4 * The contents of this file are subject to the terms of the
5 * Common Development and Distribution License (the "License").
6 * You may not use this file except in compliance with the License.
8 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
9 * or http://www.opensolaris.org/os/licensing.
10 * See the License for the specific language governing permissions
11 * and limitations under the License.
13 * When distributing Covered Code, include this CDDL HEADER in each
14 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
15 * If applicable, add the following below this CDDL HEADER, with the
16 * fields enclosed by brackets "[]" replaced with your own identifying
17 * information: Portions Copyright [yyyy] [name of copyright owner]
23 * Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
24 * Copyright (c) 2011, 2014 by Delphix. All rights reserved.
25 * Copyright 2011 Nexenta Systems, Inc. All rights reserved.
26 * Copyright (c) 2013, Joyent, Inc. All rights reserved.
29 /* Portions Copyright 2010 Robert Milkowski */
35 #include <sys/zio_priority.h>
42 * Types and constants shared between userland and the kernel.
46 * Each dataset can be one of the following types. These constants can be
47 * combined into masks that can be passed to various functions.
50 ZFS_TYPE_FILESYSTEM = (1 << 0),
51 ZFS_TYPE_SNAPSHOT = (1 << 1),
52 ZFS_TYPE_VOLUME = (1 << 2),
53 ZFS_TYPE_POOL = (1 << 3),
54 ZFS_TYPE_BOOKMARK = (1 << 4)
57 typedef enum dmu_objset_type {
62 DMU_OST_OTHER, /* For testing only! */
63 DMU_OST_ANY, /* Be careful! */
67 #define ZFS_TYPE_DATASET \
68 (ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME | ZFS_TYPE_SNAPSHOT)
71 * All of these include the terminating NUL byte.
73 #define ZAP_MAXNAMELEN 256
74 #define ZAP_MAXVALUELEN (1024 * 8)
75 #define ZAP_OLDMAXVALUELEN 1024
76 #define ZFS_MAX_DATASET_NAME_LEN 256
79 * Dataset properties are identified by these constants and must be added to
80 * the end of this list to ensure that external consumers are not affected
81 * by the change. If you make any changes to this list, be sure to update
82 * the property table in module/zcommon/zfs_prop.c.
91 ZFS_PROP_COMPRESSRATIO,
97 ZFS_PROP_VOLBLOCKSIZE,
102 ZFS_PROP_COMPRESSION,
110 ZFS_PROP_PRIVATE, /* not exposed to user, temporary */
112 ZFS_PROP_CREATETXG, /* not exposed to the user */
113 ZFS_PROP_NAME, /* not exposed to the user */
115 ZFS_PROP_ISCSIOPTIONS, /* not exposed to the user */
117 ZFS_PROP_NUMCLONES, /* not exposed to the user */
127 ZFS_PROP_REFRESERVATION,
129 ZFS_PROP_PRIMARYCACHE,
130 ZFS_PROP_SECONDARYCACHE,
134 ZFS_PROP_USEDREFRESERV,
135 ZFS_PROP_USERACCOUNTING, /* not exposed to the user */
136 ZFS_PROP_STMF_SHAREINFO, /* not exposed to the user */
137 ZFS_PROP_DEFER_DESTROY,
140 ZFS_PROP_UNIQUE, /* not exposed to the user */
141 ZFS_PROP_OBJSETID, /* not exposed to the user */
149 ZFS_PROP_LOGICALUSED,
150 ZFS_PROP_LOGICALREFERENCED,
151 ZFS_PROP_INCONSISTENT, /* not exposed to the user */
152 ZFS_PROP_FILESYSTEM_LIMIT,
153 ZFS_PROP_SNAPSHOT_LIMIT,
154 ZFS_PROP_FILESYSTEM_COUNT,
155 ZFS_PROP_SNAPSHOT_COUNT,
158 ZFS_PROP_SELINUX_CONTEXT,
159 ZFS_PROP_SELINUX_FSCONTEXT,
160 ZFS_PROP_SELINUX_DEFCONTEXT,
161 ZFS_PROP_SELINUX_ROOTCONTEXT,
163 ZFS_PROP_REDUNDANT_METADATA,
166 ZFS_PROP_RECEIVE_RESUME_TOKEN,
175 ZFS_PROP_USEROBJUSED,
176 ZFS_PROP_USEROBJQUOTA,
177 ZFS_PROP_GROUPOBJUSED,
178 ZFS_PROP_GROUPOBJQUOTA,
179 ZFS_NUM_USERQUOTA_PROPS
180 } zfs_userquota_prop_t;
182 extern const char *zfs_userquota_prop_prefixes[ZFS_NUM_USERQUOTA_PROPS];
185 * Pool properties are identified by these constants and must be added to the
186 * end of this list to ensure that external consumers are not affected
187 * by the change. If you make any changes to this list, be sure to update
188 * the property table in module/zcommon/zpool_prop.c.
199 ZPOOL_PROP_DELEGATION,
200 ZPOOL_PROP_AUTOREPLACE,
201 ZPOOL_PROP_CACHEFILE,
202 ZPOOL_PROP_FAILUREMODE,
203 ZPOOL_PROP_LISTSNAPS,
204 ZPOOL_PROP_AUTOEXPAND,
205 ZPOOL_PROP_DEDUPDITTO,
206 ZPOOL_PROP_DEDUPRATIO,
208 ZPOOL_PROP_ALLOCATED,
214 ZPOOL_PROP_FRAGMENTATION,
216 ZPOOL_PROP_MAXBLOCKSIZE,
218 ZPOOL_PROP_MAXDNODESIZE,
222 /* Small enough to not hog a whole line of printout in zpool(1M). */
223 #define ZPROP_MAX_COMMENT 32
225 #define ZPROP_CONT -2
226 #define ZPROP_INVAL -1
228 #define ZPROP_VALUE "value"
229 #define ZPROP_SOURCE "source"
232 ZPROP_SRC_NONE = 0x1,
233 ZPROP_SRC_DEFAULT = 0x2,
234 ZPROP_SRC_TEMPORARY = 0x4,
235 ZPROP_SRC_LOCAL = 0x8,
236 ZPROP_SRC_INHERITED = 0x10,
237 ZPROP_SRC_RECEIVED = 0x20
240 #define ZPROP_SRC_ALL 0x3f
242 #define ZPROP_SOURCE_VAL_RECVD "$recvd"
243 #define ZPROP_N_MORE_ERRORS "N_MORE_ERRORS"
246 * Dataset flag implemented as a special entry in the props zap object
247 * indicating that the dataset has received properties on or after
248 * SPA_VERSION_RECVD_PROPS. The first such receive blows away local properties
249 * just as it did in earlier versions, and thereafter, local properties are
252 #define ZPROP_HAS_RECVD "$hasrecvd"
255 ZPROP_ERR_NOCLEAR = 0x1, /* failure to clear existing props */
256 ZPROP_ERR_NORESTORE = 0x2 /* failure to restore props on error */
259 typedef int (*zprop_func)(int, void *);
262 * Properties to be set on the root file system of a new pool
263 * are stuffed into their own nvlist, which is then included in
264 * the properties nvlist with the pool properties.
266 #define ZPOOL_ROOTFS_PROPS "root-props-nvl"
269 * Dataset property functions shared between libzfs and kernel.
271 const char *zfs_prop_default_string(zfs_prop_t);
272 uint64_t zfs_prop_default_numeric(zfs_prop_t);
273 boolean_t zfs_prop_readonly(zfs_prop_t);
274 boolean_t zfs_prop_inheritable(zfs_prop_t);
275 boolean_t zfs_prop_setonce(zfs_prop_t);
276 const char *zfs_prop_to_name(zfs_prop_t);
277 zfs_prop_t zfs_name_to_prop(const char *);
278 boolean_t zfs_prop_user(const char *);
279 boolean_t zfs_prop_userquota(const char *);
280 boolean_t zfs_prop_written(const char *);
281 int zfs_prop_index_to_string(zfs_prop_t, uint64_t, const char **);
282 int zfs_prop_string_to_index(zfs_prop_t, const char *, uint64_t *);
283 uint64_t zfs_prop_random_value(zfs_prop_t, uint64_t seed);
284 boolean_t zfs_prop_valid_for_type(int, zfs_type_t, boolean_t);
287 * Pool property functions shared between libzfs and kernel.
289 zpool_prop_t zpool_name_to_prop(const char *);
290 const char *zpool_prop_to_name(zpool_prop_t);
291 const char *zpool_prop_default_string(zpool_prop_t);
292 uint64_t zpool_prop_default_numeric(zpool_prop_t);
293 boolean_t zpool_prop_readonly(zpool_prop_t);
294 boolean_t zpool_prop_feature(const char *);
295 boolean_t zpool_prop_unsupported(const char *);
296 int zpool_prop_index_to_string(zpool_prop_t, uint64_t, const char **);
297 int zpool_prop_string_to_index(zpool_prop_t, const char *, uint64_t *);
298 uint64_t zpool_prop_random_value(zpool_prop_t, uint64_t seed);
301 * Definitions for the Delegation.
304 ZFS_DELEG_WHO_UNKNOWN = 0,
305 ZFS_DELEG_USER = 'u',
306 ZFS_DELEG_USER_SETS = 'U',
307 ZFS_DELEG_GROUP = 'g',
308 ZFS_DELEG_GROUP_SETS = 'G',
309 ZFS_DELEG_EVERYONE = 'e',
310 ZFS_DELEG_EVERYONE_SETS = 'E',
311 ZFS_DELEG_CREATE = 'c',
312 ZFS_DELEG_CREATE_SETS = 'C',
313 ZFS_DELEG_NAMED_SET = 's',
314 ZFS_DELEG_NAMED_SET_SETS = 'S'
315 } zfs_deleg_who_type_t;
319 ZFS_DELEG_PERM_LOCAL = 1,
320 ZFS_DELEG_PERM_DESCENDENT = 2,
321 ZFS_DELEG_PERM_LOCALDESCENDENT = 3,
322 ZFS_DELEG_PERM_CREATE = 4
323 } zfs_deleg_inherit_t;
325 #define ZFS_DELEG_PERM_UID "uid"
326 #define ZFS_DELEG_PERM_GID "gid"
327 #define ZFS_DELEG_PERM_GROUPS "groups"
329 #define ZFS_MLSLABEL_DEFAULT "none"
331 #define ZFS_SMB_ACL_SRC "src"
332 #define ZFS_SMB_ACL_TARGET "target"
335 ZFS_CANMOUNT_OFF = 0,
337 ZFS_CANMOUNT_NOAUTO = 2
338 } zfs_canmount_type_t;
341 ZFS_LOGBIAS_LATENCY = 0,
342 ZFS_LOGBIAS_THROUGHPUT = 1
345 typedef enum zfs_share_op {
352 typedef enum zfs_smb_acl_op {
359 typedef enum zfs_cache_type {
361 ZFS_CACHE_METADATA = 1,
366 ZFS_SYNC_STANDARD = 0,
368 ZFS_SYNC_DISABLED = 2
378 ZFS_DNSIZE_LEGACY = 0,
380 ZFS_DNSIZE_1K = 1024,
381 ZFS_DNSIZE_2K = 2048,
382 ZFS_DNSIZE_4K = 4096,
383 ZFS_DNSIZE_8K = 8192,
384 ZFS_DNSIZE_16K = 16384
388 ZFS_REDUNDANT_METADATA_ALL,
389 ZFS_REDUNDANT_METADATA_MOST
390 } zfs_redundant_metadata_type_t;
393 * On-disk version number.
395 #define SPA_VERSION_1 1ULL
396 #define SPA_VERSION_2 2ULL
397 #define SPA_VERSION_3 3ULL
398 #define SPA_VERSION_4 4ULL
399 #define SPA_VERSION_5 5ULL
400 #define SPA_VERSION_6 6ULL
401 #define SPA_VERSION_7 7ULL
402 #define SPA_VERSION_8 8ULL
403 #define SPA_VERSION_9 9ULL
404 #define SPA_VERSION_10 10ULL
405 #define SPA_VERSION_11 11ULL
406 #define SPA_VERSION_12 12ULL
407 #define SPA_VERSION_13 13ULL
408 #define SPA_VERSION_14 14ULL
409 #define SPA_VERSION_15 15ULL
410 #define SPA_VERSION_16 16ULL
411 #define SPA_VERSION_17 17ULL
412 #define SPA_VERSION_18 18ULL
413 #define SPA_VERSION_19 19ULL
414 #define SPA_VERSION_20 20ULL
415 #define SPA_VERSION_21 21ULL
416 #define SPA_VERSION_22 22ULL
417 #define SPA_VERSION_23 23ULL
418 #define SPA_VERSION_24 24ULL
419 #define SPA_VERSION_25 25ULL
420 #define SPA_VERSION_26 26ULL
421 #define SPA_VERSION_27 27ULL
422 #define SPA_VERSION_28 28ULL
423 #define SPA_VERSION_5000 5000ULL
426 * When bumping up SPA_VERSION, make sure GRUB ZFS understands the on-disk
427 * format change. Go to usr/src/grub/grub-0.97/stage2/{zfs-include/, fsys_zfs*},
428 * and do the appropriate changes. Also bump the version number in
429 * usr/src/grub/capability.
431 #define SPA_VERSION SPA_VERSION_5000
432 #define SPA_VERSION_STRING "5000"
435 * Symbolic names for the changes that caused a SPA_VERSION switch.
436 * Used in the code when checking for presence or absence of a feature.
437 * Feel free to define multiple symbolic names for each version if there
438 * were multiple changes to on-disk structures during that version.
440 * NOTE: When checking the current SPA_VERSION in your code, be sure
441 * to use spa_version() since it reports the version of the
442 * last synced uberblock. Checking the in-flight version can
443 * be dangerous in some cases.
445 #define SPA_VERSION_INITIAL SPA_VERSION_1
446 #define SPA_VERSION_DITTO_BLOCKS SPA_VERSION_2
447 #define SPA_VERSION_SPARES SPA_VERSION_3
448 #define SPA_VERSION_RAIDZ2 SPA_VERSION_3
449 #define SPA_VERSION_BPOBJ_ACCOUNT SPA_VERSION_3
450 #define SPA_VERSION_RAIDZ_DEFLATE SPA_VERSION_3
451 #define SPA_VERSION_DNODE_BYTES SPA_VERSION_3
452 #define SPA_VERSION_ZPOOL_HISTORY SPA_VERSION_4
453 #define SPA_VERSION_GZIP_COMPRESSION SPA_VERSION_5
454 #define SPA_VERSION_BOOTFS SPA_VERSION_6
455 #define SPA_VERSION_SLOGS SPA_VERSION_7
456 #define SPA_VERSION_DELEGATED_PERMS SPA_VERSION_8
457 #define SPA_VERSION_FUID SPA_VERSION_9
458 #define SPA_VERSION_REFRESERVATION SPA_VERSION_9
459 #define SPA_VERSION_REFQUOTA SPA_VERSION_9
460 #define SPA_VERSION_UNIQUE_ACCURATE SPA_VERSION_9
461 #define SPA_VERSION_L2CACHE SPA_VERSION_10
462 #define SPA_VERSION_NEXT_CLONES SPA_VERSION_11
463 #define SPA_VERSION_ORIGIN SPA_VERSION_11
464 #define SPA_VERSION_DSL_SCRUB SPA_VERSION_11
465 #define SPA_VERSION_SNAP_PROPS SPA_VERSION_12
466 #define SPA_VERSION_USED_BREAKDOWN SPA_VERSION_13
467 #define SPA_VERSION_PASSTHROUGH_X SPA_VERSION_14
468 #define SPA_VERSION_USERSPACE SPA_VERSION_15
469 #define SPA_VERSION_STMF_PROP SPA_VERSION_16
470 #define SPA_VERSION_RAIDZ3 SPA_VERSION_17
471 #define SPA_VERSION_USERREFS SPA_VERSION_18
472 #define SPA_VERSION_HOLES SPA_VERSION_19
473 #define SPA_VERSION_ZLE_COMPRESSION SPA_VERSION_20
474 #define SPA_VERSION_DEDUP SPA_VERSION_21
475 #define SPA_VERSION_RECVD_PROPS SPA_VERSION_22
476 #define SPA_VERSION_SLIM_ZIL SPA_VERSION_23
477 #define SPA_VERSION_SA SPA_VERSION_24
478 #define SPA_VERSION_SCAN SPA_VERSION_25
479 #define SPA_VERSION_DIR_CLONES SPA_VERSION_26
480 #define SPA_VERSION_DEADLISTS SPA_VERSION_26
481 #define SPA_VERSION_FAST_SNAP SPA_VERSION_27
482 #define SPA_VERSION_MULTI_REPLACE SPA_VERSION_28
483 #define SPA_VERSION_BEFORE_FEATURES SPA_VERSION_28
484 #define SPA_VERSION_FEATURES SPA_VERSION_5000
486 #define SPA_VERSION_IS_SUPPORTED(v) \
487 (((v) >= SPA_VERSION_INITIAL && (v) <= SPA_VERSION_BEFORE_FEATURES) || \
488 ((v) >= SPA_VERSION_FEATURES && (v) <= SPA_VERSION))
491 * ZPL version - rev'd whenever an incompatible on-disk format change
492 * occurs. This is independent of SPA/DMU/ZAP versioning. You must
493 * also update the version_table[] and help message in zfs_prop.c.
495 * When changing, be sure to teach GRUB how to read the new format!
496 * See usr/src/grub/grub-0.97/stage2/{zfs-include/,fsys_zfs*}
498 #define ZPL_VERSION_1 1ULL
499 #define ZPL_VERSION_2 2ULL
500 #define ZPL_VERSION_3 3ULL
501 #define ZPL_VERSION_4 4ULL
502 #define ZPL_VERSION_5 5ULL
503 #define ZPL_VERSION ZPL_VERSION_5
504 #define ZPL_VERSION_STRING "5"
506 #define ZPL_VERSION_INITIAL ZPL_VERSION_1
507 #define ZPL_VERSION_DIRENT_TYPE ZPL_VERSION_2
508 #define ZPL_VERSION_FUID ZPL_VERSION_3
509 #define ZPL_VERSION_NORMALIZATION ZPL_VERSION_3
510 #define ZPL_VERSION_SYSATTR ZPL_VERSION_3
511 #define ZPL_VERSION_USERSPACE ZPL_VERSION_4
512 #define ZPL_VERSION_SA ZPL_VERSION_5
514 /* Rewind request information */
515 #define ZPOOL_NO_REWIND 1 /* No policy - default behavior */
516 #define ZPOOL_NEVER_REWIND 2 /* Do not search for best txg or rewind */
517 #define ZPOOL_TRY_REWIND 4 /* Search for best txg, but do not rewind */
518 #define ZPOOL_DO_REWIND 8 /* Rewind to best txg w/in deferred frees */
519 #define ZPOOL_EXTREME_REWIND 16 /* Allow extreme measures to find best txg */
520 #define ZPOOL_REWIND_MASK 28 /* All the possible rewind bits */
521 #define ZPOOL_REWIND_POLICIES 31 /* All the possible policy bits */
523 typedef struct zpool_rewind_policy {
524 uint32_t zrp_request; /* rewind behavior requested */
525 uint64_t zrp_maxmeta; /* max acceptable meta-data errors */
526 uint64_t zrp_maxdata; /* max acceptable data errors */
527 uint64_t zrp_txg; /* specific txg to load */
528 } zpool_rewind_policy_t;
531 * The following are configuration names used in the nvlist describing a pool's
534 #define ZPOOL_CONFIG_VERSION "version"
535 #define ZPOOL_CONFIG_POOL_NAME "name"
536 #define ZPOOL_CONFIG_POOL_STATE "state"
537 #define ZPOOL_CONFIG_POOL_TXG "txg"
538 #define ZPOOL_CONFIG_POOL_GUID "pool_guid"
539 #define ZPOOL_CONFIG_CREATE_TXG "create_txg"
540 #define ZPOOL_CONFIG_TOP_GUID "top_guid"
541 #define ZPOOL_CONFIG_VDEV_TREE "vdev_tree"
542 #define ZPOOL_CONFIG_TYPE "type"
543 #define ZPOOL_CONFIG_CHILDREN "children"
544 #define ZPOOL_CONFIG_ID "id"
545 #define ZPOOL_CONFIG_GUID "guid"
546 #define ZPOOL_CONFIG_PATH "path"
547 #define ZPOOL_CONFIG_DEVID "devid"
548 #define ZPOOL_CONFIG_METASLAB_ARRAY "metaslab_array"
549 #define ZPOOL_CONFIG_METASLAB_SHIFT "metaslab_shift"
550 #define ZPOOL_CONFIG_ASHIFT "ashift"
551 #define ZPOOL_CONFIG_ASIZE "asize"
552 #define ZPOOL_CONFIG_DTL "DTL"
553 #define ZPOOL_CONFIG_SCAN_STATS "scan_stats" /* not stored on disk */
554 #define ZPOOL_CONFIG_VDEV_STATS "vdev_stats" /* not stored on disk */
556 /* container nvlist of extended stats */
557 #define ZPOOL_CONFIG_VDEV_STATS_EX "vdev_stats_ex"
559 /* Active queue read/write stats */
560 #define ZPOOL_CONFIG_VDEV_SYNC_R_ACTIVE_QUEUE "vdev_sync_r_active_queue"
561 #define ZPOOL_CONFIG_VDEV_SYNC_W_ACTIVE_QUEUE "vdev_sync_w_active_queue"
562 #define ZPOOL_CONFIG_VDEV_ASYNC_R_ACTIVE_QUEUE "vdev_async_r_active_queue"
563 #define ZPOOL_CONFIG_VDEV_ASYNC_W_ACTIVE_QUEUE "vdev_async_w_active_queue"
564 #define ZPOOL_CONFIG_VDEV_SCRUB_ACTIVE_QUEUE "vdev_async_scrub_active_queue"
567 #define ZPOOL_CONFIG_VDEV_SYNC_R_PEND_QUEUE "vdev_sync_r_pend_queue"
568 #define ZPOOL_CONFIG_VDEV_SYNC_W_PEND_QUEUE "vdev_sync_w_pend_queue"
569 #define ZPOOL_CONFIG_VDEV_ASYNC_R_PEND_QUEUE "vdev_async_r_pend_queue"
570 #define ZPOOL_CONFIG_VDEV_ASYNC_W_PEND_QUEUE "vdev_async_w_pend_queue"
571 #define ZPOOL_CONFIG_VDEV_SCRUB_PEND_QUEUE "vdev_async_scrub_pend_queue"
573 /* Latency read/write histogram stats */
574 #define ZPOOL_CONFIG_VDEV_TOT_R_LAT_HISTO "vdev_tot_r_lat_histo"
575 #define ZPOOL_CONFIG_VDEV_TOT_W_LAT_HISTO "vdev_tot_w_lat_histo"
576 #define ZPOOL_CONFIG_VDEV_DISK_R_LAT_HISTO "vdev_disk_r_lat_histo"
577 #define ZPOOL_CONFIG_VDEV_DISK_W_LAT_HISTO "vdev_disk_w_lat_histo"
578 #define ZPOOL_CONFIG_VDEV_SYNC_R_LAT_HISTO "vdev_sync_r_lat_histo"
579 #define ZPOOL_CONFIG_VDEV_SYNC_W_LAT_HISTO "vdev_sync_w_lat_histo"
580 #define ZPOOL_CONFIG_VDEV_ASYNC_R_LAT_HISTO "vdev_async_r_lat_histo"
581 #define ZPOOL_CONFIG_VDEV_ASYNC_W_LAT_HISTO "vdev_async_w_lat_histo"
582 #define ZPOOL_CONFIG_VDEV_SCRUB_LAT_HISTO "vdev_scrub_histo"
584 /* Request size histograms */
585 #define ZPOOL_CONFIG_VDEV_SYNC_IND_R_HISTO "vdev_sync_ind_r_histo"
586 #define ZPOOL_CONFIG_VDEV_SYNC_IND_W_HISTO "vdev_sync_ind_w_histo"
587 #define ZPOOL_CONFIG_VDEV_ASYNC_IND_R_HISTO "vdev_async_ind_r_histo"
588 #define ZPOOL_CONFIG_VDEV_ASYNC_IND_W_HISTO "vdev_async_ind_w_histo"
589 #define ZPOOL_CONFIG_VDEV_IND_SCRUB_HISTO "vdev_ind_scrub_histo"
590 #define ZPOOL_CONFIG_VDEV_SYNC_AGG_R_HISTO "vdev_sync_agg_r_histo"
591 #define ZPOOL_CONFIG_VDEV_SYNC_AGG_W_HISTO "vdev_sync_agg_w_histo"
592 #define ZPOOL_CONFIG_VDEV_ASYNC_AGG_R_HISTO "vdev_async_agg_r_histo"
593 #define ZPOOL_CONFIG_VDEV_ASYNC_AGG_W_HISTO "vdev_async_agg_w_histo"
594 #define ZPOOL_CONFIG_VDEV_AGG_SCRUB_HISTO "vdev_agg_scrub_histo"
596 /* vdev enclosure sysfs path */
597 #define ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH "vdev_enc_sysfs_path"
599 #define ZPOOL_CONFIG_WHOLE_DISK "whole_disk"
600 #define ZPOOL_CONFIG_ERRCOUNT "error_count"
601 #define ZPOOL_CONFIG_NOT_PRESENT "not_present"
602 #define ZPOOL_CONFIG_SPARES "spares"
603 #define ZPOOL_CONFIG_IS_SPARE "is_spare"
604 #define ZPOOL_CONFIG_NPARITY "nparity"
605 #define ZPOOL_CONFIG_HOSTID "hostid"
606 #define ZPOOL_CONFIG_HOSTNAME "hostname"
607 #define ZPOOL_CONFIG_LOADED_TIME "initial_load_time"
608 #define ZPOOL_CONFIG_UNSPARE "unspare"
609 #define ZPOOL_CONFIG_PHYS_PATH "phys_path"
610 #define ZPOOL_CONFIG_IS_LOG "is_log"
611 #define ZPOOL_CONFIG_L2CACHE "l2cache"
612 #define ZPOOL_CONFIG_HOLE_ARRAY "hole_array"
613 #define ZPOOL_CONFIG_VDEV_CHILDREN "vdev_children"
614 #define ZPOOL_CONFIG_IS_HOLE "is_hole"
615 #define ZPOOL_CONFIG_DDT_HISTOGRAM "ddt_histogram"
616 #define ZPOOL_CONFIG_DDT_OBJ_STATS "ddt_object_stats"
617 #define ZPOOL_CONFIG_DDT_STATS "ddt_stats"
618 #define ZPOOL_CONFIG_SPLIT "splitcfg"
619 #define ZPOOL_CONFIG_ORIG_GUID "orig_guid"
620 #define ZPOOL_CONFIG_SPLIT_GUID "split_guid"
621 #define ZPOOL_CONFIG_SPLIT_LIST "guid_list"
622 #define ZPOOL_CONFIG_REMOVING "removing"
623 #define ZPOOL_CONFIG_RESILVER_TXG "resilver_txg"
624 #define ZPOOL_CONFIG_COMMENT "comment"
625 #define ZPOOL_CONFIG_SUSPENDED "suspended" /* not stored on disk */
626 #define ZPOOL_CONFIG_TIMESTAMP "timestamp" /* not stored on disk */
627 #define ZPOOL_CONFIG_BOOTFS "bootfs" /* not stored on disk */
628 #define ZPOOL_CONFIG_MISSING_DEVICES "missing_vdevs" /* not stored on disk */
629 #define ZPOOL_CONFIG_LOAD_INFO "load_info" /* not stored on disk */
630 #define ZPOOL_CONFIG_REWIND_INFO "rewind_info" /* not stored on disk */
631 #define ZPOOL_CONFIG_UNSUP_FEAT "unsup_feat" /* not stored on disk */
632 #define ZPOOL_CONFIG_ENABLED_FEAT "enabled_feat" /* not stored on disk */
633 #define ZPOOL_CONFIG_CAN_RDONLY "can_rdonly" /* not stored on disk */
634 #define ZPOOL_CONFIG_FEATURES_FOR_READ "features_for_read"
635 #define ZPOOL_CONFIG_FEATURE_STATS "feature_stats" /* not stored on disk */
636 #define ZPOOL_CONFIG_ERRATA "errata" /* not stored on disk */
637 #define ZPOOL_CONFIG_VDEV_TOP_ZAP "com.delphix:vdev_zap_top"
638 #define ZPOOL_CONFIG_VDEV_LEAF_ZAP "com.delphix:vdev_zap_leaf"
639 #define ZPOOL_CONFIG_HAS_PER_VDEV_ZAPS "com.delphix:has_per_vdev_zaps"
641 * The persistent vdev state is stored as separate values rather than a single
642 * 'vdev_state' entry. This is because a device can be in multiple states, such
643 * as offline and degraded.
645 #define ZPOOL_CONFIG_OFFLINE "offline"
646 #define ZPOOL_CONFIG_FAULTED "faulted"
647 #define ZPOOL_CONFIG_DEGRADED "degraded"
648 #define ZPOOL_CONFIG_REMOVED "removed"
649 #define ZPOOL_CONFIG_FRU "fru"
650 #define ZPOOL_CONFIG_AUX_STATE "aux_state"
652 /* Rewind policy parameters */
653 #define ZPOOL_REWIND_POLICY "rewind-policy"
654 #define ZPOOL_REWIND_REQUEST "rewind-request"
655 #define ZPOOL_REWIND_REQUEST_TXG "rewind-request-txg"
656 #define ZPOOL_REWIND_META_THRESH "rewind-meta-thresh"
657 #define ZPOOL_REWIND_DATA_THRESH "rewind-data-thresh"
659 /* Rewind data discovered */
660 #define ZPOOL_CONFIG_LOAD_TIME "rewind_txg_ts"
661 #define ZPOOL_CONFIG_LOAD_DATA_ERRORS "verify_data_errors"
662 #define ZPOOL_CONFIG_REWIND_TIME "seconds_of_rewind"
664 #define VDEV_TYPE_ROOT "root"
665 #define VDEV_TYPE_MIRROR "mirror"
666 #define VDEV_TYPE_REPLACING "replacing"
667 #define VDEV_TYPE_RAIDZ "raidz"
668 #define VDEV_TYPE_DISK "disk"
669 #define VDEV_TYPE_FILE "file"
670 #define VDEV_TYPE_MISSING "missing"
671 #define VDEV_TYPE_HOLE "hole"
672 #define VDEV_TYPE_SPARE "spare"
673 #define VDEV_TYPE_LOG "log"
674 #define VDEV_TYPE_L2CACHE "l2cache"
677 * This is needed in userland to report the minimum necessary device size.
679 #define SPA_MINDEVSIZE (64ULL << 20)
682 * Set if the fragmentation has not yet been calculated. This can happen
683 * because the space maps have not been upgraded or the histogram feature
686 #define ZFS_FRAG_INVALID UINT64_MAX
689 * The location of the pool configuration repository, shared between kernel and
692 #define ZPOOL_CACHE "/etc/zfs/zpool.cache"
695 * vdev states are ordered from least to most healthy.
696 * A vdev that's CANT_OPEN or below is considered unusable.
698 typedef enum vdev_state {
699 VDEV_STATE_UNKNOWN = 0, /* Uninitialized vdev */
700 VDEV_STATE_CLOSED, /* Not currently open */
701 VDEV_STATE_OFFLINE, /* Not allowed to open */
702 VDEV_STATE_REMOVED, /* Explicitly removed from system */
703 VDEV_STATE_CANT_OPEN, /* Tried to open, but failed */
704 VDEV_STATE_FAULTED, /* External request to fault device */
705 VDEV_STATE_DEGRADED, /* Replicated vdev with unhealthy kids */
706 VDEV_STATE_HEALTHY /* Presumed good */
709 #define VDEV_STATE_ONLINE VDEV_STATE_HEALTHY
712 * vdev aux states. When a vdev is in the CANT_OPEN state, the aux field
713 * of the vdev stats structure uses these constants to distinguish why.
715 typedef enum vdev_aux {
716 VDEV_AUX_NONE, /* no error */
717 VDEV_AUX_OPEN_FAILED, /* ldi_open_*() or vn_open() failed */
718 VDEV_AUX_CORRUPT_DATA, /* bad label or disk contents */
719 VDEV_AUX_NO_REPLICAS, /* insufficient number of replicas */
720 VDEV_AUX_BAD_GUID_SUM, /* vdev guid sum doesn't match */
721 VDEV_AUX_TOO_SMALL, /* vdev size is too small */
722 VDEV_AUX_BAD_LABEL, /* the label is OK but invalid */
723 VDEV_AUX_VERSION_NEWER, /* on-disk version is too new */
724 VDEV_AUX_VERSION_OLDER, /* on-disk version is too old */
725 VDEV_AUX_UNSUP_FEAT, /* unsupported features */
726 VDEV_AUX_SPARED, /* hot spare used in another pool */
727 VDEV_AUX_ERR_EXCEEDED, /* too many errors */
728 VDEV_AUX_IO_FAILURE, /* experienced I/O failure */
729 VDEV_AUX_BAD_LOG, /* cannot read log chain(s) */
730 VDEV_AUX_EXTERNAL, /* external diagnosis */
731 VDEV_AUX_SPLIT_POOL /* vdev was split off into another pool */
735 * pool state. The following states are written to disk as part of the normal
736 * SPA lifecycle: ACTIVE, EXPORTED, DESTROYED, SPARE, L2CACHE. The remaining
737 * states are software abstractions used at various levels to communicate
740 typedef enum pool_state {
741 POOL_STATE_ACTIVE = 0, /* In active use */
742 POOL_STATE_EXPORTED, /* Explicitly exported */
743 POOL_STATE_DESTROYED, /* Explicitly destroyed */
744 POOL_STATE_SPARE, /* Reserved for hot spare use */
745 POOL_STATE_L2CACHE, /* Level 2 ARC device */
746 POOL_STATE_UNINITIALIZED, /* Internal spa_t state */
747 POOL_STATE_UNAVAIL, /* Internal libzfs state */
748 POOL_STATE_POTENTIALLY_ACTIVE /* Internal libzfs state */
754 typedef enum pool_scan_func {
762 * ZIO types. Needed to interpret vdev statistics below.
764 typedef enum zio_type {
775 * Pool statistics. Note: all fields should be 64-bit because this
776 * is passed between kernel and userland as an nvlist uint64 array.
778 typedef struct pool_scan_stat {
779 /* values stored on disk */
780 uint64_t pss_func; /* pool_scan_func_t */
781 uint64_t pss_state; /* dsl_scan_state_t */
782 uint64_t pss_start_time; /* scan start time */
783 uint64_t pss_end_time; /* scan end time */
784 uint64_t pss_to_examine; /* total bytes to scan */
785 uint64_t pss_examined; /* total examined bytes */
786 uint64_t pss_to_process; /* total bytes to process */
787 uint64_t pss_processed; /* total processed bytes */
788 uint64_t pss_errors; /* scan errors */
790 /* values not stored on disk */
791 uint64_t pss_pass_exam; /* examined bytes per scan pass */
792 uint64_t pss_pass_start; /* start time of a scan pass */
795 typedef enum dsl_scan_state {
804 * Errata described by http://zfsonlinux.org/msg/ZFS-8000-ER. The ordering
805 * of this enum must be maintained to ensure the errata identifiers map to
806 * the correct documentation. New errata may only be appended to the list
807 * and must contain corresponding documentation at the above link.
809 typedef enum zpool_errata {
811 ZPOOL_ERRATA_ZOL_2094_SCRUB,
812 ZPOOL_ERRATA_ZOL_2094_ASYNC_DESTROY,
816 * Vdev statistics. Note: all fields should be 64-bit because this
817 * is passed between kernel and userland as an nvlist uint64 array.
819 typedef struct vdev_stat {
820 hrtime_t vs_timestamp; /* time since vdev load */
821 uint64_t vs_state; /* vdev state */
822 uint64_t vs_aux; /* see vdev_aux_t */
823 uint64_t vs_alloc; /* space allocated */
824 uint64_t vs_space; /* total capacity */
825 uint64_t vs_dspace; /* deflated capacity */
826 uint64_t vs_rsize; /* replaceable dev size */
827 uint64_t vs_esize; /* expandable dev size */
828 uint64_t vs_ops[ZIO_TYPES]; /* operation count */
829 uint64_t vs_bytes[ZIO_TYPES]; /* bytes read/written */
830 uint64_t vs_read_errors; /* read errors */
831 uint64_t vs_write_errors; /* write errors */
832 uint64_t vs_checksum_errors; /* checksum errors */
833 uint64_t vs_self_healed; /* self-healed bytes */
834 uint64_t vs_scan_removing; /* removing? */
835 uint64_t vs_scan_processed; /* scan processed bytes */
836 uint64_t vs_fragmentation; /* device fragmentation */
843 * These are stats which aren't included in the original iostat output. For
844 * convenience, they are grouped together in vdev_stat_ex, although each stat
845 * is individually exported as an nvlist.
847 typedef struct vdev_stat_ex {
848 /* Number of ZIOs issued to disk and waiting to finish */
849 uint64_t vsx_active_queue[ZIO_PRIORITY_NUM_QUEUEABLE];
851 /* Number of ZIOs pending to be issued to disk */
852 uint64_t vsx_pend_queue[ZIO_PRIORITY_NUM_QUEUEABLE];
855 * Below are the histograms for various latencies. Buckets are in
856 * units of nanoseconds.
860 * 2^37 nanoseconds = 134s. Timeouts will probably start kicking in
863 #define VDEV_L_HISTO_BUCKETS 37 /* Latency histo buckets */
864 #define VDEV_RQ_HISTO_BUCKETS 25 /* Request size histo buckets */
867 /* Amount of time in ZIO queue (ns) */
868 uint64_t vsx_queue_histo[ZIO_PRIORITY_NUM_QUEUEABLE]
869 [VDEV_L_HISTO_BUCKETS];
871 /* Total ZIO latency (ns). Includes queuing and disk access time */
872 uint64_t vsx_total_histo[ZIO_TYPES][VDEV_L_HISTO_BUCKETS];
874 /* Amount of time to read/write the disk (ns) */
875 uint64_t vsx_disk_histo[ZIO_TYPES][VDEV_L_HISTO_BUCKETS];
877 /* "lookup the bucket for a value" histogram macros */
878 #define HISTO(val, buckets) (val != 0 ? MIN(highbit64(val) - 1, \
880 #define L_HISTO(a) HISTO(a, VDEV_L_HISTO_BUCKETS)
881 #define RQ_HISTO(a) HISTO(a, VDEV_RQ_HISTO_BUCKETS)
883 /* Physical IO histogram */
884 uint64_t vsx_ind_histo[ZIO_PRIORITY_NUM_QUEUEABLE]
885 [VDEV_RQ_HISTO_BUCKETS];
887 /* Delegated (aggregated) physical IO histogram */
888 uint64_t vsx_agg_histo[ZIO_PRIORITY_NUM_QUEUEABLE]
889 [VDEV_RQ_HISTO_BUCKETS];
894 * DDT statistics. Note: all fields should be 64-bit because this
895 * is passed between kernel and userland as an nvlist uint64 array.
897 typedef struct ddt_object {
898 uint64_t ddo_count; /* number of elements in ddt */
899 uint64_t ddo_dspace; /* size of ddt on disk */
900 uint64_t ddo_mspace; /* size of ddt in-core */
903 typedef struct ddt_stat {
904 uint64_t dds_blocks; /* blocks */
905 uint64_t dds_lsize; /* logical size */
906 uint64_t dds_psize; /* physical size */
907 uint64_t dds_dsize; /* deflated allocated size */
908 uint64_t dds_ref_blocks; /* referenced blocks */
909 uint64_t dds_ref_lsize; /* referenced lsize * refcnt */
910 uint64_t dds_ref_psize; /* referenced psize * refcnt */
911 uint64_t dds_ref_dsize; /* referenced dsize * refcnt */
914 typedef struct ddt_histogram {
915 ddt_stat_t ddh_stat[64]; /* power-of-two histogram buckets */
918 #define ZVOL_DRIVER "zvol"
919 #define ZFS_DRIVER "zfs"
920 #define ZFS_DEV "/dev/zfs"
921 #define ZFS_SHARETAB "/etc/dfs/sharetab"
923 /* general zvol path */
924 #define ZVOL_DIR "/dev"
926 #define ZVOL_MAJOR 230
927 #define ZVOL_MINOR_BITS 4
928 #define ZVOL_MINOR_MASK ((1U << ZVOL_MINOR_BITS) - 1)
929 #define ZVOL_MINORS (1 << 4)
930 #define ZVOL_DEV_NAME "zd"
932 #define ZVOL_PROP_NAME "name"
933 #define ZVOL_DEFAULT_BLOCKSIZE 8192
936 * /dev/zfs ioctl numbers.
938 typedef enum zfs_ioc {
940 * Illumos - 71/128 numbers reserved.
942 ZFS_IOC_FIRST = ('Z' << 8),
943 ZFS_IOC = ZFS_IOC_FIRST,
944 ZFS_IOC_POOL_CREATE = ZFS_IOC_FIRST,
945 ZFS_IOC_POOL_DESTROY,
948 ZFS_IOC_POOL_CONFIGS,
950 ZFS_IOC_POOL_TRYIMPORT,
953 ZFS_IOC_POOL_UPGRADE,
954 ZFS_IOC_POOL_GET_HISTORY,
957 ZFS_IOC_VDEV_SET_STATE,
960 ZFS_IOC_VDEV_SETPATH,
962 ZFS_IOC_OBJSET_STATS,
963 ZFS_IOC_OBJSET_ZPLPROPS,
964 ZFS_IOC_DATASET_LIST_NEXT,
965 ZFS_IOC_SNAPSHOT_LIST_NEXT,
973 ZFS_IOC_INJECT_FAULT,
975 ZFS_IOC_INJECT_LIST_NEXT,
980 ZFS_IOC_DSOBJ_TO_DSNAME,
982 ZFS_IOC_POOL_SET_PROPS,
983 ZFS_IOC_POOL_GET_PROPS,
987 ZFS_IOC_INHERIT_PROP,
989 ZFS_IOC_USERSPACE_ONE,
990 ZFS_IOC_USERSPACE_MANY,
991 ZFS_IOC_USERSPACE_UPGRADE,
995 ZFS_IOC_OBJSET_RECVD_PROPS,
999 ZFS_IOC_TMP_SNAPSHOT,
1000 ZFS_IOC_OBJ_TO_STATS,
1001 ZFS_IOC_SPACE_WRITTEN,
1002 ZFS_IOC_SPACE_SNAPS,
1003 ZFS_IOC_DESTROY_SNAPS,
1004 ZFS_IOC_POOL_REGUID,
1005 ZFS_IOC_POOL_REOPEN,
1006 ZFS_IOC_SEND_PROGRESS,
1007 ZFS_IOC_LOG_HISTORY,
1012 ZFS_IOC_GET_BOOKMARKS,
1013 ZFS_IOC_DESTROY_BOOKMARKS,
1017 * Linux - 3/64 numbers reserved.
1019 ZFS_IOC_LINUX = ('Z' << 8) + 0x80,
1020 ZFS_IOC_EVENTS_NEXT,
1021 ZFS_IOC_EVENTS_CLEAR,
1022 ZFS_IOC_EVENTS_SEEK,
1025 * FreeBSD - 1/64 numbers reserved.
1027 ZFS_IOC_FREEBSD = ('Z' << 8) + 0xC0,
1033 * zvol ioctl to get dataset name
1035 #define BLKZNAME _IOR(0x12, 125, char[ZFS_MAX_DATASET_NAME_LEN])
1038 * Internal SPA load state. Used by FMA diagnosis engine.
1041 SPA_LOAD_NONE, /* no load in progress */
1042 SPA_LOAD_OPEN, /* normal open */
1043 SPA_LOAD_IMPORT, /* import in progress */
1044 SPA_LOAD_TRYIMPORT, /* tryimport in progress */
1045 SPA_LOAD_RECOVER, /* recovery requested */
1046 SPA_LOAD_ERROR, /* load failed */
1047 SPA_LOAD_CREATE /* creation in progress */
1051 * Bookmark name values.
1053 #define ZPOOL_ERR_LIST "error list"
1054 #define ZPOOL_ERR_DATASET "dataset"
1055 #define ZPOOL_ERR_OBJECT "object"
1057 #define HIS_MAX_RECORD_LEN (MAXPATHLEN + MAXPATHLEN + 1)
1060 * The following are names used in the nvlist describing
1061 * the pool's history log.
1063 #define ZPOOL_HIST_RECORD "history record"
1064 #define ZPOOL_HIST_TIME "history time"
1065 #define ZPOOL_HIST_CMD "history command"
1066 #define ZPOOL_HIST_WHO "history who"
1067 #define ZPOOL_HIST_ZONE "history zone"
1068 #define ZPOOL_HIST_HOST "history hostname"
1069 #define ZPOOL_HIST_TXG "history txg"
1070 #define ZPOOL_HIST_INT_EVENT "history internal event"
1071 #define ZPOOL_HIST_INT_STR "history internal str"
1072 #define ZPOOL_HIST_INT_NAME "internal_name"
1073 #define ZPOOL_HIST_IOCTL "ioctl"
1074 #define ZPOOL_HIST_INPUT_NVL "in_nvl"
1075 #define ZPOOL_HIST_OUTPUT_NVL "out_nvl"
1076 #define ZPOOL_HIST_DSNAME "dsname"
1077 #define ZPOOL_HIST_DSID "dsid"
1080 * Flags for ZFS_IOC_VDEV_SET_STATE
1082 #define ZFS_ONLINE_CHECKREMOVE 0x1
1083 #define ZFS_ONLINE_UNSPARE 0x2
1084 #define ZFS_ONLINE_FORCEFAULT 0x4
1085 #define ZFS_ONLINE_EXPAND 0x8
1086 #define ZFS_OFFLINE_TEMPORARY 0x1
1089 * Flags for ZFS_IOC_POOL_IMPORT
1091 #define ZFS_IMPORT_NORMAL 0x0
1092 #define ZFS_IMPORT_VERBATIM 0x1
1093 #define ZFS_IMPORT_ANY_HOST 0x2
1094 #define ZFS_IMPORT_MISSING_LOG 0x4
1095 #define ZFS_IMPORT_ONLY 0x8
1096 #define ZFS_IMPORT_TEMP_NAME 0x10
1099 * Sysevent payload members. ZFS will generate the following sysevents with the
1102 * ESC_ZFS_RESILVER_START
1103 * ESC_ZFS_RESILVER_END
1104 * ESC_ZFS_POOL_DESTROY
1105 * ESC_ZFS_POOL_REGUID
1107 * ZFS_EV_POOL_NAME DATA_TYPE_STRING
1108 * ZFS_EV_POOL_GUID DATA_TYPE_UINT64
1110 * ESC_ZFS_VDEV_REMOVE
1111 * ESC_ZFS_VDEV_CLEAR
1112 * ESC_ZFS_VDEV_CHECK
1114 * ZFS_EV_POOL_NAME DATA_TYPE_STRING
1115 * ZFS_EV_POOL_GUID DATA_TYPE_UINT64
1116 * ZFS_EV_VDEV_PATH DATA_TYPE_STRING (optional)
1117 * ZFS_EV_VDEV_GUID DATA_TYPE_UINT64
1119 #define ZFS_EV_POOL_NAME "pool_name"
1120 #define ZFS_EV_POOL_GUID "pool_guid"
1121 #define ZFS_EV_VDEV_PATH "vdev_path"
1122 #define ZFS_EV_VDEV_GUID "vdev_guid"
1128 #endif /* _SYS_FS_ZFS_H */