = zap_lookup(dp->dp_meta_objset,
|
|
|
|
dd->dd_phys->dd_child_dir_zapobj,
|
|
|
|
buf, sizeof (ddobj), 1, &ddobj);
|
2013-09-04 16:00:57 +04:00
|
|
|
if (err != 0) {
|
2008-11-20 23:01:55 +03:00
|
|
|
if (err == ENOENT)
|
|
|
|
err = 0;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
err = dsl_dir_hold_obj(dp, ddobj, buf, tag, &child_ds);
|
|
|
|
if (err != 0)
|
2008-11-20 23:01:55 +03:00
|
|
|
break;
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_rele(dd, tag);
|
2008-11-20 23:01:55 +03:00
|
|
|
dd = child_ds;
|
|
|
|
next = nextnext;
|
|
|
|
}
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
if (err != 0) {
|
|
|
|
dsl_dir_rele(dd, tag);
|
2010-08-26 21:53:31 +04:00
|
|
|
goto error;
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* It's an error if there's more than one component left, or
|
|
|
|
* tailp==NULL and there's any component left.
|
|
|
|
*/
|
|
|
|
if (next != NULL &&
|
|
|
|
(tailp == NULL || (nextnext && nextnext[0] != '\0'))) {
|
|
|
|
/* bad path name */
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_rele(dd, tag);
|
2008-11-20 23:01:55 +03:00
|
|
|
dprintf("next=%p (%s) tail=%p\n", next, next?next:"", tailp);
|
2013-03-08 22:41:28 +04:00
|
|
|
err = SET_ERROR(ENOENT);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
2013-09-04 16:00:57 +04:00
|
|
|
if (tailp != NULL)
|
2008-11-20 23:01:55 +03:00
|
|
|
*tailp = next;
|
|
|
|
*ddp = dd;
|
2010-08-26 21:53:31 +04:00
|
|
|
error:
|
|
|
|
kmem_free(buf, MAXNAMELEN);
|
2008-11-20 23:01:55 +03:00
|
|
|
return (err);
|
|
|
|
}
|
|
|
|
|
|
|
|
uint64_t
|
2008-12-03 23:09:06 +03:00
|
|
|
dsl_dir_create_sync(dsl_pool_t *dp, dsl_dir_t *pds, const char *name,
|
|
|
|
dmu_tx_t *tx)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
2008-12-03 23:09:06 +03:00
|
|
|
objset_t *mos = dp->dp_meta_objset;
|
2008-11-20 23:01:55 +03:00
|
|
|
uint64_t ddobj;
|
2010-05-29 00:45:14 +04:00
|
|
|
dsl_dir_phys_t *ddphys;
|
2008-11-20 23:01:55 +03:00
|
|
|
dmu_buf_t *dbuf;
|
|
|
|
|
|
|
|
ddobj = dmu_object_alloc(mos, DMU_OT_DSL_DIR, 0,
|
|
|
|
DMU_OT_DSL_DIR, sizeof (dsl_dir_phys_t), tx);
|
2008-12-03 23:09:06 +03:00
|
|
|
if (pds) {
|
|
|
|
VERIFY(0 == zap_add(mos, pds->dd_phys->dd_child_dir_zapobj,
|
|
|
|
name, sizeof (uint64_t), 1, &ddobj, tx));
|
|
|
|
} else {
|
|
|
|
/* it's the root dir */
|
|
|
|
VERIFY(0 == zap_add(mos, DMU_POOL_DIRECTORY_OBJECT,
|
|
|
|
DMU_POOL_ROOT_DATASET, sizeof (uint64_t), 1, &ddobj, tx));
|
|
|
|
}
|
2008-11-20 23:01:55 +03:00
|
|
|
VERIFY(0 == dmu_bonus_hold(mos, ddobj, FTAG, &dbuf));
|
|
|
|
dmu_buf_will_dirty(dbuf, tx);
|
2010-05-29 00:45:14 +04:00
|
|
|
ddphys = dbuf->db_data;
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2010-05-29 00:45:14 +04:00
|
|
|
ddphys->dd_creation_time = gethrestime_sec();
|
2008-12-03 23:09:06 +03:00
|
|
|
if (pds)
|
2010-05-29 00:45:14 +04:00
|
|
|
ddphys->dd_parent_obj = pds->dd_object;
|
|
|
|
ddphys->dd_props_zapobj = zap_create(mos,
|
2008-11-20 23:01:55 +03:00
|
|
|
DMU_OT_DSL_PROPS, DMU_OT_NONE, 0, tx);
|
2010-05-29 00:45:14 +04:00
|
|
|
ddphys->dd_child_dir_zapobj = zap_create(mos,
|
2008-11-20 23:01:55 +03:00
|
|
|
DMU_OT_DSL_DIR_CHILD_MAP, DMU_OT_NONE, 0, tx);
|
2008-12-03 23:09:06 +03:00
|
|
|
if (spa_version(dp->dp_spa) >= SPA_VERSION_USED_BREAKDOWN)
|
2010-05-29 00:45:14 +04:00
|
|
|
ddphys->dd_flags |= DD_FLAG_USED_BREAKDOWN;
|
2008-11-20 23:01:55 +03:00
|
|
|
dmu_buf_rele(dbuf, FTAG);
|
|
|
|
|
|
|
|
return (ddobj);
|
|
|
|
}
|
|
|
|
|
2008-12-03 23:09:06 +03:00
|
|
|
boolean_t
|
|
|
|
dsl_dir_is_clone(dsl_dir_t *dd)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
2008-12-03 23:09:06 +03:00
|
|
|
return (dd->dd_phys->dd_origin_obj &&
|
|
|
|
(dd->dd_pool->dp_origin_snap == NULL ||
|
|
|
|
dd->dd_phys->dd_origin_obj !=
|
|
|
|
dd->dd_pool->dp_origin_snap->ds_object));
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
dsl_dir_stats(dsl_dir_t *dd, nvlist_t *nv)
|
|
|
|
{
|
|
|
|
mutex_enter(&dd->dd_lock);
|
2008-12-03 23:09:06 +03:00
|
|
|
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_USED,
|
|
|
|
dd->dd_phys->dd_used_bytes);
|
2008-11-20 23:01:55 +03:00
|
|
|
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_QUOTA, dd->dd_phys->dd_quota);
|
|
|
|
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_RESERVATION,
|
|
|
|
dd->dd_phys->dd_reserved);
|
|
|
|
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_COMPRESSRATIO,
|
|
|
|
dd->dd_phys->dd_compressed_bytes == 0 ? 100 :
|
|
|
|
(dd->dd_phys->dd_uncompressed_bytes * 100 /
|
|
|
|
dd->dd_phys->dd_compressed_bytes));
|
2013-02-22 13:23:09 +04:00
|
|
|
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_LOGICALUSED,
|
|
|
|
dd->dd_phys->dd_uncompressed_bytes);
|
2008-12-03 23:09:06 +03:00
|
|
|
if (dd->dd_phys->dd_flags & DD_FLAG_USED_BREAKDOWN) {
|
|
|
|
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_USEDSNAP,
|
|
|
|
dd->dd_phys->dd_used_breakdown[DD_USED_SNAP]);
|
|
|
|
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_USEDDS,
|
|
|
|
dd->dd_phys->dd_used_breakdown[DD_USED_HEAD]);
|
|
|
|
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_USEDREFRESERV,
|
|
|
|
dd->dd_phys->dd_used_breakdown[DD_USED_REFRSRV]);
|
|
|
|
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_USEDCHILD,
|
|
|
|
dd->dd_phys->dd_used_breakdown[DD_USED_CHILD] +
|
|
|
|
dd->dd_phys->dd_used_breakdown[DD_USED_CHILD_RSRV]);
|
|
|
|
}
|
2008-11-20 23:01:55 +03:00
|
|
|
mutex_exit(&dd->dd_lock);
|
|
|
|
|
2008-12-03 23:09:06 +03:00
|
|
|
if (dsl_dir_is_clone(dd)) {
|
2008-11-20 23:01:55 +03:00
|
|
|
dsl_dataset_t *ds;
|
|
|
|
char buf[MAXNAMELEN];
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
VERIFY0(dsl_dataset_hold_obj(dd->dd_pool,
|
2008-12-03 23:09:06 +03:00
|
|
|
dd->dd_phys->dd_origin_obj, FTAG, &ds));
|
2008-11-20 23:01:55 +03:00
|
|
|
dsl_dataset_name(ds, buf);
|
2008-12-03 23:09:06 +03:00
|
|
|
dsl_dataset_rele(ds, FTAG);
|
2008-11-20 23:01:55 +03:00
|
|
|
dsl_prop_nvlist_add_string(nv, ZFS_PROP_ORIGIN, buf);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
dsl_dir_dirty(dsl_dir_t *dd, dmu_tx_t *tx)
|
|
|
|
{
|
|
|
|
dsl_pool_t *dp = dd->dd_pool;
|
|
|
|
|
|
|
|
ASSERT(dd->dd_phys);
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
if (txg_list_add(&dp->dp_dirty_dirs, dd, tx->tx_txg)) {
|
2008-11-20 23:01:55 +03:00
|
|
|
/* up the hold count until we can be written out */
|
|
|
|
dmu_buf_add_ref(dd->dd_dbuf, dd);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static int64_t
|
|
|
|
parent_delta(dsl_dir_t *dd, uint64_t used, int64_t delta)
|
|
|
|
{
|
|
|
|
uint64_t old_accounted = MAX(used, dd->dd_phys->dd_reserved);
|
|
|
|
uint64_t new_accounted = MAX(used + delta, dd->dd_phys->dd_reserved);
|
|
|
|
return (new_accounted - old_accounted);
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
dsl_dir_sync(dsl_dir_t *dd, dmu_tx_t *tx)
|
|
|
|
{
|
|
|
|
ASSERT(dmu_tx_is_syncing(tx));
|
|
|
|
|
|
|
|
mutex_enter(&dd->dd_lock);
|
2013-05-11 01:17:03 +04:00
|
|
|
ASSERT0(dd->dd_tempreserved[tx->tx_txg&TXG_MASK]);
|
2008-11-20 23:01:55 +03:00
|
|
|
dprintf_dd(dd, "txg=%llu towrite=%lluK\n", tx->tx_txg,
|
|
|
|
dd->dd_space_towrite[tx->tx_txg&TXG_MASK] / 1024);
|
|
|
|
dd->dd_space_towrite[tx->tx_txg&TXG_MASK] = 0;
|
|
|
|
mutex_exit(&dd->dd_lock);
|
|
|
|
|
|
|
|
/* release the hold from dsl_dir_dirty */
|
|
|
|
dmu_buf_rele(dd->dd_dbuf, dd);
|
|
|
|
}
|
|
|
|
|
|
|
|
static uint64_t
|
|
|
|
dsl_dir_space_towrite(dsl_dir_t *dd)
|
|
|
|
{
|
|
|
|
uint64_t space = 0;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
ASSERT(MUTEX_HELD(&dd->dd_lock));
|
|
|
|
|
|
|
|
for (i = 0; i < TXG_SIZE; i++) {
|
|
|
|
space += dd->dd_space_towrite[i&TXG_MASK];
|
|
|
|
ASSERT3U(dd->dd_space_towrite[i&TXG_MASK], >=, 0);
|
|
|
|
}
|
|
|
|
return (space);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* How much space would dd have available if ancestor had delta applied
|
|
|
|
* to it? If ondiskonly is set, we're only interested in what's
|
|
|
|
* on-disk, not estimated pending changes.
|
|
|
|
*/
|
|
|
|
uint64_t
|
|
|
|
dsl_dir_space_available(dsl_dir_t *dd,
|
|
|
|
dsl_dir_t *ancestor, int64_t delta, int ondiskonly)
|
|
|
|
{
|
|
|
|
uint64_t parentspace, myspace, quota, used;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If there are no restrictions otherwise, assume we have
|
|
|
|
* unlimited space available.
|
|
|
|
*/
|
|
|
|
quota = UINT64_MAX;
|
|
|
|
parentspace = UINT64_MAX;
|
|
|
|
|
|
|
|
if (dd->dd_parent != NULL) {
|
|
|
|
parentspace = dsl_dir_space_available(dd->dd_parent,
|
|
|
|
ancestor, delta, ondiskonly);
|
|
|
|
}
|
|
|
|
|
|
|
|
mutex_enter(&dd->dd_lock);
|
|
|
|
if (dd->dd_phys->dd_quota != 0)
|
|
|
|
quota = dd->dd_phys->dd_quota;
|
2008-12-03 23:09:06 +03:00
|
|
|
used = dd->dd_phys->dd_used_bytes;
|
2008-11-20 23:01:55 +03:00
|
|
|
if (!ondiskonly)
|
|
|
|
used += dsl_dir_space_towrite(dd);
|
|
|
|
|
|
|
|
if (dd->dd_parent == NULL) {
|
|
|
|
uint64_t poolsize = dsl_pool_adjustedsize(dd->dd_pool, FALSE);
|
|
|
|
quota = MIN(quota, poolsize);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (dd->dd_phys->dd_reserved > used && parentspace != UINT64_MAX) {
|
|
|
|
/*
|
|
|
|
* We have some space reserved, in addition to what our
|
|
|
|
* parent gave us.
|
|
|
|
*/
|
|
|
|
parentspace += dd->dd_phys->dd_reserved - used;
|
|
|
|
}
|
|
|
|
|
2008-12-03 23:09:06 +03:00
|
|
|
if (dd == ancestor) {
|
|
|
|
ASSERT(delta <= 0);
|
|
|
|
ASSERT(used >= -delta);
|
|
|
|
used += delta;
|
|
|
|
if (parentspace != UINT64_MAX)
|
|
|
|
parentspace -= delta;
|
|
|
|
}
|
|
|
|
|
2008-11-20 23:01:55 +03:00
|
|
|
if (used > quota) {
|
|
|
|
/* over quota */
|
|
|
|
myspace = 0;
|
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* the lesser of the space provided by our parent and
|
|
|
|
* the space left in our quota
|
|
|
|
*/
|
|
|
|
myspace = MIN(parentspace, quota - used);
|
|
|
|
}
|
|
|
|
|
|
|
|
mutex_exit(&dd->dd_lock);
|
|
|
|
|
|
|
|
return (myspace);
|
|
|
|
}
|
|
|
|
|
|
|
|
struct tempreserve {
|
|
|
|
list_node_t tr_node;
|
|
|
|
dsl_dir_t *tr_ds;
|
|
|
|
uint64_t tr_size;
|
|
|
|
};
|
|
|
|
|
|
|
|
static int
|
|
|
|
dsl_dir_tempreserve_impl(dsl_dir_t *dd, uint64_t asize, boolean_t netfree,
|
|
|
|
boolean_t ignorequota, boolean_t checkrefquota, list_t *tr_list,
|
|
|
|
dmu_tx_t *tx, boolean_t first)
|
|
|
|
{
|
|
|
|
uint64_t txg = tx->tx_txg;
|
|
|
|
uint64_t est_inflight, used_on_disk, quota, parent_rsrv;
|
2010-05-29 00:45:14 +04:00
|
|
|
uint64_t deferred = 0;
|
2008-11-20 23:01:55 +03:00
|
|
|
struct tempreserve *tr;
|
2010-05-29 00:45:14 +04:00
|
|
|
int retval = EDQUOT;
|
2008-11-20 23:01:55 +03:00
|
|
|
int txgidx = txg & TXG_MASK;
|
|
|
|
int i;
|
|
|
|
uint64_t ref_rsrv = 0;
|
|
|
|
|
|
|
|
ASSERT3U(txg, !=, 0);
|
|
|
|
ASSERT3S(asize, >, 0);
|
|
|
|
|
|
|
|
mutex_enter(&dd->dd_lock);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Check against the dsl_dir's quota. We don't add in the delta
|
|
|
|
* when checking for over-quota because they get one free hit.
|
|
|
|
*/
|
|
|
|
est_inflight = dsl_dir_space_towrite(dd);
|
|
|
|
for (i = 0; i < TXG_SIZE; i++)
|
|
|
|
est_inflight += dd->dd_tempreserved[i];
|
2008-12-03 23:09:06 +03:00
|
|
|
used_on_disk = dd->dd_phys->dd_used_bytes;
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
/*
|
|
|
|
* On the first iteration, fetch the dataset's used-on-disk and
|
|
|
|
* refreservation values. Also, if checkrefquota is set, test if
|
|
|
|
* allocating this space would exceed the dataset's refquota.
|
|
|
|
*/
|
|
|
|
if (first && tx->tx_objset) {
|
|
|
|
int error;
|
2010-05-29 00:45:14 +04:00
|
|
|
dsl_dataset_t *ds = tx->tx_objset->os_dsl_dataset;
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
error = dsl_dataset_check_quota(ds, checkrefquota,
|
|
|
|
asize, est_inflight, &used_on_disk, &ref_rsrv);
|
|
|
|
if (error) {
|
|
|
|
mutex_exit(&dd->dd_lock);
|
2014-02-28 04:11:11 +04:00
|
|
|
DMU_TX_STAT_BUMP(dmu_tx_quota);
|
2008-11-20 23:01:55 +03:00
|
|
|
return (error);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If this transaction will result in a net free of space,
|
|
|
|
* we want to let it through.
|
|
|
|
*/
|
|
|
|
if (ignorequota || netfree || dd->dd_phys->dd_quota == 0)
|
|
|
|
quota = UINT64_MAX;
|
|
|
|
else
|
|
|
|
quota = dd->dd_phys->dd_quota;
|
|
|
|
|
|
|
|
/*
|
2010-05-29 00:45:14 +04:00
|
|
|
* Adjust the quota against the actual pool size at the root
|
|
|
|
* minus any outstanding deferred frees.
|
2008-11-20 23:01:55 +03:00
|
|
|
* To ensure that it's possible to remove files from a full
|
|
|
|
* pool without inducing transient overcommits, we throttle
|
|
|
|
* netfree transactions against a quota that is slightly larger,
|
|
|
|
* but still within the pool's allocation slop. In cases where
|
|
|
|
* we're very close to full, this will allow a steady trickle of
|
|
|
|
* removes to get through.
|
|
|
|
*/
|
|
|
|
if (dd->dd_parent == NULL) {
|
2010-05-29 00:45:14 +04:00
|
|
|
spa_t *spa = dd->dd_pool->dp_spa;
|
2008-11-20 23:01:55 +03:00
|
|
|
uint64_t poolsize = dsl_pool_adjustedsize(dd->dd_pool, netfree);
|
2010-05-29 00:45:14 +04:00
|
|
|
deferred = metaslab_class_get_deferred(spa_normal_class(spa));
|
|
|
|
if (poolsize - deferred < quota) {
|
|
|
|
quota = poolsize - deferred;
|
|
|
|
retval = ENOSPC;
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If they are requesting more space, and our current estimate
|
|
|
|
* is over quota, they get to try again unless the actual
|
|
|
|
* on-disk is over quota and there are no pending changes (which
|
|
|
|
* may free up space for us).
|
|
|
|
*/
|
2010-05-29 00:45:14 +04:00
|
|
|
if (used_on_disk + est_inflight >= quota) {
|
|
|
|
if (est_inflight > 0 || used_on_disk < quota ||
|
|
|
|
(retval == ENOSPC && used_on_disk < quota + deferred))
|
|
|
|
retval = ERESTART;
|
2008-11-20 23:01:55 +03:00
|
|
|
dprintf_dd(dd, "failing: used=%lluK inflight = %lluK "
|
|
|
|
"quota=%lluK tr=%lluK err=%d\n",
|
|
|
|
used_on_disk>>10, est_inflight>>10,
|
2010-05-29 00:45:14 +04:00
|
|
|
quota>>10, asize>>10, retval);
|
2008-11-20 23:01:55 +03:00
|
|
|
mutex_exit(&dd->dd_lock);
|
2014-02-28 04:11:11 +04:00
|
|
|
DMU_TX_STAT_BUMP(dmu_tx_quota);
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(retval));
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
/* We need to up our estimated delta before dropping dd_lock */
|
|
|
|
dd->dd_tempreserved[txgidx] += asize;
|
|
|
|
|
|
|
|
parent_rsrv = parent_delta(dd, used_on_disk + est_inflight,
|
|
|
|
asize - ref_rsrv);
|
|
|
|
mutex_exit(&dd->dd_lock);
|
|
|
|
|
2012-05-07 21:49:51 +04:00
|
|
|
tr = kmem_zalloc(sizeof (struct tempreserve), KM_PUSHPAGE);
|
2008-11-20 23:01:55 +03:00
|
|
|
tr->tr_ds = dd;
|
|
|
|
tr->tr_size = asize;
|
|
|
|
list_insert_tail(tr_list, tr);
|
|
|
|
|
|
|
|
/* see if it's OK with our parent */
|
|
|
|
if (dd->dd_parent && parent_rsrv) {
|
|
|
|
boolean_t ismos = (dd->dd_phys->dd_head_dataset_obj == 0);
|
|
|
|
|
|
|
|
return (dsl_dir_tempreserve_impl(dd->dd_parent,
|
|
|
|
parent_rsrv, netfree, ismos, TRUE, tr_list, tx, FALSE));
|
|
|
|
} else {
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Reserve space in this dsl_dir, to be used in this tx's txg.
|
|
|
|
* After the space has been dirtied (and dsl_dir_willuse_space()
|
|
|
|
* has been called), the reservation should be canceled, using
|
|
|
|
* dsl_dir_tempreserve_clear().
|
|
|
|
*/
|
|
|
|
int
|
|
|
|
dsl_dir_tempreserve_space(dsl_dir_t *dd, uint64_t lsize, uint64_t asize,
|
|
|
|
uint64_t fsize, uint64_t usize, void **tr_cookiep, dmu_tx_t *tx)
|
|
|
|
{
|
|
|
|
int err;
|
|
|
|
list_t *tr_list;
|
|
|
|
|
|
|
|
if (asize == 0) {
|
|
|
|
*tr_cookiep = NULL;
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
2012-05-07 21:49:51 +04:00
|
|
|
tr_list = kmem_alloc(sizeof (list_t), KM_PUSHPAGE);
|
2008-11-20 23:01:55 +03:00
|
|
|
list_create(tr_list, sizeof (struct tempreserve),
|
|
|
|
offsetof(struct tempreserve, tr_node));
|
|
|
|
ASSERT3S(asize, >, 0);
|
|
|
|
ASSERT3S(fsize, >=, 0);
|
|
|
|
|
|
|
|
err = arc_tempreserve_space(lsize, tx->tx_txg);
|
|
|
|
if (err == 0) {
|
|
|
|
struct tempreserve *tr;
|
|
|
|
|
2012-05-07 21:49:51 +04:00
|
|
|
tr = kmem_zalloc(sizeof (struct tempreserve), KM_PUSHPAGE);
|
2008-11-20 23:01:55 +03:00
|
|
|
tr->tr_size = lsize;
|
|
|
|
list_insert_tail(tr_list, tr);
|
|
|
|
} else {
|
|
|
|
if (err == EAGAIN) {
|
Illumos #4045 write throttle & i/o scheduler performance work
4045 zfs write throttle & i/o scheduler performance work
1. The ZFS i/o scheduler (vdev_queue.c) now divides i/os into 5 classes: sync
read, sync write, async read, async write, and scrub/resilver. The scheduler
issues a number of concurrent i/os from each class to the device. Once a class
has been selected, an i/o is selected from this class using either an elevator
algorithem (async, scrub classes) or FIFO (sync classes). The number of
concurrent async write i/os is tuned dynamically based on i/o load, to achieve
good sync i/o latency when there is not a high load of writes, and good write
throughput when there is. See the block comment in vdev_queue.c (reproduced
below) for more details.
2. The write throttle (dsl_pool_tempreserve_space() and
txg_constrain_throughput()) is rewritten to produce much more consistent delays
when under constant load. The new write throttle is based on the amount of
dirty data, rather than guesses about future performance of the system. When
there is a lot of dirty data, each transaction (e.g. write() syscall) will be
delayed by the same small amount. This eliminates the "brick wall of wait"
that the old write throttle could hit, causing all transactions to wait several
seconds until the next txg opens. One of the keys to the new write throttle is
decrementing the amount of dirty data as i/o completes, rather than at the end
of spa_sync(). Note that the write throttle is only applied once the i/o
scheduler is issuing the maximum number of outstanding async writes. See the
block comments in dsl_pool.c and above dmu_tx_delay() (reproduced below) for
more details.
This diff has several other effects, including:
* the commonly-tuned global variable zfs_vdev_max_pending has been removed;
use per-class zfs_vdev_*_max_active values or zfs_vdev_max_active instead.
* the size of each txg (meaning the amount of dirty data written, and thus the
time it takes to write out) is now controlled differently. There is no longer
an explicit time goal; the primary determinant is amount of dirty data.
Systems that are under light or medium load will now often see that a txg is
always syncing, but the impact to performance (e.g. read latency) is minimal.
Tune zfs_dirty_data_max and zfs_dirty_data_sync to control this.
* zio_taskq_batch_pct = 75 -- Only use 75% of all CPUs for compression,
checksum, etc. This improves latency by not allowing these CPU-intensive tasks
to consume all CPU (on machines with at least 4 CPU's; the percentage is
rounded up).
--matt
APPENDIX: problems with the current i/o scheduler
The current ZFS i/o scheduler (vdev_queue.c) is deadline based. The problem
with this is that if there are always i/os pending, then certain classes of
i/os can see very long delays.
For example, if there are always synchronous reads outstanding, then no async
writes will be serviced until they become "past due". One symptom of this
situation is that each pass of the txg sync takes at least several seconds
(typically 3 seconds).
If many i/os become "past due" (their deadline is in the past), then we must
service all of these overdue i/os before any new i/os. This happens when we
enqueue a batch of async writes for the txg sync, with deadlines 2.5 seconds in
the future. If we can't complete all the i/os in 2.5 seconds (e.g. because
there were always reads pending), then these i/os will become past due. Now we
must service all the "async" writes (which could be hundreds of megabytes)
before we service any reads, introducing considerable latency to synchronous
i/os (reads or ZIL writes).
Notes on porting to ZFS on Linux:
- zio_t gained new members io_physdone and io_phys_children. Because
object caches in the Linux port call the constructor only once at
allocation time, objects may contain residual data when retrieved
from the cache. Therefore zio_create() was updated to zero out the two
new fields.
- vdev_mirror_pending() relied on the depth of the per-vdev pending queue
(vq->vq_pending_tree) to select the least-busy leaf vdev to read from.
This tree has been replaced by vq->vq_active_tree which is now used
for the same purpose.
- vdev_queue_init() used the value of zfs_vdev_max_pending to determine
the number of vdev I/O buffers to pre-allocate. That global no longer
exists, so we instead use the sum of the *_max_active values for each of
the five I/O classes described above.
- The Illumos implementation of dmu_tx_delay() delays a transaction by
sleeping in condition variable embedded in the thread
(curthread->t_delay_cv). We do not have an equivalent CV to use in
Linux, so this change replaced the delay logic with a wrapper called
zfs_sleep_until(). This wrapper could be adopted upstream and in other
downstream ports to abstract away operating system-specific delay logic.
- These tunables are added as module parameters, and descriptions added
to the zfs-module-parameters.5 man page.
spa_asize_inflation
zfs_deadman_synctime_ms
zfs_vdev_max_active
zfs_vdev_async_write_active_min_dirty_percent
zfs_vdev_async_write_active_max_dirty_percent
zfs_vdev_async_read_max_active
zfs_vdev_async_read_min_active
zfs_vdev_async_write_max_active
zfs_vdev_async_write_min_active
zfs_vdev_scrub_max_active
zfs_vdev_scrub_min_active
zfs_vdev_sync_read_max_active
zfs_vdev_sync_read_min_active
zfs_vdev_sync_write_max_active
zfs_vdev_sync_write_min_active
zfs_dirty_data_max_percent
zfs_delay_min_dirty_percent
zfs_dirty_data_max_max_percent
zfs_dirty_data_max
zfs_dirty_data_max_max
zfs_dirty_data_sync
zfs_delay_scale
The latter four have type unsigned long, whereas they are uint64_t in
Illumos. This accommodates Linux's module_param() supported types, but
means they may overflow on 32-bit architectures.
The values zfs_dirty_data_max and zfs_dirty_data_max_max are the most
likely to overflow on 32-bit systems, since they express physical RAM
sizes in bytes. In fact, Illumos initializes zfs_dirty_data_max_max to
2^32 which does overflow. To resolve that, this port instead initializes
it in arc_init() to 25% of physical RAM, and adds the tunable
zfs_dirty_data_max_max_percent to override that percentage. While this
solution doesn't completely avoid the overflow issue, it should be a
reasonable default for most systems, and the minority of affected
systems can work around the issue by overriding the defaults.
- Fixed reversed logic in comment above zfs_delay_scale declaration.
- Clarified comments in vdev_queue.c regarding when per-queue minimums take
effect.
- Replaced dmu_tx_write_limit in the dmu_tx kstat file
with dmu_tx_dirty_delay and dmu_tx_dirty_over_max. The first counts
how many times a transaction has been delayed because the pool dirty
data has exceeded zfs_delay_min_dirty_percent. The latter counts how
many times the pool dirty data has exceeded zfs_dirty_data_max (which
we expect to never happen).
- The original patch would have regressed the bug fixed in
zfsonlinux/zfs@c418410, which prevented users from setting the
zfs_vdev_aggregation_limit tuning larger than SPA_MAXBLOCKSIZE.
A similar fix is added to vdev_queue_aggregate().
- In vdev_queue_io_to_issue(), dynamically allocate 'zio_t search' on the
heap instead of the stack. In Linux we can't afford such large
structures on the stack.
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Adam Leventhal <ahl@delphix.com>
Reviewed by: Christopher Siden <christopher.siden@delphix.com>
Reviewed by: Ned Bass <bass6@llnl.gov>
Reviewed by: Brendan Gregg <brendan.gregg@joyent.com>
Approved by: Robert Mustacchi <rm@joyent.com>
References:
http://www.illumos.org/issues/4045
illumos/illumos-gate@69962b5647e4a8b9b14998733b765925381b727e
Ported-by: Ned Bass <bass6@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #1913
2013-08-29 07:01:20 +04:00
|
|
|
/*
|
|
|
|
* If arc_memory_throttle() detected that pageout
|
|
|
|
* is running and we are low on memory, we delay new
|
|
|
|
* non-pageout transactions to give pageout an
|
|
|
|
* advantage.
|
|
|
|
*
|
|
|
|
* It is unfortunate to be delaying while the caller's
|
|
|
|
* locks are held.
|
|
|
|
*/
|
2013-08-29 03:05:48 +04:00
|
|
|
txg_delay(dd->dd_pool, tx->tx_txg,
|
|
|
|
MSEC2NSEC(10), MSEC2NSEC(10));
|
2013-03-08 22:41:28 +04:00
|
|
|
err = SET_ERROR(ERESTART);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (err == 0) {
|
|
|
|
err = dsl_dir_tempreserve_impl(dd, asize, fsize >= asize,
|
|
|
|
FALSE, asize > usize, tr_list, tx, TRUE);
|
|
|
|
}
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
if (err != 0)
|
2008-11-20 23:01:55 +03:00
|
|
|
dsl_dir_tempreserve_clear(tr_list, tx);
|
|
|
|
else
|
|
|
|
*tr_cookiep = tr_list;
|
|
|
|
|
|
|
|
return (err);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Clear a temporary reservation that we previously made with
|
|
|
|
* dsl_dir_tempreserve_space().
|
|
|
|
*/
|
|
|
|
void
|
|
|
|
dsl_dir_tempreserve_clear(void *tr_cookie, dmu_tx_t *tx)
|
|
|
|
{
|
|
|
|
int txgidx = tx->tx_txg & TXG_MASK;
|
|
|
|
list_t *tr_list = tr_cookie;
|
|
|
|
struct tempreserve *tr;
|
|
|
|
|
|
|
|
ASSERT3U(tx->tx_txg, !=, 0);
|
|
|
|
|
|
|
|
if (tr_cookie == NULL)
|
|
|
|
return;
|
|
|
|
|
Illumos #4045 write throttle & i/o scheduler performance work
4045 zfs write throttle & i/o scheduler performance work
1. The ZFS i/o scheduler (vdev_queue.c) now divides i/os into 5 classes: sync
read, sync write, async read, async write, and scrub/resilver. The scheduler
issues a number of concurrent i/os from each class to the device. Once a class
has been selected, an i/o is selected from this class using either an elevator
algorithem (async, scrub classes) or FIFO (sync classes). The number of
concurrent async write i/os is tuned dynamically based on i/o load, to achieve
good sync i/o latency when there is not a high load of writes, and good write
throughput when there is. See the block comment in vdev_queue.c (reproduced
below) for more details.
2. The write throttle (dsl_pool_tempreserve_space() and
txg_constrain_throughput()) is rewritten to produce much more consistent delays
when under constant load. The new write throttle is based on the amount of
dirty data, rather than guesses about future performance of the system. When
there is a lot of dirty data, each transaction (e.g. write() syscall) will be
delayed by the same small amount. This eliminates the "brick wall of wait"
that the old write throttle could hit, causing all transactions to wait several
seconds until the next txg opens. One of the keys to the new write throttle is
decrementing the amount of dirty data as i/o completes, rather than at the end
of spa_sync(). Note that the write throttle is only applied once the i/o
scheduler is issuing the maximum number of outstanding async writes. See the
block comments in dsl_pool.c and above dmu_tx_delay() (reproduced below) for
more details.
This diff has several other effects, including:
* the commonly-tuned global variable zfs_vdev_max_pending has been removed;
use per-class zfs_vdev_*_max_active values or zfs_vdev_max_active instead.
* the size of each txg (meaning the amount of dirty data written, and thus the
time it takes to write out) is now controlled differently. There is no longer
an explicit time goal; the primary determinant is amount of dirty data.
Systems that are under light or medium load will now often see that a txg is
always syncing, but the impact to performance (e.g. read latency) is minimal.
Tune zfs_dirty_data_max and zfs_dirty_data_sync to control this.
* zio_taskq_batch_pct = 75 -- Only use 75% of all CPUs for compression,
checksum, etc. This improves latency by not allowing these CPU-intensive tasks
to consume all CPU (on machines with at least 4 CPU's; the percentage is
rounded up).
--matt
APPENDIX: problems with the current i/o scheduler
The current ZFS i/o scheduler (vdev_queue.c) is deadline based. The problem
with this is that if there are always i/os pending, then certain classes of
i/os can see very long delays.
For example, if there are always synchronous reads outstanding, then no async
writes will be serviced until they become "past due". One symptom of this
situation is that each pass of the txg sync takes at least several seconds
(typically 3 seconds).
If many i/os become "past due" (their deadline is in the past), then we must
service all of these overdue i/os before any new i/os. This happens when we
enqueue a batch of async writes for the txg sync, with deadlines 2.5 seconds in
the future. If we can't complete all the i/os in 2.5 seconds (e.g. because
there were always reads pending), then these i/os will become past due. Now we
must service all the "async" writes (which could be hundreds of megabytes)
before we service any reads, introducing considerable latency to synchronous
i/os (reads or ZIL writes).
Notes on porting to ZFS on Linux:
- zio_t gained new members io_physdone and io_phys_children. Because
object caches in the Linux port call the constructor only once at
allocation time, objects may contain residual data when retrieved
from the cache. Therefore zio_create() was updated to zero out the two
new fields.
- vdev_mirror_pending() relied on the depth of the per-vdev pending queue
(vq->vq_pending_tree) to select the least-busy leaf vdev to read from.
This tree has been replaced by vq->vq_active_tree which is now used
for the same purpose.
- vdev_queue_init() used the value of zfs_vdev_max_pending to determine
the number of vdev I/O buffers to pre-allocate. That global no longer
exists, so we instead use the sum of the *_max_active values for each of
the five I/O classes described above.
- The Illumos implementation of dmu_tx_delay() delays a transaction by
sleeping in condition variable embedded in the thread
(curthread->t_delay_cv). We do not have an equivalent CV to use in
Linux, so this change replaced the delay logic with a wrapper called
zfs_sleep_until(). This wrapper could be adopted upstream and in other
downstream ports to abstract away operating system-specific delay logic.
- These tunables are added as module parameters, and descriptions added
to the zfs-module-parameters.5 man page.
spa_asize_inflation
zfs_deadman_synctime_ms
zfs_vdev_max_active
zfs_vdev_async_write_active_min_dirty_percent
zfs_vdev_async_write_active_max_dirty_percent
zfs_vdev_async_read_max_active
zfs_vdev_async_read_min_active
zfs_vdev_async_write_max_active
zfs_vdev_async_write_min_active
zfs_vdev_scrub_max_active
zfs_vdev_scrub_min_active
zfs_vdev_sync_read_max_active
zfs_vdev_sync_read_min_active
zfs_vdev_sync_write_max_active
zfs_vdev_sync_write_min_active
zfs_dirty_data_max_percent
zfs_delay_min_dirty_percent
zfs_dirty_data_max_max_percent
zfs_dirty_data_max
zfs_dirty_data_max_max
zfs_dirty_data_sync
zfs_delay_scale
The latter four have type unsigned long, whereas they are uint64_t in
Illumos. This accommodates Linux's module_param() supported types, but
means they may overflow on 32-bit architectures.
The values zfs_dirty_data_max and zfs_dirty_data_max_max are the most
likely to overflow on 32-bit systems, since they express physical RAM
sizes in bytes. In fact, Illumos initializes zfs_dirty_data_max_max to
2^32 which does overflow. To resolve that, this port instead initializes
it in arc_init() to 25% of physical RAM, and adds the tunable
zfs_dirty_data_max_max_percent to override that percentage. While this
solution doesn't completely avoid the overflow issue, it should be a
reasonable default for most systems, and the minority of affected
systems can work around the issue by overriding the defaults.
- Fixed reversed logic in comment above zfs_delay_scale declaration.
- Clarified comments in vdev_queue.c regarding when per-queue minimums take
effect.
- Replaced dmu_tx_write_limit in the dmu_tx kstat file
with dmu_tx_dirty_delay and dmu_tx_dirty_over_max. The first counts
how many times a transaction has been delayed because the pool dirty
data has exceeded zfs_delay_min_dirty_percent. The latter counts how
many times the pool dirty data has exceeded zfs_dirty_data_max (which
we expect to never happen).
- The original patch would have regressed the bug fixed in
zfsonlinux/zfs@c418410, which prevented users from setting the
zfs_vdev_aggregation_limit tuning larger than SPA_MAXBLOCKSIZE.
A similar fix is added to vdev_queue_aggregate().
- In vdev_queue_io_to_issue(), dynamically allocate 'zio_t search' on the
heap instead of the stack. In Linux we can't afford such large
structures on the stack.
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Adam Leventhal <ahl@delphix.com>
Reviewed by: Christopher Siden <christopher.siden@delphix.com>
Reviewed by: Ned Bass <bass6@llnl.gov>
Reviewed by: Brendan Gregg <brendan.gregg@joyent.com>
Approved by: Robert Mustacchi <rm@joyent.com>
References:
http://www.illumos.org/issues/4045
illumos/illumos-gate@69962b5647e4a8b9b14998733b765925381b727e
Ported-by: Ned Bass <bass6@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #1913
2013-08-29 07:01:20 +04:00
|
|
|
while ((tr = list_head(tr_list)) != NULL) {
|
|
|
|
if (tr->tr_ds) {
|
2008-11-20 23:01:55 +03:00
|
|
|
mutex_enter(&tr->tr_ds->dd_lock);
|
|
|
|
ASSERT3U(tr->tr_ds->dd_tempreserved[txgidx], >=,
|
|
|
|
tr->tr_size);
|
|
|
|
tr->tr_ds->dd_tempreserved[txgidx] -= tr->tr_size;
|
|
|
|
mutex_exit(&tr->tr_ds->dd_lock);
|
|
|
|
} else {
|
|
|
|
arc_tempreserve_clear(tr->tr_size);
|
|
|
|
}
|
|
|
|
list_remove(tr_list, tr);
|
|
|
|
kmem_free(tr, sizeof (struct tempreserve));
|
|
|
|
}
|
|
|
|
|
|
|
|
kmem_free(tr_list, sizeof (list_t));
|
|
|
|
}
|
|
|
|
|
Illumos #4045 write throttle & i/o scheduler performance work
4045 zfs write throttle & i/o scheduler performance work
1. The ZFS i/o scheduler (vdev_queue.c) now divides i/os into 5 classes: sync
read, sync write, async read, async write, and scrub/resilver. The scheduler
issues a number of concurrent i/os from each class to the device. Once a class
has been selected, an i/o is selected from this class using either an elevator
algorithem (async, scrub classes) or FIFO (sync classes). The number of
concurrent async write i/os is tuned dynamically based on i/o load, to achieve
good sync i/o latency when there is not a high load of writes, and good write
throughput when there is. See the block comment in vdev_queue.c (reproduced
below) for more details.
2. The write throttle (dsl_pool_tempreserve_space() and
txg_constrain_throughput()) is rewritten to produce much more consistent delays
when under constant load. The new write throttle is based on the amount of
dirty data, rather than guesses about future performance of the system. When
there is a lot of dirty data, each transaction (e.g. write() syscall) will be
delayed by the same small amount. This eliminates the "brick wall of wait"
that the old write throttle could hit, causing all transactions to wait several
seconds until the next txg opens. One of the keys to the new write throttle is
decrementing the amount of dirty data as i/o completes, rather than at the end
of spa_sync(). Note that the write throttle is only applied once the i/o
scheduler is issuing the maximum number of outstanding async writes. See the
block comments in dsl_pool.c and above dmu_tx_delay() (reproduced below) for
more details.
This diff has several other effects, including:
* the commonly-tuned global variable zfs_vdev_max_pending has been removed;
use per-class zfs_vdev_*_max_active values or zfs_vdev_max_active instead.
* the size of each txg (meaning the amount of dirty data written, and thus the
time it takes to write out) is now controlled differently. There is no longer
an explicit time goal; the primary determinant is amount of dirty data.
Systems that are under light or medium load will now often see that a txg is
always syncing, but the impact to performance (e.g. read latency) is minimal.
Tune zfs_dirty_data_max and zfs_dirty_data_sync to control this.
* zio_taskq_batch_pct = 75 -- Only use 75% of all CPUs for compression,
checksum, etc. This improves latency by not allowing these CPU-intensive tasks
to consume all CPU (on machines with at least 4 CPU's; the percentage is
rounded up).
--matt
APPENDIX: problems with the current i/o scheduler
The current ZFS i/o scheduler (vdev_queue.c) is deadline based. The problem
with this is that if there are always i/os pending, then certain classes of
i/os can see very long delays.
For example, if there are always synchronous reads outstanding, then no async
writes will be serviced until they become "past due". One symptom of this
situation is that each pass of the txg sync takes at least several seconds
(typically 3 seconds).
If many i/os become "past due" (their deadline is in the past), then we must
service all of these overdue i/os before any new i/os. This happens when we
enqueue a batch of async writes for the txg sync, with deadlines 2.5 seconds in
the future. If we can't complete all the i/os in 2.5 seconds (e.g. because
there were always reads pending), then these i/os will become past due. Now we
must service all the "async" writes (which could be hundreds of megabytes)
before we service any reads, introducing considerable latency to synchronous
i/os (reads or ZIL writes).
Notes on porting to ZFS on Linux:
- zio_t gained new members io_physdone and io_phys_children. Because
object caches in the Linux port call the constructor only once at
allocation time, objects may contain residual data when retrieved
from the cache. Therefore zio_create() was updated to zero out the two
new fields.
- vdev_mirror_pending() relied on the depth of the per-vdev pending queue
(vq->vq_pending_tree) to select the least-busy leaf vdev to read from.
This tree has been replaced by vq->vq_active_tree which is now used
for the same purpose.
- vdev_queue_init() used the value of zfs_vdev_max_pending to determine
the number of vdev I/O buffers to pre-allocate. That global no longer
exists, so we instead use the sum of the *_max_active values for each of
the five I/O classes described above.
- The Illumos implementation of dmu_tx_delay() delays a transaction by
sleeping in condition variable embedded in the thread
(curthread->t_delay_cv). We do not have an equivalent CV to use in
Linux, so this change replaced the delay logic with a wrapper called
zfs_sleep_until(). This wrapper could be adopted upstream and in other
downstream ports to abstract away operating system-specific delay logic.
- These tunables are added as module parameters, and descriptions added
to the zfs-module-parameters.5 man page.
spa_asize_inflation
zfs_deadman_synctime_ms
zfs_vdev_max_active
zfs_vdev_async_write_active_min_dirty_percent
zfs_vdev_async_write_active_max_dirty_percent
zfs_vdev_async_read_max_active
zfs_vdev_async_read_min_active
zfs_vdev_async_write_max_active
zfs_vdev_async_write_min_active
zfs_vdev_scrub_max_active
zfs_vdev_scrub_min_active
zfs_vdev_sync_read_max_active
zfs_vdev_sync_read_min_active
zfs_vdev_sync_write_max_active
zfs_vdev_sync_write_min_active
zfs_dirty_data_max_percent
zfs_delay_min_dirty_percent
zfs_dirty_data_max_max_percent
zfs_dirty_data_max
zfs_dirty_data_max_max
zfs_dirty_data_sync
zfs_delay_scale
The latter four have type unsigned long, whereas they are uint64_t in
Illumos. This accommodates Linux's module_param() supported types, but
means they may overflow on 32-bit architectures.
The values zfs_dirty_data_max and zfs_dirty_data_max_max are the most
likely to overflow on 32-bit systems, since they express physical RAM
sizes in bytes. In fact, Illumos initializes zfs_dirty_data_max_max to
2^32 which does overflow. To resolve that, this port instead initializes
it in arc_init() to 25% of physical RAM, and adds the tunable
zfs_dirty_data_max_max_percent to override that percentage. While this
solution doesn't completely avoid the overflow issue, it should be a
reasonable default for most systems, and the minority of affected
systems can work around the issue by overriding the defaults.
- Fixed reversed logic in comment above zfs_delay_scale declaration.
- Clarified comments in vdev_queue.c regarding when per-queue minimums take
effect.
- Replaced dmu_tx_write_limit in the dmu_tx kstat file
with dmu_tx_dirty_delay and dmu_tx_dirty_over_max. The first counts
how many times a transaction has been delayed because the pool dirty
data has exceeded zfs_delay_min_dirty_percent. The latter counts how
many times the pool dirty data has exceeded zfs_dirty_data_max (which
we expect to never happen).
- The original patch would have regressed the bug fixed in
zfsonlinux/zfs@c418410, which prevented users from setting the
zfs_vdev_aggregation_limit tuning larger than SPA_MAXBLOCKSIZE.
A similar fix is added to vdev_queue_aggregate().
- In vdev_queue_io_to_issue(), dynamically allocate 'zio_t search' on the
heap instead of the stack. In Linux we can't afford such large
structures on the stack.
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Adam Leventhal <ahl@delphix.com>
Reviewed by: Christopher Siden <christopher.siden@delphix.com>
Reviewed by: Ned Bass <bass6@llnl.gov>
Reviewed by: Brendan Gregg <brendan.gregg@joyent.com>
Approved by: Robert Mustacchi <rm@joyent.com>
References:
http://www.illumos.org/issues/4045
illumos/illumos-gate@69962b5647e4a8b9b14998733b765925381b727e
Ported-by: Ned Bass <bass6@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #1913
2013-08-29 07:01:20 +04:00
|
|
|
/*
|
|
|
|
* This should be called from open context when we think we're going to write
|
|
|
|
* or free space, for example when dirtying data. Be conservative; it's okay
|
|
|
|
* to write less space or free more, but we don't want to write more or free
|
|
|
|
* less than the amount specified.
|
2014-01-20 08:39:28 +04:00
|
|
|
*
|
|
|
|
* NOTE: The behavior of this function is identical to the Illumos / FreeBSD
|
|
|
|
* version however it has been adjusted to use an iterative rather then
|
|
|
|
* recursive algorithm to minimize stack usage.
|
Illumos #4045 write throttle & i/o scheduler performance work
4045 zfs write throttle & i/o scheduler performance work
1. The ZFS i/o scheduler (vdev_queue.c) now divides i/os into 5 classes: sync
read, sync write, async read, async write, and scrub/resilver. The scheduler
issues a number of concurrent i/os from each class to the device. Once a class
has been selected, an i/o is selected from this class using either an elevator
algorithem (async, scrub classes) or FIFO (sync classes). The number of
concurrent async write i/os is tuned dynamically based on i/o load, to achieve
good sync i/o latency when there is not a high load of writes, and good write
throughput when there is. See the block comment in vdev_queue.c (reproduced
below) for more details.
2. The write throttle (dsl_pool_tempreserve_space() and
txg_constrain_throughput()) is rewritten to produce much more consistent delays
when under constant load. The new write throttle is based on the amount of
dirty data, rather than guesses about future performance of the system. When
there is a lot of dirty data, each transaction (e.g. write() syscall) will be
delayed by the same small amount. This eliminates the "brick wall of wait"
that the old write throttle could hit, causing all transactions to wait several
seconds until the next txg opens. One of the keys to the new write throttle is
decrementing the amount of dirty data as i/o completes, rather than at the end
of spa_sync(). Note that the write throttle is only applied once the i/o
scheduler is issuing the maximum number of outstanding async writes. See the
block comments in dsl_pool.c and above dmu_tx_delay() (reproduced below) for
more details.
This diff has several other effects, including:
* the commonly-tuned global variable zfs_vdev_max_pending has been removed;
use per-class zfs_vdev_*_max_active values or zfs_vdev_max_active instead.
* the size of each txg (meaning the amount of dirty data written, and thus the
time it takes to write out) is now controlled differently. There is no longer
an explicit time goal; the primary determinant is amount of dirty data.
Systems that are under light or medium load will now often see that a txg is
always syncing, but the impact to performance (e.g. read latency) is minimal.
Tune zfs_dirty_data_max and zfs_dirty_data_sync to control this.
* zio_taskq_batch_pct = 75 -- Only use 75% of all CPUs for compression,
checksum, etc. This improves latency by not allowing these CPU-intensive tasks
to consume all CPU (on machines with at least 4 CPU's; the percentage is
rounded up).
--matt
APPENDIX: problems with the current i/o scheduler
The current ZFS i/o scheduler (vdev_queue.c) is deadline based. The problem
with this is that if there are always i/os pending, then certain classes of
i/os can see very long delays.
For example, if there are always synchronous reads outstanding, then no async
writes will be serviced until they become "past due". One symptom of this
situation is that each pass of the txg sync takes at least several seconds
(typically 3 seconds).
If many i/os become "past due" (their deadline is in the past), then we must
service all of these overdue i/os before any new i/os. This happens when we
enqueue a batch of async writes for the txg sync, with deadlines 2.5 seconds in
the future. If we can't complete all the i/os in 2.5 seconds (e.g. because
there were always reads pending), then these i/os will become past due. Now we
must service all the "async" writes (which could be hundreds of megabytes)
before we service any reads, introducing considerable latency to synchronous
i/os (reads or ZIL writes).
Notes on porting to ZFS on Linux:
- zio_t gained new members io_physdone and io_phys_children. Because
object caches in the Linux port call the constructor only once at
allocation time, objects may contain residual data when retrieved
from the cache. Therefore zio_create() was updated to zero out the two
new fields.
- vdev_mirror_pending() relied on the depth of the per-vdev pending queue
(vq->vq_pending_tree) to select the least-busy leaf vdev to read from.
This tree has been replaced by vq->vq_active_tree which is now used
for the same purpose.
- vdev_queue_init() used the value of zfs_vdev_max_pending to determine
the number of vdev I/O buffers to pre-allocate. That global no longer
exists, so we instead use the sum of the *_max_active values for each of
the five I/O classes described above.
- The Illumos implementation of dmu_tx_delay() delays a transaction by
sleeping in condition variable embedded in the thread
(curthread->t_delay_cv). We do not have an equivalent CV to use in
Linux, so this change replaced the delay logic with a wrapper called
zfs_sleep_until(). This wrapper could be adopted upstream and in other
downstream ports to abstract away operating system-specific delay logic.
- These tunables are added as module parameters, and descriptions added
to the zfs-module-parameters.5 man page.
spa_asize_inflation
zfs_deadman_synctime_ms
zfs_vdev_max_active
zfs_vdev_async_write_active_min_dirty_percent
zfs_vdev_async_write_active_max_dirty_percent
zfs_vdev_async_read_max_active
zfs_vdev_async_read_min_active
zfs_vdev_async_write_max_active
zfs_vdev_async_write_min_active
zfs_vdev_scrub_max_active
zfs_vdev_scrub_min_active
zfs_vdev_sync_read_max_active
zfs_vdev_sync_read_min_active
zfs_vdev_sync_write_max_active
zfs_vdev_sync_write_min_active
zfs_dirty_data_max_percent
zfs_delay_min_dirty_percent
zfs_dirty_data_max_max_percent
zfs_dirty_data_max
zfs_dirty_data_max_max
zfs_dirty_data_sync
zfs_delay_scale
The latter four have type unsigned long, whereas they are uint64_t in
Illumos. This accommodates Linux's module_param() supported types, but
means they may overflow on 32-bit architectures.
The values zfs_dirty_data_max and zfs_dirty_data_max_max are the most
likely to overflow on 32-bit systems, since they express physical RAM
sizes in bytes. In fact, Illumos initializes zfs_dirty_data_max_max to
2^32 which does overflow. To resolve that, this port instead initializes
it in arc_init() to 25% of physical RAM, and adds the tunable
zfs_dirty_data_max_max_percent to override that percentage. While this
solution doesn't completely avoid the overflow issue, it should be a
reasonable default for most systems, and the minority of affected
systems can work around the issue by overriding the defaults.
- Fixed reversed logic in comment above zfs_delay_scale declaration.
- Clarified comments in vdev_queue.c regarding when per-queue minimums take
effect.
- Replaced dmu_tx_write_limit in the dmu_tx kstat file
with dmu_tx_dirty_delay and dmu_tx_dirty_over_max. The first counts
how many times a transaction has been delayed because the pool dirty
data has exceeded zfs_delay_min_dirty_percent. The latter counts how
many times the pool dirty data has exceeded zfs_dirty_data_max (which
we expect to never happen).
- The original patch would have regressed the bug fixed in
zfsonlinux/zfs@c418410, which prevented users from setting the
zfs_vdev_aggregation_limit tuning larger than SPA_MAXBLOCKSIZE.
A similar fix is added to vdev_queue_aggregate().
- In vdev_queue_io_to_issue(), dynamically allocate 'zio_t search' on the
heap instead of the stack. In Linux we can't afford such large
structures on the stack.
Reviewed by: George Wilson <george.wilson@delphix.com>
Reviewed by: Adam Leventhal <ahl@delphix.com>
Reviewed by: Christopher Siden <christopher.siden@delphix.com>
Reviewed by: Ned Bass <bass6@llnl.gov>
Reviewed by: Brendan Gregg <brendan.gregg@joyent.com>
Approved by: Robert Mustacchi <rm@joyent.com>
References:
http://www.illumos.org/issues/4045
illumos/illumos-gate@69962b5647e4a8b9b14998733b765925381b727e
Ported-by: Ned Bass <bass6@llnl.gov>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #1913
2013-08-29 07:01:20 +04:00
|
|
|
*/
|
|
|
|
void
|
|
|
|
dsl_dir_willuse_space(dsl_dir_t *dd, int64_t space, dmu_tx_t *tx)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
|
|
|
int64_t parent_space;
|
|
|
|
uint64_t est_used;
|
|
|
|
|
2014-01-20 08:39:28 +04:00
|
|
|
do {
|
|
|
|
mutex_enter(&dd->dd_lock);
|
|
|
|
if (space > 0)
|
|
|
|
dd->dd_space_towrite[tx->tx_txg & TXG_MASK] += space;
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2014-01-20 08:39:28 +04:00
|
|
|
est_used = dsl_dir_space_towrite(dd) +
|
|
|
|
dd->dd_phys->dd_used_bytes;
|
|
|
|
parent_space = parent_delta(dd, est_used, space);
|
|
|
|
mutex_exit(&dd->dd_lock);
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2014-01-20 08:39:28 +04:00
|
|
|
/* Make sure that we clean up dd_space_to* */
|
|
|
|
dsl_dir_dirty(dd, tx);
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2014-01-20 08:39:28 +04:00
|
|
|
dd = dd->dd_parent;
|
|
|
|
space = parent_space;
|
|
|
|
} while (space && dd);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
/* call from syncing context when we actually write/free space for this dd */
|
|
|
|
void
|
2008-12-03 23:09:06 +03:00
|
|
|
dsl_dir_diduse_space(dsl_dir_t *dd, dd_used_t type,
|
2008-11-20 23:01:55 +03:00
|
|
|
int64_t used, int64_t compressed, int64_t uncompressed, dmu_tx_t *tx)
|
|
|
|
{
|
|
|
|
int64_t accounted_delta;
|
2013-08-22 21:51:47 +04:00
|
|
|
|
|
|
|
/*
|
|
|
|
* dsl_dataset_set_refreservation_sync_impl() calls this with
|
|
|
|
* dd_lock held, so that it can atomically update
|
|
|
|
* ds->ds_reserved and the dsl_dir accounting, so that
|
|
|
|
* dsl_dataset_check_quota() can see dataset and dir accounting
|
|
|
|
* consistently.
|
|
|
|
*/
|
2008-12-03 23:09:06 +03:00
|
|
|
boolean_t needlock = !MUTEX_HELD(&dd->dd_lock);
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
ASSERT(dmu_tx_is_syncing(tx));
|
2008-12-03 23:09:06 +03:00
|
|
|
ASSERT(type < DD_USED_NUM);
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-08-22 21:51:47 +04:00
|
|
|
dmu_buf_will_dirty(dd->dd_dbuf, tx);
|
|
|
|
|
2008-12-03 23:09:06 +03:00
|
|
|
if (needlock)
|
|
|
|
mutex_enter(&dd->dd_lock);
|
|
|
|
accounted_delta = parent_delta(dd, dd->dd_phys->dd_used_bytes, used);
|
|
|
|
ASSERT(used >= 0 || dd->dd_phys->dd_used_bytes >= -used);
|
2008-11-20 23:01:55 +03:00
|
|
|
ASSERT(compressed >= 0 ||
|
|
|
|
dd->dd_phys->dd_compressed_bytes >= -compressed);
|
|
|
|
ASSERT(uncompressed >= 0 ||
|
|
|
|
dd->dd_phys->dd_uncompressed_bytes >= -uncompressed);
|
2008-12-03 23:09:06 +03:00
|
|
|
dd->dd_phys->dd_used_bytes += used;
|
2008-11-20 23:01:55 +03:00
|
|
|
dd->dd_phys->dd_uncompressed_bytes += uncompressed;
|
|
|
|
dd->dd_phys->dd_compressed_bytes += compressed;
|
2008-12-03 23:09:06 +03:00
|
|
|
|
|
|
|
if (dd->dd_phys->dd_flags & DD_FLAG_USED_BREAKDOWN) {
|
|
|
|
ASSERT(used > 0 ||
|
|
|
|
dd->dd_phys->dd_used_breakdown[type] >= -used);
|
|
|
|
dd->dd_phys->dd_used_breakdown[type] += used;
|
|
|
|
#ifdef DEBUG
|
2010-08-26 20:52:39 +04:00
|
|
|
{
|
|
|
|
dd_used_t t;
|
|
|
|
uint64_t u = 0;
|
|
|
|
for (t = 0; t < DD_USED_NUM; t++)
|
|
|
|
u += dd->dd_phys->dd_used_breakdown[t];
|
|
|
|
ASSERT3U(u, ==, dd->dd_phys->dd_used_bytes);
|
|
|
|
}
|
2008-12-03 23:09:06 +03:00
|
|
|
#endif
|
|
|
|
}
|
|
|
|
if (needlock)
|
|
|
|
mutex_exit(&dd->dd_lock);
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
if (dd->dd_parent != NULL) {
|
2008-12-03 23:09:06 +03:00
|
|
|
dsl_dir_diduse_space(dd->dd_parent, DD_USED_CHILD,
|
2008-11-20 23:01:55 +03:00
|
|
|
accounted_delta, compressed, uncompressed, tx);
|
2008-12-03 23:09:06 +03:00
|
|
|
dsl_dir_transfer_space(dd->dd_parent,
|
|
|
|
used - accounted_delta,
|
|
|
|
DD_USED_CHILD_RSRV, DD_USED_CHILD, tx);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-12-03 23:09:06 +03:00
|
|
|
void
|
|
|
|
dsl_dir_transfer_space(dsl_dir_t *dd, int64_t delta,
|
|
|
|
dd_used_t oldtype, dd_used_t newtype, dmu_tx_t *tx)
|
|
|
|
{
|
|
|
|
ASSERT(dmu_tx_is_syncing(tx));
|
|
|
|
ASSERT(oldtype < DD_USED_NUM);
|
|
|
|
ASSERT(newtype < DD_USED_NUM);
|
|
|
|
|
|
|
|
if (delta == 0 || !(dd->dd_phys->dd_flags & DD_FLAG_USED_BREAKDOWN))
|
|
|
|
return;
|
|
|
|
|
2013-08-22 21:51:47 +04:00
|
|
|
dmu_buf_will_dirty(dd->dd_dbuf, tx);
|
|
|
|
mutex_enter(&dd->dd_lock);
|
2008-12-03 23:09:06 +03:00
|
|
|
ASSERT(delta > 0 ?
|
|
|
|
dd->dd_phys->dd_used_breakdown[oldtype] >= delta :
|
|
|
|
dd->dd_phys->dd_used_breakdown[newtype] >= -delta);
|
|
|
|
ASSERT(dd->dd_phys->dd_used_bytes >= ABS(delta));
|
|
|
|
dd->dd_phys->dd_used_breakdown[oldtype] -= delta;
|
|
|
|
dd->dd_phys->dd_used_breakdown[newtype] += delta;
|
2013-08-22 21:51:47 +04:00
|
|
|
mutex_exit(&dd->dd_lock);
|
2008-12-03 23:09:06 +03:00
|
|
|
}
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
typedef struct dsl_dir_set_qr_arg {
|
|
|
|
const char *ddsqra_name;
|
|
|
|
zprop_source_t ddsqra_source;
|
|
|
|
uint64_t ddsqra_value;
|
|
|
|
} dsl_dir_set_qr_arg_t;
|
|
|
|
|
2008-11-20 23:01:55 +03:00
|
|
|
static int
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_set_quota_check(void *arg, dmu_tx_t *tx)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_set_qr_arg_t *ddsqra = arg;
|
|
|
|
dsl_pool_t *dp = dmu_tx_pool(tx);
|
|
|
|
dsl_dataset_t *ds;
|
|
|
|
int error;
|
|
|
|
uint64_t towrite, newval;
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
error = dsl_dataset_hold(dp, ddsqra->ddsqra_name, FTAG, &ds);
|
|
|
|
if (error != 0)
|
|
|
|
return (error);
|
|
|
|
|
|
|
|
error = dsl_prop_predict(ds->ds_dir, "quota",
|
|
|
|
ddsqra->ddsqra_source, ddsqra->ddsqra_value, &newval);
|
|
|
|
if (error != 0) {
|
|
|
|
dsl_dataset_rele(ds, FTAG);
|
|
|
|
return (error);
|
|
|
|
}
|
2010-05-29 00:45:14 +04:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
if (newval == 0) {
|
|
|
|
dsl_dataset_rele(ds, FTAG);
|
2008-11-20 23:01:55 +03:00
|
|
|
return (0);
|
2013-09-04 16:00:57 +04:00
|
|
|
}
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
mutex_enter(&ds->ds_dir->dd_lock);
|
2008-11-20 23:01:55 +03:00
|
|
|
/*
|
|
|
|
* If we are doing the preliminary check in open context, and
|
|
|
|
* there are pending changes, then don't fail it, since the
|
|
|
|
* pending changes could under-estimate the amount of space to be
|
|
|
|
* freed up.
|
|
|
|
*/
|
2013-09-04 16:00:57 +04:00
|
|
|
towrite = dsl_dir_space_towrite(ds->ds_dir);
|
2008-11-20 23:01:55 +03:00
|
|
|
if ((dmu_tx_is_syncing(tx) || towrite == 0) &&
|
2013-09-04 16:00:57 +04:00
|
|
|
(newval < ds->ds_dir->dd_phys->dd_reserved ||
|
|
|
|
newval < ds->ds_dir->dd_phys->dd_used_bytes + towrite)) {
|
2013-03-08 22:41:28 +04:00
|
|
|
error = SET_ERROR(ENOSPC);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
2013-09-04 16:00:57 +04:00
|
|
|
mutex_exit(&ds->ds_dir->dd_lock);
|
|
|
|
dsl_dataset_rele(ds, FTAG);
|
|
|
|
return (error);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_set_quota_sync(void *arg, dmu_tx_t *tx)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_set_qr_arg_t *ddsqra = arg;
|
|
|
|
dsl_pool_t *dp = dmu_tx_pool(tx);
|
|
|
|
dsl_dataset_t *ds;
|
|
|
|
uint64_t newval;
|
2010-05-29 00:45:14 +04:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
VERIFY0(dsl_dataset_hold(dp, ddsqra->ddsqra_name, FTAG, &ds));
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-05-23 21:07:25 +04:00
|
|
|
if (spa_version(dp->dp_spa) >= SPA_VERSION_RECVD_PROPS) {
|
|
|
|
dsl_prop_set_sync_impl(ds, zfs_prop_to_name(ZFS_PROP_QUOTA),
|
|
|
|
ddsqra->ddsqra_source, sizeof (ddsqra->ddsqra_value), 1,
|
|
|
|
&ddsqra->ddsqra_value, tx);
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-05-23 21:07:25 +04:00
|
|
|
VERIFY0(dsl_prop_get_int_ds(ds,
|
|
|
|
zfs_prop_to_name(ZFS_PROP_QUOTA), &newval));
|
|
|
|
} else {
|
|
|
|
newval = ddsqra->ddsqra_value;
|
|
|
|
spa_history_log_internal_ds(ds, "set", tx, "%s=%lld",
|
|
|
|
zfs_prop_to_name(ZFS_PROP_QUOTA), (longlong_t)newval);
|
|
|
|
}
|
2013-08-28 15:45:09 +04:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
dmu_buf_will_dirty(ds->ds_dir->dd_dbuf, tx);
|
|
|
|
mutex_enter(&ds->ds_dir->dd_lock);
|
|
|
|
ds->ds_dir->dd_phys->dd_quota = newval;
|
|
|
|
mutex_exit(&ds->ds_dir->dd_lock);
|
|
|
|
dsl_dataset_rele(ds, FTAG);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
int
|
2010-05-29 00:45:14 +04:00
|
|
|
dsl_dir_set_quota(const char *ddname, zprop_source_t source, uint64_t quota)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_set_qr_arg_t ddsqra;
|
2010-05-29 00:45:14 +04:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
ddsqra.ddsqra_name = ddname;
|
|
|
|
ddsqra.ddsqra_source = source;
|
|
|
|
ddsqra.ddsqra_value = quota;
|
2010-05-29 00:45:14 +04:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
return (dsl_sync_task(ddname, dsl_dir_set_quota_check,
|
|
|
|
dsl_dir_set_quota_sync, &ddsqra, 0));
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
int
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_set_reservation_check(void *arg, dmu_tx_t *tx)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_set_qr_arg_t *ddsqra = arg;
|
|
|
|
dsl_pool_t *dp = dmu_tx_pool(tx);
|
|
|
|
dsl_dataset_t *ds;
|
|
|
|
dsl_dir_t *dd;
|
|
|
|
uint64_t newval, used, avail;
|
|
|
|
int error;
|
2010-05-29 00:45:14 +04:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
error = dsl_dataset_hold(dp, ddsqra->ddsqra_name, FTAG, &ds);
|
|
|
|
if (error != 0)
|
|
|
|
return (error);
|
|
|
|
dd = ds->ds_dir;
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If we are doing the preliminary check in open context, the
|
|
|
|
* space estimates may be inaccurate.
|
|
|
|
*/
|
2013-09-04 16:00:57 +04:00
|
|
|
if (!dmu_tx_is_syncing(tx)) {
|
|
|
|
dsl_dataset_rele(ds, FTAG);
|
2008-11-20 23:01:55 +03:00
|
|
|
return (0);
|
2013-09-04 16:00:57 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
error = dsl_prop_predict(ds->ds_dir,
|
|
|
|
zfs_prop_to_name(ZFS_PROP_RESERVATION),
|
|
|
|
ddsqra->ddsqra_source, ddsqra->ddsqra_value, &newval);
|
|
|
|
if (error != 0) {
|
|
|
|
dsl_dataset_rele(ds, FTAG);
|
|
|
|
return (error);
|
|
|
|
}
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
mutex_enter(&dd->dd_lock);
|
2008-12-03 23:09:06 +03:00
|
|
|
used = dd->dd_phys->dd_used_bytes;
|
2008-11-20 23:01:55 +03:00
|
|
|
mutex_exit(&dd->dd_lock);
|
|
|
|
|
|
|
|
if (dd->dd_parent) {
|
|
|
|
avail = dsl_dir_space_available(dd->dd_parent,
|
|
|
|
NULL, 0, FALSE);
|
|
|
|
} else {
|
|
|
|
avail = dsl_pool_adjustedsize(dd->dd_pool, B_FALSE) - used;
|
|
|
|
}
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
if (MAX(used, newval) > MAX(used, dd->dd_phys->dd_reserved)) {
|
|
|
|
uint64_t delta = MAX(used, newval) -
|
2009-02-18 23:51:31 +03:00
|
|
|
MAX(used, dd->dd_phys->dd_reserved);
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
if (delta > avail ||
|
|
|
|
(dd->dd_phys->dd_quota > 0 &&
|
|
|
|
newval > dd->dd_phys->dd_quota))
|
2013-03-08 22:41:28 +04:00
|
|
|
error = SET_ERROR(ENOSPC);
|
2009-02-18 23:51:31 +03:00
|
|
|
}
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dataset_rele(ds, FTAG);
|
|
|
|
return (error);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
void
|
2013-08-28 15:45:09 +04:00
|
|
|
dsl_dir_set_reservation_sync_impl(dsl_dir_t *dd, uint64_t value, dmu_tx_t *tx)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
|
|
|
uint64_t used;
|
|
|
|
int64_t delta;
|
|
|
|
|
|
|
|
dmu_buf_will_dirty(dd->dd_dbuf, tx);
|
|
|
|
|
|
|
|
mutex_enter(&dd->dd_lock);
|
2008-12-03 23:09:06 +03:00
|
|
|
used = dd->dd_phys->dd_used_bytes;
|
2013-08-28 15:45:09 +04:00
|
|
|
delta = MAX(used, value) - MAX(used, dd->dd_phys->dd_reserved);
|
|
|
|
dd->dd_phys->dd_reserved = value;
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
if (dd->dd_parent != NULL) {
|
|
|
|
/* Roll up this additional usage into our ancestors */
|
2008-12-03 23:09:06 +03:00
|
|
|
dsl_dir_diduse_space(dd->dd_parent, DD_USED_CHILD_RSRV,
|
|
|
|
delta, 0, 0, tx);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
2008-12-03 23:09:06 +03:00
|
|
|
mutex_exit(&dd->dd_lock);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
2013-08-28 15:45:09 +04:00
|
|
|
static void
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_set_reservation_sync(void *arg, dmu_tx_t *tx)
|
2013-08-28 15:45:09 +04:00
|
|
|
{
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_set_qr_arg_t *ddsqra = arg;
|
|
|
|
dsl_pool_t *dp = dmu_tx_pool(tx);
|
|
|
|
dsl_dataset_t *ds;
|
|
|
|
uint64_t newval;
|
2013-08-28 15:45:09 +04:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
VERIFY0(dsl_dataset_hold(dp, ddsqra->ddsqra_name, FTAG, &ds));
|
|
|
|
|
2013-05-23 21:07:25 +04:00
|
|
|
if (spa_version(dp->dp_spa) >= SPA_VERSION_RECVD_PROPS) {
|
|
|
|
dsl_prop_set_sync_impl(ds,
|
|
|
|
zfs_prop_to_name(ZFS_PROP_RESERVATION),
|
|
|
|
ddsqra->ddsqra_source, sizeof (ddsqra->ddsqra_value), 1,
|
|
|
|
&ddsqra->ddsqra_value, tx);
|
2013-11-01 23:26:11 +04:00
|
|
|
|
2013-05-23 21:07:25 +04:00
|
|
|
VERIFY0(dsl_prop_get_int_ds(ds,
|
|
|
|
zfs_prop_to_name(ZFS_PROP_RESERVATION), &newval));
|
|
|
|
} else {
|
|
|
|
newval = ddsqra->ddsqra_value;
|
|
|
|
spa_history_log_internal_ds(ds, "set", tx, "%s=%lld",
|
|
|
|
zfs_prop_to_name(ZFS_PROP_RESERVATION),
|
|
|
|
(longlong_t)newval);
|
|
|
|
}
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_set_reservation_sync_impl(ds->ds_dir, newval, tx);
|
|
|
|
dsl_dataset_rele(ds, FTAG);
|
2013-11-01 23:26:11 +04:00
|
|
|
}
|
2013-08-28 15:45:09 +04:00
|
|
|
|
2008-11-20 23:01:55 +03:00
|
|
|
int
|
2010-05-29 00:45:14 +04:00
|
|
|
dsl_dir_set_reservation(const char *ddname, zprop_source_t source,
|
|
|
|
uint64_t reservation)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_set_qr_arg_t ddsqra;
|
2010-05-29 00:45:14 +04:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
ddsqra.ddsqra_name = ddname;
|
|
|
|
ddsqra.ddsqra_source = source;
|
|
|
|
ddsqra.ddsqra_value = reservation;
|
2010-05-29 00:45:14 +04:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
return (dsl_sync_task(ddname, dsl_dir_set_reservation_check,
|
|
|
|
dsl_dir_set_reservation_sync, &ddsqra, 0));
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
static dsl_dir_t *
|
|
|
|
closest_common_ancestor(dsl_dir_t *ds1, dsl_dir_t *ds2)
|
|
|
|
{
|
|
|
|
for (; ds1; ds1 = ds1->dd_parent) {
|
|
|
|
dsl_dir_t *dd;
|
|
|
|
for (dd = ds2; dd; dd = dd->dd_parent) {
|
|
|
|
if (ds1 == dd)
|
|
|
|
return (dd);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return (NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If delta is applied to dd, how much of that delta would be applied to
|
|
|
|
* ancestor? Syncing context only.
|
|
|
|
*/
|
|
|
|
static int64_t
|
|
|
|
would_change(dsl_dir_t *dd, int64_t delta, dsl_dir_t *ancestor)
|
|
|
|
{
|
|
|
|
if (dd == ancestor)
|
|
|
|
return (delta);
|
|
|
|
|
|
|
|
mutex_enter(&dd->dd_lock);
|
2008-12-03 23:09:06 +03:00
|
|
|
delta = parent_delta(dd, dd->dd_phys->dd_used_bytes, delta);
|
2008-11-20 23:01:55 +03:00
|
|
|
mutex_exit(&dd->dd_lock);
|
|
|
|
return (would_change(dd->dd_parent, delta, ancestor));
|
|
|
|
}
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
typedef struct dsl_dir_rename_arg {
|
|
|
|
const char *ddra_oldname;
|
|
|
|
const char *ddra_newname;
|
|
|
|
} dsl_dir_rename_arg_t;
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
/* ARGSUSED */
|
2008-11-20 23:01:55 +03:00
|
|
|
static int
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_valid_rename(dsl_pool_t *dp, dsl_dataset_t *ds, void *arg)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
2013-09-04 16:00:57 +04:00
|
|
|
int *deltap = arg;
|
|
|
|
char namebuf[MAXNAMELEN];
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dataset_name(ds, namebuf);
|
|
|
|
|
|
|
|
if (strlen(namebuf) + *deltap >= MAXNAMELEN)
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(ENAMETOOLONG));
|
2013-09-04 16:00:57 +04:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
dsl_dir_rename_check(void *arg, dmu_tx_t *tx)
|
|
|
|
{
|
|
|
|
dsl_dir_rename_arg_t *ddra = arg;
|
|
|
|
dsl_pool_t *dp = dmu_tx_pool(tx);
|
|
|
|
dsl_dir_t *dd, *newparent;
|
|
|
|
const char *mynewname;
|
|
|
|
int error;
|
|
|
|
int delta = strlen(ddra->ddra_newname) - strlen(ddra->ddra_oldname);
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
/* target dir should exist */
|
|
|
|
error = dsl_dir_hold(dp, ddra->ddra_oldname, FTAG, &dd, NULL);
|
|
|
|
if (error != 0)
|
|
|
|
return (error);
|
|
|
|
|
|
|
|
/* new parent should exist */
|
|
|
|
error = dsl_dir_hold(dp, ddra->ddra_newname, FTAG,
|
|
|
|
&newparent, &mynewname);
|
|
|
|
if (error != 0) {
|
|
|
|
dsl_dir_rele(dd, FTAG);
|
|
|
|
return (error);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* can't rename to different pool */
|
|
|
|
if (dd->dd_pool != newparent->dd_pool) {
|
|
|
|
dsl_dir_rele(newparent, FTAG);
|
|
|
|
dsl_dir_rele(dd, FTAG);
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(ENXIO));
|
2013-09-04 16:00:57 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
/* new name should not already exist */
|
|
|
|
if (mynewname == NULL) {
|
|
|
|
dsl_dir_rele(newparent, FTAG);
|
|
|
|
dsl_dir_rele(dd, FTAG);
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(EEXIST));
|
2013-09-04 16:00:57 +04:00
|
|
|
}
|
|
|
|
|
|
|
|
/* if the name length is growing, validate child name lengths */
|
|
|
|
if (delta > 0) {
|
|
|
|
error = dmu_objset_find_dp(dp, dd->dd_object, dsl_valid_rename,
|
|
|
|
&delta, DS_FIND_CHILDREN | DS_FIND_SNAPSHOTS);
|
|
|
|
if (error != 0) {
|
|
|
|
dsl_dir_rele(newparent, FTAG);
|
|
|
|
dsl_dir_rele(dd, FTAG);
|
|
|
|
return (error);
|
|
|
|
}
|
|
|
|
}
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
if (newparent != dd->dd_parent) {
|
2008-11-20 23:01:55 +03:00
|
|
|
/* is there enough space? */
|
|
|
|
uint64_t myspace =
|
2008-12-03 23:09:06 +03:00
|
|
|
MAX(dd->dd_phys->dd_used_bytes, dd->dd_phys->dd_reserved);
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
/* no rename into our descendant */
|
2013-09-04 16:00:57 +04:00
|
|
|
if (closest_common_ancestor(dd, newparent) == dd) {
|
|
|
|
dsl_dir_rele(newparent, FTAG);
|
|
|
|
dsl_dir_rele(dd, FTAG);
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(EINVAL));
|
2013-09-04 16:00:57 +04:00
|
|
|
}
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
error = dsl_dir_transfer_possible(dd->dd_parent,
|
|
|
|
newparent, myspace);
|
|
|
|
if (error != 0) {
|
|
|
|
dsl_dir_rele(newparent, FTAG);
|
|
|
|
dsl_dir_rele(dd, FTAG);
|
|
|
|
return (error);
|
|
|
|
}
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_rele(newparent, FTAG);
|
|
|
|
dsl_dir_rele(dd, FTAG);
|
2008-11-20 23:01:55 +03:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_rename_sync(void *arg, dmu_tx_t *tx)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_rename_arg_t *ddra = arg;
|
|
|
|
dsl_pool_t *dp = dmu_tx_pool(tx);
|
|
|
|
dsl_dir_t *dd, *newparent;
|
|
|
|
const char *mynewname;
|
|
|
|
int error;
|
2008-11-20 23:01:55 +03:00
|
|
|
objset_t *mos = dp->dp_meta_objset;
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
VERIFY0(dsl_dir_hold(dp, ddra->ddra_oldname, FTAG, &dd, NULL));
|
|
|
|
VERIFY0(dsl_dir_hold(dp, ddra->ddra_newname, FTAG, &newparent,
|
|
|
|
&mynewname));
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-08-28 15:45:09 +04:00
|
|
|
/* Log this before we change the name. */
|
|
|
|
spa_history_log_internal_dd(dd, "rename", tx,
|
2013-09-04 16:00:57 +04:00
|
|
|
"-> %s", ddra->ddra_newname);
|
2013-08-28 15:45:09 +04:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
if (newparent != dd->dd_parent) {
|
2008-12-03 23:09:06 +03:00
|
|
|
dsl_dir_diduse_space(dd->dd_parent, DD_USED_CHILD,
|
|
|
|
-dd->dd_phys->dd_used_bytes,
|
2008-11-20 23:01:55 +03:00
|
|
|
-dd->dd_phys->dd_compressed_bytes,
|
|
|
|
-dd->dd_phys->dd_uncompressed_bytes, tx);
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_diduse_space(newparent, DD_USED_CHILD,
|
2008-12-03 23:09:06 +03:00
|
|
|
dd->dd_phys->dd_used_bytes,
|
2008-11-20 23:01:55 +03:00
|
|
|
dd->dd_phys->dd_compressed_bytes,
|
|
|
|
dd->dd_phys->dd_uncompressed_bytes, tx);
|
2008-12-03 23:09:06 +03:00
|
|
|
|
|
|
|
if (dd->dd_phys->dd_reserved > dd->dd_phys->dd_used_bytes) {
|
|
|
|
uint64_t unused_rsrv = dd->dd_phys->dd_reserved -
|
|
|
|
dd->dd_phys->dd_used_bytes;
|
|
|
|
|
|
|
|
dsl_dir_diduse_space(dd->dd_parent, DD_USED_CHILD_RSRV,
|
|
|
|
-unused_rsrv, 0, 0, tx);
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_diduse_space(newparent, DD_USED_CHILD_RSRV,
|
2008-12-03 23:09:06 +03:00
|
|
|
unused_rsrv, 0, 0, tx);
|
|
|
|
}
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
dmu_buf_will_dirty(dd->dd_dbuf, tx);
|
|
|
|
|
|
|
|
/* remove from old parent zapobj */
|
2013-09-04 16:00:57 +04:00
|
|
|
error = zap_remove(mos, dd->dd_parent->dd_phys->dd_child_dir_zapobj,
|
2008-11-20 23:01:55 +03:00
|
|
|
dd->dd_myname, tx);
|
2013-09-04 16:00:57 +04:00
|
|
|
ASSERT0(error);
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
(void) strcpy(dd->dd_myname, mynewname);
|
|
|
|
dsl_dir_rele(dd->dd_parent, dd);
|
|
|
|
dd->dd_phys->dd_parent_obj = newparent->dd_object;
|
|
|
|
VERIFY0(dsl_dir_hold_obj(dp,
|
|
|
|
newparent->dd_object, NULL, dd, &dd->dd_parent));
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
/* add to new parent zapobj */
|
2013-09-04 16:00:57 +04:00
|
|
|
VERIFY0(zap_add(mos, newparent->dd_phys->dd_child_dir_zapobj,
|
|
|
|
dd->dd_myname, 8, 1, &dd->dd_object, tx));
|
|
|
|
|
2013-12-07 02:20:22 +04:00
|
|
|
#ifdef _KERNEL
|
|
|
|
zvol_rename_minors(ddra->ddra_oldname, ddra->ddra_newname);
|
|
|
|
#endif
|
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_prop_notify_all(dd);
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_rele(newparent, FTAG);
|
|
|
|
dsl_dir_rele(dd, FTAG);
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
int
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_rename(const char *oldname, const char *newname)
|
2008-11-20 23:01:55 +03:00
|
|
|
{
|
2013-09-04 16:00:57 +04:00
|
|
|
dsl_dir_rename_arg_t ddra;
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
ddra.ddra_oldname = oldname;
|
|
|
|
ddra.ddra_newname = newname;
|
2008-11-20 23:01:55 +03:00
|
|
|
|
2013-09-04 16:00:57 +04:00
|
|
|
return (dsl_sync_task(oldname,
|
|
|
|
dsl_dir_rename_check, dsl_dir_rename_sync, &ddra, 3));
|
2008-11-20 23:01:55 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
int
|
|
|
|
dsl_dir_transfer_possible(dsl_dir_t *sdd, dsl_dir_t *tdd, uint64_t space)
|
|
|
|
{
|
|
|
|
dsl_dir_t *ancestor;
|
|
|
|
int64_t adelta;
|
|
|
|
uint64_t avail;
|
|
|
|
|
|
|
|
ancestor = closest_common_ancestor(sdd, tdd);
|
|
|
|
adelta = would_change(sdd, -space, ancestor);
|
|
|
|
avail = dsl_dir_space_available(tdd, ancestor, adelta, FALSE);
|
|
|
|
if (avail < space)
|
2013-03-08 22:41:28 +04:00
|
|
|
return (SET_ERROR(ENOSPC));
|
2008-11-20 23:01:55 +03:00
|
|
|
|
|
|
|
return (0);
|
|
|
|
}
|
2010-05-29 00:45:14 +04:00
|
|
|
|
|
|
|
timestruc_t
|
|
|
|
dsl_dir_snap_cmtime(dsl_dir_t *dd)
|
|
|
|
{
|
|
|
|
timestruc_t t;
|
|
|
|
|
|
|
|
mutex_enter(&dd->dd_lock);
|
|
|
|
t = dd->dd_snap_cmtime;
|
|
|
|
mutex_exit(&dd->dd_lock);
|
|
|
|
|
|
|
|
return (t);
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
dsl_dir_snap_cmtime_update(dsl_dir_t *dd)
|
|
|
|
{
|
|
|
|
timestruc_t t;
|
|
|
|
|
|
|
|
gethrestime(&t);
|
|
|
|
mutex_enter(&dd->dd_lock);
|
|
|
|
dd->dd_snap_cmtime = t;
|
|
|
|
mutex_exit(&dd->dd_lock);
|
|
|
|
}
|
2010-08-26 22:49:16 +04:00
|
|
|
|
|
|
|
#if defined(_KERNEL) && defined(HAVE_SPL)
|
|
|
|
EXPORT_SYMBOL(dsl_dir_set_quota);
|
|
|
|
EXPORT_SYMBOL(dsl_dir_set_reservation);
|
|
|
|
#endif
|