2010-08-26 18:45:02 +00:00
|
|
|
/*
|
|
|
|
* 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
|
|
|
|
*/
|
|
|
|
/*
|
|
|
|
* Copyright (C) 2008-2010 Lawrence Livermore National Security, LLC.
|
|
|
|
* Produced at Lawrence Livermore National Laboratory (cf, DISCLAIMER).
|
|
|
|
* Rewritten for Linux by Brian Behlendorf <behlendorf1@llnl.gov>.
|
|
|
|
* LLNL-CODE-403049.
|
|
|
|
*
|
|
|
|
* ZFS volume emulation driver.
|
|
|
|
*
|
|
|
|
* Makes a DMU object look like a volume of arbitrary size, up to 2^64 bytes.
|
|
|
|
* Volumes are accessed through the symbolic links named:
|
|
|
|
*
|
|
|
|
* /dev/<pool_name>/<dataset_name>
|
|
|
|
*
|
|
|
|
* Volumes are persistent through reboot and module load. No user command
|
|
|
|
* needs to be run before opening and using a device.
|
2015-08-02 13:01:14 +00:00
|
|
|
*
|
|
|
|
* Copyright 2014 Nexenta Systems, Inc. All rights reserved.
|
2016-02-16 19:52:55 +00:00
|
|
|
* Copyright (c) 2016 Actifio, Inc. All rights reserved.
|
2010-08-26 18:45:02 +00:00
|
|
|
*/
|
|
|
|
|
2013-05-10 19:47:54 +00:00
|
|
|
#include <sys/dbuf.h>
|
2010-08-26 18:45:02 +00:00
|
|
|
#include <sys/dmu_traverse.h>
|
|
|
|
#include <sys/dsl_dataset.h>
|
|
|
|
#include <sys/dsl_prop.h>
|
2014-03-22 09:07:14 +00:00
|
|
|
#include <sys/dsl_dir.h>
|
2010-08-26 18:45:02 +00:00
|
|
|
#include <sys/zap.h>
|
2015-08-24 21:18:48 +00:00
|
|
|
#include <sys/zfeature.h>
|
2010-08-26 18:45:02 +00:00
|
|
|
#include <sys/zil_impl.h>
|
2015-08-02 13:01:14 +00:00
|
|
|
#include <sys/dmu_tx.h>
|
2010-08-26 18:45:02 +00:00
|
|
|
#include <sys/zio.h>
|
|
|
|
#include <sys/zfs_rlock.h>
|
|
|
|
#include <sys/zfs_znode.h>
|
2014-03-22 09:07:14 +00:00
|
|
|
#include <sys/spa_impl.h>
|
2010-08-26 18:45:02 +00:00
|
|
|
#include <sys/zvol.h>
|
2011-02-22 20:15:13 +00:00
|
|
|
#include <linux/blkdev_compat.h>
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2012-06-02 01:49:10 +00:00
|
|
|
unsigned int zvol_inhibit_dev = 0;
|
2010-08-26 18:45:02 +00:00
|
|
|
unsigned int zvol_major = ZVOL_MAJOR;
|
2015-08-18 20:51:20 +00:00
|
|
|
unsigned int zvol_prefetch_bytes = (128 * 1024);
|
Limit the number of blocks to discard at once.
The number of blocks that can be discarded in one BLKDISCARD ioctl on a
zvol is currently unlimited. Some applications, such as mkfs, discard
the whole volume at once and they use the maximum possible discard size
to do that. As a result, several gigabytes discard requests are not
uncommon.
Unfortunately, if a large amount of data is allocated in the zvol, ZFS
can be quite slow to process discard requests. This is especially true
if the volblocksize is low (e.g. the 8K default). As a result, very
large discard requests can take a very long time (seconds to minutes
under heavy load) to complete. This can cause a number of problems, most
notably if the zvol is accessed remotely (e.g. via iSCSI), in which case
the client has a high probability of timing out on the request.
This patch solves the issue by adding a new tunable module parameter:
zvol_max_discard_blocks. This indicates the maximum possible range, in
zvol blocks, of one discard operation. It is set by default to 16384
blocks, which appears to be a good tradeoff. Using the default
volblocksize of 8K this is equivalent to 128 MB. When using the maximum
volblocksize of 128K this is equivalent to 2 GB.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #858
2012-07-31 08:45:37 +00:00
|
|
|
unsigned long zvol_max_discard_blocks = 16384;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
static kmutex_t zvol_state_lock;
|
|
|
|
static list_t zvol_state_list;
|
2016-01-06 21:22:48 +00:00
|
|
|
void *zvol_tag = "zvol_tag";
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* The in-core state of each volume.
|
|
|
|
*/
|
|
|
|
typedef struct zvol_state {
|
2011-02-22 10:58:44 +00:00
|
|
|
char zv_name[MAXNAMELEN]; /* name */
|
2013-12-12 21:04:40 +00:00
|
|
|
uint64_t zv_volsize; /* advertised space */
|
|
|
|
uint64_t zv_volblocksize; /* volume block size */
|
2010-08-26 18:45:02 +00:00
|
|
|
objset_t *zv_objset; /* objset handle */
|
|
|
|
uint32_t zv_flags; /* ZVOL_* flags */
|
|
|
|
uint32_t zv_open_count; /* open counts */
|
|
|
|
uint32_t zv_changed; /* disk changed */
|
|
|
|
zilog_t *zv_zilog; /* ZIL handle */
|
2016-04-11 21:53:48 +00:00
|
|
|
zfs_rlock_t zv_range_lock; /* range lock */
|
2010-08-26 18:45:02 +00:00
|
|
|
dmu_buf_t *zv_dbuf; /* bonus handle */
|
|
|
|
dev_t zv_dev; /* device id */
|
|
|
|
struct gendisk *zv_disk; /* generic disk */
|
|
|
|
struct request_queue *zv_queue; /* request queue */
|
|
|
|
list_node_t zv_next; /* next zvol_state_t linkage */
|
|
|
|
} zvol_state_t;
|
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
typedef enum {
|
|
|
|
ZVOL_ASYNC_CREATE_MINORS,
|
|
|
|
ZVOL_ASYNC_REMOVE_MINORS,
|
|
|
|
ZVOL_ASYNC_RENAME_MINORS,
|
|
|
|
ZVOL_ASYNC_SET_SNAPDEV,
|
|
|
|
ZVOL_ASYNC_MAX
|
|
|
|
} zvol_async_op_t;
|
|
|
|
|
|
|
|
typedef struct {
|
|
|
|
zvol_async_op_t op;
|
|
|
|
char pool[MAXNAMELEN];
|
|
|
|
char name1[MAXNAMELEN];
|
|
|
|
char name2[MAXNAMELEN];
|
|
|
|
zprop_source_t source;
|
|
|
|
uint64_t snapdev;
|
|
|
|
} zvol_task_t;
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
#define ZVOL_RDONLY 0x1
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Find the next available range of ZVOL_MINORS minor numbers. The
|
|
|
|
* zvol_state_list is kept in ascending minor order so we simply need
|
|
|
|
* to scan the list for the first gap in the sequence. This allows us
|
|
|
|
* to recycle minor number as devices are created and removed.
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
zvol_find_minor(unsigned *minor)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv;
|
|
|
|
|
|
|
|
*minor = 0;
|
|
|
|
ASSERT(MUTEX_HELD(&zvol_state_lock));
|
|
|
|
for (zv = list_head(&zvol_state_list); zv != NULL;
|
2013-12-12 21:04:40 +00:00
|
|
|
zv = list_next(&zvol_state_list, zv), *minor += ZVOL_MINORS) {
|
2010-08-26 18:45:02 +00:00
|
|
|
if (MINOR(zv->zv_dev) != MINOR(*minor))
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* All minors are in use */
|
|
|
|
if (*minor >= (1 << MINORBITS))
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(ENXIO));
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (0);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Find a zvol_state_t given the full major+minor dev_t.
|
|
|
|
*/
|
|
|
|
static zvol_state_t *
|
|
|
|
zvol_find_by_dev(dev_t dev)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv;
|
|
|
|
|
|
|
|
ASSERT(MUTEX_HELD(&zvol_state_lock));
|
|
|
|
for (zv = list_head(&zvol_state_list); zv != NULL;
|
2013-12-12 21:04:40 +00:00
|
|
|
zv = list_next(&zvol_state_list, zv)) {
|
2010-08-26 18:45:02 +00:00
|
|
|
if (zv->zv_dev == dev)
|
2013-12-12 21:04:40 +00:00
|
|
|
return (zv);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (NULL);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Find a zvol_state_t given the name provided at zvol_alloc() time.
|
|
|
|
*/
|
|
|
|
static zvol_state_t *
|
|
|
|
zvol_find_by_name(const char *name)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv;
|
|
|
|
|
|
|
|
ASSERT(MUTEX_HELD(&zvol_state_lock));
|
|
|
|
for (zv = list_head(&zvol_state_list); zv != NULL;
|
2013-12-12 21:04:40 +00:00
|
|
|
zv = list_next(&zvol_state_list, zv)) {
|
|
|
|
if (strncmp(zv->zv_name, name, MAXNAMELEN) == 0)
|
|
|
|
return (zv);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (NULL);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2012-12-17 01:33:57 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Given a path, return TRUE if path is a ZVOL.
|
|
|
|
*/
|
|
|
|
boolean_t
|
|
|
|
zvol_is_zvol(const char *device)
|
|
|
|
{
|
|
|
|
struct block_device *bdev;
|
|
|
|
unsigned int major;
|
|
|
|
|
|
|
|
bdev = lookup_bdev(device);
|
|
|
|
if (IS_ERR(bdev))
|
|
|
|
return (B_FALSE);
|
|
|
|
|
|
|
|
major = MAJOR(bdev->bd_dev);
|
|
|
|
bdput(bdev);
|
|
|
|
|
|
|
|
if (major == zvol_major)
|
2013-12-12 21:04:40 +00:00
|
|
|
return (B_TRUE);
|
2012-12-17 01:33:57 +00:00
|
|
|
|
|
|
|
return (B_FALSE);
|
|
|
|
}
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
/*
|
|
|
|
* ZFS_IOC_CREATE callback handles dmu zvol and zap object creation.
|
|
|
|
*/
|
|
|
|
void
|
|
|
|
zvol_create_cb(objset_t *os, void *arg, cred_t *cr, dmu_tx_t *tx)
|
|
|
|
{
|
|
|
|
zfs_creat_t *zct = arg;
|
|
|
|
nvlist_t *nvprops = zct->zct_props;
|
|
|
|
int error;
|
|
|
|
uint64_t volblocksize, volsize;
|
|
|
|
|
|
|
|
VERIFY(nvlist_lookup_uint64(nvprops,
|
|
|
|
zfs_prop_to_name(ZFS_PROP_VOLSIZE), &volsize) == 0);
|
|
|
|
if (nvlist_lookup_uint64(nvprops,
|
|
|
|
zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE), &volblocksize) != 0)
|
|
|
|
volblocksize = zfs_prop_default_numeric(ZFS_PROP_VOLBLOCKSIZE);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* These properties must be removed from the list so the generic
|
|
|
|
* property setting step won't apply to them.
|
|
|
|
*/
|
|
|
|
VERIFY(nvlist_remove_all(nvprops,
|
|
|
|
zfs_prop_to_name(ZFS_PROP_VOLSIZE)) == 0);
|
|
|
|
(void) nvlist_remove_all(nvprops,
|
|
|
|
zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE));
|
|
|
|
|
|
|
|
error = dmu_object_claim(os, ZVOL_OBJ, DMU_OT_ZVOL, volblocksize,
|
|
|
|
DMU_OT_NONE, 0, tx);
|
|
|
|
ASSERT(error == 0);
|
|
|
|
|
|
|
|
error = zap_create_claim(os, ZVOL_ZAP_OBJ, DMU_OT_ZVOL_PROP,
|
|
|
|
DMU_OT_NONE, 0, tx);
|
|
|
|
ASSERT(error == 0);
|
|
|
|
|
|
|
|
error = zap_update(os, ZVOL_ZAP_OBJ, "size", 8, 1, &volsize, tx);
|
|
|
|
ASSERT(error == 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* ZFS_IOC_OBJSET_STATS entry point.
|
|
|
|
*/
|
|
|
|
int
|
|
|
|
zvol_get_stats(objset_t *os, nvlist_t *nv)
|
|
|
|
{
|
|
|
|
int error;
|
|
|
|
dmu_object_info_t *doi;
|
|
|
|
uint64_t val;
|
|
|
|
|
|
|
|
error = zap_lookup(os, ZVOL_ZAP_OBJ, "size", 8, 1, &val);
|
|
|
|
if (error)
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(error));
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_VOLSIZE, val);
|
2013-12-12 21:04:40 +00:00
|
|
|
doi = kmem_alloc(sizeof (dmu_object_info_t), KM_SLEEP);
|
2010-08-26 18:45:02 +00:00
|
|
|
error = dmu_object_info(os, ZVOL_OBJ, doi);
|
|
|
|
|
|
|
|
if (error == 0) {
|
|
|
|
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_VOLBLOCKSIZE,
|
|
|
|
doi->doi_data_block_size);
|
|
|
|
}
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
kmem_free(doi, sizeof (dmu_object_info_t));
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(error));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2014-01-13 22:27:33 +00:00
|
|
|
static void
|
|
|
|
zvol_size_changed(zvol_state_t *zv, uint64_t volsize)
|
|
|
|
{
|
|
|
|
struct block_device *bdev;
|
|
|
|
|
|
|
|
bdev = bdget_disk(zv->zv_disk, 0);
|
|
|
|
if (bdev == NULL)
|
|
|
|
return;
|
|
|
|
set_capacity(zv->zv_disk, volsize >> 9);
|
|
|
|
zv->zv_volsize = volsize;
|
|
|
|
check_disk_size_change(zv->zv_disk, bdev);
|
|
|
|
|
|
|
|
bdput(bdev);
|
|
|
|
}
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
/*
|
|
|
|
* Sanity check volume size.
|
|
|
|
*/
|
|
|
|
int
|
|
|
|
zvol_check_volsize(uint64_t volsize, uint64_t blocksize)
|
|
|
|
{
|
|
|
|
if (volsize == 0)
|
2013-03-08 18:41:28 +00:00
|
|
|
return (SET_ERROR(EINVAL));
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
if (volsize % blocksize != 0)
|
2013-03-08 18:41:28 +00:00
|
|
|
return (SET_ERROR(EINVAL));
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
#ifdef _ILP32
|
|
|
|
if (volsize - 1 > MAXOFFSET_T)
|
2013-03-08 18:41:28 +00:00
|
|
|
return (SET_ERROR(EOVERFLOW));
|
2010-08-26 18:45:02 +00:00
|
|
|
#endif
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Ensure the zap is flushed then inform the VFS of the capacity change.
|
|
|
|
*/
|
|
|
|
static int
|
2014-01-13 22:27:33 +00:00
|
|
|
zvol_update_volsize(uint64_t volsize, objset_t *os)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
|
|
|
dmu_tx_t *tx;
|
|
|
|
int error;
|
2016-02-26 07:33:44 +00:00
|
|
|
uint64_t txg;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
ASSERT(MUTEX_HELD(&zvol_state_lock));
|
|
|
|
|
2011-02-25 07:36:01 +00:00
|
|
|
tx = dmu_tx_create(os);
|
2010-08-26 18:45:02 +00:00
|
|
|
dmu_tx_hold_zap(tx, ZVOL_ZAP_OBJ, TRUE, NULL);
|
2014-07-07 19:49:36 +00:00
|
|
|
dmu_tx_mark_netfree(tx);
|
2010-08-26 18:45:02 +00:00
|
|
|
error = dmu_tx_assign(tx, TXG_WAIT);
|
|
|
|
if (error) {
|
|
|
|
dmu_tx_abort(tx);
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(error));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
2016-02-26 07:33:44 +00:00
|
|
|
txg = dmu_tx_get_txg(tx);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2011-02-25 07:36:01 +00:00
|
|
|
error = zap_update(os, ZVOL_ZAP_OBJ, "size", 8, 1,
|
2010-08-26 18:45:02 +00:00
|
|
|
&volsize, tx);
|
|
|
|
dmu_tx_commit(tx);
|
|
|
|
|
2016-02-26 07:33:44 +00:00
|
|
|
txg_wait_synced(dmu_objset_pool(os), txg);
|
|
|
|
|
2014-01-13 22:27:33 +00:00
|
|
|
if (error == 0)
|
|
|
|
error = dmu_free_long_range(os,
|
|
|
|
ZVOL_OBJ, volsize, DMU_OBJECT_END);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2014-01-13 22:27:33 +00:00
|
|
|
return (error);
|
|
|
|
}
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2014-01-13 22:27:33 +00:00
|
|
|
static int
|
|
|
|
zvol_update_live_volsize(zvol_state_t *zv, uint64_t volsize)
|
|
|
|
{
|
|
|
|
zvol_size_changed(zv, volsize);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2014-01-13 22:27:33 +00:00
|
|
|
/*
|
|
|
|
* We should post a event here describing the expansion. However,
|
|
|
|
* the zfs_ereport_post() interface doesn't nicely support posting
|
|
|
|
* events for zvols, it assumes events relate to vdevs or zios.
|
|
|
|
*/
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Set ZFS_PROP_VOLSIZE set entry point.
|
|
|
|
*/
|
|
|
|
int
|
|
|
|
zvol_set_volsize(const char *name, uint64_t volsize)
|
|
|
|
{
|
2014-01-13 22:27:33 +00:00
|
|
|
zvol_state_t *zv = NULL;
|
2010-08-26 18:45:02 +00:00
|
|
|
objset_t *os = NULL;
|
|
|
|
int error;
|
2014-01-13 22:27:33 +00:00
|
|
|
dmu_object_info_t *doi;
|
|
|
|
uint64_t readonly;
|
|
|
|
boolean_t owned = B_FALSE;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2013-09-04 12:00:57 +00:00
|
|
|
error = dsl_prop_get_integer(name,
|
|
|
|
zfs_prop_to_name(ZFS_PROP_READONLY), &readonly, NULL);
|
|
|
|
if (error != 0)
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(error));
|
2013-09-04 12:00:57 +00:00
|
|
|
if (readonly)
|
2013-03-08 18:41:28 +00:00
|
|
|
return (SET_ERROR(EROFS));
|
2013-09-04 12:00:57 +00:00
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
mutex_enter(&zvol_state_lock);
|
|
|
|
zv = zvol_find_by_name(name);
|
2014-01-13 22:27:33 +00:00
|
|
|
|
|
|
|
if (zv == NULL || zv->zv_objset == NULL) {
|
|
|
|
if ((error = dmu_objset_own(name, DMU_OST_ZVOL, B_FALSE,
|
|
|
|
FTAG, &os)) != 0) {
|
|
|
|
mutex_exit(&zvol_state_lock);
|
|
|
|
return (SET_ERROR(error));
|
|
|
|
}
|
|
|
|
owned = B_TRUE;
|
|
|
|
if (zv != NULL)
|
|
|
|
zv->zv_objset = os;
|
|
|
|
} else {
|
|
|
|
os = zv->zv_objset;
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
doi = kmem_alloc(sizeof (dmu_object_info_t), KM_SLEEP);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
if ((error = dmu_object_info(os, ZVOL_OBJ, doi)) ||
|
|
|
|
(error = zvol_check_volsize(volsize, doi->doi_data_block_size)))
|
2014-01-13 22:27:33 +00:00
|
|
|
goto out;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2014-01-13 22:27:33 +00:00
|
|
|
error = zvol_update_volsize(volsize, os);
|
2013-12-12 21:04:40 +00:00
|
|
|
kmem_free(doi, sizeof (dmu_object_info_t));
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2014-01-13 22:27:33 +00:00
|
|
|
if (error == 0 && zv != NULL)
|
|
|
|
error = zvol_update_live_volsize(zv, volsize);
|
|
|
|
out:
|
|
|
|
if (owned) {
|
|
|
|
dmu_objset_disown(os, FTAG);
|
|
|
|
if (zv != NULL)
|
|
|
|
zv->zv_objset = NULL;
|
|
|
|
}
|
2010-08-26 18:45:02 +00:00
|
|
|
mutex_exit(&zvol_state_lock);
|
2014-01-13 22:27:33 +00:00
|
|
|
return (error);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Sanity check volume block size.
|
|
|
|
*/
|
|
|
|
int
|
2015-08-24 21:18:48 +00:00
|
|
|
zvol_check_volblocksize(const char *name, uint64_t volblocksize)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
2015-08-24 21:18:48 +00:00
|
|
|
/* Record sizes above 128k need the feature to be enabled */
|
|
|
|
if (volblocksize > SPA_OLD_MAXBLOCKSIZE) {
|
|
|
|
spa_t *spa;
|
|
|
|
int error;
|
|
|
|
|
|
|
|
if ((error = spa_open(name, &spa, FTAG)) != 0)
|
|
|
|
return (error);
|
|
|
|
|
|
|
|
if (!spa_feature_is_enabled(spa, SPA_FEATURE_LARGE_BLOCKS)) {
|
|
|
|
spa_close(spa, FTAG);
|
|
|
|
return (SET_ERROR(ENOTSUP));
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We don't allow setting the property above 1MB,
|
|
|
|
* unless the tunable has been changed.
|
|
|
|
*/
|
|
|
|
if (volblocksize > zfs_max_recordsize)
|
|
|
|
return (SET_ERROR(EDOM));
|
|
|
|
|
|
|
|
spa_close(spa, FTAG);
|
|
|
|
}
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
if (volblocksize < SPA_MINBLOCKSIZE ||
|
|
|
|
volblocksize > SPA_MAXBLOCKSIZE ||
|
|
|
|
!ISP2(volblocksize))
|
2013-03-08 18:41:28 +00:00
|
|
|
return (SET_ERROR(EDOM));
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Set ZFS_PROP_VOLBLOCKSIZE set entry point.
|
|
|
|
*/
|
|
|
|
int
|
|
|
|
zvol_set_volblocksize(const char *name, uint64_t volblocksize)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv;
|
|
|
|
dmu_tx_t *tx;
|
|
|
|
int error;
|
|
|
|
|
|
|
|
mutex_enter(&zvol_state_lock);
|
|
|
|
|
|
|
|
zv = zvol_find_by_name(name);
|
|
|
|
if (zv == NULL) {
|
2013-03-08 18:41:28 +00:00
|
|
|
error = SET_ERROR(ENXIO);
|
2010-08-26 18:45:02 +00:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2013-12-06 22:20:22 +00:00
|
|
|
if (zv->zv_flags & ZVOL_RDONLY) {
|
2013-03-08 18:41:28 +00:00
|
|
|
error = SET_ERROR(EROFS);
|
2010-08-26 18:45:02 +00:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
tx = dmu_tx_create(zv->zv_objset);
|
|
|
|
dmu_tx_hold_bonus(tx, ZVOL_OBJ);
|
|
|
|
error = dmu_tx_assign(tx, TXG_WAIT);
|
|
|
|
if (error) {
|
|
|
|
dmu_tx_abort(tx);
|
|
|
|
} else {
|
|
|
|
error = dmu_object_set_blocksize(zv->zv_objset, ZVOL_OBJ,
|
|
|
|
volblocksize, 0, tx);
|
|
|
|
if (error == ENOTSUP)
|
2013-03-08 18:41:28 +00:00
|
|
|
error = SET_ERROR(EBUSY);
|
2010-08-26 18:45:02 +00:00
|
|
|
dmu_tx_commit(tx);
|
|
|
|
if (error == 0)
|
|
|
|
zv->zv_volblocksize = volblocksize;
|
|
|
|
}
|
|
|
|
out:
|
|
|
|
mutex_exit(&zvol_state_lock);
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(error));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2015-08-02 13:01:14 +00:00
|
|
|
/*
|
|
|
|
* Replay a TX_TRUNCATE ZIL transaction if asked. TX_TRUNCATE is how we
|
|
|
|
* implement DKIOCFREE/free-long-range.
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
zvol_replay_truncate(zvol_state_t *zv, lr_truncate_t *lr, boolean_t byteswap)
|
|
|
|
{
|
|
|
|
uint64_t offset, length;
|
|
|
|
|
|
|
|
if (byteswap)
|
|
|
|
byteswap_uint64_array(lr, sizeof (*lr));
|
|
|
|
|
|
|
|
offset = lr->lr_offset;
|
|
|
|
length = lr->lr_length;
|
|
|
|
|
|
|
|
return (dmu_free_long_range(zv->zv_objset, ZVOL_OBJ, offset, length));
|
|
|
|
}
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
/*
|
|
|
|
* Replay a TX_WRITE ZIL transaction that didn't get committed
|
|
|
|
* after a system failure
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
zvol_replay_write(zvol_state_t *zv, lr_write_t *lr, boolean_t byteswap)
|
|
|
|
{
|
|
|
|
objset_t *os = zv->zv_objset;
|
|
|
|
char *data = (char *)(lr + 1); /* data follows lr_write_t */
|
|
|
|
uint64_t off = lr->lr_offset;
|
|
|
|
uint64_t len = lr->lr_length;
|
|
|
|
dmu_tx_t *tx;
|
|
|
|
int error;
|
|
|
|
|
|
|
|
if (byteswap)
|
|
|
|
byteswap_uint64_array(lr, sizeof (*lr));
|
|
|
|
|
|
|
|
tx = dmu_tx_create(os);
|
|
|
|
dmu_tx_hold_write(tx, ZVOL_OBJ, off, len);
|
|
|
|
error = dmu_tx_assign(tx, TXG_WAIT);
|
|
|
|
if (error) {
|
|
|
|
dmu_tx_abort(tx);
|
|
|
|
} else {
|
|
|
|
dmu_write(os, ZVOL_OBJ, off, len, data, tx);
|
|
|
|
dmu_tx_commit(tx);
|
|
|
|
}
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(error));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
zvol_replay_err(zvol_state_t *zv, lr_t *lr, boolean_t byteswap)
|
|
|
|
{
|
2013-03-08 18:41:28 +00:00
|
|
|
return (SET_ERROR(ENOTSUP));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Callback vectors for replaying records.
|
2015-08-02 13:01:14 +00:00
|
|
|
* Only TX_WRITE and TX_TRUNCATE are needed for zvol.
|
2010-08-26 18:45:02 +00:00
|
|
|
*/
|
2013-02-15 04:37:43 +00:00
|
|
|
zil_replay_func_t zvol_replay_vector[TX_MAX_TYPE] = {
|
|
|
|
(zil_replay_func_t)zvol_replay_err, /* no such transaction type */
|
|
|
|
(zil_replay_func_t)zvol_replay_err, /* TX_CREATE */
|
|
|
|
(zil_replay_func_t)zvol_replay_err, /* TX_MKDIR */
|
|
|
|
(zil_replay_func_t)zvol_replay_err, /* TX_MKXATTR */
|
|
|
|
(zil_replay_func_t)zvol_replay_err, /* TX_SYMLINK */
|
|
|
|
(zil_replay_func_t)zvol_replay_err, /* TX_REMOVE */
|
|
|
|
(zil_replay_func_t)zvol_replay_err, /* TX_RMDIR */
|
|
|
|
(zil_replay_func_t)zvol_replay_err, /* TX_LINK */
|
|
|
|
(zil_replay_func_t)zvol_replay_err, /* TX_RENAME */
|
|
|
|
(zil_replay_func_t)zvol_replay_write, /* TX_WRITE */
|
2015-08-02 13:01:14 +00:00
|
|
|
(zil_replay_func_t)zvol_replay_truncate, /* TX_TRUNCATE */
|
2013-02-15 04:37:43 +00:00
|
|
|
(zil_replay_func_t)zvol_replay_err, /* TX_SETATTR */
|
|
|
|
(zil_replay_func_t)zvol_replay_err, /* TX_ACL */
|
2010-08-26 18:45:02 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* zvol_log_write() handles synchronous writes using TX_WRITE ZIL transactions.
|
|
|
|
*
|
|
|
|
* We store data in the log buffers if it's small enough.
|
|
|
|
* Otherwise we will later flush the data out via dmu_sync().
|
|
|
|
*/
|
|
|
|
ssize_t zvol_immediate_write_sz = 32768;
|
|
|
|
|
|
|
|
static void
|
2013-12-12 21:04:40 +00:00
|
|
|
zvol_log_write(zvol_state_t *zv, dmu_tx_t *tx, uint64_t offset,
|
|
|
|
uint64_t size, int sync)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
|
|
|
uint32_t blocksize = zv->zv_volblocksize;
|
|
|
|
zilog_t *zilog = zv->zv_zilog;
|
|
|
|
boolean_t slogging;
|
2012-06-06 09:30:24 +00:00
|
|
|
ssize_t immediate_write_sz;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
if (zil_replaying(zilog, tx))
|
|
|
|
return;
|
|
|
|
|
2012-06-06 09:30:24 +00:00
|
|
|
immediate_write_sz = (zilog->zl_logbias == ZFS_LOGBIAS_THROUGHPUT)
|
|
|
|
? 0 : zvol_immediate_write_sz;
|
|
|
|
slogging = spa_has_slogs(zilog->zl_spa) &&
|
|
|
|
(zilog->zl_logbias == ZFS_LOGBIAS_LATENCY);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
while (size) {
|
|
|
|
itx_t *itx;
|
|
|
|
lr_write_t *lr;
|
|
|
|
ssize_t len;
|
|
|
|
itx_wr_state_t write_state;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Unlike zfs_log_write() we can be called with
|
|
|
|
* up to DMU_MAX_ACCESS/2 (5MB) writes.
|
|
|
|
*/
|
2012-06-06 09:30:24 +00:00
|
|
|
if (blocksize > immediate_write_sz && !slogging &&
|
2010-08-26 18:45:02 +00:00
|
|
|
size >= blocksize && offset % blocksize == 0) {
|
|
|
|
write_state = WR_INDIRECT; /* uses dmu_sync */
|
|
|
|
len = blocksize;
|
|
|
|
} else if (sync) {
|
|
|
|
write_state = WR_COPIED;
|
|
|
|
len = MIN(ZIL_MAX_LOG_DATA, size);
|
|
|
|
} else {
|
|
|
|
write_state = WR_NEED_COPY;
|
|
|
|
len = MIN(ZIL_MAX_LOG_DATA, size);
|
|
|
|
}
|
|
|
|
|
|
|
|
itx = zil_itx_create(TX_WRITE, sizeof (*lr) +
|
|
|
|
(write_state == WR_COPIED ? len : 0));
|
|
|
|
lr = (lr_write_t *)&itx->itx_lr;
|
|
|
|
if (write_state == WR_COPIED && dmu_read(zv->zv_objset,
|
|
|
|
ZVOL_OBJ, offset, len, lr+1, DMU_READ_NO_PREFETCH) != 0) {
|
|
|
|
zil_itx_destroy(itx);
|
|
|
|
itx = zil_itx_create(TX_WRITE, sizeof (*lr));
|
|
|
|
lr = (lr_write_t *)&itx->itx_lr;
|
|
|
|
write_state = WR_NEED_COPY;
|
|
|
|
}
|
|
|
|
|
|
|
|
itx->itx_wr_state = write_state;
|
|
|
|
if (write_state == WR_NEED_COPY)
|
|
|
|
itx->itx_sod += len;
|
|
|
|
lr->lr_foid = ZVOL_OBJ;
|
|
|
|
lr->lr_offset = offset;
|
|
|
|
lr->lr_length = len;
|
|
|
|
lr->lr_blkoff = 0;
|
|
|
|
BP_ZERO(&lr->lr_blkptr);
|
|
|
|
|
|
|
|
itx->itx_private = zv;
|
|
|
|
itx->itx_sync = sync;
|
|
|
|
|
|
|
|
(void) zil_itx_assign(zilog, itx, tx);
|
|
|
|
|
|
|
|
offset += len;
|
|
|
|
size -= len;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
static int
|
2016-02-06 01:36:07 +00:00
|
|
|
zvol_write(zvol_state_t *zv, uio_t *uio, boolean_t sync)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
2016-02-06 01:36:07 +00:00
|
|
|
uint64_t volsize = zv->zv_volsize;
|
2010-08-26 18:45:02 +00:00
|
|
|
rl_t *rl;
|
2016-02-06 01:36:07 +00:00
|
|
|
int error = 0;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
ASSERT(zv && zv->zv_open_count > 0);
|
|
|
|
|
2016-04-11 21:53:48 +00:00
|
|
|
rl = zfs_range_lock(&zv->zv_range_lock, uio->uio_loffset,
|
|
|
|
uio->uio_resid, RL_WRITER);
|
2011-09-05 09:11:38 +00:00
|
|
|
|
2016-02-06 01:36:07 +00:00
|
|
|
while (uio->uio_resid > 0 && uio->uio_loffset < volsize) {
|
|
|
|
uint64_t bytes = MIN(uio->uio_resid, DMU_MAX_ACCESS >> 1);
|
|
|
|
uint64_t off = uio->uio_loffset;
|
|
|
|
dmu_tx_t *tx = dmu_tx_create(zv->zv_objset);
|
2011-09-05 09:11:38 +00:00
|
|
|
|
2016-02-06 01:36:07 +00:00
|
|
|
if (bytes > volsize - off) /* don't write past the end */
|
|
|
|
bytes = volsize - off;
|
2015-12-08 20:37:24 +00:00
|
|
|
|
2016-02-06 01:36:07 +00:00
|
|
|
dmu_tx_hold_write(tx, ZVOL_OBJ, off, bytes);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2016-02-06 01:36:07 +00:00
|
|
|
/* This will only fail for ENOSPC */
|
|
|
|
error = dmu_tx_assign(tx, TXG_WAIT);
|
|
|
|
if (error) {
|
|
|
|
dmu_tx_abort(tx);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
error = dmu_write_uio_dbuf(zv->zv_dbuf, uio, bytes, tx);
|
|
|
|
if (error == 0)
|
|
|
|
zvol_log_write(zv, tx, off, bytes, sync);
|
|
|
|
dmu_tx_commit(tx);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2016-02-06 01:36:07 +00:00
|
|
|
if (error)
|
|
|
|
break;
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
zfs_range_unlock(rl);
|
2016-02-06 01:36:07 +00:00
|
|
|
if (sync)
|
2010-08-26 18:45:02 +00:00
|
|
|
zil_commit(zv->zv_zilog, ZVOL_OBJ);
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
return (error);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2015-08-02 13:01:14 +00:00
|
|
|
/*
|
|
|
|
* Log a DKIOCFREE/free-long-range to the ZIL with TX_TRUNCATE.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
zvol_log_truncate(zvol_state_t *zv, dmu_tx_t *tx, uint64_t off, uint64_t len,
|
|
|
|
boolean_t sync)
|
|
|
|
{
|
|
|
|
itx_t *itx;
|
|
|
|
lr_truncate_t *lr;
|
|
|
|
zilog_t *zilog = zv->zv_zilog;
|
|
|
|
|
|
|
|
if (zil_replaying(zilog, tx))
|
|
|
|
return;
|
|
|
|
|
|
|
|
itx = zil_itx_create(TX_TRUNCATE, sizeof (*lr));
|
|
|
|
lr = (lr_truncate_t *)&itx->itx_lr;
|
|
|
|
lr->lr_foid = ZVOL_OBJ;
|
|
|
|
lr->lr_offset = off;
|
|
|
|
lr->lr_length = len;
|
|
|
|
|
|
|
|
itx->itx_sync = sync;
|
|
|
|
zil_itx_assign(zilog, itx, tx);
|
|
|
|
}
|
|
|
|
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
static int
|
|
|
|
zvol_discard(struct bio *bio)
|
2011-09-02 13:23:12 +00:00
|
|
|
{
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
zvol_state_t *zv = bio->bi_bdev->bd_disk->private_data;
|
|
|
|
uint64_t start = BIO_BI_SECTOR(bio) << 9;
|
|
|
|
uint64_t size = BIO_BI_SIZE(bio);
|
|
|
|
uint64_t end = start + size;
|
2011-09-02 13:23:12 +00:00
|
|
|
int error;
|
|
|
|
rl_t *rl;
|
2015-08-02 13:01:14 +00:00
|
|
|
dmu_tx_t *tx;
|
2011-09-02 13:23:12 +00:00
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
ASSERT(zv && zv->zv_open_count > 0);
|
|
|
|
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
if (end > zv->zv_volsize)
|
|
|
|
return (SET_ERROR(EIO));
|
2011-09-02 13:23:12 +00:00
|
|
|
|
2012-10-04 08:38:55 +00:00
|
|
|
/*
|
2016-08-09 18:22:30 +00:00
|
|
|
* Align the request to volume block boundaries when a secure erase is
|
|
|
|
* not required. This will prevent dnode_free_range() from zeroing out
|
|
|
|
* the unaligned parts which is slow (read-modify-write) and useless
|
|
|
|
* since we are not freeing any space by doing so.
|
2012-10-04 08:38:55 +00:00
|
|
|
*/
|
2016-08-09 18:22:30 +00:00
|
|
|
if (!bio_is_secure_erase(bio)) {
|
2014-10-10 15:23:23 +00:00
|
|
|
start = P2ROUNDUP(start, zv->zv_volblocksize);
|
|
|
|
end = P2ALIGN(end, zv->zv_volblocksize);
|
2015-09-18 12:32:52 +00:00
|
|
|
size = end - start;
|
2014-10-10 15:23:23 +00:00
|
|
|
}
|
2012-10-04 08:38:55 +00:00
|
|
|
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
if (start >= end)
|
|
|
|
return (0);
|
2011-09-02 13:23:12 +00:00
|
|
|
|
2016-04-11 21:53:48 +00:00
|
|
|
rl = zfs_range_lock(&zv->zv_range_lock, start, size, RL_WRITER);
|
2015-08-02 13:01:14 +00:00
|
|
|
tx = dmu_tx_create(zv->zv_objset);
|
|
|
|
dmu_tx_mark_netfree(tx);
|
|
|
|
error = dmu_tx_assign(tx, TXG_WAIT);
|
|
|
|
if (error != 0) {
|
|
|
|
dmu_tx_abort(tx);
|
|
|
|
} else {
|
|
|
|
zvol_log_truncate(zv, tx, start, size, B_TRUE);
|
|
|
|
dmu_tx_commit(tx);
|
|
|
|
error = dmu_free_long_range(zv->zv_objset,
|
|
|
|
ZVOL_OBJ, start, size);
|
|
|
|
}
|
2011-09-02 13:23:12 +00:00
|
|
|
|
|
|
|
zfs_range_unlock(rl);
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
|
|
|
|
return (error);
|
2011-09-02 13:23:12 +00:00
|
|
|
}
|
|
|
|
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
static int
|
2016-02-06 01:36:07 +00:00
|
|
|
zvol_read(zvol_state_t *zv, uio_t *uio)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
2016-02-06 01:36:07 +00:00
|
|
|
uint64_t volsize = zv->zv_volsize;
|
2010-08-26 18:45:02 +00:00
|
|
|
rl_t *rl;
|
2016-02-06 01:36:07 +00:00
|
|
|
int error = 0;
|
2011-09-05 09:11:38 +00:00
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
ASSERT(zv && zv->zv_open_count > 0);
|
|
|
|
|
2016-04-11 21:53:48 +00:00
|
|
|
rl = zfs_range_lock(&zv->zv_range_lock, uio->uio_loffset,
|
|
|
|
uio->uio_resid, RL_READER);
|
2016-02-06 01:36:07 +00:00
|
|
|
while (uio->uio_resid > 0 && uio->uio_loffset < volsize) {
|
|
|
|
uint64_t bytes = MIN(uio->uio_resid, DMU_MAX_ACCESS >> 1);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2016-02-06 01:36:07 +00:00
|
|
|
/* don't read past the end */
|
|
|
|
if (bytes > volsize - uio->uio_loffset)
|
|
|
|
bytes = volsize - uio->uio_loffset;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2016-02-06 04:10:13 +00:00
|
|
|
error = dmu_read_uio_dbuf(zv->zv_dbuf, uio, bytes);
|
2016-02-06 01:36:07 +00:00
|
|
|
if (error) {
|
|
|
|
/* convert checksum errors into IO errors */
|
|
|
|
if (error == ECKSUM)
|
|
|
|
error = SET_ERROR(EIO);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2010-08-26 18:45:02 +00:00
|
|
|
zfs_range_unlock(rl);
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
return (error);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
static MAKE_REQUEST_FN_RET
|
|
|
|
zvol_request(struct request_queue *q, struct bio *bio)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
2016-02-06 01:36:07 +00:00
|
|
|
uio_t uio;
|
2010-08-26 18:45:02 +00:00
|
|
|
zvol_state_t *zv = q->queuedata;
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
fstrans_cookie_t cookie = spl_fstrans_mark();
|
2015-09-07 16:03:19 +00:00
|
|
|
int rw = bio_data_dir(bio);
|
|
|
|
#ifdef HAVE_GENERIC_IO_ACCT
|
|
|
|
unsigned long start = jiffies;
|
|
|
|
#endif
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
int error = 0;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2016-02-06 01:36:07 +00:00
|
|
|
uio.uio_bvec = &bio->bi_io_vec[BIO_BI_IDX(bio)];
|
|
|
|
uio.uio_skip = BIO_BI_SKIP(bio);
|
|
|
|
uio.uio_resid = BIO_BI_SIZE(bio);
|
|
|
|
uio.uio_iovcnt = bio->bi_vcnt - BIO_BI_IDX(bio);
|
|
|
|
uio.uio_loffset = BIO_BI_SECTOR(bio) << 9;
|
|
|
|
uio.uio_limit = MAXOFFSET_T;
|
|
|
|
uio.uio_segflg = UIO_BVEC;
|
|
|
|
|
|
|
|
if (bio_has_data(bio) && uio.uio_loffset + uio.uio_resid >
|
|
|
|
zv->zv_volsize) {
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
printk(KERN_INFO
|
2016-02-06 01:36:07 +00:00
|
|
|
"%s: bad access: offset=%llu, size=%lu\n",
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
zv->zv_disk->disk_name,
|
2016-02-06 01:36:07 +00:00
|
|
|
(long long unsigned)uio.uio_loffset,
|
|
|
|
(long unsigned)uio.uio_resid);
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
error = SET_ERROR(EIO);
|
2015-09-07 16:03:19 +00:00
|
|
|
goto out1;
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
}
|
|
|
|
|
2016-02-06 01:36:07 +00:00
|
|
|
generic_start_io_acct(rw, bio_sectors(bio), &zv->zv_disk->part0);
|
2015-09-07 16:03:19 +00:00
|
|
|
|
|
|
|
if (rw == WRITE) {
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
if (unlikely(zv->zv_flags & ZVOL_RDONLY)) {
|
|
|
|
error = SET_ERROR(EROFS);
|
2015-09-07 16:03:19 +00:00
|
|
|
goto out2;
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2016-08-09 18:22:30 +00:00
|
|
|
if (bio_is_discard(bio) || bio_is_secure_erase(bio)) {
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
error = zvol_discard(bio);
|
2015-09-07 16:03:19 +00:00
|
|
|
goto out2;
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
}
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2016-02-06 01:36:07 +00:00
|
|
|
/*
|
|
|
|
* Some requests are just for flush and nothing else.
|
|
|
|
*/
|
|
|
|
if (uio.uio_resid == 0) {
|
2016-08-09 18:22:30 +00:00
|
|
|
if (bio_is_flush(bio))
|
2016-02-06 01:36:07 +00:00
|
|
|
zil_commit(zv->zv_zilog, ZVOL_OBJ);
|
|
|
|
goto out2;
|
|
|
|
}
|
|
|
|
|
|
|
|
error = zvol_write(zv, &uio,
|
2016-08-09 18:22:30 +00:00
|
|
|
bio_is_flush(bio) || bio_is_fua(bio) ||
|
|
|
|
zv->zv_objset->os_sync == ZFS_SYNC_ALWAYS);
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
} else
|
2016-02-06 01:36:07 +00:00
|
|
|
error = zvol_read(zv, &uio);
|
2011-09-02 13:23:12 +00:00
|
|
|
|
2015-09-07 16:03:19 +00:00
|
|
|
out2:
|
|
|
|
generic_end_io_acct(rw, &zv->zv_disk->part0, start);
|
|
|
|
out1:
|
2015-09-23 15:55:15 +00:00
|
|
|
BIO_END_IO(bio, -error);
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
spl_fstrans_unmark(cookie);
|
|
|
|
#ifdef HAVE_MAKE_REQUEST_FN_RET_INT
|
|
|
|
return (0);
|
2015-11-23 22:47:29 +00:00
|
|
|
#elif defined(HAVE_MAKE_REQUEST_FN_RET_QC)
|
|
|
|
return (BLK_QC_T_NONE);
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
#endif
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
zvol_get_done(zgd_t *zgd, int error)
|
|
|
|
{
|
|
|
|
if (zgd->zgd_db)
|
|
|
|
dmu_buf_rele(zgd->zgd_db, zgd);
|
|
|
|
|
|
|
|
zfs_range_unlock(zgd->zgd_rl);
|
|
|
|
|
|
|
|
if (error == 0 && zgd->zgd_bp)
|
|
|
|
zil_add_block(zgd->zgd_zilog, zgd->zgd_bp);
|
|
|
|
|
|
|
|
kmem_free(zgd, sizeof (zgd_t));
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Get data to generate a TX_WRITE intent log record.
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
zvol_get_data(void *arg, lr_write_t *lr, char *buf, zio_t *zio)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv = arg;
|
|
|
|
objset_t *os = zv->zv_objset;
|
2013-05-10 19:47:54 +00:00
|
|
|
uint64_t object = ZVOL_OBJ;
|
2010-08-26 18:45:02 +00:00
|
|
|
uint64_t offset = lr->lr_offset;
|
|
|
|
uint64_t size = lr->lr_length;
|
2013-05-10 19:47:54 +00:00
|
|
|
blkptr_t *bp = &lr->lr_blkptr;
|
2010-08-26 18:45:02 +00:00
|
|
|
dmu_buf_t *db;
|
|
|
|
zgd_t *zgd;
|
|
|
|
int error;
|
|
|
|
|
|
|
|
ASSERT(zio != NULL);
|
|
|
|
ASSERT(size != 0);
|
|
|
|
|
2014-11-21 00:09:39 +00:00
|
|
|
zgd = (zgd_t *)kmem_zalloc(sizeof (zgd_t), KM_SLEEP);
|
2010-08-26 18:45:02 +00:00
|
|
|
zgd->zgd_zilog = zv->zv_zilog;
|
2016-04-11 21:53:48 +00:00
|
|
|
zgd->zgd_rl = zfs_range_lock(&zv->zv_range_lock, offset, size,
|
|
|
|
RL_READER);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Write records come in two flavors: immediate and indirect.
|
|
|
|
* For small writes it's cheaper to store the data with the
|
|
|
|
* log record (immediate); for large writes it's cheaper to
|
|
|
|
* sync the data and get a pointer to it (indirect) so that
|
|
|
|
* we don't have to write the data twice.
|
|
|
|
*/
|
|
|
|
if (buf != NULL) { /* immediate write */
|
2013-05-10 19:47:54 +00:00
|
|
|
error = dmu_read(os, object, offset, size, buf,
|
2010-08-26 18:45:02 +00:00
|
|
|
DMU_READ_NO_PREFETCH);
|
|
|
|
} else {
|
|
|
|
size = zv->zv_volblocksize;
|
|
|
|
offset = P2ALIGN_TYPED(offset, size, uint64_t);
|
2013-05-10 19:47:54 +00:00
|
|
|
error = dmu_buf_hold(os, object, offset, zgd, &db,
|
2010-08-26 18:45:02 +00:00
|
|
|
DMU_READ_NO_PREFETCH);
|
|
|
|
if (error == 0) {
|
2013-05-10 19:47:54 +00:00
|
|
|
blkptr_t *obp = dmu_buf_get_blkptr(db);
|
|
|
|
if (obp) {
|
|
|
|
ASSERT(BP_IS_HOLE(bp));
|
|
|
|
*bp = *obp;
|
|
|
|
}
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
zgd->zgd_db = db;
|
|
|
|
zgd->zgd_bp = &lr->lr_blkptr;
|
|
|
|
|
|
|
|
ASSERT(db != NULL);
|
|
|
|
ASSERT(db->db_offset == offset);
|
|
|
|
ASSERT(db->db_size == size);
|
|
|
|
|
|
|
|
error = dmu_sync(zio, lr->lr_common.lrc_txg,
|
|
|
|
zvol_get_done, zgd);
|
|
|
|
|
|
|
|
if (error == 0)
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
zvol_get_done(zgd, error);
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(error));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The zvol_state_t's are inserted in increasing MINOR(dev_t) order.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
zvol_insert(zvol_state_t *zv_insert)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv = NULL;
|
|
|
|
|
|
|
|
ASSERT(MUTEX_HELD(&zvol_state_lock));
|
|
|
|
ASSERT3U(MINOR(zv_insert->zv_dev) & ZVOL_MINOR_MASK, ==, 0);
|
|
|
|
for (zv = list_head(&zvol_state_list); zv != NULL;
|
2013-12-12 21:04:40 +00:00
|
|
|
zv = list_next(&zvol_state_list, zv)) {
|
2010-08-26 18:45:02 +00:00
|
|
|
if (MINOR(zv->zv_dev) > MINOR(zv_insert->zv_dev))
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
list_insert_before(&zvol_state_list, zv, zv_insert);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Simply remove the zvol from to list of zvols.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
zvol_remove(zvol_state_t *zv_remove)
|
|
|
|
{
|
|
|
|
ASSERT(MUTEX_HELD(&zvol_state_lock));
|
|
|
|
list_remove(&zvol_state_list, zv_remove);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
zvol_first_open(zvol_state_t *zv)
|
|
|
|
{
|
|
|
|
objset_t *os;
|
|
|
|
uint64_t volsize;
|
|
|
|
int error;
|
|
|
|
uint64_t ro;
|
|
|
|
|
|
|
|
/* lie and say we're read-only */
|
|
|
|
error = dmu_objset_own(zv->zv_name, DMU_OST_ZVOL, 1, zvol_tag, &os);
|
|
|
|
if (error)
|
2015-09-23 16:34:51 +00:00
|
|
|
return (SET_ERROR(-error));
|
|
|
|
|
|
|
|
zv->zv_objset = os;
|
|
|
|
|
|
|
|
error = dsl_prop_get_integer(zv->zv_name, "readonly", &ro, NULL);
|
|
|
|
if (error)
|
|
|
|
goto out_owned;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
error = zap_lookup(os, ZVOL_ZAP_OBJ, "size", 8, 1, &volsize);
|
2015-09-23 16:34:51 +00:00
|
|
|
if (error)
|
|
|
|
goto out_owned;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
error = dmu_bonus_hold(os, ZVOL_OBJ, zvol_tag, &zv->zv_dbuf);
|
2015-09-23 16:34:51 +00:00
|
|
|
if (error)
|
|
|
|
goto out_owned;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
set_capacity(zv->zv_disk, volsize >> 9);
|
|
|
|
zv->zv_volsize = volsize;
|
|
|
|
zv->zv_zilog = zil_open(os, zvol_get_data);
|
|
|
|
|
2013-03-03 05:57:39 +00:00
|
|
|
if (ro || dmu_objset_is_snapshot(os) ||
|
|
|
|
!spa_writeable(dmu_objset_spa(os))) {
|
2013-01-18 17:44:09 +00:00
|
|
|
set_disk_ro(zv->zv_disk, 1);
|
|
|
|
zv->zv_flags |= ZVOL_RDONLY;
|
2010-08-26 18:45:02 +00:00
|
|
|
} else {
|
2013-01-18 17:44:09 +00:00
|
|
|
set_disk_ro(zv->zv_disk, 0);
|
|
|
|
zv->zv_flags &= ~ZVOL_RDONLY;
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2015-09-23 16:34:51 +00:00
|
|
|
out_owned:
|
|
|
|
if (error) {
|
|
|
|
dmu_objset_disown(os, zvol_tag);
|
|
|
|
zv->zv_objset = NULL;
|
|
|
|
}
|
2013-01-18 17:44:09 +00:00
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(-error));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
zvol_last_close(zvol_state_t *zv)
|
|
|
|
{
|
|
|
|
zil_close(zv->zv_zilog);
|
|
|
|
zv->zv_zilog = NULL;
|
2012-08-24 14:12:46 +00:00
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
dmu_buf_rele(zv->zv_dbuf, zvol_tag);
|
|
|
|
zv->zv_dbuf = NULL;
|
2012-08-24 14:12:46 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Evict cached data
|
|
|
|
*/
|
|
|
|
if (dsl_dataset_is_dirty(dmu_objset_ds(zv->zv_objset)) &&
|
|
|
|
!(zv->zv_flags & ZVOL_RDONLY))
|
|
|
|
txg_wait_synced(dmu_objset_pool(zv->zv_objset), 0);
|
|
|
|
(void) dmu_objset_evict_dbufs(zv->zv_objset);
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
dmu_objset_disown(zv->zv_objset, zvol_tag);
|
|
|
|
zv->zv_objset = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
zvol_open(struct block_device *bdev, fmode_t flag)
|
|
|
|
{
|
2016-02-16 19:52:55 +00:00
|
|
|
zvol_state_t *zv;
|
2010-08-26 18:45:02 +00:00
|
|
|
int error = 0, drop_mutex = 0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If the caller is already holding the mutex do not take it
|
2014-03-22 09:07:14 +00:00
|
|
|
* again, this will happen as part of zvol_create_minor_impl().
|
2010-08-26 18:45:02 +00:00
|
|
|
* Once add_disk() is called the device is live and the kernel
|
|
|
|
* will attempt to open it to read the partition information.
|
|
|
|
*/
|
|
|
|
if (!mutex_owned(&zvol_state_lock)) {
|
|
|
|
mutex_enter(&zvol_state_lock);
|
|
|
|
drop_mutex = 1;
|
|
|
|
}
|
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
/*
|
|
|
|
* Obtain a copy of private_data under the lock to make sure
|
|
|
|
* that either the result of zvol_freeg() setting
|
|
|
|
* bdev->bd_disk->private_data to NULL is observed, or zvol_free()
|
|
|
|
* is not called on this zv because of the positive zv_open_count.
|
|
|
|
*/
|
|
|
|
zv = bdev->bd_disk->private_data;
|
|
|
|
if (zv == NULL) {
|
|
|
|
error = -ENXIO;
|
|
|
|
goto out_mutex;
|
|
|
|
}
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
if (zv->zv_open_count == 0) {
|
|
|
|
error = zvol_first_open(zv);
|
|
|
|
if (error)
|
|
|
|
goto out_mutex;
|
|
|
|
}
|
|
|
|
|
2013-12-06 22:20:22 +00:00
|
|
|
if ((flag & FMODE_WRITE) && (zv->zv_flags & ZVOL_RDONLY)) {
|
2010-08-26 18:45:02 +00:00
|
|
|
error = -EROFS;
|
|
|
|
goto out_open_count;
|
|
|
|
}
|
|
|
|
|
|
|
|
zv->zv_open_count++;
|
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
check_disk_change(bdev);
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
out_open_count:
|
|
|
|
if (zv->zv_open_count == 0)
|
|
|
|
zvol_last_close(zv);
|
|
|
|
|
|
|
|
out_mutex:
|
|
|
|
if (drop_mutex)
|
|
|
|
mutex_exit(&zvol_state_lock);
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(error));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2013-06-03 06:58:52 +00:00
|
|
|
#ifdef HAVE_BLOCK_DEVICE_OPERATIONS_RELEASE_VOID
|
|
|
|
static void
|
|
|
|
#else
|
2010-08-26 18:45:02 +00:00
|
|
|
static int
|
2013-06-03 06:58:52 +00:00
|
|
|
#endif
|
2010-08-26 18:45:02 +00:00
|
|
|
zvol_release(struct gendisk *disk, fmode_t mode)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv = disk->private_data;
|
|
|
|
int drop_mutex = 0;
|
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
ASSERT(zv && zv->zv_open_count > 0);
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
if (!mutex_owned(&zvol_state_lock)) {
|
|
|
|
mutex_enter(&zvol_state_lock);
|
|
|
|
drop_mutex = 1;
|
|
|
|
}
|
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
zv->zv_open_count--;
|
|
|
|
if (zv->zv_open_count == 0)
|
|
|
|
zvol_last_close(zv);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
if (drop_mutex)
|
|
|
|
mutex_exit(&zvol_state_lock);
|
|
|
|
|
2013-06-03 06:58:52 +00:00
|
|
|
#ifndef HAVE_BLOCK_DEVICE_OPERATIONS_RELEASE_VOID
|
2010-08-26 18:45:02 +00:00
|
|
|
return (0);
|
2013-06-03 06:58:52 +00:00
|
|
|
#endif
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
zvol_ioctl(struct block_device *bdev, fmode_t mode,
|
2013-12-12 21:04:40 +00:00
|
|
|
unsigned int cmd, unsigned long arg)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
|
|
|
zvol_state_t *zv = bdev->bd_disk->private_data;
|
|
|
|
int error = 0;
|
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
ASSERT(zv && zv->zv_open_count > 0);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
switch (cmd) {
|
|
|
|
case BLKFLSBUF:
|
|
|
|
zil_commit(zv->zv_zilog, ZVOL_OBJ);
|
|
|
|
break;
|
2011-02-22 10:58:44 +00:00
|
|
|
case BLKZNAME:
|
|
|
|
error = copy_to_user((void *)arg, zv->zv_name, MAXNAMELEN);
|
|
|
|
break;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
default:
|
|
|
|
error = -ENOTTY;
|
|
|
|
break;
|
|
|
|
|
|
|
|
}
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(error));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef CONFIG_COMPAT
|
|
|
|
static int
|
|
|
|
zvol_compat_ioctl(struct block_device *bdev, fmode_t mode,
|
2013-12-12 21:04:40 +00:00
|
|
|
unsigned cmd, unsigned long arg)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
2013-12-12 21:04:40 +00:00
|
|
|
return (zvol_ioctl(bdev, mode, cmd, arg));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
#else
|
2013-12-12 21:04:40 +00:00
|
|
|
#define zvol_compat_ioctl NULL
|
2010-08-26 18:45:02 +00:00
|
|
|
#endif
|
|
|
|
|
|
|
|
static int zvol_media_changed(struct gendisk *disk)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv = disk->private_data;
|
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
ASSERT(zv && zv->zv_open_count > 0);
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (zv->zv_changed);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int zvol_revalidate_disk(struct gendisk *disk)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv = disk->private_data;
|
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
ASSERT(zv && zv->zv_open_count > 0);
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
zv->zv_changed = 0;
|
|
|
|
set_capacity(zv->zv_disk, zv->zv_volsize >> 9);
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (0);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Provide a simple virtual geometry for legacy compatibility. For devices
|
|
|
|
* smaller than 1 MiB a small head and sector count is used to allow very
|
|
|
|
* tiny devices. For devices over 1 Mib a standard head and sector count
|
|
|
|
* is used to keep the cylinders count reasonable.
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
zvol_getgeo(struct block_device *bdev, struct hd_geometry *geo)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv = bdev->bd_disk->private_data;
|
2016-02-16 19:52:55 +00:00
|
|
|
sector_t sectors;
|
|
|
|
|
|
|
|
ASSERT(zv && zv->zv_open_count > 0);
|
|
|
|
|
|
|
|
sectors = get_capacity(zv->zv_disk);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
if (sectors > 2048) {
|
|
|
|
geo->heads = 16;
|
|
|
|
geo->sectors = 63;
|
|
|
|
} else {
|
|
|
|
geo->heads = 2;
|
|
|
|
geo->sectors = 4;
|
|
|
|
}
|
|
|
|
|
|
|
|
geo->start = 0;
|
|
|
|
geo->cylinders = sectors / (geo->heads * geo->sectors);
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (0);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static struct kobject *
|
|
|
|
zvol_probe(dev_t dev, int *part, void *arg)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv;
|
|
|
|
struct kobject *kobj;
|
|
|
|
|
|
|
|
mutex_enter(&zvol_state_lock);
|
|
|
|
zv = zvol_find_by_dev(dev);
|
2012-09-24 17:30:18 +00:00
|
|
|
kobj = zv ? get_disk(zv->zv_disk) : NULL;
|
2010-08-26 18:45:02 +00:00
|
|
|
mutex_exit(&zvol_state_lock);
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (kobj);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef HAVE_BDEV_BLOCK_DEVICE_OPERATIONS
|
|
|
|
static struct block_device_operations zvol_ops = {
|
2013-12-12 21:04:40 +00:00
|
|
|
.open = zvol_open,
|
|
|
|
.release = zvol_release,
|
|
|
|
.ioctl = zvol_ioctl,
|
|
|
|
.compat_ioctl = zvol_compat_ioctl,
|
|
|
|
.media_changed = zvol_media_changed,
|
|
|
|
.revalidate_disk = zvol_revalidate_disk,
|
|
|
|
.getgeo = zvol_getgeo,
|
|
|
|
.owner = THIS_MODULE,
|
2010-08-26 18:45:02 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
#else /* HAVE_BDEV_BLOCK_DEVICE_OPERATIONS */
|
|
|
|
|
|
|
|
static int
|
|
|
|
zvol_open_by_inode(struct inode *inode, struct file *file)
|
|
|
|
{
|
2013-12-12 21:04:40 +00:00
|
|
|
return (zvol_open(inode->i_bdev, file->f_mode));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
zvol_release_by_inode(struct inode *inode, struct file *file)
|
|
|
|
{
|
2013-12-12 21:04:40 +00:00
|
|
|
return (zvol_release(inode->i_bdev->bd_disk, file->f_mode));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
zvol_ioctl_by_inode(struct inode *inode, struct file *file,
|
2013-12-12 21:04:40 +00:00
|
|
|
unsigned int cmd, unsigned long arg)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
2010-10-29 19:13:52 +00:00
|
|
|
if (file == NULL || inode == NULL)
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(-EINVAL));
|
|
|
|
|
|
|
|
return (zvol_ioctl(inode->i_bdev, file->f_mode, cmd, arg));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
#ifdef CONFIG_COMPAT
|
2010-08-26 18:45:02 +00:00
|
|
|
static long
|
|
|
|
zvol_compat_ioctl_by_inode(struct file *file,
|
2013-12-12 21:04:40 +00:00
|
|
|
unsigned int cmd, unsigned long arg)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
2010-10-29 19:13:52 +00:00
|
|
|
if (file == NULL)
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(-EINVAL));
|
|
|
|
|
|
|
|
return (zvol_compat_ioctl(file->f_dentry->d_inode->i_bdev,
|
|
|
|
file->f_mode, cmd, arg));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
2013-12-12 21:04:40 +00:00
|
|
|
#else
|
|
|
|
#define zvol_compat_ioctl_by_inode NULL
|
|
|
|
#endif
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
static struct block_device_operations zvol_ops = {
|
2013-12-12 21:04:40 +00:00
|
|
|
.open = zvol_open_by_inode,
|
|
|
|
.release = zvol_release_by_inode,
|
|
|
|
.ioctl = zvol_ioctl_by_inode,
|
|
|
|
.compat_ioctl = zvol_compat_ioctl_by_inode,
|
|
|
|
.media_changed = zvol_media_changed,
|
|
|
|
.revalidate_disk = zvol_revalidate_disk,
|
|
|
|
.getgeo = zvol_getgeo,
|
|
|
|
.owner = THIS_MODULE,
|
2010-08-26 18:45:02 +00:00
|
|
|
};
|
|
|
|
#endif /* HAVE_BDEV_BLOCK_DEVICE_OPERATIONS */
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Allocate memory for a new zvol_state_t and setup the required
|
|
|
|
* request queue and generic disk structures for the block device.
|
|
|
|
*/
|
|
|
|
static zvol_state_t *
|
|
|
|
zvol_alloc(dev_t dev, const char *name)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv;
|
|
|
|
|
2014-11-21 00:09:39 +00:00
|
|
|
zv = kmem_zalloc(sizeof (zvol_state_t), KM_SLEEP);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2013-07-02 18:59:10 +00:00
|
|
|
list_link_init(&zv->zv_next);
|
|
|
|
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
zv->zv_queue = blk_alloc_queue(GFP_ATOMIC);
|
2010-08-26 18:45:02 +00:00
|
|
|
if (zv->zv_queue == NULL)
|
|
|
|
goto out_kmem;
|
|
|
|
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
blk_queue_make_request(zv->zv_queue, zvol_request);
|
2016-08-09 18:22:30 +00:00
|
|
|
blk_queue_set_write_cache(zv->zv_queue, B_TRUE, B_TRUE);
|
2011-09-05 09:11:38 +00:00
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
zv->zv_disk = alloc_disk(ZVOL_MINORS);
|
|
|
|
if (zv->zv_disk == NULL)
|
|
|
|
goto out_queue;
|
|
|
|
|
|
|
|
zv->zv_queue->queuedata = zv;
|
|
|
|
zv->zv_dev = dev;
|
|
|
|
zv->zv_open_count = 0;
|
2011-02-22 10:58:44 +00:00
|
|
|
strlcpy(zv->zv_name, name, MAXNAMELEN);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2016-04-11 21:53:48 +00:00
|
|
|
zfs_rlock_init(&zv->zv_range_lock);
|
2011-02-08 19:29:50 +00:00
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
zv->zv_disk->major = zvol_major;
|
|
|
|
zv->zv_disk->first_minor = (dev & MINORMASK);
|
|
|
|
zv->zv_disk->fops = &zvol_ops;
|
|
|
|
zv->zv_disk->private_data = zv;
|
|
|
|
zv->zv_disk->queue = zv->zv_queue;
|
2011-02-22 10:58:44 +00:00
|
|
|
snprintf(zv->zv_disk->disk_name, DISK_NAME_LEN, "%s%d",
|
|
|
|
ZVOL_DEV_NAME, (dev & MINORMASK));
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (zv);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
out_queue:
|
|
|
|
blk_cleanup_queue(zv->zv_queue);
|
|
|
|
out_kmem:
|
|
|
|
kmem_free(zv, sizeof (zvol_state_t));
|
2013-06-29 11:07:45 +00:00
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (NULL);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Cleanup then free a zvol_state_t which was created by zvol_alloc().
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
zvol_free(zvol_state_t *zv)
|
|
|
|
{
|
2016-02-16 19:52:55 +00:00
|
|
|
ASSERT(MUTEX_HELD(&zvol_state_lock));
|
|
|
|
ASSERT(zv->zv_open_count == 0);
|
|
|
|
|
2016-04-11 21:53:48 +00:00
|
|
|
zfs_rlock_destroy(&zv->zv_range_lock);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
zv->zv_disk->private_data = NULL;
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
del_gendisk(zv->zv_disk);
|
|
|
|
blk_cleanup_queue(zv->zv_queue);
|
|
|
|
put_disk(zv->zv_disk);
|
|
|
|
|
|
|
|
kmem_free(zv, sizeof (zvol_state_t));
|
|
|
|
}
|
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
/*
|
|
|
|
* Create a block device minor node and setup the linkage between it
|
|
|
|
* and the specified volume. Once this function returns the block
|
|
|
|
* device is live and ready for use.
|
|
|
|
*/
|
2010-08-26 18:45:02 +00:00
|
|
|
static int
|
2014-03-22 09:07:14 +00:00
|
|
|
zvol_create_minor_impl(const char *name)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
|
|
|
zvol_state_t *zv;
|
|
|
|
objset_t *os;
|
|
|
|
dmu_object_info_t *doi;
|
|
|
|
uint64_t volsize;
|
2015-08-18 20:51:20 +00:00
|
|
|
uint64_t len;
|
2010-08-26 18:45:02 +00:00
|
|
|
unsigned minor = 0;
|
|
|
|
int error = 0;
|
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
mutex_enter(&zvol_state_lock);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
zv = zvol_find_by_name(name);
|
|
|
|
if (zv) {
|
2013-03-08 18:41:28 +00:00
|
|
|
error = SET_ERROR(EEXIST);
|
2010-08-26 18:45:02 +00:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2014-11-21 00:09:39 +00:00
|
|
|
doi = kmem_alloc(sizeof (dmu_object_info_t), KM_SLEEP);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
error = dmu_objset_own(name, DMU_OST_ZVOL, B_TRUE, zvol_tag, &os);
|
|
|
|
if (error)
|
|
|
|
goto out_doi;
|
|
|
|
|
|
|
|
error = dmu_object_info(os, ZVOL_OBJ, doi);
|
|
|
|
if (error)
|
|
|
|
goto out_dmu_objset_disown;
|
|
|
|
|
|
|
|
error = zap_lookup(os, ZVOL_ZAP_OBJ, "size", 8, 1, &volsize);
|
|
|
|
if (error)
|
|
|
|
goto out_dmu_objset_disown;
|
|
|
|
|
|
|
|
error = zvol_find_minor(&minor);
|
|
|
|
if (error)
|
|
|
|
goto out_dmu_objset_disown;
|
|
|
|
|
|
|
|
zv = zvol_alloc(MKDEV(zvol_major, minor), name);
|
|
|
|
if (zv == NULL) {
|
2013-03-08 18:41:28 +00:00
|
|
|
error = SET_ERROR(EAGAIN);
|
2010-08-26 18:45:02 +00:00
|
|
|
goto out_dmu_objset_disown;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (dmu_objset_is_snapshot(os))
|
|
|
|
zv->zv_flags |= ZVOL_RDONLY;
|
|
|
|
|
|
|
|
zv->zv_volblocksize = doi->doi_data_block_size;
|
|
|
|
zv->zv_volsize = volsize;
|
|
|
|
zv->zv_objset = os;
|
|
|
|
|
|
|
|
set_capacity(zv->zv_disk, zv->zv_volsize >> 9);
|
|
|
|
|
2015-08-28 00:01:59 +00:00
|
|
|
blk_queue_max_hw_sectors(zv->zv_queue, (DMU_MAX_ACCESS / 4) >> 9);
|
2011-09-05 13:15:45 +00:00
|
|
|
blk_queue_max_segments(zv->zv_queue, UINT16_MAX);
|
|
|
|
blk_queue_max_segment_size(zv->zv_queue, UINT_MAX);
|
|
|
|
blk_queue_physical_block_size(zv->zv_queue, zv->zv_volblocksize);
|
|
|
|
blk_queue_io_opt(zv->zv_queue, zv->zv_volblocksize);
|
Limit the number of blocks to discard at once.
The number of blocks that can be discarded in one BLKDISCARD ioctl on a
zvol is currently unlimited. Some applications, such as mkfs, discard
the whole volume at once and they use the maximum possible discard size
to do that. As a result, several gigabytes discard requests are not
uncommon.
Unfortunately, if a large amount of data is allocated in the zvol, ZFS
can be quite slow to process discard requests. This is especially true
if the volblocksize is low (e.g. the 8K default). As a result, very
large discard requests can take a very long time (seconds to minutes
under heavy load) to complete. This can cause a number of problems, most
notably if the zvol is accessed remotely (e.g. via iSCSI), in which case
the client has a high probability of timing out on the request.
This patch solves the issue by adding a new tunable module parameter:
zvol_max_discard_blocks. This indicates the maximum possible range, in
zvol blocks, of one discard operation. It is set by default to 16384
blocks, which appears to be a good tradeoff. Using the default
volblocksize of 8K this is equivalent to 128 MB. When using the maximum
volblocksize of 128K this is equivalent to 2 GB.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #858
2012-07-31 08:45:37 +00:00
|
|
|
blk_queue_max_discard_sectors(zv->zv_queue,
|
|
|
|
(zvol_max_discard_blocks * zv->zv_volblocksize) >> 9);
|
2012-08-01 08:29:59 +00:00
|
|
|
blk_queue_discard_granularity(zv->zv_queue, zv->zv_volblocksize);
|
2011-09-02 13:23:12 +00:00
|
|
|
queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, zv->zv_queue);
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
#ifdef QUEUE_FLAG_NONROT
|
2011-09-05 13:15:45 +00:00
|
|
|
queue_flag_set_unlocked(QUEUE_FLAG_NONROT, zv->zv_queue);
|
|
|
|
#endif
|
2015-08-29 16:49:55 +00:00
|
|
|
#ifdef QUEUE_FLAG_ADD_RANDOM
|
|
|
|
queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, zv->zv_queue);
|
|
|
|
#endif
|
2011-09-05 13:15:45 +00:00
|
|
|
|
2013-03-03 05:57:39 +00:00
|
|
|
if (spa_writeable(dmu_objset_spa(os))) {
|
|
|
|
if (zil_replay_disable)
|
|
|
|
zil_destroy(dmu_objset_zil(os), B_FALSE);
|
|
|
|
else
|
|
|
|
zil_replay(os, zv, zvol_replay_vector);
|
|
|
|
}
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2015-08-18 20:51:20 +00:00
|
|
|
/*
|
|
|
|
* When udev detects the addition of the device it will immediately
|
|
|
|
* invoke blkid(8) to determine the type of content on the device.
|
|
|
|
* Prefetching the blocks commonly scanned by blkid(8) will speed
|
|
|
|
* up this process.
|
|
|
|
*/
|
|
|
|
len = MIN(MAX(zvol_prefetch_bytes, 0), SPA_MAXBLOCKSIZE);
|
|
|
|
if (len > 0) {
|
2015-12-22 01:31:57 +00:00
|
|
|
dmu_prefetch(os, ZVOL_OBJ, 0, 0, len, ZIO_PRIORITY_SYNC_READ);
|
|
|
|
dmu_prefetch(os, ZVOL_OBJ, 0, volsize - len, len,
|
|
|
|
ZIO_PRIORITY_SYNC_READ);
|
2015-08-18 20:51:20 +00:00
|
|
|
}
|
|
|
|
|
2012-11-27 22:02:49 +00:00
|
|
|
zv->zv_objset = NULL;
|
2010-08-26 18:45:02 +00:00
|
|
|
out_dmu_objset_disown:
|
|
|
|
dmu_objset_disown(os, zvol_tag);
|
|
|
|
out_doi:
|
2013-12-12 21:04:40 +00:00
|
|
|
kmem_free(doi, sizeof (dmu_object_info_t));
|
2010-08-26 18:45:02 +00:00
|
|
|
out:
|
|
|
|
|
|
|
|
if (error == 0) {
|
|
|
|
zvol_insert(zv);
|
2016-02-16 19:52:55 +00:00
|
|
|
/*
|
|
|
|
* Drop the lock to prevent deadlock with sys_open() ->
|
|
|
|
* zvol_open(), which first takes bd_disk->bd_mutex and then
|
|
|
|
* takes zvol_state_lock, whereas this code path first takes
|
|
|
|
* zvol_state_lock, and then takes bd_disk->bd_mutex.
|
|
|
|
*/
|
|
|
|
mutex_exit(&zvol_state_lock);
|
2010-08-26 18:45:02 +00:00
|
|
|
add_disk(zv->zv_disk);
|
2014-03-22 09:07:14 +00:00
|
|
|
} else {
|
|
|
|
mutex_exit(&zvol_state_lock);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(error));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2013-12-06 22:20:22 +00:00
|
|
|
/*
|
|
|
|
* Rename a block device minor mode for the specified volume.
|
|
|
|
*/
|
|
|
|
static void
|
2014-03-22 09:07:14 +00:00
|
|
|
zvol_rename_minor(zvol_state_t *zv, const char *newname)
|
2013-12-06 22:20:22 +00:00
|
|
|
{
|
|
|
|
int readonly = get_disk_ro(zv->zv_disk);
|
|
|
|
|
|
|
|
ASSERT(MUTEX_HELD(&zvol_state_lock));
|
|
|
|
|
|
|
|
strlcpy(zv->zv_name, newname, sizeof (zv->zv_name));
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The block device's read-only state is briefly changed causing
|
|
|
|
* a KOBJ_CHANGE uevent to be issued. This ensures udev detects
|
|
|
|
* the name change and fixes the symlinks. This does not change
|
|
|
|
* ZVOL_RDONLY in zv->zv_flags so the actual read-only state never
|
|
|
|
* changes. This would normally be done using kobject_uevent() but
|
|
|
|
* that is a GPL-only symbol which is why we need this workaround.
|
|
|
|
*/
|
|
|
|
set_disk_ro(zv->zv_disk, !readonly);
|
|
|
|
set_disk_ro(zv->zv_disk, readonly);
|
|
|
|
}
|
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Mask errors to continue dmu_objset_find() traversal
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
zvol_create_snap_minor_cb(const char *dsname, void *arg)
|
|
|
|
{
|
|
|
|
const char *name = (const char *)arg;
|
|
|
|
|
2015-09-23 16:34:51 +00:00
|
|
|
ASSERT0(MUTEX_HELD(&spa_namespace_lock));
|
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
/* skip the designated dataset */
|
|
|
|
if (name && strcmp(dsname, name) == 0)
|
|
|
|
return (0);
|
|
|
|
|
|
|
|
/* at this point, the dsname should name a snapshot */
|
|
|
|
if (strchr(dsname, '@') == 0) {
|
|
|
|
dprintf("zvol_create_snap_minor_cb(): "
|
|
|
|
"%s is not a shapshot name\n", dsname);
|
|
|
|
} else {
|
|
|
|
(void) zvol_create_minor_impl(dsname);
|
|
|
|
}
|
|
|
|
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Mask errors to continue dmu_objset_find() traversal
|
|
|
|
*/
|
2010-08-26 18:45:02 +00:00
|
|
|
static int
|
2013-09-04 12:00:57 +00:00
|
|
|
zvol_create_minors_cb(const char *dsname, void *arg)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
2014-03-22 09:07:14 +00:00
|
|
|
uint64_t snapdev;
|
|
|
|
int error;
|
|
|
|
|
2015-09-23 16:34:51 +00:00
|
|
|
ASSERT0(MUTEX_HELD(&spa_namespace_lock));
|
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
error = dsl_prop_get_integer(dsname, "snapdev", &snapdev, NULL);
|
|
|
|
if (error)
|
|
|
|
return (0);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Given the name and the 'snapdev' property, create device minor nodes
|
|
|
|
* with the linkages to zvols/snapshots as needed.
|
|
|
|
* If the name represents a zvol, create a minor node for the zvol, then
|
|
|
|
* check if its snapshots are 'visible', and if so, iterate over the
|
|
|
|
* snapshots and create device minor nodes for those.
|
|
|
|
*/
|
|
|
|
if (strchr(dsname, '@') == 0) {
|
|
|
|
/* create minor for the 'dsname' explicitly */
|
|
|
|
error = zvol_create_minor_impl(dsname);
|
|
|
|
if ((error == 0 || error == EEXIST) &&
|
|
|
|
(snapdev == ZFS_SNAPDEV_VISIBLE)) {
|
|
|
|
fstrans_cookie_t cookie = spl_fstrans_mark();
|
|
|
|
/*
|
|
|
|
* traverse snapshots only, do not traverse children,
|
|
|
|
* and skip the 'dsname'
|
|
|
|
*/
|
|
|
|
error = dmu_objset_find((char *)dsname,
|
|
|
|
zvol_create_snap_minor_cb, (void *)dsname,
|
|
|
|
DS_FIND_SNAPSHOTS);
|
|
|
|
spl_fstrans_unmark(cookie);
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
dprintf("zvol_create_minors_cb(): %s is not a zvol name\n",
|
|
|
|
dsname);
|
|
|
|
}
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2011-02-16 17:40:29 +00:00
|
|
|
return (0);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2014-03-22 09:07:14 +00:00
|
|
|
* Create minors for the specified dataset, including children and snapshots.
|
|
|
|
* Pay attention to the 'snapdev' property and iterate over the snapshots
|
|
|
|
* only if they are 'visible'. This approach allows one to assure that the
|
|
|
|
* snapshot metadata is read from disk only if it is needed.
|
|
|
|
*
|
|
|
|
* The name can represent a dataset to be recursively scanned for zvols and
|
|
|
|
* their snapshots, or a single zvol snapshot. If the name represents a
|
|
|
|
* dataset, the scan is performed in two nested stages:
|
|
|
|
* - scan the dataset for zvols, and
|
|
|
|
* - for each zvol, create a minor node, then check if the zvol's snapshots
|
|
|
|
* are 'visible', and only then iterate over the snapshots if needed
|
|
|
|
*
|
|
|
|
* If the name represents a snapshot, a check is perfromed if the snapshot is
|
|
|
|
* 'visible' (which also verifies that the parent is a zvol), and if so,
|
|
|
|
* a minor node for that snapshot is created.
|
2010-08-26 18:45:02 +00:00
|
|
|
*/
|
2014-03-22 09:07:14 +00:00
|
|
|
static int
|
|
|
|
zvol_create_minors_impl(const char *name)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
|
|
|
int error = 0;
|
2016-02-16 19:52:55 +00:00
|
|
|
fstrans_cookie_t cookie;
|
2014-03-22 09:07:14 +00:00
|
|
|
char *atp, *parent;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
if (zvol_inhibit_dev)
|
|
|
|
return (0);
|
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
parent = kmem_alloc(MAXPATHLEN, KM_SLEEP);
|
|
|
|
(void) strlcpy(parent, name, MAXPATHLEN);
|
|
|
|
|
|
|
|
if ((atp = strrchr(parent, '@')) != NULL) {
|
|
|
|
uint64_t snapdev;
|
|
|
|
|
|
|
|
*atp = '\0';
|
|
|
|
error = dsl_prop_get_integer(parent, "snapdev",
|
|
|
|
&snapdev, NULL);
|
|
|
|
|
|
|
|
if (error == 0 && snapdev == ZFS_SNAPDEV_VISIBLE)
|
|
|
|
error = zvol_create_minor_impl(name);
|
|
|
|
} else {
|
|
|
|
cookie = spl_fstrans_mark();
|
|
|
|
error = dmu_objset_find(parent, zvol_create_minors_cb,
|
|
|
|
NULL, DS_FIND_CHILDREN);
|
|
|
|
spl_fstrans_unmark(cookie);
|
|
|
|
}
|
|
|
|
|
|
|
|
kmem_free(parent, MAXPATHLEN);
|
2013-12-06 22:20:22 +00:00
|
|
|
|
|
|
|
return (SET_ERROR(error));
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Remove minors for specified dataset including children and snapshots.
|
|
|
|
*/
|
2014-03-22 09:07:14 +00:00
|
|
|
static void
|
|
|
|
zvol_remove_minors_impl(const char *name)
|
2013-12-06 22:20:22 +00:00
|
|
|
{
|
|
|
|
zvol_state_t *zv, *zv_next;
|
|
|
|
int namelen = ((name) ? strlen(name) : 0);
|
|
|
|
|
2012-06-02 01:49:10 +00:00
|
|
|
if (zvol_inhibit_dev)
|
2013-12-06 22:20:22 +00:00
|
|
|
return;
|
2012-06-02 01:49:10 +00:00
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
mutex_enter(&zvol_state_lock);
|
2013-12-06 22:20:22 +00:00
|
|
|
|
|
|
|
for (zv = list_head(&zvol_state_list); zv != NULL; zv = zv_next) {
|
|
|
|
zv_next = list_next(&zvol_state_list, zv);
|
|
|
|
|
|
|
|
if (name == NULL || strcmp(zv->zv_name, name) == 0 ||
|
|
|
|
(strncmp(zv->zv_name, name, namelen) == 0 &&
|
2016-02-16 19:52:55 +00:00
|
|
|
(zv->zv_name[namelen] == '/' ||
|
|
|
|
zv->zv_name[namelen] == '@'))) {
|
|
|
|
|
|
|
|
/* If in use, leave alone */
|
|
|
|
if (zv->zv_open_count > 0)
|
|
|
|
continue;
|
|
|
|
|
2013-12-06 22:20:22 +00:00
|
|
|
zvol_remove(zv);
|
|
|
|
zvol_free(zv);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-12-06 22:20:22 +00:00
|
|
|
mutex_exit(&zvol_state_lock);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
/* Remove minor for this specific snapshot only */
|
|
|
|
static void
|
|
|
|
zvol_remove_minor_impl(const char *name)
|
|
|
|
{
|
|
|
|
zvol_state_t *zv, *zv_next;
|
|
|
|
|
|
|
|
if (zvol_inhibit_dev)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (strchr(name, '@') == NULL)
|
|
|
|
return;
|
|
|
|
|
|
|
|
mutex_enter(&zvol_state_lock);
|
|
|
|
|
|
|
|
for (zv = list_head(&zvol_state_list); zv != NULL; zv = zv_next) {
|
|
|
|
zv_next = list_next(&zvol_state_list, zv);
|
|
|
|
|
|
|
|
if (strcmp(zv->zv_name, name) == 0) {
|
|
|
|
/* If in use, leave alone */
|
|
|
|
if (zv->zv_open_count > 0)
|
|
|
|
continue;
|
|
|
|
zvol_remove(zv);
|
|
|
|
zvol_free(zv);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
mutex_exit(&zvol_state_lock);
|
|
|
|
}
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
/*
|
2013-12-06 22:20:22 +00:00
|
|
|
* Rename minors for specified dataset including children and snapshots.
|
2010-08-26 18:45:02 +00:00
|
|
|
*/
|
2014-03-22 09:07:14 +00:00
|
|
|
static void
|
|
|
|
zvol_rename_minors_impl(const char *oldname, const char *newname)
|
2010-08-26 18:45:02 +00:00
|
|
|
{
|
|
|
|
zvol_state_t *zv, *zv_next;
|
2013-12-06 22:20:22 +00:00
|
|
|
int oldnamelen, newnamelen;
|
|
|
|
char *name;
|
2010-08-26 18:45:02 +00:00
|
|
|
|
2012-06-02 01:49:10 +00:00
|
|
|
if (zvol_inhibit_dev)
|
|
|
|
return;
|
|
|
|
|
2013-12-06 22:20:22 +00:00
|
|
|
oldnamelen = strlen(oldname);
|
|
|
|
newnamelen = strlen(newname);
|
2014-11-21 00:09:39 +00:00
|
|
|
name = kmem_alloc(MAXNAMELEN, KM_SLEEP);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
mutex_enter(&zvol_state_lock);
|
2013-12-06 22:20:22 +00:00
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
for (zv = list_head(&zvol_state_list); zv != NULL; zv = zv_next) {
|
|
|
|
zv_next = list_next(&zvol_state_list, zv);
|
|
|
|
|
2016-02-16 19:52:55 +00:00
|
|
|
/* If in use, leave alone */
|
|
|
|
if (zv->zv_open_count > 0)
|
|
|
|
continue;
|
|
|
|
|
2013-12-06 22:20:22 +00:00
|
|
|
if (strcmp(zv->zv_name, oldname) == 0) {
|
2014-03-22 09:07:14 +00:00
|
|
|
zvol_rename_minor(zv, newname);
|
2013-12-06 22:20:22 +00:00
|
|
|
} else if (strncmp(zv->zv_name, oldname, oldnamelen) == 0 &&
|
|
|
|
(zv->zv_name[oldnamelen] == '/' ||
|
|
|
|
zv->zv_name[oldnamelen] == '@')) {
|
|
|
|
snprintf(name, MAXNAMELEN, "%s%c%s", newname,
|
|
|
|
zv->zv_name[oldnamelen],
|
|
|
|
zv->zv_name + oldnamelen + 1);
|
2014-03-22 09:07:14 +00:00
|
|
|
zvol_rename_minor(zv, name);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
}
|
2013-12-06 22:20:22 +00:00
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
mutex_exit(&zvol_state_lock);
|
2013-12-06 22:20:22 +00:00
|
|
|
|
|
|
|
kmem_free(name, MAXNAMELEN);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
typedef struct zvol_snapdev_cb_arg {
|
|
|
|
uint64_t snapdev;
|
|
|
|
} zvol_snapdev_cb_arg_t;
|
|
|
|
|
2013-02-13 23:11:59 +00:00
|
|
|
static int
|
2014-03-22 09:07:14 +00:00
|
|
|
zvol_set_snapdev_cb(const char *dsname, void *param) {
|
|
|
|
zvol_snapdev_cb_arg_t *arg = param;
|
2013-02-13 23:11:59 +00:00
|
|
|
|
|
|
|
if (strchr(dsname, '@') == NULL)
|
2013-12-06 22:20:22 +00:00
|
|
|
return (0);
|
2013-02-13 23:11:59 +00:00
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
switch (arg->snapdev) {
|
2013-02-13 23:11:59 +00:00
|
|
|
case ZFS_SNAPDEV_VISIBLE:
|
2014-03-22 09:07:14 +00:00
|
|
|
(void) zvol_create_minor_impl(dsname);
|
2013-02-13 23:11:59 +00:00
|
|
|
break;
|
|
|
|
case ZFS_SNAPDEV_HIDDEN:
|
2014-03-22 09:07:14 +00:00
|
|
|
(void) zvol_remove_minor_impl(dsname);
|
2013-02-13 23:11:59 +00:00
|
|
|
break;
|
|
|
|
}
|
2013-12-06 22:20:22 +00:00
|
|
|
|
|
|
|
return (0);
|
2013-02-13 23:11:59 +00:00
|
|
|
}
|
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
static void
|
|
|
|
zvol_set_snapdev_impl(char *name, uint64_t snapdev)
|
|
|
|
{
|
|
|
|
zvol_snapdev_cb_arg_t arg = {snapdev};
|
|
|
|
fstrans_cookie_t cookie = spl_fstrans_mark();
|
|
|
|
/*
|
|
|
|
* The zvol_set_snapdev_sync() sets snapdev appropriately
|
|
|
|
* in the dataset hierarchy. Here, we only scan snapshots.
|
|
|
|
*/
|
|
|
|
dmu_objset_find(name, zvol_set_snapdev_cb, &arg, DS_FIND_SNAPSHOTS);
|
|
|
|
spl_fstrans_unmark(cookie);
|
|
|
|
}
|
|
|
|
|
|
|
|
static zvol_task_t *
|
|
|
|
zvol_task_alloc(zvol_async_op_t op, const char *name1, const char *name2,
|
|
|
|
uint64_t snapdev)
|
|
|
|
{
|
|
|
|
zvol_task_t *task;
|
|
|
|
char *delim;
|
|
|
|
|
|
|
|
/* Never allow tasks on hidden names. */
|
|
|
|
if (name1[0] == '$')
|
|
|
|
return (NULL);
|
|
|
|
|
|
|
|
task = kmem_zalloc(sizeof (zvol_task_t), KM_SLEEP);
|
|
|
|
task->op = op;
|
|
|
|
task->snapdev = snapdev;
|
|
|
|
delim = strchr(name1, '/');
|
|
|
|
strlcpy(task->pool, name1, delim ? (delim - name1 + 1) : MAXNAMELEN);
|
|
|
|
|
|
|
|
strlcpy(task->name1, name1, MAXNAMELEN);
|
|
|
|
if (name2 != NULL)
|
|
|
|
strlcpy(task->name2, name2, MAXNAMELEN);
|
|
|
|
|
|
|
|
return (task);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
zvol_task_free(zvol_task_t *task)
|
|
|
|
{
|
|
|
|
kmem_free(task, sizeof (zvol_task_t));
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The worker thread function performed asynchronously.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
zvol_task_cb(void *param)
|
|
|
|
{
|
|
|
|
zvol_task_t *task = (zvol_task_t *)param;
|
|
|
|
|
|
|
|
switch (task->op) {
|
|
|
|
case ZVOL_ASYNC_CREATE_MINORS:
|
|
|
|
(void) zvol_create_minors_impl(task->name1);
|
|
|
|
break;
|
|
|
|
case ZVOL_ASYNC_REMOVE_MINORS:
|
|
|
|
zvol_remove_minors_impl(task->name1);
|
|
|
|
break;
|
|
|
|
case ZVOL_ASYNC_RENAME_MINORS:
|
|
|
|
zvol_rename_minors_impl(task->name1, task->name2);
|
|
|
|
break;
|
|
|
|
case ZVOL_ASYNC_SET_SNAPDEV:
|
|
|
|
zvol_set_snapdev_impl(task->name1, task->snapdev);
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
VERIFY(0);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
zvol_task_free(task);
|
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct zvol_set_snapdev_arg {
|
|
|
|
const char *zsda_name;
|
|
|
|
uint64_t zsda_value;
|
|
|
|
zprop_source_t zsda_source;
|
|
|
|
dmu_tx_t *zsda_tx;
|
|
|
|
} zvol_set_snapdev_arg_t;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Sanity check the dataset for safe use by the sync task. No additional
|
|
|
|
* conditions are imposed.
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
zvol_set_snapdev_check(void *arg, dmu_tx_t *tx)
|
|
|
|
{
|
|
|
|
zvol_set_snapdev_arg_t *zsda = arg;
|
|
|
|
dsl_pool_t *dp = dmu_tx_pool(tx);
|
|
|
|
dsl_dir_t *dd;
|
|
|
|
int error;
|
|
|
|
|
|
|
|
error = dsl_dir_hold(dp, zsda->zsda_name, FTAG, &dd, NULL);
|
|
|
|
if (error != 0)
|
|
|
|
return (error);
|
|
|
|
|
|
|
|
dsl_dir_rele(dd, FTAG);
|
|
|
|
|
|
|
|
return (error);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
zvol_set_snapdev_sync_cb(dsl_pool_t *dp, dsl_dataset_t *ds, void *arg)
|
|
|
|
{
|
|
|
|
zvol_set_snapdev_arg_t *zsda = arg;
|
|
|
|
char dsname[MAXNAMELEN];
|
|
|
|
zvol_task_t *task;
|
|
|
|
|
|
|
|
dsl_dataset_name(ds, dsname);
|
|
|
|
dsl_prop_set_sync_impl(ds, zfs_prop_to_name(ZFS_PROP_SNAPDEV),
|
|
|
|
zsda->zsda_source, sizeof (zsda->zsda_value), 1,
|
|
|
|
&zsda->zsda_value, zsda->zsda_tx);
|
|
|
|
|
|
|
|
task = zvol_task_alloc(ZVOL_ASYNC_SET_SNAPDEV, dsname,
|
|
|
|
NULL, zsda->zsda_value);
|
|
|
|
if (task == NULL)
|
|
|
|
return (0);
|
|
|
|
|
|
|
|
(void) taskq_dispatch(dp->dp_spa->spa_zvol_taskq, zvol_task_cb,
|
|
|
|
task, TQ_SLEEP);
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Traverse all child snapshot datasets and apply snapdev appropriately.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
zvol_set_snapdev_sync(void *arg, dmu_tx_t *tx)
|
|
|
|
{
|
|
|
|
zvol_set_snapdev_arg_t *zsda = arg;
|
|
|
|
dsl_pool_t *dp = dmu_tx_pool(tx);
|
|
|
|
dsl_dir_t *dd;
|
|
|
|
|
|
|
|
VERIFY0(dsl_dir_hold(dp, zsda->zsda_name, FTAG, &dd, NULL));
|
|
|
|
zsda->zsda_tx = tx;
|
|
|
|
|
|
|
|
dmu_objset_find_dp(dp, dd->dd_object, zvol_set_snapdev_sync_cb,
|
|
|
|
zsda, DS_FIND_CHILDREN);
|
|
|
|
|
|
|
|
dsl_dir_rele(dd, FTAG);
|
|
|
|
}
|
|
|
|
|
2013-02-13 23:11:59 +00:00
|
|
|
int
|
2014-03-22 09:07:14 +00:00
|
|
|
zvol_set_snapdev(const char *ddname, zprop_source_t source, uint64_t snapdev)
|
|
|
|
{
|
|
|
|
zvol_set_snapdev_arg_t zsda;
|
2016-02-16 19:52:55 +00:00
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
zsda.zsda_name = ddname;
|
|
|
|
zsda.zsda_source = source;
|
|
|
|
zsda.zsda_value = snapdev;
|
2016-02-16 19:52:55 +00:00
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
return (dsl_sync_task(ddname, zvol_set_snapdev_check,
|
|
|
|
zvol_set_snapdev_sync, &zsda, 0, ZFS_SPACE_CHECK_NONE));
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
zvol_create_minors(spa_t *spa, const char *name, boolean_t async)
|
|
|
|
{
|
|
|
|
zvol_task_t *task;
|
|
|
|
taskqid_t id;
|
|
|
|
|
|
|
|
task = zvol_task_alloc(ZVOL_ASYNC_CREATE_MINORS, name, NULL, ~0ULL);
|
|
|
|
if (task == NULL)
|
|
|
|
return;
|
|
|
|
|
|
|
|
id = taskq_dispatch(spa->spa_zvol_taskq, zvol_task_cb, task, TQ_SLEEP);
|
|
|
|
if ((async == B_FALSE) && (id != 0))
|
|
|
|
taskq_wait_id(spa->spa_zvol_taskq, id);
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
zvol_remove_minors(spa_t *spa, const char *name, boolean_t async)
|
|
|
|
{
|
|
|
|
zvol_task_t *task;
|
|
|
|
taskqid_t id;
|
|
|
|
|
|
|
|
task = zvol_task_alloc(ZVOL_ASYNC_REMOVE_MINORS, name, NULL, ~0ULL);
|
|
|
|
if (task == NULL)
|
|
|
|
return;
|
2016-02-16 19:52:55 +00:00
|
|
|
|
2014-03-22 09:07:14 +00:00
|
|
|
id = taskq_dispatch(spa->spa_zvol_taskq, zvol_task_cb, task, TQ_SLEEP);
|
|
|
|
if ((async == B_FALSE) && (id != 0))
|
|
|
|
taskq_wait_id(spa->spa_zvol_taskq, id);
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
zvol_rename_minors(spa_t *spa, const char *name1, const char *name2,
|
|
|
|
boolean_t async)
|
|
|
|
{
|
|
|
|
zvol_task_t *task;
|
|
|
|
taskqid_t id;
|
|
|
|
|
|
|
|
task = zvol_task_alloc(ZVOL_ASYNC_RENAME_MINORS, name1, name2, ~0ULL);
|
|
|
|
if (task == NULL)
|
|
|
|
return;
|
|
|
|
|
|
|
|
id = taskq_dispatch(spa->spa_zvol_taskq, zvol_task_cb, task, TQ_SLEEP);
|
|
|
|
if ((async == B_FALSE) && (id != 0))
|
|
|
|
taskq_wait_id(spa->spa_zvol_taskq, id);
|
2013-02-13 23:11:59 +00:00
|
|
|
}
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
int
|
|
|
|
zvol_init(void)
|
|
|
|
{
|
|
|
|
int error;
|
|
|
|
|
2013-07-02 18:59:10 +00:00
|
|
|
list_create(&zvol_state_list, sizeof (zvol_state_t),
|
2013-12-12 21:04:40 +00:00
|
|
|
offsetof(zvol_state_t, zv_next));
|
2013-07-02 18:59:10 +00:00
|
|
|
mutex_init(&zvol_state_lock, NULL, MUTEX_DEFAULT, NULL);
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
error = register_blkdev(zvol_major, ZVOL_DRIVER);
|
|
|
|
if (error) {
|
|
|
|
printk(KERN_INFO "ZFS: register_blkdev() failed %d\n", error);
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
goto out;
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
blk_register_region(MKDEV(zvol_major, 0), 1UL << MINORBITS,
|
2013-12-12 21:04:40 +00:00
|
|
|
THIS_MODULE, zvol_probe, NULL, NULL);
|
2010-08-26 18:45:02 +00:00
|
|
|
|
|
|
|
return (0);
|
2013-07-02 18:59:10 +00:00
|
|
|
|
zvol processing should use struct bio
Internally, zvols are files exposed through the block device API. This
is intended to reduce overhead when things require block devices.
However, the ZoL zvol code emulates a traditional block device in that
it has a top half and a bottom half. This is an unnecessary source of
overhead that does not exist on any other OpenZFS platform does this.
This patch removes it. Early users of this patch reported double digit
performance gains in IOPS on zvols in the range of 50% to 80%.
Comments in the code suggest that the current implementation was done to
obtain IO merging from Linux's IO elevator. However, the DMU already
does write merging while arc_read() should implicitly merge read IOs
because only 1 thread is permitted to fetch the buffer into ARC. In
addition, commercial ZFSOnLinux distributions report that regular files
are more performant than zvols under the current implementation, and the
main consumers of zvols are VMs and iSCSI targets, which have their own
elevators to merge IOs.
Some minor refactoring allows us to register zfs_request() as our
->make_request() handler in place of the generic_make_request()
function. This eliminates the layer of code that broke IO requests on
zvols into a top half and a bottom half. This has several benefits:
1. No per zvol spinlocks.
2. No redundant IO elevator processing.
3. Interrupts are disabled only when actually necessary.
4. No redispatching of IOs when all taskq threads are busy.
5. Linux's page out routines will properly block.
6. Many autotools checks become obsolete.
An unfortunate consequence of eliminating the layer that
generic_make_request() is that we no longer calls the instrumentation
hooks for block IO accounting. Those hooks are GPL-exported, so we
cannot call them ourselves and consequently, we lose the ability to do
IO monitoring via iostat. Since zvols are internally files mapped as
block devices, this should be okay. Anyone who is willing to accept the
performance penalty for the block IO layer's accounting could use the
loop device in between the zvol and its consumer. Alternatively, perf
and ftrace likely could be used. Also, tools like latencytop will still
work. Tools such as latencytop sometimes provide a better view of
performance bottlenecks than the traditional block IO accounting tools
do.
Lastly, if direct reclaim occurs during spacemap loading and swap is on
a zvol, this code will deadlock. That deadlock could already occur with
sync=always on zvols. Given that swap on zvols is not yet production
ready, this is not a blocker.
Signed-off-by: Richard Yao <ryao@gentoo.org>
2014-07-04 22:43:47 +00:00
|
|
|
out:
|
2013-07-02 18:59:10 +00:00
|
|
|
mutex_destroy(&zvol_state_lock);
|
|
|
|
list_destroy(&zvol_state_list);
|
|
|
|
|
2013-12-12 21:04:40 +00:00
|
|
|
return (SET_ERROR(error));
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
zvol_fini(void)
|
|
|
|
{
|
2014-03-22 09:07:14 +00:00
|
|
|
zvol_remove_minors_impl(NULL);
|
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
blk_unregister_region(MKDEV(zvol_major, 0), 1UL << MINORBITS);
|
|
|
|
unregister_blkdev(zvol_major, ZVOL_DRIVER);
|
2014-03-22 09:07:14 +00:00
|
|
|
|
2010-08-26 18:45:02 +00:00
|
|
|
list_destroy(&zvol_state_list);
|
2014-03-22 09:07:14 +00:00
|
|
|
mutex_destroy(&zvol_state_lock);
|
2010-08-26 18:45:02 +00:00
|
|
|
}
|
|
|
|
|
2012-06-02 01:49:10 +00:00
|
|
|
module_param(zvol_inhibit_dev, uint, 0644);
|
|
|
|
MODULE_PARM_DESC(zvol_inhibit_dev, "Do not create zvol device nodes");
|
|
|
|
|
2011-12-07 17:23:44 +00:00
|
|
|
module_param(zvol_major, uint, 0444);
|
2010-08-26 18:45:02 +00:00
|
|
|
MODULE_PARM_DESC(zvol_major, "Major number for zvol device");
|
|
|
|
|
Limit the number of blocks to discard at once.
The number of blocks that can be discarded in one BLKDISCARD ioctl on a
zvol is currently unlimited. Some applications, such as mkfs, discard
the whole volume at once and they use the maximum possible discard size
to do that. As a result, several gigabytes discard requests are not
uncommon.
Unfortunately, if a large amount of data is allocated in the zvol, ZFS
can be quite slow to process discard requests. This is especially true
if the volblocksize is low (e.g. the 8K default). As a result, very
large discard requests can take a very long time (seconds to minutes
under heavy load) to complete. This can cause a number of problems, most
notably if the zvol is accessed remotely (e.g. via iSCSI), in which case
the client has a high probability of timing out on the request.
This patch solves the issue by adding a new tunable module parameter:
zvol_max_discard_blocks. This indicates the maximum possible range, in
zvol blocks, of one discard operation. It is set by default to 16384
blocks, which appears to be a good tradeoff. Using the default
volblocksize of 8K this is equivalent to 128 MB. When using the maximum
volblocksize of 128K this is equivalent to 2 GB.
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #858
2012-07-31 08:45:37 +00:00
|
|
|
module_param(zvol_max_discard_blocks, ulong, 0444);
|
2013-12-12 21:04:40 +00:00
|
|
|
MODULE_PARM_DESC(zvol_max_discard_blocks, "Max number of blocks to discard");
|
2015-08-18 20:51:20 +00:00
|
|
|
|
|
|
|
module_param(zvol_prefetch_bytes, uint, 0644);
|
|
|
|
MODULE_PARM_DESC(zvol_prefetch_bytes, "Prefetch N bytes at zvol start+end");
|