Index: head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h
===================================================================
--- head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h (revision 359302)
+++ head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs.h (revision 359303)
@@ -1,886 +1,887 @@
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2011 Pawel Jakub Dawidek. All rights reserved.
* Copyright (c) 2011, 2017 by Delphix. All rights reserved.
* Copyright 2019 Joyent, Inc.
* Copyright (c) 2012 Martin Matuska <mm@FreeBSD.org>. All rights reserved.
* Copyright (c) 2013 Steven Hartland. All rights reserved.
* Copyright (c) 2014 Integros [integros.com]
* Copyright 2016 Nexenta Systems, Inc.
* Copyright (c) 2019 Datto Inc.
*/
#ifndef _LIBZFS_H
#define _LIBZFS_H
#include <assert.h>
#include <libnvpair.h>
#include <sys/mnttab.h>
#include <sys/param.h>
#include <sys/types.h>
#include <sys/varargs.h>
#include <sys/fs/zfs.h>
#include <sys/avl.h>
#include <sys/zfs_ioctl.h>
#include <libzfs_core.h>
#ifdef __cplusplus
extern "C" {
#endif
/*
* Miscellaneous ZFS constants
*/
#define ZFS_MAXPROPLEN MAXPATHLEN
#define ZPOOL_MAXPROPLEN MAXPATHLEN
/*
* libzfs errors
*/
typedef enum zfs_error {
EZFS_SUCCESS = 0, /* no error -- success */
EZFS_NOMEM = 2000, /* out of memory */
EZFS_BADPROP, /* invalid property value */
EZFS_PROPREADONLY, /* cannot set readonly property */
EZFS_PROPTYPE, /* property does not apply to dataset type */
EZFS_PROPNONINHERIT, /* property is not inheritable */
EZFS_PROPSPACE, /* bad quota or reservation */
EZFS_BADTYPE, /* dataset is not of appropriate type */
EZFS_BUSY, /* pool or dataset is busy */
EZFS_EXISTS, /* pool or dataset already exists */
EZFS_NOENT, /* no such pool or dataset */
EZFS_BADSTREAM, /* bad backup stream */
EZFS_DSREADONLY, /* dataset is readonly */
EZFS_VOLTOOBIG, /* volume is too large for 32-bit system */
EZFS_INVALIDNAME, /* invalid dataset name */
EZFS_BADRESTORE, /* unable to restore to destination */
EZFS_BADBACKUP, /* backup failed */
EZFS_BADTARGET, /* bad attach/detach/replace target */
EZFS_NODEVICE, /* no such device in pool */
EZFS_BADDEV, /* invalid device to add */
EZFS_NOREPLICAS, /* no valid replicas */
EZFS_RESILVERING, /* currently resilvering */
EZFS_BADVERSION, /* unsupported version */
EZFS_POOLUNAVAIL, /* pool is currently unavailable */
EZFS_DEVOVERFLOW, /* too many devices in one vdev */
EZFS_BADPATH, /* must be an absolute path */
EZFS_CROSSTARGET, /* rename or clone across pool or dataset */
EZFS_ZONED, /* used improperly in local zone */
EZFS_MOUNTFAILED, /* failed to mount dataset */
EZFS_UMOUNTFAILED, /* failed to unmount dataset */
EZFS_UNSHARENFSFAILED, /* unshare(1M) failed */
EZFS_SHARENFSFAILED, /* share(1M) failed */
EZFS_PERM, /* permission denied */
EZFS_NOSPC, /* out of space */
EZFS_FAULT, /* bad address */
EZFS_IO, /* I/O error */
EZFS_INTR, /* signal received */
EZFS_ISSPARE, /* device is a hot spare */
EZFS_INVALCONFIG, /* invalid vdev configuration */
EZFS_RECURSIVE, /* recursive dependency */
EZFS_NOHISTORY, /* no history object */
EZFS_POOLPROPS, /* couldn't retrieve pool props */
EZFS_POOL_NOTSUP, /* ops not supported for this type of pool */
EZFS_POOL_INVALARG, /* invalid argument for this pool operation */
EZFS_NAMETOOLONG, /* dataset name is too long */
EZFS_OPENFAILED, /* open of device failed */
EZFS_NOCAP, /* couldn't get capacity */
EZFS_LABELFAILED, /* write of label failed */
EZFS_BADWHO, /* invalid permission who */
EZFS_BADPERM, /* invalid permission */
EZFS_BADPERMSET, /* invalid permission set name */
EZFS_NODELEGATION, /* delegated administration is disabled */
EZFS_UNSHARESMBFAILED, /* failed to unshare over smb */
EZFS_SHARESMBFAILED, /* failed to share over smb */
EZFS_BADCACHE, /* bad cache file */
EZFS_ISL2CACHE, /* device is for the level 2 ARC */
EZFS_VDEVNOTSUP, /* unsupported vdev type */
EZFS_NOTSUP, /* ops not supported on this dataset */
EZFS_ACTIVE_SPARE, /* pool has active shared spare devices */
EZFS_UNPLAYED_LOGS, /* log device has unplayed logs */
EZFS_REFTAG_RELE, /* snapshot release: tag not found */
EZFS_REFTAG_HOLD, /* snapshot hold: tag already exists */
EZFS_TAGTOOLONG, /* snapshot hold/rele: tag too long */
EZFS_PIPEFAILED, /* pipe create failed */
EZFS_THREADCREATEFAILED, /* thread create failed */
EZFS_POSTSPLIT_ONLINE, /* onlining a disk after splitting it */
EZFS_SCRUBBING, /* currently scrubbing */
EZFS_NO_SCRUB, /* no active scrub */
EZFS_DIFF, /* general failure of zfs diff */
EZFS_DIFFDATA, /* bad zfs diff data */
EZFS_POOLREADONLY, /* pool is in read-only mode */
EZFS_SCRUB_PAUSED, /* scrub currently paused */
EZFS_ACTIVE_POOL, /* pool is imported on a different system */
EZFS_NO_PENDING, /* cannot cancel, no operation is pending */
EZFS_CHECKPOINT_EXISTS, /* checkpoint exists */
EZFS_DISCARDING_CHECKPOINT, /* currently discarding a checkpoint */
EZFS_NO_CHECKPOINT, /* pool has no checkpoint */
EZFS_DEVRM_IN_PROGRESS, /* a device is currently being removed */
EZFS_VDEV_TOO_BIG, /* a device is too big to be used */
EZFS_TOOMANY, /* argument list too long */
EZFS_INITIALIZING, /* currently initializing */
EZFS_NO_INITIALIZE, /* no active initialize */
+ EZFS_WRONG_PARENT, /* invalid parent dataset (e.g ZVOL) */
EZFS_UNKNOWN
} zfs_error_t;
/*
* UEFI boot support parameters. When creating whole disk boot pool,
* zpool create should allow to create EFI System partition for UEFI boot
* program. In case of BIOS, the EFI System partition is not used
* even if it does exist.
*/
typedef enum zpool_boot_label {
ZPOOL_NO_BOOT_LABEL = 0,
ZPOOL_CREATE_BOOT_LABEL,
ZPOOL_COPY_BOOT_LABEL
} zpool_boot_label_t;
/*
* The following data structures are all part
* of the zfs_allow_t data structure which is
* used for printing 'allow' permissions.
* It is a linked list of zfs_allow_t's which
* then contain avl tree's for user/group/sets/...
* and each one of the entries in those trees have
* avl tree's for the permissions they belong to and
* whether they are local,descendent or local+descendent
* permissions. The AVL trees are used primarily for
* sorting purposes, but also so that we can quickly find
* a given user and or permission.
*/
typedef struct zfs_perm_node {
avl_node_t z_node;
char z_pname[MAXPATHLEN];
} zfs_perm_node_t;
typedef struct zfs_allow_node {
avl_node_t z_node;
char z_key[MAXPATHLEN]; /* name, such as joe */
avl_tree_t z_localdescend; /* local+descendent perms */
avl_tree_t z_local; /* local permissions */
avl_tree_t z_descend; /* descendent permissions */
} zfs_allow_node_t;
typedef struct zfs_allow {
struct zfs_allow *z_next;
char z_setpoint[MAXPATHLEN];
avl_tree_t z_sets;
avl_tree_t z_crperms;
avl_tree_t z_user;
avl_tree_t z_group;
avl_tree_t z_everyone;
} zfs_allow_t;
/*
* Basic handle types
*/
typedef struct zfs_handle zfs_handle_t;
typedef struct zpool_handle zpool_handle_t;
typedef struct libzfs_handle libzfs_handle_t;
/*
* Library initialization
*/
extern libzfs_handle_t *libzfs_init(void);
extern void libzfs_fini(libzfs_handle_t *);
extern libzfs_handle_t *zpool_get_handle(zpool_handle_t *);
extern libzfs_handle_t *zfs_get_handle(zfs_handle_t *);
extern void libzfs_print_on_error(libzfs_handle_t *, boolean_t);
extern void zfs_save_arguments(int argc, char **, char *, int);
extern int zpool_log_history(libzfs_handle_t *, const char *);
extern int libzfs_errno(libzfs_handle_t *);
extern const char *libzfs_error_action(libzfs_handle_t *);
extern const char *libzfs_error_description(libzfs_handle_t *);
extern int zfs_standard_error(libzfs_handle_t *, int, const char *);
extern void libzfs_mnttab_init(libzfs_handle_t *);
extern void libzfs_mnttab_fini(libzfs_handle_t *);
extern void libzfs_mnttab_cache(libzfs_handle_t *, boolean_t);
extern int libzfs_mnttab_find(libzfs_handle_t *, const char *,
struct mnttab *);
extern void libzfs_mnttab_add(libzfs_handle_t *, const char *,
const char *, const char *);
extern void libzfs_mnttab_remove(libzfs_handle_t *, const char *);
/*
* Basic handle functions
*/
extern zpool_handle_t *zpool_open(libzfs_handle_t *, const char *);
extern zpool_handle_t *zpool_open_canfail(libzfs_handle_t *, const char *);
extern void zpool_close(zpool_handle_t *);
extern const char *zpool_get_name(zpool_handle_t *);
extern int zpool_get_state(zpool_handle_t *);
extern const char *zpool_state_to_name(vdev_state_t, vdev_aux_t);
extern const char *zpool_pool_state_to_name(pool_state_t);
extern void zpool_free_handles(libzfs_handle_t *);
extern int zpool_nextboot(libzfs_handle_t *, uint64_t, uint64_t, const char *);
/*
* Iterate over all active pools in the system.
*/
typedef int (*zpool_iter_f)(zpool_handle_t *, void *);
extern int zpool_iter(libzfs_handle_t *, zpool_iter_f, void *);
extern boolean_t zpool_skip_pool(const char *);
/*
* Functions to create and destroy pools
*/
extern int zpool_create(libzfs_handle_t *, const char *, nvlist_t *,
nvlist_t *, nvlist_t *);
extern int zpool_destroy(zpool_handle_t *, const char *);
extern int zpool_add(zpool_handle_t *, nvlist_t *);
typedef struct splitflags {
/* do not split, but return the config that would be split off */
int dryrun : 1;
/* after splitting, import the pool */
int import : 1;
int name_flags;
} splitflags_t;
/*
* Functions to manipulate pool and vdev state
*/
extern int zpool_scan(zpool_handle_t *, pool_scan_func_t, pool_scrub_cmd_t);
extern int zpool_initialize(zpool_handle_t *, pool_initialize_func_t,
nvlist_t *);
extern int zpool_clear(zpool_handle_t *, const char *, nvlist_t *);
extern int zpool_reguid(zpool_handle_t *);
extern int zpool_reopen(zpool_handle_t *);
extern int zpool_sync_one(zpool_handle_t *, void *);
extern int zpool_vdev_online(zpool_handle_t *, const char *, int,
vdev_state_t *);
extern int zpool_vdev_offline(zpool_handle_t *, const char *, boolean_t);
extern int zpool_vdev_attach(zpool_handle_t *, const char *,
const char *, nvlist_t *, int);
extern int zpool_vdev_detach(zpool_handle_t *, const char *);
extern int zpool_vdev_remove(zpool_handle_t *, const char *);
extern int zpool_vdev_remove_cancel(zpool_handle_t *);
extern int zpool_vdev_indirect_size(zpool_handle_t *, const char *, uint64_t *);
extern int zpool_vdev_split(zpool_handle_t *, char *, nvlist_t **, nvlist_t *,
splitflags_t);
extern int zpool_vdev_fault(zpool_handle_t *, uint64_t, vdev_aux_t);
extern int zpool_vdev_degrade(zpool_handle_t *, uint64_t, vdev_aux_t);
extern int zpool_vdev_clear(zpool_handle_t *, uint64_t);
extern nvlist_t *zpool_find_vdev(zpool_handle_t *, const char *, boolean_t *,
boolean_t *, boolean_t *);
extern nvlist_t *zpool_find_vdev_by_physpath(zpool_handle_t *, const char *,
boolean_t *, boolean_t *, boolean_t *);
extern int zpool_label_disk(libzfs_handle_t *, zpool_handle_t *, const char *,
zpool_boot_label_t, uint64_t, int *);
/*
* Functions to manage pool properties
*/
extern int zpool_set_prop(zpool_handle_t *, const char *, const char *);
extern int zpool_get_prop(zpool_handle_t *, zpool_prop_t, char *,
size_t proplen, zprop_source_t *, boolean_t);
extern uint64_t zpool_get_prop_int(zpool_handle_t *, zpool_prop_t,
zprop_source_t *);
extern const char *zpool_prop_to_name(zpool_prop_t);
extern const char *zpool_prop_values(zpool_prop_t);
/*
* Pool health statistics.
*/
typedef enum {
/*
* The following correspond to faults as defined in the (fault.fs.zfs.*)
* event namespace. Each is associated with a corresponding message ID.
* This must be kept in sync with the zfs_msgid_table in
* lib/libzfs/libzfs_status.c.
*/
ZPOOL_STATUS_CORRUPT_CACHE, /* corrupt /kernel/drv/zpool.cache */
ZPOOL_STATUS_MISSING_DEV_R, /* missing device with replicas */
ZPOOL_STATUS_MISSING_DEV_NR, /* missing device with no replicas */
ZPOOL_STATUS_CORRUPT_LABEL_R, /* bad device label with replicas */
ZPOOL_STATUS_CORRUPT_LABEL_NR, /* bad device label with no replicas */
ZPOOL_STATUS_BAD_GUID_SUM, /* sum of device guids didn't match */
ZPOOL_STATUS_CORRUPT_POOL, /* pool metadata is corrupted */
ZPOOL_STATUS_CORRUPT_DATA, /* data errors in user (meta)data */
ZPOOL_STATUS_FAILING_DEV, /* device experiencing errors */
ZPOOL_STATUS_VERSION_NEWER, /* newer on-disk version */
ZPOOL_STATUS_HOSTID_MISMATCH, /* last accessed by another system */
ZPOOL_STATUS_HOSTID_ACTIVE, /* currently active on another system */
ZPOOL_STATUS_HOSTID_REQUIRED, /* multihost=on and hostid=0 */
ZPOOL_STATUS_IO_FAILURE_WAIT, /* failed I/O, failmode 'wait' */
ZPOOL_STATUS_IO_FAILURE_CONTINUE, /* failed I/O, failmode 'continue' */
ZPOOL_STATUS_IO_FAILURE_MMP, /* failed MMP, failmode not 'panic' */
ZPOOL_STATUS_BAD_LOG, /* cannot read log chain(s) */
/*
* If the pool has unsupported features but can still be opened in
* read-only mode, its status is ZPOOL_STATUS_UNSUP_FEAT_WRITE. If the
* pool has unsupported features but cannot be opened at all, its
* status is ZPOOL_STATUS_UNSUP_FEAT_READ.
*/
ZPOOL_STATUS_UNSUP_FEAT_READ, /* unsupported features for read */
ZPOOL_STATUS_UNSUP_FEAT_WRITE, /* unsupported features for write */
/*
* These faults have no corresponding message ID. At the time we are
* checking the status, the original reason for the FMA fault (I/O or
* checksum errors) has been lost.
*/
ZPOOL_STATUS_FAULTED_DEV_R, /* faulted device with replicas */
ZPOOL_STATUS_FAULTED_DEV_NR, /* faulted device with no replicas */
/*
* The following are not faults per se, but still an error possibly
* requiring administrative attention. There is no corresponding
* message ID.
*/
ZPOOL_STATUS_VERSION_OLDER, /* older legacy on-disk version */
ZPOOL_STATUS_FEAT_DISABLED, /* supported features are disabled */
ZPOOL_STATUS_RESILVERING, /* device being resilvered */
ZPOOL_STATUS_OFFLINE_DEV, /* device offline */
ZPOOL_STATUS_REMOVED_DEV, /* removed device */
ZPOOL_STATUS_NON_NATIVE_ASHIFT, /* (e.g. 512e dev with ashift of 9) */
/*
* Finally, the following indicates a healthy pool.
*/
ZPOOL_STATUS_OK
} zpool_status_t;
extern zpool_status_t zpool_get_status(zpool_handle_t *, char **);
extern zpool_status_t zpool_import_status(nvlist_t *, char **);
extern void zpool_dump_ddt(const ddt_stat_t *dds, const ddt_histogram_t *ddh);
/*
* Statistics and configuration functions.
*/
extern nvlist_t *zpool_get_config(zpool_handle_t *, nvlist_t **);
extern nvlist_t *zpool_get_features(zpool_handle_t *);
extern int zpool_refresh_stats(zpool_handle_t *, boolean_t *);
extern int zpool_get_errlog(zpool_handle_t *, nvlist_t **);
extern boolean_t zpool_is_bootable(zpool_handle_t *);
/*
* Import and export functions
*/
extern int zpool_export(zpool_handle_t *, boolean_t, const char *);
extern int zpool_export_force(zpool_handle_t *, const char *);
extern int zpool_import(libzfs_handle_t *, nvlist_t *, const char *,
char *altroot);
extern int zpool_import_props(libzfs_handle_t *, nvlist_t *, const char *,
nvlist_t *, int);
extern void zpool_print_unsup_feat(nvlist_t *config);
/*
* Search for pools to import
*/
typedef struct importargs {
char **path; /* a list of paths to search */
int paths; /* number of paths to search */
char *poolname; /* name of a pool to find */
uint64_t guid; /* guid of a pool to find */
char *cachefile; /* cachefile to use for import */
int can_be_active : 1; /* can the pool be active? */
int unique : 1; /* does 'poolname' already exist? */
int exists : 1; /* set on return if pool already exists */
nvlist_t *policy; /* load policy (max txg, rewind, etc.) */
} importargs_t;
extern nvlist_t *zpool_search_import(libzfs_handle_t *, importargs_t *);
extern int zpool_tryimport(libzfs_handle_t *hdl, char *target,
nvlist_t **configp, importargs_t *args);
/* legacy pool search routines */
extern nvlist_t *zpool_find_import(libzfs_handle_t *, int, char **);
extern nvlist_t *zpool_find_import_cached(libzfs_handle_t *, const char *,
char *, uint64_t);
/*
* Miscellaneous pool functions
*/
struct zfs_cmd;
extern const char *zfs_history_event_names[];
typedef enum {
VDEV_NAME_PATH = 1 << 0,
VDEV_NAME_GUID = 1 << 1,
VDEV_NAME_FOLLOW_LINKS = 1 << 2,
VDEV_NAME_TYPE_ID = 1 << 3,
} vdev_name_t;
extern char *zpool_vdev_name(libzfs_handle_t *, zpool_handle_t *, nvlist_t *,
int name_flags);
extern int zpool_upgrade(zpool_handle_t *, uint64_t);
extern int zpool_get_history(zpool_handle_t *, nvlist_t **);
extern int zpool_history_unpack(char *, uint64_t, uint64_t *,
nvlist_t ***, uint_t *);
extern void zpool_obj_to_path(zpool_handle_t *, uint64_t, uint64_t, char *,
size_t len);
extern int zfs_ioctl(libzfs_handle_t *, int request, struct zfs_cmd *);
extern int zpool_get_physpath(zpool_handle_t *, char *, size_t);
extern void zpool_explain_recover(libzfs_handle_t *, const char *, int,
nvlist_t *);
extern int zpool_checkpoint(zpool_handle_t *);
extern int zpool_discard_checkpoint(zpool_handle_t *);
/*
* Basic handle manipulations. These functions do not create or destroy the
* underlying datasets, only the references to them.
*/
extern zfs_handle_t *zfs_open(libzfs_handle_t *, const char *, int);
extern zfs_handle_t *zfs_handle_dup(zfs_handle_t *);
extern void zfs_close(zfs_handle_t *);
extern zfs_type_t zfs_get_type(const zfs_handle_t *);
extern const char *zfs_get_name(const zfs_handle_t *);
extern zpool_handle_t *zfs_get_pool_handle(const zfs_handle_t *);
extern const char *zfs_get_pool_name(const zfs_handle_t *);
/*
* Property management functions. Some functions are shared with the kernel,
* and are found in sys/fs/zfs.h.
*/
/*
* zfs dataset property management
*/
extern const char *zfs_prop_default_string(zfs_prop_t);
extern uint64_t zfs_prop_default_numeric(zfs_prop_t);
extern const char *zfs_prop_column_name(zfs_prop_t);
extern boolean_t zfs_prop_align_right(zfs_prop_t);
extern nvlist_t *zfs_valid_proplist(libzfs_handle_t *, zfs_type_t,
nvlist_t *, uint64_t, zfs_handle_t *, zpool_handle_t *, const char *);
extern const char *zfs_prop_to_name(zfs_prop_t);
extern int zfs_prop_set(zfs_handle_t *, const char *, const char *);
extern int zfs_prop_set_list(zfs_handle_t *, nvlist_t *);
extern int zfs_prop_get(zfs_handle_t *, zfs_prop_t, char *, size_t,
zprop_source_t *, char *, size_t, boolean_t);
extern int zfs_prop_get_recvd(zfs_handle_t *, const char *, char *, size_t,
boolean_t);
extern int zfs_prop_get_numeric(zfs_handle_t *, zfs_prop_t, uint64_t *,
zprop_source_t *, char *, size_t);
extern int zfs_prop_get_userquota_int(zfs_handle_t *zhp, const char *propname,
uint64_t *propvalue);
extern int zfs_prop_get_userquota(zfs_handle_t *zhp, const char *propname,
char *propbuf, int proplen, boolean_t literal);
extern int zfs_prop_get_written_int(zfs_handle_t *zhp, const char *propname,
uint64_t *propvalue);
extern int zfs_prop_get_written(zfs_handle_t *zhp, const char *propname,
char *propbuf, int proplen, boolean_t literal);
extern int zfs_prop_get_feature(zfs_handle_t *zhp, const char *propname,
char *buf, size_t len);
extern uint64_t zfs_prop_get_int(zfs_handle_t *, zfs_prop_t);
extern int zfs_prop_inherit(zfs_handle_t *, const char *, boolean_t);
extern const char *zfs_prop_values(zfs_prop_t);
extern int zfs_prop_is_string(zfs_prop_t prop);
extern nvlist_t *zfs_get_user_props(zfs_handle_t *);
extern nvlist_t *zfs_get_recvd_props(zfs_handle_t *);
extern nvlist_t *zfs_get_clones_nvl(zfs_handle_t *);
typedef struct zprop_list {
int pl_prop;
char *pl_user_prop;
struct zprop_list *pl_next;
boolean_t pl_all;
size_t pl_width;
size_t pl_recvd_width;
boolean_t pl_fixed;
} zprop_list_t;
extern int zfs_expand_proplist(zfs_handle_t *, zprop_list_t **, boolean_t,
boolean_t);
extern void zfs_prune_proplist(zfs_handle_t *, uint8_t *);
#define ZFS_MOUNTPOINT_NONE "none"
#define ZFS_MOUNTPOINT_LEGACY "legacy"
#define ZFS_FEATURE_DISABLED "disabled"
#define ZFS_FEATURE_ENABLED "enabled"
#define ZFS_FEATURE_ACTIVE "active"
#define ZFS_UNSUPPORTED_INACTIVE "inactive"
#define ZFS_UNSUPPORTED_READONLY "readonly"
/*
* zpool property management
*/
extern int zpool_expand_proplist(zpool_handle_t *, zprop_list_t **);
extern int zpool_prop_get_feature(zpool_handle_t *, const char *, char *,
size_t);
extern const char *zpool_prop_default_string(zpool_prop_t);
extern uint64_t zpool_prop_default_numeric(zpool_prop_t);
extern const char *zpool_prop_column_name(zpool_prop_t);
extern boolean_t zpool_prop_align_right(zpool_prop_t);
/*
* Functions shared by zfs and zpool property management.
*/
extern int zprop_iter(zprop_func func, void *cb, boolean_t show_all,
boolean_t ordered, zfs_type_t type);
extern int zprop_get_list(libzfs_handle_t *, char *, zprop_list_t **,
zfs_type_t);
extern void zprop_free_list(zprop_list_t *);
#define ZFS_GET_NCOLS 5
typedef enum {
GET_COL_NONE,
GET_COL_NAME,
GET_COL_PROPERTY,
GET_COL_VALUE,
GET_COL_RECVD,
GET_COL_SOURCE
} zfs_get_column_t;
/*
* Functions for printing zfs or zpool properties
*/
typedef struct zprop_get_cbdata {
int cb_sources;
zfs_get_column_t cb_columns[ZFS_GET_NCOLS];
int cb_colwidths[ZFS_GET_NCOLS + 1];
boolean_t cb_scripted;
boolean_t cb_literal;
boolean_t cb_first;
zprop_list_t *cb_proplist;
zfs_type_t cb_type;
} zprop_get_cbdata_t;
void zprop_print_one_property(const char *, zprop_get_cbdata_t *,
const char *, const char *, zprop_source_t, const char *,
const char *);
/*
* Iterator functions.
*/
typedef int (*zfs_iter_f)(zfs_handle_t *, void *);
extern int zfs_iter_root(libzfs_handle_t *, zfs_iter_f, void *);
extern int zfs_iter_children(zfs_handle_t *, zfs_iter_f, void *);
extern int zfs_iter_dependents(zfs_handle_t *, boolean_t, zfs_iter_f, void *);
extern int zfs_iter_filesystems(zfs_handle_t *, zfs_iter_f, void *);
extern int zfs_iter_snapshots(zfs_handle_t *, boolean_t, zfs_iter_f, void *,
uint64_t, uint64_t);
extern int zfs_iter_snapshots_sorted(zfs_handle_t *, zfs_iter_f, void *,
uint64_t, uint64_t);
extern int zfs_iter_snapspec(zfs_handle_t *, const char *, zfs_iter_f, void *);
extern int zfs_iter_bookmarks(zfs_handle_t *, zfs_iter_f, void *);
typedef struct get_all_cb {
zfs_handle_t **cb_handles;
size_t cb_alloc;
size_t cb_used;
} get_all_cb_t;
void zfs_foreach_mountpoint(libzfs_handle_t *, zfs_handle_t **, size_t,
zfs_iter_f, void*, boolean_t);
void libzfs_add_handle(get_all_cb_t *, zfs_handle_t *);
/*
* Functions to create and destroy datasets.
*/
extern int zfs_create(libzfs_handle_t *, const char *, zfs_type_t,
nvlist_t *);
extern int zfs_create_ancestors(libzfs_handle_t *, const char *);
extern int zfs_destroy(zfs_handle_t *, boolean_t);
extern int zfs_destroy_snaps(zfs_handle_t *, char *, boolean_t);
extern int zfs_destroy_snaps_nvl(libzfs_handle_t *, nvlist_t *, boolean_t);
extern int zfs_clone(zfs_handle_t *, const char *, nvlist_t *);
extern int zfs_snapshot(libzfs_handle_t *, const char *, boolean_t, nvlist_t *);
extern int zfs_snapshot_nvl(libzfs_handle_t *hdl, nvlist_t *snaps,
nvlist_t *props);
extern int zfs_rollback(zfs_handle_t *, zfs_handle_t *, boolean_t);
typedef struct renameflags {
/* recursive rename */
int recurse : 1;
/* don't unmount file systems */
int nounmount : 1;
/* force unmount file systems */
int forceunmount : 1;
} renameflags_t;
extern int zfs_rename(zfs_handle_t *, const char *, const char *,
renameflags_t flags);
typedef struct sendflags {
/* print informational messages (ie, -v was specified) */
boolean_t verbose;
/* recursive send (ie, -R) */
boolean_t replicate;
/* for incrementals, do all intermediate snapshots */
boolean_t doall;
/* if dataset is a clone, do incremental from its origin */
boolean_t fromorigin;
/* do deduplication */
boolean_t dedup;
/* send properties (ie, -p) */
boolean_t props;
/* do not send (no-op, ie. -n) */
boolean_t dryrun;
/* parsable verbose output (ie. -P) */
boolean_t parsable;
/* show progress (ie. -v) */
boolean_t progress;
/* large blocks (>128K) are permitted */
boolean_t largeblock;
/* WRITE_EMBEDDED records of type DATA are permitted */
boolean_t embed_data;
/* compressed WRITE records are permitted */
boolean_t compress;
/* show progress as process title(ie. -V) */
boolean_t progressastitle;
} sendflags_t;
typedef boolean_t (snapfilter_cb_t)(zfs_handle_t *, void *);
extern int zfs_send(zfs_handle_t *, const char *, const char *,
sendflags_t *, int, snapfilter_cb_t, void *, nvlist_t **);
extern int zfs_send_one(zfs_handle_t *, const char *, int, sendflags_t flags);
extern int zfs_send_resume(libzfs_handle_t *, sendflags_t *, int outfd,
const char *);
extern nvlist_t *zfs_send_resume_token_to_nvlist(libzfs_handle_t *hdl,
const char *token);
extern int zfs_promote(zfs_handle_t *);
extern int zfs_hold(zfs_handle_t *, const char *, const char *,
boolean_t, int);
extern int zfs_hold_nvl(zfs_handle_t *, int, nvlist_t *);
extern int zfs_release(zfs_handle_t *, const char *, const char *, boolean_t);
extern int zfs_get_holds(zfs_handle_t *, nvlist_t **);
extern uint64_t zvol_volsize_to_reservation(uint64_t, nvlist_t *);
typedef int (*zfs_userspace_cb_t)(void *arg, const char *domain,
uid_t rid, uint64_t space);
extern int zfs_userspace(zfs_handle_t *, zfs_userquota_prop_t,
zfs_userspace_cb_t, void *);
extern int zfs_get_fsacl(zfs_handle_t *, nvlist_t **);
extern int zfs_set_fsacl(zfs_handle_t *, boolean_t, nvlist_t *);
typedef struct recvflags {
/* print informational messages (ie, -v was specified) */
boolean_t verbose;
/* the destination is a prefix, not the exact fs (ie, -d) */
boolean_t isprefix;
/*
* Only the tail of the sent snapshot path is appended to the
* destination to determine the received snapshot name (ie, -e).
*/
boolean_t istail;
/* do not actually do the recv, just check if it would work (ie, -n) */
boolean_t dryrun;
/* rollback/destroy filesystems as necessary (eg, -F) */
boolean_t force;
/* set "canmount=off" on all modified filesystems */
boolean_t canmountoff;
/*
* Mark the file systems as "resumable" and do not destroy them if the
* receive is interrupted
*/
boolean_t resumable;
/* byteswap flag is used internally; callers need not specify */
boolean_t byteswap;
/* do not mount file systems as they are extracted (private) */
boolean_t nomount;
} recvflags_t;
extern int zfs_receive(libzfs_handle_t *, const char *, nvlist_t *,
recvflags_t *, int, avl_tree_t *);
typedef enum diff_flags {
ZFS_DIFF_PARSEABLE = 0x1,
ZFS_DIFF_TIMESTAMP = 0x2,
ZFS_DIFF_CLASSIFY = 0x4
} diff_flags_t;
extern int zfs_show_diffs(zfs_handle_t *, int, const char *, const char *,
int);
/*
* Miscellaneous functions.
*/
extern const char *zfs_type_to_name(zfs_type_t);
extern void zfs_refresh_properties(zfs_handle_t *);
extern int zfs_name_valid(const char *, zfs_type_t);
extern zfs_handle_t *zfs_path_to_zhandle(libzfs_handle_t *, char *, zfs_type_t);
extern boolean_t zfs_dataset_exists(libzfs_handle_t *, const char *,
zfs_type_t);
extern int zfs_spa_version(zfs_handle_t *, int *);
extern boolean_t zfs_bookmark_exists(const char *path);
extern ulong_t get_system_hostid(void);
/*
* Mount support functions.
*/
extern boolean_t is_mounted(libzfs_handle_t *, const char *special, char **);
extern boolean_t zfs_is_mounted(zfs_handle_t *, char **);
extern int zfs_mount(zfs_handle_t *, const char *, int);
extern int zfs_mount_at(zfs_handle_t *, const char *, int, const char *);
extern int zfs_unmount(zfs_handle_t *, const char *, int);
extern int zfs_unmountall(zfs_handle_t *, int);
/*
* Share support functions.
*/
extern boolean_t zfs_is_shared(zfs_handle_t *);
extern int zfs_share(zfs_handle_t *);
extern int zfs_unshare(zfs_handle_t *);
/*
* Protocol-specific share support functions.
*/
extern boolean_t zfs_is_shared_nfs(zfs_handle_t *, char **);
extern boolean_t zfs_is_shared_smb(zfs_handle_t *, char **);
extern int zfs_share_nfs(zfs_handle_t *);
extern int zfs_share_smb(zfs_handle_t *);
extern int zfs_shareall(zfs_handle_t *);
extern int zfs_unshare_nfs(zfs_handle_t *, const char *);
extern int zfs_unshare_smb(zfs_handle_t *, const char *);
extern int zfs_unshareall_nfs(zfs_handle_t *);
extern int zfs_unshareall_smb(zfs_handle_t *);
extern int zfs_unshareall_bypath(zfs_handle_t *, const char *);
extern int zfs_unshareall(zfs_handle_t *);
extern int zfs_deleg_share_nfs(libzfs_handle_t *, char *, char *, char *,
void *, void *, int, zfs_share_op_t);
/*
* FreeBSD-specific jail support function.
*/
extern int zfs_jail(zfs_handle_t *, int, int);
/*
* When dealing with nvlists, verify() is extremely useful
*/
#ifndef verify
#ifdef NDEBUG
#define verify(EX) ((void)(EX))
#else
#define verify(EX) assert(EX)
#endif
#endif
/*
* Utility function to convert a number to a human-readable form.
*/
extern void zfs_nicenum(uint64_t, char *, size_t);
extern int zfs_nicestrtonum(libzfs_handle_t *, const char *, uint64_t *);
/*
* Given a device or file, determine if it is part of a pool.
*/
extern int zpool_in_use(libzfs_handle_t *, int, pool_state_t *, char **,
boolean_t *);
/*
* Label manipulation.
*/
extern int zpool_read_label(int, nvlist_t **);
extern int zpool_read_all_labels(int, nvlist_t **);
extern int zpool_clear_label(int);
/* is this zvol valid for use as a dump device? */
extern int zvol_check_dump_config(char *);
/*
* Management interfaces for SMB ACL files
*/
int zfs_smb_acl_add(libzfs_handle_t *, char *, char *, char *);
int zfs_smb_acl_remove(libzfs_handle_t *, char *, char *, char *);
int zfs_smb_acl_purge(libzfs_handle_t *, char *, char *);
int zfs_smb_acl_rename(libzfs_handle_t *, char *, char *, char *, char *);
/*
* Enable and disable datasets within a pool by mounting/unmounting and
* sharing/unsharing them.
*/
extern int zpool_enable_datasets(zpool_handle_t *, const char *, int);
extern int zpool_disable_datasets(zpool_handle_t *, boolean_t);
/*
* Mappings between vdev and FRU.
*/
extern void libzfs_fru_refresh(libzfs_handle_t *);
extern const char *libzfs_fru_lookup(libzfs_handle_t *, const char *);
extern const char *libzfs_fru_devpath(libzfs_handle_t *, const char *);
extern boolean_t libzfs_fru_compare(libzfs_handle_t *, const char *,
const char *);
extern boolean_t libzfs_fru_notself(libzfs_handle_t *, const char *);
extern int zpool_fru_set(zpool_handle_t *, uint64_t, const char *);
#ifndef illumos
extern int zmount(const char *, const char *, int, char *, char *, int, char *,
int);
#endif
extern int zfs_remap_indirects(libzfs_handle_t *hdl, const char *);
/* Allow consumers to initialize libshare externally for optimal performance */
extern int zfs_init_libshare_arg(libzfs_handle_t *, int, void *);
/*
* For most consumers, zfs_init_libshare_arg is sufficient on its own, and
* zfs_uninit_libshare is unnecessary. zfs_uninit_libshare should only be called
* if the caller has already initialized libshare for one set of zfs handles,
* and wishes to share or unshare filesystems outside of that set. In that case,
* the caller should uninitialize libshare, and then re-initialize it with the
* new handles being shared or unshared.
*/
extern void zfs_uninit_libshare(libzfs_handle_t *);
#ifdef __cplusplus
}
#endif
#endif /* _LIBZFS_H */
Index: head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_dataset.c
===================================================================
--- head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_dataset.c (revision 359302)
+++ head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_dataset.c (revision 359303)
@@ -1,5289 +1,5284 @@
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2018, Joyent, Inc. All rights reserved.
* Copyright (c) 2011, 2016 by Delphix. All rights reserved.
* Copyright (c) 2012 DEY Storage Systems, Inc. All rights reserved.
* Copyright (c) 2011-2012 Pawel Jakub Dawidek. All rights reserved.
* Copyright (c) 2013 Martin Matuska. All rights reserved.
* Copyright (c) 2013 Steven Hartland. All rights reserved.
* Copyright (c) 2014 Integros [integros.com]
* Copyright 2017 Nexenta Systems, Inc.
* Copyright 2016 Igor Kozhukhov <ikozhukhov@gmail.com>
* Copyright 2017-2018 RackTop Systems.
* Copyright (c) 2019 Datto Inc.
*/
#include <ctype.h>
#include <errno.h>
#include <libintl.h>
#include <math.h>
#include <stdio.h>
#include <stdlib.h>
#include <strings.h>
#include <unistd.h>
#include <stddef.h>
#include <zone.h>
#include <fcntl.h>
#include <sys/mntent.h>
#include <sys/mount.h>
#include <priv.h>
#include <pwd.h>
#include <grp.h>
#include <stddef.h>
#ifdef illumos
#include <idmap.h>
#endif
#include <sys/dnode.h>
#include <sys/spa.h>
#include <sys/zap.h>
#include <sys/misc.h>
#include <libzfs.h>
#include "zfs_namecheck.h"
#include "zfs_prop.h"
#include "libzfs_impl.h"
#include "zfs_deleg.h"
static int userquota_propname_decode(const char *propname, boolean_t zoned,
zfs_userquota_prop_t *typep, char *domain, int domainlen, uint64_t *ridp);
/*
* Given a single type (not a mask of types), return the type in a human
* readable form.
*/
const char *
zfs_type_to_name(zfs_type_t type)
{
switch (type) {
case ZFS_TYPE_FILESYSTEM:
return (dgettext(TEXT_DOMAIN, "filesystem"));
case ZFS_TYPE_SNAPSHOT:
return (dgettext(TEXT_DOMAIN, "snapshot"));
case ZFS_TYPE_VOLUME:
return (dgettext(TEXT_DOMAIN, "volume"));
case ZFS_TYPE_POOL:
return (dgettext(TEXT_DOMAIN, "pool"));
case ZFS_TYPE_BOOKMARK:
return (dgettext(TEXT_DOMAIN, "bookmark"));
default:
assert(!"unhandled zfs_type_t");
}
return (NULL);
}
/*
* Validate a ZFS path. This is used even before trying to open the dataset, to
* provide a more meaningful error message. We call zfs_error_aux() to
* explain exactly why the name was not valid.
*/
int
zfs_validate_name(libzfs_handle_t *hdl, const char *path, int type,
boolean_t modifying)
{
namecheck_err_t why;
char what;
if (entity_namecheck(path, &why, &what) != 0) {
if (hdl != NULL) {
switch (why) {
case NAME_ERR_TOOLONG:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"name is too long"));
break;
case NAME_ERR_LEADING_SLASH:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"leading slash in name"));
break;
case NAME_ERR_EMPTY_COMPONENT:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"empty component in name"));
break;
case NAME_ERR_TRAILING_SLASH:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"trailing slash in name"));
break;
case NAME_ERR_INVALCHAR:
zfs_error_aux(hdl,
dgettext(TEXT_DOMAIN, "invalid character "
"'%c' in name"), what);
break;
case NAME_ERR_MULTIPLE_DELIMITERS:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"multiple '@' and/or '#' delimiters in "
"name"));
break;
case NAME_ERR_NOLETTER:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool doesn't begin with a letter"));
break;
case NAME_ERR_RESERVED:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"name is reserved"));
break;
case NAME_ERR_DISKLIKE:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"reserved disk name"));
break;
default:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"(%d) not defined"), why);
break;
}
}
return (0);
}
if (!(type & ZFS_TYPE_SNAPSHOT) && strchr(path, '@') != NULL) {
if (hdl != NULL)
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"snapshot delimiter '@' is not expected here"));
return (0);
}
if (type == ZFS_TYPE_SNAPSHOT && strchr(path, '@') == NULL) {
if (hdl != NULL)
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"missing '@' delimiter in snapshot name"));
return (0);
}
if (!(type & ZFS_TYPE_BOOKMARK) && strchr(path, '#') != NULL) {
if (hdl != NULL)
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"bookmark delimiter '#' is not expected here"));
return (0);
}
if (type == ZFS_TYPE_BOOKMARK && strchr(path, '#') == NULL) {
if (hdl != NULL)
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"missing '#' delimiter in bookmark name"));
return (0);
}
if (modifying && strchr(path, '%') != NULL) {
if (hdl != NULL)
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"invalid character %c in name"), '%');
return (0);
}
return (-1);
}
int
zfs_name_valid(const char *name, zfs_type_t type)
{
if (type == ZFS_TYPE_POOL)
return (zpool_name_valid(NULL, B_FALSE, name));
return (zfs_validate_name(NULL, name, type, B_FALSE));
}
/*
* This function takes the raw DSL properties, and filters out the user-defined
* properties into a separate nvlist.
*/
static nvlist_t *
process_user_props(zfs_handle_t *zhp, nvlist_t *props)
{
libzfs_handle_t *hdl = zhp->zfs_hdl;
nvpair_t *elem;
nvlist_t *propval;
nvlist_t *nvl;
if (nvlist_alloc(&nvl, NV_UNIQUE_NAME, 0) != 0) {
(void) no_memory(hdl);
return (NULL);
}
elem = NULL;
while ((elem = nvlist_next_nvpair(props, elem)) != NULL) {
if (!zfs_prop_user(nvpair_name(elem)))
continue;
verify(nvpair_value_nvlist(elem, &propval) == 0);
if (nvlist_add_nvlist(nvl, nvpair_name(elem), propval) != 0) {
nvlist_free(nvl);
(void) no_memory(hdl);
return (NULL);
}
}
return (nvl);
}
static zpool_handle_t *
zpool_add_handle(zfs_handle_t *zhp, const char *pool_name)
{
libzfs_handle_t *hdl = zhp->zfs_hdl;
zpool_handle_t *zph;
if ((zph = zpool_open_canfail(hdl, pool_name)) != NULL) {
if (hdl->libzfs_pool_handles != NULL)
zph->zpool_next = hdl->libzfs_pool_handles;
hdl->libzfs_pool_handles = zph;
}
return (zph);
}
static zpool_handle_t *
zpool_find_handle(zfs_handle_t *zhp, const char *pool_name, int len)
{
libzfs_handle_t *hdl = zhp->zfs_hdl;
zpool_handle_t *zph = hdl->libzfs_pool_handles;
while ((zph != NULL) &&
(strncmp(pool_name, zpool_get_name(zph), len) != 0))
zph = zph->zpool_next;
return (zph);
}
/*
* Returns a handle to the pool that contains the provided dataset.
* If a handle to that pool already exists then that handle is returned.
* Otherwise, a new handle is created and added to the list of handles.
*/
static zpool_handle_t *
zpool_handle(zfs_handle_t *zhp)
{
char *pool_name;
int len;
zpool_handle_t *zph;
len = strcspn(zhp->zfs_name, "/@#") + 1;
pool_name = zfs_alloc(zhp->zfs_hdl, len);
(void) strlcpy(pool_name, zhp->zfs_name, len);
zph = zpool_find_handle(zhp, pool_name, len);
if (zph == NULL)
zph = zpool_add_handle(zhp, pool_name);
free(pool_name);
return (zph);
}
void
zpool_free_handles(libzfs_handle_t *hdl)
{
zpool_handle_t *next, *zph = hdl->libzfs_pool_handles;
while (zph != NULL) {
next = zph->zpool_next;
zpool_close(zph);
zph = next;
}
hdl->libzfs_pool_handles = NULL;
}
/*
* Utility function to gather stats (objset and zpl) for the given object.
*/
static int
get_stats_ioctl(zfs_handle_t *zhp, zfs_cmd_t *zc)
{
libzfs_handle_t *hdl = zhp->zfs_hdl;
(void) strlcpy(zc->zc_name, zhp->zfs_name, sizeof (zc->zc_name));
while (ioctl(hdl->libzfs_fd, ZFS_IOC_OBJSET_STATS, zc) != 0) {
if (errno == ENOMEM) {
if (zcmd_expand_dst_nvlist(hdl, zc) != 0) {
return (-1);
}
} else {
return (-1);
}
}
return (0);
}
/*
* Utility function to get the received properties of the given object.
*/
static int
get_recvd_props_ioctl(zfs_handle_t *zhp)
{
libzfs_handle_t *hdl = zhp->zfs_hdl;
nvlist_t *recvdprops;
zfs_cmd_t zc = { 0 };
int err;
if (zcmd_alloc_dst_nvlist(hdl, &zc, 0) != 0)
return (-1);
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
while (ioctl(hdl->libzfs_fd, ZFS_IOC_OBJSET_RECVD_PROPS, &zc) != 0) {
if (errno == ENOMEM) {
if (zcmd_expand_dst_nvlist(hdl, &zc) != 0) {
return (-1);
}
} else {
zcmd_free_nvlists(&zc);
return (-1);
}
}
err = zcmd_read_dst_nvlist(zhp->zfs_hdl, &zc, &recvdprops);
zcmd_free_nvlists(&zc);
if (err != 0)
return (-1);
nvlist_free(zhp->zfs_recvd_props);
zhp->zfs_recvd_props = recvdprops;
return (0);
}
static int
put_stats_zhdl(zfs_handle_t *zhp, zfs_cmd_t *zc)
{
nvlist_t *allprops, *userprops;
zhp->zfs_dmustats = zc->zc_objset_stats; /* structure assignment */
if (zcmd_read_dst_nvlist(zhp->zfs_hdl, zc, &allprops) != 0) {
return (-1);
}
/*
* XXX Why do we store the user props separately, in addition to
* storing them in zfs_props?
*/
if ((userprops = process_user_props(zhp, allprops)) == NULL) {
nvlist_free(allprops);
return (-1);
}
nvlist_free(zhp->zfs_props);
nvlist_free(zhp->zfs_user_props);
zhp->zfs_props = allprops;
zhp->zfs_user_props = userprops;
return (0);
}
static int
get_stats(zfs_handle_t *zhp)
{
int rc = 0;
zfs_cmd_t zc = { 0 };
if (zcmd_alloc_dst_nvlist(zhp->zfs_hdl, &zc, 0) != 0)
return (-1);
if (get_stats_ioctl(zhp, &zc) != 0)
rc = -1;
else if (put_stats_zhdl(zhp, &zc) != 0)
rc = -1;
zcmd_free_nvlists(&zc);
return (rc);
}
/*
* Refresh the properties currently stored in the handle.
*/
void
zfs_refresh_properties(zfs_handle_t *zhp)
{
(void) get_stats(zhp);
}
/*
* Makes a handle from the given dataset name. Used by zfs_open() and
* zfs_iter_* to create child handles on the fly.
*/
static int
make_dataset_handle_common(zfs_handle_t *zhp, zfs_cmd_t *zc)
{
if (put_stats_zhdl(zhp, zc) != 0)
return (-1);
/*
* We've managed to open the dataset and gather statistics. Determine
* the high-level type.
*/
if (zhp->zfs_dmustats.dds_type == DMU_OST_ZVOL)
zhp->zfs_head_type = ZFS_TYPE_VOLUME;
else if (zhp->zfs_dmustats.dds_type == DMU_OST_ZFS)
zhp->zfs_head_type = ZFS_TYPE_FILESYSTEM;
else if (zhp->zfs_dmustats.dds_type == DMU_OST_OTHER)
return (-1);
else
abort();
if (zhp->zfs_dmustats.dds_is_snapshot)
zhp->zfs_type = ZFS_TYPE_SNAPSHOT;
else if (zhp->zfs_dmustats.dds_type == DMU_OST_ZVOL)
zhp->zfs_type = ZFS_TYPE_VOLUME;
else if (zhp->zfs_dmustats.dds_type == DMU_OST_ZFS)
zhp->zfs_type = ZFS_TYPE_FILESYSTEM;
else
abort(); /* we should never see any other types */
if ((zhp->zpool_hdl = zpool_handle(zhp)) == NULL)
return (-1);
return (0);
}
zfs_handle_t *
make_dataset_handle(libzfs_handle_t *hdl, const char *path)
{
zfs_cmd_t zc = { 0 };
zfs_handle_t *zhp = calloc(sizeof (zfs_handle_t), 1);
if (zhp == NULL)
return (NULL);
zhp->zfs_hdl = hdl;
(void) strlcpy(zhp->zfs_name, path, sizeof (zhp->zfs_name));
if (zcmd_alloc_dst_nvlist(hdl, &zc, 0) != 0) {
free(zhp);
return (NULL);
}
if (get_stats_ioctl(zhp, &zc) == -1) {
zcmd_free_nvlists(&zc);
free(zhp);
return (NULL);
}
if (make_dataset_handle_common(zhp, &zc) == -1) {
free(zhp);
zhp = NULL;
}
zcmd_free_nvlists(&zc);
return (zhp);
}
zfs_handle_t *
make_dataset_handle_zc(libzfs_handle_t *hdl, zfs_cmd_t *zc)
{
zfs_handle_t *zhp = calloc(sizeof (zfs_handle_t), 1);
if (zhp == NULL)
return (NULL);
zhp->zfs_hdl = hdl;
(void) strlcpy(zhp->zfs_name, zc->zc_name, sizeof (zhp->zfs_name));
if (make_dataset_handle_common(zhp, zc) == -1) {
free(zhp);
return (NULL);
}
return (zhp);
}
zfs_handle_t *
make_dataset_simple_handle_zc(zfs_handle_t *pzhp, zfs_cmd_t *zc)
{
zfs_handle_t *zhp = calloc(sizeof (zfs_handle_t), 1);
if (zhp == NULL)
return (NULL);
zhp->zfs_hdl = pzhp->zfs_hdl;
(void) strlcpy(zhp->zfs_name, zc->zc_name, sizeof (zhp->zfs_name));
zhp->zfs_head_type = pzhp->zfs_type;
zhp->zfs_type = ZFS_TYPE_SNAPSHOT;
zhp->zpool_hdl = zpool_handle(zhp);
return (zhp);
}
zfs_handle_t *
zfs_handle_dup(zfs_handle_t *zhp_orig)
{
zfs_handle_t *zhp = calloc(sizeof (zfs_handle_t), 1);
if (zhp == NULL)
return (NULL);
zhp->zfs_hdl = zhp_orig->zfs_hdl;
zhp->zpool_hdl = zhp_orig->zpool_hdl;
(void) strlcpy(zhp->zfs_name, zhp_orig->zfs_name,
sizeof (zhp->zfs_name));
zhp->zfs_type = zhp_orig->zfs_type;
zhp->zfs_head_type = zhp_orig->zfs_head_type;
zhp->zfs_dmustats = zhp_orig->zfs_dmustats;
if (zhp_orig->zfs_props != NULL) {
if (nvlist_dup(zhp_orig->zfs_props, &zhp->zfs_props, 0) != 0) {
(void) no_memory(zhp->zfs_hdl);
zfs_close(zhp);
return (NULL);
}
}
if (zhp_orig->zfs_user_props != NULL) {
if (nvlist_dup(zhp_orig->zfs_user_props,
&zhp->zfs_user_props, 0) != 0) {
(void) no_memory(zhp->zfs_hdl);
zfs_close(zhp);
return (NULL);
}
}
if (zhp_orig->zfs_recvd_props != NULL) {
if (nvlist_dup(zhp_orig->zfs_recvd_props,
&zhp->zfs_recvd_props, 0)) {
(void) no_memory(zhp->zfs_hdl);
zfs_close(zhp);
return (NULL);
}
}
zhp->zfs_mntcheck = zhp_orig->zfs_mntcheck;
if (zhp_orig->zfs_mntopts != NULL) {
zhp->zfs_mntopts = zfs_strdup(zhp_orig->zfs_hdl,
zhp_orig->zfs_mntopts);
}
zhp->zfs_props_table = zhp_orig->zfs_props_table;
return (zhp);
}
boolean_t
zfs_bookmark_exists(const char *path)
{
nvlist_t *bmarks;
nvlist_t *props;
char fsname[ZFS_MAX_DATASET_NAME_LEN];
char *bmark_name;
char *pound;
int err;
boolean_t rv;
(void) strlcpy(fsname, path, sizeof (fsname));
pound = strchr(fsname, '#');
if (pound == NULL)
return (B_FALSE);
*pound = '\0';
bmark_name = pound + 1;
props = fnvlist_alloc();
err = lzc_get_bookmarks(fsname, props, &bmarks);
nvlist_free(props);
if (err != 0) {
nvlist_free(bmarks);
return (B_FALSE);
}
rv = nvlist_exists(bmarks, bmark_name);
nvlist_free(bmarks);
return (rv);
}
zfs_handle_t *
make_bookmark_handle(zfs_handle_t *parent, const char *path,
nvlist_t *bmark_props)
{
zfs_handle_t *zhp = calloc(sizeof (zfs_handle_t), 1);
if (zhp == NULL)
return (NULL);
/* Fill in the name. */
zhp->zfs_hdl = parent->zfs_hdl;
(void) strlcpy(zhp->zfs_name, path, sizeof (zhp->zfs_name));
/* Set the property lists. */
if (nvlist_dup(bmark_props, &zhp->zfs_props, 0) != 0) {
free(zhp);
return (NULL);
}
/* Set the types. */
zhp->zfs_head_type = parent->zfs_head_type;
zhp->zfs_type = ZFS_TYPE_BOOKMARK;
if ((zhp->zpool_hdl = zpool_handle(zhp)) == NULL) {
nvlist_free(zhp->zfs_props);
free(zhp);
return (NULL);
}
return (zhp);
}
struct zfs_open_bookmarks_cb_data {
const char *path;
zfs_handle_t *zhp;
};
static int
zfs_open_bookmarks_cb(zfs_handle_t *zhp, void *data)
{
struct zfs_open_bookmarks_cb_data *dp = data;
/*
* Is it the one we are looking for?
*/
if (strcmp(dp->path, zfs_get_name(zhp)) == 0) {
/*
* We found it. Save it and let the caller know we are done.
*/
dp->zhp = zhp;
return (EEXIST);
}
/*
* Not found. Close the handle and ask for another one.
*/
zfs_close(zhp);
return (0);
}
/*
* Opens the given snapshot, bookmark, filesystem, or volume. The 'types'
* argument is a mask of acceptable types. The function will print an
* appropriate error message and return NULL if it can't be opened.
*/
zfs_handle_t *
zfs_open(libzfs_handle_t *hdl, const char *path, int types)
{
zfs_handle_t *zhp;
char errbuf[1024];
char *bookp;
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot open '%s'"), path);
/*
* Validate the name before we even try to open it.
*/
if (!zfs_validate_name(hdl, path, types, B_FALSE)) {
(void) zfs_error(hdl, EZFS_INVALIDNAME, errbuf);
return (NULL);
}
/*
* Bookmarks needs to be handled separately.
*/
bookp = strchr(path, '#');
if (bookp == NULL) {
/*
* Try to get stats for the dataset, which will tell us if it
* exists.
*/
errno = 0;
if ((zhp = make_dataset_handle(hdl, path)) == NULL) {
(void) zfs_standard_error(hdl, errno, errbuf);
return (NULL);
}
} else {
char dsname[ZFS_MAX_DATASET_NAME_LEN];
zfs_handle_t *pzhp;
struct zfs_open_bookmarks_cb_data cb_data = {path, NULL};
/*
* We need to cut out '#' and everything after '#'
* to get the parent dataset name only.
*/
assert(bookp - path < sizeof (dsname));
(void) strncpy(dsname, path, bookp - path);
dsname[bookp - path] = '\0';
/*
* Create handle for the parent dataset.
*/
errno = 0;
if ((pzhp = make_dataset_handle(hdl, dsname)) == NULL) {
(void) zfs_standard_error(hdl, errno, errbuf);
return (NULL);
}
/*
* Iterate bookmarks to find the right one.
*/
errno = 0;
if ((zfs_iter_bookmarks(pzhp, zfs_open_bookmarks_cb,
&cb_data) == 0) && (cb_data.zhp == NULL)) {
(void) zfs_error(hdl, EZFS_NOENT, errbuf);
zfs_close(pzhp);
return (NULL);
}
if (cb_data.zhp == NULL) {
(void) zfs_standard_error(hdl, errno, errbuf);
zfs_close(pzhp);
return (NULL);
}
zhp = cb_data.zhp;
/*
* Cleanup.
*/
zfs_close(pzhp);
}
if (zhp == NULL) {
char *at = strchr(path, '@');
if (at != NULL)
*at = '\0';
errno = 0;
if ((zhp = make_dataset_handle(hdl, path)) == NULL) {
(void) zfs_standard_error(hdl, errno, errbuf);
return (NULL);
}
if (at != NULL)
*at = '@';
(void) strlcpy(zhp->zfs_name, path, sizeof (zhp->zfs_name));
zhp->zfs_type = ZFS_TYPE_SNAPSHOT;
}
if (!(types & zhp->zfs_type)) {
(void) zfs_error(hdl, EZFS_BADTYPE, errbuf);
zfs_close(zhp);
return (NULL);
}
return (zhp);
}
/*
* Release a ZFS handle. Nothing to do but free the associated memory.
*/
void
zfs_close(zfs_handle_t *zhp)
{
if (zhp->zfs_mntopts)
free(zhp->zfs_mntopts);
nvlist_free(zhp->zfs_props);
nvlist_free(zhp->zfs_user_props);
nvlist_free(zhp->zfs_recvd_props);
free(zhp);
}
typedef struct mnttab_node {
struct mnttab mtn_mt;
avl_node_t mtn_node;
} mnttab_node_t;
static int
libzfs_mnttab_cache_compare(const void *arg1, const void *arg2)
{
const mnttab_node_t *mtn1 = (const mnttab_node_t *)arg1;
const mnttab_node_t *mtn2 = (const mnttab_node_t *)arg2;
int rv;
rv = strcmp(mtn1->mtn_mt.mnt_special, mtn2->mtn_mt.mnt_special);
return (AVL_ISIGN(rv));
}
void
libzfs_mnttab_init(libzfs_handle_t *hdl)
{
pthread_mutex_init(&hdl->libzfs_mnttab_cache_lock, NULL);
assert(avl_numnodes(&hdl->libzfs_mnttab_cache) == 0);
avl_create(&hdl->libzfs_mnttab_cache, libzfs_mnttab_cache_compare,
sizeof (mnttab_node_t), offsetof(mnttab_node_t, mtn_node));
}
void
libzfs_mnttab_update(libzfs_handle_t *hdl)
{
struct mnttab entry;
rewind(hdl->libzfs_mnttab);
while (getmntent(hdl->libzfs_mnttab, &entry) == 0) {
mnttab_node_t *mtn;
if (strcmp(entry.mnt_fstype, MNTTYPE_ZFS) != 0)
continue;
mtn = zfs_alloc(hdl, sizeof (mnttab_node_t));
mtn->mtn_mt.mnt_special = zfs_strdup(hdl, entry.mnt_special);
mtn->mtn_mt.mnt_mountp = zfs_strdup(hdl, entry.mnt_mountp);
mtn->mtn_mt.mnt_fstype = zfs_strdup(hdl, entry.mnt_fstype);
mtn->mtn_mt.mnt_mntopts = zfs_strdup(hdl, entry.mnt_mntopts);
avl_add(&hdl->libzfs_mnttab_cache, mtn);
}
}
void
libzfs_mnttab_fini(libzfs_handle_t *hdl)
{
void *cookie = NULL;
mnttab_node_t *mtn;
while ((mtn = avl_destroy_nodes(&hdl->libzfs_mnttab_cache, &cookie))
!= NULL) {
free(mtn->mtn_mt.mnt_special);
free(mtn->mtn_mt.mnt_mountp);
free(mtn->mtn_mt.mnt_fstype);
free(mtn->mtn_mt.mnt_mntopts);
free(mtn);
}
avl_destroy(&hdl->libzfs_mnttab_cache);
(void) pthread_mutex_destroy(&hdl->libzfs_mnttab_cache_lock);
}
void
libzfs_mnttab_cache(libzfs_handle_t *hdl, boolean_t enable)
{
hdl->libzfs_mnttab_enable = enable;
}
int
libzfs_mnttab_find(libzfs_handle_t *hdl, const char *fsname,
struct mnttab *entry)
{
mnttab_node_t find;
mnttab_node_t *mtn;
int ret = ENOENT;
if (!hdl->libzfs_mnttab_enable) {
struct mnttab srch = { 0 };
if (avl_numnodes(&hdl->libzfs_mnttab_cache))
libzfs_mnttab_fini(hdl);
rewind(hdl->libzfs_mnttab);
srch.mnt_special = (char *)fsname;
srch.mnt_fstype = MNTTYPE_ZFS;
if (getmntany(hdl->libzfs_mnttab, entry, &srch) == 0)
return (0);
else
return (ENOENT);
}
pthread_mutex_lock(&hdl->libzfs_mnttab_cache_lock);
if (avl_numnodes(&hdl->libzfs_mnttab_cache) == 0)
libzfs_mnttab_update(hdl);
find.mtn_mt.mnt_special = (char *)fsname;
mtn = avl_find(&hdl->libzfs_mnttab_cache, &find, NULL);
if (mtn) {
*entry = mtn->mtn_mt;
ret = 0;
}
pthread_mutex_unlock(&hdl->libzfs_mnttab_cache_lock);
return (ret);
}
void
libzfs_mnttab_add(libzfs_handle_t *hdl, const char *special,
const char *mountp, const char *mntopts)
{
mnttab_node_t *mtn;
pthread_mutex_lock(&hdl->libzfs_mnttab_cache_lock);
if (avl_numnodes(&hdl->libzfs_mnttab_cache) == 0) {
mtn = zfs_alloc(hdl, sizeof (mnttab_node_t));
mtn->mtn_mt.mnt_special = zfs_strdup(hdl, special);
mtn->mtn_mt.mnt_mountp = zfs_strdup(hdl, mountp);
mtn->mtn_mt.mnt_fstype = zfs_strdup(hdl, MNTTYPE_ZFS);
mtn->mtn_mt.mnt_mntopts = zfs_strdup(hdl, mntopts);
avl_add(&hdl->libzfs_mnttab_cache, mtn);
}
pthread_mutex_unlock(&hdl->libzfs_mnttab_cache_lock);
}
void
libzfs_mnttab_remove(libzfs_handle_t *hdl, const char *fsname)
{
mnttab_node_t find;
mnttab_node_t *ret;
pthread_mutex_lock(&hdl->libzfs_mnttab_cache_lock);
find.mtn_mt.mnt_special = (char *)fsname;
if ((ret = avl_find(&hdl->libzfs_mnttab_cache, (void *)&find, NULL))
!= NULL) {
avl_remove(&hdl->libzfs_mnttab_cache, ret);
free(ret->mtn_mt.mnt_special);
free(ret->mtn_mt.mnt_mountp);
free(ret->mtn_mt.mnt_fstype);
free(ret->mtn_mt.mnt_mntopts);
free(ret);
}
pthread_mutex_unlock(&hdl->libzfs_mnttab_cache_lock);
}
int
zfs_spa_version(zfs_handle_t *zhp, int *spa_version)
{
zpool_handle_t *zpool_handle = zhp->zpool_hdl;
if (zpool_handle == NULL)
return (-1);
*spa_version = zpool_get_prop_int(zpool_handle,
ZPOOL_PROP_VERSION, NULL);
return (0);
}
/*
* The choice of reservation property depends on the SPA version.
*/
static int
zfs_which_resv_prop(zfs_handle_t *zhp, zfs_prop_t *resv_prop)
{
int spa_version;
if (zfs_spa_version(zhp, &spa_version) < 0)
return (-1);
if (spa_version >= SPA_VERSION_REFRESERVATION)
*resv_prop = ZFS_PROP_REFRESERVATION;
else
*resv_prop = ZFS_PROP_RESERVATION;
return (0);
}
/*
* Given an nvlist of properties to set, validates that they are correct, and
* parses any numeric properties (index, boolean, etc) if they are specified as
* strings.
*/
nvlist_t *
zfs_valid_proplist(libzfs_handle_t *hdl, zfs_type_t type, nvlist_t *nvl,
uint64_t zoned, zfs_handle_t *zhp, zpool_handle_t *zpool_hdl,
const char *errbuf)
{
nvpair_t *elem;
uint64_t intval;
char *strval;
zfs_prop_t prop;
nvlist_t *ret;
int chosen_normal = -1;
int chosen_utf = -1;
if (nvlist_alloc(&ret, NV_UNIQUE_NAME, 0) != 0) {
(void) no_memory(hdl);
return (NULL);
}
/*
* Make sure this property is valid and applies to this type.
*/
elem = NULL;
while ((elem = nvlist_next_nvpair(nvl, elem)) != NULL) {
const char *propname = nvpair_name(elem);
prop = zfs_name_to_prop(propname);
if (prop == ZPROP_INVAL && zfs_prop_user(propname)) {
/*
* This is a user property: make sure it's a
* string, and that it's less than ZAP_MAXNAMELEN.
*/
if (nvpair_type(elem) != DATA_TYPE_STRING) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' must be a string"), propname);
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
goto error;
}
if (strlen(nvpair_name(elem)) >= ZAP_MAXNAMELEN) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"property name '%s' is too long"),
propname);
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
goto error;
}
(void) nvpair_value_string(elem, &strval);
if (nvlist_add_string(ret, propname, strval) != 0) {
(void) no_memory(hdl);
goto error;
}
continue;
}
/*
* Currently, only user properties can be modified on
* snapshots.
*/
if (type == ZFS_TYPE_SNAPSHOT) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"this property can not be modified for snapshots"));
(void) zfs_error(hdl, EZFS_PROPTYPE, errbuf);
goto error;
}
if (prop == ZPROP_INVAL && zfs_prop_userquota(propname)) {
zfs_userquota_prop_t uqtype;
char newpropname[128];
char domain[128];
uint64_t rid;
uint64_t valary[3];
if (userquota_propname_decode(propname, zoned,
&uqtype, domain, sizeof (domain), &rid) != 0) {
zfs_error_aux(hdl,
dgettext(TEXT_DOMAIN,
"'%s' has an invalid user/group name"),
propname);
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
goto error;
}
if (uqtype != ZFS_PROP_USERQUOTA &&
uqtype != ZFS_PROP_GROUPQUOTA) {
zfs_error_aux(hdl,
dgettext(TEXT_DOMAIN, "'%s' is readonly"),
propname);
(void) zfs_error(hdl, EZFS_PROPREADONLY,
errbuf);
goto error;
}
if (nvpair_type(elem) == DATA_TYPE_STRING) {
(void) nvpair_value_string(elem, &strval);
if (strcmp(strval, "none") == 0) {
intval = 0;
} else if (zfs_nicestrtonum(hdl,
strval, &intval) != 0) {
(void) zfs_error(hdl,
EZFS_BADPROP, errbuf);
goto error;
}
} else if (nvpair_type(elem) ==
DATA_TYPE_UINT64) {
(void) nvpair_value_uint64(elem, &intval);
if (intval == 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"use 'none' to disable "
"userquota/groupquota"));
goto error;
}
} else {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' must be a number"), propname);
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
goto error;
}
/*
* Encode the prop name as
* userquota@<hex-rid>-domain, to make it easy
* for the kernel to decode.
*/
(void) snprintf(newpropname, sizeof (newpropname),
"%s%llx-%s", zfs_userquota_prop_prefixes[uqtype],
(longlong_t)rid, domain);
valary[0] = uqtype;
valary[1] = rid;
valary[2] = intval;
if (nvlist_add_uint64_array(ret, newpropname,
valary, 3) != 0) {
(void) no_memory(hdl);
goto error;
}
continue;
} else if (prop == ZPROP_INVAL && zfs_prop_written(propname)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' is readonly"),
propname);
(void) zfs_error(hdl, EZFS_PROPREADONLY, errbuf);
goto error;
}
if (prop == ZPROP_INVAL) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"invalid property '%s'"), propname);
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
goto error;
}
if (!zfs_prop_valid_for_type(prop, type)) {
zfs_error_aux(hdl,
dgettext(TEXT_DOMAIN, "'%s' does not "
"apply to datasets of this type"), propname);
(void) zfs_error(hdl, EZFS_PROPTYPE, errbuf);
goto error;
}
if (zfs_prop_readonly(prop) &&
(!zfs_prop_setonce(prop) || zhp != NULL)) {
zfs_error_aux(hdl,
dgettext(TEXT_DOMAIN, "'%s' is readonly"),
propname);
(void) zfs_error(hdl, EZFS_PROPREADONLY, errbuf);
goto error;
}
if (zprop_parse_value(hdl, elem, prop, type, ret,
&strval, &intval, errbuf) != 0)
goto error;
/*
* Perform some additional checks for specific properties.
*/
switch (prop) {
case ZFS_PROP_VERSION:
{
int version;
if (zhp == NULL)
break;
version = zfs_prop_get_int(zhp, ZFS_PROP_VERSION);
if (intval < version) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"Can not downgrade; already at version %u"),
version);
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
goto error;
}
break;
}
case ZFS_PROP_VOLBLOCKSIZE:
case ZFS_PROP_RECORDSIZE:
{
int maxbs = SPA_MAXBLOCKSIZE;
if (zpool_hdl != NULL) {
maxbs = zpool_get_prop_int(zpool_hdl,
ZPOOL_PROP_MAXBLOCKSIZE, NULL);
}
/*
* Volumes are limited to a volblocksize of 128KB,
* because they typically service workloads with
* small random writes, which incur a large performance
* penalty with large blocks.
*/
if (prop == ZFS_PROP_VOLBLOCKSIZE)
maxbs = SPA_OLD_MAXBLOCKSIZE;
/*
* The value must be a power of two between
* SPA_MINBLOCKSIZE and maxbs.
*/
if (intval < SPA_MINBLOCKSIZE ||
intval > maxbs || !ISP2(intval)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' must be power of 2 from 512B "
"to %uKB"), propname, maxbs >> 10);
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
goto error;
}
break;
}
case ZFS_PROP_SPECIAL_SMALL_BLOCKS:
if (zpool_hdl != NULL) {
char state[64] = "";
/*
* Issue a warning but do not fail so that
* tests for setable properties succeed.
*/
if (zpool_prop_get_feature(zpool_hdl,
"feature@allocation_classes", state,
sizeof (state)) != 0 ||
strcmp(state, ZFS_FEATURE_ACTIVE) != 0) {
(void) fprintf(stderr, gettext(
"%s: property requires a special "
"device in the pool\n"), propname);
}
}
if (intval != 0 &&
(intval < SPA_MINBLOCKSIZE ||
intval > SPA_OLD_MAXBLOCKSIZE || !ISP2(intval))) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"invalid '%s=%d' property: must be zero or "
"a power of 2 from 512B to 128K"), propname,
intval);
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
goto error;
}
break;
case ZFS_PROP_MLSLABEL:
{
#ifdef illumos
/*
* Verify the mlslabel string and convert to
* internal hex label string.
*/
m_label_t *new_sl;
char *hex = NULL; /* internal label string */
/* Default value is already OK. */
if (strcasecmp(strval, ZFS_MLSLABEL_DEFAULT) == 0)
break;
/* Verify the label can be converted to binary form */
if (((new_sl = m_label_alloc(MAC_LABEL)) == NULL) ||
(str_to_label(strval, &new_sl, MAC_LABEL,
L_NO_CORRECTION, NULL) == -1)) {
goto badlabel;
}
/* Now translate to hex internal label string */
if (label_to_str(new_sl, &hex, M_INTERNAL,
DEF_NAMES) != 0) {
if (hex)
free(hex);
goto badlabel;
}
m_label_free(new_sl);
/* If string is already in internal form, we're done. */
if (strcmp(strval, hex) == 0) {
free(hex);
break;
}
/* Replace the label string with the internal form. */
(void) nvlist_remove(ret, zfs_prop_to_name(prop),
DATA_TYPE_STRING);
verify(nvlist_add_string(ret, zfs_prop_to_name(prop),
hex) == 0);
free(hex);
break;
badlabel:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"invalid mlslabel '%s'"), strval);
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
m_label_free(new_sl); /* OK if null */
#else /* !illumos */
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"mlslabel is not supported on FreeBSD"));
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
#endif /* illumos */
goto error;
}
case ZFS_PROP_MOUNTPOINT:
{
namecheck_err_t why;
if (strcmp(strval, ZFS_MOUNTPOINT_NONE) == 0 ||
strcmp(strval, ZFS_MOUNTPOINT_LEGACY) == 0)
break;
if (mountpoint_namecheck(strval, &why)) {
switch (why) {
case NAME_ERR_LEADING_SLASH:
zfs_error_aux(hdl,
dgettext(TEXT_DOMAIN,
"'%s' must be an absolute path, "
"'none', or 'legacy'"), propname);
break;
case NAME_ERR_TOOLONG:
zfs_error_aux(hdl,
dgettext(TEXT_DOMAIN,
"component of '%s' is too long"),
propname);
break;
default:
zfs_error_aux(hdl,
dgettext(TEXT_DOMAIN,
"(%d) not defined"),
why);
break;
}
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
goto error;
}
}
/*FALLTHRU*/
case ZFS_PROP_SHARESMB:
case ZFS_PROP_SHARENFS:
/*
* For the mountpoint and sharenfs or sharesmb
* properties, check if it can be set in a
* global/non-global zone based on
* the zoned property value:
*
* global zone non-global zone
* --------------------------------------------------
* zoned=on mountpoint (no) mountpoint (yes)
* sharenfs (no) sharenfs (no)
* sharesmb (no) sharesmb (no)
*
* zoned=off mountpoint (yes) N/A
* sharenfs (yes)
* sharesmb (yes)
*/
if (zoned) {
if (getzoneid() == GLOBAL_ZONEID) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' cannot be set on "
"dataset in a non-global zone"),
propname);
(void) zfs_error(hdl, EZFS_ZONED,
errbuf);
goto error;
} else if (prop == ZFS_PROP_SHARENFS ||
prop == ZFS_PROP_SHARESMB) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' cannot be set in "
"a non-global zone"), propname);
(void) zfs_error(hdl, EZFS_ZONED,
errbuf);
goto error;
}
} else if (getzoneid() != GLOBAL_ZONEID) {
/*
* If zoned property is 'off', this must be in
* a global zone. If not, something is wrong.
*/
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' cannot be set while dataset "
"'zoned' property is set"), propname);
(void) zfs_error(hdl, EZFS_ZONED, errbuf);
goto error;
}
/*
* At this point, it is legitimate to set the
* property. Now we want to make sure that the
* property value is valid if it is sharenfs.
*/
if ((prop == ZFS_PROP_SHARENFS ||
prop == ZFS_PROP_SHARESMB) &&
strcmp(strval, "on") != 0 &&
strcmp(strval, "off") != 0) {
zfs_share_proto_t proto;
if (prop == ZFS_PROP_SHARESMB)
proto = PROTO_SMB;
else
proto = PROTO_NFS;
/*
* Must be an valid sharing protocol
* option string so init the libshare
* in order to enable the parser and
* then parse the options. We use the
* control API since we don't care about
* the current configuration and don't
* want the overhead of loading it
* until we actually do something.
*/
if (zfs_init_libshare(hdl,
SA_INIT_CONTROL_API) != SA_OK) {
/*
* An error occurred so we can't do
* anything
*/
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' cannot be set: problem "
"in share initialization"),
propname);
(void) zfs_error(hdl, EZFS_BADPROP,
errbuf);
goto error;
}
if (zfs_parse_options(strval, proto) != SA_OK) {
/*
* There was an error in parsing so
* deal with it by issuing an error
* message and leaving after
* uninitializing the the libshare
* interface.
*/
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' cannot be set to invalid "
"options"), propname);
(void) zfs_error(hdl, EZFS_BADPROP,
errbuf);
zfs_uninit_libshare(hdl);
goto error;
}
zfs_uninit_libshare(hdl);
}
break;
case ZFS_PROP_UTF8ONLY:
chosen_utf = (int)intval;
break;
case ZFS_PROP_NORMALIZE:
chosen_normal = (int)intval;
break;
default:
break;
}
/*
* For changes to existing volumes, we have some additional
* checks to enforce.
*/
if (type == ZFS_TYPE_VOLUME && zhp != NULL) {
uint64_t volsize = zfs_prop_get_int(zhp,
ZFS_PROP_VOLSIZE);
uint64_t blocksize = zfs_prop_get_int(zhp,
ZFS_PROP_VOLBLOCKSIZE);
char buf[64];
switch (prop) {
case ZFS_PROP_RESERVATION:
if (intval > volsize) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' is greater than current "
"volume size"), propname);
(void) zfs_error(hdl, EZFS_BADPROP,
errbuf);
goto error;
}
break;
case ZFS_PROP_REFRESERVATION:
if (intval > volsize && intval != UINT64_MAX) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' is greater than current "
"volume size"), propname);
(void) zfs_error(hdl, EZFS_BADPROP,
errbuf);
goto error;
}
break;
case ZFS_PROP_VOLSIZE:
if (intval % blocksize != 0) {
zfs_nicenum(blocksize, buf,
sizeof (buf));
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' must be a multiple of "
"volume block size (%s)"),
propname, buf);
(void) zfs_error(hdl, EZFS_BADPROP,
errbuf);
goto error;
}
if (intval == 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' cannot be zero"),
propname);
(void) zfs_error(hdl, EZFS_BADPROP,
errbuf);
goto error;
}
break;
default:
break;
}
}
}
/*
* If normalization was chosen, but no UTF8 choice was made,
* enforce rejection of non-UTF8 names.
*
* If normalization was chosen, but rejecting non-UTF8 names
* was explicitly not chosen, it is an error.
*/
if (chosen_normal > 0 && chosen_utf < 0) {
if (nvlist_add_uint64(ret,
zfs_prop_to_name(ZFS_PROP_UTF8ONLY), 1) != 0) {
(void) no_memory(hdl);
goto error;
}
} else if (chosen_normal > 0 && chosen_utf == 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' must be set 'on' if normalization chosen"),
zfs_prop_to_name(ZFS_PROP_UTF8ONLY));
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
goto error;
}
return (ret);
error:
nvlist_free(ret);
return (NULL);
}
int
zfs_add_synthetic_resv(zfs_handle_t *zhp, nvlist_t *nvl)
{
uint64_t old_volsize;
uint64_t new_volsize;
uint64_t old_reservation;
uint64_t new_reservation;
zfs_prop_t resv_prop;
nvlist_t *props;
/*
* If this is an existing volume, and someone is setting the volsize,
* make sure that it matches the reservation, or add it if necessary.
*/
old_volsize = zfs_prop_get_int(zhp, ZFS_PROP_VOLSIZE);
if (zfs_which_resv_prop(zhp, &resv_prop) < 0)
return (-1);
old_reservation = zfs_prop_get_int(zhp, resv_prop);
props = fnvlist_alloc();
fnvlist_add_uint64(props, zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE),
zfs_prop_get_int(zhp, ZFS_PROP_VOLBLOCKSIZE));
if ((zvol_volsize_to_reservation(old_volsize, props) !=
old_reservation) || nvlist_exists(nvl,
zfs_prop_to_name(resv_prop))) {
fnvlist_free(props);
return (0);
}
if (nvlist_lookup_uint64(nvl, zfs_prop_to_name(ZFS_PROP_VOLSIZE),
&new_volsize) != 0) {
fnvlist_free(props);
return (-1);
}
new_reservation = zvol_volsize_to_reservation(new_volsize, props);
fnvlist_free(props);
if (nvlist_add_uint64(nvl, zfs_prop_to_name(resv_prop),
new_reservation) != 0) {
(void) no_memory(zhp->zfs_hdl);
return (-1);
}
return (1);
}
/*
* Helper for 'zfs {set|clone} refreservation=auto'. Must be called after
* zfs_valid_proplist(), as it is what sets the UINT64_MAX sentinal value.
* Return codes must match zfs_add_synthetic_resv().
*/
static int
zfs_fix_auto_resv(zfs_handle_t *zhp, nvlist_t *nvl)
{
uint64_t volsize;
uint64_t resvsize;
zfs_prop_t prop;
nvlist_t *props;
if (!ZFS_IS_VOLUME(zhp)) {
return (0);
}
if (zfs_which_resv_prop(zhp, &prop) != 0) {
return (-1);
}
if (prop != ZFS_PROP_REFRESERVATION) {
return (0);
}
if (nvlist_lookup_uint64(nvl, zfs_prop_to_name(prop), &resvsize) != 0) {
/* No value being set, so it can't be "auto" */
return (0);
}
if (resvsize != UINT64_MAX) {
/* Being set to a value other than "auto" */
return (0);
}
props = fnvlist_alloc();
fnvlist_add_uint64(props, zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE),
zfs_prop_get_int(zhp, ZFS_PROP_VOLBLOCKSIZE));
if (nvlist_lookup_uint64(nvl, zfs_prop_to_name(ZFS_PROP_VOLSIZE),
&volsize) != 0) {
volsize = zfs_prop_get_int(zhp, ZFS_PROP_VOLSIZE);
}
resvsize = zvol_volsize_to_reservation(volsize, props);
fnvlist_free(props);
(void) nvlist_remove_all(nvl, zfs_prop_to_name(prop));
if (nvlist_add_uint64(nvl, zfs_prop_to_name(prop), resvsize) != 0) {
(void) no_memory(zhp->zfs_hdl);
return (-1);
}
return (1);
}
void
zfs_setprop_error(libzfs_handle_t *hdl, zfs_prop_t prop, int err,
char *errbuf)
{
switch (err) {
case ENOSPC:
/*
* For quotas and reservations, ENOSPC indicates
* something different; setting a quota or reservation
* doesn't use any disk space.
*/
switch (prop) {
case ZFS_PROP_QUOTA:
case ZFS_PROP_REFQUOTA:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"size is less than current used or "
"reserved space"));
(void) zfs_error(hdl, EZFS_PROPSPACE, errbuf);
break;
case ZFS_PROP_RESERVATION:
case ZFS_PROP_REFRESERVATION:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"size is greater than available space"));
(void) zfs_error(hdl, EZFS_PROPSPACE, errbuf);
break;
default:
(void) zfs_standard_error(hdl, err, errbuf);
break;
}
break;
case EBUSY:
(void) zfs_standard_error(hdl, EBUSY, errbuf);
break;
case EROFS:
(void) zfs_error(hdl, EZFS_DSREADONLY, errbuf);
break;
case E2BIG:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"property value too long"));
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
break;
case ENOTSUP:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool and or dataset must be upgraded to set this "
"property or value"));
(void) zfs_error(hdl, EZFS_BADVERSION, errbuf);
break;
case ERANGE:
case EDOM:
if (prop == ZFS_PROP_COMPRESSION ||
prop == ZFS_PROP_RECORDSIZE) {
(void) zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"property setting is not allowed on "
"bootable datasets"));
(void) zfs_error(hdl, EZFS_NOTSUP, errbuf);
} else if (prop == ZFS_PROP_CHECKSUM ||
prop == ZFS_PROP_DEDUP) {
(void) zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"property setting is not allowed on "
"root pools"));
(void) zfs_error(hdl, EZFS_NOTSUP, errbuf);
} else {
(void) zfs_standard_error(hdl, err, errbuf);
}
break;
case EINVAL:
if (prop == ZPROP_INVAL) {
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
} else {
(void) zfs_standard_error(hdl, err, errbuf);
}
break;
case EOVERFLOW:
/*
* This platform can't address a volume this big.
*/
#ifdef _ILP32
if (prop == ZFS_PROP_VOLSIZE) {
(void) zfs_error(hdl, EZFS_VOLTOOBIG, errbuf);
break;
}
#endif
/* FALLTHROUGH */
default:
(void) zfs_standard_error(hdl, err, errbuf);
}
}
/*
* Given a property name and value, set the property for the given dataset.
*/
int
zfs_prop_set(zfs_handle_t *zhp, const char *propname, const char *propval)
{
int ret = -1;
char errbuf[1024];
libzfs_handle_t *hdl = zhp->zfs_hdl;
nvlist_t *nvl = NULL;
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot set property for '%s'"),
zhp->zfs_name);
if (nvlist_alloc(&nvl, NV_UNIQUE_NAME, 0) != 0 ||
nvlist_add_string(nvl, propname, propval) != 0) {
(void) no_memory(hdl);
goto error;
}
ret = zfs_prop_set_list(zhp, nvl);
error:
nvlist_free(nvl);
return (ret);
}
/*
* Given an nvlist of property names and values, set the properties for the
* given dataset.
*/
int
zfs_prop_set_list(zfs_handle_t *zhp, nvlist_t *props)
{
zfs_cmd_t zc = { 0 };
int ret = -1;
prop_changelist_t **cls = NULL;
int cl_idx;
char errbuf[1024];
libzfs_handle_t *hdl = zhp->zfs_hdl;
nvlist_t *nvl;
int nvl_len;
int added_resv = 0;
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot set property for '%s'"),
zhp->zfs_name);
if ((nvl = zfs_valid_proplist(hdl, zhp->zfs_type, props,
zfs_prop_get_int(zhp, ZFS_PROP_ZONED), zhp, zhp->zpool_hdl,
errbuf)) == NULL)
goto error;
/*
* We have to check for any extra properties which need to be added
* before computing the length of the nvlist.
*/
for (nvpair_t *elem = nvlist_next_nvpair(nvl, NULL);
elem != NULL;
elem = nvlist_next_nvpair(nvl, elem)) {
if (zfs_name_to_prop(nvpair_name(elem)) == ZFS_PROP_VOLSIZE &&
(added_resv = zfs_add_synthetic_resv(zhp, nvl)) == -1) {
goto error;
}
}
if (added_resv != 1 &&
(added_resv = zfs_fix_auto_resv(zhp, nvl)) == -1) {
goto error;
}
/*
* Check how many properties we're setting and allocate an array to
* store changelist pointers for postfix().
*/
nvl_len = 0;
for (nvpair_t *elem = nvlist_next_nvpair(nvl, NULL);
elem != NULL;
elem = nvlist_next_nvpair(nvl, elem))
nvl_len++;
if ((cls = calloc(nvl_len, sizeof (prop_changelist_t *))) == NULL)
goto error;
cl_idx = 0;
for (nvpair_t *elem = nvlist_next_nvpair(nvl, NULL);
elem != NULL;
elem = nvlist_next_nvpair(nvl, elem)) {
zfs_prop_t prop = zfs_name_to_prop(nvpair_name(elem));
assert(cl_idx < nvl_len);
/*
* We don't want to unmount & remount the dataset when changing
* its canmount property to 'on' or 'noauto'. We only use
* the changelist logic to unmount when setting canmount=off.
*/
if (prop != ZFS_PROP_CANMOUNT ||
(fnvpair_value_uint64(elem) == ZFS_CANMOUNT_OFF &&
zfs_is_mounted(zhp, NULL))) {
cls[cl_idx] = changelist_gather(zhp, prop, 0, 0);
if (cls[cl_idx] == NULL)
goto error;
}
if (prop == ZFS_PROP_MOUNTPOINT &&
changelist_haszonedchild(cls[cl_idx])) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"child dataset with inherited mountpoint is used "
"in a non-global zone"));
ret = zfs_error(hdl, EZFS_ZONED, errbuf);
goto error;
}
/* We don't support those properties on FreeBSD. */
switch (prop) {
case ZFS_PROP_DEVICES:
case ZFS_PROP_ISCSIOPTIONS:
case ZFS_PROP_XATTR:
case ZFS_PROP_VSCAN:
case ZFS_PROP_NBMAND:
case ZFS_PROP_MLSLABEL:
(void) snprintf(errbuf, sizeof (errbuf),
"property '%s' not supported on FreeBSD",
nvpair_name(elem));
ret = zfs_error(hdl, EZFS_PERM, errbuf);
goto error;
}
if (cls[cl_idx] != NULL &&
(ret = changelist_prefix(cls[cl_idx])) != 0)
goto error;
cl_idx++;
}
assert(cl_idx == nvl_len);
/*
* Execute the corresponding ioctl() to set this list of properties.
*/
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
if ((ret = zcmd_write_src_nvlist(hdl, &zc, nvl)) != 0 ||
(ret = zcmd_alloc_dst_nvlist(hdl, &zc, 0)) != 0)
goto error;
ret = zfs_ioctl(hdl, ZFS_IOC_SET_PROP, &zc);
if (ret != 0) {
if (zc.zc_nvlist_dst_filled == B_FALSE) {
(void) zfs_standard_error(hdl, errno, errbuf);
goto error;
}
/* Get the list of unset properties back and report them. */
nvlist_t *errorprops = NULL;
if (zcmd_read_dst_nvlist(hdl, &zc, &errorprops) != 0)
goto error;
for (nvpair_t *elem = nvlist_next_nvpair(errorprops, NULL);
elem != NULL;
elem = nvlist_next_nvpair(errorprops, elem)) {
zfs_prop_t prop = zfs_name_to_prop(nvpair_name(elem));
zfs_setprop_error(hdl, prop, errno, errbuf);
}
nvlist_free(errorprops);
if (added_resv && errno == ENOSPC) {
/* clean up the volsize property we tried to set */
uint64_t old_volsize = zfs_prop_get_int(zhp,
ZFS_PROP_VOLSIZE);
nvlist_free(nvl);
nvl = NULL;
zcmd_free_nvlists(&zc);
if (nvlist_alloc(&nvl, NV_UNIQUE_NAME, 0) != 0)
goto error;
if (nvlist_add_uint64(nvl,
zfs_prop_to_name(ZFS_PROP_VOLSIZE),
old_volsize) != 0)
goto error;
if (zcmd_write_src_nvlist(hdl, &zc, nvl) != 0)
goto error;
(void) zfs_ioctl(hdl, ZFS_IOC_SET_PROP, &zc);
}
} else {
for (cl_idx = 0; cl_idx < nvl_len; cl_idx++) {
if (cls[cl_idx] != NULL) {
int clp_err = changelist_postfix(cls[cl_idx]);
if (clp_err != 0)
ret = clp_err;
}
}
/*
* Refresh the statistics so the new property value
* is reflected.
*/
if (ret == 0)
(void) get_stats(zhp);
}
error:
nvlist_free(nvl);
zcmd_free_nvlists(&zc);
if (cls != NULL) {
for (cl_idx = 0; cl_idx < nvl_len; cl_idx++) {
if (cls[cl_idx] != NULL)
changelist_free(cls[cl_idx]);
}
free(cls);
}
return (ret);
}
/*
* Given a property, inherit the value from the parent dataset, or if received
* is TRUE, revert to the received value, if any.
*/
int
zfs_prop_inherit(zfs_handle_t *zhp, const char *propname, boolean_t received)
{
zfs_cmd_t zc = { 0 };
int ret;
prop_changelist_t *cl;
libzfs_handle_t *hdl = zhp->zfs_hdl;
char errbuf[1024];
zfs_prop_t prop;
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot inherit %s for '%s'"), propname, zhp->zfs_name);
zc.zc_cookie = received;
if ((prop = zfs_name_to_prop(propname)) == ZPROP_INVAL) {
/*
* For user properties, the amount of work we have to do is very
* small, so just do it here.
*/
if (!zfs_prop_user(propname)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"invalid property"));
return (zfs_error(hdl, EZFS_BADPROP, errbuf));
}
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
(void) strlcpy(zc.zc_value, propname, sizeof (zc.zc_value));
if (zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_INHERIT_PROP, &zc) != 0)
return (zfs_standard_error(hdl, errno, errbuf));
return (0);
}
/*
* Verify that this property is inheritable.
*/
if (zfs_prop_readonly(prop))
return (zfs_error(hdl, EZFS_PROPREADONLY, errbuf));
if (!zfs_prop_inheritable(prop) && !received)
return (zfs_error(hdl, EZFS_PROPNONINHERIT, errbuf));
/*
* Check to see if the value applies to this type
*/
if (!zfs_prop_valid_for_type(prop, zhp->zfs_type))
return (zfs_error(hdl, EZFS_PROPTYPE, errbuf));
/*
* Normalize the name, to get rid of shorthand abbreviations.
*/
propname = zfs_prop_to_name(prop);
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
(void) strlcpy(zc.zc_value, propname, sizeof (zc.zc_value));
if (prop == ZFS_PROP_MOUNTPOINT && getzoneid() == GLOBAL_ZONEID &&
zfs_prop_get_int(zhp, ZFS_PROP_ZONED)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"dataset is used in a non-global zone"));
return (zfs_error(hdl, EZFS_ZONED, errbuf));
}
/*
* Determine datasets which will be affected by this change, if any.
*/
if ((cl = changelist_gather(zhp, prop, 0, 0)) == NULL)
return (-1);
if (prop == ZFS_PROP_MOUNTPOINT && changelist_haszonedchild(cl)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"child dataset with inherited mountpoint is used "
"in a non-global zone"));
ret = zfs_error(hdl, EZFS_ZONED, errbuf);
goto error;
}
if ((ret = changelist_prefix(cl)) != 0)
goto error;
if ((ret = zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_INHERIT_PROP, &zc)) != 0) {
return (zfs_standard_error(hdl, errno, errbuf));
} else {
if ((ret = changelist_postfix(cl)) != 0)
goto error;
/*
* Refresh the statistics so the new property is reflected.
*/
(void) get_stats(zhp);
}
error:
changelist_free(cl);
return (ret);
}
/*
* True DSL properties are stored in an nvlist. The following two functions
* extract them appropriately.
*/
static uint64_t
getprop_uint64(zfs_handle_t *zhp, zfs_prop_t prop, char **source)
{
nvlist_t *nv;
uint64_t value;
*source = NULL;
if (nvlist_lookup_nvlist(zhp->zfs_props,
zfs_prop_to_name(prop), &nv) == 0) {
verify(nvlist_lookup_uint64(nv, ZPROP_VALUE, &value) == 0);
(void) nvlist_lookup_string(nv, ZPROP_SOURCE, source);
} else {
verify(!zhp->zfs_props_table ||
zhp->zfs_props_table[prop] == B_TRUE);
value = zfs_prop_default_numeric(prop);
*source = "";
}
return (value);
}
static const char *
getprop_string(zfs_handle_t *zhp, zfs_prop_t prop, char **source)
{
nvlist_t *nv;
const char *value;
*source = NULL;
if (nvlist_lookup_nvlist(zhp->zfs_props,
zfs_prop_to_name(prop), &nv) == 0) {
value = fnvlist_lookup_string(nv, ZPROP_VALUE);
(void) nvlist_lookup_string(nv, ZPROP_SOURCE, source);
} else {
verify(!zhp->zfs_props_table ||
zhp->zfs_props_table[prop] == B_TRUE);
value = zfs_prop_default_string(prop);
*source = "";
}
return (value);
}
static boolean_t
zfs_is_recvd_props_mode(zfs_handle_t *zhp)
{
return (zhp->zfs_props == zhp->zfs_recvd_props);
}
static void
zfs_set_recvd_props_mode(zfs_handle_t *zhp, uint64_t *cookie)
{
*cookie = (uint64_t)(uintptr_t)zhp->zfs_props;
zhp->zfs_props = zhp->zfs_recvd_props;
}
static void
zfs_unset_recvd_props_mode(zfs_handle_t *zhp, uint64_t *cookie)
{
zhp->zfs_props = (nvlist_t *)(uintptr_t)*cookie;
*cookie = 0;
}
/*
* Internal function for getting a numeric property. Both zfs_prop_get() and
* zfs_prop_get_int() are built using this interface.
*
* Certain properties can be overridden using 'mount -o'. In this case, scan
* the contents of the /etc/mnttab entry, searching for the appropriate options.
* If they differ from the on-disk values, report the current values and mark
* the source "temporary".
*/
static int
get_numeric_property(zfs_handle_t *zhp, zfs_prop_t prop, zprop_source_t *src,
char **source, uint64_t *val)
{
zfs_cmd_t zc = { 0 };
nvlist_t *zplprops = NULL;
struct mnttab mnt;
char *mntopt_on = NULL;
char *mntopt_off = NULL;
boolean_t received = zfs_is_recvd_props_mode(zhp);
*source = NULL;
switch (prop) {
case ZFS_PROP_ATIME:
mntopt_on = MNTOPT_ATIME;
mntopt_off = MNTOPT_NOATIME;
break;
case ZFS_PROP_DEVICES:
mntopt_on = MNTOPT_DEVICES;
mntopt_off = MNTOPT_NODEVICES;
break;
case ZFS_PROP_EXEC:
mntopt_on = MNTOPT_EXEC;
mntopt_off = MNTOPT_NOEXEC;
break;
case ZFS_PROP_READONLY:
mntopt_on = MNTOPT_RO;
mntopt_off = MNTOPT_RW;
break;
case ZFS_PROP_SETUID:
mntopt_on = MNTOPT_SETUID;
mntopt_off = MNTOPT_NOSETUID;
break;
case ZFS_PROP_XATTR:
mntopt_on = MNTOPT_XATTR;
mntopt_off = MNTOPT_NOXATTR;
break;
case ZFS_PROP_NBMAND:
mntopt_on = MNTOPT_NBMAND;
mntopt_off = MNTOPT_NONBMAND;
break;
default:
break;
}
/*
* Because looking up the mount options is potentially expensive
* (iterating over all of /etc/mnttab), we defer its calculation until
* we're looking up a property which requires its presence.
*/
if (!zhp->zfs_mntcheck &&
(mntopt_on != NULL || prop == ZFS_PROP_MOUNTED)) {
libzfs_handle_t *hdl = zhp->zfs_hdl;
struct mnttab entry;
if (libzfs_mnttab_find(hdl, zhp->zfs_name, &entry) == 0) {
zhp->zfs_mntopts = zfs_strdup(hdl,
entry.mnt_mntopts);
if (zhp->zfs_mntopts == NULL)
return (-1);
}
zhp->zfs_mntcheck = B_TRUE;
}
if (zhp->zfs_mntopts == NULL)
mnt.mnt_mntopts = "";
else
mnt.mnt_mntopts = zhp->zfs_mntopts;
switch (prop) {
case ZFS_PROP_ATIME:
case ZFS_PROP_DEVICES:
case ZFS_PROP_EXEC:
case ZFS_PROP_READONLY:
case ZFS_PROP_SETUID:
case ZFS_PROP_XATTR:
case ZFS_PROP_NBMAND:
*val = getprop_uint64(zhp, prop, source);
if (received)
break;
if (hasmntopt(&mnt, mntopt_on) && !*val) {
*val = B_TRUE;
if (src)
*src = ZPROP_SRC_TEMPORARY;
} else if (hasmntopt(&mnt, mntopt_off) && *val) {
*val = B_FALSE;
if (src)
*src = ZPROP_SRC_TEMPORARY;
}
break;
case ZFS_PROP_CANMOUNT:
case ZFS_PROP_VOLSIZE:
case ZFS_PROP_QUOTA:
case ZFS_PROP_REFQUOTA:
case ZFS_PROP_RESERVATION:
case ZFS_PROP_REFRESERVATION:
case ZFS_PROP_FILESYSTEM_LIMIT:
case ZFS_PROP_SNAPSHOT_LIMIT:
case ZFS_PROP_FILESYSTEM_COUNT:
case ZFS_PROP_SNAPSHOT_COUNT:
*val = getprop_uint64(zhp, prop, source);
if (*source == NULL) {
/* not default, must be local */
*source = zhp->zfs_name;
}
break;
case ZFS_PROP_MOUNTED:
*val = (zhp->zfs_mntopts != NULL);
break;
case ZFS_PROP_NUMCLONES:
*val = zhp->zfs_dmustats.dds_num_clones;
break;
case ZFS_PROP_VERSION:
case ZFS_PROP_NORMALIZE:
case ZFS_PROP_UTF8ONLY:
case ZFS_PROP_CASE:
if (!zfs_prop_valid_for_type(prop, zhp->zfs_head_type) ||
zcmd_alloc_dst_nvlist(zhp->zfs_hdl, &zc, 0) != 0)
return (-1);
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
if (zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_OBJSET_ZPLPROPS, &zc)) {
zcmd_free_nvlists(&zc);
return (-1);
}
if (zcmd_read_dst_nvlist(zhp->zfs_hdl, &zc, &zplprops) != 0 ||
nvlist_lookup_uint64(zplprops, zfs_prop_to_name(prop),
val) != 0) {
zcmd_free_nvlists(&zc);
return (-1);
}
nvlist_free(zplprops);
zcmd_free_nvlists(&zc);
break;
case ZFS_PROP_INCONSISTENT:
*val = zhp->zfs_dmustats.dds_inconsistent;
break;
default:
switch (zfs_prop_get_type(prop)) {
case PROP_TYPE_NUMBER:
case PROP_TYPE_INDEX:
*val = getprop_uint64(zhp, prop, source);
/*
* If we tried to use a default value for a
* readonly property, it means that it was not
* present. Note this only applies to "truly"
* readonly properties, not set-once properties
* like volblocksize.
*/
if (zfs_prop_readonly(prop) &&
!zfs_prop_setonce(prop) &&
*source != NULL && (*source)[0] == '\0') {
*source = NULL;
return (-1);
}
break;
case PROP_TYPE_STRING:
default:
zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN,
"cannot get non-numeric property"));
return (zfs_error(zhp->zfs_hdl, EZFS_BADPROP,
dgettext(TEXT_DOMAIN, "internal error")));
}
}
return (0);
}
/*
* Calculate the source type, given the raw source string.
*/
static void
get_source(zfs_handle_t *zhp, zprop_source_t *srctype, char *source,
char *statbuf, size_t statlen)
{
if (statbuf == NULL || *srctype == ZPROP_SRC_TEMPORARY)
return;
if (source == NULL) {
*srctype = ZPROP_SRC_NONE;
} else if (source[0] == '\0') {
*srctype = ZPROP_SRC_DEFAULT;
} else if (strstr(source, ZPROP_SOURCE_VAL_RECVD) != NULL) {
*srctype = ZPROP_SRC_RECEIVED;
} else {
if (strcmp(source, zhp->zfs_name) == 0) {
*srctype = ZPROP_SRC_LOCAL;
} else {
(void) strlcpy(statbuf, source, statlen);
*srctype = ZPROP_SRC_INHERITED;
}
}
}
int
zfs_prop_get_recvd(zfs_handle_t *zhp, const char *propname, char *propbuf,
size_t proplen, boolean_t literal)
{
zfs_prop_t prop;
int err = 0;
if (zhp->zfs_recvd_props == NULL)
if (get_recvd_props_ioctl(zhp) != 0)
return (-1);
prop = zfs_name_to_prop(propname);
if (prop != ZPROP_INVAL) {
uint64_t cookie;
if (!nvlist_exists(zhp->zfs_recvd_props, propname))
return (-1);
zfs_set_recvd_props_mode(zhp, &cookie);
err = zfs_prop_get(zhp, prop, propbuf, proplen,
NULL, NULL, 0, literal);
zfs_unset_recvd_props_mode(zhp, &cookie);
} else {
nvlist_t *propval;
char *recvdval;
if (nvlist_lookup_nvlist(zhp->zfs_recvd_props,
propname, &propval) != 0)
return (-1);
verify(nvlist_lookup_string(propval, ZPROP_VALUE,
&recvdval) == 0);
(void) strlcpy(propbuf, recvdval, proplen);
}
return (err == 0 ? 0 : -1);
}
static int
get_clones_string(zfs_handle_t *zhp, char *propbuf, size_t proplen)
{
nvlist_t *value;
nvpair_t *pair;
value = zfs_get_clones_nvl(zhp);
if (value == NULL)
return (-1);
propbuf[0] = '\0';
for (pair = nvlist_next_nvpair(value, NULL); pair != NULL;
pair = nvlist_next_nvpair(value, pair)) {
if (propbuf[0] != '\0')
(void) strlcat(propbuf, ",", proplen);
(void) strlcat(propbuf, nvpair_name(pair), proplen);
}
return (0);
}
struct get_clones_arg {
uint64_t numclones;
nvlist_t *value;
const char *origin;
char buf[ZFS_MAX_DATASET_NAME_LEN];
};
int
get_clones_cb(zfs_handle_t *zhp, void *arg)
{
struct get_clones_arg *gca = arg;
if (gca->numclones == 0) {
zfs_close(zhp);
return (0);
}
if (zfs_prop_get(zhp, ZFS_PROP_ORIGIN, gca->buf, sizeof (gca->buf),
NULL, NULL, 0, B_TRUE) != 0)
goto out;
if (strcmp(gca->buf, gca->origin) == 0) {
fnvlist_add_boolean(gca->value, zfs_get_name(zhp));
gca->numclones--;
}
out:
(void) zfs_iter_children(zhp, get_clones_cb, gca);
zfs_close(zhp);
return (0);
}
nvlist_t *
zfs_get_clones_nvl(zfs_handle_t *zhp)
{
nvlist_t *nv, *value;
if (nvlist_lookup_nvlist(zhp->zfs_props,
zfs_prop_to_name(ZFS_PROP_CLONES), &nv) != 0) {
struct get_clones_arg gca;
/*
* if this is a snapshot, then the kernel wasn't able
* to get the clones. Do it by slowly iterating.
*/
if (zhp->zfs_type != ZFS_TYPE_SNAPSHOT)
return (NULL);
if (nvlist_alloc(&nv, NV_UNIQUE_NAME, 0) != 0)
return (NULL);
if (nvlist_alloc(&value, NV_UNIQUE_NAME, 0) != 0) {
nvlist_free(nv);
return (NULL);
}
gca.numclones = zfs_prop_get_int(zhp, ZFS_PROP_NUMCLONES);
gca.value = value;
gca.origin = zhp->zfs_name;
if (gca.numclones != 0) {
zfs_handle_t *root;
char pool[ZFS_MAX_DATASET_NAME_LEN];
char *cp = pool;
/* get the pool name */
(void) strlcpy(pool, zhp->zfs_name, sizeof (pool));
(void) strsep(&cp, "/@");
root = zfs_open(zhp->zfs_hdl, pool,
ZFS_TYPE_FILESYSTEM);
(void) get_clones_cb(root, &gca);
}
if (gca.numclones != 0 ||
nvlist_add_nvlist(nv, ZPROP_VALUE, value) != 0 ||
nvlist_add_nvlist(zhp->zfs_props,
zfs_prop_to_name(ZFS_PROP_CLONES), nv) != 0) {
nvlist_free(nv);
nvlist_free(value);
return (NULL);
}
nvlist_free(nv);
nvlist_free(value);
verify(0 == nvlist_lookup_nvlist(zhp->zfs_props,
zfs_prop_to_name(ZFS_PROP_CLONES), &nv));
}
verify(nvlist_lookup_nvlist(nv, ZPROP_VALUE, &value) == 0);
return (value);
}
/*
* Accepts a property and value and checks that the value
* matches the one found by the channel program. If they are
* not equal, print both of them.
*/
void
zcp_check(zfs_handle_t *zhp, zfs_prop_t prop, uint64_t intval,
const char *strval)
{
if (!zhp->zfs_hdl->libzfs_prop_debug)
return;
int error;
char *poolname = zhp->zpool_hdl->zpool_name;
const char *program =
"args = ...\n"
"ds = args['dataset']\n"
"prop = args['property']\n"
"value, setpoint = zfs.get_prop(ds, prop)\n"
"return {value=value, setpoint=setpoint}\n";
nvlist_t *outnvl;
nvlist_t *retnvl;
nvlist_t *argnvl = fnvlist_alloc();
fnvlist_add_string(argnvl, "dataset", zhp->zfs_name);
fnvlist_add_string(argnvl, "property", zfs_prop_to_name(prop));
error = lzc_channel_program_nosync(poolname, program,
10 * 1000 * 1000, 10 * 1024 * 1024, argnvl, &outnvl);
if (error == 0) {
retnvl = fnvlist_lookup_nvlist(outnvl, "return");
if (zfs_prop_get_type(prop) == PROP_TYPE_NUMBER) {
int64_t ans;
error = nvlist_lookup_int64(retnvl, "value", &ans);
if (error != 0) {
(void) fprintf(stderr, "zcp check error: %u\n",
error);
return;
}
if (ans != intval) {
(void) fprintf(stderr,
"%s: zfs found %lld, but zcp found %lld\n",
zfs_prop_to_name(prop),
(longlong_t)intval, (longlong_t)ans);
}
} else {
char *str_ans;
error = nvlist_lookup_string(retnvl, "value", &str_ans);
if (error != 0) {
(void) fprintf(stderr, "zcp check error: %u\n",
error);
return;
}
if (strcmp(strval, str_ans) != 0) {
(void) fprintf(stderr,
"%s: zfs found %s, but zcp found %s\n",
zfs_prop_to_name(prop),
strval, str_ans);
}
}
} else {
(void) fprintf(stderr,
"zcp check failed, channel program error: %u\n", error);
}
nvlist_free(argnvl);
nvlist_free(outnvl);
}
/*
* Retrieve a property from the given object. If 'literal' is specified, then
* numbers are left as exact values. Otherwise, numbers are converted to a
* human-readable form.
*
* Returns 0 on success, or -1 on error.
*/
int
zfs_prop_get(zfs_handle_t *zhp, zfs_prop_t prop, char *propbuf, size_t proplen,
zprop_source_t *src, char *statbuf, size_t statlen, boolean_t literal)
{
char *source = NULL;
uint64_t val;
const char *str;
const char *strval;
boolean_t received = zfs_is_recvd_props_mode(zhp);
/*
* Check to see if this property applies to our object
*/
if (!zfs_prop_valid_for_type(prop, zhp->zfs_type))
return (-1);
if (received && zfs_prop_readonly(prop))
return (-1);
if (src)
*src = ZPROP_SRC_NONE;
switch (prop) {
case ZFS_PROP_CREATION:
/*
* 'creation' is a time_t stored in the statistics. We convert
* this into a string unless 'literal' is specified.
*/
{
val = getprop_uint64(zhp, prop, &source);
time_t time = (time_t)val;
struct tm t;
if (literal ||
localtime_r(&time, &t) == NULL ||
strftime(propbuf, proplen, "%a %b %e %k:%M %Y",
&t) == 0)
(void) snprintf(propbuf, proplen, "%llu", val);
}
zcp_check(zhp, prop, val, NULL);
break;
case ZFS_PROP_MOUNTPOINT:
/*
* Getting the precise mountpoint can be tricky.
*
* - for 'none' or 'legacy', return those values.
* - for inherited mountpoints, we want to take everything
* after our ancestor and append it to the inherited value.
*
* If the pool has an alternate root, we want to prepend that
* root to any values we return.
*/
str = getprop_string(zhp, prop, &source);
if (str[0] == '/') {
char buf[MAXPATHLEN];
char *root = buf;
const char *relpath;
/*
* If we inherit the mountpoint, even from a dataset
* with a received value, the source will be the path of
* the dataset we inherit from. If source is
* ZPROP_SOURCE_VAL_RECVD, the received value is not
* inherited.
*/
if (strcmp(source, ZPROP_SOURCE_VAL_RECVD) == 0) {
relpath = "";
} else {
relpath = zhp->zfs_name + strlen(source);
if (relpath[0] == '/')
relpath++;
}
if ((zpool_get_prop(zhp->zpool_hdl,
ZPOOL_PROP_ALTROOT, buf, MAXPATHLEN, NULL,
B_FALSE)) || (strcmp(root, "-") == 0))
root[0] = '\0';
/*
* Special case an alternate root of '/'. This will
* avoid having multiple leading slashes in the
* mountpoint path.
*/
if (strcmp(root, "/") == 0)
root++;
/*
* If the mountpoint is '/' then skip over this
* if we are obtaining either an alternate root or
* an inherited mountpoint.
*/
if (str[1] == '\0' && (root[0] != '\0' ||
relpath[0] != '\0'))
str++;
if (relpath[0] == '\0')
(void) snprintf(propbuf, proplen, "%s%s",
root, str);
else
(void) snprintf(propbuf, proplen, "%s%s%s%s",
root, str, relpath[0] == '@' ? "" : "/",
relpath);
} else {
/* 'legacy' or 'none' */
(void) strlcpy(propbuf, str, proplen);
}
zcp_check(zhp, prop, NULL, propbuf);
break;
case ZFS_PROP_ORIGIN:
str = getprop_string(zhp, prop, &source);
if (str == NULL)
return (-1);
(void) strlcpy(propbuf, str, proplen);
zcp_check(zhp, prop, NULL, str);
break;
case ZFS_PROP_CLONES:
if (get_clones_string(zhp, propbuf, proplen) != 0)
return (-1);
break;
case ZFS_PROP_QUOTA:
case ZFS_PROP_REFQUOTA:
case ZFS_PROP_RESERVATION:
case ZFS_PROP_REFRESERVATION:
if (get_numeric_property(zhp, prop, src, &source, &val) != 0)
return (-1);
/*
* If quota or reservation is 0, we translate this into 'none'
* (unless literal is set), and indicate that it's the default
* value. Otherwise, we print the number nicely and indicate
* that its set locally.
*/
if (val == 0) {
if (literal)
(void) strlcpy(propbuf, "0", proplen);
else
(void) strlcpy(propbuf, "none", proplen);
} else {
if (literal)
(void) snprintf(propbuf, proplen, "%llu",
(u_longlong_t)val);
else
zfs_nicenum(val, propbuf, proplen);
}
zcp_check(zhp, prop, val, NULL);
break;
case ZFS_PROP_FILESYSTEM_LIMIT:
case ZFS_PROP_SNAPSHOT_LIMIT:
case ZFS_PROP_FILESYSTEM_COUNT:
case ZFS_PROP_SNAPSHOT_COUNT:
if (get_numeric_property(zhp, prop, src, &source, &val) != 0)
return (-1);
/*
* If limit is UINT64_MAX, we translate this into 'none' (unless
* literal is set), and indicate that it's the default value.
* Otherwise, we print the number nicely and indicate that it's
* set locally.
*/
if (literal) {
(void) snprintf(propbuf, proplen, "%llu",
(u_longlong_t)val);
} else if (val == UINT64_MAX) {
(void) strlcpy(propbuf, "none", proplen);
} else {
zfs_nicenum(val, propbuf, proplen);
}
zcp_check(zhp, prop, val, NULL);
break;
case ZFS_PROP_REFRATIO:
case ZFS_PROP_COMPRESSRATIO:
if (get_numeric_property(zhp, prop, src, &source, &val) != 0)
return (-1);
(void) snprintf(propbuf, proplen, "%llu.%02llux",
(u_longlong_t)(val / 100),
(u_longlong_t)(val % 100));
zcp_check(zhp, prop, val, NULL);
break;
case ZFS_PROP_TYPE:
switch (zhp->zfs_type) {
case ZFS_TYPE_FILESYSTEM:
str = "filesystem";
break;
case ZFS_TYPE_VOLUME:
str = "volume";
break;
case ZFS_TYPE_SNAPSHOT:
str = "snapshot";
break;
case ZFS_TYPE_BOOKMARK:
str = "bookmark";
break;
default:
abort();
}
(void) snprintf(propbuf, proplen, "%s", str);
zcp_check(zhp, prop, NULL, propbuf);
break;
case ZFS_PROP_MOUNTED:
/*
* The 'mounted' property is a pseudo-property that described
* whether the filesystem is currently mounted. Even though
* it's a boolean value, the typical values of "on" and "off"
* don't make sense, so we translate to "yes" and "no".
*/
if (get_numeric_property(zhp, ZFS_PROP_MOUNTED,
src, &source, &val) != 0)
return (-1);
if (val)
(void) strlcpy(propbuf, "yes", proplen);
else
(void) strlcpy(propbuf, "no", proplen);
break;
case ZFS_PROP_NAME:
/*
* The 'name' property is a pseudo-property derived from the
* dataset name. It is presented as a real property to simplify
* consumers.
*/
(void) strlcpy(propbuf, zhp->zfs_name, proplen);
zcp_check(zhp, prop, NULL, propbuf);
break;
case ZFS_PROP_MLSLABEL:
{
#ifdef illumos
m_label_t *new_sl = NULL;
char *ascii = NULL; /* human readable label */
(void) strlcpy(propbuf,
getprop_string(zhp, prop, &source), proplen);
if (literal || (strcasecmp(propbuf,
ZFS_MLSLABEL_DEFAULT) == 0))
break;
/*
* Try to translate the internal hex string to
* human-readable output. If there are any
* problems just use the hex string.
*/
if (str_to_label(propbuf, &new_sl, MAC_LABEL,
L_NO_CORRECTION, NULL) == -1) {
m_label_free(new_sl);
break;
}
if (label_to_str(new_sl, &ascii, M_LABEL,
DEF_NAMES) != 0) {
if (ascii)
free(ascii);
m_label_free(new_sl);
break;
}
m_label_free(new_sl);
(void) strlcpy(propbuf, ascii, proplen);
free(ascii);
#else /* !illumos */
propbuf[0] = '\0';
#endif /* illumos */
}
break;
case ZFS_PROP_GUID:
case ZFS_PROP_CREATETXG:
/*
* GUIDs are stored as numbers, but they are identifiers.
* We don't want them to be pretty printed, because pretty
* printing mangles the ID into a truncated and useless value.
*/
if (get_numeric_property(zhp, prop, src, &source, &val) != 0)
return (-1);
(void) snprintf(propbuf, proplen, "%llu", (u_longlong_t)val);
zcp_check(zhp, prop, val, NULL);
break;
default:
switch (zfs_prop_get_type(prop)) {
case PROP_TYPE_NUMBER:
if (get_numeric_property(zhp, prop, src,
&source, &val) != 0) {
return (-1);
}
if (literal) {
(void) snprintf(propbuf, proplen, "%llu",
(u_longlong_t)val);
} else {
zfs_nicenum(val, propbuf, proplen);
}
zcp_check(zhp, prop, val, NULL);
break;
case PROP_TYPE_STRING:
str = getprop_string(zhp, prop, &source);
if (str == NULL)
return (-1);
(void) strlcpy(propbuf, str, proplen);
zcp_check(zhp, prop, NULL, str);
break;
case PROP_TYPE_INDEX:
if (get_numeric_property(zhp, prop, src,
&source, &val) != 0)
return (-1);
if (zfs_prop_index_to_string(prop, val, &strval) != 0)
return (-1);
(void) strlcpy(propbuf, strval, proplen);
zcp_check(zhp, prop, NULL, strval);
break;
default:
abort();
}
}
get_source(zhp, src, source, statbuf, statlen);
return (0);
}
/*
* Utility function to get the given numeric property. Does no validation that
* the given property is the appropriate type; should only be used with
* hard-coded property types.
*/
uint64_t
zfs_prop_get_int(zfs_handle_t *zhp, zfs_prop_t prop)
{
char *source;
uint64_t val;
(void) get_numeric_property(zhp, prop, NULL, &source, &val);
return (val);
}
int
zfs_prop_set_int(zfs_handle_t *zhp, zfs_prop_t prop, uint64_t val)
{
char buf[64];
(void) snprintf(buf, sizeof (buf), "%llu", (longlong_t)val);
return (zfs_prop_set(zhp, zfs_prop_to_name(prop), buf));
}
/*
* Similar to zfs_prop_get(), but returns the value as an integer.
*/
int
zfs_prop_get_numeric(zfs_handle_t *zhp, zfs_prop_t prop, uint64_t *value,
zprop_source_t *src, char *statbuf, size_t statlen)
{
char *source;
/*
* Check to see if this property applies to our object
*/
if (!zfs_prop_valid_for_type(prop, zhp->zfs_type)) {
return (zfs_error_fmt(zhp->zfs_hdl, EZFS_PROPTYPE,
dgettext(TEXT_DOMAIN, "cannot get property '%s'"),
zfs_prop_to_name(prop)));
}
if (src)
*src = ZPROP_SRC_NONE;
if (get_numeric_property(zhp, prop, src, &source, value) != 0)
return (-1);
get_source(zhp, src, source, statbuf, statlen);
return (0);
}
static int
idmap_id_to_numeric_domain_rid(uid_t id, boolean_t isuser,
char **domainp, idmap_rid_t *ridp)
{
#ifdef illumos
idmap_get_handle_t *get_hdl = NULL;
idmap_stat status;
int err = EINVAL;
if (idmap_get_create(&get_hdl) != IDMAP_SUCCESS)
goto out;
if (isuser) {
err = idmap_get_sidbyuid(get_hdl, id,
IDMAP_REQ_FLG_USE_CACHE, domainp, ridp, &status);
} else {
err = idmap_get_sidbygid(get_hdl, id,
IDMAP_REQ_FLG_USE_CACHE, domainp, ridp, &status);
}
if (err == IDMAP_SUCCESS &&
idmap_get_mappings(get_hdl) == IDMAP_SUCCESS &&
status == IDMAP_SUCCESS)
err = 0;
else
err = EINVAL;
out:
if (get_hdl)
idmap_get_destroy(get_hdl);
return (err);
#else /* !illumos */
assert(!"invalid code path");
return (EINVAL); // silence compiler warning
#endif /* illumos */
}
/*
* convert the propname into parameters needed by kernel
* Eg: userquota@ahrens -> ZFS_PROP_USERQUOTA, "", 126829
* Eg: userused@matt@domain -> ZFS_PROP_USERUSED, "S-1-123-456", 789
*/
static int
userquota_propname_decode(const char *propname, boolean_t zoned,
zfs_userquota_prop_t *typep, char *domain, int domainlen, uint64_t *ridp)
{
zfs_userquota_prop_t type;
char *cp, *end;
char *numericsid = NULL;
boolean_t isuser;
domain[0] = '\0';
*ridp = 0;
/* Figure out the property type ({user|group}{quota|space}) */
for (type = 0; type < ZFS_NUM_USERQUOTA_PROPS; type++) {
if (strncmp(propname, zfs_userquota_prop_prefixes[type],
strlen(zfs_userquota_prop_prefixes[type])) == 0)
break;
}
if (type == ZFS_NUM_USERQUOTA_PROPS)
return (EINVAL);
*typep = type;
isuser = (type == ZFS_PROP_USERQUOTA ||
type == ZFS_PROP_USERUSED);
cp = strchr(propname, '@') + 1;
if (strchr(cp, '@')) {
#ifdef illumos
/*
* It's a SID name (eg "user@domain") that needs to be
* turned into S-1-domainID-RID.
*/
int flag = 0;
idmap_stat stat, map_stat;
uid_t pid;
idmap_rid_t rid;
idmap_get_handle_t *gh = NULL;
stat = idmap_get_create(&gh);
if (stat != IDMAP_SUCCESS) {
idmap_get_destroy(gh);
return (ENOMEM);
}
if (zoned && getzoneid() == GLOBAL_ZONEID)
return (ENOENT);
if (isuser) {
stat = idmap_getuidbywinname(cp, NULL, flag, &pid);
if (stat < 0)
return (ENOENT);
stat = idmap_get_sidbyuid(gh, pid, flag, &numericsid,
&rid, &map_stat);
} else {
stat = idmap_getgidbywinname(cp, NULL, flag, &pid);
if (stat < 0)
return (ENOENT);
stat = idmap_get_sidbygid(gh, pid, flag, &numericsid,
&rid, &map_stat);
}
if (stat < 0) {
idmap_get_destroy(gh);
return (ENOENT);
}
stat = idmap_get_mappings(gh);
idmap_get_destroy(gh);
if (stat < 0) {
return (ENOENT);
}
if (numericsid == NULL)
return (ENOENT);
cp = numericsid;
*ridp = rid;
/* will be further decoded below */
#else /* !illumos */
return (ENOENT);
#endif /* illumos */
}
if (strncmp(cp, "S-1-", 4) == 0) {
/* It's a numeric SID (eg "S-1-234-567-89") */
(void) strlcpy(domain, cp, domainlen);
errno = 0;
if (*ridp == 0) {
cp = strrchr(domain, '-');
*cp = '\0';
cp++;
*ridp = strtoull(cp, &end, 10);
} else {
end = "";
}
if (numericsid) {
free(numericsid);
numericsid = NULL;
}
if (errno != 0 || *end != '\0')
return (EINVAL);
} else if (!isdigit(*cp)) {
/*
* It's a user/group name (eg "user") that needs to be
* turned into a uid/gid
*/
if (zoned && getzoneid() == GLOBAL_ZONEID)
return (ENOENT);
if (isuser) {
struct passwd *pw;
pw = getpwnam(cp);
if (pw == NULL)
return (ENOENT);
*ridp = pw->pw_uid;
} else {
struct group *gr;
gr = getgrnam(cp);
if (gr == NULL)
return (ENOENT);
*ridp = gr->gr_gid;
}
} else {
/* It's a user/group ID (eg "12345"). */
uid_t id = strtoul(cp, &end, 10);
idmap_rid_t rid;
char *mapdomain;
if (*end != '\0')
return (EINVAL);
if (id > MAXUID) {
/* It's an ephemeral ID. */
if (idmap_id_to_numeric_domain_rid(id, isuser,
&mapdomain, &rid) != 0)
return (ENOENT);
(void) strlcpy(domain, mapdomain, domainlen);
*ridp = rid;
} else {
*ridp = id;
}
}
ASSERT3P(numericsid, ==, NULL);
return (0);
}
static int
zfs_prop_get_userquota_common(zfs_handle_t *zhp, const char *propname,
uint64_t *propvalue, zfs_userquota_prop_t *typep)
{
int err;
zfs_cmd_t zc = { 0 };
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
err = userquota_propname_decode(propname,
zfs_prop_get_int(zhp, ZFS_PROP_ZONED),
typep, zc.zc_value, sizeof (zc.zc_value), &zc.zc_guid);
zc.zc_objset_type = *typep;
if (err)
return (err);
err = ioctl(zhp->zfs_hdl->libzfs_fd, ZFS_IOC_USERSPACE_ONE, &zc);
if (err)
return (err);
*propvalue = zc.zc_cookie;
return (0);
}
int
zfs_prop_get_userquota_int(zfs_handle_t *zhp, const char *propname,
uint64_t *propvalue)
{
zfs_userquota_prop_t type;
return (zfs_prop_get_userquota_common(zhp, propname, propvalue,
&type));
}
int
zfs_prop_get_userquota(zfs_handle_t *zhp, const char *propname,
char *propbuf, int proplen, boolean_t literal)
{
int err;
uint64_t propvalue;
zfs_userquota_prop_t type;
err = zfs_prop_get_userquota_common(zhp, propname, &propvalue,
&type);
if (err)
return (err);
if (literal) {
(void) snprintf(propbuf, proplen, "%llu", propvalue);
} else if (propvalue == 0 &&
(type == ZFS_PROP_USERQUOTA || type == ZFS_PROP_GROUPQUOTA)) {
(void) strlcpy(propbuf, "none", proplen);
} else {
zfs_nicenum(propvalue, propbuf, proplen);
}
return (0);
}
int
zfs_prop_get_written_int(zfs_handle_t *zhp, const char *propname,
uint64_t *propvalue)
{
int err;
zfs_cmd_t zc = { 0 };
const char *snapname;
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
snapname = strchr(propname, '@') + 1;
if (strchr(snapname, '@')) {
(void) strlcpy(zc.zc_value, snapname, sizeof (zc.zc_value));
} else {
/* snapname is the short name, append it to zhp's fsname */
char *cp;
(void) strlcpy(zc.zc_value, zhp->zfs_name,
sizeof (zc.zc_value));
cp = strchr(zc.zc_value, '@');
if (cp != NULL)
*cp = '\0';
(void) strlcat(zc.zc_value, "@", sizeof (zc.zc_value));
(void) strlcat(zc.zc_value, snapname, sizeof (zc.zc_value));
}
err = ioctl(zhp->zfs_hdl->libzfs_fd, ZFS_IOC_SPACE_WRITTEN, &zc);
if (err)
return (err);
*propvalue = zc.zc_cookie;
return (0);
}
int
zfs_prop_get_written(zfs_handle_t *zhp, const char *propname,
char *propbuf, int proplen, boolean_t literal)
{
int err;
uint64_t propvalue;
err = zfs_prop_get_written_int(zhp, propname, &propvalue);
if (err)
return (err);
if (literal) {
(void) snprintf(propbuf, proplen, "%llu", propvalue);
} else {
zfs_nicenum(propvalue, propbuf, proplen);
}
return (0);
}
/*
* Returns the name of the given zfs handle.
*/
const char *
zfs_get_name(const zfs_handle_t *zhp)
{
return (zhp->zfs_name);
}
/*
* Returns the name of the parent pool for the given zfs handle.
*/
const char *
zfs_get_pool_name(const zfs_handle_t *zhp)
{
return (zhp->zpool_hdl->zpool_name);
}
/*
* Returns the type of the given zfs handle.
*/
zfs_type_t
zfs_get_type(const zfs_handle_t *zhp)
{
return (zhp->zfs_type);
}
/*
* Is one dataset name a child dataset of another?
*
* Needs to handle these cases:
* Dataset 1 "a/foo" "a/foo" "a/foo" "a/foo"
* Dataset 2 "a/fo" "a/foobar" "a/bar/baz" "a/foo/bar"
* Descendant? No. No. No. Yes.
*/
static boolean_t
is_descendant(const char *ds1, const char *ds2)
{
size_t d1len = strlen(ds1);
/* ds2 can't be a descendant if it's smaller */
if (strlen(ds2) < d1len)
return (B_FALSE);
/* otherwise, compare strings and verify that there's a '/' char */
return (ds2[d1len] == '/' && (strncmp(ds1, ds2, d1len) == 0));
}
/*
* Given a complete name, return just the portion that refers to the parent.
* Will return -1 if there is no parent (path is just the name of the
* pool).
*/
static int
parent_name(const char *path, char *buf, size_t buflen)
{
char *slashp;
(void) strlcpy(buf, path, buflen);
if ((slashp = strrchr(buf, '/')) == NULL)
return (-1);
*slashp = '\0';
return (0);
}
/*
* If accept_ancestor is false, then check to make sure that the given path has
* a parent, and that it exists. If accept_ancestor is true, then find the
* closest existing ancestor for the given path. In prefixlen return the
* length of already existing prefix of the given path. We also fetch the
* 'zoned' property, which is used to validate property settings when creating
* new datasets.
*/
static int
check_parents(libzfs_handle_t *hdl, const char *path, uint64_t *zoned,
boolean_t accept_ancestor, int *prefixlen)
{
zfs_cmd_t zc = { 0 };
char parent[ZFS_MAX_DATASET_NAME_LEN];
char *slash;
zfs_handle_t *zhp;
char errbuf[1024];
uint64_t is_zoned;
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot create '%s'"), path);
/* get parent, and check to see if this is just a pool */
if (parent_name(path, parent, sizeof (parent)) != 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"missing dataset name"));
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
}
/* check to see if the pool exists */
if ((slash = strchr(parent, '/')) == NULL)
slash = parent + strlen(parent);
(void) strncpy(zc.zc_name, parent, slash - parent);
zc.zc_name[slash - parent] = '\0';
if (ioctl(hdl->libzfs_fd, ZFS_IOC_OBJSET_STATS, &zc) != 0 &&
errno == ENOENT) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"no such pool '%s'"), zc.zc_name);
return (zfs_error(hdl, EZFS_NOENT, errbuf));
}
/* check to see if the parent dataset exists */
while ((zhp = make_dataset_handle(hdl, parent)) == NULL) {
if (errno == ENOENT && accept_ancestor) {
/*
* Go deeper to find an ancestor, give up on top level.
*/
if (parent_name(parent, parent, sizeof (parent)) != 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"no such pool '%s'"), zc.zc_name);
return (zfs_error(hdl, EZFS_NOENT, errbuf));
}
} else if (errno == ENOENT) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"parent does not exist"));
return (zfs_error(hdl, EZFS_NOENT, errbuf));
} else
return (zfs_standard_error(hdl, errno, errbuf));
}
is_zoned = zfs_prop_get_int(zhp, ZFS_PROP_ZONED);
if (zoned != NULL)
*zoned = is_zoned;
/* we are in a non-global zone, but parent is in the global zone */
if (getzoneid() != GLOBAL_ZONEID && !is_zoned) {
(void) zfs_standard_error(hdl, EPERM, errbuf);
zfs_close(zhp);
return (-1);
}
/* make sure parent is a filesystem */
if (zfs_get_type(zhp) != ZFS_TYPE_FILESYSTEM) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"parent is not a filesystem"));
(void) zfs_error(hdl, EZFS_BADTYPE, errbuf);
zfs_close(zhp);
return (-1);
}
zfs_close(zhp);
if (prefixlen != NULL)
*prefixlen = strlen(parent);
return (0);
}
/*
* Finds whether the dataset of the given type(s) exists.
*/
boolean_t
zfs_dataset_exists(libzfs_handle_t *hdl, const char *path, zfs_type_t types)
{
zfs_handle_t *zhp;
if (!zfs_validate_name(hdl, path, types, B_FALSE))
return (B_FALSE);
/*
* Try to get stats for the dataset, which will tell us if it exists.
*/
if ((zhp = make_dataset_handle(hdl, path)) != NULL) {
int ds_type = zhp->zfs_type;
zfs_close(zhp);
if (types & ds_type)
return (B_TRUE);
}
return (B_FALSE);
}
/*
* Given a path to 'target', create all the ancestors between
* the prefixlen portion of the path, and the target itself.
* Fail if the initial prefixlen-ancestor does not already exist.
*/
int
create_parents(libzfs_handle_t *hdl, char *target, int prefixlen)
{
zfs_handle_t *h;
char *cp;
const char *opname;
/* make sure prefix exists */
cp = target + prefixlen;
if (*cp != '/') {
assert(strchr(cp, '/') == NULL);
h = zfs_open(hdl, target, ZFS_TYPE_FILESYSTEM);
} else {
*cp = '\0';
h = zfs_open(hdl, target, ZFS_TYPE_FILESYSTEM);
*cp = '/';
}
if (h == NULL)
return (-1);
zfs_close(h);
/*
* Attempt to create, mount, and share any ancestor filesystems,
* up to the prefixlen-long one.
*/
for (cp = target + prefixlen + 1;
(cp = strchr(cp, '/')) != NULL; *cp = '/', cp++) {
*cp = '\0';
h = make_dataset_handle(hdl, target);
if (h) {
/* it already exists, nothing to do here */
zfs_close(h);
continue;
}
if (zfs_create(hdl, target, ZFS_TYPE_FILESYSTEM,
NULL) != 0) {
opname = dgettext(TEXT_DOMAIN, "create");
goto ancestorerr;
}
h = zfs_open(hdl, target, ZFS_TYPE_FILESYSTEM);
if (h == NULL) {
opname = dgettext(TEXT_DOMAIN, "open");
goto ancestorerr;
}
if (zfs_mount(h, NULL, 0) != 0) {
opname = dgettext(TEXT_DOMAIN, "mount");
goto ancestorerr;
}
if (zfs_share(h) != 0) {
opname = dgettext(TEXT_DOMAIN, "share");
goto ancestorerr;
}
zfs_close(h);
}
return (0);
ancestorerr:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"failed to %s ancestor '%s'"), opname, target);
return (-1);
}
/*
* Creates non-existing ancestors of the given path.
*/
int
zfs_create_ancestors(libzfs_handle_t *hdl, const char *path)
{
int prefix;
char *path_copy;
char errbuf[1024];
int rc = 0;
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot create '%s'"), path);
/*
* Check that we are not passing the nesting limit
* before we start creating any ancestors.
*/
if (dataset_nestcheck(path) != 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"maximum name nesting depth exceeded"));
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
}
if (check_parents(hdl, path, NULL, B_TRUE, &prefix) != 0)
return (-1);
if ((path_copy = strdup(path)) != NULL) {
rc = create_parents(hdl, path_copy, prefix);
free(path_copy);
}
if (path_copy == NULL || rc != 0)
return (-1);
return (0);
}
/*
* Create a new filesystem or volume.
*/
int
zfs_create(libzfs_handle_t *hdl, const char *path, zfs_type_t type,
nvlist_t *props)
{
int ret;
uint64_t size = 0;
uint64_t blocksize = zfs_prop_default_numeric(ZFS_PROP_VOLBLOCKSIZE);
char errbuf[1024];
uint64_t zoned;
enum lzc_dataset_type ost;
zpool_handle_t *zpool_handle;
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot create '%s'"), path);
/* validate the path, taking care to note the extended error message */
if (!zfs_validate_name(hdl, path, type, B_TRUE))
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
if (dataset_nestcheck(path) != 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"maximum name nesting depth exceeded"));
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
}
/* validate parents exist */
if (check_parents(hdl, path, &zoned, B_FALSE, NULL) != 0)
return (-1);
/*
* The failure modes when creating a dataset of a different type over
* one that already exists is a little strange. In particular, if you
* try to create a dataset on top of an existing dataset, the ioctl()
* will return ENOENT, not EEXIST. To prevent this from happening, we
* first try to see if the dataset exists.
*/
if (zfs_dataset_exists(hdl, path, ZFS_TYPE_DATASET)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"dataset already exists"));
return (zfs_error(hdl, EZFS_EXISTS, errbuf));
}
if (type == ZFS_TYPE_VOLUME)
ost = LZC_DATSET_TYPE_ZVOL;
else
ost = LZC_DATSET_TYPE_ZFS;
/* open zpool handle for prop validation */
char pool_path[ZFS_MAX_DATASET_NAME_LEN];
(void) strlcpy(pool_path, path, sizeof (pool_path));
/* truncate pool_path at first slash */
char *p = strchr(pool_path, '/');
if (p != NULL)
*p = '\0';
if ((zpool_handle = zpool_open(hdl, pool_path)) == NULL)
return (-1);
if (props && (props = zfs_valid_proplist(hdl, type, props,
zoned, NULL, zpool_handle, errbuf)) == 0) {
zpool_close(zpool_handle);
return (-1);
}
zpool_close(zpool_handle);
if (type == ZFS_TYPE_VOLUME) {
/*
* If we are creating a volume, the size and block size must
* satisfy a few restraints. First, the blocksize must be a
* valid block size between SPA_{MIN,MAX}BLOCKSIZE. Second, the
* volsize must be a multiple of the block size, and cannot be
* zero.
*/
if (props == NULL || nvlist_lookup_uint64(props,
zfs_prop_to_name(ZFS_PROP_VOLSIZE), &size) != 0) {
nvlist_free(props);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"missing volume size"));
return (zfs_error(hdl, EZFS_BADPROP, errbuf));
}
if ((ret = nvlist_lookup_uint64(props,
zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE),
&blocksize)) != 0) {
if (ret == ENOENT) {
blocksize = zfs_prop_default_numeric(
ZFS_PROP_VOLBLOCKSIZE);
} else {
nvlist_free(props);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"missing volume block size"));
return (zfs_error(hdl, EZFS_BADPROP, errbuf));
}
}
if (size == 0) {
nvlist_free(props);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"volume size cannot be zero"));
return (zfs_error(hdl, EZFS_BADPROP, errbuf));
}
if (size % blocksize != 0) {
nvlist_free(props);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"volume size must be a multiple of volume block "
"size"));
return (zfs_error(hdl, EZFS_BADPROP, errbuf));
}
}
/* create the dataset */
ret = lzc_create(path, ost, props);
nvlist_free(props);
/* check for failure */
if (ret != 0) {
char parent[ZFS_MAX_DATASET_NAME_LEN];
(void) parent_name(path, parent, sizeof (parent));
switch (errno) {
case ENOENT:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"no such parent '%s'"), parent);
return (zfs_error(hdl, EZFS_NOENT, errbuf));
- case EINVAL:
- zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
- "parent '%s' is not a filesystem"), parent);
- return (zfs_error(hdl, EZFS_BADTYPE, errbuf));
-
case ENOTSUP:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool must be upgraded to set this "
"property or value"));
return (zfs_error(hdl, EZFS_BADVERSION, errbuf));
case ERANGE:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"invalid property value(s) specified"));
return (zfs_error(hdl, EZFS_BADPROP, errbuf));
#ifdef _ILP32
case EOVERFLOW:
/*
* This platform can't address a volume this big.
*/
if (type == ZFS_TYPE_VOLUME)
return (zfs_error(hdl, EZFS_VOLTOOBIG,
errbuf));
#endif
/* FALLTHROUGH */
default:
return (zfs_standard_error(hdl, errno, errbuf));
}
}
return (0);
}
/*
* Destroys the given dataset. The caller must make sure that the filesystem
* isn't mounted, and that there are no active dependents. If the file system
* does not exist this function does nothing.
*/
int
zfs_destroy(zfs_handle_t *zhp, boolean_t defer)
{
int error;
if (zhp->zfs_type != ZFS_TYPE_SNAPSHOT && defer)
return (EINVAL);
if (zhp->zfs_type == ZFS_TYPE_BOOKMARK) {
nvlist_t *nv = fnvlist_alloc();
fnvlist_add_boolean(nv, zhp->zfs_name);
error = lzc_destroy_bookmarks(nv, NULL);
fnvlist_free(nv);
if (error != 0) {
return (zfs_standard_error_fmt(zhp->zfs_hdl, error,
dgettext(TEXT_DOMAIN, "cannot destroy '%s'"),
zhp->zfs_name));
}
return (0);
}
if (zhp->zfs_type == ZFS_TYPE_SNAPSHOT) {
nvlist_t *nv = fnvlist_alloc();
fnvlist_add_boolean(nv, zhp->zfs_name);
error = lzc_destroy_snaps(nv, defer, NULL);
fnvlist_free(nv);
} else {
error = lzc_destroy(zhp->zfs_name);
}
if (error != 0 && error != ENOENT) {
return (zfs_standard_error_fmt(zhp->zfs_hdl, errno,
dgettext(TEXT_DOMAIN, "cannot destroy '%s'"),
zhp->zfs_name));
}
remove_mountpoint(zhp);
return (0);
}
struct destroydata {
nvlist_t *nvl;
const char *snapname;
};
static int
zfs_check_snap_cb(zfs_handle_t *zhp, void *arg)
{
struct destroydata *dd = arg;
char name[ZFS_MAX_DATASET_NAME_LEN];
int rv = 0;
(void) snprintf(name, sizeof (name),
"%s@%s", zhp->zfs_name, dd->snapname);
if (lzc_exists(name))
verify(nvlist_add_boolean(dd->nvl, name) == 0);
rv = zfs_iter_filesystems(zhp, zfs_check_snap_cb, dd);
zfs_close(zhp);
return (rv);
}
/*
* Destroys all snapshots with the given name in zhp & descendants.
*/
int
zfs_destroy_snaps(zfs_handle_t *zhp, char *snapname, boolean_t defer)
{
int ret;
struct destroydata dd = { 0 };
dd.snapname = snapname;
verify(nvlist_alloc(&dd.nvl, NV_UNIQUE_NAME, 0) == 0);
(void) zfs_check_snap_cb(zfs_handle_dup(zhp), &dd);
if (nvlist_empty(dd.nvl)) {
ret = zfs_standard_error_fmt(zhp->zfs_hdl, ENOENT,
dgettext(TEXT_DOMAIN, "cannot destroy '%s@%s'"),
zhp->zfs_name, snapname);
} else {
ret = zfs_destroy_snaps_nvl(zhp->zfs_hdl, dd.nvl, defer);
}
nvlist_free(dd.nvl);
return (ret);
}
/*
* Destroys all the snapshots named in the nvlist.
*/
int
zfs_destroy_snaps_nvl(libzfs_handle_t *hdl, nvlist_t *snaps, boolean_t defer)
{
int ret;
nvlist_t *errlist = NULL;
ret = lzc_destroy_snaps(snaps, defer, &errlist);
if (ret == 0) {
nvlist_free(errlist);
return (0);
}
if (nvlist_empty(errlist)) {
char errbuf[1024];
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot destroy snapshots"));
ret = zfs_standard_error(hdl, ret, errbuf);
}
for (nvpair_t *pair = nvlist_next_nvpair(errlist, NULL);
pair != NULL; pair = nvlist_next_nvpair(errlist, pair)) {
char errbuf[1024];
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot destroy snapshot %s"),
nvpair_name(pair));
switch (fnvpair_value_int32(pair)) {
case EEXIST:
zfs_error_aux(hdl,
dgettext(TEXT_DOMAIN, "snapshot is cloned"));
ret = zfs_error(hdl, EZFS_EXISTS, errbuf);
break;
default:
ret = zfs_standard_error(hdl, errno, errbuf);
break;
}
}
nvlist_free(errlist);
return (ret);
}
/*
* Clones the given dataset. The target must be of the same type as the source.
*/
int
zfs_clone(zfs_handle_t *zhp, const char *target, nvlist_t *props)
{
char parent[ZFS_MAX_DATASET_NAME_LEN];
int ret;
char errbuf[1024];
libzfs_handle_t *hdl = zhp->zfs_hdl;
uint64_t zoned;
assert(zhp->zfs_type == ZFS_TYPE_SNAPSHOT);
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot create '%s'"), target);
/* validate the target/clone name */
if (!zfs_validate_name(hdl, target, ZFS_TYPE_FILESYSTEM, B_TRUE))
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
/* validate parents exist */
if (check_parents(hdl, target, &zoned, B_FALSE, NULL) != 0)
return (-1);
(void) parent_name(target, parent, sizeof (parent));
/* do the clone */
if (props) {
zfs_type_t type;
if (ZFS_IS_VOLUME(zhp)) {
type = ZFS_TYPE_VOLUME;
} else {
type = ZFS_TYPE_FILESYSTEM;
}
if ((props = zfs_valid_proplist(hdl, type, props, zoned,
zhp, zhp->zpool_hdl, errbuf)) == NULL)
return (-1);
if (zfs_fix_auto_resv(zhp, props) == -1) {
nvlist_free(props);
return (-1);
}
}
ret = lzc_clone(target, zhp->zfs_name, props);
nvlist_free(props);
if (ret != 0) {
switch (errno) {
case ENOENT:
/*
* The parent doesn't exist. We should have caught this
* above, but there may a race condition that has since
* destroyed the parent.
*
* At this point, we don't know whether it's the source
* that doesn't exist anymore, or whether the target
* dataset doesn't exist.
*/
zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN,
"no such parent '%s'"), parent);
return (zfs_error(zhp->zfs_hdl, EZFS_NOENT, errbuf));
case EXDEV:
zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN,
"source and target pools differ"));
return (zfs_error(zhp->zfs_hdl, EZFS_CROSSTARGET,
errbuf));
default:
return (zfs_standard_error(zhp->zfs_hdl, errno,
errbuf));
}
}
return (ret);
}
/*
* Promotes the given clone fs to be the clone parent.
*/
int
zfs_promote(zfs_handle_t *zhp)
{
libzfs_handle_t *hdl = zhp->zfs_hdl;
char snapname[ZFS_MAX_DATASET_NAME_LEN];
int ret;
char errbuf[1024];
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot promote '%s'"), zhp->zfs_name);
if (zhp->zfs_type == ZFS_TYPE_SNAPSHOT) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"snapshots can not be promoted"));
return (zfs_error(hdl, EZFS_BADTYPE, errbuf));
}
if (zhp->zfs_dmustats.dds_origin[0] == '\0') {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"not a cloned filesystem"));
return (zfs_error(hdl, EZFS_BADTYPE, errbuf));
}
if (!zfs_validate_name(hdl, zhp->zfs_name, zhp->zfs_type, B_TRUE))
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
ret = lzc_promote(zhp->zfs_name, snapname, sizeof (snapname));
if (ret != 0) {
switch (ret) {
case EEXIST:
/* There is a conflicting snapshot name. */
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"conflicting snapshot '%s' from parent '%s'"),
snapname, zhp->zfs_dmustats.dds_origin);
return (zfs_error(hdl, EZFS_EXISTS, errbuf));
default:
return (zfs_standard_error(hdl, ret, errbuf));
}
}
return (ret);
}
typedef struct snapdata {
nvlist_t *sd_nvl;
const char *sd_snapname;
} snapdata_t;
static int
zfs_snapshot_cb(zfs_handle_t *zhp, void *arg)
{
snapdata_t *sd = arg;
char name[ZFS_MAX_DATASET_NAME_LEN];
int rv = 0;
if (zfs_prop_get_int(zhp, ZFS_PROP_INCONSISTENT) == 0) {
(void) snprintf(name, sizeof (name),
"%s@%s", zfs_get_name(zhp), sd->sd_snapname);
fnvlist_add_boolean(sd->sd_nvl, name);
rv = zfs_iter_filesystems(zhp, zfs_snapshot_cb, sd);
}
zfs_close(zhp);
return (rv);
}
int
zfs_remap_indirects(libzfs_handle_t *hdl, const char *fs)
{
int err;
char errbuf[1024];
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot remap dataset '%s'"), fs);
err = lzc_remap(fs);
if (err != 0) {
switch (err) {
case ENOTSUP:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool must be upgraded"));
(void) zfs_error(hdl, EZFS_BADVERSION, errbuf);
break;
case EINVAL:
(void) zfs_error(hdl, EZFS_BADTYPE, errbuf);
break;
default:
(void) zfs_standard_error(hdl, err, errbuf);
break;
}
}
return (err);
}
/*
* Creates snapshots. The keys in the snaps nvlist are the snapshots to be
* created.
*/
int
zfs_snapshot_nvl(libzfs_handle_t *hdl, nvlist_t *snaps, nvlist_t *props)
{
int ret;
char errbuf[1024];
nvpair_t *elem;
nvlist_t *errors;
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot create snapshots "));
elem = NULL;
while ((elem = nvlist_next_nvpair(snaps, elem)) != NULL) {
const char *snapname = nvpair_name(elem);
/* validate the target name */
if (!zfs_validate_name(hdl, snapname, ZFS_TYPE_SNAPSHOT,
B_TRUE)) {
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN,
"cannot create snapshot '%s'"), snapname);
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
}
}
/*
* get pool handle for prop validation. assumes all snaps are in the
* same pool, as does lzc_snapshot (below).
*/
char pool[ZFS_MAX_DATASET_NAME_LEN];
elem = nvlist_next_nvpair(snaps, NULL);
(void) strlcpy(pool, nvpair_name(elem), sizeof (pool));
pool[strcspn(pool, "/@")] = '\0';
zpool_handle_t *zpool_hdl = zpool_open(hdl, pool);
if (props != NULL &&
(props = zfs_valid_proplist(hdl, ZFS_TYPE_SNAPSHOT,
props, B_FALSE, NULL, zpool_hdl, errbuf)) == NULL) {
zpool_close(zpool_hdl);
return (-1);
}
zpool_close(zpool_hdl);
ret = lzc_snapshot(snaps, props, &errors);
if (ret != 0) {
boolean_t printed = B_FALSE;
for (elem = nvlist_next_nvpair(errors, NULL);
elem != NULL;
elem = nvlist_next_nvpair(errors, elem)) {
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN,
"cannot create snapshot '%s'"), nvpair_name(elem));
(void) zfs_standard_error(hdl,
fnvpair_value_int32(elem), errbuf);
printed = B_TRUE;
}
if (!printed) {
switch (ret) {
case EXDEV:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"multiple snapshots of same "
"fs not allowed"));
(void) zfs_error(hdl, EZFS_EXISTS, errbuf);
break;
default:
(void) zfs_standard_error(hdl, ret, errbuf);
}
}
}
nvlist_free(props);
nvlist_free(errors);
return (ret);
}
int
zfs_snapshot(libzfs_handle_t *hdl, const char *path, boolean_t recursive,
nvlist_t *props)
{
int ret;
snapdata_t sd = { 0 };
char fsname[ZFS_MAX_DATASET_NAME_LEN];
char *cp;
zfs_handle_t *zhp;
char errbuf[1024];
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot snapshot %s"), path);
if (!zfs_validate_name(hdl, path, ZFS_TYPE_SNAPSHOT, B_TRUE))
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
(void) strlcpy(fsname, path, sizeof (fsname));
cp = strchr(fsname, '@');
*cp = '\0';
sd.sd_snapname = cp + 1;
if ((zhp = zfs_open(hdl, fsname, ZFS_TYPE_FILESYSTEM |
ZFS_TYPE_VOLUME)) == NULL) {
return (-1);
}
verify(nvlist_alloc(&sd.sd_nvl, NV_UNIQUE_NAME, 0) == 0);
if (recursive) {
(void) zfs_snapshot_cb(zfs_handle_dup(zhp), &sd);
} else {
fnvlist_add_boolean(sd.sd_nvl, path);
}
ret = zfs_snapshot_nvl(hdl, sd.sd_nvl, props);
nvlist_free(sd.sd_nvl);
zfs_close(zhp);
return (ret);
}
/*
* Destroy any more recent snapshots. We invoke this callback on any dependents
* of the snapshot first. If the 'cb_dependent' member is non-zero, then this
* is a dependent and we should just destroy it without checking the transaction
* group.
*/
typedef struct rollback_data {
const char *cb_target; /* the snapshot */
uint64_t cb_create; /* creation time reference */
boolean_t cb_error;
boolean_t cb_force;
} rollback_data_t;
static int
rollback_destroy_dependent(zfs_handle_t *zhp, void *data)
{
rollback_data_t *cbp = data;
prop_changelist_t *clp;
/* We must destroy this clone; first unmount it */
clp = changelist_gather(zhp, ZFS_PROP_NAME, 0,
cbp->cb_force ? MS_FORCE: 0);
if (clp == NULL || changelist_prefix(clp) != 0) {
cbp->cb_error = B_TRUE;
zfs_close(zhp);
return (0);
}
if (zfs_destroy(zhp, B_FALSE) != 0)
cbp->cb_error = B_TRUE;
else
changelist_remove(clp, zhp->zfs_name);
(void) changelist_postfix(clp);
changelist_free(clp);
zfs_close(zhp);
return (0);
}
static int
rollback_destroy(zfs_handle_t *zhp, void *data)
{
rollback_data_t *cbp = data;
if (zfs_prop_get_int(zhp, ZFS_PROP_CREATETXG) > cbp->cb_create) {
cbp->cb_error |= zfs_iter_dependents(zhp, B_FALSE,
rollback_destroy_dependent, cbp);
cbp->cb_error |= zfs_destroy(zhp, B_FALSE);
}
zfs_close(zhp);
return (0);
}
/*
* Given a dataset, rollback to a specific snapshot, discarding any
* data changes since then and making it the active dataset.
*
* Any snapshots and bookmarks more recent than the target are
* destroyed, along with their dependents (i.e. clones).
*/
int
zfs_rollback(zfs_handle_t *zhp, zfs_handle_t *snap, boolean_t force)
{
rollback_data_t cb = { 0 };
int err;
boolean_t restore_resv = 0;
uint64_t min_txg = 0, old_volsize = 0, new_volsize;
zfs_prop_t resv_prop;
assert(zhp->zfs_type == ZFS_TYPE_FILESYSTEM ||
zhp->zfs_type == ZFS_TYPE_VOLUME);
/*
* Destroy all recent snapshots and their dependents.
*/
cb.cb_force = force;
cb.cb_target = snap->zfs_name;
cb.cb_create = zfs_prop_get_int(snap, ZFS_PROP_CREATETXG);
if (cb.cb_create > 0)
min_txg = cb.cb_create;
(void) zfs_iter_snapshots(zhp, B_FALSE, rollback_destroy, &cb,
min_txg, 0);
(void) zfs_iter_bookmarks(zhp, rollback_destroy, &cb);
if (cb.cb_error)
return (-1);
/*
* Now that we have verified that the snapshot is the latest,
* rollback to the given snapshot.
*/
if (zhp->zfs_type == ZFS_TYPE_VOLUME) {
if (zfs_which_resv_prop(zhp, &resv_prop) < 0)
return (-1);
old_volsize = zfs_prop_get_int(zhp, ZFS_PROP_VOLSIZE);
restore_resv =
(old_volsize == zfs_prop_get_int(zhp, resv_prop));
}
/*
* Pass both the filesystem and the wanted snapshot names,
* we would get an error back if the snapshot is destroyed or
* a new snapshot is created before this request is processed.
*/
err = lzc_rollback_to(zhp->zfs_name, snap->zfs_name);
if (err != 0) {
char errbuf[1024];
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot rollback '%s'"),
zhp->zfs_name);
switch (err) {
case EEXIST:
zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN,
"there is a snapshot or bookmark more recent "
"than '%s'"), snap->zfs_name);
(void) zfs_error(zhp->zfs_hdl, EZFS_EXISTS, errbuf);
break;
case ESRCH:
zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN,
"'%s' is not found among snapshots of '%s'"),
snap->zfs_name, zhp->zfs_name);
(void) zfs_error(zhp->zfs_hdl, EZFS_NOENT, errbuf);
break;
case EINVAL:
(void) zfs_error(zhp->zfs_hdl, EZFS_BADTYPE, errbuf);
break;
default:
(void) zfs_standard_error(zhp->zfs_hdl, err, errbuf);
}
return (err);
}
/*
* For volumes, if the pre-rollback volsize matched the pre-
* rollback reservation and the volsize has changed then set
* the reservation property to the post-rollback volsize.
* Make a new handle since the rollback closed the dataset.
*/
if ((zhp->zfs_type == ZFS_TYPE_VOLUME) &&
(zhp = make_dataset_handle(zhp->zfs_hdl, zhp->zfs_name))) {
if (restore_resv) {
new_volsize = zfs_prop_get_int(zhp, ZFS_PROP_VOLSIZE);
if (old_volsize != new_volsize)
err = zfs_prop_set_int(zhp, resv_prop,
new_volsize);
}
zfs_close(zhp);
}
return (err);
}
/*
* Renames the given dataset.
*/
int
zfs_rename(zfs_handle_t *zhp, const char *source, const char *target,
renameflags_t flags)
{
int ret = 0;
zfs_cmd_t zc = { 0 };
char *delim;
prop_changelist_t *cl = NULL;
zfs_handle_t *zhrp = NULL;
char *parentname = NULL;
char parent[ZFS_MAX_DATASET_NAME_LEN];
char property[ZFS_MAXPROPLEN];
libzfs_handle_t *hdl = zhp->zfs_hdl;
char errbuf[1024];
/* if we have the same exact name, just return success */
if (strcmp(zhp->zfs_name, target) == 0)
return (0);
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot rename to '%s'"), target);
if (source != NULL) {
/*
* This is recursive snapshots rename, put snapshot name
* (that might not exist) into zfs_name.
*/
assert(flags.recurse);
(void) strlcat(zhp->zfs_name, "@", sizeof(zhp->zfs_name));
(void) strlcat(zhp->zfs_name, source, sizeof(zhp->zfs_name));
zhp->zfs_type = ZFS_TYPE_SNAPSHOT;
}
/* make sure source name is valid */
if (!zfs_validate_name(hdl, zhp->zfs_name, zhp->zfs_type, B_TRUE))
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
/*
* Make sure the target name is valid
*/
if (zhp->zfs_type == ZFS_TYPE_SNAPSHOT ||
zhp->zfs_type == ZFS_TYPE_BOOKMARK) {
const char sep = zhp->zfs_type == ZFS_TYPE_SNAPSHOT ? '@' : '#';
if ((strchr(target, sep) == NULL) || *target == sep) {
/*
* Snapshot target name is abbreviated,
* reconstruct full dataset name
*/
(void) strlcpy(parent, zhp->zfs_name, sizeof (parent));
delim = strchr(parent, sep);
if (strchr(target, sep) == NULL)
*(++delim) = '\0';
else
*delim = '\0';
(void) strlcat(parent, target, sizeof (parent));
target = parent;
} else {
/*
* Make sure we're renaming within the same dataset.
*/
delim = strchr(target, sep);
if (strncmp(zhp->zfs_name, target, delim - target)
!= 0 || zhp->zfs_name[delim - target] != sep) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"%s must be part of same dataset"),
zhp->zfs_type == ZFS_TYPE_SNAPSHOT ?
"snapshots" : "bookmarks");
return (zfs_error(hdl, EZFS_CROSSTARGET,
errbuf));
}
}
if (!zfs_validate_name(hdl, target, zhp->zfs_type, B_TRUE))
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
} else {
if (flags.recurse) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"recursive rename must be a snapshot"));
return (zfs_error(hdl, EZFS_BADTYPE, errbuf));
}
if (!zfs_validate_name(hdl, target, zhp->zfs_type, B_TRUE))
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
/* validate parents */
if (check_parents(hdl, target, NULL, B_FALSE, NULL) != 0)
return (-1);
/* make sure we're in the same pool */
verify((delim = strchr(target, '/')) != NULL);
if (strncmp(zhp->zfs_name, target, delim - target) != 0 ||
zhp->zfs_name[delim - target] != '/') {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"datasets must be within same pool"));
return (zfs_error(hdl, EZFS_CROSSTARGET, errbuf));
}
/* new name cannot be a child of the current dataset name */
if (is_descendant(zhp->zfs_name, target)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"New dataset name cannot be a descendant of "
"current dataset name"));
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
}
}
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot rename '%s'"), zhp->zfs_name);
if (getzoneid() == GLOBAL_ZONEID &&
zfs_prop_get_int(zhp, ZFS_PROP_ZONED)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"dataset is used in a non-global zone"));
return (zfs_error(hdl, EZFS_ZONED, errbuf));
}
/*
* Avoid unmounting file systems with mountpoint property set to
* 'legacy' or 'none' even if -u option is not given.
*/
if (zhp->zfs_type == ZFS_TYPE_FILESYSTEM &&
!flags.recurse && !flags.nounmount &&
zfs_prop_get(zhp, ZFS_PROP_MOUNTPOINT, property,
sizeof (property), NULL, NULL, 0, B_FALSE) == 0 &&
(strcmp(property, "legacy") == 0 ||
strcmp(property, "none") == 0)) {
flags.nounmount = B_TRUE;
}
if (flags.recurse) {
parentname = zfs_strdup(zhp->zfs_hdl, zhp->zfs_name);
if (parentname == NULL) {
ret = -1;
goto error;
}
delim = strchr(parentname, '@');
*delim = '\0';
zhrp = zfs_open(zhp->zfs_hdl, parentname, ZFS_TYPE_DATASET);
if (zhrp == NULL) {
ret = -1;
goto error;
}
} else if (zhp->zfs_type != ZFS_TYPE_SNAPSHOT &&
zhp->zfs_type != ZFS_TYPE_BOOKMARK) {
if ((cl = changelist_gather(zhp, ZFS_PROP_NAME,
flags.nounmount ? CL_GATHER_DONT_UNMOUNT : 0,
flags.forceunmount ? MS_FORCE : 0)) == NULL) {
return (-1);
}
if (changelist_haszonedchild(cl)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"child dataset with inherited mountpoint is used "
"in a non-global zone"));
(void) zfs_error(hdl, EZFS_ZONED, errbuf);
ret = -1;
goto error;
}
if ((ret = changelist_prefix(cl)) != 0)
goto error;
}
if (ZFS_IS_VOLUME(zhp))
zc.zc_objset_type = DMU_OST_ZVOL;
else
zc.zc_objset_type = DMU_OST_ZFS;
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
(void) strlcpy(zc.zc_value, target, sizeof (zc.zc_value));
zc.zc_cookie = flags.recurse ? 1 : 0;
if (flags.nounmount)
zc.zc_cookie |= 2;
if ((ret = zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_RENAME, &zc)) != 0) {
/*
* if it was recursive, the one that actually failed will
* be in zc.zc_name
*/
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot rename '%s'"), zc.zc_name);
if (flags.recurse && errno == EEXIST) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"a child dataset already has a snapshot "
"with the new name"));
(void) zfs_error(hdl, EZFS_EXISTS, errbuf);
} else if (errno == EINVAL) {
(void) zfs_error(hdl, EZFS_INVALIDNAME, errbuf);
} else {
(void) zfs_standard_error(zhp->zfs_hdl, errno, errbuf);
}
/*
* On failure, we still want to remount any filesystems that
* were previously mounted, so we don't alter the system state.
*/
if (cl != NULL)
(void) changelist_postfix(cl);
} else {
if (cl != NULL) {
changelist_rename(cl, zfs_get_name(zhp), target);
ret = changelist_postfix(cl);
}
}
error:
if (parentname != NULL) {
free(parentname);
}
if (zhrp != NULL) {
zfs_close(zhrp);
}
if (cl != NULL) {
changelist_free(cl);
}
return (ret);
}
nvlist_t *
zfs_get_user_props(zfs_handle_t *zhp)
{
return (zhp->zfs_user_props);
}
nvlist_t *
zfs_get_recvd_props(zfs_handle_t *zhp)
{
if (zhp->zfs_recvd_props == NULL)
if (get_recvd_props_ioctl(zhp) != 0)
return (NULL);
return (zhp->zfs_recvd_props);
}
/*
* This function is used by 'zfs list' to determine the exact set of columns to
* display, and their maximum widths. This does two main things:
*
* - If this is a list of all properties, then expand the list to include
* all native properties, and set a flag so that for each dataset we look
* for new unique user properties and add them to the list.
*
* - For non fixed-width properties, keep track of the maximum width seen
* so that we can size the column appropriately. If the user has
* requested received property values, we also need to compute the width
* of the RECEIVED column.
*/
int
zfs_expand_proplist(zfs_handle_t *zhp, zprop_list_t **plp, boolean_t received,
boolean_t literal)
{
libzfs_handle_t *hdl = zhp->zfs_hdl;
zprop_list_t *entry;
zprop_list_t **last, **start;
nvlist_t *userprops, *propval;
nvpair_t *elem;
char *strval;
char buf[ZFS_MAXPROPLEN];
if (zprop_expand_list(hdl, plp, ZFS_TYPE_DATASET) != 0)
return (-1);
userprops = zfs_get_user_props(zhp);
entry = *plp;
if (entry->pl_all && nvlist_next_nvpair(userprops, NULL) != NULL) {
/*
* Go through and add any user properties as necessary. We
* start by incrementing our list pointer to the first
* non-native property.
*/
start = plp;
while (*start != NULL) {
if ((*start)->pl_prop == ZPROP_INVAL)
break;
start = &(*start)->pl_next;
}
elem = NULL;
while ((elem = nvlist_next_nvpair(userprops, elem)) != NULL) {
/*
* See if we've already found this property in our list.
*/
for (last = start; *last != NULL;
last = &(*last)->pl_next) {
if (strcmp((*last)->pl_user_prop,
nvpair_name(elem)) == 0)
break;
}
if (*last == NULL) {
if ((entry = zfs_alloc(hdl,
sizeof (zprop_list_t))) == NULL ||
((entry->pl_user_prop = zfs_strdup(hdl,
nvpair_name(elem)))) == NULL) {
free(entry);
return (-1);
}
entry->pl_prop = ZPROP_INVAL;
entry->pl_width = strlen(nvpair_name(elem));
entry->pl_all = B_TRUE;
*last = entry;
}
}
}
/*
* Now go through and check the width of any non-fixed columns
*/
for (entry = *plp; entry != NULL; entry = entry->pl_next) {
if (entry->pl_fixed && !literal)
continue;
if (entry->pl_prop != ZPROP_INVAL) {
if (zfs_prop_get(zhp, entry->pl_prop,
buf, sizeof (buf), NULL, NULL, 0, literal) == 0) {
if (strlen(buf) > entry->pl_width)
entry->pl_width = strlen(buf);
}
if (received && zfs_prop_get_recvd(zhp,
zfs_prop_to_name(entry->pl_prop),
buf, sizeof (buf), literal) == 0)
if (strlen(buf) > entry->pl_recvd_width)
entry->pl_recvd_width = strlen(buf);
} else {
if (nvlist_lookup_nvlist(userprops, entry->pl_user_prop,
&propval) == 0) {
verify(nvlist_lookup_string(propval,
ZPROP_VALUE, &strval) == 0);
if (strlen(strval) > entry->pl_width)
entry->pl_width = strlen(strval);
}
if (received && zfs_prop_get_recvd(zhp,
entry->pl_user_prop,
buf, sizeof (buf), literal) == 0)
if (strlen(buf) > entry->pl_recvd_width)
entry->pl_recvd_width = strlen(buf);
}
}
return (0);
}
int
zfs_deleg_share_nfs(libzfs_handle_t *hdl, char *dataset, char *path,
char *resource, void *export, void *sharetab,
int sharemax, zfs_share_op_t operation)
{
zfs_cmd_t zc = { 0 };
int error;
(void) strlcpy(zc.zc_name, dataset, sizeof (zc.zc_name));
(void) strlcpy(zc.zc_value, path, sizeof (zc.zc_value));
if (resource)
(void) strlcpy(zc.zc_string, resource, sizeof (zc.zc_string));
zc.zc_share.z_sharedata = (uint64_t)(uintptr_t)sharetab;
zc.zc_share.z_exportdata = (uint64_t)(uintptr_t)export;
zc.zc_share.z_sharetype = operation;
zc.zc_share.z_sharemax = sharemax;
error = ioctl(hdl->libzfs_fd, ZFS_IOC_SHARE, &zc);
return (error);
}
void
zfs_prune_proplist(zfs_handle_t *zhp, uint8_t *props)
{
nvpair_t *curr;
/*
* Keep a reference to the props-table against which we prune the
* properties.
*/
zhp->zfs_props_table = props;
curr = nvlist_next_nvpair(zhp->zfs_props, NULL);
while (curr) {
zfs_prop_t zfs_prop = zfs_name_to_prop(nvpair_name(curr));
nvpair_t *next = nvlist_next_nvpair(zhp->zfs_props, curr);
/*
* User properties will result in ZPROP_INVAL, and since we
* only know how to prune standard ZFS properties, we always
* leave these in the list. This can also happen if we
* encounter an unknown DSL property (when running older
* software, for example).
*/
if (zfs_prop != ZPROP_INVAL && props[zfs_prop] == B_FALSE)
(void) nvlist_remove(zhp->zfs_props,
nvpair_name(curr), nvpair_type(curr));
curr = next;
}
}
#ifdef illumos
static int
zfs_smb_acl_mgmt(libzfs_handle_t *hdl, char *dataset, char *path,
zfs_smb_acl_op_t cmd, char *resource1, char *resource2)
{
zfs_cmd_t zc = { 0 };
nvlist_t *nvlist = NULL;
int error;
(void) strlcpy(zc.zc_name, dataset, sizeof (zc.zc_name));
(void) strlcpy(zc.zc_value, path, sizeof (zc.zc_value));
zc.zc_cookie = (uint64_t)cmd;
if (cmd == ZFS_SMB_ACL_RENAME) {
if (nvlist_alloc(&nvlist, NV_UNIQUE_NAME, 0) != 0) {
(void) no_memory(hdl);
return (0);
}
}
switch (cmd) {
case ZFS_SMB_ACL_ADD:
case ZFS_SMB_ACL_REMOVE:
(void) strlcpy(zc.zc_string, resource1, sizeof (zc.zc_string));
break;
case ZFS_SMB_ACL_RENAME:
if (nvlist_add_string(nvlist, ZFS_SMB_ACL_SRC,
resource1) != 0) {
(void) no_memory(hdl);
return (-1);
}
if (nvlist_add_string(nvlist, ZFS_SMB_ACL_TARGET,
resource2) != 0) {
(void) no_memory(hdl);
return (-1);
}
if (zcmd_write_src_nvlist(hdl, &zc, nvlist) != 0) {
nvlist_free(nvlist);
return (-1);
}
break;
case ZFS_SMB_ACL_PURGE:
break;
default:
return (-1);
}
error = ioctl(hdl->libzfs_fd, ZFS_IOC_SMB_ACL, &zc);
nvlist_free(nvlist);
return (error);
}
int
zfs_smb_acl_add(libzfs_handle_t *hdl, char *dataset,
char *path, char *resource)
{
return (zfs_smb_acl_mgmt(hdl, dataset, path, ZFS_SMB_ACL_ADD,
resource, NULL));
}
int
zfs_smb_acl_remove(libzfs_handle_t *hdl, char *dataset,
char *path, char *resource)
{
return (zfs_smb_acl_mgmt(hdl, dataset, path, ZFS_SMB_ACL_REMOVE,
resource, NULL));
}
int
zfs_smb_acl_purge(libzfs_handle_t *hdl, char *dataset, char *path)
{
return (zfs_smb_acl_mgmt(hdl, dataset, path, ZFS_SMB_ACL_PURGE,
NULL, NULL));
}
int
zfs_smb_acl_rename(libzfs_handle_t *hdl, char *dataset, char *path,
char *oldname, char *newname)
{
return (zfs_smb_acl_mgmt(hdl, dataset, path, ZFS_SMB_ACL_RENAME,
oldname, newname));
}
#endif /* illumos */
int
zfs_userspace(zfs_handle_t *zhp, zfs_userquota_prop_t type,
zfs_userspace_cb_t func, void *arg)
{
zfs_cmd_t zc = { 0 };
zfs_useracct_t buf[100];
libzfs_handle_t *hdl = zhp->zfs_hdl;
int ret;
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
zc.zc_objset_type = type;
zc.zc_nvlist_dst = (uintptr_t)buf;
for (;;) {
zfs_useracct_t *zua = buf;
zc.zc_nvlist_dst_size = sizeof (buf);
if (zfs_ioctl(hdl, ZFS_IOC_USERSPACE_MANY, &zc) != 0) {
char errbuf[1024];
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN,
"cannot get used/quota for %s"), zc.zc_name);
return (zfs_standard_error_fmt(hdl, errno, errbuf));
}
if (zc.zc_nvlist_dst_size == 0)
break;
while (zc.zc_nvlist_dst_size > 0) {
if ((ret = func(arg, zua->zu_domain, zua->zu_rid,
zua->zu_space)) != 0)
return (ret);
zua++;
zc.zc_nvlist_dst_size -= sizeof (zfs_useracct_t);
}
}
return (0);
}
struct holdarg {
nvlist_t *nvl;
const char *snapname;
const char *tag;
boolean_t recursive;
int error;
};
static int
zfs_hold_one(zfs_handle_t *zhp, void *arg)
{
struct holdarg *ha = arg;
char name[ZFS_MAX_DATASET_NAME_LEN];
int rv = 0;
(void) snprintf(name, sizeof (name),
"%s@%s", zhp->zfs_name, ha->snapname);
if (lzc_exists(name))
fnvlist_add_string(ha->nvl, name, ha->tag);
if (ha->recursive)
rv = zfs_iter_filesystems(zhp, zfs_hold_one, ha);
zfs_close(zhp);
return (rv);
}
int
zfs_hold(zfs_handle_t *zhp, const char *snapname, const char *tag,
boolean_t recursive, int cleanup_fd)
{
int ret;
struct holdarg ha;
ha.nvl = fnvlist_alloc();
ha.snapname = snapname;
ha.tag = tag;
ha.recursive = recursive;
(void) zfs_hold_one(zfs_handle_dup(zhp), &ha);
if (nvlist_empty(ha.nvl)) {
char errbuf[1024];
fnvlist_free(ha.nvl);
ret = ENOENT;
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN,
"cannot hold snapshot '%s@%s'"),
zhp->zfs_name, snapname);
(void) zfs_standard_error(zhp->zfs_hdl, ret, errbuf);
return (ret);
}
ret = zfs_hold_nvl(zhp, cleanup_fd, ha.nvl);
fnvlist_free(ha.nvl);
return (ret);
}
int
zfs_hold_nvl(zfs_handle_t *zhp, int cleanup_fd, nvlist_t *holds)
{
int ret;
nvlist_t *errors;
libzfs_handle_t *hdl = zhp->zfs_hdl;
char errbuf[1024];
nvpair_t *elem;
errors = NULL;
ret = lzc_hold(holds, cleanup_fd, &errors);
if (ret == 0) {
/* There may be errors even in the success case. */
fnvlist_free(errors);
return (0);
}
if (nvlist_empty(errors)) {
/* no hold-specific errors */
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot hold"));
switch (ret) {
case ENOTSUP:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool must be upgraded"));
(void) zfs_error(hdl, EZFS_BADVERSION, errbuf);
break;
case EINVAL:
(void) zfs_error(hdl, EZFS_BADTYPE, errbuf);
break;
default:
(void) zfs_standard_error(hdl, ret, errbuf);
}
}
for (elem = nvlist_next_nvpair(errors, NULL);
elem != NULL;
elem = nvlist_next_nvpair(errors, elem)) {
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN,
"cannot hold snapshot '%s'"), nvpair_name(elem));
switch (fnvpair_value_int32(elem)) {
case E2BIG:
/*
* Temporary tags wind up having the ds object id
* prepended. So even if we passed the length check
* above, it's still possible for the tag to wind
* up being slightly too long.
*/
(void) zfs_error(hdl, EZFS_TAGTOOLONG, errbuf);
break;
case EINVAL:
(void) zfs_error(hdl, EZFS_BADTYPE, errbuf);
break;
case EEXIST:
(void) zfs_error(hdl, EZFS_REFTAG_HOLD, errbuf);
break;
default:
(void) zfs_standard_error(hdl,
fnvpair_value_int32(elem), errbuf);
}
}
fnvlist_free(errors);
return (ret);
}
static int
zfs_release_one(zfs_handle_t *zhp, void *arg)
{
struct holdarg *ha = arg;
char name[ZFS_MAX_DATASET_NAME_LEN];
int rv = 0;
nvlist_t *existing_holds;
(void) snprintf(name, sizeof (name),
"%s@%s", zhp->zfs_name, ha->snapname);
if (lzc_get_holds(name, &existing_holds) != 0) {
ha->error = ENOENT;
} else if (!nvlist_exists(existing_holds, ha->tag)) {
ha->error = ESRCH;
} else {
nvlist_t *torelease = fnvlist_alloc();
fnvlist_add_boolean(torelease, ha->tag);
fnvlist_add_nvlist(ha->nvl, name, torelease);
fnvlist_free(torelease);
}
if (ha->recursive)
rv = zfs_iter_filesystems(zhp, zfs_release_one, ha);
zfs_close(zhp);
return (rv);
}
int
zfs_release(zfs_handle_t *zhp, const char *snapname, const char *tag,
boolean_t recursive)
{
int ret;
struct holdarg ha;
nvlist_t *errors = NULL;
nvpair_t *elem;
libzfs_handle_t *hdl = zhp->zfs_hdl;
char errbuf[1024];
ha.nvl = fnvlist_alloc();
ha.snapname = snapname;
ha.tag = tag;
ha.recursive = recursive;
ha.error = 0;
(void) zfs_release_one(zfs_handle_dup(zhp), &ha);
if (nvlist_empty(ha.nvl)) {
fnvlist_free(ha.nvl);
ret = ha.error;
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN,
"cannot release hold from snapshot '%s@%s'"),
zhp->zfs_name, snapname);
if (ret == ESRCH) {
(void) zfs_error(hdl, EZFS_REFTAG_RELE, errbuf);
} else {
(void) zfs_standard_error(hdl, ret, errbuf);
}
return (ret);
}
ret = lzc_release(ha.nvl, &errors);
fnvlist_free(ha.nvl);
if (ret == 0) {
/* There may be errors even in the success case. */
fnvlist_free(errors);
return (0);
}
if (nvlist_empty(errors)) {
/* no hold-specific errors */
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot release"));
switch (errno) {
case ENOTSUP:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool must be upgraded"));
(void) zfs_error(hdl, EZFS_BADVERSION, errbuf);
break;
default:
(void) zfs_standard_error_fmt(hdl, errno, errbuf);
}
}
for (elem = nvlist_next_nvpair(errors, NULL);
elem != NULL;
elem = nvlist_next_nvpair(errors, elem)) {
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN,
"cannot release hold from snapshot '%s'"),
nvpair_name(elem));
switch (fnvpair_value_int32(elem)) {
case ESRCH:
(void) zfs_error(hdl, EZFS_REFTAG_RELE, errbuf);
break;
case EINVAL:
(void) zfs_error(hdl, EZFS_BADTYPE, errbuf);
break;
default:
(void) zfs_standard_error_fmt(hdl,
fnvpair_value_int32(elem), errbuf);
}
}
fnvlist_free(errors);
return (ret);
}
int
zfs_get_fsacl(zfs_handle_t *zhp, nvlist_t **nvl)
{
zfs_cmd_t zc = { 0 };
libzfs_handle_t *hdl = zhp->zfs_hdl;
int nvsz = 2048;
void *nvbuf;
int err = 0;
char errbuf[1024];
assert(zhp->zfs_type == ZFS_TYPE_VOLUME ||
zhp->zfs_type == ZFS_TYPE_FILESYSTEM);
tryagain:
nvbuf = malloc(nvsz);
if (nvbuf == NULL) {
err = (zfs_error(hdl, EZFS_NOMEM, strerror(errno)));
goto out;
}
zc.zc_nvlist_dst_size = nvsz;
zc.zc_nvlist_dst = (uintptr_t)nvbuf;
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
if (ioctl(hdl->libzfs_fd, ZFS_IOC_GET_FSACL, &zc) != 0) {
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot get permissions on '%s'"),
zc.zc_name);
switch (errno) {
case ENOMEM:
free(nvbuf);
nvsz = zc.zc_nvlist_dst_size;
goto tryagain;
case ENOTSUP:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool must be upgraded"));
err = zfs_error(hdl, EZFS_BADVERSION, errbuf);
break;
case EINVAL:
err = zfs_error(hdl, EZFS_BADTYPE, errbuf);
break;
case ENOENT:
err = zfs_error(hdl, EZFS_NOENT, errbuf);
break;
default:
err = zfs_standard_error_fmt(hdl, errno, errbuf);
break;
}
} else {
/* success */
int rc = nvlist_unpack(nvbuf, zc.zc_nvlist_dst_size, nvl, 0);
if (rc) {
(void) snprintf(errbuf, sizeof (errbuf), dgettext(
TEXT_DOMAIN, "cannot get permissions on '%s'"),
zc.zc_name);
err = zfs_standard_error_fmt(hdl, rc, errbuf);
}
}
free(nvbuf);
out:
return (err);
}
int
zfs_set_fsacl(zfs_handle_t *zhp, boolean_t un, nvlist_t *nvl)
{
zfs_cmd_t zc = { 0 };
libzfs_handle_t *hdl = zhp->zfs_hdl;
char *nvbuf;
char errbuf[1024];
size_t nvsz;
int err;
assert(zhp->zfs_type == ZFS_TYPE_VOLUME ||
zhp->zfs_type == ZFS_TYPE_FILESYSTEM);
err = nvlist_size(nvl, &nvsz, NV_ENCODE_NATIVE);
assert(err == 0);
nvbuf = malloc(nvsz);
err = nvlist_pack(nvl, &nvbuf, &nvsz, NV_ENCODE_NATIVE, 0);
assert(err == 0);
zc.zc_nvlist_src_size = nvsz;
zc.zc_nvlist_src = (uintptr_t)nvbuf;
zc.zc_perm_action = un;
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
if (zfs_ioctl(hdl, ZFS_IOC_SET_FSACL, &zc) != 0) {
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot set permissions on '%s'"),
zc.zc_name);
switch (errno) {
case ENOTSUP:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool must be upgraded"));
err = zfs_error(hdl, EZFS_BADVERSION, errbuf);
break;
case EINVAL:
err = zfs_error(hdl, EZFS_BADTYPE, errbuf);
break;
case ENOENT:
err = zfs_error(hdl, EZFS_NOENT, errbuf);
break;
default:
err = zfs_standard_error_fmt(hdl, errno, errbuf);
break;
}
}
free(nvbuf);
return (err);
}
int
zfs_get_holds(zfs_handle_t *zhp, nvlist_t **nvl)
{
int err;
char errbuf[1024];
err = lzc_get_holds(zhp->zfs_name, nvl);
if (err != 0) {
libzfs_handle_t *hdl = zhp->zfs_hdl;
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot get holds for '%s'"),
zhp->zfs_name);
switch (err) {
case ENOTSUP:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool must be upgraded"));
err = zfs_error(hdl, EZFS_BADVERSION, errbuf);
break;
case EINVAL:
err = zfs_error(hdl, EZFS_BADTYPE, errbuf);
break;
case ENOENT:
err = zfs_error(hdl, EZFS_NOENT, errbuf);
break;
default:
err = zfs_standard_error_fmt(hdl, errno, errbuf);
break;
}
}
return (err);
}
/*
* Convert the zvol's volume size to an appropriate reservation.
* Note: If this routine is updated, it is necessary to update the ZFS test
* suite's shell version in reservation.kshlib.
*/
uint64_t
zvol_volsize_to_reservation(uint64_t volsize, nvlist_t *props)
{
uint64_t numdb;
uint64_t nblocks, volblocksize;
int ncopies;
char *strval;
if (nvlist_lookup_string(props,
zfs_prop_to_name(ZFS_PROP_COPIES), &strval) == 0)
ncopies = atoi(strval);
else
ncopies = 1;
if (nvlist_lookup_uint64(props,
zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE),
&volblocksize) != 0)
volblocksize = ZVOL_DEFAULT_BLOCKSIZE;
nblocks = volsize/volblocksize;
/* start with metadnode L0-L6 */
numdb = 7;
/* calculate number of indirects */
while (nblocks > 1) {
nblocks += DNODES_PER_LEVEL - 1;
nblocks /= DNODES_PER_LEVEL;
numdb += nblocks;
}
numdb *= MIN(SPA_DVAS_PER_BP, ncopies + 1);
volsize *= ncopies;
/*
* this is exactly DN_MAX_INDBLKSHIFT when metadata isn't
* compressed, but in practice they compress down to about
* 1100 bytes
*/
numdb *= 1ULL << DN_MAX_INDBLKSHIFT;
volsize += numdb;
return (volsize);
}
/*
* Attach/detach the given filesystem to/from the given jail.
*/
int
zfs_jail(zfs_handle_t *zhp, int jailid, int attach)
{
libzfs_handle_t *hdl = zhp->zfs_hdl;
zfs_cmd_t zc = { 0 };
char errbuf[1024];
unsigned long cmd;
int ret;
if (attach) {
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot jail '%s'"), zhp->zfs_name);
} else {
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN, "cannot unjail '%s'"), zhp->zfs_name);
}
switch (zhp->zfs_type) {
case ZFS_TYPE_VOLUME:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"volumes can not be jailed"));
return (zfs_error(hdl, EZFS_BADTYPE, errbuf));
case ZFS_TYPE_SNAPSHOT:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"snapshots can not be jailed"));
return (zfs_error(hdl, EZFS_BADTYPE, errbuf));
}
assert(zhp->zfs_type == ZFS_TYPE_FILESYSTEM);
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
zc.zc_objset_type = DMU_OST_ZFS;
zc.zc_jailid = jailid;
cmd = attach ? ZFS_IOC_JAIL : ZFS_IOC_UNJAIL;
if ((ret = ioctl(hdl->libzfs_fd, cmd, &zc)) != 0)
zfs_standard_error(hdl, errno, errbuf);
return (ret);
}
Index: head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c
===================================================================
--- head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c (revision 359302)
+++ head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_sendrecv.c (revision 359303)
@@ -1,3882 +1,3920 @@
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2011, 2015 by Delphix. All rights reserved.
* Copyright (c) 2012, Joyent, Inc. All rights reserved.
* Copyright (c) 2012 Pawel Jakub Dawidek. All rights reserved.
* Copyright (c) 2013 Steven Hartland. All rights reserved.
* Copyright 2015, OmniTI Computer Consulting, Inc. All rights reserved.
* Copyright (c) 2014 Integros [integros.com]
* Copyright 2016 Igor Kozhukhov <ikozhukhov@gmail.com>
+ * Copyright (c) 2018, loli10K <ezomori.nozomu@gmail.com>. All rights reserved.
* Copyright (c) 2019 Datto Inc.
*/
#include <assert.h>
#include <ctype.h>
#include <errno.h>
#include <libintl.h>
#include <stdio.h>
#include <stdlib.h>
#include <strings.h>
#include <unistd.h>
#include <stddef.h>
#include <fcntl.h>
#include <sys/param.h>
#include <sys/mount.h>
#include <pthread.h>
#include <umem.h>
#include <time.h>
#include <libzfs.h>
#include <libzfs_core.h>
#include "zfs_namecheck.h"
#include "zfs_prop.h"
#include "zfs_fletcher.h"
#include "libzfs_impl.h"
#include <zlib.h>
#include <sha2.h>
#include <sys/zio_checksum.h>
#include <sys/ddt.h>
#ifdef __FreeBSD__
extern int zfs_ioctl_version;
#endif
/* in libzfs_dataset.c */
extern void zfs_setprop_error(libzfs_handle_t *, zfs_prop_t, int, char *);
/* We need to use something for ENODATA. */
#define ENODATA EIDRM
static int zfs_receive_impl(libzfs_handle_t *, const char *, const char *,
recvflags_t *, int, const char *, nvlist_t *, avl_tree_t *, char **, int,
uint64_t *, const char *);
static int guid_to_name(libzfs_handle_t *, const char *,
uint64_t, boolean_t, char *);
static const zio_cksum_t zero_cksum = { 0 };
typedef struct dedup_arg {
int inputfd;
int outputfd;
libzfs_handle_t *dedup_hdl;
} dedup_arg_t;
typedef struct progress_arg {
zfs_handle_t *pa_zhp;
int pa_fd;
boolean_t pa_parsable;
boolean_t pa_astitle;
uint64_t pa_size;
} progress_arg_t;
typedef struct dataref {
uint64_t ref_guid;
uint64_t ref_object;
uint64_t ref_offset;
} dataref_t;
typedef struct dedup_entry {
struct dedup_entry *dde_next;
zio_cksum_t dde_chksum;
uint64_t dde_prop;
dataref_t dde_ref;
} dedup_entry_t;
#define MAX_DDT_PHYSMEM_PERCENT 20
#define SMALLEST_POSSIBLE_MAX_DDT_MB 128
typedef struct dedup_table {
dedup_entry_t **dedup_hash_array;
umem_cache_t *ddecache;
uint64_t max_ddt_size; /* max dedup table size in bytes */
uint64_t cur_ddt_size; /* current dedup table size in bytes */
uint64_t ddt_count;
int numhashbits;
boolean_t ddt_full;
} dedup_table_t;
static int
high_order_bit(uint64_t n)
{
int count;
for (count = 0; n != 0; count++)
n >>= 1;
return (count);
}
static size_t
ssread(void *buf, size_t len, FILE *stream)
{
size_t outlen;
if ((outlen = fread(buf, len, 1, stream)) == 0)
return (0);
return (outlen);
}
static void
ddt_hash_append(libzfs_handle_t *hdl, dedup_table_t *ddt, dedup_entry_t **ddepp,
zio_cksum_t *cs, uint64_t prop, dataref_t *dr)
{
dedup_entry_t *dde;
if (ddt->cur_ddt_size >= ddt->max_ddt_size) {
if (ddt->ddt_full == B_FALSE) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"Dedup table full. Deduplication will continue "
"with existing table entries"));
ddt->ddt_full = B_TRUE;
}
return;
}
if ((dde = umem_cache_alloc(ddt->ddecache, UMEM_DEFAULT))
!= NULL) {
assert(*ddepp == NULL);
dde->dde_next = NULL;
dde->dde_chksum = *cs;
dde->dde_prop = prop;
dde->dde_ref = *dr;
*ddepp = dde;
ddt->cur_ddt_size += sizeof (dedup_entry_t);
ddt->ddt_count++;
}
}
/*
* Using the specified dedup table, do a lookup for an entry with
* the checksum cs. If found, return the block's reference info
* in *dr. Otherwise, insert a new entry in the dedup table, using
* the reference information specified by *dr.
*
* return value: true - entry was found
* false - entry was not found
*/
static boolean_t
ddt_update(libzfs_handle_t *hdl, dedup_table_t *ddt, zio_cksum_t *cs,
uint64_t prop, dataref_t *dr)
{
uint32_t hashcode;
dedup_entry_t **ddepp;
hashcode = BF64_GET(cs->zc_word[0], 0, ddt->numhashbits);
for (ddepp = &(ddt->dedup_hash_array[hashcode]); *ddepp != NULL;
ddepp = &((*ddepp)->dde_next)) {
if (ZIO_CHECKSUM_EQUAL(((*ddepp)->dde_chksum), *cs) &&
(*ddepp)->dde_prop == prop) {
*dr = (*ddepp)->dde_ref;
return (B_TRUE);
}
}
ddt_hash_append(hdl, ddt, ddepp, cs, prop, dr);
return (B_FALSE);
}
static int
dump_record(dmu_replay_record_t *drr, void *payload, int payload_len,
zio_cksum_t *zc, int outfd)
{
ASSERT3U(offsetof(dmu_replay_record_t, drr_u.drr_checksum.drr_checksum),
==, sizeof (dmu_replay_record_t) - sizeof (zio_cksum_t));
(void) fletcher_4_incremental_native(drr,
offsetof(dmu_replay_record_t, drr_u.drr_checksum.drr_checksum), zc);
if (drr->drr_type != DRR_BEGIN) {
ASSERT(ZIO_CHECKSUM_IS_ZERO(&drr->drr_u.
drr_checksum.drr_checksum));
drr->drr_u.drr_checksum.drr_checksum = *zc;
}
(void) fletcher_4_incremental_native(
&drr->drr_u.drr_checksum.drr_checksum, sizeof (zio_cksum_t), zc);
if (write(outfd, drr, sizeof (*drr)) == -1)
return (errno);
if (payload_len != 0) {
(void) fletcher_4_incremental_native(payload, payload_len, zc);
if (write(outfd, payload, payload_len) == -1)
return (errno);
}
return (0);
}
/*
* This function is started in a separate thread when the dedup option
* has been requested. The main send thread determines the list of
* snapshots to be included in the send stream and makes the ioctl calls
* for each one. But instead of having the ioctl send the output to the
* the output fd specified by the caller of zfs_send()), the
* ioctl is told to direct the output to a pipe, which is read by the
* alternate thread running THIS function. This function does the
* dedup'ing by:
* 1. building a dedup table (the DDT)
* 2. doing checksums on each data block and inserting a record in the DDT
* 3. looking for matching checksums, and
* 4. sending a DRR_WRITE_BYREF record instead of a write record whenever
* a duplicate block is found.
* The output of this function then goes to the output fd requested
* by the caller of zfs_send().
*/
static void *
cksummer(void *arg)
{
dedup_arg_t *dda = arg;
char *buf = zfs_alloc(dda->dedup_hdl, SPA_MAXBLOCKSIZE);
dmu_replay_record_t thedrr;
dmu_replay_record_t *drr = &thedrr;
FILE *ofp;
int outfd;
dedup_table_t ddt;
zio_cksum_t stream_cksum;
uint64_t physmem = sysconf(_SC_PHYS_PAGES) * sysconf(_SC_PAGESIZE);
uint64_t numbuckets;
ddt.max_ddt_size =
MAX((physmem * MAX_DDT_PHYSMEM_PERCENT) / 100,
SMALLEST_POSSIBLE_MAX_DDT_MB << 20);
numbuckets = ddt.max_ddt_size / (sizeof (dedup_entry_t));
/*
* numbuckets must be a power of 2. Increase number to
* a power of 2 if necessary.
*/
if (!ISP2(numbuckets))
numbuckets = 1 << high_order_bit(numbuckets);
ddt.dedup_hash_array = calloc(numbuckets, sizeof (dedup_entry_t *));
ddt.ddecache = umem_cache_create("dde", sizeof (dedup_entry_t), 0,
NULL, NULL, NULL, NULL, NULL, 0);
ddt.cur_ddt_size = numbuckets * sizeof (dedup_entry_t *);
ddt.numhashbits = high_order_bit(numbuckets) - 1;
ddt.ddt_full = B_FALSE;
outfd = dda->outputfd;
ofp = fdopen(dda->inputfd, "r");
while (ssread(drr, sizeof (*drr), ofp) != 0) {
/*
* kernel filled in checksum, we are going to write same
* record, but need to regenerate checksum.
*/
if (drr->drr_type != DRR_BEGIN) {
bzero(&drr->drr_u.drr_checksum.drr_checksum,
sizeof (drr->drr_u.drr_checksum.drr_checksum));
}
switch (drr->drr_type) {
case DRR_BEGIN:
{
struct drr_begin *drrb = &drr->drr_u.drr_begin;
int fflags;
int sz = 0;
ZIO_SET_CHECKSUM(&stream_cksum, 0, 0, 0, 0);
ASSERT3U(drrb->drr_magic, ==, DMU_BACKUP_MAGIC);
/* set the DEDUP feature flag for this stream */
fflags = DMU_GET_FEATUREFLAGS(drrb->drr_versioninfo);
fflags |= (DMU_BACKUP_FEATURE_DEDUP |
DMU_BACKUP_FEATURE_DEDUPPROPS);
DMU_SET_FEATUREFLAGS(drrb->drr_versioninfo, fflags);
if (drr->drr_payloadlen != 0) {
sz = drr->drr_payloadlen;
if (sz > SPA_MAXBLOCKSIZE) {
buf = zfs_realloc(dda->dedup_hdl, buf,
SPA_MAXBLOCKSIZE, sz);
}
(void) ssread(buf, sz, ofp);
if (ferror(stdin))
perror("fread");
}
if (dump_record(drr, buf, sz, &stream_cksum,
outfd) != 0)
goto out;
break;
}
case DRR_END:
{
struct drr_end *drre = &drr->drr_u.drr_end;
/* use the recalculated checksum */
drre->drr_checksum = stream_cksum;
if (dump_record(drr, NULL, 0, &stream_cksum,
outfd) != 0)
goto out;
break;
}
case DRR_OBJECT:
{
struct drr_object *drro = &drr->drr_u.drr_object;
if (drro->drr_bonuslen > 0) {
(void) ssread(buf,
P2ROUNDUP((uint64_t)drro->drr_bonuslen, 8),
ofp);
}
if (dump_record(drr, buf,
P2ROUNDUP((uint64_t)drro->drr_bonuslen, 8),
&stream_cksum, outfd) != 0)
goto out;
break;
}
case DRR_SPILL:
{
struct drr_spill *drrs = &drr->drr_u.drr_spill;
(void) ssread(buf, drrs->drr_length, ofp);
if (dump_record(drr, buf, drrs->drr_length,
&stream_cksum, outfd) != 0)
goto out;
break;
}
case DRR_FREEOBJECTS:
{
if (dump_record(drr, NULL, 0, &stream_cksum,
outfd) != 0)
goto out;
break;
}
case DRR_WRITE:
{
struct drr_write *drrw = &drr->drr_u.drr_write;
dataref_t dataref;
uint64_t payload_size;
payload_size = DRR_WRITE_PAYLOAD_SIZE(drrw);
(void) ssread(buf, payload_size, ofp);
/*
* Use the existing checksum if it's dedup-capable,
* else calculate a SHA256 checksum for it.
*/
if (ZIO_CHECKSUM_EQUAL(drrw->drr_key.ddk_cksum,
zero_cksum) ||
!DRR_IS_DEDUP_CAPABLE(drrw->drr_checksumflags)) {
SHA256_CTX ctx;
zio_cksum_t tmpsha256;
SHA256Init(&ctx);
SHA256Update(&ctx, buf, payload_size);
SHA256Final(&tmpsha256, &ctx);
drrw->drr_key.ddk_cksum.zc_word[0] =
BE_64(tmpsha256.zc_word[0]);
drrw->drr_key.ddk_cksum.zc_word[1] =
BE_64(tmpsha256.zc_word[1]);
drrw->drr_key.ddk_cksum.zc_word[2] =
BE_64(tmpsha256.zc_word[2]);
drrw->drr_key.ddk_cksum.zc_word[3] =
BE_64(tmpsha256.zc_word[3]);
drrw->drr_checksumtype = ZIO_CHECKSUM_SHA256;
drrw->drr_checksumflags = DRR_CHECKSUM_DEDUP;
}
dataref.ref_guid = drrw->drr_toguid;
dataref.ref_object = drrw->drr_object;
dataref.ref_offset = drrw->drr_offset;
if (ddt_update(dda->dedup_hdl, &ddt,
&drrw->drr_key.ddk_cksum, drrw->drr_key.ddk_prop,
&dataref)) {
dmu_replay_record_t wbr_drr = {0};
struct drr_write_byref *wbr_drrr =
&wbr_drr.drr_u.drr_write_byref;
/* block already present in stream */
wbr_drr.drr_type = DRR_WRITE_BYREF;
wbr_drrr->drr_object = drrw->drr_object;
wbr_drrr->drr_offset = drrw->drr_offset;
wbr_drrr->drr_length = drrw->drr_logical_size;
wbr_drrr->drr_toguid = drrw->drr_toguid;
wbr_drrr->drr_refguid = dataref.ref_guid;
wbr_drrr->drr_refobject =
dataref.ref_object;
wbr_drrr->drr_refoffset =
dataref.ref_offset;
wbr_drrr->drr_checksumtype =
drrw->drr_checksumtype;
wbr_drrr->drr_checksumflags =
drrw->drr_checksumtype;
wbr_drrr->drr_key.ddk_cksum =
drrw->drr_key.ddk_cksum;
wbr_drrr->drr_key.ddk_prop =
drrw->drr_key.ddk_prop;
if (dump_record(&wbr_drr, NULL, 0,
&stream_cksum, outfd) != 0)
goto out;
} else {
/* block not previously seen */
if (dump_record(drr, buf, payload_size,
&stream_cksum, outfd) != 0)
goto out;
}
break;
}
case DRR_WRITE_EMBEDDED:
{
struct drr_write_embedded *drrwe =
&drr->drr_u.drr_write_embedded;
(void) ssread(buf,
P2ROUNDUP((uint64_t)drrwe->drr_psize, 8), ofp);
if (dump_record(drr, buf,
P2ROUNDUP((uint64_t)drrwe->drr_psize, 8),
&stream_cksum, outfd) != 0)
goto out;
break;
}
case DRR_FREE:
{
if (dump_record(drr, NULL, 0, &stream_cksum,
outfd) != 0)
goto out;
break;
}
default:
(void) fprintf(stderr, "INVALID record type 0x%x\n",
drr->drr_type);
/* should never happen, so assert */
assert(B_FALSE);
}
}
out:
umem_cache_destroy(ddt.ddecache);
free(ddt.dedup_hash_array);
free(buf);
(void) fclose(ofp);
return (NULL);
}
/*
* Routines for dealing with the AVL tree of fs-nvlists
*/
typedef struct fsavl_node {
avl_node_t fn_node;
nvlist_t *fn_nvfs;
char *fn_snapname;
uint64_t fn_guid;
} fsavl_node_t;
static int
fsavl_compare(const void *arg1, const void *arg2)
{
const fsavl_node_t *fn1 = (const fsavl_node_t *)arg1;
const fsavl_node_t *fn2 = (const fsavl_node_t *)arg2;
return (AVL_CMP(fn1->fn_guid, fn2->fn_guid));
}
/*
* Given the GUID of a snapshot, find its containing filesystem and
* (optionally) name.
*/
static nvlist_t *
fsavl_find(avl_tree_t *avl, uint64_t snapguid, char **snapname)
{
fsavl_node_t fn_find;
fsavl_node_t *fn;
fn_find.fn_guid = snapguid;
fn = avl_find(avl, &fn_find, NULL);
if (fn) {
if (snapname)
*snapname = fn->fn_snapname;
return (fn->fn_nvfs);
}
return (NULL);
}
static void
fsavl_destroy(avl_tree_t *avl)
{
fsavl_node_t *fn;
void *cookie;
if (avl == NULL)
return;
cookie = NULL;
while ((fn = avl_destroy_nodes(avl, &cookie)) != NULL)
free(fn);
avl_destroy(avl);
free(avl);
}
/*
* Given an nvlist, produce an avl tree of snapshots, ordered by guid
*/
static avl_tree_t *
fsavl_create(nvlist_t *fss)
{
avl_tree_t *fsavl;
nvpair_t *fselem = NULL;
if ((fsavl = malloc(sizeof (avl_tree_t))) == NULL)
return (NULL);
avl_create(fsavl, fsavl_compare, sizeof (fsavl_node_t),
offsetof(fsavl_node_t, fn_node));
while ((fselem = nvlist_next_nvpair(fss, fselem)) != NULL) {
nvlist_t *nvfs, *snaps;
nvpair_t *snapelem = NULL;
VERIFY(0 == nvpair_value_nvlist(fselem, &nvfs));
VERIFY(0 == nvlist_lookup_nvlist(nvfs, "snaps", &snaps));
while ((snapelem =
nvlist_next_nvpair(snaps, snapelem)) != NULL) {
fsavl_node_t *fn;
uint64_t guid;
VERIFY(0 == nvpair_value_uint64(snapelem, &guid));
if ((fn = malloc(sizeof (fsavl_node_t))) == NULL) {
fsavl_destroy(fsavl);
return (NULL);
}
fn->fn_nvfs = nvfs;
fn->fn_snapname = nvpair_name(snapelem);
fn->fn_guid = guid;
/*
* Note: if there are multiple snaps with the
* same GUID, we ignore all but one.
*/
if (avl_find(fsavl, fn, NULL) == NULL)
avl_add(fsavl, fn);
else
free(fn);
}
}
return (fsavl);
}
/*
* Routines for dealing with the giant nvlist of fs-nvlists, etc.
*/
typedef struct send_data {
/*
* assigned inside every recursive call,
* restored from *_save on return:
*
* guid of fromsnap snapshot in parent dataset
* txg of fromsnap snapshot in current dataset
* txg of tosnap snapshot in current dataset
*/
uint64_t parent_fromsnap_guid;
uint64_t fromsnap_txg;
uint64_t tosnap_txg;
/* the nvlists get accumulated during depth-first traversal */
nvlist_t *parent_snaps;
nvlist_t *fss;
nvlist_t *snapprops;
/* send-receive configuration, does not change during traversal */
const char *fsname;
const char *fromsnap;
const char *tosnap;
boolean_t recursive;
boolean_t verbose;
boolean_t replicate;
/*
* The header nvlist is of the following format:
* {
* "tosnap" -> string
* "fromsnap" -> string (if incremental)
* "fss" -> {
* id -> {
*
* "name" -> string (full name; for debugging)
* "parentfromsnap" -> number (guid of fromsnap in parent)
*
* "props" -> { name -> value (only if set here) }
* "snaps" -> { name (lastname) -> number (guid) }
* "snapprops" -> { name (lastname) -> { name -> value } }
*
* "origin" -> number (guid) (if clone)
* "sent" -> boolean (not on-disk)
* }
* }
* }
*
*/
} send_data_t;
static void send_iterate_prop(zfs_handle_t *zhp, nvlist_t *nv);
static int
send_iterate_snap(zfs_handle_t *zhp, void *arg)
{
send_data_t *sd = arg;
uint64_t guid = zhp->zfs_dmustats.dds_guid;
uint64_t txg = zhp->zfs_dmustats.dds_creation_txg;
char *snapname;
nvlist_t *nv;
snapname = strrchr(zhp->zfs_name, '@')+1;
if (sd->tosnap_txg != 0 && txg > sd->tosnap_txg) {
if (sd->verbose) {
(void) fprintf(stderr, dgettext(TEXT_DOMAIN,
"skipping snapshot %s because it was created "
"after the destination snapshot (%s)\n"),
zhp->zfs_name, sd->tosnap);
}
zfs_close(zhp);
return (0);
}
VERIFY(0 == nvlist_add_uint64(sd->parent_snaps, snapname, guid));
/*
* NB: if there is no fromsnap here (it's a newly created fs in
* an incremental replication), we will substitute the tosnap.
*/
if ((sd->fromsnap && strcmp(snapname, sd->fromsnap) == 0) ||
(sd->parent_fromsnap_guid == 0 && sd->tosnap &&
strcmp(snapname, sd->tosnap) == 0)) {
sd->parent_fromsnap_guid = guid;
}
VERIFY(0 == nvlist_alloc(&nv, NV_UNIQUE_NAME, 0));
send_iterate_prop(zhp, nv);
VERIFY(0 == nvlist_add_nvlist(sd->snapprops, snapname, nv));
nvlist_free(nv);
zfs_close(zhp);
return (0);
}
static void
send_iterate_prop(zfs_handle_t *zhp, nvlist_t *nv)
{
nvpair_t *elem = NULL;
while ((elem = nvlist_next_nvpair(zhp->zfs_props, elem)) != NULL) {
char *propname = nvpair_name(elem);
zfs_prop_t prop = zfs_name_to_prop(propname);
nvlist_t *propnv;
if (!zfs_prop_user(propname)) {
/*
* Realistically, this should never happen. However,
* we want the ability to add DSL properties without
* needing to make incompatible version changes. We
* need to ignore unknown properties to allow older
* software to still send datasets containing these
* properties, with the unknown properties elided.
*/
if (prop == ZPROP_INVAL)
continue;
if (zfs_prop_readonly(prop))
continue;
}
verify(nvpair_value_nvlist(elem, &propnv) == 0);
if (prop == ZFS_PROP_QUOTA || prop == ZFS_PROP_RESERVATION ||
prop == ZFS_PROP_REFQUOTA ||
prop == ZFS_PROP_REFRESERVATION) {
char *source;
uint64_t value;
verify(nvlist_lookup_uint64(propnv,
ZPROP_VALUE, &value) == 0);
if (zhp->zfs_type == ZFS_TYPE_SNAPSHOT)
continue;
/*
* May have no source before SPA_VERSION_RECVD_PROPS,
* but is still modifiable.
*/
if (nvlist_lookup_string(propnv,
ZPROP_SOURCE, &source) == 0) {
if ((strcmp(source, zhp->zfs_name) != 0) &&
(strcmp(source,
ZPROP_SOURCE_VAL_RECVD) != 0))
continue;
}
} else {
char *source;
if (nvlist_lookup_string(propnv,
ZPROP_SOURCE, &source) != 0)
continue;
if ((strcmp(source, zhp->zfs_name) != 0) &&
(strcmp(source, ZPROP_SOURCE_VAL_RECVD) != 0))
continue;
}
if (zfs_prop_user(propname) ||
zfs_prop_get_type(prop) == PROP_TYPE_STRING) {
char *value;
verify(nvlist_lookup_string(propnv,
ZPROP_VALUE, &value) == 0);
VERIFY(0 == nvlist_add_string(nv, propname, value));
} else {
uint64_t value;
verify(nvlist_lookup_uint64(propnv,
ZPROP_VALUE, &value) == 0);
VERIFY(0 == nvlist_add_uint64(nv, propname, value));
}
}
}
/*
* returns snapshot creation txg
* and returns 0 if the snapshot does not exist
*/
static uint64_t
get_snap_txg(libzfs_handle_t *hdl, const char *fs, const char *snap)
{
char name[ZFS_MAX_DATASET_NAME_LEN];
uint64_t txg = 0;
if (fs == NULL || fs[0] == '\0' || snap == NULL || snap[0] == '\0')
return (txg);
(void) snprintf(name, sizeof (name), "%s@%s", fs, snap);
if (zfs_dataset_exists(hdl, name, ZFS_TYPE_SNAPSHOT)) {
zfs_handle_t *zhp = zfs_open(hdl, name, ZFS_TYPE_SNAPSHOT);
if (zhp != NULL) {
txg = zfs_prop_get_int(zhp, ZFS_PROP_CREATETXG);
zfs_close(zhp);
}
}
return (txg);
}
/*
* recursively generate nvlists describing datasets. See comment
* for the data structure send_data_t above for description of contents
* of the nvlist.
*/
static int
send_iterate_fs(zfs_handle_t *zhp, void *arg)
{
send_data_t *sd = arg;
nvlist_t *nvfs, *nv;
int rv = 0;
uint64_t min_txg = 0, max_txg = 0;
uint64_t parent_fromsnap_guid_save = sd->parent_fromsnap_guid;
uint64_t fromsnap_txg_save = sd->fromsnap_txg;
uint64_t tosnap_txg_save = sd->tosnap_txg;
uint64_t txg = zhp->zfs_dmustats.dds_creation_txg;
uint64_t guid = zhp->zfs_dmustats.dds_guid;
uint64_t fromsnap_txg, tosnap_txg;
char guidstring[64];
fromsnap_txg = get_snap_txg(zhp->zfs_hdl, zhp->zfs_name, sd->fromsnap);
if (fromsnap_txg != 0)
sd->fromsnap_txg = fromsnap_txg;
tosnap_txg = get_snap_txg(zhp->zfs_hdl, zhp->zfs_name, sd->tosnap);
if (tosnap_txg != 0)
sd->tosnap_txg = tosnap_txg;
/*
* on the send side, if the current dataset does not have tosnap,
* perform two additional checks:
*
* - skip sending the current dataset if it was created later than
* the parent tosnap
* - return error if the current dataset was created earlier than
* the parent tosnap
*/
if (sd->tosnap != NULL && tosnap_txg == 0) {
if (sd->tosnap_txg != 0 && txg > sd->tosnap_txg) {
if (sd->verbose) {
(void) fprintf(stderr, dgettext(TEXT_DOMAIN,
"skipping dataset %s: snapshot %s does "
"not exist\n"), zhp->zfs_name, sd->tosnap);
}
} else {
(void) fprintf(stderr, dgettext(TEXT_DOMAIN,
"cannot send %s@%s%s: snapshot %s@%s does not "
"exist\n"), sd->fsname, sd->tosnap, sd->recursive ?
dgettext(TEXT_DOMAIN, " recursively") : "",
zhp->zfs_name, sd->tosnap);
rv = -1;
}
goto out;
}
nvfs = fnvlist_alloc();
fnvlist_add_string(nvfs, "name", zhp->zfs_name);
fnvlist_add_uint64(nvfs, "parentfromsnap",
sd->parent_fromsnap_guid);
if (zhp->zfs_dmustats.dds_origin[0]) {
zfs_handle_t *origin = zfs_open(zhp->zfs_hdl,
zhp->zfs_dmustats.dds_origin, ZFS_TYPE_SNAPSHOT);
if (origin == NULL) {
rv = -1;
goto out;
}
VERIFY(0 == nvlist_add_uint64(nvfs, "origin",
origin->zfs_dmustats.dds_guid));
}
/* iterate over props */
VERIFY(0 == nvlist_alloc(&nv, NV_UNIQUE_NAME, 0));
send_iterate_prop(zhp, nv);
VERIFY(0 == nvlist_add_nvlist(nvfs, "props", nv));
nvlist_free(nv);
/* iterate over snaps, and set sd->parent_fromsnap_guid */
if (!sd->replicate && fromsnap_txg != 0)
min_txg = fromsnap_txg;
if (!sd->replicate && tosnap_txg != 0)
max_txg = tosnap_txg;
sd->parent_fromsnap_guid = 0;
VERIFY(0 == nvlist_alloc(&sd->parent_snaps, NV_UNIQUE_NAME, 0));
VERIFY(0 == nvlist_alloc(&sd->snapprops, NV_UNIQUE_NAME, 0));
(void) zfs_iter_snapshots_sorted(zhp, send_iterate_snap, sd,
min_txg, max_txg);
VERIFY(0 == nvlist_add_nvlist(nvfs, "snaps", sd->parent_snaps));
VERIFY(0 == nvlist_add_nvlist(nvfs, "snapprops", sd->snapprops));
fnvlist_free(sd->parent_snaps);
fnvlist_free(sd->snapprops);
/* add this fs to nvlist */
(void) snprintf(guidstring, sizeof (guidstring),
"0x%llx", (longlong_t)guid);
VERIFY(0 == nvlist_add_nvlist(sd->fss, guidstring, nvfs));
nvlist_free(nvfs);
/* iterate over children */
if (sd->recursive)
rv = zfs_iter_filesystems(zhp, send_iterate_fs, sd);
out:
sd->parent_fromsnap_guid = parent_fromsnap_guid_save;
sd->fromsnap_txg = fromsnap_txg_save;
sd->tosnap_txg = tosnap_txg_save;
zfs_close(zhp);
return (rv);
}
static int
gather_nvlist(libzfs_handle_t *hdl, const char *fsname, const char *fromsnap,
const char *tosnap, boolean_t recursive, boolean_t verbose,
boolean_t replicate, nvlist_t **nvlp, avl_tree_t **avlp)
{
zfs_handle_t *zhp;
int error;
uint64_t min_txg = 0, max_txg = 0;
send_data_t sd = { 0 };
zhp = zfs_open(hdl, fsname, ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME);
if (zhp == NULL)
return (EZFS_BADTYPE);
VERIFY(0 == nvlist_alloc(&sd.fss, NV_UNIQUE_NAME, 0));
sd.fsname = fsname;
sd.fromsnap = fromsnap;
sd.tosnap = tosnap;
sd.recursive = recursive;
sd.verbose = verbose;
sd.replicate = replicate;
if ((error = send_iterate_fs(zhp, &sd)) != 0) {
nvlist_free(sd.fss);
if (avlp != NULL)
*avlp = NULL;
*nvlp = NULL;
return (error);
}
if (avlp != NULL && (*avlp = fsavl_create(sd.fss)) == NULL) {
nvlist_free(sd.fss);
*nvlp = NULL;
return (EZFS_NOMEM);
}
*nvlp = sd.fss;
return (0);
}
/*
* Routines specific to "zfs send"
*/
typedef struct send_dump_data {
/* these are all just the short snapname (the part after the @) */
const char *fromsnap;
const char *tosnap;
char prevsnap[ZFS_MAX_DATASET_NAME_LEN];
uint64_t prevsnap_obj;
boolean_t seenfrom, seento, replicate, doall, fromorigin;
boolean_t verbose, dryrun, parsable, progress, embed_data, std_out;
boolean_t progressastitle;
boolean_t large_block, compress;
int outfd;
boolean_t err;
nvlist_t *fss;
nvlist_t *snapholds;
avl_tree_t *fsavl;
snapfilter_cb_t *filter_cb;
void *filter_cb_arg;
nvlist_t *debugnv;
char holdtag[ZFS_MAX_DATASET_NAME_LEN];
int cleanup_fd;
uint64_t size;
} send_dump_data_t;
static int
zfs_send_space(zfs_handle_t *zhp, const char *snapname, const char *from,
enum lzc_send_flags flags, uint64_t *spacep)
{
libzfs_handle_t *hdl = zhp->zfs_hdl;
int error;
assert(snapname != NULL);
error = lzc_send_space(snapname, from, flags, spacep);
if (error != 0) {
char errbuf[1024];
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"warning: cannot estimate space for '%s'"), snapname);
switch (error) {
case EXDEV:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"not an earlier snapshot from the same fs"));
return (zfs_error(hdl, EZFS_CROSSTARGET, errbuf));
case ENOENT:
if (zfs_dataset_exists(hdl, snapname,
ZFS_TYPE_SNAPSHOT)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"incremental source (%s) does not exist"),
snapname);
}
return (zfs_error(hdl, EZFS_NOENT, errbuf));
case EDQUOT:
case EFBIG:
case EIO:
case ENOLINK:
case ENOSPC:
case ENXIO:
case EPIPE:
case ERANGE:
case EFAULT:
case EROFS:
case EINVAL:
zfs_error_aux(hdl, strerror(error));
return (zfs_error(hdl, EZFS_BADBACKUP, errbuf));
default:
return (zfs_standard_error(hdl, error, errbuf));
}
}
return (0);
}
/*
* Dumps a backup of the given snapshot (incremental from fromsnap if it's not
* NULL) to the file descriptor specified by outfd.
*/
static int
dump_ioctl(zfs_handle_t *zhp, const char *fromsnap, uint64_t fromsnap_obj,
boolean_t fromorigin, int outfd, enum lzc_send_flags flags,
nvlist_t *debugnv)
{
zfs_cmd_t zc = { 0 };
libzfs_handle_t *hdl = zhp->zfs_hdl;
nvlist_t *thisdbg;
assert(zhp->zfs_type == ZFS_TYPE_SNAPSHOT);
assert(fromsnap_obj == 0 || !fromorigin);
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
zc.zc_cookie = outfd;
zc.zc_obj = fromorigin;
zc.zc_sendobj = zfs_prop_get_int(zhp, ZFS_PROP_OBJSETID);
zc.zc_fromobj = fromsnap_obj;
zc.zc_flags = flags;
VERIFY(0 == nvlist_alloc(&thisdbg, NV_UNIQUE_NAME, 0));
if (fromsnap && fromsnap[0] != '\0') {
VERIFY(0 == nvlist_add_string(thisdbg,
"fromsnap", fromsnap));
}
if (zfs_ioctl(zhp->zfs_hdl, ZFS_IOC_SEND, &zc) != 0) {
char errbuf[1024];
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"warning: cannot send '%s'"), zhp->zfs_name);
VERIFY(0 == nvlist_add_uint64(thisdbg, "error", errno));
if (debugnv) {
VERIFY(0 == nvlist_add_nvlist(debugnv,
zhp->zfs_name, thisdbg));
}
nvlist_free(thisdbg);
switch (errno) {
case EXDEV:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"not an earlier snapshot from the same fs"));
return (zfs_error(hdl, EZFS_CROSSTARGET, errbuf));
case ENOENT:
if (zfs_dataset_exists(hdl, zc.zc_name,
ZFS_TYPE_SNAPSHOT)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"incremental source (@%s) does not exist"),
zc.zc_value);
}
return (zfs_error(hdl, EZFS_NOENT, errbuf));
case EDQUOT:
case EFBIG:
case EIO:
case ENOLINK:
case ENOSPC:
#ifdef illumos
case ENOSTR:
#endif
case ENXIO:
case EPIPE:
case ERANGE:
case EFAULT:
case EROFS:
zfs_error_aux(hdl, strerror(errno));
return (zfs_error(hdl, EZFS_BADBACKUP, errbuf));
default:
return (zfs_standard_error(hdl, errno, errbuf));
}
}
if (debugnv)
VERIFY(0 == nvlist_add_nvlist(debugnv, zhp->zfs_name, thisdbg));
nvlist_free(thisdbg);
return (0);
}
static void
gather_holds(zfs_handle_t *zhp, send_dump_data_t *sdd)
{
assert(zhp->zfs_type == ZFS_TYPE_SNAPSHOT);
/*
* zfs_send() only sets snapholds for sends that need them,
* e.g. replication and doall.
*/
if (sdd->snapholds == NULL)
return;
fnvlist_add_string(sdd->snapholds, zhp->zfs_name, sdd->holdtag);
}
static void *
send_progress_thread(void *arg)
{
progress_arg_t *pa = arg;
zfs_cmd_t zc = { 0 };
zfs_handle_t *zhp = pa->pa_zhp;
libzfs_handle_t *hdl = zhp->zfs_hdl;
unsigned long long bytes, total;
char buf[16];
time_t t;
struct tm *tm;
(void) strlcpy(zc.zc_name, zhp->zfs_name, sizeof (zc.zc_name));
if (!pa->pa_parsable && !pa->pa_astitle)
(void) fprintf(stderr, "TIME SENT SNAPSHOT\n");
/*
* Print the progress from ZFS_IOC_SEND_PROGRESS every second.
*/
for (;;) {
(void) sleep(1);
zc.zc_cookie = pa->pa_fd;
if (zfs_ioctl(hdl, ZFS_IOC_SEND_PROGRESS, &zc) != 0)
return ((void *)-1);
(void) time(&t);
tm = localtime(&t);
bytes = zc.zc_cookie;
if (pa->pa_astitle) {
int pct;
if (pa->pa_size > bytes)
pct = 100 * bytes / pa->pa_size;
else
pct = 100;
setproctitle("sending %s (%d%%: %llu/%llu)",
zhp->zfs_name, pct, bytes, pa->pa_size);
} else if (pa->pa_parsable) {
(void) fprintf(stderr, "%02d:%02d:%02d\t%llu\t%s\n",
tm->tm_hour, tm->tm_min, tm->tm_sec,
bytes, zhp->zfs_name);
} else {
zfs_nicenum(bytes, buf, sizeof (buf));
(void) fprintf(stderr, "%02d:%02d:%02d %5s %s\n",
tm->tm_hour, tm->tm_min, tm->tm_sec,
buf, zhp->zfs_name);
}
}
}
static void
send_print_verbose(FILE *fout, const char *tosnap, const char *fromsnap,
uint64_t size, boolean_t parsable)
{
if (parsable) {
if (fromsnap != NULL) {
(void) fprintf(fout, "incremental\t%s\t%s",
fromsnap, tosnap);
} else {
(void) fprintf(fout, "full\t%s",
tosnap);
}
} else {
if (fromsnap != NULL) {
if (strchr(fromsnap, '@') == NULL &&
strchr(fromsnap, '#') == NULL) {
(void) fprintf(fout, dgettext(TEXT_DOMAIN,
"send from @%s to %s"),
fromsnap, tosnap);
} else {
(void) fprintf(fout, dgettext(TEXT_DOMAIN,
"send from %s to %s"),
fromsnap, tosnap);
}
} else {
(void) fprintf(fout, dgettext(TEXT_DOMAIN,
"full send of %s"),
tosnap);
}
}
if (parsable) {
(void) fprintf(fout, "\t%llu",
(longlong_t)size);
} else if (size != 0) {
char buf[16];
zfs_nicenum(size, buf, sizeof (buf));
(void) fprintf(fout, dgettext(TEXT_DOMAIN,
" estimated size is %s"), buf);
}
(void) fprintf(fout, "\n");
}
static int
dump_snapshot(zfs_handle_t *zhp, void *arg)
{
send_dump_data_t *sdd = arg;
progress_arg_t pa = { 0 };
pthread_t tid;
char *thissnap;
enum lzc_send_flags flags = 0;
int err;
boolean_t isfromsnap, istosnap, fromorigin;
boolean_t exclude = B_FALSE;
FILE *fout = sdd->std_out ? stdout : stderr;
err = 0;
thissnap = strchr(zhp->zfs_name, '@') + 1;
isfromsnap = (sdd->fromsnap != NULL &&
strcmp(sdd->fromsnap, thissnap) == 0);
if (!sdd->seenfrom && isfromsnap) {
gather_holds(zhp, sdd);
sdd->seenfrom = B_TRUE;
(void) strcpy(sdd->prevsnap, thissnap);
sdd->prevsnap_obj = zfs_prop_get_int(zhp, ZFS_PROP_OBJSETID);
zfs_close(zhp);
return (0);
}
if (sdd->seento || !sdd->seenfrom) {
zfs_close(zhp);
return (0);
}
istosnap = (strcmp(sdd->tosnap, thissnap) == 0);
if (istosnap)
sdd->seento = B_TRUE;
if (sdd->large_block)
flags |= LZC_SEND_FLAG_LARGE_BLOCK;
if (sdd->embed_data)
flags |= LZC_SEND_FLAG_EMBED_DATA;
if (sdd->compress)
flags |= LZC_SEND_FLAG_COMPRESS;
if (!sdd->doall && !isfromsnap && !istosnap) {
if (sdd->replicate) {
char *snapname;
nvlist_t *snapprops;
/*
* Filter out all intermediate snapshots except origin
* snapshots needed to replicate clones.
*/
nvlist_t *nvfs = fsavl_find(sdd->fsavl,
zhp->zfs_dmustats.dds_guid, &snapname);
VERIFY(0 == nvlist_lookup_nvlist(nvfs,
"snapprops", &snapprops));
VERIFY(0 == nvlist_lookup_nvlist(snapprops,
thissnap, &snapprops));
exclude = !nvlist_exists(snapprops, "is_clone_origin");
} else {
exclude = B_TRUE;
}
}
/*
* If a filter function exists, call it to determine whether
* this snapshot will be sent.
*/
if (exclude || (sdd->filter_cb != NULL &&
sdd->filter_cb(zhp, sdd->filter_cb_arg) == B_FALSE)) {
/*
* This snapshot is filtered out. Don't send it, and don't
* set prevsnap_obj, so it will be as if this snapshot didn't
* exist, and the next accepted snapshot will be sent as
* an incremental from the last accepted one, or as the
* first (and full) snapshot in the case of a replication,
* non-incremental send.
*/
zfs_close(zhp);
return (0);
}
gather_holds(zhp, sdd);
fromorigin = sdd->prevsnap[0] == '\0' &&
(sdd->fromorigin || sdd->replicate);
if (sdd->verbose || sdd->progress) {
uint64_t size = 0;
char fromds[ZFS_MAX_DATASET_NAME_LEN];
if (sdd->prevsnap[0] != '\0') {
(void) strlcpy(fromds, zhp->zfs_name, sizeof (fromds));
*(strchr(fromds, '@') + 1) = '\0';
(void) strlcat(fromds, sdd->prevsnap, sizeof (fromds));
}
if (zfs_send_space(zhp, zhp->zfs_name,
sdd->prevsnap[0] ? fromds : NULL, flags, &size) != 0) {
size = 0; /* cannot estimate send space */
} else {
send_print_verbose(fout, zhp->zfs_name,
sdd->prevsnap[0] ? sdd->prevsnap : NULL,
size, sdd->parsable);
}
sdd->size += size;
}
if (!sdd->dryrun) {
/*
* If progress reporting is requested, spawn a new thread to
* poll ZFS_IOC_SEND_PROGRESS at a regular interval.
*/
if (sdd->progress) {
pa.pa_zhp = zhp;
pa.pa_fd = sdd->outfd;
pa.pa_parsable = sdd->parsable;
pa.pa_size = sdd->size;
pa.pa_astitle = sdd->progressastitle;
if ((err = pthread_create(&tid, NULL,
send_progress_thread, &pa)) != 0) {
zfs_close(zhp);
return (err);
}
}
err = dump_ioctl(zhp, sdd->prevsnap, sdd->prevsnap_obj,
fromorigin, sdd->outfd, flags, sdd->debugnv);
if (sdd->progress) {
(void) pthread_cancel(tid);
(void) pthread_join(tid, NULL);
}
}
(void) strcpy(sdd->prevsnap, thissnap);
sdd->prevsnap_obj = zfs_prop_get_int(zhp, ZFS_PROP_OBJSETID);
zfs_close(zhp);
return (err);
}
static int
dump_filesystem(zfs_handle_t *zhp, void *arg)
{
int rv = 0;
uint64_t min_txg = 0, max_txg = 0;
send_dump_data_t *sdd = arg;
boolean_t missingfrom = B_FALSE;
zfs_cmd_t zc = { 0 };
(void) snprintf(zc.zc_name, sizeof (zc.zc_name), "%s@%s",
zhp->zfs_name, sdd->tosnap);
if (ioctl(zhp->zfs_hdl->libzfs_fd, ZFS_IOC_OBJSET_STATS, &zc) != 0) {
(void) fprintf(stderr, dgettext(TEXT_DOMAIN,
"WARNING: could not send %s@%s: does not exist\n"),
zhp->zfs_name, sdd->tosnap);
sdd->err = B_TRUE;
return (0);
}
if (sdd->replicate && sdd->fromsnap) {
/*
* If this fs does not have fromsnap, and we're doing
* recursive, we need to send a full stream from the
* beginning (or an incremental from the origin if this
* is a clone). If we're doing non-recursive, then let
* them get the error.
*/
(void) snprintf(zc.zc_name, sizeof (zc.zc_name), "%s@%s",
zhp->zfs_name, sdd->fromsnap);
if (ioctl(zhp->zfs_hdl->libzfs_fd,
ZFS_IOC_OBJSET_STATS, &zc) != 0) {
missingfrom = B_TRUE;
}
}
sdd->seenfrom = sdd->seento = sdd->prevsnap[0] = 0;
sdd->prevsnap_obj = 0;
if (sdd->fromsnap == NULL || missingfrom)
sdd->seenfrom = B_TRUE;
if (!sdd->replicate && sdd->fromsnap != NULL)
min_txg = get_snap_txg(zhp->zfs_hdl, zhp->zfs_name,
sdd->fromsnap);
if (!sdd->replicate && sdd->tosnap != NULL)
max_txg = get_snap_txg(zhp->zfs_hdl, zhp->zfs_name,
sdd->tosnap);
rv = zfs_iter_snapshots_sorted(zhp, dump_snapshot, arg,
min_txg, max_txg);
if (!sdd->seenfrom) {
(void) fprintf(stderr, dgettext(TEXT_DOMAIN,
"WARNING: could not send %s@%s:\n"
"incremental source (%s@%s) does not exist\n"),
zhp->zfs_name, sdd->tosnap,
zhp->zfs_name, sdd->fromsnap);
sdd->err = B_TRUE;
} else if (!sdd->seento) {
if (sdd->fromsnap) {
(void) fprintf(stderr, dgettext(TEXT_DOMAIN,
"WARNING: could not send %s@%s:\n"
"incremental source (%s@%s) "
"is not earlier than it\n"),
zhp->zfs_name, sdd->tosnap,
zhp->zfs_name, sdd->fromsnap);
} else {
(void) fprintf(stderr, dgettext(TEXT_DOMAIN,
"WARNING: "
"could not send %s@%s: does not exist\n"),
zhp->zfs_name, sdd->tosnap);
}
sdd->err = B_TRUE;
}
return (rv);
}
static int
dump_filesystems(zfs_handle_t *rzhp, void *arg)
{
send_dump_data_t *sdd = arg;
nvpair_t *fspair;
boolean_t needagain, progress;
if (!sdd->replicate)
return (dump_filesystem(rzhp, sdd));
/* Mark the clone origin snapshots. */
for (fspair = nvlist_next_nvpair(sdd->fss, NULL); fspair;
fspair = nvlist_next_nvpair(sdd->fss, fspair)) {
nvlist_t *nvfs;
uint64_t origin_guid = 0;
VERIFY(0 == nvpair_value_nvlist(fspair, &nvfs));
(void) nvlist_lookup_uint64(nvfs, "origin", &origin_guid);
if (origin_guid != 0) {
char *snapname;
nvlist_t *origin_nv = fsavl_find(sdd->fsavl,
origin_guid, &snapname);
if (origin_nv != NULL) {
nvlist_t *snapprops;
VERIFY(0 == nvlist_lookup_nvlist(origin_nv,
"snapprops", &snapprops));
VERIFY(0 == nvlist_lookup_nvlist(snapprops,
snapname, &snapprops));
VERIFY(0 == nvlist_add_boolean(
snapprops, "is_clone_origin"));
}
}
}
again:
needagain = progress = B_FALSE;
for (fspair = nvlist_next_nvpair(sdd->fss, NULL); fspair;
fspair = nvlist_next_nvpair(sdd->fss, fspair)) {
nvlist_t *fslist, *parent_nv;
char *fsname;
zfs_handle_t *zhp;
int err;
uint64_t origin_guid = 0;
uint64_t parent_guid = 0;
VERIFY(nvpair_value_nvlist(fspair, &fslist) == 0);
if (nvlist_lookup_boolean(fslist, "sent") == 0)
continue;
VERIFY(nvlist_lookup_string(fslist, "name", &fsname) == 0);
(void) nvlist_lookup_uint64(fslist, "origin", &origin_guid);
(void) nvlist_lookup_uint64(fslist, "parentfromsnap",
&parent_guid);
if (parent_guid != 0) {
parent_nv = fsavl_find(sdd->fsavl, parent_guid, NULL);
if (!nvlist_exists(parent_nv, "sent")) {
/* parent has not been sent; skip this one */
needagain = B_TRUE;
continue;
}
}
if (origin_guid != 0) {
nvlist_t *origin_nv = fsavl_find(sdd->fsavl,
origin_guid, NULL);
if (origin_nv != NULL &&
!nvlist_exists(origin_nv, "sent")) {
/*
* origin has not been sent yet;
* skip this clone.
*/
needagain = B_TRUE;
continue;
}
}
zhp = zfs_open(rzhp->zfs_hdl, fsname, ZFS_TYPE_DATASET);
if (zhp == NULL)
return (-1);
err = dump_filesystem(zhp, sdd);
VERIFY(nvlist_add_boolean(fslist, "sent") == 0);
progress = B_TRUE;
zfs_close(zhp);
if (err)
return (err);
}
if (needagain) {
assert(progress);
goto again;
}
/* clean out the sent flags in case we reuse this fss */
for (fspair = nvlist_next_nvpair(sdd->fss, NULL); fspair;
fspair = nvlist_next_nvpair(sdd->fss, fspair)) {
nvlist_t *fslist;
VERIFY(nvpair_value_nvlist(fspair, &fslist) == 0);
(void) nvlist_remove_all(fslist, "sent");
}
return (0);
}
nvlist_t *
zfs_send_resume_token_to_nvlist(libzfs_handle_t *hdl, const char *token)
{
unsigned int version;
int nread;
unsigned long long checksum, packed_len;
/*
* Decode token header, which is:
* <token version>-<checksum of payload>-<uncompressed payload length>
* Note that the only supported token version is 1.
*/
nread = sscanf(token, "%u-%llx-%llx-",
&version, &checksum, &packed_len);
if (nread != 3) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"resume token is corrupt (invalid format)"));
return (NULL);
}
if (version != ZFS_SEND_RESUME_TOKEN_VERSION) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"resume token is corrupt (invalid version %u)"),
version);
return (NULL);
}
/* convert hexadecimal representation to binary */
token = strrchr(token, '-') + 1;
int len = strlen(token) / 2;
unsigned char *compressed = zfs_alloc(hdl, len);
for (int i = 0; i < len; i++) {
nread = sscanf(token + i * 2, "%2hhx", compressed + i);
if (nread != 1) {
free(compressed);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"resume token is corrupt "
"(payload is not hex-encoded)"));
return (NULL);
}
}
/* verify checksum */
zio_cksum_t cksum;
fletcher_4_native(compressed, len, NULL, &cksum);
if (cksum.zc_word[0] != checksum) {
free(compressed);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"resume token is corrupt (incorrect checksum)"));
return (NULL);
}
/* uncompress */
void *packed = zfs_alloc(hdl, packed_len);
uLongf packed_len_long = packed_len;
if (uncompress(packed, &packed_len_long, compressed, len) != Z_OK ||
packed_len_long != packed_len) {
free(packed);
free(compressed);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"resume token is corrupt (decompression failed)"));
return (NULL);
}
/* unpack nvlist */
nvlist_t *nv;
int error = nvlist_unpack(packed, packed_len, &nv, KM_SLEEP);
free(packed);
free(compressed);
if (error != 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"resume token is corrupt (nvlist_unpack failed)"));
return (NULL);
}
return (nv);
}
int
zfs_send_resume(libzfs_handle_t *hdl, sendflags_t *flags, int outfd,
const char *resume_token)
{
char errbuf[1024];
char *toname;
char *fromname = NULL;
uint64_t resumeobj, resumeoff, toguid, fromguid, bytes;
zfs_handle_t *zhp;
int error = 0;
char name[ZFS_MAX_DATASET_NAME_LEN];
enum lzc_send_flags lzc_flags = 0;
uint64_t size = 0;
FILE *fout = (flags->verbose && flags->dryrun) ? stdout : stderr;
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot resume send"));
nvlist_t *resume_nvl =
zfs_send_resume_token_to_nvlist(hdl, resume_token);
if (resume_nvl == NULL) {
/*
* zfs_error_aux has already been set by
* zfs_send_resume_token_to_nvlist
*/
return (zfs_error(hdl, EZFS_FAULT, errbuf));
}
if (flags->verbose) {
(void) fprintf(fout, dgettext(TEXT_DOMAIN,
"resume token contents:\n"));
nvlist_print(fout, resume_nvl);
}
if (nvlist_lookup_string(resume_nvl, "toname", &toname) != 0 ||
nvlist_lookup_uint64(resume_nvl, "object", &resumeobj) != 0 ||
nvlist_lookup_uint64(resume_nvl, "offset", &resumeoff) != 0 ||
nvlist_lookup_uint64(resume_nvl, "bytes", &bytes) != 0 ||
nvlist_lookup_uint64(resume_nvl, "toguid", &toguid) != 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"resume token is corrupt"));
return (zfs_error(hdl, EZFS_FAULT, errbuf));
}
fromguid = 0;
(void) nvlist_lookup_uint64(resume_nvl, "fromguid", &fromguid);
if (flags->largeblock || nvlist_exists(resume_nvl, "largeblockok"))
lzc_flags |= LZC_SEND_FLAG_LARGE_BLOCK;
if (flags->embed_data || nvlist_exists(resume_nvl, "embedok"))
lzc_flags |= LZC_SEND_FLAG_EMBED_DATA;
if (flags->compress || nvlist_exists(resume_nvl, "compressok"))
lzc_flags |= LZC_SEND_FLAG_COMPRESS;
if (guid_to_name(hdl, toname, toguid, B_FALSE, name) != 0) {
if (zfs_dataset_exists(hdl, toname, ZFS_TYPE_DATASET)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' is no longer the same snapshot used in "
"the initial send"), toname);
} else {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' used in the initial send no longer exists"),
toname);
}
return (zfs_error(hdl, EZFS_BADPATH, errbuf));
}
zhp = zfs_open(hdl, name, ZFS_TYPE_DATASET);
if (zhp == NULL) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"unable to access '%s'"), name);
return (zfs_error(hdl, EZFS_BADPATH, errbuf));
}
if (fromguid != 0) {
if (guid_to_name(hdl, toname, fromguid, B_TRUE, name) != 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"incremental source %#llx no longer exists"),
(longlong_t)fromguid);
return (zfs_error(hdl, EZFS_BADPATH, errbuf));
}
fromname = name;
}
if (flags->progress || flags->verbose) {
error = lzc_send_space(zhp->zfs_name, fromname,
lzc_flags, &size);
if (error == 0)
size = MAX(0, (int64_t)(size - bytes));
}
if (flags->verbose) {
send_print_verbose(fout, zhp->zfs_name, fromname,
size, flags->parsable);
}
if (!flags->dryrun) {
progress_arg_t pa = { 0 };
pthread_t tid;
/*
* If progress reporting is requested, spawn a new thread to
* poll ZFS_IOC_SEND_PROGRESS at a regular interval.
*/
if (flags->progress) {
pa.pa_zhp = zhp;
pa.pa_fd = outfd;
pa.pa_parsable = flags->parsable;
pa.pa_size = size;
pa.pa_astitle = flags->progressastitle;
error = pthread_create(&tid, NULL,
send_progress_thread, &pa);
if (error != 0) {
zfs_close(zhp);
return (error);
}
}
error = lzc_send_resume(zhp->zfs_name, fromname, outfd,
lzc_flags, resumeobj, resumeoff);
if (flags->progress) {
(void) pthread_cancel(tid);
(void) pthread_join(tid, NULL);
}
char errbuf[1024];
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"warning: cannot send '%s'"), zhp->zfs_name);
zfs_close(zhp);
switch (error) {
case 0:
return (0);
case EXDEV:
case ENOENT:
case EDQUOT:
case EFBIG:
case EIO:
case ENOLINK:
case ENOSPC:
#ifdef illumos
case ENOSTR:
#endif
case ENXIO:
case EPIPE:
case ERANGE:
case EFAULT:
case EROFS:
zfs_error_aux(hdl, strerror(errno));
return (zfs_error(hdl, EZFS_BADBACKUP, errbuf));
default:
return (zfs_standard_error(hdl, errno, errbuf));
}
}
zfs_close(zhp);
return (error);
}
/*
* Generate a send stream for the dataset identified by the argument zhp.
*
* The content of the send stream is the snapshot identified by
* 'tosnap'. Incremental streams are requested in two ways:
* - from the snapshot identified by "fromsnap" (if non-null) or
* - from the origin of the dataset identified by zhp, which must
* be a clone. In this case, "fromsnap" is null and "fromorigin"
* is TRUE.
*
* The send stream is recursive (i.e. dumps a hierarchy of snapshots) and
* uses a special header (with a hdrtype field of DMU_COMPOUNDSTREAM)
* if "replicate" is set. If "doall" is set, dump all the intermediate
* snapshots. The DMU_COMPOUNDSTREAM header is used in the "doall"
* case too. If "props" is set, send properties.
*/
int
zfs_send(zfs_handle_t *zhp, const char *fromsnap, const char *tosnap,
sendflags_t *flags, int outfd, snapfilter_cb_t filter_func,
void *cb_arg, nvlist_t **debugnvp)
{
char errbuf[1024];
send_dump_data_t sdd = { 0 };
int err = 0;
nvlist_t *fss = NULL;
avl_tree_t *fsavl = NULL;
static uint64_t holdseq;
int spa_version;
pthread_t tid = 0;
int pipefd[2];
dedup_arg_t dda = { 0 };
int featureflags = 0;
FILE *fout;
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot send '%s'"), zhp->zfs_name);
if (fromsnap && fromsnap[0] == '\0') {
zfs_error_aux(zhp->zfs_hdl, dgettext(TEXT_DOMAIN,
"zero-length incremental source"));
return (zfs_error(zhp->zfs_hdl, EZFS_NOENT, errbuf));
}
if (zhp->zfs_type == ZFS_TYPE_FILESYSTEM) {
uint64_t version;
version = zfs_prop_get_int(zhp, ZFS_PROP_VERSION);
if (version >= ZPL_VERSION_SA) {
featureflags |= DMU_BACKUP_FEATURE_SA_SPILL;
}
}
if (flags->dedup && !flags->dryrun) {
featureflags |= (DMU_BACKUP_FEATURE_DEDUP |
DMU_BACKUP_FEATURE_DEDUPPROPS);
if ((err = pipe(pipefd)) != 0) {
zfs_error_aux(zhp->zfs_hdl, strerror(errno));
return (zfs_error(zhp->zfs_hdl, EZFS_PIPEFAILED,
errbuf));
}
dda.outputfd = outfd;
dda.inputfd = pipefd[1];
dda.dedup_hdl = zhp->zfs_hdl;
if ((err = pthread_create(&tid, NULL, cksummer, &dda)) != 0) {
(void) close(pipefd[0]);
(void) close(pipefd[1]);
zfs_error_aux(zhp->zfs_hdl, strerror(errno));
return (zfs_error(zhp->zfs_hdl,
EZFS_THREADCREATEFAILED, errbuf));
}
}
if (flags->replicate || flags->doall || flags->props) {
dmu_replay_record_t drr = { 0 };
char *packbuf = NULL;
size_t buflen = 0;
zio_cksum_t zc = { 0 };
if (flags->replicate || flags->props) {
nvlist_t *hdrnv;
VERIFY(0 == nvlist_alloc(&hdrnv, NV_UNIQUE_NAME, 0));
if (fromsnap) {
VERIFY(0 == nvlist_add_string(hdrnv,
"fromsnap", fromsnap));
}
VERIFY(0 == nvlist_add_string(hdrnv, "tosnap", tosnap));
if (!flags->replicate) {
VERIFY(0 == nvlist_add_boolean(hdrnv,
"not_recursive"));
}
err = gather_nvlist(zhp->zfs_hdl, zhp->zfs_name,
fromsnap, tosnap, flags->replicate, flags->verbose,
flags->replicate, &fss, &fsavl);
if (err)
goto err_out;
VERIFY(0 == nvlist_add_nvlist(hdrnv, "fss", fss));
err = nvlist_pack(hdrnv, &packbuf, &buflen,
NV_ENCODE_XDR, 0);
if (debugnvp)
*debugnvp = hdrnv;
else
nvlist_free(hdrnv);
if (err)
goto stderr_out;
}
if (!flags->dryrun) {
/* write first begin record */
drr.drr_type = DRR_BEGIN;
drr.drr_u.drr_begin.drr_magic = DMU_BACKUP_MAGIC;
DMU_SET_STREAM_HDRTYPE(drr.drr_u.drr_begin.
drr_versioninfo, DMU_COMPOUNDSTREAM);
DMU_SET_FEATUREFLAGS(drr.drr_u.drr_begin.
drr_versioninfo, featureflags);
(void) snprintf(drr.drr_u.drr_begin.drr_toname,
sizeof (drr.drr_u.drr_begin.drr_toname),
"%s@%s", zhp->zfs_name, tosnap);
drr.drr_payloadlen = buflen;
err = dump_record(&drr, packbuf, buflen, &zc, outfd);
free(packbuf);
if (err != 0)
goto stderr_out;
/* write end record */
bzero(&drr, sizeof (drr));
drr.drr_type = DRR_END;
drr.drr_u.drr_end.drr_checksum = zc;
err = write(outfd, &drr, sizeof (drr));
if (err == -1) {
err = errno;
goto stderr_out;
}
err = 0;
}
}
/* dump each stream */
sdd.fromsnap = fromsnap;
sdd.tosnap = tosnap;
if (tid != 0)
sdd.outfd = pipefd[0];
else
sdd.outfd = outfd;
sdd.replicate = flags->replicate;
sdd.doall = flags->doall;
sdd.fromorigin = flags->fromorigin;
sdd.fss = fss;
sdd.fsavl = fsavl;
sdd.verbose = flags->verbose;
sdd.parsable = flags->parsable;
sdd.progress = flags->progress;
sdd.progressastitle = flags->progressastitle;
sdd.dryrun = flags->dryrun;
sdd.large_block = flags->largeblock;
sdd.embed_data = flags->embed_data;
sdd.compress = flags->compress;
sdd.filter_cb = filter_func;
sdd.filter_cb_arg = cb_arg;
if (debugnvp)
sdd.debugnv = *debugnvp;
if (sdd.verbose && sdd.dryrun)
sdd.std_out = B_TRUE;
fout = sdd.std_out ? stdout : stderr;
/*
* Some flags require that we place user holds on the datasets that are
* being sent so they don't get destroyed during the send. We can skip
* this step if the pool is imported read-only since the datasets cannot
* be destroyed.
*/
if (!flags->dryrun && !zpool_get_prop_int(zfs_get_pool_handle(zhp),
ZPOOL_PROP_READONLY, NULL) &&
zfs_spa_version(zhp, &spa_version) == 0 &&
spa_version >= SPA_VERSION_USERREFS &&
(flags->doall || flags->replicate)) {
++holdseq;
(void) snprintf(sdd.holdtag, sizeof (sdd.holdtag),
".send-%d-%llu", getpid(), (u_longlong_t)holdseq);
sdd.cleanup_fd = open(ZFS_DEV, O_RDWR|O_EXCL);
if (sdd.cleanup_fd < 0) {
err = errno;
goto stderr_out;
}
sdd.snapholds = fnvlist_alloc();
} else {
sdd.cleanup_fd = -1;
sdd.snapholds = NULL;
}
if (flags->progress || flags->verbose || sdd.snapholds != NULL) {
/*
* Do a verbose no-op dry run to get all the verbose output
* or to gather snapshot hold's before generating any data,
* then do a non-verbose real run to generate the streams.
*/
sdd.dryrun = B_TRUE;
err = dump_filesystems(zhp, &sdd);
if (err != 0)
goto stderr_out;
if (flags->verbose) {
if (flags->parsable) {
(void) fprintf(fout, "size\t%llu\n",
(longlong_t)sdd.size);
} else {
char buf[16];
zfs_nicenum(sdd.size, buf, sizeof (buf));
(void) fprintf(fout, dgettext(TEXT_DOMAIN,
"total estimated size is %s\n"), buf);
}
}
/* Ensure no snaps found is treated as an error. */
if (!sdd.seento) {
err = ENOENT;
goto err_out;
}
/* Skip the second run if dryrun was requested. */
if (flags->dryrun)
goto err_out;
if (sdd.snapholds != NULL) {
err = zfs_hold_nvl(zhp, sdd.cleanup_fd, sdd.snapholds);
if (err != 0)
goto stderr_out;
fnvlist_free(sdd.snapholds);
sdd.snapholds = NULL;
}
sdd.dryrun = B_FALSE;
sdd.verbose = B_FALSE;
}
err = dump_filesystems(zhp, &sdd);
fsavl_destroy(fsavl);
nvlist_free(fss);
/* Ensure no snaps found is treated as an error. */
if (err == 0 && !sdd.seento)
err = ENOENT;
if (tid != 0) {
if (err != 0)
(void) pthread_cancel(tid);
(void) close(pipefd[0]);
(void) pthread_join(tid, NULL);
}
if (sdd.cleanup_fd != -1) {
VERIFY(0 == close(sdd.cleanup_fd));
sdd.cleanup_fd = -1;
}
if (!flags->dryrun && (flags->replicate || flags->doall ||
flags->props)) {
/*
* write final end record. NB: want to do this even if
* there was some error, because it might not be totally
* failed.
*/
dmu_replay_record_t drr = { 0 };
drr.drr_type = DRR_END;
if (write(outfd, &drr, sizeof (drr)) == -1) {
return (zfs_standard_error(zhp->zfs_hdl,
errno, errbuf));
}
}
return (err || sdd.err);
stderr_out:
err = zfs_standard_error(zhp->zfs_hdl, err, errbuf);
err_out:
fsavl_destroy(fsavl);
nvlist_free(fss);
fnvlist_free(sdd.snapholds);
if (sdd.cleanup_fd != -1)
VERIFY(0 == close(sdd.cleanup_fd));
if (tid != 0) {
(void) pthread_cancel(tid);
(void) close(pipefd[0]);
(void) pthread_join(tid, NULL);
}
return (err);
}
int
zfs_send_one(zfs_handle_t *zhp, const char *from, int fd, sendflags_t flags)
{
int err = 0;
libzfs_handle_t *hdl = zhp->zfs_hdl;
enum lzc_send_flags lzc_flags = 0;
FILE *fout = (flags.verbose && flags.dryrun) ? stdout : stderr;
char errbuf[1024];
if (flags.largeblock)
lzc_flags |= LZC_SEND_FLAG_LARGE_BLOCK;
if (flags.embed_data)
lzc_flags |= LZC_SEND_FLAG_EMBED_DATA;
if (flags.compress)
lzc_flags |= LZC_SEND_FLAG_COMPRESS;
if (flags.verbose) {
uint64_t size = 0;
err = lzc_send_space(zhp->zfs_name, from, lzc_flags, &size);
if (err == 0) {
send_print_verbose(fout, zhp->zfs_name, from, size,
flags.parsable);
if (flags.parsable) {
(void) fprintf(fout, "size\t%llu\n",
(longlong_t)size);
} else {
char buf[16];
zfs_nicenum(size, buf, sizeof (buf));
(void) fprintf(fout, dgettext(TEXT_DOMAIN,
"total estimated size is %s\n"), buf);
}
} else {
(void) fprintf(stderr, "Cannot estimate send size: "
"%s\n", strerror(errno));
}
}
if (flags.dryrun)
return (err);
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"warning: cannot send '%s'"), zhp->zfs_name);
err = lzc_send(zhp->zfs_name, from, fd, lzc_flags);
if (err != 0) {
switch (errno) {
case EXDEV:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"not an earlier snapshot from the same fs"));
return (zfs_error(hdl, EZFS_CROSSTARGET, errbuf));
case ENOENT:
case ESRCH:
if (lzc_exists(zhp->zfs_name)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"incremental source (%s) does not exist"),
from);
}
return (zfs_error(hdl, EZFS_NOENT, errbuf));
case EBUSY:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"target is busy; if a filesystem, "
"it must not be mounted"));
return (zfs_error(hdl, EZFS_BUSY, errbuf));
case EDQUOT:
case EFBIG:
case EIO:
case ENOLINK:
case ENOSPC:
#ifdef illumos
case ENOSTR:
#endif
case ENXIO:
case EPIPE:
case ERANGE:
case EFAULT:
case EROFS:
zfs_error_aux(hdl, strerror(errno));
return (zfs_error(hdl, EZFS_BADBACKUP, errbuf));
default:
return (zfs_standard_error(hdl, errno, errbuf));
}
}
return (err != 0);
}
/*
* Routines specific to "zfs recv"
*/
static int
recv_read(libzfs_handle_t *hdl, int fd, void *buf, int ilen,
boolean_t byteswap, zio_cksum_t *zc)
{
char *cp = buf;
int rv;
int len = ilen;
assert(ilen <= SPA_MAXBLOCKSIZE);
do {
rv = read(fd, cp, len);
cp += rv;
len -= rv;
} while (rv > 0);
if (rv < 0 || len != 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"failed to read from stream"));
return (zfs_error(hdl, EZFS_BADSTREAM, dgettext(TEXT_DOMAIN,
"cannot receive")));
}
if (zc) {
if (byteswap)
(void) fletcher_4_incremental_byteswap(buf, ilen, zc);
else
(void) fletcher_4_incremental_native(buf, ilen, zc);
}
return (0);
}
static int
recv_read_nvlist(libzfs_handle_t *hdl, int fd, int len, nvlist_t **nvp,
boolean_t byteswap, zio_cksum_t *zc)
{
char *buf;
int err;
buf = zfs_alloc(hdl, len);
if (buf == NULL)
return (ENOMEM);
err = recv_read(hdl, fd, buf, len, byteswap, zc);
if (err != 0) {
free(buf);
return (err);
}
err = nvlist_unpack(buf, len, nvp, 0);
free(buf);
if (err != 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid "
"stream (malformed nvlist)"));
return (EINVAL);
}
return (0);
}
static int
recv_rename(libzfs_handle_t *hdl, const char *name, const char *tryname,
int baselen, char *newname, recvflags_t *flags)
{
static int seq;
int err;
prop_changelist_t *clp;
zfs_handle_t *zhp;
zhp = zfs_open(hdl, name, ZFS_TYPE_DATASET);
if (zhp == NULL)
return (-1);
clp = changelist_gather(zhp, ZFS_PROP_NAME, 0,
flags->force ? MS_FORCE : 0);
zfs_close(zhp);
if (clp == NULL)
return (-1);
err = changelist_prefix(clp);
if (err)
return (err);
if (tryname) {
(void) strcpy(newname, tryname);
if (flags->verbose) {
(void) printf("attempting rename %s to %s\n",
name, newname);
}
err = lzc_rename(name, newname);
if (err == 0)
changelist_rename(clp, name, tryname);
} else {
err = ENOENT;
}
if (err != 0 && strncmp(name + baselen, "recv-", 5) != 0) {
seq++;
(void) snprintf(newname, ZFS_MAX_DATASET_NAME_LEN,
"%.*srecv-%u-%u", baselen, name, getpid(), seq);
if (flags->verbose) {
(void) printf("failed - trying rename %s to %s\n",
name, newname);
}
err = lzc_rename(name, newname);
if (err == 0)
changelist_rename(clp, name, newname);
if (err && flags->verbose) {
(void) printf("failed (%u) - "
"will try again on next pass\n", errno);
}
err = EAGAIN;
} else if (flags->verbose) {
if (err == 0)
(void) printf("success\n");
else
(void) printf("failed (%u)\n", errno);
}
(void) changelist_postfix(clp);
changelist_free(clp);
return (err);
}
static int
recv_destroy(libzfs_handle_t *hdl, const char *name, int baselen,
char *newname, recvflags_t *flags)
{
int err = 0;
prop_changelist_t *clp;
zfs_handle_t *zhp;
boolean_t defer = B_FALSE;
int spa_version;
zhp = zfs_open(hdl, name, ZFS_TYPE_DATASET);
if (zhp == NULL)
return (-1);
clp = changelist_gather(zhp, ZFS_PROP_NAME, 0,
flags->force ? MS_FORCE : 0);
if (zfs_get_type(zhp) == ZFS_TYPE_SNAPSHOT &&
zfs_spa_version(zhp, &spa_version) == 0 &&
spa_version >= SPA_VERSION_USERREFS)
defer = B_TRUE;
zfs_close(zhp);
if (clp == NULL)
return (-1);
err = changelist_prefix(clp);
if (err)
return (err);
if (flags->verbose)
(void) printf("attempting destroy %s\n", name);
if (zhp->zfs_type == ZFS_TYPE_SNAPSHOT) {
nvlist_t *nv = fnvlist_alloc();
fnvlist_add_boolean(nv, name);
err = lzc_destroy_snaps(nv, defer, NULL);
fnvlist_free(nv);
} else {
err = lzc_destroy(name);
}
if (err == 0) {
if (flags->verbose)
(void) printf("success\n");
changelist_remove(clp, name);
}
(void) changelist_postfix(clp);
changelist_free(clp);
/*
* Deferred destroy might destroy the snapshot or only mark it to be
* destroyed later, and it returns success in either case.
*/
if (err != 0 || (defer && zfs_dataset_exists(hdl, name,
ZFS_TYPE_SNAPSHOT))) {
err = recv_rename(hdl, name, NULL, baselen, newname, flags);
}
return (err);
}
typedef struct guid_to_name_data {
uint64_t guid;
boolean_t bookmark_ok;
char *name;
char *skip;
} guid_to_name_data_t;
static int
guid_to_name_cb(zfs_handle_t *zhp, void *arg)
{
guid_to_name_data_t *gtnd = arg;
const char *slash;
int err;
if (gtnd->skip != NULL &&
(slash = strrchr(zhp->zfs_name, '/')) != NULL &&
strcmp(slash + 1, gtnd->skip) == 0) {
zfs_close(zhp);
return (0);
}
if (zfs_prop_get_int(zhp, ZFS_PROP_GUID) == gtnd->guid) {
(void) strcpy(gtnd->name, zhp->zfs_name);
zfs_close(zhp);
return (EEXIST);
}
err = zfs_iter_children(zhp, guid_to_name_cb, gtnd);
if (err != EEXIST && gtnd->bookmark_ok)
err = zfs_iter_bookmarks(zhp, guid_to_name_cb, gtnd);
zfs_close(zhp);
return (err);
}
/*
* Attempt to find the local dataset associated with this guid. In the case of
* multiple matches, we attempt to find the "best" match by searching
* progressively larger portions of the hierarchy. This allows one to send a
* tree of datasets individually and guarantee that we will find the source
* guid within that hierarchy, even if there are multiple matches elsewhere.
*/
static int
guid_to_name(libzfs_handle_t *hdl, const char *parent, uint64_t guid,
boolean_t bookmark_ok, char *name)
{
char pname[ZFS_MAX_DATASET_NAME_LEN];
guid_to_name_data_t gtnd;
gtnd.guid = guid;
gtnd.bookmark_ok = bookmark_ok;
gtnd.name = name;
gtnd.skip = NULL;
/*
* Search progressively larger portions of the hierarchy, starting
* with the filesystem specified by 'parent'. This will
* select the "most local" version of the origin snapshot in the case
* that there are multiple matching snapshots in the system.
*/
(void) strlcpy(pname, parent, sizeof (pname));
char *cp = strrchr(pname, '@');
if (cp == NULL)
cp = strchr(pname, '\0');
for (; cp != NULL; cp = strrchr(pname, '/')) {
/* Chop off the last component and open the parent */
*cp = '\0';
zfs_handle_t *zhp = make_dataset_handle(hdl, pname);
if (zhp == NULL)
continue;
int err = guid_to_name_cb(zfs_handle_dup(zhp), >nd);
if (err != EEXIST)
err = zfs_iter_children(zhp, guid_to_name_cb, >nd);
if (err != EEXIST && bookmark_ok)
err = zfs_iter_bookmarks(zhp, guid_to_name_cb, >nd);
zfs_close(zhp);
if (err == EEXIST)
return (0);
/*
* Remember the last portion of the dataset so we skip it next
* time through (as we've already searched that portion of the
* hierarchy).
*/
gtnd.skip = strrchr(pname, '/') + 1;
}
return (ENOENT);
}
/*
* Return +1 if guid1 is before guid2, 0 if they are the same, and -1 if
* guid1 is after guid2.
*/
static int
created_before(libzfs_handle_t *hdl, avl_tree_t *avl,
uint64_t guid1, uint64_t guid2)
{
nvlist_t *nvfs;
char *fsname, *snapname;
char buf[ZFS_MAX_DATASET_NAME_LEN];
int rv;
zfs_handle_t *guid1hdl, *guid2hdl;
uint64_t create1, create2;
if (guid2 == 0)
return (0);
if (guid1 == 0)
return (1);
nvfs = fsavl_find(avl, guid1, &snapname);
VERIFY(0 == nvlist_lookup_string(nvfs, "name", &fsname));
(void) snprintf(buf, sizeof (buf), "%s@%s", fsname, snapname);
guid1hdl = zfs_open(hdl, buf, ZFS_TYPE_SNAPSHOT);
if (guid1hdl == NULL)
return (-1);
nvfs = fsavl_find(avl, guid2, &snapname);
VERIFY(0 == nvlist_lookup_string(nvfs, "name", &fsname));
(void) snprintf(buf, sizeof (buf), "%s@%s", fsname, snapname);
guid2hdl = zfs_open(hdl, buf, ZFS_TYPE_SNAPSHOT);
if (guid2hdl == NULL) {
zfs_close(guid1hdl);
return (-1);
}
create1 = zfs_prop_get_int(guid1hdl, ZFS_PROP_CREATETXG);
create2 = zfs_prop_get_int(guid2hdl, ZFS_PROP_CREATETXG);
if (create1 < create2)
rv = -1;
else if (create1 > create2)
rv = +1;
else
rv = 0;
zfs_close(guid1hdl);
zfs_close(guid2hdl);
return (rv);
}
static int
recv_incremental_replication(libzfs_handle_t *hdl, const char *tofs,
recvflags_t *flags, nvlist_t *stream_nv, avl_tree_t *stream_avl,
nvlist_t *renamed)
{
nvlist_t *local_nv, *deleted = NULL;
avl_tree_t *local_avl;
nvpair_t *fselem, *nextfselem;
char *fromsnap;
char newname[ZFS_MAX_DATASET_NAME_LEN];
char guidname[32];
int error;
boolean_t needagain, progress, recursive;
char *s1, *s2;
VERIFY(0 == nvlist_lookup_string(stream_nv, "fromsnap", &fromsnap));
recursive = (nvlist_lookup_boolean(stream_nv, "not_recursive") ==
ENOENT);
if (flags->dryrun)
return (0);
again:
needagain = progress = B_FALSE;
VERIFY(0 == nvlist_alloc(&deleted, NV_UNIQUE_NAME, 0));
if ((error = gather_nvlist(hdl, tofs, fromsnap, NULL,
recursive, B_FALSE, B_FALSE, &local_nv, &local_avl)) != 0)
return (error);
/*
* Process deletes and renames
*/
for (fselem = nvlist_next_nvpair(local_nv, NULL);
fselem; fselem = nextfselem) {
nvlist_t *nvfs, *snaps;
nvlist_t *stream_nvfs = NULL;
nvpair_t *snapelem, *nextsnapelem;
uint64_t fromguid = 0;
uint64_t originguid = 0;
uint64_t stream_originguid = 0;
uint64_t parent_fromsnap_guid, stream_parent_fromsnap_guid;
char *fsname, *stream_fsname;
nextfselem = nvlist_next_nvpair(local_nv, fselem);
VERIFY(0 == nvpair_value_nvlist(fselem, &nvfs));
VERIFY(0 == nvlist_lookup_nvlist(nvfs, "snaps", &snaps));
VERIFY(0 == nvlist_lookup_string(nvfs, "name", &fsname));
VERIFY(0 == nvlist_lookup_uint64(nvfs, "parentfromsnap",
&parent_fromsnap_guid));
(void) nvlist_lookup_uint64(nvfs, "origin", &originguid);
/*
* First find the stream's fs, so we can check for
* a different origin (due to "zfs promote")
*/
for (snapelem = nvlist_next_nvpair(snaps, NULL);
snapelem; snapelem = nvlist_next_nvpair(snaps, snapelem)) {
uint64_t thisguid;
VERIFY(0 == nvpair_value_uint64(snapelem, &thisguid));
stream_nvfs = fsavl_find(stream_avl, thisguid, NULL);
if (stream_nvfs != NULL)
break;
}
/* check for promote */
(void) nvlist_lookup_uint64(stream_nvfs, "origin",
&stream_originguid);
if (stream_nvfs && originguid != stream_originguid) {
switch (created_before(hdl, local_avl,
stream_originguid, originguid)) {
case 1: {
/* promote it! */
zfs_cmd_t zc = { 0 };
nvlist_t *origin_nvfs;
char *origin_fsname;
if (flags->verbose)
(void) printf("promoting %s\n", fsname);
origin_nvfs = fsavl_find(local_avl, originguid,
NULL);
VERIFY(0 == nvlist_lookup_string(origin_nvfs,
"name", &origin_fsname));
(void) strlcpy(zc.zc_value, origin_fsname,
sizeof (zc.zc_value));
(void) strlcpy(zc.zc_name, fsname,
sizeof (zc.zc_name));
error = zfs_ioctl(hdl, ZFS_IOC_PROMOTE, &zc);
if (error == 0)
progress = B_TRUE;
break;
}
default:
break;
case -1:
fsavl_destroy(local_avl);
nvlist_free(local_nv);
return (-1);
}
/*
* We had/have the wrong origin, therefore our
* list of snapshots is wrong. Need to handle
* them on the next pass.
*/
needagain = B_TRUE;
continue;
}
for (snapelem = nvlist_next_nvpair(snaps, NULL);
snapelem; snapelem = nextsnapelem) {
uint64_t thisguid;
char *stream_snapname;
nvlist_t *found, *props;
nextsnapelem = nvlist_next_nvpair(snaps, snapelem);
VERIFY(0 == nvpair_value_uint64(snapelem, &thisguid));
found = fsavl_find(stream_avl, thisguid,
&stream_snapname);
/* check for delete */
if (found == NULL) {
char name[ZFS_MAX_DATASET_NAME_LEN];
if (!flags->force)
continue;
(void) snprintf(name, sizeof (name), "%s@%s",
fsname, nvpair_name(snapelem));
error = recv_destroy(hdl, name,
strlen(fsname)+1, newname, flags);
if (error)
needagain = B_TRUE;
else
progress = B_TRUE;
sprintf(guidname, "%" PRIu64, thisguid);
nvlist_add_boolean(deleted, guidname);
continue;
}
stream_nvfs = found;
if (0 == nvlist_lookup_nvlist(stream_nvfs, "snapprops",
&props) && 0 == nvlist_lookup_nvlist(props,
stream_snapname, &props)) {
zfs_cmd_t zc = { 0 };
zc.zc_cookie = B_TRUE; /* received */
(void) snprintf(zc.zc_name, sizeof (zc.zc_name),
"%s@%s", fsname, nvpair_name(snapelem));
if (zcmd_write_src_nvlist(hdl, &zc,
props) == 0) {
(void) zfs_ioctl(hdl,
ZFS_IOC_SET_PROP, &zc);
zcmd_free_nvlists(&zc);
}
}
/* check for different snapname */
if (strcmp(nvpair_name(snapelem),
stream_snapname) != 0) {
char name[ZFS_MAX_DATASET_NAME_LEN];
char tryname[ZFS_MAX_DATASET_NAME_LEN];
(void) snprintf(name, sizeof (name), "%s@%s",
fsname, nvpair_name(snapelem));
(void) snprintf(tryname, sizeof (name), "%s@%s",
fsname, stream_snapname);
error = recv_rename(hdl, name, tryname,
strlen(fsname)+1, newname, flags);
if (error)
needagain = B_TRUE;
else
progress = B_TRUE;
}
if (strcmp(stream_snapname, fromsnap) == 0)
fromguid = thisguid;
}
/* check for delete */
if (stream_nvfs == NULL) {
if (!flags->force)
continue;
error = recv_destroy(hdl, fsname, strlen(tofs)+1,
newname, flags);
if (error)
needagain = B_TRUE;
else
progress = B_TRUE;
sprintf(guidname, "%" PRIu64, parent_fromsnap_guid);
nvlist_add_boolean(deleted, guidname);
continue;
}
if (fromguid == 0) {
if (flags->verbose) {
(void) printf("local fs %s does not have "
"fromsnap (%s in stream); must have "
"been deleted locally; ignoring\n",
fsname, fromsnap);
}
continue;
}
VERIFY(0 == nvlist_lookup_string(stream_nvfs,
"name", &stream_fsname));
VERIFY(0 == nvlist_lookup_uint64(stream_nvfs,
"parentfromsnap", &stream_parent_fromsnap_guid));
s1 = strrchr(fsname, '/');
s2 = strrchr(stream_fsname, '/');
/*
* Check if we're going to rename based on parent guid change
* and the current parent guid was also deleted. If it was then
* rename will fail and is likely unneeded, so avoid this and
* force an early retry to determine the new
* parent_fromsnap_guid.
*/
if (stream_parent_fromsnap_guid != 0 &&
parent_fromsnap_guid != 0 &&
stream_parent_fromsnap_guid != parent_fromsnap_guid) {
sprintf(guidname, "%" PRIu64, parent_fromsnap_guid);
if (nvlist_exists(deleted, guidname)) {
progress = B_TRUE;
needagain = B_TRUE;
goto doagain;
}
}
/*
* Check for rename. If the exact receive path is specified, it
* does not count as a rename, but we still need to check the
* datasets beneath it.
*/
if ((stream_parent_fromsnap_guid != 0 &&
parent_fromsnap_guid != 0 &&
stream_parent_fromsnap_guid != parent_fromsnap_guid) ||
((flags->isprefix || strcmp(tofs, fsname) != 0) &&
(s1 != NULL) && (s2 != NULL) && strcmp(s1, s2) != 0)) {
nvlist_t *parent;
char tryname[ZFS_MAX_DATASET_NAME_LEN];
parent = fsavl_find(local_avl,
stream_parent_fromsnap_guid, NULL);
/*
* NB: parent might not be found if we used the
* tosnap for stream_parent_fromsnap_guid,
* because the parent is a newly-created fs;
* we'll be able to rename it after we recv the
* new fs.
*/
if (parent != NULL) {
char *pname;
VERIFY(0 == nvlist_lookup_string(parent, "name",
&pname));
(void) snprintf(tryname, sizeof (tryname),
"%s%s", pname, strrchr(stream_fsname, '/'));
} else {
tryname[0] = '\0';
if (flags->verbose) {
(void) printf("local fs %s new parent "
"not found\n", fsname);
}
}
newname[0] = '\0';
error = recv_rename(hdl, fsname, tryname,
strlen(tofs)+1, newname, flags);
if (renamed != NULL && newname[0] != '\0') {
VERIFY(0 == nvlist_add_boolean(renamed,
newname));
}
if (error)
needagain = B_TRUE;
else
progress = B_TRUE;
}
}
doagain:
fsavl_destroy(local_avl);
nvlist_free(local_nv);
nvlist_free(deleted);
if (needagain && progress) {
/* do another pass to fix up temporary names */
if (flags->verbose)
(void) printf("another pass:\n");
goto again;
}
return (needagain);
}
static int
zfs_receive_package(libzfs_handle_t *hdl, int fd, const char *destname,
recvflags_t *flags, dmu_replay_record_t *drr, zio_cksum_t *zc,
char **top_zfs, int cleanup_fd, uint64_t *action_handlep)
{
nvlist_t *stream_nv = NULL;
avl_tree_t *stream_avl = NULL;
char *fromsnap = NULL;
char *sendsnap = NULL;
char *cp;
char tofs[ZFS_MAX_DATASET_NAME_LEN];
char sendfs[ZFS_MAX_DATASET_NAME_LEN];
char errbuf[1024];
dmu_replay_record_t drre;
int error;
boolean_t anyerr = B_FALSE;
boolean_t softerr = B_FALSE;
boolean_t recursive;
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot receive"));
assert(drr->drr_type == DRR_BEGIN);
assert(drr->drr_u.drr_begin.drr_magic == DMU_BACKUP_MAGIC);
assert(DMU_GET_STREAM_HDRTYPE(drr->drr_u.drr_begin.drr_versioninfo) ==
DMU_COMPOUNDSTREAM);
/*
* Read in the nvlist from the stream.
*/
if (drr->drr_payloadlen != 0) {
error = recv_read_nvlist(hdl, fd, drr->drr_payloadlen,
&stream_nv, flags->byteswap, zc);
if (error) {
error = zfs_error(hdl, EZFS_BADSTREAM, errbuf);
goto out;
}
}
recursive = (nvlist_lookup_boolean(stream_nv, "not_recursive") ==
ENOENT);
if (recursive && strchr(destname, '@')) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"cannot specify snapshot name for multi-snapshot stream"));
error = zfs_error(hdl, EZFS_BADSTREAM, errbuf);
goto out;
}
/*
* Read in the end record and verify checksum.
*/
if (0 != (error = recv_read(hdl, fd, &drre, sizeof (drre),
flags->byteswap, NULL)))
goto out;
if (flags->byteswap) {
drre.drr_type = BSWAP_32(drre.drr_type);
drre.drr_u.drr_end.drr_checksum.zc_word[0] =
BSWAP_64(drre.drr_u.drr_end.drr_checksum.zc_word[0]);
drre.drr_u.drr_end.drr_checksum.zc_word[1] =
BSWAP_64(drre.drr_u.drr_end.drr_checksum.zc_word[1]);
drre.drr_u.drr_end.drr_checksum.zc_word[2] =
BSWAP_64(drre.drr_u.drr_end.drr_checksum.zc_word[2]);
drre.drr_u.drr_end.drr_checksum.zc_word[3] =
BSWAP_64(drre.drr_u.drr_end.drr_checksum.zc_word[3]);
}
if (drre.drr_type != DRR_END) {
error = zfs_error(hdl, EZFS_BADSTREAM, errbuf);
goto out;
}
if (!ZIO_CHECKSUM_EQUAL(drre.drr_u.drr_end.drr_checksum, *zc)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"incorrect header checksum"));
error = zfs_error(hdl, EZFS_BADSTREAM, errbuf);
goto out;
}
(void) nvlist_lookup_string(stream_nv, "fromsnap", &fromsnap);
if (drr->drr_payloadlen != 0) {
nvlist_t *stream_fss;
VERIFY(0 == nvlist_lookup_nvlist(stream_nv, "fss",
&stream_fss));
if ((stream_avl = fsavl_create(stream_fss)) == NULL) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"couldn't allocate avl tree"));
error = zfs_error(hdl, EZFS_NOMEM, errbuf);
goto out;
}
if (fromsnap != NULL && recursive) {
nvlist_t *renamed = NULL;
nvpair_t *pair = NULL;
(void) strlcpy(tofs, destname, sizeof (tofs));
if (flags->isprefix) {
struct drr_begin *drrb = &drr->drr_u.drr_begin;
int i;
if (flags->istail) {
cp = strrchr(drrb->drr_toname, '/');
if (cp == NULL) {
(void) strlcat(tofs, "/",
sizeof (tofs));
i = 0;
} else {
i = (cp - drrb->drr_toname);
}
} else {
i = strcspn(drrb->drr_toname, "/@");
}
/* zfs_receive_one() will create_parents() */
(void) strlcat(tofs, &drrb->drr_toname[i],
sizeof (tofs));
*strchr(tofs, '@') = '\0';
}
if (!flags->dryrun && !flags->nomount) {
VERIFY(0 == nvlist_alloc(&renamed,
NV_UNIQUE_NAME, 0));
}
softerr = recv_incremental_replication(hdl, tofs, flags,
stream_nv, stream_avl, renamed);
/* Unmount renamed filesystems before receiving. */
while ((pair = nvlist_next_nvpair(renamed,
pair)) != NULL) {
zfs_handle_t *zhp;
prop_changelist_t *clp = NULL;
zhp = zfs_open(hdl, nvpair_name(pair),
ZFS_TYPE_FILESYSTEM);
if (zhp != NULL) {
clp = changelist_gather(zhp,
ZFS_PROP_MOUNTPOINT, 0, 0);
zfs_close(zhp);
if (clp != NULL) {
softerr |=
changelist_prefix(clp);
changelist_free(clp);
}
}
}
nvlist_free(renamed);
}
}
/*
* Get the fs specified by the first path in the stream (the top level
* specified by 'zfs send') and pass it to each invocation of
* zfs_receive_one().
*/
(void) strlcpy(sendfs, drr->drr_u.drr_begin.drr_toname,
sizeof (sendfs));
if ((cp = strchr(sendfs, '@')) != NULL) {
*cp = '\0';
/*
* Find the "sendsnap", the final snapshot in a replication
* stream. zfs_receive_one() handles certain errors
* differently, depending on if the contained stream is the
* last one or not.
*/
sendsnap = (cp + 1);
}
/* Finally, receive each contained stream */
do {
/*
* we should figure out if it has a recoverable
* error, in which case do a recv_skip() and drive on.
* Note, if we fail due to already having this guid,
* zfs_receive_one() will take care of it (ie,
* recv_skip() and return 0).
*/
error = zfs_receive_impl(hdl, destname, NULL, flags, fd,
sendfs, stream_nv, stream_avl, top_zfs, cleanup_fd,
action_handlep, sendsnap);
if (error == ENODATA) {
error = 0;
break;
}
anyerr |= error;
} while (error == 0);
if (drr->drr_payloadlen != 0 && recursive && fromsnap != NULL) {
/*
* Now that we have the fs's they sent us, try the
* renames again.
*/
softerr = recv_incremental_replication(hdl, tofs, flags,
stream_nv, stream_avl, NULL);
}
out:
fsavl_destroy(stream_avl);
nvlist_free(stream_nv);
if (softerr)
error = -2;
if (anyerr)
error = -1;
return (error);
}
static void
trunc_prop_errs(int truncated)
{
ASSERT(truncated != 0);
if (truncated == 1)
(void) fprintf(stderr, dgettext(TEXT_DOMAIN,
"1 more property could not be set\n"));
else
(void) fprintf(stderr, dgettext(TEXT_DOMAIN,
"%d more properties could not be set\n"), truncated);
}
static int
recv_skip(libzfs_handle_t *hdl, int fd, boolean_t byteswap)
{
dmu_replay_record_t *drr;
void *buf = zfs_alloc(hdl, SPA_MAXBLOCKSIZE);
char errbuf[1024];
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot receive:"));
/* XXX would be great to use lseek if possible... */
drr = buf;
while (recv_read(hdl, fd, drr, sizeof (dmu_replay_record_t),
byteswap, NULL) == 0) {
if (byteswap)
drr->drr_type = BSWAP_32(drr->drr_type);
switch (drr->drr_type) {
case DRR_BEGIN:
if (drr->drr_payloadlen != 0) {
(void) recv_read(hdl, fd, buf,
drr->drr_payloadlen, B_FALSE, NULL);
}
break;
case DRR_END:
free(buf);
return (0);
case DRR_OBJECT:
if (byteswap) {
drr->drr_u.drr_object.drr_bonuslen =
BSWAP_32(drr->drr_u.drr_object.
drr_bonuslen);
}
(void) recv_read(hdl, fd, buf,
P2ROUNDUP(drr->drr_u.drr_object.drr_bonuslen, 8),
B_FALSE, NULL);
break;
case DRR_WRITE:
if (byteswap) {
drr->drr_u.drr_write.drr_logical_size =
BSWAP_64(
drr->drr_u.drr_write.drr_logical_size);
drr->drr_u.drr_write.drr_compressed_size =
BSWAP_64(
drr->drr_u.drr_write.drr_compressed_size);
}
uint64_t payload_size =
DRR_WRITE_PAYLOAD_SIZE(&drr->drr_u.drr_write);
(void) recv_read(hdl, fd, buf,
payload_size, B_FALSE, NULL);
break;
case DRR_SPILL:
if (byteswap) {
drr->drr_u.drr_spill.drr_length =
BSWAP_64(drr->drr_u.drr_spill.drr_length);
}
(void) recv_read(hdl, fd, buf,
drr->drr_u.drr_spill.drr_length, B_FALSE, NULL);
break;
case DRR_WRITE_EMBEDDED:
if (byteswap) {
drr->drr_u.drr_write_embedded.drr_psize =
BSWAP_32(drr->drr_u.drr_write_embedded.
drr_psize);
}
(void) recv_read(hdl, fd, buf,
P2ROUNDUP(drr->drr_u.drr_write_embedded.drr_psize,
8), B_FALSE, NULL);
break;
case DRR_WRITE_BYREF:
case DRR_FREEOBJECTS:
case DRR_FREE:
break;
default:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"invalid record type"));
return (zfs_error(hdl, EZFS_BADSTREAM, errbuf));
}
}
free(buf);
return (-1);
}
static void
recv_ecksum_set_aux(libzfs_handle_t *hdl, const char *target_snap,
boolean_t resumable)
{
char target_fs[ZFS_MAX_DATASET_NAME_LEN];
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"checksum mismatch or incomplete stream"));
if (!resumable)
return;
(void) strlcpy(target_fs, target_snap, sizeof (target_fs));
*strchr(target_fs, '@') = '\0';
zfs_handle_t *zhp = zfs_open(hdl, target_fs,
ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME);
if (zhp == NULL)
return;
char token_buf[ZFS_MAXPROPLEN];
int error = zfs_prop_get(zhp, ZFS_PROP_RECEIVE_RESUME_TOKEN,
token_buf, sizeof (token_buf),
NULL, NULL, 0, B_TRUE);
if (error == 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"checksum mismatch or incomplete stream.\n"
"Partially received snapshot is saved.\n"
"A resuming stream can be generated on the sending "
"system by running:\n"
" zfs send -t %s"),
token_buf);
}
zfs_close(zhp);
}
/*
* Restores a backup of tosnap from the file descriptor specified by infd.
*/
static int
zfs_receive_one(libzfs_handle_t *hdl, int infd, const char *tosnap,
const char *originsnap, recvflags_t *flags, dmu_replay_record_t *drr,
dmu_replay_record_t *drr_noswap, const char *sendfs, nvlist_t *stream_nv,
avl_tree_t *stream_avl, char **top_zfs, int cleanup_fd,
uint64_t *action_handlep, const char *finalsnap)
{
zfs_cmd_t zc = { 0 };
time_t begin_time;
int ioctl_err, ioctl_errno, err;
char *cp;
struct drr_begin *drrb = &drr->drr_u.drr_begin;
char errbuf[1024];
char prop_errbuf[1024];
const char *chopprefix;
boolean_t newfs = B_FALSE;
boolean_t stream_wantsnewfs;
uint64_t parent_snapguid = 0;
prop_changelist_t *clp = NULL;
nvlist_t *snapprops_nvlist = NULL;
zprop_errflags_t prop_errflags;
boolean_t recursive;
char *snapname = NULL;
begin_time = time(NULL);
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot receive"));
recursive = (nvlist_lookup_boolean(stream_nv, "not_recursive") ==
ENOENT);
if (stream_avl != NULL) {
nvlist_t *fs = fsavl_find(stream_avl, drrb->drr_toguid,
&snapname);
nvlist_t *props;
int ret;
(void) nvlist_lookup_uint64(fs, "parentfromsnap",
&parent_snapguid);
err = nvlist_lookup_nvlist(fs, "props", &props);
if (err)
VERIFY(0 == nvlist_alloc(&props, NV_UNIQUE_NAME, 0));
if (flags->canmountoff) {
VERIFY(0 == nvlist_add_uint64(props,
zfs_prop_to_name(ZFS_PROP_CANMOUNT), 0));
}
ret = zcmd_write_src_nvlist(hdl, &zc, props);
if (err)
nvlist_free(props);
if (0 == nvlist_lookup_nvlist(fs, "snapprops", &props)) {
VERIFY(0 == nvlist_lookup_nvlist(props,
snapname, &snapprops_nvlist));
}
if (ret != 0)
return (-1);
}
cp = NULL;
/*
* Determine how much of the snapshot name stored in the stream
* we are going to tack on to the name they specified on the
* command line, and how much we are going to chop off.
*
* If they specified a snapshot, chop the entire name stored in
* the stream.
*/
if (flags->istail) {
/*
* A filesystem was specified with -e. We want to tack on only
* the tail of the sent snapshot path.
*/
if (strchr(tosnap, '@')) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid "
"argument - snapshot not allowed with -e"));
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
}
chopprefix = strrchr(sendfs, '/');
if (chopprefix == NULL) {
/*
* The tail is the poolname, so we need to
* prepend a path separator.
*/
int len = strlen(drrb->drr_toname);
cp = malloc(len + 2);
cp[0] = '/';
(void) strcpy(&cp[1], drrb->drr_toname);
chopprefix = cp;
} else {
chopprefix = drrb->drr_toname + (chopprefix - sendfs);
}
} else if (flags->isprefix) {
/*
* A filesystem was specified with -d. We want to tack on
* everything but the first element of the sent snapshot path
* (all but the pool name).
*/
if (strchr(tosnap, '@')) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid "
"argument - snapshot not allowed with -d"));
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
}
chopprefix = strchr(drrb->drr_toname, '/');
if (chopprefix == NULL)
chopprefix = strchr(drrb->drr_toname, '@');
} else if (strchr(tosnap, '@') == NULL) {
/*
* If a filesystem was specified without -d or -e, we want to
* tack on everything after the fs specified by 'zfs send'.
*/
chopprefix = drrb->drr_toname + strlen(sendfs);
} else {
/* A snapshot was specified as an exact path (no -d or -e). */
if (recursive) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"cannot specify snapshot name for multi-snapshot "
"stream"));
return (zfs_error(hdl, EZFS_BADSTREAM, errbuf));
}
chopprefix = drrb->drr_toname + strlen(drrb->drr_toname);
}
ASSERT(strstr(drrb->drr_toname, sendfs) == drrb->drr_toname);
ASSERT(chopprefix > drrb->drr_toname);
ASSERT(chopprefix <= drrb->drr_toname + strlen(drrb->drr_toname));
ASSERT(chopprefix[0] == '/' || chopprefix[0] == '@' ||
chopprefix[0] == '\0');
/*
* Determine name of destination snapshot, store in zc_value.
*/
(void) strcpy(zc.zc_value, tosnap);
(void) strncat(zc.zc_value, chopprefix, sizeof (zc.zc_value));
#ifdef __FreeBSD__
if (zfs_ioctl_version == ZFS_IOCVER_UNDEF)
zfs_ioctl_version = get_zfs_ioctl_version();
/*
* For forward compatibility hide tosnap in zc_value
*/
if (zfs_ioctl_version < ZFS_IOCVER_LZC)
(void) strcpy(zc.zc_value + strlen(zc.zc_value) + 1, tosnap);
#endif
free(cp);
if (!zfs_name_valid(zc.zc_value, ZFS_TYPE_SNAPSHOT)) {
zcmd_free_nvlists(&zc);
return (zfs_error(hdl, EZFS_INVALIDNAME, errbuf));
}
/*
* Determine the name of the origin snapshot, store in zc_string.
*/
if (originsnap) {
(void) strncpy(zc.zc_string, originsnap, sizeof (zc.zc_string));
if (flags->verbose)
(void) printf("using provided clone origin %s\n",
zc.zc_string);
} else if (drrb->drr_flags & DRR_FLAG_CLONE) {
if (guid_to_name(hdl, zc.zc_value,
drrb->drr_fromguid, B_FALSE, zc.zc_string) != 0) {
zcmd_free_nvlists(&zc);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"local origin for clone %s does not exist"),
zc.zc_value);
return (zfs_error(hdl, EZFS_NOENT, errbuf));
}
if (flags->verbose)
(void) printf("found clone origin %s\n", zc.zc_string);
}
boolean_t resuming = DMU_GET_FEATUREFLAGS(drrb->drr_versioninfo) &
DMU_BACKUP_FEATURE_RESUMING;
stream_wantsnewfs = (drrb->drr_fromguid == 0 ||
(drrb->drr_flags & DRR_FLAG_CLONE) || originsnap) && !resuming;
if (stream_wantsnewfs) {
/*
* if the parent fs does not exist, look for it based on
* the parent snap GUID
*/
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot receive new filesystem stream"));
(void) strcpy(zc.zc_name, zc.zc_value);
cp = strrchr(zc.zc_name, '/');
if (cp)
*cp = '\0';
if (cp &&
!zfs_dataset_exists(hdl, zc.zc_name, ZFS_TYPE_DATASET)) {
char suffix[ZFS_MAX_DATASET_NAME_LEN];
(void) strcpy(suffix, strrchr(zc.zc_value, '/'));
if (guid_to_name(hdl, zc.zc_name, parent_snapguid,
B_FALSE, zc.zc_value) == 0) {
*strchr(zc.zc_value, '@') = '\0';
(void) strcat(zc.zc_value, suffix);
}
}
} else {
/*
* If the fs does not exist, look for it based on the
* fromsnap GUID.
*/
if (resuming) {
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN,
"cannot receive resume stream"));
} else {
(void) snprintf(errbuf, sizeof (errbuf),
dgettext(TEXT_DOMAIN,
"cannot receive incremental stream"));
}
(void) strcpy(zc.zc_name, zc.zc_value);
*strchr(zc.zc_name, '@') = '\0';
/*
* If the exact receive path was specified and this is the
* topmost path in the stream, then if the fs does not exist we
* should look no further.
*/
if ((flags->isprefix || (*(chopprefix = drrb->drr_toname +
strlen(sendfs)) != '\0' && *chopprefix != '@')) &&
!zfs_dataset_exists(hdl, zc.zc_name, ZFS_TYPE_DATASET)) {
char snap[ZFS_MAX_DATASET_NAME_LEN];
(void) strcpy(snap, strchr(zc.zc_value, '@'));
if (guid_to_name(hdl, zc.zc_name, drrb->drr_fromguid,
B_FALSE, zc.zc_value) == 0) {
*strchr(zc.zc_value, '@') = '\0';
(void) strcat(zc.zc_value, snap);
}
}
}
(void) strcpy(zc.zc_name, zc.zc_value);
*strchr(zc.zc_name, '@') = '\0';
if (zfs_dataset_exists(hdl, zc.zc_name, ZFS_TYPE_DATASET)) {
zfs_handle_t *zhp;
/*
* Destination fs exists. It must be one of these cases:
* - an incremental send stream
* - the stream specifies a new fs (full stream or clone)
* and they want us to blow away the existing fs (and
* have therefore specified -F and removed any snapshots)
* - we are resuming a failed receive.
*/
if (stream_wantsnewfs) {
+ boolean_t is_volume = drrb->drr_type == DMU_OST_ZVOL;
if (!flags->force) {
zcmd_free_nvlists(&zc);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"destination '%s' exists\n"
"must specify -F to overwrite it"),
zc.zc_name);
return (zfs_error(hdl, EZFS_EXISTS, errbuf));
}
if (ioctl(hdl->libzfs_fd, ZFS_IOC_SNAPSHOT_LIST_NEXT,
&zc) == 0) {
zcmd_free_nvlists(&zc);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"destination has snapshots (eg. %s)\n"
"must destroy them to overwrite it"),
zc.zc_name);
return (zfs_error(hdl, EZFS_EXISTS, errbuf));
}
+ if (is_volume && strrchr(zc.zc_name, '/') == NULL) {
+ zcmd_free_nvlists(&zc);
+ zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
+ "destination '%s' is the root dataset\n"
+ "cannot overwrite with a ZVOL"),
+ zc.zc_name);
+ return (zfs_error(hdl, EZFS_EXISTS, errbuf));
+ }
+ if (is_volume &&
+ ioctl(hdl->libzfs_fd, ZFS_IOC_DATASET_LIST_NEXT,
+ &zc) == 0) {
+ zcmd_free_nvlists(&zc);
+ zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
+ "destination has children (eg. %s)\n"
+ "cannot overwrite with a ZVOL"),
+ zc.zc_name);
+ return (zfs_error(hdl, EZFS_WRONG_PARENT,
+ errbuf));
+ }
}
if ((zhp = zfs_open(hdl, zc.zc_name,
ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME)) == NULL) {
zcmd_free_nvlists(&zc);
return (-1);
}
if (stream_wantsnewfs &&
zhp->zfs_dmustats.dds_origin[0]) {
zcmd_free_nvlists(&zc);
zfs_close(zhp);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"destination '%s' is a clone\n"
"must destroy it to overwrite it"),
zc.zc_name);
return (zfs_error(hdl, EZFS_EXISTS, errbuf));
}
if (!flags->dryrun && zhp->zfs_type == ZFS_TYPE_FILESYSTEM &&
stream_wantsnewfs) {
/* We can't do online recv in this case */
clp = changelist_gather(zhp, ZFS_PROP_NAME, 0, 0);
if (clp == NULL) {
zfs_close(zhp);
zcmd_free_nvlists(&zc);
return (-1);
}
if (changelist_prefix(clp) != 0) {
changelist_free(clp);
zfs_close(zhp);
zcmd_free_nvlists(&zc);
return (-1);
}
}
/*
* If we are resuming a newfs, set newfs here so that we will
* mount it if the recv succeeds this time. We can tell
* that it was a newfs on the first recv because the fs
* itself will be inconsistent (if the fs existed when we
* did the first recv, we would have received it into
* .../%recv).
*/
if (resuming && zfs_prop_get_int(zhp, ZFS_PROP_INCONSISTENT))
newfs = B_TRUE;
zfs_close(zhp);
} else {
+ zfs_handle_t *zhp;
+
/*
* Destination filesystem does not exist. Therefore we better
* be creating a new filesystem (either from a full backup, or
* a clone). It would therefore be invalid if the user
* specified only the pool name (i.e. if the destination name
* contained no slash character).
*/
if (!stream_wantsnewfs ||
(cp = strrchr(zc.zc_name, '/')) == NULL) {
zcmd_free_nvlists(&zc);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"destination '%s' does not exist"), zc.zc_name);
return (zfs_error(hdl, EZFS_NOENT, errbuf));
}
/*
* Trim off the final dataset component so we perform the
* recvbackup ioctl to the filesystems's parent.
*/
*cp = '\0';
if (flags->isprefix && !flags->istail && !flags->dryrun &&
create_parents(hdl, zc.zc_value, strlen(tosnap)) != 0) {
zcmd_free_nvlists(&zc);
return (zfs_error(hdl, EZFS_BADRESTORE, errbuf));
}
+
+ /* validate parent */
+ zhp = zfs_open(hdl, zc.zc_name, ZFS_TYPE_DATASET);
+ if (zhp == NULL) {
+ zcmd_free_nvlists(&zc);
+ return (zfs_error(hdl, EZFS_BADRESTORE, errbuf));
+ }
+ if (zfs_get_type(zhp) != ZFS_TYPE_FILESYSTEM) {
+ zcmd_free_nvlists(&zc);
+ zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
+ "parent '%s' is not a filesystem"), zc.zc_name);
+ zfs_close(zhp);
+ return (zfs_error(hdl, EZFS_WRONG_PARENT, errbuf));
+ }
+ zfs_close(zhp);
newfs = B_TRUE;
}
zc.zc_begin_record = *drr_noswap;
zc.zc_cookie = infd;
zc.zc_guid = flags->force;
zc.zc_resumable = flags->resumable;
if (flags->verbose) {
(void) printf("%s %s stream of %s into %s\n",
flags->dryrun ? "would receive" : "receiving",
drrb->drr_fromguid ? "incremental" : "full",
drrb->drr_toname, zc.zc_value);
(void) fflush(stdout);
}
if (flags->dryrun) {
zcmd_free_nvlists(&zc);
return (recv_skip(hdl, infd, flags->byteswap));
}
zc.zc_nvlist_dst = (uint64_t)(uintptr_t)prop_errbuf;
zc.zc_nvlist_dst_size = sizeof (prop_errbuf);
zc.zc_cleanup_fd = cleanup_fd;
zc.zc_action_handle = *action_handlep;
err = ioctl_err = zfs_ioctl(hdl, ZFS_IOC_RECV, &zc);
ioctl_errno = errno;
prop_errflags = (zprop_errflags_t)zc.zc_obj;
if (err == 0) {
nvlist_t *prop_errors;
VERIFY(0 == nvlist_unpack((void *)(uintptr_t)zc.zc_nvlist_dst,
zc.zc_nvlist_dst_size, &prop_errors, 0));
nvpair_t *prop_err = NULL;
while ((prop_err = nvlist_next_nvpair(prop_errors,
prop_err)) != NULL) {
char tbuf[1024];
zfs_prop_t prop;
int intval;
prop = zfs_name_to_prop(nvpair_name(prop_err));
(void) nvpair_value_int32(prop_err, &intval);
if (strcmp(nvpair_name(prop_err),
ZPROP_N_MORE_ERRORS) == 0) {
trunc_prop_errs(intval);
break;
} else if (snapname == NULL || finalsnap == NULL ||
strcmp(finalsnap, snapname) == 0 ||
strcmp(nvpair_name(prop_err),
zfs_prop_to_name(ZFS_PROP_REFQUOTA)) != 0) {
/*
* Skip the special case of, for example,
* "refquota", errors on intermediate
* snapshots leading up to a final one.
* That's why we have all of the checks above.
*
* See zfs_ioctl.c's extract_delay_props() for
* a list of props which can fail on
* intermediate snapshots, but shouldn't
* affect the overall receive.
*/
(void) snprintf(tbuf, sizeof (tbuf),
dgettext(TEXT_DOMAIN,
"cannot receive %s property on %s"),
nvpair_name(prop_err), zc.zc_name);
zfs_setprop_error(hdl, prop, intval, tbuf);
}
}
nvlist_free(prop_errors);
}
zc.zc_nvlist_dst = 0;
zc.zc_nvlist_dst_size = 0;
zcmd_free_nvlists(&zc);
if (err == 0 && snapprops_nvlist) {
zfs_cmd_t zc2 = { 0 };
(void) strcpy(zc2.zc_name, zc.zc_value);
zc2.zc_cookie = B_TRUE; /* received */
if (zcmd_write_src_nvlist(hdl, &zc2, snapprops_nvlist) == 0) {
(void) zfs_ioctl(hdl, ZFS_IOC_SET_PROP, &zc2);
zcmd_free_nvlists(&zc2);
}
}
if (err && (ioctl_errno == ENOENT || ioctl_errno == EEXIST)) {
/*
* It may be that this snapshot already exists,
* in which case we want to consume & ignore it
* rather than failing.
*/
avl_tree_t *local_avl;
nvlist_t *local_nv, *fs;
cp = strchr(zc.zc_value, '@');
/*
* XXX Do this faster by just iterating over snaps in
* this fs. Also if zc_value does not exist, we will
* get a strange "does not exist" error message.
*/
*cp = '\0';
if (gather_nvlist(hdl, zc.zc_value, NULL, NULL, B_FALSE,
B_FALSE, B_FALSE, &local_nv, &local_avl) == 0) {
*cp = '@';
fs = fsavl_find(local_avl, drrb->drr_toguid, NULL);
fsavl_destroy(local_avl);
nvlist_free(local_nv);
if (fs != NULL) {
if (flags->verbose) {
(void) printf("snap %s already exists; "
"ignoring\n", zc.zc_value);
}
err = ioctl_err = recv_skip(hdl, infd,
flags->byteswap);
}
}
*cp = '@';
}
if (ioctl_err != 0) {
switch (ioctl_errno) {
case ENODEV:
cp = strchr(zc.zc_value, '@');
*cp = '\0';
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"most recent snapshot of %s does not\n"
"match incremental source"), zc.zc_value);
(void) zfs_error(hdl, EZFS_BADRESTORE, errbuf);
*cp = '@';
break;
case ETXTBSY:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"destination %s has been modified\n"
"since most recent snapshot"), zc.zc_name);
(void) zfs_error(hdl, EZFS_BADRESTORE, errbuf);
break;
case EEXIST:
cp = strchr(zc.zc_value, '@');
if (newfs) {
/* it's the containing fs that exists */
*cp = '\0';
}
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"destination already exists"));
(void) zfs_error_fmt(hdl, EZFS_EXISTS,
dgettext(TEXT_DOMAIN, "cannot restore to %s"),
zc.zc_value);
*cp = '@';
break;
case EINVAL:
(void) zfs_error(hdl, EZFS_BADSTREAM, errbuf);
break;
case ECKSUM:
recv_ecksum_set_aux(hdl, zc.zc_value, flags->resumable);
(void) zfs_error(hdl, EZFS_BADSTREAM, errbuf);
break;
case ENOTSUP:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool must be upgraded to receive this stream."));
(void) zfs_error(hdl, EZFS_BADVERSION, errbuf);
break;
case EDQUOT:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"destination %s space quota exceeded"), zc.zc_name);
(void) zfs_error(hdl, EZFS_NOSPC, errbuf);
break;
default:
(void) zfs_standard_error(hdl, ioctl_errno, errbuf);
}
}
/*
* Mount the target filesystem (if created). Also mount any
* children of the target filesystem if we did a replication
* receive (indicated by stream_avl being non-NULL).
*/
cp = strchr(zc.zc_value, '@');
if (cp && (ioctl_err == 0 || !newfs)) {
zfs_handle_t *h;
*cp = '\0';
h = zfs_open(hdl, zc.zc_value,
ZFS_TYPE_FILESYSTEM | ZFS_TYPE_VOLUME);
if (h != NULL) {
if (h->zfs_type == ZFS_TYPE_VOLUME) {
*cp = '@';
} else if (newfs || stream_avl) {
/*
* Track the first/top of hierarchy fs,
* for mounting and sharing later.
*/
if (top_zfs && *top_zfs == NULL)
*top_zfs = zfs_strdup(hdl, zc.zc_value);
}
zfs_close(h);
}
*cp = '@';
}
if (clp) {
if (!flags->nomount)
err |= changelist_postfix(clp);
changelist_free(clp);
}
if (prop_errflags & ZPROP_ERR_NOCLEAR) {
(void) fprintf(stderr, dgettext(TEXT_DOMAIN, "Warning: "
"failed to clear unreceived properties on %s"),
zc.zc_name);
(void) fprintf(stderr, "\n");
}
if (prop_errflags & ZPROP_ERR_NORESTORE) {
(void) fprintf(stderr, dgettext(TEXT_DOMAIN, "Warning: "
"failed to restore original properties on %s"),
zc.zc_name);
(void) fprintf(stderr, "\n");
}
if (err || ioctl_err)
return (-1);
*action_handlep = zc.zc_action_handle;
if (flags->verbose) {
char buf1[64];
char buf2[64];
uint64_t bytes = zc.zc_cookie;
time_t delta = time(NULL) - begin_time;
if (delta == 0)
delta = 1;
zfs_nicenum(bytes, buf1, sizeof (buf1));
zfs_nicenum(bytes/delta, buf2, sizeof (buf1));
(void) printf("received %sB stream in %lu seconds (%sB/sec)\n",
buf1, delta, buf2);
}
return (0);
}
static int
zfs_receive_impl(libzfs_handle_t *hdl, const char *tosnap,
const char *originsnap, recvflags_t *flags, int infd, const char *sendfs,
nvlist_t *stream_nv, avl_tree_t *stream_avl, char **top_zfs, int cleanup_fd,
uint64_t *action_handlep, const char *finalsnap)
{
int err;
dmu_replay_record_t drr, drr_noswap;
struct drr_begin *drrb = &drr.drr_u.drr_begin;
char errbuf[1024];
zio_cksum_t zcksum = { 0 };
uint64_t featureflags;
int hdrtype;
(void) snprintf(errbuf, sizeof (errbuf), dgettext(TEXT_DOMAIN,
"cannot receive"));
if (flags->isprefix &&
!zfs_dataset_exists(hdl, tosnap, ZFS_TYPE_DATASET)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "specified fs "
"(%s) does not exist"), tosnap);
return (zfs_error(hdl, EZFS_NOENT, errbuf));
}
if (originsnap &&
!zfs_dataset_exists(hdl, originsnap, ZFS_TYPE_DATASET)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "specified origin fs "
"(%s) does not exist"), originsnap);
return (zfs_error(hdl, EZFS_NOENT, errbuf));
}
/* read in the BEGIN record */
if (0 != (err = recv_read(hdl, infd, &drr, sizeof (drr), B_FALSE,
&zcksum)))
return (err);
if (drr.drr_type == DRR_END || drr.drr_type == BSWAP_32(DRR_END)) {
/* It's the double end record at the end of a package */
return (ENODATA);
}
/* the kernel needs the non-byteswapped begin record */
drr_noswap = drr;
flags->byteswap = B_FALSE;
if (drrb->drr_magic == BSWAP_64(DMU_BACKUP_MAGIC)) {
/*
* We computed the checksum in the wrong byteorder in
* recv_read() above; do it again correctly.
*/
bzero(&zcksum, sizeof (zio_cksum_t));
(void) fletcher_4_incremental_byteswap(&drr,
sizeof (drr), &zcksum);
flags->byteswap = B_TRUE;
drr.drr_type = BSWAP_32(drr.drr_type);
drr.drr_payloadlen = BSWAP_32(drr.drr_payloadlen);
drrb->drr_magic = BSWAP_64(drrb->drr_magic);
drrb->drr_versioninfo = BSWAP_64(drrb->drr_versioninfo);
drrb->drr_creation_time = BSWAP_64(drrb->drr_creation_time);
drrb->drr_type = BSWAP_32(drrb->drr_type);
drrb->drr_flags = BSWAP_32(drrb->drr_flags);
drrb->drr_toguid = BSWAP_64(drrb->drr_toguid);
drrb->drr_fromguid = BSWAP_64(drrb->drr_fromguid);
}
if (drrb->drr_magic != DMU_BACKUP_MAGIC || drr.drr_type != DRR_BEGIN) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid "
"stream (bad magic number)"));
return (zfs_error(hdl, EZFS_BADSTREAM, errbuf));
}
featureflags = DMU_GET_FEATUREFLAGS(drrb->drr_versioninfo);
hdrtype = DMU_GET_STREAM_HDRTYPE(drrb->drr_versioninfo);
if (!DMU_STREAM_SUPPORTED(featureflags) ||
(hdrtype != DMU_SUBSTREAM && hdrtype != DMU_COMPOUNDSTREAM)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"stream has unsupported feature, feature flags = %lx"),
featureflags);
return (zfs_error(hdl, EZFS_BADSTREAM, errbuf));
}
if (strchr(drrb->drr_toname, '@') == NULL) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "invalid "
"stream (bad snapshot name)"));
return (zfs_error(hdl, EZFS_BADSTREAM, errbuf));
}
if (DMU_GET_STREAM_HDRTYPE(drrb->drr_versioninfo) == DMU_SUBSTREAM) {
char nonpackage_sendfs[ZFS_MAX_DATASET_NAME_LEN];
if (sendfs == NULL) {
/*
* We were not called from zfs_receive_package(). Get
* the fs specified by 'zfs send'.
*/
char *cp;
(void) strlcpy(nonpackage_sendfs,
drr.drr_u.drr_begin.drr_toname,
sizeof (nonpackage_sendfs));
if ((cp = strchr(nonpackage_sendfs, '@')) != NULL)
*cp = '\0';
sendfs = nonpackage_sendfs;
VERIFY(finalsnap == NULL);
}
return (zfs_receive_one(hdl, infd, tosnap, originsnap, flags,
&drr, &drr_noswap, sendfs, stream_nv, stream_avl, top_zfs,
cleanup_fd, action_handlep, finalsnap));
} else {
assert(DMU_GET_STREAM_HDRTYPE(drrb->drr_versioninfo) ==
DMU_COMPOUNDSTREAM);
return (zfs_receive_package(hdl, infd, tosnap, flags, &drr,
&zcksum, top_zfs, cleanup_fd, action_handlep));
}
}
/*
* Restores a backup of tosnap from the file descriptor specified by infd.
* Return 0 on total success, -2 if some things couldn't be
* destroyed/renamed/promoted, -1 if some things couldn't be received.
* (-1 will override -2, if -1 and the resumable flag was specified the
* transfer can be resumed if the sending side supports it).
*/
int
zfs_receive(libzfs_handle_t *hdl, const char *tosnap, nvlist_t *props,
recvflags_t *flags, int infd, avl_tree_t *stream_avl)
{
char *top_zfs = NULL;
int err;
int cleanup_fd;
uint64_t action_handle = 0;
char *originsnap = NULL;
if (props) {
err = nvlist_lookup_string(props, "origin", &originsnap);
if (err && err != ENOENT)
return (err);
}
cleanup_fd = open(ZFS_DEV, O_RDWR|O_EXCL);
VERIFY(cleanup_fd >= 0);
err = zfs_receive_impl(hdl, tosnap, originsnap, flags, infd, NULL, NULL,
stream_avl, &top_zfs, cleanup_fd, &action_handle, NULL);
VERIFY(0 == close(cleanup_fd));
if (err == 0 && !flags->nomount && top_zfs) {
zfs_handle_t *zhp;
prop_changelist_t *clp;
zhp = zfs_open(hdl, top_zfs, ZFS_TYPE_FILESYSTEM);
if (zhp != NULL) {
clp = changelist_gather(zhp, ZFS_PROP_MOUNTPOINT,
CL_GATHER_MOUNT_ALWAYS, 0);
zfs_close(zhp);
if (clp != NULL) {
/* mount and share received datasets */
err = changelist_postfix(clp);
changelist_free(clp);
}
}
if (zhp == NULL || clp == NULL || err)
err = -1;
}
if (top_zfs)
free(top_zfs);
return (err);
}
Index: head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_util.c
===================================================================
--- head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_util.c (revision 359302)
+++ head/cddl/contrib/opensolaris/lib/libzfs/common/libzfs_util.c (revision 359303)
@@ -1,1621 +1,1626 @@
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright 2019 Joyent, Inc.
* Copyright (c) 2011, 2017 by Delphix. All rights reserved.
* Copyright 2016 Igor Kozhukhov <ikozhukhov@gmail.com>
* Copyright (c) 2017 Datto Inc.
*/
/*
* Internal utility routines for the ZFS library.
*/
#include <sys/param.h>
#include <sys/linker.h>
#include <sys/module.h>
#include <sys/stat.h>
#include <errno.h>
#include <fcntl.h>
#include <libintl.h>
#include <stdarg.h>
#include <stdio.h>
#include <stdlib.h>
#include <strings.h>
#include <unistd.h>
#include <ctype.h>
#include <math.h>
#include <sys/mnttab.h>
#include <sys/mntent.h>
#include <sys/types.h>
#include <libcmdutils.h>
#include <libzfs.h>
#include <libzfs_core.h>
#include "libzfs_impl.h"
#include "zfs_prop.h"
#include "zfs_comutil.h"
#include "zfeature_common.h"
int
libzfs_errno(libzfs_handle_t *hdl)
{
return (hdl->libzfs_error);
}
const char *
libzfs_error_action(libzfs_handle_t *hdl)
{
return (hdl->libzfs_action);
}
const char *
libzfs_error_description(libzfs_handle_t *hdl)
{
if (hdl->libzfs_desc[0] != '\0')
return (hdl->libzfs_desc);
switch (hdl->libzfs_error) {
case EZFS_NOMEM:
return (dgettext(TEXT_DOMAIN, "out of memory"));
case EZFS_BADPROP:
return (dgettext(TEXT_DOMAIN, "invalid property value"));
case EZFS_PROPREADONLY:
return (dgettext(TEXT_DOMAIN, "read-only property"));
case EZFS_PROPTYPE:
return (dgettext(TEXT_DOMAIN, "property doesn't apply to "
"datasets of this type"));
case EZFS_PROPNONINHERIT:
return (dgettext(TEXT_DOMAIN, "property cannot be inherited"));
case EZFS_PROPSPACE:
return (dgettext(TEXT_DOMAIN, "invalid quota or reservation"));
case EZFS_BADTYPE:
return (dgettext(TEXT_DOMAIN, "operation not applicable to "
"datasets of this type"));
case EZFS_BUSY:
return (dgettext(TEXT_DOMAIN, "pool or dataset is busy"));
case EZFS_EXISTS:
return (dgettext(TEXT_DOMAIN, "pool or dataset exists"));
case EZFS_NOENT:
return (dgettext(TEXT_DOMAIN, "no such pool or dataset"));
case EZFS_BADSTREAM:
return (dgettext(TEXT_DOMAIN, "invalid backup stream"));
case EZFS_DSREADONLY:
return (dgettext(TEXT_DOMAIN, "dataset is read-only"));
case EZFS_VOLTOOBIG:
return (dgettext(TEXT_DOMAIN, "volume size exceeds limit for "
"this system"));
case EZFS_INVALIDNAME:
return (dgettext(TEXT_DOMAIN, "invalid name"));
case EZFS_BADRESTORE:
return (dgettext(TEXT_DOMAIN, "unable to restore to "
"destination"));
case EZFS_BADBACKUP:
return (dgettext(TEXT_DOMAIN, "backup failed"));
case EZFS_BADTARGET:
return (dgettext(TEXT_DOMAIN, "invalid target vdev"));
case EZFS_NODEVICE:
return (dgettext(TEXT_DOMAIN, "no such device in pool"));
case EZFS_BADDEV:
return (dgettext(TEXT_DOMAIN, "invalid device"));
case EZFS_NOREPLICAS:
return (dgettext(TEXT_DOMAIN, "no valid replicas"));
case EZFS_RESILVERING:
return (dgettext(TEXT_DOMAIN, "currently resilvering"));
case EZFS_BADVERSION:
return (dgettext(TEXT_DOMAIN, "unsupported version or "
"feature"));
case EZFS_POOLUNAVAIL:
return (dgettext(TEXT_DOMAIN, "pool is unavailable"));
case EZFS_DEVOVERFLOW:
return (dgettext(TEXT_DOMAIN, "too many devices in one vdev"));
case EZFS_BADPATH:
return (dgettext(TEXT_DOMAIN, "must be an absolute path"));
case EZFS_CROSSTARGET:
return (dgettext(TEXT_DOMAIN, "operation crosses datasets or "
"pools"));
case EZFS_ZONED:
return (dgettext(TEXT_DOMAIN, "dataset in use by local zone"));
case EZFS_MOUNTFAILED:
return (dgettext(TEXT_DOMAIN, "mount failed"));
case EZFS_UMOUNTFAILED:
return (dgettext(TEXT_DOMAIN, "umount failed"));
case EZFS_UNSHARENFSFAILED:
return (dgettext(TEXT_DOMAIN, "unshare(1M) failed"));
case EZFS_SHARENFSFAILED:
return (dgettext(TEXT_DOMAIN, "share(1M) failed"));
case EZFS_UNSHARESMBFAILED:
return (dgettext(TEXT_DOMAIN, "smb remove share failed"));
case EZFS_SHARESMBFAILED:
return (dgettext(TEXT_DOMAIN, "smb add share failed"));
case EZFS_PERM:
return (dgettext(TEXT_DOMAIN, "permission denied"));
case EZFS_NOSPC:
return (dgettext(TEXT_DOMAIN, "out of space"));
case EZFS_FAULT:
return (dgettext(TEXT_DOMAIN, "bad address"));
case EZFS_IO:
return (dgettext(TEXT_DOMAIN, "I/O error"));
case EZFS_INTR:
return (dgettext(TEXT_DOMAIN, "signal received"));
case EZFS_ISSPARE:
return (dgettext(TEXT_DOMAIN, "device is reserved as a hot "
"spare"));
case EZFS_INVALCONFIG:
return (dgettext(TEXT_DOMAIN, "invalid vdev configuration"));
case EZFS_RECURSIVE:
return (dgettext(TEXT_DOMAIN, "recursive dataset dependency"));
case EZFS_NOHISTORY:
return (dgettext(TEXT_DOMAIN, "no history available"));
case EZFS_POOLPROPS:
return (dgettext(TEXT_DOMAIN, "failed to retrieve "
"pool properties"));
case EZFS_POOL_NOTSUP:
return (dgettext(TEXT_DOMAIN, "operation not supported "
"on this type of pool"));
case EZFS_POOL_INVALARG:
return (dgettext(TEXT_DOMAIN, "invalid argument for "
"this pool operation"));
case EZFS_NAMETOOLONG:
return (dgettext(TEXT_DOMAIN, "dataset name is too long"));
case EZFS_OPENFAILED:
return (dgettext(TEXT_DOMAIN, "open failed"));
case EZFS_NOCAP:
return (dgettext(TEXT_DOMAIN,
"disk capacity information could not be retrieved"));
case EZFS_LABELFAILED:
return (dgettext(TEXT_DOMAIN, "write of label failed"));
case EZFS_BADWHO:
return (dgettext(TEXT_DOMAIN, "invalid user/group"));
case EZFS_BADPERM:
return (dgettext(TEXT_DOMAIN, "invalid permission"));
case EZFS_BADPERMSET:
return (dgettext(TEXT_DOMAIN, "invalid permission set name"));
case EZFS_NODELEGATION:
return (dgettext(TEXT_DOMAIN, "delegated administration is "
"disabled on pool"));
case EZFS_BADCACHE:
return (dgettext(TEXT_DOMAIN, "invalid or missing cache file"));
case EZFS_ISL2CACHE:
return (dgettext(TEXT_DOMAIN, "device is in use as a cache"));
case EZFS_VDEVNOTSUP:
return (dgettext(TEXT_DOMAIN, "vdev specification is not "
"supported"));
case EZFS_NOTSUP:
return (dgettext(TEXT_DOMAIN, "operation not supported "
"on this dataset"));
case EZFS_ACTIVE_SPARE:
return (dgettext(TEXT_DOMAIN, "pool has active shared spare "
"device"));
case EZFS_UNPLAYED_LOGS:
return (dgettext(TEXT_DOMAIN, "log device has unplayed intent "
"logs"));
case EZFS_REFTAG_RELE:
return (dgettext(TEXT_DOMAIN, "no such tag on this dataset"));
case EZFS_REFTAG_HOLD:
return (dgettext(TEXT_DOMAIN, "tag already exists on this "
"dataset"));
case EZFS_TAGTOOLONG:
return (dgettext(TEXT_DOMAIN, "tag too long"));
case EZFS_PIPEFAILED:
return (dgettext(TEXT_DOMAIN, "pipe create failed"));
case EZFS_THREADCREATEFAILED:
return (dgettext(TEXT_DOMAIN, "thread create failed"));
case EZFS_POSTSPLIT_ONLINE:
return (dgettext(TEXT_DOMAIN, "disk was split from this pool "
"into a new one"));
case EZFS_SCRUB_PAUSED:
return (dgettext(TEXT_DOMAIN, "scrub is paused; "
"use 'zpool scrub' to resume"));
case EZFS_SCRUBBING:
return (dgettext(TEXT_DOMAIN, "currently scrubbing; "
"use 'zpool scrub -s' to cancel current scrub"));
case EZFS_NO_SCRUB:
return (dgettext(TEXT_DOMAIN, "there is no active scrub"));
case EZFS_DIFF:
return (dgettext(TEXT_DOMAIN, "unable to generate diffs"));
case EZFS_DIFFDATA:
return (dgettext(TEXT_DOMAIN, "invalid diff data"));
case EZFS_POOLREADONLY:
return (dgettext(TEXT_DOMAIN, "pool is read-only"));
case EZFS_NO_PENDING:
return (dgettext(TEXT_DOMAIN, "operation is not "
"in progress"));
case EZFS_CHECKPOINT_EXISTS:
return (dgettext(TEXT_DOMAIN, "checkpoint exists"));
case EZFS_DISCARDING_CHECKPOINT:
return (dgettext(TEXT_DOMAIN, "currently discarding "
"checkpoint"));
case EZFS_NO_CHECKPOINT:
return (dgettext(TEXT_DOMAIN, "checkpoint does not exist"));
case EZFS_DEVRM_IN_PROGRESS:
return (dgettext(TEXT_DOMAIN, "device removal in progress"));
case EZFS_VDEV_TOO_BIG:
return (dgettext(TEXT_DOMAIN, "device exceeds supported size"));
case EZFS_ACTIVE_POOL:
return (dgettext(TEXT_DOMAIN, "pool is imported on a "
"different host"));
case EZFS_TOOMANY:
return (dgettext(TEXT_DOMAIN, "argument list too long"));
case EZFS_INITIALIZING:
return (dgettext(TEXT_DOMAIN, "currently initializing"));
case EZFS_NO_INITIALIZE:
return (dgettext(TEXT_DOMAIN, "there is no active "
"initialization"));
+ case EZFS_WRONG_PARENT:
+ return (dgettext(TEXT_DOMAIN, "invalid parent dataset"));
case EZFS_UNKNOWN:
return (dgettext(TEXT_DOMAIN, "unknown error"));
default:
assert(hdl->libzfs_error == 0);
return (dgettext(TEXT_DOMAIN, "no error"));
}
}
/*PRINTFLIKE2*/
void
zfs_error_aux(libzfs_handle_t *hdl, const char *fmt, ...)
{
va_list ap;
va_start(ap, fmt);
(void) vsnprintf(hdl->libzfs_desc, sizeof (hdl->libzfs_desc),
fmt, ap);
hdl->libzfs_desc_active = 1;
va_end(ap);
}
static void
zfs_verror(libzfs_handle_t *hdl, int error, const char *fmt, va_list ap)
{
(void) vsnprintf(hdl->libzfs_action, sizeof (hdl->libzfs_action),
fmt, ap);
hdl->libzfs_error = error;
if (hdl->libzfs_desc_active)
hdl->libzfs_desc_active = 0;
else
hdl->libzfs_desc[0] = '\0';
if (hdl->libzfs_printerr) {
if (error == EZFS_UNKNOWN) {
(void) fprintf(stderr, dgettext(TEXT_DOMAIN, "internal "
"error: %s\n"), libzfs_error_description(hdl));
abort();
}
(void) fprintf(stderr, "%s: %s\n", hdl->libzfs_action,
libzfs_error_description(hdl));
if (error == EZFS_NOMEM)
exit(1);
}
}
int
zfs_error(libzfs_handle_t *hdl, int error, const char *msg)
{
return (zfs_error_fmt(hdl, error, "%s", msg));
}
/*PRINTFLIKE3*/
int
zfs_error_fmt(libzfs_handle_t *hdl, int error, const char *fmt, ...)
{
va_list ap;
va_start(ap, fmt);
zfs_verror(hdl, error, fmt, ap);
va_end(ap);
return (-1);
}
static int
zfs_common_error(libzfs_handle_t *hdl, int error, const char *fmt,
va_list ap)
{
switch (error) {
case EPERM:
case EACCES:
zfs_verror(hdl, EZFS_PERM, fmt, ap);
return (-1);
case ECANCELED:
zfs_verror(hdl, EZFS_NODELEGATION, fmt, ap);
return (-1);
case EIO:
zfs_verror(hdl, EZFS_IO, fmt, ap);
return (-1);
case EFAULT:
zfs_verror(hdl, EZFS_FAULT, fmt, ap);
return (-1);
case EINTR:
zfs_verror(hdl, EZFS_INTR, fmt, ap);
return (-1);
}
return (0);
}
int
zfs_standard_error(libzfs_handle_t *hdl, int error, const char *msg)
{
return (zfs_standard_error_fmt(hdl, error, "%s", msg));
}
/*PRINTFLIKE3*/
int
zfs_standard_error_fmt(libzfs_handle_t *hdl, int error, const char *fmt, ...)
{
va_list ap;
va_start(ap, fmt);
if (zfs_common_error(hdl, error, fmt, ap) != 0) {
va_end(ap);
return (-1);
}
switch (error) {
case ENXIO:
case ENODEV:
case EPIPE:
zfs_verror(hdl, EZFS_IO, fmt, ap);
break;
case ENOENT:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"dataset does not exist"));
zfs_verror(hdl, EZFS_NOENT, fmt, ap);
break;
case ENOSPC:
case EDQUOT:
zfs_verror(hdl, EZFS_NOSPC, fmt, ap);
va_end(ap);
return (-1);
case EEXIST:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"dataset already exists"));
zfs_verror(hdl, EZFS_EXISTS, fmt, ap);
break;
case EBUSY:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"dataset is busy"));
zfs_verror(hdl, EZFS_BUSY, fmt, ap);
break;
case EROFS:
zfs_verror(hdl, EZFS_POOLREADONLY, fmt, ap);
break;
case ENAMETOOLONG:
zfs_verror(hdl, EZFS_NAMETOOLONG, fmt, ap);
break;
case ENOTSUP:
zfs_verror(hdl, EZFS_BADVERSION, fmt, ap);
break;
case EAGAIN:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool I/O is currently suspended"));
zfs_verror(hdl, EZFS_POOLUNAVAIL, fmt, ap);
break;
case EREMOTEIO:
zfs_verror(hdl, EZFS_ACTIVE_POOL, fmt, ap);
break;
default:
zfs_error_aux(hdl, strerror(error));
zfs_verror(hdl, EZFS_UNKNOWN, fmt, ap);
break;
}
va_end(ap);
return (-1);
}
int
zpool_standard_error(libzfs_handle_t *hdl, int error, const char *msg)
{
return (zpool_standard_error_fmt(hdl, error, "%s", msg));
}
/*PRINTFLIKE3*/
int
zpool_standard_error_fmt(libzfs_handle_t *hdl, int error, const char *fmt, ...)
{
va_list ap;
va_start(ap, fmt);
if (zfs_common_error(hdl, error, fmt, ap) != 0) {
va_end(ap);
return (-1);
}
switch (error) {
case ENODEV:
zfs_verror(hdl, EZFS_NODEVICE, fmt, ap);
break;
case ENOENT:
zfs_error_aux(hdl,
dgettext(TEXT_DOMAIN, "no such pool or dataset"));
zfs_verror(hdl, EZFS_NOENT, fmt, ap);
break;
case EEXIST:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool already exists"));
zfs_verror(hdl, EZFS_EXISTS, fmt, ap);
break;
case EBUSY:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN, "pool is busy"));
zfs_verror(hdl, EZFS_BUSY, fmt, ap);
break;
case ENXIO:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"one or more devices is currently unavailable"));
zfs_verror(hdl, EZFS_BADDEV, fmt, ap);
break;
case ENAMETOOLONG:
zfs_verror(hdl, EZFS_DEVOVERFLOW, fmt, ap);
break;
case ENOTSUP:
zfs_verror(hdl, EZFS_POOL_NOTSUP, fmt, ap);
break;
case EINVAL:
zfs_verror(hdl, EZFS_POOL_INVALARG, fmt, ap);
break;
case ENOSPC:
case EDQUOT:
zfs_verror(hdl, EZFS_NOSPC, fmt, ap);
va_end(ap);
return (-1);
case EAGAIN:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"pool I/O is currently suspended"));
zfs_verror(hdl, EZFS_POOLUNAVAIL, fmt, ap);
break;
case EROFS:
zfs_verror(hdl, EZFS_POOLREADONLY, fmt, ap);
break;
/* There is no pending operation to cancel */
case ESRCH:
zfs_verror(hdl, EZFS_NO_PENDING, fmt, ap);
break;
case EREMOTEIO:
zfs_verror(hdl, EZFS_ACTIVE_POOL, fmt, ap);
break;
case ZFS_ERR_CHECKPOINT_EXISTS:
zfs_verror(hdl, EZFS_CHECKPOINT_EXISTS, fmt, ap);
break;
case ZFS_ERR_DISCARDING_CHECKPOINT:
zfs_verror(hdl, EZFS_DISCARDING_CHECKPOINT, fmt, ap);
break;
case ZFS_ERR_NO_CHECKPOINT:
zfs_verror(hdl, EZFS_NO_CHECKPOINT, fmt, ap);
break;
case ZFS_ERR_DEVRM_IN_PROGRESS:
zfs_verror(hdl, EZFS_DEVRM_IN_PROGRESS, fmt, ap);
break;
case ZFS_ERR_VDEV_TOO_BIG:
zfs_verror(hdl, EZFS_VDEV_TOO_BIG, fmt, ap);
+ break;
+ case ZFS_ERR_WRONG_PARENT:
+ zfs_verror(hdl, EZFS_WRONG_PARENT, fmt, ap);
break;
default:
zfs_error_aux(hdl, strerror(error));
zfs_verror(hdl, EZFS_UNKNOWN, fmt, ap);
}
va_end(ap);
return (-1);
}
/*
* Display an out of memory error message and abort the current program.
*/
int
no_memory(libzfs_handle_t *hdl)
{
return (zfs_error(hdl, EZFS_NOMEM, "internal error"));
}
/*
* A safe form of malloc() which will die if the allocation fails.
*/
void *
zfs_alloc(libzfs_handle_t *hdl, size_t size)
{
void *data;
if ((data = calloc(1, size)) == NULL)
(void) no_memory(hdl);
return (data);
}
/*
* A safe form of asprintf() which will die if the allocation fails.
*/
/*PRINTFLIKE2*/
char *
zfs_asprintf(libzfs_handle_t *hdl, const char *fmt, ...)
{
va_list ap;
char *ret;
int err;
va_start(ap, fmt);
err = vasprintf(&ret, fmt, ap);
va_end(ap);
if (err < 0)
(void) no_memory(hdl);
return (ret);
}
/*
* A safe form of realloc(), which also zeroes newly allocated space.
*/
void *
zfs_realloc(libzfs_handle_t *hdl, void *ptr, size_t oldsize, size_t newsize)
{
void *ret;
if ((ret = realloc(ptr, newsize)) == NULL) {
(void) no_memory(hdl);
return (NULL);
}
bzero((char *)ret + oldsize, (newsize - oldsize));
return (ret);
}
/*
* A safe form of strdup() which will die if the allocation fails.
*/
char *
zfs_strdup(libzfs_handle_t *hdl, const char *str)
{
char *ret;
if ((ret = strdup(str)) == NULL)
(void) no_memory(hdl);
return (ret);
}
/*
* Convert a number to an appropriately human-readable output.
*/
void
zfs_nicenum(uint64_t num, char *buf, size_t buflen)
{
nicenum(num, buf, buflen);
}
void
libzfs_print_on_error(libzfs_handle_t *hdl, boolean_t printerr)
{
hdl->libzfs_printerr = printerr;
}
static int
libzfs_load(void)
{
int error;
if (modfind("zfs") < 0) {
/* Not present in kernel, try loading it. */
if (kldload("zfs") < 0 || modfind("zfs") < 0) {
if (errno != EEXIST)
return (-1);
}
}
return (0);
}
libzfs_handle_t *
libzfs_init(void)
{
libzfs_handle_t *hdl;
if ((hdl = calloc(1, sizeof (libzfs_handle_t))) == NULL) {
return (NULL);
}
if (libzfs_load() < 0) {
free(hdl);
return (NULL);
}
if ((hdl->libzfs_fd = open(ZFS_DEV, O_RDWR)) < 0) {
free(hdl);
return (NULL);
}
if ((hdl->libzfs_mnttab = fopen(MNTTAB, "r")) == NULL) {
(void) close(hdl->libzfs_fd);
free(hdl);
return (NULL);
}
hdl->libzfs_sharetab = fopen(ZFS_EXPORTS_PATH, "r");
if (libzfs_core_init() != 0) {
(void) close(hdl->libzfs_fd);
(void) fclose(hdl->libzfs_mnttab);
(void) fclose(hdl->libzfs_sharetab);
free(hdl);
return (NULL);
}
zfs_prop_init();
zpool_prop_init();
zpool_feature_init();
libzfs_mnttab_init(hdl);
if (getenv("ZFS_PROP_DEBUG") != NULL) {
hdl->libzfs_prop_debug = B_TRUE;
}
return (hdl);
}
void
libzfs_fini(libzfs_handle_t *hdl)
{
(void) close(hdl->libzfs_fd);
if (hdl->libzfs_mnttab)
(void) fclose(hdl->libzfs_mnttab);
if (hdl->libzfs_sharetab)
(void) fclose(hdl->libzfs_sharetab);
zfs_uninit_libshare(hdl);
zpool_free_handles(hdl);
#ifdef illumos
libzfs_fru_clear(hdl, B_TRUE);
#endif
namespace_clear(hdl);
libzfs_mnttab_fini(hdl);
libzfs_core_fini();
free(hdl);
}
libzfs_handle_t *
zpool_get_handle(zpool_handle_t *zhp)
{
return (zhp->zpool_hdl);
}
libzfs_handle_t *
zfs_get_handle(zfs_handle_t *zhp)
{
return (zhp->zfs_hdl);
}
zpool_handle_t *
zfs_get_pool_handle(const zfs_handle_t *zhp)
{
return (zhp->zpool_hdl);
}
/*
* Given a name, determine whether or not it's a valid path
* (starts with '/' or "./"). If so, walk the mnttab trying
* to match the device number. If not, treat the path as an
* fs/vol/snap/bkmark name.
*/
zfs_handle_t *
zfs_path_to_zhandle(libzfs_handle_t *hdl, char *path, zfs_type_t argtype)
{
struct stat64 statbuf;
struct extmnttab entry;
int ret;
if (path[0] != '/' && strncmp(path, "./", strlen("./")) != 0) {
/*
* It's not a valid path, assume it's a name of type 'argtype'.
*/
return (zfs_open(hdl, path, argtype));
}
if (stat64(path, &statbuf) != 0) {
(void) fprintf(stderr, "%s: %s\n", path, strerror(errno));
return (NULL);
}
#ifdef illumos
rewind(hdl->libzfs_mnttab);
while ((ret = getextmntent(hdl->libzfs_mnttab, &entry, 0)) == 0) {
if (makedevice(entry.mnt_major, entry.mnt_minor) ==
statbuf.st_dev) {
break;
}
}
#else
{
struct statfs sfs;
ret = statfs(path, &sfs);
if (ret == 0)
statfs2mnttab(&sfs, &entry);
else {
(void) fprintf(stderr, "%s: %s\n", path,
strerror(errno));
}
}
#endif /* illumos */
if (ret != 0) {
return (NULL);
}
if (strcmp(entry.mnt_fstype, MNTTYPE_ZFS) != 0) {
(void) fprintf(stderr, gettext("'%s': not a ZFS filesystem\n"),
path);
return (NULL);
}
return (zfs_open(hdl, entry.mnt_special, ZFS_TYPE_FILESYSTEM));
}
/*
* Initialize the zc_nvlist_dst member to prepare for receiving an nvlist from
* an ioctl().
*/
int
zcmd_alloc_dst_nvlist(libzfs_handle_t *hdl, zfs_cmd_t *zc, size_t len)
{
if (len == 0)
len = 16 * 1024;
zc->zc_nvlist_dst_size = len;
zc->zc_nvlist_dst =
(uint64_t)(uintptr_t)zfs_alloc(hdl, zc->zc_nvlist_dst_size);
if (zc->zc_nvlist_dst == 0)
return (-1);
return (0);
}
/*
* Called when an ioctl() which returns an nvlist fails with ENOMEM. This will
* expand the nvlist to the size specified in 'zc_nvlist_dst_size', which was
* filled in by the kernel to indicate the actual required size.
*/
int
zcmd_expand_dst_nvlist(libzfs_handle_t *hdl, zfs_cmd_t *zc)
{
free((void *)(uintptr_t)zc->zc_nvlist_dst);
zc->zc_nvlist_dst =
(uint64_t)(uintptr_t)zfs_alloc(hdl, zc->zc_nvlist_dst_size);
if (zc->zc_nvlist_dst == 0)
return (-1);
return (0);
}
/*
* Called to free the src and dst nvlists stored in the command structure.
*/
void
zcmd_free_nvlists(zfs_cmd_t *zc)
{
free((void *)(uintptr_t)zc->zc_nvlist_conf);
free((void *)(uintptr_t)zc->zc_nvlist_src);
free((void *)(uintptr_t)zc->zc_nvlist_dst);
zc->zc_nvlist_conf = NULL;
zc->zc_nvlist_src = NULL;
zc->zc_nvlist_dst = NULL;
}
static int
zcmd_write_nvlist_com(libzfs_handle_t *hdl, uint64_t *outnv, uint64_t *outlen,
nvlist_t *nvl)
{
char *packed;
size_t len;
verify(nvlist_size(nvl, &len, NV_ENCODE_NATIVE) == 0);
if ((packed = zfs_alloc(hdl, len)) == NULL)
return (-1);
verify(nvlist_pack(nvl, &packed, &len, NV_ENCODE_NATIVE, 0) == 0);
*outnv = (uint64_t)(uintptr_t)packed;
*outlen = len;
return (0);
}
int
zcmd_write_conf_nvlist(libzfs_handle_t *hdl, zfs_cmd_t *zc, nvlist_t *nvl)
{
return (zcmd_write_nvlist_com(hdl, &zc->zc_nvlist_conf,
&zc->zc_nvlist_conf_size, nvl));
}
int
zcmd_write_src_nvlist(libzfs_handle_t *hdl, zfs_cmd_t *zc, nvlist_t *nvl)
{
return (zcmd_write_nvlist_com(hdl, &zc->zc_nvlist_src,
&zc->zc_nvlist_src_size, nvl));
}
/*
* Unpacks an nvlist from the ZFS ioctl command structure.
*/
int
zcmd_read_dst_nvlist(libzfs_handle_t *hdl, zfs_cmd_t *zc, nvlist_t **nvlp)
{
if (nvlist_unpack((void *)(uintptr_t)zc->zc_nvlist_dst,
zc->zc_nvlist_dst_size, nvlp, 0) != 0)
return (no_memory(hdl));
return (0);
}
int
zfs_ioctl(libzfs_handle_t *hdl, int request, zfs_cmd_t *zc)
{
return (ioctl(hdl->libzfs_fd, request, zc));
}
/*
* ================================================================
* API shared by zfs and zpool property management
* ================================================================
*/
static void
zprop_print_headers(zprop_get_cbdata_t *cbp, zfs_type_t type)
{
zprop_list_t *pl = cbp->cb_proplist;
int i;
char *title;
size_t len;
cbp->cb_first = B_FALSE;
if (cbp->cb_scripted)
return;
/*
* Start with the length of the column headers.
*/
cbp->cb_colwidths[GET_COL_NAME] = strlen(dgettext(TEXT_DOMAIN, "NAME"));
cbp->cb_colwidths[GET_COL_PROPERTY] = strlen(dgettext(TEXT_DOMAIN,
"PROPERTY"));
cbp->cb_colwidths[GET_COL_VALUE] = strlen(dgettext(TEXT_DOMAIN,
"VALUE"));
cbp->cb_colwidths[GET_COL_RECVD] = strlen(dgettext(TEXT_DOMAIN,
"RECEIVED"));
cbp->cb_colwidths[GET_COL_SOURCE] = strlen(dgettext(TEXT_DOMAIN,
"SOURCE"));
/* first property is always NAME */
assert(cbp->cb_proplist->pl_prop ==
((type == ZFS_TYPE_POOL) ? ZPOOL_PROP_NAME : ZFS_PROP_NAME));
/*
* Go through and calculate the widths for each column. For the
* 'source' column, we kludge it up by taking the worst-case scenario of
* inheriting from the longest name. This is acceptable because in the
* majority of cases 'SOURCE' is the last column displayed, and we don't
* use the width anyway. Note that the 'VALUE' column can be oversized,
* if the name of the property is much longer than any values we find.
*/
for (pl = cbp->cb_proplist; pl != NULL; pl = pl->pl_next) {
/*
* 'PROPERTY' column
*/
if (pl->pl_prop != ZPROP_INVAL) {
const char *propname = (type == ZFS_TYPE_POOL) ?
zpool_prop_to_name(pl->pl_prop) :
zfs_prop_to_name(pl->pl_prop);
len = strlen(propname);
if (len > cbp->cb_colwidths[GET_COL_PROPERTY])
cbp->cb_colwidths[GET_COL_PROPERTY] = len;
} else {
len = strlen(pl->pl_user_prop);
if (len > cbp->cb_colwidths[GET_COL_PROPERTY])
cbp->cb_colwidths[GET_COL_PROPERTY] = len;
}
/*
* 'VALUE' column. The first property is always the 'name'
* property that was tacked on either by /sbin/zfs's
* zfs_do_get() or when calling zprop_expand_list(), so we
* ignore its width. If the user specified the name property
* to display, then it will be later in the list in any case.
*/
if (pl != cbp->cb_proplist &&
pl->pl_width > cbp->cb_colwidths[GET_COL_VALUE])
cbp->cb_colwidths[GET_COL_VALUE] = pl->pl_width;
/* 'RECEIVED' column. */
if (pl != cbp->cb_proplist &&
pl->pl_recvd_width > cbp->cb_colwidths[GET_COL_RECVD])
cbp->cb_colwidths[GET_COL_RECVD] = pl->pl_recvd_width;
/*
* 'NAME' and 'SOURCE' columns
*/
if (pl->pl_prop == (type == ZFS_TYPE_POOL ? ZPOOL_PROP_NAME :
ZFS_PROP_NAME) &&
pl->pl_width > cbp->cb_colwidths[GET_COL_NAME]) {
cbp->cb_colwidths[GET_COL_NAME] = pl->pl_width;
cbp->cb_colwidths[GET_COL_SOURCE] = pl->pl_width +
strlen(dgettext(TEXT_DOMAIN, "inherited from"));
}
}
/*
* Now go through and print the headers.
*/
for (i = 0; i < ZFS_GET_NCOLS; i++) {
switch (cbp->cb_columns[i]) {
case GET_COL_NAME:
title = dgettext(TEXT_DOMAIN, "NAME");
break;
case GET_COL_PROPERTY:
title = dgettext(TEXT_DOMAIN, "PROPERTY");
break;
case GET_COL_VALUE:
title = dgettext(TEXT_DOMAIN, "VALUE");
break;
case GET_COL_RECVD:
title = dgettext(TEXT_DOMAIN, "RECEIVED");
break;
case GET_COL_SOURCE:
title = dgettext(TEXT_DOMAIN, "SOURCE");
break;
default:
title = NULL;
}
if (title != NULL) {
if (i == (ZFS_GET_NCOLS - 1) ||
cbp->cb_columns[i + 1] == GET_COL_NONE)
(void) printf("%s", title);
else
(void) printf("%-*s ",
cbp->cb_colwidths[cbp->cb_columns[i]],
title);
}
}
(void) printf("\n");
}
/*
* Display a single line of output, according to the settings in the callback
* structure.
*/
void
zprop_print_one_property(const char *name, zprop_get_cbdata_t *cbp,
const char *propname, const char *value, zprop_source_t sourcetype,
const char *source, const char *recvd_value)
{
int i;
const char *str = NULL;
char buf[128];
/*
* Ignore those source types that the user has chosen to ignore.
*/
if ((sourcetype & cbp->cb_sources) == 0)
return;
if (cbp->cb_first)
zprop_print_headers(cbp, cbp->cb_type);
for (i = 0; i < ZFS_GET_NCOLS; i++) {
switch (cbp->cb_columns[i]) {
case GET_COL_NAME:
str = name;
break;
case GET_COL_PROPERTY:
str = propname;
break;
case GET_COL_VALUE:
str = value;
break;
case GET_COL_SOURCE:
switch (sourcetype) {
case ZPROP_SRC_NONE:
str = "-";
break;
case ZPROP_SRC_DEFAULT:
str = "default";
break;
case ZPROP_SRC_LOCAL:
str = "local";
break;
case ZPROP_SRC_TEMPORARY:
str = "temporary";
break;
case ZPROP_SRC_INHERITED:
(void) snprintf(buf, sizeof (buf),
"inherited from %s", source);
str = buf;
break;
case ZPROP_SRC_RECEIVED:
str = "received";
break;
default:
str = NULL;
assert(!"unhandled zprop_source_t");
}
break;
case GET_COL_RECVD:
str = (recvd_value == NULL ? "-" : recvd_value);
break;
default:
continue;
}
if (cbp->cb_columns[i + 1] == GET_COL_NONE)
(void) printf("%s", str);
else if (cbp->cb_scripted)
(void) printf("%s\t", str);
else
(void) printf("%-*s ",
cbp->cb_colwidths[cbp->cb_columns[i]],
str);
}
(void) printf("\n");
}
/*
* Given a numeric suffix, convert the value into a number of bits that the
* resulting value must be shifted.
*/
static int
str2shift(libzfs_handle_t *hdl, const char *buf)
{
const char *ends = "BKMGTPEZ";
int i;
if (buf[0] == '\0')
return (0);
for (i = 0; i < strlen(ends); i++) {
if (toupper(buf[0]) == ends[i])
break;
}
if (i == strlen(ends)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"invalid numeric suffix '%s'"), buf);
return (-1);
}
/*
* We want to allow trailing 'b' characters for 'GB' or 'Mb'. But don't
* allow 'BB' - that's just weird.
*/
if (buf[1] == '\0' || (toupper(buf[1]) == 'B' && buf[2] == '\0' &&
toupper(buf[0]) != 'B'))
return (10*i);
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"invalid numeric suffix '%s'"), buf);
return (-1);
}
/*
* Convert a string of the form '100G' into a real number. Used when setting
* properties or creating a volume. 'buf' is used to place an extended error
* message for the caller to use.
*/
int
zfs_nicestrtonum(libzfs_handle_t *hdl, const char *value, uint64_t *num)
{
char *end;
int shift;
*num = 0;
/* Check to see if this looks like a number. */
if ((value[0] < '0' || value[0] > '9') && value[0] != '.') {
if (hdl)
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"bad numeric value '%s'"), value);
return (-1);
}
/* Rely on strtoull() to process the numeric portion. */
errno = 0;
*num = strtoull(value, &end, 10);
/*
* Check for ERANGE, which indicates that the value is too large to fit
* in a 64-bit value.
*/
if (errno == ERANGE) {
if (hdl)
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"numeric value is too large"));
return (-1);
}
/*
* If we have a decimal value, then do the computation with floating
* point arithmetic. Otherwise, use standard arithmetic.
*/
if (*end == '.') {
double fval = strtod(value, &end);
if ((shift = str2shift(hdl, end)) == -1)
return (-1);
fval *= pow(2, shift);
if (fval > UINT64_MAX) {
if (hdl)
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"numeric value is too large"));
return (-1);
}
*num = (uint64_t)fval;
} else {
if ((shift = str2shift(hdl, end)) == -1)
return (-1);
/* Check for overflow */
if (shift >= 64 || (*num << shift) >> shift != *num) {
if (hdl)
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"numeric value is too large"));
return (-1);
}
*num <<= shift;
}
return (0);
}
/*
* Given a propname=value nvpair to set, parse any numeric properties
* (index, boolean, etc) if they are specified as strings and add the
* resulting nvpair to the returned nvlist.
*
* At the DSL layer, all properties are either 64-bit numbers or strings.
* We want the user to be able to ignore this fact and specify properties
* as native values (numbers, for example) or as strings (to simplify
* command line utilities). This also handles converting index types
* (compression, checksum, etc) from strings to their on-disk index.
*/
int
zprop_parse_value(libzfs_handle_t *hdl, nvpair_t *elem, int prop,
zfs_type_t type, nvlist_t *ret, char **svalp, uint64_t *ivalp,
const char *errbuf)
{
data_type_t datatype = nvpair_type(elem);
zprop_type_t proptype;
const char *propname;
char *value;
boolean_t isnone = B_FALSE;
boolean_t isauto = B_FALSE;
if (type == ZFS_TYPE_POOL) {
proptype = zpool_prop_get_type(prop);
propname = zpool_prop_to_name(prop);
} else {
proptype = zfs_prop_get_type(prop);
propname = zfs_prop_to_name(prop);
}
/*
* Convert any properties to the internal DSL value types.
*/
*svalp = NULL;
*ivalp = 0;
switch (proptype) {
case PROP_TYPE_STRING:
if (datatype != DATA_TYPE_STRING) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' must be a string"), nvpair_name(elem));
goto error;
}
(void) nvpair_value_string(elem, svalp);
if (strlen(*svalp) >= ZFS_MAXPROPLEN) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' is too long"), nvpair_name(elem));
goto error;
}
break;
case PROP_TYPE_NUMBER:
if (datatype == DATA_TYPE_STRING) {
(void) nvpair_value_string(elem, &value);
if (strcmp(value, "none") == 0) {
isnone = B_TRUE;
} else if (strcmp(value, "auto") == 0) {
isauto = B_TRUE;
} else if (zfs_nicestrtonum(hdl, value, ivalp) != 0) {
goto error;
}
} else if (datatype == DATA_TYPE_UINT64) {
(void) nvpair_value_uint64(elem, ivalp);
} else {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' must be a number"), nvpair_name(elem));
goto error;
}
/*
* Quota special: force 'none' and don't allow 0.
*/
if ((type & ZFS_TYPE_DATASET) && *ivalp == 0 && !isnone &&
(prop == ZFS_PROP_QUOTA || prop == ZFS_PROP_REFQUOTA)) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"use 'none' to disable quota/refquota"));
goto error;
}
/*
* Special handling for "*_limit=none". In this case it's not
* 0 but UINT64_MAX.
*/
if ((type & ZFS_TYPE_DATASET) && isnone &&
(prop == ZFS_PROP_FILESYSTEM_LIMIT ||
prop == ZFS_PROP_SNAPSHOT_LIMIT)) {
*ivalp = UINT64_MAX;
}
/*
* Special handling for setting 'refreservation' to 'auto'. Use
* UINT64_MAX to tell the caller to use zfs_fix_auto_resv().
* 'auto' is only allowed on volumes.
*/
if (isauto) {
switch (prop) {
case ZFS_PROP_REFRESERVATION:
if ((type & ZFS_TYPE_VOLUME) == 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s=auto' only allowed on "
"volumes"), nvpair_name(elem));
goto error;
}
*ivalp = UINT64_MAX;
break;
default:
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'auto' is invalid value for '%s'"),
nvpair_name(elem));
goto error;
}
}
break;
case PROP_TYPE_INDEX:
if (datatype != DATA_TYPE_STRING) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' must be a string"), nvpair_name(elem));
goto error;
}
(void) nvpair_value_string(elem, &value);
if (zprop_string_to_index(prop, value, ivalp, type) != 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"'%s' must be one of '%s'"), propname,
zprop_values(prop, type));
goto error;
}
break;
default:
abort();
}
/*
* Add the result to our return set of properties.
*/
if (*svalp != NULL) {
if (nvlist_add_string(ret, propname, *svalp) != 0) {
(void) no_memory(hdl);
return (-1);
}
} else {
if (nvlist_add_uint64(ret, propname, *ivalp) != 0) {
(void) no_memory(hdl);
return (-1);
}
}
return (0);
error:
(void) zfs_error(hdl, EZFS_BADPROP, errbuf);
return (-1);
}
static int
addlist(libzfs_handle_t *hdl, char *propname, zprop_list_t **listp,
zfs_type_t type)
{
int prop;
zprop_list_t *entry;
prop = zprop_name_to_prop(propname, type);
if (prop != ZPROP_INVAL && !zprop_valid_for_type(prop, type))
prop = ZPROP_INVAL;
/*
* When no property table entry can be found, return failure if
* this is a pool property or if this isn't a user-defined
* dataset property,
*/
if (prop == ZPROP_INVAL && ((type == ZFS_TYPE_POOL &&
!zpool_prop_feature(propname) &&
!zpool_prop_unsupported(propname)) ||
(type == ZFS_TYPE_DATASET && !zfs_prop_user(propname) &&
!zfs_prop_userquota(propname) && !zfs_prop_written(propname)))) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"invalid property '%s'"), propname);
return (zfs_error(hdl, EZFS_BADPROP,
dgettext(TEXT_DOMAIN, "bad property list")));
}
if ((entry = zfs_alloc(hdl, sizeof (zprop_list_t))) == NULL)
return (-1);
entry->pl_prop = prop;
if (prop == ZPROP_INVAL) {
if ((entry->pl_user_prop = zfs_strdup(hdl, propname)) ==
NULL) {
free(entry);
return (-1);
}
entry->pl_width = strlen(propname);
} else {
entry->pl_width = zprop_width(prop, &entry->pl_fixed,
type);
}
*listp = entry;
return (0);
}
/*
* Given a comma-separated list of properties, construct a property list
* containing both user-defined and native properties. This function will
* return a NULL list if 'all' is specified, which can later be expanded
* by zprop_expand_list().
*/
int
zprop_get_list(libzfs_handle_t *hdl, char *props, zprop_list_t **listp,
zfs_type_t type)
{
*listp = NULL;
/*
* If 'all' is specified, return a NULL list.
*/
if (strcmp(props, "all") == 0)
return (0);
/*
* If no props were specified, return an error.
*/
if (props[0] == '\0') {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"no properties specified"));
return (zfs_error(hdl, EZFS_BADPROP, dgettext(TEXT_DOMAIN,
"bad property list")));
}
/*
* It would be nice to use getsubopt() here, but the inclusion of column
* aliases makes this more effort than it's worth.
*/
while (*props != '\0') {
size_t len;
char *p;
char c;
if ((p = strchr(props, ',')) == NULL) {
len = strlen(props);
p = props + len;
} else {
len = p - props;
}
/*
* Check for empty options.
*/
if (len == 0) {
zfs_error_aux(hdl, dgettext(TEXT_DOMAIN,
"empty property name"));
return (zfs_error(hdl, EZFS_BADPROP,
dgettext(TEXT_DOMAIN, "bad property list")));
}
/*
* Check all regular property names.
*/
c = props[len];
props[len] = '\0';
if (strcmp(props, "space") == 0) {
static char *spaceprops[] = {
"name", "avail", "used", "usedbysnapshots",
"usedbydataset", "usedbyrefreservation",
"usedbychildren", NULL
};
int i;
for (i = 0; spaceprops[i]; i++) {
if (addlist(hdl, spaceprops[i], listp, type))
return (-1);
listp = &(*listp)->pl_next;
}
} else {
if (addlist(hdl, props, listp, type))
return (-1);
listp = &(*listp)->pl_next;
}
props = p;
if (c == ',')
props++;
}
return (0);
}
void
zprop_free_list(zprop_list_t *pl)
{
zprop_list_t *next;
while (pl != NULL) {
next = pl->pl_next;
free(pl->pl_user_prop);
free(pl);
pl = next;
}
}
typedef struct expand_data {
zprop_list_t **last;
libzfs_handle_t *hdl;
zfs_type_t type;
} expand_data_t;
int
zprop_expand_list_cb(int prop, void *cb)
{
zprop_list_t *entry;
expand_data_t *edp = cb;
if ((entry = zfs_alloc(edp->hdl, sizeof (zprop_list_t))) == NULL)
return (ZPROP_INVAL);
entry->pl_prop = prop;
entry->pl_width = zprop_width(prop, &entry->pl_fixed, edp->type);
entry->pl_all = B_TRUE;
*(edp->last) = entry;
edp->last = &entry->pl_next;
return (ZPROP_CONT);
}
int
zprop_expand_list(libzfs_handle_t *hdl, zprop_list_t **plp, zfs_type_t type)
{
zprop_list_t *entry;
zprop_list_t **last;
expand_data_t exp;
if (*plp == NULL) {
/*
* If this is the very first time we've been called for an 'all'
* specification, expand the list to include all native
* properties.
*/
last = plp;
exp.last = last;
exp.hdl = hdl;
exp.type = type;
if (zprop_iter_common(zprop_expand_list_cb, &exp, B_FALSE,
B_FALSE, type) == ZPROP_INVAL)
return (-1);
/*
* Add 'name' to the beginning of the list, which is handled
* specially.
*/
if ((entry = zfs_alloc(hdl, sizeof (zprop_list_t))) == NULL)
return (-1);
entry->pl_prop = (type == ZFS_TYPE_POOL) ? ZPOOL_PROP_NAME :
ZFS_PROP_NAME;
entry->pl_width = zprop_width(entry->pl_prop,
&entry->pl_fixed, type);
entry->pl_all = B_TRUE;
entry->pl_next = *plp;
*plp = entry;
}
return (0);
}
int
zprop_iter(zprop_func func, void *cb, boolean_t show_all, boolean_t ordered,
zfs_type_t type)
{
return (zprop_iter_common(func, cb, show_all, ordered, type));
}
ulong_t
get_system_hostid(void)
{
char *env;
/*
* Allow the hostid to be subverted for testing.
*/
env = getenv("ZFS_HOSTID");
if (env) {
ulong_t hostid = strtoull(env, NULL, 16);
return (hostid & 0xFFFFFFFF);
}
return (gethostid());
}
Index: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_objset.c
===================================================================
--- head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_objset.c (revision 359302)
+++ head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_objset.c (revision 359303)
@@ -1,2451 +1,2478 @@
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2017 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) 2014 Integros [integros.com]
* Copyright 2017 Nexenta Systems, Inc.
+ * Copyright (c) 2018, loli10K <ezomori.nozomu@gmail.com>. All rights reserved.
*/
/* Portions Copyright 2010 Robert Milkowski */
#include <sys/cred.h>
#include <sys/zfs_context.h>
#include <sys/dmu_objset.h>
#include <sys/dsl_dir.h>
#include <sys/dsl_dataset.h>
#include <sys/dsl_prop.h>
#include <sys/dsl_pool.h>
#include <sys/dsl_synctask.h>
#include <sys/dsl_deleg.h>
#include <sys/dnode.h>
#include <sys/dbuf.h>
#include <sys/zvol.h>
#include <sys/dmu_tx.h>
#include <sys/zap.h>
#include <sys/zil.h>
#include <sys/dmu_impl.h>
#include <sys/zfs_ioctl.h>
#include <sys/sa.h>
#include <sys/zfs_onexit.h>
#include <sys/dsl_destroy.h>
#include <sys/vdev.h>
#include <sys/zfeature.h>
#include "zfs_namecheck.h"
/*
* 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 parallization
* 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.
*/
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.
*/
int dmu_rescan_dnode_threshold = 131072;
static void dmu_objset_find_dp_cb(void *arg);
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, newval,
ZIO_COMPRESS_ON);
}
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
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);
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_OLD_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_OLD_PHYS_SIZE || 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 == sizeof (objset_phys_t)) {
dnode_byteswap(&osp->os_userused_dnode);
dnode_byteswap(&osp->os_groupused_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);
}
unsigned int
dnode_multilist_index_func(multilist_t *ml, void *obj)
{
dnode_t *dn = obj;
return (dnode_hash(dn->dn_objset, dn->dn_object) %
multilist_get_num_sublists(ml));
}
/*
* 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));
#if 0
/*
* 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);
}
#endif
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;
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;
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_FLAG_CANFAIL, &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);
}
/* Increase the blocksize if we are permitted. */
if (spa_version(spa) >= SPA_VERSION_USERSPACE &&
arc_buf_size(os->os_phys_buf) < sizeof (objset_phys_t)) {
arc_buf_t *buf = arc_alloc_buf(spa, &os->os_phys_buf,
ARC_BUFC_METADATA, sizeof (objset_phys_t));
bzero(buf->b_data, sizeof (objset_phys_t));
bcopy(os->os_phys_buf->b_data, 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_OLD_PHYS_SIZE;
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;
bzero(os->os_phys, size);
}
/*
* 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) {
boolean_t needlock = B_FALSE;
/*
* Note: it's valid to open the objset if the dataset is
* long-held, in which case the pool_config lock will not
* be held.
*/
if (!dsl_pool_config_held(dmu_objset_pool(os))) {
needlock = B_TRUE;
dsl_pool_config_enter(dmu_objset_pool(os), FTAG);
}
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 (!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 (needlock)
dsl_pool_config_exit(dmu_objset_pool(os), FTAG);
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_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;
}
/*
* 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;
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++) {
os->os_dirty_dnodes[i] = multilist_create(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));
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 (arc_buf_size(os->os_phys_buf) >= sizeof (objset_phys_t)) {
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);
}
*osp = os;
return (0);
}
int
dmu_objset_from_ds(dsl_dataset_t *ds, objset_t **osp)
{
int err = 0;
/*
* We shouldn't be doing anything with dsl_dataset_t's unless the
* pool_config lock is held, or the dataset is long-held.
*/
ASSERT(dsl_pool_config_held(ds->ds_dir->dd_pool) ||
dsl_dataset_long_held(ds));
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(const char *name, void *tag, objset_t **osp)
{
dsl_pool_t *dp;
dsl_dataset_t *ds;
int err;
err = dsl_pool_hold(name, tag, &dp);
if (err != 0)
return (err);
err = dsl_dataset_hold(dp, name, 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);
}
static int
dmu_objset_own_impl(dsl_dataset_t *ds, dmu_objset_type_t type,
boolean_t readonly, void *tag, objset_t **osp)
{
int err;
err = dmu_objset_from_ds(ds, osp);
if (err != 0) {
dsl_dataset_disown(ds, tag);
} else if (type != DMU_OST_ANY && type != (*osp)->os_phys->os_type) {
dsl_dataset_disown(ds, tag);
return (SET_ERROR(EINVAL));
} else if (!readonly && dsl_dataset_is_snapshot(ds)) {
dsl_dataset_disown(ds, tag);
return (SET_ERROR(EROFS));
}
return (err);
}
/*
* 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, void *tag, objset_t **osp)
{
dsl_pool_t *dp;
dsl_dataset_t *ds;
int err;
err = dsl_pool_hold(name, FTAG, &dp);
if (err != 0)
return (err);
err = dsl_dataset_own(dp, name, tag, &ds);
if (err != 0) {
dsl_pool_rele(dp, FTAG);
return (err);
}
err = dmu_objset_own_impl(ds, type, readonly, tag, osp);
dsl_pool_rele(dp, FTAG);
return (err);
}
int
dmu_objset_own_obj(dsl_pool_t *dp, uint64_t obj, dmu_objset_type_t type,
boolean_t readonly, void *tag, objset_t **osp)
{
dsl_dataset_t *ds;
int err;
err = dsl_dataset_own_obj(dp, obj, tag, &ds);
if (err != 0)
return (err);
return (dmu_objset_own_impl(ds, type, readonly, tag, osp));
}
void
dmu_objset_rele(objset_t *os, void *tag)
{
dsl_pool_t *dp = dmu_objset_pool(os);
dsl_dataset_rele(os->os_dsl_dataset, tag);
dsl_pool_rele(dp, 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,
void *tag)
{
dsl_pool_t *dp;
char name[ZFS_MAX_DATASET_NAME_LEN];
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, tag);
VERIFY0(dsl_dataset_own(dp, name, tag, newds));
dsl_pool_config_exit(dp, FTAG);
}
void
dmu_objset_disown(objset_t *os, void *tag)
{
dsl_dataset_disown(os->os_dsl_dataset, tag);
}
void
dmu_objset_evict_dbufs(objset_t *os)
{
dnode_t dn_marker;
dnode_t *dn;
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);
if (DMU_USERUSED_DNODE(os) != NULL) {
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)) {
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);
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));
}
timestruc_t
dmu_objset_snap_cmtime(objset_t *os)
{
return (dsl_dir_snap_cmtime(os->os_dsl_dataset->ds_dir));
}
/* 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)
{
objset_t *os;
dnode_t *mdn;
ASSERT(dmu_tx_is_syncing(tx));
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, DNODE_BLOCK_SIZE, DN_MAX_INDBLKSHIFT,
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 quescing 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) {
int 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;
if (dmu_objset_userused_enabled(os)) {
os->os_phys->os_flags |= OBJSET_FLAG_USERACCOUNTING_COMPLETE;
os->os_flags = os->os_phys->os_flags;
}
dsl_dataset_dirty(ds, tx);
return (os);
}
typedef struct dmu_objset_create_arg {
const char *doca_name;
cred_t *doca_cred;
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;
} dmu_objset_create_arg_t;
/*ARGSUSED*/
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 = dsl_fs_ss_limit_check(pdd, 1, ZFS_PROP_FILESYSTEM_LIMIT, NULL,
doca->doca_cred);
+ 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);
dsl_dir_t *pdd;
const char *tail;
dsl_dataset_t *ds;
uint64_t obj;
blkptr_t *bp;
objset_t *os;
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, tx);
VERIFY0(dsl_dataset_hold_obj(pdd->dd_pool, obj, FTAG, &ds));
rrw_enter(&ds->ds_bp_rwlock, RW_READER, FTAG);
bp = dsl_dataset_get_blkptr(ds);
os = dmu_objset_create_impl(pdd->dd_pool->dp_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);
}
spa_history_log_internal_ds(ds, "create", tx, "");
dsl_dataset_rele(ds, FTAG);
dsl_dir_rele(pdd, FTAG);
}
int
dmu_objset_create(const char *name, dmu_objset_type_t type, uint64_t flags,
void (*func)(objset_t *os, void *arg, cred_t *cr, dmu_tx_t *tx), void *arg)
{
dmu_objset_create_arg_t doca;
doca.doca_name = name;
doca.doca_cred = CRED();
doca.doca_flags = flags;
doca.doca_userfunc = func;
doca.doca_userarg = arg;
doca.doca_type = type;
return (dsl_sync_task(name,
dmu_objset_create_check, dmu_objset_create_sync, &doca,
5, ZFS_SPACE_CHECK_NORMAL));
}
typedef struct dmu_objset_clone_arg {
const char *doca_clone;
const char *doca_origin;
cred_t *doca_cred;
} dmu_objset_clone_arg_t;
/*ARGSUSED*/
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);
if (error != 0) {
dsl_dir_rele(pdd, FTAG);
return (SET_ERROR(EDQUOT));
}
dsl_dir_rele(pdd, FTAG);
error = dsl_dataset_hold(dp, doca->doca_origin, FTAG, &origin);
if (error != 0)
return (error);
/* You can only clone snapshots, not the head datasets. */
if (!origin->ds_is_snapshot) {
dsl_dataset_rele(origin, FTAG);
return (SET_ERROR(EINVAL));
}
dsl_dataset_rele(origin, 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, 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, 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();
return (dsl_sync_task(clone,
dmu_objset_clone_check, dmu_objset_clone_sync, &doca,
5, ZFS_SPACE_CHECK_NORMAL));
}
static int
dmu_objset_remap_indirects_impl(objset_t *os, uint64_t last_removed_txg)
{
int error = 0;
uint64_t object = 0;
while ((error = dmu_object_next(os, &object, B_FALSE, 0)) == 0) {
error = dmu_object_remap_indirects(os, object,
last_removed_txg);
/*
* If the ZPL removed the object before we managed to dnode_hold
* it, we would get an ENOENT. If the ZPL declares its intent
* to remove the object (dnode_free) before we manage to
* dnode_hold it, we would get an EEXIST. In either case, we
* want to continue remapping the other objects in the objset;
* in all other cases, we want to break early.
*/
if (error != 0 && error != ENOENT && error != EEXIST) {
break;
}
}
if (error == ESRCH) {
error = 0;
}
return (error);
}
int
dmu_objset_remap_indirects(const char *fsname)
{
int error = 0;
objset_t *os = NULL;
uint64_t last_removed_txg;
uint64_t remap_start_txg;
dsl_dir_t *dd;
error = dmu_objset_hold(fsname, FTAG, &os);
if (error != 0) {
return (error);
}
dd = dmu_objset_ds(os)->ds_dir;
if (!spa_feature_is_enabled(dmu_objset_spa(os),
SPA_FEATURE_OBSOLETE_COUNTS)) {
dmu_objset_rele(os, FTAG);
return (SET_ERROR(ENOTSUP));
}
if (dsl_dataset_is_snapshot(dmu_objset_ds(os))) {
dmu_objset_rele(os, FTAG);
return (SET_ERROR(EINVAL));
}
/*
* If there has not been a removal, we're done.
*/
last_removed_txg = spa_get_last_removal_txg(dmu_objset_spa(os));
if (last_removed_txg == -1ULL) {
dmu_objset_rele(os, FTAG);
return (0);
}
/*
* If we have remapped since the last removal, we're done.
*/
if (dsl_dir_is_zapified(dd)) {
uint64_t last_remap_txg;
if (zap_lookup(spa_meta_objset(dmu_objset_spa(os)),
dd->dd_object, DD_FIELD_LAST_REMAP_TXG,
sizeof (last_remap_txg), 1, &last_remap_txg) == 0 &&
last_remap_txg > last_removed_txg) {
dmu_objset_rele(os, FTAG);
return (0);
}
}
dsl_dataset_long_hold(dmu_objset_ds(os), FTAG);
dsl_pool_rele(dmu_objset_pool(os), FTAG);
remap_start_txg = spa_last_synced_txg(dmu_objset_spa(os));
error = dmu_objset_remap_indirects_impl(os, last_removed_txg);
if (error == 0) {
/*
* We update the last_remap_txg to be the start txg so that
* we can guarantee that every block older than last_remap_txg
* that can be remapped has been remapped.
*/
error = dsl_dir_update_last_remap_txg(dd, remap_start_txg);
}
dsl_dataset_long_rele(dmu_objset_ds(os), FTAG);
dsl_dataset_rele(dmu_objset_ds(os), FTAG);
return (error);
}
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);
strfree(longsnap);
err = dsl_dataset_snapshot(snaps, NULL, NULL);
fnvlist_free(snaps);
return (err);
}
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 ouside 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);
/*
* If we are not doing useraccounting (os_synced_dnodes == NULL)
* we are done with this dnode for this txg. Unset dn_dirty_txg
* if later txgs aren't dirtying it so that future holders do
* not get a stale value. Otherwise, we will do this in
* userquota_updates_task() when processing has completely
* finished for this txg.
*/
multilist_t *newlist = dn->dn_objset->os_synced_dnodes;
if (newlist != NULL) {
(void) dnode_add_ref(dn, newlist);
multilist_insert(newlist, dn);
} else {
mutex_enter(&dn->dn_mtx);
if (dn->dn_dirty_txg == tx->tx_txg)
dn->dn_dirty_txg = 0;
mutex_exit(&dn->dn_mtx);
}
dnode_sync(dn, tx);
}
}
/* ARGSUSED */
static void
dmu_objset_write_ready(zio_t *zio, arc_buf_t *abuf, void *arg)
{
blkptr_t *bp = zio->io_bp;
objset_t *os = arg;
dnode_phys_t *dnp = &os->os_phys->os_meta_dnode;
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 accounting objects).
*/
bp->blk_fill = 0;
for (int i = 0; i < dnp->dn_nblkptr; i++)
bp->blk_fill += BP_GET_FILL(&dnp->dn_blkptr[i]);
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);
}
/* ARGSUSED */
static void
dmu_objset_write_done(zio_t *zio, arc_buf_t *abuf, void *arg)
{
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(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", 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);
zio = arc_write(pio, os->os_spa, tx->tx_txg,
blkptr_copy, os->os_phys_buf, DMU_OS_IS_L2CACHEABLE(os),
&zp, dmu_objset_write_ready, NULL, 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);
}
txgoff = tx->tx_txg & TXG_MASK;
if (dmu_objset_userused_enabled(os)) {
/*
* 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 == NULL) {
os->os_synced_dnodes =
multilist_create(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_head(list)) != NULL) {
ASSERT0(dr->dr_dbuf->db_level);
list_remove(list, dr);
if (dr->dr_zio)
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 objset_used_cb_t *used_cbs[DMU_OST_NUMTYPES];
void
dmu_objset_register_type(dmu_objset_type_t ost, objset_used_cb_t *cb)
{
used_cbs[ost] = cb;
}
boolean_t
dmu_objset_userused_enabled(objset_t *os)
{
return (spa_version(os->os_spa) >= SPA_VERSION_USERSPACE &&
used_cbs[os->os_phys->os_type] != NULL &&
DMU_USERUSED_DNODE(os) != NULL);
}
typedef struct userquota_node {
uint64_t uqn_id;
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;
} 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;
if (luqn->uqn_id < ruqn->uqn_id)
return (-1);
if (luqn->uqn_id > ruqn->uqn_id)
return (1);
return (0);
}
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_int(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_int(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);
}
static void
userquota_update_cache(avl_tree_t *avl, uint64_t id, int64_t delta)
{
userquota_node_t search = { .uqn_id = id };
avl_index_t idx;
userquota_node_t *uqn = avl_find(avl, &search, &idx);
if (uqn == NULL) {
uqn = kmem_zalloc(sizeof (*uqn), KM_SLEEP);
uqn->uqn_id = id;
avl_insert(avl, uqn, idx);
}
uqn->uqn_delta += delta;
}
static void
do_userquota_update(userquota_cache_t *cache, uint64_t used, uint64_t flags,
uint64_t user, uint64_t group, boolean_t subtract)
{
if ((flags & DNODE_FLAG_USERUSED_ACCOUNTED)) {
int64_t delta = DNODE_MIN_SIZE + used;
if (subtract)
delta = -delta;
userquota_update_cache(&cache->uqc_user_deltas, user, delta);
userquota_update_cache(&cache->uqc_group_deltas, group, 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(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));
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(&cache,
dn->dn_oldused, dn->dn_oldflags,
dn->dn_olduid, dn->dn_oldgid, B_TRUE);
}
if (flags & DN_ID_NEW_EXIST) {
do_userquota_update(&cache,
DN_USED_BYTES(dn->dn_phys),
dn->dn_phys->dn_flags, dn->dn_newuid,
dn->dn_newgid, 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_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);
if (dn->dn_dirty_txg == spa_syncing_txg(os->os_spa))
dn->dn_dirty_txg = 0;
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));
}
void
dmu_objset_do_userquota_updates(objset_t *os, dmu_tx_t *tx)
{
int num_sublists;
if (!dmu_objset_userused_enabled(os))
return;
/* Allocate the user/groupused 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));
}
num_sublists = multilist_get_num_sublists(os->os_synced_dnodes);
for (int i = 0; i < num_sublists; i++) {
if (multilist_sublist_is_empty_idx(os->os_synced_dnodes, i))
continue;
userquota_updates_arg_t *uua =
kmem_alloc(sizeof (*uua), KM_SLEEP);
uua->uua_os = os;
uua->uua_sublist_idx = i;
uua->uua_tx = tx;
/* note: caller does taskq_wait() */
(void) taskq_dispatch(dmu_objset_pool(os)->dp_sync_taskq,
userquota_updates_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, **drp;
void *data;
if (db->db_dirtycnt == 0)
return (db->db.db_data); /* Nothing is changing */
for (drp = &db->db_last_dirty; (dr = *drp) != NULL; drp = &dr->dr_next)
if (dr->dr_txg == tx->tx_txg)
break;
if (dr == NULL) {
data = NULL;
} else {
dnode_t *dn;
DB_DNODE_ENTER(dr->dr_dbuf);
dn = DB_DNODE(dr->dr_dbuf);
if (dn->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;
DB_DNODE_EXIT(dr->dr_dbuf);
}
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;
uint64_t *user = NULL;
uint64_t *group = NULL;
int flags = dn->dn_id_flags;
int error;
boolean_t have_spill = B_FALSE;
if (!dmu_objset_userused_enabled(dn->dn_objset))
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;
}
if (before) {
ASSERT(data);
user = &dn->dn_olduid;
group = &dn->dn_oldgid;
} else if (data) {
user = &dn->dn_newuid;
group = &dn->dn_newgid;
}
/*
* Must always call the callback in case the object
* type has changed and that type isn't an object type to track
*/
error = used_cbs[os->os_phys->os_type](dn->dn_bonustype, data,
user, group);
/*
* 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;
} else {
dn->dn_newuid = 0;
dn->dn_newgid = 0;
}
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);
}
int
dmu_objset_userspace_upgrade(objset_t *os)
{
uint64_t obj;
int err = 0;
if (dmu_objset_userspace_present(os))
return (0);
if (!dmu_objset_userused_enabled(os))
return (SET_ERROR(ENOTSUP));
if (dmu_objset_is_snapshot(os))
return (SET_ERROR(EINVAL));
/*
* 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;
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_tx_abort(tx);
continue;
}
dmu_buf_will_dirty(db, tx);
dmu_buf_rele(db, FTAG);
dmu_tx_commit(tx);
}
os->os_flags |= OBJSET_FLAG_USERACCOUNTING_COMPLETE;
txg_wait_synced(dmu_objset_pool(os), 0);
return (0);
}
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, 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) strcpy(name, attr.za_name);
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_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) strcpy(name, attr.za_name);
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
* asssert(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, minclsyspri, 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);
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);
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(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);
}
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.
*/
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);
}
}
Index: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_send.c
===================================================================
--- head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_send.c (revision 359302)
+++ head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dmu_send.c (revision 359303)
@@ -1,3518 +1,3542 @@
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright 2011 Nexenta Systems, Inc. All rights reserved.
* Copyright (c) 2011, 2015 by Delphix. All rights reserved.
* Copyright (c) 2014, Joyent, Inc. All rights reserved.
* Copyright (c) 2012, Martin Matuska <mm@FreeBSD.org>. All rights reserved.
* Copyright 2014 HybridCluster. All rights reserved.
* Copyright 2016 RackTop Systems.
* Copyright (c) 2014 Integros [integros.com]
+ * Copyright (c) 2018, loli10K <ezomori.nozomu@gmail.com>. All rights reserved.
*/
#include <sys/dmu.h>
#include <sys/dmu_impl.h>
#include <sys/dmu_tx.h>
#include <sys/dbuf.h>
#include <sys/dnode.h>
#include <sys/zfs_context.h>
#include <sys/dmu_objset.h>
#include <sys/dmu_traverse.h>
#include <sys/dsl_dataset.h>
#include <sys/dsl_dir.h>
#include <sys/dsl_prop.h>
#include <sys/dsl_pool.h>
#include <sys/dsl_synctask.h>
#include <sys/zfs_ioctl.h>
#include <sys/zap.h>
#include <sys/zio_checksum.h>
#include <sys/zfs_znode.h>
#include <zfs_fletcher.h>
#include <sys/avl.h>
#include <sys/ddt.h>
#include <sys/zfs_onexit.h>
#include <sys/dmu_send.h>
#include <sys/dsl_destroy.h>
#include <sys/blkptr.h>
#include <sys/dsl_bookmark.h>
#include <sys/zfeature.h>
#include <sys/bqueue.h>
#ifdef __FreeBSD__
#undef dump_write
#define dump_write dmu_dump_write
#endif
/* Set this tunable to TRUE to replace corrupt data with 0x2f5baddb10c */
int zfs_send_corrupt_data = B_FALSE;
int zfs_send_queue_length = 16 * 1024 * 1024;
int zfs_recv_queue_length = 16 * 1024 * 1024;
/* Set this tunable to FALSE to disable setting of DRR_FLAG_FREERECORDS */
int zfs_send_set_freerecords_bit = B_TRUE;
#ifdef _KERNEL
TUNABLE_INT("vfs.zfs.send_set_freerecords_bit", &zfs_send_set_freerecords_bit);
#endif
static char *dmu_recv_tag = "dmu_recv_tag";
const char *recv_clone_name = "%recv";
/*
* Use this to override the recordsize calculation for fast zfs send estimates.
*/
uint64_t zfs_override_estimate_recordsize = 0;
#define BP_SPAN(datablkszsec, indblkshift, level) \
(((uint64_t)datablkszsec) << (SPA_MINBLOCKSHIFT + \
(level) * (indblkshift - SPA_BLKPTRSHIFT)))
static void byteswap_record(dmu_replay_record_t *drr);
struct send_thread_arg {
bqueue_t q;
dsl_dataset_t *ds; /* Dataset to traverse */
uint64_t fromtxg; /* Traverse from this txg */
int flags; /* flags to pass to traverse_dataset */
int error_code;
boolean_t cancel;
zbookmark_phys_t resume;
};
struct send_block_record {
boolean_t eos_marker; /* Marks the end of the stream */
blkptr_t bp;
zbookmark_phys_t zb;
uint8_t indblkshift;
uint16_t datablkszsec;
bqueue_node_t ln;
};
static int
dump_bytes(dmu_sendarg_t *dsp, void *buf, int len)
{
dsl_dataset_t *ds = dmu_objset_ds(dsp->dsa_os);
struct uio auio;
struct iovec aiov;
/*
* The code does not rely on this (len being a multiple of 8). We keep
* this assertion because of the corresponding assertion in
* receive_read(). Keeping this assertion ensures that we do not
* inadvertently break backwards compatibility (causing the assertion
* in receive_read() to trigger on old software).
*
* Removing the assertions could be rolled into a new feature that uses
* data that isn't 8-byte aligned; if the assertions were removed, a
* feature flag would have to be added.
*/
ASSERT0(len % 8);
aiov.iov_base = buf;
aiov.iov_len = len;
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_resid = len;
auio.uio_segflg = UIO_SYSSPACE;
auio.uio_rw = UIO_WRITE;
auio.uio_offset = (off_t)-1;
auio.uio_td = dsp->dsa_td;
#ifdef _KERNEL
if (dsp->dsa_fp->f_type == DTYPE_VNODE)
bwillwrite();
dsp->dsa_err = fo_write(dsp->dsa_fp, &auio, dsp->dsa_td->td_ucred, 0,
dsp->dsa_td);
#else
fprintf(stderr, "%s: returning EOPNOTSUPP\n", __func__);
dsp->dsa_err = EOPNOTSUPP;
#endif
mutex_enter(&ds->ds_sendstream_lock);
*dsp->dsa_off += len;
mutex_exit(&ds->ds_sendstream_lock);
return (dsp->dsa_err);
}
/*
* For all record types except BEGIN, fill in the checksum (overlaid in
* drr_u.drr_checksum.drr_checksum). The checksum verifies everything
* up to the start of the checksum itself.
*/
static int
dump_record(dmu_sendarg_t *dsp, void *payload, int payload_len)
{
ASSERT3U(offsetof(dmu_replay_record_t, drr_u.drr_checksum.drr_checksum),
==, sizeof (dmu_replay_record_t) - sizeof (zio_cksum_t));
(void) fletcher_4_incremental_native(dsp->dsa_drr,
offsetof(dmu_replay_record_t, drr_u.drr_checksum.drr_checksum),
&dsp->dsa_zc);
if (dsp->dsa_drr->drr_type == DRR_BEGIN) {
dsp->dsa_sent_begin = B_TRUE;
} else {
ASSERT(ZIO_CHECKSUM_IS_ZERO(&dsp->dsa_drr->drr_u.
drr_checksum.drr_checksum));
dsp->dsa_drr->drr_u.drr_checksum.drr_checksum = dsp->dsa_zc;
}
if (dsp->dsa_drr->drr_type == DRR_END) {
dsp->dsa_sent_end = B_TRUE;
}
(void) fletcher_4_incremental_native(&dsp->dsa_drr->
drr_u.drr_checksum.drr_checksum,
sizeof (zio_cksum_t), &dsp->dsa_zc);
if (dump_bytes(dsp, dsp->dsa_drr, sizeof (dmu_replay_record_t)) != 0)
return (SET_ERROR(EINTR));
if (payload_len != 0) {
(void) fletcher_4_incremental_native(payload, payload_len,
&dsp->dsa_zc);
if (dump_bytes(dsp, payload, payload_len) != 0)
return (SET_ERROR(EINTR));
}
return (0);
}
/*
* Fill in the drr_free struct, or perform aggregation if the previous record is
* also a free record, and the two are adjacent.
*
* Note that we send free records even for a full send, because we want to be
* able to receive a full send as a clone, which requires a list of all the free
* and freeobject records that were generated on the source.
*/
static int
dump_free(dmu_sendarg_t *dsp, uint64_t object, uint64_t offset,
uint64_t length)
{
struct drr_free *drrf = &(dsp->dsa_drr->drr_u.drr_free);
/*
* When we receive a free record, dbuf_free_range() assumes
* that the receiving system doesn't have any dbufs in the range
* being freed. This is always true because there is a one-record
* constraint: we only send one WRITE record for any given
* object,offset. We know that the one-record constraint is
* true because we always send data in increasing order by
* object,offset.
*
* If the increasing-order constraint ever changes, we should find
* another way to assert that the one-record constraint is still
* satisfied.
*/
ASSERT(object > dsp->dsa_last_data_object ||
(object == dsp->dsa_last_data_object &&
offset > dsp->dsa_last_data_offset));
if (length != -1ULL && offset + length < offset)
length = -1ULL;
/*
* If there is a pending op, but it's not PENDING_FREE, push it out,
* since free block aggregation can only be done for blocks of the
* same type (i.e., DRR_FREE records can only be aggregated with
* other DRR_FREE records. DRR_FREEOBJECTS records can only be
* aggregated with other DRR_FREEOBJECTS records.
*/
if (dsp->dsa_pending_op != PENDING_NONE &&
dsp->dsa_pending_op != PENDING_FREE) {
if (dump_record(dsp, NULL, 0) != 0)
return (SET_ERROR(EINTR));
dsp->dsa_pending_op = PENDING_NONE;
}
if (dsp->dsa_pending_op == PENDING_FREE) {
/*
* There should never be a PENDING_FREE if length is -1
* (because dump_dnode is the only place where this
* function is called with a -1, and only after flushing
* any pending record).
*/
ASSERT(length != -1ULL);
/*
* Check to see whether this free block can be aggregated
* with pending one.
*/
if (drrf->drr_object == object && drrf->drr_offset +
drrf->drr_length == offset) {
drrf->drr_length += length;
return (0);
} else {
/* not a continuation. Push out pending record */
if (dump_record(dsp, NULL, 0) != 0)
return (SET_ERROR(EINTR));
dsp->dsa_pending_op = PENDING_NONE;
}
}
/* create a FREE record and make it pending */
bzero(dsp->dsa_drr, sizeof (dmu_replay_record_t));
dsp->dsa_drr->drr_type = DRR_FREE;
drrf->drr_object = object;
drrf->drr_offset = offset;
drrf->drr_length = length;
drrf->drr_toguid = dsp->dsa_toguid;
if (length == -1ULL) {
if (dump_record(dsp, NULL, 0) != 0)
return (SET_ERROR(EINTR));
} else {
dsp->dsa_pending_op = PENDING_FREE;
}
return (0);
}
static int
dump_write(dmu_sendarg_t *dsp, dmu_object_type_t type,
uint64_t object, uint64_t offset, int lsize, int psize, const blkptr_t *bp,
void *data)
{
uint64_t payload_size;
struct drr_write *drrw = &(dsp->dsa_drr->drr_u.drr_write);
/*
* We send data in increasing object, offset order.
* See comment in dump_free() for details.
*/
ASSERT(object > dsp->dsa_last_data_object ||
(object == dsp->dsa_last_data_object &&
offset > dsp->dsa_last_data_offset));
dsp->dsa_last_data_object = object;
dsp->dsa_last_data_offset = offset + lsize - 1;
/*
* If there is any kind of pending aggregation (currently either
* a grouping of free objects or free blocks), push it out to
* the stream, since aggregation can't be done across operations
* of different types.
*/
if (dsp->dsa_pending_op != PENDING_NONE) {
if (dump_record(dsp, NULL, 0) != 0)
return (SET_ERROR(EINTR));
dsp->dsa_pending_op = PENDING_NONE;
}
/* write a WRITE record */
bzero(dsp->dsa_drr, sizeof (dmu_replay_record_t));
dsp->dsa_drr->drr_type = DRR_WRITE;
drrw->drr_object = object;
drrw->drr_type = type;
drrw->drr_offset = offset;
drrw->drr_toguid = dsp->dsa_toguid;
drrw->drr_logical_size = lsize;
/* only set the compression fields if the buf is compressed */
if (lsize != psize) {
ASSERT(dsp->dsa_featureflags & DMU_BACKUP_FEATURE_COMPRESSED);
ASSERT(!BP_IS_EMBEDDED(bp));
ASSERT(!BP_SHOULD_BYTESWAP(bp));
ASSERT(!DMU_OT_IS_METADATA(BP_GET_TYPE(bp)));
ASSERT3U(BP_GET_COMPRESS(bp), !=, ZIO_COMPRESS_OFF);
ASSERT3S(psize, >, 0);
ASSERT3S(lsize, >=, psize);
drrw->drr_compressiontype = BP_GET_COMPRESS(bp);
drrw->drr_compressed_size = psize;
payload_size = drrw->drr_compressed_size;
} else {
payload_size = drrw->drr_logical_size;
}
if (bp == NULL || BP_IS_EMBEDDED(bp)) {
/*
* There's no pre-computed checksum for partial-block
* writes or embedded BP's, so (like
* fletcher4-checkummed blocks) userland will have to
* compute a dedup-capable checksum itself.
*/
drrw->drr_checksumtype = ZIO_CHECKSUM_OFF;
} else {
drrw->drr_checksumtype = BP_GET_CHECKSUM(bp);
if (zio_checksum_table[drrw->drr_checksumtype].ci_flags &
ZCHECKSUM_FLAG_DEDUP)
drrw->drr_checksumflags |= DRR_CHECKSUM_DEDUP;
DDK_SET_LSIZE(&drrw->drr_key, BP_GET_LSIZE(bp));
DDK_SET_PSIZE(&drrw->drr_key, BP_GET_PSIZE(bp));
DDK_SET_COMPRESS(&drrw->drr_key, BP_GET_COMPRESS(bp));
drrw->drr_key.ddk_cksum = bp->blk_cksum;
}
if (dump_record(dsp, data, payload_size) != 0)
return (SET_ERROR(EINTR));
return (0);
}
static int
dump_write_embedded(dmu_sendarg_t *dsp, uint64_t object, uint64_t offset,
int blksz, const blkptr_t *bp)
{
char buf[BPE_PAYLOAD_SIZE];
struct drr_write_embedded *drrw =
&(dsp->dsa_drr->drr_u.drr_write_embedded);
if (dsp->dsa_pending_op != PENDING_NONE) {
if (dump_record(dsp, NULL, 0) != 0)
return (EINTR);
dsp->dsa_pending_op = PENDING_NONE;
}
ASSERT(BP_IS_EMBEDDED(bp));
bzero(dsp->dsa_drr, sizeof (dmu_replay_record_t));
dsp->dsa_drr->drr_type = DRR_WRITE_EMBEDDED;
drrw->drr_object = object;
drrw->drr_offset = offset;
drrw->drr_length = blksz;
drrw->drr_toguid = dsp->dsa_toguid;
drrw->drr_compression = BP_GET_COMPRESS(bp);
drrw->drr_etype = BPE_GET_ETYPE(bp);
drrw->drr_lsize = BPE_GET_LSIZE(bp);
drrw->drr_psize = BPE_GET_PSIZE(bp);
decode_embedded_bp_compressed(bp, buf);
if (dump_record(dsp, buf, P2ROUNDUP(drrw->drr_psize, 8)) != 0)
return (EINTR);
return (0);
}
static int
dump_spill(dmu_sendarg_t *dsp, uint64_t object, int blksz, void *data)
{
struct drr_spill *drrs = &(dsp->dsa_drr->drr_u.drr_spill);
if (dsp->dsa_pending_op != PENDING_NONE) {
if (dump_record(dsp, NULL, 0) != 0)
return (SET_ERROR(EINTR));
dsp->dsa_pending_op = PENDING_NONE;
}
/* write a SPILL record */
bzero(dsp->dsa_drr, sizeof (dmu_replay_record_t));
dsp->dsa_drr->drr_type = DRR_SPILL;
drrs->drr_object = object;
drrs->drr_length = blksz;
drrs->drr_toguid = dsp->dsa_toguid;
if (dump_record(dsp, data, blksz) != 0)
return (SET_ERROR(EINTR));
return (0);
}
static int
dump_freeobjects(dmu_sendarg_t *dsp, uint64_t firstobj, uint64_t numobjs)
{
struct drr_freeobjects *drrfo = &(dsp->dsa_drr->drr_u.drr_freeobjects);
/*
* If there is a pending op, but it's not PENDING_FREEOBJECTS,
* push it out, since free block aggregation can only be done for
* blocks of the same type (i.e., DRR_FREE records can only be
* aggregated with other DRR_FREE records. DRR_FREEOBJECTS records
* can only be aggregated with other DRR_FREEOBJECTS records.
*/
if (dsp->dsa_pending_op != PENDING_NONE &&
dsp->dsa_pending_op != PENDING_FREEOBJECTS) {
if (dump_record(dsp, NULL, 0) != 0)
return (SET_ERROR(EINTR));
dsp->dsa_pending_op = PENDING_NONE;
}
if (dsp->dsa_pending_op == PENDING_FREEOBJECTS) {
/*
* See whether this free object array can be aggregated
* with pending one
*/
if (drrfo->drr_firstobj + drrfo->drr_numobjs == firstobj) {
drrfo->drr_numobjs += numobjs;
return (0);
} else {
/* can't be aggregated. Push out pending record */
if (dump_record(dsp, NULL, 0) != 0)
return (SET_ERROR(EINTR));
dsp->dsa_pending_op = PENDING_NONE;
}
}
/* write a FREEOBJECTS record */
bzero(dsp->dsa_drr, sizeof (dmu_replay_record_t));
dsp->dsa_drr->drr_type = DRR_FREEOBJECTS;
drrfo->drr_firstobj = firstobj;
drrfo->drr_numobjs = numobjs;
drrfo->drr_toguid = dsp->dsa_toguid;
dsp->dsa_pending_op = PENDING_FREEOBJECTS;
return (0);
}
static int
dump_dnode(dmu_sendarg_t *dsp, uint64_t object, dnode_phys_t *dnp)
{
struct drr_object *drro = &(dsp->dsa_drr->drr_u.drr_object);
if (object < dsp->dsa_resume_object) {
/*
* Note: when resuming, we will visit all the dnodes in
* the block of dnodes that we are resuming from. In
* this case it's unnecessary to send the dnodes prior to
* the one we are resuming from. We should be at most one
* block's worth of dnodes behind the resume point.
*/
ASSERT3U(dsp->dsa_resume_object - object, <,
1 << (DNODE_BLOCK_SHIFT - DNODE_SHIFT));
return (0);
}
if (dnp == NULL || dnp->dn_type == DMU_OT_NONE)
return (dump_freeobjects(dsp, object, 1));
if (dsp->dsa_pending_op != PENDING_NONE) {
if (dump_record(dsp, NULL, 0) != 0)
return (SET_ERROR(EINTR));
dsp->dsa_pending_op = PENDING_NONE;
}
/* write an OBJECT record */
bzero(dsp->dsa_drr, sizeof (dmu_replay_record_t));
dsp->dsa_drr->drr_type = DRR_OBJECT;
drro->drr_object = object;
drro->drr_type = dnp->dn_type;
drro->drr_bonustype = dnp->dn_bonustype;
drro->drr_blksz = dnp->dn_datablkszsec << SPA_MINBLOCKSHIFT;
drro->drr_bonuslen = dnp->dn_bonuslen;
drro->drr_dn_slots = dnp->dn_extra_slots + 1;
drro->drr_checksumtype = dnp->dn_checksum;
drro->drr_compress = dnp->dn_compress;
drro->drr_toguid = dsp->dsa_toguid;
if (!(dsp->dsa_featureflags & DMU_BACKUP_FEATURE_LARGE_BLOCKS) &&
drro->drr_blksz > SPA_OLD_MAXBLOCKSIZE)
drro->drr_blksz = SPA_OLD_MAXBLOCKSIZE;
if (dump_record(dsp, DN_BONUS(dnp),
P2ROUNDUP(dnp->dn_bonuslen, 8)) != 0) {
return (SET_ERROR(EINTR));
}
/* Free anything past the end of the file. */
if (dump_free(dsp, object, (dnp->dn_maxblkid + 1) *
(dnp->dn_datablkszsec << SPA_MINBLOCKSHIFT), -1ULL) != 0)
return (SET_ERROR(EINTR));
if (dsp->dsa_err != 0)
return (SET_ERROR(EINTR));
return (0);
}
static boolean_t
backup_do_embed(dmu_sendarg_t *dsp, const blkptr_t *bp)
{
if (!BP_IS_EMBEDDED(bp))
return (B_FALSE);
/*
* Compression function must be legacy, or explicitly enabled.
*/
if ((BP_GET_COMPRESS(bp) >= ZIO_COMPRESS_LEGACY_FUNCTIONS &&
!(dsp->dsa_featureflags & DMU_BACKUP_FEATURE_LZ4)))
return (B_FALSE);
/*
* Embed type must be explicitly enabled.
*/
switch (BPE_GET_ETYPE(bp)) {
case BP_EMBEDDED_TYPE_DATA:
if (dsp->dsa_featureflags & DMU_BACKUP_FEATURE_EMBED_DATA)
return (B_TRUE);
break;
default:
return (B_FALSE);
}
return (B_FALSE);
}
/*
* This is the callback function to traverse_dataset that acts as the worker
* thread for dmu_send_impl.
*/
/*ARGSUSED*/
static int
send_cb(spa_t *spa, zilog_t *zilog, const blkptr_t *bp,
const zbookmark_phys_t *zb, const struct dnode_phys *dnp, void *arg)
{
struct send_thread_arg *sta = arg;
struct send_block_record *record;
uint64_t record_size;
int err = 0;
ASSERT(zb->zb_object == DMU_META_DNODE_OBJECT ||
zb->zb_object >= sta->resume.zb_object);
if (sta->cancel)
return (SET_ERROR(EINTR));
if (bp == NULL) {
ASSERT3U(zb->zb_level, ==, ZB_DNODE_LEVEL);
return (0);
} else if (zb->zb_level < 0) {
return (0);
}
record = kmem_zalloc(sizeof (struct send_block_record), KM_SLEEP);
record->eos_marker = B_FALSE;
record->bp = *bp;
record->zb = *zb;
record->indblkshift = dnp->dn_indblkshift;
record->datablkszsec = dnp->dn_datablkszsec;
record_size = dnp->dn_datablkszsec << SPA_MINBLOCKSHIFT;
bqueue_enqueue(&sta->q, record, record_size);
return (err);
}
/*
* This function kicks off the traverse_dataset. It also handles setting the
* error code of the thread in case something goes wrong, and pushes the End of
* Stream record when the traverse_dataset call has finished. If there is no
* dataset to traverse, the thread immediately pushes End of Stream marker.
*/
static void
send_traverse_thread(void *arg)
{
struct send_thread_arg *st_arg = arg;
int err;
struct send_block_record *data;
if (st_arg->ds != NULL) {
err = traverse_dataset_resume(st_arg->ds,
st_arg->fromtxg, &st_arg->resume,
st_arg->flags, send_cb, st_arg);
if (err != EINTR)
st_arg->error_code = err;
}
data = kmem_zalloc(sizeof (*data), KM_SLEEP);
data->eos_marker = B_TRUE;
bqueue_enqueue(&st_arg->q, data, 1);
thread_exit();
}
/*
* This function actually handles figuring out what kind of record needs to be
* dumped, reading the data (which has hopefully been prefetched), and calling
* the appropriate helper function.
*/
static int
do_dump(dmu_sendarg_t *dsa, struct send_block_record *data)
{
dsl_dataset_t *ds = dmu_objset_ds(dsa->dsa_os);
const blkptr_t *bp = &data->bp;
const zbookmark_phys_t *zb = &data->zb;
uint8_t indblkshift = data->indblkshift;
uint16_t dblkszsec = data->datablkszsec;
spa_t *spa = ds->ds_dir->dd_pool->dp_spa;
dmu_object_type_t type = bp ? BP_GET_TYPE(bp) : DMU_OT_NONE;
int err = 0;
ASSERT3U(zb->zb_level, >=, 0);
ASSERT(zb->zb_object == DMU_META_DNODE_OBJECT ||
zb->zb_object >= dsa->dsa_resume_object);
if (zb->zb_object != DMU_META_DNODE_OBJECT &&
DMU_OBJECT_IS_SPECIAL(zb->zb_object)) {
return (0);
} else if (BP_IS_HOLE(bp) &&
zb->zb_object == DMU_META_DNODE_OBJECT) {
uint64_t span = BP_SPAN(dblkszsec, indblkshift, zb->zb_level);
uint64_t dnobj = (zb->zb_blkid * span) >> DNODE_SHIFT;
err = dump_freeobjects(dsa, dnobj, span >> DNODE_SHIFT);
} else if (BP_IS_HOLE(bp)) {
uint64_t span = BP_SPAN(dblkszsec, indblkshift, zb->zb_level);
uint64_t offset = zb->zb_blkid * span;
err = dump_free(dsa, zb->zb_object, offset, span);
} else if (zb->zb_level > 0 || type == DMU_OT_OBJSET) {
return (0);
} else if (type == DMU_OT_DNODE) {
int epb = BP_GET_LSIZE(bp) >> DNODE_SHIFT;
arc_flags_t aflags = ARC_FLAG_WAIT;
arc_buf_t *abuf;
ASSERT0(zb->zb_level);
if (arc_read(NULL, spa, bp, arc_getbuf_func, &abuf,
ZIO_PRIORITY_ASYNC_READ, ZIO_FLAG_CANFAIL,
&aflags, zb) != 0)
return (SET_ERROR(EIO));
dnode_phys_t *blk = abuf->b_data;
uint64_t dnobj = zb->zb_blkid * epb;
for (int i = 0; i < epb; i += blk[i].dn_extra_slots + 1) {
err = dump_dnode(dsa, dnobj + i, blk + i);
if (err != 0)
break;
}
arc_buf_destroy(abuf, &abuf);
} else if (type == DMU_OT_SA) {
arc_flags_t aflags = ARC_FLAG_WAIT;
arc_buf_t *abuf;
int blksz = BP_GET_LSIZE(bp);
if (arc_read(NULL, spa, bp, arc_getbuf_func, &abuf,
ZIO_PRIORITY_ASYNC_READ, ZIO_FLAG_CANFAIL,
&aflags, zb) != 0)
return (SET_ERROR(EIO));
err = dump_spill(dsa, zb->zb_object, blksz, abuf->b_data);
arc_buf_destroy(abuf, &abuf);
} else if (backup_do_embed(dsa, bp)) {
/* it's an embedded level-0 block of a regular object */
int blksz = dblkszsec << SPA_MINBLOCKSHIFT;
ASSERT0(zb->zb_level);
err = dump_write_embedded(dsa, zb->zb_object,
zb->zb_blkid * blksz, blksz, bp);
} else {
/* it's a level-0 block of a regular object */
arc_flags_t aflags = ARC_FLAG_WAIT;
arc_buf_t *abuf;
int blksz = dblkszsec << SPA_MINBLOCKSHIFT;
uint64_t offset;
/*
* If we have large blocks stored on disk but the send flags
* don't allow us to send large blocks, we split the data from
* the arc buf into chunks.
*/
boolean_t split_large_blocks = blksz > SPA_OLD_MAXBLOCKSIZE &&
!(dsa->dsa_featureflags & DMU_BACKUP_FEATURE_LARGE_BLOCKS);
/*
* We should only request compressed data from the ARC if all
* the following are true:
* - stream compression was requested
* - we aren't splitting large blocks into smaller chunks
* - the data won't need to be byteswapped before sending
* - this isn't an embedded block
* - this isn't metadata (if receiving on a different endian
* system it can be byteswapped more easily)
*/
boolean_t request_compressed =
(dsa->dsa_featureflags & DMU_BACKUP_FEATURE_COMPRESSED) &&
!split_large_blocks && !BP_SHOULD_BYTESWAP(bp) &&
!BP_IS_EMBEDDED(bp) && !DMU_OT_IS_METADATA(BP_GET_TYPE(bp));
ASSERT0(zb->zb_level);
ASSERT(zb->zb_object > dsa->dsa_resume_object ||
(zb->zb_object == dsa->dsa_resume_object &&
zb->zb_blkid * blksz >= dsa->dsa_resume_offset));
ASSERT0(zb->zb_level);
ASSERT(zb->zb_object > dsa->dsa_resume_object ||
(zb->zb_object == dsa->dsa_resume_object &&
zb->zb_blkid * blksz >= dsa->dsa_resume_offset));
ASSERT3U(blksz, ==, BP_GET_LSIZE(bp));
enum zio_flag zioflags = ZIO_FLAG_CANFAIL;
if (request_compressed)
zioflags |= ZIO_FLAG_RAW;
if (arc_read(NULL, spa, bp, arc_getbuf_func, &abuf,
ZIO_PRIORITY_ASYNC_READ, zioflags, &aflags, zb) != 0) {
if (zfs_send_corrupt_data) {
/* Send a block filled with 0x"zfs badd bloc" */
abuf = arc_alloc_buf(spa, &abuf, ARC_BUFC_DATA,
blksz);
uint64_t *ptr;
for (ptr = abuf->b_data;
(char *)ptr < (char *)abuf->b_data + blksz;
ptr++)
*ptr = 0x2f5baddb10cULL;
} else {
return (SET_ERROR(EIO));
}
}
offset = zb->zb_blkid * blksz;
if (split_large_blocks) {
ASSERT3U(arc_get_compression(abuf), ==,
ZIO_COMPRESS_OFF);
char *buf = abuf->b_data;
while (blksz > 0 && err == 0) {
int n = MIN(blksz, SPA_OLD_MAXBLOCKSIZE);
err = dump_write(dsa, type, zb->zb_object,
offset, n, n, NULL, buf);
offset += n;
buf += n;
blksz -= n;
}
} else {
err = dump_write(dsa, type, zb->zb_object, offset,
blksz, arc_buf_size(abuf), bp, abuf->b_data);
}
arc_buf_destroy(abuf, &abuf);
}
ASSERT(err == 0 || err == EINTR);
return (err);
}
/*
* Pop the new data off the queue, and free the old data.
*/
static struct send_block_record *
get_next_record(bqueue_t *bq, struct send_block_record *data)
{
struct send_block_record *tmp = bqueue_dequeue(bq);
kmem_free(data, sizeof (*data));
return (tmp);
}
/*
* Actually do the bulk of the work in a zfs send.
*
* Note: Releases dp using the specified tag.
*/
static int
dmu_send_impl(void *tag, dsl_pool_t *dp, dsl_dataset_t *to_ds,
zfs_bookmark_phys_t *ancestor_zb, boolean_t is_clone,
boolean_t embedok, boolean_t large_block_ok, boolean_t compressok,
int outfd, uint64_t resumeobj, uint64_t resumeoff,
#ifdef illumos
vnode_t *vp, offset_t *off)
#else
struct file *fp, offset_t *off)
#endif
{
objset_t *os;
dmu_replay_record_t *drr;
dmu_sendarg_t *dsp;
int err;
uint64_t fromtxg = 0;
uint64_t featureflags = 0;
struct send_thread_arg to_arg = { 0 };
err = dmu_objset_from_ds(to_ds, &os);
if (err != 0) {
dsl_pool_rele(dp, tag);
return (err);
}
drr = kmem_zalloc(sizeof (dmu_replay_record_t), KM_SLEEP);
drr->drr_type = DRR_BEGIN;
drr->drr_u.drr_begin.drr_magic = DMU_BACKUP_MAGIC;
DMU_SET_STREAM_HDRTYPE(drr->drr_u.drr_begin.drr_versioninfo,
DMU_SUBSTREAM);
#ifdef _KERNEL
if (dmu_objset_type(os) == DMU_OST_ZFS) {
uint64_t version;
if (zfs_get_zplprop(os, ZFS_PROP_VERSION, &version) != 0) {
kmem_free(drr, sizeof (dmu_replay_record_t));
dsl_pool_rele(dp, tag);
return (SET_ERROR(EINVAL));
}
if (version >= ZPL_VERSION_SA) {
featureflags |= DMU_BACKUP_FEATURE_SA_SPILL;
}
}
#endif
if (large_block_ok && to_ds->ds_feature_inuse[SPA_FEATURE_LARGE_BLOCKS])
featureflags |= DMU_BACKUP_FEATURE_LARGE_BLOCKS;
if (to_ds->ds_feature_inuse[SPA_FEATURE_LARGE_DNODE])
featureflags |= DMU_BACKUP_FEATURE_LARGE_DNODE;
if (embedok &&
spa_feature_is_active(dp->dp_spa, SPA_FEATURE_EMBEDDED_DATA)) {
featureflags |= DMU_BACKUP_FEATURE_EMBED_DATA;
if (spa_feature_is_active(dp->dp_spa, SPA_FEATURE_LZ4_COMPRESS))
featureflags |= DMU_BACKUP_FEATURE_LZ4;
}
if (compressok) {
featureflags |= DMU_BACKUP_FEATURE_COMPRESSED;
}
if ((featureflags &
(DMU_BACKUP_FEATURE_EMBED_DATA | DMU_BACKUP_FEATURE_COMPRESSED)) !=
0 && spa_feature_is_active(dp->dp_spa, SPA_FEATURE_LZ4_COMPRESS)) {
featureflags |= DMU_BACKUP_FEATURE_LZ4;
}
if (resumeobj != 0 || resumeoff != 0) {
featureflags |= DMU_BACKUP_FEATURE_RESUMING;
}
DMU_SET_FEATUREFLAGS(drr->drr_u.drr_begin.drr_versioninfo,
featureflags);
drr->drr_u.drr_begin.drr_creation_time =
dsl_dataset_phys(to_ds)->ds_creation_time;
drr->drr_u.drr_begin.drr_type = dmu_objset_type(os);
if (is_clone)
drr->drr_u.drr_begin.drr_flags |= DRR_FLAG_CLONE;
drr->drr_u.drr_begin.drr_toguid = dsl_dataset_phys(to_ds)->ds_guid;
if (dsl_dataset_phys(to_ds)->ds_flags & DS_FLAG_CI_DATASET)
drr->drr_u.drr_begin.drr_flags |= DRR_FLAG_CI_DATA;
if (zfs_send_set_freerecords_bit)
drr->drr_u.drr_begin.drr_flags |= DRR_FLAG_FREERECORDS;
if (ancestor_zb != NULL) {
drr->drr_u.drr_begin.drr_fromguid =
ancestor_zb->zbm_guid;
fromtxg = ancestor_zb->zbm_creation_txg;
}
dsl_dataset_name(to_ds, drr->drr_u.drr_begin.drr_toname);
if (!to_ds->ds_is_snapshot) {
(void) strlcat(drr->drr_u.drr_begin.drr_toname, "@--head--",
sizeof (drr->drr_u.drr_begin.drr_toname));
}
dsp = kmem_zalloc(sizeof (dmu_sendarg_t), KM_SLEEP);
dsp->dsa_drr = drr;
dsp->dsa_outfd = outfd;
dsp->dsa_proc = curproc;
dsp->dsa_td = curthread;
dsp->dsa_fp = fp;
dsp->dsa_os = os;
dsp->dsa_off = off;
dsp->dsa_toguid = dsl_dataset_phys(to_ds)->ds_guid;
dsp->dsa_pending_op = PENDING_NONE;
dsp->dsa_featureflags = featureflags;
dsp->dsa_resume_object = resumeobj;
dsp->dsa_resume_offset = resumeoff;
mutex_enter(&to_ds->ds_sendstream_lock);
list_insert_head(&to_ds->ds_sendstreams, dsp);
mutex_exit(&to_ds->ds_sendstream_lock);
dsl_dataset_long_hold(to_ds, FTAG);
dsl_pool_rele(dp, tag);
void *payload = NULL;
size_t payload_len = 0;
if (resumeobj != 0 || resumeoff != 0) {
dmu_object_info_t to_doi;
err = dmu_object_info(os, resumeobj, &to_doi);
if (err != 0)
goto out;
SET_BOOKMARK(&to_arg.resume, to_ds->ds_object, resumeobj, 0,
resumeoff / to_doi.doi_data_block_size);
nvlist_t *nvl = fnvlist_alloc();
fnvlist_add_uint64(nvl, "resume_object", resumeobj);
fnvlist_add_uint64(nvl, "resume_offset", resumeoff);
payload = fnvlist_pack(nvl, &payload_len);
drr->drr_payloadlen = payload_len;
fnvlist_free(nvl);
}
err = dump_record(dsp, payload, payload_len);
fnvlist_pack_free(payload, payload_len);
if (err != 0) {
err = dsp->dsa_err;
goto out;
}
err = bqueue_init(&to_arg.q, zfs_send_queue_length,
offsetof(struct send_block_record, ln));
to_arg.error_code = 0;
to_arg.cancel = B_FALSE;
to_arg.ds = to_ds;
to_arg.fromtxg = fromtxg;
to_arg.flags = TRAVERSE_PRE | TRAVERSE_PREFETCH;
(void) thread_create(NULL, 0, send_traverse_thread, &to_arg, 0, &p0,
TS_RUN, minclsyspri);
struct send_block_record *to_data;
to_data = bqueue_dequeue(&to_arg.q);
while (!to_data->eos_marker && err == 0) {
err = do_dump(dsp, to_data);
to_data = get_next_record(&to_arg.q, to_data);
if (issig(JUSTLOOKING) && issig(FORREAL))
err = EINTR;
}
if (err != 0) {
to_arg.cancel = B_TRUE;
while (!to_data->eos_marker) {
to_data = get_next_record(&to_arg.q, to_data);
}
}
kmem_free(to_data, sizeof (*to_data));
bqueue_destroy(&to_arg.q);
if (err == 0 && to_arg.error_code != 0)
err = to_arg.error_code;
if (err != 0)
goto out;
if (dsp->dsa_pending_op != PENDING_NONE)
if (dump_record(dsp, NULL, 0) != 0)
err = SET_ERROR(EINTR);
if (err != 0) {
if (err == EINTR && dsp->dsa_err != 0)
err = dsp->dsa_err;
goto out;
}
bzero(drr, sizeof (dmu_replay_record_t));
drr->drr_type = DRR_END;
drr->drr_u.drr_end.drr_checksum = dsp->dsa_zc;
drr->drr_u.drr_end.drr_toguid = dsp->dsa_toguid;
if (dump_record(dsp, NULL, 0) != 0)
err = dsp->dsa_err;
out:
mutex_enter(&to_ds->ds_sendstream_lock);
list_remove(&to_ds->ds_sendstreams, dsp);
mutex_exit(&to_ds->ds_sendstream_lock);
VERIFY(err != 0 || (dsp->dsa_sent_begin && dsp->dsa_sent_end));
kmem_free(drr, sizeof (dmu_replay_record_t));
kmem_free(dsp, sizeof (dmu_sendarg_t));
dsl_dataset_long_rele(to_ds, FTAG);
return (err);
}
int
dmu_send_obj(const char *pool, uint64_t tosnap, uint64_t fromsnap,
boolean_t embedok, boolean_t large_block_ok, boolean_t compressok,
#ifdef illumos
int outfd, vnode_t *vp, offset_t *off)
#else
int outfd, struct file *fp, offset_t *off)
#endif
{
dsl_pool_t *dp;
dsl_dataset_t *ds;
dsl_dataset_t *fromds = NULL;
int err;
err = dsl_pool_hold(pool, FTAG, &dp);
if (err != 0)
return (err);
err = dsl_dataset_hold_obj(dp, tosnap, FTAG, &ds);
if (err != 0) {
dsl_pool_rele(dp, FTAG);
return (err);
}
if (fromsnap != 0) {
zfs_bookmark_phys_t zb;
boolean_t is_clone;
err = dsl_dataset_hold_obj(dp, fromsnap, FTAG, &fromds);
if (err != 0) {
dsl_dataset_rele(ds, FTAG);
dsl_pool_rele(dp, FTAG);
return (err);
}
if (!dsl_dataset_is_before(ds, fromds, 0))
err = SET_ERROR(EXDEV);
zb.zbm_creation_time =
dsl_dataset_phys(fromds)->ds_creation_time;
zb.zbm_creation_txg = dsl_dataset_phys(fromds)->ds_creation_txg;
zb.zbm_guid = dsl_dataset_phys(fromds)->ds_guid;
is_clone = (fromds->ds_dir != ds->ds_dir);
dsl_dataset_rele(fromds, FTAG);
err = dmu_send_impl(FTAG, dp, ds, &zb, is_clone,
embedok, large_block_ok, compressok, outfd, 0, 0, fp, off);
} else {
err = dmu_send_impl(FTAG, dp, ds, NULL, B_FALSE,
embedok, large_block_ok, compressok, outfd, 0, 0, fp, off);
}
dsl_dataset_rele(ds, FTAG);
return (err);
}
int
dmu_send(const char *tosnap, const char *fromsnap, boolean_t embedok,
boolean_t large_block_ok, boolean_t compressok, int outfd,
uint64_t resumeobj, uint64_t resumeoff,
#ifdef illumos
vnode_t *vp, offset_t *off)
#else
struct file *fp, offset_t *off)
#endif
{
dsl_pool_t *dp;
dsl_dataset_t *ds;
int err;
boolean_t owned = B_FALSE;
if (fromsnap != NULL && strpbrk(fromsnap, "@#") == NULL)
return (SET_ERROR(EINVAL));
err = dsl_pool_hold(tosnap, FTAG, &dp);
if (err != 0)
return (err);
if (strchr(tosnap, '@') == NULL && spa_writeable(dp->dp_spa)) {
/*
* We are sending a filesystem or volume. Ensure
* that it doesn't change by owning the dataset.
*/
err = dsl_dataset_own(dp, tosnap, FTAG, &ds);
owned = B_TRUE;
} else {
err = dsl_dataset_hold(dp, tosnap, FTAG, &ds);
}
if (err != 0) {
dsl_pool_rele(dp, FTAG);
return (err);
}
if (fromsnap != NULL) {
zfs_bookmark_phys_t zb;
boolean_t is_clone = B_FALSE;
int fsnamelen = strchr(tosnap, '@') - tosnap;
/*
* If the fromsnap is in a different filesystem, then
* mark the send stream as a clone.
*/
if (strncmp(tosnap, fromsnap, fsnamelen) != 0 ||
(fromsnap[fsnamelen] != '@' &&
fromsnap[fsnamelen] != '#')) {
is_clone = B_TRUE;
}
if (strchr(fromsnap, '@')) {
dsl_dataset_t *fromds;
err = dsl_dataset_hold(dp, fromsnap, FTAG, &fromds);
if (err == 0) {
if (!dsl_dataset_is_before(ds, fromds, 0))
err = SET_ERROR(EXDEV);
zb.zbm_creation_time =
dsl_dataset_phys(fromds)->ds_creation_time;
zb.zbm_creation_txg =
dsl_dataset_phys(fromds)->ds_creation_txg;
zb.zbm_guid = dsl_dataset_phys(fromds)->ds_guid;
is_clone = (ds->ds_dir != fromds->ds_dir);
dsl_dataset_rele(fromds, FTAG);
}
} else {
err = dsl_bookmark_lookup(dp, fromsnap, ds, &zb);
}
if (err != 0) {
dsl_dataset_rele(ds, FTAG);
dsl_pool_rele(dp, FTAG);
return (err);
}
err = dmu_send_impl(FTAG, dp, ds, &zb, is_clone,
embedok, large_block_ok, compressok,
outfd, resumeobj, resumeoff, fp, off);
} else {
err = dmu_send_impl(FTAG, dp, ds, NULL, B_FALSE,
embedok, large_block_ok, compressok,
outfd, resumeobj, resumeoff, fp, off);
}
if (owned)
dsl_dataset_disown(ds, FTAG);
else
dsl_dataset_rele(ds, FTAG);
return (err);
}
static int
dmu_adjust_send_estimate_for_indirects(dsl_dataset_t *ds, uint64_t uncompressed,
uint64_t compressed, boolean_t stream_compressed, uint64_t *sizep)
{
int err = 0;
uint64_t size;
/*
* Assume that space (both on-disk and in-stream) is dominated by
* data. We will adjust for indirect blocks and the copies property,
* but ignore per-object space used (eg, dnodes and DRR_OBJECT records).
*/
uint64_t recordsize;
uint64_t record_count;
objset_t *os;
VERIFY0(dmu_objset_from_ds(ds, &os));
/* Assume all (uncompressed) blocks are recordsize. */
if (zfs_override_estimate_recordsize != 0) {
recordsize = zfs_override_estimate_recordsize;
} else if (os->os_phys->os_type == DMU_OST_ZVOL) {
err = dsl_prop_get_int_ds(ds,
zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE), &recordsize);
} else {
err = dsl_prop_get_int_ds(ds,
zfs_prop_to_name(ZFS_PROP_RECORDSIZE), &recordsize);
}
if (err != 0)
return (err);
record_count = uncompressed / recordsize;
/*
* If we're estimating a send size for a compressed stream, use the
* compressed data size to estimate the stream size. Otherwise, use the
* uncompressed data size.
*/
size = stream_compressed ? compressed : uncompressed;
/*
* Subtract out approximate space used by indirect blocks.
* Assume most space is used by data blocks (non-indirect, non-dnode).
* Assume no ditto blocks or internal fragmentation.
*
* Therefore, space used by indirect blocks is sizeof(blkptr_t) per
* block.
*/
size -= record_count * sizeof (blkptr_t);
/* Add in the space for the record associated with each block. */
size += record_count * sizeof (dmu_replay_record_t);
*sizep = size;
return (0);
}
int
dmu_send_estimate(dsl_dataset_t *ds, dsl_dataset_t *fromds,
boolean_t stream_compressed, uint64_t *sizep)
{
dsl_pool_t *dp = ds->ds_dir->dd_pool;
int err;
uint64_t uncomp, comp;
ASSERT(dsl_pool_config_held(dp));
/* tosnap must be a snapshot */
if (!ds->ds_is_snapshot)
return (SET_ERROR(EINVAL));
/* fromsnap, if provided, must be a snapshot */
if (fromds != NULL && !fromds->ds_is_snapshot)
return (SET_ERROR(EINVAL));
/*
* fromsnap must be an earlier snapshot from the same fs as tosnap,
* or the origin's fs.
*/
if (fromds != NULL && !dsl_dataset_is_before(ds, fromds, 0))
return (SET_ERROR(EXDEV));
/* Get compressed and uncompressed size estimates of changed data. */
if (fromds == NULL) {
uncomp = dsl_dataset_phys(ds)->ds_uncompressed_bytes;
comp = dsl_dataset_phys(ds)->ds_compressed_bytes;
} else {
uint64_t used;
err = dsl_dataset_space_written(fromds, ds,
&used, &comp, &uncomp);
if (err != 0)
return (err);
}
err = dmu_adjust_send_estimate_for_indirects(ds, uncomp, comp,
stream_compressed, sizep);
/*
* Add the size of the BEGIN and END records to the estimate.
*/
*sizep += 2 * sizeof (dmu_replay_record_t);
return (err);
}
struct calculate_send_arg {
uint64_t uncompressed;
uint64_t compressed;
};
/*
* Simple callback used to traverse the blocks of a snapshot and sum their
* uncompressed and compressed sizes.
*/
/* ARGSUSED */
static int
dmu_calculate_send_traversal(spa_t *spa, zilog_t *zilog, const blkptr_t *bp,
const zbookmark_phys_t *zb, const dnode_phys_t *dnp, void *arg)
{
struct calculate_send_arg *space = arg;
if (bp != NULL && !BP_IS_HOLE(bp)) {
space->uncompressed += BP_GET_UCSIZE(bp);
space->compressed += BP_GET_PSIZE(bp);
}
return (0);
}
/*
* Given a desination snapshot and a TXG, calculate the approximate size of a
* send stream sent from that TXG. from_txg may be zero, indicating that the
* whole snapshot will be sent.
*/
int
dmu_send_estimate_from_txg(dsl_dataset_t *ds, uint64_t from_txg,
boolean_t stream_compressed, uint64_t *sizep)
{
dsl_pool_t *dp = ds->ds_dir->dd_pool;
int err;
struct calculate_send_arg size = { 0 };
ASSERT(dsl_pool_config_held(dp));
/* tosnap must be a snapshot */
if (!ds->ds_is_snapshot)
return (SET_ERROR(EINVAL));
/* verify that from_txg is before the provided snapshot was taken */
if (from_txg >= dsl_dataset_phys(ds)->ds_creation_txg) {
return (SET_ERROR(EXDEV));
}
/*
* traverse the blocks of the snapshot with birth times after
* from_txg, summing their uncompressed size
*/
err = traverse_dataset(ds, from_txg, TRAVERSE_POST,
dmu_calculate_send_traversal, &size);
if (err)
return (err);
err = dmu_adjust_send_estimate_for_indirects(ds, size.uncompressed,
size.compressed, stream_compressed, sizep);
return (err);
}
typedef struct dmu_recv_begin_arg {
const char *drba_origin;
dmu_recv_cookie_t *drba_cookie;
cred_t *drba_cred;
uint64_t drba_snapobj;
} dmu_recv_begin_arg_t;
static int
recv_begin_check_existing_impl(dmu_recv_begin_arg_t *drba, dsl_dataset_t *ds,
uint64_t fromguid)
{
uint64_t val;
+ uint64_t children;
int error;
dsl_pool_t *dp = ds->ds_dir->dd_pool;
/* Temporary clone name must not exist. */
error = zap_lookup(dp->dp_meta_objset,
dsl_dir_phys(ds->ds_dir)->dd_child_dir_zapobj, recv_clone_name,
8, 1, &val);
if (error != ENOENT)
return (error == 0 ? SET_ERROR(EBUSY) : error);
/* Resume state must not be set. */
if (dsl_dataset_has_resume_receive_state(ds))
return (SET_ERROR(EBUSY));
/* New snapshot name must not exist. */
error = zap_lookup(dp->dp_meta_objset,
dsl_dataset_phys(ds)->ds_snapnames_zapobj,
drba->drba_cookie->drc_tosnap, 8, 1, &val);
if (error != ENOENT)
return (error == 0 ? SET_ERROR(EEXIST) : error);
+ /* must not have children if receiving a ZVOL */
+ error = zap_count(dp->dp_meta_objset,
+ dsl_dir_phys(ds->ds_dir)->dd_child_dir_zapobj, &children);
+ if (error != 0)
+ return (error);
+ if (drba->drba_cookie->drc_drrb->drr_type != DMU_OST_ZFS &&
+ children > 0)
+ return (SET_ERROR(ZFS_ERR_WRONG_PARENT));
+
/*
* Check snapshot limit before receiving. We'll recheck again at the
* end, but might as well abort before receiving if we're already over
* the limit.
*
* Note that we do not check the file system limit with
* dsl_dir_fscount_check because the temporary %clones don't count
* against that limit.
*/
error = dsl_fs_ss_limit_check(ds->ds_dir, 1, ZFS_PROP_SNAPSHOT_LIMIT,
NULL, drba->drba_cred);
if (error != 0)
return (error);
if (fromguid != 0) {
dsl_dataset_t *snap;
uint64_t obj = dsl_dataset_phys(ds)->ds_prev_snap_obj;
/* Find snapshot in this dir that matches fromguid. */
while (obj != 0) {
error = dsl_dataset_hold_obj(dp, obj, FTAG,
&snap);
if (error != 0)
return (SET_ERROR(ENODEV));
if (snap->ds_dir != ds->ds_dir) {
dsl_dataset_rele(snap, FTAG);
return (SET_ERROR(ENODEV));
}
if (dsl_dataset_phys(snap)->ds_guid == fromguid)
break;
obj = dsl_dataset_phys(snap)->ds_prev_snap_obj;
dsl_dataset_rele(snap, FTAG);
}
if (obj == 0)
return (SET_ERROR(ENODEV));
if (drba->drba_cookie->drc_force) {
drba->drba_snapobj = obj;
} else {
/*
* If we are not forcing, there must be no
* changes since fromsnap.
*/
if (dsl_dataset_modified_since_snap(ds, snap)) {
dsl_dataset_rele(snap, FTAG);
return (SET_ERROR(ETXTBSY));
}
drba->drba_snapobj = ds->ds_prev->ds_object;
}
dsl_dataset_rele(snap, FTAG);
} else {
/* if full, then must be forced */
if (!drba->drba_cookie->drc_force)
return (SET_ERROR(EEXIST));
/* start from $ORIGIN@$ORIGIN, if supported */
drba->drba_snapobj = dp->dp_origin_snap != NULL ?
dp->dp_origin_snap->ds_object : 0;
}
return (0);
}
static int
dmu_recv_begin_check(void *arg, dmu_tx_t *tx)
{
dmu_recv_begin_arg_t *drba = arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
struct drr_begin *drrb = drba->drba_cookie->drc_drrb;
uint64_t fromguid = drrb->drr_fromguid;
int flags = drrb->drr_flags;
int error;
uint64_t featureflags = DMU_GET_FEATUREFLAGS(drrb->drr_versioninfo);
dsl_dataset_t *ds;
const char *tofs = drba->drba_cookie->drc_tofs;
/* already checked */
ASSERT3U(drrb->drr_magic, ==, DMU_BACKUP_MAGIC);
ASSERT(!(featureflags & DMU_BACKUP_FEATURE_RESUMING));
if (DMU_GET_STREAM_HDRTYPE(drrb->drr_versioninfo) ==
DMU_COMPOUNDSTREAM ||
drrb->drr_type >= DMU_OST_NUMTYPES ||
((flags & DRR_FLAG_CLONE) && drba->drba_origin == NULL))
return (SET_ERROR(EINVAL));
/* Verify pool version supports SA if SA_SPILL feature set */
if ((featureflags & DMU_BACKUP_FEATURE_SA_SPILL) &&
spa_version(dp->dp_spa) < SPA_VERSION_SA)
return (SET_ERROR(ENOTSUP));
if (drba->drba_cookie->drc_resumable &&
!spa_feature_is_enabled(dp->dp_spa, SPA_FEATURE_EXTENSIBLE_DATASET))
return (SET_ERROR(ENOTSUP));
/*
* The receiving code doesn't know how to translate a WRITE_EMBEDDED
* record to a plain WRITE record, so the pool must have the
* EMBEDDED_DATA feature enabled if the stream has WRITE_EMBEDDED
* records. Same with WRITE_EMBEDDED records that use LZ4 compression.
*/
if ((featureflags & DMU_BACKUP_FEATURE_EMBED_DATA) &&
!spa_feature_is_enabled(dp->dp_spa, SPA_FEATURE_EMBEDDED_DATA))
return (SET_ERROR(ENOTSUP));
if ((featureflags & DMU_BACKUP_FEATURE_LZ4) &&
!spa_feature_is_enabled(dp->dp_spa, SPA_FEATURE_LZ4_COMPRESS))
return (SET_ERROR(ENOTSUP));
/*
* The receiving code doesn't know how to translate large blocks
* to smaller ones, so the pool must have the LARGE_BLOCKS
* feature enabled if the stream has LARGE_BLOCKS. Same with
* large dnodes.
*/
if ((featureflags & DMU_BACKUP_FEATURE_LARGE_BLOCKS) &&
!spa_feature_is_enabled(dp->dp_spa, SPA_FEATURE_LARGE_BLOCKS))
return (SET_ERROR(ENOTSUP));
if ((featureflags & DMU_BACKUP_FEATURE_LARGE_DNODE) &&
!spa_feature_is_enabled(dp->dp_spa, SPA_FEATURE_LARGE_DNODE))
return (SET_ERROR(ENOTSUP));
error = dsl_dataset_hold(dp, tofs, FTAG, &ds);
if (error == 0) {
/* target fs already exists; recv into temp clone */
/* Can't recv a clone into an existing fs */
if (flags & DRR_FLAG_CLONE || drba->drba_origin) {
dsl_dataset_rele(ds, FTAG);
return (SET_ERROR(EINVAL));
}
error = recv_begin_check_existing_impl(drba, ds, fromguid);
dsl_dataset_rele(ds, FTAG);
} else if (error == ENOENT) {
/* target fs does not exist; must be a full backup or clone */
char buf[ZFS_MAX_DATASET_NAME_LEN];
+ objset_t *os;
/*
* If it's a non-clone incremental, we are missing the
* target fs, so fail the recv.
*/
if (fromguid != 0 && !(flags & DRR_FLAG_CLONE ||
drba->drba_origin))
return (SET_ERROR(ENOENT));
/*
* If we're receiving a full send as a clone, and it doesn't
* contain all the necessary free records and freeobject
* records, reject it.
*/
if (fromguid == 0 && drba->drba_origin &&
!(flags & DRR_FLAG_FREERECORDS))
return (SET_ERROR(EINVAL));
/* Open the parent of tofs */
ASSERT3U(strlen(tofs), <, sizeof (buf));
(void) strlcpy(buf, tofs, strrchr(tofs, '/') - tofs + 1);
error = dsl_dataset_hold(dp, buf, FTAG, &ds);
if (error != 0)
return (error);
/*
* Check filesystem and snapshot limits before receiving. We'll
* recheck snapshot limits again at the end (we create the
* filesystems and increment those counts during begin_sync).
*/
error = dsl_fs_ss_limit_check(ds->ds_dir, 1,
ZFS_PROP_FILESYSTEM_LIMIT, NULL, drba->drba_cred);
if (error != 0) {
dsl_dataset_rele(ds, FTAG);
return (error);
}
error = dsl_fs_ss_limit_check(ds->ds_dir, 1,
ZFS_PROP_SNAPSHOT_LIMIT, NULL, drba->drba_cred);
if (error != 0) {
dsl_dataset_rele(ds, FTAG);
return (error);
}
+ /* can't recv below anything but filesystems (eg. no ZVOLs) */
+ error = dmu_objset_from_ds(ds, &os);
+ if (error != 0) {
+ dsl_dataset_rele(ds, FTAG);
+ return (error);
+ }
+ if (dmu_objset_type(os) != DMU_OST_ZFS) {
+ dsl_dataset_rele(ds, FTAG);
+ return (SET_ERROR(ZFS_ERR_WRONG_PARENT));
+ }
+
if (drba->drba_origin != NULL) {
dsl_dataset_t *origin;
error = dsl_dataset_hold(dp, drba->drba_origin,
FTAG, &origin);
if (error != 0) {
dsl_dataset_rele(ds, FTAG);
return (error);
}
if (!origin->ds_is_snapshot) {
dsl_dataset_rele(origin, FTAG);
dsl_dataset_rele(ds, FTAG);
return (SET_ERROR(EINVAL));
}
if (dsl_dataset_phys(origin)->ds_guid != fromguid &&
fromguid != 0) {
dsl_dataset_rele(origin, FTAG);
dsl_dataset_rele(ds, FTAG);
return (SET_ERROR(ENODEV));
}
dsl_dataset_rele(origin, FTAG);
}
+
dsl_dataset_rele(ds, FTAG);
error = 0;
}
return (error);
}
static void
dmu_recv_begin_sync(void *arg, dmu_tx_t *tx)
{
dmu_recv_begin_arg_t *drba = arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
objset_t *mos = dp->dp_meta_objset;
struct drr_begin *drrb = drba->drba_cookie->drc_drrb;
const char *tofs = drba->drba_cookie->drc_tofs;
dsl_dataset_t *ds, *newds;
uint64_t dsobj;
int error;
uint64_t crflags = 0;
if (drrb->drr_flags & DRR_FLAG_CI_DATA)
crflags |= DS_FLAG_CI_DATASET;
error = dsl_dataset_hold(dp, tofs, FTAG, &ds);
if (error == 0) {
/* create temporary clone */
dsl_dataset_t *snap = NULL;
if (drba->drba_snapobj != 0) {
VERIFY0(dsl_dataset_hold_obj(dp,
drba->drba_snapobj, FTAG, &snap));
}
dsobj = dsl_dataset_create_sync(ds->ds_dir, recv_clone_name,
snap, crflags, drba->drba_cred, tx);
if (drba->drba_snapobj != 0)
dsl_dataset_rele(snap, FTAG);
dsl_dataset_rele(ds, FTAG);
} else {
dsl_dir_t *dd;
const char *tail;
dsl_dataset_t *origin = NULL;
VERIFY0(dsl_dir_hold(dp, tofs, FTAG, &dd, &tail));
if (drba->drba_origin != NULL) {
VERIFY0(dsl_dataset_hold(dp, drba->drba_origin,
FTAG, &origin));
}
/* Create new dataset. */
dsobj = dsl_dataset_create_sync(dd,
strrchr(tofs, '/') + 1,
origin, crflags, drba->drba_cred, tx);
if (origin != NULL)
dsl_dataset_rele(origin, FTAG);
dsl_dir_rele(dd, FTAG);
drba->drba_cookie->drc_newfs = B_TRUE;
}
VERIFY0(dsl_dataset_own_obj(dp, dsobj, dmu_recv_tag, &newds));
if (drba->drba_cookie->drc_resumable) {
dsl_dataset_zapify(newds, tx);
if (drrb->drr_fromguid != 0) {
VERIFY0(zap_add(mos, dsobj, DS_FIELD_RESUME_FROMGUID,
8, 1, &drrb->drr_fromguid, tx));
}
VERIFY0(zap_add(mos, dsobj, DS_FIELD_RESUME_TOGUID,
8, 1, &drrb->drr_toguid, tx));
VERIFY0(zap_add(mos, dsobj, DS_FIELD_RESUME_TONAME,
1, strlen(drrb->drr_toname) + 1, drrb->drr_toname, tx));
uint64_t one = 1;
uint64_t zero = 0;
VERIFY0(zap_add(mos, dsobj, DS_FIELD_RESUME_OBJECT,
8, 1, &one, tx));
VERIFY0(zap_add(mos, dsobj, DS_FIELD_RESUME_OFFSET,
8, 1, &zero, tx));
VERIFY0(zap_add(mos, dsobj, DS_FIELD_RESUME_BYTES,
8, 1, &zero, tx));
if (DMU_GET_FEATUREFLAGS(drrb->drr_versioninfo) &
DMU_BACKUP_FEATURE_LARGE_BLOCKS) {
VERIFY0(zap_add(mos, dsobj, DS_FIELD_RESUME_LARGEBLOCK,
8, 1, &one, tx));
}
if (DMU_GET_FEATUREFLAGS(drrb->drr_versioninfo) &
DMU_BACKUP_FEATURE_EMBED_DATA) {
VERIFY0(zap_add(mos, dsobj, DS_FIELD_RESUME_EMBEDOK,
8, 1, &one, tx));
}
if (DMU_GET_FEATUREFLAGS(drrb->drr_versioninfo) &
DMU_BACKUP_FEATURE_COMPRESSED) {
VERIFY0(zap_add(mos, dsobj, DS_FIELD_RESUME_COMPRESSOK,
8, 1, &one, tx));
}
}
dmu_buf_will_dirty(newds->ds_dbuf, tx);
dsl_dataset_phys(newds)->ds_flags |= DS_FLAG_INCONSISTENT;
/*
* If we actually created a non-clone, we need to create the
* objset in our new dataset.
*/
rrw_enter(&newds->ds_bp_rwlock, RW_READER, FTAG);
if (BP_IS_HOLE(dsl_dataset_get_blkptr(newds))) {
(void) dmu_objset_create_impl(dp->dp_spa,
newds, dsl_dataset_get_blkptr(newds), drrb->drr_type, tx);
}
rrw_exit(&newds->ds_bp_rwlock, FTAG);
drba->drba_cookie->drc_ds = newds;
spa_history_log_internal_ds(newds, "receive", tx, "");
}
static int
dmu_recv_resume_begin_check(void *arg, dmu_tx_t *tx)
{
dmu_recv_begin_arg_t *drba = arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
struct drr_begin *drrb = drba->drba_cookie->drc_drrb;
int error;
uint64_t featureflags = DMU_GET_FEATUREFLAGS(drrb->drr_versioninfo);
dsl_dataset_t *ds;
const char *tofs = drba->drba_cookie->drc_tofs;
/* 6 extra bytes for /%recv */
char recvname[ZFS_MAX_DATASET_NAME_LEN + 6];
/* already checked */
ASSERT3U(drrb->drr_magic, ==, DMU_BACKUP_MAGIC);
ASSERT(featureflags & DMU_BACKUP_FEATURE_RESUMING);
if (DMU_GET_STREAM_HDRTYPE(drrb->drr_versioninfo) ==
DMU_COMPOUNDSTREAM ||
drrb->drr_type >= DMU_OST_NUMTYPES)
return (SET_ERROR(EINVAL));
/* Verify pool version supports SA if SA_SPILL feature set */
if ((featureflags & DMU_BACKUP_FEATURE_SA_SPILL) &&
spa_version(dp->dp_spa) < SPA_VERSION_SA)
return (SET_ERROR(ENOTSUP));
/*
* The receiving code doesn't know how to translate a WRITE_EMBEDDED
* record to a plain WRITE record, so the pool must have the
* EMBEDDED_DATA feature enabled if the stream has WRITE_EMBEDDED
* records. Same with WRITE_EMBEDDED records that use LZ4 compression.
*/
if ((featureflags & DMU_BACKUP_FEATURE_EMBED_DATA) &&
!spa_feature_is_enabled(dp->dp_spa, SPA_FEATURE_EMBEDDED_DATA))
return (SET_ERROR(ENOTSUP));
if ((featureflags & DMU_BACKUP_FEATURE_LZ4) &&
!spa_feature_is_enabled(dp->dp_spa, SPA_FEATURE_LZ4_COMPRESS))
return (SET_ERROR(ENOTSUP));
/*
* The receiving code doesn't know how to translate large blocks
* to smaller ones, so the pool must have the LARGE_BLOCKS
* feature enabled if the stream has LARGE_BLOCKS. Same with
* large dnodes.
*/
if ((featureflags & DMU_BACKUP_FEATURE_LARGE_BLOCKS) &&
!spa_feature_is_enabled(dp->dp_spa, SPA_FEATURE_LARGE_BLOCKS))
return (SET_ERROR(ENOTSUP));
if ((featureflags & DMU_BACKUP_FEATURE_LARGE_DNODE) &&
!spa_feature_is_enabled(dp->dp_spa, SPA_FEATURE_LARGE_DNODE))
return (SET_ERROR(ENOTSUP));
(void) snprintf(recvname, sizeof (recvname), "%s/%s",
tofs, recv_clone_name);
if (dsl_dataset_hold(dp, recvname, FTAG, &ds) != 0) {
/* %recv does not exist; continue in tofs */
error = dsl_dataset_hold(dp, tofs, FTAG, &ds);
if (error != 0)
return (error);
}
/* check that ds is marked inconsistent */
if (!DS_IS_INCONSISTENT(ds)) {
dsl_dataset_rele(ds, FTAG);
return (SET_ERROR(EINVAL));
}
/* check that there is resuming data, and that the toguid matches */
if (!dsl_dataset_is_zapified(ds)) {
dsl_dataset_rele(ds, FTAG);
return (SET_ERROR(EINVAL));
}
uint64_t val;
error = zap_lookup(dp->dp_meta_objset, ds->ds_object,
DS_FIELD_RESUME_TOGUID, sizeof (val), 1, &val);
if (error != 0 || drrb->drr_toguid != val) {
dsl_dataset_rele(ds, FTAG);
return (SET_ERROR(EINVAL));
}
/*
* Check if the receive is still running. If so, it will be owned.
* Note that nothing else can own the dataset (e.g. after the receive
* fails) because it will be marked inconsistent.
*/
if (dsl_dataset_has_owner(ds)) {
dsl_dataset_rele(ds, FTAG);
return (SET_ERROR(EBUSY));
}
/* There should not be any snapshots of this fs yet. */
if (ds->ds_prev != NULL && ds->ds_prev->ds_dir == ds->ds_dir) {
dsl_dataset_rele(ds, FTAG);
return (SET_ERROR(EINVAL));
}
/*
* Note: resume point will be checked when we process the first WRITE
* record.
*/
/* check that the origin matches */
val = 0;
(void) zap_lookup(dp->dp_meta_objset, ds->ds_object,
DS_FIELD_RESUME_FROMGUID, sizeof (val), 1, &val);
if (drrb->drr_fromguid != val) {
dsl_dataset_rele(ds, FTAG);
return (SET_ERROR(EINVAL));
}
dsl_dataset_rele(ds, FTAG);
return (0);
}
static void
dmu_recv_resume_begin_sync(void *arg, dmu_tx_t *tx)
{
dmu_recv_begin_arg_t *drba = arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
const char *tofs = drba->drba_cookie->drc_tofs;
dsl_dataset_t *ds;
uint64_t dsobj;
/* 6 extra bytes for /%recv */
char recvname[ZFS_MAX_DATASET_NAME_LEN + 6];
(void) snprintf(recvname, sizeof (recvname), "%s/%s",
tofs, recv_clone_name);
if (dsl_dataset_hold(dp, recvname, FTAG, &ds) != 0) {
/* %recv does not exist; continue in tofs */
VERIFY0(dsl_dataset_hold(dp, tofs, FTAG, &ds));
drba->drba_cookie->drc_newfs = B_TRUE;
}
/* clear the inconsistent flag so that we can own it */
ASSERT(DS_IS_INCONSISTENT(ds));
dmu_buf_will_dirty(ds->ds_dbuf, tx);
dsl_dataset_phys(ds)->ds_flags &= ~DS_FLAG_INCONSISTENT;
dsobj = ds->ds_object;
dsl_dataset_rele(ds, FTAG);
VERIFY0(dsl_dataset_own_obj(dp, dsobj, dmu_recv_tag, &ds));
dmu_buf_will_dirty(ds->ds_dbuf, tx);
dsl_dataset_phys(ds)->ds_flags |= DS_FLAG_INCONSISTENT;
rrw_enter(&ds->ds_bp_rwlock, RW_READER, FTAG);
ASSERT(!BP_IS_HOLE(dsl_dataset_get_blkptr(ds)));
rrw_exit(&ds->ds_bp_rwlock, FTAG);
drba->drba_cookie->drc_ds = ds;
spa_history_log_internal_ds(ds, "resume receive", tx, "");
}
/*
* NB: callers *MUST* call dmu_recv_stream() if dmu_recv_begin()
* succeeds; otherwise we will leak the holds on the datasets.
*/
int
dmu_recv_begin(char *tofs, char *tosnap, dmu_replay_record_t *drr_begin,
boolean_t force, boolean_t resumable, char *origin, dmu_recv_cookie_t *drc)
{
dmu_recv_begin_arg_t drba = { 0 };
bzero(drc, sizeof (dmu_recv_cookie_t));
drc->drc_drr_begin = drr_begin;
drc->drc_drrb = &drr_begin->drr_u.drr_begin;
drc->drc_tosnap = tosnap;
drc->drc_tofs = tofs;
drc->drc_force = force;
drc->drc_resumable = resumable;
drc->drc_cred = CRED();
drc->drc_clone = (origin != NULL);
if (drc->drc_drrb->drr_magic == BSWAP_64(DMU_BACKUP_MAGIC)) {
drc->drc_byteswap = B_TRUE;
(void) fletcher_4_incremental_byteswap(drr_begin,
sizeof (dmu_replay_record_t), &drc->drc_cksum);
byteswap_record(drr_begin);
} else if (drc->drc_drrb->drr_magic == DMU_BACKUP_MAGIC) {
(void) fletcher_4_incremental_native(drr_begin,
sizeof (dmu_replay_record_t), &drc->drc_cksum);
} else {
return (SET_ERROR(EINVAL));
}
drba.drba_origin = origin;
drba.drba_cookie = drc;
drba.drba_cred = CRED();
if (DMU_GET_FEATUREFLAGS(drc->drc_drrb->drr_versioninfo) &
DMU_BACKUP_FEATURE_RESUMING) {
return (dsl_sync_task(tofs,
dmu_recv_resume_begin_check, dmu_recv_resume_begin_sync,
&drba, 5, ZFS_SPACE_CHECK_NORMAL));
} else {
return (dsl_sync_task(tofs,
dmu_recv_begin_check, dmu_recv_begin_sync,
&drba, 5, ZFS_SPACE_CHECK_NORMAL));
}
}
struct receive_record_arg {
dmu_replay_record_t header;
void *payload; /* Pointer to a buffer containing the payload */
/*
* If the record is a write, pointer to the arc_buf_t containing the
* payload.
*/
arc_buf_t *write_buf;
int payload_size;
uint64_t bytes_read; /* bytes read from stream when record created */
boolean_t eos_marker; /* Marks the end of the stream */
bqueue_node_t node;
};
struct receive_writer_arg {
objset_t *os;
boolean_t byteswap;
bqueue_t q;
/*
* These three args are used to signal to the main thread that we're
* done.
*/
kmutex_t mutex;
kcondvar_t cv;
boolean_t done;
int err;
/* A map from guid to dataset to help handle dedup'd streams. */
avl_tree_t *guid_to_ds_map;
boolean_t resumable;
uint64_t last_object;
uint64_t last_offset;
uint64_t max_object; /* highest object ID referenced in stream */
uint64_t bytes_read; /* bytes read when current record created */
};
struct objlist {
list_t list; /* List of struct receive_objnode. */
/*
* Last object looked up. Used to assert that objects are being looked
* up in ascending order.
*/
uint64_t last_lookup;
};
struct receive_objnode {
list_node_t node;
uint64_t object;
};
struct receive_arg {
objset_t *os;
kthread_t *td;
struct file *fp;
uint64_t voff; /* The current offset in the stream */
uint64_t bytes_read;
/*
* A record that has had its payload read in, but hasn't yet been handed
* off to the worker thread.
*/
struct receive_record_arg *rrd;
/* A record that has had its header read in, but not its payload. */
struct receive_record_arg *next_rrd;
zio_cksum_t cksum;
zio_cksum_t prev_cksum;
int err;
boolean_t byteswap;
/* Sorted list of objects not to issue prefetches for. */
struct objlist ignore_objlist;
};
typedef struct guid_map_entry {
uint64_t guid;
dsl_dataset_t *gme_ds;
avl_node_t avlnode;
} guid_map_entry_t;
static int
guid_compare(const void *arg1, const void *arg2)
{
const guid_map_entry_t *gmep1 = (const guid_map_entry_t *)arg1;
const guid_map_entry_t *gmep2 = (const guid_map_entry_t *)arg2;
return (AVL_CMP(gmep1->guid, gmep2->guid));
}
static void
free_guid_map_onexit(void *arg)
{
avl_tree_t *ca = arg;
void *cookie = NULL;
guid_map_entry_t *gmep;
while ((gmep = avl_destroy_nodes(ca, &cookie)) != NULL) {
dsl_dataset_long_rele(gmep->gme_ds, gmep);
dsl_dataset_rele(gmep->gme_ds, gmep);
kmem_free(gmep, sizeof (guid_map_entry_t));
}
avl_destroy(ca);
kmem_free(ca, sizeof (avl_tree_t));
}
static int
restore_bytes(struct receive_arg *ra, void *buf, int len, off_t off, ssize_t *resid)
{
struct uio auio;
struct iovec aiov;
int error;
aiov.iov_base = buf;
aiov.iov_len = len;
auio.uio_iov = &aiov;
auio.uio_iovcnt = 1;
auio.uio_resid = len;
auio.uio_segflg = UIO_SYSSPACE;
auio.uio_rw = UIO_READ;
auio.uio_offset = off;
auio.uio_td = ra->td;
#ifdef _KERNEL
error = fo_read(ra->fp, &auio, ra->td->td_ucred, FOF_OFFSET, ra->td);
#else
fprintf(stderr, "%s: returning EOPNOTSUPP\n", __func__);
error = EOPNOTSUPP;
#endif
*resid = auio.uio_resid;
return (error);
}
static int
receive_read(struct receive_arg *ra, int len, void *buf)
{
int done = 0;
/*
* The code doesn't rely on this (lengths being multiples of 8). See
* comment in dump_bytes.
*/
ASSERT0(len % 8);
while (done < len) {
ssize_t resid;
ra->err = restore_bytes(ra, buf + done,
len - done, ra->voff, &resid);
if (resid == len - done) {
/*
* Note: ECKSUM indicates that the receive
* was interrupted and can potentially be resumed.
*/
ra->err = SET_ERROR(ECKSUM);
}
ra->voff += len - done - resid;
done = len - resid;
if (ra->err != 0)
return (ra->err);
}
ra->bytes_read += len;
ASSERT3U(done, ==, len);
return (0);
}
noinline static void
byteswap_record(dmu_replay_record_t *drr)
{
#define DO64(X) (drr->drr_u.X = BSWAP_64(drr->drr_u.X))
#define DO32(X) (drr->drr_u.X = BSWAP_32(drr->drr_u.X))
drr->drr_type = BSWAP_32(drr->drr_type);
drr->drr_payloadlen = BSWAP_32(drr->drr_payloadlen);
switch (drr->drr_type) {
case DRR_BEGIN:
DO64(drr_begin.drr_magic);
DO64(drr_begin.drr_versioninfo);
DO64(drr_begin.drr_creation_time);
DO32(drr_begin.drr_type);
DO32(drr_begin.drr_flags);
DO64(drr_begin.drr_toguid);
DO64(drr_begin.drr_fromguid);
break;
case DRR_OBJECT:
DO64(drr_object.drr_object);
DO32(drr_object.drr_type);
DO32(drr_object.drr_bonustype);
DO32(drr_object.drr_blksz);
DO32(drr_object.drr_bonuslen);
DO64(drr_object.drr_toguid);
break;
case DRR_FREEOBJECTS:
DO64(drr_freeobjects.drr_firstobj);
DO64(drr_freeobjects.drr_numobjs);
DO64(drr_freeobjects.drr_toguid);
break;
case DRR_WRITE:
DO64(drr_write.drr_object);
DO32(drr_write.drr_type);
DO64(drr_write.drr_offset);
DO64(drr_write.drr_logical_size);
DO64(drr_write.drr_toguid);
ZIO_CHECKSUM_BSWAP(&drr->drr_u.drr_write.drr_key.ddk_cksum);
DO64(drr_write.drr_key.ddk_prop);
DO64(drr_write.drr_compressed_size);
break;
case DRR_WRITE_BYREF:
DO64(drr_write_byref.drr_object);
DO64(drr_write_byref.drr_offset);
DO64(drr_write_byref.drr_length);
DO64(drr_write_byref.drr_toguid);
DO64(drr_write_byref.drr_refguid);
DO64(drr_write_byref.drr_refobject);
DO64(drr_write_byref.drr_refoffset);
ZIO_CHECKSUM_BSWAP(&drr->drr_u.drr_write_byref.
drr_key.ddk_cksum);
DO64(drr_write_byref.drr_key.ddk_prop);
break;
case DRR_WRITE_EMBEDDED:
DO64(drr_write_embedded.drr_object);
DO64(drr_write_embedded.drr_offset);
DO64(drr_write_embedded.drr_length);
DO64(drr_write_embedded.drr_toguid);
DO32(drr_write_embedded.drr_lsize);
DO32(drr_write_embedded.drr_psize);
break;
case DRR_FREE:
DO64(drr_free.drr_object);
DO64(drr_free.drr_offset);
DO64(drr_free.drr_length);
DO64(drr_free.drr_toguid);
break;
case DRR_SPILL:
DO64(drr_spill.drr_object);
DO64(drr_spill.drr_length);
DO64(drr_spill.drr_toguid);
break;
case DRR_END:
DO64(drr_end.drr_toguid);
ZIO_CHECKSUM_BSWAP(&drr->drr_u.drr_end.drr_checksum);
break;
}
if (drr->drr_type != DRR_BEGIN) {
ZIO_CHECKSUM_BSWAP(&drr->drr_u.drr_checksum.drr_checksum);
}
#undef DO64
#undef DO32
}
static inline uint8_t
deduce_nblkptr(dmu_object_type_t bonus_type, uint64_t bonus_size)
{
if (bonus_type == DMU_OT_SA) {
return (1);
} else {
return (1 +
((DN_OLD_MAX_BONUSLEN -
MIN(DN_OLD_MAX_BONUSLEN, bonus_size)) >> SPA_BLKPTRSHIFT));
}
}
static void
save_resume_state(struct receive_writer_arg *rwa,
uint64_t object, uint64_t offset, dmu_tx_t *tx)
{
int txgoff = dmu_tx_get_txg(tx) & TXG_MASK;
if (!rwa->resumable)
return;
/*
* We use ds_resume_bytes[] != 0 to indicate that we need to
* update this on disk, so it must not be 0.
*/
ASSERT(rwa->bytes_read != 0);
/*
* We only resume from write records, which have a valid
* (non-meta-dnode) object number.
*/
ASSERT(object != 0);
/*
* For resuming to work correctly, we must receive records in order,
* sorted by object,offset. This is checked by the callers, but
* assert it here for good measure.
*/
ASSERT3U(object, >=, rwa->os->os_dsl_dataset->ds_resume_object[txgoff]);
ASSERT(object != rwa->os->os_dsl_dataset->ds_resume_object[txgoff] ||
offset >= rwa->os->os_dsl_dataset->ds_resume_offset[txgoff]);
ASSERT3U(rwa->bytes_read, >=,
rwa->os->os_dsl_dataset->ds_resume_bytes[txgoff]);
rwa->os->os_dsl_dataset->ds_resume_object[txgoff] = object;
rwa->os->os_dsl_dataset->ds_resume_offset[txgoff] = offset;
rwa->os->os_dsl_dataset->ds_resume_bytes[txgoff] = rwa->bytes_read;
}
noinline static int
receive_object(struct receive_writer_arg *rwa, struct drr_object *drro,
void *data)
{
dmu_object_info_t doi;
dmu_tx_t *tx;
uint64_t object;
int err;
uint8_t dn_slots = drro->drr_dn_slots != 0 ?
drro->drr_dn_slots : DNODE_MIN_SLOTS;
if (drro->drr_type == DMU_OT_NONE ||
!DMU_OT_IS_VALID(drro->drr_type) ||
!DMU_OT_IS_VALID(drro->drr_bonustype) ||
drro->drr_checksumtype >= ZIO_CHECKSUM_FUNCTIONS ||
drro->drr_compress >= ZIO_COMPRESS_FUNCTIONS ||
P2PHASE(drro->drr_blksz, SPA_MINBLOCKSIZE) ||
drro->drr_blksz < SPA_MINBLOCKSIZE ||
drro->drr_blksz > spa_maxblocksize(dmu_objset_spa(rwa->os)) ||
drro->drr_bonuslen >
DN_BONUS_SIZE(spa_maxdnodesize(dmu_objset_spa(rwa->os))) ||
dn_slots >
(spa_maxdnodesize(dmu_objset_spa(rwa->os)) >> DNODE_SHIFT)) {
return (SET_ERROR(EINVAL));
}
err = dmu_object_info(rwa->os, drro->drr_object, &doi);
if (err != 0 && err != ENOENT && err != EEXIST)
return (SET_ERROR(EINVAL));
if (drro->drr_object > rwa->max_object)
rwa->max_object = drro->drr_object;
/*
* If we are losing blkptrs or changing the block size this must
* be a new file instance. We must clear out the previous file
* contents before we can change this type of metadata in the dnode.
*/
if (err == 0) {
int nblkptr;
object = drro->drr_object;
nblkptr = deduce_nblkptr(drro->drr_bonustype,
drro->drr_bonuslen);
if (drro->drr_blksz != doi.doi_data_block_size ||
nblkptr < doi.doi_nblkptr ||
dn_slots != doi.doi_dnodesize >> DNODE_SHIFT) {
err = dmu_free_long_range(rwa->os, drro->drr_object,
0, DMU_OBJECT_END);
if (err != 0)
return (SET_ERROR(EINVAL));
}
} else if (err == EEXIST) {
/*
* The object requested is currently an interior slot of a
* multi-slot dnode. This will be resolved when the next txg
* is synced out, since the send stream will have told us
* to free this slot when we freed the associated dnode
* earlier in the stream.
*/
txg_wait_synced(dmu_objset_pool(rwa->os), 0);
object = drro->drr_object;
} else {
/* object is free and we are about to allocate a new one */
object = DMU_NEW_OBJECT;
}
/*
* If this is a multi-slot dnode there is a chance that this
* object will expand into a slot that is already used by
* another object from the previous snapshot. We must free
* these objects before we attempt to allocate the new dnode.
*/
if (dn_slots > 1) {
boolean_t need_sync = B_FALSE;
for (uint64_t slot = drro->drr_object + 1;
slot < drro->drr_object + dn_slots;
slot++) {
dmu_object_info_t slot_doi;
err = dmu_object_info(rwa->os, slot, &slot_doi);
if (err == ENOENT || err == EEXIST)
continue;
else if (err != 0)
return (err);
err = dmu_free_long_object(rwa->os, slot);
if (err != 0)
return (err);
need_sync = B_TRUE;
}
if (need_sync)
txg_wait_synced(dmu_objset_pool(rwa->os), 0);
}
tx = dmu_tx_create(rwa->os);
dmu_tx_hold_bonus(tx, object);
err = dmu_tx_assign(tx, TXG_WAIT);
if (err != 0) {
dmu_tx_abort(tx);
return (err);
}
if (object == DMU_NEW_OBJECT) {
/* currently free, want to be allocated */
err = dmu_object_claim_dnsize(rwa->os, drro->drr_object,
drro->drr_type, drro->drr_blksz,
drro->drr_bonustype, drro->drr_bonuslen,
dn_slots << DNODE_SHIFT, tx);
} else if (drro->drr_type != doi.doi_type ||
drro->drr_blksz != doi.doi_data_block_size ||
drro->drr_bonustype != doi.doi_bonus_type ||
drro->drr_bonuslen != doi.doi_bonus_size ||
drro->drr_dn_slots != (doi.doi_dnodesize >> DNODE_SHIFT)) {
/* currently allocated, but with different properties */
err = dmu_object_reclaim_dnsize(rwa->os, drro->drr_object,
drro->drr_type, drro->drr_blksz,
drro->drr_bonustype, drro->drr_bonuslen,
drro->drr_dn_slots << DNODE_SHIFT, tx);
}
if (err != 0) {
dmu_tx_commit(tx);
return (SET_ERROR(EINVAL));
}
dmu_object_set_checksum(rwa->os, drro->drr_object,
drro->drr_checksumtype, tx);
dmu_object_set_compress(rwa->os, drro->drr_object,
drro->drr_compress, tx);
if (data != NULL) {
dmu_buf_t *db;
VERIFY0(dmu_bonus_hold(rwa->os, drro->drr_object, FTAG, &db));
dmu_buf_will_dirty(db, tx);
ASSERT3U(db->db_size, >=, drro->drr_bonuslen);
bcopy(data, db->db_data, drro->drr_bonuslen);
if (rwa->byteswap) {
dmu_object_byteswap_t byteswap =
DMU_OT_BYTESWAP(drro->drr_bonustype);
dmu_ot_byteswap[byteswap].ob_func(db->db_data,
drro->drr_bonuslen);
}
dmu_buf_rele(db, FTAG);
}
dmu_tx_commit(tx);
return (0);
}
/* ARGSUSED */
noinline static int
receive_freeobjects(struct receive_writer_arg *rwa,
struct drr_freeobjects *drrfo)
{
uint64_t obj;
int next_err = 0;
if (drrfo->drr_firstobj + drrfo->drr_numobjs < drrfo->drr_firstobj)
return (SET_ERROR(EINVAL));
for (obj = drrfo->drr_firstobj == 0 ? 1 : drrfo->drr_firstobj;
obj < drrfo->drr_firstobj + drrfo->drr_numobjs && next_err == 0;
next_err = dmu_object_next(rwa->os, &obj, FALSE, 0)) {
dmu_object_info_t doi;
int err;
err = dmu_object_info(rwa->os, obj, NULL);
if (err == ENOENT)
continue;
else if (err != 0)
return (err);
err = dmu_free_long_object(rwa->os, obj);
if (err != 0)
return (err);
if (obj > rwa->max_object)
rwa->max_object = obj;
}
if (next_err != ESRCH)
return (next_err);
return (0);
}
noinline static int
receive_write(struct receive_writer_arg *rwa, struct drr_write *drrw,
arc_buf_t *abuf)
{
dmu_tx_t *tx;
int err;
if (drrw->drr_offset + drrw->drr_logical_size < drrw->drr_offset ||
!DMU_OT_IS_VALID(drrw->drr_type))
return (SET_ERROR(EINVAL));
/*
* For resuming to work, records must be in increasing order
* by (object, offset).
*/
if (drrw->drr_object < rwa->last_object ||
(drrw->drr_object == rwa->last_object &&
drrw->drr_offset < rwa->last_offset)) {
return (SET_ERROR(EINVAL));
}
rwa->last_object = drrw->drr_object;
rwa->last_offset = drrw->drr_offset;
if (rwa->last_object > rwa->max_object)
rwa->max_object = rwa->last_object;
if (dmu_object_info(rwa->os, drrw->drr_object, NULL) != 0)
return (SET_ERROR(EINVAL));
tx = dmu_tx_create(rwa->os);
dmu_tx_hold_write(tx, drrw->drr_object,
drrw->drr_offset, drrw->drr_logical_size);
err = dmu_tx_assign(tx, TXG_WAIT);
if (err != 0) {
dmu_tx_abort(tx);
return (err);
}
if (rwa->byteswap) {
dmu_object_byteswap_t byteswap =
DMU_OT_BYTESWAP(drrw->drr_type);
dmu_ot_byteswap[byteswap].ob_func(abuf->b_data,
DRR_WRITE_PAYLOAD_SIZE(drrw));
}
/* use the bonus buf to look up the dnode in dmu_assign_arcbuf */
dmu_buf_t *bonus;
if (dmu_bonus_hold(rwa->os, drrw->drr_object, FTAG, &bonus) != 0)
return (SET_ERROR(EINVAL));
dmu_assign_arcbuf(bonus, drrw->drr_offset, abuf, tx);
/*
* Note: If the receive fails, we want the resume stream to start
* with the same record that we last successfully received (as opposed
* to the next record), so that we can verify that we are
* resuming from the correct location.
*/
save_resume_state(rwa, drrw->drr_object, drrw->drr_offset, tx);
dmu_tx_commit(tx);
dmu_buf_rele(bonus, FTAG);
return (0);
}
/*
* Handle a DRR_WRITE_BYREF record. This record is used in dedup'ed
* streams to refer to a copy of the data that is already on the
* system because it came in earlier in the stream. This function
* finds the earlier copy of the data, and uses that copy instead of
* data from the stream to fulfill this write.
*/
static int
receive_write_byref(struct receive_writer_arg *rwa,
struct drr_write_byref *drrwbr)
{
dmu_tx_t *tx;
int err;
guid_map_entry_t gmesrch;
guid_map_entry_t *gmep;
avl_index_t where;
objset_t *ref_os = NULL;
dmu_buf_t *dbp;
if (drrwbr->drr_offset + drrwbr->drr_length < drrwbr->drr_offset)
return (SET_ERROR(EINVAL));
/*
* If the GUID of the referenced dataset is different from the
* GUID of the target dataset, find the referenced dataset.
*/
if (drrwbr->drr_toguid != drrwbr->drr_refguid) {
gmesrch.guid = drrwbr->drr_refguid;
if ((gmep = avl_find(rwa->guid_to_ds_map, &gmesrch,
&where)) == NULL) {
return (SET_ERROR(EINVAL));
}
if (dmu_objset_from_ds(gmep->gme_ds, &ref_os))
return (SET_ERROR(EINVAL));
} else {
ref_os = rwa->os;
}
if (drrwbr->drr_object > rwa->max_object)
rwa->max_object = drrwbr->drr_object;
err = dmu_buf_hold(ref_os, drrwbr->drr_refobject,
drrwbr->drr_refoffset, FTAG, &dbp, DMU_READ_PREFETCH);
if (err != 0)
return (err);
tx = dmu_tx_create(rwa->os);
dmu_tx_hold_write(tx, drrwbr->drr_object,
drrwbr->drr_offset, drrwbr->drr_length);
err = dmu_tx_assign(tx, TXG_WAIT);
if (err != 0) {
dmu_tx_abort(tx);
return (err);
}
dmu_write(rwa->os, drrwbr->drr_object,
drrwbr->drr_offset, drrwbr->drr_length, dbp->db_data, tx);
dmu_buf_rele(dbp, FTAG);
/* See comment in restore_write. */
save_resume_state(rwa, drrwbr->drr_object, drrwbr->drr_offset, tx);
dmu_tx_commit(tx);
return (0);
}
static int
receive_write_embedded(struct receive_writer_arg *rwa,
struct drr_write_embedded *drrwe, void *data)
{
dmu_tx_t *tx;
int err;
if (drrwe->drr_offset + drrwe->drr_length < drrwe->drr_offset)
return (EINVAL);
if (drrwe->drr_psize > BPE_PAYLOAD_SIZE)
return (EINVAL);
if (drrwe->drr_etype >= NUM_BP_EMBEDDED_TYPES)
return (EINVAL);
if (drrwe->drr_compression >= ZIO_COMPRESS_FUNCTIONS)
return (EINVAL);
if (drrwe->drr_object > rwa->max_object)
rwa->max_object = drrwe->drr_object;
tx = dmu_tx_create(rwa->os);
dmu_tx_hold_write(tx, drrwe->drr_object,
drrwe->drr_offset, drrwe->drr_length);
err = dmu_tx_assign(tx, TXG_WAIT);
if (err != 0) {
dmu_tx_abort(tx);
return (err);
}
dmu_write_embedded(rwa->os, drrwe->drr_object,
drrwe->drr_offset, data, drrwe->drr_etype,
drrwe->drr_compression, drrwe->drr_lsize, drrwe->drr_psize,
rwa->byteswap ^ ZFS_HOST_BYTEORDER, tx);
/* See comment in restore_write. */
save_resume_state(rwa, drrwe->drr_object, drrwe->drr_offset, tx);
dmu_tx_commit(tx);
return (0);
}
static int
receive_spill(struct receive_writer_arg *rwa, struct drr_spill *drrs,
void *data)
{
dmu_tx_t *tx;
dmu_buf_t *db, *db_spill;
int err;
if (drrs->drr_length < SPA_MINBLOCKSIZE ||
drrs->drr_length > spa_maxblocksize(dmu_objset_spa(rwa->os)))
return (SET_ERROR(EINVAL));
if (dmu_object_info(rwa->os, drrs->drr_object, NULL) != 0)
return (SET_ERROR(EINVAL));
if (drrs->drr_object > rwa->max_object)
rwa->max_object = drrs->drr_object;
VERIFY0(dmu_bonus_hold(rwa->os, drrs->drr_object, FTAG, &db));
if ((err = dmu_spill_hold_by_bonus(db, FTAG, &db_spill)) != 0) {
dmu_buf_rele(db, FTAG);
return (err);
}
tx = dmu_tx_create(rwa->os);
dmu_tx_hold_spill(tx, db->db_object);
err = dmu_tx_assign(tx, TXG_WAIT);
if (err != 0) {
dmu_buf_rele(db, FTAG);
dmu_buf_rele(db_spill, FTAG);
dmu_tx_abort(tx);
return (err);
}
dmu_buf_will_dirty(db_spill, tx);
if (db_spill->db_size < drrs->drr_length)
VERIFY(0 == dbuf_spill_set_blksz(db_spill,
drrs->drr_length, tx));
bcopy(data, db_spill->db_data, drrs->drr_length);
dmu_buf_rele(db, FTAG);
dmu_buf_rele(db_spill, FTAG);
dmu_tx_commit(tx);
return (0);
}
/* ARGSUSED */
noinline static int
receive_free(struct receive_writer_arg *rwa, struct drr_free *drrf)
{
int err;
if (drrf->drr_length != -1ULL &&
drrf->drr_offset + drrf->drr_length < drrf->drr_offset)
return (SET_ERROR(EINVAL));
if (dmu_object_info(rwa->os, drrf->drr_object, NULL) != 0)
return (SET_ERROR(EINVAL));
if (drrf->drr_object > rwa->max_object)
rwa->max_object = drrf->drr_object;
err = dmu_free_long_range(rwa->os, drrf->drr_object,
drrf->drr_offset, drrf->drr_length);
return (err);
}
/* used to destroy the drc_ds on error */
static void
dmu_recv_cleanup_ds(dmu_recv_cookie_t *drc)
{
if (drc->drc_resumable) {
/* wait for our resume state to be written to disk */
txg_wait_synced(drc->drc_ds->ds_dir->dd_pool, 0);
dsl_dataset_disown(drc->drc_ds, dmu_recv_tag);
} else {
char name[ZFS_MAX_DATASET_NAME_LEN];
dsl_dataset_name(drc->drc_ds, name);
dsl_dataset_disown(drc->drc_ds, dmu_recv_tag);
(void) dsl_destroy_head(name);
}
}
static void
receive_cksum(struct receive_arg *ra, int len, void *buf)
{
if (ra->byteswap) {
(void) fletcher_4_incremental_byteswap(buf, len, &ra->cksum);
} else {
(void) fletcher_4_incremental_native(buf, len, &ra->cksum);
}
}
/*
* Read the payload into a buffer of size len, and update the current record's
* payload field.
* Allocate ra->next_rrd and read the next record's header into
* ra->next_rrd->header.
* Verify checksum of payload and next record.
*/
static int
receive_read_payload_and_next_header(struct receive_arg *ra, int len, void *buf)
{
int err;
if (len != 0) {
ASSERT3U(len, <=, SPA_MAXBLOCKSIZE);
err = receive_read(ra, len, buf);
if (err != 0)
return (err);
receive_cksum(ra, len, buf);
/* note: rrd is NULL when reading the begin record's payload */
if (ra->rrd != NULL) {
ra->rrd->payload = buf;
ra->rrd->payload_size = len;
ra->rrd->bytes_read = ra->bytes_read;
}
}
ra->prev_cksum = ra->cksum;
ra->next_rrd = kmem_zalloc(sizeof (*ra->next_rrd), KM_SLEEP);
err = receive_read(ra, sizeof (ra->next_rrd->header),
&ra->next_rrd->header);
ra->next_rrd->bytes_read = ra->bytes_read;
if (err != 0) {
kmem_free(ra->next_rrd, sizeof (*ra->next_rrd));
ra->next_rrd = NULL;
return (err);
}
if (ra->next_rrd->header.drr_type == DRR_BEGIN) {
kmem_free(ra->next_rrd, sizeof (*ra->next_rrd));
ra->next_rrd = NULL;
return (SET_ERROR(EINVAL));
}
/*
* Note: checksum is of everything up to but not including the
* checksum itself.
*/
ASSERT3U(offsetof(dmu_replay_record_t, drr_u.drr_checksum.drr_checksum),
==, sizeof (dmu_replay_record_t) - sizeof (zio_cksum_t));
receive_cksum(ra,
offsetof(dmu_replay_record_t, drr_u.drr_checksum.drr_checksum),
&ra->next_rrd->header);
zio_cksum_t cksum_orig =
ra->next_rrd->header.drr_u.drr_checksum.drr_checksum;
zio_cksum_t *cksump =
&ra->next_rrd->header.drr_u.drr_checksum.drr_checksum;
if (ra->byteswap)
byteswap_record(&ra->next_rrd->header);
if ((!ZIO_CHECKSUM_IS_ZERO(cksump)) &&
!ZIO_CHECKSUM_EQUAL(ra->cksum, *cksump)) {
kmem_free(ra->next_rrd, sizeof (*ra->next_rrd));
ra->next_rrd = NULL;
return (SET_ERROR(ECKSUM));
}
receive_cksum(ra, sizeof (cksum_orig), &cksum_orig);
return (0);
}
static void
objlist_create(struct objlist *list)
{
list_create(&list->list, sizeof (struct receive_objnode),
offsetof(struct receive_objnode, node));
list->last_lookup = 0;
}
static void
objlist_destroy(struct objlist *list)
{
for (struct receive_objnode *n = list_remove_head(&list->list);
n != NULL; n = list_remove_head(&list->list)) {
kmem_free(n, sizeof (*n));
}
list_destroy(&list->list);
}
/*
* This function looks through the objlist to see if the specified object number
* is contained in the objlist. In the process, it will remove all object
* numbers in the list that are smaller than the specified object number. Thus,
* any lookup of an object number smaller than a previously looked up object
* number will always return false; therefore, all lookups should be done in
* ascending order.
*/
static boolean_t
objlist_exists(struct objlist *list, uint64_t object)
{
struct receive_objnode *node = list_head(&list->list);
ASSERT3U(object, >=, list->last_lookup);
list->last_lookup = object;
while (node != NULL && node->object < object) {
VERIFY3P(node, ==, list_remove_head(&list->list));
kmem_free(node, sizeof (*node));
node = list_head(&list->list);
}
return (node != NULL && node->object == object);
}
/*
* The objlist is a list of object numbers stored in ascending order. However,
* the insertion of new object numbers does not seek out the correct location to
* store a new object number; instead, it appends it to the list for simplicity.
* Thus, any users must take care to only insert new object numbers in ascending
* order.
*/
static void
objlist_insert(struct objlist *list, uint64_t object)
{
struct receive_objnode *node = kmem_zalloc(sizeof (*node), KM_SLEEP);
node->object = object;
#ifdef ZFS_DEBUG
struct receive_objnode *last_object = list_tail(&list->list);
uint64_t last_objnum = (last_object != NULL ? last_object->object : 0);
ASSERT3U(node->object, >, last_objnum);
#endif
list_insert_tail(&list->list, node);
}
/*
* Issue the prefetch reads for any necessary indirect blocks.
*
* We use the object ignore list to tell us whether or not to issue prefetches
* for a given object. We do this for both correctness (in case the blocksize
* of an object has changed) and performance (if the object doesn't exist, don't
* needlessly try to issue prefetches). We also trim the list as we go through
* the stream to prevent it from growing to an unbounded size.
*
* The object numbers within will always be in sorted order, and any write
* records we see will also be in sorted order, but they're not sorted with
* respect to each other (i.e. we can get several object records before
* receiving each object's write records). As a result, once we've reached a
* given object number, we can safely remove any reference to lower object
* numbers in the ignore list. In practice, we receive up to 32 object records
* before receiving write records, so the list can have up to 32 nodes in it.
*/
/* ARGSUSED */
static void
receive_read_prefetch(struct receive_arg *ra,
uint64_t object, uint64_t offset, uint64_t length)
{
if (!objlist_exists(&ra->ignore_objlist, object)) {
dmu_prefetch(ra->os, object, 1, offset, length,
ZIO_PRIORITY_SYNC_READ);
}
}
/*
* Read records off the stream, issuing any necessary prefetches.
*/
static int
receive_read_record(struct receive_arg *ra)
{
int err;
switch (ra->rrd->header.drr_type) {
case DRR_OBJECT:
{
struct drr_object *drro = &ra->rrd->header.drr_u.drr_object;
uint32_t size = P2ROUNDUP(drro->drr_bonuslen, 8);
void *buf = kmem_zalloc(size, KM_SLEEP);
dmu_object_info_t doi;
err = receive_read_payload_and_next_header(ra, size, buf);
if (err != 0) {
kmem_free(buf, size);
return (err);
}
err = dmu_object_info(ra->os, drro->drr_object, &doi);
/*
* See receive_read_prefetch for an explanation why we're
* storing this object in the ignore_obj_list.
*/
if (err == ENOENT ||
(err == 0 && doi.doi_data_block_size != drro->drr_blksz)) {
objlist_insert(&ra->ignore_objlist, drro->drr_object);
err = 0;
}
return (err);
}
case DRR_FREEOBJECTS:
{
err = receive_read_payload_and_next_header(ra, 0, NULL);
return (err);
}
case DRR_WRITE:
{
struct drr_write *drrw = &ra->rrd->header.drr_u.drr_write;
arc_buf_t *abuf;
boolean_t is_meta = DMU_OT_IS_METADATA(drrw->drr_type);
if (DRR_WRITE_COMPRESSED(drrw)) {
ASSERT3U(drrw->drr_compressed_size, >, 0);
ASSERT3U(drrw->drr_logical_size, >=,
drrw->drr_compressed_size);
ASSERT(!is_meta);
abuf = arc_loan_compressed_buf(
dmu_objset_spa(ra->os),
drrw->drr_compressed_size, drrw->drr_logical_size,
drrw->drr_compressiontype);
} else {
abuf = arc_loan_buf(dmu_objset_spa(ra->os),
is_meta, drrw->drr_logical_size);
}
err = receive_read_payload_and_next_header(ra,
DRR_WRITE_PAYLOAD_SIZE(drrw), abuf->b_data);
if (err != 0) {
dmu_return_arcbuf(abuf);
return (err);
}
ra->rrd->write_buf = abuf;
receive_read_prefetch(ra, drrw->drr_object, drrw->drr_offset,
drrw->drr_logical_size);
return (err);
}
case DRR_WRITE_BYREF:
{
struct drr_write_byref *drrwb =
&ra->rrd->header.drr_u.drr_write_byref;
err = receive_read_payload_and_next_header(ra, 0, NULL);
receive_read_prefetch(ra, drrwb->drr_object, drrwb->drr_offset,
drrwb->drr_length);
return (err);
}
case DRR_WRITE_EMBEDDED:
{
struct drr_write_embedded *drrwe =
&ra->rrd->header.drr_u.drr_write_embedded;
uint32_t size = P2ROUNDUP(drrwe->drr_psize, 8);
void *buf = kmem_zalloc(size, KM_SLEEP);
err = receive_read_payload_and_next_header(ra, size, buf);
if (err != 0) {
kmem_free(buf, size);
return (err);
}
receive_read_prefetch(ra, drrwe->drr_object, drrwe->drr_offset,
drrwe->drr_length);
return (err);
}
case DRR_FREE:
{
/*
* It might be beneficial to prefetch indirect blocks here, but
* we don't really have the data to decide for sure.
*/
err = receive_read_payload_and_next_header(ra, 0, NULL);
return (err);
}
case DRR_END:
{
struct drr_end *drre = &ra->rrd->header.drr_u.drr_end;
if (!ZIO_CHECKSUM_EQUAL(ra->prev_cksum, drre->drr_checksum))
return (SET_ERROR(ECKSUM));
return (0);
}
case DRR_SPILL:
{
struct drr_spill *drrs = &ra->rrd->header.drr_u.drr_spill;
void *buf = kmem_zalloc(drrs->drr_length, KM_SLEEP);
err = receive_read_payload_and_next_header(ra, drrs->drr_length,
buf);
if (err != 0)
kmem_free(buf, drrs->drr_length);
return (err);
}
default:
return (SET_ERROR(EINVAL));
}
}
/*
* Commit the records to the pool.
*/
static int
receive_process_record(struct receive_writer_arg *rwa,
struct receive_record_arg *rrd)
{
int err;
/* Processing in order, therefore bytes_read should be increasing. */
ASSERT3U(rrd->bytes_read, >=, rwa->bytes_read);
rwa->bytes_read = rrd->bytes_read;
switch (rrd->header.drr_type) {
case DRR_OBJECT:
{
struct drr_object *drro = &rrd->header.drr_u.drr_object;
err = receive_object(rwa, drro, rrd->payload);
kmem_free(rrd->payload, rrd->payload_size);
rrd->payload = NULL;
return (err);
}
case DRR_FREEOBJECTS:
{
struct drr_freeobjects *drrfo =
&rrd->header.drr_u.drr_freeobjects;
return (receive_freeobjects(rwa, drrfo));
}
case DRR_WRITE:
{
struct drr_write *drrw = &rrd->header.drr_u.drr_write;
err = receive_write(rwa, drrw, rrd->write_buf);
/* if receive_write() is successful, it consumes the arc_buf */
if (err != 0)
dmu_return_arcbuf(rrd->write_buf);
rrd->write_buf = NULL;
rrd->payload = NULL;
return (err);
}
case DRR_WRITE_BYREF:
{
struct drr_write_byref *drrwbr =
&rrd->header.drr_u.drr_write_byref;
return (receive_write_byref(rwa, drrwbr));
}
case DRR_WRITE_EMBEDDED:
{
struct drr_write_embedded *drrwe =
&rrd->header.drr_u.drr_write_embedded;
err = receive_write_embedded(rwa, drrwe, rrd->payload);
kmem_free(rrd->payload, rrd->payload_size);
rrd->payload = NULL;
return (err);
}
case DRR_FREE:
{
struct drr_free *drrf = &rrd->header.drr_u.drr_free;
return (receive_free(rwa, drrf));
}
case DRR_SPILL:
{
struct drr_spill *drrs = &rrd->header.drr_u.drr_spill;
err = receive_spill(rwa, drrs, rrd->payload);
kmem_free(rrd->payload, rrd->payload_size);
rrd->payload = NULL;
return (err);
}
default:
return (SET_ERROR(EINVAL));
}
}
/*
* dmu_recv_stream's worker thread; pull records off the queue, and then call
* receive_process_record When we're done, signal the main thread and exit.
*/
static void
receive_writer_thread(void *arg)
{
struct receive_writer_arg *rwa = arg;
struct receive_record_arg *rrd;
for (rrd = bqueue_dequeue(&rwa->q); !rrd->eos_marker;
rrd = bqueue_dequeue(&rwa->q)) {
/*
* If there's an error, the main thread will stop putting things
* on the queue, but we need to clear everything in it before we
* can exit.
*/
if (rwa->err == 0) {
rwa->err = receive_process_record(rwa, rrd);
} else if (rrd->write_buf != NULL) {
dmu_return_arcbuf(rrd->write_buf);
rrd->write_buf = NULL;
rrd->payload = NULL;
} else if (rrd->payload != NULL) {
kmem_free(rrd->payload, rrd->payload_size);
rrd->payload = NULL;
}
kmem_free(rrd, sizeof (*rrd));
}
kmem_free(rrd, sizeof (*rrd));
mutex_enter(&rwa->mutex);
rwa->done = B_TRUE;
cv_signal(&rwa->cv);
mutex_exit(&rwa->mutex);
thread_exit();
}
static int
resume_check(struct receive_arg *ra, nvlist_t *begin_nvl)
{
uint64_t val;
objset_t *mos = dmu_objset_pool(ra->os)->dp_meta_objset;
uint64_t dsobj = dmu_objset_id(ra->os);
uint64_t resume_obj, resume_off;
if (nvlist_lookup_uint64(begin_nvl,
"resume_object", &resume_obj) != 0 ||
nvlist_lookup_uint64(begin_nvl,
"resume_offset", &resume_off) != 0) {
return (SET_ERROR(EINVAL));
}
VERIFY0(zap_lookup(mos, dsobj,
DS_FIELD_RESUME_OBJECT, sizeof (val), 1, &val));
if (resume_obj != val)
return (SET_ERROR(EINVAL));
VERIFY0(zap_lookup(mos, dsobj,
DS_FIELD_RESUME_OFFSET, sizeof (val), 1, &val));
if (resume_off != val)
return (SET_ERROR(EINVAL));
return (0);
}
/*
* Read in the stream's records, one by one, and apply them to the pool. There
* are two threads involved; the thread that calls this function will spin up a
* worker thread, read the records off the stream one by one, and issue
* prefetches for any necessary indirect blocks. It will then push the records
* onto an internal blocking queue. The worker thread will pull the records off
* the queue, and actually write the data into the DMU. This way, the worker
* thread doesn't have to wait for reads to complete, since everything it needs
* (the indirect blocks) will be prefetched.
*
* NB: callers *must* call dmu_recv_end() if this succeeds.
*/
int
dmu_recv_stream(dmu_recv_cookie_t *drc, struct file *fp, offset_t *voffp,
int cleanup_fd, uint64_t *action_handlep)
{
int err = 0;
struct receive_arg ra = { 0 };
struct receive_writer_arg rwa = { 0 };
int featureflags;
nvlist_t *begin_nvl = NULL;
ra.byteswap = drc->drc_byteswap;
ra.cksum = drc->drc_cksum;
ra.td = curthread;
ra.fp = fp;
ra.voff = *voffp;
if (dsl_dataset_is_zapified(drc->drc_ds)) {
(void) zap_lookup(drc->drc_ds->ds_dir->dd_pool->dp_meta_objset,
drc->drc_ds->ds_object, DS_FIELD_RESUME_BYTES,
sizeof (ra.bytes_read), 1, &ra.bytes_read);
}
objlist_create(&ra.ignore_objlist);
/* these were verified in dmu_recv_begin */
ASSERT3U(DMU_GET_STREAM_HDRTYPE(drc->drc_drrb->drr_versioninfo), ==,
DMU_SUBSTREAM);
ASSERT3U(drc->drc_drrb->drr_type, <, DMU_OST_NUMTYPES);
/*
* Open the objset we are modifying.
*/
VERIFY0(dmu_objset_from_ds(drc->drc_ds, &ra.os));
ASSERT(dsl_dataset_phys(drc->drc_ds)->ds_flags & DS_FLAG_INCONSISTENT);
featureflags = DMU_GET_FEATUREFLAGS(drc->drc_drrb->drr_versioninfo);
/* if this stream is dedup'ed, set up the avl tree for guid mapping */
if (featureflags & DMU_BACKUP_FEATURE_DEDUP) {
minor_t minor;
if (cleanup_fd == -1) {
ra.err = SET_ERROR(EBADF);
goto out;
}
ra.err = zfs_onexit_fd_hold(cleanup_fd, &minor);
if (ra.err != 0) {
cleanup_fd = -1;
goto out;
}
if (*action_handlep == 0) {
rwa.guid_to_ds_map =
kmem_alloc(sizeof (avl_tree_t), KM_SLEEP);
avl_create(rwa.guid_to_ds_map, guid_compare,
sizeof (guid_map_entry_t),
offsetof(guid_map_entry_t, avlnode));
err = zfs_onexit_add_cb(minor,
free_guid_map_onexit, rwa.guid_to_ds_map,
action_handlep);
if (ra.err != 0)
goto out;
} else {
err = zfs_onexit_cb_data(minor, *action_handlep,
(void **)&rwa.guid_to_ds_map);
if (ra.err != 0)
goto out;
}
drc->drc_guid_to_ds_map = rwa.guid_to_ds_map;
}
uint32_t payloadlen = drc->drc_drr_begin->drr_payloadlen;
void *payload = NULL;
if (payloadlen != 0)
payload = kmem_alloc(payloadlen, KM_SLEEP);
err = receive_read_payload_and_next_header(&ra, payloadlen, payload);
if (err != 0) {
if (payloadlen != 0)
kmem_free(payload, payloadlen);
goto out;
}
if (payloadlen != 0) {
err = nvlist_unpack(payload, payloadlen, &begin_nvl, KM_SLEEP);
kmem_free(payload, payloadlen);
if (err != 0)
goto out;
}
if (featureflags & DMU_BACKUP_FEATURE_RESUMING) {
err = resume_check(&ra, begin_nvl);
if (err != 0)
goto out;
}
(void) bqueue_init(&rwa.q, zfs_recv_queue_length,
offsetof(struct receive_record_arg, node));
cv_init(&rwa.cv, NULL, CV_DEFAULT, NULL);
mutex_init(&rwa.mutex, NULL, MUTEX_DEFAULT, NULL);
rwa.os = ra.os;
rwa.byteswap = drc->drc_byteswap;
rwa.resumable = drc->drc_resumable;
(void) thread_create(NULL, 0, receive_writer_thread, &rwa, 0, &p0,
TS_RUN, minclsyspri);
/*
* We're reading rwa.err without locks, which is safe since we are the
* only reader, and the worker thread is the only writer. It's ok if we
* miss a write for an iteration or two of the loop, since the writer
* thread will keep freeing records we send it until we send it an eos
* marker.
*
* We can leave this loop in 3 ways: First, if rwa.err is
* non-zero. In that case, the writer thread will free the rrd we just
* pushed. Second, if we're interrupted; in that case, either it's the
* first loop and ra.rrd was never allocated, or it's later, and ra.rrd
* has been handed off to the writer thread who will free it. Finally,
* if receive_read_record fails or we're at the end of the stream, then
* we free ra.rrd and exit.
*/
while (rwa.err == 0) {
if (issig(JUSTLOOKING) && issig(FORREAL)) {
err = SET_ERROR(EINTR);
break;
}
ASSERT3P(ra.rrd, ==, NULL);
ra.rrd = ra.next_rrd;
ra.next_rrd = NULL;
/* Allocates and loads header into ra.next_rrd */
err = receive_read_record(&ra);
if (ra.rrd->header.drr_type == DRR_END || err != 0) {
kmem_free(ra.rrd, sizeof (*ra.rrd));
ra.rrd = NULL;
break;
}
bqueue_enqueue(&rwa.q, ra.rrd,
sizeof (struct receive_record_arg) + ra.rrd->payload_size);
ra.rrd = NULL;
}
if (ra.next_rrd == NULL)
ra.next_rrd = kmem_zalloc(sizeof (*ra.next_rrd), KM_SLEEP);
ra.next_rrd->eos_marker = B_TRUE;
bqueue_enqueue(&rwa.q, ra.next_rrd, 1);
mutex_enter(&rwa.mutex);
while (!rwa.done) {
cv_wait(&rwa.cv, &rwa.mutex);
}
mutex_exit(&rwa.mutex);
/*
* If we are receiving a full stream as a clone, all object IDs which
* are greater than the maximum ID referenced in the stream are
* by definition unused and must be freed. Note that it's possible that
* we've resumed this send and the first record we received was the END
* record. In that case, max_object would be 0, but we shouldn't start
* freeing all objects from there; instead we should start from the
* resumeobj.
*/
if (drc->drc_clone && drc->drc_drrb->drr_fromguid == 0) {
uint64_t obj;
if (nvlist_lookup_uint64(begin_nvl, "resume_object", &obj) != 0)
obj = 0;
if (rwa.max_object > obj)
obj = rwa.max_object;
obj++;
int free_err = 0;
int next_err = 0;
while (next_err == 0) {
free_err = dmu_free_long_object(rwa.os, obj);
if (free_err != 0 && free_err != ENOENT)
break;
next_err = dmu_object_next(rwa.os, &obj, FALSE, 0);
}
if (err == 0) {
if (free_err != 0 && free_err != ENOENT)
err = free_err;
else if (next_err != ESRCH)
err = next_err;
}
}
cv_destroy(&rwa.cv);
mutex_destroy(&rwa.mutex);
bqueue_destroy(&rwa.q);
if (err == 0)
err = rwa.err;
out:
nvlist_free(begin_nvl);
if ((featureflags & DMU_BACKUP_FEATURE_DEDUP) && (cleanup_fd != -1))
zfs_onexit_fd_rele(cleanup_fd);
if (err != 0) {
/*
* Clean up references. If receive is not resumable,
* destroy what we created, so we don't leave it in
* the inconsistent state.
*/
dmu_recv_cleanup_ds(drc);
}
*voffp = ra.voff;
objlist_destroy(&ra.ignore_objlist);
return (err);
}
static int
dmu_recv_end_check(void *arg, dmu_tx_t *tx)
{
dmu_recv_cookie_t *drc = arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
int error;
ASSERT3P(drc->drc_ds->ds_owner, ==, dmu_recv_tag);
if (!drc->drc_newfs) {
dsl_dataset_t *origin_head;
error = dsl_dataset_hold(dp, drc->drc_tofs, FTAG, &origin_head);
if (error != 0)
return (error);
if (drc->drc_force) {
/*
* We will destroy any snapshots in tofs (i.e. before
* origin_head) that are after the origin (which is
* the snap before drc_ds, because drc_ds can not
* have any snaps of its own).
*/
uint64_t obj;
obj = dsl_dataset_phys(origin_head)->ds_prev_snap_obj;
while (obj !=
dsl_dataset_phys(drc->drc_ds)->ds_prev_snap_obj) {
dsl_dataset_t *snap;
error = dsl_dataset_hold_obj(dp, obj, FTAG,
&snap);
if (error != 0)
break;
if (snap->ds_dir != origin_head->ds_dir)
error = SET_ERROR(EINVAL);
if (error == 0) {
error = dsl_destroy_snapshot_check_impl(
snap, B_FALSE);
}
obj = dsl_dataset_phys(snap)->ds_prev_snap_obj;
dsl_dataset_rele(snap, FTAG);
if (error != 0)
break;
}
if (error != 0) {
dsl_dataset_rele(origin_head, FTAG);
return (error);
}
}
error = dsl_dataset_clone_swap_check_impl(drc->drc_ds,
origin_head, drc->drc_force, drc->drc_owner, tx);
if (error != 0) {
dsl_dataset_rele(origin_head, FTAG);
return (error);
}
error = dsl_dataset_snapshot_check_impl(origin_head,
drc->drc_tosnap, tx, B_TRUE, 1, drc->drc_cred);
dsl_dataset_rele(origin_head, FTAG);
if (error != 0)
return (error);
error = dsl_destroy_head_check_impl(drc->drc_ds, 1);
} else {
error = dsl_dataset_snapshot_check_impl(drc->drc_ds,
drc->drc_tosnap, tx, B_TRUE, 1, drc->drc_cred);
}
return (error);
}
static void
dmu_recv_end_sync(void *arg, dmu_tx_t *tx)
{
dmu_recv_cookie_t *drc = arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
spa_history_log_internal_ds(drc->drc_ds, "finish receiving",
tx, "snap=%s", drc->drc_tosnap);
if (!drc->drc_newfs) {
dsl_dataset_t *origin_head;
VERIFY0(dsl_dataset_hold(dp, drc->drc_tofs, FTAG,
&origin_head));
if (drc->drc_force) {
/*
* Destroy any snapshots of drc_tofs (origin_head)
* after the origin (the snap before drc_ds).
*/
uint64_t obj;
obj = dsl_dataset_phys(origin_head)->ds_prev_snap_obj;
while (obj !=
dsl_dataset_phys(drc->drc_ds)->ds_prev_snap_obj) {
dsl_dataset_t *snap;
VERIFY0(dsl_dataset_hold_obj(dp, obj, FTAG,
&snap));
ASSERT3P(snap->ds_dir, ==, origin_head->ds_dir);
obj = dsl_dataset_phys(snap)->ds_prev_snap_obj;
dsl_destroy_snapshot_sync_impl(snap,
B_FALSE, tx);
dsl_dataset_rele(snap, FTAG);
}
}
VERIFY3P(drc->drc_ds->ds_prev, ==,
origin_head->ds_prev);
dsl_dataset_clone_swap_sync_impl(drc->drc_ds,
origin_head, tx);
dsl_dataset_snapshot_sync_impl(origin_head,
drc->drc_tosnap, tx);
/* set snapshot's creation time and guid */
dmu_buf_will_dirty(origin_head->ds_prev->ds_dbuf, tx);
dsl_dataset_phys(origin_head->ds_prev)->ds_creation_time =
drc->drc_drrb->drr_creation_time;
dsl_dataset_phys(origin_head->ds_prev)->ds_guid =
drc->drc_drrb->drr_toguid;
dsl_dataset_phys(origin_head->ds_prev)->ds_flags &=
~DS_FLAG_INCONSISTENT;
dmu_buf_will_dirty(origin_head->ds_dbuf, tx);
dsl_dataset_phys(origin_head)->ds_flags &=
~DS_FLAG_INCONSISTENT;
drc->drc_newsnapobj =
dsl_dataset_phys(origin_head)->ds_prev_snap_obj;
dsl_dataset_rele(origin_head, FTAG);
dsl_destroy_head_sync_impl(drc->drc_ds, tx);
if (drc->drc_owner != NULL)
VERIFY3P(origin_head->ds_owner, ==, drc->drc_owner);
} else {
dsl_dataset_t *ds = drc->drc_ds;
dsl_dataset_snapshot_sync_impl(ds, drc->drc_tosnap, tx);
/* set snapshot's creation time and guid */
dmu_buf_will_dirty(ds->ds_prev->ds_dbuf, tx);
dsl_dataset_phys(ds->ds_prev)->ds_creation_time =
drc->drc_drrb->drr_creation_time;
dsl_dataset_phys(ds->ds_prev)->ds_guid =
drc->drc_drrb->drr_toguid;
dsl_dataset_phys(ds->ds_prev)->ds_flags &=
~DS_FLAG_INCONSISTENT;
dmu_buf_will_dirty(ds->ds_dbuf, tx);
dsl_dataset_phys(ds)->ds_flags &= ~DS_FLAG_INCONSISTENT;
if (dsl_dataset_has_resume_receive_state(ds)) {
(void) zap_remove(dp->dp_meta_objset, ds->ds_object,
DS_FIELD_RESUME_FROMGUID, tx);
(void) zap_remove(dp->dp_meta_objset, ds->ds_object,
DS_FIELD_RESUME_OBJECT, tx);
(void) zap_remove(dp->dp_meta_objset, ds->ds_object,
DS_FIELD_RESUME_OFFSET, tx);
(void) zap_remove(dp->dp_meta_objset, ds->ds_object,
DS_FIELD_RESUME_BYTES, tx);
(void) zap_remove(dp->dp_meta_objset, ds->ds_object,
DS_FIELD_RESUME_TOGUID, tx);
(void) zap_remove(dp->dp_meta_objset, ds->ds_object,
DS_FIELD_RESUME_TONAME, tx);
}
drc->drc_newsnapobj =
dsl_dataset_phys(drc->drc_ds)->ds_prev_snap_obj;
}
/*
* Release the hold from dmu_recv_begin. This must be done before
* we return to open context, so that when we free the dataset's dnode,
* we can evict its bonus buffer.
*/
dsl_dataset_disown(drc->drc_ds, dmu_recv_tag);
drc->drc_ds = NULL;
}
static int
add_ds_to_guidmap(const char *name, avl_tree_t *guid_map, uint64_t snapobj)
{
dsl_pool_t *dp;
dsl_dataset_t *snapds;
guid_map_entry_t *gmep;
int err;
ASSERT(guid_map != NULL);
err = dsl_pool_hold(name, FTAG, &dp);
if (err != 0)
return (err);
gmep = kmem_alloc(sizeof (*gmep), KM_SLEEP);
err = dsl_dataset_hold_obj(dp, snapobj, gmep, &snapds);
if (err == 0) {
gmep->guid = dsl_dataset_phys(snapds)->ds_guid;
gmep->gme_ds = snapds;
avl_add(guid_map, gmep);
dsl_dataset_long_hold(snapds, gmep);
} else
kmem_free(gmep, sizeof (*gmep));
dsl_pool_rele(dp, FTAG);
return (err);
}
static int dmu_recv_end_modified_blocks = 3;
static int
dmu_recv_existing_end(dmu_recv_cookie_t *drc)
{
#ifdef _KERNEL
/*
* We will be destroying the ds; make sure its origin is unmounted if
* necessary.
*/
char name[ZFS_MAX_DATASET_NAME_LEN];
dsl_dataset_name(drc->drc_ds, name);
zfs_destroy_unmount_origin(name);
#endif
return (dsl_sync_task(drc->drc_tofs,
dmu_recv_end_check, dmu_recv_end_sync, drc,
dmu_recv_end_modified_blocks, ZFS_SPACE_CHECK_NORMAL));
}
static int
dmu_recv_new_end(dmu_recv_cookie_t *drc)
{
return (dsl_sync_task(drc->drc_tofs,
dmu_recv_end_check, dmu_recv_end_sync, drc,
dmu_recv_end_modified_blocks, ZFS_SPACE_CHECK_NORMAL));
}
int
dmu_recv_end(dmu_recv_cookie_t *drc, void *owner)
{
int error;
drc->drc_owner = owner;
if (drc->drc_newfs)
error = dmu_recv_new_end(drc);
else
error = dmu_recv_existing_end(drc);
if (error != 0) {
dmu_recv_cleanup_ds(drc);
} else if (drc->drc_guid_to_ds_map != NULL) {
(void) add_ds_to_guidmap(drc->drc_tofs,
drc->drc_guid_to_ds_map,
drc->drc_newsnapobj);
}
return (error);
}
/*
* Return TRUE if this objset is currently being received into.
*/
boolean_t
dmu_objset_is_receiving(objset_t *os)
{
return (os->os_dsl_dataset != NULL &&
os->os_dsl_dataset->ds_owner == dmu_recv_tag);
}
Index: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_dir.c
===================================================================
--- head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_dir.c (revision 359302)
+++ head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/dsl_dir.c (revision 359303)
@@ -1,2158 +1,2184 @@
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2011 Pawel Jakub Dawidek <pawel@dawidek.net>.
* All rights reserved.
* Copyright (c) 2012, 2017 by Delphix. All rights reserved.
* Copyright (c) 2014 Joyent, Inc. All rights reserved.
* Copyright (c) 2014 Spectra Logic Corporation, All rights reserved.
* Copyright 2015 Nexenta Systems, Inc. All rights reserved.
+ * Copyright (c) 2018, loli10K <ezomori.nozomu@gmail.com>. All rights reserved.
*/
#include <sys/dmu.h>
#include <sys/dmu_objset.h>
#include <sys/dmu_tx.h>
#include <sys/dsl_dataset.h>
#include <sys/dsl_dir.h>
#include <sys/dsl_prop.h>
#include <sys/dsl_synctask.h>
#include <sys/dsl_deleg.h>
#include <sys/dmu_impl.h>
#include <sys/spa.h>
#include <sys/metaslab.h>
#include <sys/zap.h>
#include <sys/zio.h>
#include <sys/arc.h>
#include <sys/sunddi.h>
#include <sys/zvol.h>
#ifdef _KERNEL
#include <sys/zfs_vfsops.h>
#endif
#include <sys/zfeature.h>
#include <sys/policy.h>
#include <sys/zfs_znode.h>
#include "zfs_namecheck.h"
#include "zfs_prop.h"
/*
* Filesystem and Snapshot Limits
* ------------------------------
*
* These limits are used to restrict the number of filesystems and/or snapshots
* that can be created at a given level in the tree or below. A typical
* use-case is with a delegated dataset where the administrator wants to ensure
* that a user within the zone is not creating too many additional filesystems
* or snapshots, even though they're not exceeding their space quota.
*
* The filesystem and snapshot counts are stored as extensible properties. This
* capability is controlled by a feature flag and must be enabled to be used.
* Once enabled, the feature is not active until the first limit is set. At
* that point, future operations to create/destroy filesystems or snapshots
* will validate and update the counts.
*
* Because the count properties will not exist before the feature is active,
* the counts are updated when a limit is first set on an uninitialized
* dsl_dir node in the tree (The filesystem/snapshot count on a node includes
* all of the nested filesystems/snapshots. Thus, a new leaf node has a
* filesystem count of 0 and a snapshot count of 0. Non-existent filesystem and
* snapshot count properties on a node indicate uninitialized counts on that
* node.) When first setting a limit on an uninitialized node, the code starts
* at the filesystem with the new limit and descends into all sub-filesystems
* to add the count properties.
*
* In practice this is lightweight since a limit is typically set when the
* filesystem is created and thus has no children. Once valid, changing the
* limit value won't require a re-traversal since the counts are already valid.
* When recursively fixing the counts, if a node with a limit is encountered
* during the descent, the counts are known to be valid and there is no need to
* descend into that filesystem's children. The counts on filesystems above the
* one with the new limit will still be uninitialized, unless a limit is
* eventually set on one of those filesystems. The counts are always recursively
* updated when a limit is set on a dataset, unless there is already a limit.
* When a new limit value is set on a filesystem with an existing limit, it is
* possible for the new limit to be less than the current count at that level
* since a user who can change the limit is also allowed to exceed the limit.
*
* Once the feature is active, then whenever a filesystem or snapshot is
* created, the code recurses up the tree, validating the new count against the
* limit at each initialized level. In practice, most levels will not have a
* limit set. If there is a limit at any initialized level up the tree, the
* check must pass or the creation will fail. Likewise, when a filesystem or
* snapshot is destroyed, the counts are recursively adjusted all the way up
* the initizized nodes in the tree. Renaming a filesystem into different point
* in the tree will first validate, then update the counts on each branch up to
* the common ancestor. A receive will also validate the counts and then update
* them.
*
* An exception to the above behavior is that the limit is not enforced if the
* user has permission to modify the limit. This is primarily so that
* recursive snapshots in the global zone always work. We want to prevent a
* denial-of-service in which a lower level delegated dataset could max out its
* limit and thus block recursive snapshots from being taken in the global zone.
* Because of this, it is possible for the snapshot count to be over the limit
* and snapshots taken in the global zone could cause a lower level dataset to
* hit or exceed its limit. The administrator taking the global zone recursive
* snapshot should be aware of this side-effect and behave accordingly.
* For consistency, the filesystem limit is also not enforced if the user can
* modify the limit.
*
* The filesystem and snapshot limits are validated by dsl_fs_ss_limit_check()
* and updated by dsl_fs_ss_count_adjust(). A new limit value is setup in
* dsl_dir_activate_fs_ss_limit() and the counts are adjusted, if necessary, by
* dsl_dir_init_fs_ss_count().
*
* There is a special case when we receive a filesystem that already exists. In
* this case a temporary clone name of %X is created (see dmu_recv_begin). We
* never update the filesystem counts for temporary clones.
*
* Likewise, we do not update the snapshot counts for temporary snapshots,
* such as those created by zfs diff.
*/
extern inline dsl_dir_phys_t *dsl_dir_phys(dsl_dir_t *dd);
static uint64_t dsl_dir_space_towrite(dsl_dir_t *dd);
typedef struct ddulrt_arg {
dsl_dir_t *ddulrta_dd;
uint64_t ddlrta_txg;
} ddulrt_arg_t;
static void
dsl_dir_evict_async(void *dbu)
{
dsl_dir_t *dd = dbu;
dsl_pool_t *dp = dd->dd_pool;
int t;
dd->dd_dbuf = NULL;
for (t = 0; t < TXG_SIZE; t++) {
ASSERT(!txg_list_member(&dp->dp_dirty_dirs, dd, t));
ASSERT(dd->dd_tempreserved[t] == 0);
ASSERT(dd->dd_space_towrite[t] == 0);
}
if (dd->dd_parent)
dsl_dir_async_rele(dd->dd_parent, dd);
spa_async_close(dd->dd_pool->dp_spa, dd);
dsl_prop_fini(dd);
mutex_destroy(&dd->dd_lock);
kmem_free(dd, sizeof (dsl_dir_t));
}
int
dsl_dir_hold_obj(dsl_pool_t *dp, uint64_t ddobj,
const char *tail, void *tag, dsl_dir_t **ddp)
{
dmu_buf_t *dbuf;
dsl_dir_t *dd;
int err;
ASSERT(dsl_pool_config_held(dp));
err = dmu_bonus_hold(dp->dp_meta_objset, ddobj, tag, &dbuf);
if (err != 0)
return (err);
dd = dmu_buf_get_user(dbuf);
#ifdef ZFS_DEBUG
{
dmu_object_info_t doi;
dmu_object_info_from_db(dbuf, &doi);
ASSERT3U(doi.doi_bonus_type, ==, DMU_OT_DSL_DIR);
ASSERT3U(doi.doi_bonus_size, >=, sizeof (dsl_dir_phys_t));
}
#endif
if (dd == NULL) {
dsl_dir_t *winner;
dd = kmem_zalloc(sizeof (dsl_dir_t), KM_SLEEP);
dd->dd_object = ddobj;
dd->dd_dbuf = dbuf;
dd->dd_pool = dp;
mutex_init(&dd->dd_lock, NULL, MUTEX_DEFAULT, NULL);
dsl_prop_init(dd);
dsl_dir_snap_cmtime_update(dd);
if (dsl_dir_phys(dd)->dd_parent_obj) {
err = dsl_dir_hold_obj(dp,
dsl_dir_phys(dd)->dd_parent_obj, NULL, dd,
&dd->dd_parent);
if (err != 0)
goto errout;
if (tail) {
#ifdef ZFS_DEBUG
uint64_t foundobj;
err = zap_lookup(dp->dp_meta_objset,
dsl_dir_phys(dd->dd_parent)->
dd_child_dir_zapobj, tail,
sizeof (foundobj), 1, &foundobj);
ASSERT(err || foundobj == ddobj);
#endif
(void) strcpy(dd->dd_myname, tail);
} else {
err = zap_value_search(dp->dp_meta_objset,
dsl_dir_phys(dd->dd_parent)->
dd_child_dir_zapobj,
ddobj, 0, dd->dd_myname);
}
if (err != 0)
goto errout;
} else {
(void) strcpy(dd->dd_myname, spa_name(dp->dp_spa));
}
if (dsl_dir_is_clone(dd)) {
dmu_buf_t *origin_bonus;
dsl_dataset_phys_t *origin_phys;
/*
* We can't open the origin dataset, because
* that would require opening this dsl_dir.
* Just look at its phys directly instead.
*/
err = dmu_bonus_hold(dp->dp_meta_objset,
dsl_dir_phys(dd)->dd_origin_obj, FTAG,
&origin_bonus);
if (err != 0)
goto errout;
origin_phys = origin_bonus->db_data;
dd->dd_origin_txg =
origin_phys->ds_creation_txg;
dmu_buf_rele(origin_bonus, FTAG);
}
dmu_buf_init_user(&dd->dd_dbu, NULL, dsl_dir_evict_async,
&dd->dd_dbuf);
winner = dmu_buf_set_user_ie(dbuf, &dd->dd_dbu);
if (winner != NULL) {
if (dd->dd_parent)
dsl_dir_rele(dd->dd_parent, dd);
dsl_prop_fini(dd);
mutex_destroy(&dd->dd_lock);
kmem_free(dd, sizeof (dsl_dir_t));
dd = winner;
} else {
spa_open_ref(dp->dp_spa, dd);
}
}
/*
* The dsl_dir_t has both open-to-close and instantiate-to-evict
* holds on the spa. We need the open-to-close holds because
* otherwise the spa_refcnt wouldn't change when we open a
* dir which the spa also has open, so we could incorrectly
* think it was OK to unload/export/destroy the pool. We need
* the instantiate-to-evict hold because the dsl_dir_t has a
* pointer to the dd_pool, which has a pointer to the spa_t.
*/
spa_open_ref(dp->dp_spa, tag);
ASSERT3P(dd->dd_pool, ==, dp);
ASSERT3U(dd->dd_object, ==, ddobj);
ASSERT3P(dd->dd_dbuf, ==, dbuf);
*ddp = dd;
return (0);
errout:
if (dd->dd_parent)
dsl_dir_rele(dd->dd_parent, dd);
dsl_prop_fini(dd);
mutex_destroy(&dd->dd_lock);
kmem_free(dd, sizeof (dsl_dir_t));
dmu_buf_rele(dbuf, tag);
return (err);
}
void
dsl_dir_rele(dsl_dir_t *dd, void *tag)
{
dprintf_dd(dd, "%s\n", "");
spa_close(dd->dd_pool->dp_spa, tag);
dmu_buf_rele(dd->dd_dbuf, tag);
}
/*
* Remove a reference to the given dsl dir that is being asynchronously
* released. Async releases occur from a taskq performing eviction of
* dsl datasets and dirs. This process is identical to a normal release
* with the exception of using the async API for releasing the reference on
* the spa.
*/
void
dsl_dir_async_rele(dsl_dir_t *dd, void *tag)
{
dprintf_dd(dd, "%s\n", "");
spa_async_close(dd->dd_pool->dp_spa, tag);
dmu_buf_rele(dd->dd_dbuf, tag);
}
/* buf must be at least ZFS_MAX_DATASET_NAME_LEN bytes */
void
dsl_dir_name(dsl_dir_t *dd, char *buf)
{
if (dd->dd_parent) {
dsl_dir_name(dd->dd_parent, buf);
VERIFY3U(strlcat(buf, "/", ZFS_MAX_DATASET_NAME_LEN), <,
ZFS_MAX_DATASET_NAME_LEN);
} else {
buf[0] = '\0';
}
if (!MUTEX_HELD(&dd->dd_lock)) {
/*
* recursive mutex so that we can use
* dprintf_dd() with dd_lock held
*/
mutex_enter(&dd->dd_lock);
VERIFY3U(strlcat(buf, dd->dd_myname, ZFS_MAX_DATASET_NAME_LEN),
<, ZFS_MAX_DATASET_NAME_LEN);
mutex_exit(&dd->dd_lock);
} else {
VERIFY3U(strlcat(buf, dd->dd_myname, ZFS_MAX_DATASET_NAME_LEN),
<, ZFS_MAX_DATASET_NAME_LEN);
}
}
/* Calculate name length, avoiding all the strcat calls of dsl_dir_name */
int
dsl_dir_namelen(dsl_dir_t *dd)
{
int result = 0;
if (dd->dd_parent) {
/* parent's name + 1 for the "/" */
result = dsl_dir_namelen(dd->dd_parent) + 1;
}
if (!MUTEX_HELD(&dd->dd_lock)) {
/* see dsl_dir_name */
mutex_enter(&dd->dd_lock);
result += strlen(dd->dd_myname);
mutex_exit(&dd->dd_lock);
} else {
result += strlen(dd->dd_myname);
}
return (result);
}
static int
getcomponent(const char *path, char *component, const char **nextp)
{
char *p;
if ((path == NULL) || (path[0] == '\0'))
return (SET_ERROR(ENOENT));
/* This would be a good place to reserve some namespace... */
p = strpbrk(path, "/@");
if (p && (p[1] == '/' || p[1] == '@')) {
/* two separators in a row */
return (SET_ERROR(EINVAL));
}
if (p == NULL || p == path) {
/*
* if the first thing is an @ or /, it had better be an
* @ and it had better not have any more ats or slashes,
* and it had better have something after the @.
*/
if (p != NULL &&
(p[0] != '@' || strpbrk(path+1, "/@") || p[1] == '\0'))
return (SET_ERROR(EINVAL));
if (strlen(path) >= ZFS_MAX_DATASET_NAME_LEN)
return (SET_ERROR(ENAMETOOLONG));
(void) strcpy(component, path);
p = NULL;
} else if (p[0] == '/') {
if (p - path >= ZFS_MAX_DATASET_NAME_LEN)
return (SET_ERROR(ENAMETOOLONG));
(void) strncpy(component, path, p - path);
component[p - path] = '\0';
p++;
} else if (p[0] == '@') {
/*
* if the next separator is an @, there better not be
* any more slashes.
*/
if (strchr(path, '/'))
return (SET_ERROR(EINVAL));
if (p - path >= ZFS_MAX_DATASET_NAME_LEN)
return (SET_ERROR(ENAMETOOLONG));
(void) strncpy(component, path, p - path);
component[p - path] = '\0';
} else {
panic("invalid p=%p", (void *)p);
}
*nextp = p;
return (0);
}
/*
* Return the dsl_dir_t, and possibly the last component which couldn't
* be found in *tail. The name must be in the specified dsl_pool_t. This
* thread must hold the dp_config_rwlock for the pool. Returns NULL if the
* path is bogus, or if tail==NULL and we couldn't parse the whole name.
* (*tail)[0] == '@' means that the last component is a snapshot.
*/
int
dsl_dir_hold(dsl_pool_t *dp, const char *name, void *tag,
dsl_dir_t **ddp, const char **tailp)
{
char buf[ZFS_MAX_DATASET_NAME_LEN];
const char *spaname, *next, *nextnext = NULL;
int err;
dsl_dir_t *dd;
uint64_t ddobj;
err = getcomponent(name, buf, &next);
if (err != 0)
return (err);
/* Make sure the name is in the specified pool. */
spaname = spa_name(dp->dp_spa);
if (strcmp(buf, spaname) != 0)
return (SET_ERROR(EXDEV));
ASSERT(dsl_pool_config_held(dp));
err = dsl_dir_hold_obj(dp, dp->dp_root_dir_obj, NULL, tag, &dd);
if (err != 0) {
return (err);
}
while (next != NULL) {
dsl_dir_t *child_dd;
err = getcomponent(next, buf, &nextnext);
if (err != 0)
break;
ASSERT(next[0] != '\0');
if (next[0] == '@')
break;
dprintf("looking up %s in obj%lld\n",
buf, dsl_dir_phys(dd)->dd_child_dir_zapobj);
err = zap_lookup(dp->dp_meta_objset,
dsl_dir_phys(dd)->dd_child_dir_zapobj,
buf, sizeof (ddobj), 1, &ddobj);
if (err != 0) {
if (err == ENOENT)
err = 0;
break;
}
err = dsl_dir_hold_obj(dp, ddobj, buf, tag, &child_dd);
if (err != 0)
break;
dsl_dir_rele(dd, tag);
dd = child_dd;
next = nextnext;
}
if (err != 0) {
dsl_dir_rele(dd, tag);
return (err);
}
/*
* It's an error if there's more than one component left, or
* tailp==NULL and there's any component left.
*/
if (next != NULL &&
(tailp == NULL || (nextnext && nextnext[0] != '\0'))) {
/* bad path name */
dsl_dir_rele(dd, tag);
dprintf("next=%p (%s) tail=%p\n", next, next?next:"", tailp);
err = SET_ERROR(ENOENT);
}
if (tailp != NULL)
*tailp = next;
*ddp = dd;
return (err);
}
/*
* If the counts are already initialized for this filesystem and its
* descendants then do nothing, otherwise initialize the counts.
*
* The counts on this filesystem, and those below, may be uninitialized due to
* either the use of a pre-existing pool which did not support the
* filesystem/snapshot limit feature, or one in which the feature had not yet
* been enabled.
*
* Recursively descend the filesystem tree and update the filesystem/snapshot
* counts on each filesystem below, then update the cumulative count on the
* current filesystem. If the filesystem already has a count set on it,
* then we know that its counts, and the counts on the filesystems below it,
* are already correct, so we don't have to update this filesystem.
*/
static void
dsl_dir_init_fs_ss_count(dsl_dir_t *dd, dmu_tx_t *tx)
{
uint64_t my_fs_cnt = 0;
uint64_t my_ss_cnt = 0;
dsl_pool_t *dp = dd->dd_pool;
objset_t *os = dp->dp_meta_objset;
zap_cursor_t *zc;
zap_attribute_t *za;
dsl_dataset_t *ds;
ASSERT(spa_feature_is_active(dp->dp_spa, SPA_FEATURE_FS_SS_LIMIT));
ASSERT(dsl_pool_config_held(dp));
ASSERT(dmu_tx_is_syncing(tx));
dsl_dir_zapify(dd, tx);
/*
* If the filesystem count has already been initialized then we
* don't need to recurse down any further.
*/
if (zap_contains(os, dd->dd_object, DD_FIELD_FILESYSTEM_COUNT) == 0)
return;
zc = kmem_alloc(sizeof (zap_cursor_t), KM_SLEEP);
za = kmem_alloc(sizeof (zap_attribute_t), KM_SLEEP);
/* Iterate my child dirs */
for (zap_cursor_init(zc, os, dsl_dir_phys(dd)->dd_child_dir_zapobj);
zap_cursor_retrieve(zc, za) == 0; zap_cursor_advance(zc)) {
dsl_dir_t *chld_dd;
uint64_t count;
VERIFY0(dsl_dir_hold_obj(dp, za->za_first_integer, NULL, FTAG,
&chld_dd));
/*
* Ignore hidden ($FREE, $MOS & $ORIGIN) objsets and
* temporary datasets.
*/
if (chld_dd->dd_myname[0] == '$' ||
chld_dd->dd_myname[0] == '%') {
dsl_dir_rele(chld_dd, FTAG);
continue;
}
my_fs_cnt++; /* count this child */
dsl_dir_init_fs_ss_count(chld_dd, tx);
VERIFY0(zap_lookup(os, chld_dd->dd_object,
DD_FIELD_FILESYSTEM_COUNT, sizeof (count), 1, &count));
my_fs_cnt += count;
VERIFY0(zap_lookup(os, chld_dd->dd_object,
DD_FIELD_SNAPSHOT_COUNT, sizeof (count), 1, &count));
my_ss_cnt += count;
dsl_dir_rele(chld_dd, FTAG);
}
zap_cursor_fini(zc);
/* Count my snapshots (we counted children's snapshots above) */
VERIFY0(dsl_dataset_hold_obj(dd->dd_pool,
dsl_dir_phys(dd)->dd_head_dataset_obj, FTAG, &ds));
for (zap_cursor_init(zc, os, dsl_dataset_phys(ds)->ds_snapnames_zapobj);
zap_cursor_retrieve(zc, za) == 0;
zap_cursor_advance(zc)) {
/* Don't count temporary snapshots */
if (za->za_name[0] != '%')
my_ss_cnt++;
}
zap_cursor_fini(zc);
dsl_dataset_rele(ds, FTAG);
kmem_free(zc, sizeof (zap_cursor_t));
kmem_free(za, sizeof (zap_attribute_t));
/* we're in a sync task, update counts */
dmu_buf_will_dirty(dd->dd_dbuf, tx);
VERIFY0(zap_add(os, dd->dd_object, DD_FIELD_FILESYSTEM_COUNT,
sizeof (my_fs_cnt), 1, &my_fs_cnt, tx));
VERIFY0(zap_add(os, dd->dd_object, DD_FIELD_SNAPSHOT_COUNT,
sizeof (my_ss_cnt), 1, &my_ss_cnt, tx));
}
static int
dsl_dir_actv_fs_ss_limit_check(void *arg, dmu_tx_t *tx)
{
char *ddname = (char *)arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
dsl_dataset_t *ds;
dsl_dir_t *dd;
int error;
error = dsl_dataset_hold(dp, ddname, FTAG, &ds);
if (error != 0)
return (error);
if (!spa_feature_is_enabled(dp->dp_spa, SPA_FEATURE_FS_SS_LIMIT)) {
dsl_dataset_rele(ds, FTAG);
return (SET_ERROR(ENOTSUP));
}
dd = ds->ds_dir;
if (spa_feature_is_active(dp->dp_spa, SPA_FEATURE_FS_SS_LIMIT) &&
dsl_dir_is_zapified(dd) &&
zap_contains(dp->dp_meta_objset, dd->dd_object,
DD_FIELD_FILESYSTEM_COUNT) == 0) {
dsl_dataset_rele(ds, FTAG);
return (SET_ERROR(EALREADY));
}
dsl_dataset_rele(ds, FTAG);
return (0);
}
static void
dsl_dir_actv_fs_ss_limit_sync(void *arg, dmu_tx_t *tx)
{
char *ddname = (char *)arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
dsl_dataset_t *ds;
spa_t *spa;
VERIFY0(dsl_dataset_hold(dp, ddname, FTAG, &ds));
spa = dsl_dataset_get_spa(ds);
if (!spa_feature_is_active(spa, SPA_FEATURE_FS_SS_LIMIT)) {
/*
* Since the feature was not active and we're now setting a
* limit, increment the feature-active counter so that the
* feature becomes active for the first time.
*
* We are already in a sync task so we can update the MOS.
*/
spa_feature_incr(spa, SPA_FEATURE_FS_SS_LIMIT, tx);
}
/*
* Since we are now setting a non-UINT64_MAX limit on the filesystem,
* we need to ensure the counts are correct. Descend down the tree from
* this point and update all of the counts to be accurate.
*/
dsl_dir_init_fs_ss_count(ds->ds_dir, tx);
dsl_dataset_rele(ds, FTAG);
}
/*
* Make sure the feature is enabled and activate it if necessary.
* Since we're setting a limit, ensure the on-disk counts are valid.
* This is only called by the ioctl path when setting a limit value.
*
* We do not need to validate the new limit, since users who can change the
* limit are also allowed to exceed the limit.
*/
int
dsl_dir_activate_fs_ss_limit(const char *ddname)
{
int error;
error = dsl_sync_task(ddname, dsl_dir_actv_fs_ss_limit_check,
dsl_dir_actv_fs_ss_limit_sync, (void *)ddname, 0,
ZFS_SPACE_CHECK_RESERVED);
if (error == EALREADY)
error = 0;
return (error);
}
/*
* Used to determine if the filesystem_limit or snapshot_limit should be
* enforced. We allow the limit to be exceeded if the user has permission to
* write the property value. We pass in the creds that we got in the open
* context since we will always be the GZ root in syncing context. We also have
* to handle the case where we are allowed to change the limit on the current
* dataset, but there may be another limit in the tree above.
*
* We can never modify these two properties within a non-global zone. In
* addition, the other checks are modeled on zfs_secpolicy_write_perms. We
* can't use that function since we are already holding the dp_config_rwlock.
* In addition, we already have the dd and dealing with snapshots is simplified
* in this code.
*/
typedef enum {
ENFORCE_ALWAYS,
ENFORCE_NEVER,
ENFORCE_ABOVE
} enforce_res_t;
static enforce_res_t
dsl_enforce_ds_ss_limits(dsl_dir_t *dd, zfs_prop_t prop, cred_t *cr)
{
enforce_res_t enforce = ENFORCE_ALWAYS;
uint64_t obj;
dsl_dataset_t *ds;
uint64_t zoned;
ASSERT(prop == ZFS_PROP_FILESYSTEM_LIMIT ||
prop == ZFS_PROP_SNAPSHOT_LIMIT);
#ifdef _KERNEL
#ifdef __FreeBSD__
if (jailed(cr))
#else
if (crgetzoneid(cr) != GLOBAL_ZONEID)
#endif
return (ENFORCE_ALWAYS);
if (secpolicy_zfs(cr) == 0)
return (ENFORCE_NEVER);
#endif
if ((obj = dsl_dir_phys(dd)->dd_head_dataset_obj) == 0)
return (ENFORCE_ALWAYS);
ASSERT(dsl_pool_config_held(dd->dd_pool));
if (dsl_dataset_hold_obj(dd->dd_pool, obj, FTAG, &ds) != 0)
return (ENFORCE_ALWAYS);
if (dsl_prop_get_ds(ds, "zoned", 8, 1, &zoned, NULL) || zoned) {
/* Only root can access zoned fs's from the GZ */
enforce = ENFORCE_ALWAYS;
} else {
if (dsl_deleg_access_impl(ds, zfs_prop_to_name(prop), cr) == 0)
enforce = ENFORCE_ABOVE;
}
dsl_dataset_rele(ds, FTAG);
return (enforce);
}
static void
dsl_dir_update_last_remap_txg_sync(void *varg, dmu_tx_t *tx)
{
ddulrt_arg_t *arg = varg;
uint64_t last_remap_txg;
dsl_dir_t *dd = arg->ddulrta_dd;
objset_t *mos = dd->dd_pool->dp_meta_objset;
dsl_dir_zapify(dd, tx);
if (zap_lookup(mos, dd->dd_object, DD_FIELD_LAST_REMAP_TXG,
sizeof (last_remap_txg), 1, &last_remap_txg) != 0 ||
last_remap_txg < arg->ddlrta_txg) {
VERIFY0(zap_update(mos, dd->dd_object, DD_FIELD_LAST_REMAP_TXG,
sizeof (arg->ddlrta_txg), 1, &arg->ddlrta_txg, tx));
}
}
int
dsl_dir_update_last_remap_txg(dsl_dir_t *dd, uint64_t txg)
{
ddulrt_arg_t arg;
arg.ddulrta_dd = dd;
arg.ddlrta_txg = txg;
return (dsl_sync_task(spa_name(dd->dd_pool->dp_spa),
NULL, dsl_dir_update_last_remap_txg_sync, &arg,
1, ZFS_SPACE_CHECK_RESERVED));
}
/*
* Check if adding additional child filesystem(s) would exceed any filesystem
* limits or adding additional snapshot(s) would exceed any snapshot limits.
* The prop argument indicates which limit to check.
*
* Note that all filesystem limits up to the root (or the highest
* initialized) filesystem or the given ancestor must be satisfied.
*/
int
dsl_fs_ss_limit_check(dsl_dir_t *dd, uint64_t delta, zfs_prop_t prop,
dsl_dir_t *ancestor, cred_t *cr)
{
objset_t *os = dd->dd_pool->dp_meta_objset;
uint64_t limit, count;
char *count_prop;
enforce_res_t enforce;
int err = 0;
ASSERT(dsl_pool_config_held(dd->dd_pool));
ASSERT(prop == ZFS_PROP_FILESYSTEM_LIMIT ||
prop == ZFS_PROP_SNAPSHOT_LIMIT);
/*
* If we're allowed to change the limit, don't enforce the limit
* e.g. this can happen if a snapshot is taken by an administrative
* user in the global zone (i.e. a recursive snapshot by root).
* However, we must handle the case of delegated permissions where we
* are allowed to change the limit on the current dataset, but there
* is another limit in the tree above.
*/
enforce = dsl_enforce_ds_ss_limits(dd, prop, cr);
if (enforce == ENFORCE_NEVER)
return (0);
/*
* e.g. if renaming a dataset with no snapshots, count adjustment
* is 0.
*/
if (delta == 0)
return (0);
if (prop == ZFS_PROP_SNAPSHOT_LIMIT) {
/*
* We don't enforce the limit for temporary snapshots. This is
* indicated by a NULL cred_t argument.
*/
if (cr == NULL)
return (0);
count_prop = DD_FIELD_SNAPSHOT_COUNT;
} else {
count_prop = DD_FIELD_FILESYSTEM_COUNT;
}
/*
* If an ancestor has been provided, stop checking the limit once we
* hit that dir. We need this during rename so that we don't overcount
* the check once we recurse up to the common ancestor.
*/
if (ancestor == dd)
return (0);
/*
* If we hit an uninitialized node while recursing up the tree, we can
* stop since we know there is no limit here (or above). The counts are
* not valid on this node and we know we won't touch this node's counts.
*/
if (!dsl_dir_is_zapified(dd) || zap_lookup(os, dd->dd_object,
count_prop, sizeof (count), 1, &count) == ENOENT)
return (0);
err = dsl_prop_get_dd(dd, zfs_prop_to_name(prop), 8, 1, &limit, NULL,
B_FALSE);
if (err != 0)
return (err);
/* Is there a limit which we've hit? */
if (enforce == ENFORCE_ALWAYS && (count + delta) > limit)
return (SET_ERROR(EDQUOT));
if (dd->dd_parent != NULL)
err = dsl_fs_ss_limit_check(dd->dd_parent, delta, prop,
ancestor, cr);
return (err);
}
/*
* Adjust the filesystem or snapshot count for the specified dsl_dir_t and all
* parents. When a new filesystem/snapshot is created, increment the count on
* all parents, and when a filesystem/snapshot is destroyed, decrement the
* count.
*/
void
dsl_fs_ss_count_adjust(dsl_dir_t *dd, int64_t delta, const char *prop,
dmu_tx_t *tx)
{
int err;
objset_t *os = dd->dd_pool->dp_meta_objset;
uint64_t count;
ASSERT(dsl_pool_config_held(dd->dd_pool));
ASSERT(dmu_tx_is_syncing(tx));
ASSERT(strcmp(prop, DD_FIELD_FILESYSTEM_COUNT) == 0 ||
strcmp(prop, DD_FIELD_SNAPSHOT_COUNT) == 0);
/*
* When we receive an incremental stream into a filesystem that already
* exists, a temporary clone is created. We don't count this temporary
* clone, whose name begins with a '%'. We also ignore hidden ($FREE,
* $MOS & $ORIGIN) objsets.
*/
if ((dd->dd_myname[0] == '%' || dd->dd_myname[0] == '$') &&
strcmp(prop, DD_FIELD_FILESYSTEM_COUNT) == 0)
return;
/*
* e.g. if renaming a dataset with no snapshots, count adjustment is 0
*/
if (delta == 0)
return;
/*
* If we hit an uninitialized node while recursing up the tree, we can
* stop since we know the counts are not valid on this node and we
* know we shouldn't touch this node's counts. An uninitialized count
* on the node indicates that either the feature has not yet been
* activated or there are no limits on this part of the tree.
*/
if (!dsl_dir_is_zapified(dd) || (err = zap_lookup(os, dd->dd_object,
prop, sizeof (count), 1, &count)) == ENOENT)
return;
VERIFY0(err);
count += delta;
/* Use a signed verify to make sure we're not neg. */
VERIFY3S(count, >=, 0);
VERIFY0(zap_update(os, dd->dd_object, prop, sizeof (count), 1, &count,
tx));
/* Roll up this additional count into our ancestors */
if (dd->dd_parent != NULL)
dsl_fs_ss_count_adjust(dd->dd_parent, delta, prop, tx);
}
uint64_t
dsl_dir_create_sync(dsl_pool_t *dp, dsl_dir_t *pds, const char *name,
dmu_tx_t *tx)
{
objset_t *mos = dp->dp_meta_objset;
uint64_t ddobj;
dsl_dir_phys_t *ddphys;
dmu_buf_t *dbuf;
ddobj = dmu_object_alloc(mos, DMU_OT_DSL_DIR, 0,
DMU_OT_DSL_DIR, sizeof (dsl_dir_phys_t), tx);
if (pds) {
VERIFY0(zap_add(mos, dsl_dir_phys(pds)->dd_child_dir_zapobj,
name, sizeof (uint64_t), 1, &ddobj, tx));
} else {
/* it's the root dir */
VERIFY0(zap_add(mos, DMU_POOL_DIRECTORY_OBJECT,
DMU_POOL_ROOT_DATASET, sizeof (uint64_t), 1, &ddobj, tx));
}
VERIFY0(dmu_bonus_hold(mos, ddobj, FTAG, &dbuf));
dmu_buf_will_dirty(dbuf, tx);
ddphys = dbuf->db_data;
ddphys->dd_creation_time = gethrestime_sec();
if (pds) {
ddphys->dd_parent_obj = pds->dd_object;
/* update the filesystem counts */
dsl_fs_ss_count_adjust(pds, 1, DD_FIELD_FILESYSTEM_COUNT, tx);
}
ddphys->dd_props_zapobj = zap_create(mos,
DMU_OT_DSL_PROPS, DMU_OT_NONE, 0, tx);
ddphys->dd_child_dir_zapobj = zap_create(mos,
DMU_OT_DSL_DIR_CHILD_MAP, DMU_OT_NONE, 0, tx);
if (spa_version(dp->dp_spa) >= SPA_VERSION_USED_BREAKDOWN)
ddphys->dd_flags |= DD_FLAG_USED_BREAKDOWN;
dmu_buf_rele(dbuf, FTAG);
return (ddobj);
}
boolean_t
dsl_dir_is_clone(dsl_dir_t *dd)
{
return (dsl_dir_phys(dd)->dd_origin_obj &&
(dd->dd_pool->dp_origin_snap == NULL ||
dsl_dir_phys(dd)->dd_origin_obj !=
dd->dd_pool->dp_origin_snap->ds_object));
}
uint64_t
dsl_dir_get_used(dsl_dir_t *dd)
{
return (dsl_dir_phys(dd)->dd_used_bytes);
}
uint64_t
dsl_dir_get_compressed(dsl_dir_t *dd)
{
return (dsl_dir_phys(dd)->dd_compressed_bytes);
}
uint64_t
dsl_dir_get_quota(dsl_dir_t *dd)
{
return (dsl_dir_phys(dd)->dd_quota);
}
uint64_t
dsl_dir_get_reservation(dsl_dir_t *dd)
{
return (dsl_dir_phys(dd)->dd_reserved);
}
uint64_t
dsl_dir_get_compressratio(dsl_dir_t *dd)
{
/* a fixed point number, 100x the ratio */
return (dsl_dir_phys(dd)->dd_compressed_bytes == 0 ? 100 :
(dsl_dir_phys(dd)->dd_uncompressed_bytes * 100 /
dsl_dir_phys(dd)->dd_compressed_bytes));
}
uint64_t
dsl_dir_get_logicalused(dsl_dir_t *dd)
{
return (dsl_dir_phys(dd)->dd_uncompressed_bytes);
}
uint64_t
dsl_dir_get_usedsnap(dsl_dir_t *dd)
{
return (dsl_dir_phys(dd)->dd_used_breakdown[DD_USED_SNAP]);
}
uint64_t
dsl_dir_get_usedds(dsl_dir_t *dd)
{
return (dsl_dir_phys(dd)->dd_used_breakdown[DD_USED_HEAD]);
}
uint64_t
dsl_dir_get_usedrefreserv(dsl_dir_t *dd)
{
return (dsl_dir_phys(dd)->dd_used_breakdown[DD_USED_REFRSRV]);
}
uint64_t
dsl_dir_get_usedchild(dsl_dir_t *dd)
{
return (dsl_dir_phys(dd)->dd_used_breakdown[DD_USED_CHILD] +
dsl_dir_phys(dd)->dd_used_breakdown[DD_USED_CHILD_RSRV]);
}
void
dsl_dir_get_origin(dsl_dir_t *dd, char *buf)
{
dsl_dataset_t *ds;
VERIFY0(dsl_dataset_hold_obj(dd->dd_pool,
dsl_dir_phys(dd)->dd_origin_obj, FTAG, &ds));
dsl_dataset_name(ds, buf);
dsl_dataset_rele(ds, FTAG);
}
int
dsl_dir_get_filesystem_count(dsl_dir_t *dd, uint64_t *count)
{
if (dsl_dir_is_zapified(dd)) {
objset_t *os = dd->dd_pool->dp_meta_objset;
return (zap_lookup(os, dd->dd_object, DD_FIELD_FILESYSTEM_COUNT,
sizeof (*count), 1, count));
} else {
return (ENOENT);
}
}
int
dsl_dir_get_snapshot_count(dsl_dir_t *dd, uint64_t *count)
{
if (dsl_dir_is_zapified(dd)) {
objset_t *os = dd->dd_pool->dp_meta_objset;
return (zap_lookup(os, dd->dd_object, DD_FIELD_SNAPSHOT_COUNT,
sizeof (*count), 1, count));
} else {
return (ENOENT);
}
}
int
dsl_dir_get_remaptxg(dsl_dir_t *dd, uint64_t *count)
{
if (dsl_dir_is_zapified(dd)) {
objset_t *os = dd->dd_pool->dp_meta_objset;
return (zap_lookup(os, dd->dd_object, DD_FIELD_LAST_REMAP_TXG,
sizeof (*count), 1, count));
} else {
return (ENOENT);
}
}
void
dsl_dir_stats(dsl_dir_t *dd, nvlist_t *nv)
{
mutex_enter(&dd->dd_lock);
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_QUOTA,
dsl_dir_get_quota(dd));
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_RESERVATION,
dsl_dir_get_reservation(dd));
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_LOGICALUSED,
dsl_dir_get_logicalused(dd));
if (dsl_dir_phys(dd)->dd_flags & DD_FLAG_USED_BREAKDOWN) {
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_USEDSNAP,
dsl_dir_get_usedsnap(dd));
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_USEDDS,
dsl_dir_get_usedds(dd));
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_USEDREFRESERV,
dsl_dir_get_usedrefreserv(dd));
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_USEDCHILD,
dsl_dir_get_usedchild(dd));
}
mutex_exit(&dd->dd_lock);
uint64_t count;
if (dsl_dir_get_filesystem_count(dd, &count) == 0) {
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_FILESYSTEM_COUNT,
count);
}
if (dsl_dir_get_snapshot_count(dd, &count) == 0) {
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_SNAPSHOT_COUNT,
count);
}
if (dsl_dir_get_remaptxg(dd, &count) == 0) {
dsl_prop_nvlist_add_uint64(nv, ZFS_PROP_REMAPTXG,
count);
}
if (dsl_dir_is_clone(dd)) {
char buf[ZFS_MAX_DATASET_NAME_LEN];
dsl_dir_get_origin(dd, buf);
dsl_prop_nvlist_add_string(nv, ZFS_PROP_ORIGIN, buf);
}
}
void
dsl_dir_dirty(dsl_dir_t *dd, dmu_tx_t *tx)
{
dsl_pool_t *dp = dd->dd_pool;
ASSERT(dsl_dir_phys(dd));
if (txg_list_add(&dp->dp_dirty_dirs, dd, tx->tx_txg)) {
/* up the hold count until we can be written out */
dmu_buf_add_ref(dd->dd_dbuf, dd);
}
}
static int64_t
parent_delta(dsl_dir_t *dd, uint64_t used, int64_t delta)
{
uint64_t old_accounted = MAX(used, dsl_dir_phys(dd)->dd_reserved);
uint64_t new_accounted =
MAX(used + delta, dsl_dir_phys(dd)->dd_reserved);
return (new_accounted - old_accounted);
}
void
dsl_dir_sync(dsl_dir_t *dd, dmu_tx_t *tx)
{
ASSERT(dmu_tx_is_syncing(tx));
mutex_enter(&dd->dd_lock);
ASSERT0(dd->dd_tempreserved[tx->tx_txg&TXG_MASK]);
dprintf_dd(dd, "txg=%llu towrite=%lluK\n", tx->tx_txg,
dd->dd_space_towrite[tx->tx_txg&TXG_MASK] / 1024);
dd->dd_space_towrite[tx->tx_txg&TXG_MASK] = 0;
mutex_exit(&dd->dd_lock);
/* release the hold from dsl_dir_dirty */
dmu_buf_rele(dd->dd_dbuf, dd);
}
static uint64_t
dsl_dir_space_towrite(dsl_dir_t *dd)
{
uint64_t space = 0;
ASSERT(MUTEX_HELD(&dd->dd_lock));
for (int i = 0; i < TXG_SIZE; i++) {
space += dd->dd_space_towrite[i & TXG_MASK];
ASSERT3U(dd->dd_space_towrite[i & TXG_MASK], >=, 0);
}
return (space);
}
/*
* How much space would dd have available if ancestor had delta applied
* to it? If ondiskonly is set, we're only interested in what's
* on-disk, not estimated pending changes.
*/
uint64_t
dsl_dir_space_available(dsl_dir_t *dd,
dsl_dir_t *ancestor, int64_t delta, int ondiskonly)
{
uint64_t parentspace, myspace, quota, used;
/*
* If there are no restrictions otherwise, assume we have
* unlimited space available.
*/
quota = UINT64_MAX;
parentspace = UINT64_MAX;
if (dd->dd_parent != NULL) {
parentspace = dsl_dir_space_available(dd->dd_parent,
ancestor, delta, ondiskonly);
}
mutex_enter(&dd->dd_lock);
if (dsl_dir_phys(dd)->dd_quota != 0)
quota = dsl_dir_phys(dd)->dd_quota;
used = dsl_dir_phys(dd)->dd_used_bytes;
if (!ondiskonly)
used += dsl_dir_space_towrite(dd);
if (dd->dd_parent == NULL) {
uint64_t poolsize = dsl_pool_adjustedsize(dd->dd_pool,
ZFS_SPACE_CHECK_NORMAL);
quota = MIN(quota, poolsize);
}
if (dsl_dir_phys(dd)->dd_reserved > used && parentspace != UINT64_MAX) {
/*
* We have some space reserved, in addition to what our
* parent gave us.
*/
parentspace += dsl_dir_phys(dd)->dd_reserved - used;
}
if (dd == ancestor) {
ASSERT(delta <= 0);
ASSERT(used >= -delta);
used += delta;
if (parentspace != UINT64_MAX)
parentspace -= delta;
}
if (used > quota) {
/* over quota */
myspace = 0;
} else {
/*
* the lesser of the space provided by our parent and
* the space left in our quota
*/
myspace = MIN(parentspace, quota - used);
}
mutex_exit(&dd->dd_lock);
return (myspace);
}
struct tempreserve {
list_node_t tr_node;
dsl_dir_t *tr_ds;
uint64_t tr_size;
};
static int
dsl_dir_tempreserve_impl(dsl_dir_t *dd, uint64_t asize, boolean_t netfree,
boolean_t ignorequota, list_t *tr_list,
dmu_tx_t *tx, boolean_t first)
{
uint64_t txg = tx->tx_txg;
uint64_t quota;
struct tempreserve *tr;
int retval = EDQUOT;
uint64_t ref_rsrv = 0;
ASSERT3U(txg, !=, 0);
ASSERT3S(asize, >, 0);
mutex_enter(&dd->dd_lock);
/*
* Check against the dsl_dir's quota. We don't add in the delta
* when checking for over-quota because they get one free hit.
*/
uint64_t est_inflight = dsl_dir_space_towrite(dd);
for (int i = 0; i < TXG_SIZE; i++)
est_inflight += dd->dd_tempreserved[i];
uint64_t used_on_disk = dsl_dir_phys(dd)->dd_used_bytes;
/*
* On the first iteration, fetch the dataset's used-on-disk and
* refreservation values. Also, if checkrefquota is set, test if
* allocating this space would exceed the dataset's refquota.
*/
if (first && tx->tx_objset) {
int error;
dsl_dataset_t *ds = tx->tx_objset->os_dsl_dataset;
error = dsl_dataset_check_quota(ds, !netfree,
asize, est_inflight, &used_on_disk, &ref_rsrv);
if (error != 0) {
mutex_exit(&dd->dd_lock);
return (error);
}
}
/*
* If this transaction will result in a net free of space,
* we want to let it through.
*/
if (ignorequota || netfree || dsl_dir_phys(dd)->dd_quota == 0)
quota = UINT64_MAX;
else
quota = dsl_dir_phys(dd)->dd_quota;
/*
* Adjust the quota against the actual pool size at the root
* minus any outstanding deferred frees.
* To ensure that it's possible to remove files from a full
* pool without inducing transient overcommits, we throttle
* netfree transactions against a quota that is slightly larger,
* but still within the pool's allocation slop. In cases where
* we're very close to full, this will allow a steady trickle of
* removes to get through.
*/
uint64_t deferred = 0;
if (dd->dd_parent == NULL) {
uint64_t avail = dsl_pool_unreserved_space(dd->dd_pool,
(netfree) ?
ZFS_SPACE_CHECK_RESERVED : ZFS_SPACE_CHECK_NORMAL);
if (avail < quota) {
quota = avail;
retval = ENOSPC;
}
}
/*
* If they are requesting more space, and our current estimate
* is over quota, they get to try again unless the actual
* on-disk is over quota and there are no pending changes (which
* may free up space for us).
*/
if (used_on_disk + est_inflight >= quota) {
if (est_inflight > 0 || used_on_disk < quota ||
(retval == ENOSPC && used_on_disk < quota + deferred))
retval = ERESTART;
dprintf_dd(dd, "failing: used=%lluK inflight = %lluK "
"quota=%lluK tr=%lluK err=%d\n",
used_on_disk>>10, est_inflight>>10,
quota>>10, asize>>10, retval);
mutex_exit(&dd->dd_lock);
return (SET_ERROR(retval));
}
/* We need to up our estimated delta before dropping dd_lock */
dd->dd_tempreserved[txg & TXG_MASK] += asize;
uint64_t parent_rsrv = parent_delta(dd, used_on_disk + est_inflight,
asize - ref_rsrv);
mutex_exit(&dd->dd_lock);
tr = kmem_zalloc(sizeof (struct tempreserve), KM_SLEEP);
tr->tr_ds = dd;
tr->tr_size = asize;
list_insert_tail(tr_list, tr);
/* see if it's OK with our parent */
if (dd->dd_parent != NULL && parent_rsrv != 0) {
boolean_t ismos = (dsl_dir_phys(dd)->dd_head_dataset_obj == 0);
return (dsl_dir_tempreserve_impl(dd->dd_parent,
parent_rsrv, netfree, ismos, tr_list, tx, B_FALSE));
} else {
return (0);
}
}
/*
* Reserve space in this dsl_dir, to be used in this tx's txg.
* After the space has been dirtied (and dsl_dir_willuse_space()
* has been called), the reservation should be canceled, using
* dsl_dir_tempreserve_clear().
*/
int
dsl_dir_tempreserve_space(dsl_dir_t *dd, uint64_t lsize, uint64_t asize,
boolean_t netfree, void **tr_cookiep, dmu_tx_t *tx)
{
int err;
list_t *tr_list;
if (asize == 0) {
*tr_cookiep = NULL;
return (0);
}
tr_list = kmem_alloc(sizeof (list_t), KM_SLEEP);
list_create(tr_list, sizeof (struct tempreserve),
offsetof(struct tempreserve, tr_node));
ASSERT3S(asize, >, 0);
err = arc_tempreserve_space(dd->dd_pool->dp_spa, lsize, tx->tx_txg);
if (err == 0) {
struct tempreserve *tr;
tr = kmem_zalloc(sizeof (struct tempreserve), KM_SLEEP);
tr->tr_size = lsize;
list_insert_tail(tr_list, tr);
} else {
if (err == EAGAIN) {
/*
* If arc_memory_throttle() detected that pageout
* is running and we are low on memory, we delay new
* non-pageout transactions to give pageout an
* advantage.
*
* It is unfortunate to be delaying while the caller's
* locks are held.
*/
txg_delay(dd->dd_pool, tx->tx_txg,
MSEC2NSEC(10), MSEC2NSEC(10));
err = SET_ERROR(ERESTART);
}
}
if (err == 0) {
err = dsl_dir_tempreserve_impl(dd, asize, netfree,
B_FALSE, tr_list, tx, B_TRUE);
}
if (err != 0)
dsl_dir_tempreserve_clear(tr_list, tx);
else
*tr_cookiep = tr_list;
return (err);
}
/*
* Clear a temporary reservation that we previously made with
* dsl_dir_tempreserve_space().
*/
void
dsl_dir_tempreserve_clear(void *tr_cookie, dmu_tx_t *tx)
{
int txgidx = tx->tx_txg & TXG_MASK;
list_t *tr_list = tr_cookie;
struct tempreserve *tr;
ASSERT3U(tx->tx_txg, !=, 0);
if (tr_cookie == NULL)
return;
while ((tr = list_head(tr_list)) != NULL) {
if (tr->tr_ds) {
mutex_enter(&tr->tr_ds->dd_lock);
ASSERT3U(tr->tr_ds->dd_tempreserved[txgidx], >=,
tr->tr_size);
tr->tr_ds->dd_tempreserved[txgidx] -= tr->tr_size;
mutex_exit(&tr->tr_ds->dd_lock);
} else {
arc_tempreserve_clear(tr->tr_size);
}
list_remove(tr_list, tr);
kmem_free(tr, sizeof (struct tempreserve));
}
kmem_free(tr_list, sizeof (list_t));
}
/*
* This should be called from open context when we think we're going to write
* or free space, for example when dirtying data. Be conservative; it's okay
* to write less space or free more, but we don't want to write more or free
* less than the amount specified.
*/
void
dsl_dir_willuse_space(dsl_dir_t *dd, int64_t space, dmu_tx_t *tx)
{
int64_t parent_space;
uint64_t est_used;
mutex_enter(&dd->dd_lock);
if (space > 0)
dd->dd_space_towrite[tx->tx_txg & TXG_MASK] += space;
est_used = dsl_dir_space_towrite(dd) + dsl_dir_phys(dd)->dd_used_bytes;
parent_space = parent_delta(dd, est_used, space);
mutex_exit(&dd->dd_lock);
/* Make sure that we clean up dd_space_to* */
dsl_dir_dirty(dd, tx);
/* XXX this is potentially expensive and unnecessary... */
if (parent_space && dd->dd_parent)
dsl_dir_willuse_space(dd->dd_parent, parent_space, tx);
}
/* call from syncing context when we actually write/free space for this dd */
void
dsl_dir_diduse_space(dsl_dir_t *dd, dd_used_t type,
int64_t used, int64_t compressed, int64_t uncompressed, dmu_tx_t *tx)
{
int64_t accounted_delta;
/*
* dsl_dataset_set_refreservation_sync_impl() calls this with
* dd_lock held, so that it can atomically update
* ds->ds_reserved and the dsl_dir accounting, so that
* dsl_dataset_check_quota() can see dataset and dir accounting
* consistently.
*/
boolean_t needlock = !MUTEX_HELD(&dd->dd_lock);
ASSERT(dmu_tx_is_syncing(tx));
ASSERT(type < DD_USED_NUM);
dmu_buf_will_dirty(dd->dd_dbuf, tx);
if (needlock)
mutex_enter(&dd->dd_lock);
accounted_delta =
parent_delta(dd, dsl_dir_phys(dd)->dd_used_bytes, used);
ASSERT(used >= 0 || dsl_dir_phys(dd)->dd_used_bytes >= -used);
ASSERT(compressed >= 0 ||
dsl_dir_phys(dd)->dd_compressed_bytes >= -compressed);
ASSERT(uncompressed >= 0 ||
dsl_dir_phys(dd)->dd_uncompressed_bytes >= -uncompressed);
dsl_dir_phys(dd)->dd_used_bytes += used;
dsl_dir_phys(dd)->dd_uncompressed_bytes += uncompressed;
dsl_dir_phys(dd)->dd_compressed_bytes += compressed;
if (dsl_dir_phys(dd)->dd_flags & DD_FLAG_USED_BREAKDOWN) {
ASSERT(used > 0 ||
dsl_dir_phys(dd)->dd_used_breakdown[type] >= -used);
dsl_dir_phys(dd)->dd_used_breakdown[type] += used;
#ifdef DEBUG
dd_used_t t;
uint64_t u = 0;
for (t = 0; t < DD_USED_NUM; t++)
u += dsl_dir_phys(dd)->dd_used_breakdown[t];
ASSERT3U(u, ==, dsl_dir_phys(dd)->dd_used_bytes);
#endif
}
if (needlock)
mutex_exit(&dd->dd_lock);
if (dd->dd_parent != NULL) {
dsl_dir_diduse_space(dd->dd_parent, DD_USED_CHILD,
accounted_delta, compressed, uncompressed, tx);
dsl_dir_transfer_space(dd->dd_parent,
used - accounted_delta,
DD_USED_CHILD_RSRV, DD_USED_CHILD, NULL);
}
}
void
dsl_dir_transfer_space(dsl_dir_t *dd, int64_t delta,
dd_used_t oldtype, dd_used_t newtype, dmu_tx_t *tx)
{
ASSERT(tx == NULL || dmu_tx_is_syncing(tx));
ASSERT(oldtype < DD_USED_NUM);
ASSERT(newtype < DD_USED_NUM);
if (delta == 0 ||
!(dsl_dir_phys(dd)->dd_flags & DD_FLAG_USED_BREAKDOWN))
return;
if (tx != NULL)
dmu_buf_will_dirty(dd->dd_dbuf, tx);
mutex_enter(&dd->dd_lock);
ASSERT(delta > 0 ?
dsl_dir_phys(dd)->dd_used_breakdown[oldtype] >= delta :
dsl_dir_phys(dd)->dd_used_breakdown[newtype] >= -delta);
ASSERT(dsl_dir_phys(dd)->dd_used_bytes >= ABS(delta));
dsl_dir_phys(dd)->dd_used_breakdown[oldtype] -= delta;
dsl_dir_phys(dd)->dd_used_breakdown[newtype] += delta;
mutex_exit(&dd->dd_lock);
}
typedef struct dsl_dir_set_qr_arg {
const char *ddsqra_name;
zprop_source_t ddsqra_source;
uint64_t ddsqra_value;
} dsl_dir_set_qr_arg_t;
static int
dsl_dir_set_quota_check(void *arg, dmu_tx_t *tx)
{
dsl_dir_set_qr_arg_t *ddsqra = arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
dsl_dataset_t *ds;
int error;
uint64_t towrite, newval;
error = dsl_dataset_hold(dp, ddsqra->ddsqra_name, FTAG, &ds);
if (error != 0)
return (error);
error = dsl_prop_predict(ds->ds_dir, "quota",
ddsqra->ddsqra_source, ddsqra->ddsqra_value, &newval);
if (error != 0) {
dsl_dataset_rele(ds, FTAG);
return (error);
}
if (newval == 0) {
dsl_dataset_rele(ds, FTAG);
return (0);
}
mutex_enter(&ds->ds_dir->dd_lock);
/*
* If we are doing the preliminary check in open context, and
* there are pending changes, then don't fail it, since the
* pending changes could under-estimate the amount of space to be
* freed up.
*/
towrite = dsl_dir_space_towrite(ds->ds_dir);
if ((dmu_tx_is_syncing(tx) || towrite == 0) &&
(newval < dsl_dir_phys(ds->ds_dir)->dd_reserved ||
newval < dsl_dir_phys(ds->ds_dir)->dd_used_bytes + towrite)) {
error = SET_ERROR(ENOSPC);
}
mutex_exit(&ds->ds_dir->dd_lock);
dsl_dataset_rele(ds, FTAG);
return (error);
}
static void
dsl_dir_set_quota_sync(void *arg, dmu_tx_t *tx)
{
dsl_dir_set_qr_arg_t *ddsqra = arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
dsl_dataset_t *ds;
uint64_t newval;
VERIFY0(dsl_dataset_hold(dp, ddsqra->ddsqra_name, FTAG, &ds));
if (spa_version(dp->dp_spa) >= SPA_VERSION_RECVD_PROPS) {
dsl_prop_set_sync_impl(ds, zfs_prop_to_name(ZFS_PROP_QUOTA),
ddsqra->ddsqra_source, sizeof (ddsqra->ddsqra_value), 1,
&ddsqra->ddsqra_value, tx);
VERIFY0(dsl_prop_get_int_ds(ds,
zfs_prop_to_name(ZFS_PROP_QUOTA), &newval));
} else {
newval = ddsqra->ddsqra_value;
spa_history_log_internal_ds(ds, "set", tx, "%s=%lld",
zfs_prop_to_name(ZFS_PROP_QUOTA), (longlong_t)newval);
}
dmu_buf_will_dirty(ds->ds_dir->dd_dbuf, tx);
mutex_enter(&ds->ds_dir->dd_lock);
dsl_dir_phys(ds->ds_dir)->dd_quota = newval;
mutex_exit(&ds->ds_dir->dd_lock);
dsl_dataset_rele(ds, FTAG);
}
int
dsl_dir_set_quota(const char *ddname, zprop_source_t source, uint64_t quota)
{
dsl_dir_set_qr_arg_t ddsqra;
ddsqra.ddsqra_name = ddname;
ddsqra.ddsqra_source = source;
ddsqra.ddsqra_value = quota;
return (dsl_sync_task(ddname, dsl_dir_set_quota_check,
dsl_dir_set_quota_sync, &ddsqra, 0,
ZFS_SPACE_CHECK_EXTRA_RESERVED));
}
int
dsl_dir_set_reservation_check(void *arg, dmu_tx_t *tx)
{
dsl_dir_set_qr_arg_t *ddsqra = arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
dsl_dataset_t *ds;
dsl_dir_t *dd;
uint64_t newval, used, avail;
int error;
error = dsl_dataset_hold(dp, ddsqra->ddsqra_name, FTAG, &ds);
if (error != 0)
return (error);
dd = ds->ds_dir;
/*
* If we are doing the preliminary check in open context, the
* space estimates may be inaccurate.
*/
if (!dmu_tx_is_syncing(tx)) {
dsl_dataset_rele(ds, FTAG);
return (0);
}
error = dsl_prop_predict(ds->ds_dir,
zfs_prop_to_name(ZFS_PROP_RESERVATION),
ddsqra->ddsqra_source, ddsqra->ddsqra_value, &newval);
if (error != 0) {
dsl_dataset_rele(ds, FTAG);
return (error);
}
mutex_enter(&dd->dd_lock);
used = dsl_dir_phys(dd)->dd_used_bytes;
mutex_exit(&dd->dd_lock);
if (dd->dd_parent) {
avail = dsl_dir_space_available(dd->dd_parent,
NULL, 0, FALSE);
} else {
avail = dsl_pool_adjustedsize(dd->dd_pool,
ZFS_SPACE_CHECK_NORMAL) - used;
}
if (MAX(used, newval) > MAX(used, dsl_dir_phys(dd)->dd_reserved)) {
uint64_t delta = MAX(used, newval) -
MAX(used, dsl_dir_phys(dd)->dd_reserved);
if (delta > avail ||
(dsl_dir_phys(dd)->dd_quota > 0 &&
newval > dsl_dir_phys(dd)->dd_quota))
error = SET_ERROR(ENOSPC);
}
dsl_dataset_rele(ds, FTAG);
return (error);
}
void
dsl_dir_set_reservation_sync_impl(dsl_dir_t *dd, uint64_t value, dmu_tx_t *tx)
{
uint64_t used;
int64_t delta;
dmu_buf_will_dirty(dd->dd_dbuf, tx);
mutex_enter(&dd->dd_lock);
used = dsl_dir_phys(dd)->dd_used_bytes;
delta = MAX(used, value) - MAX(used, dsl_dir_phys(dd)->dd_reserved);
dsl_dir_phys(dd)->dd_reserved = value;
if (dd->dd_parent != NULL) {
/* Roll up this additional usage into our ancestors */
dsl_dir_diduse_space(dd->dd_parent, DD_USED_CHILD_RSRV,
delta, 0, 0, tx);
}
mutex_exit(&dd->dd_lock);
}
static void
dsl_dir_set_reservation_sync(void *arg, dmu_tx_t *tx)
{
dsl_dir_set_qr_arg_t *ddsqra = arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
dsl_dataset_t *ds;
uint64_t newval;
VERIFY0(dsl_dataset_hold(dp, ddsqra->ddsqra_name, FTAG, &ds));
if (spa_version(dp->dp_spa) >= SPA_VERSION_RECVD_PROPS) {
dsl_prop_set_sync_impl(ds,
zfs_prop_to_name(ZFS_PROP_RESERVATION),
ddsqra->ddsqra_source, sizeof (ddsqra->ddsqra_value), 1,
&ddsqra->ddsqra_value, tx);
VERIFY0(dsl_prop_get_int_ds(ds,
zfs_prop_to_name(ZFS_PROP_RESERVATION), &newval));
} else {
newval = ddsqra->ddsqra_value;
spa_history_log_internal_ds(ds, "set", tx, "%s=%lld",
zfs_prop_to_name(ZFS_PROP_RESERVATION),
(longlong_t)newval);
}
dsl_dir_set_reservation_sync_impl(ds->ds_dir, newval, tx);
dsl_dataset_rele(ds, FTAG);
}
int
dsl_dir_set_reservation(const char *ddname, zprop_source_t source,
uint64_t reservation)
{
dsl_dir_set_qr_arg_t ddsqra;
ddsqra.ddsqra_name = ddname;
ddsqra.ddsqra_source = source;
ddsqra.ddsqra_value = reservation;
return (dsl_sync_task(ddname, dsl_dir_set_reservation_check,
dsl_dir_set_reservation_sync, &ddsqra, 0,
ZFS_SPACE_CHECK_EXTRA_RESERVED));
}
static dsl_dir_t *
closest_common_ancestor(dsl_dir_t *ds1, dsl_dir_t *ds2)
{
for (; ds1; ds1 = ds1->dd_parent) {
dsl_dir_t *dd;
for (dd = ds2; dd; dd = dd->dd_parent) {
if (ds1 == dd)
return (dd);
}
}
return (NULL);
}
/*
* If delta is applied to dd, how much of that delta would be applied to
* ancestor? Syncing context only.
*/
static int64_t
would_change(dsl_dir_t *dd, int64_t delta, dsl_dir_t *ancestor)
{
if (dd == ancestor)
return (delta);
mutex_enter(&dd->dd_lock);
delta = parent_delta(dd, dsl_dir_phys(dd)->dd_used_bytes, delta);
mutex_exit(&dd->dd_lock);
return (would_change(dd->dd_parent, delta, ancestor));
}
typedef struct dsl_dir_rename_arg {
const char *ddra_oldname;
const char *ddra_newname;
cred_t *ddra_cred;
} dsl_dir_rename_arg_t;
typedef struct dsl_valid_rename_arg {
int char_delta;
int nest_delta;
} dsl_valid_rename_arg_t;
/* ARGSUSED */
static int
dsl_valid_rename(dsl_pool_t *dp, dsl_dataset_t *ds, void *arg)
{
dsl_valid_rename_arg_t *dvra = arg;
char namebuf[ZFS_MAX_DATASET_NAME_LEN];
dsl_dataset_name(ds, namebuf);
ASSERT3U(strnlen(namebuf, ZFS_MAX_DATASET_NAME_LEN),
<, ZFS_MAX_DATASET_NAME_LEN);
int namelen = strlen(namebuf) + dvra->char_delta;
int depth = get_dataset_depth(namebuf) + dvra->nest_delta;
if (namelen >= ZFS_MAX_DATASET_NAME_LEN)
return (SET_ERROR(ENAMETOOLONG));
if (dvra->nest_delta > 0 && depth >= zfs_max_dataset_nesting)
return (SET_ERROR(ENAMETOOLONG));
return (0);
}
static int
dsl_dir_rename_check(void *arg, dmu_tx_t *tx)
{
dsl_dir_rename_arg_t *ddra = arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
dsl_dir_t *dd, *newparent;
dsl_valid_rename_arg_t dvra;
+ dsl_dataset_t *parentds;
+ objset_t *parentos;
const char *mynewname;
int error;
/* target dir should exist */
error = dsl_dir_hold(dp, ddra->ddra_oldname, FTAG, &dd, NULL);
if (error != 0)
return (error);
/* new parent should exist */
error = dsl_dir_hold(dp, ddra->ddra_newname, FTAG,
&newparent, &mynewname);
if (error != 0) {
dsl_dir_rele(dd, FTAG);
return (error);
}
/* can't rename to different pool */
if (dd->dd_pool != newparent->dd_pool) {
dsl_dir_rele(newparent, FTAG);
dsl_dir_rele(dd, FTAG);
return (SET_ERROR(EXDEV));
}
/* new name should not already exist */
if (mynewname == NULL) {
dsl_dir_rele(newparent, FTAG);
dsl_dir_rele(dd, FTAG);
return (SET_ERROR(EEXIST));
}
+
+ /* can't rename below anything but filesystems (eg. no ZVOLs) */
+ error = dsl_dataset_hold_obj(newparent->dd_pool,
+ dsl_dir_phys(newparent)->dd_head_dataset_obj, FTAG, &parentds);
+ if (error != 0) {
+ dsl_dir_rele(newparent, FTAG);
+ dsl_dir_rele(dd, FTAG);
+ return (error);
+ }
+ error = dmu_objset_from_ds(parentds, &parentos);
+ if (error != 0) {
+ dsl_dataset_rele(parentds, FTAG);
+ dsl_dir_rele(newparent, FTAG);
+ dsl_dir_rele(dd, FTAG);
+ return (error);
+ }
+ if (dmu_objset_type(parentos) != DMU_OST_ZFS) {
+ dsl_dataset_rele(parentds, FTAG);
+ dsl_dir_rele(newparent, FTAG);
+ dsl_dir_rele(dd, FTAG);
+ return (error);
+ }
+ dsl_dataset_rele(parentds, FTAG);
ASSERT3U(strnlen(ddra->ddra_newname, ZFS_MAX_DATASET_NAME_LEN),
<, ZFS_MAX_DATASET_NAME_LEN);
ASSERT3U(strnlen(ddra->ddra_oldname, ZFS_MAX_DATASET_NAME_LEN),
<, ZFS_MAX_DATASET_NAME_LEN);
dvra.char_delta = strlen(ddra->ddra_newname)
- strlen(ddra->ddra_oldname);
dvra.nest_delta = get_dataset_depth(ddra->ddra_newname)
- get_dataset_depth(ddra->ddra_oldname);
/* if the name length is growing, validate child name lengths */
if (dvra.char_delta > 0 || dvra.nest_delta > 0) {
error = dmu_objset_find_dp(dp, dd->dd_object, dsl_valid_rename,
&dvra, DS_FIND_CHILDREN | DS_FIND_SNAPSHOTS);
if (error != 0) {
dsl_dir_rele(newparent, FTAG);
dsl_dir_rele(dd, FTAG);
return (error);
}
}
if (dmu_tx_is_syncing(tx)) {
if (spa_feature_is_active(dp->dp_spa,
SPA_FEATURE_FS_SS_LIMIT)) {
/*
* Although this is the check function and we don't
* normally make on-disk changes in check functions,
* we need to do that here.
*
* Ensure this portion of the tree's counts have been
* initialized in case the new parent has limits set.
*/
dsl_dir_init_fs_ss_count(dd, tx);
}
}
if (newparent != dd->dd_parent) {
/* is there enough space? */
uint64_t myspace =
MAX(dsl_dir_phys(dd)->dd_used_bytes,
dsl_dir_phys(dd)->dd_reserved);
objset_t *os = dd->dd_pool->dp_meta_objset;
uint64_t fs_cnt = 0;
uint64_t ss_cnt = 0;
if (dsl_dir_is_zapified(dd)) {
int err;
err = zap_lookup(os, dd->dd_object,
DD_FIELD_FILESYSTEM_COUNT, sizeof (fs_cnt), 1,
&fs_cnt);
if (err != ENOENT && err != 0) {
dsl_dir_rele(newparent, FTAG);
dsl_dir_rele(dd, FTAG);
return (err);
}
/*
* have to add 1 for the filesystem itself that we're
* moving
*/
fs_cnt++;
err = zap_lookup(os, dd->dd_object,
DD_FIELD_SNAPSHOT_COUNT, sizeof (ss_cnt), 1,
&ss_cnt);
if (err != ENOENT && err != 0) {
dsl_dir_rele(newparent, FTAG);
dsl_dir_rele(dd, FTAG);
return (err);
}
}
/* no rename into our descendant */
if (closest_common_ancestor(dd, newparent) == dd) {
dsl_dir_rele(newparent, FTAG);
dsl_dir_rele(dd, FTAG);
return (SET_ERROR(EINVAL));
}
error = dsl_dir_transfer_possible(dd->dd_parent,
newparent, fs_cnt, ss_cnt, myspace, ddra->ddra_cred);
if (error != 0) {
dsl_dir_rele(newparent, FTAG);
dsl_dir_rele(dd, FTAG);
return (error);
}
}
dsl_dir_rele(newparent, FTAG);
dsl_dir_rele(dd, FTAG);
return (0);
}
static void
dsl_dir_rename_sync(void *arg, dmu_tx_t *tx)
{
dsl_dir_rename_arg_t *ddra = arg;
dsl_pool_t *dp = dmu_tx_pool(tx);
dsl_dir_t *dd, *newparent;
const char *mynewname;
int error;
objset_t *mos = dp->dp_meta_objset;
VERIFY0(dsl_dir_hold(dp, ddra->ddra_oldname, FTAG, &dd, NULL));
VERIFY0(dsl_dir_hold(dp, ddra->ddra_newname, FTAG, &newparent,
&mynewname));
/* Log this before we change the name. */
spa_history_log_internal_dd(dd, "rename", tx,
"-> %s", ddra->ddra_newname);
if (newparent != dd->dd_parent) {
objset_t *os = dd->dd_pool->dp_meta_objset;
uint64_t fs_cnt = 0;
uint64_t ss_cnt = 0;
/*
* We already made sure the dd counts were initialized in the
* check function.
*/
if (spa_feature_is_active(dp->dp_spa,
SPA_FEATURE_FS_SS_LIMIT)) {
VERIFY0(zap_lookup(os, dd->dd_object,
DD_FIELD_FILESYSTEM_COUNT, sizeof (fs_cnt), 1,
&fs_cnt));
/* add 1 for the filesystem itself that we're moving */
fs_cnt++;
VERIFY0(zap_lookup(os, dd->dd_object,
DD_FIELD_SNAPSHOT_COUNT, sizeof (ss_cnt), 1,
&ss_cnt));
}
dsl_fs_ss_count_adjust(dd->dd_parent, -fs_cnt,
DD_FIELD_FILESYSTEM_COUNT, tx);
dsl_fs_ss_count_adjust(newparent, fs_cnt,
DD_FIELD_FILESYSTEM_COUNT, tx);
dsl_fs_ss_count_adjust(dd->dd_parent, -ss_cnt,
DD_FIELD_SNAPSHOT_COUNT, tx);
dsl_fs_ss_count_adjust(newparent, ss_cnt,
DD_FIELD_SNAPSHOT_COUNT, tx);
dsl_dir_diduse_space(dd->dd_parent, DD_USED_CHILD,
-dsl_dir_phys(dd)->dd_used_bytes,
-dsl_dir_phys(dd)->dd_compressed_bytes,
-dsl_dir_phys(dd)->dd_uncompressed_bytes, tx);
dsl_dir_diduse_space(newparent, DD_USED_CHILD,
dsl_dir_phys(dd)->dd_used_bytes,
dsl_dir_phys(dd)->dd_compressed_bytes,
dsl_dir_phys(dd)->dd_uncompressed_bytes, tx);
if (dsl_dir_phys(dd)->dd_reserved >
dsl_dir_phys(dd)->dd_used_bytes) {
uint64_t unused_rsrv = dsl_dir_phys(dd)->dd_reserved -
dsl_dir_phys(dd)->dd_used_bytes;
dsl_dir_diduse_space(dd->dd_parent, DD_USED_CHILD_RSRV,
-unused_rsrv, 0, 0, tx);
dsl_dir_diduse_space(newparent, DD_USED_CHILD_RSRV,
unused_rsrv, 0, 0, tx);
}
}
dmu_buf_will_dirty(dd->dd_dbuf, tx);
/* remove from old parent zapobj */
error = zap_remove(mos,
dsl_dir_phys(dd->dd_parent)->dd_child_dir_zapobj,
dd->dd_myname, tx);
ASSERT0(error);
(void) strcpy(dd->dd_myname, mynewname);
dsl_dir_rele(dd->dd_parent, dd);
dsl_dir_phys(dd)->dd_parent_obj = newparent->dd_object;
VERIFY0(dsl_dir_hold_obj(dp,
newparent->dd_object, NULL, dd, &dd->dd_parent));
/* add to new parent zapobj */
VERIFY0(zap_add(mos, dsl_dir_phys(newparent)->dd_child_dir_zapobj,
dd->dd_myname, 8, 1, &dd->dd_object, tx));
#ifdef __FreeBSD__
#ifdef _KERNEL
zfsvfs_update_fromname(ddra->ddra_oldname, ddra->ddra_newname);
zvol_rename_minors(ddra->ddra_oldname, ddra->ddra_newname);
#endif
#endif
dsl_prop_notify_all(dd);
dsl_dir_rele(newparent, FTAG);
dsl_dir_rele(dd, FTAG);
}
int
dsl_dir_rename(const char *oldname, const char *newname)
{
dsl_dir_rename_arg_t ddra;
ddra.ddra_oldname = oldname;
ddra.ddra_newname = newname;
ddra.ddra_cred = CRED();
return (dsl_sync_task(oldname,
dsl_dir_rename_check, dsl_dir_rename_sync, &ddra,
3, ZFS_SPACE_CHECK_RESERVED));
}
int
dsl_dir_transfer_possible(dsl_dir_t *sdd, dsl_dir_t *tdd,
uint64_t fs_cnt, uint64_t ss_cnt, uint64_t space, cred_t *cr)
{
dsl_dir_t *ancestor;
int64_t adelta;
uint64_t avail;
int err;
ancestor = closest_common_ancestor(sdd, tdd);
adelta = would_change(sdd, -space, ancestor);
avail = dsl_dir_space_available(tdd, ancestor, adelta, FALSE);
if (avail < space)
return (SET_ERROR(ENOSPC));
err = dsl_fs_ss_limit_check(tdd, fs_cnt, ZFS_PROP_FILESYSTEM_LIMIT,
ancestor, cr);
if (err != 0)
return (err);
err = dsl_fs_ss_limit_check(tdd, ss_cnt, ZFS_PROP_SNAPSHOT_LIMIT,
ancestor, cr);
if (err != 0)
return (err);
return (0);
}
timestruc_t
dsl_dir_snap_cmtime(dsl_dir_t *dd)
{
timestruc_t t;
mutex_enter(&dd->dd_lock);
t = dd->dd_snap_cmtime;
mutex_exit(&dd->dd_lock);
return (t);
}
void
dsl_dir_snap_cmtime_update(dsl_dir_t *dd)
{
timestruc_t t;
gethrestime(&t);
mutex_enter(&dd->dd_lock);
dd->dd_snap_cmtime = t;
mutex_exit(&dd->dd_lock);
}
void
dsl_dir_zapify(dsl_dir_t *dd, dmu_tx_t *tx)
{
objset_t *mos = dd->dd_pool->dp_meta_objset;
dmu_object_zapify(mos, dd->dd_object, DMU_OT_DSL_DIR, tx);
}
boolean_t
dsl_dir_is_zapified(dsl_dir_t *dd)
{
dmu_object_info_t doi;
dmu_object_info_from_db(dd->dd_dbuf, &doi);
return (doi.doi_type == DMU_OTN_ZAP_METADATA);
}
Index: head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ioctl.c
===================================================================
--- head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ioctl.c (revision 359302)
+++ head/sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_ioctl.c (revision 359303)
@@ -1,7340 +1,7338 @@
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2011-2012 Pawel Jakub Dawidek. All rights reserved.
* Copyright 2013 Martin Matuska <mm@FreeBSD.org>. All rights reserved.
* Copyright 2014 Xin Li <delphij@FreeBSD.org>. All rights reserved.
* Copyright 2015, OmniTI Computer Consulting, Inc. All rights reserved.
* Copyright 2015 Nexenta Systems, Inc. All rights reserved.
* Copyright (c) 2014, 2016 Joyent, Inc. All rights reserved.
* Copyright (c) 2011, 2017 by Delphix. All rights reserved.
* Copyright (c) 2013 by Saso Kiselkov. All rights reserved.
* Copyright (c) 2013 Steven Hartland. All rights reserved.
* Copyright (c) 2014 Integros [integros.com]
* Copyright 2016 Toomas Soome <tsoome@me.com>
* Copyright 2017 RackTop Systems.
+ * Copyright (c) 2018, loli10K <ezomori.nozomu@gmail.com>. All rights reserved.
* Copyright (c) 2019 Datto Inc.
*/
/*
* ZFS ioctls.
*
* This file handles the ioctls to /dev/zfs, used for configuring ZFS storage
* pools and filesystems, e.g. with /sbin/zfs and /sbin/zpool.
*
* There are two ways that we handle ioctls: the legacy way where almost
* all of the logic is in the ioctl callback, and the new way where most
* of the marshalling is handled in the common entry point, zfsdev_ioctl().
*
* Non-legacy ioctls should be registered by calling
* zfs_ioctl_register() from zfs_ioctl_init(). The ioctl is invoked
* from userland by lzc_ioctl().
*
* The registration arguments are as follows:
*
* const char *name
* The name of the ioctl. This is used for history logging. If the
* ioctl returns successfully (the callback returns 0), and allow_log
* is true, then a history log entry will be recorded with the input &
* output nvlists. The log entry can be printed with "zpool history -i".
*
* zfs_ioc_t ioc
* The ioctl request number, which userland will pass to ioctl(2).
* The ioctl numbers can change from release to release, because
* the caller (libzfs) must be matched to the kernel.
*
* zfs_secpolicy_func_t *secpolicy
* This function will be called before the zfs_ioc_func_t, to
* determine if this operation is permitted. It should return EPERM
* on failure, and 0 on success. Checks include determining if the
* dataset is visible in this zone, and if the user has either all
* zfs privileges in the zone (SYS_MOUNT), or has been granted permission
* to do this operation on this dataset with "zfs allow".
*
* zfs_ioc_namecheck_t namecheck
* This specifies what to expect in the zfs_cmd_t:zc_name -- a pool
* name, a dataset name, or nothing. If the name is not well-formed,
* the ioctl will fail and the callback will not be called.
* Therefore, the callback can assume that the name is well-formed
* (e.g. is null-terminated, doesn't have more than one '@' character,
* doesn't have invalid characters).
*
* zfs_ioc_poolcheck_t pool_check
* This specifies requirements on the pool state. If the pool does
* not meet them (is suspended or is readonly), the ioctl will fail
* and the callback will not be called. If any checks are specified
* (i.e. it is not POOL_CHECK_NONE), namecheck must not be NO_NAME.
* Multiple checks can be or-ed together (e.g. POOL_CHECK_SUSPENDED |
* POOL_CHECK_READONLY).
*
* boolean_t smush_outnvlist
* If smush_outnvlist is true, then the output is presumed to be a
* list of errors, and it will be "smushed" down to fit into the
* caller's buffer, by removing some entries and replacing them with a
* single "N_MORE_ERRORS" entry indicating how many were removed. See
* nvlist_smush() for details. If smush_outnvlist is false, and the
* outnvlist does not fit into the userland-provided buffer, then the
* ioctl will fail with ENOMEM.
*
* zfs_ioc_func_t *func
* The callback function that will perform the operation.
*
* The callback should return 0 on success, or an error number on
* failure. If the function fails, the userland ioctl will return -1,
* and errno will be set to the callback's return value. The callback
* will be called with the following arguments:
*
* const char *name
* The name of the pool or dataset to operate on, from
* zfs_cmd_t:zc_name. The 'namecheck' argument specifies the
* expected type (pool, dataset, or none).
*
* nvlist_t *innvl
* The input nvlist, deserialized from zfs_cmd_t:zc_nvlist_src. Or
* NULL if no input nvlist was provided. Changes to this nvlist are
* ignored. If the input nvlist could not be deserialized, the
* ioctl will fail and the callback will not be called.
*
* nvlist_t *outnvl
* The output nvlist, initially empty. The callback can fill it in,
* and it will be returned to userland by serializing it into
* zfs_cmd_t:zc_nvlist_dst. If it is non-empty, and serialization
* fails (e.g. because the caller didn't supply a large enough
* buffer), then the overall ioctl will fail. See the
* 'smush_nvlist' argument above for additional behaviors.
*
* There are two typical uses of the output nvlist:
* - To return state, e.g. property values. In this case,
* smush_outnvlist should be false. If the buffer was not large
* enough, the caller will reallocate a larger buffer and try
* the ioctl again.
*
* - To return multiple errors from an ioctl which makes on-disk
* changes. In this case, smush_outnvlist should be true.
* Ioctls which make on-disk modifications should generally not
* use the outnvl if they succeed, because the caller can not
* distinguish between the operation failing, and
* deserialization failing.
*/
#ifdef __FreeBSD__
#include "opt_kstack_pages.h"
#endif
#include <sys/types.h>
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/conf.h>
#include <sys/kernel.h>
#include <sys/lock.h>
#include <sys/malloc.h>
#include <sys/mutex.h>
#include <sys/proc.h>
#include <sys/errno.h>
#include <sys/uio.h>
#include <sys/buf.h>
#include <sys/file.h>
#include <sys/kmem.h>
#include <sys/conf.h>
#include <sys/cmn_err.h>
#include <sys/stat.h>
#include <sys/zfs_ioctl.h>
#include <sys/zfs_vfsops.h>
#include <sys/zfs_znode.h>
#include <sys/zap.h>
#include <sys/spa.h>
#include <sys/spa_impl.h>
#include <sys/vdev.h>
#include <sys/dmu.h>
#include <sys/dsl_dir.h>
#include <sys/dsl_dataset.h>
#include <sys/dsl_prop.h>
#include <sys/dsl_deleg.h>
#include <sys/dmu_objset.h>
#include <sys/dmu_impl.h>
#include <sys/dmu_tx.h>
#include <sys/sunddi.h>
#include <sys/policy.h>
#include <sys/zone.h>
#include <sys/nvpair.h>
#include <sys/mount.h>
#include <sys/taskqueue.h>
#include <sys/sdt.h>
#include <sys/varargs.h>
#include <sys/fs/zfs.h>
#include <sys/zfs_ctldir.h>
#include <sys/zfs_dir.h>
#include <sys/zfs_onexit.h>
#include <sys/zvol.h>
#include <sys/dsl_scan.h>
#include <sys/dmu_objset.h>
#include <sys/dmu_send.h>
#include <sys/dsl_destroy.h>
#include <sys/dsl_bookmark.h>
#include <sys/dsl_userhold.h>
#include <sys/zfeature.h>
#include <sys/zcp.h>
#include <sys/zio_checksum.h>
#include <sys/vdev_removal.h>
#include <sys/vdev_impl.h>
#include <sys/vdev_initialize.h>
#include "zfs_namecheck.h"
#include "zfs_prop.h"
#include "zfs_deleg.h"
#include "zfs_comutil.h"
#include "zfs_ioctl_compat.h"
#include "lua.h"
#include "lauxlib.h"
static struct cdev *zfsdev;
extern void zfs_init(void);
extern void zfs_fini(void);
uint_t zfs_fsyncer_key;
extern uint_t rrw_tsd_key;
static uint_t zfs_allow_log_key;
extern uint_t zfs_geom_probe_vdev_key;
typedef int zfs_ioc_legacy_func_t(zfs_cmd_t *);
typedef int zfs_ioc_func_t(const char *, nvlist_t *, nvlist_t *);
typedef int zfs_secpolicy_func_t(zfs_cmd_t *, nvlist_t *, cred_t *);
typedef enum {
NO_NAME,
POOL_NAME,
DATASET_NAME,
ENTITY_NAME
} zfs_ioc_namecheck_t;
typedef enum {
POOL_CHECK_NONE = 1 << 0,
POOL_CHECK_SUSPENDED = 1 << 1,
POOL_CHECK_READONLY = 1 << 2,
} zfs_ioc_poolcheck_t;
typedef struct zfs_ioc_vec {
zfs_ioc_legacy_func_t *zvec_legacy_func;
zfs_ioc_func_t *zvec_func;
zfs_secpolicy_func_t *zvec_secpolicy;
zfs_ioc_namecheck_t zvec_namecheck;
boolean_t zvec_allow_log;
zfs_ioc_poolcheck_t zvec_pool_check;
boolean_t zvec_smush_outnvlist;
const char *zvec_name;
} zfs_ioc_vec_t;
/* This array is indexed by zfs_userquota_prop_t */
static const char *userquota_perms[] = {
ZFS_DELEG_PERM_USERUSED,
ZFS_DELEG_PERM_USERQUOTA,
ZFS_DELEG_PERM_GROUPUSED,
ZFS_DELEG_PERM_GROUPQUOTA,
};
static int zfs_ioc_userspace_upgrade(zfs_cmd_t *zc);
static int zfs_check_settable(const char *name, nvpair_t *property,
cred_t *cr);
static int zfs_check_clearable(char *dataset, nvlist_t *props,
nvlist_t **errors);
static int zfs_fill_zplprops_root(uint64_t, nvlist_t *, nvlist_t *,
boolean_t *);
int zfs_set_prop_nvlist(const char *, zprop_source_t, nvlist_t *, nvlist_t *);
static int get_nvlist(uint64_t nvl, uint64_t size, int iflag, nvlist_t **nvp);
static void zfsdev_close(void *data);
static int zfs_prop_activate_feature(spa_t *spa, spa_feature_t feature);
/* _NOTE(PRINTFLIKE(4)) - this is printf-like, but lint is too whiney */
void
__dprintf(const char *file, const char *func, int line, const char *fmt, ...)
{
const char *newfile;
char buf[512];
va_list adx;
/*
* Get rid of annoying "../common/" prefix to filename.
*/
newfile = strrchr(file, '/');
if (newfile != NULL) {
newfile = newfile + 1; /* Get rid of leading / */
} else {
newfile = file;
}
va_start(adx, fmt);
(void) vsnprintf(buf, sizeof (buf), fmt, adx);
va_end(adx);
/*
* To get this data, use the zfs-dprintf probe as so:
* dtrace -q -n 'zfs-dprintf \
* /stringof(arg0) == "dbuf.c"/ \
* {printf("%s: %s", stringof(arg1), stringof(arg3))}'
* arg0 = file name
* arg1 = function name
* arg2 = line number
* arg3 = message
*/
DTRACE_PROBE4(zfs__dprintf,
char *, newfile, char *, func, int, line, char *, buf);
}
static void
history_str_free(char *buf)
{
kmem_free(buf, HIS_MAX_RECORD_LEN);
}
static char *
history_str_get(zfs_cmd_t *zc)
{
char *buf;
if (zc->zc_history == 0)
return (NULL);
buf = kmem_alloc(HIS_MAX_RECORD_LEN, KM_SLEEP);
if (copyinstr((void *)(uintptr_t)zc->zc_history,
buf, HIS_MAX_RECORD_LEN, NULL) != 0) {
history_str_free(buf);
return (NULL);
}
buf[HIS_MAX_RECORD_LEN -1] = '\0';
return (buf);
}
/*
* Check to see if the named dataset is currently defined as bootable
*/
static boolean_t
zfs_is_bootfs(const char *name)
{
objset_t *os;
if (dmu_objset_hold(name, FTAG, &os) == 0) {
boolean_t ret;
ret = (dmu_objset_id(os) == spa_bootfs(dmu_objset_spa(os)));
dmu_objset_rele(os, FTAG);
return (ret);
}
return (B_FALSE);
}
/*
* Return non-zero if the spa version is less than requested version.
*/
static int
zfs_earlier_version(const char *name, int version)
{
spa_t *spa;
if (spa_open(name, &spa, FTAG) == 0) {
if (spa_version(spa) < version) {
spa_close(spa, FTAG);
return (1);
}
spa_close(spa, FTAG);
}
return (0);
}
/*
* Return TRUE if the ZPL version is less than requested version.
*/
static boolean_t
zpl_earlier_version(const char *name, int version)
{
objset_t *os;
boolean_t rc = B_TRUE;
if (dmu_objset_hold(name, FTAG, &os) == 0) {
uint64_t zplversion;
if (dmu_objset_type(os) != DMU_OST_ZFS) {
dmu_objset_rele(os, FTAG);
return (B_TRUE);
}
/* XXX reading from non-owned objset */
if (zfs_get_zplprop(os, ZFS_PROP_VERSION, &zplversion) == 0)
rc = zplversion < version;
dmu_objset_rele(os, FTAG);
}
return (rc);
}
static void
zfs_log_history(zfs_cmd_t *zc)
{
spa_t *spa;
char *buf;
if ((buf = history_str_get(zc)) == NULL)
return;
if (spa_open(zc->zc_name, &spa, FTAG) == 0) {
if (spa_version(spa) >= SPA_VERSION_ZPOOL_HISTORY)
(void) spa_history_log(spa, buf);
spa_close(spa, FTAG);
}
history_str_free(buf);
}
/*
* Policy for top-level read operations (list pools). Requires no privileges,
* and can be used in the local zone, as there is no associated dataset.
*/
/* ARGSUSED */
static int
zfs_secpolicy_none(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
return (0);
}
/*
* Policy for dataset read operations (list children, get statistics). Requires
* no privileges, but must be visible in the local zone.
*/
/* ARGSUSED */
static int
zfs_secpolicy_read(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
if (INGLOBALZONE(curthread) ||
zone_dataset_visible(zc->zc_name, NULL))
return (0);
return (SET_ERROR(ENOENT));
}
static int
zfs_dozonecheck_impl(const char *dataset, uint64_t zoned, cred_t *cr)
{
int writable = 1;
/*
* The dataset must be visible by this zone -- check this first
* so they don't see EPERM on something they shouldn't know about.
*/
if (!INGLOBALZONE(curthread) &&
!zone_dataset_visible(dataset, &writable))
return (SET_ERROR(ENOENT));
if (INGLOBALZONE(curthread)) {
/*
* If the fs is zoned, only root can access it from the
* global zone.
*/
if (secpolicy_zfs(cr) && zoned)
return (SET_ERROR(EPERM));
} else {
/*
* If we are in a local zone, the 'zoned' property must be set.
*/
if (!zoned)
return (SET_ERROR(EPERM));
/* must be writable by this zone */
if (!writable)
return (SET_ERROR(EPERM));
}
return (0);
}
static int
zfs_dozonecheck(const char *dataset, cred_t *cr)
{
uint64_t zoned;
if (dsl_prop_get_integer(dataset, "jailed", &zoned, NULL))
return (SET_ERROR(ENOENT));
return (zfs_dozonecheck_impl(dataset, zoned, cr));
}
static int
zfs_dozonecheck_ds(const char *dataset, dsl_dataset_t *ds, cred_t *cr)
{
uint64_t zoned;
if (dsl_prop_get_int_ds(ds, "jailed", &zoned))
return (SET_ERROR(ENOENT));
return (zfs_dozonecheck_impl(dataset, zoned, cr));
}
static int
zfs_secpolicy_write_perms_ds(const char *name, dsl_dataset_t *ds,
const char *perm, cred_t *cr)
{
int error;
error = zfs_dozonecheck_ds(name, ds, cr);
if (error == 0) {
error = secpolicy_zfs(cr);
if (error != 0)
error = dsl_deleg_access_impl(ds, perm, cr);
}
return (error);
}
static int
zfs_secpolicy_write_perms(const char *name, const char *perm, cred_t *cr)
{
int error;
dsl_dataset_t *ds;
dsl_pool_t *dp;
/*
* First do a quick check for root in the global zone, which
* is allowed to do all write_perms. This ensures that zfs_ioc_*
* will get to handle nonexistent datasets.
*/
if (INGLOBALZONE(curthread) && secpolicy_zfs(cr) == 0)
return (0);
error = dsl_pool_hold(name, FTAG, &dp);
if (error != 0)
return (error);
error = dsl_dataset_hold(dp, name, FTAG, &ds);
if (error != 0) {
dsl_pool_rele(dp, FTAG);
return (error);
}
error = zfs_secpolicy_write_perms_ds(name, ds, perm, cr);
dsl_dataset_rele(ds, FTAG);
dsl_pool_rele(dp, FTAG);
return (error);
}
#ifdef SECLABEL
/*
* Policy for setting the security label property.
*
* Returns 0 for success, non-zero for access and other errors.
*/
static int
zfs_set_slabel_policy(const char *name, char *strval, cred_t *cr)
{
char ds_hexsl[MAXNAMELEN];
bslabel_t ds_sl, new_sl;
boolean_t new_default = FALSE;
uint64_t zoned;
int needed_priv = -1;
int error;
/* First get the existing dataset label. */
error = dsl_prop_get(name, zfs_prop_to_name(ZFS_PROP_MLSLABEL),
1, sizeof (ds_hexsl), &ds_hexsl, NULL);
if (error != 0)
return (SET_ERROR(EPERM));
if (strcasecmp(strval, ZFS_MLSLABEL_DEFAULT) == 0)
new_default = TRUE;
/* The label must be translatable */
if (!new_default && (hexstr_to_label(strval, &new_sl) != 0))
return (SET_ERROR(EINVAL));
/*
* In a non-global zone, disallow attempts to set a label that
* doesn't match that of the zone; otherwise no other checks
* are needed.
*/
if (!INGLOBALZONE(curproc)) {
if (new_default || !blequal(&new_sl, CR_SL(CRED())))
return (SET_ERROR(EPERM));
return (0);
}
/*
* For global-zone datasets (i.e., those whose zoned property is
* "off", verify that the specified new label is valid for the
* global zone.
*/
if (dsl_prop_get_integer(name,
zfs_prop_to_name(ZFS_PROP_ZONED), &zoned, NULL))
return (SET_ERROR(EPERM));
if (!zoned) {
if (zfs_check_global_label(name, strval) != 0)
return (SET_ERROR(EPERM));
}
/*
* If the existing dataset label is nondefault, check if the
* dataset is mounted (label cannot be changed while mounted).
* Get the zfsvfs; if there isn't one, then the dataset isn't
* mounted (or isn't a dataset, doesn't exist, ...).
*/
if (strcasecmp(ds_hexsl, ZFS_MLSLABEL_DEFAULT) != 0) {
objset_t *os;
static char *setsl_tag = "setsl_tag";
/*
* Try to own the dataset; abort if there is any error,
* (e.g., already mounted, in use, or other error).
*/
error = dmu_objset_own(name, DMU_OST_ZFS, B_TRUE,
setsl_tag, &os);
if (error != 0)
return (SET_ERROR(EPERM));
dmu_objset_disown(os, setsl_tag);
if (new_default) {
needed_priv = PRIV_FILE_DOWNGRADE_SL;
goto out_check;
}
if (hexstr_to_label(strval, &new_sl) != 0)
return (SET_ERROR(EPERM));
if (blstrictdom(&ds_sl, &new_sl))
needed_priv = PRIV_FILE_DOWNGRADE_SL;
else if (blstrictdom(&new_sl, &ds_sl))
needed_priv = PRIV_FILE_UPGRADE_SL;
} else {
/* dataset currently has a default label */
if (!new_default)
needed_priv = PRIV_FILE_UPGRADE_SL;
}
out_check:
if (needed_priv != -1)
return (PRIV_POLICY(cr, needed_priv, B_FALSE, EPERM, NULL));
return (0);
}
#endif /* SECLABEL */
static int
zfs_secpolicy_setprop(const char *dsname, zfs_prop_t prop, nvpair_t *propval,
cred_t *cr)
{
char *strval;
/*
* Check permissions for special properties.
*/
switch (prop) {
case ZFS_PROP_ZONED:
/*
* Disallow setting of 'zoned' from within a local zone.
*/
if (!INGLOBALZONE(curthread))
return (SET_ERROR(EPERM));
break;
case ZFS_PROP_QUOTA:
case ZFS_PROP_FILESYSTEM_LIMIT:
case ZFS_PROP_SNAPSHOT_LIMIT:
if (!INGLOBALZONE(curthread)) {
uint64_t zoned;
char setpoint[ZFS_MAX_DATASET_NAME_LEN];
/*
* Unprivileged users are allowed to modify the
* limit on things *under* (ie. contained by)
* the thing they own.
*/
if (dsl_prop_get_integer(dsname, "jailed", &zoned,
setpoint))
return (SET_ERROR(EPERM));
if (!zoned || strlen(dsname) <= strlen(setpoint))
return (SET_ERROR(EPERM));
}
break;
case ZFS_PROP_MLSLABEL:
#ifdef SECLABEL
if (!is_system_labeled())
return (SET_ERROR(EPERM));
if (nvpair_value_string(propval, &strval) == 0) {
int err;
err = zfs_set_slabel_policy(dsname, strval, CRED());
if (err != 0)
return (err);
}
#else
return (EOPNOTSUPP);
#endif
break;
}
return (zfs_secpolicy_write_perms(dsname, zfs_prop_to_name(prop), cr));
}
/* ARGSUSED */
static int
zfs_secpolicy_set_fsacl(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
int error;
error = zfs_dozonecheck(zc->zc_name, cr);
if (error != 0)
return (error);
/*
* permission to set permissions will be evaluated later in
* dsl_deleg_can_allow()
*/
return (0);
}
/* ARGSUSED */
static int
zfs_secpolicy_rollback(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
return (zfs_secpolicy_write_perms(zc->zc_name,
ZFS_DELEG_PERM_ROLLBACK, cr));
}
/* ARGSUSED */
static int
zfs_secpolicy_send(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
dsl_pool_t *dp;
dsl_dataset_t *ds;
char *cp;
int error;
/*
* Generate the current snapshot name from the given objsetid, then
* use that name for the secpolicy/zone checks.
*/
cp = strchr(zc->zc_name, '@');
if (cp == NULL)
return (SET_ERROR(EINVAL));
error = dsl_pool_hold(zc->zc_name, FTAG, &dp);
if (error != 0)
return (error);
error = dsl_dataset_hold_obj(dp, zc->zc_sendobj, FTAG, &ds);
if (error != 0) {
dsl_pool_rele(dp, FTAG);
return (error);
}
dsl_dataset_name(ds, zc->zc_name);
error = zfs_secpolicy_write_perms_ds(zc->zc_name, ds,
ZFS_DELEG_PERM_SEND, cr);
dsl_dataset_rele(ds, FTAG);
dsl_pool_rele(dp, FTAG);
return (error);
}
/* ARGSUSED */
static int
zfs_secpolicy_send_new(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
return (zfs_secpolicy_write_perms(zc->zc_name,
ZFS_DELEG_PERM_SEND, cr));
}
/* ARGSUSED */
static int
zfs_secpolicy_deleg_share(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
vnode_t *vp;
int error;
if ((error = lookupname(zc->zc_value, UIO_SYSSPACE,
NO_FOLLOW, NULL, &vp)) != 0)
return (error);
/* Now make sure mntpnt and dataset are ZFS */
if (strcmp(vp->v_vfsp->mnt_stat.f_fstypename, "zfs") != 0 ||
(strcmp((char *)refstr_value(vp->v_vfsp->vfs_resource),
zc->zc_name) != 0)) {
VN_RELE(vp);
return (SET_ERROR(EPERM));
}
VN_RELE(vp);
return (dsl_deleg_access(zc->zc_name,
ZFS_DELEG_PERM_SHARE, cr));
}
int
zfs_secpolicy_share(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
if (!INGLOBALZONE(curthread))
return (SET_ERROR(EPERM));
if (secpolicy_nfs(cr) == 0) {
return (0);
} else {
return (zfs_secpolicy_deleg_share(zc, innvl, cr));
}
}
int
zfs_secpolicy_smb_acl(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
if (!INGLOBALZONE(curthread))
return (SET_ERROR(EPERM));
if (secpolicy_smb(cr) == 0) {
return (0);
} else {
return (zfs_secpolicy_deleg_share(zc, innvl, cr));
}
}
static int
zfs_get_parent(const char *datasetname, char *parent, int parentsize)
{
char *cp;
/*
* Remove the @bla or /bla from the end of the name to get the parent.
*/
(void) strncpy(parent, datasetname, parentsize);
cp = strrchr(parent, '@');
if (cp != NULL) {
cp[0] = '\0';
} else {
cp = strrchr(parent, '/');
if (cp == NULL)
return (SET_ERROR(ENOENT));
cp[0] = '\0';
}
return (0);
}
int
zfs_secpolicy_destroy_perms(const char *name, cred_t *cr)
{
int error;
if ((error = zfs_secpolicy_write_perms(name,
ZFS_DELEG_PERM_MOUNT, cr)) != 0)
return (error);
return (zfs_secpolicy_write_perms(name, ZFS_DELEG_PERM_DESTROY, cr));
}
/* ARGSUSED */
static int
zfs_secpolicy_destroy(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
return (zfs_secpolicy_destroy_perms(zc->zc_name, cr));
}
/*
* Destroying snapshots with delegated permissions requires
* descendant mount and destroy permissions.
*/
/* ARGSUSED */
static int
zfs_secpolicy_destroy_snaps(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
nvlist_t *snaps;
nvpair_t *pair, *nextpair;
int error = 0;
if (nvlist_lookup_nvlist(innvl, "snaps", &snaps) != 0)
return (SET_ERROR(EINVAL));
for (pair = nvlist_next_nvpair(snaps, NULL); pair != NULL;
pair = nextpair) {
nextpair = nvlist_next_nvpair(snaps, pair);
error = zfs_secpolicy_destroy_perms(nvpair_name(pair), cr);
if (error == ENOENT) {
/*
* Ignore any snapshots that don't exist (we consider
* them "already destroyed"). Remove the name from the
* nvl here in case the snapshot is created between
* now and when we try to destroy it (in which case
* we don't want to destroy it since we haven't
* checked for permission).
*/
fnvlist_remove_nvpair(snaps, pair);
error = 0;
}
if (error != 0)
break;
}
return (error);
}
int
zfs_secpolicy_rename_perms(const char *from, const char *to, cred_t *cr)
{
char parentname[ZFS_MAX_DATASET_NAME_LEN];
int error;
if ((error = zfs_secpolicy_write_perms(from,
ZFS_DELEG_PERM_RENAME, cr)) != 0)
return (error);
if ((error = zfs_secpolicy_write_perms(from,
ZFS_DELEG_PERM_MOUNT, cr)) != 0)
return (error);
if ((error = zfs_get_parent(to, parentname,
sizeof (parentname))) != 0)
return (error);
if ((error = zfs_secpolicy_write_perms(parentname,
ZFS_DELEG_PERM_CREATE, cr)) != 0)
return (error);
if ((error = zfs_secpolicy_write_perms(parentname,
ZFS_DELEG_PERM_MOUNT, cr)) != 0)
return (error);
return (error);
}
/* ARGSUSED */
static int
zfs_secpolicy_rename(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
char *at = NULL;
char *pound;
int error;
if ((pound = strchr(zc->zc_name, '#')) != NULL) {
*pound = '\0';
error = zfs_secpolicy_write_perms(zc->zc_name,
ZFS_DELEG_PERM_RENAME, cr);
if (error == 0) {
error = zfs_secpolicy_write_perms(zc->zc_name,
ZFS_DELEG_PERM_BOOKMARK, cr);
}
*pound = '#';
return (error);
}
if ((zc->zc_cookie & 1) != 0) {
/*
* This is recursive rename, so the starting snapshot might
* not exist. Check file system or volume permission instead.
*/
at = strchr(zc->zc_name, '@');
if (at == NULL)
return (EINVAL);
*at = '\0';
}
error = zfs_secpolicy_rename_perms(zc->zc_name, zc->zc_value, cr);
if (at != NULL)
*at = '@';
return (error);
}
/* ARGSUSED */
static int
zfs_secpolicy_promote(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
dsl_pool_t *dp;
dsl_dataset_t *clone;
int error;
error = zfs_secpolicy_write_perms(zc->zc_name,
ZFS_DELEG_PERM_PROMOTE, cr);
if (error != 0)
return (error);
error = dsl_pool_hold(zc->zc_name, FTAG, &dp);
if (error != 0)
return (error);
error = dsl_dataset_hold(dp, zc->zc_name, FTAG, &clone);
if (error == 0) {
char parentname[ZFS_MAX_DATASET_NAME_LEN];
dsl_dataset_t *origin = NULL;
dsl_dir_t *dd;
dd = clone->ds_dir;
error = dsl_dataset_hold_obj(dd->dd_pool,
dsl_dir_phys(dd)->dd_origin_obj, FTAG, &origin);
if (error != 0) {
dsl_dataset_rele(clone, FTAG);
dsl_pool_rele(dp, FTAG);
return (error);
}
error = zfs_secpolicy_write_perms_ds(zc->zc_name, clone,
ZFS_DELEG_PERM_MOUNT, cr);
dsl_dataset_name(origin, parentname);
if (error == 0) {
error = zfs_secpolicy_write_perms_ds(parentname, origin,
ZFS_DELEG_PERM_PROMOTE, cr);
}
dsl_dataset_rele(clone, FTAG);
dsl_dataset_rele(origin, FTAG);
}
dsl_pool_rele(dp, FTAG);
return (error);
}
/* ARGSUSED */
static int
zfs_secpolicy_recv(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
int error;
if ((error = zfs_secpolicy_write_perms(zc->zc_name,
ZFS_DELEG_PERM_RECEIVE, cr)) != 0)
return (error);
if ((error = zfs_secpolicy_write_perms(zc->zc_name,
ZFS_DELEG_PERM_MOUNT, cr)) != 0)
return (error);
return (zfs_secpolicy_write_perms(zc->zc_name,
ZFS_DELEG_PERM_CREATE, cr));
}
int
zfs_secpolicy_snapshot_perms(const char *name, cred_t *cr)
{
return (zfs_secpolicy_write_perms(name,
ZFS_DELEG_PERM_SNAPSHOT, cr));
}
/*
* Check for permission to create each snapshot in the nvlist.
*/
/* ARGSUSED */
static int
zfs_secpolicy_snapshot(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
nvlist_t *snaps;
int error;
nvpair_t *pair;
if (nvlist_lookup_nvlist(innvl, "snaps", &snaps) != 0)
return (SET_ERROR(EINVAL));
for (pair = nvlist_next_nvpair(snaps, NULL); pair != NULL;
pair = nvlist_next_nvpair(snaps, pair)) {
char *name = nvpair_name(pair);
char *atp = strchr(name, '@');
if (atp == NULL) {
error = SET_ERROR(EINVAL);
break;
}
*atp = '\0';
error = zfs_secpolicy_snapshot_perms(name, cr);
*atp = '@';
if (error != 0)
break;
}
return (error);
}
/*
* Check for permission to create each snapshot in the nvlist.
*/
/* ARGSUSED */
static int
zfs_secpolicy_bookmark(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
int error = 0;
for (nvpair_t *pair = nvlist_next_nvpair(innvl, NULL);
pair != NULL; pair = nvlist_next_nvpair(innvl, pair)) {
char *name = nvpair_name(pair);
char *hashp = strchr(name, '#');
if (hashp == NULL) {
error = SET_ERROR(EINVAL);
break;
}
*hashp = '\0';
error = zfs_secpolicy_write_perms(name,
ZFS_DELEG_PERM_BOOKMARK, cr);
*hashp = '#';
if (error != 0)
break;
}
return (error);
}
/* ARGSUSED */
static int
zfs_secpolicy_remap(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
return (zfs_secpolicy_write_perms(zc->zc_name,
ZFS_DELEG_PERM_REMAP, cr));
}
/* ARGSUSED */
static int
zfs_secpolicy_destroy_bookmarks(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
nvpair_t *pair, *nextpair;
int error = 0;
for (pair = nvlist_next_nvpair(innvl, NULL); pair != NULL;
pair = nextpair) {
char *name = nvpair_name(pair);
char *hashp = strchr(name, '#');
nextpair = nvlist_next_nvpair(innvl, pair);
if (hashp == NULL) {
error = SET_ERROR(EINVAL);
break;
}
*hashp = '\0';
error = zfs_secpolicy_write_perms(name,
ZFS_DELEG_PERM_DESTROY, cr);
*hashp = '#';
if (error == ENOENT) {
/*
* Ignore any filesystems that don't exist (we consider
* their bookmarks "already destroyed"). Remove
* the name from the nvl here in case the filesystem
* is created between now and when we try to destroy
* the bookmark (in which case we don't want to
* destroy it since we haven't checked for permission).
*/
fnvlist_remove_nvpair(innvl, pair);
error = 0;
}
if (error != 0)
break;
}
return (error);
}
/* ARGSUSED */
static int
zfs_secpolicy_log_history(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
/*
* Even root must have a proper TSD so that we know what pool
* to log to.
*/
if (tsd_get(zfs_allow_log_key) == NULL)
return (SET_ERROR(EPERM));
return (0);
}
static int
zfs_secpolicy_create_clone(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
char parentname[ZFS_MAX_DATASET_NAME_LEN];
int error;
char *origin;
if ((error = zfs_get_parent(zc->zc_name, parentname,
sizeof (parentname))) != 0)
return (error);
if (nvlist_lookup_string(innvl, "origin", &origin) == 0 &&
(error = zfs_secpolicy_write_perms(origin,
ZFS_DELEG_PERM_CLONE, cr)) != 0)
return (error);
if ((error = zfs_secpolicy_write_perms(parentname,
ZFS_DELEG_PERM_CREATE, cr)) != 0)
return (error);
return (zfs_secpolicy_write_perms(parentname,
ZFS_DELEG_PERM_MOUNT, cr));
}
/*
* Policy for pool operations - create/destroy pools, add vdevs, etc. Requires
* SYS_CONFIG privilege, which is not available in a local zone.
*/
/* ARGSUSED */
static int
zfs_secpolicy_config(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
if (secpolicy_sys_config(cr, B_FALSE) != 0)
return (SET_ERROR(EPERM));
return (0);
}
/*
* Policy for object to name lookups.
*/
/* ARGSUSED */
static int
zfs_secpolicy_diff(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
int error;
if ((error = secpolicy_sys_config(cr, B_FALSE)) == 0)
return (0);
error = zfs_secpolicy_write_perms(zc->zc_name, ZFS_DELEG_PERM_DIFF, cr);
return (error);
}
/*
* Policy for fault injection. Requires all privileges.
*/
/* ARGSUSED */
static int
zfs_secpolicy_inject(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
return (secpolicy_zinject(cr));
}
/* ARGSUSED */
static int
zfs_secpolicy_inherit_prop(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
zfs_prop_t prop = zfs_name_to_prop(zc->zc_value);
if (prop == ZPROP_INVAL) {
if (!zfs_prop_user(zc->zc_value))
return (SET_ERROR(EINVAL));
return (zfs_secpolicy_write_perms(zc->zc_name,
ZFS_DELEG_PERM_USERPROP, cr));
} else {
return (zfs_secpolicy_setprop(zc->zc_name, prop,
NULL, cr));
}
}
static int
zfs_secpolicy_userspace_one(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
int err = zfs_secpolicy_read(zc, innvl, cr);
if (err)
return (err);
if (zc->zc_objset_type >= ZFS_NUM_USERQUOTA_PROPS)
return (SET_ERROR(EINVAL));
if (zc->zc_value[0] == 0) {
/*
* They are asking about a posix uid/gid. If it's
* themself, allow it.
*/
if (zc->zc_objset_type == ZFS_PROP_USERUSED ||
zc->zc_objset_type == ZFS_PROP_USERQUOTA) {
if (zc->zc_guid == crgetuid(cr))
return (0);
} else {
if (groupmember(zc->zc_guid, cr))
return (0);
}
}
return (zfs_secpolicy_write_perms(zc->zc_name,
userquota_perms[zc->zc_objset_type], cr));
}
static int
zfs_secpolicy_userspace_many(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
int err = zfs_secpolicy_read(zc, innvl, cr);
if (err)
return (err);
if (zc->zc_objset_type >= ZFS_NUM_USERQUOTA_PROPS)
return (SET_ERROR(EINVAL));
return (zfs_secpolicy_write_perms(zc->zc_name,
userquota_perms[zc->zc_objset_type], cr));
}
/* ARGSUSED */
static int
zfs_secpolicy_userspace_upgrade(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
return (zfs_secpolicy_setprop(zc->zc_name, ZFS_PROP_VERSION,
NULL, cr));
}
/* ARGSUSED */
static int
zfs_secpolicy_hold(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
nvpair_t *pair;
nvlist_t *holds;
int error;
error = nvlist_lookup_nvlist(innvl, "holds", &holds);
if (error != 0)
return (SET_ERROR(EINVAL));
for (pair = nvlist_next_nvpair(holds, NULL); pair != NULL;
pair = nvlist_next_nvpair(holds, pair)) {
char fsname[ZFS_MAX_DATASET_NAME_LEN];
error = dmu_fsname(nvpair_name(pair), fsname);
if (error != 0)
return (error);
error = zfs_secpolicy_write_perms(fsname,
ZFS_DELEG_PERM_HOLD, cr);
if (error != 0)
return (error);
}
return (0);
}
/* ARGSUSED */
static int
zfs_secpolicy_release(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
nvpair_t *pair;
int error;
for (pair = nvlist_next_nvpair(innvl, NULL); pair != NULL;
pair = nvlist_next_nvpair(innvl, pair)) {
char fsname[ZFS_MAX_DATASET_NAME_LEN];
error = dmu_fsname(nvpair_name(pair), fsname);
if (error != 0)
return (error);
error = zfs_secpolicy_write_perms(fsname,
ZFS_DELEG_PERM_RELEASE, cr);
if (error != 0)
return (error);
}
return (0);
}
/*
* Policy for allowing temporary snapshots to be taken or released
*/
static int
zfs_secpolicy_tmp_snapshot(zfs_cmd_t *zc, nvlist_t *innvl, cred_t *cr)
{
/*
* A temporary snapshot is the same as a snapshot,
* hold, destroy and release all rolled into one.
* Delegated diff alone is sufficient that we allow this.
*/
int error;
if ((error = zfs_secpolicy_write_perms(zc->zc_name,
ZFS_DELEG_PERM_DIFF, cr)) == 0)
return (0);
error = zfs_secpolicy_snapshot_perms(zc->zc_name, cr);
if (error == 0)
error = zfs_secpolicy_hold(zc, innvl, cr);
if (error == 0)
error = zfs_secpolicy_release(zc, innvl, cr);
if (error == 0)
error = zfs_secpolicy_destroy(zc, innvl, cr);
return (error);
}
/*
* Returns the nvlist as specified by the user in the zfs_cmd_t.
*/
static int
get_nvlist(uint64_t nvl, uint64_t size, int iflag, nvlist_t **nvp)
{
char *packed;
int error;
nvlist_t *list = NULL;
/*
* Read in and unpack the user-supplied nvlist.
*/
if (size == 0)
return (SET_ERROR(EINVAL));
packed = kmem_alloc(size, KM_SLEEP);
if ((error = ddi_copyin((void *)(uintptr_t)nvl, packed, size,
iflag)) != 0) {
kmem_free(packed, size);
return (SET_ERROR(EFAULT));
}
if ((error = nvlist_unpack(packed, size, &list, 0)) != 0) {
kmem_free(packed, size);
return (error);
}
kmem_free(packed, size);
*nvp = list;
return (0);
}
/*
* Reduce the size of this nvlist until it can be serialized in 'max' bytes.
* Entries will be removed from the end of the nvlist, and one int32 entry
* named "N_MORE_ERRORS" will be added indicating how many entries were
* removed.
*/
static int
nvlist_smush(nvlist_t *errors, size_t max)
{
size_t size;
size = fnvlist_size(errors);
if (size > max) {
nvpair_t *more_errors;
int n = 0;
if (max < 1024)
return (SET_ERROR(ENOMEM));
fnvlist_add_int32(errors, ZPROP_N_MORE_ERRORS, 0);
more_errors = nvlist_prev_nvpair(errors, NULL);
do {
nvpair_t *pair = nvlist_prev_nvpair(errors,
more_errors);
fnvlist_remove_nvpair(errors, pair);
n++;
size = fnvlist_size(errors);
} while (size > max);
fnvlist_remove_nvpair(errors, more_errors);
fnvlist_add_int32(errors, ZPROP_N_MORE_ERRORS, n);
ASSERT3U(fnvlist_size(errors), <=, max);
}
return (0);
}
static int
put_nvlist(zfs_cmd_t *zc, nvlist_t *nvl)
{
char *packed = NULL;
int error = 0;
size_t size;
size = fnvlist_size(nvl);
if (size > zc->zc_nvlist_dst_size) {
/*
* Solaris returns ENOMEM here, because even if an error is
* returned from an ioctl(2), new zc_nvlist_dst_size will be
* passed to the userland. This is not the case for FreeBSD.
* We need to return 0, so the kernel will copy the
* zc_nvlist_dst_size back and the userland can discover that a
* bigger buffer is needed.
*/
error = 0;
} else {
packed = fnvlist_pack(nvl, &size);
if (ddi_copyout(packed, (void *)(uintptr_t)zc->zc_nvlist_dst,
size, zc->zc_iflags) != 0)
error = SET_ERROR(EFAULT);
fnvlist_pack_free(packed, size);
}
zc->zc_nvlist_dst_size = size;
zc->zc_nvlist_dst_filled = B_TRUE;
return (error);
}
int
getzfsvfs_impl(objset_t *os, vfs_t **vfsp)
{
zfsvfs_t *zfvp;
int error = 0;
if (dmu_objset_type(os) != DMU_OST_ZFS) {
return (SET_ERROR(EINVAL));
}
mutex_enter(&os->os_user_ptr_lock);
zfvp = dmu_objset_get_user(os);
if (zfvp) {
*vfsp = zfvp->z_vfs;
vfs_ref(zfvp->z_vfs);
} else {
error = SET_ERROR(ESRCH);
}
mutex_exit(&os->os_user_ptr_lock);
return (error);
}
int
getzfsvfs(const char *dsname, zfsvfs_t **zfvp)
{
objset_t *os;
vfs_t *vfsp;
int error;
error = dmu_objset_hold(dsname, FTAG, &os);
if (error != 0)
return (error);
error = getzfsvfs_impl(os, &vfsp);
dmu_objset_rele(os, FTAG);
if (error != 0)
return (error);
error = vfs_busy(vfsp, 0);
vfs_rel(vfsp);
if (error != 0) {
*zfvp = NULL;
error = SET_ERROR(ESRCH);
} else {
*zfvp = vfsp->vfs_data;
}
return (error);
}
/*
* Find a zfsvfs_t for a mounted filesystem, or create our own, in which
* case its z_vfs will be NULL, and it will be opened as the owner.
* If 'writer' is set, the z_teardown_lock will be held for RW_WRITER,
* which prevents all vnode ops from running.
*/
static int
zfsvfs_hold(const char *name, void *tag, zfsvfs_t **zfvp, boolean_t writer)
{
int error = 0;
if (getzfsvfs(name, zfvp) != 0)
error = zfsvfs_create(name, zfvp);
if (error == 0) {
rrm_enter(&(*zfvp)->z_teardown_lock, (writer) ? RW_WRITER :
RW_READER, tag);
#ifdef illumos
if ((*zfvp)->z_unmounted) {
/*
* XXX we could probably try again, since the unmounting
* thread should be just about to disassociate the
* objset from the zfsvfs.
*/
rrm_exit(&(*zfvp)->z_teardown_lock, tag);
return (SET_ERROR(EBUSY));
}
#else
/*
* vfs_busy() ensures that the filesystem is not and
* can not be unmounted.
*/
ASSERT(!(*zfvp)->z_unmounted);
#endif
}
return (error);
}
static void
zfsvfs_rele(zfsvfs_t *zfsvfs, void *tag)
{
rrm_exit(&zfsvfs->z_teardown_lock, tag);
if (zfsvfs->z_vfs) {
#ifdef illumos
VFS_RELE(zfsvfs->z_vfs);
#else
vfs_unbusy(zfsvfs->z_vfs);
#endif
} else {
dmu_objset_disown(zfsvfs->z_os, zfsvfs);
zfsvfs_free(zfsvfs);
}
}
static int
zfs_ioc_pool_create(zfs_cmd_t *zc)
{
int error;
nvlist_t *config, *props = NULL;
nvlist_t *rootprops = NULL;
nvlist_t *zplprops = NULL;
char *spa_name = zc->zc_name;
if (error = get_nvlist(zc->zc_nvlist_conf, zc->zc_nvlist_conf_size,
zc->zc_iflags, &config))
return (error);
if (zc->zc_nvlist_src_size != 0 && (error =
get_nvlist(zc->zc_nvlist_src, zc->zc_nvlist_src_size,
zc->zc_iflags, &props))) {
nvlist_free(config);
return (error);
}
if (props) {
nvlist_t *nvl = NULL;
uint64_t version = SPA_VERSION;
char *tname;
(void) nvlist_lookup_uint64(props,
zpool_prop_to_name(ZPOOL_PROP_VERSION), &version);
if (!SPA_VERSION_IS_SUPPORTED(version)) {
error = SET_ERROR(EINVAL);
goto pool_props_bad;
}
(void) nvlist_lookup_nvlist(props, ZPOOL_ROOTFS_PROPS, &nvl);
if (nvl) {
error = nvlist_dup(nvl, &rootprops, KM_SLEEP);
if (error != 0) {
nvlist_free(config);
nvlist_free(props);
return (error);
}
(void) nvlist_remove_all(props, ZPOOL_ROOTFS_PROPS);
}
VERIFY(nvlist_alloc(&zplprops, NV_UNIQUE_NAME, KM_SLEEP) == 0);
error = zfs_fill_zplprops_root(version, rootprops,
zplprops, NULL);
if (error != 0)
goto pool_props_bad;
if (nvlist_lookup_string(props,
zpool_prop_to_name(ZPOOL_PROP_TNAME), &tname) == 0)
spa_name = tname;
}
error = spa_create(zc->zc_name, config, props, zplprops);
/*
* Set the remaining root properties
*/
if (!error && (error = zfs_set_prop_nvlist(spa_name,
ZPROP_SRC_LOCAL, rootprops, NULL)) != 0)
(void) spa_destroy(spa_name);
pool_props_bad:
nvlist_free(rootprops);
nvlist_free(zplprops);
nvlist_free(config);
nvlist_free(props);
return (error);
}
static int
zfs_ioc_pool_destroy(zfs_cmd_t *zc)
{
int error;
zfs_log_history(zc);
error = spa_destroy(zc->zc_name);
if (error == 0)
zvol_remove_minors(zc->zc_name);
return (error);
}
static int
zfs_ioc_pool_import(zfs_cmd_t *zc)
{
nvlist_t *config, *props = NULL;
uint64_t guid;
int error;
if ((error = get_nvlist(zc->zc_nvlist_conf, zc->zc_nvlist_conf_size,
zc->zc_iflags, &config)) != 0)
return (error);
if (zc->zc_nvlist_src_size != 0 && (error =
get_nvlist(zc->zc_nvlist_src, zc->zc_nvlist_src_size,
zc->zc_iflags, &props))) {
nvlist_free(config);
return (error);
}
if (nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_GUID, &guid) != 0 ||
guid != zc->zc_guid)
error = SET_ERROR(EINVAL);
else
error = spa_import(zc->zc_name, config, props, zc->zc_cookie);
if (zc->zc_nvlist_dst != 0) {
int err;
if ((err = put_nvlist(zc, config)) != 0)
error = err;
}
nvlist_free(config);
nvlist_free(props);
return (error);
}
static int
zfs_ioc_pool_export(zfs_cmd_t *zc)
{
int error;
boolean_t force = (boolean_t)zc->zc_cookie;
boolean_t hardforce = (boolean_t)zc->zc_guid;
zfs_log_history(zc);
error = spa_export(zc->zc_name, NULL, force, hardforce);
if (error == 0)
zvol_remove_minors(zc->zc_name);
return (error);
}
static int
zfs_ioc_pool_configs(zfs_cmd_t *zc)
{
nvlist_t *configs;
int error;
if ((configs = spa_all_configs(&zc->zc_cookie)) == NULL)
return (SET_ERROR(EEXIST));
error = put_nvlist(zc, configs);
nvlist_free(configs);
return (error);
}
/*
* inputs:
* zc_name name of the pool
*
* outputs:
* zc_cookie real errno
* zc_nvlist_dst config nvlist
* zc_nvlist_dst_size size of config nvlist
*/
static int
zfs_ioc_pool_stats(zfs_cmd_t *zc)
{
nvlist_t *config;
int error;
int ret = 0;
error = spa_get_stats(zc->zc_name, &config, zc->zc_value,
sizeof (zc->zc_value));
if (config != NULL) {
ret = put_nvlist(zc, config);
nvlist_free(config);
/*
* The config may be present even if 'error' is non-zero.
* In this case we return success, and preserve the real errno
* in 'zc_cookie'.
*/
zc->zc_cookie = error;
} else {
ret = error;
}
return (ret);
}
/*
* Try to import the given pool, returning pool stats as appropriate so that
* user land knows which devices are available and overall pool health.
*/
static int
zfs_ioc_pool_tryimport(zfs_cmd_t *zc)
{
nvlist_t *tryconfig, *config;
int error;
if ((error = get_nvlist(zc->zc_nvlist_conf, zc->zc_nvlist_conf_size,
zc->zc_iflags, &tryconfig)) != 0)
return (error);
config = spa_tryimport(tryconfig);
nvlist_free(tryconfig);
if (config == NULL)
return (SET_ERROR(EINVAL));
error = put_nvlist(zc, config);
nvlist_free(config);
return (error);
}
/*
* inputs:
* zc_name name of the pool
* zc_cookie scan func (pool_scan_func_t)
* zc_flags scrub pause/resume flag (pool_scrub_cmd_t)
*/
static int
zfs_ioc_pool_scan(zfs_cmd_t *zc)
{
spa_t *spa;
int error;
if ((error = spa_open(zc->zc_name, &spa, FTAG)) != 0)
return (error);
if (zc->zc_flags >= POOL_SCRUB_FLAGS_END)
return (SET_ERROR(EINVAL));
if (zc->zc_flags == POOL_SCRUB_PAUSE)
error = spa_scrub_pause_resume(spa, POOL_SCRUB_PAUSE);
else if (zc->zc_cookie == POOL_SCAN_NONE)
error = spa_scan_stop(spa);
else
error = spa_scan(spa, zc->zc_cookie);
spa_close(spa, FTAG);
return (error);
}
static int
zfs_ioc_pool_freeze(zfs_cmd_t *zc)
{
spa_t *spa;
int error;
error = spa_open(zc->zc_name, &spa, FTAG);
if (error == 0) {
spa_freeze(spa);
spa_close(spa, FTAG);
}
return (error);
}
static int
zfs_ioc_pool_upgrade(zfs_cmd_t *zc)
{
spa_t *spa;
int error;
if ((error = spa_open(zc->zc_name, &spa, FTAG)) != 0)
return (error);
if (zc->zc_cookie < spa_version(spa) ||
!SPA_VERSION_IS_SUPPORTED(zc->zc_cookie)) {
spa_close(spa, FTAG);
return (SET_ERROR(EINVAL));
}
spa_upgrade(spa, zc->zc_cookie);
spa_close(spa, FTAG);
return (error);
}
static int
zfs_ioc_pool_get_history(zfs_cmd_t *zc)
{
spa_t *spa;
char *hist_buf;
uint64_t size;
int error;
if ((size = zc->zc_history_len) == 0)
return (SET_ERROR(EINVAL));
if ((error = spa_open(zc->zc_name, &spa, FTAG)) != 0)
return (error);
if (spa_version(spa) < SPA_VERSION_ZPOOL_HISTORY) {
spa_close(spa, FTAG);
return (SET_ERROR(ENOTSUP));
}
hist_buf = kmem_alloc(size, KM_SLEEP);
if ((error = spa_history_get(spa, &zc->zc_history_offset,
&zc->zc_history_len, hist_buf)) == 0) {
error = ddi_copyout(hist_buf,
(void *)(uintptr_t)zc->zc_history,
zc->zc_history_len, zc->zc_iflags);
}
spa_close(spa, FTAG);
kmem_free(hist_buf, size);
return (error);
}
static int
zfs_ioc_pool_reguid(zfs_cmd_t *zc)
{
spa_t *spa;
int error;
error = spa_open(zc->zc_name, &spa, FTAG);
if (error == 0) {
error = spa_change_guid(spa);
spa_close(spa, FTAG);
}
return (error);
}
static int
zfs_ioc_dsobj_to_dsname(zfs_cmd_t *zc)
{
return (dsl_dsobj_to_dsname(zc->zc_name, zc->zc_obj, zc->zc_value));
}
/*
* inputs:
* zc_name name of filesystem
* zc_obj object to find
*
* outputs:
* zc_value name of object
*/
static int
zfs_ioc_obj_to_path(zfs_cmd_t *zc)
{
objset_t *os;
int error;
/* XXX reading from objset not owned */
if ((error = dmu_objset_hold(zc->zc_name, FTAG, &os)) != 0)
return (error);
if (dmu_objset_type(os) != DMU_OST_ZFS) {
dmu_objset_rele(os, FTAG);
return (SET_ERROR(EINVAL));
}
error = zfs_obj_to_path(os, zc->zc_obj, zc->zc_value,
sizeof (zc->zc_value));
dmu_objset_rele(os, FTAG);
return (error);
}
/*
* inputs:
* zc_name name of filesystem
* zc_obj object to find
*
* outputs:
* zc_stat stats on object
* zc_value path to object
*/
static int
zfs_ioc_obj_to_stats(zfs_cmd_t *zc)
{
objset_t *os;
int error;
/* XXX reading from objset not owned */
if ((error = dmu_objset_hold(zc->zc_name, FTAG, &os)) != 0)
return (error);
if (dmu_objset_type(os) != DMU_OST_ZFS) {
dmu_objset_rele(os, FTAG);
return (SET_ERROR(EINVAL));
}
error = zfs_obj_to_stats(os, zc->zc_obj, &zc->zc_stat, zc->zc_value,
sizeof (zc->zc_value));
dmu_objset_rele(os, FTAG);
return (error);
}
static int
zfs_ioc_vdev_add(zfs_cmd_t *zc)
{
spa_t *spa;
int error;
nvlist_t *config, **l2cache, **spares;
uint_t nl2cache = 0, nspares = 0;
error = spa_open(zc->zc_name, &spa, FTAG);
if (error != 0)
return (error);
error = get_nvlist(zc->zc_nvlist_conf, zc->zc_nvlist_conf_size,
zc->zc_iflags, &config);
(void) nvlist_lookup_nvlist_array(config, ZPOOL_CONFIG_L2CACHE,
&l2cache, &nl2cache);
(void) nvlist_lookup_nvlist_array(config, ZPOOL_CONFIG_SPARES,
&spares, &nspares);
#ifdef illumos
/*
* A root pool with concatenated devices is not supported.
* Thus, can not add a device to a root pool.
*
* Intent log device can not be added to a rootpool because
* during mountroot, zil is replayed, a seperated log device
* can not be accessed during the mountroot time.
*
* l2cache and spare devices are ok to be added to a rootpool.
*/
if (spa_bootfs(spa) != 0 && nl2cache == 0 && nspares == 0) {
nvlist_free(config);
spa_close(spa, FTAG);
return (SET_ERROR(EDOM));
}
#endif /* illumos */
if (error == 0) {
error = spa_vdev_add(spa, config);
nvlist_free(config);
}
spa_close(spa, FTAG);
return (error);
}
/*
* inputs:
* zc_name name of the pool
* zc_guid guid of vdev to remove
* zc_cookie cancel removal
*/
static int
zfs_ioc_vdev_remove(zfs_cmd_t *zc)
{
spa_t *spa;
int error;
error = spa_open(zc->zc_name, &spa, FTAG);
if (error != 0)
return (error);
if (zc->zc_cookie != 0) {
error = spa_vdev_remove_cancel(spa);
} else {
error = spa_vdev_remove(spa, zc->zc_guid, B_FALSE);
}
spa_close(spa, FTAG);
return (error);
}
static int
zfs_ioc_vdev_set_state(zfs_cmd_t *zc)
{
spa_t *spa;
int error;
vdev_state_t newstate = VDEV_STATE_UNKNOWN;
if ((error = spa_open(zc->zc_name, &spa, FTAG)) != 0)
return (error);
switch (zc->zc_cookie) {
case VDEV_STATE_ONLINE:
error = vdev_online(spa, zc->zc_guid, zc->zc_obj, &newstate);
break;
case VDEV_STATE_OFFLINE:
error = vdev_offline(spa, zc->zc_guid, zc->zc_obj);
break;
case VDEV_STATE_FAULTED:
if (zc->zc_obj != VDEV_AUX_ERR_EXCEEDED &&
zc->zc_obj != VDEV_AUX_EXTERNAL)
zc->zc_obj = VDEV_AUX_ERR_EXCEEDED;
error = vdev_fault(spa, zc->zc_guid, zc->zc_obj);
break;
case VDEV_STATE_DEGRADED:
if (zc->zc_obj != VDEV_AUX_ERR_EXCEEDED &&
zc->zc_obj != VDEV_AUX_EXTERNAL)
zc->zc_obj = VDEV_AUX_ERR_EXCEEDED;
error = vdev_degrade(spa, zc->zc_guid, zc->zc_obj);
break;
default:
error = SET_ERROR(EINVAL);
}
zc->zc_cookie = newstate;
spa_close(spa, FTAG);
return (error);
}
static int
zfs_ioc_vdev_attach(zfs_cmd_t *zc)
{
spa_t *spa;
int replacing = zc->zc_cookie;
nvlist_t *config;
int error;
if ((error = spa_open(zc->zc_name, &spa, FTAG)) != 0)
return (error);
if ((error = get_nvlist(zc->zc_nvlist_conf, zc->zc_nvlist_conf_size,
zc->zc_iflags, &config)) == 0) {
error = spa_vdev_attach(spa, zc->zc_guid, config, replacing);
nvlist_free(config);
}
spa_close(spa, FTAG);
return (error);
}
static int
zfs_ioc_vdev_detach(zfs_cmd_t *zc)
{
spa_t *spa;
int error;
if ((error = spa_open(zc->zc_name, &spa, FTAG)) != 0)
return (error);
error = spa_vdev_detach(spa, zc->zc_guid, 0, B_FALSE);
spa_close(spa, FTAG);
return (error);
}
static int
zfs_ioc_vdev_split(zfs_cmd_t *zc)
{
spa_t *spa;
nvlist_t *config, *props = NULL;
int error;
boolean_t exp = !!(zc->zc_cookie & ZPOOL_EXPORT_AFTER_SPLIT);
if ((error = spa_open(zc->zc_name, &spa, FTAG)) != 0)
return (error);
if (error = get_nvlist(zc->zc_nvlist_conf, zc->zc_nvlist_conf_size,
zc->zc_iflags, &config)) {
spa_close(spa, FTAG);
return (error);
}
if (zc->zc_nvlist_src_size != 0 && (error =
get_nvlist(zc->zc_nvlist_src, zc->zc_nvlist_src_size,
zc->zc_iflags, &props))) {
spa_close(spa, FTAG);
nvlist_free(config);
return (error);
}
error = spa_vdev_split_mirror(spa, zc->zc_string, config, props, exp);
spa_close(spa, FTAG);
nvlist_free(config);
nvlist_free(props);
return (error);
}
static int
zfs_ioc_vdev_setpath(zfs_cmd_t *zc)
{
spa_t *spa;
char *path = zc->zc_value;
uint64_t guid = zc->zc_guid;
int error;
error = spa_open(zc->zc_name, &spa, FTAG);
if (error != 0)
return (error);
error = spa_vdev_setpath(spa, guid, path);
spa_close(spa, FTAG);
return (error);
}
static int
zfs_ioc_vdev_setfru(zfs_cmd_t *zc)
{
spa_t *spa;
char *fru = zc->zc_value;
uint64_t guid = zc->zc_guid;
int error;
error = spa_open(zc->zc_name, &spa, FTAG);
if (error != 0)
return (error);
error = spa_vdev_setfru(spa, guid, fru);
spa_close(spa, FTAG);
return (error);
}
static int
zfs_ioc_objset_stats_impl(zfs_cmd_t *zc, objset_t *os)
{
int error = 0;
nvlist_t *nv;
dmu_objset_fast_stat(os, &zc->zc_objset_stats);
if (zc->zc_nvlist_dst != 0 &&
(error = dsl_prop_get_all(os, &nv)) == 0) {
dmu_objset_stats(os, nv);
/*
* NB: zvol_get_stats() will read the objset contents,
* which we aren't supposed to do with a
* DS_MODE_USER hold, because it could be
* inconsistent. So this is a bit of a workaround...
* XXX reading with out owning
*/
if (!zc->zc_objset_stats.dds_inconsistent &&
dmu_objset_type(os) == DMU_OST_ZVOL) {
error = zvol_get_stats(os, nv);
if (error == EIO)
return (error);
VERIFY0(error);
}
error = put_nvlist(zc, nv);
nvlist_free(nv);
}
return (error);
}
/*
* inputs:
* zc_name name of filesystem
* zc_nvlist_dst_size size of buffer for property nvlist
*
* outputs:
* zc_objset_stats stats
* zc_nvlist_dst property nvlist
* zc_nvlist_dst_size size of property nvlist
*/
static int
zfs_ioc_objset_stats(zfs_cmd_t *zc)
{
objset_t *os;
int error;
error = dmu_objset_hold(zc->zc_name, FTAG, &os);
if (error == 0) {
error = zfs_ioc_objset_stats_impl(zc, os);
dmu_objset_rele(os, FTAG);
}
if (error == ENOMEM)
error = 0;
return (error);
}
/*
* inputs:
* zc_name name of filesystem
* zc_nvlist_dst_size size of buffer for property nvlist
*
* outputs:
* zc_nvlist_dst received property nvlist
* zc_nvlist_dst_size size of received property nvlist
*
* Gets received properties (distinct from local properties on or after
* SPA_VERSION_RECVD_PROPS) for callers who want to differentiate received from
* local property values.
*/
static int
zfs_ioc_objset_recvd_props(zfs_cmd_t *zc)
{
int error = 0;
nvlist_t *nv;
/*
* Without this check, we would return local property values if the
* caller has not already received properties on or after
* SPA_VERSION_RECVD_PROPS.
*/
if (!dsl_prop_get_hasrecvd(zc->zc_name))
return (SET_ERROR(ENOTSUP));
if (zc->zc_nvlist_dst != 0 &&
(error = dsl_prop_get_received(zc->zc_name, &nv)) == 0) {
error = put_nvlist(zc, nv);
nvlist_free(nv);
}
return (error);
}
static int
nvl_add_zplprop(objset_t *os, nvlist_t *props, zfs_prop_t prop)
{
uint64_t value;
int error;
/*
* zfs_get_zplprop() will either find a value or give us
* the default value (if there is one).
*/
if ((error = zfs_get_zplprop(os, prop, &value)) != 0)
return (error);
VERIFY(nvlist_add_uint64(props, zfs_prop_to_name(prop), value) == 0);
return (0);
}
/*
* inputs:
* zc_name name of filesystem
* zc_nvlist_dst_size size of buffer for zpl property nvlist
*
* outputs:
* zc_nvlist_dst zpl property nvlist
* zc_nvlist_dst_size size of zpl property nvlist
*/
static int
zfs_ioc_objset_zplprops(zfs_cmd_t *zc)
{
objset_t *os;
int err;
/* XXX reading without owning */
if (err = dmu_objset_hold(zc->zc_name, FTAG, &os))
return (err);
dmu_objset_fast_stat(os, &zc->zc_objset_stats);
/*
* NB: nvl_add_zplprop() will read the objset contents,
* which we aren't supposed to do with a DS_MODE_USER
* hold, because it could be inconsistent.
*/
if (zc->zc_nvlist_dst != 0 &&
!zc->zc_objset_stats.dds_inconsistent &&
dmu_objset_type(os) == DMU_OST_ZFS) {
nvlist_t *nv;
VERIFY(nvlist_alloc(&nv, NV_UNIQUE_NAME, KM_SLEEP) == 0);
if ((err = nvl_add_zplprop(os, nv, ZFS_PROP_VERSION)) == 0 &&
(err = nvl_add_zplprop(os, nv, ZFS_PROP_NORMALIZE)) == 0 &&
(err = nvl_add_zplprop(os, nv, ZFS_PROP_UTF8ONLY)) == 0 &&
(err = nvl_add_zplprop(os, nv, ZFS_PROP_CASE)) == 0)
err = put_nvlist(zc, nv);
nvlist_free(nv);
} else {
err = SET_ERROR(ENOENT);
}
dmu_objset_rele(os, FTAG);
return (err);
}
boolean_t
dataset_name_hidden(const char *name)
{
/*
* Skip over datasets that are not visible in this zone,
* internal datasets (which have a $ in their name), and
* temporary datasets (which have a % in their name).
*/
if (strchr(name, '$') != NULL)
return (B_TRUE);
if (strchr(name, '%') != NULL)
return (B_TRUE);
if (!INGLOBALZONE(curthread) && !zone_dataset_visible(name, NULL))
return (B_TRUE);
return (B_FALSE);
}
/*
* inputs:
* zc_name name of filesystem
* zc_cookie zap cursor
* zc_nvlist_src iteration range nvlist
* zc_nvlist_src_size size of iteration range nvlist
*
* outputs:
* zc_name name of next filesystem
* zc_cookie zap cursor
* zc_objset_stats stats
* zc_nvlist_dst property nvlist
* zc_nvlist_dst_size size of property nvlist
*/
static int
zfs_ioc_dataset_list_next(zfs_cmd_t *zc)
{
objset_t *os;
int error;
char *p;
size_t orig_len = strlen(zc->zc_name);
top:
if (error = dmu_objset_hold(zc->zc_name, FTAG, &os)) {
if (error == ENOENT)
error = SET_ERROR(ESRCH);
return (error);
}
p = strrchr(zc->zc_name, '/');
if (p == NULL || p[1] != '\0')
(void) strlcat(zc->zc_name, "/", sizeof (zc->zc_name));
p = zc->zc_name + strlen(zc->zc_name);
do {
error = dmu_dir_list_next(os,
sizeof (zc->zc_name) - (p - zc->zc_name), p,
NULL, &zc->zc_cookie);
if (error == ENOENT)
error = SET_ERROR(ESRCH);
} while (error == 0 && dataset_name_hidden(zc->zc_name));
dmu_objset_rele(os, FTAG);
/*
* If it's an internal dataset (ie. with a '$' in its name),
* don't try to get stats for it, otherwise we'll return ENOENT.
*/
if (error == 0 && strchr(zc->zc_name, '$') == NULL) {
error = zfs_ioc_objset_stats(zc); /* fill in the stats */
if (error == ENOENT) {
/* We lost a race with destroy, get the next one. */
zc->zc_name[orig_len] = '\0';
goto top;
}
}
return (error);
}
/*
* inputs:
* zc_name name of filesystem
* zc_cookie zap cursor
* zc_nvlist_dst_size size of buffer for property nvlist
* zc_simple when set, only name is requested
*
* outputs:
* zc_name name of next snapshot
* zc_objset_stats stats
* zc_nvlist_dst property nvlist
* zc_nvlist_dst_size size of property nvlist
*/
static int
zfs_ioc_snapshot_list_next(zfs_cmd_t *zc)
{
int error;
objset_t *os, *ossnap;
dsl_dataset_t *ds;
uint64_t min_txg = 0, max_txg = 0;
if (zc->zc_nvlist_src_size != 0) {
nvlist_t *props = NULL;
error = get_nvlist(zc->zc_nvlist_src, zc->zc_nvlist_src_size,
zc->zc_iflags, &props);
if (error != 0)
return (error);
(void) nvlist_lookup_uint64(props, SNAP_ITER_MIN_TXG,
&min_txg);
(void) nvlist_lookup_uint64(props, SNAP_ITER_MAX_TXG,
&max_txg);
nvlist_free(props);
}
error = dmu_objset_hold(zc->zc_name, FTAG, &os);
if (error != 0) {
return (error == ENOENT ? ESRCH : error);
}
/*
* A dataset name of maximum length cannot have any snapshots,
* so exit immediately.
*/
if (strlcat(zc->zc_name, "@", sizeof (zc->zc_name)) >=
ZFS_MAX_DATASET_NAME_LEN) {
dmu_objset_rele(os, FTAG);
return (SET_ERROR(ESRCH));
}
while (error == 0) {
if (issig(JUSTLOOKING) && issig(FORREAL)) {
error = SET_ERROR(EINTR);
break;
}
error = dmu_snapshot_list_next(os,
sizeof (zc->zc_name) - strlen(zc->zc_name),
zc->zc_name + strlen(zc->zc_name), &zc->zc_obj,
&zc->zc_cookie, NULL);
if (error == ENOENT) {
error = SET_ERROR(ESRCH);
break;
} else if (error != 0) {
break;
}
error = dsl_dataset_hold_obj(dmu_objset_pool(os), zc->zc_obj,
FTAG, &ds);
if (error != 0)
break;
if ((min_txg != 0 && dsl_get_creationtxg(ds) < min_txg) ||
(max_txg != 0 && dsl_get_creationtxg(ds) > max_txg)) {
dsl_dataset_rele(ds, FTAG);
/* undo snapshot name append */
*(strchr(zc->zc_name, '@') + 1) = '\0';
/* skip snapshot */
continue;
}
if (zc->zc_simple) {
dsl_dataset_rele(ds, FTAG);
break;
}
if ((error = dmu_objset_from_ds(ds, &ossnap)) != 0) {
dsl_dataset_rele(ds, FTAG);
break;
}
if ((error = zfs_ioc_objset_stats_impl(zc, ossnap)) != 0) {
dsl_dataset_rele(ds, FTAG);
break;
}
dsl_dataset_rele(ds, FTAG);
break;
}
dmu_objset_rele(os, FTAG);
/* if we failed, undo the @ that we tacked on to zc_name */
if (error != 0)
*strchr(zc->zc_name, '@') = '\0';
return (error);
}
static int
zfs_prop_set_userquota(const char *dsname, nvpair_t *pair)
{
const char *propname = nvpair_name(pair);
uint64_t *valary;
unsigned int vallen;
const char *domain;
char *dash;
zfs_userquota_prop_t type;
uint64_t rid;
uint64_t quota;
zfsvfs_t *zfsvfs;
int err;
if (nvpair_type(pair) == DATA_TYPE_NVLIST) {
nvlist_t *attrs;
VERIFY(nvpair_value_nvlist(pair, &attrs) == 0);
if (nvlist_lookup_nvpair(attrs, ZPROP_VALUE,
&pair) != 0)
return (SET_ERROR(EINVAL));
}
/*
* A correctly constructed propname is encoded as
* userquota@<rid>-<domain>.
*/
if ((dash = strchr(propname, '-')) == NULL ||
nvpair_value_uint64_array(pair, &valary, &vallen) != 0 ||
vallen != 3)
return (SET_ERROR(EINVAL));
domain = dash + 1;
type = valary[0];
rid = valary[1];
quota = valary[2];
err = zfsvfs_hold(dsname, FTAG, &zfsvfs, B_FALSE);
if (err == 0) {
err = zfs_set_userquota(zfsvfs, type, domain, rid, quota);
zfsvfs_rele(zfsvfs, FTAG);
}
return (err);
}
/*
* If the named property is one that has a special function to set its value,
* return 0 on success and a positive error code on failure; otherwise if it is
* not one of the special properties handled by this function, return -1.
*
* XXX: It would be better for callers of the property interface if we handled
* these special cases in dsl_prop.c (in the dsl layer).
*/
static int
zfs_prop_set_special(const char *dsname, zprop_source_t source,
nvpair_t *pair)
{
const char *propname = nvpair_name(pair);
zfs_prop_t prop = zfs_name_to_prop(propname);
uint64_t intval;
int err = -1;
if (prop == ZPROP_INVAL) {
if (zfs_prop_userquota(propname))
return (zfs_prop_set_userquota(dsname, pair));
return (-1);
}
if (nvpair_type(pair) == DATA_TYPE_NVLIST) {
nvlist_t *attrs;
VERIFY(nvpair_value_nvlist(pair, &attrs) == 0);
VERIFY(nvlist_lookup_nvpair(attrs, ZPROP_VALUE,
&pair) == 0);
}
if (zfs_prop_get_type(prop) == PROP_TYPE_STRING)
return (-1);
VERIFY(0 == nvpair_value_uint64(pair, &intval));
switch (prop) {
case ZFS_PROP_QUOTA:
err = dsl_dir_set_quota(dsname, source, intval);
break;
case ZFS_PROP_REFQUOTA:
err = dsl_dataset_set_refquota(dsname, source, intval);
break;
case ZFS_PROP_FILESYSTEM_LIMIT:
case ZFS_PROP_SNAPSHOT_LIMIT:
if (intval == UINT64_MAX) {
/* clearing the limit, just do it */
err = 0;
} else {
err = dsl_dir_activate_fs_ss_limit(dsname);
}
/*
* Set err to -1 to force the zfs_set_prop_nvlist code down the
* default path to set the value in the nvlist.
*/
if (err == 0)
err = -1;
break;
case ZFS_PROP_RESERVATION:
err = dsl_dir_set_reservation(dsname, source, intval);
break;
case ZFS_PROP_REFRESERVATION:
err = dsl_dataset_set_refreservation(dsname, source, intval);
break;
case ZFS_PROP_VOLSIZE:
err = zvol_set_volsize(dsname, intval);
break;
case ZFS_PROP_VERSION:
{
zfsvfs_t *zfsvfs;
if ((err = zfsvfs_hold(dsname, FTAG, &zfsvfs, B_TRUE)) != 0)
break;
err = zfs_set_version(zfsvfs, intval);
zfsvfs_rele(zfsvfs, FTAG);
if (err == 0 && intval >= ZPL_VERSION_USERSPACE) {
zfs_cmd_t *zc;
zc = kmem_zalloc(sizeof (zfs_cmd_t), KM_SLEEP);
(void) strcpy(zc->zc_name, dsname);
(void) zfs_ioc_userspace_upgrade(zc);
kmem_free(zc, sizeof (zfs_cmd_t));
}
break;
}
default:
err = -1;
}
return (err);
}
/*
* This function is best effort. If it fails to set any of the given properties,
* it continues to set as many as it can and returns the last error
* encountered. If the caller provides a non-NULL errlist, it will be filled in
* with the list of names of all the properties that failed along with the
* corresponding error numbers.
*
* If every property is set successfully, zero is returned and errlist is not
* modified.
*/
int
zfs_set_prop_nvlist(const char *dsname, zprop_source_t source, nvlist_t *nvl,
nvlist_t *errlist)
{
nvpair_t *pair;
nvpair_t *propval;
int rv = 0;
uint64_t intval;
char *strval;
nvlist_t *genericnvl = fnvlist_alloc();
nvlist_t *retrynvl = fnvlist_alloc();
retry:
pair = NULL;
while ((pair = nvlist_next_nvpair(nvl, pair)) != NULL) {
const char *propname = nvpair_name(pair);
zfs_prop_t prop = zfs_name_to_prop(propname);
int err = 0;
/* decode the property value */
propval = pair;
if (nvpair_type(pair) == DATA_TYPE_NVLIST) {
nvlist_t *attrs;
attrs = fnvpair_value_nvlist(pair);
if (nvlist_lookup_nvpair(attrs, ZPROP_VALUE,
&propval) != 0)
err = SET_ERROR(EINVAL);
}
/* Validate value type */
if (err == 0 && prop == ZPROP_INVAL) {
if (zfs_prop_user(propname)) {
if (nvpair_type(propval) != DATA_TYPE_STRING)
err = SET_ERROR(EINVAL);
} else if (zfs_prop_userquota(propname)) {
if (nvpair_type(propval) !=
DATA_TYPE_UINT64_ARRAY)
err = SET_ERROR(EINVAL);
} else {
err = SET_ERROR(EINVAL);
}
} else if (err == 0) {
if (nvpair_type(propval) == DATA_TYPE_STRING) {
if (zfs_prop_get_type(prop) != PROP_TYPE_STRING)
err = SET_ERROR(EINVAL);
} else if (nvpair_type(propval) == DATA_TYPE_UINT64) {
const char *unused;
intval = fnvpair_value_uint64(propval);
switch (zfs_prop_get_type(prop)) {
case PROP_TYPE_NUMBER:
break;
case PROP_TYPE_STRING:
err = SET_ERROR(EINVAL);
break;
case PROP_TYPE_INDEX:
if (zfs_prop_index_to_string(prop,
intval, &unused) != 0)
err = SET_ERROR(EINVAL);
break;
default:
cmn_err(CE_PANIC,
"unknown property type");
}
} else {
err = SET_ERROR(EINVAL);
}
}
/* Validate permissions */
if (err == 0)
err = zfs_check_settable(dsname, pair, CRED());
if (err == 0) {
err = zfs_prop_set_special(dsname, source, pair);
if (err == -1) {
/*
* For better performance we build up a list of
* properties to set in a single transaction.
*/
err = nvlist_add_nvpair(genericnvl, pair);
} else if (err != 0 && nvl != retrynvl) {
/*
* This may be a spurious error caused by
* receiving quota and reservation out of order.
* Try again in a second pass.
*/
err = nvlist_add_nvpair(retrynvl, pair);
}
}
if (err != 0) {
if (errlist != NULL)
fnvlist_add_int32(errlist, propname, err);
rv = err;
}
}
if (nvl != retrynvl && !nvlist_empty(retrynvl)) {
nvl = retrynvl;
goto retry;
}
if (!nvlist_empty(genericnvl) &&
dsl_props_set(dsname, source, genericnvl) != 0) {
/*
* If this fails, we still want to set as many properties as we
* can, so try setting them individually.
*/
pair = NULL;
while ((pair = nvlist_next_nvpair(genericnvl, pair)) != NULL) {
const char *propname = nvpair_name(pair);
int err = 0;
propval = pair;
if (nvpair_type(pair) == DATA_TYPE_NVLIST) {
nvlist_t *attrs;
attrs = fnvpair_value_nvlist(pair);
propval = fnvlist_lookup_nvpair(attrs,
ZPROP_VALUE);
}
if (nvpair_type(propval) == DATA_TYPE_STRING) {
strval = fnvpair_value_string(propval);
err = dsl_prop_set_string(dsname, propname,
source, strval);
} else {
intval = fnvpair_value_uint64(propval);
err = dsl_prop_set_int(dsname, propname, source,
intval);
}
if (err != 0) {
if (errlist != NULL) {
fnvlist_add_int32(errlist, propname,
err);
}
rv = err;
}
}
}
nvlist_free(genericnvl);
nvlist_free(retrynvl);
return (rv);
}
/*
* Check that all the properties are valid user properties.
*/
static int
zfs_check_userprops(nvlist_t *nvl)
{
nvpair_t *pair = NULL;
while ((pair = nvlist_next_nvpair(nvl, pair)) != NULL) {
const char *propname = nvpair_name(pair);
if (!zfs_prop_user(propname) ||
nvpair_type(pair) != DATA_TYPE_STRING)
return (SET_ERROR(EINVAL));
if (strlen(propname) >= ZAP_MAXNAMELEN)
return (SET_ERROR(ENAMETOOLONG));
if (strlen(fnvpair_value_string(pair)) >= ZAP_MAXVALUELEN)
return (E2BIG);
}
return (0);
}
static void
props_skip(nvlist_t *props, nvlist_t *skipped, nvlist_t **newprops)
{
nvpair_t *pair;
VERIFY(nvlist_alloc(newprops, NV_UNIQUE_NAME, KM_SLEEP) == 0);
pair = NULL;
while ((pair = nvlist_next_nvpair(props, pair)) != NULL) {
if (nvlist_exists(skipped, nvpair_name(pair)))
continue;
VERIFY(nvlist_add_nvpair(*newprops, pair) == 0);
}
}
static int
clear_received_props(const char *dsname, nvlist_t *props,
nvlist_t *skipped)
{
int err = 0;
nvlist_t *cleared_props = NULL;
props_skip(props, skipped, &cleared_props);
if (!nvlist_empty(cleared_props)) {
/*
* Acts on local properties until the dataset has received
* properties at least once on or after SPA_VERSION_RECVD_PROPS.
*/
zprop_source_t flags = (ZPROP_SRC_NONE |
(dsl_prop_get_hasrecvd(dsname) ? ZPROP_SRC_RECEIVED : 0));
err = zfs_set_prop_nvlist(dsname, flags, cleared_props, NULL);
}
nvlist_free(cleared_props);
return (err);
}
/*
* inputs:
* zc_name name of filesystem
* zc_value name of property to set
* zc_nvlist_src{_size} nvlist of properties to apply
* zc_cookie received properties flag
*
* outputs:
* zc_nvlist_dst{_size} error for each unapplied received property
*/
static int
zfs_ioc_set_prop(zfs_cmd_t *zc)
{
nvlist_t *nvl;
boolean_t received = zc->zc_cookie;
zprop_source_t source = (received ? ZPROP_SRC_RECEIVED :
ZPROP_SRC_LOCAL);
nvlist_t *errors;
int error;
if ((error = get_nvlist(zc->zc_nvlist_src, zc->zc_nvlist_src_size,
zc->zc_iflags, &nvl)) != 0)
return (error);
if (received) {
nvlist_t *origprops;
if (dsl_prop_get_received(zc->zc_name, &origprops) == 0) {
(void) clear_received_props(zc->zc_name,
origprops, nvl);
nvlist_free(origprops);
}
error = dsl_prop_set_hasrecvd(zc->zc_name);
}
errors = fnvlist_alloc();
if (error == 0)
error = zfs_set_prop_nvlist(zc->zc_name, source, nvl, errors);
if (zc->zc_nvlist_dst != 0 && errors != NULL) {
(void) put_nvlist(zc, errors);
}
nvlist_free(errors);
nvlist_free(nvl);
return (error);
}
/*
* inputs:
* zc_name name of filesystem
* zc_value name of property to inherit
* zc_cookie revert to received value if TRUE
*
* outputs: none
*/
static int
zfs_ioc_inherit_prop(zfs_cmd_t *zc)
{
const char *propname = zc->zc_value;
zfs_prop_t prop = zfs_name_to_prop(propname);
boolean_t received = zc->zc_cookie;
zprop_source_t source = (received
? ZPROP_SRC_NONE /* revert to received value, if any */
: ZPROP_SRC_INHERITED); /* explicitly inherit */
if (received) {
nvlist_t *dummy;
nvpair_t *pair;
zprop_type_t type;
int err;
/*
* zfs_prop_set_special() expects properties in the form of an
* nvpair with type info.
*/
if (prop == ZPROP_INVAL) {
if (!zfs_prop_user(propname))
return (SET_ERROR(EINVAL));
type = PROP_TYPE_STRING;
} else if (prop == ZFS_PROP_VOLSIZE ||
prop == ZFS_PROP_VERSION) {
return (SET_ERROR(EINVAL));
} else {
type = zfs_prop_get_type(prop);
}
VERIFY(nvlist_alloc(&dummy, NV_UNIQUE_NAME, KM_SLEEP) == 0);
switch (type) {
case PROP_TYPE_STRING:
VERIFY(0 == nvlist_add_string(dummy, propname, ""));
break;
case PROP_TYPE_NUMBER:
case PROP_TYPE_INDEX:
VERIFY(0 == nvlist_add_uint64(dummy, propname, 0));
break;
default:
nvlist_free(dummy);
return (SET_ERROR(EINVAL));
}
pair = nvlist_next_nvpair(dummy, NULL);
err = zfs_prop_set_special(zc->zc_name, source, pair);
nvlist_free(dummy);
if (err != -1)
return (err); /* special property already handled */
} else {
/*
* Only check this in the non-received case. We want to allow
* 'inherit -S' to revert non-inheritable properties like quota
* and reservation to the received or default values even though
* they are not considered inheritable.
*/
if (prop != ZPROP_INVAL && !zfs_prop_inheritable(prop))
return (SET_ERROR(EINVAL));
}
/* property name has been validated by zfs_secpolicy_inherit_prop() */
return (dsl_prop_inherit(zc->zc_name, zc->zc_value, source));
}
static int
zfs_ioc_pool_set_props(zfs_cmd_t *zc)
{
nvlist_t *props;
spa_t *spa;
int error;
nvpair_t *pair;
if (error = get_nvlist(zc->zc_nvlist_src, zc->zc_nvlist_src_size,
zc->zc_iflags, &props))
return (error);
/*
* If the only property is the configfile, then just do a spa_lookup()
* to handle the faulted case.
*/
pair = nvlist_next_nvpair(props, NULL);
if (pair != NULL && strcmp(nvpair_name(pair),
zpool_prop_to_name(ZPOOL_PROP_CACHEFILE)) == 0 &&
nvlist_next_nvpair(props, pair) == NULL) {
mutex_enter(&spa_namespace_lock);
if ((spa = spa_lookup(zc->zc_name)) != NULL) {
spa_configfile_set(spa, props, B_FALSE);
spa_write_cachefile(spa, B_FALSE, B_TRUE);
}
mutex_exit(&spa_namespace_lock);
if (spa != NULL) {
nvlist_free(props);
return (0);
}
}
if ((error = spa_open(zc->zc_name, &spa, FTAG)) != 0) {
nvlist_free(props);
return (error);
}
error = spa_prop_set(spa, props);
nvlist_free(props);
spa_close(spa, FTAG);
return (error);
}
static int
zfs_ioc_pool_get_props(zfs_cmd_t *zc)
{
spa_t *spa;
int error;
nvlist_t *nvp = NULL;
if ((error = spa_open(zc->zc_name, &spa, FTAG)) != 0) {
/*
* If the pool is faulted, there may be properties we can still
* get (such as altroot and cachefile), so attempt to get them
* anyway.
*/
mutex_enter(&spa_namespace_lock);
if ((spa = spa_lookup(zc->zc_name)) != NULL)
error = spa_prop_get(spa, &nvp);
mutex_exit(&spa_namespace_lock);
} else {
error = spa_prop_get(spa, &nvp);
spa_close(spa, FTAG);
}
if (error == 0 && zc->zc_nvlist_dst != 0)
error = put_nvlist(zc, nvp);
else
error = SET_ERROR(EFAULT);
nvlist_free(nvp);
return (error);
}
/*
* inputs:
* zc_name name of filesystem
* zc_nvlist_src{_size} nvlist of delegated permissions
* zc_perm_action allow/unallow flag
*
* outputs: none
*/
static int
zfs_ioc_set_fsacl(zfs_cmd_t *zc)
{
int error;
nvlist_t *fsaclnv = NULL;
if ((error = get_nvlist(zc->zc_nvlist_src, zc->zc_nvlist_src_size,
zc->zc_iflags, &fsaclnv)) != 0)
return (error);
/*
* Verify nvlist is constructed correctly
*/
if ((error = zfs_deleg_verify_nvlist(fsaclnv)) != 0) {
nvlist_free(fsaclnv);
return (SET_ERROR(EINVAL));
}
/*
* If we don't have PRIV_SYS_MOUNT, then validate
* that user is allowed to hand out each permission in
* the nvlist(s)
*/
error = secpolicy_zfs(CRED());
if (error != 0) {
if (zc->zc_perm_action == B_FALSE) {
error = dsl_deleg_can_allow(zc->zc_name,
fsaclnv, CRED());
} else {
error = dsl_deleg_can_unallow(zc->zc_name,
fsaclnv, CRED());
}
}
if (error == 0)
error = dsl_deleg_set(zc->zc_name, fsaclnv, zc->zc_perm_action);
nvlist_free(fsaclnv);
return (error);
}
/*
* inputs:
* zc_name name of filesystem
*
* outputs:
* zc_nvlist_src{_size} nvlist of delegated permissions
*/
static int
zfs_ioc_get_fsacl(zfs_cmd_t *zc)
{
nvlist_t *nvp;
int error;
if ((error = dsl_deleg_get(zc->zc_name, &nvp)) == 0) {
error = put_nvlist(zc, nvp);
nvlist_free(nvp);
}
return (error);
}
/* ARGSUSED */
static void
zfs_create_cb(objset_t *os, void *arg, cred_t *cr, dmu_tx_t *tx)
{
zfs_creat_t *zct = arg;
zfs_create_fs(os, cr, zct->zct_zplprops, tx);
}
#define ZFS_PROP_UNDEFINED ((uint64_t)-1)
/*
* inputs:
* os parent objset pointer (NULL if root fs)
* fuids_ok fuids allowed in this version of the spa?
* sa_ok SAs allowed in this version of the spa?
* createprops list of properties requested by creator
*
* outputs:
* zplprops values for the zplprops we attach to the master node object
* is_ci true if requested file system will be purely case-insensitive
*
* Determine the settings for utf8only, normalization and
* casesensitivity. Specific values may have been requested by the
* creator and/or we can inherit values from the parent dataset. If
* the file system is of too early a vintage, a creator can not
* request settings for these properties, even if the requested
* setting is the default value. We don't actually want to create dsl
* properties for these, so remove them from the source nvlist after
* processing.
*/
static int
zfs_fill_zplprops_impl(objset_t *os, uint64_t zplver,
boolean_t fuids_ok, boolean_t sa_ok, nvlist_t *createprops,
nvlist_t *zplprops, boolean_t *is_ci)
{
uint64_t sense = ZFS_PROP_UNDEFINED;
uint64_t norm = ZFS_PROP_UNDEFINED;
uint64_t u8 = ZFS_PROP_UNDEFINED;
ASSERT(zplprops != NULL);
+ /* parent dataset must be a filesystem */
if (os != NULL && os->os_phys->os_type != DMU_OST_ZFS)
- return (SET_ERROR(EINVAL));
+ return (SET_ERROR(ZFS_ERR_WRONG_PARENT));
/*
* Pull out creator prop choices, if any.
*/
if (createprops) {
(void) nvlist_lookup_uint64(createprops,
zfs_prop_to_name(ZFS_PROP_VERSION), &zplver);
(void) nvlist_lookup_uint64(createprops,
zfs_prop_to_name(ZFS_PROP_NORMALIZE), &norm);
(void) nvlist_remove_all(createprops,
zfs_prop_to_name(ZFS_PROP_NORMALIZE));
(void) nvlist_lookup_uint64(createprops,
zfs_prop_to_name(ZFS_PROP_UTF8ONLY), &u8);
(void) nvlist_remove_all(createprops,
zfs_prop_to_name(ZFS_PROP_UTF8ONLY));
(void) nvlist_lookup_uint64(createprops,
zfs_prop_to_name(ZFS_PROP_CASE), &sense);
(void) nvlist_remove_all(createprops,
zfs_prop_to_name(ZFS_PROP_CASE));
}
/*
* If the zpl version requested is whacky or the file system
* or pool is version is too "young" to support normalization
* and the creator tried to set a value for one of the props,
* error out.
*/
if ((zplver < ZPL_VERSION_INITIAL || zplver > ZPL_VERSION) ||
(zplver >= ZPL_VERSION_FUID && !fuids_ok) ||
(zplver >= ZPL_VERSION_SA && !sa_ok) ||
(zplver < ZPL_VERSION_NORMALIZATION &&
(norm != ZFS_PROP_UNDEFINED || u8 != ZFS_PROP_UNDEFINED ||
sense != ZFS_PROP_UNDEFINED)))
return (SET_ERROR(ENOTSUP));
/*
* Put the version in the zplprops
*/
VERIFY(nvlist_add_uint64(zplprops,
zfs_prop_to_name(ZFS_PROP_VERSION), zplver) == 0);
if (norm == ZFS_PROP_UNDEFINED)
VERIFY(zfs_get_zplprop(os, ZFS_PROP_NORMALIZE, &norm) == 0);
VERIFY(nvlist_add_uint64(zplprops,
zfs_prop_to_name(ZFS_PROP_NORMALIZE), norm) == 0);
/*
* If we're normalizing, names must always be valid UTF-8 strings.
*/
if (norm)
u8 = 1;
if (u8 == ZFS_PROP_UNDEFINED)
VERIFY(zfs_get_zplprop(os, ZFS_PROP_UTF8ONLY, &u8) == 0);
VERIFY(nvlist_add_uint64(zplprops,
zfs_prop_to_name(ZFS_PROP_UTF8ONLY), u8) == 0);
if (sense == ZFS_PROP_UNDEFINED)
VERIFY(zfs_get_zplprop(os, ZFS_PROP_CASE, &sense) == 0);
VERIFY(nvlist_add_uint64(zplprops,
zfs_prop_to_name(ZFS_PROP_CASE), sense) == 0);
if (is_ci)
*is_ci = (sense == ZFS_CASE_INSENSITIVE);
return (0);
}
static int
zfs_fill_zplprops(const char *dataset, nvlist_t *createprops,
nvlist_t *zplprops, boolean_t *is_ci)
{
boolean_t fuids_ok, sa_ok;
uint64_t zplver = ZPL_VERSION;
objset_t *os = NULL;
char parentname[ZFS_MAX_DATASET_NAME_LEN];
- char *cp;
spa_t *spa;
uint64_t spa_vers;
int error;
- (void) strlcpy(parentname, dataset, sizeof (parentname));
- cp = strrchr(parentname, '/');
- ASSERT(cp != NULL);
- cp[0] = '\0';
+ zfs_get_parent(dataset, parentname, sizeof (parentname));
if ((error = spa_open(dataset, &spa, FTAG)) != 0)
return (error);
spa_vers = spa_version(spa);
spa_close(spa, FTAG);
zplver = zfs_zpl_version_map(spa_vers);
fuids_ok = (zplver >= ZPL_VERSION_FUID);
sa_ok = (zplver >= ZPL_VERSION_SA);
/*
* Open parent object set so we can inherit zplprop values.
*/
if ((error = dmu_objset_hold(parentname, FTAG, &os)) != 0)
return (error);
error = zfs_fill_zplprops_impl(os, zplver, fuids_ok, sa_ok, createprops,
zplprops, is_ci);
dmu_objset_rele(os, FTAG);
return (error);
}
static int
zfs_fill_zplprops_root(uint64_t spa_vers, nvlist_t *createprops,
nvlist_t *zplprops, boolean_t *is_ci)
{
boolean_t fuids_ok;
boolean_t sa_ok;
uint64_t zplver = ZPL_VERSION;
int error;
zplver = zfs_zpl_version_map(spa_vers);
fuids_ok = (zplver >= ZPL_VERSION_FUID);
sa_ok = (zplver >= ZPL_VERSION_SA);
error = zfs_fill_zplprops_impl(NULL, zplver, fuids_ok, sa_ok,
createprops, zplprops, is_ci);
return (error);
}
/*
* innvl: {
* "type" -> dmu_objset_type_t (int32)
* (optional) "props" -> { prop -> value }
* }
*
* outnvl: propname -> error code (int32)
*/
static int
zfs_ioc_create(const char *fsname, nvlist_t *innvl, nvlist_t *outnvl)
{
int error = 0;
zfs_creat_t zct = { 0 };
nvlist_t *nvprops = NULL;
void (*cbfunc)(objset_t *os, void *arg, cred_t *cr, dmu_tx_t *tx);
int32_t type32;
dmu_objset_type_t type;
boolean_t is_insensitive = B_FALSE;
if (nvlist_lookup_int32(innvl, "type", &type32) != 0)
return (SET_ERROR(EINVAL));
type = type32;
(void) nvlist_lookup_nvlist(innvl, "props", &nvprops);
switch (type) {
case DMU_OST_ZFS:
cbfunc = zfs_create_cb;
break;
case DMU_OST_ZVOL:
cbfunc = zvol_create_cb;
break;
default:
cbfunc = NULL;
break;
}
if (strchr(fsname, '@') ||
strchr(fsname, '%'))
return (SET_ERROR(EINVAL));
zct.zct_props = nvprops;
if (cbfunc == NULL)
return (SET_ERROR(EINVAL));
if (type == DMU_OST_ZVOL) {
uint64_t volsize, volblocksize;
if (nvprops == NULL)
return (SET_ERROR(EINVAL));
if (nvlist_lookup_uint64(nvprops,
zfs_prop_to_name(ZFS_PROP_VOLSIZE), &volsize) != 0)
return (SET_ERROR(EINVAL));
if ((error = nvlist_lookup_uint64(nvprops,
zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE),
&volblocksize)) != 0 && error != ENOENT)
return (SET_ERROR(EINVAL));
if (error != 0)
volblocksize = zfs_prop_default_numeric(
ZFS_PROP_VOLBLOCKSIZE);
if ((error = zvol_check_volblocksize(
volblocksize)) != 0 ||
(error = zvol_check_volsize(volsize,
volblocksize)) != 0)
return (error);
} else if (type == DMU_OST_ZFS) {
int error;
/*
* We have to have normalization and
* case-folding flags correct when we do the
* file system creation, so go figure them out
* now.
*/
VERIFY(nvlist_alloc(&zct.zct_zplprops,
NV_UNIQUE_NAME, KM_SLEEP) == 0);
error = zfs_fill_zplprops(fsname, nvprops,
zct.zct_zplprops, &is_insensitive);
if (error != 0) {
nvlist_free(zct.zct_zplprops);
return (error);
}
}
error = dmu_objset_create(fsname, type,
is_insensitive ? DS_FLAG_CI_DATASET : 0, cbfunc, &zct);
nvlist_free(zct.zct_zplprops);
/*
* It would be nice to do this atomically.
*/
if (error == 0) {
error = zfs_set_prop_nvlist(fsname, ZPROP_SRC_LOCAL,
nvprops, outnvl);
if (error != 0)
(void) dsl_destroy_head(fsname);
}
#ifdef __FreeBSD__
if (error == 0 && type == DMU_OST_ZVOL)
zvol_create_minors(fsname);
#endif
return (error);
}
/*
* innvl: {
* "origin" -> name of origin snapshot
* (optional) "props" -> { prop -> value }
* }
*
* outnvl: propname -> error code (int32)
*/
static int
zfs_ioc_clone(const char *fsname, nvlist_t *innvl, nvlist_t *outnvl)
{
int error = 0;
nvlist_t *nvprops = NULL;
char *origin_name;
if (nvlist_lookup_string(innvl, "origin", &origin_name) != 0)
return (SET_ERROR(EINVAL));
(void) nvlist_lookup_nvlist(innvl, "props", &nvprops);
if (strchr(fsname, '@') ||
strchr(fsname, '%'))
return (SET_ERROR(EINVAL));
if (dataset_namecheck(origin_name, NULL, NULL) != 0)
return (SET_ERROR(EINVAL));
error = dmu_objset_clone(fsname, origin_name);
if (error != 0)
return (error);
/*
* It would be nice to do this atomically.
*/
if (error == 0) {
error = zfs_set_prop_nvlist(fsname, ZPROP_SRC_LOCAL,
nvprops, outnvl);
if (error != 0)
(void) dsl_destroy_head(fsname);
}
#ifdef __FreeBSD__
if (error == 0)
zvol_create_minors(fsname);
#endif
return (error);
}
/* ARGSUSED */
static int
zfs_ioc_remap(const char *fsname, nvlist_t *innvl, nvlist_t *outnvl)
{
if (strchr(fsname, '@') ||
strchr(fsname, '%'))
return (SET_ERROR(EINVAL));
return (dmu_objset_remap_indirects(fsname));
}
/*
* innvl: {
* "snaps" -> { snapshot1, snapshot2 }
* (optional) "props" -> { prop -> value (string) }
* }
*
* outnvl: snapshot -> error code (int32)
*/
static int
zfs_ioc_snapshot(const char *poolname, nvlist_t *innvl, nvlist_t *outnvl)
{
nvlist_t *snaps;
nvlist_t *props = NULL;
int error, poollen;
nvpair_t *pair;
(void) nvlist_lookup_nvlist(innvl, "props", &props);
if (!nvlist_empty(props) &&
zfs_earlier_version(poolname, SPA_VERSION_SNAP_PROPS))
return (SET_ERROR(ENOTSUP));
if ((error = zfs_check_userprops(props)) != 0)
return (error);
if (nvlist_lookup_nvlist(innvl, "snaps", &snaps) != 0)
return (SET_ERROR(EINVAL));
poollen = strlen(poolname);
for (pair = nvlist_next_nvpair(snaps, NULL); pair != NULL;
pair = nvlist_next_nvpair(snaps, pair)) {
const char *name = nvpair_name(pair);
char *cp = strchr(name, '@');
/*
* The snap name must contain an @, and the part after it must
* contain only valid characters.
*/
if (cp == NULL ||
zfs_component_namecheck(cp + 1, NULL, NULL) != 0)
return (SET_ERROR(EINVAL));
/*
* The snap must be in the specified pool.
*/
if (strncmp(name, poolname, poollen) != 0 ||
(name[poollen] != '/' && name[poollen] != '@'))
return (SET_ERROR(EXDEV));
/*
* Check for permission to set the properties on the fs.
*/
if (!nvlist_empty(props)) {
*cp = '\0';
error = zfs_secpolicy_write_perms(name,
ZFS_DELEG_PERM_USERPROP, CRED());
*cp = '@';
if (error != 0)
return (error);
}
/* This must be the only snap of this fs. */
for (nvpair_t *pair2 = nvlist_next_nvpair(snaps, pair);
pair2 != NULL; pair2 = nvlist_next_nvpair(snaps, pair2)) {
if (strncmp(name, nvpair_name(pair2), cp - name + 1)
== 0) {
return (SET_ERROR(EXDEV));
}
}
}
error = dsl_dataset_snapshot(snaps, props, outnvl);
return (error);
}
/*
* innvl: "message" -> string
*/
/* ARGSUSED */
static int
zfs_ioc_log_history(const char *unused, nvlist_t *innvl, nvlist_t *outnvl)
{
char *message;
spa_t *spa;
int error;
char *poolname;
/*
* The poolname in the ioctl is not set, we get it from the TSD,
* which was set at the end of the last successful ioctl that allows
* logging. The secpolicy func already checked that it is set.
* Only one log ioctl is allowed after each successful ioctl, so
* we clear the TSD here.
*/
poolname = tsd_get(zfs_allow_log_key);
(void) tsd_set(zfs_allow_log_key, NULL);
error = spa_open(poolname, &spa, FTAG);
strfree(poolname);
if (error != 0)
return (error);
if (nvlist_lookup_string(innvl, "message", &message) != 0) {
spa_close(spa, FTAG);
return (SET_ERROR(EINVAL));
}
if (spa_version(spa) < SPA_VERSION_ZPOOL_HISTORY) {
spa_close(spa, FTAG);
return (SET_ERROR(ENOTSUP));
}
error = spa_history_log(spa, message);
spa_close(spa, FTAG);
return (error);
}
#ifdef __FreeBSD__
static int
zfs_ioc_nextboot(const char *unused, nvlist_t *innvl, nvlist_t *outnvl)
{
char name[MAXNAMELEN];
spa_t *spa;
vdev_t *vd;
char *command;
uint64_t pool_guid;
uint64_t vdev_guid;
int error;
if (nvlist_lookup_uint64(innvl,
ZPOOL_CONFIG_POOL_GUID, &pool_guid) != 0)
return (EINVAL);
if (nvlist_lookup_uint64(innvl,
ZPOOL_CONFIG_GUID, &vdev_guid) != 0)
return (EINVAL);
if (nvlist_lookup_string(innvl,
"command", &command) != 0)
return (EINVAL);
mutex_enter(&spa_namespace_lock);
spa = spa_by_guid(pool_guid, vdev_guid);
if (spa != NULL)
strcpy(name, spa_name(spa));
mutex_exit(&spa_namespace_lock);
if (spa == NULL)
return (ENOENT);
if ((error = spa_open(name, &spa, FTAG)) != 0)
return (error);
spa_vdev_state_enter(spa, SCL_ALL);
vd = spa_lookup_by_guid(spa, vdev_guid, B_TRUE);
if (vd == NULL) {
(void) spa_vdev_state_exit(spa, NULL, ENXIO);
spa_close(spa, FTAG);
return (ENODEV);
}
error = vdev_label_write_pad2(vd, command, strlen(command));
(void) spa_vdev_state_exit(spa, NULL, 0);
txg_wait_synced(spa->spa_dsl_pool, 0);
spa_close(spa, FTAG);
return (error);
}
#endif
/*
* The dp_config_rwlock must not be held when calling this, because the
* unmount may need to write out data.
*
* This function is best-effort. Callers must deal gracefully if it
* remains mounted (or is remounted after this call).
*
* Returns 0 if the argument is not a snapshot, or it is not currently a
* filesystem, or we were able to unmount it. Returns error code otherwise.
*/
void
zfs_unmount_snap(const char *snapname)
{
vfs_t *vfsp = NULL;
zfsvfs_t *zfsvfs = NULL;
if (strchr(snapname, '@') == NULL)
return;
int err = getzfsvfs(snapname, &zfsvfs);
if (err != 0) {
ASSERT3P(zfsvfs, ==, NULL);
return;
}
vfsp = zfsvfs->z_vfs;
ASSERT(!dsl_pool_config_held(dmu_objset_pool(zfsvfs->z_os)));
#ifdef illumos
err = vn_vfswlock(vfsp->vfs_vnodecovered);
VFS_RELE(vfsp);
if (err != 0)
return;
#endif
/*
* Always force the unmount for snapshots.
*/
#ifdef illumos
(void) dounmount(vfsp, MS_FORCE, kcred);
#else
vfs_ref(vfsp);
vfs_unbusy(vfsp);
(void) dounmount(vfsp, MS_FORCE, curthread);
#endif
}
/* ARGSUSED */
static int
zfs_unmount_snap_cb(const char *snapname, void *arg)
{
zfs_unmount_snap(snapname);
return (0);
}
/*
* When a clone is destroyed, its origin may also need to be destroyed,
* in which case it must be unmounted. This routine will do that unmount
* if necessary.
*/
void
zfs_destroy_unmount_origin(const char *fsname)
{
int error;
objset_t *os;
dsl_dataset_t *ds;
error = dmu_objset_hold(fsname, FTAG, &os);
if (error != 0)
return;
ds = dmu_objset_ds(os);
if (dsl_dir_is_clone(ds->ds_dir) && DS_IS_DEFER_DESTROY(ds->ds_prev)) {
char originname[ZFS_MAX_DATASET_NAME_LEN];
dsl_dataset_name(ds->ds_prev, originname);
dmu_objset_rele(os, FTAG);
zfs_unmount_snap(originname);
} else {
dmu_objset_rele(os, FTAG);
}
}
/*
* innvl: {
* "snaps" -> { snapshot1, snapshot2 }
* (optional boolean) "defer"
* }
*
* outnvl: snapshot -> error code (int32)
*
*/
/* ARGSUSED */
static int
zfs_ioc_destroy_snaps(const char *poolname, nvlist_t *innvl, nvlist_t *outnvl)
{
int error, poollen;
nvlist_t *snaps;
nvpair_t *pair;
boolean_t defer;
if (nvlist_lookup_nvlist(innvl, "snaps", &snaps) != 0)
return (SET_ERROR(EINVAL));
defer = nvlist_exists(innvl, "defer");
poollen = strlen(poolname);
for (pair = nvlist_next_nvpair(snaps, NULL); pair != NULL;
pair = nvlist_next_nvpair(snaps, pair)) {
const char *name = nvpair_name(pair);
/*
* The snap must be in the specified pool to prevent the
* invalid removal of zvol minors below.
*/
if (strncmp(name, poolname, poollen) != 0 ||
(name[poollen] != '/' && name[poollen] != '@'))
return (SET_ERROR(EXDEV));
zfs_unmount_snap(nvpair_name(pair));
#if defined(__FreeBSD__)
zvol_remove_minors(name);
#endif
}
return (dsl_destroy_snapshots_nvl(snaps, defer, outnvl));
}
/*
* Create bookmarks. Bookmark names are of the form <fs>#<bmark>.
* All bookmarks must be in the same pool.
*
* innvl: {
* bookmark1 -> snapshot1, bookmark2 -> snapshot2
* }
*
* outnvl: bookmark -> error code (int32)
*
*/
/* ARGSUSED */
static int
zfs_ioc_bookmark(const char *poolname, nvlist_t *innvl, nvlist_t *outnvl)
{
for (nvpair_t *pair = nvlist_next_nvpair(innvl, NULL);
pair != NULL; pair = nvlist_next_nvpair(innvl, pair)) {
char *snap_name;
/*
* Verify the snapshot argument.
*/
if (nvpair_value_string(pair, &snap_name) != 0)
return (SET_ERROR(EINVAL));
/* Verify that the keys (bookmarks) are unique */
for (nvpair_t *pair2 = nvlist_next_nvpair(innvl, pair);
pair2 != NULL; pair2 = nvlist_next_nvpair(innvl, pair2)) {
if (strcmp(nvpair_name(pair), nvpair_name(pair2)) == 0)
return (SET_ERROR(EINVAL));
}
}
return (dsl_bookmark_create(innvl, outnvl));
}
/*
* innvl: {
* property 1, property 2, ...
* }
*
* outnvl: {
* bookmark name 1 -> { property 1, property 2, ... },
* bookmark name 2 -> { property 1, property 2, ... }
* }
*
*/
static int
zfs_ioc_get_bookmarks(const char *fsname, nvlist_t *innvl, nvlist_t *outnvl)
{
return (dsl_get_bookmarks(fsname, innvl, outnvl));
}
/*
* innvl: {
* bookmark name 1, bookmark name 2
* }
*
* outnvl: bookmark -> error code (int32)
*
*/
static int
zfs_ioc_destroy_bookmarks(const char *poolname, nvlist_t *innvl,
nvlist_t *outnvl)
{
int error, poollen;
poollen = strlen(poolname);
for (nvpair_t *pair = nvlist_next_nvpair(innvl, NULL);
pair != NULL; pair = nvlist_next_nvpair(innvl, pair)) {
const char *name = nvpair_name(pair);
const char *cp = strchr(name, '#');
/*
* The bookmark name must contain an #, and the part after it
* must contain only valid characters.
*/
if (cp == NULL ||
zfs_component_namecheck(cp + 1, NULL, NULL) != 0)
return (SET_ERROR(EINVAL));
/*
* The bookmark must be in the specified pool.
*/
if (strncmp(name, poolname, poollen) != 0 ||
(name[poollen] != '/' && name[poollen] != '#'))
return (SET_ERROR(EXDEV));
}
error = dsl_bookmark_destroy(innvl, outnvl);
return (error);
}
static int
zfs_ioc_channel_program(const char *poolname, nvlist_t *innvl,
nvlist_t *outnvl)
{
char *program;
uint64_t instrlimit, memlimit;
boolean_t sync_flag;
nvpair_t *nvarg = NULL;
if (0 != nvlist_lookup_string(innvl, ZCP_ARG_PROGRAM, &program)) {
return (EINVAL);
}
if (0 != nvlist_lookup_boolean_value(innvl, ZCP_ARG_SYNC, &sync_flag)) {
sync_flag = B_TRUE;
}
if (0 != nvlist_lookup_uint64(innvl, ZCP_ARG_INSTRLIMIT, &instrlimit)) {
instrlimit = ZCP_DEFAULT_INSTRLIMIT;
}
if (0 != nvlist_lookup_uint64(innvl, ZCP_ARG_MEMLIMIT, &memlimit)) {
memlimit = ZCP_DEFAULT_MEMLIMIT;
}
if (0 != nvlist_lookup_nvpair(innvl, ZCP_ARG_ARGLIST, &nvarg)) {
return (EINVAL);
}
if (instrlimit == 0 || instrlimit > zfs_lua_max_instrlimit)
return (EINVAL);
if (memlimit == 0 || memlimit > zfs_lua_max_memlimit)
return (EINVAL);
return (zcp_eval(poolname, program, sync_flag, instrlimit, memlimit,
nvarg, outnvl));
}
/*
* innvl: unused
* outnvl: empty
*/
/* ARGSUSED */
static int
zfs_ioc_pool_checkpoint(const char *poolname, nvlist_t *innvl, nvlist_t *outnvl)
{
return (spa_checkpoint(poolname));
}
/*
* innvl: unused
* outnvl: empty
*/
/* ARGSUSED */
static int
zfs_ioc_pool_discard_checkpoint(const char *poolname, nvlist_t *innvl,
nvlist_t *outnvl)
{
return (spa_checkpoint_discard(poolname));
}
/*
* inputs:
* zc_name name of dataset to destroy
* zc_defer_destroy mark for deferred destroy
*
* outputs: none
*/
static int
zfs_ioc_destroy(zfs_cmd_t *zc)
{
objset_t *os;
dmu_objset_type_t ost;
int err;
err = dmu_objset_hold(zc->zc_name, FTAG, &os);
if (err != 0)
return (err);
ost = dmu_objset_type(os);
dmu_objset_rele(os, FTAG);
if (ost == DMU_OST_ZFS)
zfs_unmount_snap(zc->zc_name);
if (strchr(zc->zc_name, '@'))
err = dsl_destroy_snapshot(zc->zc_name, zc->zc_defer_destroy);
else
err = dsl_destroy_head(zc->zc_name);
if (ost == DMU_OST_ZVOL && err == 0)
#ifdef __FreeBSD__
zvol_remove_minors(zc->zc_name);
#else
(void) zvol_remove_minor(zc->zc_name);
#endif
return (err);
}
/*
* innvl: {
* vdevs: {
* guid 1, guid 2, ...
* },
* func: POOL_INITIALIZE_{CANCEL|DO|SUSPEND}
* }
*
* outnvl: {
* [func: EINVAL (if provided command type didn't make sense)],
* [vdevs: {
* guid1: errno, (see function body for possible errnos)
* ...
* }]
* }
*
*/
static int
zfs_ioc_pool_initialize(const char *poolname, nvlist_t *innvl, nvlist_t *outnvl)
{
spa_t *spa;
int error;
error = spa_open(poolname, &spa, FTAG);
if (error != 0)
return (error);
uint64_t cmd_type;
if (nvlist_lookup_uint64(innvl, ZPOOL_INITIALIZE_COMMAND,
&cmd_type) != 0) {
spa_close(spa, FTAG);
return (SET_ERROR(EINVAL));
}
if (!(cmd_type == POOL_INITIALIZE_CANCEL ||
cmd_type == POOL_INITIALIZE_DO ||
cmd_type == POOL_INITIALIZE_SUSPEND)) {
spa_close(spa, FTAG);
return (SET_ERROR(EINVAL));
}
nvlist_t *vdev_guids;
if (nvlist_lookup_nvlist(innvl, ZPOOL_INITIALIZE_VDEVS,
&vdev_guids) != 0) {
spa_close(spa, FTAG);
return (SET_ERROR(EINVAL));
}
nvlist_t *vdev_errlist = fnvlist_alloc();
int total_errors = 0;
for (nvpair_t *pair = nvlist_next_nvpair(vdev_guids, NULL);
pair != NULL; pair = nvlist_next_nvpair(vdev_guids, pair)) {
uint64_t vdev_guid = fnvpair_value_uint64(pair);
error = spa_vdev_initialize(spa, vdev_guid, cmd_type);
if (error != 0) {
char guid_as_str[MAXNAMELEN];
(void) snprintf(guid_as_str, sizeof (guid_as_str),
"%llu", (unsigned long long)vdev_guid);
fnvlist_add_int64(vdev_errlist, guid_as_str, error);
total_errors++;
}
}
if (fnvlist_size(vdev_errlist) > 0) {
fnvlist_add_nvlist(outnvl, ZPOOL_INITIALIZE_VDEVS,
vdev_errlist);
}
fnvlist_free(vdev_errlist);
spa_close(spa, FTAG);
return (total_errors > 0 ? EINVAL : 0);
}
/*
* fsname is name of dataset to rollback (to most recent snapshot)
*
* innvl may contain name of expected target snapshot
*
* outnvl: "target" -> name of most recent snapshot
* }
*/
/* ARGSUSED */
static int
zfs_ioc_rollback(const char *fsname, nvlist_t *innvl, nvlist_t *outnvl)
{
zfsvfs_t *zfsvfs;
char *target = NULL;
int error;
(void) nvlist_lookup_string(innvl, "target", &target);
if (target != NULL) {
const char *cp = strchr(target, '@');
/*
* The snap name must contain an @, and the part after it must
* contain only valid characters.
*/
if (cp == NULL ||
zfs_component_namecheck(cp + 1, NULL, NULL) != 0)
return (SET_ERROR(EINVAL));
}
if (getzfsvfs(fsname, &zfsvfs) == 0) {
dsl_dataset_t *ds;
ds = dmu_objset_ds(zfsvfs->z_os);
error = zfs_suspend_fs(zfsvfs);
if (error == 0) {
int resume_err;
error = dsl_dataset_rollback(fsname, target, zfsvfs,
outnvl);
resume_err = zfs_resume_fs(zfsvfs, ds);
error = error ? error : resume_err;
}
#ifdef illumos
VFS_RELE(zfsvfs->z_vfs);
#else
vfs_unbusy(zfsvfs->z_vfs);
#endif
} else {
error = dsl_dataset_rollback(fsname, target, NULL, outnvl);
}
return (error);
}
static int
recursive_unmount(const char *fsname, void *arg)
{
const char *snapname = arg;
char fullname[ZFS_MAX_DATASET_NAME_LEN];
(void) snprintf(fullname, sizeof (fullname), "%s@%s", fsname, snapname);
zfs_unmount_snap(fullname);
return (0);
}
/*
* inputs:
* zc_name old name of dataset or bookmark
* zc_value new name of dataset or bookmark
* zc_cookie recursive flag (only valid for snapshots)
*
* outputs: none
*/
static int
zfs_ioc_rename(zfs_cmd_t *zc)
{
objset_t *os;
dmu_objset_type_t ost;
boolean_t recursive = zc->zc_cookie & 1;
char *pos, *pos2;
boolean_t allow_mounted = B_TRUE;
int err;
#ifdef __FreeBSD__
allow_mounted = (zc->zc_cookie & 2) != 0;
#endif
zc->zc_name[sizeof (zc->zc_name) - 1] = '\0';
zc->zc_value[sizeof (zc->zc_value) - 1] = '\0';
pos = strchr(zc->zc_name, '#');
if (pos != NULL) {
/* Bookmarks must be in same fs. */
pos2 = strchr(zc->zc_value, '#');
if (pos2 == NULL)
return (SET_ERROR(EINVAL));
/* Recursive flag is not supported yet. */
if (recursive)
return (SET_ERROR(ENOTSUP));
*pos = '\0';
*pos2 = '\0';
if (strcmp(zc->zc_name, zc->zc_value) == 0) {
err = dsl_bookmark_rename(zc->zc_name,
pos + 1, pos2 + 1);
} else {
err = SET_ERROR(EXDEV);
}
*pos = '#';
*pos2 = '#';
return (err);
}
/* "zfs rename" from and to ...%recv datasets should both fail */
if (dataset_namecheck(zc->zc_name, NULL, NULL) != 0 ||
dataset_namecheck(zc->zc_value, NULL, NULL) != 0 ||
strchr(zc->zc_name, '%') || strchr(zc->zc_value, '%'))
return (SET_ERROR(EINVAL));
err = dmu_objset_hold(zc->zc_name, FTAG, &os);
if (err != 0)
return (err);
ost = dmu_objset_type(os);
dmu_objset_rele(os, FTAG);
pos = strchr(zc->zc_name, '@');
if (pos != NULL) {
/* Snapshots must be in same fs. */
pos2 = strchr(zc->zc_value, '@');
if (pos2 == NULL)
return (SET_ERROR(EINVAL));
*pos = '\0';
*pos2 = '\0';
if (strcmp(zc->zc_name, zc->zc_value) != 0) {
err = SET_ERROR(EXDEV);
} else {
if (ost == DMU_OST_ZFS && !allow_mounted) {
err = dmu_objset_find(zc->zc_name,
recursive_unmount, pos + 1,
recursive ? DS_FIND_CHILDREN : 0);
}
if (err == 0) {
err = dsl_dataset_rename_snapshot(zc->zc_name,
pos + 1, pos2 + 1, recursive);
}
}
*pos = '@';
*pos2 = '@';
return (err);
} else {
#ifdef illumos
if (ost == DMU_OST_ZVOL)
(void) zvol_remove_minor(zc->zc_name);
#endif
return (dsl_dir_rename(zc->zc_name, zc->zc_value));
}
}
static int
zfs_check_settable(const char *dsname, nvpair_t *pair, cred_t *cr)
{
const char *propname = nvpair_name(pair);
boolean_t issnap = (strchr(dsname, '@') != NULL);
zfs_prop_t prop = zfs_name_to_prop(propname);
uint64_t intval;
int err;
if (prop == ZPROP_INVAL) {
if (zfs_prop_user(propname)) {
if (err = zfs_secpolicy_write_perms(dsname,
ZFS_DELEG_PERM_USERPROP, cr))
return (err);
return (0);
}
if (!issnap && zfs_prop_userquota(propname)) {
const char *perm = NULL;
const char *uq_prefix =
zfs_userquota_prop_prefixes[ZFS_PROP_USERQUOTA];
const char *gq_prefix =
zfs_userquota_prop_prefixes[ZFS_PROP_GROUPQUOTA];
if (strncmp(propname, uq_prefix,
strlen(uq_prefix)) == 0) {
perm = ZFS_DELEG_PERM_USERQUOTA;
} else if (strncmp(propname, gq_prefix,
strlen(gq_prefix)) == 0) {
perm = ZFS_DELEG_PERM_GROUPQUOTA;
} else {
/* USERUSED and GROUPUSED are read-only */
return (SET_ERROR(EINVAL));
}
if (err = zfs_secpolicy_write_perms(dsname, perm, cr))
return (err);
return (0);
}
return (SET_ERROR(EINVAL));
}
if (issnap)
return (SET_ERROR(EINVAL));
if (nvpair_type(pair) == DATA_TYPE_NVLIST) {
/*
* dsl_prop_get_all_impl() returns properties in this
* format.
*/
nvlist_t *attrs;
VERIFY(nvpair_value_nvlist(pair, &attrs) == 0);
VERIFY(nvlist_lookup_nvpair(attrs, ZPROP_VALUE,
&pair) == 0);
}
/*
* Check that this value is valid for this pool version
*/
switch (prop) {
case ZFS_PROP_COMPRESSION:
/*
* If the user specified gzip compression, make sure
* the SPA supports it. We ignore any errors here since
* we'll catch them later.
*/
if (nvpair_value_uint64(pair, &intval) == 0) {
if (intval >= ZIO_COMPRESS_GZIP_1 &&
intval <= ZIO_COMPRESS_GZIP_9 &&
zfs_earlier_version(dsname,
SPA_VERSION_GZIP_COMPRESSION)) {
return (SET_ERROR(ENOTSUP));
}
if (intval == ZIO_COMPRESS_ZLE &&
zfs_earlier_version(dsname,
SPA_VERSION_ZLE_COMPRESSION))
return (SET_ERROR(ENOTSUP));
if (intval == ZIO_COMPRESS_LZ4) {
spa_t *spa;
if ((err = spa_open(dsname, &spa, FTAG)) != 0)
return (err);
if (!spa_feature_is_enabled(spa,
SPA_FEATURE_LZ4_COMPRESS)) {
spa_close(spa, FTAG);
return (SET_ERROR(ENOTSUP));
}
spa_close(spa, FTAG);
}
/*
* If this is a bootable dataset then
* verify that the compression algorithm
* is supported for booting. We must return
* something other than ENOTSUP since it
* implies a downrev pool version.
*/
if (zfs_is_bootfs(dsname) &&
!BOOTFS_COMPRESS_VALID(intval)) {
return (SET_ERROR(ERANGE));
}
}
break;
case ZFS_PROP_COPIES:
if (zfs_earlier_version(dsname, SPA_VERSION_DITTO_BLOCKS))
return (SET_ERROR(ENOTSUP));
break;
case ZFS_PROP_RECORDSIZE:
/* Record sizes above 128k need the feature to be enabled */
if (nvpair_value_uint64(pair, &intval) == 0 &&
intval > SPA_OLD_MAXBLOCKSIZE) {
spa_t *spa;
/*
* We don't allow setting the property above 1MB,
* unless the tunable has been changed.
*/
if (intval > zfs_max_recordsize ||
intval > SPA_MAXBLOCKSIZE)
return (SET_ERROR(ERANGE));
if ((err = spa_open(dsname, &spa, FTAG)) != 0)
return (err);
if (!spa_feature_is_enabled(spa,
SPA_FEATURE_LARGE_BLOCKS)) {
spa_close(spa, FTAG);
return (SET_ERROR(ENOTSUP));
}
spa_close(spa, FTAG);
}
break;
case ZFS_PROP_DNODESIZE:
/* Dnode sizes above 512 need the feature to be enabled */
if (nvpair_value_uint64(pair, &intval) == 0 &&
intval != ZFS_DNSIZE_LEGACY) {
spa_t *spa;
if ((err = spa_open(dsname, &spa, FTAG)) != 0)
return (err);
if (!spa_feature_is_enabled(spa,
SPA_FEATURE_LARGE_DNODE)) {
spa_close(spa, FTAG);
return (SET_ERROR(ENOTSUP));
}
spa_close(spa, FTAG);
}
break;
case ZFS_PROP_SPECIAL_SMALL_BLOCKS:
/*
* This property could require the allocation classes
* feature to be active for setting, however we allow
* it so that tests of settable properties succeed.
* The CLI will issue a warning in this case.
*/
break;
case ZFS_PROP_SHARESMB:
if (zpl_earlier_version(dsname, ZPL_VERSION_FUID))
return (SET_ERROR(ENOTSUP));
break;
case ZFS_PROP_ACLINHERIT:
if (nvpair_type(pair) == DATA_TYPE_UINT64 &&
nvpair_value_uint64(pair, &intval) == 0) {
if (intval == ZFS_ACL_PASSTHROUGH_X &&
zfs_earlier_version(dsname,
SPA_VERSION_PASSTHROUGH_X))
return (SET_ERROR(ENOTSUP));
}
break;
case ZFS_PROP_CHECKSUM:
case ZFS_PROP_DEDUP:
{
spa_feature_t feature;
spa_t *spa;
/* dedup feature version checks */
if (prop == ZFS_PROP_DEDUP &&
zfs_earlier_version(dsname, SPA_VERSION_DEDUP))
return (SET_ERROR(ENOTSUP));
if (nvpair_value_uint64(pair, &intval) != 0)
return (SET_ERROR(EINVAL));
/* check prop value is enabled in features */
feature = zio_checksum_to_feature(intval & ZIO_CHECKSUM_MASK);
if (feature == SPA_FEATURE_NONE)
break;
if ((err = spa_open(dsname, &spa, FTAG)) != 0)
return (err);
/*
* Salted checksums are not supported on root pools.
*/
if (spa_bootfs(spa) != 0 &&
intval < ZIO_CHECKSUM_FUNCTIONS &&
(zio_checksum_table[intval].ci_flags &
ZCHECKSUM_FLAG_SALTED)) {
spa_close(spa, FTAG);
return (SET_ERROR(ERANGE));
}
if (!spa_feature_is_enabled(spa, feature)) {
spa_close(spa, FTAG);
return (SET_ERROR(ENOTSUP));
}
spa_close(spa, FTAG);
break;
}
}
return (zfs_secpolicy_setprop(dsname, prop, pair, CRED()));
}
/*
* Checks for a race condition to make sure we don't increment a feature flag
* multiple times.
*/
static int
zfs_prop_activate_feature_check(void *arg, dmu_tx_t *tx)
{
spa_t *spa = dmu_tx_pool(tx)->dp_spa;
spa_feature_t *featurep = arg;
if (!spa_feature_is_active(spa, *featurep))
return (0);
else
return (SET_ERROR(EBUSY));
}
/*
* The callback invoked on feature activation in the sync task caused by
* zfs_prop_activate_feature.
*/
static void
zfs_prop_activate_feature_sync(void *arg, dmu_tx_t *tx)
{
spa_t *spa = dmu_tx_pool(tx)->dp_spa;
spa_feature_t *featurep = arg;
spa_feature_incr(spa, *featurep, tx);
}
/*
* Activates a feature on a pool in response to a property setting. This
* creates a new sync task which modifies the pool to reflect the feature
* as being active.
*/
static int
zfs_prop_activate_feature(spa_t *spa, spa_feature_t feature)
{
int err;
/* EBUSY here indicates that the feature is already active */
err = dsl_sync_task(spa_name(spa),
zfs_prop_activate_feature_check, zfs_prop_activate_feature_sync,
&feature, 2, ZFS_SPACE_CHECK_RESERVED);
if (err != 0 && err != EBUSY)
return (err);
else
return (0);
}
/*
* Removes properties from the given props list that fail permission checks
* needed to clear them and to restore them in case of a receive error. For each
* property, make sure we have both set and inherit permissions.
*
* Returns the first error encountered if any permission checks fail. If the
* caller provides a non-NULL errlist, it also gives the complete list of names
* of all the properties that failed a permission check along with the
* corresponding error numbers. The caller is responsible for freeing the
* returned errlist.
*
* If every property checks out successfully, zero is returned and the list
* pointed at by errlist is NULL.
*/
static int
zfs_check_clearable(char *dataset, nvlist_t *props, nvlist_t **errlist)
{
zfs_cmd_t *zc;
nvpair_t *pair, *next_pair;
nvlist_t *errors;
int err, rv = 0;
if (props == NULL)
return (0);
VERIFY(nvlist_alloc(&errors, NV_UNIQUE_NAME, KM_SLEEP) == 0);
zc = kmem_alloc(sizeof (zfs_cmd_t), KM_SLEEP);
(void) strcpy(zc->zc_name, dataset);
pair = nvlist_next_nvpair(props, NULL);
while (pair != NULL) {
next_pair = nvlist_next_nvpair(props, pair);
(void) strcpy(zc->zc_value, nvpair_name(pair));
if ((err = zfs_check_settable(dataset, pair, CRED())) != 0 ||
(err = zfs_secpolicy_inherit_prop(zc, NULL, CRED())) != 0) {
VERIFY(nvlist_remove_nvpair(props, pair) == 0);
VERIFY(nvlist_add_int32(errors,
zc->zc_value, err) == 0);
}
pair = next_pair;
}
kmem_free(zc, sizeof (zfs_cmd_t));
if ((pair = nvlist_next_nvpair(errors, NULL)) == NULL) {
nvlist_free(errors);
errors = NULL;
} else {
VERIFY(nvpair_value_int32(pair, &rv) == 0);
}
if (errlist == NULL)
nvlist_free(errors);
else
*errlist = errors;
return (rv);
}
static boolean_t
propval_equals(nvpair_t *p1, nvpair_t *p2)
{
if (nvpair_type(p1) == DATA_TYPE_NVLIST) {
/* dsl_prop_get_all_impl() format */
nvlist_t *attrs;
VERIFY(nvpair_value_nvlist(p1, &attrs) == 0);
VERIFY(nvlist_lookup_nvpair(attrs, ZPROP_VALUE,
&p1) == 0);
}
if (nvpair_type(p2) == DATA_TYPE_NVLIST) {
nvlist_t *attrs;
VERIFY(nvpair_value_nvlist(p2, &attrs) == 0);
VERIFY(nvlist_lookup_nvpair(attrs, ZPROP_VALUE,
&p2) == 0);
}
if (nvpair_type(p1) != nvpair_type(p2))
return (B_FALSE);
if (nvpair_type(p1) == DATA_TYPE_STRING) {
char *valstr1, *valstr2;
VERIFY(nvpair_value_string(p1, (char **)&valstr1) == 0);
VERIFY(nvpair_value_string(p2, (char **)&valstr2) == 0);
return (strcmp(valstr1, valstr2) == 0);
} else {
uint64_t intval1, intval2;
VERIFY(nvpair_value_uint64(p1, &intval1) == 0);
VERIFY(nvpair_value_uint64(p2, &intval2) == 0);
return (intval1 == intval2);
}
}
/*
* Remove properties from props if they are not going to change (as determined
* by comparison with origprops). Remove them from origprops as well, since we
* do not need to clear or restore properties that won't change.
*/
static void
props_reduce(nvlist_t *props, nvlist_t *origprops)
{
nvpair_t *pair, *next_pair;
if (origprops == NULL)
return; /* all props need to be received */
pair = nvlist_next_nvpair(props, NULL);
while (pair != NULL) {
const char *propname = nvpair_name(pair);
nvpair_t *match;
next_pair = nvlist_next_nvpair(props, pair);
if ((nvlist_lookup_nvpair(origprops, propname,
&match) != 0) || !propval_equals(pair, match))
goto next; /* need to set received value */
/* don't clear the existing received value */
(void) nvlist_remove_nvpair(origprops, match);
/* don't bother receiving the property */
(void) nvlist_remove_nvpair(props, pair);
next:
pair = next_pair;
}
}
/*
* Extract properties that cannot be set PRIOR to the receipt of a dataset.
* For example, refquota cannot be set until after the receipt of a dataset,
* because in replication streams, an older/earlier snapshot may exceed the
* refquota. We want to receive the older/earlier snapshot, but setting
* refquota pre-receipt will set the dsl's ACTUAL quota, which will prevent
* the older/earlier snapshot from being received (with EDQUOT).
*
* The ZFS test "zfs_receive_011_pos" demonstrates such a scenario.
*
* libzfs will need to be judicious handling errors encountered by props
* extracted by this function.
*/
static nvlist_t *
extract_delay_props(nvlist_t *props)
{
nvlist_t *delayprops;
nvpair_t *nvp, *tmp;
static const zfs_prop_t delayable[] = { ZFS_PROP_REFQUOTA, 0 };
int i;
VERIFY(nvlist_alloc(&delayprops, NV_UNIQUE_NAME, KM_SLEEP) == 0);
for (nvp = nvlist_next_nvpair(props, NULL); nvp != NULL;
nvp = nvlist_next_nvpair(props, nvp)) {
/*
* strcmp() is safe because zfs_prop_to_name() always returns
* a bounded string.
*/
for (i = 0; delayable[i] != 0; i++) {
if (strcmp(zfs_prop_to_name(delayable[i]),
nvpair_name(nvp)) == 0) {
break;
}
}
if (delayable[i] != 0) {
tmp = nvlist_prev_nvpair(props, nvp);
VERIFY(nvlist_add_nvpair(delayprops, nvp) == 0);
VERIFY(nvlist_remove_nvpair(props, nvp) == 0);
nvp = tmp;
}
}
if (nvlist_empty(delayprops)) {
nvlist_free(delayprops);
delayprops = NULL;
}
return (delayprops);
}
#ifdef DEBUG
static boolean_t zfs_ioc_recv_inject_err;
#endif
/*
* inputs:
* zc_name name of containing filesystem
* zc_nvlist_src{_size} nvlist of properties to apply
* zc_value name of snapshot to create
* zc_string name of clone origin (if DRR_FLAG_CLONE)
* zc_cookie file descriptor to recv from
* zc_begin_record the BEGIN record of the stream (not byteswapped)
* zc_guid force flag
* zc_cleanup_fd cleanup-on-exit file descriptor
* zc_action_handle handle for this guid/ds mapping (or zero on first call)
* zc_resumable if data is incomplete assume sender will resume
*
* outputs:
* zc_cookie number of bytes read
* zc_nvlist_dst{_size} error for each unapplied received property
* zc_obj zprop_errflags_t
* zc_action_handle handle for this guid/ds mapping
*/
static int
zfs_ioc_recv(zfs_cmd_t *zc)
{
file_t *fp;
dmu_recv_cookie_t drc;
boolean_t force = (boolean_t)zc->zc_guid;
int fd;
int error = 0;
int props_error = 0;
nvlist_t *errors;
offset_t off;
nvlist_t *props = NULL; /* sent properties */
nvlist_t *origprops = NULL; /* existing properties */
nvlist_t *delayprops = NULL; /* sent properties applied post-receive */
char *origin = NULL;
char *tosnap;
char tofs[ZFS_MAX_DATASET_NAME_LEN];
boolean_t first_recvd_props = B_FALSE;
if (dataset_namecheck(zc->zc_value, NULL, NULL) != 0 ||
strchr(zc->zc_value, '@') == NULL ||
strchr(zc->zc_value, '%'))
return (SET_ERROR(EINVAL));
(void) strcpy(tofs, zc->zc_value);
tosnap = strchr(tofs, '@');
*tosnap++ = '\0';
if (zc->zc_nvlist_src != 0 &&
(error = get_nvlist(zc->zc_nvlist_src, zc->zc_nvlist_src_size,
zc->zc_iflags, &props)) != 0)
return (error);
fd = zc->zc_cookie;
#ifdef illumos
fp = getf(fd);
#else
fget_read(curthread, fd, &cap_pread_rights, &fp);
#endif
if (fp == NULL) {
nvlist_free(props);
return (SET_ERROR(EBADF));
}
errors = fnvlist_alloc();
if (zc->zc_string[0])
origin = zc->zc_string;
error = dmu_recv_begin(tofs, tosnap,
&zc->zc_begin_record, force, zc->zc_resumable, origin, &drc);
if (error != 0)
goto out;
/*
* Set properties before we receive the stream so that they are applied
* to the new data. Note that we must call dmu_recv_stream() if
* dmu_recv_begin() succeeds.
*/
if (props != NULL && !drc.drc_newfs) {
if (spa_version(dsl_dataset_get_spa(drc.drc_ds)) >=
SPA_VERSION_RECVD_PROPS &&
!dsl_prop_get_hasrecvd(tofs))
first_recvd_props = B_TRUE;
/*
* If new received properties are supplied, they are to
* completely replace the existing received properties, so stash
* away the existing ones.
*/
if (dsl_prop_get_received(tofs, &origprops) == 0) {
nvlist_t *errlist = NULL;
/*
* Don't bother writing a property if its value won't
* change (and avoid the unnecessary security checks).
*
* The first receive after SPA_VERSION_RECVD_PROPS is a
* special case where we blow away all local properties
* regardless.
*/
if (!first_recvd_props)
props_reduce(props, origprops);
if (zfs_check_clearable(tofs, origprops, &errlist) != 0)
(void) nvlist_merge(errors, errlist, 0);
nvlist_free(errlist);
if (clear_received_props(tofs, origprops,
first_recvd_props ? NULL : props) != 0)
zc->zc_obj |= ZPROP_ERR_NOCLEAR;
} else {
zc->zc_obj |= ZPROP_ERR_NOCLEAR;
}
}
if (props != NULL) {
props_error = dsl_prop_set_hasrecvd(tofs);
if (props_error == 0) {
delayprops = extract_delay_props(props);
(void) zfs_set_prop_nvlist(tofs, ZPROP_SRC_RECEIVED,
props, errors);
}
}
off = fp->f_offset;
error = dmu_recv_stream(&drc, fp, &off, zc->zc_cleanup_fd,
&zc->zc_action_handle);
if (error == 0) {
zfsvfs_t *zfsvfs = NULL;
if (getzfsvfs(tofs, &zfsvfs) == 0) {
/* online recv */
dsl_dataset_t *ds;
int end_err;
ds = dmu_objset_ds(zfsvfs->z_os);
error = zfs_suspend_fs(zfsvfs);
/*
* If the suspend fails, then the recv_end will
* likely also fail, and clean up after itself.
*/
end_err = dmu_recv_end(&drc, zfsvfs);
if (error == 0)
error = zfs_resume_fs(zfsvfs, ds);
error = error ? error : end_err;
#ifdef illumos
VFS_RELE(zfsvfs->z_vfs);
#else
vfs_unbusy(zfsvfs->z_vfs);
#endif
} else {
error = dmu_recv_end(&drc, NULL);
}
/* Set delayed properties now, after we're done receiving. */
if (delayprops != NULL && error == 0) {
(void) zfs_set_prop_nvlist(tofs, ZPROP_SRC_RECEIVED,
delayprops, errors);
}
}
if (delayprops != NULL) {
/*
* Merge delayed props back in with initial props, in case
* we're DEBUG and zfs_ioc_recv_inject_err is set (which means
* we have to make sure clear_received_props() includes
* the delayed properties).
*
* Since zfs_ioc_recv_inject_err is only in DEBUG kernels,
* using ASSERT() will be just like a VERIFY.
*/
ASSERT(nvlist_merge(props, delayprops, 0) == 0);
nvlist_free(delayprops);
}
/*
* Now that all props, initial and delayed, are set, report the prop
* errors to the caller.
*/
if (zc->zc_nvlist_dst_size != 0 &&
(nvlist_smush(errors, zc->zc_nvlist_dst_size) != 0 ||
put_nvlist(zc, errors) != 0)) {
/*
* Caller made zc->zc_nvlist_dst less than the minimum expected
* size or supplied an invalid address.
*/
props_error = SET_ERROR(EINVAL);
}
zc->zc_cookie = off - fp->f_offset;
if (off >= 0 && off <= MAXOFFSET_T)
fp->f_offset = off;
#ifdef DEBUG
if (zfs_ioc_recv_inject_err) {
zfs_ioc_recv_inject_err = B_FALSE;
error = 1;
}
#endif
#ifdef __FreeBSD__
if (error == 0)
zvol_create_minors(tofs);
#endif
/*
* On error, restore the original props.
*/
if (error != 0 && props != NULL && !drc.drc_newfs) {
if (clear_received_props(tofs, props, NULL) != 0) {
/*
* We failed to clear the received properties.
* Since we may have left a $recvd value on the
* system, we can't clear the $hasrecvd flag.
*/
zc->zc_obj |= ZPROP_ERR_NORESTORE;
} else if (first_recvd_props) {
dsl_prop_unset_hasrecvd(tofs);
}
if (origprops == NULL && !drc.drc_newfs) {
/* We failed to stash the original properties. */
zc->zc_obj |= ZPROP_ERR_NORESTORE;
}
/*
* dsl_props_set() will not convert RECEIVED to LOCAL on or
* after SPA_VERSION_RECVD_PROPS, so we need to specify LOCAL
* explictly if we're restoring local properties cleared in the
* first new-style receive.
*/
if (origprops != NULL &&
zfs_set_prop_nvlist(tofs, (first_recvd_props ?
ZPROP_SRC_LOCAL : ZPROP_SRC_RECEIVED),
origprops, NULL) != 0) {
/*
* We stashed the original properties but failed to
* restore them.
*/
zc->zc_obj |= ZPROP_ERR_NORESTORE;
}
}
out:
nvlist_free(props);
nvlist_free(origprops);
nvlist_free(errors);
releasef(fd);
if (error == 0)
error = props_error;
return (error);
}
/*
* inputs:
* zc_name name of snapshot to send
* zc_cookie file descriptor to send stream to
* zc_obj fromorigin flag (mutually exclusive with zc_fromobj)
* zc_sendobj objsetid of snapshot to send
* zc_fromobj objsetid of incremental fromsnap (may be zero)
* zc_guid if set, estimate size of stream only. zc_cookie is ignored.
* output size in zc_objset_type.
* zc_flags lzc_send_flags
*
* outputs:
* zc_objset_type estimated size, if zc_guid is set
*
* NOTE: This is no longer the preferred interface, any new functionality
* should be added to zfs_ioc_send_new() instead.
*/
static int
zfs_ioc_send(zfs_cmd_t *zc)
{
int error;
offset_t off;
boolean_t estimate = (zc->zc_guid != 0);
boolean_t embedok = (zc->zc_flags & 0x1);
boolean_t large_block_ok = (zc->zc_flags & 0x2);
boolean_t compressok = (zc->zc_flags & 0x4);
if (zc->zc_obj != 0) {
dsl_pool_t *dp;
dsl_dataset_t *tosnap;
error = dsl_pool_hold(zc->zc_name, FTAG, &dp);
if (error != 0)
return (error);
error = dsl_dataset_hold_obj(dp, zc->zc_sendobj, FTAG, &tosnap);
if (error != 0) {
dsl_pool_rele(dp, FTAG);
return (error);
}
if (dsl_dir_is_clone(tosnap->ds_dir))
zc->zc_fromobj =
dsl_dir_phys(tosnap->ds_dir)->dd_origin_obj;
dsl_dataset_rele(tosnap, FTAG);
dsl_pool_rele(dp, FTAG);
}
if (estimate) {
dsl_pool_t *dp;
dsl_dataset_t *tosnap;
dsl_dataset_t *fromsnap = NULL;
error = dsl_pool_hold(zc->zc_name, FTAG, &dp);
if (error != 0)
return (error);
error = dsl_dataset_hold_obj(dp, zc->zc_sendobj, FTAG, &tosnap);
if (error != 0) {
dsl_pool_rele(dp, FTAG);
return (error);
}
if (zc->zc_fromobj != 0) {
error = dsl_dataset_hold_obj(dp, zc->zc_fromobj,
FTAG, &fromsnap);
if (error != 0) {
dsl_dataset_rele(tosnap, FTAG);
dsl_pool_rele(dp, FTAG);
return (error);
}
}
error = dmu_send_estimate(tosnap, fromsnap, compressok,
&zc->zc_objset_type);
if (fromsnap != NULL)
dsl_dataset_rele(fromsnap, FTAG);
dsl_dataset_rele(tosnap, FTAG);
dsl_pool_rele(dp, FTAG);
} else {
file_t *fp;
#ifdef illumos
fp = getf(zc->zc_cookie);
#else
fget_write(curthread, zc->zc_cookie, &cap_write_rights, &fp);
#endif
if (fp == NULL)
return (SET_ERROR(EBADF));
off = fp->f_offset;
error = dmu_send_obj(zc->zc_name, zc->zc_sendobj,
zc->zc_fromobj, embedok, large_block_ok, compressok,
#ifdef illumos
zc->zc_cookie, fp->f_vnode, &off);
#else
zc->zc_cookie, fp, &off);
#endif
if (off >= 0 && off <= MAXOFFSET_T)
fp->f_offset = off;
releasef(zc->zc_cookie);
}
return (error);
}
/*
* inputs:
* zc_name name of snapshot on which to report progress
* zc_cookie file descriptor of send stream
*
* outputs:
* zc_cookie number of bytes written in send stream thus far
*/
static int
zfs_ioc_send_progress(zfs_cmd_t *zc)
{
dsl_pool_t *dp;
dsl_dataset_t *ds;
dmu_sendarg_t *dsp = NULL;
int error;
error = dsl_pool_hold(zc->zc_name, FTAG, &dp);
if (error != 0)
return (error);
error = dsl_dataset_hold(dp, zc->zc_name, FTAG, &ds);
if (error != 0) {
dsl_pool_rele(dp, FTAG);
return (error);
}
mutex_enter(&ds->ds_sendstream_lock);
/*
* Iterate over all the send streams currently active on this dataset.
* If there's one which matches the specified file descriptor _and_ the
* stream was started by the current process, return the progress of
* that stream.
*/
for (dsp = list_head(&ds->ds_sendstreams); dsp != NULL;
dsp = list_next(&ds->ds_sendstreams, dsp)) {
if (dsp->dsa_outfd == zc->zc_cookie &&
dsp->dsa_proc == curproc)
break;
}
if (dsp != NULL)
zc->zc_cookie = *(dsp->dsa_off);
else
error = SET_ERROR(ENOENT);
mutex_exit(&ds->ds_sendstream_lock);
dsl_dataset_rele(ds, FTAG);
dsl_pool_rele(dp, FTAG);
return (error);
}
static int
zfs_ioc_inject_fault(zfs_cmd_t *zc)
{
int id, error;
error = zio_inject_fault(zc->zc_name, (int)zc->zc_guid, &id,
&zc->zc_inject_record);
if (error == 0)
zc->zc_guid = (uint64_t)id;
return (error);
}
static int
zfs_ioc_clear_fault(zfs_cmd_t *zc)
{
return (zio_clear_fault((int)zc->zc_guid));
}
static int
zfs_ioc_inject_list_next(zfs_cmd_t *zc)
{
int id = (int)zc->zc_guid;
int error;
error = zio_inject_list_next(&id, zc->zc_name, sizeof (zc->zc_name),
&zc->zc_inject_record);
zc->zc_guid = id;
return (error);
}
static int
zfs_ioc_error_log(zfs_cmd_t *zc)
{
spa_t *spa;
int error;
size_t count = (size_t)zc->zc_nvlist_dst_size;
if ((error = spa_open(zc->zc_name, &spa, FTAG)) != 0)
return (error);
error = spa_get_errlog(spa, (void *)(uintptr_t)zc->zc_nvlist_dst,
&count);
if (error == 0)
zc->zc_nvlist_dst_size = count;
else
zc->zc_nvlist_dst_size = spa_get_errlog_size(spa);
spa_close(spa, FTAG);
return (error);
}
static int
zfs_ioc_clear(zfs_cmd_t *zc)
{
spa_t *spa;
vdev_t *vd;
int error;
/*
* On zpool clear we also fix up missing slogs
*/
mutex_enter(&spa_namespace_lock);
spa = spa_lookup(zc->zc_name);
if (spa == NULL) {
mutex_exit(&spa_namespace_lock);
return (SET_ERROR(EIO));
}
if (spa_get_log_state(spa) == SPA_LOG_MISSING) {
/* we need to let spa_open/spa_load clear the chains */
spa_set_log_state(spa, SPA_LOG_CLEAR);
}
spa->spa_last_open_failed = 0;
mutex_exit(&spa_namespace_lock);
if (zc->zc_cookie & ZPOOL_NO_REWIND) {
error = spa_open(zc->zc_name, &spa, FTAG);
} else {
nvlist_t *policy;
nvlist_t *config = NULL;
if (zc->zc_nvlist_src == 0)
return (SET_ERROR(EINVAL));
if ((error = get_nvlist(zc->zc_nvlist_src,
zc->zc_nvlist_src_size, zc->zc_iflags, &policy)) == 0) {
error = spa_open_rewind(zc->zc_name, &spa, FTAG,
policy, &config);
if (config != NULL) {
int err;
if ((err = put_nvlist(zc, config)) != 0)
error = err;
nvlist_free(config);
}
nvlist_free(policy);
}
}
if (error != 0)
return (error);
/*
* If multihost is enabled, resuming I/O is unsafe as another
* host may have imported the pool.
*/
if (spa_multihost(spa) && spa_suspended(spa))
return (SET_ERROR(EINVAL));
spa_vdev_state_enter(spa, SCL_NONE);
if (zc->zc_guid == 0) {
vd = NULL;
} else {
vd = spa_lookup_by_guid(spa, zc->zc_guid, B_TRUE);
if (vd == NULL) {
(void) spa_vdev_state_exit(spa, NULL, ENODEV);
spa_close(spa, FTAG);
return (SET_ERROR(ENODEV));
}
}
vdev_clear(spa, vd);
(void) spa_vdev_state_exit(spa, NULL, 0);
/*
* Resume any suspended I/Os.
*/
if (zio_resume(spa) != 0)
error = SET_ERROR(EIO);
spa_close(spa, FTAG);
return (error);
}
static int
zfs_ioc_pool_reopen(zfs_cmd_t *zc)
{
spa_t *spa;
int error;
error = spa_open(zc->zc_name, &spa, FTAG);
if (error != 0)
return (error);
spa_vdev_state_enter(spa, SCL_NONE);
/*
* If a resilver is already in progress then set the
* spa_scrub_reopen flag to B_TRUE so that we don't restart
* the scan as a side effect of the reopen. Otherwise, let
* vdev_open() decided if a resilver is required.
*/
spa->spa_scrub_reopen = dsl_scan_resilvering(spa->spa_dsl_pool);
vdev_reopen(spa->spa_root_vdev);
spa->spa_scrub_reopen = B_FALSE;
(void) spa_vdev_state_exit(spa, NULL, 0);
spa_close(spa, FTAG);
return (0);
}
/*
* inputs:
* zc_name name of filesystem
*
* outputs:
* zc_string name of conflicting snapshot, if there is one
*/
static int
zfs_ioc_promote(zfs_cmd_t *zc)
{
dsl_pool_t *dp;
dsl_dataset_t *ds, *ods;
char origin[ZFS_MAX_DATASET_NAME_LEN];
char *cp;
int error;
zc->zc_name[sizeof (zc->zc_name) - 1] = '\0';
if (dataset_namecheck(zc->zc_name, NULL, NULL) != 0 ||
strchr(zc->zc_name, '%'))
return (SET_ERROR(EINVAL));
error = dsl_pool_hold(zc->zc_name, FTAG, &dp);
if (error != 0)
return (error);
error = dsl_dataset_hold(dp, zc->zc_name, FTAG, &ds);
if (error != 0) {
dsl_pool_rele(dp, FTAG);
return (error);
}
if (!dsl_dir_is_clone(ds->ds_dir)) {
dsl_dataset_rele(ds, FTAG);
dsl_pool_rele(dp, FTAG);
return (SET_ERROR(EINVAL));
}
error = dsl_dataset_hold_obj(dp,
dsl_dir_phys(ds->ds_dir)->dd_origin_obj, FTAG, &ods);
if (error != 0) {
dsl_dataset_rele(ds, FTAG);
dsl_pool_rele(dp, FTAG);
return (error);
}
dsl_dataset_name(ods, origin);
dsl_dataset_rele(ods, FTAG);
dsl_dataset_rele(ds, FTAG);
dsl_pool_rele(dp, FTAG);
/*
* We don't need to unmount *all* the origin fs's snapshots, but
* it's easier.
*/
cp = strchr(origin, '@');
if (cp)
*cp = '\0';
(void) dmu_objset_find(origin,
zfs_unmount_snap_cb, NULL, DS_FIND_SNAPSHOTS);
return (dsl_dataset_promote(zc->zc_name, zc->zc_string));
}
/*
* Retrieve a single {user|group}{used|quota}@... property.
*
* inputs:
* zc_name name of filesystem
* zc_objset_type zfs_userquota_prop_t
* zc_value domain name (eg. "S-1-234-567-89")
* zc_guid RID/UID/GID
*
* outputs:
* zc_cookie property value
*/
static int
zfs_ioc_userspace_one(zfs_cmd_t *zc)
{
zfsvfs_t *zfsvfs;
int error;
if (zc->zc_objset_type >= ZFS_NUM_USERQUOTA_PROPS)
return (SET_ERROR(EINVAL));
error = zfsvfs_hold(zc->zc_name, FTAG, &zfsvfs, B_FALSE);
if (error != 0)
return (error);
error = zfs_userspace_one(zfsvfs,
zc->zc_objset_type, zc->zc_value, zc->zc_guid, &zc->zc_cookie);
zfsvfs_rele(zfsvfs, FTAG);
return (error);
}
/*
* inputs:
* zc_name name of filesystem
* zc_cookie zap cursor
* zc_objset_type zfs_userquota_prop_t
* zc_nvlist_dst[_size] buffer to fill (not really an nvlist)
*
* outputs:
* zc_nvlist_dst[_size] data buffer (array of zfs_useracct_t)
* zc_cookie zap cursor
*/
static int
zfs_ioc_userspace_many(zfs_cmd_t *zc)
{
zfsvfs_t *zfsvfs;
int bufsize = zc->zc_nvlist_dst_size;
if (bufsize <= 0)
return (SET_ERROR(ENOMEM));
int error = zfsvfs_hold(zc->zc_name, FTAG, &zfsvfs, B_FALSE);
if (error != 0)
return (error);
void *buf = kmem_alloc(bufsize, KM_SLEEP);
error = zfs_userspace_many(zfsvfs, zc->zc_objset_type, &zc->zc_cookie,
buf, &zc->zc_nvlist_dst_size);
if (error == 0) {
error = ddi_copyout(buf,
(void *)(uintptr_t)zc->zc_nvlist_dst,
zc->zc_nvlist_dst_size, zc->zc_iflags);
}
kmem_free(buf, bufsize);
zfsvfs_rele(zfsvfs, FTAG);
return (error);
}
/*
* inputs:
* zc_name name of filesystem
*
* outputs:
* none
*/
static int
zfs_ioc_userspace_upgrade(zfs_cmd_t *zc)
{
objset_t *os;
int error = 0;
zfsvfs_t *zfsvfs;
if (getzfsvfs(zc->zc_name, &zfsvfs) == 0) {
if (!dmu_objset_userused_enabled(zfsvfs->z_os)) {
/*
* If userused is not enabled, it may be because the
* objset needs to be closed & reopened (to grow the
* objset_phys_t). Suspend/resume the fs will do that.
*/
dsl_dataset_t *ds, *newds;
ds = dmu_objset_ds(zfsvfs->z_os);
error = zfs_suspend_fs(zfsvfs);
if (error == 0) {
dmu_objset_refresh_ownership(ds, &newds,
zfsvfs);
error = zfs_resume_fs(zfsvfs, newds);
}
}
if (error == 0)
error = dmu_objset_userspace_upgrade(zfsvfs->z_os);
#ifdef illumos
VFS_RELE(zfsvfs->z_vfs);
#else
vfs_unbusy(zfsvfs->z_vfs);
#endif
} else {
/* XXX kind of reading contents without owning */
error = dmu_objset_hold(zc->zc_name, FTAG, &os);
if (error != 0)
return (error);
error = dmu_objset_userspace_upgrade(os);
dmu_objset_rele(os, FTAG);
}
return (error);
}
#ifdef illumos
/*
* We don't want to have a hard dependency
* against some special symbols in sharefs
* nfs, and smbsrv. Determine them if needed when
* the first file system is shared.
* Neither sharefs, nfs or smbsrv are unloadable modules.
*/
int (*znfsexport_fs)(void *arg);
int (*zshare_fs)(enum sharefs_sys_op, share_t *, uint32_t);
int (*zsmbexport_fs)(void *arg, boolean_t add_share);
int zfs_nfsshare_inited;
int zfs_smbshare_inited;
ddi_modhandle_t nfs_mod;
ddi_modhandle_t sharefs_mod;
ddi_modhandle_t smbsrv_mod;
#endif /* illumos */
kmutex_t zfs_share_lock;
#ifdef illumos
static int
zfs_init_sharefs()
{
int error;
ASSERT(MUTEX_HELD(&zfs_share_lock));
/* Both NFS and SMB shares also require sharetab support. */
if (sharefs_mod == NULL && ((sharefs_mod =
ddi_modopen("fs/sharefs",
KRTLD_MODE_FIRST, &error)) == NULL)) {
return (SET_ERROR(ENOSYS));
}
if (zshare_fs == NULL && ((zshare_fs =
(int (*)(enum sharefs_sys_op, share_t *, uint32_t))
ddi_modsym(sharefs_mod, "sharefs_impl", &error)) == NULL)) {
return (SET_ERROR(ENOSYS));
}
return (0);
}
#endif /* illumos */
static int
zfs_ioc_share(zfs_cmd_t *zc)
{
#ifdef illumos
int error;
int opcode;
switch (zc->zc_share.z_sharetype) {
case ZFS_SHARE_NFS:
case ZFS_UNSHARE_NFS:
if (zfs_nfsshare_inited == 0) {
mutex_enter(&zfs_share_lock);
if (nfs_mod == NULL && ((nfs_mod = ddi_modopen("fs/nfs",
KRTLD_MODE_FIRST, &error)) == NULL)) {
mutex_exit(&zfs_share_lock);
return (SET_ERROR(ENOSYS));
}
if (znfsexport_fs == NULL &&
((znfsexport_fs = (int (*)(void *))
ddi_modsym(nfs_mod,
"nfs_export", &error)) == NULL)) {
mutex_exit(&zfs_share_lock);
return (SET_ERROR(ENOSYS));
}
error = zfs_init_sharefs();
if (error != 0) {
mutex_exit(&zfs_share_lock);
return (SET_ERROR(ENOSYS));
}
zfs_nfsshare_inited = 1;
mutex_exit(&zfs_share_lock);
}
break;
case ZFS_SHARE_SMB:
case ZFS_UNSHARE_SMB:
if (zfs_smbshare_inited == 0) {
mutex_enter(&zfs_share_lock);
if (smbsrv_mod == NULL && ((smbsrv_mod =
ddi_modopen("drv/smbsrv",
KRTLD_MODE_FIRST, &error)) == NULL)) {
mutex_exit(&zfs_share_lock);
return (SET_ERROR(ENOSYS));
}
if (zsmbexport_fs == NULL && ((zsmbexport_fs =
(int (*)(void *, boolean_t))ddi_modsym(smbsrv_mod,
"smb_server_share", &error)) == NULL)) {
mutex_exit(&zfs_share_lock);
return (SET_ERROR(ENOSYS));
}
error = zfs_init_sharefs();
if (error != 0) {
mutex_exit(&zfs_share_lock);
return (SET_ERROR(ENOSYS));
}
zfs_smbshare_inited = 1;
mutex_exit(&zfs_share_lock);
}
break;
default:
return (SET_ERROR(EINVAL));
}
switch (zc->zc_share.z_sharetype) {
case ZFS_SHARE_NFS:
case ZFS_UNSHARE_NFS:
if (error =
znfsexport_fs((void *)
(uintptr_t)zc->zc_share.z_exportdata))
return (error);
break;
case ZFS_SHARE_SMB:
case ZFS_UNSHARE_SMB:
if (error = zsmbexport_fs((void *)
(uintptr_t)zc->zc_share.z_exportdata,
zc->zc_share.z_sharetype == ZFS_SHARE_SMB ?
B_TRUE: B_FALSE)) {
return (error);
}
break;
}
opcode = (zc->zc_share.z_sharetype == ZFS_SHARE_NFS ||
zc->zc_share.z_sharetype == ZFS_SHARE_SMB) ?
SHAREFS_ADD : SHAREFS_REMOVE;
/*
* Add or remove share from sharetab
*/
error = zshare_fs(opcode,
(void *)(uintptr_t)zc->zc_share.z_sharedata,
zc->zc_share.z_sharemax);
return (error);
#else /* !illumos */
return (ENOSYS);
#endif /* illumos */
}
ace_t full_access[] = {
{(uid_t)-1, ACE_ALL_PERMS, ACE_EVERYONE, 0}
};
/*
* inputs:
* zc_name name of containing filesystem
* zc_obj object # beyond which we want next in-use object #
*
* outputs:
* zc_obj next in-use object #
*/
static int
zfs_ioc_next_obj(zfs_cmd_t *zc)
{
objset_t *os = NULL;
int error;
error = dmu_objset_hold(zc->zc_name, FTAG, &os);
if (error != 0)
return (error);
error = dmu_object_next(os, &zc->zc_obj, B_FALSE, 0);
dmu_objset_rele(os, FTAG);
return (error);
}
/*
* inputs:
* zc_name name of filesystem
* zc_value prefix name for snapshot
* zc_cleanup_fd cleanup-on-exit file descriptor for calling process
*
* outputs:
* zc_value short name of new snapshot
*/
static int
zfs_ioc_tmp_snapshot(zfs_cmd_t *zc)
{
char *snap_name;
char *hold_name;
int error;
minor_t minor;
error = zfs_onexit_fd_hold(zc->zc_cleanup_fd, &minor);
if (error != 0)
return (error);
snap_name = kmem_asprintf("%s-%016llx", zc->zc_value,
(u_longlong_t)ddi_get_lbolt64());
hold_name = kmem_asprintf("%%%s", zc->zc_value);
error = dsl_dataset_snapshot_tmp(zc->zc_name, snap_name, minor,
hold_name);
if (error == 0)
(void) strcpy(zc->zc_value, snap_name);
strfree(snap_name);
strfree(hold_name);
zfs_onexit_fd_rele(zc->zc_cleanup_fd);
return (error);
}
/*
* inputs:
* zc_name name of "to" snapshot
* zc_value name of "from" snapshot
* zc_cookie file descriptor to write diff data on
*
* outputs:
* dmu_diff_record_t's to the file descriptor
*/
static int
zfs_ioc_diff(zfs_cmd_t *zc)
{
file_t *fp;
offset_t off;
int error;
#ifdef illumos
fp = getf(zc->zc_cookie);
#else
fget_write(curthread, zc->zc_cookie, &cap_write_rights, &fp);
#endif
if (fp == NULL)
return (SET_ERROR(EBADF));
off = fp->f_offset;
#ifdef illumos
error = dmu_diff(zc->zc_name, zc->zc_value, fp->f_vnode, &off);
#else
error = dmu_diff(zc->zc_name, zc->zc_value, fp, &off);
#endif
if (off >= 0 && off <= MAXOFFSET_T)
fp->f_offset = off;
releasef(zc->zc_cookie);
return (error);
}
#ifdef illumos
/*
* Remove all ACL files in shares dir
*/
static int
zfs_smb_acl_purge(znode_t *dzp)
{
zap_cursor_t zc;
zap_attribute_t zap;
zfsvfs_t *zfsvfs = dzp->z_zfsvfs;
int error;
for (zap_cursor_init(&zc, zfsvfs->z_os, dzp->z_id);
(error = zap_cursor_retrieve(&zc, &zap)) == 0;
zap_cursor_advance(&zc)) {
if ((error = VOP_REMOVE(ZTOV(dzp), zap.za_name, kcred,
NULL, 0)) != 0)
break;
}
zap_cursor_fini(&zc);
return (error);
}
#endif /* illumos */
static int
zfs_ioc_smb_acl(zfs_cmd_t *zc)
{
#ifdef illumos
vnode_t *vp;
znode_t *dzp;
vnode_t *resourcevp = NULL;
znode_t *sharedir;
zfsvfs_t *zfsvfs;
nvlist_t *nvlist;
char *src, *target;
vattr_t vattr;
vsecattr_t vsec;
int error = 0;
if ((error = lookupname(zc->zc_value, UIO_SYSSPACE,
NO_FOLLOW, NULL, &vp)) != 0)
return (error);
/* Now make sure mntpnt and dataset are ZFS */
if (strcmp(vp->v_vfsp->mnt_stat.f_fstypename, "zfs") != 0 ||
(strcmp((char *)refstr_value(vp->v_vfsp->vfs_resource),
zc->zc_name) != 0)) {
VN_RELE(vp);
return (SET_ERROR(EINVAL));
}
dzp = VTOZ(vp);
zfsvfs = dzp->z_zfsvfs;
ZFS_ENTER(zfsvfs);
/*
* Create share dir if its missing.
*/
mutex_enter(&zfsvfs->z_lock);
if (zfsvfs->z_shares_dir == 0) {
dmu_tx_t *tx;
tx = dmu_tx_create(zfsvfs->z_os);
dmu_tx_hold_zap(tx, MASTER_NODE_OBJ, TRUE,
ZFS_SHARES_DIR);
dmu_tx_hold_zap(tx, DMU_NEW_OBJECT, FALSE, NULL);
error = dmu_tx_assign(tx, TXG_WAIT);
if (error != 0) {
dmu_tx_abort(tx);
} else {
error = zfs_create_share_dir(zfsvfs, tx);
dmu_tx_commit(tx);
}
if (error != 0) {
mutex_exit(&zfsvfs->z_lock);
VN_RELE(vp);
ZFS_EXIT(zfsvfs);
return (error);
}
}
mutex_exit(&zfsvfs->z_lock);
ASSERT(zfsvfs->z_shares_dir);
if ((error = zfs_zget(zfsvfs, zfsvfs->z_shares_dir, &sharedir)) != 0) {
VN_RELE(vp);
ZFS_EXIT(zfsvfs);
return (error);
}
switch (zc->zc_cookie) {
case ZFS_SMB_ACL_ADD:
vattr.va_mask = AT_MODE|AT_UID|AT_GID|AT_TYPE;
vattr.va_type = VREG;
vattr.va_mode = S_IFREG|0777;
vattr.va_uid = 0;
vattr.va_gid = 0;
vsec.vsa_mask = VSA_ACE;
vsec.vsa_aclentp = &full_access;
vsec.vsa_aclentsz = sizeof (full_access);
vsec.vsa_aclcnt = 1;
error = VOP_CREATE(ZTOV(sharedir), zc->zc_string,
&vattr, EXCL, 0, &resourcevp, kcred, 0, NULL, &vsec);
if (resourcevp)
VN_RELE(resourcevp);
break;
case ZFS_SMB_ACL_REMOVE:
error = VOP_REMOVE(ZTOV(sharedir), zc->zc_string, kcred,
NULL, 0);
break;
case ZFS_SMB_ACL_RENAME:
if ((error = get_nvlist(zc->zc_nvlist_src,
zc->zc_nvlist_src_size, zc->zc_iflags, &nvlist)) != 0) {
VN_RELE(vp);
VN_RELE(ZTOV(sharedir));
ZFS_EXIT(zfsvfs);
return (error);
}
if (nvlist_lookup_string(nvlist, ZFS_SMB_ACL_SRC, &src) ||
nvlist_lookup_string(nvlist, ZFS_SMB_ACL_TARGET,
&target)) {
VN_RELE(vp);
VN_RELE(ZTOV(sharedir));
ZFS_EXIT(zfsvfs);
nvlist_free(nvlist);
return (error);
}
error = VOP_RENAME(ZTOV(sharedir), src, ZTOV(sharedir), target,
kcred, NULL, 0);
nvlist_free(nvlist);
break;
case ZFS_SMB_ACL_PURGE:
error = zfs_smb_acl_purge(sharedir);
break;
default:
error = SET_ERROR(EINVAL);
break;
}
VN_RELE(vp);
VN_RELE(ZTOV(sharedir));
ZFS_EXIT(zfsvfs);
return (error);
#else /* !illumos */
return (EOPNOTSUPP);
#endif /* illumos */
}
/*
* innvl: {
* "holds" -> { snapname -> holdname (string), ... }
* (optional) "cleanup_fd" -> fd (int32)
* }
*
* outnvl: {
* snapname -> error value (int32)
* ...
* }
*/
/* ARGSUSED */
static int
zfs_ioc_hold(const char *pool, nvlist_t *args, nvlist_t *errlist)
{
nvpair_t *pair;
nvlist_t *holds;
int cleanup_fd = -1;
int error;
minor_t minor = 0;
error = nvlist_lookup_nvlist(args, "holds", &holds);
if (error != 0)
return (SET_ERROR(EINVAL));
/* make sure the user didn't pass us any invalid (empty) tags */
for (pair = nvlist_next_nvpair(holds, NULL); pair != NULL;
pair = nvlist_next_nvpair(holds, pair)) {
char *htag;
error = nvpair_value_string(pair, &htag);
if (error != 0)
return (SET_ERROR(error));
if (strlen(htag) == 0)
return (SET_ERROR(EINVAL));
}
if (nvlist_lookup_int32(args, "cleanup_fd", &cleanup_fd) == 0) {
error = zfs_onexit_fd_hold(cleanup_fd, &minor);
if (error != 0)
return (error);
}
error = dsl_dataset_user_hold(holds, minor, errlist);
if (minor != 0)
zfs_onexit_fd_rele(cleanup_fd);
return (error);
}
/*
* innvl is not used.
*
* outnvl: {
* holdname -> time added (uint64 seconds since epoch)
* ...
* }
*/
/* ARGSUSED */
static int
zfs_ioc_get_holds(const char *snapname, nvlist_t *args, nvlist_t *outnvl)
{
ASSERT3P(args, ==, NULL);
return (dsl_dataset_get_holds(snapname, outnvl));
}
/*
* innvl: {
* snapname -> { holdname, ... }
* ...
* }
*
* outnvl: {
* snapname -> error value (int32)
* ...
* }
*/
/* ARGSUSED */
static int
zfs_ioc_release(const char *pool, nvlist_t *holds, nvlist_t *errlist)
{
return (dsl_dataset_user_release(holds, errlist));
}
/*
* inputs:
* zc_name name of new filesystem or snapshot
* zc_value full name of old snapshot
*
* outputs:
* zc_cookie space in bytes
* zc_objset_type compressed space in bytes
* zc_perm_action uncompressed space in bytes
*/
static int
zfs_ioc_space_written(zfs_cmd_t *zc)
{
int error;
dsl_pool_t *dp;
dsl_dataset_t *new, *old;
error = dsl_pool_hold(zc->zc_name, FTAG, &dp);
if (error != 0)
return (error);
error = dsl_dataset_hold(dp, zc->zc_name, FTAG, &new);
if (error != 0) {
dsl_pool_rele(dp, FTAG);
return (error);
}
error = dsl_dataset_hold(dp, zc->zc_value, FTAG, &old);
if (error != 0) {
dsl_dataset_rele(new, FTAG);
dsl_pool_rele(dp, FTAG);
return (error);
}
error = dsl_dataset_space_written(old, new, &zc->zc_cookie,
&zc->zc_objset_type, &zc->zc_perm_action);
dsl_dataset_rele(old, FTAG);
dsl_dataset_rele(new, FTAG);
dsl_pool_rele(dp, FTAG);
return (error);
}
/*
* innvl: {
* "firstsnap" -> snapshot name
* }
*
* outnvl: {
* "used" -> space in bytes
* "compressed" -> compressed space in bytes
* "uncompressed" -> uncompressed space in bytes
* }
*/
static int
zfs_ioc_space_snaps(const char *lastsnap, nvlist_t *innvl, nvlist_t *outnvl)
{
int error;
dsl_pool_t *dp;
dsl_dataset_t *new, *old;
char *firstsnap;
uint64_t used, comp, uncomp;
if (nvlist_lookup_string(innvl, "firstsnap", &firstsnap) != 0)
return (SET_ERROR(EINVAL));
error = dsl_pool_hold(lastsnap, FTAG, &dp);
if (error != 0)
return (error);
error = dsl_dataset_hold(dp, lastsnap, FTAG, &new);
if (error == 0 && !new->ds_is_snapshot) {
dsl_dataset_rele(new, FTAG);
error = SET_ERROR(EINVAL);
}
if (error != 0) {
dsl_pool_rele(dp, FTAG);
return (error);
}
error = dsl_dataset_hold(dp, firstsnap, FTAG, &old);
if (error == 0 && !old->ds_is_snapshot) {
dsl_dataset_rele(old, FTAG);
error = SET_ERROR(EINVAL);
}
if (error != 0) {
dsl_dataset_rele(new, FTAG);
dsl_pool_rele(dp, FTAG);
return (error);
}
error = dsl_dataset_space_wouldfree(old, new, &used, &comp, &uncomp);
dsl_dataset_rele(old, FTAG);
dsl_dataset_rele(new, FTAG);
dsl_pool_rele(dp, FTAG);
fnvlist_add_uint64(outnvl, "used", used);
fnvlist_add_uint64(outnvl, "compressed", comp);
fnvlist_add_uint64(outnvl, "uncompressed", uncomp);
return (error);
}
static int
zfs_ioc_jail(zfs_cmd_t *zc)
{
return (zone_dataset_attach(curthread->td_ucred, zc->zc_name,
(int)zc->zc_jailid));
}
static int
zfs_ioc_unjail(zfs_cmd_t *zc)
{
return (zone_dataset_detach(curthread->td_ucred, zc->zc_name,
(int)zc->zc_jailid));
}
/*
* innvl: {
* "fd" -> file descriptor to write stream to (int32)
* (optional) "fromsnap" -> full snap name to send an incremental from
* (optional) "largeblockok" -> (value ignored)
* indicates that blocks > 128KB are permitted
* (optional) "embedok" -> (value ignored)
* presence indicates DRR_WRITE_EMBEDDED records are permitted
* (optional) "compressok" -> (value ignored)
* presence indicates compressed DRR_WRITE records are permitted
* (optional) "resume_object" and "resume_offset" -> (uint64)
* if present, resume send stream from specified object and offset.
* }
*
* outnvl is unused
*/
/* ARGSUSED */
static int
zfs_ioc_send_new(const char *snapname, nvlist_t *innvl, nvlist_t *outnvl)
{
file_t *fp;
int error;
offset_t off;
char *fromname = NULL;
int fd;
boolean_t largeblockok;
boolean_t embedok;
boolean_t compressok;
uint64_t resumeobj = 0;
uint64_t resumeoff = 0;
error = nvlist_lookup_int32(innvl, "fd", &fd);
if (error != 0)
return (SET_ERROR(EINVAL));
(void) nvlist_lookup_string(innvl, "fromsnap", &fromname);
largeblockok = nvlist_exists(innvl, "largeblockok");
embedok = nvlist_exists(innvl, "embedok");
compressok = nvlist_exists(innvl, "compressok");
(void) nvlist_lookup_uint64(innvl, "resume_object", &resumeobj);
(void) nvlist_lookup_uint64(innvl, "resume_offset", &resumeoff);
#ifdef illumos
file_t *fp = getf(fd);
#else
fget_write(curthread, fd, &cap_write_rights, &fp);
#endif
if (fp == NULL)
return (SET_ERROR(EBADF));
off = fp->f_offset;
error = dmu_send(snapname, fromname, embedok, largeblockok, compressok,
#ifdef illumos
fd, resumeobj, resumeoff, fp->f_vnode, &off);
#else
fd, resumeobj, resumeoff, fp, &off);
#endif
#ifdef illumos
if (VOP_SEEK(fp->f_vnode, fp->f_offset, &off, NULL) == 0)
fp->f_offset = off;
#else
fp->f_offset = off;
#endif
releasef(fd);
return (error);
}
/*
* Determine approximately how large a zfs send stream will be -- the number
* of bytes that will be written to the fd supplied to zfs_ioc_send_new().
*
* innvl: {
* (optional) "from" -> full snap or bookmark name to send an incremental
* from
* (optional) "largeblockok" -> (value ignored)
* indicates that blocks > 128KB are permitted
* (optional) "embedok" -> (value ignored)
* presence indicates DRR_WRITE_EMBEDDED records are permitted
* (optional) "compressok" -> (value ignored)
* presence indicates compressed DRR_WRITE records are permitted
* }
*
* outnvl: {
* "space" -> bytes of space (uint64)
* }
*/
static int
zfs_ioc_send_space(const char *snapname, nvlist_t *innvl, nvlist_t *outnvl)
{
dsl_pool_t *dp;
dsl_dataset_t *tosnap;
int error;
char *fromname;
boolean_t compressok;
uint64_t space;
error = dsl_pool_hold(snapname, FTAG, &dp);
if (error != 0)
return (error);
error = dsl_dataset_hold(dp, snapname, FTAG, &tosnap);
if (error != 0) {
dsl_pool_rele(dp, FTAG);
return (error);
}
compressok = nvlist_exists(innvl, "compressok");
error = nvlist_lookup_string(innvl, "from", &fromname);
if (error == 0) {
if (strchr(fromname, '@') != NULL) {
/*
* If from is a snapshot, hold it and use the more
* efficient dmu_send_estimate to estimate send space
* size using deadlists.
*/
dsl_dataset_t *fromsnap;
error = dsl_dataset_hold(dp, fromname, FTAG, &fromsnap);
if (error != 0)
goto out;
error = dmu_send_estimate(tosnap, fromsnap, compressok,
&space);
dsl_dataset_rele(fromsnap, FTAG);
} else if (strchr(fromname, '#') != NULL) {
/*
* If from is a bookmark, fetch the creation TXG of the
* snapshot it was created from and use that to find
* blocks that were born after it.
*/
zfs_bookmark_phys_t frombm;
error = dsl_bookmark_lookup(dp, fromname, tosnap,
&frombm);
if (error != 0)
goto out;
error = dmu_send_estimate_from_txg(tosnap,
frombm.zbm_creation_txg, compressok, &space);
} else {
/*
* from is not properly formatted as a snapshot or
* bookmark
*/
error = SET_ERROR(EINVAL);
goto out;
}
} else {
/*
* If estimating the size of a full send, use dmu_send_estimate.
*/
error = dmu_send_estimate(tosnap, NULL, compressok, &space);
}
fnvlist_add_uint64(outnvl, "space", space);
out:
dsl_dataset_rele(tosnap, FTAG);
dsl_pool_rele(dp, FTAG);
return (error);
}
/*
* Sync the currently open TXG to disk for the specified pool.
* This is somewhat similar to 'zfs_sync()'.
* For cases that do not result in error this ioctl will wait for
* the currently open TXG to commit before returning back to the caller.
*
* innvl: {
* "force" -> when true, force uberblock update even if there is no dirty data.
* In addition this will cause the vdev configuration to be written
* out including updating the zpool cache file. (boolean_t)
* }
*
* onvl is unused
*/
/* ARGSUSED */
static int
zfs_ioc_pool_sync(const char *pool, nvlist_t *innvl, nvlist_t *onvl)
{
int err;
boolean_t force;
spa_t *spa;
if ((err = spa_open(pool, &spa, FTAG)) != 0)
return (err);
force = fnvlist_lookup_boolean_value(innvl, "force");
if (force) {
spa_config_enter(spa, SCL_CONFIG, FTAG, RW_WRITER);
vdev_config_dirty(spa->spa_root_vdev);
spa_config_exit(spa, SCL_CONFIG, FTAG);
}
txg_wait_synced(spa_get_dsl(spa), 0);
spa_close(spa, FTAG);
return (err);
}
static zfs_ioc_vec_t zfs_ioc_vec[ZFS_IOC_LAST - ZFS_IOC_FIRST];
static void
zfs_ioctl_register_legacy(zfs_ioc_t ioc, zfs_ioc_legacy_func_t *func,
zfs_secpolicy_func_t *secpolicy, zfs_ioc_namecheck_t namecheck,
boolean_t log_history, zfs_ioc_poolcheck_t pool_check)
{
zfs_ioc_vec_t *vec = &zfs_ioc_vec[ioc - ZFS_IOC_FIRST];
ASSERT3U(ioc, >=, ZFS_IOC_FIRST);
ASSERT3U(ioc, <, ZFS_IOC_LAST);
ASSERT3P(vec->zvec_legacy_func, ==, NULL);
ASSERT3P(vec->zvec_func, ==, NULL);
vec->zvec_legacy_func = func;
vec->zvec_secpolicy = secpolicy;
vec->zvec_namecheck = namecheck;
vec->zvec_allow_log = log_history;
vec->zvec_pool_check = pool_check;
}
/*
* See the block comment at the beginning of this file for details on
* each argument to this function.
*/
static void
zfs_ioctl_register(const char *name, zfs_ioc_t ioc, zfs_ioc_func_t *func,
zfs_secpolicy_func_t *secpolicy, zfs_ioc_namecheck_t namecheck,
zfs_ioc_poolcheck_t pool_check, boolean_t smush_outnvlist,
boolean_t allow_log)
{
zfs_ioc_vec_t *vec = &zfs_ioc_vec[ioc - ZFS_IOC_FIRST];
ASSERT3U(ioc, >=, ZFS_IOC_FIRST);
ASSERT3U(ioc, <, ZFS_IOC_LAST);
ASSERT3P(vec->zvec_legacy_func, ==, NULL);
ASSERT3P(vec->zvec_func, ==, NULL);
/* if we are logging, the name must be valid */
ASSERT(!allow_log || namecheck != NO_NAME);
vec->zvec_name = name;
vec->zvec_func = func;
vec->zvec_secpolicy = secpolicy;
vec->zvec_namecheck = namecheck;
vec->zvec_pool_check = pool_check;
vec->zvec_smush_outnvlist = smush_outnvlist;
vec->zvec_allow_log = allow_log;
}
static void
zfs_ioctl_register_pool(zfs_ioc_t ioc, zfs_ioc_legacy_func_t *func,
zfs_secpolicy_func_t *secpolicy, boolean_t log_history,
zfs_ioc_poolcheck_t pool_check)
{
zfs_ioctl_register_legacy(ioc, func, secpolicy,
POOL_NAME, log_history, pool_check);
}
static void
zfs_ioctl_register_dataset_nolog(zfs_ioc_t ioc, zfs_ioc_legacy_func_t *func,
zfs_secpolicy_func_t *secpolicy, zfs_ioc_poolcheck_t pool_check)
{
zfs_ioctl_register_legacy(ioc, func, secpolicy,
DATASET_NAME, B_FALSE, pool_check);
}
static void
zfs_ioctl_register_pool_modify(zfs_ioc_t ioc, zfs_ioc_legacy_func_t *func)
{
zfs_ioctl_register_legacy(ioc, func, zfs_secpolicy_config,
POOL_NAME, B_TRUE, POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY);
}
static void
zfs_ioctl_register_pool_meta(zfs_ioc_t ioc, zfs_ioc_legacy_func_t *func,
zfs_secpolicy_func_t *secpolicy)
{
zfs_ioctl_register_legacy(ioc, func, secpolicy,
NO_NAME, B_FALSE, POOL_CHECK_NONE);
}
static void
zfs_ioctl_register_dataset_read_secpolicy(zfs_ioc_t ioc,
zfs_ioc_legacy_func_t *func, zfs_secpolicy_func_t *secpolicy)
{
zfs_ioctl_register_legacy(ioc, func, secpolicy,
DATASET_NAME, B_FALSE, POOL_CHECK_SUSPENDED);
}
static void
zfs_ioctl_register_dataset_read(zfs_ioc_t ioc, zfs_ioc_legacy_func_t *func)
{
zfs_ioctl_register_dataset_read_secpolicy(ioc, func,
zfs_secpolicy_read);
}
static void
zfs_ioctl_register_dataset_modify(zfs_ioc_t ioc, zfs_ioc_legacy_func_t *func,
zfs_secpolicy_func_t *secpolicy)
{
zfs_ioctl_register_legacy(ioc, func, secpolicy,
DATASET_NAME, B_TRUE, POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY);
}
static void
zfs_ioctl_init(void)
{
zfs_ioctl_register("snapshot", ZFS_IOC_SNAPSHOT,
zfs_ioc_snapshot, zfs_secpolicy_snapshot, POOL_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_TRUE, B_TRUE);
zfs_ioctl_register("log_history", ZFS_IOC_LOG_HISTORY,
zfs_ioc_log_history, zfs_secpolicy_log_history, NO_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_FALSE, B_FALSE);
zfs_ioctl_register("space_snaps", ZFS_IOC_SPACE_SNAPS,
zfs_ioc_space_snaps, zfs_secpolicy_read, DATASET_NAME,
POOL_CHECK_SUSPENDED, B_FALSE, B_FALSE);
zfs_ioctl_register("send", ZFS_IOC_SEND_NEW,
zfs_ioc_send_new, zfs_secpolicy_send_new, DATASET_NAME,
POOL_CHECK_SUSPENDED, B_FALSE, B_FALSE);
zfs_ioctl_register("send_space", ZFS_IOC_SEND_SPACE,
zfs_ioc_send_space, zfs_secpolicy_read, DATASET_NAME,
POOL_CHECK_SUSPENDED, B_FALSE, B_FALSE);
zfs_ioctl_register("create", ZFS_IOC_CREATE,
zfs_ioc_create, zfs_secpolicy_create_clone, DATASET_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_TRUE, B_TRUE);
zfs_ioctl_register("clone", ZFS_IOC_CLONE,
zfs_ioc_clone, zfs_secpolicy_create_clone, DATASET_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_TRUE, B_TRUE);
zfs_ioctl_register("remap", ZFS_IOC_REMAP,
zfs_ioc_remap, zfs_secpolicy_remap, DATASET_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_FALSE, B_TRUE);
zfs_ioctl_register("destroy_snaps", ZFS_IOC_DESTROY_SNAPS,
zfs_ioc_destroy_snaps, zfs_secpolicy_destroy_snaps, POOL_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_TRUE, B_TRUE);
zfs_ioctl_register("hold", ZFS_IOC_HOLD,
zfs_ioc_hold, zfs_secpolicy_hold, POOL_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_TRUE, B_TRUE);
zfs_ioctl_register("release", ZFS_IOC_RELEASE,
zfs_ioc_release, zfs_secpolicy_release, POOL_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_TRUE, B_TRUE);
zfs_ioctl_register("get_holds", ZFS_IOC_GET_HOLDS,
zfs_ioc_get_holds, zfs_secpolicy_read, DATASET_NAME,
POOL_CHECK_SUSPENDED, B_FALSE, B_FALSE);
zfs_ioctl_register("rollback", ZFS_IOC_ROLLBACK,
zfs_ioc_rollback, zfs_secpolicy_rollback, DATASET_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_FALSE, B_TRUE);
zfs_ioctl_register("bookmark", ZFS_IOC_BOOKMARK,
zfs_ioc_bookmark, zfs_secpolicy_bookmark, POOL_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_TRUE, B_TRUE);
zfs_ioctl_register("get_bookmarks", ZFS_IOC_GET_BOOKMARKS,
zfs_ioc_get_bookmarks, zfs_secpolicy_read, DATASET_NAME,
POOL_CHECK_SUSPENDED, B_FALSE, B_FALSE);
zfs_ioctl_register("destroy_bookmarks", ZFS_IOC_DESTROY_BOOKMARKS,
zfs_ioc_destroy_bookmarks, zfs_secpolicy_destroy_bookmarks,
POOL_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_TRUE, B_TRUE);
zfs_ioctl_register("channel_program", ZFS_IOC_CHANNEL_PROGRAM,
zfs_ioc_channel_program, zfs_secpolicy_config,
POOL_NAME, POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_TRUE,
B_TRUE);
zfs_ioctl_register("zpool_checkpoint", ZFS_IOC_POOL_CHECKPOINT,
zfs_ioc_pool_checkpoint, zfs_secpolicy_config, POOL_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_TRUE, B_TRUE);
zfs_ioctl_register("zpool_discard_checkpoint",
ZFS_IOC_POOL_DISCARD_CHECKPOINT, zfs_ioc_pool_discard_checkpoint,
zfs_secpolicy_config, POOL_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_TRUE, B_TRUE);
zfs_ioctl_register("initialize", ZFS_IOC_POOL_INITIALIZE,
zfs_ioc_pool_initialize, zfs_secpolicy_config, POOL_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_TRUE, B_TRUE);
zfs_ioctl_register("sync", ZFS_IOC_POOL_SYNC,
zfs_ioc_pool_sync, zfs_secpolicy_none, POOL_NAME,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY, B_FALSE, B_FALSE);
/* IOCTLS that use the legacy function signature */
zfs_ioctl_register_legacy(ZFS_IOC_POOL_FREEZE, zfs_ioc_pool_freeze,
zfs_secpolicy_config, NO_NAME, B_FALSE, POOL_CHECK_READONLY);
zfs_ioctl_register_pool(ZFS_IOC_POOL_CREATE, zfs_ioc_pool_create,
zfs_secpolicy_config, B_TRUE, POOL_CHECK_NONE);
zfs_ioctl_register_pool_modify(ZFS_IOC_POOL_SCAN,
zfs_ioc_pool_scan);
zfs_ioctl_register_pool_modify(ZFS_IOC_POOL_UPGRADE,
zfs_ioc_pool_upgrade);
zfs_ioctl_register_pool_modify(ZFS_IOC_VDEV_ADD,
zfs_ioc_vdev_add);
zfs_ioctl_register_pool_modify(ZFS_IOC_VDEV_REMOVE,
zfs_ioc_vdev_remove);
zfs_ioctl_register_pool_modify(ZFS_IOC_VDEV_SET_STATE,
zfs_ioc_vdev_set_state);
zfs_ioctl_register_pool_modify(ZFS_IOC_VDEV_ATTACH,
zfs_ioc_vdev_attach);
zfs_ioctl_register_pool_modify(ZFS_IOC_VDEV_DETACH,
zfs_ioc_vdev_detach);
zfs_ioctl_register_pool_modify(ZFS_IOC_VDEV_SETPATH,
zfs_ioc_vdev_setpath);
zfs_ioctl_register_pool_modify(ZFS_IOC_VDEV_SETFRU,
zfs_ioc_vdev_setfru);
zfs_ioctl_register_pool_modify(ZFS_IOC_POOL_SET_PROPS,
zfs_ioc_pool_set_props);
zfs_ioctl_register_pool_modify(ZFS_IOC_VDEV_SPLIT,
zfs_ioc_vdev_split);
zfs_ioctl_register_pool_modify(ZFS_IOC_POOL_REGUID,
zfs_ioc_pool_reguid);
zfs_ioctl_register_pool_meta(ZFS_IOC_POOL_CONFIGS,
zfs_ioc_pool_configs, zfs_secpolicy_none);
zfs_ioctl_register_pool_meta(ZFS_IOC_POOL_TRYIMPORT,
zfs_ioc_pool_tryimport, zfs_secpolicy_config);
zfs_ioctl_register_pool_meta(ZFS_IOC_INJECT_FAULT,
zfs_ioc_inject_fault, zfs_secpolicy_inject);
zfs_ioctl_register_pool_meta(ZFS_IOC_CLEAR_FAULT,
zfs_ioc_clear_fault, zfs_secpolicy_inject);
zfs_ioctl_register_pool_meta(ZFS_IOC_INJECT_LIST_NEXT,
zfs_ioc_inject_list_next, zfs_secpolicy_inject);
/*
* pool destroy, and export don't log the history as part of
* zfsdev_ioctl, but rather zfs_ioc_pool_export
* does the logging of those commands.
*/
zfs_ioctl_register_pool(ZFS_IOC_POOL_DESTROY, zfs_ioc_pool_destroy,
zfs_secpolicy_config, B_FALSE, POOL_CHECK_NONE);
zfs_ioctl_register_pool(ZFS_IOC_POOL_EXPORT, zfs_ioc_pool_export,
zfs_secpolicy_config, B_FALSE, POOL_CHECK_NONE);
zfs_ioctl_register_pool(ZFS_IOC_POOL_STATS, zfs_ioc_pool_stats,
zfs_secpolicy_read, B_FALSE, POOL_CHECK_NONE);
zfs_ioctl_register_pool(ZFS_IOC_POOL_GET_PROPS, zfs_ioc_pool_get_props,
zfs_secpolicy_read, B_FALSE, POOL_CHECK_NONE);
zfs_ioctl_register_pool(ZFS_IOC_ERROR_LOG, zfs_ioc_error_log,
zfs_secpolicy_inject, B_FALSE, POOL_CHECK_NONE);
zfs_ioctl_register_pool(ZFS_IOC_DSOBJ_TO_DSNAME,
zfs_ioc_dsobj_to_dsname,
zfs_secpolicy_diff, B_FALSE, POOL_CHECK_NONE);
zfs_ioctl_register_pool(ZFS_IOC_POOL_GET_HISTORY,
zfs_ioc_pool_get_history,
zfs_secpolicy_config, B_FALSE, POOL_CHECK_SUSPENDED);
zfs_ioctl_register_pool(ZFS_IOC_POOL_IMPORT, zfs_ioc_pool_import,
zfs_secpolicy_config, B_TRUE, POOL_CHECK_NONE);
zfs_ioctl_register_pool(ZFS_IOC_CLEAR, zfs_ioc_clear,
zfs_secpolicy_config, B_TRUE, POOL_CHECK_READONLY);
zfs_ioctl_register_pool(ZFS_IOC_POOL_REOPEN, zfs_ioc_pool_reopen,
zfs_secpolicy_config, B_TRUE, POOL_CHECK_SUSPENDED);
zfs_ioctl_register_dataset_read(ZFS_IOC_SPACE_WRITTEN,
zfs_ioc_space_written);
zfs_ioctl_register_dataset_read(ZFS_IOC_OBJSET_RECVD_PROPS,
zfs_ioc_objset_recvd_props);
zfs_ioctl_register_dataset_read(ZFS_IOC_NEXT_OBJ,
zfs_ioc_next_obj);
zfs_ioctl_register_dataset_read(ZFS_IOC_GET_FSACL,
zfs_ioc_get_fsacl);
zfs_ioctl_register_dataset_read(ZFS_IOC_OBJSET_STATS,
zfs_ioc_objset_stats);
zfs_ioctl_register_dataset_read(ZFS_IOC_OBJSET_ZPLPROPS,
zfs_ioc_objset_zplprops);
zfs_ioctl_register_dataset_read(ZFS_IOC_DATASET_LIST_NEXT,
zfs_ioc_dataset_list_next);
zfs_ioctl_register_dataset_read(ZFS_IOC_SNAPSHOT_LIST_NEXT,
zfs_ioc_snapshot_list_next);
zfs_ioctl_register_dataset_read(ZFS_IOC_SEND_PROGRESS,
zfs_ioc_send_progress);
zfs_ioctl_register_dataset_read_secpolicy(ZFS_IOC_DIFF,
zfs_ioc_diff, zfs_secpolicy_diff);
zfs_ioctl_register_dataset_read_secpolicy(ZFS_IOC_OBJ_TO_STATS,
zfs_ioc_obj_to_stats, zfs_secpolicy_diff);
zfs_ioctl_register_dataset_read_secpolicy(ZFS_IOC_OBJ_TO_PATH,
zfs_ioc_obj_to_path, zfs_secpolicy_diff);
zfs_ioctl_register_dataset_read_secpolicy(ZFS_IOC_USERSPACE_ONE,
zfs_ioc_userspace_one, zfs_secpolicy_userspace_one);
zfs_ioctl_register_dataset_read_secpolicy(ZFS_IOC_USERSPACE_MANY,
zfs_ioc_userspace_many, zfs_secpolicy_userspace_many);
zfs_ioctl_register_dataset_read_secpolicy(ZFS_IOC_SEND,
zfs_ioc_send, zfs_secpolicy_send);
zfs_ioctl_register_dataset_modify(ZFS_IOC_SET_PROP, zfs_ioc_set_prop,
zfs_secpolicy_none);
zfs_ioctl_register_dataset_modify(ZFS_IOC_DESTROY, zfs_ioc_destroy,
zfs_secpolicy_destroy);
zfs_ioctl_register_dataset_modify(ZFS_IOC_RECV, zfs_ioc_recv,
zfs_secpolicy_recv);
zfs_ioctl_register_dataset_modify(ZFS_IOC_PROMOTE, zfs_ioc_promote,
zfs_secpolicy_promote);
zfs_ioctl_register_dataset_modify(ZFS_IOC_INHERIT_PROP,
zfs_ioc_inherit_prop, zfs_secpolicy_inherit_prop);
zfs_ioctl_register_dataset_modify(ZFS_IOC_SET_FSACL, zfs_ioc_set_fsacl,
zfs_secpolicy_set_fsacl);
/*
* Not using zfs_ioctl_register_dataset_modify as DATASET_NAME check
* won't allow a bookmark name.
*/
zfs_ioctl_register_legacy(ZFS_IOC_RENAME, zfs_ioc_rename,
zfs_secpolicy_rename, ENTITY_NAME, B_TRUE,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY);
zfs_ioctl_register_dataset_nolog(ZFS_IOC_SHARE, zfs_ioc_share,
zfs_secpolicy_share, POOL_CHECK_NONE);
zfs_ioctl_register_dataset_nolog(ZFS_IOC_SMB_ACL, zfs_ioc_smb_acl,
zfs_secpolicy_smb_acl, POOL_CHECK_NONE);
zfs_ioctl_register_dataset_nolog(ZFS_IOC_USERSPACE_UPGRADE,
zfs_ioc_userspace_upgrade, zfs_secpolicy_userspace_upgrade,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY);
zfs_ioctl_register_dataset_nolog(ZFS_IOC_TMP_SNAPSHOT,
zfs_ioc_tmp_snapshot, zfs_secpolicy_tmp_snapshot,
POOL_CHECK_SUSPENDED | POOL_CHECK_READONLY);
#ifdef __FreeBSD__
zfs_ioctl_register_dataset_nolog(ZFS_IOC_JAIL, zfs_ioc_jail,
zfs_secpolicy_config, POOL_CHECK_NONE);
zfs_ioctl_register_dataset_nolog(ZFS_IOC_UNJAIL, zfs_ioc_unjail,
zfs_secpolicy_config, POOL_CHECK_NONE);
zfs_ioctl_register("fbsd_nextboot", ZFS_IOC_NEXTBOOT,
zfs_ioc_nextboot, zfs_secpolicy_config, NO_NAME,
POOL_CHECK_NONE, B_FALSE, B_FALSE);
#endif
}
int
pool_status_check(const char *name, zfs_ioc_namecheck_t type,
zfs_ioc_poolcheck_t check)
{
spa_t *spa;
int error;
ASSERT(type == POOL_NAME || type == DATASET_NAME ||
type == ENTITY_NAME);
if (check & POOL_CHECK_NONE)
return (0);
error = spa_open(name, &spa, FTAG);
if (error == 0) {
if ((check & POOL_CHECK_SUSPENDED) && spa_suspended(spa))
error = SET_ERROR(EAGAIN);
else if ((check & POOL_CHECK_READONLY) && !spa_writeable(spa))
error = SET_ERROR(EROFS);
spa_close(spa, FTAG);
}
return (error);
}
/*
* Find a free minor number.
*/
minor_t
zfsdev_minor_alloc(void)
{
static minor_t last_minor;
minor_t m;
ASSERT(MUTEX_HELD(&spa_namespace_lock));
for (m = last_minor + 1; m != last_minor; m++) {
if (m > ZFSDEV_MAX_MINOR)
m = 1;
if (ddi_get_soft_state(zfsdev_state, m) == NULL) {
last_minor = m;
return (m);
}
}
return (0);
}
static int
zfs_ctldev_init(struct cdev *devp)
{
minor_t minor;
zfs_soft_state_t *zs;
ASSERT(MUTEX_HELD(&spa_namespace_lock));
minor = zfsdev_minor_alloc();
if (minor == 0)
return (SET_ERROR(ENXIO));
if (ddi_soft_state_zalloc(zfsdev_state, minor) != DDI_SUCCESS)
return (SET_ERROR(EAGAIN));
devfs_set_cdevpriv((void *)(uintptr_t)minor, zfsdev_close);
zs = ddi_get_soft_state(zfsdev_state, minor);
zs->zss_type = ZSST_CTLDEV;
zfs_onexit_init((zfs_onexit_t **)&zs->zss_data);
return (0);
}
static void
zfs_ctldev_destroy(zfs_onexit_t *zo, minor_t minor)
{
ASSERT(MUTEX_HELD(&spa_namespace_lock));
zfs_onexit_destroy(zo);
ddi_soft_state_free(zfsdev_state, minor);
}
void *
zfsdev_get_soft_state(minor_t minor, enum zfs_soft_state_type which)
{
zfs_soft_state_t *zp;
zp = ddi_get_soft_state(zfsdev_state, minor);
if (zp == NULL || zp->zss_type != which)
return (NULL);
return (zp->zss_data);
}
static int
zfsdev_open(struct cdev *devp, int flag, int mode, struct thread *td)
{
int error = 0;
#ifdef illumos
if (getminor(*devp) != 0)
return (zvol_open(devp, flag, otyp, cr));
#endif
/* This is the control device. Allocate a new minor if requested. */
if (flag & FEXCL) {
mutex_enter(&spa_namespace_lock);
error = zfs_ctldev_init(devp);
mutex_exit(&spa_namespace_lock);
}
return (error);
}
static void
zfsdev_close(void *data)
{
zfs_onexit_t *zo;
minor_t minor = (minor_t)(uintptr_t)data;
if (minor == 0)
return;
mutex_enter(&spa_namespace_lock);
zo = zfsdev_get_soft_state(minor, ZSST_CTLDEV);
if (zo == NULL) {
mutex_exit(&spa_namespace_lock);
return;
}
zfs_ctldev_destroy(zo, minor);
mutex_exit(&spa_namespace_lock);
}
static int
zfsdev_ioctl(struct cdev *dev, u_long zcmd, caddr_t arg, int flag,
struct thread *td)
{
zfs_cmd_t *zc;
uint_t vecnum;
int error, rc, len;
#ifdef illumos
minor_t minor = getminor(dev);
#else
zfs_iocparm_t *zc_iocparm;
int cflag, cmd, oldvecnum;
boolean_t newioc, compat;
void *compat_zc = NULL;
cred_t *cr = td->td_ucred;
#endif
const zfs_ioc_vec_t *vec;
char *saved_poolname = NULL;
nvlist_t *innvl = NULL;
cflag = ZFS_CMD_COMPAT_NONE;
compat = B_FALSE;
newioc = B_TRUE; /* "new" style (zfs_iocparm_t) ioctl */
len = IOCPARM_LEN(zcmd);
vecnum = cmd = zcmd & 0xff;
/*
* Check if we are talking to supported older binaries
* and translate zfs_cmd if necessary
*/
if (len != sizeof(zfs_iocparm_t)) {
newioc = B_FALSE;
compat = B_TRUE;
vecnum = cmd;
switch (len) {
case sizeof(zfs_cmd_zcmd_t):
cflag = ZFS_CMD_COMPAT_LZC;
break;
case sizeof(zfs_cmd_deadman_t):
cflag = ZFS_CMD_COMPAT_DEADMAN;
break;
case sizeof(zfs_cmd_v28_t):
cflag = ZFS_CMD_COMPAT_V28;
break;
case sizeof(zfs_cmd_v15_t):
if (cmd >= sizeof(zfs_ioctl_v15_to_v28) /
sizeof(zfs_ioctl_v15_to_v28[0]))
return (EINVAL);
cflag = ZFS_CMD_COMPAT_V15;
vecnum = zfs_ioctl_v15_to_v28[cmd];
/*
* Return without further handling
* if the command is blacklisted.
*/
if (vecnum == ZFS_IOC_COMPAT_PASS)
return (0);
else if (vecnum == ZFS_IOC_COMPAT_FAIL)
return (ENOTSUP);
break;
default:
return (EINVAL);
}
}
#ifdef illumos
vecnum = cmd - ZFS_IOC_FIRST;
ASSERT3U(getmajor(dev), ==, ddi_driver_major(zfs_dip));
#endif
if (vecnum >= sizeof (zfs_ioc_vec) / sizeof (zfs_ioc_vec[0]))
return (SET_ERROR(EINVAL));
vec = &zfs_ioc_vec[vecnum];
zc = kmem_zalloc(sizeof(zfs_cmd_t), KM_SLEEP);
#ifdef illumos
error = ddi_copyin((void *)arg, zc, sizeof (zfs_cmd_t), flag);
if (error != 0) {
error = SET_ERROR(EFAULT);
goto out;
}
#else /* !illumos */
bzero(zc, sizeof(zfs_cmd_t));
if (newioc) {
zc_iocparm = (void *)arg;
switch (zc_iocparm->zfs_ioctl_version) {
case ZFS_IOCVER_CURRENT:
if (zc_iocparm->zfs_cmd_size != sizeof(zfs_cmd_t)) {
error = SET_ERROR(EINVAL);
goto out;
}
break;
case ZFS_IOCVER_INLANES:
if (zc_iocparm->zfs_cmd_size != sizeof(zfs_cmd_inlanes_t)) {
error = SET_ERROR(EFAULT);
goto out;
}
compat = B_TRUE;
cflag = ZFS_CMD_COMPAT_INLANES;
break;
case ZFS_IOCVER_RESUME:
if (zc_iocparm->zfs_cmd_size != sizeof(zfs_cmd_resume_t)) {
error = SET_ERROR(EFAULT);
goto out;
}
compat = B_TRUE;
cflag = ZFS_CMD_COMPAT_RESUME;
break;
case ZFS_IOCVER_EDBP:
if (zc_iocparm->zfs_cmd_size != sizeof(zfs_cmd_edbp_t)) {
error = SET_ERROR(EFAULT);
goto out;
}
compat = B_TRUE;
cflag = ZFS_CMD_COMPAT_EDBP;
break;
case ZFS_IOCVER_ZCMD:
if (zc_iocparm->zfs_cmd_size > sizeof(zfs_cmd_t) ||
zc_iocparm->zfs_cmd_size < sizeof(zfs_cmd_zcmd_t)) {
error = SET_ERROR(EFAULT);
goto out;
}
compat = B_TRUE;
cflag = ZFS_CMD_COMPAT_ZCMD;
break;
default:
error = SET_ERROR(EINVAL);
goto out;
/* NOTREACHED */
}
if (compat) {
ASSERT(sizeof(zfs_cmd_t) >= zc_iocparm->zfs_cmd_size);
compat_zc = kmem_zalloc(sizeof(zfs_cmd_t), KM_SLEEP);
bzero(compat_zc, sizeof(zfs_cmd_t));
error = ddi_copyin((void *)(uintptr_t)zc_iocparm->zfs_cmd,
compat_zc, zc_iocparm->zfs_cmd_size, flag);
if (error != 0) {
error = SET_ERROR(EFAULT);
goto out;
}
} else {
error = ddi_copyin((void *)(uintptr_t)zc_iocparm->zfs_cmd,
zc, zc_iocparm->zfs_cmd_size, flag);
if (error != 0) {
error = SET_ERROR(EFAULT);
goto out;
}
}
}
if (compat) {
if (newioc) {
ASSERT(compat_zc != NULL);
zfs_cmd_compat_get(zc, compat_zc, cflag);
} else {
ASSERT(compat_zc == NULL);
zfs_cmd_compat_get(zc, arg, cflag);
}
oldvecnum = vecnum;
error = zfs_ioctl_compat_pre(zc, &vecnum, cflag);
if (error != 0)
goto out;
if (oldvecnum != vecnum)
vec = &zfs_ioc_vec[vecnum];
}
#endif /* !illumos */
zc->zc_iflags = flag & FKIOCTL;
if (zc->zc_nvlist_src_size != 0) {
error = get_nvlist(zc->zc_nvlist_src, zc->zc_nvlist_src_size,
zc->zc_iflags, &innvl);
if (error != 0)
goto out;
}
/* rewrite innvl for backwards compatibility */
if (compat)
innvl = zfs_ioctl_compat_innvl(zc, innvl, vecnum, cflag);
/*
* Ensure that all pool/dataset names are valid before we pass down to
* the lower layers.
*/
zc->zc_name[sizeof (zc->zc_name) - 1] = '\0';
switch (vec->zvec_namecheck) {
case POOL_NAME:
if (pool_namecheck(zc->zc_name, NULL, NULL) != 0)
error = SET_ERROR(EINVAL);
else
error = pool_status_check(zc->zc_name,
vec->zvec_namecheck, vec->zvec_pool_check);
break;
case DATASET_NAME:
if (dataset_namecheck(zc->zc_name, NULL, NULL) != 0)
error = SET_ERROR(EINVAL);
else
error = pool_status_check(zc->zc_name,
vec->zvec_namecheck, vec->zvec_pool_check);
break;
case ENTITY_NAME:
if (entity_namecheck(zc->zc_name, NULL, NULL) != 0) {
error = SET_ERROR(EINVAL);
} else {
error = pool_status_check(zc->zc_name,
vec->zvec_namecheck, vec->zvec_pool_check);
}
break;
case NO_NAME:
break;
}
if (error == 0)
error = vec->zvec_secpolicy(zc, innvl, cr);
if (error != 0)
goto out;
/* legacy ioctls can modify zc_name */
len = strcspn(zc->zc_name, "/@#") + 1;
saved_poolname = kmem_alloc(len, KM_SLEEP);
(void) strlcpy(saved_poolname, zc->zc_name, len);
if (vec->zvec_func != NULL) {
nvlist_t *outnvl;
int puterror = 0;
spa_t *spa;
nvlist_t *lognv = NULL;
ASSERT(vec->zvec_legacy_func == NULL);
/*
* Add the innvl to the lognv before calling the func,
* in case the func changes the innvl.
*/
if (vec->zvec_allow_log) {
lognv = fnvlist_alloc();
fnvlist_add_string(lognv, ZPOOL_HIST_IOCTL,
vec->zvec_name);
if (!nvlist_empty(innvl)) {
fnvlist_add_nvlist(lognv, ZPOOL_HIST_INPUT_NVL,
innvl);
}
}
outnvl = fnvlist_alloc();
error = vec->zvec_func(zc->zc_name, innvl, outnvl);
/*
* Some commands can partially execute, modify state, and still
* return an error. In these cases, attempt to record what
* was modified.
*/
if ((error == 0 ||
(cmd == ZFS_IOC_CHANNEL_PROGRAM && error != EINVAL)) &&
vec->zvec_allow_log &&
spa_open(zc->zc_name, &spa, FTAG) == 0) {
if (!nvlist_empty(outnvl)) {
fnvlist_add_nvlist(lognv, ZPOOL_HIST_OUTPUT_NVL,
outnvl);
}
if (error != 0) {
fnvlist_add_int64(lognv, ZPOOL_HIST_ERRNO,
error);
}
(void) spa_history_log_nvl(spa, lognv);
spa_close(spa, FTAG);
}
fnvlist_free(lognv);
/* rewrite outnvl for backwards compatibility */
if (compat)
outnvl = zfs_ioctl_compat_outnvl(zc, outnvl, vecnum,
cflag);
if (!nvlist_empty(outnvl) || zc->zc_nvlist_dst_size != 0) {
int smusherror = 0;
if (vec->zvec_smush_outnvlist) {
smusherror = nvlist_smush(outnvl,
zc->zc_nvlist_dst_size);
}
if (smusherror == 0)
puterror = put_nvlist(zc, outnvl);
}
if (puterror != 0)
error = puterror;
nvlist_free(outnvl);
} else {
error = vec->zvec_legacy_func(zc);
}
out:
nvlist_free(innvl);
#ifdef illumos
rc = ddi_copyout(zc, (void *)arg, sizeof (zfs_cmd_t), flag);
if (error == 0 && rc != 0)
error = SET_ERROR(EFAULT);
#else
if (compat) {
zfs_ioctl_compat_post(zc, cmd, cflag);
if (newioc) {
ASSERT(compat_zc != NULL);
ASSERT(sizeof(zfs_cmd_t) >= zc_iocparm->zfs_cmd_size);
zfs_cmd_compat_put(zc, compat_zc, vecnum, cflag);
rc = ddi_copyout(compat_zc,
(void *)(uintptr_t)zc_iocparm->zfs_cmd,
zc_iocparm->zfs_cmd_size, flag);
if (error == 0 && rc != 0)
error = SET_ERROR(EFAULT);
kmem_free(compat_zc, sizeof (zfs_cmd_t));
} else {
zfs_cmd_compat_put(zc, arg, vecnum, cflag);
}
} else {
ASSERT(newioc);
rc = ddi_copyout(zc, (void *)(uintptr_t)zc_iocparm->zfs_cmd,
sizeof (zfs_cmd_t), flag);
if (error == 0 && rc != 0)
error = SET_ERROR(EFAULT);
}
#endif
if (error == 0 && vec->zvec_allow_log) {
char *s = tsd_get(zfs_allow_log_key);
if (s != NULL)
strfree(s);
(void) tsd_set(zfs_allow_log_key, saved_poolname);
} else {
if (saved_poolname != NULL)
strfree(saved_poolname);
}
kmem_free(zc, sizeof (zfs_cmd_t));
return (error);
}
#ifdef illumos
static int
zfs_attach(dev_info_t *dip, ddi_attach_cmd_t cmd)
{
if (cmd != DDI_ATTACH)
return (DDI_FAILURE);
if (ddi_create_minor_node(dip, "zfs", S_IFCHR, 0,
DDI_PSEUDO, 0) == DDI_FAILURE)
return (DDI_FAILURE);
zfs_dip = dip;
ddi_report_dev(dip);
return (DDI_SUCCESS);
}
static int
zfs_detach(dev_info_t *dip, ddi_detach_cmd_t cmd)
{
if (spa_busy() || zfs_busy() || zvol_busy())
return (DDI_FAILURE);
if (cmd != DDI_DETACH)
return (DDI_FAILURE);
zfs_dip = NULL;
ddi_prop_remove_all(dip);
ddi_remove_minor_node(dip, NULL);
return (DDI_SUCCESS);
}
/*ARGSUSED*/
static int
zfs_info(dev_info_t *dip, ddi_info_cmd_t infocmd, void *arg, void **result)
{
switch (infocmd) {
case DDI_INFO_DEVT2DEVINFO:
*result = zfs_dip;
return (DDI_SUCCESS);
case DDI_INFO_DEVT2INSTANCE:
*result = (void *)0;
return (DDI_SUCCESS);
}
return (DDI_FAILURE);
}
#endif /* illumos */
/*
* OK, so this is a little weird.
*
* /dev/zfs is the control node, i.e. minor 0.
* /dev/zvol/[r]dsk/pool/dataset are the zvols, minor > 0.
*
* /dev/zfs has basically nothing to do except serve up ioctls,
* so most of the standard driver entry points are in zvol.c.
*/
#ifdef illumos
static struct cb_ops zfs_cb_ops = {
zfsdev_open, /* open */
zfsdev_close, /* close */
zvol_strategy, /* strategy */
nodev, /* print */
zvol_dump, /* dump */
zvol_read, /* read */
zvol_write, /* write */
zfsdev_ioctl, /* ioctl */
nodev, /* devmap */
nodev, /* mmap */
nodev, /* segmap */
nochpoll, /* poll */
ddi_prop_op, /* prop_op */
NULL, /* streamtab */
D_NEW | D_MP | D_64BIT, /* Driver compatibility flag */
CB_REV, /* version */
nodev, /* async read */
nodev, /* async write */
};
static struct dev_ops zfs_dev_ops = {
DEVO_REV, /* version */
0, /* refcnt */
zfs_info, /* info */
nulldev, /* identify */
nulldev, /* probe */
zfs_attach, /* attach */
zfs_detach, /* detach */
nodev, /* reset */
&zfs_cb_ops, /* driver operations */
NULL, /* no bus operations */
NULL, /* power */
ddi_quiesce_not_needed, /* quiesce */
};
static struct modldrv zfs_modldrv = {
&mod_driverops,
"ZFS storage pool",
&zfs_dev_ops
};
static struct modlinkage modlinkage = {
MODREV_1,
(void *)&zfs_modlfs,
(void *)&zfs_modldrv,
NULL
};
#endif /* illumos */
static struct cdevsw zfs_cdevsw = {
.d_version = D_VERSION,
.d_open = zfsdev_open,
.d_ioctl = zfsdev_ioctl,
.d_name = ZFS_DEV_NAME
};
static void
zfs_allow_log_destroy(void *arg)
{
char *poolname = arg;
strfree(poolname);
}
static void
zfsdev_init(void)
{
zfsdev = make_dev(&zfs_cdevsw, 0x0, UID_ROOT, GID_OPERATOR, 0666,
ZFS_DEV_NAME);
}
static void
zfsdev_fini(void)
{
if (zfsdev != NULL)
destroy_dev(zfsdev);
}
static struct root_hold_token *zfs_root_token;
#ifdef illumos
int
_init(void)
{
int error;
spa_init(FREAD | FWRITE);
zfs_init();
zvol_init();
zfs_ioctl_init();
if ((error = mod_install(&modlinkage)) != 0) {
zvol_fini();
zfs_fini();
spa_fini();
return (error);
}
tsd_create(&zfs_fsyncer_key, NULL);
tsd_create(&rrw_tsd_key, rrw_tsd_destroy);
tsd_create(&zfs_allow_log_key, zfs_allow_log_destroy);
error = ldi_ident_from_mod(&modlinkage, &zfs_li);
ASSERT(error == 0);
mutex_init(&zfs_share_lock, NULL, MUTEX_DEFAULT, NULL);
return (0);
}
int
_fini(void)
{
int error;
if (spa_busy() || zfs_busy() || zvol_busy() || zio_injection_enabled)
return (SET_ERROR(EBUSY));
if ((error = mod_remove(&modlinkage)) != 0)
return (error);
zvol_fini();
zfs_fini();
spa_fini();
if (zfs_nfsshare_inited)
(void) ddi_modclose(nfs_mod);
if (zfs_smbshare_inited)
(void) ddi_modclose(smbsrv_mod);
if (zfs_nfsshare_inited || zfs_smbshare_inited)
(void) ddi_modclose(sharefs_mod);
tsd_destroy(&zfs_fsyncer_key);
ldi_ident_release(zfs_li);
zfs_li = NULL;
mutex_destroy(&zfs_share_lock);
return (error);
}
int
_info(struct modinfo *modinfop)
{
return (mod_info(&modlinkage, modinfop));
}
#endif /* illumos */
static int zfs__init(void);
static int zfs__fini(void);
static void zfs_shutdown(void *, int);
static eventhandler_tag zfs_shutdown_event_tag;
#ifdef __FreeBSD__
#define ZFS_MIN_KSTACK_PAGES 4
#endif
int
zfs__init(void)
{
#ifdef __FreeBSD__
#if KSTACK_PAGES < ZFS_MIN_KSTACK_PAGES
printf("ZFS NOTICE: KSTACK_PAGES is %d which could result in stack "
"overflow panic!\nPlease consider adding "
"'options KSTACK_PAGES=%d' to your kernel config\n", KSTACK_PAGES,
ZFS_MIN_KSTACK_PAGES);
#endif
#endif
zfs_root_token = root_mount_hold("ZFS");
mutex_init(&zfs_share_lock, NULL, MUTEX_DEFAULT, NULL);
spa_init(FREAD | FWRITE);
zfs_init();
zvol_init();
zfs_ioctl_init();
tsd_create(&zfs_fsyncer_key, NULL);
tsd_create(&rrw_tsd_key, rrw_tsd_destroy);
tsd_create(&zfs_allow_log_key, zfs_allow_log_destroy);
tsd_create(&zfs_geom_probe_vdev_key, NULL);
printf("ZFS storage pool version: features support (" SPA_VERSION_STRING ")\n");
root_mount_rel(zfs_root_token);
zfsdev_init();
return (0);
}
int
zfs__fini(void)
{
if (spa_busy() || zfs_busy() || zvol_busy() ||
zio_injection_enabled) {
return (EBUSY);
}
zfsdev_fini();
zvol_fini();
zfs_fini();
spa_fini();
tsd_destroy(&zfs_fsyncer_key);
tsd_destroy(&rrw_tsd_key);
tsd_destroy(&zfs_allow_log_key);
mutex_destroy(&zfs_share_lock);
return (0);
}
static void
zfs_shutdown(void *arg __unused, int howto __unused)
{
/*
* ZFS fini routines can not properly work in a panic-ed system.
*/
if (!KERNEL_PANICKED())
(void)zfs__fini();
}
static int
zfs_modevent(module_t mod, int type, void *unused __unused)
{
int err;
switch (type) {
case MOD_LOAD:
err = zfs__init();
if (err == 0)
zfs_shutdown_event_tag = EVENTHANDLER_REGISTER(
shutdown_post_sync, zfs_shutdown, NULL,
SHUTDOWN_PRI_FIRST);
return (err);
case MOD_UNLOAD:
err = zfs__fini();
if (err == 0 && zfs_shutdown_event_tag != NULL)
EVENTHANDLER_DEREGISTER(shutdown_post_sync,
zfs_shutdown_event_tag);
return (err);
case MOD_SHUTDOWN:
return (0);
default:
break;
}
return (EOPNOTSUPP);
}
static moduledata_t zfs_mod = {
"zfsctrl",
zfs_modevent,
0
};
DECLARE_MODULE(zfsctrl, zfs_mod, SI_SUB_VFS, SI_ORDER_ANY);
MODULE_VERSION(zfsctrl, 1);
MODULE_DEPEND(zfsctrl, opensolaris, 1, 1, 1);
MODULE_DEPEND(zfsctrl, krpc, 1, 1, 1);
MODULE_DEPEND(zfsctrl, acl_nfs4, 1, 1, 1);
MODULE_DEPEND(zfsctrl, zlib, 1, 1, 1);
Index: head/sys/cddl/contrib/opensolaris/uts/common/sys/fs/zfs.h
===================================================================
--- head/sys/cddl/contrib/opensolaris/uts/common/sys/fs/zfs.h (revision 359302)
+++ head/sys/cddl/contrib/opensolaris/uts/common/sys/fs/zfs.h (revision 359303)
@@ -1,1229 +1,1230 @@
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2011, 2016 by Delphix. All rights reserved.
* Copyright 2011 Nexenta Systems, Inc. All rights reserved.
* Copyright (c) 2012, Martin Matuska <mm@FreeBSD.org>. All rights reserved.
* Copyright (c) 2014 Integros [integros.com]
* Copyright 2017 Joyent, Inc.
* Copyright (c) 2019 Datto Inc.
* Copyright (c) 2017, Intel Corporation.
*/
/* Portions Copyright 2010 Robert Milkowski */
#ifndef _SYS_FS_ZFS_H
#define _SYS_FS_ZFS_H
#include <sys/types.h>
#include <sys/ioccom.h>
#include <sys/time.h>
#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_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_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 usr/src/common/zfs/zfs_prop.c.
*/
typedef enum {
ZPROP_CONT = -2,
ZPROP_INVAL = -1,
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, /* not exposed to the user */
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_REDUNDANT_METADATA,
ZFS_PROP_PREV_SNAP,
ZFS_PROP_RECEIVE_RESUME_TOKEN,
ZFS_PROP_REMAPTXG, /* not exposed to the user */
ZFS_PROP_SPECIAL_SMALL_BLOCKS,
ZFS_NUM_PROPS
} zfs_prop_t;
typedef enum {
ZFS_PROP_USERUSED,
ZFS_PROP_USERQUOTA,
ZFS_PROP_GROUPUSED,
ZFS_PROP_GROUPQUOTA,
ZFS_NUM_USERQUOTA_PROPS
} zfs_userquota_prop_t;
extern const char *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. 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 {
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_COMMENT,
ZPOOL_PROP_EXPANDSZ,
ZPOOL_PROP_FREEING,
ZPOOL_PROP_FRAGMENTATION,
ZPOOL_PROP_LEAKED,
ZPOOL_PROP_MAXBLOCKSIZE,
ZPOOL_PROP_BOOTSIZE,
ZPOOL_PROP_CHECKPOINT,
ZPOOL_PROP_TNAME,
ZPOOL_PROP_MAXDNODESIZE,
ZPOOL_PROP_MULTIHOST,
ZPOOL_NUM_PROPS
} zpool_prop_t;
/* Small enough to not hog a whole line of printout in zpool(1M). */
#define ZPROP_MAX_COMMENT 32
#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
/*
* Dataset property functions shared between libzfs and kernel.
*/
const char *zfs_prop_default_string(zfs_prop_t);
uint64_t zfs_prop_default_numeric(zfs_prop_t);
boolean_t zfs_prop_readonly(zfs_prop_t);
boolean_t zfs_prop_visible(zfs_prop_t prop);
boolean_t zfs_prop_inheritable(zfs_prop_t);
boolean_t zfs_prop_setonce(zfs_prop_t);
const char *zfs_prop_to_name(zfs_prop_t);
zfs_prop_t zfs_name_to_prop(const char *);
boolean_t zfs_prop_user(const char *);
boolean_t zfs_prop_userquota(const char *);
int zfs_prop_index_to_string(zfs_prop_t, uint64_t, const char **);
int zfs_prop_string_to_index(zfs_prop_t, const char *, uint64_t *);
uint64_t zfs_prop_random_value(zfs_prop_t, uint64_t seed);
boolean_t zfs_prop_valid_for_type(int, zfs_type_t);
/*
* Pool property functions shared between libzfs and kernel.
*/
zpool_prop_t zpool_name_to_prop(const char *);
const char *zpool_prop_to_name(zpool_prop_t);
const char *zpool_prop_default_string(zpool_prop_t);
uint64_t zpool_prop_default_numeric(zpool_prop_t);
boolean_t zpool_prop_readonly(zpool_prop_t);
boolean_t zpool_prop_feature(const char *);
boolean_t zpool_prop_unsupported(const char *name);
int zpool_prop_index_to_string(zpool_prop_t, uint64_t, const char **);
int zpool_prop_string_to_index(zpool_prop_t, const char *, uint64_t *);
uint64_t zpool_prop_random_value(zpool_prop_t, 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_VOLMODE_DEFAULT = 0,
ZFS_VOLMODE_GEOM = 1,
ZFS_VOLMODE_DEV = 2,
ZFS_VOLMODE_NONE = 3
} zfs_volmode_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_type_t;
/*
* 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
/*
* When bumping up SPA_VERSION, make sure GRUB ZFS understands the on-disk
* format change. Go to usr/src/grub/grub-0.97/stage2/{zfs-include/, fsys_zfs*},
* and do the appropriate changes. Also bump the version number in
* usr/src/grub/capability.
*/
#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.
*
* When changing, be sure to teach GRUB how to read the new format!
* See usr/src/grub/grub-0.97/stage2/{zfs-include/,fsys_zfs*}
*/
#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
/* 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 "<reverse-DNS>:"
* (e.g. "org.open-zfs:") 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_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 */
#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_REMOVING "removing"
#define ZPOOL_CONFIG_RESILVER_TXG "resilver_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_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_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 */
/*
* 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_DATA_ERRORS "verify_data_errors"
#define ZPOOL_CONFIG_REWIND_TIME "seconds_of_rewind"
#define VDEV_TYPE_ROOT "root"
#define VDEV_TYPE_MIRROR "mirror"
#define VDEV_TYPE_REPLACING "replacing"
#define VDEV_TYPE_RAIDZ "raidz"
#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"
/* 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_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"
#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"
/*
* This is needed in userland to report the minimum necessary device size.
*
* Note that the zfs test suite uses 64MB vdevs.
*/
#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/zfs/zpool.cache"
/*
* 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 */
VDEV_AUX_SPLIT_POOL, /* vdev was split off into another pool */
VDEV_AUX_ASHIFT_TOO_BIG, /* vdev's min block size is too large */
VDEV_AUX_CHILDREN_OFFLINE, /* all children are offline */
VDEV_AUX_ACTIVE /* vdev active on a different host */
} 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_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;
/*
* Initialize functions.
*/
typedef enum pool_initialize_func {
POOL_INITIALIZE_DO,
POOL_INITIALIZE_CANCEL,
POOL_INITIALIZE_SUSPEND,
POOL_INITIALIZE_FUNCS
} pool_initialize_func_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_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_to_process; /* total bytes to process */
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 scurb pass */
/* cumulative time scrub spent paused, needed for rate calculation */
uint64_t pss_pass_scrub_spent_paused;
/* Sorted scrubbing new fields */
/* Stored on disk */
uint64_t pss_issued; /* total bytes checked by scanner */
/* Not stored on disk */
uint64_t pss_pass_issued; /* issued bytes per scan pass */
} 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_NUM_STATES
} dsl_scan_state_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;
typedef enum {
VDEV_INITIALIZE_NONE,
VDEV_INITIALIZE_ACTIVE,
VDEV_INITIALIZE_CANCELED,
VDEV_INITIALIZE_SUSPENDED,
VDEV_INITIALIZE_COMPLETE
} vdev_initializing_state_t;
/*
* Vdev statistics. Note: all fields should be 64-bit because this
* is passed between kernel and userland as an nvlist uint64 array.
*/
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[ZIO_TYPES]; /* operation count */
uint64_t vs_bytes[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_self_healed; /* self-healed bytes */
uint64_t vs_scan_removing; /* removing? */
uint64_t vs_scan_processed; /* scan processed bytes */
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_fragmentation; /* device fragmentation */
uint64_t vs_checkpoint_space; /* checkpoint-consumed space */
uint64_t vs_initialize_errors; /* initializing errors */
uint64_t vs_initialize_bytes_done; /* bytes initialized */
uint64_t vs_initialize_bytes_est; /* total bytes to initialize */
uint64_t vs_initialize_state; /* vdev_initialzing_state_t */
uint64_t vs_initialize_action_time; /* time_t */
} 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)))
/*
* 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 elments 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_NAME "zfs"
#define ZFS_DEV "/dev/" ZFS_DEV_NAME
#define ZFS_DISK_ROOT "/dev/dsk"
#define ZFS_DISK_ROOTD ZFS_DISK_ROOT "/"
#define ZFS_RDISK_ROOT "/dev/rdsk"
#define ZFS_RDISK_ROOTD ZFS_RDISK_ROOT "/"
/* general zvol path */
#define ZVOL_DIR "/dev/zvol"
/* expansion */
#define ZVOL_PSEUDO_DEV "/devices/pseudo/zfs@0:"
/* for dump and swap */
#define ZVOL_FULL_DEV_DIR ZVOL_DIR "/dsk/"
#define ZVOL_FULL_RDEV_DIR ZVOL_DIR "/rdsk/"
#define ZVOL_PROP_NAME "name"
#define ZVOL_DEFAULT_BLOCKSIZE 8192
/*
* /dev/zfs ioctl numbers.
*/
typedef enum zfs_ioc {
ZFS_IOC_FIRST = 0,
ZFS_IOC_POOL_CREATE = ZFS_IOC_FIRST,
ZFS_IOC_POOL_DESTROY,
ZFS_IOC_POOL_IMPORT,
ZFS_IOC_POOL_EXPORT,
ZFS_IOC_POOL_CONFIGS,
ZFS_IOC_POOL_STATS,
ZFS_IOC_POOL_TRYIMPORT,
ZFS_IOC_POOL_SCAN,
ZFS_IOC_POOL_FREEZE,
ZFS_IOC_POOL_UPGRADE,
ZFS_IOC_POOL_GET_HISTORY,
ZFS_IOC_VDEV_ADD,
ZFS_IOC_VDEV_REMOVE,
ZFS_IOC_VDEV_SET_STATE,
ZFS_IOC_VDEV_ATTACH,
ZFS_IOC_VDEV_DETACH,
ZFS_IOC_VDEV_SETPATH,
ZFS_IOC_VDEV_SETFRU,
ZFS_IOC_OBJSET_STATS,
ZFS_IOC_OBJSET_ZPLPROPS,
ZFS_IOC_DATASET_LIST_NEXT,
ZFS_IOC_SNAPSHOT_LIST_NEXT,
ZFS_IOC_SET_PROP,
ZFS_IOC_CREATE,
ZFS_IOC_DESTROY,
ZFS_IOC_ROLLBACK,
ZFS_IOC_RENAME,
ZFS_IOC_RECV,
ZFS_IOC_SEND,
ZFS_IOC_INJECT_FAULT,
ZFS_IOC_CLEAR_FAULT,
ZFS_IOC_INJECT_LIST_NEXT,
ZFS_IOC_ERROR_LOG,
ZFS_IOC_CLEAR,
ZFS_IOC_PROMOTE,
ZFS_IOC_DESTROY_SNAPS,
ZFS_IOC_SNAPSHOT,
ZFS_IOC_DSOBJ_TO_DSNAME,
ZFS_IOC_OBJ_TO_PATH,
ZFS_IOC_POOL_SET_PROPS,
ZFS_IOC_POOL_GET_PROPS,
ZFS_IOC_SET_FSACL,
ZFS_IOC_GET_FSACL,
ZFS_IOC_SHARE,
ZFS_IOC_INHERIT_PROP,
ZFS_IOC_SMB_ACL,
ZFS_IOC_USERSPACE_ONE,
ZFS_IOC_USERSPACE_MANY,
ZFS_IOC_USERSPACE_UPGRADE,
ZFS_IOC_HOLD,
ZFS_IOC_RELEASE,
ZFS_IOC_GET_HOLDS,
ZFS_IOC_OBJSET_RECVD_PROPS,
ZFS_IOC_VDEV_SPLIT,
ZFS_IOC_NEXT_OBJ,
ZFS_IOC_DIFF,
ZFS_IOC_TMP_SNAPSHOT,
ZFS_IOC_OBJ_TO_STATS,
ZFS_IOC_JAIL,
ZFS_IOC_UNJAIL,
ZFS_IOC_POOL_REGUID,
ZFS_IOC_SPACE_WRITTEN,
ZFS_IOC_SPACE_SNAPS,
ZFS_IOC_SEND_PROGRESS,
ZFS_IOC_POOL_REOPEN,
ZFS_IOC_LOG_HISTORY,
ZFS_IOC_SEND_NEW,
ZFS_IOC_SEND_SPACE,
ZFS_IOC_CLONE,
ZFS_IOC_BOOKMARK,
ZFS_IOC_GET_BOOKMARKS,
ZFS_IOC_DESTROY_BOOKMARKS,
#ifdef __FreeBSD__
ZFS_IOC_NEXTBOOT,
#endif
ZFS_IOC_CHANNEL_PROGRAM,
ZFS_IOC_REMAP,
ZFS_IOC_POOL_CHECKPOINT,
ZFS_IOC_POOL_DISCARD_CHECKPOINT,
ZFS_IOC_POOL_INITIALIZE,
ZFS_IOC_POOL_SYNC,
ZFS_IOC_LAST
} zfs_ioc_t;
/*
* 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.
*/
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_VDEV_TOO_BIG,
+ ZFS_ERR_WRONG_PARENT,
} 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;
/*
* 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_DSNAME "dsname"
#define ZPOOL_HIST_DSID "dsid"
#define ZPOOL_HIST_ERRNO "errno"
/*
* The following are names used when invoking ZFS_IOC_POOL_INITIALIZE.
*/
#define ZPOOL_INITIALIZE_COMMAND "initialize_command"
#define ZPOOL_INITIALIZE_VDEVS "initialize_vdevs"
/*
* 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_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_CHECKPOINT 0x10
#define ZFS_IMPORT_TEMP_NAME 0x20
#define ZFS_IMPORT_SKIP_MMP 0x40
/*
* 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)
/*
* 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"
/*
* Sysevent payload members. ZFS will generate the following sysevents with the
* given payloads:
*
* ESC_ZFS_RESILVER_START
* ESC_ZFS_RESILVER_END
* 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"
#ifdef __cplusplus
}
#endif
#endif /* _SYS_FS_ZFS_H */