1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265
|
// SPDX-License-Identifier: CDDL-1.0
/*
* 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 https://opensource.org/licenses/CDDL-1.0.
* 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) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2017 by Delphix. All rights reserved.
*/
#include <sys/dmu.h>
#include <sys/dmu_tx.h>
#include <sys/dsl_pool.h>
#include <sys/dsl_dir.h>
#include <sys/dsl_synctask.h>
#include <sys/metaslab.h>
#define DST_AVG_BLKSHIFT 14
static int
dsl_null_checkfunc(void *arg, dmu_tx_t *tx)
{
(void) arg, (void) tx;
return (0);
}
static int
dsl_sync_task_common(const char *pool, dsl_checkfunc_t *checkfunc,
dsl_syncfunc_t *syncfunc, dsl_sigfunc_t *sigfunc, void *arg,
int blocks_modified, zfs_space_check_t space_check, boolean_t early)
{
spa_t *spa;
dmu_tx_t *tx;
int err;
dsl_sync_task_t dst = { { { NULL } } };
dsl_pool_t *dp;
err = spa_open(pool, &spa, FTAG);
if (err != 0)
return (err);
dp = spa_get_dsl(spa);
top:
tx = dmu_tx_create_dd(dp->dp_mos_dir);
VERIFY0(dmu_tx_assign(tx, DMU_TX_WAIT));
dst.dst_pool = dp;
dst.dst_txg = dmu_tx_get_txg(tx);
dst.dst_space = blocks_modified << DST_AVG_BLKSHIFT;
dst.dst_space_check = space_check;
dst.dst_checkfunc = checkfunc != NULL ? checkfunc : dsl_null_checkfunc;
dst.dst_syncfunc = syncfunc;
dst.dst_arg = arg;
dst.dst_error = 0;
dst.dst_nowaiter = B_FALSE;
dsl_pool_config_enter(dp, FTAG);
err = dst.dst_checkfunc(arg, tx);
dsl_pool_config_exit(dp, FTAG);
if (err != 0) {
dmu_tx_commit(tx);
spa_close(spa, FTAG);
return (err);
}
txg_list_t *task_list = (early) ?
&dp->dp_early_sync_tasks : &dp->dp_sync_tasks;
VERIFY(txg_list_add_tail(task_list, &dst, dst.dst_txg));
dmu_tx_commit(tx);
if (sigfunc != NULL) {
err = txg_wait_synced_flags(dp, dst.dst_txg, TXG_WAIT_SIGNAL);
if (err != 0) {
VERIFY3U(err, ==, EINTR);
/* current contract is to call func once */
sigfunc(arg, tx);
/* in case we're performing an EAGAIN retry */
sigfunc = NULL;
txg_wait_synced(dp, dst.dst_txg);
}
} else
txg_wait_synced(dp, dst.dst_txg);
if (dst.dst_error == EAGAIN) {
txg_wait_synced(dp, dst.dst_txg + TXG_DEFER_SIZE);
goto top;
}
spa_close(spa, FTAG);
return (dst.dst_error);
}
/*
* Called from open context to perform a callback in syncing context. Waits
* for the operation to complete.
*
* The checkfunc will be called from open context as a preliminary check
* which can quickly fail. If it succeeds, it will be called again from
* syncing context. The checkfunc should generally be designed to work
* properly in either context, but if necessary it can check
* dmu_tx_is_syncing(tx).
*
* The synctask infrastructure enforces proper locking strategy with respect
* to the dp_config_rwlock -- the lock will always be held when the callbacks
* are called. It will be held for read during the open-context (preliminary)
* call to the checkfunc, and then held for write from syncing context during
* the calls to the check and sync funcs.
*
* A dataset or pool name can be passed as the first argument. Typically,
* the check func will hold, check the return value of the hold, and then
* release the dataset. The sync func will VERIFYO(hold()) the dataset.
* This is safe because no changes can be made between the check and sync funcs,
* and the sync func will only be called if the check func successfully opened
* the dataset.
*/
int
dsl_sync_task(const char *pool, dsl_checkfunc_t *checkfunc,
dsl_syncfunc_t *syncfunc, void *arg,
int blocks_modified, zfs_space_check_t space_check)
{
return (dsl_sync_task_common(pool, checkfunc, syncfunc, NULL, arg,
blocks_modified, space_check, B_FALSE));
}
/*
* An early synctask works exactly as a standard synctask with one important
* difference on the way it is handled during syncing context. Standard
* synctasks run after we've written out all the dirty blocks of dirty
* datasets. Early synctasks are executed before writing out any dirty data,
* and thus before standard synctasks.
*
* For that reason, early synctasks can affect the process of writing dirty
* changes to disk for the txg that they run and should be used with caution.
* In addition, early synctasks should not dirty any metaslabs as this would
* invalidate the precondition/invariant for subsequent early synctasks.
* [see dsl_pool_sync() and dsl_early_sync_task_verify()]
*/
int
dsl_early_sync_task(const char *pool, dsl_checkfunc_t *checkfunc,
dsl_syncfunc_t *syncfunc, void *arg,
int blocks_modified, zfs_space_check_t space_check)
{
return (dsl_sync_task_common(pool, checkfunc, syncfunc, NULL, arg,
blocks_modified, space_check, B_TRUE));
}
/*
* A standard synctask that can be interrupted from a signal. The sigfunc
* is called once if a signal occurred while waiting for the task to sync.
*/
int
dsl_sync_task_sig(const char *pool, dsl_checkfunc_t *checkfunc,
dsl_syncfunc_t *syncfunc, dsl_sigfunc_t *sigfunc, void *arg,
int blocks_modified, zfs_space_check_t space_check)
{
return (dsl_sync_task_common(pool, checkfunc, syncfunc, sigfunc, arg,
blocks_modified, space_check, B_FALSE));
}
static void
dsl_sync_task_nowait_common(dsl_pool_t *dp, dsl_syncfunc_t *syncfunc, void *arg,
dmu_tx_t *tx, boolean_t early)
{
dsl_sync_task_t *dst = kmem_zalloc(sizeof (*dst), KM_SLEEP);
dst->dst_pool = dp;
dst->dst_txg = dmu_tx_get_txg(tx);
dst->dst_space_check = ZFS_SPACE_CHECK_NONE;
dst->dst_checkfunc = dsl_null_checkfunc;
dst->dst_syncfunc = syncfunc;
dst->dst_arg = arg;
dst->dst_error = 0;
dst->dst_nowaiter = B_TRUE;
txg_list_t *task_list = (early) ?
&dp->dp_early_sync_tasks : &dp->dp_sync_tasks;
VERIFY(txg_list_add_tail(task_list, dst, dst->dst_txg));
}
void
dsl_sync_task_nowait(dsl_pool_t *dp, dsl_syncfunc_t *syncfunc, void *arg,
dmu_tx_t *tx)
{
dsl_sync_task_nowait_common(dp, syncfunc, arg, tx, B_FALSE);
}
void
dsl_early_sync_task_nowait(dsl_pool_t *dp, dsl_syncfunc_t *syncfunc, void *arg,
dmu_tx_t *tx)
{
dsl_sync_task_nowait_common(dp, syncfunc, arg, tx, B_TRUE);
}
/*
* Called in syncing context to execute the synctask.
*/
void
dsl_sync_task_sync(dsl_sync_task_t *dst, dmu_tx_t *tx)
{
dsl_pool_t *dp = dst->dst_pool;
ASSERT0(dst->dst_error);
/*
* Check for sufficient space.
*
* When the sync task was created, the caller specified the
* type of space checking required. See the comment in
* zfs_space_check_t for details on the semantics of each
* type of space checking.
*
* We just check against what's on-disk; we don't want any
* in-flight accounting to get in our way, because open context
* may have already used up various in-core limits
* (arc_tempreserve, dsl_pool_tempreserve).
*/
if (dst->dst_space_check != ZFS_SPACE_CHECK_NONE) {
uint64_t quota = dsl_pool_unreserved_space(dp,
dst->dst_space_check);
uint64_t used = dsl_dir_phys(dp->dp_root_dir)->dd_used_bytes;
/* MOS space is triple-dittoed, so we multiply by 3. */
if (used + dst->dst_space * 3 > quota) {
dst->dst_error = SET_ERROR(ENOSPC);
if (dst->dst_nowaiter)
kmem_free(dst, sizeof (*dst));
return;
}
}
/*
* Check for errors by calling checkfunc.
*/
rrw_enter(&dp->dp_config_rwlock, RW_WRITER, FTAG);
dst->dst_error = dst->dst_checkfunc(dst->dst_arg, tx);
if (dst->dst_error == 0)
dst->dst_syncfunc(dst->dst_arg, tx);
rrw_exit(&dp->dp_config_rwlock, FTAG);
if (dst->dst_nowaiter)
kmem_free(dst, sizeof (*dst));
}
#if defined(_KERNEL)
EXPORT_SYMBOL(dsl_sync_task);
EXPORT_SYMBOL(dsl_sync_task_nowait);
#endif
|