diff --git a/include/sys/dmu_objset.h b/include/sys/dmu_objset.h index 9f6e0fdd601b..a9123e862af7 100644 --- a/include/sys/dmu_objset.h +++ b/include/sys/dmu_objset.h @@ -1,276 +1,277 @@ /* * 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, 2018 by Delphix. All rights reserved. * Copyright (c) 2013 by Saso Kiselkov. All rights reserved. * Copyright (c) 2014 Spectra Logic Corporation, All rights reserved. */ /* Portions Copyright 2010 Robert Milkowski */ #ifndef _SYS_DMU_OBJSET_H #define _SYS_DMU_OBJSET_H #include #include #include #include #include #include #include #include #include #ifdef __cplusplus extern "C" { #endif extern krwlock_t os_lock; struct dsl_pool; struct dsl_dataset; struct dmu_tx; #define OBJSET_PHYS_SIZE_V1 1024 #define OBJSET_PHYS_SIZE_V2 2048 #define OBJSET_PHYS_SIZE_V3 4096 #define OBJSET_BUF_HAS_USERUSED(buf) \ (arc_buf_size(buf) >= OBJSET_PHYS_SIZE_V2) #define OBJSET_BUF_HAS_PROJECTUSED(buf) \ (arc_buf_size(buf) >= OBJSET_PHYS_SIZE_V3) #define OBJSET_FLAG_USERACCOUNTING_COMPLETE (1ULL << 0) #define OBJSET_FLAG_USEROBJACCOUNTING_COMPLETE (1ULL << 1) #define OBJSET_FLAG_PROJECTQUOTA_COMPLETE (1ULL << 2) /* * This mask defines the set of flags which are "portable", meaning * that they can be preserved when doing a raw encrypted zfs send. * Flags included in this mask will be protected by os_portable_mac * when the block of dnodes is encrypted. No portable flags currently * exist. */ #define OBJSET_CRYPT_PORTABLE_FLAGS_MASK (0) #if defined(__clang__) #pragma clang diagnostic push #pragma clang diagnostic ignored "-Wgnu-variable-sized-type-not-at-end" #endif typedef struct objset_phys { dnode_phys_t os_meta_dnode; zil_header_t os_zil_header; uint64_t os_type; uint64_t os_flags; uint8_t os_portable_mac[ZIO_OBJSET_MAC_LEN]; uint8_t os_local_mac[ZIO_OBJSET_MAC_LEN]; char os_pad0[OBJSET_PHYS_SIZE_V2 - sizeof (dnode_phys_t)*3 - sizeof (zil_header_t) - sizeof (uint64_t)*2 - 2*ZIO_OBJSET_MAC_LEN]; dnode_phys_t os_userused_dnode; dnode_phys_t os_groupused_dnode; dnode_phys_t os_projectused_dnode; char os_pad1[OBJSET_PHYS_SIZE_V3 - OBJSET_PHYS_SIZE_V2 - sizeof (dnode_phys_t)]; } objset_phys_t; #if defined(__clang__) #pragma clang diagnostic pop #endif typedef int (*dmu_objset_upgrade_cb_t)(objset_t *); #define OBJSET_PROP_UNINITIALIZED ((uint64_t)-1) struct objset { /* Immutable: */ struct dsl_dataset *os_dsl_dataset; spa_t *os_spa; arc_buf_t *os_phys_buf; objset_phys_t *os_phys; boolean_t os_encrypted; /* * The following "special" dnodes have no parent, are exempt * from dnode_move(), and are not recorded in os_dnodes, but they * root their descendents in this objset using handles anyway, so * that all access to dnodes from dbufs consistently uses handles. */ dnode_handle_t os_meta_dnode; dnode_handle_t os_userused_dnode; dnode_handle_t os_groupused_dnode; dnode_handle_t os_projectused_dnode; zilog_t *os_zil; list_node_t os_evicting_node; /* can change, under dsl_dir's locks: */ uint64_t os_dnodesize; /* default dnode size for new objects */ enum zio_checksum os_checksum; enum zio_compress os_compress; uint8_t os_complevel; uint8_t os_copies; enum zio_checksum os_dedup_checksum; boolean_t os_dedup_verify; zfs_logbias_op_t os_logbias; zfs_cache_type_t os_primary_cache; zfs_cache_type_t os_secondary_cache; + zfs_prefetch_type_t os_prefetch; zfs_sync_type_t os_sync; zfs_redundant_metadata_type_t os_redundant_metadata; uint64_t os_recordsize; /* * The next four values are used as a cache of whatever's on disk, and * are initialized the first time these properties are queried. Before * being initialized with their real values, their values are * OBJSET_PROP_UNINITIALIZED. */ uint64_t os_version; uint64_t os_normalization; uint64_t os_utf8only; uint64_t os_casesensitivity; /* * The largest zpl file block allowed in special class. * cached here instead of zfsvfs for easier access. */ int os_zpl_special_smallblock; /* * Pointer is constant; the blkptr it points to is protected by * os_dsl_dataset->ds_bp_rwlock */ blkptr_t *os_rootbp; /* no lock needed: */ struct dmu_tx *os_synctx; /* XXX sketchy */ zil_header_t os_zil_header; multilist_t os_synced_dnodes; uint64_t os_flags; uint64_t os_freed_dnodes; boolean_t os_rescan_dnodes; boolean_t os_raw_receive; /* os_phys_buf should be written raw next txg */ boolean_t os_next_write_raw[TXG_SIZE]; /* Protected by os_obj_lock */ kmutex_t os_obj_lock; uint64_t os_obj_next_chunk; /* Per-CPU next object to allocate, protected by atomic ops. */ uint64_t *os_obj_next_percpu; int os_obj_next_percpu_len; /* Protected by os_lock */ kmutex_t os_lock; multilist_t os_dirty_dnodes[TXG_SIZE]; list_t os_dnodes; list_t os_downgraded_dbufs; /* Protects changes to DMU_{USER,GROUP,PROJECT}USED_OBJECT */ kmutex_t os_userused_lock; /* stuff we store for the user */ kmutex_t os_user_ptr_lock; void *os_user_ptr; sa_os_t *os_sa; /* kernel thread to upgrade this dataset */ kmutex_t os_upgrade_lock; taskqid_t os_upgrade_id; dmu_objset_upgrade_cb_t os_upgrade_cb; boolean_t os_upgrade_exit; int os_upgrade_status; }; #define DMU_META_OBJSET 0 #define DMU_META_DNODE_OBJECT 0 #define DMU_OBJECT_IS_SPECIAL(obj) ((int64_t)(obj) <= 0) #define DMU_META_DNODE(os) ((os)->os_meta_dnode.dnh_dnode) #define DMU_USERUSED_DNODE(os) ((os)->os_userused_dnode.dnh_dnode) #define DMU_GROUPUSED_DNODE(os) ((os)->os_groupused_dnode.dnh_dnode) #define DMU_PROJECTUSED_DNODE(os) ((os)->os_projectused_dnode.dnh_dnode) /* called from zpl */ int dmu_objset_hold(const char *name, const void *tag, objset_t **osp); int dmu_objset_hold_flags(const char *name, boolean_t decrypt, const void *tag, objset_t **osp); int dmu_objset_own(const char *name, dmu_objset_type_t type, boolean_t readonly, boolean_t decrypt, const void *tag, objset_t **osp); int dmu_objset_own_obj(struct dsl_pool *dp, uint64_t obj, dmu_objset_type_t type, boolean_t readonly, boolean_t decrypt, const void *tag, objset_t **osp); void dmu_objset_refresh_ownership(struct dsl_dataset *ds, struct dsl_dataset **newds, boolean_t decrypt, const void *tag); void dmu_objset_rele(objset_t *os, const void *tag); void dmu_objset_rele_flags(objset_t *os, boolean_t decrypt, const void *tag); void dmu_objset_disown(objset_t *os, boolean_t decrypt, const void *tag); int dmu_objset_from_ds(struct dsl_dataset *ds, objset_t **osp); void dmu_objset_stats(objset_t *os, nvlist_t *nv); void dmu_objset_fast_stat(objset_t *os, dmu_objset_stats_t *stat); void dmu_objset_space(objset_t *os, uint64_t *refdbytesp, uint64_t *availbytesp, uint64_t *usedobjsp, uint64_t *availobjsp); uint64_t dmu_objset_fsid_guid(objset_t *os); int dmu_objset_find_dp(struct dsl_pool *dp, uint64_t ddobj, int func(struct dsl_pool *, struct dsl_dataset *, void *), void *arg, int flags); void dmu_objset_evict_dbufs(objset_t *os); inode_timespec_t dmu_objset_snap_cmtime(objset_t *os); /* called from dsl */ void dmu_objset_sync(objset_t *os, zio_t *zio, dmu_tx_t *tx); boolean_t dmu_objset_is_dirty(objset_t *os, uint64_t txg); objset_t *dmu_objset_create_impl_dnstats(spa_t *spa, struct dsl_dataset *ds, blkptr_t *bp, dmu_objset_type_t type, int levels, int blksz, int ibs, dmu_tx_t *tx); objset_t *dmu_objset_create_impl(spa_t *spa, struct dsl_dataset *ds, blkptr_t *bp, dmu_objset_type_t type, dmu_tx_t *tx); int dmu_objset_open_impl(spa_t *spa, struct dsl_dataset *ds, blkptr_t *bp, objset_t **osp); void dmu_objset_evict(objset_t *os); void dmu_objset_sync_done(objset_t *os, dmu_tx_t *tx); void dmu_objset_userquota_get_ids(dnode_t *dn, boolean_t before, dmu_tx_t *tx); boolean_t dmu_objset_userused_enabled(objset_t *os); void dmu_objset_userspace_upgrade(objset_t *os); boolean_t dmu_objset_userspace_present(objset_t *os); boolean_t dmu_objset_userobjused_enabled(objset_t *os); boolean_t dmu_objset_userobjspace_upgradable(objset_t *os); boolean_t dmu_objset_userobjspace_present(objset_t *os); boolean_t dmu_objset_incompatible_encryption_version(objset_t *os); boolean_t dmu_objset_projectquota_enabled(objset_t *os); boolean_t dmu_objset_projectquota_present(objset_t *os); boolean_t dmu_objset_projectquota_upgradable(objset_t *os); void dmu_objset_id_quota_upgrade(objset_t *os); int dmu_get_file_info(objset_t *os, dmu_object_type_t bonustype, const void *data, zfs_file_info_t *zfi); int dmu_fsname(const char *snapname, char *buf); void dmu_objset_evict_done(objset_t *os); void dmu_objset_willuse_space(objset_t *os, int64_t space, dmu_tx_t *tx); void dmu_objset_init(void); void dmu_objset_fini(void); #ifdef __cplusplus } #endif #endif /* _SYS_DMU_OBJSET_H */ diff --git a/include/sys/fs/zfs.h b/include/sys/fs/zfs.h index f84cb7aade7f..4329e4e86f2d 100644 --- a/include/sys/fs/zfs.h +++ b/include/sys/fs/zfs.h @@ -1,1865 +1,1872 @@ /* * 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) 2011, 2024 by Delphix. All rights reserved. * Copyright 2011 Nexenta Systems, Inc. All rights reserved. * Copyright (c) 2013, 2017 Joyent, Inc. All rights reserved. * Copyright (c) 2014 Integros [integros.com] * Copyright (c) 2017, Intel Corporation. * Copyright (c) 2019 Datto Inc. * Portions Copyright 2010 Robert Milkowski * Copyright (c) 2021, Colm Buckley * Copyright (c) 2022 Hewlett Packard Enterprise Development LP. */ #ifndef _SYS_FS_ZFS_H #define _SYS_FS_ZFS_H extern __attribute__((visibility("default"))) #include #include #ifdef __cplusplus extern "C" { #endif /* * Types and constants shared between userland and the kernel. */ /* * Each dataset can be one of the following types. These constants can be * combined into masks that can be passed to various functions. */ typedef enum { ZFS_TYPE_INVALID = 0, ZFS_TYPE_FILESYSTEM = (1 << 0), ZFS_TYPE_SNAPSHOT = (1 << 1), ZFS_TYPE_VOLUME = (1 << 2), ZFS_TYPE_POOL = (1 << 3), ZFS_TYPE_BOOKMARK = (1 << 4), ZFS_TYPE_VDEV = (1 << 5), } zfs_type_t; /* * NB: lzc_dataset_type should be updated whenever a new objset type is added, * if it represents a real type of a dataset that can be created from userland. */ typedef enum dmu_objset_type { DMU_OST_NONE, DMU_OST_META, DMU_OST_ZFS, DMU_OST_ZVOL, DMU_OST_OTHER, /* For testing only! */ DMU_OST_ANY, /* Be careful! */ DMU_OST_NUMTYPES } dmu_objset_type_t; #define ZFS_TYPE_DATASET \ (ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME | ZFS_TYPE_SNAPSHOT) /* * All of these include the terminating NUL byte. */ #define ZAP_MAXNAMELEN 256 #define ZAP_MAXVALUELEN (1024 * 8) #define ZAP_OLDMAXVALUELEN 1024 #define ZFS_MAX_DATASET_NAME_LEN 256 /* * Dataset properties are identified by these constants and must be added to * the end of this list to ensure that external consumers are not affected * by the change. If you make any changes to this list, be sure to update * the property table in module/zcommon/zfs_prop.c. */ typedef enum { ZPROP_CONT = -2, ZPROP_INVAL = -1, ZPROP_USERPROP = ZPROP_INVAL, ZFS_PROP_TYPE = 0, ZFS_PROP_CREATION, ZFS_PROP_USED, ZFS_PROP_AVAILABLE, ZFS_PROP_REFERENCED, ZFS_PROP_COMPRESSRATIO, ZFS_PROP_MOUNTED, ZFS_PROP_ORIGIN, ZFS_PROP_QUOTA, ZFS_PROP_RESERVATION, ZFS_PROP_VOLSIZE, ZFS_PROP_VOLBLOCKSIZE, ZFS_PROP_RECORDSIZE, ZFS_PROP_MOUNTPOINT, ZFS_PROP_SHARENFS, ZFS_PROP_CHECKSUM, ZFS_PROP_COMPRESSION, ZFS_PROP_ATIME, ZFS_PROP_DEVICES, ZFS_PROP_EXEC, ZFS_PROP_SETUID, ZFS_PROP_READONLY, ZFS_PROP_ZONED, ZFS_PROP_SNAPDIR, ZFS_PROP_ACLMODE, ZFS_PROP_ACLINHERIT, ZFS_PROP_CREATETXG, ZFS_PROP_NAME, /* not exposed to the user */ ZFS_PROP_CANMOUNT, ZFS_PROP_ISCSIOPTIONS, /* not exposed to the user */ ZFS_PROP_XATTR, ZFS_PROP_NUMCLONES, /* not exposed to the user */ ZFS_PROP_COPIES, ZFS_PROP_VERSION, ZFS_PROP_UTF8ONLY, ZFS_PROP_NORMALIZE, ZFS_PROP_CASE, ZFS_PROP_VSCAN, ZFS_PROP_NBMAND, ZFS_PROP_SHARESMB, ZFS_PROP_REFQUOTA, ZFS_PROP_REFRESERVATION, ZFS_PROP_GUID, ZFS_PROP_PRIMARYCACHE, ZFS_PROP_SECONDARYCACHE, ZFS_PROP_USEDSNAP, ZFS_PROP_USEDDS, ZFS_PROP_USEDCHILD, ZFS_PROP_USEDREFRESERV, ZFS_PROP_USERACCOUNTING, /* not exposed to the user */ ZFS_PROP_STMF_SHAREINFO, /* not exposed to the user */ ZFS_PROP_DEFER_DESTROY, ZFS_PROP_USERREFS, ZFS_PROP_LOGBIAS, ZFS_PROP_UNIQUE, /* not exposed to the user */ ZFS_PROP_OBJSETID, ZFS_PROP_DEDUP, ZFS_PROP_MLSLABEL, ZFS_PROP_SYNC, ZFS_PROP_DNODESIZE, ZFS_PROP_REFRATIO, ZFS_PROP_WRITTEN, ZFS_PROP_CLONES, ZFS_PROP_LOGICALUSED, ZFS_PROP_LOGICALREFERENCED, ZFS_PROP_INCONSISTENT, /* not exposed to the user */ ZFS_PROP_VOLMODE, ZFS_PROP_FILESYSTEM_LIMIT, ZFS_PROP_SNAPSHOT_LIMIT, ZFS_PROP_FILESYSTEM_COUNT, ZFS_PROP_SNAPSHOT_COUNT, ZFS_PROP_SNAPDEV, ZFS_PROP_ACLTYPE, ZFS_PROP_SELINUX_CONTEXT, ZFS_PROP_SELINUX_FSCONTEXT, ZFS_PROP_SELINUX_DEFCONTEXT, ZFS_PROP_SELINUX_ROOTCONTEXT, ZFS_PROP_RELATIME, ZFS_PROP_REDUNDANT_METADATA, ZFS_PROP_OVERLAY, ZFS_PROP_PREV_SNAP, ZFS_PROP_RECEIVE_RESUME_TOKEN, ZFS_PROP_ENCRYPTION, ZFS_PROP_KEYLOCATION, ZFS_PROP_KEYFORMAT, ZFS_PROP_PBKDF2_SALT, ZFS_PROP_PBKDF2_ITERS, ZFS_PROP_ENCRYPTION_ROOT, ZFS_PROP_KEY_GUID, ZFS_PROP_KEYSTATUS, ZFS_PROP_REMAPTXG, /* obsolete - no longer used */ ZFS_PROP_SPECIAL_SMALL_BLOCKS, ZFS_PROP_IVSET_GUID, /* not exposed to the user */ ZFS_PROP_REDACTED, ZFS_PROP_REDACT_SNAPS, ZFS_PROP_SNAPSHOTS_CHANGED, + ZFS_PROP_PREFETCH, ZFS_NUM_PROPS } zfs_prop_t; typedef enum { ZFS_PROP_USERUSED, ZFS_PROP_USERQUOTA, ZFS_PROP_GROUPUSED, ZFS_PROP_GROUPQUOTA, ZFS_PROP_USEROBJUSED, ZFS_PROP_USEROBJQUOTA, ZFS_PROP_GROUPOBJUSED, ZFS_PROP_GROUPOBJQUOTA, ZFS_PROP_PROJECTUSED, ZFS_PROP_PROJECTQUOTA, ZFS_PROP_PROJECTOBJUSED, ZFS_PROP_PROJECTOBJQUOTA, ZFS_NUM_USERQUOTA_PROPS } zfs_userquota_prop_t; _SYS_FS_ZFS_H const char *const zfs_userquota_prop_prefixes[ ZFS_NUM_USERQUOTA_PROPS]; /* * Pool properties are identified by these constants and must be added to the * end of this list to ensure that external consumers are not affected * by the change. Properties must be registered in zfs_prop_init(). */ typedef enum { ZPOOL_PROP_INVAL = -1, ZPOOL_PROP_NAME, ZPOOL_PROP_SIZE, ZPOOL_PROP_CAPACITY, ZPOOL_PROP_ALTROOT, ZPOOL_PROP_HEALTH, ZPOOL_PROP_GUID, ZPOOL_PROP_VERSION, ZPOOL_PROP_BOOTFS, ZPOOL_PROP_DELEGATION, ZPOOL_PROP_AUTOREPLACE, ZPOOL_PROP_CACHEFILE, ZPOOL_PROP_FAILUREMODE, ZPOOL_PROP_LISTSNAPS, ZPOOL_PROP_AUTOEXPAND, ZPOOL_PROP_DEDUPDITTO, ZPOOL_PROP_DEDUPRATIO, ZPOOL_PROP_FREE, ZPOOL_PROP_ALLOCATED, ZPOOL_PROP_READONLY, ZPOOL_PROP_ASHIFT, ZPOOL_PROP_COMMENT, ZPOOL_PROP_EXPANDSZ, ZPOOL_PROP_FREEING, ZPOOL_PROP_FRAGMENTATION, ZPOOL_PROP_LEAKED, ZPOOL_PROP_MAXBLOCKSIZE, ZPOOL_PROP_TNAME, ZPOOL_PROP_MAXDNODESIZE, ZPOOL_PROP_MULTIHOST, ZPOOL_PROP_CHECKPOINT, ZPOOL_PROP_LOAD_GUID, ZPOOL_PROP_AUTOTRIM, ZPOOL_PROP_COMPATIBILITY, ZPOOL_PROP_BCLONEUSED, ZPOOL_PROP_BCLONESAVED, ZPOOL_PROP_BCLONERATIO, ZPOOL_NUM_PROPS } zpool_prop_t; /* Small enough to not hog a whole line of printout in zpool(8). */ #define ZPROP_MAX_COMMENT 32 #define ZPROP_BOOLEAN_NA 2 #define ZPROP_VALUE "value" #define ZPROP_SOURCE "source" typedef enum { ZPROP_SRC_NONE = 0x1, ZPROP_SRC_DEFAULT = 0x2, ZPROP_SRC_TEMPORARY = 0x4, ZPROP_SRC_LOCAL = 0x8, ZPROP_SRC_INHERITED = 0x10, ZPROP_SRC_RECEIVED = 0x20 } zprop_source_t; #define ZPROP_SRC_ALL 0x3f #define ZPROP_SOURCE_VAL_RECVD "$recvd" #define ZPROP_N_MORE_ERRORS "N_MORE_ERRORS" /* * Dataset flag implemented as a special entry in the props zap object * indicating that the dataset has received properties on or after * SPA_VERSION_RECVD_PROPS. The first such receive blows away local properties * just as it did in earlier versions, and thereafter, local properties are * preserved. */ #define ZPROP_HAS_RECVD "$hasrecvd" typedef enum { ZPROP_ERR_NOCLEAR = 0x1, /* failure to clear existing props */ ZPROP_ERR_NORESTORE = 0x2 /* failure to restore props on error */ } zprop_errflags_t; typedef int (*zprop_func)(int, void *); /* * Properties to be set on the root file system of a new pool * are stuffed into their own nvlist, which is then included in * the properties nvlist with the pool properties. */ #define ZPOOL_ROOTFS_PROPS "root-props-nvl" /* * Length of 'written@' and 'written#' */ #define ZFS_WRITTEN_PROP_PREFIX_LEN 8 /* * VDEV properties are identified by these constants and must be added to the * end of this list to ensure that external consumers are not affected * by the change. If you make any changes to this list, be sure to update * the property table in usr/src/common/zfs/zpool_prop.c. */ typedef enum { VDEV_PROP_INVAL = -1, VDEV_PROP_USERPROP = VDEV_PROP_INVAL, VDEV_PROP_NAME, VDEV_PROP_CAPACITY, VDEV_PROP_STATE, VDEV_PROP_GUID, VDEV_PROP_ASIZE, VDEV_PROP_PSIZE, VDEV_PROP_ASHIFT, VDEV_PROP_SIZE, VDEV_PROP_FREE, VDEV_PROP_ALLOCATED, VDEV_PROP_COMMENT, VDEV_PROP_EXPANDSZ, VDEV_PROP_FRAGMENTATION, VDEV_PROP_BOOTSIZE, VDEV_PROP_PARITY, VDEV_PROP_PATH, VDEV_PROP_DEVID, VDEV_PROP_PHYS_PATH, VDEV_PROP_ENC_PATH, VDEV_PROP_FRU, VDEV_PROP_PARENT, VDEV_PROP_CHILDREN, VDEV_PROP_NUMCHILDREN, VDEV_PROP_READ_ERRORS, VDEV_PROP_WRITE_ERRORS, VDEV_PROP_CHECKSUM_ERRORS, VDEV_PROP_INITIALIZE_ERRORS, VDEV_PROP_OPS_NULL, VDEV_PROP_OPS_READ, VDEV_PROP_OPS_WRITE, VDEV_PROP_OPS_FREE, VDEV_PROP_OPS_CLAIM, VDEV_PROP_OPS_TRIM, VDEV_PROP_BYTES_NULL, VDEV_PROP_BYTES_READ, VDEV_PROP_BYTES_WRITE, VDEV_PROP_BYTES_FREE, VDEV_PROP_BYTES_CLAIM, VDEV_PROP_BYTES_TRIM, VDEV_PROP_REMOVING, VDEV_PROP_ALLOCATING, VDEV_PROP_FAILFAST, VDEV_PROP_CHECKSUM_N, VDEV_PROP_CHECKSUM_T, VDEV_PROP_IO_N, VDEV_PROP_IO_T, VDEV_PROP_RAIDZ_EXPANDING, VDEV_PROP_SLOW_IO_N, VDEV_PROP_SLOW_IO_T, VDEV_NUM_PROPS } vdev_prop_t; /* * Dataset property functions shared between libzfs and kernel. */ _SYS_FS_ZFS_H const char *zfs_prop_default_string(zfs_prop_t); _SYS_FS_ZFS_H uint64_t zfs_prop_default_numeric(zfs_prop_t); _SYS_FS_ZFS_H boolean_t zfs_prop_readonly(zfs_prop_t); _SYS_FS_ZFS_H boolean_t zfs_prop_visible(zfs_prop_t prop); _SYS_FS_ZFS_H boolean_t zfs_prop_inheritable(zfs_prop_t); _SYS_FS_ZFS_H boolean_t zfs_prop_setonce(zfs_prop_t); _SYS_FS_ZFS_H boolean_t zfs_prop_encryption_key_param(zfs_prop_t); _SYS_FS_ZFS_H boolean_t zfs_prop_valid_keylocation(const char *, boolean_t); _SYS_FS_ZFS_H const char *zfs_prop_to_name(zfs_prop_t); _SYS_FS_ZFS_H zfs_prop_t zfs_name_to_prop(const char *); _SYS_FS_ZFS_H boolean_t zfs_prop_user(const char *); _SYS_FS_ZFS_H boolean_t zfs_prop_userquota(const char *); _SYS_FS_ZFS_H boolean_t zfs_prop_written(const char *); _SYS_FS_ZFS_H int zfs_prop_index_to_string(zfs_prop_t, uint64_t, const char **); _SYS_FS_ZFS_H int zfs_prop_string_to_index(zfs_prop_t, const char *, uint64_t *); _SYS_FS_ZFS_H uint64_t zfs_prop_random_value(zfs_prop_t, uint64_t seed); _SYS_FS_ZFS_H boolean_t zfs_prop_valid_for_type(int, zfs_type_t, boolean_t); /* * Pool property functions shared between libzfs and kernel. */ _SYS_FS_ZFS_H zpool_prop_t zpool_name_to_prop(const char *); _SYS_FS_ZFS_H const char *zpool_prop_to_name(zpool_prop_t); _SYS_FS_ZFS_H const char *zpool_prop_default_string(zpool_prop_t); _SYS_FS_ZFS_H uint64_t zpool_prop_default_numeric(zpool_prop_t); _SYS_FS_ZFS_H boolean_t zpool_prop_readonly(zpool_prop_t); _SYS_FS_ZFS_H boolean_t zpool_prop_setonce(zpool_prop_t); _SYS_FS_ZFS_H boolean_t zpool_prop_feature(const char *); _SYS_FS_ZFS_H boolean_t zpool_prop_unsupported(const char *); _SYS_FS_ZFS_H int zpool_prop_index_to_string(zpool_prop_t, uint64_t, const char **); _SYS_FS_ZFS_H int zpool_prop_string_to_index(zpool_prop_t, const char *, uint64_t *); _SYS_FS_ZFS_H uint64_t zpool_prop_random_value(zpool_prop_t, uint64_t seed); /* * VDEV property functions shared between libzfs and kernel. */ _SYS_FS_ZFS_H vdev_prop_t vdev_name_to_prop(const char *); _SYS_FS_ZFS_H boolean_t vdev_prop_user(const char *name); _SYS_FS_ZFS_H const char *vdev_prop_to_name(vdev_prop_t); _SYS_FS_ZFS_H const char *vdev_prop_default_string(vdev_prop_t); _SYS_FS_ZFS_H uint64_t vdev_prop_default_numeric(vdev_prop_t); _SYS_FS_ZFS_H boolean_t vdev_prop_readonly(vdev_prop_t prop); _SYS_FS_ZFS_H int vdev_prop_index_to_string(vdev_prop_t, uint64_t, const char **); _SYS_FS_ZFS_H int vdev_prop_string_to_index(vdev_prop_t, const char *, uint64_t *); _SYS_FS_ZFS_H boolean_t zpool_prop_vdev(const char *name); _SYS_FS_ZFS_H uint64_t vdev_prop_random_value(vdev_prop_t prop, uint64_t seed); /* * Definitions for the Delegation. */ typedef enum { ZFS_DELEG_WHO_UNKNOWN = 0, ZFS_DELEG_USER = 'u', ZFS_DELEG_USER_SETS = 'U', ZFS_DELEG_GROUP = 'g', ZFS_DELEG_GROUP_SETS = 'G', ZFS_DELEG_EVERYONE = 'e', ZFS_DELEG_EVERYONE_SETS = 'E', ZFS_DELEG_CREATE = 'c', ZFS_DELEG_CREATE_SETS = 'C', ZFS_DELEG_NAMED_SET = 's', ZFS_DELEG_NAMED_SET_SETS = 'S' } zfs_deleg_who_type_t; typedef enum { ZFS_DELEG_NONE = 0, ZFS_DELEG_PERM_LOCAL = 1, ZFS_DELEG_PERM_DESCENDENT = 2, ZFS_DELEG_PERM_LOCALDESCENDENT = 3, ZFS_DELEG_PERM_CREATE = 4 } zfs_deleg_inherit_t; #define ZFS_DELEG_PERM_UID "uid" #define ZFS_DELEG_PERM_GID "gid" #define ZFS_DELEG_PERM_GROUPS "groups" #define ZFS_MLSLABEL_DEFAULT "none" #define ZFS_SMB_ACL_SRC "src" #define ZFS_SMB_ACL_TARGET "target" typedef enum { ZFS_CANMOUNT_OFF = 0, ZFS_CANMOUNT_ON = 1, ZFS_CANMOUNT_NOAUTO = 2 } zfs_canmount_type_t; typedef enum { ZFS_LOGBIAS_LATENCY = 0, ZFS_LOGBIAS_THROUGHPUT = 1 } zfs_logbias_op_t; typedef enum zfs_share_op { ZFS_SHARE_NFS = 0, ZFS_UNSHARE_NFS = 1, ZFS_SHARE_SMB = 2, ZFS_UNSHARE_SMB = 3 } zfs_share_op_t; typedef enum zfs_smb_acl_op { ZFS_SMB_ACL_ADD, ZFS_SMB_ACL_REMOVE, ZFS_SMB_ACL_RENAME, ZFS_SMB_ACL_PURGE } zfs_smb_acl_op_t; typedef enum zfs_cache_type { ZFS_CACHE_NONE = 0, ZFS_CACHE_METADATA = 1, ZFS_CACHE_ALL = 2 } zfs_cache_type_t; typedef enum { ZFS_SYNC_STANDARD = 0, ZFS_SYNC_ALWAYS = 1, ZFS_SYNC_DISABLED = 2 } zfs_sync_type_t; typedef enum { ZFS_XATTR_OFF = 0, ZFS_XATTR_DIR = 1, ZFS_XATTR_SA = 2 } zfs_xattr_type_t; typedef enum { ZFS_DNSIZE_LEGACY = 0, ZFS_DNSIZE_AUTO = 1, ZFS_DNSIZE_1K = 1024, ZFS_DNSIZE_2K = 2048, ZFS_DNSIZE_4K = 4096, ZFS_DNSIZE_8K = 8192, ZFS_DNSIZE_16K = 16384 } zfs_dnsize_type_t; typedef enum { ZFS_REDUNDANT_METADATA_ALL, ZFS_REDUNDANT_METADATA_MOST, ZFS_REDUNDANT_METADATA_SOME, ZFS_REDUNDANT_METADATA_NONE } zfs_redundant_metadata_type_t; typedef enum { ZFS_VOLMODE_DEFAULT = 0, ZFS_VOLMODE_GEOM = 1, ZFS_VOLMODE_DEV = 2, ZFS_VOLMODE_NONE = 3 } zfs_volmode_t; typedef enum zfs_keystatus { ZFS_KEYSTATUS_NONE = 0, ZFS_KEYSTATUS_UNAVAILABLE, ZFS_KEYSTATUS_AVAILABLE, } zfs_keystatus_t; typedef enum zfs_keyformat { ZFS_KEYFORMAT_NONE = 0, ZFS_KEYFORMAT_RAW, ZFS_KEYFORMAT_HEX, ZFS_KEYFORMAT_PASSPHRASE, ZFS_KEYFORMAT_FORMATS } zfs_keyformat_t; typedef enum zfs_key_location { ZFS_KEYLOCATION_NONE = 0, ZFS_KEYLOCATION_PROMPT, ZFS_KEYLOCATION_URI, ZFS_KEYLOCATION_LOCATIONS } zfs_keylocation_t; +typedef enum { + ZFS_PREFETCH_NONE = 0, + ZFS_PREFETCH_METADATA = 1, + ZFS_PREFETCH_ALL = 2 +} zfs_prefetch_type_t; + #define DEFAULT_PBKDF2_ITERATIONS 350000 #define MIN_PBKDF2_ITERATIONS 100000 /* * On-disk version number. */ #define SPA_VERSION_1 1ULL #define SPA_VERSION_2 2ULL #define SPA_VERSION_3 3ULL #define SPA_VERSION_4 4ULL #define SPA_VERSION_5 5ULL #define SPA_VERSION_6 6ULL #define SPA_VERSION_7 7ULL #define SPA_VERSION_8 8ULL #define SPA_VERSION_9 9ULL #define SPA_VERSION_10 10ULL #define SPA_VERSION_11 11ULL #define SPA_VERSION_12 12ULL #define SPA_VERSION_13 13ULL #define SPA_VERSION_14 14ULL #define SPA_VERSION_15 15ULL #define SPA_VERSION_16 16ULL #define SPA_VERSION_17 17ULL #define SPA_VERSION_18 18ULL #define SPA_VERSION_19 19ULL #define SPA_VERSION_20 20ULL #define SPA_VERSION_21 21ULL #define SPA_VERSION_22 22ULL #define SPA_VERSION_23 23ULL #define SPA_VERSION_24 24ULL #define SPA_VERSION_25 25ULL #define SPA_VERSION_26 26ULL #define SPA_VERSION_27 27ULL #define SPA_VERSION_28 28ULL #define SPA_VERSION_5000 5000ULL /* * The incrementing pool version number has been replaced by pool feature * flags. For more details, see zfeature.c. */ #define SPA_VERSION SPA_VERSION_5000 #define SPA_VERSION_STRING "5000" /* * Symbolic names for the changes that caused a SPA_VERSION switch. * Used in the code when checking for presence or absence of a feature. * Feel free to define multiple symbolic names for each version if there * were multiple changes to on-disk structures during that version. * * NOTE: When checking the current SPA_VERSION in your code, be sure * to use spa_version() since it reports the version of the * last synced uberblock. Checking the in-flight version can * be dangerous in some cases. */ #define SPA_VERSION_INITIAL SPA_VERSION_1 #define SPA_VERSION_DITTO_BLOCKS SPA_VERSION_2 #define SPA_VERSION_SPARES SPA_VERSION_3 #define SPA_VERSION_RAIDZ2 SPA_VERSION_3 #define SPA_VERSION_BPOBJ_ACCOUNT SPA_VERSION_3 #define SPA_VERSION_RAIDZ_DEFLATE SPA_VERSION_3 #define SPA_VERSION_DNODE_BYTES SPA_VERSION_3 #define SPA_VERSION_ZPOOL_HISTORY SPA_VERSION_4 #define SPA_VERSION_GZIP_COMPRESSION SPA_VERSION_5 #define SPA_VERSION_BOOTFS SPA_VERSION_6 #define SPA_VERSION_SLOGS SPA_VERSION_7 #define SPA_VERSION_DELEGATED_PERMS SPA_VERSION_8 #define SPA_VERSION_FUID SPA_VERSION_9 #define SPA_VERSION_REFRESERVATION SPA_VERSION_9 #define SPA_VERSION_REFQUOTA SPA_VERSION_9 #define SPA_VERSION_UNIQUE_ACCURATE SPA_VERSION_9 #define SPA_VERSION_L2CACHE SPA_VERSION_10 #define SPA_VERSION_NEXT_CLONES SPA_VERSION_11 #define SPA_VERSION_ORIGIN SPA_VERSION_11 #define SPA_VERSION_DSL_SCRUB SPA_VERSION_11 #define SPA_VERSION_SNAP_PROPS SPA_VERSION_12 #define SPA_VERSION_USED_BREAKDOWN SPA_VERSION_13 #define SPA_VERSION_PASSTHROUGH_X SPA_VERSION_14 #define SPA_VERSION_USERSPACE SPA_VERSION_15 #define SPA_VERSION_STMF_PROP SPA_VERSION_16 #define SPA_VERSION_RAIDZ3 SPA_VERSION_17 #define SPA_VERSION_USERREFS SPA_VERSION_18 #define SPA_VERSION_HOLES SPA_VERSION_19 #define SPA_VERSION_ZLE_COMPRESSION SPA_VERSION_20 #define SPA_VERSION_DEDUP SPA_VERSION_21 #define SPA_VERSION_RECVD_PROPS SPA_VERSION_22 #define SPA_VERSION_SLIM_ZIL SPA_VERSION_23 #define SPA_VERSION_SA SPA_VERSION_24 #define SPA_VERSION_SCAN SPA_VERSION_25 #define SPA_VERSION_DIR_CLONES SPA_VERSION_26 #define SPA_VERSION_DEADLISTS SPA_VERSION_26 #define SPA_VERSION_FAST_SNAP SPA_VERSION_27 #define SPA_VERSION_MULTI_REPLACE SPA_VERSION_28 #define SPA_VERSION_BEFORE_FEATURES SPA_VERSION_28 #define SPA_VERSION_FEATURES SPA_VERSION_5000 #define SPA_VERSION_IS_SUPPORTED(v) \ (((v) >= SPA_VERSION_INITIAL && (v) <= SPA_VERSION_BEFORE_FEATURES) || \ ((v) >= SPA_VERSION_FEATURES && (v) <= SPA_VERSION)) /* * ZPL version - rev'd whenever an incompatible on-disk format change * occurs. This is independent of SPA/DMU/ZAP versioning. You must * also update the version_table[] and help message in zfs_prop.c. */ #define ZPL_VERSION_1 1ULL #define ZPL_VERSION_2 2ULL #define ZPL_VERSION_3 3ULL #define ZPL_VERSION_4 4ULL #define ZPL_VERSION_5 5ULL #define ZPL_VERSION ZPL_VERSION_5 #define ZPL_VERSION_STRING "5" #define ZPL_VERSION_INITIAL ZPL_VERSION_1 #define ZPL_VERSION_DIRENT_TYPE ZPL_VERSION_2 #define ZPL_VERSION_FUID ZPL_VERSION_3 #define ZPL_VERSION_NORMALIZATION ZPL_VERSION_3 #define ZPL_VERSION_SYSATTR ZPL_VERSION_3 #define ZPL_VERSION_USERSPACE ZPL_VERSION_4 #define ZPL_VERSION_SA ZPL_VERSION_5 /* Persistent L2ARC version */ #define L2ARC_PERSISTENT_VERSION_1 1ULL #define L2ARC_PERSISTENT_VERSION L2ARC_PERSISTENT_VERSION_1 #define L2ARC_PERSISTENT_VERSION_STRING "1" /* Rewind policy information */ #define ZPOOL_NO_REWIND 1 /* No policy - default behavior */ #define ZPOOL_NEVER_REWIND 2 /* Do not search for best txg or rewind */ #define ZPOOL_TRY_REWIND 4 /* Search for best txg, but do not rewind */ #define ZPOOL_DO_REWIND 8 /* Rewind to best txg w/in deferred frees */ #define ZPOOL_EXTREME_REWIND 16 /* Allow extreme measures to find best txg */ #define ZPOOL_REWIND_MASK 28 /* All the possible rewind bits */ #define ZPOOL_REWIND_POLICIES 31 /* All the possible policy bits */ typedef struct zpool_load_policy { uint32_t zlp_rewind; /* rewind policy requested */ uint64_t zlp_maxmeta; /* max acceptable meta-data errors */ uint64_t zlp_maxdata; /* max acceptable data errors */ uint64_t zlp_txg; /* specific txg to load */ } zpool_load_policy_t; /* * The following are configuration names used in the nvlist describing a pool's * configuration. New on-disk names should be prefixed with ":" * (e.g. "org.openzfs:") to avoid conflicting names being developed * independently. */ #define ZPOOL_CONFIG_VERSION "version" #define ZPOOL_CONFIG_POOL_NAME "name" #define ZPOOL_CONFIG_POOL_STATE "state" #define ZPOOL_CONFIG_POOL_TXG "txg" #define ZPOOL_CONFIG_POOL_GUID "pool_guid" #define ZPOOL_CONFIG_CREATE_TXG "create_txg" #define ZPOOL_CONFIG_TOP_GUID "top_guid" #define ZPOOL_CONFIG_VDEV_TREE "vdev_tree" #define ZPOOL_CONFIG_TYPE "type" #define ZPOOL_CONFIG_CHILDREN "children" #define ZPOOL_CONFIG_ID "id" #define ZPOOL_CONFIG_GUID "guid" #define ZPOOL_CONFIG_INDIRECT_OBJECT "com.delphix:indirect_object" #define ZPOOL_CONFIG_INDIRECT_BIRTHS "com.delphix:indirect_births" #define ZPOOL_CONFIG_PREV_INDIRECT_VDEV "com.delphix:prev_indirect_vdev" #define ZPOOL_CONFIG_PATH "path" #define ZPOOL_CONFIG_DEVID "devid" #define ZPOOL_CONFIG_SPARE_ID "spareid" #define ZPOOL_CONFIG_METASLAB_ARRAY "metaslab_array" #define ZPOOL_CONFIG_METASLAB_SHIFT "metaslab_shift" #define ZPOOL_CONFIG_ASHIFT "ashift" #define ZPOOL_CONFIG_ASIZE "asize" #define ZPOOL_CONFIG_DTL "DTL" #define ZPOOL_CONFIG_SCAN_STATS "scan_stats" /* not stored on disk */ #define ZPOOL_CONFIG_REMOVAL_STATS "removal_stats" /* not stored on disk */ #define ZPOOL_CONFIG_CHECKPOINT_STATS "checkpoint_stats" /* not on disk */ #define ZPOOL_CONFIG_VDEV_STATS "vdev_stats" /* not stored on disk */ #define ZPOOL_CONFIG_INDIRECT_SIZE "indirect_size" /* not stored on disk */ /* container nvlist of extended stats */ #define ZPOOL_CONFIG_VDEV_STATS_EX "vdev_stats_ex" /* Active queue read/write stats */ #define ZPOOL_CONFIG_VDEV_SYNC_R_ACTIVE_QUEUE "vdev_sync_r_active_queue" #define ZPOOL_CONFIG_VDEV_SYNC_W_ACTIVE_QUEUE "vdev_sync_w_active_queue" #define ZPOOL_CONFIG_VDEV_ASYNC_R_ACTIVE_QUEUE "vdev_async_r_active_queue" #define ZPOOL_CONFIG_VDEV_ASYNC_W_ACTIVE_QUEUE "vdev_async_w_active_queue" #define ZPOOL_CONFIG_VDEV_SCRUB_ACTIVE_QUEUE "vdev_async_scrub_active_queue" #define ZPOOL_CONFIG_VDEV_TRIM_ACTIVE_QUEUE "vdev_async_trim_active_queue" #define ZPOOL_CONFIG_VDEV_REBUILD_ACTIVE_QUEUE "vdev_rebuild_active_queue" /* Queue sizes */ #define ZPOOL_CONFIG_VDEV_SYNC_R_PEND_QUEUE "vdev_sync_r_pend_queue" #define ZPOOL_CONFIG_VDEV_SYNC_W_PEND_QUEUE "vdev_sync_w_pend_queue" #define ZPOOL_CONFIG_VDEV_ASYNC_R_PEND_QUEUE "vdev_async_r_pend_queue" #define ZPOOL_CONFIG_VDEV_ASYNC_W_PEND_QUEUE "vdev_async_w_pend_queue" #define ZPOOL_CONFIG_VDEV_SCRUB_PEND_QUEUE "vdev_async_scrub_pend_queue" #define ZPOOL_CONFIG_VDEV_TRIM_PEND_QUEUE "vdev_async_trim_pend_queue" #define ZPOOL_CONFIG_VDEV_REBUILD_PEND_QUEUE "vdev_rebuild_pend_queue" /* Latency read/write histogram stats */ #define ZPOOL_CONFIG_VDEV_TOT_R_LAT_HISTO "vdev_tot_r_lat_histo" #define ZPOOL_CONFIG_VDEV_TOT_W_LAT_HISTO "vdev_tot_w_lat_histo" #define ZPOOL_CONFIG_VDEV_DISK_R_LAT_HISTO "vdev_disk_r_lat_histo" #define ZPOOL_CONFIG_VDEV_DISK_W_LAT_HISTO "vdev_disk_w_lat_histo" #define ZPOOL_CONFIG_VDEV_SYNC_R_LAT_HISTO "vdev_sync_r_lat_histo" #define ZPOOL_CONFIG_VDEV_SYNC_W_LAT_HISTO "vdev_sync_w_lat_histo" #define ZPOOL_CONFIG_VDEV_ASYNC_R_LAT_HISTO "vdev_async_r_lat_histo" #define ZPOOL_CONFIG_VDEV_ASYNC_W_LAT_HISTO "vdev_async_w_lat_histo" #define ZPOOL_CONFIG_VDEV_SCRUB_LAT_HISTO "vdev_scrub_histo" #define ZPOOL_CONFIG_VDEV_TRIM_LAT_HISTO "vdev_trim_histo" #define ZPOOL_CONFIG_VDEV_REBUILD_LAT_HISTO "vdev_rebuild_histo" /* Request size histograms */ #define ZPOOL_CONFIG_VDEV_SYNC_IND_R_HISTO "vdev_sync_ind_r_histo" #define ZPOOL_CONFIG_VDEV_SYNC_IND_W_HISTO "vdev_sync_ind_w_histo" #define ZPOOL_CONFIG_VDEV_ASYNC_IND_R_HISTO "vdev_async_ind_r_histo" #define ZPOOL_CONFIG_VDEV_ASYNC_IND_W_HISTO "vdev_async_ind_w_histo" #define ZPOOL_CONFIG_VDEV_IND_SCRUB_HISTO "vdev_ind_scrub_histo" #define ZPOOL_CONFIG_VDEV_IND_TRIM_HISTO "vdev_ind_trim_histo" #define ZPOOL_CONFIG_VDEV_IND_REBUILD_HISTO "vdev_ind_rebuild_histo" #define ZPOOL_CONFIG_VDEV_SYNC_AGG_R_HISTO "vdev_sync_agg_r_histo" #define ZPOOL_CONFIG_VDEV_SYNC_AGG_W_HISTO "vdev_sync_agg_w_histo" #define ZPOOL_CONFIG_VDEV_ASYNC_AGG_R_HISTO "vdev_async_agg_r_histo" #define ZPOOL_CONFIG_VDEV_ASYNC_AGG_W_HISTO "vdev_async_agg_w_histo" #define ZPOOL_CONFIG_VDEV_AGG_SCRUB_HISTO "vdev_agg_scrub_histo" #define ZPOOL_CONFIG_VDEV_AGG_TRIM_HISTO "vdev_agg_trim_histo" #define ZPOOL_CONFIG_VDEV_AGG_REBUILD_HISTO "vdev_agg_rebuild_histo" /* Number of slow IOs */ #define ZPOOL_CONFIG_VDEV_SLOW_IOS "vdev_slow_ios" /* vdev enclosure sysfs path */ #define ZPOOL_CONFIG_VDEV_ENC_SYSFS_PATH "vdev_enc_sysfs_path" #define ZPOOL_CONFIG_WHOLE_DISK "whole_disk" #define ZPOOL_CONFIG_ERRCOUNT "error_count" #define ZPOOL_CONFIG_NOT_PRESENT "not_present" #define ZPOOL_CONFIG_SPARES "spares" #define ZPOOL_CONFIG_IS_SPARE "is_spare" #define ZPOOL_CONFIG_NPARITY "nparity" #define ZPOOL_CONFIG_HOSTID "hostid" #define ZPOOL_CONFIG_HOSTNAME "hostname" #define ZPOOL_CONFIG_LOADED_TIME "initial_load_time" #define ZPOOL_CONFIG_UNSPARE "unspare" #define ZPOOL_CONFIG_PHYS_PATH "phys_path" #define ZPOOL_CONFIG_IS_LOG "is_log" #define ZPOOL_CONFIG_L2CACHE "l2cache" #define ZPOOL_CONFIG_HOLE_ARRAY "hole_array" #define ZPOOL_CONFIG_VDEV_CHILDREN "vdev_children" #define ZPOOL_CONFIG_IS_HOLE "is_hole" #define ZPOOL_CONFIG_DDT_HISTOGRAM "ddt_histogram" #define ZPOOL_CONFIG_DDT_OBJ_STATS "ddt_object_stats" #define ZPOOL_CONFIG_DDT_STATS "ddt_stats" #define ZPOOL_CONFIG_SPLIT "splitcfg" #define ZPOOL_CONFIG_ORIG_GUID "orig_guid" #define ZPOOL_CONFIG_SPLIT_GUID "split_guid" #define ZPOOL_CONFIG_SPLIT_LIST "guid_list" #define ZPOOL_CONFIG_NONALLOCATING "non_allocating" #define ZPOOL_CONFIG_REMOVING "removing" #define ZPOOL_CONFIG_RESILVER_TXG "resilver_txg" #define ZPOOL_CONFIG_REBUILD_TXG "rebuild_txg" #define ZPOOL_CONFIG_COMMENT "comment" #define ZPOOL_CONFIG_SUSPENDED "suspended" /* not stored on disk */ #define ZPOOL_CONFIG_SUSPENDED_REASON "suspended_reason" /* not stored */ #define ZPOOL_CONFIG_TIMESTAMP "timestamp" /* not stored on disk */ #define ZPOOL_CONFIG_BOOTFS "bootfs" /* not stored on disk */ #define ZPOOL_CONFIG_MISSING_DEVICES "missing_vdevs" /* not stored on disk */ #define ZPOOL_CONFIG_LOAD_INFO "load_info" /* not stored on disk */ #define ZPOOL_CONFIG_REWIND_INFO "rewind_info" /* not stored on disk */ #define ZPOOL_CONFIG_UNSUP_FEAT "unsup_feat" /* not stored on disk */ #define ZPOOL_CONFIG_ENABLED_FEAT "enabled_feat" /* not stored on disk */ #define ZPOOL_CONFIG_CAN_RDONLY "can_rdonly" /* not stored on disk */ #define ZPOOL_CONFIG_FEATURES_FOR_READ "features_for_read" #define ZPOOL_CONFIG_FEATURE_STATS "feature_stats" /* not stored on disk */ #define ZPOOL_CONFIG_ERRATA "errata" /* not stored on disk */ #define ZPOOL_CONFIG_VDEV_ROOT_ZAP "com.klarasystems:vdev_zap_root" #define ZPOOL_CONFIG_VDEV_TOP_ZAP "com.delphix:vdev_zap_top" #define ZPOOL_CONFIG_VDEV_LEAF_ZAP "com.delphix:vdev_zap_leaf" #define ZPOOL_CONFIG_HAS_PER_VDEV_ZAPS "com.delphix:has_per_vdev_zaps" #define ZPOOL_CONFIG_RESILVER_DEFER "com.datto:resilver_defer" #define ZPOOL_CONFIG_CACHEFILE "cachefile" /* not stored on disk */ #define ZPOOL_CONFIG_MMP_STATE "mmp_state" /* not stored on disk */ #define ZPOOL_CONFIG_MMP_TXG "mmp_txg" /* not stored on disk */ #define ZPOOL_CONFIG_MMP_SEQ "mmp_seq" /* not stored on disk */ #define ZPOOL_CONFIG_MMP_HOSTNAME "mmp_hostname" /* not stored on disk */ #define ZPOOL_CONFIG_MMP_HOSTID "mmp_hostid" /* not stored on disk */ #define ZPOOL_CONFIG_ALLOCATION_BIAS "alloc_bias" /* not stored on disk */ #define ZPOOL_CONFIG_EXPANSION_TIME "expansion_time" /* not stored */ #define ZPOOL_CONFIG_REBUILD_STATS "org.openzfs:rebuild_stats" #define ZPOOL_CONFIG_COMPATIBILITY "compatibility" /* * The persistent vdev state is stored as separate values rather than a single * 'vdev_state' entry. This is because a device can be in multiple states, such * as offline and degraded. */ #define ZPOOL_CONFIG_OFFLINE "offline" #define ZPOOL_CONFIG_FAULTED "faulted" #define ZPOOL_CONFIG_DEGRADED "degraded" #define ZPOOL_CONFIG_REMOVED "removed" #define ZPOOL_CONFIG_FRU "fru" #define ZPOOL_CONFIG_AUX_STATE "aux_state" /* Pool load policy parameters */ #define ZPOOL_LOAD_POLICY "load-policy" #define ZPOOL_LOAD_REWIND_POLICY "load-rewind-policy" #define ZPOOL_LOAD_REQUEST_TXG "load-request-txg" #define ZPOOL_LOAD_META_THRESH "load-meta-thresh" #define ZPOOL_LOAD_DATA_THRESH "load-data-thresh" /* Rewind data discovered */ #define ZPOOL_CONFIG_LOAD_TIME "rewind_txg_ts" #define ZPOOL_CONFIG_LOAD_META_ERRORS "verify_meta_errors" #define ZPOOL_CONFIG_LOAD_DATA_ERRORS "verify_data_errors" #define ZPOOL_CONFIG_REWIND_TIME "seconds_of_rewind" /* dRAID configuration */ #define ZPOOL_CONFIG_DRAID_NDATA "draid_ndata" #define ZPOOL_CONFIG_DRAID_NSPARES "draid_nspares" #define ZPOOL_CONFIG_DRAID_NGROUPS "draid_ngroups" #define VDEV_TYPE_ROOT "root" #define VDEV_TYPE_MIRROR "mirror" #define VDEV_TYPE_REPLACING "replacing" #define VDEV_TYPE_RAIDZ "raidz" #define VDEV_TYPE_DRAID "draid" #define VDEV_TYPE_DRAID_SPARE "dspare" #define VDEV_TYPE_DISK "disk" #define VDEV_TYPE_FILE "file" #define VDEV_TYPE_MISSING "missing" #define VDEV_TYPE_HOLE "hole" #define VDEV_TYPE_SPARE "spare" #define VDEV_TYPE_LOG "log" #define VDEV_TYPE_L2CACHE "l2cache" #define VDEV_TYPE_INDIRECT "indirect" #define VDEV_RAIDZ_MAXPARITY 3 #define VDEV_DRAID_MAXPARITY 3 #define VDEV_DRAID_MIN_CHILDREN 2 #define VDEV_DRAID_MAX_CHILDREN UINT8_MAX /* VDEV_TOP_ZAP_* are used in top-level vdev ZAP objects. */ #define VDEV_TOP_ZAP_INDIRECT_OBSOLETE_SM \ "com.delphix:indirect_obsolete_sm" #define VDEV_TOP_ZAP_OBSOLETE_COUNTS_ARE_PRECISE \ "com.delphix:obsolete_counts_are_precise" #define VDEV_TOP_ZAP_POOL_CHECKPOINT_SM \ "com.delphix:pool_checkpoint_sm" #define VDEV_TOP_ZAP_MS_UNFLUSHED_PHYS_TXGS \ "com.delphix:ms_unflushed_phys_txgs" #define VDEV_TOP_ZAP_VDEV_REBUILD_PHYS \ "org.openzfs:vdev_rebuild" #define VDEV_TOP_ZAP_ALLOCATION_BIAS \ "org.zfsonlinux:allocation_bias" /* vdev metaslab allocation bias */ #define VDEV_ALLOC_BIAS_LOG "log" #define VDEV_ALLOC_BIAS_SPECIAL "special" #define VDEV_ALLOC_BIAS_DEDUP "dedup" /* vdev initialize state */ #define VDEV_LEAF_ZAP_INITIALIZE_LAST_OFFSET \ "com.delphix:next_offset_to_initialize" #define VDEV_LEAF_ZAP_INITIALIZE_STATE \ "com.delphix:vdev_initialize_state" #define VDEV_LEAF_ZAP_INITIALIZE_ACTION_TIME \ "com.delphix:vdev_initialize_action_time" /* vdev TRIM state */ #define VDEV_LEAF_ZAP_TRIM_LAST_OFFSET \ "org.zfsonlinux:next_offset_to_trim" #define VDEV_LEAF_ZAP_TRIM_STATE \ "org.zfsonlinux:vdev_trim_state" #define VDEV_LEAF_ZAP_TRIM_ACTION_TIME \ "org.zfsonlinux:vdev_trim_action_time" #define VDEV_LEAF_ZAP_TRIM_RATE \ "org.zfsonlinux:vdev_trim_rate" #define VDEV_LEAF_ZAP_TRIM_PARTIAL \ "org.zfsonlinux:vdev_trim_partial" #define VDEV_LEAF_ZAP_TRIM_SECURE \ "org.zfsonlinux:vdev_trim_secure" /* * This is needed in userland to report the minimum necessary device size. */ #define SPA_MINDEVSIZE (64ULL << 20) /* * Set if the fragmentation has not yet been calculated. This can happen * because the space maps have not been upgraded or the histogram feature * is not enabled. */ #define ZFS_FRAG_INVALID UINT64_MAX /* * The location of the pool configuration repository, shared between kernel and * userland. */ #define ZPOOL_CACHE_BOOT "/boot/zfs/zpool.cache" #define ZPOOL_CACHE "/etc/zfs/zpool.cache" /* * Settings for zpool compatibility features files */ #define ZPOOL_SYSCONF_COMPAT_D SYSCONFDIR "/zfs/compatibility.d" #define ZPOOL_DATA_COMPAT_D PKGDATADIR "/compatibility.d" #define ZPOOL_COMPAT_MAXSIZE 16384 /* * Hard-wired compatibility settings */ #define ZPOOL_COMPAT_LEGACY "legacy" #define ZPOOL_COMPAT_OFF "off" /* * vdev states are ordered from least to most healthy. * A vdev that's CANT_OPEN or below is considered unusable. */ typedef enum vdev_state { VDEV_STATE_UNKNOWN = 0, /* Uninitialized vdev */ VDEV_STATE_CLOSED, /* Not currently open */ VDEV_STATE_OFFLINE, /* Not allowed to open */ VDEV_STATE_REMOVED, /* Explicitly removed from system */ VDEV_STATE_CANT_OPEN, /* Tried to open, but failed */ VDEV_STATE_FAULTED, /* External request to fault device */ VDEV_STATE_DEGRADED, /* Replicated vdev with unhealthy kids */ VDEV_STATE_HEALTHY /* Presumed good */ } vdev_state_t; #define VDEV_STATE_ONLINE VDEV_STATE_HEALTHY /* * vdev aux states. When a vdev is in the CANT_OPEN state, the aux field * of the vdev stats structure uses these constants to distinguish why. */ typedef enum vdev_aux { VDEV_AUX_NONE, /* no error */ VDEV_AUX_OPEN_FAILED, /* ldi_open_*() or vn_open() failed */ VDEV_AUX_CORRUPT_DATA, /* bad label or disk contents */ VDEV_AUX_NO_REPLICAS, /* insufficient number of replicas */ VDEV_AUX_BAD_GUID_SUM, /* vdev guid sum doesn't match */ VDEV_AUX_TOO_SMALL, /* vdev size is too small */ VDEV_AUX_BAD_LABEL, /* the label is OK but invalid */ VDEV_AUX_VERSION_NEWER, /* on-disk version is too new */ VDEV_AUX_VERSION_OLDER, /* on-disk version is too old */ VDEV_AUX_UNSUP_FEAT, /* unsupported features */ VDEV_AUX_SPARED, /* hot spare used in another pool */ VDEV_AUX_ERR_EXCEEDED, /* too many errors */ VDEV_AUX_IO_FAILURE, /* experienced I/O failure */ VDEV_AUX_BAD_LOG, /* cannot read log chain(s) */ VDEV_AUX_EXTERNAL, /* external diagnosis or forced fault */ VDEV_AUX_SPLIT_POOL, /* vdev was split off into another pool */ VDEV_AUX_BAD_ASHIFT, /* vdev ashift is invalid */ VDEV_AUX_EXTERNAL_PERSIST, /* persistent forced fault */ VDEV_AUX_ACTIVE, /* vdev active on a different host */ VDEV_AUX_CHILDREN_OFFLINE, /* all children are offline */ VDEV_AUX_ASHIFT_TOO_BIG, /* vdev's min block size is too large */ } vdev_aux_t; /* * pool state. The following states are written to disk as part of the normal * SPA lifecycle: ACTIVE, EXPORTED, DESTROYED, SPARE, L2CACHE. The remaining * states are software abstractions used at various levels to communicate * pool state. */ typedef enum pool_state { POOL_STATE_ACTIVE = 0, /* In active use */ POOL_STATE_EXPORTED, /* Explicitly exported */ POOL_STATE_DESTROYED, /* Explicitly destroyed */ POOL_STATE_SPARE, /* Reserved for hot spare use */ POOL_STATE_L2CACHE, /* Level 2 ARC device */ POOL_STATE_UNINITIALIZED, /* Internal spa_t state */ POOL_STATE_UNAVAIL, /* Internal libzfs state */ POOL_STATE_POTENTIALLY_ACTIVE /* Internal libzfs state */ } pool_state_t; /* * mmp state. The following states provide additional detail describing * why a pool couldn't be safely imported. */ typedef enum mmp_state { MMP_STATE_ACTIVE = 0, /* In active use */ MMP_STATE_INACTIVE, /* Inactive and safe to import */ MMP_STATE_NO_HOSTID /* System hostid is not set */ } mmp_state_t; /* * Scan Functions. */ typedef enum pool_scan_func { POOL_SCAN_NONE, POOL_SCAN_SCRUB, POOL_SCAN_RESILVER, POOL_SCAN_ERRORSCRUB, POOL_SCAN_FUNCS } pool_scan_func_t; /* * Used to control scrub pause and resume. */ typedef enum pool_scrub_cmd { POOL_SCRUB_NORMAL = 0, POOL_SCRUB_PAUSE, POOL_SCRUB_FLAGS_END } pool_scrub_cmd_t; typedef enum { CS_NONE, CS_CHECKPOINT_EXISTS, CS_CHECKPOINT_DISCARDING, CS_NUM_STATES } checkpoint_state_t; typedef struct pool_checkpoint_stat { uint64_t pcs_state; /* checkpoint_state_t */ uint64_t pcs_start_time; /* time checkpoint/discard started */ uint64_t pcs_space; /* checkpointed space */ } pool_checkpoint_stat_t; /* * ZIO types. Needed to interpret vdev statistics below. */ typedef enum zio_type { ZIO_TYPE_NULL = 0, ZIO_TYPE_READ, ZIO_TYPE_WRITE, ZIO_TYPE_FREE, ZIO_TYPE_CLAIM, ZIO_TYPE_IOCTL, ZIO_TYPE_TRIM, ZIO_TYPES } zio_type_t; /* * Pool statistics. Note: all fields should be 64-bit because this * is passed between kernel and userland as an nvlist uint64 array. */ typedef struct pool_scan_stat { /* values stored on disk */ uint64_t pss_func; /* pool_scan_func_t */ uint64_t pss_state; /* dsl_scan_state_t */ uint64_t pss_start_time; /* scan start time */ uint64_t pss_end_time; /* scan end time */ uint64_t pss_to_examine; /* total bytes to scan */ uint64_t pss_examined; /* total bytes located by scanner */ uint64_t pss_skipped; /* total bytes skipped by scanner */ uint64_t pss_processed; /* total processed bytes */ uint64_t pss_errors; /* scan errors */ /* values not stored on disk */ uint64_t pss_pass_exam; /* examined bytes per scan pass */ uint64_t pss_pass_start; /* start time of a scan pass */ uint64_t pss_pass_scrub_pause; /* pause time of a scrub pass */ /* cumulative time scrub spent paused, needed for rate calculation */ uint64_t pss_pass_scrub_spent_paused; uint64_t pss_pass_issued; /* issued bytes per scan pass */ uint64_t pss_issued; /* total bytes checked by scanner */ /* error scrub values stored on disk */ uint64_t pss_error_scrub_func; /* pool_scan_func_t */ uint64_t pss_error_scrub_state; /* dsl_scan_state_t */ uint64_t pss_error_scrub_start; /* error scrub start time */ uint64_t pss_error_scrub_end; /* error scrub end time */ uint64_t pss_error_scrub_examined; /* error blocks issued I/O */ /* error blocks to be issued I/O */ uint64_t pss_error_scrub_to_be_examined; /* error scrub values not stored on disk */ /* error scrub pause time in milliseconds */ uint64_t pss_pass_error_scrub_pause; } pool_scan_stat_t; typedef struct pool_removal_stat { uint64_t prs_state; /* dsl_scan_state_t */ uint64_t prs_removing_vdev; uint64_t prs_start_time; uint64_t prs_end_time; uint64_t prs_to_copy; /* bytes that need to be copied */ uint64_t prs_copied; /* bytes copied so far */ /* * bytes of memory used for indirect mappings. * This includes all removed vdevs. */ uint64_t prs_mapping_memory; } pool_removal_stat_t; typedef enum dsl_scan_state { DSS_NONE, DSS_SCANNING, DSS_FINISHED, DSS_CANCELED, DSS_ERRORSCRUBBING, DSS_NUM_STATES } dsl_scan_state_t; typedef struct vdev_rebuild_stat { uint64_t vrs_state; /* vdev_rebuild_state_t */ uint64_t vrs_start_time; /* time_t */ uint64_t vrs_end_time; /* time_t */ uint64_t vrs_scan_time_ms; /* total run time (millisecs) */ uint64_t vrs_bytes_scanned; /* allocated bytes scanned */ uint64_t vrs_bytes_issued; /* read bytes issued */ uint64_t vrs_bytes_rebuilt; /* rebuilt bytes */ uint64_t vrs_bytes_est; /* total bytes to scan */ uint64_t vrs_errors; /* scanning errors */ uint64_t vrs_pass_time_ms; /* pass run time (millisecs) */ uint64_t vrs_pass_bytes_scanned; /* bytes scanned since start/resume */ uint64_t vrs_pass_bytes_issued; /* bytes rebuilt since start/resume */ uint64_t vrs_pass_bytes_skipped; /* bytes skipped since start/resume */ } vdev_rebuild_stat_t; /* * Errata described by https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-ER. * The ordering of this enum must be maintained to ensure the errata identifiers * map to the correct documentation. New errata may only be appended to the * list and must contain corresponding documentation at the above link. */ typedef enum zpool_errata { ZPOOL_ERRATA_NONE, ZPOOL_ERRATA_ZOL_2094_SCRUB, ZPOOL_ERRATA_ZOL_2094_ASYNC_DESTROY, ZPOOL_ERRATA_ZOL_6845_ENCRYPTION, ZPOOL_ERRATA_ZOL_8308_ENCRYPTION, } zpool_errata_t; /* * Vdev statistics. Note: all fields should be 64-bit because this * is passed between kernel and user land as an nvlist uint64 array. * * The vs_ops[] and vs_bytes[] arrays must always be an array size of 6 in * order to keep subsequent members at their known fixed offsets. When * adding a new field it must be added to the end the structure. */ #define VS_ZIO_TYPES 6 typedef struct vdev_stat { hrtime_t vs_timestamp; /* time since vdev load */ uint64_t vs_state; /* vdev state */ uint64_t vs_aux; /* see vdev_aux_t */ uint64_t vs_alloc; /* space allocated */ uint64_t vs_space; /* total capacity */ uint64_t vs_dspace; /* deflated capacity */ uint64_t vs_rsize; /* replaceable dev size */ uint64_t vs_esize; /* expandable dev size */ uint64_t vs_ops[VS_ZIO_TYPES]; /* operation count */ uint64_t vs_bytes[VS_ZIO_TYPES]; /* bytes read/written */ uint64_t vs_read_errors; /* read errors */ uint64_t vs_write_errors; /* write errors */ uint64_t vs_checksum_errors; /* checksum errors */ uint64_t vs_initialize_errors; /* initializing errors */ uint64_t vs_self_healed; /* self-healed bytes */ uint64_t vs_scan_removing; /* removing? */ uint64_t vs_scan_processed; /* scan processed bytes */ uint64_t vs_fragmentation; /* device fragmentation */ uint64_t vs_initialize_bytes_done; /* bytes initialized */ uint64_t vs_initialize_bytes_est; /* total bytes to initialize */ uint64_t vs_initialize_state; /* vdev_initializing_state_t */ uint64_t vs_initialize_action_time; /* time_t */ uint64_t vs_checkpoint_space; /* checkpoint-consumed space */ uint64_t vs_resilver_deferred; /* resilver deferred */ uint64_t vs_slow_ios; /* slow IOs */ uint64_t vs_trim_errors; /* trimming errors */ uint64_t vs_trim_notsup; /* supported by device */ uint64_t vs_trim_bytes_done; /* bytes trimmed */ uint64_t vs_trim_bytes_est; /* total bytes to trim */ uint64_t vs_trim_state; /* vdev_trim_state_t */ uint64_t vs_trim_action_time; /* time_t */ uint64_t vs_rebuild_processed; /* bytes rebuilt */ uint64_t vs_configured_ashift; /* TLV vdev_ashift */ uint64_t vs_logical_ashift; /* vdev_logical_ashift */ uint64_t vs_physical_ashift; /* vdev_physical_ashift */ uint64_t vs_noalloc; /* allocations halted? */ uint64_t vs_pspace; /* physical capacity */ } vdev_stat_t; #define VDEV_STAT_VALID(field, uint64_t_field_count) \ ((uint64_t_field_count * sizeof (uint64_t)) >= \ (offsetof(vdev_stat_t, field) + sizeof (((vdev_stat_t *)NULL)->field))) /* * Extended stats * * These are stats which aren't included in the original iostat output. For * convenience, they are grouped together in vdev_stat_ex, although each stat * is individually exported as an nvlist. */ typedef struct vdev_stat_ex { /* Number of ZIOs issued to disk and waiting to finish */ uint64_t vsx_active_queue[ZIO_PRIORITY_NUM_QUEUEABLE]; /* Number of ZIOs pending to be issued to disk */ uint64_t vsx_pend_queue[ZIO_PRIORITY_NUM_QUEUEABLE]; /* * Below are the histograms for various latencies. Buckets are in * units of nanoseconds. */ /* * 2^37 nanoseconds = 134s. Timeouts will probably start kicking in * before this. */ #define VDEV_L_HISTO_BUCKETS 37 /* Latency histo buckets */ #define VDEV_RQ_HISTO_BUCKETS 25 /* Request size histo buckets */ /* Amount of time in ZIO queue (ns) */ uint64_t vsx_queue_histo[ZIO_PRIORITY_NUM_QUEUEABLE] [VDEV_L_HISTO_BUCKETS]; /* Total ZIO latency (ns). Includes queuing and disk access time */ uint64_t vsx_total_histo[ZIO_TYPES][VDEV_L_HISTO_BUCKETS]; /* Amount of time to read/write the disk (ns) */ uint64_t vsx_disk_histo[ZIO_TYPES][VDEV_L_HISTO_BUCKETS]; /* "lookup the bucket for a value" histogram macros */ #define HISTO(val, buckets) (val != 0 ? MIN(highbit64(val) - 1, \ buckets - 1) : 0) #define L_HISTO(a) HISTO(a, VDEV_L_HISTO_BUCKETS) #define RQ_HISTO(a) HISTO(a, VDEV_RQ_HISTO_BUCKETS) /* Physical IO histogram */ uint64_t vsx_ind_histo[ZIO_PRIORITY_NUM_QUEUEABLE] [VDEV_RQ_HISTO_BUCKETS]; /* Delegated (aggregated) physical IO histogram */ uint64_t vsx_agg_histo[ZIO_PRIORITY_NUM_QUEUEABLE] [VDEV_RQ_HISTO_BUCKETS]; } vdev_stat_ex_t; /* * Initialize functions. */ typedef enum pool_initialize_func { POOL_INITIALIZE_START, POOL_INITIALIZE_CANCEL, POOL_INITIALIZE_SUSPEND, POOL_INITIALIZE_UNINIT, POOL_INITIALIZE_FUNCS } pool_initialize_func_t; /* * TRIM functions. */ typedef enum pool_trim_func { POOL_TRIM_START, POOL_TRIM_CANCEL, POOL_TRIM_SUSPEND, POOL_TRIM_FUNCS } pool_trim_func_t; /* * DDT statistics. Note: all fields should be 64-bit because this * is passed between kernel and userland as an nvlist uint64 array. */ typedef struct ddt_object { uint64_t ddo_count; /* number of elements in ddt */ uint64_t ddo_dspace; /* size of ddt on disk */ uint64_t ddo_mspace; /* size of ddt in-core */ } ddt_object_t; typedef struct ddt_stat { uint64_t dds_blocks; /* blocks */ uint64_t dds_lsize; /* logical size */ uint64_t dds_psize; /* physical size */ uint64_t dds_dsize; /* deflated allocated size */ uint64_t dds_ref_blocks; /* referenced blocks */ uint64_t dds_ref_lsize; /* referenced lsize * refcnt */ uint64_t dds_ref_psize; /* referenced psize * refcnt */ uint64_t dds_ref_dsize; /* referenced dsize * refcnt */ } ddt_stat_t; typedef struct ddt_histogram { ddt_stat_t ddh_stat[64]; /* power-of-two histogram buckets */ } ddt_histogram_t; #define ZVOL_DRIVER "zvol" #define ZFS_DRIVER "zfs" #define ZFS_DEV "/dev/zfs" #define ZFS_DEVDIR "/dev" #define ZFS_SUPER_MAGIC 0x2fc12fc1 /* general zvol path */ #define ZVOL_DIR "/dev/zvol/" #define ZVOL_MAJOR 230 #define ZVOL_MINOR_BITS 4 #define ZVOL_MINOR_MASK ((1U << ZVOL_MINOR_BITS) - 1) #define ZVOL_MINORS (1 << 4) #define ZVOL_DEV_NAME "zd" #define ZVOL_PROP_NAME "name" #define ZVOL_DEFAULT_BLOCKSIZE 16384 typedef enum { VDEV_INITIALIZE_NONE, VDEV_INITIALIZE_ACTIVE, VDEV_INITIALIZE_CANCELED, VDEV_INITIALIZE_SUSPENDED, VDEV_INITIALIZE_COMPLETE } vdev_initializing_state_t; typedef enum { VDEV_TRIM_NONE, VDEV_TRIM_ACTIVE, VDEV_TRIM_CANCELED, VDEV_TRIM_SUSPENDED, VDEV_TRIM_COMPLETE, } vdev_trim_state_t; typedef enum { VDEV_REBUILD_NONE, VDEV_REBUILD_ACTIVE, VDEV_REBUILD_CANCELED, VDEV_REBUILD_COMPLETE, } vdev_rebuild_state_t; /* * nvlist name constants. Facilitate restricting snapshot iteration range for * the "list next snapshot" ioctl */ #define SNAP_ITER_MIN_TXG "snap_iter_min_txg" #define SNAP_ITER_MAX_TXG "snap_iter_max_txg" /* * /dev/zfs ioctl numbers. * * These numbers cannot change over time. New ioctl numbers must be appended. */ typedef enum zfs_ioc { /* * Core features - 88/128 numbers reserved. */ #ifdef __FreeBSD__ ZFS_IOC_FIRST = 0, #else ZFS_IOC_FIRST = ('Z' << 8), #endif ZFS_IOC = ZFS_IOC_FIRST, ZFS_IOC_POOL_CREATE = ZFS_IOC_FIRST, /* 0x5a00 */ ZFS_IOC_POOL_DESTROY, /* 0x5a01 */ ZFS_IOC_POOL_IMPORT, /* 0x5a02 */ ZFS_IOC_POOL_EXPORT, /* 0x5a03 */ ZFS_IOC_POOL_CONFIGS, /* 0x5a04 */ ZFS_IOC_POOL_STATS, /* 0x5a05 */ ZFS_IOC_POOL_TRYIMPORT, /* 0x5a06 */ ZFS_IOC_POOL_SCAN, /* 0x5a07 */ ZFS_IOC_POOL_FREEZE, /* 0x5a08 */ ZFS_IOC_POOL_UPGRADE, /* 0x5a09 */ ZFS_IOC_POOL_GET_HISTORY, /* 0x5a0a */ ZFS_IOC_VDEV_ADD, /* 0x5a0b */ ZFS_IOC_VDEV_REMOVE, /* 0x5a0c */ ZFS_IOC_VDEV_SET_STATE, /* 0x5a0d */ ZFS_IOC_VDEV_ATTACH, /* 0x5a0e */ ZFS_IOC_VDEV_DETACH, /* 0x5a0f */ ZFS_IOC_VDEV_SETPATH, /* 0x5a10 */ ZFS_IOC_VDEV_SETFRU, /* 0x5a11 */ ZFS_IOC_OBJSET_STATS, /* 0x5a12 */ ZFS_IOC_OBJSET_ZPLPROPS, /* 0x5a13 */ ZFS_IOC_DATASET_LIST_NEXT, /* 0x5a14 */ ZFS_IOC_SNAPSHOT_LIST_NEXT, /* 0x5a15 */ ZFS_IOC_SET_PROP, /* 0x5a16 */ ZFS_IOC_CREATE, /* 0x5a17 */ ZFS_IOC_DESTROY, /* 0x5a18 */ ZFS_IOC_ROLLBACK, /* 0x5a19 */ ZFS_IOC_RENAME, /* 0x5a1a */ ZFS_IOC_RECV, /* 0x5a1b */ ZFS_IOC_SEND, /* 0x5a1c */ ZFS_IOC_INJECT_FAULT, /* 0x5a1d */ ZFS_IOC_CLEAR_FAULT, /* 0x5a1e */ ZFS_IOC_INJECT_LIST_NEXT, /* 0x5a1f */ ZFS_IOC_ERROR_LOG, /* 0x5a20 */ ZFS_IOC_CLEAR, /* 0x5a21 */ ZFS_IOC_PROMOTE, /* 0x5a22 */ ZFS_IOC_SNAPSHOT, /* 0x5a23 */ ZFS_IOC_DSOBJ_TO_DSNAME, /* 0x5a24 */ ZFS_IOC_OBJ_TO_PATH, /* 0x5a25 */ ZFS_IOC_POOL_SET_PROPS, /* 0x5a26 */ ZFS_IOC_POOL_GET_PROPS, /* 0x5a27 */ ZFS_IOC_SET_FSACL, /* 0x5a28 */ ZFS_IOC_GET_FSACL, /* 0x5a29 */ ZFS_IOC_SHARE, /* 0x5a2a */ ZFS_IOC_INHERIT_PROP, /* 0x5a2b */ ZFS_IOC_SMB_ACL, /* 0x5a2c */ ZFS_IOC_USERSPACE_ONE, /* 0x5a2d */ ZFS_IOC_USERSPACE_MANY, /* 0x5a2e */ ZFS_IOC_USERSPACE_UPGRADE, /* 0x5a2f */ ZFS_IOC_HOLD, /* 0x5a30 */ ZFS_IOC_RELEASE, /* 0x5a31 */ ZFS_IOC_GET_HOLDS, /* 0x5a32 */ ZFS_IOC_OBJSET_RECVD_PROPS, /* 0x5a33 */ ZFS_IOC_VDEV_SPLIT, /* 0x5a34 */ ZFS_IOC_NEXT_OBJ, /* 0x5a35 */ ZFS_IOC_DIFF, /* 0x5a36 */ ZFS_IOC_TMP_SNAPSHOT, /* 0x5a37 */ ZFS_IOC_OBJ_TO_STATS, /* 0x5a38 */ ZFS_IOC_SPACE_WRITTEN, /* 0x5a39 */ ZFS_IOC_SPACE_SNAPS, /* 0x5a3a */ ZFS_IOC_DESTROY_SNAPS, /* 0x5a3b */ ZFS_IOC_POOL_REGUID, /* 0x5a3c */ ZFS_IOC_POOL_REOPEN, /* 0x5a3d */ ZFS_IOC_SEND_PROGRESS, /* 0x5a3e */ ZFS_IOC_LOG_HISTORY, /* 0x5a3f */ ZFS_IOC_SEND_NEW, /* 0x5a40 */ ZFS_IOC_SEND_SPACE, /* 0x5a41 */ ZFS_IOC_CLONE, /* 0x5a42 */ ZFS_IOC_BOOKMARK, /* 0x5a43 */ ZFS_IOC_GET_BOOKMARKS, /* 0x5a44 */ ZFS_IOC_DESTROY_BOOKMARKS, /* 0x5a45 */ ZFS_IOC_RECV_NEW, /* 0x5a46 */ ZFS_IOC_POOL_SYNC, /* 0x5a47 */ ZFS_IOC_CHANNEL_PROGRAM, /* 0x5a48 */ ZFS_IOC_LOAD_KEY, /* 0x5a49 */ ZFS_IOC_UNLOAD_KEY, /* 0x5a4a */ ZFS_IOC_CHANGE_KEY, /* 0x5a4b */ ZFS_IOC_REMAP, /* 0x5a4c */ ZFS_IOC_POOL_CHECKPOINT, /* 0x5a4d */ ZFS_IOC_POOL_DISCARD_CHECKPOINT, /* 0x5a4e */ ZFS_IOC_POOL_INITIALIZE, /* 0x5a4f */ ZFS_IOC_POOL_TRIM, /* 0x5a50 */ ZFS_IOC_REDACT, /* 0x5a51 */ ZFS_IOC_GET_BOOKMARK_PROPS, /* 0x5a52 */ ZFS_IOC_WAIT, /* 0x5a53 */ ZFS_IOC_WAIT_FS, /* 0x5a54 */ ZFS_IOC_VDEV_GET_PROPS, /* 0x5a55 */ ZFS_IOC_VDEV_SET_PROPS, /* 0x5a56 */ ZFS_IOC_POOL_SCRUB, /* 0x5a57 */ /* * Per-platform (Optional) - 8/128 numbers reserved. */ ZFS_IOC_PLATFORM = ZFS_IOC_FIRST + 0x80, ZFS_IOC_EVENTS_NEXT, /* 0x81 (Linux) */ ZFS_IOC_EVENTS_CLEAR, /* 0x82 (Linux) */ ZFS_IOC_EVENTS_SEEK, /* 0x83 (Linux) */ ZFS_IOC_NEXTBOOT, /* 0x84 (FreeBSD) */ ZFS_IOC_JAIL, /* 0x85 (FreeBSD) */ ZFS_IOC_USERNS_ATTACH = ZFS_IOC_JAIL, /* 0x85 (Linux) */ ZFS_IOC_UNJAIL, /* 0x86 (FreeBSD) */ ZFS_IOC_USERNS_DETACH = ZFS_IOC_UNJAIL, /* 0x86 (Linux) */ ZFS_IOC_SET_BOOTENV, /* 0x87 */ ZFS_IOC_GET_BOOTENV, /* 0x88 */ ZFS_IOC_LAST } zfs_ioc_t; /* * zvol ioctl to get dataset name */ #define BLKZNAME _IOR(0x12, 125, char[ZFS_MAX_DATASET_NAME_LEN]) #ifdef __linux__ /* * IOCTLs to update and retrieve additional file level attributes on * Linux. */ #define ZFS_IOC_GETDOSFLAGS _IOR(0x83, 1, uint64_t) #define ZFS_IOC_SETDOSFLAGS _IOW(0x83, 2, uint64_t) /* * Additional file level attributes, that are stored * in the upper half of z_pflags */ #define ZFS_READONLY 0x0000000100000000ull #define ZFS_HIDDEN 0x0000000200000000ull #define ZFS_SYSTEM 0x0000000400000000ull #define ZFS_ARCHIVE 0x0000000800000000ull #define ZFS_IMMUTABLE 0x0000001000000000ull #define ZFS_NOUNLINK 0x0000002000000000ull #define ZFS_APPENDONLY 0x0000004000000000ull #define ZFS_NODUMP 0x0000008000000000ull #define ZFS_OPAQUE 0x0000010000000000ull #define ZFS_AV_QUARANTINED 0x0000020000000000ull #define ZFS_AV_MODIFIED 0x0000040000000000ull #define ZFS_REPARSE 0x0000080000000000ull #define ZFS_OFFLINE 0x0000100000000000ull #define ZFS_SPARSE 0x0000200000000000ull #define ZFS_DOS_FL_USER_VISIBLE (ZFS_IMMUTABLE | ZFS_APPENDONLY | \ ZFS_NOUNLINK | ZFS_ARCHIVE | ZFS_NODUMP | ZFS_SYSTEM | \ ZFS_HIDDEN | ZFS_READONLY | ZFS_REPARSE | ZFS_OFFLINE | \ ZFS_SPARSE) #endif /* * ZFS-specific error codes used for returning descriptive errors * to the userland through zfs ioctls. * * The enum implicitly includes all the error codes from errno.h. * New code should use and extend this enum for errors that are * not described precisely by generic errno codes. * * These numbers should not change over time. New entries should be appended. * * (Keep in sync with contrib/pyzfs/libzfs_core/_constants.py) */ typedef enum { ZFS_ERR_CHECKPOINT_EXISTS = 1024, ZFS_ERR_DISCARDING_CHECKPOINT, ZFS_ERR_NO_CHECKPOINT, ZFS_ERR_DEVRM_IN_PROGRESS, ZFS_ERR_VDEV_TOO_BIG, ZFS_ERR_IOC_CMD_UNAVAIL, ZFS_ERR_IOC_ARG_UNAVAIL, ZFS_ERR_IOC_ARG_REQUIRED, ZFS_ERR_IOC_ARG_BADTYPE, ZFS_ERR_WRONG_PARENT, ZFS_ERR_FROM_IVSET_GUID_MISSING, ZFS_ERR_FROM_IVSET_GUID_MISMATCH, ZFS_ERR_SPILL_BLOCK_FLAG_MISSING, ZFS_ERR_UNKNOWN_SEND_STREAM_FEATURE, ZFS_ERR_EXPORT_IN_PROGRESS, ZFS_ERR_BOOKMARK_SOURCE_NOT_ANCESTOR, ZFS_ERR_STREAM_TRUNCATED, ZFS_ERR_STREAM_LARGE_BLOCK_MISMATCH, ZFS_ERR_RESILVER_IN_PROGRESS, ZFS_ERR_REBUILD_IN_PROGRESS, ZFS_ERR_BADPROP, ZFS_ERR_VDEV_NOTSUP, ZFS_ERR_NOT_USER_NAMESPACE, ZFS_ERR_RESUME_EXISTS, ZFS_ERR_CRYPTO_NOTSUP, ZFS_ERR_RAIDZ_EXPAND_IN_PROGRESS, ZFS_ERR_ASHIFT_MISMATCH, } zfs_errno_t; /* * Internal SPA load state. Used by FMA diagnosis engine. */ typedef enum { SPA_LOAD_NONE, /* no load in progress */ SPA_LOAD_OPEN, /* normal open */ SPA_LOAD_IMPORT, /* import in progress */ SPA_LOAD_TRYIMPORT, /* tryimport in progress */ SPA_LOAD_RECOVER, /* recovery requested */ SPA_LOAD_ERROR, /* load failed */ SPA_LOAD_CREATE /* creation in progress */ } spa_load_state_t; typedef enum { ZPOOL_WAIT_CKPT_DISCARD, ZPOOL_WAIT_FREE, ZPOOL_WAIT_INITIALIZE, ZPOOL_WAIT_REPLACE, ZPOOL_WAIT_REMOVE, ZPOOL_WAIT_RESILVER, ZPOOL_WAIT_SCRUB, ZPOOL_WAIT_TRIM, ZPOOL_WAIT_NUM_ACTIVITIES } zpool_wait_activity_t; typedef enum { ZFS_WAIT_DELETEQ, ZFS_WAIT_NUM_ACTIVITIES } zfs_wait_activity_t; /* * Bookmark name values. */ #define ZPOOL_ERR_LIST "error list" #define ZPOOL_ERR_DATASET "dataset" #define ZPOOL_ERR_OBJECT "object" #define HIS_MAX_RECORD_LEN (MAXPATHLEN + MAXPATHLEN + 1) /* * The following are names used in the nvlist describing * the pool's history log. */ #define ZPOOL_HIST_RECORD "history record" #define ZPOOL_HIST_TIME "history time" #define ZPOOL_HIST_CMD "history command" #define ZPOOL_HIST_WHO "history who" #define ZPOOL_HIST_ZONE "history zone" #define ZPOOL_HIST_HOST "history hostname" #define ZPOOL_HIST_TXG "history txg" #define ZPOOL_HIST_INT_EVENT "history internal event" #define ZPOOL_HIST_INT_STR "history internal str" #define ZPOOL_HIST_INT_NAME "internal_name" #define ZPOOL_HIST_IOCTL "ioctl" #define ZPOOL_HIST_INPUT_NVL "in_nvl" #define ZPOOL_HIST_OUTPUT_NVL "out_nvl" #define ZPOOL_HIST_OUTPUT_SIZE "out_size" #define ZPOOL_HIST_DSNAME "dsname" #define ZPOOL_HIST_DSID "dsid" #define ZPOOL_HIST_ERRNO "errno" #define ZPOOL_HIST_ELAPSED_NS "elapsed_ns" /* * Special nvlist name that will not have its args recorded in the pool's * history log. */ #define ZPOOL_HIDDEN_ARGS "hidden_args" /* * The following are names used when invoking ZFS_IOC_POOL_INITIALIZE. */ #define ZPOOL_INITIALIZE_COMMAND "initialize_command" #define ZPOOL_INITIALIZE_VDEVS "initialize_vdevs" /* * The following are names used when invoking ZFS_IOC_POOL_TRIM. */ #define ZPOOL_TRIM_COMMAND "trim_command" #define ZPOOL_TRIM_VDEVS "trim_vdevs" #define ZPOOL_TRIM_RATE "trim_rate" #define ZPOOL_TRIM_SECURE "trim_secure" /* * The following are names used when invoking ZFS_IOC_POOL_WAIT. */ #define ZPOOL_WAIT_ACTIVITY "wait_activity" #define ZPOOL_WAIT_TAG "wait_tag" #define ZPOOL_WAIT_WAITED "wait_waited" /* * The following are names used when invoking ZFS_IOC_VDEV_GET_PROP. */ #define ZPOOL_VDEV_PROPS_GET_VDEV "vdevprops_get_vdev" #define ZPOOL_VDEV_PROPS_GET_PROPS "vdevprops_get_props" /* * The following are names used when invoking ZFS_IOC_VDEV_SET_PROP. */ #define ZPOOL_VDEV_PROPS_SET_VDEV "vdevprops_set_vdev" #define ZPOOL_VDEV_PROPS_SET_PROPS "vdevprops_set_props" /* * The following are names used when invoking ZFS_IOC_WAIT_FS. */ #define ZFS_WAIT_ACTIVITY "wait_activity" #define ZFS_WAIT_WAITED "wait_waited" /* * Flags for ZFS_IOC_VDEV_SET_STATE */ #define ZFS_ONLINE_CHECKREMOVE 0x1 #define ZFS_ONLINE_UNSPARE 0x2 #define ZFS_ONLINE_FORCEFAULT 0x4 #define ZFS_ONLINE_EXPAND 0x8 #define ZFS_ONLINE_SPARE 0x10 #define ZFS_OFFLINE_TEMPORARY 0x1 /* * Flags for ZFS_IOC_POOL_IMPORT */ #define ZFS_IMPORT_NORMAL 0x0 #define ZFS_IMPORT_VERBATIM 0x1 #define ZFS_IMPORT_ANY_HOST 0x2 #define ZFS_IMPORT_MISSING_LOG 0x4 #define ZFS_IMPORT_ONLY 0x8 #define ZFS_IMPORT_TEMP_NAME 0x10 #define ZFS_IMPORT_SKIP_MMP 0x20 #define ZFS_IMPORT_LOAD_KEYS 0x40 #define ZFS_IMPORT_CHECKPOINT 0x80 /* * Channel program argument/return nvlist keys and defaults. */ #define ZCP_ARG_PROGRAM "program" #define ZCP_ARG_ARGLIST "arg" #define ZCP_ARG_SYNC "sync" #define ZCP_ARG_INSTRLIMIT "instrlimit" #define ZCP_ARG_MEMLIMIT "memlimit" #define ZCP_ARG_CLIARGV "argv" #define ZCP_RET_ERROR "error" #define ZCP_RET_RETURN "return" #define ZCP_DEFAULT_INSTRLIMIT (10 * 1000 * 1000) #define ZCP_MAX_INSTRLIMIT (10 * ZCP_DEFAULT_INSTRLIMIT) #define ZCP_DEFAULT_MEMLIMIT (10 * 1024 * 1024) #define ZCP_MAX_MEMLIMIT (10 * ZCP_DEFAULT_MEMLIMIT) /* * Sysevent payload members. ZFS will generate the following sysevents with the * given payloads: * * ESC_ZFS_RESILVER_START * ESC_ZFS_RESILVER_FINISH * * ZFS_EV_POOL_NAME DATA_TYPE_STRING * ZFS_EV_POOL_GUID DATA_TYPE_UINT64 * ZFS_EV_RESILVER_TYPE DATA_TYPE_STRING * * ESC_ZFS_POOL_DESTROY * ESC_ZFS_POOL_REGUID * * ZFS_EV_POOL_NAME DATA_TYPE_STRING * ZFS_EV_POOL_GUID DATA_TYPE_UINT64 * * ESC_ZFS_VDEV_REMOVE * ESC_ZFS_VDEV_CLEAR * ESC_ZFS_VDEV_CHECK * * ZFS_EV_POOL_NAME DATA_TYPE_STRING * ZFS_EV_POOL_GUID DATA_TYPE_UINT64 * ZFS_EV_VDEV_PATH DATA_TYPE_STRING (optional) * ZFS_EV_VDEV_GUID DATA_TYPE_UINT64 * * ESC_ZFS_HISTORY_EVENT * * ZFS_EV_POOL_NAME DATA_TYPE_STRING * ZFS_EV_POOL_GUID DATA_TYPE_UINT64 * ZFS_EV_HIST_TIME DATA_TYPE_UINT64 (optional) * ZFS_EV_HIST_CMD DATA_TYPE_STRING (optional) * ZFS_EV_HIST_WHO DATA_TYPE_UINT64 (optional) * ZFS_EV_HIST_ZONE DATA_TYPE_STRING (optional) * ZFS_EV_HIST_HOST DATA_TYPE_STRING (optional) * ZFS_EV_HIST_TXG DATA_TYPE_UINT64 (optional) * ZFS_EV_HIST_INT_EVENT DATA_TYPE_UINT64 (optional) * ZFS_EV_HIST_INT_STR DATA_TYPE_STRING (optional) * ZFS_EV_HIST_INT_NAME DATA_TYPE_STRING (optional) * ZFS_EV_HIST_IOCTL DATA_TYPE_STRING (optional) * ZFS_EV_HIST_DSNAME DATA_TYPE_STRING (optional) * ZFS_EV_HIST_DSID DATA_TYPE_UINT64 (optional) * * The ZFS_EV_HIST_* members will correspond to the ZPOOL_HIST_* members in the * history log nvlist. The keynames will be free of any spaces or other * characters that could be potentially unexpected to consumers of the * sysevents. */ #define ZFS_EV_POOL_NAME "pool_name" #define ZFS_EV_POOL_GUID "pool_guid" #define ZFS_EV_VDEV_PATH "vdev_path" #define ZFS_EV_VDEV_GUID "vdev_guid" #define ZFS_EV_HIST_TIME "history_time" #define ZFS_EV_HIST_CMD "history_command" #define ZFS_EV_HIST_WHO "history_who" #define ZFS_EV_HIST_ZONE "history_zone" #define ZFS_EV_HIST_HOST "history_hostname" #define ZFS_EV_HIST_TXG "history_txg" #define ZFS_EV_HIST_INT_EVENT "history_internal_event" #define ZFS_EV_HIST_INT_STR "history_internal_str" #define ZFS_EV_HIST_INT_NAME "history_internal_name" #define ZFS_EV_HIST_IOCTL "history_ioctl" #define ZFS_EV_HIST_DSNAME "history_dsname" #define ZFS_EV_HIST_DSID "history_dsid" #define ZFS_EV_RESILVER_TYPE "resilver_type" /* * We currently support block sizes from 512 bytes to 16MB. * The benefits of larger blocks, and thus larger IO, need to be weighed * against the cost of COWing a giant block to modify one byte, and the * large latency of reading or writing a large block. * * The recordsize property can not be set larger than zfs_max_recordsize * (default 16MB on 64-bit and 1MB on 32-bit). See the comment near * zfs_max_recordsize in dsl_dataset.c for details. * * Note that although the LSIZE field of the blkptr_t can store sizes up * to 32MB, the dnode's dn_datablkszsec can only store sizes up to * 32MB - 512 bytes. Therefore, we limit SPA_MAXBLOCKSIZE to 16MB. */ #define SPA_MINBLOCKSHIFT 9 #define SPA_OLD_MAXBLOCKSHIFT 17 #define SPA_MAXBLOCKSHIFT 24 #define SPA_MINBLOCKSIZE (1ULL << SPA_MINBLOCKSHIFT) #define SPA_OLD_MAXBLOCKSIZE (1ULL << SPA_OLD_MAXBLOCKSHIFT) #define SPA_MAXBLOCKSIZE (1ULL << SPA_MAXBLOCKSHIFT) /* supported encryption algorithms */ enum zio_encrypt { ZIO_CRYPT_INHERIT = 0, ZIO_CRYPT_ON, ZIO_CRYPT_OFF, ZIO_CRYPT_AES_128_CCM, ZIO_CRYPT_AES_192_CCM, ZIO_CRYPT_AES_256_CCM, ZIO_CRYPT_AES_128_GCM, ZIO_CRYPT_AES_192_GCM, ZIO_CRYPT_AES_256_GCM, ZIO_CRYPT_FUNCTIONS }; #define ZIO_CRYPT_ON_VALUE ZIO_CRYPT_AES_256_GCM #define ZIO_CRYPT_DEFAULT ZIO_CRYPT_OFF /* * xattr namespace prefixes. These are forbidden in xattr names. * * For cross-platform compatibility, xattrs in the user namespace should not be * prefixed with the namespace name, but for backwards compatibility with older * ZFS on Linux versions we do prefix the namespace. */ #define ZFS_XA_NS_FREEBSD_PREFIX "freebsd:" #define ZFS_XA_NS_FREEBSD_PREFIX_LEN strlen("freebsd:") #define ZFS_XA_NS_LINUX_SECURITY_PREFIX "security." #define ZFS_XA_NS_LINUX_SECURITY_PREFIX_LEN strlen("security.") #define ZFS_XA_NS_LINUX_SYSTEM_PREFIX "system." #define ZFS_XA_NS_LINUX_SYSTEM_PREFIX_LEN strlen("system.") #define ZFS_XA_NS_LINUX_TRUSTED_PREFIX "trusted." #define ZFS_XA_NS_LINUX_TRUSTED_PREFIX_LEN strlen("trusted.") #define ZFS_XA_NS_LINUX_USER_PREFIX "user." #define ZFS_XA_NS_LINUX_USER_PREFIX_LEN strlen("user.") #define ZFS_XA_NS_PREFIX_MATCH(ns, name) \ (strncmp(name, ZFS_XA_NS_##ns##_PREFIX, \ ZFS_XA_NS_##ns##_PREFIX_LEN) == 0) #define ZFS_XA_NS_PREFIX_FORBIDDEN(name) \ (ZFS_XA_NS_PREFIX_MATCH(FREEBSD, name) || \ ZFS_XA_NS_PREFIX_MATCH(LINUX_SECURITY, name) || \ ZFS_XA_NS_PREFIX_MATCH(LINUX_SYSTEM, name) || \ ZFS_XA_NS_PREFIX_MATCH(LINUX_TRUSTED, name) || \ ZFS_XA_NS_PREFIX_MATCH(LINUX_USER, name)) #ifdef __cplusplus } #endif #endif /* _SYS_FS_ZFS_H */ diff --git a/lib/libzfs/libzfs.abi b/lib/libzfs/libzfs.abi index 771c7407e9da..8bedfe72294c 100644 --- a/lib/libzfs/libzfs.abi +++ b/lib/libzfs/libzfs.abi @@ -1,9548 +1,9549 @@ - + + diff --git a/man/man7/zfsprops.7 b/man/man7/zfsprops.7 index e3674b1f8a8d..59f6404379af 100644 --- a/man/man7/zfsprops.7 +++ b/man/man7/zfsprops.7 @@ -1,2136 +1,2153 @@ .\" .\" 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) 2009 Sun Microsystems, Inc. All Rights Reserved. .\" Copyright 2011 Joshua M. Clulow .\" Copyright (c) 2011, 2019 by Delphix. All rights reserved. .\" Copyright (c) 2011, Pawel Jakub Dawidek .\" Copyright (c) 2012, Glen Barber .\" Copyright (c) 2012, Bryan Drewery .\" Copyright (c) 2013, Steven Hartland .\" Copyright (c) 2013 by Saso Kiselkov. All rights reserved. .\" Copyright (c) 2014, Joyent, Inc. All rights reserved. .\" Copyright (c) 2014 by Adam Stevko. All rights reserved. .\" Copyright (c) 2014 Integros [integros.com] .\" Copyright (c) 2016 Nexenta Systems, Inc. All Rights Reserved. .\" Copyright (c) 2014, Xin LI .\" Copyright (c) 2014-2015, The FreeBSD Foundation, All Rights Reserved. .\" Copyright 2019 Richard Laager. All rights reserved. .\" Copyright 2018 Nexenta Systems, Inc. .\" Copyright 2019 Joyent, Inc. .\" Copyright (c) 2019, Kjeld Schouten-Lebbing .\" Copyright (c) 2022 Hewlett Packard Enterprise Development LP. .\" .Dd August 8, 2023 .Dt ZFSPROPS 7 .Os . .Sh NAME .Nm zfsprops .Nd native and user-defined properties of ZFS datasets . .Sh DESCRIPTION Properties are divided into two types, native properties and user-defined .Po or .Qq user .Pc properties. Native properties either export internal statistics or control ZFS behavior. In addition, native properties are either editable or read-only. User properties have no effect on ZFS behavior, but you can use them to annotate datasets in a way that is meaningful in your environment. For more information about user properties, see the .Sx User Properties section, below. . .Ss Native Properties Every dataset has a set of properties that export statistics about the dataset as well as control various behaviors. Properties are inherited from the parent unless overridden by the child. Some properties apply only to certain types of datasets .Pq file systems, volumes, or snapshots . .Pp The values of numeric properties can be specified using human-readable suffixes .Po for example, .Sy k , .Sy KB , .Sy M , .Sy Gb , and so forth, up to .Sy Z for zettabyte .Pc . The following are all valid .Pq and equal specifications: .Li 1536M , .Li 1.5g , .Li 1.50GB . .Pp The values of non-numeric properties are case sensitive and must be lowercase, except for .Sy mountpoint , .Sy sharenfs , and .Sy sharesmb . .Pp The following native properties consist of read-only statistics about the dataset. These properties can be neither set, nor inherited. Native properties apply to all dataset types unless otherwise noted. .Bl -tag -width "usedbyrefreservation" .It Sy available The amount of space available to the dataset and all its children, assuming that there is no other activity in the pool. Because space is shared within a pool, availability can be limited by any number of factors, including physical pool size, quotas, reservations, or other datasets within the pool. .Pp This property can also be referred to by its shortened column name, .Sy avail . .It Sy compressratio For non-snapshots, the compression ratio achieved for the .Sy used space of this dataset, expressed as a multiplier. The .Sy used property includes descendant datasets, and, for clones, does not include the space shared with the origin snapshot. For snapshots, the .Sy compressratio is the same as the .Sy refcompressratio property. Compression can be turned on by running: .Nm zfs Cm set Sy compression Ns = Ns Sy on Ar dataset . The default value is .Sy off . .It Sy createtxg The transaction group (txg) in which the dataset was created. Bookmarks have the same .Sy createtxg as the snapshot they are initially tied to. This property is suitable for ordering a list of snapshots, e.g. for incremental send and receive. .It Sy creation The time this dataset was created. .It Sy clones For snapshots, this property is a comma-separated list of filesystems or volumes which are clones of this snapshot. The clones' .Sy origin property is this snapshot. If the .Sy clones property is not empty, then this snapshot can not be destroyed .Po even with the .Fl r or .Fl f options .Pc . The roles of origin and clone can be swapped by promoting the clone with the .Nm zfs Cm promote command. .It Sy defer_destroy This property is .Sy on if the snapshot has been marked for deferred destroy by using the .Nm zfs Cm destroy Fl d command. Otherwise, the property is .Sy off . .It Sy encryptionroot For encrypted datasets, indicates where the dataset is currently inheriting its encryption key from. Loading or unloading a key for the .Sy encryptionroot will implicitly load / unload the key for any inheriting datasets (see .Nm zfs Cm load-key and .Nm zfs Cm unload-key for details). Clones will always share an encryption key with their origin. See the .Sx Encryption section of .Xr zfs-load-key 8 for details. .It Sy filesystem_count The total number of filesystems and volumes that exist under this location in the dataset tree. This value is only available when a .Sy filesystem_limit has been set somewhere in the tree under which the dataset resides. .It Sy keystatus Indicates if an encryption key is currently loaded into ZFS. The possible values are .Sy none , .Sy available , and .Sy unavailable . See .Nm zfs Cm load-key and .Nm zfs Cm unload-key . .It Sy guid The 64 bit GUID of this dataset or bookmark which does not change over its entire lifetime. When a snapshot is sent to another pool, the received snapshot has the same GUID. Thus, the .Sy guid is suitable to identify a snapshot across pools. .It Sy logicalreferenced The amount of space that is .Qq logically accessible by this dataset. See the .Sy referenced property. The logical space ignores the effect of the .Sy compression and .Sy copies properties, giving a quantity closer to the amount of data that applications see. However, it does include space consumed by metadata. .Pp This property can also be referred to by its shortened column name, .Sy lrefer . .It Sy logicalused The amount of space that is .Qq logically consumed by this dataset and all its descendents. See the .Sy used property. The logical space ignores the effect of the .Sy compression and .Sy copies properties, giving a quantity closer to the amount of data that applications see. However, it does include space consumed by metadata. .Pp This property can also be referred to by its shortened column name, .Sy lused . .It Sy mounted For file systems, indicates whether the file system is currently mounted. This property can be either .Sy yes or .Sy no . .It Sy objsetid A unique identifier for this dataset within the pool. Unlike the dataset's .Sy guid , No the Sy objsetid of a dataset is not transferred to other pools when the snapshot is copied with a send/receive operation. The .Sy objsetid can be reused (for a new dataset) after the dataset is deleted. .It Sy origin For cloned file systems or volumes, the snapshot from which the clone was created. See also the .Sy clones property. .It Sy receive_resume_token For filesystems or volumes which have saved partially-completed state from .Nm zfs Cm receive Fl s , this opaque token can be provided to .Nm zfs Cm send Fl t to resume and complete the .Nm zfs Cm receive . .It Sy redact_snaps For bookmarks, this is the list of snapshot guids the bookmark contains a redaction list for. For snapshots, this is the list of snapshot guids the snapshot is redacted with respect to. .It Sy referenced The amount of data that is accessible by this dataset, which may or may not be shared with other datasets in the pool. When a snapshot or clone is created, it initially references the same amount of space as the file system or snapshot it was created from, since its contents are identical. .Pp This property can also be referred to by its shortened column name, .Sy refer . .It Sy refcompressratio The compression ratio achieved for the .Sy referenced space of this dataset, expressed as a multiplier. See also the .Sy compressratio property. .It Sy snapshot_count The total number of snapshots that exist under this location in the dataset tree. This value is only available when a .Sy snapshot_limit has been set somewhere in the tree under which the dataset resides. .It Sy type The type of dataset: .Sy filesystem , .Sy volume , .Sy snapshot , or .Sy bookmark . .It Sy used The amount of space consumed by this dataset and all its descendents. This is the value that is checked against this dataset's quota and reservation. The space used does not include this dataset's reservation, but does take into account the reservations of any descendent datasets. The amount of space that a dataset consumes from its parent, as well as the amount of space that is freed if this dataset is recursively destroyed, is the greater of its space used and its reservation. .Pp The used space of a snapshot .Po see the .Sx Snapshots section of .Xr zfsconcepts 7 .Pc is space that is referenced exclusively by this snapshot. If this snapshot is destroyed, the amount of .Sy used space will be freed. Space that is shared by multiple snapshots isn't accounted for in this metric. When a snapshot is destroyed, space that was previously shared with this snapshot can become unique to snapshots adjacent to it, thus changing the used space of those snapshots. The used space of the latest snapshot can also be affected by changes in the file system. Note that the .Sy used space of a snapshot is a subset of the .Sy written space of the snapshot. .Pp The amount of space used, available, or referenced does not take into account pending changes. Pending changes are generally accounted for within a few seconds. Committing a change to a disk using .Xr fsync 2 or .Sy O_SYNC does not necessarily guarantee that the space usage information is updated immediately. .It Sy usedby* The .Sy usedby* properties decompose the .Sy used properties into the various reasons that space is used. Specifically, .Sy used No = .Sy usedbychildren No + .Sy usedbydataset No + .Sy usedbyrefreservation No + .Sy usedbysnapshots . These properties are only available for datasets created on .Nm zpool .Qo version 13 Qc pools. .It Sy usedbychildren The amount of space used by children of this dataset, which would be freed if all the dataset's children were destroyed. .It Sy usedbydataset The amount of space used by this dataset itself, which would be freed if the dataset were destroyed .Po after first removing any .Sy refreservation and destroying any necessary snapshots or descendents .Pc . .It Sy usedbyrefreservation The amount of space used by a .Sy refreservation set on this dataset, which would be freed if the .Sy refreservation was removed. .It Sy usedbysnapshots The amount of space consumed by snapshots of this dataset. In particular, it is the amount of space that would be freed if all of this dataset's snapshots were destroyed. Note that this is not simply the sum of the snapshots' .Sy used properties because space can be shared by multiple snapshots. .It Sy userused Ns @ Ns Ar user The amount of space consumed by the specified user in this dataset. Space is charged to the owner of each file, as displayed by .Nm ls Fl l . The amount of space charged is displayed by .Nm du No and Nm ls Fl s . See the .Nm zfs Cm userspace command for more information. .Pp Unprivileged users can access only their own space usage. The root user, or a user who has been granted the .Sy userused privilege with .Nm zfs Cm allow , can access everyone's usage. .Pp The .Sy userused Ns @ Ns Ar … properties are not displayed by .Nm zfs Cm get Sy all . The user's name must be appended after the .Sy @ symbol, using one of the following forms: .Bl -bullet -compact -offset 4n .It POSIX name .Pq Qq joe .It POSIX numeric ID .Pq Qq 789 .It SID name .Pq Qq joe.smith@mydomain .It SID numeric ID .Pq Qq S-1-123-456-789 .El .Pp Files created on Linux always have POSIX owners. .It Sy userobjused Ns @ Ns Ar user The .Sy userobjused property is similar to .Sy userused but instead it counts the number of objects consumed by a user. This property counts all objects allocated on behalf of the user, it may differ from the results of system tools such as .Nm df Fl i . .Pp When the property .Sy xattr Ns = Ns Sy on is set on a file system additional objects will be created per-file to store extended attributes. These additional objects are reflected in the .Sy userobjused value and are counted against the user's .Sy userobjquota . When a file system is configured to use .Sy xattr Ns = Ns Sy sa no additional internal objects are normally required. .It Sy userrefs This property is set to the number of user holds on this snapshot. User holds are set by using the .Nm zfs Cm hold command. .It Sy groupused Ns @ Ns Ar group The amount of space consumed by the specified group in this dataset. Space is charged to the group of each file, as displayed by .Nm ls Fl l . See the .Sy userused Ns @ Ns Ar user property for more information. .Pp Unprivileged users can only access their own groups' space usage. The root user, or a user who has been granted the .Sy groupused privilege with .Nm zfs Cm allow , can access all groups' usage. .It Sy groupobjused Ns @ Ns Ar group The number of objects consumed by the specified group in this dataset. Multiple objects may be charged to the group for each file when extended attributes are in use. See the .Sy userobjused Ns @ Ns Ar user property for more information. .Pp Unprivileged users can only access their own groups' space usage. The root user, or a user who has been granted the .Sy groupobjused privilege with .Nm zfs Cm allow , can access all groups' usage. .It Sy projectused Ns @ Ns Ar project The amount of space consumed by the specified project in this dataset. Project is identified via the project identifier (ID) that is object-based numeral attribute. An object can inherit the project ID from its parent object (if the parent has the flag of inherit project ID that can be set and changed via .Nm chattr Fl /+P or .Nm zfs project Fl s ) when being created. The privileged user can set and change object's project ID via .Nm chattr Fl p or .Nm zfs project Fl s anytime. Space is charged to the project of each file, as displayed by .Nm lsattr Fl p or .Nm zfs project . See the .Sy userused Ns @ Ns Ar user property for more information. .Pp The root user, or a user who has been granted the .Sy projectused privilege with .Nm zfs allow , can access all projects' usage. .It Sy projectobjused Ns @ Ns Ar project The .Sy projectobjused is similar to .Sy projectused but instead it counts the number of objects consumed by project. When the property .Sy xattr Ns = Ns Sy on is set on a fileset, ZFS will create additional objects per-file to store extended attributes. These additional objects are reflected in the .Sy projectobjused value and are counted against the project's .Sy projectobjquota . When a filesystem is configured to use .Sy xattr Ns = Ns Sy sa no additional internal objects are required. See the .Sy userobjused Ns @ Ns Ar user property for more information. .Pp The root user, or a user who has been granted the .Sy projectobjused privilege with .Nm zfs allow , can access all projects' objects usage. .It Sy snapshots_changed Provides a mechanism to quickly determine whether snapshot list has changed without having to mount a dataset or iterate the snapshot list. Specifies the time at which a snapshot for a dataset was last created or deleted. .Pp This allows us to be more efficient how often we query snapshots. The property is persistent across mount and unmount operations only if the .Sy extensible_dataset feature is enabled. .It Sy volblocksize For volumes, specifies the block size of the volume. The .Sy blocksize cannot be changed once the volume has been written, so it should be set at volume creation time. The default .Sy blocksize for volumes is 16 Kbytes. Any power of 2 from 512 bytes to 128 Kbytes is valid. .Pp This property can also be referred to by its shortened column name, .Sy volblock . .It Sy written The amount of space .Sy referenced by this dataset, that was written since the previous snapshot .Pq i.e. that is not referenced by the previous snapshot . .It Sy written Ns @ Ns Ar snapshot The amount of .Sy referenced space written to this dataset since the specified snapshot. This is the space that is referenced by this dataset but was not referenced by the specified snapshot. .Pp The .Ar snapshot may be specified as a short snapshot name .Pq just the part after the Sy @ , in which case it will be interpreted as a snapshot in the same filesystem as this dataset. The .Ar snapshot may be a full snapshot name .Pq Ar filesystem Ns @ Ns Ar snapshot , which for clones may be a snapshot in the origin's filesystem .Pq or the origin of the origin's filesystem, etc. .El .Pp The following native properties can be used to change the behavior of a ZFS dataset. .Bl -tag -width "" .It Xo .Sy aclinherit Ns = Ns Sy discard Ns | Ns Sy noallow Ns | Ns .Sy restricted Ns | Ns Sy passthrough Ns | Ns Sy passthrough-x .Xc Controls how ACEs are inherited when files and directories are created. .Bl -tag -compact -offset 4n -width "passthrough-x" .It Sy discard does not inherit any ACEs. .It Sy noallow only inherits inheritable ACEs that specify .Qq deny permissions. .It Sy restricted default, removes the .Sy write_acl and .Sy write_owner permissions when the ACE is inherited. .It Sy passthrough inherits all inheritable ACEs without any modifications. .It Sy passthrough-x same meaning as .Sy passthrough , except that the .Sy owner@ , group@ , No and Sy everyone@ ACEs inherit the execute permission only if the file creation mode also requests the execute bit. .El .Pp When the property value is set to .Sy passthrough , files are created with a mode determined by the inheritable ACEs. If no inheritable ACEs exist that affect the mode, then the mode is set in accordance to the requested mode from the application. .Pp The .Sy aclinherit property does not apply to POSIX ACLs. .It Xo .Sy aclmode Ns = Ns Sy discard Ns | Ns Sy groupmask Ns | Ns .Sy passthrough Ns | Ns Sy restricted Ns .Xc Controls how an ACL is modified during chmod(2) and how inherited ACEs are modified by the file creation mode: .Bl -tag -compact -offset 4n -width "passthrough" .It Sy discard default, deletes all .Sy ACEs except for those representing the mode of the file or directory requested by .Xr chmod 2 . .It Sy groupmask reduces permissions granted in all .Sy ALLOW entries found in the .Sy ACL such that they are no greater than the group permissions specified by .Xr chmod 2 . .It Sy passthrough indicates that no changes are made to the ACL other than creating or updating the necessary ACL entries to represent the new mode of the file or directory. .It Sy restricted will cause the .Xr chmod 2 operation to return an error when used on any file or directory which has a non-trivial ACL whose entries can not be represented by a mode. .Xr chmod 2 is required to change the set user ID, set group ID, or sticky bits on a file or directory, as they do not have equivalent ACL entries. In order to use .Xr chmod 2 on a file or directory with a non-trivial ACL when .Sy aclmode is set to .Sy restricted , you must first remove all ACL entries which do not represent the current mode. .El .It Sy acltype Ns = Ns Sy off Ns | Ns Sy nfsv4 Ns | Ns Sy posix Controls whether ACLs are enabled and if so what type of ACL to use. When this property is set to a type of ACL not supported by the current platform, the behavior is the same as if it were set to .Sy off . .Bl -tag -compact -offset 4n -width "posixacl" .It Sy off default on Linux, when a file system has the .Sy acltype property set to off then ACLs are disabled. .It Sy noacl an alias for .Sy off .It Sy nfsv4 default on .Fx , indicates that NFSv4-style ZFS ACLs should be used. These ACLs can be managed with the .Xr getfacl 1 and .Xr setfacl 1 . The .Sy nfsv4 ZFS ACL type is not yet supported on Linux. .It Sy posix indicates POSIX ACLs should be used. POSIX ACLs are specific to Linux and are not functional on other platforms. POSIX ACLs are stored as an extended attribute and therefore will not overwrite any existing NFSv4 ACLs which may be set. .It Sy posixacl an alias for .Sy posix .El .Pp To obtain the best performance when setting .Sy posix users are strongly encouraged to set the .Sy xattr Ns = Ns Sy sa property. This will result in the POSIX ACL being stored more efficiently on disk. But as a consequence, all new extended attributes will only be accessible from OpenZFS implementations which support the .Sy xattr Ns = Ns Sy sa property. See the .Sy xattr property for more details. .It Sy atime Ns = Ns Sy on Ns | Ns Sy off Controls whether the access time for files is updated when they are read. Turning this property off avoids producing write traffic when reading files and can result in significant performance gains, though it might confuse mailers and other similar utilities. The values .Sy on and .Sy off are equivalent to the .Sy atime and .Sy noatime mount options. The default value is .Sy on . See also .Sy relatime below. .It Sy canmount Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Sy noauto If this property is set to .Sy off , the file system cannot be mounted, and is ignored by .Nm zfs Cm mount Fl a . Setting this property to .Sy off is similar to setting the .Sy mountpoint property to .Sy none , except that the dataset still has a normal .Sy mountpoint property, which can be inherited. Setting this property to .Sy off allows datasets to be used solely as a mechanism to inherit properties. One example of setting .Sy canmount Ns = Ns Sy off is to have two datasets with the same .Sy mountpoint , so that the children of both datasets appear in the same directory, but might have different inherited characteristics. .Pp When set to .Sy noauto , a dataset can only be mounted and unmounted explicitly. The dataset is not mounted automatically when the dataset is created or imported, nor is it mounted by the .Nm zfs Cm mount Fl a command or unmounted by the .Nm zfs Cm unmount Fl a command. .Pp This property is not inherited. .It Xo .Sy checksum Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Sy fletcher2 Ns | Ns .Sy fletcher4 Ns | Ns Sy sha256 Ns | Ns Sy noparity Ns | Ns .Sy sha512 Ns | Ns Sy skein Ns | Ns Sy edonr Ns | Ns Sy blake3 .Xc Controls the checksum used to verify data integrity. The default value is .Sy on , which automatically selects an appropriate algorithm .Po currently, .Sy fletcher4 , but this may change in future releases .Pc . The value .Sy off disables integrity checking on user data. The value .Sy noparity not only disables integrity but also disables maintaining parity for user data. This setting is used internally by a dump device residing on a RAID-Z pool and should not be used by any other dataset. Disabling checksums is .Em NOT a recommended practice. .Pp The .Sy sha512 , .Sy skein , .Sy edonr , and .Sy blake3 checksum algorithms require enabling the appropriate features on the pool. .Pp Please see .Xr zpool-features 7 for more information on these algorithms. .Pp Changing this property affects only newly-written data. .It Xo .Sy compression Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Sy gzip Ns | Ns .Sy gzip- Ns Ar N Ns | Ns Sy lz4 Ns | Ns Sy lzjb Ns | Ns Sy zle Ns | Ns Sy zstd Ns | Ns .Sy zstd- Ns Ar N Ns | Ns Sy zstd-fast Ns | Ns Sy zstd-fast- Ns Ar N .Xc Controls the compression algorithm used for this dataset. .Pp When set to .Sy on (the default), indicates that the current default compression algorithm should be used. The default balances compression and decompression speed, with compression ratio and is expected to work well on a wide variety of workloads. Unlike all other settings for this property, .Sy on does not select a fixed compression type. As new compression algorithms are added to ZFS and enabled on a pool, the default compression algorithm may change. The current default compression algorithm is either .Sy lzjb or, if the .Sy lz4_compress feature is enabled, .Sy lz4 . .Pp The .Sy lz4 compression algorithm is a high-performance replacement for the .Sy lzjb algorithm. It features significantly faster compression and decompression, as well as a moderately higher compression ratio than .Sy lzjb , but can only be used on pools with the .Sy lz4_compress feature set to .Sy enabled . See .Xr zpool-features 7 for details on ZFS feature flags and the .Sy lz4_compress feature. .Pp The .Sy lzjb compression algorithm is optimized for performance while providing decent data compression. .Pp The .Sy gzip compression algorithm uses the same compression as the .Xr gzip 1 command. You can specify the .Sy gzip level by using the value .Sy gzip- Ns Ar N , where .Ar N is an integer from 1 .Pq fastest to 9 .Pq best compression ratio . Currently, .Sy gzip is equivalent to .Sy gzip-6 .Po which is also the default for .Xr gzip 1 .Pc . .Pp The .Sy zstd compression algorithm provides both high compression ratios and good performance. You can specify the .Sy zstd level by using the value .Sy zstd- Ns Ar N , where .Ar N is an integer from 1 .Pq fastest to 19 .Pq best compression ratio . .Sy zstd is equivalent to .Sy zstd-3 . .Pp Faster speeds at the cost of the compression ratio can be requested by setting a negative .Sy zstd level. This is done using .Sy zstd-fast- Ns Ar N , where .Ar N is an integer in .Bq Sy 1 Ns - Ns Sy 10 , 20 , 30 , No … , Sy 100 , 500 , 1000 which maps to a negative .Sy zstd level. The lower the level the faster the compression \(em .Sy 1000 provides the fastest compression and lowest compression ratio. .Sy zstd-fast is equivalent to .Sy zstd-fast- Ns Ar 1 . .Pp The .Sy zle compression algorithm compresses runs of zeros. .Pp This property can also be referred to by its shortened column name .Sy compress . Changing this property affects only newly-written data. .Pp When any setting except .Sy off is selected, compression will explicitly check for blocks consisting of only zeroes (the NUL byte). When a zero-filled block is detected, it is stored as a hole and not compressed using the indicated compression algorithm. .Pp Any block being compressed must be no larger than 7/8 of its original size after compression, otherwise the compression will not be considered worthwhile and the block saved uncompressed. Note that when the logical block is less than 8 times the disk sector size this effectively reduces the necessary compression ratio; for example, 8 KiB blocks on disks with 4 KiB disk sectors must compress to 1/2 or less of their original size. .It Xo .Sy context Ns = Ns Sy none Ns | Ns .Ar SELinux-User : Ns Ar SELinux-Role : Ns Ar SELinux-Type : Ns Ar Sensitivity-Level .Xc This flag sets the SELinux context for all files in the file system under a mount point for that file system. See .Xr selinux 8 for more information. .It Xo .Sy fscontext Ns = Ns Sy none Ns | Ns .Ar SELinux-User : Ns Ar SELinux-Role : Ns Ar SELinux-Type : Ns Ar Sensitivity-Level .Xc This flag sets the SELinux context for the file system file system being mounted. See .Xr selinux 8 for more information. .It Xo .Sy defcontext Ns = Ns Sy none Ns | Ns .Ar SELinux-User : Ns Ar SELinux-Role : Ns Ar SELinux-Type : Ns Ar Sensitivity-Level .Xc This flag sets the SELinux default context for unlabeled files. See .Xr selinux 8 for more information. .It Xo .Sy rootcontext Ns = Ns Sy none Ns | Ns .Ar SELinux-User : Ns Ar SELinux-Role : Ns Ar SELinux-Type : Ns Ar Sensitivity-Level .Xc This flag sets the SELinux context for the root inode of the file system. See .Xr selinux 8 for more information. .It Sy copies Ns = Ns Sy 1 Ns | Ns Sy 2 Ns | Ns Sy 3 Controls the number of copies of data stored for this dataset. These copies are in addition to any redundancy provided by the pool, for example, mirroring or RAID-Z. The copies are stored on different disks, if possible. The space used by multiple copies is charged to the associated file and dataset, changing the .Sy used property and counting against quotas and reservations. .Pp Changing this property only affects newly-written data. Therefore, set this property at file system creation time by using the .Fl o Sy copies Ns = Ns Ar N option. .Pp Remember that ZFS will not import a pool with a missing top-level vdev. Do .Em NOT create, for example a two-disk striped pool and set .Sy copies Ns = Ns Ar 2 on some datasets thinking you have setup redundancy for them. When a disk fails you will not be able to import the pool and will have lost all of your data. .Pp Encrypted datasets may not have .Sy copies Ns = Ns Ar 3 since the implementation stores some encryption metadata where the third copy would normally be. .It Sy devices Ns = Ns Sy on Ns | Ns Sy off Controls whether device nodes can be opened on this file system. The default value is .Sy on . The values .Sy on and .Sy off are equivalent to the .Sy dev and .Sy nodev mount options. .It Xo .Sy dedup Ns = Ns Sy off Ns | Ns Sy on Ns | Ns Sy verify Ns | Ns .Sy sha256 Ns Oo , Ns Sy verify Oc Ns | Ns Sy sha512 Ns Oo , Ns Sy verify Oc Ns | Ns Sy skein Ns Oo , Ns Sy verify Oc Ns | Ns .Sy edonr , Ns Sy verify Ns | Ns Sy blake3 Ns Oo , Ns Sy verify Oc Ns .Xc Configures deduplication for a dataset. The default value is .Sy off . The default deduplication checksum is .Sy sha256 (this may change in the future). When .Sy dedup is enabled, the checksum defined here overrides the .Sy checksum property. Setting the value to .Sy verify has the same effect as the setting .Sy sha256 , Ns Sy verify . .Pp If set to .Sy verify , ZFS will do a byte-to-byte comparison in case of two blocks having the same signature to make sure the block contents are identical. Specifying .Sy verify is mandatory for the .Sy edonr algorithm. .Pp Unless necessary, deduplication should .Em not be enabled on a system. See the .Sx Deduplication section of .Xr zfsconcepts 7 . .It Xo .Sy dnodesize Ns = Ns Sy legacy Ns | Ns Sy auto Ns | Ns Sy 1k Ns | Ns .Sy 2k Ns | Ns Sy 4k Ns | Ns Sy 8k Ns | Ns Sy 16k .Xc Specifies a compatibility mode or literal value for the size of dnodes in the file system. The default value is .Sy legacy . Setting this property to a value other than .Sy legacy No requires the Sy large_dnode No pool feature to be enabled . .Pp Consider setting .Sy dnodesize to .Sy auto if the dataset uses the .Sy xattr Ns = Ns Sy sa property setting and the workload makes heavy use of extended attributes. This may be applicable to SELinux-enabled systems, Lustre servers, and Samba servers, for example. Literal values are supported for cases where the optimal size is known in advance and for performance testing. .Pp Leave .Sy dnodesize set to .Sy legacy if you need to receive a send stream of this dataset on a pool that doesn't enable the .Sy large_dnode feature, or if you need to import this pool on a system that doesn't support the .Sy large_dnode No feature . .Pp This property can also be referred to by its shortened column name, .Sy dnsize . .It Xo .Sy encryption Ns = Ns Sy off Ns | Ns Sy on Ns | Ns Sy aes-128-ccm Ns | Ns .Sy aes-192-ccm Ns | Ns Sy aes-256-ccm Ns | Ns Sy aes-128-gcm Ns | Ns .Sy aes-192-gcm Ns | Ns Sy aes-256-gcm .Xc Controls the encryption cipher suite (block cipher, key length, and mode) used for this dataset. Requires the .Sy encryption feature to be enabled on the pool. Requires a .Sy keyformat to be set at dataset creation time. .Pp Selecting .Sy encryption Ns = Ns Sy on when creating a dataset indicates that the default encryption suite will be selected, which is currently .Sy aes-256-gcm . In order to provide consistent data protection, encryption must be specified at dataset creation time and it cannot be changed afterwards. .Pp For more details and caveats about encryption see the .Sx Encryption section of .Xr zfs-load-key 8 . .It Sy keyformat Ns = Ns Sy raw Ns | Ns Sy hex Ns | Ns Sy passphrase Controls what format the user's encryption key will be provided as. This property is only set when the dataset is encrypted. .Pp Raw keys and hex keys must be 32 bytes long (regardless of the chosen encryption suite) and must be randomly generated. A raw key can be generated with the following command: .Dl # Nm dd Sy if=/dev/urandom bs=32 count=1 Sy of= Ns Pa /path/to/output/key .Pp Passphrases must be between 8 and 512 bytes long and will be processed through PBKDF2 before being used (see the .Sy pbkdf2iters property). Even though the encryption suite cannot be changed after dataset creation, the keyformat can be with .Nm zfs Cm change-key . .It Xo .Sy keylocation Ns = Ns Sy prompt Ns | Ns Sy file:// Ns Ar /absolute/file/path Ns | Ns Sy https:// Ns Ar address Ns | Ns Sy http:// Ns Ar address .Xc Controls where the user's encryption key will be loaded from by default for commands such as .Nm zfs Cm load-key and .Nm zfs Cm mount Fl l . This property is only set for encrypted datasets which are encryption roots. If unspecified, the default is .Sy prompt . .Pp Even though the encryption suite cannot be changed after dataset creation, the keylocation can be with either .Nm zfs Cm set or .Nm zfs Cm change-key . If .Sy prompt is selected ZFS will ask for the key at the command prompt when it is required to access the encrypted data (see .Nm zfs Cm load-key for details). This setting will also allow the key to be passed in via the standard input stream, but users should be careful not to place keys which should be kept secret on the command line. If a file URI is selected, the key will be loaded from the specified absolute file path. If an HTTPS or HTTP URL is selected, it will be GETted using .Xr fetch 3 , libcurl, or nothing, depending on compile-time configuration and run-time availability. The .Sy SSL_CA_CERT_FILE environment variable can be set to set the location of the concatenated certificate store. The .Sy SSL_CA_CERT_PATH environment variable can be set to override the location of the directory containing the certificate authority bundle. The .Sy SSL_CLIENT_CERT_FILE and .Sy SSL_CLIENT_KEY_FILE environment variables can be set to configure the path to the client certificate and its key. .It Sy pbkdf2iters Ns = Ns Ar iterations Controls the number of PBKDF2 iterations that a .Sy passphrase encryption key should be run through when processing it into an encryption key. This property is only defined when encryption is enabled and a keyformat of .Sy passphrase is selected. The goal of PBKDF2 is to significantly increase the computational difficulty needed to brute force a user's passphrase. This is accomplished by forcing the attacker to run each passphrase through a computationally expensive hashing function many times before they arrive at the resulting key. A user who actually knows the passphrase will only have to pay this cost once. As CPUs become better at processing, this number should be raised to ensure that a brute force attack is still not possible. The current default is .Sy 350000 and the minimum is .Sy 100000 . This property may be changed with .Nm zfs Cm change-key . .It Sy exec Ns = Ns Sy on Ns | Ns Sy off Controls whether processes can be executed from within this file system. The default value is .Sy on . The values .Sy on and .Sy off are equivalent to the .Sy exec and .Sy noexec mount options. .It Sy filesystem_limit Ns = Ns Ar count Ns | Ns Sy none Limits the number of filesystems and volumes that can exist under this point in the dataset tree. The limit is not enforced if the user is allowed to change the limit. Setting a .Sy filesystem_limit to .Sy on a descendent of a filesystem that already has a .Sy filesystem_limit does not override the ancestor's .Sy filesystem_limit , but rather imposes an additional limit. This feature must be enabled to be used .Po see .Xr zpool-features 7 .Pc . .It Sy special_small_blocks Ns = Ns Ar size This value represents the threshold block size for including small file blocks into the special allocation class. Blocks smaller than or equal to this value will be assigned to the special allocation class while greater blocks will be assigned to the regular class. Valid values are zero or a power of two from 512 up to 1048576 (1 MiB). The default size is 0 which means no small file blocks will be allocated in the special class. .Pp Before setting this property, a special class vdev must be added to the pool. See .Xr zpoolconcepts 7 for more details on the special allocation class. .It Sy mountpoint Ns = Ns Pa path Ns | Ns Sy none Ns | Ns Sy legacy Controls the mount point used for this file system. See the .Sx Mount Points section of .Xr zfsconcepts 7 for more information on how this property is used. .Pp When the .Sy mountpoint property is changed for a file system, the file system and any children that inherit the mount point are unmounted. If the new value is .Sy legacy , then they remain unmounted. Otherwise, they are automatically remounted in the new location if the property was previously .Sy legacy or .Sy none . In addition, any shared file systems are unshared and shared in the new location. .Pp When the .Sy mountpoint property is set with .Nm zfs Cm set Fl u , the .Sy mountpoint property is updated but dataset is not mounted or unmounted and remains as it was before. .It Sy nbmand Ns = Ns Sy on Ns | Ns Sy off Controls whether the file system should be mounted with .Sy nbmand .Pq Non-blocking mandatory locks . Changes to this property only take effect when the file system is umounted and remounted. This was only supported by Linux prior to 5.15, and was buggy there, and is not supported by .Fx . On Solaris it's used for SMB clients. .It Sy overlay Ns = Ns Sy on Ns | Ns Sy off Allow mounting on a busy directory or a directory which already contains files or directories. This is the default mount behavior for Linux and .Fx file systems. On these platforms the property is .Sy on by default. Set to .Sy off to disable overlay mounts for consistency with OpenZFS on other platforms. .It Sy primarycache Ns = Ns Sy all Ns | Ns Sy none Ns | Ns Sy metadata Controls what is cached in the primary cache .Pq ARC . If this property is set to .Sy all , then both user data and metadata is cached. If this property is set to .Sy none , then neither user data nor metadata is cached. If this property is set to .Sy metadata , then only metadata is cached. The default value is .Sy all . .It Sy quota Ns = Ns Ar size Ns | Ns Sy none Limits the amount of space a dataset and its descendents can consume. This property enforces a hard limit on the amount of space used. This includes all space consumed by descendents, including file systems and snapshots. Setting a quota on a descendent of a dataset that already has a quota does not override the ancestor's quota, but rather imposes an additional limit. .Pp Quotas cannot be set on volumes, as the .Sy volsize property acts as an implicit quota. .It Sy snapshot_limit Ns = Ns Ar count Ns | Ns Sy none Limits the number of snapshots that can be created on a dataset and its descendents. Setting a .Sy snapshot_limit on a descendent of a dataset that already has a .Sy snapshot_limit does not override the ancestor's .Sy snapshot_limit , but rather imposes an additional limit. The limit is not enforced if the user is allowed to change the limit. For example, this means that recursive snapshots taken from the global zone are counted against each delegated dataset within a zone. This feature must be enabled to be used .Po see .Xr zpool-features 7 .Pc . .It Sy userquota@ Ns Ar user Ns = Ns Ar size Ns | Ns Sy none Limits the amount of space consumed by the specified user. User space consumption is identified by the .Sy userspace@ Ns Ar user property. .Pp Enforcement of user quotas may be delayed by several seconds. This delay means that a user might exceed their quota before the system notices that they are over quota and begins to refuse additional writes with the .Er EDQUOT error message. See the .Nm zfs Cm userspace command for more information. .Pp Unprivileged users can only access their own groups' space usage. The root user, or a user who has been granted the .Sy userquota privilege with .Nm zfs Cm allow , can get and set everyone's quota. .Pp This property is not available on volumes, on file systems before version 4, or on pools before version 15. The .Sy userquota@ Ns Ar … properties are not displayed by .Nm zfs Cm get Sy all . The user's name must be appended after the .Sy @ symbol, using one of the following forms: .Bl -bullet -compact -offset 4n .It POSIX name .Pq Qq joe .It POSIX numeric ID .Pq Qq 789 .It SID name .Pq Qq joe.smith@mydomain .It SID numeric ID .Pq Qq S-1-123-456-789 .El .Pp Files created on Linux always have POSIX owners. .It Sy userobjquota@ Ns Ar user Ns = Ns Ar size Ns | Ns Sy none The .Sy userobjquota is similar to .Sy userquota but it limits the number of objects a user can create. Please refer to .Sy userobjused for more information about how objects are counted. .It Sy groupquota@ Ns Ar group Ns = Ns Ar size Ns | Ns Sy none Limits the amount of space consumed by the specified group. Group space consumption is identified by the .Sy groupused@ Ns Ar group property. .Pp Unprivileged users can access only their own groups' space usage. The root user, or a user who has been granted the .Sy groupquota privilege with .Nm zfs Cm allow , can get and set all groups' quotas. .It Sy groupobjquota@ Ns Ar group Ns = Ns Ar size Ns | Ns Sy none The .Sy groupobjquota is similar to .Sy groupquota but it limits number of objects a group can consume. Please refer to .Sy userobjused for more information about how objects are counted. .It Sy projectquota@ Ns Ar project Ns = Ns Ar size Ns | Ns Sy none Limits the amount of space consumed by the specified project. Project space consumption is identified by the .Sy projectused@ Ns Ar project property. Please refer to .Sy projectused for more information about how project is identified and set/changed. .Pp The root user, or a user who has been granted the .Sy projectquota privilege with .Nm zfs allow , can access all projects' quota. .It Sy projectobjquota@ Ns Ar project Ns = Ns Ar size Ns | Ns Sy none The .Sy projectobjquota is similar to .Sy projectquota but it limits number of objects a project can consume. Please refer to .Sy userobjused for more information about how objects are counted. .It Sy readonly Ns = Ns Sy on Ns | Ns Sy off Controls whether this dataset can be modified. The default value is .Sy off . The values .Sy on and .Sy off are equivalent to the .Sy ro and .Sy rw mount options. .Pp This property can also be referred to by its shortened column name, .Sy rdonly . .It Sy recordsize Ns = Ns Ar size Specifies a suggested block size for files in the file system. This property is designed solely for use with database workloads that access files in fixed-size records. ZFS automatically tunes block sizes according to internal algorithms optimized for typical access patterns. .Pp For databases that create very large files but access them in small random chunks, these algorithms may be suboptimal. Specifying a .Sy recordsize greater than or equal to the record size of the database can result in significant performance gains. Use of this property for general purpose file systems is strongly discouraged, and may adversely affect performance. .Pp The size specified must be a power of two greater than or equal to .Ar 512 B and less than or equal to .Ar 128 KiB . If the .Sy large_blocks feature is enabled on the pool, the size may be up to .Ar 1 MiB . See .Xr zpool-features 7 for details on ZFS feature flags. .Pp Changing the file system's .Sy recordsize affects only files created afterward; existing files are unaffected. .Pp This property can also be referred to by its shortened column name, .Sy recsize . .It Sy redundant_metadata Ns = Ns Sy all Ns | Ns Sy most Ns | Ns Sy some Ns | Ns Sy none Controls what types of metadata are stored redundantly. ZFS stores an extra copy of metadata, so that if a single block is corrupted, the amount of user data lost is limited. This extra copy is in addition to any redundancy provided at the pool level .Pq e.g. by mirroring or RAID-Z , and is in addition to an extra copy specified by the .Sy copies property .Pq up to a total of 3 copies . For example if the pool is mirrored, .Sy copies Ns = Ns 2 , and .Sy redundant_metadata Ns = Ns Sy most , then ZFS stores 6 copies of most metadata, and 4 copies of data and some metadata. .Pp When set to .Sy all , ZFS stores an extra copy of all metadata. If a single on-disk block is corrupt, at worst a single block of user data .Po which is .Sy recordsize bytes long .Pc can be lost. .Pp When set to .Sy most , ZFS stores an extra copy of most types of metadata. This can improve performance of random writes, because less metadata must be written. In practice, at worst about 1000 blocks .Po of .Sy recordsize bytes each .Pc of user data can be lost if a single on-disk block is corrupt. The exact behavior of which metadata blocks are stored redundantly may change in future releases. .Pp When set to .Sy some , ZFS stores an extra copy of only critical metadata. This can improve file create performance since less metadata needs to be written. If a single on-disk block is corrupt, at worst a single user file can be lost. .Pp When set to .Sy none , ZFS does not store any copies of metadata redundantly. If a single on-disk block is corrupt, an entire dataset can be lost. .Pp The default value is .Sy all . .It Sy refquota Ns = Ns Ar size Ns | Ns Sy none Limits the amount of space a dataset can consume. This property enforces a hard limit on the amount of space used. This hard limit does not include space used by descendents, including file systems and snapshots. .It Sy refreservation Ns = Ns Ar size Ns | Ns Sy none Ns | Ns Sy auto The minimum amount of space guaranteed to a dataset, not including its descendents. When the amount of space used is below this value, the dataset is treated as if it were taking up the amount of space specified by .Sy refreservation . The .Sy refreservation reservation is accounted for in the parent datasets' space used, and counts against the parent datasets' quotas and reservations. .Pp If .Sy refreservation is set, a snapshot is only allowed if there is enough free pool space outside of this reservation to accommodate the current number of .Qq referenced bytes in the dataset. .Pp If .Sy refreservation is set to .Sy auto , a volume is thick provisioned .Po or .Qq not sparse .Pc . .Sy refreservation Ns = Ns Sy auto is only supported on volumes. See .Sy volsize in the .Sx Native Properties section for more information about sparse volumes. .Pp This property can also be referred to by its shortened column name, .Sy refreserv . .It Sy relatime Ns = Ns Sy on Ns | Ns Sy off Controls the manner in which the access time is updated when .Sy atime Ns = Ns Sy on is set. Turning this property on causes the access time to be updated relative to the modify or change time. Access time is only updated if the previous access time was earlier than the current modify or change time or if the existing access time hasn't been updated within the past 24 hours. The default value is .Sy on . The values .Sy on and .Sy off are equivalent to the .Sy relatime and .Sy norelatime mount options. .It Sy reservation Ns = Ns Ar size Ns | Ns Sy none The minimum amount of space guaranteed to a dataset and its descendants. When the amount of space used is below this value, the dataset is treated as if it were taking up the amount of space specified by its reservation. Reservations are accounted for in the parent datasets' space used, and count against the parent datasets' quotas and reservations. .Pp This property can also be referred to by its shortened column name, .Sy reserv . .It Sy secondarycache Ns = Ns Sy all Ns | Ns Sy none Ns | Ns Sy metadata Controls what is cached in the secondary cache .Pq L2ARC . If this property is set to .Sy all , then both user data and metadata is cached. If this property is set to .Sy none , then neither user data nor metadata is cached. If this property is set to .Sy metadata , then only metadata is cached. The default value is .Sy all . +.It Sy prefetch Ns = Ns Sy all Ns | Ns Sy none Ns | Ns Sy metadata +Controls what speculative prefetch does. +If this property is set to +.Sy all , +then both user data and metadata are prefetched. +If this property is set to +.Sy none , +then neither user data nor metadata are prefetched. +If this property is set to +.Sy metadata , +then only metadata are prefetched. +The default value is +.Sy all . +.Pp +Please note that the module parameter zfs_disable_prefetch=1 can +be used to totally disable speculative prefetch, bypassing anything +this property does. .It Sy setuid Ns = Ns Sy on Ns | Ns Sy off Controls whether the setuid bit is respected for the file system. The default value is .Sy on . The values .Sy on and .Sy off are equivalent to the .Sy suid and .Sy nosuid mount options. .It Sy sharesmb Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Ar opts Controls whether the file system is shared by using .Sy Samba USERSHARES and what options are to be used. Otherwise, the file system is automatically shared and unshared with the .Nm zfs Cm share and .Nm zfs Cm unshare commands. If the property is set to on, the .Xr net 8 command is invoked to create a .Sy USERSHARE . .Pp Because SMB shares requires a resource name, a unique resource name is constructed from the dataset name. The constructed name is a copy of the dataset name except that the characters in the dataset name, which would be invalid in the resource name, are replaced with underscore (_) characters. Linux does not currently support additional options which might be available on Solaris. .Pp If the .Sy sharesmb property is set to .Sy off , the file systems are unshared. .Pp The share is created with the ACL (Access Control List) "Everyone:F" ("F" stands for "full permissions", i.e. read and write permissions) and no guest access (which means Samba must be able to authenticate a real user \(em .Xr passwd 5 Ns / Ns Xr shadow 5 Ns - , LDAP- or .Xr smbpasswd 5 Ns -based ) by default. This means that any additional access control (disallow specific user specific access etc) must be done on the underlying file system. .Pp When the .Sy sharesmb property is updated with .Nm zfs Cm set Fl u , the property is set to desired value, but the operation to share, reshare or unshare the the dataset is not performed. .It Sy sharenfs Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Ar opts Controls whether the file system is shared via NFS, and what options are to be used. A file system with a .Sy sharenfs property of .Sy off is managed with the .Xr exportfs 8 command and entries in the .Pa /etc/exports file. Otherwise, the file system is automatically shared and unshared with the .Nm zfs Cm share and .Nm zfs Cm unshare commands. If the property is set to .Sy on , the dataset is shared using the default options: .Dl sec=sys,rw,crossmnt,no_subtree_check .Pp Please note that the options are comma-separated, unlike those found in .Xr exports 5 . This is done to negate the need for quoting, as well as to make parsing with scripts easier. .Pp See .Xr exports 5 for the meaning of the default options. Otherwise, the .Xr exportfs 8 command is invoked with options equivalent to the contents of this property. .Pp When the .Sy sharenfs property is changed for a dataset, the dataset and any children inheriting the property are re-shared with the new options, only if the property was previously .Sy off , or if they were shared before the property was changed. If the new property is .Sy off , the file systems are unshared. .Pp When the .Sy sharenfs property is updated with .Nm zfs Cm set Fl u , the property is set to desired value, but the operation to share, reshare or unshare the the dataset is not performed. .It Sy logbias Ns = Ns Sy latency Ns | Ns Sy throughput Provide a hint to ZFS about handling of synchronous requests in this dataset. If .Sy logbias is set to .Sy latency .Pq the default , ZFS will use pool log devices .Pq if configured to handle the requests at low latency. If .Sy logbias is set to .Sy throughput , ZFS will not use configured pool log devices. ZFS will instead optimize synchronous operations for global pool throughput and efficient use of resources. .It Sy snapdev Ns = Ns Sy hidden Ns | Ns Sy visible Controls whether the volume snapshot devices under .Pa /dev/zvol/ Ns Aq Ar pool are hidden or visible. The default value is .Sy hidden . .It Sy snapdir Ns = Ns Sy hidden Ns | Ns Sy visible Controls whether the .Pa .zfs directory is hidden or visible in the root of the file system as discussed in the .Sx Snapshots section of .Xr zfsconcepts 7 . The default value is .Sy hidden . .It Sy sync Ns = Ns Sy standard Ns | Ns Sy always Ns | Ns Sy disabled Controls the behavior of synchronous requests .Pq e.g. fsync, O_DSYNC . .Sy standard is the POSIX-specified behavior of ensuring all synchronous requests are written to stable storage and all devices are flushed to ensure data is not cached by device controllers .Pq this is the default . .Sy always causes every file system transaction to be written and flushed before its system call returns. This has a large performance penalty. .Sy disabled disables synchronous requests. File system transactions are only committed to stable storage periodically. This option will give the highest performance. However, it is very dangerous as ZFS would be ignoring the synchronous transaction demands of applications such as databases or NFS. Administrators should only use this option when the risks are understood. .It Sy version Ns = Ns Ar N Ns | Ns Sy current The on-disk version of this file system, which is independent of the pool version. This property can only be set to later supported versions. See the .Nm zfs Cm upgrade command. .It Sy volsize Ns = Ns Ar size For volumes, specifies the logical size of the volume. By default, creating a volume establishes a reservation of equal size. For storage pools with a version number of 9 or higher, a .Sy refreservation is set instead. Any changes to .Sy volsize are reflected in an equivalent change to the reservation .Pq or Sy refreservation . The .Sy volsize can only be set to a multiple of .Sy volblocksize , and cannot be zero. .Pp The reservation is kept equal to the volume's logical size to prevent unexpected behavior for consumers. Without the reservation, the volume could run out of space, resulting in undefined behavior or data corruption, depending on how the volume is used. These effects can also occur when the volume size is changed while it is in use .Pq particularly when shrinking the size . Extreme care should be used when adjusting the volume size. .Pp Though not recommended, a .Qq sparse volume .Po also known as .Qq thin provisioned .Pc can be created by specifying the .Fl s option to the .Nm zfs Cm create Fl V command, or by changing the value of the .Sy refreservation property .Po or .Sy reservation property on pool version 8 or earlier .Pc after the volume has been created. A .Qq sparse volume is a volume where the value of .Sy refreservation is less than the size of the volume plus the space required to store its metadata. Consequently, writes to a sparse volume can fail with .Er ENOSPC when the pool is low on space. For a sparse volume, changes to .Sy volsize are not reflected in the .Sy refreservation . A volume that is not sparse is said to be .Qq thick provisioned . A sparse volume can become thick provisioned by setting .Sy refreservation to .Sy auto . .It Sy volmode Ns = Ns Sy default Ns | Ns Sy full Ns | Ns Sy geom Ns | Ns Sy dev Ns | Ns Sy none This property specifies how volumes should be exposed to the OS. Setting it to .Sy full exposes volumes as fully fledged block devices, providing maximal functionality. The value .Sy geom is just an alias for .Sy full and is kept for compatibility. Setting it to .Sy dev hides its partitions. Volumes with property set to .Sy none are not exposed outside ZFS, but can be snapshotted, cloned, replicated, etc, that can be suitable for backup purposes. Value .Sy default means that volumes exposition is controlled by system-wide tunable .Sy zvol_volmode , where .Sy full , .Sy dev and .Sy none are encoded as 1, 2 and 3 respectively. The default value is .Sy full . .It Sy vscan Ns = Ns Sy on Ns | Ns Sy off Controls whether regular files should be scanned for viruses when a file is opened and closed. In addition to enabling this property, the virus scan service must also be enabled for virus scanning to occur. The default value is .Sy off . This property is not used by OpenZFS. .It Sy xattr Ns = Ns Sy on Ns | Ns Sy off Ns | Ns Sy sa Controls whether extended attributes are enabled for this file system. Two styles of extended attributes are supported: either directory-based or system-attribute-based. .Pp The default value of .Sy on enables directory-based extended attributes. This style of extended attribute imposes no practical limit on either the size or number of attributes which can be set on a file. Although under Linux the .Xr getxattr 2 and .Xr setxattr 2 system calls limit the maximum size to .Sy 64K . This is the most compatible style of extended attribute and is supported by all ZFS implementations. .Pp System-attribute-based xattrs can be enabled by setting the value to .Sy sa . The key advantage of this type of xattr is improved performance. Storing extended attributes as system attributes significantly decreases the amount of disk I/O required. Up to .Sy 64K of data may be stored per-file in the space reserved for system attributes. If there is not enough space available for an extended attribute then it will be automatically written as a directory-based xattr. System-attribute-based extended attributes are not accessible on platforms which do not support the .Sy xattr Ns = Ns Sy sa feature. OpenZFS supports .Sy xattr Ns = Ns Sy sa on both .Fx and Linux. .Pp The use of system-attribute-based xattrs is strongly encouraged for users of SELinux or POSIX ACLs. Both of these features heavily rely on extended attributes and benefit significantly from the reduced access time. .Pp The values .Sy on and .Sy off are equivalent to the .Sy xattr and .Sy noxattr mount options. .It Sy jailed Ns = Ns Sy off Ns | Ns Sy on Controls whether the dataset is managed from a jail. See .Xr zfs-jail 8 for more information. Jails are a .Fx feature and this property is not available on other platforms. .It Sy zoned Ns = Ns Sy off Ns | Ns Sy on Controls whether the dataset is managed from a non-global zone or namespace. See .Xr zfs-zone 8 for more information. Zoning is a Linux feature and this property is not available on other platforms. .El .Pp The following three properties cannot be changed after the file system is created, and therefore, should be set when the file system is created. If the properties are not set with the .Nm zfs Cm create or .Nm zpool Cm create commands, these properties are inherited from the parent dataset. If the parent dataset lacks these properties due to having been created prior to these features being supported, the new file system will have the default values for these properties. .Bl -tag -width "" .It Xo .Sy casesensitivity Ns = Ns Sy sensitive Ns | Ns .Sy insensitive Ns | Ns Sy mixed .Xc Indicates whether the file name matching algorithm used by the file system should be case-sensitive, case-insensitive, or allow a combination of both styles of matching. The default value for the .Sy casesensitivity property is .Sy sensitive . Traditionally, .Ux and POSIX file systems have case-sensitive file names. .Pp The .Sy mixed value for the .Sy casesensitivity property indicates that the file system can support requests for both case-sensitive and case-insensitive matching behavior. Currently, case-insensitive matching behavior on a file system that supports mixed behavior is limited to the SMB server product. For more information about the .Sy mixed value behavior, see the "ZFS Administration Guide". .It Xo .Sy normalization Ns = Ns Sy none Ns | Ns Sy formC Ns | Ns .Sy formD Ns | Ns Sy formKC Ns | Ns Sy formKD .Xc Indicates whether the file system should perform a .Sy unicode normalization of file names whenever two file names are compared, and which normalization algorithm should be used. File names are always stored unmodified, names are normalized as part of any comparison process. If this property is set to a legal value other than .Sy none , and the .Sy utf8only property was left unspecified, the .Sy utf8only property is automatically set to .Sy on . The default value of the .Sy normalization property is .Sy none . This property cannot be changed after the file system is created. .It Sy utf8only Ns = Ns Sy on Ns | Ns Sy off Indicates whether the file system should reject file names that include characters that are not present in the .Sy UTF-8 character code set. If this property is explicitly set to .Sy off , the normalization property must either not be explicitly set or be set to .Sy none . The default value for the .Sy utf8only property is .Sy off . This property cannot be changed after the file system is created. .El .Pp The .Sy casesensitivity , .Sy normalization , and .Sy utf8only properties are also new permissions that can be assigned to non-privileged users by using the ZFS delegated administration feature. . .Ss Temporary Mount Point Properties When a file system is mounted, either through .Xr mount 8 for legacy mounts or the .Nm zfs Cm mount command for normal file systems, its mount options are set according to its properties. The correlation between properties and mount options is as follows: .Bl -tag -compact -offset Ds -width "rootcontext=" .It Sy atime atime/noatime .It Sy canmount auto/noauto .It Sy devices dev/nodev .It Sy exec exec/noexec .It Sy readonly ro/rw .It Sy relatime relatime/norelatime .It Sy setuid suid/nosuid .It Sy xattr xattr/noxattr .It Sy nbmand mand/nomand .It Sy context Ns = context= .It Sy fscontext Ns = fscontext= .It Sy defcontext Ns = defcontext= .It Sy rootcontext Ns = rootcontext= .El .Pp In addition, these options can be set on a per-mount basis using the .Fl o option, without affecting the property that is stored on disk. The values specified on the command line override the values stored in the dataset. The .Sy nosuid option is an alias for .Sy nodevices , Ns Sy nosetuid . These properties are reported as .Qq temporary by the .Nm zfs Cm get command. If the properties are changed while the dataset is mounted, the new setting overrides any temporary settings. . .Ss User Properties In addition to the standard native properties, ZFS supports arbitrary user properties. User properties have no effect on ZFS behavior, but applications or administrators can use them to annotate datasets .Pq file systems, volumes, and snapshots . .Pp User property names must contain a colon .Pq Qq Sy \&: character to distinguish them from native properties. They may contain lowercase letters, numbers, and the following punctuation characters: colon .Pq Qq Sy \&: , dash .Pq Qq Sy - , period .Pq Qq Sy \&. , and underscore .Pq Qq Sy _ . The expected convention is that the property name is divided into two portions such as .Ar module : Ns Ar property , but this namespace is not enforced by ZFS. User property names can be at most 256 characters, and cannot begin with a dash .Pq Qq Sy - . .Pp When making programmatic use of user properties, it is strongly suggested to use a reversed DNS domain name for the .Ar module component of property names to reduce the chance that two independently-developed packages use the same property name for different purposes. .Pp The values of user properties are arbitrary strings, are always inherited, and are never validated. All of the commands that operate on properties .Po Nm zfs Cm list , .Nm zfs Cm get , .Nm zfs Cm set , and so forth .Pc can be used to manipulate both native properties and user properties. Use the .Nm zfs Cm inherit command to clear a user property. If the property is not defined in any parent dataset, it is removed entirely. Property values are limited to 8192 bytes. diff --git a/module/zcommon/zfs_prop.c b/module/zcommon/zfs_prop.c index 3db6fd13f4ae..29764674a31b 100644 --- a/module/zcommon/zfs_prop.c +++ b/module/zcommon/zfs_prop.c @@ -1,1101 +1,1112 @@ /* * 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) 2011, 2018 by Delphix. All rights reserved. * Copyright (c) 2013 by Saso Kiselkov. All rights reserved. * Copyright 2016, Joyent, Inc. * Copyright (c) 2019, Klara Inc. * Copyright (c) 2019, Allan Jude * Copyright (c) 2022 Hewlett Packard Enterprise Development LP. */ /* Portions Copyright 2010 Robert Milkowski */ #if defined(_KERNEL) #include #endif #include #include #include #include #include #include #include #include "zfs_prop.h" #include "zfs_deleg.h" #include "zfs_fletcher.h" #if !defined(_KERNEL) #include #include #include #endif static zprop_desc_t zfs_prop_table[ZFS_NUM_PROPS]; /* Note this is indexed by zfs_userquota_prop_t, keep the order the same */ const char *const zfs_userquota_prop_prefixes[] = { "userused@", "userquota@", "groupused@", "groupquota@", "userobjused@", "userobjquota@", "groupobjused@", "groupobjquota@", "projectused@", "projectquota@", "projectobjused@", "projectobjquota@" }; zprop_desc_t * zfs_prop_get_table(void) { return (zfs_prop_table); } void zfs_prop_init(void) { static const zprop_index_t checksum_table[] = { { "on", ZIO_CHECKSUM_ON }, { "off", ZIO_CHECKSUM_OFF }, { "fletcher2", ZIO_CHECKSUM_FLETCHER_2 }, { "fletcher4", ZIO_CHECKSUM_FLETCHER_4 }, { "sha256", ZIO_CHECKSUM_SHA256 }, { "noparity", ZIO_CHECKSUM_NOPARITY }, { "sha512", ZIO_CHECKSUM_SHA512 }, { "skein", ZIO_CHECKSUM_SKEIN }, { "edonr", ZIO_CHECKSUM_EDONR }, { "blake3", ZIO_CHECKSUM_BLAKE3 }, { NULL } }; static const zprop_index_t dedup_table[] = { { "on", ZIO_CHECKSUM_ON }, { "off", ZIO_CHECKSUM_OFF }, { "verify", ZIO_CHECKSUM_ON | ZIO_CHECKSUM_VERIFY }, { "sha256", ZIO_CHECKSUM_SHA256 }, { "sha256,verify", ZIO_CHECKSUM_SHA256 | ZIO_CHECKSUM_VERIFY }, { "sha512", ZIO_CHECKSUM_SHA512 }, { "sha512,verify", ZIO_CHECKSUM_SHA512 | ZIO_CHECKSUM_VERIFY }, { "skein", ZIO_CHECKSUM_SKEIN }, { "skein,verify", ZIO_CHECKSUM_SKEIN | ZIO_CHECKSUM_VERIFY }, { "edonr,verify", ZIO_CHECKSUM_EDONR | ZIO_CHECKSUM_VERIFY }, { "blake3", ZIO_CHECKSUM_BLAKE3 }, { "blake3,verify", ZIO_CHECKSUM_BLAKE3 | ZIO_CHECKSUM_VERIFY }, { NULL } }; static const zprop_index_t compress_table[] = { { "on", ZIO_COMPRESS_ON }, { "off", ZIO_COMPRESS_OFF }, { "lzjb", ZIO_COMPRESS_LZJB }, { "gzip", ZIO_COMPRESS_GZIP_6 }, /* gzip default */ { "gzip-1", ZIO_COMPRESS_GZIP_1 }, { "gzip-2", ZIO_COMPRESS_GZIP_2 }, { "gzip-3", ZIO_COMPRESS_GZIP_3 }, { "gzip-4", ZIO_COMPRESS_GZIP_4 }, { "gzip-5", ZIO_COMPRESS_GZIP_5 }, { "gzip-6", ZIO_COMPRESS_GZIP_6 }, { "gzip-7", ZIO_COMPRESS_GZIP_7 }, { "gzip-8", ZIO_COMPRESS_GZIP_8 }, { "gzip-9", ZIO_COMPRESS_GZIP_9 }, { "zle", ZIO_COMPRESS_ZLE }, { "lz4", ZIO_COMPRESS_LZ4 }, { "zstd", ZIO_COMPRESS_ZSTD }, { "zstd-fast", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_DEFAULT) }, /* * ZSTD 1-19 are synthetic. We store the compression level in a * separate hidden property to avoid wasting a large amount of * space in the ZIO_COMPRESS enum. * * The compression level is also stored within the header of the * compressed block since we may need it for later recompression * to avoid checksum errors (L2ARC). * * Note that the level here is defined as bit shifted mask on * top of the method. */ { "zstd-1", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_1) }, { "zstd-2", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_2) }, { "zstd-3", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_3) }, { "zstd-4", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_4) }, { "zstd-5", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_5) }, { "zstd-6", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_6) }, { "zstd-7", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_7) }, { "zstd-8", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_8) }, { "zstd-9", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_9) }, { "zstd-10", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_10) }, { "zstd-11", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_11) }, { "zstd-12", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_12) }, { "zstd-13", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_13) }, { "zstd-14", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_14) }, { "zstd-15", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_15) }, { "zstd-16", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_16) }, { "zstd-17", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_17) }, { "zstd-18", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_18) }, { "zstd-19", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_19) }, /* * The ZSTD-Fast levels are also synthetic. */ { "zstd-fast-1", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_1) }, { "zstd-fast-2", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_2) }, { "zstd-fast-3", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_3) }, { "zstd-fast-4", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_4) }, { "zstd-fast-5", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_5) }, { "zstd-fast-6", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_6) }, { "zstd-fast-7", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_7) }, { "zstd-fast-8", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_8) }, { "zstd-fast-9", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_9) }, { "zstd-fast-10", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_10) }, { "zstd-fast-20", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_20) }, { "zstd-fast-30", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_30) }, { "zstd-fast-40", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_40) }, { "zstd-fast-50", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_50) }, { "zstd-fast-60", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_60) }, { "zstd-fast-70", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_70) }, { "zstd-fast-80", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_80) }, { "zstd-fast-90", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_90) }, { "zstd-fast-100", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_100) }, { "zstd-fast-500", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_500) }, { "zstd-fast-1000", ZIO_COMPLEVEL_ZSTD(ZIO_ZSTD_LEVEL_FAST_1000) }, { NULL } }; static const zprop_index_t crypto_table[] = { { "on", ZIO_CRYPT_ON }, { "off", ZIO_CRYPT_OFF }, { "aes-128-ccm", ZIO_CRYPT_AES_128_CCM }, { "aes-192-ccm", ZIO_CRYPT_AES_192_CCM }, { "aes-256-ccm", ZIO_CRYPT_AES_256_CCM }, { "aes-128-gcm", ZIO_CRYPT_AES_128_GCM }, { "aes-192-gcm", ZIO_CRYPT_AES_192_GCM }, { "aes-256-gcm", ZIO_CRYPT_AES_256_GCM }, { NULL } }; static const zprop_index_t keyformat_table[] = { { "none", ZFS_KEYFORMAT_NONE }, { "raw", ZFS_KEYFORMAT_RAW }, { "hex", ZFS_KEYFORMAT_HEX }, { "passphrase", ZFS_KEYFORMAT_PASSPHRASE }, { NULL } }; static const zprop_index_t snapdir_table[] = { { "hidden", ZFS_SNAPDIR_HIDDEN }, { "visible", ZFS_SNAPDIR_VISIBLE }, { NULL } }; static const zprop_index_t snapdev_table[] = { { "hidden", ZFS_SNAPDEV_HIDDEN }, { "visible", ZFS_SNAPDEV_VISIBLE }, { NULL } }; static const zprop_index_t acl_mode_table[] = { { "discard", ZFS_ACL_DISCARD }, { "groupmask", ZFS_ACL_GROUPMASK }, { "passthrough", ZFS_ACL_PASSTHROUGH }, { "restricted", ZFS_ACL_RESTRICTED }, { NULL } }; static const zprop_index_t acltype_table[] = { { "off", ZFS_ACLTYPE_OFF }, { "posix", ZFS_ACLTYPE_POSIX }, { "nfsv4", ZFS_ACLTYPE_NFSV4 }, { "disabled", ZFS_ACLTYPE_OFF }, /* bkwrd compatibility */ { "noacl", ZFS_ACLTYPE_OFF }, /* bkwrd compatibility */ { "posixacl", ZFS_ACLTYPE_POSIX }, /* bkwrd compatibility */ { NULL } }; static const zprop_index_t acl_inherit_table[] = { { "discard", ZFS_ACL_DISCARD }, { "noallow", ZFS_ACL_NOALLOW }, { "restricted", ZFS_ACL_RESTRICTED }, { "passthrough", ZFS_ACL_PASSTHROUGH }, { "secure", ZFS_ACL_RESTRICTED }, /* bkwrd compatibility */ { "passthrough-x", ZFS_ACL_PASSTHROUGH_X }, { NULL } }; static const zprop_index_t case_table[] = { { "sensitive", ZFS_CASE_SENSITIVE }, { "insensitive", ZFS_CASE_INSENSITIVE }, { "mixed", ZFS_CASE_MIXED }, { NULL } }; static const zprop_index_t copies_table[] = { { "1", 1 }, { "2", 2 }, { "3", 3 }, { NULL } }; /* * Use the unique flags we have to send to u8_strcmp() and/or * u8_textprep() to represent the various normalization property * values. */ static const zprop_index_t normalize_table[] = { { "none", 0 }, { "formD", U8_TEXTPREP_NFD }, { "formKC", U8_TEXTPREP_NFKC }, { "formC", U8_TEXTPREP_NFC }, { "formKD", U8_TEXTPREP_NFKD }, { NULL } }; static const zprop_index_t version_table[] = { { "1", 1 }, { "2", 2 }, { "3", 3 }, { "4", 4 }, { "5", 5 }, { "current", ZPL_VERSION }, { NULL } }; static const zprop_index_t boolean_table[] = { { "off", 0 }, { "on", 1 }, { NULL } }; static const zprop_index_t keystatus_table[] = { { "none", ZFS_KEYSTATUS_NONE}, { "unavailable", ZFS_KEYSTATUS_UNAVAILABLE}, { "available", ZFS_KEYSTATUS_AVAILABLE}, { NULL } }; static const zprop_index_t logbias_table[] = { { "latency", ZFS_LOGBIAS_LATENCY }, { "throughput", ZFS_LOGBIAS_THROUGHPUT }, { NULL } }; static const zprop_index_t canmount_table[] = { { "off", ZFS_CANMOUNT_OFF }, { "on", ZFS_CANMOUNT_ON }, { "noauto", ZFS_CANMOUNT_NOAUTO }, { NULL } }; static const zprop_index_t cache_table[] = { { "none", ZFS_CACHE_NONE }, { "metadata", ZFS_CACHE_METADATA }, { "all", ZFS_CACHE_ALL }, { NULL } }; + static const zprop_index_t prefetch_table[] = { + { "none", ZFS_PREFETCH_NONE }, + { "metadata", ZFS_PREFETCH_METADATA }, + { "all", ZFS_PREFETCH_ALL }, + { NULL } + }; + static const zprop_index_t sync_table[] = { { "standard", ZFS_SYNC_STANDARD }, { "always", ZFS_SYNC_ALWAYS }, { "disabled", ZFS_SYNC_DISABLED }, { NULL } }; static const zprop_index_t xattr_table[] = { { "off", ZFS_XATTR_OFF }, { "on", ZFS_XATTR_DIR }, { "sa", ZFS_XATTR_SA }, { "dir", ZFS_XATTR_DIR }, { NULL } }; static const zprop_index_t dnsize_table[] = { { "legacy", ZFS_DNSIZE_LEGACY }, { "auto", ZFS_DNSIZE_AUTO }, { "1k", ZFS_DNSIZE_1K }, { "2k", ZFS_DNSIZE_2K }, { "4k", ZFS_DNSIZE_4K }, { "8k", ZFS_DNSIZE_8K }, { "16k", ZFS_DNSIZE_16K }, { NULL } }; static const zprop_index_t redundant_metadata_table[] = { { "all", ZFS_REDUNDANT_METADATA_ALL }, { "most", ZFS_REDUNDANT_METADATA_MOST }, { "some", ZFS_REDUNDANT_METADATA_SOME }, { "none", ZFS_REDUNDANT_METADATA_NONE }, { NULL } }; static const zprop_index_t volmode_table[] = { { "default", ZFS_VOLMODE_DEFAULT }, { "full", ZFS_VOLMODE_GEOM }, { "geom", ZFS_VOLMODE_GEOM }, { "dev", ZFS_VOLMODE_DEV }, { "none", ZFS_VOLMODE_NONE }, { NULL } }; struct zfs_mod_supported_features *sfeatures = zfs_mod_list_supported(ZFS_SYSFS_DATASET_PROPERTIES); /* inherit index properties */ zprop_register_index(ZFS_PROP_REDUNDANT_METADATA, "redundant_metadata", ZFS_REDUNDANT_METADATA_ALL, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "all | most | some | none", "REDUND_MD", redundant_metadata_table, sfeatures); zprop_register_index(ZFS_PROP_SYNC, "sync", ZFS_SYNC_STANDARD, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "standard | always | disabled", "SYNC", sync_table, sfeatures); zprop_register_index(ZFS_PROP_CHECKSUM, "checksum", ZIO_CHECKSUM_DEFAULT, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "on | off | fletcher2 | fletcher4 | sha256 | sha512 | skein" " | edonr | blake3", "CHECKSUM", checksum_table, sfeatures); zprop_register_index(ZFS_PROP_DEDUP, "dedup", ZIO_CHECKSUM_OFF, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "on | off | verify | sha256[,verify] | sha512[,verify] | " "skein[,verify] | edonr,verify | blake3[,verify]", "DEDUP", dedup_table, sfeatures); zprop_register_index(ZFS_PROP_COMPRESSION, "compression", ZIO_COMPRESS_DEFAULT, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "on | off | lzjb | gzip | gzip-[1-9] | zle | lz4 | " "zstd | zstd-[1-19] | " "zstd-fast | zstd-fast-[1-10,20,30,40,50,60,70,80,90,100,500,1000]", "COMPRESS", compress_table, sfeatures); zprop_register_index(ZFS_PROP_SNAPDIR, "snapdir", ZFS_SNAPDIR_HIDDEN, PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "hidden | visible", "SNAPDIR", snapdir_table, sfeatures); zprop_register_index(ZFS_PROP_SNAPDEV, "snapdev", ZFS_SNAPDEV_HIDDEN, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "hidden | visible", "SNAPDEV", snapdev_table, sfeatures); zprop_register_index(ZFS_PROP_ACLMODE, "aclmode", ZFS_ACL_DISCARD, PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "discard | groupmask | passthrough | restricted", "ACLMODE", acl_mode_table, sfeatures); zprop_register_index(ZFS_PROP_ACLTYPE, "acltype", #ifdef __linux__ /* Linux doesn't natively support ZFS's NFSv4-style ACLs. */ ZFS_ACLTYPE_OFF, #else ZFS_ACLTYPE_NFSV4, #endif PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT, "off | nfsv4 | posix", "ACLTYPE", acltype_table, sfeatures); zprop_register_index(ZFS_PROP_ACLINHERIT, "aclinherit", ZFS_ACL_RESTRICTED, PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "discard | noallow | restricted | passthrough | passthrough-x", "ACLINHERIT", acl_inherit_table, sfeatures); zprop_register_index(ZFS_PROP_COPIES, "copies", 1, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "1 | 2 | 3", "COPIES", copies_table, sfeatures); zprop_register_index(ZFS_PROP_PRIMARYCACHE, "primarycache", ZFS_CACHE_ALL, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT | ZFS_TYPE_VOLUME, "all | none | metadata", "PRIMARYCACHE", cache_table, sfeatures); zprop_register_index(ZFS_PROP_SECONDARYCACHE, "secondarycache", ZFS_CACHE_ALL, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT | ZFS_TYPE_VOLUME, "all | none | metadata", "SECONDARYCACHE", cache_table, sfeatures); + zprop_register_index(ZFS_PROP_PREFETCH, "prefetch", + ZFS_PREFETCH_ALL, PROP_INHERIT, + ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT | ZFS_TYPE_VOLUME, + "none | metadata | all", "PREFETCH", prefetch_table, sfeatures); zprop_register_index(ZFS_PROP_LOGBIAS, "logbias", ZFS_LOGBIAS_LATENCY, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "latency | throughput", "LOGBIAS", logbias_table, sfeatures); zprop_register_index(ZFS_PROP_XATTR, "xattr", ZFS_XATTR_DIR, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT, "on | off | dir | sa", "XATTR", xattr_table, sfeatures); zprop_register_index(ZFS_PROP_DNODESIZE, "dnodesize", ZFS_DNSIZE_LEGACY, PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "legacy | auto | 1k | 2k | 4k | 8k | 16k", "DNSIZE", dnsize_table, sfeatures); zprop_register_index(ZFS_PROP_VOLMODE, "volmode", ZFS_VOLMODE_DEFAULT, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "default | full | geom | dev | none", "VOLMODE", volmode_table, sfeatures); /* inherit index (boolean) properties */ zprop_register_index(ZFS_PROP_ATIME, "atime", 1, PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "on | off", "ATIME", boolean_table, sfeatures); zprop_register_index(ZFS_PROP_RELATIME, "relatime", 1, PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "on | off", "RELATIME", boolean_table, sfeatures); zprop_register_index(ZFS_PROP_DEVICES, "devices", 1, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT, "on | off", "DEVICES", boolean_table, sfeatures); zprop_register_index(ZFS_PROP_EXEC, "exec", 1, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT, "on | off", "EXEC", boolean_table, sfeatures); zprop_register_index(ZFS_PROP_SETUID, "setuid", 1, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT, "on | off", "SETUID", boolean_table, sfeatures); zprop_register_index(ZFS_PROP_READONLY, "readonly", 0, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "on | off", "RDONLY", boolean_table, sfeatures); #ifdef __FreeBSD__ zprop_register_index(ZFS_PROP_ZONED, "jailed", 0, PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "on | off", "JAILED", boolean_table, sfeatures); #else zprop_register_index(ZFS_PROP_ZONED, "zoned", 0, PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "on | off", "ZONED", boolean_table, sfeatures); #endif zprop_register_index(ZFS_PROP_VSCAN, "vscan", 0, PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "on | off", "VSCAN", boolean_table, sfeatures); zprop_register_index(ZFS_PROP_NBMAND, "nbmand", 0, PROP_INHERIT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT, "on | off", "NBMAND", boolean_table, sfeatures); zprop_register_index(ZFS_PROP_OVERLAY, "overlay", 1, PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "on | off", "OVERLAY", boolean_table, sfeatures); /* default index properties */ zprop_register_index(ZFS_PROP_VERSION, "version", 0, PROP_DEFAULT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT, "1 | 2 | 3 | 4 | 5 | current", "VERSION", version_table, sfeatures); zprop_register_index(ZFS_PROP_CANMOUNT, "canmount", ZFS_CANMOUNT_ON, PROP_DEFAULT, ZFS_TYPE_FILESYSTEM, "on | off | noauto", "CANMOUNT", canmount_table, sfeatures); /* readonly index properties */ zprop_register_index(ZFS_PROP_MOUNTED, "mounted", 0, PROP_READONLY, ZFS_TYPE_FILESYSTEM, "yes | no", "MOUNTED", boolean_table, sfeatures); zprop_register_index(ZFS_PROP_DEFER_DESTROY, "defer_destroy", 0, PROP_READONLY, ZFS_TYPE_SNAPSHOT, "yes | no", "DEFER_DESTROY", boolean_table, sfeatures); zprop_register_index(ZFS_PROP_KEYSTATUS, "keystatus", ZFS_KEYSTATUS_NONE, PROP_READONLY, ZFS_TYPE_DATASET, "none | unavailable | available", "KEYSTATUS", keystatus_table, sfeatures); /* set once index properties */ zprop_register_index(ZFS_PROP_NORMALIZE, "normalization", 0, PROP_ONETIME, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT, "none | formC | formD | formKC | formKD", "NORMALIZATION", normalize_table, sfeatures); zprop_register_index(ZFS_PROP_CASE, "casesensitivity", ZFS_CASE_SENSITIVE, PROP_ONETIME, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT, "sensitive | insensitive | mixed", "CASE", case_table, sfeatures); zprop_register_index(ZFS_PROP_KEYFORMAT, "keyformat", ZFS_KEYFORMAT_NONE, PROP_ONETIME_DEFAULT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "none | raw | hex | passphrase", "KEYFORMAT", keyformat_table, sfeatures); zprop_register_index(ZFS_PROP_ENCRYPTION, "encryption", ZIO_CRYPT_DEFAULT, PROP_ONETIME, ZFS_TYPE_DATASET, "on | off | aes-128-ccm | aes-192-ccm | aes-256-ccm | " "aes-128-gcm | aes-192-gcm | aes-256-gcm", "ENCRYPTION", crypto_table, sfeatures); /* set once index (boolean) properties */ zprop_register_index(ZFS_PROP_UTF8ONLY, "utf8only", 0, PROP_ONETIME, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_SNAPSHOT, "on | off", "UTF8ONLY", boolean_table, sfeatures); /* string properties */ zprop_register_string(ZFS_PROP_ORIGIN, "origin", NULL, PROP_READONLY, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "", "ORIGIN", sfeatures); zprop_register_string(ZFS_PROP_CLONES, "clones", NULL, PROP_READONLY, ZFS_TYPE_SNAPSHOT, "[,...]", "CLONES", sfeatures); zprop_register_string(ZFS_PROP_MOUNTPOINT, "mountpoint", "/", PROP_INHERIT, ZFS_TYPE_FILESYSTEM, " | legacy | none", "MOUNTPOINT", sfeatures); zprop_register_string(ZFS_PROP_SHARENFS, "sharenfs", "off", PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "on | off | NFS share options", "SHARENFS", sfeatures); zprop_register_string(ZFS_PROP_TYPE, "type", NULL, PROP_READONLY, ZFS_TYPE_DATASET | ZFS_TYPE_BOOKMARK, "filesystem | volume | snapshot | bookmark", "TYPE", sfeatures); zprop_register_string(ZFS_PROP_SHARESMB, "sharesmb", "off", PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "on | off | SMB share options", "SHARESMB", sfeatures); zprop_register_string(ZFS_PROP_MLSLABEL, "mlslabel", ZFS_MLSLABEL_DEFAULT, PROP_INHERIT, ZFS_TYPE_DATASET, "", "MLSLABEL", sfeatures); zprop_register_string(ZFS_PROP_SELINUX_CONTEXT, "context", "none", PROP_DEFAULT, ZFS_TYPE_DATASET, "", "CONTEXT", sfeatures); zprop_register_string(ZFS_PROP_SELINUX_FSCONTEXT, "fscontext", "none", PROP_DEFAULT, ZFS_TYPE_DATASET, "", "FSCONTEXT", sfeatures); zprop_register_string(ZFS_PROP_SELINUX_DEFCONTEXT, "defcontext", "none", PROP_DEFAULT, ZFS_TYPE_DATASET, "", "DEFCONTEXT", sfeatures); zprop_register_string(ZFS_PROP_SELINUX_ROOTCONTEXT, "rootcontext", "none", PROP_DEFAULT, ZFS_TYPE_DATASET, "", "ROOTCONTEXT", sfeatures); zprop_register_string(ZFS_PROP_RECEIVE_RESUME_TOKEN, "receive_resume_token", NULL, PROP_READONLY, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "", "RESUMETOK", sfeatures); zprop_register_string(ZFS_PROP_ENCRYPTION_ROOT, "encryptionroot", NULL, PROP_READONLY, ZFS_TYPE_DATASET, "", "ENCROOT", sfeatures); zprop_register_string(ZFS_PROP_KEYLOCATION, "keylocation", "none", PROP_DEFAULT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "prompt | | | ", "KEYLOCATION", sfeatures); zprop_register_string(ZFS_PROP_REDACT_SNAPS, "redact_snaps", NULL, PROP_READONLY, ZFS_TYPE_DATASET | ZFS_TYPE_BOOKMARK, "[,...]", "RSNAPS", sfeatures); /* readonly number properties */ zprop_register_number(ZFS_PROP_USED, "used", 0, PROP_READONLY, ZFS_TYPE_DATASET, "", "USED", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_AVAILABLE, "available", 0, PROP_READONLY, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "", "AVAIL", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_REFERENCED, "referenced", 0, PROP_READONLY, ZFS_TYPE_DATASET | ZFS_TYPE_BOOKMARK, "", "REFER", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_COMPRESSRATIO, "compressratio", 0, PROP_READONLY, ZFS_TYPE_DATASET | ZFS_TYPE_BOOKMARK, "<1.00x or higher if compressed>", "RATIO", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_REFRATIO, "refcompressratio", 0, PROP_READONLY, ZFS_TYPE_DATASET, "<1.00x or higher if compressed>", "REFRATIO", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_VOLBLOCKSIZE, "volblocksize", ZVOL_DEFAULT_BLOCKSIZE, PROP_ONETIME, ZFS_TYPE_VOLUME, "512 to 128k, power of 2", "VOLBLOCK", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_USEDSNAP, "usedbysnapshots", 0, PROP_READONLY, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "", "USEDSNAP", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_USEDDS, "usedbydataset", 0, PROP_READONLY, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "", "USEDDS", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_USEDCHILD, "usedbychildren", 0, PROP_READONLY, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "", "USEDCHILD", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_USEDREFRESERV, "usedbyrefreservation", 0, PROP_READONLY, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "", "USEDREFRESERV", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_USERREFS, "userrefs", 0, PROP_READONLY, ZFS_TYPE_SNAPSHOT, "", "USERREFS", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_WRITTEN, "written", 0, PROP_READONLY, ZFS_TYPE_DATASET, "", "WRITTEN", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_LOGICALUSED, "logicalused", 0, PROP_READONLY, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "", "LUSED", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_LOGICALREFERENCED, "logicalreferenced", 0, PROP_READONLY, ZFS_TYPE_DATASET | ZFS_TYPE_BOOKMARK, "", "LREFER", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_FILESYSTEM_COUNT, "filesystem_count", UINT64_MAX, PROP_READONLY, ZFS_TYPE_FILESYSTEM, "", "FSCOUNT", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_SNAPSHOT_COUNT, "snapshot_count", UINT64_MAX, PROP_READONLY, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "", "SSCOUNT", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_GUID, "guid", 0, PROP_READONLY, ZFS_TYPE_DATASET | ZFS_TYPE_BOOKMARK, "", "GUID", B_TRUE, sfeatures); zprop_register_number(ZFS_PROP_CREATETXG, "createtxg", 0, PROP_READONLY, ZFS_TYPE_DATASET | ZFS_TYPE_BOOKMARK, "", "CREATETXG", B_TRUE, sfeatures); zprop_register_number(ZFS_PROP_PBKDF2_ITERS, "pbkdf2iters", 0, PROP_ONETIME_DEFAULT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "", "PBKDF2ITERS", B_TRUE, sfeatures); zprop_register_number(ZFS_PROP_OBJSETID, "objsetid", 0, PROP_READONLY, ZFS_TYPE_DATASET, "", "OBJSETID", B_TRUE, sfeatures); /* default number properties */ zprop_register_number(ZFS_PROP_QUOTA, "quota", 0, PROP_DEFAULT, ZFS_TYPE_FILESYSTEM, " | none", "QUOTA", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_RESERVATION, "reservation", 0, PROP_DEFAULT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, " | none", "RESERV", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_VOLSIZE, "volsize", 0, PROP_DEFAULT, ZFS_TYPE_SNAPSHOT | ZFS_TYPE_VOLUME, "", "VOLSIZE", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_REFQUOTA, "refquota", 0, PROP_DEFAULT, ZFS_TYPE_FILESYSTEM, " | none", "REFQUOTA", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_REFRESERVATION, "refreservation", 0, PROP_DEFAULT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, " | none", "REFRESERV", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_FILESYSTEM_LIMIT, "filesystem_limit", UINT64_MAX, PROP_DEFAULT, ZFS_TYPE_FILESYSTEM, " | none", "FSLIMIT", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_SNAPSHOT_LIMIT, "snapshot_limit", UINT64_MAX, PROP_DEFAULT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, " | none", "SSLIMIT", B_FALSE, sfeatures); /* inherit number properties */ zprop_register_number(ZFS_PROP_RECORDSIZE, "recordsize", SPA_OLD_MAXBLOCKSIZE, PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "512 to 1M, power of 2", "RECSIZE", B_FALSE, sfeatures); zprop_register_number(ZFS_PROP_SPECIAL_SMALL_BLOCKS, "special_small_blocks", 0, PROP_INHERIT, ZFS_TYPE_FILESYSTEM, "zero or 512 to 1M, power of 2", "SPECIAL_SMALL_BLOCKS", B_FALSE, sfeatures); /* hidden properties */ zprop_register_hidden(ZFS_PROP_NUMCLONES, "numclones", PROP_TYPE_NUMBER, PROP_READONLY, ZFS_TYPE_SNAPSHOT, "NUMCLONES", B_FALSE, sfeatures); zprop_register_hidden(ZFS_PROP_NAME, "name", PROP_TYPE_STRING, PROP_READONLY, ZFS_TYPE_DATASET | ZFS_TYPE_BOOKMARK, "NAME", B_TRUE, sfeatures); zprop_register_hidden(ZFS_PROP_ISCSIOPTIONS, "iscsioptions", PROP_TYPE_STRING, PROP_INHERIT, ZFS_TYPE_VOLUME, "ISCSIOPTIONS", B_TRUE, sfeatures); zprop_register_hidden(ZFS_PROP_STMF_SHAREINFO, "stmf_sbd_lu", PROP_TYPE_STRING, PROP_INHERIT, ZFS_TYPE_VOLUME, "STMF_SBD_LU", B_TRUE, sfeatures); zprop_register_hidden(ZFS_PROP_USERACCOUNTING, "useraccounting", PROP_TYPE_NUMBER, PROP_READONLY, ZFS_TYPE_DATASET, "USERACCOUNTING", B_FALSE, sfeatures); zprop_register_hidden(ZFS_PROP_UNIQUE, "unique", PROP_TYPE_NUMBER, PROP_READONLY, ZFS_TYPE_DATASET, "UNIQUE", B_FALSE, sfeatures); zprop_register_hidden(ZFS_PROP_INCONSISTENT, "inconsistent", PROP_TYPE_NUMBER, PROP_READONLY, ZFS_TYPE_DATASET, "INCONSISTENT", B_FALSE, sfeatures); zprop_register_hidden(ZFS_PROP_IVSET_GUID, "ivsetguid", PROP_TYPE_NUMBER, PROP_READONLY, ZFS_TYPE_DATASET | ZFS_TYPE_BOOKMARK, "IVSETGUID", B_TRUE, sfeatures); zprop_register_hidden(ZFS_PROP_PREV_SNAP, "prevsnap", PROP_TYPE_STRING, PROP_READONLY, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "PREVSNAP", B_TRUE, sfeatures); zprop_register_hidden(ZFS_PROP_PBKDF2_SALT, "pbkdf2salt", PROP_TYPE_NUMBER, PROP_ONETIME_DEFAULT, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "PBKDF2SALT", B_FALSE, sfeatures); zprop_register_hidden(ZFS_PROP_KEY_GUID, "keyguid", PROP_TYPE_NUMBER, PROP_READONLY, ZFS_TYPE_DATASET, "KEYGUID", B_TRUE, sfeatures); zprop_register_hidden(ZFS_PROP_REDACTED, "redacted", PROP_TYPE_NUMBER, PROP_READONLY, ZFS_TYPE_DATASET, "REDACTED", B_FALSE, sfeatures); /* * Properties that are obsolete and not used. These are retained so * that we don't have to change the values of the zfs_prop_t enum, or * have NULL pointers in the zfs_prop_table[]. */ zprop_register_hidden(ZFS_PROP_REMAPTXG, "remaptxg", PROP_TYPE_NUMBER, PROP_READONLY, ZFS_TYPE_DATASET, "REMAPTXG", B_FALSE, sfeatures); /* oddball properties */ /* 'creation' is a number but displayed as human-readable => flex */ zprop_register_impl(ZFS_PROP_CREATION, "creation", PROP_TYPE_NUMBER, 0, NULL, PROP_READONLY, ZFS_TYPE_DATASET | ZFS_TYPE_BOOKMARK, "", "CREATION", B_FALSE, B_TRUE, B_TRUE, NULL, sfeatures); zprop_register_impl(ZFS_PROP_SNAPSHOTS_CHANGED, "snapshots_changed", PROP_TYPE_NUMBER, 0, NULL, PROP_READONLY, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME, "", "SNAPSHOTS_CHANGED", B_FALSE, B_TRUE, B_TRUE, NULL, sfeatures); zfs_mod_list_supported_free(sfeatures); } boolean_t zfs_prop_delegatable(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); zprop_desc_t *pd = &zfs_prop_table[prop]; /* The mlslabel property is never delegatable. */ if (prop == ZFS_PROP_MLSLABEL) return (B_FALSE); return (pd->pd_attr != PROP_READONLY); } /* * Given a zfs dataset property name, returns the corresponding property ID. */ zfs_prop_t zfs_name_to_prop(const char *propname) { return (zprop_name_to_prop(propname, ZFS_TYPE_DATASET)); } /* * Returns true if this is a valid user-defined property (one with a ':'). */ boolean_t zfs_prop_user(const char *name) { int i; char c; boolean_t foundsep = B_FALSE; for (i = 0; i < strlen(name); i++) { c = name[i]; if (!zprop_valid_char(c)) return (B_FALSE); if (c == ':') foundsep = B_TRUE; } if (!foundsep) return (B_FALSE); return (B_TRUE); } /* * Returns true if this is a valid userspace-type property (one with a '@'). * Note that after the @, any character is valid (eg, another @, for SID * user@domain). */ boolean_t zfs_prop_userquota(const char *name) { zfs_userquota_prop_t prop; for (prop = 0; prop < ZFS_NUM_USERQUOTA_PROPS; prop++) { if (strncmp(name, zfs_userquota_prop_prefixes[prop], strlen(zfs_userquota_prop_prefixes[prop])) == 0) { return (B_TRUE); } } return (B_FALSE); } /* * Returns true if this is a valid written@ property. * Note that after the @, any character is valid (eg, another @, for * written@pool/fs@origin). */ boolean_t zfs_prop_written(const char *name) { static const char *prop_prefix = "written@"; static const char *book_prefix = "written#"; return (strncmp(name, prop_prefix, strlen(prop_prefix)) == 0 || strncmp(name, book_prefix, strlen(book_prefix)) == 0); } /* * Tables of index types, plus functions to convert between the user view * (strings) and internal representation (uint64_t). */ int zfs_prop_string_to_index(zfs_prop_t prop, const char *string, uint64_t *index) { return (zprop_string_to_index(prop, string, index, ZFS_TYPE_DATASET)); } int zfs_prop_index_to_string(zfs_prop_t prop, uint64_t index, const char **string) { return (zprop_index_to_string(prop, index, string, ZFS_TYPE_DATASET)); } uint64_t zfs_prop_random_value(zfs_prop_t prop, uint64_t seed) { return (zprop_random_value(prop, seed, ZFS_TYPE_DATASET)); } /* * Returns TRUE if the property applies to any of the given dataset types. */ boolean_t zfs_prop_valid_for_type(int prop, zfs_type_t types, boolean_t headcheck) { return (zprop_valid_for_type(prop, types, headcheck)); } zprop_type_t zfs_prop_get_type(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); return (zfs_prop_table[prop].pd_proptype); } /* * Returns TRUE if the property is readonly. */ boolean_t zfs_prop_readonly(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); return (zfs_prop_table[prop].pd_attr == PROP_READONLY || zfs_prop_table[prop].pd_attr == PROP_ONETIME || zfs_prop_table[prop].pd_attr == PROP_ONETIME_DEFAULT); } /* * Returns TRUE if the property is visible (not hidden). */ boolean_t zfs_prop_visible(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); return (zfs_prop_table[prop].pd_visible && zfs_prop_table[prop].pd_zfs_mod_supported); } /* * Returns TRUE if the property is only allowed to be set once. */ boolean_t zfs_prop_setonce(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); return (zfs_prop_table[prop].pd_attr == PROP_ONETIME || zfs_prop_table[prop].pd_attr == PROP_ONETIME_DEFAULT); } const char * zfs_prop_default_string(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); return (zfs_prop_table[prop].pd_strdefault); } uint64_t zfs_prop_default_numeric(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); return (zfs_prop_table[prop].pd_numdefault); } /* * Given a dataset property ID, returns the corresponding name. * Assuming the zfs dataset property ID is valid. */ const char * zfs_prop_to_name(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); return (zfs_prop_table[prop].pd_name); } /* * Returns TRUE if the property is inheritable. */ boolean_t zfs_prop_inheritable(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); return (zfs_prop_table[prop].pd_attr == PROP_INHERIT || zfs_prop_table[prop].pd_attr == PROP_ONETIME); } /* * Returns TRUE if property is one of the encryption properties that requires * a loaded encryption key to modify. */ boolean_t zfs_prop_encryption_key_param(zfs_prop_t prop) { /* * keylocation does not count as an encryption property. It can be * changed at will without needing the master keys. */ return (prop == ZFS_PROP_PBKDF2_SALT || prop == ZFS_PROP_PBKDF2_ITERS || prop == ZFS_PROP_KEYFORMAT); } /* * Helper function used by both kernelspace and userspace to check the * keylocation property. If encrypted is set, the keylocation must be valid * for an encrypted dataset. */ boolean_t zfs_prop_valid_keylocation(const char *str, boolean_t encrypted) { if (strcmp("none", str) == 0) return (!encrypted); else if (strcmp("prompt", str) == 0) return (B_TRUE); else if (strlen(str) > 8 && strncmp("file:///", str, 8) == 0) return (B_TRUE); else if (strlen(str) > 8 && strncmp("https://", str, 8) == 0) return (B_TRUE); else if (strlen(str) > 7 && strncmp("http://", str, 7) == 0) return (B_TRUE); return (B_FALSE); } #ifndef _KERNEL #include /* * Returns a string describing the set of acceptable values for the given * zfs property, or NULL if it cannot be set. */ const char * zfs_prop_values(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); return (zfs_prop_table[prop].pd_values); } /* * Returns TRUE if this property is a string type. Note that index types * (compression, checksum) are treated as strings in userland, even though they * are stored numerically on disk. */ int zfs_prop_is_string(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); return (zfs_prop_table[prop].pd_proptype == PROP_TYPE_STRING || zfs_prop_table[prop].pd_proptype == PROP_TYPE_INDEX); } /* * Returns the column header for the given property. Used only in * 'zfs list -o', but centralized here with the other property information. */ const char * zfs_prop_column_name(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); return (zfs_prop_table[prop].pd_colname); } /* * Returns whether the given property should be displayed right-justified for * 'zfs list'. */ boolean_t zfs_prop_align_right(zfs_prop_t prop) { ASSERT3S(prop, >=, 0); ASSERT3S(prop, <, ZFS_NUM_PROPS); return (zfs_prop_table[prop].pd_rightalign); } #endif #if defined(_KERNEL) #if defined(HAVE_KERNEL_FPU_INTERNAL) uint8_t **zfs_kfpu_fpregs; EXPORT_SYMBOL(zfs_kfpu_fpregs); #endif /* defined(HAVE_KERNEL_FPU_INTERNAL) */ extern int __init zcommon_init(void); extern void zcommon_fini(void); int __init zcommon_init(void) { int error = kfpu_init(); if (error) return (error); fletcher_4_init(); return (0); } void zcommon_fini(void) { fletcher_4_fini(); kfpu_fini(); } #ifdef __FreeBSD__ module_init_early(zcommon_init); module_exit(zcommon_fini); #endif #endif /* zfs dataset property functions */ EXPORT_SYMBOL(zfs_userquota_prop_prefixes); EXPORT_SYMBOL(zfs_prop_init); EXPORT_SYMBOL(zfs_prop_get_type); EXPORT_SYMBOL(zfs_prop_get_table); EXPORT_SYMBOL(zfs_prop_delegatable); EXPORT_SYMBOL(zfs_prop_visible); /* Dataset property functions shared between libzfs and kernel. */ EXPORT_SYMBOL(zfs_prop_default_string); EXPORT_SYMBOL(zfs_prop_default_numeric); EXPORT_SYMBOL(zfs_prop_readonly); EXPORT_SYMBOL(zfs_prop_inheritable); EXPORT_SYMBOL(zfs_prop_encryption_key_param); EXPORT_SYMBOL(zfs_prop_valid_keylocation); EXPORT_SYMBOL(zfs_prop_setonce); EXPORT_SYMBOL(zfs_prop_to_name); EXPORT_SYMBOL(zfs_name_to_prop); EXPORT_SYMBOL(zfs_prop_user); EXPORT_SYMBOL(zfs_prop_userquota); EXPORT_SYMBOL(zfs_prop_index_to_string); EXPORT_SYMBOL(zfs_prop_string_to_index); EXPORT_SYMBOL(zfs_prop_valid_for_type); EXPORT_SYMBOL(zfs_prop_written); diff --git a/module/zfs/dmu_objset.c b/module/zfs/dmu_objset.c index f8bd7422a5df..ef966d6703c9 100644 --- a/module/zfs/dmu_objset.c +++ b/module/zfs/dmu_objset.c @@ -1,3083 +1,3102 @@ /* * 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, 2020 by Delphix. All rights reserved. * Copyright (c) 2013 by Saso Kiselkov. All rights reserved. * Copyright (c) 2013, Joyent, Inc. All rights reserved. * Copyright (c) 2014 Spectra Logic Corporation, All rights reserved. * Copyright (c) 2015, STRATO AG, Inc. All rights reserved. * Copyright (c) 2016 Actifio, Inc. All rights reserved. * Copyright 2017 Nexenta Systems, Inc. * Copyright (c) 2017 Open-E, Inc. All Rights Reserved. * Copyright (c) 2018, loli10K . All rights reserved. * Copyright (c) 2019, Klara Inc. * Copyright (c) 2019, Allan Jude * Copyright (c) 2022 Hewlett Packard Enterprise Development LP. */ /* Portions Copyright 2010 Robert Milkowski */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include "zfs_namecheck.h" #include #include /* * Needed to close a window in dnode_move() that allows the objset to be freed * before it can be safely accessed. */ krwlock_t os_lock; /* * Tunable to overwrite the maximum number of threads for the parallelization * of dmu_objset_find_dp, needed to speed up the import of pools with many * datasets. * Default is 4 times the number of leaf vdevs. */ static const int dmu_find_threads = 0; /* * Backfill lower metadnode objects after this many have been freed. * Backfilling negatively impacts object creation rates, so only do it * if there are enough holes to fill. */ static const int dmu_rescan_dnode_threshold = 1 << DN_MAX_INDBLKSHIFT; static const char *upgrade_tag = "upgrade_tag"; static void dmu_objset_find_dp_cb(void *arg); static void dmu_objset_upgrade(objset_t *os, dmu_objset_upgrade_cb_t cb); static void dmu_objset_upgrade_stop(objset_t *os); void dmu_objset_init(void) { rw_init(&os_lock, NULL, RW_DEFAULT, NULL); } void dmu_objset_fini(void) { rw_destroy(&os_lock); } spa_t * dmu_objset_spa(objset_t *os) { return (os->os_spa); } zilog_t * dmu_objset_zil(objset_t *os) { return (os->os_zil); } dsl_pool_t * dmu_objset_pool(objset_t *os) { dsl_dataset_t *ds; if ((ds = os->os_dsl_dataset) != NULL && ds->ds_dir) return (ds->ds_dir->dd_pool); else return (spa_get_dsl(os->os_spa)); } dsl_dataset_t * dmu_objset_ds(objset_t *os) { return (os->os_dsl_dataset); } dmu_objset_type_t dmu_objset_type(objset_t *os) { return (os->os_phys->os_type); } void dmu_objset_name(objset_t *os, char *buf) { dsl_dataset_name(os->os_dsl_dataset, buf); } uint64_t dmu_objset_id(objset_t *os) { dsl_dataset_t *ds = os->os_dsl_dataset; return (ds ? ds->ds_object : 0); } uint64_t dmu_objset_dnodesize(objset_t *os) { return (os->os_dnodesize); } zfs_sync_type_t dmu_objset_syncprop(objset_t *os) { return (os->os_sync); } zfs_logbias_op_t dmu_objset_logbias(objset_t *os) { return (os->os_logbias); } static void checksum_changed_cb(void *arg, uint64_t newval) { objset_t *os = arg; /* * Inheritance should have been done by now. */ ASSERT(newval != ZIO_CHECKSUM_INHERIT); os->os_checksum = zio_checksum_select(newval, ZIO_CHECKSUM_ON_VALUE); } static void compression_changed_cb(void *arg, uint64_t newval) { objset_t *os = arg; /* * Inheritance and range checking should have been done by now. */ ASSERT(newval != ZIO_COMPRESS_INHERIT); os->os_compress = zio_compress_select(os->os_spa, ZIO_COMPRESS_ALGO(newval), ZIO_COMPRESS_ON); os->os_complevel = zio_complevel_select(os->os_spa, os->os_compress, ZIO_COMPRESS_LEVEL(newval), ZIO_COMPLEVEL_DEFAULT); } static void copies_changed_cb(void *arg, uint64_t newval) { objset_t *os = arg; /* * Inheritance and range checking should have been done by now. */ ASSERT(newval > 0); ASSERT(newval <= spa_max_replication(os->os_spa)); os->os_copies = newval; } static void dedup_changed_cb(void *arg, uint64_t newval) { objset_t *os = arg; spa_t *spa = os->os_spa; enum zio_checksum checksum; /* * Inheritance should have been done by now. */ ASSERT(newval != ZIO_CHECKSUM_INHERIT); checksum = zio_checksum_dedup_select(spa, newval, ZIO_CHECKSUM_OFF); os->os_dedup_checksum = checksum & ZIO_CHECKSUM_MASK; os->os_dedup_verify = !!(checksum & ZIO_CHECKSUM_VERIFY); } static void primary_cache_changed_cb(void *arg, uint64_t newval) { objset_t *os = arg; /* * Inheritance and range checking should have been done by now. */ ASSERT(newval == ZFS_CACHE_ALL || newval == ZFS_CACHE_NONE || newval == ZFS_CACHE_METADATA); os->os_primary_cache = newval; } static void secondary_cache_changed_cb(void *arg, uint64_t newval) { objset_t *os = arg; /* * Inheritance and range checking should have been done by now. */ ASSERT(newval == ZFS_CACHE_ALL || newval == ZFS_CACHE_NONE || newval == ZFS_CACHE_METADATA); os->os_secondary_cache = newval; } +static void +prefetch_changed_cb(void *arg, uint64_t newval) +{ + objset_t *os = arg; + + /* + * Inheritance should have been done by now. + */ + ASSERT(newval == ZFS_PREFETCH_ALL || newval == ZFS_PREFETCH_NONE || + newval == ZFS_PREFETCH_METADATA); + os->os_prefetch = newval; +} + static void sync_changed_cb(void *arg, uint64_t newval) { objset_t *os = arg; /* * Inheritance and range checking should have been done by now. */ ASSERT(newval == ZFS_SYNC_STANDARD || newval == ZFS_SYNC_ALWAYS || newval == ZFS_SYNC_DISABLED); os->os_sync = newval; if (os->os_zil) zil_set_sync(os->os_zil, newval); } static void redundant_metadata_changed_cb(void *arg, uint64_t newval) { objset_t *os = arg; /* * Inheritance and range checking should have been done by now. */ ASSERT(newval == ZFS_REDUNDANT_METADATA_ALL || newval == ZFS_REDUNDANT_METADATA_MOST || newval == ZFS_REDUNDANT_METADATA_SOME || newval == ZFS_REDUNDANT_METADATA_NONE); os->os_redundant_metadata = newval; } static void dnodesize_changed_cb(void *arg, uint64_t newval) { objset_t *os = arg; switch (newval) { case ZFS_DNSIZE_LEGACY: os->os_dnodesize = DNODE_MIN_SIZE; break; case ZFS_DNSIZE_AUTO: /* * Choose a dnode size that will work well for most * workloads if the user specified "auto". Future code * improvements could dynamically select a dnode size * based on observed workload patterns. */ os->os_dnodesize = DNODE_MIN_SIZE * 2; break; case ZFS_DNSIZE_1K: case ZFS_DNSIZE_2K: case ZFS_DNSIZE_4K: case ZFS_DNSIZE_8K: case ZFS_DNSIZE_16K: os->os_dnodesize = newval; break; } } static void smallblk_changed_cb(void *arg, uint64_t newval) { objset_t *os = arg; /* * Inheritance and range checking should have been done by now. */ ASSERT(newval <= SPA_MAXBLOCKSIZE); ASSERT(ISP2(newval)); os->os_zpl_special_smallblock = newval; } static void logbias_changed_cb(void *arg, uint64_t newval) { objset_t *os = arg; ASSERT(newval == ZFS_LOGBIAS_LATENCY || newval == ZFS_LOGBIAS_THROUGHPUT); os->os_logbias = newval; if (os->os_zil) zil_set_logbias(os->os_zil, newval); } static void recordsize_changed_cb(void *arg, uint64_t newval) { objset_t *os = arg; os->os_recordsize = newval; } void dmu_objset_byteswap(void *buf, size_t size) { objset_phys_t *osp = buf; ASSERT(size == OBJSET_PHYS_SIZE_V1 || size == OBJSET_PHYS_SIZE_V2 || size == sizeof (objset_phys_t)); dnode_byteswap(&osp->os_meta_dnode); byteswap_uint64_array(&osp->os_zil_header, sizeof (zil_header_t)); osp->os_type = BSWAP_64(osp->os_type); osp->os_flags = BSWAP_64(osp->os_flags); if (size >= OBJSET_PHYS_SIZE_V2) { dnode_byteswap(&osp->os_userused_dnode); dnode_byteswap(&osp->os_groupused_dnode); if (size >= sizeof (objset_phys_t)) dnode_byteswap(&osp->os_projectused_dnode); } } /* * The hash is a CRC-based hash of the objset_t pointer and the object number. */ static uint64_t dnode_hash(const objset_t *os, uint64_t obj) { uintptr_t osv = (uintptr_t)os; uint64_t crc = -1ULL; ASSERT(zfs_crc64_table[128] == ZFS_CRC64_POLY); /* * The low 6 bits of the pointer don't have much entropy, because * the objset_t is larger than 2^6 bytes long. */ crc = (crc >> 8) ^ zfs_crc64_table[(crc ^ (osv >> 6)) & 0xFF]; crc = (crc >> 8) ^ zfs_crc64_table[(crc ^ (obj >> 0)) & 0xFF]; crc = (crc >> 8) ^ zfs_crc64_table[(crc ^ (obj >> 8)) & 0xFF]; crc = (crc >> 8) ^ zfs_crc64_table[(crc ^ (obj >> 16)) & 0xFF]; crc ^= (osv>>14) ^ (obj>>24); return (crc); } static unsigned int dnode_multilist_index_func(multilist_t *ml, void *obj) { dnode_t *dn = obj; /* * The low order bits of the hash value are thought to be * distributed evenly. Otherwise, in the case that the multilist * has a power of two number of sublists, each sublists' usage * would not be evenly distributed. In this context full 64bit * division would be a waste of time, so limit it to 32 bits. */ return ((unsigned int)dnode_hash(dn->dn_objset, dn->dn_object) % multilist_get_num_sublists(ml)); } static inline boolean_t dmu_os_is_l2cacheable(objset_t *os) { if (os->os_secondary_cache == ZFS_CACHE_ALL || os->os_secondary_cache == ZFS_CACHE_METADATA) { if (l2arc_exclude_special == 0) return (B_TRUE); blkptr_t *bp = os->os_rootbp; if (bp == NULL || BP_IS_HOLE(bp)) return (B_FALSE); uint64_t vdev = DVA_GET_VDEV(bp->blk_dva); vdev_t *rvd = os->os_spa->spa_root_vdev; vdev_t *vd = NULL; if (vdev < rvd->vdev_children) vd = rvd->vdev_child[vdev]; if (vd == NULL) return (B_TRUE); if (vd->vdev_alloc_bias != VDEV_BIAS_SPECIAL && vd->vdev_alloc_bias != VDEV_BIAS_DEDUP) return (B_TRUE); } return (B_FALSE); } /* * Instantiates the objset_t in-memory structure corresponding to the * objset_phys_t that's pointed to by the specified blkptr_t. */ int dmu_objset_open_impl(spa_t *spa, dsl_dataset_t *ds, blkptr_t *bp, objset_t **osp) { objset_t *os; int i, err; ASSERT(ds == NULL || MUTEX_HELD(&ds->ds_opening_lock)); ASSERT(!BP_IS_REDACTED(bp)); /* * We need the pool config lock to get properties. */ ASSERT(ds == NULL || dsl_pool_config_held(ds->ds_dir->dd_pool)); /* * The $ORIGIN dataset (if it exists) doesn't have an associated * objset, so there's no reason to open it. The $ORIGIN dataset * will not exist on pools older than SPA_VERSION_ORIGIN. */ if (ds != NULL && spa_get_dsl(spa) != NULL && spa_get_dsl(spa)->dp_origin_snap != NULL) { ASSERT3P(ds->ds_dir, !=, spa_get_dsl(spa)->dp_origin_snap->ds_dir); } os = kmem_zalloc(sizeof (objset_t), KM_SLEEP); os->os_dsl_dataset = ds; os->os_spa = spa; os->os_rootbp = bp; if (!BP_IS_HOLE(os->os_rootbp)) { arc_flags_t aflags = ARC_FLAG_WAIT; zbookmark_phys_t zb; int size; zio_flag_t zio_flags = ZIO_FLAG_CANFAIL; SET_BOOKMARK(&zb, ds ? ds->ds_object : DMU_META_OBJSET, ZB_ROOT_OBJECT, ZB_ROOT_LEVEL, ZB_ROOT_BLKID); if (dmu_os_is_l2cacheable(os)) aflags |= ARC_FLAG_L2CACHE; if (ds != NULL && ds->ds_dir->dd_crypto_obj != 0) { ASSERT3U(BP_GET_COMPRESS(bp), ==, ZIO_COMPRESS_OFF); ASSERT(BP_IS_AUTHENTICATED(bp)); zio_flags |= ZIO_FLAG_RAW; } dprintf_bp(os->os_rootbp, "reading %s", ""); err = arc_read(NULL, spa, os->os_rootbp, arc_getbuf_func, &os->os_phys_buf, ZIO_PRIORITY_SYNC_READ, zio_flags, &aflags, &zb); if (err != 0) { kmem_free(os, sizeof (objset_t)); /* convert checksum errors into IO errors */ if (err == ECKSUM) err = SET_ERROR(EIO); return (err); } if (spa_version(spa) < SPA_VERSION_USERSPACE) size = OBJSET_PHYS_SIZE_V1; else if (!spa_feature_is_enabled(spa, SPA_FEATURE_PROJECT_QUOTA)) size = OBJSET_PHYS_SIZE_V2; else size = sizeof (objset_phys_t); /* Increase the blocksize if we are permitted. */ if (arc_buf_size(os->os_phys_buf) < size) { arc_buf_t *buf = arc_alloc_buf(spa, &os->os_phys_buf, ARC_BUFC_METADATA, size); memset(buf->b_data, 0, size); memcpy(buf->b_data, os->os_phys_buf->b_data, arc_buf_size(os->os_phys_buf)); arc_buf_destroy(os->os_phys_buf, &os->os_phys_buf); os->os_phys_buf = buf; } os->os_phys = os->os_phys_buf->b_data; os->os_flags = os->os_phys->os_flags; } else { int size = spa_version(spa) >= SPA_VERSION_USERSPACE ? sizeof (objset_phys_t) : OBJSET_PHYS_SIZE_V1; os->os_phys_buf = arc_alloc_buf(spa, &os->os_phys_buf, ARC_BUFC_METADATA, size); os->os_phys = os->os_phys_buf->b_data; memset(os->os_phys, 0, size); } /* * These properties will be filled in by the logic in zfs_get_zplprop() * when they are queried for the first time. */ os->os_version = OBJSET_PROP_UNINITIALIZED; os->os_normalization = OBJSET_PROP_UNINITIALIZED; os->os_utf8only = OBJSET_PROP_UNINITIALIZED; os->os_casesensitivity = OBJSET_PROP_UNINITIALIZED; /* * Note: the changed_cb will be called once before the register * func returns, thus changing the checksum/compression from the * default (fletcher2/off). Snapshots don't need to know about * checksum/compression/copies. */ if (ds != NULL) { os->os_encrypted = (ds->ds_dir->dd_crypto_obj != 0); err = dsl_prop_register(ds, zfs_prop_to_name(ZFS_PROP_PRIMARYCACHE), primary_cache_changed_cb, os); if (err == 0) { err = dsl_prop_register(ds, zfs_prop_to_name(ZFS_PROP_SECONDARYCACHE), secondary_cache_changed_cb, os); } + if (err == 0) { + err = dsl_prop_register(ds, + zfs_prop_to_name(ZFS_PROP_PREFETCH), + prefetch_changed_cb, os); + } if (!ds->ds_is_snapshot) { if (err == 0) { err = dsl_prop_register(ds, zfs_prop_to_name(ZFS_PROP_CHECKSUM), checksum_changed_cb, os); } if (err == 0) { err = dsl_prop_register(ds, zfs_prop_to_name(ZFS_PROP_COMPRESSION), compression_changed_cb, os); } if (err == 0) { err = dsl_prop_register(ds, zfs_prop_to_name(ZFS_PROP_COPIES), copies_changed_cb, os); } if (err == 0) { err = dsl_prop_register(ds, zfs_prop_to_name(ZFS_PROP_DEDUP), dedup_changed_cb, os); } if (err == 0) { err = dsl_prop_register(ds, zfs_prop_to_name(ZFS_PROP_LOGBIAS), logbias_changed_cb, os); } if (err == 0) { err = dsl_prop_register(ds, zfs_prop_to_name(ZFS_PROP_SYNC), sync_changed_cb, os); } if (err == 0) { err = dsl_prop_register(ds, zfs_prop_to_name( ZFS_PROP_REDUNDANT_METADATA), redundant_metadata_changed_cb, os); } if (err == 0) { err = dsl_prop_register(ds, zfs_prop_to_name(ZFS_PROP_RECORDSIZE), recordsize_changed_cb, os); } if (err == 0) { err = dsl_prop_register(ds, zfs_prop_to_name(ZFS_PROP_DNODESIZE), dnodesize_changed_cb, os); } if (err == 0) { err = dsl_prop_register(ds, zfs_prop_to_name( ZFS_PROP_SPECIAL_SMALL_BLOCKS), smallblk_changed_cb, os); } } if (err != 0) { arc_buf_destroy(os->os_phys_buf, &os->os_phys_buf); kmem_free(os, sizeof (objset_t)); return (err); } } else { /* It's the meta-objset. */ os->os_checksum = ZIO_CHECKSUM_FLETCHER_4; os->os_compress = ZIO_COMPRESS_ON; os->os_complevel = ZIO_COMPLEVEL_DEFAULT; os->os_encrypted = B_FALSE; os->os_copies = spa_max_replication(spa); os->os_dedup_checksum = ZIO_CHECKSUM_OFF; os->os_dedup_verify = B_FALSE; os->os_logbias = ZFS_LOGBIAS_LATENCY; os->os_sync = ZFS_SYNC_STANDARD; os->os_primary_cache = ZFS_CACHE_ALL; os->os_secondary_cache = ZFS_CACHE_ALL; os->os_dnodesize = DNODE_MIN_SIZE; + os->os_prefetch = ZFS_PREFETCH_ALL; } if (ds == NULL || !ds->ds_is_snapshot) os->os_zil_header = os->os_phys->os_zil_header; os->os_zil = zil_alloc(os, &os->os_zil_header); for (i = 0; i < TXG_SIZE; i++) { multilist_create(&os->os_dirty_dnodes[i], sizeof (dnode_t), offsetof(dnode_t, dn_dirty_link[i]), dnode_multilist_index_func); } list_create(&os->os_dnodes, sizeof (dnode_t), offsetof(dnode_t, dn_link)); list_create(&os->os_downgraded_dbufs, sizeof (dmu_buf_impl_t), offsetof(dmu_buf_impl_t, db_link)); list_link_init(&os->os_evicting_node); mutex_init(&os->os_lock, NULL, MUTEX_DEFAULT, NULL); mutex_init(&os->os_userused_lock, NULL, MUTEX_DEFAULT, NULL); mutex_init(&os->os_obj_lock, NULL, MUTEX_DEFAULT, NULL); mutex_init(&os->os_user_ptr_lock, NULL, MUTEX_DEFAULT, NULL); os->os_obj_next_percpu_len = boot_ncpus; os->os_obj_next_percpu = kmem_zalloc(os->os_obj_next_percpu_len * sizeof (os->os_obj_next_percpu[0]), KM_SLEEP); dnode_special_open(os, &os->os_phys->os_meta_dnode, DMU_META_DNODE_OBJECT, &os->os_meta_dnode); if (OBJSET_BUF_HAS_USERUSED(os->os_phys_buf)) { dnode_special_open(os, &os->os_phys->os_userused_dnode, DMU_USERUSED_OBJECT, &os->os_userused_dnode); dnode_special_open(os, &os->os_phys->os_groupused_dnode, DMU_GROUPUSED_OBJECT, &os->os_groupused_dnode); if (OBJSET_BUF_HAS_PROJECTUSED(os->os_phys_buf)) dnode_special_open(os, &os->os_phys->os_projectused_dnode, DMU_PROJECTUSED_OBJECT, &os->os_projectused_dnode); } mutex_init(&os->os_upgrade_lock, NULL, MUTEX_DEFAULT, NULL); *osp = os; return (0); } int dmu_objset_from_ds(dsl_dataset_t *ds, objset_t **osp) { int err = 0; /* * We need the pool_config lock to manipulate the dsl_dataset_t. * Even if the dataset is long-held, we need the pool_config lock * to open the objset, as it needs to get properties. */ ASSERT(dsl_pool_config_held(ds->ds_dir->dd_pool)); mutex_enter(&ds->ds_opening_lock); if (ds->ds_objset == NULL) { objset_t *os; rrw_enter(&ds->ds_bp_rwlock, RW_READER, FTAG); err = dmu_objset_open_impl(dsl_dataset_get_spa(ds), ds, dsl_dataset_get_blkptr(ds), &os); rrw_exit(&ds->ds_bp_rwlock, FTAG); if (err == 0) { mutex_enter(&ds->ds_lock); ASSERT(ds->ds_objset == NULL); ds->ds_objset = os; mutex_exit(&ds->ds_lock); } } *osp = ds->ds_objset; mutex_exit(&ds->ds_opening_lock); return (err); } /* * Holds the pool while the objset is held. Therefore only one objset * can be held at a time. */ int dmu_objset_hold_flags(const char *name, boolean_t decrypt, const void *tag, objset_t **osp) { dsl_pool_t *dp; dsl_dataset_t *ds; int err; ds_hold_flags_t flags; flags = (decrypt) ? DS_HOLD_FLAG_DECRYPT : DS_HOLD_FLAG_NONE; err = dsl_pool_hold(name, tag, &dp); if (err != 0) return (err); err = dsl_dataset_hold_flags(dp, name, flags, tag, &ds); if (err != 0) { dsl_pool_rele(dp, tag); return (err); } err = dmu_objset_from_ds(ds, osp); if (err != 0) { dsl_dataset_rele(ds, tag); dsl_pool_rele(dp, tag); } return (err); } int dmu_objset_hold(const char *name, const void *tag, objset_t **osp) { return (dmu_objset_hold_flags(name, B_FALSE, tag, osp)); } static int dmu_objset_own_impl(dsl_dataset_t *ds, dmu_objset_type_t type, boolean_t readonly, boolean_t decrypt, const void *tag, objset_t **osp) { (void) tag; int err = dmu_objset_from_ds(ds, osp); if (err != 0) { return (err); } else if (type != DMU_OST_ANY && type != (*osp)->os_phys->os_type) { return (SET_ERROR(EINVAL)); } else if (!readonly && dsl_dataset_is_snapshot(ds)) { return (SET_ERROR(EROFS)); } else if (!readonly && decrypt && dsl_dir_incompatible_encryption_version(ds->ds_dir)) { return (SET_ERROR(EROFS)); } /* if we are decrypting, we can now check MACs in os->os_phys_buf */ if (decrypt && arc_is_unauthenticated((*osp)->os_phys_buf)) { zbookmark_phys_t zb; SET_BOOKMARK(&zb, ds->ds_object, ZB_ROOT_OBJECT, ZB_ROOT_LEVEL, ZB_ROOT_BLKID); err = arc_untransform((*osp)->os_phys_buf, (*osp)->os_spa, &zb, B_FALSE); if (err != 0) return (err); ASSERT0(arc_is_unauthenticated((*osp)->os_phys_buf)); } return (0); } /* * dsl_pool must not be held when this is called. * Upon successful return, there will be a longhold on the dataset, * and the dsl_pool will not be held. */ int dmu_objset_own(const char *name, dmu_objset_type_t type, boolean_t readonly, boolean_t decrypt, const void *tag, objset_t **osp) { dsl_pool_t *dp; dsl_dataset_t *ds; int err; ds_hold_flags_t flags; flags = (decrypt) ? DS_HOLD_FLAG_DECRYPT : DS_HOLD_FLAG_NONE; err = dsl_pool_hold(name, FTAG, &dp); if (err != 0) return (err); err = dsl_dataset_own(dp, name, flags, tag, &ds); if (err != 0) { dsl_pool_rele(dp, FTAG); return (err); } err = dmu_objset_own_impl(ds, type, readonly, decrypt, tag, osp); if (err != 0) { dsl_dataset_disown(ds, flags, tag); dsl_pool_rele(dp, FTAG); return (err); } /* * User accounting requires the dataset to be decrypted and rw. * We also don't begin user accounting during claiming to help * speed up pool import times and to keep this txg reserved * completely for recovery work. */ if (!readonly && !dp->dp_spa->spa_claiming && (ds->ds_dir->dd_crypto_obj == 0 || decrypt)) { if (dmu_objset_userobjspace_upgradable(*osp) || dmu_objset_projectquota_upgradable(*osp)) { dmu_objset_id_quota_upgrade(*osp); } else if (dmu_objset_userused_enabled(*osp)) { dmu_objset_userspace_upgrade(*osp); } } dsl_pool_rele(dp, FTAG); return (0); } int dmu_objset_own_obj(dsl_pool_t *dp, uint64_t obj, dmu_objset_type_t type, boolean_t readonly, boolean_t decrypt, const void *tag, objset_t **osp) { dsl_dataset_t *ds; int err; ds_hold_flags_t flags; flags = (decrypt) ? DS_HOLD_FLAG_DECRYPT : DS_HOLD_FLAG_NONE; err = dsl_dataset_own_obj(dp, obj, flags, tag, &ds); if (err != 0) return (err); err = dmu_objset_own_impl(ds, type, readonly, decrypt, tag, osp); if (err != 0) { dsl_dataset_disown(ds, flags, tag); return (err); } return (0); } void dmu_objset_rele_flags(objset_t *os, boolean_t decrypt, const void *tag) { ds_hold_flags_t flags; dsl_pool_t *dp = dmu_objset_pool(os); flags = (decrypt) ? DS_HOLD_FLAG_DECRYPT : DS_HOLD_FLAG_NONE; dsl_dataset_rele_flags(os->os_dsl_dataset, flags, tag); dsl_pool_rele(dp, tag); } void dmu_objset_rele(objset_t *os, const void *tag) { dmu_objset_rele_flags(os, B_FALSE, tag); } /* * When we are called, os MUST refer to an objset associated with a dataset * that is owned by 'tag'; that is, is held and long held by 'tag' and ds_owner * == tag. We will then release and reacquire ownership of the dataset while * holding the pool config_rwlock to avoid intervening namespace or ownership * changes may occur. * * This exists solely to accommodate zfs_ioc_userspace_upgrade()'s desire to * release the hold on its dataset and acquire a new one on the dataset of the * same name so that it can be partially torn down and reconstructed. */ void dmu_objset_refresh_ownership(dsl_dataset_t *ds, dsl_dataset_t **newds, boolean_t decrypt, const void *tag) { dsl_pool_t *dp; char name[ZFS_MAX_DATASET_NAME_LEN]; ds_hold_flags_t flags; flags = (decrypt) ? DS_HOLD_FLAG_DECRYPT : DS_HOLD_FLAG_NONE; VERIFY3P(ds, !=, NULL); VERIFY3P(ds->ds_owner, ==, tag); VERIFY(dsl_dataset_long_held(ds)); dsl_dataset_name(ds, name); dp = ds->ds_dir->dd_pool; dsl_pool_config_enter(dp, FTAG); dsl_dataset_disown(ds, flags, tag); VERIFY0(dsl_dataset_own(dp, name, flags, tag, newds)); dsl_pool_config_exit(dp, FTAG); } void dmu_objset_disown(objset_t *os, boolean_t decrypt, const void *tag) { ds_hold_flags_t flags; flags = (decrypt) ? DS_HOLD_FLAG_DECRYPT : DS_HOLD_FLAG_NONE; /* * Stop upgrading thread */ dmu_objset_upgrade_stop(os); dsl_dataset_disown(os->os_dsl_dataset, flags, tag); } void dmu_objset_evict_dbufs(objset_t *os) { dnode_t *dn_marker; dnode_t *dn; dn_marker = kmem_alloc(sizeof (dnode_t), KM_SLEEP); mutex_enter(&os->os_lock); dn = list_head(&os->os_dnodes); while (dn != NULL) { /* * Skip dnodes without holds. We have to do this dance * because dnode_add_ref() only works if there is already a * hold. If the dnode has no holds, then it has no dbufs. */ if (dnode_add_ref(dn, FTAG)) { list_insert_after(&os->os_dnodes, dn, dn_marker); mutex_exit(&os->os_lock); dnode_evict_dbufs(dn); dnode_rele(dn, FTAG); mutex_enter(&os->os_lock); dn = list_next(&os->os_dnodes, dn_marker); list_remove(&os->os_dnodes, dn_marker); } else { dn = list_next(&os->os_dnodes, dn); } } mutex_exit(&os->os_lock); kmem_free(dn_marker, sizeof (dnode_t)); if (DMU_USERUSED_DNODE(os) != NULL) { if (DMU_PROJECTUSED_DNODE(os) != NULL) dnode_evict_dbufs(DMU_PROJECTUSED_DNODE(os)); dnode_evict_dbufs(DMU_GROUPUSED_DNODE(os)); dnode_evict_dbufs(DMU_USERUSED_DNODE(os)); } dnode_evict_dbufs(DMU_META_DNODE(os)); } /* * Objset eviction processing is split into into two pieces. * The first marks the objset as evicting, evicts any dbufs that * have a refcount of zero, and then queues up the objset for the * second phase of eviction. Once os->os_dnodes has been cleared by * dnode_buf_pageout()->dnode_destroy(), the second phase is executed. * The second phase closes the special dnodes, dequeues the objset from * the list of those undergoing eviction, and finally frees the objset. * * NOTE: Due to asynchronous eviction processing (invocation of * dnode_buf_pageout()), it is possible for the meta dnode for the * objset to have no holds even though os->os_dnodes is not empty. */ void dmu_objset_evict(objset_t *os) { dsl_dataset_t *ds = os->os_dsl_dataset; for (int t = 0; t < TXG_SIZE; t++) ASSERT(!dmu_objset_is_dirty(os, t)); if (ds) dsl_prop_unregister_all(ds, os); if (os->os_sa) sa_tear_down(os); dmu_objset_evict_dbufs(os); mutex_enter(&os->os_lock); spa_evicting_os_register(os->os_spa, os); if (list_is_empty(&os->os_dnodes)) { mutex_exit(&os->os_lock); dmu_objset_evict_done(os); } else { mutex_exit(&os->os_lock); } } void dmu_objset_evict_done(objset_t *os) { ASSERT3P(list_head(&os->os_dnodes), ==, NULL); dnode_special_close(&os->os_meta_dnode); if (DMU_USERUSED_DNODE(os)) { if (DMU_PROJECTUSED_DNODE(os)) dnode_special_close(&os->os_projectused_dnode); dnode_special_close(&os->os_userused_dnode); dnode_special_close(&os->os_groupused_dnode); } zil_free(os->os_zil); arc_buf_destroy(os->os_phys_buf, &os->os_phys_buf); /* * This is a barrier to prevent the objset from going away in * dnode_move() until we can safely ensure that the objset is still in * use. We consider the objset valid before the barrier and invalid * after the barrier. */ rw_enter(&os_lock, RW_READER); rw_exit(&os_lock); kmem_free(os->os_obj_next_percpu, os->os_obj_next_percpu_len * sizeof (os->os_obj_next_percpu[0])); mutex_destroy(&os->os_lock); mutex_destroy(&os->os_userused_lock); mutex_destroy(&os->os_obj_lock); mutex_destroy(&os->os_user_ptr_lock); mutex_destroy(&os->os_upgrade_lock); for (int i = 0; i < TXG_SIZE; i++) multilist_destroy(&os->os_dirty_dnodes[i]); spa_evicting_os_deregister(os->os_spa, os); kmem_free(os, sizeof (objset_t)); } inode_timespec_t dmu_objset_snap_cmtime(objset_t *os) { return (dsl_dir_snap_cmtime(os->os_dsl_dataset->ds_dir)); } objset_t * dmu_objset_create_impl_dnstats(spa_t *spa, dsl_dataset_t *ds, blkptr_t *bp, dmu_objset_type_t type, int levels, int blksz, int ibs, dmu_tx_t *tx) { objset_t *os; dnode_t *mdn; ASSERT(dmu_tx_is_syncing(tx)); if (blksz == 0) blksz = DNODE_BLOCK_SIZE; if (ibs == 0) ibs = DN_MAX_INDBLKSHIFT; if (ds != NULL) VERIFY0(dmu_objset_from_ds(ds, &os)); else VERIFY0(dmu_objset_open_impl(spa, NULL, bp, &os)); mdn = DMU_META_DNODE(os); dnode_allocate(mdn, DMU_OT_DNODE, blksz, ibs, DMU_OT_NONE, 0, DNODE_MIN_SLOTS, tx); /* * We don't want to have to increase the meta-dnode's nlevels * later, because then we could do it in quiescing context while * we are also accessing it in open context. * * This precaution is not necessary for the MOS (ds == NULL), * because the MOS is only updated in syncing context. * This is most fortunate: the MOS is the only objset that * needs to be synced multiple times as spa_sync() iterates * to convergence, so minimizing its dn_nlevels matters. */ if (ds != NULL) { if (levels == 0) { levels = 1; /* * Determine the number of levels necessary for the * meta-dnode to contain DN_MAX_OBJECT dnodes. Note * that in order to ensure that we do not overflow * 64 bits, there has to be a nlevels that gives us a * number of blocks > DN_MAX_OBJECT but < 2^64. * Therefore, (mdn->dn_indblkshift - SPA_BLKPTRSHIFT) * (10) must be less than (64 - log2(DN_MAX_OBJECT)) * (16). */ while ((uint64_t)mdn->dn_nblkptr << (mdn->dn_datablkshift - DNODE_SHIFT + (levels - 1) * (mdn->dn_indblkshift - SPA_BLKPTRSHIFT)) < DN_MAX_OBJECT) levels++; } mdn->dn_next_nlevels[tx->tx_txg & TXG_MASK] = mdn->dn_nlevels = levels; } ASSERT(type != DMU_OST_NONE); ASSERT(type != DMU_OST_ANY); ASSERT(type < DMU_OST_NUMTYPES); os->os_phys->os_type = type; /* * Enable user accounting if it is enabled and this is not an * encrypted receive. */ if (dmu_objset_userused_enabled(os) && (!os->os_encrypted || !dmu_objset_is_receiving(os))) { os->os_phys->os_flags |= OBJSET_FLAG_USERACCOUNTING_COMPLETE; if (dmu_objset_userobjused_enabled(os)) { ASSERT3P(ds, !=, NULL); ds->ds_feature_activation[ SPA_FEATURE_USEROBJ_ACCOUNTING] = (void *)B_TRUE; os->os_phys->os_flags |= OBJSET_FLAG_USEROBJACCOUNTING_COMPLETE; } if (dmu_objset_projectquota_enabled(os)) { ASSERT3P(ds, !=, NULL); ds->ds_feature_activation[ SPA_FEATURE_PROJECT_QUOTA] = (void *)B_TRUE; os->os_phys->os_flags |= OBJSET_FLAG_PROJECTQUOTA_COMPLETE; } os->os_flags = os->os_phys->os_flags; } dsl_dataset_dirty(ds, tx); return (os); } /* called from dsl for meta-objset */ objset_t * dmu_objset_create_impl(spa_t *spa, dsl_dataset_t *ds, blkptr_t *bp, dmu_objset_type_t type, dmu_tx_t *tx) { return (dmu_objset_create_impl_dnstats(spa, ds, bp, type, 0, 0, 0, tx)); } typedef struct dmu_objset_create_arg { const char *doca_name; cred_t *doca_cred; proc_t *doca_proc; void (*doca_userfunc)(objset_t *os, void *arg, cred_t *cr, dmu_tx_t *tx); void *doca_userarg; dmu_objset_type_t doca_type; uint64_t doca_flags; dsl_crypto_params_t *doca_dcp; } dmu_objset_create_arg_t; static int dmu_objset_create_check(void *arg, dmu_tx_t *tx) { dmu_objset_create_arg_t *doca = arg; dsl_pool_t *dp = dmu_tx_pool(tx); dsl_dir_t *pdd; dsl_dataset_t *parentds; objset_t *parentos; const char *tail; int error; if (strchr(doca->doca_name, '@') != NULL) return (SET_ERROR(EINVAL)); if (strlen(doca->doca_name) >= ZFS_MAX_DATASET_NAME_LEN) return (SET_ERROR(ENAMETOOLONG)); if (dataset_nestcheck(doca->doca_name) != 0) return (SET_ERROR(ENAMETOOLONG)); error = dsl_dir_hold(dp, doca->doca_name, FTAG, &pdd, &tail); if (error != 0) return (error); if (tail == NULL) { dsl_dir_rele(pdd, FTAG); return (SET_ERROR(EEXIST)); } error = dmu_objset_create_crypt_check(pdd, doca->doca_dcp, NULL); if (error != 0) { dsl_dir_rele(pdd, FTAG); return (error); } error = dsl_fs_ss_limit_check(pdd, 1, ZFS_PROP_FILESYSTEM_LIMIT, NULL, doca->doca_cred, doca->doca_proc); if (error != 0) { dsl_dir_rele(pdd, FTAG); return (error); } /* can't create below anything but filesystems (eg. no ZVOLs) */ error = dsl_dataset_hold_obj(pdd->dd_pool, dsl_dir_phys(pdd)->dd_head_dataset_obj, FTAG, &parentds); if (error != 0) { dsl_dir_rele(pdd, FTAG); return (error); } error = dmu_objset_from_ds(parentds, &parentos); if (error != 0) { dsl_dataset_rele(parentds, FTAG); dsl_dir_rele(pdd, FTAG); return (error); } if (dmu_objset_type(parentos) != DMU_OST_ZFS) { dsl_dataset_rele(parentds, FTAG); dsl_dir_rele(pdd, FTAG); return (SET_ERROR(ZFS_ERR_WRONG_PARENT)); } dsl_dataset_rele(parentds, FTAG); dsl_dir_rele(pdd, FTAG); return (error); } static void dmu_objset_create_sync(void *arg, dmu_tx_t *tx) { dmu_objset_create_arg_t *doca = arg; dsl_pool_t *dp = dmu_tx_pool(tx); spa_t *spa = dp->dp_spa; dsl_dir_t *pdd; const char *tail; dsl_dataset_t *ds; uint64_t obj; blkptr_t *bp; objset_t *os; zio_t *rzio; VERIFY0(dsl_dir_hold(dp, doca->doca_name, FTAG, &pdd, &tail)); obj = dsl_dataset_create_sync(pdd, tail, NULL, doca->doca_flags, doca->doca_cred, doca->doca_dcp, tx); VERIFY0(dsl_dataset_hold_obj_flags(pdd->dd_pool, obj, DS_HOLD_FLAG_DECRYPT, FTAG, &ds)); rrw_enter(&ds->ds_bp_rwlock, RW_READER, FTAG); bp = dsl_dataset_get_blkptr(ds); os = dmu_objset_create_impl(spa, ds, bp, doca->doca_type, tx); rrw_exit(&ds->ds_bp_rwlock, FTAG); if (doca->doca_userfunc != NULL) { doca->doca_userfunc(os, doca->doca_userarg, doca->doca_cred, tx); } /* * The doca_userfunc() may write out some data that needs to be * encrypted if the dataset is encrypted (specifically the root * directory). This data must be written out before the encryption * key mapping is removed by dsl_dataset_rele_flags(). Force the * I/O to occur immediately by invoking the relevant sections of * dsl_pool_sync(). */ if (os->os_encrypted) { dsl_dataset_t *tmpds = NULL; boolean_t need_sync_done = B_FALSE; mutex_enter(&ds->ds_lock); ds->ds_owner = FTAG; mutex_exit(&ds->ds_lock); rzio = zio_root(spa, NULL, NULL, ZIO_FLAG_MUSTSUCCEED); tmpds = txg_list_remove_this(&dp->dp_dirty_datasets, ds, tx->tx_txg); if (tmpds != NULL) { dsl_dataset_sync(ds, rzio, tx); need_sync_done = B_TRUE; } VERIFY0(zio_wait(rzio)); dmu_objset_sync_done(os, tx); taskq_wait(dp->dp_sync_taskq); if (txg_list_member(&dp->dp_dirty_datasets, ds, tx->tx_txg)) { ASSERT3P(ds->ds_key_mapping, !=, NULL); key_mapping_rele(spa, ds->ds_key_mapping, ds); } rzio = zio_root(spa, NULL, NULL, ZIO_FLAG_MUSTSUCCEED); tmpds = txg_list_remove_this(&dp->dp_dirty_datasets, ds, tx->tx_txg); if (tmpds != NULL) { dmu_buf_rele(ds->ds_dbuf, ds); dsl_dataset_sync(ds, rzio, tx); } VERIFY0(zio_wait(rzio)); if (need_sync_done) { ASSERT3P(ds->ds_key_mapping, !=, NULL); key_mapping_rele(spa, ds->ds_key_mapping, ds); dsl_dataset_sync_done(ds, tx); dmu_buf_rele(ds->ds_dbuf, ds); } mutex_enter(&ds->ds_lock); ds->ds_owner = NULL; mutex_exit(&ds->ds_lock); } spa_history_log_internal_ds(ds, "create", tx, " "); dsl_dataset_rele_flags(ds, DS_HOLD_FLAG_DECRYPT, FTAG); dsl_dir_rele(pdd, FTAG); } int dmu_objset_create(const char *name, dmu_objset_type_t type, uint64_t flags, dsl_crypto_params_t *dcp, dmu_objset_create_sync_func_t func, void *arg) { dmu_objset_create_arg_t doca; dsl_crypto_params_t tmp_dcp = { 0 }; doca.doca_name = name; doca.doca_cred = CRED(); doca.doca_proc = curproc; doca.doca_flags = flags; doca.doca_userfunc = func; doca.doca_userarg = arg; doca.doca_type = type; /* * Some callers (mostly for testing) do not provide a dcp on their * own but various code inside the sync task will require it to be * allocated. Rather than adding NULL checks throughout this code * or adding dummy dcp's to all of the callers we simply create a * dummy one here and use that. This zero dcp will have the same * effect as asking for inheritance of all encryption params. */ doca.doca_dcp = (dcp != NULL) ? dcp : &tmp_dcp; int rv = dsl_sync_task(name, dmu_objset_create_check, dmu_objset_create_sync, &doca, 6, ZFS_SPACE_CHECK_NORMAL); if (rv == 0) zvol_create_minor(name); return (rv); } typedef struct dmu_objset_clone_arg { const char *doca_clone; const char *doca_origin; cred_t *doca_cred; proc_t *doca_proc; } dmu_objset_clone_arg_t; static int dmu_objset_clone_check(void *arg, dmu_tx_t *tx) { dmu_objset_clone_arg_t *doca = arg; dsl_dir_t *pdd; const char *tail; int error; dsl_dataset_t *origin; dsl_pool_t *dp = dmu_tx_pool(tx); if (strchr(doca->doca_clone, '@') != NULL) return (SET_ERROR(EINVAL)); if (strlen(doca->doca_clone) >= ZFS_MAX_DATASET_NAME_LEN) return (SET_ERROR(ENAMETOOLONG)); error = dsl_dir_hold(dp, doca->doca_clone, FTAG, &pdd, &tail); if (error != 0) return (error); if (tail == NULL) { dsl_dir_rele(pdd, FTAG); return (SET_ERROR(EEXIST)); } error = dsl_fs_ss_limit_check(pdd, 1, ZFS_PROP_FILESYSTEM_LIMIT, NULL, doca->doca_cred, doca->doca_proc); if (error != 0) { dsl_dir_rele(pdd, FTAG); return (SET_ERROR(EDQUOT)); } error = dsl_dataset_hold(dp, doca->doca_origin, FTAG, &origin); if (error != 0) { dsl_dir_rele(pdd, FTAG); return (error); } /* You can only clone snapshots, not the head datasets. */ if (!origin->ds_is_snapshot) { dsl_dataset_rele(origin, FTAG); dsl_dir_rele(pdd, FTAG); return (SET_ERROR(EINVAL)); } dsl_dataset_rele(origin, FTAG); dsl_dir_rele(pdd, FTAG); return (0); } static void dmu_objset_clone_sync(void *arg, dmu_tx_t *tx) { dmu_objset_clone_arg_t *doca = arg; dsl_pool_t *dp = dmu_tx_pool(tx); dsl_dir_t *pdd; const char *tail; dsl_dataset_t *origin, *ds; uint64_t obj; char namebuf[ZFS_MAX_DATASET_NAME_LEN]; VERIFY0(dsl_dir_hold(dp, doca->doca_clone, FTAG, &pdd, &tail)); VERIFY0(dsl_dataset_hold(dp, doca->doca_origin, FTAG, &origin)); obj = dsl_dataset_create_sync(pdd, tail, origin, 0, doca->doca_cred, NULL, tx); VERIFY0(dsl_dataset_hold_obj(pdd->dd_pool, obj, FTAG, &ds)); dsl_dataset_name(origin, namebuf); spa_history_log_internal_ds(ds, "clone", tx, "origin=%s (%llu)", namebuf, (u_longlong_t)origin->ds_object); dsl_dataset_rele(ds, FTAG); dsl_dataset_rele(origin, FTAG); dsl_dir_rele(pdd, FTAG); } int dmu_objset_clone(const char *clone, const char *origin) { dmu_objset_clone_arg_t doca; doca.doca_clone = clone; doca.doca_origin = origin; doca.doca_cred = CRED(); doca.doca_proc = curproc; int rv = dsl_sync_task(clone, dmu_objset_clone_check, dmu_objset_clone_sync, &doca, 6, ZFS_SPACE_CHECK_NORMAL); if (rv == 0) zvol_create_minor(clone); return (rv); } int dmu_objset_snapshot_one(const char *fsname, const char *snapname) { int err; char *longsnap = kmem_asprintf("%s@%s", fsname, snapname); nvlist_t *snaps = fnvlist_alloc(); fnvlist_add_boolean(snaps, longsnap); kmem_strfree(longsnap); err = dsl_dataset_snapshot(snaps, NULL, NULL); fnvlist_free(snaps); return (err); } static void dmu_objset_upgrade_task_cb(void *data) { objset_t *os = data; mutex_enter(&os->os_upgrade_lock); os->os_upgrade_status = EINTR; if (!os->os_upgrade_exit) { int status; mutex_exit(&os->os_upgrade_lock); status = os->os_upgrade_cb(os); mutex_enter(&os->os_upgrade_lock); os->os_upgrade_status = status; } os->os_upgrade_exit = B_TRUE; os->os_upgrade_id = 0; mutex_exit(&os->os_upgrade_lock); dsl_dataset_long_rele(dmu_objset_ds(os), upgrade_tag); } static void dmu_objset_upgrade(objset_t *os, dmu_objset_upgrade_cb_t cb) { if (os->os_upgrade_id != 0) return; ASSERT(dsl_pool_config_held(dmu_objset_pool(os))); dsl_dataset_long_hold(dmu_objset_ds(os), upgrade_tag); mutex_enter(&os->os_upgrade_lock); if (os->os_upgrade_id == 0 && os->os_upgrade_status == 0) { os->os_upgrade_exit = B_FALSE; os->os_upgrade_cb = cb; os->os_upgrade_id = taskq_dispatch( os->os_spa->spa_upgrade_taskq, dmu_objset_upgrade_task_cb, os, TQ_SLEEP); if (os->os_upgrade_id == TASKQID_INVALID) { dsl_dataset_long_rele(dmu_objset_ds(os), upgrade_tag); os->os_upgrade_status = ENOMEM; } } else { dsl_dataset_long_rele(dmu_objset_ds(os), upgrade_tag); } mutex_exit(&os->os_upgrade_lock); } static void dmu_objset_upgrade_stop(objset_t *os) { mutex_enter(&os->os_upgrade_lock); os->os_upgrade_exit = B_TRUE; if (os->os_upgrade_id != 0) { taskqid_t id = os->os_upgrade_id; os->os_upgrade_id = 0; mutex_exit(&os->os_upgrade_lock); if ((taskq_cancel_id(os->os_spa->spa_upgrade_taskq, id)) == 0) { dsl_dataset_long_rele(dmu_objset_ds(os), upgrade_tag); } txg_wait_synced(os->os_spa->spa_dsl_pool, 0); } else { mutex_exit(&os->os_upgrade_lock); } } static void dmu_objset_sync_dnodes(multilist_sublist_t *list, dmu_tx_t *tx) { dnode_t *dn; while ((dn = multilist_sublist_head(list)) != NULL) { ASSERT(dn->dn_object != DMU_META_DNODE_OBJECT); ASSERT(dn->dn_dbuf->db_data_pending); /* * Initialize dn_zio outside dnode_sync() because the * meta-dnode needs to set it outside dnode_sync(). */ dn->dn_zio = dn->dn_dbuf->db_data_pending->dr_zio; ASSERT(dn->dn_zio); ASSERT3U(dn->dn_nlevels, <=, DN_MAX_LEVELS); multilist_sublist_remove(list, dn); /* * See the comment above dnode_rele_task() for an explanation * of why this dnode hold is always needed (even when not * doing user accounting). */ multilist_t *newlist = &dn->dn_objset->os_synced_dnodes; (void) dnode_add_ref(dn, newlist); multilist_insert(newlist, dn); dnode_sync(dn, tx); } } static void dmu_objset_write_ready(zio_t *zio, arc_buf_t *abuf, void *arg) { (void) abuf; blkptr_t *bp = zio->io_bp; objset_t *os = arg; dnode_phys_t *dnp = &os->os_phys->os_meta_dnode; uint64_t fill = 0; ASSERT(!BP_IS_EMBEDDED(bp)); ASSERT3U(BP_GET_TYPE(bp), ==, DMU_OT_OBJSET); ASSERT0(BP_GET_LEVEL(bp)); /* * Update rootbp fill count: it should be the number of objects * allocated in the object set (not counting the "special" * objects that are stored in the objset_phys_t -- the meta * dnode and user/group/project accounting objects). */ for (int i = 0; i < dnp->dn_nblkptr; i++) fill += BP_GET_FILL(&dnp->dn_blkptr[i]); BP_SET_FILL(bp, fill); if (os->os_dsl_dataset != NULL) rrw_enter(&os->os_dsl_dataset->ds_bp_rwlock, RW_WRITER, FTAG); *os->os_rootbp = *bp; if (os->os_dsl_dataset != NULL) rrw_exit(&os->os_dsl_dataset->ds_bp_rwlock, FTAG); } static void dmu_objset_write_done(zio_t *zio, arc_buf_t *abuf, void *arg) { (void) abuf; blkptr_t *bp = zio->io_bp; blkptr_t *bp_orig = &zio->io_bp_orig; objset_t *os = arg; if (zio->io_flags & ZIO_FLAG_IO_REWRITE) { ASSERT(BP_EQUAL(bp, bp_orig)); } else { dsl_dataset_t *ds = os->os_dsl_dataset; dmu_tx_t *tx = os->os_synctx; (void) dsl_dataset_block_kill(ds, bp_orig, tx, B_TRUE); dsl_dataset_block_born(ds, bp, tx); } kmem_free(bp, sizeof (*bp)); } typedef struct sync_dnodes_arg { multilist_t *sda_list; int sda_sublist_idx; multilist_t *sda_newlist; dmu_tx_t *sda_tx; } sync_dnodes_arg_t; static void sync_dnodes_task(void *arg) { sync_dnodes_arg_t *sda = arg; multilist_sublist_t *ms = multilist_sublist_lock_idx(sda->sda_list, sda->sda_sublist_idx); dmu_objset_sync_dnodes(ms, sda->sda_tx); multilist_sublist_unlock(ms); kmem_free(sda, sizeof (*sda)); } /* called from dsl */ void dmu_objset_sync(objset_t *os, zio_t *pio, dmu_tx_t *tx) { int txgoff; zbookmark_phys_t zb; zio_prop_t zp; zio_t *zio; list_t *list; dbuf_dirty_record_t *dr; int num_sublists; multilist_t *ml; blkptr_t *blkptr_copy = kmem_alloc(sizeof (*os->os_rootbp), KM_SLEEP); *blkptr_copy = *os->os_rootbp; dprintf_ds(os->os_dsl_dataset, "txg=%llu\n", (u_longlong_t)tx->tx_txg); ASSERT(dmu_tx_is_syncing(tx)); /* XXX the write_done callback should really give us the tx... */ os->os_synctx = tx; if (os->os_dsl_dataset == NULL) { /* * This is the MOS. If we have upgraded, * spa_max_replication() could change, so reset * os_copies here. */ os->os_copies = spa_max_replication(os->os_spa); } /* * Create the root block IO */ SET_BOOKMARK(&zb, os->os_dsl_dataset ? os->os_dsl_dataset->ds_object : DMU_META_OBJSET, ZB_ROOT_OBJECT, ZB_ROOT_LEVEL, ZB_ROOT_BLKID); arc_release(os->os_phys_buf, &os->os_phys_buf); dmu_write_policy(os, NULL, 0, 0, &zp); /* * If we are either claiming the ZIL or doing a raw receive, write * out the os_phys_buf raw. Neither of these actions will effect the * MAC at this point. */ if (os->os_raw_receive || os->os_next_write_raw[tx->tx_txg & TXG_MASK]) { ASSERT(os->os_encrypted); arc_convert_to_raw(os->os_phys_buf, os->os_dsl_dataset->ds_object, ZFS_HOST_BYTEORDER, DMU_OT_OBJSET, NULL, NULL, NULL); } zio = arc_write(pio, os->os_spa, tx->tx_txg, blkptr_copy, os->os_phys_buf, B_FALSE, dmu_os_is_l2cacheable(os), &zp, dmu_objset_write_ready, NULL, dmu_objset_write_done, os, ZIO_PRIORITY_ASYNC_WRITE, ZIO_FLAG_MUSTSUCCEED, &zb); /* * Sync special dnodes - the parent IO for the sync is the root block */ DMU_META_DNODE(os)->dn_zio = zio; dnode_sync(DMU_META_DNODE(os), tx); os->os_phys->os_flags = os->os_flags; if (DMU_USERUSED_DNODE(os) && DMU_USERUSED_DNODE(os)->dn_type != DMU_OT_NONE) { DMU_USERUSED_DNODE(os)->dn_zio = zio; dnode_sync(DMU_USERUSED_DNODE(os), tx); DMU_GROUPUSED_DNODE(os)->dn_zio = zio; dnode_sync(DMU_GROUPUSED_DNODE(os), tx); } if (DMU_PROJECTUSED_DNODE(os) && DMU_PROJECTUSED_DNODE(os)->dn_type != DMU_OT_NONE) { DMU_PROJECTUSED_DNODE(os)->dn_zio = zio; dnode_sync(DMU_PROJECTUSED_DNODE(os), tx); } txgoff = tx->tx_txg & TXG_MASK; /* * We must create the list here because it uses the * dn_dirty_link[] of this txg. But it may already * exist because we call dsl_dataset_sync() twice per txg. */ if (os->os_synced_dnodes.ml_sublists == NULL) { multilist_create(&os->os_synced_dnodes, sizeof (dnode_t), offsetof(dnode_t, dn_dirty_link[txgoff]), dnode_multilist_index_func); } else { ASSERT3U(os->os_synced_dnodes.ml_offset, ==, offsetof(dnode_t, dn_dirty_link[txgoff])); } ml = &os->os_dirty_dnodes[txgoff]; num_sublists = multilist_get_num_sublists(ml); for (int i = 0; i < num_sublists; i++) { if (multilist_sublist_is_empty_idx(ml, i)) continue; sync_dnodes_arg_t *sda = kmem_alloc(sizeof (*sda), KM_SLEEP); sda->sda_list = ml; sda->sda_sublist_idx = i; sda->sda_tx = tx; (void) taskq_dispatch(dmu_objset_pool(os)->dp_sync_taskq, sync_dnodes_task, sda, 0); /* callback frees sda */ } taskq_wait(dmu_objset_pool(os)->dp_sync_taskq); list = &DMU_META_DNODE(os)->dn_dirty_records[txgoff]; while ((dr = list_remove_head(list)) != NULL) { ASSERT0(dr->dr_dbuf->db_level); zio_nowait(dr->dr_zio); } /* Enable dnode backfill if enough objects have been freed. */ if (os->os_freed_dnodes >= dmu_rescan_dnode_threshold) { os->os_rescan_dnodes = B_TRUE; os->os_freed_dnodes = 0; } /* * Free intent log blocks up to this tx. */ zil_sync(os->os_zil, tx); os->os_phys->os_zil_header = os->os_zil_header; zio_nowait(zio); } boolean_t dmu_objset_is_dirty(objset_t *os, uint64_t txg) { return (!multilist_is_empty(&os->os_dirty_dnodes[txg & TXG_MASK])); } static file_info_cb_t *file_cbs[DMU_OST_NUMTYPES]; void dmu_objset_register_type(dmu_objset_type_t ost, file_info_cb_t *cb) { file_cbs[ost] = cb; } int dmu_get_file_info(objset_t *os, dmu_object_type_t bonustype, const void *data, zfs_file_info_t *zfi) { file_info_cb_t *cb = file_cbs[os->os_phys->os_type]; if (cb == NULL) return (EINVAL); return (cb(bonustype, data, zfi)); } boolean_t dmu_objset_userused_enabled(objset_t *os) { return (spa_version(os->os_spa) >= SPA_VERSION_USERSPACE && file_cbs[os->os_phys->os_type] != NULL && DMU_USERUSED_DNODE(os) != NULL); } boolean_t dmu_objset_userobjused_enabled(objset_t *os) { return (dmu_objset_userused_enabled(os) && spa_feature_is_enabled(os->os_spa, SPA_FEATURE_USEROBJ_ACCOUNTING)); } boolean_t dmu_objset_projectquota_enabled(objset_t *os) { return (file_cbs[os->os_phys->os_type] != NULL && DMU_PROJECTUSED_DNODE(os) != NULL && spa_feature_is_enabled(os->os_spa, SPA_FEATURE_PROJECT_QUOTA)); } typedef struct userquota_node { /* must be in the first filed, see userquota_update_cache() */ char uqn_id[20 + DMU_OBJACCT_PREFIX_LEN]; int64_t uqn_delta; avl_node_t uqn_node; } userquota_node_t; typedef struct userquota_cache { avl_tree_t uqc_user_deltas; avl_tree_t uqc_group_deltas; avl_tree_t uqc_project_deltas; } userquota_cache_t; static int userquota_compare(const void *l, const void *r) { const userquota_node_t *luqn = l; const userquota_node_t *ruqn = r; int rv; /* * NB: can only access uqn_id because userquota_update_cache() doesn't * pass in an entire userquota_node_t. */ rv = strcmp(luqn->uqn_id, ruqn->uqn_id); return (TREE_ISIGN(rv)); } static void do_userquota_cacheflush(objset_t *os, userquota_cache_t *cache, dmu_tx_t *tx) { void *cookie; userquota_node_t *uqn; ASSERT(dmu_tx_is_syncing(tx)); cookie = NULL; while ((uqn = avl_destroy_nodes(&cache->uqc_user_deltas, &cookie)) != NULL) { /* * os_userused_lock protects against concurrent calls to * zap_increment_int(). It's needed because zap_increment_int() * is not thread-safe (i.e. not atomic). */ mutex_enter(&os->os_userused_lock); VERIFY0(zap_increment(os, DMU_USERUSED_OBJECT, uqn->uqn_id, uqn->uqn_delta, tx)); mutex_exit(&os->os_userused_lock); kmem_free(uqn, sizeof (*uqn)); } avl_destroy(&cache->uqc_user_deltas); cookie = NULL; while ((uqn = avl_destroy_nodes(&cache->uqc_group_deltas, &cookie)) != NULL) { mutex_enter(&os->os_userused_lock); VERIFY0(zap_increment(os, DMU_GROUPUSED_OBJECT, uqn->uqn_id, uqn->uqn_delta, tx)); mutex_exit(&os->os_userused_lock); kmem_free(uqn, sizeof (*uqn)); } avl_destroy(&cache->uqc_group_deltas); if (dmu_objset_projectquota_enabled(os)) { cookie = NULL; while ((uqn = avl_destroy_nodes(&cache->uqc_project_deltas, &cookie)) != NULL) { mutex_enter(&os->os_userused_lock); VERIFY0(zap_increment(os, DMU_PROJECTUSED_OBJECT, uqn->uqn_id, uqn->uqn_delta, tx)); mutex_exit(&os->os_userused_lock); kmem_free(uqn, sizeof (*uqn)); } avl_destroy(&cache->uqc_project_deltas); } } static void userquota_update_cache(avl_tree_t *avl, const char *id, int64_t delta) { userquota_node_t *uqn; avl_index_t idx; ASSERT(strlen(id) < sizeof (uqn->uqn_id)); /* * Use id directly for searching because uqn_id is the first field of * userquota_node_t and fields after uqn_id won't be accessed in * avl_find(). */ uqn = avl_find(avl, (const void *)id, &idx); if (uqn == NULL) { uqn = kmem_zalloc(sizeof (*uqn), KM_SLEEP); strlcpy(uqn->uqn_id, id, sizeof (uqn->uqn_id)); avl_insert(avl, uqn, idx); } uqn->uqn_delta += delta; } static void do_userquota_update(objset_t *os, userquota_cache_t *cache, uint64_t used, uint64_t flags, uint64_t user, uint64_t group, uint64_t project, boolean_t subtract) { if (flags & DNODE_FLAG_USERUSED_ACCOUNTED) { int64_t delta = DNODE_MIN_SIZE + used; char name[20]; if (subtract) delta = -delta; (void) snprintf(name, sizeof (name), "%llx", (longlong_t)user); userquota_update_cache(&cache->uqc_user_deltas, name, delta); (void) snprintf(name, sizeof (name), "%llx", (longlong_t)group); userquota_update_cache(&cache->uqc_group_deltas, name, delta); if (dmu_objset_projectquota_enabled(os)) { (void) snprintf(name, sizeof (name), "%llx", (longlong_t)project); userquota_update_cache(&cache->uqc_project_deltas, name, delta); } } } static void do_userobjquota_update(objset_t *os, userquota_cache_t *cache, uint64_t flags, uint64_t user, uint64_t group, uint64_t project, boolean_t subtract) { if (flags & DNODE_FLAG_USEROBJUSED_ACCOUNTED) { char name[20 + DMU_OBJACCT_PREFIX_LEN]; int delta = subtract ? -1 : 1; (void) snprintf(name, sizeof (name), DMU_OBJACCT_PREFIX "%llx", (longlong_t)user); userquota_update_cache(&cache->uqc_user_deltas, name, delta); (void) snprintf(name, sizeof (name), DMU_OBJACCT_PREFIX "%llx", (longlong_t)group); userquota_update_cache(&cache->uqc_group_deltas, name, delta); if (dmu_objset_projectquota_enabled(os)) { (void) snprintf(name, sizeof (name), DMU_OBJACCT_PREFIX "%llx", (longlong_t)project); userquota_update_cache(&cache->uqc_project_deltas, name, delta); } } } typedef struct userquota_updates_arg { objset_t *uua_os; int uua_sublist_idx; dmu_tx_t *uua_tx; } userquota_updates_arg_t; static void userquota_updates_task(void *arg) { userquota_updates_arg_t *uua = arg; objset_t *os = uua->uua_os; dmu_tx_t *tx = uua->uua_tx; dnode_t *dn; userquota_cache_t cache = { { 0 } }; multilist_sublist_t *list = multilist_sublist_lock_idx( &os->os_synced_dnodes, uua->uua_sublist_idx); ASSERT(multilist_sublist_head(list) == NULL || dmu_objset_userused_enabled(os)); avl_create(&cache.uqc_user_deltas, userquota_compare, sizeof (userquota_node_t), offsetof(userquota_node_t, uqn_node)); avl_create(&cache.uqc_group_deltas, userquota_compare, sizeof (userquota_node_t), offsetof(userquota_node_t, uqn_node)); if (dmu_objset_projectquota_enabled(os)) avl_create(&cache.uqc_project_deltas, userquota_compare, sizeof (userquota_node_t), offsetof(userquota_node_t, uqn_node)); while ((dn = multilist_sublist_head(list)) != NULL) { int flags; ASSERT(!DMU_OBJECT_IS_SPECIAL(dn->dn_object)); ASSERT(dn->dn_phys->dn_type == DMU_OT_NONE || dn->dn_phys->dn_flags & DNODE_FLAG_USERUSED_ACCOUNTED); flags = dn->dn_id_flags; ASSERT(flags); if (flags & DN_ID_OLD_EXIST) { do_userquota_update(os, &cache, dn->dn_oldused, dn->dn_oldflags, dn->dn_olduid, dn->dn_oldgid, dn->dn_oldprojid, B_TRUE); do_userobjquota_update(os, &cache, dn->dn_oldflags, dn->dn_olduid, dn->dn_oldgid, dn->dn_oldprojid, B_TRUE); } if (flags & DN_ID_NEW_EXIST) { do_userquota_update(os, &cache, DN_USED_BYTES(dn->dn_phys), dn->dn_phys->dn_flags, dn->dn_newuid, dn->dn_newgid, dn->dn_newprojid, B_FALSE); do_userobjquota_update(os, &cache, dn->dn_phys->dn_flags, dn->dn_newuid, dn->dn_newgid, dn->dn_newprojid, B_FALSE); } mutex_enter(&dn->dn_mtx); dn->dn_oldused = 0; dn->dn_oldflags = 0; if (dn->dn_id_flags & DN_ID_NEW_EXIST) { dn->dn_olduid = dn->dn_newuid; dn->dn_oldgid = dn->dn_newgid; dn->dn_oldprojid = dn->dn_newprojid; dn->dn_id_flags |= DN_ID_OLD_EXIST; if (dn->dn_bonuslen == 0) dn->dn_id_flags |= DN_ID_CHKED_SPILL; else dn->dn_id_flags |= DN_ID_CHKED_BONUS; } dn->dn_id_flags &= ~(DN_ID_NEW_EXIST); mutex_exit(&dn->dn_mtx); multilist_sublist_remove(list, dn); dnode_rele(dn, &os->os_synced_dnodes); } do_userquota_cacheflush(os, &cache, tx); multilist_sublist_unlock(list); kmem_free(uua, sizeof (*uua)); } /* * Release dnode holds from dmu_objset_sync_dnodes(). When the dnode is being * synced (i.e. we have issued the zio's for blocks in the dnode), it can't be * evicted because the block containing the dnode can't be evicted until it is * written out. However, this hold is necessary to prevent the dnode_t from * being moved (via dnode_move()) while it's still referenced by * dbuf_dirty_record_t:dr_dnode. And dr_dnode is needed for * dirty_lightweight_leaf-type dirty records. * * If we are doing user-object accounting, the dnode_rele() happens from * userquota_updates_task() instead. */ static void dnode_rele_task(void *arg) { userquota_updates_arg_t *uua = arg; objset_t *os = uua->uua_os; multilist_sublist_t *list = multilist_sublist_lock_idx( &os->os_synced_dnodes, uua->uua_sublist_idx); dnode_t *dn; while ((dn = multilist_sublist_head(list)) != NULL) { multilist_sublist_remove(list, dn); dnode_rele(dn, &os->os_synced_dnodes); } multilist_sublist_unlock(list); kmem_free(uua, sizeof (*uua)); } /* * Return TRUE if userquota updates are needed. */ static boolean_t dmu_objset_do_userquota_updates_prep(objset_t *os, dmu_tx_t *tx) { if (!dmu_objset_userused_enabled(os)) return (B_FALSE); /* * If this is a raw receive just return and handle accounting * later when we have the keys loaded. We also don't do user * accounting during claiming since the datasets are not owned * for the duration of claiming and this txg should only be * used for recovery. */ if (os->os_encrypted && dmu_objset_is_receiving(os)) return (B_FALSE); if (tx->tx_txg <= os->os_spa->spa_claim_max_txg) return (B_FALSE); /* Allocate the user/group/project used objects if necessary. */ if (DMU_USERUSED_DNODE(os)->dn_type == DMU_OT_NONE) { VERIFY0(zap_create_claim(os, DMU_USERUSED_OBJECT, DMU_OT_USERGROUP_USED, DMU_OT_NONE, 0, tx)); VERIFY0(zap_create_claim(os, DMU_GROUPUSED_OBJECT, DMU_OT_USERGROUP_USED, DMU_OT_NONE, 0, tx)); } if (dmu_objset_projectquota_enabled(os) && DMU_PROJECTUSED_DNODE(os)->dn_type == DMU_OT_NONE) { VERIFY0(zap_create_claim(os, DMU_PROJECTUSED_OBJECT, DMU_OT_USERGROUP_USED, DMU_OT_NONE, 0, tx)); } return (B_TRUE); } /* * Dispatch taskq tasks to dp_sync_taskq to update the user accounting, and * also release the holds on the dnodes from dmu_objset_sync_dnodes(). * The caller must taskq_wait(dp_sync_taskq). */ void dmu_objset_sync_done(objset_t *os, dmu_tx_t *tx) { boolean_t need_userquota = dmu_objset_do_userquota_updates_prep(os, tx); int num_sublists = multilist_get_num_sublists(&os->os_synced_dnodes); for (int i = 0; i < num_sublists; i++) { userquota_updates_arg_t *uua = kmem_alloc(sizeof (*uua), KM_SLEEP); uua->uua_os = os; uua->uua_sublist_idx = i; uua->uua_tx = tx; /* * If we don't need to update userquotas, use * dnode_rele_task() to call dnode_rele() */ (void) taskq_dispatch(dmu_objset_pool(os)->dp_sync_taskq, need_userquota ? userquota_updates_task : dnode_rele_task, uua, 0); /* callback frees uua */ } } /* * Returns a pointer to data to find uid/gid from * * If a dirty record for transaction group that is syncing can't * be found then NULL is returned. In the NULL case it is assumed * the uid/gid aren't changing. */ static void * dmu_objset_userquota_find_data(dmu_buf_impl_t *db, dmu_tx_t *tx) { dbuf_dirty_record_t *dr; void *data; if (db->db_dirtycnt == 0) return (db->db.db_data); /* Nothing is changing */ dr = dbuf_find_dirty_eq(db, tx->tx_txg); if (dr == NULL) { data = NULL; } else { if (dr->dr_dnode->dn_bonuslen == 0 && dr->dr_dbuf->db_blkid == DMU_SPILL_BLKID) data = dr->dt.dl.dr_data->b_data; else data = dr->dt.dl.dr_data; } return (data); } void dmu_objset_userquota_get_ids(dnode_t *dn, boolean_t before, dmu_tx_t *tx) { objset_t *os = dn->dn_objset; void *data = NULL; dmu_buf_impl_t *db = NULL; int flags = dn->dn_id_flags; int error; boolean_t have_spill = B_FALSE; if (!dmu_objset_userused_enabled(dn->dn_objset)) return; /* * Raw receives introduce a problem with user accounting. Raw * receives cannot update the user accounting info because the * user ids and the sizes are encrypted. To guarantee that we * never end up with bad user accounting, we simply disable it * during raw receives. We also disable this for normal receives * so that an incremental raw receive may be done on top of an * existing non-raw receive. */ if (os->os_encrypted && dmu_objset_is_receiving(os)) return; if (before && (flags & (DN_ID_CHKED_BONUS|DN_ID_OLD_EXIST| DN_ID_CHKED_SPILL))) return; if (before && dn->dn_bonuslen != 0) data = DN_BONUS(dn->dn_phys); else if (!before && dn->dn_bonuslen != 0) { if (dn->dn_bonus) { db = dn->dn_bonus; mutex_enter(&db->db_mtx); data = dmu_objset_userquota_find_data(db, tx); } else { data = DN_BONUS(dn->dn_phys); } } else if (dn->dn_bonuslen == 0 && dn->dn_bonustype == DMU_OT_SA) { int rf = 0; if (RW_WRITE_HELD(&dn->dn_struct_rwlock)) rf |= DB_RF_HAVESTRUCT; error = dmu_spill_hold_by_dnode(dn, rf | DB_RF_MUST_SUCCEED, FTAG, (dmu_buf_t **)&db); ASSERT(error == 0); mutex_enter(&db->db_mtx); data = (before) ? db->db.db_data : dmu_objset_userquota_find_data(db, tx); have_spill = B_TRUE; } else { mutex_enter(&dn->dn_mtx); dn->dn_id_flags |= DN_ID_CHKED_BONUS; mutex_exit(&dn->dn_mtx); return; } /* * Must always call the callback in case the object * type has changed and that type isn't an object type to track */ zfs_file_info_t zfi; error = file_cbs[os->os_phys->os_type](dn->dn_bonustype, data, &zfi); if (before) { ASSERT(data); dn->dn_olduid = zfi.zfi_user; dn->dn_oldgid = zfi.zfi_group; dn->dn_oldprojid = zfi.zfi_project; } else if (data) { dn->dn_newuid = zfi.zfi_user; dn->dn_newgid = zfi.zfi_group; dn->dn_newprojid = zfi.zfi_project; } /* * Preserve existing uid/gid when the callback can't determine * what the new uid/gid are and the callback returned EEXIST. * The EEXIST error tells us to just use the existing uid/gid. * If we don't know what the old values are then just assign * them to 0, since that is a new file being created. */ if (!before && data == NULL && error == EEXIST) { if (flags & DN_ID_OLD_EXIST) { dn->dn_newuid = dn->dn_olduid; dn->dn_newgid = dn->dn_oldgid; dn->dn_newprojid = dn->dn_oldprojid; } else { dn->dn_newuid = 0; dn->dn_newgid = 0; dn->dn_newprojid = ZFS_DEFAULT_PROJID; } error = 0; } if (db) mutex_exit(&db->db_mtx); mutex_enter(&dn->dn_mtx); if (error == 0 && before) dn->dn_id_flags |= DN_ID_OLD_EXIST; if (error == 0 && !before) dn->dn_id_flags |= DN_ID_NEW_EXIST; if (have_spill) { dn->dn_id_flags |= DN_ID_CHKED_SPILL; } else { dn->dn_id_flags |= DN_ID_CHKED_BONUS; } mutex_exit(&dn->dn_mtx); if (have_spill) dmu_buf_rele((dmu_buf_t *)db, FTAG); } boolean_t dmu_objset_userspace_present(objset_t *os) { return (os->os_phys->os_flags & OBJSET_FLAG_USERACCOUNTING_COMPLETE); } boolean_t dmu_objset_userobjspace_present(objset_t *os) { return (os->os_phys->os_flags & OBJSET_FLAG_USEROBJACCOUNTING_COMPLETE); } boolean_t dmu_objset_projectquota_present(objset_t *os) { return (os->os_phys->os_flags & OBJSET_FLAG_PROJECTQUOTA_COMPLETE); } static int dmu_objset_space_upgrade(objset_t *os) { uint64_t obj; int err = 0; /* * We simply need to mark every object dirty, so that it will be * synced out and now accounted. If this is called * concurrently, or if we already did some work before crashing, * that's fine, since we track each object's accounted state * independently. */ for (obj = 0; err == 0; err = dmu_object_next(os, &obj, FALSE, 0)) { dmu_tx_t *tx; dmu_buf_t *db; int objerr; mutex_enter(&os->os_upgrade_lock); if (os->os_upgrade_exit) err = SET_ERROR(EINTR); mutex_exit(&os->os_upgrade_lock); if (err != 0) return (err); if (issig(JUSTLOOKING) && issig(FORREAL)) return (SET_ERROR(EINTR)); objerr = dmu_bonus_hold(os, obj, FTAG, &db); if (objerr != 0) continue; tx = dmu_tx_create(os); dmu_tx_hold_bonus(tx, obj); objerr = dmu_tx_assign(tx, TXG_WAIT); if (objerr != 0) { dmu_buf_rele(db, FTAG); dmu_tx_abort(tx); continue; } dmu_buf_will_dirty(db, tx); dmu_buf_rele(db, FTAG); dmu_tx_commit(tx); } return (0); } static int dmu_objset_userspace_upgrade_cb(objset_t *os) { int err = 0; if (dmu_objset_userspace_present(os)) return (0); if (dmu_objset_is_snapshot(os)) return (SET_ERROR(EINVAL)); if (!dmu_objset_userused_enabled(os)) return (SET_ERROR(ENOTSUP)); err = dmu_objset_space_upgrade(os); if (err) return (err); os->os_flags |= OBJSET_FLAG_USERACCOUNTING_COMPLETE; txg_wait_synced(dmu_objset_pool(os), 0); return (0); } void dmu_objset_userspace_upgrade(objset_t *os) { dmu_objset_upgrade(os, dmu_objset_userspace_upgrade_cb); } static int dmu_objset_id_quota_upgrade_cb(objset_t *os) { int err = 0; if (dmu_objset_userobjspace_present(os) && dmu_objset_projectquota_present(os)) return (0); if (dmu_objset_is_snapshot(os)) return (SET_ERROR(EINVAL)); if (!dmu_objset_userused_enabled(os)) return (SET_ERROR(ENOTSUP)); if (!dmu_objset_projectquota_enabled(os) && dmu_objset_userobjspace_present(os)) return (SET_ERROR(ENOTSUP)); if (dmu_objset_userobjused_enabled(os)) dmu_objset_ds(os)->ds_feature_activation[ SPA_FEATURE_USEROBJ_ACCOUNTING] = (void *)B_TRUE; if (dmu_objset_projectquota_enabled(os)) dmu_objset_ds(os)->ds_feature_activation[ SPA_FEATURE_PROJECT_QUOTA] = (void *)B_TRUE; err = dmu_objset_space_upgrade(os); if (err) return (err); os->os_flags |= OBJSET_FLAG_USERACCOUNTING_COMPLETE; if (dmu_objset_userobjused_enabled(os)) os->os_flags |= OBJSET_FLAG_USEROBJACCOUNTING_COMPLETE; if (dmu_objset_projectquota_enabled(os)) os->os_flags |= OBJSET_FLAG_PROJECTQUOTA_COMPLETE; txg_wait_synced(dmu_objset_pool(os), 0); return (0); } void dmu_objset_id_quota_upgrade(objset_t *os) { dmu_objset_upgrade(os, dmu_objset_id_quota_upgrade_cb); } boolean_t dmu_objset_userobjspace_upgradable(objset_t *os) { return (dmu_objset_type(os) == DMU_OST_ZFS && !dmu_objset_is_snapshot(os) && dmu_objset_userobjused_enabled(os) && !dmu_objset_userobjspace_present(os) && spa_writeable(dmu_objset_spa(os))); } boolean_t dmu_objset_projectquota_upgradable(objset_t *os) { return (dmu_objset_type(os) == DMU_OST_ZFS && !dmu_objset_is_snapshot(os) && dmu_objset_projectquota_enabled(os) && !dmu_objset_projectquota_present(os) && spa_writeable(dmu_objset_spa(os))); } void dmu_objset_space(objset_t *os, uint64_t *refdbytesp, uint64_t *availbytesp, uint64_t *usedobjsp, uint64_t *availobjsp) { dsl_dataset_space(os->os_dsl_dataset, refdbytesp, availbytesp, usedobjsp, availobjsp); } uint64_t dmu_objset_fsid_guid(objset_t *os) { return (dsl_dataset_fsid_guid(os->os_dsl_dataset)); } void dmu_objset_fast_stat(objset_t *os, dmu_objset_stats_t *stat) { stat->dds_type = os->os_phys->os_type; if (os->os_dsl_dataset) dsl_dataset_fast_stat(os->os_dsl_dataset, stat); } void dmu_objset_stats(objset_t *os, nvlist_t *nv) { ASSERT(os->os_dsl_dataset || os->os_phys->os_type == DMU_OST_META); if (os->os_dsl_dataset != NULL) dsl_dataset_stats(os->os_dsl_dataset, nv); dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_TYPE, os->os_phys->os_type); dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_USERACCOUNTING, dmu_objset_userspace_present(os)); } int dmu_objset_is_snapshot(objset_t *os) { if (os->os_dsl_dataset != NULL) return (os->os_dsl_dataset->ds_is_snapshot); else return (B_FALSE); } int dmu_snapshot_realname(objset_t *os, const char *name, char *real, int maxlen, boolean_t *conflict) { dsl_dataset_t *ds = os->os_dsl_dataset; uint64_t ignored; if (dsl_dataset_phys(ds)->ds_snapnames_zapobj == 0) return (SET_ERROR(ENOENT)); return (zap_lookup_norm(ds->ds_dir->dd_pool->dp_meta_objset, dsl_dataset_phys(ds)->ds_snapnames_zapobj, name, 8, 1, &ignored, MT_NORMALIZE, real, maxlen, conflict)); } int dmu_snapshot_list_next(objset_t *os, int namelen, char *name, uint64_t *idp, uint64_t *offp, boolean_t *case_conflict) { dsl_dataset_t *ds = os->os_dsl_dataset; zap_cursor_t cursor; zap_attribute_t attr; ASSERT(dsl_pool_config_held(dmu_objset_pool(os))); if (dsl_dataset_phys(ds)->ds_snapnames_zapobj == 0) return (SET_ERROR(ENOENT)); zap_cursor_init_serialized(&cursor, ds->ds_dir->dd_pool->dp_meta_objset, dsl_dataset_phys(ds)->ds_snapnames_zapobj, *offp); if (zap_cursor_retrieve(&cursor, &attr) != 0) { zap_cursor_fini(&cursor); return (SET_ERROR(ENOENT)); } if (strlen(attr.za_name) + 1 > namelen) { zap_cursor_fini(&cursor); return (SET_ERROR(ENAMETOOLONG)); } (void) strlcpy(name, attr.za_name, namelen); if (idp) *idp = attr.za_first_integer; if (case_conflict) *case_conflict = attr.za_normalization_conflict; zap_cursor_advance(&cursor); *offp = zap_cursor_serialize(&cursor); zap_cursor_fini(&cursor); return (0); } int dmu_snapshot_lookup(objset_t *os, const char *name, uint64_t *value) { return (dsl_dataset_snap_lookup(os->os_dsl_dataset, name, value)); } int dmu_dir_list_next(objset_t *os, int namelen, char *name, uint64_t *idp, uint64_t *offp) { dsl_dir_t *dd = os->os_dsl_dataset->ds_dir; zap_cursor_t cursor; zap_attribute_t attr; /* there is no next dir on a snapshot! */ if (os->os_dsl_dataset->ds_object != dsl_dir_phys(dd)->dd_head_dataset_obj) return (SET_ERROR(ENOENT)); zap_cursor_init_serialized(&cursor, dd->dd_pool->dp_meta_objset, dsl_dir_phys(dd)->dd_child_dir_zapobj, *offp); if (zap_cursor_retrieve(&cursor, &attr) != 0) { zap_cursor_fini(&cursor); return (SET_ERROR(ENOENT)); } if (strlen(attr.za_name) + 1 > namelen) { zap_cursor_fini(&cursor); return (SET_ERROR(ENAMETOOLONG)); } (void) strlcpy(name, attr.za_name, namelen); if (idp) *idp = attr.za_first_integer; zap_cursor_advance(&cursor); *offp = zap_cursor_serialize(&cursor); zap_cursor_fini(&cursor); return (0); } typedef struct dmu_objset_find_ctx { taskq_t *dc_tq; dsl_pool_t *dc_dp; uint64_t dc_ddobj; char *dc_ddname; /* last component of ddobj's name */ int (*dc_func)(dsl_pool_t *, dsl_dataset_t *, void *); void *dc_arg; int dc_flags; kmutex_t *dc_error_lock; int *dc_error; } dmu_objset_find_ctx_t; static void dmu_objset_find_dp_impl(dmu_objset_find_ctx_t *dcp) { dsl_pool_t *dp = dcp->dc_dp; dsl_dir_t *dd; dsl_dataset_t *ds; zap_cursor_t zc; zap_attribute_t *attr; uint64_t thisobj; int err = 0; /* don't process if there already was an error */ if (*dcp->dc_error != 0) goto out; /* * Note: passing the name (dc_ddname) here is optional, but it * improves performance because we don't need to call * zap_value_search() to determine the name. */ err = dsl_dir_hold_obj(dp, dcp->dc_ddobj, dcp->dc_ddname, FTAG, &dd); if (err != 0) goto out; /* Don't visit hidden ($MOS & $ORIGIN) objsets. */ if (dd->dd_myname[0] == '$') { dsl_dir_rele(dd, FTAG); goto out; } thisobj = dsl_dir_phys(dd)->dd_head_dataset_obj; attr = kmem_alloc(sizeof (zap_attribute_t), KM_SLEEP); /* * Iterate over all children. */ if (dcp->dc_flags & DS_FIND_CHILDREN) { for (zap_cursor_init(&zc, dp->dp_meta_objset, dsl_dir_phys(dd)->dd_child_dir_zapobj); zap_cursor_retrieve(&zc, attr) == 0; (void) zap_cursor_advance(&zc)) { ASSERT3U(attr->za_integer_length, ==, sizeof (uint64_t)); ASSERT3U(attr->za_num_integers, ==, 1); dmu_objset_find_ctx_t *child_dcp = kmem_alloc(sizeof (*child_dcp), KM_SLEEP); *child_dcp = *dcp; child_dcp->dc_ddobj = attr->za_first_integer; child_dcp->dc_ddname = spa_strdup(attr->za_name); if (dcp->dc_tq != NULL) (void) taskq_dispatch(dcp->dc_tq, dmu_objset_find_dp_cb, child_dcp, TQ_SLEEP); else dmu_objset_find_dp_impl(child_dcp); } zap_cursor_fini(&zc); } /* * Iterate over all snapshots. */ if (dcp->dc_flags & DS_FIND_SNAPSHOTS) { dsl_dataset_t *ds; err = dsl_dataset_hold_obj(dp, thisobj, FTAG, &ds); if (err == 0) { uint64_t snapobj; snapobj = dsl_dataset_phys(ds)->ds_snapnames_zapobj; dsl_dataset_rele(ds, FTAG); for (zap_cursor_init(&zc, dp->dp_meta_objset, snapobj); zap_cursor_retrieve(&zc, attr) == 0; (void) zap_cursor_advance(&zc)) { ASSERT3U(attr->za_integer_length, ==, sizeof (uint64_t)); ASSERT3U(attr->za_num_integers, ==, 1); err = dsl_dataset_hold_obj(dp, attr->za_first_integer, FTAG, &ds); if (err != 0) break; err = dcp->dc_func(dp, ds, dcp->dc_arg); dsl_dataset_rele(ds, FTAG); if (err != 0) break; } zap_cursor_fini(&zc); } } kmem_free(attr, sizeof (zap_attribute_t)); if (err != 0) { dsl_dir_rele(dd, FTAG); goto out; } /* * Apply to self. */ err = dsl_dataset_hold_obj(dp, thisobj, FTAG, &ds); /* * Note: we hold the dir while calling dsl_dataset_hold_obj() so * that the dir will remain cached, and we won't have to re-instantiate * it (which could be expensive due to finding its name via * zap_value_search()). */ dsl_dir_rele(dd, FTAG); if (err != 0) goto out; err = dcp->dc_func(dp, ds, dcp->dc_arg); dsl_dataset_rele(ds, FTAG); out: if (err != 0) { mutex_enter(dcp->dc_error_lock); /* only keep first error */ if (*dcp->dc_error == 0) *dcp->dc_error = err; mutex_exit(dcp->dc_error_lock); } if (dcp->dc_ddname != NULL) spa_strfree(dcp->dc_ddname); kmem_free(dcp, sizeof (*dcp)); } static void dmu_objset_find_dp_cb(void *arg) { dmu_objset_find_ctx_t *dcp = arg; dsl_pool_t *dp = dcp->dc_dp; /* * We need to get a pool_config_lock here, as there are several * assert(pool_config_held) down the stack. Getting a lock via * dsl_pool_config_enter is risky, as it might be stalled by a * pending writer. This would deadlock, as the write lock can * only be granted when our parent thread gives up the lock. * The _prio interface gives us priority over a pending writer. */ dsl_pool_config_enter_prio(dp, FTAG); dmu_objset_find_dp_impl(dcp); dsl_pool_config_exit(dp, FTAG); } /* * Find objsets under and including ddobj, call func(ds) on each. * The order for the enumeration is completely undefined. * func is called with dsl_pool_config held. */ int dmu_objset_find_dp(dsl_pool_t *dp, uint64_t ddobj, int func(dsl_pool_t *, dsl_dataset_t *, void *), void *arg, int flags) { int error = 0; taskq_t *tq = NULL; int ntasks; dmu_objset_find_ctx_t *dcp; kmutex_t err_lock; mutex_init(&err_lock, NULL, MUTEX_DEFAULT, NULL); dcp = kmem_alloc(sizeof (*dcp), KM_SLEEP); dcp->dc_tq = NULL; dcp->dc_dp = dp; dcp->dc_ddobj = ddobj; dcp->dc_ddname = NULL; dcp->dc_func = func; dcp->dc_arg = arg; dcp->dc_flags = flags; dcp->dc_error_lock = &err_lock; dcp->dc_error = &error; if ((flags & DS_FIND_SERIALIZE) || dsl_pool_config_held_writer(dp)) { /* * In case a write lock is held we can't make use of * parallelism, as down the stack of the worker threads * the lock is asserted via dsl_pool_config_held. * In case of a read lock this is solved by getting a read * lock in each worker thread, which isn't possible in case * of a writer lock. So we fall back to the synchronous path * here. * In the future it might be possible to get some magic into * dsl_pool_config_held in a way that it returns true for * the worker threads so that a single lock held from this * thread suffices. For now, stay single threaded. */ dmu_objset_find_dp_impl(dcp); mutex_destroy(&err_lock); return (error); } ntasks = dmu_find_threads; if (ntasks == 0) ntasks = vdev_count_leaves(dp->dp_spa) * 4; tq = taskq_create("dmu_objset_find", ntasks, maxclsyspri, ntasks, INT_MAX, 0); if (tq == NULL) { kmem_free(dcp, sizeof (*dcp)); mutex_destroy(&err_lock); return (SET_ERROR(ENOMEM)); } dcp->dc_tq = tq; /* dcp will be freed by task */ (void) taskq_dispatch(tq, dmu_objset_find_dp_cb, dcp, TQ_SLEEP); /* * PORTING: this code relies on the property of taskq_wait to wait * until no more tasks are queued and no more tasks are active. As * we always queue new tasks from within other tasks, task_wait * reliably waits for the full recursion to finish, even though we * enqueue new tasks after taskq_wait has been called. * On platforms other than illumos, taskq_wait may not have this * property. */ taskq_wait(tq); taskq_destroy(tq); mutex_destroy(&err_lock); return (error); } /* * Find all objsets under name, and for each, call 'func(child_name, arg)'. * The dp_config_rwlock must not be held when this is called, and it * will not be held when the callback is called. * Therefore this function should only be used when the pool is not changing * (e.g. in syncing context), or the callback can deal with the possible races. */ static int dmu_objset_find_impl(spa_t *spa, const char *name, int func(const char *, void *), void *arg, int flags) { dsl_dir_t *dd; dsl_pool_t *dp = spa_get_dsl(spa); dsl_dataset_t *ds; zap_cursor_t zc; zap_attribute_t *attr; char *child; uint64_t thisobj; int err; dsl_pool_config_enter(dp, FTAG); err = dsl_dir_hold(dp, name, FTAG, &dd, NULL); if (err != 0) { dsl_pool_config_exit(dp, FTAG); return (err); } /* Don't visit hidden ($MOS & $ORIGIN) objsets. */ if (dd->dd_myname[0] == '$') { dsl_dir_rele(dd, FTAG); dsl_pool_config_exit(dp, FTAG); return (0); } thisobj = dsl_dir_phys(dd)->dd_head_dataset_obj; attr = kmem_alloc(sizeof (zap_attribute_t), KM_SLEEP); /* * Iterate over all children. */ if (flags & DS_FIND_CHILDREN) { for (zap_cursor_init(&zc, dp->dp_meta_objset, dsl_dir_phys(dd)->dd_child_dir_zapobj); zap_cursor_retrieve(&zc, attr) == 0; (void) zap_cursor_advance(&zc)) { ASSERT3U(attr->za_integer_length, ==, sizeof (uint64_t)); ASSERT3U(attr->za_num_integers, ==, 1); child = kmem_asprintf("%s/%s", name, attr->za_name); dsl_pool_config_exit(dp, FTAG); err = dmu_objset_find_impl(spa, child, func, arg, flags); dsl_pool_config_enter(dp, FTAG); kmem_strfree(child); if (err != 0) break; } zap_cursor_fini(&zc); if (err != 0) { dsl_dir_rele(dd, FTAG); dsl_pool_config_exit(dp, FTAG); kmem_free(attr, sizeof (zap_attribute_t)); return (err); } } /* * Iterate over all snapshots. */ if (flags & DS_FIND_SNAPSHOTS) { err = dsl_dataset_hold_obj(dp, thisobj, FTAG, &ds); if (err == 0) { uint64_t snapobj; snapobj = dsl_dataset_phys(ds)->ds_snapnames_zapobj; dsl_dataset_rele(ds, FTAG); for (zap_cursor_init(&zc, dp->dp_meta_objset, snapobj); zap_cursor_retrieve(&zc, attr) == 0; (void) zap_cursor_advance(&zc)) { ASSERT3U(attr->za_integer_length, ==, sizeof (uint64_t)); ASSERT3U(attr->za_num_integers, ==, 1); child = kmem_asprintf("%s@%s", name, attr->za_name); dsl_pool_config_exit(dp, FTAG); err = func(child, arg); dsl_pool_config_enter(dp, FTAG); kmem_strfree(child); if (err != 0) break; } zap_cursor_fini(&zc); } } dsl_dir_rele(dd, FTAG); kmem_free(attr, sizeof (zap_attribute_t)); dsl_pool_config_exit(dp, FTAG); if (err != 0) return (err); /* Apply to self. */ return (func(name, arg)); } /* * See comment above dmu_objset_find_impl(). */ int dmu_objset_find(const char *name, int func(const char *, void *), void *arg, int flags) { spa_t *spa; int error; error = spa_open(name, &spa, FTAG); if (error != 0) return (error); error = dmu_objset_find_impl(spa, name, func, arg, flags); spa_close(spa, FTAG); return (error); } boolean_t dmu_objset_incompatible_encryption_version(objset_t *os) { return (dsl_dir_incompatible_encryption_version( os->os_dsl_dataset->ds_dir)); } void dmu_objset_set_user(objset_t *os, void *user_ptr) { ASSERT(MUTEX_HELD(&os->os_user_ptr_lock)); os->os_user_ptr = user_ptr; } void * dmu_objset_get_user(objset_t *os) { ASSERT(MUTEX_HELD(&os->os_user_ptr_lock)); return (os->os_user_ptr); } /* * Determine name of filesystem, given name of snapshot. * buf must be at least ZFS_MAX_DATASET_NAME_LEN bytes */ int dmu_fsname(const char *snapname, char *buf) { char *atp = strchr(snapname, '@'); if (atp == NULL) return (SET_ERROR(EINVAL)); if (atp - snapname >= ZFS_MAX_DATASET_NAME_LEN) return (SET_ERROR(ENAMETOOLONG)); (void) strlcpy(buf, snapname, atp - snapname + 1); return (0); } /* * Call when we think we're going to write/free space in open context * to track the amount of dirty data in the open txg, which is also the * amount of memory that can not be evicted until this txg syncs. * * Note that there are two conditions where this can be called from * syncing context: * * [1] When we just created the dataset, in which case we go on with * updating any accounting of dirty data as usual. * [2] When we are dirtying MOS data, in which case we only update the * pool's accounting of dirty data. */ void dmu_objset_willuse_space(objset_t *os, int64_t space, dmu_tx_t *tx) { dsl_dataset_t *ds = os->os_dsl_dataset; int64_t aspace = spa_get_worst_case_asize(os->os_spa, space); if (ds != NULL) { dsl_dir_willuse_space(ds->ds_dir, aspace, tx); } dsl_pool_dirty_space(dmu_tx_pool(tx), space, tx); } #if defined(_KERNEL) EXPORT_SYMBOL(dmu_objset_zil); EXPORT_SYMBOL(dmu_objset_pool); EXPORT_SYMBOL(dmu_objset_ds); EXPORT_SYMBOL(dmu_objset_type); EXPORT_SYMBOL(dmu_objset_name); EXPORT_SYMBOL(dmu_objset_hold); EXPORT_SYMBOL(dmu_objset_hold_flags); EXPORT_SYMBOL(dmu_objset_own); EXPORT_SYMBOL(dmu_objset_rele); EXPORT_SYMBOL(dmu_objset_rele_flags); EXPORT_SYMBOL(dmu_objset_disown); EXPORT_SYMBOL(dmu_objset_from_ds); EXPORT_SYMBOL(dmu_objset_create); EXPORT_SYMBOL(dmu_objset_clone); EXPORT_SYMBOL(dmu_objset_stats); EXPORT_SYMBOL(dmu_objset_fast_stat); EXPORT_SYMBOL(dmu_objset_spa); EXPORT_SYMBOL(dmu_objset_space); EXPORT_SYMBOL(dmu_objset_fsid_guid); EXPORT_SYMBOL(dmu_objset_find); EXPORT_SYMBOL(dmu_objset_byteswap); EXPORT_SYMBOL(dmu_objset_evict_dbufs); EXPORT_SYMBOL(dmu_objset_snap_cmtime); EXPORT_SYMBOL(dmu_objset_dnodesize); EXPORT_SYMBOL(dmu_objset_sync); EXPORT_SYMBOL(dmu_objset_is_dirty); EXPORT_SYMBOL(dmu_objset_create_impl_dnstats); EXPORT_SYMBOL(dmu_objset_create_impl); EXPORT_SYMBOL(dmu_objset_open_impl); EXPORT_SYMBOL(dmu_objset_evict); EXPORT_SYMBOL(dmu_objset_register_type); EXPORT_SYMBOL(dmu_objset_sync_done); EXPORT_SYMBOL(dmu_objset_userquota_get_ids); EXPORT_SYMBOL(dmu_objset_userused_enabled); EXPORT_SYMBOL(dmu_objset_userspace_upgrade); EXPORT_SYMBOL(dmu_objset_userspace_present); EXPORT_SYMBOL(dmu_objset_userobjused_enabled); EXPORT_SYMBOL(dmu_objset_userobjspace_upgradable); EXPORT_SYMBOL(dmu_objset_userobjspace_present); EXPORT_SYMBOL(dmu_objset_projectquota_enabled); EXPORT_SYMBOL(dmu_objset_projectquota_present); EXPORT_SYMBOL(dmu_objset_projectquota_upgradable); EXPORT_SYMBOL(dmu_objset_id_quota_upgrade); #endif diff --git a/module/zfs/dmu_zfetch.c b/module/zfs/dmu_zfetch.c index 3439f9bddf4e..ed50f1889b59 100644 --- a/module/zfs/dmu_zfetch.c +++ b/module/zfs/dmu_zfetch.c @@ -1,768 +1,773 @@ /* * 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 2009 Sun Microsystems, Inc. All rights reserved. * Use is subject to license terms. */ /* * Copyright (c) 2013, 2017 by Delphix. All rights reserved. */ #include #include #include #include #include #include #include #include #include /* * This tunable disables predictive prefetch. Note that it leaves "prescient" * prefetch (e.g. prefetch for zfs send) intact. Unlike predictive prefetch, * prescient prefetch never issues i/os that end up not being needed, * so it can't hurt performance. */ static int zfs_prefetch_disable = B_FALSE; /* max # of streams per zfetch */ static unsigned int zfetch_max_streams = 8; /* min time before stream reclaim */ static unsigned int zfetch_min_sec_reap = 1; /* max time before stream delete */ static unsigned int zfetch_max_sec_reap = 2; #ifdef _ILP32 /* min bytes to prefetch per stream (default 2MB) */ static unsigned int zfetch_min_distance = 2 * 1024 * 1024; /* max bytes to prefetch per stream (default 8MB) */ unsigned int zfetch_max_distance = 8 * 1024 * 1024; #else /* min bytes to prefetch per stream (default 4MB) */ static unsigned int zfetch_min_distance = 4 * 1024 * 1024; /* max bytes to prefetch per stream (default 64MB) */ unsigned int zfetch_max_distance = 64 * 1024 * 1024; #endif /* max bytes to prefetch indirects for per stream (default 64MB) */ unsigned int zfetch_max_idistance = 64 * 1024 * 1024; /* max request reorder distance within a stream (default 16MB) */ unsigned int zfetch_max_reorder = 16 * 1024 * 1024; /* Max log2 fraction of holes in a stream */ unsigned int zfetch_hole_shift = 2; typedef struct zfetch_stats { kstat_named_t zfetchstat_hits; kstat_named_t zfetchstat_future; kstat_named_t zfetchstat_stride; kstat_named_t zfetchstat_past; kstat_named_t zfetchstat_misses; kstat_named_t zfetchstat_max_streams; kstat_named_t zfetchstat_io_issued; kstat_named_t zfetchstat_io_active; } zfetch_stats_t; static zfetch_stats_t zfetch_stats = { { "hits", KSTAT_DATA_UINT64 }, { "future", KSTAT_DATA_UINT64 }, { "stride", KSTAT_DATA_UINT64 }, { "past", KSTAT_DATA_UINT64 }, { "misses", KSTAT_DATA_UINT64 }, { "max_streams", KSTAT_DATA_UINT64 }, { "io_issued", KSTAT_DATA_UINT64 }, { "io_active", KSTAT_DATA_UINT64 }, }; struct { wmsum_t zfetchstat_hits; wmsum_t zfetchstat_future; wmsum_t zfetchstat_stride; wmsum_t zfetchstat_past; wmsum_t zfetchstat_misses; wmsum_t zfetchstat_max_streams; wmsum_t zfetchstat_io_issued; aggsum_t zfetchstat_io_active; } zfetch_sums; #define ZFETCHSTAT_BUMP(stat) \ wmsum_add(&zfetch_sums.stat, 1) #define ZFETCHSTAT_ADD(stat, val) \ wmsum_add(&zfetch_sums.stat, val) static kstat_t *zfetch_ksp; static int zfetch_kstats_update(kstat_t *ksp, int rw) { zfetch_stats_t *zs = ksp->ks_data; if (rw == KSTAT_WRITE) return (EACCES); zs->zfetchstat_hits.value.ui64 = wmsum_value(&zfetch_sums.zfetchstat_hits); zs->zfetchstat_future.value.ui64 = wmsum_value(&zfetch_sums.zfetchstat_future); zs->zfetchstat_stride.value.ui64 = wmsum_value(&zfetch_sums.zfetchstat_stride); zs->zfetchstat_past.value.ui64 = wmsum_value(&zfetch_sums.zfetchstat_past); zs->zfetchstat_misses.value.ui64 = wmsum_value(&zfetch_sums.zfetchstat_misses); zs->zfetchstat_max_streams.value.ui64 = wmsum_value(&zfetch_sums.zfetchstat_max_streams); zs->zfetchstat_io_issued.value.ui64 = wmsum_value(&zfetch_sums.zfetchstat_io_issued); zs->zfetchstat_io_active.value.ui64 = aggsum_value(&zfetch_sums.zfetchstat_io_active); return (0); } void zfetch_init(void) { wmsum_init(&zfetch_sums.zfetchstat_hits, 0); wmsum_init(&zfetch_sums.zfetchstat_future, 0); wmsum_init(&zfetch_sums.zfetchstat_stride, 0); wmsum_init(&zfetch_sums.zfetchstat_past, 0); wmsum_init(&zfetch_sums.zfetchstat_misses, 0); wmsum_init(&zfetch_sums.zfetchstat_max_streams, 0); wmsum_init(&zfetch_sums.zfetchstat_io_issued, 0); aggsum_init(&zfetch_sums.zfetchstat_io_active, 0); zfetch_ksp = kstat_create("zfs", 0, "zfetchstats", "misc", KSTAT_TYPE_NAMED, sizeof (zfetch_stats) / sizeof (kstat_named_t), KSTAT_FLAG_VIRTUAL); if (zfetch_ksp != NULL) { zfetch_ksp->ks_data = &zfetch_stats; zfetch_ksp->ks_update = zfetch_kstats_update; kstat_install(zfetch_ksp); } } void zfetch_fini(void) { if (zfetch_ksp != NULL) { kstat_delete(zfetch_ksp); zfetch_ksp = NULL; } wmsum_fini(&zfetch_sums.zfetchstat_hits); wmsum_fini(&zfetch_sums.zfetchstat_future); wmsum_fini(&zfetch_sums.zfetchstat_stride); wmsum_fini(&zfetch_sums.zfetchstat_past); wmsum_fini(&zfetch_sums.zfetchstat_misses); wmsum_fini(&zfetch_sums.zfetchstat_max_streams); wmsum_fini(&zfetch_sums.zfetchstat_io_issued); ASSERT0(aggsum_value(&zfetch_sums.zfetchstat_io_active)); aggsum_fini(&zfetch_sums.zfetchstat_io_active); } /* * This takes a pointer to a zfetch structure and a dnode. It performs the * necessary setup for the zfetch structure, grokking data from the * associated dnode. */ void dmu_zfetch_init(zfetch_t *zf, dnode_t *dno) { if (zf == NULL) return; zf->zf_dnode = dno; zf->zf_numstreams = 0; list_create(&zf->zf_stream, sizeof (zstream_t), offsetof(zstream_t, zs_node)); mutex_init(&zf->zf_lock, NULL, MUTEX_DEFAULT, NULL); } static void dmu_zfetch_stream_fini(zstream_t *zs) { ASSERT(!list_link_active(&zs->zs_node)); zfs_refcount_destroy(&zs->zs_callers); zfs_refcount_destroy(&zs->zs_refs); kmem_free(zs, sizeof (*zs)); } static void dmu_zfetch_stream_remove(zfetch_t *zf, zstream_t *zs) { ASSERT(MUTEX_HELD(&zf->zf_lock)); list_remove(&zf->zf_stream, zs); zf->zf_numstreams--; membar_producer(); if (zfs_refcount_remove(&zs->zs_refs, NULL) == 0) dmu_zfetch_stream_fini(zs); } /* * Clean-up state associated with a zfetch structure (e.g. destroy the * streams). This doesn't free the zfetch_t itself, that's left to the caller. */ void dmu_zfetch_fini(zfetch_t *zf) { zstream_t *zs; mutex_enter(&zf->zf_lock); while ((zs = list_head(&zf->zf_stream)) != NULL) dmu_zfetch_stream_remove(zf, zs); mutex_exit(&zf->zf_lock); list_destroy(&zf->zf_stream); mutex_destroy(&zf->zf_lock); zf->zf_dnode = NULL; } /* * If there aren't too many active streams already, create one more. * In process delete/reuse all streams without hits for zfetch_max_sec_reap. * If needed, reuse oldest stream without hits for zfetch_min_sec_reap or ever. * The "blkid" argument is the next block that we expect this stream to access. */ static void dmu_zfetch_stream_create(zfetch_t *zf, uint64_t blkid) { zstream_t *zs, *zs_next, *zs_old = NULL; uint_t now = gethrestime_sec(), t; ASSERT(MUTEX_HELD(&zf->zf_lock)); /* * Delete too old streams, reusing the first found one. */ t = now - zfetch_max_sec_reap; for (zs = list_head(&zf->zf_stream); zs != NULL; zs = zs_next) { zs_next = list_next(&zf->zf_stream, zs); /* * Skip if still active. 1 -- zf_stream reference. */ if ((int)(zs->zs_atime - t) >= 0) continue; if (zfs_refcount_count(&zs->zs_refs) != 1) continue; if (zs_old) dmu_zfetch_stream_remove(zf, zs); else zs_old = zs; } if (zs_old) { zs = zs_old; list_remove(&zf->zf_stream, zs); goto reuse; } /* * The maximum number of streams is normally zfetch_max_streams, * but for small files we lower it such that it's at least possible * for all the streams to be non-overlapping. */ uint32_t max_streams = MAX(1, MIN(zfetch_max_streams, (zf->zf_dnode->dn_maxblkid << zf->zf_dnode->dn_datablkshift) / zfetch_max_distance)); if (zf->zf_numstreams >= max_streams) { t = now - zfetch_min_sec_reap; for (zs = list_head(&zf->zf_stream); zs != NULL; zs = list_next(&zf->zf_stream, zs)) { if ((int)(zs->zs_atime - t) >= 0) continue; if (zfs_refcount_count(&zs->zs_refs) != 1) continue; if (zs_old == NULL || (int)(zs_old->zs_atime - zs->zs_atime) >= 0) zs_old = zs; } if (zs_old) { zs = zs_old; list_remove(&zf->zf_stream, zs); goto reuse; } ZFETCHSTAT_BUMP(zfetchstat_max_streams); return; } zs = kmem_zalloc(sizeof (*zs), KM_SLEEP); zfs_refcount_create(&zs->zs_callers); zfs_refcount_create(&zs->zs_refs); /* One reference for zf_stream. */ zfs_refcount_add(&zs->zs_refs, NULL); zf->zf_numstreams++; reuse: list_insert_head(&zf->zf_stream, zs); zs->zs_blkid = blkid; /* Allow immediate stream reuse until first hit. */ zs->zs_atime = now - zfetch_min_sec_reap; memset(zs->zs_ranges, 0, sizeof (zs->zs_ranges)); zs->zs_pf_dist = 0; zs->zs_ipf_dist = 0; zs->zs_pf_start = blkid; zs->zs_pf_end = blkid; zs->zs_ipf_start = blkid; zs->zs_ipf_end = blkid; zs->zs_missed = B_FALSE; zs->zs_more = B_FALSE; } static void dmu_zfetch_done(void *arg, uint64_t level, uint64_t blkid, boolean_t io_issued) { zstream_t *zs = arg; if (io_issued && level == 0 && blkid < zs->zs_blkid) zs->zs_more = B_TRUE; if (zfs_refcount_remove(&zs->zs_refs, NULL) == 0) dmu_zfetch_stream_fini(zs); aggsum_add(&zfetch_sums.zfetchstat_io_active, -1); } /* * Process stream hit access for nblks blocks starting at zs_blkid. Return * number of blocks to proceed for after aggregation with future ranges. */ static uint64_t dmu_zfetch_hit(zstream_t *zs, uint64_t nblks) { uint_t i, j; /* Optimize sequential accesses (no future ranges). */ if (zs->zs_ranges[0].start == 0) goto done; /* Look for intersections with further ranges. */ for (i = 0; i < ZFETCH_RANGES; i++) { zsrange_t *r = &zs->zs_ranges[i]; if (r->start == 0 || r->start > nblks) break; if (r->end >= nblks) { nblks = r->end; i++; break; } } /* Delete all found intersecting ranges, updates remaining. */ for (j = 0; i < ZFETCH_RANGES; i++, j++) { if (zs->zs_ranges[i].start == 0) break; ASSERT3U(zs->zs_ranges[i].start, >, nblks); ASSERT3U(zs->zs_ranges[i].end, >, nblks); zs->zs_ranges[j].start = zs->zs_ranges[i].start - nblks; zs->zs_ranges[j].end = zs->zs_ranges[i].end - nblks; } if (j < ZFETCH_RANGES) { zs->zs_ranges[j].start = 0; zs->zs_ranges[j].end = 0; } done: zs->zs_blkid += nblks; return (nblks); } /* * Process future stream access for nblks blocks starting at blkid. Return * number of blocks to proceed for if future ranges reach fill threshold. */ static uint64_t dmu_zfetch_future(zstream_t *zs, uint64_t blkid, uint64_t nblks) { ASSERT3U(blkid, >, zs->zs_blkid); blkid -= zs->zs_blkid; ASSERT3U(blkid + nblks, <=, UINT16_MAX); /* Search for first and last intersection or insert point. */ uint_t f = ZFETCH_RANGES, l = 0, i; for (i = 0; i < ZFETCH_RANGES; i++) { zsrange_t *r = &zs->zs_ranges[i]; if (r->start == 0 || r->start > blkid + nblks) break; if (r->end < blkid) continue; if (f > i) f = i; if (l < i) l = i; } if (f <= l) { /* Got some intersecting range, expand it if needed. */ if (zs->zs_ranges[f].start > blkid) zs->zs_ranges[f].start = blkid; zs->zs_ranges[f].end = MAX(zs->zs_ranges[l].end, blkid + nblks); if (f < l) { /* Got more than one intersection, remove others. */ for (f++, l++; l < ZFETCH_RANGES; f++, l++) { zs->zs_ranges[f].start = zs->zs_ranges[l].start; zs->zs_ranges[f].end = zs->zs_ranges[l].end; } zs->zs_ranges[f].start = 0; zs->zs_ranges[f].end = 0; } } else if (i < ZFETCH_RANGES) { /* Got no intersecting ranges, insert new one. */ for (l = ZFETCH_RANGES - 1; l > i; l--) { zs->zs_ranges[l].start = zs->zs_ranges[l - 1].start; zs->zs_ranges[l].end = zs->zs_ranges[l - 1].end; } zs->zs_ranges[i].start = blkid; zs->zs_ranges[i].end = blkid + nblks; } else { /* No space left to insert. Drop the range. */ return (0); } /* Check if with the new access addition we reached fill threshold. */ if (zfetch_hole_shift >= 16) return (0); uint_t hole = 0; for (i = f = l = 0; i < ZFETCH_RANGES; i++) { zsrange_t *r = &zs->zs_ranges[i]; if (r->start == 0) break; hole += r->start - f; f = r->end; if (hole <= r->end >> zfetch_hole_shift) l = r->end; } if (l > 0) return (dmu_zfetch_hit(zs, l)); return (0); } /* * This is the predictive prefetch entry point. dmu_zfetch_prepare() * associates dnode access specified with blkid and nblks arguments with * prefetch stream, predicts further accesses based on that stats and returns * the stream pointer on success. That pointer must later be passed to * dmu_zfetch_run() to initiate the speculative prefetch for the stream and * release it. dmu_zfetch() is a wrapper for simple cases when window between * prediction and prefetch initiation is not needed. * fetch_data argument specifies whether actual data blocks should be fetched: * FALSE -- prefetch only indirect blocks for predicted data blocks; * TRUE -- prefetch predicted data blocks plus following indirect blocks. */ zstream_t * dmu_zfetch_prepare(zfetch_t *zf, uint64_t blkid, uint64_t nblks, boolean_t fetch_data, boolean_t have_lock) { zstream_t *zs; spa_t *spa = zf->zf_dnode->dn_objset->os_spa; + zfs_prefetch_type_t os_prefetch = zf->zf_dnode->dn_objset->os_prefetch; - if (zfs_prefetch_disable) + if (zfs_prefetch_disable || os_prefetch == ZFS_PREFETCH_NONE) return (NULL); + + if (os_prefetch == ZFS_PREFETCH_METADATA) + fetch_data = B_FALSE; + /* * If we haven't yet loaded the indirect vdevs' mappings, we * can only read from blocks that we carefully ensure are on * concrete vdevs (or previously-loaded indirect vdevs). So we * can't allow the predictive prefetcher to attempt reads of other * blocks (e.g. of the MOS's dnode object). */ if (!spa_indirect_vdevs_loaded(spa)) return (NULL); /* * As a fast path for small (single-block) files, ignore access * to the first block. */ if (!have_lock && blkid == 0) return (NULL); if (!have_lock) rw_enter(&zf->zf_dnode->dn_struct_rwlock, RW_READER); /* * A fast path for small files for which no prefetch will * happen. */ uint64_t maxblkid = zf->zf_dnode->dn_maxblkid; if (maxblkid < 2) { if (!have_lock) rw_exit(&zf->zf_dnode->dn_struct_rwlock); return (NULL); } mutex_enter(&zf->zf_lock); /* * Find perfect prefetch stream. Depending on whether the accesses * are block-aligned, first block of the new access may either follow * the last block of the previous access, or be equal to it. */ unsigned int dbs = zf->zf_dnode->dn_datablkshift; uint64_t end_blkid = blkid + nblks; for (zs = list_head(&zf->zf_stream); zs != NULL; zs = list_next(&zf->zf_stream, zs)) { if (blkid == zs->zs_blkid) { goto hit; } else if (blkid + 1 == zs->zs_blkid) { blkid++; nblks--; goto hit; } } /* * Find close enough prefetch stream. Access crossing stream position * is a hit in its new part. Access ahead of stream position considered * a hit for metadata prefetch, since we do not care about fill percent, * or stored for future otherwise. Access behind stream position is * silently ignored, since we already skipped it reaching fill percent. */ uint_t max_reorder = MIN((zfetch_max_reorder >> dbs) + 1, UINT16_MAX); uint_t t = gethrestime_sec() - zfetch_max_sec_reap; for (zs = list_head(&zf->zf_stream); zs != NULL; zs = list_next(&zf->zf_stream, zs)) { if (blkid > zs->zs_blkid) { if (end_blkid <= zs->zs_blkid + max_reorder) { if (!fetch_data) { nblks = dmu_zfetch_hit(zs, end_blkid - zs->zs_blkid); ZFETCHSTAT_BUMP(zfetchstat_stride); goto future; } nblks = dmu_zfetch_future(zs, blkid, nblks); if (nblks > 0) ZFETCHSTAT_BUMP(zfetchstat_stride); else ZFETCHSTAT_BUMP(zfetchstat_future); goto future; } } else if (end_blkid >= zs->zs_blkid) { nblks -= zs->zs_blkid - blkid; blkid += zs->zs_blkid - blkid; goto hit; } else if (end_blkid + max_reorder > zs->zs_blkid && (int)(zs->zs_atime - t) >= 0) { ZFETCHSTAT_BUMP(zfetchstat_past); zs->zs_atime = gethrestime_sec(); goto out; } } /* * This access is not part of any existing stream. Create a new * stream for it unless we are at the end of file. */ if (end_blkid < maxblkid) dmu_zfetch_stream_create(zf, end_blkid); mutex_exit(&zf->zf_lock); if (!have_lock) rw_exit(&zf->zf_dnode->dn_struct_rwlock); ZFETCHSTAT_BUMP(zfetchstat_misses); return (NULL); hit: nblks = dmu_zfetch_hit(zs, nblks); ZFETCHSTAT_BUMP(zfetchstat_hits); future: zs->zs_atime = gethrestime_sec(); /* Exit if we already prefetched for this position before. */ if (nblks == 0) goto out; /* If the file is ending, remove the stream. */ end_blkid = zs->zs_blkid; if (end_blkid >= maxblkid) { dmu_zfetch_stream_remove(zf, zs); out: mutex_exit(&zf->zf_lock); if (!have_lock) rw_exit(&zf->zf_dnode->dn_struct_rwlock); return (NULL); } /* * This access was to a block that we issued a prefetch for on * behalf of this stream. Calculate further prefetch distances. * * Start prefetch from the demand access size (nblks). Double the * distance every access up to zfetch_min_distance. After that only * if needed increase the distance by 1/8 up to zfetch_max_distance. * * Don't double the distance beyond single block if we have more * than ~6% of ARC held by active prefetches. It should help with * getting out of RAM on some badly mispredicted read patterns. */ unsigned int nbytes = nblks << dbs; unsigned int pf_nblks; if (fetch_data) { if (unlikely(zs->zs_pf_dist < nbytes)) zs->zs_pf_dist = nbytes; else if (zs->zs_pf_dist < zfetch_min_distance && (zs->zs_pf_dist < (1 << dbs) || aggsum_compare(&zfetch_sums.zfetchstat_io_active, arc_c_max >> (4 + dbs)) < 0)) zs->zs_pf_dist *= 2; else if (zs->zs_more) zs->zs_pf_dist += zs->zs_pf_dist / 8; zs->zs_more = B_FALSE; if (zs->zs_pf_dist > zfetch_max_distance) zs->zs_pf_dist = zfetch_max_distance; pf_nblks = zs->zs_pf_dist >> dbs; } else { pf_nblks = 0; } if (zs->zs_pf_start < end_blkid) zs->zs_pf_start = end_blkid; if (zs->zs_pf_end < end_blkid + pf_nblks) zs->zs_pf_end = end_blkid + pf_nblks; /* * Do the same for indirects, starting where we will stop reading * data blocks (and the indirects that point to them). */ if (unlikely(zs->zs_ipf_dist < nbytes)) zs->zs_ipf_dist = nbytes; else zs->zs_ipf_dist *= 2; if (zs->zs_ipf_dist > zfetch_max_idistance) zs->zs_ipf_dist = zfetch_max_idistance; pf_nblks = zs->zs_ipf_dist >> dbs; if (zs->zs_ipf_start < zs->zs_pf_end) zs->zs_ipf_start = zs->zs_pf_end; if (zs->zs_ipf_end < zs->zs_pf_end + pf_nblks) zs->zs_ipf_end = zs->zs_pf_end + pf_nblks; zfs_refcount_add(&zs->zs_refs, NULL); /* Count concurrent callers. */ zfs_refcount_add(&zs->zs_callers, NULL); mutex_exit(&zf->zf_lock); if (!have_lock) rw_exit(&zf->zf_dnode->dn_struct_rwlock); return (zs); } void dmu_zfetch_run(zfetch_t *zf, zstream_t *zs, boolean_t missed, boolean_t have_lock) { int64_t pf_start, pf_end, ipf_start, ipf_end; int epbs, issued; if (missed) zs->zs_missed = missed; /* * Postpone the prefetch if there are more concurrent callers. * It happens when multiple requests are waiting for the same * indirect block. The last one will run the prefetch for all. */ if (zfs_refcount_remove(&zs->zs_callers, NULL) != 0) { /* Drop reference taken in dmu_zfetch_prepare(). */ if (zfs_refcount_remove(&zs->zs_refs, NULL) == 0) dmu_zfetch_stream_fini(zs); return; } mutex_enter(&zf->zf_lock); if (zs->zs_missed) { pf_start = zs->zs_pf_start; pf_end = zs->zs_pf_start = zs->zs_pf_end; } else { pf_start = pf_end = 0; } ipf_start = zs->zs_ipf_start; ipf_end = zs->zs_ipf_start = zs->zs_ipf_end; mutex_exit(&zf->zf_lock); ASSERT3S(pf_start, <=, pf_end); ASSERT3S(ipf_start, <=, ipf_end); epbs = zf->zf_dnode->dn_indblkshift - SPA_BLKPTRSHIFT; ipf_start = P2ROUNDUP(ipf_start, 1 << epbs) >> epbs; ipf_end = P2ROUNDUP(ipf_end, 1 << epbs) >> epbs; ASSERT3S(ipf_start, <=, ipf_end); issued = pf_end - pf_start + ipf_end - ipf_start; if (issued > 1) { /* More references on top of taken in dmu_zfetch_prepare(). */ zfs_refcount_add_few(&zs->zs_refs, issued - 1, NULL); } else if (issued == 0) { /* Some other thread has done our work, so drop the ref. */ if (zfs_refcount_remove(&zs->zs_refs, NULL) == 0) dmu_zfetch_stream_fini(zs); return; } aggsum_add(&zfetch_sums.zfetchstat_io_active, issued); if (!have_lock) rw_enter(&zf->zf_dnode->dn_struct_rwlock, RW_READER); issued = 0; for (int64_t blk = pf_start; blk < pf_end; blk++) { issued += dbuf_prefetch_impl(zf->zf_dnode, 0, blk, ZIO_PRIORITY_ASYNC_READ, 0, dmu_zfetch_done, zs); } for (int64_t iblk = ipf_start; iblk < ipf_end; iblk++) { issued += dbuf_prefetch_impl(zf->zf_dnode, 1, iblk, ZIO_PRIORITY_ASYNC_READ, 0, dmu_zfetch_done, zs); } if (!have_lock) rw_exit(&zf->zf_dnode->dn_struct_rwlock); if (issued) ZFETCHSTAT_ADD(zfetchstat_io_issued, issued); } void dmu_zfetch(zfetch_t *zf, uint64_t blkid, uint64_t nblks, boolean_t fetch_data, boolean_t missed, boolean_t have_lock) { zstream_t *zs; zs = dmu_zfetch_prepare(zf, blkid, nblks, fetch_data, have_lock); if (zs) dmu_zfetch_run(zf, zs, missed, have_lock); } ZFS_MODULE_PARAM(zfs_prefetch, zfs_prefetch_, disable, INT, ZMOD_RW, "Disable all ZFS prefetching"); ZFS_MODULE_PARAM(zfs_prefetch, zfetch_, max_streams, UINT, ZMOD_RW, "Max number of streams per zfetch"); ZFS_MODULE_PARAM(zfs_prefetch, zfetch_, min_sec_reap, UINT, ZMOD_RW, "Min time before stream reclaim"); ZFS_MODULE_PARAM(zfs_prefetch, zfetch_, max_sec_reap, UINT, ZMOD_RW, "Max time before stream delete"); ZFS_MODULE_PARAM(zfs_prefetch, zfetch_, min_distance, UINT, ZMOD_RW, "Min bytes to prefetch per stream"); ZFS_MODULE_PARAM(zfs_prefetch, zfetch_, max_distance, UINT, ZMOD_RW, "Max bytes to prefetch per stream"); ZFS_MODULE_PARAM(zfs_prefetch, zfetch_, max_idistance, UINT, ZMOD_RW, "Max bytes to prefetch indirects for per stream"); ZFS_MODULE_PARAM(zfs_prefetch, zfetch_, max_reorder, UINT, ZMOD_RW, "Max request reorder distance within a stream"); ZFS_MODULE_PARAM(zfs_prefetch, zfetch_, hole_shift, UINT, ZMOD_RW, "Max log2 fraction of holes in a stream");